diff --git "a/data/test-jira-sentences.txt" "b/data/test-jira-sentences.txt" new file mode 100644--- /dev/null +++ "b/data/test-jira-sentences.txt" @@ -0,0 +1,434899 @@ +Fixed +The test cases depend on conversion of data into other data types which is in the dummy operator + will tell Olga all the necessary information and Olga will add it to the guide +The problem now is that it fails when I pull out the network cable which means that it still relies on some sort of network access even if I use offline +released with +Fixed on allow on the does not get set as passed in as an optional parameter to the do you mind adding some documentation notes about this new allow option? +Do you want to open a YARN JIRA for the changes +Thanks for looking at the patch Rick! I expanded the abbreviationCNF to case normal form in the comments you indicatedHi RickRick Rick Hillegas commented onRick Rick Rick Hi DagRick Rick Thanks for the patch + This bug has been marked as a duplicate of +Unfortunately since were not based in India we cant easily get this phone to test on +import readline breaks on I recall we werent going to make an effort to support windows here yet just making sure we didnt paint ourselves into a corner for when we doBut yeah a missing readline should probably be caught and handled regardless +Verified on RHEL +Vaibav Is this patch ready for review? If so please attach the patch to this ticket and click Submit Patch +constructors removed and replaced with readable static builder methodsSending srcjavaorgapachecommonslangSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +There are releases using this in production +Thanks issue Oozie is released +RC release +So any fix would need to made consistently also in +The docs probably need to be updated +added a safe patch to avoid all the problems +Did you happen to run through the code to ensure that throwing exceptions is the correct behaviour +Just looked at the runs output again this is related to this patchSo Teds right either commit an addendum or revert the patch please +The notion of separate user and group mechanisms was meant to describe the current situation and obviously it makes no sense to have a group file delivering the groups when authentication is done via active directory sayI believe there is a need for this when external authentication mechanisms are used for precisely the reason above it is a possible security issueThe external group file mechanism is meant to work in combination with the current external password file decoupling groups fromHope that explains things betterAndrew +I was try to move from to however because of the new dependency I was reluctant. +Thanks changes applied. +svn revision Thanks Joydeep. +Can you post the result of full test run +this is a blocker for some stuff we are doing in production +Could have incorporated this into some other patch but okay looks good checking it in +No test case or additional details provided. +I already enumerated a number of properties we have and would like to have in +Marking as resolved +Update version affect for both and Sure it can be simulated on while you are investigating this issue I am going to unset the Fix Version for this jira +I have opened a new JIRA to implement the EBNF grammar that will correctly determine how to identify invalid location strings. +If you have the code checked out you could step it through we are planning to release soon so if you could supply a patch I will make sure it goes in quickThx reimplemented combinedCategory queries +Also please look at previos discussions already discussed about the same issue +The Java files remain unchanged only patch has changed +Ill comment further when I know more +Machine details provided assigning back to with with latest automation run +The patch I attached should take care of as well +Max Ive looked into deploy dir and is there +I just committed this. +This is basically just test code itself and I dont think its worth supplying unit tests for it +equals upfront +Also there are quite a few test cases doesnt do appendNosync between and so they are needed to be adjusted +Mind posting a separate patch Andrew +supposing the messaging pattern doesnt change between superstep and superstep this looks like a great improvement great work +Assuming that I am remembering things correctly +Now the core of the problem may not be in it may be in the way the Application server interacts with and uses the API but this is a serious critical bug that should not be ignored +ve just committed this +Both memory usage andreuse of images which have changed on disk are a big problemCurrently I am disabling the cache altogether even commented out the variableto be sure +I would never authorize a pull request to Wildfly on a version + +Eric looks like you might have put the keyname in the passphrase fieldNo Stephens latest wiki edit explains it You cant have a passphrase set and use the agent +Given Matthews comment I think I should have provided more motivation for the issue first +Initial and final patch. +Thanks Owen! +The sub tasks are pending but the core implementation is done ad the samples are working hence closingThanks to Nabeel for his initial contrubutions to make this possible +Thanks. +Production server config would be a good candidate minimum would be server startup and running a sample app like the would be sufficient +Thanks Closed. +What exact versions of java are you using? I tried invoking the script above a few hundred times and didnt observe the behavior you describe +Great work! Thanks again for putting so much time and thought into this +Any comments or suggestions are welcomedThe cancelation task is drivenmaintaned in class instanceA brief explanation of changes in classesComments to xarollbackcommitprepare added stating that these functions shouldnot be called directly but commitprepare shouldbe called instead because these do also timer task maintanancecancallationcancel added a method to cancel the current running statementcalling commitprepare insteadof commitprepare +Please also note that my comment about own transaction management from EJB also include delegation to JTA subsystem +That is used by both wsprovide and wsconsume scripts to define the tools classloader +I agree with you will post the design document soon +ThomasI am reassigning this to you are you are going to be looking into memory leakissues +Attachment has been added with description reset for each have committed this patch +You are not the first who found that one out and wont be the last +in CVS as far as I can tell +great in +Hi Sushanth Thanks for your comments about the unit test case +I really like what I am seeing so far +Documented as a known issue in the EAP release notes +Seems like the same number of files might get created anyway when you split the L file? Or do you plan on doing some logical striping across the L boundary as Nicolas says above where the L files are never truly splitLike Stack mentions longer term I think well need to split memstore while in use and those splits should probably have some alignment with these stripe boundaries +It should have errors for validate that configurations +Ray is finishing up needs updating +ALR Actually should this be applied to as well +Added patch for force property +How JBM examples are packaged and delivered with EAP? Is it a separate distributable? +I usedto use SSH to remember my password when using CVS +Uploading a patch that uses checks the length to verify a being added and removed from it + +A comment on the viability of this request would help a lot +Seems like missed this one +Simply adding and jars to my war lib causes thisThanks it works when I put all of the jars in my +I kind of disagree first I do not mess with any API +Ill poke around and see if i can figure something out +Thanks +I just committed this +Contains simpletype for replacing xsnormalizedString types +docx sample file was not the same as that from the +The section about adding a logj appender looks already complete +Okay +I have not set up the PL SP to interoperate against testShib +CommittedPlease verify +fix is a minor behavior change that may require user app changes which are documented in release note +I guess the best thing to do is to put it to a hour vote on the list +Minor comment does not seem appropriate in this scenario +What is happening when you try to paginate? Is there an error? Or what happens +patch update small refactoring excuse for blanks in only addTables removeTables were added +Please use this latest patch +Could just be Fedora weirdness +No +Cannot reproduce your screenshot +straight forward Ivan for the patch and looks pretty nice +Ive just committed thisI opened for the larger issuebq +Patch for unused results +Thanks AlejandroYou are faster than a rocketAvner + Fixes the NP +Instead enter your anwser as a comment to the corresponding ticket +Question on thisWhat is happening when you try to updatecreate an object? Notcompatible? Is this a issue or an actual Axisissue? Has anyone tried to work with the API directly usingAxis +Thats a good point Sameer +I keep this issue opened for the advanced parameters suggested by Freddy +Thanks for looking into this Alfonso +Screenshot from the reporting tweet +But is it going to be tied to checkpointing +completes this fix. +tmit looks great +for committing the patch +Submitting patch +Example extracted from the email report and checked to act as described +Lets touch base on the next week to see whats Bruno I agree +I say commit Werners patch +It will contain few main classes like Delete SQL Qualifier +Canceling the patch to fix test failures +Thanks Ted for taking a look +Could you think this over? +Note the latest greatest is +The howto is updatedNOTEDo not use directories with +This should avoid an getting generated due to the messages getting spooled the the Memory Store without went into trunk in rev +Can you post some sample code that shows how you handle the reconnect? doesnt itself do anything to try and reconnect so Id like to see how you are dealing with this so I can get an idea of how to reproduce the issue +But since its not possible to say whether the table or the index was meant if theres an index that has the same name as a table I guess thats not an option +This is current DFS behavior +Commit it tomorrow if no objectionsThanks for the review Ted Ram free to commit Chunhui +How do I reopen the issue +Hi Rishi Ill close out this issue for a wrap up of the project and raise a new issue for the Velocity docs +This is issue is void +Thank you for a quick fix Stepan. +Whats the meaning of not capable of filtering returned values and how this can be switched off? +tribute to document van a tribute to rondjes puzzelen tussen rijen en gilze van tot to simon +fixed in +Ill take care for these next +XML will now show in the result view for both deployed VDB testing and preview +It should be fine to build keys before processing any other elements +Should be removedYes it is removed in this patch +It is a very common idiom in C to declare the copy constructor and the assignment operator as private and to provide no definition for them if copy and assigment semantics make no sense for a class +how do I reproduce this issue +And this wont enable IDE to hook into receiving test STDOUT and STDERR +xml +Thanks Hairong! +Also is based on inventory unitCosts +Moving back to for now +It is unfortunate that suffers from this particular issue +In particular the second focuses on what the property name is while the first is a general introduction on the concept of getter +Hi MiceWe recreate the system VMs only when there is a change in system templateIs there any change with the upgrade instructions nowThanks with non oss build +Deleting the old files caused new ones to be created correctly without the versioning info however the deploy plugin really should update the metadata correctly without having to delete +Fixed by + prevents us applying the change without forcing the users to specify the sign implementation in the config +Please let me know your thoughts +Thanks for digging into this one JarekIf we swallow the exception in the shim and return it would be misleading to the userInstead what if we catch the exception during the call to compute in and log a message about how the counters could not be fetched for this job +So The should have been triggered before that shouldnt it +We need to look into it. +The unix stat uses b for blocks and s for size in bytesI agree that adding s for size in bytes is a good idea but changing the semantics of b is an incompatible change which seems risky +I have the same problem and cant just improve the input +This issue will be finally closed but first Ill do some reviews before +Does this happen after the session has been invalidated or at other times +But to this patch at least I can acquire locks to all tables on the database +Not sure if it is easy to add a unit test +All of them successfully pass on RI +Should not be a problem +Mark this as closed if that was sufficient. +Avoiding placing more than one data node of the same logical cluster on the same physical host will increase reliability even if the new topology algorithm is in place +Bulk closing stale resolved issues +Thanks +I try this now for days to find a suitable way to make it workI have on my search page results but the pager thinks he has to draw steps because the numFound result says +Will add a patch +This looks good to me +Greatthanks! And I appreciate your taking the time for the dialog. +Im not sure why this is put in the solrj source treeIn the x the zookeeper can be removed +Whoops comment race sorry back to Eli +hopefully it will back soon +This avoid reinitialization during the web app startup which may include a conflicting commons logging version INSTALL INSTRUCTIONS Copy the attached jars to the JBOSSHOMEserverdefaultdeploy and JBOSSHOMEserveralldeploy and COMPATIBILITY BY Scott Wed Jun PDT t this be going into CP +to record porting to to branch at subversion revision +You can work around this by turning off deleting the table on the test and using the shell +done in a test in revision to detect potential backward incompatible changes to this behavior + +Thanks +seems its not a hbase issue but related with hdfs project please close this jira and create a new hdfs issue if youd likeps please give more detailed info +English and Finnish updated translation updates are needed for other languages +Good making it required would be Id prefer simple telling how to avoid the conflict and if possible relaying the exception with an indication about the possible cause +From the point of view optimizing IOPS in HDFS is very important +Im not against exposing something package private for tests anyone that jumps the fence to use that should know what they are getting themselves into +The problem appears to be a simple matter of the name for the Jetty Web Container being incorrect inIt currently is specified as public static final String WEBCONTAINEROBJECTNAME nullorgapachegeronimoServergeronimojeeTypenamewhen it should be public static final String WEBCONTAINEROBJECTNAME nullorgapachegeronimoServergeronimojeeTypenameThis is causing an attempted lookup of a invalid +The work around is implemented in +Thanks Mridul for the contribution committed in trunk at r in at rAshish Vijaywargiya +Verified on EAP test present in with soapUI modified encoding property +I see +And honestly I want this in but thats a different story +Ive NOT made any modification to the code in SVN trunk and Ive tried running cargo with both jbossx and jbossx on port +updated patch with those changes +I would go with b + +When I think about it I think it could be because this test case is one of few where the underlying tabledata is changed between each execution of the +I will commit this to trunk in rev +Resolved by +There are some misprints in Please use new patch to fix it +on the review +Thanks DagCommitted revision Also backported to and committed revision . +Assigning to Todd for completion +Each distributed index is seven total shards on two servers +Minor code change +Lucene must be ableto repair a corrupt index +Not sure how it does on tomcat +Unfortunately extendeded is too imaturebuggy to be usable at this time it would be nice to see consolidated stable and consistent support for xxx +mrepositoryorgapachecommonsUserstucu +This does not indicate anything is wrong with the test or the patch per se +the file came back again +Excellent news thanks Werner! Credit goes to Gax for originally finding the problem and posting the patch +Attributes have been added to the tld but not to the Tag class +Thank you +This should work noThe other is a feature request to be able to disable CRC creation in get +Its now implemented as a group operationThis patch contains the implementation +I will try and find a way around needing a new release but probably this wont happen +Patch for to handle missing parameter value +If the heap has nothing in it then matcher should be set to the end? Or start? There are no Readers +Committed revision +Yes agreed +Hi Suran I have started going through the test but will be out tomorrow so may not finish my review until Monday +Since s testing is suspect at the moment ran unit tests locally and all passed +He said he would comment on this JIRA today +yes this is expected as this is not implemented + +I had a problem applying the patch where ended up to be an empty file +They are just the files downloaded from the CSP the zip file of EAP +Thanks Evanthika +How are others working around this +This has been done +Best even to try the latest from couldnt reproduce the problem with the given process and test +Ill leave this issue open for reporting back on that +It works fine for me now +is the outstanding issue for the linux CI build +We have a lot of other toollibraries from which we can copy +This last change affects only the configuration factory +Do we need extra output in or was it a debug changeReverted some code clean up and check the others +d for generator d for the remaining tasks +Committed revision . +As Mohammad mentioned it will not affect release right now +Thanks Elliott +As far as I can see no changes in the installer are needed +close schedulerattach a patch following this sequence. +Changing affected version to and fixed version to the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Thejas Is the patch ready for review +the example uses which we dont support in +Blindly guessing maybe related to +As of revision generic method invocations can infer parameters on anonymous closures. +Its not in there anywhere +Its a very simple fix +Applied in bd +Changed fix version to sub tasks completed +however that relatively obscure bug in the database constraints took a days work to solveFor me and for several hundred developers in the forums I have seen the error reporting IS an issuePlease I admire what the hibernate team are doing but please listen +And in this case I really cannot do anything with it in the maven plugin and you will need to adress this issue to the aspectJ mailing lists. +Ill try to clean it up +load order attributes order item attributes the list name to order the previous patch had order + +One story would be to add this as an option to the +Btw that was my first guess so I guess this is intuitive enough +Without them I would have to close this issue as Can not reproduce +Ill post minutes here so that people who can not attend can see and participate +Committed revision +Scheduled for +If it needs to be changed it should be a different issue +I deployed the sender and listener on EAP in RHEL but was not able to reproduce the error +I believe that for now its safe to assume that all scripts executed with this command are +GuoSo do I wait for your work on this before commiting +Sorry about running the release here +No time for it in the the fix to +New patch added +Fernando please review and close +It wasnt until yesterday that I moved the mirror back to and it seems to go connection or speed crazy with that host +please close this bug +Look for all of the +What it didnt pick up though was the new tab +as I was going home after submitting the last comment I realized I was mistakenI still would like to understand the motivation behind creating a separate rather than having the class provide the shorthand methodss response that the exception factory is in preparation for the JDBC work seems reasonable to me +Is there any way I can help with this change? We are blocked until it is fixed +Are there other show stoppers in this designimplementationI wont know till I look closer but I dont doubt we will get this in +behavior is exactly the same +contains only the fix for URN authority not found when searching for all authorities +Patch applied with slight modifications +Thank you! +Cancelling the patch broke the unit tests needs more testingGrant could you please assign this to me +Revised patch for fs resolver +Method cannot be made synchronized because it can result in a deadlock +Regressions ran ok committed as svn +Im closing this for now as it is not reproducible +This is now fixed +Is this a global map or per connection +Bring the existing patch against latest trunk +It uses a new persistence capable package order with two classes Order and +This problem effectively prohibits Forge to discover plugins in anything but the topmost in turn effectively prohibits anyone from using anything but a single project for defining Forge plugins +Though the spec could still be more specific here +Implemented in rev +I wonder if there is anything special about your setup +Ciro thanks +Create new issues one leave it as is +Please let me know if somebody else wants to review this patch +Thanks Ashwin! I committed this to trunk and. +I have ran into same issueIt is common way of developing starting projects and samples storing classes in default package +Why this take so long to do? May we have a response about when we can have this done or not +This attachement contains test data used to verify the fix for issue worked as expected +Thanks Andy for reviewing it. +Because of speed we do not do this +Can you your documents as individual files? That makes it a lot easier to read you dont have to download and zip you can just go to it straight from JIRA +Please do share a patch with your changes +Sid please ensure that this patch is committed to too +Thanks +What we probably need is something like nested cast operators for complex types +So youll be writing your own version so nothing to be done by me on this task +fix is even more simple than the closing stale resolved issues +Rebuilt the patch on top of the latest from svn +Hi Andrew New tests committed in revision +There will be two bound to the same transaction one for the logical handle one for the physical handle which has the positive effect of reusing the logical handle for the entire transaction not recreating it for every getConnection you seen an actual negative effect here? Or were you just wondering about the two mentioned in the debug logve revised this for RC explicitly checking for a in now +I tried to reproduce the bug with any known situation where it occurred +Will commit if tests pass +The schema says the soapBody should have a single child called which your message does notStuck a better error message in there. +Testing this patch now +fixed in cvs +I backed that out in revision +Verified in Red Hat Docs for JBDS +Why not try and reproduce it with a test? Theres already a large number of them in to work from +reopening for possible backport +checked in trunkOleg +Michael could you tell me if will this ticket take you long to finishThank you so much for your development +Yea I just tried changing my environment to filedevurandom and it doesnt work correctly I see Jetty hanging at startup +implementation is done +Thanks +I checked that all the places in query parser that require this change are covered +btw please note that in different interesting +Patch by Alexander Peters +for the review Jimmy +Patch applied +WDYT +Thanks for making the change +Test failures and are unrelated to this patch +which version are you using please? Also you mention you have numerous independant fixes in your local working copy +Christian honestly not sure in either case this did what I wanted for my app and I thought it would be useful for other folks +I ran the attached example and was able to reproduce it and indeed appears to endlessly grow +could you please also attach the new and sources those are the only two which arent useable after applying the had various problems applying patches provoked by the Id +has been marked as a duplicate of this bug +For me most cases seem to be more related to the ability to exclude certain paths based on a given depth +I have a slight preference for Network over IO and Protocol over Amqp +For user implementations of Chunker I expect them to create their own Chunks as well that do Tuple creation in next +This is finally making its way into Spring now! See my comment on +Line import Should be removed +Interested parties please review and improve +If the client fumbles the authentication fails +In addition I added additional unit tests for minus values to the updated patch +can confirm this is an MVEL bug submitted failing test to mvel author +Can you add a patch for documentation too +Would you please try this patchBest regardsSpark patch applied at revision r with minor changes thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +call for such classes hibernate annotation order to implement that Iave patch grails source +Hadoop QA picks the newest attachmentYou can attach trunk patch one more time +further improved ops per sec to fixed a bug +If we made descriptors to be agnostic to this then yes descriptors becomes a DSL for any type of format +I am labeling this ticket as newbie with a caveat +Identify features for flexibility that are never used and remove them leaving a single implementation which is easier to understand +patch looks good +its just like the sitedeploy plugin really +Sateeshs comments indicate this is not complete implementation for +yes this is for LATER not +The attached patch disables on the branch when running Java or higherCommitted to the branch with revision . +newcomers wont know what to put there without doing some research and we should default the version box to the latest stable release +The colors for the ink are white and brownCheyenne +Yes I added a +zk why cannot master read zk nodes and apply rules too? ZK nodes may need to be made smarter for this purpose +Thanks! +Missing is how to start up the bugger +All patches have been submitted +We already do this +Thanks for the feedback John +Thanks Ravi +But we need to implement validation too +So my suggestion would be rather useless. +Sorry forgot to do svn add before +bq +There is also the related performance concern of caching the resolvedvalidate language object form of the respective expressions which is hard to generalize for an m more inclined to go with the latter approach and work out any details of modifications to the policy later +There is selection and it works +Second the second level batch writer will have to rebin the mutations even though the first level batch writer has already done this +Ive found the problem +Triaged for +Sorry for my mistake on previous commit cbeddcdbbacbed +If this needs to be reconsidered please raise the issue with the SOA PM team. +Closing issues. +However I see no problem in its being acceptedand when it is this bug can close +ah yes probably the quickfix I rememberHaving it at code completion makes perfect sense too though +We will need to add new tests to the testsuite as we add dependencies to in the Embedded will add a comment for each mbean that we add to this Jira +This is an API docs issue +I ran and CDH integration tests successfully +in should the r be the same height as the s or is it intentionally shorter +Which works fine if you serialize only one column but generates incorrect results in the general case +Also your reference to use the onclick attribute is an insufficient solution as it refers to a script expression and not to an think both node and node would be useful +If is found in classpath it is loaded emitting a warning +Attached updated version of the patch which is now compiled and tested with Java +commontests +With this patch I think this issue can be closed +Haha looks like our comments raced with each otherThanks for filing the JIRA to address the comments +Im backporting this to and branchesMichi +Im sending a patch to rubyspec to add some specs +Emmanuel please confirm that the attached diff is the complete fix for +svn should just be is apparently a regression since was already marked as fixed +You should definitely check this with legal +This is on trunk only +Simply the fact that there is an enum value normal for and the default is not pt suggests that youre probably right and normal may indeed have a different meaning +He will take a look +If any other problem exists with soapr they should have separate bugs reported +Hi Tian HongAs a general rule their rates of change are different cpu changes faster than memory and memory changes faster than disk +Now parse the URI using String +Happy to contribute to such a fantastic project +The customer will need to add this package to the install instructions if they want to use +Pradeep will take a look at fixing this +Changing these files in the sar would require introducing a different build artifact for AS testing as opposed to your functional testing +The URL patterns are changed due to the admin console navigation tree reorganization +for the patch +better lookup procedures Are there many classes with such a big number of members to benefit from hashing vs linear search? Do you have some profiling data pointing to this bottleneck or it is a wild guess +I believe this issue was fixed in commit r +These issues appear to be all related to the simple scheduler implementation +If you finish it before we release beta please correct the version +Can any guys help me out or provide me code as how to integrate the jquery function in the page +The issue is not about changing the structure of the managed resource +Len please resolve this +Replaced with instead +I suppose we could modify those methods to make copies and only allow them to be initialized once +I found the problem! The parameters being passed into the method callare being filled into the params attribute of the +new patch that refs instead +In every case it is accompanied by a message like DEBUG Internal protocol error message delivery ackednacked multiple times where the actual count displayed could be different depending on how many previous successful tests were run +So Im not sure if them ending up at the same cache slot is actually a bad thing Yes same here +The is +Ill close this ticket given it duplicates +That is a bug in which incorrectly erases column ids previously set +It works because properties are injected by Java into the Rails yeah I should have thought of it +In the query doesnt compile +This patch also fixes tests that depended on the incorrect parameter name +I am still not convinced it should be enabled per default +In this case template size is zero so volume and snapshot size also was set to zeroTemplate size was reported in and is closed currently +Thanks for your support. +WDYT? +I was using open ssh to test this +Solved ages ago +This looks stable and eliminates the worst of the bouncing around scrollbar behaviourIt also fixes the dynamic sizing bug originally reported here. +So I think creation of this file can be moved back to the compile phaseHowever the new structure is such that the hbase project shows up last in the reactor list +Looks like work is going to keep me busy this week +Also I think it makes sense to check in the commit part of this patch while working through the rollback issues in another patch but it would be good to do so with the regression test for the commit portion +Daniel could you track these tests on hudson +Scott Our Public API is an extension of the Spec +So please take it in your list of soon implemented features Do you have any tips and tricks where I can get more information about Grunt and cordovaThanks! +Actually this is s bug +There is a potential +Is this ticket stil valid +bq +Thanks for helping to keep the docs honest! +Patch looks good +For the metadata file you need to setup automatic cant you just upload it to the repo with mvn which should generate the metadata +Another solution would be to have a separate entity for just the party ID and the invoice number +is not a valid reply status as far as I can thats been reported in albeit my fix used incorrect reply codes +My guess would be hierarchical Brett can you confirm? directory structure should be hierarchical +We are using several map implementations in both JavaC and each implementation is a bit different +It is still for earlier versions of Hadoop and not for commit here +Thanks +And the patch file +Hi Wing and JacobI want to request that Will it be possible for scomp util to identify such huge coded value source file to split into Kb JVM standard before doing the compiliationMay be this woud be additional feature for Please adviceThanks Wing Yew Thanks for the advice +Feel free to mention this on the mailing list +Patch committed to the trunk +Fixed failures in moduleslunisrcmainnativeluniwindows at r +Eclipse configuration +I just committed this +Then and it keeps running like that until I have rows the same rows that never get deleted which are for the only artifact I have added after the migration to Archiva Couldnt be something about the +Thanks +It requires some changes but Ill get there before +Heres a proposed fix to downgrade the timeout exception from to +Therefore has to create an message context and inject the response once it comes +To ensure a client is not requesting an excluded version +Doing so will create a huge performance degredation for the operation for tables with lots of rows +Jong this was not your Issue please open a new one rather than alter someone elses +files to reproduce to since is released +Lahiru I had a sample code which I tried out to do this +I have tested the fix and I can confirm that it is working in my environment +Attaching patch +In instead of duplicating code in shallowIterator could we rename deepIterator to internalIterator and add a flag to control whether we want to do shallow iteration or deep iteration? In general we dont want to expose the shallow iterator externally +Or am I missing something +Use the +I have merged this to. +Thanks Paul looks good to me +At the moment we only save policy at the service level and this can cause problems in the case where we try to invoke different operations of the same service +could you rebase against trunk +bq +A version for converters as well might be still needed +Fixed in Maven Publisher +committed to head +push to +classes from srcaction and srcmodel end up in one jar anyway as far as I know +Bulk moved to +That solution would probably involve some pretty big changes though and Im not sure how feasible it would be to actually implement +Thanks for your work. +If im doing a stress test against a cluster and restart a node the entire cluster still gets blocked +how about +Might end up just closing all of these as Out of date +agreed and thanks for the feedback fortunately there were a reasonable explanation behind it +This is causing us big problems since we lose the replicated session state when we run into this ll validate and hopefully take care of this before the next major release +The right patch is +I propose to commit this +pdf instead of showing it in a popup windowCan you try with ? It works fine for me +I committed this. +Attempted to reproduce this on the table that was having issues originally but scans seemed to work fine +So first of all it should work with the fully supported maven repo that is supplied as a zip no +Thank you +Eric if you arent done here you should reopen. +According to QA reports the caused rev +So the O tag seems to be important +Previous rev of the release notes had an unbalanced paragraph tag +did you try to execute without compiling the groovy parts? if it works in that case then it confirms that you used the wrong compiler +This was done because it looked like the compiler was not thread safe +If we can use dcontrols mailbox we dont need it +As far as I know debug mode works fine if you run ofbiz in debug mode from antshell and then connect to it from your IDE +I say lets start with fixing in and depending on how the fix looks like we can expand to other places +The above handling for the case of region min should be bounded by the number of regions in the table minus +The internet accessing is caused by xml validating fix was applied at r please verify if it works for you +Thanks Boris +It sounds you are right +Also use instead of just using throw because it is more readableThanks +theres a snapshot dependency that cant be the version having the snapshot there can cause updated snapshot version for removed +So when I created the woden branch they werent checked in +I tried it +I have no philosophical objections to adding these annotations +See the attached this patch in place the richfaces and seam booking demo will start working again +the problem is more general encoding support for encoding support is fixed for and every xdoc file fml and so onto be precise BOM is supported for and but not + for this feature +Ive removed some of the info from section and moved it into the attached +The weakness of the spooling is that you can lose the spooled data if you lose the client filesystem but it has the advantage of lower resource requirements overall I think +Using the latest I now see the same output as with the patch applied to. +This jira will be resolved once Richard merges to trunk +We will fix up this soon in So above resolution is anyway correct to proceed +We can then leverage that for our constraints +is required as well I posted a patch for that + +And in html you cant just put anywhere you want +Linked new issues +Modified to and +This time it immediately fails with broker full even though the consumer has emptied the broker +and +I fixed both exception and refresh problems +Fixed and ported to +Using Scout and j the Registry Guide in CP CR +There are ways to restrict where the script comes fromAlso all there is a seperate for each document +Some of them were actually not used and those fields and methods were removedSome methods had to stay public because they were interface methods +Fixed by adding nulalblefalse on unique diff for branch vGACP +It is not intrusive for +Patch appliedThanks! +Hack away +The produces one XML Schema file per namespace +I can update once a decision has been made +I would be okay marking this not valuable enough to justify the potental +Updated the patch a little mostly for the patch reverted the caching default using the scan attribute to specify prefetching instead of a global flag +When we have multipart payloads we do expect with the type if we relax it then we can capture by mistake the data meant to be processed byIf you disagree then let me know please what do you think can be improved +I think we want to added similar handling in the +I guess Daniel meant to say that Zebra isnt in active development which is true +deleted the wrong ones and moved to new group will show up in some can you reopen this issue? The is referring the artifacts at the old location and it needs some relocation for version and to fix it +When I apply this and revert the patch for unit tests still pass +I dont have a tool to open the rar file for this issue +Since youre planning to fix the messages in a patch Im to both a and bAs to the messages perhaps we could reduce the number of messages needed by using the same message for encryption and decryption +That commit did not add it it simply moved it from a implicit transitive dependency to a dependency that was listed in the main pom +Ive detailed a possible use in the Cordova Plugin Upgrade Guide +Replaced attachment +Another batch of private finals +Its been a month or so just checking in +The move has been done see revision We need to add a build script +It works in this way in JSFStrutsSeam +Committed to in r. +The attached source with applied changes are from release now made in the Xerces versions of the programsAlso added the grammarpool property inFix may be stiil incomplete because Michael Glavassevich noted that it should cover all properties +Can you please subscribe using and drop us a note so we can discuss it? Its very low traffic +Looking at the patch it is not clear to me how this change would resolve the different behaviour experienced when a naming suffix is specified to when it isnm concerned by the fact that we are now experiencing different behaviour Ill continue to experiment to see if I can shed any further light on the me attach the test case I am using later on so that we have a common base to discuss +The underlying crypto engine is throwing an exception for the lack of an IV with the cipher specified +It is now possible to invoke the ee tests from the top level +How much memory is the process actually using +There is a comment in the test saying create and write a file that contains three blocks of data +All of my executions have unique id elements and this behavior still occurs +Per the spec the mixetag mixin types jcretag property is not mandatory but Section states that an implementation must create a jcretag property with an implementation determined value +That has all been the basis for what all needs to come next +Please try this patch +Duplicate to. +This is a great discussion +Should I somehow modify the file +In the meantime would you mind to remove all the formatting changes? And also move back to? I guess it would reduce several thousands lines from the patch +Encountering a block with an invalid check value is a corruption +But if we press this tool bar button dialog for global settings appears not for the current page +Heres a patch which clears the instance variable on task attempt failure +If were going to do all these whitespace changes its going to make changes to both and trunk difficult +However this is only a minor issue because it does not corrupt anything +Mentioned to Javier that our current code drop does not currently build installers +What happens if there is no hostname +patch which adds the origin of an artifact to the default +Fix looks good +Ryan refers here to the documentation chapter +Its not set thats the issue i think +If it does not then I think its best for the NM to terminate the resources utilization it is responsible for and leave the node in the state it had been upon startup +Close after release +This bug means you cant use Groovy mocks for checking the arguments passed to the mocked method +pending jenkins +You might also want to update the warning text replacing you classpath with your classpath +Thanks Alessandro! + reports the problem is fixed +This code was adapted from the data table +Attachment has been added with description Page breaker work is a collaboration between myself and Peter MatthiasThanks for pointing out the flaw in the current logic +Sorry about this +The algorithm used by is also used by Spring framework for instance +the order works as expected. +Was fixed in the series. +Wont fix +Need to see the startup log + +modulessrctestjavaorgapacheairavataxbayainterpreterApply the patch to trunkmodules +Patches committedThanks! +Thanks Andrew! +Fixed in head revision +Hi Richard Patch committed in revision +Just in case any preferences whether to use or +One remark though are also printed to the file when running with the service wrapper +Assigning to Laura from the docs team +Committed to trunk and x branches +The earlier patches for the new tests excluded +bq +Let me know if it causes you any problems +The next question would be why this empty nodes are hanging around? Lets shut them all down +I decided to directly remove copy of classes as they had been added for internal use by the hashbelt caches only +Very clear writeup Army thank you very much +Can you provide a clean patch +I recommend you to upgrade and have look +i fixed a bug with the completion processor which i introduced +Thanks Andrzej +if the download fails for some reason log messages but dont fail the built in case developer is in a disconnected state +Closing since this looks resolved for now +simple files open easier in browser +Hadrian the tests in TC looks have green light to build the release +Tried all those +just a quick questions i notice that operations on sender in initiateConnection are not synchronized +to make sure publish request doesnt return wrong message seq id +please click on the line below it in the error log it has the stacktrace +I committed the patch to trunk in revision since I dont think we have any other nicer solution right now +In this scenario the would map only one of those pages to the new index and in the next call the not mapped page will cause the page expired exception + lgtm +I think we have plans for a facility that fixes the underlying issue here without the proposed way of solving it +That means your is not installed so as to include the Lists Permissions or Webs servicesHave a look at the URL you are using as the base URL and append vtibin using a browser +Im beginner and if its all right I add other If youre planning on building out the entire module or at least the majority of it I would suggest assembling everything together into a single large contribution +Multiple Password annotations in a bean are allowed now +made this redundant. +After some interesting discussion in chat proliferating here is not the way to go +Shaun I have removed this unused settings from Seam Settings page at all +I will clarify this on the next version of the document +defer all issues to +Looks like JET compare divider with bit zero +Already done +Hi PaulexThe patch looks fine +imo there is no big sense to enhance something here +Francesco Your patch looks pretty good +Thanks for the overview it was helpful +Could this be too late? I mean if you dont lock the node then it is possible that interceptor can detect a fqn node but before it can lock it the fqn is removed by another thread +The difficulties I ran into were due to maintaining +Please elaborate on htat issue +attach a new patch to address Ivan and Flavio comments +All of the delegates that fail to respond to during initialization would be specially tagged +Attaching two files one for common project and one for mapreduce for review +It looks good! Now Im inclined to commit this now for +Sorry for the mistake +Thanks Hitesh +Regenerated following project split +Dup +Sorry I was so late. +Attaching a first attempt at a release note +I will start looking at this bug +Committed. +bulk close of all resolved issues. +Reviewed +There is an interaction of reordering for triples in a BGP or quad block and filter placement +I just looked at the protocol documentation and I dont think currently the btree scan will do what you want with respect to OPENMODELOCKNOWAIT +Eli it sounds good +I just have created a new issue to deal with authentication service and user management service +With these definition the date string generated by the javascript calendar is always a valid date that can be parsed by javas dateparser +But still unclearMay be i will check the defect to see if it helps +Fixed in in the iPOJO runtime +Please test and get back to with which works so for noticed that Enums are not handled by so my guess is that has to be modified as well +Not really +Im not sure what you mean about adding the file to the stack +Committed revision . +Ill take a closer look when I get home but minus the config stuff I think this is gonna be pretty close to good +Fixed in ESB codebase will be in next tag for platform +Id leave it for a few days so that some of our people who are using it can try it out and report any issues +I committed the patch to trunk and +Just the mention of EC not the mention the test environment setup and cost issues put me off +Committed +I guess we should be safe and just use the old stream copying for nowJust out of curiosity thoughbq +Scott Im looking into the case where the client could get a CNFE and Ill see if it needs some additional handling in the EJBCLIENT project +You can achieve this by Describe your problem in terms of what you observe and what you expect to observe Provide a sample application with one bean and one facelet reproducing this error +Hiconsider that in my application I created two jars one with annotations and another one with xml and the two jars contain separate persistence unitsHope this helps. +batch transition resolvedfixed bugs to closedfixed +This allows us to parameterize the corresponding permission grantM javadrdaorgapachederbydrdaM javadrdaorgapachederbydrdaAdditional permission grants +Thanks for fixing this. +Devarajs test results indicate this is NOT a backwards compatible patch to the extent described in the release note +Closing this record since is now released. +It is very important that for small queues are long enough +I wont be working on this any time soon +srcmainjavaorgapachehadoopyarnapirecords +So I started to think this is an overengineering +I mentioned this issue to my manager who reports that the problem exists currently on IE too +This test belongs to the kind its not a mandatory threshold defined from earlier precise settings so updating it when the code changes is often fair +GeertYour issue is different +I suppose we could use a JMS consumer to do the download and have a static hashmap with the key as some kind of file ID and the value as the progress percent +TRUNK doesnt have the prob +and where did you get your launch command from? Does not look like anything we ever used +edson is this fixed +Fixed in r +Updated the wiki page for it +It is still race condition if the operation on the right of the assignment operator can return different values for different callsIn the example above native heap may be exausted at the time of race condition in this assignment and one of calls to new will return NULL white the other will return normal object +When looked at the generated stub class while parsing the server response it is getting errorWe are using Axis +Ive created a quickstart and upgraded and to and it works +Tests passed for me yay +See Section Message Selector Syntax for more informationand Identifiers An identifier is an character sequence that must begin with a Java identifier start character all following characters must be Java identifier part characters +no partial blocks in the middleI assume that this means the block in the middle should be of size block length +Thanks for the comments well try the new maven version +Proposed fix for this issue +Verified in +The same black backgropund is seen on WTPs palettethumbnail +So as in the first case where the character does not appear in the target charset then any behavior could be acceptable character replacement or perhaps something else +Patch to fix the two test failures in +the test was not correctly registering the extension used to check the impl +Attaching a patch for this issue +Would be good to have this fixed before we start consider cutting a new release of any of the Camel branches +Thanks GaganAaron +Thank youLukasz! it works now +I suggested this and can do this for you if you like +I think I prefer the approach for two reasons it relies on configuration to guarantee the hostname is reported as expected rather than how we may be able to decipher and also its a smaller change with fewer moving parts +committed inNot sure about Im a fan of the simplifications we get from letting CFRR only need to deal with splits +Bob was going to add these yesterday +The plugin will contain the code that is not user interface related +Create Folder Name foobar fails if there is no folder foo Same as point implemented in accordance with the explanation provided in my earlier comment on this issue +Thanks Chris +I do not want to reduce robustness in places where we document null handling and throw IAE with informative messages where not allowed nulls are provided +This issue has been fixed and released as part of release +added initial review comments +I dont have anything to addit sounds like youve nailed down the issue and have a clear statement of the problem +I have a patch for Original Message +Ive attached a screenshot of what a typical project looks like +defer all X issues to +This pacth is also targeted for the release so it needs to be on the branch too +Unless you are suggesting that we remove the file and move the contents into as a static inner class +Hi Mubarak and thanks for the patch! This is definitely a feature we want in production environments so its greatly appreciated +If no class is specified all tests are run and the test fails +I think that this can go in if we get one more pair of eyes on my latest patchIt may need p to apply +But iterating on the znodes and deleting them is ok +Left m version in place I dont want to write an m build for thisrev can you please try this +Weve not been able to reproduce this issue +Attaching marked up header content of the rat report will fix up the NO EXCEPTION marked lines and repost when reduced to only EXCEPTION lines +commit the fix +I was never sure if this concept was interesting to the gang +lets agree to be confused you mean +Attached is a patch that fixes the issue by changing some settings in the configuration in the root sitemapThe patch also turns on locale negotiation in the +I dont think the jline version has changed which could be the only thing I can think of affecting ve never seen this problem on the +Though it is not possible to use DAS on this caseI also dont see how we could improve DAS to support this kind of situationRegardsAdriano on the last comment from wangful marking this as resolved +This will make testing setups waaaay easierThanks for tackling this +yes implementing orthogonality check in does affect the verification job in general +You could put the executions in a release profile so we still get value doing a release build via +Im excited for your feedback +domeModel +So either make depends REQUIRED inthe DTD or figure out how to allow it to be skipped +Yes I am sure +Yes that would do it +Any easier ways to do thisRunning tests and doing some manual inspection +The path value is specified by a developer configuring Shiro not an application end user if the url the dev specifies is invalid startup will fail fast showing what went wrong no need to sanitize developer configuration input in this case. +The approach Ive taken should support something like of the formats possible with Issue is that the datepicker currently supports only the notion of hour time +Heres a patch which should address all of your feedbackbq +Ill handle the documentation for the component +Updated the Deprecated methods usage with successorsPlease review +New wizard issue cant be reproduced in JBDS and in the on last build from trunk the same exception assigning back to my to verify it on real build from hudson not local I think you referred us to a different issue however this influences the entire New you Denis for the progress update the issue in case you need more information from me but I see you reproduced the issue completely and know where the problem is +Not applicable to standard jspwiki +Submitting a patch for a JDK patch +Anyone know where were at with this issue +does not work + +committed +Take a look at src Ive written utility methods to parse addresses. +Wowo! Greate job +heres my trivial changes +jconsole screenshot that shows perf counters in nanoseconds +Looks good to me +Committed to trunk branch and branch of branch +Im attaching version of the patch +If they end up wanting it we can move it but I think its fine on for now + +Closed due to the release of Tiles. +Then Ill create link from my side after firewall will be set correctly +OK I made the few changes David wanted locally and committed. +BumpIf we could get this committed it will help with the entity documentation effort +JDBC is correct on the branch because JDBC is not fully implementedFor trunk the JRE JDBC line was removed with the fix for +If you ever get around to fixing this try to lift the requirement for not supporting denoting a different namespace than its parents +Committed revision +Usageant rpm to create rpm packageant deb to create deb package + st patch +We think the problem has to do with a typo on this line var rawData It should corrected to var rawData +puppetmodulestemplates at line uses taskTracker port +Today I have used a Hex Viewer and there i saw that every character was converted to Unicode +I didnt find the time to implement the fix on Friday and feel like its not a good idea to quickly push it into the release and potentially sacrifice quality +What do you think Scott +Committed to trunk and +patch looks good +So is it OK to leave the fix as it is? If so we may close the issue +expression could be as complex as you want it to be +Jon did you ever get to try this with the final release +Another would be when upgrading a service orCurrently services are started with deploy so we need a matching undeploy that would be run from the same machine that would cause the service to stop processing input requests and shutdown when the current request has been completed +I think this came down to not closing +It does however capture stdout during tests so if your logging is configured to log to stdout then it will be captured and will be in the test report +I think it is reasonable to call this not a bugThanks for your help and I am sorry about the confusion. +Ok I can see the exception +Data that was supposed to be gone could reappear which feels semantically different to meI tend to agree +Not sure why wed want to delay changing the status code until +This test in special has not to do much with the vendor +Here is a example of a output from clusterdump of JSON +It was fixed in +Oh yeah sorry in fact i have a value binding instead of Some Value +Refactored start and row functions out to reduce code duplication +Committed v at subversion revision +As Mircea said in we could pass the hash around and if conflict request and compare the actual XML or configuration objects to give clues to the user on whats diverging +For the same reason Jerkins wont be able to test the patch +I just committed this +So there seems to be issues before we ever get to the individual bindings +Will bug separately +The user has access to the through a JSP page so they should be able to set and hivehadoop variables from HWI +Thanks Roy! +is there a way to reduce this test case to something that does not use? That would give us a more solid base to find the have attached a shortened version of the testcases which includes only the test methods that do not use +Will this work for you +to +I typically do a inplace development where property to point to srcwebapp is also a already as well +testing should still not be assumed +closing because related issues are resolved. +latter maybeformer they patch +Yes Utf is correct +PatchI extended the algorithm file syntax to allow for double and and nested parenthetical expressionsI converted the shingle benchmark to use the new task and removed +minor some and FIXME will go away from after this patchSid Do the testcases pass for you? Some testcases are failing for me with the patch +Fixed the parameter usage with +addr port +The metrics thing would become clear as we implement more of HA functionality +Importexport will generate insertselect statements with quoted table names schema name and column names to be execute on the database after this patchRan debyall test suite all tests passed +is a mature branch which should only be accepting very and conservative changes at this point +Yeah sorry I mustve messed up something earlier when posting +Bart maybe a better approach is to change the implementation to not need the socklent parameter? Its implied by the address family so should not be necessary +Changed the page title with appropriate page title and also changed the screenlet ids Added ui labels for that for workeffort componentThanks to Ratnesh Upadhayay for helpThanks and RegardsAwdesh Parihar +I committed this +Ill be ready to help albeit from europe +something like which would be zero by defaultAfter your table becomes populated you set it to M so that old splits can be erasedThis would also make be a parallelizable operation and not a serialized utility like it is now +Please attach a project that I can run to see this error. +So I dont think that it should not be necessary to include a jaxp am I wrong? If not I would vote against closing this issue as fixed +Ill take a look now +Hi here is a patch that resolve the problem and the relative test case +This IS a bug +Attach My patch +I am not saying we wont want to enable this but it has to wait for if we do of the testing I did indicated that ordering had anything to dowith the behavior of +Wont backport to for the transition all resolved issue that did not see any further comment in the last month to closed status +According to Dave Grove the JIT in Jikes RVM is smart enough to eliminate all traces of the object allocation and virtual calls +The class uses its own constants which leads to an issue if these were compared to those which are created by the Parser +The patch for applied thanks Marius. +The patch is hacky and is a temporary fix +I am not overwriting but supplementing the validation +I just looked at the patch very closely +correction +If a javamodule has a bundleDir entry this value is taken +Already fixed with as confirmed by reporter. +Will do but seen it on master from yesterday +Sowmya how exactly does it fail to query the job result? In the youve attached to the bug I dont see any failures for this job no exceptionserrors eitheralenamacbookpro grep iE excerrorunablefail alenamacbookproCan you please elaborate +still should be uploaded mean if your users want to render the PDF table theyll need itAnd also uploading it will set the correct groupId and artifactId for your and anybody elses am going to withdraw my request then +I have and idea for lot simpler example +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +could we use this for having separate phases like described in and have them complete even if one failserrors +Spelling police told me occured should be occurred +This is why the for the value of is quick note I can also confirm that this is a regression as we hadnt encountered this issue before this release +I think the intent would be for the VFS provider to be GPL as well and so the whole provider should be shipped from outside the ASF +Committed as svn +zip of the app to recreate the issue + +For performance purposes should probably be removed eventually but wanted to make sure no one saw any issues first +Thanks +Committed to trunk and +The answer is that I just had to set the version of the findbugs plugin to in order to get things working again +Thank Kathey for all the good points +I have restarted a new build +bulk close of issues as part of making release notes. +Seems fine +Should that all pan out we should merge into master bump the version and deploy to npm then resolve this issue +The latest issue about Unable to find affinity group is not reproducible +Digging +were backported to as well adding the latter fix version. +It is sufficient to pass one parameter like a to compile for the report! +Hmmm still no luck with the archetypes +It would be simpler to use a direct binary encoder in this case +Verified on EAP ER +Attached is a unit test demonstrating the issue +Jason Im not sure I understand your last post correctlyAre you saying that AS isI mean its just been released with a and I would assume that a will come at some point with bug fixesAre you saying that for a production environment we should then use AS? Im already struggling with a move from AS to ASsI hope Im wrong but can you please let me know what to do hereCheers tried with latest snapshot from this morning and I still have the same problemIve also tried defining the dependencies on my MDB and it didnt workIm not sure what else to doThanks for the helpCheers can see its still waiting in the publishing queue +Again is for tests +Perhaps related to a or JMX? Something else that captures the signal? Have you tried echoing the command uses to start and then running it manually removing arguments until you find on that could possibly trigger this? Does the same happen when you run the minimal server? Have you tried to remove modules until you could perhaps find one which without it everything worksIf I had a Mac I would have a look at it but +Here is the Benoits suggestion +bq +Attaching a functional patch including tests +It used to read like apidoc +Find attached a patch with a cleaner version of the assigned this one to you since youve already picked it upE +We are missing things like adding a deployment URL to keep the programmers deploy location clean from the server +Checked in patch. +Duplicate ofNo matter if this will be fixed its annoying that it is logged as an ERROR +Triaged for assigned normal urgency marked as newcomer and repro attached +Hence marking this bug as closed. +all subtasksbugs were fixed so closing to set correct Fixed version +Martin said he would try to check this last week +You probably eventually ended up with dead topic subscriptions that would remained attached to the topic even after the websocket had been killed +Dear allEnvironmentTomcat should be Tomcat Could someboday help me to update the Environment infoThanks a lotMichael Xiong +Ill be sure to do that check before I submit any future patches. + +doh +Here is an initial patch for this one up for review +It looks like youre using Forge Please upgrade to the latest version and try againThanks! Btw +batch transition resolvedfixed bugs to closedfixed +But that doesnt seem to be the caseIf we would build the code with Apache Hadoop I am more worried about the fact that we expect users to use it with some other distribution +Thanks and sorry for causing the confusion +However by some strange reason it doesnt work without this +we have ways to get bundles started in a clean server listed in mentioned in a boot featureso along with the flag to copy the bundle into the system repo we need something to do one of these +patch looks good to meMy only suggestion would be to add a test for the side of the house to make sure that that isnt affected by this issue as well though you could also certainly do this in a separate JIRA if you wanted +Id merge it to both and +Uploaded a trivial patch for fixing this issue +What version of the sample are you using? Should be fixed in master +Thanks for the patch! +This is a tough one to crack as error handlers can be quite flexible how its configured and how its woven into the routes +Proceed with your original plan +and lower leak for the Ruby Thread case but and seem to be ok +It would have to be configured per action +Thanks for your time and let me know if anything else comes up +Im going to go find and vote for any tickets that mention testing nodes in degraded states +Per the JIRA committers meeting March assigning this to Yash +I adjusted the target milestone +Currently this is limited but handles all of the requirements +Platform jira for numbers seem to match up for me added a comment for a recheck with +Curious +Heres the regression Check out project Revert srcmainjavaorgxdoclettestappweb to version Build projectYoull get an error in the test there +Attaching a testcase on the offchance it may save some debugging time +Jucawould you please help me set up this db profile on our integration job +Same case descriptions of first post +The +agreed +Instead of changing can you create the tablespartitions you need in the new test only +After discussion we decided not to fix row limitation so theres nothing to fix in this issue +I would guess this would not be until though +Mike thanks for investigating thisResolving issue as Wont fix not a bug of our product +Need to wait for them to release the fixes + +Modified the description of the option required by the tools +I hope I interpreted your instructions well +My suggestion would be to simplify the log collection by collecting and writing the raw log files into a directory structure as followsI agree that approach would be simple but it has a lot of issues at scale +I combined an example of Geronimo testcase and a JPA example in the Please access the httplocalhost +Passes tests on IA Linux with a Sun JDK +can you attach the relevant broker and client logs and xml configuration +There when navigating from NN to DN NN http address is sent as request parameter to DN which is used by DN to construct the back urlI have implemented the navigation from NM to RM by reading the configuration +Ah thanks for this +Hi BryanYou should see in your classes tree +Have you tested against a default implementation for is the Java based one so this shouldnt be an issue any longer. +Its not perfect someone writing a malicious servlet can return back data at the rate of a few bytes every seconds and get past the timeouts locking up the WFS schemas but better than nothing I guess +Sounds good but dont see point of soatools +the patch looks good to me +does NOT raise an exception does NOT raise an excepton +gitignore Rename containerSubmitter ? What is the need for ERRORFILE? LOGFILE should be enough no? I see that all the messages you made to go to ERRORFILE are going to LOGFILE anyways +Still would be better to start writing a User Guide +Ill attach another patch +Tested the second patch but basically the same thing WRT performance +Ill take that out and resubmit the patchIll change the patch to make releasesOne for versionAnd one for the new trunk version that contains the new method signatures for +If there is a problem with this please contact me +Anoop Its late but what you say makes sense +Nice idea thanksMaybe a new javadoc plugin goaloption could be created that uses the dependency plugin to simplify the configuration for this +Will make patch laterPlease test this w at high load +changed to for and replace with TextCould you check again +Id suggest poking through their source code snapshot +Fixed updated to latest rampart and axis issue is resolved +Initial patch forThe annotation now has an additional parameter alias which allows to set the name under which the evaluated object should be put into the script contextI decided to concentrate on a solution for JSR compatible languages for now +Simple patch uses regexp to parse out a single int param +fsck and format are different it is okay to accidentally click y for fsck but not for formatOK another comparison mkefs doesnt ask for confirmation at all +Is permissionalteringstatement related or just a refactor that happened to be included in this patchRest LGTMNB I removed from in beedcfaececafdefdfe so you can drop that catch block +Fixes unicode conversion above UFFFF being done into charactersSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +So for older JDK versions it would work like before and for the current version there would have to be a no args constructor +I marked the wrongly named class as deprecated so we can be api compatible. +Anyhow some numbersOn my bb running OSkB wxhr mskB w msMB wXHR msMB w errord out +updated patch includes imports change +merge trunk changes that fixed the misspelling of apended +Some of that can possibly be cleaned up over time in the upstream sosreport project +Please let me know if a more involved detailed example is required +I think we decided on the dev list but I dont consider this a big enough issue to hold back a release +Since the class is an adapter for the type the PR should solve the issue +Attached is a portlet which is affected by this issue +change in bugfix in release +This is one of the goal of the master redesign. +This allows external client to access the database which is not the case for the pure embedded mode +patch code changes look goodCan you add the test as mentioned in the jira +TedSeemingly I cant +It would be great if youd stop linking to your blog on our systems +It is for standalone tests only +Bulk close of resolved wont fix cant reproduce etc +Maybe we should implement this as pending deletes per segment rather than using a generational system because with we may need to maintain the per queryterm docidupto per segment +This clearly has to be committed into and above +But what is the problem to apply that on the trunk + might change things over to mvn +Where it says The first two statements run and return correct results when translated to SQL and run against the Postgresql database in +I see what happened in the supplied patch I failed to remove two inheritAlltrue parameters +The placeholder in the reference documentation is +committed as r in branch. +I am successfully using these helper classes to get around the issue I reported in AXIS +noticed the property file is now located under mahoutconf so I created a patch that moved to conf +So its OK with me to commit as Samyes indeed i didnt add USA to the table +seems to be good now it works for youThanks for reporting back to us. +Failures are not related to directory service +file and remove the spring nature from the +This appears to be stuck +Patch applied +Aaah great Ive been meaning to do this for a while +In addition Ive added to in order to audit the source codes automatically +Looks good +I have checked for the presence before sending message to the client it does not show the correct presence the other way it works fine +Backup is also in the same dir +The gui is not well not the jbpm developers at least +buildivylibcommongriddevbin +Sorry for the noise +Attaching latest patch incorporating viveks comments +This error can be reproduced only when using a separate database for each cluster instance +You should report this against the product you are using and not Synapse unless you use an Apache Synapse distribution directly +Instead you should start to modify to filessrcmainresourcesmainresources attach a patch with the changed files here +JimmyLast days i was not in town +Jaimin is this needed for +Thanks Lincoln +After that is in place I would like to commit v if you agree +Ill add it there and this can be closeddeleted +See +If youre doing the collision check in both schemes the random assignment keeps things simplerThe possibility of collision with sequential assignment of is very remote but why expose ourselves? Im probably being paranoid so ignore me on this one if you want +Currently we cannot cast to those subinterfaces through a proxy +IE doesnt place the popupPanel correctly FF does +updated patches with fixed merge patch looks good in general but I could not run it +Move minor issues to a next release +Yes I do have in my classpath + on this one +This has a side affect of possibly causing objects to be instantiated twiceThis is something we obviously do not want but so far we havent had any sideeffects may be because its now a dead code altogether which just kick in by swallowing an exceptionIMHO it would be good if a quick build could be setup without this second retry +Regarding building the whole thing thats really a task for a tool like Hudson +Any idea when this might be done? I want to know whether it is worth using yet or +Most of the stuctural changes are in line with what I have already put in placeYou will see that I have deprecated class in favour of the newAs you pointed out this update comes at a time when there is much change and much discussion about further change including the suggestion of not using the junit parameterized testing feature and providing similar behaviour where required through java features alone +They simplify it a lot +Maybe there should be two Full Interface pages? The current Full Interface page has the handing a bomb to a baby problem for very large indexes +New patch with html which fixes the html encoding of fallback patch Uses filename from new field resourcename with copyField to text and included in qf Handles HTML escaping of toggle all fields Show file type icon before title first detected from filename then from contenttype Do not show author contenttype resourcename if empty Refactored the toggle explainallFields section into own fixes to icons for docx and pptx as well as file prefix for solrcell filesThink this is getting ready for committing +Metamer facilities dont influence the sample +In any case the modification operation has to be reviewed entirely the schema code for it is overly complex +I am not sure how many people rely on this default behavior +you know all fields coming out of your call to the Tika API will be Tika fields +Also could you provide a patch from within the trunk so the patch gets the fixed class and the test in one shot +I need some help reproducing this to decide if my fix works +Lets call these the first level CRC +The blocks level are read so infrequently compared to the level blocks that I suspect the cost of the occasional random read for a level block is amortized away. +Tagging for family +Im agree about CDI spec contradictions but Im testing Seam with OWB in hope running it on Websphere V +This patch appears to fix the problem +Ive reimplemented the synchronization to use local variables for lockingunlocking instead of the document field +If its not at least the patch from matches mine +will be released in +I already set the following properties in And copied two jasper jars into hivelib But above didnt work for this issue +let us know once populated +Fix typo in the description for +Code something Paul +Group with null values werent handled properly +The Changes HTML is added to the source tarball and is not existent in SVN +Yes suggesting improvements are best done directly through JIRA otherwise they tend to get misplaced + +If not Ill look at this tomorrow + patch applied in r +I know you sent those characters in the original post on the users list butcould you add them here so we might have a better look +Will look at options there when I get back to updating that patch soon +Personally I prefer the approach recommended in the forum to that anyways +We can rename it to +pushing out to theres no patch for this yet and its release is already resolved as discussed above. +I think its good to go now +Please see demo attached to +Yeah I think the above code would work looks kinda like the code that I have at the moment +I cleared the estimated time as I misunderstood what the field would indicate +Closing this record since is now released. +Can someone give me more detail on this projectThanks +If compression can be added in without modifying or then all I need to do is add a test case +Whenever team A release incremental version it should picked up by team B without changing anything +I can agree with i think today anyone that wants to do things the way you describe can just create the from scratch alreadythey can do this and set everything from scratch add the field twice whatever they wantBut if we remove the ability to do simpler things like i want a with term vectors enabled or I want a with boosts then I think thats a big loss to less advanced users with no gain to the experts who can already do things from scratch anyway if they prefer to do that +It would be cool if you could some various uses of the catch integration in a jaxrs example +I believe this is a Pluto problem + +we now have an alternative property index so it doesnt make sense to keep this issue around. +This is just temporary until the real fix is ready +Thoughts + +Additional comments The main pig script file name should be available from Antlr parser which can be set in the AST node +Hm +Fixed in Rev +Zhou did not start to work on this I can contribute this issue +Note that I uncovered while developing this though this isnt affected by that bug +Vishal We havent figured that out yet +Twilight zone indeedI uninstalled what I downloaded today and installed back what I downloaded on October +This is going to need some further investigation +But given the importance of audit logging if the error condition associated with that logger is fixed we need a mechanism to reinstate that logger back without having to restart the namenode +I think weve agreed that while design principles are going to inform the approach for some parts of the API they are not appropriate for all parts of the API +If I run just mvn tomcatdeploy then it authenticates +I am marking this issue as wont fix since there are no plans to make any new releases in the +Committed revision committed to trunkthanx alex +Fix verified. +Thanks +The only way i can test it is to wait for an exception to occur this usually happens while im accessing the mail store through a web client at the same time ofbiz is trying to access it so Ill just have to wait for them to clash again +in +Change the import statements for the last set of filesN Delete and as well asany references to the package itselfHow many total different patch files do you think you needThe patch you provided for the java directory looks fine to me +Thanks for release the fix. +Thanks Hitesh! +The problem concerns embedded in +Thats GREAT! + +What about this one? I have encoded the version with QCM messages which has no compatibility issue I think +Issue fixed +Patch which addresses the issue +Also in +I still think it may be a good idea if we could detect that we were heading for full disk and abort everything before things crashedThats a good feature to have but out of scope here +Done +Thanks again +Linux Firefox +But to date its not yet looking like there is support for a commit of something like this so wanted to reflect that in JIRA +Further browser refreshes should deliver the app though +Same issue also exists with Xerces IRIX +the Projet Id is used as the ticket number instead of the requestIDMCA is used to link the return messages to the correct Project as The messages are timelined in the project +looks like it is not shortened +If you have some time and can generate the trunk patch Id appreciate it +These patches fix the createlist instances issue and make the client code more robust and flexible +Yeah go ahead Bilgin and work on a patch with the abstract class components guys if you think that the abstract class in the patch is fine then I start adjusting the components +Please reword fail message to be something likeFailed to enable table tablename after waiting for sec +Oops the patch didnt add an obj directory which the make file expected to already exist +Let me think more and come up with a patch if possible +Its copied from Oracle JDK and I admit I dont know about license things +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Hiattaching patch for this issue +Hi BrettYes this fix works for my testcase so far and we are still waiting for customers final confirmationWhats your concernRegards good to know thanks +The attached patch demonstrates the conceptThere is an issue preventing it from being committed The XML validator reports errors because the schemas dont contain a target namespace +Please verify itwith the latest code +patch the file to retain the exception text and the underlying cause for a bind failure during name node startup +Since all mentioned Rubyspec failures are fixed now Im going to resolve this bug no need to keep it open. +Regarding having it in two places is an example that is present in both test and examples in current and older hadoop versionsAlthough I think we need it in the examples I am fine if everyone agrees we shouldnt +This had been the next thing I would have asked for +We have at least jobs on the go atm +Ive unassigned this from myself +I promise to commit it as soon as testing is restored +Committedtrunk rbranchx rlucenesolr r +will investigate tomorrow +bq +What do I need to do +This discussion has gone stale +got it +Reopening the issue until the correct status code is does one subscribe to that mailing list? I couldnt figure it out from the ietf pages +Im seeing the same bizarre behavior here so I should be able to track it down +Replace these files with the uploaded versionsThen apply the +Merged from cycleclean branch. +Some minor nitsTo be consistent let all properties start with TT in +For some reason git prepends a bogus directory to all the paths you can apply it with patch p instead of the svn patch p +Thanks. +Each has entries when blocksize is MB +On what filesystem in your example are tmpfoo and tmpbar? Are you suggesting that the patch should limit interoperability to the local filesystem? Or that the default unqualified file system should always be the local filesystem and that people should have to use fully qualified paths if they want something else +Alejandro your version does not support you could patch this to work I would be issue of inverse is worrisome +This bug is my mistakeThe patch looks good +It might be worth adding a are my filenames going to be too large check against all KS CF combinations before starting to migrate data files around though +I had some trouble merging from trunk to and needed to just copy a few files i hope that does not muck stuff upThanks Olivier +my bad +to Jyotsna Rathore and Pranay Pandey for working on this issueChanges are in rev Ashish Vijaywargiya +hi Chris Yonikdid you check out? it is trying to achieve the same thing +Added as blocker +To reproducePut the attached files in a app and access requiredParam then click Bang + +Hi Mit the patch updateditens fixedItem fixed by adding a timeout and checking of error absence during the server startupBesides the following is fixeda added check for exception in server threadb server start method in not Before since the server thread is needed only for testcases of c server runnable class made static and renamed accordingly to its typed server thread is now joined in After +Can you please try reproducing it +Yes it will be a bit of a pain to get it over to the FUSE side but not impossibleFor now Ill wait for some beforeafter performance numbers before doing anything on this. +server is now listed as child. +Currently this is only copy of old Session interface +Yes the test runs the dirlist example +Since we made as public findbugs triggered +Using this patch the shuffle mostly keeps up with the maps during my sort benchmarkThis patch also allows the task trackers http servers to try multiple ports until it finds one that works +Rob can you explain where this Deployable is? I think the conclusion is that by default the Deployment tab of server editor shouldnt display modules that are children of another module correctBut all I can see in the menu is All Deployed and By Module Name +Im closing this as out of date +Attaching atch for trunk +Maven Assembly plugin? I searched around for others with similar problems found nothing which points at some local quirk in my setup +both pushpull and metadata can be updated to when it is released +the line is no longer supported +Thanks for tracking this down! Im not sure Id use the word simple here but your descriptionis quite clear and your patch looks good to me +Which manifest field are you talking about +Just added the Apache license header to which I forgot to add before +Implemented +Fixed. +I add splitsize as a parameter +Tom this looks to be the same as which went in a while ago +Easiest way to start the maven in debug modemvnDebug jettyThere is another way to do the sameWindows set MAVENOPTSXdebug XrunjdwptransportdtsocketserverysuspendyaddressOsx MAVENOPTSTXdebug Xrunjdwptransportdtsocketserverysuspendyaddressand then mvn jetty admit then that Im a little stumped by this one +Stan didnt you fix this in? Do you mind take a look +I forgot to reset the last when we sync and also prohibit back to back sync marks +Can anyone merge the diff with the trunk? So what happens here is simple if someone adds the to an eclipse project and when exporting a Runnable jar choses to package jars instead of extract jars then any pig script will fail with the exception rsrc +But if no one is signing up to review the pieces splitting it up just slows down progress in my experience +Big Im not sure but it could be that atm m is used for generating the site so moving this to m should be done as well +So just review docs for moving +Just wanted to clarify no target directory is created in application directory until a command like grails or something other than is run +Verified with JBT. +I the working fix from the branch back to only files are required frameworksrcjavaorgapachetapestryservicesimpl frameworksrcjavaorgapachetapestryservicesimplNote that the above two files are using the JDK class so if youre compiling in youll need to replace it with the version of. +Hadoop loads native compression libraries +I attributed Lucian in the commit +It appears that this is not a documentation bug because the option is deprecated and should not be documented +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +I have a patch that solves the immediate problem +Initially local build failed when doing a complete build after removing local maven repo +Thanks for this enhancement +Is the nio selector stuff changed only on trunk or on alsoRegardsUlhas patch for component on trunk. +I expressed some concern to Richard about the generality of the patch + +committed the one line change to the trunkmsvn commitSending javaengineorgapachederbycatalogTransmitting file dataCommitted revision +Therefore to fix the problem the sorting should apply to +Yes requesting a test case was not the right thing here +Ill give the timeout option a try locally and see if I can reproduce the problem +In fact this cant be fixed in M because M includes Teiid +Perhaps its all necessary but it seems a bit much to me +test on all tagged platforms platforms for now +Just uninstalled the GRE restarted Eclipse and installed the snapshot version +bulk defer to defer to +One last thing can you include a unit test to verify the fix +This requires some hacky logic to fix the paths in the case that the user set a non default location for the classes or directory in the surefire config +Moving to TBD for fix dont think Ill get to this for Beehive ship +The Namenode should remove and replace corrupt replicas when they are reported by clientsdatanodes probably on read to start off withThats already done +Thanks Philip! +The fix was included in +I have applied it +Patch attached +Ive added a test and committed +handlres were ported to an Axis moduleThe interop scenarios are included in the integration test suit +All the classpath entries contain absolute path of your machine +Hence closing this issue. +The attached is a patch that fixes the can you please provide a unit test +The sql script should be auto generated by JPA provider +We need evict invalid objects on the preUpdate event on the hibernate event listeners so this should no longer be an issue +Marking as critical since this prevents usage with SOA platform +Same patch just different messageNote that I manually tested this +Also delivered the change to to add the xstream +Thanks for the comments! Ive uploaded a new patch which should address all of your comments. +This issue has represented several things but what we agreed to do is to make JDO compatible with Maven but deferred the decision on how to build JDO and run the TCK to later +minutes is good enought +Thanks +Thanks for the review Suresh +So this requires some maven work at the build time but this is a MUST fix to me +Asokan +Is this happening on restart of the server after the bundles had already been resolvedstarted properly? If so please attach the debug log and tell us what rev of master youre using +excellent +It looks like the Configuration Admin service extensions have been dropped from the release in favor of bringing more substantial changes in a later releaseThus I resolve this issue for changes have made it into the specification and some didnt + for the patch +I dont think is relevant here since it landed at +Modeshape has been writing a json parser they have plans to release as separate module as well as many projects in AS use Jersy +Hi tmjeeI have created a test for the issue +It creates a new float array but does not free it +I originally planned for Official and Nightly but Chris and Doron have suggested other versions as well +Im working on an contribution but the public release process is slow at MITRE +If you want something else be specific +Please use rather user ML to discuss such topics you should get much more answers. +Cfr +Thanks Todd! +Applied TRUNK +Since moving its local mode to Hadoop local mode Pig no longer needs this layer +Alexey I cant think of a good reason to scan derived folders +Our paths should be relative to the esbcontent folder +ok Ill do another rev taking account these changes +I can volunteer for this one but I would need to know what you are planning for this +Please keep this out of the core +The patch does not introduce any new Findbugs warnings +Verified fixed in BRMS ER +As rev is reverted and being reworked +is fixed in Hibernate ORM +The exception is thrown prior to the data extraction. +Committed to trunk at subversion revision and to the branch at . +It is important from a debugging perspective that we see something like a warning data import did not launch because of mutex acquisition failure +Have you already made progress on this adaptor +Thanks for your review +Thanks Zheng! +This is just confusing since no view is rendered in that case. +The screen will be located inAccountingGeneral LedgerAccountingReportsIncome! Thanks +I have a slight refactoring and tests to prove that it works +Still no visiting yet in this patch +PiRunning faster is part of it +We recently hit the same issue +Dont see how we can do anything about this +Fixed with close of all resolved issues as part of closing items. +Would it be an alternative to the current implementation? Or done as an enhancement to the current implementation +Need to fix before Thanks in both trunk and where can I obtain new distribution of this fixed? I see this fixed on Thanks a lot Deepal +This is only available in webhcat I suppose +This is an important documentation task but not a show stopper for release have documented how to deploy gfac on tomcat +Also after trying to reproduce this I think the issue has to do with performing a join that has an implicit type conversion +Updated I cant see any fix in the file you attached have you uploaded the correct file? +Hmm not sure about this although the copyrights are already in the LICENSE file +Verified on work properly. +Attachment has been added with description Patch to fix licensing has been added with description For clarity attaching the entire source has been added with description For clarity attaching the entire source applied thanks +Youd have to pull them out of the component tree save state somehow though then reinject them +Patch from rb +Oh I spoke to Pete about this a while ago +Closing issues which have been resolved for some time +But in theory since the lifecycle of those objects ends with the request we can ignore it if they are created at startup +fixed as well most probably gets fixed by reverting to DWR +This snapshot is the crashed scene +From reading the bug at eclipse I assume this wont be fixable until we start using Eclipse bits which wont reference update sites at all +I will be away for the next few weeks so I wont be looking at this for a while +This issue needs to include a unit test that uses binary keys and many regions + complete. + +Perhaps changing it to +Guys I was thinking about approaching the Tinkerpop community to see if anybody there is interested in contributing this one +The cause is the same as for +Spring configuration of logginglevel should be easier currently its an enum that cant be set using a String +In other words you want to make sure when youre at through the transition it roughly amounts to of the blocks not of each block +Mass close all issues in resolved state that have not received further comments in the last month +I didnt think that a bunch of delegate methods needed to be expanded completelyForthcoming patches will be consistent with what Ive seen here. +But these tests were unable to replicate the reported problem +Are there any others +The opaque uri would not work with how Path works right now and I think it would be difficult to get it working as well +Fixed into rev +Thanks Pam and Owen +Edited the a bit to deal with some incorrect paths +I think the form is already pretty crowded so we should replace the current estimated unit run time with estimated total run time +Is this solution accepted as final? From my understanding any analysis that falls beyond the pages cap wont be browsable via a filter +Well look into to do releasediscovery of plugins w SVN over https and possibly SVN +Thank you Nicholas. +The jsp pages then use this to get both the machine name and http port +Ok that is clearer +but a clearer test case can be found in the duplicated issue linked from this one +Define method to clear cache in +Ive taken that out for now and am making some progress +Thanks Vandana +Didnt read build is fixed submitting patch for Hudson to verify +release bulk close +That was my mistake sorry about thatI shall commit this shortly +Committed nodetool version with minor naming changes and removed describe ring from CLI. +So I guess that should make your use case a lot easier +Thanks Nadav +The other half of the fadvise support is in +I think you have to start thinking about reordering and partitioning strategies +Can you please try it +I checked for a BOM marker and as far as I can tell there isnt one +Some extra infoI added a properties field to classes +Thanks Marek +Has it been reported to the Eclipse folks +Enter +Can you paste in examples of it working? How will we doc this facility in the shell? Needs a nice fat release note too +Not necessarily related but can you think of a way that we can keep and in sync for these kind of things +If someone can indicate that the bug does exist please reopen + is attatched +The patch contains an abridged version of the getting started document featuring rewritten intro sections maximum links to other relevant materials repplaced section with cmd options with a link to the wiki page abridged scenarios to decrease the level of detail and remove screenshotsthis is not the final version of the page because several are left that require others attention the new section about using cmd options to modify VM operation is empty +One developer out of three objected and thats a consensus? That developer basically said this class should only be Serialized if X Y and Z were true +They should be isolated youre correct +assigning back to Rob since I did not get in a quick look how the cathegory for the adapter is set +Committed revision +So all issues not directly assigned to an AS release are being closed +The patch passed srv on Win + +Why did you add in +However it will be interesting to see how the Eclipse Designer can cope with this kind of shortcuts + + +AdamGotchya +Is there any reason why we dont implement our own with default permissions? Then we could just pass in the jar names to its constructor instead of relying on the system properties +I have in there for both currently which is whats been done elsewhere in the code +Does that make sense +Thanks VasiliyI a waiting forPleas write me I help for you modelling testing debuging +Feel free to add more tests if you have other scenarios that you think need tested +Fixed in nightly build +Of course using test or disabling Spring runner is not an optionThe problem I am facing started from migration from older surefire spring junit +linking to an identical report +Please go ahead and merge it to dev +The real unanswered question I have is this must we really create the page map when the link is first renderedno only a pagemap name should be appended to the url +Otherwise when you redeployyou end up using the ear classloader from the previous deployment which has been undeployed +Can we contain the statement at all though? What if you are just working with essentially ascii text? Basic English chars? That type of thing +The fix is simple +Of course this wont help people that use other +Fixed by Brian. +Fixed on +JieshanPlease provide patches for other branches after addressing review comments +Fixed +The plan is to add a new entryto each job which has parents in addition to log the job DAG of a Pig script to the Pig log +I think your patch will get in very well +Keith whats your views on this +moving to +was this issue resolved with jboss as release +This patch looks good to me +one could do more with the base class I believe. +A is shut down session on B is now in BUDDYBACKUPADEAD +The component already offers this feature +This is an updated version of the file that fixes the condition that led to the AIOOB exception +buildivylibcommongriddevbin +DuhTesting you patches now but looking at them in the raw they look pretty good already +Documentation tested it before posting and I thought it was a bugThanks! +Does not look so in everything but pubring now works across all platforms +manipulation vpn reason do not appear in the logI do not know how to make that logging was in debug modelog attached +Unless someone jumps in with a good reason not too Ill apply this one sometime soon +Thanks Raghu! +Is this problem still present with a recent checkout of Cocoon branch +Ill send you an email +I think a similar patch would be applicable to those connectors as well +jvm core dump occurs after applying the you using the option socket +I will change the test accordingly or is this already shown in some test +Thanks for the test case Jeff! It is very helpfulBTW the description of the problem is not correct +This patch makes it possible to create aggregates on user defined types +Here is the diff +See comment above. +Were using lucene in a large distributed search environment and we would hugely benefit from this patch being available in the main Lucene build +Need move the checkbox up and disable table when it selected +Does anybody disagree +Yes why not but only if it can be done with little effort. +Post we still use xml encoder clone the query plan which doesnt work with Java +Niels please resolve and close this issue if you are satisfied with it +This JIRA covers stop copying hythr library on classlib build and is not related anyhow to changes in VM codebuildIll file separate JIRA to cover that +This is correct has never been used in the ESB integration +If you can confirm that JAXB likes the above then I dont see any issue with that change +Unless Maven has a way to specify different dependencies for different Java versions thats not going dependency on individual object models is scheduled to be fixed in when well debundle the individual models +Since its configurable it can be turned off for clusters having lots of users +See for further details. +We will not use a replication strategy to maintain the schema keyspace +I moved the feature request to its own issue this issue since this was about being compatible to the latest in terms of existing features. +for the patch with from Jenkins +Untestable means untestable. +found. +You should not have to call setServer on anythingWith my changes I was able to run up the server with the normal contextsSo I will look at your change but will probably back it out and replace with mine +Your questions are relevant to + tcpdump log excerpt from ldapauth pGINA +OlegI think it should be OK as we wrap bundles for in this way for about more than half a year no one complain about it +I will add this as a known issue to the release notes +Nico wroteok ich kann mir noch vorstellen dass ein falsches Encoding benutzt wird wenn dein Java nicht standardmig benutzt +Working on this issue will entail adding a number of new GROUP BY tests Ill open a of this issue to convert the existing lang tests to so that the new tests can be added as part of a more modern framework +Fixed in V +I was mainly asked to return stats on nodes in the form of map data structure much like how the prints out its stats +Just updated the feature file with thanks to +You took so long JenkinsI just committed this to trunk and branch +We can reuse the core and ui plugin better for our new ws wizard change the facet definition in add a definition for to support add a new jboss ws facet add a icon for jboss ws facet set a default value for the jboss ws in the preference ws runtime page add a Null validate for in the +Not sure thats a good idea +You should also not be able to set the in NFIMO i would keep and only remove Fieldable as interfaces are not wanted in Lucene +HiI like to solve this problem but can not reproduce it +Workaround sufficient. +Spoke with Tom and have done this for him. +updated component and priority +I have changed the to given solution +Hi TomCould you take care to update the FAQ on the homepage accordingly or close this issueThanksBernd +Can I deprecate the method now? Or we should wait for bigger review +Thanks Romain +We just need to ensure that were not doing too much work to get those each time as the is called on a TON of requests +it will require two resolve steps in a build where the first will resolveunpack then something would remove the result of that unpacking then the second resolve step would avoid unpacking +These could all be fixed together IMHO +The patch is the correct note that the ramfs doesnt perform checksums +Done +bulk assign all unscheduled issues to next release +This v adds a fix so minihbasecluster can run w more than a single region server +The reverts were committed +I think it is very easy tofix +make your changes +with CVS See also testCase +Gerhard do you have a fresh patch to offer at this time that will resolve and? +Making instance variables private +testCase chat with ralf a more simple testCase attached the former is too complex i think that the trouble is with only complex object graph but its occurs with more simple object resume if you query Product that have extends in two diferent transactions you get same timestamps for both but if you query a object that references to Product the timestamps of this referenced Product is diferent than those one queried before +Thanks Carl! +Thanks Mike +Ill contact Danesh and see if he remembers this issue. +Reopen the issue since the commit was reverted +Closed after release. +updated affected for a possible workaround +Ive moved it up into common +but I this wouldnt be an issue by has been fixed +A s is not validated because the s bean class cant be determined you sure that this issue still exists in Spring IDE because it should be handled by and +Im not married to the current name but Im slightly inclined to leaving it as isAs for the second point I agree that the chance of this condition being violated is remote +After all the recommendations will be for those movies that should be rated highly +can you attach your pom +No specific fix for this has been added as of now +Reopening for discussion +The f is still not implemented here +The patch also include a couple of updates for as well +Im not sure this issue is totally fixed +Lets do one thing at a time If we get that in no reason we cant move on to the next bit +Resolved in revision +Hi Thejas let me do that +Could you please also try that out? If it is good we may make it a default setting +It might be worth looking at whether CXF can process references to that are not in the SOAP request +and no noticeable reduction in performance in small datasets or close up also added a parameter to the datastore allowing a client to specify whether to use gzip or not +You are correct It should be qname and not namespace +Caused by r +Ive just committed this +Verified in as part of +I was using the javadocs from version on not from the trunk +Thanks Paul +Done in trunk +You owe me a beer! +I was a bit premature in closing this +because its a single object being returned +The additional javac warning is a deprecation warningCommitted to trunk and +the line looks like the fix I had in mindyes the invoice is used in accountingcreate should be modified to to call capture only if the order type a lot of efforts but nothing cordinated to fix this +Committed revision Thanks Matt +If this is an urgent matter please send email to +LeonidCan you please attach a patch and resubmit +Thanks Bikas that makes sense +Things like In this chapter we will discuss how to manage installed AS servers from within the JBDS +Hi JerekI think it is fair to say that the attached pacth and test were intended to fix the problem described by this jira We currently convert by value artefacts after generating the which means we will not pull in converted artefacts when no is provided +Disabling CDI support means unchecking CDI support in the project properties CDI tabIf the Facet is manually removed the me configurator has no way to know it shouldnt enable it on the next project update +Updates patch for a multitenant changes in +Please review and apply +Shouldnt this be OSTYPELINUXTruth changed it to LINUX +This patch does not affect currently used codeIs there anything I can do to help speed up integrating the global data store changesIts a bit frustrating for me if I have to wait and cant do anything +Certainly Jeffs example passes +Hi LahiruMany thanks for your patience on this issue +Its fixed now thanks for your report. +Patch applied on branch +Especially I reverted the patch from which solved this issue you can guess using this code base my problem comes back +If anyone has time to review and has comments I will address those in a subsequent checkinI plan to backport this change to and +Can you suggest any alternative +updates +Its really poor design by whomever wrote those servicesMarking invalid as its not something we would support +I would like to ultimately see the tests removed from the classpath +Ming Ma on patch +It will remove the user when the rpm is uninstalledPer feedback from Eric Yang on the useradd call has been modified as follows Long form of some useradd switches used for readability +I realized long didnt map to the type I wanted so I changed the generator to int but forgot to do the same on the object id +I prefer to use compressed archives and had set my environment accordingly +i would help working on this but before should be committed +Thanks for finding this +if you think its not please explainng why not please +Fixed and merged to branch +rename choose to chosenBlocks Done The new test looks great Thanks Eli +This may be at any time if there are more tests to contribute. +The new importer does though allow hooks to be provided for the purpose of transferring state from the existing module in memory to the new module when it is being loaded. +Patch for branch +The new testkit based mail transport tests use Greenmail and the old tests have been removed. +Please provide an update on your issue as a comment +Looks good to me except that the missing semicolons have not been added +A patch commit is either functional or reformatting the code but not both please tidy up your code +I mainly think about implementing a Controller with just one Script +No Steve we want a paragraph to add in the release notes to warn a customer about this problem with escape syntax in LIKE clauses in case they try and do that with a PGSQL +we still want to detect network errors so lets not skip those checksums in this case. +to update fix version +Oh Mighty Bartosz please show us your mercy and prioritize on this task +The JT is a long running process so it could detect the change and maintain a mapping +looks good +It will have to be MXML +Notice there is no register command in the testing scripts +generate a simple junit xml report file indicating whats issue requires further discussion +What do people usually do about this? Do they resort to the trick of instantiating multiple versions of the function with differing numbers of arguments +Its MS SQL Server +So in fact if we want to keep behave like today we should use and +Yeah this is a pretty big deal that effects protocol correctness +setting to for the new mock thanks again +The only difference to is that you have to parsethe chunks and can not just skip a known number of bytes +See for further details. +Which still have gaps but there is not much that can be done there for those because the renderer simply cannot imagine how join and caps should be drawn using that svg element +Im hoping we will see a more elegant solution later +appreciate your patience +In other words we should not have to modify our core modules or the application itself to pick up this dependency +Changes committed in revision +Bulk move of M to is moved to next move from to to to +with password +It also supports tez being deployed on all nodes +I am on this jira and have a general idea how to do it +Thanks Karthik +That is bad API design it should return an FST with nothing init +bulk close of resolved issues as part of JIRA housekeeping for release. +I think we can not do the second point mentioned here +Good but we should not change the method +First note to thread on jboss community forumIn example given by reporter is used ajajax with listener attached to selectitem event +Lets run it through againIf successful Ill commit +Fork an new dialect? +Reopening so I can resolve it as hopefully this will be the last change until the RM closes it. +I guess the only problem is that we have actually changed to not be regex so we are still impacting end users +Well this is not the first time it happens +Good catch Todd +. +Also the underlying feature is automatically supported with the use of message builders in Axis. +The mrecursiveGuard property ensures that lognet discards log events as soon as the recursion is detected if it is configured badly by linking one appender into another +It seems to be a good idea +It will be great if a new version of Java DSL is somewhat more strict to reflect your expert knowledge of what is a good route design +Can you tryright clicking on a row in the table when the menuitem appears just move the mouseaway so that the menu item hides then try right clicking on a row again? I mostlyjust get the browser right click menuThanksBrendan. +The ctor of takes a parameter +It always throw a now Im looking the karaf codes and trying to figure it out +patch looks good +It isnt possible to fix it in because we use new maven artifacts and for that it was necessary to modify some part of continuum +EnjoyI noticed acouple of ex in the file not in your patch should that not be px +I think its bug of model implementation +The reason is simple few people follow JIRA tickets whereas many users can contribute to the discussion and share ideas on the user mailing list +Fixed the overwrite for copy +Im closing the ticket for nowPlease reopen if you find a use case where the applied fix breaks the functionality +Also note that this jira was created for the branch and weve been recommending that users move to the branch. +Verified on upstream serialization does not fail any longer +batch transition resolvedfixed bugs to closedfixed +locking on app in case of allocate call wont work +Resolved conflict with +A API could substitute though +replace the patch improved +As a data point to that question we hardly ever had CMS collections on and maybe one full GC ever that I can think of for what was years of cumulative uptime +Hopefully that will pop it +Im on committing this if others are cool with the design +its a different object +Perhaps yes if only because otherwise you could never quietly migrate forward and drop a field +okay fixed +bq +OK Ill submit a patch +IMHO elevating the tickets status to Blocker wouldnt be a bad idea +Anil Chrisin rev +in +I just discovered this bug and I wonder if it will ever be resolved +Thanks +Indeed thats the behavior Im seeing. +Then what you said work you can inject the datasource with the java +Hence I will remove it and work on getting a reproducible testcase +Here is the like this is fixed? +Can you verify +Software ext mmorin Original Message +Ill try to commit his patch before I leave Vienna tomorrow for a week +looks good +I will report back shortly +Ive added the dependency to eduoswego to modulemain and have put the jar up on the m repository on +Shall we add the user who started the to the admin ACL +Max channel capacity is another one that is worth exposing +In this case and comparator should be mutually exclusive. +Interesting +Did you mean to attach anything already +Id encourage those bugs to get reported. +It will still have to be a preference or at least easily disabled so that for apps that dont want it can have it off +If there is a way to do this in Ant then please do so. +Apologies +Improvement of the with First action is added in the Chain History API to view the flow Parameter allowing the chaining to a previous action +Whenever the is set to a size smaller than the current anyone care to argue for either of these two or for a third option such as setting the position to when the is changed +This is replaced by the epic. +Committed at r in trunk r in Jacques now the calendar have the right image and i removed th X +Writing new concurrent scenarios should be easy with this to a future release as a pointer to check if new scenarios have popped this release we should at least find out how this will fit into the overall CI hudson t close this issue until concurrency testing actually runs on the ci for all supported dbs. +Putting up a patch for overall review though it is not completely polished doesnt have all possible testcases and documentation is coarseincomplete +Please make sureyou have at least one vector layer and at least one style configured +Deferring post console work +I just stumbled into this ticket today +must definitely be fixed +Also includes some other performance enhancements now uses LIMIT when making against tables for purposes +Attachment has been added with description Test case to reproduce the issue +Ive tested It will build +Do you agree +backported to with revision . +This was fixed in. +Method has been removed will not be there in ER +No statistics was being very much +OK the last test had a silly bug that made update QPS slowdown even low indexing reopen rates look worse than it should beThe test ran on a fully optimized index ie it had no deletionsSo the pure searching add tests had no deletedDocs vector to check but the update test after the very first doc was indexed had to check the deleteDocs +Jeremy Any response on this? Would be good to get some resolutionThanks to appears this one has been inactive for over a year and the patch as submitted was not agreed upon as something commitable. +Didnt incorporate the rowcolumn dividers fix since that is contained in the yarn css file and will affect all pages +Reverted changes previously committed by this patch. +X produces bit binaries only +the class was just one example +Fixed both on and trunk +This is programmed to the public API that Hibernate supports +First thought would be to decouple mapreduceyarn from hdfs then +When I said basic I mean the default setting +As Alan commented we need to fix the parser to disallow this use case +Ill open a subtask for that +submit patch +XHTML page that showed performance issue while rendering HowardTrunk was updated to owb trunkNot sure the behavior is exactly the same +Ill give another patch shortly +Unless you can get me a patch to fix these issues in the now building branch Ill have to engage you later after is out +I think we will want doubles but perhaps we can have two implementations and people can pickIf we need something more than small integers it is likely that we will need doubles +As far as I understand the problem it isnt specific to a database system or version +No this is by design +Please stand by while I consult the phoneME experts +Forgot to remove my debug code +Infinispan doesnt allow inserting keys with null values how did you end up with null values in the cache? I think this makes sense as a general requirement +A more general enhancement would look at for each row plans containing nothing but a series of update statements and turn that into a batched update plan +This issue has a few votes already Id love to hear some specific use cases for this issue +Karthik I was going to open another ticket regarding this issueIts quite confusing to do post development on a branch while is done on trunk +At least a JIRA should be filed to add some automated tests for this functionalityAaron please see the label all these jiras will need subsequent tests +Now note that has a method called get where it could deal with any number of to determine the dominant DTD + test this and then close the bugUpayavira +Andrew can you get a quick fix for this issue +MattI guess what I wrote WAS quite confusing +But the asynch annotation is part of the EJB standard and I think that this compansates the code changes in the beans and clients code +So maybe that didnt fix it +Then its obvious that theres a time for the blancer and that there are rpc calls into master +Ive written quite a few tools like this before and I hope to be able to hack an independent tool today that is tuned to James +Ive taken care of all the I could find in +bq +Ive just committed this +combined the tow fixes into the new patch +We need to enable studio to setup without having to use the +Is this documented somewhere +thanks +Added on revision +I think this is a really bad plan +Thanks Adrian +That is one you should be using as well as platform bindings for launch shortcuts +Committed to trunk. +This has the same root cause as +Please let me know if something else was required I will update the patch +We might be able to avoid this by have the target remove the cached plugins tooOther problems dont seem to have recurred since bumping the memory settings. +This just means that you need to install a version of mina to be able to install +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +This patch gets rid of the separate test class and combines all test in +The patch has been applied +committed +I think caching the SMD rather than the token would be cleaner and less +Definitely a dup of which has a nice patch attached. +this has been due to classpath clashes + for the to trunk +Patch to apply on the arch projectChange the code to use the new arch format and transform the command lineChange the version number from to +I have no problem with a dependency on Commons Lang or on Commons Collections for that matter +Patch regenerated for use with Hive +This is not just an issue about +I couldnt merge down from there because was not reverted from +Done thanks +Closing all issues fixed previously after a brief review of each. +Good caught +Note that once the column is created you cannot change the column attributes +Maybe a separate depclean target should make things Really Really Clean +Hi ChrisI would be glad to see this coming trough is it possible to volunteer my time to try and fix itWhats the procedure for such offerKind regards +I applied Denis patch with success +Having separate RM vs NM configs allows ops to use them as an enforcement tool if they desire a homogeneous YARN version across the cluster but that requires reconfigs of the RM and nodes before the rolling upgrade can start and reconfigs after the rolling upgrade is complete +Perhaps it could be fixed in the native code +Not really a solution but it let me compile itMy vote is to get this fixed as well +Suresh talked with IBM hadoop dev team internally we will push the patchset later after internal testing with other hadoop components successfully at first +This was fixed in the version of in +Please ignore and Im sorry for the trouble +So should we go with patch +This should hopefully correct that problem +Will commit the change later today +Ive just committed this +Will be fixed in the stage by COB today. +I was mistaken +The default should handle most normal use cases. +I think the existing syntax is already good +Ill send the fix soon +The patch uses a system property for MR application jar location +Now tiles definitions specified in old format can be read through the use of +I find easier to have all Help related changes in one Jira +The profiler is not donated to Apache +This is a conflict with jdk +Sorry for the extra overhead +Hi Stack are the patches ok for the commits? Then we may could integrate in our solutionproducts releasing as JIRA backport +Utility class created +the nightly build now completed having to wait hrs to see if it will get triggered by svn updates as have been being triggering automatically the last few days consider it fixed +I am going to close it however if it fails again can you provide more details on your environment +Peter has kindly agreed to do some preliminary investigation +Thanks Sylvain! I reformatted the comments so the lines didnt exceed characters and committed the patch to trunk with revision . +committed close for Solr +You can use Version to work with back compat no? +Whats the status +The test suite didnt fail bc of this. +Committed to trunk with revision +Adds a ctor w to and its extensionsFixed tests and codeFixed tagsAll tests looks good Shai! Only change I made was to have writer be final in +Patch applied +Verified thanks +Logged In YES useridPlease provide a testcase or detailed instructions toreproduce +Using maven +Thanks for the reviewIf other committers have followup happy to address later on branch +Closing all resolved tickets from or older +Lets schedule this for Camel if the best is to wait for AHC to be released first +Some work needs to be done to get the final solution by the people who know the code better than I +I am closing the issue +Need some sort of test case here to validate this +fixed +Not really related to but changes to the way layer IDs are generated will affect the Widget. +Ok +It doesnt seem like this should be too hard from either the sending or receiving side +Patch attached +In my usage Ill be calling into this code very often so every little bit countsAs far as the transcoding is concerned I agree with you +The fact that you see an intersection and I dont could very well be related to small improvements in the interpretation of dist distErr max that were done a couple months agoIm going to commit this patch tomorrow +So if you can post against that would be didnt we get you commit rightsmake you of arcsde in geotools? Did we never get you a password +Why Derby doesnt implements an offsetTheres a workaround solution to paginate in the database? I get the UNIQUE CONSTRAINT name by? Just like I get PRIMARY KEY name by and get FOREIGN KEY name by +For now use the above +miro I have committed a fix for this Can you try it out please +as we tried things mentioned also +Running it through Jenkins +Hi BJ I have removed plenty of them at r I had no time yet to think about your proposition but looks like a good idea to me so far. +This had been fixed in embedder +Since I am not hearing objections and cognizant that people are waiting on this going to commit +As Tom suggested Ill start a discussion on the developers forum. +Got green light from jms team. +Is there a problem in the test or just slipped over +Lets revisit ifwhen someone else bothers fixing it +JinchaoI find V has also a problemIm sorryCould you use vThanks +Set Fix version to +In order to stay compatible the distributed code has to support them too although our distributed matrix operations are always keyed by ints +Thanks Devaraj! +This may have unforeseen consequences with the java build so Ill hold off for now until theres more time available or if someone asks for it +Chandan since there are various fixes can you log separate defects or update the ones that are taken care of and resolve when all items are can you check off which ones are done and log separate defects for remaining items +Im not sure Ive tested all of the many scenarios of classloaderenvironmentsecuritymanager execution paths that are possible and Im hoping that with a couple more eyes on it we can ensure that we get the maximum useful information out of sysinfo with the fewest possible permissions granted to it +Btw this test showed that returned bytes without checking CRC for full bytes +fixed in trunk for a bit. +Made a few changes newer version +Closing all resolved tickets from or older +Previous file removed and +I have made a patch which addresses this in the way suggested above by Martin +Which is why proposing an configurable parameter is a nicer middle ground +Committed version b of the patch as svn resolving +I am unable to use the plugin with CDATA sections +Another fix has been checked in at revision +My Maven script is producing ejb jars for both and Weblogic and I want to distinguish between the two with a classifier but unfortunatley im also getting the problem mentioned above +Ok so I will keep the original definition but add the two bullet points +It refers back to the exampleAnd in the there is some stuff about +Or major compact first in it make sense to build on top of an existing utility like SqoopI dont know how Sqoop could be expanded to consider split points anditerators +The PEP process is based which more naturally keeps it from becoming a +It is still a test case issue since the retry count of the test in is set to +We see exactly the same error with in high load condition but only one timeWe suspect a problem with the SAN filesystem that cause a filestore corruptionAfter server restart all seems OKDo you have any test case to reproduce this error bug +Yeah you are right +How could we break surefire? How would we find what broke it? A bisect would take for ever +Ran using security profile on Mac but didnt reproduce the test failure +Thanks Simone! +This seemed basic enough to me to do eyeball testing and patch without a unit test +public static final String DFSNAMENODEDURESERVEDKEY reserved public static final long DFSNAMENODEDURESERVEDDEFAULT MB +Here is the patch for above described functionality +After I hoodwink someone into applying and validating Ill push it into trunk +The get has similar code where the result type of the project operator is set to instead of the type of the projection operator +Partial archetypes would be an option too +I am reading it +No objections recorded so wontfixing. +Fixed on waiting for Justins feedback before committing the forward to trunk on trunk as well +Just found this issue +Dont think the really needs to be setup +JulienThanks for fixing this clustering related bug +xsb classes +Mike +verified by Vladimir +Its been a while but as I recall the cleanupImpl is called from within the same old cleanupFailure that was already there still in place +likely +Putting this rough version here for until I get time to look again +Thanks +In my eyes that is a workaround for a missing feature more than a long term solution +Removing fix version of for those that are not in branch. +Looks goodI just committed this +To track the scripts testing Ive created +Thanks Bryant +I have commented those tests for now and Ill investigate why they fail +On sqlserver it might be a permission problem +Please review it +Closing. +LarsI need to rebase the patch once again related to test caseCurrently the performance numbers am not able to get it because our clusters are busy with other tasksIf it is ok i will rebase the patch now and upload it so that it can be committedIf not i will wait to get the results +this tiny patch solves the bug and includes a new test case for in cafd. +Thanks for this workaround it must have taken a lot of figuring out +Therefore we are not very keen to expose a lot more entry points +The setting of timezone in is not necessary as it seems +Thanks Marek for such a prompt response +Im leaving this for post freeze because its not really a new feature per se and has basically no change of screwing things up +Hmm +Done right +This is harmless +Justin +Users createedit function models and add them to VDB Editor like other models +Not seeing this any longer Farah please if you are still getting this +Audits motivation includes the observation of the administrators behavior +Thx for all the patches and inspiration. +To have a Compacted file around dont mean the sstable can be deleted and it will never be +The close method is called with immediate true which bypass sending the CHANNELCLOSE message +Maciej Alejandro I discovered an issue when looping tasks the executions are ended but the associated tasks remain active +The latest patch was against a branch so it didnt apply quite cleanlyI fixed this findbugs error already on the HA branch which we will merge soon +Separated changes into logical pieces to make verification easier +The existing test cases ran code may be overly complex as portions were designed to handle more generic the similarity between and they should probably inherit from a common base class +couldnt we avoid using both and just have one +Might take a couple of days to get around to it. +Test project which exposes the again you are using the dropdown choice incorrectlyyour model object is of type Long while your choice objects are +Also we have some problem during replication just when we try to replicate some deletion and it might prefectly explain why we get those issues +modified as suggested +Locus JBTIS TP TP update site JBTIS all updated +Aforementioned patch +Hive and Cassandra? Where exactly do the latest Hive handler sources live +Ias did some changes for this +Please could you check that it has been applied as expected +Igor is right +Unlinking pull request to address further enhancements as noted . +mikeyou can apply your patch and close this applied. +and if there other information to display that would be before the forms +If you are using a massive JS library like dojo that really sucks +Lets say you had the vlan already set +Ive just committed thisThanks a lot Uma +There is not such attribute for datasources +Yes Matt there are many other apis which are not checked for authorization in +This patch will fix this issueThis patch was tested manually by running distcp to upload data to sn and verifying that renames still worked as expected and that log messages confirmed creation and destruction of the background progress thread +That would also explain why Im not seeing any special logic in MRI +We shutdown the socket server and enqueue a shutdown request to the queue +pushed to nexus +thanks for the patch +If we get enough votes for this to be backported to we can do that at a later SP. +Thanks Harsh +Are these patches more complete? Would it help if I downloaded them applied them and then reuploaded a patch file +Committed to branch +We are now trying to figure out how to resolve this +I have not a clue about what has changed and what could have been causing the problem. +One last thing I forgotThese files are generated using the Thrift compiler +On Trunk you can use the as Lucene trunk is Java +Current state is considered satisfactory. +Wouter did you get a chance to create a sample application to reproduce this +for an extension point +Amit you cant make that assumption +this is a duplicate and the other issues are now marked as resolved +Please review +I assume you meant to assign this to me and not John +Any suggestions for workarounds until RC is released +Thanks for the review stack +What is your gut on the stability of the RC and when it will be moved to stable? I wont hold you to anything +Done we now support inbound resolved issues for already released versions +Hi DominikThank you for the fix +Marking closed. +I switched this out to use primitiveMkdir which is perhaps the better fix +See for more discussion +This is my DB and the problem is with the viales table the style used was the polygon default style of geoserver and the wms query was the layer preview of a layer created from that data please help is my DB and the problem is with the viales table the style used was the polygon default style of geoserver and the wms query was the layer preview of a layer created from that data please help closing all resolved issue that have not gotten any more feedback or comment in the last month +Should be smootherredgreen boxes should be check boxes many people have redgreen color blindnessrequired components a radio button with one choice is odd should be a checkboxoption components should also be a checkbox but if you leave as radio buttons dont install should be the initial valuelog window the close button is over the resize window iconWould be nice if the text was easier to read +ThanksPranav +Example UA that causes the NULL bug for usRCB Browser Profile Configuration +Assigning to Benn to help integrate the latest plugman implementation into +Version adds fix for Incidentally that IF statement will report table found even when that is not the case +This service actually works +This patch fixes the ec tests +Find attached the patch that corrects this minor just committed this +Completed. +Anyone hav a more elegant solution? Should we sort at allAlso the error code from the fetch response needs to move out into the send class +Yeah +IR spun with latest version of by using a new version of Seam FP. +Committed to trunk. +Makes senseThe number of slots per node is a at best and it just means in normal case the TT can run that many tasks of typical jobs concurrentlyHowever tasks of different jobs may need different resources +the CRC test failure is unrelated +ported fix back to all branches through +FWIW its fine to have support for Velocity or XYZ in core +I think only public fields should be disallowed see also +I will write the info on how to create a diff file to put on again! +Could you post them again +I have created to track the issue +Ok Ill ask in dev if its a problem for people +I dont understand what problem the suggested code solves +sld is also getting renamed when a style is renamed +Cummins Jr +Shoulda seen this here you go +Because RSZKREGIONSPLIT and RSZKREGIONSPLITTING will not be handled in processRITI think in that case we can just say break in default +And the extra bonus is to reduce the number of dependencies in Tapestry. +Bulk close of old resolved issues. +This patch is based on Todds patchIntroduced an environment variable to control the addition of debug option +So DDL is db specific and thus the timer service config needs to be updated for the db +Small X program using structs and the C code that we will generate to implement said program +in current svn +Although this is an improvement its an incompatible change so Ive marked it so and updated release notes +Or rather why cant it be null after the test for null +The default entries in the hit queue must be different instancesThe problem you are seing may be related to the fact that you are making your too big by passing a too large number of hits to be returned +However I think we need to its integration into the accumulo name +Please reopen if you experience this again. +This change would be backwards compatible so Im fine not marking it as a blocker +The same behavior was noticed during changing options on Visual Page Editor preference pagewhen you are on the Source tab and after changing preferences Visal Pane will appear +Is this issue waiting for anything further from me +simulates automatical address bar hiding on scroll events +Its output should help greatlyCiaoAny other thoughts on this topic? As Larry mentioned we do have a workaround in place but it creates a severe performance penalty +I am attaching the policy example for symmetric binding that works now with the fixesRegards should be fixed with the updates to to support issued tokensAlso fixed ThanksRuchith +Patch against trunk +committing to branch +Patch checked in. +Is there a clean way around it +We could also update the EMF such that it reads the root element registers the correct for the version and then proceeds to load the model +Well my local copy is quite different from b now +I will post a solution when its you please notify me as soon as a SNAPSHOT version with the change from your side will be availableThank you and stay in touch +Patch updates documentation to point to SVN +bq +I am going to close the jira +My previous comment was incorrect +Thats probably the reason things are faster with r profile + +Portability Will this work on Windows as set up now via? New external code +linked issue to make sure we dont forget to backport this too once we backport to +Neither does it timeout in seconds which would indicate a connect timeoutI have also tested the default configuration wherein the RPC should wait forever or at least till minutes when YARN kills the task attempt and tries again elsewhere +This is the guarantee that loss of a single rack will not result in data lossThe second condition that the third copy is placed on one of the previous racks is just an optimization an attempt to minimize network traffic across the racks during writes and there is no need to enforceit after the write is completeThus fsck should verify that each block is replicated on at least two racks if there is more than one rack in the systemDoes it work for balancing +Patch getting rid of extraneous M chars +I am fixing this better sooner than later +patch file with simple fix for this issue +bulk defer of bugsminor issues from to +Eli you forgot to revert this from +These should be scope test +Fixed and test case included +Also I have question Will this summation works on all the collection of objects or only objects which are displayed on the new decorator for automatic display of totals +Changed the title description to expand the concept to also cover +However for a topic publisher it needs to know which subscribers havent got the message +corrected as described +Hisorry for the delay in getting back to you +Location of schema file file is in schema directory of distribution +Ill take a look +DavidAm not sure if i applied this correctly +Struts does not and will not support HTML +A candidate for fixing in M I suggest +Committed this to trunk and +Agreed +java file in the project tree + +Whats your actual concern here +It was decided to turn it off for M +new files added in revision . +Does this need to be configurable? Wouldnt it be sufficient to catch the Exception and then fall back to the new approach +Attaching patch a which removes and adds to the list of supportedcompatible JRE versionsCommitted to trunk with revision +When I comment out the wagon extension the dashboard runs I the Wagon extension and try to run the dashboard I get this error +commit notifications configured + + +This bug is still present in the Xalan version included with Java SE update FYI The developers here have no bearing on what Oracle releases with regard to Xalan +will be in is not fixed in XDT release +asankhaWith the Dec patch I am not seeing any unit test failures +Also included in the patch is a minor cleanup related to +Patch for +Unfortunaly I leave Friday for vacation and I am no where near wrapped up so know I wont have time before thenI dont know if it helps but Ive added it to my calendar to do upon my return August st +Attaching a patch for +The release plugin depends on the interactive flag being set on the Settings object so I will see what I can do to mimic +I will keep a look out for further detailsBut sorry I cant give you a on the name +This issue related to +I think the RPM plugin covers almost anything you would want to do with and is configured entirely within the POM +I assume that is unintentionalIf you are really going to ignore my patch and not fix this until the end of the year +The second counter is handling the intermediate merge which creates its own Writer independent of the counter for records emitted from the +Thanks +Added support for pagination based upon using a +It could happen that the view id is updated before the node has added itself to the address cache +Fixed the one +You are correct +Thanks Amar +Other than that the patch looks great +Hi BenJustin patch has been already committed as forThis ticket might be complementary to that +Moving out of they lose them it could be very bad +Best way to make sure would be to test with the patchBut since you have ttl I would say that yes there is a good chance thats related +Sorry for the trouble. +Verified on Build +Patch committed thanks for your contribution +StefanoTwo different messages can have the same hashcode +you could add it back when this is ready. +Ill get it in shortly +See +Trunk is not released yet so there should be no rush getting things to work with existing releases +New patch including schema changes +I think you might need to write that book because the commit history is tough to follow +Thanks Namit! +Also I would probably name it to compiler or better only extensionLoader +i suppose that is what machines are for +For artifacts we publish this requires having a correct pom file which if we need anyway we might as well use for dependency resolution to to keep things simple +Is that a mistake +So it might be the combination that causes a problem +Perhaps as a ctor argument +skh we are taking inconsistent approaches to connector naming in dqp and the server + +Im looking at this pretty late so feel free to ignore them +Awaiting final verification that it does not break a user with a correct setup although this would be extremely unlikely. +this is just an XML file I fail to see that it should pose any risk to you environment though +Changed logging to FINER committed in r and r to +How are the local storage pools simulated? Looks like a storage pool already exists with the same UUID and thats why another cannot be created +Can you provide your thoughts on this + +Additional note to my set of patchesIm applying these patches over and patches applied locally +buildclassesUsersjoaDocumentsworkspacescompanyworkspacebin +Pretty much all new classes now have unit testsOleg +Fixed with revision +Reopening for handling +There may still be some issues with ancestral namespaces I need to look at +Small change +Thanks for your hint +The feature promises adding datanodes if a user configures it but datanodes cannot be added for some reason I think we should fail the operationThanks for the review +Attaching and with changes as followsA srcrefM srcrefM srcrefHope the terminology is not too far off +Should we close it +So the Gateway listeners take something from outside the ESB context and bring it in as an that can then be dealt with by components and listeners +Thanks for getting back to me +And we also have tests that verify that default values are used if attributes are not specified so such a change may break applications that rely on this functionalitySince the alternatives seem to be either disabling a security feature by default or breaking a different set of applications Im leaning towards option with a workaround in the affected test as implemented by the patch +fixed already +Committed at r +Checked again +I have run out of problem queries to test against this patch +I suppose its ok +I have attached screenshot of the original unintuitive dialog Makes sense if you know what to enter but completely baffling to users who have never seen it before +Bob its as if the svn server on beaver no longer has access to homeprojects +I guess we have to dig deeper into it +Completed At revision +Thanks Albert! +This is a just documenting issue +This issue is fixed and released as part of release +Merged the patch into MR in rev to merge r from trunk to fix unit test failures which is done in r on branch. +The suggestionBox you see on the top left corner should actually be right below the inputText for Postnummer +Test case demonstrating the Map field +Reopen per user ended up fixing it in +I am having some cases when it takes too much time to roll and I am having a soft real time environmentAbout the decimal + +Initial patch uploaded +The password being a property value Hudson treats it as any solution not to get the password appearing on the log files is to set it in the file +Since the patch does not touch any other component there was no need to run any other testsI just committed this +Pretty barebone model but Works with +Can you increase the Java Heap size when running the build Build works okey to meRegards tried building with this environment variable setset MVNOPTIONSXmxmBut I got out of memory exceptions instead +pigcontribzebrasrctesteestreaming +Before commit we can disable our component in the configuration file so you could safely inspect it and see if it suits your needs +Id rather fix this properly so you dont have to do the hack you have listed I provided the patch for this plugin to get product support happening I only needed one config to had not yet given enough thought to how to attach multiple the os part of the config tuple as the classifier would do the m not sure I have the time to create the correct you want to make the changes with appropriate tests it may get done faster +We are making explicit effort to make Apache the place for hadoop releases +Im taking over this issue to implement the repository backup and migration tool we need to have in Jackrabbit so we can better handle upgrades to the upcoming Jackrabbit release +Why the changes to +I suppose there should be a bound on how many times we try split logs or a particular server +This will take some time and discussion and doesnt need to be fixed for so Im deferring to or later. +Sorry the patch available status was my failure and I didnt know how to cancel this operationI will upload my example in the next days +James I did a commit that for me works +Not sure why though +Possible resolutions Reimplement in such a way that new API is not needed Create a default definition so that only a recompile is needed Mark this as that will be the recommended action if we mark this behavior as acceptableIt also seems that Decompressor is an interface preventing option of providing a default definition +I am just studying them + Test Case is fixed +We should do the same for all the String references these are usually much more expensive to convert +We could farm out conversion to the tserver work queue +No nothing in the its the + to trunk +Hi StefaxI just attempted an implementation as discussed above but it doesnt give the desired effect especially when using margins +So there is really no need to worry doing a change because it could break something +Verified as fixed. +You two guys are the best great work! I had no time to completely review it but I am happy that the new API seems so successful +When a method is public it also has to be maintained if you write a component that uses some private methods or members thats your prerogative but there is no guarantee from Apache that its existence and behavior will remain in the futureIn this issue Im stating that exposing Properties in the is hindering future evolution because they will have to be maintained that the same information should be available from what I consider a narrower scope location that does not reduce the functional possibilities +Does the ADG need changes as well +You may want to avoid svn checkouts and updates until Im done. +They are attached as a base for a Droids implementation +Does it make sense to commit this and close the issue +Created a model extension definition that encompasses the extension properties +Problem was indeed solved +Patch file is it to Blocker as the associated DEV JIRA issue has been changed to Blocker by customer +This is a specific war file that ships in SOA for this purpose? filename? the esb archive in this context is theThis is only an issue if the esb archive has failed to deploy cleanly +No When I log for the first time I am performing authentication and when I am inside the application and being idle for seconds then I am getting this issue and I am redirected to the login page +Can anyone comment on which approach would be more welcome XML change or a simple jobconf entry that does it +question answered by Dave C +Snuck in while I was testingSuresh et al +As I said the subtree user class has neither refinement nor ldap filter component +Yah but give it a whirl first +The sstablejson attempt is throwing the followingThat looks a lot like sstablejson couldnt find the Cassandra schemabq +if tab renaming was supported i could use short descriptive tab names like the following correlated subQ uncorrelated subQ and restricted join to quickly and easily tell the difference without having to actually view the tab contentsfurthermore the fact that every query execution creates a new tab might be part of the problem +I noticed that the workaround involved changing the concurrency level to a much smaller value +The current implementation includes the code Sean suggested to delegate handling of null values and seems to have no ill effects so Im declaring this one as fixed. +Embedded fonts not for the report but this is a duplicate of works with as of yesterday. +I agree with the proposed approach but it is an incompatible behavior change +I didnt think it is critical to fix this issue inHowever please let me know if anyone wants me to commit it to. +The jobs fail consistently with reducer NONE +Assigning to Jeremy who has volunteered to review patches contributed by Jeff +So from my perspective the software works as designed an inline with what Adobe Reader doesBR AndreasMaruan Thanks for your feedback +It does not compile +My previous patch left out and +afaik you can define security stuff in those files is that really intended just for earor was diff naming vs +I would much rather prefer Cookie constructors to throw but we cannot change API at this pointIll rework the patch in a minute +This looks reasonable to me +There is one important consideration which isnt so much philosophical as practical traditionally we test for whether a job is still running by looking for a null end time so it really has to be set as late as possible to prevent anyone from drawing a conclusion from the status field in a job that appears not to be running +Now that it actually honors the ignore contract all tests pass +Wed like to make sure that this fix gets into the releaseThanks +Its been a couple of days now +Just a small remark note that kickstart also contains some drawing logic but is currently pretty simple because of the processes that is capable of generating +In Our Vpe Plugins displays only debug information when in we sets debug mode +Currenly with the latest version a crashed database will itself +Patch applied +The JPA example seems to be working ok now +Any update on this +This issue doesnt occur any more +Revision the query is included in the exception message +As we alway say in opensource Itll be released when it will be releasedHonnesty youll have to wait as there is no major changes yet and few bugs fixed maybe a release in few month. +I believe this issue is resolve now +Rebased matches import code style +See. +When it stopped working with I tried adding an explicit dependency but this made no difference +This pull request fixes item but not so the bug should remain open +This dependency seems to have been introduced in the unsupportedarcsdedatastore module +delete a replica with the least free space +These are parameters +As a resolution note I validated the current reference documentation and the required explanation s there. +Why not to enable Seam facet if facets are enabled +Its definitely a little unintuitive that a memtable could be flushed before the write to the commitlog but its harmless as far as the semantics are concerned +Ive attached code for a hive UDF that implements the substringindex function described as well as some unit testsPlease review +I dont know the reason needs to know the current Thrift port if the assumption is that all the nodes are running on the same Thrift port then it will break when one node doesnt +Count me in as someone who got bitten +Attached is the most simple EBWS that I could create and then ll try to JBESBGACP revision to change Fixed field +Just tried it out and now it works. +The problem is most likely on machines with enough other work running such that thebackground thread does not get a chance to run +Attached fix for the other error +If we go using system cf for this we need to do the same and doing so requires flushes during one compaction one for inserting log and one for removing at the endI think plain file under cf data dir is easy to implement both for writing log and recovery +for me means that it returns an instance of determined by a int property see code +Have you tried recent +What was the partitioner? Were there secondary indexes involved +Hi JarekI think you are using release +It seems at one time the header blocks were deleted in the destructor but for some reason was moved out to the Stub destructor which is strange +I attached a new patch using wrapDefaults and wrapAppended and I took care to retain the import statements according to LuceneSolrs strict standards I think its still an if we want to more aggressively take the approach and create some methods that return a and +What i want is a simple ant target testjar so that projects depending on will simply use this +rev the WAR ZIP now gets installed to the maven repo at build time +You can copy that as is to your repoThere is an abstraction of a which I believe you also regression first appeared in still fails with the same cause +Please never run TP with get latest that makes no sense besides for debuggingchecking out new things IMO +I tried restarting server still happening +Moving out of +patch applied +Reported to me by Christopher Eckhardt +It appears as if with ibm we just get one more exception in the stackIf that is all then maybe the string more +For example my application knwo when to clear caches and when just remove one entity from the cacheSo please make an option to disable clearing all caches +The cause of race conditions are fixed in patch +Yes the separate seem necessary to meIve added a link to issue about the incompatibilityIn general its easier to review patches that do small incremental changes +Hence i still dont know the underlying exception for my case +Why it is minor? This is breaking change for +Xavier could you verify and close this one? Thanks +Please extend for your integration tests +OK I think I have an idea but Sid I would like your opinion on this +Beside the already provided patch from David Fischer thx for it I propose to remove additional the entry for in infrastructurednszonesThanks in advance for taking care of this request +Heres a patch that removes the unused constructors +Updated patch according to genesis resource best practice move resources that need to be filtered to keep resources that do not need to be filtered to resources +When I tried html format for output it did +Otherwise potential for the timer task thread to be starved and not be able to execute timer tasks that have fired in a timely manner. +project filein source repository instead use me to auto make these files on import +This makes disk shutdown fasterI added unit tests for the composite cache manager shutdown that verify the behavior +Yes go ahead revertreopen Ill check on mapreduce +So hold you horses a bit on this one +Put an instance inside s methods +I applied your change to and verified that it fixes the problem for me thanks + +Attached the changes based on the looks good +Wolfgang are you sure this is not a duplicate of +all issues to the CR + Fixed maintenance of cardinality values +For example precision steps are represented as xaa +Hi JoachimMy name is Antonia Horincar and I am currently studying Computer Science at the University of Sheffield +thanks +Doing more testing today and it appears that there are writes occurring but it looks like a definite reduction +Secondly Id like the uploaded diff to be compared to the branch +We got the March th and all is well. +Therefore calling close from onMessage ought to mean that the call blocks forever waiting for onMessage to complete +The failed test was reported at +As far as I can see that option only chances the delegation order for loading the classes but still uses the isolated classloader +This is not planned to be changed for but I will look into this definitely after +fo which shows the regressions +If I see it again I will post a screen shot but as I do not use Windows much I am not sure I will see this +Sorry my mistake surefire does not fail however does +So all issues not directly assigned to an AS release are being closed +The failed tests are not relatedI have committed this. +Make sense +Youre right I hadnt noticed that it was marked as a Feature Request +It is really annoying +The global listener list is synchronised on the list itself both for changes and for the list run through itself +Closing the issue as it is fixed in latest builds +But its a bit unweildy right now as you have to know about bit manipulation stuff to specify the correct levels +Let me work on it and you will get even Javadocs! Ccan i ask a small additional favor and have you name the implementations and? I like pretty names when possible and Impl is always extraneous +But we also have and Are those supposed to be part of solution as well as Distributed? Do we have test projects for those two +All of for EJB endpoints has WSDL importinclude handling has changed +bulk close of resolved issues. + assonance up weirdo r a chicken +Can you give this a try +It is inconsistent behavior for cpp project +Is that acceptableTry it? What you have currently is going to stay v fixes the findbugs warning +was closed before jira lynching +cacheRequest should include expiration +Improved patch which avoids copies when passed that are +what fileConstant lookup errors are sometimes tricky to investigate but usually easy to fix +dims +Pushed to with JBDS. +This issue is due to changes introduced in where the default protocol was switched from remote to remotingReverting that change would not be appropriate as it is needed to be able to enable GSSAPI supportThe change for does allow the protocol name to be specified in configuration however doing that would require the same config on the server and in all clients it would also only be a short term fix until we enable support for GSSAP +Edit release note for publication. +I addressed the nits on commit. +Tomasz What portal are you running +This issue was closed by mistake +Could someone with appropriate karma please close this out? I dont seem to have that karma +Right issue this timeMostly functional patch +Maybe related to +Need more information if we are to investigate this any further broker logs client logs information on your config etc +tests included +No new unit test cases have been added since this is only a packaging change +Committed with minor changes such as removed unnecessary return and ifelse statements changed output format to return statements without any header +Hi Kohsukewith this patch you have an elegant solution for this problem but it is unfortunately not the way we got in +I know that you added the license tags to the Java files already so I suppose you have no issue with it but I do need that comment in this issue will ArjenMr +Thanks a lot for this enhancementBest regardsRichard +The Tomcat guys changed the naming convention between and which broke the connector metrics +What we faced was that the minutes task was too slow to clean the memory under heavy load and the VM was reporting us many +I also removed the package and moved its contents into +Can restart the master which is less than ideal but doubt Ill get to this issue soon enough to do better fix. +Unless there are objections Ill commit this soon +In my bookonly after it is voted at least a beta would a link on thevalidatots web site for download be createdI know it is a little confusing +bq +Last time I booted up AS it seemed to pick up the host name and node name properly +jar file? Or howAt any rate I have some changes that Id like to have you test but Im not sure how to get you the changes +fixed +Verified on revision +Koushik also related to efcacaeeedaccfff youve made +We take the finally block and replicate it on all normal exits from the try block and at the start of all catch blocks +I agree this fix is not that urgent and doesnt need to be included in +Apparently a lot of people depend on this behavior +Use as key or a and provide the if you want the keys to equate when their content is the same +Committed at subversion revision +After I change SchoolID to SchoolId this exception does go away +I will add simons suggestion trying to also think of the scariest shortest verbage we can use +I cannot reproduce it anymore +change the just like mingjian said +Ive applied your patch and added a unit test which now passes successfullyThanks a lot for your contribution! +I should be able to encrypt all passwordspassphrases in the. +I assumed it was since the trinidad tag document said it would cycle through the access keys +Can you please provide a test class which is independent of your appplication code +closing issues for released version +Looks good +The class file must be found by the classloader from the class path +Ive attached the diffs from version that will allow you to set the Restlet server parameters from the Endpoint +A fortiori it returns in case header value is null +Thanks Dag +Verified by Aleksei. +Attaching another patch +to in commit dbabaeeeedfaef + can also do ps ux instead of aux just to list our processes +bq +For what its worth Im seeing the same issue doing a simple +Thank you +learn the packunpack message functions from Mikes transportsrcmaincpp +move it to CXF DanAre you able to characterize the work that remains before this issue can be closedThanksGreat news +schema of the latest code I found many final classes those classes will be unique in JVM and in case of Application servers or Eclipse could produce the interference because they have the separation of the application based on classloader +fixed in abecdfdadcbabceb +A patch file eclipse classpath is included +Will your proposed behavior makes sense +Im sure that the sufferers on the users list will be happy to see us get somewhere with this stuff +Kicking Jenkins again now that the trunk build should be fixed again. +Not sure if it is possible to add new menu items to the Browser context not then probably it is better to leave everything as is because standard items like Copy Link are quite useful +toFor we have a conservative configuration which is correct though slow +I would like similar functionality which works with a Map as opposed to a +Thanks Konstantin and Tom for the pointers +At revision +Reporters of these issues please verify the fixes and help close these found this NPE in last automation runs +Reopening then closing to deal with migration bug + patch look just committed this. +log does it works on +i guess it is not really now becomes but now can also be? because else we break much +It is possible to handle strings and gstrings which contain meta characters andor +ForkingThis would be a pretty major change +In other words what is the problem you face SD JPA could help you withimprove +I would hate to ask the cluster admin to run multiple commands to generate the info I needRegarding the locking question you are right that we might be a getting a consistent picture of all the data structures that are being dumped +You may have broken the build +Nate please take a look this looks like an Native limitation +equals is not needed it could have been just equals +We didnt see any need for the return from a check op so we ignore it completely +I thought we had it in the SOA branch but it is not file must be from the SOA build. +Im going to try out your patch on a test cluster that loads +Cos you need to move the record about this jira in on trunk under section +this can be bumped if there isnt some icons to svn to use for. +Would be nice to get the other clean up in but wont let it hold up an we move this jira toOne of the subtasks is done the other three moved to +I also implemented a fall back to the default configuration when the provided configuration does not contain a configuration for the app namei am not convinced that this is the right thing to doa regular setup should go along with a proper configuration and fallback to someimplementation specific default should only occur if there is no configurationpresent at allin addition i didnt experience any failing tests during integrationTestingso i would like to know what configuration was present with the query teststhat caused the failure +Thats a bit confusing for end users to look at the code +Number formatting shouldnt happen at this level as Claus pointed out it screws up a bunch of other clients of the code +I will try to have a look before the end of this week but Im really short on time +looks fine thank you Nathan +Ive arranged this now through catching all Exceptions during preparation of the handler method and redirecting it to the handleException that makes more sense now +The fix addresses the indexing issue in and enhances to preventminimize when this situation can occur +Id like time out for too it is all we needTo implement application specified strategies of pooling there will be the JMX supportWith the JMX we could monitor and manage the pool state for strategies +Currently I think its working well enough but I would like to explore Rolands comment in the future so will file a new bug to track that +It also prevents from being added as dependencies even if they are not marked optional +Network failures are kind of handled in my proposal where we STOPCONT trackers at will +seems like the previous one was corrupted +Integrated fix to still does not work with collapsible as of version should be. +It would be good to have. +This is to support the volume migration feature on KVM +Attached is first version of patch +We can them when weve got a solution +Do not commit +Once you start a server in the servers view and it completes startup it should immediately become expandable in the +please ignore testing how attachments seems that some databases have tighter restrictions on where this escape can go hold off on applying this patch until I can resolve that +This is a sample maven project +I just thought Id let you know the idea before these names are set in stoneMatt +Test case am also attaching the output of the test case. +you get all the necessary information regarding this issue +its a only issue +Martin has anticipated what I wanted to write +Please confirm that it works for you +Ie hdfshostxlink filefile is ok even though Hdfs doesnt support wo authorities foohostlink hostfile is only ok if file system foo supports with just an authority I should also mention that even in the link file case we know the file system of the link since uses the default file system to make the paths it creates fully qualified ie the default file system is never used when resolving a link target but is still used when creating paths +This is fixed in trunk rev +I stand by the claim that multiple server principal HA is insane +Im on the my mail server does the same +Please reviewThank you +Maybe it is just my lack of understanding why the assert is needed +the problem seems to be unqualified columns on queries with the joined subclass strategy may it manifest on the WHERE clause or the ORDER BY clause +I did some further investigation and it seems that both errors were caused by tmp being symlinked to privatetmp on +Probably because it is unusual from the of the current java spec but thats just my opinionSure all new things feel unusual in the beginning +No response assuming ok. +But the cluster will take advantage of only half of the resources when executing sequential jobsIs it possible to have a configuration schema that allows to use all resources for sequential jobs but not more than available resources when parallel job executionsDoes it make sense to have a that limits the maximun total number of tasks per node but forcing to be smaller than for skip the possible deadlockThanks for your amazing OS project +As part of the processor thread we continuously invoke process on each iteration +You have a valid case +I share your frustration with these brittle noisy tests +Were working to speed up Hive with vectorized query execution and are thinking ahead about reducing process overhead by processes and introducing +Both BSP master and slaves classes can be implements by two interfaces extended by +furthermore the test case does not include required libs such as domj etc +Also affects per Grzegorz Kossakowski on maven +I ran the example several times and it seems to clean up +Assigning to Henry Z +Attached example demonstrates the problem it creates threads which sleep for a millisecond and then exit +I will revert r and backport r to where it makes sense since r was done in July Thanks for your appreciated help +This is more involved than I expected +Use jdk +Committed revision fixes the issue the various cleanup changes for helped determine the issue. +Corrected doxia skin +The patch allows for POM files to yield Eclipse project files in case the Maven project has no child projects itself +I am guessing it has something to do with the temp file +As an aside streaming from multiple sources and discarding data in based on streaming data you have and comparing overlaps in mechanism with what you might want to do for repair in order to avoid the size explosion problem completely +Sounds good to me +I still have one nocommit since I want to add more tests for the persisted global map +This issue has been hanging around for a long time now +but punting until we can get resolution on how to replace the symlink since that will not work on Windows +I dont know this part or state machine +I hope someone would care about this since this problem have been announced for a year +Thanks for the patch but you must not change the default behavior to limit the max messages per poll to +The patch should merely add ! imo +If we modify +bq +Is there some way I can get git to generate something acceptableI havent written full yet Id be happy to write them if people are ok with the code changes in generalFinally there was a problem with unrelated classes and which were not declared in their own files which seemed to break trunk +Keith +Now the upgrade script moves configs from service level to the cluster +thanks for the explanation +Please close it. +My mistake +Thanks Claudio I gonna take a look +thanks for looking at this +Thank you for your reactionI understand that you will not change this forMaybe you could apply it for the next release +To add to this I have noticed an issue with the wheel zoom that when you zoom in or out the image shrinksenlarges for visual feedback but sometimes it remains at that new size and you see it refresh at that to use Richards technical explanation sometimes it refreshes to the proper size but the image is all screwed up In both cases a refresh of the map brings things back to is a map that has WFS and WMS layers +Stack changes looks good +Thats easier than modifying JDBM to enable cache wo writing to disk as well +this issue is duplicated with +The comp bean is an internal bean. +The screenshot without localization may be the key for the issue +branchSending configsSending configsclientSending configsSending configsSending configsSending configsSending configsSending configsSending configsdirectorySending configsSending configsSending configsSending configssrcplanSending configsSending configsSending configsjettySending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configstomcatSending configsSending configsSending configsSending configsSending configsSending etcDeleting modulesservicemixDeleting modulesSending Transmitting file dataCommitted revision Not closing yet want to check I got it all +Hi MikeWould it be possible for you to take a look at this for me pleaseTomPS It could be the case that this is something we dont need to support but can you update the Jira with details why +For connections returning after password reset are still not closed immediately +Any objections to merging this work to Jackrabbit trunk? The branch looks pretty good to me now and I think any remaining issues are best solved in trunk where its easier for more people to try out and look at the code +Edits going to META are more problematic you could see old regions coming back and completely mess up your data +revision Refactored classpath properties for clarity and to avoid duplication +Sorry this is cuase by forgeting the property +Thanks Dmytro +If there is any issues we can reopen this JIRA +FIXED revision +No nothing +Attachment has been added with description Changes to to enable String anybody out thereNi guys Can somebody create the patch check it and close this bug? I includethe method to do it so it cant be so dificult +I think Shalins asleep Ill get this committed in a bit suppers almost ready +The hardware clocks themselves do not go back but if two updates come with seqnum X and seqnum X then X might get a smaller ns because the hardware thread might observe a different clock than the other thread +We specifically steer clear of defining our own Castor is used inside of other containers that do define grabbing the were simply trying to play nice with the hierarchy +Ive fixed an issue in passivation handling that I believe will reduce maybe eliminate the problem +Cause of infinity loop in broadcasting OK +However it is harmless +That look great! thanks a lotI will try this as soon as I can. +What do you mean by does not load? When running mvn jettyrun in web module or youre deploying in a standalone jetty? mvn jettyrun works fine here +eh you show zero hibernate related code or test case in here +this works only with cfml sessions JEE Sessions are managed by the servlet engine railo has no control over it +I have attached the signed grant form to this issue +Im unsure how this changes things unless its some fu to get around antjunit spawn semantics +Actually I now see the ability to set the full jarfile in the macro +Any progress +I see that locking is done both with the lock and in one place the synchronize keyword +Maybe you can testwith your file where the umlauts were brokenThanks +ThanksGood suggestions +Committed at subversion revision Currently the release manager constructs the download page by hand +committed at r create page with workspaceId instead of ws name properly look up namespace info to populate datastore namespace parameter and add unit tests +pls +You are asking for an improvement +Doesnt look like hadoop QA is coming back + for the patch +There is no change with from profiler +tomcat +Anders will you make publicIm greatly interested in it too +I have confirmed all of the remaining jars in Stargate lib on branch are required +I do not think that this is related to the problem accessing on jdk. +We have the following setup Using we dont cache the service locator object +I will +Fixed since Richard applied the closing all of these resolved issues. +Beans are looked up through the bean resolver registered on the evaluation context +Well my first comment works for the file but doesnt provide any help for the filesTake a look at a similar problem +I will need a bit more information than that +Anybody there? please help me to resolve this issue +as a cheatsheet and Committed to the trunk +fixed. + +RajeshMind running your patch through test suite and post the result? +Added comments for classes methods +Can the status be changed +If a user is not authenticated she will get error +Also added new methods to JMS message service test case +So it can be accessed by multiple threads and each thread should get the model when the method returns +John could you please attach in question +cool +Did you test your second patch ? It looks good for me +Reopening to edit release notes text for release notes. +Huh thats am obvious Java faultIt says one thing in the Registry and something different is installedI suppose we should not trust the key in the Registry after allHave you checked with SunOracle if there is some ticket opened for thatI would hate to fix something that might be already characterized as a bug in JRE installationIll leave the case open for now till we figure out how to handle this + +The target should depend on the new one as cpa wont compile without the generated classes later also have to add cpatargetjavacc in eclipse build path for eclipse to be able to compile cpa care that the new target can be executed when using build scripts and from within the bin additional note the ant build scripts use blank instead of tabs by intention +In that case this jira is a and can be closed +That gave me a little understanding about this attribute so all I did was to make sure that if for some reason such an merged artifact was also selected in a normal way the merged artifact wasnt returned from and so wont get retrieved by the +for the patch +patch applied see revision thanks! +Im no maven expert but doesnt this patch just replace with Any way we can have both installed +The simple A B table names were conflicting with the same names from previous tests so making the tables unqiue resolved the problem +After applying the patch to the trunk you can run bundle exec rake in the librb folder and it should create the gem for strlcpy from into its own file and patch to place strlcpy in its own file and added define of HAVESTRLCPY when using from here +Any suggestionsI dont see this as a big problem +A flag for extracting the value from models is now available in +This patch improves the repository upload instructions and adds a link to it in the main. +Here is a new patch file that should allow the manglers to run +I added parameter to that method +Anyway at this stage having to interleave two frameworks in the code should be fine +Can this be closed? +Kill eclipse +that class is used as scope but i am not sure its classloader is used. +Fixes findbug warnings +Then I copied the app to Simulator again and fired it up +After manually adding +Great thanks for putting it together Stephen! And youre correct about the deployment it should be out of the scope of this JIRA +The new attachment contains a junit testcase +So without the txmethod name it was rolled back if a write error happened within the first chunk +Native libs +Making it binsh forces installation node to have bin etc +Test comment +RTC work item move of all unresolved issues from to +Agreed +patch Uses JAVA in instead of relying on PATHMakes sure that JAVA is appropriately set from bincassandra and debianinitChanges debian initscript so that etcdefaultcassandra is sourced before etccassandra +I just tried maven on the command line and it also works +Done the structure is now available +A fix for this ticket was rolled into. +They encourage people looking at the code to email the individuals directly instead of contacting the mailing list +Thanks for testing that +But the proxied ones are not since then +Worst case can be bad +The report has merit dont like to much to see the code duplication either +Please discuss in forum referenced above +Sorry Max will try to look at this and other issues soon +cygdrivecjavarubylibrubyin mkdircygdrivecjavarubylibrubyin mkdirpcygdrivecjavarubylibrubyin reverseeachcygdrivecjavarubylibrubyin mkdirpcygdrivecjavarubylibrubyin eachcygdrivecjavarubylibrubyin mkdirpcygdrivecjavaserverdamagecontrolcorein newprojectcygdrivecjavaserverdamagecontrolwebin storeconfigurationwhen I try to create a new ve not searched to see if this is a known issue or not +The page is intentionally low on info for security concernsfix in d +Ok but as I said above you cant guarantee that an LDAP directory is configured to be consistent with the approach +missed your first comment yes we have a gradle plugin that builds the poms when we cut a release. +It would not be required by the AM +I foresee problems with running compactions might have to get a compaction lock on all the regions before starting all of the above +Both and these to generate a simple so the same +Sanchit this is much better now but still has a conflict in appviews +Im not quite clear what the issue is +Actually ignore my last comment as it is not valid +DE applied in trunkNow we still need the patch for the IT and zhCN languages +bq +Configuring ehcache from logs +Seems like we can attack it in two parts +My bad +Note that the exception appears in the log only application reacts as it is expected redirects to login screen +Expanding everything could be problematic for large deployments +Committed revision . +tests from patch added to the new toolsTest project +Test failure is unrelated to this patch +Ive just spent the last hour trying to reproduce in bare bones project with no success +Um you have heard of method overriding right +I think ability to annotate methods with a list of filters names which should be applier to response can be useful extension to this change request +Picture of multiple repos for one artifact looks like your outbound connection keeps getting a different IP +It seems to me that the classloading configuration is broken in Jboss but again I could be wrong I tried to insert some with classloading instruction to load first the one inside grails but that didnt I found workarounds replace the JBOSSHOMEcommonlib and its dependencies comment out the bean inside JBOSSHOMEserverdefaultdeployers use in production the second one because we had another issue with the quartz plugin which conflicted with the one bundled in is also present in had to use the fix from Graeme Rocher +Create a patch to add the static string + +We do not use the resolved status. +headerClass header header contentClass attributes dont work in VPE +I have committed this patch at subversion revision +Thanks for the review Chris! Attach new version of the patch +N configurations into the equation here? They shouldnt play a role in the job settings +unixdos converting file +Without threads blockings I could save minutes in a process that takes minutesIs there any chance the correction will be integrated? +I just did the following Ganymede meclipse testng jboss tools all using update sites +A landscape example with page rotation ! Patch applied in revision +Unfortunately this bug does not appear to be fixed by the above changes +r +Ive committed the patch. +It is not a regression and nobody seems to be working on it so it wont make it into + +I will test it with the sort doesnt work in the case of of reduces the jsp throws a divide by exceptionOther than that it works great +Sishen I added you as a contributor and assigned you this issue +Users set up ideal state as described in previous comment along with the state model +Particularly the pole bugs +Can I request this to be to +IIRC Dave Grove was interested in this option too +We have a base war file that we override selected elements from in dependent projects +Fixing failed test cases +The original problem was with which triggered a but there was no value change +Please Verify and Close +Such administration interface is normally part of a CMS or a web application and not the web framewok itselfIt requires significant amount of resources to implement an administration interface like Django did +Thanks UmaThere are thousands of them +This patch makes GS switch to the new classes +To be last week with the refactoring of AVA and +Please review the code and let me know if it should be done differently +Once I have that Ill apply and check it out locally then it should be pretty quick to get inThanks for all your work on this ticket +Ill prepare the patch for this +On mysql the failing ones are TC TC makes me wonder is that the failures occure at different tests +This patch serializes into the regionserver ephemeral node instead of passing the load over the heartbeat +Since Xerces has been namespaced since if I make sure to put it first in my +Since we are throwing the exception from init context loading is failing +Have you tried rvmrootbranches +Resolves the properties file location issueThere are also internal improvements to the is currently still Java but your patch depends on Java +New enum value called IMPORTINVALID for corner cases +Please confirm that it works now +Sorry the stack copy doesnt show it but It is null pointer exception as the title says +Im using XML schemas only and no DTD so the only option is what you are suggesting +Please check. +bz streams +David Any thoughts on how we can handle thisthanks current spec does not define any copying of the wsaFrom on to response messagesThe described scenarios is really mixing layers in an uncomfortable way it would seem that a better way of dealing with this is either to put the replyto address in the EPR or to carry the reply to address in a transport header. +Also please add an overloaded method which takes no filterstring most cases users wont be interested in a filter string so it makes sense to have a method that caters for this case +I guess I was confused by why Lucene was picked in the first place as I thought we dont test writing indexes with old Codecs +I think it would be of more usefulness if the object values were not but bean names of +youre right +One other effect of this is that the startup activators do not need to add these global listeners which also helps our other goal of moving logic out of our activators and into more reasonable locations +I will put that to as Scala must be GA +I dont see them contraversial too lets keep them here in MR instead of creating tiny tickets in each projectOverall the patch looks good to me +JeroenThe objective of getting XML file support into was to move away from using +Patches were too broad and have gone stale +Rebased and merged into the master branch. +But is valid +on changing HSL to do float +I think some users would ask what the relationship between and is +Patch contains the same +Note that if you attempt to Finish the web service again it finishes successfully +Hrm OK I figured it out +log output and other details for a getFeature request with a spatial filter. +When a region server is choosing peers it prints out a message like this Considering rs with ratio +solved this by allowing you to pass a list a of to set. +Manik not sure you understood what I meant If you use a repl queue youll use a separate thread to send the replication but the replication message underneath will still be sync due to the replay transformation which happens further down the code correct? The only difference is that instead of blocking the user thread and getting the exception to bubble up to the client thread youd see the exception in the repl queue thread +Jesse No problem +I have not had time to understand the relationship of the security data model to the SSO data model +ivy was released on Sept and should fix this issue +as for the job directory changes this is the directory structure in HDFS +The further error about exiting ij is dumped in the main function where ij is caughtThere might be few more things The attached patch does not exit if there is a failure in connection + +Thanks for squashing those Snjezana +Moved to reference to pull request with changed patch +It consists of two servlets that provide the RSS and Atom feeds and a little utility class for extracting the project information +done +Going back to +In your sources do you have a comisbrigelgaiajmxThanksFreddy +Here are my observations +Is that correct? It seems to me that assuming your fix went into right away we still have and will always have an issue with the metadata statements getting updated properly when you go back to a version prior to the fix forThis would mean that none of the existing statements could have new stored procedure calls added +Missed the input file again +I now see why the gadget content is not getting rendered on Android browsers +Reopened created resource and took a look atI changed port configuration from my console to so that I could monitor network traffic +Closed. +I cant edit the original report text +This was an invalid issue. +after release project page will show all fixed problems human mistake safe code sometimes feature request for version can be fixed without correct fix version sometimes tester do not set correct version sometimes programmer do not ser correct fix versionTrue this is strange and maybe should be reported as additional part of JIRA report named like tasks incorrectly defined in JIRA +It also needs to not be case sensitive in comparing table return and key +conflicts this needs to be regenerated. +Maven Integration for Seam Projects also specific for JBT +ahhh +Can others w access to a cluster try to get the other fix on +I merged from trunk into ran the libhdfs test and committed +In addition to the changes made to reduce the times we also made a few minor changes to reduce the time required to register the node types upon the initialization of the repository +Pavel thanks a lot for your noteIMO it is reasonable +Justin +Bugs in the current RPC SASL negotiation are blocking this jira +Ah stupid me +If a user is only interested at some of them he can do so in his own implementation of these listeners +new tests were added to the test framework to test choice construct as follows to test string type choice elements to test Simple type choice elements to test simple type unbounded choice elementsAll these test are document style tests +Sorted lists of WAL entry reach destination RS for final merge or replay +As you probably know you may need to have your build force a new fetch of the snapshot build since youve already done it once today +When m is specified it should probably just call copy in a serial fashion across the list and not run a job +Hive QA to trunk +As this is a corner case that should not happen so often it is postponed for the moment +Pushing this further down the lane +Do we want to spend the time on to fix this? We can leave that to Carlos teamJims comment EAP is on the way to end of life not really but its getting old I would not wait weeks for this but I would hopefully be able to put it the startup script a test that fails fast +Added headers and sent a pull request +Updated +Does not even compile +the first bit is by design but the FNFE is a fixed +Patch applied in r +There are many application nodes in the version cluster and many in the version cluster +I guess it is pretty straight forward and does not bring any side affects +Generate the code coverage report for the mycomponent module from the in mycomponenttarget +AFAIK maven itself orders its dependencies from top to bottom in the dependency list and for transitive deps the shorter path wins +Please close. +NP I had a boring plane ride home and got most of the structure in place +Javadocs are greatly missing +I check whyAdrian +Cant reproduce it either on Windows +The code changes look good but I will wait from a confirmation byMac before checking it in +The exception handler in only to catches Exceptions that are cause by my actions +I think it makes sense to add this method to Database interface and your patch seams to be reasonable but I would prefer to see a minimal test case +I have added a releases section to Validator that should helpUnfortunately when I tried to rebuild the site with MavenMaven created a series of recursive directories that exceededthe maximum directory depth which then could not be deletedSo well have to wait for someone else to rebuild or untilI have time to look into it further. +This deserves more attention +IT shows this working on trunk leaving issue open in case we want to backport to +It didnt generate fields for a half a day +Dominic you wouldnt want to zip up the entire project since thatll include log files compiled classes and other cruft +Thanks! Ill get it done first thing in the morning and Ill issue the rc +Fixed in SVN +to let evaluateSample return a boolean +Reporters of these issues please verify the fixes and help close these test passed +It cannot find the foreign reader index +Can you check that the selenium integration test in the monitoring works in your environment? Can you come up with a selenium test for the problem you are seeingI created the selenium test by recording my actions setting up a view and then putting it into a java test case +In this version of the path all old functionality of the build works as beforeHerriot artifacts arent produced yet but this seems to be pretty minor fix +attach a runnable test case +Morevoer why these nbsp around the uilabel +Just a quick addendum the problem persists with the snapshot as well +User error by the sounds +I think should be ok in terms of space +Or at least Hadoop +I think the changes are what I want them to be so I will leave review to post commit and change whatevers needed +Lucene is tested with true +changes replacement logic and removes a spacial resolution case +only run when it is needed or at least only enter maintenance mode when really updating something +Thanks +Is this still a thing? I note that the window for changing in a minor release has passed +Please let me know if the name is acceptableWhat do I need to do for forrest docs? I can see the settings documented in the builddocsI would prefer to keep the tests that individually enable the limits in addition to the test that enables both +Datanodes that join the system after the upgrade is finished are handled using above +I would like to get a new bug and thread dump from you for a application +Change required for only two files and +If the multiplication produces a number max filesize for a region well take maxfilesizeIf region for a given table on a regionserver well split on the first flushIf regions from same table on a regionserver well split at M and so onWe could have the size grow more aggressively by squaring the count of regions that might make sense if cluster has lots of small tables in fact it might be better altogether +Not marking resolved until we have an official JNA release incorporating the change +Not sure why no tests are failing +Is that correct? For empty table minmax doesnt sound right +Thanks Henri! +Is this intended just trying to understandTalking to Elliott his plan was to extract all the actions and then change this test to make alters happening while destroying the cluster +Committed revision . +Its a pleasure working with you +What if we keep it in our repo and at the appropriate time spin it off into a sub project or another incubator projectI just think for getting started and building momentum it would pay to keep it close +should be resolved using assignedContaienrs as a repository for containers which at some point have run a task but havent been released yet +This seems like an obscure error to me +One more questionbq +The shell build runs fine no more +Thank you for your analysis and explanation +It seemsthat only textxml and textcsv content WFS GML and WFS CSVoutput is getting gzipped only when the settingtrueIf you go to our geoserver now and do a WMS whatcomes back to browser is single gzipped content which seems OK as a workaround for now but I need to check everythingelse and make sure there are no side effectsStill not sure if this double gzip is due to a change with our Apachesetting or whether its some change with Geoserver from toI will check Monday with our Sys +Attaching v Cleaned up the join between primary and secondary indexes now accomplished by the iterator Turned on lightweight compression by default since it gave faster numbers on my CPU bound machine Fixed lots of bugs related to larger than memory indexes Added to skip short byte arrays rather than allocating any benchmarks with a baseline run for comparison +However this can be a different JIRA +The UI code which does the update throws an exception +I dont have the time currently maybe Suraj has something to come up with it +Released as part of. +Unless the ESB wizard is stupid then it should simply be using the defaults as specified in the IDE are you saying it it is enabling project specific settings and forces Java compiler compliance even though your IDE is set to Java ? Or is the global default also Java for you +Id also propose that we keep PROXY as a separate value and the service needs to check the real users to determine the underlying basis for the authenticationI think it is also important to avoid abbreviations in the interfacesHere is a quick modification of your patch that changes the api as suggested +Wiki or other external documentation +Ive attached a patch and HTML file for review for this issue for an expanded file for the bin distribution that includes references to the manuals documentation javadoc and other resources +Ill need to reproduce your error in order to fix it +Patch looks good to me too +bq +Is this sufficient for your needs +created its own S browser +Attempts to set security mode to Derby SQL Standard Authorization mode without first requiring authentication will raise a warning +Hadoop was released +The same goes for exportingThe easiest rule of thumb is this if you cant do it in the process diagram then you cant do it in the BPMN either even if Activiti engine supports itIf you find a construct you can edit in the process diagram but doesnt get synced to the BPMN then thats a bug +At this point Im really suspicious that of this one is different between the two what else could there be? Do I have from now on explicitly test append on my clusters because the code path and applicable policies are different between the two? I would certainly hope not +Patch +Patch applied +switchType server will be fixed with for release +Committed to trunk! +I guess this means we should then return false +I will double check the submitted patch to avoid this problemNice Catch Jonathan Hsieh +Ive submitted issue with a patch that improves the accessibility of this class and others in the package +Ive tried run portlet in pluto and +This is great news! Thank you know when it will be available through maven +No worries +Wrapping it in an option implies were being good Scala citizens and avoiding null which we arent +Jun is this still relevant +Is this in trunk as wellYes in a sense +Seems like a timing related issue +Thanks Kev +the tar file with the sources +Lets target end of day wednesday to complete testing +and for looking for other similarly misleading messages +Im gonna update in this way +I had to rollback the fix which looks like bad change +igor good to know what defaults are like onI renamed to and dropped the other one +Rebased on +Got the same problem +Thanks Kan for the trunk patch. +Getters for the main are added with rev aefPlease attach a patch with any further improvements you may see. +Seems not completely fixed yet. +I give it thumbs up +No response from reporter. +Since this is a fix Im marking this open to get it out of our long patch build queue on Hudson +Correct I never got this work with the bridge I have a lot of JSF portlets working with the and page navigation has no problemOf course I am here to help +Ill commit this +Moving to +Hi Spark Patch applied at r +Ah +Please also note that Hbm application is not Java and Hbm application is on Java +It seems like an undesirable dependency to have a txt file bundled in the that has to be read at runtime somehow +Also versions of the foolowing servers was changed according to Supported Java VersionIBM BEA +This is addressedfixed by the rpm needs to be validated +reopened and moved as a feature request of the resources plugin +This patch breaks several tests +In there is the notion of partialResults +You said By default workspace resolution should be enabledIs this implemented this way in the current SVN trunkIf it is then there is a bug because for me this option is disabled by default +Properties in the capacity scheduler conf should be defined in and should be read from the capacity scheduler conf object and not from JTs configuration object +If this test fails even if there is nothing wrong we should probably just disable them for now until we can make sure that the test starts after the agent is completely up +I added the part where attempt id is passed as part of the child task environment and circulated back for taskkill +The patch uses a couple of oracle mechanisms to determine the minmax from the stats rather than issuing an aggregate query +ModifiedIt seems that the modified test doesnt fail with version! +Sure thing +This issue is a duplicate of specific +Closed as of could just drop the test for the port number as we are testing other config items as well that should pass +Next question How do I set the name conflict reporting strategy? Is it in the file or a argument to +I have overlooked this setting in that test +It seems that the problem is only performing the action on a remote node +Stack does not want to do contrib for any more +see thread entitled hbase master is not starting on new ubutnu system on userhbaseYou can leave a comment or TODO stating the need to revisit when support is added +IIRC this is a new API method so we can change it at this point +I also spiced up a bit so we can trash individual or all connections to test +Committed to trunk only since its an optimization +Reopening issue +the patch looks good +I applied the patch and also bumped the ibrowse version number in the build system. +For backward compatability could we enforce this during current dfs dfs upgradehmm +Verified in JBT Betae +lgtm should be safe and respect the dfs settings +Lets work on the miscellaneous dirs in another issue +Committed patch for +Can we close this +that would each handle instances of their particular concepts actors operators etc +Well if you define the spec merely by what passes the TCK then yes its spec compliant +This patch still requires itests + is root srcjavaorgapacheluceneanalysis is root srctestorgapacheluceneanalysisIm sorry about the patch +but I dont want to complicate things unnecessarily +Yes this is used in IBM +Sorry about at issue is gone when using RF +Each site has a daily volume of documents and threads per process +files within the dist hook when removing the +Attaching the patch to fix this ambiguity issue +If you build from the root or from it works fine +I am willing to offer a hand in fixing this. +Resolving this please reopen if there are still issues with Ubuntu and their FW defaults. +New unit tests added for the above +Do you want me to remove this from the patch or can you simply disregard the fragment that applies to the main method +Thanks for the patch Michael +I will have a lookat the details of throwing an unimplemented exception in SQL Auth modeI agree that it is desirable to get the current code into the trunk so that the communitycan start gaining experience with it +One minor nit can we write a unit test for this so that this does not break again +attach a patch to enable delivery throttle which is required for multiplexingthe fix is quite straightforward which checks the gap between last and last whether it is over the throttle value +I suspect the cause of the NPE to be that the root was null +This patch takes into consideration Konstantins comments +Turns out this is only a problem for where were operating on granularity was correctly retrying until it got a consistent set of references +bulk move from M to edit pu should be Browse button with dialog which look like Select java project dialog in case if there are pu to suggestif there is no pu to suggest just message box with message There are no Persistence units in current JPA configuration +Moving items not being worked on for M afaik out of I did this ages ago. +We just had a fix as proposed above and it works fine +upated to trunkmoved files to respective src folderfixed default resource loading as mentioned in above commentmade minor changes in failmon and is a concern +files? deferralOr failing that if its really important that we have packd files we can publish to like we do for everything else in Tools land +Baruch can you give an example of code that would use the Autowired annotation? Im not sure how that would resolve the issue so Im hoping see the code youre thinking of will help +patch looks good +Yeah I did already and everything looked ok +Thank you for this suggestionit should make it easier to study this problem +To the extent that those promotions is a function of fragmentation the pause times might be variable over time as a function of old space +Fix made on branch revision . +Jar with Source +this method will ensure security and protect data +bq +I ran ant Dtestcase true DargsXX and with the Oracle JVM I get Compressed Oops true but with IBM JVM I still get false +John means after attached the source to the of ESB classpath contaienr Eclipse editor still can not open the source the classes in the jar but if attach the source to the jars in AS classpath container it works its little weird +ChangesParameter names now paramNames and paramValuesNow can use multiple values to check againstChanged to commons logging +Daytrader Revision fix pingservletsessionjdbc failureRevision change to eager to fix pingservletsessioncmronemany failure Daytrader Revision fix ping primitives failures Daytrader trunkRevision fix ping primitives failures +But I do +Is there any reason not to make this a blocker for or even? This is really causing us out here a lot of pain and really needs to get fixed +Was resolved several versions ago. +Or just make it package private? This flag is only used by +Although actually Im not even sure if that rewrite is really a problem I almost dont think it will tickle the same issue as the rewrite that was happening before the search +Hi Joe! Thanks much for the Compatibility option to enable the code in +thanks a lot +If there are two producers of the same type with the same qualifier then you will get an +trying to remember +Is that a bug or do I need to run some other command +The log from my previous message is reproducible on debug builds only +This patch is not a fix for this problem but attempts to capture more detailed log message +Todd not having sendfile will be held against netty since the bulk of the data transported by uses it +Unfortunately there never seems to be enough spare time. +If there are concurrent clients accessing the server they will be just as exposed to the original problem +This patch +Seems to be fixed +Forgive my ignorance but where exactly do I go to download these releases? Thanks +I will be fixing that as part of the same ticket +Now a Hadoop job which uses Jakes trick of caching the row widths +Attached is the patch that changes the drivers and build script to enable parallel execution of ee tests +But you do need a way to push out configuration changes to all the task trackers which classic configurations dont have +As you said we dont always need to sortA comparison between the two complexities should be done at the beginning to decide which way to go +Can you reference the ticket number here for that change please? +Inline comments for jcc usingdbclient removal have been updated as per review +Are there any objections to my releasing the contents of? To summarize the patch will have Applications iTests run against Equinox for the Aries build in order to avoid a recurrence of the NPECME +but illustrates my point that this stuff is complicated and I think we need to take the safe option here and back it out +This issue showed when testing the implementation for in Rev +The idea of config sets for multcore had its origins in Right it would probably to make the most sense to name configsets as configs to match the layout in ZK +since this is done outside HDFS it is simpler for users to start experimentingSay a file has blocks with replication of total blocksWith this tool replication could be reduced to with one block for parity total blocksThis is a savings of space +Please confirm the current status of this issue and if its incorrect change it as required +So my action form was never called +To be able to remove B and C from As view we need to install a new view in C and that requires FLUSH wed have to analyze what happens in this case. +So I am going to drop the priority to critical for now +Unified patch against srcbugs with working binding file included +See Zhengs earlier comments for details +This is similar to the initial version you had given +Added logic to handle siblings +Our usecase is to intercept the method call add transactions to the call perform throttling tracing everything configurable for each endpointoperation +the problem here is that an artifact with type isnt automatically being assigned the classifier client +Thanks for this Martin! Ive committed your fix in revision Peter would you mind trying out the latest code from trunk +In addition Im a little bit afraid of changing the pom structure for a micro release branch tbh +Thanks for the review +Currently the recentTasks is a set of task ids that are running +You are awesome +I understand from the comments that this patch is not ready for commit because the functionality is still changing so Im switching off patch available +Test exception? Or is it just ignoring the attribute override +I am getting the same thing on Solaris once I change my password to be encrypted +Fixed in CVS. +What about if you make usage optionalplugable +fixed incorrect copying of special properties during checkinCommitted revision +JoshThis is a corner case with very low reproducibility + for the patch +This patch also adds a flush if stalled control that hijacks indexing threads if the DW is stalled and there are still pending flushes +Applied the patch with a test case +The screen is showing the location where the probem shows up and the error in the Problems view +Verified by Mikhail. +Will keep watching to see if there is some way I can help out. +Also please add a test +hibernatetools is checked +WHERE deletedfalse +I think it is best to include the JAR into the endorsed directory + +I have tested to verify that this worksNote To make use of it you will need the parameter set in your The way this key works is by specifying which keys are to be inherited and uses as a wildcard to specify that all keys are to be inherited +More tea leaves and entrails to sift through though +new controller automatic name +Patch file that includes fix for the issue +IIRC this doesnt show up when running locally but does come up when running on a hadoop cluster +He replied a few minutes later from a cab and his colleague asking him to do this so if stars align you may see zkclient out and you can remove it from KafkaAnd then I really really hope you can publish Kafka to a Maven repoQuestionscomments How come this is not Fix Version? Should it not be? Is it not planned for the upcoming? Nobody is assigned +I agree with the refactoring mentioned by StackBut I would approach a JIRA at the difficult part first namely the filter +It looks like a hostname DNS problem to me +When a datanode starts do we expect any thing to be in the cache at all? Also seconds cache report means quite a lot of messages from datanode to namenode +Thanks for all the hard work Eugene +So the solution will be a bit more complicated but we are on the right way +Why cant it be represented as a +If you have an IMAP mail server and set your s to false it will keep mail on server but mark them as SEEN so they dont get downloaded againWhat is a UDIL +In just added one info log from the previous patch +svnwc svnwc B Jan +Is this how you obtained your new copy or did you use the standard download +As to if we should do upsert for Puts +I dont know of a way of changing the initLimit dynamically though +The patch looks good to me +please discuss things in the userforum first +Im not actively working on this so if someone else needs it done feel free to take it +the problems found by Alexei Varlamov was fixed +The benefit of is the ability to have a global to share among all the instances correct? +Thanks for patch v +It would also be nice if the extraction would support something similar to Xpdfs pdftotext layout flag +I have created a ticket which has a fix for this +In the current code the namenode sends out only block deletion requests per heartbeat +I will also update it on review board +DSL rules had the same problem +Patch made compatible with and added some more integration tests for the changed functionalityThis patch also makes the writing and filenaming of output files from consistent the given output path is taken as a directory and a new file is created in the directory +Thanks! +What does your Gemfile look like +Those are written into kvbuffer bypassing kvmeta +Im attaching and with that changeThanks for reviewing this +Looks like yes that did fix it. +I am still learning Pig internalThanks +We could potential develop a DB engine on top of Cassandra to enable real time joining on multiple tables or other comprehensive queries +If intentional np Ill back your test out of my patch +Yes we have to create a new every time since some of the old values need to be cleared +That wasnt it +I thought the incubator site was managed like the other TLP sites and could only be modified by the PMC +In case it is not ready yet update the Jira with justification comment and preferably with date when it will be resolved +Can you please comment on this +Fixed javadoced and logged the return values +You may want to comment why in windows the output can not be assertAddedbq +Also there are several places where a simple synchronized block on around a contructor can cause this issueintegrationEAPsrcmainorgjbossjmsserverendpoint xsrcmainorgjbossjmsserverendpoint xsrcmainorgjbossjmsserverendpointmainorgjbossjmsserverconnectionfactorymainorgjbossjmsserverconnectionmanager xsrcmainorgjbossjmsclientcontainer +AchimNo disrespect intended +version since now were only working on dependency on to which is required to have all to be then able to process a URI as argument +Which part do you needsince this bug crashes the whole container I think its a blockerMichele +Solution for remove and replace the headers +Please create new ticket if there are bugs or improvements to be made +code committed thanks Ishan. +Thanks Corinne! +Please take a look at and comment your viewWe may need to revert the change for this ticket. +Do you plan on implementing it in the future +Not test case provided which would be needed to reproduce the issue in question +Patch enables VEH processing on windowsemtGeneral functions were moved into separate file in utilwiniaemt directory! Patch includes changes suggested by Ivan +Duplicate of. +Setting this to Server component +Outputs junit xml the patch doesnt convert the testng output files if the build fails back to the drawing board +Thanks Ning +Ive just committed this to trunk andThanks a lot for the contribution Colin. +Attached are files to reproduce the problem +Committed trunk revision Merged revision closing for +What else do I need to do to get wiki write access +rebased patch as looks good to me +Yes thats fine. +hmmm +Fixed in rev in branchesVerified that the fix in branches branches. +However I dont feel that this is a priority at this time +We could do a better job with the tosinspect result for arrays like elementData though. +will have the same +change listApplications api to take in a list of application types +This patches throws a which is better than a Stackoverflow +I have a sld with discreteIs there some way to remove the Red X that appears for No Data valueThanks +Patch with only the bug fix +I dont really like having different behavior on Windows and posix platforms but if the platform itself behaves in a specific way I doubt there is anything we can doIn case of the dual checkpoints we might be ok because we actually dont open the files +Sounds good to me +Isolating now +It would not be the first time during my work on plugins that a naughty third party plugin got in the way +for the patch +The trunk failures are fixed as part of the project split +I assume you did +Canceling patch pending review updates +This will allow us to have a simple Kernel bootstrapper +Seems ok now resolving +Ill figure out the view support now +Chris please upload a patch or let me know what I am missing in the merge if you want it to be merged to +Hi HenningIm NOT using +This pull request addresses and by wrapping the WISE project jars and using their SOAP message generation approach +I can help out by answering any questions that might arise +Ok + +Thanks Rohit for your time and kindly replyThe fix you mentioned will bypass the issue but some others occuredI will submit a new bug if possible +ophostplannerreservation table is meant for internal house keeping and it doesnt hold the information on who the resource is dedicated too +Its more general really +I went ahead to commit Shawns patch to trunk at revision and it did not bring any new failed junit caseThanks Shawn. +Thanks Pat + this patch address all the issues especially the thread safe ones? Im not the expert with OJB and Im not sure if the returns thread safe instances at allApart from that the proposed changes look good +Patch includes several critical changes New security scheme ledger and mac keys Optimized reads client sends request to just a quorum Fixes the problem with reads that triggered this jira +Closing this record since is now released. +Done +Mice can you review this issue +This is a big disadvantage if you would like to refactor and restructure your java client code +The bugfix that was added depends on which itself depended on the original +Are we able to release without this fix I guess so but wanted to check +Mike I merged the branch with our current trunk and fixed all code to compile and run with Hope that helps +Woud you mind attaching a diff with the original versions as its easier to understand the changes youve doneYou could as well issue a pull request through github if you prefer +was thrown while evaluating an expressionERROR S Column not found +And if you only could remember where +Reopened to make sure all information was updated correctly +Just wondering has there been any update on this +This is a complete implementation for Java and updates the spec documentNote that a new unreleased version of qdox is required to fix a bug +It should be applied from the geronimoapplications dir +This patch should solve it +bq +If it dont work on single retry then it deserves to fail Id say +No magical assumptionsbtw +MarshallPlease help to apply the patchIf the work in is done please back to me for testing +Yeah Im sure it doesnt +Previous patch didnt actually work +This felt easier than porting heapsizemetrics over to Todds proposed patchI got tired of one race condition fix adding in another race condition +yes I cant reproduce it in Fedora +thanks Sijie +However I dont know how to apply the patch +Thanks +This can happen for a few reasons some of the more common are +Maybe we could give column families the same treatment +You will see the response delivered in a single packet like the attached +These are some tests which fail from this defect +The problem is they were designed for just one time import +possible fix as per attached filed and testcase Yes the snapshot works for for the fix +the error cannot be HA! Perseverance pays off +Hopefully that patch will do the trick +I used the gmail address youve given in your JIRA profileAlso im using or + Implemented Socket Data Loader REST API and Hadoop Activity widget +I was hoping you could spare the CPU cycles and solve your specific communication problem this way rather than inflicting extra CPU cycles on everyone elseWould be interesting to know exactly what removing final is going to cost +Failsafe has been added to the version but it is not available in version which is currently in use +The problem also occurs with a CXF snapshot built from sources checked out this morning +The runs without the patch were and seconds +part of release. +new more easy to understand examples +I dont need the connections remote shutdown +please check you tested that +All changes needed to change from targettype to type +Committed revision +This is another symptom of you running instead of +I can understand the desire for short cuts but in this case it really is meaningless and completely outside the spec +Moving to off topic but I think logging which remote URL Archiva is about to request would be immensely helpful +It looks like the version clashes are properly detected but the database is updated nonetheless +Thanks for fixing it +Its from groovy beta +Thanks +I confirm the issue +In that case if control is attempted an error Control not available for JBNEM will be creating a new resource type for the server resource this is the best we can do +Ill look deeper into it however I want to release Roo before the end of the year +Verified that this issue has been resolved correctly +an ephemeral port is now used in tests. +This patch contain the implementation for the new feature and some other modification to C source code generation +Marking as invalid since new Cocoon features solve the problemFeel free to reopen if Cocoon still has this problem. +patch looks good to me +bq +merged and pushed upstream branch and master +After h I have also the OOM errorI changed my settings to use svn in seems to be fixed with my latest changes but need more tests on machines where it occured +the best solution is the db driver can support pull clob data with cursor but it cannt maybe it will in the future +We should clean that up for sureRegarding the build hangtimeout that could be because the requires correct loop backlocalhost configuration setup +The seek to row isnt implemented and it is still a standalone project but it will let people start looking at it + Mirror maker shares the producer data channel and producer thread from migration tool +I will check it out and report back +Gokhan +Closing as valid work around is to use Windows or Linux +By Looking at the code looks like if there is race condition between and in event dispatcher This condition may arrise when Nodemanager tries to register itself and puts this node in the Nodes list and schedule the event for recoonect however in the mean time there is an unhealthy event come first to RM and it deletes this Node from the Nodes mapThanks patchThanks patch does not need rebasingThanks good to me +The actual problem is to decide whithin a DB context how can I tell to my adapter to do an Update instead of an I have to finish it I hope Ill be able to make it generic enough when everything will be finished Ill send you the you very to unless contribution makes it possible to release to esperio db output adapter +I think we should open new bugs about specific bad names +Bruce +fixed in r +We use the same property in than inThis needs to be set in not since we dont want to tie the user with Nutch +I already made the same changes two weeks ago in Revision per Keiths comment this has already been fixed. +Windows doesnt have function which appeared only in XP +thanks Uwe! will remove and commit shortly +Thanks Andrey +I forgot to update my tree. +Now I create a new snapshot edition of Mod make some changes test them and finally want to release Mod I should be safe to call the release on P rather than on Mod +Updated metadata accordingly in revision . +Sample project +This is actually the reasoning behind my initial solution pooling contexts marshallers and unmarshallers which in hindsight is somewhat overkill for common usecases not everbody needs a highly scalable solution catering for high load as we are currently dealing with +Whats the status of this? I could take a look if its still open +Looks like the interface was moved to collections but remains in math +Solved in jbossweb CP +is the only thing you need to verified +I dropped them an email but I didnt get a clear answer I ended up hardcoding in +Is there a new version that solves this +pluginssrcgroovyorgcodehausgroovygrailspluginsspringsecurity +The model and UI changes for Camel SQL were incorporated into the changes done for +Works for me on Sabayon with Oracle JDK +This issue could be due to defect +We should move to a GA before AS goes out +I just ran my tests with in place and I no longer see +committed in trunkrunning tests for +patch applied with thanks to Art. +bq +applies only to the trunk applies only to the trunk +Given that its config change and not a code change I dont think retest of everything is necessary so Gary pls pass it on to customer +Same problem resolved in +So your google results are off topic +Because of random ranges on the whole range +One last patch fixdoc null null searcher issueremove a few unused +I have committed this patch at rev +Ok then I will work on do we use right now for code generation? wscompile? Some Axis tool +This patch protects against requesting maps +My first intuition was that we were eating an exception somewhere when we couldnt create a directory +Lets focus this jira on parallel hflush which is what hbase really needs +We can roll a release with the when it is fixed +I seems to me a user would simply select a printer from their browser +Your customised Wookie RTC workflow scheme would need to be changed to allow you to the issue and change its status andor then to not allow this to happen in the first place +If you have a patch for this issue please attach it and it. +See. +Attaching rebased vWould be nice to get some review on this before the freeze +But how it was used in the case I built it for was to show a two dimensional display of the dataIm not tied to any particular name +I tend to lean more toward it being a server side problem tooI can also confirm that the garbage collection trick suggested elsewhere works but it is by no means elegant and definitely affects the normally passive JVM garbage collection routine + +seen this plenty of times +Assigned to m on migration +Is that correct +This fix is about being able to supply an arbitary userID and password rather than accepting the OS userID credentials +Patch with trivial formatting doc changes committed in r +I am going to attach a test case that demonstrates the issue with Felix +For RPMdeb it will use etchadoop as default and create symlink for PREFIXetchadoop point to etchadoop +Can you attach one of the offending files +I will of course make sure to pick up the fixes from your patches as well +Thats weird Mike +Fixed in commit dfdffbadfdbbbffafe +Is this true are we abanoning the jbosgi umbrella projectI thought we keep offering a stand alone pure runtime but remove AS as a supported remote test target +Trivial patch +Thanks Hernn +Moved to SPR as this seems to be a OXM improvement +and are all part of your +Previous patch was do you think this makes sense +Closing until someone cares about +proposed can confirm this issue we are using heavily in unit tests and we get these warnings all over the place +Slava why this dialog doesnt apper +Tried it with the most recent build and it is still the same behaviour. +Accepted for is not the proper solution we have to distribute only one file with fixed issue and provide the upgrade script for customers to upgrade to latest dbOtherwise there is a risk that customer will create a new db using sql scripts provided that lacks bugfix and then switch hbm file when meet the bug which in turn means that db wont get updated and jBPM can go mad +I agree that the default should be moved to Futon code as I think should not be just Futons library +This needs checking and may be cause of this issue +I believe that this is now fixed in branch +Just need to make up some junit code to test it as well +Now that the various code bugs have been fixed it seems safe to proceed with this documentation improvement +Here is the patch for changing Create return to Create Return for maintaining the consistency +There are additional problems with the way the overlaying is done +For other containers like glassfish jboss and tomcat I get failures even with plain cargo trunk which I hadnt had the timemotivation to look further into +Thanks Chris that looks goodTrunkCommitted revision branch xCommitted revision +Svn moved +The deployment context creates a private +Same here a bunch of methods were removed from the Gandi drivers +upload a new patch +Yes my last successful commit was later than June +i want use the featureplease handle this bug. +The functionality is essentially a reworking of the lucene document reconstruction functionality in Luke but with improvements to handle the problems listed in the jira issue description aboveIll be pushing up a patch soon and will look forward to any additional recommendations after others have had a chance to try it out +Fixes tests +DEADLOCK +OK I think I can do something about this +Daryoush are we trying to fix this before v +If you have a failing test case that you can attach please dont hesitate to reopen this issue +Good scalability backup namenodes can be plugged into the network on demand +Committed small patch to and trunk +It will probably run even faster with instead of +Im not sure why that would be usefulA distributed secondary index would allow you to query one machine to figure out what other machines have columns matching a predicate as opposed to a local secondary index where you immediately query every machine to figure out whether a predicate matches +Whats the inconsistency in Jochens proposal? It is the operator for a latch +Ive committed a fix for this +Since the parent is still needed but the dependencies are now correctly handled by Stanbols parent POM this issue is resolved. +No other links to GEOS found +Works for inherited profiles in Maven only. +Please rerun upgrade from +Attaching patch made against trunk +Howards comment might be a useful improvement to the page should be dealt as a separate issue closing. +My solution for DB on OS Maximiliano key generation is very different in +note that if you use the provider eucalyptus you will need to set the property httpyourhostservicesEucalyptus when creating the Adrian for letting me know +Done with closing stale resolved issues +That mention of trial key is rudiment from the past actually never used +I could not create a test case for the same +bq +There are some such import in the current code base of but there are only here to support the context menu on the Ivy classpath container +I wonder if this has more to do with dynamically changing the switchType after having created some tabs +Fixed in the branch +From the test result log it looks like the testcase failedI manually ran this test case against the Pig trunk patch and it passed +I think the problem here is thread context switchingIn your thread stack you have threadsThis seems to be caused by two problems +Alexey please verify +DaveHave you had time to work on this at all +DONEShould not use DEBUG logging for Camel DONE +Actually I take it back +I made a temporary change to the to avoid this specific test if running on Mac + +Nahi I dont think its a problem either its actually a little surprising to me that MRI is able to raise at that point +Looks like we need the also committed since I moved some prints from the to here +Closing. +note same happens when using and you have a server starte on which is a common need to handle this somehow +The whole ideamechanism is built based on the paradigma that all the involved parties know NOTHNING about each They dont +hdfs pathCache hdfs cacheadmin rename all classes to fix exception translation issue with write permission denied +I am keeping this issue open because I think there is lot more that can be done to improve this screen + +Yep I confirm this +Correct the error that the old patch did not check if the lastupdatedblock is null +OkayYou are right that the documentation must be clear +It does not appear in all the back and forth over the past few days it appears to have been revertedI reapplied the patch. +There really needs to be an automated integration test for this +I will convert it into a component and if objects I will add it to extensions +Thanks Scott! +I wonder if we could get back to forking only once if we would then see some real benefit +Even though it is a new page sometimes a free page is really a reused page and we need some information from the old page to properly create the new page +A good example is that column in one row might map to a column named foo while column in another row might map to a column named bar +I will add another patch after resolving some issues that I am having now +release bulk close + +talks about older clients +All set +Resolving this issue I believe the changes in address the above and changes were made for this JIRA and included in which is now released os no furhter change should go against this JIRA either way. +Thanks Ive attached a revised patch with the fixes you suggested +Upon further consideration I think this would be a useful feature and went to try and find the discussions you mentioned in JIRA or the forums but I could not find anything with mappedBy and inheritI would appreciate any additional help on the search language that I could use to find themThanksMichael +Btw +Thanks for verifying Andrew +In some cases the task got timeout or lost +It should usd exists API from seam tools for that +bulk fixing the version info for and all affected issues have in s very likely this could have been +The problem seems to stem from having a dependency on junit in a subproject +OK +Agreed this needs to addressed +The new branch build needs to have a module alias similar to what has in to ensure that previous build instructions continue to work +Closed? Are you for real? NPE is a default way to report parsing errors in your coding style +Of course in the future we should support using multiple but it adds some complexity to the initial release +Let me know if there is some problem +My steps to repro show the bug +A pragmatic solution would be to use that is work in a sort of saturation logic +I think we are drifting away from the original goal to have a Quaternion classIf this class cannot implement Field because of commutativity problems then dont implement the interfaceFor now there is no use of Field +Changed test project added data grids which show the correct behavior but not the spark list +Ill vote for this +Thanks again +hiI am still facing this issue even if I am using +Done for RHEL +An alternate name formcommand? for looking into this so deeply +Whilst investigation this i noticed that pressing the main Refresh button in the toolbar causes the images within the tree to update to reflect open closed folders so I replaced the above call with an asynchronous call for the tree to refresh which results the the images updating as you use the tree and prevents the runtime exceptions. +Need to be by QA +nd try after an svn up +suggested was just released. +Has anyone actually looked at the code of this dependency to see how robust and reliable it will be? I would not like uDig to absorb core elements in its very JRE which have been written in haste and not yet matured +I can see why Werner would want to check to make sure that the object isnt expected to be handled in long transactions as well as why Stein thinks it shouldnt matter or at least shouldnt be enforced so thought is that we should we should have it both waysIf instead of throwing an exception about the object being timestample what if we were to always make sure the instead the object timestamp was always set to or maybe even to make sure that it would be an invalid timestamp when we have the none cache +However Ive run into this issue myself on a number of occasions + +That is right +These are only used in examples +Is that OK +Committed revision Thanks for the contribution Jason! +Roger same behavior on master +As written on the forum you should set the confidential transport guarantee to force https address +New patch which Leaves as Private Unstable +Found another thing Id liked tweaked we dont have bit CDH packages anywhere but so if were not explicitly supplying an image ID we should have the choose a bit image +Im not sure but I suspect a bug +I believe the Xerces implementation is conflicting with the built in Java parser +I think you can close this issue +Admittedly I should have run the test on Linux VM +If there arent very reasonable arguments to add this or you want to do that yourself I think we should close this issue as wont fix +good catch pat! Ill fix it +I will run regression testsTouches the following filesM javadrdaorgapachederbyimpldrdaThe method which wraps streams around LOB output parameters now detects when the underlying LOB is nullM javatestingorgapachederbyTestingfunctionTeststestsjdbcNew test case for this bug +trunk +ran more tests against a large body of real world emails with the patched and unpatched version of mimej and got consistent results + +Right all are fixed now in trunk +You could move this to JIRA if you want +right you have a point +I changed the superclass of the servlet and it looks like Sling is finding the servlet again without issue +Please reopen if this is still an issue +which hbase version are you targeting? +beta will bundle the library and add it to a projects classpath when it is done or +But I think it is OK to just leave it on +Closing all resolved tickets from or older +thanks +Im not sure that it can go into a release unless we can make it not break backwards compatibility +Earlier I uploaded we currently use in Ocj under solaris env +Let me try it on my home machine +Setting jvmContext earlier +Attached patch fix the issue +If the thread has fallen behind what confidence do we have that the client will be able to renew the lease any easier? Moreover this approach sounds like it would place more load on the namenode which it probably doesnt need if it is being slow to respond to lease requestsPerhaps we could instead change the client to wait for the thread to sucessfully renew things +Thanks Thomas! +I am trying to get httpclient debug logging work +doesnt document what happens when empty Seed dont document whether they take Ipv addresses I had intended the scope of this ticket to be formattingstyling +Do you know what happened? I am still trying to figure out what went wrong this time +Unfortunately Im right in the middle of a big refactor of the Axis binding and this isnt going to apply cleanly anymore +In that case I was expecting Removal of mixin no longer automatically removes invalid properties or similar as bug title +Spaces in DESC are allowed +Ill fix the give me some feedback on and + +Please review the markup in svn +the resulting code was committed in revision Im leaving this issue open for further improvements like the Numbers support you mentioned +IIRC these kind of queries are special cased because it basically means traversing the hierarchy +I forgot the critical last step of running svn update in wwwchanges should be live today all ResolvedFixed issues lacking a Fixed Version +Im usually a fan of output options having grep on the raw output as a fall back +should we split this jira to address just hte java shell? we can add a new jira for addressing the c shell +update the patch with Anotation +Where are the frequent updates +Do we need a db? Wouldnt a queue per destination work? Plenty of good queue implementations out there +Still would like to hear that its working for the original reporter though +Ajay I am curious if the database in Derby was created with territory based collation enabled? In other words when the database was created did you use a JDBC url property like collationTERRITORYBASED +For what I can see this method is only used from the test classesFurthermore as it is the status quo now this will work only with theThats why IMHO is not much useful to have it in the Oak still dont see why the method should be somewhere else +Fixed in revision +Removed following file from the patch +A project with one class +should work with the existing bridge +Good suggestion +Nice catch selckinI was finally able to repro the OOMEI think the attached patch should fix it +Earlier the refactored tests implemented it in a similar but slightly different way +You remember the status of this? If you say it is still present I will fix it right away +Logs of the applet when connecting to HTTPS site through NTLM proxyHost names and login have been replaced by XXXXX or similar for confidentiality reasons +None of the patches attached have been granted license to the ASF therefore we cannot use them +Too bad we cannot put post install hook into gems itself for in commit +Did this job run for a long time before unregistering or not? How long between unregister and final app exit? did the app exitI have seen a case where the job spent a long time running +Returning as invalid per the issue meeting on May . + is not active maintained anymore IMHO we should not depend on such projects in components out of the box at ApacheWe have previously debated this and decided to not do a component at Camel +Try replacing the jar in JRUBYHOMElibrubygemsgemslib clear the tmpwar directory of your app and +Commented this API method +Could you please provide your patch? That would be great +The click does cause a new request however and the folder is then shown as open +Ive set it to as this blocks which is marked could you take a look +Im suggesting that lockTimer itself is a separate thread +Thanks +Viraj can you please confirm if you can reproduce it or some variant of it +This way if there are not very many events we do more writes and the events are output quickly but if we start to fall behind in the writes then we start batching them up into bigger chunks which are more efficient +Ill set one up if someone doesnt get to it first +I grabbed the latest source and built the assemblies +I marked it as critical since it caused a blocker regression in but I guess since it turns out a workaround is available from the implementing class you can lower the priority +We love contributions +lock at usage site is brittle andAgree +Small update to the headings +we have tested it on plain railo version without problems testcode in additional info here? Otherwise we will close this ticket within the next days. +Need add changes in the model for working with is the status here? have we removed the feature +The doc should explain the message and the workaround for the error +Ive attached a JAR file that you can plop into your eclipse installation that should fix this JAR file is +Then I can listen to the and then ask the editor about the caret position +The JBAS error only comes if I deploy using the +Thanks Hong +In other words how many people will be surprised by this change and how many environments will it turn a working instance into a broken one +Binglin +Ok added mcr as a committer and removed +This certainly looks like a deadlock and afaik there was some recent work done in this area which may have introduced thisDo you happen to have a reproducible testcase? or even a smallish private testcase you could send to Richard andor me +Second version of the patch +The fact that your program works with cdoes not gives us any clue about whats going on +I now get the same problem Stefan describes where srcmainjava is missing from the URL generated for each change +Candidate patch for trunk +Much of the work Ive done on this project has been around this object and I know it pretty well now and when I originally proposed the enum I had a feeling that it really should have been a fix for the but took the quick path +It was only used within hdfs code for optimization which we now know causes data loss +GreatI updated the to be of version to avoid future confusion +m directories along with using a fresh checkout of the I am unable to reproduce itCan you confirm if this problem is reproducible and if so why is it that in your case the internal resolver chain is not going beyond the fs config +ok its done have fun +Im going to only apply the changes to trunk to avoid the cleanup work +so Ill hold off applying this particular patch until the one youre working on is ready +Im attaching my current patch so that anyone else can take a look +Looks like wasnt updated to include this JIRA +Hi JacquesFollowing are my comments a +I still think it would be useful for local particularly when combined with the libvirt integration youve been working on +Yes earlier I have opened this issue under Mapreduce specific and now I Changed to commonVinay +This patch remove all never used imports and is patch for SUID warnings +attach new patchadd testing closewrite on read only on test +It is not a conflict with other Sonar plugins +It would also be good to add first class Solrj support for this Ill break that into a new issue though +Christian suggestion seems to be in the right direction but I need to do more research +The stencil mask is missing completely +Patch based on latest the same patch to trigger changes look fine to me too +The same is true for +To sum up check whats going on a little better on my end I cannot see anything wrong Ill close the bug as cannot reproduce unless you provide more evidence of the avoid logging the stack trace to make sure people dont quickly assume its an error +If there are no further objections Ill commit this once Colin posts an updated patch +After looking at evaluation I think the bug should be moved to class library. +The user is meant to select one + +Id doc it as subject to change and unsupported myself +Thanks +We would like to use the to support this +Thanks I think I got both of these +Both the colours and order issues have been taken care of +Thank you +Thanks Jakob +though could be problematic due to parallel fetchesRight +Sounds like a bug in svngit conversion toolAnyway it is not a blocker +for the new patch +Yep lets get things working well in trunk first +This ensures that there is a change in the znode version +I dont think resolutions involving unions are handled well +The issue is if the server response returns a dynamic YUI element like a dialog the init script never gets run and the dialog is never created so we end up with an dialog on the page +This seems to be a useful feature so it should be true by default no +The exception handling seems to be backwards compatibleBy the sound of it you changed the runtime behavior in an incompat way with a single thread you have very tight control and knowledge of what docs got accepted in and what docs failed and the exception for every fail +This is different from a class that knows that it is or might be configured to be at the modeling level +Good point +Filtering enhancement applies to key as well +If anything then I would say classes behave my prototype I was able to carry class definitions and variables between if this isnt really possible with Groovy then it may mean a no go for the language on our project +But I agree it is not internal anymore +Hey Sebastian Hudson Thank you so much for on pushing things that hard +applied some formatting needed fixing +Here is the fix +good idea whould be to add a request without reply counter for detect dead connection +Then this might work in +Does tis plugin work with nutch +Sure +Please review the updates in this repo gitbaileyjeI have added support for filtering imports in dependencies +First step write a test + revision on trunk and on branchxThanks Yury +Here is a set of two utilities to validate the frameworks sorta Checks the records in both the input and the output to the sort bm and ensures they are consistentb Checks to ensure that the records in each split are sorted correctly +I realized that the issue is almost fixed by s +Should contain CXF version +I have done just that in the attached patch +As you note existing implementations would not be able to read this until theyve implemented this codec while with multiple files they would +Fixed now could u please reopen if you still see an issue +Fixes the following Remove the serial number column from being imported to CSV Display the time in days and hours for expected and actual duration instead of showing total minutes Adjusted the height for Child Job tab +Just to be clear +What does QUERYXPATHDOCORDER return in that caseJackrabbit should probably enable or disable document order on all workspaces and not selectivelyJukka wrote it does mean that a client would need to use some yet to be implemented custom extensionmy intention was to provide a default configuration where you wont have document order but if you wish the repository to do so you could enable it in the configuration +Can you try with that +As per Tonys comment on the scrubbed export should be placed in your home directory securely with or more restricted permissions + +For general use nowadays private is preferred because the DS lifecycle and bind methods should not be part of any conceived API. + +properties svnnative +See +It looks like using maven there is only one big datanucleus log file +Got it I misunderstood the for i alone as meaning i rather that for i but not for triggers +This is directly related with and not at all there are some circustances where I need to inject another bean in my REST service bean and the injected bean is fully customizable by using property point is should work exactly like standard where I have singleton beans completly customizable via property placeholders but this isnt true by by itself is working pretty fine but its spring support has this minimal issue +AFAICT this is addressed by the kar packaging +abc the same asabc work in our test like excpected also the same way as with additional info here? Otherwise we will close this ticket within the next days +If you finish it before we release beta please correct the version +Closing issue. +We need to be able to update a Subject with update Kerberos tickets and Tokens and changing that would break a lot of other codeIt would probably make sense to make a that does a doAs and then removes all filesystems based on the ugi since clearly most of the Hadoop ecosystem servers have related problems +Committed revision . +Added sufficient margin to fix layout problems for Designer Editor preferences contributions +we already have a way to specific a default Action for a package +Obviously it is a different protocol so it wouldnt be trivial +With the postresponse creator it could be presented a lot better to the client +Set correct fix in JBT L +Will satisfy this ticket? Can I close this as a duplicate +I raised this issue on the jacc spec list and an extended behavior that could be added is to make a call to the with a role name of to allow a provider to handle a to any role mapping if they wanted +Been open long enough now +Still just focused on editable contentdisplayed content +TimLet me make you feel better the hack has saved my talk for tomorrow at Groovy Grails eXchange +version that checks true before executing our custom will try it out in the next few days +Committed a fix to throw a when makeObject returns a null. +I believe I was having this issue with version Maven +add unit tests +We have to fix it for final +This only requires those who have extended Encoder to implement the string variation and that could be backed out and simply overriden in the child classes +Updated patch addressing Elis comments +Heres a couple apps that Check the query time Check the last replication timeThey exit with error code on failure on s a cleaned up commitable if this spams things however its unlikely that Ill work on these. +This is for Branch right +so there is an additional bug here why is the client not getting the session moved error +changed the artifactId for the latest persistence bundle to be consistent with other glassfish artifacts in this issue +V looks good overall +Leak that Aslaks test shows is gone +I guess we can close this one now +If you finish it before we release beta please correct the versionThanks +Its definitely B +Waiting on response from Mark Little +Wow thats important thanks for catching thatOn my running EAP instance production profile httpsservernameinvoker and httpservernameinvoker both do not connect to a running serviceAre you saying that the CN of the certificate has to be the same as the SERVERNAME that a client uses to connect with the server? If so where is this configured on the client +Thanks for jumping in Koen +I wont fix this for now since its upgraded on trunk and a real decoupling would be a hassle due to classloader issues and such. +Thanks for doing all that +These are my only libraries +Yes Ramesh this is another alternative thank you +if no error found submit the form again or just execute the business logic by the same ajax request +This has always workedI checked that the class is actually in the work dir so why doesnt jbossweb find it +This was done for EAP and AS. +Thanks Chris! I might have more questions for you when I restart working on thisRegardsVarun +Patch testing now that addendum has been suite passesWill commit later today if no objections +Could you add TODO comments in the code if we dont have it yet +Set usecapturelobs on symtrigger +I implemented example fixtures for postgis oracle and db +Verified by Svetlana +Thanks Shai! +I think that problem is unavoidable if you dont flush the Hibernate session in your +Now selection rectangle is redrawn when we scroll a page using mouse or scroll bar +The whole point of this ticket is that we should test the archetypes +Patch +is there a reason that this shouldntcant work or is it a known bug +test case failure of test test from test is not related to this patch +Sean Thanks for marking this resolved. +Its clear to me that the zip timestamp changes the filesize changes but the app server refuses to pick up the change unless I do a full publish +I recommend not pursuing this as a legacy enhancement and instead rely on +Just attached the patch now but need the review of we indent the if block properly? In addition could we do more to check the URL redirect method in a while loop with a maximum of retries? We can add synthetic logic to pause and seconds upon each retry +How did my name get onto this? +Committed to closed. +Should the datanode send an error code to other parties +My original thoughts were that this would be a pretty thin wrapper that would mutate a single document when putting +Should have mentioned the contribution is in repo revision r +I think were all done thanks all +Maybe check in the changes commented out for now +Didnt even go into work so not to take out my entire floorIll keep you posted and definitely give you an update +This is no longer a trivial issue since the order component does use the approachThis appears to solve the problem by using the component syntax instead of a relative pathAll instances of the relative path in bsh files would need to be modified as well and I will be happy to do that after feedback +Thanks stack +Thanks Tim! +The versioning in this example is although we should discuss whether that is the right one +Sure every feature of Hive should be tested by hudson but I do not think replaces real testing +Thanks for the see fix ChrisRobert youre probably right +Boundary events now on correct activities. +To carry this through we really need a class per operation +PhilCould you please review commit for the JIRA +It seems that there are some problems in build +It did not appear the test really needed to have the JRE CPE before the MAVEN CPE it was just asserting the order +ToddDo you think the attached proposal belongs to? +I have done some changes to the exception handling +The we use need to remain independant from JBT parent because we want them to be almost independant +I just committed this +problematic also +Current precision formula p as the output precision gives some room +Definitely to get docs by id it would be cleanerThe real purpose of the patch is to execute the individual shard requests as concurrent threads +When I manage to deploy the JNI bindings to Maven Central all we will need to do is to upgrade the JAR packaged with Cassandra +Ill answer all your questions there! JIRA is good for reporting new features or bugs AFTER they have been still believe youre calling cargostart twice but lets discuss it on the list +Test to Andy since he already has done some work towards addressing this in his Scratch area +contains a small correction +We can just make it easy for them to get going on that customization and the first step is having an exploded deployment so they need not expand a war file +already approved once +This is an addition to workaround does not work if Continuum is run as a service because is not used + +For the service factory manager synchronize access to the map +Can you just let me know when you post it online itd be good to link it from Apache site and maybe blogtweet about it +I ran the unit tests and they passed +Patch committed +Looks like the problem didnt go away +Fixed in revision r +Verified in JBDS build. +When I tried to deploy something it wasnt picked up +Ill review this +Mark want to do a proposal for this one +The script will compile and run the example once with v on the classpath and a second time with v on the classpath +javatestingorgapachederbyTestingfunctionTestsmaster +It makes using Railo for a webfarm very difficult to so annoying having to set a root mapping for each instance +There is a race between the shutdown of the preceding test and the start of the am modifying the tests to use different ports and also the to close down the connector on exceptions +all in one patch +I will use the same technique to sort out DBTYPE thank you for the suggestion +So adding one or two methods is better in that sense rather than a dozen +Before going with the implementation Id like to explain how Im planning to implement itThe default behavior will be to remove the headers if they are processedIf a user wants to override the default behavior he can do so with a property +Yoav please confirm and if so resolve as duplicate +bq +Extensions should only be used in one direction +again + +Great cant wait to see it +Confirm I can reproduce with Keiths stripped down test case onIt occurs in both and modes takes about twice as long in +Im closing this for now +AddedIve merged your patch with some minor style changes into trunkFor future changes please open a new JIRA ticket. +Since the message is INFO do we need to have the exception logged as well? Im not used to seeing green bar and an exception +waiting on maintenance to open before committing +Plenty of memory is allocated for objects of class +Assign back to Andre to verify in about hrs when the respins done + +outputtestswstoolsjavaorgjbosstestwsjaxrpcsamplesdocstylewrapped you can see the name ends following SVN commits contributed to the fix . +This has now been fixed for the trunk and my docs branch +No exceptions are thrown and since only the formal parameter itself is changed it is easy to see that the refactoring did not javadoc refactoring will be very hard to do because javadoc is not currently parsed by the groovy compiler +Nightly release is just a build that is provided as is without any warranty +I think we should add AR and AP as two independent components under special purpose folder +ChunhuiYour suggestion is interesting +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I have no idea of what you are talking about seriouslyIs there a bug you want to report or +committed to trunk and merged to xtodo close for release +I have added the checks to the tests so hopefully db will be good now +svn directories when deploying a cluster +This seems to be an architectural decision which requires careful thought and experiment +Knut my mistake that I was not attentive enough to notice that I was using single quotes around the String I actually wanted searched +Id like to keep the two issues completely apart as they are somehow not related +Ive fixed this before inside our cocoon based CMS +I have been planning to move this to outside the lock for doing less holding the lock +I have no objection to your analysis +That seems a cool idea +Thanks Sandy! +I would really be sorry to be obliged to drop maven because of connection issues that seem not impossible to fix +It depends for real testing against a real cluster its probably just best to use the remote url feature I think +I am marking this as a Blocker since it is a regression from earlier releases +Verified with JBT. +Please fix this +Patch applied thanks Andrea +It will send the LOB instead and the client should be intelligent enough to handle this +There are a bunch of classes which will be used later on +Thanks guys for the reviewIve just committed it. +Already retracted that comment +So the Shindig java side is all about implementing both restful and the js +Otherwise looks great +Can you suggest a patch +This change should go into file +I dont think that the Derby community can answer your question about Apache POI +Final Patch with Release fixing some tab and idention issues at Ahmads patch I recognized some areas where code could be improved a bit and decided to go for them +Change reconnect to reconnect and post patch here +Reducing the priority as this seems to be a user errorThanksIsuru +Reopen if needed. +Done thanks for the patch! +Ive added a testing configuration that extends the global configuration and includes junit as a dependency +In terms of the projectjbt it is used in Savara to display the generated BPMN models +Please quickly check if this is still an issue with +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Changes made Added a boolean flag called anonymous to Added necessary getters and setters Updated endpoint serializers to not add the name attribute for anonymous endpoints marks endpoints as anonymous when auto generating names for on the trunk +Add a bit error check code and tested with branch +Model demonstrating should have noted that the demo zip unpacks as a demo java package location just add demo into your classpath +I committed this yesterday + upgrade affects also documentation due changed openidjava dependencies +Why would someone want to store file data into Accumulo instead of directly into HDFS? Or into a super file +HiIll be taking care of this +Applied edits as suggested by Lev +I am currently struggling with actually moving the perisistence into a separate module +bulk defer to +Juergen +I will change the modifiers to remove duplication that was just a quick and dirty try to see if any of the existing tests were to do about Map check? Any specific things you want me to try or we respond when we get kicked? I ran the tests and nothing broke +Thats rare though Id say +Maybe that might give some clue +This is the eclipse project in which I am seeing the error +Would it be possible for you to run this with DEBUG logging turned on for the server and attach the server logs from after the test run +added test it fails without patch and passes withoutAlso removed dead overload from +For eg the order of tuples in group the top tuples selected by limit etc +Thank you +This is a prerequisite for +Youd still be able to use it in single user by which I assume you mean a single database user via JDBC +Hi Robbie could you review this commit please +Please check that the bug is fixed for you +Committed to trunk thanks for taking a look Stack +It is not that I dont trust you to write good code here it is more that both Alex and myself need to see that first hand to make it better in the future +Thanks JasonYes I failed to stage and commit the new file with the next patchShould have tried to reproduce in a fresh clone instead of dev environmentThanks again +A to ZK is expensive +But as you scale up in either or complexity of documents it becomes a bottle neck +I have done this now +If this is still an issue anyone have a test case to attach? +Attaching patch for now +Im going to try to carve out some time to look this overOverall it seems quite good and fixes a number of big issues +I just committed this +More things after looking at the code doesnt handle the leave of multiple nodes well even if the notification is done once per leaver if theres an running then it is canceled the list of leavers is then updated to also include the new leaver the stateProviders and receivers is calculated based on the last leaver only it should be calculated based on the entire set of leavers +I am committing the first iteration of the decoration API +We would need to add compiling of single files to maven dont know how +So probably not +Thanks +Correct me if am wrong but DFS fsck does not fix bad imageedits file on startup +Thats work great with bypass the volatile image buffering no more system property has been added to support disabling of the volatile image buffer When set to true the volatile buffer will be bypassed and standard image buffering will be used +Submitting patch for Jenkins testing +Your patch is correctSo the thing is as follows on trunk I made the caching work against the sole name for speed reasons but then I forgot to amend the LRU cache cleanup procedureWhen I did commit the fixed LRU cache cleanup I was thinking I made the same change on but in fact I did not so I back ported a patch for a problem that was not thereApologies for jumping the gun +Committed the changes to to include Wookie WAR in cargo config +I am not sure at all that its at fault here Are you using SSL? If yes are you sure SSL is properly configured for JDK? Can you provide me with the complete wiredebug log? all I can see from these logs is that the server fails to send back any sort of response +Thanks +to replace above mentioned updated after aidans configuration changes. +lowering priority a bit +just got word from Steve great that someone with a linux box could get around fixing thisone thing though with the change in signature are you completely sure that will guarantee to work with respect to classloaders etc +It doesnt matter if it takes more time were committed to implement it as a way to give back something useful to the library +I think this solution may work just need to remember that we support runtime registration therefore the cache must be dumped if a new provider is added +So maybe it would be possible to add some code in the to supply the custom editors. +another patch which does a bit more changes but makes bookmarkablepagelink more flexible +Fixed since version +has been marked as a duplicate of this bug +MERGE statements combining all three kinds of actions INSERT UPDATE and DELETEWhen setting an identity column to the literal DEFAULT value the UPDATE action of a MERGE statement behaves in the same incorrect way as an ordinary UPDATE statement +Hi MartinThanks for providing the patchIt looks correct but its hard to say without further investigation +Will file a JIRA to fix this +A solution will really help +In there is a long discussion about pros and cons of serializable configurations. +This change makes pass again nowJuergen +They demonstrate the problem they fail without the code changes andthey succeed when the code changes are applied +This issue has been resolved for over a year with no further movement +Its just a matter of style and future direction of the components. +That fix solves also this issue. +I havent finished with Unix yetsee +Also this patch shoudl merge well with the current trunk +This trivial patch only affects commentspatch p i Please review and let me know if there any issues +fixed in ivyde +Ditto for replacement is to use an +No testcase added as change is done only in log message +Good stuff guysstack assign to u +Okay heres a patch! It adds and annotations to their respective generatorsSince annotations are I considered something like and but this felt a bit like overkill +Reassigning to move of all unresolved issues from to +pdb files are handled in +A new packaging structure is being made which will allow for exploded deployment as well as a new and improved API and UI for packaging configurations +Add Stack widerow mode we werent decoding the key since we have to track it to recompose the row into a bag +In that case we can revert to the original proposal introduce a new feature specific to the extenderIt could be more flexible indeed +this filtering merge reader approach also fulfills other requirements i have for migrating old indexed content to use new features so it would be a for me to use this filtered merge reader approach to ensure consistency and conformance with my schemaI just have to repeat myself We dont need to modify Lucene to make this migration work! See my last comment it lists the whole workflow to do this merge on the directory index without the need for temporary index instances completely +Thank you for spotting this. +But IMHO a formatter is almost useless without dozen of configuration options Thats why I suggested earlier to take these options from the Java preferences when they exist +Mike I think Marvin did some optimization on long queries by breaking them up +I was able to reproduce this behaviour also on trunkThe following classpath was okCLASSPATHwhile this one failed the same wayCLASSPATHThe problem seems to be coming before +Do you think this problem relates to HEADER HTTP request +that would be much better +I have opened for the remaining test failures +the invalidation should happen after the applied the changes +This issue is fixed and released as part of release +IMO writing the schema in HDFS or in Zookeepers namespace should be the same operation so it should not be a major problem can be the same code but with a different output +If you installed the Beta then you also need to update this to CRYour requests tells that you are only updating from Beta to CR without updating which includes this feature +As I outlined a year ago these adapters are simply not intended for JNDI binding +Can you please provide a revised svn diff that does not tabcharacters on all the source lines? Besides violating the normal Strutsconvention of indents and no tabs it makes the actual changes veryhard to determine +write UUID as a writing out a resolve any unknown Object to with a test to verify the encodingdecoding of UUID with the changes to +Lowering scope to found a couple of errors in the test cases along the way. +I probably would just opt to comment out the entire two offending classes and modify the pom to not export at all since the modified io package is not really usable anyway and could cause some confusionEither way though the dependency will be removed until the Alliance makes the necessary classes available +STARTMISS ENDMISS was shown wrongly even for cases SLA was met +Here are the test logs for a run that stalled for me +The already contains the neccesary information sincePropably the site hasnt been deployed since then +committed v +has this been fixed now +So throwing an exception about what is logically partof the next transaction seems a bit to meSo it seems like the engine could be altered so that rather than throwing anexception in this case it could just quietly leave the connection in the stateof clean but cannot be used for any future transactionsThat way there would be no exceptions thrown so long as the connectionwas never used again +Ill try to add a test for it too +Site update +Attaching utility jar for removing unused resources roles and related data directly from the database +sorry about that +The Create Relationships actions were removed in so they wont be part of the pictureStill need a clearer picture of what you are looking for Van +on v and you add a short to describing this feature +This helps w MTTR +which repository are you using +You might want to try uploading that again +I just committed this +I had to do an SVN add to get it to show up in the diff +So an access control subentry subordinate to ousystem naming context cannot control access to other naming contexts +If you need the repository xml instead let me know and I will strip everything out and it all for There was no error in the log +Works for urls with explicit file protocol. +Want to make a new issue J? Lets keep this one for repeats of the substitution depth too large exception +There is a Jira in JBAS that talks about confirming that properties that are in components are the properties that are needed and not missing any or have any extra it doesnt need. +If Cocoon trunk synchronizes on top level http requests the same issue could occur from my understanding +Attached which shows changes I think need to be made to the CBR tooling +Therefore I do it now +This would break UDDI v compliancy? I guess it would be a vendor specific feature +It seems your complaint would apply even more to the existing instructions and Id assume you would support a patch that removes them but they were added by someone else who thought using eclipse without the sonatype plugins was a viable route +This part should work normally with the current patchIn case that turns out to make a real difference it might also be considered to deprecate all the Searcher methods that take a Filter argument and indicate the preferred alternative implementation with a Filter as a clause to in the javadocsNow if I could find the time to get this last bug out of the current patch +Oh sorry read the rest of my mail +SOA Platform is the product name is it not? And Data Services overlaps with the name used to also include hibernate and in the future infinispandatagridThus I suggest we just keep it simple for nowSOA and Data Services ToolingBusiness Rules ToolingOk +The patch should be applied to the trunk and the branch +Add minmax split size settings to unit tests +Commited to trunk and to. +Is there any progress on this issue or at least a time estimate? Wed like to have a central entry point for the community and start documenting +Committed trunk revision Committed revision Thanks Shay! +Ah this isnt a problem on trunk +Removing from. +fo file is has been added with description fo file demonstrating list ve dove into the first testcasethis issue is related to within a simpleThe nd testcase isolates the causing the has been added with description shortened transition to closed for remaining resolved bugs +Hi IainDid you get a chance to make a patch +I have just tried Erics patch and it works for me +We have many Jira projects and we need to map one sonar project to one jira project +I did some optimization +Thanks for the patch Bryan. +I think this patch wont apply until is in +Tom can you look at it to see if it is fineIll also upload the example given by Simon as a +The reason might be not everybody wants to use XHTML but might want to be able to use HTML +Ive committed this and merged to +batch transition to closed remaining resolved bugs +This sounds like a good suggestion to me +I was just investigating the possibilities of importing your Bugzilla stuff +See the link I included +The limit that I encountered is characters +Each class has one property name +That would ensure that my suggestion of a collector with two separate mincounts would return the correct result for counts as well as terms as long as it did not exceeded the given limits +Ill try making the change for Java client +Patch file for the improvement with sequence mediator + +Thank you very much! Looks good +Thanks for the quick feedback Bryan and Kathey +Username woodydiwu +Hello GeorgeThe fix looks good +Changes as a patch +Perfect to have a maven goalstart +The trick about closing and reopening project is an old workaround for any kind of dependency resolution issues +If we want to support enablingdisabling log aggregation as a rolling change through the grid then we need a way to refresh it +Interesting +The failure of was not related to this +The advice he gives here is the same as he gives a couple of years later in +So I am very much in favour of your proposed solutionOne warning I already hear complaints from those using an IDE to run the whole test suite by right clicking the test folder that all tests are being executed +Thanks Sam +Please see the mailing list discussion about backwards incompatibility +Thanks Jukka +Does not contain or HADOOPCONF variable +Heres a patch that seems to solve the problems for me +truekeepOriginals isnt actually quite correct as what it actually keeps is the unzipped unversion version of the concatenated bundled file +Instead the observation task of each index component should be wired up directly with the underlying and the execution schedulerbq +bq +Is it possible to merge changes to Grails branch in order to include this fix to Grails +error messages to be more descriptive +Why not +I noticed that the latest branch has been updated with +I am currently on vacation from thru +Patch with fixed tests +Applied in rev +Anyway Chunhui has said he would update the patch +You are limited to what is retrieved from hbase and any other data is lost along the wayThere are ways that can be used to forward that information using for instance the dynamic router pattern to route the message to its destination based on the hbase result but this does not provide an ability to propagate the hbase data to the destination unless you add all of that information on the required end point string returnedLooking at the content enricher pattern I see the use of this technique but in the prior case +But file didnt have any references to jcc +Ive been working on it for a while and I made it to work for jpdl definitions +However this has not been extensively tested and some further work is necessary in +Maybe it would be nice to spawn separate compaction process? It is quite operation so maybe it make sense to separate it from serverIt would also be nice to have cli tool to compact files without cassandra server for backup purpose why not spawn such tool from server +Thanks for the patch Michael! +Fix for freemarker plugin is not reason to respin +create a new session on authentication unless is disabled +Since Dominik just reopened I am reopening this JIRA as well +Can you provide the object graph? Is an instance of a holding the CDI factory? or is the field holding the reference? Or is it one of the delegates within the mapThe is supposed to be just a shell +Is a duplicate of this issue +Applied the patch +The Java Debug UI is expecting that whatever source files it is looking for will always end in +yes the issue is with +And thanx much for the incredibly quick turnaround. + on patch +I might change it back if jvm exit codes prove consistent across implementations. +for most applications you will already have something in place for failure detection +Verified on JBDS +So ill go for that instead of forcing the const to a static value +I begin with the first one +Not exactly the same way but very similar +Use the overloads. +Closing the resolved issue. +The is outputting + +Ive granted admin rights to all on all Jetty plans also. +Without a unit test I have a very hard time committing the proposed change +Looks like we might be doing similar things +What was the issue with emma and the other tests just in case this failure is related +This can be easily fixed in by having it write two class names for Writable parameters both the declared parameter type and the instance type +true exist or is it more per component allowed? The spec ist unclear in this point but no serves as please select a valueitem +Maybe its good to provide units in the description field +Just wanted to update this to let you know I just added support for the file into the jaxb databinding +moving to +I looked into this today and realized that it was a problem with a particular Jenkins slave +Thanks JessePatch applied to LUNI module at repo revision rPlease check it was applied as expectedI suspect there may be more cases like this in the Harmony code +Is there anything that needs to be done to expedite the scm release? My companys cvs will actually hang when using compression which is another issue entirely but having a good way to configure that would be incredibly useful +Users account deactivated +It should include maps type as wellGood catch +It looks like this was already done in. +Assigning to Josh that the need for projectscaleVersion suffixes for everything is obnoxious but is the least bad option and would be appreciated by everyone +respected posincr and poslen. +Fixed with the new console snapshot +There are and +Full publish will not build the project +Given that we have an Ant expert on the team and given how far Ant has come the unofficial concensus is to continue with Ant in the future +Found an issue buildr tries to update regular artifact if it is installing this is expected functionality according to specs but seems need to rewrite method for downloading artifact to not break existing artifact +Removing a non existing property succeeds vacuously +Ouch this one bit us hard we have a lot of final classes and classes with private ctors +Unit test which reproduces the am able to reproduce this when I connect stop and connect consumers many times +for the patch +do you have an example that allows us to reproduce the problem +I searched before but solution was not there alternative you have given and if you know any pdf properties that have to change then let me know think will fix this problem since the garbage seems to come from the +Thanks +Triangle appears next to headers +AIUI there is no intention to resolve this +update to sdo javawsdl and wsdljava poms to increment their emf version dependency to +is there a simpler reproducable for this ive tried running on my work station but just get OOMs and hangs +As requested via IRC here is a reminder comment to add sugar methods like set to +Thus difficulties in finding a suitable seed are an indicator for problems in the code +committed +Have a look at for further details +woops! my mistake +Then make a new class to use in your round robin table create + + +inBy the way if youre using you might want to consider upgrading to +Priorty was inadvertently changed to Blocker +So apparently this works fine for current dojo +AFAIK an authentication cache is already used +I ran this a few times on cluster its handy +Hi OlegBelieving it is resolved is good enough for me thanksThe issue of adoption of is complex +This Tests naming convention is contrary to naming conventions +Tested on a Windows VM with Maven Java u and Git. +Thanks! +Thanks for your contribution looks promisingAfter a quick test it seems that all rotate and mirror samples give a black image +Yes this can go into once finished +Attached test case that demos this issue +HiOne solution is to use a listener and set the Time stamp or set the time property at the constructor levelBoth suggestions above would require to remove Generated annotations +Logged In YES useridHellodoes anybody care about this mistake in the documentationIt is actually in the paid J? In YES useridIts been fixed. +There is one more thing I should add +Baseline performance number was hours. +This module simply isnt meant to be used outside of an EE Server +Your patch has been integrated +After all it only had a test project a patch and a user that confirmed it was if it doesnt work for you then at least provide a test project that can be used to prove that it doesnt work +just synced the latest with cvs +version of the patch +Jimcould you please try if the cahnges in the attaced are working for you +In fact one account can have multiple networks so I think network name make more sense here +It looks OK for meThere is only a mismatch of plugin version +there seems to be a problem with the file before since i cannot apply the patch +Not sure about this thoughI didnt implement the check deref nodes +And note just to be clear the testcase this JIRA uses does actually run successfully +Change to blocker +Gotcha so a faster way to reveal the full timestamp and copying is less important +I was saying that does not +and here it is displaying correctly BTW Ill have a look at it again. +Improvements to Javadocs can be done later of course +This makes it much easier to have new users Google for the website and follow the Documentation link +Yanked the code from Harmony +RolandAt the very least I will try to document it somewhere as a possible cause for a socket reset at the server side +No further code changes are necesary for yoko + for fixVersion +Will give that a try. +I think is may actually be more of an issue with the performance of the weld portable extensions implementation rather than weldx itself +No worries +Just writing the documentation +Submitted a patch to trunk +More research shows that the problem is that you cannot calculate statistics for tables that are part of the app or default schema +Note that this is dependent on the masterinstance heartbeat rate and not on the slave instance heartbeat rate +Hi Steve what do you think of this patch +reopen to add fix to backport to revision together with revision for to with revision . +Fix in place +Reopening to deal with error in chain when a DOM is processed twice Unknown Line Column HIERARCHYREQUESTERR An attempt was made to insert a node where it is not permitted +No the topology was not defined +Perhaps youd like to go over that JIRA first +I have tried to reproduce this with a local build of the project but was unsuccessful +I am going to work through the deployment issues with are four changes to be made The identified change The identified change Update of the to use the app server jars Update of the war file creation to remove would have attached the generated ear file but the new jira spent mins trying to upload even before I had pressed the attach button before telling me that it was larger than MbThe new ear can be regenerated by removing the directory and replacing it with the current attachment +AFAIK nobody has looked at Jersey integration +Closed all tickets +bulk defer of unresolved bugs to +I just dontthink term diversity is so important that it qualifies as a base litmustestThere are other ways of choosing good fragments and IDF is one of them +The only solution available in bridge is wrapping original request by the wrapper +Are you aware of any ftp commands which can helpIMO there is not much we can do about it using capabilities of ftp protocol. +Both Firefox and chrome have the problem on Windows Tomcat +Marking closed. +I think this is now ready to commit +The source editor will also freak out sometimes +Jan looks like you are just waiting for feedback for the to trunk hereNothing remains for +Heres a version of that uses ant get to grab the dependencies +added +Im still learning Scala so please provide any feedback +Patch looks good +The are additional targets in but the old targets are still there +My teory is that I have some xml documents inside the project which are causing this reaction but I need to investigate further. +ah sweet irony +So ability to restart them sounds unrelated +Part of a bulk update all resolutions changed to done please review history to original resolution type +Change binsh to usrbinenv bash to ensure being used in these scripts will work correctly +to trunkThanks for the review Anoop +Likely a jar conflict or parent classloader configuration thing in WAS but not really sure. +Agreed +This also affects version ve just applied the change to my local version of xstream which solves the problem for me +I just committed this +Dima please check is all correct and if possible close the issue +Well that is the only way to ensure that the escaping is precisely the sameThe method is saveConvert from and is a private method +Hi MridulWhy cant it stays side by side with Side Deep Category +Found another example where was required +Sid I just put this into and trunk +Hit enter too quickly +Maybe thats good enough? Are there other options to consider beside the following? add options to list scans for selecting columns to print do nothing and let users use awk or like the csv option +Lets consolidate +fixed with commit defcffcfdbaaeapplied to and master as well +Added and +You are right +Just verified that this has been fixed in +Ive committed the patch +So in my opinion it will be better that when is false the default values of memory and disk are set to MB per task and respectively as in the previous configuration +Since this were still discussing this Im moving this to only +Yes as far as I know this can be closed +Patch to patch to cover the target as patch to run native builds by default but disable when false is provided +Actually looking at this more closely it appears that your problem is overflowing of the holidayid column not any of the date columns. +If there is a problem please open a new jira and link this one with that. +Yes installer that just adds runtime paths should behave just like adding runtime paths in the UIUser should be asked on startup if the installer just added pathsOnly if the installer explicitly added a specific named server do we do autosetup +branchesJBESBCP was supposed to be an AOP only change but has also been modified +and trunk as noted above layerStyle should also remain in use then +The patch is reviewed no issues found so far +I have added an overloaded onRegister method with as well as instance argument though so its easy enough for a subclass to keep a Map of actual to isntances itself checking it in +Hi AslakWe have blocked test development because of this issue +Tested +Good luck! Leo +Patch for. +Also I dont think we have really ran into the problem as long as I can remember +Need to think through how +Bulk closing stale resolved issues +A different approach to solving the problem +The Proxy class is chiefly intended as a superclass for classes which wrap a streamPossible improvementsMake it abstract to avoid confusion +On the mailing list it has been suggested to add enumerations to +Revision +Changes look fine +Thanks for the patch! Is more work needed to incorporate this into the next Derby release +Ive temporarily the change because it breaks the get the latest snapshots have you tried building withbuild U? +This test is not failing with JDK so it doesnt look like a test related issue +Verified in r closing this issue. +The mirror is some days. +The tokenstream backcompat was a bitch as well but we had to slog through itMark honestly I do not yet know how this one could be reasonably doneThe problem is the behavior is dependent a lot upon the users JVM how in the hell do we know which JVM version was used to create some given index +I will change and report back with the results +ACS feature freeze date is November th +on these changes from meJonathan maybe you could port the same changes to since you run on windows +task level edges is always on +Attaching a simple patch to use the behavior of only in case of wildcard addressesI hope this solves the patch appears to be working for me as well +for me on this patch as is below I dont see this causing any breaking changes to anyones existing codeI dont think there is any real reason for as opposed to +bq +Patch looks good +Patch applied +Basically you need to go and build the src +Ignore this comment cos its not right +b iendtifying instead of identifying in the same chapter c is that is allows instead of is that it allows in the chapter +The problem is a defect with +the handler is not activated fast enough to be useful for version number. +Modified sample Activity Stream entries in +It happens when importing projects using xmlrpc api using addI give an url without credentials but when continuum tries to get the pom from viewvc dont uses the credentials in Ive made the subject more specific well need to take a also fails to use when importing a project by uploading the pom file +Changed to minor status +What are you talking of TedIll commit unless objection +The Java type that is contained in the result of a query using an aggregate function is as follows +Exposing a good one that a hosted provider can trust to obey CORS not hog server resources etc is probably more tricky +load does not work correctly +bug in the +NOTE however that the default number of threads is and the default maximum number of connections is parameter is still used though its default is now parameter is no longer used +perhaps it would be less pain and less confusion to do this switch now with this major version number bump +Adding the attached file called in servicesorgactivemqstorejdbc of removes the warning WARN Database driver NOT recognized +Resolving for now since my PR has been merged hence it should be explicit to the user that they have incorrect credentials +assign to Daisy and pushed +attaching my test file that i was using its derived from ran some quick numbers using the syn file example here just best of runsImplBuild timeRAM usagems mb ms mbI modified the builder slightly to build the FST more efficiently for this will upload the updated patchSo i think the and RAM consumption are really improved the next thing is to benchmark the runtime perf +Nick can you take look at this? Also reproduced log pretty sure this worked previously as long as you installed into a user writable dir +thanks tucu +This is exactly what the enforcer plugin can do + Looks good to me +Added jbossbuild componentinfo files +It had nothing to do with the merits of validating vs non validating +Thanks BabakI got your point as I didnt get the Javadoc update progress automatic it could be handy if we have a maven plugin to do this kind of job +Ive committed this to trunk and +Well we do not lookup registry to get resources when building the proxy service policies this should be fixed +This patch excludes the bzip and gzip files from the change +It looks like the crashes have stopped after disabling HTTP range support +Without this in my test other threads that perform namenode operations could not get the lock frequently during deletion +The author tag was existed in Logj +Thanks a lot for the help Jun! Resolving. +Would you mind to commit this? I am currently not on my linux box +Works fine with xsd and +There are lots of dependees but the obvious change for them would be users and thatll be too confusing +The files could be removed as well if they are found to be confusing +Bruce thank you for good comment +What should the reader threads synchronize on? and correct publication is only guaranteed if the writer and reader threads use the same lock so the reader thread needs to know what lock to use. +This will be due to the fact that Findbugs is now launched in a forked JVM to support memory configuration and time outs +The controller registers a listener to the deleteTopic path and when the watcher is triggered Send stopReplica requests to each relevant broker +Bug has been marked as a duplicate of this bug +All tests pass +Tested no error on tries for me +yeah I mean think the fix from might resolve this issue +go ahead and ignore my comments about parser it was not based on reading the code +The patch is complete and ready for regression testing +My apologiesIll submit an updated patch +Youre right +Gordons comment on confirms the headers exchange now compares the bytes regardless of encoding. +we can use current instead of you for the patch Terry +uid should be username above +Doc looks good to meThanksAnshul +I use Oracle and had issue with boolean type before using JDBC a lot +The build farm shows this working on +In the discussion we dont want to use weak root for tagged object support I will keep the discussion in mind as a future enhancement feature and close this issueI think it could be useful to have regression test committed +The resolutionIds is for all outstanding issues +Any progress on this issue? Patch attaches but still not included within the trunk +So how bad is this? Will this prevent secure operation? In that case Ill sink the current RC +Feel free to take it over +This would be a very valuable addition +Any specific reason this should be in rather than +Given the protobuf change where optional field no longer carries the previous default value this should be marked incompatible +My reason isI have reflective proxy of that measures query execution times +This test was executed on AWS s with heap settings of +The emphasize why this is important every class in your app is on the classpath in Eclipse yet when the test runs especially in a remote container if the class isnt in the deployment fail +I threw them out there in the hopes that someone would counter with something better +Would it be possible for you to recreate against the current source treeThanks ZoeI am now attaching a new versionof the patch based on revision +Ive just this minute added some comments on this issue to the mailing list. +Simple patch which fixed the log spew in the case I was testing +MRA phonetic codec patch +I will submit the patch for that in the next hours +Let me ping the RHQ guys to see if they sort this out +Thanks Philip +So far I am unable to reproduce this issue +thanks for all your help! +jar to and trying to register new user exception happen described in issue +If this is done the archetypes should be updated not to specify the home config +the code is not exactly straighforward +Work around like problems the class allows different vectors to be different classes and so making a like Matrix with more rows is impossible +I also have verified the ability to kill a running job and have it propagate down to stopping the child processes at the OS level + +This is a candidate but not committed to EAP d say its not needed to put this in Release Notes since it doesnt block any functionality +Nigel are your tests OK now? Thanks +On RI it is always located in the center of the window +Did you have a chance to check the code above that handles the reactor dependency? I am not sure how to test the reactor code +Hey Michaelsure thing +So I think only if the set to the test case will be valid +As those are brew builds they can be built from the need to add this to the Release Notes +How about generifying the content or just automating the download +I didnt get a NPE when setting it to trueCould you give more information like console log +Please give a few daysSorry I was only focusing on Japanese languageIt is so great to support another CJK languagesPlease go ahead +Struts has many ways to pass configuration information to your Action and it is designed such that you shouldnt need to know that doesnt even exist +Attaching +This will override any look for the latest behavior maven may otherwise use +Maybe it doesnt exist the requirement that switching document from template to +I am of the idea that in order to solve this problem we need to have a base class for complex type classes that we generate out of WSDL +I agree that the does not work with timeout but then make it clear in Javadoc +jython is easy enough to integrate +bulk close of all resolved issues in preparation for release. +Some candidates test case not provided. +So there seems to be a bug here +Before committing Id like to make movedata optional +Having received no comments committed patch to documentation trunk at revision +I attempted to use the solution proposed above but it had no effect + has been implemented for the next it will fit your needs I thinkAnd we will probably have a release within this month +Thanks and Regards + as I said before on IRC I dont see any benefit of this +This is unrelated to your changes as I confirmed this is also the case when using +It also clearly updates resource capability instead of being an implicit harder to follow changeHitesh had the additional idea that such normalization could happen within validation before application submission +Then why dont you check in the second patch first so we can use scpexe meanwhile +Attach a patch +looks ok to. +What could be done quite easily if its not easy to make sure that xooki always use the same eol char would be to use ant fixcrlf task on the doc before commiting +I think rest server has to get the data at first then to find its size +I would like to take up and work on this bug +T E S T SRunning run Failures Errors Skipped Time elapsed sec +patch looks good +More interesting i the fact that the thread doing recovery has gone but apparently the database still thinks its in a slave state +It can be used to specify field cache values while indexing with RT turned on modally +Only one unit test failure with this patch +Currently the dpi works but has some issues do you need just to start testing or to deploy +This fixes my problem thank you Mikko + +And since you committed the fix yourself I assign the issue to you. +Ugh the patch is full of reformats +Please check that they are applied as expected +Alejandro saisI created a branch for the fixed a broken and set the version numbers in the poms took the console for a test drive with the websale process and itworks just fine +release bulk close +Patch to move into and the rest into examplespig +that last one should read sure if this fixes all issues but adds a test case for some shim validity +This is fixed +Well the problem was getting all the stuff to work at once +Ill do something cleaner later todayPut this in too +HiI have attached a patch containing unit testsThanks +Now when no class loader is specified at construction the class loader of this class will be used +libUsersrjurneypigbin +Perhaps its never used as such? I note that has a similar line commented out perhaps masking this bugWould fixing this be an incompatible change? If so would it be worth it +The current patch only contains the renamed bug fixes +Both of these issues can go as a subsequent patchThe aggregate methods that are covered by the patch are COALESCE CONCATENATE NULLIF attached patch using revision +Thanks +Without implementation for it seems to be wasteful to implement these tests too much effort is needed to access the needed packets information +Setting the Patch Available flag +It looks like the current code will sort the fields by field id before serializing which is at least correct behavior if potentially suboptimal in the current implementation +I just dont like replication to part of the name for it then +Ok I let you know +There are several ways this could be implementedand Im not sure what best fits with other post changes +Do we still need this JIRA opened? Isnt it addressed by one of your lock refactoring? If so please link to it and close this +Thanks Gunther! +Lets keep it simple and extend it in the future if neededIll make the changes you propose create the new unit test and resubmit the patch +This patch removes and keeps creating Path +But this will work only when padding of the counter is in effect once you exceed the padding the typo issue could happen right? you could do the typo killing for when trying to kill granted the likelihood of still being around when is around is minimal +I just forgot the langjavaavro when I was typing that inDoing an old fashioned mvn clean test works from the root with your first patch +Eks the original patch looked good I did not change anything in your code I only added the above mentioned tests and other changesOne thing As checks the buffer sizes the test in the for Token must also be changed +Moved the based orderinglist from the sandbox to the easiest way to see what attributes are currently impled in a component is to build the dist module with the release +replaced by +needs this feature in Rumen to be able to emulate CPUMemory usage +This works fine for me using TomcatA couple of notes though The build file is fine now +Retrying hudson +This check is kinda historical actually I just need to check whether it makes sense to add it or not +Yes I write error it is can you give it a reasonable valueI wonder that times is enough +Patch v adopts Lars suggestionI am looping locally +New bug logged there +Will use both to implement this feature +Attaching a patch that moves to the actual Solr release and simplifies DEB rules +To day that I simply aded a classifier when deploying the jarmvn deploy Dversion true Dpackagingjar reggie Dfile maven false rio Dclassifierdl Durlscpmvn deploy Dversion true Dpackagingjar reggie Dfilelib maven false rio Durlscp +There are placeholders in the code for for the following classes in the module +Patch to support USRIDONL security for network server +Isnt it better to fix the bug instead of the symptoms? you are right cutting corners +Just wanted to make sure where to apply js file within sandboxgsocjs +Ning hit the broken dependencies he may be able to tell whether this list is complete +Initial patch +Well deal with that in a round of usability tweaks well make for the next this one as DONE! +The comment is wrong as the Itanium has a relaxed consistency model +I have no idea why its not being documented in the documentation though at this point you can no longer claim its internal api since everyone is using it +and any potential work is the work for this issue complete +for the quick review +I would like to correct this constraint in my local sandbox but its not clear to me how to build the part of the validation framework +Thanks Alexei the patch was applied as expectedVerified closing. +Only comment is that we should include the V for version info as well +Order of build doesnt matter it just needs to be used somewhere as the extensions are is to deploy a snapshot of the plugin so that Maven downloads the snapshot uses that to load the extensions builds the plugin then reloads the new plugin when it is used +verified by Vladimir. +I am going to look to package a couple of other fixes and then release a version to the dev site for beta testing +svn ciSending assembliesDeleting assembliessrcmaindemoscommandTransmitting file dataCommitted revision . +Thanks +The nature of the API is that its easy to extend and difficult to removeThe question is What does it buy us to have separate configuration for for stuff that can be gathered from existing will be changes in identity in but for this part it should be rather evolution than revolution. +Your help is much appreciatedI may not have understood parts of your previous comment +Lets make it simple submit what we have and build upon thatI dont think thats simple The patch can be iterated on outside of trunk as easy as in +with this patch +Sample Ant build file to showcase the problem + +That shouldnt be too difficultSteve does need to make sure nothing unintended happens +Committed +I am working on it +Created a separate patch for branch only have schema why do we need the SQL scripts +bq +Id still go with factoring out the differences +griffonversionprojectsappnamepluginsscripts next release of the wizard plugin will fix this automatically +Thanks +btw does the patch still need to go in +This windows xp notebook is years old without reinstalling + +It seems something is wrong there +completed patchcode changes testsfull build passes successfully +Committed to sqoop +Let me see if I can at least get the current site into staging so we can flick the switch. + should probably also unwrap before passing them to listeners +Is there a chance for getting this patch integrated into? +Ill post also a bug report for this in Xalan + in revision Vladimir please confirm that my fix works for thanks +In the interest of moving this forward Im going to mark it resolved since the issues fixed by the patch have already been fixed on both branch and trunk + +Here is a patch that I think will wrap the Socket exception with a localized DRDA Exception and then the test should get the exception it expect and not print to stdoutI am not sure how to test the patch +I will put a git patch with the fix soon +for fixing this ticket says Fix Versions +Running on Windows os the s delimiter becomes backslash hence the count is Yes +We can just do it properly on trunk and live with this minor performance hit for consumer rebalancing onChecked in the latest patch to +This is fixed covered by the customerinfo sample +The patches on supersede the ones attached here. +Okay its in! +I know whats happening already +Given that Java will be giving us Closures at some point in the future it does seem that the Generics version of Collections should consider changing the name +OK good to hear that it works for you now! And of course thanks for the kind words on Spring +To be honest Im not sure this is the best approach anymore +zhihongDo you think it is best to add it as a module as opposed to keeping it completely independent and just having hbase depend on it +I currently just need two categories unit and integration but it would be nice that the concept is I said compared to running unit test integration is way more involved and it doesnt execute in the same box so I do need to be able to import it in the same of course since both are about the same its useful to be able to see separated and aggregated code the categorization support more than one key per covered linebranch we could push to Sonar the name of the test so that we can then answer questions like Which tests cover for this classmethodlinebranch and then a developer or automated tool can focus on running only those tests. +you mean the parameterYes that has been enabled +srcjavaorgapachecommonslang +Attached patch fixes at least the behaviour you are seeing +Excellent +Made a simple jsp page which displays the user profileProfile page has tabs for the time being which on click will have a panel slide down to show information +Thank you for analysis when you have test changes ready I can check it on +This works as designed +I am open to suggestion on indexing chukwa data +Committed the a patch to trunk with revision +Ok as we dont have any clue about how to reproduce the problem I mark it as cannot reproducedIt does not mean the bug does not exist +Hardcoding of current rbw finalized etc fordirectory path names is removed +Adding warnings for these conditions is a great idea +what is the status of this patch? +Steve Yes +Prasad can you attach a sample application which reproduces this +Committed this to trunk and + +You mean that we should not try to replace BSH snippets in widgets and minilang because there could be side effect doing so if its not done manually for each instance thats itSo we would try to create a Jira by kind of replacement to do +This issue has as fixversion +Duplicate of. +Ive never used Maven since Maven is still the default on any version of Ubuntu + +Why increase the initial sleep in join +Not needed for now well open separate tickets if tck tests need update +Done + +There isnt an easy fix to this patch a different approach is required +The original attachment lacked the which is really the heart of the problem +This bug eliminates most large organizations from using the site due to the windows platform +I am running the tests now +binhbase Configurationrows columns per row versions size of size of qualifier size of value Populating dataTestingNext us us usYou can see here that next is much cheaper than reseek +Fixed by patches in . +I dont think they should be part of the user APIcan I delete them entirely until they actually get used? patch that removes the timestamp methods +RolandIn my opinion this is such a corner case that has hardly any relevance in the real world +Please verify it was applied as expected and solves the original problem +Moved the SSH key related problems toMalteplease continue to discuss the key related problems on +We need to have them first to prove that the current implementation is not working then apply your patch and run the tests again to prove that it works +EJB Proxy has tests to ensure all proxies are unbound EJB Core is faking it until Proxy is fully integrated +The code change looks good but we need to make sure this is documented somewhere as an incompatible change +Is it ok if the repo snapshot is only kept temporarily for these DR builds? For example months +The first is and the last is RC with believe I at least know what we are doing wrong +Applied. +Patch applied to r +Thanks Jonathan the patch looks good +backported from trunk to at r merged to at r +The HTTP method is POST +update italso has this problem +for the idea +Awesome AndrzejAs distributed cache Hazelcast is my favorite +Fixed in revision +Besides for any query it could be interesting to know how long it took to parse it to validate it and to execute it +Thus it has a different content handler thing which actually handles the CDATA things correctly +The idea is to make overview without refer to Bigtable paper +That is similar to adding a JAR with duplicated classesAdding Hoop to the config file would be a point solution for Hoop but not for other FS impls +Next time could you please also add a patch for the documentation file? As Im not a expert could you please describe exactly what your aggregator does +It did not seem particularly useful elsewhere +Isnt that behavior better than an client retaining the lease on the renamed file blocking everyone else yet is unable to do another with the file because it doesnt know where it is? For daemons it may retain the lease indefinitely +I think making them strictly ordered is less portable across languages +Can you create a SVN patch to flesh out your solution +The guys can choose to patch or notIve created should try both KSSL and SPNEGO when authentication is required +Each database module in modulespluginjdbc extends those two classes to do the database specific portion of the testI guess you will have to explicitly tell the db to use a specific metadata table instead of the default one that might be a bit tricky +Collecting finish does require braces + +Need more time to fix this so moving to +I will come up with a solution to make it possible to some extend and more configurable +Unfortunately it seems that trunk or Hudson is broken +and so the additional classes which should be serialized cant be foundIf a is added then the problem is solvedSee the attached patch +please download the file and see yourselfWell thats what I did and couldnt spot the defects you describe +how does adding new libs to webcontentlib end up in deleting them before deploying? Files living inside your webcontentlib should never be deleted by our tools +Reapplied using v +If you prefer some other strategy just the looks good to me In the old code there were various dump outs sprinkled liberally over Throwable based on certain conditions that have proved useful for debug in the past +Alexey thank you for your support. +This is a four year old Bugzilla issue against Xerces +If there is a problem please open a new jira and link this one with that. +See for details +Ill raise a new jira for thatThere will be a change to the way single line comments are handled in this new refactoring +Severe Error Filename for extracted file wont map to target directory +This patch was used to resolve the issue +WDYT +OK +Could you be more precise? does the portlet retrieve the title from the resource bundleDoes it apply also to the main trunk +When a very large binary value is input to a LONGVARBINARY parameter it may be more practical to send it via a object +Committed to JBESBGAFP with revision to trunk with revision +A good place to start would be to look at the patch which Dag attached to +Since this is only impacting one or two fields Ill likely skip the work around and just live with it until goes final +The exact same thing is done there so it would at least save some duplicate code +I added you in as Space Admin Jacques now YOU have the ability to add others in if needed +Closing due to lack of response +Ill like to get consensus here first +Updated based on Sureshs comments +Thanks! Fixed in revision . +Hm +Local Eclipse project got corrupted because of unknown reason +I believe this can be resolved now? +Sorry for the misunderstanding I got it wrong with the EAP repos +So any well written LDAP client needs to be able to work with the fact that the schema may or may not be browsable +build failed on why? its second time I see this failure and it seems completely not connected with change +My test run for this patch finished and I see some new tests failing +after no further comment has been given. +Thats good news +Postponing due to CP related issues +Helious build is not ready +but that can happen later but cant I glean the absolute url by using a combination of script src path +It was fixed from other ZK fixes namely not asking for ROOT location every time +comment on the with you completely KevinHad not done a submit patch since I wanted to run all the tests first +Well probably go with something more like Thrift or +Are you still having thisproblem? If you are please respond by posting a test case +Info how to reproduce + +usability we need to a confirm dialog after user click the remove link +Alena should this be resolved with update to release a documentation bug and yes it will be resolved if the release notes get updated +Doing frequent compactions should resolve this problem +Ill make a new JIRA to discuss if that is a good idea or not +Pushed to +commit in masterCan we mark this bug as resolved then? wait a moment wscript should also be modified or rpmbuild will ecdbddfaeff for and wscriptbuildverified in my own build environment +Patch for addressing the for release +At the very least I remember this problem and dont actually remember it getting fixed +This should be fixed in ca +Ok +This should fix it +Attached apps for for both platforms +Verified. +I didnt check Smalltalk +Thanks Jing for reviewing the patchI have committed this. +What do you thinkShall we close this bug +ldap server is down +Is there a stack trace for the CNFE? We should be detecting if the is there and using our own if notAlso is this reproducible with of CXFAsked for stack traces and more information +I think this has been fixed so close it. +The specification needs to be updated. +laid down the framework for the TT to do this + +the patch in rev +Instead it leaves those blocks for the lease recovery process to prompt them +Martin can you provide a list of the ones that need to be updated? Or is it pretty much all of them +We have Project Verifier builder +Im just curious about something Jeff said earlier about there being more places than where the integration with Spring LDAP should be investigated + +NamitWhats your suggestion? which sub task shall we start? +please commit +But it is still expermental and has to be imporved +When theres a new feature in Hives trunk that we need how do we get a new snapshot uploaded? I modified Hives Hudson job to automatically run the and tasks which should hopefully push artifacts to the snapshot repository on every successful build +Can you please document this useful feature +Setting worked for me too +I believe the Reference Manual topic SYSCSDIAG diagnostic tables and functions should be updated to include this new information for the four functions named in this issue +Let me raise a couple of points about it If the bug youre reporting is legitimate then I wonder why is not failing +Under strace or gdb all works fine +Thanks for the submission +Just wrapping up AS reports +I use maven but it happened with too +The current patch contains user name which should not be part of the patch +The patch is a modified version of Todds patch for you be willing to please post a patch wo all the build changes that moved the files? Even if only for sake of review? Otherwise its hard to figure out what actually changed +Hence created a patch forWith this patch I ran derbyall in with Sun jdk on Windows XP +I expect it works the same on trunk but may be its differentLastly I changed some tableAvailable to tableEnabled in a few cases +Heres the implementation with test cases provided +I plan to commit in a day or two +We can track that problem by a different jira +Patch for commit from the patch on review boardThe only changes from the review board patch are a merge up to the latest trunk for a clean diff some whitespace and line wrapping fixes and bumping the RPC version number to since the connection header now contains the protocol interface name +This bug has been marked as a duplicate of +He has written a test which uses conventional machinery to get a connection +Derby fails with No current connection if some previous operation has failed for some reasonCan you set the following in a file and see if any errors are logged true +Fantastic! Does this include the WAR too +Reopening for review +Great but report Galder +I will change the param javadoc as proposed before I check in the patch +Id suggest something in between like Give the new template a name +Updated patch based on review as per some more Python version compatibility issuesIve fixed most of them but one lockfile doesnt work with Python +Unless range scans are used at +We are currently using POS and CCs as card not present +Very good idea +It makes no sense that maven should be looking to an external server rather than at the build tree itself +I filed a separate jira to track this but cannot reproduce the test error on my local environment +Id prefer not to add a new component type to since it wouldnt be backwards compatible within the minor versions +If we can implement our own it could go in otherwise we may want to pull in a better library for abitrary precision math + +This issue no longer produces an error as of r with or without STATICCALLS +Changed priority and description +This version removes some superfluous cleanup I did and adds Javadoc for Chunk and Chunker +It may make sense to use a custom Resource impl for this to make it easier to do things like insert in the middle +We might change each subsystem to use their own project code which would make this mostly irrelevant +Hello +All issues have been corrected in the EAP and EWP tags +the jnp or are you thinking about the webservicecontainter you are runningKnut In YES useridThe ability to bind JNP was added to +of course is the place to change this And I have done some changes to enable circualr references but its seems they were not enough +I think we need a separate JIRA +Verify that the thread cleanup listener is not notified +If you could make the bug reproducible with trunk standard fonts and no images we could find out whats wrong +I can send you the Utility source if you wishCan you clarify if Velocity supports Maps or not? I need a solution to this problem +Will commit when tests pass +Logs with Trace and new patch +The goal is to move the parsing code to. +Until that point and without this fix it will crash +Thanks Mark + +Yeah Isis has been broken a while we have just now got a replacement machine which I need to prep and build hang on a bit longer and Ill have it sorted for you +We should not put a release candidate out with at least putting the new user docs in +I just committed this to. +Attached the backwards branch patch +Actually DRLVM has been throwing not +Resolving as invalid the issue is too old to be useful and if the issue is still there it would have a different incarnation +Thanks Amar. +Thats the last iteration from me and passes all tests +The floats I introduced were intended to fix problems with some screens that had their layouts go wacky with the conversion to ems +The class will then log a warn message but continue to index the remaining itemsFixed in revision to the branch in revision . +ACK thanks for the patch! There is another place that references the supported drivers I attach a second patch here for that +comments +Fixed in the latest code +Thats what I thought +Now it outputs all the timings to a csv file which looks likeTimings grep select rankings select uservisits aggregation joinTrial HivePIGHadoopTrial HivePIGHadoopTrial HivePIGHadoopThe first line shows the queries followed by query timings from different trials +None of related issues will be worked on for EAP +Its not necessary + I wasnt sure if a diff patch should be created for a new class but I figured not +Filippo which jars is it that you want to add to the server classpath? And why is putting it in lib not right for you +On benchmarks it seems like about even performance with trunk code so the impact is at best very additional features bugs updates should be reported as new bugs. +Ive added the unit test as well to the camel module +Oh and please if you could document the mechanism for the way to add new partitions +It will be the responsibility of peoplewho implement their own Similarity then to take care of this in a similar wayIll open a new issue for fixing the. +Have you tried with L enabled +bq +See attached +vm file is updated and and then can runmvn changesmv targetannouncement Commons Parent has a profile for thismvn changes but this requires the vm file to be located undersrcchangesrather thanarcmainresourcestemplatesOK to move the file accordingly +If you see nothing in next hour yes +I posted on rb too +Still awaiting a response from Logan +Less elegant than could be users may have to wait but way better than corruption +i agree with Alex with this patch will no longer depends on Ognl to determine the property and class accessed +Will upload the patch momentarilybq +The running of a zk instance inside the VM when in local mode was incomplete +Another option here is to simply replace the call to ensure with a call to ensure +And this unit test in demonstrating the mixed usage of annotations in same POJO with Spring and Camel annotationstrunk +This is also likely to be backported to a Spring release which will probably be released around the time of there something we can do in the meantime? seems this affects autowiring across the the change minor enough that i can simply compile my own version and use that as a work around until such time that the versions that you mention are releasedthanks +Here is a sample +Manage Extended Properties new extended properties are the relational objects that I am allowing extended properties forTable column procedure procedure parameter procedure result primaryforeign keys +This would cause sending Bootstrap stream sessions to itself +I will cook up the patch once I learn the right way +Ehm I am not sure about this +This also has the advantage of not conflating client access errors with snapshot operations exceptions which I think are semantically different types of errors +I am resolving the issue for now +Somebody needs to poke around to figure the real cause and devise the fix +But I am still not sure how you expect this feature to work and how user would interact with such search. +Using a privatePath attribute on a probing element in your applications config you should be able to load both versions +If you have a script tag that is not of type textjavascript then injecting a into the element may have undesired consequences on the template +The comments in this issue lead me to believe that this is resolved +I dont think this is quite right +But seems to rely on the globals being set so I can not get my ASO before the call and if I try to get it after it is too late +This looks good Laura I just have one suggestionInstead of thisCheck with Apache for any updates to the copyright wordingChange it to this when you commit itCheck with for any updates to the copyright wordingOne tricky thing about committing web site changes is sometimes more pages get modified than you explicitly changed +See also +Ive just committed this +Behavior was discussed in details in +Three test cases were covered make binding address troublesome by changing the config file to bind two name node service address to the same +For this issue I will add Derby alongside HSQLDB as a random option for testing building on the new tests added with +Hi I can not find BPEL Project Examples in JBDS +Please schedule for a later release +But then we have to be sure for thingsis onblur always before onfocus of the next oneis onblur called when we replace the component if that is the case then those onblurs should be ignored +bq +Can you try with Camel trunkAnd the error you see with could be some remote server issue at plosone +Doesnt belong here +Any ideasCheers you pasteattach the stacktraces please +Unassigning this as there has been no activity on it for quite some time so others can pick it up. +This patch is intended for the trunk after is resolved +Patch against branches Ill commit this tomorrow in the trunkWould it be helpful for you to have it committed in the branch as well +Its simply a matter of getting demo app from right +fixed with patch from issue +I may be able to get it into RC today +Good work Ted and David +Basic Configuration Changes +patch looks good to me no more commentsI was surprised the sort avoidance did not work for such a simple query probably bad costing on the sort side +fixed the character +Scanned components now comply with all defaults as defined in the beans tag of the configuration file where the element itself is addition to this applies to autowiring and +Marking as resolved. +Ok I see your point well see if we can get this change in for agree that the dynamic scaffolding and similar code should work without adding a flush everywhere including the that flush should happen automatically before a I am seeing this problem in using and is sometimes but not always the save hasnt completed before the show happens and the update doesnt didnt notice it in so I wonder if it has become noticeable due to performance improvements +Other question I have is that you dont seem to fallback to mochijson? Or did I miss something +If we were out of fds we would see Too many open files in the logs + +My goal was to avoid the context lookups and use CDI Inject insteadI just noticed that I still hat a on the classpath containing username and password which seemed to get picked up on the second call +And indeed this is a issue with evaling yield to the supplied block for more detail +Need to get the server side client and server are working even with the Java implementationThe liked issues are related and have to be fixed those will be looked into post can be considered done now +Uploaded another new SNAPSHOT with several additional shutdown and exception handling fixes to prevent exceptions and segfaults as well as avoiding possible leaks +Hi ManikCan you tell me which groupIds will come from this repo? Just +But the statistics are currently being worked on so what is shown does not mean anything +The ARQ managed container starts the server associated with the BEFORESUITE event so once for all tests that run +I am quite sure that currently there is no method like createFilter returning Criteria interface. +Tested and working on m row dataset +Smooks editor is no longer being developed +Attaching a new patch that fixes the failure. +I see +Anyhow see my comment about the connected cannot expect that it will be zero if all sessions times out becauseit represents AJP protocol connection count and AJP protocol per se is constantwith disconnection treated as error in most cases +Attached a patch that removes lookaheadchaining +I am interested in hearing your thoughts on this I can add the code in the attached zip to the code base I need your formal approval to releasing this code under the Apache License version +Here it is +I plan to delete old and from branches +we need to fix that! No idea why it was done that way to start +I will address and and post a follow up againSome related responses Removed the spaces between the names of the functions and the parenthesis +Now that work is underway with AS all previous community releases are +Accidentally selected C backend +fixes a silly error in the last patch +Not a big deal +Rewritten to build against CVS +That makes sense +These appear to share the same root cause +done in svn rev JXR integration is only partial since the undelining program does not output line it does clicking on method lead to the whole class rather than directly on a method +Updated fix with a better varbinary mapping to fix +If theres additional work needed that wasnt solved by this patch open another bug +If network rate is mentioned in the service offering then it considers the value in service offeringThis also applies for the uservm default network +Code ported from the legacy +This is duplicate with per Ivars comment this is being resolved as a duplicate of. +Were trying to shift everything towards so all that stuff has to disappear at some point +I added two more test cases in this patch +v is much betterI tested with it and worked as expected +I am only seeing this same failure when reducing the sec sleep period that takes place after shutting down the cluster and before starting +Right now the only reason we parse the header is to translate xxx style destination names to actual destination names +Do you agree +Is this what you are trying to solve +Manually tested! +Committed test cases exhibiting the problems in this issue in Rev +I need to get better hardware +Done +Thanks for reporting the issue I do not see an attachment but I understand what the issue is +So my current patch will not be effective +This is also during a index closing then opening +I also refactored the admin css a bit to support using dd and dt with the same style as the statistics panel but in other place in the admin +I tried it on other layers from the same source and they worked fineI used ogrogr to create a new shapefile with the same features and it worked fine as wellThere seems to be something about this particular shapefile that is triggering the problem +So all issues not directly assigned to an AS release are being closed +Im thinking that we just make HTD and HCD be maps +What is the objection to the temp file +Possibly some cached stuff? No thanks for the fix +workaround updated. +It is only a problem durring the load +I would disagree +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Makes sense +the details are in m functionally the code that got rolled back because I suspect that this was not the root problem +I noticed that didnt have the svn property +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Fixed +Im wondering if we want to switch to it for to keep in line with an updated form of a java +Even if we change the engine so you can provide a class loader +For the future avoid the IDE to change the way imports are managed change newlines and avoid renaming variables in the code +Thanks +Record all dates in UTC +Any reasons not to do that +Something for the project to discuss +Ill look into it when we get a new regression report if were still getting the same problems +Fixed +But with my approach the logic is clean and easy to implement +If thats the price you are willing to pay you may want to extend the and make all its methods synchronizedWhat do you think? seems to me that it shouldnt be possible through proper use of the API to cause assertions to fail +I read Attached patch fixes this problem +This is also happening when I try to extend I get the mysterious from within Solr +I can work around for now but itd be nice to have +fixes the example code to build with updated patch cleans up the tutorial as well +Configuration in configuration file +Moving to JCL would maybe be even more consistent with the Spring flame war please +In the best world the date time localized works with the user locale but I prefere in first draft use a propertie to have a useable system +It seems like unwrapping Acegi Security exceptions from might be a good idea after all. +Hi Victor I was able to generate service archive from the project that you have attached with the service archiver plugin +A heartbeat response from the namenode is lost +Integrated to Thanks for the patch SergeyThanks for the reviews Stack Lars Ram Rajesh and Enis. +Indeed looking in the beta archive I see both cp and Proxool +committed patch in +Submitted fix with pull request from tmalaskapig github +For the time being code only identifies sequences +bq +bulk close of issues as part of making release notes. +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in +Can you test with my patch +Hi KevanIs it possible that the tests failing with update relationships before the instance is deleted? With the current implementation dependencies wouldnt be detected +Maybe the bigger issue is that this got out +It will be released soon +The project has been created and code has been moved +please verify on the updated ER build +group is not visible when there is no Extension installed +makes sense +The crashes no longer happen marking as resolved. +Verified in to set up fix verson +Like I said earlier this is acceptable to me +I cannot reproduce this +They are received by the broker then received by the consumer +I will commit this patch in some time today +Server side we need scripts in the bin folder of tomcat to support encrypting credentialsClient side the same time except the classpaths are Ive added the server side scripts +The best approach to this issue may simply be better documentation of the assumptions of +What will happen if we seting major priority between and blocking compactionThreshold may be like this for example blocking compactionThreshold we set major compact priority to If some region have or files its priority is lower than the major compact the major compact will do first +just committed this +I was wondering about thisOK +Verify if precompiling of groovy template could improve significantly starting me know for templates are being reworked as part of webui removal +Maybe we could split the immutable and mutable parts up into two different classes +Thanks Chris and Arun +Other notes Underlying String indexOf methods all take String as the delimiter +Hi Olegthe problem is not the +The same regular expressions used by inclusion and exclusion filters for build pathsThis list is to include scripts srcmainresources and srctestresourcesInside of I added a few lines of code that tests each file coming in to see if it is a script or not +For one there is no uniform way to build the project and it wont work in a assembly plugin already has all the functionality you need to package and deliver zips and tarballs +I am reasonably confident about that +This the shell script that will reorgnaize the source locations +Attached patch applied to trunk and. +Note Locks indicates Once a failure occurs the corresponding TIF file cannot be deleted manually either +Thanks +Resolving as fixed. +Lets change Show All +Patch for latest applied +I would rename the project in eclipse to something else +After patching with I also observe no more failures or hangs of after local invocations or more runs up on EC jenkins +Ive committed this to trunk but not yet to branch because Im awaiting Maxs input on whether to wait until after GA is to have patch but did we get any wiser on when this is triggered +The additional commit has been merged to eap. +Yes. +Rishi +Use the attribute description text to explain why undefined may be are precedents for that kind of thing +proposed tried to apply the patch but the class is missing +I also question whether the size of RPC headers is a significant factor in Hadoop performance +Later tservers complain that the files do not exist +Moving to since I already have a patch ready it involves only the oracle datastore and has been requested a few times already +You cant reconstruct Indexable by definition as it is and providedYour average Document may be one implementation but I can have an implementation that just reads stuff from the socket and shoves it into the index howre you going to reconstruct that +Patch looks good overall to me +I think this is with the new iOS SDK correct +Perhaps the term flow is being used for two different concepts here? Can someone expand on what is meant exactly in each circumstanceGiven that there is an actual Rules feature called perhaps the term shouldnt be used in the Rules documentation at all outside of that feature to avoid Conflict Resolution states As a general rule it is a good idea not to count on the rules firing in any particular order and try and author the rules without worrying about a flow where the order matters +Incorporated feedback from Arvind and Ralph +The Eclipse guys dont have that problem I assume since they use a different way of reason we do not use for example asm to read the class from disc is because we want Groovy to support runtime compilation +Ill fix the typo for than and commit +pushed to master please test and close thanks. +Sorry I havent noticed your comment from nd May stating where it was fixed +So in this first pass it is fine to mark those features while MR plans compilation is in progress +It starts with the default locale of the server +I like it but I do worry about the complexity for new users in having to use different interfaces for setting up the users of a system vs +Attaching which backs out this feature +IMHO the is a good idea +Thanks Thomas! +Attaching +By convention we do not suppress this warning in Pig +More like hacked around +I would like to make the error messages as follows Cause Error setting property set of +We have more inclusive header settings than what I believe you are trying to create in the proposed pom though I am unsure whether that will be a problem +Yes it does indeed +It is applied and when svn is back up i will commit itRegarding the panels stuff yes things have changed since this doc was created +This is already mostly handled +Ive committed the patch +Now it works fine! +We could place a link on the website to itI will move the other tools into a new project in our sandbox and create a lemmatizer addon module +You could be experiencing the problem raised as that is targeted to be included in AS +Unfortunately since this build does not run unless is set we did not catch it when running test and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +We will fix it on the planned on next weeks +Closing old issues +pom has already been deployed to the remote host +wrong component +Thanks RegardsAmol Verma +Right the ipc module has become the home of most tests that depend on the compiler maven plugins +This causes the applications to send cookies which expire as soon as the browser receives them +Setting urgency to None since it has not been triagedShrikanth can you provide some more information please? Which derby version are you using? OS Java versionCan you provide the log files andor the database files? Did this happen only once +Reopening to set Fix In to also. +The API in has been improved recently which allows a Service to be notified about a shutdown is taking place. +Fixed the behavior of and added regression tests +Closing this one +Delete your and run then coberturacobertura twice +I updated to the latest code +Attaching the correct patch +The slowdown is tightly correlated with the above commit +The open request takes the client host name as a parameter +Because our hardware crypto modules only send one line of response with a rterminator there is no next character +Fix as described in original posting of this bug +Should have noted this is the fix +However I dont think that will materially add to the build time +Updated the log formatter and added a test to verify this util from now on +Workaround add an to the start event of the receive task which copies the variables needed for correlation to the local executionThat way the execution waiting in the receive task also references a snapshot of the variables +Heres a patch which addresses your nit +Maybe modifing the getPhysical methods return type which should be the type defined in the WSDL file would be better +I dont think I have permission to assign myself +This is resolved is being resolved by the new deleted docs support +OK marking as fixed +The fix is in SVN +My patch to this issue +Ported to trunk +for the patch +When it comes to misspelled or exceptions formats they come perhaps most often from somehow hard coded or hand written requests and not by selecting one format from the list in +Thanks +Id suggest we discuss other mechanisms +If this is a preexisting bug then assign it to since is now released +per the JIRA meeting July G +Im missing any comment why this issue has been reopend +Heres an implementation and test case +This can be as soon as the mirror conforms +All seems to be working just fine. +This has already been fixed and will be part of stale resolved issues +Using compression compouded the heap charge already taken by several copies from RPC to Cell to etc +Changed build rules to make to be on platforms other than windows response assuming ok. +Ruel youll take care of it? +Alexey I seem to remember this issue from earlier WTP bugAnis could you attach the log so we can get the full stacktrace of those +another year ago r doesnt actually fix mentioned issue +Jonathan you are welcome +Dennis has status Rejected +please try a newer version is the latest atm +I rm the dir from tmp then start the master create a table with a single column family then restart the master with time binhbase master +Attached from ant v seems that richfaces and overwriten the specified version in build file +Want me to commit +I have made a commit to trunk regarding steps basically disabling the show in mbean explorer action unless the mbean connection actually worked +EricI can take up this one +It takes IBM about months to do that workHope this helpsRick +Cool any chance of hosting this somewhere convenient? We would also ideally need a namespace +Thanks for the suggestion! Ive added such support for system property placeholders for import location similar to the placeholder support for config that is correct +Patch submitted +maybe we can safely assume minor changes in the pom are ok +Extracted replaced all parseInt with radix argument and packed back +Moving all issues to the CR +I verified that the nightly build succeeds with this and committed it +First attempt at a patch this code is hairy +The fix should be there no additional javadoc warnings during +Claus when you say I am thrown off +fixed by revision Included fixing the bug in +Will? Did this fall through the cracks? +I updated my patch as follows I removed the counter logic from but I didnt push it into +Will close it after it has been merged. +I dont believe we can easily change the current design of and from what I can tell Grails now has workarounds in place +Need sync with trunk +I will start a cluster tomorrow and inspect the config files on the remote machines +not sure about it myself +After reading your and Dicks mail in the I understand now better what is happening +I just committed this +There is no way to distinguish being given a bogus host from a good host +Ill take care of it +I would suggest posting questions on the mailing list if you are continuing to have problems as this appears to be an installation and configuration error. +This one has been failing pretty frequently +nobody +IT! +Hi AaronI modified the thrift interface hooked up the controller and shard server and stubbed the class +Need to add the Debian Wheezy in the list of supported guestOS +The other implementation idea was having it stored temporarily in zookeeper +also needs to be applied to branch please +provides a workaround +this patch wont handle that +Adding support for enumeration type hashset +management server log should be fixed by commit bbadabedfbebeeecccbcee +Ben fixed this one a few days ago closing all issues that have been in resolved state for more than a month without further comments +The more I know jackrabbit internals the more I see that this kind of approach is discouraged +Now invalid fids are stripped before they reach the closing all issues that have been in resolved state for months or more without any feedback or update +Resolved + +Attachment has been added with description Build Glenn +The use of the word selected is wrong here now that i see it realized +However I am not that confident with opening the implementation at all edges +Closing issue fixed in released version. +bulk close of all resolved issues. +Change to feature request for need bit version detection so that error is reported if wrong one is used during postinstall step +It would be more meaningful to return an IAE in this case +It completed without any problems so that issue seems to be related to your setup not to the Mahout code. +Not for beta since the issue is upstreamNot critical since it only happens if your project is within the workspace which is bad practice anywayMoved to for investigation by Fred for fix in me core +Code checked in +If it isnt you have to provide more information than just a stack trace with no code configs what you were doing at the time etc. +Lukas is this possibly related to the resource optimizations we underwent in M +Lets add basic interfaces first +managment server blockers and critical issues are resolved but not verified +I Agree with After reinstalling Cassandra properly is working might had mixed the versions while upgrading. +HeyIs this patch in a state where it could go into? If its not going to be polished off please let us knowThanks JeffSorry for the delay +Only close once it is verified +You probably would want a switch in the module to handle the versions differently +Jackrabbit trunk uses Lucene +So at least over here the test suite never terminates +I have no way of recreating this problem +ports that server complains about are not jbpm but AS itself +The RIT that was got in the testcase should be updated everytimeWill update the patch and commit it once i reach home +MD and +Im not able to reproduce your issue +Okay here is a revised version of the patch +Revert changes should not be applied. +Do you agree +It it does by far not end here +Can we increase priority +I havent tried to verify the bug but it should be fixed for if it exists +I attached the wrong file previously +Thanks for the patch Nick. +I have moved this to M to allow more time for research +Thanks very much Dag and Rick +Will test and commit +Best wishes +Compression would be enabled by default +testcase for input + +It looks like there are a few more unit tests that need to be fixed +ini parametersLet me know if this patch is fully ok +not sure what is going on +Thanks Ryan +I think the only way to fix this test is to change it to actually get the first message prior to canceling the consumer and publishing the second messageFixed by moving get before cancel. +This is causing issues as mentioned in the chances to get this fixed +I cant read this patch +Some might need to be allowed to change others need to stay the same once an input or output stream is created +If someone would like to port the patch to please domy build environment no longer works and the effort required to repair it does not seem justifiedI am inclined to resolve this issue if there are no objections +Thanks Alex +review you are +The patch looks corrupted seems that you were moving from top level directory a to b or created the patch from the wrong directory +Added SPNEGO support +Thanks as far as I can tell this fixed our issues! +If the node doesnt have a token it doesnt matter if its a gossip member it wont be part of get and wont be hinted anyway +reverting now to and undoing the release +The attachment contains all the Xerces files affected by this patch +I have only been using FINER as the logging level since you created this after i removed the contents of the coverages directory it is working +The saved file is the printable version of the temporary wiki pageWhich html does not show the links all items in the table of contents are linksReview is appreciated +If it turns out to be something quirky that is doing please reopen this JIRA +Central will use preferences set in the Project Examples preferences pageWould you like me to add a link to this preferences page to CentralWe could remove the Seam preferences page icon that isnt used anymore and add a Project Examples preferences page icon +Forgot to mention zip file contains code that reproduces issue with aggregate reports causing release failure +Do you think I should move it to contrib? I just wanted to park it somewhere in svn so that after is release I could reference it from the Accumulo web page +Should be fixed in +I see frequent concurrency errors due to the dispose method not synchronizing on the deliveredMessages linkedList +Patch Fix of compiler warning +Ive updated in commit bb to fail upon connect if the Derby version is less than +Test source took it +Please open a new issue if this is still a problem +Thanks LauraMy to commit patch +Apparently this issue is actually caused by Subclipse +Hi Nick for some reason I forgot all about this patch when I integrated your previous work +bq +Ill see about adding a test for it. + +gtr invalid backslash escape at end of stringecho i tr sed s +Many jobs override defaults in parent pom to pass in different for composite sites or parameters for testing +Passes unit tests but still fails in Solr ville to generate sort strings +And can you provide an sample route or what else to demonstrate this +roger all issues that have been in resolved state for more than one month without further comments to closed status +Imho its stil there as I see some warnings when using it myself but I cant help you as Im also quite confused about how the abstract classes are managedLooking forward to understand more of it from your fix +Attachment has been added with description transition resolvedfixed bugs to closedfixed +I think that they are totallyFractional dimension norms may be useful in some situations but this paper really doesnt say anything useful on the topic +We are glad to hear that however in future it would be better to use Sonar Users Mailing List before creation of an issue. +I dont think it was fixed in or if it was I think it has regressed +closing since is now released +I actually addressed that in but it hasnt seen the light yet +Also the DDL page has more information about options they support with examples +o Camel o Camel NMR o Camel o CXF Camel NMR o CXF NMR o Loan Broker Bpel Please note that several new cxf examples do not include files as such i can not determine if they are testable under current capabilities +What are the different use cases for this and +You dont need a testcase for this +Maybe only reset if the current value matches the value in the class +Verified by Nathan. +Thanks for the reply +Nice so you want to be able lets say to distribute things under and we should let you just because you like it and its possible with package names +a multiform to show list of skilltypes and its updation in partymgr and List partySkills in HR +Sorry I hadnt hit watch on this one when I got here last time so did not get your request in time +Daves analysis seems right to me marking this bug as invalid accordingly +If we want the improvements separate issues can be opened up. +Backported patches a and a to the branch with revision +Cool nice feature addition +Hi Davanum I want to know what fixed this issue i am looking at the comment and it says try switching to how do you make this switch +Propogated as an removed the deprecated method since we can still break the API in this alpha state without affecting other software removed additional logging of parameters in the site mojo. +Or just use SELECT FROM list +why would master think its on A when A already sent long ago +It looks like not much has happened in the last year at least regarding +I found a fix that worksIf you change the ownership of the directory varlibcassandra to cassandrasudo chown R cassandracassandra varlibcassandraThen Cassandra will be able to create the directoriesWhen installed from the Debian package by default varlibcassandra has these permissions root root So either the directory has to be set so users besides the owner can read and execute in the directory or the user that cassandra uses needs to own the directoryvarlogcassandra also has this issue as the complains that cant be createdThat issue can be fixed by runningsudo chown R cassandracassandra varlogcassandraSeems like it should be an easy fix in the Debian package +the fact there is a +Merged into the master and branches. +but I dont think changing existing code is worth the effort +Changes Fixes a possible connection leak in getStream method +This was resolved with the patch for +Done +Hi jiraDid you find any solution for your connection timeout issue +Just wish I could recreate this +For the LF the output still breaks and two fetches have to be done to get the extended plan +Please respect our guidelines and do not create JIRA tickets on your own +and +We have several sortable overviews in our application in which the user can search these overviews are backed by Compass and the code that implements the sorting and searching is pretty much abstracted +Pushed to master +So we should have a way to convert between the model type that we need for the renderer and the that the useYou could register such a converter on the like the normal converter businessConverter +That is done its just the +Inject it via Context. +I have added the docs tests for the new parameter +Apply the patches from the geronimoapplications revision +The field actually means all the brokers in the cluster +I will say that I saw this issue again yesterday morning when I logged in +Fancy that. +To be applied on +Postponing to the next in. +Dot releases are for bug fixes not new features +the mrCommand function I dont see how the options list is helping much +to applying it. +attaching that was missing from patchset + is based on that is build for a single endpoint +One of the new files was missing the license header and docs +New Version includes all like we re missing the +Both and now use so consistency is achieved +Changes +The best information that we have now is that this occurs when there are inner classes in the file +I also set the and system properties to fully match the geronimo start scriptserver starts successfully +It really looks like the other one +Creating a new data Saved information on Saved information on Saved information on Loaded information on Saved information on Saved information on Error reading file homejdoecontinuumwork invalid stream headerCobertura Coverage data file homejdoecontinuumwork either does not exist or is not readable +Try running dosunix on the patch and then apply it +Were experiencing this issue too it happens quite frequently during our integration test suite when web tests are performed faster than a human user could possibly perform the same alwaysReauthenticate to false is our current workaround +Was the fix on ported to? Which reported issue dealt with problem exactlyThanks Parameter directory is requiredDirectory containting property filesIm only a user but when I considered this further and have since used later versions of Geoserver +Maxim thats right we did invented the bicycle +But I do not have a OSX on hand could you please help to review it? Thanks +Between file A is iterated and is opened on A a compaction can remove A assuming no one else is reading A at that time +See mine above +I guess Ive integrated all changes that seemed fine to me +Uploaded a patch that makes it possible to use Vendor API even when the Vendor based connection does not support the is or unwrap ve delivered a modified version of the patch which does the best it can to allow the transaction wrapper to function on a Java VM +This probably will not be too hard to resolve once investigated furtherCurrently my idea is to have one bundle providing the james and its dependencies for our fragment and another bundle that gets its from our fragment but Im not sure whether this really is the right approachI certainly will have another look at it +I often store objects in the cache that have as a cache key +Ive encountered this in an attempted upgrade +This does not work as the dependent feature is not also uninstalled +Applied suggestedPlease test this fix in your environment. +This means doc may be reordered since Tiered MP can merge segmentsI think this is a very hard break and it should depend on the Version you pass to IWC +So Im just going to close this as Cannot Reproduce Bug. +Dominik again this is an old JIRA that describes what youve been doing +This just calls for a custom simplified parser that would read the header and forget the rest +for example is not +It is probably a good idea to take this discussion to the mailing list at this point +I could probably arrange a test of a patch in production +see attachment patch with some minor changes and one bugfix +Could anybody edit this issue and delete code stuff? Ive already attached it +vfsfile is recognized as a file in a Jar and its handled that way +Copying var is the mechanism for creating multiple server instances on +I will work along that line to make the loading of ecj optional except for specs for which it will be loaded in sandbox but disabled for guesscompiler and reenabled for the its specs +Hi NicolasCould you please review the patchThankful +Tested +Ive committed this to trunk and +Two remaining files to fixtrunktrunkbuildresults +Actually no it isnt just loaded on startup +There are no actual duplicate classes just split packages +But we also have to consider the risk of foot shooting for users and Im not yet bought on the fact that its acceptable in that case +Working now +to commit. +For example if we change the quorum port we might get into trouble +This is for the for the patch. +My output using trunk for Toms case matches Ruby +The hbase component does not provide the ability to forward any data from the to the except for the hbase specific data +Will commit when tests pass +Cant seem to reproduce this on my hardware +I agree with this course of action +Did you try it with in the end +applied fix and because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I imagine because the forking of the vm for the tests +I had started fixing this locally yesterday +Marking for and +Committing in a bit +Added dependency +transitioning is a mouthful though it goes w our region that is all over the code base +Thanks to the reporter and to the contributors of the patchD + to patch at trunk +Sorry I overloked must be issue with qdox ordering of ll nivestigate this +Thanks +whats the status of this issue? Its still set to BTalked with Burr about his case his issue has more to do with timeout handling and dns not being available or detected in a timely fashion +Please commit the fix if there is any +trunk r and x after release. +The reason is response message context is not freed in case of sendrobust +And we can also insert null values into the unique columnDerby reference guide mentions this about UNIQUE constraintSpecifies that values in the columns must be unique +Thanks for your help +Consistent with the move away from the abstract factory pattern that we are making elsewhere in commons math we should deprecate +Thunderbird should behave reasonably provided that RFC is deliminators are now implementation specific +Committed to thetrunk as revision +This seems like a recipe for complex bootstrapping issues +Yep its a fairly new development see +See footer image is not rendered +rebased patch probably wont work either thoughthe error is quite strange had a quick look and the files should all be closed before renaming +Wiki text got the better of my last comment +I can believe that reuse is still not working optimally + +Seam does not have it +I agree on +patch against the branch attachedThis does a mark on the buffer saves off the token value and then resets the buffer to back to the mark +I agree that I dont think it makes sense to use different names for the binders +Now that I finally got a chance to look at it this it turns out that this was fairly easy to do +EAP +We could of course change the brand to include a direct link to the KCS article but I dont think its worth the effort +Fixed +Update the implicit conversion of datetime values to CHAR VARCHAR is available in Derby also in call contexts +This will anyway will get skipped later while replicating right +Any way this can be addressed separately +JacopoYou can close it I will submit a new one issues like what youre said if I try to implementing more macro +A fix has been applied to +That should do the trick +I just added a test for it in camel trunkIt looks like the problem has be fixed. +bq +This is only searched for under classpath root current working directoryIf one placed his under classpath Java will not be recognized any more +replaced log with proper log +Attachment has been added with description Test +I think it is better to take as default the version of of imported project +Please put your thoughts there and close this one as duplicated or provide details why this issue should be tracked separately and how it is different from +Set Known Fix and Newcomer assuming the suggested fix is the right one +Reopening the bug to add in a patch to fix the tests +What is usually suddenly jumps to few hundreds and some in thousands +Thanks Simon I appreciate your working on a target +and the implementationso that means you basically inline your dependencies within a config file? why is that necessarywhy not doing it by simply using a repository to create those descriptors rather than hardcoding all that in your own ivy filewhat is assumePublished forif its needed do you really need to set it or would simply the definition be enough +Indeed the numerator was correct denominator was not +I didnt fix any such warnings and I dont see any such warningsDid you look at the updated patch I uploaded? Do you agreeunderstand the problem those tests had before and why the cast is indeed redundantCould you also let me know what parts of the patch you decided to omit? Ive spent a lot of time clearing those warnings and Id hate to see a large portion of them come back for no good reason +will test this issue according the attached specification +Revision . +By the way the API could be still be using a float +Hello Morristhank you for your fast answerI made test files that produce the same problemThe real input XML and stylesheets would have been too largeIf you cant download the attachement then writes to me and I send the zip file to youBest greetingsDieter dont know how big your document is +Passed +Patch committed to trunkThanks for the contribution Vivek! +I also agree with having some service support from the beginning but this should probably go on another issue rightIts hard if not impossible to design a general interface without having a concrete implementation so its better to write a simple implementation as a part of the initial JIRA IMO +I think considering the schema with a possible configuration override is a very good solution +Chris thanks for the patch +Resolved as duplicate +The patch extends the class to hold classifier information +Made testsuite sping quite a lot +Committed patch with workaround in revisions and +I think we should take this patch on trunk +GDM and GDSM play a important role so if we find a pattern or code is really quite useful why not add it to Groovy and share our effort and wisdom with otherspsYour code is more groovier +I think that is fixed in with debaaaaecdcbebdeba by Leif in master +The file lists this issue under the section titled NEW FEATURES +Please use unified diff format to submit changes +All calls come via the boss now +Just to be sure if the direct integration is off then the problem does not appear +this could be a reason of the patch fix the issue +Im not against this but just a note that some users have scripts written that depend on the order of the ls resultSome even failed when we just added a timestamp +closing self reported issues resolved on versions already released +Thanks +Please verify +does the string need to be formatted as Ljavalang +I have updated the protocol wiki to note that these are not available in +There are other test programs that use local viz +I think we should have already managed to get rid of adaptx as a result of some work by Martin a few months ago +Thanks for the help +Is there any nicer way of using number of tasks? for moving the server out of JVM +Created to discuss ways of making the proposal application more +Once you delete the resource adapter recycle the server and wait for it to restart and continue previous operation +So we removed the mirrorCan you set up an alias like mirrors we want that our mail is sentto people who really read their mail and respond +i thought I already commented this should be fixed afaik please validate +process tree is standalone component for some time +We typically at the time of the release +Thisis not the case for env entries +patch looks correct +Adding backward compat as suggested in CR commentsChange some comments rebasePhabricator review updated wcorresponding patch +Looks resolved to me +on I have committed it +Updated the patch for trunkChanges Rebased the patch to trunk +both inserts and deletes are committed in the same transactionOK cool +Fixed. +Please mark it as a blocker for if required +The exclusion fixes the for pointing this out +Its providing log monitoring for massive indexing rightI think we can go wo it for a while +LGTM to trunk +I get what youre saying it is odd but its what im gettingIm running my app on an embedded tomcat because im deploying to Heroku and thats how the web app gets deployed there. +Instead it relies on acquiring the aiomutex for every request enqueue +Thanks for patch Jon +geronimo properties and cannot be set to function has been fixed via trunk will be investigated later +This patch merely removes one unused import under the tests module +thanks that explains it +Bumping urgency +Jacques I am sorry for my commentsWe are attaching the patch in which we implemented the Straight Line and Double Decline Balance Depreciation algorithmsTo verify the algorithm we prepaired the data for it +You cant call shutdown when libraries are in Tomcat shared libs because calling that method will shutdown the shared thread pool and all other web application deployed to the same Tomcat will be unable to send any JMS messages +I still need to add some +reopen to edit +It should go to as well +I see you were using +Logged in as root +The patch contains the diff against the +im not sure if throwing an exception by default is a good idea changing the default behaviour may interfere with for the long delay to get this added +Unfortunately looking up dependencies in children and in particular creating the bean in a child factory would severely affect these semantics because both the parent and the child would hold an instance of the bean in question without the child having a definition for that bean +It has grown to a very large patch and changes a lot of stuff in core too +carl formatted the describe output through describe formatted +Currently the Fix Version for this issue is None +Please check that it applied as expected + +Corrected the typo thanks +Great work thanks! +Zheng +Namely that wed need support from the Microkernel if we want to be able to control the life cycle of binaries + +The higher I set the session timout the sooner I get the am sure of that the problem is not that the session has not timed out yet that the number of instances are growing +verified by Tony +I also added logging of the full bad parameters +added to install it to get the behavior +I havent test this myself but I also think this wont happen in latest releases please test with too and let us know your results +Rafi I wasnt aware of the LIBDIR macro in Cmake +Mirror is online again +Constraining the queues memory usage should block message producers for slow brokersHave you had any success with memory usage or disk usage constraints that will cause a send to block pending space? +I will committ this patch tomorrow if there are no other commentsNicholas Brandon do you have any comments to address +Thanks Thomas! +I am going through them +Here is the rub I would be happy to wrote hundreds of test cases if I knew someone was going to actually help me get this done +Derby will not try to see if it can find any other privilege to replace the one being dropped so the object does not have to drop itself +User can always use the two events and find out the cases which task is deleted without completion +Added thanks will appear soon. +verified against r +People will use any of whirr pallet chef or other CM method to deploy kafka I went with pallet +Great that makes this patch very simple +If someone external opens a bug about this we will happily revisit. +The real test we should be doing is that after the patch NM is not accepting the same event say more than once +I have a couple questions +I did this some weeks ago missed to close this. +Were looking to upgrade that to or +My thinking is the followingFor a given proxy object when we first have a successful RPC we can set an internal flag in the failover proxy provider indicating that it has connected once +Vincent Ill apply the patch this weekend to the branch and trunk +There is no support for this in at this point +Committed to branch and trunk. +Resolving for +It dont seems to create the inboxes folder +Thanks +Lets documentAddedbq +It looks fixed in the lates beta build will verify and close the JIRA with the final beta bits. + +We could hook into JSF DOM update process and replace reference in time of updating JSF would probably bring more issues than savings +Major compact will call bloom filter builing when doing merging +The deploy and install plugins use this for goals that are meant to take some random project and install or deploy it like rd party jars +Hi Yip thank you for the patch! I will have a look at your changes this weekend +binary is largely unchanged here +This updates the to a more recent version that does not add the case insensitive duplicates +I believe in this case the text is wrapping when you type in more text or delete existing text +Thats a good idea +Correct + +could you please test? All unit testsintegration tests pass but they dont try with +Thanks Yan! +As you noticed when running Igor +This will also make the delegate fields final +Repro page flow +bq +patch that adds a check to assign meta in cas the meta znode is null +Reverted the accidental proto changes +This issue also affects the Java backend in that some flags are processed upon loading of the Configuration class +I was unable to measure a performance difference but it is less code and uses less memory so it seems a safe change in any case. +GC now uses the Trash if configured +You will not get any error at the backend and Xbaya +It will be better to reboot this discussion with a questionWhat additional features is missing that you need +Thanks for the guys who work on this case +This also fixes at the same also add no unit tests because its just straightforward null checks +My last patches passed this test tho +The only bearing snapshotting has on cleanup is that the snapshot dir must be avoided when cleaning up +Right ok +See attached test case +Regarding the goals I will explain those on the list later this evening +If I cant Ill send you a public DSA key +is it? we were going to do some changes to BCEL to clear up other issuesGarvin + +None of these affected the meaning or intent of the document +I separated out theclauses expressions join operations and query informationPlease let me know if there is anything that is not in the proper location +FYI I suspect is responsible for this regressionThis is a showstopper issue for us +actually I see this is just internal should be ine +Ill build an RC +Fred thanks for the explanationEverything works fineVerified in +fixed in revision Dmitry please verify that it fully resolves the Mikhail everything works as expected +If I build the webapp the code gets checked twice and deployed twice as is identiy module +First patch for New tag update for samples to use patch agaist HEADNew task for and update on samplesservletsrcscriptsshare to use cargo with cactusPrevious patch uses incorrect files +this might be a point we want to use the wwtoken which is supposed to allow you to prevent double submission in webwork +I will retest on and try to wait for reclaim of the disk space +At moment the ipc Server use is suspect +After removing the duplicate entries and the schedule example runs fine +Committed to trunk branch +Im mistaken it does use the proper subdirectory +If the plugin isnt a singleton P will load all the versions of the plugin +artifact works in both Maven and Maven +That seems to exist in trunk only so maybe the reference should just be removed from or we add in too +similar feature already exists. +One more option which is the one I actually implemented is to implement a dirty bit on the application so that it is aware of modifications that should cause reindexing +Resolving see for more refactoring. +Is that what you prefer +New test processorinfoerr will cover this situationIn processorinfoerr there is a declaration for the prefix so its justan unknown property +So wed have to migrate to a table to allow maps +Hi MaryI dont think this is a issue since the forums feature of the forge has been totally removed for now +Tom can this be removed now? I think youve since removed its use in that QS right +You can try that out and see if that resolves the issue +The resource adapter can now be configured with Wsdl and properties to use a specific WSDL serviceport +A fix +I just fixed this +I added support for the follow atribute in revision +If the plugin has dependency jars do you expect it to be a part of the jar itselfYes Im thinking something along the lines of Hadoop Job Jar +Works fine now +It looks niceBut why keep project reports in the home pageAre they so important for new usersTo me they look confusing I would put them somewhere else deeper in the site +There is a serious bottle neck on obtaining the size of the display area +Multiple can now be deleted in the Build Results page and individually in the Build Result page +I will run it again and figure out + Bug has been marked as a duplicate of this bug +In Hadoop stack it can be installed only when Hive or is installed +Updated patch that also removes some blanks from the applied on trunk thanks for the contribution +Patch against objections against backporting this improvement to the branch? +They are statistically very rare and if done right it will not affect performance of BMP content +the nfs mount point mnt Good Can write to mount point is properly mounted and this mount point is used by ssvm as the secondary storage nfs mount point +this includes manytoone composite fk and but the declarations has to be in the same order as the declaration in the componentI still need to do a to work around that +Thanks for reviewing ColinYes we should change maxCapacity to MAXCAPACITY as usual. +Attachment has been added with description Test file which causes the to reproduce in trunk marking fixed per JM comments below +Some questions in my mind Should check be outside Do we need to visit defaultValue in or Is there a better way rather than pushing defaultValue into Should we be handling multiple assignment cases better And the big TODO of dealing with local variables that dont have an initial patch excludes method parameter logic as that is already there any plan to fix this? +I know Java very well though +if this is really an issue it should be fixed in +If we go that route we should use something like Googles Caja framework +I can find this library is in SUN distribution from of JDK +the view extent is something not coherent with the imported shapefiles +The fundamental problem here is that we cannot reliably determine whether a given resource corresponds to a directory or not +TatuI could really use your help here +From the description it seems you just want to know how to have Spring inject your actions with properties +Also the order should it be of type integer instead of string? I think you should combine condition and mask on the same panel +In either case it might be quicker to make the line change than to apply this patch +Fixing Affects and Fix in Version +And when the patch is committed please put into the commit comments so svn will back link to this pageActually if you put in the comments JIRA will automatically create a link to the commit in svn which allows you to see a patch +Ive committed the new patch +It was actually to switch the browse interface to not use +I would rather see it go in to a not so ideal place and then we can worry about splitting it up later when the rest of the documentation gets the same treatment +Over time we hope to make embedded much better and more flexible +I am unable to the issue +The only method that does is the log because it already has two lists and had no room for the permissions in the object +I will review a bit more and commit in hours +I have done as you say and couchdb seems to start nowI must say that the problem was happening because I was starting couchdb as a normal user but the directory usrlocaletccouchdb was only readablewritable by the couchdb userls lrtd usrlocaletccouchdbdrwxrwx couchdb couchdb usrlocaletccouchdbI got couchdb to start by doingsudo chmod usrlocaletccouchdbSo that the permissions werels lrtd usrlocaletccouchdb couchdb couchdb usrlocaletccouchdbI was hoping that couchdb could start by reading the conf files in usrlocaletccouchdb even though I was explicitly telling couchdb not to look thereThat seemed to do the trickNow to repeat the test with your instructions I set the permission back to but the problem did not occur againNow I can not get the problem to happen again with e or with ex with or with +My suggestion was based on other similar guis Ive seen like the windows services control panel which use automatic manual disabled +It should rather say round to N chars ie +Thanks Derek +If I use version that uses Anonymous channels the Master and the message received is null +Bumping the priority +fixed in trunk +running test +Its something I hope to improve before is releasedYou can try the unreleased version in SVN HEAD or wait a few weeks for the release +ja is very important +Latest update to the docs update API to sync with latest updates on qmfv wiki add tagging to attributes +Vlado can you verifed it using tools beta to see if the issue is still exist at your side +Hi ChristianIve committed your fix +Most people will need business logic added +added the use of in the filter method merged. +Added for andNote that there is an issue with described in +I need to fix our tests as well and it would be great to have it before another build +Ive updated the unit test to have the workaround so the test will no longer fail +Patch and documentation all tickets +So the patch now includes and +Thanks +Fix committed +Hi Tuomasthx for your work so far +Also when shutting down Datanode make sure to stop and join all the BP threads and RPC proxy +Fixed +as per Dogacans comments +Should we kill it +It would be great if we could keep one review request and update it with new revisions +The outcome could be skip the record stop the whole process +You can then publish it from this draft state +An easy approach is to generate an ascending set of keys and positions apply the MD hash and add them to both data structures and compare the RAM usage +apply the patcho +Rohini and Mona for your help +Thank you! +This code has not changed in trunk since the release +If we have an average of geohash fields per doc and we have M rows how do we set these cachesWe will get back to you on when it occurs +I think its safer for Jerome to comment if this would cause any harm +Max I dont really understand why that would be a problem at all? Users shouldnt be using the POM for their builds +Any coders in your house that can hack some on this script +We can just put the lastmodifieddate of the file in ZK +Hi Abner I think it would be a good idea to create corresponding issues for each offending plugin compiling all the or +Its up to Konstantin to commit to if he wants he has a committers to do soI know that +Nothing to do with the WS binding this is the old problem of support only working with remotable services when the arguments are way to fix this for now that i think would help a lot of users would be for the interceptor to not be inserted when the java component is wired fromto a binding or just the axis binding or when the databinding is different or the axiom databinding +please provide a testcase or quickstart +I checked out the source made a change to check javaProject for null and did a build locally and that seemed to fix the issue +Net clients which were removed from the repository months ago and not the current +The workaround to strip the document type declaration is good to get us past the release +First is this replicatable under any server or only under server? From what I can tell this is only replicatable under server +Is your ban also true forNo snapshots are fine staging too its the release repo we should not pollute +Can you make sure you have full readwrite permission on that directory and that the process that is running the batch job also has? We have uint tests that check that this works as expected in normal circs so the only thing I can think of is file permissions +But for very complex cases the patch not a best solution +Ok I removed the counting fix +yeah correct but its just an initial test devWill take care of this in th enext will be removed once we ship with the possible idea is to implement a registration promptdialog on the On first launch we present the dialog and ask for username i was trying to implement this and it looks like the enroll endpoint is secured so currently you must be logged in to create a either we change that which could be a security hole or we try creating a user with a script that a user has to the august timeframe lets just go with a default admin user and force a password change +I was able to reproduce itWill work on fix as soon as I can +Liaise with Edson Tirelli to update this to handle index configuration for s important that we abstract System properties away from main code via these Configuration objects which have hard coded defaults that can be overriden by either a system property or a programmatic setter +It isnt the most perfect or easiest solution out there but you can use the file from the cvs module to render the Velocity site in a printable formatjon +Thanks for the patch. +go do your vodoo +However in an appserver environment why arent you just doing postprocessing via the JPA contract for contributing a? We are trying +For example seem to recalculate fairshare based on how much resource demand is passed by applications +correcting fix view selects the first item +They have escalated it through to Cenqua who are looking into the problems we seem to be continually having with the mule and grails d wager therell be an upgrade that partially fixes the problem over the weekend +NathanThank you for trying out the patch +Version has correct sequence of elements +Justin I think at one point we had an export configuration command in the Wicket UI prototype +Hi whats the status of this issue +Galder yep. +For this to be done following things needs to be done +The RPCommand will be configured with class name of the locale resolver to instantiate and use on each request +I will therefore need to change the artifactId of the archetype to archetype now is properly named issue is now fixed and ready for review +To you it may be but not to everyonebq +Ill post back when I have that working +The ACL model is also fine though it is tedious to specify that and pass it to the +AaronI believe and related change how work +Added Integer Long handling for just in case +Todd +I hear you have some SC slides on this topic +copy +Thanks a lot for this great patch with unit tests! for the slowliness of applying the patch. +Josh I have made changes to to copy the file to cache +template files for current filters of this nature +The functionality is implemented and added but most of the code was copied from the +Patch attached to remove assumption that replay message contains message headerThis removal is fine since nodes already reject streaming from different versions +Do you plan to contribute this improvement? If not would you contribute a testcase? thanks +If you want to make the access to handlers thread safe you should synchronize other methods as well +This has something to do w rev +Thanks Mike +This was due to an invalid assumption in the prepared plan logic that the analysis record would have already been set +Im not sure whether the patch is done properly since it is my first time +ps I dont know if this makes a difference but I have boundproperties enabled in my castor build properties file +There is a jca test in qa when we run hook up the tests that Bill created and add emma to them. +It works as expected +Ive just tested the attached war on the latest trunk version and the portlet title displays just fine +Then we can close the issue. +The main problem now is a performance issue +It would be nice if the could print binary without printing jargon for values +It looks like this is a problem with the branch +Please verify and close. +I encountered the same problem and found that setting LANGC solves the also found does not work in environment either +The example using pax web just for historical forget to say it was also committed some fixes to make myfaces code more friendly with osgi including add a header so bundle can find using the bundle classloader +If commit is required all the output files will be moved from task output dir to job output dir and after that temporary directory will be deleted +Excellent +Thanks for the review Devaraj +justa note I would find unfortunate to implement a custom own user management +In additional to that problem and produce a dll with the same name +However this would be a much bigger refactoring with a much bigger risk IMO +Discussion on the process for adding language profiles to TIKA should be continued in I have a plan to add profiles for the Norwegian and Sami languages when time allows +Jessica can you check this? It might be fixed with latest build fromNot able to reproduce this issue anymore. +Yes I think this is fine +Will upload as soon as it is ready +Shouldnt the default behaviour be enough for thisYes we do not need to make any changes +should really stay away from implementations are much as possible logj avalon all have this feature +You can close the ticket +Is this patch ok to apply? Your last comment seens to indicate it was not ready +this ticket has the broken link in the attribute Bug Testcase +Cant we at least deprecate the cleanupJob now? And define a new one like commitJob that cleanupJob calls +I am looking at the issue to see what existing query plans will change because of the patch +Implemented by +Sebastian can you take a look at it? I dont know german and can only merge it as is +Attached is a patch adding this +Fixed on closing all issues that have been in resolved state for months or more without any feedback or update + +Hi All Please find the updated patch file +If there are any other issues that stem from this fix Ill be sure to report dont understand why it tries to drop the table GROUPS +What do you mean concretely? The maxmemory proprty is not set at all by default in the javadoc plugin +Is improving the data locality improving overall job performancedecreasing runtime? What is the other is it rack local or off rack +Committed the a patch to trunk revision Committed the a patch to trunk revision +same old WTP validator bug or? did we actually create a eclipse bug these kind of issues. +Option renamed to to be consistent with the message +In our case if the logj jar is in the containers class path then every app will initialize whether it actually uses logj or not and whether it has a configuration or not +Fix has been confirmed by henks fix +is there any tests i need to include here? or some examples +Attached patch implements the support for custom bootstrapperIt also updates snapshot test suite to use this feature +There is also problem that the datanode has written the data but yet not reported to the NN +Fixed in +Replaced occurrences of with Date +But maybe I just do not see the forest for the trees right now. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This is then available via create in or getConfig in +I dont have one +Alternative patch that adds the serveroptions directive for setting advanced options on the server socket and deprecates nodelay +Personally Im to commit +the control is editable but its covered by the latlon bounding box label +I believe this is stating that you cant update the relationship from the other side because if you try to do that you get This relationship is managed from the other side +patch looks good +Test reply with image attachmentJeff +I created a version of the header based on the project banner you used +bulk close of resolved issues. +working as expected. +why form attribute was not added in but only in? +Rebasing post changes to Srikanth Sundarrajan +Its dynamically generated on installMaybe theres a way to enumerate all the special folders in Livelink and then search for Recycle Bin but I would have to research that +I know that multiple providers besides EC allow you to stop a node and have a separate stopped state so adding a new STOPPED seems reasonable to meAs far preserving the old mapping I think the attribute should be called oldstate or something like thatOn a related note when this change lands in a release it should be documented on the Upgrading Instructions page because it could break or change the behavior of the existing code which depends on the current behavior +Works for me too! I echo Matt Bensons request +thats why I suggested compiler option but no pressure +to Erics suggestion +Thanks! +copyOf attribute handles this now. +I did a really crude trial of adding the SUM keyword to the definition and handle it in the same way as COUNT is done +Working one Im not sure about? +Can you post a complete sample app that demonstrates your consumer code cant tell from this snippet what your client is doing +Marking noob in case some one wants to have a go at it. +This seems to be MAC OS specific or Idea dependent? I just tried on windows with idea and is not crashing +Any progress on getting the new patch based on liblzma ready +Do you think I should commit what Ive fixed already? We can let Hudson catch more of these its funSimon I will fix as you suggest just noticed its internal +I am not sure yet they really fix the issue at hand though would be nice if that could be confirmed +I have confirmed that this issue is happening with other quickstarts and is not specific to +If i try to grab everything for the build config which extends default I dont get the dependencies for webwork +fixes the streaming tests to use ranges +I changed to use and it fixed the issue can be closed based on this new info +Regarding the simple jdbc driver I will submit a patch next week +Does it return or throw? good point clarified getLength and read methods in svn r +Verified on. +Havent heard anything on this for a while +Now that work is well underway with AS all previous community releases are +How would I restrict access to non authenticated users +Heres a patch that implements this +Thanks Andreas Andreou for warning me +The testing has been done combining with timeout condition check is moved into the catch block and just before. +Working on adding logic to to distinguish between device and emulators +Looks folder structure has been changed +In this case the portlet has not been accessed through a portlet container so the portlet example wont work at all +Patch applied to SVN rev . +However it would be nice if the build scripts degraded gracefully in this circumstance +Just means I have to get onto it and nut it outThanks for the example markup its a big helpWhats the Servlet engine youre running +The corba specs are being referenced via system properties +Can anyone confirm if contains the fix tooIt seems to work its just not clear wether fix made it there or its just a coincidenceUPDATE sorry Ive tested against so nevermind +It will work if the is put first in the chain as its added to the session chain when its created +Anyone object to marking this one as fixed in documentation +If the cert is not present then the agent should not do ssl +It should do no harm since Weld internals are not supposed to be deployed in a bean archive anyway +Programs that assume that all address types can be stored safely in a struct sockaddr need to be changed to use struct sockaddrstorage for that instead +I think this issue is resolved with all issues +Wonder if theres an OSS project out there doing transactions on files +Remove the namenode specification get the namenode configuration from class path +Revisit open bugs older than a year as Out of Date +Bulk close for Solr +Theres too many warning to fix before this release +Additional fix for determining the common path which was failing in windows was committed to trunk r +Originally the XSLT transformation was done in the testsuite for so that all tests would run against the same IP configuration then the groups of tests would do other config changes and run on their specific configurationIm checking whats the state of this concept now +reopening since were not talking about here +Maybe enough to sink the RC I put up last night +added configuration patch +Thanks Navis! +Please test in your environment. +Jonathan you are right I just tried CP and the issue is gone + +A small change to the docs to instruct the user to first run the CLI version using the help switch +This original jar in also has unreplaced macros in the manifest +Simple patch +r rfscholte text for annotation with comment +At that point you should be able to clearly see why your prependand would never fire. +In general we do not promote new features to earlier releases +Readline is essentially singleton in that case which means it wont work correct for we can do something like Charles to fix this specific problem and check if the input stream and output stream is connected to the need to think about it I guess +Fixed the issue and added some unit GuillaumeDo you want to check what I have commited? Can I make a release? +HiCan you include the stack trace so I can see where this error is occurring? Thanks +When will JIRA provide field level security? IT IS NOT FOR YOU TO SET THE PRIORITY ALWAYS TEST THE LATEST VERSION BEFORE REPORTING A +Now that work is well underway with AS all previous community releases are + +Except for the performance issues reported by everything else seems to be Ok +That is implemented +attached +Forexample is this just a question of what JDK Im using? Did we invent theterminology JDBC environment or is this part of the JDBC specification +Patch +The attached file contains all the hbase related logs +Turns out that the fix doesnt work on AS so either we need a proper fix for AS or a different packaging of WSRP for AS and AS +Anyway why is not an Enumto the second paragraph This sounds good one would move the to a single spot +I took nkeywals patch and cleaned up the test a little bit +I think its all in including the macro changes +The file fixes the issue. +am unassigning in case someone else can figure this one out at my wits end hereperhaps its just something wierd about my environment or I left ant clean test running in a loop and just checked it with this patch no problemsso perhaps its my own incompetence +Thanks +Set Component correctly +Ah +I think account would be OK +Waiting for review to Mar +Our tests seem to indicate that this causes performance regressions across the board not just when the feature is used +This plugin requires the and the sun tools jar to be used +Very trivial patch +Copying the master and slave log files +Simon if you have a moment maybe you can review this one for me +The idea of keeping it available for internal employees was to prevent breaking old product builds +Please accept this patch that allows weave style execution of maven reactor patch is the result of a long community discussion and serves as a fully functional patch that is targeted at a specific subset of builds typically enterprisey reactor builds +Are you tested it with GNU unzip on Linux? Can you provide a buggy zip filegenerated by cocoon? I will be glad to test it on LinuxMy question is because there is a bug in winzip very close to the problem thepatch fixes +Committed to trunk and. +Thus the new behavior is java sysinfo cp checks all the various components of the classpath java sysinfo cp client just checks whether the network client is in the classpath java sysinfo cp dbdriver just checks whether the JCC driver is in the classpathI think this patch provides a much better solution to the original problem observed by DavidPlease have a look +Sorry we made a mistake in the command line to generate the patchThis one is correct. +But how to check expected warnings information messages etc +also has issues when running directly from warI hope to find some time soon to hunt this down +I have no opinion on MsgMessage but that style change would be large than the bugfix so if we want to do so lets do it in another JIRA +Committed to trunk +Id be just a little worried about all those calls +But this has not changed for a long time +Helps some +Jon So you want me to review whats over in github and commit that +To achieve that I made each data table with a unique id +The release audit warning is due to a new public API in +fixed still fails due to another. +Yes like the current INENTITY option +We dont want to kill containers that receive EXPIRE in the RUNNING state +Devaraj and Harsh right now you can specify that subset of mapsreduces be run with special options to debugprofile only a subsetIm pointing out that allowing special denominations for ports etc +Appreciate if Ayhan and Phil could confirm on whether this fixed the issue then we can close the bug report +See also is broken with as well +In this case we should disable EJBEAR project name fields in Seam Facet PageThese fields should be available only for New Seam Web Project wizard +Kevin please make a comment and reject this as decided in SOA PM jira triage +If you think they are critical for CR please them +Good work with the testsIve merged your patch into trunk and branch +If anyone else has seen this please add a comment +Personally I think the number of dependencies is already high Id prefer notto check them into the main lib we can have a testsjdolib directory for specific dependencies for thetest cases +The attached script is uses the datetime datatype for several columns and this is not a valid ORACLE datatype +Once an entry is added takes care of calling free method hence clearing fileplease see the description of for more info able changes +Fixed as revision closing all issues in resolved state that have not been reopened nor commented over in the last month +I worked on it this weekend nad hope to commit it today or tomorrow +absolutely +We have the patch ready and will be committing the code today +updated path that fixes same exception in unpack mojo and also includes two unit tests +In our case it is not and in general it probably will not be +Patch committed to branch +This feature is yet to be implemented and now if you select a different type of database the SQL properties are not changed +Probably needs more tests +Since its the last defect in the way of the release well consider it completely resolved unless otherwise pointed out. +Condtitional Task needs some rework +for the patch +Fixed in r +After this JIRA for retired jobs nothing will be missing. +Abstract common framework for used on applicationContext files invoked via place holders +for sure for sure I will just ignore the plug in links in the translator and later on we will have to link them to actual plugins +One thing of using timestamp is that producer clients may start at roughly the same time and I do not know if there is a correlation between numeric difference of seed values and the numeric difference between the generated values +I split crash on Linux in separate JIRAThanks +Since the issue is a little complicated I decide to apply several patches to complete the function with one step at a time +Marking closed. +While creating the we can first kill the zk if anything is running in the default port and then create a new oneIs there any possibility of getting this in production code also +The updated sample output for queue dump +Built the with the commit above and it did fix the issue at hand +At this time it looks like were not going to implement this feature +With a clear failure did you have something other in mind than logging the error? Since this is running in a separate thread I dont think theres much else we can do +Ill commit this +Ah thanks. +Need this for patch as with bug fix in which makes all the tests pass +Fixed unit tests to not have path for trash paths +Can you review your patch for the unit test and add a new one for the new method the Query object it has to manage the node name to make this kind of query +About failures on RI everything pass on my +Patch uploaded +Tested in FF IE SaschaI dont have IE thats why i didnt checked it i checked it in FF chrome opera there its working fineThanks for your help patch +Also this just checks to see if the directory does not exist only if it gets listings from listStatus +Thanks +For files its sec and sec +If a writer already exists when inform gets called why make a new one +Updated patch addressing Todds comments +Should we wontfix thenYes done +Crud + +I have kicked off a hudson build again to see if the failing tests was temporary +doesnt take an interface to a so i cant just implement that If I continue to access the REST API via in the future if the REST API changes hopefully ill get the changes because im using the API from the project instead of my own client API We have shared code that is used in MR and Web Clients and access data via so id be back to having to implement the whole implementation This seems like something that other enterprises will run into as becomes more mainstream so why not contribute back +Aha! Brilliant +Code looks good but I am getting an error with this patch on compile +If region is still on this server we should be fixing up the recently moved cache +Done. +Hi Robert Andrewattached patch implements your suggestion now always represents a valid and constant represents the invalid one +Ill update the attachment here as well +Reporters of these issues please verify the fixes and help close these pl reopen is you see this issue again +ok got it +Maybe we can also add the files path info as well as the timestamp when it is moved to trash +I had to use MRI +Also tests need to be written + issues seems to be solved +Patch of blocking support +Uhmm +Looks like creating smoothly applicable patches that rename files is beyond capabilities of both tortoise and subversive +since the index can be easily corrupted by a forced shutdown or an I think there is no way to be totally sure that this will never happen a index could probably be the better solutionThanks good work +I was basically told that v was pretty awful and not to use it by folks at Yahoo +Id suggest to have this viewer as a new bundle rather than integrate it in account +are we not persisting in Zookeeper while creating a table +This is a good thing to do +Changes in the patch +same as before but this time with a comment to show which removal of entry will break the reason why you have a mix of getting things from the classpath and getting things from external resourcesThis should of course work but just wondering why you dont get it all from your classpathbtw +OkI think I have to update it in SonarWhen I go to Available Plugins then I get Not connected to update center +It is more limited in that it can only wrap essentially a request by having the opportunity to intercept the parameters beforehand and modify the response afterwards +Sorry I meant the dependencies of clients and contrib +Yes +Table supplied by Tom has been added to Chapter of the Programmers Guide. +I used clean install to fix the problem +On each prediction it calculates the probability of each outcome +Ive attached one possible fix +Sorry for the noise +The patches have all been committed and produces some numbers that without deeper analysis appear to be reasonable for all the algorithms +Because are held by the writerfiles remain open whereas before this patch they did not +As far as I remember it was the reason to introduce this propertyWe are running all tests on our local Jenkins instance and I think its sufficient enough for now +instantiates its asynchronously +The current idea is borrowed from hive use the input file size to estimate the reducer numberTwo parameters can been set for this purpose the number of bytes of input for each reducer the max number of reducer numberThis only work for hdfs wont work for other data source such as hbase or cassandra +Assigning to Gord for comment butRipple is known not to work in +As of now the hudson windows builds are failing on timeout waiting for the ci server to start +Committed to main trunk with revision . + released to rubyforge +Is there a forum thread on this +It will create a problem if the user isnt fact we dont have to add this in the pluginits possible to do it in the project settings. +I dont think its worth trying until we succeed under concurrency well only fail the second offer occasionally and dropping the update in that case seems reasonable +it would be a mess +However I just returned home and will have a patch release out tomorrow fixing the issue +If the user has specified a mix of nodes and racks even then we need to ensure that racks are added for nodes whose racks are missing +So when we release we can include wikisearchIf its in apache svn or on github it would be nice if stayed compatible with the apache license so it can be included w an Accumulo release +Just want to state that theres absolutely no need to require a special namespace for this +to trunk and + +I close the issue. +This seems to solve the issue by placing the registration of the hook in a different Paulbased on your description this might be the same issue +is there a in that directoryYes there is in the dir +its fixed locally but require more. +Bulk move of M to to should at least get this started in the next rev. +Moving this to +Yes its important to doc how this AOP stuff really works and how it should be configured +AnilI commited your patch in trunk rev +Fixed with and LB tests are now passing +When you can repair when you can release the new version +has been marked as a duplicate of this bug +Each thread do Create a Single Query +The new file will be called +Yes it can be considered as patch ready + test case that shows the problem fails with an unpatched but works as intended using a patched XSD file as needed for the test alternative patch for the part of the showing problem with substitution in multiple there are a lot of attachments now so just for the sake of clarity the patch referred to in the last comment by me is +Updated patch by your suggestions +The real problem is that if youre pushing schema changes and have to wait for a compaction even RINGDELAY or longer probably wont be enough +Just added a fix to the Main class +Assigning to Lucas for final review +I hope not +does not show attribute description which is a bit of a bummer +No we use delayed callable +The extension should tie in to Messages API in order to provide in support +fixing NPE in +ok first target fault tolerant resumable upload but protocol should be flexible enough to support parallel chunks upload +Attaching the updated request logger that logs the collected data in JSON formatIt hasnt been included in Wicket core until we decide how actually to log the data +This is what is great in an open source project we learn everyday from each other +We need the fix in EWS as well +Seems like RINGDELAY would be a better choice for this is the longest I should go wo hearing from any given node +Steve I created to request the cp version upgradeCan you link or merge it with +settings directory and content. +winutils is a native executable that gets built on Windows +Verified in GA version of Admin Guide section +On further examination the issue is that there are a bunch of +The test patch has been applied the string you searched for is two lines after the comment Regression test for +bq +OK then Ill merge it into upstream +How does it scale to x? Perhaps we should have a different related logo for that size perhaps just the binary without the feather and straightened up +Fixed in revision now ignores all JMSX properties unless they are client ones. +memstore size is than blocking sizeeventually the RS just devolves into repeatedly writing WARN Region +Canceling patch until Nicholas comments are addressed +Changed the code to check that condition before invoke pc +This looks like it is fixed can it be closed +looking into that +What config are you runningIm curious what it is about this environment thats causing this +Patch applied. +Would you like me to do the following create a new project examples site Developer Studio move examples that work on EAP and Teiid to that site remove the remaining sites from JBDS +pthreadcondsignalwait is also used by AIO module but we cant simply replace it with eventfd as AIO code use the mutex to protect other stuffSo I dont change AIO code in V patch +to +you are right this makes more sense +Any plans to include that? +In my machine I got patch apply failure +committed thanks Peace! +should we close this now +to look at those +If you finish it before we release beta please correct the version + +The exceptions were caused by the not being deployed +A possible fix would be to change the fields get to return a clone of the fields array however this is likely to have a performance cost inside the DGADG and elsewhere. +The original example is obviously no longer valid +Seems to be some missing event handling +hdfs dfsadmin printTopology is a good way to dump topology state +Yes +Rather than email it please attach it to this issueAlso dont forget that there is some basic documentation in a forrest seed sample site +The NPE should also be fixed already +but youd still have to name things with the ecsnitch conventions for things to not break still hold true with the above +Im not sure I like keepPercentage as an option name feel free to bikeshed over that and all the defaults I have chosenIt might be ready to commit now if nobody finds fault with it +Fix and test for all the issues the original patch had some unused imports in so I have attached an updated one +Just set up at Moodle with no issuesI guess you need to contact user list according your I made a silly mistake +still need more Mike There is a conversation happening on Reviewboard which discussed the same issue +Since the fallback with the property is missing too there is no way for the script in evaluate to find the all behaves as specified +The same effect can be achieved using a attached to the Usermanager servlets since this does not requrie any changes to sling I am closing this issue as invalid. +We will modify the patch to accommodate this and post the details. +Same problems on Ubuntu with +Adding a new bundle to the standalonedeployments folder will make it appear againAny ideas on how to make the JMX beans appear consistently +Thank you for the stat file +Its more from a providers perspective +Sybase Limitations chapter of Administration And Configuration Guide for both EAP and +Sure it applies it does not compilebq +Ive tested this in a number of ways and it seems to rotate my logs fine +Attachment has been added with description Patch for you very much for submitting this patch +This is skipped for nowThe failure in contrib tests seems unrelated +This issue will resolve itself on the next release when the term vector functionality will be available +Either that or simply remove both items in preference of the config page +The patch in also fixes +Maybe if I open a CAS with the wrong type system the editor then remembers this one and dont let me open it with another one +I attached the source of what the rendered looks like +Verified on revision +Most likely there are other problems as well +Version has been releasedClosing this issue. +Can we open an issue with INFRA to resolve this +theres now a Main in the module for booting up modules using a file to load a all issues +Fix has been verified. +Perhaps this is a feature being inserted and udig is producing that value +Is it possible for you to go into some more detail please? Perhaps you could outline what problem youre trying to solve? +feel free to address the nits directly in the commit since its just indentation +The equivalent of removing them would be to just not commit the deprecated methods to trunk right +Feel free to reopen if this is still an issueDavid +Id just nix it and use the annotationMany of these apply to the as well +Im not clear on that +I cant even think of how I can work around this issue +We can remove that attribute from the config schema +Thought about that but then IIUC you wont have anything to show for the backup master +I just committed this +Tested on HEAD looks like a bug in Welds modular classloading logic +Similar could be done for mvrm operations +Is it possible to add this support in the plugin configuration? +Perhaps david can comment as i believe he is the author of the Attribution patch +Disregard my patch then +after deciding to go with markdown and converting the zoo guide +Today I will make a patch base on new codesIf I change in that it pass the review this code will merge into trunk versions +We need to review the document to see if this is just an doc issue or if it is best to change behavior of the updated in an earlier release. +Fixed on trunk revision +Well I was doing it from memoryThere are lot of concepts so it is difficult to get everything organized perfectly +Luckily the XML Schema does not enforce that which allows us to take this shortcut +Please follow the instructions given in the Steps to reproduce section to reproduce this problemThanks Tyronne +AS upgraded to on master branches. +ACS feature freeze is Nov th +Hopefully the memory issue does not come surefireStill another error on the console the time reported for small tests seems to be wrong +Uploaded vhd and after the upload completion deleted from CS +The SNAPSHOT ftpserver and mina jars should be removed from srctestlib as well +In any case PM need to be involved +Are there any other comments or its ready to go in the snapshot branch +This is fixed +We can retest this +Patch to compile against api jar +bq +I need to check if this works with Pig the order of optimization rules in the patch the limit optimizer would push a limit between the load and filter and that would prevent the from pushing filter into load function +Patch for this issue depends on the protocol versioning in HADOOP +The reason is pretty simple binary compatibilityThere are quite some different versions of various commons logger etc around +The region is switched +Many thanks for fixing that fast + +olamy is that the package you really need +Thanks for the commit +Thanks +Interesting topic +Thanks KnutCommitted patch to documentation trunk at revision I agree that we dont have to duplicate documentation about JVM limits +looks good. +phkim checked in a fix for this problem already +The port handler delivers the request but no inbound message activity is prepared to pick it up +I can move pass it by issuing publishSchema however I run into the fix in which explains the problem here too. +thanks newline in xmlsec signature uses the Java system property falseto control the use of newlinesRegardsWernerThanks Werner Ill have a look at that as wellMarking this as fixed +The redacted part of the backtrace looked normal no missing fields no empty values +no one expects an client to talk to an server or vice versaThough I also agree that in this case even if we had that guarantee this wouldnt break anything since it is an additional protocol rather than a change to an old one. +Ping? Anyone? Shall we remove +The source of some benchmarks not passing successfully is somewhere else +This shows that DOES use only one query to fetch orders and their customers in case the relation is not bidirectional +Thanks as usual for theclear patch description the careful commenting and attention to detail in the codeand the test cases +However requests ending with will be discarded without any response messageIs the request youre talking about falling in that use case +and havesome switch somewhere to turn on classloading for references that are null +I thought about putting it support for autobinding to containing component parameters based on informal parameter name but that could be confusing and it can be put in later. +Is this resolved as part of the CLI work possibly +bulk close of all resolved issues prior to release. + +Has the fix for modproxyajp been identified? It seems the best would be to backport it and continue using since thats what the users will be doing too +refcount with the current layout. +stack looks like a wrong page is being pulled out of the diskstore? also it is interesting that it is pulling the topframepage that page is statelessi am also attaching a missing pom to make the project easy to setup and goingeven given that he is not using a different pagemap per frame this kind of think should not happen. +Unpack the attachment and run lein immutant test from the dummy directory +I wasnt saying i was totally against the idea btw i was just trying to invite some discussion +Turns out that we now need to include jline in the shades jar for the +do the same for stable war launcher is not having any config not even at wrong place copying from other launcher +Would you be open to patches that try to work around that or is that beyond reasonable portability goals +I suppose since we are talking on large data sets one could argue that the defaults should be much smaller but it is easy enough to pass in your own values in that caseWhat would you propose +I do alot of load data infile to already existing partitions +Not sure what you mean about the test all pass in HEAD now and there was no breakage before or after the change +taste Tofu +Many of the tests are not passing though. +Observe that the build completes do mvn mavenbugsfrobnicate +Normally this is configured automatically +Similarly no tests attached for the changesIbis depends on these changes and it works on top of this so it has been indirectly tested +Closer look should have revealed that the component is WMS +We are also planning to add some features to the Fair Scheduler ourselves such as fairness that might require major changes to other schedulers even if we were to wait on them adopting the existing features +Hum +As for the behavior with the jars I am debugging to see why the code path might be different when going through the jars rather than classes directly +Other issue i noticed is we cannot disable or make the textfield as readonly +can some hdfs contributor pl review this one +No response from original poster closing. +Overall most of the tools and interfaces are already available in common via the HDFS HA project +The problem appears as well in +Thanks for the update Andrew +Ive passed on our contact for central sync issues to Olivier as there are broader issues to work through as well +revised the banner file to use the logo as its image with the logo being a link to the project home page center the document name and document links in the approximate middle of the banner add another cell to the formatting table to create white space on the side which is what pushes the title etc to the center +Specifying a different user via the user switch fails with the same error +whoops account is ryanNot sure if CVS was the right place to file this so move as necessary +We should check all of our ZK queries I have a feeling this is not the only place well find such a thingGood catch Ryan +And once we choose to promote then should honor itDoes the current patch make sense with respect to making honor set api +I have checked the problem and codeassist works ok +Agreed! I assume that is defined along with the in the frameworkcommonwidget is actually a placeholder for the real application screen nameam I correct +HelloId like to work this issue + +I apologizePicked the wrong project from the dropdown and never looked back +Here is what I applied to and trunkMinor formatting change only w some change to descriptions removing statement of what the default is +Patch to work with latest master +Fixed with is expected as the fix was on API layer deletion fails on acl as were not searching among accounts which are already marked deleted +I came across couple of instances where potential users enquired about this and this limitation deterred them to use AvroShould we make the reflect default representation of String schema as itself? I agree that performance is the concern but I think for reflect that is fine +Sorry clicked enter too early and was assigned to you by mistake +Thanks Hiram +It still exists with Hibernate Annotations +Need to check if FC is subject to the same fact that the Follower was stuck after a of the other Follower was not caused by MFC but by TCP blocking the sending thread when attempting to send to C. +Is it an expected side effect that any Java object is now accepted where an Avro string is requiredThis appears as a type safety regression to meC +In this case I will close the issue +Its not correct we should put all info into transfearable object +Attached my change as the patch work + changes the return types and also removes the update parameter from some methods +That is a tech preview repository not a release +svn and +In this scenario the users would be well aware of the change and therefore have to make the necessary config changes too +What do you think +Just had a quick look at this test to see if it could easily be validated for I notice the use of a pooled connection factory +should work in. +I dont see this problem any more. +You can also set the explicitly +We shall drop GZIP handler since this is not a plugable solutionIF any projects needs a zip handler then this will get applied via a reusable and not via hardcoding this functionality into each and every of our applications +Could you verify it whether it can be patched or notSorry this statement does not make any sense to me +Which collation is proposed to be used in and MSSQL dialects +It can be achieved by several way +We still need a way to provide a to a separate host gbean for jettyAs Jeff mentions this still requires more thought for tomcatAdding modulesjettysrcjavaorgapachegeronimojettySending modulesjettysrcjavaorgapachegeronimojettySending modulessrcjavaorgapachegeronimojettydeploymentSending modulessrcschemaSending modulessrcschemaTransmitting file dataCommitted revision +I just committed this +In the case we go with exponential only then we dont need the modification to rightAnd last comment it doesnt look like the classes implementing need to be public +added section about JDK t documentation issues be assigned to me? +This is not an acceptable way to implement the sanitizer because it only works for the Java build +Yes it does work if I lose the S +What is the status here +The correct test which actually works when manually launched is indeedmvn clean install javadocjavadoc assemblyassembly deploy Preleasewhich correctly reproduces the goals defined in the release plugin configuration +Karaf branch uses +Committing momentarily to and trunk +but instead of an im just seeing the immutable check failing +I guess the levels of crazy hacking would need to be stepped up to support Avro especially Avro mapred rather than mapreduceIll try and give the patch a go early this week and come back with some more comments +I hope using the jmx api directly instead of the but that does not provide any API can help do the bundle persistence and also record it for a clean start +Last night we tried to deploy changes to upgrade this to +rpms are added back to our repository and they will be shipped in our releases +A file does not exist in the repository at the moment +JieshanCan you a new test case show this possibility +Still I am happy with the final resolution +looking good +Thrift server and test case updatedThe test only fails now due to assertion failure not NPE +Comments Please add this functionality in a separate class instead of +Thats right +Thanks for your help +This means that this issue only needs to be documented in one set of Release Notes and not the FP Release Notes as issue will be closed again straight after the fix version edit has occured. +Thanks for spotting this! +I found the same is true for update so I will try to find out why this has been disabled for DELETE and UPDATE +Attached a patchCould someone give me contributor permission and assign to me +Great +To compile them go to the accumulo home directorysrcserversrcmainc and run make to compile them for your system +Committed revision +I dont think a static variable is the way to go +From the stacktrace it looks like the AM was trying to setup the common portion of the task launch contexts and encountered an while processing distributed cache files because they were deleted +errr I meant now instead of not +Closing all resolved tickets from or older +Thanks JacquesModifying the url in comment. +another option is to use the German stemmer from snowball which is a variation on the german stemmerdesigned to handle these casesIf you use in it uses this stemmer by default. +but thats software for you +Thanks. +There must have been a different explanation for the problem I saw +But yeah I definitely should have been returning false instead of letting it fall through to true +It can be populated by iterating through all the blocks at the very beginning +I dont think so +Someone has to be first to use Struts Wiki for this +I didnt actually removed the Jira component +but it is a kind of fix +The first patch had in merge issue with line in but I manually fixed that +I like using generics but chose to forgo it is this instance to make the code cleaner and not force a bunch of compiler directives to be added +Ill fix +I think we can implement this only for single selection +I committed this. +Interesting +Reason nested exception is I tried adding but that didnt help +Does this indicate a problem in parse +Another is a scheme where users are mapped to roles which are mapped to HDFS users which are aggregated as HDFS groups user role HDFS user HDFS groupThis can provide some flexibility for various configurations from to +Ted whats the status here are you comfortable with committing? and does this patch need updating in order to be good for inclusion +If the custom sorting comparator orders things such that all the keys which should be grouped together fall next to each other then yeah youre fine +I will have a look on you code as soon as possible and hope to be able to merge mycode into your API for not loosing my information about the registration +Apparently it cant be reproduced on the latest +Added question in the FAQ How to make html scrollbars in modalPanel +Verified with EAP with JBDS B +I dont know why we have this two step process but looks like this patch needs to clear the datanodes set in recent as wellI dont think add is necessary +Resolving issue. +Vasily patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +This is somehow related to are needed to make full use of a folder +trying to catch up on mail i completely missread almost everything about this issueyes yes +See lets continue the discussion there. +thanks for the +Naming is the hardest part +These methods can be by classes that extend RedirectThis is internal development nothing changes for the end user +Committed with revision +But definitely dont want to remove the concept +No action on this in years and no other PIG has stuck in the project +In addition I added release notes +maven it really missing from offline +I think the roo t cause is better than exception stack trace +Thanks for addressing the comments +I committed the patch to and +doh! +Having said that this was a mess according to other changes to that had been committed in the mean test suite works without failures but I need to run JDO test suite later on before would appreciate if no changes to would be committed befroe this one is resolved +Ship it +Patch for the script the echo statements +thanx that works for me maybe it should be noted in the docs that you must add a x to the end of each java file before running groovydocs over that package +I can try with the patch there if you think it is the same issueKathey +I agree +The updates have been verified in the ER build +Moving this to doc +This issue reverts back what has been done in +You have to put +This version has no anomalies +Export with Copy API can be realized in a straitforward way +Im happy to settle with +Reviewed and tested +Working on this one +This change has no tangible benefits that I can see but does have a downsideWell I dont particularly want to argue indefinitely on this one so well probably have to agree to disagree +I am closing the issue out but if there are any more problems please open a new issue and attach wsdl and soap response +In you execute asadmin commands rather than sending extra deployments +Thanks EricLets get it into CamelI wonder if we should introduce a DSL for it? eventLogWhich also allows you to use any Expression so you can use JUEL OGNL Groovy etc +removed the debug messages and fixed all unchecked warning in Configuration +Wherever possible iBATIS favors leaving the SQL as it is typed +Attached patch fix this problemThank you +Add a repository through the web interface so it will have some new configuration to save +there is no BIGINT datatype +If Gora does not support that and is nevertheless a paradigm its not possible to use it for +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved +Any reason to further change the archetype have you found any other issue or problem? +Overall Im still unclear about why we need this when we already have tasklogs in HDFS and they are accessible via etc +Im not sure whether this is na issue but please check whether the file has the correct permissions before you check in +The code sent in my first comment is representative of my problem but i will soonsend you a complet test caseCan you please send in the test case you mentioned? Thanks +reverted the code change with rev +Only the Hibernate Session should be kept open if necessary in a customization regarding Hibernate flushing can be achieved via Interceptor +Lets get rid of those remaining bytes as wellReplaced UUID instance with two longssvn revision +I suggest we apply Daniels patch +But than weve changed it +Thanks for your help with this +My mistake +We should reduce the nuber of sites exposed to a minimum and nit expose tools sites since they have other lifecyclecontents +Fault paths are now handled although requires the creation of explicit transformers as the jaxb object is contained within the thrown exceptionRemaining work is to createregister the appropriate transformers +wrong file again +thats a very basic topic and without portability we dont need at all +Committed as svn +I wonder if there is a way to delete an attachement +Please tx pull and retry +It does test for it +The wiki space key will be SANTUARIO +I like the idea of adding a new flag to the getListings call +Its working fine and I have attached the file +Justin how is the user supposed to fill the Feature type for a shapefile thats not thereIf you want to allow editinginsertion the shapefile should be there just empty +Weird +Id prefer we address this as follows Remove references to since Avro cannot correctly handle all subclasses +Thanks Aaron. +If we do not put that load then it is working fine +patch against trunk with minor changes +Following should be added into uses uses ity uses angusesor uses usesororgorgor uses pectionusesorg +patch to handle SQL identifiers correctly +When invoking the List All Subscriptions operation on a Topic that has multiple subscribers Im now seeing the error described in. +You have an old Galaxy S on Android right? Can you check this one +Note that single property needs to be set which is not a case in your exampleRe we have utility class but I guess one more utility class is needed Ill have a look thanksAlso on the client side also register your provider explicitly by passing it to factory Sergey The custom jaxb provider extends and its set in server side bean as well as +Probably quite easy now +updated EPR serializationdeserialization returns a vanilla EPR +Thanks Deveraj +Are you seeing it in? We did a big refactor between and +In this case when JPA is starting if weld has not been started yet MSC may decide that the optional dependency is not satisfiable and start JPA anyway even though Weld will be started later on +The bacon project feeds the eggs project +Nick this is fine +The failover is a Karaf feature not an app feature +Clausto reuse these things for the mail attachment handling is absolute okBut again this is only the handling of mail attachments +The pruning indeed appears dumb +ThanksIll have a look at it and fix the test +The patch is about cleaun resources merged with jira +pure upstream. +Ok thanks then thats what Ill do once I get to it. +Lets see what Elliott and hadoopqa tell about the patch before I resubmit another one + is no longer needed +Heres the external process wiki page with relevant edits +Could someone add the right abbreviation for grams in Russian Thai Chinese Japanese Vietnamese and other languages? Thanks +See my comment at the end +I have attached the file for your reference +Yes I would love to have access to the patch +whoops +The difference between and is that the server is started using the +Resubmit patch to resolve two build dependency issues using IBM jvm +Karel shall I open a JIRA issue regarding Arquillian +updated with references to mail mapping files +I agree since the view is from the coordinator describealivenodes isnt very helpfulCommitted Mcks most recent patchbq +as well +The pull requests above have been rebased against +bq +The advantages that I see Aside from Java developers most other developers dont know ant as well as bash +Duplicate of +the suggested fix probably wont hurt anything. +Added a small application to test this behaviour +Done +If necessary I can supply the pom file for reference +My +Anyway supporting inheritance would be a great feature. +It doesnt work +I tried creating some but assembly depends on the sandboxed that has compile errors + what happened to the other patch? Is it still accessible or does it just get deleted +Our behavior is correct the spec needs to be clarified. +See my latest comment but essentially I missed one tiny place where XML documentation support is required +Use always +The biggest downside is that it might prevent me from ever getting a reproducible scenario +Sorry didnt mean to grab it away from you was doing five things at once +Doesnt fail anymore +bulk fix version change +v +Real world latency are expressed in milliseconds not in retry counts and we should let people push those in their operations +I am having this issue as well +The first patch broke the build in had to put the validation code at the HRS level +Thanks Alex Im looking into this right now +ill see if i can make another test upgraded unfortunately we still have the hack +Have you tried the sourceSchemas param +Please map root to the name of the person that committed as rootI am pretty sure they arent the same animal as our svnadmin +This issue was resolved by Cristis comment this issue duplicates so i mark this as resolved for the same version as the duplicated one. +I ended up needing to change the to no longer add the types after keys when converting back to properties +Will you resend it +The character doesnt appear in a password hash +Think this code has been extensively anyway. +BenoitThats exactly what Im looking for +Ok Ill test it too +Have tried many configurations but can not reproduce a bug on localhost +Found the problemWhile looking at your implementation and the diff to mine it became clear that i am creating my client based on a WSDL on the SSL enabled server +For this reason has been removed in cannot be removed in because should be as compatible as possible with its been renamed to. +Rescheduling to back to unassigned until there is a concrete request for it +for scripts for closures leaving Field for future use? Hmmm +Status update +FYI removed the global directory that has the in it and excersize the repro step again and again and also do the redeploy again and again it never produce any problem +It was applied at the revision +new feature example of a beanshell script converted to groovy +Committed revision +Some stuff depends on and others onWe need to clear this up +Im not aware of anyone trying this in a real setting but the issue is legitimate and we have verified both through reasoning and tests +It is because the checkstyle requires only return statements that is why i had changed it to using a flag +This is just a temp fix will refine it later +Deepak +Attach a full patch +I get a Forbidden message +about apply the patch you posted above as a temporary workaround +What seems to be happening is that the client launches another thread to send and receive message +HiOK great! Yes the patch required only minor rebasing +I applied few modifications to cover relevant report entries +I just updated to and using your route above I see connectors from convertBody to the choice and the jmsorders thats what you meant to happen right +These two scenarios were tested with only using the updated +Groovy script used to create the images and SLD documents +We should be able to safely skip the analysis though when we find no classes to analyze +I just committed this +Im going to keep it in contriblangid but have it use the Tika libs from contribextraction so that we wont have to package them twice +Simple class removal it should be fine without additional tests +The journal node is tricky to clean all the way up since it registers with the default metrics system which allows registration but no easy way to unregister +I dont think it should go into the code base as is as per above +Fixed in revision the commit +vinay thanks for explanation +I added the tests with ASF licence +Without an XSD I can type up a file with the XSD I need to find that horrible chunk of text to put at the top of the file every timeSurely there is some way to turn off validation in the IBM parser right +NET code from other +Proxy rewrite to Hudson added +Please reviewOleg +Unfortunatly I believe it will only work using the latest versions of Tomcat as the servlet portlet spec provides the requirement that a portlet and a servlet that share a web application context should also share the that backs the +IIUC the problem is that Tika in this case does not properly normalize the type names which leads to the mismatch between the detected and supported types +I just upgraded my instance to and it works great! Now I can finally focus on real analysis failures +Maybe it would be better if Railo didnt support any cfscript tag without parenthesis if that causes problems. +Its the same way to fix both log spams so we should fix both at the same time if its thought to be a good idea +accepting +locally tried +implements lazy fetch and am not too sure about remote check because it will be set to true for any kind of exceptionThe patch looks ok to me otherwise +Jersey is also just the implementation of a Java standard +Bumped to Critical priority since it looks like a simple fix +This is due to the fact that we hold the lock while calling out to bundle provided service factories +Otherwise only deserialize position +JBTJBDS running on Juno all new issues will be handled in separate Jiras +New failure this time in a different test +how to get involved to implement Template currently im doing heavy development project using i glad to contribute for the distribution there is now an example based on the familiar northwind database +must enforce void return type which is simple to do with refmethod and not possible with one still needs to benefit from expression while maintaining the same behavior as use with I also refactor and fix in this issueAnd also how about to rename these both channelAdapters to Default instead of +This simple patch fixes the problem and now the processes are found in the bar file +Thanks Bikas! +Baskar there is something odd going on +I created a new without the hidden +I still suggest you use my patch but perhaps uncomment the lines that were still debating +And note that the central Maven repository is +lgtm good to reduce the hbase version dep for compat +Yes that did the trick +So If there arent any further objections Ill set this issue to resolved +Probably one of those cases that some previous test did not stop some serversAnyway it passes in my own builds and it has nothing to do with at all +This is likely due to an invalid soap message but wed need to see it to be sureThe other part of the question is whether CXF is being used on both side of the wire or just on the client side? From the stack it LOOKS like the service is sending back an invalid message +Sushanth and Ashutosh updated the patch to reflect the comments +I installed xalan and xerces and when I try to run Xalan from the command line I get this error Exception in thread main javaxxmltransformThis class is not in any of the jar files shipping with these products +Should it be criticalSaw your mail saying HBCK will bring the cluster back so just askingLet me also think of a soln and comment here +Checked in new SBS and works OK. +For clusters this will be very important so were hoping that we can help get this done prior to the release +I still fail to see how a sax style parser and json could work together but thats another story +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +on additional hooksDifferent usecases can be supported +DTD file be documented to with this feature +I think that I somehow copied the file from an older version at the time that I was installing apologize for the inconvenience +running tests +Wont fixYou dont find it strange that the exemplar code for people to follow is confusingwrong that the comment associated with the code to create a time describes a currency select and is clearly a copypaste errorThis is still an issue in the current docoThis long discussion is all about word! Trivial but still wrong +With trunk solr all cores with loadfalse in without any changes in the codeK cores server startup in ms minK cores server startup in ms min with the changeK cores server startup in ms secK cores server startup in ms secThis little change is a big boost for the server startup. +Merged all of the either into the master or the branches. +Patch to allow configurable progress interval +fieldName article type +bbbcdcfbacdbf +At the same timethere could be a start or a or whatever local forward that pointed tothe entry page of a workflow +Please wait or test in trunk +Hi Doug I read through the Forrest documentation and played around with it +These are the patch files with log level changed to finest +Ill add the same options to the all the LM which will be +Having them in the master pom your update works +Irritatingly enough Eclipse does not show the duplicates +Verified with CR. +Tested in to during tidy up prior to +bq +Fixed at close of resolved issues. +I now separated the usecases active data fetching a view gets the data it wants to display passive data fetching a view gets notified if its data was changedThe former implementation did treat both cases in the same way +If you disagree please let me know +The package is used if no package is specified +FWIW I would rather few languages were supported and were fast than support a lot of languages that are all unusably slow +I committed the patch to trunk and +Thanks Doron +Attaching the patch +I think we should still always look for and load but not +Sample for more detail +r. +Pls check the latest patch +The failed tests are not related to the patch +see this new linked issue update the new functions for suggestionbox and is in progress task is assigned to Sean Rogers as it affects the FUTURE version +Moving to M can be moved back if it will be done for M +They should have setters and getters and the setter should be locked after type system commit +Think about EAR scenarios again +I havent pinpointed yet where exactly the problem is but I think its related to floatsIm using BETA at the moment. +The patch looks good the tests in particular +Yes there is no harm to do thatI will test and commit it +Yes I think it did help the original case is now closed so not currently relevant. +The dependencies in the pom for example also have to drastically change +Ok I think I got it +Attached Test case to replicate NPE +I will add a diff in the morning +Documents are now split when they reach a threshold of k and contain at least commitsRun time of TCK tests has already improved quite a bit but there are still some performance bottleneck which need to be addressed +NB for the unit tests has been renamed to because else Surefire tries to instantiate to do a Pojo test set + +There has never been a report of this problem +but Steve if you get a chance can you review it? Thanks +Attached the repository the model jar the zipped source code of model Tihomir can you take a look at this one please? I stepped through compiling the package with and logs a stating unable to parse xml +Nice +HiFirst of all please do not create tickets before discussion in users mailing linkgSecond plugin wasnt migrated and correct groupIdartifactId is still +Benjamin If we need shade could you roll a release +If not please add a commentlet me knowTo see the diff changes you can unzip both zips and then use your favorite diff tool +Fixed the vectorizer unit test to accept as valid vectorization RemusOverall this looks good! Please see my comments onEric +Type checkingstatic compilation at package level has many drawbacks like impossibility to perform incremental compilation +If you would like me to include any links let me know. +I am closing the issue +Yes passive monitoring does go some way to help alleviate the concerns but as I say we want it simple for nowPerhaps a simple script run weekly emailed to the dev list will suffice for the time being? Tony +Applied to CVS +Cheyenne could we please make the following changes Change ApplicationService Runtime to AppService Runtime Add another small cloud underneath AppService Runtime called Runtime with a speech bubble as the small imagethanks could we now just move the Rules Processing Process Management text above the flowchart image to match the rest of them +Also Cassandra is now in maven central so that may be of use as well +Very nice indeed +Please am not able to assign this issue to sathish +If however you go back and them raise an event on the previous page that transitons to the view again a new will be allocated +The test verifies that just nothing breaks +Attaching screenshot as a regular file +The core test failed is not releated to this issue +Thanks Jitendra +KeithThere is a difference between process instances and process definitions +sorry for the delay +Thanks Ole +This needs to get done with high priority +This issue has been resolved for over a year with no further movement +For now I simply disenable the running of the test under Cygwin +Add check for null pointers +The custom icon is applied when only one icon attribute is set. +Added my patch +I have changed some attributes of this issue according to latest evolution reported in above comments +please watch the recording +As long as the syntax checking is good and will throw errors when something is mistyped having code thats easy to write and understand is all good +The release notes will be updated in time for the release after this JIRA gets resolvedIve restarted the at problems are fixed. +Unfortunately it simply looks for whitespace to determine a word boundary +Honestly I think that the spec that requires the jaxws proxy to implement is not very good +But the basic idea behind the patch will remain the same most modifications are just to make things a bit more flexible +Of course one can also document the setting on as well if it changes +The patch is attached besides the reported one other differences have also been stated and interest I have also attached the and the of the. +See for request to enable Nexus access for Lucene and Solr +First cut patch there could still be some issues +Thanks +Thank you +Why is this being marked? The change is backward compatible older clients will ignore the extra optional field and structure +Im sorry I mistook something inBut my suggestion here is use xffff only for assertionTo achive this we should add textLength property to and use it in modify compareTextIll submit a patch +For developers they just have to care about and or replace the do you call send a diffwhat tool? +The point being please vote for this issue so it can be fixed properly +Will post a list to and create an associated jira for adding these +will think we should simply do nothing +I like that we are now conforming correctly to the Iterable interface +We happily take patches +They are there just for the first time +Fixed in head. +I am positive on both trimming and decodingI dont think anybody will be setting say replication to But having an opportunity to set to x rather than to is an advantage imo +As I see it artifacts are just a way of pointing to files in maven repositories lets keep them simpleSo if you need to add metadata on artifacts you could add it to the jars manifest +To elaborate consider the case of the client gets a token from which is a CNAME to +I will write a migration document +The project sitemap bug stated above is fixed +can confirm that the alterations to my patch do indeed appear to workSubmitting patch to Hudson +Suggest you try using and see if this same problem exists +Would you please try it +Addresses Nobles thoughts +Perfect +Then the bug is most likely in the initscript itself +settings +Closing. +Patch for trunk addressing Mahadevs comments +verified with CR +If it doesnt it will at least give us an exception indicating which type definitions in your system are leading to the issueShould be in the build dated or latest snapshot build has fix the problem for the compilation thanks +i am sooooooooo much happier doing it this way +As I mentioned in my first comment classes are already described in the documentation +Im going to try something out which I then want to use in to support further configuration options +I have created JIRA for documentation coverage Could you please attach your sample application +It will not take me time to make this into a generic header based authenticator. +So far I had assumed you had no influence on the suppliersActually Im an independent consultant and in some circumstances I am Supplier B trying to decide which version to use and how to communicate about possible conflicts to my users +Ive attached a leaner patch file +Obviously there could be other bugs with Xercess handling of this complex schema but until the schema is cleaned up a bit itll be very hard to find them +This is cleaner than hacking to create +Checking back with parent the specific performance characteristic is not clear +This introduced regression in the CTS +Based on the comments from Mamta I will need to generate new patches for th Developers Guide and the Reference ManualIf you have comments on the updates to these guides please make them now +this supports both a gateway and a process manager API component. +I have working but will make one more improvement I think +If you dont hear anything from me I didnt find any other bugs gg +mojo +The new code should be also the zk cluster key really should be using the format that wants and it should be used else well have as many ways to point to a cluster as the number of tools we have +Thanks Arun! updated with the review comment +This doesnt include any dom objects that may be created previously to producing the report +The stuff which Ive used before looks reasonable to me +actually Ive changed it back to use +Let me know if I should run any other test on this patch +Either the messages are being cast to their incorrect types or the is not working due to classpath changes or class version changes +Im looking into ti +Better to get a compile time error +Patch for that has the datanode preserve the original hostname when registering with the namenode +Most code is pushed to implementation classes +Configuration still references CQFS filesystem inI suggest we configure the derby persistence manager and add the derby library as dependency +Im afraid you have to provide some more information on how to write test cases +Im stealing this since I suspect that its core not tools +When gets put together well backport the changes as appropriate +Committed there +Several new capabilities are listed here +Going forward we will be using the wiki +This idea also came from Raghu +Any user code that implements an will no longer compile after this change is made +OK lets move topology state transfer out of the equation for this JIRA Ill try to configure the lazy retrieval +Removed deprecated UTF parameter to some methods +One thing less to think about +But it is probably not worth ensuring +Its a real pain when Eclipse happily builds a project and you think everything is fine but then executing Maven from the command line fails because the project relies on a dependency that is not listed in your POM but just happens to be available in the workspace you currently have me several times now +QA tried to apply the +This one sorts by relevance and does no filter no deletions +Thanks AnkitYour patch is in trunk at r +When the browser requests a widget from the server it must already have the right configuration otherwise it could not download the widget in the first place +This looks good +has been moved to devNext is to merge working on that nowI note that you havent mentioned anything at this point about the user and commit lists +this makes sense +So actually this issue turned from a new feature to a bugMike thanks for doing the work to add it to FRN it made the fix very easy I just pulled it to instead + Added overrides to all the request object to make logging more readable and ignore things that we dont care Errors are handled by each request and removed the case statements in kafkaapi +First implementation inspired by code in patch applied +Rcommend you upgrade +Why it takes so long? It is because the writer only writes to a single replica in the first place +So for me the only thing that we could do here would be to allow creating anonymous sessions and for this I believe that we need to add a new method such as get in that will internally call internalLogin +unscheduled issues to the next release +branch also has a error in swtbot requirement +This first version disables validation of certificate +Applied in r +So I have changed the subject +Even with the implementation of the ability to add rules to the you still have to configure them + +Here ya go one on white and one transparent background Cheyenne! +Sorry there is no problem my mistake +Ditto No errors with if changed as either +Fixed javadoc warning and a bug in +Patch to integrate PIG with Hadoop +Hi Dave Sorry for the spam +for separate heap size Fix Version to Hadoop +This tutorial is specific to direct programmatic use of Cache +New classes in blue +The issue should be left in Patch Available state until a committer commits the patch at which time they will mark it Fixed +Assuming goes into I think we can close this unless Todd do you want to pull out that particular fix to a patch for this jira + are implemented with a hack +Thoughts +I notice you are on Tomcat which has never been tested with the bridge so I am curious what you are using +LGTM +We dont use USER anymore therefore lines and were death codes and I removed them. +The same bug appear with just checked that this also occurs with so I conclude this is not a change due to +uploaded under I am the owner and original developer of the library on and I am unaware of any request to us by Davide to upload this project to any rd party Maven repository +All unit tests passed with the committed thanks pradeep +I removed class and updated the assertion spreadsheet +is not going to be resolved until somebody takes the time to write and automated way of updating the rhinojs +ant test Dtestcase Dqfile DoverwritetrueCan you regenerate the patch after that +John I fixed the problem with javamrexample +if that is too much of an overhead we can also disable this by default via IWC and make it optional this is actually very straight forwardreviews comments are very appreciated +What do you meanHow the Hibernate Tools are related to Hibernate Anntoations compilation +This is a very simple change +will +I was going to use this one instead of jQuery but it was not done yet +Ive just committed this +I get the same behavior with Camel as with I figured it out +Its likely that this isnt serving a useful purpose anymore +Pardon me lars hofhansl why +buildgriddevbin +Patch closed this one out a while back +Thanks for the helpHans +Weve decided to let branch with ant as main build engine and focus on mavenization on branch. +But it seems that a more general external parameter as suggested by Si and Daniel is still missing + +Are you sure you didnt have an old copy of interfering +I have fixed part of the issue that is related to the Tools preference page +and we use replaceAll for this +I attach a patch file about two images of the former comment +I forgot to resolve this issue for so it will not show up in the changelog for that releaseI am also setting as a fix version so it will be listed in the changelog for that version +All the sentences should be short and clear +We have open issues in there we dont want to loose access to +I think the related code there is atget inqlmetadataI think we should be sorting the files there as wellThis is picked up in to create the paths needed +Create a shelf and an +All big boys do it is a weak argument in my opinionWe can do betterAnyways I have the patch almost ready +Thanks for the review Suresh! Ill commit this later today to unless there are any objections +Fixed in r +Some lines in the new test code The new test is mostly dup code from +Any ETA on when this might be implementedNot really +A B do you think it is OK to commit now that Dyre has verified that the regression tests pass +Ill review this JIRA as the highest priority then +Im getting an NPE at a different location not at the extensions. +Hmm didnt quite realize there was already a patch when I started this +Unfortunately Ive just done a couple of tests and this doesnt solve theproblem in all casesThe fundamental problem is that the utility class used in the sampleprograms frees memory allocated by the DLL which is something that shouldgenerally be avoided if possible +This task will be done as soon as the linked task is done +If we go with the forced order it might be nice to break things up into fixtures along the lines of the original test numbers +Hi YousufHow many times did you triedIm running my local test environment with +You can file another jira and I can commit that separatelyCommitted +sample project and steps to reproduce will let us provide fixes faster +Rest the internal hash map every time readFields is called on sounds reasonable +Can you create Review Board or phabricator entry +Done. +moved to since this is a +The Revision History since is missing from and January pages at least +Hadrian had the karma so he have assigned karma to me and James so we can help Camel committers to grant karma when needed. +breaks backward compat with previous trunk to trunk +Chinna this slipped through the cracks and is not generating a lot of merge conflictsI think this is a really good idea Can you refresh? I will definitely go over it this time and try to get it in +Testing old revisions it seems that worked on ppc when poisoning was committed but did is on my list to look at this soon but naturally will not be before +All the Connection options specified on the Uri are now applied even when failover is used. +All the subsystem tests pass with this patch applied +Committed at subversion revision +this was dependent on to work in all cases which is the case now. +See the enqueueCallback wait execute and releaseCallback methods +And should the test if the is closed? Reason for doing this is of course when using the that depends on the Id keep it simple and just throw an exception if bespoke situationhappens +This is also with the reference as an external library +Heres the patch that addresses the problem which is in the if you name your property anything other than the class name the link from the edit page to the create page will break. +Heres a patch with the changes in my previous comment +What do you think +The problem was that when the keyspace creation was reattempted identical objects were created that obliterated the older correct cfidsI could have put a check in the private constructor that threw an exception but decided Id rather have explicit control over cf instead +Committed revision . +patch looks goodI tested the patch manually +The property in for telnet referred to that oneSecond I was not aware that Remote Shell was using properties on the osgi +If it happens again please post the relevant part of the log +Jakob any more thoughts +Dont think they are neededMy bad +This patch looks good to meIt also passes mvn clean verify +Now can be used for both and +I would add master to NUMZKINPUTSPLITTHREADS and I dont see that you solved from the task +Our release procedure requires VPATH to work +Now all logs are split +I have tried upgrading to and then upgraded to but it did not resolve the problem +Updating with style changes requires efficient pruning of outdated tiles need a way to find them quickly +So this goes to trunk +I still need to make sure its doing all the necessary and proper Mainframe resource utilization +Reopening to add fix version +Please I consulted the work in progress version with Eva +Adding a link to you want that feature if you want multiple masters +Trying to set the default cache size but there is no knob to tune itUse x on the command line or set it in your as wont fix as it looks like this isnt an issue. +However someone could check the code and handle the pottential null value +Im not able to build this branch +This will not get fixed for +That were addressing in other issues in and elsewhere +You can make it warning or ignore it in Seam Validator preferences +As Ive written on the list I think the entities should be evicted from the persistence context to really solve the memory leakIt is quite tricky to detect when to evict them +Bulk move issues to and +Exact same compile errors +Patch committed +I found the fix i had missed some things in fileTested on IE Chrome Safari Opera with weather widget using numeric key +The original resolution may not be wont fix but as the bulk change required it that is what it is +Thanks Oliver indeed it seems s end + for the to trunk. +trunkCommitted revision +For the record there was in fact no available version for +Hi EugenioThe paging feature of is not just for browser scenarios where someone has to click a link +YuckWhoops cross posted +No problem though this just takes a bit longerWhat should I do with the old mirror that no longer points to anything +Xue you write Done but there is only a commit for visibleIf already fixed by other means in trunk then all is fine what is left is just to resolve this issue in jiraIf not fixed then it should also go into trunk and then you can resolve this issue +Lars Up to Ram +You can open another JIRA to reverse these changes and link it to the yoko JIRA +committed in revision thanks for the testsAndrew please check it was applied as MikhailThe fix looks good +I think the should not send a at all if syncfalse +Does anybody know how IRC archiving works and who initially created the channel +Updated comment in about updating style on SBS spaces as well +Eric I looked at the attached thread dumps and in my opinion the deadlocks are the same as in +The disable command worked fine but the drop afterwards failed with the for METAScanning ROOT succeeded scanning META threwYou mentioned you are seeing the same problem in a fresh cluster aswell +Apply the following patches inside modules Apply the following patch inside modulesdistribution very nice work on the core improvements to the adding the AMQP support +updated version of the patch to work against current trunk +Nope the framework is in for anything that derives from the base class but only the uses it yet +has been added with description New patch taking most of Andreas comments into account +i try both and +made the hard per thread RAM limit configurable and added DEFAULTRAMPERTHREADHARDLIMIT set to MB renamed added set to that is checked each time we add a delete and if set we flush the global deletesthis seems somewhat close though +The original stepwise traveral jumped to the root while going back and from there going down to the starting point +It fixed the bug reported in +Fixed in rev +If no fix shows up I can create a patch for it but was hoping this issue would shed some light on the fix that jason mentioned +Basically it has to deal with object mesh with potentially circular references +patch for xdocs and docs page with css file to show how this is going to look +Ivo please verify from stage +This issue is addressed in +Excellent but I dont think its a minor problemMy own skin originally based on pelt is now far from the original one generally for translation reasonThe plugin has now the same limits and I also have my own copy of itThe configuration management begins to be complex for me when a file changes +This was fixed with. +PDF document created by is a duplicate of +And by the way I tested that it is OK to run icupkg on the file even if it is alreadyI find it a strange concept to have two binary file formats one for and one for only one of which is usable +Lets keep this option in mind for later relies on the to be added crawldatum metadata viaAnyway i agree +batch transition resolvedfixed bugs to closedfixed +The current svn image of the +please review this +method internally reads one doc at a time +tests failed with testbalancer +Fang since you are already working on this issue +has been marked as a duplicate of this bug +Ill handle this issue using the following behavior +And root should not do it either because the packaging system will not know about these there a way to pregenerate them and have them installedOr can Grails be configured to generate them in the users home folder? Something like +Not strictly related in terms of bugs but are related in the fact they are issues with the same class +If you say so I was just concerned about additional amount of fds open +closed. +Committed revision which implements for All comments welcome +Ive committed this and merged to and +s reproducible with betarc +Just to note the ode dependency was updated for +We do not attach event handlers to touch or anything like thatIf it didnt happen on Mobile Safari it may be a difference between and Mobile Safari tooCan anyone narrow this problem down to any specific executing inside +Yeah I read thatWhat I was saying is besides the interface where do you store the flag? As a field in the feature source I gather? Is the feature source available to all usage points that need to deal with a PK then? That was mainly my question +im a bit afraid on sychronizing the repository instance +Thanks for taking a look CheolsooA test would be awesome +I also needed to test something on JIRA +This is already a good step forward IMO +Hmmm interesting thought though +We expect to have dev build released in a few days and will post announcement to the user mailing list +Thanks for the patch! +If you need or more you have to do it the old fashion wayUnfortunately sflj doesnt support varargs +Attaching patch which fixes this issue +and this doesnt even factor in posLength +Proposed architecture for Mumak +Ill apply your patch for the time being +Especially lastUpdated as it is not set when an object is created +Please close if satisfied +Closing this seems to be related to the Hudson machine somehow and not the test everyone Ive asked to run this have had no problems with the test. +We have a few XA tests in our testsuite but these tests are using XA directlyYou should probably just use a MDB as a consumer and that will kick XA underneath +Isnt this too close to JSTLI dont think thats really a problem JSTL and our jst templates are fairly different in change implemented in revision the is now and uses the jst extension +I tried with IE hack important but it doesnt help +Thanks Jay! +Moving the along with tests into coreserialization Also retaining a dummy in morphline sink dir that forwards calls to coreserialization you please upload this to rb? +Fixed on Cellar trunk revision +It is replaced by an occasional NPE in which Ill try to narrow down separately +Introduce a new which extends +I no longer see the problem +but it has no log for s a JVM bug nothing to do with MahoutI would normally concur with Sean but this does look like it might beWhat zip file is this that is being read multiply +In my case even register fail +This looks like a regression since it works on +We did +Please verify and close +Hi all will there be any progress on this? Is there anything more I can help with to make this issue be resolved? Thanks +I think yes as it is part of Maniks rework of rehashing +for the patchcommitted as r in branch +Hence assigning to Abhinav +js files generated using maven resource filtering. +bq +Its come a long way Alexey +pending +corrected the file translator and made the source required logic similar to the metadata source required in that a connection must exist if connection required is true and that if a connection name is defined but source required is false then we will still try to obtain a connection but wont throw an exception if one isnt available +Will commit this in a bit +moved fix version from final to RC since this is quite an old issue and is cluttering up the release notes. +In your are you checking that the Menu item is selectedThe behaviour you are describing regarding A +UI should provide users to add NFS cache storage after adding S secondary storage +Punting issues from RC to final +but this as much as I hate to have scripts seems like an overkill compared to the bash solution +All unit and integration tests pass +Im going to change the patch and attach it today +The Bean Validator spec is missing +Ok this is the patch with my modifications that affected code generationIve made other ones which are related more to and self but they are not includedI also included the modifications since this is my preferred way to work with HibernateHope this helps +The Aperture simple mail extractor was replaced by a new one that in addition to the text also keeps HTML content and lists attachments +Thanks! Please check the new patch out +This book is no longer handled by the JBDS docs team and is now apart of the SOA docs set +More REST related issues will come up but the overall picture is there. +Attaching an example screenshot +Fix ant targets Also set appropriate permissions for bin Sebastian +is on the and target platforms but I have as yet not added it to JBDS for an OOTB experience +Thanks Sijie and Matteo +MattiasThanks for the code +missed cutoff for R + +Raising priority as this is both effective data loss and an occurrence in the wild of a condition the author of thought should be impossible +for the patch. +Fix attached as a patch +update the IP given is the correct oneLewis you can now login with your keySee +Fine if there is a lack of test coverage already there +Looks like service status dumps for each offset? Is there a pointer to how I can try this + +Hi BJ and fixed issues you reported here +attached again for naming convention +Verified the calculator sample works. +bulk move of all unresolved issues from to +I also concur this is a reasonable way to model things if the module you are including is the one which uses the interface. +Im still not sure how I should implement zkmaxfiledescriptormismatch +How about reducing expected region count to +Fixed in rNow managed x respects the directory specified with buildxlib option +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content issue is out of date and will not be fixed +ok +The user needs a way to discover which layer is likely to be the one they want and a FACC code does not help here +Update the original line patch does not seem to address all of the problems +I did a fresh checkout and the old df code was not removed from core main and test it was removed successfully from examples +I just committed this +That makes us no longer schemaless +Could you please tell me where you expected to put this menu or if I have missed something? long time I submitted this +Question about If it is not legal to write an unset union should it be legal to read one +Thanks Richard for committing the patch +You should see the same error in Tomcat +Moved to This will go away once we nail down the deployment part is +I made others test and it work for all +Eclipse XML editor will be used for SAVARA v +b Include the tests in the production jars +I guess checking in lib is fine +all the ant core tests now pass +Steffen Does this fix your issueA better patch would add logging to this class and log a warning when a row when a NULL key is added +Attachment has been added with description Good has been added with description Bad problem exists in both trunk and Temp branch +This should fix the issue +Ive just checked that the dependency jar should be generated by a project on the same workspace +We might want to think about just making this a row in the Core table but this will at least star that discussion +Im terrified of this too +bulk close of all resolved issues prior to release. +It isnt clear to me what you believe the issue to be +committed this. +I think we can use transferRanges there even though we do blocking flush before streaming +was a problem in the maven embededer +but it makes sense to convert this failure to a sure about whether warning is really the appropriate log level +Isnt removing protected variables from the iterator in the of axiom considered an? I use Abdera which subclasses some iterators +Authentication headers however were apparently overlooked +Closing all tickets +I think i will stick to for nowThey do indeed +So I made a new one using cvs diff u +Downgrading to critical since weve decided to move off the Java backend for nowEventually I believe this will be a blocker for MR but we can make progress for now without it +I committed the patch +Minor nit is test mentions has beeb instead of has been +Is something wrong with Jenkins? Pretty small patch +Per cedric Ill come back to this once has finished and can figure out why the code was inserted in the first place +Anton what do you mean? Do we need the doc folder +So it is a very valid bug +Ill get your patch applied today +At least the issue has not been reported before +Marking as closed. +This issue is fixed and released as part of release +They will be there years from dont really see the advantage of making a tool to take some and putting it in a local repository when in the short term of a pilot project just check the to a repository +Nick could you see if this is still a problem? +Denis heres another one Im not quite sure where it should go or who should look at it +Oops +Ignore the jpa s the simplified roo scriptproject setup provider HIBERNATE database POSTGRES databaseName foo hostName localhost userName fooadmin password foo transactionManager entity class +Changed all affected files under jobhistory testcase is included as the change is trivial +reopen to set fix version to again with correct fix version +If you finish it before we release beta please correct the version +It doubt the framework hung it looks more like it cannot resolve any bundles thus the shell never started +Jason yes definitely the same underlying issue +Moving over a bug fix from the G sandbox +The patch you attached seems to be the whole file +WDYTOTOH there is also +Take a look at this for. +Could you possibly post a reproduction for this issue or more information to help us reproduce? If not I think we should close this asThanks for +Also there should be a trunk patch for this jira since trunk has the same issue +This issue has been fixed and released as part of release +ThomasCan you supply more information maps config source code java pieces +Ill probably introduce a strategy interface or template method for this +Does this mean that JDBC assumes SQLATTRODBCVERSIONSQLOVODBC then? And what about JDBC? Does it assume SQLOVODBC also? Umm +Eli suggests that if this is the language we expect most people will write their schemata in we should just call it Avro Schema Language +filtered from instance list added code to stop fw in destroyfirewall +By default its the that is used by the component did you try using this skin instead of the old mobile +We are using JBPM and we have a requirement to genrate report for which we need the searching for process instances that contain variables set to a given value feature in JBPM +patch to achive this +Bulk close after release of +Saved us a ton of time. +so we dont have to make any change for this issue other than just update the manual. +Fixed in revision this as a bug +patch looks good +Michele In my case I have some web services deployed on Project Web Access which return me cookies for authentication +Sure we can move it there +He already identified the fix it seems like a no brainer to me +Suggested patch to add a timeout attribute to rest binding and pass it on to wink API before making a rest for this update Anuj this looks good to me and ive committed it in trunk and nice to see that it came with test cases too +Some test cases are changed as a result +Okay that was couple notes this is a HD display so I am not totally surprised Java calculations are being thrown off +Heres my previous comment in patch form +Do you plan to fix this in branch +Closing all resolved tickets from or older +Committed as is. +This avoids having to force true and mess up other tests +Perhaps add a note of this to the README or the +Thats why we have the debug log level right? Devs can toggle it if you want to see all AM transitions but doesnt inflict huge logs on normal usersThe in the rpc code definitely needs to go so lets not revert that +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Hummm +Rebased PR +It could be easily implemented for we are going to release +However I think I was oversubscribing CPU quite a bit which caused this race more likely to be hit +See Target for location of annotations +Jan please try to reproduce this issue +Marking as unknown release date since there is no patch for thisHow about we mark this as wont fix due to inactivityAs far as I can tell Hama isnt moving very fast and hasnt attracted much community or code +This is the functionality bug +Is there a technical reason this needs to be done in the prepare phase rather than in the process phaseThe idea of this security filter forcibly inserts security Filters before executing query +bootstrap pleaseThanks +will bundle webtools stable. +I can proceed now +Oh Im allergic to XML +comments on updated with code for stripped +Reopening for the backport +Just leave it to the user +I am running on clean workspaces every time +I forgot to mention that I dont know do know Ant pretty I got this job I was mainly a mainframe programmer +I used your acct as thats the one you log into the other hasnt been used for over a year +The final patch to the issue has been submitted +If you find further issues please create a new jira and link it to this one. +This produced an assymetry defect that I think also exists in canopy clustering and kmeans +This works properly in AS +When its all lower case it works fine +Simple fix the light of recent discussions I would not call this a backport +META +If you find further issues please create a new jira and link it to this one. +Looks good +Cant reproduce myself anymoreWTF is this nowI always reckoned linux and java the most unspooky systems +Maybe if special sorter is written +Almost doneWe are all happySTART gio mag CEST messaggi non messaggi non messaggi non messaggi non messaggi non gio mag CEST Total number of messages not yet translated Saluti sab mag CEST messaggi non messaggi non messaggi non sab mag CEST Total number of messages not yet translated Bye sab mag CEST messaggi non messaggi non tradotti +Can you backport it to as well +I would prefer to use a mixin class because in the current patch there is a lot of copy and pasted code which just makes it harder to maintain +This one is against trunk +I agree it should have a release note +I assume this is something to do with the maven metadata +Modified from release from release believe this issue is solved with the control +The connection could also get closed before the response can be sent which would mean that producing it was a waste of resources +commited +Closing a repo used to take a matter of seconds if that +Please review the patch and apply it +Thanks +LGTM +Hi Claus happy to but it will have to wait a couple of weeks +Rich could you review and apply this patch to rollback svn +However since we cant really patch we have to workaround in our daemon management scripts +The latter while causing a slight constructor explosion doesnt turn out to be as bad as I thoughtIs this patch acceptable +New screenshot +AFAIK this was fixed a while ago so not sure what if anything I need to do +issues +what kind of error do you get when it is not in classpath? Posting it here might at least give a search hit for someone running into issue +We could imagine adding similar checks to make sure that the jdk and jsr classpaths are sensible too +has been too long that this has been open and with no action + also created the relocation for +I now have a clean Mavenized build structure that tidy modular and intuitiveIf I can figure out a good place post a public version of this project Ill post it there +patch looks good +and it works fine. +Just in case I dont find time in a hurry Whats the release schedule for +But submitting patches is a straight way to commitership +The fix looks good +bq +Applied patch to CVS +Well if possible I suggest you use scp via Maven JSCH +Hi Uma thanks for the review +Maybe seam can handle similar functions that apply only to jBPM processes +Hi Raghu can you pl take a look into this one and suggest what could be done? Thanks +Patch fixing the boxing issues +Assigning to Shelly to pick up for ER +Gail +I just committed this. +I ended up leaving the assertion in place making normalize support overlapping ranges is handy but it is more complicated than we truly need and Id rather not open that can of worms until its a requirement +if calling onsubmit at all makes sense +Its now filtering as expected +Sateesh can you check on this +Hadoop was released. +Issues is fixed in a slightly different fashion in looks like this is closeable thanks Robin +Duplicate of +It depends how the ejb release bundles are ejb uses the ejb timer implementation of the jboss appserver in which case fixing the problem there should fix it for ejb as should know +Thanks for the fix Felix +That sounds good to meThanks! +Should be all fixed now. +fixed +Try that instead of the CR before filing the regards youd better open a JIRA in the JBAS project for that +Fixed in git master and branch +The majority of people using Solr are not using MCF though Id venture to say that the majority of folks using MCF are using Solr +Recent changes must have introduced this bug +This patch calls get before calling a mkdir and then checks to see if an exception is thrown in that case proceeds with mkdir else errors out with the corresponding error messages +Ill rework the fix to implement Jeannes first suggestion +I think thats it I added a second pause and havent encountered it in tries +But +Attaching a patch that fixes the issue due to partial double encoding of the access token value +Fixed again +Status update I am working on this right now +I dont see why batching would necessarily help in this situation +So it looks like on windows x the bug remains +I see that Tika may be just around the corner so waiting a few more days to see if it materializes then we can jump directly to which has a bunch of more bugs fixed a newer and more flexible configuration of plugin parsers +In fact it could be refactored to be a real plugin fairly easilyIf it looks like a plugin it should behave like one +Thanks for the patch Elliott! +patch for the branch to fix the ANT download +I suggested annotations but I dont think they want to go that way +I committed the patch to trunk +This should really be done in the JBC testsuite +I understand now that it should be going in the CP and would be released on July since this has not been integrated into branch is not going to be possible to apply this patch as using static loggers would create leaks and other classloading issues in standalone Tomcat +If you find further issues please create a new jira and link it to this one. +file prev attempt had some issues +This is not a bug +Now T is interrupted +Peter can you please the patch as I believe that theres missing code in the test case +Can you tell me how to open uploaded file that is not an image +I dont see a problem with having a space in the middle like San Antonio but youre right about trimming +value is in my current hadoop version +Maybe set the true system property by default for Beta +You can look at any of the sample programs to see how you can catch them +Checked into develop branch. +And it would be consistent with how listeners work in the SwingAWT world +Thanks Gregory +Resubmitting patch to Hudson since the last test run seemed to be very brokenThe changes look good as far as I can tell +BTW the Remoting version does not need noting there is no guarantee that components will always have their versions exactly in sync +This issue is no longer relevant now that Messaging can be deployed without scoping. +Our Andrew Purtell has set me up w a clone of the boxes used when ec build runs +Actually for you are right offsets are returned in descending order +In the current head version of the calendar popup functionality has been completely reworkedCan you test again with the current headthanks tested again using the SNAPSHOT build created on Jan +Need a separate file for each image and icon +Fixed +Ravi it is not about the ticket type +AFAIR I set the as host to bind on all interfaceshostIf you dont mind David I will make the change +Yes for sure please close the issue as I could not it +It comes with an old ASM version unmaintained and buggy +Superseded by. +Passes with the patches in and JPOX CVS HEAD + +Add configuration parameter to allow the crawling of sites where a is returned when accessing +PLUS the everything jar in the xalan binary distribution is missing some classes and thus doesnt actually work +Revision I leave this Jira open while bundles have not been released +Error in test application +depends on message smtpsession Peter we need this +Fix attached as a patch +Im not sure if this is an issue that needs to be fixed as well +Ill look into the master branch as if you want to pursue a change to the spec you could try this mailing list how aboutAccept applicationq applicationjsonq available media types applicationjson and applicationxmlapplicationxml should be picked as it has a higher qualifier +Hadoop QA test failures were unrelated so patch committed to branch and trunk. +Note that you can safely use Cdevworkspace +Created. +I agree that Vishal can help in filling up the table +There are a bunch of binaries not handled by svn diff during normal process contributor will submit patch for things in srcdocs however committer is responsible for running forrest which updates truckdocs and then committing resulting trunkdocs along with trunksrcdocs +Closing the values in a kvp request must always be properly url escaped. +Can you again +Fixed on and close all issues in resolved state that have not received further comments in the last month +All of them works fine +Where is the data stored into? A remote WFS layer? Or a databaseIf its a database I fear there is little we can do its the JDBC driver that controls the timeouts and normally thats not exposed as a parameter to the by wfs all issues that have been in resolved state for more than one month without further comments to closed status +downgrading severity since theres a workaround to to did not run across this for quite some time now +committed +from your stack trace is not a valid hostname +CoolAre these a different set of optimizations than those already proposed and the rather unspecific +It sounds like it is somehow related to the VM reboots I dont know why yet +Patch committed +Heres a patch for trunk +Hi I haver tried with +This might be a thing tagging +Any update on thisCheers project does not have a SVN repo and is intending to start on a new canonical GIT repo and bring in the donated code which is currently in the donors git +It is designed to map to a schema not to allow users to pick and choose +Gallery Sources +This is a candidate but not committed for EAP introduced JVM property that is enabled by default so users can turn it offSince now remoting in transport is configuredto always close URL connections by default +This problem was also fixed in the release +The height will then be scaled automatically +Lets see if we can squeeze this in for +its not clean but heres a workaround to implements an Asynchronous interceptor. +Works with Java fails if you use the MVEL dialect +Im up for whichever you think would be best +Proposed patch +Im hopeful this issue is resolevd in as well be using Java s concurrent libraries and not rely on issues in +Open class and check TIMESTOITERATE variable if variable allow change repeat times the issue is fixed +In that case thats actually a bug since you havent set up any advanced zone and that button should be hidden +EPP was built with early EAP that was CR in that moment +Same problem with trinidad +Users may type in machine name of connelly but over on host it may resolve as +I just committed this to branch +ChrisWhats the bottom line here with CRLF and ? The repository is keptin Unix LF form isnt it +I think it is ok to go with the option for now with the understanding that we clean things up as part of the transition to Hadoop +. +Perhaps a bug in the summary code +Comments can you please create a patch that only includes relevant changes? it seems there are several changes in the patch that are just line breaks at different positions +This shouldnt be problematic since anyone will be able to override it by specifying the property explicitly in the framework configuration properties +Reproducer for the underlying memory leak +Tagged for the release. + +Seems that RI allows is there any way to verify your assumptions +holding it +And most of the local methods has been made static +That works +Moving out of +Not seeing this problem on with Java +YES +Fixed checkstyle issues and hasPrevious +I would like to give up this issue to someone which is more familiar with the camel build system and could testapply my patch +This state transition is now UNSPECIFIED +This doesnt seem possible for this release +Add missing license good thanks second thought this still has a problemwhen running it on my Mac the test fails with End fdcount is +I can dig if necessary +Maybe this is a stupid question but what is the reason to index a very large docIf I indexed a whole book as single document It would appear in almost every kind of searchs result setssearch computer this booksearch sport this booksearch politics this bookDIGY +Believe it or not the sample is right +By the way I want to make sure its clear Im not trying to blame Mark hereI suspect he takes some of this personally because he is the main guy actually working on fixing these problemsI just opened an issue for the general problem that we need to stabilize the solr test suite so we can trust itPart of that means that more people than Mark have to somehow be involved fixing this stuff +was done as part of +Neglected to resolve this issue after the two were merged into the branch. +So I think it would be better to refactor the doc creation in into a method and add that extends and overrides the creation of the doc to be added calling super and then add the facets into it +Sergey Im not sure what you want me to do +What if of the JVM versions suffer from this issue? Also like I said beta did not have this problem so it might turn out that it is not a virtual machine problem it only looks like oneBut heck I dont care a bunch of people might upgrade to and their nodes wont restartJust like the Xss thing we can ignore it and hope it goes away +In this case it had defaulted to the Marketing department instead of the company like Im used toPlease close this issue although you might want to glance at the second part where I get an error receiving an order if I pick the radio button Purchase Shipment but not if I pick to receive the entire POThanksDaniel +We should probably come up with a strategy of giving shorter unique filenames to the job history +Test program and JLDAP library +If it wasnt clear in the above comment I am looking for feedback +I also get a strange order when building under windows mvn vMaven version Java version OS name windows xp version arch x Family windowsI think we should change the app project setup back to normal dependencies again and distribute these amongst the build levels +Patch looks good +wont fix +If there are two tools for related things some admins are going to curse us when they have to remember which one to use for particular task But we can have two tools not a problem for me +Do you think a release is possible before August th? How could I help youThanks in advanceBRABANT Benjamin +Thanks Cos +This is clearly a good feature and I fully agree with you Andreas branch coverage must not replace line coverage but we need to have both of them +looks like the new patch is for not +Sorry we did not get to this sooner +This is a thread safe implementation and there is only one additional io hit per system instance +Im closing this as a duplicate since the functionality is effectively available through +Updated patch with javadoc and naming so seems to work when I step through it slowly with a debugger but fails pretty consistently when just running the test +Draft complete and committed to SVN revision . +added this support as discussed updated release notes and Aggregator docs. +this fixed it for will be incorporated for the change for for release to branch for inclusion in new release candidate +Tested the current patch on a index containing M documents that has M unique groups +You can probably get another x out of the algorithm pretty easily and possibly x with some cleverness +These blockers and critical issues are resolved but not verified +This is covered under. +Karaf Trunk svn ciSending jaasmodulessrcmainjavaorgapachekarafjaasmodulespropertiesTransmitting file dataCommitted revision +I only made it defer the fetching of the token +Updated the setup script to use unique key tabs for master and region servers +Why dont we use an enum instead of a boolean? and +You need facelets for and has +Sounds reasonable to me +I agree with you +Thanks Robert Im taking a amd +I have seen it only with IBM +If they dont well fill our in files full of duplicate text that will quickly fall out of sync if we some particular problem with the text is reported against one instance of it +It looks to me like everything in Struts was updated to call the new methods except fo theMartinThanks for the reponse +Having this field makes sense if you Update an existing item but it would be better to leave it blank and let the system work out the next seq number if you are adding a new one +Add detail to exception thrown from undefined syntax +And after all there is only two stylesAlso in we will have ANT files to start the samples with thanks to Peter for the hard work of creating the ANT files and provide it as a patch. +This is the patch I am going to commit +Hi AlanThat patch needs to be there +Currently the JPA API is bundled within the Webapp anyway so that wouldnt make a big difference +In general a secure server should check that directories and files should be configured with the minimum accessto run the the service securely +I have the reset method check if the cache is null before creating a new Iterator because it would not throw an exception before if you called reset before calling next +The problem is that user jobs are written by users and therefore are not really suitable for detecting unhealthy nodes +I agree lets pass the reused in +Proposed patch attached +sorry call +wed be able to remove them in provided we formalize that is a mandatory stop before upgrading to We should clean up a lot of other things in as well if we require going through on the way to. +We can do with +OK here is the patch with the macro +My bad I didnt quite get the Optimistic Concurrency feature it would indeed do what I was describing +I dont think that its will be fixed because its can be fixed only by removing these assertsClosing as Wont Fix +Reopen to upgrade to to to +I think all I really had to do was replace the call that copies a C string to a Pascal string +Thank you. +Any update on this? Do you need any additional information +Forgot to remove unused imports +svn commitSending quickstartswebserviceproxybasic quickstartswebserviceproxyrouted quickstartswebserviceproxysecurity quickstartswebserviceproxyversioning file data +we can fix it if it turns out to be a problem when we move to a real serialization api +Hi Aaron I dont think it is a good idea +this patch adds a continue button to the box head that submits the form for you +This is too much of a price to pay for a for what is essentially a bug in a broken CGI script +Updated checkstyle from to and moved configLocation from reporting to the plugin definition +Facebook JS SDK was attached in the earlier patch. +thx imre for fixing mojarra +xml as the location +Theres no patch here +Sorry I did not know that it was included in the cassandra source +As its not valid XML on the wire I would also classify it as ran wsdljava s t on the and rewrote the server impl and client +Proper fix delayed until cr +Yes this is communicated more clearly +One option would be to make the callback handler an abstract class and not force users to implement methods like get +You installed into asas? Running? Did you update the endorsed jars +Committed at revision +Do you mean to switch this default to false instead? I thought it was set to false by default +I just did a mvn clea install and it works fine you provide more details on the failures youre seeing +I assume user can just go to jira and lookup change report +Sorry for my late response +This issue has been fixed with the attached patch fileAdditionally searching on identifiers has been added with this patchThis patch invalidates the orginal patch that fixes issue +The test is very small and intended to be backwards compatible +What if we throw an exception if a value set attempt occurs where a different value already exists? We could include messaging about how if probably what they want +Applied to Committed revision Committed revision Thanks! +Ive lost track of this now + +Endpoints are now children under hostxserverydeploymentsubsystemwebservicesendpoint can we close this issue? + Username user +There was a seperate issue due to lazy creation of the session that was masking the +hi the attached zip has the the port type is and the operation is UnsubscribeAs you can see I do have a response message for that operation which points to an Empty type in the schema +Any chance you can relevant items from that discussion over here +thats fine for testing but what about later versions? for wanted to know if there is no better way +Committed the change +The initial documentation has been done +This is the previous patch updated to reflect the Path changes that went in today +An example project with a junit test containing the all resolved issues due to a successful release +May very well cause other problems but after reading a great deal of annotation related jvm source code and other items it starts to make more sense +This is also a problem for +I am not sure but it looks like the patch of bug might haveintroduced the problem described above +This may be a bug of set seems it should set last when setting AMPM +Attaching updated patchThanks just committed this after running affected tests +Bug has been marked as a duplicate of this bug +commited in rev +I just need from another committer to commit them because I modified the patchesThen I will open follow up jiras to implement missing features including multiple schema support etc +Committed a with revision . +I want to remove marking interface as well why do we have it +This patch bundles the previous patches includes a junit test and some other changes to to make the testing easier +Unfortunately the machine running thesetests had a number of problems so I dont know exactly when the issue started but it has been happening at least since +Ralf assigning this to you to grab your attention +Assigning to Kevin to sort out the related issues + was fixed in +Ive found usages of CFSQLVARCHAR in our codebase vs for CFSQLVARCHAR so clearly either works in ACF and in previous versions of Railo +Any suggestions on this +Actually I didnt believe when I saw that the a solution to solve that would be an instalation path no spacesYes! That solved the issueBut whyAs Glensk said I am also using version but here is versionRegards +When you are ready it would be great to iterate on this in a branch +Once is released the svnexternals would be removed trunk moved to futuremain and future renamed to trunk +If this isnt the desired behavior please let me know +I think youll get the result +Im not sure theres a right answer so well stick with the status quo and hope a better answer emerges sometime. +Ahem +After a close look it appears to be default behaviour for NONE isolation level so this issue may should be closed +Verify and update +I am currently running tests but am interested to know if I am on the right track since this is my first collation bug fix +Ill go ahead and close this issue then. +backported fix from trunk to branchsibmsvn commitSendingSending javatestingorgapachederbyTestingfunctionTeststestsstoreTransmitting file dataCommitted revision This test was created for a new feature in so this change need not be backported to any version before addied the backport reject label +Could we get the version for from either or via replacement as with +I think Satheesh andor Tomohito should look it over to make sure but if they have no comments then I say commit it +Thanks Jake and Ben +Will close after verifying in. +The patch iterates the dependency list to develop an optional artifacts +If you need assistance please contact another person who is currently at the company +Evgeny could you please create a review for this patch at? Let me know if you need assistance doing so +fixed using your suggested all resolvedfixed issues of already released versions of Roller. +if not please main reason for this to happen is does not return after logout but continues with doChain does not invalidate sessionplease take a loolk at the source code +FAIL if a bad checksum is downloaded consider the downloaded file to be bad fail the attempt +this one seems to for uddi +Attaching the patch to fix the issue +Currently this isnt the case +It is still not binding to the with id drinks or to the repeater with id contacts +Fresh build works for me now too +Patch applied +I do not think it should block any longerOleg +Thanks Ivan for the reviews +This change got the resolver completing consistently in about second with testcaseIt is not clear if these fixes are general or specific to testcase but my intuition says they should provide general improvements +What kind of is the super class? Is it a primary class node as well or is it something else? And in which phase are you doing that? Because if it is a precompiled class then there wont be anything in the if the node was not resolved yet +This was fixed in by +except block so that a single problematic file wont stop +I changed the test to be less brittle in this regardWhile I generally like both of these last two patches I am still unsure of the wisdom of this last change +Patch that makes limits into percentages +Patch with the just committed this +I made an hadoop from head of with the patch that Todd just loaded up into +Youre absolutely right about it being Zipfan I was just trying to discuss the upper bound on error without having to do the math for confidence with different distributions +Patch with review comments incorporated +Attaching issue observed in fresh installation also +This patch will resolve the issue +They just confuse people +If you upload a zip containing multiple shape files and specify configureall it configures all of them +My mistake on the last perf regression +New patch fixed a s a nice little test you wrote there! will commit this once Jenkins says okay +Committed revision +Did anyone look at the solution I proposed in JIRA PIG for partitioned rank computation on billions of rows? There may be better solutions but I do know that that one works without need for serialization of all rows only on the histogram. +Arent we allowing multiple threads to do put which may result in inconsistencyYes this was my intention +Closing this as fixed in deltacloud git commit bdadbbfdbbadffdeOn Fedora Ruby +This is probably one for Dan +commit +Initial patch lots of nocommitsPatch is against the branch +Thank you for you commenti will try to fix in the code +I think this was already covered by in trunk +I dont really know ivy but here is a patch that moves dfsminicluster dependencies from to core +this will only work with s right +can we do the forrest doc as a separate patch? its already quite large as it is +and their definitions? wondering if we can reuse one or need to introduce a new kind of marker for sites we do not controluse +There are commits happening in the branch ATM +Hi CharithI have checked in the basic message store API and the message store impl +As test issues dont need to go into the release notes it doesnt matter from that point of view +Reopening for handling in seems reasonable +Unscheduling since was reverted. +I am happy to supply any more information if neededThanks is really difficult for me to say if the above scenario description is correct or not since I am not able to reproduce by creating bundles for the above three manifestsI mentioned two situations above where it might choose a lower version instead of a higher version however I do see a potential issue when selecting the best candidate +for the sI can still make it OOME here locally if I use small cells of size bytes and if I put up lots of clients +Here is not about append and create +thanks for fix +I see no reason to not do thisHere is what I proposeA implementation that by default does document routing with a single merged response +This is complicated as this string would have to deserialize any maps or conditions +Use cases are described on the page . +It is basically the same patch with more testcases +Another instance is the and classesI agree that prior to having the Audience annotations it wasnt clear who the intended audience is +fix is OK for meThanksPH +Though theoretically it will be better to find out the features on the fully compiled and optimized MR plan it will be hard and may not be worth the complexity doing it +In fact if there are GPL leftovers it is because I didnt notice +Proposed resolution to Changes are contained to the class only +Section Authentication contains a procedure that could be duplicated with specific settings for the info from the Security Guide +The reason of keeping in mapred package was that it works only for tasks in a mapred job as of now +Feedback from Hudson +Roman Ive created the fix and applied it at the revision +Took care of most of the productservices directory need to work on the productlibext directory next +Which version of Maven are you using? My guess would be +It disables the test cases that fail with client client gives wrong and protocol error inserting a GB clob +We can use this as a starting point +Thanks Todd +So exact resulting combination needs to be carefully defined order matters +Running on a test node I changed the CF to and scrubbed +It does not refer to the default block size of the filesystem +Isnt this mostly about? The patch is also named and not +i would like to be able to put Fix for values that people can rely on. +Upon deployment the web service module finds the latest version of the process definition at that time +txt and +Transparent at plugin level Passes was saying to oposed to doesnt have regex +I run the tests again and its actually AS profile where I can reproduce the problem +Thank you +As applications grow in size so will the complexity and number of indexes to support it Im concerned creating this many could cause serious issues +If parent is not alive it kills the c process after three retriesChristian Can you use this patch and let us know if you see improvement +Checked in in Teiid Designer +The idea here being that any class which wants to do messaging can ask for access to a instance but it must supply a locale that the messages will be inOnce this is done we can begin touching up the rest of the code to properly get an instance of the class in the locale that should be usedIll be committing some of this code pretty soon so please feel free to review and offer feedbackSide Note a number of the classes jsps originally listed have already been as part of the struts struts migration and therefore are no longer of concern +At least the version should not change maybe this helps +Very true +Theres a problem with that patch in that it hardcodes the absolute path to the artifact on the local system in the remote OBR +Great! Ill try it asap +Attached patch basedon source. +Madhumita is this a issue? Does work from command line +this patch was made at HBASEHOMEsrcjavaorgapachehadoopclient +This issue should be deleted +Marking as duplicate ofIf youre willing to attach a +Is it +What badness happens if you try to release and the topic isnt thereif a topic is evicted it means the topic entry is not in the topics cache right now +If you are not seeing it perhaps it is JVM version specific +just looks kind of silly and is trivial to fix +You means all of the four values of the ConfigurationsIn JBDS the server eap is wrong? the web service runtime axis is wrong I will check other two values are from wtp framwork if you use the new web service wizard we cant customize it +Thank you Eric +For me it was quite comfortable moving between the Java and +This is done but I need to create the wiki page for it +Todd can you use the code in DNS to get localhost reliably on more systems +the bulk of your code could be example the same you would just need a simple which impliments and wraps asomething like that might even be a common enough use case to inlcude in the core +Thats correct +I will have your questions answered hopefully by Wednesday +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Regarding zlib I see no need to make it a requirement especially since few people use the zlib transport +I think this shouldnt be a big effort and it would pave the way for new kinds of custom constraints by users and BV providers +I will be looking into this shortly as I apply the inferencing engine for code navigation +Another option is in the etc dir of karaf +I broke it +This means that if the latest timestamp is and we know we want to keep at least entries then if we run across any timestamps greater than that its guaranteed to be in the top entries and we dont need to consider it further +Though confirmation from someone who can reproduce it would be nice +True always a hassle! bq +I borrow from CXF sources +It is likely that the extra jbpm jars found in the are the believe the DS defs are OK now but there is still the problem of ending up with different databases +This is not a problem for partition inserts because the input and output is a single directorypartition +I did mvn U Pcodehaus webstartjnlp and now I dont get the error anymore so for me it seems to be working +Program to reproduce the fixed string matching bugI compiled it like this g o x +bq +It seem that the iso doesnt add the Taiwan here for us +Chandan is right + +The update of Norman confused me +This is now fixed +Patch for version of the patch is being validated by s build for stack to make sure that it solved the problem there +Fixed some test errors +Reassigning +Yep that was it! Is that port listed in the docs? I cant believe I would have missed that! Thanks +how these tests are different +As per Jacob applying closure. +Yeah as you can see the git mirror is not tied in with svn and so when moving code around one must also ask for any mirror to be updatedI have started this workinjg from revision when the current location was first createdWill let you know when it is completed +ReynaldFor Maven Job type we reuse exactly the same version of Maven as defined for JobFor Freestyle Job we cant guess and thats why there is selection option. +im confused by this issue +Freddy Dont know how you can test as this change is invisible for end user +Patch which reorders the arguments +I refactored a little and added a some logging but the core logic is the same +class from bundlekarafroot featuresaddurl mvnxmlfeatureskarafroot featuresinstall cxfThen deploy this isnt a SMX JBI component issue its a SMX issue so move to the correct jira instance +Seems fair +Id be happy with having Java as a prerequisite for users requiring the Bugzilla side of maybe worth taking to the dev list for discussion +added information about Phase Listener for is done. +r Committed on winXP +Spun off a separate JIRA for project wizard enhancements +Committed to trunk +Please look in history for what the actual resolution was. +Proposing this for as well +Confirmed that it happens without debug enabled as well +replaced complete preference for fixVersion +We also needs to fix the same typo in accordingly +I am not sure whether this is a good thing to do since it may break client scripts which read log messages as input +Manual tests run created a har file as follows hadoop fs put test tmp hadoop archive archiveName p tmp test tmp ran the following manual tests wordcount on a couple of har files streaming on the same har file with hadoop jar input hartmptestaa output tmp mapper cat reducer wc lBoth of the above jobs completed successfully and had outputs in the corresponding output directory +I will try to persuade the owner of the server to adjust their service accordingly +for previous HADOOP version not for just committed this +Patch a adds a basic test for database decryption +I dont see this ever becoming a priority the code will likely be dropped as part of the Infinispan move +bulk assign all unscheduled issues to next release +I cant remember though the use case for this and forgot to add the forum thread at the time so lets leave it as it is +I think I might be a little confused as to how the code deals with that dictionary +The code in both projects are almost the same +Check it in +Im assigning this back to you Heiko to verify this is still an issue +Im happy to review +Should we reopen this issue or create a new one +See next exception for detailsCaused by Illegal key size +as in JSP the default syntax works also inside html tag attributes. +More over Translator overrides are exceptions not normal usecase +Yes this would be very useful for the Lucene migration +a lot for the review Todd +Fixed since revision . +Go to bed Ill review later +Maybe this issue should be set as duplicate to since the resolution of would resolve this issue too +Yes I can work around this which is why it is optional priority +Unfortunately most of the JBI component unit tests only test synchronous usage +What is the source svn url for the website +dims +Does this test provide any value by being in the suites when SUR is not implemented yetRather than trying to make it run clean in ibm where the problem seem to be related to exceptions we could wait until it passes cleanly testing real functionalityThen the SUR patch that makes the test that runs clean could put it back into a suite +I merged it into see commit adacbbbdcfeff +going to commit v that fixes the ensure row per region t fail anymore after matteo jujitsu +I just figured since the doc method was already there it would be a time saver +Maybe what I am missing is the context of what state means +But why would this starve other? Shouldnt the RM offer some other containers to these other jobs if the cluster is idle? I can see how some containers may be just tossing back and forth between the RM and the picky job +Ive heard good things about both Gradle and Rake but as I say Not something Im familiar with and until the IDE can run off of the gradlerake script it doesnt offer much more than what Ant with Ivy would +The are new and replacement files for the xalancbranch treeThese files incorporate the work of Samuel and are built using MS +Are there any projects with replicated brokerconsumerproducer management out therethanks Carey completely agree with your here +workaround implemented +Currently updated Preference Page looks like on Current Updated VPE Preference Page screenshot +Then a validator can read the file while generating random objects from the seed and schema and check that the two match +Go and commit it +Everything looks good for me +It would still be good to have a full varbinary type and associated literal escape syntax +done +Considers all changes and links are OK. +hostname other network configurations home directory etcThat seems reasonable to me +Further investigation required to determine whether this is a application or a component for this issue are passing in +Values that contain such characters are now always written to sub elements and. +Aleda thanks for the stack trace +is a second patch for the srcmainjavaorgapacheservicemixnmrspring file in the NMR that does the same thing. +My CFLAGS are fPIC static libgcc shared z initfirst the appropriate httpd mpm settings the MTOM sample can be ran for long hours without any issue. +Two different composite components one nested inside the other +The partial allocation approach actually isnt workable +SergioThis seems like a simple trivial fix and problems getting this into +Closing since changes appear in Latest Alpha Manuals. +I had to change so that it wouldnt have an empty fileset of asm file on win because empty fileset matches all files so build tried to assemble all interpreter sources even onesPlease check that patch was applied as expected +Thanks for your report. +I also ran this on my CI hardware and I no longer see the issue there either +Same for me when running ant with the proposed patch in +So lets close it and open a fellow up issue to resolve this at some point. +Verified in JBDS Linux JDK +Reopen if you disagree. +Could you please quote the relevant parts of your Spring configuration as well as the exact Spring version that youre running against? XSD files that you refer to in your bean definition files seem not to be found there +ok but that is an exception but what advantage does it have that we have to maintain the list of IUs both on connector side and in the exampleim fine having the option to be fine grained but things becomes a lot easier if we dont have to keep multiple lists in sync +Would definitely comment on the jira after readingthinking through this +test failures are ve committed this +Fixed this in SVN +AM doesnt interpret this right? User code interprets the payload and should probably know the underlying layout +I think John is looking into making arc do this automatically +After using it you will get a schema containing all of the common typeselementetc +Uploaded some new images with color options and monopoly guys more feedback and ideasrv I like many of these the color really helps this onerv Not crazy about many of these if you go for bling it needs more bling rv None of these really popped for me just did not really like the colorsrv I like the original best +Not a problem +I went ahead and added the semicolons and committed the patch with revision +The testcase aforementioned passed on my machine with classlib r +Changing priority to major from Blocker +The more thought I put into this the more it seems a separate operation name is better +for the change +Will fix assuming same is true for recently released +Carl are you working on thisWe really need it asap +in M +Committed at subversion revision +But so far the only experience Ive had with the Eclipse libraries is checking out somebodys +However im pretty sure there is already a way that these metrics can be turned off +x for the bug report Andrea +any chance you can do a little more work to chase this down? the fact that it hasnt saved the model is clearly the problem +Thanks will test this by next week +For we developed a fix for it that works in most cases +Closing with the release. +Brian could I close this issue and start to another one +Richard could you please test if it works for you without any special settings in your browser +thanks for the patch hopefully it will be committed soon +The issue as described is invalid changed to reflect error reporting in the UI +Cool thanks for the recap +I do not want to make it default unless the latest release of hadoop totally works with uber settingThe making default code should also have logic to ignore it for hadoop clusters else we will pay the cost of pattern matching done for adjusting AM size +But lets see if Im able to implement it +Fact that tool has created table in workspace itself is actually helpful +verified. +Patch updated to this one has gone stale +I may be able to supply a version in the near future as itmay be critical that I develop it for our internal needs +The problem has been reproduced and will be fixed in Sonar. +Test case is checked into trunk. +Improved with r. +Hive already has infra for plan walking +Uh +please ignore my previous comment +The nodes on a cluster should be able by your network configuration to talk to each other +This leaves an empty indexTechnically the logic is still correct +Output logs for php tools are sent to +wont make it into M worries +for patch. +Yeah I have it in mind +I dont understand why you combined both patches into one +The wildfly console is the only supported tooling interface to the transaction product +The first problem is simple the nd one is more interesting +We probably need to have a boolean parameter to turn this on with off being the default behavior +Mostly making the benches easier to run +OK +Hi BrettYes i didnt get this email in my inbox sorry about thati think i am fine at this moment as i already made my own repository code If i will find any issue in future i will contact you againThanks for all your help. +Sorry I posted the wrong link +Patch attached for review for DProjectsJEEgeronimomodulessystemsrcjavaorgapachegeronimosystemmain a committer please apply this simple patch and close the issue +Added test suites plus minor fixesThe test suites are because they caused memory issues during a complete mvn test run in +Isnt this safe +I can do it every time in my application but the object graph is very complex and I dont have time to reduce it down to the bare minimum at the moment +unreleased NPE was fixed in trunk as part of +One more test which has the same behaviour on zOS ran the lang suite on Zos machine machine with the Xjitcount option and found the following are still failing with the syntax error +Thanks Ron! +Committed to +Ive removed it from and marked it as deprecated in +closing resolved issue for release +Or to make the instructions as clear and complete as possible For example I dont remember reading about the need to install the is precisely aimed at making all this immediate and painless isnt it? +The original proposal was to discard location information completelyHmm +I was leaving that for later while I add more tests and expand the API to allow for more inspection of the state of the pool +An alternative is to maintain a global unique term countsuch that when a term is added every other deletes is checkedfor that term and if its not already been tallied we increment the numberof buffered terms +Only tested on Linux and Windows to make sure I didnt make things worse +Documented as a known issue in the EAP and release notes +First rather than using the I suggest you just use the provided in Logj +The WSDL is then read from this URL +Thanks. +One important thing I noticed is that the test was not added to the store Suite +tested and already fixed in axis current trunk. +Spotted a bug in Camel and created ticket on Fuse Camel branch now +Closing stale resolved issues +Im confident that the test failure of is unrelatedIm going to commit this in the next few hours unless there are any more comments in the mean time +Tested and closed. +If you wish to retrieve JSON data from the REST endpoint you need to call one of the two HTTP GET methods for JSON +This improvement has been already implemented a long time ago and so we can close this issueThanksMarco +A good rule is to never store applications in paths with spaces on windows +Thanks Daniel for the contribution +Some months ago I upgraded to Solr and I made some small changes as part of the port +Therefore it is a standard maven jar without embedded dependencies +Contains a new parser for the files that uses DOM API +Mmk I think this has toiled long enough without any progress +Sorry its not a major thing +Hi I met JAVAOPTS is not recognized as an internal or external command problem tooFix is already in branch under this JIRA number +Catenated +NotYes thanks for the correction +There was one scenario that we missed Inner beans with nesting like in your case which led to the reuse of the same inner bean name and hence to an misplaced cache hit +Thanks for the patch BJ + to send QMF command creating a durable Ken GiustiThe result while not ideal cannot be prevented because the cluster cannot be guaranteed to operate correctly in this configurationThe host environment differs between the clustered brokers one host has more available diskspace than the other +So I think we should go the full length and completely remove the use of getResource +In the latest version I added three more datetime related and related test cases according to Russells suggestion +This is in the sandbox for now under +Maybe it happened because of my messy codebase +I like the idea of assembly but not everyone is going to wantlike another dependency so we adding a dependency for running the server without providing another option +Get over your control issues and make it a real open source project with real releases already +Ill commit your patch to trunk +Although +I have looked at the problem and I dont see how we can address that they are just invalid +Perhaps they should be in instead of +Updated license file from Atlassian +Stepan are you sure the failures are caused by this patch? Im able to reproduce the failures on SLES without applying the patch +HelloWe had the same problem when using Continuum and Oracle we used to solve it changing superior to bytes to value in the following filessrcsrcmainmdosrcsrcmainmdosrcmainmdo I think that the way that Raul is a better +Closing issue as it has been released as part of release +I agree +Committed. +patch for trunk +In order to keep under max file size had to remove jar files +Did I miss something or did they break this morning +Of not Im not going to get round to it soon so patches welcome +Thanks Srimanth +Well actually at this point the plugin is not available anymore so it cant redirectAnd its not confusing because at the top the bundles tab will be rendered as selected +When I ran the repro with encoding using ibm I got same as what I get for jdk +I would be for trying to get this and stuff like it into a implementationI do like that idea too if is not a core concern +gwtGadgets needs to be included as a dependency in the file which configures +Done +Also mention this jira number for more reference +And since we had those debates the industry has moved even more to an Inversion of Control philosophy that the logger approach follows. +i dont forget itputting pressure doesnt speed up thingsi will work on the security stuff in the near future again +Good point we can just do the read on the RS first +extends but uses the same namespace +This was completed +The file includes the previously posted fix and an update to the unit tests to verify the behavior +push out to +dont what the query syntax would be for this so toString is not properly implemented +I also replaced the use of sudo with su +Yes I believe more compile time can be saved +From it started returning rows +Lets just abandon this version. +Each request also prints the request name now +Attachment has been added with description FO to reproduce issue +New revision includes doc update +No you dont need to explicitly set it to as the default value is +I dont think we want to keep adding more options to its algorithm but we should describe more precisely its split policy +As far as API compatibility just make sure you dont change any existing public or protected class or method signatures and you should be good +I would delete the code rather than commenting out in +Attached with this change +Renamed the issue +Good catchIt is a bug +If a designer adds that conflict or have an then they have nobody to blame except themselves +Ugh +User provisioning systems like portals can directly create accountuserdomains across regions with out need of event busEven if one uses event bus there are other implementation options with which once achieve this +Ive just committed this +It is impossible to know what kind of corruption might have happened but all the following are possible because derby relys on the info in the log file to maintainconsistent data +Oh ha nvm thanks for rebasing +mjdksvn commitSendingSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestsstoreTransmitting file dataCommitted revision +It would be nice to have a solution in Roo quickly then when people are comfortable with it to port it to Spring +Vinod looks like the patch is out of sync again +The answer is no if the error was introduced by a wrong setupconfiguration + +Upgrade to +Not sure exactly what happened +I think the most correct wording would be Syntax for in ij commands +It wasnt immediately clear to me that it was correct to set the variable once and never update it or invalidate it but it seems like the object isnt changed when the underlying Clob changes even without your patch so it looks safe to me +I agree to what Owen and Dhurba suggested +Fix for issue which was found internally +Test file still appears to have tabs +Attached patch fixes the bug +Then I iterate over the entries using a new transaction each +Thoughts +It seems there is something wrong with the can do things Remove the in Add directive Does that help +This patch fixes the shell and makes it behave more consistently with other shell commands +The following pom reproduces the error +Campbell it sounds as though you are working on the +Almost there with this one +This patch enables multiple different implementations running on the infrastructure +Rescheduling remaining issues for Beta +I applied changes to the class according to your ideas +wsplugins +A on a system package is +Minor patch update factoring common unittest code into utility methods +The instance on is not Serializable +Im resolving it as a duplicate for now +There is a race condition between shutdown hook and the engine side closure of the last request +Will you wait or will you create your own project with these two domain classes +Taobao people Whats the verdict here? If you feel strongly that passing those configs in cluster response messages is required please let me know and I will undo this commit. +This is a second version of the patchThe count attribute is now for the occurence of errorsIf it works fine I will check in +Now update in v patch +This is a patch to add shard support to the +The new version works it just needs to be available from everywhere +the for this patch KristianDain thanks for having fix that +Inpute file name and press finish Problems occurred when invoking code from +push to +This is exactly what well be adding in Spring Integration using Spring s EL +maybe Im missing something is the thread needed for the entries inserted with a limited lifetimeI can of course disable the thread creation using reaperEnabledfalse +Nailed this +The problem seems to be that the doest find a file referenced in your m not sure that the c comes from the C drive you reproduce it with the debug flag problem is due to the dynamic with use windows and have no if you define the to the directory where the is generated +although Im torn on this a little since our custom is simpler and it does provide a good example of how to create such a thingPlease close this issue if you are satisfied +Changed JIRA type from feature request to bug Sybase is tier supported DB +you mean code format? I tried to restrict this as little as possible +On the other hand if I give repair a specific keyspace name it takes much longer to reproduce if at all +In the meantime I believe Ive found a trunk bug that seems pretty serious +This is fairly trivial in Ill take a crack at this +See comment above +Thanks for checking Ram +Indeed thanks for the reminder +Please attach a test project if you cant track it +It also adds additional tests to the model load tests we have for the tooling and added the new camel schemas to the XSD used for validation +Changed the to no longer keep the subgraph read requests in the which can cause when the workspace content is very large or when it contains large binary valuesAll unit and integration tests pass +Ive committed the fix to the issue I mention in the above comment +Isnt node with the higher and more loaded? Can you try to collect statistics on the same node or relatively loaded nodes +The JIRA for is +Whether we create a column family using cli or a table using cqlsh arent the same things happening under the covers? So why is an object created using one interface not visible in the other +With the correct approach none of these poms would have a version in them at allNow you could go right now to all these poms and remove the versions and then create a master pom with all the correct versions in it +Just a random thought would it be possible to not use aliases? Would that have a big impact on the hibernate codebase +I can repro with v patch but not v +Converted tests to and added timeouts to prevent windows hangs +IS NULL and IS NOT NULL are operators +Moving out of after chatting w +QA Closing dev task +Sample application that deletes nodes below the root configuration that might be used for both create and remove sample application +I suggest we set the whole result set as a message context property and we provide another extension function for class to retrieve the required information form the results set passing row and the column index to the function +Generated Java code for records implements the interface so these records could be used as keys and values for the Hadooop framework +Upgrade to revision Im gonna upgrade to SLFJ now +Verify that all instances created are also deleted +The validation of artifacts in Nexus just checks that the deployed files match the expected mime type and that their checksums are correct +No dont know if ODE registers anything we will checkHowever if this is the case then if you were to deploy any other JAXWS service while ODE is installed then it should cause the same error is it possible for you to give this a try +If nobody objects I am going to commit this later today and backport to this will break bw compat for time limiting collector but I think we should do that in this particular patch to to trunk backported to close after release of +Average users wont be able to assess what are the best choices there and advanced users are able to change this directly in the source anyway +Detecting the in entrylogger during addEntry and flush +and it still has redundant castsI dont think at this time we should be upgrading any of our code generators though so I think a workaround for our is the best solutionAnd I dont think we should in general hide any warnings even to users for the reasons i mentioned above +This one was first but I already resolved the other one as Done so this is the duplicate +doesnt let the application control it As a result I think the test for and the conf field should be removed +Good to see the comments taking shape +Recommend you create a unit test that reproduces your issue also you should try testing a later release is a bit old these days +The following classes added in the patchare public classes +We switched to PPR recently and thats when the issue started occurring +Great Ill update a patch for this jira +Added internal map and updated documentation to reflect changes +I understood changes using but I was understanding incorrect part of your explanation about the pom profile +In practice the code that writes and reads data would rarely share a Writable +This can certainly be improved +bq +r rfscholte fix for +Please ignore my earlier comment +DO NOT USEThanks +Ill close the issue if we find a problem in SSL mode we can always it. +Will be included in the next release +pretty small project with no external dependencies +That would be really helpful +file names are based on the most recent stalled for at message logged by the junit runner at the time i took the thread dump +Added new option ignore to ignore invalid exchanges +Final +Or even to be even more consistent with the JPA naming how about I know its a small thing but naming is always important +Hello SeanThe patch has been applied at revision r with minor modifications thanks a lot for this enhancementBest regards RichardVerified by Sean at revision r +batch transition resolvedfixed bugs to closedfixed +However it will probably result in just using the default which for Maven is probably OK +Will do +Split commit and optimize work out since optimize has a stronger need to lock and be consistent +Patches were applied at the revision +Update the patch to address Brandons commentsbq +Changed name of method to rollbackMemstore +So could you please provide the exact URL for downloading +If we drop a new version of the same +Unlike in C isset is not respected when writing optional fieldsbq +It might be better to add one to +Fixed +It displays OK on not tested on other browsers yet +Joebq +All the problems occured because there is a new Eclipse SP released and some of our pluginsfeatures we use for IDE are no longer available under the old version from the eclipse update site +Has default parallel already shown up in the job conf? For it should be same as so do we still need it +Just as a side note the patched release plugin failed during site generation because of cobertura failing to find the class +This just includes in the job jar and assumes it will be unpacked on the other side +does anyone have any thoughts on the proposed methods +commit fix in rev of branch will fix in and trunk soon +Attaching a patch with a similar cleanup for the client code +and here is the patch without committed +Ah of course! My mind was thinking too narrowly on this issue +Patch for Mikes issueAny objections to commit +off by default now Ive only ever seen people complaining about attached makes it configurable this is not currently integrated +Went over all projects once again aligning dependencies and sneaking in an upgrade to Apache Karaf + Verified +I will see though if its worth to continue Marcos work I cant believe kb of work is totally useless + +out of sync with trunk Declare more concrete exceptions beyond renameTo unprotected get addSymlink Remove IO exception declaration not thrown unprotectedConcat getBlocklocations create concat setTimes Throw more detailed exception startFile appendFile i delete File Jira to cleanup a better exception can be thrown removeBlock remove get plus a few others + +Ive removed all the usage of trim in the content handlers which fixes the problem +Youre not talking about the port on which requests are processed but the internal tomcat port +The attachment is corrupt can you please upload it again +Problem is fixedError message is added into log file for both deployment with console and with sh +There are other bugs due to the lack of strong synchronization on the ZK nodes among and +Closing as invalid. +Well we wouldnt be modifying the enumerator implementation we would be modifying the builtins to call the enumerators reset function after were done fiddling with them +Thats right +fixed in apache by fixing the layout and providing an error page +In the meantime Im preparing a patch for this +If the user has only adjusted his firewall the diagnostics packets might not get through and he may wonder why works but probe does firewall issue may come into play with other defaults which are getting picked up if there are any +Ah my had indeed the parameter overridden to map +Fixed +Committed to trunk a minute ago. +and see whats the accuracy s the confusion matrices for a untrimmed run against and run against with minDf and minSupportThe trimmed version did not do as well as the untrimmed in this caseUntrimmedSummaryCorrectly Classified Instances Incorrectly Classified Instances Total Classified Instances TrimmedSummaryCorrectly Classified Instances Incorrectly Classified Instances Total Classified Instances +Sorry please ignore first attachment and look at this one +Updated patch for previous versions +META +This is fixed in the latest rampart codebase. +In case when war is packaged within EAR it is even worse because it may be necessary to put dependencies into EAR in order to be loaded by common classloader between war ejb jar and rar +More than likely this issue is the cause of the problem with Dependent scoped beans as well although the paths followed would be different +For now lets retain DNASHUTDOWN command +With this configuration file and using a Ingres on linux all examples work fine except for the example +i would say for the most part the has been completed +updated license on am fine with the patch! I think you can commit it +But if we know in which version it has been fixed we can at least put a comment to remove that hack +Although Start is maintained for compatbility with alpha versions of Tapestry the preferred approach is to use an Index page at the root level +Im not involved in Struts development so assignment to me does not make sense +Once we have and in place this should not be difficult to addWe also need to factor in +Its a CCE in the same place +Abstract entities are no longer mirrored as of commit dfdfabbbecadefaaa +Minor fixup of assert message +Committed r trunk r x +I guess we will have to revisit once we want to persist the data +after i make sure the tests pass +Thanks for debugging this! Could we have some additional details Do you see this with all clients or is it with some specific one? Do you get the same behavior when running our unit tests? Do you get it also without the patch for? Is anything logged at debug level during the seconds? Do you get similar behavior when transferring files? Your using do you use a firewallNAT between the client and the server? If yes do you get the same behavior without that intermediary +Verified on +s I wasnt aware that signs would get converted +Do you have a test case that reproduces the behavior you are having issues with that you can post +First wed need to figure out if its a broker or client issue +Applied Christophs patch with minor changes Changed all constructors to protected Removed unused constructorsCommitted in revision +that could take a whileDepending how adventurous you are you can compile the branch and run with it for a while +Final summary report is not pretty +Resolving for verification by QE and anyone else who wants to critique the categorization description icons etc +the patch looks good +Thanks this has been applied. +This is the unicode license that all of their data and code comes from +My bad for not tagging it so itd go in before I cut the branch +bq +Okkecan you check if this issue can be resolved? If so can you resolve the issue If not can you change the fix version? in +Was away on a our interaction with you and some on our end we arrived at the conlcusion that no cofiguration solution existed +Raffaele can you please create a separate issue for the second problem? Wed like to keep things as much separated and as possible +The limit of the open files is already set toa big number about or so +Closing as invalid issue and updating Tiles support test suite accordingly. +This has been done and should be resolved right +Attached is output from an http live headers capature session. +Since its been about months before this issue was really responded to Id appreciate a few days to review your patch before you commit it +I just need this with a very small spoon +This patch additionally fixed the typos that Suresh pointed out +I can run it and not shutdown and it will just stay inactive after it hits Creating instance of filechannel +I am unable to reproduce the issue in the JBPAPP workspace +Verified +Not planned for +committed rev afcf +Once that happens then releases are allowedIve not been involved with CSV but interest seems to have waned so I would be surprised if it comes out of the sandbox any time soon +Andrew could you please take a look at this +Attached a patch that conservatively checks all child nodes and properties of changed nodes before they are saved +What if well detect a corruption related to a certain Linux kernel version would we prevent Lucene from running there too +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +I am surprised this bug even existsIn the meantime I am going to try to get around the issue by storing parameters in +upstream pull request merged + +go ahead and commit NPE is gone on +Turns out the patch fails in trunk with on tests +The fix was to add a in the stop method before the stopChannel call. +Attached version tested on +Ive been watching and code reviewing that one +The use of as a test was just me being lazy +If AM fails inside unregister as zhijie mentioned job history files should already be flushed and move to intermediatedone dir but we do have an orphan staging dirIMO making AM behaves like a normal AM failure in case of unregister failure is in the same sense on RM side that an application that fails unregister is deemed as failure +Igor Thank you for your support this has been added now and should appear in the mirrors list within an hour or so. +In order to set pool size per disk we need to create new config option +This is probably because of the problem in srcdocsuser The script binsqoop is failing due to an illegal character inCan you please fix these and rebase the patch? It will be good to do some sanity testing on the build to ensure that nothing else is affected +I suspect an integration layer for Rack will not be the sole attempt at using thrift and rails +The FKUK constraint areas have changed quite a bit recently +It discards all messages sent by myself +A similar class is in the branchTo use the workaround instead of mapping typebooleanuse type + +So it should not be related to the work +It works but it has some side effects like make EL expressions context dependant which evidence the initial problemIn few words once the component tree is created by facelets engine the page context get lost +I have modified the handleException method to return the transmitterobject that is createdThe log shipper bases its decision on whether it should continue with thelog shipping based on if a valid transmitter object is created +derby triage +during off peak hours everyday +I am also receiving this issue with a maven +Ive defined a factory for creating that gets the object you can implement a facade if you have different parties interested in different aspects of the interface. +Opened to capture Ravi and Elis suggested enhancement for future work +Ive tested with Mojarra since early beta and it always worked fine with integrated JSF faceletsAnd again the other question What could be the possible use of an algorithm beyond the that disables facelets +What is the jdk you use? can you try with the earlier jdk with which you could generate the code correctly +Implemented will be in +I am saying all this because I think we dont need any major change in current model +We have patched our own calling code to add extra synchronization around our calls to matches which we will remove when this fix is released and we move our dependencies onwardsIf we do recreate the problem on our machines I will try out the fix hereThanks for your help on this issue. +Would you please attach a diff file on this post for the codeThanks +The output of the above java file once run +Thanks Uma +HiIt was deleted at the codehaus server crash I attach it to this issue +Sounds reasonable thanks for reviewing and correcting them Robert! Please commit the patch +looks Ivan for working on the patch and Ben for reviewing it +I have introduced a new stream class to handle the switching from dvd stream to +Yeah its in the scans +This step should be between In Progress and Resolved +I agree for the option indicating the Marshaller wether to marshal or not the default values since when using the Source Generator all the default values have to be marshalled since they are created from an XML Schema that specifies default or fixed values that need to be reflected in the XML +bulk move to next stream +There might be quite some code already executed before the RH gets invokedPS did you already try to debug into the various methods? +This patch keeps a clean copy of the that the was started with and uses that when it is reinitialized +Chrome on OS X Lion +myTest or whatever +Can you try the attached patch +This would be great along with Apache Enums +Affects also +I will add a design document for better profile overlaying on the lifecycle and multiple rebuilds. +Issue may already be resolved may be invalid or may never be fixed +This is a SLSB that calls the example webservice +Jaredplace them behind the list of current set of specific +on disallowing colons in usernames for consistency +isAnonymous Y +This has been fixedAlso updated the testsMany thanks to Jing and Nicholas for all the help and for authoring the most complicated parts of the patch +Im guessing some level of correlation will be necessary in the rtgov layer to relate an initiated event and a completion event +Reasonable in this case would be a few hours if not several minutes +Andy Thanks for through reviewUpdated with my comments in review board +This issue moved to. +tomcatconf +if there is nodecontentfooand userA has following permissions on above nodejcrversionManagementjcrlockManagementrepwritejcrreadFor this user checkin and checkout calls are successful but it fails while adding version labelIf I add permission to jcrsystem version store for userA then add call also succeed +note check done on trunk not +I have done this before committing the fix but more testing is good +Maybe it also has to do with the default locale +However the potential that the search order is wrong remains. +Its another issueJava validator has limitiation too +This one is not worth adding an elaborate unit test the fix +Delayed review looks good to me although I still dont see a benefit in storing the timestamp + +Is there any discouragement for what Im doing from the framework author perspective? ThanksBilguun +This is question you should ask on user mailing list +dims +remove because we dont support it +I just pushed some changesI hope that it will fix your issue +please review the got icon image from Delete Application actionNew screenshots and new patch attachedPlease delete icon but was not necessary sorry +I works fine if I simply extract the text of the whole pdf +It works fine +It should be very useful to include the unallocated disk space in the global and jsps so as to provide an easy way for operator to diagnose if and when tasks cant be allocated due to lack of diskspace +Will upload patch tomorrow +However the next read doesnt check to see if currentNode null but instead only checks to see if the current position is valid when deciding whether to open another block reader and update currentNode +What about creating a tag and commiting the patch to the trunk? like the idea of tagging with and then getting the patches in order +Issue fixed included in the iPOJO release. +Thanks very much Joe for ungating the SIS site +this patch fixes this is needed cause focus can truly break in many different ways and it may not be tapestrys faultA recent example of mine was with dojoEditor its focus may end up being called before it manages to loadits internal iframe +There are two places in the Hibernate Config view which consistently gives the exception noted earlier +So in your case the user youre using can actually access the USERSDO views but you setup a schema other than the user name so the information can only be found in the ALLSDO views? Absolutelty true +ShiThank you very much for all of your hard work! There are a couple of problems though +The mechanism I use is bundled with a number of other useful services including updating the index by modifying field values of selected documents managing synchronization of delete write and update managing the periodic refreshing of the reader used for search etc +Hi I think Gavin popped on last night and responded to my irc request +Tested on macosx and on linux before applying to branch and trunk +Not a blocker for +I posted the link to this issue to the Geb user group mailing list and asked Luke to pile on +Related to this prior for change moving the isDirty set flag to a position after a +I will fix it and adding some unit test to avoid the regression failure +WFM +Have you tried aut with NRQ too? If not I will run a comparison with a integer index beforeafter this patch and measure query times +Yes the behaviour on result sets not created by jdbcdefaultconnection is intentionalSQL part states Section clause bb If any element of RS is not an object returned by a connection to the current SQL system and SQLsession then the effect is +looks good I see that all have interfaces and default implementations +and where added for these two failures +Something is better than nothing +I think this is being worked on and might arrive soon +Ive fixed the plugin to work with sure wherehow to upload or even if ive got permission +If all you want is to check the size of items after they have been uploadedand youre OK with setting sizeMax to which you say you are then you dontneed anything further from Struts in the first place +Client observed latency analysis parallel minicluster load patch +can we move this out to? I dont think its a regression or is it +This would also remove potential name ambiguities due to how child catalogsschemas can be added in the relational modelThe only usefulness of these constructs is in having designer attempt to represent the exact form of the source metadata +I was under the impression that it was actually at +Should be postponed IMHO and not be taken into account for +Logged In YES useridCurrently the bundles are hand built and prone to error +In the latest document it says Current assumption is that a Custom Command cannot change the state of the Service or Component +Looks good to meWill commit this change +Does this include the possibility of having the code under test and the test cases in separate modulesWeve got some very large projects where integration tests are located in a separate module but we still would like to have coverage reports on that +Since the code is in a test I wont update it for the nits but I will beware them for my future code. +The more drools examples the better +Additional testing Fixes the case when the node you are hitting does not have a piece of the first collection in your alias list Updates aliases and does a retry on alias miss Tries to wait for alias changes as mentioned in my comment above +Chosing the right one is an issue +If someone has time they should probably take this up +This looks very good and makes it a whole lot easier for users to get started using the inherent language capabilities +I am not an hdfs sink expert but I would expect the idleTimeout to close these files out +my understanding is that geronimo run will still pop a windows like before for the user to track startup progress +Then it just hammers one node updating the same document over and over againFairly shortly you can start seeing duplicates in the changes feed +A savvy scripter using curl or some perlgroovyjavarubypowershellpythonetc +This has been superceded by +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Please try latest nightly +This can be documented in the guide or maybe in a Readme +I think as the classfiles does only contain the file name but no path this is a requirement +Akash how are things going with this one? Im starting to work on for +Why was this moved out from being a of +Tom just FYI the scrolling that Dart has done is similar to what I was recommending when you and I were talking about this a while back +I will apply this as soon as the release is final +New patch updated with deserialization code fix was needed in the Double and Float Array Deserializers to allow string based doubles to be passed in arrays + will be injected during bean instantiation any call on its methods will be aware +Thanks Ivo! I totally understand +feasibility for beta but s responsibility is it to start the new shard creation process? isnt it the sharding strategy itself which should be able to demand for newIn this case could we see the as another client of the lazy initialization lifecycle API just that it might trigger cration of more +I can lend a hand if youd like +Alex clarified this in +whereas throws an exception +Thanks Ravi +Will be in the code drop. +This is an area of Derby which is largely untested +Tyrone Where did this come from? Support caseif so you should be using JBAPP +Also I think this is internal and so is safe to do beyond +Otherwise we must patch this by ourselves what we want to avoid +do we need to add a unit test +I see one of them was allocated a normal priority container although Im not sure how from looking at the codeThink this could happen if theres no node or tasks for a container + +This work has some refactoring in it to and provides a base for other jira work +Sorry but we cannot accept the patch as its tied to Spring +So if yourinterested in deriving this further it would be great to see it approachedwith those higher moments too +This patch implements most of what I had in mind +Sounds like the response on this was wontfix heres how you can do it yourself +Bulk close after release of +Maybe someone with more experience with LE can comment on this? ThanksAlex +However we still really need unit tests for these +Ted Updated the patch with above here my concern is close is not success let me try second time to close +Postponing to wont have time to implement this feature in +any startstop mechanism must allow separate control of DFS and and separate nodes for namenode jobtracker datanodes and tasktrackersjvm options the pid matter etc +The cause of the issue seems to be that the Content Type of the HTTP Response is set to texthtml for CSS JS and Images +This patch introduces the specificImports property under osgiManifest to allow specifying additional dependencies in +Do you think you could open a pull request with the unit test and the fix +Committed the latest patch hopefully that finally fixes this issue +Lars can you check if this works better for you now +Example this should be validcreate column family TI with comparator and columntype Super and keyscached and rowscached and comment Stores term information with indexNamefieldterm as composite keyvscreate column family TI with comparator and columntype Super and keyscached and rowscached and comment Stores term information with indexNamefieldterm as composite key +Applied +Cool thanks! I am glad easymock is gone tooThis is a good step forwards to better structure test dependencies in Solr +New iteration of the patch handles canceling checkpoints if a failover happens in the middle of the processThere is still one very narrow race where the cancellation doesnt work but it wont cause a crash or anything just a delayed failover +I dont see this a problem with +Thanks Sergey +I think returning is fine +would it be possible to link the node back to the sequenced nodes via a correlation of related +There is some code duplication in the test case +This of course is not right +So the same configuration should be possible via the API for variables which are created during Why is the history disabled by default? What are the reasons against enabled by default +I have no customizations in or +Patch described in ealier comment unassigning so that committers can review and to fix hotdeploy revision in trunkVerified on by redeploying in deployRakesh thanks for the patch. +Bryan I agree this looks like a serious bug +One item of note is that your attached ears contain exploded ejbmodules which is not supported +Maybe someone out there could test it +Marking this as as much as I would love to get it in for +bq +Even if doesnt know how to deal with schemas meclipse should not remove it +talked to max and he disagrees he wants the current naming scheme +i think columnar storage might already be doing this +This way It is scalable +Treating them as data sources where they are loosely coupled can be done but that needs to more workCurrently when a translator is removed the vdbs that depend on it stopped and once they are available then the service starts again + +Assigning this to myself since the DOMDTM experiment is mine +Published version of the guide Using in FUSE ESB which contains a chapter on this topic. +Testing on Linux and iTerm have completed successfully +I committed this +I thought it was readyI am surprised that has a problem as it is the that reads from the +removed that and recompiled an it was fine +I know how to use these apps for testing +Looks like this was committed back in April so closing this JIRA +I suppose we could deprecate +Hmmm +It would be an easy fix +Can I propose this patch it simply removes the extends Writable bit from the output and output parameters +Committed this +Thats the reason why the doesnt implement Serializeable +This cant function as a remote aux. +I need clarification from Mubarak that that is indeed the case +m already working on pluggable sendersThilina +it helps now the example is workingmay be definition of in to make example is a best solutionor it is possible to add in like comment and make a note at startup pagewhat in case of exception which I got the user should Mark as Deployable the file or uncomment inso it will not necessary to search why login doesnt work from the start +Ok I could be wrong but IIRC the idea of having a default Project Group was that no Projects would live outside a we allow renaming the default instead of deleting it +Weve have tested it in a with the following string that debugging it returns a float with the value which clearly is wrong +Raja I am attaching an updated patch that incorporates your advice to use abspath and where applicable to improve readability +Update the parent and feature for it in revision +Patch is for Branchx may require some manual attention for will not be a moving to for in AS trunk as part of +The aspecjtcompile works fine but when I call the mvn install assemblyassembly with a descriptor it doesnt includes the correct sources +LCM should not be used +Closing as duplicate of. +for this patchIn terms of why we have findbugs I agree we should hook this into a CI build +ahhh + +the radio needs to tweak its markupid in a certain way mainly to support javascript traversal of all radios in a group +As you pointed out this was just a compiler glitch. +We will try but it would be better to provide war sample +please remove the LOG +committed to rev as part of closing issues resolved in JEXL. +I fixed Slava please fix issue opened for +unless the constructorfactoryprovider does something special to keep track of them they wont know anything about eachotherok so I think we can go with the second option of having the serve the to Ill post a patch later today +We need to give the keystore portlet a normal gbean name not something hardcoded +Since we intend to release soon and since we think the change is too big even for I move it to make the issue as blocker for so to show that we want to absolutely find a solution for the issue till code properties and default values arent really linked those two issues are about a problem of the joint compiler to properly handle Delegate +You need to add those changes to app aswell +Changing to also contains a lot of unnecessary of existing code which isnt ideal +The WSDL is a standardized XML document with a few public +I have verified it +Id actually like to drop them down to Debug level unless theres some objection since under normal circumstances these generally dont hurt anything and can cause some confusion especially around failover which is a commonly used feature +I can confirm the bug +Thanks Valentin +This patch is to applied on iPOJO project +Yes in pool +I think it is not needed +I put them into the first array out of order into the second array in order including the two that are lexically equivalent +Sorry first invalid exampleselect e from e Entity e +Or if there is even a problem +Merge with +My query is I want to download any of the following versions and MSo please provide me the details about these versions +Thanks for the confirmation AleksFixed + is a repro of the bug in the form of a junit test to be run from the codeline +Proposed patch +Added to the features +Tested it on am resolving this can close this issue after verification +Im happy to help updating the wiki but would need some tips and pointers on where to get started +Problem is with flow definionts and XHTML pages +I ran the sort benchmark with nodes java heap size mband the results are as followssourcerun timetrunk hr mpatched hr m +This chapter has now been included in the guide. +Without a test case its difficult to provide you with any assistance +Care to expound a bit +Proposed patch +And then with patch applied will make sense +I am tempted to suggest that we make them stateless but I know that will cause some trauma in the DBCP JNDI setup +Adding in the base issue I just filed +yeah we will move + +Fixed. +Any files in that directory will automatically be included in the jar without any additional pom config +This is a very useful test +Need a for the container for as need a beta of the mc project before we can release aop alpha +I have full access to write files in the install directory and I am not behind a proxyWhile my preference is to be using the update site I will also try the workaround above using the zip versionThanks +for documenting this layout on wiki +If this is not the behavior you are seeing then this is likely a bugIn a negative number will result in an +I will try to reproduce on the latest and repostSorry for the confusion +Also there is a jira that proposes to deprecate Hadoop recordIO so marking this Jira as wont fix. +Thanks for patch Jeffrey. +Will play with that in trunk +Committed revision +removed from affected versions the problem is resolved in. +This can of course be debated but will not change the past +Looks great Andrzej! to commit +The fix contains refactoring of interior pointers to static fields support in inNow both IA and EMT version works identically +Sorry for the blocker but we will soon decide whether logj will be dead for the next few years or not. +Depending on how fine grind we want to go to we can treat them both as file reading error or only catch and throws +Permissions within are a whole other kettle of fish so I think Ill keep quiet until is committed then start another JIRA +Yes it is forecast temperature data over the United States. +commit fceadfba fixes apis list list list list list listCounters listConditions list list list listConfigurations list listClusters listPods list list list list list list list listProjects list list qa all commands mentioned in prev comments have to be tested +Do you really have GBsec of monitoring data +Now that work is underway with AS all previous community releases are +Can you explain it more detailly and I also want to know what is consequences of the race +Verified on EAP ER +Ok I think there is another ticket for postpre hooks for configuration. +BrancheActaully I have committed the codes to bumped to for now +We will get that into so we have a jboss and jboss component for ppl to use +catching up on this one +I checked on Linux HT machine in all modes +sigh +Thats probably due to test methods being executed in a separare thread +Also the test finishes with in mins of time by when datanode will be marked dead +Right +See the PR for the initial changes regarding the Configurarion need some test cases for the configuration of custom identity stores +Yes this is meant to be a feature and the definition of parsing normal form is in the Avro spec as a resultTo aid in assuring interoperability this patch includes test data that should be used across languages +After both of these errors are fixed an primary key violation error occurs as a result of create being called from create of +If we can fix these that would be great the table names is an obvious issue to fixin any case Ill create a task to update the README and documentation with your steps above so that other users dont need to go through the same troubleshooting process +Im in transit at the moment but can send a patch tonight +Id love to take a crack at it great exercise to ramp up onAre we over any sort of feature freeze for? I can put this on the front burner if its only a question of timing +Move feature requests to the next issues to next release +im attaching a test right now i just did a fresh checkout of xwork and used issue is still broken well a variation of it basically it just gets stuck in an infinite loop try the attachment i put up and use the newest xwork from cvs and youll see it get stuck in a dowhile for the testcase +Will wait a beat to see if anyones got comments aware this is holding up other changes though so will try to get this sorted tonight +Paul can you have a look at this +No objects are shared between threads +is this still an issue in +for svnignore sounds fine as committed +He told me to update the issue with this request +I just reverted the changes to so that the regression tests would pass +Heres an update the compiles against trunk +Michael you still the lead for right? If so Ill assign these to you when the pull request is ready +comitted +Im convinced +We should say explicitly something like with no options given prompt the user if there is a name dir in the local file system +Fixes javadoc and javac warnings +Committed the changes and requested the pull request +Thanks +This is implemented and the tests work except on win systems where locking issues with the vfs code prevent removal of the uploaded test content + for fixVersion +Jan we need more information about what exactly you input into the Security panel +Sorry should have been in FP release +This actually gives us a chance to try something out I have been curious about and that is how hibernate deals with schema changes +There is only the dependency +For example is stuck using metrics until all of our users move to since we need to support both and and the metrics API is incompatiblebq +Since annotations are not available in other compilation units we cannot know if a particular type has for example a singleton annotation on it or if any of its members have possibility and I only just thought of this now what about setting some bit flags in the if it has certain kinds of standard AST transforms on it? Could you add a custom modifier to the modifiers field to store this information? Its a bit risky but it could save a lot of time because I dont have to go through all annotations on a unless I know that it may have an AST Transform associated with it +The integration has been completed in AS through +I also added the source jar to the distribution +I spent some time looking at this again yesterday +Thanks Colin now I remember Ive been bitten by this before +Documented in the devguide. +new patch coming that adds more parameters and fixes a problem with previous patch +Laura you are right about If a required privilege on one of the objects referenced by the view is revoked then +Thanx a ton for the review and the commit Rick! has been submitted and committedthanxNarayanan +Alexey sorry but I had to revert your changes at +Everyone Its great that this has gained some interest +For this issue Ill move available and unused to +I will take a look at this for m moving it to future as I cannot someone prepare an example application +The linked issue is now solved +This behavior should be corrected when we fixTouches the following filesM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileBulk of the changes needed to support UPDATE actionsM javaengineorgapachederbyimplsqlcompileChanges to distinguish MERGE actions from driven by +Verified in JBT +Like you suggested above I syncronized struts so that only one could run at a timeI think this is a good design practice in general but I dont know how itll end up affecting AJAX componentsIf nothing else Id love to see this as a configurable option for +I had to remove the stlport include and i believe i also removed the XALANNONAMESPACES define +What do you think +Ive moved your attachmemt to that issue +The main purpose of this issue is to get the proper number of partitions +Gridmix V needs to have support for Sleep jobs +I need to think about this some more +It would need us to study the new API and understand what parts of it can we leverage to our advantageRight now I really dont have the time to dig in to this +SVN +Can you please create a new ticket under the website component and attach the how to in markdown format for this +I will hold off with this patch until is committed dont wanna interrupt roberts work with this cleanup here +Yes agree on the dirty fix for thanks +This may be a follow on issue but imacat can probably handle it. +Therefor go ahead and commit +Pooling of course exactly gets you in such a situationSo what do you think about removing the pooling of the objects +Is this really a bugAs a user I would personally not complain about a documented limitationwarningthat this method should not be used for maps that may contain cyclic objectgraphs +Does the pig build system set up some audit rule that do not allow to change the constructor of public classno +ThanksIve tried the nightly build and the problem seems to be for letting me know marking the issue as transition all resolved issue that did not see any further comment in the last month to closed status +One corner case we need to consider is the interaction between regular expressions and topic names +Patch handles zk node in RSZKREGIONSPLIT or RSZKREGIONSPLITTING state +If I use a troutputText then it works +A subsequent select issued from the first SQL WorkbenchJ window completed successfully and I was able to see the results +Is run the Thrift server okWhat do you mean by experience? Do you mean examples of usage in different languages +Attaching the US States example WAR file +I will change new +Remove fix version as this is dependent upon getting fixed +Thanks for the reviews! +Thanks for spotting this +Thanks for reporting this +rev added a couple of files that maven wont process since the normal maven processing strips out all the useful comments I put in the original poms +I wait until tomorrow before I commit this fix +If no input row is present the output collector is not present check for that in reduce sink operator +Log file attached +I raised the priority of this issue +Im first going to try jacking the timelimit up a little bit +I happen to disagree with both of those assertions +I will fix the problem for both applet and applications ifif it would be OK +please reopen if you think there might still be a bug but again I havent seen issues in a very long close for release +Id rather it was implemented as an extension of where wed have a partitioner that sends to all SOLR instances which is I believe what is about +Thanks TimThe patch is applied OK boost on the attached microtest achieved. +Im not really concerned about adding a few ms to a scenario +its created before we know how big it should be so it grows as needed i have tried both actually +Rolled +back ported fix from trunk to and setting original owner back +The split in requires to propagate changes between projects +Moving to next release. +Made s sessionKey property configurable +Patch for making Multi Part Parameters available in the Groovy Event Too +attaching a new draft patch based on Owens suggestion +Please review this +Do you think it is okay to implement in on TT? Please suggest I ll upload a patch accordinglyIf interested the user should be able to configure this timeout to be shorter than the to ensure a single attempt. +Applied. +Perhaps part of the process of deleting a User requires any ownerships to be transferred to another user first +I just committed this. +This change of behavior breaks existing applications that happen to have persistent business object attributes with names like Category Reference Sequence and several a very minimum Groovy integrators should be able to control how names are resolved in the script so that we can maintain upward compatibility with all of the production apps that will now run into this issue +The bugs have already been fixed +Apparently my first patch was brokenHere a fixed one +If there is anything I would add it will be the actual link of deregister attribute in the Derby Reference Manual +How would Hibernate determine which decimal representation it should useHow is this SQL generated? As a workaround couldnt be bound as a parameter +This has already been discussed elsewhere + +Im going to post this question on the Java Lucene mailing list for some inputErik Maybe you already know about this issue can you comment? ThanksRegards DIGYIm going to close this issue as Invalid since it works as designRegards i believe that small design change is needed this issue can be closed +Patch for issue +Its a separate issue and i think jev already did those +I just committed this +TESTS derbyall test suite passed on Windows XPJDKIt would be great if some can review and commit this patch +Anildid you get a chance to review Abhijits patchIf you need help let me knowJacopo +Tom could you or one of your team please supply me with this information so I can add it to the docs for Rick +I agree with your comments about not penalising all clients because of IEs behaviour however I first experienced this issue with a Flash client and not from a browser XHR request +But if theres no other choice +But I do agree we shouldnt be generating things that xml parsers wont be able to deal with +Thank you +Nevertheless the primitive types are compliant with IEEE +includes some example classes under +If it is not the case please record it and close the bug +I vote of leaving the temporary location to blank giving a chance for the user to think and type +If all threads synchronize before the method stuff in cache wont be picked up until another thread is done looking for a another row +Commit improved tests and reload command in a momentAnother thing we will need before too long is a way to get a response I think? Right now the client cant learn of the success or failure of the command +Is that proxys reference null for some reason? Anything special about how you create that proxy? looks like that proxy is your reference since it is being used as a key for transaction resource binding +Thanks Christian +jdtcorebuilder fails +I forgot to mention there are commented out tests which demonstrate this in the JMS integration test module +libexec and use the property name in the env element +It should be possible to generate tables automagically based on what the demux actually produces +This should to be added to as seen at the region server right +This seems to be working OK now. +Thank you! +Here is a patch version of what is up onAttaching so that hadoopQA can take a crack at it +the close for Solr +Hi MahdevI ran ant test there are no failures because of this patch +I am currently working on implementing this and plan to post a patch when I have some basic functionality that passes the basic test suite +Thanks Dave +Ideally there should be some kind of a version attached to +I left some comments on RB +We see this again +If it saw rebooting and it had zero sessions then it restarted itselfIs there some time delay before tablets are reassigned? Can the tserver restart within that window of time +doesnt solve our problem thought +The fix has been tested by IJ test suite and by JDBC TCK tests and committed to the branch +For instance the tag element in the scm configuration in the pom isnt updated it doesnt track things properly but it does solve the releasecheckout issue +I must confess that I dont use Spring +Thx Willem will change all resolved tickets from or older +Thought about this againAnd I dont think this solution is validBy convention all crypto functions throughout the SDK API accept explicit providerIf you can find another exception to this rule I may understand and drop thisBut if xmlsec is the only exception than it should sync up +Reduce progress seems to be +Shall the version number just be tacked onto the end + for option +Many methods are updating the value of dst is should also be updated in those places +Thanks Vinod! +Most of test classes in the patch pass successfully few have failures which may be bugs in MR or further changes to be done in the testsTo run the tests please apply first the patch from +Great timing After an update of the sources I was able to build the snapshot release ofUnfortunately the error still exists +I believe this is the same issue as that reported in where we are prematurely closing streams we have adopted into Ruby IO objects +ping +At revision remark is a good point fyi I checked where the oro jar might be to find +I downloaded the newest CR and yes I have zero projects +Make your proposal public and post a link to this JIRA +Im now gonna try with the attached remoting jar file and see what I get for this test +Patch looks simple enough +Andre can you please verify this and close? Or will you yell again when you see any more problems and I can just close this for now? +Fixed in revision by changing to use Flumes +Confirmed +I think you should include only the RPC port in the notes attribute and not the entire URL +I build my at my own only using the stripped src without additional resources and got a handy size between kb and MB +I tried locally and passed +Thanks Brett Ill get this fixed in the CVS with your patch +Please comment on implemenation and usefulness and on the fact that it only deletes existing records +The log file which Ive described +as a package separator and inner class separator +All duplicates have been removed from the Maven repo +projectgroupmembers list has broken links in the name of the project and the remove link +This will get fully fixed after we refactor the Jar file support for +Will this change the readpoint? Should we add it to snapshot as well? +Ok I just disabled the exact string assertions and log a message to stderr if the match does not exist +This issue is also still present in the latest release +or shouldnt need to change at all +I also want to use OFL as part of Font Awesome a Twitter Bootstrap plugin +I added kristians patch to make us less bad +This is possible with the camel jdbc component +I can imagine that it can be some specific usecase maybe someone needs a Disposer that logs registrations or allows to add callbacks to be called on disposal of specific objects +Isthere something fundamental going on or did the original code just pick anarbitrary constant to use as a search chunk size +I will take care of the white transparent Jacquesthanks so muchYour work is in revI just changed the of and the tabs still have a white background +Couldnt reproduce the problem Ill open another ticket when Ill find the root cause. +Would need to backport the fix to in order to fix it for the editor +Ive tested on two servers with self signed certificate and with a good certificate +Verified. +i wanted to get this change into trunk since i have another patch dependent on it but am happy to address any review comments in a subsequent change +Agreed +probably too early to does not have regressions for winx +Ive implemented something similar override and extend the component implementing the get and put methods +What about a of an exisiting workspace? Will it still contain the variables or are they replaced by their values when the is created +the result of a query may be different because nodes from the version storage now match the primary type from the nodes in the workspace +This said Im completely fine with having such short term solution in place. +No comments from the Go experts around +If this callback interface is encountered elsewhere an exception will be thrown on bean creation +What you get is the maven build infrastructure for a webapplication ready to use shindig +Patch attached +If it is declared as Object array the JDKs varargs handling already passes it in as array +Patch committed +So thankful to Flavio and everyone else here for helping me sort that out +Please verify + +Thanks Harsh for your inputsUpdated the patch against to latest trunk +dont merge partial filters after closing the output but instead when opening a Reader read all partial filters and pretend they are one +The new state Im not mad about not without our first having done a review in toto of all possible states to figure what new states were missing but Im grand with this going in +Marking as duplicate +I was not able to reproduce the stuck in loop trace cursor got duplicate is expected and many of these are expected as all of the messages in the store are replayed when a batch is refilled this is because with priority ordering the index of the last message replayed has no bearing on the next message in priority the current code base I was able to produce messages and consumer all in proper order +Attached is the new patchTested the same at httplocalhostmaincharts patch with recommended changesTested the same on the UI +After making this change both Count and Time Last Update are getting updated appropriately now +Okay Ill work on thatOddly I dont see your request to subscribe + thanks +Thanks a lot good job +I just committed this. +Apply the changes in patch + +There are help pages for the DNA Servers view DNA Message Console all the dialogs and wizard pages the preference page and the context menu +Im emailing the geotools list to determine what can be done about this +They were flushed in a visible way and I dont recall anything saying that one shouldnt open a reader before the writer was closed +Thanks Mike! +Incidentally I am also using acegi on myEclipse JDK upd Project JDK upd Groovy plugin MarcusI have done exactly as you have suggested but no luck +This means the output is basically a mail message with the sad side effect that encoded output is around larger than the original binary +Improving the logo should be addressed in a separate issue +This one got overlooked +Currently HDFS provides only guarantee so priority covers all rack requirement break casesIn methods add remove startDecomission and mark in put both and rack blocks into the neededReplication queue +Instead of printing stacktrace we have used info method of Logger +So Im not quite sure whether we really got a mismatch thanks for looking into it +Stevo just a comment about the last sentence about mvn package and mvn siteI dont think that unpack goal is run and it succeeds but I think that the report executor notices that the build has been already done and then it does not retry to build it + +I wonder Dave wanted to say that catch as generics exception type cannot be implemented naturally with erasure +it +If this is because theres a desire to have functionality lets just use spring +Verified in ER +It can be passed as a boostsharedptr to the and classesThe handles X cert verification and is inherited from both and classesMultithreading is left out as I feel that it is already handled at the levelBy default only is supported now and both servers and clients supply certificates +Hi i think this code could be a starting development point for metrics +Thank you Arpit Chuan and Chris for the review +Sorry I hit the create button too directory structure of concern is configuration configuration configurationWalt s an integration test for this issue +I added one +index scan depending on isolation level often one had to set a TABLE lock rather then individual row locks this is a particular problem with serializable isolation +This issue has been resolved for over a year with no further movement +The method is very similar to get method in +This should be fixed now +It is not random although it may not be completely obvious +could you pls try with a more recent version +on patch +In order to get automatic setting of the compiler properties you need to add the following line to yourautosetPropsonOnce were confident that this patch addresses the major platforms we use we can make automatic property setting standardThis patch also adds another diagnostic so that the wont stumble on an array bounds exception in the situation that Myrna tripped across +Moved to subsequent release not subsequent patch +I think that we can make it a little bit more general to cover other since others are sure to come up even on cases +I added some additional and could scroll the entire list +I couldnt reproduce this bug in. +with all changes we discussed in one patch? orB as an incremental patch that just adds the check for the single patch addressing all the needs of this jira generated as svn diff from your checkout would be +I dont have this problem on linux +The ability to Override is now essential for these classes. +I will run the completed test to see the problem +Due to the fact the does not demarcate class startstop in any sensible way in the notification api getting this issue right is a lot harder than it seems maybe even impossible for parallelclasses +Removed and from Fix Versions +Mike thanks for giving us a great clue for this issueI checked in the fix and deployes JAR +I hope you read it +I separated bulk startup assign and general bulk assign +OK I committed Ignore of +Reviewed that patch from Paul and it gives an error dialog if the zookeeper password is wrong +Attached the modified class corresponing SVN patch +Moved to upon release of +committed trunk and x +Added a first implementation with revision s now possible to add a and to the config directoryThe webapp +Clearly as you said cos these threads will block youll need to size the pool correctly to avoid starvation +Fixed in Apache Karaf. +Thanks for your help Juergenchris Ive tried latest nightly snapshot and it fixes the problem +admittedly this isnt the exact version we ran on our cluster so there could be a difference but it passes tests Im a little stymied +We use the base name with common directories but the full name with directories into jars +Move to to harder than you think +the green bar is only shown for your own video the green dot is shown on all video viewsSyncing the green bar to all video views is impossible as it would mean you sync events per minuteFor the users own video view you might be right the green dot is duplicated however I tend to say that it might be good for the user to see the same what everybody else can see +Closing all issues. +There is no longer a separate capabilities class +add a jar file the new projects lib directory +Will prepare a new should be renamed as isnt using a proper properties overriding order when reading the is now fixed on top of that one +Your approach is better +Seeing the version every time Remoting gets used is annoying so it would be nice to log the version at DEBUG level instead of using +I dont know the details of new similarities so its up to you Robert to come up with suggestions +Paul is there such a test in AS clustering suite I can copy +autocompleteThese files are not modified and still contains prototypejs +Alternatively use CXF specific Multipart annotations or just enumerate the content +point of description on review boardbq +I would imagine its taken care of and can be closed +See the attached +Patch against trunk that adds licenses to most files that lack themSome patches look bigger than expected as I created them on Linux and many files lack the svn Subversion property +I think I have finally gotten full text search to work given Razvans comments +Justin +I tried to use from SVN branch which claims to hold but that is not working +I understand the delay when the artifact is written directly to the file system +But throwing an exception there is not backwards compatible +But perhaps snippet generation might still be useful for long fields +verified on to set release note note is not required since the entire component is new to issue now that release note status has been set +duplicate of +JBOSS completes initialization and I am able to shut it down normally +In most cases the JVM font manager just throws an exception but this one leads to a crashThe goal of is not to use the fontmager anymore +We have seen that type of thing a lot recently we fix something like this and it just breaks a couple months later in a new refactoring +I dont have any particular concerns I just like to understand motivation behind spending time backporting things into rather than spending time on forward development +Making it for since it seems a bit risky to change the way we deployconsume stuff so close to release +We just tried the patch and it seems somewhat ugly without the fonts +Added couple more fixes and unit tests Error handling logic in puppet executor to gracefully handle if java home is not available in config Remove unnecessary error logs as DECOMMISSIONDATANODE was not recognized as an action +However +I wonder if the SQL spec says anything at all for such implicit and their collation type +Checked +Here is the approach that works for union collections +this looks good + addresses the same issue but due to invalid test fixtures of contribstreaming of not deleting files properly +Ill comment there +Please help me +patch applied and committed +reassigning to Laurent +OK I committed the missing its not insane if a field has both Bits and array entries just missed this on backport for +Maybe putting a flag inside the polling endpoint and adapting all extending endpoints of components to evaluate this flag +Since we have some valid programs that arent parsed correctly can this be documented as a limitation and deferred +The patch +I looked through the code and found almost no examples of its usage +Hopefully I can dig into this tomorrow or Thursday +failures are irrelavant to the current issue +I forgot to mention that this causes an exception on every watch fixed in the released version. +The basic use case is +Please help to review this new one +Thanks Namit +Committed +Patch with section improved and Link from environment to just committed this +Thanks anyway for the work. +They cant just be on the hibernate properties since there can be multiple configurations for one project and the ui wont fit but we could definitely do with having a Details +The exception trace gave it away +Isnt this issue for the WAR Plugin for Maven +We should fix this +Committed to trunk and +This is actually implemented but the problem is now the displayed information isnt responding accordingly +visiting or causes their child paths to have forcetrue +Very small change to the patch to make the path abeff rather than abe +Done +Brian youre right +It seems the person uploading the JAR never creates the are we supposed to use those if no pom exists +Could you please add some javadoc to the state enum values explaining the purpose of each state and what the transitions are between them? Or augment the design doc for with this state machine and reference it from the codeSure +trunkapplicationsHogstromdevgeronimotrunkapplications hogstrom svn commit m Database portlet version issueSending applicationssrcjavaorgapachegeronimoconsoleinternaldbTransmitting file dataCommitted revision +See also and +His email isUser parrtDomain then changing the directory structure this code is the exact same as provided in the original it be easier if I asked Terence to attach this version to the wiki page as wellIt doesnt really make any difference to me which version makes it in for version +I would be in favor of removing the current concept of rack awareness and better documenting the way to achieve distribution among racks +Jochen perhaps we could have a look at this issue to see whether we can put the fix in +Attached patch for trinidad +Hi Yongqiang I fixed +Grid can you review my patch? I just want to make sure I didnt break anything else +Assigning to Robert to commit the patch +This works for me now with trace testing enabled +I currently have no idea how to implement that so you can really depend on it across all situations +In case I didnt got the message across The idea is to ease the step +HiI didnt wrote an LDIF in the reportIm using iPlanet Directory Server service Pack and I can delete the attribute using an LDIF or Softerra Ldap Administrator for instance +I went with Bernhards solution because it did not require me to add any additional files to my project as a workaround. +Attachment has been added with description Version has been added with description Version modified to basically do the review and comment +This is an expository cut at a Sampler library +Sebb I have added the test cases in rev +Patch was created on Windows XP from geronimo root Sending modulessecuritysrctestorgapachegeronimosecuritynetworkprotocolTransmitting file dataCommitted revision trunkSending modulessecuritysrctestorgapachegeronimosecuritynetworkprotocolTransmitting file dataCommitted revision +Mike has advised me he is manually running a script to update this panel +Thanks +Given these numbers its almost like we should start considering splitting writes which doesnt make intuitive sense +What you described at your last comment may get a bit difficult and may need some more discussion +Perhaps should implement Configurable and call getConf in getProxy? It should still probably make a copy of the Configuration object though since its mutating the client retries setting +I forgot to include two new Windows build files +Attaching a patch tested it on HTTPS enabled environment +I think its very useful having it on its own page as users look at the naviagation and see that heading +There really isnt a good way to break up the patch its just going to be large eg we cant separate out the terms dict from the RT postings etc +Maybe someone with commit permission of project can fix this problem +updated the order of to successfully run with Java was tested with java and java with resultsTests run Failures Errors Skipped Thanks +If we switch to this configuration we need to make sure that archetypesquickstarts will also be released this way and that openshift and the target release support this +lgtm + push out to +Strange Ive been running for a few weeks now and its hung together +This list seems reasonable +Based on the discussion at todays contrib meeting it sounds like we can drop this patch and instead backout and +to the latest patch +Applied the patch that updated some of the content in the Getting Started and Reference Guide documents +After discussing with Brian and Jason rejecting given that if its not going to get done for Id just as soon see it rejected bumping the schema in a later release for something like that is IMO a worse crime than having some attributes on the subsystem element +also need to add documentation for flexible block size and part file sizes in archives +Not sure what but cant reproduce it any more after upgrading the test cluster to. +Testdata to run with this is related to +Patch applied +file contains sources and configurations placed relative to the path +What we need to do here is to first fix the error so that if a Job runs over the counters limit it will not cause the AM to crash violently +I just committed this thanks Ahmed. +This is probably related to +In the cforms block on is the same filethat means there is already fixed too +We can always change the number of release date later +A BPEL implementation must rely on a security framework that guarantees the integrity of the exchanged messages +I dont see any obvious reason why debugging should affect the user experience in the way you described +I will note that we have serious issues keeping running at our ve tried without luck to reach the developers or locate where the project landed +As part of the tests listed above were all to support search and all pass. +The change was committed to the latest development code +I looked through the new patch and I am still a on this change + to the master +If someone set the number of taskstip to it would be a big problem +Ive been experimenting with GB memstore flush size on a LZOd table and saw radically better write performanceI think it would be good that we bring the flush size up as we bring up the split size since this means less regions per RS leaving room for bigger and more efficient Memstores +Christian great work yeah its a good plan +Hope I made the patch correctly +Can you verify it works? After you do please update this issue to closed. +Looking into adding TRACE logging to the REST suite the errors no longer happen +fail in trunk and due to this +It turns out that the DB and Derby dialects will default to a maximum length of bytes unless you explicitly set a parameter on the Column annotation +I wont be too bummed if you disagree with the role of Hibernate Tools as I view it and decide to close this one out +Bulk defer all frontend issues to +currently there is one uber bundle that contains along with the brokerclient etc +Yes it makes sense +Were there any prior exceptions before you hit this oneIts hard to see how could have both its indexReader indexWriter null unless it had hit an exception on one of them +Let me know if youd like me to add some testsKathey is this what you had in mind? If so Ill commit +Agree making the parameter types use generics can only be done if the lower layer also supports it +Ive attached a new version of the logo that should match the blue used on the wiki and in jira +Hello! after some FP research Im going to open this at least for the foldr implementation +Left a comment in beforeClass saying so +I see problem only with attribute style +The attached patch is not a real solution to the problem but may be a hint to what is really going wrong +Also includes a patch for this issue +I originally saw the issue in in January but then in. +The only issue my current project would have with this is the Map which prevents multiple fields of the same name from being added +We can open an issue in due course to address the additional issue commented here and link tickets +If not you need to pick one Fixed Version +I was waiting to make sure this worked before applying to all our files +We develop using Git and the Git mirror and so were unable to produce a patch fileIs there are way around this problemThanks +Is supposed to be +I went through the classes in package and things look OK there +The now uses the highest version of all pdfs for the destination documentThanks for the report! +closing since this is checked for this Ill try to incorporate this into the next version +Seems there is no activities on this JIRA for almost a year now and seems to be close to be releasedShould we close thise one? +Simone This is far from a trivial task +duplicates +To the next release +It should just work if you run it with the right Hadoop configuration +Blocker as the env has and going to be removed +Right so this and are the same issue then? We should close one of them I guess +I didnt have any problem finding my way around the Spring Framework and Tomcat source codes when I first started working on them +call +Only my web project shows the problem +wsu and wsse which seems like it is correct + version deprecates trunk version removes +the merging loop has a magicMarker which helps with setting the offset when copying the arrays +What is your custom authorize endpoint url? If you can see how to fix it even better +I guess this is my fault Ill fix it upload a new version and post back here +Committing this patch to trunk for now while we figure out as to how to interpret the output of streaming kmeans clustering +Run the runxalan +Will work on backporting this to and in trunk with revision in with revision +Agreed +If we have many candidates it can take a while before we find a valid result so if we check the time for every candidate it will be ok and we wont need an extra threadthoughts +Fixed. +I forgot to mention that files to were declared bad crc +Ive just committed this to trunk andThanks a lot for the contribution Andy +XD is deprecated +Seemed to start after the +We ran into the same problem after switching from mongodb version to against a Mongo +Thanks for the patch Ren +There are no get methods yet +I even was able to update from the update site this time +Nice work! We should have this long time agoBTW do you want to also check characters per line? But it is arguably if it is the rule we want to enforce +API documentation still needs to be written additional tickets to be created. +IvanI am not saying the reordered write performance is not good no merit compared with approach +I actually built pig from branch using the instructions on the wiki and that rpm has usrconf where as the released rpm does not. +Now disign mode causes a crash of eclipse on resizing so its should be fixed +Resolving as duplicate of +patch withdrawn review patch update is available at Review Board +It doesnt seem to be related to server load +Hi Han Thanks for reporting this issue +patch looks fineWe dont have consensus on how to handle coprocessor API changes between even minor releases + +Hmmm I do ship the image io now though so perhaps this issue has been resolved +Ready to you update to include it in the tarball +This was a big chunk of work thanks everybody +not urgent +And thanks for this patch +I filled as are right +It would be nice to fix it in this being looked at? Is there a dependency on another component + will commit if tests pass +correct document attached +Ok Ive uncommented the asked Gavin why he had commented it and hes not sure but we need to keep an eye on this +Any comments on my changes to? have added a few test cases to the latest Mohammads patch +Lets move the discussion of this over to more relevant +But it would still be nice to hear more details on this especially with WAS this time I am resolving it as Cant reproduce and there is not enough information from the realizing that this is one of the original functionalities of Sprong Integration that have been used with WAS MQ by many clients we can safely assume that if this indeed was an issue we would have many reports about it. +Thanks for the commit ClausI will add the documentation as part of this issue +It is no more appening on current master +In addition it mentions coercion between BOOLEAN and rather than +The patch for was reverted due to issues in +Pushing to since they are not critical for and because were approaching the end of +Please ask other questions on the mailing listthanksdims +In Derby authorization mode this value is not used anyway +thanks +Im observing an interesting behavior +Between those two I dont see an AS installer like the one described here happening. +derby to trunk with revision +Sounds like a good idea to provide a warning at least + +Ill commit my patch next weekend if nobody objects before then +noow I cant upload testcases in that format but i will do it in advance +I plan to commit this later today +Something to try ifwhen theres a next time then +HiDo you have an idea for a solution yet? Regards didnt have time to work on this yet +Committed to branchx. +Well be working around the bug for Grails to get this working. +Thank you! Well investigate further on this end and report back +Updated patch +Once we have a stable implementation if you could include a FAQ entry referencing the so that if people wish to use oauth with http client they can find out how to do it +added the if statement that checks if nameValues is null to the isChecked method of class in the pluginI built the plugin and all tests passedRunning my form tags example application with the version of the plugin fixed the NPE when the nameValues is null +Is it okay to add the extened to? I am still working on buddy testing JDBC +Only try to access the manifest attributes if the jar has a deployment of Jars without manifests no exceptions reported + +I dont think we should do this by default IAC users should configure whether this needs to be done +Looks useful thanks! Patch committed in revision Instead of copying Lucenes code to Jackrabbit its probably a better idea to directly call that class from +is still relevant and it would definitely bring benefits and the point I am trying to get my heads around is what is the costbenefit of the two approaches +The changes in do not seem to harm but are not strictly necessaryI will attach a patch though in case you want to apply it +another part would be to print the annotations only if the java field is true +WhewI agree that the test should exercise the protection mechanism to also ensure that the production code does what it should +Looks good to me +Ill see if I can gain consensus on lowering the entry barrier I believe there is a slackness test +But so far it hasnt been committed to either +seconds with retriesBut still this should not be blocking download should be a jobbackground thread +Seems that HTTP protocol talk only in USASCII and other charsets are supported at higher levelill fix my codebut do you think that this patch can be useful! do you mind creating a proper patch against sandboxtrunk? Do that and Ill review m a bit confused in Delft I have not found a clear distinction between the decoding of header and body then what is the best charset! Deft can be multithreaded so it would be safer to instantiate a new every request +Thanks a lot for the contribution Andrew. +Ok Done anyone with a Confluence acct can create pageseditdeleteYou still need more admins I would say note that as a current admin Chris could have done all this for you +too many bugs for you +I searched for past discussions on the question of including the old examples and I found but nothing else +Closing as this is in the tree now. +Do you want i build a test only for the javasource part? Else everything compile fine and the unit test you provide pass +hm crapbut it doesnt matter that much since i wanted to provide a patch that covers even more date patternssorry for wasting your time though +the protected method Its funny because literally this morning I was looking at for other reasons and I saw that method and briefly wondered what it was for it just return what was passed to it by default +configure homejfclereJAVA simple stub this still needed? can we close this? +There are optional portions of the url for Postgres which I left off +Thanks for reporting +I definitely agree I think it would be quite helpful +I tried trunk just now on Windows + the bulk of the changes are to not use the default filesystem object in many code paths and instead use the one derived from the pathconf +Probably the test code needs to be updated +The tmp file in seems to look like as though the test exited in between this behavior seems to me like the one seen in +This looks like a good change to me +so I assume that the other end of the pipesocket now is some application unrelated to whats in subversion yesWould it be possible to replace the explicit use of an fd with a virtual interface that can be implemented out near the wrappers? Maybe the action of qmfcommon notifying is simply a method call through a defined interface +JasonI did not realize that it is already fixed in trunk will update the JIRA +Jay just took a look at your sorting classUnfortunately I cannot accept that +I see +Thanks +added a new test changed to junit same as but reverted simplified codes in etc +This one actually works! The first one I uploaded was building the Search DB with absolute paths +to all marked branches. +Thanks! +Hi SamindaAs Heshan says we cant go only with that change +Attaching the test case a simple configuration flag in would sufficeWhat exception do you get? I would have expected that each new container would and all would be fine +Therefore this particular issue is no longer relevant +patch for trunk +I refactored the continuum rpc client in order to use the xmlrpc am going to apply Mathias patch and then modify it in order to use the old continuum client for continuum we still need the http client dependency for continuum +After making sure that the directories exist both tests the Groovy jar from the link above and Ant from Gradles lib directory Ive done the followinggroovyc cp +Requiring jruby will provide all methods those that need java and those found in inUtil as well for backward compatibility +thats correct +You will find attached the output of mvn X clean and a test case to reproduce the the sh script and next run mvn clean to see the issue tmptesttoto is deleted with default configuration! +I propose creating a new loadstoreFunc and require strings to be quoted in that and apply the changes that hc busy proposed in this patch +Patch doesnt apply to trunk thats why fails +This patch enables CRUD functionality for JCR nodes exposing them as cmisfolder and cmisfiles +Fix confirmed. +No need for IT +Resolved for a while +Sounds great nkeywal +Maybe you have not packaged something +there is a huge problem w the current implementationThe whole process is built like an after thought +Its expecting something its not getting +I dont know the answer to Dans question +I think some synchronization will probably be required +Different searchers may come back with different sort types and then there isnt a good way to merge the resultsThis could probably be fixed in the straight case by getting the sort type from the first searcher that returns results and using that for all the other searchers +Looks like we dont need a connectionstate either +properties files comitted to CVSregars Malcolm Edgar +You can take care of this at your convenience +In this case please submit a new patch for the current SVN with the changes I mentioned + branch already fixed and release accomplished +Sorry for delay +I will resubmit the patch for trunk and by fixing all the comments given above +just applied the for as wellProbably should also be fixed for but I dont have that checked out right now +The length of these ranges can be +Thanks ArunYour patch is in rev +If you have other patches please open a new bug and attach the patches there +Bigger value of connectionsession timeout fixes socked closed on benchmark test side. +Niklas could you close this issue after your review +screens in were updated according to the new view of VPE +We have no useful information for diagnosing this bug nor can we reproduce it. +That may help us understand what you want +I combed through the file and removed all duplicates +Actually there is already a floor on boxes level which cannot be in the bestFit functionFiner radiusmiles param is for filtering afterwards +By using another name for the property than does the effects of this issue is minimized +Meeting the needs of both and is not a simple thing and this is how Ive accomplished this in the pastand it works swimmingly +Bug fixed in Xalan +Hi Justinit seems that work on this is done +Patch for trunk +Seems this is still not scheduled for a release although only two other issues have more votes +This patch will make the archetype use the current version of camel +The problem is still there +The exact cause of this issue is unknown and we will wait until a user reports the error again with the additionally logged information +Although the source datanode now waits for a response from the target datanode before closing the stream it should not affect for the cluster +We dont have much time to fix this now kerberos will be our next priority after releasing +htaccess errors think certain pages need absolute +In particular the expressionnull not matches any valueis always evaluated false in and true in +However it might be a little ambiguous for usersThanks +Postponed to has been releasedMoving all related issues to the next version. +Can we move the validation into the methods and then just invoke a method from a runnable thereI tried to optimize it as much as possible because I still think that there is a reason to keep it because it encapsulates all announce apply and validation logic pretty good +This also removed loop before event send +This way the client doesnt get more permissions than they are approved for and they also only get the permissions they ask for. +Should These commands throw an if unable to save or write to the file rather than succeeding but logging a warning +Thanks for the contribution +When I run mvn findbugsgui with the source encoding set to I launch jvisualvm and see the property matching +And I can certainly test it for youRegards Richard According to you the messages were probably being sent over an already dead connection within a tx context +looks good +Ill try them +Hi Craig Yes we have tried the second workaround so we add http and it works! +The should be used to reconstitute Subject instances across VM boundaries +It does not need any additional MR +Or just release and be done with it +Hmm seems like the regex got a little too specific this time aroundI see tags left untouched now +Triage for +Similarly for the client and service wrappers +introduced native exec features and this bug uncovered a problem with the new implementation +Syntactically its a correct query but uses features that are not required in +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +What I have initially discovered is the writer is actually Null +ImplementedFreddy I believe that we should add this rule to the Sonar way profile +Sorry we misunderstood each other! In we need to fix this if we release +I built the project with the ccode changeThen i reproduced the same error and the required change has been reflected +verified and working fine hence closing this issue +Thanks HiteshHitesh can you create a separate ticket for the remaining tests? Ravi and I can put up the list of tests that are pending and one of us can take it from there +First issue wont be fixed since Jackrabbit is not being used now +I removed it because the full path causes a http and not a https request +Test again when is out next week if it we can +This is very urgent as lots of works are pending after this +OkEverything merged as of edebccceddb +Fixed with issue +If you finish it before we release beta please correct the version +This issue has been fixed and released as part of release +yup those work with and without only thing that the convention plugin does at request time happens in the rest is just configuration loading on start up you might want to put a breakpoint in handle and see what is going on +AS Perspective are are updated +I know that some of our ITs are in the top package which is my bad +You can check our examples there is a code which show how to use hashed and plain form +can you explain better what kind of configuration youre Kris perhaps a better message in the exception or a message in a logger help to diagnostice the causes +The only risky part is implementing a proper control file generatior but that shouldnt be much of an wont have time to rework the for another week or two but after that I might have time to do the entire Dpkg refactoring +all the various typed implementations we have now what do we see as the role of Field? Is it just serving as a parent class to the implementations or do we expect users will be using it too +bq +thanks +Oh sorry I havent update the local copy completelyEverythings fine. +It is great you added in the XID DATABASE DRDAID etc +Yes and IMO those were mistakes that should be fixed in Struts +Patch is good with some minor whitespace issues which I am happy to fixWill apply +aborted IO error parsetreeLooking at the source it seems that it looks for the big question Is rubyruby even supposed to work with at all +The metadata stuff can be handled under for filing that patch as svn resolving +Please supply us with a runnable project that shows your problem +If we modify delete objects to be sorted by priority and in addition by the order they are marked for deletion it should work for both use cases +Using long values for dates would be a significant improvement over the current String implementation +Looks to me like you found a bug in the GNU XML library wouldnt you say? I dont think theres anything we can do about this +And when it does so the domain objects which it references are committed also even if they were not supposed to +Removing JON and CP from fix versions since it affects the embedded console not shipped with +Here is patch for our pom +The fix looks fineBest regardsRichard +Done the basic integration +Thanks for the tip and comment Harsh +Once I did the artifacts were downloaded appropriately. +Problem does not concern in r test pending +Ive found a similar issues with the code that I had to use some workarounds for +HiYou could move those files out of srcmainwebapp srcmainresources which would put them relative to classes and therefore on the classpath +the test is for the case where unions write to a common tabledirectory +Thanks +me commit it +Closing issue Oozie is released +I was assuming it would only be going into +Committed revision +The designator is just the official build number? Its not a version of the release +I plan to commit tomorrow or soon after unless I hear otherwise +A new patch adding a graceful option to enable the new gracefully rolling restart facility +rand is horrid +This does not help much though as no JVMTI is implied by this test scenario +There is actually already such functionality get I just added a nd version where you can pass a given with +Thanks thanks a lot +By the way the error is thrown from +Hibernate Validator exposes the bean itself as the invalid value then despite the registered path pointing to a specific property +More will have to wait till after the release. +Icertainly dont want to create unneccessary noise or distractions but thoughtit would be best to document any problems I encountered one by one while puttingthose testcases together +Exact location is serverdefaultdeploy +Terrier has it though so I guess there should be at least one method that depends on it +So this issue has to be fixed separately from. +And you cant embed HTML in a Build Name only in the Build Description +Thanks Myrna for helping me sort out those test failures somewhat +This is no longer a simple config file fix +BTW what is the traditional ofbiz process for getting translations? Even as much as some of you are multilingual there are some rather exotic languages in there +bq +Defer to +Good job! is low resolving as Done. +This was an AXIOM issue rather than a Rampart issue +Maven should be taking care of that +IE does not display the server memory usage graph +m +But I think that unit tests should test minimally what it is supposed to test +As discussed on the dev list pushed for too +Imagine for example you have a very large data structure and it fits into memory only once using up almost all of the available memory +Committed to issue as it has been released as part of release +Sending mail for these because these should be reviewed to see if they are still valid issues +Thanks Adrian I had wondered the same thing when I initially came across it +Make it a blocker too +Ive committed this to trunk and. +Yup +And be documented of course +Would be great to get it fixed for release +We really should use the of the application i guess because thats our hook into this patch work for youWe already have interface for OSGI to load the right classesSo that one should also be used for loading resourcesThe default implementation is what you did +Can you produce a unit test to reproduce +The patch is fine was error in the failing tests +Nested placeholder expressions in placeholder keys will be supported in Spring using a more general expression I misunderstood the issue +Thats not the case here +Its the first time I did this so I hope you can use the file +Our users dont just have names but they also have an ID for which domain they are from +We seem to be boxed into the corner which Bryan identified +Jacques We will have some comments and suggestions on this for you shortly +They need some prettyfying if the stay there so this is an interim solution at best +Attaching proposed patch +Attached patch contains all converted examples and changes in some beans in order to get the examples work +Yes if you create me a issue focusing on just that area +pushed +BTW once we nail down the formatting and everything I will toss reposync up on a github repo or something +Add meta data support to webdb +Thank you for the clarificationWould it make sense to deliver this with Chainsaw? dont know anything about Chainsaw but hopefully someone else can contribute a useful opinion +Your comments were really helpful +The way the work around the issue is by keeping a separate set of loggers per classloader +Hi Vasily Contribution committed in revision under enhancedclasslibtrunkmodulesrmi where it will reside while we merge it into the modulesrmi folder +I wasnt the patch I was saying that the use of as a magic value is a bit hackyI think inserting or null and checking for it on the other end instead of checking for is cleaner +Code looks good +problem is with. +I agree with Jon +Not really sure why this +svn diff would not include it in the patch so Im putting it up like thisI dont know whose image this is and if it is copyrighted +and revision to set the Fix Versions field to JDO beta. +The bug does not exist in its a result of work done for fun to be on the bleeding edge. +I am sorry I am busy these days and I will give my feed back in a few days +attached patch +Im trying to write tests and on all the stuff that was done over on previous issue +No need for IT. +The kvs list is now final and cleared when reading fields +This IT deals with plugin alias remapping so I strongly suspect its not related to mercury will try to confirm +Of course Im assuming the TP is versioned independently from JBT. +Block locations change is +Updated patch to fix test +Thanks Erwan +Added Javadoc description to plugin parameters +anything that it makes means nothing for me +On the one hand side I can understand that there is an interestes of not bordering users with stack traces +Snjezana could you please provide some hint howto verify this? Thank you +i disabled cvs and forums +I also have some changes that Ill submit soon that consolidate some of the Thrift handling to make it easier to plug in the SSL and SASL code as there are multiple things using Thrift RPC in Flume +Lastly almost everything applied except for a couple changes +There are two new properties you can set JMX invoker authenticationadmin resolved issues for already released versions +The first block is written to worker who pipelines it to worker who pipelines it to worker +o I thought about renaming the current to but then on second thought it really has no dependency on the API it could potentially be used by any application to communicate with a Solr server +Could you please link the respective jira that applied the change to security branch and or trunk +Lukaszcould you test please and let me know if its working for you +Lists start with or was not touched or added as you dont need build instructions anymore once you have the full documentation was minimalized and only points to the HTML docuWe still need a system requirements page +needs to go away +It binds the port before clearing the system directory +committed thanks +Attaching unit test for patch thanks +My concern here is tokens +Some details followo trunk Created a single target for making the directory +Patch to add xerces dependency to when the ibm jdk is detected +Fix with +I seeI think opinion of Knut is properI will rollbackThen this task is regarded as blocking We need to solve this issue before those issues +Is this with or a nightly build? ignore that last one +Did I miss any thing? If not will delete the method +Do we still need this issue open? It seems we have two other issues for the problems now +Was Improved namespace handling. +think that? like +Hi Sandro +Thanks Terence. +The is for auth between client and AM and clearly MR AM only uses it now +This patch should really have associated MR and YARN jiras patches as we want to make sure all their tests work Jenkins for Hadoop common doesnt build and test them which we need to do so as to ensure that their build and test runs dont fail +Also uncomment set in and will be is no problem hereThe cookie is set in the first response the one that makes the redirectAnd it stays in the browser until expired or deletedCheck it with real browser and see +Not sure what happened since Apache Hadoop but we have to add some other path libjvm to LDFLAGSTested on Fedora and Mageia +on committing the patchMaybe it is good to commit the tests from too just to be sure that we have a lot of tested scenarios +resetting P open bugs to P pending further status from ASSIGNED to NEW for consistency +Do we know about the impact on query performance? All this array creation and compacting +will be included with AS the fix in ER. +Yes that would work really nice +Granted +I would vote for it once I get clarity on the and RegardsHafiz +I will close this once is committed +This is good enough for me right captures the result of aspectjcompile and creates an xdoc with improvements can be do the same for aspectj format the output neatly link to the jxr for other issues if you provide docs ill gladly apply the for the docs +Pavel Mack is writing a new performance test for this issue +Note that I changed the build slightly to leave generate xsd file in classes +We could revise this so the timeout is recalculated to commit time is no issue with absolute timeouts nevertheless +Verified that the test passes with the patch +Fix applied to master and +Of course Im fine w doing the excludes no need to guess on future functionalitySean does this work for you? I think the patch above keeps everything happy in Maven land and actually makes it a lot easier to run the demo ala my goal inIf it does I can commit and update the taste docs +Not every app has high QPS requirements and might never expect to either for that matter +We have a URL that points to a nested jar and the new implementation vfszip doesnt seem to deal with that the right way +Please confirm they have been applied as expected +has this issue. +I dont have a Linux box to test on +I changed the file extension +Can this be jbossxb instead to incorporate +Well discuss this and how it related to even propagation will be addressed in the frameworks leave the bundle in the ACTIVE state even though there was an exception in the bundles activator +Hi Erik Should we remove the on or wait untill +Please attach the whole stack trace so that we have an idea what to look for +DC is piping +I will note that in practice this was not really a bug because the deletionInfo used were those of the columnIterator and at that point those deletionInfo should only contain whole row tombstone not range tombstone +Hi Jitendra Fixed the comments and provided the patch +Rejecting this for testing infrastructure for the EAP download is not going to be ready in time for us to implement this for we are punting this to. +Removing on these issues until we have somebody committed to fixing them. +Seems I figured it out +i suggest that we close the issue. +Let me handle that +fix will be in to branch +If thats not correct please remove both xml and html versions of the file there +So it should be an admin cluster event with the target destinations +Many corrections of and a new example +Thanks Zheng +BULK EDIT These issues are open to be picked up +Each ssl unit test will throw the skip exception if the openssl libraries are not present. +Hi HansSeems like my interpretation of entire property was incorrect +Fix is included in tag and is included in EAPThank you Carlo. +Please keep the svn repo for our website +Committing this +See Welcome is committed to trunk and branch action are fixed for Create New ESB and Teiid projects in trunk and branch +Likewise should rename sourcecluster to simply cluster +Bug has been marked as a duplicate of this bug +both for EAR and WAR it is the old I had an old launch +Committed thanks for the feedback and the reviews +Hi folksCould someone help me to review the patchchanges and would appreciate the efforts +OTOH it might be that we just want to add an endpoint type within a pipeline. +This bug is about a request to provide binaries +I dont see anything obviously wrong you excised this code from the same code thats used when merging the postings during flush +Username pilhuhn +Attached a H module adapted from HSQL by Andrea +HIVEHOMEeval echo HIVEUSER +Closing resolvedwont fix issues. +Onclick events will be processed before the tab contents are displayed. +Since is done this is no longer an issue +It depends on patch for +Now it throws where before it just soldiered on +I removed and from my CLASSPATH and tested XML queries they executed fine +Fixed in both and The options is named private and private +Created module structure +Thanks Sharan +Rob any comments +This makes it possible to report the current methods fileline but it also means that arity checking always happens regardless of whether it is neededIn mismatched arity generally doesnt even make it into the target method because theres no path with that arity to even make the call +Ah of course I had forgotten that the async thread may never be invoked at allI will put in the additional sleep then test against trunk if that still shows up clean +Right! Sorry for overlooking your comment Owen +This new one holds the good changes to apply +I think rebalance is not really possible because once a group is assigned and the consumer is valid moving it will break the expectation of getting the full batchgroup of messagesBut what we can fix is the allocation of new groups to the full complement of consumers + +Id rather flush out any bugs here sooner rather than later +I saw that and later forgot about it that method didnt need to be there anymore +Ill backport to x and then look at making from so we can support other directories in tests +It would be good to know exactly which version of jruby fails here +I am running some console testcases then upload the patchThanks +Offline I started engaging some of PigHive community folks +If we need to set up a chat one day to hash it out Im willing but this might require discussions with Max Nick etc to understand the roadmap +The regression patchThe case is very rare but possible as we can at r +Could someone close it +The new patch thus has only the following changes Modified signature of update in to include the Modified to pass in the RPC port of the jobtracker +The files in this amended submission areM toolsantpropertiesM javabuildorgapachederbyBuildM javatestingA javatestingorgapachederbyTestingfunctionTeststestscompatibilityA javatestingorgapachederbyTestingfunctionTeststestscompatibilityA javatestingorgapachederbyTestingfunctionTeststestscompatibilityA javatestingorgapachederbyTestingfunctionTeststestscompatibilityA javatestingorgapachederbyTestingfunctionTeststestscompatibilityA javatestingorgapachederbyTestingfunctionTeststestscompatibilityA javatestingorgapachederbyTestingfunctionTeststestsjdbcapiA javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTestsharnessA javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestssuitesA javatestingorgapachederbyTestingfunctionTestsutilM javatestingM javaclientorgapachederbyclientnetM M patch was scraped from a subversion client at revision +Great to see this planned for +I wonder with if and will get set since they are system wide properties +Committed revision in trunk begins to describe more verbosely an accurate class description for the base Gora tests +I have incorporated your suggestions in the next version of this patch which I am attaching +Like I said the main risk is client side where you have less control over parser versions +bq +The urlDownload parameter is now optional. +Vasile we still need some additional details on this issue or well have to close it for lack of information +patch to return an iterable from RSRR instead of a list + +Thanks again! +but it doesnt show all zeros right? otherwise whats the best way to iterate thru elements of a sparse vectori just want not to iterate thru millions of zeros but i am ok to iterate thru few of them +Id need to test that further I havent tried with that kind of configuration +Obviously know more about Apache way +I have a patch for this but some reviewers have voiced concerns that it adds unnecessary complexity +About the priority of the issue I didnt want to set it to Major I just forget to set it correctly +But if any of them are active the replication slows downDisabling the caches isnt an option for me since the query times gets way too long. +Thanks Lars this looks better +Revoke of Global permission is not taking effect without restartBetter to make a new JIRA than reopen Id say +Thank you sir! +However table compression certainly takes us through code which has given rise to a lot of bugs +Id like to see an authenticate method implemented +Assigned to service pack patch attached which patches this issue +I committed this +You definitely need to say what version of each one of those is running +how about putting it to a new interface say ChecksumableIf any might implement this interface then I dont see how this is functionally different than adding this to +I spoke to the submitter who reviewed the itest case and feels it covered the scenario +thanks for your also updated the JANINO web site with the new repository address +Test for it seem to be ok with new lib +Can the patch +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Closing this issue. +Committed +We should build this via mvn clean install Next we can buildgenerate the eclipse setting for the project using mvn clean install or mvn eclipseeclipse commandThe artifacts generated can be imported in eclipse +Likitha can help you if you have any questionsthanks LikithaPrachiFrom is packaged as +Hi ColinIf talks to multiple DCs it is nevertheless mediated by the main DC otherwise it would have no idea what it should trust +So I would vote for +Sure Ralf +So replacing random with sequential sounds like the right direction to me +About the patchSince all fileschanges were in the same package the patch was created in the srcnetsfhibernateexpression directory + +Ok John please reopen if needed +There is some discussion on Feb here is the summary what is the memory cost of having to init number ofEach one has at constrcution an instance of all that require to manage their own state +This is totally awesome +Looks like again the Apache license headers might be creating issues +of +this check here would violate DRY +There are currently three diagrams on the front page +This patch is against this was my fault +I find this incredibly incredibly difficult to believe +I reran the heavy read test and it still did the same as beforeCan you please try running the test yourself +I just committed this +Would someone please provide a fix for this? Its preventing me from doing what I need to do in my build and I dont see any workarounds to the bug +I am working on another patch thanks +Should be fixed in next snapshot release +have the same line spacing as the code in defines up in Removed inconsistent blank lines in s the version number for +Note that you will need to change the and files to inject these new handlers into the security component +bq +think youre mixing up trunk and +Some minor comments +did the rest +See for more details +Vikas +In keeping with this philosophy it would be beneficial to those of us who are watching these threads to close the issues as By Design in order to not cause clutterthx +Minor patch +bq +Prof of Servlet should already do this +You may still need to manually terminate the server process in the console view to stop it +Please consider reopening this issue +changing the fix version to Unknown as we dont know what the issue is and whether there is an issue +But only a few ESTABLISHED connection from some other datanode clientsRecheck on the client side the connection may still remain CLOSEWAIT +BTW has a similar problem +For each test run k messages were written to mytopic and mytopic each with no consumers running +Todd agree that the restructuring gives an opportunity for better solutions by looking at operations across components. +In CR we should only fix the NPE the rework should happen in +Id suggest going ahead with the patch and also releasing a new version of the stylus skin in advance to fix the problem +Might be a threading problem +Hey guys just wanted to say thanks for implementing this feature requestI cant comment on this particular case since Im not familiar enough with the code but in general it makes sense for exceptions to be used outside of the packages they live in +Should be easy to execute and see the problem +Which jdk are you using which version of the pdf pluginThe error that you get was a known issue in xalan versions that shipped with older jdk releases +Otherwise its just confusing respectively wrong. +I skipped playing a video game to get this patch merged up to trunk Im very happy to have patches that are targeted and include good tests those are much fastereasier to commit without taking much time or thinking too hard +This patch adds Thrift Thrift and Rest metrics implementationsIt also added tests classes for metrics +Sure Kathey go ahead and make that change +Marking as closed because I ran into this issue and it was fixed when I used the latest jar +However IIUC the RM wouldnt schedule another attempt until AMEXPIRY is met +Less than week to pushing off more items to +Cool I see now where the is doing thatIve uploaded which implements your suggestionsIs this the best way for me to submit this changeThanksThank you looks like the two implementations will produce identical behaviorsFor the choice of class name I like your name betterAlso I think your choice of positively identifying the idempotent methods rather than checking for like I did makes the code much cleaner to read +Maximum number of allowed references exceeded +Is test failure because of this patch +Have you configured Hibernate for +A new maven project that builds a service wrapper +Before jackson was not packaged as part of +Last night before rev my test case for this was failing +But hey if you want to go back to having tftp client code go right ahead +Configuration error. +But feel free to try and remove the headers and bullets decorations or modify it if necessary +Can be closed +This error was actually the cause of your error in +The job tracker does know which machines that each task has failed on and prefers to run other tasks on that node +btw GROOVYHOMEconf contains the additional classpath entries so you should maybe take a look at this +Defer all unresolved issues from to +Are we talking about disabling Thrift access entirely to columnfamilies which use named metadata with composites or are we just talking about not supporting Thrift addressing columns inside composites by their CQL namesThe second seems eminently reasonable +Then the user name for the hadoop job should be the same as the one with which the oozie job was submitted +Also passing through the orignal params to each request +This is the patch with the overriding of the two methods in +Take a look I think this is a better solution than what I posted I noticed that the command option was not being handled in the case statement in the same way as S +patch committed to trunk. +button Verify Fileset is an option If you really want verify adding a fileset actually gets deployed too Fileset functions similar to adding a folder mapping except you can exclude resources +Hi HowardIs this something you could find out for usIf it wont take you too long we might be able to push it into the release of the Messaging GuideCheers JaredThere are three clustered examples All others are +Is this still a problem? I am thinking of taking this one on +samplesnesting +I think it would be good to have in the when it talks about LDLIBRARYPATH have it point to +Comments from triage on April Henry Z +Im working on figuring out which one it is but this is not a CXF issue so Im resolvingThanks for your help gentlemen +Bug has been marked as a duplicate of this bug +Uploaded the updated patch. +I also bumbed it to scala which has some minor bug fixes over +as Plugin provider +Attached a prototype patch that integrates the data store concept in Jackrabbit +Ah yes +Thanks Mark +In our case we want to get hold of the task participants and take action based in them when it startsI understand the technical reason for this but it seems to be a restriction imposed by the current designWe can work around it by storing state in process variables but we are then throwing away the benifit of using the jBPM API to query this information in a way +Attached reports +Hugues does this actually mean that you are building Castor completely usingMaven? If thats the case its about time to switch + ship version of mina because the SSHD projects requires allow both and versions to be deployed concurrently and the error shown above it exactly what we should have in this case +My mistake +PM has decided to have in EWS and EAP to track the builds +if fajax is on a page it should be created automatically a component that register and render javascript with target head +The usecase is pretty much implemented and committed +Now clobber does +New patch for updated patches for and trunk +Ive apparently reproduced with the decoupled scripts +And I should use version driver instead or +I have verified this fix against the latest snapshot. +sqllogin etcThanks Ashish If it is used for fallback case I think it should be handled by not in the itself +Per the JIRA triage meeting on Oct Yash T +Distinct bug fix support +Verified by Ilya. +Verified by Svetlana +We could inherit from +Verified by Ruth. +Seems ok to me waiting for a to commit +looks good to meOne thing though we should also write a translation table from KW token types to more user understandable token names +You should now use the importCertificate mojo to have all parameters of the keytool command availablein the mojo +Applied patch with commit revision dceddeebaafcbccecbcThanks for your contribution Chris! +This patch is simply adding one alias line to +Also made another change to update metadata request handling +The previous patch had git prefixes +Since is a but not a it is combinable +Fixed title +Nightlytrunk has no composite metadata +I have the following options when I right click on my context xml file and select open with +Set a Websphere web container custom property true +To be dealt with afterAttaching new that compiles modulo this error +truncating is probably fine but a couple of points to ponder The locations per split to keep should probably be a config limit? Should we pick first n locations or pick randomly? Id lean towards randomly picked splits in light of features such as We should do truncation on both the and to be wary of DOS if a malicious client submits too many locations per splitThoughts +All tests pass +Marking closed. +Because Im not good at Chinese can I ask someone to checkIve not update the transifex +Warning idea ahead Emit rate and size readings on an interval out of hbase features such as memcache +Please provide a reproducable test +Bulk close for Solr +Normally it should never be a requirement that partition need to know table properties +r. +When running mvn datanucleus does not appear to receive the schema information thus tables are not found +Thanks Nirmal for the patch and thanks Kim and Knut Anders for the reviewsand suggestions +Patch that the socket used to send the shutdown request to perform the ping to see if the server is shutdownAlso fixes a bug where the message if the server did not shutdown in time could never be seenDrops around a second off the shutdown time +on punting hereWe do not want to introduce more fragility into the just so it can be testedMaybe this is a case for a more powerful mocking framework so that the RPC engines can be mocked with failure insertion. +Hmm OK one thing I could do then is to publish a and version it as +The orange graph display the new content model where group membership is recorded in a like node structure +I think Ratha has uncovered a serious bug here +So at least there seems to be some kind of pattern +aar WSDL file used for generating serviceskeleton SOAP response containing the error the issue from release nightly to as it affects both versions and I want to be using version because its a production system +If you have a cron job to snapshot and it provides a snapshot name it will fail after this change because there will be no timestamp appended and the snapshot will complain the snapshot already exists +A prototype has already been started in the project +With the current build order in the reactor a build with an empty local repo seems to fail upon reaching the Enhancer Sling package. +compression appears to be ANSI +when using jruby as a wrapper to execute a program and keep track of the pid +Andy you said that there is a spec for this work +Attached a simple patch now replicate will throw exceptions out +Caused by an XLC regression see +But you still need to choose from which account you want for example import an application +Since snappuller is a bash script the revised patch uses the noclobber option to prevent that +Note Im using Connect to server function on Ubuntu to access JCR via all steps to reproduce including where I can get the bundle +AS automatically terminates the process if a management port is not found +The problem lies in the form servlet which prepares the form but does not include the path to the authentication handler triggering the form request +I can reproduce this issue +I think we should just remove that part +Random access file is used to readwrite to the file +Also move the note text out of the code area +It will be more clear once all compilation errors are addressedI can see still in the hierarchy which extends +Ive reviewed your patch and it seems very good to me +does not test for the more complex SQLjoinFK test from David but verifies that simple null relations work + Adds exceptionCaught method that forwards exceptions to nextFilter +I also added the ability to specify a wild card for the compile source artifacts which I will submit as part of the PR +It is suitable for a limited retry case +I was just asking in case there is something relying upon this +Resubmitting patch files together +I will push one with the other tests I will write +I think that this is some thing the core can do rather than asking users to do +Too late in the game to do this change +changed and updated +Fix is applied at +bq +Patch checked into trunk +Use this instead of attachment instead of +bq +The detailed concept how to obtain a new master password is described in the user guide section upgrading +Done. +I am working on throwing when it fails an authorization check +comment I see +fixed +Heres the error undefined method availablecommandsfor for It fails both on Safari and after +Verified Patch for Trunk +The streaming tasks dont timeout since is set to infinity +This issue was tested on the latest camel release as well as on the release we are currently using. +Please review the HTML file with the date January + well but during the migration the shared filters seems to be duplicated whereas this should not be the case +Also Im not sure what other use cases there are for system scope so Im not certain if there should be any other checks here +is fixed by +youre quite right +Note The entry is used to define an email alias for use in Postfix mail server +We can add that feature there but I guess it would be even better to provide that in the core module +Check script is not loaded due to DEFAULT loading output in demo still doesnt helped +thanks for reviewing! Ill comment on your suggestions on the MLRegardsBenedikt +I added the standard blurb about the property in the developer docs +I tested this patch by deploying artifacts locally and to the Apache staging repository +Its actually not the annotation part but he whole fails +yes objection the content type doesnt always contain the charsetdepends on the servlet container it seemsso just check em both +Thanks Emil +Thanks! +Thanks Harsh +So go back to the previous method of putting the assembly descriptors into the common parent shared build resource bundle. +closing issues resolved for final +I just committed +Ive been misled by the Jira http error code +The idea is to get rid of login accounts meaning that the email forwarding database should be something separate from the + some source of +thinking about mirroring META in ZK adding new locationsredirects in and other optimizations and availability improvements +I will check +Ive attached our source to ll do a release sometime soon +The patch fixed the LITERAL flag cant be identified problem +Fixed Typo and committed as revision rCheersAshwin +The WS layer rewrites the schema set to a single composite schema will create a dedicated jaxb test for this issue +Yes I have been aware of this +I suspect that simply caching a fixed number of connections will not provide a significant performance enhancement +Add so that Test class can access internal members of ZK handle +Attached is the changeset that made these additions and modifications +Modified to read the automatic configuration property and expose access to the property within its interface +Now reads may be migrated back to it manually which I believe is correct +I guess I was used to discriminate by the value in order to check only one single and well identified parameter but with Spring MVC your approach is not cumbersome at all +Im considering to log a new Jira for the deadlock linking it to the other issueFYI I can reproduce the assert by running in a loop +In both files are comments to indicate the necessary changes +all tasks done + +Talked to Vinod +Should be easy enough to check and I have worked with the Shape code recently as part of hunting down problems +This is related to the need for both the libraries and to setup the GRIFFONHOME the issue of the java compiler and groovy compiler sharing an output directory +Instead of creating the context on the fly I propose we let the user register the context as an service with a property name +moving to doc version for better separation +Does anybody have a suggestionCouldnt you just extend the and make use of something like COMPONENTNAME or PARAMPREFIX in the param calls instead of the static string in? That way the autosuggestcomponent would have COMPONENTNAMEautoSuggest and the spelling would have it set to spelling then let all the common code just live in the base classJust a are we on this do people feel its ready to commitWe probably want to add some unit tests too and some documentation on the wiki at some pointAFAIK were limited to one spellcheck component per request handler that should be OK though since presumably this is meant to be used on its own right? What is the recommendeddefault configuration? We should probably add it as a autocomplete handler in the example serverDoes this currently work with phrases +Thanks Robert for fixing this! +Committed revision +Closing wont fix for now +say if you want to have Eclipse just add one module should be easy to get it to workIn this task I am working on migrating this work into branchRegards +issue still occurs in IE and not if Firefox should be opened as a new issue if it was fixed in but there is an additional fix to be made +Closing RN contain text for +Remote wildfly has many issues +Further the OBR spec inside the Alliance is slowly reaching a close and it doesnt extend the filter syntax either +looks fine thank you Tim +Paul do you have any experience with setting up jobs on the QE cluster +This was caused due to a lower version of Netscaler that was used with +Is this related to the work on +Can we have some indication of a on this please +Thank you Marko for your quick responseI tried the test by removing finalize method in unfortunately that doesnt helpI dont think it will have file handle leak every instance only has one instance +So there is no need to adapt license files +Will let you know how it goes in a few minutes +The concern is that derby only expects an out of disk space error when growing the file +If you only show products in the categories the result is very counterintuitive +The abbreviation for Hectoliter is hL or hl +Id thrown the idea only to describe that an alternative design was considered than what is implemented in the attached patch +Thanks Victor! Nice fix! +This is a Spring thing +so v or v plus fix can be committed back +We need to fix it in +The description for was not having useful information due to having parent issue tag +The rebased patch is on trunk +For the record I still get this issue on master +If that doesnt help well likely need a testcase +This includes lss default behavior which I think is +You should be all set nowConfirmed working. +Download the example code Use the quick fix setting up a Jboss tests workThen Change the target runtime from ejb project from jboss to a jboss runtime I do not work! If I try to change the target runtime of ejb project back to Jboss apply +studioeclipseplugins +Any chance a quick note could be added to the javadoc for the annotation +reopened +bq +I have made some modifications to the tool so that it now correctly handles simple type restrictionsI have tested this using the provided WSDL and the code now compiles and matches what I would expect based on the WSDL defintition including consistent type naming in prototype and implementation correct use of pointers for nillableoptional or if a type for input and output parameters +The fix was tested in geronimo +META +Thanks for the patch +fixed in as now. +Theres no need to have borrowreturn semantics +Tom Fennelly thinks that an ESB fix in the will solve in is still an issue +Updating the snapshot. +Expected behavior as per JDBC specifications revision +Patch that fixes the typoAlso changed the constant in Balancer to private it is private in +If you can remember you had told that once rolling restart is tested we can take it to version +Thanks for the help + +Distributing binaries under this license is problematic may be OK to distribute scripts dtds and the like depending on the circumstance +Thanks +these jars do not exist they are only to illustrate what is required to create a module for a messaging provider +Please modify the attached application so that it shows the problemI cannot reproduce it so far +wait until it goes red in the UI +I would suggest following up with this third request at a later point in time please +What version of maven are you using? Try using +Quick update since I am currently working on JMX graphing and Nagios support for I saw that Hadoop actually has it in place the exposesnumINTEGERnumINTEGERwhich are two JMX operations querying the numbers +AFAIK this is already fixed in the Cook menu can now use absolute +Sorry I forgot to say detail are at. +Hopefully you will be able to attend this +fixed formatting and committed +Move the old word count to the test jar and rename the to +Sorry my mistake in environment conditions +If it doesnt we might want to fix that +Thanks looks betterYou still have icons missing see if you can fix that sometime I have added your mirror and it will appear soon +Fix released in. +Attaching the trunk patch +The combiner code path doesnt report progress at some configured interval so it can timeout when map output is large +Since the EJB plugin doesnt exist anymore Im going to assume this is outdated +The usecase is to create a kind of content management system or WIKI +Please email me your full name to have it includedin the Castor Projects contributors document +We wanted to actually override it using a newly sent jar without upgrading and restarting the clusterwe saw that as a general enough case to warrant a change to the task trackerI guess that an alternate solution would be to unbundle the external programs from the main jar file +The probably wont work on Java at all Sun made API changes in JDBC that are not backwards compatible in violation of their own policy on compatibility +I will add a comment with all the classes you need to preload soon +Updated bndlib to and deployed issue now that has been released. +If things are not clear I would very much ratherchange the docs somehow agree +In Server runtimes step use Find button +Verified by Richard. + is designed to process dependencies not plugins +Thanks MadhukaraI put this into trunk +I might be wrong but I think we tried that and bytecode enhanced classes had issues maybe it worths a second see where youre coming from now + +This is as per XSLT spec. +So Ill go ahead and mark it as resolved. +Viola should have fixed this after updating the daytrader ear structure +In particular just limiting the number of slots wont do any good +Arnaud determined that this is due to the provided scope we use on the applet to keep it from being included in the suggested adding a dependency on to the config for the dependency plugin +If you find further issues please create a new jira and link it to this one. +I know that is a patchShould have been I know its a pain +Linking the current PR +Somehow something is messed up there so that when I open the archive there the libnative folder is emptyWhen I however unpack the archive the old fashioned way everything is right where it is supposed to beLuckily I wont be bothered with this no more as I just moved to RVM to manage my various RubiesThanks for the support and sorry for the fuss +When we made the change away from commas I was a little concerned about validity +When created they can be marked durable which will cause their configuration to be serialized to the store module +Due to updates in command syntax work needs to be done to bring the back to functional ve added a sub task to this Jira to track this issue +is a different issue +We are attempting to set the execute permissions on the script +Code looks good +Add the above comments to the patch Id say +Solved by configuration see last comment +Issue closed with release. +Initial Patch +We couldnt in good conscience do this so we added this mechanism for users to customize paths that should be added +I must add also that I try to unmarshall the same file each time +Fixed in and patch attached for in is better to use instead of adding special constant to constant removed from use instead. +Maybe we should make a logo competition like on solr et al +Could you attach to this ticket a sample Maven project and do you confirm that you manage to reproduce this issue starting from an empty Sonar DB? started with sonar and we upgraded from sonar +At this point this new patch is the only pending patch on this Jira entry and I will highly appreciate somebody taking the time to review and committer commiting it +adjusting componentversion so it shows up that I know +A patch against fixing the problem +assumed issue did not apply to the onError method will fix that soon +So it seems a bug has been fixed in proj lately + +Manually tested! +For some reason attribute CLASSID with a value like java causes to JVM heap error and crashes eclipse +Did you have a case where this wasnt done + +Maven Integration dont want it to use dependent projects at all +In can we get away with just increasing the timeout to mins +Changes committed +Thanks RamyaPS I am trying to assign this issue to you but unable to find your name in the drop down box. +bq +In other words we cannot fix this completely in the MR AM with a subsequent attempt because it is highly likely the subsequent attempt cannot be launched successfullyAs for trying to protect the MR AM from a after success isnt this already a solved problem? See +A really nice feature would be to allow for boosting for fields for exampleqidmlttruetitleauthortopicThis would find items that are more similar to the title over the author etc. +What we could perhaps do is add a to say the server is shutting we cant rely on any logging framework not being shutdown I think would be appropriate in this case +Streamed using transfer method among with ranges +Great +Sorry for delay so long and thank you very much for all your attention on this long opened issue especially to Sian for all your patches +I altered the test case to run within the test framework the modified sources are attached +We can remove the static loadClass method now +That should be a very small number of casesI am not aware of many changes to test files which are not related to mavenization +Rose importer is no longer supported +Manually tested! +connection is still default. +For a project Im working on Ive implemented a nested set approach with Hibernate +Fixing warnings +Im commiting the current patch if its not the final solution its at least an improvement +Verified with JBDSVersion Build id +This test case works as I expect and tests which depend on Bidi stopped failingThank you very much +type check can be added when Pig types are implemented +I really dont see why youwould want to remove this header anyways +Sorry I make a mistake for issue workflow +Both when the user null and user null the request reads it as null +Thanks Brock +Not sure I agree qualified Parts example SUPPLIERID The VDB when used as a source treats each Model Name as a schema so if I had named my XMI model PARTSSUPPLIER then the fully qualified name would have been SUPPLIERID +Creates manifest jars flattens the dirs a bit etc +FixedChanges Instead of throwing out error return the response containing the info about the instances that were launchedIn masterCommit hashdbbdedecedaabfeacCommit hash this record since is now released. +Type conversion logic in methods prohibits use of generics to eliminate methods. +Pushed to back to has been releasedMoving all related issues to the next version. +Ive committed this +Also updates to bpmn +fixed in that the fix solves the problem described at the mentioned mailing lists. +To this end code was added to do some tricky stuff with join equivalent analysis +Ok +The poms must be interpolated before they are deployed to a repository +Ignore the patch of Mar it is misguided +It works fine on trunk +Thatd be my guess as well +several large content records get flushed into one file in one process which means more efficient IO usage +missing feedback moving of for follow ups and not reproducible so assumed fixednot a free to reopencreate new +To make it clear the response body send to the browser is also empty the filter causes the generated view to disappear +But per Nicholas I thought we were going to change the of to from Private +so a general tune up would be great +to recall that originally I did use but there were some really good reasons why I chose not to use it probably related to thread safety or security +It would be very nice if an uncaught exceptions stack trace is printed as a diagnostic +Also reproducible on Nexus with stock gallery app +on Patch to trunkThanks for the patch ChunhuiThanks for your work Sergey +Upgrade to CXF revision . +Any known workarounds around this! Im getting this too when Im trying to run unittests directly from the IDE where these jars might not yet be packaged up but just be on the global classpath +I noticed this was unused in the regressions suite coverage so presumably out tests do not test this error condition at least not on the client +attach a patch to support rolling journals should use marked lastMark to safely ensure the entries before lastMark are all persisted +We could however itd work the same as commit? Ive never actually used NRT so I dont fully understand it +FIXED was inserting cloned nodes in the wrong location when you had a number of unbound collections in a series +Attached patch for issue changed to in save method of having spaces in the name cause problems somewhere in the appIs the decision documented somewhere either in a design doc or in the user doc +Is it worth adding this wizard to New menu? +Next what about the server? Doesnt the embedded driver get loaded when running the server too? If so shouldnt we warn against customized server policies needing updates as well? Cf +Removes all the documentation folders from the codebase +Marking as closed since Brian reports the problem has been fixed. +Done +Sebastian I did not know that it is possible to assign one mapper to each core in Hadoop I will try that for sure +if score isnt in the sort but is in the field list scores are still not returnedIm looking into this +Daniel it looks like the corrupt page source doesnt include the rich editor js. +Thus need to revisit this problem when importer is fixed +closed +Please note that in order to make the patch work fully still has to be added to the repositoryCheers is a single file version of the patch from MarcThanks Marc look forward to using this with Synapsecheers just noticed that in my summary patch I switched on interest ops queueing cause I was playing around after posting the regular patch which has the queueing disabled +Hi MarkHow about the new version test case +This is same as Pls invalidate this issue +Can you also attach the trace of the commands your client sends when try to attach it? I would be interested to see the APPEND is my status update +Can you test it on your system with a real Kestrel serverAlso we should consider to see what it takes to build a test with can embed a Kestrel server +Sorry couldnt get started on the patch sooner +The rationale here is to bump a bit beyond thedefault produce sizeCoincidentally it seems that the property is usednowhere throughout the code +I saw Martha has upgraded common and natvie to corresponding version +I think this is a request that the security realm wizard should generate a new plan and allow you to select whatever dependencies you want for it + +Testcase works fine with r +Attached is a patch that fixes the visibility issue +Users frequently trip up on this one and the fact that we silently truncate at tokens is unexpected +The attached patch is for internal Yahoo release and trunk patch will be add here shortly +The exception is silently ignored and the code behaves in a different way +What is the risk of the fixFor some reason we seem to be seeing the stressmulti error quite a lot in nightly runs tracked under have the fix for this jira in for a long time now so I feel comfortable backporting it to +in client stress tests we used on the client side so the code for timeout wasnt therein jobs where we use the snapshot version property defaults to secs +Do you have an ETA in mind for +Ill commit the patch shortly +renato please create an issue specific for the documentation that is a subtask for this one +the extra xsd on the end is due to the library we are using +Apparently I got knocked off the cool guy list and am not on the Seam dev list for the jira project +Oops +Ill check the updates in branches move the schemas over and regenerate the schema docs +I will be out of the office on PTO until Friday August nd +I made a couple minor revision +The JTOC also needs its size increasing +Patch with lets Java on Linux +Committed the debug to trunk and. +I will look at the countryid issue +Tracking it in a separate jira +If it makes incompatible changes and doesnt make then it wouldnt go out until probably sometime in the first half of +Rock n roll +Pushing off until requesting change to could help +I dont mind researching this but I cant see anything that in the so if I am right about the process then someone needs to point me to it +FYI this patch incorporates as mentioned above +The title and description for this issue need to be updated +Attachment has been added with description SVG Stroke Font According to ISO +Sorry for not responding earlier. +This issue has been fixed and released as part of release +No problem +I have checked the situation with a current build the problem does not occur anymore +Thanks for the patch +But I got the below exception The bundle could not be resolved +Jesse is correct that patch in does not fix the problem +I agree one standard file for SQL databases would be preferable but one example of why I couldnt stay with one file for both hsql and is the text column was being turned into a blob not text at larger sizes +Im not at all sure about the fact that fetcher doesnt store redirected pages +mrepositoryorgcodehausjacksonUserstucu +The only alternative is an ejb release without iiop included +That may mean that the number needs to be available twice one format to be shown in the admin GUI and both formats available from the mbeans handler for scriptingI think the system info handler does exactly that I remember working on it at some point +Thanks for this report Mirko there are a few of these issues lingering after I moved the source from Hadoop contrib and went through restructuring and renaming things +Try with latest code from trunk as there was a fix in shutting down which could affect your passed the unit test +Any update on this serious bug +Seems to be a duplicate of +this issue is no longer current +I am attaching a patch for this issue +We want to work with Yoko to get a new engine built the right way and shipped with before we start including anything and set an expectation on the existing Ill tell ya what +I dont see consensus being reached on how the Strategy API should look so Im going to leave the patch +The javaee specs have been rebuilt to use the new which does not contain dependency management +Im suprised we didnt run into this earlier +Bug has been marked as a duplicate of this bug +Added component to the issue because a test with a runtime definition and a preparer must be provided +committed with a couple changes default wait changed to flat s schema propagation does not use gossip and is not expected to be proportional to number of nodes in cluster added validation to keyspace update and cf createupdatedrop +h H If applied the existing entries should perhaps be changed to use H as that is how they have been interpreted until now +It also defaults to the tooltip tooltip styleDont be discouraged youre on the right track +has been marked as a duplicate of this bug +The test failures seemed random and unrelated +I will close this bug report because weve never received a test case +Patch looks overall good to me +Maybe the browser cached the older version +Did anyone find a solution to this problemI have this problem too. +Olivier we removed all +We can revisit this once the release is out the door. +Im going to add them as +Initial patch just for review +So I recommend to create subtasks and incrementally commit small patches +Attached the fixed for the pacthes we are going to put the fix in AOP. +Replacement for the originally submitted patch + +Attachment has been added with description Small standalone test that demonstraights problem with Mac OS JVM +The RM decreasesincreases the size and then the AM informs the NM about the change +Can you please confirm if it is still relevant and if it is provide a quickstart + bundle is now part of theWaiting for a fix from Aries I will wrap asm in a bundle in order to limit the size of this bundle +The current of this plugin has had the SCM component upgraded to version +Same problem occured for which includes and which is fixed +Thanks Arkadi! +Ive the trademarks list againWill check back after week on this issue +As already created the patch submit the patch not a big change thoughThe patch moves the register to and modified a few test cases accordingly +cancel the patch until comments addressed +This patch stops generation of new work and sending of remaining work +I agree Charles +Other implementations of besides are certainly possibleSo it seems to me either the user of this functionality has to manage the versioning external to the Lucene index or they cant use the functionalityWould strong warnings in the javadocs be enough to allow people to take appropriate precautions +Alexey please confirm its thanks! Everything is ok. +I want generate POJO DAO for code generation I need Hibernate Console Configuration so we will cover this user case +Cannot reproduce problem with JON bits +The timeout is set to seconds +I think explicitly resolving is a good idea +Resolving the issue as there were no more comments +I guess so +I looked at your first shot and it looks good aside the points you mentionYes we the ideal would be to lock the source folder but as we dont either lock for COPY it could be committed as suchI would be good to have the implementation for at least maildir and jpaI dont think the legacy MOVE should be supported but if you do it why notThx Eric +Note IBM JDK and both exhibit the same compile error +I think it failed due to other failures +I created a Solr ticket for the problem +with some modifications Fix minor whitespace issues +Randall selected version +The file compiles with the change +These scripts are ripe for improvement lets do that in another JIRA and just fix bug thats preventing the server from starting here +Deferred until. +Hi Dave I think this is not a question it s a bug! Because the same code gets the different result at different environment +I dont think the issue should be converted into a project at this point because the problem stated here has a simple solution and because it is important to fix it +Not everything is working +A is stopped +Wow I totally lost track of this one +I guess Linux can handle this +Thanks for fixing it +Once a density is computed it forces the rest of the average density calculation to return too +Johnyour mirror is not compliant icons arent shown properlyThis is important and the fact that some other existing mirrorsare not compliant in this respect is irrelevantDo you want to fix the problem or retract your offerRegardsHenk Henk PenningFolder Icons are in place and ready to go I hopeThanksthanks for the fix your mirror will be added soonregardsHenk remove our site from mirrors +You should expect something next week +Fixed in didnt use the attached may be missing something but its not clear to me how to enable this feature. +I will fix the bugs and run ant test before submitting the next patch +The user files will be assumed to be securely owned by the user and so TT has the onus to use the jobs configuration and hence the users identity to download files from the file system +I think the documentation should explain this too + +seems to be fixed +Reopen due to report by David +Can you supply examples that would illustrate how you think this should work +Ill take your patch as the foundation to start that effort and will update this JIRA issue when I have more information +I havent actually seen it with but did with and see no reason why wouldnt be affected +Incorporated Nicholass just committed this +Anything else to do here? Maybe documentation +I too would like to see spock testsspecs in the same dir as testunit +The schema change is OK because it does not break created with if not done +If you can share your little client app that would be helpful you can attach it to this jira task and people can download it +I have removed the resteasy a Technology Preview of a portable implementation of Specification reference from the Admin and Config guide +on Ryan +The joinpoint classes are generated based on interceptor chains +Any opinion related to my previous comment +No tests since this will be called from other code in HDFS +Ill commit this in the next days if there are no objections +Thanks for committing RickSorry for the compile warning I did not even see itCould a committer please delete the file javatestingorgapachederbyTestingfunctionTeststestsjdbc? It was deleted as part of the patch but svn del was not run for it +datadatacurrentVERSION +If this is unacceptable please the issue +resolving this as its committed +is for and versions +sure +Let me see if I can do add that change too +This was changed in January of +So I will have to find out why it doesnt work Regard this issue as closed or at most minor +Run the testPerformance on my local PCBeforeFinished test took timems AfterFinished test took timems In the test I only create regions for the table if the region is much more the result difference will more exec result is incorrect if cached region location is wrong m bumping the priority a little because imho it means that anything that happen on +Updated to work with ll take this one +Of course you can always reopen it +I primarily tested on Mac OS +Hey JodyIn theres a known problem where the input become invisible when the finish button first become active +task is done +This one would be tricky to implement +Patch applied under r +I think I wasnt clear in my previous commentThe Karaf Config Service is only responsible of the storage and it implements the interface +Thanks Pete! +Jonathan your last test is not correct and happens to work because you already had the object loaded in your session +Changed the release profile to skip javadocs by default. +Your domain dependent stuff doesnt have to be involved at all as long as its not vital to reproduce the problem +Grammarparser must be updated before the type system and the emitter parts can be implemented +This should be fixable without any additional charset code shouldnt it? Quick glance through the source suggests that it can +That should be enough to this +This work has been completed with svn revision The sample and associated sample doc have been updated for building with the API and implementations jars of +I think it would be nice if we could do both macro and micro benchmark +Thread dump from the hung Solr +Integration tests were updated. +This patch addresses the issues raised by ystein in his review of previousonline backup patches +Yes +The registration handling mechanism needs to be to handle things better +Tag close of resolved issues. +AbhinavCan you please check your setup with the above findings and share if it resolves the issue +Somehow it is not new patch should if this caused any inconveniences +Did lots of small changes to both the Ant files and the Bnd files in order to get the various tests to run +Example code and associated junit test that demonstrate this are experiencing the same issue with on linux +I couldnt find any further references to in the Maven site +Currently the can be given to a GCL and this decides what extensions will be used +shame on me for this mistake +lgtm +More fixes and doc update will be done in +Should have been smarter and used ITMI can go back through svngit history to jog by my old brain if you need more than the above +Resolved as wont fix +If I understand correctly couchspawnkillable will terminate couchs children when couch explicitly asks but should couch go away silently and suddenly these children may continue to exist +Vincents working on it as its currently bugged +Consider for +It causes a lot of class loading problems in +Currently I think that list is pretty short but its worth consideringThanks for the quick feedback Im trying to discuss this issue with all the interested communities so that the fix we put into ubuntu is the best possible one and compatible with debian and anywhere else that uses etcdefaultcouchdb +Then we can check if the hostport are the same +assuming not for beta +Seems like the patch has gone stale can you rebaserebasedbq +I replaced the link widget by a hyperlink custom widget +issue is fixed +Again the loading of globals is independent of the runtime bootstrapping mechanism +Now that work is well underway with AS all previous community releases are +This is because of a race in +Will commit if tests pass +Looks good overall this had to be doneSome comments Move to be in for now? We can move it to public api when needed +What I am saying is provide a simple GUI that uses the client to pull the data back +I think I can get a workaround by a different approach not starting with my Java classes but start with WSDL and generate Java with wsdlcode + +Adds the ability to load multiple jar files for a single job +Additionally map output copy connections are taken into block to assure they are closed properly +Ifyou can create a small index that causes this error and put together some codethat we can run and see the error then we can step through Lucene and find thesource of the problem +This is very scary even the download of attached files appears to get corrupted +In order to compile has to be applied. +The new bundle list in the trunk define a lot of dependencies to other bundle lists to avoid the described duplications +Verified in JBDS h Fedora +The fact that the Eclipse IDE displays the classpath entries in the same order as the configuration file is not a problem of this plugin +ThanksBecause region server interval is s So I think should be s +Not unless theres already something existing that does this +I dont disagree with the result here +Made changes to remotingx branch +Your example deploys this one ve reproduced the problem with your project +If fine I will also upload it for +I generally favor including stuff as an intact library rather than cracking it open and modifying it all else equal +Slightly modified version of the suggested fix implemented in rThanks! +Assigning to AS too. +Having the accepted connection queue in the server helps if we want to limit the total number of connections that the server wants to accept +It ends up being less work for them +I think that I resolved this although I did not test itTo work around this issue in older versions of you can set the supports property to false confirm that this is fixed in +I dont want to use Spring WS at the moment and prefer using a runtime +Hrm I just tested on device with android with the branches and it seems to work okay +According to the thread Freeman linked to apparently the s still need to be removed from all the other samples as antbuild is to remain our lone Ant example +Carlos if you want I can change the version number to be +I got it twice on the weekend once on RHEL and once on Ubuntu +You can easily merge the commit today +Gavin fixed and I tested the change by updating the front page of the Whirr website +Thanks for catching it +But the patch looks fine and makes sense Oh and it compiles clean +And standard classworld jar names are or +I would commit it +Adds srcdata with a new zip file of a small hbase dataset with two small tablesAdds to reading the data file off the CLASSPATH +Fixed in rNow Java backend always generates Java file for XDT +Pushing back to the next release +Added patch that makes path constraints support transient namespace remapping and modified test cases to verify this +Any comments on thisIf no objections i can close this out +Set to back to has been releaseClosing all related resolved issues. +and why would theyBecause not all historical ZIP files use DEFLATE +Patch attached requires patch +Please check it out and let me know if I got everything into the tagThanks +You will need to use svnlook to determine if the commit is in the correct path +Im not sure how I ended up closing this one +hmm +Version with it works both with GF and and its very usefulI cannot imagine a good way to integrate beans without such a feature so because I dont want to switch to none of those two container for my IT do you have any suggestion on how can I simulate it with the embedded container? The remote one is still far slower in startup time and I love in my development lifecycle because of the very very fast startupCan I change this from issue to enhancement +Going to remove this and see if it fixes the issue +hum quick thought was that we maybe see a difference in the output if the +I am using which downloads deltaspike jars before test execution and save them into requirements folder within target folder + +I will investigate +How about filing jiras for the other HDFS protocolsissues so we get a sense of the scope of the remaining effortSure + change from static to private final change a few methods in to added javadoc and some methods +thanks Uma +Use support in order to generate needed code + We should also make it work for local file state persistenceSounds good +With those methods in place we just gonna cast around which should be done by the user of the API +Cool stuffPeter committed that in Trunk so itll be there inWe can open further issues for the idea of wrapping expressions into closure parameters transparently as discussed on the list etc +Patch for trunk +Working on a fix may not be done until next week as Im on travel starting tomorrow +As of runtime vs +Fixed rev . +The EXPIRE event is a watchdog for containers that are acquired but dont ever get up and running +Filters can only return INCLUDE SKIP NEXTCOL NEXTROW and SEEKNEXTUSINGHINT +Attached v with fixed validation +Changes related to our clustering code +All the info in till is out of date because Smooks Editor was completely the changes described in can be already found in +Applied the patch on trunk +Fixed the first two issues and cant find a reason for the third +As I said above this was not really a support patch so there is no file to upload to the CSP. +Ok check rev for me +Should be ready to go +Nick something you can do +If you find further issues please create a new jira and link it to this one. +Talked with Denis this looks like a really great patch from user perspective +It is not being fixed for or EAP because it would break backward compatibility and change users expectations of code +Thanks +Also added a test to ensure we check for these annotations +Sure after we do that the Codec can put whatever attributes that it wants they will be recorded in the newBut maybe we can solve these two problems by gening FIS Add +Hi Jessica I have gone through the doc +The attached testcase should show this pretty well +Yes youre right and its a shame I didnt think of it at the timeHowever any break in binary compatibility will force a package rename for IO here in commons +Looks like a duplicate of +Clears after we have used the contents of it +This issue has been fixed and released as part of release +The message is only used in this one sysinfo fileI intend to commit this fix in the next few days +Committed to with revision +I can send a copy of the replacement class file but will need to ask someone else for the patch used to create it +Im guessing weve gotten stricter on registered block lengths and do not consider reads or checksums beyond known lengthIm inclined to call any block that has been modified in any way outside of HDFS as a corrupt block and that it ought to be invalidated and +JACC is not transitive anyway so does not matter really which coordinates we use. +best solution is in general to use a CDN like Akamai and includes +Not yet +I think I solved itPlease try the latest code in master branch +attaching the transcript from the irc session in case its is fixed in. +Were lower technology than that now but as you all know Im happier. +LiyinThanks for the quick turn aroundPlease let us know the performance improvement compared to what you collected on May th +Attachment has been added with description updated patch with has been added with description mylyncontext has been added with description image used in patch did not work for me +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Attached new rev of the patch +Cancel the current patch for taking Asokans advice +Committed to trunk and +Makes it to be chosen currently +For both branch and trunk +The warnings have been suppressed manually so lets hold on with this proposed fix and see if we can justify avoiding synchronization here +I tried against tag and trunk +Question about the Backfiller does it require a Collector to be running somewhere and then the Backfiller sends to that Collector +BTW it would be nice to have Run code generation in context menu in Package explorer when pointing at +The only thing that we can do to restrict it is what the OS allows us to do +Yeah I understand that +the racing threads in create are trying to publish the same values to the mapyeh you got it +One would expect that since theres no descriptor simply touching the ear or ejb folder would restart the module but a simple test shows this is not the case +Ta was fixed in sometime in the last development cycle +Any way we can use something else? This will make it impossible to make an IT for this since we cant redistribute the jar +Rather I suggest a simple wrapper class with a default constructor and setters for the parameters required to instantiate and startstop felix +There is even a unit test for it +You can change it on project propertiesJPA +a minor could be to amend the error message with use a normal and connect to httpyourhostws +I suggest making some small changes now to allow it to scale horizontally with the understanding that we need to seriously address this in +This new patch is built based on the new SVNTo apply the patch please enter top level drlvm directory workingvm +Ok sure +forgot a file +Applied the patch to the Post Servlets in Rev +Proposed patch for we should ensure that the c client doesnt have similar issues +I updated the patch so it applies to trunk +Therefore the test failure seems to be caused the mixture of old jar file compiled with protobuf and new jar file compiled with protobuf +Could you provide some example strings + already has a patch for this can you review +Ill make the change and attach a new patch +It would be great if you could post a patch to fix it +The revision that I pulled down is at the javawsdl moduleThanks +Your third bullet is valid but not for a +Hi Jason this patch is not that good +Attaching possible fix and corresponding supplied test case and used the concurrent map approach over the synchronization option +The internals of the package would be byte manipulation code to translate from bytes to usable objects like doesAll the codecs could probably share the same module though i also dont mind having them separate +The SPI interface is called and it is located in the package +Add a little more to the triplets javadoc that specifies element prev next +Reopen while I investigate attached test fix I put in is deababefcacaafbed +Thanks for the work Ivan +The patch file did not apply cleanly for me perhaps the underlying source has changed since the patch was releasedThis appears to be an API breaking change since the Evaluator property is removed +Here are some corrections to points and in the description +Moving all issues to the CR +But all db was populated correctly and if we do a load the set will be build up correctly +Went back in and looked around it looks like I am for some reason building up more and more runtimes each time I run through this +Fix verified in +Tested on server with Apache Directory StudioThanks for the report John +Because formatting of the code was lost here it is again +Alsonow its possible to specify path to FOP user config relative to context. +as r +That makes sense patch with the suggested change +Attribute was removed because it doesnt make sense with ajax requsts +This is not the same as +Andrei confirmed that the build works again with JDK please ignore my comment +I see +As you can see its one of the simplest mixins that you may think ofThen have a look at the page +Patch will be forthcoming +Added a check to make sure the new WAL path is not a file already exists +Very if I understand correctly the response is still stored in memory just not parsed +Install graphviz plugin into Confluence +AFAIK there is no way to have Netty automatically limit the number of active client connections +Committed to trunk. +There is another constructor available that you can pass the connection manager through so that more complex uses of can still be used +and make this the default +Ill wait a bit for any feedback thoughant test passes after the ant Thanks everyone +I will try it with jetty. +My apologies if this is all obvious to the Hibernate team or if Im completelyThe problem is that the Loader for the entity calls end too early +I dont know in doubt I would have contacted Ben or open a ticket in the HAUS JIRA project +the securityStore wasnt calling the registerListenerIts been fixed now +Well take this up +Thanks for getting this going Adrian +Maybe only cover the class +Youre correct +Thanks Laxman and UmaAaron thanks for the commits +what should we change +This is not related to which is mostly about removing the temporary file that accumulates user datauntil it reaches the size of the block +But I think the root cause is the cache codes in Currency how about just remove the cache and delegate everything to icu Currency +srcjavaorgapachecommonslang +Fixing anything here would require some heavy handed reworks so I opened to address upping the timeout and Im moving this to out of down priority +Patch committed to trunk +Thank you for the patch looks good +Thanks to Rick and Andrew who tried this test out in their cygwin environments and reported that it was not an issue +going to rename the ect example to +Checked in to trunk +Im not familiar on how to roll back the changes but they are very easy to do + is issue for trunk so is right fix to M +Id be +and there is already such a map in I that would not resolve the issues you get when the map is used for queries +Integrating in into Serializers removes the need for building a comprehensive catalog for use in the final step before the final transformationserialization each XMAP handles its unique in translations and leaves unknown translations for later processingThis suggestion does not expect the removal of the inTransformer +Commit bfddcabddcaaecc +But what about using archives tool? Seems that the archive isnt deployed to at all +One quick clarification isnt the assertFalse supposed to be assert in test +I have applied the patch to my sandbox and I am making some changes to itCurrently the patch only applies to marshalling +Are you running with a special jdk version or a security manager active +Honestly I dont see any way round thisregards was fixed in JSF with the replacement of to +Justin can you have a look +Vitali check please if you have the same open open hebernate tab select DB dialectsee in error logIve checked this with plugins from trunk a didnt see the may be it is not your error +Fixed on revision . +Just one minor comment +Primarily a new compare function that is much simpler +As I said I need to be able to reproduce the issue to be able to fix it +Talked to Kevin about this and a solution might be to set our own ESB reply to header with a different name +Fix applied in trunk +Regardless it has the potential to break a lot of stuff via so Id rather do that as part of a separate JIRA if its alright with you +Hi AndrewI am uploading the Test files which have the SOAP request message in it +I think your performance testing has been superb detailed and thoroughIm comfortable with the performance impact of this changeThe deadlock is a serious problem experienced by many users and I thinkthis proposed fix is great +REALLY? The patch is so trivial! But the realPagepage properties have always been part of s Deeper Voodoo and I confess I dont understand their real function +Still not sure how to do it though +Thanks issue is fixed and released as part of release +nice idea to not having to know this magic DfilerBTW the first goal was more to remove the HUGE list of archetypeWe can certainly do bothAgree on the format? groupIdartifactId? Samples display all archetypes which contains in groupIdjee or jee display all archetypes which contains jee in artifactIdjee display all archetypes which contains in groupId AND jee in artifactIdmakes sense +This command line tool uses addIndexes and therefore merges all segments into a new index +Naming of the themselves will be addressed afterwards +Closing +Thanks for reviewing +Change of test result is too big for a phabricator entry +Sounds good Ill get to it after nailing a few more bugs but still in the timeframe +Ill work on this +Maybe include? It could be more meaningful + released +See for full explanation and the options available to you for working around it until we get a proper fix. +Lets see what comes back before deciding whether or not proceed with a fix +Added the failing test and applied the patch +Reopened to set the Fix Versions field to JDO beta. +So maybe that part of the problem is simpler than I thought it was +The user then has to cancel the wizard and check if the application was created +Are you using +Committed to tez branch. +may be i am missing some jar file +adjust code and configuration to point to the new API classes +Yes it sounds like a bug but the question is whether its a bug in the Struts framework or your application? Without knowing the cause the etiquette around here is to first post a question to the user list since often these type of things turn out to be a user error rather than a bug in Struts rather than doing what youve done which is assume its a Tiles bug firstAnyway as I said before what matters is what html is being generated if I was you I would start looking there comparing whats being produced by forwarding directly to the jsp vs using Tiles +Update Clover tests are up now as well +If this is an actual reported EAP customer issue AND productization is unable to help fix this I would patch otherwise any bug for not meeting the above criteria will not be fixed +Well not for Beta I assume Can it be separated out without too much problems? Then yes please. +Resolved per pull request +It shouldnt be the job of the user to teach the tool how to do its code generation properly so that it integrates nicely into +Attach patch v +Fixed at revision +The documentation is updated +EricThanks on the and fixesI am experiencing many more PDFA valdations than bugsKeep up the great workBRBill +No it cant +Seems everything here is resolved closing the issue scream if anything is missing. +Note this is also related to that one refers to the same problem with +Sorry covered by +I used the menu to create a patch so I dont know why it didnt work +pushed in Developer StudioVersion Build id Build date +Im going to close this for now. +What was the fix +With this patch I have a new simpler version of and have made it in such a way that it enables us to shutdown a number of leaders forcing a new election +Well some people send scripts with svn move along with patch files if they want to rename or delete something +Thomas Ill look into this +I dont know whats causing itBut the assert that is generally tripped on delete is tripped because its trying to uncache a file and finds it in the delegate and its trying to assert the file is not in both and I can seem to cause that condition kind of easily in a simple multi threaded nrtdir testIm also not sure that removing files underneath this dir would cause that situation it would seem not +AshwinMind creating the patch again? Looks like everything got duplicated in the patch file +Related to this I see that there are no node types for Long and Double only Integer and Float +Patch that should resolve this problem and now open to discussion +It should be part of Service layer see +The main difference is I think that otherports takes some descriptive information telling TS what it means +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Exception is at line as deployed and zip of exploded tmp you use the latest nightly build of IDE +Im not sure what kind of configuration you have in place in your project but Maven does not build a JAR with source code or dependencies by default +If there is a need then these need to be documented +This is on purpose so as not to log any security related things +Ive added additional messagesclarifications in v of this patch file. +Please dont comment out code but remove it instead +I just added this rule to file in the project attached to JIRA removing the other rule +Because in the issue I mentioned you could put the in the bootloader and it would behave well but that wont be easy using the groovy command alone +The patch is against trunk and only shows changes to GOP +Patch with all tests and code converted to not use the deprecated API and move to use IWC +The download instructions have been fixed +I cant confirm for any other platforms but it may affect those as well +Ive attached a quickstart demonstrating the fact that when posting to the event bus all clients registered to that resource receive the message +The rules of the repository are now correctly exported +Patch is applied at +Looking good now thanks will add over next day or so +And this breaks +If a custom request handler explictly calls the set method then it will override the configuration in Do we need tests for thissvn stat followsM srctestsolrconfM srctestsolrconfM srcjavaorgapachesolrhandlerM srcjavaorgapachesolrhandlercomponentM srcjavaorgapachesolrhandlerM srcjavaorgapachesolrhandlerM examplesolrconf the patch with the trunk +srccontribvertica and package +Null is fine but there should be a default value constraint set on the DB otherwise the assumed not null functionality of Version will fail eventually +Can you try the latest from trunk and see if you still experience problems +i tried different applications but no youre still able to reproduce the problem please send a +If you finish it before we release beta please correct the version +That is the only difference +Note that there is nothing to run and no unit tests or other verification of transactional is just enough configuration to enable gutter markers showing up for configuration and configuration that should cause gutter markers to appear when this issue is let me know if there are any questions +trunk +A third digit rev shouldnt break users apps unless we had no choice due to major server problems and changing Dojo levels just to be current doesnt register as critical in my book +The default frequency on linux is seconds +all the crap you want to use a object for +is missing and the stuff is also missingCan you please attch a complete patch that can be applied withoutcompile errorsThx m sorry for my missing of the dependency +A disadvantage of nabble is that we do not have control over it +Both of the above can be done in two or three lines of script but and for the congnescenti +The report incident example unit tests are failing now +There was a broken spec which I fixed +All uncompleted issues are now moved to the new To be determined for future releases category +maybe the xml updates via are sufficient +bulk close of all resolved issues prior to release. +Instead I think the best direction for would be to invest effort to support pluggable authentication of clients at the RPC layer using the same mechanisms under development for HadoopHowever if someone does want to invest the effort to support an impersonating proxy for shell commands as an optional service that is completely up to them as long as it does not undermine core security +Thanks! This must be the same issue as but marked as related for now +Uploading a patch very similar to the one posted for +Applied the initial this one still failing for you? I wasnt able to reproduce it today +mrepositoryaopallianceaopallianceUserstucu +SVN version . +is it having same root cause +Anything in your +Thanks Amar! +I think the code could be cleaned up a bit can you take a look at the attached patch and comment +removing assignement from versions as it didnt appear to affect them. +to trunk +Good argument I didnt think about thisThe most secure and user friendly solution would be to catch and the thrown by setObject +Closing as requested. +In general I think that this isthe right thing to do as restart processing needs to close all the sessions not just those that arecurrently waiting for a free threadHowever the new code is too powerful because it also closes the current Session which is thesession which is performing the shutdownClosing the current Session terminates it before it has a chance to format and return the Shutdown Exceptionmessage back to the client +i applied it to my workspace looks goodwaiting if there are no for this change +Attached patch makes the change +Thanks for the patch Colin! +Patch that makes create private and changes the remaining callers to create +If report should be opened to regular users it should not be a command but rather a regular shell command + patch looks good +Youre welcome! There may be some more coming Im having further trouble actually executing the query once it makes it through the parser but I need to do some more digging before Ive got anything conclusive +We want to lock down the versions used at the top level and enforce it among subprojects +We were working on the createTable part of this issue internally as well +What makes you think Im using EclipseI managed to get the repro running in by generating a freeform project with Maven and then creating a Run classfile targetThe repro demonstrates the problem and I guess the first thing to do is to figure out if the repro or Hibernate is doing something wrong or if Derby has a bugBTW Ill be away for two weeks +Looks good to me. +Martin I think Ill drop section completely +The bug could probably be fixed by faulting in the metadata only when it is needed rather than reading the catalogs when the AST is createdI also agree with Knut that the devil is in the details +Also if there are multiple filter predicates after TS there are all merged together and a new filter predicate is generated under the TS node so looking for TSFIL should be sufficientRegarding the optimization I will keep the old pruning code around and toggle this new functionality only when predicate pushdown in enabled +Would you please try the patchBest regardsSpark attached patch contains changes to which isnt mentioned in this issue at all +Maybe its just a problem with Eclipse I dont knowIf anyone can improve upon the patch they are welcome to do so +Heres a fix which uses Guavas before creating the iterator +Check out the examples that are maintained in SVN +I also dont think were in any immediate hurry for this since weve got other options in +I think its a result of +Committed revision Moved the term up to the abstract classImplemented Hosss Expose Writer option +Fixed in revision +Addressed cos comments +Update daytrader trunk to use JettyPlease check and help commit thanks +So I think we can close this +yes thats more or less what i meant to say if you put it this wayi mean the point of the should be to set an example of the conventional usage of the framework +Not sure what happenedIngest ran for hours +I agree +AFAIK touch should create the file if it doesnt exist +Waiting for your suggestion on that +Some java tests are added in +Updating the patch now that the blocking JIRA is in +This task needs updating to say what has been done for what will be deferred until looked again at all cases and there is no remaining reference to sun classes. +TODO Add wiki docs for this +I mixed up name and referenced elements +Eli Have seen my previous comment? Actually there was another potential issue here and we have raised mainly to address that functional issue +Assume closed as resolved and released +You are right this behavior can cause problems +The system packages extra should be commented out but we already made a mod to support it being blank in forNot sure about this latter issue +Think this could be linked to the problem I recieved about not working with the plugin +Updated test to move the start of the timing to the right place and ensured coprocessor configured correctly so that its used in. +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +You would help me if you could provide me with a sample project that fails so I can dive into the problem without worring to get the configuration exact as you have it +Deferred to close of resolved issues. +Can you please open a new JIRA so they wouldnt be lost and add a comment with the number hereDhruba opened for handling this +And thats how I realized that these are required for client also +thanks applied in cvs +Other files to fix in devstudio repo +a bug in the way the points were handled in the transformd method. +Ive verified this is still a problem in +Some more missing deprecations +We can may be should a need arise +So we should better avoid using property names like b in future +I can introduce a new expection called that is a subclass of +bq +Attached is the test case for reproducing the problemThe instructions for executing the test case are outlined in +This is the new quote item screen with the correct calculation +on the patch +Just to clarify this is intermittent +Also in a large task the cost of having every mappertask scan all the messages could be fairly prohibitiveId suggest making it available in contrib or some other mechanism until we see how much uptake it gets +did you try theses settings? wow +Dale I am closing this because I applied your patch thanksFor the other issues if you want to create more feel free. +Thanks Rich for the patch +there is no java code in the patch so the for javadoc isnt a problem with this patch +It is not possible at this point to know whether it is stateless or not +modified zhengs patch by adding an logj error message in addition to stdout +Ive committed the patch + +Earlier versions did not do this and you could use it without a mail server +Patch for the described patch Sending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavaorgapachegeronimomailutilSending srcmainjavaorgapachegeronimomailutilAdding srctestjavajavaxmailinternetTransmitting file dataCommitted revision . +HiChecking this part it works on but stopped working on +Waiting offline for the reply from Paul Gier concerning the maven file. +Piotr what is the status about it? Need any help with it +In my case the only unique combination of columns contains a nullable column +Since the contended case is likely to be where people want the isolation that seems reasonable +The correct version of the patch +committed to branch and trunk thanks for the work Gary! +Verified +I verified that if you create a +When an occurs either rethrows it as an or logs it at the WARN level +comments having variables one named fname and the other named filename is confusing +JamieI was running into serveral issues to do with class and javax +Commit hashaefccfedccChanges There is no good mechanism currently to figure out if the deployment failed due to affinity groups only We can just hint the user that the deployment might have failed due to the affinity groups and ask to review the inputContained in branches masterContained in no on Build. +Created issue for fixing the found discrepancies between spec v +Fixed by +Please provide a quickstart and we will reopen the ticket. + +It should fail until goes patch couldnt possibly fail any of those tests because it doesnt have any srcmain code +Adding Matteos Patch all tests passEdell +Let me see if can get better reporting a reporting that will expose clumping of syncs or syncing even though only a little amount of data has been written +Jukka are you still working on this? Otherwise I can take over since I stumble upon this as wellThanks feel free to take over +That would be a nice step towards integrating with the core codeThis is something I have been remiss in I wrote a job to do the of diffs for slope one but never wrote an implementation of that would read this output and operate based on those results +This Bug was meant to be raised Against NOT DBUTILSPlease to be raised against not Fix Version since dupesinvalidwontfix bugs clutter the release notes. +I dont think the Struts trunk build is doing it automatically yet +Propagating runtime exceptions with no warning is a bad practice IMO but I understand that other developers do not want to add null checking everywhere and I am not going to insist on it +BecauseDerby optimization is costly best practice is to use ? parameters +PS While this license could be seen to be GPL compatible I cant imagine it being LGPL compatible +I switch back to the latest HEAD build and it deadlocked again +Is the patch file a suitable way to check the numeric type? Comparing two different types are equal in for these tips Ill change it and send a pull request soon. +Applied patch marking as closed. +The current implementation uses a Map containing every indiviudal value +Patch applied some time ago already +I suspect the author of that Implala code just made something up since theres never been a documented or implemented boolean representation in TrevniThat said a packed representation might be better +Ive thought Ive bumped up the ignore count for javadocs in might call it a bug in javadoc but there is a at the end of the override parameter name so javadoc thinks it is called override I have a patch and will upload it shortly +I prefer the second I think its good to user to read on which elements the assertions is based +NormanThanks for quick fix +It was probably just an issue for me in what Im trying to achieve with my generation process +Patch applied with an additional test case +John +This is s test PDF which contains descendant font +This is a VERY different way of doing things than most of WTP does it which is why it doesnt matter how WTP does it +That exact length for columnc will trigger the edge case youve found +After a detailed offline conversation with Owen we decided to not do this for Hadoop +Mumak uses a simulated JT for telling TT some informations of a reproduced by rumen through heartbeat +Sean and Leo seems this is duplicated with and theres a more complete patch provided there so Id resolve this as duplicated please let me know if you are unhappy with this. +Replace occurences of by creating a for each test no version fix in is more usefull if it can use an existing +to trunk and branch. +stuff in parent +I didnt understand your comment Harry +Probably fixed by Andrea in. +filed to fix issues. +will take a look +I am going to resubmit this patch to verify +After running through this again I came up with a different idea +I have noticed in a recent update from master that this issue has been corrected. +So probably the difference you found in the two machines is due to one having the native JAI libraries installed while the other doesnt. +Fixed on trunk with svn revision +It would be nice to have as well bu its more mandatory +Galder I moved this one to Backlog until we have a better idea when it will happen +But a few questions +debug to master and +Username thetooth +Made by Antoine Bourre Louis Briolet Alexandre Lassalle and Amady Diop +I think they make its use cleaner and I dont think it breaks anything but please verify +There are other occurrences in the books +Please check it under XR isnt reproducible under in +Ive seen nodes with loads over timing out all sorts of requests from other hosts and Ive never seen Parent died logged when a tasktracker was really still alive +Since this is J specific portlet it should not be added to +At first I thought we didnt need the property in Hadoop due to but actually it would allow users to mark a job as on a basis +If this is not OK please fix it or exact tell me what to do +Thanks for the patch Matteo. +Did you happen to include em +Ill see what I can do about this in the Jackrabbit timescaleInstead of synchronizing things on the instance Im thinking of adding a class that encapsulates all state that needs to be guarded from concurrent access +Rejecting this may be superfluous with MVCCStriped locking designs for. +Wasnt based on jdk +Note that this patch is to be applied to Castor and not the Spring classes for Castor XML +Committed at subversion revision +bq +This is just such a hassle when it seems that an easier solution could be implemented +This is how the +Simple example of the we include this intoThis sounds for russians pretty if you try thTH? that is also a +I didnt mean to say implementation was wrong +This is the image from the silk icon set +Thanks! +should be put under test filed. +I wonder if we should add RPC and connection counters for each protocol +The children problem and its not immediately obvious to me why ZK does not support it probably needs a recursive solution since children can be nested? OR in our case we can make the assumption that nesting depth is ? +Just to be sure that it was not fixed since +not sure why this didnt get closed with last action. +Applied at revision . +Committed the backport to the branchNot tagging with a version since there are no plans for an official release. +Rerun tests to either pinpoint problems or close as to +committed to TRUNK +although it built without any errors or warnings it quietly omitted the Servlet logging components resulting in no file log output +And then resolve + +Done MaxAttached a patch of onlyThe others mods will be put in separate issues like you suggested. +Like yonik mentions above though +The problem pointed outI feel should be addressed as part of the stop implementation hence I will move the issue pointed out to +We should really avoid that to prevent bugs from the beginningI would simply add the commit that and leave the rest unchanged +yes +Looks good +It also adds denials Permissions to Role +If it breaks other code were going to need to justify it +When generating the patch the revision it wrote was Then when submitting it to it fails saying theres no file with that revision numberI then ran svn update and the revision number increased +This is a revised patch that refactors the constructors of both java and scala into a common API in +Thanks Ralf I am closing this issue for now +Fixed in +In this case we would have to change the concatenation of the two expressions with +CNFE fixed +Committed to +You have a patch Jieshan +Prasad can we close this out? I support this has already been fixed? +I can confirm this problem still exists on +Cool just saw that you committed this +Patchset to conditionally write row and to use always when a custom filter has not been written for the row + title should beMeta Characters t n r in String LiteralM was missed there +Such components should have Task oriented tools for target Professional roles +Confirmed as fixed by the reporter. +Changing this to an improvement as it is not really a bug though I agree the current implementation is not that pretty +I know very little about Objective C and just use xCode to build my phonegap Sencha Touch app +This patch contains a fix for the described issue +You shouldnt need to remove this node it should be removed after a minute or so since it is a fat client +I keep the issue assigned to you so that you can provide a usecase I can reproduce +Something like is probably clearer +One more proposal +Ill dig into this test case +Where can I vote? +Committed. +Quick starts have been created from the examples and committed to git +OTOH they probably care more about successfully migrating their app to and this is just getting in the way +Do we really? Are there required features that cannot be supported by delivering HTML? Will those features be guaranteed not to change from potentially compromising compatibility +Let me correct my previous comment where i hit the add button accidentally +Theoretically we still violate the specification but its in users hand to use the FID Filter or not +is a now +I just wrote it down here for posterity +Will run tests and commit +So far I am unable to reproduce this +We should try to port this to the new web ui at some point +Thanks Sijie. +Updated apt the dependency file +Id be happy it you applied it +good catch +Duplicate of a JXR component issue +Tests passed cleanly for me on +With this patch itd only be possible to benefit in all implementations from a new encodingdecoding method if you add it to the Output class directly which is only possible by changing the classes in Lucene not from outsideThe problem here as so often is +Reading the description of this ticket I cant get that to match the patch +Committed to trunk +We spent almost a full day hopping between and and finally cornered it to the patch at +could be solved in cooperation +should I submit a separate patch for this? what do you suggestregardsJoe +Please go on +It isnt comprehensive but gives a brief summary of all the operators as far as I understand them +Unless they will move to flume in next release +agent restart would pick it up for sure +Use the forum for usage issues + +Rob turned a doc issue into a renaming issueI definitely agree we need consistent terminology here across documentation tooling and the runtime +Addressed through changes in the +This means this Gant problem should go away if any of Groovy or are this is the case then this bug can be marked as Not A Bug +I did the same but copying the patched jars from the JIRA issue above +When JBOSS framework is used the service sends the response to the REPLY URI and it also attempts to send an HTTP response with the full payload back to the client and throws the exception +Were spitting out a warning on first access to such an invalid property now in case of ambiguous setter methods found +If you find some files without a proper license header feelfree to this issue +The system is now properly assuming that an unconstrained T is of type Any instead of type Object and there have been some changes in C codegen as well +Compared to the overall download times the seconds should not be too much of an addition +Do not backport this to Branch +This sounds like an acceptable behaviour +It is complicated basically you have to remap some to preserve them for the dom and then unmap them on the javascript side and do some CDATA statement suppression on other partsStill working out the details to get it really right +does not handle the exceptions correctly +has been released +It appears to me that this bug has been fixed by the work on related issues. +Problem appears to be gone in build from current HEAD +I dont think theres any need to declarecreate the Configuration object statically and doing so might be asking for trouble +We should also update the corresponding mapping in servicesrcjavaorgapachehiveservicecli +Findbugs version in public repositories has the correct published works for me Ive not gone through the report findbugs will run it +We probably need to that approach +Fixed by +In addition to what you point out as well. +One place were I differed is that I made it possible for the user to choose different ways of encoding the data in Accumulo +Patch to delegate target url resolution on successful ignore the first patch this is the right one +Lets try to get this into +Sounds pretty useful +There might be another issue around native Java exception handling +This will require N vs N RPCs as compared to the proposed method +My second attempt at this patch +get the Location response header +Applied the new jars +Willem as I mentioned on the mailing list +Cloudera repo integration for CDH dependency +for the patch +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +This gives correct highlights +Yeah sounds good +So if both the type and extends attribute are absent throw the error? Thats a good business rule +Correct I was testing with the release +Thanks for the review +LiDo you have time to address Todd and Liyings comments? so right now +New patch file that allows the DBIxxx types to be bound into a statement as expected. +I think your right that theyre closely related +added the build number to the mail template but Ill leave this open for the ant script the request was actually completed closing +Here is an updated patch addressing your commentsKathey +These menu items are a clone of a static held menu tree so it passed into the Velocity context and be rendered in a thread safe mannerregards Malcolm GeoffHave you resolved this issueregards Malcolm I am still seeing an issue +This include should be specific to each instance of so package should be modified to +Now that work is well underway with AS all previous community releases are +Uploading which lifts restriction on role names to chars it is now as for other identifiers cf +No problem +This patch follows the pattern of other implementations of the dependent drop down but does not display the StateProvince field correctly +my patch was quicker but you comment appears first +Verified in ER +I think the full statement is needed only in the autogen file +ivy on purpose because of the chrooting +patch checked in +Thanks for the fix. +hover classes cannot be removed because we need to control buttons in disabled mode +Ill resolve it for now. +Attached patch with a testcase for apis +This doesnt work with yet +Try navigating the jboss pom forest using find and grep +One final statement I still believe that in the ambiguous cases where core classes read arbitrary bytes and write bytes Accumulo should be agnostic and not make encoding assertions +Is there a workaround available for thisHaving simple apps which require all the user permissions will make users hesitate to download and install the appImo this issue should have a higher agree this is a show stopper +From the peanut gallery of users Whatever the solution is there must not be any arbitrary limits along the lines of no one is ever going to write a program larger than KB +This depends on work being done for specifically support for subsystem specific deployment configuration. +Please it or consider opening a new one if you still experience. +We know Roo works on a Mac as I have one +Create an entry favorite +Anyone have an update on this? Its driving me crazy I have half a mind to just go through the code and change all to +Hi LuisI can not connect to the mirror have you removed it do you still want it adding +I added more logging to +Can this be closed? Looks like its been committed +It turned out that inherited create fromThus when I invoked was called actuallyAfter replacing with new it workedIll replace with something else when comes aroundI guess it is safe to close the issue +Fixed in r and r. +Patch available. +Went through the latest patch one more timeDidnt find where is passed as readpoint +Exactly Im having the same problem +Same patch only w better name +QA Report has been generated could you please have a look at the report +Remove background thread from and let it called by the monitor thread in Beside this there are some small changes Use exponential backoff when query job status +In a clustered ESB environment multiple Aggregator instances are running on the different nodes of the cluster each with its own memory space + + this causes problems with git +use saveAs mode close this mode is not convience although I can select directory but there is a saveTo textInput on the top +The patch is ok for me the only trouble is that it seems you changed the indentation of the class so the patch shows that everything has changedBTW I wont have more time than you do to improve the feature currently most of my open source dedicated time goes to Ivy and the move to ASF but Ill integrate your patch later this week +Checked in without review since the changes were trivial +Has to be installed under the codestyles configuration in here is an Eclipse code style for Lucene and its children +Those email notifications will be lost but at least the new ones will be sent in time +M +Correction +Ill put your name in the commit log message +Thanks Steve. +Is this already part of the security auditing features the security team has added +Ive started trying to make the styles more unified with the rest of the UI +Next iteration +See the forum thread for a suggested workaround if you havent already +This meens we cant use the component on Windows platforms with one of the most used internet browsers! +That will indeed help as I find time to look things over +We didnt change the behavior at all and we just spit out a couple of extra lines of logging +Yongqiang please notice that I updated a latest one that fixed a couple of contrib tests +Before making request to ganglia server ensure that heartbeat wasnt lost for this host + +My web project has the Tools natures and builders that are expected +Actually we cant add new param in without a manual operation +Added a workaround to each +Ive tested your fix in CVS and it works greatTime permitting I should have a test case together by the end of the week which I will attach to this bug for submission +unable to reproduce it with trivial cases +Do you meant that I should call afterfinished post the request +Fix for indicates it is targeted to be released in version +yohoo! we re going closer to code +Convert status to enum instead of String +This is because processing the request attempts to get the first child of the soap body which turns out to be the indented text +However Im not sure what wed do instead returning null or empty string seems only slightly worse than the current default +I do intend for it to go across all branches +Comments +This patch seems to resolve the problem +Error messages repeats every second +This is the fixed which I used to workaround this issue +something like instead of having a dir includeset you can say Includeset or something +But one mistake is there are other tests to add Test annotationThanks have just checked it will be applicable for and versions also +This issue applies to codebase prior this issue scheduled to be fixedAny reason why it applies only to versions prior to? issue was created for the codebase +Thisleaves the class as an extensible class that canbe extended as described in its and replaces the implementation with your implementation which I agree is a far better solutionI would prefer that the interface remains as it providesa generic way to easily identify a class as being able to add andremove listeners +Seeing as else has expressed interest in this Ill put it together following the same style as was used for sinks and the warn mentioned in the review to for not getting to this patch earlier +This has been resolved in trunk +I view the sparseness of Thrift as a key feature and am more than willing to pay the price of using accessor methods to maintain the isset for me +Removing from as this is not really a priority at the moment +Ive not done any performance benchmark yet +Youre right I think I got wires crossed in my head +Ill play around with it and try to find a starting point for implementing my own container +What would you suggest? The code is not checked in yet so we can make changes as necessary +Daniel can you handle of this please +Suggestions Add a subclass of named something like and add your lock versions to this subclass +Patch applied +Patch committed to trunkThanks Hien! +These blockers and critical issues are resolved but not verified +More importantly the heartbeat and the HRS closing the region is not logged anywhere +I hope you are not teaching users to use this syntax +Sure it is given that could be created on the disconnected nodes and have the same cfId assigned after the schema merge system would crash with RTE +Could this be addresses or WONT FIXED before the release +Try the latest again +Script to set EOL LF on Unix script filesSome projects prefer to use native for these and fix the EOL when creating the archives. +Patch has been applied to trunkI dont have a good environment to test the changes right now +The patch also fixes an unchecked assignment warning in the same applied. +what is the most for such a file? In a +Ah gotcha +New patch which removes ignore improves javadoc and fixes a test bug for affix fileI wonder why equals did only lowercase text and not text in the compare +The same approach is used in the case of multiple sstableskeyspaces +thanks +comments on comments +Closing dupe. +This is failureAnd this required the spring components test first attachment testCasee will be more clear and simple than the second one though they work sameThe second one contains some assertion that I forget to by all resolved tickets from or after hsqldb upgrade +Ill figure something out +Can someone close this as fixed in is resolved as fixed and I suspect the actual problem was that our faking of exec in is probably to blame +Added patch that implementations described behavior +tests passed +fixed +About the bug that wrongly resolve the hostname starting with number yes +This action has been deprecated for some time now we can safely remove it from the codebase +Dont close the issue I have too add french bundle +Im very sorry about these extra commits and late changes +Will commit this shortly +Merged to the branch in revision . +As promised the test client I have been using to I tried to reproduce this issue with and I was unable to do so +Briefly +Ive pushed up to npm to resolve this. +generated javadocs also show the parents javadoc when availablemvn javadocjavadoc showed a lot of warnings +This was closed by mistake +Thanks. +Without the prefill it actually has in the pool after startup as you would expect +New patch updated the regex to ignore JMX for patch v +I think we can close this as the test is excluded anyway +Task does not know the real user +Copy the template to HADOOPCONFDIR if it does not exist +Drew does it even run on a cluster? I dont see any part that copies the content to HDFS +Since this is there even in there is no rush to fix it in +well ok then in +I think we may have to gossip them at some point however to deal with ever increasing contexts but Im not yet completely clear on that and I dont think its an urgent matter +Just a couple of comments +Yes +bq +Patch does the following Sets the correct value for the property cursor +Adrian mentioned this should be good now +Added a second pull request which fixes the issue and contains a test case to validate the fix based on the attached file +Aliyep +Since Command Line Tools to be installed is a requirement issue fix will need to be pushed to so we can properly test document etc +Thanks for the patch +new patch removes redundant logj logging when exceptions are thrown +Ill commit shortly +Link to moved the issue to HHH +Regd buffers for readers will remove one out of BUFFERSIZE buffer +Further work for reporting the diff to users will be done in separate jiras +So yes there seems to be a similar issue with Undertow. +Bob these two bpel examples have been there +You also mentioned that Jira needed to be customised tell me what you mean by that +I know this doesnt really help in any way but to leaving feedback to the eg that this is a silly idea +I have investigated that closer and after the changes the global element list on the one hand contains more elements than before +classes +I intentionnally ommited the tests to reduce burden +From what I can tell by reading the Sun javadocs everybody should have needed the get permission and certainly when I wrote a simple standalone test program the program failed unless I gave it the get permission +Sorry about the last comment it was for another reached that page clicked on download but it just seems to sit there and wait doing nothing. +Most of the Spring Gemfire factories simply invoke the native and register Gemfire objects as logging to +Updating priority to Blocker +No time is my enemy at the worked on it tonight but I decided to stop +HelloWould you please try this patch? Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +verified by Mikhail +Note that no new unit tests have been added +If your file is encoded your code will convert a file to in jedit select utilitiesbuffer optionsencoding the compiler should convert input files or strings to encoding if necessary before passing them on to antlr +A similar test will be run again soon using the releasee in a random fashion is to mean that the key within the single database is a Perl random number +Vincents original comments look reasonable to me but Im not that familiar with daytrader +IFEXISTS javadoc seems to be incompleteIf there is no active context for the scope to which the bean declaring this observer method belongs +or so ignore directory entries beginning with those + +if you need any help just ping me +Thanks for the patch Han Hong Fang +I will remove then. +Thanks for sending over this final pass and sorry we didnt have a place for it earlier +The proposed patch performs a check on the number of nodes that get added by the update and prepares an index reader when the number is higher than the buffer size +This way we may also add it to the later on +Patch of refactor into XML plugin +Conversion should be part of even though it from +Did I get it correctly +Tenzin since your issue seems different I opened a new issue under +We should probably at least give the run method a chance to finish before we hard abort the httpMethod question is how long do we wait +Hi ThomasShould we release note this for CP? If so can you provide some Cause Consequence Workaround text if we arent going to fix for CPCheers there was no ER we should include this issue as a documented issue like you said Jared +Ok so what can i do to workCan u give me a wayI really need finish itAny tricksTkz +Committed. +In furtherance of which References to articles are alternately to Wikipedia or Wolframs Mathworld +Submitting the same patch again to get Hadoop QA running and see if the findbug issue was because of this patch or not +Also updated error messages to be more descriptive +pushing off remaining issues in order to release version in JIRA and create release notes +I related the other JIRA with the build file changes. +Since it has API changes +Patch against patch deals with the issues in without the nasty +patch looks good +I believe it is done +i pulled latest trunk and applied patch init seems that communication between nodes and works as expected +I have attached a zip file with my a diff on and my annotation class. +Sorry guys +Thanks Sascha +on Alexs suggestion +GridI did not create any servlets +I will plan to commit this by the end of the day +It seems to me that geronimo should not be doing this kind of stuff the osgi framework should +Thanks +from cvs is running perfectly in our Xeon HT processors with +He obviously wouldnt have taken the time to post here if it was clearIts not clear to me either +We collate the policy information from the composite component referenceservice binding hierarchy onto Endpoints and +I have also left code commented out just to illustrate what I have in mind +Both I and Vladimir said that much on theTeds analysis is only the story though +Indeed the Source Generator uses reserved names like Item Choice Sequence or Types and the only work around is to specify in a binding file the name to use when encountering menuItem +Jakob This tool has been added to support migration to HA NN for Hive customers +Because this is governed by a property this should not effect the ability to release Trinidad Added a profile that will allow you to run the Trinidad Demo in jetty using Pluto +Why? Id like to hear more of this discussionMainly I feel the Abacus package proved useful and fits into nicely +Its hard to reproduce and hard to fix when it does happen to the point where I try to avoud using OXM from both Spring MVC and in the same application +If this behavior is desirable for Drools we may want to apply the fix to +I like your proposal +Would it be make sense to use to determine the home dirYes someday +got tired of seeing +How do you make the redirectCan you provide a quickstart application please +Already want to make big changes to what I wrote. +I dont think the bug actually applies to trunk since it looks like the latest successfully handles the else condition in move +Next I am planning on pursuing the broken case cost estimates during different join permutation because it seems like we are ending up with scenarios where the cost estimate shows the cost and row count to be and hence we end up picking that plan because it is costing nothing to pickup that plan +I will review and commit +I do not see the hang when running tests with IBM +Ok ill take a look +does this affect too +Patch should apply to or +warning message? +r woden branchr woden trunk +bq +Thanks Andrew that seems reasonable to me too +With the patch the exception no longer occurs +has a problem with lookahead that causes exceptions from the parser with attempting to parse poorly formed SQL +For p s of or less a single long can be used +Reworked to fix of the findbugs issues +This flag is confusing and might cause problems if its turned off and on back and forthIt looks better to me to just drop it and support the whole feature +We should stick to double quotes when we need identifier disambiguation +Many thanks for this patch Scott humblest apologies for not applying it sooner +I believe the patch Im attaching achieves the same goal and is even simpler but Id like to make sure it suits your needs VishalIf you agree with the modifications I can work on a test +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Ill update the file to include it +Kind of invalid since so old kind of a duplicate of. +This is not a complete replacement for as intended so no need to rush it in +I did not change the base classes of surround only tests +Supporting annotations on expressions and LHS of assignments should be taken up as a language extension post +Is the special case shown here worthwhile or could it be accomplished in another wayAlso added some additional filtering for the header in the response if the handles it then we should probably remove the existence of that header from the client application +The file is there now all tests are running +new grails project +Since row filtering is probably the most common case it could possibly extend a more general iterator for ease of use +Moving to in preparation for release +I had every intention of having this in but you can see it took two commits for me to get it right +I have actually produced a patch to solve the issue initially spotted in this jira and uploaded it to +If you dont know the size of the input you can use the interface that allows you to signal the end of the third batch at even if the batch size is set to could take a look at this again for the release to see if we can come up with an alternative implementation that would be more flexible +The order is pluggable so it could be completely customized +The test ran cleanly for me after I made the following change toI am running the full regression tests now +So there you go +Im using Maven think I have found the issue +Incorporated hemanths comment +See for some historical background +replaced by as of r +Also the bug seems to be fixed however I dont know if it is just a side effect and there is still the mentioned problem or it can not occur anymoreSo far it seems good we havent test it yet on our application currently there is no spare time for this +Shane please review my changes or reassign appropriately +Please confirm that they have been applied as expected and close this JIRA +Could you give me the XML of the profile sonar where you exclude api please +On the suggestion API changes should be avoided in CLI better to look at CLI and consider them there +Yes but what about all the devices that do not support JS? Youve changed a to input for everything not just for PDA. +removed unwanted classes +Here is the patch that will be committed to Yahoo! Hadoop +I checked in the fix to the trunk and deployed +I will report when a test run finishes +I think if you said everything level is repaired that would work +Added some more clear instructions on how to installAlso tried to have a go at transforming it into an extends class but failed to figure out how to assign the onopen onmessage onerror onclose listener functions to javascriptIf anyone who knows more on phonegap plugins then me wants to have a go it would be nice to have this available on official list of plugins +For the record do you agree to include these test in jaxen under the BSD license +If you have any idea what I can do please let me know +Thank you want to get to the bottom of this just as much as you +The outside phrase and then the inside phrase +I reviewed the code and it is almost completely rewritten in this area. +I think this is fine with the browse handler since all of that is example data schema specific but select should stay pretty general eh +Perhaps we can make a Temporal utility class for main to make creating these things a bit easier +If you stopped resubscribing the subscriber got nothing notified about its subscription is stopped unless it registered listener about it +warning a lot +Ill try to create a test to reproduce the issue +Tx +I assume you mean there is no need for groupbys anymore since we have aggregates +ThanksJust one thought I am not aware of if the putMethod closes its +Andrew committed the master update patch as revision +v +See for some discussion +Move all main classes from to mpt Done some time ago +Patch changing both equals and hashCode +Committed variant in r as the empty map is by far the most common this avoids the use of a hash map. +jar to fix the to QE for verification +Committed to trunk. +Well this doesnt solve the problem of the person whos on the BCC receiving the message regardless of the ToCc which Ive checked with a number of different email programs and I ALWAYS get the BCC even if the To and CC do not +Why would you want to relax the verification +I just committed this +The output is still not printed instantly but within seconds +Please go ahead +Here is a new patch +This failure isnt easily reproducibleWe can move it to report of failing unit test + +Not your change but would be less error prone if used eg an enum CORRUPTIMGXFER instead of + +Daniel Perhaps Im not thinking hard enough but I cant come up with a scenario where youd have unintentional or duplicate relationships +Moving all issues to the CR +But I wouldnt mind a review Johan +We should target for fixing this +is for the general case and it was all written in which we have dumped for Jasmine for some reason +I also added a link to stable in the served up by most mirror ftp serversIt will take a day or so for these changes to be visible on mirrors +Marking Patch Available +In fact It would have been even better to write the tests without any database interaction at all +The java contribution Open Type +I have tried to run traffic for a couple of days which does mixed operations and cassandra is still runningHowever i just started to run a traffic testcase which does only reads and the issue is back directlyI dont know if it matters but all my reads returns empty responses which is expected since these keys should be deleted The testcase does a lookup on the key only +probably better fit into JBDS since JBT isnt really for adding rd party bundles +All prerequisites updated to and up to date repository profile added to +Hi Felixwow many thanks for handling this issue +Brian is missing from the patch +Either we integrate the class which works perfectly into tree so no additional jar is needed +addressed comments by Devaraj +running tests +Can you post an rb +I did say was also failing with this but today it seems fine and i cant get it to fail again or see any rogue dependencies that might have caused the is working fine now +So it should be will port the code to ehlo and add a patch here +It wasnt swapping and the CPU usage was pretty example I provided was oversimplified +Still stuck to early Seam CR builds +Option is the logical solution since the expectation is is not producing any output but there is still an expectation of cleaning up after committing +The postgres team was very helpful and theres a much better implementation than the original one checked into CVS +test would need to be written in the esb section +Going I think would break a lot of tests since they set small block sizes and would be an incompatible changeFor the max number of blocks per file I chose million +However I dont understand why this is useful and it is not a regression +New version of patch with code review changes +Is it know which JIRA issues caused the extra code to be written they must have the use cases needed to test +Archives should stay but list shouldnt be mailable anymoreIf that equals remove then go ahead +This patch requires patch +Thanks +This is being fixed as part of the problem is that Apache snapshots repo is not being declared and the build does not find the the JAR isn the release repoIn the mean time if you build from trunk things work because is part of the build and maven reactor uses it from the built itself +unclear to me if this was a regression in or never worked will check it built with a clean repository and cant reproduce this issue +This is a newer version +Thanks LangI applied your patch to the imageio module at repo revision rI merged the test case into the class created for and updated the manifest with a new module dependencyPlease check it was applied as you expected +You were absolutely right +The fix is a bit trickier +If I can get something consistent Ill run in DEBUG and if it still triggers Ill upload the logs +Its been two days and these updates have not made it to the central repo +Emmanuel please add the URL for the relevant JIRA issue +Thanks for this contribution I was thinking about implementing something like this myself before +to not reproduce in +Attachment has been added with description Beforeafter diff u of ve also attached in case there are problems with my diff file +I will need to fix it as its blocks some MTTR tests +Updated the patch +Maybe we should discuss this problem at the dev list and hopefully somebody will have a better idea how to deal with this +Part of the problem is that Java doesnt give us access to the raw fd values but another part is that can wrap an IO around streams that dont have an IO associated with some cases we can dig out the real FD but it involves code that secured environments might not permit +Possible patch +will do thanks +Try setting a time to live on the messages +It is reproducible there +The fullYear needs to be set before the date in the change handler +Support of AJP is postponed. +The jars Im using are logj and the jar containing my test code +By the way client got no exceptions +Unit version compatible with RC +It is private and used only in get +July I would suggest to commit as is +Weve started to extend the review workflow in Sonar but of course the ultimate goal is to let a Sonar administrator to define his own workflow to be able to implement such requirement +Sigh this is because the python version started using the web version of IPC instead of sockets +Committed to trunk. +extract a server helps with this +I like the lower bound cutoff condition + +Does this still apply +Also ran into this +We spent many hours doing post mortem analysis and only after we ruled out lots and lots of corner case possibilities did we come to find out that the users application had set the write flag false in a small number of their puts +Marking this as out of date since we have moved on to AS and have a separate documentation site for it. +If not I suggest we close it cannot reproduce +Now resolving this on the basis of tests via on +Thanks Arun + +I have observed the same problem +Hi FreddyBut is already the default value under Settings CheckstyleThe problem still persistsKindly addressThanks +I would say that should be happening just once when constructing the final ue of ants chmod task might provide simple codebase +I tried turning off watching topic advisories but I got another problem Temporary destinations didnt work after that +Its a matter of the good news is that the Sonar plugin remains compatible with Sonar tested +The patch provides support for Antlr Grammar Inheritance +all issues to the CR +Moving this over to a release +The replacement order for the return created is in Completed status +This is a major headache for me +IIRC it adds this functionality to hbck and to the offline meta rebuilder +OFFSET clauses so the patch uses HSQLDB which has a license +ive marked it as patch so it gets noticed +solution see wait until the is available +a and c have already been implemented in +In quite a few patterns its not unreasonable to put your secured code in a protected abstract method implementation +Is that code in the Hue github repos? Having support for running Pig jobs in Hue would be a big win +Might be easier might not +The last dependency was in not in +However I previously made the same changes and my changes were similar to the ones Adrian made in so they should be correct +Preferences Dialog under Mac OSthumbnail +is a sample PDF created with and the attached diff to sources +iBATIS doesnt parse SQL never has never will +Domain specific debugging to make diagnosis easier +How can I do this +yes Id identified that problem too so Im continuing on will definitely take that into account thanks for the reminder +I have looked through the latest patch and I see that Sureshs comments are addressed +Right is only really cost effective when the number of Tokens that are teed is roughly less than according to my basic tests +I just downloaded the nightly ant build to test with it +New version of the adding an entryto NOTICE + is the default welcome screen in eclipse +bulk close of all resolved issues as part of closing items. +Of course I will review your patch ASAP +Can you submit the patch to the JIRA and grant it with Apache licenseSo I can apply it into trunk +Patch looks good to me +Thanks +Patch for dealing with is version of thx to Zdenek Sochor for providing this useful patch +No response assuming ok. +Attached example for earlier version not to be committed +affects version? shouldnt it be fix version +So I figured I should document how I fixed the problem +As Arun put it we dont want to introduce a generalized solution when the issue here is a very specific Jetty bug the generalized solution might have other ill effects that would be hard to pin down making the change hard to verifySo instead the approach we will take is to apply a very specific fix for this very specific Jetty issue start a thread inside the TT which monitors for a spinning Jetty selector thread and if detected shut down the TT +Just some Jeff liked The idea of yellow mahout and since i hated a blue elephant just trying a more suitable shade for the lukas only two primary color +I think instead of we could change API so that instead of receiving reader it receives a struct that has parent reader sub reader ord of that subso I changed the Weightscorer API to use a class called that holds the parent reader the current sub and the subs ord +Resolving the issue as duplicate of +fix in rNow runjava command preloads the Java main class by default +DUP of. +Dan First of all thanks for looking into this +Would probably end up embedding that in html + +We have hopes of making use of off the shelf formal SLD documents + +Reporters of these issues please verify the fixes and help close these fine now +anything else we can do +Youre right +OTOH perhaps a select style for would be appropriate +Sure +Patch for trunk +Not resolved in time for +The idea is to make the resource bundles available in the object just like the other is available in the do closure so we can add all available resource bundles to the messageSource bean. +Closing issue out as part of JIRA cleanup +I have added an extra check that may or may not fix it +The real issue here is the logging message that gets printed out for sub deployments +First draft +I like the idea of adding test support for mac and win to the parent pom and disabling it in the xulrunner root pom +I have to move the s to the region server level +That would cause JRUBYOPTS contents to not be processed using that same logic +The class I used to validate the xml file against the did not uploaded the lib dir to keep a small size +Oleg +Integrated patch v to and trunkThanks for the patch ChunhuiThanks for the review Ram Lars Stack and Ming. +In one corpus doc +Then I removed some arraycopy around that was superflous +tmp file and copying the previously aggregated logs into new +I guess we should also document the CAST possibilities I am not sure of the restrictions here off the top of my head Kristian is our local lob maven +So the result is not for char +build comments match those in +Doesnt make the cut given the set of things we have to deal with +Committed to. +I dont understand because when I committed the fix the latest daily snapshot was instead I checked for the incovenience +The patch affects a lot of tests and setting timeout for all of them shouldnt be done as part of this patch +false alarm still not working +obviously wasnt looking hard in revision . +for the patch Israel! Pushed to trunk and branches. +To be clear Im not which I think is perfectly fine for managing plugins +i dont think that should lead to a problem right +Fixed in intentionally use erasure to implement co of function type +the overflow channel could itself overflow into another channel +Ive attached the patch +Patrick Crocker +Im guessing the issue is the improperly escaped m +The rest of the patch includes a refactoring of the unit tests +Change the sentence to something like The ALL and DISTINCT keywords determine whether duplicates are eliminated from the result +DuhAdded more to the test +Script to benchmark this optimization +Can this be closed now +I have resolved this by adding an extra check in the ESBjBPM integration code to explicitly check for terminated tokens +Okay the JMX initialization logic was a bit more elaborate and thus decided to keep it on master only to keep existing branches stable as is. +Patch that removes all refs of the said property in MapReduce +fixed in subversion actual fix is r +I signed up for the User list as ben and Im not able to send to the am using +The Cipher interface has been renamed to in the release to reflect its stateless nature +forgot the grant ASF license reattach it +Latest patch also on +i thought i had fixed it yet +CommittedDIGY +Actually the real problem here is that scaffolding cannot let you specify a null value for a constraintssize +There is only a small window when the error is not detected after typing import and before typing any useful code +err +Thank you for the review! I havent decided on default values yet I am only working on the general structure or hierarchy. +Made the suggested fix and added another example +Fix made by Jack checked in to SVN +Then the next loading of got an error +But I have not found anything obvious or anything that stands out so far +Ill commit this soon unless someone objects +All tests pass and it is back compatible +Or to be exact I was right +fix with javadoc comments +I think this will fix it +Im running on BTWLet me try and dig deeper and figure out the cause +Ive upgraded ant to and put the Xalan jar into the the command line it seems to build ok and produce data in the report dir +A new patch for and this time meeting the new specifications by +Attached a new patch I believe addresses Sureshs comments +Buttonicon arent applyed but all bugs were fixed before the latest build and all work correctly in trunk for now +Forgot to mention this patch both introduces the new methods and per deprecates the ones that dont take a path to help people transition from Hadoop +Ive produced a color version with PMS colors for use by the screen printer +I am also having problems getting the with putpost on rowscolumns workingI keep getting a return code of and the row is not inserted +I like the last image +Moved RPCADDRESSCONFKEYS to to trunk and +all issues to the CR +Im really sorry for this but it seems I cant find time to have a look at how git github with eclipse svn works so I decided to at least upload the fixes I mentioned in November +I actually started working on this last week +Thats sounds really intresstring +Sounds good +Apparently this is not working as expected and the test test case does not accurately verify the functionality. +Thanks Chinna found a typoshould be + +Now the will use the classloader of cls when it cant create the proxy with the thread context classloader +Ive applied the patch to CocoonAs this is a new functionality I will apply to Cocoon branch only on user request as branch is concidered a bug fixing branch. +settingsto ignore committed in revision +I repeated running the tests from the root of the repository and also the individual child modules +Is everything else done +Fantastic thanks for this ability +It contains changes to the logic where iterating over the nodes and the query +Both cases can probably be fixed in the same manner but just wanted to point out all scenarios which I can think of that can cause this problem +The NPE issue on linux shown in previous comment is observed in trunk. +Looking through our manual were kind of being a bit goofy on this as BOOLEAN doesnt exist in so maybe BIT should be left or you can somehow work in a comment in your docs about the issue +two points to make +This issue duplicates +Tomcat since doesnt allow symbolic links inside your webapplicationDo you have symbolic links in your webapplication +Youd need a Isa proxy totest it +There are two of them however ant test runs only one +did you test the nightly buildthis component is in change +We plan on running Asokans patch through a machine gridmix run soon hopefully today +Resolving as Not A Problem. +If its not fixed could you send a test case? thanks + created for removing from still need a way to differentiate between timestamp and increment counter in +Merged into trunk +I dont see a listOrganismsoperation in that wsdl +A more robust solution would be to address the entitlement mapping code however we do not have the cycles available currently +Meaning traditionally a query is sent to N servers whose responses are merged and X results are returned +But since super columns dont exist in CQL and since thrift dont allow to define that applies to only of composites the patch just always use the else branch that does the right thingbq +Using methodPOST to submit the form instead avoids creating the URL and solves the problem +Thanks for updating patch Dmytro +I had the same issue so I took a shortcut I skipped the conversion and referenced directly svg in the htmluploaded new version +Sure it will + MAPREDUCE to +XHTML requires that javascript that uses and other xml characters be contained in a CDATA section +Thanks Joerg for enlighten me +The logic is functioning correctly in the pushdown allowed case in which its entirely expected for you to see the exception that you are getting from h +I am pretty sure that above mentioned change in exception handling is causing above change was done on +These tests dont fail with so is specific +Hi AaronI hope you will be able to post the files todayRegards patch as it exists now +This file causes problem when deployed to and works ok on This file does not scope database drivers and works ok on version of jboss +The initialize method is where they are getting created at the moment +Does this work for you +This would help everyone trying to follow Solr development better understand what is going on and why +Revision . +Marking as closed. +Thanks anyway! +revised patch added +Still need to determine why the tck tests are failing but they are commented out for now +Dont like the solution +Attachment has been added with description Set of test for leaders of type dots +I can take care of subversion but wanted to discuss as weve not removed avalon from there yet +Can you please open a YARN jira for this +However in support of this example a note has been added which points the reader to the online version of the Tree Cache Guide that contains a more detailed look into the topic +It has the concept of message chunks with fixed boundaries +These are easy enough to fix +Ive just tried to update the website with a commit but the updates dont show up on the website is there something still holding back the transition from completing +Thx +Sysadmins updated JRE on Saturday but forgot to restart JIRA which lead to missing timezone files on filesystem in location of old JRE where running old JIRA JVM looked for them +seems there are some problem in Hudson +Its not possible to use a generic JDBC driver to connect to Drill +Unfortunately i disagree on this one +linking to the original bloom filter jira in case implementer wants to inspect patch for code +Do you have done more work on this provider +This patch made the changes suggested above +Look fine +we are using eclipse are deploy the war file in the unix box for are deploy the war i file in the Windows box for running the application in local +bulk defer to defer to +this file is patch for combobox in Shah can you provide a patch fileThat make things MUCH easierThanks +Im currently on travel and offline but I will start hacking on a test case Thursday morning +Move it back in if you think differently. +Its up to Michael when he wants to commit this code its not critical for Seam +Hi Matthias Im not sure that I really support this change +jclouds could break this test up slightly to only bother checking the hostname command if the field is set +Keith could you review this change please +So now that we have we could consider this what would setting the clipping area based on the extent do? +This is a feature of eclipse and has been discussed here a few times +Oops +I wanted to go ahead and introduce the change requested introduce an init parameter for If specified servlet init parameter will take preference else it will fall back on the system property as it currently uses sees any issues +Manually tested +sample believe we actually have this fixed via the patches associated with +Ive committed this +Good suggestion Luca this is a good way to save some memory +Test failure couldnt be reproduced on Mac +Information attached to the deployment for later processingby the deployers should be in the also issue is cannot be closed until the test see my forum post +Ideally there would be separate source directories for global assets versus supplementary HTMLimages +btw there will be a lot more issues as i didnt yet fix all the issues i foundduring testing +Thanks for the report +Depends on respin of the staging version of the discovery plugin to point to staging site instead of GA respun the disco plugins by hand so this is working now. +I recommend not trying to configure this as I dont really see any added value +Please add specific items that may have been missed +It is listed as unresolved in the list of issues but marked fixedclosed +I filed to track the issue. +Moved fix version to in preparation for release. +Thanks Jakub +These pages are handled by Product Management or Marketing +This shows it is struggling and timing out with the funcclause rule +Can you just attach the exact EAR youre using and tell me how you set up the database so that doesnt cause problems? Also if you can attach the full stack trace that would help +The Downloads page is now only for official releases +I tested the latest patch +Did you mean to say the customer must specify a replyDestination? If so then yes I agree this could be an interim workaround +seem like a very nice way to create strings and appear to work as strings in all other cases +let me take a look + +It instead stays open until apparently garbage collection occurs + it sounds like youre complaining that the server uses single quotes to denote the input it couldnt parse +You can use these to build independent tarballs then created a single unified tarball using the script attached to this issue +Really though this shared NFS mount destroys the utility of this toolkit for me +I created for that +Good stuff! Ive made the fix and also fixed us from assuming all content are +Hey Ted sure thing + +I tend to agree with Satheesh If we could figure out the right advice to give the user then maybe we should just make Derby default to doing the right thing in the first place +You are right that webhdfs currently does not work with multiple namenodes +Is the lost connection what is about +Lukas please assess and schedule to M this is simple changebut tests needs to be written. +Safe mode is still there +Resolving. +Checking on flush seems as good a time as any to look for split +Committed to trunk and Alejandro the patch seems causing some problems +Thanks Campbell +Marco Im having a hard time picturing you panel issue +The internal implementation and the actual working will come with a follow up s an important discussion no doubt but I wonder if it matters hereAgree +Are you running simpleaxisserver with a given sample? What are your valgrind command line options used +Here is the patch for this issueThanks to Rishi Solanki for guidance +Allbeiit were not full time on the issue due to work for but we hope to have the problemm resolved soon +But breaking down the tasks of each page +The conditions where only noted in the pack objects but not actually read in and put into the packconditions list when initializing the Rules note that every pack has to have an id attribute to work with conditions +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +If it is in jruby we can do so interactively +Feedback is welcomeBR Maruan +Its adequate for many applications +Finally I dont see why find would need anconfiguredcheck since those are only used on methods that modify storage structures +forge +Applied patch in revision into branch in revision +Marking Patch Available to get a test run +looks good pending to mention Colin tried to add a test but ran into Ill ping those jiras +Hi Rohini the bug you discovered is not specific +Perhaps this is not there anymore in the latest codebase +Thanks Francis for explaining +This issue has been resolved for over a year with no further movement +It seems the is flag is not correctly initialized +Looks like needs to be true for this to fail +If you call truncate while a compaction is currently going on it hangs both the truncate and the parallel compaction iterator +Please clarify meaning of previous comment +Updated patch to fix commitWithin parameter in and failing unit test +This is really minor as richfaces and seam are not targeting to be in any inheritance of another product +And it really looks like the test itself is the broken part running the same code on MRI breaks as well +Thanks +Raising the priority as it is important to fix the issue here for release +libdeltacloudhelpersin sendusrsharebin +If you would do this this issue wont exist +Was able to clone see remote branches and whole history +This certainly looks like a spec violation to me +Also this should probably go into +Since is always called with adopt set to true lets remove the newly added parameter adopt +Works like a charmCheersand have a good evening SaschaYour patch is commited at r +Remove from fixed versions this tag seemed to be set mistakenly +Patch committed given Alans . +HI Vital Trizna youre most welcome +I dont see why it doesnt work from continuum because continuum start the process from the command line like a user BUT Continuum is an old version and weve done lot of improvment to run external processes since you try the latest beta +Make sure that you can run the target in ant successfullyThanks +More to improve the work what can I catch the points to make this deamon? If I do I want to get the best choice by taking other opinionsI want to get your good opinions +It should be like is API in thatWhat do you say +Editing with the Ivy editor did the trick my Eclipse was defaulting to editing using the old xml editor +Test passed +IMOthis is a bug in Word +has moved and the changes to and do not apply cleanly +I guess I can. +the perf issues here are really from our contrived tests +you know with CoresThreads and allIm pretty sure that ATS would deliver some performance on those machines +Something definitely broke in a recent commit +Fixed by generating default deserialization constructor +However the Manifest for the second jar did not include in the Classpath entry +tests pass for me +I know this was a problem in older versions of but has since been fixed +It doesnt work +Please reviewKathey +A new version has replaced it and it works againSo much for TDD. +Please review and add missing parts +An initial untested implementation of the +Test source from the bug is the patch +Screenshots of rendered source code for ecommercecontrolmainThe DOCTYPE declaration was added to so that we can be sure it is the first element in the page +Could somebody who has enviroment where it is failing with JBDS verify this Jira please +Hello VikasHere is my updated patch with your suggestionsHere i have used cancel service because after canceling the order I need to send order cancellation notification and this service is doing it +This can be done in a fashion +In the event the version cannot be resolved the users is asked if they wish to continue with project creation +It should use +Rob I am assuming the one you corrected is the only correction required +backported to and +New patch removed useless dependency +For we can at least remove that file prefix from the example in the reference manual +these commits breaks the treeListbox for +i think the limit should be fixed in release and we can open a new jira for getchildren api +Chaining is there all closing all resolved issue that have not gotten any more feedback or comment in the last month +Fix applied to please to because the bug was already partially commited there. +this is happening now could take hours +Committed fix ajax redirect case +I try presentation another computer the page controll is successfullThe next computer Ubuntu Mozilla +Changes now applied to branch are no specific unit tests +In this case the intention of the exception is different +but the transaction will be rolled back +can you try a later version say version this version is known to work ok + +Will do with the final build so I will keep this jira open +yes you cant instantiate a object without a it is mandatory +Still need to figure out the Javascript for submitting the form etc +Josema some of the code snippets are making the page really wideCare to make any adjustments for ? Thanks +Good after commitThanks anita! +Thanks Harish! +Actually Chris all the flags can be set via exportsSee man erl and look for ENVIRONMENT VARIABLES at the bottomERLAFLAGS or ERLZFLAGS are what you want +change the comment from cleanup history to cleanup local conf +If anyone disagrees with this let me know and I will roll back instead +Need to get with Prod Ops to give him was notified of the issue but I have not received a response +says done pending review. +time to fix the test case and get it running +One thing that struck me is that the patch adds quite a bit of implementation details in the for a public JDBC method +Attached patch to add warning on works for now +Test whitespace covers this case +I had a look at the code changes and they look good to me +Keep those patches coming +Charles asked me to look into how long shutdown takes for the war so we can decide if we should look into optimizing that in addition to startup +This patch was applied of and is available in the latest and published +thanks Ivan. +I just committed this to trunk and +The formauth just extended servlet and switching its parent class to a sling servlet caused the servlet to be found again +Thanks for posting your solution Matthew +Hello All Here is patch for event to create account in SFA +This is a duplicate of +Adds another optimization + Properly throw timeouts from decrement the count of waiters on timeout fix in taskCount and test all of it +If we really need a config file we can just modify the start up script to read a file and set up JVM property on start up +Please verify +Ok Ill try to reproduce it with Tomcat and Apache HTTP +Thanks for picking up this issue any update on it +I checked in a couple changes last night the inputStream was getting nulled out you +on to as well. +This issue is resolved by +Created for the improvements +Tomorrow Ill try to pass more junit tests +Not related to this issue +This is the patch and images for the new backend admin theme to match the new frontend look and feel +I cant remember my use case now +Mine was still rough using the old property and readString methods rather than defining a new property and method as you didDoes this look reasonable? If so Ill commit it +Test failure appears to be looks like both ATMs comment and mine have been addressed in the latest patch +Summarized all previous comments in description +Reopening for handling +The issue had to do with the parent pom containing configuration for the sourceDirectory srcmain but some of the modules had their sources located in srcmainjava +Let me add that and update the patch here +Hi MarkRe your comment about Maven duly notedIf I end up using this library it will be a required dependency +the task is moved to the next s necessary to translate the description of validator component provided by dev team and to add it to the guide +It would be appreciated if vm guys can help to review this patch +We will make it available on ASAP making it completely transparent to users +Looking into the PR +Wasnt there a similar problem with another test reported not so long ago +Small incremental progress and then we can file further improvements later! +Committed revision +Does the focusstack thing makes any sense to you? Seems that is directly related to the bug +This also moves the logic of logging the Problems into the Problems interfaceimplementationsThis does not fix the underlying problem but merely adds logging so that we can hopefully diagnose the underlying cause +m new to your build process so if Keith pushed this does it mean the fix will be in the nightly buildAlso how can I track when this fix also gets migrated into the Switchyard Eclipse tooling +Apache Camel will be part of SMX +Hi WestonThanks for all your efforts to make it happenWhat we need now is very simple +The JCS logging is set such that you must run at info or hihger in prod +Works issue but when trying to sort the list of dashboards +added method where we also pass in the parameter Map so we can do the expansion for placeholders +there is no remote way to access them afaik since its all right +Thanks Mike for committing this change +Please review +bq +Patch that solves the issue +test logs its the number of marshalledunmarshalled classes you have the number of media types +I have seen the did not start error before while working on for BYONThe problem occurred with scripts with a very short execution time +I could not reproduce the issueHowever I added a null pointer check +It is checked into all trunk and +Weve already checked that size so position is the only possible place for it to be +In Section I inserted a combination of the SOA service instructions from the GettingStartedGuide and the extra information contained in this JIRAMade some minor textual changes to Test your Installation and Appendix A Disabling Authentication +bulk defer of featurestaskswishes from to +Resolved for a while +I guess we have git cherrypick or may be Jeff simply merges the same patch to the trunk and then we can only record a backmerge in SVN properties. +Reattach to wake up hudson with the right patch file +Ah right okay + +J sorted the signing routine to handle the zips within zips structure ready for whenever brms is going to be signed. +has been replaced with fixme for +I then override the find to first check the list and if not found use the +Similar problem to inserts but for unionsTreating union like and performing set at the added a couple of tests +I played a bit with EOL preferences and now I cannot reproduce it no matter what style is selected Windows or UNIX +This might require some changes in Arquillian Drone core. +Per phone triage with Burr and Max on +Yes I got it after this time +Dns configured website up +It even has a patch +I am attaching an updated the which will create the based on the endpoints schema +beg you not to move it out into a separate action profile whatever + and my apologies. +Needed to up FD limit again +Then we will have the option of using the parsed of tika or parsed simpler than now exists +OK +Mubarak I will take a look at the issue +Thanks PaulexThe fix looks good +Did not mean for that change to go in just wanted it as a stumbling block to thinking about good sort field data +Its happening on t look like it actually got solved +New Spring Data test suite class with corresponding Spring Data context configuration file reliably reproducing the error +the trickier partwill be the tools that use Velocity logging and the + +I think we should change also to use this but we can scope it into a separate ticket depending on this patchs sizeWhat do others think +verified that customized entity names work now +thanks +The Mevenide plugin doesnt seem to be able to interepret this custom property +But Ill wait to check it in a couple of hours to give anyone else time to comment if they want to +Odds are slim + +Thanks to Dyre for the workflow reminder today. +Patch for this issue attached +Reporters of these issues please verify the fixes and help close these. +Sorry about the spam +Have you discussed your issue in the forums first +Resolving as being fixed in +There are a coupla features Id like to see from maven to make use of the CMSmore palatable for large sites like +Is there going to be a patch release axis in the future including such fixesBTW seems to be another duplicate of this issue +Thanks Mickael for this link to this Tycho issue. +Would disabling this capability by default answer the security issue? Or is it still too much of a back door? Thats the central question IMO and Ill gladly defer to your expertise in this areaWhether a would be a better way to address the possibility is certainly a valid point and one I entertained early on +We should at least implement the subtask and maybe that will provide information in the field why this is happening +Nice find! If you pass all unit tests just ship it +I suspect error here where you configured a wrong DS name that is not available but VDB got flipped to active thus you could make connection and issue a query +The tests will not fail instead print a warning message +Fix with importFrom with a class realm. +Both nl and de applied to +This patch copies too much as the samples and the directory should not always be copied +In patch v the is created inI think it should belong to region server +RickThanks for the quick responseI will test and let you know soonYou dont have permission to access rickmcguirejavamail on this serverTrying to download the snapshot +lzma format is a legacy format superseded by the +Updated patch with Hairongs comments Removed the ability to specify a duration in script file +Archetypes in central are working properly +How about adding a and putting there +Jakub I just figured that there was a small hickup during deployment of release + committed to trunk +Could you try changing the code to create a new object every time and then run your load test again? We original did that but it was slower at least according to this commit comment from two years agoDont objects use static ones instead +In fact for Cloudera it would seem to be smarter to specifically make that happen at RPM install time rather than burden the Hadoop base with a dir that will ultimately need special permissions +I think it makes sense to add the header to but Im not sure if the +The test case class and corresponding application context file in Eclipse patch format created against HEAD +Repro steperror message same as above +Also please include the full stacktrace +It uses x from a couple days ago +On bit linuxwindows machines this problem never occurs. +The only reason we actually have an is so we can capture member structural information for the benefit of other and in the process it makes it easier to emit +Sample maven project that replicates the issue +The interface would be generated based on the properties defined in the entityUnless Im mistaken theres one pass to generate super and subclasses for the server and another pass to generate for super and subclasses for the client +The following requested updates are ready for your review changes requested by Milosz changes requested by Jeremy changes requested by for Donald under revision Committed for Donald under revision The issue regarding the use of serial will be resolved under another JIRA +nago as mentioned emailing user is the correct forum for user questions. +indeed did not mean to violate etiquette when I asked if I could submit a patch +We saw these issues on the build machines a while back and Giri fixed it by fixing some filedirectory permissions +I will review the patch and post comments soon +I have been recentlygoing over the relevant while refining multipart related classes +Sure Matei what you suggest sounds really simple and Im testing it now +Apparently I did not find that ticket +linked to the geotools arcsde plugin as fixed as per the related geotools issue +Please find the attached patch which will take care of the issues you have pointed +Many thanks to both of you for working on this issue. +works but its annoying to have to declare subtybes +Will be fixed in jPDL . +Thanks Sandy +Changes to so it doesnt crash +I have the same problem +If you test this code on Yahoos clusters Ill appreciate your help +Discussions should take place in the forum and this issue is incorrect +It can also run over Equinox +integration tests are now running and passing on +To resolve this bug you should install +Is this still valid +Downgraded the priority level from major to minor until we revisit this issue +This seems to be some remnants of EAP processes +New patch moves namespace to and changes indent to match as discussed +Integrated to branchThanks for the patch JeffreyThanks for the reviews Stack Lars and +provides a working solutionSet to resolved. +It doesnt appear to break anythingUnresolved questions Im not sure how best to test the new functionality +I added Serializable to the class declaration +This one is ready to be committed +Forcing orphans to on the outermost block would definitely create the situation that is described in the comment +RFC Implementations MUST be prepared that the same short name might be used in a subschema to refer to the different kinds of schema elements +I believe that this issue is present under all versions and against Spring +rewrote description for data +Sorry please do not use email tomkralidis +I just wanted to point out that this bug seems to happen whether is set to or notThe current release that is available on the site is not very usable because of this bugThe current trunk does fix this problem though which is great +Thanks Vladimir the patch was applied to BEANS module at r +They successfully pass on Wiindows XP in opt mode as well as in jet and int +Fixed by allowing the is active test in to wait for the initial advisory message +This is really a duplicate of +Luis has been working on a raw PNG Image Loader that addresses a couple of performance issues +Im not quite sure what that means but its doubtful my class is container managed +Reopening since has been reopened +Sounds good to me +Attached a partial patch for UI requirements +the idea is that you should be able to deploy rpcpojo using just the min distro +I committed the new QR decomposition +This issue is a duplicate of closing as such. +I provided a quick and dirty patch that did the the same for restconfig in +Thanks Ning and Carl! +Fix doesnt seem to have broken anything else so closing this issue +in JBDS +It doesnt seem to no +yep I was thinking recommendation +I strongly advice to wait till or use the nightlies +Thus having a method interceptor run at INFO level is excessive +the patch looks good to me. +Sounds good +The test above was using a ms wait but I went up to ms to no avail +Sanne what if when the merge was done the services files were not included +Added sample for nested variant but doesnt work due to samples which doesnt work due to Need to cover these by tests +Please check if it works with this project +I think should list all by default and perhaps take an argument to list blacklistedactive ones +In general Id rather have threads interrupt other threads rather than set a volatile boolean flag but I havent looked at your patch in the real context just standalone +Wayne said that this fix should be backported to branch as well +Complete +This regression was introduced when was fixed +I honestly have no idea what that code is about. +A Spring Rich Client for displaying a generated usage notes including application context configuration snippet +No new tests neededI have committed this +CommittedSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaSending javaengineorgapachederbyiapireferenceTransmitting file dataCommitted revision +Is there a known issue opened for the Hadoop race condition +I dont know if you have those fixes in the build you are testing with though because I have been forced to build from upstream vs +Second version of the documentation +I have tried your option +This issue was a partial fix for however still fails to store the object graph now because of bad order of updates and inserts of parents and children +Moving to and rolling back to these as out of date as we will soon be looking at integrating a new version of wise. +form +vs +We can adjust as needed +We can change this when we adapt the method +is addressed testhassequence failure after applied the patch it could pass either standalone and batch run. +While were at it would you mind adding the config option to? Now that its mentioned by the release notes I think its worth putting in there with an explanation that it should be overridden if not on the PATH +Im a third year student and Im looking for a project to contribute in terms of I have already learned a basics about bloom filters from the internet and usefull links from this thread +Ill just verifyvalidate that the port be an integer when the user enters data for that field +Ill know tomorrowI already logged a patch for this inSince the summary line is so small Im trying to keep each bug separatePatch applied thanksColm. +Are you saying that the portlets mocked servlet request does not itself +What you want is the hub information who is the owner of a topic +Hi Please let me know the status So that i can search for an alternative +Arrg! I should have known better than to change something at the last minuteThe latter attachement is the correct one +Issue fixed but may want to raise new issue about using values in framework. +Test now passesI need to change the patch instructions and get those two commit messages ll do that on Monday does that fit with the CP releaseAFAIK it should be released the nd Tuesday of the problem if you do not change the source for branch because I will tag it after the testsuite is finished +Hopefully tonight +Do we care +Wont get to this for. +stack JSP page seems to be invoked directly there despite being intended as a view behind a controller +metadata subdirectory of your Eclipse workspace and the log file of the Geronimo server when this error occurs? Thanks your memory parameters seem fine +Verified and closing issue. +This would basically be the same as the wrapped script just with pid files managed by startstop wrapper +we need a whole team of people just to handle bug accounting these days +HiThe patch is verified +The common way that people ship scripts for streaming is using the file argument +I think were good here at this point +I could reproduce the bug Metamer Mojarra AS Runtime Environment LinuxChrome Linux that creates issues with scrolling +I should have used +Refactored the code added xdoc more extensive testing +classpath file to include missing libs +If this is in error please just reassign as you see fit +Submit patch triggered Hadoop QA to apply the patch to trunk and hence all the s +Camel deployed as plain war files on is not doableI have only tested on Windows XP platform +If possible could you please provide a patch to fix the translation +properties files to change text strings? If I show you a patch to add tooltips to one element would you be able to address the others +lgtm +Fixedthanksdims +But surely beta was a lot weaker so could you try a nightly +Committed revision +Thanks for your patch Steven Ive applied it in revision +The current behaviour is the one that is expected by this featureSharing a filter that keeps the favourites of the one who initially shared the filter would be a new feature +b +should give us this for free +In latest SAAJ impl we dont have this problem So it is fixedthx Azeez +Yeah +More investigation is being done in the upstream RHQ Bug and the findings should be updated here when available +Some base functionality merged into master. +the above run timed out but it seems to pass here reliably +Planning to cut a release of Enunciate +I made a small modification in the assertion to better understand the output at . +Sigh I cant read properly +We should have reverted this for +now chance of getting this committed +bq +Ran to latest ran in QA +This has been done in the bramley branch and should be in. +This is all we can do w current architecture +the patch only one truly required +I did a release of htrace to that just has the dependency scope change for junit and no other changes +Which does lead to potential problems with duplicate columns for those who are just aggregatingWhile clients can aggregate on their end the is no convenient mechanism to run a tables configured aggregator against a mutation +It would contain a few comment lines with instruction on how to configure the JVM and one commented out line with an example configurationLet me know if you need a test pilot for the JRE detection +fixes are available in only checks the pk of the top object to be removed +hg folders after the IDE boots up or when you open a project +If messing with orientation in this way isnt supported would you happen to know of a way that I CAN allow rotation on Android tablet devices? Unfortunately its not as simple as it is with the iPad and iOS in. +In revision I updated the diagnose command to perform deeper analysis of some resolver problems + please committed. +It all or nothing +This is pretty critical one +This probably means that the blocking mechanism when memcache twice the limit we stop taking on updates no longer works +To fix the problem we should release and we make a decision on what our byte code level should be? I thought that we just agreed that our code should compile on JDK which does not necessarily mean that we target byte code +I used Weblogic to access ejbs in a POC some time ago +So this bug does not exist in and above +No no reload +At upgrade readTransactions had been set to null which caused lots of NPE at cpactf +Or we could use a Properties object with key value pairs by defaultnamex typey timestamp +Thank you MayankI assume this is not applicable for trunk. +Ralph would you be willing and able to provide me with a small test case +The registry deletion was being done by which doesnt take the flags that let you specifically choose between the hive or the one +We can alter a column defined as UNIQUE to null +zipOne test is missingHowever in the Surefire run its SKIPPEDJakube could you pls have look at that +Fixed in branch at revision +Thanks Ben +Thanks Wu I made a change in convertloggingcategory and now verbosegc works for me! Committed at +Theres still a specific MBW named returning the files length for getSize but as expected I cant find any call to this method in Jersey codeIll give it a try with a sample when I have some time to see if the is set somewhere else now +Reclassifying as a bug since this is actually blocking a desired use case for Pageflows +The difference in sizes is explained by +Attaching an interim patch for this issueThe is now unregistering itself from as follows Check for Autoloaded driver before DB shutdown Testing with Getting Registered Drivers Shutting Down Database Database shut down normally Check for Autoloaded driver after DB shutdown Getting Registered Drivers I appreciate your commentsreviewThanks looks good +Trace C represents a place in Remoting that slipped through my tests +It is tested on branches +Only bug fixes are down ported to the release branches not new features +Atul Sure +The tasktrackers uid would need sudo privledges in order to run chroot so we would want to run user tasks as a different uid but all user tasks could run as the same uid but each with a different root filesystem +So it works exactly the same way as the Download button in runtime detection but here one would not expect it + +It looks good to me +The server knows nothing about JMS producers so we wont be able to provide that information +Candidate for to the EWS docs lead +close it. +Actually I think we should do this right before code freeze for +If you want to use IDE get the bundle that is certified to work together +i had the labels in the application itself at first +RightThis work includes deserialization of all kinds of as well as Target and so onIt will take quite long time. +This patch fixes the problem +Now DN network comes back +This patch applies to both and trunks +I used the attached script to test this against a clusterThe script uses the fetchdt tool to create a delegation token and verifies that it works correctly through failover renewal and cancellation +What does u in uf stand for +Logged In YES useridYour last bug report said you are working with jboss b +The useof the Lang classes is not meant to be a replacement but just a behind thescenes fix until we can use the real dealIf we want to keep this functionality but without the other classes we couldjust move the handling into and delete Nestable andIm pretty flexible on this one +This is the fixed version of it +i could still see them listed on the web admin console even if the broker was restarted. +Please check +Dhruba Can you please have a look at the patch? Thanks +I think is a good name +So maybe move the links to the README and add a little context +The rest of the list is simply skipped +At this moment I wasnt able to find out what encoding has been used with the file so I wasnt able to reproduce back to Finnish characters +Artifacts promotion is not completedIn the TODO I have to update the Karaf website for both release news and add Cellar section send an announcement on dev and users mailing of course the release vote has passed +Thank you Benjamin and Jeorg +Closing issue out as part of JIRA cleanup +Ill update it +Thanks Sijie. +Thanks lets change +commit some change in r but headers were correct to me before +I guess it should simply be a Subject with a Principal with the unauthenticated identity name +Bulk move as +The refactoring of the via SPI is still missing stopping issue progress but keeping it open +Modified the to delay setting of the frozen flag in the superclass until just before the proxy is created +Please ignore my comment about toObject I made earlier +That is to say that you cant subclass one Enum with another Enum +thanks Tim +Frankly speeking upgrading our App to this version was bit tricky this version still referring to core spring framework will it be possible that this version start using the latest core spring framework +One point of confusion is why there is a section in srcmainresourcesorgspringframeworkrooaddongwt +workaround is to disable mails for ERROR state +Closing issue +Have you already looked at that code too +dataQueue lock and ackQueue lock are merged into oneIs this done to make the code simpler? Or was there some race condition that this fixes +The compatibility is very fragile in that can only read an if they have the class since the key and value classes are contained in +This way youll get both when issueing build clean +I forgot to mark this as In Progress +Patch committed +Nice one thats exactly what I had in mind +bq +with every parameter filled out with custom +If I understand Junit correctly the object should be created when the class file the loaded and destroyed when the jvm for the class exists +Hi Saschathank you for the great job you didI have reviewed your patch +Not sure how you would do B + +META +The message files had to be manually mergedrevisions merged were +this only messes up the buildfile even more +Hi manojThe WODEN API changed yesterday +Done that case was already covered +Summarizing the description Ask for tag attributes during tag insert dialog always appears while dragging to the source part regardless of the checkbox The dialog never shows up while dragging to the visual part regardless of the checkbox option tasks are Read the checkbox option while dragging to the source Restore the dialog showing for the visual part +Moving away from Singleton might have some other unintended consequences and I am not sure about that but it might be the right way to do it +Thanks for going through the patch +This is quite strange +The test case files contains a basic test case to check whether the object map can be partitioned among different slices and retrieved successfully in case of eager unidirectional join +The provided fix reduced number of problems significantly +Duh +I started thinking about a case where you have two pieces of information in the family +Since the stats are disabled by default I had to either go to webserver page to enable the container stats or set collectStatistics attribute via +Imagine you have a module A in your repository with a module descriptor you depend on it everything works fine youre happy and release your product depending on it +Will add a unit test in +As per argumenst above +Of course the advanced version of this just saves the openlayers state and lets them start up a map based on that +Migration of VR along with storage across the cluster is not supported yet +The issue is in when it tries to decide whether to render the join dereferenceEntitysee test +When working through this it might be worthwhile to look at to make sure these two issues dont share the same problem +Talluri pl review this again if you do not agree wiht the assessment +But the above comment stands +Ill iterate and see how far I get +Yes this is definitely a bug I will commit the fix tomorrow thanks for reporting itPS Since you are from the icefaces team I would love to have some feedback on what you could need for improvement +It doesnt matter whether you have the localhost entry in the configuration file or not +thats fine i marked to say it fixed in branch no regression because of I only changed the test codeD +Thanks for the latest patch Dag +In CRX we expose the cluster node id through an extra repository descriptor +This got delayed but I should be able to make progress on it again next week +Dont use the old maven zip +attaching patch here that was sent to the dev mailing listthis also addresses +Now in your latest patch you avoid two flushes happening in parallel if it comes through flush caseYes in our case there was an explicit split call also that happened which will do one more flushVery good analysis +Format date per Dicks specification. +can now be configured to submit a form when clicking on a pagination or sorting is the versionStill testing since +Let me try and look at this for the release +Logged In YES useridOk looking closer all that has changed is that the jsp page is displayed as though it was an html page but there is no server side processing so the source still available on the client +Also added a warning stating that proceeding may cause issues with model generation. +Tobyyes will do later this week +Orca design looks goodIs there a reason behind one fin being bigger than the other +Yes we know about this cyclic problem but this should only happen if you install the plugins not the base hibernate tools the logging is there something that does not work +otherwise it was too much dark in dark and not easy to see the difference between attribute and attributes valueIll go ahead and commit this soon +Committed for rev +And its still very hard to get everything right +That was was I suspected by my cursory investigation yes +Heres another try addressing the review comments +Now fixed. +No problem sir +Patch and code AshwinAs is feature complete and to keep scope to getting out of the door we will scheduled this new component for Camel is this working with network printers +Of course I should have just let the exception bubble up and fail the test directly rather than catching and failing +Another suggestion would be to do this by default if is configured +no I installed using install +I dont like having my application extend an class to expose this functionality +Ive created the Interface assertion class and added the assertion to the registration method in the +Its not byet committed because I was hoping at least one other person besides Drew could verify it builds properly +I pulled this into +Depending on the underlying executor implementation those threads may be created when work queue fills up and may timeout when unused +pushed +Committed. +That is equivalent to setting the default Lucene query operator at the boolean level but has no effect for terms that get split down at the analyzer level +Any chance you could addsome tests for this area first +I found something out The SLD contains two +I am not able to find class but there are two different classes and +Please keep your RBE preferences now o +of EWS with EAP is a very common use case for users who want to use apache with EAP +Still investigating +Its better to revert and release right now even if we release a when those special test failures ar fixedSorry Endi +Hi Joel I will try to patch it into and let you know if it DavidIts probably easiest to put two files in place rather then try to patch +The implementation has to assume that the bean is only properly initialized if any dependency introduced by setters is you have not a proper in your situation +Attaching AllAttached the proposed fix based on my analysis please help me to validate the proposal +Pushing this to for longer term resolution +Result of autoreconf i f v v v +Limited fix version to series +With patch v I ran unit tests from tip of Phoenix and found no regressionNo change in Phoenix code is needed +Fixed for verified with Felix console +And again for GA which will also end up in extras +I suppose we should mark this as fixed for unless theres still something youd like to add in the next few daysweeks +Vinay code should be done separately from code modifications unless they are a part of some fixes +New patch adds fixes to the schema that resolve the issues around the unit tests +This is a duplicate of and should be closed +Perhaps better to use +I have attached them also +Triaged July Assigned normal urgency +Scratch Jeff +With the most recent changes against this ticket including the bug fixes I just committed I am seeing merge operations get stuck during random walk test. +And the atomicity can be guaranteed at the server side +Alejandro My fear is that if we swerve off the route maven has mapped for us well end up in the maven ditch and well spend the rest of our lives patching and shimming keeping our build in shape +Defer to +According to Charles L the initial implementation of this was in the HQ drop +In order to accept such a change we need an encoding test that actually uses UTF and shows that the patch is did however make a change that caches the default encoder in issues should be addressed related to for pointing this resolved issues for already released versions +I am afraid this will have to wait until the completeredesign targeted for the release thoughOleg +I can only guess what you have done but you should make sure that for example if you change ports you do that in all files and that you set the param externaluserId in the SOAPREST API methodsPlease use the mailing list to further discuss your problems +The domain is unlocked and the auth code now sent to root at +Everyhing in OWB has been written as plugin +I still get the same since we have not seen a reproducible case +zip is created +do you want the equivalent of rsyncing a remote repo +The tests have been disabled in the common pom file +In each bundle MUST have a symbolic name because its the unique identifier +File to out to +I tried to replicate it after reboot and it works for me now. + +Please dont use inheritance for this +Closing issues which have been released. +Thanks Raghu! +Im most concerned with some of the timeouts printlns etc in +Observed the messages however I didnt see such a failure in soap messages +See comments on RBRenamed the issue because this couldshould be common to the regionserver and master right +Raising priority so it doesnt get missed for the release +By the wayIn the meantime a possible workaround in my own application is to detect when Derby is the connected database and try to enable a scrollable when creating the Statement but continue to open the simpler forwardonly variety for all other database types +forrest run open localhost select the All and then Whole Site HTML from the navigation +I believe the should work for +jacegisecuritycheck etc are named to match their corresponding equivalents in the servlet parameter has no servlet spec equivalent and only needs to be uniqe + pending Jenkins this issue as incompatible given the change from to +this is a good shutdown test but does not really match the original bug where a daemon thread was getting terminated inside a critical region causing the shutdown thread to hang +Ill dig a bit more +Hi Dmytro thanks for the patch +Ciao CharlesI forgot to report back to you +If we first update the implementation without stopping it it will try to resolve against the new API and fail because that has not yet been installed +I can change it for the aliases which are just a single token +These are designed for synchronous use thereby assuming that there would never be a message that was rolled back and resent later +Patch applied thanks for the fix Greg. +bq +Added a test case and created tests cases which passed +FintonOn what platform are you running the sample? RegardsUlhas can see that pom refer to apache repos +Thanks for the update +Scott you have had fun with this before could you verifyfix it +Thanks for the patch Raimund +Verified at +Closing resolved issues for already released versions +Please review +For m the goals could just be the lifecycle though +But I do think it would nicer the broker make them more consistent +Therefore this is a configuration issue rather than a problem with random writer. +We all know that POST messages usually have a message body but I dont think we can read from the spec that it MUST have oneIf in doubt might I suggest that a better solution could be to handle an entity enclosing request without message body by using a and a length of +Closing the bug for now +I have a working example application attached to the JIRA that uses a custom legacy version of + the WSSJ interceptors change the request content getting rid of security headers +another simpler idea you just handle this yourself in the Analyzer you pass to the thingThis is currently how Kuromoji works it has a stopfilter +v attached which doesnt fail to in one case +on Cyrils comment +But on a recent incident ti does not look like there are many blocks being written +Attaching think this is also part of a bigger picture with the Karaf shell +Any schema containing a type would have exhibited the described behaviour which didnt allow to resetclear the values of such a field in the generated class. +I just tested by handI added a line to aa bb fired up the example server and then executed the following queryhttplocalhostsolrselectqaadebugQuerytrueI then verified that the synonyms were in effect in general viahttplocalhostsolrselectqfooaaadebugQuerytrue +tested and verified +The patch looks good to me +marking the issue fixed in as the repro shows it was broken in and fixed in +The test is not the bugId say its both a bug in some unknown class and a bug in testAssuming you can instantiate random classes in a random order wo any side effects isnt a valid +Fixed in +Could it be because my default locale is French +Some thoughts it would be great if the classpath could be generated by the ivy files +Attaching a application in which it appears that this is working as designed +orig files on and Solaris are now fixed on CR build +small updateOmar Tazi doesnt need an account +Its kind of a messy fix but it at least keeps the dependencies in the correct just run into this issue and used Anders Lis workaround as its the best solution t this have been fixed already after years +Resolved. +Ho Andrei Im so ashamed and so happy +On second thought maybe we can add code to make it impossible to override the default namespace g in which case we would printer a warning to the console +Errata corrige please ignore my previous message I need to have a look to your patch more deeply +This should use the version of getLog taking a class not a String to avoid exactly this issue +Were suffering from this as well +Verified with JBT. +I changed it to and it seems works fine +Part of this issue is related to which relates to filters not executing for default controller actions +It may be the IDE Im using or somethingIll put together a BATCH file for windows and based on the release +I just hope others will not shed tears as well and we can make Solr Java compiled +execute APL actions based on APL would have access to the JBN inventory so it should be able to take the name of an existing script service and execute it +This is hard to guess readHowever I do think that deletingtrue is a bit to harsh maybe its best to force to configure this to true if the need itHowever only processing unread mail is what I would like to be default as I would not suspect consuming the same mail message again on next polling +How can we divide the work Erik? I can +Resolving obsolete issues. +But as I know currently a lot of source code still use ASCII encoding +Is this correct +Unfortunatly im not a Java developer +Associated this issue with the inception of the. +Issue resolved with the the fixed and resolved status for an issue is only meant to be set by developers after they have actually changed the code in svnWhen you attach a patch to an issue you just leave it as open. +Merged to the branch in revision . +I disagree that should make a HTTP call since it is unnecessary +NicholasIf it is too much of a pain to separate them that is OK +This could be related to +I am looking at this some more +The error appears when I use OPENJPAMYSQL and OPENJPAHINMEMORY not with HIBERNATEMYSQL +The only open Point is Port the indexer It does not use RDF but directly readsprocesses the DB dumps +Glad I could help! +Merged +Then I have added a new option for the called dist that is the distribution used to model the +is for the test +I applied the patch to the code base +Closing issues which have code changes and have not been modified for a while +After seeing thousands of these exceptions I got rid of them by always using the mvn urls and checking the locations of each installed bundle for duplicates before trying to install +Try this script +That would definitely screw things up when reading through it +I dont mind to spend more minutes to change the licence to the latest oneFor the question As covers many areas which I am not good at I think the translations can be improved persistently changes are inevitable +Agree that fixing antivy at this point is a waste of time and removing it is better +So how big is the performance difference between the two +testwsdl sample worksPlease send your if it is differnt +Roller log file with errors on resin +Until it happens it doesnt care what is being executed down the stack +pushed +Especially with HTML looming around the corner this is rather essential IMO +Ive verified mvn clean install. +Committed to trunk and branch +branches is now what trunk looked like at the end of the day on +Id suggest you try the user forum +Closed upon release of Hadoop. +The casts between BOOLEAN and CLOB are inherited from the old Cloudscape days +Also the getters and setters were already available and the interface is marked unstable +Add one additional test case +Thanks Eric! +I looked at your logs and could not see a database lock there I only saw a timeout to acquire a lock +Also with EL support there will be more flexibility for binding message content to method arguments in am marking this as resolved after the latest changes click on the Fisheye tab for this issue and explore the various commit logs for more detail +NPE isnt the clearest way to convey this obviously but seems more correct than silent acceptance +Works! +PatrickMy current approach is to checkin the and files +A spambot could easily exist to automatically log in to JIRA sites +There is also a readme with an explanation of what they do +Merged in branch +There is no point in executing additional statements +That is why I reverted the patch +There is a problem with parkunpark functionalityThis is a patch to fix the problem +Checked into the branch. +martin was added to infrastructure by mads +Implemented feature and tested manually using dnsmasq +I agree that its not our business to test the webconsole itself but when everybody votes on a release without noticing that something as basic as the config console is broken we have a problemThose tests could manual tests part of a release test procedure document on the wiki they dont necessarily need to be automated +totally agree here we will change all files with extension cfm and cfc are now used independ of the setting in ther railo admintnx for input +Cool +How do I fix it? is a transitive dependency of the tester project +is there anything in the error log when this happens +What is the motivation to put the unimplemented methods within the classes rather than in the base classes +Committed to and +This bug has been marked as a duplicate of +has been released with the appropriate dependency upgrades +No problem that means there is no bugTx for the quick response! +Committed +Committed in rev +A few commits later were left with only Listener classes I cannot get rid of easily because of serialization issues +From this information we will write the final release notes for all issues +In terms of actually treating +Have added and now returned by createXX methods on and correctly implemented the methods to throwAdding tests now +Thanks for the new patch +it looks like the mail server has troubles as well even the email address is valid +That would eliminate some of the problems related to order of operations between andOr you could simply avoid using base classes +Well is just passing cp on to the JVM so I think its fine with my change +Any update about the test gadget for this +I do confirm there is a single process running +The children can anyway receive other types of events +Some fixes in r +I though we had settled this for +The text looks ok to me but maybe assign next to Jaime to review the text +This is great information but a bit overwhelming +Changed both the factory and the style builder as well as so that it recognizes and finally added a test +hmmm +Attached zip file containing patch and keystore for operation +Thanks for the concise bug report and an excellent test project +we probably need this to simplify test and to do cross everything smoothly +Closing until someone wants to submit a patch +got it + +Move issues to next release +Ok rebased +Fair enough +Please retry +I think the RPC layer is the best place to transparently make the change which has the nice trait of providing the capability for other projects RPC serversRecent changes have already laid the groundwork for multiplexing RPC connections +attach a new patch which rebased to latest trunk + to and trunk thanks for the review +mingzym Still an issue? +Some better Filter API is required which lets us know of the contextTheres an opposite opinion +last comment +Already in master +Now that work is underway with AS all previous community releases are +Cool thanks for the pointer +Marking EWS ER as fix version +committing this one thanks. +Here is the patch that makes MapReduce to work with new RPC compatibility protocol +Thats correct the commented code was a victim of nearness to what I was fixing +Please try the support forum and explain clearly there how you are using Spring Security and what is going wrong +what about using +c all tmp Run EJB testsshell +Thank you so much. +The servlet can be registered using the +So now we have the code for the tool in you have commit rights why not you commit the stuff diret yourself +i like the idea of doing the FOLLOWERINFO and adding the server id in there +Updating component to since this is not an Eclipse specific issue +Its much better if is flexible in the response data so that it can adapt to their whatever model systemThree solutions I mentioned could solve this +So i guess this is a behaviour in Safari + +It seems describe extended command does not output in my test +New version after refactoring of Solr sourceThe patch installed fine +The causing issue is following is working for me docck have fixed +This new patch old patch reformatting based on code convention +Sorry for the noiseKind regardsSiamak error in the current building +Thanks chackaravarthy! +Nit its probably time to add javadoc with returns to forceFlush to help prevent this with the nit addressed thanks. +on Jenkins is not failing because before running this test plugin has been already accessedBut localy there is a failure +Not fixed in EAP ER thus Im reopening this +correction from Mark Proctorjxl is used in the decision table table it can be exercised by using the decison table junit is fixed I have increased the unit test to have a number of doesnt appear to be an ESB issue at present can we close this please +Has been fixed in the ESB codebase will be in the next merge +A patch with some simple documentation to keep consistency across the plugins +I just need some other devs to test and hopefully approve +Ill get the London thing in but the rest would make sense as a separate JIRA if thats ok + +I still have two tests left to figure out why theyre failing but Ill get them working later this week +Alexey for beta lets reduce the clutter in File NewRemove all the CDI annotation wizardsLeaving CDI bean interceptor and decorator wizard +So it isnt that merely dirtying the pom triggers something so start freshAlso Ive been running this using mvn U X clean verify every time +I could swear we fixed this one +Following title might make it more clearTitile Encrypting an encrypted database with a new key and then under thatEncrypting databases with a new boot passwordEncrypting databases with a new external key +But if the impact is low we could make an exceptionIm not sure about adding those two libraries to the engine as we try to keep the number of dependencies as low as possibleBut there might be a way to make the dependency optional and document how to add the dependencies when users want to use this feature like the Drools task? +Works for me +I just committed this. +The fact that browser do or do not do certain things does not prove or disprove anything +FWIW I say go for it +This would mean that we are currently only able to run expanded as the path matcher we use in other places is not working unexpanded +Closing. +Please verify +ok cool thanks for your helpi think we have an idt error here if there is no other issue shall I mark it as Not a Bug +Fixed for M +Your suggested approach looks good to me +committed to trunk +It puts all moves into the patch but still thousands of files and files Useless for review +changed some indentation and added a patch +So we can make it resolved status since youve already updated it +I think instead of adding a new configuration option this should simply use perhaps changing that to be rather than and implementing +Add patch to leave off on https for engine service applied thanks +the console tree was never optimized for large databases so for that specific usecase this is not Im more worried about is if tables in the db also takes ages from +Thanks a lot Ivan for the reviewI have updated the patch by addressing the comments above +Fixed and deployed as snapshot now that is released. +Is there anything a can do to helpIm afraid theres relatively little to do +I have no idea +I still feel like its easy enough to ignore but I understand the point others have made +The edit history is tracked so you can always see what the original comment was and what were the changes + +We will most likely add support for listening to task changes +Hi Jordi please consider that task execution is asynchronous +This would also help eliminate the many confusing Container killed by messages that are appearing in task attempt diagnostics for tasks that are otherwise operating normally +Ive tried to do an advanced view browser for setting query params but never got the UI worked outI do like the autocomplete +Just a FYI who is waiting for this fix +Thanks Navis +correct no differences in jars these files are in Branchbuilddocs and are added to the zip by the ant target in please follow up with Shelly +If the writer of a given ledger L removes C from its configuration and writes to then we can safely recover the ledger fragment of C for L +If all session types derive from should the existance of the class be hidden and +I find this very usefulI hesitated between using or as a separator and I went for since it worked nicely with the rewrite handlerI think either should be allowed or we need a more powerful rewrite handler since not using slashes will make thing much more complicated for meThanks for your answer +Either way Ill close this for now and let someone other than the author request addition +Its not perfect but a direction for future filename is uploaded as attachment under +and maybe try a clean install instead of just install in case there happen to be leftover classes in your output directory +Encoding of sasl messages could also be delegated to the but I would prefer to do that in a separate jira +Or would you rather partition things into simply mapred and hdfs? Personally I dont think of those as components so much as and think that those existing classes are appropriate for these accessors +Migrated all code to work with Castor +ThanksBest regards new patch applied at revision r please verify the problem is fully resolved as you expected thank you for the enhancement +Indeed there is an example with the patch +Nevertheless we should change the implementation of the connections so that the password is stored when its edited not when the plugin shuts down +This patch modifies a +Please detach it before deleting +What is changing is the RPC layer and I agree that we effectively have a tagged union between the link result and the actual result +Could you attach unit tests that demonstrate each problem? Thank you +Happens to me as well +I used public mode initially but thought that public is quite explicit about what it binds +The versions are negative +The state only happens when the bundle is started and thats by design +Ill commit it shortly +Move the majority of and blocker issues to the next point release +Its saying youre using the wrong version of Netty not the one we tested the release with so dont expect everything to workIf you can provide a test case which shows this wrong version string being printed can investigate +We already have dependencies on other Services so you can use those as an example +Applied the patch thanks Matt +I have tried to install OO and tried the scenario again with correct configurations +check the server logs +Closing this as invalid please reopen if this is still an issue. +Reopen as per request of Mathias +And youre right +Good stuff +I believe there is some small restructuring of in the works +What do you think +Let me know if you need any help +I cannot reproduce this bouncycastlejar is not a required library on my box +Attaching updated patch and benchmarkThe benchmark is from running ten times each with the patch and without the patch and throwing away the first two runs to ignore JVM warmup +This is not a really critical case because in practice the servers would just move to a new round of leader election and the likelihood that this keeps happening forever is small It adds a unit test that basically mocks a server to force a particular sequence and interleaving of messages It makes some methods and variables public in to enable the unit test to access them directly +Attached patch fixes the three broken unit tests and passed all unit tests on my computer +I think this issue is sorely mislabeled +work around set JAVAHOME environment variable like set JAVAHOMEcProgram FilesJavajre then you can run bit jruby or use it with jre on windows +The fix for this ffdefcededefbfdfee uncovers a race condition in security realms +fieldName commentFit sizeMax add field date jpa class +The UI looks great +Sorry but you lost your bet I can confirm it is the levigo plugin +so has started to strict up the argument processing +Okay now I understand your approach a bit better +Please post a list of known issues or a design discussion if you have one +bundle exec rake clean +It looks to me like is using reflection incorrectly and not checking whether the method was inherited +Resolving. +I have verifed that with and without the manifest now contains the header +The proposed fix executes a second time the original request which is most likely not what you want when youre trying to prevent a double submit. +Many thanks Carl! +Thanks for your response Arun +Nightly builds of Community and Geoserver works +Committed to trunk and +Huh that is peculiar +Hibernate expects method get in the class +Im willing to make a run at this but not until my mass of other changes go in +Changes staged modified sources committed to SVN +Also what should be if single map input split is constructed from multiple filesMilindWhat contract are you referringThe current way of choosing output file names is purely incidentalThe output file names do not bear any meaning except that they are all distinctI dont think we support the case where a single split is constructed from multiple filesIf we do in the future we can choose any one of them as long as they are distinct +Note that this same technique will work with as well. +Catalina has provided a new private patch for this Issue which does not introduce the multiple database dictionaries for the various DB zOS versions +done in a sec +I like bk as well +Outstanding issues is not executed by the build as it does not fit the Surefire pattern fails in the IDE Please review my changes to not rely upon the TCCL Tests is take up to s to complete why +That class is part of jboss modules +should this maybe be incorporated in to +Ah +There are common elements between the two vectors Their distance is indeed +Can you give me an example of a resource dojo has been unable to resolve when packaged into the Tapestry jar? Ive not had this issue so far and would definitely like to investigate further +Kathey I agree with what you are saying +Maybe one simple solution is to send a timestamp with the report about when the health checker was last run +Correct +Hi I can see this is fixed but right now I am stuck between and this issue + +Seems to be a bug of Eclipse Data Tools +This is working fine in +Michael good but can you please provide a plain text patch made against the latest committed trunk? That would make much more easier understand how to integrate your modificationsI havent had a look yet at your contribution my question is does needs to be integrated directly in the DM core or it is something pluggableTIA contains svn patchs to apply to svn head +Since a PO can be drafted to a supplier with lines to be distributed to various facilities one user may be able to view purchase orders but only be able to view lines and because theyre sourced to their facility the same PO can have all their lines viewed by another user because their view is ALL system facilities +Oh and to answer the other question Im happy for you to host it on your internal one if you want to +This is a patch for TDB which allows to configure and use LARQ if available in the classpath +since the issue doesnt exist in close it. +There may be some other items that need updating as well +Use with patched with attached patch +This info needed to getinvalidated when object for a column was reused to process the next rowAdded a test to based on bug reportmibmsvn commitSending javaengineorgapachederbyiapitypesSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +bq +The same as the previous include but the attachment is made with Grant license to ASF for inclusion in ASF Search feature has been improved +justiterate and print the map entries eliminating the quoted line above from thejavadoc. +I guess the intention was to move this filesrcmainjavaorgapacheuimatmdltktextmarkerconsoletosrcmainjavaorgapacheuimatmdltkconsoleAs far as I understand the patch the source file is deleted and a change is applied to the destination file but I cannot see an instruction to add the file at the new location does someone know how that works in a patch file +If you finish it before we release beta please correct the version +Wouldnt be better if both properties can be used in the same time? This wont break the previous functionality. +Ive eliminated this issue by caching the descendant font within the instance of that classSo there is still the left and of course all the differences within the as already mentioned in my former post +It manages the Hibernate environment itself +Ive tried to repro this with zookeepers running on the same machine and zookeepers running on virtual machine and I cannot get it to repro +we hope +I have no objection for revert +Good question actually +Close after release +Gdal has normalized my data column to kronendurc +Documenting this in the changes report and in the upgrade article +If I want to do theright thing and format dates using the current then I need to havecanned format patterns in all combinations required for each of my localizedlanguages but this is just rewriting functionality thats already available inJava through and its unlikely to be as accurateEven if you dont support languages other than English datetime format variesconsiderably in the English speaking world +Oh and I dont think we need the commented out code that used the Random +I worked around the above issue by manually copying the war file into runtimewebapps from the target directoryStill appears to be failing +rvmppclnx was full again +I think Patrick said hed add the Freemarker examples +I think the compare method should NOT be called if no docs are in the scores arrayI would expect that every docId has a score +wouldnt that ideally be the transient spaceWe in any case need some way to represent uncommitted content in as otherwise theres no way to do things like node type or access control validation before a commit +Includes protocol details for management messages +Ive just committed this to trunk andThanks a lot for the contribution Colin and thanks a lot to Chris and Suresh for the reviews as well. +So Im struggling a bit to like just this change the new implementation is mostly a copy +Alternatively you can just wait for the final release which should be available soonThanks for catching this. +This is not the same as +I think youd be better off having your Action classuse an object or collection to store the things you want to write messages forand when you read an entry off that list to create an for then alsoremove it from the list +As is a major new ability for VPE it will be mentioned in depth in the Components section of the Release Notes document and thus does not require another individual release note +Thanks for discovering that LeonardoWhere do we go from here +Dont we just call getChildren against RMAPPROOT once and from there we get the list of application nodes and iterate through them? Thats exactly what Im thinking if we have lots of applications the extra remote call cost is proportional to the number of applications +I am planning on backporting this fix let me know if you dont think it should be +Thanks Kristian +interfaces a patch that disables the failing tests on JDK for now so that we wont see all this noise when testing the release candidate +No body is seeing my request still now +I dont think it makes a lot of sense to pass a tagbranch name into the HG scm providerProbably the easiest way to fix this is allow users to pass Dtag on the +Id appreciate hearing if anyone feels that alphabetical order is betterHere is the order that I propose for the system procedures +Would a patch for that be considered +The JSTL core tags are tested only +This sample demonstrates how to invoke stored procedures from Synapse +jbpmfeatures + last exception happens to all extensions at shutdown +No this has not been fixed yet. +I distinctly remember that roo was running and it updated the domain files and not much said I tried it all again tonight and with pretty much every combination of running not running it running from Jetty running from gwt dev mode even dev mode via the Intellij plugin and the same actions all worked perfectly +On the other side people not familiar with WFS might be surprised about why their reqeusts and resposes are not schema valid +All tests are green +thanks after looking at the code I can see will fail too if encoded chars are included +So uses both H and prefixes +has been marked as a duplicate of this bug +Use the attachment +Patched committed with svn revision and I have verified that the repro given in the description for this issue has been fixed +closing as Wont fix since I suspect you ran into a special error the error I can reproduce looks pretty expressive +Linking this issue to +Thanks for the suggestionI think you are referring to the nodes classpath +Jim can I convince you to use Jira +My bad here is a patch that has all the new files in it +HelloWould you please try the patch? Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +The problem is not only that we need to omit them at execution they also cause compile errors that I dont see a chance to omit +I have fixed failures and masked failure and commented out failure +Since the JVM doesnt unload native libraries the only way youd be able to get this working across app deployments that I can see is to create an AS module for the jmupdf option would be to isolate the interaction with native code in a separate service that has no real logic of its own +Do you see a way to do this? The only way to use substitutions I saw was to include a component in the config file that +Patch applied +Username rafciom +shalin mentioned on irc that hed take a look at mentions that it works with +Are you just looking for the notice of successful end of build +Attached patch file patch in commit bccdfbebacabaaffThanks +Works OK from my point of view +Thanks MikeYes +I just committed this. +signededsteiidadminshell jcoleman soa Dec +This problem and the attached patch are similar the previously fixed and its patches +File srcmainxrtsorgapachemyfacestrinidadresourceKey Message The date and time is in not the correct format + +The patch looks good +Actually Im referring to Maven dependencies and not AS module dependencies +file scheme to point to checksum scheme to point to This is backward incompatible +Changed the component to java client Thx Emmanuel for spotting itI would assume its probably ok to rename the package name from commons to qpid +This seems to be an enhancement request due the current support for checking long ranges +Though I wouldnt wait too long to take a crack at it because I might get bored and write it +Ill try and find time to look at is no doubt that Struts code should be changed to make best use of beanutils +Remove the queued concept since should take care of thatAs mentioned earlier the job runtime will be useless once the JT restarts as the job could be running for more time than the time displayed +I vote that we take the latest brew build of the installer and put the on the CP +Yes anonymous cvs is broken for almost a month now and we wont be migrating the core to Apache until goes final +Wow this is finally done dont reopen please! Just a typo in one of the messages listed more then one time +How did you test it? What setup do you have +Yes indeed there should be reads I have a guess that its somehow related to the LRU part of the cache because the interesting is that it returns a valid key position testing this theory +JesseYou can decide what to put in this JIRA as long as future enhancement can be easily accommodated +Patch for test suite in over the latest All tests passed +Marking closed. +IP Clearance file for s about the same I found when I had investigated the name +Hi GonzaloPlease see for those featuresWoonsan +it seems like making the nodes doing the work is a better encapsulation +Id be much more comfortable with that policy if a patch had been approved by consensus that actually added the annotations to all unstable public classes +Created a new nightly directory in jbosstoolsjbosstoolsexamples +Please reopen if new evidence emerges +Another testcase for this anyone mind backporting the fix for this issue to the branch? +Im not sure about n +Not committing this yet for fear it will break TCK +Thanks again Nicholas! +Attachment has been added with description Short that demonstrates this P open bugs to P pending further spotted Matthiasattached test case is now rendered as expected. +But you can easily write your own validator +We can make the wrapping collection class do wraps on puts tooThe approach is costly and unnecessary since it adds multiple bytes to the serialized form and extra references to the structure +Ive created and +Enabling Declarative SecurityThanksPeter +Its abnormal situation if this method is not found in the library and in this case we will be notified with an error instead of silent return +I still need to add some tests and I believe I am missing something with the metric registrationinitialization +I will collect all data and send it to ESRI as we have a for issues like this +Shouldnt it be something like Attribute maxRequests is invalid for element type settings +Also I dont see any assert conditions specific to retired job in the code +It may be no one is maintaining them so they may not beWhy dont you just delete that source file from the project using the solution explorer? The other possibility is to use your version of Visual Studio +After that period if a different client request a lease for the same file the resouces associated with the original lease gets reclaimed and the new is satisfiedA clients lease trasitions from stale to expired after a hour period +Same for send why two language IdsI think if the invitation email is in english the inviation link could also set english as default lang for the application +Upgrade for branch works successfully +closed +I dont think I like the idea of reading current structure because this will make it difficult to run against concurrent versions of HadoopI think we need to do further investigation and find other alternatives or just say that this will resolve once we have a reasonable support form Hadoop +Paolo yes true makes the problem go awayI note that the setting affects but does not affect nor anything else in ARQ +The outcome of that discussion was this first a deployment gets deployed resulting process definitions can be used +as it can be seen there are two before ua for first set of logs hdfslocalDCrcbillingua also replication SUCCEEDS in the above case but no data is actually transferred +Scheduling for +The difference is explain does the optimization but describe does not +Would you pease verify? +JobsServicesDone. +TedI will do the testing as you have told + is another Patch committed to trunk +was meant to track processes which involve the version strings of products and the places they are created at +It is great that we can now generate the source from the build script but if we dont include them in the Maven bundles no one will be able to take advantage of them without downloading the entire source tree and running Ant which kind of defeats the purposeThe script lives in the infra branch +and committed to the Giri +Ive added a that shows how to code a client that doesnt depend on Tuscany at all just uses the standard SCA spec programming model +Admittedly extracting the renewer is a bit dubious +Distance calculation +Because of it the original patch was reverted and replaced with an update +Marking closed. +Cant reproduce in +because QA returns after comitted +Indeed enforcing a char month and a or char day seem to go through all present tests and works with the case i submitted +Sorry I didnt mean remove the with the cores on them I meant the constructor functions +Patch for this JIRA comment for the patch +Its worth noting that during readsinserts you would expect the op rate to be equal to the key rate but this isnt always the case due the concurrent processes involved however they should always be fairly close +Attached revised patch file +I was worried that we were passing to it +Manoj Could you provide a patch relevant to these changes please +Bulk close for release +Patch committed in r +META +Reassigning to +I have to reformat which is easy but renaming all the variables is a bit tedious. +fixed +There is a need to update chapter +When I unloaded the leader neither replica became leader +isnt this same as +Try running just cordova serve and then hitting httplocalhostandroidwww +? Thought that there will be an exception thrown when loading the mapping of +When tasks of the job change state we could store the update immediately to a log +FYI in my test cast the exception doesnt bubble up to the original caller unless tracing is turned on +Itd be nice to get this into +alt attribute for was still missing +Filed for +I ran with the following combinationsI precreated the database systemwombat with to test soft upgrade modeI filed issue related to this testingI tested the tests with client server and did not get any interesting failuresI tested the tests with client and server and did not get any interesting failures. +Perhaps if it traps and there are some native threads elsewhere it can hang dont know +Changing the priority +Will make a test case for xbean and try to get resolved +I would hae thought Spring beahvior would be to automatically use the existing bean of the given id instead of failing +If so extra JSON Object is added root name comes from value class or forced static type by default but JAXB annotation introspector adds based override as necessary. +woo +Thanks a lot Todd +is a bit more involved so I guess its why its been left as a future issue +One to investigate on the POI bugzilla entry +that would ask for making a WCS client +this is an issue that i know exists but its very difficult to replicate +You are saying that we still need to do the checks in and cannot simply rely on Hives checks correct? Hence your patch has still doing many checks +Wrong issue +Yes I know the date format system +here is the sample the second sample code +ok +fixes in standalone up sloppy to trunk revision Thanks Bob. +would hold the conf keys and defaults for the Job History Server +Previous patch contains only test changesSorry for thatThis is a full patch +I have created a sample workflow to test and attached it here +I am not sure if they are sufficient to fix the test but it is worth a try +bq +Tried this on my Mac and it passes every time on my Linux box where Ive been running the test it fails every time +AndrewTheyre exposed through jmx too +The attachment contains the patch for this samplePlease note that this is dependent on +either way you are going to need to send the port to the workers probably as an environment variable +The and jobs were tied to the ubuntu slave I have removed this restriction to see if that helpsAlso removed the Sling lock on those jobs if they execute on different slaves its not needed and even on the same slave that should not be a problem +I cannot replicate the issue with the latest SVN revision please give it a try and if the issue still exists provide more details +Application not an infinite loop it is just working hard to find an answer in a big search space +Some of these issues are awaiting retesting after upgrade of ODE version and merging of the JPA work from the ODE trunk to the branch +You can develop your solutions based on actions as opposed to looking for data orrelying on completely synchronous behaviourProviding infrastructure for messaging across large distances leverage the network connectors and broker networks to integratefor example geographically different data centres. +Closed at the request of the original reporter. +Err +Previous was not synced with trunk +Attached a patch that upgrades the version in the main POM +It actually was used concurrently before was introduced at some point in +Please reassign to Bolek if thats still an checked +Removed unused please check my work to make sure that everything was applied +Enis first comment my patch is adding those jackson artifacts to the dependency management section this means that they are not added to Hadoop classpath unless somebody add them to the a dependencies section explicitly or they are pulled as transitive dependency +What version of Java are you running? Show us java version output. +I think we should simply do nothing +Checked on on stax +Id still like to keep this one specifically for +Could you attach the sources to the test client +made one method static package level to be able to get raw type. +added a reverted the change of removing since it needs more thought +I tried using and after putting entries into node none of them is replicated to I create another JIRA for this +Jenkins is running into the same problem I did the bzip devel package isnt availableCan we change this so it builds with bzip native support upon request? That way we arent adding another required dependency for Hadoop native builds rather an optional dependency if people want the native support +This issue is caused by a bug in jUDDI +Can you give more information? Otherwise Ill close the bug +We have thanks +Changed external commit interface to not have locking but rather query if a commitrollbackoptimize is already running to accomplish the same goal +Hi Ashwinmany thanks for integrating the fixCheers +Applied in thanksdims +The should return a until the point that the service becomes available +Fixed on and trunk +Yeah I was thinking this as a possibility but thought people may want to rely on the ordering even if they mark it as unordered +Another ping +Dont think we can do anything moreShould we close this now or do you have something in mind to fix this? I think that sending a fault message is good enough here instead of trying to encodeescape the invalid stuff +Tests should be reproducible by definition +This version of significantly beats or equals the native implementation +My understanding is that returning causes all browsers to create a modal login dialog box that cannot be styled by the application +Could that have been a bw break since it did not do what it claimed to doI dont understand the question +c In phase a command allocates a new blockpoolId and a namespace id +MM th Oct GMT Added detail from and cancelled it +Is it the normal Finnish way to separate thousands +I would categorize this issue as CRITICAL since in AR fixtures scenario it stops rest of the specs from running +This is working as designed +Fixed in trunk. +Another attempt to fix the warning +Thanks for the prompt response +Thanks for the sample project to reproduce the issue +Thanks +Version of API matching proposal document has been checked inrActual wiki API document will require extensive update obviously +I plan to look into adding a regression test for this Ill upload that as a separate patchI ran derbyall on windowsjdk oksvn statM javaengineorgapachederbycatalogCan some please review this +I originally named it instead of +All code that addsremoves resources from inventory eventually make their way can you confirm if thats true +As far as I remember thats not the case +Were working on how to remedy this + is still not out because were still having problems with WCS tests so its still in time to be added +Nightly build issue caused the html books on the website not to use the updated stylesheet +The tasklet always connects as a Hive client which just like the Cli in remote mode doesnt perform variable substitution locally but sends the requests as they are to the that the cli supports! or OS commands where it applies var replacement since these are not sent to the server this is not something that we currently you reproduce your problem with the CLI in remote mode +Ill use that in the future +I committed this to the branch too. +Thanks Harsh. +I will pull this into. +Resolved in the pull with +Can you post the link to the documentation you are referring +Done. +I wonder if we should handle that case or leave it as is at some point when more documents are added that segment will be converted to a CFSI think that means that the first few segments that will be flushed will remain in non CFS format +Im with Jarcec +Any word if inclusion of Camerons patch is up for consideration in future Hibernate releases? The purpose of filters is to uniformly apply constraints to queries across your application but if you have to manually remember to do it for subqueries kind of defeats the purpose a little bit +Effectively the original issue reported in was never resolved +Is this to consume rest +I was able to upgrade to without changing the signatures and it all worked fine. +Let me follow up +It would be nice to use the same logging libraries in james projects +bulk defer to defer to +It doesnt have any hooks to trace rpcs or collect annotations that would work with or Hadoop +This is turning out to be a big problem the default IE which is being deployed right now is having issues with mapbuilder +Let me know if it pops up again glad it got working +Er the version is wrong +Creator forgot to link +Attaching a manually edited patch +Attached is a patch +In s at the bottom you will find a line that looks like this WARNchange that to DEBUGAlso in you will find a logj statement that determines what your log file is varjspwikilogsRecycle and you should see this file appearing +Patch looks pretty clean +The contains a snapshot of my User Guide development effortWeb Page trunkbuilddocshtml describes how to build the documentationWeb Page trunkbuilddocshtml describes how to build the sources on many platformsPlease post your comments to the mail list +Do we need to keep the check against the hardcoded classpath values so the test will fail if they were changed in +It will work either wayThanks it seems ok to me +This has now been done. +Kurt I will close this ticket later today as I will assume its ok if I dont hear back from you +I see that currently a VM is running in some pod we are always trying to start it in that pod +Now Camel detects this problem and fails starting +So seems sufficient to provide a way to drop all statistics and have the implementation handle this unexpected row also +Ive made sure that close can be called more than once without issue +I would also really like to add my voice to this is hard to get right and it usually scares me when I find yet another security framework +Do you think it would be useful to have this corner case in +This is just a sample trace that I made with my instrumented version of hbase +I did a minor modification in to avoid get on a implicit closed I checked the use of scrollable result sets +Its metter of our JSF Tools model + +sorry for the long silence I am currently busy with some other more important stuffI hope I will have some time in the next weeksThx in advance for staying tuned. +We can discuss it in the users list +We are not in need of something as extreme as a separate classpath per module +Frank you do not own the rights at this source code and therefore we cannot add it to our code base and relicense it +Where did you place the in your EAR archive? It seems that Spring simply cannot see the class it that it is supposed to load there probably because it is running in the wrong class would recommend to analyze the class loader hierarchy that your code is running in +for the patch +I will apply your changes to the trunk +I attached a patch to backport and +rNew tickets will be created for documentation changes +Thanks for the clarification JD Jonathan. +I missed refactoring outside of Common and HDFS +The unit tests will be fixed by +Whether that is appropriate is a topic for debate +and test all passed now +Attaching a patch for this issue clearing the tables used for test in on method +Added to Affects Versions +Attaching patch +yes setting it to null should work +There is also the component which is a mina codec +Fixed +I think I can collaborate with Yexi on this projectThe only thing is I cannot apply CPT for GSOC neitherIf it is ok for me to work with Yexi what would be our next stepLooking forward to your earliest replies +It might be worth making the decision via a keyword so you can add new keywords laterAbout the multiple languages in one field problem you cant solve everything at once +When the target broker is alive on startup thinks are ok for staticfailovertcp and I see the problem with staticfailoverssl but this use case opens a bunch off issues for the independent start case +Obviously a certain risk isinvolved when using FOP Trunk as it may not be stableBTW you should maybe pay a little attention to what FOP warns you about +Great Karel thanks for your insightJura could you please adapt this property for your needs +This was fixed with the merging of the regexpoptions branch to master +trunk +Patch supplied as attachment. +we should reopen this +fixed pom descriptor in the bundle +I ran derbyall and without failuresThe description of the patch is in file +Patch incorporating the review comments +Thanks for the prompt fix! Ketan +Edward can we get this in +Resolving +Im attaching a patch which provides a based on the +I have implemented action chaining but with little testing +The remaining work for this issue is getting tracked in +Preferable to upload the whole war +I reproduced it again +Error messages now include possible alternatives. +Closing resolved issues. +I am wondering if there is some mismatch of encryption because of the few odd characters in the url +I opened for the YARN on Rackspace bug. +stack Jun AM is about offliiningonlining not being reliiable on branch +I guess the attribute is an important one +Now we only ask the EPR from the transport so in addition to that we need to ask the binding object from the Transport ListenerSo that going to have API change in the Transport Listener API and I do not think we should not do that at this moment. +Lukas agree this should be incubated in the wiki first +Fixed in +It completes now +That sounds good to me +Still awaiting clean SMX Features build for nightly to test out AIX modifications +Eclipsethumbnail! Apple Application Support is not found on Windowsthumbnail! Failure on Linuxthumbnail! Another failuresthumbnail +Id love to have this feature in Wicket Will you be able to commit the code before Wicket beta +I like this idea a lot +Ah thanks sorry I missed that +I will try but I cant be sure that same exceptions or code pathes are used +Patch attached +Perhaps the best way to proceed is to move add plugin functionality elsewhere on the editor and let the addremove buttons do only xml manipulation +One question is are you positive the characters youre passing are in the encoding you declared in that document? The fixes proposed will be investigated but we also need to be concious of performance in this area +I had some earlier concerns with the patch that I had communicated to Sandhya privately +The only time that the send method can be used to send to particular destinations is when you create a without a default destination +IMHO users know the problem best +How can we address it? As says we cannot trap INT which would interfere with current VM behavior +patch pushed as commit decabeefabfeegoing to close this issue for now +So it is best to avoid maps as input params I guess +have a look at the users mailing user wasnt testing with KML it was google maps. + +This was fixed in. +development team will accept the request +Minors must always retain deletes whether they are actually deleted in the compacted files themselves or not +I just committed this +This seems to be a manfiestation of the same issue as but in a different query language +Uses couchdbkit because is broken +bq +But again no objections from me on it being a Solr contrib if others feel strongly about itI personally dont see a distillation of this anytime soon into a common security framework within Solr so unless someone already has this generified and a second implementation of it I dont find it a compelling argument to try to make this generic from the start +Merged in RSS and Wiki connector ports into trunk +This issue is fixed and released as part of release +Draft Revision removed added This wasnt necessary and is sufficient to handle Entrys +If it matches only we will try to get a lock on the file and process it +if we convert the question is to what +Anushathis is a clone of in which I explained that is not supported by the web user interface + +Do you have more detailsinput about this +next iteration +So the debug execution with GROOVYHOME set explicitly and calculated by using the location of the groovy executable should be different +but if you get validation error on your first change password attempt when getting there off of that key the whole thing blows up cause now it is expecting that existing password form element +The other solution is to write an append only log implementation that knows how to truncate. +Can you please commit that to trunk as well +Attaching new patch with the suggested changes + committed to trunk +Or if you are happy with this approach please close this issue +Fixes throws not invented here in core as wellPatch applied. +Just noting that this is a mandated change +I would say this would not be an issue if is fixed +Perhaps that should be +packages manually +Mostly minor errors +C depends on B and B depends on A then surely both B and A should be overlaid on CLet me know what you thinkThanksAlex +committed to on JBT +edited pom patch to get rid of trailing Ive done only minor changes to your patch +Using similiar technique to lchmod on I can fix this for linux xstat methods +Yep its only those env vars whose names are not natively all +too late kicking hudson +See. +Should be fixed in SVN trunkCould you give it a try +In any case were not disablingsorbs for you its far too effective +I have raised to get it removed. +Daniel we need some details otherwise we cannot do is not sufficient for us to determine if there is an issue and what it might be +You guys are awesome +Perhaps the DTD should only be used for the stricter structure and not for the more freeform structure +The hierarchies of existing filters cannot be changed for passivity no improvement can be done since all filters wont be changedrefitted anyway +Wow ok lots of pain +Some code changes in the object structures has not been reflected in +The last ZIP has an incomplete Reading reopen method + the test after the upgrade to +I think that your WSDD should say reverseReturn somewhere in it +you probably dont want the synonym filter to be after the stemmer +Verified on metamer. +Hi ChristopherIm currently working on a new version I hope to attach later this eveningIm adding faq examples updated unit test and integration testMaybe you can better wait for new zip +At the very least there should be a way to disableCan you share your Checkstyle and Checkstyle plugin configuration with us? It should be possible to make this work +VasileIf you can reproduce at will and can provide a tests case please open a new jira issue to track your use case +We found the problem in our code One of our data access points was using the wrong VDB version +what if get later on returns null for non existent paths? So unless provides a guarantee that an exception will be thrown for paths id like to go in the lines of what Amar mentioned in the code snippet +The top is always there +SO this issue can be considered as fixed +I really cant accept this decision +Taking either of these approaches is a risk for as people are used to the way it works now +A patch that would use the old methods and possibly fall back to the new methods if not available would be OK for +bq +Matt Id have already attached a patch if not for my employers patch reviewapproval process +secondary index references +Is it not a good idea to work from the branches? I was working off the latest branch and did a pull again but dont see the changes that made to itWhat am I missing +defining the RDF formatneed to update the Salesforce repository object to hold all the appropriate dataneed to update rssj to produce the appropriate patch RDF salesforce repository object on Devadded new fields to still need to load the products repaces replacedBy and requires fields I need to look up the values in salesforce and need to find an efficient way of doing it +Tried surefire Works as expected +so I removed the create methods entirelyit was not easy I misplaced one of those exists checks and it all blew upattaching patch tests pass feedback needed +Pulled in fix for the bottom of the patch the value is true but the description says set to falseIs it correct that the default value should be false +Its better to consider the load information when placing the daughter regions +bulk defer minortrivial issues to +fixed failing would be a nice patch to have in +Thanks Scott I will try this out either later today or tomorrow and let you know what happens +Did you run test in a loop to ensure it does not fail +If all callers would be changed the ? extends could be removed +I am happy to open another ticket if that makes more sense +Please close this JIRA +SVN +What the cache requirements are specifically is that a cache object must be retained until we are finished with it +That Nullpointerexception occurs because the postoffice is null +This works using the RC can you +Unchecking Regession checkbox as this is not a confirmed product as Cannot Reproduce +Anyone knows reason behind such a restriction +Yep I know +If you need any other assistance please let me for this issue Dan +Heres a simple patch +The changes have been committed to the product branch and should be in the ER build +Its about money + +Hi LucasSorry its taken me a little while to look at thisIm having a little difficulty applying the patch to check it +This new patch should fix the patch after stripping out some superfluous code. +I just kicked the build server to do a fresh checkout for the IBM Java build but I have not seen the results have no theories why this would be failing +appsshowcase +I assigned to myself by mistake +Thanks Kay Kay. +as Id really like to commit this asap +Attaching in case anyone is in +Im sure I will get thereD +yes it fails consistently whenever one tries to compile EAP testsuite on entirely true +any news from the gump folk +bulk close of resolved issues as part of JIRA housekeeping for release. +Hi ChrisPlease review the patch attached for the same +Did you even take half a look at the patch? Explain to me how that patch can ever have any impact at all on any existing hibernate mapping +This points to the JVM but without much information IBM will not be able to help +I have requested an upload on maven repositories of the version in issue is fixed we will proceed with updating this plugin pom +add Wizard to file to TRUNK +If that is so you should modify the ant file +adding new configuration and change am configuration to +I suggest that we create a small test set with articles and then use these for training and testing +My actions is that I am trying to purge an overloaded queue where some consumers are already active and then I have something like messages +I also added exact match promotion to the top of the suggestions list regardless of the score of the exact match +I think it is actually a for because otherwise the feature implemented in would be incomplete +Juca the patch to test is which is signed +The file contain revised files with more changes than corresponding ones in file +Please ask questions in the optimal way +Error occurs without patch +sorry was moving through these a little too fast I guess +Thank you for quick verification! +Hash or code or something +Ive asked Snjezana to point out which to use +One more thing for the release checklists Ensure that the release date is in add a new release section to add a locationmap section +You are free to use this code in any way that is helpful to you and if there is anything I can do I would be vary happy to +The configured job just reads a file a puts everything into the database +Id suspect itd be something that would have to be run into once users want to start rendering large complex pages anyway. +oops setting priority to minor +This is fantastic stuff thanks Armaan! +I guess that you might add a webresource section with an include for the files you want and a targetPath. +yes it was during the +Although there was a failure in tests of shims the build continued to test ql +So now you should not see the same errors in your script files but you will also not get any content assist outline views or eager parsing for these filesResolving this as wont fix but perhaps there are some UI improvements we can do here +Will be available in ibiblio in the following hoursAbout mirrors we dont support them we only publish them for users convenience +I saw a failure in in running which fails also in a client without this change so I dont think it is related +I think its ready to commit +I think the way Named qualifier is defaulted on field injection point is reasonable and could be also used for bean definitions +No tests are added as the protocol related changes are tested using existing tests +Thanks +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +Of course this only works for methods but that is a problem of the programmer I guess +Parallel rendering on the server is on approach +Lets do it more convenient +Attached is a wsdl that can help replicate the problem +The patch needs to be updated its not applying cleanlySome minor style comments on Phabricator +It looks like the messages for Spanish have changed in the environment used to run these testsWith just the clean svn co I do not see these failures +move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +But thisshould either remain as a bug or a feature request +done in Rev +Attached a second version of the prototype as which supersedes the prior patch. +Does workaround exist? +Any reason this has to be targeted at rather than or even? Seems like it would be a pretty applicable patch +If already set dont go through with second close +Thanks Vinod +Please reopen if you can provide more information. +Added the requests and responses. +Verified +Great! Thanks. +There is another jira for this +moved to archive +Do not forget the libendorsed to the java command line in these scripts or Tomcat will not run +Chunhui Shouldnt the comments in the properties file explain why there are two hbase contexts that can be loaded +I can confirm that this bug treats us very often +Patch updated due to recent configuration changes +Hi ChrisAS long as you dont persist the CV code in DB there is no problems to keep it in a session +Im closing this issue because this functionality can be achieved by using a combination of and urlscoring filters +This is the heap setting in our toolXmsg Xmxg XX XX BTW do you see lots of rebalances in your consumer log +Resolving as fixed +Assume closed as resolved and released +That wont work the onRehash attribute is not exposed by the subsystem +added simple repro +Pig PatchAddressesMathstringother functionsMisc updates +You will receive new numbers now when running ortpasswd in the form of jsxxxx where xxxx is random first time but stays with you until reset again +Any life here btw +m sorry +Should have mentioned this patch includes both patch for and this issue +Committed revision Sorry my bad +Thanks Namit and John! +Merged to doc branch at revision Leaving open for the next little patch +thanks Andy +I am currently performing analysis of whether replacing CXF jars with Jersey jars has performancefootprint benefitsI will provide an update shortly +I think having II implement slice is much cleaner than Directory having to implement createSlicer returning an with only one method +Talking about it at hackathon +He would also like to change the name from Create project to Propose project or some other text that is clearer that it is being submitted for approval not automatically created +Better to just add a flag to the syslog sources to use the whole line as the Event body instead of just the parsed content portion +I believe the above showsome of the casesAnother case I have seen is an order by of being satisfied by join on when it is known is a single value +Changes merged to the release branch. +So I changed the assignee to unassigned to not block others who may want to resolve it as well +What was different then? And is it something you can possibly go back to +please review and commit +BTW it would be better to run the consuming client on another machine so that the consuming clients is not contending for resources +i have been very busy with trying to understand workflow of castor source generator i should have notified everyone so anyone else could have picked it instead of me +Again dont close it until it is ported to trunk +This code is so brittle I dislike changing it +Turned up some good stuff +Rerunning through hudson +I made a preliminary patch for and verified that it fixed the test +You are responsible for closing all of these in finally blocks after you are done with would be possible using the wrappers to close result sets and statements when a connection is closed or when they become unreachable something of this nature may be implemented in jboss +Its a one line change to the package header +The first cobertura run will produce a report I think because it doesnt write data until the VM exits +Clicks table always renders one cell per heading +module does not have tests +But if we move the code to Apache SVN we have to to fix this issue +I want of it +Its been tested locally and indeed jobs only go to a task tracker with sufficient space +thanks santhosh +A DNA context menu appears in the context menu for all +I think it would be good to have a class comment with the summary of the handling of values in JME and an explanation of how tests that use values should be written so that they work with JME +Log file of the test execution at debug level +look to have been present before application of this patch +We dont need both +Display what message? Youre throwing an exception which because its inside of an EJB rolls back the transaction that its in +Time hours +Hopefully the new version of the jar plugin is to be released soon +So in a way i didnt change much +adding Patch to resolve the issue +Id say we should leave this issue open +Glad you liked and thanks for your continued feedback it really helps +Will try to test again +is not a suported platform +I need to reproduce the issue locally +bq +As to your authentication problemplease post it as a question to the listOleg +we wont support this construct in the foreseeable futurethanksdims +The reason I was considering it was that because as of now if you use a Hibernate xml or properties based configuration within the the exposed managed attributes do not reflect these settings +As youve said this is because the seed data for the base delegator doesnt contain a reference to the tenants +Am i missing something here or is this really a bug +Attaching diff for and with all those newly addedLet me know if anything is missingHow can I check this in +Seems that i dont have the permission to reopen. +But I definitely have a UI working with all the KML reflector on the nightly from last night I wasnt able to get a broken WMS reflector +bq +I committed a fix for this and deployed a snapshot +We have our own Eclipse compiler adapter +I think this issue would involve picking up job jar from a dfs path directly instead of local fs +Mark Can I have a copy of the saml response from salesforce? You can remove any sensitive information that you may have +pending +currently no unit tests pending a unit test strategy for the whole mechanism +Because we have until Beta for that +Otherwise it is under PreferencesHybrid m not sure what do you mean exactly +Thanks Lars and Ted I will try to reproduce the failure locally first and then to see if any logical bug of the schema metrics flushing +IMO we should put the effort in a working solution than a more or less complex workaround +Works fine. +And that fails in your setting stepwisefalse should prevent this +If this is only about MR jobs isnt setting LDLIBRARYPATH in enoughYes it is enough +Most possibility in my thought is validation progress in unfortunately they used different parameter name in each SinkCould somebody give me your advice? +I think we need to have a property that controls the level of output to the log file +Thanks for the add +Good point +I am not able to reproduce your error +enough I was only looking at the online docs. +Ive seen this happen with custom workflows and also I think from JIRA desktop clients. +Ill be releasing as soon as I finish my own testing which should be in a day or twoAnd as usual feel free to reopen +While that happens you could have updates coming in the memtable and if you are unlucky have this memtable flushed almost right away +Patched on HEAD Revision Changes Path testsjunitorgjgroupstests Revision srcjunitorgjgroupsblocks +Thank for the fix Igor +just committed this +I updated to use the getOptionhasOption methods and it is nice to use the same constants for those that are used to define the option themselvesAs far as is concerned I agree it appears that it could be merged in but Im uncertain where its used outside of the latest work +Correct +A Message is arriving at the same time the transaction is being made +Done +I tested it and it replicates as places blocks as intended +and have no precedence issues +I cannot reproduce with the released versions but I can with my chechout +Thanks for the response! If you check the latest attachment is contains only source files and nothing to do with VS +Thanks for reporting + shows that there are more situations than just this one here where such things could be automatic list expansion on method calls is something a static type checker cannot really do with the types available on the JVM and at the same time allowing overloaded methods +Today when I was reproducing the issue wsdl file which was generated contains this URL of WS service httplocalhost but as I remember in Beta it was httplocalhostprojectServices +Touches the following filesM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompile +Thanks very much! +One thing my PC is bit you please attach zipped plugin of your distribution +I am still thinking doing more on mailjdbc side to avoid deadlock situation +Cant reproduce in V of the plugin. + for reusing I dont see any problemIve fixed initialization problem of record readerwriter objects for the next version +We still need to change the code calling get and get since the existing code is wrong assuming the default setting must not be from the table specific setting +settings files? Or somehow via the +Correct +sorry my fault then +Looking more into it it does look like its an issue with the producer +I got the initial thing from Axis +This parameter should also be a List not a comma separated list +One note I tried manually removing the deps from the archive but deployment failed as if only the archive is searched for deps and not the normal maven mechanisms +This bug has been marked as a duplicate of +NTLM auth to Sharepoint succeeds with JDK and crashes the VM on JDK +Kristian is there still more work to come under this JIRA? +If it should migration is pretty easy its just a new parameter to addNow that it seems to work well like this the next step is to make work and to handle and test module descriptor aware version matcherThen the last step will be to make it pluggable which shouldnt be too difficult. +The gist is simply that we had the wrong principals configured in a few places and in a few other places we were making or authenticated HTTP requests without providing any credentialsNo tests are included since security must be enabled to test this +Obviously thats quite a huge break for clients though +rename the message into should now fail +Committed patch revision and patch with revision +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Commit the implementation at revision +we shouldnt the around but rather make it a task +Merging of subcontainers needs to be handled in the implementation +So I think expecting rewind to work regardless of the wrapped stream type is overzealous +Committed revision I need to test this on Windows +Would it be worth opening an issue with the Chromium folks to see why theyre doing this +This is committed only to trunk +Seeing as Hive is nowing moving to its own TLP this could be a chance to move to Confluence if you really are still unhappy with Moin or Ill close this issue soon moinmoin updates will happen when they happen and are not project dependent +Thanks Chris! +Cleberts fix looks ok in have applied a fix in branch too +You may now set default values or callbacks on types. +updated with some +Sure thing +Yes I was unable to reproduce with the patch applied +Also I wouldnt worry about that lazy sync compared to any IO cost its probably nothing +Just for fun you can stop iptables to see if it is the culprit +Verified in CR. +But adding the extra compare of every kv would be a big cost to take onCould the merge be thrown out of whack when a new file is added in after a flush +verified the patch with rpm installation on a single node clustersinglenode setup script requires a fix to create the required mapred dirs in the hdfs fs +Hi Colm the patch still required a bit of work +I guess this one is probably for you is the difficulty the feature collection is only allowed one kind of feature type for its contents +The modification replaced a couple of bad for +Yep confirmed renderer support is still +i just committed this +Committed to trunk and branch +bq +Im not at all surprised that it doesnt compilepass tests under you can get the VS VS and example projects integrated with tests passing that should be sufficient for you to issue the pull request +Sorry for not very responsive I created to track the issues. +The global stuff needs either to be stored in ZK or we are going to propagate this back to the via statusreports +Attaching patch with all the changes incorporating the comments mentioned +Fixed on and on and trunk +The test which broke was one of the jBPM integration ones there was no explicit test added +Without the guard if you declare the method in the interface you get an unintelligible exception +Ive attached a maven project that contains code to implement this feature + +Thanks for rooting out the SPENGO +The patch looks a good fix to meI also verified on both Windows and LinuxPreviously both failedWith the patch the test passed on both Windows and Linux +As the name of the method implies add adds the headers configured on the to the given +Sorry Jacques I have not seen that it was a grouped bugs +New patch resyncd with are in streaming +The OS JVM and Eclipse are bits +bulk fixing the version info for and all affected issues have in +Is an Apache Felix already available and including these fixes +gitkeep and removed unneeded imports +Can I get someone to take a peek at this folks are asking to build against +Ill post more when I find out definitely +I just wanted to verify with a colleague before getting back to reject this issue and I will find a better solution that allows adding custom fonts +Yes description says listener does not fire +or do a moveAll of those things will take a different amount of time so we cant really estimate the amount of time an iteration will take +Changed the description to include the SCA samples as they have the same problem +Yes we operate under a heavy write load and we do have frequent compaction as a result but under normal conditions I never see this exception or at least it doesnt happen often enough for me to catch +Todd Lipconbq +FYI none of us know anything about nor do we have access to copies of it so this is going to need community help +Todd! Ive committed this. +hey want to break the benchmark out into a separate jira issue for simplicity? Done seeDeleted benchmark patches from this issue +I was surprisingly far with the plugin with when I patched Ant myself. +I must have copied the old sample configuration +Done used your gmail address seeing as you didnt Gavin! +Housekeeping due to oddities in the migration. +Resolved as fixed thanks for pointing this out! +So where do we stand with this issue? Is it a documentation issue or should we attempt to make future upgrades use ZK transactions? Furthermore will this issue be present in due to the upgrades involved +Where is this? Is it done +Ok then Ill close this issueThanks for testing Costin. +Maybe the JMS expiration value should also be carried with the Message object +We will always try very hard to avoid heuristics +Infact though it was static the scope was private previously +This seems to fix the problem +The reason is the adb does not support the minOccurs and maxOccrs in sequence level +Sending srcmainjavaorgapacheservicemixbeanAdding srctestjavaorgapacheservicemixbeanSending srctestjavaorgapacheservicemixbeanbeansTransmitting file dataCommitted revision . +Maybe this will work also for the repository +Generally speaking not +I have the basic functionality working for reads now +Quite frankly this sucks +Thanks for the info! Do you have any more details on exactly where I can download this from and the steps I need to go through to rebuild the necessary Much +I have a few ideas that Im playing with +Also notice this patch is only for cuz is gone in I will coordinate with Xuefu to see whether to solve it in parser or +Do you think the new version worth revoting +Ie we wont know if someone introduces a bug that causes same problem again and causes the first location to be associated with each of the partitions + +It should perhaps be broken into two classes one in solrj to be used by clients which is designed to balance across a known set of solr servers defined at construction time and one in solr proper that deals with the distributed requests which are dependant on the query itself +quite the opposite. +By reducing the heap size and running with high number of map tasks I could reproduce the OOMWith the attached patch objects are being garbage collected and I am not seeing the OOM with the same settings +I tried the patch together with andThey eiminate recent classlib tests pass rate degradationdrlvm smoke cunit kernel pass +Attaching Michalis fixed patch from for your review +Of course anything can bind a instance into JNDI it does not necessarily need to be the any rate Ben Wang and I will talk next week to figure out a long term solution +Basic functionality with documents are now committed. +removed master coprocessor changes from patch +That is a geronimo message indicating it cannot remove a directory as part of undeploy +It may not be a good idea to depend on a single developer project +Actually I have raised the same question to both Ehcache and Hibernate but no response from both party +All outstanding issues resolved or noted for next release +I propose we split this patch into two +I dont need to set renderall he already checks this +Sergio this back to you as you are updating it anyway +Seems to me that Byron already made the switch to for +Thats either microsecond based time or its past AD +AFAIU the script expects that +This means we will have to encourage users to update to either or +Thanks for the rapid audit +Actually your patch is in a different format from mine and the patch command doesnt seem to know what to do with it +This would be a nice to have fix +I didnt mean to ignore it +Patch applied can you verify? Thanks +I changed to for scm +Sorry forgot to set the version +Does it work without the task executor? Incidentally you shouldnt use a task executor with a flat file item reader unless you take special care of a few things like setting saveStatefalse and restartablefalse in the step +Of course the Thread is having a strong reference to it so I doubt this will actually happen +all tests passed patch committed +The only reason is lack of time Please do add the additional functions +Bumping to critical so this gets fixed by the end of next week Feb the Property and Setter Method fields on the bean m still seeing the Data properties on the bean properties +This is what I applied to Trunk. +Thanks Brock +should be removed +implemented requested feature based on a new config parameterset use config parameter to true to use action message bundlesThe default behaviour is not changed at all so this change is fully backwards compliantcheersRainer +Closing this one as we have a new general issue for memory leak checking in the release code. +Regarding versioning IMO it seems cleaner to do that at prefix level +Lucene supports serialization explicitly by implementing Serializable +Id like to add unit tests benchmarks +Closing resolvedwont fix issues. +I think we need to be consistent and quote all of the separators +Hi Ivan I have attached an initial patch here +Will this fix go into EAP? EAP +Assigning to ALR for review +As far as I know all of the bundles produce JSON not JSONP so I dont see how this applies +Committed revision +Do i need to migrate to or fix is available in any patch ticket is fixed in and similar problem was fixed in in so should be finePlease create a quickstart application that reproduces the problem and attach it to a new ticket. +Done +It seems this failure was related to after applying the modified supplied by Marek Novotny this failure disappeared in Hudson +HiThis is already fixed on apache side during Middle of Sep next fuse side release will pick up this fixSorry I mismatch another fix in I saw this problem nowThanks for pointing it out this issueBtw wanna append a patch to fix it? tests pass on trunk from my testsSo were ok with the commit +Ive got a proposal for how to modify the format for it to easily support encryption +Right now this is accomplished by having the throw away the Pages instance causing the files to be +Is the problem that you dont know what the last column is? Normally an application should be able to insert a fake column thats always larger than any real columns +I first thought we could do even better by registering one servlet per endpoint +Patches were too broad and have gone stale +But we will likely modify this when using a namespace +Fixed in EWS +since Mojarra allowed me this I thought it was a bug here +changed a bit for beta +But why do we want createFile to throw an IOE in this case instead? Im to this if we resolve this in a follow on jira that removes the check and fixes up the test thats creating and is expecting an IOE instead of an NPE +Bulk assign all unscheduled issues with priority major or less to target +But still if for whatever reason they do not exist any more or if the paths are changed in the configuration file the user will never guess what the problem is as there is no helping error messageI am not saything that it should not crash or something +The code was removed from the observer method implementation but this had no effect on existing tests +Oh yes I forgot +Fixed thanks for your help +Thanks for patch Kannan +I dont see any indication of this problem going back as far as the current branch +I will take a look at it right away +Comments on the patch +its in the distribution +This patch fixes the missing payment type in accounting +If you finish it before we release beta please correct the versionThanks +I use own maven repository for latest development snapshot of iBatisAnd I doesnt use dont reopen the feature requests +This is a tradeoff as is more important than orderingHowever the current system has a bug which is that the http api is trying to use unicode collation on the endkeys when it should be using binary comparisons +Ill spend some more time on it next week and submit a patch once I get something working +attach a patch to implement a multiplexing channel manager +This is very minor but it enforces consistency between the schedulersThe more I think about it the more I want to see the ability to request only part of a core +Jeff Mesnil managed to replicate this on Linux +PS please excuse lack of bug details in the bug header +Does anyone at AlibabaTaobao want to take this? +Put that back to true and see if that not then there could be a bug in JTDS and the BTM pool does not know how to handle it so it leaks connections +Package is the first package that has children or doesnt have any configurators are called aggressively +I am tempted to go ahead and ignore it instead of making it an and suchbq +test and report so that I can close this issue +the task is permanent as these links should be updated every time when new guides versions are placed do we have links that changes on every release in the docs? That is a maintanence are those links? Cant we change them to something more permanent +It doesnt looks like was updated with this can you please update. +This set the PATH appropriately +it looks like we should hold until we commit this then backport commit toI agree with that +is incorporated into the URI specified by the next link relation + +Committed revision . +locations are in the docuementdocs is complete +Also prefetch is a good feature that I probably want to use so switching it off just to get messages redelivered to another consumer in the same process seems like it must be a workaround for a bug +So something is keeping a reference to database objects havent investigated it yet +Postgresql logging shows nothing unusual at the time the reindex fails to complete +Added steps in spec to update default local regionDefault local region will be added with name Local and endpoint httplocalhostclient +We were racing between the main thread and after interrupting it +Bulk closing stale resolved issues +Patch produced with svn diff on a Windows PC +Seems failed with this patch so need to fix +Attachment has been added with description has been added with description successful has been added with description failing problem also occurs when requesting postscript output but it does not occur when requesting png output +Canceling request for first patch which hasnt run yet +thx +Allows a couch node to handle ssl validation and pass ssl certificate to the replication for such nodes +A url can have inlink from different segments +Patch patch doesnt seem to fix the problem of absent genstamp +I think this causes problems with source generation plugins +any ideas here +So we need to check bytesRemaining by default +JDK is not officially supported +The feature will be proposed only if it doesnt affect too much the performance on if you need twice more time to launch I really believe that it is useless so we need to check it first +Looks like you have rather than +It should start withUsing these catalog descriptors +Small change +Looks reasonable to me +Shouldnt the same change be made for javadoc jars +The savenamespace command clears out the edits log +But still has problem to activate +Closing the resolved issue. +A framework was composed and a handful of companies implemented the specification and then found its way into projects +Krasimir what a version of your bundle +Patch attached +thanks for the confirmation sorry it took me attempts to fix this issue +My initial thoughts Im not sure I agree with not returning a response for a regardless of the count +Can I reopen this JIRAThe intention is to make transition from to easier +Factories are now functions and not classes +Okay I started with a fresh Fuse ESB and installed the eip example +fix is in place on both and trunk branches but needs a unit test for completeness +cant reproduce on with JBT +Vasily could you please add a regression testcase for this issue? I think it is better to add a testcase to kernel tests for this issue rather than to regression suite +I made those changes and then ran the whole buffer through a little untabify so this latest patch should be good +All the people I talked to onthere were very helpful in giving feedback on how to do it in a waywhich would be most useful to othersFrom what I have experienced if you need an open source library to dosomething the best thing to do is +updated in rev believe this work is complete now + +I havent modified the packaging part of the build +Patch submitted thanks a lot +However this is not possible Being present is specific for configurations but an optional configuration could be anything +Several individuals were disappointed to see this and were discussing it in a forum +I dont know the grails framework ist not able to evaluate the actionname +The patch looks good +The seems awfully polished to properly represent Hadoop +Ted the changes in cef look great to me +I will send it now and we will see what the community think +Regards merge I was saying that the help text on the historian page says that region runs merges +Patch looks good to me +Ill close this JIRA now +Patch looks good to me +Missed aom Test annotations on the Hibernate tests +Worked fine for me a lot Garvin. +I have reopened this JIRA +ah good catch +this should fix the Thanks for the patch +Its not just semantic its common sense +The left two warning in Findbugs is not a issue as getset primitive is atomic if the size of the variable is less that or equal to bits +Attaching patch which changes the DUPSOK queue consumer behavior to match that of AUTOACK +Decided to set up a separate compile task which copied the source files to a tmpdir replacing the token JBOSSBINDURLADDRESS with the actual bind address +The problems you find may be a minority in eclipse +Policy files do not allow access to environment variables +Basically the intent is to avoid the situation we had noted in where two different versions of the JPA exist +Do you also want to remove here +The attached patch includes the new gridmix versionThis patch should be applied to trunk and branch +This is very interesting you are definitely delving into areas never before tested +One question on the code I am looking at why is the run method of infinitely looping creating destinations? Arnaud +The problem has been corrected in the release combined with the unreleased EJB RC +Sandy have you seen this in +It should do this once for a given Configuration and then use to store this object +Made the changes and rebased +Thanks for doing this fix +Thanks for the extra legwork on this one! +The new patch now has test case that runs standalone using Apache mina FTP server embedded in its code +my earlier comment was more about avoiding to use Stream in more places than strictly necessarySure +We fixed this in +I can also declare a context file containing helper beans that do the work that Im trying to put in the s and have applications import this file +thanks for pointing it out arun +Included are patches for Trinidad and Trinidad trunk +When someone has the time to look over it and test it we can decide if we plan on including it or not +She specified the svn command to extract the front matter files +Updated patch to reflect ideas discussed on patch from Jason Porter. +Dont think it would have nulled but your code is clearer +Heres the translated labels +Improved patch +Improved logging in a few places +Thank you +I have run the tests shown as comments at the end of the file and the values are correct +Thanks Ashish. +Looks like a sensible suggestion but this also looks like one of those things that existing devs will do one day in passing through that file since they are not affected by itSince you see the problem in your own work any chance of providing a tested patch +Marking closed. +This issue has been resolved for over a year with no further movement +Attaching a screen shot of the Directory project group and the scm result +Regarding the DOC issue there will be a wiki for I just need to convert it to wiki markup +Wont main need for tracking be your Struct? How hard to add a into your lib +Perhaps we could imagine to support a new parameter to deactivate this behavior but in any case I think this should remain the default behavior +Closed per reporter request +This also potentially fixes an issue we have been seeing intermittently when a reduce fetcher keeps asking for task completion events even after getting all of them +So my original statement was entirely wrong a type with no analysis chain is perfectly correct it was entirely a problem with the code +Oh yeah Ive seen it thats why I wrote you to change exception +Im on the case +thinking more about this doesnt have to do with async caches but just the way two messages are ordered responses to get invalidation messages +Change looks fine. +Ok Ill talk with the guy who set this up for me next monday +What should happen if the editor is dirty and the refresh button is pressed? Best would be to popup a dialog Do you want to save changes? +Ok done! +I actually thought I had already marked it resolved but looks like I forgot to click the button. +ebdc on master branch by upgrading to Launcher +it is because your provided in your application a services file with a non existing file +I believe an ordinary pig user already knows about doing such things in unix shells +ColmOne thought we could copy all the non +I dont know the details of WHY they have to serialize mutations +If we are happy with this requirement then I will just close this issue as fixed +Yeah I think this came from habit and different people working on it +Ulf can you provide some sample code to duplicate this behaviour +Dylan the jars are missing from repolocal probably due to a +Any ideas? can add a new aspect to the test +srcmainjavaorgapachehadoopmapred +Testing needs to be done to measure the actuall performance benefits between these versions +Fixed in trunk and branch +I just committed this +Why would the name node not keep the timestampversion? Seems to me it will need it to disambiguate new block reports and such + Looks great for me +I have corrected all example Java sample sourcecode and most XML example code +Ive attached the image which causes the crash although the image is so simple that Im not sure how it specifically could be the issue +I can confirm this on my end as well +HelloPlease apply the patch run the shell in modulestextThanks a lotBest regards RichardFixed in TEXT module at repo revision rPlease check it was fixed as you expected +The steps necessary to take advantage of this app are in the main issue description +NewJavaNew Service Provider Class +Do you also have the XML of the WPS request in question +Setting locale in constructor and added test caseCommitted revision . +Upload the file for reference +Has there been any work done on upgrading to the latest version of Thrift? Recent versions of Cassandra use a newer version of Thrift +I thank we need to watch and make sure the split check is not launched on a minor compactionExampleSay we have a new region fresh split and it starts getting heave updates and memcache flushes +Unpacked +Jarek just find the reason why the Provider failed on the tomcat axisI will create another JIRA for it so I temporary removed the testcases +Id stay with a standalone interface for now and see how people react +Closing unless someone can reproduce this. +The last deadlock above should be resolved by the fix applied in r for +I will create an another jira to rearrange the coprocessors tests on the znode +Fixed in branch and in all resolvedfixed issues of already released versions of Roller. +This seems not an urgent issue I dont think we need to back port to +Will update the patch with Teds comments +There is a extra debug trace in the code that I will remove as part of the committed to the trunk +Pavel where were these examples being drawn from for EWPThe version in the repo is now correct +So one single classifier might not solve the problem +Adding explicit dependencies to the relevant cxf features fixes the validation problem although it shouldnt have been necessary +Looks like an issue in application of a bulk update all resolutions changed to done please review history to original resolution type +Jay If you can on a current JVM we would appreciate it this one is getting stale +The JSR early draft only allows XML charactes in JCR names to solve this issueI think we should adopt the same policy already now since JSR allows implementations to place custom restrictions on repository content +Thank you turned out that the bug is of Javassist +Will wait till tests pass before closing +Where should notifications go to +Its contained in the sfp in the sffooter as an sf +But as Jochen said trying may be interesting as we fixed some bugs in this area. +Ive created for the example rename + +Thanks Vemund! Committed revision . +Hiwould you like to create a patchwed be happy to apply it for gzip only starts to win on larger IO transfers as the compression overhead will add cost. +Looks ok to me as soon as you have a test or sample that shows it works we can look at moving it out from the sandbox +Hi again I think i found a possible origin for this problemWe have a dependency on another maven project which depend on the version of cxf +Someone needs to add more information regarding the return values and the fact that detached instances are not returned but values are merged from the detached instance to the persistent instances +Sometimes it just works sometimes I get the expected broken in Tools to add release after having added release notes +Konstantin checked exceptions can be used when a particular layer does not have an immediate response to the condition causing the exceptionFirst in this case I consider HDFS to be a single layer or module +Committed as r thanks for reviewing guys +I think this relates to the same problem in +for +Externalizable works seamlessly with native object serialization and Serializable implemented classes meaning it works with a number of existing Java classes in addition to Externalizable classes +Thanks Luke! +there is no duplicate icons as far as I can see in Betatrunk. +Now suppose there is some functionality which at best is an effort level of right now +I suppose that the debugger than takes over the F shortcut but I have not clue how to fix that +This sounds great +Seems that the issue was just test thing timing issue +I just noticed something in the chain +So no it shouldnt fail according to the current contract +The patch I attached is ready for review +more examples of behavior for reference +Wouldnt it just be simpler to include into the binary distribution? does that and I think thats the way we should do it as well +to assess new some of the issues in the patch +Ok the patch looks fine except for the namespacing our util etc +Heres the rebase sorry for the delay +Since Adam has asserted that this change shouldnt do any harm I will add the testcase and resolve this issue +Ill take a look at these broken tests +filterFunction removedclient is the right version +lib was totally there etc +plugins folder of my workspace I destoyed all +Ive change the guththilawritetobuffer method which is now specific for axiomxmlwriterwriteraw method +Uma do you have a why you definitely need this in +Well be looking out for the release +This also adjusts to note that Java is now required in order to build Derby +So if your transaction manager does not do this youd never see this problem +Please let me try to set the interval and run another testThanks +If you are going to start suggesting some kind of totalitarian rule over what goes in your repository then I think Ill be getting off this ride right nowregardsbenps +JSF version of the patch +Stephen It doesnt look like we got to this +The slight modifications +In addition to that a test method required by the lang is also included in the patchRan derbyall +Fix some corner cases relating to transaction +Ill commit this en un segundo. +Committed in rev and trunk . +That plus of versions restriction plus puts and deletes will make version management +I forgot to add the implementation for is the same as youve proposed with the delete action not executing a whole repo and db scan +I just committed this +Changing the issue name to avoid further confusion +As stated flushing w cache would be virtually the same but consume the memory +Thanks MartinFYI we now use this feature a lot to generate HTML emails and PDF files using. +Changes look good to me +to both trunk and +Use a virtual user table like pretty much every other mail server +Thanks AndrewI was able to reproduce it +probably checks +No it should use the Marshalling factory methods instead +There are validation errors too even if the is fixedSeverity WARNINGSection Description Invalid forCode Class Property failJoin Type booleanSeverity WARNINGSection Description Invalid forCode Class Property Type Severity ERRORSection Description A must have a default constructor Severity WARNINGSection Description Invalid forCode Class Property sleep Type longSeverity WARNINGSection Description Invalid forCode Class Property sleep Type longArchives from the distribution comply with the spec. +I will try to reproduce our issue without and with your patch as soon as poosible since I am only this week this may take until next week +Done +Whoops yes +Done +Committed to +Oops that rationale slipped my attention even though it was explicitly documented thanks for fixing that! +An interesting of the proposed changes is that one can a Gant instance with different scripts +I believe the patch can be applied already +Mac OSX Mountain File in JBDS Help Check For just noticed the stack trace states in some of its listed that actually on the file system +Ill continue to debug the issue further and provide a fix if possible +Closing this ticket. +Also tried with Groovy and +This behaviour is changed to log an LOGINFO message for all unsatisfied referencesIn addition a LOGINFO message is logged when a service is registered which causes a component to be activated +the parts where configure checks for zliblzo headers and so on +cs file which lives under the +Hey Alex please do not reopen an old issue but file a new jira instead +Im sure we can do something better and I continue my investigations +Thanks +Something like this would work as a just the servlet +If its wonky it goes wo saying that you didnt do it +I have committed the patches Joe thanks for all of your hard work in this + issues to Nick and had to set resolution of all of them to done +Wilson please watch your tone and remember to treat other developers with respect +Does the Derby store behavior change in anyway using +Is this information required at all for the response generation? Else this may as well be kept in the +same issue +via a testcaseI dont understand what you want the alias transformer to do when there is no alias is has to have one for setting it +Patch and committed +Fixed by +The movement of the assertion while redundant is the correct way to check this +New file +Implemented without immediate full build after modification of +Creates a Component named which needs to be started comments it looks like the classes are calling just getRequestResponse which returns the wrapped Response which of course cannot be castSo we have to overwrite the getResponse methods in the adapters to ensure the baseclass getRequestResponse mehod is called which then is the RequestResponseApplied a fix in Rev +Maybe it is fixed by a newer version of graphiti +Should we call it something other than to prevent confusion +I dont want to wait on tika to release +This issue should have been fixed with. +Checked in svn to resolve this issue. +another way to handle it is perhaps to have a flag that is off by default and that enables last read precision inheritance for subsequent writes which is what the author seems to desire here +Several of the failures should be fixed by the duplicate cleanup that was done for ER +Please have a look at the generated code now uses a utility method in Axiom and this method should work even if the underlying parser is non coalescing +Mine for service only and will require many more changes to make it work for bundles +We can either apply Nonbinding transitively or we can say its not applied transitivelyDiscuss at EG meeting +Great +It makes me sad to introduce a schema dependency since that means you cant run sstableloader except on a cluster memberWhat if we added key count and cfid to the statsmetadata sstable component +The current poll finishes but a new one does not start +Anyway you need to create it how else will you refresh it +here is my ve uploaded my key +ThanksBest regardsAndrew +Thats it thank you for the tipBut nevertheless its a strange error message +This benefits the increment operation tremendously especially when there are hot rows +Thanks. +Initially I used for the implementation but was having problem with JAXB attachment marshaller in the server side response to the client applicationFrom further debugging the JAXB Attachment marshaller uses when marshalling the byte array request from the client application and hence for reasons I was not able to determine the jaxws attachment marshaller implementation using was failing on the server side response message to the client applicationApart from the fact that JAXB attachment implementation uses Also from reading the mime part It seem that the usage is the best approach for messages with mime headers and attachment. +EULA +Closing old bugs. +correct spelling of eclipse release Helios not Halios +Would be good to have a test case. +James you are right +Im not sure what to do about this I dont see it fail +was committed so this is no longer needed. + Example of file of your web applicationsuggestExample Add in the file of your web application a specific listener classsuggestAdd a specific listener class to the file of your web application +Test plan +This is highly undesirableThe solution for project roles are currently available is great +In the issue that you linked to jlage implies that he experienced the slowdown only after switching to annotations +Closing xml and console issues. +it seems like I just need to get from the database object that my game logic produced back to the session that mina is holding for me +BTW sometimes this test just hangs most probably because of +Calling it a string implies there are semantics beyond a bunch of bytes for which its the applications responsibility to derive any meaning fromCalling it a string when it may or may not be is holding languages that do have real string support hostage to those that dont +We are in the same situation +Verified at +Poor users +Attaching patch +Seems like straight forward porting to me +Bulk change of Fix Version for all issues fixed or closed for +This is a delta patch and goes on top of the latest patch that is already uploaded +I have updated the mentioned files +Also there is no need to include the public keyword in interface method declarations +There is really only the minor change that Thawan has requested everything else is the same +I am not familiar with JSF core code but I dont think it is implemented like that intentionally since it seems not reasonable accepting List as value but failing when trying to get the selected values +This depends on +I think this needs to be exposed along with other metrics about what is going on +Any hints on this one? +Heres an LDIF export of the directory contents so you can see theres not much in thereversion dn uidpbaylissouusersdcexampledccomobjectClass topobjectClass inetobjectClass personobjectClass organizationalPersoncn Paul Baylisssn Baylissuid pbaylissdn ouusersdcexampledccomobjectClass topobjectClass organizationalUnitou usersdn dcexampledccomobjectclass topobjectclass domaindc a blocker and need to be investigated fast +Checked the code +It was applied to and +Also the use of SBM should be attempted but SBM will not have entries for test beans +How does the database URL look like? Is the Derby server running +If someone wants to order messages they need to enrich the headers explicitly with some business knowledge +If I get a Conflict response from I assume the database rejected the request but its not the case here +Not a bug +Do we even need to call encode in? From what I understand encode is primarily for ensuring one thing That key value pairs encoded in urls dont get messed up by user supplied textSince this is sent over post in json format it shouldnt matter right? Cant we just skip the uriEncoding step +I will wait to other comments +Just a matter of exposing the position from the writerreader +Ive tested AJAX with deactivated Cockies and it looks good so I close this issue. +This shaved a few seconds off the total test runtime for a typical test run +for the response +As a hack trying switching back to a saved version or the original thread context class loader gets past this and the from call works +is badFor +Committed the patch after all unit and integration tests pass +srctestresultspositivesvn mv srctestresults +Hi Matteo thanks for the patch +for the patch +I have allready attached sample code by creation of this issue +In my case the isnt called for some reason +Patch is ok to me now +sleeps with a bit of random jitter I hope possibly even exponential s a patch which addresses the issue +It is not a wicket bug +See mail list thread for details. +Many Thanks +Thanks for the review Stack +Yes one of the reasons I wanted to discuss these patches is that they uncovered some of the underlying ugliness +Im going to push this back to +This seems to be a hitherto unseen corner case that is not adequately addressed write is not always interrupt safe +We could either extend on of the Map implementation directly or implement Map and wrap each callI would prefer implementing the interface via composition instead of inheritance +Merged into for quick fix! Verified that its now working for async and also JDBC cache store so I am closing the issue. +done +I will update a patch soon +The NPE is not seen in the JON build an error message is displayed instead +The mutable objects should be ready to track addingremoving children and addingchangingremoving properties +Areas of improvements +etcI dont know why you saying like that +its the cleanest way to be correct w gets wo introducing explicit locks +The importer that this Jira refers to will interrogate the metadata served up by an service and will be custom tailored to any generic server including SAP Gateway +We can borrow some design techniques etc +price USDIts not pretty to look at but it at least would be a correct solution by removing the dependency upon locale altogether + behaves differently only because it was easier to do like that +ill need to think it over +Thanks a lot for all the help Matei +html for this fix looks good to me +Here is the simpliest patch fixing the current directory implementations to use the new +rar i attached which shows proper nesting for your usecase. +Found a bug where a path without a slash will not be prefixed with +Attached unit test which demonstrates the problem +However noone has ever actually asked for itTry building from source with this modified to use a smaller limit +If your email is urgent please contact Rick Hobbs or Nigel Atkinson on otherwise I will respond on my returnThanksNicko +Updated old exception message to match that of Brock Noland +resolved new runtime exception is being thrown to publish the error upon! +I believe that U should work here or configuring updatePolicy to always for the repo +has no such problems +Yes the test was wrongSee attached correct unit test +This patch fixes that issue for us +and this just to have an lazy iteratorThis sounds kind of weird to me +Latest patch looks fine +The problem still exists when you validate the XML sample document using the Xerces XML Schema branch and use the schema version +Ill commit a fix if everyone is ok with it +This looks like it could be the same thanks for spotting this one JasonIve added your test case to SVN as as well as resolving the issue. +I dont see anything special about Opcodes +Binary with Konstantins patch applied +You will have to provide a Roo script so we can reproduce problem is that there is no example when I want to create a new controller it gives me this error +This script will upgrade all the existing bridge name to new bridge name +This patch changes the id in the xdoc parser to macroId and the id in the snippet macro to snippetId +to Charles +Implemented for +Modifying the sleep logic to use works fine as you m going to run test suites and see how they look but this is probably going to be fine +Maybe for debugging purposes at first view it is enough to know the submittedValue and localValue +Verified on revision . +Patch is applied at +I have looked again attest just in still fails but the failure is similar to the failure ontest are passing +username MarkI was wondering if we can try manually editing the current export to say its from in case that works +this one really has a huge impact on development performanceworkflownow using Eclipse JEE Package meclipse am trying to upgrade to WTP if that does not fix the problem I will downgrade to Eclipse installed Eclipse SDK WTP meclipse everything works fine again +Committed to trunk +Fixed with revision + also shows this will disable the tests for now so please mark them broken locally until we do another TCK intent of these tests is to validate that a library installed into lib with a is available for injection in the application + assign and schedule +I guess it could differ by size of splits and number of tasks +MaxHow is installed? Is it being installed into the same directory as your eclipse install? Or is your eclipse install directory and being installed in a separate location +Andrea I think this may be happening as a result of changes to the custom converter stuff we were doing before +m works and Ill close the issue then +jo no posaria or so ja que escaig sempre vol dir que safegeix a la quantitat esmentada en canvi or so vol dir que shi acosta tamb per sota +Can you point my to the UI classes and jsps +If we extract the name from the middle of a normalmixedCase style Java name then the name will by default begin with a capital letterJava programmers are accustomed to having normal identifiers start with lower case lettersVigorous reviewer input has convinced us that we should follow this same conventional rulefor property and event names? IntrospectionThus when we extract a property or event name from the middle of an existing Java name wenormally convert the first character to lower case +Its fine for compression but I suspect that more awesome compression will come along that will make that API not be reasonable +bq +It might be a karaf issue. +Well actually a variation on the patch but the overall effect seems to be the same. +Radio set workspace file system has to be in second column aligned to left and buttons Variables +I did not notice my configuration was using expressions +Is someone can and Anty Rao into the assignee listAntySeems we should add junit code and make patch for a right versionThe bugs i reported are very clear on a glance for anyone who know the nuts and bolts of mapreduce for the attached patch is also very simple and straightI have done A LOG OF tests comparing the overall mapreduce performance in and +The problem was caused by the data buffering +Updating Fix Version for tracking purpose +Canceling patches until concerns are sorted out +So I still pretend that this is not a bug but an enhancement +I was working on reproducing the bug but wasnt seeing the error +Adding dependencies to the profiles section would work arround the issue with osx and possibly be generally useful particularly if there was a library of profiles dependencies for different platforms. +The original goal was to deprecate in and remove in correct +Rakesh Looks like there is a findbugs warning +Regression tests passed without failures on Solaris Java SE +branchesJBESBCP +New files are reattached +However I wasnt sure if we should do that right away in lieu of the current numbering scheme +This one fixes yet another race where a ZK node does not disappear in timeI think these changes are uncontroversial +I take it that is exactly what your patch was supposed to do? Pleasecheck and if so close this one +Could you post your complete xsd file so that we can reproduce your problem? Thx +Although this implementation doesnt give uniform performance it is better than the old implementation or the application doing its own implementation +There are no activities for the bug for a long time Im closing the defectThanks +Oleg +Well set up Bamboo on an internal server +Petr you can take a try for the jboss tools nightly build +You can verify it byAccounting select the Asset Id from the above depreciation generate the report to display the depreciation from acquired date to current yearIf every thing is correct then we will implement other depreciation calculation servicesThanks Santosh MalviyaSumit Pandit +Have you tested the app? No Problems here on a Windows professional SP system here +The issue seems resolvedOf course be free to reopen in case something is still to fix. +I like option the best for a final fixThat being said it all depends on how long you think opt will take you +I believe that we should try to make friendly for backward compatibilitybq +I then added a new mapping without the suffix so my application config matched +Koen do you mean when you run forge in the same VM or when you do not run forge in the same VMRunning CDIForge classloading tricks in the same VM as Eclipse has serious implications needs to be extremely careful that no classes gets loadedaccessed in both forge and eclipse to avoid alot of problems +This bug has been marked as a duplicate of +I added as a dependency of the project and it now compiles without complaintI think having the AP silently stop processing would be best +Patch looks good +Another option is to add a settings dialog to ShowHide particular groups of widgets +If the user was created by a system admin the account should be validated by default the email validation should no longer be sent +Lets also disable pmem no point in doing it for tests +If I use element mode I am not going to see a class called Trade I wonder how it would work but in any case let me try to write the test the way you have it +Unit tests and have passed with +However if the state format does change between versions and we end up only supporting a conversion from the old format to the new format then that would be a case where we support a rolling upgrade but not a rolling downgrade between those versions +Anyone has the same experience or ideasRex +Max code completion is ok +I reviewed the patches and they are good +And its easy to adapt et al +bq +This has been working for a long time I think by M +Fixed +In other words it would seem that this change breaks the builds of everyone who had followed the doc +Dont know if it applies to too +Why not just Mavenize the project and be done with it +My suggestion would be to throw an exception and fail if the resources already exist so that people are not blindsided +Instead of creating and closing could you stick one instance in insteadSure np new patch coming soonThanks for looking at the patch +Thanks for the patch Madhukara! +build has revs in it +Here is a patch with unit test to fix it +updated datafile +see previous Ive seen in source code of it always removes independently of configuration. +Thanks Francis +Me and Max cannot reproduce the dependent issue +This has been resolved +Heres a first pass at this +Revised summary. +Closing old resolved issues +Hi Bhanu Can you please enlighten us on how you secured this service? As it seems this service is not deployed in Axis right? This policy is not written according to WS Security Policy specification +Robert +Here are the code changes tests incoming on and! Yes it does funny I searched for a issue before opening this one but did not see it +I am reviewing this patch +The behaviour in the new version is that service channels work in a style +With this it works seems that the other workspace i used has a problem with an old install of roo? I think Ive used that workspace with an older version of sts too +Ive committed this and merged to +sorry +changing this to because only uDig will care +I just committed this to trunk +Committed revision Thanks +I did not change the priority for issue because it had already been solved in the branch which is targeted to the EAP CR. +mike pumphrey is reviewing +Btw was this happening from the UI or by using the REST configuration extensions +replaces andThis new patch contains the revised license headers and makes some improvements to the codeShort intervals with a calculated height of no longer render a row to prevent double bordersIntervals are now sorted internally and therefore can be added in any orderOverlapping intervals are truncated to ensure only one is rendered at any particular time +Should fix the problem with multiple escapes by using +New patch fixing to track initial past end +Thus some of the ws stuff will start working with the dispatch clients +Closing in favor of. +I think the user has to be responsible for uniqueness in that case +Reviewed and tested by me +This means we should also be able to specify a maximum row length +Any update on a patch for this +This does sound like incorrect behavior to me +They all are DD related +I am definately missing somethingjust dont know whatI hate to ask you to explain ant task as that seems like something basic but I do not understandThank you so muchBrian Nettles +I dont like thisThis class has too many methods even now +Hi Jacquesship doesnt have any style specified +We cant reproduce it on provide additional your environmentAjaxjsfFacelets provide your build version +In the regression test you find a lot of tests exercising the features you are interested in +At the very least it will validate invalid utf which is impossible to test completely but it should not blow up for any valid utf + +The only way they can lock things down is setting up firewall rules on every Solr node allowing traffic from frontend and the of the other Solr nodes +Do you think that perhaps why I created an issue in JIRA of type BUG? Hmmmm +change the name or package of the class in the schema to make such a change easy to detect +Id thought that on replication of local source db queries alldocs list or similar but he is not. +perfectly fine +Im using JDK and Scotts with G and s with and without the attached patch +When I say move I meant replicating to the destination and removing from the source +This is the only one with the constraint not null that why this one doesnt to +patch applies fine for me on a different machine +see for comments +I retested this with my mk test and now xsltc passes agreeing with xalanexcept for indentation +Attachment has been added with description Sample containing a simple and its PDF result that cannot be verified correctlyThanks for reporting! This was actually already reported last week so follow up at +We should not always create a new if its not changed at all +Would be nice if the gateway could be more intelligent as to how many actual executions to expect +And good point about several bundle caches having different versions of the same content +Appliedthanksdims +It is up to you if you want to do this for some reason but it is not good to force everyone to do it +I see no way to get the right connection or even the before seednext is calledCan this issue pls be reopened +The issue went into and + added the job counters to the Web UI but forgot to add it to the Web can you please make sure this works for the AM also +Im not sure its its clear what the implementation of CDI will do in this case +a lot for the reviews Eli and Todd + in Jira add the Hadoop PMC chair to in svn add the Hadoop PMC chair to DoneAnd two more for the queue update mailing list names in hadoops website update wiki with new project domain name and mailing list names +BJThis is because the xsd file points to the last version which now use DBCP in place of Miverva which is always used in + +I guess one question I have is all of this required before acceptance? It was my understanding that these issues just need to be resolved before release right? GrantThe requirements for notice and license files are after the code is accepted and before release +Thus this is clearly a bug +Ive tried running the attached patch on both trunk and branch without the fix in but the test doesnt fail +the API can be done by adding a type as in to allow a single unified batch op +Attached first cut patch just moving the +Ive got to look at the code but I wonder if this should be a shared instance that is reused the cleaner uses a thread that has a special call related to shutting it down +Ive linked to this report +Attaching v of the JR Query test suiteIt is and contains the references to all the created issues +Since it just updates the text in an assert statement for one test and gets it passing theres no need to run jenkins. +anyone who wants to do this in can just take the two mapreduce classes and include them in their mapreduce jobThat could work as long as we put the necessary hooks into itself the small changes to the existing classes +gwtplugins seriously dont have this file in my trunk +No semantic change from previous patch +Build demonstrated both and of changes to avoid spinning when the root server goes down and that is the only region server synchronization in general +Mass closing all resolved issues that have been closed for more than one month without further comments +What would be the best way to fix this +Does this happen with codegen service or just POJO service +Victor can you look into this +Una altra cosa en informtica i en electrnica coax com a substantiu vol dir coaxial referit a un cable de coure +as its working on other versions. +lets deal with this when gets reported on a FF GA the same error with Windows box using FF and +it shouldnt leak memory right? given that the cache eviction listener in unreferences everything that gets evicted from the would be nice to have some kind of optional leak detection finalizer but in the meantime youre right that this is unnecessary and a performance hit +I forgot to mention my patch also include a table resource leak fix in the where we dont close instance +So all we need is to upgrade the lucene jars in the solr lib rightIf there are no objections Ill do that today +The above sentence is for network offering for +In addition to a quick code review of the bits I was interested in related to fetching map output fragments I did a quick and dirty test on trunk on a tiny cluster to make sure that this change had the same effect as the fix I apply to on production for similar benefits +Defer past. +My last patch removes that as per s suggestion +At this stage we have merely promised to maintain wire compatibility going forward +I think there is a relation between these two bugs +Florian please review ajpush page in order to be sure that It fully covers functionality requested by the community and introduces real will check the Developer Guide to be sure it provides enough info to understand ajpush +OK thanks for the links +you wouldnt be able to jump from page to Im convinced now could you submit a patch w your column offset code +Makes sense +looks +Patch committed rev eaafadceecccac +could you work up a patch but on not +The only down side is that the with this kind of scheduling the outbounded pools used for other operations cannot be used here +Ill keep you like Marcels idea to use the creation tick to identity the base index reader without caring about the index reader hierarchy +Would it be possible that Hibernate Team works together with Ehcache team for this problem as I am following the documentation from Hibernate to enable the nd level cache +Thanks JLR that was it +Ive just committed this +I waiting for a resolution yet +Well monitor the builds and attach traces for the pessimistic test the next time it fails +I havent seen issue or complaints on the mailing list that it was too difficult to understand how to get started either +Patch applied +Changes are in release +Fixed and added test caseI lowered the priority because critical for us means something like doesnt work at all while this is clearly a bugThanks for the Jira post this was really a tricky bug +With that being realized this issue could be closedPierre +attaching junit test +Just reran this and you are correct the no values is working +The only question I have remaining about the change is the change in definition of numItems and size +sorry +If there were later revisions you may be able to patch them back together from the commit +The developers are busy at the moment +Thanks for the catch Dan +Ill try to use a time to look at this and review by monday morning worst case scenario +Im unable to apply the patch possibly because of the line endings changes + +I have the code done and will start on the tests +Mostly looks straight forward and passes all the tests +Both of these are now on the and branches and a new is being deployed. +I think Raghu is right +Please revert it and so if we ever do decide to port and others it will not be so confusing or difficultThe same goes for +Closed upon release of Hadoop. +Perhaps some addition information is needed on this +This affects both the eclipse plugin and the users +I add to the helloworld example homepage got the same stack trace but just once +Can you share in a comment how you do that? what do you skipI am interested in the result +The patch has changes for in a new package +Also your changes to wreck the entire file +does this mean that we now always create a? Are there any cases we still want to useAlso did all the unit tests pass +Thanks Eamonn +Im thinking of making it so compactions go on but not splits +meanAll I did was run the goal and it didnt produce the md +The proposed functionality is not necessarily a superset of the existing authorization functionality but subsumes some of the existing functionality +Eclipse crashes on startup +The executable should consider whitespace +I committed this +Added code to that tests new method +Also added junit to test new changes from Wills should be ready for review +See resolved issues for already released versions +Installation for Upgrade went it to Radhika to document extra steps needed to upgrade CP on centos needs to be documented so bumping down + +I dont think changing the output from sselectItems based on its parent is a good idea hence this jira should be a list populated with objects that have a attached to them or attach a Converter to the comboBox itself with the converter attribute. +Thanks SianPatch applied to LOGGING module at repo revision rPlease check it was applied as you expected +Thanks +Triaged forFor Derby and later the same failure is seen with the client driver so Im marking this as a regression. +Tyronneany update on this +There is only one committer to my knowledge that is able to look after patches that is Vincent and he is the lead +I always like to have fewer jars as possible +Im going to guess that youre running into the limitation that we dont support interface inheritance at the moment +Attached patch fixes not to copy VT and objinfo files from the source object +Not planned for +Thanks alot for all the help. +Its what happens when you rotate the logo by pi +FredI think you are more involved in the archetype stuff +patches the build system to understand the new source locations +If this is not one of our goals then yes the embedded schema is a lot simpler +I set this back to unassigned. +Heres an updated patch that should work correctly +bat suffix +Feel free to switch reschedule if you feel appropriate +Attached files for the related the resolves the correct libraries when is not also listed as a dependencyWe have been able to reproduce this across several computers in our environment so it does not appear to be an environment problem. +Note that my previous comment was ill thought +Local tests show these now passing on +Your second patch looks reasonable and the help message is OKI was thinking though that each connectors have a target and the root file invoke all connector targets +Will be back to work on this for rpc after little markers from end of blocks let end instead be found because length was provided elsewherePurged Keyvalue from protos +I think the default setting in your on install time should be gotten from locale setting of your computerI agree with you this is not a locale issue but a language setting issue +Also there seems to be a reasonable workaround proposed that would not require Lucene to handle mock empty documents to preserve the other doc +I didnt do this mostly due to laziness +includes the remaining bunch of header files +While it is unfortunate that there is this limit on String literals in groovy it is much worse to get this kind of error message as opposed to something like String literal too long +Right update description +So this doesnt affect me any more although it may still be a bug +This patch adds an interface to that is used to initialise an +Sounds like the right solution is to add a apply the way we do for normal Mutations +before version working with JCA and +Adding RB link +Eli I actually used to create the diff +Ill change those specs to match on value instead of object identity. +We could move the statics to another class that is instantiated on a per name node basis +You can also see this if you look for tests in the patchbq +Please let me know if you need any help in that regard I can collaborate with youThanks Robert thanks for offering to help +Ill try and see if I can replicate it sometime this week +you have to have a variable in a rule to reference it +Its a pathological implementation wouldnt it be better to provide a more useful example rather than one thats explicitly not meant to be used +I am syncing from the trunk and trying it out +I am running to see we dont produce too many javacore files +I would say that once a new update center is added the list of modules should update. +Committers can you please review +testing by submitter and no reports of problems I am comitting this patch +I might add that the disk is not full and in contrast to the original poster it doesnt seem to be a threading issue +Ive submitted additions to s but thats not likely to be incorporated any time soon. +Will have to remember how I was able to generate this exception I think it was because I didnt have all our Writable classes in the classpathIll get it together but may need help on how to create the test as I would have to create a file with a custom Writable key but then remove that writable class from the running test and try to read the file +correct +It works on the prefix of the classname so if you have these Excludesorg +As you mentioned if it is a typo error and a required jar is missing the program may getSince is a system level configuration we should not fail all jobs +unit test +Im against the change you did Trustin since it removes the ability for the user to control SOLINGER behavior at all +ThanksBut please dont mark the issue resolved until the patch has been applied to the repository +rename +The waiting interval is about of the assignment time out +But in debcontrol they are set to Depends +I just committed this. +Very true Vikas +Did using NULL for the readers schema work for you +TODO Same as st TODO but for EJB with in the case of EJB because all this logic happens in the proxy itself theres no chance of putting something in between the proxy itself and EJB because the failover and cluster family update logic happens in you can put an interceptorin between and +So I had to implement our own version of to handle the tab indentation action +More time likely +How is that? If bootstrap fails the new node will not join the ring +I dont have an account for that repository +Ive just wrapped the exception with to fix the problem you pointed out + iBATIS test suite results with changes +Could you attach the Eclipse crash log? It is a file named where XXXX is the PID of the Eclipse processIt is in your working directory or in the ECLIPSEHOME directoryCan you reproduce this issue when using the VPE editor +I dont think we should put docs there in those classesin trunk you can use however many bytes you want for thisso the text would be wrong anyway +Then copy the war to the servers deploy directory and point your web browser to httplocalhosthello +See and + are mixing different version of Struts and and any other related libraries +The previous patch went stale it was unable to automatic merge +Tomee snapshot building process seems to be down the last snapshot is of +I have to agree with Stevo searching and not finding anything is not an exceptional situation +NOT CRITICAL for +DonaldI have found that IBM support PKCS see end of appendex A of httpdeveloperworksjavajdksecuritysecguidesjceDocs experience with PKCS keystores +Yes it was fixed +So there is no regression +The patch itself is quite small and simple too +Will run more tests with hadoop and submit new patch +Is there a way to rollback without closing? That would seem far more useful +I also added an explicit dispose method to the deferred responses in the event that the caller wants to clean up resources yet abort the receipt and processing of the response data +Applying patch and successfully run the described test and the included sampleId wish for automated tests but this seems impossible for now +Most solutions do appear to have some outstanding issues so I hope Yongcheng can add the appropriate configuration and documentation of the behavior of his settings so people can understand the consequences +By using the X switch i get all kinds of irelevant info. +moved to +batch transition to closed remaining resolved bugs +I need to look into something else +Completed the link list +The scripts are already in a location +is a patch incorporating the comments from Eddie +This is a duplicate of and +Patch committed to trunkThanks Woody! +Since the topology has effectively not changed all you need is the token gone but if you removetoken it unhelpfully tries to restore the replica count needlessly +I need a review given how this patch is actually revert of two commits Ive made one recent and another a couple of months ago +Once a user gets that exception once or twice he or she will know how the works and so can ensure that getters are available. +bq +Heres a modified patch +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Attaching the from which the layer is being generated +Ok +Markus dont you have space in the path to your JAVAHOME or your KARAFHOMEI tested on Windows XP Ubuntu and without problem +Is there a similar round up in the actual allocation code which may cause additional containers to be allocated to a queueChecked this before +Updated on trunk revision +I dont understand +Neat I didnt realize Java had this capabilityWhy couldnt we catch the and rethrow as a like we do the other runtime errors? Or is that not what you mean by Unchecked cast which might be ignoredSorry about the other test failures I should have been running all of them +This is a test case for this issue +This is workwise basically a duplication of the EJB spec in this area +What are the example queries in subqueryFlattening that involve flattening of NOT EXISTS +This patch includes all the previous onesNote It includes the functionality solving +bq +Null is ok I just given a name for testing +Currently the version is always passed to that function so the default never gets used but I guess we should update that to default to insteadI think Id rather not have a default so its always explicit +If I read your proposal correctly for launcher jobs you will set to and +Hmmm +Running test in Xint mode against debug build of DRLVM may lead to throwing assertion Ive submitted +this is very cool contribution +Thanks I can confirm its working +Added checkbox for this +use and the user forum in the s your mean? to post this questtion in user forumwhat about? I need more information +At compile time transitive dependencies are in most cases not needed +In most cases this shouldnt matter but there might be applications which expect +App server users requiring transactions over RMIIIOP will need to download and drop in the JTS +By the way for the latest Tomcat codes are not imported I reverted the uncomment createWrapper method on my own machineOnce the deployment could work we will be able to try more things +I will integrate this tomorrow if there are no objectionscomments +I uploaded my signatures last Thursday but Im still getting the same error +It should be put under provisioning directory in VCL +Please close +The file adds junit tests for the inner class support +The toc is not part of attached but it builds ok in me sandbox +from the current trunk revision r it works fine. +BTW +we need to have a better workaround for this in CR as it stands now AS is released enforcing that jars cant be both on manifest and in lib directory for this case +If there is still any code please reopen the! +If merging forward is a pain let me know and Ill generate patches for other lines +Sorry Brett blew by the first comment +OK +Lots of exceptions I didnt bother to chase upSo I guess this should only change with major version changes in Cayenne +This type of can already be achieved using +Yes I will fix it in next milestone +Build log is ostensibly successful kernel panic at runtime for +if im wrong sorry for waste your have done the same tests using and results are similar so performance doesnt depends on glassfishOf course using Inject Interceptor not EJB Interceptor +Name follows package name which I believe is the base platform interface definitions for the Jini starter kit are in the +Seems that nobody is interested to bring this further. +Fixed all other operations and checked that the responses were valid +The UI doesnt send the updated vlan in update API call +fixed in branch rev applied to trunk in rev +Patch on top of and patch using tables generated by there any specific reason why the checksum update functions are in the headers +assigning it back +It does affect NRQ but to fix NRQ correctly half open ranges must be modified to end at Positiveinfnity but then NRQ can never matchIn my opinion is made for and the raw method is an intrinsic we should use it +sorry Max I cant understand the views you refered +Thank you for immediate response first of all +v failed after iterations for me +Hi Davidthank you very much for your feedback +If not you need to get the library into your project +Use the JSON plugin. +Please leave Incubator management out of this issueAs per this issue description these pages were a very useful resource for all projects but now not +Waiting until is released before merging back into master +So as long as you have Named annotation on your post processor it will apply to all beans +Weird +Patch updated with APIs look right on this +On the other hand having different layouts for the different deployment vehicles is pretty broken too +For maven there is property. +We have one column that holds a byte id that is written frequently. +also there are a lot of tests in the core that use formtester so why dont those exhibit the classcast? this is why we need a quickstart +FTR created because in the is misleading with regards to the classpath support. +A test case +Bulk defer all frontend issues to +I just committed this +Done by Humbedooh! Thanks! +Once my changes have been reviewed and everything looks good Ill move the changes into TRUNK + +junitpickseed is cool +I should note that it looks like you are using a patched framework while I am testing on the trunk +Fail +Oops thatll teach me to not even compile before submitting +Well thats my point of view as a user I let the developers choose +Thanks for your understanding. +I appreciate the helpI agree that has lots of benefitIt kills the tasks quickly before the task starts doing bad things and it is more predictableBut also wastes some memory because the processes does not actually use that much of physical memoryFor example there are tasks acquire of virtual memory but use only mb all the timeIf you cumulate the total the memory that is not being used is a lotThe other main reason that we want to do this is that per task is an API change for our usersThe jobs they used to submit everyday may failed and they have to learn to set the limitThe good thing about this patch is that it is backward compatibleI think it is may not be that bad that we kill the taskIt will be scheduled again on other machine and the job can still be finished +Same issue exists in many other templates +Fixed with zk +Please verify whether the problem is resolved as you expected +Hi Ashish since the testsuite only runs successfully only windows we decided to turn it off by default so as to not adversely impact the build for Mac and Linux user +Nellya Ive fully rebuilt the classlib and now the test starts working for me after your patch +Are you disabling it simply to reduce the network traffic or does it not work as expected when left on +Replication log shipping thread gets CPU time and ships all buffered log to the slave +I used the version installed on my Mac +feel free to reopen or create new issue if you got more details +I cant reproduce this issue +Very thanks +Let me try to reproduce it on fresh generated seam projectSo far I tried on plain web projectWhat version of Seam are you using +Applied at from must fix the problem +Yes the typical result is that youll end up with downloaded data filesthat are multiples of the buffer sizen in this example +Screenshot of this bug +Yes the javadoc is wrong +I dont feel overly confident with the second one but it seemed to workThanks for your help guys +JrgDid you only change from tmessages to hmessages or did you also switch from the core jars to the JSF RI jarsIs the problem only with tmessages or is it with both hmessages and tmessages under +Closing the at the end of each test fixes the failure +Will be in ER +Punting to and assigning to myself +Stated differently the current row lock feature can be implemented on the client side through use of an atomic compare primitive +The computation of the space previous and space next features in line only considers a space character as a whitespace but I think all kind of spaces should be considered here +Could you please check your logs to see if there are any other suspicious traces +Dirkwhats your test setup configuration? I suspect you are talking about doing sort of integration testing having defined the providers in your? Can you plz provide a stacktraceTx loaded by fixtures deriving from dont get registered with the global +You can try turning on so on the tcp transports by appending sotrue to the transports specified in the Failover transport URI +Bruno is a manifestation of this problem +Closed after release. +Looks good +Does will solve this issueNo such calls will produce even more invalidate calls +But I guess there could be a some tangles in thereAlso we actually have a number of converter classes in +patch looks good +Bulk assign all unscheduled issues with priority major or less to target +Attaching propsed solution as stated in propsed solution as stated in you provide a patch rather than the modified codeHi AndreasNetwork restrictions here mean that I cannot provide a patch +That brings the run time for down to a few seconds +Applied in rev +What about a normal slash? But this isuncommon too +Only two bugs are actually fixed between RC and +Listing nodes +. +FWIW I cant reproduce this with and thrift HEAD so good that this was resolved not an issue +It also avoids sorting the spans twice +just committed this. +Thats the wrong URL +Im going to commit this momentarily. +I have committed this change as of revision +bq +I have attached a copy of the fixed jar if anyone cares to upload it into previous comment +Looks good to me +In this patch Ive changed storeApplication from blocking API to API +committed +I think its ready. +Put a form in your jsp +Thanks for the patch Jon. +I keep to the joda methods +move like this was actually resolved long ago +The approach looks alright to me but I suspect this will result in updates in most +Applied the patch and unit tests thanks for the contribution +Marked as blocker since if we keep clean in here we keep having slow startup could you replace this plugin +This jira is closed as part of a review and archiving of the aj jira project. +HiApproach Added the finally block for the this will do the cleanup activities and do graceful shutdown +For the release needs to remove the and a standalone jar will be created +Committed to +The motivating is the task of converting a Derby based to +Added HCFS into stacks for and stack from HCFS to so this isnt relevant anymore +Can you test the latest SNAPSHOT with a? As we fork a JVM to run a setup there is no reason for testSuite not to be supported. +Been intending to get to this but couldnt find a good chunk of time +Closing this issue. +I would ignore the discussions about the accuracy of our histograms and just use the for now +And a comment this is mainly because JRE do not have server argument +This is a general question not related toIm working on Apache Thrift and would like to shorten the LICENSE Headers if possibleRoger +Reverted changes from and trunk +This issue should be reopened it is not fixed +This patch takes the second approach minimizing reverse DNS lookups +closing since is now released +Maybe we should have hbase trunk include hadoop? Have mvn depend on a till its released? Id assign you if I could figure out how to add contributors in the new jira interface +Modified previous fix to look only for Immutable annotation +Hi LewisI should have checked on the main page of CC before asking over jira +This patch is missing one for srcjavaorgapache hadoopnet which is why it got a +Something like use +Okay will be providing element and attribute there +So I dont yet see block joining as an argument for this +We moved it to the +Hey Doacan to introducing a object +Looks like this patch may have broken compilation of since doesnt include which removed throws from reassign and is therefore a for this patchAlso not a big deal but in the future please get a from a committer before committing a patch +If it doesnt pick up the lastest SDK then when you import the project into Eclipse the has two errors as I mentioned +Before this change an admin could change the Hadoop or Accumulo config to point to the new namenode and push the configs out +Messed up the first patch +comp att +Stuart is thinking about whether this is only applied to proposal is that this should be a class level annotation possibly called to match Servlet +Agreed this should be fixed +adding new patch +why this cant be handled by class inheritance? and it is more natural using an OO language +This is so we keep the count consistent with Java Lucene +So rather than declaring max number of maps or reduces we should allow large task jobs where each task is given multiple think we should revert this before it is released +Updated affected and fix versions +I didnt resolve this issue because we must discuss about the features which must be available in the editor for other files +Preferably I would like to be able to register aliases for an element to specify that the instance vars should be attributes not child have looked at writing a new but it seems more complicated than that as there is something else that builds the xml String +This really REALLY REALLY slow things downwe might build them in module but its just because they are a file with a timestamp +Please add any follow ups to the issue above. +Attachment has been added with description Attempts to draw a has been added with description Attempts to display an has been added with description The image used by has been added with description The resulting PDF not reproducible for meWhat java versionvendor do you useCan you provide a trace log +good idea also nice to have things like JAVAHOME detection mvn git and so on into the shell environment +Updated patch files for XML support based on all of the comments received so far +Ive updated the patch in revision t the problem here that we have different EULA files not that we havemultiple EULA files? If we remove the EULA files under the directoriesother that then those directories have no licence file associated withthem +Duplicate of. +This is in order to make things work better on WindowsI have tested this in standalone and operation on both Linux and Windows with unit tests and with the Nutch crawlerBarring objections I will apply this tomorrow +Fixed resolved issues. +Indeed this is a duplicate of +fixed warning messages +If someone feels strongly otherwise it can be changed by removing a few linesWe will still save a lot of memory because empty expressions are sharing a common class instance +Perhaps the common elementis that in all these scripts the extra column is in both the WHERE and ORDER BY clauses +in transition after stopping the RSIs there any easy way to identify a single node cluster? Or you mean to say if online servers are keep waiting till it becomes atleast +As discussed at JBW Vegas we would like to have a dedicated mailing list +The attached now fast fails with an +Follow the steps in at the root of the Seam distribution to get your AS cluster started +I added an alternate https access path jetspeedsecure +Have you examined the POST to see if its being sent as JSON? AFAIK function makes a regular post not a JSON request but I could be remembering incorrectly +This seems like a bad example to set in a template script +So it seems fine +sorry +If your broker is supposed to used JDBC persistence then you should modify the to meet your needs and run with that if is what you want then the makes a good starting point +I pasted the here would you pls check if I used the maven plugin of openjpa and dependency correctlyWhat I am running on is simple environment just from cmd line of IDE to run a junit testAnd just let me know if I missed anything +Im unassigning myself for now +for a commit +Ok +Committed +No that was a mistake +Can you check that one please +Map an attribute of the session to a new object. +bulk move of all unresolved issues from to +To reproduce this issuecd astestsuiteintegrationmvn test Dtestas Ppreview think this is a duplicate of +Thanks Alex! +Maybe its a Windows vs Unix line ending issue? I had another problem though +feel free to change it if duplicate does not sit right. +moved db creation scripts from modules to bin updated scripts to work with derby version packaged with Geronimo and +s methods like get return objects +No I dont want to use aggressive releaseMay be you mean need +I have uploaded the new bundle in the same location with groupId of +explain why we need to restore workarounds are Use packaged web archive instead of unpackaged Replace all files with if such file exists in jboss start solutions that comes in my mind areDo not modify at all and do just memory changes only as suggestedby one user in the forum but I dont know if this could be possible +I wonder if this causes this particular issue as well because Im not sure theres a default user location for SYSTEM to write the Karaf history into +Couldnt resist chiming in hereThe first question is why would the call to send the message from inside the onMessage of the message sucker fail? Can we have a concrete example of this that doesnt use bytemanIf it does fail probably the message sucker just needs to close its session and recreate it to force redeliveryRegarding XA this is unnecessary since with JBM all nodes share the same DB so theres really only one transactional resource +Yeah we could add aliasing +As TTL is already tested we may even think about getting rid of this test +Doing a rows is unpredictable so weve previously done a loop of incrementing start by a and getting rows each time +a modified Chris suggestion to break the issue into two partThis Jira will be for the gridmix part only +Attaching proposed patch made for trunk +I wonder why this happens only when recovery is enabled as even the recovery is disabled the logic is the same +Ill take a look at thisAlejandro can you please provide some more detail if you have on where the extra element to the principal is getting added +Thanks all +tmp and then rename it on successful create +So now it is only necessary to fix more modern distributions +It is not clear from the bug report alone what the underlying issue is +The does not specify the deploy directory unless it is running within workaround is not correct +Will wait until a test case is created before closing +C ext support is deprecated as of so Im marking this Wont Fix. +Id also like to know how this would play into and. +Closing the issue as the API is working fine but there is another issue raised for the UI. +Tom I really like this ideaJust to be sure Id like to run all the tests after this patch is applied +Thanks for merging the timeout part +I wasnt sure what the protocol on changing the expected behavior for was for +Quick note about the patch the tests fail unless is prepended with false + +One thing we could do is add an activator to the zookeeper jar and have it use the OSGI Configuration Admin to control the zookeeper server +The current behaviour throwing an exception when all guarded transitions have conditions that evaluate to false is correct +Same bug for some reason this defaults behavior +I think that the relocations are simply exposing a latent that isnt being handled properly +In my test excel exporting only export header of each column to spreadsheet +Ill take care of that immediately +Thanks +Thomas +Ok what I was actually looking at was the monitoring view on Amazon RDS showing the number of connections I didnt know the lsof example would not work +A couple commentso I would slightly reword the first sentence under New Features This is a bug fix release which also provides localizations of new messages +Can you describe the scenario where this bug occursIm not sure how wed ever have a null connection at transaction commit time +But I actually tried to be careful not to change whitespaces +Can somebody confirm this +The whole Validator and code is to be soon when a new pluginextension mechanism is available for validation +I will commit this tomorrow afternoon if there are no objections +It is the same state after a drag an item from aviation folder to autos folder +pleas use z to struts no is it is related to could you check with version as well +Will commit ASAP. +Closing this one as this is a duplicate. +We must be sure we dont mess with that in case an exception or something happens +I dont have cluster access yet might be a while on my end +Everything else will get just copiedTBD in trunk +If this issue is fixed please resolve this ticket +Sergey Ive tried debugging this a little on my side because Im still not sure how to replicate this in my project that produced the original errorBut heres what I can tell youIn the project that works after your patch the instance of resourceObject on line of is an instance called Testb which is basically aIn my more complex project that fails the instance of resourceObject on line of is an instance called Proxy which is aIm still trying to figure out how to get my project to fail but I thought this information may help in the meantime +Please verify +But of course feel free to reopenWe have unscheduled issues in JIRA I wanted to start cleaning up a bit as were almost in the code freeze phase now. +Thanks Paulex Applied patches to archive module and to at repo revision along with changes for +for the patch +Could you also remove property indeed its not needed because key would be transformed from artifactId +for the patch +Currently this patch makes things WORSEBit of work to do yet +Both of these would be good to have and came up during design discussions but we have not addressed eitherFor in addition to your points there are other locations where storage directories are assumed to be +see my comment in Rgds compiler now emits fully qualified class names so the is not required to located unqualified classes. +James are you following the conversation Case sensitivity of a dictionary loaded from the model in Dev listIn my opinion we should stop using this +Thanks for the new patch Yonik! It doesnt apply cleanly because of the way you generated the test files but after those have been removed it looks good +Triaged for +We tried running the tests with marvin from master branch and it the run was successful. +This is because when I used my own system installed version of maven to build the project it worked fine with no problem and it worked with all the versions that I used but when using the Tool Suite mEclipse then I have that error +I actually thought that this one had been fixed and committed already but I guess not +Clearer to have a separate issue +The converted app I tested adding to as Roger did according to forum post I see the same behaviour +Register as public +No idea what you are requesting here +If you want to go ahead and commit this for thats fine +Fine by me + fix javac this is the correct file +I tried to replicate the issue again but was not able to do it +Committed patch. +I assume this is still an issue +I had problems applying this patch +any insights? +The first idea I had was to build a reference interpreter alternative that worked with super sonic +No need to deferfix At revision I also replaced the old ugly msg iconsve tried it on build Aug AM +Sorry to add my opinion to this however I think that while deletes are quite fancy it seems they are open to various bugs such as this +RTC work item defer minortrivial issues to +nice issue Oozie is released +Includes Dans fix suggested above +Give me a shout when you commit and I will try to run cite tests over here to verify that nothing has broken +adding a save puts it there and fixes this bug +AliakseyCould you check if you really replaced your plugin with this patched pluginThe size of the original plugin is KB +Suggested fix is attached +Closing this is working as designed as far as I see. +Can I request a code review and commit of this patch so that it gets integrated in to branch in the mean time it is ported to? Thanks +I can reproduce it only with Conversation that has id conversation +So far the current cases are only the ones found on the Catalog Main page Im sure there are more +This is my mistake +Committed at subversion revision +Somehow I didnt get it. +Does the latest patch solve the original problem as expected +New design could be described as Fair Share in the RPC layer +I do not think we should spend anymore time on code thatshould have never been there in the first place applied. +It does not matter if I override the default duration or notI will try to reproduce it in Wicket beta like this has been inactive for a while so i am resolving it as incomplete for nowalso to note lookslike the thread is being interrupted so it must be something outside of wicket most likely the servlet container that is timing out what it thinks to be a runaway request thread +Sorry but this is a limitation of Android +We have jira for supporting workbench artifact resolver with Maven launcher +to see this bug fixed +In addition there were a couple of bugs that were fixed +Also none of our automated tests reproduced this so closing. +Thank you for committing it +Have service +Thanks for all the help Rick and Dag! Im attaching and modifying one topic each in the Admin and Dev guidesM srcdevguideM srcadminguidePlease let me know what changes are needed +Please attach a sample project which reproduces the problem +This patch is dependent on +Ack nice catch +So the EJB removal has to be somewhere between these values or one of them +Note that this patch does not put the license text into the CSS files generated by the DITA toolkit +The fixes seem very trivial but we need more eyesopinions on what is the best thing to doThanks +The backport was done with +So the case for this happening was really rareHowever even though being a really rare condition I preferred fixing it now since the system would be left into compacting state and the system would probably fail if this happenedThe fix here was to close files synchronously +x +I have an audit job looking for this so if we see it come up well let you know once we move prod to Any idea why compaction run right after schema creation seemed to fix this? +I am working on a patch for this +The problem is that the ports are also hardcoded in the fileThe easiest solution is probably to use a embedded database stored in the parent directory if the cluster nodes +Created a second round of banners for rv is the winner +I am not too concerned about the failing test +A fix is in SVN +It was committed after the branch. +With the recent lack of developer resources I think closing out issues that are not reproduceable issues that people are no longer experiencing or issues resolved by recent changes in Hadoopetc +Ok +Sounds like this is not an Undertow issue is gone when using a snapshot with the latest Undertow for change of resolution + I was facing the problem that Im not able to schedule calculated dates in a workflow +passing to you since your name is all over the classes +This example contains two parts +tested locally +The commit protocol is strongly intertwined with the way in which the server has to behave +The interesting thing was that I enhanced the jar URL mechanisms inside of Java to create recursive jar handler capabilities +Additionally sample binary marshalling code already exists for any language with a Thrift driver +It would particularly help Spring Roo as JSPX is the preferred template language +bq +Id originally intended this only for had just thought to investigate the impact of a potential fix before pulling the trigger on to so we can then rename to +In REST world it select the longest URI match resource therefore our approach best fit for that model +Set the number of remaining hours to + +The user is correct but the distcp command specified a different dst directory than user +That is indeed odd +It should only happen once since those files wont be generated in an updated checkoutThe files Myrna found in bin are the scripts we generate at build time after +for the last version as wellPartly unrelated question for standard scanners would it makes sense to merge the rpc calls for open and the first scan +There appears to be an issue with this patch when using operations with MEP +patch looks goodI have committed this +Closing per above comment. +Stack I found this on build reading it is known issue on this branch +Drools Tools Reference Guide +lgtm +Could you please help review the code I revision for and revisedStill with +Verified by Svetlana +Can you add it there or should I write a small patch? I think it should at least be mentioned there for what it is for but the are much more informative and the corresponding paper code credits are cited thereThank you very much for helping to get this into Lucene +I am closing the issue as Maven already provides useful mechanisms for that. +Fixed in way dialect does +Now that work is well underway with AS all previous community releases are +Initially I thought that will be covered as part of or some other sub task +This was fixed in more recent jiras +Reassigning back to Brett +More progressThis is not yet ready for integration +However I cant configure Spring Security with never because of this issue +verify +Otherwise we might need to use different uris for web services then for the web UI which would be ugly +Please feel free to reopen if you still see same issue with the latest version. +I think we should create a new JIRA to the pom as a child and leave this one to at least get it working +The webserive operation itself is not honoring the Oneway annotation on th webmethod +Even TCP should not work +That would be a visible issue that someone can take and improve +But actually not working in and I cant figure out why +Would it ask too much to commit this? It is very stable and works very wellI know that there might be a new version coming out but isnt that always the caseI would love more people the weigh in +The message is also logged to the console but the exception from the process method is not logged to the console +But I think it is unfair to refuse this patch when support of the same type already exists +This really should be a separate config though +How you are using the interceptor though legal isnt really a good practice +Completed in Git eaeedbdeeafe and also the command line option for DB was changed to DB for better tab completion support between the two variants of the DB entry from the parent wrapping pom temporarily in Git ID daceefccfdeacfe due to licensing reasons. +One should expect to have to do iterative refinement after any significant change in this area +As Ivan mentioned abovethe expected behavior is that the security realm file is not to be deleted when you delete the datasource it depends the realm file will stay in the ropositorybut we should also keep the consistency on consolewhen click Security Realm on consolethe security realm file is expected to be showed. +What do you think +You make that choice +Unfortunately I dont have any time to work on a test until Monday +Alexander shall we handle that somehow +This patch includes strlcpys source code and uses it instead of clean looks fine to me +Defer to +The project is specified locale and current locale on the machine +Its ugly as is +Attaching patch against is not impacted +Changed to improvement corrected affects and fix versions +Right now it expects to be able to have a call back mechanism which is a huge impedance mismatch for the control flow of iterating viewsIve been working on a python port of webmachine the last couple days and as part Ive been trying to think how to patch webmachine to return a that could be passed to our streaming output +this is good for demo but not clear for the first propose create one more Tab here special for generation in external process or generation in external process should be in up part of Main tab + + +For HEAD were integrating +Thanks Lars +Thanks for the patch +I think this is gray area yet according to +If it supports EHLO and supports BITMIME then the message is sent without conversions and is not corrupted +Content Assist for ESB XML file +Tom any ideas on this one? Is it simply because its an older Smooks version in the config +It is WAY too hard to stop a Whirr clusterI just booted an elasticsearch cluster on EC +Well Ill take another look for +It attempts to ping the driver genserver with a reloaddriver message but the driver genserver just ignores any info messages +Im still not too sure about adding implementation or data into annotations but the qualifiers do add an extra level of indirection so it should be ok in practice +Likewise I think time to give up on this one +So if someone tries to export a table in the trash can they should get an exception +Doing this I was able to replicate the issue reported by the streamy fellas ISE region offline +Second patch is to add this test to exclude list for JPDA module due to know bug +It is not working +Fixed thanks for the patch! +Kevin Looks like you registered with but now are trying to deregister with an all lower case email address is wrong at our end Ill make it case insensitive like it always should haveve made the required changes to Xircles you should be able to view your subscriptions with any casing of your email address +Committed to trunk. +In I implemented the balanceCluster method in the +Am I missing some thing +Okay just ran the tests one more time and looked over everything +This patch disables variable scope checking in extension elements which is the best we can do right nowCommitters can you please review? +Ive committed a fix for this +Looking at the options shown there I would say it should stay since many of them arent yet available +Forgot to move an error logging in the right place +Cutting over to and letting dictate how compares sounds greatWill CMS still have the ability to pause running big merges in order to let smaller ones complete? Seems like these changes should still allow that to work correctly +Move container init after wont start the manager I have decided to use a valve instead of a filter because valve has a simpler life cycle +So I changed that to use undescore +Fixed in SVN +Isnt that a core mandatory swc +Looks like Amila has already fixed thisthanksdims +My plan was to split up the existing docs into Getting Started this is really about getting started with the project and not an install guide +Applied patchThanks Johan +Workaround Copy the file from the folder orgapachekarafassemblies to orgapachekaraf +The color of the bar changes to orange if the used capacity goes over and it doesnt streak off the right end of the page as it did previouslyAlso added a legend so its clear what the colors indicate and added a used capacity column for quick reference +A few notes on the patchThere is some duplicated code that should be factored out in a few locations due to a quick copy and did not update a unit test that is affected by the additional method m going to dig into to see if they actually close resources acquired when a connection is used in this manner +Currently I am tending to close this bug as incomplete since I dont have any original example to try this onAs for your Options +Theres some funky model stuff going on under the covers +Thanks for your help on this one. +Want to commit to +Restructuring and renaming is done +The documentation has been updated to reflect the new attribute +I believe this one is still relevant yet its probably a bit too lax the code should check if the extra attributes are the GML ones if they are not an exception is imho warranted +I want to review trust verification in general and incorporate your patch as part of this reviewJust to clarify are you after a fix for the branch or only trunkColm +we wait for some time before we consider RS discussion might be better to have on the dev list then here +BUMPNow that is part of Hive the need for this is more apparent and this also blocks Sqoop integration +The Total Shipped column now displays the quantity of this shipment plus earlier shipments if anyAlso to get the requested count to be correct when one product code is on multiple order lines and is then split across multiple packages and shipments I had to change the look up of issuances to be against the order rather than just the shipment otherwise it wouldnt count all the requestsIve looked at this in quite a few scenarios and it looks OK to me now so any comments welcome before I commit it shortlyRayPS Found more problems with requested counts on the packing page but Ill create a new Jira for that +You will want to change logging level and then restart since there is some initialization done on startup and if an error occurs the inline handling disables itself +Thanks for the heads up Anish +Would that be acceptable +Fixed +Even though both are Linux machines +This is directly related to the default Native to JS bridge on Android being a HANGINGGET +Pls review the someone please review the attached patchI am not an expert on the code but I thought that this was by design +I will try to find out what is the problem +Thanks! Ill look into the batch writer when I get a chance and let you know +Here is a patchThis patchs using a system properties +has been marked as a duplicate of this bug +The file appears to download OK when accessed manually however the script does not download it correctlyThe downloaded file contains the content twice prefixed by bb and suffixed by nn +What he seems to be talking about is that the two commands that he tried dont work anymore but he didnt break discussion around build breakage will be there something that needs to be done here to make these commands work again from your end +In accordance to the roadmap when is this version expected to be released +Thats actually a good idea +not really just saw some stray resources in a heap dump +Also the comments on reviewboard are from M IS not me +Note that this bug is not about maven support but about running this code inside of Eclipse +bq +Database connections are thread bound if necessaryChecked again checkstyleNot thoroughly tested +Marecel Maatkamp has reported problems when processing Treo profiles which contain errors +What is that? +Thanks Divesh Pranay for the contributionYour changes are in trunk at revision Ashish +meta will be written to many times + code targetted to be refactored in with Milindas help on his work +Reading the result columns in order is actually recommended by the JDBC javadoc for for portability so while it would be good to not have this restriction I think it would be ok to require it in some form if it made a streaminglocator LOB implementation easier +Forgot to mention that this testing was done with the patch for applied as well +Closing all resolved tickets from or older +I went ahead and checked this in to trunk but would appreciate review +We configured a batch job to run with a of threads and the job never finishes +The component must be declared with this namespace for the return functionality to be available +Do you mean you want by default the libraries to be under lib +Comments indicate that this is a duplicate of +Do you know the version resolves this issueThanks for your you very muchI need the two encodingsI do not see many work on this api + +Fixed some things I looked over +Merged to +folded into +However the code is searching for all link in the HTML page so I dont see the relation with a specific format +bulk close for +This has a back compat break that we should address somehow or at least mention in changes previously you could specify explicit shard ids and still get distributed updates now if you do that you wont get distrib updates as shards wont be assigned ranges +Thanks Liu for the patchThanks to Ted and Stack for review +shouldnt be major +If there are new problems according to CORS support please create new specific issues. + +Thanks for seeing these issues and fixing them +One is that we access DAOs from different threads and in some cases from constructors +Testsuite run Failures Errors Time elapsed secTestcase testEnvironment took secTestcase testInjection took also passes in EAP CR version +Added the change to update the instance post handshake so that the session in use contains accurate data about the cipher suites etc that are in play for that current session +I can not see a requirement for this code gets more difficult to understand and there is no added value I can see +I have asked Burr to update and try again +This is indeed the case +Theres an open issue whether Signature attributes have to besupplied for augmented JDO methods that carry the type of managedfields or the PC class as parameter or return +this will be address if the discussion in the dev bares fruit. +Hi KeithIve added my comments on the double click issue to issue. +The reporter posted a small webapp that had a crazy recursive tile insert thing going on IIRC. +revising titlecomponent to better reflect real think it is that fixed this one +It is still possible for different jobs to modify the same process original work queries a shared database to retrieve a list of jobs thereby opening the potential for concurrent executors to execute the same job +Above all Im sorry for delaying this work +bq +The latest snapshot is up now and should contain my changes +In our case the first datanode confirmed the block and the block is inserted into the blocksMap +Removing unrelated change from +All the people who have been assigned know the branches they are working on and have the mandate to commit in the branches they maintain +Thank you MichaelI have it targeted to be fixed for the release +the fix for the issue should take into account note take a look at comment +Attached with patch tested MSI working properly. +For uniformity this patch should use the existing convention +now Im ready to be pedantic I refuse to let file handling get complicated for this stuff +and fail to compile + +Ill take a look at this +I raised the Jira because of the discussion on the user list and the problem that I cant access my service with the generated clientBut as far as I understand you found the error and raised another Jira thanks for thatRegards Michael +That looks great to me! Ill commit in a day or two +to commit to trunk and +this is currently just added on top of JCRAPI which will not work in the long run +I made to do this better in. +is there a reason we cant house those javadoc resources inside the AS javadoc aggregate project rather than attempting to deal with file collisions during aggregationAt least then we could be more explicit WRT which resources are used and which are not +Feature request +Same issue just different element +Nope I have no more context at this timeI suspected that is what you might say +I used to have a bit more understanding on how warpath works but I believe it unbundles the war for you and manages this aspect +This is a patch which adapts Spring RCP classes to be compatible with the new Looks library +Committed to the trunk as revision I intend to merge this change to and commit it there as wellI also intend to mark as a duplicate of this issue +I just committed this. +Sorry I cant answer your question with full knowledge +Again like mentioned in the other issue you specify the controller only this will automatically result in a GET request which will take you to action +Just committed +lets commit to trunk for the next major release and ignore untilunless someone is motivated enough to backport it and update converter. +Patch attached +This problem could be fixed either by doing a contains check before adding a datanodeDescriptor to the heartbeat queue or using a collection type that disallow duplicate entries for the heartbeat queue +Moved to t make that this JIRA did not make it to +has been applied to the trunkShindig PHP now fully supports Atom and RSS feeds through +Please use the consolidated patch available in instead of the one available here +back the old breaking on +a new version of the C file to also support file streams logs in new file streams handling introduced with the above patch should work with both and so the code could probably be simplified removing the conditional compilation preprocessor directive +Mladen does this mean that by default a iaw build is done? I do not think that this is a good step because default compile target fo CC and GCC on is always +Ah I see. +But two packages with the same name defined in different class loaders are indeed distinct packages and do not allow accessibility between them because such a check is performed on the identity of the Package instance defined for the Class +Assuming this gets accepted Ill fix the code style +I am marking this issue as superceded by based on discussion at Summit +To reproduce the failureant test Dtestcase Dtestmethodtest random +Checked into branch. +Heres a new version of the patch with the wording we discusses +Norman should review this +Please reference for further details +Regarding the issue in Indexer a word of warning you are running with mapred speculative execution set to true +yes followed exactly these you have configured ivy to use its probably an environment configurationim not working behind a firewall. +Removing on these issues until we have somebody committed to fixing them +Would you please also postyour test files so I can easily reproduce what you were after +OK I applied the patch with revision . +Is that not sufficient +The patch applies and the feature makes you be able to provide a documentation updatepatch as well? I will handle this soon +I will check in the code shortly and update the documentationPlease let me know what you thinkCheersAshwin +For sure ryan +Core tests ran successfully on my machine and looking at the test report the failures seem transient +We also think about the possible solution in which libwebdhfs uses its own preference files to determine the default hostname and port +Any recommendations +Probably you mean end of stream condition not a timeout? There is basically no point fixing as it is nearing its end of life +Our build is done against Spring not +This isnt a mediocre API that should be improved later its a bad API that shouldnt go in at all + this is how should patch fix the please reviewtestcheck the patch and it back to me with review notes +Thanks for the patch JohnI put this into trunk and +Candidate patch +Sorry about the formatting +The other issue is that the peer needs to be cleared before a list of peer brokers is created for propagation to a client or to a peer as the second level information is not relevant at the momentSome of my tests are showing blocking with duplex network connector start and I want to get to the bottom of that before progressing further +to Oct +Configurations are observable since the release with the introduction of and +I have no idea what is going on inside velocity but suspect you can use either logj or logback +Sorry I meant after all regions were moved +The test failure is un related to this patch +Switching to then +only after seeing your comments above will could you try again +fixed in svn +What is it that is causing it to take so long anyways? Is it because we increment the replication counter +Looks good thanksI note that you didnt check the Grant license to ASF box when attaching the files +However ill to that for the all linkrank settings prior to commit +Btg per we shouldnt allow people to easily create symlinks until at a minimum theyre supported in and weve got a release with that code +Simple patch that implements the proposed change to guaranteed capacity computation +you think I should raise this with the Tomcat guysRichard +Could you help me submit it to Jenkins again +Heres a patch with the proposed refactor +Eclipse is not exporting anymore the file in the +Sweet thanks for the quick response +does run with ant test on Hadoop but not with so its important to run tests against both configurations +Im not afraid of native +on this idea +the same comment about missing the Spring specifics would apply to the client side but you are right we still have Spring in scope even after the wrapper returns except that in the client case we have to postpone the initialization while in the server case we can get the wrapper do it for usCheers Sergey +Merged thanks for the pull request! +Added to MATH branchContinued fractions convergents were diverging because of infinite intermediate computations +In response to Alans comments +Raise warning instead of exception and continue with code generation +Can you provide the configurations and platforms? Also since there is always the possibility that I can not replicate the same config I need close collaboration on debugging the far I have been quite relaxed with our and Android SDK requirements +the wizard now initializes the page as incomplete +Google seems to have done it! +Im going to commit this change and roll a new release candidate today +Good point then maybe your patch should be still neccessary +It seems I was over confident of my little hack to cure the println problem +If that is true is there a work around that we can give the users to disable the workers in some other way +Which rails plugin has set to map? +Committed revision +Is this not what does +it doesnt work that is the problem +Indirectly +Now that work is underway with AS all previous community releases are +Although it still does not work on Silverstream I will take the issue up with them because they use proprietry classes when generating java code from the JSP. +These are not jrubys fault but just the way java works +but i can also see the advantage of allowing more specific control +you have several options +Done +Updated patch for for latest state of the branch addresses Arpits comments +some support added in but need to complete it before closing +made a small change in for D testcaseCommitted in trunk thanks Vandana +Proposed code is the correct +Looks good +API changes needed to expose this to clients by default +thanks for letting me know +Found a bug when the imported file has an import +Otherwise an inconsistent value may be read while totalLoad is still being accumulated +The same when tested on another machine with default python version as it has thrown the error mentioned +I only mentioned it as one more thing Hadoop could use Native system call access for + patch is uploaded +Lior CostinAny progress on this? +Thanks LeonYour patch is in trunk at r +Ensured that activerecordversion is loaded by jdbcadapter in. +I committed the patch to branch. +Someone may have built a system that depends on it being like that +Moved matrix classes to matrix package +Im reopening this. +Curious if you see the same on +What is the status of thisI recall seeing some mail about if we needed this plugin or not +Patches look OK to me. +Looking forward to haveing this feature thanks! +Updating the pull request after finding some issues with the config model when playing around with forge +Previously it contained the +original ve attached a little patch for kernel unit verifying that Evguenis patch restores the hierarchy broken by the launchers implementation +Reuploaded with unsigned jars to the same locationCommitted revision and +I agree with Nadav the problem is with the Queries that use to build the priority queue its just not good use of the queueI also agree with Doron that its really two distinct issues though +That would ease reviewing and testing as well +is this a critical ticket? I think its a request for enhancement +I redid my patch to handle the case of slashes and added a couple of unit tests as you suggestedWhen running the test suite I had some tests hang due to timeouts but no failures +Net seems to parse the betterI would leave it as isDIGY +This is a good idea +Odd that the SHA would not match since its all one operation +Attached pull request based on and merged the second into the master branch. +Ill fork the repo and will come back with a pull request to get it merged when Im done with it +Feel free to close this issue if you think that there is no further work needed +Thanks for the high quality patch +So as I mentioned above the problem was that templated beans were only being translated at creationNow setting a new locale will force a of all elements in the DOM from templated widgets as well as any fragments rooted at an beanNow at template creations the root DOM element of templated bean stores a attribute and all elements which are translated will explicitly store attributes +Vladimir could you please try the patch +Sorry for the wrong level when I raised the issue +Please leave the Fix Version empty +Can you change the encodingThe method is an abstract method which is overridden by the actual class that implements the certificate +Its an optional element to the specification and its a technology thats disallowed by the Basic we see a large enough demand for it we may change our mind +Igor So the error is actually there and its a fatal one +Net clients which were removed from the repository months ago and not the current +Patch to search for Central Directory signature when ZSYNCFLUSH has been used +That would mean creating an empty concrete client that extends an abstract because has a complete default implementation of the interface +Not for commit here +My server dont have internet conection whit internect conection dont have errors +As long as the implementation honors the isPassword flag for the parameter thats it +Hi Ronthe code you mentioned uses the doReceive function of the +is not an official GCC release +So what to do in this caseShould this be raised to Ruby core folks to get the clarification +Rick said in a comment on that he saw a deprecation warning in on trunk now +There is a subtle bug in the handling of failed moves of the files to its final resting place +The grep is available using vfsgrep +Note To simulate badly configured proxy behavior you can remove your computer from the network and run off line +Rototilled tree under javaengineorgapachederbyiapiservices at subversion revision +I believe that this patch is now out of date with current build +And it adds more questions than solves problems +Right now this leads to stack overflow error +So all issues not directly assigned to an AS release are being closed +verified fixed +Patch committed rev bb +I think we can close this JIRA +I plan on uploading to SVN within days so if you can try it out and comment it would be helpful +Forgot to clean up conf changes from the last patch +have been fixed by some other work on the code +I think if you look closely at the logs youll see that its the forked build failing and not the main one +How about the javadoc warnings +I have tried to reproduce your bug but its a bit hard to do it preciselywithout having the date sitemap flow etcCould you attach a zip file which can be expanded in a directory underbuildwebapp including all required files to reproduce the problem +Implemented Mahadevs suggestion +Pavel I believe private static final String JAXRSSERVLETNAME has confused you about the usage +Patch review for Javadocs is a little much +To reproduce this do a and use currentbuild in your webapp +Committed r +Our internal format for compiling metadata from Teiid Designer is exactly in the format youre describing +Test fix attachedI removed the initDone boolean flag +Since we are going to be modifying table and column descriptions for this set of enhancements we should take this opportunity to make max versions a per column attribute as well +Thanks Bobby +Will upload updated patch soon +Simply attaching the code for your application will not be enough. +this is a patch which errors out immediately on a rename or copy failure and then logs it but also throws the which should open the dialog saying there was a failure +Actually I would not do it like this but rather have the servlets set the encoding themselvesWill do that +They have private svn. +Preferred userid gbrownAlternate userid gkbrownFull name Greg BrownForwarding email address gkbrown +additional code for a and b +Let me retry it +I fear this all gets much more complicated than a simple line producer for theFor now Ill provide the and we can think about other ideas later on + Directory Studio has now been releasedLets clean Jira and close this issue. +Ta +trying patch v didnt use binary when generating patch more time +Patch from Shubham I can test and then commit +BTW can you provide your test service for us? Better to create simple service without those database access etc + +Checked in the fix that just lets the Properties View cell editor display the first allowed value initially for property definitions that have allowed values but no default value +At least now I know to patch the XML files and not the HTML ones. +Its just that everyone wants support for their particular configuration needs in the namespace and we want to keep it as simple as possible and not rush into adding things which may not be widely required or which have straightforward enough alternatives +The takes hour So I am commenting it out before checking it in +Exception in the Error Log Window of eclipse +Wont there be a sizable performance impact by switching it to a given how we check users Authorizations as we iterate over key value pairs +added by class inside method on +The workaround is to update to Spring Security you +We do want to move to +My idea was to move it to the monitoring system +It doesnt apply cleanly currently I think because its missing +Exposing an EJB implementation as a Web Service would mean you would want to leverage the EJB container services for the exposed web service endpoint like Transaction support instance pooling Java EE DI etc +Second is whether split should assume uniform data or data +Does not include unit tests this is an existing open issue +It should be trivial to resolve though so Ill make some changes and submit a new patch +The main reason for this is that may register further beans which may even turn out to be further +Depends on the patch posted on +js stuffthanks to volker! +and a disclaimer about sentence breaking +I put this on critical because its a major time killer +I like defaulting to false and keeping backwards compatibility for empty marker file files +It has an even cleanerAPIimplementation separation +No +Thanks Vlasta this is useful. +Since the parameter types are able to be inferred there are no casting requirements +Package uploaded +OK fixed +Marking closed. +Thanks +The driver does have the service provider stuff but its just not being picked up for some reason +right +All the job can be done through the wizardThanks +The trunk and snapshot have already been updated to use Bnd as part of +and seem to have been moved to the listmap package which seems odd do you have a reason for this? you seem to have removed some classes +Im confused +It used to bejob SUCCEEDED SUCCEEDEDjob KILLED KILLEDjob FAILED FAILEDNow its gettingjob SUCCEEDED SUCCEEDEDjob ERROR ERRORThe test hangs after job +And anyway for Id prefer a more local change just to do +So requiring the slash to be encoded is more consistent IMHO +Instead there are lots and lots of changes mostly arbitrary line break changes +Thanks for the latest patch +Thanks for the analysisPlease put the above analysis in a new JIRA +Hi SimoneSorry about the rookie mistakes +fixed committed +Thus the monohtml docs are not a derived work + includes a fix to generate flat repository layout as I change the orginal proposed APPENV to SETUPNAME +Change of plan we aspectise the jar to handle vfszip schemes so will only do the jar +This backs the port of this bugfix out of the branch +Tested with SNAPSHOT from Thu Dec GMT To see the XML in question in action is rather ugly run the unit test the HTTP message with XML is written to the console then the test times out +Do you think it would be acceptable to figure out whether the PPR capability is enabled BEFORE forcing is executed? I mean when I see a force method I assume its actually going to Force us into a Partial Rendering state regardless of capabilitiesScott +The patch I created for this has been applied and a new SNAPSHOT version of the plugin has been published +Also I think we are mixing with +I ended up with compile errors in becauseSorry about that +I have confirmed that the messages are actually there by then retrieving them off the queue +I believe this is closed now. +Propertie files ending with +We would need development time plus community bake time for this and would target WFK. +The fix I proposed here overlaps with the fix I proposde for +Also you are missing a newline in the at the end of the named constructors +This is nuts but since both hdfs and mapreduce files are changed I have to open for each for proper accounting +I tried building the source from a separate system +Thanks Chris +I have created ticket Claus +As a result the build fails and the forked DC hangs as a should investigate why the forked DC doesnt die and try to address that +provides isolation to clients +Reopened to address the problem reported by Craig +The integration of Unitils with Spring would be very helpful for many projects and it shouldnt wait forever to be done. +may be deleted by another query it should be checked before access itplease confirm attached patch. +Verified fixed as of rev. +Didnt find any +Lets see if any tests need to be fixed +We somehow need to know which users are using +I had to make a slight modification since there was a simple dependency on a Java method +assume could try to load the codec +so the deployer has to take a look at the location of the jar to find the filehow about exploded jars? how would you handle this +The Tx subsystem contains four new services The service are registerd PASSIVE +Added a feedback indicator in the top right corner +link is between different table otherwise is just a reference file +The staging dir will not be cleaned +Massive will be on Monday sorry for the lag +I didnt notice before but it looks like we require sstable files to be named according to the convention cassandra is using? If that is true then yeah getting the keyspace name from the actual sstable file name is probably the right approach here. +The filtering in the war plugin was introduced in +Also deployed and tested to a cluster and a cluster +When shutting down the framework in this situation the thread cannot be stopped and the framework just seems to be hangingInterestingly when removing the bundle this does not happen and everything works just fineMaybe someone more knowledgeable in windows networking has some clueUnfortunately at least in my case applying the patch wont help because it is the constructor blocking the stuff. +I appreciate your persistence here and hope I havent deterred you +With what version of do you observe this behaviour +Moreover I can see a lot of +Imagine a stream that identifies phrases captures them and then tees them to a second field +After that the bookie keeps using this identifier you could let new bookie inject So if I understand correctly here the idea is to keep the identifier as hostnameport format +Then at least folks can see them and ask the right questions etc +bq +Cant get back to whatever state I was in that caused this error +small bug in PPDfor workaround set false +Hopefully well have RB or Phabricator automatically do this for us in the future. +Ive attached two patches +Sorry took me a bit longer than anticipated +Committed the fix at rUpdated the dependencies from to +Corrected the record iteration and timestamp value return issues +Added patch that sets the symbolic name to be the artifact ID +Deleting the only blip in the thread is more complicatedThe other thing is that if we allow the conversation to have no blips then we need some UI concept to create the first Blip +Ive commented out the source that marshalls and unmarshalls the element for as this is a UDDI v feature and jUDDI does not support v at this time +As a workaround now you can always refactor the Dod ITD and modify the methods yourself to supply test data +Thanks Colin! +This probably has something to do with character encoding but as far as I know all logs are encoded +Fixes the problem under Tomcat +Thanks Ted +Added new patch according to discussion with Doug Cutting and Thiruvalluvan M +Also the current branch shows +Please give tonights build a shot and +Checking availability of ports should be done prior to changing ports of transports +Thanks +So I think you modify that because I do not know how to add a note you could just mention in the tutorial that before running the program its better to make sure not to include the old from the lib in am going to work on geotools further so I would be in constant touch with you +Attached is a patch that converts all existing unit tests to style +Testcase seems to be passing +Do we call it for the test or this patch +Are you saying conflicts should just grow infinitely until they are purged +probably the test puts serverdefault in the classpath +If not I will reopen the ticket reuse the old release note text for CP +I think the fact that the boot fails isnt that interesting the question of why some messages do get logged however is +Has no issue finding +Brian can you please clarify if you are getting that error when launching Maven from Eclipse or when importing project or when running Update Dependencies action +I wrote this up as an issue with the textXML wireFormat handling +A subset of source files that fails to compile in a way similar to whole benchmark +There is no change to service client +Please use the user forum +Closing stale resolved issues +please check it one more time +versioning issue +Igor thanks for the review +Given there is now the and as well as are specialized maybe we can just advocate using this combination if the user is interested in the input line? I guess theres no special API needed +Build Fresh true which is why it build +Word of caution though that screen was implemented a very long time ago and appears to have seen little in the way of improvement +removing official JCR support +Im trying very hard to get back to this but we have done a new beta release of our own software that has required all hands +! although I seem to be late to the voting booth +Imagining the implementation behind p and n as a I think you should simply expose this functionality via the regular couchdb script instead of having a new command +Ill raise a separate issue for thisCheers Iain No virus found in this outgoing messageChecked by AVG Free EditionVersion Virus Database Release Date +Is this supported in hadoop? Sure +Which quickstart did this appear in +Chunhuis can reproduce failure on my laptop +changing wo breaking back compatI wonder if we shouldnt also pass along the request parameters on this so that if one is extending a for spatial which I think will be somewhat common they can have access to the params +Good design +Second removed the extra text sorry about thatAs far as I can tell nothing in the scriptsplugins are hard coded +With the above mentioned change I am able to disctp data to S when security is enabled also +I will contact Devsubversion and look at the source of svnpubsub too. +I applied the patch here and it now works fine +Looks good Tom +I agree with Eli that this looks good except for a couple of improvements +Just to make one thing clear Velocity Engine is NOT moving to the maven build system anytime soon +Fantastic thanks very much for thatPlease note that timer execution is not the same as job execution in the current codebase only jobs are protected by exclusivity as I mentioned earlier +catch exception set nowrap to true to clarify this is in fact the fix for this bug +slip to next dont see that could you please verify with the latest version + +to trunk +PrasannaI dont seem to have any rights to assign an issueRegardsSimon +Can you take a look at and let us know what is different about what you are doing +At least thats the explanation weve claimed in the past +Any changes here will have to wait till the next major version as they could break existing apps +I dont mind if it is a bit more work to get it running inside of WAS +I foresee cases where we wont be able to put a together with a +I made changes as per your suggestions +Oops patch is against r +Should readX and Y intercept a class annotated with A but not a subclass of the bean annotated with A +The type of backupNode and namenode should be not right? Otherwise looks great +Updated to apply to new svn +it must be loading classes that it shouldnt +Initial patch was missing imports statements +Before deployments are processed the auto start bundles are processed +This is just the calculator algorithm updates to hbck to follow +both openjdk and version checks should be +But Im not really convinced we really need a config at all +Currently we are using in trunk r. +Hi Emily At the time I added that comment I didnt know if there would be changes to the sample assemblies such as +looks goodIm attaching one minor improvement that I didnt get around to in my originalpatch a slightly more efficient way of accessing initCause +Guess it would be a INFRA jira +I have the same feeling +show me the resolved issues for already released versions +will add more while uploading final patch +Hi Carl I have already done some work with regard to review comments +Its important Id say +As soon as the code stabilizes somewhat Javadocs will be provided +Ill commit shortly +The permissions should now be applied to all the projects matching the search criteria + the to use now adds to classpath for to cool down the got dependent to work so that we dont have to upload a special version of the demo to the Maven repo +Thanks Cameron! +Bug exists at Grails and with new public member domainClasses has been fixed my problem +Actually I am not sure what the test shall demonstrate +No I compiled the code with java in Eclipse IDE +should be added to the the same way we have added bundles from Orbit mirrors +server only way you have ejb is if you selected this in the installer and this installer states jdk is required to run ejb. +If we are parametrizing the types should we also be unifying all the integral types +Did you managed to translate frameworksprojetcsrpc and frameworksprojectsspark bundles yet +Sanjay Choudhary verified the patch hit the mark. +I am always thinking what was it +Please have a look at this patch and suggest if it needs some modification +Someone later may add some other tags from other library and then it becomes very confusing +Nice work Andrey +Patch applied to CVS +Let me check +Dont know how possible that is if its easy +Backported to r +It only accurs when is used and uses that +So for creating network include vlan cidr private gateway ip and gateway parameter or do not allow same private gateway with vlan and cidr twice +My dev VM server went into maintenance this night and I dont want to run the tests on the laptop as its going to be painfully slow +Good point Stuart Ill either do that or keep a reference to the UT around to return +I find this useful for quickly determining which regionservers have been restartedresurrected +Triaged July Assigning normal urgency +Fixed for +Closing RC issues. +Setting to resolution state before reopening to change affects. +However It would be nice if the user could set the selector path separator globally and on a project level +Its not an integration test its a true unit test +I can live with things as they areOleg +The latter would better permit a cluster to be shared by mapreduce and tasks +Given its the config directory and not the config itself +No need for IT +so its not like certain sstables refuse to get upgraded +Applied fix to as well. +So its probably best to leave this for now until we can determine that its not needed +Fixed by +Ive been running the test times in a row without any problem +We already committed a fix forYeah +Is this the only thing that stops the release of +Those are not unit tests however if anyone will change the interfacesabstract classes in a way that it breaks well know it right away +I cant reproduce the problem but I think should fix the encoding issue +Thanks +This should read skip step +This one is going to take quite awhile to close +We will have two versions in CP +good point yonik +If possible submit the files in a patchAs a side note please tell us of any +I think for znodes that store more than one value could benefit using a simple json format instead of combinations of and + +I use SAC and looking in the source I think a iscreated with type SACEM which extends +There has not been allot of interest in the clay specific namespace so Im changing the target to The JSF jsp knock off will most likely be the preference +Committed at subversion revision +Ive added the comment to the code +But even when that database is sitting on the same machine optimized and so on its orders of magnitude slower +Presumably IDEA and Netbeans are the same +Since only deserialization needs special handling the attached patch adds a private method to readPlease comment +Is this a duplicate of +I will add it and +JPA data I understand the benefits of being database agnostic and having the compiler to help catch any issues but Im still not sure Id want the initial data to be defined in Java code +Thanks everyone this is fixed in SVN +I create manually to help learn code and to illustrate the idea of visualize EVENT relationships between components +Looks good to me. +Ill explore a bit and see what is going on here + +is the correct file reverts the changes causing the findbugs warning +Problem solved +Thank you for commenting on m going to close this issue +Ive attached a patch with a modification to s equals method so that it is similar to the one in +for the review Knut AndersI committed patch a to trunk with revision +It doesnt change the underlying data as of yet though +The same test passes though if run under the debugger +I didnt like to find weird bugs just because someone duplicated a flag that he didnt know it that method is simpler Im changing it +component handling is done in a different way than it is described here +I am adding a new user to get access to the admin console using script failed on me times because I subsequently tried to use the username a simple word a number letter combo etcIt would be nice to get the list of constraints a password has to follow to be valid upfront or at least the first time an error occurs +Flavio +Ill create a separate bug report. +this has nothing to do with our SVN +One aspect that I wasnt sure which way to go was whether to check the exception against the cause we are looking for +Assigning to Alex as this needs fixing in RF +Heres a quickn dirty patch fixing the case where the failure doesnt have an error message separated from the stacktraceI didnt write a proper test case cause Im in a bit of a hurry but I successfully unit tested this patch with the problematic gallio report +We would represent those sets as parallel arrays of stored procedure parameters +I could be wrong but I thought we decided it would be mapped to +Why is this Deployers specific hence getting assigned to me +Committed! this still a problem or was it fixed by the stuff +Above patch checks the cacheType in the constuctor of and throws a if a cache type of none is specified for a class that istime +I removed it in both places +But I think we should be promoting best practice +All docs now build without any expanse errors thanks for the contribution Kim +On Windows Server R Bit with Oracle JDK the fix works thanks very much for sorting it this issue was beginning to frustrate our clients +How stable is the code? Actually having said that it does not really make a difference to upgrade a source file or a JAR right +Lets let it run a few more times +The imported were not setting the output method at all and were assuming it would come from the main stylesheet +Thanks for the extra info Denis +SrinivasJayasheelI have attached the wsdl file Please help me to migrate to axis +Personally I more envisioned EL exceptions but I guess the spec suggest any sort of runtime exception +would you describe exactly what is the scope of this patch +Changed the propertycount name in a new patch. +bq +expunged all uses of Text in the timeframe +Attaching from branch +Patch was applied +OK I finally got an assertion failure after running all night +Could you please check why they are not picked up +Ive created to expose row index size so this can be tuned more accurately +for the +Thats why it has not been done that way. +Which is not to say that I have a much better idea +Created an build profile in Revs +This is for novice users who want to use just WFGEN out of box and users just need to expand tarball run bin and ready to goif there is no objection for this as well the patch is ready for commit +Yes It would be better but it would impact thousands of Ofbiz classesHad I started it from scratch I would have done it as you suggest but now it is too impactingThank youRegardsAre there any remarks regarding this workThank youRegardsPhilippe +Because user may addmapping for several security realm but each time only one securityrealm can be used user need to specify the one to be used intothe security realm name field of generalpage first +Note that the +Please provide description to this issue I cant figure out the problem from test and your changes in codeNote test is in Protected part of it should be moved into Public section +bulk close of resolved issues. +updated stress client not to print false thread idle warnings at the start of the test +Who were you asking AmilaI personally cant right now unless you attach a packaged up system +Thanks for your document! +I am not sure if it is possible to influence the order hereBut it looks like that your code will load these into memory and then load the with the proper serializer later +My assertion is that this issue is very rareWe saw the frequently so perhaps not +The error is in work ok since our patched source code no longer shows any diff on this +It would be good to get it fixed earlier +I had forgotten to change those testcases in the earlier patches +Im trying to add index support to build using maven and Im facing exactly the same problem +The interceptor is invoked when an exception is thrown from an action +The Netty dynamic buffer does not resize on a setIndex +Proposed patch that allowed to configure Cobertura with Groovy plugin for Maven +Fix is simple and safe to be merged into +Say the boot hangs for some reason it would not be possible to use the http api to try and address itBut If the boot hangs due to some bug the odds are very high the hung thread would hold the exclusive model controller lock +Fixed in the CP in revision . +ThanksActually I think you guys should review the related API usage on this project because I just patched what I usedThe coverage is NOT +note as patch available +I checked it again and it looks good +Great +could you try putting a synchronized on loadClass and confirm it solves the issue +committed +I will commit it if everything passed +Hope that this patch should solve the problem +Im not sure how best to specify it though +This is on my todo list and will be fixed before +Trivial patch that uses to avoid concurrent modification +A more comprehensive patch addressing Hiteshs comment and includingAdded new context objects for easy evolution of +What defines wrong from your perspective? Remember the Maven Central artifact ids are different than EBR artifact ids +Any commentsconcerns with the approach outlined above +Bhupendra has applied the patch. +Let me give it a try and then Ill let you know what you if you will be able to roll this fix into the release +Problem is fixed in it or cause of problem was another +Out of interest how did you fix the problem of specifying multiple D options +Logged In YES useridFixed on the current CVS Head as well as on Branch +Committed the code for HTTP Basic Authentication +that Im running the unit tests I can see that the Master doesnt handle the case where its assigning a region to a region server that is shutting down +The performance problem Ive seen is not dataset specific Ive seen similarily bad performance on a variety of datasets when queries yield large results +I had it about done but during the past month I had several hardware problems starting with a hdd failure so Ill have to redo it +Within a major upgrade say to throw exception and indicate user to use a separate upgrade tool to upgrade RM state +In other words theres not much we can do about this from Springs +Thanks for the patch! +In our use case we create a new session each time so it doesnt make sense to keep these sessioninfo data around +It would be much more pleasant if we can easily identify premature time out situations and adjust max time out values dynamically by a system property according to different environment needs +Applied patchthanks issue +So sources attached to Maven Dependencies entries are there but not the projects from workspace? Some simple test projects would really help +But that was just thinking out loud +Thanks Freeman +So I think we need at least two baselines micro and macro +Here is a patch based on the latest version of the project +For me this is gmail vs +I would be fine if you decide to move it +JST component should not depends on AS component +Is there a snapshot available somewhere +What I get out of this thread is that we have a general problem with the performance of from Java code +I also created Issue fro difference and sumitted a patch please check it thanks +Builds and passes tests on both trunk and both autotools and build scripts +I also the backport as it is slightly different for consistency +Sorrythat cannot work because the segments can share the same terms so the sum is always the real unique term count +Thanks Chuan! Ill commit this after Jenkins gives on the latest patch +But you are right the Reference Guide doesnt define what an Expression isThanks for the patch Kim + fixing the condition +This version of compress which I refer to as offline just reads through the used pages of the file and inserts into the new table and never looks at the problem empty and not on disk pages at the end of the file +Resolving this issue +bq +Added the section file selection to Doc Smooks checked the screens. +I will ask Andrej to take a right now integration tests contain both soa and +Great Job Dan! +Please go ahead and apply itthanks the uploaded patch. +It looks like the continuation field is only created in the register method but not dropped in the unregister method +This slow down tests +Any commentsreviews at that time are appreciated +passed in my local workspaceI am looking at the test failures in +follow the steps of Amit Kumar if it doesnt work try the steps by Sebastian EngelIts working now on my envi using Eclipse Galileo +added patch for error looks strange +Added the call to set the defaults before the connection is created. +Patch looks fine to me +Added try catch checks for all the modulePrefs element as get metadata was failing if any of these elements are catch MarijanThanks +KeeganSuresh the patch does not apply +last patch before merge +docsexamplesjms have not been updated to reflect this fix +The patch adds a mixin which can be added to a zone to periodically update the mixin to load the zone by default without the need of an event handler +But I do have a theoryCould you please test the last patch once again with debug logging enabled and send me the log +Closing this as fixed +we need to clean up the rsync list first otherwise broken artifacts might be brought in fixed +The following component were corrected calendartrunkuicalendarsrcmainresourcesorgrichfacesrenderkithtmlscripts contextMenutrunkuicontextMenusrcmainresourcesorgrichfacesrenderkithtmlscriptstrunkuisrcmainresourcesorgrichfacesrenderkithtmlscripts tooltiptrunkuitooltipsrcmainresourcesorgrichfacesrenderkithtmlscripts treetrunkuitreesrcmainresourcesorgrichfacesrenderkithtmlscripts these files are not corrected yettrunkframeworkimplsrcmainjavascriptdndtrunkframeworkimplsrcmainresourcesorgrichfacesrenderkithtmlscriptstrunkframeworkimplsrcmainresourcesorgrichfacesrenderkithtmlscriptsdnd trunksandboxuisrcmainresourcesorgrichfacesrenderkitscript datatrunkuidatasrcmainresourcesorgrichfacesrenderkithtmlscripts trunkuisrcmainresourcesorgrichfacesrenderkithtmlscript panelbartrunkuipanelbarsrcmainresourcesorgrichfacesrenderkithtmlscripts scrollabletrunkuiscrollablesrcmainjavascriptcontrolsgridtrunkuiscrollablesrcmainjavascriptcontrolsgridtrunkuiscrollablesrcmainjavascriptlayoutstrunkuiscrollablesrcmainjavascripttrunkuiscrollablesrcmainjavascriptcommonprototype +Took an approach similar to to spawn a sessionrequest for the refresh. +It breaks for older versions of hadoop +Committed to and trunk +Note that is an unsupported and release I suggest try your code with a recent version +What if returns false for hasMore and then returns true for hasMoreThis kind of chaining is a typical pattern for CP hooks +Ive tried to repeat this error but gcstress either passes or dies with different errors for me +Tagged for does this problem still occur on the latest development version +For now are we good with the conversion of some vs active servicesLet us do that change these on a basis. +The changes tests are based on the latest Qpid changes +Issue is currently with the Spring guys targeted for rolling issue forward to +However test fails with or without the patchThen the patch doesnt fix the issue +Hopefully will commit on monday +Lack of time to resolve moving to TBD +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved +Please choose the one that has the better description and close the other as duplicate +WillemAh yeah of course +Sorry for the spam +Includes Simons patch to create the filter I think you can trash the offset attribute reference in there entirely just dont mess with them at all +Ill see if I can keep both easily assuming its ok to keep them exclusive +I fixed thatDick Can you put this patch on? We can get it in faster as this jira would take more time for reviews +Bryan can you please shepherd this bug +So as my previous remarks indicated we are running without validation for the Sun JDK and with validation for all other +The patch is small enough that Im tempted to just commit it +To me it seems the right way to go +Ari Can you add a test case for that? with test code +The target repository can check that the message comes from the same underlying data store so anyone who has access to such a code already has access to the related binary through one of the repositories attached to that data store +Will extract the section home vs base for +This will be avoided now since small chunks will be sent to the aggregate functions initial methodThe code checks for a combine plan and if it is present it replaces the and in the combine plan with which is a combination of the two customized for streaming small bags between the package and the foreach +Oops read it too quickly +Hibernate must internally translate this to ordinal parameters which seem to be but that should not affect the user API level at could you please specify more details for your scenario? What exactly are you passing in for which HQL query string? did a little more work trying to track down exactly what the problem was and it turns out the problem was Hibernates bad error messages +I was asked by Ted Husted to permit including in the roadmap for Struts and as a measure preliminary to that to include the cookbook examples on the FAQ page +closing all resolved issues for previous releases +Patch applied thanks. +Verified on JBDS. +Upgrading to did not solve the problem and gives me this error patched version of AS with still has this issue. +Im very interesting in helping you to resolve this issue because it is blocking a part of our development +Undo +Whats the serde for the case +Anyone had a chance to try this out yet +Latest patch generated with git diff binary which can be applied with git apply +The new patch looks good to me +This was working with richfaces and is not working anymore with richfaces which is a regression from my povI cannot create any simpler sample app and I am fairly confident that the sample is not too complex for you +Initial patch to add the to the related explained on +Ill make the changes youve also have an additional wish on the list Can this remain public API so I can use it from within my own app? I still need these classes and wouldnt mind moving over to this its cleaner than my own +If you DO specify a bundle name we dont retryThat said we should catch the and pull the out and throw that instead +probably not wise to close an issue when not yet completeUntil the mailing list is created notification scheme can not be created +Thank you NOMURA Yoshihide for reporting the issue and the initial patch. + +Yes pls close it +This comes in handy when you have databases created with different permissions or if the warehouse directory has permissions like +Ill leave it up to the committer to decide whether it should only go on trunk or whether it should be backportedColm +If this needs to be reconsidered please raise it with the SOA PM team. +Now everything is correctThanks again for commit +Welcome to volunteer driven open source +Patch committed to and trunk +I loaded it into as WSDL is available at httplocalhost custom validator is at httplocalhost the FAQ I have found the problem +I am trying to get images to put them inplace of the corrupted ones +Can we resolve this as FIXED now? +This is interesting +To change the order an RFE with Eclipse should be created +there we go no need for the beforeConfigure hackery now +Added detector class in repository to detect class with static method to use a shared instance of this class and added an extra method to to detect andor retrieve the +What happens if you add default click propertiesthanks Malcolm it works with file thank you +The code that Jian wrote is working on the RM as well as the AM I tested it +works +This patch removes snappy from which allows avro to build install and function on mac os x +patch looks good +It works at least occasionally +Moving to if its now possible still a problem since its rebinding to old name also after edit the parameter +I agree with Jeff +Bumped off +Doesnt sound to difficult +new mantis jar with the diff output for for fix +Hey Alan +Why notThats fine too +patch doesnt apply anymore +Can you try CXF? should be gone in exception is samethe class at line is same than version +See linked for more details +Patch works for me +The fix eliminated the deadlock. +Please reopen it if you think otherwise or even better reopen new with new specific issues or new patches as necessary +Resolved with spec update in r. + +However this means we cant create symlink in install for RPM build +Sadly Ubuntu LTS uses B +The alternative is a global structure as you have currently implemented but you have to be careful that there is no possibility of of session cookies between different connections and I am not sure there is a good mechanism in to achieve that +Steve +this sounds indeed very like that the issue you mentioned +I checked the other tests by hardcoding maybe to always wrap with at the end +Thanks +Feel free to close this issue. +Weve been hit by this problem too +Alternatively maybe we should use + is Eclipse WTP HTML validator +Do we want that too +ok +OK +Note Whether you declare the dependency on with provided scope or not the JAR will not be part of the plugin class loader +fixed in rev +looks like we put out against a HEAD build of +Required to be applied first +No Ann and this will even be the opposite indeed the goal of this new algorithm is to match as many new violations as possible with old ones +Committed to trunk +A patch with the +That isnot the first approach needed. +I still think we should convert from error codes to exceptions +I decided to go ahead with moving initialization of RTT fields from init to Java constructor so that we can eliminate constructor call node after inlining it +Ill do this fix as an additional item after getting in +Im sorry but it is completely short sighted to have a single use env var like this +reading a presumably longer table and parsing a presumably larger number of octets for each double would result in a longer loading timeWhy should the literal case need parsing? The compiler already parsed the source when the class was compiledI dont see why the literal case should use more octets eitherI agree it would be nice to know the source of the conventional wisdom maybe the assumptions it was based on no longer hold +Tested with password and that works fine +Five datanodes with fusionio +This is really quite important for my project so hope to get to the bottom of it +Can you please add an additional check to see if ANSI is abled and if not display a warning that this command will be +Currently I dont see the gccjunitant folders under tools directory +If it cant wait then we can just roll things back and add the feature in +We have the hell that is the +Carl Updated the patch with the suggested change +Thank you for passing that along +Also Im not sure why youd ever want to go from a later version to an earlier version unless youve mistakenly upgraded in the first place +For JSF project there are two similar nodes with labels Web Resources and +Contrib test failures are again unrelated +Hi JacquesI have uploaded a new patchI have tried to do those steps copy the patch into the root folder of patch p copy the APACHEHEADERFORXML under frameworkwebtoolsconfigLet me know if you get some more issuesThanks n not intended to be committed yet +I think I fixed it but I need to be careful since this could have consequences elsewhere +Can now run with mvn jettyrun +I tested using the nightly build and it worked very wellthanks +To be revisited if we want to remove the dependencies and fix the other hardcoded sentences +Awesome to the proposal listed above +Yes I do +Patch attached +Since we use java queues for internal messaging vm is obsolete +Attaching a patch that updates the exclude map with additional interfaces and methodsIt still fails but complains about fewer unsupported methods +Tez Engine is definitely data type aware +No I dont want to change the existing API the idea is to provide an utility method which can be used by all resource provider implementationsAnd change the jcr implementation to support this version handling +The basics The part of is in srcjruby like the process logic I mentioned +Ive found the default Intellij code style passes the AS checks +Attached server soa er server logs +Patch attached for the factored out component and the updated demo +It will be useful to bring that info up to date +Andre is on PTO issues moved to PR fixes the issue by cleaning s cache in the resource change listenerThe listener is only active for WindowsThe fix will also work for a git project if the plugin is started +Just ran into this issue +Rescheduled for because a workaround exists +uploaded a patch for a code review +Thanks Ted MatteoAttached v to add back the comment line +Is there a reason not to create an attribute similar to the attribute to the element? Im just curious at this point +Since this problem is in betaand not xalan please follow the same workaround incase you are using themilestone builds of or Regards I patch the? Will this work if I edit its content by navigating in it with and replacing its content with the new properties file? Or will Xalan complain about JAR file alteration if theres some signature checking +I take it it can access this new functionality from the latest download of cocoonregardsAndrew +Shravan thanks for providing the backgroundI still believe that we need to make work like any other tuple +If it should work +The issue can be closed now +I also see this problem when running these testsuiteintegrationbasic tests checked and this does not happen in was introduced by ve attached a simpler WAR that reproduces the issue +fixes this issue for this patch applies only to the x credited Uwe in the file +I have resolved this issue as both the documentation and the EAP Release Notes have been updated +Another related issue that still needs to be solved is Im closing this issue as all the work in here is done +Sorry it took us two days to set up a working build environment for the castor project itself so that we could apply the source patch but now we could try the fix and it seems to work fine +Updated groupId for all files with per Carlos request +It was discussed to add this to the Corpus Server +Si Quite right thanks for catching that +bulk close of resolved issues. +It will be better to have the job tracking URL and the task ID instead +In your forms the to attribute in transitionButton elements should be called transition +Changed the logging logic +That means if a client invokes a RPC to read some portion of a block from the datanode the calling thread will block for the entire of the RPC +I only had one attribute and I worked around it by using an extra element instead of the attribute +I would like to have a working example maybe a quickstart done in this style +Thats unexpected but I can imagine that a cleaned path could lead to different treatment of the have moved the path cleaning into itself which will consider a equal if either the URL or the cleaned path matches +Please could you send a message the developers so as we can continue the conversation thereYou could indicate what are your skills and what those ideas are +I have tested response code on Firefox and Safari in Mac and the general behaviour is for the browser to die and present a blank screen instead of rendering the HTMLI am not sure a HTTP Status code is required in this instance as the action and field errors map which is sent back will automatically give an indication of validation issues on the screen +Please have a look so that we can continue the discussionI havent made any test at all so I doubt it really works but thats not the point here +Gary logging really helped me in identifying the issue +o s your problem with TC iima? And forget that mug of mulled wine +Im going to mark it as closed now +its ok now +This makes sense +Id rather document the tool in GAE Javadoc and FAQBest Practice +I will close this open and reopen if I see it again +fixed with revision able to repro this on rev Downloaded attached schema and invoked d +bq +What about refactoring messages from resource bundles +Committed +fix version needs to be changed +doing that now +I looked at the code and found that in fact it leaves the children versionless if you give them the same versions as the parent when asked +Could you please run and test targets and post the results here +do not specifiy a solution at this the current code should work +Sorry meant to link to +Just comment out the cp and uncomment the dhcp me know if that solves your problem +Also in the application changes required section Ive added a recommendation that both the client and the server are upgraded to or later if the error code is significant to the application +I cannot think of any side effect by this change +Thanks for the clarification Shane +At lease Spring Framework shouldnt be there +Fix for adding versions and apis +OK +But since same logic was observed and therefore for safer side applied the check there as wellBTW for now I will prepare a patch for changes +There are two options for progressing this further Have the oracle datasource enable statement caching +Im confused +I think we should document the SSL config as well as a plain HTTP config and a plain HTTP config with BASIC Auth +I just committed this +RS is stoppedT +here is the ported patch +Subselects are not supported within select target lists +When I click on the browsersim icon I get the operation not available message +Meaning there will be C versions which support bothOnly if they use thrift and they ever want to define a that apply to one of the component of a CT +Ive attached an amended version of which recursively unwraps the response if necessary +svn metainfo was broken so after a completely new checkout I finally know what you are talking contains the missing files +I think HS has similar problems +this is issue is actually invalid +adj +You can also change and back to subclassing if you prefer thatI only changed them to use a reference because I moved the implementation to an impl packagecheers Roland +Is this for both and +I will add a spec for it +Not sure we can fix this in this release without downgrading the testng plugin in Eclipse which doesnt seem to actually work +Any help would be appreciated the API has been reworked recently to take advantage of Java features +flt files before uploading to Guvnor +Resolve comment since bug fix was moved to separate jira and port will closed. +on switching it downIve opened for the same please review +Jody I belive you are incorrect wants to prod the service to fetch the capabilities into memory again +I think what would be ideal is to create an authoritative documentation as part of our website on Bigtop deployment and configuration +The in each subproject contains scope +We have the same problemTwo different JDK on the same OS and when maven is configured to compile on the old one jasper plugin still compile on the active one +When it was discussed on the mailing list I was thinking about checking if we write or localhost as the server address in ZK +I reused an existing function which updates the counter if initialized +Committed as revision +Let me know if it doesnt work for you +Well thats a very bad design decision but its your decision to make and I will accept it +Assigning to myself +Please reopen if you still have issues. +I will try to fix this bug +Its also something Im planning to do in the next few months to reduce the unnecessary work and garbage littering +The Job did not kill off the map task that it had running before exiting +Although I dont know where to find the sources. +Can we explore a hybrid approach? Suppose the client passes a representation of its viewfs mount table in NN RPC requests +Im still in favour of the left oneIf we go down the route that the box is synonym for dropbox we could well question the name as one could consider that to be an online storage for PDF files +I have committed this +This is because it still uses +Here is the diffMy apologies on being behind to test the actual jruby jar as built from antStill looking for a light way to test during the build +Removed from the patch because it is not related with this bug and moved to +Looking back as Pete suggests catching the seems reasonable especially with the upcoming logging improvements +It will render features for you and you can put them on top of other layers and I do believe that geoserver renders each layer in order no matter what labels have been rendered +Ive applied the unescape patch with the interpretationClosing this issue as there is currently nothing open on it. +When one css file is already opened in the editor and there is enabled toolbar icon selecting another css file in Package Explorer disables that icon and its ok +It doesnt appear to be there anymore but I highly doubt I browsed their classes and found this random action and decided to insert it +So if I understand this will let you read sequence files but not use it for map reduce input right +Thanks for the patch! +Ive updated the documentation to make no more reference to the fixed conflict manager in the list of available conflict managers since it cant be used by any other mean than the one documented on the conflictmanager page. +Currently that detail is being delegated to the classI tried to apply your patch but there are some conflicts could you please create an updated patch? this is handled differently in see for instance has been done recently see +For extra credit we could add a test modelled on or which calls all methods on closed objects and verifies that they either raise the exception or +Marking this bug for hackish work around to the class loader issueThis patch tries to use findClass to load the Driver class +This is a side effect of fixing. +release bulk close +Has there been such proposals for Struts? +You are correct in that it will not work for lower case values +In RS key columns will not be removed column +I have been using the patched version of the bridge in production since start October last year + +for the patch +Committed work with revision to add check for overriding the default checks for the default with revision Files effected A rosettatestssrcorgjbosssoaesblistenersconfigmappers rosettatestssrcorgjbosssoaesblistenersconfigmappers rosettasrcorgjbosssoaesblistenersconfigmappers +When we translate UUID to internal id in it needs to include those removed entries as well +Thanks Carl +I will try it later at home +Please verify +configure prefixusrlocalts CXXFLAGS it still a problem now with the latest code +verified +In an effort to get unicode characters working a long time ago we started hardcoding string conversion using and since then its become more apparent we should be using the encoding +Closing since the params are set to enable windows support and can be changed as needed for longer names +The problem is more that it might not be able to pick any of them because the none of these interface should be preferred by the method selection +Its not ideal surely but it would not currently be much of a burden and will provide substantial benefit +In In +All methods need javadoc once agreement exists in and +Attached v +Old JIRA because the old spatial contrib module referenced here no longer exists as of Solr . +cancelled +While probably not fixed I think this problem is mitigated by the work I did inPeople upgrading to should not meet this problem or at least they can work around it by setting the thread configuration of the mass indexer. +Same steps as above with the additional trace enabledI tried the steps with a clean workspace +Unfortunately I am near the end of my release hence very busy so it will take a few days +and will cause difference in compatibility of types and which is the issue around difference in compatibility of types between Network and Embeddedis required to be solved in order to accept that differenceNow it is not completely decided extent of modification in and I will post my plan in soonIf you have some opinion for please comment it +Thanks! +I just realize that the uses statefull page information to generate the URL +NoIt is not really appropriate to migrate subscriptions across organisations and could well breach a range of privacy lawsTherefore unfortunately you must request folks toSorryUpayavira +Linking to separate issue for rollback +I think I see a solutionIn your precondition if Ctrl is used then change to true and return false to stop the Ajax call +Those files were added to the HEAD branch after the branch was created +Tell me if this is required in a different format +JoseAbout the patch its ok dont worry about that +Closing the issue now that it has been resolved. +Added test case to regression tests. +According to the JDBC specification a value of indicates that the operation was successful but the number of rows affected is unknown +My patch was for not Hibernate +Im working a way to remove unreferenced segments from the pool +Thats the reason xmlnsdwr namespace validatin error is displayed +If still relevant reopen or create new issues. +Patch for improvementMind attaching patch for trunk +Patch applied to to a lack of knowledge I havent close the bug after fixing the issues in the last open active branch +Closing old resolved issues +Theres been a bunch of debate on this topic +Remove the sentence There are rules of precedence +bulk close of all resolved issues. +A separate context is now created for each session and these race conditions no longer occur. +JB is right +This is going to bite many newcomers +Defer all open issues to +Should copy row passed in as well +Are there any junit tests for it? +Sorry Ive marked it as a major +If is acceptable you might want to apply that patch before this one or just the order of the Remoting chapter later +Feel free to assign tickets and to me +Never blame yourself for SVN commit issues +Right now its not working like that unless you turn off the cache as explained in the patch +This implies that the project defines itself in what profile they areThis is no longer configurable server side work around of specifying has stopped working in sonar +does the google plugins not work on kepler +it is then +port to + +Controller to Broker RPC added with unit for the patch +This patch has some potential bugs +Also sudden enforcing is going to break a whole lot of code out there in production +This seems to have gone away Im not familiar with what configuration changes have been made on sigma but the request works fine for me. +Please check on XS to make sure if this is something specific to KVM +Committed to trunk in r x in r +but if we read description we dont have to update it when new statuses are addedremoved +Now it works fine +Initial patch for feedback on progress +Does v pass the tests +We can tag a tagging is easy +Problem Has Been FixedAnd Mirror Has Been Restored +good plan +Do you agree Thejas +I also cannot replicate the issue in +log showing error when viewing page with portlet no longer was introduced in commit babacfcdfdefcfd for manager feature +be aware that needs this additional private package to but i think the patches would screw up at this point because they interfere withSo has to be changed to privatepackage +Can you please review this ince more and maybe have a close look at the two sample XML instances given as I cannot see any difference +I will review your comments more tomorrow +Im transferring this to Xerces since the parser should be catching the error +This should be fixed with both break and continue implemented +But I dont see any better option here +Tim could you please apply minor fix to make all test pass +Well need to revisit this later +am not sure if shutdowntrue is useful when running in server mode +So all issues not directly assigned to an AS release are being closed +I committed this +The updated patch includes a test case for the new configuration points +Thanks for fixing them Kim +The file must be overwritten +Some getters and setters could be consolidated into if that makes sense +So that could be spit back to the user in this case +The patch lgtm btw +The problem described in the last comment is actually linked to and not to this issue +In the case thatd be a newly opened for every segment in the index every time a new segment is flushed +If you just need the raw stream use the jettyhttp component +Part of a bulk update all resolutions changed to done please review history to original resolution type +Please close or retarget to a different fix target as appropriate +For me Michaels patch worked too +we already bumped Blocks version in to fix other bugs so we dont need to do it again +If you disagree let me know +the XMLRPC should be tested before release +I really dont have a specific use case +Sounds good. +Thanks for looking into it. +How do you think +Needs more research for the server side +Duplicate of +but I can say this error has existed since at least +This patch removed it +A separated JIRA created for tracking the unimplemented functionalities in JMS portlets +If an app cannot handle this situation it should create its own config and set the correct value of on submission +As a I think it can go on as well as TRUNKFor future reference its helpful that the file names for patches match the issue number or describe the content rather than +Its just adding headers to code +Maybe it relates to this just a guessI dont see this problem of the time its about of the time +We can just run the patch on solaris machine and see if it works +Downside of this approach is that it would potentially bloat the reports even when rebalancing isnt required +Maybe not tying the reader to has some value but Ive never heard anybody ever took advantage of that +Yes I downloaded the latest snapshot version and now the example works also with the referenced from within the camelcontext +Please log the exception instead of doing nothing +My feeling is that these failures are unrelatedCan this patch pls be reconsidered for a commit +I also had to update the logic a bit that did the actual bundle update and package refresh since it actually was starting and stopping the bundle twice. +No patch attached +Oops sorry about that +is available at the time when the exception is thrown it would be useful to throw a new with page instance instead ofCould that be done +There could be a problem with the wsdl +Both on jobtracker and jobhistory webapps +Ok with me to check it in +Canceling and resubmitting patch to trigger the QA build +Added to trunk +By the way +Verified in JBT. +Marking Fix In and leaving open to make sure this gets merged to +Thanks GwenI applied the patch in r with some minor modifications changed scm url used for testing added assertion for checking the build result page when building the project group in the selenium test +Unless you veto this of course Hemanth can you quickly respond +looks good +Please do not use the Fix Versions field as has been described in the forms unless a fix has been committed already to that version +Similar failures also in jaxwseewjrpcliteralsec resolved issues for already released versions +Its working and its fasterFor the patch itself I think HRS +Its a big patch since it required me to move all the Object methods up into and add stubs to to bind them +As I said in the discussion in the mailing list thread this is something that needs to be implemented in a consistent way once and for all at the Axiom level see +Ill leave this issue open for a few more days for any further improvements or fixes +My bad I fix it. + +So I am trying to see if there is a bug in or we should be sending a POST request after the form auth success redirect rather than GET +The swamps streams etc +Thanks Rishi +Thanks for the report +I will integrate the patch later today if there is no objection +So the same set of resources is reused for the next test case class and so forth +It is a service and it needed content permission checking not just entity permissionPlease let me know how I can helpThe new permission code should be implemented as simple methods and be documented for customization +I think the leak is circumvented +Instead of subverting this jira maybe we can open a new jira and discus how we can get scattergather into the APIAlso it is better to do the mmap only once for the entire file instead of mmaping only the relevant offsets on demand +lets make it a different JIRAashish results shouldnt need to be changed since ppd is not enabled for them +yeah no problem +Ill post my results +Committed toRavi John thanks for adding those two new test cases they are just what the doctor orderedCan you please port them up to v trunk before we close this jira? Thanks +Great less is moreNow weve done a release cycle we can release v at leisure and the rest catches upWe could have a module which depends on which is a ed POM artifact +I am able to reproduce tomcat jboss as +I committed this to trunk and +Its really too good to be true that returns a string +Runping could you please try things out with this patch? This implements the pseudo code you proposedThen milliseconds timeout is too short for connection setup +Any chance to add an option for reporters +In particular Im thinking of MS which I dont think are exposed in the same way UNIX groups are +I just published a candidate for release which well promote unless well find any show stopper issues +When using the annotation the description is completely ignored and only shows the method name as the is minor please +I shall check and find a fix +Please post full complete code for two pages and reopen the issue +The only thing I can think of is some kind of class loading problem caused by bundling a different version of jackson in the app +Adds logging +Yes +Of course the connector must be http +is this not going to happen? I am looking to open source something that depends on thrift and it uses maven to build it would be a lot nicer to depend on thrift via the pom file vs my providing an icky shell script to do a mvn install with the distro +camelfrom wwwdistThis can be done by anyone with the camel group id which should be all Camel PMC members at least +Once i reverted them the npe went away +to commit +We still have too many archetypes that could be consolidated +Will keep looking andor try to get together and start cluster to make sure the this is a real problemPatch is pretty much ready for verification +Yes you are right Knut +defer all issues to +fine +Code was committed in ! +Fixed manually in maven +Now do a couple of restarts without cleaning the data directory +this is yet another reason to use crawler commonsKen I wonder if you would be so kind to start a thread over on devnutch regarding the atmosphere going on over CC +Please kindly review it +I have a fix and test ready +OK I will dupe this to and then merge it intoThanks Ivan and Andrey for looking into this and finding the solution +This issue is fixed and released as part of release +Still the solution is to use JMS selectors which offers this out of the box +Right now it does indeed swallow all errors and return null +I just committed this +I believe this works Steve can you look over my usage of the DNS class +Thanks HarshaI finally reproduced and fixed locallyYour patch is intrunk r r r r +For me it is reasonable only for isVisible conditionSince when I want to know whether an element is visible I do not want to struggle with the fact whether it is present or not +Could possibly reconsider at +Keeping as this was referenced later in the cant reproduce the bug +also the namenode ui does not come up if the configs for the kerberos do not have the correct values +Just added a log line in xml and added encoding attributes for the copying +One thing to keep in mind when using is that the grouping separator may lead to unwanted results +With the other one thats probably enough reason to pull the latest mochiweb +Not a bug in Spring LDAP as far as we can determine +Cordova v returns +Please verify whether the problem is resolved as you expected +HiramI have been very busy with some other items around here so Ive not worked on this for about a month +Fixed. +released issues. +Source modified +temp assigned whilst removing from In Progess state +it stays where it is meaning the comment is not above the th dependency now +I hit the problem under Jboss with Hibernate Annotations +The cleanstate target is still used by the release machinery when building both sane and insane distributions +I think it is something that needs to be done but perhaps needs some design work especially in relation to how we may want to do dependency isolation in the future with or something else +I just committed this +GerritSorry this fell through the cracks! Just noticed this ticketThe ability to specify just one column seems very limited +In is not introduced +Would it be possible for you to put an explicit include into thefile? If that fixes the problem I will commit it to cvs +Master will not take because it is obvious that large changes are afoot +It works fine if I use any of the deprecated goals just not the new single goalIt does sound similar however they are trying to execute the assemblyassembly goal which does work fine for me as I stated above +Thank you Ivan +Thanks Peter +alexey ist this still relevant +The testHandler and interrupt method are being invoked on two different threads correct? If you want to maintain the as a instance variable youll have to synchronize access to it only one thread can be interacting with it at a time +Add test to verify is set on allocated containers +we will welcome contributions +Ok will check as the handling that is now being addressed in where a descriptor is being introducedI have also created for the descriptor as in addition to the handling there are also additional changes required +work done +As noted Ive fixed most of the tests but is still open and I cant reproduce. +I do not get javadoc warning when I run testpatch +I just committed this +For that reason I would not be inclined to hold up the release for this bug fix +I agree +This occurs outside of Eclipse +It would be better that to leave Chukwa Archive as it is and create a new archiver class to extend on top of this one +This one is correct +if no mount options are specified then it is mounted readwrite +I was working on completing the github setup over the weekend but unfortunately didnt yet have time to complete it +The two parameters used for paging are pageScore and pageDoc +Attached should add this was looked at as part of but I think in that use case we ONLY looked at classes that JAXB already knew about +Arent users able to edit the actual filename so we can fix it already +Hi HansWould it be worth including some documentation that describes how to include the jspcssjavaetc +Thanks for the report StephenIn snapshot +Resolved as part of +I continue to see this intermittently on Mac OS X running +But for dev it would be ok +This patch conflicts with the patch pending on so whoever loses the commit race will need to rebaseLooks like Ive an opportunity to show a bit of my evil side +Updated to lock usage of selector keys +v +Does that seem OK to you +I think we should make use of this in Oak as well +Issue is ready to be closed. +This task requires the newly updated spreadsheet to already be applied for this patch to work properly +Updated version with correct arrow and more contrast borderlines +This is good +Client side use server side configuration that seems counter intuitive +Tested and works well +Please reassign to Prabhat when component is ready +reopening to fix fix version +The mini cluster tests exercise this change so no need for new testsIll commit in the next day or so unless there are any further comments. +The Directory interface already has an in the function signature so that would not break back compat but would yield the correct behaviour +When the new SPLIT BY FILE syntax is used I inform that it shouldnt split files which is the same as its behavior in prior to this patch +Back in the days when using Java security you could hit issues when key lengths got too big +Also with timestamps per second testing result it looks very promise to me +Great! Shall I take steps for this issue? Would write a list of components I see and their as a comment and after we discussed and adjusted that structure if needed I can start to map it the current trunk +Committed to trunk +Integrated Web Beans +invalid +With it is now possible to register a through the MVC Java Config +The initial code base is now in trunk +Apologies for the report +Can you tell me why you claim that does not implementWhat I see is that is implemented +A value of true just means yes this is auto create +Patch applied at r thanks for your contribution please verify +Hmm we already have a class +Upgrading to struts on introduced the problemAs I posted earlier the debug code suggests there are some class cast issuesSet property public to trueI specifically specified a custom className on the attribute and expected to see that custom class name in the debug statement. +I was suggesting you could pick that code out from the similar call for the first one +But If the fsimage is not compressed It will be better for limit transmissionWe use the method of limit and compression transmission in Baidu +Moving items not being worked on afaik out of M Fix issue out as part of JIRA cleanup +Doug ok I will split things in different patches and open a set of new bugs +Are there any other concerns over this design? A review of the DFS side of things would be appreciated +This has a rather high ranking on my TODO list +This is a file containing information on the compilation errors reported in the Eclipse Problems view +Enlight me if you have a better way +Please review the patchRegards this patch looks correct +if it is still a problem +Normally he doesnt see the issue when running the Gradle build so we thought the failures were OS specific +what if its proxying to something other than +Attached great Shakeh! Thanks +The simple solution would be to just turn off the declodeBlanks option but Im not sure why it was added in the first place +If reopened it shows only a blank diagram +objects are cached in an externally configurable cache +Kabir what am I missing here? I dont get why we are using aopc and javaagent +Maybe I just posted to +Its always been unsupported to bootstrap a second node into the same token arc while a previous one is ongoing +Also discovered its in just a different location +what about using the id central +i had figured that out already +This is a candidate for SOA +How we handle them when user upgrade it from to +Checked in new and modified classes for annotations tests Rrevision +Moved from to just after merge of old branch into TRUNK +Feel free to submit a for the that should certainly go in if the Ant dependency is soft Id like to consider Wills changes forVelocity with the stipulation that an appropriate warning message isdisplayed when those who dont have the necessary dependencies satisfied try andinvoke the parser target +Backport complete +Will commit after tests pass +Linked to underlying Android issues +Youll find the complete patch in the attachement + The clone is far from perfect but lets see how hadoop qa reacts +Its taken at the acegisecuritycore level in the source tree +The patch looks good +Max I think that deprecated tag is used to alert users that something will be removed in the future +Thanks Doug +Doh just realized it was to improve and not core property value semantics of the xml config +I think one persons dislike of generics is not a sufficiently good reason for rejecting this request +I honestly have no idea because I havent been following Oak dev for the last months +Yes please love to have it for +Will apply cleanly after your patch toSNAPSHOT will deployed by jenkins buildReopen if any issuesThanks for the patch +I also agree with the your last statement +fixed in rev! +Committed to trunk. +Do you think we should create another ticket about these quotation marks as wellYes I thought Id fixed all of those but I must have missed some +Committed revision +It is not even possible to instantiate the class with JDK +As a part of service start it will renew all the tokens +and please do not forget to upgrade the Maven POM dependency as well +Applied at r +Forgot to add this +Eclipse project for the EJB test program +Please close the issue when we are done here + +Committed revision . +Our app was stillyet again rejected +wars lib that should fix all of thisId not spend too much time on this weve lost a fair bit of time on this already +Pull request merged +HiI have the same problem and would love to have the patch accepted +Here is the patch that does itI also added assertions to make sure that blocks are really corrupted +I am glad that filters are something which is used atleast somewhere +though note that srcmaven is still showing Lucene dependencies +Added the information that VPE supports custom tags to Doc +Fix ok +Let me know if it works for you +This was really a big problem for me because my boss want to release an alpha on monday +This important part is that depends on CL +go to core and type mvn clean package jettyrun +Checkout latest revision of ofbiz and try it +For me it was a problem with the transactions +Update the patch to be based on will change and As to is more general exception not specific for this issue +Heres a patch for trunk +This should be better for our users because users can still specify any hadoop versions on the command line we just strip it off to get the major version and append +getsets are being created when people have seen usage requirements to do soThis appears to have been more along the lines of when not enough function is being exposed rather than too muchAre there are specific areas where you think we are exposing too much or are you looking for more information to be exposed +Version Released +Other option was to make a small change in clients +on patch vMinor comments on logging +Thanks Patrick! +lets keep it simple +This should explain why the build works fine on Unix systems but not on Windows +Please discard all the other patches + for the patch +for previous version not to previous version not to commitadds ivy dependency for jackson json the contrib patch the eclipse classpath hasnt been updated leading to s +I was not setting to +Thank you for taking the time to read the overview Kai +based on Shalins suggestion to add an attribute for the locale I prepared another patch implementing those changes +This is not really a bug fix +So the updated version of Sergeys patch and patch for the test was committed at the revision +Superseded by +I just committed this. +Even if the solution shown would solve the problem I do not like to cache resultset inAs SQL queries for all entities of an extends hierarchy are almost the same already and we already know the correct class wouldnt it be better to always return the Object array needed for this class with the first call to load +The goal here is just to make checksums independent of the files block size +Please open a new one +on commit to TRUNK +I cant see a way to resolve this in a generic way +yeah youre right +Attach my patch to remove duplicates +This happened the first time I tried to run a groovy file after downloading the plugin +it is possible that the checkpointed HW of one of the replicas is lower than the others +The visual theme code provides default values +Although it might look wrong the current code is correct set must set the stream on the content node of the current node + +under Tuscanyjavasdoimplsrctestresources directory should be the serialization format based on the +Heres the patch for anyone can review both the content and my english +Changed the behaviour will be included in the next release +There have been a lot of bug fixes in this area since +Currentlythe double separators were not introduced by the VME alone +If that fails I may need some if we could get rid ofmavengroovyjars that does seem to be unusable that would be Paul +Thanks. +Thanks for the clarification +I just tested it on trunk and and it worked as expected +Useful for both ptest as well as for running standalone tests +Hope vacations going well +So I still think we need thisAgreed also we need a fix before is resolved +We could document that this is the group to which the tasktracker belongs and qualify that if is used it takes more meaning in that it should be the group owner of the binary etc +Attaching patch which marks the running job row as green. +This fix and other recent cache fixes will be backported to other releases some time in the near future. +namespace test for top level selector option. +If format was successful it was returning false +Fixed +If this patch is committed Ill resolve this issue as INCOMPLETE +This is actually caused by the that you refer to from the trigger +This is because the same check happens inside get which is always called before createLog in get no +I submitted a patch on +I created a new issue for the to accept the +Yeah I see the issue +No changes to key objects +Faron has fixed this I believe. +If a is thrown we then know that the variable name is not a system variable and we then try the other cases such as page variables and request headersThis change greatly increases the readability of the getValue method and trivializes adding new system variables +Hang on let me purge everything and try again +Revisit any notion that said region is at said location +mrepository the build runs fine +Can you reproduce this by just deleting the log directory in broker ? +Since people are unaware of what the remote resources plugin actually does wrt to NOTICE files Im attaching an example NOTICE file that was generated for Commons Validator by that it would be good if you could find a link to that email thread re LICENSE because that is not my understanding of how it worksAIUI we always distribute software under just the APL +Reproduced on test case something about handle +looks good +But if you could I would really appreciate it if you could test with your current checkout and verify that it does in fact fail to properly handle the property generation and then try the latest code to verify it does in fact address the situationI apologize about the snapshot repo +I just saw some old stuff I know form kodo so most probably either JDO or homegrown stuff +I will need to take a deeper look in trunk +In this model if the parameter is not present the Search Component looks for a array of parameters insteadRight now modauthzannotate puts the requisite tokens into the AAAGRP header +bq +is checking the job tracker if its ok to start a new job so there is no running job instance available +Not a good idea when replicating between hosts that dont support one form or another +While the config for jetty builds the config for tomcat just sits there Scanning for projects +In the future we may want to detect if this has previously been done and execute the attainGoal based on the results +You can add explicit imports in the template I think +The first phase of the view implementation has been +Committed modified patch revision . +Appears to have been a Rails issue now resolved. +We dont even have information how much memory is needed for journal with xy messages +Core connection classes working +Fixed in rev +Cool youve isolated the cause so it should be trivial to come up with a small test caseI mean if the issue is really just the red herringness I can easily come up with a more appropriate failure messageIf you want to argue that this should be allowed I need a test case showing what it is that you are trying to do and believe to be valid +code duplication somewhat lesser complexity +I need to take care of this but it is independent of aint gonna happen +Super thanks Uwe! +Suggest picking up the coding convention of surrounding class +Ill commit to trunk and if it looks good Ill also push it to! This is awesome +Id express the start times as absolute times and the others as relative times in hours minutes and seconds +One Line code changeat line after else blockstart end +Correct me if Im wrong but a single user shouldnt expect to get more than maybe messages queued up +bq +As it stands this is not an issue though so Im going to close it here you are free to open an issue on the broker if you think it appropriate. +This is not specific to CCE moving to Teiid +properties to +I had previously reported that banners were not being used in child we get this resolved before the site plugin is released as its preventing us from taking advantage of this new inherited feature? Strikes me it shouldnt be too tricky to fix but thats easy for me say +We really need this too! We would like to have for instance a product url to just contain major versions +May be this could be done in a separate JIRA +Please use that jira so that we can keep focussed on the MR framework +Heres a patch that uses the default chain for both log replaying and peer sync replaying +Could you have a look at what Ive done and see if you have an idea of how to reproduce the problem with a unit test? The most interesting added test is test +Sorry for inconvenience +Thanks +AR then tries to repeatably use the closed connection never recovers from the situation. +seems I had the right intention and then forgot about it later in the implementation +The patch is the attached file with name ending in +Updated with the new signature also updated files that implement use those functions +The name my kid gave a stuffed yellow elephant +Thanks Mr +So many options + +Cloned for Resteasy +how would we provide the option to? I suppose we could inaugurate a contrib section if you want to write such a tool +We also expect equalshashCode to disregard order +Anduns patch is applied. +A snapshot will be good +mark i applied ur changes and refactored a bit more to use separate connections for each consumer but to no avail +So we need not start cluster balancer right +Fix is in CVS please verify if possible +hmmm +But you might not be able to make be a subclass of +MilindIssue is not the complexity of implementation but that I am not sure we want to support command line aliasing and I want to discuss and understand the use cases for it separately +When I tried the version the same problem occurs as is described in this ticket so it appears that the bug is in Tika and that it is already resolved in the releaseIll try to update the version of Tika in use in my installation although its something that has caused more problems than it has solved when Ive tried it in the past +What do you think we should do hereColm +Ill make the patches for the other branches and fix the tabs +t know wether commenting a closed jira makes sense but I think that this made matters worse because now bootstrapping is not distinguishable anymoreI thought that would signal that a read cannot be served due to CL +This issue was reproducible using some of the update sites described inIt cant be reproduced anymoreHowever if you succeed to reproduce it please reopen this jira +yes the http session is easy to expire so the logout filter catch this exception will be better +Please provide more info or lets close the bug as not reproducibleI used the following commandcProgram FilesJavabin cp +examplesbinwork +If Tajo supports various kinds of catalogs simultaneously it will be usefulBut I think that it will significantly increase the systems complexityFor example when a worker tries to access to a catalog to build a query plan it should get the information of which catalog stores the meta information of the relations relevant to the query planI think that the complexity problem outweighs the benefitsHDFS federation is designed and implemented to achieve the scalability and to handle the single point of failures +I have committed this +The issue is fixed +Implemented this change and fixed default configuration and Metatype Service descriptive texts in Rev +What is woostox? And are they both wanting to become bonafide haus projects? If so can I get more details on them please +Committed to trunk +OK Matt I give in +I think we should change it and make it ANSI compliant +The virtue of your reliance on the isDisabled invocation is that it neatly side steps the reliance on state change events and puts everything into the runtime state evaluations +Is fixed by the latest mina patches that has been + looks good test failure is unrelated +wont be modified for this reason and will still exist as a maintenance tool to delete s without segment information +It might be little hard to read because I have not put the comments in the code yet +patch for the right patch for trunk +It works as described by Nick +Attached is a screen shot of the platform inventory page +Each screenlet has its own expandcollapse parameter +Will update minotaur and +Does EL makes any sense for PHP? I dont think so +submitting +Looks seemless +Heres a first cut +the job option has nothing to do with the libjars and others +Maybe instead of redirecting we can just show a link +of for patch v +The modified version of this document will be available on the next EAP release . +could you provide a patch? With some luck it could then still make it into the severity to critical because it renders BCEL useless on Java ornewer +Tx +If not then we construct a new one and register it with the mbean server +Should we ask Mark to agree to the Apache License? He only sent a code snippet not a patch +Im not going to have time to work on this +Its a bit like the custom namespace configuration I guess +A consolidated patch that Jenkins can test +The last part of this task is to extend this to the eCommerce applicationwhen the customer submits the userpassword in the ecommerce site the system will check if there are agreements in place and if they are foundif there is more than one agreement display a drop down box similar to the one of the order component but force the user to select at least one if there is just one agreement notify the user that the agreement will be usedThe selected agreement will be set in the cart in the sessionNote that this is mostly the same behavior of the backend order entry +Patch to fix +Thanks Jacques for committing the patch. +Oops missed this one off the list needs to be related to a revision of the startup code where most threads are created +Thanks Jeff! I close this issue. +That should work +You are correct in your theory +Attached patch implements as described and includes a unit test +Would you expect me to save it for you for over a yearPlease feel free to play with Suns blackboxIn your report to said you would attach a project for demonstration but you did not +Prasad if you have some time to look at that one it would be great +This will make it easier for me to validate +The change itself looks fine to me +Change log is generated from release note is in srcdocs +Fixed with revision +merged in preparation for RC +NickCould you please pick this up as wellthanksThis is a blocker for release +Updated trunk patch with extended test and CHANGES entry +I switched the dependency of the regression module to the new compendium jar +Applied to branch and trunk. +It can now display a very simple error message and supports editing the of the CASNext step is to implement a status control which supports switching to another view +was just thinking about that +The definition of zero is an application specific thing and thus should not be included in the DM itself +Could you please look at it since I would like to release CR at the end of this week +We need to validate all of wait +for the patch +The issue is still present on karaf but not on +Thanks +modified which i used to generate pdf from +Cool its okay Irina thanks +I just pushed this to trunk! +Ive added a method in the test case which shows exactly the issue encountered when generating the whole documentThe use case is what Ive described in my precedent comment so here is a code example +Thanks. +Otherwise there will be frontend exception +Machines werent upgraded at the same time to prevent cluster downtimeYeah makes sense just wasnt sure how you went about it +to trunk +Looks good +Changes committed to trunk revision to switch from static methods to class instance its a huge task but it should be done soon +Patch against trunk rather than x +Let me know if you need me to collect any debugging info +Willemcould you please update the svnignore accordingly see the attached screenshot for details +Please provide a quickstart application +Updated in revision +So effectively if you specified a custom implementation at the moment youd see the initial exception that triggers recovery there +HiSo you set the bindaddr parameter when you ran GR? I usually just run the GR with all the default values maybe if I set the GR up as you have the issue that I was easily able to reproduce would go away +No I dont have any great suggestion for improvement +Really I cant see the benefits of using the localized symbols in the resourcefile especially when Sun didnt even care to document them since itsintroduction in the JDK +I dont think so +Applied the patch with modifications as described at revision I suggest to resolve this issue as fixed and follow up with new issues if necessary +I prefixed these moved constants with HFILEBLOCKI also had to make depend upon prefix tree so I could get that class +We may need to analyse for validity of the bugs which reported by Jcarder carefully +Its easy +Candidate for to the EWS docs lead +I do have the same issue on linux using chrome on Sqoop project +iocoreaio +After looking at the logs a little more it appears that this error might be thrown from tdataScroller as it appears to show up when there are less than numRows rows in the table and the dataScroller does not need to be rendered +This would be really be nice to get in and make it cleaner to embed ZK. +Thanks +Hi Adrien thanks for the information +hence the mapping from jcr to the file system overlaps for names that are equal when ignoring thanks +And now only some hours later it IS fixedI cloned the git repo and gave it a try Now the works in the contacts example as well as in our real world applicationThanks a lot you saved my day +It will also catch null data passed to handleRegion in AM and I improved comments around the catches in AMunassign enunciating presumptions around how the recovery of a failed rpc close is expected to happen +This has been fixed in +Done or fix later or not an issue now +Time to remove the +After that each recovery pass calls hasget +Would need to track him out +So it may only beMondayTuesday of the week following that I manage to get something backto youregardsAndAndrew feedback from Andrew Dinn +Verified fix with EAP ER. +we randomly suggest names and the +done +Maximal delay should be one second just as long as it takes to create a +Sorry I dont know what happenedThe title is fixed and is for now enough descriptive to in JBT Betae. +bq +I will file a separate jira to write more unit tests for the patch looks good to me +you add a new styleclass that you want to skin you need to document it in +Finally ensuring the colocationAm i clear Ted +thank you for your immediate and easily understandable explanations +A graph created from a memory profiler +You can do this by writing the data to an HDFS directory that Hive is watching as a table +This issue was fixed in roller trunk revision +I created dedicated JIRA issues CAS protocol logoutYour patch cant be applied because it does not follow style conventions +The reasons for that are several +Here is the patch to fix the issue +Fixed in revision I have moved all the test documents to srctestresources and added the supplied new onesFor I used a different file as yours contained a persons nameThanks for your contribution! +Distributions are not going to be changing their defaults +If that change is an urgent problem for you just make your private code change for the meantime and use it until it is m watching this bug as well so I can help if required +The resulting command is different which is why the unit tests were failing +A couple minor documentation nits +Could you attach a document which reproduces this problem? ThanksI think you could see the problem from the source alone but anyway heres a test case to reproduce the problem +But say so if otherwise. +If someone else has the changes fire away +Whats needed here is to document the above info into the Roller Confluence Wiki. +Patch applied +This one appears to be right up your these out for for now. +Bulk close for +hi Fergus The issue here is that when there is a deleteQuery deleteId is present the document is still tried to be inserted +properties fileseg +Renamed issue to match the removal of support for Password Policy interceptor. +You can look at the code its in class +This patch is based on trunk codebase +Is there an alternative? Otherwise a callback will work since I can make it a singleton +It should be shared across all instances +I think it is ok that we have to regenrate each mapping file for version +Temp fix now is not treating valueId as +Tested +This issue is maybe related to. +It has nothing to do with code +Yes I think will fix this issue +It turned out that the jaitools operation was not subject to the same bug but the relevant part of the code was needlessly complicated and Ive replaced it with code adapted from the fix applied to +Hi Army +Same problem here +Thanks for the quick +You are on this already +Pinged MY to see if he knows more +This is almost back to the original code with the minor change to how attribute nodes are handledRafal please take a review of this and run locally +Hi LucianoThanks for that info sounds ok to me +On that page it says Be aware that if a process started under another username than the one configured here also has the shortcircuit enabled it will get an Exception regarding an unauthorized access but the data will still be read +to trunk +Assume closed as resolved and released +But I believe there is still some space for beanutils to improve tooI think the getInstance is unnecessarily synchronized as the is already synchronized +Now that work is underway with AS all previous community releases are +Assert that you get an error message which includes the name of the JVM +Error marker is put by WTP framework that validates xml by schema +regioninfo +The default value is the in cvs +Unfortunately there seam a fix had been attached to this issue that we lost at the conversion from bugzilla to jira about a year help us to resolve the issue can you please write up a short test cases that showcases the problem and allows us to debug where things go wrong. +I have looked into this too a while ago and came to the same conclusion as earwin there seem to be no real gain in refactoring this to use instead fo sync blocks +To me it looks like a duplicate of and passes on my local environment +Looks like this could be close to done so marking it for +I can get this done by early next week +In any case do not cache the pluginrepository is a bad idea thinkabout writing a own plugin that solve your problem that should be a cleaner solution for your problem +Is the code working as you expected nowPlease let me know if yes I will close the issue +I cant reproduce it with the rpm installed on RHEL JBOSSHOME is either varlibjbossas or varlib so JBOSSHOME +Fixed on trunk revision . +Sascha could you please try latest snapshots? You are probably running into +Has a custumer who had licensed normally also had licensed the missingIts maybe a good idea to leave this issue open until its relevance becomes more clear +ClosingI added a note to the Derbysnapshotorrelease page. +It is readonly hence it does not appear in the documentation. +has noted that OPP is deprecated in favor of BOP for +Thanks. +In a case if a project doesnt have only KB nature it will successfully add it to a project path in spite of we cant call it from context menu + +Yes this is intentional +You didnt answer my question regarding your first test dont mix things up and take one step at a time + +I think a patch that allows us to store more than one instance in should solve the problem +If the tests pass I plan to commit this versionIm still interested in the communitys advice about how thread interrupts should interact with login timeoutsIve also looked into the failures in which appear when you put it in a suite with another test and set +The older version required the fields and would throw an error if not present +This caused more than required headache already +It effectively prevents that a port is being reused too early because that may result into the problem that it receives data from the earlier session in case that the remote peer suddenly manages to send packets or packets were routed oddly and travelled longer than expectedTherefore the decision of how long tcpfintimeout should be chosen depends mostly on the network infrastructure and should be lowered only to the worst case round trip time of a client +You need to have actual consumer for this to work +Is it ok +I think you should just stick with the default +Hardy I see other issues when I search on mappedBy reference an unknown target entity propertyI believe is a duplicate +Ive just committed this +Then we indent for different parts of the command and SQL statement +The tar A lone zero block at X error can be ignored +Ill take a look at it +Its apparentsimplicity lulled me into submitted it too quickly +The attached patch adds counter and size info to the debug log +It is much easier to split then it is to merge +We dont need to maintain such a file in two locations as they can be sourced from the same location during the build +Installer Filename +was referring to +BTW I have a question +The next time when war is redeployed the view remains serialized in the application scope cache which conflicts some how with the new view tree created? please comment +I am just afraid of extending form the old Testcasewe already extend this! Have you looked at latelybq +Minor note seems you edited latest manually it did not apply + +Ive attached a patch which demonstrates the fixes Im proposing I think the actual implementation would probably want to be a little cleaner report more accurate debug messages and reduce code duplication +Fixes the call does some refactoring and gets rid of typosElliott and I considered getting rid of and merge it with but it would just create a mutant +Marking as Documented as Known Issue To change to Fixed Issue in Release Notes for +This is not a bug but mistake in configuration +Is the bug against JET or against OPT +Trying out hudson with the real patch +Thats the nature of the model +Please place any further comments in the other issue +So there is no native only java +However if you use STRICT cleanup mode the whole database should be cleaned for you +this is lingering since march is there some code or shall we move it to +Patch makes sense +Closing as duplicate. +There is no way to shut it off +CORRECTION the change IS in the Response class of the file +For me this modification is not necessary +I do not know if its a regression +The rework in the upcoming should solve this as there is now a separate success flag returned which is based on the actual return code from the compilerJochen Could you please test of on a problematic project +At the moment in the serif examples on both the pages the blue at the top and the orange text is san serif and the rest is serif +So if you sent the xml file out to n translators and got n variations back you would have to merge them all together take into account any new messages that have been added in the meanwhile and make sure they all get merged correctly +Where is the usage doc? +We should fix the test +From database +Update the license header in the Adobe Glyph List source file to match the new one. +That one was a possible test script issue which Ill be fixing +besides some javadoc errors the patch looks ok +Thanks Yip! The new Javadoc is clear and detailed and very usefulThe new tests are also very clear and easy to follow +They will be easier to add once I introduce some more language +Sure this could also be used +It runs as part of the regular test this is sufficient Scott please resolve this case +This problem has been resolved +It wont work with and I agree that its not really worth it for Ive seen more than once +I would rather expose useful settings over JMX than say heres a big wad of text good luck + ilustrating this bug +This patch adds unit tests for a few of the gms classes mainly failure detector and related +Letting my test cluster run for a couple more hours then will commit it +I think option is easiest I also believe there is another option which would require slightly less storage but would require more work +The new created class full name is like +Actually I thought to take a look on this weekend +Making it patch available +Ill have a look at it today +The problem was discovered while writing a test case involving a database on a classpath running under a security manager +I will have this committed in the next hour or so after my final testing completesThanks Tom for this submission TomRevision is +Committed at subversion revision +trunk +the new patch looks good +Just making sure Im reading this issue correctly +Empty layer group +Are they still usable +Unfortunately I dont have Safari to test it in so Ill ask Martijn to test if this hack can be removed +It works when the IPC client is a server with a principal like hdfsREALM + +It looks like Adam applied it to the branch only +How about create instead of usingWhats the benefit of doing thisbq +I moved JIT part in separate JIRA ve committed temporarily to let JIT investigation to continue. +It still appears a bit slower after a lot of IO but not really enough to matter and probably worth the lower cpuThe windows server results are still scary though perhaps we should only use transfer on when detecting non windows +Now or never +Fixed the example as this seems to be the case where as per JPA spec section should indeed throw an exception +defer all issues to. +Changes applied to product documentation +So just replace the one that is defined in conforganizationThere is also no documentation available but this is just a couple of interfaces to cover and it will replace usage of with some custom solution +Seems fine now +Part of patch in all issues not touched in a month to has been in for a while just added a changes after release. +thanks +Thanks Roshan! Committed to trunk and! +on the new patch +Basically an abrubt failure in the parser is not correcting the parsers state for reuse +Works for me +If we want to allow data to be anywhere in the data dir theres no more a way to recognize stuff thats local to it from the absolute m committing a fix that first check if a url taken as absolute exists if not it tries the data dir the workaround spots absolute paths provided they do exist +I dont mean to be pedantic as well +Please have a look at am going to bump this to + While the Eclipse window is not fullscreen Start Central via Help Maximize the Eclipse Restore the Eclipse Window to its original state +BlehCommittedx You missed to add the new dependencies to +Pull request please if this is going to be supported at all anymore that is +Neither one indicates a final solution +Im running the tests again +Exactly +AlexDo tests no longer set logger level in trunk now? +Looks like the testcase starts trackers and then waits for of the mappers to finish +Fixed +Manual test pass +Keep in mind that what comes back from Solr may or may not be the full document that was added originally due to the fields not being requested or the schema not configured to store them +It as simple as without an additional configuration only default hosts are allowed additional hosts can be configured either by listing them in the allowed hosts property or by adding a regex in the allowed regexp property +BrunoYour patch is in trunk rev +Thats great guysCheers +One minor enhancement can you add to and modify to replace the value for target +Charles Ignore my previous comment sorry +As I specified in the attachment comment my patch fixes only Java files +get it to stop writing to HDFS +Closing all resolved tickets from or older +Good candidate for merging to branch +Updated documentation to reflect shouldnt have been canceled resubmitting +Ive take a look at it and the good new is it has nothing to do with the node restartIf you start firs the Reader and then the Writer and abort a transaction the bug will be bug is quite simple +Done +trivial patch attached +May be an issue in the Maven version +Committed as proposedsvn revision tests to daily integration test suitesvn revision +Bulk closing all resolved issue that have not gotten any more feedback or comment in the last month +Actually its exactly how I imagined works +lets pull these up from since they have nothing to do with being sortedinstead with being deduped +Cheers Jonathan +Unfortunately we need to run behind a firewall and need to use a suggested to add the proxy details to the pom +please have a look at with it +Not sure how to fix this oneIE is sending a header for the file input field which is causing us to recognise the field as a file upload field +The violation message should be updated as it does not match our new standards +Patch for branchx +Please crosslink to +However that problem no longer is occurring so it is hard to explain why the fails to return the code signers every so often +Ill give it another stab and I dont get it done this weekend Ill revert it +Does CDI really prevent you from defining private methods having final modifiers or is this a mistake in the check which should actually apply to public methods onlyIll post the same question in the openwebbeans forum +I want to wait for jQM as there are real improvements compared to +Te most current upstream contains which fixes this issue. +tgz artifact right awayThanks! +Thanks for the explanation about Hadoop RPC +Reopening until Chiba has confirmed that the patchis the correct one +Clausmany thanksI updated the ftp component page and the release pageDo you interested in a separate patch with the test improvementsRegards yeah please feel free to submit a new patch with the latest test Christian I have applied the polished patchtrunk . +Fixed +Could that work here as well +Verified +Fixed in all branches +see lines onwards +Fixed on and the other reason builds appear to be faster on with forked javac is that each subprocess gets a new heap in effect doubling the ram available to the build +Joe If you dont have time to maintain beetlejuice any more let me know and well take it over +Maybe we will be able to work around the issue in our Seam please report this issue in eclipse bugzilla too +I the process method and hopefully simplified the calculations +Hi AntonPls find attched the jar list in the attached screen and Regards must put in your path three jars and remove library is in the PATH +but even better if you give a hand +Raintung updating with the storm of small messages is not common for search engines world +is a reasonable suggestionAnother idea is create a separate file JUST for this this crash informationSince its similar in notion to a crash dump file how about a uniquelygenerated name for a standalone dump file of some sort as in Where the YYMMDD is todays date and the N isa number incremented each time we produce a dump onthat dateA nicely packaged standalone dump file produced on a system crashwould provide a very helpful tool for users to interact with thecommunity when trying to diagnose Derby problems +profiles take care of this +Resolved several issues with the connector and model extension creation. +I ran into the same problem indexing into ES from Nutch and had to gut out to useTo add my cents I think that a solution that takes advantage of the less powerful but more configurable would be more broadly useful where adding support for multicast discovery would be an added bonus +Surprisingly the proceeds without any EOF exceptions +Also I use capital letters to name the report files with extention txt +how we behaveThanks +pushing this to feel free to put back if there is been activity on this one for sometime +What if we forced to either Column or on the reply +I have not found a solution yet and I know this will take more time than expected +ok now I know how to attache a file +Alexander thanks for the improvements +If you cant include in the source tree then I would suggest that it is disabled by default and that you have to use yes to enable it +I have posted an enhancement to this patch as +Fixed in commit ffafcddddca and fdfbfdcebdacc in and blockers and critical issues are resolved but not verified +What if maxTasks is zero +How does the appender find a collectorSame as the a list of collectorportHow do we name the chunksOne recordType per appender so we can keep the Chunk metadata +For initial testing reasons this is being moved back to the branch +Hi AsankhaCould you please have a look at this +Ill send you an email on how to update our internal build. +Thanks Adrien new patch looks great +I let Leon comment but I believe we should revert this line indeed +X +It could be you need to start your jvm with some bigger initial heaps +One of the advantages of handling the different cases in the is that the exception may as well occur during the parsing process +corrected patch is attached +Redo as class +WARN action Command output WARN action Provider message The cvs command failed +NET Binding for C messaging + +Also I needed +Hi folks I just wanted to note that were really interested in the netty integration esp +Tested and fixed rpfilter in eacddfdeedbfdc +small diff svn diff format +This behavior seems to trigger a failure in our application though +Attached patch fixes the issue so that the DTSM only runs when the NN is both in active mode and not in safe mode +Can you share the full SLD +Justin and Rolando I know you have been working on a mockup and definiting what the GUI should look like +It should be ok now +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Thatll do it +I can confirm that I have been hitting this issue too and other people at my company also hit itIt would be great to see this patch in an upcoming releaseThanks! +Every thing looks good +Then only subsequently might we add checking on the datanode itself +Strange thing is that I thought there were explicit tests but I could not find them +The problem is some of our supported compilers do not yet support stdstring and stdstringstream +Fixed the authorization validation visitor for proc relational queries +Notice that Host is selected but what you see in the list are the servers defined in Host +Id like to backport this to Will wait to make sure all runs fine on trunk a couple of days +This particular query was optimized for a different SPARQL implementation +r to add exploration of connection editing and authority editing +Exception swallowing is from ve just written a quick test application for this and Im seeing the same behaviour with Seam from as Christian said did this +jar now as well which breaks all of my testing code +With partial state transfer we can choost to ingonore initial state transfer +Hi guyswhat do you think about moving the debug interceptor to the bottom of the stack? Beyond the issue I think it is more reasonable +The HTTP spec doesnt appear to directly address this scenario so its open to interpretation +Test failures are unrelated to this patch +Yes Im on it. +It would theoretically be nice to add to jar as building composite columns in thrift is tedious presently but given that higher level libs probably have this functionality and also the dependency problems that already exist in probably isnt worth it +It allocates a single int value and then references and dereferences that value times +Iterating through that array was causing the slowdownAll tests still pass running time for the resolved array tests are now comparable with the array tests +Thanks Nate for the elaborate explanation on this +Makes use new mapreduce packageChanges example input to be a CSV file which I think is fairly common +Attaching patch incorporating Hemanths Comment +works for me s not working on Fedora bit +Im wondering if we should cache the is now +Is it possible to customize error message when file is not found +The fix is already committed to MDR and will be part of next AS now Im attaching you the fixed mdr jar +But what if I wanted Axis to generate it? How should be the signiture of my java method + +Im not sure if I have open a Jira for that but I will fix it early next week +committed to +I might have forgotten to include a file as I had several patches going on at once +can you confirm if the problem is fixed +I remember a few little weird errors basically it would say there was a featureType but it would be empty and would always cause errors on the requests +but it seems everybody including me of course is needing it +I can turn off MTOM for specific services +I have attached the updated test case +Thanks. +Thank youBest regardsRichard +In my application I use even if it get raised more than one time for each login or autologin +Ill take a shot at this + +This seems to have fixed the problem I was seeing in the trunk +I am suggesting to move noop and noopwithmap to contrib and then register them in +Register dependency on Bug thans for the patch +I could not reproduce this error +The default configuration can now be overridden in more than one way for instance by specifying a prop file in a env var +This was in trunk from yesterday +thanks a lot jacek I have subscribed to dev mailing list +It passes with a locale +if yes then the rest is easy +rewrite and optimize to address the issues ofdivision by zero and overflow +Glad to see that someone had finally time to look into this +Could someone please try this patch? The fix is fine +Had several HRS go down in succession with OOME +The class already has a constructor which takes an Iterable so that already covers Lists and more +Verified by Leviev. +The manifest contains a mandatory import but the and list it as looks like the same issue applies for the Spring Security and packages should have optional resolution as well. +And drop Maven +Istill dont see what the code might be falling over on +Until then I think we can use the last commit sha in git for the revision number in git +If someone can help +Applied the latest patch and everything seems good +and provide integration tests but for the refactoring package spliting etc I need feedback and help from the othersI am currently working to test the integration of the bundles with apache karaf apache aries other frameworks should follow +In beta works fine +The new file needs license header +Uploaded new patch that properly handles requests that have contentTRACE isnt allowed to have payloadPOSTPUT must specify either or see bug +Ill fix this and create a new ticket +bq +I suppose that the solution to this is to create the builders with the objects then to pass them all along +Oh! Are you sure its just that one setting? If so it could be a timing thing +Anybody who has the permission to change it can you change the priority +Rajesh I believe this issue is resolved +I dont think its intended to be included in just proposed as is +This argument isnt about priorities it is about an incremental change versus a whole new set of apisAs someone who both worked on all parts of the framework and as some who maintains lots of it I do not see any clear benefit of adding a whole set of new apis we need to support and maintain since the incremental api change that Owen proposes gives all of the control necessary to accomplish all of the stated goals being proposed by Tom at a significantly smaller cost of maintenance +Fixed in changeset for Release +Version has been releaseClosing all related resolved issues. +Addition its unhandy right now because our user are actually but okay we can change thatThere might be some effects but I dont know exactly how internally the database will compare strings and I dont know what database youre using +Ive just written a pure sample that discovered that it is my fault not Xercess. +this +was added in +Updated patch changes the comments no code changes +But before we move files in from tmp we should know that they were on datanodes at least when we did the moveMostly this came from a user in hbase irc +SVN is installed on the machine it was occurring on +As far as I understand the issue is a pure DRLVM issue +Thank you +we can perhaps use this to avoid the read before write +Ahh yes that does indeed address it +Community wont be patching youll have to bring this up with the productization team if you want it fixed in +As an alternative workaround for jboss cant we just disable that thing if we dont need itOr even that whole +Just create a basic text encoder and have your handler write your messages withwithout linefeeds as needed +Bulk closing stale resolved issues +This results in directory structure trunkspringpythontestspringpythontest so test classes now have the global module name like +I keep the jira opened for now +If you set the enviroment variable TUSCANYHOME the installer looks inside TUSCANYHOMEmodules and TUSCANYHOMElib for the jars +No tests as this is a refactor +YES! please +When there is an error thrown from the service it is properly communicated to client as a SOAP Fault +Integrated to TRUNK and branchThanks for the patch Ramkrishna. +Moving to and rolling back to these as out of date as we will soon be looking at integrating a new version of wise. +Could someone attach an example that reproduces the issue +utilsrun script had the wrong variables substituted +Reopening issue to reassign them from version to or before removing version from Jira +Seems that the problem is tranferring the conversions map into a new comparator compares the lengths of package names loading the new with jdk appears to exhibit different behavior where packages that have the same length will transfer only entry jdk behavior is that both entries are transferred +eventThe container must fire an event for each Java class or interface it discovers in a bean archive before it reads the declared annotations + I have seen this flakiness too +can someone commit this for me? Attached is the patch +Just pumping data into a pipe may be not good enough +Thank you for your replyI have not changed anything in the listening interfaceswhat I do not understand is that on the same server I use it the and it doesnt work but it works with localhostFor mailinig list you can give me the linkBest regards +It shouldnt be too hard to plug in an entity resolver +Arno do you want to work on a patch +Can you provide us with an example of this? Preferably a complete project with and files +then trying to keep it updated when basepath ect are being set +Can you try running the tests on a machine where you get this problem? +Any progress for this issueIt is hard to determine when to call method since is shared +In the proposed patch it is tested whether the hash is actually set and if so AXISSUCCESS is returned +Yes its not completely solved but significantly reduces the race because of the seconds wait in MR +The driver will not be backwards compatible +Manually tested! +Ok heres what Ill do tomorrow +Please give one of the next nightly snapshots a try and let me know whether my changes work for you on Oracle AS! tested this with and it work fine +Here is contents of my bindingsphpusrbincc bundle Wlflatnamespace Wlundefined Wlsuppress o +bq +missed that one out +How will this work exactly? Where will the STS get the public key of the client from? Does it assume the use of TLSColm +I am closing the issue even if there are a lot of applications that still dont use it +I think the albums without owners should be visible to everyone and only the Super Admin can make changes to those albumsCurrently those albums are not visible to the users other than Super Admin +Confirmed the problem is gone withShould I mark it closed? Anyway I think it is +Nice analysis Ram +Also both dtds are now included in the distribution +The clean up function mentioned in item f is not implemented yetTwo properties are introduced and both of them could be configured in two waysa +Exhaustive tests need not be I think +made new jira for the ChineseJapanese shortcuts should fix this for ER if we have time +The following works for mepython c print hey cat and thats pretty much what the script does +updated the patch to fix a problem with the integration unit tests messing up the auth as resolved since the patches were committed in revision +When I do it I try to add some extra value by fixing comments maybe adding some javadoc where it is missing +window losing focus may possibly be related to Groovy executing dialog box so linking it with we go ahead with removing the dialog box and replace it with a status bar message both these issues may get fixed together +Since a JVM can run more than broker the WSN layer just needs to route requests to different brokers based on the root +Here is the patchfilePlease let me know if its the correct formatThanks +bulk defer to a bug at r +This looks good now and seems to work fine +A pretty clear win on the scan workload and no Scanner timeout were seen in the logs +The integration tests runs with the value found in +It is bad ass +Even the number of databases might be configurableHowever all the database resources should be cleaned just after the test finishes +For the moment the bpmn deployer will scan for +I think the move went rather smoothly can we close this one +You might want to do the same +It is still present in UIMA release though it does not cause any problems +I modified the code to retrieve the user theme in the also have added security so only users having a role that will match the personalize action on the portal object will benefit of default the default portal has the permission personalizerecursive as an unchecked permission allowing everybody that is logged in with a profile to personalize the theme. +We do need to be very careful that whatever exceptions are used that they do not reveal the existence of items that the caller should not know about +Maven like to create its on the was looking to the Tools code and meclipse code in SVN to figure out a solution to include this information to WTP and I guess I have an idea what I have to tweak in Tools +The pending ranges calculation needs to include moving nodes +As Tim says the gadget developer may assign whatever ID is desired +Created a patch +Can you let me know what is broken so we can put some checks in place to detect that fail this way in the future and report them rather than crashing the internals? If its internal plumbing causing it please let me know how to reproduce it and Ill look into it +Feel free to close this out. +Verified in website Fixed + for the patch +If you find further issues please create a new jira and link it to this one. +What if we just created the index with cachingNONE instead of adding a special flag +The state is really only needed for the Restore View phase +I reverted the patch from trunk +Thanks for setting this up +bq +This might have been the reason why was returning false prior to the submitted patch +Change the version of s content and the test pom files +the version that is queried is thus the +I am going to check this in +Fixes the issue +Not sure why hadoop QA wouldnt run +I dont think I need to solve that part of the problem for this jira +Finally applied. +Just a note +Thread B calls initializeClass finds class have its value and returns +Juraj However Im still not able to find the root cause of aforementioned error + +comments indicate that this is a duplicate of which has been fixed + +I wrote a silly little integration test in the attached patch would you mind checking it out to see if it fails on your machine +Working on a refreshed proposal and code patch +This is really stopping any meaningful use +The trunk revision is +Style classes +Thanks +but the test is broken thenI am sorry +Actually if we want to expose the dcrack information to clients we need to add it to the local table too +You can run the second apk I sent and see the logcat +JDK? OS? Ive tested this with Suns on Windows and Linux and also with Blackdown on Linux and do not have problems +Can we just use scrub for that? Its a bit but its not a super common scenario +This is only empirical but we try to keep more than TB free just to avoid itWithout any knowledge of whats going on internally it seems that theres certainly correlation between HDFS space left and likelihood of a job on roughly the same input size failing with No space left +It certainly works when I run a CPEEven if putting the paths in the classpath works its the wrong solution +But one of the args is not used in the class +This issue was first reported under albeit with a different use case +What were your results with +The test has been fixedtouchz still creates files without x +What exactly are you trying to use mysql for? In general any kind of persistence can be used you just need to provide your own implementations for interfaces responsible for persistence and inject them via Guice. +committed to and trunk +Regression test. +using put it needs to set if first and clear it later for concurrent operations at same time +It looks like you missed two includes on the CXXFLAGS variable from the original Makefile +Updated in revision . +New Features is now This is a bug fix release which also provides localizations of new messages + to trunk +Merged the changes locally and the build after the merge was successful +Closing +I have attached a patch that fixes conflicts for randomly generated +Given that jars doesnt exist it just fails at that point +Misdetecting the split isnt as easy as retrying because it usually means you get an are you may have already returned bad data +Heres a patch for the issue +This problem is not only major it is a blocker +The Release Notes explain how to build the JDBC functionality with Derby +If the concept is reasonable more performance analysis of the change is required +This provides a max lag mbean for both consumer fetcher manager and replica fetcher manager although I think it is more useful for monitoring consumers +Alan note that this issue causes we cannot run tests with current SNAPSHOT for example we cannot check +It also only ever test compact storage so it would be good to also allow testing compact storage +Pi heres the perl script I use to generate data for testing large and small +Agreed +OK tidied up +Thanks Luke +Thank you +would it be possible to get a test case or such source here so to see that things are working +Thats good to know that having an exploded directory should work +Merged to +It may be or +version and applied the patch but still get a error if I try to POST the networkname parameter +just keep the behavior same as before for simple client +Today table statistics is stored as part of the tableparams +I see tcnative working +Can you please add more description to the jira +There isnt a new Javadoc warning its a false positive +The code path was moved out of the path that updates the data expression +Julia you gotta track down this issue on your end as you are the one which has access to a MS server +I am sorry James but your assertion is incorrect +merge this please to branch demosite +You could require explicitly to ensure the +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Even though the constraint generation class lives in Core it is really a Validator issue +Need to check +sorry for yet another changeHoping to see these changes go into SVN in a near futurePS please let me know if any support is needed from me about this patchRegards per feedback from Khaled I have modified the assertions samples with few improvements keeping in view the recent improvements in psychopathI am attaching +are you sure about the versioncause still uses of of Karaf uses of +No response so well assume its okay to close +Added to provide improvements when there is proper time for it +Also could you please name the patches something like OS X seems to get confused with a extensionThanks +As discussed on the user list I ran into an issue running with +backported the renaming to x as well +No problems thanks for the confirmation +In practice most servers or clients should work well with it +I attached the results of my jenkins run +Ill throw in my in short resist the temptation to handle certificate provisioning +Thanks for the patches! +Fixed patch applied along with fixing the other two methods that had this same issue +Done + Username djencks +I just committed thisThanks for the comments. +intermediate relative url is missing port and and protocol is now preserved +Dontknow if there are some legal issues for this +I did it +We are waiting for them to come up so that we can get back on track with release +Resolved for a while and just confirmed Couch returns a +Already appliedThanks Suran +defer to +Totally agree with you forgot about that JIRAHere is a patch that is very much along the lines of but forI wonder if we need anymore +As far as I am concerned this can be closed +Hello FreddyThanks that would be greatI have attached the first pdf report +Marking as closed. +Applied on to compile the plugin against the target +This has already been fixed + Issue +I think KILLED would be more appropriate then COMPLETED +Resolved but still have to merge changes into Trunk. +We want to remove before opening by reading compaction log system CF +Normally if you are on Windows please use the +I agree this can be closed +Heres something I hacked together tonight which maps the memstore maps hierarchical +this will move the contribcollation ICU based components to contribiCU and this is where I want to bring the unicode support +Thanks RavindraYour patch is in trunk Thanks Vikas for Wiki Upgrading from earlier revisions page update +bq +For that reason this issue will be postponed until after that unification +Also Ive attached an unittest +so how are they showing up +no other change +Well it apparently does if you use JAXB annotations and not a castor XML mapping file +Instead the test is fooled by always falling back on classpath loading which finds the resourceThis is exactly what I had in mind when I mentioned the responsibilities of RF and RSLThis test is trying to test Path but is actually testing RSLs classpath loading +Im not sure there is a for changing the CSV Strategy once it has been createdI would make all the fields final and only allow them to be set via the constructor +We should finish discussions and if changes come out of it we could make further changes to the protocol +do you have a simple test case project + +Is this the root cause of the problems being experienced at Digg and by Lu Ming on the ML +Fixed byMaster doesnt edit +Patch for new hdfs trunk +But I can tell you all the attributes we changed +I didnt want to affect the obviously high performance of the +We could perhaps add it to a standard lib namespace or somesuch +The freezeunfreeze task used to work on rake by loading a specific version of Buildr first thing in the rakefile +Fixed a bunch of failures due to bugs in the compareTo hashCode impls for the specialized tuples +Ive attached a sample Maven project that follows your guidelines with println commands in init cleanup and testSomething +thanks for reporting thisCan you point us to the specifications for these directories please +this seems good to do though are there any other base test classes that need the Ignore tooIm pretty sure making these classes abstract would work for too I just thought that Ignore was a less invasive change +refactored code let FWD hints be bytes per address as jbellis requested avoid fwd hints when already in local DC remove fwd hint header when we send to final destination to avoid infinite loopsIm not sure how I can unit test this yet there are very few tests overallI will fully test on a live cluster soon +Looks like nice addition and would make it easy to try out different poolingsWhats a little odd is your adding of new functionality but wo unit tests +Excellent +Is this patch going to be included in the next Janino Release +Currently nothing is copied +If this is in trunk we should do a trunk patch at the same time though +Set Fix version to +Thanks +These patchesconflict with and and can be committed only after these patchesare committedI request for these patches to be considered for reviews and comments +Hi AhsanIf your run is only running tests then I think you are not running your newversion of as the patched version has many new tests +I think it should work out quite well +For example I recently upgraded a dependency in a complex application from cayenne to cayenne +you mean min right +Should be fixed dont know the version +I just ran these tests on my machine both with and without this patch applied and they all passed just fineWould someone mind reviewing this ASAP? MR Jenkins runs will be broken until this is committed +Please reopen if it doesnt work with your scenario. +This adds the test to the jdbc test suite +Or at least move the definitions into a Terms section so its clearly separated +I will post more comments after debugging +If the query is malformed then a BAD REQUEST is returned +Thanks for spotting It seems Codehaus upgraded to a new version of Bamboo with different link structure +Even if default gets changed +Most times the client just verifies the server as in https because its impossible for public websites to know each client that connects +Does anybody in the grails team care to comment on this. +I can confirm it works +Yes methods take conf and protocol class as arguments +Want to work on a patch +attached the nd file please remove the structure should be the ZIP file minus the CVS have been added to thirdparty cvs directory. +Removed the accidental change from the previous patch +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +Lance I tried just this but I still get compile errors +Committing. +Thanks Wei! +As long as they are compiled with but not usedas default +Thanks! Patch applied. +It is probably quite rare almost everything in a jar is in its own namespace of directories +bq +Verified +I dont think that we should configure me any different for projectsMax WDYT +Added a +updated CHANGES and closed +The plan is to split out the Seam stuff in its own plugin and then this can be done +Could we use the Avro file format to store the fsimage? Weve designed configurable compression into the format and tools will automatically be available for inspection of the file +This patch is a first step toward writing the new test +Hi Sherry does this issue not also affect trunk? If so would you mind preparing a trunk patch as well +Ive fixed the nit and committed to branch. +afaik this is not easy to tom +Patch LGTM but doesnt apply cleanly +I just committed this +Rather than use state listeners it is easier to simply lock on an object and wait until a notifyAll is called when the state is changed +Perhaps thats the case here? Do other libraries seem to load in reasonable time +whyDid you had a look on that mentioned rpm spec file? super outdated could be removed from my perspectiveWhy do we need debian as a top folder here? many years ago people decided to do so building packages is straight forward no folder moving before building +The only place in any history I can find with an isSplitable method but without taking splittable codecs into account is in +bulk close of all resolved issues as part of closing items. + but it breaks the current testsmake C testnodejs server make C testnodejs clientsorry i didnt have much time to find out the reasonCheers any updates on this +Move the issue to AS CR +The patch looks fine to me in the sense that it solves my issue Im not sure though if this behavior isnt even less intuitive than before +But our users expect this behavior and have pig jobs that depend on itIf the current patch works I am inclined to leave it until I get time to do a better +Now I think Kabir is outThose dependencies look correct and I assume that Kabir got something working but I wont be working with it for at least several days so there is no hurry +I commit shortly +The same patch for branch +This is not a trivial task to be tackled given the current state of affairs +This should be resolved now as the is not longer added by default but instead all EE APIs are added individually +bulk close of resolved issues as part of JIRA housekeeping for release. +I think it should be removed but I realized I had forgotten to do it +skh targeting to preferred approach to this defect will be to utilize an eclipse based plugin utilizing the DQP +if the property is declared as being of a type having the annotation then all is okHowever IF the generics are not defined on the list or the bean property is defined as being of a super type of that having the annotation then no class is injectedI was expecting that it would be independant of the declared type and only depentant on the actual type being supplied. +Please attach the schemas used plus enough other information to be able to reproduce this problem +Attaching +Go compiler should get the patch as well +The approach I took was extremely consistent given what I had to work with +Please +Through this JIRA I got a chance to finally look at a bit of how joint compilation works +I obtain the same error by creating new project group and assign a template got that Continuum can eliminate this type of projectgroup but when they have problems with the generation of projects and Ant Console +Do we need to create a new issue +We could check if the second part is HOST in the key and try binding to local address +Verified Fixed +Also is the protocol we expect to communicate with RM on also it is more than just the protocol for just getting resources from the Scheduler since it has registration eventually well need to add cluster information etcThose were my reservations about the names I proposed too +Copyright laws differ internationally +How do you deploy and run it? Do you use some Jee server standalone from an unit test tomcat etc? And can you try creating an unit test to see if it then worksI wonder if there is something that proxies that bean? After all Camel is just asking the method which annotations it has and then it uses that information +Full logs of with the timeout exception +I think Emmanuel may have +Im not sure of a way in EMF to preserve the copyright on modified generated code +My strategy to solving this was to install the latest build create a new clean project and add to the new project bit by bit until I could reproduce the issue +Sam are you working on a patch for trunk HDFS here or just the branch append feature +Fixed at close for the release +Problem is in our Struts actions we catch exceptions and not runtime errors +A name making clear that the page number is eventually calculated + +There should be no perf hit like walking the hash now. + you attach a patch instead of a new version of the file +Well I started writing a new to handle this logic but Ive hit a road block since Im now discovering that Plexus isnt recognizing my class as an additional component when I add my project as a plugin dependency of or in my pom +Hi can a filename contain characters +Do we need a list of sequencers to display in JONJopr? +test looks good and produces the problem so I will use that +It confirms what we thought and that is helpful. +Weve attempted to handle it as much as we can and hopefully AS will address this better which were supporting in. +bq +The major issue is that wickethead is rendered by a different code path than other header contributions +Heres a working version +I think the problem was with maven version +If there are improvements please open up new issues and be specific. +If not Ill attach it when I have anKeith +We now have LSMR and other techniques that deal with this well +Will keep you posted +we can create a job with poll SCM enabled with value hourly +I have not tried using it +So for me it sounds like a feature request +I also tried to clear the browser history it do sign out the current user +Yes the IP isThanksEric +Not a blocker. +Its the default file name that the uses for its own +Martin please the patch with Apache license reattached +but thats just a workaround and shouldnt be necessaryAnd it turns out even a private accessor method is sufficient +Discussed this with Tom and he said I could take this one +I did think in order to get things working as currently there seems to be a mismatch between the assembly plugin doc examples and the implementation +on adding as a configuration optionEdit Meaning on UUID support +something is really broke in this userrole thing is a red herring + +committed to trunk revision +to both and trunk +Illcheckout a clean one and apply the patch this weekend and then check it into theCVS +Agreed I see now that would be the best option +public static instead of static public class make some of the instance variables of for checking on this +With the addition of and a few other pieces it will handle the other pieces as well it should handle the last of large cases thus combining usability with scale +Included +Dimscould you tell if you could reproduce the problem at least? I mean that you need to stopstart a webapp with Axis added before xsdstring instead of soapencstring gets generated to the WSDL? Its really not a showstopper but its not nice for me I have to force users of my web service after the deploy step to restart my webappThanx! Merten +bulk close of all resolved issues. +The flip side of this though is that if a ZK issue were identified it could be fixed and redeployed wo requiring a of Hadoop +no more than overages in a minute +Thanks +So patch against trunk will help and would allow to cleanup outdated attachments +I didnt have the set so Im trying that and hopefully thats the workaround Brett was referring to +Missed the fix while creating the patchThis one passes just fine on my machine +However I think that processing positionIncrements with is necessary +Okay I guess we can fix this as well so if the target is part of the same source as yourself then you should not be able to I got this partly fixed +NoelThanks for all your work to fix this +For it should suffice to bump up the default hardcoded value +bq +A quickstart demonstrating this issue +We accomplish this by doing a walk over the childsibling chains but missed synchronizing attribute nodes since they are not children of elements +Additional validation and messages have been added +Verified by Mark. +Yeah saw this in todays build too +evenOk will change and update the patch +Revision I provided a set of default comparators file name size and last modification dateThe users can use these comparatorsI updated the wiki page with informations about comparator usage. +add clean before does not fix the entire problem +had to be merged manually +Thats what current code already does +We may need to replace this test data if details are not forthcoming. +I dont have a use case for it currently but I think it would be nice to allow any even if they do not implement the full API +Thanks Vinod! +Looks good to me +For the script running for we will get something like shellcommand echopwd shellargs YarnHadoopMapreduce shellargsnum +In this version accessToken is also unused +Pushing to future Barry will evaluate whether to move to Teiid. +Compiler issue. +It does a few extra CQL requests to get to the next pageCF rows +This is the version I committed +you mentioned the apt plugin +My proposed patched +form a first look the patch makes sense to me +Patch adds three things Support for file globbing Support for summary mode that just counts Chunks instead of dumping them +It whole the document each time the document partitioning is changed +Patch that moves and +Prompted by I have resynchronized the Gant script with the Groovy scripts and ensured the patch attached to is met even though it was not appliedThe changes I have made are UNTESTED as I have no Windows system to try the revised scripts +This solves my case thank you +Then we could have an option to show implicit variables as compilation errors so we are forced to write better code +can you please add to release notes the that were added as well as their syntax and usage examples +In this case first checkpoint will try to use all the settings from the default configuration instead of fixing the failed onesTaking this into account I suggest we go back to the original propose namely remembering failed replicas and trying to restore them on every checkpoint +logic test result as well as design improvement based on that +We do have jbossws common framework and spi besides the main jbossws build which also produces dont have any integration module +So we really need to get a relocation pom generated to point people to the new groupId +Revision r From within trunkclownfish move libBoilerplater to libClownfish and lib to lib +Roberts patch looks good +The only thing Im not sure about is whether the +I seem to have fixed it +I will get the from the the ticket as you suggested but it would be good if you can attach to this issue all the relevant classes for reproducing the problem +Just needs tests now +Here is the trivial patch +fixes this issue +See revision +I was favoring a signaling mechanism but this might work just as well +Thanks Ashutosh! +can somebody please review this one +This patch applies to the trunk +Ill add it to svn and exclude it from the itest as it is not working yet +The issue seems to apply to Ruby in general +Im OK with a solution that uses the repository thread pool but uses the current thread if the pool is busy +The current implementation of the classifier algorithms in Mahout may require different Java or Hadoop file formats but from the command line users point of view the requirement of these algorithms are the same records with predictor and target variables and predictor variables may be of type numeric word or text the target variable may be binary or category with more than values I think there are two appoaches +This allows multiple runtimes to share the jitted representation so long as the code is logically identical. +Im afraid we cant do anything about this for the moment +My errorresolved not closed yet until or released +Patches applied. +I agree that its a matter of perspective +Yes thats one way but its not secure +Phew! This took more effort than I anticipated +Did we ever remove the goofy code? Do we care? Is this resolvable +try with contacts just updated and tried with contacts no problems +I believe it is required since all style transactions would fail without this fix +In open source there are two choices Gora +Hi againI discovered that setting LDLIBRARYPATH will get the clients to run +Patch looks good +I will fix this I am not aware of a bugid that covers this +mgmt server upgrade and kvm agent upgrade +handling file corruption is less critical but it is something Im starting to worry about +Jesse Sounds good +Changes are fine with me. +Alessio do you have a test case for this issue +OK +it works. +However I want to believe that they have same implementation so if one works then another also should work +I think this may be specific to the CLI but the Cordova CLI may be affected as well +TedSo by doing so we will get a considerable delay which will ensure that is completed +this works now with psi +You need them since data gets set from the methods in +Yes Slava everything is OK +Gary I understand the value and we should definitely recognize the matching versions im just saying that if you do it the way you are suggesting it will currently result in the server type being hard to maintain for us within the current wtp framework +Are and being archived as per the usual setupThanks again saw the note re the in the infra list I assume is doing the same thing so please ignore item in the comment immediately above +I suggest you to have a look at echo sample and get some idea +Attached the patch file with the small change +Pl +This issue was related to a jar unsigning issue and has been resolved +comitted to master and with build +Should be read from the Maven Manifest included in all where we should log? I think if we get version of an implementation the version couldnt log as soon as gatein is what Im not sure abouteXo Kernel does all the bootstrapping but its not the place to add that codeIt could be added in the with an additional we wont fix on Gatein +my eyeballs got crossed a little bit when I stared at this Is the net effect of this JIRApatch just to explicitly define the xmlns? or does the current usage info provided in the console not conform to the deployment schema +Thanks I agree with you about the correct behavior +Thanks! I will take a look at +Axis is not RPC middleware its a SOAP processor +Actually it looks like the main work is to cut the center out of and make it separate but included +Attaching patch fixing the issue +Assaid in previous comment this is done with another approach on + a foolish mistakeAlso limited to the classes sepcified in the ticket +rebasedNOTE system tests are currently failing +The patch is an elegant solution to the issue +This pretty much looks like a break in the server side API +Back ported to r +If not the process is stopped in a dedicated server for this issue so that I can attach gdb and inspect any variable in any thread any time +I dont know what is the side effects +Attaching patch +Marking the issues as fixed for +This has the server side and the Java client implementation +you started working on the at all? It cant use outside doc ids like that +It seems to me that this should be used when efficiency is needed and the user can take an appropriate explicit action to enable it +It suggests using tls model on emt which gives up to boost on jbb as it avoids tlsaddrget calls +merged to +Restarting the cluster node with the problem appears to clear the problem with no bad effects to the JCRThere are no deadlocks reported on any threads on the locked nodes and other https requests that do not use the JCR are operating correctly +Thanks for this +This can be if you can show this issue repeatedly occurs on another also got the same error +Thanks Stack for review +And now making sure I check the Grant License box +We make another jira for that. +This is a duplicate of fixed the issue +Ah +Static fixtures couldnt be handled with a rule so Ive decided to rewrite Runner instead of subclassing it +It looks like you are using it to try and read to the end of the but I am not sure that it is guaranteed to work that way +Whats the status of this issue? It doesnt look like its been touched in months +still exist and will try to fix that. +Could you please attach a full debug log of your Maven build +This version should be correct. +p applied +After digging it should be in the boot classloader. +That did not workIt would be interesting to see the details on this to either allow others not to waste their time trying a similar solution or to allow others to get the solution workingThe benefit of an open source community is to throw code out there and ask opinions on why it doesnt work usually the whole community benefits through the exposure +Yes We have to wait till we implement multiple shared edits support +provide information if the or M fixes the issues +Generating the sources in targetxmlbeans would IMO be the best solution +Is there a use case for the Sling API where an outside party would be interested in getting the scripts value as a Reader? It seems you are suggesting not but I want to make sure +Resolved by invoking method and url will expire after one hour +The bug is always present into +Rajesh could you have some EAP folks look into it? If not I will handle it +Its not really needed for this to work correctly +Will link PR for maintenance if possible +This works for me now +The original scala issue is still open +Currently this is only on and merged to +Thank you for thisSometimes problems can also be caused by old Jackson jars in classpath this could cause symptoms like you describe. +In terms of when I can get the patch done Ive scheduled time next week +Setting initial value for next back L instead of L +to JBTIS from someone explain me how can I verify this bug please +Your email message has been forwarded to that address so I will get it and deal with it +I can try removing it from my env script though +Ruth patch applied at r thanks a lot please verify +This was an interesting bug +Could you say more about how it would be done differently between and +Tika dependency is still present in the components below and should be removed as these are dependent on which is the only reason they had tika listed bogusly in the first placeAs for that should remain the same for now as it depends on +Also I would give it a name to indicate that its a seq file of Records not of raw Chunks +Hi ChrisSorry aminor correction to my previous comments +Kickstart is no longer in the scope of Activiti +This is provided so that Oozie users get the requisite hadoop jars to start off with +This should be fixed now +Kevin will provide input and then hand over to ECS +bq +Fixed with help from federate to intercept Designers UUID SQL version during resolving. +A fix was committed +may be the right place to add default locale preferences for the user +Nice work +Instead of cached is it possible to instead keep a separate array of the names in byte form? If were doing all access positionally that would allow us to skip the hash map lookups +This changes the existing approach of keeping a map of all files +I am digging through those logs a bit more to see what else I can come up with +Its obvious to me that several generic adapters are disappearing and most are to be downloaded separately now +Moreover this breaks the existing user interface +Ok probably notStill Id think this belongs into set somehow +any update on release progress? I would be happy to do some testing on the snapshot when a new build is done +Perhaps we should consider revisiting things to make this easierFor instance a simpler algorithm might look like such +Defer all open issues to +These aggregations are going to be minutely +Closing issue as it has been released as part of release +just some javadocs +performance is still good +Therefore it cannot be the default +It may seem obvious but what are the steps to reproduce this +So I introduced annotationDepth to hold its patch from Robert Scholte +If this last thing is something you need then a need mode would make sense +I think +Note the source of this error was so the fix is specific to +The fork configuration works with maven but not with maven +Woops thanks Billy Ill merge w my patchSorry test failures are expected until I get payloads working +The old format still works +pig complained about a missing class from guava which wasnt included any where in the tarball +groovy c groovy c UTF groovy c groovy c utf groovy c groovy c uTf All the above commands work fineif the file unicodeNamegroovy contains no Windows BOM +verified on +You can overrun a cluster by running a PE upload into a small cluster +Henning I noticed you do not have a shell on ajaxIf youd like to look into this a bit more closely I cancreate an account for you there +it should be possible to just install a stock solr with no modifications on a node fire it up without having any files and be done with it +Allan what are the chances that this improvement makes it into MM +Grrrr why can I not set the component when creating new issues +In the endpoint class then do the lookup in the doStart method and keep the gettersetter as plain eg the set method +Attached Patch +Jackson is moving really fast with releases +The workaround for this issue is to use the Sun JDK insteadIs this workaround still currentThanksRebecca +Santiago has committed a patch which fixes the testcase I attachedIll leave this open until we get around to adding the testcase to the conformance bucket +Ill probably create one map abstraction in Spring +Designer was used timestamp for each Sink when it had been created +It would go to the check of the required flag +reopening since with an XA datasource the exception is properly logged but not so with a non XA data source +Thats all +hasnt hung for a while. +Architecture Diagram for the is the complete standalone solution +Cant this redundancy be avoided +Need to add the side menu +Guess we can slip it in for round two now +Sorry this went to the wrong project +for branch +on commit +If Im understanding you correctly I think youre saying thatthis feature could be implemented efficiently but that doing sois more work than is in scope for the next that correctIf so Im very glad the feature request isnt totally dead +When we output applicationReport we print out diagnosis be good to make the behavior consistent +verified in on Maven Configurator ve opened separate issue for it +thanks for reporting + uses which always call with must set to false +For instance using proftpd you can setup if the account is chroot or not. +Attached patch has additional unit tests could not create review board entry because this patch is on top of patch +DavidIm not gettting the cache errors +Examples of displayModefancy and displayMode support in via the new module +Resolved as Wont Fix +On one call to retrieve the map it will have a thread context class loader set for one application +I too dont know how to get it too run standalone as the Antrun task doesnt let you specify the id to run +Ha! Gotcha +Whoops replace Term Level on the stack trace +In the UI this could be implemented as a or something to make it easier to add this purpose +Im still fixing the failedskipped tests as of the moment +Easy fix +Manik did you try increasing the OOB thread pool +Do we want to specify in terms of or absolute key count +See discussion in comments +On unregisterclassloader set the static aopclassAdvisoraop field to null +Sorry. +Thanks. +We cannot use any GPL code +the URL is redirected and then changed by the as you ll take a look at your suggestion over the next day or two as I have to make some progress on some other pieces of the app +Closing this oneOleg +I just committed this +Ive excluded the asynchronous fork test from the HSQLDB test suite a limitation which has been discussed investigated and made known way too many times +Thanks Cheolsoo +removing files which are corrupted confirm it works now for a long time! +this is going to be very popular! catch me on freenode whirr or jclouds when you are ready would be good to have +If youd like to help it would be a welcome contribution +But if we delete from the page all works again correctly + +They tell me that stuff that I wrote works the way I intended it +I applied the patch to the trunk +Ive committed it against. +Sorry I didnt get time to fix this been swamped with so many projects at the momentThats great you got the thread local NOW included as well +Just wondering if it would not be bettercleaner to create a CLI command for this instead of the scriptI know that assumes an initial connection from localhost to be able to add the first user +Attachment has been added with description Text as shapes workaround in Adobe SVG has been added with description Text problems shown in Adobe SVG all open bugs to the development listSorry for the mass mailing. +bq +Itd be a class that wed hide JVM particular references +It returns from a loop to iterate on excludes +If that field is dirty we need to ensure that the referenced side of the relationship is kicked out of the cache. +PrashanthNot sure if you still have those questions in the deleted comment +Yeah it works fine for me locally as well +can be done separately if needed +Yeah Id say just hook it up for actually dont really like the and OGC stuff since it assumes a weird structure which is not in line with our assumptions +Though recoverablezookeeper does retry after errors is special and normal retry wonrecover the error no matter how many retries +Otherwise we have a dependency that is never used +I fixed up the other in this commit modulo the mr code thanks for catching that! +Patch applied in SVN revision by SVN revision had accidentally left a file out of the previous patchIve attached a patch for that missing file which should take care of the unit test regressions. +Working on a fix. +After thinking about it further I think we need to stay in the logical realm +Thanks a lot Jarek! +The updated Unit Test with test can show this fix +removes memory meter and does changes to fix the issue +Attaching a patch for the threaded version of plugin hook point dispatch +BTW this example requires Tomcat the Servlet mapping to doesnt appear to work +Please check that it works for you +Dunno no one has resolved it yet +Here is a first triage pass through these classes +Attempting to use databases from earlier or community versions of the components may not work and is not the use of existing databases is needed then contact Red Hat Support for advice. +sure why the filed m isnt used for this. +Property are processed by JPOX +My fault for not writing it down close this +After we improve things upstream Ill make this a nicer demo. +It would still make sense to have a test so that we dont have any regressions +Tiny additional patch update fixed imports on parent class to eliminate warnings in eclipse +Linking to +And I modified the jute and version gen stuff so that you can specify an output directoryAlso several files got deleted like all the jar files in the lib dir and the checked in javacc generated files since maven is now running javacc to generate them from the +Thanks +Im planning on committing this later today +Using a for the prepared statements was overkill +Now you can do PUT on a table name to create the table GET to describe it etc +So committing only to branch whenever that is cut is fine +fixed now +Likely the client API will change dramatically in +Good findId vote for reverting that particular config var rename +It can enhance the replication We can reopen if necessary +Stopwords and stoptags for Solr are now tracked in and a patch is available + +Hi DanielIve tried running your code against the latest code and it works as LeeI am using am happy to hear that the issue has been fixed in +Can you create a one with the new trunk +We are addressing this in Scout. +Resolving as WONTFIX +Hithx for providing this improvementThe interface changes that comes with will break old implementations +Subtasks done ok for closing +Note that identifier paths can also occur in property values in which case we need to be able to roundtrip them through OAK +testAutomaticallyfield string fieldName nameentity class +tested +I see that there was a commit to two days ago so I was wondering if someone on the Hibernate team would mind taking another look at this patch +Patch that resolves the issue by only setting the default if it was not set already +Click on the name to open the popup +sighyou and i are officially done on this line of closing stale resolved issues +Fixes the Error in semantic analysis +bulk defer to close of wont fix etc resolved issues. +Thanks for the patch! +Sorry +I recall implementing some dirty workaround but looks like it will need more research +I dont recall them being in different orders just that one returns the full stack and the other doesnt +Thanks very much for the contributionThe fixing is available in all builds after r +Got itI could try to prepare a custom Franken Cassandra from both versions and give it to our customer to try in a test environmentJust let me know what customized jar you would want me to replace in a stock installation +last is pointing to From the logbq +A few questions Why arent the credentials used as intended to store each secret individually? Joining the keys and values with a comma may be subject to issues if either contains a comma +please note the current build has multiple poms configured with servlet for the patch +Heh +Patch attached +Good to hear + looks good to revision +All is well. +Are the issues within session scope manager having to do with attached hibernate objects and lazy loading locking of objects? Im not saying the problem shouldnt be solved somewhere Im just wondering if Tapestry really needs adjusting in this instance +for applying + already has set exposed what would be the switch context I didnt understand what functionality you need besides the property functionality you needed on your factory was because it would be easier to implement it that way +I use the distribution method and I dont see any reason youd want to use the other +Trash in my hdfs home directory for disabling trash in my account +NathanYes DRLVM extracts attributes from manifests of jars files specified on its boot class path and appends them to boot class path +cheers +However if what you wantneed is a VM that you run this on yourselves then that is a different request which we might well be able to help with +Could this be a problem of fault tolerance? Like one groom isnt responsive anymore and the others are just waiting for this groom to reach the barrierWhats the error log say? it seems similar with problem +This patch coerces the solution to produce and that compile andor run on or systems interchangeably +Reworked and simplified the design of the patch method is not used anymore by which is now just proxying every aspect dependency callbacksI think the patch is now finalized +Although this functionality is new there is little chance that this transaction will break other existing thingsI proposed to commit it to both branch and trunk +I added a Java build target to javaengineorgapachederbyimplsql because requires varargs in order to compile +Ted When you say Patch from Matteo what does that mean? He sent it to you to upload? Or you are doing something else? Is Matteo indisposed unable to contribute his own patch to Apache for some reason? Just wondering +We shouldnt count pointless moves in the selection count so they really should be filtered through a. +winutils patch for launching task as a domain user + version of Maven made no difference I am still getting the same you please do the following your existent maven repo maven mvn versionAnd report back if you still see the issue +Wade appreciate the help and intent with this but this should be handled by the PMC so the client library distributed is correctly signed and proper ASF guidelines are followed +However Sannes code that checks that a key is present until just before those seconds and then waits for a much bigger time for the data to disappear works for the moment + +Example Eclipse project demonstrating the you expect your array to be filled with the data by the jbossxb? Or would you expect it to beI dont think we are going to just fill the already initialized array +If it works for youplease close this +Good news +I wish youd document stuff in the source code so I know when omissions are deliberate and not erroneous +Many thanks for the fix! +Can this issue be also assigned to meThanks +Updated description to reflect my current thoughts. +The patch does the following +No we need a different solution for this +Target for branch also +With versions of the protocol there is either a or chance of collision +Please review itIf there is any issue please let me know if you see any issue or more fixes will be added by +OK LeonI will put a comment for other langages as well +Hi OlivierWhat this issue means and how it is related to duplications +Thanks Aaron! Ive committed this to and trunk. +It seems that this breaks see also. +issue based on Gwens comments +Apache infrastructure prefers this +Thanks for the info Vinod +please reopen if you can reproduce. +Im not entirely against a manual trigger but this suggests that our current tuning knobs are insufficient +Quick fix for bad +Good catch John +the jar file build from the CVS HEAD today. +The search term is retained in the search box +the more commits go on the more difficult to merge +What if I want to output this stuff as xml or as html Id have to redo this patch fundamentally rather than alter a single method +Id rather stick with what is established in Java than make something up on our own +No need for IT +Ok also backported to. +Raj could you check if this patch solves the bug +No Freeman added the attribute before the creation of +Pushed to with JBDS. +We can discuss changing the implementation but I tend to agree with Dain that it is better for passivate to prevent idle connections from holding open database transactions +I assume its not a big issue otherwise Id expect bootstrap to come with some other solutionThanks for the patch. +I was thinking the same thing as Paul are right +patch applied +I removed the Hide Fields link and added some inline styling +Work in progress patch for this new error handler +conf should not be touched +Could youplease verify +I dont think gives reply count so probably want to reuse the loging for the branch The timed out after receiving all replies message could probably use a bit more explanation for people unfamiliar w the digest logicI also note that the patch is reversed which confused me a bit but looking at the branch cleared it up +This is happening from Pig onwards if I use Pig I can see the files savedI believe this is because the schema object is null in but for there is an empty schema created +Thanks a lot Robert for taking look at the patchI have updated the patch with the above comments fixed +Sorry I see my previous comment was ambiguousWhat I meant was that the code could be AMENDED to keep track of the current userpassword combination and use that to invalidate connections which were set up using a different passwordAs far as I can tell that would prevent the old password from being used again when getting a new connection from the poolIt would not stop an active connection from being used but IMO thats really up to the database engine whether or not a password change has any effect on an open connection +Tried ant and ant with the codes posted in +Fileupload doesnt but thats simply because I am aware of the problem +r +This is an old issue and it has not been affected by +The failures appear intermittently but frequently enough to be a concern +Im not sure that is the case +I shall post the results soon +Correct +Thanks for the contribution Joe! +Given the above do you think that this patch needs to touch at all? If so what is ur proposalNo proposal +However the TOC remains that of the manual containing the link not the manual linked to +As requested logs from the test run +works for me youll need to attach a reproducible example +Option is not a possibility since it increases the possibility of encountering the errors reported in like modules failing to load +The check on the broker to not check live leaders will allow the broker to circumvent itBut the good news is that after we make the fix you proposed for brokers handling of does not need to contain aliveLeaders +We have to explain entries we are using in client side +Supun Why do we need to separate out the two conditions to two blocks in implementing the double lock? I mean the recreate condition and the cachedTemplates null conditionAnyway the issue that you are addressing seems to be an actual issue and is a good catch +I just added a comment to the page that says Outputs the content of the Value Stack +maybe? apply from root of geronimo is the interface ? Couldnt it be Reporter with no I at the beggining? Is it required by M +Time to go back into the code to see what Im missing +Some automatic formatting change done by my editor caused the previous patch virtually unreadable +It looks like theres two remaining issues that cause about compilation errors +This is just wrong +Its about removing the special handling of the String typeproviding an setting also will not work as it will break component libs that expect empty strings to be converted to nullsDo you have an example case where this would be problematicit is not a job of a converter to perform valuenull conversions +Waiting to discuss the fix with patch with unit tests +I sometimes see quite a few commitlog segments +Committed to trunk and thanks for reviewing Aaron. +the ones who can view the security vulnerabilities? Currently all hive security issue seem to be in public domain on JIRA +Add query hints where need +Attached is a patch where is replaced with and is replaced with mean is replaced with i was running the build without specifying the version +Okay the icon needed by the cmlenz patch wants the image with the arrows not the one with the plusminus symbols +No obvious solution so marking as rather than +I think at least initially is preferable +fixed xml file +Small snag on this +Closing released issues. +The results produces is an empty list of vertices when it should be the outer circle with a hole where the inner circle is +now ignores extension elements that are null +So anything that will work with vanilla ZK will work with Curator +care to provide links to the reports that is being discussed herebut on the surface im agreeing with mickael it does not hurt having code coverage included for tests. +Thanks for committing this! +Im attaching a patch for the leak in +Ive updated the title to reflect my opinionBut basically we need to fix all the cqlsh help for release anyway so I dont see a point in creating many different tickets +Thanks that the patch is fine samples were not building before applying then samples need to be solution is to break up Acegi into separate modules +This patch includes the hostnameport on a failure and modifies to check for both the address and the nested cause in the exception text +Thanks +This said we may add the repo there after when that repo has stabilized +Hi BabakNo reason other than time right now +I see +We need not null to be a separate token to be able to add a default value after the column is added but before setting not nullPlease report back if this is ok Scott +Couldnt agree more Todd +Thank you very much. +Thx +Can be resolved after that +JPAs join syntax +Lets table this for later +confSystemLibraryFrameworksVersionsHomelibUsersjoaDocumentsworkspacescompanyworkspacebin +attached the patch with the approach we ended up taking +Right +Until I didnt look in source I didnt realize what was wrong +I havent seen these and the QE testing of EAP isnt finding them either so Im going to assume a change in Beta took care of this +And kill the term slice if it does not give extra value over shard +Shouldnt we leave the stream open so that a response can be written back to the client at some point +Also why are hibernate properties on an infinispan config +This revised spec is excellent John from the perspective of documentation +Rev +Closed +Lets push it off and wait until they fix it in WTP +I prefer option as well unless anyone has an objection +mUsers who really like to compile those snapshots from source just need to enable this profile with mvn clean install +This is a duplicate of +I am interested in using DateTime category API for a Groovy DSL I am writing +this isVersion Build id Build date for label was done by dynamically trying to detect that the server containsIf it does we add the flag at server creation time +and the server does not understand GOhow about disable the block mode +It temporarily includes the codes in + for this its really disturbing seeing the same exact user +Then we can get feedback on that and get verified what those examples canwill targetbtw +r committed to release branch +It looks promising +To me downloading to view the source and then looking through the docs Id be confused when the links didnt work but that doesnt mean anything has to change +Added information about EL support to user guide sub process section revision +Okay can you capture as text the output of the following commands +Do folks think this should force a release +Also I think an cannot have a monopoly on properties of a specific type so it would have to reanalize the data to find out which properties it would add and remove only those +That would be very costly +Updating the list in the client is idea +Yecht added this functionality in last fall +In an RVM install is present but is not +Simple whitespace management works screenshots of the same code one with disabled whitespace management one with simple and one with smart +Shortcuts XDG Menu applicable +Please see this ASAP +In the GUI has changed significantly reopen if the bug is still reproducable against +Thanks Alexander for reportingtrunk If possible then please test the trunk code in your all tickets +Ive changed the summary line +Have removed this misleading note does this fix the issue +Axis client sample request that works compare against +Ahhh sorry didnt realise that set e took care of it for you +probably does not work if the mapped superclass or embeddable object is reused multiple time +in the case of the user just didnt read the documentation several timesI agree we can add the same warning thats in and the same warning thats on the solr docvalues wiki page to the codecs but I am going to absolutely call out the fact that it would have done nothing for that users case I dont see how they could have configured what they did without intentionally choosing to ignore the documentation +Only products shown and no way to get to the th product +Verified with JBT. +Sorry this seems not a memory problem. +initialstate +Attached test case +Attaching latest patch from the corresponding Github pull request +So this issue is resolved for me +Updating cannot be done as it would overwritten changes made by the third transaction and compromise the data integrity +I am not sure why this is and whether this affects +C +Oops youre right +The latest patch updated to trunk +Thanks for the patchIve committed a slightly modified version of it could you please give it a try and post your feedback herethanks that I think of it a bit further Im not sure that my commit will change anythingPlease feel free to give it a try but Im reopening it for further investigation +Maybe that was fixed as but I dont think its been fixed on purpose +we have convertTypetrue in our dih config +Changed the JIRA title to reflect the Stans operation name suggestion. +We can reopen it in future if needed +Un simulacre de justcia una aparena de moralitat +tested your fix against eclipse and with jdk and +Damn it sry lost the orientation. +Seems fine in JBT +I think that is only true for Trinidad + I have this same need +bulk close of all resolved issues prior to release. +As the correlation id of any message from the Conduit should start with the given string we should get all messages with only one consumer +Thanks for the patch! Try and use the unified format next time +buildivylibHadoopcommongriddevbin +reference guide +These changes were just some initial work that needed to be done to clean up some of the failover handling code +Brill is this using the checked in eclipse project? It isnt maintained since none of the committers are using you have a patch I would be happy to commit it +batch transition to closed if someone wishes to restore one of these to resolved in order to perform a verification step then feel free to do so +Updated in patch +Doh! Looks like the final patch I uploaded to was slighlty older which missed the current changes in these patches to correctly log the for failed jobs +Was crypto +Can someone with the necessary bits please commit this? Thanks +I have the same libcairo version +Currently the category is the path + +This must be done for site launch +f +Rob forgot to close thisFixed in. +add an automatic test case to the test suite +Thanks for the patch +Hey Dan I wonder is it the XSL Processor +Got a patch pull request +Plumbing issues to replace portions of our code with existing improvements in WSDL importer has addressed this issue +If the DB is READCOMMITTED simultaneously a DB readcache PFER operation could be going on that caches a now invalid entity +Committed to trunk. +Thanks +On a KVM environment I cannnot create snapshots at all see +unused imports arent vote for next step is to add checkstyle rules for ve created an Eclipse Code Formatter profile based on these conventions +Thanks Sergey but I didnt receive my credentials yetI may do my first commit on endpoints clients exception monitoring my project will need it very very soon +It sounds like the real question is were other nodes directing requests away from the poorly performing ones and if not what did their dsnitch contents look like +REST is popular these days +Nicholas since you have gotten good at reviewing this patch can you also look at the patch +Thats a rather common scenario under I would like to move the entire +Also see +bulk move of all unresolved issues from to +Should we increase the error level for ERROR to be higher to avoid the javacore for IBM JVM instead of the workaround fix by setting to +I guess if there are validation errors the User object should not be persisted +good finds so once weve updated to this it should be fixed +Theres no point in referencing Job as a bean +If bottom right works for you Ill process those out for you +Fixed on. +This bug may not be a large issue but Im curious if anyone can validate or invalidate it so it can be closed andor resolvedThanks. +Fixed at r thanks +Ted thanks for review and integration. +Closing this issue. +So I am not sure why is the wiki page? Enabling hardware virtualization is something that you have to do with or without bigtop if you want to run a VM with any decent performance. +Thanks for the continued development +thats a good pointI will ping Tim to attach the original patch and grant the necessary rights +forksjoins MUST always come in pairs +and I have send this feedback to the company but get no response. +Test again and this time event got fired and just one time every rotation in or Perfect! +This issue might supersede but we can decide after some investigation +Maybe the patch is also outdated i goes against Revision maybe you can downgrade your checkout using svn up r patch and upgrade againI use s patch tool which is more intelligent and also applies very old patches wizthout problems +It seems that this issue was already fixed upstream in +Moving out of +The is extended to be able to issuevalidate which consist of a nonce signed with one of the block token secret keys +Reopening as Java reflection is NOT nice. +Thanks for the update +It also makes the startup commands even more verbose to have lots more solr +Rebase to recently code base on YARN +I will try to do the same thing from a remote machine and testThanks committed zip archives related to WC interoperability tests also modified to utilize local copies +no it wouldnt +Patch against patch for catch +like sec test case sort of slow +Well this is not completely unreasonable but if we are going to add filtering to the command why would we only filter on state? Why not bundle name location etc? What we really need is a way to do something like piping with grep or something similar +easier than forrest and more people could keep it up to Glen and Robert! +Fixed in subversion repository as of rThanks for the report and for the patch +Will this be added to the Criteria API once is outHow do you handle that is compatible to various mongodb versions +Looks like you remove the timeout exception which we still need to track +Applied patch in r +A better way would be to register a or probably a new type of Response Listener for the Reply +Btw David I mentioned it above but just to remind you Im using the context parameter to turn entity expansion by the JAXB providers on and off +I find that the daemon code is a bit hard to read so if you plan to have a look I would wait for the next rev +This seems to be a subtle race condition +Could you create a pull request with that changeIm usually always using JAVAHOME on my Mac so Ive not faced that problem myself +I will try to get this patch finished off and into trunk in the next few days +Commit revised patch to snapshot At revision trunk At revision thanks Rex +All tests pass. +Ive updated the design doc to reflect the changes of the patch I proposed +patch uploaded +looks to trunkThanks for the review Sergey. +Reverted by Shane who reported problems with other builds +Ive upgraded to version GA and Ive come to realize this patch has been included +Does this mean namenode will keep on trying until someone access the file and notice that its corrupted +Committed at subversion revision . +Either way it would be good to document why such a trick is being used +Ill take a look in more detail later +is the patch committed +Test steps +DIGY Understood will do so in the future +Documented as a known issue with workaround in the release notes +That would be better +How many fields and how many documents are in your indexCan you provide a stack trace from the? This would be veryuseful +This issue is a problem with both deletes and selects using multiple ORs on different columns +This implements the change Dhruba was suggesting +just as in the case with controllers and the OSIV interceptor +converted the dynamic import of the package into a static but optional import +Given that it would appear to me that the doc is wrong about being able to specify custom class types +If we want to add auditing in the handler it is useful for both sync and async producerThe new design made the following changesa +O sorry should clean your project and select Build automaticallyAnd if the partnerlink has not error you can get the operation correctly displayed +This script is maintained in the JBPAPP branch +I committed this. +I dont see no need to differ from upstream maven as well as from JPP inthis case. +Thanks Arthur! It is now fixed +Patch upmerged to trunk +This was committed by Jonathan as ee while JIRA was down. +Reopening since it seems we need to be sure this is actually expected behavior +So as a prereq for this what do we need to install into AS? Is there a single JAR distro we can depend upon in Maven and deploy directly into the server? Or some more manual patching first +The problem can fixed once is fixed +I dont think that would really work if it is a one to many relationship but thank you for your responseI dont see why not +So this can be closed +Add more unit tests +Due to some issue between the RA and SJSAS our consumers prefetch is set to +something weird though it generated a brand new seam prefs file in +if get that htd from cache create a new +Ill look into seeing if I can reproduce + +This looks like a Java bug not a bug. +namecurrentVERSIONAs we introduced clusterID and blockpoolID we need to add two new entriesclusterIDblockpoolIDinto the above listed VERSION files +This one has smapreduceversionMAPREDUCEVERSION too +Actually the solution does not apply for everywhere sometimes we can not determine exactly how to process a file according its path +Marking this as duplicate of per comments. +Confirmed in r +Hibernateadd JPA annotations view section was added to the complete +The original patch did not tell the Indexer to read crawlparse too so that it can pickup fetch datums +Which version is this with +Looks good +It was later not used because we decided not to use preprocessing for changes to validateInput and instead decided to rely on reflection +now just uses a instead of an. +See my last comment getExcludes is used both in copyResources and performPackaging +Hello AndreasThank you for the promptness I attached to post the source code of our Parser which we use for text extraction +Thanks Zheng +It would seem to me that maybe supporting java might not make sense so if you can demonstrate that in a bug thatd be great +Hmmm can you expand on whats wrong +Are there any more we should be changing +I think this can be committed +Passes against SVN head +which release of hadoop has this behaviour been seen +With Seams conversation context we used to have the same kind of problem +This happens on every key press because structure is changed on every press +I may break this out to a separate JIRA for commit so that the patches are manageable to reviewStill need to implement the rest of the above proposal +Whats the consensus on this? Id love to commit it so you guys can stop rebasing it all the time but I lack the personal experience to give it a review +Service works fine and returns the expected response with no errorsTherefore resolving the issueThanksIsuru +We now only have a CLOB and a VARCHAR column with he short two tables and reusing the keys from the owning table is a simpler solution no need to generate additional keys so no ned for additional sequencesequence table +The first one is set at startup by each server to indicate the class that will handle security +Theget method does not currently fulfill this +Im looking solely at the larger tarball and not the tarball +I have added direct references for the source class to wherever it was needed +Please do so after you address the latest comments +Ill see if I can get a test coverage for you +that it is a surprise the optimizer can handle that case +I will look at the test failures shortly +for me this will attach my working looks like in JDT they just hardcoded zip jar as for example and +Change pushed to trunk. +containing sample for review +Implementing as to should fit this fix version +Are we sure thats the fix? Because as of Friday we still had to change to in order to get any meaningful output +Tracked this down to how RF handles resources lifecycle + +Thanks +Can we have more info on thisI cant see any param number in the pasted config +Can you please close this as a duplicate? Thanks +is useless for git users as far as I can seeIs this related to? Is it fixed by change +I guess val is another special field to account for +Delete the subdir +It is interesting that according to that section in the SE spec the limit could be less than if some parameters are large in size Note that a method descriptor is defined in terms of a notion of method parameter length in which a parameter of type long or double contributes two units to the length so parameters of these types further reduce the limit +Please review the attached converted test +Thanks for this fix! +Difficult to understand but you probably have good reasons +Ive opened a bug for this some time ago perhaps we should disable this test for now? for example if one of the Java tests fails the C tests dont run at all +The POM packages the code into an executable jar so DDL for a given model in a given schema would be generated with this syntaxjava jar dialect model Basic out tmpId appreciate any feedback on the patch +and call the service client with the that +Hi Matteo Thanks for putting this proposal up I really like this direction +I generated them wrong +Patch for for Branch +Well he belongs to and they have rights to comment on issues and create attachmentsI think he might have forgotten to Login +Marking it as later if necessary reopen it depending upon the discussion +Also the SVN links are now incorrect +They would like to directly access their own dashboard as their homepage +no problem Im glad youre on top of things +I hope that you are not avoiding itIm going to quote myself to answer the questionbq +Either rejected or a dupe +libhomeedwardworkspacebin +Kannan would make sense +Completed At revision +Comparators or things that extend them as some Writables do +bq +They want their regexes and wildcards to workbut you have it backwards +I have now updated the axis engine to support all XSD type for deserialization meaning the Axis engine can now serialize and deserialize all the XML Schema defined simple typesI shall now work on updating so it is possible to generate stubs for these additional types from WSDL +Instead of always sampling the first record of the block index now contains the entry of first record in the block for which join may happen thus saving time spent in fetching right tuples over the network which couldnt be joined in any case +Building a regression test for this scenario is not easy without a fault injection framework +Committed revision . +Thank you to Ivan for the original patch and thank you to Chuan for the additional help reviewing and testing on trunk. +jco Ha thanks for cluing me into that +Thanks for the comments +Everything works well as long as camel dont try to lock entity before removal operation +This should cover the main thread as well as threads owned and launched by Ruby +Im getting an error Server string invalidWhen removing my proxy declaration all is fine +Because it is customary to install stuff on Windows platforms under cprogram files this bug creates great difficulty for shipping products under Windows +More bed time thinking to come +bq +Just the gpg plugin part of Dougs patchTested this with a push of rc and seems to work fine +needs to be added +The JDG plugin solution works fine. +the TCK passed with the integrated patchestrunk the and TCK passed with the integrated patchesWill leave it to Mike to determine if the patch should be applied to the release +bq +Vikas +Robert I gave you edit rights for mimej +Sending srcmainjavaorgapacheservicemixbeansupportAdding srctestjavaorgapacheservicemixbeanTransmitting file dataCommitted revision +Tested! Well done! Please add an IT to prevent any new regression +If I shutdown broker B the network channel on both ends +Invoke it like thisant pathtoWhere pathto points to the root directory of an unpacked hibernate distribution +Has been fixed in the nightly build +It says ok +Test is jaxwsjbws resolved issues for already released versions +hii +I think it should pass with this +Pull merged on master and on +Oh yeah they are dependent on +So i have not added a new test for this +Ill have to try out the new version +OK +Also modified the meesage to be displayed for the case of with empty list of users causing stack trace with to be displayed +This new patch makes two changes in my newly added unit test +There are some very minor tweaks Id like to do some time but not worth risking breaking the website again. +HiI am facing the same issue in Eclipse Indigo Java EE distribution on OS X lion +Thanks for the patch +The svn command that fails issvn copy file tmp +Ive tested and on linux +Thanks for spotting this. +Depending on how quickly this patch is completed we may start hosting RPMs sooner than later. + last Vera the bug was fixed in AUTH module at rPlease check that the fix fully resolves your problem +The SQL standard is single quotes are for string literals and double quotes are for identifiers that need escaping +Thanks FrancisYour patch is in at revision In I also added the default values I suggested +I like Henris second suggestion better +The patch seems to fix the issue for me +If you find further issues please create a new jira and link it to this one. +Attachment has been added with description sample input has been added with description sample has been added with description sample current that fixed it +Its easy to move the dependency to extensionwms +We will be addressing this problem soon but this error log entry does not negatively impact the plugins looks like when you tried to update the plugin P failed to install it properly +Thanks for sharing background here + +We might need to reconsider the way we whipped this together +No test no doc +Ben is looking at system test as part of system env before we can expand these tests +Works so nowThe attached report shows current grouping in +I hope this is ok for you +Attaching my presentation on Configuring Derby for Performance and Durability given at US in AustinJean thanks for making these presentations available on the Derby web site +this were confirmed to be correct to do +Reproducible on Android on Samsung Galaxy tiny on handler that routes all button keypresses to +MatthieuI think the best solution in this case would be if the formmodel had a switch that enables this behavior +For clarification the Value Expressions in the Expression language are parsed by a custom parser and the Logic Expressions are parsed using +If some problems are there Ill file a compilation method will be a great feature which surely improve think global variable manipulation of JSR is bit weird and should be rewritten +This patch is huge and I would like to view at it mre in detail so my proposal would be to include this in a releaseOk on code you are using the errorgetlast method +Thanks you have a very fast response rate! Ill look at it soon +List closed. +Since the old user forum is lost I found my old code is an concurreny example using different in the hibernate mapping files for how to create a table in the the readme to general this kind of functionality should be pushed into the connection pool since there is no really elegant and performant way to do this via a decorator +I think the problem here was that particular message has the title Error +This is using the snapshot from march +Here is a newer version that fixes the commented issues is now +The issue here is just one symptom of the underlying problem in. +any chance you can post a link or a zip of the results of this test +Have to fix WSDL +If youre willing to help track down the locale specific format list for the various built in formats the POI will happily accept them and Tika can then use them! You might be able to find it in a Microsoft spec somewhere to save you having to work it out manually by switching locales. +committed to trunk and +Upgrading to did not have an effectUpgrading to from the tarball you uploaded worked! Im able to get all the way through ant clean test repeatably +Committed to TRUNK +bulk close of obsolete debugger issues +Right so we have to fix the schema location generated so that it includes the schema prefix +I have marked this issue for Newcomers +Ported the roundFloat method from +verified by Vladimir. + +Yep this was a bug in but should be fixed in SVN trunk +I dont know why but in my mind was an +Okay well based on the current discussion on picking an impl if the ibm impl is picked then I will work with Nick to get my mavenized version of the ibmjasxrs codebase ready for a wink drop +Thanks Rossen pleased to help such a great projectMy CLA is . +This could be implemented without messages but performance would be bad so Im marking this as dependent on them +One possibility would be that there would be a mix of old and new files +I am going to fix it +Fixed in r +Maybe theres something different with ThriftIve seen people having trouble because of the message size too though +in +However if it only synchronizes block reports or not sync all heartbeats If we do this then it will have some impact on the failover times right? At the time of failover the backupnode will not have all the information it needs it will have to gather those missing pieces of information before it can start servicing request isnt itI think all block related information should be forwarded to BN and the only different between us is how the info get to BN isnt it? You could dedicate a link for NN to backup node communication to improv if all communication from datanodes to backupnodes flow via the NN then in the case of NN death the backup nodes might not have all the that was sent from the datanodes isnt itBN should keep up with NN otherwise the failover cannot be seamless +Documented for +Unable to find any more cases changes merged to trunk last night. +The rollback occurs +This occurred during building of Greenhouse with a empty +Logged In YES useridFixed in CVS by the import of. +I think that in general this is something good to have for Lucene overall however I only took care of the search +Thats what happened in the create case +The object that is reference has hashCodeequals methods and they are not called +As discussed in IRC channel with development team to provide a forge plugin is required git as project repository and to publish in a intranet needs a new file +ThanksHoward +Seems OK to me too +I am referring to the Sonar Jenkins plugin +Clearly that second constructor is currently not very useful +There were some whitespaceindendation changes in that werent neededIve attached a new version of a patch that fixes these details + +Ill look at it on Monday and fix it up to where it was before my changes +Also I dont see any content in the site or test +Please be more specific about what you are actually trying to do and why you believe theres a problem +LiluI committed the patch to SVN trunk +If Tools is not responsible for the Java Model Exception above then we can resolve the issue +bq +Changes +We extract simbol by simbol with rispective coordinates for each simbol +i was was grepping for +the issue is clarified as development but not just a bug fix +I have changed the summary of this issue +Here is a simple example that shows the error +Can you please confirm +As you understood TP makes dependency resolution independent of IDE so I cant see any reason why this would not work +As a side note we still need Java for forrest in the site docsIs there an ant target for this? Otherwise Im not quite sure what youre talking about +If we have to introduce new classes methods in otder to fix the problem itwill have to wait until it is just an exception +The patch is awesome! The directed graph interface is very simple and can represent any complex query plans +This looks cool it looked tricky to me to implement that iterative normalization api Im glad you tackled itIts a good idea for some tokenizers like Ngrams etc you really need normalization as a charfilterIll see if i can help with the integration +Over a month old now if this is going to be backported to do it soon or were going to close it as a fix +There is a difference between corrupt and invalid replicas +The people that own the build did not really put a suitable value in the in the POM she feels so she just entersa value in manually in the RPM generator project is completely separate from the original s artifact is an RPM and SRPM in shortShe has to do more work because the POM does not have what she if the POM had a summar element it could be that the owners of the POM filled it in and she could use suppose they did not +Hi Alena Yes we are going to drop this table and start using volumetemplatestoreref now +Felix can you check if it is ok +Invalidating the cache when the exchange is deleted wouldsolve both standalone and cluster issues +Ryan any updates on this issue +No new tests because its a pom change +I think this was just a refactoring omission +Hmmm I dont think its advised to mess with the default charsetAt least do not commit it directly but attach a patch we can take a look and discuss +For the attached patch I have tested it with followings Tested platforms Redhat AIX Susu Linux Enterprise Server Ubuntu Tested scenarios startstop a module with correct module ID startstop a module with incorrect module to trunk At revision and At revision +This helps factor out complexity in the compilation pipeline where in this case performance is not something were afterThe only comment here is that get in is Hadoop specific so it should be moved under hadoop backend package +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Flush should not mean that your bytes have reached the disk +Have you tried to restart the node and execute update or dropping mt keyspace recreate with with initial values and update? I think that assertion from could have stalled the executor involved in saving schema to disk +Works for meThe code was testedd both on JVM and On an exception is always logged as INFO which I assume is theintented behaviorThanks guys this is great +Yes it is +API that version is for a more intentional API probably based on JODA or JODA in JDK +Applied +It is not possible to extend it by overriding one of the methods +The upload form should be in the More Actions menu right? It isnt there maybe because the bug is closed? And I dont see a way to reopen it +The last column header does not stay the width it was given when its partly behind the edge +to Darran for that purpose. +This eliminates my need to use the patchThank you. +If I understand it correctly the number of slots cant change for a given deployment and every server has to agree on that number +Also added stronger testingsvn commit m Check for possible corruption usign Javas since its stricter than rubyzipThese checks would have caught early +Sure I think we could commit this to trunk as an interim solution +Need to update the book could be too late Ive already sent chap but will ask it back +The patch looks correct to me +It stands forAgencias the of the project to release. +Closed this issue and created a clone for the proper fix at +Closing this issue since it is no longer needed +We know the system wont boot properly today if the master is not started first and initialize the directory structure +Blank values are also ignored appropriately +Most importantly for versions Check also made on any event WS API updated done. +Can you try with that +This lets us produce intermediate results without finishing off the is likely improved as well since it wont have to digest all the data anew every time a digest is requested +Attached the patch for +I have attached a test case that show how to do this. +I am currently out of the office +on ant build file revision . +Running tests and thanks Jarcec! +Attached committed +I was! thanks for posting +I usally did that for testting on eclipse and i forgot to restore it when i was trying to make the patchjoin and join are reporting error now will check and repair +If you feel strongly that this is something that we need to do please post your proposal to list so that others can comment on it +This is the patch I threw together to fix this issue +Willthe equivalent code be marked as deprecated? Same plan as with base +I updated everyone to depend on the released version +dude go report to Oracle why they report for this queryThe solution for and the workaround on for it have already been explained to youLast comment from me here go to the user forum if you want more help +ok +Changing priority to minor +But following the changes happening in jira etc +Youre right +Some people will be very happy about this change +Applied to trunk and +Thanks! +We are also willing to help them fix it but I dont think it belongs in our tracker right now +Does this handle files with a single lineI also dont understand the change to +If the calls throws a or a then bail out otherwise continue to retry the register call over and over again +I have updated the createcase functionality so that it sets showissue to t when a case is created. +Then when a process actually needed devrandom there wouldnt be any entropy left and it would block +This is mostly to support multiple unwrapped objects without requiring deserializers to declare things they accept +Now that is in heres a version with just the changes. +Any one of the subclasses should work as the proper converter + Performs password strength checks using following rules Number of characters Presence of upper case letters Presence of lower case letters Presence of numbers Presence of symbols Repeat characters Consecutive upper case letters Consecutive lower case letters Consecutive numbers Minimum requirements minimum characters in length contains of the following items upper case letters lower case letters numbers is there but it has been postponed to +Added a diff of the above applied commited Completed At revision +Marking closed. +Yes DFS and any like all already cache file poition somewhere +That is really odd +Hello AndersDo you feel like solving this issue? Ill handle this +In wrong jobconf is getting passed which is clear from the context +if there is no assignments then is requiredbut if there are associations then from element of association is used and it seems that is never usedSo it is possible to have scenarios without +Hi SergeyCould you please attach your or other project file so we understand what youre trying to do and how youre trying to do itThank version works well +Thank you for solving the problem +So I think we shouldnt change such functions +decision was made to leave the current name as is and only introduce Remote +Ive moved this issue to Sonar Plugins project because we want to decouple Sonar core from Maven in future. +Opened for defining a proper SPI. +Not very nice but you can understand what they mean +Committed revision . +thx +In line with this I spent quite a bit of time thinking about pros and cons to having Path object support backslash VS not +conflict seems to be popping up in object +Draft version of a Mac OS X wrapper can now build Harmony without any etc +Which forkmode are you running these tests with +I dont think we need to worry about this too much in the samples or test packages but if someone feels like fixing those too they can +v and v are the same +Disregard my old comment +The grammar has been tweaked so that the it now accepts a single optional trailing semicolon at the end of the BIND or DO clause +it can also detect ivy cache corrumptionits really unrelated to actually how we get the jars thats an impl detailwe should be checking that its the jar we tested against or fail hard +Attaching final patch +See attached README file find a new proposed implementation of STARTTLS for James in the zip file named STARTTLS is now configurableRegards submit an diff against the branch +I think the HTML looks better with the example titles and with the code blocks reformatted to avoid unnecessary line splitsPlease have a look at the latest attached HTML pages +just committed this +verified in Tools to set release note now release note status has been set +Ah sorry about that failureWould returning an empty suite in when running with JME fix the problemI can have a look at that tomorrow unless someone else beats me to it but then I have to setup an environment for running with JME to verify the fix +fxp file here +and session replications are working with Jetty. +I looked into the failure and it had nothing to do with what this patch is all about the failure was just demonstrating a known fact which is that even with synchronous replication a client can fail over to another server before the latest session replication I wrote test I was worried about that kind of spurious failure +I however will dig deeper here and revert back +When I did some simple tests I noticed that the pig parser borks when confronted with default instructions maybe you already have a ticket for thatIll try to tidy up the code and attach it sometime tomorrow +Hopefully you agree that this area is very user unfriendly +new cleeaned up and updated I fixed small bug in HG log consumer when empty lines were reported as warnings +Working a fix to cmakelists now +Revision addresses this by in AS trunkejb this was a flaw in my yesterday +Editing signal name of intermediate catch dont think it was changed recently but give the latest version a try and see if what I do in that screen shot is what you BZ vs JIRA and links yes we do prefer a JIRA in the community and linking that to the BZ is useful +in a few days i can upload a new patch +Junping Luke I have concerns with the proposed approach I will comment later today +But it does not belongonCraig +Added keyAlias as an attribute +Hi AndreasIve had a look at this and have a few suggestions to a solutionAt first I thought that I could just check if the menu had any items and skip it if it didnt have any +I dont have a strong opinion on this what do others think +the back button wont work once youre in Sourceforge downloads as it redirects forward via javascriptThats essentially why I consider this particularly important for the download buttons but no strong idea on the other links +Yes decorators package seems bestI was thinking it would be the exact reverse of holding both a Set and a List internally just implementing List this time +Not including it in is the release managers job +Divesh The changes looks okay +To elaborate currently is used if the is set to +I forgot to mention that compiles this patch successfully +If you get a chance can you try it out and let me know if you still see the exception +Patch updated +The updated patch looks reasonable I am assuming the test failure is unrelated +The file should be not +Maybe Im asking for something different instead of a change to. +The application works well when deployed to though +It locks us into that field and I cant see the benefit myself +Have a look at +What were the real paths you encountered? It would help isolate the problem to know whether the source and target paths happened to be within a single underlying source or whether the source and target were in different underlying sources +Bug confirmedRonald had you tried with previous versions as well or is it the first time youre using the Sonar Runner +Arun how is this different from option in +Thanks Adrian. +I cannot delete the file until thehandle is garbage collectedRegards +James I cannot reproduce this bug +Implements the new feature which allows a key attribute to be specified in place of the name value and label attributes on ui tags +Removes overridden method no longer required by Graphiti that was causing the issue +Im gonna add file in Karaf bundles +Passes all unit tests including those on linux +Sorry +not all tests executed even when test system property is set +This will only be added to master +to trunk and +We should lock and we dont do it in trunk or head +Amir any further comments? Something else to mention is have a look at the interface added in Spring RC +Offline is only for parents of splits not yet clearedStarted in on a bulk assigned for enabledisable of table so can have an executor to run assignunassigns in bulkDTH is a messTODO Add disablingenabling of table +NDC should also be copied +so this issue is not urgent +That schema works for meIm unsure about the alternate encoding thing but my gut is telling me no +I see that the Sun JME tests have not run since September +Because the idjit who runs this project hasnt been updating the release versions +Regression tests run cleanly for me +Resolved with latest SNAPSHOT hour regression test No access violations and no hangs +Afaik classpath multimodule and simple should each contain such a script +Not sure on the semantics we should use maybe a new filter or something and clients can choose which one to use +Set your terminal to ignore the terminal bell +Hadrian Jonathan what do you say +Im almost certain this is a problem in your application configuration +SMALLBUFFERSIZE should do +How do people like the format for output? Its basically broken down by doc then field then term then term information See the unit tests for some samples +Id prefer to supply something for clients that works correctly +related +I had suggested that earlier but you pointed out that it would have this problem +Since we dont store conversation id in the page context anymore maybe the usage of viewMap introduced in can be reverted now +We also need to mention that this is not a supportedcertified configuration +False alarm MERGE doesnt get INFO messages from different clusters because The transport discards messages from different clusters If MPING or TCPPING is used the carries the cluster name and discards messages from different clusters too +changes looks good but why not actually bump the version number of the plugin instead of adding in the end? sicnce the discovery plugin is to be released independent of the specific JBDS version bumping version are the correct that can be done for future updates for now no need to spend more time on redoing it for this small fix +Fixed under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +the process of signing off on a release +Fix version is to M +Andrew is this still valid for +Thanks +In Cocoon it has been fixed by implementing universal Object Model that is initialized independently of flowscript. +Incidentally startup is slow is definitely on our radar +I had to do a sudo npm install g cordova +Committed to trunk and branchx +We are in agreement +The attached patch just does what is proposed in this issueAdditional thing we can think about is to add fields to the website which are available in the ofbiz fields in like the local entity and mainwith this patch the web will be taken from the entity when they are not defined inOr should be better use as the default which can be overridden in the Website entity +Finished refactoring of the first of CTF test cases +Yes +The pom seems to be missing K K +Long term API stability is very important and this simply boils down to a documentation issueIf we changed the external API every time we thought of a slightly better name things would be quite a mess +And yes there might be some incompatibilities with Maven should I create an issue for that or did you already do it Brhave you tested the changes? RegardsLars +Josh that makes a lot of sense! much better than changing an interface +another test and another fix a single correctly spelled word will come back with correctlySpelled revision Thanks Mark +Oops! Fixed in new rev of patch +Thats fine +Reading from JMX seems fineThis also isnt DRY +Will work towards integrating that into now. +I closed this myself because I found out that this had to do with dependencies and was stated in other bug reports +Juhani please attach the latest patch to the Jira make sure to grant license to Apache +Addressed the nits +This test case under Tuscanyjavasdoimplsrctestjavaorgapachetuscanysdotest directory would demonstrate the problem +You should also complete the issue description and explain how to reproduce the issue +was created just to fix +I can still not see this in the latest files updated build restarted. +Done by using an event based the version +At least the GNU windows patch doesnt like will recreate using svn diff +Tom just to confirm setting the message body of the resulting message eliminated the oddunexpected manner in which the DB was affected +We already know we have some issues with deploying into tomcat thought we still havent had a chance to work on you provide us the log of the failed tomcat deployment? I want to make sure it matches with the one weve already seen +fixed the same problem in trunk +Added performance test +This would mean most likely saving off a codepoint for an ungotten character rather than saving off a byte value as in mode +is a Runtime exception and private to YARN +As penance for being slightly pedantic and because Mikhail seems to be catching up with my JIRA backlog +Provided patch for trunk as per the above comments +Hi HowardWith Joined table strategy any update in Child will involve two update because the version column is in the parent table and the child field is in the Child table +Ronald you are right +Ive been running it and there is no traffic to the disk while the clients are watching +When will snapshot have changes which was done? +Nobody apply this patch though because doing so will just create some merge conflicts as I take the necessary steps to get the http stuff into common where it belongs +This includes the followingarchetypessrcmainresourcesexamplesexamplesexamplesOther readme files in the source will be reviewed when I return from vacation +This issue is well known and only occurs in JBDS +Ross Im glad to see you like the work +Putting something in a release which we are not very sure of does not seem like a good idea +However I dont see this cruft in my subversion client +Maybe a good solution would be to have catch any exceptions that occur during the rendering of the string and fall back on emitting just the component ID plus caught while retrieving component details +I forgot to a bunch of tests in that I had commented out to reduce the time it takes to run while I was debugging the test failure +Ill point to the commit log shortly +Assume closed as resolved and released +Worst case scenario there is an unmatched ack exception and the message is eventually redelivered and should be seen as a duplicateSo failover among a network of brokers will always be a little more lossy in terms of message ordering than master slave +Danny +Hi JustinPlease review the new patch as per our recent discussion +feast or famine I see +Code changes look good +This affects XA transactions and probably could be release noted as we now support +I created this issue by mistake +Can you please also post a reviewboard diff Maja +Is it possible to post the contents of to see how it is implemented? This is probably the key to the problem +Thought you wanted a softer failure +Thanks +Can you provide more detail about what you tried and what happened +HiI am running the example of Xerces provided using my own Memory Manager +Thoughts +This patch looks good to me +Thanks John for showing how to do this with pure +This is meant to address something Juhani raised in that jira +If an id is not assigned then it generates one +The scarborough is the standard theme for nowThe is the hello world exampleI will add a section to the faq +Ive noticed other places where this problem manifests On disconnection of a subscriber youd expect the underlying queue to be removed from the list checked by SCD +The purpose is very great and definitely necessary for us +This would break everything +I think we would have to move and to different package say on the same lines as java which has it in different packageGood idea Sharad +This is very interesting work and Im looking forward to playing with the Phi failuredetector +Steve +Bulk move of M to blocking M +I think I will buy deceptive but deceptive in the way that passing the encoding around the way client generally does in other places is deceptive +Currently no unit tests for this. +I dont think we need this +Agree Ill amend the description appropriately. +There are probably more efficient ways to do this + +If reporter is always null for feedback messages in the session we can remove the clear entirely +I can be wrong but this change makes thread unsafe +In which case you dont want the equivalent of development reminders peppering your log do you? +We probably should do the same +This was pretty easy its right there just is not implemented in the mapping +For example make the enqueue count match the queue size on restart +not an issue then. +Ill close it myself +If it helps Ive been using the patch for a while no issues since Ive done that +Attaching and with the following new and changed topicsM srcadminguideM srcadminguideA srcadminguideA srcadminguideA srcadminguideA srcadminguideA srcadminguideA srcadminguideA srcadminguideA srcadminguideM srcadminguideA srcadminguideA srcadminguideA srcadminguideA srcadminguideM srcdevguideM srcdevguide +Well let it go this time but lets try to get rid of those in the future +If there is a patch for the branch please upload it here so that others may apply the patch manually +Tom and I have had a chat about this issue and there are a number of reasons why this could need you to provide us with more information as to your intent and to work out the message payload contents at the time the action is would be better handled on the forums +Unsetting Fix Version for unassigned issues +Sorry didnt see your comment +New version of patch attached +I thought close wouldnt affect the flush and sync of +Hii dont understand why we can do it with Standard Java and not ith CXF +Is that ok? +Hmm +Attaching +If you can get forrest to work with a subdirectorygreat in general though weve found it better to keep all the xml files in onedirectory +Hrm I agree and apologize that was silly +Correct operation has its costs +Applied at revision thanks +Correction It should be misunderstood what you where asking for +I propose to add a test query using an implict parameter in the having clause +Currently that method is not working for beans of the same class out of the boxShall you justify your previous comment and clarify why the Date is exeptional in this case please +sorry I got the sql queries wrongIf you want to search for abc in you have to use value like abc or value like abc escape In Oracle or HSQLDB you have to use value like abc escape and not value like abcAs you can see simple escaping with does not work with Oracle and HSQLDB so I wanted to use another escape sequence +Adds to source assemblyIts also missing from the dist assembly though this may be intentional? If not Ill update the patchSince content is trivial skipping dont think it was intentional +As discussed last night the Infinispan integration may make this outdated +A few addtional tests were addedRunning derbyall and +Kim validated the fix against his test setup +Ive created to track the release +Thanks for the advice Rick +Work around on r will report to ICU bug trac later +Ill test redback with Continuum as well to verify its behaviour +If desired make this a configuration option +No problem +Beehive from SVN is currently buildable. +committed this to branch +Verified +Sorry can you summarize what the results were and whether you used averaging and stepsize reduction? Ive been using max iterations though models tend to stop anywhere from iterationsFor NER do you have a breakdown of PRF for each type of entity? edited the original posts and extended them so just scroll up and look at my first posts where I reported the resultsNER was only tested on person entities and I posted Precision Recall and +After that point it will switch to always using the proxy server even for remote repos configured to use direct connect +This resulted in special processing of GDB attaching in case of SIGABRT because SIGABRT cant be rethrown by returning to unchanged signal context +Why not simply change it to Commit by server adapter +I dont even know what scout is let alone how I got listed for it +lineBreak didnt know about that one thank +Whats the latest on this? I think Jason had some work with FJP he was testing out +The problem in indeed in the OOMEs +Brian I dont see a attribute on the element forSo does the offset only apply to standalone or can we add a attribute in +Problem with the beta build of IBM JDK. +Fixed typoAdded since timeout is used in testWe have tested this in cluster when the active nns nw broken +No change to running code +this path would put the hadoop logs to buildtestlogshadoop dir and the test logs to buildtestlogs dir +I think have not done it yet in any of the poc +Please try modifying your sample code to run out of loop declaration first and then run for in loop declaration +The old distribs have been removed +Thanks everyone and this is really cool +and should correspond to each other +Yeah I tried to reproduce this several times on different machines +But this is how the resolver works in the framework we have an entity that has a set of requirements associated with it we dont just have requirements +Thanks Dhruba +This was causing most of the fails waiting on recovery in tests +IT IS NOT A BUG +Please note revision should be a part of this as well +The from Hadoop QA was because the uploaded branch is for patch +Thanks RegisI will separate this patch +I have fixed this in is not a bug +That issue should have been rejected anyway as a new decision was made +bq +Could you help to review it? thank you very much +Dima investigate and if issue outline how to reproduce and if not reject it +Its done by Thomas +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first. +All these are node level information and are common for all the storages that a node managesIf you look at the proposal from and treat say a datanode as collection of storages all the above information is needed only once at the datanode level and not per storage level +Go ahead and deliver it to the branch for and to Denny +So can we close this bug +conversions but not for the boolean +This patch discards the connection prior to hive script generation +Attached is the report for the derbynetclientmats run after the fix for +Broken after if you use then the delta parameter is ignored +Committed +When I said earlier that we should deprecate I didnt realise how difficult a move this would be as I didnt take into account that people used it in their own custom couchapps on top of Couchdb +I faisl on the missing dependency +reopened since alankila wants to add a patch for more improvement +It confirms the analysis I based my patch on +now both derbytools and derbynetmats pass the tests +Linking that issue to this +Everything seems to be working properly +But I agree on the rest just sayingRight +Just a bug. + but lazy dont works when constrainedfalse hibernate does eagerly fetch for association. +Is it possible that this is happening when you didnt actually rebuild the APK for some reason? Maybe nothing changed so it didnt rebuild or maybe the command failed and you tried to sign the old APK againIf you delete the old file cordova build android and sign it again does it work reliably +Lowering Priorty delay until unassigning issue +We dont need to change the comparator in any way just to add a bit of extra logic to prioritize these needy jobs +Checking it in +classloader issue +Marking this issue as resolved +I tweaked the output to be more informative +Well no offence but to me this plugin is in a really bad shape +It seems like something regressed for loading FFI from in RC +how about because trailing spaces are ignored for character comparisionsPS +If you happen to find it send it my way pleaseIm almost certain that the tika error is a problem with ppt docs. +Lets target with this change +Thomas can you look into this and address it +I have this fully implemented locally now +The intent of the issue was to really clarify version support +is a quickstart that show how to use CDI together with Producer and Consumer methods to inject a session is a quickstart which contains a couple of CLI scripts that are used to setupclear the repositoryAll quickstarts contain remote Arquillian based unit tests +Amedment I should have also said that the updated release notes will go live tomorrow +Its probably added as part of refactor +Please provide me link to +I converted the named anchors to so that the entity names dont look like links and I also made a small change to make the code compatible with other languages +I am not really thrilled to be breaking out new you like me to rename the method? I always figured was named based on the fact that it was a ReferencedEnveloped we were adapting to +Ive done all I want to on this for now. +Feel free to log other issues if you see a need though. +we should do that in Bill s branch +Great Job +It shud check and make sure the vectors have the proper data in them as well +Moving to so that can be released +Ugh +I believe a Store interface can help here which provide access to data by a key Looks better +Sorry somehow missed it when I looked at it firstCool then we came to the same conclusion +for the fix +I am now working on this the single big commit is just too large +Since there is no feedback I attach it once more +Thanks +The link to the code sample is broken +step See if compiling and all supporting classes with gcj into a SO is viable +Fix turned out to be a good thing even for use caseAnd best of all this can go in. +If its true we never exposed fields we could repurpose this as an RFE +We jacked up the debug level to no avail +ill make a to go with this patch +Cheolsoo already committed +Fix test failure add whitespace remove irrelevant changes from trunk +I just did a test patch it works patch with thankstrunk +Looks good +excluding condperm now the launcher is broken +The classeswill be removed inPlease post about that on the dev list. +patch attached +Patch makes sense +Binary compatibility relies on the API signature which includes the return type but does not include the throws clauseSo you can have a binary compatible jar which is not and. +fixed in CVS for Im reopening the issue I will add a test case with a single page and fix it as added and fix committed for the problem +Ill take this as part of Bug +The fix was to use a path location instead +Well this is the correct issue for node import for federation I assume let me switch it to great merged into the branch and into the master branch. +meanwhile why dont you just pick up the from and make the necessary changes and test it out +Can you please try the latest snapshot and let me know if this is resolved for you +Among other things I had removed documentation from the working copy +Kindly let me know the wicket version in which this issue has been fixed +stat shows more files than appear to be in the +I think i will attempt to revisit this in my spark solver where MR passes dont suffer form intermediate persistence or restart costs so solver fusion within same step is not much of a problem there. + +I couldnt replicate the first bug but I could the second one +This has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Maniax says it needs more tests + to trunk! Thank you Edward! +What was the ldap query issued? From looking at the code it appears this could be expected if a limit is pushed down +ThanksAs you commented write a document for this issue please +I did it myself now that I can commit! +In a way the current behavior is more secure +on patch then +Can Apache projects rely on components whose licensing affects the Apache productApache PROJECTS cannot distribute any such componentsHowever if the component is only needed for optional features a project can provide the user with instructions on howto obtain and install the work +Patch to make parse fields tag added the patch in revision as proposedThanks for the contribution! +Contains files ssltsl capabilities for Go as well as the tutorial files to utilize it +After talking to Dart +Cant hurt to create the bugzilla and see what they have to say about it. +Applied the patch with thanks to Claus +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +This should either be HTML or XML of if we wish to use a WSYWIG editor then we should store the document in ODF or some other open format +In it gives the error Multiple artifacts of the module are retrieved to the same filewhen using ivyretrieveThanks am using version and until now everything was fine +Ive just changed to +Having said that if you happen to have some code changes to Id like to know of those +seems to have been fixed in +Plus Ive been kind of missing it + +fixed in subversion repository as of rthanks for the report +Attached file with new functionality +I think it is actually a good thing to hardcode this as it provides degree of consistence across configuration files which means that when you go from project to project Spring configuration sections will always look the same +the component that is loading has an hourglass but not the full window +Where are you trying to change thatCheck the Shell start script +I have been debugging this issue and found that for some reason for a value we have a which does the conversion for char to territory based char equivalent at execution time +Verified that the bug is present all the way back to +There was a simple workaround for this put in place months ago and Ive yet to find another place where this behavior would be needed +srcblockstypemapperspostgrestarget +If there is no new version then OBR could just return the same input stream +Logging containerId to jobHistory inFixes a couple of issues with trackingUrl and hostname for attemptFinished events +Closing resolved issues for already released versions +Ill have a look at it +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Attaching and with Knuts suggested fix for this topic +This is how views in Eclipse generally work. +Attached patch resolves the conflict with AMBARI AMBARI AMBARI +Code spaces enabled by default performance looks acceptable to me. +Just set your maven version for SI to can also download the xsd and modify your namespaces to target your local xsd instead +Thanks for the patch Yang +Is this the right behavior of current design? +The idea of the interceptor sounds good +Thanks stack +It is probably a good idea to document what exactly is being computed +Change to newly created Sandbox component +This seems to be fixed in the latest code +Please let me knowof any additional comments or suggestions +I got why it failed +What exactly does that mean? Could you clarify this a bit +Attaching +Committed revision +Appears to be fixed +Since config for these is not listed and both are turned off by default I assume you are not using the idle object evictor or pooling prepared statements +Good patch +I also altered them to conform our checkstyle policy +Thanks Tom! +The error was the default EMF value for nullable is NULLABLE which has been that way for if no nullable value is defined in DDL it was ending up NULLABLE independent of any PKUC now check the PKUC to make sure its column is NOT NULL. +SamindaCan you please confirm the observation and comment or resolve this issueThanks +Do we need something for selecting package as new Java Class wizard does? I mean Action package text field with browse button +Nearly all projects that use Confluence did not take us up on the offer of testingWill come back to you when I find something +add support to etc in row format +We need to figure out how we will enable extended request handling and the registration of extended request hooks +Looks good to me +On top of all this MRI does not behave this way +Ok so What is occuring is that we start a Retry operation +Thanks for reviewing Bikasbq +Thanks Rudi +Is there another benefit that Im missing? I have no doubt that will take advantage of AJAX approaches but again its not needed hereI am all for creating more ways to incorporate technology into +But I maybe wrong about that +Has anyone looked at this? Any reason not to default the jsp pages to +bq +Thanks. +to v the fix in the delta to branch +Heappears to be making heavy use of the NNTP package and has contributedimportant patches +better wording for hdfsUtime +The JSTL dependencies are now declared and creates the views. +I hope it ok to do that? Thanks once and please report back the results thanks in advance +I needed the last build where can I download it +I think for the external users of merge it would be fine +Yeah almost +Some random shamanic dancing to maybe appease is going to fail with the same error message! See my earlier comments +I think I can try whipping up a smaller test case seems like you are disabling and so existing conflict should not cause exception +Once IDM is integrated with AS we can leverage the AS directly. +Thanks Dennis +It seems that unit tests were already broken because of this issue +You say its broken +since the price is being embedded in a dynamic javascript I was kinda forced to hardcode the for currency +Just for a heads up I think this still occurs in karafI tried installing the in karaf and get the No matching native libraries found +What problems does this cause? It could be intentional +In order to do this a new would need to be returned for each segmentOne possibility modify set to return a +When the Oracle Java file is not present either because the user is building from a source release while disconnected from the network or because the file for Oracle Java javadocs is not downloadable for some other reason javadocs will be built and the build will not fail though an error will appear in the build log +For img tags having a wicketmessage attribute I have to use the relative path +I tried something like that a while back +Create a new patch after some other updates +Message still very odd though +So lets look at what wed need to do to define a date format +In fact it looks like we dont cleanup any of the xyref tables on network cleanupaccountnetworkrefdomainnetworkref etc +A good ide with code completion takes all the pain out of using the DSL no matter how rich it is +Thanks AndreyI pulled this into trunk and. +this just makes support easierI think for starters I will just log the places and the reason where we change to in where we werent before and leave the NPE +you may want to use block mode where you can send those special insert directly to database server unalter this feature is in do I do to use the block mode? is that a parameter to put in the config +apply your patchb +Note that you should not use those bits in production since its a snapshot of the development branch which has not been through QA +This would be an improvement on two levels the polled source may take an input Map the trigger Messages headers the trigger and poller will be decoupled so that other events at runtime may also trigger polling not just scheduled tasksThe basic idea would be Trigger ve done some prototyping but moving this to M to allow for further exploration +Ill see what I can do to get this patch and committed by next week and maybe start up a patch for deprecating logOnce etc +The dlwmin prefix for instance seems to indicate that the is being used +updated v to include new that just means ant got confused +thanks santhosh for contributing and alan for reviewing. +Moving the and into HDFS is not clean for other filesystem implementors +I tried apply to TRUNK and it fails +META +Please see should inject a different into the +I mean the that will be modified to support requests should be made to work in a way that ensures these invalid requests do not happen +bulk defer of items from to +This patch fixes some oddness with how the enabled setting worked and removes the test file that was addedWiki has been updated as well +Ok I can fetch a file off our ftp server but I still cant get a directorylisting to check to see if any files are there to download +What do you thinkFor splits IMHO this does not need to change +I am using the nightly from today +I am in need for this fix and would like to have a go on making a fixplugin for itI am normally working in Eclipse +Merging them seems like it makes sense +David already committed the fix in trunk on Feb +Thats great! Thanks for the help Gary! +please close this issue after applying the patch +Adrianhaving the URL in the property let us have different web pages for different locales +like the still refers to where it should refer to +Youre right +Added preclean and postclean flags to that overrides the clean flag when specified +that looks like a problem wit the maven embedder +This is on trunk +I found out how to create a JAR +Hmmm this patch causes test failures because doesnt handle null values +If it looks good I think we canshould drop the old mirror to avoid confusing people +So were the initial results you posted were with this hacking or were they with an unalteredJust trying to understand the impact of the changes +This should now be fixed in x can you try it +Server seems to have a common cause +Is it not possible for the meclipse plugin to add this builder automatically when you enable m on a project +No matter the profile is empty +Attachment has been added with description Patch has been added with description Another test has been added with description Result for the lastIs there an updated patch for version? I would like to use it if possible +Thanks for the feedback Suresh! I have updated the patch +Is this still an issue +Accepting for of. +Hi All What is the best way to work around this? The resource location is environment specific for me +I just committed this +Verified at. +Released +Looks good let us go over the changes in offline +I recognize Libors edge case however a failproof way to stop the second server is to check the Console view and terminate the process +Id want the new Client impl contributed to the project +We want to make sure the object after serializationdeserialization loop are still the same as it was before when comparing the explain extended results +Installs new for lifecycle test patch and added up the few files that werent added. +But I dont see the source for the doc in project +Hi DagThanks for the patch +Tested with jboss +SamHairongDhruba keeps failing +A copy that doesnt reference the original name of Breaking +One thought there is to base the choice of speculative candidates on the rate of increase of the relevant SystemFramework Counters in each task phase +Ported to branch at subversion revision +Thanks Thomas. +Ill look further but this may be more up Lucas alley to figure outYou can figure out what is going on by opening a UI browser window connected to localhost while the test is running +Same problem applies to +The branch is a good place to evolve it +you could read the code +Will add a simple test and commit +I think this issue is safe to be closed as out of date according to its age and the status of the linked issues. +It also includes specs +So it wont get a from not adding testsObviously the core and contrib tests are not related to this +This works for the indexer and clean tools not the deduplicator +may be you want to remove as affected version +of +Before giving you a feedback I read you lastest source code into Subversion +Also use the Apache version of legal +What we have is good enough for and this is irrelevant for anyway. +I still think that adding the actual source for small components makes more sense but I can see why that would be a hassle for larger components with changing the licenses and all +disables some tests in +Same patch with constants +the situation on gt is becoming untenable I cannot hold commits anymore its not only our playground other people do need it as well +You are right +Add preference for Description LongShort such that short will display rightmost Description characters from the last with hover showing entire description +The same problem occurs in line +Reopening for until it is done +It is only meant to be called from Derby code not from application code +Attaching a screenshot for inspiration how it could look like +The image logic or something related to layout of images appears to be broken in the trunk ATM as placing either test file into throws FOP into an infinite loop +Would it be possible to have run the tests so we can make sure to catch things that break it earlier +In our scenario the engine gets bootstrapped via spring using postgres as database +for global limits +A couple of general comments The indentation use of tabsspaces does not seem to match the surrounding code +WALVERSION is used to indicate compression type +Thanks Jrgen +Havent read through the whole discussion yet but it looks to me that the following will solve the issue What we need to ensure is the final is logged and flushed before changing the to SUCCEEDED +FYI after I worked around this problem as per the previous comments I then hit this after I worked around this problem and ignored I hit this problem Note that all three can use the attachment found in to replicate the bugs. +JonathanCan you address QA report Mar Therere outstanding review comments on review board +I renamed the variable to initial and I broke out a separate double random in +I can try it again and but Ive created patches this way before and theyve worked so Im not really sure +I was just thinking that while youre in this code you need to be aware that we shouldnt always do this. +Updated patch for use on SP its probably the same as the previous one but I thought Id at least give this one a bump and make it as painless as possible to have it integrated +I cut and pasted them and it worked just beautifully so maybe we need a look at the actual filesAlso I dont see where you are using the XSLT +These issues have been marked as resolved for some time now with no recent activity so I am moving them to the closed state. +bq +Ive added to the refguide explaining the current state of things +The fix for looks good too +Thanks Suresh for your suggestion. +Furthermore can you tell me which Eclipse Distribution you have downloadedThanks +verified on +The use of a native API is more a at this time +Resolving since is in +cvsignore etc and shouldnt be Im missing something +Marking as closed. +I thought it would be a common function for generic udf error prompt +r Vikas +done hotfix integrated into trunk. +Is there any specification for that font type available +I have merged this to. +To be merged into +Personally I hate that coz the code base just bloats when number of MR jobs increaseI have been trying to setup my idea using provided by mahout cwiki +So I am fine with the current way of ours +Created for publishing the h artifact to maven +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Certainly not undoable but the test cases alone are daunting +but now we have one coming with +Hmm there are a lot of changes between these versionDid you compare your profiling results to the old version? The invocation of cmp is probably a good hint but perhaps it was already a bottle neck in the old version +Fix attached as a patch +Could we add a wiki page with Repository Structure and describe the structure that is being used for the all the work you have been doing? Ill ask the others to update the page as well +So recovery would be generic +So effectively vlanidsvlanid and vlanidsvlanid would be valid if created under the same account +This test used to work on Foundation before it was converted to so I dont think it should be excluded +I will do this just to ensure that this failure is not a result of that +Hi JodyI would like your opinion on the patch for the issue +Thanks Corinne +ignore for now + Bug has been marked as a duplicate of this bug +unit test writes on dev list I have reviewed the patchand have done test on my litter luster all is good +Hudson builds will fail due to this so folks should refrain from marking thinngs Patch Available until this is resolved +Checked out trunk of compiled and replaced the cached classes both scp and scpexe work fine +Hairong Excellent +I have not tried it on Linux but it should be the same so I cannot se how it could work on any unix flavour +It use to work without cygwin in the path varI added it and it work yesBut I dont understand why it use to work before thats weird +Can maintain mapping from hash of logDirs toIf hash of logDirs isnt found in this mapping we createPrakashWhat do you think +Passing to hudson +Not sure what the relative value is but no time to code it right now unfortunately +The text format has an issue with XML structures +A few have been filed around this problem with little progress so I wanted to put forth a concrete proposal +HI Bernt +I think I agree as otherwise its just including the dependency falsely to ensure it is managed just need to check the original discussion +We are getting ever closer to a RC hereThank you great work Ferdy +My favor is ajax +Patch applied +subroto Are you working on fixing the above mentioned issues +serverdefaultdeploy reopening to update release note info. +Hi TatuWe created a standalone reproducerThe output I get isclass outside cdataCould you please review and let me know your thoughtsThanksKalyan +This will be part of a plugin +It will just cause additional debug statements to print to the +I suppressed the warnings since they are unavoidable +intepreted also yields some differences I believe +Its something like the followingAdd the new component in the following files parent srcmaindescriptors Build the kit by running mvn clean install from the directory +Committed +This patch makes several package privateIt also adds a deprecated and constructors that are compatible with +It is neither here nor one cannot use grails binding if theres a property that wont accept the empty string +Thanks Jason and Jie for patchWill resolve when JIRA settles +I have fixed. +For example there is a bug in which I want to patch Ive checked s source code and I want to change all of the POM versions to so that I can create a version which is equal to except for having my one change in it +Email exchange with Scott at +Sorry but I cant help any more hereSo far I havent looked into the C code at all and I think it would take me around a month to figure out what is going on +However Id like to find out whether there is a deeper issue somewhere in the instances lifecycle state handling +But the only issue here is that that destroyed volume is still showing up in the UI due to removed column is not set for this volume +m close to setup the quickstart in mvn for youthanks for your patiencem unable to reproduce the problem in a clean quickstart wicket app +on commit +At present migration off to Mockito looks easierOleg +Scatterbrained refers to ones nature a person who does things sense ordre ni concert +The fix version should be GACP +cant you define the default value in +Stefan Bunciak wonderful and thank you. +Saphen pls help make this patch against Geronimo also need patches for and trunk +Cocoon for example is often distributed with mutually incompatibleBatik and FOP jars +Then can be mutated in the same way and metric naming can be handled by static helpers only +revision . +I only wonder if the testDIRAPI test with the loop should be kept as the test takes seconds +thanks! +maybe some internal timeout is too short etc +applied here works for me +Fix on the web sentence in the Advertise description was already telling it cant workOf course at least one modadvertise must be in the to allow modcluster to find the right IP and port to give to the +Implemented rec block part +Well had some but not satisfied with itDont have enough to do this thing sorry +Heres a simple call graph from illustrating who calls how the calls to findVariable are recursive and how their invocations are passed through Hibernates putting great stress on Hibernate as a side effect +This was not well done Jacopo had to put a default ManagerI have a better idea +Attached is a patch to upgrade the component to use cometd +You can already get a pretty decent representation of a structure by serializing it with the +Hi Chris thanks for this ticket +Ill try a few things and get back to you +Am considering doing a gem release +Attach examples that reproduce the problem +I know there are still a lot of projects Ive been working on one the whole dayIMHO no new Cocoon project would be happy with FOP made me think he was starting a new Cocoon project +Pull request includes changes for both rules PR per feedback from rcernich to simplify listener scheme a bit on the details dialog +please adjust the priority to indicate that the Jira should be considered for CR +A file is missing from the version of the patch addresses most of the crash issues +This has been added. +How would this interact with the ability to define aGood question +Attaching the junit case that confirms the bug +in the title this template is just part of the bundle a core update will not solve this problem for existing installs +failed because of timeout +Resolved in by this as the library is adding false precision and we should be rounding its result +verified on Version +You could try posting to the Hibernate or forums to follow up with a wider audience +Please change line to match your IP address and line to reflect your node +I wonder if creating a new module for just the Execute class is really necessaryWhat would you think of moving this class inside the existing bundle +So we can add the API saying wether to start md and finish mds +Sushanth Alan Gates Olga Francis and I had a meeting over this at Yahoo +Resolved and closed. +Ignore my last sentence I misread the description this JIRA is about a diff between +Sorry to be a pain but I thought to silently remove an attached file after sending was not a good idea +Sounds like the best solution to me +Added a replacement test using +Thanks a lot for the quick reviews Todd and Andrew +I receive the same exception on my JMS client +committed to. +Heres the patch! Since we dont have a place in the repo for modules yet youll notice that everything in here is inside a modules directory +The paths to WAL files are collected in the queue of theI expect that it tolerates rather long suspension because the WAL rolling period would be relatively long +Attach to address concerns and comments by Pradeep +now can be removed from the exclude listGood luckLeo +The fix would probably resolve the sync send case as well +Heres a patch based on the Map Factory me what you passes on or its backward compatible and falls back on the impl for and no see no problem in pushing this into release +RickYou probably need to chase Alef down for the original diagram m almost done updating the digrams +Dominik as far as we know jbpm on the tooling side is working need to verify with recent jbpm console in still uses FORM based authentication so it does not the deployer URL is appupload whereas JBT still defaults to uploadFrom the comments its not clear to me what was the the distribution going to change auth method to BASICAre we going to sync the URL with +Sure Ill see if I can dumb down the src assemly descriptor to copy over the source tree I guess those are then +i also get mad with this problem +The to use is now a required parameter is now called The ceiling of FFFFFFF is now FFFFFFFF and tests were changed to accomodate thisNo changes are necessary on the branch so the v patch for is still current +I have changed the names of the classes I already had previously in an attempt to make more clear what they do +Can we leave this bug as FIXED please and create a new enhancement ticket for any newadditional requirementsThanks. +Fixed findbugs warnings +such that it is easy to keep track of all the different ways things get run over time +Verified in JBT build. +Thanks Rick John +Thanks for reviews lads +Is it practical to this? We like to get new unit tests for each nontrivial change to the Java code +I found a way to solve the problem that is I put in ejb archive +you should do something similar to what is doing but with setPrecision and need muchas gracias me sirvio de muchoThank you so much. +While on the topic of recompilation I think there might be an issue with views as well +Really sorry for the slow responce I forgot to add myself as a watcher to get notifications backThe server side is running an interactive shell +Mostly looks good + +Same with a +expected behavior of the WCM +And if one of the contributers is a himself who better to handle the issue? +hmmm +Done +Is this still happening in? +The PR is integrated in both master and. +The problem is Eclipse is pointing to but in the classpath there was entry for jdk +But if the is a pointer then it should be fineKelvinCan we have this reviewed and committed soon please +This applies to the attributes and childList fields as well +Its not allowed to use contextMenu with tabPanel in such way in the +Uploaded +Im on vacation now and probably wont have time until I get back +Added printing row count to output +This seems to be working correctly on my work machine +In the short to medium term I am fine with launching containers and letting them wait until there are enough to run the entire topology +By the way just want to made a correction to the but in that it wasnt the that got updated it was the that was getting updated + problem is that the servlet container should always throw an abort exception +Unfortunately many contributions come in which havent had any of this effort put it and it takes me many hours to clean up to get them to the required standardBob as I think developing this control will be a work in progress for a little while I would like to grant you SVN access to get this stuff committed +The tests themselves dont really need gig +Sometimes it is not easy software such as Apache Solr or other run on or AIX +Jon Yes null check I also dont like to putAlso what about +Im not a big fan of the temp file idea so Ive had a try at a different way of solving it in r +bq +Thank everyone +and r. +not like WAR deployer WAB extender does not rewrite the bundle manifest to add the caculated import packages +Agreed +I guess both versionless copy of all the schemas should be published as well since they are the latest ones +No in this testing I was using the ant script to undeploy + +The numbers published above are using +EugeneConsider using the formatter from +I tried to google this but was stymied by the Rowling fanboys seo noiseDoes Voldemort make the fat clients part of the gossip network or do the clients periodically refresh from the server nodes? Does Voldemort even use gossip +regarding did they cause this test failure +Before this under very low traffic the producer ended up waiting for a timeout proportional to number of events added to the queue the upper bound being the batch size + +The latter ones are supported via there own simple sm authentication protocol the former are a much simpler form + +If I understand correctly this patch move the snapshot step from election to after synchronization and before broadcast phase +Not likely IANA reserved I see connects from aa nevermindSorry forgot to mention our IPV adressedThanks for addingRegardsMike Hulsman +And when i log in JIRA today i found you have replied +And then we ran into this exception +I would be surprised if that allocation would be a significant RPC bottleneck but we should of course test that with a benchmark +All references to should now be removed. +That is what I am currently doing but I consider it a hack +We are still on hadoop and we have already lauched trunk into production +Could you tell me what should be the expected result for the two scenarios above? If the expected result is the same there is no need for me to by using production server +In Java jobs if it either reads input writes output or explicitly calls the Reporter object +Attaching patch with changes needed to allow AS to compile properly with Cache does not include the change to needed to use Cache and +Thread local is not perfect but surely better then static +Removing the attribute means that particular security domain will not cache authentication information. +If its trivial for you to these artifacts let me know so I can drop this repo and you can try again +committed the have merged this to. +Is it ok for you? Did you review the codeI had a look at the code and it does solve our problem +The plugin will be deployed as an independent webapp on the server +On a clean jvm shutdown this happens on restart +Did you reproduce this exception or not +Ok +Committed the patch to and trunk +VERIFIED +Thanks +A number of people at my workplace see the same thing +moved the postcode list test into Hibernate Validator +I will try to do small programexample to reproduce this problem +Will make it more messy than present timeOne alternative is that we can use Plan rather than Query and consequently be +Rob sar projects are properly deployed as standalone components but not when part of an EAR +derbyall testrun just finished with the proposed fix and it did not show any new failures +I verified by reverting that patch and checking that the test failed +Now I attach a correct one. +Hello AllI hope everyone is doing well! Havent seen any movement on this project for a few weeks! I wanted to check in and see the viability of adding this feature to the next release +doesnt use context class loader +Latest status +It was fixed with. +Unscheduling this from +Des what is indicated in the issue declares defaultTransformers as private and static +Committed. +B updates the database but A is unaware +Subscribers need to be vetted as either PMC members or ASF members so this MUST be a moderated list +The issue is probably due to not being prompted for proper password does something need to happen in order to that happen +Pig gridmix part will be addressed in a separate JiraA lot of common codes were factored out +Ive merged this to and trunk Ill leave it to Sylvains judgement as to whether or not this should be merged to as wellThanks. +spelling and validation is checked for birt and as and validation is checked for smooks drools ws and jmx and validation is checked for jsf ref jsf tutorial struts ref struts tutorial visual web tools portlet gsg converter and migration guides +However I dont see that logic in the patch +Fixed with svn commit to SQL component. +The same withWith the snapshot of I could not reproduce this issue no errors show in the server console or the log file +Resolved indirk +Hi yeah I think I fixed this already in current CVS +Stack trace of the Stefan for fixing this one so fast +requires this features to setup path environment when forking a test +The know issues section in the FAQ only contains things that we know about but cant or wont fix +Hi AdrianCouldnt this synchronization slow down things +This has to do with the way that grails generates its +Please note that also the code assist is wrong just like the highlighting if you try to invoke code assist at owner +My apologies +Attached a quickstart +Icon development should be separated into its own JIRA task +There might be a good reason for it but I fail to see what extra information names of column provides where type and position of columns should be sufficient +Thanks +Asha uses code very close to the snippet you gave +This was just to provide an example +Has anyone done that? Thats probably a good place to startHas anyone tried to see what happens when a JAR that has a manifest with attribute is put on the classpath of the RI? I think this would be informative. +Can you provide more information how the query was executed and what entity type the results were projected toAnd the resulting would be root node of the Neoj database has no type information a lot. +please provide testcase for what is actually failing for you +Ok I see yes I agree we should try to make this a friendly pluginNow I think we could simply create a separate descriptor xml for each component and then use the to just process changed files and regenerate the xml for thoseIm still not against going the annotation processor way for the future but right now I think its the harder way to goWDYT +Closing the bug assuming it was fixed when we updated to xalan. +I addeddefine LOGCXXLOCALEENCODINGUTF to the +Whats your suggestion here? remove the usage of keywords? or remove the property in special cases andor file types +Makes sense +Yes very much so. +Also it removes a test which had got in by mistake the previous patch +I can conform this +Standalone still has problem. +applied thanks +Putting the url param last is a simple defensive measure to insure that the url param is properly decoded +Pete can you look at the issue? Empty string goes into validation of email this issue repaired with? +we might as well handcode all the soap messages and claim maximum perf Ajith If you remeber we are doing the first tentitive data binding before M we take lot of effort to make sure OM is not build from the Objects if the is not accsessed by Handlers +Patch for trunk where I introduced with default value of secondsIf provides API that returns recovery status for given Path we can utilize that to reduce the number of times we request lease recovery +We should consider this patch to be included in Pig +Hi KevinI think you missed my point +Agreed. + Add support for Ideographic Space to the revision Thanks Luis. +I added nullpointer check after every usage of that method +There may be a more clever way rather then enforcing this in the GUIbasically when someone tries to save a package config or check in an asset or create a snapshot etc if the packageasset being dealt with is already a snapshot then we throw back a with the message in it of hey numbnuts you cant change this one! means less work and its validated at the back for historical assets they are already read only so no work to do there +Webtools didnt even give you the option to force a full publish via its API +Could you please specify which version is affected +The whole c +Please specify exactly what has to change in the example code to place Asynchronous correctly +Not sure if I understand what you mean but I think the only thing that would happen is that you cannot run the Derby client code with a VM at level assuming you want to produce Java class filesMaybe you get a better answer if you explain which value you would use instead of and which DalvikAndroid related issue you are trying to addressThanks +Sorry that was ambiguous +Here is the modified patch +Blocker for SOA note +This is necessary for the second part of +Problem fix asserting that tree adds the new item to the dirty since there is no parent node at client to be recreatedProblem fix asserting that tree dont add an not presented root node to the AJAX response by invalidating it +Thanks for your patience +Soon as I saw the diff colorized it was an apparent conflict +This should take care of the findbugs and test issue +To build it use trunk by opening which uses an older version of this dll or build it manually +I am ok if you think the unit test can be tracked by a different JIRA +Sending manualTransmitting file dataCommitted revision +It sounds like this issue was already fixed from the forum and is actually an AS issue +With this parameter added this plugin could replace the jarsign and jar I am planning to put this plugin into the release profile of some parent poms soon to be released what steps does it take for this plugin to get released +The document should transition to an error state given that the target db is missing or to a success state if created with createtargettrueThe replication is not cancelled or halted so that would be inappropriate tooWhat if someone deletes the target db specifically to get it repopulated from scratch via the replication job? Its unreasonable to conflate the two things. +The patch doesnt incorporate the latter comments as above is this something which would be required +You can reach me by email or mobile +So all issues not directly assigned to an AS release are being closed +The attached file is a patch +No work done for a long time no response to change anything +Can you run adb logcat while you are reproducing the problem? I want to see the stack trace to figure out what line is causing the problem +It should be well documented in the code that it is important that your value objects implement equals method correctly otherwise cleansing code will not work +Hi WaiNot sure when exactly but should be soon +The patch depends on some keytabs which I will upload separately because they are binary files +unit Vladimir the bug was fixed in AUTH module at rPlease check that the fix fully resolves your problem +A quick google search for write security finds many people with similar problems +Thank you for your review +the expected result is that the process is ended normally and no async continuation should be scheduled +Closed as wont fix because of introduction of email notifications in Sonar. +Hi EricQuestion on the patch +The trouble is that the groovy compiler does not support the idea of multiple locations and it seems to me that to break up the groovy scripts into different compilation units based on their output location would just open up a nice dependency can of worms +Also logged to improve document generation performance. +HiI got the same issue on Tomcat with Jencks when using +Can you please give it a try and test that it fixes this issue for you +Im wondering if there is something going on correlated to the DDL statements executedby the repro script which sometimes results in tables that return reasonableestimates and sometimes results in tables that return infinite estimates +I do not have a Linux VM +I ran into the same issue +The same runtime errorwarning messages also occure on version +Others Agreeing with Suresh +Patch committed. +I was running mvn clean install +This proposal like ISO itself relies on the geometry objects +er + +Does it render ok on your server when you have a single instance +Ok I raised this issue because I thought that it was causing the other tests not to run but now I realized that it is causing no problem +I didnt want to introduce a way to inadvertently void a bunch of items etcLet me know what you think +The original reported problems have been addressed and we have separate reports for others +This issue duplicates please see the workaround in that as we have to wait for sitemesh +Re if there is any ambiguity then its clearly wrong already +occasionally fails +Youll need a to +I am attaching the initial proposal of the way to do thisI would love to hear your thoughts on this document I am currently experimenting with the design described. +Im updating the component and the summary to reflect the fact that this isnt a Convention plugin issue its a more general Struts feature +Your checkout seems to be a few weeks old as the patch no longer applies cleanly to the latest trunkAlso once I resolved the merge failures I get a number of configuration warnings about duplicate glob patterns and some of unit test failures due to changed MIME type settingsCan you check those issues? Otherwise the patch looks greatPS +I also like the changes in hdfsproxy +vfszip redeployment is OK so Im closing this case the problem is still present in ASdeployers reopen the issue. +Actually the problem is doesnt work with found exception here cannot be cast to One more question Why we have to care about the number of constructor of +TrygveDoes maven really understand that the wsdl artifact will be built by the API project? The wsdl is attached to the build by the plugin code of javvaws +Can you try again with the latest snapshot release and let me know if your problem remains +The offset feature will probably be implemented later +Youre right +I appreciate your comments and tests! Keep going on with your complaints +A new module has been imported +r and deployed after release +Assigned to release +Maybe ask on the geoserver developer list too +Yes combinations of things sometimes have rough edges +Committed this with so marking this as duplicate. +This is why I suggested the use of MXJ +Ive applied the patch with version Thanks to Adam for the contribution +Claudius Im finally working on your patch but the class is missing +Workaround for the automatized test switch to tab Servers and invoke the New server definition wizard +I fixed the nd issue under in the st issue under r +Did you run the following testSeems like it would fail +Still fails two test cases +Seems likely +Runping did you see any timeouts? Or was that all of the reduces logsI assume you are using theIm marking this as a blocker till we understand what is happening +that is the beauty of using bookeeperbookkeeper has a ledgerid +Submitted fix as svn based upon patch from Mridul +Attaching patch +I have not added method or class coloring but couldIf I understand correctly the actual missing features here are class and method coloring +Ted Yeah I took a look at BAOS +Gordonwhy do you think that the trace calls pose a significant overhead onDid you make any specific measurements? Please post figures +Not being mentioned in the spec doesnt automatically mean it shouldnt be done +Is this still an issue? Im going to move this to unless I get a patch by tomorrow morningColm +This testing is done for both as part of this Jira and other testing performed by ips joseph and others. +I think is too big a feature add to do for deferring it to +Once depends on we should remove the class ISO and use the one from the commons module +Whats the log like? We were working on similar issues recently +Temporarily assigning to Peter as this is more of a Bindex OBR spec question +I believe it should require running both Mesos and HDFS +Thanks +Its again IO errors +If there continues to be a problem with this issue please reopen +If they were a properties file it would have been extremely simple to read them +We combined it with and forgot about the original issue +or actually a simple AND on a preexisting query depending on where the faceted data searches are not really relevant to Hibernate search bu their strictest definition +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +LenCould you replace your plugin with the attached plugin in JBDS and try to reproduce the issue again +I tried to get Dr Porter to commit those changes for years but its still the same +This is in progress +You say compile but Im wondering if you really meant link +reopen for setting proper fix proper fix version. +Tested +The purpose of this attribute is not to rewrite any passwords but to allow the use of digested passwords in place of clear text passwords +repopen for backport +Have a few minor questionscomments +Attaching a patch that fixes the root issue +Tested successfully on AS +I dont see a way in gettin this doneonly way that I see here it to pass the location of the as an argument to the fileMahadev said he will be looking into this +It is optional though +I am trying to puzzle through what we should do here +If a downstream projectagreed +Do we need throw exceptin in api addPeer +This way I only need to manage version changes at my root pom rather then on my multiple the formatting I used the codehause eclipse formatter but perhaps it didnt import correctly +Here is the patch which makes Port shared libraryThe Port library also provides APR and logcxx functionality to other componentsThis patch is checked on Linux only for now but AFAIK Linux is now more tough on linkage +I had a quick look at the test case +synchronized I think has been used for a simple implementation of LRU policy +A problem was discovered in which is a showstopper so Im going to drop it and prepare a new release dropped staring with dropped is around the corner +An external tool would be fine does one exist that fixes precisely this problem +SDO Type Integer maps to not +I would suggest that this editor works like the current one + to commitworks for me +the patch looks fine +Ill try to get to this again in near future +Thanks again. +Closing xml and console issues. +If it turns into a concrete set of requirements we can engage with then great but at the moment the discussion seems to be discussion +Patch that applies cleanly to branch +This was fixed by Ant +Attaching a patch that solves the problem that is caused by the lack of name resolution +Except for test the other test failures were due to Too many open filesI ran test a few times and didnt encounter test failure +However I dont think that its appropriate to expect a first time contributor to make all of those changes before adding a smaller fix +Thanks Chris ill follow up on the way DNS and mail has been set up to easily follow the transition from podling to tlp this bug may hit sooner than you think. +Not sure how implementing N different resolveXXX methods would be easier on other EE container implementations +Any suggestions on how I can use the script to get rid of this or alternative ways to make the output from this test match the master file +thanks! +Now this is the correct one +This version already seems to be working with maven but maven requires still the MSHARED issue to be fixed in one way or another +New patch with fix for issues found after more tests +I tried setting up a simple test case in which case I cant reproduce the problem it only happens in our production codeThe class tested has collections +Thanks +yes I am aware that a normal trace will contain only the shown part many times +I close the fix as resolved if it didnt do the job please reopen +I think as dhruba said it really is a blocker since it is a bad bug when opening a file in append modethanks pete +This is most likely a Dojo issue youll have more luck talking to the Dojo folks but Im guessing support will be a little thin +Created a dev discussion forum post +Can you see if you can reproduce this? I am not sure what could have done this although there have been refactorings that could have straightened out things +Ill upload a v patch +For a simple usecase like ours we dont want to depend on additional layersIt is a much simpler change to support in hive rather than fixing a lot more inEventually if moves into Hive these should be merged +Jaroslav can you tell me if this worked with JBDS against EAP ? Or is this a new test +I guess Im just unable to properly reproduce it outside of my live app. +Looks like I misread Chris credit line +I wonder why is not thrown in MR task + +The failed tests are all part of the same target as so even if they startstop their own server will run first and mess up the environment +Thats in there for supporting lower level functionality but thats not the typical pattern used in as we do use strings for sequencing +There are only arrowrighttophalf and arrowhookleft left for which I cant find any unicode mapping +It looks to be behavior +Some more work is neededDont bother to review for now +Reviewed all these issues and marking them as closed. +Duplicate of the issue +Amar this is because true true is not set +Curious how did you come up with the parser? Did you copy and modify it from LucenesYep and modified the grammar to slurp a localParams block as one big tokenI should be able to further simplify some of the parser by cutting out the cruft Solr never needed since it drives from the schema. +I guess you also need to update test which just runs on Hadoop as well +I wont have issues if the team settles on by default I just cant make that decision given that Ive not been actively involvedThanks reasonable +Please could you provide a description of the steps taken to reproduce the issue or the configuration options that lead to it +In my case I wrote a custom filter fucked around with the configuration getting it to work which means you really have to get into it +Adds checks for interrupted in getFile +We can consider this for a future + +The Weld team seem to have improved their deployment code hence marking it as enhancement +Attaching a new patch with slight modifications to the test so that all the information present in scheduler information is tested. +for the hotfix patch +Yes I think Ive also seen that name defaulting for producer methods is broken +With some trepidation I just committed thisThere were a number of problems with this patch +Folks could not upgrade to a version of Avro that includes this change without also installing +However it would be good if you could add some regressiontests for this issue to the replication test suiteBefore committing this patch I want to get one question clarifiedIn stop if stopMaster fails you will still writeReplication master role stopped for database to +Jesse you once told me you have a WIP for a new shapefile store +classpath files and tell people that they should use the eclipse ui to build itYes thats correct +Currently only ever works reliably for final classes +Setting the default scan wait only thing that is configurable is the scan rate and that doesnt seem like something that really needs to change +No more deadlock so far +address a string length logic for the quick response and update +Merged into the master branch. +Once youre ready for a review click Submit Patch +No error appears in the consoleIf you are still seeing this error please post how you are loading the composites or give the test case that shows the error. +bq +Here is the patch for given functionalitySteps for testing Go to Purchase Order screen +Dont you think this is very confusing for contributers? The env should compile when we check it out thats a minimum requirementMaybe we can add the instructions for manually fixing this to the installation wiki +ssh directly instead of calling java to create directory +Flavio thanks for editing the documents +Second report on the same day +I am seeing the issue on Wink R with Java and on Windows XP with Java but NOT on RHEL with Java +Reinstalled Fedora again from scratch same errorInstalled the following packagesyum y install yum y install libxml yum y install libxslt yum y install sqlite yum y install ruby yum y install git +Jerome has nothing to do with rules its about importing source code +Confirmed Brett +Tammo Im not sure I see the difference other than the line being moved upwardDo you want to reattach a full patch? Or point to a github branchAlso how do you trigger the circular dependency? By running the +Here a patch that fixes this issue +I just committed this +But after your changes in only the allocating thread should be releasing the connection right? To catch this one would need a second consistency check after the lock is obtainedAnother more simple explanation is that the route stored in the connection is incorrect at release timecheers Roland +Needs picking to blockers and critical issues are resolved but not verified +This is not justifiable at this point and probably wont be justifiable for years to come since we need to remain compatible with Java EE runtime environments where only JPA is present and where the JPA API cant be updated without a full server doesnt seem to actually call those new methods or only calls them within a catch block in order to be compatible with JPA +I had incorrectly mentioned that VM was trying to load Factory +OK I can reproduce here as well with the ASUS Camera +How this ticket is different thanI just closed with description where is the problem in the users code +yeah the patch looks good +old bugs. +Looks good in +This is fine for running it on the system on which it was built but not on other systemsThe util file is included in the distribution +IMHO it is ok to throw NPE when one of parameters is null +The Reresolve References action has no implementation behind it so itll be removed from the Metadata Add To Metadata Favorites Open In Relationship Navigator actions really shouldnt be there either so removing themAlso moving those actions out of Eclipses Extended actions into the Modeling XXXXXX submenu to clean up the context menu on selection +The new logical plan must handle implicit wrapping tuples differently than used to be the case + resolved issues for already released to comment on a closed post +Id hate for an implementation change on the JR side to cause this to break again +Allowing custom templates with markers would allow any kind of downstream system to be used for bootstrapping from simple scripts to classworlds to perl scripts that can input an XML file and invoke jvm via exec bypassing command line length restrictions +I dont know the dataflow to that changed script but I felt forced to ask if its ok to use there as I could quote more than one problem involving this practiceIs it? + +It doesnt look like this Issue will be doable before the cutoff +Are you sure your application properties file is named the same as the application itself? I have one site in particular that has only an applciation property file with no page properties that works correctlyAny more information on your specific circumstance would be helpful so it can be verified and tested +Heres another patch to allow JAVAHOME to be set properly on OS X even in +So all issues not directly assigned to an AS release are being closed +This doesnt look like a bug to me forward references are forbidden in initialization blocks +It is also said that for statements executed only once one should not use a prepared statements +Tests have been enhanced so that if a failure is received the error message is printed in the failure message giving more clues about the error. +Patch file with full example showing the usage of the Webservice Provider with the client using Dispatch API +Martin would you be interested in testing the fix by building from source? I can point you to the branch +Bump to chance of reviving this patch Bob D? +trunksystestssrctestjavaorgapachecxfsystesthttpresources +Just wondering is there anything we on the Spark side need to do for this? It would be great to have the web space up so we can begin migrating our old site +Removing Patch available status to incorporate feedback and refactor the change into a static method in +If the endpoint doesnt have access to the Token itself I cant do that +SuneelWhat do you mean by on the second iterationStreaming is a single pass algorithm +roberti think the patched ISM Locking Class has nothing to do with the stale item exception you get nowI have also found a bug in the with Transactions see Maybe you test the attached patch and share your experiencesthanks have implemented a extension to the current Locking Strategy that binds the lock to thread +Yes +But when i invoke my rest url the path variable i get is still removing the +nevermind +Hey Deepak Id be surprised if this wasnt working its one of the things that Bigtop runs to verify Crunch works with their release +Thanks for reporting +Take for instance We are using structured tokens that have claims and metadata built into the tokens itself +The other three test cases show the problemI didnt know if it is Ok to change the bug subject so I left it as is +For x installation the user should disable the repo +Could somebody please have a look? I currently lack a to try that +Builds build with bytecode +not done for likely I can look at this anytime soon +Closed after release. +Could you provide more information please if possible with a link how other systems hash database passwords? If its a cryptographic one way hash how should jackrabbit the password to connect to the database +Thanks +Actually I dont care how the syntax is +moving to v +Yeah I agree with more descriptive +I thinkOKI think it could be hierarchical wo so much work ie on reopen as itwalks the terms it should be able to easily build up the parentchildarrays since the terms are in sorted order +I dont believe our jars even ship with the javadoc included +It was mostly authored by NicholasMore rename tests to follow soon +waiting on JBT build to test and Delete project does not throw NPE +stack +Feel free to switch reschedule if you feel appropriate. +I also changed the get method to return a instead of taking a object and adding the namedlist to it +Fixed with svn revision I added a JMX property so it can be viewed in as +HiThe only way to post the files seems to be to post them each as comments +Ive fixed it by adding mentioned dependencies to Augment of development classpath because they dont really needed at +Do you know what browser and version is used by eclipse in jboss tools +Mikhail the is not really in JIRA it reports HTTP error +The value type for is added to the +Our site is still hosted on me know if you see this as a +Have you been able to reproduce this case? Its possible we have a hashing issue and are weighing these two methods equally but any sort of reproduction would really help us fix it quickly +I see trying to make this a bit faster doesnt make sense at allSorry but I am confused about the epoch handling in the initial negotiation between the Follower and the LeaderThe current FOLLOWERINFO seems to contain already the acceptedEpoch isnt it already possible for the Leader to check that value and reject the connection if it is wrong +Looks like it could be a bug as long as youre using correctly +this would be fixed by persistence of cookie it has been done byFor having a unique url for each languagepage +trying hudson again +He did not respond probably since I said I was going to put this in. +While this patch does add the ability to have multiple jars it does not change any script options +Yes I do have a test case +Closing the issue now that it has been resolved. +Uploaded patch which implements the proposed changes +I think this is a good point to leave it at the moment +Feel free to adjust if you disagree +I know Stale Pedersen has commented that he thinks it would be better if we required escaping in AS +Concerning read speed of map applications a calculation comes up with an estimate of MBs per map application +The problem I see is that in case ip based delay is configured it can happen that we use the crawl delay of one host for a other host running on the same ipFeedback is welcome +I can add more excludes for the components listed there but that may erase the benefit of shading this stuff in the first place +I dont think we need to worry about this an request for a location will override the previous one thus an invalid mix cannot happenHow about this case +The patch looks fine to me +Other than that it has no programmatic uses +couple of candidates +Working fine +Yeah codewise I agree the patch looks nice +Thanks! +I guess the main problem with the example I attached is that the plugin only runs on the toplevel pom which makes the property only available in the toplevel pom is in fact different behaviour from declaring properties in the toplevel pom the properties that are declared in the toplevel pom are propagated down to all child projects +Ive attached a second patch which has the Serialization abstraction and the object reuse facility for Deserializers that Doug suggested +Applied + Sandy +cleaned up addedit repositories screen should change the label from Indexed to Scanned too +This wont happen because the stop method intterupts handlers and listener +I tried the snapshot but wasnt able to make things work +In short what this JIRA introduces without any exposure on theSounds good +Even if we have one Id rather prefer turning off this feature completely in as without and related issues this issue is practically useless +It was applied in and +Wouldnt it be better to make plugable with a separate issue prior to the introspector stuff? Just to make patches smaller an easier to understand +Does it really need to throw an exceptionAside from back compat I dont see why it would need to +duplicate to existing I want to keep the visualization part separate from the actual management and will use your issue for the former. +Bagsie not think that if we are adjusting the prefetch we should make it +Does it make sense +PhilippeDavid has commited changes at r +It works with complete disregard of any text encoding considerations +Sorry for clarification purposes my comment applies to the mentioned in Ericks comment and not the problem stated in the original bug report +All tests pass which failed before is reliable now +Attached the patch as sub task. +Create tests manualautomatic for supported servers +I will wait for a day or so before committing it to +Id like to know too what the release schedule is for +Code gets dirty and a is at least recommended if not required +Im not trying to use works for me argument but I cant reproduce this so logs would be really helpful +So we dont need to limit positions to +Im ok with this +Fix NPE add unit test +Thanks for the quick fix Kim +Im looking into this +fixed in my original intention was to add these files to the source tree for metro so we could have the metada files for all the possible configurations +This has been fixed for PL +Could someone please post an integration guideas you said returns but or returns fromI set the requestHandler to but the queryComponent still return theCan you give me some advicesThanks +Will be back on Monday morning +Thanks for reporting it I will also need to add verification to the unit test suite. +Patch is attached +Notepad never lies +will add a comment that tells how to duplicate the problem. +I forgot to push the change up to the repo +Feel free to enter jiras if you notice a problem. +I think there is a problem +I think using directly and getting rid of would probably be bestAgreed opened for tell me youre at least seeing this less often than with +Think Ive finally got it +push out to +Please to current trunk and Ill commit +A lot of the imports were changed from explicit single class references to wildcard imports +seems the grails script generate the test class extending from and not would that be classed a bug or is desired feature +Part of the patch wires into the Groovy ant build +The easiest fix would probably be to either replicate from the old version to a newer version or go through and purge and reinsert each document +Hi Christian you make some suggestions here but I wonder if you can provide a patch to fix this issue? It would be greatly appreciated +Is this a cleaner fix than the local variable +This incorporates an updated version of allowing the return of partial search results +I would find it unlikely on that any of the threads were CPU starved with only msgs +Set correct fix version +Having the same issue any workarounds +If Nitay follows that example this stuff should work the same way +Marking closed. +Done +I am attaching a patch which adds files for gora and all its modules +Added a servlet to tabulate queue stats +pushing to to addressed but I wonder if multiplication or other operations are also affectedWell have to find some workaround for . +done for now +rUbuntu smoke kernel +I suggest that this change be made to that file as well and will post a patch shortly for that +I agree David +Did you have a particular question about the git mirror? Im not aware of any issues with it and anecdotally it appears to be what is most commonly developed against +Thanks for the review and completing the patch Andrea! +Everything looks like localhost to a minicluster +The issue from above is simply a case of using Java features when we should still be sticking with whats included in Java +Just to be clear +Revised batch a bit +Moved to not fixed the problem +This separation is a hang over and needs refactoring so that the verbosity is held only in the instance +I have added number of partitions and the minmax vertices and edges count metrics but I am not sure on how to make them showI mean should I add something into the class? or in some howAny suggestions are highly appreciated +attached patch adds one more option to Maven Integration to control GWT how does this play with GPEs own configuratorHow does it know what to set +Thats great cheers! +This is about selecting tests for execution when you have multiple surefire executions with different JVMruntime requirements +might take a hour or two before the cvs commit is visible in the public cvs +HiIve just tested on Fedora and I had the same issuesThe fix is the samerm rf +Both the thrift interface and the bloom filter could be extracted and compiled under +Approximate Counts and Quantiles over Sliding Windows is more desirable for its ability to do arbitrary percentiles for sliding time windows +I have not seen this issue in a while. +The way it will work is that that the remote repo will contain the version info within the artifact file name but upon resolution will remove the version and place it into the local repo +push to +Couldnt we integrate the journal querying approach which Marcel proposed with thehash property approach? The Microkernel could generate that property on the fly querying the journal internally +javadocs too +Hi PeterI agree that the tests should follow a consistent naming convention +Checking this in +bq +Committed additional patch to trunk and branch +for the change +Im not in a hurry so I can wait on this one its not fun to constantly work on the configuration +closing as cleanup +ask away +Thanks Elliot for pointing this +Okay can reproduce this reliably in Fuse ESB just have to install feature and restart the happens every time +That way rather than recalculating the entire for the node we only have to add or remove a subtreeEDIT Ignore this comment using midpoint to select tokens for Nodes isnt really helpful at all +That was a typo +Added back in the code added in with a patch to get the parameter from the parameter map instead of the is just the function is the patch against trunk I did not get the time to review and test properly yet +Seems to be the same bug +So they want us to rename table A in cluster Y to be and then copy table A from cluster X into cluster Y +Brian if we dont need to create a new xml file site so just over copy the two files to the directory above but max said we should create a new one to avoid the overriding +The reference implementation is with fonts properties +yes i was running on windows +Thanks Alex the patch works for me +Improved patch. +This copies and into the esbs lib next to and +I committed this +At this point all you can do is to vote on it +Would you mind to post the type system pieces so it is easier to reproduce the error +Good catch MarcoThis is a type that somehow escaped my automated renaming process from Java convention to CThis is now fixed in ver and +thx for the email +Revision +due security bug we will upgrade to +OKOur upgrade was successful when run nohup d u cloud p password c r tail f How about your upgrade +Fix to by utilizing cURL to to support changes to +Thanks for resolving this so quickly +and are flaky and pass on local +Thanks Jitendra! +Test m reopening this issue as its basically incomplete +Attached patch should fix this +Thanks for the report Tom +I dont think it that basic +You dont have to use the local build +Hi MyrnaI think this is a big improvement + Adapter property public +The tables of attributes are checked for the release. +This seems to be a specific integration issue with in Axis? We need to have a discussion with the Axis community about this +It is effectively putting a magnifying glass on the area where the change has been made +If this came up out of flushRegion call then we add to the queue you just addedGood stuff +Sorry for the trouble. +Ill work on this +Sorry Im still confused +While is a wee bit larger it is much more flexible than Modello is and has better error reporting so we will continue to use it for out XML inout muck. +Verified by Mark +Also its not the only case there are plenty of issues pending +I just tried installing node via the linked to instructions on a ubuntu desktop image worked fine installed nodejsUnfortunately this did not also install npm which the and install from nodejs Mac package on nodejs site do +IMO we should always return empty child objects on the last level +Im working on the cutting module of this project and want to commit +Updated the issue title and the description +Is jira compatible with firefox? I get nothing when I hit to me on the assign use firefox as well +bq +Thanks for confirming its not issue Jean Andre. +Closing issues +If we are unable to get to support this we can always make a gem which exposes an API which just does what MRI would do today and we would call our enhanced cheap syntax +Hi AllI just attached a patch to add this link to the in the News section of the website +Verified in Fedora +We decided to support it for SELECT through comments so that we do not violate any SQL standards thus keeping our SQL syntax compatible with other database products +Committed to trunk +I will add a patch as soon as possible +Yes indeed good catch I did copy it to the wrong location I was testing with a JVM lately I m going to undo and copy it totrunkjavatestingorgapachederbyTestingfunctionTestsmasterjdkassuming I dont also have to copy it under the jdk master directory but am guessing not or it would be there already change to be in the next changes revision package +I too require this for +New patch against trunkAdded support for commitWithin as request parameter on deletes for XML and JSON loaders +Ill put it on the list to fix +The attached patch adds background thread which pings the parent java mapreduce task every seconds to check if it is alive +Closing as per previous comment. +But my current code just works with MRI FFI and FFI on OS X and Linux +I am resolving as invalid per the response +This needs to be either fixed by or delayed to since we want to release the last rc unchanged we cannot schedule issues against this does not seem to be trivial depending on unfixed gt issues +Thanks Bobby +Thats good less processing for configuring your project +Many of the tests failed with the c patch +But I did not follow your change for the default constructor because I wanted to preserve the behaviour we had before and that was that the default constructor asks some properties but not all and uses defaults instead +Patch applied +The user expects certain semantics from rm +Merged this to. +In addition weshould use a consistently. +has just been released +Since some of the commands return results users probably expect something or an operation to be complete when it returns +I would care as well cause many plugins are there for metrics and quality so should be prevented. +Im going to attach his +I thought i had reduced the size to but the node it was testing on had a size of configured and redistributed the config over the cluster via config bootstrapThis still leaves the problem that Solr itself should run out of memory and not the OS as the cache is part of the heap +i am on this since it makes the build to run run test twice +changed to blocker since its blocking the upgrade to Hibernate ORM idea how long do we have to wait for this? has already been resolved +we are not going to put time and effort into what may very well turn out to be bugs in chrome +Note that the Data Bindable field will be removed from the table +if we narrowed the usecase we should note that in the migration guide if we havent already +Thanks Kevin. +updating the patch to trunk +Fix +After a bit more research and testing I was able to create a patch for the revision +But looks like that I accidentally deleted it when I was managing the attachments +This dialog has been removed. +for putting this in it doesnt break anything and I have need of it +Go ahead commit it +Thanks Anil +If someone has optional resource refs perhaps they can look them up in the global jca context where they are all now bound +We are specifying conf which does not work but also putting that directory on the classpath which does +Committed to trunk and branch +XML doesnt mandate we encode all characters into some encoded entitiesXML payloads can very well use characters from any unicode character plans +There is another defect which covers the same topic +These warnings do not appear when building the community branches +Seeing that something like Email was put in Im kind of wondering why Url was not put in at the same time +Actually as this needs to be included into maven via canI request that be deployed with this change +See big issue that Ive been hacking around with is the problem that objects have a type and a name but this does not necessarily correspond with what is going on in the text +Ill commit this shortly +Removed unused variables from Please file the to track the metrics aggregation and the config naming convention +Nor have I written any documentation on this yet. +Ive tightened it up so that create performs the same validation on injection points that a bean gets +As mentioned in the previous comment if its getting closer to the release and we dont have enough confidence in the new code it can be disabled by commenting out four lines in +I have the same problem +Usage of in +bugfix for get inand has been added with description fix two wrong in svn +Verified in DR +Now I need to generate Like I make it? because the command ANT generate but not Best Regard download the source and also the binary The environment variable XERCESHOME set to the folder The command ANT generate the library but I need to generate the library This is because the client that I made in AXIS C takes a long time seconds in Windows NT but in Windows XP is quick The only DLL that not this generated in Windows NT it is it will be this the reason of the because it is slow in Windows NT +The found concepts are shown differently than the concepts chosendoes this mean that you want chosen concepts to be displayed in identical manner to available concepts? I cannot remove concepts I just added I can only remove after I savedI confirm Ill fix it +Thanks Vinod +We can reopen it and close as wont fix if community votes for moving pack back the archive module. +Now roles template and preparer can be overridden in a consistent mannerRefactored and to hold template preparer and roles +So where is the page +Committed to branch too +Ping! +ssh root root anthony return exit bs when you click into a field! Argh +but I still have to perf tes +fixed on +Polished code a bit in rev +what if you set JAVAHOME to the jre? +When the name node receives block reports containing blocks from the previous file system it throws the exceptionTo avoid this we can either make the Block class public or explicitly load the class in +Patch attached and applied + +This should be strongly considered and discussed for SOA +Once its there it stays +The indentations etc +If we cant do this with junit maybe we should investigate testng +Takuya Sorry about that +Just downloaded the source from the activemq site and it seems to work fine when I build itBtw make sure to do the maven build inside the src seems to be there now +David +Patch using approach +one will be startedloaded + +Show an example of how you would use reason it is would you code without it +revised some sentences according to kans comments +However it seems like a good idea to keep all ZK specific logic bundled inside +All tests for and passI am testing in SVN revision . +So I think we should do the timeout processing +The real issue here is startup time +The quickest and most immediate fix is to just add the jars to designer +We tracked it down to a JNI issue if when using libhdfs the thread that started the JVM isnt the same thread that first uses libhdfs it will fail with the error described in the JIRA +Some tests in trunksolr are not passing but my putty scrollback wasnt large enough to see what went wrong +Sending content CTrinidadTrunksrcsitexdocdevguide Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadinternalstylexmlparse Completed At revision +Patch v added configuration to control the size of max number of prefetcher threads and max prefetched result size +This is probably the correct behaviour needs documenting be better to have this fail with a meaningful error we dont know our own hostname please fix your underlying system configuration +trunk + +Sean all are correct +I think that EXPRESSIONS field should contain a single value for range and hash +The test and associated files in the coreWebminiTeststiles test directory have been updated to include an example for this issue +Fixed a findbugs warning and attached the new patch +RTC defect all remaining issues so I can release in JIRA +Olivier Gaudin said that it will be available in near future +Please let me know if theres anything I could do to assist you with the migration +so attaching patch for and closing. +Does this also work for downloading from repositories? Or is it only for uploading +as part of +However its pretty rare that this happens and it has no side effect to the user. +I agree that sometimes this setting isnt needed but I am not seeing how this would grow into multiple sets of checking rules +Is it the expected functionality +METARegards the calculations I searched in the whole source and it doesnt seem to be used only referred to in the text so I think I will just remove it +I only used the releaseThanksSumit +Maybe the underlying Reader should only get Output no filenames so the delegation can completely done in the directory +If it finds a lesser version it fails with return code and with proper message +I have the same problem +on committing this as it is. +If nobody complainsloudly Ill commit the patch tomorrow around GMT committed +Sorry for the broken description but someone pressed my enter key Here are some additional infosThe checkbox Prompt for Workspace on startup is checked in the preferences +Im getting FNF on indexes on my cluster +I can committe this patch and it works fine +Here are my Producer and Consumer JMS examples so you dont have to create them yourself +A simple rule could put a Defaults bean inthe digester stack and other rules which could peek back through the digesterstack looking for the instance +Here is a program to generate files with different sizes +branch and on trunk. +The problem is that some of the Enunciate mappers and perhaps other code make assumptions about the format of those GWT classes +yeah you might as well +I am so far unable to duplicate with the release candidate +Thank you +Good faith reverts are not a problem and can easily be reverted themselves +Yeah not surpised Dave Blasby created the relative reference trick for many years ago and the only input to care for was an XML file +Would this be relatively easy to implement +Feel free to close but this does limit functionality +bulk close of resolved issues from releases prior to +For speed reasons there are no checks in serializers +My apologies and thank you for your patience +The +IMO for applications which do not respect the limits instead of reducing their defined requirements to the max value we should throw an error as we are not sure if the app really needs that high amount of resources and whether it will actually if we reduce that amount to the max valueDoes that make sense +take snapshot of volume vscenario +Please do not forget. +Can somebody please give here a feedback if this possible to be fixed +I get the same error message I tried replacing it while the app was running then restarted the app then did grails clean and started the app +my real problem is that whomever created these lists forgot to setup the main mbox archives which feed eyebrowse and the rest of our archiving systemWe need to find out how so that we can fix this issue structurally +has been releaseClosing all related resolved issues. +If this is not right I agree with Mike +This patch is just a start +Do you have a full stack trace Ricardo that you could attach to the issue? I gather the underlying problem is fixed in so Ill work on integrating that into the version of the plugin +Indeed! +Can you atach a file and spevify which line or part are you select +Pushing to to logic in to make an explicit hint less of an issue +In the case I am working with the consumer was discarded without being closed so it doesnt much matter what state the client end of the consumer is in but I wonder what should happen if the consumer is actually being used +returns a map of URI and statistics but the URI key contains only scheme and authority +For your eyes only the explicit stack traceJust a note I think both issues are fixed in the latest snapshotsFirst we added wstx as optional to the bundle imports +We should make sure that it looks good +I have not been working on in a week or so but on I did not see it happen + +Can you please confirm? If so we should resolve this issue as a duplicate +Patch works for me +Filed a bug with major as priority so would be taken post Good point +No stacktrace. +However the change here is really simple There was a Transaction operation that was waiting the previous operations to complete on the + +Hy RobertoLooks fine for me you could updated in SVN I will close this issue after +Right patch this time +FixedNow if the precondition doesnt match and there is an indicator then it will be decremented. +backport this issue the fix +Maybe this is a stupid question but why isnt it possible to sort rows by Long Type +The comparisons between the google results are invalid because of the different result sizesScaling the results by the size factor is not valid clearly the server is responding differently +Ill force push branch rebased on top of current master +Needs retesting after realizing still had references +Patch uploaded +This feature is stale and crustly like a well worn piece of american style bread propped up against a flimsy cabinet panel +synced nodetypes with spec in revision with some minor adaptions +Please submit a patch +Comment not needed any longer +With the size of each column and the number of rows the driver can compute the absolute maximum size of the data returned in a single fetch +Opened a bug for the additional work thats still required on this doc section +MikeI fixed a few more issues based on your patch +is not supported +I am very close to having a patch for this issue +IMO Bind and Unbind are useless even for a VD perspective +I just did a grails and version is downloaded +Ok the problem seems to be when the labelPosition is set to left the required and tooltip icon are rendered to the left of the label when they should be to the right of the label +the test grouping alone seems worth it +I will try to fix first then revisit this jira +The port of our build to Maven is complete. +Now it depends on the selected of the selected element when open SourceMapping file test to check that all elements from Mapping file found well added +That contains the test changes for +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Are we committing on release at this point or just trunk +Werner we use JDBC +is now complete +Possibly you could vote for your favorite one +Minor bug fix of the first as part of +Resolving. +If a source has a problem with a correlated reference as part of a expression in an aggregate then well open a new issue for that. +It appears there is not a loss on my setup or its very small if there is +Applied as r +I can now get Windows command line to execute bash +Does that work for you +Again no problemsCan you be more specific as to what you did to trigger this bug +take a snapshot of a root volume +We seem to be updating only once +Reboot of the computer fixed the issue +With such a small change would you like to attach a patch or send a pull request +Should have been fixed in on +From this information we will write the final release notes for all issues +Thanks Tatyana DanielPatch applied to MATH module at repo revision rPlease check that it was applied as you expected +This does make sense to implement +to import and used the instead of +OK button is enabledResolved +This means in future embedded is going to have one version of and jopr is going to have a different version of +Do you need to pass instance to? It seems it is only needed for setting the abort status +My Pleasure +Will commit if tests pass +The release is already built and being voted on with good chances of becoming the final release +thats all. +This has been fixed on trunk + XA datasource definition from the AS docsexamplesjca the driver is taken from the defaults the in the UI can needs to be filled with a dummy +All of the code in one commit and then the book additions in a separate commit +That seems like a much easier API for Hadoop to maintain compatibly +This might be addressed by Carlos work on is also related to +Based on this further optimizations are possible that improves throughput by reducing the synchronized sections +Why should be necessary for buildingMoreover this bug duplicates only partially +Looks good generally some minor commentsIn you always generate a new blank row such a holder could be across multiple insertRow operations from the same The SQL state you added is CANNOTINSERTATCURRENTPOSITION which in my thinking is a little misleadingThe text of the message describes the error better so I wonder if the constant should beNOTPOSITIONEDONINSERTROWThere seems to be potential for code sharing in with the updateRow and deleteRow case maybe once this patch is in that could be a further improvementTypically for tests when an expected exception is caught its message and sql state are displayed to ensure the correct message is being returnedAnd in all cases when you catch and display a SQL exception you should display the and the message and walk down the chain using get reporting all of them +I think you should discuss this on the dev list first +Next time +There is an important point here +Yes only the first query is slow +Thanks for the feedback Pavel. +Ive attached the file and the that I modified to make the test fail to use as a diff for what is in the baseline +Agreed +an updated patch +Sorry about that +Modified if there is no buffering then the position in the inout stream can be used for now +Since ordering is an attribute of a field order attributes must immediately precede the field nameTests included documentation updated +Quick answer IMAP queries can be give me all mails with foo in the body +I assume you are going to hunt down the rpoblems or at least post a JIRAgeir +Do we really need this? It seems to me that most people wouldnt have a different address for thrift and the native protocolThus far we avoided gossiping any ports because using different ports on different machines is a misfeature +I believe Norman was using the wrong CVS repo to generate this? Norman care to comment here +UI has to do with the tool +Im going to press ahead in but ifwhen we resolve this we should enhance the tests in to test more STATUS calls +Reopened for backporting to and for general experience this and same happen in +The partial data Ive been able to gather is close enough to the current data we can live without updating it for DNA gathered and committed in r +there are probably a couple of problems left in the tck. +Please find here my IRAD conf +Thanks Thomas! +on s patch for the patch rebased on top of r +Yonik could Solr pull only Lucenes contrib analyzers JAR frequentlyWould it help if we also consolidated all of Lucenes core analyzers into contribanalyzers so thered be a single place for all analyzers + Databinding itests for primitive types and array of primitive types +Problem still exists with version can provided a broker name other than localhost in the vm url for one of the use unique broker names. +If we could get this commit from Apache SVN into the FUSE release that would be greatAuthor davsclausDate Thu Jan New Revision It helps with this commit as Committed revision helps withI dont actually do any syncing with the Apache SVN for releases I just kick off the release once I get the go ahead that everything is in the FUSE SVN +Hey Sam thats a fair argument +Patch includes DirTerm combination in PPP as well as proper jdocs about concurrency concerns +Yes I want to include interceptor in core struts API +For the race again but different place +Bulk assign all unscheduled issues with priority major or less to target +It is very hard to go back the other way though and add precision to an int +Pull back if you disagree. +This was using ant undeploy +The extension class has the following importsimport import Thus I think that CNFE will be thrown by Weld when it tries to load the class +It appears that something is wrong on the solaris box +Oops! got caught for disabling rat check in my +yep looks totally fine to. +There were already in trunk. +OK even better! +Patch with the noted improvements +Thread dump when the server is stuck on the thread dump and the state of the server when it blocks it seems that something funky is going on with the WSDL deploymentreading +The example is staying its the example that was removed. +Both repos work they look like what Id expect +The timer task was added as part of +I committed the newer version +Maybe I hadnt properly saved the +The default should remain the fifo schedulerIm closing this jira +Who cares? Isnt it just used for a test of processing the file +I would like to see a new accordion set for components the related VE support +Please dont commit this +It is changing the assert of the testcase good Ram +Trivial patch that adds the class name to the cfm thrift thanks! +That patch works well +Im seeing the error off and on in different test casesI will put together a patch separating the two dir infos in IW +The tests are running without error so far +Let us know the outcome of your tests and I will apply your patch +Update the solution file format to visual studio +Sorry julien could you tell me what I am missing here? do you mean that I should have only on file for both branches +I should make clear that its not well tested or even very carefully reviewed but if someone else is looking at a port it might make a helpful reference point +this issue does not affect EAP. +pushed +a slower but more memory efficient implementation of larger strings +Committing changes described above also updated the docs +Are you able to run a modified Lucene JAR? If so Id like to add an assert at the point when Lucene commits deletes so we can catch earlier when this happens +patch is against trunk yes +Added sample data and unit test +FWIW a separate test case would be fine here but note that the actual fix is part of +I hope you know about the geronimo monitoring portlet which gives quite good amount of information +So itll be hard to correctly separate this +Seriously I just tried it again on +verified at subversion revision +Enabled now and works fine tested local file upload to and passes fine +Could refactor start sequence but thatd be a bigger changebq +Its a backend to +just edited the patch to update the boolean within the syncrhonized block +Yes this is only older versions will not look for the adder. +It is necessary to disable the disk write cache to ensure a commit is truly persistent but that is not the only source of this problem +Terra +Checking it in +bq +I believe TT absolutely needs to set ugorx for dirs containing expanded archives +Any thoughts on this? +Not fixed in CP +is a part of mapreduce now +So Id like to see what exactly you are suggesting for behaviour here because I for one am pretty well stuck +This is a common issue to native threads in VM +Dart its tough to do anything with this +Ah youre right thats a good point +Since DRLVM is the only VM which works on x I assume that you were using it +This is the sample WSDL to try reproducing it with +So I would rather run one large compaction once a day or two and have to gzip the biggest part of each region only every few days in place of doing every day or more then once a day +By which the consumer thread can pickup the next message +I will fix it later this week +This is actually already in the latest CVS if you want to pull it downNote this is actually a GCC compiler bug and not a bug in the source code +This should be solvablefixed with +It would be a huge help +Usage of in +I might have created an issue too early +Thanks for volunteeringYou may want to take a look at the changes Kim made in as similar changes will be needed for this issue +The real fix is not to starting a thread in the constructorsIn this issue is changed back to package privateI have committed this +I just committed this +I believe have addressed the listed issues +Thanks for the new documentation Kim +This is an awesome change +Creating new long and short texts works as expected now though +Next generation and update to the sitedocs should fix the site itself. +Also a side note we arent a web server but a database using HTTP as a media +This turned out to be easy enough to fix hence Ive backported it to as be available in the next nightly M and snapshotsJuergen +Syed failed theres no +Closing. +Sorry Ali just fixed missed that file in my environmentThe build is still broken dont understand why let me know if its related to my changes +Also updated testcase to add +I was too optimistic in thinking that my JVM build of Avalon was okay +Invalid field value for field selectedOperations +KoenDo you have a plan to update the docs in the next weeksMaybe the doc team can rebuild the docbook from the HTML pages and maintain the docbook in the jbosstools SVN from now on +Test case in works ok but this new issue also affectsI updated Affects Versions to include Mikko +Dont know why it worked once +Thus when we kill the process its up to the kernel to make sure all is well +hmm +Update meansVerify the getting started guide are actually updates from the dev team about new features and incorporate this into the doc +Will comment more on this issue later on +Thanks for the reviewfeedback! +I backed out the fix I had after because I made a mistake +how are you going with this +So even if you had access to the corresponding methods in it would not help you because the strings passed to these methods have already been modified by the parser +Note that for all operating systems the easiest thing to do with the zips is to extract elsewhere and copy the correct folder to the correct place in the EAP distro +If there are operations in flight when the learner handler code we are talking about is executed then the ones in to will be sent but the learner might still need the snapshot +Ive separated out about s AES approachWe wont be able to provide an implementation of PKWAREs strong crypto because of the licensing restrictions set forth by PKWARE but we can provide hooks that allow people to implement it themselves this isThis leaves the traditional encryption for which Bears code is a good start + +As with there was a clear cut decision made to support cmd line options so this patch would also need them added to work correctly? Additionally we know that many people find this convenient and comprehensive especially if they are documented well on the wiki Apart from this Im also ! Ill add a command line option and update the tool description on the wiki +I had to fix one other issue in my test with how I was setting up the servers in the first place but this updated patch should work +Will open new issue if this comes up again. +I disagree with the premise of this issue +I just committed this thank you +Hi JoergThanks for the clarification +Ah yes +This doesnt give you admin privileges in reviewboard +Also we should just use over rather than copy code for readLong +The hadoop version seems strange +The reason is because this seems like the one that Roller users would be most likely to want to extend so that they could add in support for a language that Roller doesnt support out of the box +Assuming this works as expected resolving. +Package is resolved issues for already released versions +Committed +HiSame issue +b parameter used when starting server is reason for itUsecase we should have is this +I havent been able to reproduce this eitherHowever whats clear is that a recursion can occur because to build the path we want to know whether the parent allows siblings so we need that node definition calculation the node definition inside JCRSPI can lead to an exception and the through SPI requires knowledge of the node id whose construction may need the index +I ported this fix from branch into branch and trunk +Still cant reproduce +So I am closing this issue +postponing since this is not not available in rest api yetCould use hardcoded info but not happening for +I just committed this to trunk and +Sometimes it executes successfully but sometimes you will get this exception +RC patch gives to MTPresumably it gives the boost everywhere across usersPatch cleaned up upon the discussion on +I have also seen this fail on the very first INSERT +So since is meant for automatic dereferencing of resources one might as well make it general enough to work on all the resources +Theres mins on followed by mins of general scalability so should be a good issue to transition the discussion +This issue has been found also on Solaris x and Solaris x +Update with refactoring and performance improvements +When changing direction the table is not properly updated and operation shown is incorrect +I was not able to reproduce the problem by running grails sanitizer because the flag wasnt respected +This class has been modified slightly to parse nested expressions +In other words please do add test case once has been committed +o +Ill look forward to reading your message when I return to the office Friday June. +Ive uploaded the second patch that removes the modifier public from +Liyins patch from our internal repository +As well as a workaround. +Trev to get more info for the +Somehow it got left out during forward porting +Uploading patch for Hudson +The outstanding issue with knowing the type in the any was. +read the patch it sounds really clear it is an optimization to improve performances while processing detecting nested node paths +Is this the hri saved off as part of the snapshot manifest or is it only in the meta hri? Is the problem only that the meta entry has SPLIT as an attributeCould we have a unit test where we force meta to have the split marker and the restore +Fix verified with latest srcs closing +TYPO Hi AdrianAnswered on dev ML +can i use this patch +Please test +This patch is applicable to trunk and branches +but the functional results of this patch of this patch as you describe them sound great to meKathey +Its a general issue with the rules files not specific toIve raised this as and made a temporary fix by putting copies of the rules files in srctestresourcesetc +It is working great now +Also since code puts the instantiated Object in the threads threadLocal the accumulation increase with increase in thread in +Ensure that bidi level is assigned to generated blocks +Ive reviewed these integration scripts and fixed them to be more BTI compliantAttached patch contains improved version +Approve +Much earlier my code failed an assertion which caused one of the worker threads in the to die +It is indeed broken +Otherwise lets close it. +Thanks Dawrani Im aware of the call bat inside bat trick but I forgot groovy is actually a batch file +I know its maybe too late but moving the error messages to a resource bundle would be a quite desired task as well + patch with +Some tests such as the one for explicit interposition are definitely not showing up on emma coverage at least as far as the is concerned +Um using catalogs seems like itd be useful if geoserver was processing the schemas and was unable to find them +But anything that should not be overridden should be declared final in the clusters configuration and otherwise the users configuration including defaults should be observed no +In fact the patch will delete the complete job directory inside in DFS if option a or all is set to true +Ill give it a quick look since I touched the delayer last +Many thanks for fixing this Andreas +Can we augment to test the scenarios where feed path is not world readable +Refactors the initialization code into two functions and uses an annotation on get to remove the warnings +Thus the tps numbers above really need to be multiplied by to get the total number of times we pass into the JPA code base +Youre right making it isnt feasible without writing the streamed row out to a temporary file first since we dont have a way to rebuild the bloom filter block index +I ran the unit tests manually +The text replace code in the pear installer is completely broken +reopen if necessary +In general Jackrabbit is full of such side effects +The only problem is that now saving the fsimage needs to touch two files and its hard to guarantee atomicityAnother solution is to use a suffix to the image file name to indicate the compression algorithm +spec file from Fedora Core and put in here unless someone else has a better patch +bundle exec rake warprod +To be safe I didnt add any of them to the repo +The test class name must be unique across both unit and integration tests +one nit the check needs to consider blocks with single replication +The problem in his environment was that claim space kerberos did not coexist amicablyIn retrospect the plugins architecture is reliant on one web service being able to reach another and that is turning out to be a repeated problem under +Solr sets the servlet containers POST buffer size limit +But all the other also inherit this class +The get for was removed from. +The only reason it doesnt is because there isnt a test for this issue. +This may be related to and the issues encountered there +btw +Thanks Isaac +To me it looks like simply changing to follow the new naming convention should be the fix +Ive added small version of code where I have error +Should have something by cob tomorrow +Moved back to way we handle the schema right now forbid the injection of a wrong schema +I think it will be a great relief to many users and support peopleI wonder if we could avoid the reference to by just making it so that implicit in setting any of the properties +It has tokenstream reuse bugs that cause offsets from last token of the previous document to be applied to the calculations of the next document because it reads dirty attributesIts not just release manager being an asshole here there are technical problems that need to be fixed +Oops +After building a brand new dist ZIP and running through the Page Flow tutorial from scratch the bug still repros for me I still get the error after I run ant f BEEHIVEHOMEant Cbeehiveprojectspageflowtutorial pageflowtutorial this bug back +Closed upon release of +And it only seems to happen if these consumers are busy processing messages when the bundle is shut down +Attached addendums to be committed +Were they to do this the sync order would again be undefined and the shared dir might not get logged first and thus we might hit this caseSo it seems like we either need to push down the shared distinction into the or we need to somehow push down a generic sync order concept into the journal set +Screen shot with Firefox is attached +has been committed +For JBDS BYOE is not necessary a feature just a list of features you can install from eclipse marketplace +I can take a look at what would it would take to port the current impl to the latest librariesMight be worth trying what Claus mentions +Not with the current state of things +Test case +factorypath we dont need at all then or do we +why build is expected to happen before build the needs to read mappingsto me it is a little more reasonable to have Metadata and for simple use cases we provides went ahead and pushed this +It is what you would call the method that returns this variable +Yes a snapshot path is supposed to be a path under an snapshot +mdsumacdbfbdcbbc +Dart does the updated option page mean that you cant reproduce this? I wonder if its a issue +Ill make sure to be more clear in the future +Rebased the patch to the current trunk +branch a long time ago will not backport. +It looks like a class called collects all the fields from a class +Ivan you changed get to get it would break reading old version factory metadata +Schemas for distinct filter and sort handle input expression operator +All of the MR tests passed +So as to avoid this we enclosed these things inside the Context +Hugo Im glad this is so easy to do this we eagerly await your patches +I was able to implement test cases by using ANT +I dont like C style declarations because they tend to introduce bugs +Please a provide a quickstart that reproduces the problemThanks +I can reproduce what you are being telling and I am happy that you fixed itSee it is worth to have a reviewPushing your last patch and set the issue to FIXEDThanks issue. +Id like to mention that patch fixes JIRA bugRegarding patch the patch is incorrect due to it introduces some more bugs in verification process +Thanks DanielI think its a very interesting feature I wish I would have more time to work in it +I actually tried the same thing that Richard and Sahoo describe +OSAPIESAPI Brain typo on my partOther conversion functions I wouldnt say ISO entities specifically as it depends on the format as to the notion of entity but Id say thats generally true +Fixed in svn revision + +for the failing tests +We open a new ticket in the support platform +Changing the lock file directory wouldntdo it for would itI am not being sarcastic the questions are for realRegarding the release yes I think it will be a while before is released +thanks Sid. +What project should I pick +The JS changes might make more sense if they are paired with the corresponding functionality on the rendering server that takes advantage of themThere are two end goals from these changes +The failure in profile is jenkins memory issue. +but if you have another suggestion please let me know +patch to add client certificate all Thrift ssl socket implementations support thisPythons doesnt for example +Patch checked inOleg +I believe your reasoning is current on the WSDL paths not sure if you want to do the remote file support +Committed to trunk +Thank you for following up Karthik +Think it is link this what release did you see this bug in +I will investigate how we can make that automatic +Please verify +Yes Id forgotten to check in a couple of changes to give public access to the async shutdown option +and the other half in +So the fix for this issue will not cancel the use of +Release audit warnings are unrelated see +I have same issue to deploy site reports from windows to linux it is a blocker for me +Sure +I think its good to keep the out of the box performance as close as possible to regular Tomcat so that people do not see as bloated +DhrubaSince you seem to be in agreement with the changes for this JIRA can you make it a +bulk cleanup of Jira versioning +I am going to make some changes to the design doc in light of your feedback and the discussion + +Hi SorinWhat is the logic behind removing theTo me this seems to only add a large amount of duplicated code giving the potential future option for to deviate from the defined logging patternCheers Martin There is no good logic for removing the class this was an intermediary patch that should not have been attached +I was wondering if it will be possible to add any regression test in upgrade suite to show how prior releases would favor index to unique index without the fix for and how trunk will use unique index instead +situation cause lazy init eclipse technology but I think this is suitable solution so this is FIXED issue +These tools could use REST if we havethat or JMX but they need to exist before we can consider removingthe old UI +I asked in how to decide between those but I got no answer so far +Xuan Gong are you fine if I takes part from your patch of? This is because we want the branch is always compilableInstead of taking it from different patch cant we just depends on that JIRA and commit that JIRA first +Seems good verified in JBDS Linux +not blocking eclipse just wont support this out of box needs custom launch script on osx or upgrade to +I like those sort of comments +Have done some change based on Zhengs comment +OK just subscribed to dev replying on the Nascif +Committed to trunk. +The delete calls to are gone Im not sure whats going to be left of itbq +Thats right +Need to work a bit more on this but I hope this week will provide some solution for discussionJust to mention the problem is with repeatable timers +Otherwise users have to download those jars before they can use the simple server +Do you also have time to extend the existing unit tests with one that fails before the fix and works when the fix is applied? This is a necessary condition for the change to be performed we dont make bug fixes without a test going with them +patch committing close resolved issues +attached committed the fix which will go out in the Framework release +Tested on windows linux +My routes are working correct without my workaround +I have removed the name from the TLD taglibrary but it seems Myfaces puts in a default name anyway +Im attaching a fix for this issue +Enjoy +If that solves it for you then please close +But though when I was writing the test cases I could only think of one +fix in +There is a bug report from but that has since been fixed +Every time we navigate from page to page we should be redrawing +Changed get to return a list of directories and added a check to ensure that exist in v. +Manually added. +Ok to target this low hanging fruit to Beta +Have the same name for the error and debug pages is confusing +Ref precision for DOUBLE is confirmed correct +Functional requirement It is a kind of feature which is good to have for user security within the system +What are you thinking? Adding the processor as a parameter toI didnt have a clear alternative +Sscoping items for early release of Java Eclipse Mgt Console +we just need to tweak the format and check the files +ThanksAjay +Thanks for the update can you try again? I have just committed a fix +Fixed on and trunk +The cartridge is implemented +Earlier I did checkout using Subclipse +Can you try whether that helps you any? Thanks +HiI am receiving the same error as aboveIs there a known cause to this? And is there a fixworkaround to resolve this issue? Thanks +I agree this should be reopened but not as originally described +I hope Ill have some time to look at this but it may be a bit later than with the other issues as I must finish some other things +Closing issue fixed in released version. +Havent regretted though +This is the implementation I am using currently for Multithreadedtablemapper which is a modification of from +Thanks for the patch Ed! +I didnt branch yet so master is both and +It seems that and classes have similar problem +Applied your patch Oleg +I dont think the negative value for maps in the cluster summary has anything to do with the recovery process +Thanks for the patch Ali +I believe that this jira is ment for you to get the file size of the attachment that is already on a case +Thanks for taking care of this for usGlen +now uses the rules described in the pdf specs to select the correct glyph idThanks for the report +Close after release +theres no finish around your async +It looks like this bug was in limbo +updated docs +The patch doesnt apply with todays code +Sorry about that +I still think its more important to figure out how to do configs and then work back from there +Added new module to global build reactor in Rev +Are we going to try to get this done for? Do you need any help with that +Decided to put the release note with so Im closing the issue +Minor thing The hamcrest dependency I added is probably no longer needed with your fix of adding junit as a dependency +HiIm fine with changing it but can we get an authoritative reference from some Apache legal page or email thread? Or does your description comes from such a source? If not can you get one? in revision +When hitting it from a browser I need to do basic authentication with the message Artifactory Realm +Im looking forward tosome more test conversions this summerPlease verify that the updated code in the trunk looks correct in your sandbox afteran svn update and verify that you can build and run the new test from the trunk inyour environment and then please mark this issue closed +looks good +This is the last drop wed like to make before working directly in Apache +If we could extend the Microkernel to support journals for branches I think this might the best way for the time being +Ravi I checked these six tests and it seems they were removed as part of +For the scp problem can you try to run the deploy with the X option? This should give you the detailed reason why it failed to deploy to a particular repository +Hey Ted a package private setRequests seems to be the way to go rather than pass in on construction just for the case of passing a request count on daughter split +Made new issue to see if other znodes we should rename +This was fixed by closing this as duplicate. +Ive moved the exclusion from surefire to nothing done about the include is better to close this issue and open another in the surefire plugin about that +I have committed this +Looking at the above code the order will besql setUp builtinAuthentication setup setup this to a major test issue from critical bug +Remove unused variable not really related but I didnt want to open a separate report just for this +Patch v integrated to TRUNKThanks for the patch DavidThanks for the review Stack +the change as attachment Jira clobbers patches in on trunk. +Im attaching a trunk patch you can try out +Proposing a fix for this issue + +This is mainly caused by the fact that the MTQ in trunk can do seeking while the the MTQ in x can only increment to the next term. +the map is a json pseudomap not a collection +again everything is Maven version are you using if I may ask +as I recall the are precompiled and put in a jar file because allowing them to be compiled at runtime ended up generating long pathnames that caused problems on windows +Same issue as the binary target in the build should depend on the target +It would be nice if you could backup and restore the dashboard layout also +run forrest seed +I committed this. +The file is a backport to of this change +PrakashCan you provide your opinion +to having ant make Chukwa runnable immediatelyIm open to saying the chukwa scripts should assume jars live in build by the way as an alternative to moving them to CHUKWAHOME after the build +Then deployed to a jboss as server +So it has and should have a higher priority +I did add a hudson job to upload maven snapshots of our projects to the apache repository on a nightly basis +I think this has already been fixed in RC. +Sorry for the waste of new correct archive is my apologies +Yeah that makes sense +first patch version +Ok one more correction after implementing a patched version +Committed revision . +Thanks +thanks pat! +As the patch fixes one more issue additionally to the reported one I attached a with a test class which should be added to drlvm kernel testsThe test class contains two tests +I dont see any reason not to proceed its a net improvement to Derby and wecan continue to test it in the trunk prior to releasing it in the next releaseTwo comments on the proposed final patch It seems like there are inconsistencies in the new code +Previous patch was diff taken from wrong location in tree +Bulk defer all frontend issues to +I still get this on the Samsung Galaxy Camera but thats a pretty special case where Im taking massive images and throwing them in intents +I dont think this change is necessary Sijie exactly because of the point you make about not violating any of our guarantees +Can you help me understand how to trigger a data corruption? Thanks +It would indicate a corrupt DB +This is why this fix only applies to Debian family +Using Jays patch I can still get this to work but hasnt fixed it. +for backport toI verified this with branch +It shouldnt be the RM who must do this on release +Also check as it may already have the gettersetters +I just committed this +Attaching the patch +Yes I think the idea is to use derby so that its lightweight and can work out of the box Im definitely on board for that +so my ability to troubleshoot and offer advice is limited +Be sure to configure the embedded transport with before you can do that Added an implicit rest service at URI query that takes a parameter named sql with inbound SQL statement that takes a sql command that must generates a XML document or JSON document +That is for. +If were going to support a user setting a low limit Roberts idea of putting it into or its replacement strikes me as the right thing to do + +I used java to delete the file +Implementing the Lease Recovery cleanup some a completed version for reviewing +If these dependencies are not needed for running the testsuite then they can be excluded in the testsuite poms +Effectively this is the same issue as with filter definitions which we already support at the level +patch applied at r for and r on at r on and r on trunk +Sorry I put up the wrong patch! It was the same as the last one +Id prefer to not close the issue until the change is available at least in the snapshot repo as until that time I cant commit changes to Axiom to use the pluginThanks for your quick fixve deployed a snapshot so Ill close the bug +Okay fine youve convinced me that the idea is not worth the effortP I will just assume that is as close to the trunk as possible +fixed in rev +It looked like files are not transferred correctly +will simply grant ALL access +bq +I will post a comment by the end of the day today +Sonar is the solution in corporate environment to check all projects are following the rule +When creating a Stateless beanand launching an app using glassfish a serialization exception will always be thrown unless you make your Stateless is also a Remote +Can we promote the packages +I removed entirely the part that checks for nocamelstyle since the thrift documentation defines nocamel as Do not use field accessors with beans and Im not producing accessors here +I ran a complete test run no errors +Use sources to begin javassist usage without spring patching +Committed to trunk +Id love to see in the trunk so that I can use it for +Verified for issues to Nick and had to set resolution of all of them to done +Thanks Oliver the patch is fine +metadata +The D should be immediately followed by the name of the property no space +It is probably the same issue +Perhaps something like this could work in a future minor release of seam but I see now this is too risky for a bugfix release. +defer all issues to +We could discuss making unions contain zero or one field but as previously mentioned that essentially puts a literal null into the thrift vocabulary which Im against +I think it would be better to fix the place where this code is getting called so that the locale parameter is never null +But Ill go do it now +Committed +test demonstrating problemwhere lives in the chain doesnt affect the outcome +Why do you keep rejecting adding default filters? Like for logj file? +Verified fixed in CP fixed in CP CR +Talked with Ashish offline about this a batch retrieval method for tables is definitely required as the RTT would be excessive with s of tables +Moved JIRA to common as the changes are in common and hudson barks at the patch if in Hadoop HDFSThe failing HDFS tests go back to green with the patch +To disable secondary key optimization use option true +Hi John Could you please confirm which branch this has been fixed onThis behavior still persists for me under Maven attached +Resolved since user cant reproduce it in the last nightly build. +Could not reproduce in rDid the following rm rf +Overhauled aggregator so the timeout is now inactive all resolved tickets from or older +My guess would be the parser was never passing inModeAttribute so the parameter was effectively unusedLooking at the code briefly it looks likeModeAttribute is never used either +One use case of using our custom wsdl is when we need extraClasses for our wsdl +Regenerate the patch from the right directory +Sorry it is there now +Ok +So kindly suggest should I continue in this way or you have some other suggestion regarding Quality Check of purchased material +But its related to the discussion we are having about messages +Stephan I looked at and it looks good +It looks like a patch was applied to maven to fix +Qpid proton is a separate distributioncheckout so Im not concerned about that one for the moment +And if I hit refresh everything gets resetI am not a UI designer but it might be nicer to have both and on each row +note that this is actually simpler than bootstrap since we dont have to do anticompaction first just move all data and were donethe tricky part is not serving invalid data while this happens +I wouldnt argue against doing this but personally I think the current model is fine as wellif there is a back compat stated great if not there are no back compat commitments +Fine by me +Users can still setup AS inside JBDS +Anybody with a clear idea on how to implement this +I tried to apply the patch and some hunks fail +Id be happy to pull the pertinent parts out of my current changes for a new patch if someone thinks this is worthwhile applying +Hi HowardWe set timeout to seconds and conduct the timeout test under JBM Remoting SP +At worst I will change the default configuration to use validate and document this for users +I scheduled it for which is due out by the end of February +TimThe issue is not reproduced on vm classlib version +I believe this is fixed at trunk r +Attached an updated patch generated from SVN +Doc patch looks good Yip thanks for the patch! I built the doc successfully and viewedthe page and it looks fine +Thanks for the review Todd +srcjavaorgapachecommonsconfigurationimport +This one is correct +Anyone wishing to pass hbase values through a job conf to will have them unexpectedly them overridden +Vinods comments are addressed the newest patch +Fix is contained in +A few items we need to resolve +is not available through the maven repos yet +But I see now that it isNow fixed in Metawidget +I noticed that create passed the generic type intoThis patch includes an equivalent change to create to pass the generic type of the field +We should also have something that says that all companies should be treated equally +Can you please confirm +The replyExclusive is exclusive per +It looks like this bug is back +Thanks for confirming youre OK with this Suresh +varcachehttpd does it exist? Yes +and then hashcode should also be calculated by recursively calling hashcode +I will adapt this for the upcoming cr +You are correct I did misunderstand what you were looking for +I am working on some code i will submit over the weekend to set a max value for hits per pageI discovered this to be a serious issue with the opensearch as well since some people were sucking down wayyyyy too many records +Its all Apache licensedItll be a few days swamped with other stuff +How about this patch +Note that packages need also to create the relevant directories and set relevant configs in the pseudo conf the split packages this is no longer a problem. +rpms then we need everything that will be packaged distributed +There are lots of use cases where delete by Id isnt practical this is just one more of those casesPerhaps you should start a thread on explaining your full usecase a little better and clarifying what it is you wantneedexpect to do when using deduplication that you dont feel you can do right now +Can we get this into the next build +Closing this issue so it will show up in the release more work needs to be done you can either refer to this issue even though it is closed for minor changes or open a new issue for anything you feel should show in the release notes of a future version. +Peoples contributions are recognized within the version control logs or possibly some kind of CONTRIBUTORS or COMMITTERS type of file +it looks good +confirmed this is still a bug at r +You have update in there I assume it should be delete +Justin nope I did not do any load test lately so I dont know +Verification is done by EAP QE you just fix it in AS codebase resolve JBPAPP issue saying Fixed in upstream commit abcdefg +If any element is used as fault then it also needs to implement +Someone with a particular interest may want to pick this up +BTW Ambigious should be Ambiguous +The tests are here jbossxbtrunksrctestjavaorgjbosstestxbbuilderobjectjbossxmlmapentrytest there are commentssuggestions it can be closed +Added user input handler for when interactiveMode true to get a yesno on whether to use the discovered version over the installed version andor no version at all +Assigning to Mauro Bartolomeoli the extension transition all resolved issue that did not see any further comment in the last month to closed status +So resolving as it is working fine +when that error happens the error will be sent back to the follower ordered after all of the completed transactions +Ive looked into your environment +Targeting this to since its upstream but might not even make that +Thanks Vamsi I believe that was an accidental holdover from the patch created for +Committed in revision +I did not see it either +The abstract decorator code in and is really ugly thus Id like to release soon and also move to +Only the will get picked up and all will be completely ignored +Things like set the metadata tablet size bytes and make a table with splits +After working with Richfaces and Facelets packaged with the application for a little while I have found a problem with it I have to restart our development server at least twice a day due to space errors +Marking this as resolved since the ability to add a custom request mapping condition is now available +I mean the fairies promised theyd take care of it +what abvt this? how about Out of date? +Catalog will find matched partitions with filter condition +Thanks Amar +After looking at it again I dont see a way to backport this functionality to easily it involves major changes to the configurationstartup +Done +net code sample +Not nice but it +Here you go +zip files for the binary distributions have been uploaded +Scratch +Mirroring to Github seems to have stopped +Closing all resolved tickets from or older +IIRC the meaning of isTokenized was taken from lucene long ago True if this fields value should be analyzedLooking at the current uses of isTokenized in Solr its been a bit abused and actually may no longer be needed +Also X axis was using seconds instead of milliseconds +Tricky bash! committed +It sounds to me more like the requirement ofst levelnd levelhivemapreduce tezpigmapreduce tez +Changed version to SVN revision +Whats the protocol here +Thanks +This may fix the issue or am I bark up the wrong tree? this patch creates a new domain instance which is incorrect +Thanks! patch applied in rev +Bug has been marked as a duplicate of this bug +Thanks Ram. +Committed to trunk and. +Just committed this to trunk and +Today in our application we are creating collections by creating numerous individual shards across our cluster of Solr nodes +error will be logged in the build +bq +The thread safety problem is inside this method +From TCP illustrated section last paragraph The shift count is automatically chosen by TCP based on the size of the receive buffer +By coincidence the what was missing in one replica from domaina was replaced by an extra doc from domainb and vice versaThe collection of a couple of million records has one replica thats missing one document +Here is a patch for this issue + +Resolved the bug in rev +I can read Java code but not with great confidence +The classloader that loads thecontainer jars might contain instances of the containers copy of +Ill commit soon +Active user is wanting this would be nice to get the fix in +I also merged in the latest changes to the files from the repository before generating the patch +Thanks Nick! +If the client is just stashing the result in memory maybe but in the code I write the client operates on each batch of results it gets +I agree that spot instances are annoying in testing +It is only used in +Those could be removed +Added zip file +It doesnt get interpolated correctly and breaks everything +When will be a good time to check this all in +Put duration time into clientThis patch is useful for us to do timing statistics on our cluster without it we cant tell how many MBs were doing via pure HDFSCompiles and works for me tests pass +Fixes the config files to be only readable by owner and hadoop group +With the current test code I dont expect any errors +Does anyone have any insight for me? I added Stack Traces to our error emails so Im hoping that provides everyone with enough details +This does work in our test cases so it doesnt seem that theres a general problem in s workflow +Verified +Please go ahead and commit this +However so far I have not been able to reproduce this myselfThe two bugs that you have uncovered tell me that we are not testing hard enough for enums that are written in Groovy +To suppose that flatten might return an array rather than a collection also means that you would have functions like collect find and grep return an array rather than a collection when performed on an array +And sometimes I get some exceptions while importing +Enter some text in the editor +Not intentional was just missed as part of the port of the rest of the docs +However it should be nice if we checked unless it is a performance hit or add yet another context parameter +Fix by excluding the split packages from the +This issue is fixed and released as part of release +Please resolve and close this issue if there is no further work to be done. +Since we have a workaround which is to detach the volume created from a snapshot and resize it and then reattach it to the VM downgrading the priority of this bug from critical to major +This changes the port to be which causes the system to pick a port +That makes sense +The ydist patch doesnt work for trunk as there is a reversal of commits as went in before this into apache svn while the order was the opposite on ydistI will try finding some cycles for updating this patch ASAP +Thanks PraveenYour patch is intrunk r r r rWe no longer support older releases +Fixed as you suggested +I would guess there are conflicting versions of Hadoop on the classpath +Seems fine now +are simple the UI should be too +Also added commands for suspend and resume +I added the missing hook for inczip in +Closed. +In this case with up to indexes maybe thats a factorI didnt fully check that does scan the table sort when creating the table it was just a quick glance at the code +Committed revision . +rev in r have taken care of moving the and to only two remaining deployers in are the and +HiSpent some time working on this even coded a solution +Bulk closing stale resolved issues +The classes are the the and the build worked fine before adding this dependencyI suspect that having a profile with a pluginmanagerment section that only adds the dependency will work +My is the tests to use a more robust technique that doesnt require a web browser such as +But locally they pass +applied in revision +It is certainly the admin who uses quotas but trash is hardly ever queried by the adminThe problem is breakage of +lines versus lines for the logic comments versus commentThe first is a pseudocode +TRUNK build passed. +I have read your initial description looked at your code but Im still bewildered as to why you thought everything had to beIf youd like to incorporate the functionality you desire leveraging the existing code base and supplying tests then youll be very welcome +Any status on this issue? I saw it was downgraded to being a new feature although this functionality is supported in the Maven version of the anybody tried any work arounds? Im going to explore the to see if I can accomplish the same thing that way. +I fixed this in the last week +Can you please reproduce this and post here? Im planning to add some unit tests along with the patch +The JIRA subject is Number to Boolean conversion hence the suggestion of allowing null +Hello PavolEven Im experiencing that onclick onlistclick etc are not working in +retest and not thess issues out as they relate to the old and +PavelSorry about that +Integrated to trunkThanks for the patch Sergey +Damn +Patch attached +This is a patch that I am uploading for sharing with Patrick +User is good we can delete the old apache mirror +some more details added to the design document +If youll notice theres a new checkbox that says zip webtools projects for deployment or something of the like +So I think we still need a file which can tell karaf about these information +im quite sure this is a very old and duplicated config should get the default config name from AS when it is a AS included ESB +Ive attached a test project that works as you describe +Perhaps we could have table parser referenced to by annotation whereby each row of table represents a of the same scenario with different parameters +Cant reproduce on Windows +Its a Dojo issue but since its the dojo that accompanied roo i had to log an issue here +Hi Andy looking into it now +verifying spec language structure of a bulk update all resolutions changed to done please review history to original resolution type +BC is very helpful if data is accessed multiple times +It hasnt been part of code but that doesnt mean it doesnt have users to take into account + test patch +I quickly looked at the hibernate code in svn but it seems to have the same issue +checked into trunk and branch +Andrea any comments as original author of the class +If gets committed before this one does this change should also make sure that the safemode status field gets set to SAFEMODEAUTOMATIC +Thanks Martinyou were right the menu bar was included in wicketstuff yui already +I have ran absolutely no tests on it other than the repro for which now doesnt hang on the compilation +Thats not the trace I was thinking of +Even if you get you might have to do some local modifications to enable +Verified +shifting to M but I think with the planned changes to have launch be optional this will be need to add some UI into the flow +Sorry that was confusing +I would like to replace the wsdl first example +Attaching v with that change +Issue may be complex and is lower risk now that some time has passed with the incompatibility. +Not a proper fix since I still dont understand why this wasnt failing for everyone +I have uploaded a new patch with the changes you suggested +ok this fixes branchxI refactored into a and extended it by the core and thisReally we should clean this up more for a bunch of the other but this is a start and fixes the build +Please say if you can commit or not Seb +Maybe we shouldnt register thesePerhaps a Connection should have an attribute which identifies the remote Endpoint? A timestamp for when it was created +As of right now the submitted documentation is potentially of an uncertain IP clearance status +Ill close this as fixed +Thanks Namit for svn add of qlsrcjavaorgapachehadoophiveqlutil before +Rather than serialise the throwable and hope the far end has its classpath in perfect sync right down to JDBC drivers this turns the exception chain into a chain of which can then be written and readWe could add Avro support to this when the time comes +Patch supporting JPA with +remove filter from as future version +Thanks! +will commit once tests pass +And I really dont see any advantage of the separationIf the intention is still to split HDFS and Mapreduce into separate projects then we should reduce their interdependencies +We have repair it in our internal version +If you are ok with those changes I will merge this to and resolve this issue +Filed for the part +Here I tried to keep only the state and channel under the lockAlso Ive added a test case but this will only forcing to establish new client connections +But I do not have the overview like you do so I probably just miss something +See +on patch +This is not a supported installation option. +Fixed in requiring this version +To make everything look consistent I can delete the aggregator objects from my v patch and create a class that accepts a list of meters +Id recommend scheduling it for the next milestone if nobody else can tackle it +The functionality is same but look and feel is quite different +Thanks VinodNote that since the default value of is false this might appear like an incompatible interface change for the servlet UI +Switched to then closing to deal with migration bug +If you want to speed things up you can help with those. +Setting the workspace as the default wont make the default namespace listener should make sure that the two always stay in synch no matter how you change them in theory I should be able to change the workspace prefix uri and default flag completely and still have the namespace associated to it follow +Applied the patch +Ideally wed hoped that we could use virtualhosting once we moved to SSL even where would be the root of the repo but even that is failing due to mixture of wildcards certs and yep you can even use for all of your svn using the projectname though leaves us a little flexibility for moving projects to other boxes +unless of course you are happy with implementing the support based on uimaFIT for the time being +Im not sure about this lets discuss a little first +Marking LATER pending a course of action +Includes javadoc and warnings fixes +this isnt how I had intended to allow users to create aliases and well since it does not work its IMO more harm than good +Sounds like we need to declare a dependency on the Datatools project from Seam +I have successfully run both and with this patch +No build changes are included at the moment +so here it is uncompressed +All three seem to work +Closing out as a duplicate of although the manifestation is different the root cause is the sameNote that you may also need to remove the +Committed to branch and trunk +close issue prior to next release +Committed to trunk +Maybe there is a diff between JDK on Linux versus on Mac? Im pushing this to unless some contributes a fix stat on this because we want to roll an RC soon +Zhenxiao Please see my comments on phabricator +for isnt is a patch +Added check for missing thumbnail information in the extracted Horia +Seems like the issue is with the SNAPSHOT logic +Certainly we didnt see that perhaps for not knowing deeply the code the logic seemed simple and feasible +Closing resolved issues. +Perhaps we should put the fix version as +Done in. +Another spookiness I noticed but didnt yet make a test for to confirm if you try to add a Document that has the same field w docValues twice you get a scary exception +Thanks for the submission DavidFor Ive added this to in a slightly reworked version allowing to turn off parameter detection etc! Please see where Im suggesting additional improvements for this feature. +Committed to +Attaching patch which handles the case where a field is present in the index but not defined in the field by logging a warning instead of throwing a +Its already been a month and a half since something linked here has been updated and there arent any open or linked issues here for any remaining work +It is not critical for moved to OS are you using +Ive attached a patch that changes the plugin in the way I described +Ive added Waynes workaround with a modification to use SCCLKTCK in r +Patch containing proposed changes by Matt and to verify works in test now +Important issue but wont be done for +Another thing that could be useful for the testsuite is around those timing related tests +Why cant clients simply look at time? Jobs that finished in the last one houror last one day +thanks for the to branch thanks Aaron! +Hope to +HiI have just built a kit from trunk at apache and the up arrow is now working for me +Yeah I had some issue with Apache for Git and had to try a simple commit and push to see if it worked +commit the patch to website +But the actual cause of the second slowdown is not revealed +Tom attached is a modified patch that makes the module child of +FYI I will commit the patch attached above in due time as it happens to be a very isolated change +for the patch +defer to +There status doesnt change +Ths is deliberate +The offset should be not +Fixed in revision +Plenty to review +StackYes David discussed this with me too +has a simple test job that can reproduce this problemYou may use that job to validate your patch +Check not being components also would mess with inheritable models like and such. +It still isnt the clearest message for the but it is less misleading +Closing resolved issues for already released versions +Looks good when gaining a first insight +Thanks for the infos! BTW the funny thing was the longer it ran the faster it got probably cause of jit. +I tried it times a mvn testb mvn installBoth from the jms subdirandc mvn install for entire bookkeeperWas there any other env difference when it failed? Like hdd space constraint mem constraint etcThe basic test is actually pretty basic send msg and receive the same +Merged to trunk +will take a close look +Added fix for job column name change +What functions are we using that are currently being hidden that are only declared via gnusource? If the above is true we should be able to use these via including std headers +Gerhard how long that build takes on startup? Can you also try to run refresh and then clean on your projects before shutting down eclipse and see if it will run update dependencies after restart +Committed revision . +Velocity is trying to find the default VM library and when it cant it reports as an error +ill be submitting the updated patch for this when I publish the Thrift continued on +A few things I would like to see in fluent api maven compliant structure move to junit +Good one +Run the unit test and see the source generated youll see that the file is based on the Request complex type from instead of the anonymous inner type in that it should be based on +As you pointed out may be simple to just unlock the conversation at the end of the request via our existing +Fixing loading of old feature files by converting the in a filter +The check of the container could also be used for other purposes +This can be done in a separate jiraCreated a new class in the snapshot package and move fsimage readwrite methods to the new class as static members + +I dont know if the browser has any meaning +Looks good +bq +I guess this is not the intention of the project owner and I will be glad to get this patch as the official fix for Client +Its a git diff +Changed the as mentioned above +What wiki page youre referring to +Provide options for the current project so if its a war it would be the this should work also inside a library +Lets revisit this once more in the +Thanks +access a jr repository remotelyAnd you cant do that with +Not exactly sure why thay might be +My bad included in this patch +Thank you. +Eds is useful here since you really need self describing data for documents the mapslistssets are you talking about a standard way of using composite columns to implement them or a new column type that has a maplistset inside one column and new operations to work with them? Similar to countersI think this Issue started about having a standard way for CQL to store document attributes in multiple columns using multiple columns and composites with slices to getset themSo if you want to make a standard way of setting up composite columns for mapslistssets I think this issue can be for that +Ivo can you verify in +Pushing it to way of accomplishing it would be simply remove and add marshalling and unmarshalling logicattributes to while the benefit of is that it builds on Spring WS API +Raghu Will that ensure all threads see most recent change to forceSync +Ive removed the ftl template as well. +I dont knowbut some customers requesting this feature and I made it in a couple of custom branches +Could you please review +Thanks carl! +What are your settings for heap size? Are you actually getting an exception? Is it possible that garbage collection isnt being called bc it doesnt need to? I would suggest hooking up to a profiler +If you pass the verify flag youll see that even w just three threads sequence ids are out of order +Hi is used for and is based on soap message +References does not use the ID from the bean but the generated ID. +If UI forSG enabled Advance zone was somehow merged to branch it should bereverted as there is no backenddb upgrade support exist thereAlena +will address that issue +Often in a multiproject environment a master build project with no subproject dependencies is created +After nuking my +Patch available +Now that work is well underway with AS all previous community releases are + +is available here +and committed. +I just committed this +According to Teiid this has been resolved and will be in the ER build +Committed to as rNeeds performancestress testing to ensure no regressions before committing to trunk +HI MauroThanks for checking in the fix can you advise me when can i expect to see fix be in the release? As now i have to patch every release that i downloaded +patch against revision Thanks Sami! +Marking closed. +I confirm that this issue is still in version somebody please clarify what the issue really is here? The description is not clear +Sounds like this hash thing should be implemented in Hadoop instead +Id like to leave it open for a few more days until its clear whether the passing or failing results is in fact literal lookup has gotten slower to the point where its causing a measurable difference in prototype execution times then we probably need to address that issue +includes a mechanism for making it compliant so it is one of the existing implementationsBasically were saying that compliance should come via plugins regardless of the underlying implementation +Was given the to include Hibernate into so these changes are included in that release +I went ahead and posted it on +The bug is that a label should never appear if it is part of the body content of a and that has children +The installed file is called if you think its fixed resolve the issue +the newDo we really need to include additional em config file into the distribution? It serves only for ABCD depelopment and does not represent any official and supported VM mode +I updated the patch +Yes the are all failing for me too with same exception message +This patch fixes a bug in overseer where when two cores were registered close to each other the edits for the latter would have gone to a stale cloudStatethe zk nodes the overseer requires are now done in single method call before the live node is createdsome old cruft is also last patch missed some edits heres a new one +Resolved for a while +I submitted a patch for +Wont fix +If any effort should be spent on THIS bug that effort should rather be added to the throng of volunteers called for by Jacopo at the end of +Attachment has been added with description PDF file showing a split f CC +POJO calls to from HTTP requests +Attaching updated patch which resolves the build failure noted in the previous comments and adds setting log to the errorStream test as suggested by Kathey +I changed the title as not only the various ODF documents have to receive the ODFDOM version and the timestamp but as well the the fileChanged the version and timestamp in of the download section +Weve been missing tests for Bean methods before which is why we havent found out earlierIve backported the qualifier annotations fix to as well since it turned out to be based on our existing decoratedDefinition mechanism with limited effects on the rest of the codebase + relates to but isnt part of them as in English locale is set to a separate java command rather than to a test case +Thanks Kamil but would it be possible for you to provide a flag image +Now that we have an we can maybe refactore the code even more to support both issues with one code base +Uses classservice locator code in yoko cribbed from specs service locator work +as I said Id be happy to do more work on this if we identify the right way +chown must require admin privs just like in unix unless you restrict it so that you can only chown files to yourself +Hi Julienyes you right about thatI just add it to my test and the same result turned outIll close the issue +I think however because of the size potential impact to my system and license issues I wont be installing them +Be advised that while there is no prior proof if this build sets fire to your computer you were sufficiently warnedWe cant guarantee anything for the nightly builds no need to be dishonest about the link then users most likely wont even click there just as well put the info on that first page +Anyway your request might motivate them to rollout a release of the updated parent POM +Any way it can leak into the release +i think the problem has gone +The actual test is attached +trunk +Here we have translator classes currently +Can someone review the code and tests in their current state and comment +So if LDAP authentication works with username without space then this problem doesnt relate to LDAP plugin +Ive committed the patch with the commented out sinceit wasnt included in the definitive FOM proposal +This would allow for end to end reliability against other sinks other than hdfs +The solution is to replace found control chars by simple spaces during file reading to avoid to make the parser crash +Ok this second patch should now be backwards compatible +currently I cannot install JBDS on any windows due to closing for now will eventually reopen later +Geir showed me how to update this page and I did it +This will make the CR release +HiIt might seem to be a naive question but how to use get from this patch I have not been able to get it workingI tried writing an independent java program as well as wrote tests for the same but to no success +having troubles with my svn lately +Let me try it +Are you sure that those two functions are called from different threads? I thought that all operations for a ledger are executed by the same thread +I am unable to reproduce the issue +Trying once more +The issue fixes threadid reuse fixes fat monitors reuse +Ill the tests pass and commit momentarily. +is not what it was supposed to be it is more my testbed for than anything else and will be adequately renamed and reorganized before its checked IS a demo application of choice. +I think a new request should be opened to pass it as a property to ant. +But I have locally installed ant version for my hadoop user +could you please give us a hint about what happens when you start groovysh in cygwin? Because this wa we can only try it in cygwin and if it works there for us then close the issue +do you mean Hadoop Job and? Hama BSP has totally different Job interface. +committed there with formatting cleanupcan you post version of r for trunk +how to verify this? QA +What happens if someone asks for but doesnt have JNA installed +Wanted to use the same scan for recover with other JIRA +This is the version in trunk +Resolving this I believe it now works as desired for both situations and nothing has been done with it in a approaching a couple years we can open a new JIRA if something else happens. +Investigated and the changes made do not affect doc bundle +At my last job we switched back from to our previous JMS provider due to issues and at my current job well likely switch to something else soon as doesnt do well under high load in large systems +Attachment has been added with description Patch obtained with the command git JeremiasThanks to you +The file is a patch for Xalan interpretive to support the secure processing feature +added documentation change. +Please review and provide your comments +Heres my proposalThis patch also removes the troublesome call to +Updating the file +Maybe Ill use my extra disk space to store the backups of my computersDont reply to this if you dont want to I just wanted to let you know that I read your comment. +This issue is fixed and released as part of release +Merged to and for release +I forgot to commit this will do for +A better short term goal might be to produce a combined tarball using a assembly +I verified it +podem afegir sense pressa sense destinaci fixa +by making a plugin it can pull openejb onto the minimal server likewise MEJB and the derby database +May I close this issue +Thanks! We have started to use the jdkim library against our production mail stream I will let you know if I see any other issues +Added ArubaFirst thanks for this contribution I believe your driver will be nice addition to Deltacloud API drivers +It doesnt look like this is possible +The only suggestion I have is to check with Justin +lib +This looks good +Having said that I agree with Werner that compiling against LGPL or comercial code is a +Yes it seems to be a good to have which can prevent the problems like the ones that occured with the test case +It will be quite easy to write the serializer but the parser might be more complicated +Patch implements functionality of having all tracked stats displayed in the default serverview page +With an AMQ server and a client we have no problems +Patch for the issue +Patch file attached for this +Hence closing the bug. +Otherwise I think these changes look good +Committed at subversion revision +We can definitely move it from contrib onceit is maturesafeWhy is development in contrib such a bad idea +So either declare the dependency first or explicitly wire SD Commons to +INFO Running for rows took msjmspaggihbasetesthbase cat grep took grep INFO Running for rows took ms +test close if verified fixed. + +Can you confirm that your bit install is otherwise valid +This last patch looks good to me +can you include the for this run? If at all possible it is best to either zip up the entire stress dir and its childrenor if it is available publicly provide a pointer to where it is +There is no common parts with +Therefore this JIRA bugHarryThe page with dependencies or base requirements should preferably be rebuild on the new wiki site as it is key info both for users as well as committers +The phone and voice agents are always ignored +I removed the signatures from the repository since I theyre bad weve moved on to now and I cannot verify that those binaries are correct +Yes I think we can close this ticket +Fixed on revision +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Glad to know if this has any help on this issue + +The is textxml charset and thus the encoding SHOULD be +Ive applied the UNIX part of the patch not sure why the windows changes are required +I think it is more likely that they are caused by two people running the tests at the same time on the same box +I remember one major drawback of Xvfb it is not possible to connect to the session and see how is build doing which was really helpful for us sometimes in the past +Things like eclipse project nature can be configured in +After a adrian remark on dev I propose a patch to move some Party help Label from Common to Party +Max not sure where else to go with this one +Is the first test useful on its own? Not sure it also kills master but doesnt really test any particular scenario +The reduce input type flag should always be the same as the map output flag and thus it is completely independent of the input type flag for the mapper +patch fixing the DWPT issue +Ill have it fixed for +Can someone attach a testcase otherwise I doubt well be able to reproduce +I can confirm the pactch doesnt work and the exception stated is thrown +But as and uses internally its not guaranteed that the of a is the same Object as the corresponding of a +Master Rewrite should make it so assignment is testable outside of region context +Please your application against SVN trunk and the issue if the problem still persistsOleg +Yip I see that test changes still have the SYSTABLEPERMS checks to see if the test is behaving as expected +But i not test the Fetchmail here +updated the workaround to make it clear it applies to a SLSB +Ne patch addressing the issues found during building with zookeeper trunk +Thus embedded libaries are unpacked into jar files that is unpacking is not recursive +As I mentioned in IRC rather than using read repair for smaller differences between nodes I wanted to make more efficient for the use case of small ranges and continue to optimize that repair pathThe attached patch adds a Predicate that will apply to the key of each input +Otherwise they wont be garbage collected if they never trigger +set user expectations +Thanks for the spec! +Shows in the address bar how the region names are now encoded properly +Ive updated the tab to be Details and left the field as Binding Name but added a tooltip to read Optional name to override the default generated at runtime to uniquely identify a binding used with a service or reference +Closing the bug as I couldnt reproduce it on the latest build +Committed revision . +Closing issues +Can you please reply to the comment I posted? was an incompatible change +Followed Cenqua instructions just took us a while to get it integrated into Xircles. +Escaping should be done conditionally depending on the fact whether were outputting scriptstyle text or notua should be represented as for common text but as ua for stylescript tags body +Both have been done +If the replica with the lower HW becomes the leader then it will truncate its log upto this last checkpointed HW and start taking produce requests from there +In my opinion in such a case it is more useful to have the complete logBTWIf you have such issues with tests create the tests so that they wait for Central to be initialized +rUbuntu +Current version started on image will say unformatted or OLD version +misse a except findbugs warning +The empty avro directories should also be removed +Any progress on this +Browsing the SVN source for the configurationexperimental branch it looks like one change will be new package names for commons configuration +GC would have some problems with that but I think a FULL gc could avoid itThe point I raised the comment is to figure out if two closed ledger hold different version of ledger metadata in memory is it OK? if it was OK then I think the fix looks goodI think this issue is a know issue which would affect correctness of ledger recovery +Attaching the patch +attaching proposed code changesthese were tested also and proven to resolve the deadlockcredit goes to The Dong Pham Thanh for coding and you please provide a simple test case which is able to reproduce the dead locksee in for some examples +This appears to be a packaging problem with not a bug in ESB +Applied with some changes +The is not the client +This patch should fix the issue +Iyappan identified a few bugs +Id guess this is caused by the test added as part of which interrupts the main thread +I reproduced it with Infinispan recently +why do you need to make any change? I thought there is no way to express snapshot diff in distcp right +Part of the epic +The closest you can come is to install an entity resolver which recognizesattempts to load that DTD and returns an empty stream instead +May need some assistance with this since it seems directly related to the changes introduced by and personally Im skeptical about those changes +FYI it looks like an issue has been created for this on the Hadoop end at +ok i see we will check the possibility to extends maplist by not losing performace tnx +I dont think we should apply this to the branch +Dan wroteId thought that this issue might be related to the security manager not being setup correctly for decorators but that doesnt appear to be the caseDo you have an objection to putting the Encryption suite before All to resolve the nightly failuresKathey +I created the previous patch on the release +The restore system workspace in runtime is impossible +I committed this to. +I would like to propose this issue as a blocker for AxisThis issue is making it impossible for me to get a working webservice +Can you try again now the hang is no longer an now on the ordering of the nodes causing connections to cross +You are correct that no suffixes were added but this could be addressed if the need s what Francisco has in his example so clearly the need +Itshould be an easy one to commit before the release +I think I agree +Running on Mac OS and Eclipse will not be fixed +Committed to trunk +Fixed the problem with forms that dont have a name or an id they just get skipped for validation +I took a look at it a while back and came to two discoveries First that theres no point in replicating all tests in all four possible recognition methods second that getting it to work as a testmapping is something I just dont know how to do in that bloody test suiteTest suite desperately needs some documentation as to how to add new tests with new mapping files and new classes for folks like me who dont want to spend all day trying to figure out how to reverse engineer the bloody thing +Corrected the relative links to reflect the actual file locations in theStruts documentation application +Is this you scenarioMichele +The issues is the second step of jthread and hythread libraries clean up +This was mostly fixed in sdk release +But I have some reservations about all the different possible discovery options into the main schema +Code review comments addressed +I think we need to fix this beforeAre u able to recreate this problem? Am not sure i agree with the priority if we cant recreate the problem +Upload a patch to add implementation and unit test +Patch fixing both issuesCommitting shortly +Ok this is getting ridiculous an automated reply triggers an automatic update of the issue which triggers the reply againCan anyone turn off the automatic updating of issues via email replies? +Hi allHere is a design doc for the Workflow AM +The way solves this issue is that it fetches all the events in one go and latest events are fetched faster +That looks great DaveLang is just about getting released but as soon as thats out Ill go ahead and commit the patch for the releaseThanks for the patch +And maybe doing this better consistency check on startup would also prevent any of those why is this counter negative we have had reported in the past. +Thanks. +Yes possible unrelated to your changesLet me look at your work over in git +Are you sure that this problem is related to plugin? This is probably a issue +Yes Harsh you are rightI remember Cygwin is required for du df +So if you decide you cant use you can still use this patch +Happens here too especially during development using Eclipse IDE ToolsA Maven WAR project having a dependency on a Maven JAR project the WAR cannot be deployed successfully by Tools and AS +Nothing more to do for the moment. +fixed. +Irrational fear of a second return has lead to lots of unnecessary code branches in my experience +This would be a wonderful change +Woden has fixed this issue +Resolved right now tlibVersion will base on Struts version +However the order they are presented appears to be the order that they were created +Try to insert this value without apostrophes +Max DartWhen youre talking about more clearly indicating whats Selected +It is working well for us +Patch is andor Robert please test and then close. +Journal flushing throwsImmediately switch to mode +New default WS Tester body section screenshot +Comments on reasonable thanks Sam. +on IE Safari FF FF Opera +Tests are now passing. +compactStores are there still to support synchronous compaction need by utilities and testingGiven these I basically go with the original proposals and fix it so that test code can get split key with a separate method +Resolve is all that is needed until then. +Short of that attach a full mvn X output log so I can get more detail please +There should be no reason why you cant use Spring +But I would argue that bar is present and completely valid in this case +I will fix the test case you uploaded + is different for each platform so its copied into the output www directory during buildTry typing cordova prepare and thenfor Android look under helloplatformsandroidassetswwwfor iOS helloplatformsioswww +Im the maintainer of and I acknowledge this is a bug in my library +How can I connect to the update center? +Thanks +just committed the latest patch to trunk +So patch is welcome the same as the additional test +What is the concern for a tag being a valid unicode string? If queried via rest API the values would be urlencoded thus no harm +So once comes out may be a case of updating the doco +This is a copy of the main one and needs to go in the collectionconfvelocity directory +Trunk patch just ran dosunix to remove Windows line endings +I have a concern +ant l yes testKilled after a couple of minutes +My RUBYOPT is r rubygems +Thanks +Since you have already gone ahead and assigned it to yourself why dont you take a shotIf gets solved then my guess is that this jira might get automatically resolved +Need to rebase patch on trunk +Verified +Including source code +Thanks for the quick turnaround guys +Need to backport to Camel branch when createdFabien thanks for reporting. +is using Jetty as this is easily resolved by upgrading to Jetty or did this as part of WELD will never require a session to exist until it needs to put something into the Session or Conversation meant +applied in and +I manually verified that it corrects the log message +primary Smix server or an http request comes over to a HTTP unit deployed in the +This is also fixed by the patchPlease someone review and apply the patchThank youDumindu +I think that Jimmis solution is probably best and with an extension to the you could probably handle almost all cases for AJAX login +Fix in r. +Issue resolved and committed +I pushed that change already so Ill close this and the see you actually did rebase and that your comment was about using the artifact The maven artifact is only used when setting up the classloader for the legacy controller in the transformers tests in all other cases the artifacts are indeed from the current build. +Hi Ralph +Not unlimited scrolling and zooming is fixed +Hi OleI dont see this test fail on the latest Tinder box tests results on the list is it failing on some platfrom you ran seperately +Or you may want to have a go yourself of courseIf you need production level support you could always inquire with Oracle +Does the same test work if you pull the plug instead of disabling network connection +Please keep this issue open because we have just converted one screen from the list +Issue reopened to change Release Notes field. +In this case when function getipforinvokecallip was called it tried to disassemble memory that could be at the same time patched by a different thread +Thanks TomIt looks like the upstream wildfly update to and is still Alpha as it does not been merged into the upstream +No need to check there +It would be nice to push them to dont see a reason not to +This is to make and trunk consistent +Assigning a thread pool to a component wouldnt have any affect on that +It certainly seems a better option than the one on Id prefer to over +. +Attached is a proposed patch for review +This patch passes mvn clean install + for apache is not actively worked on +In addition the admin UI was changed a little bit to make it easier to navigate between different released! +Brilliant thanks! Looks like jclouds is expected imminently as well +The conversion of +Temporary solution is to implement Continuable for the inner classThis can be fixed if another selection was found to determinewhich classes must be transformed instead of implementing ContinuableMaybe some wildcard statement like transform all mycasses +Proposed fix and use Carlos fix as a proper CVS Carlos fix as a proper CVS patch in the hope that someone will apply this before is released +Now hopefully fixed +See comments for how +And what OS are you usingAnd can you report the JDK version number as well +consolidate +Verified with JBDS +Looks good committed to trunk and branch. +attached is an improvement to make the error more friendly +Am I missing something? +I think its a very great technique and it spares you lots of boilerplate codeUnfortunately this will by no means get into the JSF spec I think +Also description of how exactly maxActive maxWait etc are enforced after the patch would be useful +Closing old resolved issues +Pushing this in +Force the patch available status again +We can open separate issues for and later for additional improvements +The JUNIT passes after change the typo. + +Aright +As simple as this +Hi MichaelMany thanks for this input +David have committed this to svn +Thanks +Erick my understanding is that requires a physical presence on the same host where Solr is running +Whats more as long as we dont have a tracability for modifications Im not in favor of being able to completely modify the history of comments +JPOX CVS has been updated to support in the result clause +Thanks to Bertrand who suggested to change the name to something better +Committed revision . +It also will decide theescalating set of repairs to use against a particular node which is in aunhealthy stateFinally most of the repairs short of are fairly straightforward to implement +Semantically I thinkem semblaem pareix that the translation is fine but to think and semblarparixer have different syntax +Lets continue implementing it in separate issues +Rather than laving trunk unusable I am going to revert the changes for reopen that issue and close this one +When I get a free moment Im going to look at the code and see how easy it would be to slot in a check on formatting to include timezone a major bug I missed that field in the submission form realise uses UTC timzones Im saying the modifier from UTC could also accept the format as well as HOUR as the former is what the ISO spec mentions. + +Yikes set to I better revamp this to current code +Consider a fix as a subtask of +changing status to closed as has been released +Perhaps you need to do stuff like that for truly massive clusters but until we proof point everything on a few thousand nodes it doesnt seem worth this kind of change to mebq +Second the reverse filter totally makes sense +Related? The original reporter did not convey the hierarchical organization of the projects he has +Thanks Milosz for raising this issue +Fixed +Ill update you when we come to a conclusion about it +I have worked on API docs for previously so happy to take this one +Integration of maven with starteam would be a great help currently we have to use the stcheckout ant task which seems impossible to call from is planned +I have confirmed that the repo works and accepts push from at least one committerI dont see a mirror at yet +what hooks were you thinking of here +Sorry I havent gotten to it either. + in + +Progress Committed revision No longer hanging on the openejb issues after new snapshots were published +When this in the future we need a solution that would not only maintain backward compatibility but also allow us move away from the fast message formatting altogether +Good catch +Any further tests will need to be written on a basis as a bug crops up we add a new test +Additionally the implementations even better take care that the initialization of a converter will not break the setup of itself. +Reporters of these issues please verify the fixes and help close these issues + to this idea +Ironically the major reason that version will be exciting is that it allows for compatibility with old APIs +is tracking issue to integrate the final tagged version for M. +I am classifying the issue as cannot reproduce +Joydeep Yes is just a parser for thrift DDL +Someone needs to confirm it though. +As I suspected the issue is that the conversion from to is inexact +We are putting our main efford into JAXWSJAXRPC tools issues may handled by community contribution +Did you have Prompt for Basic Authentication checked when you try this? Cause I get the UIDPWD dialog once and it happily goes off to do the invocation +extend issue to the whole package +kmz file containing the raster image if we want it to be standalone +The change makes all the sub classes of call initialize for contained making it incompatible +This is the normal migration process for a project but Im unfamiliar with any subprojects of Tomcat +I still in +Can you please try again or provide details about what exactly is going wrong for you +I think it would help if we produced a similar design doc for review Ill work on that +We should perhaps create a bug to track this optimization +I got so excited about ISOANSI error code standards I forgot to actually look at the patch +Updated patch to trunk changes +I prefer to use blank lines to separate logically different groupsbq +Thanks alot +I have the sample building and creating a WAR that seems to run however it doesnt appear that the binding is creating any servlets with the servlet host +Any update here +Steve this issue is more about how the various distributed parts of the system work together in the event of failures and under stress +Get a from hudson so can get this committed +This is an EAP issue they duplicated it +when the gateway is loaded always the same instance is used this is very importend otherwise you would loose the inner state of the instance +The default should be +interesting +since the test repository is already considered scanned and the time stamp of the artifacts are older than the last scanning +So that its easy to decide if something is missing because of an issue or because of beeing not supported +I have generated a diff patch for branch and trunk need I provide it? By the way can I have the honor to be the assignee to fix this issue? Thanks +will be necessary for a patch against trunk +Hi Les WilliamI believe your above comments are currently being supported in the component +Im going to commit this +Closed upon release of. +Thank you Paul +For our ORG project works fine +David can you take a look at this test failure +Comitted in trunk rev +Apparently this patch breaks other things in WSSJ +It is not needed this could be done in a separate method but that does not change anything +Now in +passed on my box +Closed for release +has been marked as a duplicate of this bug +JSR artefacts version are in the main Maven repository and has been switched to using them and passes all its tests +Patch for just checking for both and failing attached +weird +Youre not really supposed to be able to tinker with DNS the fact that you are is a bug that will be fixed next are you trying to achieve +Fixed in version and. +Changed abstract to reflect that this issue has really become about improving dependency behavior of manifest processing not getting basic manifest processing working on the C backend. +Could you please upload it againPropose to take this in +Thanks Ning +Thank you ByranYour test helps me lotI think reporter of may mistaken at server side for error at client side +Thanks for taking the time to test the fix KatheyI havent been able to reproduce the problem myself but Im marking the issue as resolved based on your testing +nit max seems a bit wordy +and +Using cqlsh in +Once the mapjoin task fails we replace with the backup task +Running and still have this error +fixed the patch +Marks advised me that porcelli and jliu have almost completed the new design and we can start to integrate Guvnor soon +Currently testing with SLFJNote that its a different issue but Im not happy with how was resolved +in fact I have used hibernate years and it is the first time I have decided to make public my suggestionsI think it is very specific to suggest that there be a lax mode and a strict mode that in lax mode these problems get logged as errors but no error gets thrown assume that the user knows what they are doing +Upon further review this probably isnt a bug +Mike suggested I change the the title of this issue to accurately reflect that the data cannot be fully recovered +Ok this is a good point +doesnt the answers provided by Phil Steitz suit you? This issue is marked as closed do you want to reopen it +Ive seen some developers forget to dispose of application context when coding this up will provide an implementation of a base class that can be subclassed with appropriate configuraiton information and lifecycle methods. +Think this might have been fixed as part of Python test work by RG +Intend to add some minor code improvements and fix remaning checkstyle warnings +Attaching another patch for this +The tomcat cpu usage is caused by the same thing as in +trying to run hudson javac warning +Makes doesnt list the first parameter as optional +I note that +Closing as Invalid is more appropriate +Lets keep it as is +Not sure if this is intentional +Patch committed to both trunk and branch. +is true but I dont think it belongs in there since it will go stale very fast +bq +Guys Im pretty sure that this issue depends on on right now +I totally agree with Corby that this should be would like to be able to treat the the same way as other regular spring fix thanks! +Look like you mean add this? if this resolve the build on server problems please check +There are three sources Partition All AM metrics come from the sourceAll metrics come from the Partition sourcesAll metrics coming from objects shared by all in a come from the source +Description of the patch Continue reading from Reader when return value of read is +Master used update +I have removed the disabled national character code from trunk +I like cause If it can grow it will growAnd overall Tks again for your help +Each new attempt to get a connection from a failed pool will trigger an incremental recovery on it which can either succeed making the pool available for use again or fail again which makes the pool throw an exception when a connection is requested by application codeThis state is completely internal and automatically managed it is just exposed in JMX for completeness +From LeeFrankI did some digging to try to understand Francois test results from yesterday +And it will only stop the current firing rule from activating itself +Closed as suggested +Patch to special case the empty columns request +New feature EJB server portlet issues in AG is not a defect with the Ejb Server Portlet +I can record a video of the bug and then upload to you tube +While the reported issue could not be reproduced nevertheless added a missing validation to prevent this exception from being thrown if a bad input path is specifiedPlease feel free to reopen this issue with steps to reproduce the problem if I have missed something +And now it seems that the problem has resolved itself on the forum + +Reproduced with EAP branch r +This is expected behaviorWe save the encrypted password only if the VM got assigned with a SSH public keyThe password is encrypted with ssh public key and stored in the DB + we perform some validation against the sum of and so that we can remind the user of the proper mixThe reminder can be displayed on master UI +Postponed to tests has been added and now workHowever I have modified the test in order to have it working when checking for the userCertificate in the returned entry I have removed the binary postfix as JNDI isnt able to handle this optionThis is not perfect but at least one can search for userCertificatebinary and get back the attribute value +May need to push this to in DNA will rely upon JAAS enabling integration with and use of Security +We pass this information to the security code which does the macro substitution using the information weve gleaned +Bulk close for Solr +OK +the new patch works + all I can see is that we add the following Z to the pattern and timezone to the inputThis seems pretty weak +So the problem may have already been you please double check? Thanks +Just saying +But there is really nothing and it is correct +had its default value for increased from to Jeff closing jira +I have made this change to with subversion revision +BTW s resources are duplicated in basic and smoke +Curious why is the resolution status invalidDunno its not invalid +Appended the patch and a small example to reproduce the problem +Thank you Chuan! + in works just fine + title to reflect underlying issue +Let me commit this monster +We attempted to fix in but we just made the old faulty behavior doable again and added cruft to the API that will encourage further abuse +We then do two deletes wo ts +The new patch shouldaddress a lot of concerns expressed above +Attached patch +The setting is simply meant to customize where it looks for all the templates +The XML config file of the object were getting this exception sorry I put on a wrong version this is in fact a issue sorryAlso noticed that problem disappears if I remove sortnatural from the collection +On javaI will add that this seems like a on Windows +Waiting how? Meaning that theyre not creating a profile until hbase finishes their release +So superficially it appears that this patch breaks RPC tracingIt would be nice to the trace tests by default +Interesting +Good catch +This patch add a workaround +Defaulted attributes on the other hand do not behave like this. +Please close if satisfied thanks. +HiI added the in there originally because I thought I could get a refcounting scheme to work +I guess this was not deemed generally useful enough so this issue should be closed +Is the plan for those diffs to be relative to each other +Id love to see some annotation attribute or configuration item to allow the behavior of populating the reference with a new object instead of a null +Trying against hadoopqa +This would be awesome +Fixed in r +This patch fixes the findbugs warning about circular static dependencies +has mostly been covered by latest patch +Verified on JBT. +From tooling Im more interested to knowing actual status of deployments such asIs it runningDid it deploy any warnings? any errors? stacktracesInnernested details that arquillian is asking for is nice but im much more interested in getting status for deployment overall first with access to error info +Resolving +i think this is either a mysql only issue or some setting where sessionfactory is creating the foreignkey on every start resulting in duplicates +Ill note that while is similar in spirit that is information we do NOT want to use so I would prefer this be in with the tokens and have everything else in a different CF so the old method of blowing away LI when something goes bad is still valid +Thanks Todd. +getClosest needs to carry around the so can know if candidate is within a regions name +I can narrow it down to the inclusion of five rows in the selected set these rows dont seem to cause any different logic to be followed but they possibly cause a loop that executes SQL in the proc to be run a few extra timesFor now Im going back to +Modal panel size is the same as setted in width height attributes. +Can you link to the repository so people dont have to copypaste from the repo? I have no idea if the initial problem would ever be solved since the original author abandoned this bug +Sure +I think its already a great start if we have visibility of all the beans in UEL expressions and the next step should be +Attaching v which removes all schema agreement validation but leafs SDE in place for Thrift API backward compatibility. +The JIRA that Dmitriy referenced was incorrectly marked as resolved in since it was committed. +Yeah and the have different purposes in my mind +Rather than do manual versioning why not switch this to a protobuf? Then you avoid the manual serialization and you dont have to worry about versioning +If youve got questions about our use of the HTTP client code Ill do my best to answer em +With the completion of the Portal the forums are fully functional. +Good point +While I agree that it would be nice to have sysview support creating sysview manually is more than just pain So I suggest we investigate whether it would be possible to support support sysview in addition to what we have +Gianny I appreciate all your help on checking out this patch too bad it didnt work out in Safari otherwise it would have been sweet +diff for spec will be in EWS as a fixed issue in the EWS release notes +Docs +Tests look good +Also this PR adds custom in on Windows +I this suite separably below test cases still fails testrebootteststopteststoptestcreate +Sorry we do not provide this service any more as it is technically infeasible to do so. +But the goal is to remove all jars +So this is not going to be fixed in JBDS release am I right? Because this is still problem in JBDS +We now only have to get the list of listeners once during execute and once during render. +I think we could do that and possibly add a getter and setter for BURL as wellRobbie I believe Weston addressed the other concern you had +It doesnt appear to work properly with colspan however so Ill see if I can fix that too before committing +If you have other workarounds Im sure people finding this page via Google would appreciate a example +Shame on me can you reject please +While the default AS seem to distribute the and +Best regards by Andrew. +Thanks for the links +Text patch file +a etc +The new mirror is now available at git this resulted in a complete lack of history +Mass closing all resolved issues that have been closed for more than one month without further comments +it seems the pyramid plugin on its broken but I have to investigate deeper to see where and +robert +Can you check the logs to see if there is any indication of the problem +Resolving this issue because the work on prevents the additional bad syntax +I am sure that nothing is going to be broken +Hi ClausI checked the code of and they are singleton endpointand we hit the JMX exception as current Camel only register the singleton endpoint into JMX agent to avoid the situation that you describedIt could be easy to let camel skip these endpoint by checking the annotation but I still want to know if the singleton endpoint still have the potential JMX issue that you mentioned +Can you double check the jboss stuff to make sure that works fineThanks! DanIve verified it in Jboss it works great ThxJeff +Coding is done +Why is in the semicolon patch? I dont see a change from a quick look at the patch file +However tests with Codec show the followingcfafbbfa Windows XP x Cpcfafbbfa x ASCIIcfafbbfa amd ISOIt looks like this is a problem that was fixed in Codec please can you try again using Codec +The persistent data sources are produced by a standalone program +Actually this issue is already fixed in geotidy the question is more when this work can be linked to trunk. +Im going to try and get this one reviewed and patched on +The removal of the package prefix in +Ive spent some time looking at the code and trying to figure out a way to made case and valid and case invalid +Going to commit +Snapshot follows this evening +Bug has been marked as a duplicate of this bug +I did not merge changes to the class though as I hope they will not be necessary +Well like Ive said it is not a major problem +not a blocker +move to +Could do with some more tests but that might have to wait +Invalid issue no need to initialize in unless theyre being used for io +Mass closing all issues that have been in resolved state for months or more without any feedback or update +This is a more detailed design document for the rebalancing tool +How can you start deploy earlier during boot +Fair enough. +Asked about this on dev list no comments +closed for +In a circumstance where we throw before we release the lock its still unlocked in the finally block +I changed the to target and ran maven ant +Patch and script to apply revision . + +Closing resolved issues for already released versions +This change has resolved the problems that I was having and seeing +Ah still new to I was looking through the live stuff to try and change the analyzerSo that will still leave the other settings unchangeable but its a much cleaner solution and the analyzer is really the deal killer here so Ill give that a shot +option deprecate the old apis but implement it in terms of the correct one api back compat only but we drop the buggy code so maintenance is easier +The test has been committed to HEAD +This should be a blocker for +Including the steps here to reproduce the problem of out of memoryFirst create the database with populated tables using the attached java program as followsjava Some information about the tables aTable t has a BLOB column along with other columns +Looks by Rob +Also need to review if this is still applicable for the Netty integration +upon if we want to take version in account +I would like to verify further why the total is for pushes whether it is because of session closing run the example for a longer time like one day and see the problems with investigations and tests are needed in order to resolve this issueShifting to M +Harsh you tried it on windows +Thats a good question +I close this issue. +Well i seem to have xstat also strings lib grep xstatxstatlxstatlxstatfxstatfxstatxstatfxstatatfxstatatAhh think Ive figured it out +Fixed for now +Ive updated the description to aim for I was able to reproduce it myself now with Java +Now the Solr integration test passes too +I also ran ant docs and successfully built the documentation +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Its not fun +finish backport to back port to Lily +satisfying bringing order to all of that +When you have a common header in all of the Callee that header has to be a param in each and every method of the API +Scheduled for to publish certification results for that release in the official project location +committed to HA branch. +duplicate of +Merged. +Note that I do not want to release integration +Sooner or later conversion of data willbecome an issue it would be good to have the mechanism in place by that time +Snapshot related methods should be moved to an inner class or separate class +Nice +So I think we should use a dynamic web project and give a warning +Yes that would be helpful +Batch subsystem as been added. +Patches look good to me +I made some tests with my compressed and the problem is that it sometimes wastes space +So I am ok with this changeSo to apply this +Version Released +A might be to set the Java system property to the particular sequence youd likeSo on Windows platforms rather than a xD xA you could set the line sparator to just xAThis doesnt mean Im not going to fix this bug +This will avoid any issue of that sort +However it just misses one case +The scripts I created have been documented in the wiki discussed earlier +We need to test the coreference component when we update that so it will take a bit time to do that +Please reassign yourself if you would like to work on this bug +The solution is to disable atime checking in +Yes this was by design +This bug has been fixed in MSVC beta. +True +StevoDan can this be reproduced +The objects did however have endobj tags so our fix basically ignores the fact that they dont have endstream tags and terminates parsing those objects at their endobj tags +This feature was deployed to a production cluster running code base very close toFrom earlier discussion my understanding is that it is easier for a feature to be accepted by Apache if someone has put the feature in their production environment. +Partial fix applied to master +The location of them is srctestlib under src +If that works then we can figure out the best way to pare down the WAR file +Guyz bear with me +I have a heavy use financial information server thatsuffered from that bug in and testing we discovered that the better wayis not to synchronize but to instantiate newCiao +It seems that this can work but the relative gain may not be that much to justify it + +The second patch contains the new exception file which is simply a subclass of +This would have also affect message which originated from Openwire or MQTT protocols +Snjezana could you look at this +Thanks! +should simply sublass Object and simply instantiate a parser on ctor to parse and store all parsed content in itsself +Patch will make it into here +Hmm why is the Spring trying to read your Hibernate mapping files to begin with? Are you possibly using a context pattern that accidentally matches Hibernate XML files as well whereas it should actually only match Spring configuration files? good to hear that a change of location pattern worked for you +HiI uploaded a patch for +Or something between with the limited amount of attemptsSolution like omitting if is damaged would be probably usefull here +so the point of this fix isnt to get the best block but to maintain sync semantics +According to the comments I have reopened the issue +So if this can be observed in every type of SSO integration this means that something seriously differs in the flow there +Id be happy btw for the scope to default to test +As you can see the test failed with the tried to access method Can you reproduce the same result with this project +Current C tests all pass except for a couple that wait for disconnected state +The code generated for the attached wsdl generates a call to as well +Martin Im assigning this to Beta per your comment +Hadoop is the primary user of Avro reflect but I dont think the only user +I believe this has to do with the unwanted xsanyAttributes in the which Ive removed +RichIs your concern mostly just in which module the code lives? Or is it more where the code is called fromBecause of the thread switch that occurs for async invocations we have to called this before leaving the layer +This is going to be a common problem and a simple solution +I can also get certain keyspace information from describekeyspace for whats needed inHowever there are two other things that may require some more configuration also needs the Authenticator and the Paritioner +The in module verifies for this +Ill make a standalone test +ProgressCommitted revision +update applied in s correct Gary thank you +It will probably have to wait until since I called a vote for a staged already +with a being planned perhaps that could be slated into there +Use of keyAlias is not possible +Patch and testcase patch also fixes also the Patch many thanks +BTW rest assured that this new version of the TCP transport also supports endpoints configured at transport level +This is a patch illustrating the changes I was hoping for +If I can find half a day to investigate and confirm I will do but i cannot guarantee this. +Removed the white spaces and Committed to trunk andThanks for reviews +Minor clarification of the newThank you for taking this on! Before we hack on this can we put a few counters in to show the cost of the linear search versus the moveit is likely to be I assume you mean trivial +Unzip into the folder test in the webroot Add H datasource test Run testYou will see the stack overflow error +The user should be aware that using a small timeout this would happenI think it is a normal behaviorCan we mark it as wont fixHoward +has been improved with actual resolution +I would rather see an exception raised so the developer can decide what to do with the message +If that is not the recommended formatting style please let me know how to format +does it depend on a specific implementation? dont know +LenCould you please provide more details? +Im moving all bugs out to and we can move some bugs out to to get some release action going +Attached patch file containing tests to reproduce problem both as HBM and using annotations +Current hold up is to figure out how to deal with viewsWhile not a complete write up of this specific bug contains some narration on the background of the bug +If the assumption is that in general we are not going to require user intervention to gather stats what is a better guess as number of rows changes we wont change number of unique values as number of rows changes the number of unique values will stay in the same proportionObviously either assumption can be made wrong for a specific application but what would a good default assumption be +This has however been filed with platform specified as OS Solaris sxuwosa X bits JVM Sun Microsystems Inc +Please try to create a unit test to demonstrate your issue +Actually I think the port option does not work at all for rackup +Now nightly test passes +All these files have headers +Heres a patch that implements this including tests +war to AS M as it is without any change to AS it deploys fine and I see the Hello page +Since I have access to the repo so Im sending zipped project compiled using ant +Thanks AdrianI created which version interactive or static? it seems the problem is only in static but its not quire clear to what the problem is exactly +Ok if no other comments Im going to commit it in a couple of hours +Can you create a sister JIRA in Hadoop Common and propose that there? If that doesnt go well we can always do specific checks in MR as the last resort +Patch applied Many thanks +I just committed this +Link to parent issue able to reproduce the original issue +This fixed the bug exposed by aboveThanks to Raghu for these review comments +This is a code refactoring +You should set stepwisefalse also +Added and removed the old valves +pending Hudson + for patch +If it used to work before its because previously the schema that we printed out would be more like the schema that originally came in but thats the wrong behavior if youre printing out a subschema. +Fixed by properly reset autoscroll checking condition when clearConsole is called +As mentioned above there is a Spring Framework RC which will be resolved with the release of Spring RC +Anyway sometimes one is not allowed to modify anything in a document +I test again this example on and now I cant reproduce problem +Seems like this has been fixed +One thing to consider here is resources used per task +As a workaround you could explicitly create launch configuration for your project and explicitly enable workspace dependency resolution +The fontname from the will do it too +I think that if that is required it would make sense to be another JIRAThanksJoe +Will leave open another while but no longer seems critical +since I am off to the Himalayas in April for holiday and the last tarball was rather buggy I am so free to put another +libgriddevbin +Please DO NOT apply it +quick fix changed string and updated docs +Log from the driver node perf +The issue has been resolved as you pointed out for Alfresco +Thanks Carl! +Are further informations neccessary or is the bug really desired behaviour +We have been getting Spring Integration out this should be merged soon and we hope to get out shortly. +I dont think you have to call a vote for releasing the sources since theyve already been blessed by the Tuscany PMC +If there is other work involved to resolve this issue it should be external to the PM layers +I brought over some more material from the release notes and weve added some more Struts and yes its looking good now +Thanks Lohit +Updated to the new OASIS namespace +I think this is due to the way the parent docs are marked using a bit array for the ENTIRE indexBut arent filters held and evaluated the within the context of each sub reader? Are you sure the issue isnt limited to a parentchild combo that is split across segments +Is it possible to make this switch also allows for unexpected enumeration values +Resolving this straight away as this is a retrospective JIRA to reference a change that was made to without a JIRA +Just install all extras from m extras site and try the steps above +I suppose we need a synchronize in there +Thanks Keegan! +Closing all resolved tickets from or older +Fix committed to branch in revision +Not a bug. +There is no point then getting sonar to do that again as we can reuse them +I dont think we can make it behave the way the real deal is supposed to. +On the other hand test passes on my linux box at home that have +Done with merge in of biased locking code +I found itThanks! +please advise if this is approved or has internal a csv parser we could make this available for users +I think we should assume that a Master failure is promptly resolved so holding back compactions isnt that bad +Going to double check logic again +The error reported above when running in an empty directory may not be related to this patch anyway +This ties very well to the Netty initialization problem as well +Please commit the fix +The patch is coming soon +changes merged I accidentally squashed my changes on into your commits sorry for that! hope you dont mind next time Ill do better +there is a logical error +It is against to +Thanks MichalI have not seen this problem before +I expected to see the direct bindings to be faster as well but the tests didnt show that +Hi Johanneswhy did you close itThe link seems to be still broken +Thanks for the explanation Michael +Ive gone with Carstens suggestion and committed it to scratchpad so it canevolve to a proper replacement for the current first. +Im submitting a patch that looks just for the initial part of the string which should be commonRunning regressions as well +Fixed in rev +Until is fixed the fix for this Jira will cause validation errors in cross document references +From version we do require Java +Should be covered by the new tests. +The scripts evaluate the position of the lib directory them selfs and create the needed classpath as needed +I expect it to increment the current value for deltaV no matter what it is and that is exactly what happens if you try it +Its not strictly necessary for Lucene but I think its useful to keep it there so the analyzer can easily be run from the command line +Solutions to that issue may take advantage of work done on aggregates +Once the patch for is committed the entry can be moved back to. +I think almost work is done +MikeIm new at this +Doug commented many times on this +The axiomnodegetnextsibling method is working perfectly well +If youre using a Grails filter it should definitely be inside the Springcache filter as Grais filters are implemented as Spring iterceptors rather than servlet filters whereas Springcache does use a servlet filter +will do ASAPRobert please have a quick look +review according to M state +Adds support for all request methods specified in the goodThanks! +This bug describes the problem thats causing the overall automated comments on this bug +Paranoia is good +I intend to commit the patch soon +Please review and add anything that is missing +Patch checked in +Not sure what you are trying to do +I also verified that ifI take out the change to the test then fails with the expectedfailure so I think this is looking very goodHoux have you run the complete test suite with this patch? Or have you just run the new test +I noticed this once as well once you understand why it is like it is you know a lot about the XML guys and why they should be fired into the sun. +Ive tweaked the mimetypes file with your change and a few related ones thanks for your patch has tests which show that with the container aware detector we detect correctly +renaming of methods inner class name +I think the ultimate fix is more word obscurity in GWT proper. +should just be deleted since this feature is never used. +It is fine to implement recovery in another JIRA +moving to +id name revision starttime perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf perf in r. +I dont think that this should be an issue since the cache is not persistent +Changed file svn statusM srcref +Should this be closed as not a problem +replication from to latest trunk was the cause of error +Here is a start at a better summary +Thanks Brian + +Let me file a new issue for the original NPE +Fixed with revision +I can get it to use the example testschema but that doesnt have a file configured and I dont want to screw around with the example configs Id rather adjust the stuff and use it +Grr +Fixed on closing all issues that have been in resolved state for months or more without any feedback or update +I added a few questions on related code to that issue as well +I still dont understand why not do it if the plugin already exports core project properties. +Ive applied a fix to version of the Site Plugin +No just changed the test on the javatypename recompiled it and used it in eclipse and it works fine +This would also fix the known issue where we cant replace a bad character with multiple byteschars because Charset logic can only replace with a single characterI may play with this a bit tonight see if I can get something working +If you use your own build from branch it could be difficult to reproduce issues and it brings new aspects to environment +Ill have to plug in my touchscreen and start poking at itNumber is a particular problem for me as it seems that navigation by finger requires a different set of widgets or navigational style than navigation by mouse +Not in love with the layout but design has never been out strong suit +amazing the previous one slipped through really +Ive created a new Mesos Issue Types screen as per above but with the addition of you had currently and so that was neededIll sort out the reviewable status next +Could you please provide steps to reproduce it +Patch makes sense +Ping +Patch looks good Uwe +and extend and implement all the raw methods +Correction! We dont enforce unique timer names +On it was working +The last two comments refer to patches for trunk +I believe the command is msg set aerogear guard on +I tried and it is true +Is there any logstraces available from this incident? Would be mighty helpful +Removed unhelpful asserts and added see tags to copy from the interfaces with revision +requires the sum to be under +xhtml page each component comes from by location of the component right? Nope this info is not included at the momentRegardsJakob +Jee that was a while ago but I think the in question was from jasper as in the jsp compiler from tomcat +Slot millis being computed by the MR AM isnt really correct +Hi BJ could you look into this when you get a chance? was unable to figure out exactly where the clash was occurring so I went the second route +Anyone else still seeing it fail +by the way I do see this issue with windows os unsure if this is also with linux have to test +The problems in Hudson appear related to obtaining the correct version of which is a SNAPSHOT version to verify this revert +Verified +Has this got anything to do with Any +I read the title wrong +bq +If a separate target is required for productive development in some subtree then it should be placed in a separate or Makefile in that source subtree +Raymond Id really like to resolve this issue as I think the problem at hand has been addressed +nd version makes rowdigest fields +No I think the consensus is to have a DWPT max flush trigger not a global one +Thomas could glean anything from the logsAlso can you check if any NM was lost +Since has now been committed this patch is ready for commit. +But in this case its not an IO object and we end up blowing the stack when we try to display an error because it keeps causing more fix would be to modify our use of stdout to handle it correctly if its not an IO +thats an out of memory happened again today + to trunk. +but i ma not sure what version of Roller is used in current +sound like a consensus +This would probably be good enough +bulk close of resolved issues for release. +tests already exist +Most of the reflection in has to do with version compatibility not accessing private +It is for debugging a reduce crash in place on the production machines +Making a file patch and changing to coding with trunk version +Cool so looks like the code I suggested could be a go + for the +Marking as for later now since im pretty sure this is very easy to sure if a feature will be relevant for this beyond verifying that ti works to set this in the launchoperations +This issue has been resolved by the patch of jira +For example when both stable and trunk are building against the same TP jobs or pulling from the same Drools upstream +Version has been rolled back in branch. +Not a bug +This is not an issue and does not represent anything other than an inefficiency the MD hash is tested with the and works +Will look into it +Enhanced patch with generics and Java features +Please note the database log directory will need to be changed to match +Closing the issue as it has been resolved for some time now and nobody commented on it anymore so Im assuming everything is okay. +in theory a client will only need the client and ipc directories and maybe a few more +The latter +Care to say where in the code the issue is since you dug in +Attached this cleans up the architecture of FSHQ with no real functional changes +The documents advice to not use a datetime column inside a TIMESTAMPADD or TIMESTAMPDIFF function is correct but its reason wront it is missing a not +Although Ive not researched exactly how the axutilurigetxxx accessors are used it appears that generally throughout the get accessors return axischart or similar references without a const declaration and with no intention of allowing the caller to modify the indicated string +Moving items not being worked on for M afaik out of not thess issues out as they relate to the old and +The existing code would hardly change +Please review and commit the patch +How to change the status so as to make QA run +The default could be disabled! People should be directed in the first instance to the AOP solution if they want circles and then perhaps to this global configuration option if they are feeling truly lazy and feel like having an incoherent architectureP The current flag in is very hard to access without intruding enormously on the my older JIRA of +It works perfectly for meThanks a lot for this nice feature +Ive reverted this change locally but dont think its acceptable to break a common development environment in this way +But that seems like a lot of work to me +Given there has not been a reply by Ate I think we should try to come up with a test case first so that everybody has a complete understanding of theproblem should be required is a mapping for a class that has a compound identityand then try to run an OQL query against it +I will update the example to use the new code +If the problem persists I guess we have another problem to identify and fix +lib +Yeah I noticed this earlier +Adding Release Notes flagIs there a bug against RHEL to fix this? Does anyone know if a fix will be available before +the trunk patch looks good +Adding that information to the message would be a tremendous benefit +Workaround is to use Derby or put HSQLDB in clientserver mode +Havent found any more specs in a while +trunk I have added the needed spring jars to the kit +Changes made and published +Also in issue Then the console output should be rephrased +For performance btw we dont wanna construct a Camel Exchange Message until we really really need one and preferably avoid doing conversions if we can help them +on no retry for socket timeout +The reason for removing is that there is no clear criteria as to which semantically rich description should be there and which shouldnt +Removed superfluous flag Finalized SPNEGO example and tutorialI was simply not aware of your work on SPNEGO support for +RonghaiCan you upload a patchThe improvement is measured in execution time right +Now since we have multiple threads to process assignment events I think it is good to run these handlers synchronously +Patch attached +Still running tests +Sorry to come in late the patch has gone stale +Same issue +Can we only include the commons lang classes we actually use? Also this should only be done for and is it really critical +Running tests again +This pull request addresses and +The problem is due to the before read the thread of writer is already dead +So all transformations using only a DTD subset will failRegards is certainly right that if a is handed in we should avoid setting things like dynamic validation +If I dont hear any complains by the end of the day Illapply it to HEAD +WAR based project that injects CDI beans into +This would require more public apis to changes +FATAL Solr returned an error Error parsing Lucene query version appears to have been fixed in ive added a test to verify no regressions moving forwardCommitted revision Committed revision +What happens if after the check for the channels existence a subscribe succeeds and sets the the channel for the topic? Youll be put in the same situation again +preparing end up moving this off releases the past few times so completely unscheduling. +Since your are interested in this idea I would like to recommend you to go through those issues mentioned as optional and provide one or two patches before you start the project that will provide more insight on this project +is an interface so they can inject a proxy between the interface and the default +Afterward it can be decided what we do with it +Thanks for pointing out this problem +This corrected the and preserved the performance for the issueWhile this approach seems to work fine I somehow have the feeling that the indirect binding through both findType and recast is a bit unclear +Just want to clarify before committing things +in those cases probably we need to change the hints format away from using IP as youre not going to use as keys how are you going to replay hints efficiently? You need to consider the read path as well as the write when modeling something +Well have to get that fixed once we figure out the proper location for the Documentation +Just any pluggable way to receive events from test script suite startedfinished test startedfinishedfailed and test output +bulk defer of issues to +The name of the method was what was confusing me +or you create an equivalent +That was me +The ticket can be in the future if there is stronger motivation for it. +Includes unit tests for proper authentication against the repository and for node path creation when value of the exchange header contains one or more path separator characters +For example it relies on JDK property descriptors which only look for public methods which obviously wont work when annotations can indicate publicness +NimGood point about although theres nothing magic about integration you can always use paramsNo it hasnt been folded into any x code yet were letting it bake for a while on trunk +It is not substantial enough to warrant inclusion in the platform +Hi reviewers any more review comments +Simple t the supplied patch incorrect? It was saying the presentValue and the absentValue for r is false +I did also have the losing contents patch applied at the time as well +Are you usingThanks Jurgen! as invalid as per last issue which was declared invalid long ago +Committers please weigh in if you have an opinion on this +looks good to me +I have attached a patch for the changes I have made +RayeesPlease test it in automation run and reopen it if this issue still exists +You must explictly specify a value for Named if you want to use it as a qualifier +Pl note that Release Candidate will be built on the patch being reviewed +If someone with very large term vector files wanted to test this new format this would be great! Ill try on my side to perform more indexinghighlighting benchmarksMy indexes are pretty big with termvectors taking up a lot of that +If we use it runs well +bug of a bulk update all resolutions changed to done please review history to original resolution type +Rolling this over to +Committed as svn +Thanks Sergey +Agreed about splitting out a separate Iterator and making the Iterable only work with Descriptions +Please let me know if you need anything else +This is working on but it broke some other esoteric block forms +Thanks Mark +If we really want to break backwards compatibility here would be the fix for backwards branch +Ive just tested with a cluster that was not receiving any reads nor writes +Under what circumstances might an application want to call this method +There are really two major issue being solved here new client code and ledger aggregation support +Here is the patch with secured urls on this ftlPranay PranayYour patch is in trunk rev +The EULA has been added to the root of the source tree and is copied into the installation structure when built +So I wanted to check with you how we can best work together what your area of expertise is and how much time you can spend on this to some technical restraints with the current infrastructure we cant use much advanced web technology unfortunately so we are stuck with HTML CSS and maybe server side includes +will need to be added by the user in a separate WSDL file +Same issue a trx is not being used +Works for meLooks reasonable +Im attaching the test code here + +It was released with the fix in +I think we need to kill the ONEWAY state from the interface as it is not being used +Ryan anything doing on this +The component will get the first version number with the first release +Any chance you add the bean configuration you had for Shiro? +Either way the error the died before writing anything does not seem to be an appropriate error message in all cases that it is being returned for +The bug shows here +Merged to +Added a few comments in review board +Fixed on trunk +cool one more thing to remove from my list of bugs to file +works fine +Sorry about that Ralph ill clean this patch up and resubmit +The prior code remains in the file +Delegator is an interface so you cant add abstract methods to that that are common to all implementations of +The directory implementation should accept a as parameter and use it to prefix the index for a mailbox to limit the search space to a single mailbox +I would classify this as a trivial difference the are written such that implementations are free to optimize as they see fitFeel free to bring this up for additional comments on the dev mailing list to see what the rest of the community thinks +This is a more familiar construct for most sysadmins +The system already checks to ensure that the original copy is clean before proceeding why not just use it? for an option to not do anything at all and just build in place +Here we go +Does fix for belong in here Cosmin +Just check the classpath file whether your bean class taking it properly +No test case attached. +snapshot in ZFS and WAFL respectivelyYeah I think we should support this +Could you please include your client code +Removed date facet macroThink this is ready for could someone take it for a ride +If you are using git create the patch with git diff +It should show the possible data format types such as bindy csv custom jaxb etc +As Musachy Tom and Don pointed outkeeping the annotation definition classes in module with retention SOURCE has impacts is needed at runtime taglib code is not extensible externallySolutionmove back to core and set retention to in CVS annotation definitions moved back to core package annotation retention set to RUNTIME +Fixed in revision +So what is happening is the OR rule is eating the expression aA +Do you have any ideas on how to handle that during an updateAdding a separate shardKey definition to the schema would also cascade the change to the get handler which currently only uses the unique document ids as an inputRegarding sharding I look at that as being handled differently +commandsThis seems to be a basic bug which should be addressed by someone +He has been working on this code most recently +If its required it should be set required in the mojo +Easy fix for thisyour patch is in rev +Ill update my changes for the above mentioned ASCII char defines too +Duplicate of +According to brett we need to refactor before we can reasonably tackle this d like to start looking at exactly what needs to be done here to be ready for offline mode +Verified at changes required in +Your input file shows negative user and item good so that is as expected +Not test provided to reproduce the issue +The operations are at the bottom of your navigation area on the leftThe issue with adding pages has been fixedThe issue about the html export requires you to update your template +Could someone please attach a screen shot +Verified at r +Start of test patch +Hi VadirajI will ASAS I will get a chance +Should we also share this using nodetool +I think this is a dupe of patch fixes this issue. +Fixed at +Lets give it a try +contrib tests failures are due to chukwa not related to this patch +target findbugs warnings suppressed +According to Martin this has been fixed. +In my understanding they have designed their stuff to take over any scopes or deployment issues +This ensures that any child errors can be reported on the input CAS and that aggregate satisfy the CM contract of not processing the parent until all children have been returned +Test application having with List of in the deploy can see active on server +Once I added the rest all of them get enabled causing exceptions. +Start the studio disable any Seam validation +is this still an issueI close it for now +sbinI Build by everpeaceI Starting Mesos masterdyld lazy symbol binding failed Symbol not found Referenced from Userseverpeacemesoslib Expected in flat namespacedyld Symbol not found Referenced from Userseverpeacemesoslib Expected in flat namespaceTraceBPT trap +Just committed this to trunk and +I should also point out the reason for this program is purely to test that I can call FOP libraries from a stand alone application +configure no tries to build rb as wellWhen I ran bundle exec rake it complained that it couldnt find mongrel in any of the gem sources so I called gem install mongrel +Patch updated for the fixes +Uploaded patch that makes the discussed changes +PR applied on master and with JBT +Is there not more logging in the WARN? eg a stacktrace or some more details +jms and activemq with different connectionfactories +But I dont have a ddl or vdb file +Made fix for dependencies order +Fixed at least for instances of the styleClass attribute +Removed the obsolete source file from the project file. +I would appreciate any advice you could give me +Imnot sure if this is the appropriate way to do it +We can close this one i guess +Am asking you to look at the code in and do the same from your JMX thingy +Looks good after a first quick testIll remove en from after some more testing +Raised for improvements +See for further details. +commit resolves the problem as expected +Let me know if you are interested and I will change the packaging and get the comments up to snuff +MaxCould you please review +The second bullet describes the original autoloading which was introduced by JDK +Im not sure I agree with throwing a in all these cases +Only one question +Lets not wave away Joshuas concerns above just because they sound like more work +I fixed that in all the samples +Fixed in the CP branch +This is current snapshot of my adapter that Im testing on the application were porting from Access +Maven should be assigning the label addressbook to the project but its attach you file +Minor change to previous patch +Assume closed as resolved and to reopen to remove spam comment +The output looks suspiciously like +This patch should address the issues described aboveIt should be applied from the trunk directory +It would be better to add license headers to the configuration files +Ive not yet tried to debug these +As it turns out this error is totally unrelated to the exception that is displayed is just an erroneous print +Patch to refactor resolvers into cascading you regenerate the patch with option and without the email headersAlso can you explain why maven is both on internal and external resolverCanceling patch for the moment +I should be able to wrap this up before the pencils down deadline though with an example of POS tagging to follow +Thank you +leaving open for now though +So implementing this feature only requires changes to the class advisor class +Fixed remaining issues around das sample distribution Make das sample war files to include source code Make das sample files to have dependencies in lib Remove access to using file set structure in the assembly distribution plugin +If you have two different tables with identical column names and create a query that joins these two tables and select tthose identical columns you will see the same problem +Unless I am missing something +Patch including no changes to as per option above use relative paths in with all open issues this patch I think all above issues are addressed +We are only waiting for the to synchronize with the central. +Attached the patch which just change the package version +There is no compile error +Any change to the diagram will regenerate the BPMN for all constructs supported +I wasnt able to this problem in a JSE environment so Im not able to provide junit tests +Unless theres a reproducible problem we should leave this closed +Yeah I agree with you +The problem I was seeing was that the conversion of the exception data was not handled properly in which case you wouldnt get the failure message youre showingI believe I tried this with the svn code last Friday well after RC but if it works for you now it sounds like the problem has been corrected +Shall we close this or keep it open? +Please let me know if you encounter any further problems +by the way this issue affects all documentation that uses inline source this means all tutorials of picocontainer the documentation on adapters and and +DavidWould Rodrigo and Victor need to add the ASL header and them or is it OK if Jacques or I added them +Any activities should be created through. +The client is running JDK on Linux +svn ci Sending srcmainwebappindexcommunitydownloadTransmitting file dataCommitted revision +See the attached patchYou can change the logging form ERROR to WARNAnd remember to change the message if you do NOT commit theClaus +Hi Brian sure here you are +Hm +Namespace mappings registered through the workspace would not be reflected in the query managers namespace registry +Using this WSDL the error disappears +There are random readers who are randomly reading from all files +slip to any updates +With the release of the came packaged with wagon ssh external +ResolvedThe changes are committed in branch as well +for patch +Assign to during tidy up prior to resolved issues. +The problem was that the default language file was assigned to the and thus was not used for the DEFAULTLOCALE which maps to en +I cant reproduce this with the war plugin and a filtered web resource +Ill commit sometime within a few days +Id be fine with this proposal if it sees as generally better +but thats a ways off still +The problem was that hadoop jars were included before the webapps in the classpath for dev environments +I just tested the nightly build and it works perfectlyThanks! +Lets see if it fixes Ryanss issue +As discussed today please attach a new patch that includes unit tests and is updated against SVN trunk +I have committed this +This issue has been resolved for over a year with no further movement +If there is a problem with this patchthen well address it inSo if you can kindly submit updated patches Ill apply them +Uncached statements are now closed when a connection transitions to STATENOTACCESSIBLE +This should hopefully be fixed with the newer Infinispan and AS versions in +I quickly scanned through all app menus on all components plus a few data entry screens everything still looks the same +Thanks to Max Heimel Marc Hofer Qiuyan Xu and Van Long Nguyen for contributing the patch. +moving to POI +Thanks for the patch! +verified on +QA Closing Code change. +So all issues not directly assigned to an AS release are being closed +Please dont it is not my intention to ridicule anyones effortsAs you can see this issue has been open for some time now and a major reason is that we have never found a good use for an XSD +will do it +However there are some issues involved We need to copy the realm and deploy it manually using Deploy New or may be the command line tool An entry for the newly created realm is not reflected in if the realm is only deployed and not started +applied! +I am working on reviewing the full patch this week +If for any reason I add an Ignore I will also send email as you suggested +I did it when I uploaded the patch +Added patch for branch vGACP fixed on revision +Issue raised for version +Patch incorporating review commentsCan somebody please review this +Current semantics at Namenode is when operation is received Takes the lock Operation updates the inmemory state Adds an editlog entry Gives up lock Logs an audit event Returns response to the clientThat is by the time response is returned auditlog is updatedWith this change it is possible to return a response to the client and auditlog may not have been updated +Verified with IE IE +. +I hadnt investigated compliance with my comment about adding obfuscation +We will be improving the om tutorial in the next release +Starting the bundles in the correct order fixed the issue. +done +Ooops +OK this patch fixes the problems mentioned in the comments +The logic would be given a new classpathdatapath if the resourcemanager existed then use itKey would be that a resource manager instance was thread safe in that it could be accessed by multiple threads +These two should be included in the so that the rpc protocol is fully in protobuf +did you upgrade CGLIB with the Hibernate dist one +ok patch forthcoming shortly +HiI also think these constants might be useful for other stuff where margins borders widths and heights of some components are the same but with different property names +Thank you for the suggestionsThe patch now includes the ASF license and javadoc +Also have you done performance measurements with and without to see how much of a change there isAbout the synchronization problem that is a significant performance issue that I have seen while working with clients +Ill see if I can squeeze this one in before. +Thanks! Its now changed in equals +it sounds like what hes saying is that your Scope is breaking his application even though hes not using your component +Fixed rare potential deadlock issue with aggregate task not being given time to run due thread pool overloaded when running in parallel mode on multicastsplittertrunk . +Actually it looks like buildr still runs integration setup before compiling test classes which I find annoying especially if the compile fails but this may be considered a feature for some +OK looks great Arjohn! Do you have an example RTF doc with bin that we can use as a test case +Will change the error to and mark the worker in error after thinking to it that is probably worse that what we have nowThe error looks like a network problem like a firewall cutting the connection we need a kind of CPINGCPONG in HTTP to fix that +Suspecting this issue can be closedThese arent nightly builds by the way just nightly source tars +Hidden fields values inside such form tag should not be url encoded +though it wont make it into reopen this issue with a test project I can use to express the problem if you have one. +Hi Would you please try my patchThanks! applied +Done +Burr please could you close this Jira if you are ok with it +Since the spec is ambiguous Ill change the test to UNSPECIFIED for element and for the branches but not for the trunk +Marking as blocker and must be fixed for good catch yes the test now passes for me too +Thanks Vinod! +But I think its still worth committing it saves a redundant call to getNameIt was just a small code improvement I came across while digging into the. +Slightly modified patch +Added comments to logj properties files to alert customers to potential loss of log data if the value is changedSending assemblysrcvarlogSending assemblysrcvarlogSending assemblysrcvarlogTransmitting file dataCommitted revision . +I have been persuaded that there is in fact no benefit to using contentEquals +That said doesnt add dependencies all that often so maybe it doesnt matter in practice +Referring to the listener like this works Getting an error when using this ERRORinvalid component definition cant find will create a testcase this weekend and send you the code +Update the Jetty version in the patch into camel. +This represents a major violation of the contract for url generation within connectorsIf there is no such proxy that you are aware of then Id much rather generate a real url which in its raw form would not send you to anything other than the archive itself but which has enough information to be interpreted properly by using the anchor trick I alluded to earlier +since there should generally be only one +I would like to close this longstanding issue +Hi Gavinadmin thorsten ryan gsingers dev olegkThanks +If so the Enum converter is returned explicitly +It has no impact on execution however +Carlo this had been filed as a JTS issue back in January +If the same wants to open new file it will take care of adding client to renewerHow does this sounds to you +This jar is not located in the SEAMlib directory of the distributions +similarly roo currently requires a to explicitly specify a version or it gives a Version is required error +Maybe it would be enough to print out the file that causes a application issues that arent going to make it into to out the files name sounds good but even more developer friendly would best +im on the patch on the branch revision thanks you. +Another workaround to overcome some of the shortcommings in the and ODP is to use a custom type handler +Same here +Almost there! I think there is one more file missing from the patch uddidatasignatureCan you add it? This is a sample signature right? Is it using the default keystore or a new one? Im worried that if it is the default one the test will fail if the user has changed the password to something other then changeitThx KurtI have attached the signature +findbugs are bogus +I would still want to leave the option to stream from the closest even if the strict best node is available +The definition above is not reflexive so cant be implemented as equals in Java +This is how it is supposed to work after have to be careful with bindings and sometimes +I dont think the changes to loses any important functionality from Forrest I didnt see any warnings to that effect in the bug reportBut I dont have a lot of experience with ForrestI ran the link checker but decided not to make any additional changesto the faq at this time +Use Cases of Developer StudioText allows you to view your web applications on a variety of simulated mobile devices so that you can ensure they will be correct formattedCommentThis is a small point +will submit the patch tomorrow morning +Xueqiang is looking into this as part of the groovy web console addition +It has updateLog enabled and the distributed processor is in the chain +which of the above is it? all of them this is the ACL does your only apply to the properties of contentyesi think the check for an item would calculate the match path eg contentjcrtitle and then match against content +I am in the middle of creating the remaining files that I need to FTP over +So I guess we should consider including them in the lib section? Ideally just for the war since theyre already in the binary releases +Filip Thanks for adding this to the Core API Audit this is a very much needed feature and it really should be implemented soon given the maturity of Apache you so much for adding it to the Core Audit +Minor comments Add javadoc to +Patch that addresses the Welcome and Proven sections +stack Any problems with a rebase then commit +Martin +Random thought on this +Ive tested against the supplied example project and all seems to be working now. +just committed this +Thanks for the nice patch PriyoHere is a preliminary review Like I mentioned in our email exchange Im hesitant to make the impl classes public and use them directly in the sink +So needs further investigation. +Thanks for nailing the bug +I did not add an automated test +Patch added +Sorry about the typo +Even on restart some kind of message sits in MS for the host that I havent tracked down yet +Probably because Ive never read the Java documentation and didnt realize I could throw a there +Closing. +How is the trunk patch going for this one + +However AFAIK we are not shipping in any of our products +customer cannot upgrade to SP +Priority raised +Added C implementation of the classes +Thanks a lot +If you are interested please check out the release candidate for same questionThanks everyone +You are right this interface must be implemented in the order managementSorry for inconvenienceRegardsJulien. +is just a module that builds OBRs from the modules in the repository directory +I have committed this +I believe this is because of annotation scanning issue with container +I update the test to use an existing column with standard are all the commands that were run +The current implementation is not actually broken very badly over the weekend I discovered exactly what the issue is and the fix is trivial +From Mirkos attached to his mail to dev list +Thats right +Committed to the docs trunk as svn revision +Yes it is still the sameMaybe you are suggesting that I upgrade to a newer versionoI would love to but not in mood to do such +Thanks I didnt know about that workaround +I have become very untrustworthy of time across nodes +But the version was changed on every module +After adding an includes statement +Rightbq +Thrift request timeout should be interesting as well +Aha you want something like but a hashmap? What is the ratio of keyvalue iteration to regular ops any estimates +Cli you just put in the snapshot version +If not Ill probably back it out and we can revisit later +Not critical moved to + +Verified on work properly +A better LDAPDirectory icon can be used for this think this would be better packaged as a plugin than integrated into the console since the same is true of our Apache DS integration +it still does not allow other characters in the namespace +usrlibcloudagentscriptsstoragesecondary in page the path should be usrlibcloudcommonscriptsstoragesecondary +As to the option to constrain mutations we just control the percentage of rows to update for now +It did help now on Linux after was committed +This should be fixed with and +adds a missing serial to changed oddness checks from x to x ! to acocunt for negative values +thanks for letting me know +I code around it by overriding the jpa remove method and pulling everything out of the collections there rather than using the lifecycle method +New patch that includes unit tests formatting according to Hadoop code standards and documentation +Triaged for +Same as Earwins patch just syncd w current trunk +Thanks for looking into the issue +Closed after release. +Alex Can you please upload a patch that applies to trunk and branch here? Id like to get this done tonight +So since it would pull all of the hdfs reading code I removed it +The problem disapears with +Please reopen the issue if you can come up with a patch +I didnt reproduce that on my machine +SWT Builder is running in an production environment and working made one major change to I changed the package name to and to adapt the eclipse style package let me know if thats ve attached my sources +Great! One thing I would add to update the comment for commitJob reflecting this code is intended for instead of +I will look into this +Fixed correlation ids in changes in coresrctestresources are not intended right? Other than that on the latest patch +There may be a lot of room for improvement which I might have missed +As above could not reproduce +can you explain the motivating scenario in more detail? Is it just to avoid creating the outputstream? Hong you arent going to be getting the schema from the hadoop jobconf youll be getting that from pig +agree with Neil can someone post about a solution +Thanks Derek +Reopened because at least release note is needed for behavior change +You may be interested in voting on which is a more powerful approach to the the filter + +Thanks Devaraj +Yes this is not in scope of A +Turns out AES was specified in which caused a SSL exception that was silently swallowed during job submission +Patch applied +Attached patch will resolve this issue +Actually there is a solution with the current API +And attach it to this and the bugzilla ticket +im on it thanks +A client application has no business of lurking into source code of a platform just to build itself? +Closing. +Thanks for researching it Andy that helped a we no longer even have a custom implementation of in Spring Integration +EJB method permissionsSomething like this +This means we can piggyback off of that variable to figure out where our real landing path is at +We can tackle the script updates afterward since they arent part of the official release +Thank you Bill +New patch addresses comments from stacks review it was just clarification of some comments in the code so Ill go ahead and commit +Removing from and demoting to minor +Another svn patch + Such internal content should probably also not be indexed for searchThats certainly possible +JinchaoDo you have master log that would support your comment about +So why is there an option to set a selector on a queue in the web console +So lack of use of a feature within Hadoop itself does not seem like a reason to reject a feature +Jonathan I can easily handle this one if Sylvain has a lot on his plate +Ive added an additional Configuration parameter called with a default of and Ive changed the max retries from a hardcoded int to the conf parameter +Yes Ive just been making sure that all the tests are passingI will separate the timeout issuethough I do prefer having tests time out with meaningful diagnostics rather than killing the process +When handling JMS Connection failure while sending a reply msg to a queue that no longer exists the cleans up resources associated with an outgoing CAS releases the CAS and continues processing +BTW no need to support AS in the tooling any longer as we are ditching support for AS in +Im ok by removing the date from that equation would that work for you? +Next time I will discuss my question first +I made the change to the as you did in Branch and reran the test target failure occurs just as before and again only for the Java Serialization configuration not the configuration +Juergen +pushing out to theres no patch for this yet and its release time +As soon as Git support is ready at Apache we might switch to it +a backward compatible change is possible because of the version number in the headerSwitching on version number means that the community will be supporting some old cruft for a long time both in client and server +Its been fixed since thenAnyway patch applied +I was just pointing out that this has come up before +Fellas have been running into this in various guises +Okay +bulk defer of open issues to +Apply this patch from srcmainjavaorgapacheuimaducc transport patch from common and transport components are verified correct in the sandbox. +in already addressed same issue +Add index column to table region list in ok can you provide trunk patch +No feedback from user closing it. +Heres the patch +Address reviewers comments and make it actually pass tests +Thanks for the explanations Julo + +Unfortunately no license has been granted to Apache and so a test case cannot be added. +Seems like the patch on this issue would work but needs to have an accompanying unit test to make sure nothing gets broken +This is workaround for Too many files error +This looks like a good change +HiIve separated the issue related changes from this issue in patch But I couldnt find any time to look at Knuts points in his one before last commentIve unassigned my self from this issue since Ive to concentrate on my final year research projectThings to be done Run regression tests on new patch Consider Knuts above mentioned nd rd points Thanks for all the help provided on this issue +I dont have NFS handy to test with however his broker didnt shut down in this situation as he had an exceptionHandler configured to ignore +I hope all this will be solved cleanly with s use of jQuery +This class is in but does not depends on +yes it seems to be passing now +Luke would you please take a look at this +Feel challenged yet? Voted +OK branch updated should be good to go. +Leader is setting cnxAcceptor just prior to setting readyMy concern is that cnxAcceptor will be null from the view of any other thread than the leader given there is no longer any explicit barrier +FAQ chapter is created +Uh +A patch to read autoflush setting from Configuration +Can I ask a quick question? Does the implementation in the look ok? I wasnt completely sure if I had captured what you and Emmanuel wantedThanks again +Code changes look fine to me +Previous patch passed on my system +This is fine for local tests +We moved repos in and we lost the history before then +Yes +Can you verify that everythings OK? I wanted to get rid of as many of the defines as possible for the newer compiler +Uncomfortable removing all these tests +Implementing this requires too much work and refactoring of existing codeThis is only a problem when providing the file name in the endpoint uri +JoshPleased to meet you feel free to call me Bobby or really whatever you want +EL expression language belongs in could close this issue because of WTP is updated +Possibly the correct solution is to make both cases exceptions rather than +Antonio I tried to apply patch to but it was rejected +Added round for review by community +The executor service is the class that I specifically dont want to rewrite +Ive included the missing fileDanielCan you include what error youre getting? As is there should be no issue as long as the pig udf has require pigudf +The reason for failure is that changes the job initialization code in but failed to throw Exception upon init failure +why does the ui need to be different from the AS one with respect to deploy location +Bleh +ZuhayrThanks for the response +Those where not spaces but Hadoop doesnt put the encoding in the file so I didnt know what where exactly +Initial reported problem of the object problem solvedBut more to come so keeping this issue open until Ive got J properly running on a Tomcat tcp cluster +works fine now +I am attaching this test to this ticket so that you can look at it +Log file of with error output +But searching from a repository group only the artifacts from the first repository of the group are being returned +Set the umask before creating the file to give make it group change looks good to me +Thanks! +java beans property editor is no longer used for type conversion with Camel +This has solved the problem +What I discussed with Timo Boehme was the possibility in using an as an input to the parser in addition to a file +Applied patch thanks! +Subversion also needs to ignore these files +Marked Not Required as per Mistys request for feedback and lack of it +This is ok to do so long as the participant is known to be a participant +Will commit it to the branch. +Committed to subversion as revision Thanks Kathey for the reviews and suggestions! +I dont think mixing SSL protocol with bookie protocol is a good idea +Well that should be okay since the jsp file actually belongs to the dfs package +This patch solves my issue too + +Jonathancould you confirm that the optworkspaceumsUMSmainbuilddeveloper +Could you attach the? What is yourColmResolving this as per the JIRA commentColm. +How the mapping is selected Try country language winenus Try language winen Try language enThe picker will try one by one if none if found it will load the default mappings for thought about doing the same for parsing files not only loading mapping classes +I committed the patch and the Solr Cookbook image and published the site +Thx for the confirmation vm will be created along with dns etc +In my opinion do not belong to the +Attaching in a much smaller patch in svn revision +I tend to agree here Im not sure inheritance is a good default and theres already good mechanisms in RubyBuildr to make this painless where needed. +Supporting the full set of allowed characters would make mess of the lexer mainly due to some of them being already in use with other meanings +Strangely enough doesnt check for thanks a lot for pointing this out +Removing wrong query result flag because the case here is a limitation not wrong data returned from a query +Its probably was an upstream issue +Does it need more loops before optimisations kick in + +Thanks. +Ill add links on once I verify things work +looks like plugin is disabled so maybe upgrade failed +Then end users in the future may disable this if they hit some issue +You can consider this closed. +The patch didnt apply cleanly against +Diff against to didnt notice there was an attached patch before making this change myself +However I strongly disagree on replacing the int with a float +Application that tests this problem +Thanks for the explanation +My current Ofbiz customer has pentiums that he is not going to get rid ofThe point is that there are lots of businesses pathetic machines and our market is wider if we write software that will run on them +Helloattatched you will find a file which reproduces the error +Found a related problem where the content type was null when performing a partial page render +so claus moved this task to Camel + be nice to see it +Ill port my test back to to find out whats going on with this parameters +Will update the documentation soon +Initial code to process a service implementation class with an annotation referencing an SEI and build the description classes from the annotations +It also seems that a problem could arise with prepared plans if the visibility flag is toggled to private after a plan is created we do not revalidate on plan execution +You see that a big refactor of REST module has already be implemented in and some apis are changed +I seem not to be able to change the title of this issue +This assumes the following use case for +After playing around with this a bit I was able to consistently reproduce this issue when tapping very close to the edge of the +Oops forgot to svn add the so I can diff it +er that should read its not clear why the system property winning is any more unintuitive than the other way around +Attach a fix +PS I wouldnt mind being able to make a Reference Data method which didnt have to return any named attribute but could just populate the implicit model +Without src it will look for jar and js file one directory up and use those to build example project +I believe this approach has the following benefits Allows any dependencies in the root model without versions to be resolved along with the other resolution +Made changes to encodeEnd render and render to conform with changes to encodeEnd to conform with changes to was indirectly solved when fixing +thanks for the update +It seems odd that the engine would get closed without us having an error pop up somewhere else +We should have done this long ago +Hi AllThis seems to be working properly now since Leos build system updates from a few weeks back +Not sure if this is better +They can just remove the from the list or start w a fresh install when CR or Final come outSo do you prefer the idea of adding a new connector for the stuff +I can help stick in some backward compatible code if we decide it should be there +I just committed this to trunk and +Searching through the patch I dont see being parsedunwrapped out of +Ive translated the messages but didnt change the codes +These schemas can be picked from those but castor is able to stream these during code generation +Misunderstanding of how the perform worked. +This has been open for ages +BTW the version has the same effect on maven if it helps +no other revision +Waiting for Tool +This update is will become a empty value holder when created in the next SICS release +Did not know there was an hosted official build we could use +Then again needs a memory barrier as far as I know so could potentially add a noticeable slowdown +Note if I change the code to always do the refresh i works so its just the logic in detecting if its necessary that is broken +Was the change in binhbase related +by Julie Zhuo closing on her behalf as she does not have Jira rights. +I will leave that to committers who are more knowledgable about fonts. +The test was excluded at r +to hisMaven is really great in reducing redundancies with all the POM inheritance features and philosophy etcI can not see a reason why this feature should not be added +I dont even get the option to select as soon as a I scroll down to Run As or Debug As this exception is thrown before the to select the launcher type is shown +Martin can you double check on a clean install withbinrunand ant deploy +Rebase patch on trunk +This sounds like an to me +Ju will be the minimum recommended configuration for Pivot +Fix for the deleted files problem +Not applicable anymore +I am curious why it just started failing for people the codes been around for a year and I got reports of the test failing here and in Elephant Bird a week apart +In check that our genStamps are always increasing +Not sure what to do here now a binary hbase package can turn into a build environment +Thanks Gang Tim Liu +I have disabled these tests so the code can compile +Attaching patch for this issue +Thanks for the continued effort VarunPlease post your findings on as well +I am inclined to get rid of the default and make getTolerance abstract so tests have to set it explicitly +Maybe then would be the time to revisit this issue +Forgot to note Of course I am using the latest snapshot and not +Archived +Committed to. +looks good and works both before and after ant jar +Lets disable the parser here +The TLD for declares an attribute object which isnt supported by the class +Complete inasmuch as ROCG is available via +Create a branch original for solving the issue +If the person submit the work via JIRA heshe must tick the Contribution Agreement button to upload the work and that is said to be enoughCLA on file is of course better +Thx +Am I overlooking something? I happened to only have HADOOPHOME set and starting up Accumulo failed with your patch applied where I expect to have been warned +It seems to be a combination of RHEL jQuery Mobile +allEnvironmentTomcat should be Tomcat Could someboday help me to update the Environment infoThanks a lotMichael Xiong +Fix a number of compilation problems + Replace implementation with the one from Hadoop but retaining old class from nutch Change to the test because the test assumes broken behavior in +from river where +Any thoughts on this or should we close it as wont fix + of a bulk update all resolutions changed to done please review history to original resolution type +looking +Checked into branch in revision +Can we have someone look into the integration demos +in a replication setup information about a update succeeding on leader but not on one or more of the replica +The issue seems to stem from the +Thus its absolutely not mandatory to inject the subschemaSubentry attribute in all entriesNow the reason we dont add it to entries is simple we have only one schema for the whole directory atm +Patch to patch the above fix +I ommited the code for dynamic assignment of the homePage depending on the user role +that should be attached to a new issue however Im not sure if the situation warrants that it could generate a lot of load on the browse pages just validating the directory every time +The NPE appears to be an unrelated minor bug +The same information is repeated in the name Ideally the name would contain all the info and you wouldnt have to use the data bit here at all but the Solr URL also looks like a ZK path so I escaped the string for the name and also stored it on the data +just FYI EWS CR the tomcat problem is still the issue is still present in EWS CR +I should have asked you how to add tests earlier. +Added reproduction sample file and reproduction steps +All tests passed +Mass closing all issues that have been in resolved state for months or more without any feedback or update + update and new valid potm file +One way to fix it is to have the generate the token rather than since the former is the one registeringunregistering the key and knows when its appropriate to generate a tokenOh yeah that is the correct way to do it +Argh! Thanks Steven I will fix those jdocs nd time around +Patch proposal see comment Starting with Again I think +Thanks +Note that the patching produces not really good looking +It would be nice to get this testsuite passing and make it a requirement for release +looks good +One thing I have considered for making iterators more stateful is HTTP cookie like mechanism +Please it is it shows FAILED BINGER TRANSACTION on Android and it keeps going but crashes in iOS? Im going to bump this back up but I cant make any promises +Ill take this for +If only those where drops were detected are reported why are there the zeros? +Thanks for the info +Is there a different way to perform the required periodic maintenance that nodetool repair gives but with a way of knowing if its successful or not? I understand I can have something parse log files but that feels quite suboptimal. +Thanks Christian +Ive added a FAQ entry for now well return to this issue in the next think this issue may have been fixed as a side effect of other work or changes in this and it works for me too +Test cases being worked added and implemented cleaned up comments fixed bugs in segmentmerger and there be any performance measurements for this? It might be quite good for terms that occur in very many documents an area in which some improvement is possible I thinkBtw for this case it might also be good to use a instead of anI had a look at todays patch but I stopped at because it contains a lot of layout changes so its hard to focus on the functional differencesAre there any index format changes involved in this +It keeps the directory structure +I understand! I posted broken english on a jira before I had my coffeeAs for the patch Im puzzled by the seemingly contradictory statement that adding d isnt incompatible but changing the behavior of ls is incompatible +git repo is current +There is a unit test failure +Surprising that it works on but I suppose the characters are in extended ASCII range +Final patch attached and committed to the trunk +XMLCTF runs successfully again +What if we the specs by hand and then check them into branches +I could probably just loop through the results and get the distance by doing a simple geo spatial calculation but the sorting needs to be in your patch +Add tests to ensure new are called and test for good +Thanks for doing some research on this Mohan +I believe this JIRA broke as stated in +On second thought dont close this JIRAThe latest attached shows the problem without having to introduce new document contentsThe method import in is overridden in such a way that the JVM default parser is used to extract events directly from the DOMProblem is that Sund Apache Xalan parser delivers different events to the handler than does +were processing a good deal of transactions and if were not cleaning up the entries then this seems like its causing the problem were seeing +The issue is that is a derived property from other global properties +Also if we want to fix this we should just put a declaration in name only at the topstruct sockaddrOTOH I dont see how we will eliminate dependency on netinet for something that deals heavily with IP addresses +with the following change make read method of protected for use by extending plugins and fix a minor issue in the script preventing mutliple service details to be opened at the same timeAlso fixed the same rendering issue in in Rev +for that behaviour +This extractor is executed twice each time I put a jpg file into the repository + +bq +This would be good to have +What I applied to to trunk and the smaller patch to. +Tom could you please reassign appropriately +Will put it up on the JIRA asap +I assume the is optimized just to check whether the ACI is valid or not +If the replication factor is smaller than it is set to be +Committed into SVN trunk with r and into the branch with rMarking as complete now that the documentation is completed. +This needs to get fixed it is very limiting that our xhtml completion only work with whatever weve updated kbase to a minimum TLDs should be considered together with kbase and not solely kbase +The task estimation is hr so because youve never done it before Ill expect the task to be finished in Guide Install Enterprise Application Platform uncommented instructionsSVN revision +The change for was lost in a recent refactoring +Hmmh +The attached file fixed the problem +Without your help this wont be easy and hence take longer +Can you always reproduce the problem +Looks good +The logging is already in warning and I cant think of a better logging message +That is the version that I have and the location that it was downloaded from +Im guessing that your session is transacted? You should provide a complete unit test cant really do much with the snippet +Will make Tab view functional when methods are in +I like the plugin must support search in classpath too +see +Out of the box I get a console tab when I start a g +Resolving. +I was running the scripts directlyFor docs its super painful to have to svn up again to bring them back so we should figure out how to just not delete the directory +I checked and the latest Hadoop docs are accurate on this. +The information is not enough to solve this issueCould you give us more info about the error please +Converted to groovy into trunk rev +From what I can tell this is a client issue +Actually I already solved this problem in my particular case by and providing custom implementation of +all related issues tasks were all resolved issues for previous releases +The behavoir has been this way since objects returned from finders criteria etc +is for Alternate resource bundles +If it is fixed in there should be a jira to backport the changeWhen I did I found and quite different in the way it handled packet shipping from the so the work might be +Yes its fixed in r. +I have updated to Ant and yes the build works fine and I have the output +svn ci m Applying my patch from based on Tomasz Blachowiczs original patch and adds get and get sets of methods srcSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +resolved at r on ed up at r +I am not at a computer with my SVN credentials +All seem to check out fine +It would be incredibly helpful if you can attach the other threads backtraces as well +The hive part is straightforward so Ill ignore that + +I dont see any reason to wait for to expose this so heres a rebase of Sureshs patch for with a few minor changes +Kasper +On reflection it shouldnt really be in Seam but in aj +Cool I will definitely defer to you +Closing all resolved tickets from or older +Changed title to reflect actual issue lazy open with no solution might be to synchronize with the and make the call to close in a +Up on IRC Bryan suggests we look into using for Memcache +Desnis please try to reproduce it on +I usually used to profile with echo sample and I am not sure if this call in included in the call graph of echo sample +An apmail person will have to look into exactly why thats happening +Bulk move from to to to +following the reasoning in we add a field and use that to relay the protocol problem since is the only exception class clients know how to read from the server +So then we can throw a guess that the issue is related to concurrency in the new code based on NIO +no no +Volunteers are very much welcome +I do not see this problem now as we upgraded to Jackrabbit which uses POI +Ive looked at test failure +bq +Sorry Marcel I actually fixed a similar issue on Windows and mistakenly closed this issueI will try to get this merged for since it was closed prematurely +I remove the check +In an offline discussion Dhruba made me realize that the fact that the binds to a fixed port is probably the reason why the tests were failing on Hudson +from me as well +Kathey pinged me on IRC and mentioned that the diff for this test was not very clear +Takes some time for the html website to be updatedad contains components with scheme name rng and rncIf you look at the components page they list the artifact as ad Fixedad FixedNote and is fixed on the ongoing release page +Now run under ijYou will see that the ALTER TABLE DROP COLUMN succeeds +Test against current m all for speedIm just not sure I see whywhere the optimization comes in over using existing Highlighter Can you post the client code that does the performance comparisons so I can see exactly what it is you are comparingI believe the performance gain is mostly because iterates all terms in the field noIve tested against a real index we have with theses and dissertations in fulltext +I have been running Jeffs fix in stage for days and the log files have not grown +Before hacking away can we please confirm no more support for Symbian Bada Tizen? the latest cordova drops support for symbian webOS bada tizen +Since the build works Im considering it fixed. +Not sure where is best to place that info probably where we introduce the shell and also wherever else we have some json examples +Bela can you make sure this is solved in next CR? We have a workaround in TESTPING for this but itd be better to go with your fix + +shades of here +Turns out is there courtesy of the base methods so ignorance was bliss in this case +Can you please elaborate on the errors produced? Thanks +Cool +Attaching +Can this issue be resolved +Thanks Tom! Ive removed those in this new rebased patch +Is this patch going to be applied soon? Are you waiting for justification on why it doesnt include new tests +DSP now supports and has been tested on tomcat jbossas and of example has been simplified just need to specify a platform property when building the war +Test case for changes to Service +Translation before presenting the information +Fixed in geotools thought the fix is more reinstating an old workaround the real fix would be to fix the Crop operation in +attached xml is opened as in picture recognized action +Does this also squash +I was expecting to see somethingsee you monday +Btw Im not advocating for keeping just saying that there is sometimes a difference from security standpoint +What we want is Product trunk and Product branch +Ive committed this small patch. +This would simply make hibernate a better programmer friendly product +The previous version located in the wrong package +Such a change breaks backward compatibility in my view as this permission has not been required before +Please take a look at the attached patch +If you finish it before we release beta please correct the version +We can now use this information as the basis for implementing the two changes you recommendedWell use this issue to implement those two changesThanks you so muchOne more thing these exceptions keep occurring after each redeploy of my applicationKind regards +Yes the combined naivety of and mine contributed to this error +Here is the updated patch that includes the +I dont have the details just a copy of the mail from dotster +Will commmit shortly +It wouldnt be a defect but we take performance seriously +Bob can you look into this for the next release +Canceling while Amars comments are addressed +The changes look fine to me +Another option would be to do something like what Jenkins does when configuring a Java installer in the Jenkins management UI you have to click through and accept the licensetermsetc +I checked it and it is not necessary +Ill review this patch +If this is the case then Im not sure why the file is being sent to the compiler +No matter how much the C community may despise a lot of people have a lot of data stored in them for valid reasons so there needs to be a migration path +Committed. +added header job to samples its a job where header from input file is copied to output and footer with item count is added +Assigning to Martha to mark this as resolved when the prod build is complete. +module appears in the dashboard from the previously successful m sorry but I cant manage to reproduce this issue Liya +If set to it will use a permerror +Got from Jay on RBAll tests pass +File guys this is my bad + +Sorry to resurrect this old issue but lists service seems to be down +Investigated this jira +To have this half way efficient we would indeed require a custom set +Added to the SOA release notes as resolved searched for associated by sequentially iterating over all +fixed on rI also added a junit so we make sure the problem does not come back in future +Patch file for DB tests are looks good +The custom resource was intended to define the protocol order not as a attribute as it is now but rather as a field of the custom resource hidden to the user +the original request is to ask for a system property to define context path and such system property is existed already +Added a compilation failure when no messageType is found on an alias. +Any thoughts on this +It is also in use in production environement for over a year without problemsSince Jetspeed now require Java I have ported this patch created for to +Fixed by checking whether an is closer than the subsequent +This seems to be working fine with and Grails. +maybe but I think some places this one is neededused +As a result of we need to explicitly specify the OS to run on for CDH since we currently only support Ubuntu and RHEL +Oups! Same problem here +Verified +How can we get applied the patch provided by JakeThanks in advance +Thanks! +Two different wizards +would have to copy that instance variable over when creating a new FiberThoughts on this solution +There is no change to the actual patch +Maybe between the two of us we can get some documentation on the Wiki and work out any kinks we encounter. +I should have something soon +is the rest of the Summary supposed to be left as an exercise for the reader or are you still wordsmithing? +Committed to +I wonder if this issue isnt just a special case of many many bad things that could happenWith the current approach I actually dont see any benefits at all to bundling the replication process with the kafka broker +It is my connection provider +Perhaps Thomas Diesler knows this +Closing stale resolved issues +Using a filter like Claus suggests in a comment above might be a good approach +Thanks Philip Ahmed +As clearly outlined in the documentation use of implicit joins unequivocally resolves to an inner joinUse the explicit syntax and specify an outer join + +What is the status of this? Is it reproducable by others +Committed to trunk. +When I run your patch it fails two of the unit tests and +bq +I will commit it as it is +Thanks for nice patch Nicolas. +Regarding Installing and Configuring Sun JDK on a generic Linux platform Ive been asked to check if this configuration is supported? And further to that are other databases +This is problematic if you have the following bundle configuration Bundle Contains and exports the and Bundle Imports EMF and and contains a repository and the runtime the creates the transaction and binds an entity manager holder to the using the as the key +Patch for Y which we forgot to upload at the time +Thanks be nice if you could test the latest changes with your application +a bug created by me during the migration from JSP to Screen widgetI hadnt the time to correct itThanks Jacopo to solve this issueMarco +Any updates +is an improved version of the same code against more current code but hasnt been tested due to issues building the latest source from git +You can probably fix and teesink to behave correctly +So chances are they are very well testedGiven performance figure requirements I tested these while accessing the repository using Gnome VFS and browsing around in the repository and creatingdeleting files! So this is certainly kind of +Good catch Todd! Is it possible that pos bytes total in case that the last byte to read does not align at the chunk boundary +Seems to be the same sort of thing again +on the new unit tests +At the moment I not understand the exact problem +All you would need to do +What else are you worried about +Stack and Ted Ill just remove that test +Thank you Haohui for the contribution +Should look better now +You are right Aditya +reclosing this issue +It works in but it may be removed in later releases +Also Im pretty sure acl exceptions were being thrown a few months +If there is no Accept header then the default charset is used +I just committed this +Oleg could you please retry with +IE would not surprise me but IE should be fine +I will look into it a bit more and will file a JIRA about the test failureI did not include any new tests because it is a really small and the existing tests should all pass +thanks for your help was very supportive +Committed to trunk +Wow this is great! Thanks Grant +Applied a variation of your patch locally but with +It appears that due to lack of discussion this bug should be closed as wont fix +So all issues not directly assigned to an AS release are being closed +Veirfied at +I spent some time testing this latest build today and everything looks you very much. +I have added the following javadoc to theFor Microsoft SQL Server there is a problem reading large +so reads on following entry logs will failed even these logs are correct log files +I am going to leave this open until we see a new release of and mark it against our next point release to not lose track of it. +With the recent changes in Lucene trunk we can get rid of the wrapper over +Packages now validate that only rules belonging to the same package can be added and raises an exception otherwise +no it does not make sense to create the listchoice without a model because then it cannot save the selection anywherethe constructor without a model exists to support compoundpropertymodels which are implicitly resolved when usedthe exception you have pasted is incorrect there should have been another exception before it stating that the component could not save its value because it could not find a model +No +Help in resolution is appreciated +good to have this +I will do this in a day unless anybody has any objections +committed +reopened +I am pretty busy but I can take a look +Thanks Freeman. +new patch for Carls review comments +Can add more than this but we do big schema review and update in and can add all the required CLI and Studio tooling then +I can spend my time elsewhereWish you a happy +Having a hard time reproducing I added a couple small changes to trunk +This should probably look like the safemode stuff with get and wait operations +Does it refer to a connection retry in which the server actually receives both requests? Im seeing the defect mentioned here +First how about committing to trunk in order to avoid further confusion on this issue? For validation the current Gora requires because of Avro complex data type support +Minor improvements to the patch +Sorry to hear that +heres the trivial patch which reflects the approach from my anyone that has access to the filesystem is admin +In such configuration the second request used the same session as the first request usedAre you using set to true or to false? +It only affects anyway so I left it. +the link I think youre right its only support for confluence backed sites that is being phased outIve long been in favor of a move to confluence but no real time to do it myself +Kate Thanks for the if I undestand you correctly your main problem is with the keyvalue mechanism +patch is ready +Why doesnt it kill the overseer yet? Sounds like a great test scenario to have +resetting P open bugs to P pending further review +Actually do the resolve QE if you find an issue please reopen with link to the relevant issue +We will resubmit it later after whole testing +Nevermind I found the plugin documentation +Issue not to CR since its not a blocker issue but still a tricky one which may involve work on the that as opposed to what the video shows projectclean reveals the problem markers as expected so we have a workaround here +check as well as +What is in this Metrics column family? Is it normal columns some expiring ones super columns counters +pending Jenkins +However im uncertain if these dojo pieces would allow for this attribute to work +You need a certificate that has a specific but optional v extension +perhaps it is a permission problem? +Get a fresh checkout from subversion merge the changes from the old into it and then create a patchJust make sure you dont copy any old +problem can not be reproduced with the current nightly +We still manage with and only use for TLS connections +With Java based replication being in the trunk now we can also plan for an early release replication alone is a huge user facing feature +In the client code the two identical anonymous inner classes could be combined into one to reduce footprint by havinga private method in that gets the system property passed into it +Since applications dont have a descriptor this is held in the metadata +Could you find out if a transaction is running and join it instead of starting a new one +Fixes issue with the NPE +If you try to upload more the GB the IE will silently upload only the part of the last GB without any comment or error notification +Pull requests sent +Added link to fix of that issue led to the issue described in do not use when is not found +Good spot should have added the amount in the flow to looks good to me. +I think we need to modify the wite method in such that if is true we set maxOccursunbounded for all serializable types +svn ci Sending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexSending srcmainwebappTransmitting file dataCommitted revision . +Since the appender references the test case then the test case instance will never be reclaimed until the logging system is removed and that is when the application ends +I have tested the above on our cluster and they work fine with these changesI verified that has similar issue as mentioned by Jitendra +Thanks a lot for this enhancement please verify the issue is fully resolved as you expected +Aligned and files between and SOA distribution of jBPM is tuned for production deployment and maps this servlet to a URL pattern in the secured area +In addition as the execution is asynchronous an Callback interface is exposed to AM +Please close this jira if it is done I dont have enough information to do so +Committed to TRUNK +Consequently some attachments are owned by root and cannot now be deleted by JIRA +Commited by Dprojectcamelsvn commit message Added filter option to file consumer to allow end user to use a filter to skip files +Can I close this issue +Here is the rest result file +I was not really proposing anything I was just pointing out that it didnt sound right to me to give keys to bookkeeper and expect it to manage it for the application +Its fixed at Apache trunk we just need to backport it into the branch to include it in ESB to branch +I just tried testing with the attached archetype and with and still get the same error as documented in this would seem that the multimodule feature does not currently workCan this be fixed? I would very much appreciate it since I need this feature quite +I closed the wrong issue +Thank you too +Moving these to for now +Its not readily replicatable anymore but i want to leave the issue open for a few days to make sure +What test suite should be run to verify that this patch does not break existing code +Producers each send Poison once they are done sending useful data +Thanks very much +I love spring even if this is not feasible with spring framework +Maybe it would be better to workout common list of describing user property and extend current +If no should the default value of the configuration option be false at least for nowIt can be set to false if the client dont use it in container managed transactions +Sure +Im grasping at straws here but maybe it has something to do with setting or not setting AMMAINTAINERMODE in +By doing that I got better dequeue performance but the enqueue performance is still dominatingLooking to see what can be done +Hopefully this will solve the issue +INFO Running for rows took ms +I think I have a possible explanation of this behaviour +I suppose the way this class is used should not change +for the patch +So it looks like the Log transformation happens in canonicalization phase which is after the semantic analysis phase where the static verifier kicks in and throws the exception youve been seeing +With this update I see and are unquoted and all others are quotedThis is because above those metrics build their own JMX object name instead of auto generating from coda metrics nameWe can force JMX name as before and I think we should to avoid confusion +Or perhaps the better approach would be if the attribute map contained keys as absolute paths +Moved from to just after merge of old branch into TRUNK +The patch resolves the problem and the patch looks good +I assume the classes will be duplicated in with originals deprecated right +I think a fix for this is needed in m +One thing I love is having diffs that factor out renames so I just see the lines that change and not lines deleted and added +Despite many tries cant get this patch to apply without reject on still with the same block to apply manually +Patch and test files +One could switch the configured and HDFS would still work +My words in the specs didnt quite match my intent +Perhaps I should have been more specific what I was asking for is a workaround that can be applied outside Weld without upgrading Weld to an unreleased versionFor various reasons it might be difficult to upgrade +However I dont have a value for the seam runtime +update description with seed value that works with s one possible solution rather than have a hardcoded floating point delta we can use a relative one for score vs +Seems to work as anticipated. +On my box the direct invocation of the delegate takes for iterations the version takes where the overhead is mainly caused due to hashtable accesses +However Flakes use of the MAC address as a machine identifier saves on config file entries and might be adoptedLeft to my own devices I would stick with the current proposal but a more one would suit me as well +My cluster was only one node and any node that launches a container for the new AM attempt will smash the old shuffle token with the new one +The problem with this case is basically the inconsistency that exist with regards to handling the request status for different payloads +As noted in the Description this affects as well as the head of trunk +Ive added another patch to this thread which fixes both problems for us +I will try to review this again tomorrow +Configuration issue fixed that causes this problem. +Franscisco can you look at this before we tag the platform on Juneth or so +Based on this I think we should be checking that we cannot modify a UNIQUE column to nullI see that in the test you have a comment in the form of a questionij Now that B has a null value can I modify it to NOT NULLI think the current behaviour is correct +This will allow us to test the diagnostics tool against the real life use cases +Let us know why you feel that not having a greedy approach works better here +My comment on needs to be resolved too but thats a different issue +With the global inital counters the counter itself can be moved into each of the bind methods +I hope its ok with you +Some of the lines above do not seem to be valid groovy code and some entries were will add tests for each of this. +I am noticing a number of bugs were taken off of the project release and placed back into the pool including this one +ended up covering two CME issues +Thinking about this one shouldnt we move the integration tests to in this case current trunk? Since the intention is create new tests shouldnt those tests be for branch? If we found a bug we can solve in and but sooner or later the work here should be moved to that location anyway and maintain two branches seems to be an unnecessary burden. +This looks good +Fixed instructions in the first comment block to include how to update the reqs mirror and +If class is not found error marker is created +It isnt selectable and list is closed as always after click on it and nothing is put to input +this will be resolved by committed to cvs +Many thanks to the community for testing the patch and to Steve to make the effort to rewrite the quickfix componentI am integrating the patch now and replace the old component +That the patch did no apply is probably because Florian already applied the patch for which deleted the line just below this have uploaded a new patch for the current trunk version +The previous patch didnt work +Seems to no longer be a problem +Sure enough it clears the cache just like you wanted +as Im not familiar with action script code fixing it will require a patch from somebody who knows it better +Those other calls also happen to work +Nice one +I have found a way to make this work +Please set back to me once info provided +me will hopefully be resolved as a separate task. +still the case like logging output to me +Yes thinking more about it in theory in JSF it should be possible to show more than one exception so other alternative we could consider in this case is use publishException only if an exception was already thrown but just logging them is ok +Yes youve been very clear +Do you want to take a shot at itIf you want to work on it together just create a branch for this issue +For now I just deployed it guess we have to wait for the snapshot build to be put in before this can officially be closed. +None of related issues will be worked on for EAP +bq +Scott the patch looks good but you need to generate it with Once it passes through hudson Ill commit +Now catching the runtime exception and then continuing to paste without formatting +released +I speak from Experience +In the short term djencks might be able to help you through this +Hi Permaine please check properties filtering in build process for CXF +r of embjoprcore fixes this +for putting this into just committed this to branches and trunk +Reproducing on AJAX tree +Unrelated test case failure +Thanks Ashish! +Will let you know as soon as I know +Thank you for reporting +I think this is fixed in the current versions of +I agree that the process can be with the retry logic changed +Manually verified that this solves the two issues noted above +shipped with +OK so it seems the whitespace problem is only in which has got pretty bad whitespacing compared to the other sources +trunk fix coming in next day or so +Also I believe various servlet container implementations may require you to put your logging classes in different places for it to work +When million k docs are added in a random fashion the +I guess well have to adopt the differential strategy I was talking about above if we want to keep in both performance and is a patch that handles this the way depicted above it basically falls back to vector rendering mode if the image to be rasterized is too big using pixels as the threshold for doing so +Would be great if this would come up as a debug message somewhere + +The patch looks goodId have to look more carefully but generally by the specification its not illegal to declare parameters that are not used in the query filter +But I may be wrongIn any case its absolutely mandatory that we check this point and if it does not work we have to fix itThanks for having filled this JIRA +Including +I also fixed up and so they are in pretty good shape +Also when I run dist m to build the jdf quickstart root README HTML file locally it not only contains the above summary problem +Axis now attach policies to the binding using policy reference elements and to the portType using wsp attributes +Looks like a worthy addition to as patchPlease have a look +In reply to comment Ok that is the behavior I would it it expected for the properties not to be set when executing mvn package? +Yes Ive replaced but didnt think the same would work withThe thing with is that it is just used to define a cache key for some container but its not possible to enforceLoad from there +The change is simple to do I should have a patch for it very shortly +what is the status of this issue? can we resolve it or this there anything left to do? +Im not committing it as it appears that discussion needs to continue +It would chain to the positions readerThen would hold the thread private cache using this API +We could really use this patch at Korrelate +I dont have a clue about the cause but since corrupted files are index column families I think work around is to remove all those corrupted index sstables upgrade C then rebuild index using nodetool rebuildindex +will all of this code need to duplicatedredone for XSL? That said +I did some inspection with Yourkit +can you give a more real world example +Thanks for the commitSorry Bikas +Is there a particular reason for keeping subcolumns sorted by time instead of by name +If you finish it before we release beta please correct the version +How about CtrlYAlso it could be the same short cut for both operations +It would be really helpful if you could take a stab at creating a test to reproduce the issue this will also ensure that if it gets fixed it wont break again +There is one more that fails all the time when using XML parsingIve been running tests all day and the only one left that fails a lot for me was +Labels in must be improved + looks is the patch I made for for mozilla crew +Christian Dupuiscan you confirm this bug +I actually did this several weeks ago but the unassigned JIRA fell off my radar screen. +Whats the thinking on that +I think other subprojects use Pax Exam not sure what they are doing +Patch for for you create a phabricator entry +Can do will be done tomorrow +Schema necessary for code the code is generated producing the error +Im going to provide a sample and a test project containing the iOS code so we can all be on the same page +lgtm +Multiple catch clauses would be required to handle path exceptions FNF ACE etc +The problem seems to be that synchronization is independent of the locking of the +What about if you use true +On top of what youre saying it seems to me that the scheduled repl queue flush thread name is not the best? pool +Code for reviewSome things to fix are the code style and probably the regexp logic that is a bit messy given my regexp limitations +Hmm +Close after release +This will be fixed by. +I do not know how to know the answer either +As of RC we also ship and classes as a convenience for typical setup in Java EE environment pointing to the default JNDI names as defined in the final specification +Is this jira still relevant? Hbck fixes this no +Added v comments In dev mode we no longer invoke mvn when is not found +Marcelo have you tried upgrading to? Also if are you interested in contributing I would suggest you work with what we have in the trunk so that all the changes we have been making are incorporated +not blocking b but relevant to get fixed +Joao thats exactly my reasoning for not doing it in the component +No rule without an exception you knowAlso having a look at how the scm part works in Maven should be a good idea +Dont see this fail anymore. +Another example could be where there is no reference in the LICENSE or README file but the other licenses are in a top level directory named that seems quite clear and perhaps even easier to spot than pointing to them from some sentance burried within a READMEIMHO it seems reasonable to allow a little flexibility so could the policy be written to allow for that with just something like rd party products and their licensing files must be prominent and clearly locatedAgreed with Sebb +So since it is the next web build I dont think it will be in CR +So I guess both are not related to each otherId a look at the code +Im getting the deployment for the JDBC driver before it attempts to deploy the WAR +its Javadocs sayDeserializes a CAS from a format +Or we need to make this trimColums to true by default in the Teiid +we could consider safe +Its not clear which if any of them is correct +Thanks for submitting! +Adian asked this to be removed he has created to set this as a target for release. +please also fix the tabs for spaces +No worries Larsbq +I just committed thisthanks flavio! +Many thanks Sainath for this critical work. +I tested with C broken down into supplementary uDuDC and the new code seems happy for the numeric entity escaping +I have asked for clarifications andor current state on the labs mailing list +I didnt thoroughly test the attached patch but the code attached should store to the httpcontext when available otherwise it uses the thread context +are you sure you just dont hit the css entry limit in IEBecause if you do it in the main page itself then you say it works +Vasily the commit r does not include a regression test +I havent heard back from the user with information on how to reproduce this +This is a weird legacy thing by Google +Resolving as fixed because is not the issue subtask now +Regression testing prior to commit +Forgot to mention that I am using Eclipse Ganymede JEE distribution updated version +More specifically with this patch a object keeps a list of objects +Reassigning to you for verification Dimitris +Ill eagerly wait to see the results of your talents +So lets release the stable version as and do these changesFYI The correct solution for this is to implement the in our parser wrapperSo shall we agree to fix this for and wait for pleaseWhich means you will have to help otherwise it is impossible +Ive just committed this to trunk and. +Heres a simple patch that fixed the iconv and catgets linker errors for me diff u homembrownetcconfig homembrownetcconfig homembrownetcconfig homembrownetcconfig MULTICPPFLAGSPOSIX mthreads MULTILDFLAGSPOSIX mthreads link with libcatgets and libiconv LDLIBS catgets liconv set the assembler file suffix to +It sounds reasonable to return the same JDBC version as does +I was getting out of memory errors in some unit tests because clear was not creating a new root node +Patch to attempt deletion without caring about success +Delay already is implemented on the javascript side not on the behavior side resetValues is pending on the js side +The distance data does not have to be persisted as you say upon restart an RS would just relearnGenerally do you like this better than option ? would store too much dataAs for +for easier should fix this for too if possible +Victor please correct the fix version +Correct +should make debugging easier +I couldnt access those logs from the QA botS +Confirmed to not be an issue on behaviour seems to be that public methods are callable and private methods raise an undefined method error which is fine +The Entity ClassThe entity class must have a constructor +First thing this weekend is to create a for this thing + was released yesterday +I found the statement that says there is currently no way to delete a user so this should not have been reported as a bug +Committed revision . +since startDeliver could be executed only when a subscription channel is connected it is hard to trigger startDelivery in test thread after reconnecting succeed and before reconnect task call startDeliveryAny suggestion on how to add the test case +Id say go ahead and commit at least to trunk +ASF will generate related comment to hereThere are only one issue about stats Im cooking a new patch to fix it +Now the remote server is actually polled to get a list +I tested this manually by starting up a NN and browsing to jmx +Slightly cleaned up the test code +If this is a Drools issue then please create cross link in the Drools project +Xalan is built inI verified that it passes with Sun with the Xalan jars copied to the jrelibendorsed directory according to the instructions +push to +Apparently is null when it shoulnt be or the null case should be handled in +See for further detailsThanks for the report +The test case +The patch include change to both old and new parser. +re it wouldnt have issues so it could be added any time no need to link it to this issue or to rush it +Can someone have a look at it +set mini to be true so that it ignores Metrics source already exists in unit sure if this the correct fix or if it is sufficient or notRavi can you still reproduce this error +Currently CFX modules are using +from me +Not sure when Ill get round to it but definitely want to at some point. +This could have multiple problemsHadoops Writable interface is hacked in all of our classesSo it would be quite hard for other protocols to deal with this +I installed the latest TP and things look good so far the TP Is resolving fine from a dev point of view +I dont see what is inherently wrong with long lived active connectionsPushing this to as if this is implemented I dont see it happening any earlier due to the extent of the changes required to pool. +Should have been resolved in bugs with invalid fixed for release as fixed in. +use logical addressing via some +I partialy agree +Please reopen if you still see a problem. +Can you post your client code as well +This change takes care of increasing the default heap size for JDBC channel tests +Nice solution now on branch of repoEdit Correct branch name thanks mgoldmann +If its addressed in a pull request please feel free to close the issue. +Didnt see above comment +Why cant this be solved with the Workflow Manager? Im seeing us conflating these two components +targets the visual page editor so lets make this one the issue for XML Editor which also seem to not have enabled the source menu +they are already all defined in actually but they do not get used yet +Ive already read that page and I am using latest SNAPSHOT of +So if a leader is bounced some acked data could be lost +The pdf attached to triggers the for the hint thats a good work around to avoid the exception +The test in still fails +Sorry for forgeting to format the log +I had some time today to look at this again and have determined the root causeIf you enable application scope for your service Axis caches the service in the configurationContext so there is only one instance per service +Ill fix that tomorrow +Ill try and make some time to try out the patch this week +Saves resolved class in field +It looks like the esrver dropped connection during the bind call eh? And it worked in? HmmmI think well have to set up an ldap server locally to reproduce this +public class extends but the package is and you do not have an import That will lead to compile errors +But in HA that should not be the case as explained aboveWhat about others opinion on this proposal +This issue can be marked as resolved as we now have a proper licensed crypt implementation +Attached there is a patch implementing the new line added on renderkitdocs +will load the file from class path +Proposed patch +I guess sometimes I like to state the obvious +To be committed to branch +Assume splitLog failed and the deleteNode is queued up then the splitLog is retried thencreate is called the failed task is still in the tasks map so its batch info is updated and a createNode is queued up +Bulk close for Solr +stop compiling +Block with one replica is still bad not so much bad as at exceedingly at risk which is why its almost max priority in the replication queue +And so youre looking for +Although for different motivation also talks about changing the serialization mechanism for plan +While reviewing the server side metadata code I found another problem +Good stuff I plan to commit this soon if nobody objects +Backported the fix to both the as well as branches. +EAP Seam Reference documentation was as a fixed issue in EAP release notes. +chrome +doesnt this achieve what is mentioned in the fencing paragraph +fixed by a note in chapter +passes +Updated patch to change to +Patch that cleans up all warnings from +Same errors a serious problem I would say +And Jasha has the sameThere were plenty of other changes of course besides the shindig upgradeAt any rate I didnt see this before on earlier releasesTo me this seems like an issue which needs further investigationAnd as were seeing plenty of other issues right now I think we might have changed a little too much in too short a time before the release to call it stable enough +I suspectthat the parsing would fail +Carbrera added as and. +you can already customize the export link adding a targetblank in is the export link in? +Quick question + +Since this is certainly not an technical issue we can discuss to eventually split documentation inlater versions +The goal is to shift into a managed environment so you can leverage the CDI programming model in a native way +I will open a new JIRA issue to track the real cause of this problem +Heres a unit test which demonstrates the NAKACK missing window problem +Add support for getting list composition and column metadata +Thanks Harsh. +I think Im more or less tracking now +Thanks JyotsnaChanges are in trunk at revision Ashish Vijaywargiya +Logging out back in causes them to switch back +Patch soon +Are you sure that making flush unsynchronized will not cause any race? What exactly makes you think it is an obvious mistake other than you being the author of the patch +This attachment shows the patch broken up into three separate commits first two refactors and then the actual parallel sync feature +This issue seems to be fixed +I polished a few minor issues such as removing some not needed logging as well logging to a file during testingThe Apache Mina is already bundles so no need for a SMX wrapper +Len see duplicate JIRA +However the availability measurement still stated available +Thanks guys! +Forgot to attach the + +Good catch +This is an improved patch fixing the obvious thing Konstantin pointed out and contains a test caseWill submit it as a patch +patch is providedits over months and even not assigned yet +Fix included in release. +doh good catch! in trunk and for +I committed a fix to both trunk and branchx after Yonik asked me about the problem on IRC +In the check for identifier is not needed + define constants with enum +I am having the same issue running Cordova on LG Lucid ICS +Can you update the as hudson is not responding +Please take a look +The resolveProperties method could just be invoked every time and just let it +A patch that creates an includedocs test for the erlang views +Patch migrating away from commons logging to slfj +Do you know how this change should affect query performance? Should it improve or will it degradeBest regards Serge Huber +Thanks Aleksey v works +It should not happen in RPC right? I would be good to review RPC code once more to confirm itIn a related note we should think about what backwardforward compatibility for an alpha release means and how to handle it when we find issues +If there are no objections Id like to backport this to +This will ensure formatting has been done successfully +Its worth that sockets will always be supported +Verified in ER +No change code coverage is still at +Travis have you been able to narrow down the problem described in your comment? I have not been able to reproduce +I have compile xulrunner and run vpe on it +Changed fileM srctools +my point about backward compat was about the storage IIRC we dont support backward compat for backwards +I just figured its worth mentioning and fixing if its a deal +I am on vacation from to If you have any question on deployment and JEE bugs please contact Saurabh Arora or my manager Maruthi NuthikattuFor emergency contact me at +If this a recovering master doesnt update the dates the nodes would remain in the previous state which seems to be the desired behavior +Thanks Martin. +The distinction for ischeck makes sense imho +Applied. +The NMS STOMP client is fine to use as your Apollo client library +If someone on the other hand is sure they would like to use SP in their code then theyre likely to need those dependencies anyway +Hi ChatreeCan you please share any article on OSGI working that you are trying to implementThanksVikas +If you want it to be yes I think it should be moved to contribIm also a little fuzzy on how a user would use Instamo +This will require some changes to the although I agree they should be made +Maja you can start looking at to see how feasible it is to use the same ideas from the RPC implementation to the Netty implementation +Anything else we should support +in the Stylish browser plugin +Fix committed to awaiting application to trunk due to release freeze +Hopefully this will remove some of those issues that it is reporting though +CorrectThe Flexible can now be found at +Working patch minus tests at the momentIntroduces the interface like that described aboveAlso introduces a generic that can be used for visitor like +Then bitCount shift and number can be used to determine the length of the trie array and the individual precision steps +If we have immutable properties we can just return a and a clear error message when someone tries to change them +Can you please update? Tx +Closed again now properly recorded against Fix Version as well +install javadocaggregatewhich is what we use for building packages on Fedora +I am studying the version +I adjusted the maxprefetch and everything works fine now with the large messages +Net version only then I would have made the change but its not Java Lucene has the same codeI will post this JIRA issue to the Java mailing list and see what they have to say about it +See my notes on how to recreate the issue +I am able to proceed +Gavin Sorry for the lag +Yes this has been observed on linux too but only on machines +Just to keep everyone updated I checked in some initial code for this +And examples given how to configure this like the old way etc +Did you change the files that were included in both? If so Ill need you to regenerate the patch because trying to handle the diffs will be a mess on my end +The testcase +Things will work better that way +Martin can you please attach a standalone test showing the performance degradationThanks +also fixes subcolumn queries being counted twice in readStats +Any ideas or advice on this issue? Do you need more information? Is there some other place that I should post thisThanks +If the original data lives in a +Alessio + +Removed the hash and compare functions +The same occurs with C client stubsThe file is this file contains all classes need to run the jme application +Doesnt the deployer setup a dependency set for each deployment which includes components? In that case I would expect the apps class loader to have visibility into the classes that a component module uses for things like the message payload +Thanks Marshall Ill get onto this now +Which executes the requests in one thread in blocking mode +One thought that comes in my mind Is it a good idea to remove wrong date inputs? Its obvious when you type in the DateTime field +you make this change on BranchStable too +This issue has been fixed and released as part of release +Mass closing all issues that have been in resolved state for months or more without any feedback or update +now implements the new interface +After doing a bit more research I believe this to be an axis bug +So I tried with the latest snapshot on windows and verified that it works +I took a different tack +I was planning to read up on how this all works but havent got round to it yet +Im actually planning to record the sample from Javapolis and push on the blog in the future so that people can see what is all about +is definately gone from the list does this also mean that there will be no more weekly email updates +Resolved + +So if any other library has such file it can be read instead of yours and probably cause m not sure about what should be done +should hopefully have a fix soon +Patch to add required method to applied and package version incremented appropriately +jse deployment with an associated deployer that simply copies the +Ill look into this issue +pushed +instead of the value +AlexCan you attach to this JIRAPlease briefly describe how you generated the tar ballwas used to generate this passes with the binary provided by Alex +Thanks for the patch Devaraj +I will mark this as resolved +The pdefaultEncoding attribute fixes the issue for Hungarian but it appears to corrupt German unlaut characters +Yes probably. +All of the classes in and have been tagged as deprecated. +Also having a specific shell command for setting it puts a layer of indirection between the implementation and the user +I will create another jira for shutdown issue and look into thatthe commit comments were as follows and As part of shutdown command started accepting optional user name and password +StephenI tried to reproduce the issue using the scenarios you have mentioned but in all tries the VM started successfullyScenario Deploy a new Vm Take Snapshot While the Snapshot is still Backing Up stop the VM After the Snapshot is Backed up try to start the VM VM started without issuesScenario Deploy a new Vm Take Snapshot After the Snapshot is Backed up try to restart the VM VM started without issuesIf the steps you tried were different can you provide what they were I will try thatAlso the is a generic error that is thrown when the VM deployment fails it indicates something went wrong while trying to place the VM and the hoststoragepool suitable for this deployment cannot be found +Thanks Daniel for taking care of thatUnfortunately I just implemented the automatic fixing of version numbers and groupIdsBut its ok I can remove this code again since it might not have been correct in every caseIll add a new version of the conversion script later today which will takes care of the poms and mocks tooIts already implemented but I want to do some tests before adding it to JiraIt would be nice if the Questions in the post above could be answered so I can choose the right solutions for the conversionAndreas +not present in jb not present in +I like your idea about system properties being imported for use in perhaps Ill get a chance to implement that sometime soon to replace this patch +This issue is related to +Patch is in CVS +In case this is not JDK related you can change ur command line encoding to English to see if this helps +Right now we can work around this by building them locally every time +Ill let you know when its there +so of the four i like your home brew +And some more only install a plugin if the source and target version matchesThis concludes my MVP phase +it is interesting that you point to a change in SAXDOM in Apache as being a possible culprit here +New code monitors input lines +This allows users to reduce the size of their process tables without losing their process historyb In general users dont want a process instance to be automatically deleted as soon as it completes +Hi Paulex Verified ThanksLeo. +I understand the benefits of a DB but its yet another dependency + +This appears to be a bug in the Jacob library as of +That sorting may incidentally be handled on the outside is one case and I question that because if cancellation would only return result sorting after the fact is useless anyhow +Need to get fixed if its still there +Is it path relative to or it can be relative to esbcontent as well or it can be relative to the folder where is locatedThe file typically will be in either the root of the esbcontent folder or the folder +I agree the issue as an improvementThe improvement being eliminating the show style in preference for not showing a separator if the heading property on is not setAt the very least having the show style on Form default to true. +Some additional detail information from the nested exception +OK Ill take a look if there is an elegant way around that +In we found that patch but trunk +This seems like the same issue as +Thanks for the review Nathan +A fix has been delayed by other committments Ive had but I am starting work on the solution now +Fine tune propagation and inheritance of resolved FO bidi level to area trait. +I agree with Tims proposal. +So if gwtrun failed or anything related to maven I can help youBut I cannot help you for eclipse issue +The dictionary attached to fixes this problem by setting supports false +one minor change +I am moving this to the Maven Enforcer Plugin project +Patch changes thanks KamranI wouldnt mind revisiting this whole infrastructure to even make it more efficient or perhaps more modular +resolving as done +Thanks JessePatch applied to ARCHIVE module at repo revision rPlease check it was applied as you expected +This patch converts to and adds dependency to in files this should fix the bugs have been introduced in revisions and came back again +not a supersetConsiderselect +So far Ive been able to determine that the content record is created but not the record +It looks like a bug right? +This is not not a bug as such +xml descriptor to return to old logic right +Thus postponing to to to LATER since it is unlikely that well have the time to implement this in the near future there are many more items with higher priorities. + for the to trunk. +Merged into master branch. +certainly thank youI added an error check +bq +Please check commit efeeddcfbbdaadbThanks +studioeclipseplugins +Thanks also for summarizing all the problemsDaryn and Kihwal also pointed out to me that another problem with using create is that the existing behavior of some may be to depend on the recursive create in setupTask +Like Samba and Squid structure. +Fixed on and closing all issues in resolved state that have not been reopened nor commented over in the last month +Applied patch for this but Im going to leave it open for a bit because I think we still need to change some behavior by not escaping attributes at all +A proposed solution would be tospecify a different location through a property where Lucene could write itssearch lock file to +prep you please update dependencies of in Maven repositories? The current version still refers the buggy version of hibernate coreRegards Ollie +Since to test this requires that we be able to mount and futz with filesdirectories at the root of the target file system I put the test to exercise this in HDFS +Closing as requested +Rebased patch to trunk +Here is a patch to add support for find +Closing. +Could you please review the proposed fix that I had made in the patch +demonstrates the issue +What are your thoughts on marking create as Deprecated eventually removing it from the public API in favor of going through the Tool interface +That would be is deprecated in the upcoming release +The changes look fine +I guess we are waiting for ZK changed ZK to version +The should never hang if the server itself fails to boot some services +We will have to deal with the two stage comment authenticator in the next release +subsystem was converted to as result support for expression on path attribute and also most of other attributes was implementedthere was also some cleanup and clarifications done in the code. +Do this book store idea still exist? I like the idea having such a demo system with a real world focusBut the post is no four years old and the links are dead +I know this was committed already and I missed it but Im not so sure are ideal hereThis seems to deal with the case that one cant have two plugin definitions for the same name +Anyway if someone know about that point. +Well do our best to fix this ticket in Sonar +Note that the Logj project has a Flume Appender which can either be an embedded agent or merely connect directly +Ive Ignored these tests at least for the moment +I have changed the project so there is no fferbatim instead of to reproduce the issue +I understand your focus is on production +Glad to hear that SLFJ helps. +Deleting the file from the deploy directory has always been a clean way to undeploy an application and its really strange that now we dont have a clean undeploy +adding a separate testcase to test for with no defaultstestcase tests that a job submitted with a with no defaults is submitted and executed properly by Hadoop as the JT injects its own defaults +Going to commit close for +Code and for not hitting this one earlier David my attentions been taken up with releasing some other componentsWhat was the use case for this +Tested and committed to branch and broken TRB was fixed by it was +What I did was to add the spark namespace and its manifest in the to see all those could not be resolved errors go away though Im not sure if thats the right fix +Patch applied. +Please note in my case the Data folder did not contain any model artifacts +We need to update the jobs tutorial as well +Added patches for removing from code +OK I merged the original patch with my commit from +updated patch to trunk uwe on heavy think if indeed there are valid reasons to have throw away the holes then we shouldnt hide this setting behind Version +This has been active for too long +We shouldnt exclude logj from just because logj is included in its own section +Expect to always get a realm and a host or null for both and just provide backwards compatibility for setting just the realm +Thank you +bq +Regards Caius +I havent heard any further reports of this so Im assuming its resolved +We need to add a lot of new test cases to cover all the scenarios +Attached are updates to the Dev Guide Reference Manual and Tuning GuideIn the Reference Guide the new topic contains a section on Cascading object dependencies +Im also not sure why we cant just post back some type of blob that would be used on the server to create the proper context and register the write URL rewrittersIts been a long time since I looked at this code but it seems like it would be very hard to rewrite on the client without requiring people who create URL rewriter to implement them both on the server and on the client in JS +Sorry I have been distracted by Bronco mania here in Denver but now I have my tickets for the game so back to solution makes sense and I am sure is the best solution for now +Resolved by +The new file should be added svn add workingclasslibsupportsrctestjavaorgapacheharmonytestframeworkutilMay be it worth to look into Ant macros to make a solution suitable for all modules +This issue seems to be solved in beta +Modification of previous patch so that the schema creation syntax is delegated to the dialect object +I think the current test case also provides better coverage because we are checking the environment variable in the child tasks +Moved SVN location from xmlsecurity to santuario rcreated mailer entry for commitsDist area coming nextThanks for all your help! Everything looks good there are just two things I want to clarify New LDAP based unix group and committee group created +This is a patch file based off of trunknot fully tested patch file +This patch adds support for the handling of special characters names +Happy enough to take the risk that nobody was using the old bean lifecycleobserver mechanism. +Committed. +Committed revision . +Attaching +Marking for so we keep this on radar for the upcoming release +If it does cause a problem then we can always release a fix in +Its now +Attaching a patch for the test jdbcapi +attaching screenshot +Apparently java has to resolve all possible symbols or references when a class is loaded even if the exact code is not being exercised +If the carrier put a custom rom we cant replicate this issue and Ill have to put it as a Cannot Reproduce if I cant on the Sony Xperia Mini as well +I tried onkeypress event and everything was I recommend it as a work around + to remove them +To avoid the copy we could instead store the checksums in the reversed order starting from the dataStartOr is it good to interleave data and checksums +A screen shot of the in support for Console Navigation and Portlet header +The factory method on does not require security since it affects only theMethods to addremove multiple members of the have been added +Just in case you would like to test it again long do you guess will take to get those bugfixes into the GA release? Which one will it be +closing +looks good thanks Bobby! +I hope this is the end of thisThanks for looking at this Ram and Ted +As Steve mentioned I investigated the change in more detail and narrowed it down to just un that mathMagic method +It should be also noted that this works fine if the are page based rather than panel based +When run as a standalone web application this issue is not actually resolved this in the beta release when it was hosted +Marcus this should be a simple job for you +From a community point of view the fix for this is coming in next CXF version +Hi +It was because of the null connection supplied to the WS translator +Issue does not appear in TomcatJDK or TomcatJDK +Possibly +So I suppose this issue is no longer applicable to the branch trunk +Marking as duplicate of bug that suggests a plan for replacing the locale attribute before +removes mkdir parameter for Unix ifdef for crossplatform mkdir ugly but it worksSorry for breaking the build with my previous patch +But how configure Spring to enhance code I havent got succeed +The can notify a listener when a fatal error occured and the connection should be discarded + +Easy to say we can come back to this easy not to +push out to +for Daniels patch +I cant move on this until we have more information about how publicprivate sources are to be marked and at what actual we have specific information on that we can create a java annotation processor that will generate a list of files to be processed by the javadoc tool based on some annotation or other can help with this once we have more information but Im not plugged into the deadline discussions for this work so Im mostly unaware of timeframes being discussed for having this tooling free to back to me when we have more information +This is being rescheduled to save time for other CXF fixes required to pass the TCK +Test patch +I may not find the time to look at the patch before next week though +Verified on work properly. +Can we not pass down? Can we pass a narrow Interface instead +Ive attached a test case that never finishes +Remove it now? Good removing tests tooI took a look at test failure +Note need to make jpaconfiguration use the to override any JEE connection settings in template patches to generate crud files from used seam release as reference +update we can put regionoperationID into the table and enforce new updates region operation ID has to be greater than the previous version for a given region +Set the bindingKey to the empty string if the key is null or is null +David Sean Taylor wrote Another solution to consider loading User info and preferences when the user logs in I dont think that will be enoughConsider an external portlet in which I want to displayreport certain User info of several users +If the client side methods hash is the same as the server side methods hash the server does not send the list of methods back to the client +If this feature is implemented would it allow for importing using eclipse project dependencies so I dont have explicitly create config sets that include configs from all other projects transitively +HiAs I can see your installation path contains whitespace so this is a well known issue Could you please use Users Mailing List before creation of issue in future +However the number of reducers based on the above cannot be more than the one specified in the configuration parameter +Fixing it seems to require changing the behavior of implicit and explicit casts +patch looks this is a bug in DFS it is likely to affect other users as well +Heres a simplified version of Solrs +Id say move it out of but keep it alive at a low priority +wrong status +By the way some of those changes were made via my extension which would definitely not have any validation aside from whats in the catalog itself +Pavel what do you think +Patch with stream buffer as suggested +changes to trust claims and trust rst in processing trust namespace errors in trust claims dialect creation corrected. +Thanks +Not sure what is going on you try downloading Eclipse directly from instead of using the package managerYour feature version is different from the one shipped with Eclipse +This can be added to the vendors corner on the Workflow Patterns website +Well it has been like this since several yearsIm not sure whether this behavior is correct +Resolving this issue as InvalidThe library is a dependency of Xerces thats a dependency of +However to NIO Connector Same exception happened +for me too +Attaching management server the bug after validating the fix +They just use fake RM to simulate something in the test +Yes these are WTP proposals +Would anyone object to my making a branch for this +Custom karaf distribution with axiom and axis features installed +IPC template patch for our website records +The result of the code that is in the bug description can be seen in the attached PNG image +HansWhen talking about iDeal you are talking about a different set of functionalities +Thanks Chao! Looking forward to it +fixed the problem for me +moved the name to be the first column applies initial table sorting on the name fixed to catch exception when getting the for providing the patch +Hi Luke could you please take care of this +This looks like good changes to me +Closing +Second patch only on still limited to Maven managed to do some additional tests and it seems that my patch also fixes +So its not dead weight +push out to +Contains unit test to reveal possible bug +Thanks +Another point +Downgrading to resolved the issue +I fixed it in trunk Dan might backport. +Thanks Renato +For UI like the help output from a command can we spend a few minutes making the text grammatically correct before checking code inThe owner and group names may only cosists of digits alphabetmay only consist of digits alphabetical charactersFor future reference anyway +I assume the problem is fixed +SiI looked at invoiceItems form +Implemented Robs fix in this +committed this. +Is this really expected not to work when the numSegments? Thanks for the explanation +I do not know what their other operations that will require? Or are there only two casesIt would be good to have a unit test demonstrating the failure +This would probably be more difficult for environments where bytecode scanning is not available +In testdistload data infile into table studenttabkshould be updated toload data local infile into table studenttabkmysql errors out with the first syntax and setting permissions could not fix it +Screenshot which shows currently supported server profiles +Changed to feature request since the jmx timer routine that guaranteed this behaviour wasnt available fixed ratewasnt available until fact that s timer already had this guarantee is a quality of service issue. +Jaikiran this is really up to BrianJason it causes failures in CDI TCK. +The error is generated when creating the doesnt seem to appear upstream as Jozef says +I will commit the fix tomorrow evening otherwise someone else can do it in between +We have been using the patched variant ever since and are producing our product documentation with it day out and day in + +Lets make some subtasks to do the id first and sorted is a web console issue so I changed the component +I guess somehow the synchronization between the page file and memory container info goes wrong which results invalid page entries on the disk +new in JIRA is using this hash framework and will benefit from better hash function broken murmur hash as just tried and its way slower +Ive done the import and mailed lucene to see if they wish to proceed +where is the patch file? Because in this error already ocurrs really +Switching all issues that have been in resolved state for more than one month without further comments to closed status +But no one is arguing there that we shouldnt share as much logic as possible rather that discussion was about how the logic is sharedFinally I think its okay to throw an exception in the client when the configured blocksize is not a multiple of the configured bytes +admitting the fix is easy does it make sense? It should never be null at that stage we could be hiding something else + +I like it so farThanks Shinsuke +we target this for too +Hope that it will fit everyones needs. +if noone objects Ill apply the fine be my guest. +I havent seen this capability outside of basic research prototypes +Any chance of a unit test for this? Seems like a pretty easy case to test +Else +and committed. +Committed on Nov . +Ill give it a run this afternoon. +The decoder is probably just doing the best that it can with this but the result isnt going to be what you want in any case +That makes sense +grep r commons +But Araceli find the problem when viewfs is on though not using any viewfs in the query +Yeah I saw this myself as I was writing the code +Joey Thanks for the clarification I agree for RHEL youd want to enable SPNEGO just pointing out that there are people using RHEL KSSL today +I thought about it Id rather wait for JPA to define the value to minimize the migration pathI like position. +verified on! +Note that however if a new conflict for the same module is encountered at a higher level then full conflict resolution will be performed at the level of the higher module in which case youll still have to add the conflict management directives to the higher module too. +If the filter doesnt match because the property is absent the rule should be skipped and things should proceed normally +IlyaPlease review and comment +And while the configuration and startingstopping of brokers is different the general principles are the same so Im a bit wary +jbpm modifications for SEAM team to verify if this customizability of the job executor can be configured and used effectively +From my experiment this seems not a big concern +Added a subtask for the JMX stuff feel free to add others for other possible improvements of the pooling lib +I have a draft patch that looks at the namespaces used in the element name and attributes passed to startElement and automatically calls start for any namespaces that have not yet been declared +I think this feature shouldnt be documented till it works +tiny change committed to trunk and mouse caused incorrect provide any steps to verify this issue. +Sometimes but not always a newly generated EAR project will have validation errors in +This +That is one of the errors I have hit +add tests and validation it +The proof will be on solaris though this is reproduceable on solaris with the original test setThanks Mahadev +Deferred to +Replacing the previous repro attachment to add some more detail +fixes the parsing of the with this patch it should propagate the errno correctly +Currently I have only example PDF that uses this operator +Yes some people can change the section that they are editing by changing the URL but surely not that many would do that? There again if the return if based on the section that someone is editing currently rather than where they came from then it might be easier to just go to the top of the current sectionSorry mixed message here +Will check for other test casesCan you please list them down? Which of the test cases failedThis issue is different from the earlier issue that was present now resolved accountname should be passed with domainid +Need to follow up with Sun on the status of the issue filed +Downloaded yesterday and tried it out this works great! +Our web application accesses the process definitions and forms from guvnor +Encoding issue +Looks like that when Solrs synonym parsing was moved to the analysis module it was also rewritten introducing escaping bugsExamplesaa is no longer treated as a single tokenaa is no longer treated as a single tokenata is treated as ata instead of containing a tab characterI didnt do a full review so Im not sure if there are other differences in behavior +I dont know why I didnt find it as I always try to search for related issues before creating new confirm that this bug can be closed as a duplicate of + Username groovesoftware +JimmyThanks for your reply. +Hi Robbie Ive reviewed the change and have no comments. +Add +Thought about it and am still investigating some bugs related to topic metadata request timeouts and large number of topics +I cant imagine something from still being relevent +with msvc +If it works I would find it to be right handy. + +logging SPI was ignored because not more +Patch looks good to me +This patch does nothing except add a from one topic to the otherM srctuning looks good to me +I used to stay in contact with Bob Lee for such matters but he stopped answering my emails some time agoIm Bobs successor on the Android team +It does share some code which is why it does not have the compress bug +v rebased to trunks patch which applies on top of +I came to know that patch has fewer failuresStill working on similar issue for ve test case +It will not take long but it is getting late here Optionally I would like to open a ticket and attach a patch to fix Apache Hadoops daemons priorities as done by my boxgrinder appliance +Dag thanks for providing how the query in question is by walking through different derivatives +So a separate issue would be better +Inside the archetype given as an example the files contains directory as child of fileSet where one only is added a log which will appear in the next version of the plugin as the result of the xml parsing of the metadata file +Oh thats good news then +we should drop the stuff we have an get this in shape ie + +I just cant understrand how this can happen +I think that there is a bit more to supporting wtp than is dealt with in the file referenced above including updating the file within referenced projects in a build +Committed to master +I wonder though whether the method names should be renamed to show the closeness to a stack better +I will open a new JIRA which requests a policy of addingremoving blocks from failure is not because of this patch +report back with testing results shortly +The agreed approach is to check a value in the database that signals interruption so using aspects no longer makes sense. +We can always change it after we get this feature in +I removed these jars and everything came up with no problem +File upload example with RF file attached +We most likely need to make this work for any channel with a dispatcher as well +Can you add a sample project showcasing the issueI strongly suspect theres an issue when we try to check if the identified dependencies can be resolved +subversionconfig was not configured in my environmentI set attributes to those files in next patch +To me the solution doesnt look generic enough because the observer doesnt get any information about which module has been engaged +Can you do me a favor and link this one to and do the same for any other validation problems that you findAlso I think we can close +Your question is not clear to me +Is it ok +Im not sure about an transition seems to be pretty elegant and most people are happy with it + Developer id date Can not reproduce issue +We can use JNAJFFIFFI but then we lose functionality in environments that disallow native there a change we could make to fileutils to handle this better +im not sure what sort of object is returned by the client call but it seems to work just fine for getGreeting +I assume you got that sorted out +and ESB needs to work with JDK +Yes it is and the issue I was thinking of happens to be. +Thanks Arun +I guess I missed to replace somewhere in our build system the version +This issue is correct +And Ive committed now +Given the way the reconnect delay works it might be simpler to just have an option to warn after a certain number of attempts +In my opinion things such as Utility class and everything in +Hi Ronald did your production server use a fresh install of postgres or was it an upgrade from previous version? In the second case did the issue still occur after projectsThank you +Committed as svn closing. +Committed revision +how does your POM lookWe successfully do this in so I think this works +Also added one more check for containers belonging to finished applications from reaching the schedulerI also added the which fails without the patch and passes with +Ive added another patch that works the way Gianmarco described +This is because with depending on the value of the config item and the number of task failures at the various trackers a tasktracker can be blacklisted across jobs sooner than later +Will it continue to derive from? It should not IMO +Reopening in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +is this going to leak streams? +Retaining smaller buffers can also cause a problem depending on an applications memory usage characterstics +Both of the return null statements are removed in this diff am I missing something +The patch looks good to me but just curious what is the advantage? The other way it would have a field initialized in construction +Postponed to to Brad since hes already got and +And I think we should backport DWPTReally? Its a fairly large restructuring +I merged the patch to. +Closed all the resolved issues +pasted +Sorry +Applied in r. +Thanks +Im wondering if this might be due to the fact that one of the nodes in the cluster is currently down +At this point it looks we should take the shot of morphine andmove on. +Can you run some numbers? Also maybe reordering if statement conditions can be helpful +User and group modules are bad +Maybe it was from Ecommerce +Interestingly a patch went to production to address that exact issue but had to be pulled as it broke several consumers +Applied in commit Manula +It can match more documents than the standard WITHIN will match for those cases where the indexed shape is comprised of multiple disjoint pieces +Now we are merging docs manually from click trunkMerging docs will not be automated because we have to maintain. +This needs to be investigated +The BG thread priority is not finding its way down to the parallel threads and is causing nightly build to sometimes hangIve disabled the testcase for now +Hi Markus could you please provide a sample maven project to help us reproduce this error locally? sure because the project where the problem occurs is of a customer but I will try to put something together +Guys it happened to me as well the workaround works well but to get to the workaround you must kill your brains searching the net +I just committed this +. +was trying to check the default behavior in in case of these values which took some time +Jacquesis it possible that this is happening just because the POSs store has the is flag set to Y + +It is only available in the the version +Allowing the exception is the right thing in that case +Hi Rob Can you review pleaseThanks good to me solution has obviously already been tested as Windows doesnt properly support half and approved for +any other sensible default converters to add +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Heres a fix +Because it will destabilize branch +Patch that fixes both of those issues described in description +fixed on master in commit bd +Yet we should commit this regardless! I would be happy to make additions to on another issue as long as we havent released this code we can still change the API while I dont think we have to +But did not do any tests +Michael +Im no longer hitting this error +Attachment has been added with description Enable configurable PDF has been added with description corrected some a few mistakes javadoc in Added to the has been added with description Dynamic PDF versioning has been added with description Dynamic PDF versioning support +Unfortunately there isnt yet an easy way to do this from the web gui +locale is a variable and is available only for the current uirepeat +I have committed it with some changes +this patch activates the TCK on +Defer all open issues to +Have a proposed patch? Doesnt look easy I dont think is necessarily attached to any so you would have to trap it lower and rework the exception +The correct patch is +verified in +havent actually tested that though Add class makes sense to me and yes the lookup should probably be limited to the project but then the wizard needs a project field the New ORM metadata then we got another jira for that not sure if they should be merged into the same wizard but at least the flow should be samesimilar. +but I have a difficult other you +The Lucene project has published Maven artifacts since the release +I use de download +Perhaps we need a build job so that the QA team can do that type of testing at some point as well? Just my cents +Waiting for merge with branch before to close as fixed +no objection +Added asterisk to classpath in Git ID dbddabbdfadde +I can update the patch with suggestions +When my patch no longer applied to trunk cleanly I realized it had been fixed! +If its an we absolutely want to log it +Initial patch sorry about my formatting I havent got checkstyle quite playing nice enough to just reformat all the time correctly for meThis patch has an which implements the iterators and map functions as well as faster iterators in and +Changed the input validation to only do basic input validation +I think its best if we continue to require a FBS so users get a clear exception instead of silent performance hitThat seems reasonable +Hi Alexander +bq +To be refined for RC if +I might be able to perform tests on a Linux box in a few days from nowI might add that we are using JDK for the momentNote that if you put some load on the server then Jetspeed will start and run +Thank BrunoYou point main scenario of this JIRA +The days logs from the cluster hope thats helpful. +The reason the follower logs are so much smaller is because they are batch compressedNot sure what the best thing to do here is +sigh +Sorry +But if we just get a fix in for the bug first with a documented workaround for those who see an unacceptable performance hit we could always revisit the issue to improve the performance for the holdable case later +But I was trying to get the file just with thewith out invoking the action +v addressed chunhuis to know whats the exact javadocfindbugs warnings from jenkins +MirekJust uploaded the patch for +Version Released +I actually got an OOM at some point but could not reproduce it at all in subsequent open a separate issue for it if I stumble into it again. +Tests did pass +Jira migration bug. +Itd be better to convert the deprecated usage instead of just adding suppression +bq +This was as important as any other valuable hint +It should be IMO All our AS setupis based on a configuration not good practice to install in the unless the user really want to +For testhackwhatever the jvmRoute system property can be set +but I think that this can be done up in the tool using common sense checking if storefiles have references etc +committed. +libcassandrabin +latest iteration is stable on my machine +This changes the int argument to a smallint argument in the SYSCSCOMPRESSTABLE system procedure bringing it into agreement with whats documented in the Reference Manual +Jakub do you know where could be images not freed from memory which leads to this problemIf not we will need to address it later and for now run with smaller test suites +Committed revision +There is no in +Except the two that are marked since Struts in +So its as optimal a strategy as anything +Increment sum them up on read +RonCan I get your insight into how you think this should work? Ive been chatting with Sergey a bit this week about this trying to figure out what it would mean to support the continuations and how it would work where the create continuation call would be put how the resume occurs etc +Ran through fetch cycles +I keep the logfile validating but remove the layoutPattern validating because I cant find where its used +I think it was an important addition to FOP +Ive attached a test for this feature +I remembered to put the license text in the file this time I swear. +You may also want this on EIP +At any rate anyone using toString in productive code must be mad +The last patch is just fine +Put up a patch to disable this test for now +But now I dont know how to configure remote also tried to convert the General Project to a Faceted Project with the Facet Java in version +I just did a simple test on the camel trunk the endpointUri will be set with the endpoint name like my +if email field is empty +This attribute works only if youre the first member of a cluster +The shell script looks good and does what we want +Typo accross across +Patch based on the tag +I used what has been in the patches so far +Thats a effective data loss +The first is which reflects the changes necessary to the Shared +Could you attach a simple test case that demonstrates this behaviour as Ive been unable to reproduceDoes this still exist in the latest CR release +on patch v +The spec clearly says to return a mutable map which implements the entire contract of which inludes putAll and clear +I have added unit tests for Manhattan and Euclidean distance review the codeThanks implemented the times optimizations and an associated unit test that demonstrates a improvement when used with cardinality random element vectors typical of text clustering applications +defines the wire protocol that used by the while HDFS might have a dramatically different implementationI think it is okay to go with the current approach and to revisit it when we start to implement the logic in the +For custom components it really shouldnt be much of an issue because the highlighted template will show the users problematic code +Sorry Kevin can you refresh one more timeIt is not applying cleanly any more +The word encounters should be deleted in this message +With version Ive removed the encryption testfiles I forget yesterday + cummulative patch only the git mv part code changes +They are custom generated using a tool I wrote +I will add a seed and see if I can find one that induces a failureTo avoid confusion +Memory has been fixed in patch +right now it ignores order since says that exported packages from required bundles should not overlap but perhaps this is too optimistic and will need to be more strict in order +Im setting this environment variable ANTOPTSXXm XmxmThat works around the issue but requires configuration of the machine your running on +Ill leave the refactoring to you since you have a better view of the high level stuff +I am trying to find a way to generalize it a bit as mentioned +Thanks Andrei +completed and confirmed +and this will fail with some of the test case where it sends multiple requests per route +fencing command ensure that prior can no longer access the JD after it completesFence command will include a version number that we got from the ZK node +Attaching test failure fails first attempts and passes in thrid attempt +Try the nightly build changes relating to Try the nightly build changes relating to Bug may resolve your issueNiall may resolve your issueNiall +I mean it does not help you much but this sounds not like a groovy bug then +Expect a new release in days. +Please try to apply the patch again. +Thanks Hari +Any more thoughts on this? Otherwise I am going to apply this patch before I forget about it +It stillhappens months after I first used Struts +And so long as the class name does not start with Test then theyre not picked up by +StepanEverything works wellThank you +Seems quite serious +Hi Andy the final date for this document to go to QE is August +We used in the earlier version a db filesystem so that the locks were stored in the database +Then when you check out the ones you want to change they are set to writable +for to trunk. +Andree The selection weirdness is still happening to me is it supposed to be in the Beta or should I file an issue for CR +Some closed some still open +Depending upon timings a second timeout may not be enough time to allow recovery to be performed +Old VC files at the top level have been removed +If you use references this change is fundamental +Whats the output of mvn dependencylist or mvn dependencytree +Attaching patch fixing the issue +Thanks Vikram! +Attached is an initial patch +I will reopen this if I ran into it again thanks. +added check for hdfsWrite return being less than and if so break out of loop +As the base URI when calling the resolving function in the Sources he uses a base of null +It seems reasonable to have a UUIDBYTELENGTH constant in itself and set the RPC constant based on that +David this needs to be a JIRA on MCHANGES at +After a restart of the openmeetings application you can login and use the app fineIt really seems like we have a memory leak in the installation process maybe something is loaded in the memory but never freeed +It should also patch cleanly on a fresh checkoutSorry for the confusion Avery +Tried using Tomcat Context element antitrue without success +However added descriptions to both so that it explains it is about plugins +My mistake Im fixing it now +gccchecking for C compiler default output file name +It is no longer the case in the branch and onwardsIs it ok for you to close this ticket then +If not I will note it as a known issue +Which curly brace looks misaligned? They all look ok to me and I had eclipseI can fix the line length it wont match anymore but thats ok + + +This is a configuration issue +I should note that the workaround is to not use a field named data +I am unable to deploy generated Seam portlets to EPP as well +Okay it turns out I also need to publish the forrest changes I made +Just tried this using the SNAPSHOT libraries with the attached model and got the same result +Changed form method from GET to AbdullahYour patch is in trunk at r at r +Perhaps most of the patch can remain but the part that handles how a table is represented as enabled in ZK can be reverted or worked around +Youd want to either use s or your own LRU style system +No worries +FWIW this patch makes the docs target build dependent on perl being installed and in the PATH +One thing is missing Set the Wiki in a mode +We had made require only Java simply for broader compatibility +This is managed by JCA so it would appear that JCA is setting a TCCL and then not clearing it +Should be with modello stuff in it to a later release is fine by me +For the contents of the file you have streams and reader +You shouldnt have to explicitly stop and restart an in order for attribute changes to take effect +running the thread at all seems like a recipe for unexpected behaviorOr are you saying you to temporarily disable the syncer thread while a bigger transaction is in progressIn our case we didnt want the to be there at all as all the Mutation to all tables need to be sync immediately +Will let you know once I have it workingChandraHi MaartenFind attached the debug log for httpclientThanksChandra + +Committed which is better than the current state but not perfect yet some integration tests still fail in working on that +Its still not clear why was putting stuff in the local filesystems +New patch +Verified in JBDS +Patch applied in deprecated in favor of the felix plugin anyway +there are many reasons why a modal has to be inside a form tag +Instead it will monitor the folder for changes +Robert Can you do performance tests with the old and new flex patch I do not want to commit beforeUwe I will run a benchmark on both versions +Theres +Otherwise shutdown paths have nothing in common for and CtrlC +When this happens the temp files for the task are never cleared +Will reopen if seen again +Patch added grammar change bug fixes etc +But I expect more when http session is tested +I will look into solving this tomorrow unless Braden beats me too it of the things we discussed originally was to embed the JSON right in +Got it makes senseWhats interesting is that CDI made a point about introducing a new type of interceptor a decorator +I think this one fall under the same edge case +Still I am not so sure if the new config should directly contain a +Ah +Shay as far as and it seems more powerful to just let apps create a new query type themselves? Thats the nice part of lucenes openness to user query types start with the code for and hook up your own caching logicYea but it would be great to know when an has decided to actually close so caches can be eagerly cleaned +Blocker status should be reserved for issues which prevent applications from functioning at all +BOOLEAN to VARCHAR FOR BIT DATAI did not commit the documentaton changes +But it will not make todays release cut. +Example visualized dot file based on pipeline plan the integration looks super cool +Switching all issues that have been in resolved state for more than one month without further comments to closed status +And I agree with you about the failed tests +these are new tests so didnt exist by GA was fixed by hardcoding the tests to run with JDT since they were JDT specific test cases. +Maybe hacking the order of refresh could be the trick? +This sounds ok to me +Fixed in rev two attachments that were screenshots of a browser viewing this issue with no comments or other information. +which means that it wasjust a forward not a redirectas for elements are beans storing the configuration yes that wasmy assumption as well this is why I see it as a problem that they do notrepresent the setting of the XML config file +Futures look syntactically nice but tricky to work into an explicitly coroutined IO model without additional support from something like or kind of rules them out for shipping productIt looks like for my application I may need to take a similar approach to Erik Bernhardssons +This is the same problem that is described in issue +For different return type you could take a look at the and look at how it defines +Looks fine +IE version But there should be some way out +Hey JukkaHow is this different from theThanksCheers Chris +Better ask in the mailing lists of Tomcat +New issue? Or can we cover that under this +Just for your information weve run into this issue with on our dev environment and the workaround workedHope the fix could be included in the next patchThanks a lot for your help +Primary issue is easySecondary issueIt appears the Interrupted is coming from minitest +Not sure how we could do this +Either way the problem is clearly not with the Struts code itself so Immarking this as worksforme +I think there was a snapshot in the latest Enterprise spec draft. +System tests are great but are they really a replacement for unit tests? Especially as a method of confirming if a patch introduces an unwanted side effect or not +the revision Thanks Noble +Ill leave this open for all of next week and then close it out after then if there are no objections +some app directly launch AM without client AM directly accept shellcommand pwdls +Im not intimately familiar with this area of the codebase but I filed the issueThe approach seems simple and straightforward to me +Would it be possible for you to create a patch against SVN? If no Ill try to integrate your Java source file attached above +Thanks Alexei for clarificationResume should be excluded from build path of bothers me here is that if is not included into TPTP will have problems for sure vfcreateerrormessage is defined there +patch whcih checks if the transaction is active before trying to roll it back +Fix for tests Authentication patch looks good! Thanks for providing it +Aside from down loading the jars and putting them in the classpath do yo know how I go about changing out the implementation? There isnt much on the Woodstox site +Final patch that was committed to Shindig +Actually im quite a fan of geoserver and use it quite regulary +But when the apps in our tools these jars are all in the projects classpath so maybe have some conflict +I have retested this with the current version of +When job makes some action as writing is file this is not acceptable +It works with code in +resetting P open bugs to P pending further review +Ive committed the skin with this issue should be fixed with the addition of the sticky footer +Any changes youd like before committing this? Lance what config param changes did you have in mind +Please verify and apply the patch if appropriateThanksHasini +Breaking backwards compatibility is a big issue yes +The patch was used only on windows +Is this the same problem Ive heard of before when deletes and adds happen in the same request or is this a problem with entirely separate requests? +The main need here is to be able to add criteria on the clause of an outer join! Because in those cases the semantic of a criteria in the ON clause or in the WHERE clause is completely differentToday its not possible to do that and thus some kind of queries can not be expressed in HQLIm not sure the current issue will address that but it appears clearly is the heart of the problem +BOM is just an extra file that now gets included in SDEBSRPM with a complete Bill Of Material for a particular build of Bigtop +Previously attached patch file includes a file addition of a png +I was asking about the code deciding whether to exclude a particular key from the batch load +Ok after helpful comments from Wolfgang and some investigation I think I know what is happening and how to fix it +Hmm something went wrong +Alexey please could you close this Jira or provide steps how to verify it +Simon +Unless of course Im missing somethingI already implemented the first job now I should start on the second +I have recently written code for an enforcer rule to do this but it needs to be abstracted out so I can use it here and via a help goal +you are using a runtime right +I still get the failure in which without knowledge of how all this stuff works internally is really really mystifying +What is this simple line of code that users of the library need to include in order for SNI to workSee patch +class inheritance and removed all duplicated methods +committed in rtested on ubuntu smoke and passs lingering bugs in kernel +Verified as complete. + +At a st glance I see the following obvious issues with the patch +Maybe also take a range of allowed ports and scan through them until a free port can be used which will allow ops teams to restrict the open ports on a firewall to the same rangeThis is important for starting JMX in YARN containers as you cant hard code any port bindings without risking port collision +Well it would be great to migrate it toBut in current shape it may cause null pointer exceptions after upgrading ODE since is binary serialized inside instances execution state +It doesnt seem possible to distinguish between existing and keys currently would it be the good idea to revert this one? Or just close it and move all discussion to +But it doesnt metter for Seam Tools anymore because the fix for will copy any jar for Seam if it matches testng +issue didnt reproduce in the JSE environment after replace the with the apache harmony libraryAfter compared the native implementation between apache download package and android system and I found Android Froyo version will have this issue but Android Gingerbread version already fixed this issue +Please review the code and let me know if any changes need to be doneThanks we consider putting this in +I dont have agent log available and I am not hitting this issue looks as per your comment this is not reproducible I am reoslving it for now. +Reviewed all these issues and marking them as closed. +Hi JochenAIUI you have karma nowIMHO the best approach would be to go ahead and apply your patches to bring the JAMES version close to your current design +This patch looks good +Attaching patch that makes both the REST API and the RPC API accept multiple node states +Applied to master and +Added to test a three broker setup producerconsumer +Well this is even worse +I updated both +Velocitys quiet notation to check value exists and is equivalent to option +Its working now. +I came up with some implementation of the Cassandra SCA component with some changes to the original code +Running through Hudson +we use a comma in the region server log folder +David reopen if you can still reproduce. +The file defines the Buffering module which has all the methods in question build atop sysreadsyswritesysreadnonblocksyswritenonblockSo the simple answer is that we need to implement the missing nonblock methods above and then update the +Just one question Does anybody know why there is this extra payload cloningThis was introduced at the end of without patch in JIRA +Duplicate of +goffinet to commit +This should only be possible if the Member Group that the GH event belongs to is linked to a Facebook group +I can only reproduce it with a maven project and even then only sometimes +Committed to and trunk. +Verified included in installation from Pending spin in M branch +What is the value of in your configuration +Well see how it goes +bulk defer of features from to +I committed the patch +My work dir is clean each time i try to deploy the applicationI tried to deploy the application on another computer and this time i get this errorERROR Exception starting filter Class does not implement the requested interface I also recompile by removing the line in and get a similar error in when trying to log +Especially for single sourced query +Very clean and labels which used to draw at a degree or degree angle now appear the same way that they appear when they are drawn at any angle in between +We should coalesce if possible +is too old? Guess what I was usingAnyway was a false alarm I can reproduce your error +Left this issue open in case there are more changes +A patch for EAP would be nice until in Q +You may also want to try the recent nightly build of Lucene as it includesfixes for read past eof errors +The suggested change is a change to the behavior or public interface methods that we can only make in a major to release +ops that should be true instead of v patch +Do we have this in as well? if not s not in there do we want it since we turn it on? Or do we act like we always had it +Thanks Dan and Army for the suggestions and fiendishly good testsI cleared the patch available flag Ive got more work to do here +We have reverted back to XR so for next version +nit worth adding a comment to explain wtf all the clauses of that if statement are so we dont have to dig through ticket history next with comments added +added pagerank to +If I stay on the Location test page the icon does not disappearMartin is your application a single page app +If we are going toremove this capability we need to add some other toolsto get the same functionality +The numeric change looks correct to me +Reopen this JIRA as an improved patch is available +Good question +fire the with an Jsf qualifier +Added dep +Behavior aside wouldnt this patch break if the default message happened to be the same as the message returned by theIm also curious as to the usecase such that there were that many on the stack above the action +Locator sounds good +Commited in r +libsinatrain callusrsharebin +Hi RikkolaI will try to fix you could tell me how you reproduce great phew this isnt as bad as I thought +Changing fix version to +We no longer send messages back to the sender +It is unreasonable to expect that developers cannot edit the project with an IDE and then run a buildI also have to say that I really dont understand why this fails the build +Reran locally tests that failed tests that usually never fail and they passed +I set DISPLAY value to null then used installer to install EAPjava jar consoleand selected as the stackInstallation and server startup was successful alongwith generation of shortcutsHowever it seems steps to reproduce require small changeThe release notes clearly states this problem was faced while trying to install from command line In order to install from command line we must add console to the command in Steps to reproduceTom could you please confirm this and edit the textI will close this after your change +attaching patch and explanation for the embedded side +Looks spurious +Actually I did several measurements with a slightly changed test which uses the return value calculated in the loop and got an interesting reult on Sun JDK server modeCuserspozhdikhjiraCbin server st run out nd run out Other can not show such a good resultCuserspozhdikhjiraFtoolsjdkbin st run out nd run out CuserspozhdikhjiraFtoolsjdkbin server st run out nd run out CuserspozhdikhjiraCbin st run out nd run out CuserspozhdikhjiraCuserspozhdikhdrlvmtrunkworkingvmbuildwiniamsvcreleasedeployjdkjrebin st run out nd run out CuserspozhdikhjiraCuserspozhdikhdrlvmtrunkworkingvmbuildwiniamsvcreleasedeployjdkjrebin Xemserver st run out nd run out I see ways to achieve a nearly zero execution time recognize calculation pattern at calculate and substitute the result cache return value at and substitute it during More ideasBTW I agree that a dead loop elimination is a good way to optimize silly loops whose result is not used. +updated the you sure this is the output from latest patch +Ive had to do a couple of renames including renaming toIve created an package to store the JSON stuff and the couple of conflicting services +Will commit this shortly +Patch is applied at +Due to n problem it is really not feasible to use the constructor form at the moment +Resolving as wont fix per Ted comment above that it is no longer needed. +Thanks Benoit! +But if we put backup task into task directly we can solve this problem +Nick can you please answer my questions and then assign back to me? Thanks +Patch with several bug fixes for the utils project +The process should have started in a terminal now to have it working on AIX either someone provide an installer for this system or yopu can just write a daemon for itWe dont have an available AIX machine available. +Im not seeing this at allI just deployed a brand new config with webgrid curator filemgr extractors pcsws caswswebgrid works fine on Tomcat with JDK running on Ubuntu +Logged In YES useridIf you try the same SQL with direct JDBC from will itreturn the data +changed so that the default respository moderepository is changed to repository updated docs references from moderepository to repository and renamed to +Note to self Dont forget to check the site stuff +I read Spolskys original article and strongly disagreed with it although I do think disabled menu items are sometimes used when they shouldnt beIt always seems to me to be a good idea to give users relevant information as early as possible +it will require introducing some kind of provider SPI +Stepan it looks like scenario is it requires CRLF in the end of etalon files +I see that looks niceAre you sure you would be able to use named parameters? + cause is martins patch +Thats the current are you expectingThanks it appear when the smooks config file has the name by chance +The encoding is SQLLatinGeneralCPCIAS +However your main contribution of the I cannot take as long as you have not made any statement that you publish this code under BSD license correct fix version +Issue closed with release. +Yeah the CME will just cause you to not have the latest metrics but they should come back later +unfortunately the patch doesnt apply to trunk now +I forgot to mention I will merge this in CR not CR +This means that code compiled against will fail with if it uses these methods or extends the class howeverAny other opinions on what to do? Is the documentation in this ticket enough? Should we add these back in? Any other ideas +New patch with hung up fix +Would be great in case you can provide a patch +This is completely dependent on the domain model design +Actually this is very needed for people who would like to create embedded application with clean url with the rewriter for example +should play well with the rest of hadoop +Looks good to me +It seems to be that we might need a rd interface to cover these fields +Well have to pull this code into our repo at some point +is my version of addendum +I think that testcase was never enabled because it kept timing out +Attached v one more place same fix +An issue for GRECLIPSE project +Ok so I was able to ssh but now I tried to run this command to publish a release candidate and it is not working againscp p olgapublichtmlAt this point I cant ssh at all and I asked one of my to try and is not working for him either +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Ilya thank you for the patch +JBM provides clustered strict message ordering by using HA Singleton as discussed on the is in maintenance mode and there are no plans to implement strict ordering without using HA new features are being added in JBM not JBM +Hey Noah surely we can close this now but it looks like you reset it to reopened +Also test is checking for DST issues +You are correct Xerces does not provide a way to specify a type for attributeelement +Can anyone instruct me how I can provide a patch for it? Do I upload it here +A design document for implementing creation time and modification time for files and directories +The name of the variable does not reflect the actual underlying behavior +This jira is solved already +Patch with the Todd fix and I swapped an unorthodox public AFTER a static qualifier +SGTM +I meant to file a JIRA but then went on vacationSince depends on it should pull in the latters dependency graph as well +Lets keep this open for a while to see if the problem happens again +I feel would need more discussion on this +bq +As far as I can tell the problem still exists +no longer exists +Patch is comitted is created due to remove the patch after the issue is fixed +I need to fix this one too +d +Edelln would be a good person to ask about these remaining demos as she ran them all recently for thought the demo had been removed +Is there a reliable way to open the Recycle Bin object to obtain its ID +A new test to add to the plugin which fails under the current code base and passes under the proposed changes +closing since there does not seem to be any more interest in this +So you should have a proper dialog to input password if it was not stored in secure store beforeAlso you should not loose any user when restarting workbench whether password was stored or notThanks to was not able to reproduce on Beta and Ive not seen this issue on CR +Let me know if anyone have any questions or comments +then that is fine +Optional sounds reasonable to me +I replaced them with buttons it seems to work in my environment lets see if it passes QA +Thanks DIGY +Given the limited server support the fact that the RFC draft for VLV has been discontinued and the low number of users requesting this we keep scheduling this in the future +Seems to be the case for and as well +We have also a patch for the proposed structure and maybe they mixed up. +if Im reading this correctly the back compat break is for those providing their own custom highlighter? Thats gotta be almost no one +Patch looks good to me +The way of building the docbook has changes as far as I understand it we are using maven tooling now to create the docbook and not the ant scripting anymore + Roman +Any way to create a for this feature +Ive pushed an additional commit on the branch with that +why would we want to support both row and column mean subtraction? I need to the motivation of thisI think a lot also resides on a question if we actually also want output the mean +r removes the index +With and includedOnly question remaining is regarding +Updated documentation as well. +easy enough fix for +patch for yahoo security dist branch +If this is desired please open a new enhancement request +Yes very strange +so its only in presence of the old keys we do this approximation? works for me then +But comparing with XA this should be less expensive +I will take care of this later today. +Er in my example isreplace netlinksafeonlinewicketweb by mypackage it is outside the war it might be handled by a different classloader then the one that loaded wicket in which case there is not much we can do because only the classloader that loaded can see next to it +Vinu updating this to resolved + for adding events via also prevents from working correctly in Internet Explorer +Havent you committed it already? Or do you mean for trunk +Verified +We can work out the patch work out the tests and then decided its not good enough or perhaps another committer will look and decided that +If we are just doing back port we should follow it strictly without making more variants that can potentially lose track +Thanks Ahmet in my opinion this is good we just have to figure out where to place it +Attached is a patch that makes this change and adds a test +Couple of comments on the patch Move the last check to be as to close to the heartbeat as possible before we take any actions In this test heart is getting incremented irrespective of incoming ID +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +that property is protected and mandatory and in addition part of the version store +when you formatted the metadata shouldnt you format the local bookie? it is supposed to the logic for a clear consistence instance setupfor example if the ledgers exists before but the admin didnt run metadata format so there is no instance id in cookies file +Hey Edward Does the patch fixaddress any of the issues I raise above? If so which? And Im not sure I understand your comment correctly Are you suggesting that issues w the patch be addressed as? Or does other things mean something else +NET +Ill cook a patch tomorrow. +Is everything self contained within the example? Is it possible that I have something stale in my jboss server thats causing this? I manually removed the previous ear from the deploy directory and made sure I was on the tagged version of seam then rebuilt seam trinidad then the example +you need to save firstReject? +Im not working on this now please close thanks. +I am not running trunk but rather the stable branch +Pure blazeds configuration works +It is resolved as an duplicate bug there are no reason to have two issues open for the same thing +Although we dont support as a BG host operating system at this point well make sure it works well on it after will be out finally +With this patch ee works +using the same parmeter is prefer to geting another. +OK +Component upgrade reflected in the EWS release notes. +I checked in the change to get the tests running again but will leave this issue open for a little bit I try to understand that better +See also which was filed to lift this limitation +are doing +Attached patch adding final modifiers to all names for standard parsers as mentioned in your suggestion. +Using such an object an giving it a toString method could solve the problemCurrent memory usage of zxid and time in a bytes reference to bytes object header bytes for two timestamps bytes for three zxid longs bytesMoving everything from directly in the and reuse same Zxids across bytes for references to each Zxid bytes object overhead for a Zxid bytes for actual zxid and timestamp values in the Zxid object bytesSavings Around bytes for each in the system assuming that there are only as many zxid objects as there are datanodes +CVS patch containing the mentioned refactoring +groovy files are still doesnt resolve the entire issue I am seeing +A disadvantage of the tests is that they increase the total time needed to run the testsAbout the test failure it seems that the test hangs when we error out add keys but Im not completely sure +Same bug fix to be committed shortly +This seems to have been the behavior all along which is surprising because I could have sworn the initial poms included postgresql test exclusion code +I found options +Reproducer contains HQ installer built from HQ EAP branch +This way attributecomparisions where order is significant works properly when snickers encoders are used +Where is it? Would I have commit privileges for it +Would it make sense to allow directories as well in? That would make it easier to swap out a jar without rebuilding the tarballThe problem with directories is that officially they are unsupported in the distributed cache +There are many nocommits and loose ends so havent really gone to the perf aspect +Hiro The fix looks good and I pushed it in bfd but I excluded the tests +Awaiting deployment. +I think anyone serious about using replication should upgrade to anyway +This thread may not even be an X thread +It seems that there is currently an issue with Nexus where the files and folders are not showing up but the files actually exist and can still be downloaded when requested +Defer all open issues to +So will it ever be implemented? If so when? Thanks +Even if namenode were to do more work this kind of big deletes are such a rarity trying to make things performant for it seems unnecessary to me +Its should be closed as we have some different implementation in CXF. +Turns out there was a big bug in this which made upsert basically not work at allNew patch coming soon +Looks very useful thanks Myrna +I did not do that as it makes the code ugly and it does not cause any harm +This is an awful lot of work that adds quite a lot of code for what appears to be relatively little benefit +Thanks a lot Natty. +This should be available in the next RC +The parser now fully supports Unicode text. +Why does thrift need the hashcode option instead of always generating itIf Thrift needs the option youd be much better off throwing unsupported operation rather than using +BSD or ASL either is fine with another ASL project +The problem was that it wasnt being calledI was setting the URI resolver on the which generated theTemplates but it wasnt being passed through +No it fixed itself with a later cvs update so I assume you fixed itIll have to investigate it if you cant reproduce it +See the results on httplocalhostIt should just be working but notice that resources are loaded by blockWARNING This changes break branch +issue resolved in latest versions of Chromedirk +Thanks and apologies for the junk on the jira +Misread position as precisionStep should would have been using the default which I guess is +probably need to move this ticket to AS but i dont have Jira permissions to do it +It could be interesting to have this for CARGO +The patch attached +JessicaCould you look into this one as well? Not sure how advance search works +Whats the reason for thisInteresting to hear about Teeda never heard about this JSF implementation before +This problem seems to be causing of eclipse caching the previous versions of the pluginClosing eclipse and deleting the existing plugin and start eclipse and reinstalling the plugin can solve this problem +Bumped this to critical as I think it looks pretty bad +Adding it will really help as clicking on a link in a text of the exception or log is way easier than clicking through the packages in an IDE just to find out that youre clicking in a wrong module please +Thanks a lot Eli +I think Otis was just suggesting we start with those changes +enableassertions should be added to MAVENOPTS see trying to install the bundle in the local OBR file for the hudson builder machinesby setting the build flag falseon all buildsEvery now and then builds have failuresproblemsnusances due to the local file on the hudson build machines being corrupt +Yes I think supporting and here would do the trick +It would be nice if we had a very easy interface to run drivers and validate the output +HiFYI CXF will also have this fixFreeman +I guess it would be helpful to have a continuous integration system building Activiti with JDK if all developers are using JDK on their machines +NoteI am testing on Linux with JDK +Thanks Tim. +Patch is in trunk rev +Marking incompatible change. +Can you add the new variable in +The mlock stubbing will need to be rebased a bit for which was just committed since it was still stubbing with Mlockerrebased in version bq +SteveYes I agree with you +Ill go ahead and commit the current patch +Manually tested +So only that map entry is updated +One of the tests fails I thought I checked that before checking in the code +If we were to take that tact then it would seem like a Multimap would be just fine I thinkOr is there another reason to have the soft values that Im not thinking of +Maybe rename to +OK +So just get back to the image picker switch to the tab with the uploaded pictures and select the picture you just uploaded +A common case for a client is that if a insert fails the client will the latest version and then take some action accordingly +And then simply install rosetta into the jboss Riftsaw put all the artifacts into its own svn repo and then ask users to check out it manually into its local would prefer to use the jBPM way as it is simple both for users and developers +The attached patch adds an annotation and xml schema for a named fetch plan +So I guess running out memory is the main issue +out file changes and upload the full patch? You may need to run both and +derbyall ran cleanly with jars after applying this patch +I would therefore propose be a maven dependency of broker +what is the exact algorithm to find out tdl files in a jardoes the same jar included in netbeans web project get resolved correctly +When in a servlet environment just serve back the stylesheet otherwise take the tokenized one find all the tokens encode the URLs on the fly and resurn +Thanks + +Let me see if I can do thatThanks HiramLet me know any feedback you may have + +I can see that for encryption purpose you may still want to use SSL +Thanks +httpclient can do this in JVM no need for any extra dependencies works anywhere that proxies are not in the way +I feel like widening byteshortint to long value if not necessary might be overkill but I dont feel strongly enough that it shouldnt be done +The patch fixes DDL issues for various databases +Should have a patch for up soon +AndrewIts my first time to raise issue in Jira so I dont know the convention thank you for the adjustment +the patch only bumps the version of tomcat no need for testcasesthe line longer than chars is the download URL for tomcat +But again this is for deployment correctYesbq +Ted +Ok this makes zero sense +Id love to see some traction here soon. +This is actually really bogus as it seems to be using the current HTTP request as part of creating the SLSB +Thanks for spotting this! Ive removed that outdated +reattaching patch for I dont appear to have permissions to twiddle the Patch Available state on this issue +Nah sorry that wont work just realized it only extracts the tarball the untar task is deficient +Changing the extension to +Trying the ij sql that you tried in your comment gives me the same error as you +The current practice ofmarking things fixed in a hypothetical next release is provingtroublesome the result being that we now have things fixed in and but not +OK I committed the new test +r on trunk +But Im not completely sure +So I updated mine apps to and I ran the test after a while I noticed that are increased to a big didnt see the taken a was actually planning to upgrade in next few weeks looks this will be a show stopper for me do we have ETA of +If any commiters out there can check this in quick thatd be great +The escape string now depends on version +I agree on the package split that probably belongs in this JIRA +Why did you remove themThe point was to have default impl on Directory and one onOk lets wait for your patch +clutch is now displays no issueswill close this Category and notification emails are all issues seems to be resolved +You are right again with JEE bundle everything is OK +Getting an unexplainable from deep within is VERY confusing and I spent a day trying to find the bug in MY code +bq +But today we detected an file Exception inspite the delay eventually +The changes should be backwards compatible +Ive created patches for the service based example +If anyone starts this style elsewhere I am ready to change that later +Let me know if this ok +Corrected typo in description on JIRA numberThis issue is related to what is discussed in in the sense that the task tracker should decide how many tasks based on the availability of its resources +Unit test to show for +patch added +Got one conflict in which I resolved +Render pages using routing +use this for every html page of my app this was my reason to not release an unstable to devlopers +lets focus on the refactoring that is mentioned in the bug first then we can improve later +Ive seen this issue come up before in server environments so this request is definitely not unusual +Id suggest releasing a new version of xdoclet plugin which uses antrun +The trunk patch conflicts with +But this will be a behaviour change So lets discuss with IvanFlavio +Seems like you have a VC binary package and trying to buildlink with VC +Sure I will upload a patch for +Moving the Word document into the current Castor documentation standard format is a better idea of course. +Deep copy must be assignes not only when merging collections but also when merging +for now we can have both but I would love to see folks continue to bring the sh script into parity with the bash script +Tests still failing so I suspect theres something more interesting happening here +Just had a glance at this +Is officially being built with JDK or JDK Perhaps a more dynamic bit of logic in the script might be called for there to use the appropriate ones based on the javac version +I will try to reproduce it and if Im not successful I will close this issue +Includes the symlink change +The trunk version will remove them howeverI have more work patch where I remove some more code gratuitous constructors reducing the visibility of KV +and overload backup configuration to mean standby it is not clean +Setting fix version to as we need to understand the cause of the failures before the release +The patch allows you to have xalan in your without having tomanually put it into MAVENHOMElibendorsedIts just an extension of Joes patchDo we really need the property? I mean it only points to doc and is chars longer than doc +Is there any UUID associated with a table? On the other hand since cloned tables are mutable does it matter +As such this version of Templeton should only be installed by expert developers +Possible +Why dont people take the time to resolve their assigned when they merge? Why? +Rather it goes ahead and the entire pipeline +Patch applied +one more thing i missed is that should still return a String +Indeed Evgeny but we dont want to fix this ticket as long as we dont have any integration test. +Note that earlier releases can workaround by manually running flush against system KS before drain +Renamed summary +Manually tested +Looks like Grant privileges scripts are missing +per the issue meeting on May +Ive attached the error message contains the full console output and contains only the error message +Please check if you can serialize hashmap last time I talked to someone I was told to write my own writable object +Test case for the bug +Service createPerson follows suit but updatePerson doesnt require those fields +Yes i know the max files per dir is by default +Although as Gregg has pointed out we must be careful to consider compatibilityI was reading the website I think Chris can see where Id like to go thanks for the info linksInvert Dependencies The goal of using is to ultimately invert package dependencies such that packages depend upon packages +Id love too +But its not clear whats going on here since thread can be +Thanks +This also appears when when Railo and MS SQL Server R on Windows Server x +I agree with David about measures of string type +Its a new major version! Even steps major and three steps to close for +This problem goes away with the patch for the issue resolved with the patch for +Future improvements in the queuing me +It would be great if you can check that this works correctly +I am thinking about a configuration via an file +No +push out to +lgtm +the patch expects a configuration in where the value should be size of the global pool in MB +Thanks Juan Jose. +Please verify and close +Mikhail Thats the usual set of three that fail on hadoopqa fyi +Not reproducible in +So IMHO we do not need to document them seperately in user guide +This is a different resource leak Ill file later if it hasnt been fixed on trunk. +Thanks It was actually reported to me by a coworker but hes even lazier than I am. +Manually tested! Well done +I dont know how often it refreshes +I just committed this +So here goes +test failure seems unrelated +If we cant guarantee which message is going to be deleted or even detect which one were going to delete I dont think we should offer thisJust because weve always let people shoot their foot off in the past doesnt mean we should continue to do so +Could you please push Seam Render into Maven? The Forge build is broken at the moment I think. +I think it is a bug when the final variable is static it should be allowed to be set in static constructor same as instance level final variables are allowed to be set in constructors +I added arquillian to the master branch but removed the fix version since theres nothing to fix. +Leaving this JIRA open until they are merged to branch +would be good to get these removed from the jmx and webconsole for as they are now obsolete + for the to trunk +This will most probably be issue was fixed on HEAD +patch in a different format +These changes all have synergistic effects which makes them work even better +Please see +Hiis anybody interested in this featureThank you BrunoYes but I did not get a chance to have a look yet +EAP it is. +All nightly builds run to completion. +I will work on it during this week and update you +Thanks for the reviews. +Benoit I wanted to add briefly as I believe we have already discussed this but it may be fundamentally impossible to improve the serialisationview server communication structure except by migrating to a completely different architecture +build build test pass on winxp and linux rhel w gcc +Per my comments on Im fine appropriating this +The leak is not occuring now +fixed in rdo you have any idea how to set editor encoding in netbeans? I have project encoding set to and still properties are saved like you noticed. +LGTM committed +Thanks Christian for taking the time to provide a test app and pin point the issueI have looked elsewhere and have fixed a few other data formats as well where resources should be closed as well +Can you upload a new patch pat +is now final in +Okay try the latest snapshot +The deployments wont disappear from the when they are stopped +Introduce web application name as prefix of portlet preference s identity +FOR UPDATE statement but the stack trace in the bug description looks like a DELETE statement +Rob can you attach a of your patch and can you also check to see if Brents patch solves the issuefor you as target milestone +superceded by new Jetty bundle +jBPM is already tagged +Any chance it could be fixed in the dot release +Gregory +lgtm +More fully implemented and added unit tests that cover mode code and cases +EugenThis seems like something that can go into if you would like to champion the issue +would add that I fixed and classes to properly handle Blog unit test cases. +Name of the attribute will be configurable +Moving to +Thanks Robert +patch uploadedCombines and too. +Other than that this more of a question suitable to be posted in forums than a bug really +Is this still occurring? I was unable to reproduce this +Unfortunately I cannot share the code with you +Having reliable Modal Panel would be a great feature but in its current state it practically cannot be used +Fortunately all the affected methods are not part of the interface and most importantly get is notpublic +for the change +Would be great if you have something to put up even if unfinished +Test case as referenced by problem report + +I dont think this is a regression +looks fine thank you George +Does this still valid +Here a flex patch for automaton +There is some unused junk in the C version that has never caused any headaches +Not sure how this would be a issue +This means that this subclass doesnt have the required write access to the field variable +Committed revision Thanks for the review Stack +Implement registration in class +I dont know which icons youd like me to use +Thanks JakeBoth clerezza and are ok +My apologies for missing this at first +Closing the issue now that it has been resolved. +Could one of you create a simple war test for this please? It would greatly help us in determining the bug and fixing it for the Beta release +the unit test introduced a side effect as I set a System property +Im lowering the priority +Thanks for patch Jimmy. +Robert I think is related but different +passes +It is a major piece of work but something we can consider for the final release +joe just to be clear we didnt add this in a maintanence version we added it in a new release of it and we added it since there were the limitation that there could only be one process per folder +Hi Daryn This subtask is primarily about being able to compose and expose token endpoints for authentication or federation of users credentials or tokens from +Perform above steps to few other services too +Changing component from Test to Tools since the bug was in the plan exporter tool +All work fine on our side. +Investigating errors in the patch on +On windows got code page identifier from locale and map it to the name which is in the added file Would anyone want to try this +Would we want the new API to be enabled by default or something you have to specifically turn on in the configIf we want it always enabled then we should log a warning if the users schema is incompatible with automated reindexing and any attempt to use the API should fail spectacularly +Runs fine for me too the test only needs MB so I think its fine +so does it mean to say that s will not run on jruncf mx +I really like the idea of using the binding annotations for service handlers +Hi ErikSorry I missed your original message asking me if I could test out the latest patch Id be happy to help +This feature is already supported in the framework +Will you be able to deal with in time for releaseYeah i have a proto type working just have to do a quick benchmark to see if it will make sense +Thx +Im putting this into the AS releases as its required for EAP +Waiting for verification once JBT isIm not able to replicate that +People know and recognize it even if most laugh at it +Thanks to Christian for reporting +v updates to create a controller w forceDeserialize and removes the echoing special case from +In that case Im going to call this resolved +I dont know anything inside itself that would cause this +This jira is not needed any more +The patch has been applied by Mark +It is a test case which can reproduce the RayWhat does the RI behave if one is accepting and another is trying to change the block modeFrom the native aspect when accept is called for a blocking socket it should be blocked +Resolving this issue per Felipes request. +Closing per Alexeys comment. +without having looked at it in detail i assume that the problem is caused by an inconsistency adding a new child node retrieves the new child node using an different method +Have you considered upgrading to the latest version of lognet? I know that lognet doesnt release often by was released in +the work here is done +Updated diff per Namits comments +Finding out which one actually fixed this would let us mark this as fixed in the release notes rather than cannot reproduce +A shortcut for publishing would not be feasible because the shortcut wouldnt know which server to be publishing to + by +Patch applied via revision . +Modified the preparedplan logic to add a limit to the command +Should this be reslovedclosed +Changed the component to jetty rather than console since it appears to be a jetty only issue +Update +So Im closing this issue it is verified. +Updated to good +Gute Arbeit Uwe +We pass the deployment level role mapping via the context map. +Patch to fix the wsdl useage and also the case where the proxy is not able to find an applied this patch +Want to reapply with mighty Jon +Wo my update +Stephan any update on this +My last comment should readOne possible improvement might be to use for the toplevel package to keep the prompt length down to the required minimum +Sample is missing in the section +Fixes to Hftp to support token renewal +I dont htink we should be committing any more rb scripts like this to the source code they just get people into trouble fall out of date etc +HI WallaceCan you please check if the patch in fixes this issue +Reopening as we have now decided not to use File but String instead +Weird +I am not sure the OEM feature is even supported in which case the flag should just be removedCurrently I have just turned off the calls to restore remove +Fix for it on Mac OS X +Not sure what to do with this bug report +Sweet thanks +The issue seems to be with secretkey field of the user that needs to be decrypted +Actually Hibernate has a slightly different semantic than EJBThe flush cascaded operation is save in H and Persist in EJBHEM use a different set of to reflect this difference +We can use CRC +I can reproduce at will +Can be used this to implement generic DUAL table +Submit a word count example on a two file input directory +That said my point here is that it wouldnt be too difficult to get simple clients done in a couple other languagesUnfortunately protobuf impls for perl are not being maintained and it would be difficult to write clients for perl which is the favorite language for many sysadmins +Therefore I am reopening this ticket +this has nothing to do with dynamic vs static model +Want to change that Daniel +Thats what I thought you were asking for +Its due to faulty logic in the resolver that is order dependent and must have the entire dependency tree the same to guarantee the same results +Here is a first pass at fixing this +I dont have a clear idea about how people use it +Would it be possible for you to try version which has the fix? Thanks +Patch checked in +dont ask me what I have done +Similar but not entirely the same youre right +gitignore +Dennis version must be set at least to for the publishDate macro to work +AlexAre you working on this issue? seems Alex is not working on this +will if further problems found. +Thanks for the reviewsCommitted the addendum to trunkCommitted the patch to +Seemed ok so committed +Ill backport your changes to stablebranch jobs +I dont recall any code in patch that addressed this issue +It should have more clarity +This patch adds support for such that it will report the usage string of available XJC plugins +This one do trueFine Anoop +Since totalLoad no longer includes decomissioning nodes I wonder whether this could create a situation where all nodes are considered overloaded and we cant allocate new blocks +in SVN there was a slight problem in Initialization with non components being installed before components +The prob +You must run your operations inside a transaction otherwise its like having wild sex wo condom youre on your own +Ive fixed the type with version Thanks you for the hint JohannesSet this issue to +Plus it could be argued that it increases the element of surprise or at least the potential for side effectsIf a solrconfigschema refers to a variable that can be superseded in a the behavior of a core is explictly dependant on whether it is loaded in a multicore configuration of not +I dont know what the native launcher is and I dont understand your secondquestion +Hey can I take this over? I have a pull request ready to go +I have the same problem with the if developers will not fix the problem you guys should add some instructions in the document at least +I also think wed want to write sync points much more often than every checkpoint +Add a Throttler interface to allow data transfer throttling +defer all issues to +We were referring to the same commentDo you preferI can the patch and analyze test failure from there +Thanks patch applied +svn ciAdding Sending srcmainjavaorgapacheopenejbexamplesinjectionSending srcmainjavaorgapacheopenejbexamplesinjectionSending srcmainjavaorgapacheopenejbexamplesinjectionSending srcmainjavaorgapacheopenejbexamplesinjectionSending srcmainjavaorgapacheopenejbexamplesinjectionSending srcmainjavaorgapacheopenejbexamplesinjectionSending srctestjavaorgapacheopenejbexamplesinjectionTransmitting file dataCommitted revision +Should I upload this or youll take care of it TrygveIf so please assign to yourself +lgtm +Both client and server use the same to do this stuffUsing the servicebindings description for this is kind of ugly but I couldnt get the other mechanism for metainformation to work +If in a war deployment this is on the appears to be the case with DVD store as well +meaning it is an expected behavior to return null when the merged restrictions are conflicting +I took a very quick look and its interesting +In this patch I removed some tests modifications that were not intended to be committed +It provides an operation channel so for example one can send a message to startstop endpoints +VDB reuse has been a long requested feature +So not jdbc +This should be fixed by the fix to +I am verifying thats the case and I will take a look as well +Reopen if you see it again. +Ive made the a bit more verbose +thank you for remind +Test case has been submitted to MVEL author. +Thanks Alessandro Objects get updated from time to time and I missed this one +Attaching a patch the implements the basic idea +Found a fix for us and raised to propose the fix in the compiler +It depends on whats inside the group and what the namespaces areI have tried a simple example using that Schema fragment and I got only good error messages +You need to explain better what you want me to look at or what the result should be instead of describing the general and then giving me concrete example that is working as expected +Not sure why yet +Incorporated most of Petes and Craigs comments +Install version build on the cluster +Fixed in rIf super class does not have default constructor call the constructor with the fewest parameters with all the parameters null or zeroThis fixes error but it may still cause runtime error +As such it does not seems to be a priority for the existing CXF community +HelloIve tested the Script i currently have it installed in a Development Server +These both sound like good changes +The reason for this was that I could inspect runtime artifacts to ensure the behaviour instead of having into inspect files or jars if I had done the test using the pluginThe upshot of this is that I think the behaviour in question for this JIRA is better tested in the project where we can just add the schema to a stanza in the and exercise the generated classes in a small test program +spaces new lines etc +It looks like the same NPE withIf you are sure its a duplicate can you mark this as so and create the. +Sorry about that I did not make the issue clear enoughThis should only apply to automatic optimizationoptimizations triggered manually should not be reorganized +I think Brian Fitzpatrick as a good idea of the current issue +another rule be added to pick up completely attribute values? +This can probably be closed +war download since were using it directly in multiple places +integrationcomponentintegrationcomponent found this issue in latest runs +The question is whether an is expected as you say to determine the matching semantics based on matchingRules for the attribute +Thanks Richard +Each server will fetch the token from when the client establishes a metastore connectionAFAIK ZK currently doesnt have a security +Trivial change to incorporate Vinods comment on error code +The other reason is taking the info message to the users console should probably be done after +This error is when you try to invoke the CP or the loading itself +HUGE portions +This should be with the Web component owner until te tag is created +The patch looks goodOne question with the patch cannot be used as valid id because the code interprets id as undefined id +All of these issues have to do with redirection not updating the original URL in the a consequence of this issue a crawl could be permanently blockedImagine the top mio of crawldb are all redirect urls whose targets has already been fetchedThen you can successfully generate mio fetch mio and parse mio but after an update of the crawldb nothing has happened +Committing this shortly +David thank you for showing the right wayI created a new patch which uses and entities instead of + +This will be very useful for advanced mojos and reports. +Once the reference is gone the session should get closed as a part of the garbage collection process +contains a proposed implementation for need for this change in AS and later. +java and things are much better +Triage Not for Out of memory problems after an application is running for to days are very difficult to diagnose +In the case of drive failure this means you lose a chunk of swap +I agree +Could you adviseEric +jieryn Maven isnt that opinionated especially on such religious issuesBut do I agree that to be truly tidy it makes more sense to reformat the spacing than to try and preserve it +Ive just committed this. +I dont know why perhasp come kind of weird setup in the poject because I have another project were it works fineI have tried removing the ivyde container in the project and enabling DOM Runtime and the it works fineTo recap I have one project DOM Runtime and ivyde all of these components together results in the problem +HeikoOne more question is a relative link good between the process definition file and the image file? Given the process I can look up the resource name of the process file + +Applied at revision +parent which is the correct way of denoting directories see +Right +it works Olivier thanks +Minor cleanup to the +That is rough draft which I created during fast jira review +Blocked by +Am I correctIf my assumption is correct then I need to find all the functions that invoke Queuerequeue and insert some code before calling it +Anything else I can do to make this patch commited +BCEThanks for the quick response! +Filter controls are absent +is stateful and can not be shared so subclassing to change seems to be reasonable +Yes I agree good idea BJ +The cause of the issue seems to be that the BPL tag is incorrectly having the invoked on it +How large is your cluster +thanks +cold vs really cold data from any extra processing that might happen when data goes from one classification to another +this issue actually duplicates +It looks good but I think it has an issue +Ill commit that tomorrow if no one objects +added the testcase provided by Siying Dong +Integrated this fix from John Howard +Not verified but works for me +We need to get identity columns back to a stable state on trunk really as soon as possible +Committed revision Thanks Mahadev! +I dont think we should touch this code +The patch for H is included in the above V patch +Please verify +Fixed +Thanks BjornThis looks as what we need +In general my opinion this should be fixed in may be Im wrong it is necessary to get some other opinion may be from developers +Hmm +will wait for a dont see these in the build today either +This causes right part of central inaccessible when central is very narrow +A hook factory method on to create the would also be helpful +pushing out rolling RC today +please review +Patch to support to Encrypted Header +Origin server connections are not encrypted +Camel uses Apache +I think this is a more general discussion than justAgreed +Its a bug which exists before my patch +Applied patch from Joakim +Oh +It has not dawned on me until now what the actual difference is between process instances and executions +yeah +why is this better to move off of a public mailing list where others can contribute information? the problem is that sybase apparently makes no distinction between permanent and temporary tables +Just send a message to the list +Does this patch also do IO to the block and files +If the message cannot be removed the Server cannot return with a certain successfailure to the consumer +uses viewindexdir while Refuge on indexdir side +I have a workload that has the hot dataset in cache and the bottleneck for this load is that writes to the HDFS transaction log have long latency +In order to get rid of the limitation entirely need to change compaction algorithm on a different one +Dan I assigned this to you +Performance +Call your loop variable something other than locale +The rest should be done by was done for GA fix version and will be reviewed in future is done +In the trunk we have a loadable modules for commands I think I could add a conditional loading support do wed be able to check if server supports indexed stuff and enable indexed commands when needed +If this patch isnt implementing NRTwhat is the point +have the right proportions or so and this is considering that the website is MOJO which is part of codehaus +Attaching patch for time has flown +Committed thanks Namit +and the forkModes like thisnever once always of +If Affects Version and Component are not set the issue will be rejected +Please see duplicate +Thanks Andrey will try it to trunkOnly removed the ok return from terminate function +I was testing on Windows R +For JSP JSF this is even enforced at the JSP level +Alright added some specs for thisAssuming foo artifactns do ns abc endthen you can write eitherbar or +Eli suggested I post a proposed amendment as a patch on JIRA so thats what I didFWIW I dont see this as a change in bylaws just a clarification of existing language +Apologies I find that my report was incorrect +I do digress +from me too! thanks +Upstream fixes were merged with the latest rebase +The reviewer will need to verify manually as part of the review process +I havent even had a chance yet to thoroughly review the updated design doc +Ive committed the test case to trunk with revision +it seems we should use GMF +Closing this as fixed +ThanksI applied the suggested with rev In addition i added a comment explaining the reasons for replacing a null dav property value by empty string +Andrew thanks fixed incorrect in my earlier comment +trunk +bulk close of all resolved issues prior to release. +Mass closing all issues that have been in resolved state for more than one month without further comments +Agree on the general principle of introducing table dependencies later +fix required to validate this change. +some improvements on the previous patch mostly changing the test to work in a similar way to refining stopwords list javadocs etcI think this one is ready +Harish and Ashutoshs comments +Patches are the patch with Revision +RESOLVED Now if a is not found the last version of the library is assumed tests for richeditor and ajqueue have been added +The case is now closed but the tests still discussed on IRC it is probably the case that the tests need some form of rewriting. +Anyway this current patch is still valid and usefulThanks again for those explanation +All the TCK tests pass so the changes appear to be safe from that context +Brian could you update AS with a new release for us when youre ready +Let me check +Blocks development andor testing work production could not runThis is not quite how I would describe this issue +Wrote a but the bug is not reproducible in needs a single node hadoop clusterThe file spray with the is within the single reducer but the sorting bucketing of each of those files spawns off a separate job ending up with potentially parallel tasksThe tasks are being executed in the local mode in a single thread and do not have the same timing to trigger the race condition +Please let me know if there is an interest in committing this flavour to sonar +You bet +If you still cannot download the iso open a new issue +No commits have been made but has been released +Im committing it so we can start to have it tested in the builds and give us more of a chance to discover if any sneaky issues are lying around +This issue has been broken up into separate JIRA issues and +JSF Tools Tutorial is checked +This new compiler keeps track of all dependencies between files and across projects +Backported to r +codetest fixes to use predictable field orderCommitted revision +you mind raising an issue on this? thnx +Committed at subversion revision +I had a logj setup once which aggregated into a mysql db via a rsyslog hierarchy +would work without SFLJ thats correct but you would need JCL anyway +Good point! The absence of the attribute does not necessarily mean the request is not a forwarded request +Incorporates patch for incorporates the patch +Isnt the user the one that is responsible for providing the type of the entity to use when serializing? Meaning that If the generic type is an instance of class it should be passed through the same checks the raw type has been passed on the first place +Changed to use the same +But i only have a Core Duo CPU in myIt might also depend on the data file I usually used the English K sentence file from the leipzig corpus +Please review +Do we have feature freeze in progress for already +IBM JVM is working Harmony is still broken although this appears to be an NPE with mavenClosing this generic bug +Hey Dave! Thanks a million! Let me know if you need anything +Patch for trunk x doing the cloning correctly +Deploy this file first +Thanks for reporting +Create a new issue for that please explaining why you need it and what it provides that the current setup does not. +for the point It is ok I didnt know that +nodrop on table level should not block us from dropping partitions +HiAny ETA on that? is clearly up and running now +For sure +Used an approach similar to the last proposed fix but with the try catch happening in the btree code +This should work in trunk +Fix is committed +with the JMX console unsecured etc +Do you clean your repo at night +I did above modifications you can find the patch with in r +Sorry for moving quick trying to get at least an initial implementation of auto failover committed quickly and we can continue to improve and fix it upWhats the hurry +KelvinThanks for reviewing and applying this patch for me +Committed the patch in revision and removed some no longer used code in revision +This will provide flexibility to work with different use casesAnother test case can be added to test field values with the fragsize param +These blockers and critical issues are resolved but not verified +Verified in. +Note Roo will automatically reorder attributes in an alphabetical order but it should not remove customizations such as test attributes or renderfalse +We dont see the same slowdowns perhaps since we arent using a shared ZK instance just the one for this job +Its only mappingResources which doesnt work properly +for the tip +Traing a model with the UIMA Sentence Detector Trainer +are you shure you get this error with railo we have run this example on this version without can see version for example in the error print or on the startpage of the shot of the argumentcollectioncfinvoke error Im seeing +That was FAST +Ich bin gerade am berlegen was hier fr ein Vorgehen das Wnschenswerte wre +Changes Per Henris comment now keeps a reference to the Properties instance instead of keeping a reference to guess what is already there meets most of the requirements +Ok for now just implemented support with adding deserialization support should be relatively simple will add separate Jira entry. +As we decided to get rid of PHPCPD I rename this issue as the title doesnt make sense any longer +Why doesnt it work with JDK? Is there any compatibility brokage between Java and + ThiloI applied your patch with some small changesThanks should work for start events as well +But the larger fix to be smart about compilation and able to fall back is nearly impossible with current and too much effort for such a narrow addition various fixes have come in to allow compiled code to be debuggedtrace with more to come +You can close the enhancement. +Agreed Ill keep an eye on devcommons to see if I can help push out the release +From the top of my head I think server shutdown is implemented by first shutting down all booted databases and then shut down the server +avoids conversion of EMF and URI altogether due to UNC path issues +I am seeing the same problem with Eclipse +I think its because you marked it as updatable issue for me with H and HibernateEven with updatabletrue it does not work. +IMHO if its breaking we should revert for now and Greg can look at it when he gets back +I am looking at the latest snapshot and it has the suffix of the snapshot and it is showing +Thanks Willem +Here is file to test the is a file to test the just realized you are talking about Adobe Flex +Definitely a bug +Would anyone like to try this +this looks resolved +This is the case since all columns as nullable and timestamps are not unique even when paired with any other columns +committedDIGY +The change implies to the tests which extend +Addendum for fixing a typo +Candidate patch for the issue +I just checked out from the trunk cleanly did a recompile and Im still getting the same result as I did are my propertiesfalseIve tried other options combinations it matter that I am running under Windows using Cygwin for the compileI havent made any modifications to the Castor source +So the Scoobi rec seems to be dont use static references +Jonathan has a master ticket for all the gaps in spring XML vs the java DSL +sure is missed that one +At some point the core should be possibly used with other setups so we should be careful not to mix those things +Reverted to +So long as the timeout applies just to this test Id agreeThere are a couple of tests that take about two minutes or a bit more so applying a universal timeout would be unacceptableHowever if we could specify test timeouts on a per test basis that would be a +I have added support for a ranking attribute and also changed the code internally to use Dictionary rather than Properties +I verified that the test passes with this fix and fails without itMy only suggestion would be to add a comment in the OPCLOSE case to mention that this code will only be reached when loading old edits log versions +for patch but you missed +committed this change to. +In case there is a to using yarn kill app if the AM proxy is not available +After installing EDS with CXF following spring jars are copied to the client directoryManifest says they are jars of spring versionSo I am closing the issue. +The attached diff was run in the plugin patches the class by changing the type of the s jarResources collection from List to class was also changed to accept a Collection of instead of a List +YukI guess we just have to replace the groovydoc tag with a java tag so as to get a new VM with more memory +Patched build +I actually had just found that +Committed to trunk and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +This will affect not only but all benchmarks including DFSIO and especially +This patch sets default values for the three system properties that are used in the default property filesINFOstdoutIt also includes some examples of increasing logging levels +Fixed long ago just tidying up. + +In order for the error pages to have a decorator applied to them I need to add an ERROR dispatcher tag with the filter mapping. +The right approach is therefore to get these two parameters in alignment +Sorry I thought the other jira was still open +This should fix jetty in our code but is not tested +This has actually been resolved in CR +I made some changes to and classes to make use of as a temporary solution +Puts are already synced +throws +integration to work with current jetty trunk +Furthermore after applying all instances of java are gone which makes all the substantive differences between the two patches irrelevant +Committed +Note AUTOINCREMENT could be used in needs another algorithm +So if we dont do protected ie +Sorry my fault I thought it was but as you point out it is always +Unfortunately that the refactoring menu is not extensible and so Groovy Refactoring must be separate +updates to variables are atomic is a pretty pointless comment now +Info from Brian F +Let me know if we would also like it against +Close issues after release +Marek Im pretty sure the issue could be reproduced with several entities +testcase for for fixing it! +Im wondering if someone has ever been able to successfully use Archiva with TomcatDebug is ongoing on my side but if someone could find a fix before i resign it would be fine! I will also test with the current archiva from svn under Windows to see if this work better +Attempting to close again with further removal efforts. +previously. +Thanks for Patch. +This patch adds for demuxChukwa metrics are defined as the following column family chukwaAgentchunkQueue chukwaAgenthttpSender chukwaAgentmetricsHadoop metrics are defined as the following column family jvmmetrics mapredmetrics dfsmetrics dfsnamenode dfs dfs dfsdatanode mapredjobtracker mapredshuffleInput mapredshuffleOutput mapredtasktracker mapredjob just committed this. +Anyway since only is was used only super classes where checked and not interfaces +Bulk close of old resolved issues. +Start the bundle at a higher run level when the system is ready to accept requests +Tags updated updated test added. +When CR comes out try it out and if the issue still present create a new also create a new jira for the alphabet issue. +I already have a prototype of this in jbossmx trunk which handles the mainstandard mbean currently doesnt handle the descriptor annotations or the stuff. +I checked in initial support for this +Thank you. +is for branch +Could you please update issue with proper state +Also it seemed more related to server not being registered as started +This is the core ofThe following extracts from comments on are very relevantThe network client will request LOB columns to be sent as LOB Locators instead of data value bytes +I tried building on Ubuntu and it successfully built the binary. +Ill file a separate JIRA +Its time to close this out and make the remaining tasks standalone. +Committed with minor changes in revision file to upload was missing in patch added some comments and TODO notesThanks for your contribution its sooooooo easy when people provide tests with their patches! +I connected several consumers to different topics +Hi Suma is not found in the patch +Feel free to reopen if there is any additional work that needs to be done. +Ongoing work post AS GA +I dont like I committed by why dont you pull my changes and see if your system still worksUnless you bring more arguments in favor of installing in participants directly Im going to go with the NOTOTALORDER solution. +This is related to +Also fixed which didnt compile +Those need to be handle on container +Would it be hard to make a unit test of it +Removed manual handling of code pages in CVD +The context is then used by all layers from there +How may i assign that ticket to me? could not find the field +Im away until the weekend +YCSB was used to perform all tests +Does this warrant a change to the documentation +For completeness sake an attached patch with the missed Content static ref taken outIf there are no further objections Id like to commit this sometime in the next hrsThanks for the thorough review AndrzejIt feels good to do some Nutch development again +tnx for input +I think what we are seeing here is a timing issue +Yet the current situation is that this handler is always run in the client inspite of the fact that it is not neededIt is also correct that someone can explicitly set the relatesTo field of a response message but I feel that we should be encouraging users to do this by using the standard instead of going off to invent their own way of doing this +Also added row count to Query hisory +Yes Ill and it +updated diff with tests just committed this +would you please try my patch? thanks a lot +If this patch is accepted and a new release is pushed out I will submit the patch to Spring. +I think the fundamental problem is that its driven by the write traffic not the read traffic +If frozen the node will stay in place when the layout is refreshed or changed +What prevents this working is the lack of a defining project when setting up the groovy parser with the appropriate groovy class loader for transforms +MathiasYou made the review private +Shifting to ALPHA as we dont have time for it to do this after M is released +This patch passes in the real mapreduce reporter as the progressable for get +batch transition to closed remaining resolved bugs +Move to +Moving out of +return should always be created otherwise will throw NPE as demonstrated here +I cannot reproduce an error with the code that is shown in +I think the bug is from my side +The quantities will of course will be calculated based on the parent product quantity and automatically setAn entity such as with orderId order productId product price quantity and description as fields could be created to provide such functionalityMy other thought was to have BOM components be added to the order as order items and have records for them relating each BOM component order item with its parent +Glad you like it +Increment for batch loading through +This would be very convenient indeed +I added itThanks for the patch with +When we change the Throwable code we must make sure that it is not possible to recursively cause an OOM exception +Added annotation +Thanks Awdesh for your Workeffort patchYour changes are in trunk rev +should follow into performance optimization although not as big as +It crashes when transforming this file into +Solr always used real docidsEven when you sort only by score? in this case the lucene docid is the tiebreaker no +Ok second variantStore flag mconn and sending test query if and only if only when flag missed button +Made blocker on +We dont know if the underlying bug is fixed since we dont have a repro but there is a possibility that it has the same root cause as and +I reached the decision to include the extension methods as you want them +But for now I think the patch fixes this bug and that those issues can be addressed though subsequent bugs as we gain experience +Patch including refactorings and test casesIf there are no objections I will commit the patch very soonIf you have any questions about the implementation please take a look at the provided image first +so basically if I understood correctly is that the overseer has the capability of doing periodic checks without an explicit action request from a client which can help on cleaning states check for failures etcYeah basically either every n seconds or when the overseer sees a new node come or go it looks at each collection checks its replication factor and either adds or removes nodes to match it given the nodes that are currently up +Some further information on this am seeing the value for the builder is being initialised by the thread main as part of the deployment process for a POJO endpoint in deploy at the time the server is started if this happens then all threads spawned to process incoming requests are sharing the same document builder +Are you sure youve got the right product? We dont mention anywhere in the hierarchy +Issue replicated on Ubuntu have no idea how Tools can break that +Reproduced with Xjitcount Xjitcount the same as XnojitI see Xnojit when I use X to show all the options but I dont see Xjitcount described +Applied the fix in the formexport IFSbecause I feared that n is specific for bash +Fix is now available +Single has just been superceded by the fix that will cover ESB now signals the jBPM process in an asynchronous manner any such failures should be retried by constraint violation is a symptom of the optimistic locking in jBPM and may still occur. +Please let us know when there is a snapshot build available with this fix as I would like to test it +lowercase schwa was missing and thus added +Those would be very old logs from before replication was added thoughIf we break compatibility how are we going to make sure that there no old logs to read when a user upgrades to? In this case we will guarantee that old logs cannot be read any longerIn addition there would be no way to check whether a log is old as the first byte can be anything +Oh that was easy +Lets leave it that way. +Thanks Shubham your patch is in rev +Can I close that issue +We are using that code in production and it works fine +bq +Ran it timesI will try one last time to trigger Hadoop QA +a temporary workaround to id problemI think there is another issue about this somewhere +My rationale being that the point of kmplacemark is to be able to view attributes without overloading google earth with a lot of geometries +javadocs are generated just like jar files +binnutch webtable +regionserver for exceptions etc +Thanks +The OOME is happening in the RPC deserialization +what messages is waiting for that are causing the problem +Patch applied +We should probably move the documentation to some other system +Moved the fix version from +This is the main reason that I made the directUpdate functionality optionalThe exception handling seems to be backwards compatibleA separate implementation makes sense too +committed with minor modifications +Hi Im okay in principle with adding something to for this but it doesnt look like either of the two patches currently posted are appropriate for adding directly to +But on the current Master code Im not able to reproduce this issueCould you retest against the latest Git Master version +Hi YuriI am sure others would like this functionality as well +Bad luck this bug hit in so late +Yeah! Nice workMind sharing some code? I hope I get my SSD up and running soon well see how much improvement your buffers yield on it +Fixed on and trunk +Verified working +Wonder if it would have gone faster wo the slow node +And the direction with Codecs is to handle stored fields too and I also think custom files +Excellent idea! Sponsored +Not a problem +Yup that did it! +to previous comment +So were saying that the priorityQueue is essentially unlimited in size +My issue might be related to explicitly defining MAKEINSTALLPREFIXI think it might be an incorrect way to run it. +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +In we talked of subsequently testing with larger values for +I updated the docs as well +Thanks Paulex. +I do not have the impression that these attributes are applicable to We do not implement any particular javax specification or something similar +Right then it should not be generated in the top level type as it should contain only whats common between all spec versions +If its then you dont have this extra component causing fragility +Anwyay I will create some testc ases and fix it up +Let it be built as part of and Geronimo will simply be a consumer of it +Thank you very much TinnyIm looking forward to using the next releaseTobi +Committed thanks +AshutoshThe reason it is off by default is because the default compression is gzip which is really slow and most of the time not what you want +Committed revision Thanks to everyone who helped out especially Carrot creators Dawid and Stanislaw +Thanks for making these changes +Updated dependency to version +Sorry Nick no clue who owns them +See though +I modified interface to remove shared fields which could cause the problem when processing multiple request in the same time +could be a browser cache issue for meWell anyway your version of the patch is much more safe +I suspect Lucene releases are a bit far off at this point but would appreciate being corrected +The project is the same as it was in snapshop area +I get no problem to enter and process data in my applications +Taking a few bugs to work on. + To run the unit test you will need to do Dtest Dtest +Could you confirm that it does in fact resolve your issuekind regards BobYes you are right is same as this issueThank you the information! and will you please close this issue as duplicatedbest regards Itou Akihiro +Bulk closing stale resolved issues +with wrong builder order for WTP JSF projects were fixed in for M +Because if the file is there but its location well create a new +However using a parameter in a select clause without a select may require casting and should be investigated with a follow up + fix the file checksum url + +Thanks VinayUpdate patch as your commentsExcept get do throws I deleted it in this patch +o Added PUT and DELETE to request method o Date select box does not refresh properly upon changing the year +Im working on adapting the bootstrap code in Seam now to use that probably wont make it into Seam though. +Going to commit when local tests passThanks for the patch Bruno +Can I get a +Perhaps we should make this objective more concrete +For the moment I cant tell why +Fixed in the current svn trunk +verified on Developer Studio +In addition logEnd must not be reset to INVALIDLOGINSTANT when a scan is closed +This issue has been covered by issue +Patch has been applied to the trunk of SVN +Thanks so much for the information Dmitry that has been a great help and allowed me to write the release note +s parent task and are logged against +Yes I think that makes sense. +Apologies I have mixed up releases +I was using a database using MSSQL than the company decided to use a server so what I did? I just changed the Dialect connection variables and on my happy thought it would work just as it supposed toBut what I saw is that Ill have to change all my source code to add this annotation Type +The patch wouldnt apply to trunk where heart beat has been removed +We have to change the messaging system again I guess the is sucking up all the memory +Hi ViknesThe patch looks good for the current requirements you have +This is not a problem for rpms because according to the installation command from the latest Release Notes the modssl is not installed +Thanks for taking a peek +It was impossible for me to comment on the proposed implementation without understanding whatthe actual change wanted to accomplish +I see a big need for using cloud hosted web apps but keeping ones data on ones intranet +Awesome guys! I wanted this to be done sometimes backAlso if we can put the pkg such that all the deps are in namespace so say if a system had preinstalled it wont conflict with which would be part of +For the smaller examples I did indent +And even worse a search for something like will find all documents with an a which is most of them if any largish text field is searched +Created for the contention problem +to allow for secure communications +The question is if we should apply the same change there? It would be a slight change in configuration but in practice you cant run secure without both and set to kerberos +Im not sure where I was looking that it wasnt present. +You are right in that we should probably attach a tag to the nightlies such as DEV or SNAPSHOT prior to the release being rolled +I have ran the test once again and its okI have committed this +SBS uses canonical link to help search engines understand which is URL number s exactly what Google to webmasters suggest +The current lock strategy is based on what was available from the lock manager when it was implemented +The above two are the only ways to change the repo location I think +Resolved in +Current should be named as Graph +Committed to and trunk thanks Colin +At the beginning I only focus on theis function +libs +Nevermind it suddenly works +Bulk close for +I think I understand why the custom wms kvp adapter is needed +Yes its a showstopper for us as wellWe probably need some official response on this breakage to see whether this was an API improvement of some sort or simply a bug. +Alex says it all +Well leave open and see how things go. +The delete does the sameThen a TOLEADER add pops back up nothing went to control so this came out of nowhere +According to your documentation iterate should use the query cache just as list does right +Since there doesnt seem to be any objections to Daniels suggestion Ill markthis one as WONTFIX for now +I think Yip has a reference to SQL specification section about the handling of this issue +bq +And I perhaps should have filed this under docs +Thanks for the cool test Suresh! pending Jenkins +Marked as an incompatible change since existing code that relies on bucket creation will need to be changed +hdataTable is working with Set In xdataTable are problemsAnd I have the same problem with xaliasBean in level Cristi +s great the addition to will be useful +I found the docs on the section +reviewing patch now +EXITONCLOSE is perfect for the geotools examples +Maybe we could just add a boolean islocal for that +I often do that too +EricIf using thunderbird it is possible to get authenticated and if James has strict configuration mail could be blockedBut my test code still can send mail anyway +I just noticed that is the same issue +Ill redeploy the site. +or maybe there is another more appropriate way to prevent the thanks for your excellent assistance with +Thanks very much my software is now happily updated to these latest versions! +was proposing that we remove that flexibilityWhat is the simplification if we remove this flexibility can you pl explain? thanks +That was the correct errorI applied your patch to webrick locally and the error is now gone +I just think that SQL should used with care dont perform a query if you know that it will fail anywayAt least this sould be configurable +but the bad news is it uncovered a deadlockhang! Im attaching thread stacks +Am I missing something here +Could you please explain the usecase of autowiring the job into another bean? Im just curious +As part of the release process the versions in the maintenance branch should be updated to remain compatible with the CXF integration branch +FYI I cant reproduce this on Linux +We should verify that this doesnt occur there anymore either. +Jenkins builds have failed times in a rollqueueFailover failed in builds and It failed consistently on as wellRolling back the patches +Test better test case +I just tried to build the continuum ffrom trunk and replace the jpox from to this problem still persists +patch for Trinidad the instead of the first one submittedWith further testing I saw that the webkit xml agent file did not list webkit as having the capability of supporting HTML fieldsets so fixed that too +this patch includes createmode in the end user client api docsrun ant javadoc to verify output is in builddocsapi + be nice for and will open new ticket to do the same for and +Hi Martin can you review this change please thanks +Here is the patch that implements everything like youve described +STEP srcqpidlog Apply the patch inside QPIDCPPROOTsrcqpidlog +Patch contains target and collapse attributes in item element +to commit and leave to other ve just committed this. +also tab selection doesnt work in FF +For compaction strategies it would allow us to have more informed compactions immediately after startup +invalid +This changes our parse of CF names but I dont think CF names w a slash or forward slash would have ever worked anywaysThank you for the patch Ian + is in just a proxy to the +Integrated to trunkThanks for the patch Shrijeet. +From what Chris Nokelberg said this sounds like its actually a Spring bug +Committed to trunk +Keep them comingRajika +what exactly you use to create new I saw is a archetype I select some other but dont found this to really need to report this issue against +It seems OK by me +Okay then lets resolve this and release what we have so far +On the patch Ram should we even try unassign if server in is null +To summarize my opinionNo +Im definitely interested in if this is the way it is intended to work per the spec as this could cause issues if switching between implementations if it isnt part of the spec +I dont have time to investigate this issue more at the current time +It may be due to Network proxy caching +Please feel free to reopen if youll see same problem in build +patch pushed +bulk close of all resolved issues. +Hello Brien I commited some better support and a test case which would fail on the check it out so we can mark it as solved +The general strategy Im going to take is to limit exception nesting to one level +Will check this in after running a couple of sanity tests +Both serialization and deserialization of session content uses the application class loader as the TCCL +So what are the options now? Weve been trying quite a few over the past few releases and it seems to me that no matter we do its always inconvenient to somebody +As long as you dont invoke javah goal in this pomit shoul never look at your +I was looking at this ticket thinking this rings the bell +Other than that great app Ive been hoping somebody would take the time to do this for a long while thanks for doing it +Perhaps an assert would be better here? Is there a legit reason for this function to ever return null instead of throwing +Uploaded a design document that describes changes related to and +Patch is attached +Verified with JBDS +Addressing review comments in pacth the code in and add two new test comments on the review stacks comments on review m waiting to commitCurrent patch use to provide consistency view of META entries of table it is similar with +I am not sure whether it is an issue of test case if not that means that the transaction did not work as expected +revision in JBDS beta to add newandnoteworthy Koen Im unsure of what the release note text for this issue should say +I also fixed this one +Uploaded as orgjbossdeployers +The workaround here is to provide event handlers for the prepare event and to explicitly instantiate the bean in code rather than relying on Tapestry to do it +Thanks for raising it Doron +Reassign to a release that you feel comfortable resolving this issue in or leave as is +need to fix the IP range is prompting for gateway and net closing the bug. +Were synchronize additionally to check whether updated vertices are exist +See for an overarching issue that tracks other such changes +Please have a also i make the Message Store a synapse Artifact +ip mail event adapters now have. +All the regression tests ran cleanly with this patch +Ok I understand how this fixes the issue now +Heres a working patch for safekeeping +Attaching the patch +Will this slow down assignmentsFor trunk it is not a problem since we have up to zk event workers by default +With Chromium under the new Ubuntu the quickstart project works fine +seems that any change in the archive definition adds a new entry in the dialog Ive just tried it with adding a new fileset +Its actually quite important because most hard disk marketing uses so we are out of the norm and there have been class action lawsuits over the misuse of GB vs. +I too have noticed this +You might even discover the cause of the problem yourself this way +Perhaps you didnt shutdown orderly +changes entryRight I always forget to include it in the patch and add it only afterwords should change thatAlso I am not comfortable with the use of a config property in to tell that facets should be added +nobody wiped any groovy artifacts what seems to have happened is that they added a pom for groovy that never had one +Thanks +will catch above and return null for get I think it is okYes +I can live without it after all and that problem doesnt exist anymore +Version has been releasedLets close this issue. +Temporary Ignore in +classpath for the EJB project +Uwe thanks I think though there was an issue even trying to use that +The Default implementation ignores that parameter but anyone providing their own factory may need it as I do to vary my black list behavior based on the identity of the client +is a task and has several sub tasks to collect the issues which are more or less the same +Unless someone can provide a reason why this is important I also feel an adaptive mechanism is overkill +This is a known problem with having an incompatible Batik jar in theclasspath +Bulk defer all frontend issues to +This directly delegates the responsibility and task of dealing with the Bytes to the client code +Not a bug. +I think Philips and Sureshs suggestion makes sense +Rebase the patch + +I test the case against the +With we should use and together with the for transforms. +IanJust for your information I have started updating the launchpad description +For this reason JBPAPP can be closed only by QA +This shouldnt occur but Ive now had this permission problem reported twice so its possible a umask has changed for the main apache ve fixed it for now will close the issue when I have a permanent fix +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Getting closer + +ANTLR seems like a good candidate +japanese translation update +It would be better to override the property rather than changing the componentid to do this attaching a patch +Which would be a better solution? To log t and throw t? Or to log t and throw t +Hey Richard Ive just uploaded a snapshot taken from my branch of the issue +Working on updating the patch +Manually tested! +If ever you change the client library interface then the examples may prove useful still by exposing the changesWhen you do this can you also bring williams example in jiri +It looks to me that we have some conflicting dependencies +fixed in recent trunks. +Note that my patch circumvents the dependency on the JRE +This could ultimately sour them on Struts and being a Struts supporter I would hate to see that happen +Committed this to trunk and +Behaviour by design. +How about adding a page to the wiki to keep track of the DSL development progress and documenting it as we move along? I guess we could use JIRA as well but the list might grow very long and it would probably just be cluttering JIRA and the mailing list +Previous we had an explicit sizing +It could be that the cache files cannot be written to or are not being closed for some reason which may eventually lead to too many open files and a system you see any errors in the error log +Everything in the plugins directory will be automatically be package for inclusion in the plugins directory and same goes for the feature +Similarily AND would yield all docs with the term in the bounding box rightAgain I am likely missing something so bear with me +Therefore Im going to close this as WORKSFORME. +I dont like the use of the name clone here +I just ran an analysis on the title field in the trunk example and the first step in the output was in this order text start end type position +While it certainly is a nice addition I have some reservations because running a GUI application on the server side is problematic unless the server happens to be your own boxTherefore I would see this rather as a temporary fix for issues debugging server sideIn this respect the TODO is not a better configuration mechanism but a better debugging implementation +Hairong also mentioned that she conducted a test where the fault was observed without the patch and the fault was not observed with the patch +In my situation I had to dig into thrift to figure out why my ruby client couldnt interface with a php server +Attaching an initial patch regenerated for latest trunk +No reply closing this as a wontfix +Mark as fixed and send to Svante for review +screen edited Now its extending +emailed again on Oct. +And futhermore connection should be surely closed though Database was earlier closed and Exception happens in rollback +Just for the purpose of discussion I think suffixes could be better than prefixes +replicated +Are stream object neededDoe this apply to includeTool as well as includeTargets +In both cases is the message could be set to the original exceptions +Minas swallowing of the exceptions prevents fully handling this problem +Here is a patch for revision +move along +Added support in for and +Id tested out latest build today the problem still exists +Could you add to the test a check for the order of the records +Setting the leftright margins equally in all fo fixes the error +Ill take a look for + for Peters first proposed alternative solution +I havent gotten around to this one so its better to get this one rolling +I fixed the problem yesterday afternoon by suppressing the test if nodelocalhost +Public Domain +Bulk move of M to as duplicate since we should make sure resources are aligned and hopefully by using the same codebase +csmatches all +Is this still a problem I think this problem has been fixed in open source some time backIf yes can you close the jira +fixed by using objectClass index asnd having it present by +For gang scheduling I can request all of the containers and just have them sitting waiting until the final one shows up +if no is present in a request servers must respond +Thanks Vikram +Will have to investigate that but you can try to manually add it and see if it would resolve your issue +Thanks + +Congrats! +What made you go down the Thrift way as opposed to packing this information as Flume Attributes? I had envisaged packing all the MDCProperties with prefixed keys so a foo MDC value would be keyed as a Flume attribute to disambiguate it +I see that someone else doesnt think that numeric entries are not bad +Changing the paradigm is thus a more global affair and also it is in accordance with taste +PengYou should be able to create a relatively efficient container structure using techniques such as are used in the Mahout collections library +Shouldnt be any need for the CSV if this works +reopened to remove tbd from fix versions +Final patch. +excellent! +In terms of the server option for Windows +The error came from your +I tried out the v patch and encountered no problems or test failures +Sorry I cant reproduce that it seems like a file could not be loaded I actually implemented this dialog to have at least the URL so you can simply copy the URL and then check what file exactly was requested +I always feel a bit icky myself when I start mucking with threads and typically only do so as a last resort +Now its fixed +Hi amar I marked this jira as being blocked by +Thanks! +Does it still make sense +Please consider the attached patch for inclusion in the next release +Ill do that in the next few daysAnother work item is to automate the generation of the Eclipse Ganymede repository metadata from the update site with the Maven build +The plugin in the generated is no longer needed and can be removed all together +Hi For the purposes of release notes would a fair summation of this bug beWhen an interface defined both in an EAR and in another jar in the deploy directory is called from within the EAR the wrong classloader was called causing the external interface to be used and causing a Exception +You are trying to use json path format in the cli to communicate with the resource? If so this is not a bug +Looks OKTested +So thats out of the questionMINAIn MINA you could pass an implementation of the Executor interface when you create theSo its possible to do it +With the menu expanded +HiI made the patch of fairness connection pool for trunkI ran tests with following settings +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +lgtm +You probably need to use munge to solve this +Thanks Marc +In this case the affected version is and the fix version is. +Any idea where the problem could come from and how to resolve itThanks + +This patch is for the new functionality for shared render parameter +for too +Thats nice but the current reality is the files are published independently +if it is stripe will be split into and fractions etc +on patch v +I have attached an improved javadoc for the Rule class +I will verify this when I return form vacation +Since items puled from disk are put into memory if the memory cache is full and you get an item off disk the lest recently used item will be spooled to disk +The only thing its seems to help with is avoiding walog recovery +I know this is a super old issue but there is referenced a Patch file that is not a valid patch file any longer +assignment invalidIm looking into the above some +Do you have an example project that we can use to verify this issueIt would also be good is you could supply more information about your environment like Java version Maven version etc +Its not giving me the option so can you please follow through +Hi AaronI changed the protocol did some testing and pushed the changes to and the specific commits into Let me know if you see any issues +Do we try to attack this before lightweights by wrapping the exception in an appropriate exception that mirrors the Java exception hierarchy or do we wait for lightweights to allow just handling all exceptions the same way +Updated patch adds a test suite support for configuring max in the URL and configureing as one of the default patch +fix +Most likely I will keep both as Google might decide to switch back to the old format +For information the problem doesnt exist with the the zip attached filesThe eclipse console log with this projects CET BBjar in Eclipse Workspace DworkspaceB CET Removed model B BB CET Updated model B BB CET Reading B CET AAjar in Eclipse Workspace DworkspaceA CET Removed model A AA CET Updated model A AA CET Reading A CET BBjar in Eclipse Workspace DworkspaceB CET Removed model B BB CET Updated model B BB CET Reading B CET AAjar in Eclipse Workspace DworkspaceA CET Removed model A AA CET Updated model A AA CET Reading A CET BBjar in Eclipse Workspace DworkspaceB CET Removed model B BB CET Updated model B BB CET Reading B CET AAjar in Eclipse Workspace DworkspaceA CET Removed model A AA CET Updated model A AA CET Reading A least I think it should be configurable +Patch +Yes exactly +I think its considered good JSON practice to have key names not contain state but I cant remember where I saw that +At the moment the jars have different package structures due to a cvs movement +Just musing that it would be nice for this patch and the other upgrade code because I dont think we explicitly test that stuff anywhere +I did some tests by changing the comments +Hope I explained it better +committed thanks for your patience on this one Eli. +changes made +This was discussed on list subject Eviction maxEntries analysis +dfgdf +Since the entire index is getting rebuilt I can think of no reason not to gather the statistics and record them at this time +Refreshing the patch as it was getting old for a long test for the new patch +Fixed in in release. +Thanks for your attention anyway +It shouldnt be our goal to fix all test issues on windows here so it is great if it already solves some +is now supported in latest version please try again +Thanks! +In our current project we have more than violations but due to this limitation we are unable to retrieve them s unfortunate that this limit applies to paginated searches as well +Cheers Notes Version Version container uses xml parser rather than war deployment included parser +The Kwalify parser gives us which is far more useful to the user +looking good over here tooWill commit +a patch with following changes +Could you verify that +Thanks Frank! +Lesson learned run the interop testsAlso once goes in I can replace all of those named imports with from avro import +Now that work is well underway with AS all previous community releases are +Im no longer using this in the application where I was having problemswith it so I cant help retestdebug it in production any more +ThanksSo given that we are both aware the zookeeper jar wouldnt be picked up as is because of the versionless zk jar in hadoop +I wrote unit tests in our own local test harness to verify the fix +As someone on the mailing list pointed out my workaround is a poor one +Youll need and libs in the lib directory +Bugs affected are this point we have no plans to create this patch. +Thx +Let me know if you have issues with getting SP on the build machine +Therefore the future date timestamp is something that needs to be checked on the server side +Punting to + +Lily please try this program with your modified server to see if it pops the bug +OK is it an existing rule or a new oneCan you also provide a repository export so I can replicate your issue +Was fixed as part of +Unzip +To reenable GPD deployment map the servlet to the path expected by the GPD namely upload +Thx again Derek +Thanks Jitendra! +This change did not get in time for +access control implies either an incredibly large amount of per file or implies HDFS having semantic understanding of the files it stores +What Im doing here is compatible with that I think +Basically allows you to hook code into preexisting code if I understand it correctly +However I am not totally sure it justifies setting isolation to TRANSACTIONUNKNOWN +I just committed this. +I was not sure how to adjust for that +thanks for the again for patch +Even therere K containers of one the application the file size seems to be still acceptable +Seems hudson is down +Shane a very good question +Sorry for the confusionThanksGeorge +To use straight octal we either need a flag like a leading o or use a new attribute and handle the old attribute in decimalMy other concern in this area is against supporting octal via a leading on all integer attributes +You think a large if statement is fragile and yet think a mapping between event names and listener interfaces isnt! Well given that it would be backed by proper regression tests yesAt least it would be less fragile than the currently codeIll be the first to admit that my proposed solution was myopic in that it was only concerned with solving and preventing the regression identified in this bug reportRegards +is present in the buildlib directory +If you agree we can rename the subtask you already created +I think the lacking space Kim referred to was in the shutdown engine message which looks like this in the attachedTue Oct PDT Shutting down Derby EngineI agree with Kim that there should be a space between the colon and Shutting so I took the liberty of checking in the attached patch +I propose we commit some variant of it and move the discussion of generalizing serialization over to a new jira +I just committed this +Hopefully this gets fixed inIs there an official reopen or is it fixed in trunk right now +How would a user of LARQ find the examples +Do you have an example where the current code causes problems? It might be helpful for the reviewers to understand what kind of problem the patch fixes +Is this issue still valid? If not can we close it out? I think the related bug was addressed by and the other part of this issue was a misconfiguration +I agree with Jun on the info log level for the delete messages +branch revision +The end goal is definitely to have users be able to just turn on encryption +So only the CREATE TABLE statements are desired in this scenario +Alex My bad I was doing a quick sanity comparison and ended up comparing against version instead of the rd version of the patch +Im starting to work in this have done the first commit really this isnt a plugin yet because is not installable in Sonar when we decide that is prepared I have to adapt some things +I also verified that the IP put into the token file is correct ie the one in the config file in the case your host has IP aliasing turned on +This is the patch in unified format +As Camel is using CXF there are some difference of the bus between CXF and CXFIf you are using the CXF you have to wire the CXF components with the bus yourself so I didnt back port the patch into. +I am for closing it +fieldName city notNull sizeMax add field reference jpa class +Strange not sure why it would take seconds if the Callable sleeps for or does your real code wait for seconds? Either way if you could please set to log level TRACE and provide log output from the execution of the request +If you could copypaste a quick example of such a file that works with the above test case that would be great +Thanks Amareshwari +Run ant to verify that evaluation works in restricted mode +Thanks for this fix +So I am running tests now to see all is okay before committingThe fix in core is because you use in the unit tests which didnt let the component be started before it creates the endpoint +The script just takes the lastest patch +Ill be sure to test with RC. +ThanksDo you think the patch could be committed in the different branchesIf I can help just ask +Committed to trunk +In other words it does use a bit more memory so its not necessarily surprising that is not enoughbqtest fails because of junit timeout when I run it together with all other suits but passes when I only run suit +context URI +brian +Fixed with commit +I cant remember but I think the problem was that it wasnt possible to define relative directories for lib element + second this since I use the deprecated and cant migrate my code since I need the serialization +never mind +A similar issue was resolved in version let us know if you have answers to the questions +which seems to break it +Problem is that Jives plugin incorrectly take referer attribute +Fix applied in r +Seems like a few things broke or changed behaviour due to +Grid we use is not dynamic resizable +needed to be fixedAll test successful as of noon +I made a mistake on local testing +Jason has said that the system properties are already passed in so only need to restore backwards compat for env turned back on IT reactivated for. +The change looks OK but the patch isnt generated with respect to trunk +we dont duplicate the classpath from the project if you use the default of just looking up the classpath up from the project which is stored in +Awating feedback before closing +DaveThis is our usecase We have master and slaves running remotely on separate +IMO the escaping should be done where its needed in the JSP siteThe link of Mario seems to be a nice feature +Works for me in Firefox on +Carl It is working after removing the symlinks from +FYI Im tagging this CR later this evening as it is very close to to note the interop tests should not be run by CC wondering why they turn up +may carry RETAINASSIGNMENT and SKIPTABLESTATECHECKWhen hbck detects table stuck in enabling disabling state it can use the new API to fix the problem +Mind if I commit this to fix the test for the time being and file for these hadoop SCR issues and the perf regressions in +Doesnt seem necessary I removed from the fixVersion +I will review my work +New binary build of with fix applied +This same problem exists in Java and probably Ruby as well +Clearly the docs are now inconsistent with the behavior of ESB +Also note that I havent spent much on this so it might be anything from Arquillian unable to deploy to which has a zombie thread serverUUID +That issue should already be fixed as part of +What do you think of this Daniel? Is there a good reason NOT to make thesedata members protected or provide settersgetters for themAn alternative idea if Jose is willing might be for Jose to make localmodifications and then once hes gotten his working submit it as apatch to be added to +I agree we should not be duplicating anything which the OS does anyway +Removing the status as resolved for +People would be enabled to provide more information on issues they findIf you could create an Eclipse plugin with source and debug classes now that would be really absolutely greatThe Eclipse plugin doesnt need to have a small footprint anyway +Open the web console and see if you can spot the difference +bump +piggyback lac when reading entries for reading lac. +We only removed the row completely so CS was able to start againSo it seems CS sets the value unencrypted here. +Thanks Mridul for the contribution Done at rAshish +We need to verify on these machines +Attached in the current HEAD. +This seems like it would be easy for the user to understand and specify in a test +Assigning fix version to the next milestone +Looks good to me +In this issue the it test project is used as it demonstrates the bug +If it does please zip up your workspace if theres nothing proprietary in it and sent it to me and Ill take a look at it +hmmm +We already have a bunch of ideas on how to do it +Verify if there are snapshots in state in the secondary storage and not deleted prior to upgrade appropriate entries get created in snapshotstoreref table +commited in r. +Duplicate with was also closed as a dupe +Assigning to Viliam to get the nav portlet to work after a ajax refresh +I know how to use Writeable interface +If someone wants to pick up the task of changing the localization system I think it would be good to create a separate Jira for that. +where can I see the updated release notes +Contegix was unable to fix this without Atlassian support +No problem +thought I cannot say I exactly remember why I went for the standalone page + is the timeframe for the upstream fix? Im not familiar with the Apache voting process +Can have a look during the next weekendif that helpsWerner +patch committed +You will need to have at least the coreplugin as a dependency +Please let me know if you have further concerns patch against works for me +I was surprised that I got this error message +If you have any questions please post them to the user list. +CDH Hive uses because at the time the Thrift JAR was unversioned and we needed a version numberHive currently uses Thrift as far as I know +We have possibilities to handle this issue Add to documentation that we do not support maps together with lazy loading Implement a that handles above caseAs Im not sure if lazy loading of maps would be required I tend to follow Werners suggestion to go for and see if someone requests this feature again +Commit changes to branch at revision revision at branch case has failures caused by changes to jndi names need to finally check it before release +Enables hsynchflush paths by default Adds a entry to Removes explicit enablement of from tests so we cover that appendis enabled by default +We dont currently plan to make further changes to other than security fixes so hopefully will address this for you when it is released +no i understood correctly +Colinthe lack of support for ajax is due to the browser security I know there are a bunch of tricks for overcoming these restrictions Ill take a look at them for the release. +I dont think we necessarily need an annotation if we simply expect that its a CDI bean registered with the bean manager +Tree created +you can change the title of the issue so that it reflects what is the actual move fixing the version info for and all affected issues have in +I dont know why its dependent for JDO thats something we need to look at +Let me try it one more time +Ive got the same error when I was trying to run the example on the computer with two virtual machines installed Sun Java and Sun Java this is the problem with Java? Or maybe this was associated with mismatch of +Looks like this one made it into +Hi Please release a binary version with this fix +Patch with changed impl +this isnt going to work except when the key requested happens to land on the local node according to RUSreally we need a if were going to expose this over Thrift +Anybody got any issues concerns in applying this patch for release? +Anyone knows what the cause could be +Applied with encoding +Im having the same issue as Sergii above had to downgrade to get Transfer working again +I am concerned about breaking those applications with this change +Note also that whoever wrote the code to assign a default mode also got bit by this bug because when no mode is specified you still get mode +Attaching the patch with fixes +Do you have Cygwin installed? Is df working in your cygwin installationDo you run crawl from cygwin shellNutch requires cygwin on Windows +After it should be possible for to not be rooted at +I just released an artifact where I was asked for the proper version of the SNAPSHOT parent but the given value was simply ignored and the release contained still the parent POM as SNAPSHOT version making the complete release invalid +Thats very odd +It seems like this is actually an issue with the Widget Signer toolWhen I extract the +It looks like the most desirable solution +configurable is a though it probably doesnt change the way that the underlying conveyor belt API and only affects the cache implementation itself the bigger question is how do you want this configured and where should it go? Im keen for this to remain configurable on a basis +Conversion has already begun on trunk +The problem is that Trinidads wraps the as well allowing functionality for a special resource resolver on the dispatch method +Done +Not too different from previous patches +This works with the same method and the same process file with jBPM s manifest says that it is and dont have versions in their the jaxp webpage I shouldnt even need these with java so Im going to try again without them and see how we note my process definitions normally dont have a namespace so shouldnt these simply skip schema validation altogether +Right now the order of execution is as follows +Hi WoonsanYou are right in fact it seems that the Spring project is moving to Maven +It should be required field in case Yes option selected and if Java was selected on th step this field should has the same path by default +Both of them can be optimized to use one query when adding JOIN FETCH +updated patch has been committed thanks for the fixes! +as I stopped to work on this patchI also need to write a test for the distributed timeout scenario as that is the same problem +What you post should most definitely work with EL post a patch against one of the examples so we can replicate locally. +What unused parameters +New commands should have their own separate issues. +Rebased version with several bugs fixes +But I think its equally important to have an outline of the code changes being made especially given that they were made very close to the release +Hi BJId be happy to commit I guess its only used with the long help page since there are no access from eCommerce right +Retrievinguploading build numbermetadata of a simple project needs between and minutes during deployment of a snaphot version +However I did not commit the test reuters data the jscalenderRather then include the large example data in the distribution we should add a script that will download and import it for people who really want to +There is a difference between red and currentColor though the latter uses the paintTypes SVGPAINTTYPECURRENTCOLOR and SVGPAINTTYPEURICURRENTCOLOR not any of the RGBCOLOR paintTypes +Attached is a proposed patch +Version d did not run properly and I agree with Bill that it may not be a good idea +Attached file with whole stack +Do we know what issue fixed this failing test? We should backport it to branch +All tests passed +to Eddie for in SVN +After the release maybe we spend some time backporting the git ref stuff to cordovaThen we change the docs to download using refs effectively killing our reliance on master +Want to raise a conversation out on mailing list? +While I have Woodstox on my classpath I have only recently figured out that its not being usedDo you know if there is a way to have more than one implementation on the classpath and direct CXF to use the Woodstox oneCannot reproduce with latest code using Sun parser. +Included in. +The latter can be done without waiting for the former yet still provides great usability benefits +Both do the same +rev hooks up the injection framework to. +My understanding was you can not quickfixmodify files external to the +code in the cvs handle anyType +consolesrcmainwebappviewsecurityrealmmanagerderbygroups +yep +Thanks for the reply +But there are still some test cases that come with this patch that fail without the bug fix +I will look at getting this fix in but plan to change the patch such that only XL is returned in either case and somehow the text of the error is changed based on the diagnostic level +updated fixfor to by unarchving the version in the project admin JIRA settings +Committed revision patches for this JIRA. +Ill update the issue with the commands I used to create the clustershould try the following in case your keys arent working right t rsa P f +RickCommitted patch to documentation trunk at revision +Bulk closing stale resolved issues +Looks like the property is not in the correct format which is why the is not skippedIf we use property for this that means that someone who tries to disable another test on commandline would accidentally enable the beeline tests +The search feature was meant to become the core part of the frontend to make it more useful for users +yet to upload the latest changes but whats there +Ill check it once more +Sorry not a defect +I ran the test times with the changes and it passed each timeThe test still failed with just the waits increased the added waits were necessary +should be recognized +Thanks for the help has been resolved for some time the pear channel is working +WillHeap dump is around GB i dont think i can attach it here? Can i put it in any dropbox +Looks great! Im testing the change now on +There were a lot of changesfixes in and its possible something you familiar with Git? If so you could clone our repository and start bisecting revisions between and to see which commit broke it +I added new tests in but should probably add more before we commit this +Ready for release. +but the jackrabbit implementation is based on rmi? and the jackrabbit webdav client seems to exist mostly of interfaces +For support of data roles in Teiid take look at the Data Roles chapter in the Reference Guide. +Its in +Ive committed a new patch to which should fix the problem reportedhere and should fix the two xsltc failures mdocs and mdocs which the first patch introducedIm leaving the bug report open so I can continue to investigate +I dont have good access to the SVN right now +Thanks Vinay! Ive committed this to trunk and +Sets work except when there are enumsThis is very confusing +Im actually only removing the in +I was just adding a key to the doc called updatedat with the current time LanceI will give the updatedat a try and add it to the API documentation this weekend as a workaround +Well ensure logging for the events but from the thread in cases where something is going to be programatically triggered off of the event it seems appropriate to register a listener instance +On the other hand I can not simply assign this issue to someone else +In my opinion it is not wrong to save a link listed as a file +It looks fine +deffered to this planned for +Any issues please open a new bug report +treats the reply from async as a response and then throws a this gets translated in deliverAsync to a is nothing to do with the pipeline although that would certainly see it as this would happen with any client sending a fault through use of the response to indicate a successful delivery is as you say very dodgy coding +no FO input file provided insufficiently specific problem transition resolvedwontfix to closedwontfix +Designed the queries to retrieve unfinished Receive and Pick activities +You didnt close the issue as wont fix or invalid its because there is room for that rule to change +updated patch +Marking closed. +Ivan can you check if my change also solves your issue +We need to add a test for this issue +Closing. +Accepted +I can fix them on commit +is missing the behaviour is like before +Clearly I cant tell you remote search people apart dont know what it was I just with current trunk after the refactorings of the most recent few patches +This is more or less a duplicate of +I will run some tests and commit it +Tagged for that is regardless of whether we take it in or not it should be resolved by +was this fixable +It is! No more dependency on Java +If this happens concurrently then we get an exception +I do not want to go into discussion where is really low probability that change would be done +Attaching patch for typo in in revision + +This is an example +Joshua I applied the patch after testing stuff reasonably thoroughly. +Patch is really simple and nice +is another patch which disables shortcircuit read +Not sure if Drools willl support varargs at the moment give it a ago +Karl did you ever proceed on this patch? Im interested in adding autosuggest to Solr +Thanks for the debugging work on thisIs it ok for us to put the test files into under the ASL license +But it looks like that jar might be might be no +Any other value will cause the image to be compressed which leads to the issueIn order to fix the compression problem Im going to open another ticket where we will need to write our own jpg compression algorithm. +Usability and customer sat is important thus this is made a blocker for SOA +configuration file changes are easier to test so we can see that the file gets checked. +Got it thanks for the clarification +I am currently adding this featureBut still the classes should be loaderd from the CCL or youshould get an CNF exception +So I am reluctant to make this change unless I am convinced there are several developers who track trunk changes using a shHow hard is it to install bash on +Committing +Attached is a patch to that might work +Attaching patch to fix +just committed this to trunk +Heres a backport I did to a few weeks back with intent of doing some benchmarking but havent had time to follow through with yetAttaching as +Even with the pagination filter eventually you have to retrieve the rows at the client +Thanks! +Maybe some dynamic runtime binding +I cant help that I can just keep helping make them hardierAnd if contributors have questions about a fail I can give some guidance depending on the test and failI mean hopefully we get to very few fails but its a constant moving battle +works as expected. +still broken for me +Verified and closed. +Hi Claudio great stuffA couple of questionsI downloaded and think you are missing and from the diff +Revision +Page n of nRegards has been added with description example for page x of y +I almost have completed a patch for this Gertjan to test it before committing it +This patch is just to update cookiesIn case of replacement of disks data copy should be done by admin itself before calling the admin command. + +So the implementation does not just copy and paste metrics from Hadoop +I have tested this patch in a manual way as described in my earlier comment by using and iptables to simulate network problems +Will commit if tests pass +Applied to branch and branch of branch +Corrects the parent artifactId grants license to ASF +I can not exactly say that it will work good enough +Setting resolution to Wont Fix. +implementation for for the patch +You mean btw agree thats the right approach +Im not sure if this worked even before the recent changes +Shouldnt it be a integer so that we get validation done for free +I wish I could confirm this patch but my notebook couldnt handle shuffling of G data +This should apply to XML output only +Seems that the solution is in a more expressive registration system +No I have put the logger in Debug mode but nothing in Archiva logs +i do not believe that this is an issue that has been introduced in the release +It will require to work quite hard on the AST translator phases changing the way queries are generated leading to lots of QA before releasing that +patch +Fixed this in a previous SVN revision now a test is there as well as of SVN +But for now its duplicated +Better startI guess if we are then we dont really need tointegrate at indexing time? Ie app can just add the facet valuesusingbq +I would have just submitted this but subversion is down +Yes thats true +Sorry my mistake +Well candidates should not be null at this point so it is difficult for me to diagnose without being able to reproduce +Im also not sure I understand the comment about the NFE propagating back in the call chain +Please try installation again and let me know if this works around the need for either install upstream first or add upstream site to resolve deps +Hi Guillaume those two tickets have been closed a long time ago +I would like to retain the current behavior where arguments are checked and an IAE with informative message is thrown when an argument is nullThere is value to checking arguments and throwing IAE when arguments not expected to be null are null +the first rough cut of the packaging +In general the read method needs during the indexing process +Your startup method is private +That involved many tweaks throughout the class +Thats why there was no feedback from dev team +organizing fix versions for helios bug is replicatible still with ESB and Dynamic web projects +Levina there is no way we can know what happened on your machine and it may well be some error on your sideBefore opening a ticket please discuss it on the users mailing list and provide more details about it so that we can all issues that have been in resolved state for more than one month without further comments to closed status +Ill help you if I can +updated to current trunk +Im having the exact same problem with +I cant tell if its right or not but I leave out entries that produce exceptionsI also just commented out the tests around the id field they didnt look valid +for this +Right sorry for overlooking this issue! The proposed solution is indeed included in the changes so resolving as a duplicateI committed the test case with slight modifications in revision +This will avoid any future accidental duplicate topics bug and also avoids the conversion from Seq Set Seq +And it works in second case +Are there any other commands dicovered to be changed in similoar way +Im off today but will test it some more later this evening +Excellent thanks its working nowColm +With this patch a client has only one per Bookie +It will be done shortly. +It should be fixed in +Wouldnt be the obvious place? It iterates through and could add PERSIST each time the end of a bundle was reached +Id like to remove the fault flag in the object type message +There are some more to be replaced with for this part to be complete +NN keeps the block on pendingReplications +uploading patch from Nicolas this patch has been marked sa Resolve but I do not see any svn commits + +Let me know if there are any problems +For some reason I couldnt upload the straight patch +has the attributeso i think to put the prefix of the metric name to the group namefor example metric groupadd group attribute which use the prefix of the metric name +In this test case receiving a is expected behavior +Hi Raf +Prevent default node for Dashboard +Have you looked at the used by the deployment scanner +Should be definitely fixed on final +Its messy but thats why we have tag parameters to let you sort it out +Comment out all the Collector related methods including the new ctor with a TODO to reenstate in +Thanks! +Theres no expectation that a TX would exist in that context +Would it be possible for you to include a Java driver program that demonstrates the problem? Henry +Several years ago when I looked at this there was only a check in to see if the limit was exceeded but otherwise Lucene was in no way dependent on the value of that limitSo I recommend removing support for the limit from Lucene as wellActually I would recommend deprecating the max feature of both Lucene and Solr in and removing the feature from trunkAs a backup plan if the Lucene guys dont want to remove the limit Solr should simply set the limit to but otherwise remove all references to the feature in Solr itself +If I remove the static attribute of that class the memory footprint is greatly reduced and everything works fineMy document count is around millionAny help would be greatly appreciatedThank you +Attached v with unit test +This is exactly what I was thinking of +Im sorry for the noise +The Feature List chapter is created for all guides +I found a better solution which worked for me I was using Internet Explorer on the machine that showed this problem I upgrade this to IE and the problem disappeared +This issue is fixed and released as part of release +I can reproduce it in as it really works if you change url pattern from +I agree that accessing the request instance object in the session scoped bean outside of a request would cause this issue +I think were all on the same page then +If someone from packaging can take a look on how it is done as per new packaging for other scripts or share the same we will use just as a wrapper which does some processing and executesLastly whats the point of a code freeze and ml discussion if we decided not to touch it for and we keep pushing commits around it +Open Internet Explorer and access httplocalhost +The real underlying problem is that tracking URL is a string and not a URL Too late to make incompatible changes but Ill file a ticketFiled +Showing a long line command that has been up arrowed too +I think this issue conflates two aspects +Thanks can you apply the patch to? Thanks +It is getting late here so I dont have a chance to do it tonight +If you agree then can I ask why we moved to the more generic Process what is the benefit to our customersThanks thanks for that +Then we can add the solution to the check of indirect resolving bundles later on +As it turns out the support was limited to particular test cases +Makes test pass +mrepositoryorgapachemavenplugins do you still haveproblems? It seems like a core bug not an antrun bug +Not a blocker pushing to think with this should not be an issues anymore. +Interesting if youre able to reproduce this while running with mvn X clean test we might be able to get somewhere on this +Affects Spring DM resolved through a version import for that package +The test has been passing. +Depends on my patch in +Ilya please check the and highlight unnecessary attributes +Here is one such scenarioTowards the end of the reduce phase speculative tasks were launched for some reducers +Thanks Olivier +It is necessary to change the file +There is still an issue where the manifested jar can not see user defined modules +patch applies cleanly on trunk has minor conflicts in imports that can be address on commit +my projects are the Castor projects +Indeed the docs use WFS and Im not entirely crazy +Already solved by is not exactly the same as this one +This is the version of the patch rewritten to use the new mapreduce API in hadoop +The other backward incompatibility change is Pig will not runnable without a local hadoop installation +and then use the same password in all three boxes I am shown the error stating you cant use the same password +I also tried thesechancurrentccurrentcurrentefwerwerwerrrrcurrentall of these produce the same complete changes list +Ill tweak it shortly +Somehow a line was accidentally deleted +Cool +looks good +thanks gavin +Added for +Could you review the usage doc for JSW i just cook something up and need a doc corrections attached +I just committed this +The fix doesnt seem to fix the issue just extend the number of extensions it supports shouldnt this be done via contenttypes +Committed at subversion revision +can be reloaded at runtime as well +Sorry but I dont understand the error message should be shown on the attempt that it gets locked +Dunno how best to do this +Bulk move issues to and oh +persists after flags are removed +We changed it because if Bar had for example also a p property then your output would in still show the one in Foo p property +The spec does not provide any mechanism for the bean to identify if it can call set so it has completely prohibited a bean requires to do something transactional then it should not be using the transaction attribute of Supports Supports just lets an existing transaction pass through the bean without forcing a transaction to exist to call the bean +Just to be clear you will look into the tests failing on Chrome? +Fix attached so that unacknowledge messages are tracked individually by each consumerUnit test case for previous issue attached +I think it would be better to do it in series so that we dont overload the server with a lot of read requests at once +created +Mass closing all issues that have been in resolved state for months or more without any feedback or update +v deleted some valid code from some Python files handled a few of the csharp files weirdly removed some package declarations from some php files added a header to etc +Small change Moved the check for child from forward to initialize +Ive been looking at recent updates to the ref guide via the commits alias +Windows FlexBison appear to be a problem combination +fixed +This does not work for jbossws as long as is not Serializable +And for example the unit testing template code is impossible to edit with an IDEIt seems the reason behind using the preprocessor is to support multiple Hadoop versions? Would it not be simpler and easier for both developers and the end user if for example Hive only worked with Hadoop? That way there wouldnt be a need for preprocessed code +You could give the latest snapshot a try to confirm or add and as explicit dependencies +the book was updated in the release. +The title should be more descriptive of an action such as Update docs based on script initializations It would have more clear +Not a big pb + +If you think the Napkin LAF has issues just think of the issues Ive run into with Spring Rich +actually I do have the metadata information in the mysqldump +Also if the chain history is a Linked List instead of a Set we could know the flow +No problem +This simple patch makes save and retrieve the halt value +I have added a property for ignoring the operation set +This is exactly why I said this configuration creates a powerful language rather than complicating it +finder class +If so we can close this issue and open a new one +We can replace the repository with another one which also contains the version updating to is not really an option without testing the coref component on itAny suggestions for a repository which contains? Otherwise we might want to get it into maven central +Hi I have made change to copy the core available to the maven repository +I have used official jar but i think android classloader dont expose resource files insideEdit yes it could be the reason i will also try with a more recent version of android +Should probably put some verify in there +looks good will commit if the tests tests are failing +Should be restored +Then we should use facets +But maybe it might prevent othersfrom running tests? It takes time to get M docs build indexeswith different settings get good queries +Im sorry but I have some mistakes in attached files and so use files and instead of old files +Thinks are working as they should for me +Obtain guide from ve added these changes to the EAP and EWP tags so these changes will make it into future versions of EAP and EWP +I think I misunderstood how the and the work internally +Attached the right patch now +What steps led to the problem? Wheres the stack trace +I confirm that this is a problem in the decoder +I have started to work on a file locking mechanismIts not activated per default but you can activate via the ui +Thats why I raised the ll withdraw my objection then +Ok I seeWell well see if we get someone to do the hard work for this wish Isnt one of the two of you willing to supply a patchregardsMartin +without seems to be building and tests passing +El Value field field was marked as mandatory field +Rod has offered to do this so Ive assigned it to him +Just my a radically different way. +r on trunk by applying the patch by Stefan Lang +I am in the process of trying and will update the ticket as soon as I get the I confirm this applies to trunk as also find another patch which should apply against trunk +SanjeevIf I see the latest master for all hypervisors except for KVM you would see cluster as the default scope with host in the drop down +In order to find the I ported all test cases +Thanks Mahadev for looking into this issue +I liked the way you isolated football for the SAP demo of spring batch +merged into dev commit cefda +Fixed in SVN rev +Thats the plan yes +We could workaround it in a Alpha release by using some simple reflection on the Assets until there is a API in place for it +We can get the changes here after we finish the initial code donationAnd yes Gordon TLF does not compile cleanly +This looks interesting +patch +Would need a full test case for this +Passing a null as a value throw an Exception when using jettyri which is a different issue than this one. +seems to be targeted to rc should see this fixed after a TP upgrade +Yeah your repo might change but its more of a cosmetic problem than anything else +user clicks on Add Spring Aspects Tools +test some different combinationsUNIX domain socketsUNIX domain sockets short circuitUNIX domain sockets short circuit skip speed issue with reads seems to be resolved in this latest patch +removed from the closed the pull request since Hibernate is now available +Sample wsdl that generates error +Integrated Averys suggestions +lets first make this Apples to Apples +Completely arbitraty indeed +Im still working through a test case showing the issue clearlyErik based on your followup comment i ignored your patch and attempted to write a test to reproduce the general problem you described and could not do so see attachedif there is a bug i suspect it must be something subtle in the way the defaults of a particular param are defined +I have no objections +have you tested this with +I will put the other nodes into debug mode tomorrow as well and post the DEBUG output of and +Thanks +New patch folding in Uwes ideas FCRF always uses random access remove matchDoc and just override fixed CWF jdocs to not reference +Runnable code +Byrce they tagged this your way but did not actually assign it +it actually sends an RPC to the ZKFC associated with the current active +Why dont we just ignore the houtputScript for on the server if the current request is an ajax request? This is just so much easier isnt it +Thanks for the link on Topics vs Queues +What the heck +It is unresonable to expects that Axis will work with a pluggable +Im also proposing that we create a testtools directory and put the and associatedfiles in that directory and remove it from the java section as jtidy is being used for testing purposesonly +Ive addressed this through the introduction of a jta bean property on now explicitly indicating JTA awareness +Right + +Still needs to be cleaned up a lot and at a lower level somehow you still have to wait for another thread to finish before writing your own stuff in the block right? So what youre telling me is that basically takes the grouping at a lower level +marking as patch available so that it gets to the attention of other committers +Please update the jira summary and description accordingly +See the nexus unzip plugin doesnt work on snapshots +In case of a network issue where the AM is running but cannot talk to the RM or say the NM on which the AM was running goes down what knob would control handling these situationsFor these two cases I would use the knob because the AM is perceived to have failed +chukwalib lead to some javadoc warnings +It doesnt look as hard as I first thought! +Ben sorry to be pestering Id like to get rc out today +on the other hand streamlining is not a bad thingthx for the test heads up! thats definitely a good one to work inyeah likely over refactored the packages a bit +Thanks +It this the solution that is going to be implemented in a next release or are you going for the of the transaction +Consider this for inclusion in of patch that will work w to and to TRUNK +Moreover the local variable fromclient is set to a value but is not used at all because a return is called immediately after +I added some minor cleanups and removed invalid nocommits +If it does not we should complain that HADOOPHOME needs setting +I guess we could switch the build to use the upstream version of the artifact the reason we didnt before was to have the same dependency as AS +has not received much interest lately but I am sure that Ken would be interested +if there was no overriding method then will return the ID of a method in the super class rather than one that belongs to the class in question +This topic will be covered in the introductory session where we introduce the various inventory concepts in the product +for the life of me I cannot remember how things get grabbed from the pico context into a binding +Thanks +I think I can get it to work with the old methods and constructors inthe factory classes +In my previous comment I meant file not +Hdfs throughput data calculation should process data for one extra time window in case the dfs datanode metrics arrives late +In Debian case it always runs regardless there is a user specified or not +Ive also included an integration test for this issue and some validation for the integration m not terribly pleased with the code but I dont want to refactor too much +Typically set to a prime several times greater than number of available hosts +I will try to make the page never render and submit a new patchThe issue with this is the check to see if the acceptdecline box should pop up for the user viewing the shared page for the first time doesnt happen until viewPage is called on that pageThanksDaniel is a second fix which does NOT render the content of the page at all instead of hiding and showing it +I wonder if it is easy to do to create temp directory on demand rather than upfront +bq +I further updated the test for both normal and faster iterators +Id vote for Not a Problem on this I think there are more pressing issues currently +This patch moves the optional tool out of the tools jar and into the engine jar +thankls bob +This was committed a while back. +Makes sense + +Resubmitting same patch to kick the build since it applies fine for me +I have confirmed performance increase from our QA team for POJO endpoints benchmarks +So currently its running +Hi Jim updated patch here +The point is a small paragraph in the manual can save a developer hours of work +Since we have added back a method cache in it seems that the performance difference has narrowed somewhat +Verified on EWS CR + for the patch +Patch to address the trivial task of substantiating comments to accompany +Hi Aidan Can you review this change to the assertion +The patch works! Thank you Arun +Note that Ive so far only tested the layout with Chrome and Firefox on Linux +changes for this were committed within jira revision +I think thats local to your wc +Trunk can be a little broken from time to time but this is pretty big +How do we specify the clearcase source repository for use with the changelog plugin +David go ahead and commit this +Thanks +Thanks for the patch Enis! And thanks for the reviews Stack and Ted +So Ive modified the name to which seems fair as it extends from an +Ive the same problem but with standard SNAPSHOT dependencies not using classifier and with +c is to specify which server config to start such as default production allb allows you to bind your webserver to IP otherthan localhostu allows you to find jgroups to certain udp is for the gtn mgmt alpha CLI tool not the app server +Fixed for +Probably related according to a Cas dump done from inside the service the is created during service processing although it has no content +looks great thanx for sticking with it +workaroud for another problem +My analyse was a little bit too much K feet high +to trunk +So for now I dont think there is much more to add in the driver page you patch looks fineThanks news on this? +This basically gives you both and +Focus has changes from creating a T to providing transaction id and named ledger support +Hi Christophyoure patch is in trunk Rev Many thanksSascha +Depending on how important it is we do have alternatives however +Ok closing. +also lists ivy code to add the dependencies +I am going to close in hours if there are no objections +Can you tell me a few things about your workload +Eric the command will be removed from the shell so it will not appear under activemq in the console +Added test case which is not enabled to run because it fails cache was not updated when version is removedFixed in revision to the branch in revision . +It would perhaps be helpful if the plugin could detect duplicates and only add them once +Leave it open until the JPA tests are fixed this is a known hole +to provide relevant functionality +Attaching an alternative patch that does that however +We can commit the patch +quick look it looks like this test is not expected to run in classes only environment vs +The build for petclinic is currently and has some tricky bits requires the spring distribution of petclinic to be copied in integrates groovy sources etc +Reuploading the patch to trigger the patch build +Yes +We should probably also look out for that case +This issue cannot be resolved yet since we need to correct some Japanese sentences and add a link to Solr wiki +Do you have a small testcase that demonstrates the problem and verifies the integrity of the patch +makes as discussed remote heartbeats will be issued only after startup was finished +Removes the dead code and fixes up some references to protocol in comments that are no longer valid +Furthermore Im still struggling to understand how the classpath config could cause hangingI do not remember the specifics +CXF plans to have it on its roadmap for CXF +It is up to Cassandra users to decide what is or isnt correct given their respective workloads and requirementsI understand that L improvements such as would greatly alleviate the pain but in the meantime implementing this simple option could greatly benefit some folks +This is problematic only because the Connection has no corresponding API +Are you familiar with the Hadoop api? Seems like thats the right way to do this +looks great! for type A +The test throws an exception running in embedded mode and runs fine in network mode +cancel last comment I was a bit there +With r I committed the patch provided by Martijn Dashorst with minor fixes and without JSON based request loggerThe updated JSON based logger is attached to the ticket and can be used by any user application +WHERE ANYWHERE ANY +As Botond mentioned in Zookeeper clients that use Kerberos must be able to handle mutual authentication with more than one Zookeeper server and also must cope with ticket expiry +Thanks for the patch +Sweet +Hi you give a but more info here how to replicate etc +Im working with the same hadoop version +It applies to the cocoa build as well can you take a look +I have a question regarding fixing our existing data +So you should review your configuration maven installation hudson configuration and so on +Any plugins that modify the classifier would just have to append to it instead of replacing it. +Arron does this sound reasonable? +It works for me +How many of the currently relevant wiki pages have been updated in that? Considering how little has changed since Apache was released almost two years not many because there was no needThis JIRA is a great example of something that will likely be long fixed before an actual Apache release with the fix sees the light of day +also fixes length bugs in all the testUtil string generation methods too +Fixes checked into the svn repositoriesAlbert Lee. +Just a quick note we cant really encode this sort of string in the database it really needs to be done in code that passes it through various types of limited string representations +Hi are there any solutions to this until now? mosaic background by default is set to a default value on each of the bands of the original mosaic which depends on the data type of the mosaic itself for byte and unsigned short for integer for Short for Float DoubleNAN for that if the original tiles are RGBA the whole and background in the resulting mosaic will default to transparent +Thanks Himanshu +Fixed in rAdditionally I have added a first unit test and fixed more javadoc issues in the class. + for the change +Im not sure if we are going to fix this on +vwt as a project name would be somewhat obtuse +Note that those configuration flags should only be set on configuration time not modified at runtime +This content types has the extension +The modules is picked up from files +Marking closed. +The is not related to this patch it is covered byTheres a reviewer for the patch so please considering committing this patch fixes the test its intended to fix +Ill resolve this one as dup given the other one has patches attached etc. +Your changes fix the issue using the attached unit test +While it seems not too hard to fix the parser the consequences of enabling those to be deployed are rather severe as all otherolder Maven versions will not be able to parse those during dependency resolution +isnt a jdk class +nothing present locally on node manager present on remote hdfsGood point +done +If this issue has been Resolved why is the Fix Versions field still TBD EAP ? I would have thought this would now have been changed to EAP +Oops my patch generation skills are rusty +you for pointing out this issue and providing the scripts +bulk close of resolved issues. +This issue was resolved on upstream +To register specific editors for action properties the action should be included into the set +from meI would like to find out about the previous rationale of placing the HBASECLASSPATH last though + +Mass closing all issues that have been in resolved state for months or more without any feedback or update +This patch does not apply anymore as Filters no longer use but +Mixed lists in the way that you are doing are not supported so this is not a bug but an improvement request +Please test and reopen this issue if necessary +Sure Ill do it thanks for pointing this +This change seems problematic +and marking the resource nondeployable does not necessarily mean you want it removed from a server +Patch to extend to multivalued fields +Forgot to unassign +hi Sergiycan you please attach file used for those test? peter were using that parameter anyway even changing it to true doesnt change have installed my own sybase server on my local machine and i have run tests not being able to simulate the error with missing i have downloaded the newest versin form sybase web like you have +Im sorry to say that I do currently not have the time to help with this +And how many bytes are really received is not to see directly +This is a nice feature since it allows you some flexability on what you want to do with the exception instead of just forwarding the exception to a page +Default on sounds fine to me its definitely in the intent of the rule +Iterating backwards through the names should resolve the problem. +static imports and static imports are now never removed from the import container +Might make sense to have this per CF or maybe per CF and a global default setting +I think this is just a documentation issue right +That might be true right now but if the was removed from all the car modules this still should continue to work because the classloading is delegated to all the bundles in the list instead of just one +Thanks for reporting and providing the. +Ah pressing ENTER then shows the jboss fuse welcome page +Ill forward a tracking number when they ship out on theCheyenne +I have not visited this page in a very long time since I assumed there was no interest after my initial upload of this code +If you have any questions when you try bytecode weaving feel free to ask on the forum. +So whats stopping you from using jar packaging +Patch without pattern override +Closed as duplicate of. +The tables no longer can be sorted and also the next pages cannot be viewed after running a little over jobs +After longer investigation Ive discovered that of the jar was filled with description fields content +Randall Leeds wrote a test case and a patch which Ive committed to both trunk and the branch. +Needs some testing +Sorry for the false alarm +Im pretty sure we discussed the bulk submission approach on the mailing list and we were OK with it +Can I just make sure I understand what Certified means here so I get it right for the release notesCertified means This has been tested to work with the rest of Developer Studio and associated technologies +On Windows I could frequently see exactly as belowWaiting for debugee exit +with with comments Thanks! +I think you are going to need to patch SOLR +Heres a new patch with the above changes +The fix +Regarding clone +Thx for your suggestions Andrew +Can we either get this committed or reverted? trunk build is currently broken +suggestion from cp authormy initial commentsdependent on so not backwards compatible works with cp +for list of places to test see linked case +I watch all the issues I create So where do we move the project +Can we close thisOk. + +If you want to subvert the spec the best way to do it is by changing the name of your setter so something like set +Well already have our own thats generating our own tokens so I was assuming that wed have our own that expects our versions of the token classes that it knows how to manipulate +Talk to +Attaching RM logs + patch needs to be tested very carefully +Changes look fineSubmitting for for Yahoo! distributionRan ant test and +Please verify +You may vote and comment on the other issue if youd like to see this feature included in iBatis +Though your tests indeed indicate otherwise +Written chars are encoded to system default encoding +Hi Kris Yes you are right it is taking time before applications are actually deployed +Attached a patch +Already fixed in in JBDS +I think we may enhance WAL compression using dictionary in the futureSo for DICTIONARY compression type it is desirable to introduce versioning as wellI dont have strong opinion about WALVERSION actually +Your test case comments looks good +Includes a few domain classes an integration test using the builder with the domain classes a file that builds a few of the domains controllers for each domain used to manually verify that using in the workedTo simplify and make the example app a file this also includes a copy of under src ScottOnce Groovy is out ill have a look at this as it extends which is not in the current version of Groovy in I think the package should be one thing +I can test snapshot too +Can you add a test for above +Does thatmean it can be committed soon +I was thinking along the same lines as you +Pull it back in if you think different. +Incorrectly assigned this to me +One question in a fault scenario lets say its trying to download some edits and one of the dirs fails +and note the most important thing is just to be able to get the named paramters names the position should just be an added bonus if its easy performance wise +Instead of whatThe returns the URL +Will fix this when I change for a related rpc jira +The patch already creates a new resultCache map each time a new permutation is checked +So I guess that there is something unclear here dont you think so? Sorry I have no time to look at it furtherJacquesJacques +To fix this issue I have to synchronize these methods such that only one can be executed at the same time for a given component instanceFixed in Rev +is rejected for mepatching file vmvmcoresrcinitHunk FAILED at Hunk FAILED at Hunk FAILED at +Category rename done +Probably more correct name would be OSTYPEWINSUPPOTINGFSUTIL but it is confusing +Your patch was for ecommerce component inside application folder but we dont have ecommerce component there now Ashish +We should define the dependencies better and move toward to more fine rule in the build files and have easier life every time we need to change thing to make Derby better +No dependencies on SNAPSHOTS so it should be close to being a candidate for a release in alpha form +Thanks Bryant could you help push a new wink snapshot to repo so that we could pick it up in our CI system? +Thanks Ning! +Test jsp +Anyway I managed a deploy with +Thanks for quick fix and response +Hi All The test case given seems to be obsolete as it sends a text message and tries to typecast it to an object message +The attached patch felt a little wrong as I dont know if you want a reference to the kept by the Module +corrected spelling while searching for another believe this has been fixed with the set flag in. +Will check with Nick +Besides representing an overhead this might introduce a bug +In fact I have used a sun JVM instead of IBM JVM and I didnt get the error anymore +This looks like a bug in +Clausthis resolved the issue +whatever your preference is +Applied at r +Let me know how this looks +This problem could be related to or a side effect of +Since its strictly additive no existing behavior would be affected +Working with compressed gifs is such pain in the a. +Will it be possible for someone to commit this in the SCM or allow me to do converter has been used to fix the current HEAD version of is using +Please close if this suits your needs +Mass closing all issues that have been in resolved state for months or more without any feedback or update +A variant of this code appears to have already been integrated +Please review again and let me know what you think +Removing tmp might be also very dangerous since there are many files there opened by other programs and things like pipes sessions etc +The problem could not be reproduced in +Thanks! +I just noticed that it wasnt used anymore +Do you have a stacktrace +We could either morph this one into that by changing the title or one could open a new one to enable the optimization and backpoint to this item to remind us what stress testissue needs to be looked at before deciding the optimization is safe to +I started to make copies of parts of export but I noticed most of the structures were svn checkouts so I stopped +Back in a bit +My only concern with this is that the slight differentiation from Python might confuse some programmers but the javadoc explains it clearly so I guess that shouldnt be much of an issue +That is where an additional official parameter should be introduced +Maybe Ive just been programming applets too long +builds are an orthogonal issue +Thanks to Cloves Carneiro Jr for providing the perfect test app for me to try it out and confirm both the breakage and the fix. +Committed +The dependency JBWS issue doesnt have an assignee as well +I think we should commit after changing the test output +Ill take care of that +Just some random thoughts on thisCould add some sort of date to be associated with ddl objects +might be a problem wherein a delay in detecting the process is started might give incorrect results +I attached proposed patch +Since the tests were passing even with that glaring mistake the coverage definitely wasnt good enough +yep hdt is do I need to do to make accountsOn the incubator wiki I used the wiki name so go ahead and use that for an admin accountOnce Im setup as admin will I be able to make other people adminsCheers my account name is once Im in the admin group I should be able to take care of things from there +i will double check that +This does not happen with another installation of designer on a windows machine where there is a different dialog with two annoying fields for specifying the TEIID core and version with the error isTeiid Designer Version and Tools Version working was news or idea? +Copying Specific Artifacts Copying Project Dependencies Unpacking Specific Artifacts Unpacking the Project you manually generate site? I dont see your mentioned broken links any more +yes +I attached a patched version which supports bind variables +The patch includes changes to sorry about it will commit new patch shortly +Duplicate of J which has been verified and closed. +bq +Im told that my issue CVS data not pulled in for files added since initial CVS integration is a result of this +ive opened to consier returning a wraped around the init error + +Ive tested it all with and and it now works fine +Your above points makes sense +Looks good to me +I suggest that if as seems to be the case this exception type represents a bug in the compiler and not the users code it would be kind if the message made that clear +the difference is that we are accumulating requirements if one already exists for that product +Dropping priority and deferring until later +fixed in svn revision +So there can be more than oneboundary string and we must ensure they are all unique within the same requestI know we had a request when somebody wanted to control the boundary string +Would you rather I rebuild the branch to be based on trunk as of the date ? I believe that should be straightforward to do if that helps +Attached is the patch file +I implemented the other mentioned improvements in followup commits and added a few extra unit tests +This patch should solve the problem but Im not pretty sure if thats the right place or it should be after parsing the command line arguments +removed from affected version this is a unreleased version unreleased version cannot be affected +Committed revision . +Attaching a single patch with the previous two combined +Currently you have to manually update the references to and when buildingThe usual rules having the and the boost dlls in the right place or in your PATH apply at runtimeSince also launches the native it needs to be in the path or with the other dlls too +null has a specific meaning which is not present in the database. +it isnt like threads are expensive though if they are not usedWe should add a comment at the compression option that it would be best to add one ormore task threads +Not an optionD cf +bulk defer to defer to +I am running regression tests nowTouches the following filesM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapitypesAdd the missing methodM javatestingorgapachederbyTestingfunctionTeststestslangAdd test case for this bug +Runtime exceptions are the way to go +may be add this warning after create column family +bulk close of all resolved issues. +I just committed this. +batch transition resolvedfixed bugs to closedfixed +if it has not been changedShutdown of RS seems extreme on socket timeout +Well how many versions do you think you want? Maybe we could split the request field up into say the first or bits instead of for the versions +However that does not mean that the usecase in the example of this issue is actually valid. +Thanks for the comments Todd +needs a unique local username to run the processes as +This problem should go away permanently once using an external connection pooling by commenting out for now +bulk move of all unresolved issues from to +If you really want to do this change is this layer I suggest that you seperate out this feature to a new filter that modify the position incrementAttaching alternative patch as suggested for and its testThe first token always maintains its original positionIncrement but subsequent tokens in the has their positionIncrement set to match the value of I still fail to understand why Karl and Steve would rather see this logic in the +where would you like to use these references exactly? in a file set for example? Or what +This is a bit of a hack but it should allow builds to succeed on JDK. +Even if resulting file does not exist client will get a standard error response saying +Looks good +Use this second version of the test case instead of the first one +If you have other changes in mind Id be happy to them +Example project with classes and configuration +Im using the same version +The class could be considerably simplified it just needs to delegate to the +Just attaching latest. +A similar note was also added to the user guide +ok cool +I wouldnt mind +HSQLDB will not be in AS Final release +and will take care of this +morgspringframeworkroo and then mvn clean install from wrapping then mvn clean install eclipseclean eclipseeclipse from root checkout location. +Now that is about to release with this feature Im going to mark this issue as resolved +Also it looks like we already include slfj in the client via is a slfj binding +Why does this pull in dependencies from someones private github account? That seems broken +The corresponding file for datastore identity is missing because single field identity test cases are only performed for application identity +Based on the stack traces it seems to me that therere two issues hereFirst introduces a bug so that it is passing null as the configuration key of principals keytabs into the +Thanks Arnoud! +in preparation for committing +This area requires some more attention +Since Bob reports it is working for him closing. +This does not seems to be a Core bug +Thomas indicated it would be corrected I believe in +A patch for trunk version + +The problem comes from Maven +Bit by this again today +I think its fixed. +Fixed in r +Currently all codecs default to use CFS for docvalues +Patch has been applied to SVN HEAD +I am attaching the google java codei renamed the war file to zipit has the java code in it as weil as the classesa lot of thier code is not needed +Hey guys reserving inventory for products with negative atp or atp is the desired behavior +I will create a doc patch I just was not sure if it was in a different place +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +For Section and please add an eye catching warning that This feature is a technology preview +Thans Patanachai for the patch +Will commit when tests pass +One thing I dont agree with is custom config schema for this +That code is never actually called so the test case will always fail +Filed to address the javadocs +Can you send patch? Thanks +But I guess this affects as well +Ran through multiple search cycles without issue +Weve still never seen a case of this in the wild and since the impact is pretty low and pretty hard to cause Im punting to post +I dont agree to including the javadoc sources in the release +Except maybe for C tools it does not make sense for most of plugins because batch is supposed to fail when tools failThis widget would not target but system administrators +Since I didnt look into it yet it probably still exists +Talked to pmacik on irc and he agreed. +Now the count should be for findbugs warnings +resolved in r +Ive committed this to staging but I dont know how to publish yet +Trivial patch +Marking incompatible per. +When I originally filed the bug and this still hold true for me I assumed that the instance of the object would be serialized in such a fashion that it could be recreated from the JSON +just tried a full build +The error is triggered hen a cutoff value provided by user is so large it is greater than the greatest singular value +We have seen similar issues with some tests that use to fail where before and after timestamps for an operationwould return the same timestampIf this turns out to be the problemIf upgrade were not an issue seems like fix would be to make the index unique and handle a duplicate key error on insert and retrywith timestamp +This appears to be a Derby appears to be a Derby bug +If it is then please provide a use case for reproducing this +Added some line breaks to stay under the limit +Logged In YES useridLet me clerify The get is not implemented in the +Return issue to you. +included type +I just committed this. +Red dotted border is not behoves toolTip +I will upgrade when possible and report back if there is an issue + on +ErikThanks a lot for working on this +Test to this issue +Patch submitted +Daviddid you try to rebuild your systemant cleanant buildJacopo +changes to store such arbitrary data in named after the Text keys +Should also renamed the activemq modules at this time too configsactivemq configs modules modules modules modulesI always get confused about the configsactivemq +The security manager subsystem has been implemented and integrated. +it seems confusing to go halvsies here. +I need to study it a bit more and see ifhow it meshes with a few ideas Ive been pondering with regards to tokens and multiple hostnames and ips +Patch attached +Verified +OK the problem is that it will try and open the DSL for the content assistance fail and then not let you get any further +Something seems wrong here +So it looks like its not problem in jetty itself +This issue is resolved. +thanks again +If I use samplecontainer with rpc then also it works fine +This would be good future work to make transparent archives possible + +Use the one above that one for this issue from +I found the following problems when testing +would be the first release that has such a timeout +Im strongly against doing that +This is already covered in so I wont convertTests moved Tests for mixed type NULLIF statements will be moved to new test +Its best to help first on the mailing list to keep the noise down a bit on JIRA +changed some test log fixed a bug in sleep too short heartbeat too long time outPassed the tests locally with same random number generator seed used in build +So the UDFS likely compilefine but the test cases will not +Thanks +this seems close the question is if we want to backport this to tooWhy dont we get it committed to trunk and let it chill for a while let it hit random testing for a while get used by adventurous users and then make the decision +With were really just trying to fix the obvious like a subclass Bean method not being picked which is even wrong in the old model and not go much further than +I am sorry to hear WAS is only supported I must have missed that in the documentation + is a valid patch +Added wire format and simple because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +resources should go into in the ear in all earwar projects +Please Try the hudson build +But before that I must do things list below +Theyre faster and simpler if run entirely in memorybq + +Now there is no way atleast presently that you can override this into using your own namespace definitions +Committed with revision +Thanks a lot +Committed. +This defect was resolved in +The constraint system has to ensure that it retains the information of what used to be a this reference so that we can reconstruct it when generating expressions from constraints +I propose this go to +That sounds perfect +Obviously you need a primary key if you want to use it in a relation via foreign key +Its all dfs +patches as described above +This looks good thanks Uwe +However because of a and b above they cannot be referenced in EPL queries unless that are lowercase +Feel free to contribute a patch. +JFCRemy release please +Im speaking about using it when loading markup filesOutputting xml prolog to users is handled by strip in +Ugh no +block on the table flush +Done +Such a allows for flexible access including different cases for column names +Since the physical placement of the heap is ultimately up to the OS I dont think the JVM has any real control over where that heap lies +Forms need to be added in a separate patch after Adrian finishes his refactor on the webtools UI in. +just committed this +Ive tried in but with no success +Or is this a knownexpected behavior +Uploading a patch that checks access of a job to user in steps +I will look into this +If I open the IE developer tools and put a breakpoint on line and examine the form local variable there is no elements variable so it causes the elements is null or is not an object error +Hmm +explicitly makes some of the protected methods public just to do what you are asking for +OK so here is a patch with meaningful names and protected constructors +Correction confirmed +This actually causes changes to server or vm arguments not to be saved between Eclipse restarts +i have this bugon java all because I doubt this is still an issue +The problem in the patch for was the visibility of the code +No issues have been found when installed and used with JBDS issues found +Hi Sascha I did it at rev +not breaking api that badly in to be fixed in how? i dont see a reverse lookup in ichoicerenderer yet. +Closing old bugs. +Thanks Brock! +I think it would be better to display three sections Default global dashboards Default project dashboards Shared dashboards +combined patch attached +looks good to me Knut +Chris upgrade the version of Jetty to +The trunk patch still works too but I am it because the previous one had paths to the files that were specific to my own installation the new patch does not have this issue +and with changes to MAX and MIN in addition to the other filesM srcdevguideA srcdevguideM srctoolsM srcrefA srcrefM srcrefM srcrefA srcrefM srcrefM srcrefM srcrefM srcrefM srcrefI hope this falls short of overkill +on the code assuming that does not generate any apply +I can see cases where the all at once is actually worse as it will spend more time on a single heartbeat and potentially not get to other things in the queue like apps added as fast +In addition there are unit test cases in and All unit tests pass +To clarify the spec actually says that the attribute needs to be explicitly annotated as Enumerated in order for the to not be applied +Documentation will be fixed for Ashish for the review +Ill write up something for the dev list when we reach that point +Not found with latest runs +The XML fragment you are trying to unmarshal and the full class mapping for User +Incorporating findBug warnings related fixes +Any objections +The documentation for nodetool getendpoints says that the key must be in hex format +But not consistent +Hi Daniel did that get done yet +Please ask this question on User Group using address userYoure probably missing some jars at the beginning Regards jars +Documented in Version . +Will work on this +Since Im not sure about possible side effects I will only commit it after the release +To use itthere are a few parameters to adjust to the users settingso set the element to the right valueo the should be the path to the users own fonts the parameter is the path to the afm files for the base fonts that Ive put in confbasefonts +Could still think about the exception as well I suppose +Not a It should be possible to customize the control with topControl and top facet that replaces the default control with facets content +I dont expect to work because theres nothing checked out underneath the parent I do expect Continuum to be able to build the entire project group and figure out the correct flat structure is a typical of Eclipse projects and is supported by Maven at the command line +That would be OK too I implemented it this way because it seemed like the most obvious solution for my particular use case but I suppose it makes sense to make the smart enough to do that on its own +I think its a must have for the shell though +checked only into athusoo committed this into branch as well +Actually I wanted the a test for the issue but both would be better +AFAICT this is not yet excluded from +I believe this was fixed some months back +Accordingly to javadoc canonical pathname is both absolute and unique so this might be a problem in impl I did not look closer +Committed to branch +All of this depends on how long it will take for the downstream projects to move to a more recent version of HadoopUntil then I would rather be safe than sorry +I will fix it +Java properties files must be +Therefore some guidance can be given how to minimize the impact +Tested on a production job agreement on the resultsCommittingThanks Avery for the review and everyone else for contributing to the discussion +This patch will only apply to was this discussed with the Oozie people? Im not aware of this I did not see any discussion in the alias +The examples in the trunk are already updatedHelp is welcome! +Using the patch on trunk load time goes from to seconds +Attached patch for trunk with all feedback addressed +Found itDont revert +Can you tell me how you guys actually use it +Add them as deprecated and a warning that they are expensive! because they make a copy +The factory bean is a decent compromise but things like your not holding the right data because it wasnt rolled back due to incorrect transaction propagation scares me +for the patch +We should think some other way to resolve the transitive deps. +Reassign to Yura to verify this is fixed +Thanks! +Dont you have a Karaf or SMX instance running during your tests? +Therefore it is guarantied that the interceptor is installed +Any objections +The interesting thing is why did it complete without any problems? Im trying to run that job once again with the latest version from trunk +Have changed the GWT controller code which invokes the service to use the method in a transactional the patch ve released version of the plugin with this fix in it +Id love to have this working is pretty much everywhere in my project. +It probably works fine at the command line because you did not write dependencies when exporting the makefilesNote that this will only happen once unless you change code in Xalan. + for patch +I often have people tell me look at job on nodeThat would change to look at job on node +This is now done yes? If so please mark fixed for beta +testing on had only instance but i think it does not matter +And hopefully them at a later time once they have been converted +Now that work is underway with AS all previous community releases are +source project still believes that it is checking out +Thank you for fast reply. +This would also allow the RO regions to be served by all three favored notesThis would be a nice thing to have +I guess I should paste it in one line at a time to make it less +That revision also addresses the TODO about incremental backup performance by keeping a reference to the checkpoint used for the earlier backup and doing the diff against it if its still available +RTC workflow prevents me from reclassifying this as a copyright bug fix but it doesnt look like anyones about to vote on it despite it being applied. + while running the mingle instance inside of the Debian image and attaching from the outside we get a totally different failure mode that seems to be more consistent +We can either close this issue and add another one for or just reuse this one +Definitely resolved +Looking closer at this there is an exception with an action in the node it does not have to be the standard handler class it can be calling a custom class so the current implementation of scriptTask for these and the insertion of scriptTask in the series for events like is just to add a comment and list the class name for migration investigation by a consultant +If another state does get added to the we dont know if this is a final state or not +Moving target to +It also covers commitWithin from xml or whatever but it seems we do also have to worry about the case where its just a request params on an add or delete +Hi RobIt seems the fix is ok +Follow up through the jBPM issue +Ive modified the test on my local machine and found that this bug has nothing to do with the system scope +thanks. +Not sure if we canshould rely on that or not and I dont know if other vendors do the same thing +This should preserve the multipart documentation on request parameters and the Sling Post Servlet with respect to this new parameter ordering support +to trunk and branches +Committed to trunk and +I think would we a good analogy it depends on spring core libraries and offers mock objects and other functionality for building integration tests +As Vishal said that ephemeral owner is xbe +Maven core cant know how the do version resolution +FOPs AltDesign already relies on themsee its class +These extra mappings have been added to SVN trunk +I thought Cassandra was built for system engineers who knew the risks of what they were doing and had the flexibility to make those choices? I can wait until for a better caching mechanism but this seems like something that could have stayed with a disclaimer and more documentation on how to use it appropriately. +Seam refactoring should not scan all folders +There is no jsch dependency here is it needed +The test case that fails to that you added a testcase but could you make it minimal include the processdefinition inline or attach the issue after finding the problem cause +What exactly are you building against there? started off building with Sun JVM +But I dont want to be free I like break stop and return +Checkboxes should definitely support enforcing the rendering of labels on the right side just like the other controls too Radio and. +This is a screenshot of a richeditor that is working just fine +Doesnt meant this is a bad feature just means you have to be responsible +The is a general problem rather than one specific to Cygwin +Until your proposal has been formally documented including an IDL which I think is critical to the stated goal of backward compatibility Im not going to be comfortable seeing it committed +Updated release notes fields +Taking out the results for client side oneway invocation tests gives the following ratiosbisocket latter two numbers should bebisocket the tests are for the version of Remoting that will be tagged +I have a test case for KEYSET and will be working on tests for others soon +The patch was submitted by me years ago +What is the state of the file during compression? How do you deal with data that was deemed cold that could become hot at a later point? How does Datanode block scanner and directory scanner internal datanode data structures that track block length Append interact with this featureGiven that based on the approach taken this could result in changes to some core parts of HDFS please write a design document +Adding an after scanning the beans would most probably create kind of a problem +This is an ongoing task used to track time +on patch +Yes I tried if the camel editor can be also used with as this is not a bug +we are not going to invest a lot of time in our own impl any more since that is going to be replaced with quartz +This could be a dupe of has been mentioned a lot of times in the mailing list as well so will leave this is. +you also posted this issue to the wrong project. + on change +This ticket was already marked as fixed since January +Renaming this JIRA to Custom Server assemblies using Apache Geronimo Administrative you Hernan for the comments +Then we can use the internal code to do the read locks +Mike I believe youve already updated the sql error codes in +It is used only in production envinroment with PWTEXT and everyone is happyBut in it was problemYou can resolve this as invalid. +Are the CXF tests using the +Fixed a remaining dependency in geotiff as of revision +Closed after release. +Perhaps coming at this from a different direction would be good +I concur that tallying is not really necessary certainly not when considering the performance boost this change grants usWill commit this when there are no objections +should do nicely I have updated it accordingly. +Thanks Ashish! +removing strict flag from +I guess your patch files were created using diff command +Bug still present +It seems to be a derby problem +I didnt noticeBut split packages are just different kind of evil +Regarding that Stack moved this issue to but he agreed we can move it back into since we need it at SU and its going to be disabled by default +Hope its ok now. +The attached script runs correctly under MRI and +Way +Keep the original jvm config +The patch introduces some refactoring around token selector and also probably code cleanup in but it would be good if that is done in a separate jira +If a client submits several big jobs she has to either launch a new JVM for each submission or somehow patch the leak from outside +I think it reads better +But in that case the additional step needed should be mentioned somewhere in the documentation +with Elis comments addressed +Moved force include management in abstract request classes +V +Committed revision +For SYS it complains that ALTER TABLE is not allowedIn ij SHOW SCHEMAS does not list all the schemas Ive created eitherTABLESCHEMAPPNULLIDSQLJSYSSYSCATSYSCSDIAGSYSCSUTILSYSFUNSYSIBMSYSPROCSYSSTAT rows selected +current solution has bugs + is working closing all issues that have been in resolved state for months or more without any feedback or update +Is it okay if I file a separate Jira for thisSure +I agree on deleting these interfaces +But then there should be some documentation for this whats the name of the properties file and where does it have to be stored which jars have to be put into the classpath for the several data bindings +Can you explain whats the copy on write ? Exactly when do we make a copy +Hopeful fix in r +That way we might encourage users to exercise sequences more in preparation for ultimately transitioning identity columns to use this code +The contains a Pdeploy to tell mvn it should deploy during the build +Dimitris not sure if you were going to help with this +It certainly makes this bug very easy to fix +Bulk close of resolved issues of +Patch applied at +missed this for assigning to applied many thanks Alexey! +Also I made this issue into a container issue for the query implementations and created separate issues for the individual tasks for the respective components Define query API Implement query builder Implement a query parserAt some later point we should also create an issue for the actual query engine which evaluates the AST +Will move back to core +Image showing the additional mapping adding two input fields to one output field +I stopped looking at it a while back but I owe it to your efforts to look into it againIll be back with details +Thanks Kristian and BryanAll regression tests passed with jars built with the patchCommitted revision The machine where I saw that the invocations of splitmessages took seconds was a laptop running LinuxOn various desktops running Solaris Ive seen that it takes between and seconds depending on how powerful the machines are +They never did though +cannot start ide integration without being fixed +since this does not yet appear to be fixed + +This is a duplicate of consider merging +I have committed this +Committed to native and framework branches and verify the fix in the latest build +Proposed SOA TP Release NoteThe SQL listener serialization is now implemented via to be more usable with more databases +there is a lot of changes in these classes +David are you planning to apply that extra style patch you attached? I have no problem with you just committing it. +for the new file committed to trunk. +The only way to fully detach an LRS field though would be to read it all into memory which would defeat the purpose +As with previous Solr the was invisible to MLT this had no effectBy the way There is a commented out part with explicitely the MLT field but I dont understand it +Thats actually how the old hilo generators worked they had a config parameter named maxlo which is really what JPA terms allocationSize but less +Yea I got the reasoning for Set we can use that CHM with PRESENT +Ive got a large app I was testing but no or isolated deployable will be a few days before I can get around to making a smaller reproducer +Ideally wed also be able to specifically kill whichever server is the quorum leaderAlso Im unclear on the meaning of candidate in this context +This patch is not ready for commit +All subtasks are closed. +Verified with japitools +Closing issue. +Thanks Yin +Applied +Actually if a header is not Serializable should the behavior be similar to the s? Specifically do we let it be configurable the behavior of do nothing log a warning or throw an Exception if the header is not Serializable? But maybe we default to ? Just brainstorming here +Good idea +This value will be greater for the newest file resulting in this scanner coming on top +Port from Change from loading entities one by one to loading all using the IN clause +Fixed in the project and verified when Designer importer was used to create the new model and vdb. +It is still an improvement and doubles the size of the files that can be written from GB to GB on bit machines +Goal was to show that other applications are already using nano instead on milli so we might be able to do that tooRegarding Mononotic documentation saysCLOCKMONOTONIC Clock that cannot be set and represents monotonic time since some unspecified starting point +From top of my head JPA mapping defines how to handle or more queries being defined with same name in mapping +I think Blakes suggestion has some merit +Core and contrib tests pass when run manuallyThe release audit has warnings one is about the new golden file for test case which does not have license header other two are about javadoc html filePatch is ready for review +As an outside observer to this list i think this issue has gotten a little testy and I am hoping not indicative of SOLR as a project +Most of this got implemented in issue is fixed and released as part of release +Paul any news +You can now run mvn install multiple times for any of the quickstarts without having to first clean +Applied the patch in Rev +The better solution of course is to make sure that settings that should be controlled by the server are stored in the server while settings that are are stored on the project +That will be for or I forgot +So basically you need to strip out collections from entities before posting them back +You have to run mvn install for mvn site to work now +Outstanding formatting issues have been moved to for progression via POI in a later version +I have created a personal account for use +Ill take a ve let the error collector collect errors after logical expression is deserailized +However theres already an external taglib project which I believe supports JSF and JSF and which people are already using successfuly so Id prefer not to add support for additional view technologies to the main project at the moment. +Ill try and do a thorough review tomorrow before the big news day on Wednesday +You can add the exclusion to the POM of the maven plugin +Request is tracked on issue +This means all the logic for hadoop jars extjs +Do you mind helping out with the wiki documentation? We need to add about this in the idempotent consumer EIP wiki page. +Moving to so that can be released +Now throwing a no message which is consistent with the situation when no is available. +The are included in the job parameters to guarantee a new job instance for each run +see the comment with the test and fix with svn revision +Right now this is only possible if you set up an OGC client that will consume several WMS services yours and for example the one from the nasa +bulk defer of items from to. +But it can sure use improving +Works as advertised +hence closing the ticket +Verified by Jeremiah and Xibin +Hi Spring The bug happens only when i refer my jars as user library the test gose well when i refer the same jars as individual jar +Without relying on Spring Data the application works fine regardless if you have the id and a different field acting as the id +Committed to and trunk +Ill have to see if I can get this working too +I will retest now to confirm the exact steps +Ill apply the attached patch in the master and of each repo +Is that the issue youre seeing? If so can you provide more details about the circumstances where this occurs +Chandan +Its running here +Not at this point +See the attached dont need to add any library +Fixed under other Jiras as part of general sandbox build cleanup +But again all this is for later +But commented out all tests +Not sure what are the plans for Scan version and when it is ok to change it by initial design +Make sure it is properly documented +There are a few changes in this patch that can be safely removed +However the New index page is sorted +buffer sizeavg seek time in mskkkI also tried scanning through Accumulo w k and k buffers and did not really see a difference. +CharlesCould you check my latest changes onI introduce a new operation service in URI for looking up the service class from camel context registry +Thanks Hitesh! +oops attached wrong files before. +iconnlicons +It could really do with a unit test +I can reproduce it on my computer with JDK Mac OS X and Tomcat +change the signature of method jobComplete +Thanks Uma for the reviewI have committed this +AFAIK the same needs done for the +Thanks for working on this Murray +we are using file and file contains all module context files and when ever invkoing SI module context file we are getting these provide is urgent dont know what file I am saying is that you have a bundle in your environment that depends on SI but attempts to use SI +bulk close of all resolved issues prior to release. +Mike any update for your patch? +you very much Sylvain I will add that information and post a revised patch +It is gibberish to me and of no obvious legal effectIf we ignore this nonsense preamble this software seems to be just a GPL with no viral effect whatsoever on Apache code +Some minor comments use LOG instead of to log the exception +If we instead had a shared container that tracked by thread id and then change and to fetch the right map this should work +As I dont use eclipse to compile the grails projects I hadnt noticed this +Robert which version of p are you using? The code works fine for me and others and Im on +I cant see any case where the wouldnt be exposed outwardly either with Out or with a getter +Those should simply be removed fromAbout the includeSourcestrue in unified I think its good to keep it +Ruby changes this behavior and gets the last or first value directly +Scott I have included a testcase on lines with what you have suggested and its working now +to next version +The earlier patch failed with latest trunk due to some changes in so I redid some part of it +I just committed this +If not it might be easier to just fix sstablejson to not output that commaOr maybe they should both be fixed +Thanks I have no commit access +on kernel tests still fails under sh +Pushing to to remaining work with relation and procedure caching to +bumped to the fix for were now warning users that phpmyadmin may not be embedded into a scalable application +This means it wont be included in +Regards guysNot sure why this broke again but I have attached a patch +Upon inspecting further it seems the patch for added the check for the base server at region server start code +adding this was also broken in that release +Its the only one in that package but a quick search suggests there are some in other ll create another JIRA issue and list them all. +and will be fixed in the next major release +Is the intention to commit this both to and trunk? Or just trunk +Im nervous about adding extra NN operations per taskcurrent solution doesnt add any NN operations +Whitespace inside elements are usually disposed of by the parserAnother way to solve this would be to insert an explicit space between the comment prefix and the actual commit message +Hmm seems I either reverted the previous patch or the patch wasnt thereAnyhow will make the needed changes +This issue is partially solved with +I have some doubts about mutation testing but I dont have any practical experience with it so Ill just sit and watch +Great plan close this consideration and maybe the problem will go away +But the tradeoff is a longer warmup time when a new searcher is openedDo you have indexing requirements or can you live with some time +Good catch on the EOL test +bq +It seems like the regression occurred while fixing +The +This example could be enhanced with more code in the futureLet me know if there is a more suitable place for this and I can make the changes if required. +People may well not know they are using itI am guessing that using runtime compression means calling the or procedure let me know if Im off track here +I think this is more of a JAXB issue than a Spring issue and I wouldnt be surprised if using the native JAXB will issue the same there is something we can do Id be happy to fix it but as of now Im closing as Invalid. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Applied to trunk +Bug has been marked as a duplicate of this bug +in addition if this could also be used to trigger start of wizardsimport of quickstarts would be interesting too +Classifying this as a new feature. +For it will include the ability to set control properties from Spring bean definitions +Ill follow up on for now Im resolving this issue as fixed. +I should add that because of the debug log I noticed this in some test failures and it is quite easy to reproduce +There is two phase shutdown through JMS if failed then kill processI think there should be added something like this if process was killed in then show if we allow test suite to continue execution then how will we notice that a server failed to shutdownI think we should do one of +Patches Modifies the removeToken operation to follow a pattern of NORMALREMOVINGLEFT rather than the current pattern of a coordinator node setting its own status to a special cased version of NORMAL +Updated patch +Patch for review and discussion of next steps +There is no property for documentation on the user task node in the designer +Unfortunately I havent been able to reproduce yet though Im not doing the firewall testing like I was doing which may have precipitated thisI would think just checking for negative numbers might help on the server side +When creating the for parsing apply the digit year start trunk patch based on code review comments +RI does not support this format it is beyond java spec +this isnt really complete support for relative urlsthe reason relative urls are hard with this is that the url ends up looking like thiscontextresources +any progress on this? we REALLY REALLY REALLY need that stateless rulebase API as this will solve a whole lot of support problems +Ive checked in your version of the png file +is the problem that adding Property causes the setter to not get called or that Parameter doesnt call the setter either way +Also tested and reproduced on trunk Ivy +this is a dupe of this since I am sure it is fixed and I even have a test for it +Sorry for any confusion +Since jirb at console and in Swing works ok for me Im going to call this issue resolved and look for additional smaller issues for platform bugs that remain +Given poorly toned comments on and on public forums like twitter I am asking why this is not a dupe ofSuresh With this issue I had hoped to engage on discussions about tiered storage concerns specifically and volunteered some of our thoughts for consideration +Thanks +The tag index needs the new segment docmap and a callback for each merge in +But thanks for submitting it anyways the fact that more than one person found the same problem added to the urgency in investigating it. +This patch includes the CMIS Connector Java attachment includes all the Apache Chemistry dependencies needed to build the CMIS Repository Repository Connection Connector used in an ingestion job with Solr Connector Job Configuration with a CMIS is the Java code for the CMIS new Java code for the CMIS Connector +this just fixes the +Thanks +bulk close of resolved issues. +the last one from rb here +Here is a fixed patch +Got it thx James +has no tests +Some of the patch threw me off +It is in +Updated patch which gaurds against NPE if importedFiles is not patch also displays the same symptoms as patch +The fact that Xalan knows to output namespaces for tags but not attributes and so it works for most schemas but not all makes this look very much like a bug to me +Thanks for looking at the patch Rick! I enclose a new version which fixes the comment you pointed to plussharpened another comment +found the regression should be out the source and run mvn ve checked revision all works like a charm nowThank you! +Hey Oliver! Can you try a latest nightly build? I fixed some issues around the annotation processing lately and I dont see the warning when I try to reproduce it +that has been there for so longFor the general code +Altough it was discovered in a different way this issue in effect duplicates +Patch is here for those that need it though + +DonaldCan you either close this or upgrade the howl version in trunk +does this fix the describe problem too or just the update one +I guess you mean referencing a custom implementation using a bean problem also exists for the end state output that the subflow state attribute mapper does support thing to keep in mind here is that currently the cannot lookup an only a so this improvement might be bigger than just adding the bean flow builders can of course do this easily +it just became a bit of a rework of the original patch +Michael any luck on the unit tests +Other containers need to be verified more whether it is not the case for them +i am missing something +Method had to be renamed due to type erasure cannot have two methods that are only different by a parameter of of the same class template used as a parameter +Firstly it appears the behaviour of delete no longer corresponds to its contract +A patch to to assign a system ID to objects when they are created +Verified also with +why isnt the client reconnecting to another server? this seems similar to the asynchammer test failure in the sense that it looks like the server io pipe gets blocked up for some reason +My prototype works on windows and linux +A longer chars wont get user any information +The desired behavior not closing notify when user is reading it is not working on FF it works for I can see that behavior isnt consistent nowwhen using Showcase sample Attributes of Notify notification sometimes disappears sometimes notBut since this logic is encapsulated in jquery pnotify plugin I wont recommend to dig into it more now. +Hi attaching patch +tested by hand and works +Resolved with AOP and MC +Patch applied +It was the intension of the original author to allow one wsdl file to be executed +Thanks for cooking it up +Which documentation did you read on Dynamic mode? Its in the docs +The patch provided by Srinath did resolve the issue for me +Any problems here +Could you please let us know how you did the test? Maybe there are some differences between the way we did the test +Thanks Vinod! Addressed your comments +I would recommend creating an associated unit test for the new attribute +However the rest of the jbedsp integration related tests need to be evaluated for moving to Teiid +A test case simplified just for reproducing the regression +Should verify the fix is in for and then take off the fix version list and leave this to be checked in HEAD after the EJB refactoring +DML takes information saved by DDL and uses it +I can also add to unless you beat me to itAlso does your patchissue wish to include the necessary change required to update? IHMO I think these new changes should be made only to as this seems to be a more detailed schema representation which reflects newer Nutch Solr versions people seem to require +just my two cent but Im sure right after we switch to JDK with Karaf a Java will be available or shortly after and we might just hit the same issue again with JDK specific stuff +It should probably through a runtime exception so it doesnt loop +Thanks to Dennis Byrne for his efforts this is a major improvement for Apache. +Strange on windows as far I can remember doesnt break anything +good enough for me with the flexible signature handling +if the connection is managed the trackers are setup only at construction and during abort in the specific case of +Fixed at revision +enolan Applying Alexey patches all tests pass +Thanks! +Thanks RegisI agree and the patch was applied to the SQL module at repo revision rPlease check it was applied as you expected +What wed have to do to get this to work otherwise would be unnecessarily complicatedIve added the css and checked it in for our next releaseCheersve added the css to the template for now as well so the changes should be in effect immediatelyThanksCheyenne +What was the connection string? It is also possible for the client to throw an exception in the prior to failing over if the failure occurs mid query +Steps to reproduce +Fixed along with a test case. +NathanNo the issue still exists +Patch is now available in CVS. +Updated my cluster with patch and tested +I dont think that asking people to find a jar in the geronimo repository in order to use is appropriateI think this issue illustrates that we are using manifest classpath too much and that it is not appropriate for large scale software +Can you send me the files or a patch so I can test on my local machineThank There is a bug in Drools to do with OR the unbalanced trees +If you do neither of these the popupPanel will not be in any form and the inputs in the popup panel will not be submitted. +Is this on Windows? The Path code behaves differently on WindowsCan you provide a standalone test? Adding something to might be appropriate +Please take a look and let me know if you are to commit +I have reverted rev on the branch so the API is compatible with Jetty again + seems to fail with stock serialization the last comment is incorrect +Shall we just close it +Yepp ill do so patch same functionality using lucene revision +There isnt that much to do +Can we make the dialog change depending on the permissions of the user? Editable by the sysadmin uneditable for the users +Youre asking good questions and Im just trying to provide some answersor at least some indication as to why things are working the way they are +It looks like a minor change is needed to make all the mapreduce tests pass +I forget the exact details but it had to do with Javassist using a weak reference to keep track of classloaders +oh and postflight checking that codecs went in +bulk close for +But as you observe this single clause DNF case can be further simplified to a boolean and a count of required terms +Tests passed cleanly for me on +these ismptyStub project in my wokspace butI replace english e with russian e as a result i got error +Ah Ok +a new patch with comments for the code change +I can give you more refs later when I havebetter bandwidth +Not sure when the EGT stuff is rebuilt +I have been following this JIRA and also discussed with Anoop +Thank you! +Leo patch applied at revision r thanks a lot for this enhancementVladimir please verify that the problem is fully fixed as you expected thanks +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Im looking at our code for this area and I dont see that we made any changes to the way that map expressions have their source locations calculated +To be clearer this is what you get if there is a security error trying to access the catalog +I have committed a revised patch for this issue +Thanks Naela +I would like to work on this issue +Will limit this one to just HDFS and trunk +cfe etc? Seems like it wouldnt have a real cost and be worth the effort sincewe are slurping these things in anyway +What if we extends from the the original and overwrite some methods + +So the pick up I saw with nodes I believe could be linked to these errors that made request return faster +The only problem I can think of is if someone already have a custom comparator named or but thats a problem we face each time we add a new comparator +I suppose could be an interface and we the base code which is the main factor I chose to derive from a base classGiven that the new exceptions format the exception message in certain way making the following change Move the message formatting to a static method Have subclass +Basic hooks already available most notably the property in the jobs table +Local build was successfull with this new version the JBPAPP branch updated to use this new versionCommitted revision +Great thanks atm +Heres a p patch that fixes this issue + +fixed on will be released along with +Fix appliedI tested running a single sample on a Windows machine where the maven repo has a space in itI also tested running the launcher on OSX from an install directory with a space in itPlease confirm this fix works for which version is it resolved? I have and it seems to be still not solved. +Fixed +If you cant reproduce any more probably time to close +Attached new patch incorporating review comments +That would be bad I was under the impression that clones were actually closed but the close method just checked if it was a clone and if so didnt actually do anything +If you run the application and go to httplocalhost submits the page everything works fineThen change the struts version from to and the same test case The map returned after the submit is null +Hes on another machine the request does not work as expected he gets no feedback whatsoever +So far only the java build has been updated +This is fixed in so Ive cloned into for branch +But lang return . +I agree with Matt +i assume the problem actually came from the JAVAHOME part +added the JIRA for testcase Python client testcases for JIRA has been +jsp file in deploy still causes redeployment of the +Thanks Daniel! +Add alias support for other metadata integration callback items +Oh well both patches are roughly the same +I was not blaming this patch in any way just thinking perhaps now would be right time to consider timestamp refactoring to make it easier to add these new variants +Im in a workshop and will be back onThank you +Please cancel this issue +commit preparations move exception message to patch including new CTF test case. +Found one more problem +Ive fixed the malformed documentation issue which was coming up last time Brian tried to apply the patch +Something is upbq +It becomes clear for will check java code where we get authorization code and change it that our authorization request is approved +no will still need this patch to reduce the number of lazy seek but do not need extra check whether scanner already do a reseek or sreseekreal seek +lets see QA compile per chunhuis latest v patch should address the above style problem +There are also chances that the type of resources your are filtering are linked to the guess you want the issue resolved +If you dont want sub projects under single source hierarhy youll have to create separate eclipse projects +Doing this without impacting realtime workload on the cluster is the tricky part in my opinionSorry for the mistake that rather than write lock a read lock instead should be acquired when the snapshot is requested +If this operand is a constant then we can simplify the REFLECTIVECALL to a CALL +Will apply once is rolled out +HiYou are right +Absolutely Eric +This needs to be put into its own subproject +I believe this project will demonstrate the problem. +Seems like the change is not so trivial +setting it to assuming it is an regression +Any chance you could do so +Upgrading to the JRE seemed to have fixed it +I do not have internet access during this time +Genuine bug +Looks like the problem has died of the old age. +Fixed! Please test if you can Juergen +I will send it thereThanks +Brian patch for this issue is now applied to Apache CVS +This is fixed with SVN with a BVT added in SVN +Can you explain your design in more detail first and then update the patch with the latest revision +Solved by as described in the link +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Brian do you know which version of EAP would include this Feature and when we can expect that release +a small program to parse output +and are referring to fields in the bagThis behavior is inconsistent with other types because we insert type casting for other typesThe error message is confusing too because the bags schema is unknown rather than of size +Truezips JAR is only KB big +Attached is a copy of the Java files +It may be possible to extend it for range partitioning or we may have to write a new manager for it +It is not a bug in Camel +Thanks MarkusYour patch is in trunk at revision Actually it did not apply in Eclipse but was easy changes by CR +Fix to appear in. +Yes it does this issue is outdated and can be closed +Thanks! +gettersetter after construction +Now that we got maven integration we should actually be able to support this by having meclipse configurators set it up or is that not doable cleanly? Does users need to add something to their maven for us to spot it or how +should be corrected toThe following is an example section from a file +It is not being used +This issue is fixed by setting TEXTSIZE config parameter not by changing page sizeAlso I would include the note that the issue can occur even on SQL Server if explicetly set but by default it should not +Looks like this was cleared by +The issue with the properties its its not because the property is there that its used +Committed +Its the snapshot +This is a Documentation issue that affects all prior versions. +Hi I cannot see anything wrong with your code +The parts of an application that depend on ordinary scaffolding work the parts of an application that depend on all fail +Closed Issue +Its a serious regression I thinkPlugin cannot be used because of for example in a am seeing this with maven and jdk +Wouldnt it be possible to do that check at the end thenCheck the contextPath check the classes and then fall back on the s unfortunate that the doesnt have some sort of supports method +Hi Amarit is possible that a progress bar might be useful +Jordan Any update? Do we need to look into this ourselves +At best it will be misleading to someone reading the code + +So it seems like the patch is adding quite a bit of complexity to hide some sort of issues with hadoop progressI winder if the easiest change for now would be to check the progress and if it adds to not to report it until we check that the job successfully finished +It will get fixed in the next releaseThe workaround is to enter a newline character at the end of fileI will also commit this change in the so that future usersdo not run into this same propblemRegardsBhakti +This only occurs in on and I know that this is not a parser problem but an internals issue for Ruby support +Other than having also published a layer having a Style attached to the Layer I dont think there were any other stepsRepeating as far as the layer I get no crash nowPerhaps related was the fact that I encountered memory issuesUnlike for I had to increase theJAVAOPTStrue XXm all issues that have been in resolved state for more than one month without further comments to closed status +Is there a problem treating Selenium like shark and the other libraries that are LGPL +I am running above classes from my Eclipse IDE only after starting Message Broker +I just committed this +remote DB multi and single db All tests are Ok +This also leads to better code reuse as any named boolean operator could have an obvious and simple negation +Achim have you run Maven Update Project Configuration action on your projectIf it fail for you can you please provide a test project that would allow us to reproduce this? Thanks. +Heres an updated version of the patch that also updates the example NOTICE file to be the current NOTICE from httpd +Definitely +Marking as resolved since I dont expect any more work on this issue +Patch so that the consumer is added to the consumers hashMap before the call to consume is made to the broker +For instance is it the sun jdkjava version will give you the output i am interested in +i just committed this +Already fixed. +for eclipse its a tad more cumbersome but doable +Ok I can see value in tracking the number of validations processed and am opento a patch however since it works as advertised Im marking this as anenhancement. +In the meantime they have provided a PDF doc on how to get working with httpsupportuidswgIf you have any luck let us know with the steps you had to go through to get it to several attempts this is waht it worked +Yes everything works right and we can run server and other actions +Are there any other packages that you would like to make optional +was actually fixed in Im not sure if this is even bug +That seems okay and makes the bigtop smokes much more useable +So it was my bad to be inaccurately +Jitendra thanks for the review +How did you find the documentation to add these settings +Will it be fair to mark as a for +This bug has been marked as a duplicate of +this was added for was closed a while ago by Alex for RC. +The current fix breaks moving issues between productsI need to do some testing of an alternative fix on my local BZ instance +Added a add method to and used it in the pick and ordering lists to add units to numerical height closing +Honestly I just think this is bad practice on many levels +verified in JBDS +This issue was resolved in scope of +Can we use the SCANVERSION? Up it to and have version able to deserialize both versionsExactly +Please refer to +patch with applied on and trunk +I can confirm this issue +The fix is do something similar to tdataList does +I can revert the taskUrl changes +Looking at your PATH a bit morePATHhometejasDesktopapachebinhometejasDesktopapachebinusrlibjvmbinusrliblightdmlightdmusrlocalsbinusrlocalbinusrsbinusrbinsbinbinusrgamesusrlocalgamesusrlibjvmbinusrlibjvmbinusrlibjvmdbbinusrlibjvmjrebinhometejasDesktopapachebinhometejasDesktopapachebinhometejasDesktopapachebinhometejasDesktopapachebinhometejasDesktopapachebinhometejasDesktopAsterixcurrasterixbinhometejasDesktopScalasbtbinhometejas +Im using EAP +Otherwise the patch is good +Report correct num of unapproved best way to fix this issue is by resolving +Uploaded repro seems this is the same bug that is mentioned in test case ten inThere it is being referred to as number probably predating JIRA. +It would prob make sense to fall back to this if getFields returned an empty set anyway which almost makes it not even worth it to do getFields as things dont get any cleanerWe def want the multitermquery clone thats for sure Uwe recently mentioned that as well and Id been meaning to get around to it myself +But then I would also get rid of +In particular if your images are dynamically createdsuppressing the rewriting is absolutely NOT what you want to doSearch the andor archives for messages about usingmodrewrite to deal with session ids appropriately andor look at thedocumentation for the modxxx connector you are using +About point it is fairly easy to add a unit test for that +Anyway do you have a patch +Otherwise your just altering a current test class and it is not obvious how hive is effected or what was broken +Ill check it in +Link has been. +We do have test clusters that run two datanodes on the same machine +Great. +Ok I will take a look at the jardir directive + +this only work on unixes not windows windows i tried the same scheme the logic seems to work however on bit system eval the launcher results in popup dialog and there for the start script hang +Sorry this problem concerns the version of the plugin +My guess is that are not sent to the client as a security concern +Added new docs to SVN and updated site on minotaur will be live in about minutes. +Jeff there are issues about library dependencies for examples +at least document this see the +This is the attachment that I have referred to in my previous comment +xml in the repository and remove the build target thats causing this problem +I added in an overwrite option +Sounds like ulimit but seems like an odd number +What if we add extraClasses and want to use a custom wsdl? Is that too much to ask from axis +I really like the idea of revisiting the way we ledgers +In theory somebody could have implemented its own protocol doing something similar as cocoonWhat about cocoon vs +Ive attached a new patch which doesnt have this problem +Test case for multithreaded access and find the attached patch for trunk +SamThe connection management code is so complex it can easily cause a severe hair loss +Well have to define the situations where this context switching make sense and where not and try to come up with something that feels intuitive +So im not even sure it could be done before Fuse ESB at Apache and the fix has been integrated in the release +Im guessing it is the test test that is failing +So it looks like doesnt help in this caseIm was not able to quickly check trunk build because my project is using maven repository +Assiging to Yash Talwar who agreed to look at this one at the JIRA triage on December +was doing nothing if remote wasnt declared. +Lukas in Metamer I have to remove name attribute to make it work. +After taking and backing up snapshot our code will call remove snapshot from vcenter I find that after removing snapshot is issued another event called disk consolidate happened then root +So this now might be possible +Please revise. +I will have no access to email and voicemail +The only differences between the zip and repo are the known todo jars +you must be a mind readerAlso the client funding the work will need this on geoserver +Is this amultithreaded app? That could be part of the problem +. +Here is a screenshot showing the problems on Windows +jetty requires java this is OK for trunk but needs to work with close resolved issues +Ive been playing w goraci +This should solve the problem +Moving to CVS now +I once tested this with couple of RDBMSs and most of them didnt support DISTINCT with Clob or Blob +Note I have removed the marketing parts of your comment paraphrased it +Whats wrong with it +Emmanuel can you explain the reasons +Also removed unused imports +I will take a look at the patch in this week and we will see where this leads +There are too many differencesrequirements between the html and xml so I agree that a separate goal would make d like to support jdiff for testsources too so Ill have to introduce at least one abstract ll keep this in mind while refactoring +Looking at initialization log I cannot find any kind of plugin included +verified in JBDS +It is not used for the second one +After checking amd spec I realized that double argument for param has to be passed in xmm register + +These should now all be done except references in the test cases artifacts in the scratchpad area +What you can do at the moment is get the listener registry find the listener on then use that to get the server name then lookup the Server service and from there get a list of all hosts +It changes mvn compile to be mvn package in our documentation around mvn +thx. +Can you clarify Who is the author of the file? Simon Kelley the Debian packager or the original software author +Closing as a workaround has been identified and alternate ticket has been created +Craig can you confirm that null is a valid value +Does this node have ECC memory? If so memory errors are unlikely +Scratch that this fails horribly after the change to unparsing negative numbers as hex +I am getting tons of compile errors +Would it be better to check the contents of the version object instead of checking the offset? I think we have a way to run the upgrade tests against a specific set of old versions in which case may be at a different offset +All tests pass. +Ive submitted a patch for this issue in +Also it leak a thread each time in +This patch requires a big rework to get it to work +I will fix this asap possibly tomorrow +Do you mind making the file name change in SVN as I dont have commit rights +Good to add a test that the force option works specifically wo specifying it we dont blow away your existing shared edits dirs +If you read the above and replace moderator with owner it should make more senseIt looks like there were a few hiccups whilst the lists were being changed +Until Geronimo has some sort of agent capable of managing multiple servers that we can use there really isnt a mechanism available in geronimo to allow this capability +Ill start to work on a new version against trunk and will do it with more focusused mindset +I do agree the graceful downgrade strategy has value outside of just +Ive had it happen to coworkers too so Im pretty confident it is reproducible +Hi thanks for your report but I cant find any point in the code that looks as you describe +Seam contains examples +has been marked as a duplicate of this bug +yes now it is included +thanks I forgot to do that part myself +stack and jdcryans Thanks for the explanation +If the call returns saying it is well then all you know is that the service thinks it is well but still may not be capable of useful workThough at both times the only one who knows about the trouble is the health checker and not the rest of the world +This is a security issue as a bad server could exploit the failure in a client at least for. +Can be closed +Adding to identify missing CR issues +Thanks FrancisYour patch is in trunk at revision +I managed to replace with a patch file when I was testing +Thanks for your feedback +So a gatewayprovider for JMS would have a gateway listener then listening to the data coming from the gateway and bringing in those messages as they appear on the JMS queues +It returns a soap faultThanksSenaka +Of course you need to replace in the documentation example with +Leaving open as Im missing permissions for the site update +So this is very likely to be a Commons Logging issue Check out their docs on how to make it work properly with JDK course if you discover any issues and special solutions for Commons Logging we can mention them in the Spring docs +Do you mean when there are keys that are including timestamp +Konstantin had confirmed with me that my prior merged patch was now working and asked that I create a patch for trunkThis patch is a patch for trunk exact same changes just mapped to trunks code +looks like I have similar problem +looks good to me +In MEP we are supposed to send a reply +Patch to support System well get this fixed +Do you think we should do it in trunk as well +Attaching patch with following modification from the previous patch Added a new class called which implements the Command Line Interface methods for related operations +To provide a safe valve for any input fomats that might dislike the combination of their splits a boolean property of is to be provided to allow for disabling this feature +What does round robin scheduling with a throttling value per stream partition even mean? The expectation of the app would be that behaves as +Here is an exampleis indented +Patch with comments incorporated +The condition under which we query whether the app is active is if the app just got in between the time the first modification time check and writing the cleaner lock +Ah the problem was originally with Camel +Oops added missing volatile tried running the tests on linux and got failures +orig file that were not supposed to be part of the patch +Good thanks +Ill upload a small patch in a sec +We should use it +With serialization you have objects serialize and them +Very resizable +Committed revision +patch looks good +There is no method for comparing two versionsShould we request such method to be added to so that we know whether certain release is newer than the known releases +Marking closed. +Affects also Sonar version +This issue can be reproduced on desktop with reduced users rightsI created patch that resolves this compatibility issue please check itinstead of as Harmony threw before there will be warning messages printed to the system output +You can then restart AS and mvn again to see how the exclusion breaks the deployments classpath +Because of that we only have the branch name in the default verson stringWhen a release is prepared we explicitely set the version number using Dversionfoobar +the reason for the fixed clientid is durable topic consumers +Locale Policy configuration I think that Kernel should be replaced with eXo Kernel +Hi GavinI posted a message in a Hibernate Forum but dont get answerthen making tests I created a table in a Oracle with same columns but instead of CHAR I put VARCHAR and it works without change anything in my code +trunk Fixed osgi all resolved tickets from or older +More precisely when there was multiple column slices the logic was returning no intersection as soon as one of the slice was not intersecting but it should do the exact reverse it should consider it an intersection as one as any of the slice intersect but wait to have test all slices before saying it doesnt intersectAttaching simple patch to fix +this file does not seem to be present in the src control +This bug is not fixed in +Let me enable logging from here on out or maybe better temporarily bring down a version of Server into hbase adding logging of outstanding connections to see what more I can learn about the Server endgame. +Patch for review +Improved handling of environment variables +is indeed an SLSB +Reported this to the subproject +Affects ER as well but I still did not figure out if this is a possible way +This seems like a very good connector to have as it could be used in tests to effortlessly provide content +Awesome! Good work +Thanks Stefan +util and removes its default listIts just a generic you must supply the listI moved its list to where it belongsAlso this filter no longer makes passes through the string. +This approach will also move the ensemble to a new epoch as youre proposing but it will correctly move to this new epoch by running the recovery phases +None of the are created yet by that point and hence the need for that change +would you please try my patch? thanks a lot +Closed after release. +Thanks problem fixed test is ok now +Thats why the auto build in Hudson cannot apply it +Working on the split log tests +Shouldnt provide equalshashCode impl so that created for the same module would be considered equal +HelloWe have the same issue with as Niko Kilpi reported +Id like somebody else opinion before committing it +I guess this was allready done in +Workarounds can imagine that by feature installation Karaf will install configurations at first and wait until configurations are available via service + + +Ok will fix +It is peculiar all the effort that goes into backwards compatibility of the index but for Serialization it is ignored +If the user wants to configure the server heshe will have to go through the edit Committed revision Thanks Viet. +We should look at implementing the Audio tag JS interface as it is more along the lines of the regular monkey patching we do IMHO +Spec needs to be changed if it says otherwise +Attaching which added an additional assertion in of the test cases +Could you please checkEli In future I would very much appreciate if you could check with an active MR committer before committing particularly since this changed lots of code in guts of MR +done closed. +We cant modify a tagged versionNow in the current trunk all plugin versions are locked and all tests are fine. +Ill be the first to admit that I wasnt sufficiently prudent but I still think this is a bit of functionality that many developers could benefit from andthat it should be included with ll do the trick for me +I would greatly appriciate thatThanks this still an issue or was it resolved? +Suspect something going wrong in the +Cant attach file +These serialization plugins should be stand alone or at least easily in a way that doesnt involve recompiling all of Hadoop such as a contrib module +I cant imagine a is an acceptable exception to throw +Code reviewed +renamed in +Forgot to mention that I tested the above fix in my test environment and verified that it parses my resource files correctly +committed and tested on linux not tested yet on Windows +Actually its some javascript feature which are not responding for instance also the lookup description featureI used Windows XP with Firefox Chrome IE and OperaAnybody an idea +Thanks again I added a test for this to your patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Two GC took seconds and the timeout is secondsThe way things work now if someone searched during the GC theyd get all the results back the search would just take longer +I dont expect that this would be the feature request but I think it would be generally usefulThanks +This issue fixes a code change which treated an unspecified scope as equivalent to a local scope. +The change individually looks goodHowever it exposes a bigger issue that the tests haveIdeally should make sure that the tests do not have any and initialize theparameters everytime +i see it +This one was a tricky one. +Lets wait for others comments +Local +Verfied passes +Geronimo doesnt deal well with url schemes at the start of jndi names +The new class name is in the package +Black was used instead of calculating the color value based on some pseudo postscript commands +The way I work around this is to add a m builder which calls warinplaceThen every time my project builds maven gets invoked and copies the jars out to the correct location +Could you patch for trunk as well +become ampnbsp in textField during for me Thanks some tests would be nice +Ive created. +if you really need this fixed for please submit a patch +Hi JacquesFYI geronimostart is going to install geronino in serverbranchestarget +To executeant a couple small niggles for clarity the new target should also just be so it aligns with the new task and the line at L should not be removedIf you update the patch Ill commit it in the morning +why part of the patch is in project manager and part of it in work effort? I would like to suggest to first implement it on a work effort in the workeffort component and then to make this available in the project managerI also miss the function to enter the resulting total time into a timesheet referring to the taskworkeffortAnother point is the updating pe second +Angel I will apply your patch and retry locally Isabel is it possible to test Angels patch outside of a local environment on a hadoop cluster before it gets to Jenkins +Oh sorry I thought that was only for when the final patch gets approved +JSON can only be encoded as one of several UTF encodings +Wanna see something cool? Try doing an svn add for all of the new files and then do the svn diff to create a patch with both the new and changed files +I honestly have no clue how WLS could ever pass the TCK serious this is pretty obviously a bug in WLS c and whatever we do as workaround in will cause problems later on +That makes sense +Makes sense +I see the screen shot but Ive tested it in recent builds and it seems to work +to see if a test trips this well need to check the console output for the run and search for +I have a solution for this bug which also solves bug +This sounds like a good idea to ensure teardown gets called and things get cleaned up +This one apears to have been applied a while back just never closed +Checked workaround attachedThis seems rather fixed and should probably be closed +This one seems to be related to the whitespace trimming reported in for the changes needed in label rendering gives us no time to reach CDATA section works just well +Thanks Andrew! +DOM L serialization performance test. +This is actually a problem with Javas +The same needs to be done for this issue is that if you are shipping then shouldnt you be shipping regexp as it is one of its dependencies? This is probably another issue altogether but the POM for this artifact is invalid and needs fixing. +Got it +Reassigned to myself to resume work on this issue +Lets avoid themistakes of the past rather than repeating them +Looks like its completed great work! +IMHO the usage of should be reaplced with and should be used more directly +I echoed Heikos comments here +have been build in Brew. +Ok then let me see if we even have our Min +Id be quite interested to know what sort of fsync options were being used on the databaseserver in question where this occurred +Please close it or provide additional info +Added missing plugin +ORO users have been informed and a notice sent to about the list closing shortly +Patch looks good +See if that NPE can be intercepted by Seam CatchI see there is an example in the release +Ah did you test on windows +In the java social api the Address object is incorrect with a formatted field which should be unstructuredaddress to be compliant +I propose that we change the timeouts to behave as follows +Running patch build through hudson +It replaces the that everything writes to with an that records what is being written +Moving this issue out of into since its not completely baked yet use it! Send comments +maybe its not this test +This is not a problem anymore. +what hudson thinks +Could this be generated by the fact that servicescdh depends on serviceshadoop? Classpath issue? Broken local configs +Previously we only look at first recordsCan you do some benchmarks to see if there is any noticeable difference in runtime because of the delay in turning map off +There ought to be unique id validation already +Im updating the docs to include an example of how to do it. +The third attempt worked +Also Ive tried shutting down one node moving its files to another box and starting it up with different IPhostname node was up and serving properly but problem persisted +If we put it in then wouldnt have access to it unless we allow a call from one of those setup classes to the other +The magic scares me +should be made public just for testing purposes +Attached a patch +Abbreviated the css class names for the pickList and orderingList +I prefer to think of it as the boyscout helping the old lady across the street but not making sure she has two legs +looks good me to +Thank you for giving me access +I apologize but the test passes on my machine +Some UNIX boxes are configured to purge old files from tmp in order not tu run out of free disk space in tmp partition +This page needs to get updated by someone very familiar with the Tomcat integration +I can try this and see if I get a clean build +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Would you be willing to supply a patch for thiswould be great as not many of us are using both and the RI so we are always in need for people helping out with testing and debugging on the RIregardsHave you tried putting a hform tag around the navigation menu? These are links you are using here and in the RI you have to wrap every commandLink with a formregards I already tried putting a hform tag without success +One of the purposes of functional testing with FEST is to simulate user input +I did find the solution to this on my own system as Tomasz says there was more than one jar with the implementation in the classpath +The javadoc also states that future merging might not preserves blocks +The above mentioned issue is fixed in this patch +who is our designer +trunk Removed some of the +This is a regression +Or removing For example if the transformation converts Strings to Integers verbiage? To the casual eye current implementation is confusing +It is working Sharzon for your helpThis window error is pretty usefulThanks have changed a easyXDM to usual ajax calls. +This patch is for branch + +No it does not +Caching Host patterns so we dont have to rework them each time +time for extensive testing is falling short so closing for now. +anything in the error logs or exceptions +Personally I dont think backwards compatibility is critical here +This issue occurs only with transactions to work around it use transactions instead +Unless is going to upgrade to a newer Hadoop version soon +Attaching a proposed fix for this issue +Yes absolutely its a real life scenario! Weve warned people about starting concurrently before recommend adding a s delay but that was just to avoid a bit of messiness around nodes forming singleton groups and then merging +If theres a work around I recommend that we push it to which will be out in +Hence we should keep the unit tests as is. +It is optional but only if the pushdown type is set to required +Another updated patch +This issue has been fixed and released as part of release +! for the little patch +Chris could you plz post your patch here +BrianThere is no plan for another release at this time and this is slated for the. +Unless you have explicit limits on the sizes you have the potential for a DOS attach whether intentional or otherwiseIf the protocol does change you would need to also have the AM be able to limit the maximum number of requests that would be allowed +Same place as before +Attached is a patch proposalThis change modifies the behavior of cp as follows if you run java sysinfo cp then if you have either the jcc driver or the derby network client in your classpath that is treated as success +Thanks for the update +It will be easier to maintain Repository pattern with ROO during process of development +I havent checked on Windows bits yet +I am running all test cases about Trunk +I absolutely consider this issue to be blackwhiteI agree +we support Dali JPA via Hibernate JPA platform. +Leaving unresolved in case theres a release as this should be included +I would like to get back for a moment to the in out interceptors do you think we should spend some time to figure out mechanism for them as well or well address it later on if there would be positive feedback from community about this featureThanks +the entire layer so as not to render that one feature which can be slow and if you do lots of switching between features this is a useability +I was using a reference implementation +For the patch the pthreadyeild problem seems to be that the autoconf tools arent setting the correct flags on that OS as there is a branch for using schedyeild so the correct fix is to determine why that isnt happening +Integration tests were updated. +i didnt hit this on latest blockers and critical issues are resolved but not verified +Closing issues solved in released versions + +inlinetrue would display a PDF right within the browser while inlinefalse woudl open acrobat reader +Shouldnt we care about the correctness and order of execution in file systems? +i just tried again from HEAD the AI error occurs but the server not found problem doesnt happen +committed review would be appreciated +I can reproduce it too for in +The mutex would be for code internal to perhaps used by the +removes the useless timeout entry from for symmetry +I believe the implementation to be done but cant confirm until is fixed +But ADB is coming handy +seems to be an time out +I think its reasonable to wait + to trunk and +Patch does what it says in the tin and was tested to work fine onDarwin Darwin Kernel Version Fri Apr PDT rootRELEASEI iThe Glorious Glasgow Haskell Compilation System version Should be committed before releasing as code will not compile without +The current XML Sequencer will be changed +The tracking library is moved into commons area +hope I did everything the This is my first patch for GT +Low risk patch +Ah sorry +There might be tons of side effects waiting for youMost of the do NOT take a relative file url but only absolute ones +Released. +Would be lovely without this dependency so people can use it in non Spring environments +Thanks! +Patrick Ill pull your patch and well test it out here today on the benchmark +The reason why the persistent reply to example at the nabble discussion is because Camel will have to use a Dummy value for when there is no expected replies to receive +lgtm +Unit test resultsant package testAll tests pass for default hadoop versionsant clean package test testreport tests pass +Actually I think you missed my point perhaps we should discuss on IRCFrom a customers perspective it is better if handles this cleanly rather than have the previous situation where the app server had to be bounced so that the stalled messages could be delivered +The only question I had was whether it was worth it to embed a call to decompose in the method +OK assuming no one objects I am going to commit this tonight +Upping the time did the trick for me as well +Im sure it gets cached except you explicitly disable it for the Component +Also the state of the connection makes this not which a connection per request would fix +Running from my IDE is giving my different results than what you say hereI see errors about not being able to find the test to answer your question joining a transaction for JPA is only valid in the case of CMT transactions with app managed there was an issue with transaction joining whcih has been fixed +Sorry meant to mark as resolved +What would also help is if you were able to share the Exception traces you were encountering before reconfiguring the column definitionsOn this issue +If all tests pass then I should commit only its a pretty big change so I just sent a to the list warning of pending commit +good point that is another reason I dont think the pool of connections is a good choice it makes it a lot harder to reason about ordering +Temporarily reopening to update release note info. +Please backport it toBTW the encoding is a little too strict x xa and xd need not to be encoded +Verified with IR. +hibernate tools trunk is already working against Ganymede +This is my latest patch doing unit tests of rolling restarts +What branchtag and revision should I patch againstIv tried trunk and tag HEAD and the revision listed in the patch files but I always get a malformed patch errorMostly around version on theIf I manually patch the pom file compile and use the jar in my project build process I get an related to the class orgapachemavensharedfiltering +Ive tried +I probably messed up a merge. + failed +Changing the method nameIn the splitter split Doneresequencer resequence Doneaggregator aggregate Donedelayer delay Donethrottler throttle DoneIn the delay setDelay Done +We should confirm this before the release +Username kimerinn +although I will believe all of that when I see it +ok +rev Resolve and groupsURI relative GERONIMOBASEThis allows each instance to update its own copy of and +Too late to initialize issue +Patches have been applied +Thats the best part noticed it +tracks this. +No idea how to display that properlyOr trace INSERT or SELECT there isnt as much output but its pretty useless and inconsistentAlso unrelated to that requesting tracing info immediately after the query doesnt show all the events +This issue should be closed +Attaching which bumps the rev to +Attached a patch with a RobbieDo you mind to review and commit the patch +omg +catching and destroying the external process seems to be the way to go +I gave this issue a very uninformative title and description +I suspect it was a glitch in the migration process or something like that +but that wont stop people from doing it +Remove module editing enabled! FAIL it should still be disabled unless you click Publish on the serverCreated for that +Im looking forward to your work +Most important is that defaults dont change the current way it works +Unfortunately its not as convenient as I hoped forI did as you recommended attributed my ant target based on with javasource in order generate my schema based xmlBeans sourcesUsually I do compile with Xlintunchecked in order to get maximum quality at source code levelIts not a problem to avoid the massive warnings attack from javac I have to separate the schema induced xmlBeans sources from my project and have to compile them apart without XlintuncheckedIts just inconvenient and I thought javasource is supposed to fulfil that gap +Committed. +aaceefbddedfabe trafficserver script error message in ubuntuIt seems like theres a lot going on in this ticket but its pretty old and only the rc script issue seemed soluble +This is a development feature that should be turned off in production for optimal performance +v patch works with Mac a Linux amd commons daemonDont have SolarisiWin to verify the osarch properties to resolve the TAR URL +This entry needs to have field protocol avail for config +Moreover Chris we could imagine to add a new boolean metric to inform if the code coverage information is available or not +So a question is what tests are putting stuff in the heap that isnt cleaned up with a full gc after the test is overYoud mentioned adding the pre and memory usage data to the test reports +Task is done +I wanted to update most of of the stuff is unclear +Resolved with +Namit thanks very much for the valueble feedbacks + +yes if you try and access the state of that proxy you will get an error +This patch also fix these problems as well +Thats correct we would want to shut down the cluster even if thats in safe modeThats a good catch to remove the TIPS which are from recovered jobs only +No the last we tested it wasnt working yet because of some missing token propagation +consider connecting to the slave database server for reports think you are almost done with this task just assigning it to you so you can close it when you are +New patch because of changes from chance in getting this committed This is rather annoying for windows users +one detail the enitymanager on the singelton is not an extended entity manager but the function is called from a stateful session bean with an extended entity manager +I have merged this to. +Did we upgrade the surefire version recently +I have forgotten to mention that is happened only when I specified location in plugin configIt seems plugin ver doesnt have this bug +course if were the default then this would be a my preference but its too late to argue that now. + +I have committed the patch only fixes the issue with the hardcoded portlet tld jar name +I dont think it needs a test case +Rather my point is that theres room for both +Verified and tested this code +I have successfully run the svnadmin verify command on that new repository +Implementations of that dont adhere to the spec have been arecurring problem see the linked issues +The jars were renamed to match their existing names so no changes to other files were needed +Build test successfully passed on Win x +Tim has a good point about the default namespace though +Deployment manager API has been rewritten to allow for multiple content items +Since r the exception stracktrace is logged making this more annoyingThe attached patches makes s list item model fail save +Thanks spotted. +Didnt realize there was a hardcoded test in +Thank you +Len is this with JDK or JDK? The title says but the description only mentions +Venkatesh +Also how the heck would you test this there are no tools that I have seen to build and send extended requests. +So the package referred to the auto classes not the interface +I just committed this thanks Terence. +All I did was add a test dependency which was previously included by accident until Andrea fixedChristian do you think this fix is satisfactory? If so I will resolve this regardsBen +The test case works i think its fine +Fixed the problem its due to the unpredictable thread scheduling +duplicated +Command line isoptbin of xc Value cn Server +We have kernel on Fedora Core and the problem is there with client option does not affect the error +Retargeting to and +Try +All types have already had a serial added in +Committed to trunk +fixed +Failing tests for numerical issues are testFailing testFailing testFailing testFailingFailing tests with very wrong results are testFailing testFailing +We will merge Twix and Snickers messages into a single set of classes +Screenshot for Can you comment if this is a UI issue or API +I agree with Murray though the docs should never refer to an unlreased rev +As I created a for the enumerate support please allow any discussions related to this to take place at. +So if we do not nullify the the Context would linger on the thread in the pool +Indeed +Committed to trunk revision that hides an accidently visible method with Directory and againCommitted in trunk revision will post a patch later adding the needed deprecations to! This issue is kept open until the deprecations are solved +It means we didnt fully implement changes to support typed maps +It seems as though the most surgical way to do this would be to introduce a formatting that delegates to another and has a configurable to find the pattern that triggers a parse from a configurable metaFormat +That would really be great if we could reproduce the problem in order to fix the underlying issueAs for well its not actively developed these daysIf youre willing to contribute to it that would be great for those relying on +can retrieve the other instance +Mubarak After looking at FATE and whats involved I think it a bit much to expect that we build that as a prereq +Ah! I am running +I doubt seriously this will be a problem at all in practice so lets hold off on doing so until it does +Im somewhat astounded that youre managing to run out of DTM even so current code can have DTM open at once though if you have large documents a single DTM may require multipleSince you said youre using Xalan Id suggest you try and see whether the situation persists +The Serviceclient make use of the to create the +Alas +could you post the stacktrace as well +Thanks for the review and commit! +Thanks Stack for the review +Please against trunk branch +Cancel patch to wait for +Patch to bin thanks for the report +demonstrated a npe during this operation because Derby was relying on database context to get the correct Collator object +Moving issue to a next issues to next release +escalated to bug GaryI have asked the user to take a look at this we will hear back from them has been released +I think this addresses all the feedback except the setting +I would think there should just be a link on the side panel +Hadoop was released. +For some reason the new file srcref was committed without its svn set to native +We have in the past made some exceptions in URI syntax to simplify platform compatibility +You are using on these +Id say dont commit unless all pass +Improved testing is always much welcome. +I mark this issue as upstream because there is no method to get monitor client area which consider Ubuntu bars +Michael I guess thats a valid point +The introduction of types in will shake things up enough as it is lets see how people and projects use them before promoting this stuff to its own projectYes the serialization formats defined in are designed to be language agnostic +Form sign in field first name for examplenataliya surname surname last name last name +I am keen to get this bit for Cargo Maven plugin Any ideas on we could achieve spawning Appserver launch in a new process using Maven +yes +Im working on abstracting out the bits from now +This test corrupts replicas of a block and checks to see that the Namenode detects this situation and replicates the remaining good copy +ant is is done +Please confirm or elaborate otherwise. +deals with a problem in the display on the use of standard properties such as and so on but they are correctly evaluate by Velocity in the generation step of the archetypeThis bug wants to manage correctly the use of properties in the same way standard properties are such as the description of this bug spotlights +The query below results in XML thats not and the HTML generation fails with a select from where ibmreqd a b c +Need to be verified re not developing Deltacloud any longer +Ive made a patch to make the extraction of the version of a more robust +Im not really able to reproduce this issue since this issue was reported weve had +Java Data Objects plugin Apache resolutionoptional resolutionoptional +Patches sync the code with committed the improvements back to YARN. +I also noticed that the value in state is case sensitive +committed to as commit acaedafabfeab +see and +Together this gives the major status of the clusterSo for a cluster that had a with one collection shards and nodes for each shard you might imagine seeing circles arranged in a square under the heading collection +I have attached the project I am trying to load +Merged into the master branch +bulk defer of unresolved bugs to +Anything else should be considered a bugenhancement request and separate issue filed +I wish we didnt have to rely on encrypt and decrupt being idempotentIf no one gets there first Ill apply jacks suggestion in the morning +The default encryption algorithm used by Derby is DESCBCIt would be great if this change can be made +Bit of a hack but it now worksThanks +It supports most of the attributes and subelements available for the java ant task. +After starting over with a fresh installation this worked as expected +It works now thanks! +No worries +Take a look at resources for a suggested +Are you trying to do some kind of code review or reporting a bug? If it is the latter please provide the following exact steps you do what you seeget after those steps what you expect to seeget +Did you accidentally forgot to attach the patch or did you accidentally changed the status to Patch Available +Backport to branch +A patch for this for the report and the patch well take a look. +Douglas should I create a new patch? Or you correct it by yourself +This passed tests so Im submitting it as ready +Resolved for now might revisit in a later release + uses a which is bound to the calling thread +or is that just a accidental deletion +I dont feel too good about adding something to the grammar I dont understand +Marking it as a known issue is fine as far as Im concerned +Very close to being ready to submission +This fix broke the by only enhancing the test classes +This issue should be fixed in the upcoming release + +I can open a new issue at a later point of time if more work needs to be done +working with Consumer etc + only takes inInterceptors is a array list you should pass a list object reference instead of an interceptor object reference +as I think there is no reason why I should declare the bean definitions in the fragment bundle and could not keep the implementations outside +Can anyone attach a testcase for reproducing the problem with the Norman was having problems reproducing this +deploymssql passed through the whole integration testing with MSSQL database configured in the aforementioned way +Todd can the mock object implement a new interface that extends and Closeable? Will that solve the problem +Do not exclude those +What is the intent? +Craig yes youve summarized my proposal correctlyWe might wait until to file the JSR though +cfecbcffbfecd SSL certificate reconfiguration only works once +Sorry +Yeah I agree it can be confusing since action has a specific meaning for forms +Would you please check again? Thanks for your help +minor code improvements were madeNo changes to search algorithm itself were introduced +Attaching logs +Damien I had to rework the code logic a bit +reopening to use it as the root issue to track progress over the implementation of the changes fully landed on trunk +Hi OmanandI have not had the chance to work on it yetProbably in about days I will be able to do so +Thanks for your input +Attaching an Excel file with examples of some of Excels formats and a couple of custom formats +This isnt the first time Ive seen this +Any update on this bug? The fix is trivial it should take min +In example +Obviously it is not big deal to move the whole validation process into actions and configure the interceptor not to accept any parameter +Thanks Jackie. + +I believe SOCKS is toggled on by default + +Closing this issue +Now the query engine should support this xpath query however there needs to be an index that supports the repexcerpt +Hence this is the exact problem I think we should fix +BTW i created a issue for creating a new portal template its not but PS i have not right to edit my comment +patch for branch +If the bug still appears Ill reopen the issue +A simple patch that just retries incase of META +Thank you for the report +Removing JON and CP from fix versions since it affects the embedded console not shipped with +Since the need for has come up a couple times it might make sense to move it to add it to the public api and document it as you suggest +Uploading +Works fine here +What does the code that does this workWhat does it checkAs the fix was not to the mail subsystem but how the paths are handled in general +bq +That said I have to ask what features are available in JDK and are not available in JDKI can only think of better apt integration with java compiler +If they dont it is considered an only cases I know that would want to automatically create an is machine learning +this has been bugging me for a while +Im going to have a look and see if it can be incorporated +I suggest using that script but with the small change of the default value JBOSSCONFdefault instead of production +Let me know and Ill change it +Minimal failing to fix for based on issue reported on mailing list but that turned out to be user error. +Version M has been releasedClosing the issue. +Thanks for the comments +verified on Mac OS X +Need to add this to the docs +Jochen Im only proposing that the setter should not be generated +If you want this you could create a new ticket and assign it to me +Let me document that in a separate file and have the +Marking for for remaining work of actually setting up a GAE platform that doesnt assume things we assume for real platforms +When this is fixed remember to turn the forrestbot cron job on again +Same fix plus a modification to which tests the conditionThe test succeeds with the patch and fails without +Maybe the term wire format is wrong I should retitle this +I will also upload the performance number file +I had jobs with tasks expected to run minutes +used to just have one now it has get and get +I added the rewrite trick to both highlighters remove the field requirement in the and hacked a reusing that simply delegates to the shadowing the field +Code repacked +Transitioning to waiting for has been bought by a rd party +This is a rocket science issue +What I saw is that in trunk we treat that table name differently than in the latest version of so my original patch for trunk was working ok but porting it to was failing +Also changed a few places to make sure the error message looks nice +Someone else will have to patch the controlhaus stuff though once becomes +However the Boost bits would be included instead of existing in a DLL +I filed to do that. +Attaching a unit test that shows that it we can beef this test up a little to try and better replicate what happens in the tck env +What happens they introduce a lighter ident? I think that until the value is computed it is appropriate to leave the idents as idents +Hi All just a side note may be a Long parameter can be replaced with String on the server side so that the custom code can validate and throw a more exception in cases where a parameter gets in somehow though what we can also do is to either use a path expression limiting the value of user id to digits or wait till CXF integrates with Bean Validation sorry Francesco was doing this bug on behalf of some one elseWe have over come the blocking point now. +If you find further issues please create a new jira and link it to this one. +Any user who started the tomcat service for oozie is what the oozie guys told me +Test logs +We need to understand why this happens and fix or wontfix for +Revision the target version to that this improvement will be shown in the changelog +Sept Skype callDesign PointsSimple clean particular create clear distinction between App Server to avoid confusionVisual Metaphorsgrid structuredistribute loadscloud computingCompetitors to compare couple more commentsre the cross section looks like the letter A +There was another fix from Bugzilla for the same problem Ive tried my best to merge yours with that +Looking forward for more suggestionscomments as we are talking about on the list now +Thanks a lot for the review Darynbq +Alright +Note the webapp wont deploy with that +Is this issue still open now that were using jetty +Carl are you working on itWe might need it internally at facebook so wanted to know if you are working on it currently +bq +I am trying to passon the DTO with the hibernate domain object on it +This will make consider this as fixed for the time being since Im not aware how you could obtain instances from the otherwise +The javadoc warnings are unrelated to this patch + tests pass for me when using based implementation of the HTTP request factory +patch applied to the current version. +GaryWhat exactly do you want to see here? We already have a mechanism for remote deployment +I modified the NOTICE file to include sbt +And why do you think that isAnswer cos youve made it so bloody polymorphism is a basic feature of OO +According to the customer this patch doesnt appear to be thread safe +And for handling both the cases I think entity is the best place to handleBecause after reading Ratneshs comment and looking into the patch I found that Read status is changed only for the records which are specific to parties of contact list and not the parent communication event + +added to +This is a good start +Found the problem +I would fix that and see if you get stability that way. +Thats what we do for the Synthetic Control stuff +Fixed in CR +Sorry disregard that last comment of mine +overwrite should just not worry about blocks in underlying read cache +just check if my example and so short description is clear from your side +First open an issue on user forum describing the problem far better then its done then if we establish that this is really a bug should you open a JIRA issue +cant remember the ticket into +Thanks to you for the implementation! Looking forward to this new feature becoming available and really hope it will make it into +Thankssvn rev +Moved to where Mitosis will be to we wont release a is unsupported and has been removed from trunkThere is a new replication system in the trunk based on RFC +Closing it for now +OSX isnt a support deployment platform so Im not really worried about it holding up +Committed to trunkThe contrib tests pass in my machine +From now on that is initialized as +We have decided to postpone this indefinitely. +Alexey Yes the provided test passes for now +Thanks for the fix +Mmm this is not straightforward to use a nightly in eclipse mainly because we have a naming convention for nightlies which conflict with eclipse version naming convention +So Test Case indicated this problem. +All verifications are working as expected now +If schemas are in a subfolder you need to specify an pattern like +Moving to +Based on user input well call this fixed as of +Please ignore the above comment +Committed revision +It wont be today but I will start porting over some of the tests from the existing based structure to Pax Exam +passing junit test for junit test for tests applied +yes I think this should be kept intuitive for general usageI take a glimpse of the grep result strange that we use doSharefalse in patch tests pass +The correct error for validation seems to be but even this is not supported by browsers +Verified by Vladimir. +I think Laura was having some commit password problems and may be thats why its not committed yet +What are the bounds on the reasonable amount of timebq +And this is very true +We would need to know the contexts where the references were created to track down where the leak s take this discussion to the profiler design forum +NPE is not observed now +Reattaching the patch +Ill update the Wiki page tomorrow. +committed +Thats a disaster +Just continuum users will have access to the page which display the queues +Since we dont want to block the UI until markers are updated we can remove the job that builds projects and waits for builders and validators to finishA quick fix will trigger incremental building of required projects +Committed the enhancement without the cache in +Confirmed on Mac +There have been no comments from the bug reporter in ten months to elaborate on the problem so Im marking this as invalidFrom the few lines of text attached it seems like this is a user query and not a bug report +m folder and it started to work +After chatting with John we can turn this on because a proposal is forth coming that will limit downstream profile activation to os jdk and active +And thus migrating currently state to be global state would be a significant behavior I guess what Im getting at is Id recommend that all the currently global state should be associated with the TM state and not kept global to preserve current behavior +Another option is to pass a parameter to which indicates to include hadoophbase classpath or not +Change made which should deal with this situationThanks for the report and test +This is probably related to my fix for +Included in release. +Are there existing test cases in this area +Committed to trunk +Thanks Amareshwari! +Please confirm it has been applied as expected +Only with a bookmarkable request the should be hold on to in the request cycleBut what should happen in your case? Why are you adding the args? Isnt it supposed to add args +Patch that fixes the resolution of the class name in type mode +One extra thing regarding I also have to test that when merging like you describe above happens the cached docNumber for an index that did not change might have a different position in the after segment merging +Can you reproduce with +Im closing this as fixed now. +Implemented +Do we all agree that we should go with whatever name Lucene selects +For exampleJun PM mkWARNING No type for variable Jun PM mkWARNING No type for variable Jun PM mkWARNING No type for variable Do you know what causes this +Is this patch essential for or can it be moved to and? Id like to freeze as soon as possible so we can make a release candidate +But sometimes we need to setup the IDE before createimport any projects. +This should correct the error TedThanks for taking a look +I needed this in the branch and created a patch for it +simple after release. +this would be helpful for plugin to next to next version +What happens when an observer defines an parameter but the event was fired using instead of Event? The spec does not currently seem to define this +Please reopen if necessary +Rebased and merged into the master branch. +bq +here it correctly shows the matching imports +That is why I want to to talk to few folks in the design review +Hi AmreshwariThis issue seems to be invalid to me as the constructor of will make sure that the is initialized +not sure how to add a test for this +Please upload a corresponding patch once the trunk patch passes Hudson +I see this again current trunk build from fresh checkout fails in after warnings when starting jettyA simple java jar target c sling f in launchpadapp will not succeed and produce the exactly same message from top of this issue +I added the tests with Ignore in revision I see your point +See the patch I will add to + +Since weve decided the colocation check is not ve put the proposed solution in the forums for discussion +Add a repairedAt fieldSince I appear to have confused virtually everyone on this point were just using the presence of repairedAt as a boolean to mean dont bother repairing this again unless +fieldName city notNull sizeMax add field reference jpa class +The caller component does not need to do its own datastore supporting the hint GENERALIZATIONDISTANCE and PRESERVETOPOLOGY MUST be able to do a generalization preserving topology +Fortunately Javadoc is released which reverts to the behavior and we recommend everyone update to it +In looking at the patch I cant see that this actually applies to +As far as I remember is trying to upgrade the old config file to a new syntax and the code wont work if that does not happen +IMO doAs is a easier not to get confused +I see around a increase in reads against a single node with this patch +It would be nice if via the plugin configuration it could be possible to add As srcjava directory to Cs classpath enabeling live editing of sources in a library while using the devlopment mode goal gwtrun +For the proxy users you shouldnt set the +IMO the noExit flag should be turned on at all times and not made configurable you should be throwing a if there is a original patch did this and it also allowed you to continue the build even if the weblogic operation failed +Thanks Alejandro +I have verified it +What process should be followed to update this entry +Attached batch file to run the test case +Duplicates +Linux mvn +We also have a tool +Thanks Claus I had a sneaking suspicion that this had been addressed in +Freddy if I correctly remember then this will not help because only quality profile loaded from server but not settings +AndyMay be can be used here +I dont know enough about jexl and jxpath to knowwhether this is reasonable +That should solve this automatically without adding the retry window +Ive done the same for the LDAP account mapping code. +I committed this to trunk and +will open a new ticket instead. +Thomas can you please have a look at this for already deciding whether well proceed with this or JuergenI wrote this template for Spring +Please open a new one if you still observe this behavior. +Reported as being fixed. +That means it doesnt have the information about the second node +The reason I ask is because I see precisely ONE Field in the entire trace +I have made a couple small edits to clarify it a bit more +Even I could not understand it + where fall in the hierarchy is inconsistent between as and asThats not true +But if you are digging around look at the gmaven bits +close for the release +The only change is in and that passed on my box + +I just realized that this works just fine on our Hudson server +Fixed for Apache build +the file name would be something like Shouldnt we have a separate bundle for each enumeration so that applications can easily drop in new bundles for their properties? Thus the resource name should perhaps be orgapachehadoopexamples in this case +Will commit if tests pass. +This zip contains Eclipse Projects bundle which has no dependencies to Spring DMfragment for Eclipse Gemini Blueprint usingfragment for Spring DM usingIf you like this idea I will create clean patch with Angelo I think wed need to come up with the interface build around the Blueprint concepts exampleimport wont work with AriesMay be we should introduce two interfaces +Up until this is how hinted handoff worked +Can I go ahead and apply the patchI will apply the patch commit the changes deploy snapshot version ok +Cool please watch the mailing list where well hold a vote to add the plugin and yourself to the to the sandbox +Backported to Apache branch on th Nov +trust me hibernate itself does not ignore the value inTry for yourself outside of seam +I believe this test version better reflects the expected reaction of in case of blocks length +From memory I attempted to upgrade this about a year ago but it caused problems with weld +They are not necessary +A new revision of the heap patch based on Scott Cantors original +As Heiko and bob wrote jrubyutil is a handy workaround +maybe it just misses some conventions +Please check that the patch was applied as you expected +The ability to include Jira in the general search +No different than any other feature or bug present in certain plugin versions +Ive made some fix to acceptJson behavior now it should work with missing attributes +First cut at a doc for writing yarn applications +Thanks! +Im going to ahve to look over this one in more detail maybe later tonite Ihad done some modifications in my checkout on my machine at home +See the mailing list archives of last week where I answered this last week and gave more details +here is a patch against turns out that the standard query parser doesnt group expanded terms internally while our core parser does +the optional parameters are not set in the submit call in r +In fact the very first rationale for us to introduce pluggable is that we want to isolate read and write ops +Its probably caused by some edge condition interaction with my own local event bridge which is what Ive long suspected +the bug also exists in a careful review of the problem the conclusion was this problem depends on Handling is unconsistentThe code proposed in practice is just the initial step already proposed there to fix that issue +Test performed with negative results for both and +I will take care even if I cannot assign to myself +There is access of this information +What I meant was that when run inside chrooted processes the ip address remains the same but the same path on dfs client and datanode refers to different things +Sorry for late reply +The class just propagate the thrown exceptionOne more thing we may want to encapsulate the user codec excpetion in order to give a better information about what happened +I like itDepending on the size of the index it seem like keeping the range query option would be good the Trie fields do remarkably wellUsing would also make it easierfaster to get minmax values if we want to support open ranges and percentages +SPRINGCLEANING JIRAS As per Marks last comment. +Some aspects may be for this too +we can try to get that data from another source if possibleWe already do that. +BabuDid you try any of the RCs yet? you attach your complete wsdl file? the wsdl you have attached seems to be incomplete +Have resolved the conflicts and uploading new patch after running the tests locally +Runping it doesnt get to minutes until it has failed times +I verified the fix r +Rescheduling for +The reason in particular that this lazy loading would be effective in a production environment however is due to the previous behavior loading all application resources into memory that are described by large widgets libraries even if an application might only use a small subset of these +I applied it and did tests against the console didnt see any issues. +Why embed these options in the code rather than use the string provided in +to hudson +The architecture is changing in and will bring significant performance improvements +This almost certainly is not related +I think I was using AS or but I may have been using the version of the hornetq plugin that shipped w JON which is not the latest version +If I have no plan to use thin at runtime why should Thrift require me to install it +Could equally apply the same code to if this was acted upon at a JDO spec level. +ab is a String not a Character +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +Thanks Bruno. +Bulk changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited + Grid Im still seeing the issue when using the build +Was trying to make the minor changes stack suggested and found that tfile no longer applies to trunk due to the removal of Lzo codec byGiven the performance advantage of lzo it is probably not a good idea to abandon lzo support completely +Closing the bug with by design resolution +Cool +Ill go ahead and commit this to and trunk +Sounds cases found for a slightly generalized enrichmentaugmentation api resources with a common egparent should be egsibling for every foafPerson there shall be a foaf every non literal resource shall be subject of rdftype rdfsResource statemet every egCity shall have a egcurrentWeather property pointing to a bnode with temperature and enrichers as examples how this could look like +Thanks Adam! +Needs a code review +Patch obviously doesnt apply to trunkTested patch locally +Excellent +Committed to trunk and +Fixed in r added null checks +I dont think trying to use the same POM really makes sense here +Same patch just keeping it with the branch of patch failed on one chunk on latest revision +Upgrade Meaning the old hosts would show incorrect capacity so you need to correct them in ophostcapacity table and probably the host table as well +The activatedeactivate were already asynchronous but the enabledisable were too which is contrary to the spec and IMO a pretty bad idea +Would a solution to this issue solve? If so it might be good to say so there and link it to this issue +Need to fix this reopening +If someone really needs these they can use JAXP +Looks good to me. +Major thanksThe only change I made was to add a flag to the base to allow any subclass to specify whether they need the request caching or not +If I revert back to pre everything works fine +Thanks Jarcec! +Creates a parameter declaration file for the Working with Derby Guide so that version numbers can be substituted in that guide too +Hi MahadevSorry for the delay in responseYes +All tasks completed +That is why I dont have the errorI tried to train the model using the Train tool and it failed to loadLoading POS Tagger model +I am backing out these changes until the cause has been fully investigated and fixed +I did not think about an uninstall will take a look at what it will take to extract out the same build file during the uninstall and execute it there +Yes I thought about that but for some reason didnt end up doing that +Could I ask ESB developers to help me with that list and suggest what should be validated in the first place? With the set of rules to go we can decide what can be implemented for and what moved for a later version +sql scripts from the storetests directory and suite to the and removes +Because what to do with documents that have that integer valuein general im just trying to make sure we keep things sane here so that the underlying shit is efficient +Added clustering component. +It is almost the same but not a duplicate +This is actually not that common +Another Eclipse mystery +Viewget has been updated to require the return of a Serializable object and appropriate changes have been made to and. +META + problem has already been resolved with +Committed to trunk and. +Updated patch +These code paths are not just related to append +i have tested this withACF no idea about or +I can only works on this over weekends or at night +It is important we get something there before launch because of how we focus on it in the videos and now have a link to it from the project page +That issue backs out most of the work done for +Please let me know +Wifi woes +Thanks Sagara Few more questions +No change planned +All tests pass w it and I went and disabled the same tests that avoid codec +Branch now contains deploy server shuts down cleanly with the test. +HadrianIf we can make sure the deploy box is running with JDK the can be deployed without an issueIf we cant do that hosting the component in the forge could be a better solutionDavidWe can host the component wiki page in wherever the module is + +OlegCan you try specifying a field value anyway for your bug up above? I think this is actually a bug in the Lucene Spell checker +So any ETA on a possible fix? This is going to impact dates for us soon +A new patch added the ORDER BY clause +But now some tests fail because they expect a MK implementation will consider concurrent sibling creation asRemoving the child node list in the parent node as proposed in would resolve this issue +Didnt notice that it managed to make an actual assignment reverting +Applying this fix can not affect those subclasses and only allows for new subclasses to be implemented in a correct manner without having to equals and hashCodeIt may not be a big deal to have to override those methods but it is annoying error prone and unnecessary +If you need help let me know +Do you just use diff to make the patches +Thank you for reporting +code that is proven to be incorrect via user testing although without automated testing + +Good points +other jiras have been created and linked as appropriate +If it is not something that needs to be done then we dont need a TODO for it +Hi Ben +Yes we need a separate HICC user guide to address this issue +Ok that makes good sense Ill work up a patch. +Wayne to the rescue! Thanks very much +taking this off the list because nobody is signed up to fix this +Raising priority to BLOCKER to get a firm decision on this +new patch against the current Client after There are not new tests in this patch +The Remoting RMI transport is a little used transport and is not relevant to any version of the Application Server. +Knut thanks for the pointer +Ive removed some unused imports from the patch and then committed it. +thanks Shai +So I think its safer to release first and after that apply the patch to not get bad reasonable +We have hundreds of them so them all to manage their own is definitely not something we look forward to +get the resource type from the resource above it contains the old rt +AFAICS I addressed the most important places +added javadoc and fixed +Anyway command are optional it is just option nowAnd i think porting that to windows are not hard +We only test with and support versions fyi +It is strangehaSo I checked the related code I found that datasource attributes have several gbeans to make the change serilized when the server shutdownhowever the connection pool does not have oneWhen I dig into the codeI found the code intended not to save the changes? Way is thatDo we need a new to save the changes to itAny comments? can I fix this issue? Using +Its like if the master was dead when the split took place the master receives only the result +Thanks Mark for the update and for caring about SEAM users! Im sure everything is getting much better with DELTASPIKE Just one more question Are there any snapshots to download? Didnt find any jars neither in the download section nor in the maven repository +Hi Vince This is related to new feature that is recently added to trunk so it should not go in +this is fine +will commit if tests pass +I let is in the issue only for demonstration purpose +for now ive removed openshift feature from eclipse marketplace +This changes the so that it passes the whole path to the +If there is a problem please open a new jira and link this one with that. +So initial triage would be issue at the xrt layer +This looks like a really nice feature to have but all things considered its pretty low priority +Guava is not in any target +I just committed this +Fixed problem with collecting method proposals from interface hierarchies +Tried with MAC OSX and was able to remove the +Attach a patch that adds the offset support +Rebased to current same patch applies to trunk too + Removed the requirement for read permission on for backup torun under security manager +The minor issue with +itests are any tests that has cyclical dependencies or requires that the packages be built +Fixed in Felix +done +It sets up no authentication either way so obviously that limits the issues that can be linked +actually is and will still be logged as an ERROR level log with the patch I committed +Is derby meant to support booting from a jar file using using the classpath subprotocol or should one always use the jar subprotocol +By default it is false no validation is done +This modification was done in derbynet and derbynet which is subclass of formerTest Execute derbynet and Shutting down exception was printed to altered log file +So I think they are apples and oranges and belong to different I didnt have to spend the time upgrading jackson in to work with Avro I might have actually agreed with you +If called without encoding attributethe default encoding is always +No luck +After the fix is obviously to do as Doug suggested and the crc chunks at the block boundaries +This is exactly the way a pool should work so I dont believe this is an architectural flaw +The remaining tests failed due to issues with attachment handling + +Is already all resolved tickets from or older +Ill fix this in the next day or so Im going to give Android like iOS a fixed version for each release and have Coho update that version number when the release is tagged +Hi Alexandre you need to execute the script extrasdatabasemysql which is available in the Sonar archive +patch applied +It doesnt seem like a commonly required need +Neither MREPO nor MAVENREPO Classpath Variables seem to be respected nor is the global Maven or a +That removes and replaces with just one class that extends and is reusable +With that fixed read fails because in checking whether the stream is readable it sees that it is also writable and tries to force a flush again by seeking to the same position it is already in +The patch is still valid and should be tested after these two Pig issues are resolved +Memory is not the issue right now but I have plans to trying a group of VM instead +Step by step +Good because I tried again to reproduce and could not +We continue to research this with JPA providers and will provide appropriate support for vendor extensions there +Anyway it is the user convenience that matters +It reduced errors from to +I am using a dialog with formtrue and replacemy so is being used to replace with the server response +the patch has been appliedPlease verify +I tried these links in which we try now on staging and links work correctly +Merged to the branch in revision . +Attached rebased to trunk latest +Making it a blocker for this JIRA +so no explicit ordering supplies after last added item whatever that is +Thanks a lot +I am still working out a good way to have send the updates for the user information +Committed +looks good +It is a copy of already in the tests directory +Since this is so simple for developers to code not sure there is much value +The old OGNL version is deprecated and will be replaced with the commons one ASAPSo again the question is is there a way to reproduce the issue such that I can test if it affects the Commons version of OGNL? +No example currently refers to the GWT Designer connector so you just need to merge it with +Hmmm seems to be a svn sure if it is client side or server side but files that had been removed previously are still in the repository +String for any solution +Actually we found that this warning is indeed needed since it may be confusing to users who use MVC scaffolding and see date conversion issues as a result of not using Springs annotation +Can anyone take a look at this one +These properties were not removed from derby and need to be documented +I just committed this +I have opened a new bug to fix this in +Committed to trunk and branch. +the module is used ok on the client now but we need to remember this as a general principle. +Ive got aboutfour or five test cases that run successfully iow with regards to usingtransienttrue on all kind of field types everything seems fine +I think that it is the one that needs to be fixed by specifying that the iterator should iter in sorted orderYou mean just as a javadoc comment +In the meantime Ive changed it and replaced the focusSkin to get the same behavior from beforeScrolling with the mousewheel is an issue though +Note that this issue only applies to GWT projects so it is no longer relevant to Errai WFK since these now require use of GWT +Also to allow to be able to return more information about a file in the future better change the interface name to be +The patch indeed helps +Thanks Kim +Given how much various other companies are proud to tout their patch counts these should be some easy pointsBTW I also think there is some confusion around how the patch gets the name +Solved using cfadmin updatecontext action in the installation process. +It is fixed in the axis nightlies. +As I am not sure where the problem is I have put both Network Client and server as the component +done +Expect one more update from Jitendra +Thats what I meant by messy code we would need both the and the method depending on an option +It does it whenever there is a JTA transaction active +That does make the test more concise +Test webapp to demonstrate that richeditor does not work with RF it turns out to be an issue in AS code then we can transfer there some or or something like that to make the richeditor workIm not sure if dependencies loaded as a module are guaranteed to invoke those listeners or not +Closing released issues. +Hello StephenThe error is self descriptive Keystore or password are nullThis means confkeystore file is invalid or password is not questionwill try to MaximDid you manage to reproduce the error I was seeing with thisDo you need any more debug logs from me to helpCheers +Reopening for review +Thanks for the nice patch Ming. +The mapping in is the default expected for each protocol our problem is that we were overwriting that on each invocation +whitespace is probably the safest of all whitespace rules in CSS +Could you provide a test case +Patch is available for review +This break centos since gcc version is by default +bulk defer to +Conflicting per discussion it would be nicer if we could improve the instead of adding a new one to piggybank + +In addition to normal configuration update events special error events could be generated +Can we get this fixed? +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Good idea Todd +Without the patch script authors need to cast a variable every time they need to access it on an interface which is not explicitly exposed +The version plugin takes care of updating both when you want to rev +Ive heard amigable amb lusuari quite a few per a lusuaria vegades tamb es tradueix com amigable amb lusuari i ds fcil +Its a little nasty but it the EIP context I always understood the bridge to imply moving messages from one channel to another +The fact that its not supported by meclipse leads to environmental differences between maven and eclipse I thought that eliminating that gap is all meclipse is about +The moderation messages Im getting do not seem to have anything in the from or cc list that corresponds to this +Also fixed a more problematic javadocs issue due to broken links from the morphlines code to extraction code due to extending extraction classesIve added to the extraction dependenciesI dont like that tests can pass when some necessary jars are missing we will likely need to look into adding simple tests that cause each necessary jar to be used or even just have hack tests that try and create a class in the offending jars or something +the url of the last cloud was storedrestored but this only happend on creation +After shutting servicemix down and deleting the +It makes build just return if there is no indexed columns +Merged the fix further down to the branch in revision +Seems to be fixed in the CVS head now +No Freddy we are able to fix this limitation +In Safari I see it for blog but it should be on home page +Patch to include annotation processing in all the ModuleNaming a simplified version of the patch in rev +can you please explain where to implement the Listenervisitor +looks as fixed +The name node ought to come up faster if its a RAID FS but as it has to do playback it will take a while to go live +In addtion this example runs on any platform though you must update the file to specify jbosshome and the location of your Velocity jar file +Now that work is well underway with AS all previous community releases are +Committed to trunk +Then uses this mechanism +I assume running this with either MRI versions works too. + mavenized datajoin but there is remaining to move the files to correct locationnote jenkins is going to fail on this since its a script +I guess more I am wondering if the double collections used internally are needed +please investigate if you can were getting really short on time with this +works in DR closing +We are are hitting this too +Very much appreciatedI have applied it slightly modified in Rev +Suitable workarounds provided so lowering the priority +Herres my shot at the. +Cant manage to reproduce it. +maven find +But i wouldnt mind some feedback on the changes before putting more work into fixing the failing plugin will now support glob style matching using the functionality of the plugin has slightly changed as it can now find multiple files and it will sum up the total size +Maybe instead of a new imprint goal there could just be an option to the existing deploy goal which would create the list instead of uploading the artifacts +What is the purpose of this flag? and why is not effecting zone wide primary storage? i am able to take snapshot of a volume created on zone wide primary storage with this flag set to false +does not maintain a buffer +runs into a +Either we do this for or we never do it + can you please create a quickstart application that reproduces the problem and describe the steps to reproduce it +I can not make the final decision +why do you point out +Our implementation of the test expects to throw an Exception and hence only those methods are considered by Jacoco which give normal return +Im glad it works in real environments. +Ive also changed the temporary file handling a little and do not seem to close or delete the temporary file which in a long running process may prove to be problematic +Proper fix should be not to catch the at all +Closing this since has been released +It was part of the errant poorly documented commit +It would be nice if init also works with nested fields +Not the installer but the shortcuts created by the installer expect the JAVAHOME environment variable to be set in System Properties on the machine +This may be the same problem but its difficult to tell from the issue description +v also added a bit of info to other log messages +Maven on OS X just mvn test true on a project created from the quickstart archetype +Whats the high level on the perf numbers? Does more help? How much? Thanks +Done at trunk r r and r +yes +Unless the output is filtered itis easy for a malicious user to inject code that would get executedby the browser of anyone viewing the pageThus the filtering is there for security purposes not for formatting purposes +Moved the Text entry panels to the top of the wizard page +bq +Maven should not be a mandatory prerequisite for building SDK period +Andre ah great! Then closing wizard is displaying current job correctly +Here is an update patch to apply the same recipe to Hadoop Cassandra LucenePlease dont hesitate to let me know if there is anything I can do to help getting this patch into trunknext release +There its called a plugin adding a jar to the classpath we call it dependency for the plugin +All planned docs to be migrated are now there +settings select from there the JRE to use depending on the configured executable in the +Commit id Oct nd is Holiday in IndiaThanks +Other mitigation strategy start the servicerole before looking to see if the host is also running appears to be the cause of this NN startup was pulled ahead of JT startup to ensure the dirs were valid +Hey Bobby can you elaborate on how this issue is for security reasons? Thanks a lot +replacing with same patch from up level v patch but created from root to fix issue in tag in apache license header in new times a charm +NET as I generate both +The actual message orderIs not what the test is expectingFull log from this run attached +seems this bug is fixed +Need to check and update archetype module had one dependency from API +v moves the stats write after the descriptor is renamed because persist takes care not to store stats for temporary files +add code tags + fixes the javadoc and moves to +The patch only provides a solution for when the is not set and sets it to +Mmmm +If scattered the bandwidth is the of blocks of the file minus other work going on against the same blocks +Im looking into it now +we will not BETA media kit installer +Still certainly worth fixing thoughThe patch looks good to me +All seems to be fine although Ill need to try to remember to get some caching statistics in a few days to make sure we havent lostAs an asside we should probably think about upgrading the other httpd instance to and recombining the two +The default in LDAP Studio is which is infinite so somewhere you have set the time limit +Joerg sure but its a question of conveniencesanity +Patch to compile on OS X have applied these patches to trunk +Actually it misses a vertical scrool bar so the screen looks you have a URL that demonstrates the problem? Using the entity and Firefox I dont see any truncation +Thanks for continuing to review these patches Dan +It is buggy code +Instead its created per request. +Sebastian does the patch solve the problem for you? Is it okay to commit? looks fine +great work +Actually problem was that the antrun plugin which does the xslt transforms to create the wsdls ran after the wsdljava plugin with the ibm jdk +I cant believe weeks has passed that quickly +Fixed up the patch according to the comments given +A patch is attached +Maybe we can report it to eclipse project +these issues are connected +Ignore my last +I have implemented the samplecontainers code for the same +Here are the images I used see attachment +This patch makes ensures that no paths in +Ive committed to the trunk and I have a backport of this but I started to wonder if its a good idea to apply it on x this may cause issues with backport testing if not anything else +Added EXPIRATIONCACHEUNSET expiration cache default in to branch in SVN rev +That is a forked version in a sense that replaces the logj log manager with Log Manager +Probably the problem in specific mappings? Could you please attach generated classes hbm and +I also removed unused import statements +But the resource listener code that detects new groovy projects looks a bit fragile assuming that the critical change it is expecting should only be changing the + +say +Small change +Didnt realized the link to JBDS actually changed +By maintaining the copyright I can issue licenses to other parties as I see fitThe ASF software grant agreement doesnt ask me to assign copyright to ASF but to mearly agree to specific license terms +I should have done that myself when making the other changes +It is just twisting the same idea +It was that had the two? has the endpoint uri code that should have better the new get methods in all our DONE +Hey LanceAny more changes required for this patchThanksm close to committing +it. +Youll need to be a lot more specific than recent xalan versionWhich jars where do I get the archive +Committed as is +I have reviewed Kevins patch and I believe that it safely resolves the problem +Generally we need the option to have a blank first element anyway even if it is not a valid value for the field +I have verified it by running all targets used byThe HDFS issue you have came across is caused by and should be committed shortly +This is for the ARM path right? Nice catch +What do you think about renaming it to +Stripping a and b prefixesand incrementing PLEVELLooks like this is a patch +Please extract the latest CVS source and verify the changes +Editorial pass over all release notes prior to publication of +John if this testing works correctly please apply this to HEAD as well +set up +Prefer opening a new issue with the dedicated issueYes I know I it to CR while it was released in beta +Closed the wrong issue +We would also not expose it to users unless they do run into trouble +ve added an explicitly check there throwing a clear exception message Errors argument declared without preceding model attribute +Charles please take a look at this Frank or Sohil should be able to assist on this +and are cleanup +Closing issue as version has been released on. +HDFS support for something like hard links would make this even faster no? One could to blocks in a tree to checkpoint it +Thanks +This patch fixes the issue +You may want to open a separate issue for better dealing with that dependency +A lot more of the other clustering unit tests are working tooStill a work in progress but representative points are being calculated for Canopy Kmeans and now +I am unfortunately unable to test this at the moment and wont be able to do so until early next weekIf this is worth having let me know otherwise Ill revert this and delete the second patch +This does not fully address the issue of this patch so perhaps someone should start a new bugs for further aspects. +Any chance you can include the wsdl? Theres a good chance they arent supposed to be thereIn anycase how is this a blocker? In no place in the jaxws frontend should the be used at allWe just delegate to JAXB for all the JAXB generated types +Can you post a repro? Ill take a look at it. +David did you manage to reproduce this one anymore? +I ran test with and all test passed +Alternatively the traditional syntax will continue to work in which case the name will be autogenerated via MD of the parameters +Create the demanded directory and try againDo note that as of we are not supporting C extensions. +However Im wondering if AMQP is going to have a problem with that and also with modifying the logo +Have you tried running without a security manager to see if that makes a difference +IT added +The runtime field always seems to contain the default Drools runtime and this one if used if the checkbox is checked +I see the link pointing to now. +I hope that would work with you +Hi Renato +Account created for jmAFAIUI the binaries should work +Attaching a patch to fix the issueCheersAnjana +Im not sure yet how to handle the possibile extra dependencies on the client when using an external transport layer but as a default there should be no extra jars +It looks like someone got the and checkin comments wrongOops it was my mistake +From my point of view major compaction should include the flush process +Any peculiarities in your custom +I wanted to this before movingforward with the project template +Weve a duplicate here Quoting it so in case you want more logs and discussion around this bug I faced the same today searched and found two +No tests appear to use those methods +There is not enough info int he report +Need to monitor the CI runs +Charlie did compiler part of this yesterday +Sounds good +Adding in the ability to choose a jdk can be added but maybe a bit later +Not too bad +We dont support the use of the RI +Currently the cache name is embedded within the commands parameters which means that it is marshalled by if the cache name is amongst the parameters +Can we refactor the code to make this much simpler obvious and make it more likely to be correctly maintainedYes things are messy +Ill do that and commit all this stuff tonight +This is different to the other ESB components that use SQL scripts +there are even more problems to solve for example when railo and the cache is not on the same server how the cache can trigger the eventnot only a master can execute on every context that participate the session can when there is a master and the master is down a other context has to jump in the gap +It was never a supported syntax and was accidentally allowed in earlier x releases +rename to and +Stores sstable accesses per read in an per and exposes it via JMX +third changing node type definitions used to be very troublesome in the past + +Permissions are Add Projects Edit Projects Delete Projects Build Projects Show Projects Add Build Definitions Edit Build Definitions Delete Build Definitions Add Notifiers Edit Notifiers Delete Notifiers Manage Continuum Configuration Manage Schedules Manage UsersGroups +Committed latest improvements of Udai as they are +trunk First spike of the Now beforeafter process and added breakpoints for exchange Added single step is now a basic API for debuggerWe can improve it later when we get more hands on experience using the API with a rd party Camel tooling +fixed in the latest snapshot with the move to jettison +When counts went negative available messages were not dispatched hence consumers can hangNow negative count is rectified on the next dispatch +bq +It looks good to me +Anyway I think I will handle them when I will put Ruperts work in +Good suggestion updated patch +In fact the proposed patch will only have any effect when f happens to have more elements than f as far as I can make out a fix perhaps but it seems to me that it is dangerous to rely on this code at all +Committed at subversion revision +Hi Tomare there any configuration changes which should added to test case provided in +The links are available at the right places but the images dont showWe introduced these changes for the calendar to show and if we remove them the same images is used for both inside lookups and calendars +It will reveal the how put doctype in jspx using faceletsWhat do you mean by reveal the problemWill adding doctype declaration solve this issue +This looks to me like a reasonable way to proceed +Patch adds in an object keyword +Ive modifiedPlease review it +Its probably not a bug then +Filed for the failure +I like your plan +Shitty is its older brotherjason +So changes will be present in EAP to be present in EAP they should be commited to JBPAPP branch too +The only provide a way to get and set commandService +Im just looking for a solution because I have a demo to give in a finite amount of time +Attaching +Use inversetrue and use the user forum +should we just silently and slowly phase it out or get rid of it with a big bang +Ping +committed revision is rthanks Mridul for great work. +If you wanted to take a stab at it submit a patch and test it +Going to Peter for implementation assignment +Serialization of on is not implemented yet but the rest has been committed in revision +Thanks for contributing. +Filled Sun JDK bug +Therefore the exception is because of two balancers run at the same time on one namenode +Marking as Cannot reproduce. +Thanks Zhijie +A patch is attached +Actually the image is found the is the image that is sent +It seems like something is using the maven repository style +Patch that adds the functionality as well as making the plan method visible on +Will post a comment shortly showing that once Jason advises what the name of the release will be +patch looks good +testcase still ned the manual change in the reproduce the enabling is picked up imminently the disabling not because the the session is only tested when it get initialized. +Have you tracked down how the faulty jar was generated as faulty in the first place? If we have a problem with the HTTP Server build that should be fixed as well otherwise this may come back on the next release +This patch uses the pluginRoot javascript variable to create the correct URL regardless of the current URL +defer to +Manik thats just the default configuration provided by Hibernate Search which is using but this report is indeed referred toAnd I recall from the code that theres nothing in place to limit the number of open file handles I dont remember if there is some attempt to keep the number of files limited this seems to me the way to go having regular compaction of fragments into bigger files +If you hit this again can you check the logs and see if that users password was recently changed +A more complete patch with a st demonstration of use in Party Profile +Good grief +Drew its assigned to you so Ill leave it for you to testresolve +Therefore Im guessing that also in CLI use case compiler completely replaces the regular java compiler +Forgot to mention I ran derbyall with on Windows XP with Sun and there were no new failures +A patch for is submitted please check it thanks +Fixed for upcoming release. +The third field does not define a range of lines foo Input size for the large fields can be set to instead of And maybe more to come soon +defer to +Patch looks good to me +patch on behalf of Harsha for I have reviewed the patch and it is working fineThanks to Ratnesh Upadhyay for his supportThanks Regards Parimal the Display problem of the clients and Providers in the ScreenThanks Ratnesh for the Support +This will need to be rebased to current code though +Wont get access to the machines sooner +Fixed in r +Thanks for pointing it out +Updated patch file +Got v to run apply on Windows no changes necessary except for renaming the script to +Depends on how you read things it must be able to handle null for backward compatibility but contrib analyzers have no explicit back compat requirement so not supporting such a thing while inconvenient is not necessarily wrong +Will be in RC thanks for the thanks +If it shouldnt be there then thats also a bug in +Well reopen if theres a request for this target. +No further comments on this oneAgain note the patch offers a driven alternative to using the +Sorry Im using the latest snapshot build +Bump to +Works as expected +committed in rev +getHostname javadoc states that it returns the hostname part of the address +Heres a patch that simplifies. +Hi ClausI have adapted the documentation file according to Colms changesRegardsI forgot to remove the provider parameter +what configuration are you using? any custom codemodifications involvedare you able to reproduce it with an jackrabbit release? could you perhaps provide a simple test case +It seems OPREADMETADATA operation on DN is not used by any clients +Support for assembling a list of war files is complete +Dont change to native without understanding what was good for and then report an error. +Isabel you touched recently do you have thoughts +Patch v is on the right trackThere were some spelling mistakesThere was unnecessary formatting in Its a pity that mapred doesnt use +Nice one Shrijeet +This patch addresses this issue +Hmm thats odd +Id like to track this against the release to allow community exposure prior to Final +Using the above at the very least we can catch issues related to where the shuffle service is not configured +The ids for each counter were used to speed up access within each group +ok +If tools is generating such a URI and AS accepted this then you were just luckyPassing in a URI string containing a space doesnt work either on AS or on AS +Correct +Ah yes for the barrier it makes sense I misunderstood as well +This needs to be a release to be compatible with jboss messaging +The patch is appliedThanks! +An even simpler and smaller shapefile Essentially a polygon in outline that crosses the am experiencing this issue on version as well +Ive added the fix to the main Android repo. +Bulk close Solr issues +A log cleaner scheduler wakes up and decides to remove the only log segment +Done? Or more still to come +Unable to reproduce +Trying to figure out an elegant solution +If this is a correct assessment simply tell Oracle which column you want as the primary key and the will be updated and the domain class the next time you do a reverse engineer +Since its wrapped in json and in a string the browser should not be executing it +This is not a bug +than share it with the rest of us and we will see where we are +this is definitely a regression on we cant use code from the is definitely a regression I think I added tests time ago for windows probably they were incorrectly if you go back in the revisions you could get my test it was pretty complete for windows regression of this is a broke the windows supply patch from the working revisions +r on trunk +I understand someone in the community uploaded the files to the repository so would seem that this task is done +Hi Mike are you able to patch this? Thank you for patch for this issue +This is a bug in the version of the HTTP client +Could you please keep this issue open for a while longer though? There are still a few tweaks I would like to do to the +I dont disagree with you +Also occurs with RC +Ill make a note in +Wouldnt be better to call it readme? is something you should be actually able to run isnt it +No new test code changes to documentation only +branches and trunk +Pinned alloc support was required for older prototype of class unloading support +in you create a new quorum verifier and update it with a the current zxid +For convience reasons I created the process definition and unit test in the examples project coming with jBPM +Ported to branch at subversion revision . +Hithe mirror looks good now it is added to the listThanks regardsHenk Penning +added to fixVersion per and above comments. +my old Fedora I mentioned in Environment was +Dont want to go back there +Patch with minor refactor of handling the deprecated state parameter +So lets pretend that that was the plan all along +Does not extend the concept of underreview etc +Updated to CPCR doc changes now made +Bulk close for Solr +This is not a blocker but will do done in time anyway +Sorry the image is cropped +Should be the last feedback Stas The message Your browser is out of date and does not support this widget +If Derby provides extensions to JDBC then applications using them will be locked into Derby +This seems like a brainfart on my end. +Fixed on trunk Patch provided by Andy in revision part of this fix failures regarding cluster timer task were also fixed +Unfortunately I dont have a server to test this nor am I all that familiar with +Please let me know if it looks good now +This patch has the automated system test for local file permission check +Can you please enumerate the problemsdisagreements between these two plugins so we can address them? Currently this is more of a general stylistic complaint than an actionable issue +I made a few changes and committed this +Work on refresh DSLD job now pushed +I think somebody broke the javadoc warnings +fixed +Or to be more precise if you timeout with counters youre screwed them being in batch or not so allowing batch of counters is not worst than allowing counters in the first place so why forbid them? Especially since with CQL you often have to use a BATCH even to insert data in the same Cassandra row and so disallowing BATCH of counters would be a major regression compared to what you can do with thriftIts true though that we need to decide if we allow the BEGIN BATCH for counters or only the BEGIN UNLOGGED BATCH one or both +I didnt look deeply into your specific issue but can be loaded twice according to JSF formumif you have a in your directory it will be found and parsed automatically by the framework +check with Hudson +Thanks for providing the patch +Admins can still change job priorities through the standard command line +This is a blocker for releasing so I would really like to get someones input who was involved in the technical +I think this is great though +I can deal with this if you want Henry +Rescheduled for should be resolved as there is now only one way to initialize an object factory and that is through the dependency injection container. +may change in future +Ok +Looking at the stack trace above you can see clearly the is the one that gets would recommend removing the +Thanks + doesnt compile for me right now I get a ton of Avro related errors +I was commenting specifically on +This wont make +This error is difficult to quantify does it only occur with certain Spring versions with certain configurations and application JAR dependencies does it relate the application server etcregards Malcolm appreciate the though around this issue but risk to production applications using various Spring version is too high for the reward of removing this class +we should backport this to do you intend this to go into? I think perhaps it shouldIf it is going into then should also go in since it fixes +but both are not clients and receive stanzas targeted to them uncensored by the server +Seems to be duplicate of already fixed issue +We want to leave deploy name also attach since it is very please review I can apply +Hi jaikiran Im sorry that those log files are not available on my side for this week +Ive committed a patch to jthreadsuspend so that this function would always suspend the target thread +Think this is good to commit +I now see the problem You dont have the metatype service installed which also prevents the from bein registeredIf you install the Felix service implementation and refresh the web console you should get the configurabilityTaking over this issue to split the dependency on the meta type service from the dependency on the configuraiton admin service +Added section Other usefil atributes with description of status and focus attributes. +Additionally yes it may make sense to have a permanent VM setup just for Sqoop contributions +It failed once for me +bq +Patch applied on and trunk +Hi Does this issue need merge +hi AntonioAs I mentioned earlier the problem is that if there are element with the same name already exists in the same name space where creates new elements then there is an error but does not report that properlyHere I agree with you that this is a limitation comes with the Axis implementation and IMHO that is not something we can fix now +Hi RegisI think and are related to this issue so shall I modify the patchand open another JIRA to track ? RegardsRay Chen +The fix involves minor additions to SPI interfaces +We will start building CR next Monday so well need something today or tomorrowThanks Howard just want to see if you will be providing a patch or a new tag? Well need to build this ASAP and also commit the changes into the branch as well start building CR on Monday +No we dont just have to +I am foremost interested to make the only source for projection constantsIf we later need a new interface for the document builder is a different story and can be addressed then +Are you experiencing the same problem with the latest snapshot +Closing. +The last comments from Ramesh and Paul led me to believe it needs to be added back and added to the file deployment as well +Diff of logj file was included by mistake +I meant the code in the patch sry +Feel free to give it a try + for the patch +Any idea when this will be resolved? +Can you tell me how to do it? Richard +No matrix is neededI am working on designing an interface for network topology and an interface for replica placement policy +Add Zhihongs and Jinchaos comment +Moving all issues to the CR +See RFC for more info on what constitutes a valid absolute URI. +of webapps now occurs correctly +Committed a patch that fixes the compilation issues on. +Hi Alexey What I fixed here is about the different prefixes for file on windows and linux +bq +Thanks Branden +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +looks good +This system call works just fine without the hook so Ive got no idea whats wrong +Other than that there is no difference +David Can you run your test case against the lastest nightly build to see if this problem persists? thanksRgds The fix is in cvs please help verify and close the bug if it works for u thanksRgds fixed sorry it took so long. + +Awesome! Finally should be resolved along with this issueIll open new issue for refactoringThanks +I do not think there is much we can do about it +When its finished will you commit that +Verified on EAP ER +Would be good to close this out as resolved and create an enhancement issue documenting the needed improvements to cover this use case so it can get added into the road map +Manual de Referncia do Derby Documenta a implementao da linguagem SQL no Derby +Im taking a look at this one +Any chance to have it fixed in a foreseeable future? Quite surprised it has low priorityBest regards it still open or fixed? + Username stephanenicoll +Will fix them tonight and submit a patch againWill work on the INLN to get all the messages out now +What is the reasonable table lock timeout for snapshot? min +I believe this was fixed in? Can you confirm its the same. +Stuff for Pavels patch with minor modificationsVerified the fix by testing manuallyDidnt add an automated test for the moment. +Can you point to the changes regarding the filter handling? The patch is rather longJust noticed a typo in the enum for request is just Reques +Hi yesJAMES version betaThunderbird versionNow that I clicked Help About it starts downloading an update +Exception semantics are correct configure would fail regardless of which exception is would actually be cured by the patch in if you dont want to apply the entire patch just remove defaultOntrue for these metrics +suggested update the dependency as suggested as the reactor pom already refers to that version +This issue will be fixed when is completed +I agree it would greatly benefit development +Im not sure that overlapping diminishes debuggability +Fixed at revision Thanks Antonio for the patch +I think we should put it in but Ill also take a crack at fixing the failing tests as part of this issue. +Thats strange cause on my machine all tests pass or maybe this testskipped for some reasonNikola +Wasnt that fixed inAnd I had this on my wishlist for a while now +Extended jelly script of the plugin +Committed revision Committed revision +By not setting the Domain to MITLL you are not enabling NTLM auth +Please check that the bug is fixed for you +unscheduled issues to the next release +In my case it seems that the problem is caused by Shindig expecting parameter method to exist in the request path thus appending the parameter fixes the problem. +If you can provide a patch it would be nice +Add testing for side effects to the new test Add missing since tags Fix formatting of if statements Other small improvements +marked as minor issue since and getType on function values are arguably corner case issues +This is confirmed by looking at the existing implementations of +Apparently the previous fix did not fully resolve the issue +Can you verify again after Nicks latest fix of. +OK looks like all is now working +In the case where we cant find any Default JRE installed we display a page where the user is informed that a JRE needs to be installed to run Apache Directory StudioWDYT +We would have to maintain build and distribute this code somewhere else +seems to be fixed +Yes we are definitely getting very close and its looking really good +will commit if tests pass +Sorry about that +They are referenced from so they should eventually collect but you are right they will linger for some timeI imagine the real issue is that if youre loading webapps that cause this softref to happen theyll remain softrefed themselves since Thread references references references references webapps classloaderMaking it a would cause it to get collected more quickly but perhaps too quicklyA bit of info on soft references are garbage collected if they remain unreferenced for some amount of time +Changed filesrdbdistributionbinarydistributionbinarysrcmainassemblyrdbsrctestjavaorgapachetuscanydasrdbtestrdbsrctestresourcessamplessamples samplesdbconfigdeleted because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +First part is for you +Ive just committed this. +Create new Translators on every usage instead of getting a singleton instance +the idea not to store any ide meta data such +AppliedThanks! +I have only looked at the Local provider and there may be additional issues in all of the other types of providers since for the rename it was an issue with the +I did code to evaluate the ideia and it was easy to do the patch but the annotation interface needs to checked if it conforms with your naming convention +I have added the ability to put an annotationPlease now use attachments +I think setting max to would be more of a than anything else +This is an odd oneSpeculative execution is in some sense a pure overhead +Thanks Mickael build seems to have the correct Eclipse platform +Will commit if tests pass +I just got the hang again when running the jdbc suite separately so it seems we dont need to run the full to reproduce it at least +Thats fine for typical formfields but if the item represents a GB upload things dont look so goodSomething along these lines that streams the content from or to a DFOS mightwork but I hate to think of the performance implications of serialisablesessions that contain multiple gigabytes of data. +I remember some time ago I saw the exception but after doing some enhancements the problem dissapear +Thanks Owen. +did you attached a patch to the jira that fixes the problem +I have modified the patch to avoid having a fixed sleep period instead it now retries to start the cluster every sec until it starts +Will checkin when they are done +Maybe we should just provide a default implementation for advance this would save +Thanks IvanPatch applied to jdktools at repo revision rPlease check it was applied as you expected +no docs +But this way the supervised thread is not interrupted when you interrupt the supervising thread +The fix has been verified +Is this really an issue? Should we not close it as not an issue? +Verified in Sun JDK Linux s not possible to install JBT CR update with Sun JDK v +its necessary to add a link to the old Ajaxjsf guide to a new one and tell that this guide is no more updated +Seems to work +Attaching logs and DB seems caused by commit aaaffbededabea +I see a slight improvement with this setting +I will move it further in the release schedule also because I am unsure as to how useful this is to other let me know if you want this functionality in the main release by voting for it +It was released with the fix in +I know Java broker supports a user create call +Thanks CharlesPatch was applied at r please verify +I think option is the best option and is simple enough to get get done for a release +Given that this is a minor change why not just do in this +Attached patch for compiler warnings are because in is deprecated +I just downloaded the latest thrift and I cant even compile thrift with make because it cant find ERRORIn file included from +Can you verify or kill this report please +Leos patch is fixing a lot of bugs butalso adding some new functionality +Perhaps we are not properly closing simple consumer in certain cases +Its just saying that we cant do nice GC stats because our statically allocated stats buffers has overflowed because were doing lots of + +OKI got that +Please let me know if there are is any reason to stop me +We assume that the targeted release for Spring final is November are having our production release on st of November +I have checked out the code and will investigate myself but Im not familiar with the code so not confident Ill be able to offer up a patch any time soon +Max While youre right that new APIs are a difference the problem is that theyve backported the fixes to stream which means that if they do come out with a or stream the new APIs will still be all honesty If I were ever to rewrite this stuff entirely I could probably have only server type and make everything absolutely dynamic reading the servers version from the installation folder and then calculate everything from classpath container information to startup information to deployment stuff +Committed revision +Hmm DEBUG? INFO +Removed an old ambiguous Sis comment that was not in trunk at revision +from me thank you + for the patch +Albert I see that the axcflagsoption is licensed under the GPL +Doron sounds like this is ripe for a commit now to take care of both this and +mappers now spread evenly across clusterI should be able to submit sleep jobs maybe job has mappers job has mappers +Closing the bug again since Mazz said its okay if groups dont send the event +There is a Hudson failure likely due to a configuration issue +Attached the patch for trunk and security branchaddressing the review comments +Can this be pushed to next release? We are almost at RC and not sure if this change should go in now +As you can see it is not exactly trivial +SureshWe no longer have rpm scripts in trunk and +This version seems to fix the problem +html page with javascript the reproduces this bug this would help us immenselyThis project has access to a large library of devices that we can test onStep one to solving this issue is we must reproduce it locally +Having LICENSE and NOTICE files in the top level directory would result in conflicts for shared dependent projects +Over time people have been more willing to add features rather than refactor +What exactly needs to be done still? We have a migration guide for both s and ww +Now that the project has split should this include generating and promoting updated jars up the dependency line? Thats not always getting done +and are not merged into codeline yet because they depend on merge of into +th +Issue is still there +Working on and features didnt seem that noticeable +Kelven can you check on this vmware working with latest build +If possible lets try to give this commit its own performance run to see what the impact is. +This is a very important feature IMHO as it can make Forrest be able to generate spec documents like the wc ones. + fix to please with resolving to both versions to our QA. +Thanks for helping me sort this out +For example we might want to support a TRUNCATE flag in the future +I believe its called and was only able to do a few benchmarking tasks +will not be necessary to build release tarballs since well include the configure script. +Also it removes real element style from style attribute after selection is chaged as it says here sounds fair lets have it only for youre right about selection problem with using one div for selection four divs would be na option but lets wait for we have an access to flasher sources? Do yo have a link? Id look at it +Please see the Reference Guide section titled system procedure +I want to look at the and why it is needed in detail + +Some now recreate the entry by default or we use the one provided in the config +Max Please comment +I thought we had a similar issue open that would use standard udfs inside the vectorized ones +looks pretty nice and has an apache license +ok for managed beans renderers etc +The redudancy will be remove once we add the client side validation on form submission +Thanks Dave +This would be good when the new jts wkb reader is available +bulk defer of unresolved bugs to +Tom isnt this the same as isnt this a duplicate of +You can download stuff in a few days with the complete mappings and this test +Setting the number of reducers to a larger value helps me to resolve the issue for now. +Fixed in svn revision thanks has been checked in and seems working well +imgsamplecoverages know that is lame but its the way the restlet dispatcher matches uris +JPA uses exception sprin data chose null which is fine in more cases + to community need to verify +Today we will come up with a patch for this +HiI have reproduced this and after debugging I can see why its putting in the soapenvelope if soap is set to true in the httpprovider it will attempt to write out the soap envelope and body and then will wrrite out the contents its just in this case the contents already have a soapenv and soapbody so we need a way to check the contents or use another work on trying to get a fix in to remove the +Include + doesnt really address this one way or the other though this is hardlyunique +When I run mvn clean test I see the Tests run +If you have an svn command line client you should be able to just do svn diff and then post that +Regression tests run cleanly on the previous version now that the initial patch for has been committedThis new rev adds some additional test cases to verify the following can be used as input parameters to procedures including as? parameters +ok Im resetting the password and will set it thought commons decided a long time ago to move to the groupid +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Hello Dimitarthak you very much for your patch that was very helpful indeedAlthough our solution worked in our environment your solution is a lot smarter and fulfils the standardGood workRegardsRainer +Status updateDevelopment of the major part of the feature is complete UI intergration and Junit tests are in pendingCurrently developer testing is in progressExpecting to propose the patch in review board soon +If is fixed Im on my way again using SD JPA + now you further elaborate the use case? In particular what is the use case for? If you need to directly write and read instances from streams then you can use and directly no? Thanks +This has been fixed by +It should fail with Job Submission FailureModified the exit code to which is the closest we have to job submission failure +bulk move of all unresolved issues from to +Not important right now push to t remember the benefit to doing this +I have tried with S but will try with soon I have discovered strange issue with my changes stress cant read portion of the inserted keys after one of the inserts working on that now +Thanks for the patch +Patch file to start publickey thanks +Martin for change is looks good to me +Hi GailI am not sure that you have undestoud my requestIn this case the string consists only numbers and representing a integer number of +Im mostly done with the code and busy with QA and junit modifications +The rest is covered by separate issue +Still getting around to it +Committed to trunk +So how that logout that you provided will help with that case? + +I tried to do a soft upgrade run on trunk but that doesnt seem to be possible without the upgrade logic in yet +Before the patch the delete tests ran the ACL query first and if it fails then it returns a UNAUTHORIZEDACCESS error +Add a brief note about how to roll back to the old log rolling policy in the file +Here is a new patch that comes with specs and also includes the includeexclude support +The way I did find of achieving this was to add a bloom set into my single BloomLucene instance used for all fields +Hi SaschaThe patch is outdated could you please update itForget it I was able to merge it +Lets target this for trunk +This of course can close a filechannel which can cause a during a subsequent log append +Regards can confirm that using the interceptor that Freeman worked out resolves the exception +Similarly in section there is thisQuoteConcurrent requests to are always serialized by Seam +Thanks for confirming Christoph +Commit the patch to both trunk and branch. +Also however a new server created in the new stream and then used inside an old workspace should automatically have the defaults to clog up the server object again +I havent checked the other Errors but there may be similar pitfalls around +Lets see what breaksD +Sorry for confusion i did test it with wrong JBT installation +At the time we solve we can very easily rollback this additional fix +LincolnI think its ok just as long as its not top right +Added some unit tests and did some testing by placing some bundles in subfolders in the OBR and trying to load those artifacts into ACE +Please set the affect will be hard to fix it for any versions of JCR since it will require an improvement in the kernel which could cause regressions +Heres a patch that reverts to the previous behavior of using the set provided +my opinion this issue should not be resolved but closed + BOM released with DS merged +last commit in above post by Brock fixed missing changes +Ah +One more thing it would be sweet if the above were not hardcoded but instead was a set of steps described elsewhere and malleable or even better if we could describe the steps to run on top of some generic operations framework as per FATE but that would be a bunch more workHow many regions are we talking of mergingdeleting at any one time? I think above should work for a big table as long was we did stuff in parallel closes and file moving +To indicate that the configuration file tests are done separately from the view definition and taglib tests +There seems to be really no good way to implement this +I dont think this happens anymore though we have introduced notion of major compactions which might do same thing +I think that the other warning is one that already existed but some how is showing up now from because all of the other warnings are in code that was not touched by this patch +Patch checked inOleg +at least I hope I have no way to reproduce them to start fixed on my buildbot local Maven and Eclipse +so as not to break the QA bot patch set looks good +well its not wo modifying hftp but a client can verify the content length if its there and proceed as it does now if its not presentwe could also have a config option that say enforce content length which would cause missing content length throwin this way if both the client and server are on the latest hftp this works else it will work as beforeoffhand im not sure how to do this without either changing hftp or wrapping it in some other protocol that does length checking +rebased to the lastest initial comments This change disallows moving when RF N +I realize it will take time to go through these +Fixed indentation Fixed incorrect comments Rebase with trunkI only print warning message since I use the same code block that we use to print warning message when we get txn +Use Instance Per Transaction container +Looks good to meAndrew to answer your original question yes its possible for client to fail +New option added Defaults to true Commit after index +could you attach a svn diff with the whole patch pleaseThanks! is a patch for the http provider + fix version to next CP upstream issue and will need to be handled there first +There is no encoding inherent to unicode objects +The code should assume things work and catch exceptions for the exceptional cases +it is a configure option If the OS is linux we can use uselinuxnativeaio option to replace the aio thread mode +Committed at subversion revision +If a bean id is used just to reference from one object to another in xml config there is no need for Java constant surely +Merged two forms in one which does the same thing except a small change and the purpose of both the same +I also compared to the PDF and everything looked fine +I will do my best to finish my changes tomorrow and pass it onto you for review +I will keep in mind + +and Environment annotations can probably be used to customize things like the address etc though they would be used in a subclass I guess +It might have to wait until Monday until I can have a deep look into the problem +Earlier it silently ignored adding files to the patches that were in a new folder added via svn addThe new version still does not allow it +A one should get removed at some point once the sh one has parity +Depends whats you definition of SPI +Len can we close this issue then +Yes I think Heiko is happy enough +yes I hope i configured it righton the forth step when i hit finish the dialog will not close i see the project created on the background but then when i hit cancel it is gone +Thank you for your comments +its ready for use +Implemented as specified +We should be moving off to newer serialization techniques like Avro. +I see the same problems laso tested with Firefox Beta and Safari +Hithis issue is not open about MonthI provided a complete fix on of July this yearWhy is there no progress? You only have to execute the patch command +I will write commentsalgrithm +This means need to be replaced by a scan which will have a little performance loss +Bec can you remove this reference +The only thing we can do is to try to port further classes from over to provided that it is possible and that there is no better equivalent in the Hibernate m converting a project from Hibernate to +See +So every time we reassign uid we need to make sure operator has the knowledge how does that uid get generated and convey it to +Thanks Rodger for this patch! for branch and for trunk +bq +Hmm so is it merely a documentation issue +We can discuss things there and it will be easier to trackrecord +AttilaAre you sure that you took the right log entry? This one talks about rockmongo +Thanks for checking +This appears to not be a problem with +I need a quickstart to be able to debug the problem and try to fix it +There is already a count of storefiles and of memory help up in store memcaches but I was thinking that a count of stores themselves would also be useful? I was thinking of the streamy case where there are a couple of tables each with different counts of stores +My patch this to Galder to make sure this is properly thought about documented maybe blogged about and folded into the release +moving out +I managed to trim this to kb by removing gradients +Ya I thought about disabling the because it is not really needed +Is this ready to be committed +Moving to wagon for review +Ill try to get a fix in by tomorrow +What I mean is pig script in trunk is not working correctly even before patch +If we try and cater for things such as clickclick we would have this issue because clickclick uses quite a number of these templatesWith lazy loading this problem goes away because we will only load resources that are requested by the browser +The one supplies the credentials to the so that it can authenticate with the source with simple useridpassword +Is there a scedule for this feature? This would be really helpful +I will apply this patch to HEAD and branch if there are no objections +It is different from failing a task +The bug reporter used a very old version of and did not add any proof that this was still a problem with the latest release. +W options is partly undocumented and kind of +fixed +xc is currently broken due to change in +This can be committed as soon as karaf switches to aries blueprint +issue evolved +here is a new patch for the release with release notes to warn about the incompatible at it here is a patch for the outdated licenses headers found by brett +This patch is not valid +Ah yes +honestly i really doubt that is a reasonable thing to do +Okie +Some comments youre breaking api compatability with previous version changing constructor arguments youve added a new feature with no documentation and please dont use tabsIm not sure either if adding custom strings to resolve different things is the solution next time will we add again a if to check if the property starts with xxx and do some other thing +usrliblibstdc +Changed to task from bug since this is a bug in JBR not JBM +Doh! Server keystore for testcase attached +It looks like it was just copied from kitchensink +It will take a really long time to cycle through all ids +Also theres a comment in the test saying repeat the call which looks incomplete or redundant +PaulRado you guys should add to your automates testsuite to avoid future surprises +Lets consider the new at mojo +Stefano please consider adding two parameters to the jdkim method you have been referring to instead of adding respective setters to the interfaceOleg +It is not mine +I took the example of the jar which is the jar bundled with all our standalone distributionsIt explains why it bundles also because we want a faade for SLFJ to be able to log eventsNow if youre to embed into your own code with your own requirements of course isnt mandatory and youre safe to decide whatever SLFJ faade you preferOn the Jetty part is it mandatory because has a feature to launch an embedded HTTP server +This documentation will not be ready for moving to +I guess they could be used in Oracle as well yes +Trivial patch +Enabled for MODULES space you need to approve current content. +How do I kill the process properlykill on Unix or process explorer on windows +Thanks. +There seems to be some inconsistency across the board with all the methods here +Patch to provide a default lookup of libDir in the conf if the service doesnt have a reponame at loading +This is not too far out there for production usage with a high number of producers +Due to switch to Maven builds this works now for Jackson +Binary distro for Hadoop with documentation and examples +are kinda unpleasant +It is much more visible. +Please reopen if you want to finish this someday. +Attachment has been added with description PNG image to reproduce the transition to closed remaining resolved bugs +This is nice! I also had to exclude the abstract test cases in the Maven build file so that doesnt try to run them directlyAvailable in the nightly build +Thanks you very much for the patch +Ive put the text recommended by Jacopo on the front page +We should not let destabilize +Updated Test case for +I think our modification of extended attributes could also have this issue +Updated the DD and uploaded a new screen shot. +Upon looking at the central repository and the uploading guidelines I do have a question about the groupId +This is because I think they are basic factors to evaluate the behavior and performance of some message delivery system +Ah yes good thinking +By indexing again with Solr i got term vectors in my responses +with default pkglocalstatedir and with default patch p i Please review and let me know if there are any issuescheers an updated replacement patch for +fix broke testthis patch tests all slices but keeps old intersection logic all components of the composite need to this fixNevermind you put in a comment saying it probably was +Thanks R. +Im giving another try to this one tomorrow +Thanks Dib! +This is a trivial change to the logger statement thats still a problem on trunks so am going ahead and committing it +Heres a simple test program that opens a server socket with a backlog of then starts spawning threads that try to connect to it until an exception is thrown +Maybe the interface could be called or another may look older issues for Apache since were no longer actively working on these at the moment. +I bet you wont see any of the internal native bitvectors in the Java class libraries switching to +Bugzilla does not show the correct file name +jarExtracting systemorgapacheariesproxy systemorgapacheariestransaction systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamel systemorgapachecamelkaraf systemorgapachecamelkaraf systemorgapachecommons systemorgapachecxf systemorgapachecxfkaraf systemorgapachecxfkaraf systemorgapachefelix systemorgapachefelix +The interesting use case is tasks that spill since these make no progress while the sortspill is happening +It change the bytecode manipulator +Lets defer worrying about this for a followup ticket +Additions for Environment and Java +echo httplocalhostservicesechoMTOM usrlocalaxiscbinsamples +reworked to sync up with trunks latest line ve just this after all the tests +Hi AndymoduleCoretypeServer list shows prepared transactionsI waited more than minutes I did kill restart cycles at the end I restarted server with MDB too +Agree +We currently really only want this for authentication on web interfaces rightThe filter mapping will solve the problem +once for each webappThis addition would eliminate that need +looks good Thanks Henry! +This should fix the problem with configure not being able to find the script temlates +I havent implemented the server side at the moment +Applied thanks Dan +Hi Wouter which binding are you using? JAXB +Patched clover Hudson plugin works as expected +Please review andOleg +committed to trunk and branch +It is quite possible that they are related the stacks are very similar +Its been sitting here monthsThere is even a test case attached for you to review which is a modified version of the test case shipping with HibernateThanks +You can see right now tpstats shows all time blocked XXX and completeed YYY but if the node has been up for weeks you dont know if those blocked showed up all at once and when you were in the bad things happen time or if they accumulated over the full weeks +Thanks LucGlad to hear it is working well for you +This issue is related to the filtered collection in the first place +Internally isnt an arraylist on my you double check +The problem with databaseName looks like a bugAs to the replication properties I think you are right that we havent done anything actively to support replication of databases +Yea Adrian if you have some time to work on it itd be great +Do you mean they fail or abort? We havent been seeing that in our nightly runsThe tests should be run against the older version of Pig the default behavior when not verifypigscript is provided +bulk defer to close of wont fix etc resolved issues. +Attaching a screenshot and assigning back to andre who can suggest improvements +Issue also reproduced with +While Im not a lawyer I am all but certain there is no distinction between distributing a +I cant reproduce this with +Can you give it a tryI used Firefox with Quick Locale Switcher extension under Linux +It is not getting updated marked dirty after remapping +Do we need to apply both the patches attached +I forgot to mention in the Environment box that I see this on both JDK and JDK +If need a particular feature or bugfix to be included in you are welcome to get involved and contribute +All test secenarios passedfiles optionwithout with filenames as numbers Special charecters Multiple files hdfs file Multiple files Combination of hdfs and lfsCombination of hdfs and lfs without negative scenarios like file does not existlibjars normal multiple files with libjars and libjars hdfs libjars path archiveswithout without and multiple archives with and multiple archivesWith numeric tag with special charecters negative scenario of archive file +Close if not a bug +bulk defer of unresolved bugs to +exmaplesdocs should help +Committed to trunk. +I tried to shutdown the broker and restart it but the Temporary queue was still valid +you know forever I have been trying to avoid juggling potentially s to s of watchers per node like that +Could not test as well but a customer reported us that he successfully added NTLM support by updating to which is actually done when upgrading to latest version of common emailsSo we can consider this is fixed. +Jim Peterson We should be able to make this work without enabling MTOM +Patch v compilesAlso addressed two of my earlier comments tonight +Covered by JIRA shouldnt have a target because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Fixed the typos and the return code bugs +Just to clarify the above statement we are requesting the inclusion of in to EAP. +Thanks Neil +But ideally as most of us pointed out it would be nice to have inout channels on the it as wont fix based on the discussions in the PR where we all seem favor a solution at the gateway level by possibly adding inputoutput channels to the gateway +dodeploy after copying but then it doesnt remove it when doing more copyingHere is what I did Opened a file explorer to monitor what is in deployments + +The annotation declared point cuts have the same issue and will therefore need a similar solution +Alexey P thanks for applying this patch +There are known JDK and zookeeper issues +This feature has been implemented inPete if its still doesnt work then please reopen it as a bug with step to reproduce the issue +I am going to close this as Wont Fix but you can reopen if you really think we need to code around this +Id like to get some clarifications You stated above tat the Hibernate Core doesnt support such a contract +Because this is such a specialized taglib maybe it would be a good candidate for the sourceforge Struts project +Committed for in rev +Then user is free to use that metadata in any other view transformation a their usecase dictates +Also do the existing tests already verify that the correct replication factor and block size are returned and that a replication and block size are honored +What I was originally suggesting was that perhaps we could just remove them as we encounter them via bug fixes etc but they might never get removed doing it that way +but it seemed nobody wasreally interested +They should be fixed +Hi Strong Liu and othersI am also in need of this fix +As I said the workaround is removing the field +Going off what I was seeing when I had Hibernate in DEBUG mode lead me to believe that this was more a request for improvement +Am gonna mark this one as fixed for now if anyone has any good ideas for how to improve this we can reopen it +So your code will work for a time +In my situation an xml file is embedded in another xml file as CDATA section when the Xalan serializer is used to write the parent xml file the rn in the CDATA section is expanded to rrn then the CDATA section is extracted and parsed by Xerces as you said the rrn become nn +This is a deck of slides we prepared for meetup in san the modularization discussion on hasnt really had any conclusions yet we have the options to a make the patch jre compatible or b commit it as a contrib module and move to core with +Reopening for handling in +If no seeds are upgraded there will be no harm the replace will simply fail. +FTR this is happening because in a Java vm without a javaagent cannot intercept field accesses so fields that are not eagerly loaded are not available +Now if there are no available apache releases for a given version your fix seems like an excellenthelp to those users stuck at that release +But we still want its functionality we still want to check for doubling up insanity +Sorry i cant seem to reproduce this issue +Just attach your monkey path and Ill take a look at it +Need to upload jar compatible with hadoop +So may be in far future +Thanks Nikolay PaulexPatch applied to LUNI tests and NIO impl and tests at repo revision rPlease check that this fully resolves your problem +This is one of the longest and most watched issues Ive seen. +I added some docs but the template part is not complete +I think this patch can be improved Ill update it after I get back from vacation +After looking this over a bit Im not sure it will be easy or valuable enough to make the ir package removableI was able to make a simple change to let boot without the IR package present by lazily allocating +I am not sure this bug is fixedThis fix is perhaps not yet submitted to CVS since the latest night build from Aprilstill produces the same bad result +It would render the entire concept of the temporary deploy folder irrelevant thus confusing users +For example if the NN has started but no edits have been made yet and the NN wants to start a checkpoint it will ask the NN to roll the edits log +I suggest to rename the propertiesto and otherwise from mecheers Roland +Committed this to trunk andStack TedIs it ok? I verified the patch anyway +This test case exercises the get code path using gzip created this as a new test case as I couldnt use the base test case because HTTPD Plexus component cant be configured to setup Jetty to perform responses +Workaround for is to push in the Managedbean annotation from the ITD to the java class. +Will if it is encountered again. +could you review this pleaseThanks look good to me. +bq +I believe we should put this only on +The feature is similar to and depend on Tools and SOA Tooling and have to be built separately +As long as users are aware of these constraints there are plenty of use cases demonstrated by the interest on the mailing list +Committed +Tomorrow I will start with more real new version include a more flexible API to create In line process +rev trunk rev This change still does not allow trying to set properties but tells you about all the problems at once rather than just the first one. +I like idea of an Increment class could it subclass Put +There have to be changes made to enable users application to run. +Proposed your patch introduces about failures of various SWING tests a kind of some deadlock issue +Replaced old patch with a new one due to changes in package names of patched source file +In particular it is not versioned or released +Patch attached +Resolving this issue since the original problem of having an OOME appears to be fixed +Is this problem only happening on the branch or is it simply already fixed +I would prefer not changing the language +I gotta disagree with the to Exception change but taking advantage of exception chaining is good +One nit can you also define the default class as a constant like DEFAULTHBASECLUSTERMANAGERCLASS and do the Stacks comment +Your proposal leaves your package versions at but hopes that they get handled as being the same as the bundle version + Cargo container is not started or stopped when skipTests is the changes into the master branch and into the branch +bq +did you change the server host name in server editor to beits in upper left corner +Im sorry my fault +Looks good +All issues resolved in released versions +But if we should then lets open siporate issue where we will disscus all posible cases and how we should handle them +This simply removes the bugI would like to commit soon +Adding a primary key required a special migration script for Oracle +Verified JBDS Betad +Ill grab another copyof it as I cant seem to find the one I had back in +Now because webapps resource is not defined + closing all issues that have been in resolved state for months or more without any feedback or update +There may be a bit of memory growth +new ZIP exists and new issue to automate this +In particular it should test that this works stand alone that multiple work together and that it works with other accumulators +This appears to be related to this old jBPM the issue as a Known jBPM issue in the j +Logged In YES useridI am experiencing the same problem +Running rake in such a weird envirenment like from within on Windows wont probably be significant enough to change Rake +There is the related issue still but it is not required +The when the correct creds are supplied was caused by +I like it +Again as Ryan mentioned it would bebest to have them properly published somewhere but in lieu of thathappening this is not a bad way to gothanksste +iOS has been updated +Reporters of these issues please verify the fixes and help close these still exists +Im not sure that changing the s method will work with a slave but adding another method for NRTreplication masters that take a writer might work +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +So +problem with throwing an exception is that it is hard to tell if we need one +Method is has an if statement at the end which you can skip +I double checked in how they use the timeouts +Im running the Groovy plugin on top of IDE and the startup is agonizingly slow at times +Bulk close resolved issues +Patche seems good we also need a test case to show that we now return an error if the passive ports are used +Shouldnt we try to make this idiom work well with HDFSI am not sure why this would not work now +Unassigned issues any interest in bring this patch forward to trunk +is the PDF file Im printing is whats actually misplacement of the last char in the line occured because the last char was printed individually although the contained the complete line +This patch let each replica return its length on disk in response to a block recovery init request +Instinctively it seems to me they should +Making it critical as it is blocking QE +line and the +Please let us know as soon as it is ready to go +With the default security token decoder this token is perfectly valid +Please see the attached files +I changed the axis scripts in bin which use binsh to use binksh the korn shell and this problem went away +This issue can be closed down. +Please ignore attachment I made a mistake in version number +Suggested configuration changes for adding netscaler to the automation environment +No current plans to work on this +On the other hand it worryies me that the problem is reproducible for me but not for youOn first opencleanbuildtest I get test and test hangs this consistency is good +Under what circumstances if any ought the M surefire report plugin be invoked by the m pluginIf one of you would comment I would attempt to wire this up +I agree that if we do remove the dependency that can come out of the hierarchy and only things like are needed in the factory bean hierarchy +In my interpretation this issue questions the necessity of implementing a protocol on top of hedwig as you suggest Ben to implement this functionality +I shall add the test and address the potential deadlock that you pointed out +As I understand all valid conversions for DECIMAL type specified in the JDBC spec have to be covered +Thanks +How does a copyField necessitate storing the fieldsYonik suggested his approach helps with stored values for the metadatabq + +Added integration tests +patch was applied to CVS +Please report whether it happens on Mobile Safari as well +Now that work is underway with AS all previous community releases are +Brandon says Something is wrong w the upgrade of the system KS not gossip +Our use of Spring dependencies should always pull those from Maven Central +I agree with Jody either fix strConcat or use the concatenate function dont mess with it is used for far more than just labeling and youre certainly going to break other use cases with that change +Can the content returned by such a tool be redistributed in original or modified form in an Apache projectMy guess is that the answer will be it depends and you would need to either contact the owner of the service or find out where that service is licensed +Feel free to reopen when i missed something +I added in the subclass but I am not sure what changes to can be dropped +bq +Moving to +And a delicious patch for Hadoop QA to patches for all branches that remove test and remove unused imports +At the moment it just mentions the line number without any details of the actual perhaps mention in the Roo docs that JDK isnt fully supported yet +Cleans up the line in Scan class comment notes as Expert +This looks like a high risk change +Fixed in. +on +Java does not support Windows neither does you said we will clarify this and link to this issue from the documentation so that at least people that would really want compatibility could get your patches +Please do check on the possibility +We use That would explain getSplits affecting it. +So sorting them is pointlessI think the proper fix here is just to use some heuristic like keeping all of the hosts that are within of the highest total host +to NOT support features from milestonesThere is allready enough breaking changes between gwt versions! +for the patch +Im not sure about what exception should be thrown from +java Handle localThe same result as in the nd case handles on million threadsIt looks like the problem in GC working with thread +I have to remove parent instanceof check Im using own repeaters that dont extend you dont extend in which way does this affect you at all? +Everything seems to work fine +Ran tests for the current Implementation of the memory sensitive and compared them to andAdding to timer for addsSize for map Reading from LRUtimer for readsDeleting from LRUtimer for deletesAdding to timer for addsSize for map Reading from Hashtimer for readsDeleting from Hashtimer for deletesAdding to timer for addsSize for map Reading from timer for readsDeleting from timer for deletesI think that the number for reads and deletes looks ok but there seems to be something weird going on for putsWill dig deeper to find out what +Anybody want to pick this up and commit it +OK Ill merge todayIt sounds like you have some test tools? Are you able to contribute those for regression testing? +patch fixes the hanging for me tooThe fix for should go into the and branches too +There are they now +Fixed typo in the acquired write is a bug in the code that leads to a memory leak when used in a multithreaded context in startingrelease identity hashes are store in the threadHash map based on the current thread key +I deleted all contents of +I am working on resolving it. +backported fix from branch to the and branches +Will make changes to the runtime in a separate jira +Manually verified by running mapred command without arguments to see the usage statement and by running mapred mradmin mapred tasktracker and mapred jobtracker to verify the pointer to yarn is printed +Patch appliedThanks Gareth +Ah I see why you are skeptical in accepting this solution +We should catch log and ignore them +All artifacts for this build are complete +Or at least weed out a security violation from a server side error +New patch +If we assume that every processor like yours follows the extender pattern like yours then maybe this is sufficient since they will just jump to action once a resource of their type is installedHowever I was wondering if there is some generic layer here like what Deployment Admin offers with Resource Processors +looks good +Curator passes the connection string to unchanged +Text in other CDATA elements are lostThanks just checked in a new test case which can be used to reproduce this scenario +Done +Same way as indexOf for now use toString on to get a String and use its API +Ted Thanks for the review on this patch +It needs some cleanup documentation and unit testsI will not be working on refining this further at this time. +I am running without was having some trouble understanding the attachment insert code to try to eliminate the PK contraint violations so myrna suggested as a first step I remove the insert of the attachments all together and run for a while to see if there is any growth in the INBOX +The text covered by a sentence annotation is extracted and passed to the underlying sentence detector trainer +Updated version for calling parse instead of implementing its own code +The nightly tests are coming up +Ill do that although my performance measurements with TCP and UNICAST didnt show a difference to TCP without unit tests fail when xmitofftrue reverted to false until we find out why +Nothing happens when a date is picked from the popup calendar +Agreed get a foundation fix in now and get in for the next version I guess +Patch making private fields and methods protected +Perhaps it can be done as a separate JIRA though +I spent a little time looking for cases which this patch breaks and didnt find any +Test cases found in. +Good Ill merge the branch into trunk now +To be clear I mean in the last paragraph above that XML files have that sophisticated logic but the separate code paths for plist evidently dont. +Small fix for documentation +Im retiring this issue +Hi DamithaIf you can avoid hardcoding these +Ive updated your patch with unit tests +It doesnt inherit annotations from superclass methods either +Fixed with Roller +Of all the times Ive done custom logs Ive never used a filter +adjust component its most probably an issue with the in the simple dav implementation project rather than at revision +It is already fixed on the SVN +Logged In YES useridWhat is the stacktrace? This exception is thrown by thedriver not the right +Also confirmed the app freezes if it remains open and the phone sleeps overnight on any phone android and android has way too many moving parts for me to troubleshoot this bug +These have been checked in. +Updated hivedocs and javadocs for related services. +Hey Avery I hear through the grapevine that you and your colleagues have such a impl on Giraph already wanna donate it to show off how well it works +The fix is in SVN. +Applied +Marking this issue as deviation from standard +This causes the parent process to think that the test run is complete and the remainder of the output is lost +Thanks JasonI put this into and trunk +Trying to figure out if it can be easily changed +i consider the current case of wiquery setting something like that a misuse of the initializer +bq +When specifying localCheckout the in the path seems to be eaten +Its necessary to use the new keyword with Date in order to construct a date from a string +Thanks John the tag seems now ok thus Im marking the issue as resolved +Ive got a patch in progress +Fixed in JBT on with Developer StudioVersion Build id Build date +thanks. +I verified that using the two updated classes allows messages to be sent to activemq +Since there is one example of upperlower +Comments are welcome +It needs links to all books as well as release notes etc +I should say more +I applied the default table styles we developed for the jopr project for consistency +if you want dumps or verbose output from treceval I can do that too but I think its less useful to start with +On the other hand the existing code base seems to handle a lot of possible use cases through variations on this class +The in Adams patch looks like exactly what I need for this so thatll help +this ticket still relevant or its supposed to be abandoned +Theres nothing we can do without a better description of the problem so Im closing this report +If I cant get an example and I cant reproduce it using Im going to have to mark this as Cannot Reproduce +So Im not sure what you mean. + +Thanks Peace Alexandrecommitted Alexandres patch with one differencelibcppsrcwindows contains define VERSION and not this is for the release +ear +just committed this +I did not mean to look up the locale itself that should indeed be used to determine the locale +With your patch the cursor might still be at the end of the grunt window after new lines +If the and subscriptionName should be same for a durable subscription how can the same client subscribe to a second durable namespace? I have tried this and saw that broker sends error message indicating there is a durable subscriber already with the same and subscriptionNameCan you please suggest us how to fix this +I will do this if I get no objections to proposed on the mailing list +I consider it an error to continue using a file after writing to it had thrown an exception +md file and tests it I can get the bad one replaced on +The second issue was probably also related to renamed resources +I think the module needs an additional +Using JMS for spooling is something I really would like to see in james for a long time +I wanted to clarify There are NO files in your +Actually it is not reported in testing although it may not ultimately eliminate the bug +You are right thank you very muchAnd yes an improved XML parsing would be a useful feature. +BTW there seems to be a problem with the referenced related issue +Can you try this again once is released +So it can handle the old format zk data easily +reassigned to Paul so he can resolve it after he merges the fix into EAP +I am going to work with Ilya this week so that Sean will get the information. +I wish Ivy API was actually documented +I guess it depends on where and when you like to findout you screwed something upI personally like to find out something is broken when I go to use it +This is because the broker being shutdown still has alive fetchers to the new leader before it receives and acts on the +Attachment has been added with description A test case to demonstrate the also encountered this problem when using batik SVG library +Setting latest EAP +At least this is what I think but I may be wrong +OopsIf were misusing the dfs api sure we need to fix that but I dont see Hadoop lost data! as an Accumulo problem +test case preventing the since Pedro attached something after it was closed +The merge hasnt been so easy since LTC needs sometimes a Directory while other times a +Retry +Committed patch to trunk with revision and with +A person with more knowledge about hdfs would be able to decide where the tests should be placed +Commited r lets see what the CI says on Java . +Tom or Daniel are you guys still interested in making this happen? I know little about myself so its unlikely ill pursue this +Agreed +Agreed will report and patch. +Glad its working for you now. +That doesnt need to hold up this issue it justreminded me we should move it because its not really +See the docs for your container for more informationJust so that you know we do consider serialization an important feature +okVerify an account can own multiple projects and can belong to multiple projects +Patch that just adds unit test showing the issueWe already have a +moving open issues in X DT to X DT as discussed with Bob +I have ended up grabbing the patch and building the src locally and deploying that instead of the standard distro +I guess some maven person went through and modified the poms possibly automatically +In production thats always the case +The important thing is that we need to be consistent +i am currently filling in all the major permissionrelated in order to allow for parallel working with antonio +So far I have been running this every night and things have gotten stable in the last runs +This jira is about user pinned so this is good to go +Added a handler called to retriew the Http Basic Authentication details from the massage context and to put user name and password as seperate parameters +In rev I committed the above patch as isIn rev I added the modifications agreed in the discussion +update my patchthis patch is created from trunk +Uploading the binary SSL certificate file used for doing junit tests +I have no idea about why the connection is rolled back either +Implements Services and Screens for Entities mentioned in this issue +This result looks like the original result I got when it worked for a while +Hi Bobthis issue is from Jeff Delong +new jira is please explain how to verify this JIRA +I have this fixed just need a chance to get the patch reviewed and pushed. +I will spend some time on Sunday afternoon going over the code drop that you have attached to this issue and then I will report back +I had a discussion with Gavin a long time ago in the forums around something like this +Thanks for all the information about coding changes that makes it easier to understand the diffIdeally this will become comments in the new diff which I can commitYonik mentioned this in email +The cannot be scoped statically on the +Thanks for the fast reply and good explanationClears up a lot of confusion because we recently upgraded to the and where wondering why it suddenly stopped working +I actually found the common container just today after twiddling with the examples +When I download the result as a csv file an extra comma is appended to each end of the lineWould you fix this bug please +Thanks for marking the duplicateI thought I had already created a JIRA for but searching failed to find it so I created the new one. +Isnt it us that actually receives the security exception? Cant we just show a dialog asking for the details +jar to the classpath for JBDS or otherwise +Failed to mention but has extra whitesp too +But i havent tried it in a larger cluster yet +Can stateful clustering ever really scale well in cases other than read? What about maybe also having a mode where the have the smarts to redirect to the master? +Im thinking this is not enough that caches are not being cleared fast enough +A similar error was logged as +You mean remove the catch blocks from Master and main or rethrow the exceptions? I guess those catches were added for the extra logging +Your example is pretty sure cause exception +Not sure what happened to the last attempt I must have got sidetracked +Does this imply that there is something in that needs towill change and that your suggestion is a workaround for nowThanks for your attention and input here +That means this feature has wasted my timeMuch work was done to actually figure out what settings each class is fetching from the property file thats good +We implemented VFS as attached formTestcase to reproduce Us locale and create a file now use lastModifiedtime to get its long value +Thanks +Tested on a few jobs I can confirm that it solves the issue +patch applied at r thanks Jan for contributing I also added your name in the contributors re welcome +No we should be using the normal JAAS configuration for these have already rewritten the test case and am currently running a full test before checking be done soon +v of the patch for which removes the unnecessary comments which were included in the last patch +can you please show the error? This should work +Also added one more test case with async writes +Lets fix the underlying problem +Patch for trunk will commit soon +The tiny patch for orghibernatemapping fixing this +Typically running of these in a loop in parallel +It needs some more investigation I dont think removing the check as you suggest actually fixes it +Thanks Ryan! +I am running unit and ee tests now +Im marking as solved again +Why should user know that files are being merged behind the scenes +H dependency has been added in +personally I like this idea +It seems it isnt and this it a bug I created agree re silent failureMarek is there an issue for failing the build when post commands fail +Its not clear to me now what takes priority the maxEvals in the optimize function or the number I give in the Simple +Thank you. +Nightly build results dont show the warning anymore the problem must have been resolved by a fix for one of the linked issues. +The ngram things are serious so also backportWe get the java version for solr for free +hit return too soon there +is another regression that has been reported +Realized the last patch was missing some parts +Any updates on this? You could just manually trigger the hook from to see if it works +Hi EmmanuelGreat job at nailing the bug! ThanksAny idea when the release of? Would it be possible that i could build a snapshot while waiting for the release if it is not going to release any sooner +Add a parameter to limit the number of partitions that can be inserted into + is a candidate for SOA Designer checked in the fix and should be in the next release of JDBS +Patch applied +The maven eclipse plugin already supports setting different output folders for maintest source folders +Maybe that issue prevented this one from surfacing +Whats the reason for the mismatch? Isnt it dangerous to just ignore it and synchronize? +I think part of the issue is that the default value for is still set for when map and reduce slots were completely separate +As far as I could tell the second patch was a subset of the first? In the future make sure your patches are all seperate or clearly identify which ones are Changed all the s to use a LOG object +Have the latest as of +Closed. +The branch is tests are disabled in +it could be set for TBD +Sounds like you are already controlling for that +So as the clock rolls forwards and more tombstones expire you will usually get a few more rows written that will be cleaned out by the next scrub +It would be so nice if someone could do this and also implement the which is part of Level so we have at least some kind of compression on bitmaps +This is a support question that belongs on the user list. +Committed to +Why cant you do like we recommend and install the driver in the repository +The snippets you provided dont seem broken +Ideally when we have a failure like corrupted hfiles wed transition to QUARANTINED instead of FAILEDOPEN and stay there until an admin fixes the problem +Patch for to make it work with out of the box +ill close once final testing is for to think mazz volunteered to QA this one +The code from the renderer is quite different since it mixes back buffer generation with rule filtering and rule selection and filter injection are in two different parts of the code +Fixed by fix applied at r +Code reviews are welcome +Deferring post console work +Also using to make the output prettier +will address that in this JIRA +Resolving as Wont Fix +Isolated test case with all deps except jruby +Did you apply the patch from my previous comment? Can you supply more context for the next exception is it happening with the same query or is it an unrelated issue +Fixed in it for closing all issues that have been in resolved state for months or more without any feedback or update +This patch assumes that if bytes where received and no header was specified the read data is alrightThis change definately fixes read time out problems caused by badly configured servers but still relies on the connection to time outPlease comment +Thanks +I dont want it to cause a duplicate import +And maybe then add source libraries correctly labelled +Alternatively the scaffolding could use the last fall back option of layouts thus giving maximum room for overriding. +Great ideaThis new patch has this implemented +Backported to branch revision Backported to branch revision +on an index in these circumstances will fail as is not honouring the Identity of the applications +Having copied across files to fix the adopted suite of tests it turns out that this was sufficient to return the optional suite to state where there are errors and failures so I believe the CTS is now independent of the Tuscany impl resources +move of since edgecase +Stacktrace attached +The purpose of my task was to check how does examples of migration work in DR +Checked in a fix in svn revision thanksdims +also addedrequiresProject falseto the archetype plugin +Heres my implementationI agree that the JAXB javadocs statement about extending being implementation detail implies that my solution is the intended oneThanks for the speedyness of your responses +ticket for getting the TM working againGeorge +Please review and commit +Patch submitted +So you retrieve such result set as an iterator exhaust it than you can retrieve the same exhausted resultset again and get that its empty +If its resolved as a Wont Fix this should just target the duplication of HADOOPCONFDIR and then be resolved. +applied a fix for the defect to both SVN trunk and branch +Elliott You going to add comments? We should backport to especially if we are going to update to hadoop etc and make jmx work +A mistake in A above +Did I understand this correctly? Are other threads involved which also do something with converters +OK Im for addressing that +We have to decide which way to go here +The version of POI am currently using isI can try to extract the test with no other libraries but Im not sure if it will work + added the integration and CT tests all pass +Looks like I had a stale jar the last time around +Btw +I disagree with GordonGenerating bindings is a client development activity in my view +It will have to handle cookies anyway because they need tobe stored outside of the HTTP Clients and I dont want tocreate a new object for each request +Looks good thanks Keith +Issue closed with release. +believe this is now fixed in TRUNK +Fixed for included in. +Attachment has been added with description Demo of failure to process attribute of fo in CVS on sep should be released in +Hi for checking this I have noticed some differences between my local version and the latest one from may be due to some extra functionality we added +patch looks good +Lets also move the code to a new package +Does it matter that the cert is using MD and the bob is using SHA? Are there any changes that need to be made to PHP +I have these changes staged in git ready for a commit just putting a patch here for review +Question should a region carry all the information of a parent region? Example regions a and b are the production of a split from c so both would have in their meta row all c events +Rado please add ER in the affects version if this issue still exists on the ER build +I see now that this is not the final patchFor the slavemaster side of things currently the slave cluster does not have a consistent view of the clusters it is receiving edits from +patch matches current for spotting this +Screen shot +Checking the Seen in production box +Thanks Justin and Andrea +Missed to clarify that Im uploading on Vinods behalf +Arjen I will try it out this week +I will try to resolve at runtime which version is used and bootstrap cassandra with correct config file +The semantics of namespace qualified names is solely determined by the namespace URI and the local name +This is just standard as described in the custom validating for my AJAX form is just another level of validation thats why Im using an Emum for the different transition values +Code fixes and added a new test for negative block verification as Todd suggested +But the latest patch still hardcodes the logger names in and which will unconditionally override custom env var +Sorry for the confusion. +Fixed in revision +When can I see changes in production release +Something like +Ill email you about availabilityAs for the +Cool thanks for the patch Jonathan +You could size the messages and configuration of flow control such that you know exactly when the credits have been depleted and you could drop messages to simulate retransmissions that were dropped +Everything works fine on Amazon +Ivo does this mean that the first two paragraphs here can be removed from the release notes? Should all users commend out exactarch or does this option not have an effectOn Red Hat Enterprise Linux check that the Tomcat packages are not already installed if not then install the packages by executingyum install tomcat tomcat httpd in order to install the Tomcat packages you need to make the following modification to the etc file +I am reluctantly okay with enabling it by default +Added test in Rev +SVN copied files from trunk that are not in the branch in SVN revisions and +This enhancement will make our JBM cluster solution more youMirek +There were many other things configured in the parent pomAfter some investigation I found out that the problem was the version +Yep +Hey BrunoCan you generate one patch that applies against trunk so I can reviewThanks +Oh and thanks for taking a look Jieshan +I realize this might be intentional +If I understood correctly the following header should be interpreted as very long header instead of verylongheader rightHeader very long checked in + need to discuss a suitable solution for this with Christian +from me for code made into two files for putting the streaming part under streaming directory. +I think that the attachments are better to see the result send POM that produced the screenshots +is a race condition since it works on my machine +Will it be possible to perform a hive query that joins against anyall values in a ListSetMapAnswer should be the same as for existing Hive arraysmaps +Hithanks for prompt reply +fixed in febfacabdcbfffaefda unintentionally be refactoring of +As the changes are only in class which is unused for now I havent reran good +The flag simply controls how they send the attachments +See for a bit more argumentationSo I must admit that Im not sure I understand what it is you suggest exactly concerning Cassandra itself in this ticket so I dont want to close this right way but I have the feeling that whatever needs to be done probably concerns the jdbc driver not CassandraAs a side note I wouldnt disagree that CassandraCQL handling of dates is a tad limited and it might probably be worth beefing it up at some point but thats largely a different issue +then submitting a patch +replaced patch again install it from looked good so I committed it +Interpretation of those entries are entirely up to the implementation +I dont this feature anymore +doesnt have this problem false alarm +merged fix from trunk to codelinemjdksvn commitSendingSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestsstoreTransmitting file dataCommitted revision +add tests to verify listener registerunregister event callback and explore the behaviour of listeners when listeners themselves fail +General support in CDI Tools model is implemented +Please describe the problem encountered by the user the cause of that problem how the problem was fixed what the user now experiences instead of the problemThanks! Im watching the issue so no need to assign to me +That is correct +The problem was just that uploading the file is very slow from here +I saw you create one connection and close it for each row +see previous test pom which shows the problem with two entries +Whats the difference to the testing resourceresolver mock module we have +This version of the patch is a rewrite +code looks good +Works perfectly +Updated userguide in revision +Works like a charm for me +hitting the submit button for is a patch and doesnt apply to trunk +I gave that a try and it just changes the scan period to seconds before it hot deploys +When I took it out in the actual script things started working no duplicate there a reason to use updatetrue instead of just building a whole new WAR? When in fact it should be assumed that the WAR should be brand new since a new build is taking place +Moved to subsequent release not subsequent patchThis improvement request has been open for years and has stepped up to implement it +Closing issues that I resolved in already released versions +Aside from the obvious missing tests theres some further work to be done to complete the implementation but Im sure Mike can point anyone with more free time in the right direction +If you have some numbers ready that would be great otherwise I will try and find some time this week to see if I can come up with anything +need to mark this as since +Yes it is + +Id say it would be more common to have one of those chars in the username than a username with length greater than +The side is the owner of a association +sorry not sure what you mean +I forgot to mention that now you canoverride everything using D flags also +Jonathan do you have a test case that forces this +Ill keep trying +Since there was no feedback on this I closedthis ticket You may reopen it to disscuss +I am having alook at implementing the standard Listenerinterface with and right inside the +the idea is you can wipe zookeeper and hbase recreate the statebut in this case is not a real persistent information +I agree with Jukka that a prominent listing of all licenses should be encourages if not requiredThey definitely shouldnt be in the NOTICE fileThey could be stuffed in the LICENSE file but I can appreciate ant elders concerns about readability +Also I think that a change that breaks compatibility should be in noWith respect to the issue of this jira I believe we would return true in the case the ledger has already been deleted and return false if the operation really fails right? Throwing an exception compared to just returning a boolean enables us to convey more information about the error even if we are not conveying a lot right now +Hiram can we mark this resolved for now +All we added should be a kind of raising I think +Linked to and +Patch for license violation +Thanks for the reviews guys +Similarly setting field values to defaults when writing may harm performance when the application overwrites the default +the heap is not the problem +Closing the issue. +Done. +It would be better to add it back to the Wiki +You cant never write enough tests +Have you seen this a lot in production? We run into a lot but havent seen the broken acceptors in +That will basically fix both aforementioned limitations +There is an outstanding issue that documents this issue +Kishore one of the guiding thoughts is that core should support this feature natively rather than use another infrastructure. +With this patch JAXWS will cache per service +Most recently I updated my jdk to SR and started tests the tests completed and continued but I came back to the failures and the Symantic screen asking to give the new jdk permission to acess the network +I just sent you an email with the two files I was working on +All of the old tests pass as do a few new ones including a test of behaviorThe JIRA patch problem prevents me from attaching the patch to the issue now but I will try to do so tomorrow +Your analysis sounds correct +I am fine with removing the memory handling and using OOM +bq +Discussed this with Werner and decided to stay with Nabble at the moment +Can you please raise a new issue and attach a minimal test that allows me to reproduce things +Fixed well spotted thanks +Please try again +You didnt say what grant types you needed to support +part of release. +Patch is updated +And how we expect write to work withThanks a lotRui +This patch ensures that the intermediate C build directory is not packaged with the Avro release patch ensure that C object files stay out of the Avro release just committed this +Post the patch anyway +Sorry it has taken me so long to respond +However Iam sorry for not having understood the desription in the pdf that HadoopFUSE could be used to provide an NFS interface for HDFS +I found the right place in line We need to check if we are creating complex java enums there +Committed v to and trunk +This issue has been fixed and released as part of release +Attached. +Includes a new timing class plus timing data inserted into the dismax handler +Eric I wasnt proposing we ignore Windows users jeez! Just trying to diagnose what the problem is so we can figure out a was asking because it may not be a bad idea to ignore Windows users +PaulAll fine for me +is confluenzad so I think this is closable. +I suppose this is what you had in mind? But I dont see any purpose in giving a relative href to the same file +In setting up the sampling job we should know how many reducers we are going to use +When the merge commit is being generated the current head revision is used as base revision id +Its makes it much simpler concise to users in their specification of a method binding expression and reduces our dependency on all known inconsistencies. +is bundled with the ear in my case +Default container cache size is +Under unix there was a number of source file macro preprocessors +Makes sense i think +I have done a lot of web projects with Spring and Acegi for clients so I will certainly volunteer to write somethign up for this +code does not work when feature is enabledLots of failing unit testsSee my additional mail for more information. +So it seem that for now we are stuck with doing FS caching +Schema required for manual patch of + +verified in +Closing resolved issues. +Add description for pzxid in Leader our docs source is in srcdocs there is a source document in there for the programmers guide +Im thinking about cutting the release branch for after getting this patch in +Please feel free to modify the project and reattach it to this senseIn that case maybe you could attach a simple test case where your skins dont show up properly? I think you can use the same project and apply your skins and attach itThanks have checked in fixes for Callout and skins for dpi +on the back burner +Ive attached a patch which falls back to using to locate a if no custom has been registered which can handle the conversion +I dont think wed need to change the client facing API I was thinking of just moving more of the work to the clientWe could make a java based importtool that a user can pipe the CPT format serialized data into to which in turn would encode it to BMT or serialized rows to be streamed to the nodesThat way most of the work of encoding decoding happens locally and users can write their loaders in any language +Please forget about the previous applied +The code is only about showing the classes in camel transport for move Christian +Update Summary to reflect the tag that well be building from +As to and if the patch applies clean I will just do some unit testsOtherwise I will file a separate jira do backport and test it on a corresponding live cluster +Does Philips solution work for your all +I am using I dont know about wrapper +All unit tests passPlease close if this fixes your problem +This is the incubator requirement +You might be interested to introduce the same customMethod pattern in to allow more flexibility +On two occasions minutes after I see this in the logs Cassandra on this node is at usersystem on all cores +Attaching a new version of the example Visitor which implements the changes in the patch a new version of the database procedure which pokes Visitors into the compiler Knut +thanks for the feedback. +This was not the problem +Reassigning as it is Remoting only way to get around this is to apply the new XNIO and Remoting versions to the ER side +Ive finished fixing all the pages and added some small improvements +I see that on the elasticsearch website and docs that they do sometimes capitalize the E + used for text drawing on linux instead of when current font library is used +Maven should always deploy what it installed +After successful fighting against a jaxb issue i have a problem with unmarhalling the events at theAt this this time i havnt the right classloader for the event payloads +Ill have another look at this later today +Thank you Han Hong testsuite is OK now on Geronimo +As suspected it was a trivial can you guys cut a release +can you please test comment from the person that I originally opened the ticket forFYI I gave it a quirk whirl and ran into some new compile errors relating to a missing java project +Thanks Raja +ok will try and checkout issue can be closed checking out again fixed the problem +All queued puts are made at the end of the file +In the meanwhile Im doing my test runs on my trustworthy Ubuntu VM just to be safe +nfsfe +Moving to by the patch submitted for which was committed in trunk r. +See for more on modules. +My only concern is that the impl will suppose to function on top of class where to obtain one instance of it we have to provide a registered an apikey and an apisecret +Many apologies its Christoph not Christofer +I think supporting both modes would be optimal +It should always be called but if there arent any sources it shouldnt generate and the site plugin should not include it in the menu +Proposed fix which stops the bundle to be updated before the update and starting the bundle again if everything is donePossibly the should be modified to not abort on any exception from bundle update or installation +Attached is a snapshot which was taken against a clean build of AS x branch for default server configuration today +Can we do that +Im uncertain +Attached you will find a maven project for the bundle that is throwing the error +to trunk +Patch committed after further tests were added. +duplicate of +Nice +Thanks John! +Just committed another fix which fixes tab completion when in mode to only show those commands once with no global scope m going to mark this one as fixed as pretty much all the issues are resolved in karaf just need to fix up in the next release. +I am keen to participatethanksThe following issues have been moved from to and I am sure it will be moved to cos its seems to me that the comitterasignee doesnt have the time to fix it +I am using offline payments only and I couldnt create an order until I installed this patch +The formula is about the number of terms visited by the NRQ and is not the number of terms in the index +Fixed in trunk with revision +The attached each contain the same virtualTable and virtualProc +i will take this work server version api version support to CQLcqlsh select serverversion apiversion from dual serverversion apiversion cqlsh select from dual serverversion apiversion Thanks approach seems very special case with positional keywords not handled in the parserlexer +Deprecated probably can be compatible with r +See +Is any further action by me required? Thanks +The changes to make Pig and later work with Hadoop was committed in the meantime +Any ideas? Or I should just wait a little bit +Just saw your comments +Patch file for applying the above changes to the class +Marking closed. +Other code was adjusted accordingly +oh ok brilliant i did have a look threw all the issues but i could not see anything that looked like itso it is set in but is this not an option that would be better placed in +Would you please help to try it +EmmanuelThanks for your reply +is removed totally +dev testing productivityintegration +I tested this on centos and ubuntu and made sure ambari server setup and started successfully +Thank you for the quick note + lets keep VTL simple and quick to learn and not turn it into java +If the problem is that the has to be loaded by the same classloader as the Struts classes well that is certainly the case. + push to +committed faceting part and closingif a separate component has an issue with no docs matching we can open a separate issue. +It was released with the fix in +If you have suggestions to improve the README the job the ant script etc +nanoTime cannot be used as absolute timestamps it can only be used to compare times over a relatively small interval +It seems like reasonable code +Overriding defaultStack is also a bad idea +Thomas thanks for this hint +Now So the dbmdwrapper tool wouldnt live in as an exception? If our first use case for an optional tool doesnt fit there maybe we will have to solve this ad hoc going forwardRecapping a bit in the description you say tools which wouldnt be checked in the to main code line but if they are to be bundled with our jars shouldnt they be? Sorry if I am a little confused here +with a lot of configurationWe should support this cloud in better terms + +I finished reviewing the format +Thanks Willem I exposed the underlying attribute as ant on please rename as necessary +Every partition has its own Storage Descriptor object which contains information about file descriptors serdes and other storage related information that a table can have +Ive committed the patch. +Nice that sounds good +dims +I can definitely see the value in this suggestion +xhtml files? +Closed by request. +I dont remember anymore in I should probably install CVS HEAD somewhere +keeping this open as a general placeholder for any more reference manual updates +Fix version was since I close tag Jira move all issues to +patch to apply after the commit +Updated patch incorporating Olgas comments +This isnt about making it secure +Yes New Other jBPM jBPM Action should be put to can implement it exactly as New Other ESB ESB Action but implementation depends on plugin +would we be able to keep using a stable release or would things like make us decide to go back to revision specific versionsmaybe it depends on whether thrift releases are frequent enough +Users who want to run on different filesystem can change config to load different implementation +Patch applied +Didnt notice any phantom connections but I didnt really look very logging level TRACE is too verbose +hasnt been updated for until we can do in release of June +Patch with fix for javacjavadoc and findbugs warnings +BJI totally agree and I reopened issue but I cant do nothing about it +You wont be able to touch the message as another thread will be sending +If you compile it they why you did with the same name and then look up the class then there will be a class loader for the new class which is a child of the loader that contains the calling code there being the other class +Following that the admin console now appears to be working for my original user login as theory Safari was associating hostport with a userpassword combination but perhaps it was screwing this up somewhere and to a with another servers ve had a lot of different local servers running on my Macs over the years and the localhost credentials had piled up somewhat +Maybe the failing ones are new tests that do rather than use the test Configuration which will have the proper zk port set +Knut is right the following statements should fail because we are trying to override a generation clauseinsert into a select from binsert into a select i from bThanks for finding these problems +The pom file had bad syntax for multiple resource inclusion +JayCould you please confirm whether this has been fixed in the branch? Please commit the changes in this branch and resolve checking in code that will fix this issue. +Yes still the same problem +When I ran the same exact test as before with the exception of the lucene core update I still get a failure +Maybe a new jiraThoughts +Opened new JIRA as +? Looks like to meInvestigating further as to why log recovery was determined to be necessary +login like script +I will commit itThanks +Logs are attached for review +As a side effect makes it possible again to call inside an method while the framework is still in the STARTING state +patch +What do need to if the version is and which components need to do the work? Thanks jpa EE deployment sectionWhen the container finds a file it must process the persistence unit definitionsthat it contains +My fix with test cases +I have implemented my approach in +Attached a fixed version +Fixed on revision . +I just added a for these two appenders are applied with rv +Ive tried the second solution without this only affect test classes and we have a fix using Id be happy to close the issueand readdress it when we have a specific + preferably with the ability to produce test results for CI +Yes I will prepare something to show the problem within the next days. +Fixed with minor is the change youve requested preserving the original exception +Based on discussion from our meeting Im setting this jira as resolved +Hi Jakethe last one is different +I finally managed to find the OBVIOUS function +This patch addresses the issues pointed +Moving to +Updating javadocs or enriching hbase book would be good for the correct way to orchestrate a region and compactions +I havent heard back from APIm fine with releasing as is and saying logging doesnt work correctly yet and that itll be fixed soon +fixed a bug +committed thanks! +ThanksRegarding there should be no real impact this is just another extensionIll run the tests on Derby and probably Postgres and commit if they look good +New modules for ASquartz +Note this will require a change in the test is tested by CDI TCK +It also does not mean there were not problems with very long varchars in original cloudscape code +Just gone through all the protos that we have +You dont need to use dynamic imports +I probably didnt need to assertEquals the entire exception text in test +The work to resolve the issue still needs to be undertaken and landed in +Right on! Thanks. +So I think Download additional server adapters link didnt install the adapter with P manger +Note I modified the demo to use asynchronous invocations but same result +I believe the browsers in thiscase have implemented the most reasonable interpretation of the spec +never +Will commit soon +Please feel free to reopen the issue if needed +I did read it last time I have requested you say what I have to change +Release audit warning is being generated by which is unrelated to this patch +Applied to TRUNK +Unit test and run are in progress +We should probably make a separate RPC for du +We only really need to keep the result in the maven repository. +But in you have to click on the notification bar +Pl +Thanks Anila modified version of your patch is in svn with rev +Either that or the destinations on the main? +I changed the name of the property to follow an already existing +Should this be assigned to Fernando or someone else +It explains how everything in would work in practiceIts difficult to break this work up like one person working on deletes because to do this stuff right it requires a much bigger overall rearchitecture of things +HiThe error was occurring because in order field value being passed was that of reason +is there a working test case for this? right now im getting two stepScope Objects yes i know dont mix namespace and stepScope bean but i dont know how to set the proxy to true while using the namespace and getting stepScope implicitly +Please test again with latest code for letting us know +Would it bepossible to allow for a certain URL to be excluded from the security check +I need to be honest +If there is no enthusiasm for a release and switching to it as the supported line then features like this need to be retrofitted to the line and perhaps a issued with all the new stuff +Committed first part of the patch in revision Is anybody having any suggestion for the notifier part +Are you sure that the other connection does not need any admin rights and we could drop those rights from it +IMO its absolutly useless +dupe meant dupe to +No test because were removing the option +Would you have the opportunity to test it before we port to the branch +Pavel please verify this fix on the CR docs bundle and close the issue +Thanks for fixing the bug +Google finds these in preference to whatever is newer +This Jira was resolved in +I am using the embeddable Groovy JAR and I dont use any other libs which use Groovy so not sure where this is coming from its certainly not written by me +Removed the code earlier today to understand dependencies etc +Thanks Arpit! I applied the patch locally and built and tested full distros on Mac and Windows +Rescheduled for later +I tried myself and I ddint get the download messages when setting log to quietCould you post your console log +I the tar tasks and added the longfilegnu parameter meaning that long filenames will only be handled correctly by GNU complient tarOther possibilities are to fail truncate omit or warn but that doesnt help muchThis change has been done in patch +reopen to fix curly brackets in the clear text has been implemented in which is a dependency of +This way the files can be packaged together as an external jar in the ears classpath +I dont think there is any spec issue here a passivated EJB with CDI services should just work +I am sure that you are not implying improving it is a bad idea +Removed fix version its cloned by. +I will this patch so that can pick this up +Could you also confirm that the file hasnt been renamed recently? Freddy we use Mercurial to version control the project and it clearly shows the last revision to be weeks old so there were no recent renames or other kinds of changesHowever youve given me another idea we are tight with space on our Jenkins instance so job workspaces are wiped automatically after a job completes +Will attach a test case soon + +What we should do here is allow users to add class prefixes to the exception list +Could you please change DFSSUPPORTAPPENDDEFAULT to be true +Well not strictly a bug +Got it nowThank revision +Thanks. +Closed with the release of. +Is it OK to upload sql files that total about Mb +TTL defaults to implying infinity the values are propogated down to contains updated logic to caclulate expiry Unit test to verify behaviorSince TTL defaults to which implies infinity this patch should have no effect if the optional arguments are not provided to the constructorsThis patch is submitted with the permission of my employerThanksMichajlo +Here is a patch ready for a code review +NETAnyway Ill reopen this issue and would like to hear what everyone else thinks +There is a ticket in FUSE MR as well +In revision I added a configuration option by which the strategy can be switched from the default writer preference to reader preference which at least works around this issue also for the case +I have changed the xmlsec dependency to use the Apache version which is equivalent to +v patches which hopefully fixes the remaining flaky tests +Folks on have been reporting problems with and is not a installation +Jan Kuzniak What are your preferences on how should I bring it here? I am thinking of one batch file or one ant file that when executed will copy files here and thereId most prefer a sequence of unix commands to execute as Id like to execute the changes manually if possible +all the code working on +Resolved +Please close if satisfied +Fix for broke packaging + for the patch +Could you post a little test case or example of how your using the paho client +Ive got a patch but I cant quite figure out how to disable ansi to test it +We found this useful in environments where you cannot access the web console via browser +Let me know if you need details on anythingBest regardsRemko havent had a chance to look at the attachment yet but Im pretty sure that you arent really measuring the time to log the events but rather the time before control is returned back to the application +Asynchronous events and race conditionsbq +Marking this for +The connectionString field in seems to be incorrect + +the question is how a list containing itself should be you would want to display is just a preference +Its not anymore +But do you perhaps also have some comments on the implementation? Im sure it can be improved at lot +Seems like the openstack rubygem already supports this feature by setting the region parameter within the connection string but its impossible to set this parameter through deltacloud +So this isnt an issue +I will submit a bug report and try to find a short term fix +This is the same experience I have with every maven project Avro or otherwise with MEI havent used the maven eclipse plugin in years so I dont know much about that. +is that right +What do you think should happen instead +Thanks +Thanks Bo Alder for the earlier patches +In any case you NEED to find another way to expose this +I tried to run the cluster tests on the latest code from JBPAPPGACP as suggested using this command +The match functionality never did work as expected in many situations +A new test case tries to create the new entry race condition + is a different failure in different testAnyway it is not a result of fixing +added skins for panelBar panel fixed errors in description of classes +I will update the bug status accordingly +Thanks +Here is a patch for branch +Fairly straight forward to do +testcase attached change web admin password in for input +close for +However could you look at adding a unit test for this +Added convenience utility functions for setting parameters using naming conventions for the static fields so user does not have to specify them +I assign the same css class to both cssClass and double attribute of optiontransferselect but in the generated HTML code double does not show at all in the code +Even though the current throughput of this patch is similar to the existing one it will be a key feature for efficient IO when we fully optimize itLater we could also improve this to something like cooperative scans for multiple queriesIll take a look at this patch today night +We could in theory merge a fix to the and branches too if we wanted. +You cant starve replication too long by doing deletions as there are only so many replicas to delete but is it OK for replications to wait a minute? An hour? It is perhaps best to just follow Hairongs suggestion + to trunk +I could experiment to see if I can get the same NPE problem with a database +Attached a patch unit test to fix the problem for +So for the example mentioned above the is set as the job jarI have manually tested this fix on a single node cluster and confirmed that the example runs fine +However will still need to be excluded +Tested English installation same crash chance which LN scripts broke the strings that JBI rely on to unpack is not high enough because the installation progress could be finished in by a few times +Add scheduling hints on RPC timeout through on v You are still waiting on hints in the client I dont think we need anymore +It looks like it has resolved itself over time and now Im able to post to dev sorry for the noise +I am investigating this +This patch assumes the following HADOOPHOME has precedence over HADOOPPREFIX and both have precedence over the hadoop that is available via PATH installation of Hadoop is expected to have working binhadoopI think this change gives us the most flexibility in supporting all sorts of versions of Hadoop without trying to second guess their layout etc +what new utiltities will be providedIll update the spec the proposal is to add a switch to dfsadmin and a similar one to a new command mradmin +go for it seems to be quite pretty compared to the current trunk version. +Added doc and test targets and updated to call configureWill test on bit platforms and improve +For the API would it helpaffect +They are designed to be object model independent +This old issue goes away if the extension can be specified on the action which I think is more robust than allowing it on the form +antlr and antlr have different package namespaces looks like the antlr jar has not been included in the deployment +I dont think the above failures are related all of them have been flaky of late for various other reasons +Thanks RichardPatch applied to LUNI module at repo revision Please check that it was applied as you expected +And then I founder another issue to fix as well +I removed the directory skin and have the default look on +Indeed I reported this against Maven the version meant version of the checkstyle plugin +we should probably find a better solution for that one Ill look into it +bq +Might it be worth it to name the Mutation fieldssubstructs closer to the action they are performing? Deletion and make sense but Counter might be better as IncrementI dont necessarily disagree but what I dont like with Increment is that it suggests than you can only use a positive delta +Ouch! Anyway I ran a few unit tests they still pass on my machine +Its not related bug +While it doesnt mean there cant be some hashing scheme to map arbitrary values to ints thatsBut yes the docs could be adjusted to reflect realityThoughts +Now it is not possible to achieve it +But on JDK the following libraries are needed to be in the classpath +Thanks for the patch and the note N +I am not about the cause but just printing and closing client socket for a is making the test to fail +which is not totally optimal but Im probably missing something +Refresh of the patch its currently running on all clusters at +I agree with you though that greater flexibility makes it wellworth paying the priceIll incorporate your changes into the new patch which I intend to submit tomorrow +In jmx it will be under Hadoop The new metrics has context and jmx context +Committed this. +Is the problem that I didnt run svn diff from the root? If so Ill do that and resubmit +Thanks Oliver +Otherwise we will create two configurations +Does changing the manifest entries fix the problem +The javadoc for set saysSets the designated parameter to the given input stream which will have the specified number of bytes +Works for meThank you Andrey +I see +Yeah we want this too +Can you confirm you are running eclipse with the JAVAHOME set correctly in environment variable +I am using a custom writer which is default response writer and I cant select it using an admin interface +Thoughts? So it looks like the css linking to the icon just appears to be broken +With the current bundle configuration for it can not access that class +That failure was supposed to happen + s formal +it would have been better but it also woud break existing user apps if we decide to go that route now and we dont want that if thats avoidable +Ha! does the job +On a edit log failure throws exception and Primary fails or shuts down On a edit log failure continues +Yes we may be able to use for primitive multuiparts and possibly save on typing required if is should work with simple multiparts can you double check please? If so then wed consider this issue being thanks for validating it cheers Sergey +Still need a way to know that a fieldfieldtype is going to output multiple fields so that we can detect them when searching etc +Is this what you need +Youre right sorry it should be listed down as an improvement so as to allow multiple gateway methods having differing uris use the same http gateway +Looks like new ones are so this means runningchmod gr homeapmail + +If we dont see this again after a while we can then close the issue out +I was just throwing out a strawman proposalThe pros I see for the current updatepick Picker api is that it is more generic +I did verify that it sets the expiration of the timezone cookie to about years from now +Havent seen such case myself but Ill have a look +Im open to improvements but Im having a hard reconciling how to add this capability +Instead of adding these checks all over it would probably be better to add one configure check against the compiler version and then add them if its greater than patch as it is breaking build for current systems with default compiler less than +By default there is no menu button on any devices running ICS or higher +In we have actually dropped FFI support for PPC largely because the build now depends on OS X +Then I guess we dont really have an issue +Applied patch to both the branch and trunkDon thanks for the patch. +It was fixed in trunk by r need to be ported to by r +Heres my vote to do the copying with NIO instead of buffer slinging if at all possible +What system are you on and what gcc version are you using +Committed. +Sorry Im confused againMy ultimate goal is to get a patch for which will allow triggers on tables with blob columns if none of the blob columns are referenced +So the cache control directive would allow me to do that +Attached patch for validating User and updated patch file that contains validation for the ff +Now diving into the depths of jQuery +Pls use with at least and set your JAVAHOME correctlyIf it still occurs pls reopen. +In theory to use the as key you should do something similar to the hack done in +Heres a patchTests pass and Solr startup doesnt trigger the resource loading exception +sorry cfstats is not the stats +test believe this issue was fixed in +You dont have to clean old attached files +Removed the references to this configuration and related code from both archived installer and ambari +For most of them the problem is solved by making it by doing equal compare as a string the same way JDO does +Paul Gier is watching the issue +h +BobYou cant see the man behind the curtain if your gaze is focused on the smokey image in front of you +As an alternative what we could do is lookup the namespace prefix which matches the workspace name and if it does not exist maybe use the default namespace? on checking if the datastore factory actually accepts a namespace parameter good I do not have a problem with the patch +I just pushed this to trunk +Yes were very aware of this issue +How would you execute this? Plugins defined in different profilesWouldnt you have to execute the project multiple times with different profiles +You mean there is a stored in the headerThe file component ought only to store String headers with file name details such as the file name size parent relative path absolute path etc +RTC item +I think this is the best way to solve this issue +I had already installed the VS it wouldnt build via running or rebuild solution in visual studio but it starts fine if I hit F +Adding details +For example we specify we want the most recent versions of explicit columns +Ive tried to reproduce but failed so if you succeed Id like to take a look too +To do so requires a fair bit of surgery to the shutdown code +Fixed for EAP didnt see any incorrect reference of JBOSSHOME even before the change +One for successful connection if there is no RDBNAM on ACCSEC but we send it on SECCHK +If anyone chooses to fix this issue they should look at optimizes creation of the statistic row at least for single column indexes +This is because of the erasure of information in the Maven repository format +Verified +patch looks great pending jenkins +Now to test and do the MR version +In what? There are like a dozen things that use these fields +Updated with a new patch that includes some testsFor fairly obvious reasons its difficult to test most of this functionality in an automated fashion +I checked the test case into head etc testsfunctionalorgjbosscacheaopcollection testsfunctionalorgjbosscacheaoptest added the test case to the known failures list in are still looking at this issue for the release although we may not solve it for +Please commit it boss +Small update the availability check of a tag must be done in the sequential block of target +Sorry I hadnt noticed that the issue was typed as a feature request and not as a bug when I created it +If need a particular feature or bugfix to be included in you are welcome to get involved and contribute +But Zip and windows built in zip both create a directory named U with a file named in it. +Moving to and +Linkin PR which I have sent against upstream repository +This has been fixed due to a small block of code missing in wherein the principal was not getting set to the principal that corresponds to the existing in the stack +Jacopo Those sound like just odd bugs rather than the need of a +The agent refused to start and Agent already in use showed up in my logAfter I ran MCF started to work againOf course the locks may have been caused by a kill I had to perform when I tried to solve the problems I encountered +nice I like it! Is it possible to fix the same way for the other highlighterthis one is tricky since it involves getting the field from the query in order to fetch a reader +I forgot to say that I agree that the relative URL case for icons isnt especially usefulit would only really work if all of the pages were in the same idea for a fix +I thought to expose the project working directory value in the project page in a future version to allow the user to modify it for example to recreate a specific directory structure that is needed on some projects +For instance I plan to make another patch to camelcase the arguments in the function topics +We can leverage Springs light wight transaction framework to support the transactions on the server side +I will put some effort into scripting the problem +I have moved all code relating to messaging into a new delegate class +Smooth upgrades through rolling restarts are an expectation that has always maintained +Use the user forum before opening +Should this be ok to commit then? I assume we dont need so I am not submitting patch to not waste its time on it +I pushed a second attempt in ffdc. +I appreciate you taking the time to do this +Ohh +Im working off of +What do you think about it +We cannot afford to ship a point release that is less stable than those that precede it especially one made from an old branch +Reload the patch for to try again +If that is the case taking thename from sender or reciever become tricky +Is there still interest in doing it +Doh you guys sort the opposite way we do +It includes visual examples of the AST and queryplan as well as some details on the path from SQL parsing to code execution +Currently people need to read the javadoc to get thatBasically we should distinguish this feature from pagination support +I am reviewing the patch +svn rev +be used for automated testing of the tests module? We could remove the test I added and put this one in tests instead +The rest are required +Ill move my comment to a new issue +But if it unfortunately needs to wait until CXF then so be it. +Update to in module +Matthijscould you provide a code snippet or a real example of a complex javascript that triggers a parse errorperhaps there are several different root causes so if you have several real examples feel free to add them all +If Im logged in as mamburn and create an issue then I should show as the owner of the ticket +bq +Just up the wait +Thanks +I reverted for now +I am going to provide a fix for both problems soon +I know that it is all fashionable to dis on the Java cos and sin functions but really +Will post a new patch soon +Incorporated Konstantins suggestions +No a separate bug report is unneeded +I was unable to reproduce this problem given the information provided +Its a regression from +The test test in demonstrates this issue +Ill keep you postedCheers Chris that failure was due to reboot starts even before stop fully completesUploaded a new patch tested locally +But in my opinion this is not the case and the described problem still exists in trunk +Closed a misunderstanding in where the going through the proxy needs to be kept in mind This is a limitation of AOP in general +Any chance of providing quickstart that could reproduce this behavior? Does this also happen for wicket examples +Thanks Jackie I just ran into this issue myselfIve checked in your patch to CVS +closing +But is a little bit complicated +Completed At revision +It cannot make a connection to the embedded activemq broker +bumped to to LATER since not sure we have enough time to do this in the near future even though it would be highly advisable. +Attachment has been added with description test FO to reproduce exception in has been added with description resulting PDF in transition resolvedfixed bugs to closedfixed +I havent been able to accurately test this as I have never seen a If someone has a test application that can demonstrate this then it would be great if I could use it to solve the problemI still get a consistent increase in speed with this over the standard Mina have had a look at the stack traces but I dont see any deadlock? Which threads to you think are deadlockedAlso could you try this with an unmodified MINA so that we can rule out your modificationsWe will also try running your tests +I have update itDhruba I have removed the sleep in the test +too +Added options to treat inconsistencies ignore move to lostfound delete +getFrom gets the IP from the message header and the message header is set by the caller which uses and the receiving machine sets it +Im at the Strata conference this week so my availability might be spotty but let me know how I can help +You should write a blog on it and your other findings +I have an important english exam this weekend +But in the UNION part of the statementwhat do we mean by t and tIs this statement supposed to be legal? If so how does it work on other databasesystems? Does anyone have access to another DBMS implementation to testhow this statement behaves there +I changed to not hold a ref to the file and added error handling logic such that problems writing the file should not prevent server startup +opened new jira upon release of Hadoop. +This is very strange +Hey Alois graphbase looks quite useful +All known documentation issues for Pig have been resolved closing. +Correct thats a problem that Rob identified few days ago and it can be solved if we gen because will detect that case and add a new as well as create a new gen for this segments have two tests in which currently test that this is not supported once we gen FIS well change them to assert it is supported +Ok terminated and the build failed but it still posted overall +Ryan is on the right track here +This is already done isnt it? +It could work +where do you want to configure such a thing? directly in the bat files? as a system var +removed until the test is finished + +They are just convenience jsf elements +I am not using the undocumented Java other than the test XML files and dtd files +Nope implemented yet +The argument has always been that there is insufficient evidence to support this direction but neither is there sufficient evidence against it to reject it entirely hence revisit later +Ive started working on this +Adds a proper long type indicator +I took a look and thought it looked good though on my second look I have a question +It will need to be applied to the branch when created +We have the autoexport snippet working so the HTML version should be available in real time on the cwiki host and then updated to the main site within the hour +Merged thanks. +My only concern is that the user now has to define HADOOPCOMMONHOME as an env variable even for running the unit tests +The latest failures actually look like +We already have an plugin which is for making mbeans and it would make sense that mbeans live in a sar project so I think this should live in the plugin +We have to work w CDH too +To expand on that it would be nice not to require to need multiple +is this resolved +An example would be that projects B and C both contain a file and you want to always pick up the nearest one +doesnt need +But the solution should be to delete Produces and Consumers Annotations in class +And I added a test case to the test showing the problem the fixIll commit shortly +Attaching a wsdl for this problem is not happening with the latest build. +It has all of the external dependencies +Spring has a habit of swallowing exceptions and it looks like its doing that this time as well +Fixed under r +ok now it is committed sorry. +was the same problem as reported and resolved in. +Added this map for as well +Should be a simple fix +Hibernate already lets the user specify a custom user transaction jndi name in property +M +So until JIRA adds the setting to say that we want to delete comments from closed issues I am not going to turn that back on +This is not an active collection problem its a the wrong classloader being used by the active collection so no components are returned +and are fine as is +Hmm right +bq + +Could you check +So no I do not intend to turn off the test +I know this is marked as wont fix +Correct bootclasspath handling enables RVM to runxalan with Harmony on r Harmony classlib r build fix Gentoo x +Awesome! Thanks Francesco! Back from my vacation I will test your changes today +Added a validation to make sure schedule end date is greater than the schedule start date +I just committed this +I think we can use the subsystem to avoid such issue +Hi Thorsten sorry about that bad miss though I have to admit I did not have a chance to validate myself problem i checked the sources of all classes and members before implementing so nothing happened. +The code has been committedResolving the issue. +Ive moved the success message to the top left corner enlarged and increased the display duration +IMO this is a bug and needs to be fixed for otherwise an upgrade is not possible +This issue is not reproducible locally +However aims to allow stats collection on the partitioning key +It is also now fixed in and I updated the patch so the next rhds build should have this fix can you fix the part in the wizards +Please ensure that this fix is in master if needed then resolve the bug +Its best that we fix this issue in the branch where we have the freedom to refactor for it +The issue will be closed +An implementation is free to remove the possibility of spurious wakeups but it is recommended that applications programmers always assume that they can occur and so always wait in a hope that this is useful to youKind regards all tickets +The jobs never finish +I think the best approach would be to use a port which is not and use a special account like ldap with a binnull as shell and into a chrooted space +I agree it is limited and part of it is limits in eclipse others bugs in the theme but for now no plans on fixing the experimental themesIf someone wants to try maintainimprove them they are very welcome! but for now they are removed from the release. +Interceptors can also entity conent so these names would be somewhat misleading would be my favourite +However bitwidgets sounds nice +Latest patch fixes an issue with also requiring concat due to exclusively checking the implementation to match s methods +This is not a problem on and I would recommend people just use instead +Please fix your IDE to not make whitespace changes and then submit a fresh patchthanksdimsCreated the patch without formatting the source in the latest svn. +But this looks very very solid indeed +The vs difference is weird looking at it Im guessing that reintroduced the issue though Ive not dug into the changelogs yet +The patch to eliminate race condition in which also fixes the hanging rails problem. +Thanks Zhijie +Any suggestions how to verify this issue +Use the userdeveloper forum for by generating default values for configure properties in default constructor if the default value is specified by user in the plugin. +There are now two calls to enableCaja in +There was some code reorganization around Xerces which would have caused this +geotools version patch cannot be committed as is it lacks a test +Thomas feel free to commit +Yours would look to squash the exception stack to just the message +By making a test project I just cannot transform without network even with the catalog file +I will pull your patch and give it a try +Hi ClausThat is correct +Thanks Brian +to is right +Despite what the documentation says the pear builder adds all jars in the lib directory to the classpath it generates for the +If you dont specify an ID you will override the default instance +Nicholas Isnt the excepton trace different from the one reported +This should not have any side effects but some out side testing may be needed to verify +Good idea Jeremy thanks +There by this can be optimized for Sparse by using vector operators we are hiding the implementation details and reusing the existing code in Sparse Vector Dense Vector classesWhy was cloned to produce this issue? It seems like the same thing +Possible duplicate of +Fixed in branch at revision +But again not a great thing we could do is add a maxVersion parameter attached to the service +Workaround for VC variable scope problem +ThanksShameera +I committed by mistake +Comments welcome +so add an option to haveit auto createdthere are a couple of places the hard coded passwords are +Adding patches for quick review +This would make plugging other compression techniques hard +How critical do you deem this fix to be for the coming CXF release? I was about to close release vote based on RC which does not include this fix +Hi PrajaktaI got caught up with some other work Ill try to publish my comments before next week +However some of the OXM clases are present in both Spring and and yet are slightly different +willem I would have thought it was possible to do this in a backwardly compatible way essentially you would just be promoting the visibility of a specific type to a globally declaration levelI think It would still be possible to support exisisting users spring config files +The only way to have success with cp client is to have bothThis is not exactly what David asked for +per usecase since a usecase means one page access and within a page we use the to get the current please note that in your doRun method you should not call more than remark from Julien about our current I quoteBeside that based on my personal experience I see one thing missing it is the usage of non existing properties or child nodes that often code has to do because it cannot assume that it always existIn context of all optional properties in general a few examples the mop the mopicon the gtnstartpublicationdate and gtnendpublicationdateI think he is right since trying to access to missing nodes or properties mean database accesses since the value null is not kept into the cache yet so we should add parameters to and Read +If something comes up in the associated forum post this will be reopened. +As a small bonus if we promote HOSTID then representing that in binary is trivial too +we have set default read accress to from open to protected in Railo Server Admin to open and it will will discuss about this +The build trick would be fairly easy to implement +Closing xml and console issues. +I agree that the current parser sucks in that respect +Oh +Fixed the issue by manually suspending and then resuming any potential existing transaction during repository startup +Since the Element is coming straight from the XML parser +The UI changes were just adding copies of the schemas to a location for my own sanity and we can clean those up eventually +heres a patch with Uwes testThe float comparison is wasted cpu for as you already know its accepted by the automatonBut the deprecated in sandbox needs this because it has crazier logic +Ah cool with your finding +This is missing in the jetty docs +This is in the middle of being fixed already with the new validation client side system ala dojo +Worrying about Basicis nice and all but personally I dont agree that Axis should break legacy code just to promote the Basic profile +If client dies and you verify you may see lost dataThis is an interesting test case +Remove the JAXB file generation used for the sample for the test cases +I think Ive finally got this +To get that youd need all the intervals to be wholly contained by the next larger one and none intersecting the center point of any larger intervals +Indeed I should have read your description more carefully +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Then the attachments would not be handled the maven way but more the ant way +Which componenti suppose its the dashboard plugin +Justin I am pretty sure you wanted this marked as fixed for +Hey Koji I just tracked down a changed behavior to this issue +Attaching patch which should fix it +Moved the fallback mechanism to a separate issue + +Its ok since the validator only checks if a project has +Now wonder whether it would be better to reduce the size of the API by removing the Headers interface and using a list of value objects instead +This is a necessary evil as we only know the length of the compressed body after compressing all the body and we need to set the header for non chunked responsesYou cant buffer attachments in RAM +I will report again when it is finished +Resolved as a side effect of since fixes are in released code +fixed for inclusion in version rev +When ran Cbin p Connecting to Geronimo server localhostUsername systemPassword Connection establishedStopping Geronimo serverThe command never returns and the server never stopsThank you +Please close this issue as it is not a bug after all +fix in trunk of common module +If need to provide as default then can remove this optionFsck includes only those snapshottable directories which are owned by a user +Windows XP SP rev + +Im now explicitly setting it to +That way I got the Launchpad Builder running +How were these chars introduced? Did your svn client create locale specific values for the date fieldIf so this is a svn bug as the XML files are marked with encoding and should not contain any charsIf you set your locale to enUS before running the svn co does that fix the problem +Patch applied Robbie could you review please +This is still a WIP of course and a few unit tests are failing +Lets fix this a slightly different way +Alan i didnt look closely but could be used to avoid the hack? that way you only take up space relevant to the bits you are actually using +Changed it to sec in the latest patch uploaded +Fixed the code to throw right kind of exceptionModifies the testcase +I am finding a lot of things that need to be cleaned up or fixed in the product store emailsShould I make a sub task of this for those or start a new Jira +doug this is a fair concern +Punting issues from RC to final +Removed AMQ specific jars from set scripts modified deploy and run to use executable the previous version of UIMA AS the CVD application could be used to run a service via a descriptor +The only place in the code where I see such a message sent is in handleFailover +however seems promising and good to get into Solr as well. +Tried adding the to the maven central repo but ran into several issues with lack of javadocs unverifiable source issues among othersUnable to add the at this time +Is there a reason to include it in this release notes +get same addressesgatewayassign ip address of on KVM hosts cloud so I have something to test connectivity to vm with +assigning to notice that the docs are done +Does that sound right to you +Something to inject into this discussionIt would be nice if the element binding of the Any tag was optional +Concrete bugs and changes can be filed in Jira +HiThe fix version is? NotJust asking +Done +Applied at with the release of. +If you attach an updated sample that reproduces the problem I will take a look to see whats going wrong +Is memory usage the concernIm interested to know more details about the performance test itself and the result +Need to look at it closer. +regionservers +Will commit it in shortly +but i am also fine with an exception +Somebody needs to write a test for classloading in the destroy methodof a context listener for for me. +Still +Sure as Ive said before +As a first task can I ask someone with a Windows machine to validate this problem? +Patches for allowing CFS to accept a recovered from which to retrieve the indexed columnsHaving this on CFS allows for other uses such as added indexes after the fact and providing mbean hooks into rebuilding indexes +I mixed maven dependency management with the standard eclipse classpath too +Mayby someone from Sonar team could commit this patch to repository +Not sure if we should require a test +We dont expect this will mean emulation but rather a minimal custom object +Rebased and merged into the master branch. +Moving this from M to M since we are still designing the new that will become the base class for the new Aggregator itself +Ill remind our Swiss friends that the Semweb was invented by a British Subject +Just FYI Hadoop version of HOD will this if it doesnt exist +Perhaps you can look into implementing the exception model for this defect +Then I realized that the keys were delimited themselves and used the same approach for the values +I imported existing models and vdb into another project for which the translator override that already existed pointed to a type that wasnt in the new server I setup +Ill change the name back though I think given that there are now going to be more than one of these a more descriptive name makes senseAdded ASL to Jaro class +Some of these issues are awaiting retesting after upgrade of ODE version and merging of the JPA work from the ODE trunk to the branch +the bulk of the lines in the patch are to rename to and to these test failures appear to be from this +I logged this enhancement with the hope of have a way to put protection around abstract entities to limit the abuse outside of the package but this could be achieved but modifying the java class +It never works on my computer and I always compile the code after editing it to read shapefiles with Chinese DBF field may try the attached shapefile and see if you can get the right text +created as follow up to further improve the tasklet implementations. +Thanks Jeff +The forum thread has the details +Please provide a descriptive error message when thisPlease attach service aar fileThanks have to have all rampart jars in the +You can apply it after you get svn access +Java Collection +I feel like Codehaus people are sleeping as we did not get any reply for eitherPostponing this issue +Just so long as the build does not fail in that case +It says Fix Versions +Closing this issue +The row key is still sometimes read outside of the classes but I dont see us making the sstable format pluggable anytime soon so while this is relevant this would largely be cosmetic +I actually realized UINON in tez duplicates records +Audit logs are added when the user complete the mergingThanks +Oh of course +As long as patches contains test coverage and they dont seem totally off from a perspective Im all for flexible solutions +the behaviour is that if i have a bean that is that is only referenced in a element then the task will never actually run because the bean will not be initialized unless it is reference elsewhere +Taking this down to an individual row value level leads to a couple of observations +The was created against current trunk but should be applicable for many previous version branches as well as this code hasnt been changed in a long while +The likelihood of this happening is very low given how little we can do with the Android +I tried this out and I can reproduce this error +All of these would require renewalThe RM takes care of renewing tokens for HDFS it can do this since the HDFS token renewer class is in the RMs classpath +Es una mica i potser t un s ms aviat literari +Currently alterations to the schema requires a +So it cant handle xmlbeans or jaxb or know to use jsr annotations for hints +Any more thoughts +But gets very nervous if the produced xml is more verbose than it needs to be +If this works any changes made by the user and confirmed by actual saving will be checked against and updated to circumvent this bug in GraphitiThis approach also entails removing the current solution with an update feature +Im interested in this issue and I think its a good point for a new comer +Thanks Brett for the patch. +Thanks again Eli. +In both cases the compilation went ok +It probably was not implemented +I couldnt reproduce the exact issue described but I did refine the return URL generation strategy to produce a better URL +The web console plugins supporting this are calling this service and add the links or help dialogs where information is availableThe service can get a key to define the element which is kind of a url like helpconfig or helpconfigpropA +The attachment contains test cases +Partially was resolved +git commit dfbdbadfcfabecbf +OK I did make a start on so I know whats ahead of now I have to pollute my nice new Air with Maven +All the properties defined in and should be available to use +This will effectively act as a valid lower boundIt might also be worthwhile to write a throw away torture test that has one thread do appends and another thread do reads and check that this condition is not violated in case there are any more of these subtleties +But if I make the fix I plan to which is to make the warning a hard error and yet you were never seeing the warning then I wont have fixed your problem +I have also included a zip of my tomcatconf files with the database connection details conspicuously removed in case that will you build the project and deploy it to tomcat the error will occur if you attempt to navigate to profileshow within the deployed application +Thanks for this but it also needs all the schema versions removed in the test configuration files as well as the poller pulled this PR changed the SI version in the parent pom and then mvn clean test fails with many of us can take over if you dont have the bandwidth but if you want to do it thats ok too +Igor what are your thoughts on this? mangle? specify language rule that this cannot be done +Do you think you need to add doc of the new variable to the README? Done. +richinplaceInput behaves wierldly +So Refresh button shoudnt depend on definitions in the file +I think this issue has been resolved with the newer versions of BND +Patch is updated to apply after r +Works fine with +I dont think you want a split that is escapable you have to use the new split from to do this correctly. +Okay +I took a stylesheet named which has a variable name and ran it with XSLTCs +? in it +In case that provides a clue for this particular bug I might have switched between different install of openmeetings without remembering to drop the database and reinstall openmeetings +Vlado what tag doesnt work exactlyIve checked richpanelMenu richpanel richcollapsible richtabPanel richsimple they work fine +Assigning an unreleased version as. +The attached file uses threads to write to for seconds +If some connection pooling is intended to be used could be helpful to implement this. +This issue not due to +I cant find the way to download the guicyfuit without help of mavenMaybe we need to add a assembly module in guicyfuit to generate the kit +bq +This issue needs to be resolved before can be enabled +After investigation this jira is different from GThis exception is reproduced when build geronimo sample although its built successfully finallyand can be deployed to server can be found under repository but cannt find this application under Application EAR porletSo reopen it +This patch ups logging on master mostly so region transitions do not log their name only but now log the whole +Heres the patch +I wont start looking into it until the beginning of the February though +Fix typo as well as removing a unnecessary word +right now the same branding is used on the ssh login as on the regular startup +I guess an argument could also be made for putting the whole tutorial page in the example serverBut its certainly something that warrants discussion +Incorporated Todds review comments +Here is the patch to add a new profile as well as the complete pom file +We cant override the abstract class methods effectively because any calls within the base class ignore metaClass unless the code is executed within a ClosureSo theres an ugly ugly workaround at least + +Stuck keyboard +See Im reading this properly Cosmin is shutting the table down nicely it aint crashing out + Bug has been marked as a duplicate of this bug +The JMS client now shutdown all the threads and can be unloaded by the classloader etcAbout the AMQ release then this should be discussed at the AMQ mailing list such as either the dev or user +Thanks to the vote from myself Aaron and Guillaume Ive applied the final patch in rev +Thanks Jerome. +However using it shows the same result +Fixed on and closing all issues that have been in resolved state for months or more without any feedback or update +I had typed most of the above before your response came in so its probably moot. +This may all be true +As explained by Herb +I wrote a totally new thats not the one thats supposed to be in there +Waiting on this before continuing +I found the origin of the issue i am using nlogj instead of logj but even by adding an exclusion for logj in my files my application was package with logj and nlogjbecause there were two logging services i had this issue i remove nlogj dependencies and add a new one for logj and my application is deploying normallyI have to check on the maven side for thisthanks for your help +Thanks Owen. +Marking for so that it can tracked accordingly +This is why they wrapped it as a Tuple +I am still not brave enough to use it on our production sitethough +These were split up specifically to reduce the dependencies required in an environment +The reason has been only the possibility to easily implement a Long Session with JTA +otherwise +I dont quite understand what this means +This seems to be a logj issue which tries to create a in the wrong solution place the and logj into the +Added filter for STARTTIME and test case to remove STARTTIME +ive been using min quantiles with a fairly ok results +Spring can be installed into JBDS Beta but wont be added to Central until CR +svn ci containerSending containersrcmainjavaorgapacheopenejbSending containersrcmainjavaorgapacheopenejbassemblerclassicSending containersrcmainjavaorgapacheopenejbcoreAdding containersrcmainjavaorgapacheopenejbcoretimerAdding containersrcmainjavaorgapacheopenejbcoretimerTransmitting file dataCommitted revision . +This patches two files servicelibpythriftand servicesrchiveIt looks like is generated automatically +When I had the version everything worked fine but when I upgraded the deploy started to fail +Hm I do not need clastering as well as I do not need to use +This would solve the risk of leaking infos +Updated for the new JOB creation capabilities +TRACE +This should be pulled into the branch +just committed this +Any update on this? This is more inline with bigtop as well +yup good idea +I just committed this thanks Corinne +PatchLooks like this effects a few more methods inPatch applied thanksColm. +bq +It only addresses the login user concern which would now be your userid which would have sudo access to root which is running the cluster +Since many of these improvements have already been implemented as part of Im marking this issue as resolved. +I got the exact same problem as Justins with grails +I have an with a method annotated with this method removes files from the file system when an Entity is deleted +heres the patch as an attachment +Just needs confirmation by PMC of receipt of transfer by ASF secretary +Looks as expected +It will name the snapshot directory only using the passed in value and will default to a timestamp if nothing is supplied +If new server client sends a request getting with locations back together Otherwise client first get then sends separate RPC for block location +I think there are several motives +Click Add New Portal button to create new portal +Please provide it as soon as possible +Modified title from Refactor Redirect view to have a method to handle flash map logic +My fault +now sorting auths in check auths +My use case involves code similar to Roos finders but the advise was having adverse side effects I did not explore since I believed the annotation was a copy paste mistake +Ill close this issue as there isnt much which could be done from our part besides documenting the possible workarounds +Proposed SOA TP Release NoteMTOM support now functions correctly +The patch must be diffed from the root of the svn checkoutIve submitted a new patch with these fixes and will commit once QA gives a again +Problem was in bean inheritance from not serializable class. +Ive also checked in the additions to the PMF +Reporters of these issues please verify the fixes and help close these issues +Is this really still the case with? are they still leaking their Antlr usageSandeep Are you using too? Is that why you need to use the classic parser? Otherwise why not just use the Antlr based parser? +I checked the patch as submitted by Andrew into branchesAt the moment I have a couple of nits about the UI but I havent looked at the actual code yet either +For setting there is set which is internal and set which is user facing +But for service offering it is no where to be found for a user +Unsoundness bugs are never blockers because there is always an obvious all open issues to +HelloIs there any way I can help to get this up and running? Our Tizen contributors are getting antsyCan we get an repository set up next to the other repos +should now work without archives however it will be missing the filesets section of the Servers View +Thanks Kevin +Thanks Parimal Pranay your changes are in trunk at rAshish Vijaywargiya +SergioI just did some tests with multiple boolean fields in the domain object and can not see the problems you are describing +Committed revision in Joan for reporting +Also does the accept ACCESS as an option? By looking at the code I dont think so +Committed to TRUNK +Thanks! +resolved togehter with the support for finding by this a duplicate +According to the Fix Versions it should be contained in and +Removed the warning when annotations were available in private fields +Let me know if you had the same problem +For the client side we are already taken care of this in which will be fixed in PigFor the backend side there is no as hadoop does not run tasks in threads and as far as I know there are no plans to do soIf the need does arise I would be better to open a new bug that addresses that particular use case +bq +Dont have time for this one now feel free to grab it +In other words since most monitors are landscape and people are starting to have more then one of these on a desk side by side as a possible way for development makes more sense then top and bottom +It is only the resultset that is closed +But it cant tolerate exceptions that caused by JBM itself +Does the plugin retain comments now +I feared that something like this might happen +The point is that the messages which are displayed to a user should be tailored depending on whether they have successfully entered their credentials or not +Its running on Fedora but httplocalhost returns a blank screen +ok based on comments it appears that allowing +Lets implement using my initial suggestion which Lars and Stack concurred +Thanks to Jon for enlightening me. +fixed in +A patch for those final two upgrades to Camel and Specs would be helpful +Instead we take the method name base approach +Feels like thats the right model for this too conceptually +I think this is beyond the scope of the bug report +Good on you Lars +So yes when you close your session the cached statement goes away +I committed the patch to and +Code looks excellent +is this something more than variable declaration +Marking as cannot reproduce +It looks like this JIRA was accidentally closed +Hes getting me some details for how they diagnosed the problem and hopefully that will lend itself to the solution +Willem do you mind looking into thisREST support is important +Maybe XZ needs to be changed tooI think that in most cases concatenation support is wanted +This patch solves this issueNote the issue was initially raised by Mike Kienenberger while trying to integrate Velosurf in VPP +Applied some time ago I forgot to close this report +Thanks Liyin +Yeah but doesnt reload dependents either + integrated with trunk. +tnx for this still an issue? +Which is why we never saw it before. +Ah this may be a bug in store vs +Nice change JingNit mark a block as corrupt should be mark a couple blocks as corrupt we can fix this on commit +Could you please remove redundantimport if you changeI think this is the right fix for the time being although Im not happy that we should keep different identifications for the nodes and that we have different ways of node identification in different components of hadoopMy proposition would be to return back to hostnames instead of ip addressesBut this of course belongs to a different issue +Sorry for the +Being built at binding time would leave just one entry in the outline view +This is blocking a release candidate for if we can get this in soon Ill start running through the release process +depends added to clean target in revision . +the same trick use in can be done for the configurator +already has that manage logj log levels +In the constructor it should beinnermost +Committed in . +Fixed on trunl revision Fixed on branch revision Fixed on branch revision +Hi MarkYes i too noticed the problem + Upgrade to Alpha Rename of Arquillian SPI Update to Alpha API changes Update to use Maven Resolver Effected Modules gettingstarted Aslak +What does that mean Was due to a broken custom loader +Hi Surresh +But file might not exist ok will return default values in this case +log or EDSSOA files +Lets do some more testing but I think its ok +perhaps i should write down everything i think about the next time o i was also talking about the two +Id much rather see us focus on writing enough tests that it would be an issue +In the serialized log splitting case is set to true +Committed at revision + Looks good + +It is very unlikely anyone else would do that for you +back port this fix +It would be great if we can avoid remote reads as much as possible +Using multiple criteria for ORDER BY was what was breaking + released issues. +Im not volunteering but is there a way that the form tag could startappending an index number to the forms if there were more than one +threads quantity threads quantity threads quantity threads quantity threads quantity PASSEDCtemp classpath +Ill change the connector Id to and modify the example descriptors similar change to devstudioproductplugins in trunk +Which list is the correct list of initial committerswhats listed in or whats currentlyin the tuscany group in +FYI the issue for the migration was and github all showing same commits with last commit as of hrs ago closing ticket. +on the patch +Update patch against latest noted by chris the is not related to this patch further passed on my desktop +Could you please point me to the jira tracker where this belongs to? +Thanks Mahadev +Im not a database expert I just tried to achieve stable work of ODE with Oracle and it is tricky +Think about a production environment thats running months without restart +I was taking a look at patches marked LATER it seems you put alot of work into this patch +fixed with r in branches and r in trunk +See. +But Im not totally satisfied that will fix this +No also AUTOREWRITE uses the new +What would be the relevant partsIve posted detailed steps to +I have difficulties to find a meaningful name for these alternative implementations +I recommend instead opening a jira in MAPREDUCE for the port to trunk and referencing as well as the bug number in the commit message so the commit shows up here +I dont think a video will help +Thanks very much +I have just looked at the code on Git +I am not seeing how that can be done dynamically +This needs to make +Also any chances to get it in + +reopening this to let in further +Its like files stand in directory while node is running causing folder removal to fail and other tests to block +The piece of code that reveils the error has been changed recently +verifiedThanks for the quick verification! I just committed this to branch. +The resource switching based on locale is already in the code so the new patch just needs to be committed +I also added an additional optimization for next row case too. + +Where would you put the warning explaining message +Im not seeing any parse exceptions just an extra in the output +Progressive enhancement then goes off and in the java script adds in more functionality piece by piece based off of what the browser supportsYes the top level view does need to be able to help with the initialization but I think a lot of these problems are coming from trying to delay instantiation of a until it is rendered +Committed lets make further improvements in new issues. +Any chance someone can test this? +So what do I have to do now? Write a? But the wiki further states Furthermore there is an alternative configuration module which allows to resolve configure values from or via a custom from any other configuration formatsource +Thanks Arun! And sorry I missed this out after that review Tony my apologies. +Did not get fixed in more planned releases +So the user adds a logger at the DEBUG level but that logging does not appear in the log as the handler is set to INFO level +I think were doing this +The error is gone in XP +Proposed path to change so it retrieves Digester instead of +Sorry I forgot to add the in link in the nav bar +Put back in the to . +Some characters were slipped into the previous patch + +Richard do you think we can mark this as solved too? Thanks +Let me know what you think +Im moving this since I dont want to hack around it just before we drop jbossweb in favor of undertow where it may be fixed or require a different hack. +Will be in JPOX builds dated onwards +This appears to be an issue with encoding +For example the chache and another patch for the global error handling +Im not trying to rewrite the whole test system just clean up the tests +phabricator is also updated +I dont mind changing the modifiers of the get method to private static +I agree with Teds assessment +Is is OK to say that checkpoint would be same across all dirs? If so here is a patch which forces saving of image if there is a difference in checkpoint times +Something is causing Tomcat to use JSSE when run as service +added new patch that closes the about this? Super simple +Thanks Sanjay! +That way the log files of the jobtask contain the reason for the failure +MarcoSorry I did not check before my answer and I see your seed data know +Group by extension queries require significantly high overheadThus the query optimization especially for the distributed plan is very importantStatistics such as histogram are very useful for the query optimization +jenkins is green looks like an unrelated failure +stdout running ant test after the change +Fixed in mind itll be the definitions on my server I guessgoing to find out this evening +Im going to wait for one of the other committers to come by and also give this a since this is a substantial change +Reverted as it should not be in the source tree should be in dist. +Christian hold your horses on this oneThe JMX is scheduled for a total overhaul in CamelYou are still welcome to submit a patch but we might not commit it +For various reasons we dont currently default to itThat actually seems to work much better +See my comments. +Sorry for this long comment but I think componentization of web apps is one of several drivers for using SWF +and scheduled as Luke has a valid think this issue is a bit messed up right the beginning it was asked for the initial file +txt files as SVN tried to add SVNnative +Manage Portlets and Gadgets but also in section +Updating description and priority +However the Jackson JSON library also flushes the response internally by default a behavior that can be turned off but the response may still be flushed when the response buffer fills up +jvmRoute is set now +Committed for in rev +You right it does not look like Graphene issue will need to play with that deeply and then I will see. +However we can never send the JOINREQ to the coordinator because it is queued +There are some extra tests that demonstrate where parse exceptions are going to be thrown which is niceIt would be nice if we could perhaps be more liberal in what we accept +changed patch into baseline and opt +This would take care of the case for example when the network connection goes down and comes back again +Simple patch to escape the row key when it is printed +I would imagine that is where it had wrapped +Ive opened for this +Sorry about this but it seems I wont be able to look into this issue for next few weeks atleastI will submit my patch once I get free unless someone else had placed it before me +It looks like the call to s Enhancer at line throws an because my pojo has a private empty constructor +Thanks +No problems foundTested with svn revision +To add on that the fs cache isnt the best thing to rely on + +Attached project used for replicating the issue with Spring as invalid as discussed +When we started implementing inner classes in Groovy we said they would more or less get the same semantics as our open blocks +Thanks cant figure out the scenario please help recreate it with some samplethanksdims +Im going to do +I individual mailing on all jobs. +SergeyEven if you install protobuf locally tests based on hadoop would still fail because hadoop and use incompatible protobuf versions embedded in their RPC engine +test to reproduce added a note to the docs about this clash. +if I recall I dont think it was anything serious and perhaps something that seemed intuitive but wasnt and I had to work around it +For still dont want anything there either as were still discussing these things in JSR +Fixed Transactional is both Inherited and Documented you please provide sample snip of bean configuration file for transcation using annotation? I cant really find any sample any where in the net or books +I already added a implementation to +unified to the current trunk for customizing servicename and as part of +There is a different fix for this in CVS +Only localetrue causes a session to be created +The name finder training always stops by itself after around iterations not sure what is the exact number since you changed the reporting +So we could omit tests but we should keep plugins features and doc +Get a thread dump using kill +Verified this by running findbugs locally +Of course if you dont use cell then its not an issue either way +Drools suffered from at least infinite loop bug +There is no new code +Think this issue is invalid in a the way that tablestate saving is probably not configured since grouping also is wrong and even column width +Ive revised both and accordingly properly supporting REQUIRESNEW transactions in afterCompletion this fix has side effects Ive removed the backport from the codebase +So now when I get this I should not care as its rather a rare case +So I wish to strongly support the plea to remove the unnecessary requirement to put the parent version number into all restriction can make quite some trouble with version control systems +Im pretty sure this is an improvement that everyone will appreciate +Actually makes almost no sense in Spring Batch +FWIW +I did some additional testing and it looks like this issue only occurs if the service scope is set to application in +Ok to make a final decision +Basically we sign up now to allow certain DV type changes in the schema just like how we allow omitNorms to change from false to true but not viceversa +I just checked using of the plugin and this patch does not seem to be applied +We are going to continually have these discussions about what we can and cant do in CQL that we can do in thrift +or is the new lookup that takes the fate OK since the core extending the fate should work in the new methodIt seems like the new method would be ok with the proposed extension +moving this enhancement to +Attached is the binary files + XML file is not good for streaming through Chukwa because the data are ordered by line sending small bits of xml over Chukwa collectors in parallel do not guarantee the original ordering +I could not come up with an elegant way to handle constant detection and therefore a clean way to determine the Busse R squared +No Xavier I am no longer able to reproduce the issue +modified for the existing unit and integration tests to pass +Please verify and close +Hi appreciate your contribution on this issue +tests and basic documentation added +moving it to this problem and another problem that stated that watches get removed on disconnect +I think the consumers should use the Async API by default +Thanks WillemDavid +you might be right on the second one +This is a severe restriction +A typo I do not understand what is the idea under partitioning the set of commands into Administration and User groupsAll administrator commands are called using dfsadmin so these are real administration commandsMay be you want to change the name of section or just remove this level of section hierarchy making each commanda section rather than a +run quickstart and click on update link +This patch creates a class mapping a block to which contains a reference to the Block +The collections of field names and stored field names are exposed through getters +which browser +If this is still happening with please with the and a of the console log attached. +It is a bug +After the tests are done this dir will be deleted so that no temp files are left behind +Dynamic imports are not considered by the resolver at all they are a issueFurther since your bundle imports it is not eligible for dynamic importationThe issue is the bundle in question imports version but there is no provider of this package with that version +using Map interfaces +What about s sharedlib setting? Should it support the same syntax options as the solrconfig lib settingHopefully I can do some testing too +Updated config +Committed to trunk and +I need to provide a new revision of patch as I forgot to modify +It has its own mailing lists and jira instance +I just committed this to trunk and +Attachment has been added with description patch has been added with description list of improved javadocs refactored in parts +Are we going to have a separate INSTALL file for this +Hurray! First commit +The failures are not supposed to be fatal they just result in the server printing a warning to +I do not see you listed on the PMC for James if I am wrong please feel free to reopen the issueRegardsTim +The patch file has been attached +Other inner classes with unique names could be extracted as is +Other than an deprecated tag what is the problem with this issue +I would not be in favor of changing the default blocking behavior I think the exiting default is sensible +Marking as wont fix as we now question the usefulnes. +This is the expected behaviour +Im also against the idea of having version rename tool +Scott I too couldnt see any relation between Validation annotations and +Minor corrections in the patches +Im reopening to fix the JAVAHOME issue and commit to the branch +By wrong path are you saying that +Change to the interface which should have been included with yesterdays submission +This patch fixes the issueNote should be fixed first +It seems running on only single node +sounds looks like was accidentally merged into trunk +Nothing happening here so moving to wishlist +regioninfo since the table name change +By default release mode is used for federated build and +Most of the time projects are extremely badly described in terms of dependencies +for Daniel committing +Bug is marked as enhancement but I agree with you +Closing +Should the search return only the in this case? Or does the application need to pass additional info to disambiguate the search +Daniela Is that from a separate test youre running? I dont see that in my directoryb What is your command for running it? Are you running it locally or in mr modeIm just curious why we get different results when running + to trunk aededeaabfdbbdfc +I am not sure if this is the same issue but I also have a problem with converting pdfs to images +See servicecomdemo and controllerscomdemo +pull request thanks! +Add a intRange and floatRange functionality. +Patch for based on v from to go methinksWe can remove the bulk assignment code in another jira or not bother with that since fixes it +See for further details +How do you start Jetty I mean who you know the OPTIONS value when starting JettyI thnik CARGO uses OPTIONSServerjspjmxresourceswebsocket I spotted the difference you are using JettyCodehaus can you please try with JettyEclipseD Oh jeez never would have thought that the zip at codehaus could any be different from the one at eclipse +Joe Agreed +Ugh! FieldFieldable strikes again +Im working on this enhancement now and just about ready to submit this as a patch +Looks good to me + +Will be in the +And I think it would be better to move all the codes related with graph data loader to unit test +its fixed with attached would be indeed the best +Please review +Will commit unless voted down +Fixed by the BDC guys +Please also follow the general guideline for patch submission do one solid patch for issue relative to VM root not some directory insideI hadnt looked into the patch yetThanksAleksey +That caused version conflicts in rails apps +How do you install pdflatex on Ubuntu +Patch ping +Integration tests already exist +How specific do you think this is to C? Do you have any interest in documenting this as a cross language API guideline +This wasnt in JBAS so it should be prioritized below getting feature parity in the JBAS adds support for installing jar libraries into the EAP lib folder +Attached is the I have deployed in is not a bug report +I like this patch and forgot to vote for it +Patch fixing the issues listed + patch with additional +This is not a major issue if each map task is run under a new JVM but it might become an issue if the JVM is +BTW if the solution doesnt work for you we can definitely consider making that field protected. + +Changing this to an improvement as support for or relationships is limited at present +A further thought on pooled target objects Those would have to be conceptually equal to each other in order to qualify as pooled objects +Native code is designed in this way +Thanks for the fix Nirmal that was fast +Or if we went that route we might as well add some instructions for all examplesarchetypes available out there +but for the branch I guess we would have to check also +Just got back from vacation +Not for commit here +Or do you have a column filter configured +I think were good with Nicks changes my additions your simpler locus build +If you look at the Serialized Form of and you can see it gives away an implementation detail which is that they both derive from a +xlsx files as wellRegards UweI read and and saw that they really solve this problem but only if you output XHTML +The default mechanism seems adequate at the moment you can contribute anything you like you just need to convince a committer that its worthwhile +I took another look and apparently I had looked at but missed which also uses the c +It took longer than it did before so maybe it helped but that may also be coincidental +I dont think we should move jmocks jar to some central place + +bq +Possible blocker for release +But ratcheck gives me an error for and +Treat unscheduled issues as Native SOAP stack is not supporting JAXWS any more +No as this is a standard use of an existing Sonar API. +This change should also take care of the hbaselibs introduced as part of +This change is already in as part of +All good +Removed incorrect copyright statements +Can you please attach your patch as a SVN diff file against Axis trunk? +And +But using ! to say go boom when null seems really unnatural to me as a Velocity user who is used to ! meaning this can be nullHonestly i think it would make more sense to have foo throw an exception when undefined or null and have foo still throw an exception when undefined but render empty string when null +Awesome thanks +Actually for some reason this feature has not considered when codegenerating and the extension was not called in order to set the package name in +Martin could you provide an example BPMN of the problem described? Thanks +For example step can be probably done in file TOMCATHOMEwebappsopenssoclasses +I dont see this on any comboboxes on the server and runtime selection in New Seam Wizard do they have the same issueIf yes then this sounds like a feature of the GTK theme you have enabled +Without it its nearly impossible to figure out why it might be useful +Available shortly +Thanks Jacques for committing the code. +Restart JBTERROR server is running but status of server is set to stopped +Let me write something up and post it to today + tested! +this works as expected here +Right now it sounds to me that the problem is with the way youre injecting faults +Fixed with blockers and critical issues are resolved but not verified +if thats possible +All completed successfulattachment MS log MS log DB server +Here are the logs with lines of context around all the ERROR to me so this has an owner +Thanks +I think this problem may appear on Linux too +And as Shai says if style compliance is going to be a continuing issue we should agree on a style and automate compliance checking. +Username schaefera +I dont think this will make it for Final if we dont absolutely need itI agree it needs to be solved but more and more keeps getting added to Final and its adding up We had originally said that we would only ensure with Forge deploy against AS now were talking about making sure Forge works on every Archetype without any extra config and I think this is a bit much creep for FinalHowever if you have someone that has time to work on this for Final Ill be glad to assist as much as possible! If we really want to hack it we could add a facet override that allows manual installednot installed settings for individual facets by name +Will respond to other points but the intention of the timeout was that it was milliseconds +Id expect it is required +Looks Good To Me +Hi +Good pointIll have a with three different properties in revision I also cleared up the add remove change handling which was slightly broken forthe rare case that an add and remove event for the same path arrives in one set +Trivial patch +Humm +I dont think this is a dublicate since refers to an older version and is marked as releasedI will try to use the latest version of the framework +If the user pass this parameter the batch framework would generate the timestamp as +Please use the use some deprecation policy before remove or rename some classes or are enough addons developed +Should also mention this behaviour also happens on iPhone details after further can also comment that we are experiencing the same issue +adds CORS support for components and removes this be seen as resolved? I think it was tested intensively by Melanie +The exception here is from the script not from the export logic + committed to both trunk and branch. +Please feel free to contact me if you are interested in my proposal +The message in the exception can still use the utility to format the messageI am leaning towards the second option +Fixed on +First attempt at a patch +resolved via +Do you think we might remove this feature? The problem I see with it is that its just not obvious that these constraints refer to the number represented by a given String and not its length +Running tests +fixed at r +Behaviour updated to prevent NPE while logging and abort binding recoery if the exchange doesnt exist rather than continue and have the binding factory use the default exchange instead +I have committed the fix without changes +Fixed +Verified in ER +This patch supersedes the original file by enabling the code to continue using but simply appending the extra context info before invoking the parser whereas the original patch bypassed for some cases +Granted Tobias administrative karma +I pulled this change into as well since it was missed when was merged recently to +Assigning to Laura as she is the new lead for EAP docs +during locally builds +context based help were never integrated into the build thats why +We can implement it for a couple of days +Lets go ahead with to accept user and groupsThe other public API that uses UGI should be fixed later +Attached patch +I will get to it a bit later +javaengineorgapachederbyimplio has two references to +This script can be used like the script to execute example classes +He feels that the current MR behavior is correct and we should just document the differences +it is hard for svn to get track all the changes after merge thats painful +Hi AnkitIm just waiting for the final version of jsTree to integrate your patchCheers +But this intermediate source file now lives in the generated source directory +The is what should be used so let merge the definitions to it +For document style we should allow null as parameter and return value +I agree with the sentiment of this report howeverthrowing an exception seems wrongReturning null is probably valid altho Im concerned that some client code will start breaking with this fix +Confirming that fix works in. +I wrote a simple servlet to check what was received after submission +LatencyTypeMeanSDththththththPSnannannanBBnannannannanThe indicate periods where nothing happened likely caused by CSLM garbage from the higher column countsh +Documented as a known issue in the EAP Release Notes. +Call the getter for the association inside a trycatch block and in case of an exception assign a null to the association. +Fixed in trunk and branch OlegI just ran into this on for http core again did it get regressed +HI ClausYes this fix is built against the cxf component in Camel version +It seems I have the same problem +This has been fixed with the use of the Widget in the Source Code page of the Schema OC AT editors +no worries erik! +Yes it is Thanks Rajesh I did a JIRA Derby search on on all available fields before creating this new JIRA and it had returned no matches sigh +lets apply the patch in to +ClausI just submitted a patch for adding an assembly module for guiceyfruit +in M +Works great! Thanks Fil! +So I am guessing the problem is in network server not using the query timeout set initially +Is that from Jakes branch? Im kind of surprised if you didnt need more than a recompile +As you said Richard an API change between and +patch v +The files were sorted my last post I used the option srcFiles +Regressions ran ok +This relational operators are working and not an issue i forgot to prefix the effectiveDate with an at +So this issue is going to be open for at least more years +Attaching another screenshot +pressing submit too early +I will try to write a standalone driver to generate this xml which can help us verify the fix +Patch contains only js file changes +The patch looks good to me but I agree with you on the name +The Dusername parameter or the workaround proposed by Niclas is not working in as far as I can see +This will probably not go into the next pdf release as we have to wait for another doxia release +close off release +Updated filelistD jbideployersrcmainjavaorgapacheservicemixjbideployerimplM jbideployersrcmainjavaorgapacheservicemixjbideployerimplA jbideployersrcmainjavaorgapacheservicemixjbihandlerA jbideployersrcmainjavaorgapacheservicemixjbihandlerA jbideployersrcmainjavaorgapacheservicemixjbihandlerA jbideployersrcmainjavaorgapacheservicemixjbihandlerM jbideployersrcmainresourcesspringM jbideployerWith this patch we can install jbi artifact with jbi protocol url +Now mvn clean verify runs fine +I am going to commit the jobid part of the patch first since it is a very useful feature +That is certainly a case that could cause a problem +Got some local compilation issues with the query builder in contrib so assigning to Grant to get this in +I was thinking we want a configuration file for two reasons +Patch seems ok Does it pass +Correct +We I have customers who are using the workflow you mentioned +Reflect current trunk +When the background computation thread wakes up to work it removes blocks in pending from the namespace if there is any +Assuming a coding error in your application leads to a deployment failure marking the deployment as republish needed would and has confused users into continuously clicking publish expecting something different to happen and wondering why it didnt occur +That makes applying patch brokenNow it should work +tested using q s as well as via JMXShould we warn users that JMX monitoring requires a Sun JVM? +Its should not be a race condition +it looks to resolve the issue on my local build +Thats why I said instead of +If I patiently wait for the projects to finished checking out of subversion I never have a problem +I think the packagestructure could do with some revision while we are at it +Since you included test code that demonstrates this bug it shouldnt be too hard to integrate to an existing JDBC testThanks bug is only checked in on trunk +I have a strong feeling that the supplementary channels would also relate to RHEL as well +Now sure how though +Any chance this patch can be cleaned up and committed to trunk soon +In our internal Jenkins build which uses hadoop there was no such exceptionI last merged from Apache on Nov +Can you explain in more detail what you mean by incorrect values +Ill also get the related revision number inCheers to show ofbiz version for and hierarchy. +Committing this shortly +Is there anything more we need to do here or can this be closed as fixed +This is something we have to do +I needed to add enableassertions to my test case java command + for patch +I do not see these warnings with trunk itself +Attached simplified files that demonstrate the failure +we replace objects while serializing them so that we can replace references to pages with pointers to the pagestore so we do not have two different instances of a page with the same id +Patch added +In the one case my VM crashed and I attached the hserr file +Its a well known fact that when you pull off the cable the socket will only expire after minThe manual also explains that TCP is a reliable protocol hence we check the If the user requires faster failover in the advent of power offs he can just setup faster pings etcThis wasnt an issue of any of our current customers +We can even improve this and only record the time of the first unlogged edit made + +The monkey patch onBufferingdowrite is working for me so far. +So there no shared config since its another process when run on a real cluster. +Chris you can also see that table with is broken as well in +Yes I think the patch is correctI also confirm that wicketstuff issue is solved with it +Could you comment on why you want to enable redirecting +revision +But Im not sure if this is the best solution for this +It would be good to see this fixed +Committed to branch as well +done +Heres another patch that removes references to pk from deleteAll +There are still issues opentestCopytestThe complete list of failing issues is in the pom of the project +Dont know if there is a way to configure the externally to utilize some of the functionality of the but if there is id love to try it out +Attachment has been added with description example rounded corners I have made a few bug fixes since I uploaded the patch but I can provide them as a new patch against this feature branch without any problemsPlease go ahead +updates +I tried to change the text of these preferences to Default Maximize Source Maximize Visual but the change broke older have more than VisualSource and Preview instance if a user chooses Source a new file will be opened with the source tab maximized +Aside from this fact theres a bug in that the plugin is improperly taking commas and breaking apart the URL +zip once I have reopened this issue +All scripts I provided here I grant rights to Apache as per Apache dont think this is a newcomer issue +Moving all issues to the CR +One more quick comment looking at the patch please add brief description to every test on what it is testing +The current default implementation is to ignore the referralIf one add this property to the request both tests are passingI guess that we should change this behavior to default to THROW instead of IGNORE +Could we call it something else though? As you describe it really could be applied to other types of objects +There hasnt been activity for a while I will take over if you dont patch +That is because to get the to be instantiated and populated we need an empty constructor on the instance I am extending +After some thinking the relationship between and should be but not +Im not sure why as touch is definitely ending up calling utime but the mtime doesnt seem to get updatedprobably just needs another short debugging session to complete the patch +Its the async routing engine that waits for a reply that is not coming back +Rerated this is not major +defer to +and remove the other why did you resolve this issue? Nothing had been done with it yet +Would suggest we do it +Closing stale resolved issues +Unscheduling due to inactivity +Are you guys still interested in getting this done? We can take a look at how to get it into the main spring repo which is automatically syncd +on ignoring write for now +After some discussions weve come to some decisionsDo you mean the discussion above or some other discussion? Decisions are made in public +Is committing this enough Andrew +Discussion on JOIN performance also to the forum +Particularly since to tackle the underlying issue will be and most likely introduce a new crop of issues +Lets keep it open for one more week to see if anyone can think of anything additional +closing this issue. +My own basic unit test works however does not and it is probably due to the file pointer not being on the correct position during enumeration +This problem and the attached patch are similar the previously fixed and its patches +Once we got a finalized plan i will get the cpulatency numbers +we should move this to make sure to submit a new JBPAPP issue if there are any other diffs discovered with new has been resolved and verified. +Thanks! +The NN creates a daemon thread of itself and implicitly relies on other threads to keep the process alive is this a case of two wrongs make a right? Or is there a technical reason why the NN shouldnt simply do its work in the main threadWe should follow the same pattern as the namenode and wait for maint threadthreads to end +but in the all the responsewriters are specified with FQN +perhaps we should just treat data from listscans as though it were results from a table scanI like it +occurs at the begin of the sentenceCould also be caused by encoding issues +old enough to close +Just out of curiosity again +I also thought to remove usage of Class from Class but never found the time to do soI only looked at the patch yet which gave me the impression that the changes are quite big +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Moving content from Moin in progressth Oct almost complete with formatting changes nearly there +Not a blocker pushing to it to +UpdateEverything works fine with the work around not using redeploy target +I hestitated to port the fix to because Dan had some concerns with the fix +I presume the issue has been fixed as per Cleberts comment +looks like a marker for tracking the related issues and incorporating into Continuum +Sending srcmainjavaorgapachecommonslangSending srcsitechangesSending srcsitexdocSending srctestjavaorgapachecommonslangTransmitting file dataCommitted revision . +PS +There were quote character issues actually Our installer displayed them similar to what you saw +Therefore this ticket is going to be on hold until we add a configuration to the time machine to choose which metrics should appear +This probably wont make it into clearing fixVersion +But apart from this convention problem I dont see any other reason for depreciating this method +What do you mean by? my mind was thinking and i ve written +Usually they have two or three items in common among the first recommendations +Thanks. +Thanks for the cleanup Lars. +I am testing your modifications right now I hope to let you know good news ASAP +The Patch is by but since he is on vacation I took the liberty to add it under my name +Was unable to recreate the issue but I think this should solve it +JacquesI will test this tonight and get back to you tomorrowRegards JacquesMy apologies for not responding sooner but other activties prevented itI have tested the new patch and it checked out ok with derby mysql and postgresql as db backendsSo you can commit +Both you and Ersin are Confluence Adminstrators now so you should be able to handle other administrative tasks from here +Thanks Joshua and Ayende +Hi Jukka the repos can all be closed down now +Proposed project pom v committed in rev +done just miss the announcement on the and karma added +rar ran the attached sample through FOP Trunk without any problemsWhich version do you use exactly +applied in revision Anton please check it was done as the patch is applied as expected thank you +When I deleted all the content of this location CDocuments and SettingsmyUser +I can now drop a new into a running inventoried App Server and have it picked with the descriptor field set correctly +Default will be but can be overriden by each DB Dictionarys constructor +With this patch all tests passed +I have checked in an update that changes all the source build doc file headers over to the new requirmenets. +Were also running the full suite of tests in parallel +We may as well fix the broken tests now when theres a patch that applies and passes and worry about style separatelyOnly some of the tests this is updating are broken +Its most likely because you need to subscribe a second address +fixed deserialize and to remove thou shalt not stream across version changes special case in from bc of +I was mostly curious whether they used some sort of utility to do it or if they rolled their own. +If its not at proton your encouraging proton users to implement it themselves in potentially incompatible ways +Patch committed thanks Aniket +Attached I attached the patch to and have verified with the unit tests and a deployed guvnor +Im not ready to comment on a release date for this fix right now +Heck why shouldnt we store the entire config in zk and the only config you need is the address of the zk clusterYeah long time plan is to put all config +Patch is fine +Verified in version +Yes Ive verified the functionality in HEAD +JIRA is displaying the patches with incorrect timestamps for whatever reason +for the patch +Fixed in rTranslation of field access should be applied only for the fields of thisGuarded translation in Java backend by checking the target of field access +Sample page after clicking after the initial login +svn ciSending Transmitting file dataCommitted revision +It is just an inconstancy so what should we fix? The javadocs or the default +Passed my test case but want others to verify +Sorry for the trouble +Issue can be closed. +javabin transport +Please submit an updated patch +With this in place you have geospatial filter sort for indexed points +Youre working on writing a schema extraction tool +It seems that a feature can not be deployed as a jar +is quite reasonable IMO +I have this problem with externals that implement update semantics +Ah perfect +This issue is only interesting if we decide to change the reference architecture +It sets the version information etc on the events + b Another Server thread that waits for pending commits to be synced and replies back to clients +I am canceling this patch because it is sadly out of date and we do not yet have consensus on what to change +Thank youvery much. +Take a look at and then to and and you will notice that the call to saveState is deffered in different ways according if we are using client side or server side state saving +From an architectural perspective Id opt to move perform to an upper layer +We have to add an execute call which will return a immediately and keep running the operation in background +this was applied now resolving +We should although check we cant delete this attribute +I committed the patch now in revision +Note that SVG isstill in the Working Draft stage +Id like to improve collection creation outside of Solrj as well but I think it also makes sense to improve this here as well +modifying changed every dependancy to +please try sth like thisselect b +Sound too much involved and too late for +By incantation I mean bundle package bundle install local pathvendor S rake torqueboxfreezeaccomplishes exactly the above iff a Gemfile isarchive implies torqueboxfreeze also +you will never encounter a partial write more than MB from the end of the filebq +This patch adds the package and tests for this class +The patch hardcodes password and passphrase to be considered as secrets +Ill check with John but I think we should just remove libev entirely from the source. +I dont think theres anything to be done here +ChrisAs I said you on IRC I agree with you one command will be better than two separatedRegards you think you could rework this patch to combine both commands into a single one? +The request might get larger but this is more precise +And to take Einars example into account this will only happen if the value contains exactly quotes +Reopening to detach from Backlog +Did you find your patch +Enrique I found the fix but the file loading problems still seem to perplex me +Also had to add a fixup addendum to fix test build +Screenshot Dali does not like it when we overwrite says its fixed in WTP +I also did not take over your changes in the Generic type of arguments of methods. +This same issue is happening again with release can I this issue to be looked at? + +ex if RW added the ledger to pending replication and client reformed the ensemble immediately after that +studioeclipsefeatures +conf tried above two method all can not work +Thanks Gunther and Jason! +Heres the updated patch with q testing +x and then merged to x just that small fix though have not had a chance to review this patch fully yetBut is it also going to be backported to lucenesolr which is actually the branch I am working on top of +add storage type to block placement +No its not we should move Core JS resources such as to resource library and resource packaging +Attached is a patch for karaf which works for me +I will put it there as soon as possible at least for using it in in my local copy of the trunk +Server log with deadlock to Charles to take a look +Previously mapping was done in the main code +Now we have to read inside so that processes dir is placed into correct path for working +I will make the changes in trunk after GFAC is done +Ill add a test +Note that Soylatte JDK appears to work correctly +I was able to reproduce it with but it seems like its already fixed by one of the other issues since now I dont see them fail anymore. +Attachment has been added with description Patch against revision +Verified fixed in BRMS ER. +The attached file is a patch that fixes the attached file is a patch to the patches in r +Resolving as not a problem +True not necessarily a blocker +even though I dont think it will stop the app it makes sense to allow easy access to the startstop operations +If someone does let me know and we can reopen. +I guess that would work as well +i actually added a menu for +What does the say about levels ofIn the above example if has been tried and then and there is no should this go down to and try with +Fixed in +Changes in dbcp to use proposed implementation in pool +I think this is a duplicate of +One for automated found bugs in sonar then manual in Crucible and thats not what we wantOne system to rule them alldasamuelsson on good thanks +Its deployed as a war file +Its still present in M and probably has been there ever since we moved to git + +Thus in an hour or so a new snapshot should be deployed +Matt please consider this for +I can set one set of servers to LFU and the other to LRU but its production so the two sets of servers never receive the same queries and I dont really want to try any isolation tests on production equipment +Modify patch to work with r +I remember that rule for line breaking in Japanese was very different from that of EnglishIn English sentence there exists space between words andthat space tells where the line can be breakedHowever Japanese sentence does not have such a useful space between wordsAs I knows only punctuation tells where the line can not be breakedWell +I am not sure what sql standard says for such scenarios but atleast mysql instead of throwing exceptions does correct math and returns valid results to users +had failure which was not reproducible when it was run on a standalone basis +Please mark it as resolved +I finished the code for this +Namit just committed which should fix the test errorsCan you the tests +Just need to get that assigned +Hi +Thanks PeterIf we can get someone who knows more DIH stuff to add some tests we can get this committed +What was the reason to provide both delays in DELAY +The installed js is not being detected if the js headers are in usrinclude This is the case for js in RHELCentos This even fails using usrinclude branch aeechecking for JS +Ill file a separate issue for mode being too permissive. +bulk close of resolved issues. +The parent would point to the first child node That would only work if we adjust the contract and require the MK implementation to do thatAgain I think there is a misunderstanding +I am out of office from th April to th April I will reply yourletter when I come back +Thanks Manish +I like this idea +Also seeing the changes it does not introduce new states nor changes the order of the statesJimmyI agree that there may be some cases hidden +for this reason can we call the rule secondary +Ive often been building several projects in several windows simultaneously so I wonder how maven handles the conflicts +Removed the patch containing tests and +This was obtained by fixing up the log directory location in adding the following property to and then connecting +We should make it consistent and to allow for conversion will make groovy much more powerful just like has taken out the hassles of conversion and casting from the script users +Id like to commit the patch +So the TBD actually means that we have decided it needs fixing +do not set the specialized Partitioner +The patch has been committed the new revision number is +I presume thats clear enough +Ive created for your last point Evgeny +Closing resolved Release +This version of the patch includes a spec to test the bug is solved and just uses the glob method when we ask for a jar file otherwhise it uses the current implementation +For JIRA can you give me an issue prefix email address of an administrator and an email address to send notices to? Thanks +Is this ok with you Francisco +I am not yet sure how to do it exactly but the basic idea is to let the endpoint start up without waiting for the servlet and let the servlet later pick up the connections +for latest objectcreationmaybecomearealprobleminabigclusterwithmanywritesifthispatchisappliedIfwasnotaninterfaceitwouldhavebeenagoodplacetoputthiscopystuffsinceitsonlyusefulforMROntheotherhanddoingitincommitpreventsweirdsituationswhereauserwouldhavereusedthesameoverandoverandonlythelatestupdatewouldhavebeenappliedmanytimes +Attaching a patch which takes care of the NPE +So we should process such string by harmony code or mark it as diff with RI +That is the real bug hereHowever Yash Talwar is looking at adding support for XML to +QE please for pointing out the duplicate issue Pavol. +are welcome to the issue provided you submit a patch +patch uses instead +Thanks for reporting the issue +ERROR fixed +A patch fixing the issue +Any progress +Updated the connector to use the old metadata scheme as a fallback supporting +is this because of java compatibility? This could be clearer in the description +I havent looked at the changes in guys but I dont want to see a new feature in a branch which is replaced in a +The issue I described uses two top level http requests which deadlock each other because they depend on the same resources which they acquire in a different order +Haha thank goodness for JIRA I am so glad well be able to easily track this issue +In something like though it is a significant amount of the time spent when requests are doing small reads out of cache +Second i assume search may take fairly long +Make a third patch with all of the fixes. +on patch +Maybe someone else is depending on this nuanceD +Cluster access tokens on the other hand are longer lived and are used to represent the authenticated user and the event of authentication itself +Now the test should be successful +when the datanode restarts remove all blocks from the tmp directory of the datanode +This is a new bug +but this seems like shorter path +But all the other should be fine +the machine is joined to the domain I think you might need to use the complete machine name rather than just the BIOS name +Sorry for breaking it for you again +Tested with wrong endpoints of clients +example should work fine with the +Ive attached what is probably a naive attempt to fix the issue Iveessentially copied what in the main Velocity package does +Any feedback on this one? Will close as cannot reproduce unless some way to reproduce the problem with the current sources is all issues that have been in resolved state for more than one month without further comments to closed status +Clint I backed this out for the +I couldnt make head ornor tail of it +This is not such an unusual occurrence an unexpected power loss for the SNN would cause exactly this situationNote the following + +Things like are used by pipes but only part of it are used by pipes so ripping out the unused code I think would be preferable +Still need to integrate file name line number into the parser tree +It will add complexity to helping macro users on the mailing list as it adds another setting to ask them about +If we really want to try and improve the performance of that code we could try generating the default value stuff inline rather than storing it in a map which requires dups +A Labs project has been created and apinkin is working on the content +we relocated repos +Nothing was actually processing any data or deadlocked holding some locks +The server stacktrace would be more useful than the client one +Works as per the requirement +Thanx. +Incorrect evaluation on previous patchPlease reviewThanks +If someone really wants it we can add it later +New patch attached +i am working on it +RESOLVED fixed createdThe patch has beed committed in branch and trunk +Can this ticket be closed +But we want to extend the quick start example we may want to. +Rationale for commit before review all new code doesnt change any existing features so no impact on other users. +It needs packaging for remote containers and how does it effect the containers +Per discussion w rruss myarboro bfitzpat ldimaggi moving down to Major and setting fix for +Stack Sorry for the bevvy of conversation +Patch from astitcher applied to fix topic exchange also. +I dont think thats what we want to do here +After discussion with Lee uploading an updated patch is clearer now about which C were excluding and why +Will fix that +Closing stale resolved issues +Moved to upon release of +Thanks for reporting this +I mean generate the code twice once withsn pn and seconde withsn pn follow the instructions from amila. +I hope to get to this in the next week or so +Anymore that columns and a horizontal scrollbar is created on an px width browser viewport +Also we provide a class that will read either +Patches +Fixing this is on the top of my list for +I also tested it with corruptdb database and see the thread dump information +Merged to. +If you beleive that should be please and contact the program team. +This fix will be viewable in the latest round of CP releases +thanks for the fast responseplease find attached the log +This blows my mind +push it Sebastian +Committed to trunk and m still seeing the same problem as last time when you create a server with a jre that doesnt support server it is ommitted OK +Im working on documentation and test cases for the this documentation will be part of Overall mapred documentation +I attached the corresponding wsdl and binding files to reproduce this issue +Weird that the beanshelldeployer was not consistent +I use displayed tab width use spaces for tabsIve not found out how to stop Eclipse adding leading spaces to blank lines so I tend to delete those as I go along. +I have committed a patch for this +Made some modifications to the method ser of the class Stub and tested +It proposes adding caches to each datanode using mmap and mlock managed centrally by theAny feedback welcomed +Adding MD SHA hashes for convenience +we raised the deletion semantic discussion when introducing improved gc algorithm +However there may be complex pattern which contains y as a sub pattern +David I was hoping do document each item as sub tasks to track them when they get doneI also believe that it helps undestand the magnitude of the projectHowever I will refrain from this here and use my Project management software +there we have the opportunity to fix it right +The bug isnt reproduced after class refactoring at revision +Suppose uses the same approach for JDK +Sample all for the input +Patch looks good to me but could include a couple more unit tests to verify absolute paths and paths with schemes +Thanks Andrey! I committed this to trunkThe patch did not apply to so if there is interest in parallel testing for well need a separate patch +Here is a patch to add the be nice if you could add some unit tests along with these changes +The class of the local exception is now also so that it can be distinguished from other. +yes hence trying out jboss tools would be the fastest way to verify +Yeah I dont think this method should be trimming the value by default but this is the current behaviour so I dont think we can change this +Instead it needs to use the one available in the +Removed generated columns from the bug summary since the NPE can be reproduced without such columns +svn directories so you should be able to easily replace all source files while keeping your svn wc dataLet me knowThanks +Small bugfix for currency drift not considering asymmetric conversions correctly +The new patch looks good +That would only work in the common case of streaming from one replica +Here is the patch updated to go against current state of hdfs branch. +It looks like no patch is needed for trunk +The Server Runtime Environment property page shows the runtime type name +Lets phase out writable over time +The reason is simple most of resources are loaded using houtputScript or houtputStylesheet or hgraphicImage so this feature is mostly used for css filesIn that sense I think it is better do not apply the previous patch because it requires some changes over but I have found some issues that needs to be fixed to make contracts work as expected so part of the code in the patch needs to be committed but in a different way. +The issue was fixed in Arquillian AS adaptor which is available upstream on Central. +unit for theThanks for these tests +A solidcore framework with satellite projects seems to be working well so far +Sun dont think it is a bug to be fixed because the result meet the JMX spec so could you update the test caseto the right asseration to avoid misundertsandingWhen we run the AS testsuiteit is reported errors and failedwe must pay much attention to this test case +The has to be changed for this to happen +HenriYoure right +use the Inject +Id like to have an account too berndfThanks for providing this great tool Bernd +It appears to be related to this issue. +We should remove and at least in or +Android uses keyDown Android uses key up +Removing Jenkins works properly +Will take a look later using +Do we really need this dialog? Would be much better to have lust list +Further information on how the component breaks accessibility +It works ok for single modules but now can be used in multimodules as well if someone d like to get this plugin wrapped up so I can go through the process with my Kodo plugin +This regression with respect to problem was fixed by doing a better check for unique attributes in the class +Thanks Rex Wang for the patch +the modification you made to the patch sent is incorrect if cacheGlobal is true and cache is false nothing should happen as it was done before but with your commit it now removes the coplet temporary attributes from the cacheIMHO it is simpler to keep the nested tests as is +It also removes unused function +Specifically now handling of gettersetter pairs should work much betterSo might as well try to tackle all open issuesI will close this issue for now can be reopened or recreated at a later point as need arises +Hmm the current behaviour will probably break old versions where have forgotten to add synctrue +Patch for branch +You can get that list from the class itselfI will apply the part of the patch to my local copy and play with it +not necessarily needed sounds like optional and it should workto know the naming i have to know what jar are you talking about where to get it. +If jars with invalid versions are run in the container will that cause problems +Excellent work my friend! Ill be sure to test this next week! Hopefully it all works out fine and i can rewrite the other indexing patches with easeCheers +trunk +Manually tested! +Anyone know how i can get it submitted +Updated test to explicitly verify that JDBC result set contains the partition column on a select from t +I dont believe that we should assume that every multiple database application is going to use a code generator or that Derbys syntax should force application developers into using one +Eg of is attached to this jira +Not returning null does not prevent any errors from being thrown at the time of calling the script of course +Patch for the fix +Simple straightforward to the point. +i think d should be fine +As I sad before there are Warnings in XHTML because WTP knows nothing about facelets and recognize facelets page as are JSP warnings because WTP JSF Tools cant recognize managed Studio had the same issues +Patch includes and s ok for me closing. + +Im going to try it with Postgres I think they have named sequences +What if we were to expose cardinality through the Bits interface? We could then define some point in where it decides which Query Filter execution Strategy to useHmm that makes me nervous since computing cardinality isnt that cheapIe its better if this is done once up front and recorded on the like does when the filter is cached +There slipped in quite some tab indents +Fixed in trunk. +As of now this is only fixed in the branch +It will be read by DN and we dont need a new config var +But thats not the point of my initial comment +I think this should be the same for Portal +Were not communicating hereWhat do you think does +a little warning I had this done in the past and we reverted it as it introduces a dependency to Javabut Infinispan requires it so Id be favorable to upgrade but maybe this should be limited to a separate module if we could create the Infinispan module we could call it clustering extensions or something like that and include the Infinispan and integrations +Reopened so that we dont forget to update the File System Connector chapter in the Reference Guide +I suspect youll want to change it though so Ive uploaded the whole thing to make life easier +Tests passed cleanly for me on +wont all components get this support when that is done +slip probably shouldnt be done in the UI tag but in the parser +This bug is component related so I move it into component product +working on it +Yes as shown in the patch which I just uploaded +sounds good to was planning to start today +Set the correct resolution +Better patch that matches whitespace conventions and also improves stream handle exceptions in write operations +Would like to start with streaming soon so I need to finalize the protocolI have a question about the implementation of the read and write from and to sequence filesWhy arent you using the plain WRITEKEYVALUE method? You can configure the job to output sequence files there is no need to implement the stuff on your own +In Rev +One server is executing the batch job and the other servers could check for the heartbeats +bq +Your concern for my happiness is heart warming +then assign back to +Not sure what you mean about changing the META scanner +Updated the JIRA admin to chrismattmann based on recommendation that JIRA admin should be the Champion to help incubator project. +So lets do it +Al though the code is very complex and each optimization needs to be thought through deeply and comes with extensive testing +Please update the bug if you get an answer from the WG +Both the exception sorter as well as the should allow for the Universal CLI and AS drivers +Great workIll startThanks. +or we could change it on to postgisng and leave it postgis on trunk +Picky I knowI meant to set priority to trivial +committed as r in branch. +I have took it for failure is not +Hi Charitha any reason you didnt grant the license to the ASF on your docs +GerryCould you submit a Pull request with the changes? If done in time we can still make it to to be released tonight or tomorrow RegardsGeorge fixed +placing a class of same name and package in our bundle +It contains version of Birt component is built against and tested against it seems that it requires a version of Birt and ICU that is no more in Kepler M then installation is failing +Alan is not an error of hibernate +The main purpose is to not add the y option +So people could still install it with earlier Ivy +I suppose it could have been a dirty workspace +The flattened array stuff is only on trunk and not on +attachments are all empty guys any chance you could upload it again please + +When running interactive query through the CLI CtrlC will interrupt the query and go back to the hive prompt instead of terminating the process +For this issue Id suggest to focus on XML side best way to start with should be to write some testcases marshal proxy with mapping marshal object with relation to proxy with mapping marshal proxy with introspection marshal object with relation to proxy with introspection +Is there a sample somewhere? Did a quick grep in the and beta but did not find be useful to have a sample file. +hard to replicate and just a stray error message +Gary helloAs you said to James to make appropriate changes for XSD in the I have little question now Im working under improvement Groovy Control Bus so should I also make changes for XSD in the Groovy Scripting componentsOr are you going to do this for all modules at a time? questions GaryI understand you +After all in most cases the links will display right away anyway and the user wont notice that they were gone. +Your patch works but is not actually the right way to resolve this +Changes commited branch only for now +Cant change it just for +I can see still in the hierarchy which extends +So add an addendum patch to remove the unnecessary code +patched the version of the bridge and retroweaved so that it can deploy in AS +Scott has worked on it recently and I had worked on it some a while ago +Please use instead +is that workaround good enoughor should it still be addressed +This as you righly pointed out earlier seems to be consistentwith RENAME TABLEINDEX +In addition there are more objects to be created in that case which would make parsing slower even if that is not important in that case +Christian looks really great +Maybe the error is different though +Something like localizerid randomnum is a better destination path than plain random number +New Revision I added in for both values and predicatesSee tickets +Please feel free to create a new JIRA with more details if you experience any additional issues +I tried this with and it still behaves as reported my Max +Its a simple solution for now I think we could optimize it more later? +Fix the warning +Hadoop has a release has anybody on our side done due diligence API and code stability +is just a wrapper +I will proceed with the proposed restructuring tonight +java source but could be set as a controller property in little bit more of history that is more present than we would like sameproblem on Mac OS with Netscape and MSIE also see also Bug for Pocket Explorer for Windows CE on a Pocket Loox fromSiemens for Pocket Explorer for Windows CE on a Pocket Loox from in the nightly build +It can be removedWe can also bump up to its not a big deal +Enjoy +Typically printing it as hex bytes feels appropriate to me crashing less so +Not a bug but normal behaviour +The same behavior was found on both and Mojarra +Attaching +What version of Weld is being used with GFWe had a lot of problems with proxies being leaked +Obviously this solution only works for building when connected to the RH VPN +Felix can you check if it is ok? for applying the patch +Note that does not matter much since that is for all the processes together +This spec will not compile a Fedora compliant binary +That would be a nice solution for many users I think +I cant access the Spring MVC archetype +The same way idempotency is broken when someone send puts or increments to a table from into a mapreduce task +If its really useless then remove it +Committed revision made Shape serializable and cleaned up the code in slightly + +Ill update title to reflect the problem +Weve upgraded our requirement regarding ASM +Thanks for taking a look Feng! +I have applied a fix to the code in case anyone is using this before we use. +Nicholas We need to add Fix Version information to this ticket so it will show up in the right place +CTF test case and release notes +Adds new filters +I have a patch for this issue but it doesnt directly address Adils options +Fixing a couple of issues +Part to the issue wasnt addressed +The way the fuse opts macros work is not to supply a default but rather a value if the flag is set +Do you still need these tests written +that a smoother partitioning may improve reduce time in some cases and should harm things in no cases that this is suggested with Integer and Long keys with simple patterns in their distributions +Maybe as a configuration option but you should at least be able see the dbs you have access to when you hit alldbs +It however lacks the ability to tell if the resource really needs to be recreated as Martin stated in his first answer +Note that the patch is a bit rough on the edges if dev agrees with it illpolish it up with docs etc +Robert that seems to be an issue related to the jacoco plugin +If is good to go and fixes this then lets commit it +Perhaps using colours here may help +bq +All buttons to the right are greyed out +I committed the patch +If fixes are necessary then they will be applied to the branchIn terms of the pommanifest +With it was just a note that the exception occurred in some instances but it wasnt reproducible and youre able to continually hit this error with a given sequence in the UI can you provide an example portlet of this error what portlets are on the page and the steps taken to create the error? appears on my pages after authorization +The event admin as created during GSOC with the md have attached the event admin in a zip with the md has as requested by Pepijn. +When it comes to consume this valueinspector combo the code bombs as the inspector is not actually appropriate to crack an +So if the repo location is equal to a value of any of the system properties then the delete wont be allowed +But if rollWriter is blocked all the following roll will be skippted +Lets please settle on v of the patch +right now its listed in the as a table with all the other given attributes +Does this problem exist on a simple clusterI do not have a simple cluster nowCould someone verify this on a simple cluster +Version has been releaseClosing all related resolved issues. +This looks goodbq +I think it covers most of the aspects of the API and mapreduce +bulk close of all resolved issues. +Steve status update +If no columns are specified the columns available in the tabular view are implementation determined but minimally include for each selector a column for each property of the selectors node type +Shankar SoniThank you for the patch! I tried using it on my local copy but it doesnt work +Yup same cause +on committing this +How do you pass the value to check in option ? You cant pass it as attribute because it could be binary right? Maybe in the body of the put you have two values first the one to check then the one to set if it pans out +Same patch but with another improvement +are permitted to contain characters such as and are mapped to using the usual XX escaping +I also get the same exception with the following codefoo as uint version however works as expected for xfffffffffoo as uint xffffffffBut again fails for foo Number overflow +ResolvingM binhbase Had a name for the hbase jar +Moving to for fix +We wont fix this in any of the ways described +Please review +Oleg +fixed as suggested in svn rdavid thanks for reporting this issue! +But my patch does not clean up the code for blockReceived +Attached compiled ear sources and standalone configuration file +Daniele I gave it a try +This causes the buffer that contains that structure to be closed +And SNAPSHOT happened to be a copy of an older timestamp after which has changed old Mercury behavior order timestamps but choose SNAPSHOT if one is present lastModified on the stale SNAPSHOT jar is after the latest timestamp jar so even fixing the sequence in which local snapshots are processed well be getting the wrong need to investigate why policy always did not deal with stale cache the last piece of the puzzle cache was checking condition ts ts while for always it should have been ts ts as diff was explicit check for policy always in order to save some nannos +Btw this is in AS trunk +patch appliedThanks! +Therefore it is no longer necessary to invoke the API in submitApplication +Thank goodness! Thank every bodys help! Thanks +Once I have the final version created Ill update the SMX NMR user guide documentation +Arg just spent several hours trying to figure out why when I upgraded from to suddenly my file was no longer found +Cant see that build issue locally and cant figure out what caused it on the build server +Applied to trunkWould be good if the PGP data format could have a table like the other which lists all the options and explains them +It may thus suffer from the same failure in case of extreme values +Itd be great if this can be fixed relatively easily +howeverI consider this task finished +Even I am not familiar with this code +Merci beaucoup Olivier +Attaching a screenshot of the dialogue and with nilAuthenticator configured +This is a start tests for COS level objectsalso bugfix made to more tests this invalidates the previous apllied the patch in revision as proposedMehdiThanks for the contributionAdam Thanks for testing +It worked as expected. +Thanks +the presentation of the wsdl has to take into account the xsdimport was changed so that the wsdl is correctly shown with xsdimport statements +the newly added variable DEFAULTYARNAPPLICATIONCLASSPATH isnt being used anywhere? The point of adding this was to use it at least inAlso not sure whether this was introduced via but we should use rather than do a manual to parse this out +that does sound like it could be a good optimizationsimplification of the code +To investigate in Metamer +for the check and confirmPatch committed r +Quite likely +I dunno this is a bit easier said than done +after all the discussion I just ended up committing this in r +Does it work in Ruby if theres spaces in the path +This way it automatically gets deprecated along with the mapred package +Attachment has been added with description The proposed has been added with description Enhanced patch with the check for the graphics transform +Hi SaschaIts an convention to close Jira issues once resolved +XSLTC and Xalan now both escape attributes the same way so it is fixed in current CVS and in Please if you find that this is not soRegardsBrian MichaelIm Closing this bug down +Im attaching an updated patch base on your comment +Im not sure Eclipse lets to select which JVM to use by default +batch transition to closed remaining resolved bugs +defer all issues to +Working Sets should solve this issue +I just checked in some changes which I hope will fix the problem +The one thing that i can see that we are doing different is we are using +I will pull that branch and test it in my project this weekend +I manually added to. +you +For the single node zones I dont see that being the issue though +Marking this as resolved as we have upgraded to modcluster which included the fix for was open for Release Notes review +Hi Lily with IBM and I tried connecting to this database and doing SELECT FROM It completed normally and returned rows without the exception +You need to be under the VPN to connect to the postgresql instance in the QA lab +Alternatively has only been verified by making sure that the feature can be installed but not that it can actually be usedIn other words what has been done on Camel should also be done on otherwise it wont work +However I dont know how to this issue +It was Guillaume that suggested it while we look at an issue with Camel and related to discovering type converters in bundlesBut yes we should take a look at this again when G +The index now uses the from LuceneFixed in revision +should be fixed +Patches to the dispatcher look fine to me +Actually ignore my last comment +Corrected url in +This bug also has the effect of erroneously deleting logs +Fixed for now as well +Adding the documentation and release notes flags +You want to have some fun and try to make it a pluggable module? Or you want me to tackle that +Committed revision patch to trunk +No test available it changes a array during installation of the ganglia server + for patch +However you are right that one shouldnt cache null in a HashtableIve supplied a patch above for you to try +Thank you I will test it as soon as possible +might be a side effectCan you attach your patch I can have a look at it when I have itThanks +If there are no objections ill upgrade trunk to use as well +bq +This issue is closed now +to WONTFIX as Dan notes +This should be fixed now +I was just sitting down to fix that so I was happy to see an email from you that youd already taken care of it. +Proposed fix for cases where there is no +Sent +I will file a bug to investigate in Wagon as it may be symptomatic of not properly waiting for a command to finish +As this is an API change it should go into this for as part of our effort at making server bytes only at least when it comes to keyvalues +So I think its best to ask himAs for method names is perfectly fine for me +Now the resolution of the namespace war resources system and app is resolved +Ill also note that we discussed adding the cache commands to the public class for easier accessibility + for the patch +Thank you Knut! I will do that but leave the issue open for further refinements +I attach a screenshot sorry but this is running on my PC so you cant see it Aime sent me an image back in June showing a fixed version promising it would be in as the fix was too extensive to backport to +But you can close this for now and I will report a new ticket if I have problems when we try it again +Clarification on why I filed this issueIf is enabled and you have a distributed index you can currently open up the admin interface and send queries to one core +I dont see a graceful migration path that doesnt involve forcing people to upgrade toDepending on the impact maybe there should be a released that itself when running on jruby +Committed Thanks! +If Designer takes upon this in their then depending upon which Teiid Embedded it depends upon this needs to be moved into that bucket +I think the checksummed signatures are goofy but harmless +Also patch shows errors on my system when running in direct mode as well +Verified in CR +See. +I agree me should provide an option to get raw JRE classpath without any access restrictions +I had talked to Brian Stansberry on this and I believe they have incorporated an unixdos scrub of the text files in AS s buildpackaging but I havent double checked with him on that +edThanksEdmund Wong +We already use that are sealed if possible and Im not entirely certain on the interplay there +RE +This is likely not a bug but lets keep it open for a couple of days just to make sure +Hopefully I didnt forget any +Distributed log splitting is not working any more +Im off to see if I can figure out a workaround for it +Attached which has changes suggested in the review in +lastModified will be only indexed if you switch on the pluginIf you think you should change the way lastmodified and date is stored in the index please submit a patch for +As the CTF suite still runs without any problems I think this can be committed once all your tests have been converted to proper CTF tests +I spent several hours figuring out how to exclude an item from the WAR in the end I found out by reading the source code +Sorry to come in late the patch has gone stale +Maintaining separate ticket for that functionality +We would also need to overwrite code to not generate the field and IMHO at that point you may as well just write an overriding setter that generates the events +Removed JSR dependency with commit addfdefdefafa +Hey Stuart thanks for the reply +I have the changes isolated waiting for a full unit test run so I can figure out if I need to update any +My concern with this is that it embeds documentation into the error message that may change +The updated patch looks good +I agree that detection of could be much better as the seem to return different values on different operating systems on the same architecture +The old are still interpreted but produce a WARN log message stating support will be removed in a future release +Would prefer to separate this from the leaking dependency though +The issue has been fixed on the trunk and will be merged to and branches shortly +buildCycleConstructs the cycle for back edges part of DFS path back edge cyclecalculates the cycle for a back edge cycleOnce a forward edge is found the graph is walked to find a path fromthe end of the forward edge back to its beginning node findFinds the position of the edge starting from a node in a continuous list of edges +It would be nice to handle this cleanly behind the scenes if there is an easy solution to thisAnyway it seems Im thinking about this from a more automated point of view where everything starting from virtual computing nodes are managed automatically +Sorry Im a Bugzilla newbie ignore the comments about the edits of the patch file +My preference would probably be either a named file or standard inout if thats not providedYoure not using the container file format but rather just a file containing a single record +See +Ah that makes sense +patch for xmlschema applied Need a test case to gurantee that it does not case the in the given thread and some Synapse sample tests with the +After a removing the domain project and clean of dmServer in STS and a new deployment to dmServer it works via STS correctly +Whats with the blanktransparent area in each imageI made them transparent like and +Closing this bug. +We need to allow for this because customers will be doing the same thing +This test fails on AIX with ibm and On zOS with issue didnt show up in the recent tests ran on the Zos with derby build number +committed +Clearly theres a NPE here +I reviewed an early patch from Karl related to this and had some comments but it was a long time ago and possibly just on the mailing list so I dont know where those areI havent reviewed this enough to vote yes or no but before we do anything with something really low level like this a few people should thoroughly review it and vote before we do it and each reviewer should make sure they understand well this patch and all of the code it touches as well as apply it and run it firstIt would be great to get this reviewed and committed +Builds OK and test passed on my clusterI just commit this before make other changes. +Source code patch on to apply in Eclipse +This occurs because the following classes require methods to be added in order to fully implement JDK interfaces to implement If the application you are running requires a method missing from the above classes a will be generated +Ive removed the reliance on the sleep method from this test so you shouldnt have to muck with the sleep time value any more if you grab the latest code from trunk +Ill try to reply to all of this soon +To be more accurate if happens to one application this application will not be completed crash or anything while other applications that are running simultaneously should not be affected +The issue about using first splittable familys midkey for the whole region is still there +I think if core does not depend on contrib its tests shouldnt also +We can add some simple rule for downloading resource files such as starts with s schema in this case +Fix version was updated +Now I have a workaround it should just be a major +ToddBasically its for better and error handling +Somewhere on the wiki I once read a tip to a file in order to make the page thats why I thought this would have to be checked +I cannot update them +Closing resolved issues for already released versions +Its a change that we should also mention in the section +Sysprops should be a fallback though +details in issue +I removed bench from the README +Please close this issue +What version was the attribute added +Normally all developer use the same shared installationwhich should contain the same Groovy plugin version +Does this currently work? if so i think we should close the JIRA as invalid +Bring it on +Ive seen this exception too however it went away in my case when removed the asm bundle from orbitIll have a closer look and see if I also need to make changes in the tests tooThe fact that the groovy integration is done via scriptengines makes me optimistic about being able to upgrade groovy but this should be our last resort +Youll have to dig further and start investigating what the actual errors are +Added a test +Can the shared exceptions be anonymized +A useful tutorial would be writing an application state object manager that used a dirty flag within the ASO to only write it into the session when needed +Anyone agree with this? I disagree because I think the structure is good as is as it leaves open the possibility of splitting out tests for different impls making the tests abstract +I will incorporate your comment and update the patch soon +I forgot to fix too +I think the problem is when I pass controller and connect to the remote client jar +Im not sure if Hibernate other JPA vendors do the same +New patch that addresses Colins debug log comment +I am not using the command line groovyc instead using the class +The in the client allows for clients to have created the +If it is decided that we dont want this feature then it is just easily taken out +Probably all the SVN links need updating +Ubuntu i will try it +Should this issue be resolved or should be removed from the repositoryDoes it serve any useful purpose now that we have the tests +I still think that having a float for the API is preferable but until we actually start using it in practice we will not know what the real issues are +This bug is only fixed for IE +You can follow the pattern of CXFs unit tests without much trouble +Events handling was more difficult but we have already factored some of the codeAnother point is I really need adaptive step size so while I was setting up the method I extended it at the same time +Thanks for review Claus here is the unit test chunk +The SASL implementation is a wrapper around GSSAPI we are already using SASL for multiple mechanisms so it makes sense to stick with SASL for Kerberos +Thanks PaulAfter some thought I modified the code to print out a single task link for now thats good enough for debugging purposes and users wont get confused by too many outputsWe can change it if needed in the future. +yeah Ill have a look at m sticking the bit Linux JDK that I was able to reproduce this with on in x that will help with debugging the issue +Refresh the changes against current trunkMahadevs comment on post install changes is still outstanding +I asked if you had added the since if you had that could explain the problem +added new j which differs from j in that it copies the J files to the target build directory as opposed to the project itself +see last agree with Stefan that there is an issue somewhere and since it can be demonstrated by just help its not specific +this is weird on my dev machine and in the patch it is not synchronized +Can youplease be more elaborate. +mvn o Ddatabasepostgresql Dtest testRunning run Failures Errors Skipped Time elapsed resolved in +IT added +Siiigggghhh +Thanks for that! Id resolve the issue myself but unfortunately I cant +Committed revision for branch. +fix anything you think should be fixed +Even a developer editing the files on Unix and then copying them to a Windows VM ought to maintain CRLF +No time for me it happens all the time after I save something and then hit Full Publish on module in server view +on adding the random sleep seens like a low cost solution to implement + +I just had somthing similarTry the followingrun ant on each of your tasktrackers machines antthan restart your nutch and try againI think there is a problem with the classpath +Includes updated Hive benchmark queries +The problem stems down to the Java line it is calling RJB before we had the opportunity to run +Thanks for the report! +As for where these are placed others have related files in their own directory +I feel its not good +dog an offensive way of describing a woman who is not considered attractiveIm sure there must be a similar word in Catalan? incansable +Can you please verify? I had to apply the diff for manually for some strange reason +it would be better to return no such core exists +It is a single variable to track in the and is easy to explain +Added patch with Bob Copelands fixes and updated to Bob my patch still doesnt have tests looking at that now + +BTW release never became stable +Trivial patches for and trunk +In the case of transportsession the session is managed by the transport and Axis does not know anything about it +Lets add subtasks to for all of the related docs to keep trackWe need to add code to plugman probably to handle permissions from and add them to app manifest +rename issue to reflect is blue +Well I was going through +bq +The problem got introduced in r +Unfortunately that change kills performance on large file reads as the generic read path and copies the output buffer many think I fixed the problem in ebfddfacebed there is now a file size check and for files of size the slow read path is used for files where the size is known the fast read path is used +Do a ant to copy the jar file to jboss deploy directory +Since this is about block reports in the context of appends the following question should be appropriate here +Thanks for fixing it +From Snjezanawe were able to install multiple examples +thanks! Id like to report a codehaus user named mariadev this person just added some spam pages to the jetty wiki +I attach my patch +Yes thats exactly what happens when you just paste in the code +ESB will work with the default defacto lib hurray! +Can you add a debug statement that logs the exception if there was a failure opening the file system +Bringing into and +screenshot of login issue is also present in can get around this by using a filter to put the current user on the model allowing the layout to read it +We seem to keep delaying such important issues +Thanks Johan! +Targeting for +Resolved in trunk At revision +The concurrent check worked but the tablet load failed and retried +PeteHave you also edited on pointThanks so much +Anyway the timeout should be increased and made configurable to overcome those failing cases +or new ticket for omissionsoversights. +Dunno whenhow to do this +This patch does not apply I will change the line directly later +Akira review the changes and see if there are any changes that require action run is a complete buld and test of Hadoop including HDFS YARN Jenkins doesnt do that for a change to +Sending srcmainjavaorgapacheservicemixhttpSending srcmainjavaorgapacheservicemixhttpSending srcmainjavaorgapacheservicemixhttpTransmitting file dataCommitted revision +I would maybe put the list into as per the list +Leave about the jasper +Added all ddlgen source files +Here is the updated patch +Fixed on trunk and all issues that have been in resolved state for more than one month without further comments to closed status +So I was thinking this would be a good feature to offer the user in +Then once we have all of the entries updated we can look at cleaning up the distributed cache +I find it odd that youd still need Maven to build JIRA today +Let me know if it seems ok +FWIW we should probably add Uri to changes as well as he filed the issue +I agree +when you ran the test on did you use the tests or the tests? I think there may be changes in in what tests we run +what about stealing the scheme from popular companies and introduce or bg +Sorry that I havent gotten back to you yet life has a way of slowing down developmentAs for you comments from a couple of weeks ago +Still not working +I created to build on this package for a better user install experience + unscheduled issues to the next release +reproduce ant package test Dtestcase DqfileChanged directory and test passed +I just committed this to trunk and +No problem Ive removed the blanks the patch should be fine now +Then GAV specified in the BOM is correct as your link you need to be using Maven which module are you trying to build +I made a test and everything works fine +Not sure how I overlooked that! Ill wait for your sign off on and then update this on the live site +I just committed this +I just hope the ghosts of jslib dont come home to roost +Fix is in place for M. +I verified that both version are setting debugfalse and that the log output shows that seam is setting this correctly to false +Is finializePartition message what Ive been calling mark set done? If yes I dont see the string specifying the filter that specifies which partitions are done +The change that I suggested involves having the datemath syntax in regular Solr query parser +The mgmt option was considered ambiguous in boost +I am beginning an implementation of it +Dates get formated as for st July but for st November +DB specifically was tested extensively w no failures +Schema to use to show think bug and this bug are duplicates +This looks fixed in revision +Geotiff samples which dont work sent to Daniele +Good stuff Jon +Agreed I meant to use that patch as a base and we definitely need to revise it +not in this particular issue since it is more on service level and we already talked about it and a draft for it is on jbpm wiki jBPM services there is session service that will allow that +The issue is valid and resolving it should take a few hours +I see a bug in maintaining this counter +It doesnt affect any code but the documentation template and the of the class which is invisible anywhere but in the source code of the classHowever as Im a very new to the project Id appreciate if someone can make a quick review of at the patch to the internal javadoc to make sure it is accurate sufficient for future maintainability +This issue has been fixed and released as part of release +Hi ElenaBug fixed at r please check if it fixed as you expected +The bundle class loader ait atop of this boot delgation class loader so the order in your patch should not cause majot headaches insodeSorry for setting the issue to resolved so fast issue in preparation for release. +So the task is to make sure the same is done for all other modules that include schemas +Ill have a look at the patch shortly +However I double check the status of module all of the associated with a same IOR are started finally +Applied in revision in in . +As mentioned were discouraging use of the Struts datasource manager in favorof more mainstream solutions like JNDI managed through the containerTed +Updated Patch file +You want to set character encoding for request and response you say +Hey Carlos this sounds like you are running mvn tomcatrun +I just noticed that you attached another war but it still does not deploy as it is missing and is referencing an unkown module ecafemain in +Throwing a checked exception might cause backward compatibility issues and returning an empty string may have an issue that some user code may be parsing the service assuming it is ipport +It also didnt make any allowances for different behaviors that should be configurable +You can now see that you have over issues in your JIRA project +If this bug is still an issue please it. +Uwe I am talking about this patch too +I just committed this thanks Cheng +On trunk precommit passes and solr tests are underway +The patch helps in this case +No problems with the patch eithercheers Roland +Brian thank you for the details +Resolving issue as fixed +Makes sense to me +Otherwise you may request stuff that is already there + +Just want to mention that has also an awesome file upload component. +for Sis commentCheerAlex D + +Puts ibiblio first +Yep with extra gem +Updated patch to resolve the concerns in moved the saving of the result to the finally blockThe rest we can create new jiras for once this one is resolved +If the testsuite requires certain dependencies cant they be added directly to the testsuite poms +Since the Microsoft will not fix this bug Im closing the issue with resolution Wont Fix. +ran cleanly the second time with the patch too +Shoulf I just change release to debug in the following line of or filedrlvmtrunkbuildREM Note is always complied in release mode otherwise it makes VM debug too slowCALL ANTCOMMAND f make release +Build is done +I agree with Jonathan that the prefetch entire META should be optional default false +Hi DougWere you able to reproduce the problem? Please let me know if you have any issue with itThanks will commit this bugfix soon unless there are objections +Thanks for the patch Nicolas + +This is excellent documentation Nirmal +You could use an ephemeral port rather then limiting the set of ports needed +Attaching +I fixed this too +The API has changed on them +Argh! Another raw +Wait that isnt even the redirect admin classes which are causing the NPE but mine +The problem is The header for the compacted index is written much later than it could be +Let me kick off the tests tonight and Ill reply back +has been marked as a duplicate of this bug +Changes according the the latest and +I bet it is just a matter of deploying the site without disabling the benchmark profile +The heavy heartbeats cannot be much slower that what we have currently +I just have the fileHow can I have both HTML and XML reportsIf you use for example the xml option for the xml it is possible to initialize it by default with the outputDirectory option valueThanks it works well for the xml reportOn the other hand I cant generate the html report any you have a look to see if its your case too? should be in the targetsite directory by default unless you override the value in the my tests are showing it to this is not working for you still I may need to get a specific use case from you that is displaying this +Please try again with the latest trunk build +That explains it Cost me hours figuring it outD Could this be solved by adding the checkreplacement to as well +Extended to test filter functionalityApplied space changes +And with Jeffreys I would expect this becomes even better +If it helps steps to reproduce Set time to something like seconds Add an entry in a route Wait seconds Attempt to make several around the expiry time +release bulk close + +Should we use JIRA and diffs to edit the release notes or just pound away at em and do svn commits without formal discussion +Sorry I got things wrong +Wouldnt it make sense to drop start method completely then? The sole purpose of start is to allow consumers be created first and then starting all of them in one go +targetted for Apr th? That is pushing close to our beta ship date +Good catch! Fixed for +oops my bad Updated the patch +In any case Id still like to see the patch. +Hello Jilles Id still prefer no log message at all +I agree +I used am using both at first I liked Maven better but if you just want the dependency mgt feature I would go for ivyspecializing on dep mgtsmaller and simpler to useant is a first class citizencon ivyneeds to use maven repo as public reposmaller communityjust my two centsJoris sure you use jboss maven repo to prevent library duplication if you are to use maven +Instead of catching exceptions it calls super directly for generic and specific records +Many Thanks. +Resolved by Peter Lynch. +you noticed the general idea of the classes is to be compatible with JMS providers +In a recent ticket I changed the naming of fields to have the underscore at the end +Deepa you are right about testing the switch for both of following casesAs part of db upgradeA database created in wo SQL Standard Authorization gets switched to SQL Standard AuthorizationI havent given much thought to what kind of testing is involved +I think you are not the developer of apache camel so the issue cannot be assigned to youWe will release the camel soon if you want this patch to be a part of camel you may need to hurry +Verified and closing issue. +In the meantime it would perhaps be worth documenting the current behaviour +verified closed +Attached screen in trunk and site +Will it be in CPThanks this fix will be in CP. +A crude patch to fix from a pipe raises ErrnoESPIPE if passed an offset with both and +recovery is not enabled with the AS configuration of JTS +hallothanks for the feedbackthe resource filtering works! thats not the issue +The improvement has been committed to the trunk +I will try to have a look +Committed to branch and trunk +But Ive got design issues related to applications settings services beans and user connection relationship and lifecyclesI close the ticket and about my problem +Case seems valid to me but core does not seem to be ok with loading a PK object composed of reference to the initial entity +Should it not be to make it consistent with other MR configs +We need to add value to our seam model for and that please reasign this issue to me +Attached is the example process modified to demonstrate this error +Lets move on! +It would be good to add a single node secure deployment instructions as well +IMHO eliminating overlap between the specs is crucial +If I remember correctly the idea was to have differentchains at differentenginesfastchainenginesmytestchainsomething like that +Im therefore going to postpone it unless I see it happen on a second system anywhere +This is in addition to +I would add that explicitly specifying the primary key in the annotation does work +We have started to compile using so the problem if any should go away. +If VM is not running then always shows snapshot option +Tested using both the default and setting a custom one +Then the exception apparently wasnt handled well from the +bq +classpath template as well when you commit +CP branch upgraded as well +Looks like this is due to works good if this is disabled +This would imply that we want to persist when the current scan started +You are using +test need to be changed accordingly as well + should be a future release not a past release +And we are not keen on people opening tickets +Resolved in Git ID bccbfbbaaffccbfeffd +So I think it disappears with Spring you have a release plan for spring? am experiencing this on Tomcat with Spring +Yep should not be affected +Create keyspace with rf and also create a column family +I expected it to check the feature itself to cover cases when the feature content is not valid with respect to the feature type +I think a new issue is more appropriate +Well it works when one starts the installer as root user +I m still struggling with the Firefox search plugins though +Since is renamed as +In fact if I pull down the from the above location and do the exact same command I get jar tf have jarred a inside so it gets overwritten when extracted +Also configuration using the approach in provides more flexibility +We will need to define a extension for those files that we will search for and ignore the name +Such project examples files arent being cached because they are local files +We are seeing the extra eventpoll and pipe handles on jruby on +Here is what I applied to trunk and branch +Maybe with a little lighter grey +a fix to perform the saidIs the patch against the latest revision in SVN trunk? I am getting multiple conflicts applying the patch + +Sorry for the dupe! +Setup an instance of named on your PMCs zone and have that be a hidden master for This means the PMC can manage the updates zone how it sees fit but allows us to manage the main zone inline with others +We dont have to loop through resources and check them one at a time +Thanks for addressing thatVerified tests pass on Windows and are skipped on OS X +BTW the test passes with JET pure OPT RI etc +Hi HadrianIm glad you like it and yes sure I can wait till then +Regarding your issue looks like either ran out of ZK connections or restarted zk while deleting its data +Tested and closed. +The generated xml output contains a duplicate attribute xmlns and this is wrong +I didnt see any tests that check that it can properly read null data +Verified jobs worked in both casesTest procedure for the error case is the same as steps to reproduce above now however the NM will NOT show up on RM node webui if nm doesnt start properly +I still dont see how the actual use of both firstname and lastname in indexInfo though +Also configuration of the and other parameters +Applied in rev Thanks! +XutingThanks for the contribution! +Now one of the monitor servlets just displays it directly +But since this operation is synchronous we can use a didicated node inside the repo where we can keep the progress of such operation +Fixed patch +UGI in RPC and in some test code +then youd have a block index file for a duplicate of block headers in a nd file for redundancy and probably a BF file +Thanks Dyre +for the trunk patch +Otherwise reopen. +re comment A +will commit +We need to update the OWLAPI and remove the embedded Maven repo in commonsowlapi ASAP +Lets document the need for jdk in and make this part of +The ci servers use rspec which is why we switched the ruby build over to using bundler to automatically handle the dependencies +Patches are always welcome +Should this be back ported to +I believe it would be more sensible to extend the Maven SCM Plugin with such a check such that users can bind the SCM Plugin into the build +Committed revision Thank you very much for your patch CheolsooJarcec +Thanks for taking the time to create a test case that demonstrates the problem +Im currently looking at this for CE via +Created +Heres a patch which converts and also adds a basic unit test +No AUTOMATIC VERSION updates on the VDB as before are supported +Thanks a lot +Cool +buffersync reads into chunks +for the patch +Fixed in rev +Stefan Yes I agree +Do you have in the folder that JPA requires +Checked into by Matt Hogstrom +Therere only two weeks before we branch I think v would be in containing this feature and +uses Writable implementation +took a look on linux and got the failure +Denis has created which is the same should fix this dialog but developers still can use old one without any problems with shared absolute path is used only for dialog and Tools dost save it in project issue was fixed some time ago +Please disregard the review request I forgot it would automatically add it to the bug +Good question +can one of the committers please review +Opps spoke too soon +A few IBM specific encodings are supported already in sbcs looks like this one is missing thoughWell need to find some suitable detection ngrams which shouldnt be too hard as I seem to recall that EBCDIC puts and in a very different place to ascii the iso Nick! That actually sounds promising +What does Java here +I am open to changing the definition but we should make sure that the javadoc is modified to reflect whatever change we make +Ive tested this by hand and it works as expected +I implemented a filter reverted a change from markWorks without problems in my environemnt now +Unassign this JIRA since the project was not accepted. +No further input and this shouldnt be an issue as far as I can see. +as mentioned in a sophisticated caching is not implemented yet +weblogicservereclipseplugins No match is available for the required execution environment +this seems to be particular to your mojo since you are using setters for the injection +Could you give me a link where I can find this information because I found anything related to this exception except my requestAnyway you can agree that the error message is not so explicit owner not associated with session In my example I have only one collectionThanks +Diff and files as version +Not a bug +Attached some notes in that the patch is first small step +What context were you when you needed such feature? +The changes between and do have an impact on MB lib and demo when reviewing the changes it became clear that the current schema is not I will put the updating on hold for now +Thanks a lot Christian and Jacques! I will surely check this out. +It would make sense to add a note about standalone bundle to the installation docs and also have default userpassword noted in some obvious place like install docs and readme +So if you go via string the correct class is generated for the root object +defer remaining issues to +It implements the Luhn algorithm which is effectively Mod +The screenshots on cases C and D are attached +I think a deep code is required here +Compiler warnings generated by Sun Studio +Consequently Ill close this if there arent any further is a JSP peculiarity not something that Spring can address +the resources should always be on the classpath +it is also in jpox so its easier to build continuum from source changing the value in is possible to add more files into the classpath as well but I havent tried it directly +Hi Waynethe collapsed state can now be found in the attribute value of the component best if you want to set the collapsed state is to bind this value attribute somewhere to your managed beansThis is to make sure that a collapsiblePanel also works in a dataTable one of the many bugs fixed in the collapsiblePanel in the current head versionregards a for the attribute value if you want to know the collapsed state of the component. +like aws does +If the tests fail it fails the build +Running test with useprocessfalse made no difference +Where I disagree is with the assertion that we cannot get to it starting with a simple implementation +I just committed a HUGE patch that removes the static singletonThis does not yet support configuring and using multiple cores +were the other issue startup speed +This is the patch for to add above described behavior to axis +As far as I could tell it worked without problem +The has a siliding window protocol to send data to datanodes +I moved your issue to MAPREDUCE since the isnt a component of HDFSA few minor comments Please rename Statics to Statistics in the code +Confirmed +Im not sure if rewrite never really added much to Lucene though +Reopening as this give more information about the problem and how to reproduce it +Marking as resolved since most of this work has been completed +No progress has been made on this to date. +If thats the case you are oversimplifying things +Attached the patch to fix the issue +If it is a bug with firebird please let me know how I may report it +Thanks for fixing It would be good to have session +I see it done both ways in source +but unlikely to happen for unlikely to happen for +The avro specification doesnt indicate that the values must be Strings so they could be arbitrary JSON +thanks to Paul Rivera for provide us this patch +IvanIt looks like exception occured in agent code +Funny a bug in a method that was written to expose bugs +Thanks +The reported bug is about using MEDIUMBLOB which is not in the default sql mapping file of NutchWhether MEDIUMBLOB ought to work or not is not my decision but a better error message and some documentation about supported values would indeed be nice from a user perspectiveThe Data truncation exception in the comment is indeed a problem with Nutch assuming that gora does not promise to support arbitrary length data when no length is given +Is this fix suitable for backporting onto +I am going to start making a patch for these changes but I would like a review or comments on the config changes as I work on the patch to try and speed things up +Yep +Related might be worth to check whether theres a newer POM which would define the newer plugin versions +how do and Order currently get around this problem? My understanding is that they force the preceding data to be written to disk then run a sampler job and use memory estimates to determine how many reducers sampled keys need to go to +OK this has been committed to If you want to to the branch go ahead. +And the minus method is declared using a type parameterI can work something special for this case because it is common but I need a more robust solution for method type parameters + +Hi Todd thats painful but workable +It will cause performance issues +I thought of illustrating the duplicate registration of udf using this example and got into wrong directionThis was a bad example +We could revert to the eclipse generated default of bq +Account created. +The generated Java Source +Is there another workaround? +I am runnning it as a standalone application which contains hibernate as the jpa provider +Please upgrade to FOP and use the Batik version that comes with FOP +bug of a bulk update all resolutions changed to done please review history to original resolution type +Thank you +Thanks Andrey patches were applied at the revision +The next step would be to launch in debug or add debug lines to for when it is called with +Fixed in trunk. +The initial implementation will consist of the following components SOAP Endpoint Impl +the stack trace weve seen so farThose traces are suspect for the problem described I think + +TedYes that is the problem that causes the error +NeatThis looks much better +Oops you slipped in there +was barely for your help here +Patch currently applies against and Josh for posting patch +Dont assume it just works that is the attitude I use after seining so many issue with it +OTOH the current usage of EMPTYALLOCATION +I thought this would get addressed with but i guess not +I am more than happy to get the latter done today would you be able to help me out with the first one +Shame. +Marking as resolved based upon the new integration tests +I was able to rewrite parts of JAXB annotation introspector to resolve this problem and I hope it would actually help with other related issues that were due to incomplete coupling of settergetter pairsOne remaining problem is that of fields I dont know how much JAXB allowsexpects field and method annotations to be used together but current introspector will not combine these. + +Looks like good idea to me +v makes suggested improvements +thats why I was suggesting a as well +Sonar runs are triggered by our CI system Bamboo at night and take between up to minutes per project + LGTM +Attach a patch for review and we can help resolve otherwise provide an example with an attached +The EJB beans project I used in another JIRA should show this +So dont throw exceptions in the methods +An example or two wouldnt hurt eitherI agree that a signing component would also be nice and it should be very easy to put together on top of what is there +I have made the fix to an internal branch off the line that I maintain and am looking to get it onto the and lines +This should then also work for the Tika parser and the i think +Ill go ahead and setup the system though +Potential issues the GS on the linux box is older the startup script is not really passing the parameter to the virtual no worries I always use Java +That makes sense and if it is still the case it would mean the fix needs to occur in the and not +I will address this soon +Fixed for. +Is it possible for someone here to test it? +This works the same way as the servlet +I already know that I need to rename a method before I check it in +Patch for versionNow I have changed like the check will compare the rowkey only +Also converts the test to junit style +Their literate syntax is very easy to use +Lowered Priority and deferred until. +looks more like duplicate foreign keys exist in the db + +I really think the CHAR value should be a String just like the VARCHAR + will commit after test! thanks siying +Committed the updated patch +Is that what you would like me to do? This would involve the creation of the correct roles and moving any release artifacts from the old repo to the new one +it should be treated as a program errorIMO listing the exception encourages people to catch it suggests that they can depend on the exception being thrown and also makes it more difficult to remove it from certain methods in the future +Full thread Kellerman Do you have an opinion on this since it seems to be issue with root region assignment +I went ahead and changed it +How can we provide an invoiceId when we are still creating the order? Are you referring to another invoice? should be the invoiceId which is an IN parameter of capture +This is not error this is the answer to your question +ant test and ant passed on my local machine for this fix of Y! +Looking forward to see you in Berlin +Thanks for working with me on this +Just imagine a portal which have lets say users how do you think you can force all of them to drop the cache and still have creditability? Scary +The details of the exception should be logged in the +Partial solution adds back the features and versions but is missing the categories from the HTMLThey are present in the site metadata however +Ran some of the rumen tests locally and they passedThe patch doesnt apply to though +Forgot to mention Ive committed this to trunk and +I would like to review this feature with the other table components as part of the release plan +Merge problem +Need to investigate the odd failure of that may be related + +This is related to and +This makes it really easy to enabledisable authentication depending on the Stanbol use caseBy default authentication is available for the full launcher disabled for the stable launcherWill commit later today +I dont want maven +AndreasSorry for the delay +Whats a good way to specify the tokenization and analysis to do on each of the fields? We could go with a version of or maybe use Solrs reader +Apply the attached SVN diff file +I agree with your comment. +passes locally and the failure in the logs appears to be unrelated +I might have to open some in those projects. + +Hi MansourWere you having a problem with a +In the end I think well just have to remove the ifdef altogether although we may have problem on other platforms where the ompilers are doing the right thing +For example a column with validatorclass and a value of is printed as +Whoops! Missed adding an implementation of the new abstract method to one of its subclasses +This fix the build on windows +Committed +NET web services most probably due to a bug in +So the source tree is only populated with language X if the end user needs itBut since were moving to Avro this will become much less important +bulk close of all resolved issues. +Its the datanode that crashes on me +Just run the integration test included in the project +It is clearer now +flushData checks and returns false if theres more data +zips are too large to be attached to the JIRA but arehereOn In homebostonldimaggi +This patch adds Dougs suggestion to use exec for svn +Implies a check for existence as the issue to focus on making sure the the bootstrapURL fully supports all including +On Android it goes to the default browser because this is how every single app on Android behaves and if we were to suppress it it would look stupid and tip the end user off that theyre not running a real native app + +Thanks Dag! Is this what you had in mind revised? Might as well make it as correct as possible while were at it +This looks good to me can we commit +The calendar internationalization would be a great improvement not a hurry though it waits for some time now SaschaIt is very weird that I cannot drag and drop portlets after I refresh data +Corresponding changes in reflected here including merge with copypaste bug +We are interested in providing SSL for but we need open source SSL engine to do so +Is anyone planning on looking at these failures I see them consistently on my machine +Hi +the new partitioner first reaction is that we probably dont want to encourage people doing this vs a single wide row +Ill add a text field and a browse button to the Templates Preferences Page that will contain the path to storeload templates file +Changes the to HashtableThis patch may be committed in this form presumably its correctHowever more beautiful solution will be declaring PRIMITIVECLASSES asThis patch gives up to boost on running at Clovertown +Sure Ill see what I can do +If a file does not exist the OBR resource is deleted +It may be an API change but it is binary compatible as it just adds a method so surely could does not require a major version bump +This issue is fixed and released as part of release +SighPerhaps Avros reduce could be run in a separate thread that reads from a queue fed by Hadoops reduce +Did the workaround help in the mean timeHavent tried it yet tracking down another Shiroapplication problemIll let you knowAndymeanof +This means there is never anyway to differentiate +is a rendering issue that is perhaps out of our control + +Hi SiWithout knowing too much about how this works at the moment couldnt we add a split payment methods screen to the checkout process? That way the customer could go to the screen select a payment method and amount click add payment and it then gets added to a list of payments +I will test it out but wont be able to get to it for about a weekThanksdk +the last dangling issues associated with this have been resolved +Added a number of new unit tests for both and and changed the to use a configuration with sequencer path expressions that use both repository names and workspace names +we will include it in for finalI see the topic branch is it ready to merge +for the patch +What we do is after we make the necessary changes do a reload +When? Where can I find info or docs about the spec +If the problem does exist in current versions of Grails and we can isolate some details we can address the issue at that pointThanks again. +Right +Verified on revision +Hadrian I wonder if we can add a warning to the release notes about the missing hash and sign filesNobody would download and use release where there is or better +However this will mean the fields will be reported in the feature type description and returned back in is your reaction to this +We must restart tomcat to stop the increase of the log sizeWe got this problem since we put instead of and we dont find any fix so if someone have an idea we can try it +Thank you again. +I would appreciate if someone can commit this change +I am working in this improvement +On each hostvi etcselinuxconfig This file controls the state of on the system +The problem with signed tags is that you need to make your signature available somehow +Same patch fixing the typo in test +Should be verified in gatein master and reopened if still valid +This makes sense from wording but isnt there a need for putting arrays in there? Any other comments +verified on JBDS M +start from timepunted to is a year old bug +Hi ArunI am not very old with hadoop So I have two questions +Database Information is still relevant to EAPThere is also another problem with chapter JMS Recovery which duplicates the XA Recovery chapter from Messaging Guide +Could also because of +Id like to see Java supported as well +I dont know if this has any unexpected side effects +Any more info Tony or can this be closed +This make sense +Fixed by. +Where did this plugin go? The page is down +Willem will take care of the partBTW the flatpack also has these options but they have already been documented +Velocity is tried and tested so would be a good choice in my opinion +If you wanted to say script X is not allowed being loaded by user Y then we get into a problem here since the loading could be done with cached scripts only +AFAICS the shutter is very fine +Hope this helps to find out the changes affect this +Well looking at I guess we should not add yet another but modify the existing one +Please mark it such in release notes +I forgot a detail +Fixed in todays build. +OK good to hear that the snapshot works for you Thanks for all the debuggingThat old in the classpath sounds like a pain +is already at least one rd party plugin it is not available from a Maven repo yet as far as I know but it can be downloaded and installed locally +As per the commit msg there is a reinstated test that should be removed if changes for are undertaken I will post over there accordinglyThanks no comments from me. +I fell comfortable that the FB crew will produce an awesome feature but Carl is justified to suggest that if we not have at least the core tasks broken out into or jiras it might be too much In FB we trust +Its already possible to disable a Roller user so that they can no longer loginWhat we need is a way to mark an old weblog as inactive +Should be part of a separate discussion than this one whose original purpose was to recognize an existing policy +Basically it is identical to the afterInterceptor but only fires if an exception is thrown from the controller action +Can you try with version? is very outdated +I am not sure what the problem is but the difference between IE and FF makes me think this is one of those tricky IE problems +In my implementation the data is distributed among the nodes equally and each one of them perform the processing on the data on their data set only +This bug has been marked as a duplicate of +This will be a welcome m hoping this doesnt get bumped from Spring its affecting our ability to completely remove cglib from our environment which is causing us to restart our production server after every redeployment +reopening to fix fix for version +Looks like someone complained about chapter as well +Attaching contribution and test case +I am so sorry for the delay Gavin having small difficulty with versioning of the site SVN +look at userservice to see if you can do this +patch fixed the search in bsh to collect all the parameters provided by user +This is where converters for domain types are managed from now rather than from individual controllers +Is there any update on code review of the patch? Thanks +Removing the version from this bug +Probably best to change the production configuration to purely +fixed for Jboss tools. +The option will come inhandy for the script +The patch is split into two +Nevermind your work can be used by other persons for their own projects if they wantcan and maybe we will find ASL compatible fonts in the future and then we will be able to commit your workThanks +Also capitalizes the SQL keywords for in plan to commit the v patch unless there are further comments +Im still kind of surprised this would happen we should be retrying on connectionloss up to an expiration which would make us the leader no longer +More concept +Unfortunately its risky our project is already in LIVE +So all issues not directly assigned to an AS release are being closed +Reopening to move to as was issues after move to in +Or are you saying you see this not fixed in master +The issue introducing autogrowing was +A solution would return an ordered delimited list of all values +Depends on for the new CP hook +patch for fixing committed the additional null check thanks Saeid +On Linux everything works just fine +Not sure if i did something that really slows it down +Benoits rant after Hadoop made this type of change on a minor release is a fun read +Delete to trunk and +Thank you for sticking with this Francis +Bulk move a null check before calling or ported to branch revision on trunk and on close for +trunkCommitted revision +We are currently working on version which will break binary compatibility so the patch can be applied +And while the JAXP XML parser library setup may work for your general out of the box application I think it has serious shortcomings +It might be best to either investigate what common container is doing with the error and submit a patch to Shindig +If users complain about this in we can consider back porting +changes in JBDS +Logged In YES useridSorry there is no requirement that mbean attributes beserializable however convenient it might be for particular problem you are having I hope to address soonin jboss by exposing the itselfas an mbean the mcf properties all of which areserializable will appear as mbean attributes +bq +The patch might break things in the situation where different instances should go into the cache for the same field and same prefix but a for a different implementation despite different both instances end up at the same cache slotAgreed what I cannot understand is the relation between the and the provided field comparator implementation +to have too much on ones plateexample Id like to help you but I have too much on my plate at the moment. +My schema is a legacy schema on DB AS +I think I know what is happening hereOur current server is not suitable for async io yet from the client +Just build off the branchescommunitySeam branch? +Kunalkumar can you please illustrate the potential parent POM problem with an example? I cant see a problem there but perhaps I am missing the obvious +Integrated and fully working +Thanks +The feature I am addressing is that SDK on the one hand comes with endorsed standards and implementations and on the other hand allows one to override those by specifying endorsed dirs + clean +These may be removedmodified but need a starting point +IMO the issue is not chunk vs +Rather than deprecating this could we just notate that it is an expensive operation and not for normal operations? Or even only allow it to work on ROOT and META +This is fixed in the line +I think it should be expected in the path so my vote here is wontfix +Hi HenriThanks for reporting the issue and taking the time to propose a fixI opened a separate issue on the specific JIRA project for Shared where the DSML code is located +This problem is still there +Typo +I have started the regression tests +IMHOThanks anyway Beans Validation and check the goals of JPA +Wed need to change the version string if we changed the schema in some critical way thats impossible to detect by examining the schema itself but I have a hard time imagining a change like that and we could always instead +Most failures seem to be due to limitationsbugs in derby not problems with the dialectOf special concern at this time it appears that special characters in table names do not work I was unable to get identity columns to work +Ive attached the patch +However to get the build working and to fix some other potential issues I had to modified your patch a bit and hope that did not make any changes to the way it worksWhen you submit a patch next time try to do in small scale that would help both you and the communityThanks DeepalThis patch provides a way to invoke a POJO with generics +Diffs file that corrects the DITA errors in two recently modified adminguide files. +This is because our httpd config is large and reloading it all the time isnt delay is around minutes from when synchronisation starts which is normally immediately. +This will lead to very unmaintainable code +Specifically around your ideas for heterogeneous cores and +Backported to java and applied. +Well update when its available which should be soon +I still remain a few of lazy approaches on some classes that deal with some protobuf types being updated frequently during the runtime +bq +OK +OK I opened w my current patch. +Should be done in since this is backwards incompatible +is the datasource are the jms dependencies was still not looking to the correct meta data +Pretty small changes passes unit tests locally +How exactly would split differ from the TOKENIZE function if split returned a bag? TOKENIZE returns an unordered bag of words +hoping that cmd C would remove the quotes but again service had C +Duplicate ofThanks Rick! +Note No new test code required changes to documentation only +Thanks Shami. +Fixed! Thanks for spotting this +New patch changes thread model for as it was making ZK calls from a watcherAlso cleans up some code in replication tests +Yes Mike you are right + there are two parameters to control outliers one is threshold and the other one is emitIf possible I would have liked to get rid of emit in future +Close issue in preparation for release. +The correct solution here is to use a transaction to rollback the messages +currently the link is not shown in the admin page if the handler is not registered +Understood +Since theres some work recently on the Emacs support perhaps this one falls into that category +Switching to in as well +This way if the enqueue operation gets interrupted the fetch offset is not advanced +Approved for inclusion in. +Add webui change in this jira itself since it requires very small code change +The compass index case itself was not working and I have fixed that alone in the attached patchRename the file props to package run and you have a jdbc compass index +Thats a good point +Please ignore the last patch +Update as per the resolution of it shouldnt be needed to create a separate layer anymore +Yeah +Thanks for already adding the samplesbuild is the target forDo you need it? I can attach it +Patch for review +Confirmed it is working as intended now +fixed at great +The previous patch contained some mistakes +Keith are you already using short names for the types mentioned above anywhere? If so can you please send me a list +Im running tests locally for now +what is the Identifier what is the Properties what are the associations what columnstables are these properties mapped to etcSorry if i did not make that clear enough earlierAnyway im fine with committing this but i would prefer its focus would be shifted more to the ORM part of the classes +Changing fix versions to since it is now part of for move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues looks to have been resolved by the various XML parser changes +got this out of jcarder looks like the same thing right? also can confirm its gone with +File with exception test just create a jpa project in workspace with Hibernate platform and rerun the workspace +We needed to update the repository location in and the Axis handler API now requires us to return a flag to control further processing +Thank you for reporting this issue and I can reproduceidentify the problem now +the patch file that I uploaded to Review Board +vs +javassist has big problems with deserializing +When is it planed to load the dump file in to svn? And please let us know what support we need to provide +is old and it showsThis is something we should fix in later versions by getting rid of and making reference counting something done in the repository +passed on my local machineI just committed this to +We had public holidays Friday and Monday +BTW here is another patch +Your approach of parsing the commit messages from an SCM doesnt fit well into that +Thanks HyunsikI just committed it to master branch. +This is marked as resolved and the github PR closed as merged but the change is not in AS master and the git history indicates it was never committed +when i use query select from cmisdocument can query only document thank i use query select from cmisdocument can query only document thank is nothing we can do inRefer to the documentation of the repository you are using. +SOA Doug +I should have it ready in the next day or RossenIve discovered why I couldnt reproduce the issue it is because Im using hibernate and the sample I uploaded was based on an class +As per op. +Hi IlyaRebasing the patches against master instead of branchAs the feature is NOT targeted for it makes sense to merge the patches into master rather than branchCurrently testing the rebased bits would as soon as verification is doneRegardsSateesh +Im happy to see that someone cares about my jobAny help is welcome +Nick Ralph Gary is this in scope for the GA release +Tested +Can we close this issue? +For the non blocking IO feature a new issue will be created! +Class needs to be public according to the JAXB spec +Hi AtulenGB deCH ptPT and zhCN are fine +But I cant think of what we should use as the keyword in the configuraion string +The resolver is able to pick up the bundle if the exported package is versioned same as the bundle version +After reading the comments above and doing some research on the Jakarta site it appears that the bug was fixed in Tomcat +Directory Structure of the Generated +Current unit test code coverage of SAAJ according to clover is Please apply thisNice workI applied the patchThanks a new dependency in the SAAJPlease review and commit this patch +Should we replace them with ping for consistencyIm also seeing some tests failing like and but Im not sure if this is related to this patch +While this code might get rewritten in the future under the current preemption mechanism when MR is explicitly told that a container was preempted it should not count it as failed +Temporary workaround to set early in the startup process +yeah a bit smaller sounds goodand maybe a transparent background? Right now its really catching the eyeand that it shouldnt +In general you still have the class prelude in many cases where you dont need need it +know whywhere they came fromi have rebuilt them havent uploaded them to nor as the hib components maybe need be rebuilt again once JBC get upgrade +Logic behind or Searcher should simply think about the best logic how to optimize what the user wants to doMaybe I should create an new JIRA issue out of my suggestion to merge Filters and Queries? In my opinion this is something nice to have in +one more think I think this is a general problem that exists before so we might need a entry +I like your suggestion better +NITS can we remove conf field from as it is an unused field? remove unused imports from Isnt it better to put state as a last statement in the delete method? or any reasons +Reproduces the error this ticket describes +I couldnt find any when I started on tranql You should be able to make an similar to the but using locat txI dont think that only rejecting a few known exceptions is necessarily a good idea +Could you share the functional tests that your QA team wrote? How will other developers know whether they broke this featureIn your experiments does a machine with only a single functioning disk warrant staying up? I suspect tasks on this machine will perform poorly +I think that at least javadocs should be updated to reflect that if you want to use speculative execution the base must not be a path but a name +we should wait for the rpc client to be created before doing UI +it is not a to future as not a regression +Is there anyreason you reverted this back to OPEN? In my eyes it is read for review +This one changes the default behavior to make the service respond to any valid hostnameip address +Ill take a crack at updating the patch such that too many entries arent removed +svn ci m Added javadoc to show that can be implemented by default and proved it in the unit testSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +This one changes only contents of the log +If is applied I can try to make a diff for this issue only +Attached patch fixes this issue so long as the shared dirs are marked required +For all tasks a +Obviously it would be preferable not to have to do this +Thanks for the patch Gwen! Could you also provide tests for the fix? +The spec has removed this API +Can you tell me if when youll work on the vnode support? in a few days? a few weeks? a few hoursThanks may take a few of weeks +I really think we should fix this before releasing any version that has a potential of becoming Hadoop +Verified following scenario with latest build fromCreate Zone with Xenserver using SRegister xen templateAdd zone with Vmware using the same SRegister vmware template +Attachment has been added with description FO input FO document containing a nested table with two columns and a link withineach of them +I committed to Tez branch not trunk. +Ill look into it +Actually I wonder if weve tested cluster restart sufficiently +Im not closing this JIRA now lets wait until the end of your investigation +Configure Hibernate to employ Cache as its nd level cache +in + +ack +for the patch +On the other side I agree that its more nicely when user can drag and drop palette elements to visual part of VPE +Otherwise unchanged +failure which I think might be caused by my test enviroment +I will commit this in a bit +Maybe then we can start to see why this is happening +I dont think we can easily trim off long records in binary streams +Could the original reporter please split it into separate issues +this didnt make it into so as blocker since the default created server will be without this fix in +true +trunk xThanks Hoss for taking your time for this issue +Now group Search do not support distributed queryanyone else has already been meet this +One common failure mode is that some nodes fail to get the update of a new jar file +had to the commit there were some test failures on the of the changes are in Im now working on a patch for that should bring forward anything I might have forgotten +We wont migrate to Confluence Or what exactly did you mean +it is up to the user to make the Dispatcher be applied to the request which is a requirement for the Sitemesh plugin. +Temporarily + +Failed tests and are unrelated to this patch +Ill work on those +Expect an update sometime tomorrow night ticket title to more accurately reflect the root cause of the issue +xuhongboYou can get the Spring Application Context Classloader from theBut I dont think you can get the from the nowMaybe we can do some improvement in +Committed revision +If not I will check as soon as I get back in early September +Sending srcjavaorgapachecommonslangtimeTransmitting file dataCommitted revision . +Deferring this will always seem like the pragmatic choice until it becomes all too clear that it isntDeferring to for now. +should be easy to fixIf you know you are reading a ZIP you can certainly skip the autodetection version of create as a workaround but you aleady knew that +Its sloppy coding +bq +Confirmed no other method sets in r. +The does not seem to have too many issues please give it a try in case you are interested of a bit more advanced restart script +Logged In YES useridI do not have commit access to anymore please to someone else +This because we use in the +the logic could also be moved to tearDown if I use an instance variableReady for initial reviewcommenting +The most recent commit was there is an issue tracker a somewhat active Sharpen user forum and have accepted outside patches in the not so distant pastForking might not be appropriate unless its just for the purposes of working on a complex patch +As long as you document that explicit in the INSTALL and the configure output this is fine +I really like that you ran all the code in your path with this mode as default maybe worth doing that one last time when you have everything checked inif it makes sense for you i will wait for your next patch to do detailed review of the rest +Godmar +Attached patch updates the README +Linking this to the JIRA to keep track of the FS work +It changed the meaning of from the actual bytes to the percentage +New variant of old logoThanks +Patch added +Should we upgrade to for the EAP BETA? Dimitris says it contains only this fix +Ive just committed this to trunk andThanks a lot for the contribution Andrew. +Another proposed patch If the service is not currently bound bind if the filter matches and binding makes sense If the service is currently bound check whether the filter matches and bind if it matches +Hi Suhas LohithWe are students from University of Minnesota looking for a project for our distributed systems class +Oops +You have toAnd a nextInternal in an Interface like seems wrongI am for speedup but as you say this one is ugly +Now the question arises Should AXIS always use SOAP by default even when the WSDL parsed to generate the stubs indicated that its bound to SOAP? I think the generated stub should take whatever steps are necessary to enable the proper SOAP version specified in the motivating WSDL document +I am concerned that we might not even be able to write a Unit Test for this bugThe essential sequence of events for this bug is this +Alessio do you know what has to be done? Will this be OK for +It should only be returned on a shindig query for activities and has nothing to do with any action in rave +Ted committed this +I am attaching a new patch that has all the newly added files in it +Camille I actually like the change in the first part +The problem is EAP specific the AS repo still contains the that correctly references. +but I thnk the last solution was a good one +Switch return type in factory method and compare console output +OlegI agree the patch is pretty much ready to go +well by real code i also mean something small instead of a three page code example +We are going to integrate Apache CXF when it becomes available +Actually I spoke too s sorta working but the aspect isnt being used were getting closer + +This patch should do it +There are also in the code on properly handling persistence of changes and things like updating jcrallThe reason why I opted for the bare minimum change for now is that I didnt want to start making more significant modifications while you were on vacation +Attaching patch to +I dont know enough about the class to determine if its a bug +My error it works fine +Thanks for the help with testing thisCommitted to trunk as svn + is in the jbpm tag +JacquesI think that we are in the process of standardizing the UI all over Ofbiz +Do you have a specific use case where you cant call instead +As Ben stated above Tomcat native clustering support login info replication while Geronimo cant +This appears to still be an issue in hitting home or end outputs or to the console respectively. +What is the license of? Can we included it at Apache or should it be at +FWIW I have not hit this in my Ubuntu dev environment +defer all X issues to +to the doc changes +Anyway Im doing a fresh checkout from dev CVS to verify the state of my code checked the developer CVS and the new classes are present +Committed to +We could add ESB Drools and other runtimes +I dont think they should be part of the user APIcan I delete them entirely until they actually get used +Ill have a look at later this week +can we provide a convinience method on watcherevent to get watchereventwrapper from it +Its actually smaller than it appears +Work in progress patch +This patch contains implementations for assertion and assertion +It doesnt work in JBDS runtime with name Runtime exists before mentioned ESB example is downloaded only Targeted runtime for helloworldtestclient project has to be name of the runtime is different then there are lot of things to server runtime and esb runtime are defined pointing to SOA attached files +It seems that the committed patch causes some javadoc warnings +Something is then mishandling it leading to a parsing error! Further with logging on it seems as if the entire transaction happens successfully and then the server somehow sees an extra entire input message with no contents and thats what blows up +This part of the file is obsolete and is not involved in the deployment process +The customer is happy to test the patch for us but neither the upstream or apache fix has a test case for this either +Jans suggestion can also be used at runtime +Jeff Barrett will be contributing the patch for this issue +Updated the skinning document also +impossible since the is ignoring the hashcode method and is going directly to a native method to resolve the hashcodeIf you discover a way to test this and get it consistently to fail I would like to see how you did it +moved out as DB not intended for this is not supported with +I think at least we should not squelch the error we should catch the FNFE as well and probably use a better way to find resources +This week Im planning to have a working first cut +I have the refactoring essentially done this required a little more refactoring of since controlling timeouts requires manipulating parametersBtw it may be useful to provide a default set of that can be modified on a global levelIve added some additional tests for HTTP auth that verify that the revised framework does work as expected and will be checking these in with my latest changes shortly +Im gonna bump this to since Im not sure what fixed it on trunk and its too late to do the hunting now +No tests are needed as adds test for this functionality +Okey dok heres the patch Ill post some sample queries and response writer config to show how its used in one sec +Please move under if your target is +I am not familiar with Karaf +I am having this problem with Maven but it looks like it will be resolved by. +Good point +FYI +Thanks Aaron! +Pat yourself on the back and take a well earned rest +It seems there is a bit of a flaw in the way this JIRA was done +Proposed how do you distinguish from other repository errors like session closed +Thanks for committing Ryan. +Array of groups positions was initialized with wrong size what breaks references when try to sort +Some edge cases still produce warnings +But the exception is gone +Wow totally forgot that there is a class +Bobby So this is an issue with the events being logged in the job history file +for each Extension class there will only be a single Extension loaded +This patch asks the user for a keystore password if it is not configured in The case of a missing key password is not included in this is a second patch which asks the user for both storepass and keypass without echoing them. +Anyone have any opinions one way or another about this one +Correct my username is fx +Does the new reflection impl rely on code patching at all +I think that we may just check that jar file name starts with +It should be available in the RC snapshot releases +Committed to as well +So Ill definitely factor out a few template methods that allow for easy overriding +Will commit soon as trunk build restabilizes +once hudson test pass +lgtm thanks +As suggested here is the related deploy plugin issue is now addressed in Maven +Attached patch checks the existence of vertex or edge input path before it kicks off a job +Once more with licence grant + +I didnt think about it +One run took seconds the next seconds +This is an initial patch for a JWT SSO format and basic authority implementation +Thank you againNOTE the attached patch was not the one applied +Thanks! See follow on jira here released issues. +I removed my copy so now it should compile +Better yet would be a +IIRC I removed it coz it caused some issues when verifying trust of a cert that was issued by a CA in the cacerts list +In case of shared classloaders those static objects are still reachable even if the container tries to unload the app +chapter is updated +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Please check and close the issue +What happens on startup? Any exceptions or error log snippets would be fine to see +This is just a generalization of the local flag but now you can specify dc dcdc for instance +we wont be doing anything like this in HB so moved to would be also very useful for overriding the behavior of for trimming padded spaces for database column type CHAR in several databases +I debated hardcoding the Hive types in to avoid the dependency but since serde is already required by hivejdbc elsewhere I used +Thanks +passes individually but I dont actually see it as part of +Better fix might be to explicitly exclude it in +Standards compliace is a good thing so I would say we should change it and have a release note that mentions these changes and the changes in togetherOne problem I have come across is that the network server calls these methods on result sets +jpeterka Snjezana asked which examples you had problems with please let those then im sure we can get this fixed quickly +Looks good +Steven patch looks good to me +Fundamentally there is zero value to an app developer from triggers vs running the same code from a storage layer +Patch with test case for new allow flag +It will be included in +The blocking test failure has been fixed in other JIRA resubmit v patch as again. +Ive asked Emmanuel whats up with the versions of the JPA API +The patch is applied with minor improvements +I dont understand why exploded deployment should be affected +I have also seen this test failing inconsistently +Problem is that as part of the issue we have decided to change the way we assemble JAR files +The crash occurs on my production code any time is linked in as a shared object +Okay its done thanks for your help! What must I do next? Resolve the issue +Bulk close for +The core functionality is done for the only remaining work is to expose an API to allow the user to set which state transfer version they want +I agree with the assessment +If the error message is shown on the attempt the account gets locked it may seem that the account was already locked prior to the then just mys not relevant to the user getting their account locked how many attempts it took. +Ill reassign it back latter to me or to Max Andersen cause the issue Hibernate Tools related but Alexey we need your opinion and possible fix in other part of JBDS cause this look like bug in several places of JBDS +and fixed +fixed +Yea not sure why we forcedly overwrite the user defined timestamp when its present will look into this +Can you please your Spring version? Spring simply does an instanceof check +Ok sorry for this +I actually think that Snappy compression belongs in Hadoop along with the other compression codecs +Let us know how it works for you Michael. +ill commit this one shortly if there are no objections for in rev +I need to add a few more changes to make the patch complete will supply a patch once done +Well done Julien manually tested! +b +If the would keep the original token and add a lowercased token on same posIncr with tokenTypelowercase we could support case insensitive match with preference for correct caseIf user needs different boost for different fields perhaps the name could be configurable on each filter +bq +Fresh Findbugs report for previous patch +Actually I think its more major +I have experienced various problems with the cygwin cvssvn clients on windows +Current development is located at the trunk area of the SVN area +Reasons to let meta split other than because the cluster has millions of regions Distribute the load on meta +Is it possible to close this ticket? Ive replaced Spring data with service objects +When you say its intermittent what exactly do you mean? Is this happening under continuous use by users? Is there a chance youve got session timeouts occuring on the server? When you say concurrent requests are hitting the server do you mean from multiple users all with single browser sessions +We will try and get that changed or introduce custom namespace features that work around it but thats another story +Could some please temme as to when this will be implementedAlso can I move this feature from Minor to Major as this is a very important feature +Sorry +Thanks very much Dyre and to Bryan also for reviewing this. +Moving all of this to +Just downloaded and fully rebuilt the SVN version and the carselector bug cannotbe reproduced +As for the deliverables and quantifiable results your goal is to contribute your code to the Apache SIS community and hopefully to become a member on the project management committeeAnd of course to have fun +It has the nice advantage of removing lots of code that is now generated by +Thanks for catching thatCommitted patch to resolve defect +pending Jenkins +Closing this record since is now released. +A class would have to be added so that the tools are added when the view is visible +There is only one thing to do drag the right bottom corner +bq +Can this be addressed soon +Note doesnt build on OS X so avro wont buildinstall either +Attached DSA public key installed +But it also comes with the cost that each mappers then has to deal with larger portion of the large table +Ah ok i got the issueI created a new Security Group which have only the PROJECTADMINALL permissions +Could you please review and commit it +VanPaulAny final thoughts on this? Or will having additional debug logic be enough +Works better for me +Splicing done. + method was removed in therefore I removed the javadoc markup as that comment is still useful for legacy purposes without the link +In its used to load AST node classes I think this is less of an issue since they are only groovy core do you think Andrew? Would using the SU classloader be ok for you +Thanks +Sorry I didnt mean remove the with the cores on them I meant the constructor functions +Attaching the patch for review +settings file + Added conf folder and example of on wiki +Update the archive seems to be fixed Matt +In that case I agree that this should be addressed before the issue is closed +The default interceptor should have just been the registry this is my fault +Not that nice but def +zip and +I doubt they will have M usersNot just M users I can envision probable applications with M users actually +Oh I see thanks Bilgin +This leaves Collectionarray merging methods then method useful here would be array and collection cloning useful when creating defensive copies for getters. +Use plugin manager and interfaces from and configuration plugin changes to implement updated ACL handling +get is protected in and can be only accessed by classes in the same runtime package of by relativesThe second is not our case so putting test class into bootclasspath will solve the problemSo specifying Xbootclasspatha instead of cp is the solution +A sample app that demonstrates the issue would be helpful along with information about what you are using to detect memory leaks and any leak reports that you particular tool has generated +Could you please check that revision contains an adequate fix? Thanks +bq +I also need to double check and make sure that are skippable as well +Patch for review +might be different between the two versions because any patches since release are going only to the new codebaseSo any changes you make that allow general hive storage handlers to work from hcatalog must happen in the new packagesHowever the hbase hcat storage handler is part of the deprecated set and is retained as part of the old +Marking as resolved since and have been closed +Not a bug in X +No job can be run because the jobtracker stops as soon as the namenode comes out of the safemodeThe jobtracker log is attached +Rebased but without the changes from Sylvains review +librecords +thanx for the fix ivan +Yes we should probably rename it in the and even allow for a configuration in case the user wants to control the name +Simple UI only patch +! for the patch +Ill update it soon +good find Daniel +Negative unit tests now work correctly +Ive looked into my last jar because of the missing class +I dont think this is a good idea +This is superseded by the HA task +If this status is raised to error then the application will not run +The good news is it does work maybe someone can take these ideas and clean up the implementation? +Dont know when Ill be able to look into this +I have a fix for this +Ive also added a regression test to ensure strftime its working think this could also be marked to add to this patch doesnt solve the problem completely +for doing this presently +Thanks for catching that +AlexanderI think well need to use Hadoops Mini Cluster in order to have a proper unit test +Else using class +if informationabout column collation was addedAnd does each indextable have a set of properties associated with it that could be used to store metadata +That feature remained unimplemented in our search participant +We want might to try and remove that limitation but it requires non trivial changes to the code and my initial attempt led many unit tests failing +Fixed in trunk r somebody just needs to back port it +I find this ironic with regards to your response to +please attach a sample with steps to t gotten this issue in at all +This one is hooked into instead of +The only ports that jbpm uses is that is used by hornetq task server and the rest is what application server needs +Committed to branch +Please take a look +I am inclined to keeping the modules separate as it is currently instead of combining the source treeI am counting the no of modules to be +Thanks! +If it is switch of we will loss the whole history +This seems to be working for Burr on Windows so maybe it is a Linux problem since Im running Fedora ? If so it may not be so critical since Im guessing windowsmac are the most common IDE platforms +We could probably comment better or mark issues as in progress but I looked at the issue pretty soon after it was report though it did take me a few weeks to getting around to fixing it. +Help in resolution is appreciated +You meant the scala libs are bundled not the whole thing that is about M in size right +Yes +So if I read that right KV compares take ms without the patch and ms with the patch + +In particular x id that signals predefined object type can only be written for a if at writing time anything other than null is returned from the predefined object type lookupSince were dealing with a concurrent collection Ive played around with a test that marshallsunmarshalls a CHS while another thread addsremoves to it in case the issue is due to concurrent modificationmarshallingWhen you run your tests enable TRACE logging for class +This bug has been marked as a duplicate of +Many thanks for pointing it out +Fix these and unused imports unused imports unused import just committed this to trunk. +But after this I will spend some time on converting more tests +I just committed this +Im a bit confusedThe jaxws frontend and the SOAP binding should definitely be importing the packages fromThe frontend should definitely be importing everything fromThus the description of the problem is certainly wrong +Actually it seems like the outer select just doesnt return anything to constrain on when the inner select returns zero rows which is true when there is no data in the db for the specific schedule +In fact I think we can just have setcqlversion complain if a client asks for CQL +Earlier the AM would receive proper reboot command from the RM and would shutdown properly based on the reboot flag being set +However as Santiago points out this problem can be worked around by changing the way you define global variables +a good new contributor start doable indeed I will look into it +Also Thank you Lars George for supporting Thrift +bulk close of resolved issues. +According to Chinthaka it is not a bug it is a valis REST case +Resolving this issue tests no longer failexist. +I backed out a change to until we understand how to make it build correctly in those circumstancessvn +Yeah works great +Hi JBDo you mean remove all jaas bundle from and put them into a featureIf so +I could add some black magic into the daemon plugin to make the path to the log file absolute and set by the installers but this is a real PITA +Im not aware of any bugs in that are liable to cause a JIT optimizer to crash +However this was fixed for observer methods by changing to remap the method when a proxy is being used +actually passed in my machine +I only have a chickennegg bootstrapping issue to fix but I badly need a bit of luck +Hmmmm +I updated the test and attached the patch +Merged into patch for just committed this to branch +This patch is to be applied after the first one +The second properties file has mixed Our castor jar has lowerDepending on which one gets loaded up first in thecontainer we get different behaviour +Issues resolved in + fixes in +Based on the performance test you did the write performance is not better +If the user didnt want it repeating for every page they can move the footnote into the first after the +committed v +Test case attached +This is a duplicate of +Are you referring to +Thanks EduardoI think you were using illegal syntax +I will commit this work soonThe behavior implemented is not exactly the behavior described in this bug +Can we remove the tildes from debug outputThe logs with s are going to be removed completely +This should work in your scenario as well will be available in tomorrows +The tooltip plugins taglib was moved to the package +Ajaxjsf does not exist anymore +without breaking compatibility +Since were introducing RPC version via I think we should do this via +Sorry for the inconvenience but I did not know that jspringsecuritycheck is an official endpoint for starting an interactive can patch the datarelasesecurity to switch the default value to true +But to clarify the intention suppose theres a emulator device A that fires up by default +This patch adds two encoded in two different ways +Not necessary but harmlessOver in I will deal w the ugly ERROR when I try running against cluster +committed. + remove the deprecated methods +Ive applied your patches thanks a lot for your contribution and welcome to the list of contributors! +May not be specific +we have testing the patch in trunk for hours without any crash Id like conclude that we have fix the issue +Types are defined using array so those are already ordered +Trying another approach +Not only when you deploy Ruta projects to other machines but also with the svn changes of the descriptor folder when several people develop the scripts +I vote for a test module in +As of May using the artifacts Im still seeing the same error as originally reported +Will send you the dump when I see the crash again +Regression tests added +And by the way I ran into another issue with the Groovy jUnit Tests +Can you run test and test with single setup of +Colin thanks very much for the comment and the cleanup! I agree wed better get rid of the JNI code for libwebhdfs +closing +I wrote gkesavan to ask him what itd take to get this working for hbase +But some still are not fixed in this realease for now +It will be a change right +In my testing this has prevented +Unfortunately this changes the default rendering behaviour of all Menus which will require applications to change there css style sheet if this is not the behaviour they wantWe need to give this some consideration maybe change the selected class name to menuSelectedregards MalcolmAre you saying that the class selected clashes with other defined stylesIve been wonder if we shouldnt start namespacing our and CSS to avoid this issueFor example Ive already started using the Click namespace for +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Another option would be to walk the entire tree looking for dirty inodes +dcdebebdfdbababecdb SPDY plugin crashes on connection closeThanks for testing! Fixed the crash in this bug lets open a new one for the SPDY connection issues +This can be dropped into a cluster and referenced at httpServerPortitpa +Dave +for change +I have to move this into common because it depends on common +Closing. +Still doesnt pay attention to position but I dont know enough to know if that is important +My take is that we have injection already via +This issue must be resolved for all releases created and distributed on or after the st of November +The bug has been reopened once because it wasfixed incorrectly +I did some testing and it used quite a bit less memory on equivalent patterns than and a small but variable improvement in time +Heres the patch Im going to commit its dead simple +Because as I realize now asType coercion is used for the as operator while the maximum we want to have would be a groovy cast +I think for the sake of this feature its best to stick with the established Solr Cell convention +Thanks for detecting the bug Fixed in SVN headTwo things can you send as a test case that exercies this code pathAnd the last one is only me but near line class seems very big +Updated patch for POI beta patch including version bumped to beta and patch applied in r. +These are the installation instructions to install the additional jars required by appfuse into your m repository +This looks good and is very similar to the design I was going with +I have merged the queries from and in the new tests +I dont need to sign off +Since this is cannot be reproduced in trunk the issue is closed. +I was just working with the Oracle driver for some of the Apacheconexamples and all worked fine +Just found your repository for the poms +Added a highlight box on the getting started page. +for trunk and for branch. +The stacktrace does not refer to trunk +The issue appears only when using the blankfalse constraint +Until then we can keep it to say that we gave up on the new Dojo tree because the version is not compatible with the previous +reopening so I can modify fix target to +Done +Im reducing the priority this is not a major issue +Can you give an example of what is not bing handled? The code consumes a complete HTTP header in the code between the +OK I see +If you are generating an URL that is going to be used by to make an Async Request then dont let the s get encoded +Konstantin I would be good if you add test for in hours +And getting the results you get at is not surprising +However numericTypesnumeric have package scope not public scope so I think we should be ok here +Brandons advice I moved the entire warning into the logj call even though this makes it unwieldy and we dont have perfect information as to what the cause is at that point +Will include in +Verified in +I was just too lazy to search for it +Commited into trunk in revision Adding srctestjavaorgdrools srctestjavaorgdrools srctestjavaorgdroolsintegrationtests srctestresourcesorgdroolsintegrationtests srcmainjavaorgdroolsreteoo srcmainjavaorgdroolsrule file data +Heiko can you take a brief look at it? If you get into trouble with reproducing it please ping me on IRC +The package private model does not work anymore +It just must have been a new codepath I was executingDropping priority and changing the type from bug to improvement as this was what has always been the case. +Paulex thanks + provides a patch to that can work around this issue by setting the class attribute equal to the empty string +A JIRA has been raised to provide a generic way to do this and we also intend to submit a proposal for this to be included as standard inThis will be dealt with in a later JIRA as will documentation of the new config options raised by this change. +Fixed now in subversion +However theres a question +By default there is no prefix or suffix for error messages +So Its not that the +older issues for Apache since were no longer actively working on these at the moment. +patches added +Logging in the normal return case got downgraded to debug in +Im currently running out of my git repo but I just pulled today for the first time in a few weeks so Ill test it out again and get back to youThanks pulled down the latest stuff from git removed my change then tried to upgrade my project to +to use the Provide Path button Huh? Maybe I am missing the obvious but where do I find such a buttonI have synchronized with trunk this morning but dont see any differences in licence header +All reasonable attributes of richeditor should be available in supposed to close +push to +Actually never mind +Thanks +Fix a test case that was leaking a file handle +However when another component comes along that needs this causes to be activated +The integration test can be used for a manual verification that the aggregation indeed aggregates +Applied to branch and trunk +Actually once we have networks of brokers working wed be able to solve this easily by using an embedded broker in the JMS client which talked via the VM transport to the broker and used JDBC for message persistenceWed just need to use a which returned the same JDBC Connection thats associated with the current transaction then wed have database JMS XA without using would say this is a nice to have +comments on is available +appreciate any feedback +Im sorry I wasnt very clear +We will no longer check that injection points can be injected into decorators +Manik Im pretty sure thats not the case here because it uses the same Transport instance throughout sending the request and reading from it and all socket instance variables in Transport are final and were logging what transport is reading each time +Testing first before commit +Im going to commit this unless objection +Agree with Neha that is not aproblem for replicas that are in ISR however we do need to theISR change listener for those replicas that are in ISRFinally we should probably open a separate jira to implement a feature tocancel an ongoing reassignment given that it is a operationThe option reduces the need for this but nevertheless I think its agood feature to support in the future +i didnt refresh the jira before my last comment +Seems very reasonable +Ill get it working eventually +Found issues in the test so the real issue is a dependency on +ThanksWe need a few tests for this +Thanks Arpit +Thanks for pointing it out Siddharth. +So the thinking was perhaps the new version of GSE that was reworked for might make this not a problem or need to have a slightly different solution at least +It is the responsibility of the protocol handler to terminate the session properly in case the channel has been closed by the peer on the opposite endpoint +Blah blah patch using mdo file and refactoring per in rev all is fine for you with this +bq +Attaching simpler solution consisting in not transforming expired columns into tombstones so that the second phase of sees the same columns than the first phaseI would be happy to come up with a unit test for this but this is a subtle race condition and Im really not sure how to write a test that capture it +Renamed slingujax module to slingRenamed ujax to sling package to +Tair What use case mandates that? I think if we do value equality only based on GAV that should suffice do you see evidence to the contrary +Thanks Tim +Also there are some concerns about using events nanos as a default row key +Thanks David +This time the language model factory includes support for Chinese +This issue may have been fixed and code may already be present in the old trunk +This is still a problem +Just committed this. +Soright +And you need to take special care about Algebraic funcs and their return types +It definitely does not behave like the fixed version I tested even though according to the dates the fix should be in +I seem to recall a discussion that should not report JDBC yet +My comment in this JIRA has the details on what the issue is and at this point I havent investigated for a fix +Thanks Gary we do need them to propagate across the network. +I will take a look +Ok I got the Decorator to work for Excel by d like to do the same but is not in the +java and +Also the patch is quite simple +Is this still a problemCan we have a unit test +pending Jenkins +Please try again with +Ive added a final piece of information to the FAQ +As I cannot figure out a way to I created this clone issue +Ill commit the simple fix +Passing back to on JBT CR +Fixed sample branchdrseussbpmorchestration tcunning svn commitSending bpmorchestration bpmorchestration file data +It can reduce time which spend on TLS access on some machine +bq +Carlos is this what you did today +It is currently just using the default but is now very strictwith regards to the license +Inadvertently closed the issue +rebased the patch with the branch Brandon you dont have to submit the patch since it is not for trunk +But how come the pom was originally deployed incorrectlyIs it a question of ensuring the correct SVN properties +remove it from as no urgent need in +Ok Ill start going that direction then +Its currently not clear in the javadocs what is expected +I looked at the webwork stuff tried to patch it but couldnt +That is if you deployed after the other services had been discovered the cache used by the plugin was stale +I am not able to deploy a project even after marking it as only workarounds to this issue are When creating a new ESB project set the ESB version to After a new ESB project is created edit the projects file to change references to can probably take this issue if you link me to places to get all the runtimes and clearly list a step by step +Do I understand correctly that the test case listener was commented out +The critical task here is to validate the behavior is correct if the reload operation includes the false parameter +In everything is done automatically but the user will have additional file in his project folder +Please use the updated file +Interesting +It is not clear to me if it addresses my issue or not +Please watch if you are interested. +Yes the decode splits on comma +See my new comments on the test as it is now depends on and the patch itself doesn far limited the test for cases where deletion actually seemed to have passed on my box yesterday +The next step is to have the creation of store files honor this region placementIs this patch useful without giving hints to about block placement + Bug has been marked as a duplicate of this bug +Hello? Is anybody aliveIve seen same bug under russian +properties +Hi DagThanks for all of this great work +Would love more feedback from other Camel riders about this +if the table of the OUT JOIN is not placedimmediately before the OUT JOIN phrase +Hadoop was released today. +integrate feedback from early testers +Solved in CR. + +The number of times a JMS provider will redeliver the same message before giving up is +Set fix version to assigned to me +Yes we can definitely make it work with the being added only to deployments that need JSFHowever I think we have uncovered a bigger problem with removing the from the server classpath +If in the future someone would like to tackle this feel free to open is and attach a patch. +bulk move from M to M +Andreas Why is this necessary? Can you provide an example of when youd ever want to do this +NOTE Sorting doesnt yet work with my modified but that is not the important issue in this caseIt doesnt appear that the attribute is ever used during rendering. +What happensIf the system controls ids this never happens because they are unique +All the other items should be fine for me to change +Bulk move of M to to this issue no more actual as we have css editing perspective with two views. +In the meantime if you have a chance to try it with latest release I will be interested to know how that worked +Ok I replace UTF with TextShould I also remove its use in other log methods logRename logDelete and log +Closing this issue. +i committed a fixing the described buginto cocoon treei have tested the fixed version and for meit is fineyour reported is not fixed yetif you dont mind can you update your Cocoon CVS tree and recheck tooFix for Cocoon follows if your rechecks succeeds +Added a deprecation d like to log a request for you to consider keeping it but leaving it deprecated +I noted this morning that the testsuite is still failing with the same error so Im not sure if its running the latest code +Ive committed to the snapshot branch + for the patch +Ram Thinking on it the way to plug the hole you and Rajesh have identfied is by having the RS update the znode to OPENING form OFFINE before returning from the open rpc callThe way I see it weve found problems in our +Reattached after adding State of the Unionby Marco RietveldThe schedule needs to be reviewed and finalized + +Saminda was trying his ssh key password and not the apache one +I did some timings loading the example schema and found that of the time is spent loading the fieldtypes analyzersOn my Macbook Pro reading in the whole schema takes over ms the first time gradually reducing to about ms after or trials +If I remove the local repository and +how about calling the extra attribute description? description is fine too +Youre right it should be documented +backup of enc pw file to auth gpg file in svn +They also say there that user code should not rely on bootclasspathI guess it is feasible to add usrlibjni to the list of system jni paths and it should work for pretty much everything installed by Debian but thats not a general solution +The MD and SHA hashes are OK however the format of the files may perhaps be causing a problemFor example contains B BE B E C F E ED CD F CFA more usual format is eitherBBEBECFEEDCDFCForBBEBECFEEDCDFCF Unfortunately the error message from Artifactory does not say what hash it detected so its difficult to know if this is the caseIf you are able to test against Artifactory a local repo perhaps you could try reformatting the md to see if that fixes the problem? +Thanks Stack for review +Can you give more detail +Regarding and I am investigating the best way to marshalunmarshal lists containing abstract types using JSON in Andrei I would also prefer a solution without additional dependencies in our code +OK I opened +This is fairly minor especially because tr should be on the path for just about everyone and tr has not really changed in a long timeI am fine with checking it in as is but it would probably be best to just add it in +in the core +Can someone please have a look at this and commit itThanks +I incorporated changes that I myself had been using and I had a feeling it might be a bit much to process on a short timeline +the patch so that HQA can pick it up +What browser are you using +Thanks. +If either job or cron are missing for a job entry an exception will be thrown. + +Hi Lukascan you share your findings? In my case it seemed to be a dictionary problem but Im curious to hear what you experienced +plugins also give gather reports about this generate a report that can be shownintegratedaggregated by hudson like possible for emma today +Documented as a known issue with workaround in the EAP release notes +The code can still work with older versions of the jUDDI client as well +Its really not clear to me what the best approach is here although embedded JSON would require fewer moving parts in the clientIf this approach is to fly somebody will need to document these opaque configuration structures which will mean that these structures must remain backwards compatible as they evolveThoughts +Are there use cases that would help us think this through better +Hudson wont run this patch +If you want to change that you can modify value in GERONIMOHOMEvarconfig file +This patch here appears to do that and allow for an override which would make it possible to have unit tests that simulate failures +What repository and version number do I need to use? I tried and SNAPSHOT but neither resolved +ok if it is only to implement Serializeable i will fix it +Checking the javadoc report +So there really is a fundamental difference between the way the Jackson introspector works and the way the Jaxb introspector should work +dont want to do the special gyrations I have to do now to create my own release with the release plugin + +In Beta therell be Restart which was done in Alhpa and is imho the most crucial is done +Unfortunately configurations defined in secondary config stores cannot be loaded so they can not be sorted +declare all bindings at or near root node independent of whether that particular nodes has any references +Commit it Id say +Complete Packing process and go the shipment Rout Segment Send UPS Shipment Confirm Request Accpet UPS Shipment Requestupscie is not returning this at this moment +It is possible to work around this issue by using ampamp in the third times a charm to work around use this ampampamp +You can find the export tool in the branches +Thanks for your patch +trunk Now uses mandatory body +Committed in r +There is still a problem with trying to add AJP connectors +If you got a google account then I can invite you in on the team so you can commitI think we need help to ensure the project is maintained and kept working with the Camel trunkAnd then to do releases maybe in sync when we do Apache releasesSo when we do a Camel it would be nice to do that for as wellI think the has some stuff in there for doing the release and uploading to a maven repo at +You can configure the build output directory in Configuration screen under that case can the default location be within the webapp directory +Release notes checked on stage theyre fine +Attached the escript test that does a PUT of document with a multipart +Any luck recreating this? I have to keep this change local until its committed +Fix this for for branch +If I can give you one advice on this one Id say this fuck Derby use H instead as the default embedded database +I wasnt having much success reproducing on my machine by just running the test but was able to forcea path through the code in the debuggger that got a null at the point in your stack trace +Final patch also tested against mysql and postgresql. + +I hate it that I feel driven to make the following statementConsider Spindle for T on hold indefinitey until something moves on this issue +This should also work out of the box and ensure the obviously expected behavior +New patch to be posted patch after it has been split between this and attributes of a macro have to be initialized at invocation +The issue is caused by some dependency issue with FMC +Apr render WARNING Bookmark with target has a nulland the resulting PDFs table of contents has only B listed +defer all issues to +For my part you may close the issue because we will gain no new insights from keeping it open Best regards Wolfgang +Due to side effects found Ive revised this to use the fully qualified class name for Imported configuration classes instead of suffixes +Renamed as Brock suggested moved files as mark suggested +The pressing problem is that the current implementation predates dynamically created keyspacescolumn families and as a result its not possible to addremove keyspaces with auth enabled +So any job that has a greater than MB will fail on submissionThis patch is for branch with tests +Kishan could you help to take a look at it +Attached the patch after fixing the review comments +This has been sent to Jason Greene and incorporated. +How so +If found ok I will fill up the testcases and only then commit +Remy JFC have you worked on this +Attached an initial filter file with exclusions for some false positives found in the package + +Please wait for a build before verifying +Verified on +Not sure this is really a bug +I am not sure if I am fixing this correctly it sounds like there should be a default for HADOOPSNNHOST +Looks good +builds are available that support debugging now. +Reopening due to a misconfigured local repo +Updated Poseidon model for name space proposal +Moving to the XML schema Structures component. +Also create a public that is needed for of we will make the current class public +That makes sense to me the start time should be the end time not +Jason can you pls look at the javadoc warnings +The native versions use the library to do the binding which in turn depends on jffi to provide the actual libffi binding through JNI +If you can provide trunk patch and patches Ill commit them +The C backend issue is basically +fixed in revision +Attached the exception logic to check for selected objects then disable EditClear actions +So until the line is uncommented the user would not be able to access the console which I guess is as planned to make sure the user explicitly sets up valid usernamepasswords +It also includes a tool that can be used to determine if a password will of this defect +Thanks for your information! Gradle plugin for eclipse can work currently +Add import support to skinnings css filesTo test if these kinds of issues are resolved correctly make sure you save the generated css file before your change and compare it with the generated css file after your change and it should be identical +looks good +Retested and verified this is fixed in current build +Why not make it return an empty list instead? Returning null instead of empty collections is just begging for bugs like this +works! How on earth did you figure out the work aroundI actually changed to order in the project my junits depended on not the project which contained my tests! I didnt try it the other way thats weird +The values were in seconds this patch changes them to +Closing +My thinking is that while we assume protobufs will last us forever it will provide a path if we ever change our mind down the road +Eric mentioned to use HDFS abstraction for S +think could be postponed +Check rendered component property before create javaScript string +Ive just greyed everything when mylar is on and it looks ok +Directory listing is indeed very CPU intensive at NN +bulk defer of features from to +I believe so +Is it possible to do the check with only +For each word which also exists in the training data we assign it with the corresponding Pos tag that appeared most frequently +for the patch as wellThanks Stephen +Im already working on this +Lets continue this discussion in +Just wishful thinking +Any update on this +bulk close of resolved issues as part of JIRA housekeeping for release. +Changing title to better reflect the user request for better documentation for BEFORE triggers +Lets see what I can do +My suggestion would be to go into trunk and and once all the umbrella changes are in we can consider and +There are no policy documents for the logo +I guess both should be addressed for +If you find further issues please create a new jira and link it to this one. +I checked +Please closeFor the dont open a JIRA according to the trace it is normal error the request is sent but no response and the request cant be replayed there should be only a few in the tests. +This definitely worked in my sandbox when I tested it I will work on this ASAP +Jonathan Can that be done for as part of +Coming right up +This bug has been marked as a duplicate of +surely one can just create another to talk to it +your for a great productdavidvps You might want to correct the typos in the issue title +Committed as a part of +One good example is +I added flag to disable tests that is not supposed to be run from maven +A patch is attached +Michal were waiting for upgrade to fix this +Yes please +I ran into this when upgrading to +zing +Thanks Todd for your commentsWe have tested in our cluster in long run scenarios alsoBut i will check the defect also +Verified in CR +This heap exhaustion test resutls in nearly times of GCAfter all I think this is not a perf problem any more given the same Xmx of DRLVM and RI +So Ive just committed it and resolved it as Fixed. +Todd fixed this and a retest confirms. +The original test had a bug that got exposed by the changes I made +Requires the ability to distinguish if a value is a Null String in aAdded required method to interface class and used the method to return the correct exception inUnit test amended to test for this applied svn revision +Heres a patch for this +Thanks ChristopheIs there a chance to see OCM released as bundle? +However I will continue to test different scenarios to see if this happens in other cases +The change looks good to me +So you mean only conf should have the codebase setI know the codebase specified by the other is redundant I thought it was there sort of a dependency reference but I can remove it +If there are no objections Ill commit this soon and treat the javadoc issues as part of +Thanks +Emmanual put in the solution so that at the very least the were all acting the we need to do is allow for control over the query that selects the columns from the database +However if its cool to include BSD licensed build scripts Ill probably refactor this in a different way +Ill roll another patch with that goodFinal nit dont do something boolean options are not as easily understood as do something would rather call this durablewrites or similar. +When I get a sec Ill look at using the console +Ill think on this a bit +Whats the purpose of this feature? Seems strange to fail a test based on a message rather than type +Thanks Prashant for roping me into the discussionYes Ambrose relies on the unreleased Pig or the patches that Andrew referencesAmbrose has only been open sourced for about a month has been well received +implemented and tested +New patch merged with +thanks +We dont have an LDAP Importer but maybe add a subsection in section that includes some guidance about how to model LDAP sources. +I committed a fast and simple solution +Will be backported along with all other issues that have a +AM can decide whether to reexecuterelaunch that task too many times or not based on the type of failure and number of times that task failed +Uploaded the patch with only changes of Common +Would you be able to test your code against a downlevel JDK to see if youre seeing the same problems +So I didnt commit this part of change +Could you be more specific? Its not clear what you are referring to +Attaching patch to reenable tests +It seems strange we would catch an EOFE on deserialization when we have the version available to us +Ive added a section on doctypes to the template documentation. +Actually further tests are fixed by this patch and +StanWould you mind taking a look at this oneThanksll try to get to this sometime today +Thanks xieguiming! Ive committed this to trunk and. +The images are included separately and should be placed in hwisrcweb the is not modified by me +This is because the callback object wont have been callIed the chains wont have been created and hence the endpoint references wont have been resolved +I think this is a valid concern +the patch looks goodI put this into and trunk +This issue is occasionally reported by Jenkins logs for a failed test +For completeness the log when running under Eclipse is attached as for points +committed in would love for this to be to because it breaks things in the line too. +no problem found in the search problem found +Changing the assignee back to Bryan +Ive applied and committed the rest of the patch +Thanks Supun I have applied the documentation patches as wellRuwan +This file is modified by the committer when the patch is committed +Can you look at this issue further and determine if your changes affect these remaining two tests +Merged trunk revision to branch and committedto subversion as revision . +ran clean +Alter Hibernate dialect in file +Its just a warning application should work +It passes on your suse box now Nick? I triggered hadoopqa build +Attaching patch soon +If the way to get it is to have the hadoop permission scheme applied it is fine for me +You can close the issue +attaching see the repprincipalName is added by the but I cant tell if this is already included in the repo init or not +Bundles prepared with jar executable next time scmurl i think should be the http url the other is scmconnection +I made the examples extend which made them far less verbose +Ive done a yourkit profiling against it all the time is spent in ZK in my test scenario +we should get this in +Code is checked in having trouble deploying a snapshot on the new nexus repo +Backported to branch as svn +Making provided in all the poms and adding missing dependencies in some poms should work in both apply trailing whitespacewarning line adds whitespace errors +And the dont know can means not coded or I really dont knowI see +I will be careful next time firing a bug +Thus the class as it now is implemented has the best result for positive cases and provides a good for the negative case +I think the bundle info table should be pure html table instead of clickable dojo table widget +I am not certain but I think when ooking at the puppet service i saw a couple of other things happen so I thought the profile was already positioned that way as it stood but I could be wrong Ive been digging through a lot of these recipes and services the last few days +Remains to be fixed for fixed in tck +Not sure why it was allowed to run as root +This is not the case +Yeah looks good +commit is not present in +Please note that the new patch is SunithaThanx for the reviews and comments +Thanks Chris! +Any other idea +Attached is a first patch adding the fnnormalize functionSo far Ive only been working on the part but I may provide a patch for SQL as yes b well the ticket description says youre looking for collation support Im not convinced that adding something that actually doesnt collate properly is a good idea +Does not happen in all cases +Changing patch Ean SchuesslerThank you for your suggestion +The name should be configured on the appender +The previous solution allready used the Muncheun techique just a slightly different way of doing it +Sorry for any confusion! Im waiting for my admin to send me the file +Tests should be operational for Alpha +Pls give if finds useful so that I can prepare a patch +I dont follow +Once Eric gets his JIRA account issue resolved hopefully he can provide exact steps to reproduce the error +Same I am wondering if the comparator is the problem here +Removed the two class files from the SVN repo +I suspect that the logic in Methodtoproc is to blame here +Can you post how you have setup this +Jochen this contains modules inside this +Could a comiter take a look at this changes +This is trueeven if you are using Option D for caching which expliciltypromises to resynch the cache with the persistent data storeperiodically allowing for the type of manual sql changesdescribed D promises to resynch the cache with the data storeat an interval specified by the +Can you try building and installing the in your environment that I have mentioned? That one causes the problem and you can debug into it and see whats going a reality check on my end we dont need bundle activatedactivate methods in any of the components of an unless we are using type registration rightIm going to go for broke and update the other jars that the components in the stack have in common +The Cant find resource problem is solved by the patch for a similar bug +Perhaps we dont need a splitter at all withLet me know if there is any comment about these steps +This bug has been around since the introduction of Flex Mobile Project +Thanks Dennis +Actually closing this as this is a duplicate of. +Assume closed as resolved and released +Patch committed in revision +bulk close of all resolved issues. +in my case to have the javadoc jar for jdk named properly we need to set to but shouldnt the javadoc part already part of the file? Similarlysources should be a builtin for sources jar filenames should be a builtin for test sources jar filenamesjavadoc should be a builtin for javadoc jar filenames should be a builtin for test javadoc jar filenamestest should be a builtin for binary test jar filenamesthis way we just need to specify one and only one classifier jdk to have this classifier attached to all the names of the generated files. +Looks like we do not need to update the output for deleteSnapshotrenameSnapshot when the number of arguments is wrong since the USAGE msg is finally output in the end +r on trunk +New patch to remove null check for other required fields +This was broken due to the switch to using bundler for the ruby gem +wicketsrctestjavaorgapachewicketresource +Done +Matthias please see my last comment over at to get an idea about progress being made and ideas about testing etc + is the better of these t do it properly +Do you know where those tests are +It would be interesting to see some use cases for in aggregators. +I think its ok to go with the current patch for line and update the patch to match the zxids for and trunk +Attached is an updated reproduction with the Network Server reproI changed several things from the version but I think the key points were +Let me know if you find some ways of doing this +This is really an RF kind of changes are we talking about here? We need to understand the impact of changes to make sure there is not too much risk. +Ok but this is NOT a good idea for a production profile since it will reduce our performance in out of the box benchmarks +is in Hadoop so this patch should allow Mahout to work against that version of Hadoop or a later one. +Sorry the IDE did that and then hid the header +Thanks for your input and I are discussing it +Made this a feature request +Hi guys in fact this problem was incorrectly classified as a bug in JVM whereas this is a bug in which primary affects Java see +This is just the first draft and I will be expanding more on it next week +When we execute Checkstyle directly it worksWhen is expected this issue to be resolvedThanks +No problem +The itemPrice parameter will have the price the user entered which might be different from the list price or any other standard price +updated test with additional output and alternative sca binding problem was that the callable reference was serializing the wrong conversationId to the output XML when being passed over a web service binding +Reopen and resolve in correct version +Hi Jessica I just tried this on with the latest from branch works OK here +Closing out old issue. +I think allow would be more descriptiveAgree +According to my ongoing refactoring in that area I think it will be better if I take a look at this myself before I need to search how your changes fit into my said that I would appreciate if you could add a test to XML test suite dealing with enums +Committed the refacting patch into revision +isThe paging should be shown asPage of +workclusters o +Costs for moving a region vs data locality etc is not trivial and I think that it is not static as well +I updated review board as well + +But this is maybe stuff for the future +Shoule we eliminate redeploy altogether as deploy will anyway redeploy whereever necessaryWe can if we want to it but I think this need to be decided by community +HiAre there any plans to fix the DSML exportYes I plan to fix it forRegards. +This wasnt a high priority issue so I didnt want to pull you away from more important stuffIm not too worried about using an external build farm as doing so doesnt take anything away from us +Odd because the safe leader test that fails is hard coded to not use ramdir I think +This looks good to me +I have not run any other tests with the patch but I noticed that a number of both positive and negative test statements for the describe command were added to the tools test script +The bug underlying this issue was that eclipse updates could break existing Teiid preview connections +sad excuse +Good eyes! Another bug fromHowever +Ill commit tomorrow unless there are objections thenRegards in r +substantially done for +Similarly persisting a PC instance that is known to be then only uses are permitted +Closing resolved issues for already released versions +Here is a first cut at forward port +Sorry revision for my initial commit was on still have to port to trunk along with which hits some of the same to trunk too +I guess its a similar issue were having thereIts not simple but given the above points I think that even if does of the work for the remaining of the work we need a deep cooperation between HDFS +bulk close of all resolved issues prior to release. +I dont know if it qualifies as a blocker or not I just know that this test cluster is unable to deploy vms after what shouldnt be too uncommon of a scenario +html and +if Rahul has time to monitor and apply patches from him that might be better but if Pikus just wants to have a go at this at google he can always propose to donate it back once its release ready or continue it there +The next best place is in get but it needs to be called on each new connection rather than when the Proxy is created +The code should allow record number begin with disable test for record number throw an exception +Thanks for input +Closing resolved issues for already released versions +The guide is not created yet for V so this is on hold for now +Havent been able to reproduce this error. +Also I hate properties pretending you arent exposing state when you are +With the current event mechanism this use case cannot be achieved because a reload only happens if a configuration is accessed. +updated and to implement Skippable plus added testcases for skip +mostly fixed now though we do need an improvement to increase the palette size if a route wont fit on the available browser screen as a big route can overlap look crunched up a we resize the palette to ensure we dont overlap routes though we dont always show the scroll bars yet. +this is one low hanging revision with unlock use max moved to moves the xml parsing toThis introduces a class which can be used for all pluggable classes +Solr uses a single Filter +Patch rebased after +This style probably works better with +Source for component along with I do not have commit rights so I will write it up and attach it to this issue +Doing a bunch of other column namings because looks like new filters can just drop Row from name and then theyll be easy to recognize +I dont see a patch +Attached patch reuses the config originally defined under and gets rid of the wrong referencePlease assign this to me + +it doesnt have to be single threaded to drop messages it just has to not run the verbhandler until RPCTIMEOUT after the message was receivedIn practice though I dont see any place where the verbhandler would be blocking or otherwise slow including the ones running on MISC +Although it looks like a boilerplate template we have to take this serious +Findbugs however complains about not being presentMy only other thought is that it might be nice to update the packaging to also include the datajoin jar +xthanks chris +Your code appears to run cleanly with current CVSIt is possible that my fix to URL resolution for fill etcfixed this bug as wellPlease test and let me know if you agree. +It only makes sense for me to take this one +Protected would be much better in my opinion +Please close this issue if it work for you. +Okay while creating the test case I realized it was PEBKAC pilot error +collecting bug reports like these are no good especially since groovydoc is handled in the groovy project but eclipse is handled in GRECLIPSE +as we can know is early than P +I did manually test to verify an exception is thrown when the layout versions are not the same +This patch deprecates the setHeadervalue attribute DSL method and changes a bunch of usages so any snippets on the wiki will be updated +patch appears to have disappeared +The payloads will be specific to serialization engines and their serialization engine will be named in headers +Added an include property to and which defaults to true +Im sure its a path problem but apparently slightly different or related to my specific setupI hate Windows +use an interception like feature on the portlet container which intercept all request made to a portlet and react on error by providing an appropriate error page +This is +Linking which abstracts the distance filtering process and index formatsCurrently the nd step in Spatials filtering algorithm calculates the distance between the central point and each document that has passed the tier filtering stage +If there is a patchout there that hasnt been severely affected recently its smallenough not to matter and Ive swalled so many merges on the flexbranch that I dont much care about swallowing this patch too +The help files dont contain code +This bug has gone stale +Ill draft a PR to fix this +No tests since this only comes into play in situations with a very specific set of environment variables +Even for the users in the easy category this is a cost incurred per instance of Accumulo which is generally higher than the cost of including it in the primary build +I especially like the first idea which wouldnt require any declaration at all of a dependency on the sources ll look into it +Modified component and event classes for syncing with. +Remove widthheight values in your meta tag the template app be updated +JasonJaikiran I can see that this issue has Fix Version +Issue is resolved +It would be helpful to know what does +this is why signing jars is not needed when building jnlp packageI think it is use case that is worth implementing since it is probably very simple to achieve +In that sense makes sense as a fixed value +studioeclipsefeatures +unit tests for issue and +afaik I dont think we will have this material in timeframe +Depends on +Thanks for spotting and fixing this Galder. +Server A contains HQ backup for server B +Say MB +Wow! That was fast +Generates the pom and swc files based on the current java version +created launch configuration prematurely +enforce retryLimit in the was actually wrong +I think thats really bad because it means you stop relying on the IDE to tell you whether a project is broken or not +Fix a unit test failure of +I would love to see disk IO usage and page churn data +Thanks! +Just a thought Great work +So we should fix that separately? The reason why I say this is once we do that we will no longer need ivy configuration of these common jars in each of the contrib projects +Ah check that the files too big too attach +This needs to be fixed first which will involve slight changes to the tests in trunk which will them be ported to +But shouldnt there be ways to edit view the settings in it or are the hidden in some of the other metaphors that are in the current console? I would think the values of the need to be exposed somehow or am I just missing something +Its long time no updatedDo we still need this feature nowI can try to add it +So apart from having to change a lot of Ids when installing the files in my new repository I have to check all of the s for now invalid dependencies which is a bit my main question is still unanswered Why is the validation so restrictive in the first place? In there was a request that no should be allowed in an Id which is ok but the validation that was created goes far beyond that and I dont really see why this was necessary +There is no summary or a design doc that explains whats been done or what exactly the vision here is +It could use a few formatting and style tweaks and should be linked to from some introductory topics +Fix looks good. +Please remember to add documentation when you add new functionality options +We are seeing the same problem +This was fixed in all resolved issues for previous releases +Just fixed the NPE for Identity transform +I do think we should move to gradle for building which might make this easier. +attached +Please find attached test caseTo run make sure you pass aop path and javagent parameterssrcmainresources javaagenthomegalder +Ill see if I can port these to reproduce in AS as well +Therefore I dontt want to change back to it +Latest patch +GeoffThis is held back on my end +The above is fine please commit +Model project set and two message construction when has error and cant complete +Whoa +Looks like IT one of the test case created under user testtest was created in the backend database with an internal status as Waiting on Support actually +We will move the JIRA past to revisit in the future if a need arises. +I would go with for the gate as well its symmetric and provides more options +Attaching the latest SQL authorization versions of the examples and +webOS updated +Lets aim for with this +could we figure this from the start and end key somehow it might be a little jumpy on the done as it covers any range of keys where there is no rows but it would give some output +Working on this +You should use the DSL from and not really the api in eg as shown above as there is a nd parameter now for the resultType. +done on the +but that looked very overloaded and so resorted to thisPlease take a look and let me know your opinionsThanks Venkat +Many times the users that are affected by this dont know the actual cause. +Set the resolved as the remaining issues concerning shading patterns are handled inThanks for the contribution? +seamgen problem in Hudson job fixed by Seam teamphotoalbum and demo on tomcat verified manuallytodoregression tests were run manually by Lukas. +The order of cleanup steps should be Bring down the engine +LGTM +Increased server side action timeout to minutes +I committed this +Thanks for the tips they are very helpful +Just note that is going to ship with jQuery. +This was committed with. +Committed revision +what is does context mean here +While it appears that current trunk Mahout LDA correctly converges on this toy problem Im reopening this to track the need for this unit test to verify that this is the case +Oh Ive made a mistake +Master should not start because otherwise we end up with data lossIf cleaners are not present in the conf and therere no snapshot we can still use hbase without the snapshot feature +Thanks Knut. +I am running tests nowI made one change I made the a token rather than a quoted string +Should be marked as resolved +committing now so that you can recreate the patch for +Lets wait till we run into a stuck issue that is other than because of modd code. +ok thats more like failure with those settings is actually a failure to create repository +The cause of this bug is because of the path problem +Yes which branch should it be committed to +There are tons of places to host projects these days and it may make more sense to be hosted as a separate project +If the extends attribute is declared and no templateType is declared simply set templateType to null +At the end of the next week I should have it done. +The patch is to do with invalid configuration of publishing the pom file +Updated patch with requested changes +message +It was also included in the release +upgrading to blocker to make sure this is resolved before we release. +OK for starters then what version of ffi do I need to compile? It looked like the current version is but the stuff included with is +The code should skip this token this before before checking the supported options +Good catchHoward +The path which is being tested is generated by the library and it uses a config file if it exists in that place +Unfortunately there are still clients relying on that attribute being accessible everywhere +When do you think this will be added to Spring Integration +So I just added the port that makes it consistent with the other IPC threads. +Affects and good though do you have a patch +I think the correct term would be public prime +Ralph can this issue be resolved now +jar file in my folder +serveralllib +The project was probably messed up by my previous you can close this bug +Otherwise I withdraw my objection. + Username hanming +This patch should do the trick +the virtual machine metrics look different now and dont have this problem anymore +It appears only if I touch mouse and move it a little bit away +It seems the size should be bigger than by r +Cycle test is now anyway. +It affects both binaccumulo and the Main class +I am closing this issue since Griffon support is not planned on being a part of on making STS open source is progressing but it is still in the hands of our lawyers +I think that it is more important for Thrift to conform to the standard idioms of the individual languages than to be identical across languages +Thanks also to Shanyu and Bikas for additional code reviews. +Thanks Mayank +The solution seems to be a bit complicated +it is working on with that mean you think you fixed it or that it happens to work with? If it did indeed work with what was the SQL that it executed +work c +I hope Ive fixed s also missing a test to prove that it works +Just FYI I was waiting the customary hours for feedback on taking the approach before enabling it and removing tests from the old harness +The relativePath suggested does not work here +Sure I will wait for your review +The fix is very simple just compare total size to thrift message limit. +part of the log files +Are you saying your SERVERs default locale is DEIf you make that change and recompile the source does the problem go awayTed. +reopen due to it mojo will now check if the directory contains xsd not it will failed by can be skip by setting to false the new mojo parameter fail. +Fixed in latest master please do a pull test and confirm this is fixed +Thoughts +claus I tried your fixes but my route is still unusable after is restarted +Is there currently any work in progressAte Could you provide more detailed information what the problem was in your last fix? Thanks in advance +I think they made it in there from a copypaste +Jira issues and patches are pretty pointless if reviews and comments are going to be missed your message scott and we wil look into itRegards ScottYes but what if someone prefer null tomorrow etc? Should we not avoid redundancy which we now may lead to inconsistencyAnd Im sure one day someone whill ask the differnce between nullField and emptyId prefer to revert +Applied patch to CVS +why +Heres the verbose output from a failure w the patch +defer to defer of issues to +Im using the to divide a GB index into roughly equals parts and deploying into production +Resolving again. +Now waiting for release +If indeed objects are being moved during enumeration it would be good to provide a costbenefit analysis +Ive attach a patch that special cases the in +Heres the patch as great +To create a meta class we need full reflective information +patch applied and checked works great +Reassigning to compiler +has build in handling for this situation +Exactly when will the method names be consulted? On each call or only when a optimizedFoo is being called instead of the regular fooIf we assume that not all new methods are the optimized ones we have to do it for each call right +Here is my portlet with out directory lib +I observe the problem with both and +Updated patch with admin page support and numerous fixes and through this patch and I think its a good improvement on our old warpwhitelist implementation and as such should be committed +It is a limitation of Mavens project object model +I upgraded I keep getting caught by new plugins being new installations rather than upgrades so I missed this the behaviour is the same +Talking with Brandon in IRC he suggested I stop cassandra move the index files to another directory and restart those nodes +Every client thread must have its own port instance +So even if you have k servers running maps each with threads each that is still only k pending calls which wont be enough to blow out memory +Reassigning all open bugs to the development listSorry for the mass mailing +Hadrian bundle and bundle have the classes of andSo we dont need to ship these two artifacts in the distribute kit indeedAs there are some user request to use the camel context without spring or blueprint as traced I just added made the as the osgi bundle and camel kit will ship this bundle +If enabled this add some Cache Control Response HeaderAdded a new Transaction Token patch +Hi Sameer I would like to complete the append work as described in the design paper associated with this JIRAAwesome! Thanks Dhruba +Even if this does not cover all the cases it is still usefulbq +Thanks Knut AndersI added a link to that issue +Punting to hoping for community help +I am working on this and some more testing and analysis is going on +how can i get the latest cvs +Please dont stop +This is roughly what I had in mind +I know +Validator is required by Seam to do all the validation logic including JSF at I dont know is if HV is compatible with whatever Seam version is present in EAP +The highest ordinal defines the configuration + +Thanks a lot for taking time to review the patch Army +If possible i want to improve and complete mobile template things this summer and Roller can have available useful mobile device template in change +You said release nightly builds but I guess you meant make nightly builds available for developers + +I am going to hold off on this until after goes out +I have a watch on this now and will reply right away in the future +rightbut this is a better solution because across imports this will give better revision Thanks RyuuichiI changed the patch to remove put since we cannot save the object creation with or without it. +The lease database locker seems to work as expected now + +update for property and exclude lists +IMO its better to wait for JSF which will include the ability to handle any exception that propagates through JSF. +Fixed in r +bulk defer of unresolved bugs to +Yes +If it still fails Ill need a more complete test case attachedMarking resolved as of as I confirmed it worked with that version and a request for more information was not returned. +BoilerplaterTypeFloat represents floating point types of various widthsCurrently only two are supported float as r. +Yeah Christian please add that +Can you attach a stack trace? Go to Window Show view Other Error Log +Hello TomazI changed my script and now also the new way is working a lot is working fine. +Patch applied to trunk +Without this we were silently ignoring the failure to stop the proxy +The logic for this has changed quite a bit since and the given patch is no longer sufficient to ensure that types which normally coerce to Comparables will pass correctly +Use the already exists unit test through the s +Thanks Sijie marking it for. +cleaning up Open Issues list +on the suite of tests to be included +We only have to wait until gets out and makes it to our TP +Rails any otherbut with option debug all works fine +But we also want to know if it is practicable to provide a allocation mode +UML expert what next +Hey AaronThanks for doing the investigation to fill in the tableThe behavior as youve described it all seems pretty reasonable I wish the maybe were false but it sounds like it does require a pretty significant overhaul of failure tracking throughout the TTetc so Id be inclined to say its out of scopeSo from me on the current patch +If the problem doesnt show up thats the issue + javadoc fixed a spurious whitespace change +More interesting is how to get the configuration into ZK in the first place enhancements to? Or? OtherI think this is one of the most important pieces must still be easy to edit I dont think we should commit anything that leaves this piece outThe rest of this seems totally straightforward the biggest downside is going to be the increased pain in working with and so we need to minimize that as much as possible +Thanks. +also the convention for section id naming in the Configuration section is that the ids all start with +please reopen when you have more information and if this request still makes sense to you +Hi Petar this missed the cutoff Im afraid + has been marked as a duplicate of this bug +Just consider if people started regularly using + +This might be a good case to reintroduce the +Refactoring to make the match method replaceable looked like more than I had time to take on at the moment +with JBT. +Need to add to our tools kicking region out of regions before going on to mark it enabled etc +Actually thename of the configuration file was different than the name of thesolution and for this the configuration file was not foundPlease delete this bug because is not correctThanks for answering so fastRegardsCatalin +Please be sure to have checked Synchronize user status onto your synchronization taskCan you verifyAll of the checkboxes are selected including Synchronize user status and full reconciliationColmThanks for the explanationColm +Assign this to me if you want me to take it over +Hi ThomasI agree in general with your observation but I do not understand your statement because the specified element is not only added an other element is possibly removed during the invocationLooking at the add method I fail to see how this may happen +As Alex noted also mentions this problem I think it makes sense to fix it in one go +I have attached a patch that provides the necessary translations +I could not find a solution that could get the resolution of NULL vs IDENTIFIER done in itself +This bug has been marked as a duplicate of +We can change the needsArch param to false for but we still need to build and provide the native libraries for RPM users. +Im inclined to not add the procedural access pattern as a separate concern +If you are utlizing the default classloader behavior I would expect to see problems like this +Smoke tests on windowsxp and Linux w gcc passes. +Can you try again +I also manually tested on Windows by running and killing several jobs +It also mentioned potential problems with releasing of latches and those problems have now been logged as and +Please disregard the prior one +Willem ya I realize why I couldnt assign it I just wanted to let people know I was going to fix it +Committed revision +If there are for the release candidate I recommend grabbing those +I havent looked but in re I ran into some javacores being generatedKnut said Perhaps you see the javacore files because of? Since the interrupts cause session severity errors I think that code will kick in and generate thread dumps +Thx. +Fixed now by requiring the TLD to be a alphabetic character and or morecharacters +Committed to trunk. +is for those following along +Want a separate jira in seam for itShould I try and set this flag or let it be a known limitation for Seam and Seam will fix it or +Ah I see there is an example in the nabble thread +closing as no plans to fix in. +Thanks Patrick! +Did you mean to include the changes in +Investigate and resolve for +Attach followup patch v +the description doesntlook like the one ive found in the EAI is no content filter or content enricher the messageStore in place i can build a claim check by a with a new transformer and subclass that overrides get i understand the book correct the the claim check is also about information hiding and the message may be transformed multiple would not work with a payload scenario is intended? +Split this megamethod up into about or so different cases changed around testStartup to be several cases on its own that exercise directory locking more explicitly many of the tests now run with datanodes instead of so they should also run faster +submit patch +the new section has been task can be closed because all necessary info has been added +We wanted to decorate a number of standard classes to add our own logic to the normal processing +Ill check the other ones there are only about unit tests and associated code +One thing we could for the the fault injection is to move all of it to a module by doing that we wont have to publish the instrumented artifactsbq +Committed to trunk thanks Ahmed! +We are still awaiting completion of the core Perspective +Am I correct +That test was changed to compare to JDBC version values. +In the difference is not so dramatic as in but still existsdefault K rows per sec rows per secIt seems that there is performance regression in scan filters in +The patch is not applying over the latest patch at which is a blocker for this JIRA issue +In this case safemode is entering to extention after reading the complete edits during failover +Closed + +Reopen to update Fix Version +Lots of desktop applications use this concept and it seems to work wellSo the idea would be that people who are in simple editing mode would have the interface trimmed down to the bare essentials so that its as easy to use as possible while advanced mode offers all features for users who know what they are doingThis approach may or may not replace the need to make just the summary field collapsable but in general there will always be a problem with cluttering the interface with advanced features that someone doesnt really want or need and we cant provide indivdualized ways of addingremoving every field on the editing page on its own +Attached is the patch +Better patch that also sorts design docs in alphabetical order and reduces number of HTTP requests +for what it is worth I cant reproduce this on OSX +blocker? +suitesAll show no new problems +Issue marked as resolved. +This issue is important to me has it been reviewI think the solution of create another dialect a unicode SQL Server dialect is pretty easy unless Im missing something here in the bigger pictureThank youIdo +The final patch has not been reviewed or committed patch resolves from the mvn repo through ivy and avro from the local lib dirWhile submitting this patch to svn we have to add avro jar to the lib dirtnx +doesnt get aggregate config +More over some files should be excluded by default from demos +sign means use the JVM policy as well as the policy that is passed as part of the sign + + +However there is still a problem with the general url rewriting rules +In any case I stand by my original conclusion that meaningful interfaces and intention revealing methods are better than a generic and weaker listener contract. +Ive committed this +New patch stat output and derbyallreport based on comments from David Van Couvering +Now has FINISHED and FINISHING +Suggestions +It makes a degree of sense that is the case creates a new table so presumably it creates new indexes stats at the same time +The feature is developed at the revision +In addition polish the code added the tests of empty params and invalid state and added the document of the new rest API +originally the DIR property was taken by the Dictionary properties and the POLL and DEBUG from the contextWhats the rationale behind this choiceWhats the correct way +I do not know how the synchronization between geotools and geoserver take place +The first sentence in description wrongly implies that both target directories are created at the same time but only the one in the directory where is run is created when command is used +Hi MikeI am using VS express edition at home and VS unit test is not available +But I have seen that unless this status is set a job that is expected to fail does not fail +Resolved in scope of +pushed +Are overrides valid for hostnamesports of services would the cluster config be the better source of truth? For other things such as say configuring the no +Part of a bulk update all resolutions changed to done please review history to original resolution type +fixed for sqlserver and openbase +lgtm +can be considered as dead today. +If I remove the usetrue attribute the example icon is displayed correctly else the icon is not appear so i cant use my plugin +There is a balance between verbose name and completely cryptic name +In same conversation with support and PM today was recommended to list in some easily readable form on this page or one linked from it +In both screens User Name is from to Last Name First Name is from to characters +Im currently porting an application from AS to AS which uses Byteman for testing +Patch fixing the closing span. +Ive a couple of questionscomments This fix also affects the NLST command +Ok v ODIRECTORY changed to ORDONLY and checks for linux are removed fromODIRECTORY just causes open to fail if given path is not a directory so it seems we can live without it +There were a couple of issuesaprsignal is not supported on Windows which was used in +Lev verified that the issue is now fixed. +So there i think we should simply deprecate just so we dont have to reimplement hairy back compat +GWT not compiling in the default small gear sized application but to matter of tools they allow to create bigger sized applciationsVerified in Developer StudioVersion Build id Build date +I hope to be able to rewrite the install configure scripts by the end of the week +So are you suggesting that the IO reactor should itself automatically if an IO exception is thrown? In my opinion the recovery mechanism should not be part of the reactor itself +I doubt theres anything that can be done about the problem from +This patch provides the implementation for ldap Compare Modify Modfiy Distinguished name Modify Abandon and Extended any committer take a look + +Rebuild! actually rebuilding the project or something +For one case it is the location of the filefiles from where the camel route came from and for others it is used to correlate several is also possible that there can be multiple camel contexts in the same file +For the first pass this will bring up a ensemble only +Patch patch +Can you provide some details please? What exactly doesnt work? Can you try to track down whether the is even invoked by the JSF providerFYI we had some JSF providers that couldnt work with in the early stages due to bugs in the JSF works just fine for me using for the time being since it seems to work in general +That means the main issues with this bug are with and failures each off another build seems the problem was in the regenerated certs that had expired +Alex please check these icons on conformity to current RF item icons from Alex +Patch with latest common changes to match +Please could you verify that this application of your test case enhancements is OK? Thanks for this enhancementBest regards fine thank you George +Attached patch that replaces String with Text in IO +Bulk close for Solr +Works fine under IE +Some last issues is missing in the core bundle theres still the uppercase S in Change severityOtherwise its working great +I see they are in CDH. +Proposed for the patch +not we are interested in the externalization features +Rebased after was committed +I mean four characters of null are a legal JSON null token whereas unquoted string of foo is not if were used it too would be a legal JSON String value +Duplicate of close of all resolved issues prior to release. +Increased maximum post size to MB +Canopy is subject to getting just the right settings for parameters and is slower running as well +IMHO we have two options simply unpacking the kar and add the found features files to feature index handle kar files separately from features and only store which features where packed in which kar fileI personally would prefer the first option +I tried to put the URL of the SCM in the child without any effet +Please let me know if there are any problems and I will fix as this feature it quite important for me so I am keen to get it included +It depends on the first issue +Maybe we should remove that optional tag from the Maven dependency? WDYT +It will be too easy to break it right now no fun for you +Closing it now. +I guess we can call this issue fixed +I think the bug can be Thanks for truncating my initial comment I didnt realize that JIRA would include it with every mail +This is part of AS +Sorry for the delay +Patch at rev +No I mean log which produced during project analysis +Generate the WSDL file with +Hi Eddie +trying something on my internal builds didnt run this test to report failure +The regressions ran without issues for +Thanks a lot +Hey all +No rudeness is intended there just arent any other choices +a proposition diff for the list method +I will hopefully post a patch soon +If we had to choose between those too I would take the ones on the there any rule regarding plugins locations? That would be good if all JEE stuff would be on one location wouldnt itLet me know and Ill finish this +One effect of TIB constant folding is to change a load from the object header into one from the JTOC +I enabled the debug log for the class I can see this entrySkipping provider because required class is not availableThis seems strange because I can see the package being imported for Java in etc whats up here +I will work on making this did not see this bug I have a patch attached to + lgtmThanks +The idea is to detect badinaccessible data directories and shutdown the datanode if this occurs rightThere is going to be a periodic checker for all the blocks +added in the test start testing it +Alex I tested modifyRN operation with but it looks like it is working OK +from me too +Sorry for breakage +This module appears to contain nothing more than a README file which hasnt changed since +Im out of town until April st Ill check when I get back +The wrapping is removed and the real cause should popup in the error message +Reviewed by Rob +Please read the comments about enabling reloading and that covers the actual requirement +and OOODEV it is +The job jar does not seem to do what it needs +I think this issue is stale +The Mappers now allow setting of a custom like the Transformers +Right +I will close the issue on STS but will if you can give me some more specific instructions on how to reproduce +Let me know if you guys see anything screwy come up again +Heres a patch that implements this +Pull it back in if you think different. +moved to stream because were were approaching feature freeze for stream +Still lacking tests that use JCR implementation but that will follow shortly +Also attached is an excel sheet with some numbers +Comments from MaxWondering if it is seamgen that is using the wrong reference method or if resource should be updated to not only look in classpath and if that fails lookup in servlet contextThis is a portability issue between AS and AS The problem is that in AS web container started behaving correctly by not having the root of the war on the classpathIts if resource is supposed to look in the web content classpath +marking this as fixedthanksdims +In order to fix this I think we should do the following In spinnerStop as part of the if where it checks for null then it should also check to see if the spinner isShowing +Of course such an approach might be more difficult in the future if we move to a more sophisticated resolver but technically I think it would still be possible +Isaac are you going to document this issue in Release notes +Updated patch to work with latest changes to +aaah this is caused by jrebel +Could you please explain in a few words what is this all about +There are minor differences about the amount of detail regarding licenses and authors but the submitted has a fontbox copyright +You would actually also want autowired dependencies to be considered too +If still relevant reopen or create new issues. +resolved after build started to use Eclipse binaries +looks to be an issue with using +Refer to the ESB Administration Guide and the Getting Started Guide for more information on its use +This is clear enough. +Redeployed the site +Good to know. +I dont see particularly good reasons to advise Bean methods but well for auditing or instance tracking or whatever people might be doing it +Lets verify that event handlers are excluded from this rule! +If SPTP is true we now wait for a cache to bestarted for a while even if it is not defined +Sorry the patch is indeed reversed I want to throw the exception +Unified patch incl +Just made issue going of same complaint a few times in a row up on mailing list +Cool! Do you mind attaching your patch here? Thanks +Currently the CSP does not gracefully fail back to the default language of English for IE users if the browser language of preference is not English Japanese or Traditional recreate Launch IE Set browser language preference to French Go to CSP front pageAs I understand it IE is not defaulting to English in the same way that Firefox +Attached you find a new patch +I just noticed that the text was not well formated +Also added tests to RM app transitions to make sure they get killed +But that too would appear to be dangerous since there are unknown side affects of not running the ECANot sure what the bestgood compromise is for the bulk vs normal operation is +if you think the patch is hte right way to go then from me +I would like to know any review comments on the approach in patchWill the check the test failures and Javac warnings +bq +No screen captures necessary at this time +server file in the workspace metadata or it can store it in the Servers project in the main workspace +Also this patch was still in trunk and not part of any release +Also adds unescaping of the value for row cache provider +The Python issue seem to be fixed +In a meeting we said it should be should also disallow validate before ABD +let the project tree is long and has a vertical scroll +This patch supports majorminor versions of a RPC protocol +these issues are obsolete since the version of the plugin +Weve got an idea were testing out here and will hopefully post a patch in a day or so +Good point Todd +I see +is failing since we changed how compactions are done +Where are you seeing these in the systemCan you provide the full path where javadocs for the above where found and do a rpm q whatprovides for them please +Incorporate Konstantins comments +Hey Mark H any chance you will have some time to look at this soonNow that the test classes have been merged any change to the current contrib test class will break this patchI think everything is good except that we might want to alert which Highlighter version caused the Junit test to fail +Developer sample creation is done +Are pictures with zero width even possible? In any case love the comment you added for the check +The workarounds being put in place are uglying up my code a lot more than Im happy with +So what is the solution here? We could process the provided style and guess the size and then render a point symbol? The other thing we could do is let the point symbolization run and then take the entire resulting graphic and draw it in a funny color +I just committed this +A is required to be able to set modification time on files +Added property can you please attend to the following items Fix typos +patch applied ver +Remove javadocs target from facetI will commit this shortly +But I prefer we just change the docs so that users relying on receiving only jobs by default wont see a change +I think this is a good idea to decouple image and edits +Maynot be the final one +Sent from my iPhone +The repo is up again and we can deploy. +Code test and documentation updates applied to trunk and branch +Thanks for looking into this +Ive uploaded a new patch with the following changes based on your feedback I added comments to the regex to help clarify +Im either missing something or my environment is messed upOn test cases the wiki instructs By default do not let tests write any temporary files to tmp +Its resolved as duplicate but the second issue referred is not resolved either +That is the stacktrace of the exception that causes the expression payload to fail +thanks guys +Database owners alone can create multiple schemas in that database or create a schema to be owned by another user +Applied branch and trunk +I forgot to mention that if I dont touch the metaclass of the Script class everything works fine +Default match mode should be ANYWHERE not EXACT +This might be useful for regions without many items +I have added the missing srctest directory and also removedLet me know if there is anything else that you need me to do to resolve thisticket +Is the intention of this JIRA that in Hadoop common we provide an API that takes a list of paths and expected permissions that will be called by Daemons at startup and the common API implementation will throw out if there is a mismatch +Im just wondering about the testing both manual and automated +Remove local parameter and determine from the path URI given on the command line +I have tested the tutorial in the EJB RC seems as if it works +RuchithThe problem is at server who recieves the empty password +Does your implementation work for all activity typesI do have an idea on how to implement it but the test coverage grows exponentially if you want to test all the cases of course +This child +Im fine with it +on patch v for committed addendums for trunk to move the to the package + to trunk. +Ill see if I can make any simple changes to the current code to get the desired +I shouldnt have closed this as duplicate +Thanks +Committed to sqoop +Patch submitted as per Patricks contribution +I just committed this to the branch +Today distlibrequired contains Hibernate Core as well as Hibernate Core specific think our main distribution and use case will always be the Hibernate Core users or we should split the name and the documentation tooShouldnt we keep this simple and keep the distlib directories focused on the Core integration? using engine only is really hardcore and requires additional coding like the implementation and direct invocation of the indexing API +This patch makes the improvements suggested by Knut Doesnt fork while compiling the JDBC support Compiles the JDBC support to Java byte code if possible Removes the optional prefixes on the JDBC modules in Touches the following filesM javaengineorgapachederbyimpljdbcM javaengineorgapachederbyiapijdbcM javaengineorgapachederbyjdbcM javaengineorgapachederbyM javaengineM javatestingorgapachederbyTestingfunctionTeststestsjdbcM javaclientM +If you prepare a new patch with bugfix please dont use the funny test method names with spaces in them +index stores need to use the data types defined by system store to match with uuid based look ups +Currently is the most widely used Wicket version at this moment reverting will cause pain to all upgraders +Now that I think of it minfailures or maxfailures depends on how you look at the problem +the exception stack you try the same with? If it works then there is something new that you have to face +I commented them out in trunk in the fix for and havent seen any bad effects dependency is needed +new patch with more this done now Simone +The fix is to ensure that the update validator uses the proper view symbolsThis should be rare on as the metadata validation will properly initialize the cached metadata but since the entries can expire from the cache its possible that it could occur +So basically you dont have to do the checks for the timestamp because you know that none of the storefiles after the storefile where you encountered the entry are deletedRegards Erik +Could somebody take a look at committing this patch +basic tests to reduce the need for validation to run + +Usage of getKey requires users to provide a security token +Thanks Dag +Kazakov got an idea of what is causing this? is the examples outofdate for AS +Is there still some question as to whether or not cloudstack wishes to pursue this +That doesnt work because is a test classIm also wordering if it is possible to fix this issue without introducing too much changes to the journalcluster code +The build scripts are already awkward to maintain and what little maintenance occurs happens as in this issue enough to fix the immediate problem +Eli can we please open a single jira to fix findbugs warnings? Why bother with so much noise? Thanks +If not ill run it anyway whenever you commitCheers +In the merged uber jar replaced the split setup so the javamail jar file is the api replacement for the old javamail spec jar +It will go out with the next release. +Convert the unit test to a proper unit test without needing a +Patch attached to instead init the vars as default then parse from the uri and finally allow overriding with env vars +I confirm Im on Eclipse running with Java +Looks like this is related to is a workaround by adding T false when enter gshell +Thanks for your comments +Stuff like +I made one small change prepended an sq to all the gif files to group them all together for this paperCommitted revision +Yeah it would be nice if there were a test somewhere for the combinations of new open arguments +Updating fixed it +to remove Fix Version +enoughben +I was unable to come up with a good fix for this in timeframe so were going to punt to +Along with the fix for I also attend to fix this issueThe patch is ready for reviewI run and derbyall on insane and sane build +Did you check the CLASSPATH? If binary tgz then at top level there will be hbasejar +Ill investigate +Im working on it +In your case you use the XercesXalan bundled with part of JDK you might want to try updating them +have been closed. +It needs to be downloaded from the Server +Generating random data with the command line tool has many other uses tooDoes that sound reasonable? If we do this I think it makes sense to put the Tool portion into a different ticket since it will need its own section in + +Working with it +with +If you havent done so already I suggest you enable checkstyle integration with your Java IDE and check the checkstyle report prior to submitting any patchesNo need to submit a new patch just for checkstyle as I have corrected the checkstyle warnings locally +Yeah this sounds like a workable solutionI feel that in the long time as configurable parameters and annotations accumulate in Jackson it would be nice to have central document describing in detail the exact configuration space of the library +Maybe Im not looking at the right thingsWhich file is getting truncated for you? Maybe theres an issue with older versions of gnu tar +Marking closed. +IO Stats for +Did miss that +the followingThe jobtracker should send the following two messages KILLTASK which is per task message +So it would be better to read out the classpath of the project and include whatever jars are found in the classpathHowever this could be quite a bit and some jars may already be available on the cluster in hadoops classpath so there needs to be some more configurability exclude jars or include classpath entries in the manifest files if were certain that some jars are available on the cluster +I will update the online Installation Guide for and but I will leave this issue open and set the fix to and so it can be recorded in those release notes for when this fix is included in the documentation bundle that is shipped with the software +I would say it this way Whats the right fit What role could ldap play What are the specific goals we are trying to solve How do we make opssyseng guys comfortable with our decisions +code with agreement to allow the Apache for the patch +Here static synchronized method is replaced by a method in a class object +So perhaps we do need this for the release +Given that current rpc can handle MB arrays this is something to look forI will report back the default limit soon +Assigned to LucianNow you have to provide more patches Lucian +Then were okbq +I will be out of office until January th +Maybe it should be deleted from SVN +Can you provide a patch if you fixed the bug? tried with Jetty and Sun JDK update and it worksI am resolving this as incompleteCan you provide your environment +ChunhuiAny updates on this testing +XHR is already supported on BB as to not implement this on Android back in February and closed the subtask +concrete example every queued exceptionEvent should provide not null component form if exception occured in placeThe main goal of whole is to unify error handling so factionListener and actionListener and their consistent behaviour are part of the goalPlease see attached if it breaks something the purpose is queue component with and thats all +on the project instead of manually starting and typing in the demo should be speed up a hint should explain that archives can be used with any project not just JSFWTP project +Under current behavior it is attempted regardless of what pollers you use +I couldnt find a changelog for recent releases of the gems to determine if their changes would have an impactPerhaps Charles Nutter or Jamis Buck would have further insight +I got it PASSED times in a row with this cmd linecp +ThanksSupun +Missing Signature javaxinject does not exist for +I only need the headers for client +Christian the transaction is NOT committed just because resources agree to prepare +Patch fixes both Fetcher and Parser gets a new fs only if it is version of the latest patch runs smoothly +fixes a case where ROOT isnt recovered after a regionserver hard kill type crash +If this necessitates a then bump to +Please remember to set the Security Level +Also a bit of process in when you upload a patch please mark the box indicating that it is intended for inclusion in Apache also please then hit the submit patch button +Has this issue been seen sinceIf not Id like to see this closed as either a duplicate or cannot reproduce +I believe you havent updated since the first of August +Its a very simple js interface into Couchdb and works reasonably well until there is a better alternative +So I only have the good shapefile which is working +waiting for Manik to verify this and for Vladimir to add his changes regarding service views and Multiplexer so I can release +in the tarball and Ive verified that ant eclipse succeeds from the tarball +Verified the fix with the latest build +If you could kindly take stab at the patch +Attaching and various other changes to to turn it into a system whereby the user can pick and choose what escapeXyz means for themStill a work in progress +In other words well remove the API part right +This issue seems to be fixed together withAt least I cant reproduce it since r. +Move the call to method can inside obtain +Moving Steves note from relnotesbq +sbt update in order for and to appear in +Still its not obvious what the intent is and it needs to be fixed +Seems similar to. +Simple rpc performance tester +Thanks! +AlejandroWhich problem you are referring to? If you use hive as is you should not see any such exception +OK I removed the zero length filesShould should the empty src directory hierarchy also be removed or should it be left in place when other tests become deprecated +Lance still want to implement this +Thank you Raja Aluri Davio Lao Sumadhur Reddy Bolli Ahmed El Baz Kanna Karanam Chuan Liu Ivan Mitic Chris Nauroth and Bikas Saha. +Thanks Sijie and Gavin +I still plan to do it during this or following week +The attached patch is not intended to be commited +This needs to be fixed in a hdfs +So we must move our users to some org unit and apply class to all our user accounts if we want to useAnd there is another big problem +Committed +limiting for a specific resource +These problems have been solved in afaik. +Apache currently has other attorneys besides me +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +Thanks JMThe current one does no harm so there is no reason to revert it imho +Are you guys saying that this patch needs to go into as well? But this is a new feature isnt it +Notification scheme created and associated +This would give the ability to route all the data formats defined by specification through Camel +Corrected at rev +thanks. +Patch with YARN only changesIt looks good to me will check it in once Jenkins says okay +Patch submitted +Was this just missed +An would be preferable however if that is done then there is technically still a race condition when the baseUUID field is initialised +This seemed to be the best approach because we did not have to copy over or change the +Thanks for taking this up! I have some ideas on the commons packages Ill follow up on dev +Great point Todd +Alright Ill make matchVersion determine the default value of legacy which should allow this patch to be applied to pre code lines at the users riskbq is branched +Two thoughts that might be relevant seems to have started failing much more often recently but I dont have evidence for this +Fixed in r. +I will fix the comment before I commitI will wait for hadoopqa and give other people couple days to review as well +I know how to do this for sure but thanksMy question was dedicated trustStore file reloading without the need of restarting applicationAny ideas +we should at least move the package from toYes +Thanks for responding Nick +AdityaCan you remove the sentence in Release Notess the general thought here? Reapply the patchThe timeout after the patch is not really a timeout any longer so at the very least the config option is misnamed nowOr in different words What is the point of this timeout now +I believe the former would mean adding an override that doesnt actually override and then when the resource DS is enabled removing it and replacing it with the current then reversing that during doable but tricky +since i was dealing with node types anyway for the privilege mgti can take care of this +Report bugs in SQL Server to Microsoft +While looking and testing more I found one more additional point to discussed here with respect to OOTB and that is making the product url as business defined +bq +New patch added unit tests +file is available +The Legendary Former Leader assigned the domain for safekeeping +I too like the idea of having binhadoop dfs refreshConfig that will reload the config from the conf file diff the newly read ones with the ones in the +released closing. +No problem +Fixed in commit ededf by upgrading to Apache HTTP Client which was just released today including the fixes for and. +Jeff is it possible to repeat the load for Turbine? If it did get done for Torque that is +If you change the bundle version than the contents of the bundle changes as the version is included in the jar +The Hadoop code is included in the patch +Attached to correct this +Same basic issue occurs with groovyshWhen using groovysh load at the command line UI is closed using groovysh at the command line and then entering in load in the shell UI stays t the behavior of groovysh load and groovysh load be consistent +Working on Seam Catch +Im not a fan of +Committed this to branch +From beta find +Marking as regression +Its possible that the leak is deliberate +Aruns patch worked for meThanks +would be nice if someone could test this issue with cvs head at last from with JDK and it works there any way that we could have a unit test for this +sorry poor distinction in the request In one case the user is asking I would like to open a support case with RHT In another case the user is asking Heres a problem got any solutionsI think the ideal workflow would be something integrated like + Im understanding this correctly then it looks ok to me +This is only suitable for development and demonstration scenarios +Marking as resolved nowWent through this again in depth with Eugene Schava. +Verified in CR buildfind CR name modcluster printCRmodclusterCRmodclusterfind CRsigned name modcluster printCRsignedmodcluster +batch transition resolvedfixed bugs to closedfixed +This is expected behaviour +Ill backport this one to after release. +bq +passes +Looks fine thanks +Ill see if theres a way to fix that even if we cant make it match the original files encoding exactly +I will wait for Francesco to merge his branch and then redo my refactoring before attaching a patch here +Hopefully this will address some of your slowness issues +Hi folks now that has been committed I would like to resurrect the discussion about fileids for HDFS files +Hi GrgoryThis is a feature in which is not released yet +Please check that it was applied as expected +As commented on at this point I dont plan to convert this +Following bytes are invalid +After getting your latest Archive and Segment I see the same error you see +Closing resolved issues for already released versions +Framework should clear flash scope when pausing arguably always regardless of whether a view renders or not +Guido would be me Now as for the setup could you elaborate what do you have in mind? I can see how signing would be one of the things that this new service would have to handle but what else? +Looks goodcommitting now +Just added a corresponding query property to +Waiting till Enis gets an answer before committing +closed on trunk and assume you mean the current Map? Ie you would like to export what is on screen labels and all? +Thanks in advance +Also this patch breaks recovery mode unless toleration is turned off +Please verify +Changes look fine to me +If a boot with failed authentication shuts the database down then atleast it has to ensure that no valid user has connected to it since itwas booted +I swear to christ I clicked the for contribution button +I do think its because there is no type converter that converts the list of string to something that can be streamed to the file +The clipboard content is actually not duplicated but the preceding rule +Most of the users never have to write custom components for Solr +Id prefer a naming which would omit the Groovy version for the compliant artifacts and keep Groovy version in the compatible jar file names +The team has decided to abandon the effort of getting this to work +Committed to trunk and +first patch +Is it alright if I check this inPerhaps I should add a note about this somewhere in the docs +Perhaps later +BTW I patched the ruby client to perform client side redelivery so its a now for ruby this will require some thought to get right and does not seem like its critical feature users are requesting +Where can I try this fix? Is this in a new version of harmony or in a snapshot build? I checked the subversion commits for this issue and didnt find any +Thanks! Ive also updated the BCB in the same directory +Good point +Have you changed the repository info to report that getDescendants and get are now supported? The server framework checks the repository info if it has to generate the descendants and folder tree linksFrom the client perspective it looks like thisThe folder Atom entry should have two down links if getDescendants is supported +I just committed this. +Can you check installer from JBDS build? It is built with last version +Certainly in a loop with heap dumps enabled is not a thing we really ought to be doing +I havent been able to fully test this on a broken eclipse but looking at the source yesterday Im certain it will recreate the issue and I just dont have time right now to go back and break my eclipse again to prove it I already lost quite a bit of time on this one +Closing all resolved tickets from or older +Committed revision +I had to initialize maxPos to in the constructor of that does not accept length as parameter +I will give it a try and report back with the resultRegardsMauriceVerstuurd vanaf mijn iPhone +I think it is more appropriate for users to optionally specify inputoutput data types as a part of UDF implementation rather than in Pig Script +sureThis was part of a bigger effort to produce a decent logging facility for native code JS code coexisted with weinre etc +New copy of ports used by JON +After looking further we decided that we will investigate the possibilities of providing our own as a caching layer over a persistent store instead. +Leonardo responded before this change +Findbugs defines equals but not hashCode is while addressing comments from patch addresses above warning is unrelated to this patch +bq +r on the future reference use the attached patch with every future update of to check gem install gem update gem installupdate on windows. +Can someone explain what is the workaround +Committed to revision +After this is implemented evaluate whether is even possible with the way shareable nodes are implemented and how the indexes are designed +Stop the configuration while its failed to startPlease help to review it and thanks for any comment +Dan Im assigning to you to back out the Bijection changes +This blocks deployment Tap apps to GAE +The best course of action so far suggested is to have moderators forcibly unsubscribe individual addresses that are used to send spam form the lists that recieve them but this seems less then ideal since +Later we could do the scan only in the deployment process which means no need to do it again while starting up the web application +hmm not sure if we really should add this +Attaching the patch named incorporating Andys review and let me know +dont compile them +Oh I follow you +Remy does this fix look ok to you +One of us must be missing something +looks good mahadev +This came up on IRC today +attached a patch that adds a check if the referred file exists in the snapshot +You either have an account or you dont +Tomas could you please create simple test project that would allow to reproduce this issue? Thanks +Use md when available +No not critical +No problem delaying for a while +No further changes +Committed patch to trunk at subversion revision +Attaching patch which fixes this problemand also and the testcase in the body of this issue I have these remarksTemplate makeElement makes sure that the default namespace is used rather than generating ns mapping to the same URI and using thatTemplate makeElement also makes sure that ns ns +Weve worked around this issue but I dont think its resolved yet since I dont think all these properties should be mandatory +I strongly agree this advice +We can also see that the Total blocked and Total waited on the Reference Handler just keeps climbing without time you see State WAITING on Lockdccc as you please see if the latest code from Git solves this? There were changes since had a chance to grab the nightly snapshot on the st and run it over our project it works several timess faster than without hangups +This patch resolves this memory leak by destroying completion entries in deserializemulti functionPlease find attached patch for the sameCould you please review it and let me know if you have any commentsThanks Regards Thanks for the patch +Not sure what you mean? Where is that data held? As far as i know does not contain it +helps when I have too many clusters floating around +Currently they all pass +Duplicate of +The problem was a bug in the system test it wasnt actually rebuilding the see abovev applies cleanly after rebase and passes +Corrected Components and Affected Versions. +Id be more than happy to assist in testing or performing experiments to resolve this issue +Sure +We experienced some problems with both and going downThings should now build properly with a recent version of maven +Until now we have been trying to avoid adding TEZ layers over stuff that YARN already does +You have to build the server from trunk if you want to gibe Kerberos a try +Checked in to multiquery branch +Will close this jira after getting in +It has been fixed in the meantime +There is no STS involved just with a symmetric binding +introduces new design for reading flat files so well figure out how to handle headers there. +I committed these tests +I believe I just tested this last week when doing a test for an unrelated open JIRA +The sum could easily pass +This should include a tutorial on making apps run in a cluster for the booking to a cluster couldnt be simpler now +Patch committed in revision +bulk defer of unresolved bugs to +Leaving it in the bug is probably good enough +When creatingdeleting the list of users on Set RightsPassword and Delete User are not updated until you click View Users +What exactly is the relationship between this and the class in the java package? looks like there is a file +we should have an alternate sql statement for derbyI did research and seems that derby doesnt support REPLACE +Did I miss a discussion on one the mailing lists or something +TDD for the winAnyway thank you for your acknowledgmentKind Regards test case +start a deployment using conductorget instance via dcloud and it worked +Let me know if there are any issues +I attached here for reference +A fixed reply queue will work but has more adminsitrative overhead +Committed revision Fixed this issue of doubly patched files +The snapshot does work with both JBOSS AS and tomcat +Ill have one or more additional patches in a day or twoEric +Of many the I have seen I have not found the ASL in themAlso several of the +The still asks for the password +Do you have the same experience with selectNaive vs +Would one thread win and the other Tag would never be addedPessimistic locking has its own quirks too so well have to think about this a little more +Justin something odd is going on +bq +powerflasher people answer about early to start thinking of such the future of the Flex environment is mostly behind us Im closing this ticket wont fix +I saw that edits going to since deleted regions will still create a directory and put the there +I have added one testcase for this scenario +I just committed this +Damn I let this one slip simply forgot about it +I think you can mark this resolved +In the end we just support and warn the user that the rest arent available for managed beans +Updated for Helios M +Comments +now writes term vectors in order. +Please dont forget to checkin in too +Jason in case youve got some spare cycles you can have a look at it otherwise Im scheduling it for for the moment +It should return Test passed +In what context has this occurred? Which version of Xerces? Do you have a test case +It seems to take more than minutes now though it used to be less than minutes +Committed as part of the commit. +You should get a with the cause message now +I will explore and get a fix in +Are you using any container such as Apache Karaf or something +Woops +just gets a new nrt reader from the writer and passes this along to the Searcher a lean searcher with no state +Attachment has been added with description applied in r +Im using a combo of on trunk and reverting to the java comparator it might the combo of the two that is required +Closing this issue. +type f name +Fix the problem by changing the function compareAdd a unit test to verify itPlease review +Then what we need is judging whether the oldSchema is a instance of if its not we need to directly return it +I did not mean the class itself but rather its parameterless constructorthe one that causes an exception in my patchOleg +One of the most common questions we hear from people migrating to MR is how they can restrict the number of tasks per node +This should make that testing a lot simpler in the future +Michael can clarify but it seems that this code is meant to deal with thinks like logs directories in Hadoop +The patch +Using a more sophisticated algorithm that uses a decaying moving average over some period of time seems like that might have a better impact +backported to the branch with rev . +Half of the is fixed +I got quite a few leads +Out of date +email +Can you either leave this TODO in there or file a JIRA for itFiled +Hi DennisIve seen the comments related to the and also the reason why the Mylin layer could not be will look today on Maven SCM to see how the abstraction layer is m planning to come back with a design proposal any time soon +Fixed in +This is the the messy statebq +In the test is it better to add the case of empty type string +What I applied +Last report is this is fixed. +I think this is ready to be committed +you were watching iostats and CPU usage while a large file was being streamedThere is definitely merit to supporting concurrent streaming for a given sourcetarget Im not arguing against that +Uploaded the patch with the above suggested modifications done +would mean all columns not referencedIn this initial implementation I am planning to support only all columns in range +Pull request merged +The scope hasnt been used by the Maven plugin for some time so it was equivalent to not specifying the scope +Applied +what are the rules within the ACL +seems like this was some transient or user type error so closing +Appologies there is a slight error in my sample application +Troubleshooted with Kannan +Will change the test to add the serialization spec rather than put back the edit +Thanks Ray +Your latest changes fixed the PRG problem I was having with data redisplaying +See issue revert the great split in jackrabbit mailing list +CommittedThanks Vladmir! +What do you mean by support? The apache project is supported by the community +queue queue xqueue all legal +Let me know how I can helpThis may not be a Mahout issue though right? Sounds like you intend to modify Drupals +batch transition resolvedfixed bugs to closedfixed +Seems like this is done +Are these correctInstall the Management Server on the First HostThe Management server can be installed using either RPM or DEB packages + +Refactored the api to allow better access and used contextualize to get the real plugin container +Lars The version deletes are some what like a grey area for me +its just an svn diff against the trunk +And also a fix that makes amy dfsclient methods return error if the dfsclient was unable to communicate with the dfs servers for a configured period of time +Thanks +Your function promises to do one thing but it also does other hiddenthings +You are right calling mkdir on +Thanks +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Namit Ill make another pass through the patch later today +Hope this helps +This is not good in that many requests may be firedoff and the singleton plexus component will continue to funnel their output to the samelogger +Thanks +ds instanceof I guess thats simple enough Assumes application is running with Java and is but youre right thatd be the easies route +We can monitor ensemble changes by watching the node in +it ensures bottom up BA recovery is also patches another recovery related problem where the is field of the BA recovery record was incorrectly initialized when persisting a coordinator completion participant +Im sorry! Maybe someone can use my code anywayHere is a list of the characters causing different resultscharValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValuE false true charValuF false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValu false true charValuA false true charValuB false true charValu true falsecharValuE true falsecharValu true falsecharValu true falsecharValu true falsecharValu true falsecharValu true falsecharValu true falsecharValu true falsecharValu true falsecharValu true falsecharValuA true falsecharValuB true falsecharValu true falsecharValu true falsecharValuF true falsecharValu true falseGreetings as WONTFIX as there would be functionality changes. + +There was a release of many things several days ago all present on the DN repo yet has nothing more recent than October +Committing +Thanks thanks Jingwei +Attaching my config is code comment describing the only case when model tries to create linkOnly create link if actual resource exists and all other means to find resource in workspace failed +As for well that is a big API in itself and Im really only looking to take on a couple of common use cases here not develop a HTTP API for Groovy Plus Groovy doesnt depend on at the moment does it? So in any case I think the Groovy API should have something for those simple HTTP POST cases +Thanks Chirag and Sumit for fix. +Update the JIRA status +I will have the patch uploaded very soon +Was just trying to Andrea Thanks for helping +Himanshubq +A unit test should have caught this? See the same forum post for the result +Any progress on resolving this issue +if that itself is batch oriented then that may explain a large number of deletions at the same time +Just comitted this thanks Patrick. +There youll have an opportunity to set a username and password to ones accepted by your it seems your username and passwords are incorrect Its receiving your username admin and is most likely receiving the password as well however its refusing to authenticate +Works now thanks Hardy +Currently balancer does not check if the cluster is finalizedUpgrage d or not +This issue will be fixed by in. +Which section of JDBC spec indicates these methods are required for compilanceI see no reference to in sections or of JDBC +These test classes should be marked abstract but the test suite needs to be cleaned before making this change because these test classes are being direcly instantiated by other tests. +Actually +This is basically a duplicate of +I need to just add a dependency on and all is well +Fixed by updating the file in trunk to depend on as perAlso updated the myfaces dependency to match the current tooThe alternative would be to generate a new ant script from the current but that would need more testingThe commons project now uses maven almost exclusively as its build tool so ant build scripts can easily get nowMaybe we should just start removing antmaven scripts nowThe release appears ok to meThanks for reporting this Isaac. +Thejas Thanks for the commit to branch issue has been fixed and released as part of release +Im pretty sure weve never disallowed system keyspace writes intentionally +validates this +patch applied in failed and was excluded looks good thanks Mikhail +bulk defer to defer to +Attach patch v that requires either or to be present +Verified on the latest Build id Build date in +I forgot! this issue was on the Release release +Updated the patch to add a test case and rename check to for the reviews Alejandro and Arun +is updated. +Its weired +Ill put the in a gistrepo for us to update as we see fit +Getting better isolation guarantees for the cost of a simple lookup in the application tier is a good thing IMODowngrading to Minor +Its requirement for websphere specific deployment descriptors like +It looks like the patch requires order clauses to contain either ASC or DESC I think itd be good to make the direction optional +Note that the collection does not include only objects like the default insert sorting algorithm assumes +Im looking at reviewing committing this patch +Issue may already be resolved may be invalid or may never be fixed +I will commit shortly backport to +Want to double check it before I go +Thank you very much! SincerelyMatthew RichtermrichterMDVBS +Sorry if this spams things however its unlikely that Ill work on these. +Proposed patch +I will have a look at issue is a duplicate of which got fixed in Upgrading to should resolve the issue. +but all other things being equal lets keep the behavior working the way it has to avoid suprisesThis attitude makes me sick +Just found a problem +It maintains consistency with out build system that sets them in the first place +Will work on git brancnes in the in Code that I had forgotten to check in +So maybe the more appropriate location for this test is in the class +It will create databases bigdb and smalldb + +Running HDFS loads a lot of those classes +I updated the Dependency Guide to reflect this and in our next distro well make sure to include the activation jar +Good stuff thanks! Ive committed the patches to Tika trunk in revisions and Note that I took the liberty of moving the class to the package to avoid too many small packages +But after that Im gonna add commands for describingreading resources in a friendlier format so this could be included +Im saying that the problem isnt the parameter but that the error message in the description indicates that the parameter is the problem. +It is going to introduce some dependency on a new library +This is going to require some extreme redesign and currently we dont have any that this would support +In general however it should be noted that the client is not well behaved and can get protocol errors and other symptoms when the server exits while it is still connected +Im not happy with the default situation being that service events return before the response to them is complete +Just did some analysis under normal conditions +You shoudnt be using your functional tests to assess performance as the functional tests are looking at the corner cases trying to break things not simulate code +Lets commit this fix for now +SincerelyTomohiro Ito +it looks like the modcluster configuration is incorrect like it is using the default keystore and that keystore is not the same in all nodes +Patch looks good committing to trunk +Attach your notes to this Jira and reassign to me when done +The pdf uses an indexed colorspace and pdfbox didnt extract the base colorspace to calculate the correct parameter needed for creating the colormodelI fixed this issue in revision + +Although I do not think there are many of these and a test run will show +Triaged for assigned normal urgency. +To complete this work the EJB metadata referenced in needs to be completed +Change has been made but wont filter through until next Savara milestone release +Using I guess the valuable line is the one ending with +The patch has been applied with thanks to Marc +The customer I was helping is using Camel within SMX +With I could not find a code path that could possibly cause this NPE +While committing the svn executable flag is usually lost + now sets proper status to underlaying when beeing altered +OK this should be fixed now through the use of a unified mutex for singleton access +The workaround is not needed anymore +Attaching version of the spec for this first increment +Let me know +Yeah +this bug was fixed in just checked under online demosite +I suggest that you upgrade to Xalan where template inlining is off by default +EPP ships with so I assume that it will require some kind of patch release of PLIDM for bug fixes and can go to +Ganglia graphs work with this in place +Ideally the new version is a search component such that one doesnt have to send separate requests either +If someone has some thoughts feel free to. +I cant reproduce with Apple JVM and Oracle JVM on OS X +Then an additional round of negotiation would optionally happen hence the idea for using invalid calls after the initial connection header processing +document as fixed issue existing release note and made it suitable for inclusion as a Resolved Issue +Hi Antonio Enhancer is required to make openjpa work +Please review patch for jce install after security JDK location is initialized in the and made a part of every Action +Working with git repo URL provided by Gorkem +not only shows that you can inject into CMT bean but also you can actually use it +Fixed the problem and added a test case +Im making some tests +And then the would not be able to overload the parents static methodUnless I am missing something +Instead it should be done as a separate JIRA +Attaching +My apologies for not updating the issue with this information earlier +Apart from we have to just add more documentation instead of adding any methods we already decided to freeze the API +Nope sorry +Im not sure if we should make a module for it +So any new snapshjots should sort if you can check this is ok then we should be able to close this JIRA +Im not sure when it was reintroduced but I am currently running the snapshot + it will still be in SVN for reference if anyone needs it +Turns out the test basically duplicated the same faulty logic +Committed revision +Here the problem is that it sends a reply which is not according to its wsdlThen the other options are +The configuration required is already covered in documentation +when making injection by name I took the name of the file as shown in the Project view and not the name of the class inside good grails ide should show the class name +It is rather annoying if you use nodes with fixed it BTW so closing theExpressions is the easiest way to access Seam components thats why Im using it so muchbtw it works for me perfectly with the latest Jboss tools plugin for eclipsebtw when the SP release should be outMany thanks for taking care of this issueAsaf +Does that seem rightLet us know so we can get this resolved +Failed test case is not related to vaidya contrib project and not caused by current patch + test +I will look into it this weekI will also look in the dependencies +Resolved on HEADRevision Changes srcorgjgroupsmux on branchRevision Changes srcorgjgroupsmux test to be symmetric +The response is received from the server and the main method exits but there is still a thread hanging around probably the thread launched to handle the sendreceive +form was too long for the ODE build +I also implemented the memory table by just skipping the combine phase as there is indeed no guarantee this is combining anything anywayAs for the that would of course also work +this can be closed. +sound good for me +I can do it but there is an extra commit in Branch related to which I am not sure if it should be included in the tag or not as is still open. +Yes its the two issues ive and i suppose James have the same +The problem in the has been fixed before introducing your impl. +Can you please check the scrollbar problem on the Mac too +what does it meangetRow can gets data by rowKey this data exists in table but the scanner doesnt see this row and return rows +Which means that you dont get the performance gain that promises for expressions +As soon as Ive written them Ill add them +Everything should be set up +Good luck you just have to live with it or spend the time to understand the Maven sources by Martin Todorov comments for a work around +Thanks Suresh for the reviewResolving this. +Thanks everyone +I was actually able to trigger this on as well I misunderstood some of the other devs testing +Attached a quickstart showing the problem +bq +I tested the combination of the latest Axis and the latest StubThe latest Axis compiled the latest code in CVS +Fixed broken patch +For the same I have added one more parameter in argument list so we can get the lowest level workEffots on the basis of TypeThanks Ratnesh Upadhyay for discussing implementation plan Awdesh for valuable suggestion and Harsha for extensively testing it +Fixed +of course without a Would you in this case find it better +Thanks Ravi +Were supposed to drop support for Hadoop versions anyway +Michal Pavel Rajesh given shouldnt we be able to resolve this as fixed in +Committed to issue as it has been released as part of release +jansteymojoxapache svn ci m stop routing slip when exchange is failedSending srcmainjavaorgapachecamelprocessorAdding srctestjavaorgapachecamelprocessorroutingslipSending srctestjavaorgapachecamelprocessorroutingslipTransmitting file dataCommitted revision +After looking at this its not coming from the webview at all +Surely two ASF products cannot be rd party to each otherAlso I think the LICENSE file needs to be specific about which rd party products are presentIt should itemise each product and point to its license +Abhijeetactually columns parameter was extended to be supported in Export mode in addition to import inThe reasoning behind is quite simple Sqoop on export checks the table in database and assumes that input data have all the columns +The right patch this time +Jerome could you be more specific? I dont understand the question +I created issue for your bug report +examplesfobasic foo The following doesntrichardbirkenhead cygdrivefdir with spaces cd +Next step is to write more junits and updatewrite javadocs +Thanks Tanping. +The same for htmltext plain if this is the content we want to send then we shouldnt accept to return other content types +yeah yeah +Thanks Isaac +This would be cool +Please report this to the author of the plugin +We should remove the variable and always use for that port since it absolutely doesnt matter if they are on consistent ports or not +move to think the src packages are ok now +We dont want to set a system property but rather add a property in +Simplest test case I could find +Nadya as I see you omitted my vmdoc mods in particular enabling of predefined macro expansion this should badly affect generated docs +Youd still need to device a test to tell you its making a mulivalue rather than concatenating as well +Your patch has a bunch of unneeded yaml changes +From point of view that sounds good +Im not expecting to betreated differently +The file trunkexamplessrcmainjavaorgapachemahoutregressionpenalizedlinear could not be found in the repositoryI always get this error message + +Still need to think about how to handle this +There is one single thread pool with threads in the pool +What didnt you understand about thatbq +Confusing +Remove small interval codes which will be digged later +Anyway I added a check for on update because it is possible in jsf to reach that point with a non serializable datamodel and the are not side effects if that so +Edited issue title to broaden marked as no longer valid +What do you mean by all Community contributors? Do you mean everyone or a subset +Getting into the security area is not something we should take lightly +Ive committed at r please verify +Since there was no publicly hosted jar for Transfer API I moved the code of during testing +Is it possible to have in those areas the same problem reported inThe idea is to use a pool package or to build our own + to trunk. +Please open an issue if you find bugs or for improvements +I cant reproduce this with trunk as of it seems to be fixed +Verified by Oleg. +OK I just committed this thanks Eks! +The patch looks good but this patch has gone stale due to +Thanks! +This might have caused the thread to block on the socketWrite +Can you post a test case showing the issue +so messages dispatched to you and no ack +Its not clear to me yet why it lands where it lands though so far +Im seeing the same behavior +Tested locally the issue is fixed. +Certainly sounds like an IBM JVM bug or perhaps more likely an issue in their XML parser that makes it safe +HiWeve never used +The bug occurs in multiple places in Loader +Updated comments about fs and gs registers +What happened to these patchesIm especially interested in the jvm reuse since we not only run large jobs but also a lot of small ones where the setup time really kills of these things except for the jvm reuse have been done in other patches +although it looks strange foo is a valid expression on its own so putting that on a new line separates it from the line before +No HEAD is supposed to return exactly the same headers as GET just without the bodyIt can be used by clients to determine the size of a resource before downloading it so the is definitely useful +Wow that was fast +Thanks Shai +Done +I started now with the selection view but it will take longer than expected +what is the best way to fix this bug? I am clueless to which part of the source to look I prefer this fix rather than fixing staging plugin since it may make the stage pluginfix kind a http specific +Also thanks for entering this in JIRA +Issue closed. +Otherwise the stacktrace is ok +It indeed sounds like a problem with how the richeditor is retrieving its resources +Verified by Oliver and someone has applied the svnignore updates. +see +OK +Could you add a key shortcut to the menu Clear Script ContextSure +That seems like a genuine bug yes +Maybe moving the Driver creation to the thread will fix this +Please fill a Jira in Felix +I feel bad for not being able to help more with actually doing this and really appreciate your quick work so far +Merged to +Improved with rThanks! +Thanks. +Ill go do that now. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Nicholas fixed this with. +I dont remember the issues number. +Alex is trying to make it work on Solaris sparc atm +Thanks for the fix Ivan +The code in is at the moment adding an UPPER to every LIKE operation +tested! +lock file was blocking all good now. +Should havejust let the exception propogate +Just got the same problem in the same configuration without running anything under a profilerINFO Service Sync error occurred Java heap space Java heap spaceINFO Service Sync error occurred Java heap space Java heap spaceINFO Service Sync error occurred Java heap space Java heap spaceINFO Service Sync error occurred Java heap space Java heap spaceINFO Service Sync error occurred Java heap space Java heap spaceI guess I could specify some larger xmx value but because the back end is postgres I dont think the number of messages in the queue should have anything to do with the amount of memory the server needs +I have just seen an open issue in Jetty as well similar to it +In theory a release shouldnt be required not everyone uses the release feature of Continuum and they shouldnt be forced to care about as it apparently is required sending them to the configuration screen as we do on startup is better than throwing +I had to wait a while does recovery occur if you restart the server +Ah I hadnt checked for trailing whitespace on the comment lines +Thanks to Thomas Enbo for pointing me in the right direction +Fixed in +Can the scheduler look for modification time of the config file to decide whether to read it in or not +Id love inputIm on this approach because we need to keep things for existing monitoring solutions but ultimately Im unhappy with the current state of JMX and starting fresh like you have here is a good way to solve it +The becomes bloated if the application has many externalized queries +That looks good thanks Tim +I misread the spec +Patch to remove old classes applied to branch r +All done bar inheritance over composition See +AlejandroTom Any pointers +Thanks for reviewing the patch +Except that I have no idea when would be used or useful +JrmePlease create a separate issue for the error handing ideas +I was looking into merging the patch in and it conflicted with this change +At some point we might want to think about an Extras subproject where we candistribute goodies like and other cool conveniences not essentialto Struts Core +HiI confirm since I have this error on many places where there is no Thread at allThe only workaround I found is disabling this rule since it is false positiveAlso is in in +Thorsten this is assigned to you what is the status now we have dispatcher +Could you please suggest where exactly we can fix this issue. +Why is the cleanup important +They are coming from client +Where would a plugin fit into this or how would a plugin help this? Also what kind of plugin do you mean +is that trying to show that is was sorted +Closed Not a bug. +no strong feelings +Closing old resolved issues +Use aboveIf think its better not what we want to do here is to the cluster +I was thinking the assemblies in bin were for some reason +Nope we have plenty of free to improve logging for initLimit being reached +Its IE +simplify ShowHide ExpandCollapse functionality so the first selected item is a pattern to toggle selected items state +Fixed on and closing all issues that have been in resolved state for more than a month without further comments +Running run Failures Errors Skipped Time elapsed +A small app that illustrates the problem +The patch fixes the seen on Windowsin the tests +So the error can happen even when there are failed tasks +Greatthis feature can resolve the aggregate functions? now we use the redis to do the aggregate in the memorybut when data increasedthe memory cannot load the we meet the bottleneck +I dont know how to reproduce this issue for sure +Patch applied thanks. +This may explain the differences +Committed as r r r r and r. +It occurs randomly on Continuum +Annoyingly its going to take some thought +I was able to simulate and fix it only at the trunk not at that is the promoted RC +Looks great! I tested the back compatibility by applying the patch then reverting the changes to and +That the console is in reality from groovy +Unless we prefer rcuprer +Posting the patch that augments the API as in the description +Chris thats a good pointThe current design seems to be that Field can do everything and the others are simply sugar on topPersonally I think this is confusing andthats why i wrote such a huge test but its sillyIn my opinion if i have a it shouldnt have a setReader method +patch looks good +BTW the implementation of this is relatively simple the put set method would Check if a Map were the latest object pushed to the stack +Any progress on this +I created another issue based on well formed issue created +Yep and we need to change to using a +I just put up the class +setting to CANNOT REPRODUCE since we now both cannot reproduce +Now my fails with could not initialize proxy no can use JPA if you wish +Reading the patch again there is a minor bug the document cache is updated with the parsed document even if it was fetched from the cache +Let me know if I missed onesome and Ill remove itthem. +It would be nice to have this functionality out of the box + chapter were reviewed add the remarks to the guide +Thanks +push to +The warning is more for information gathering are we frequently abandoning and killing processes? I see no reason not to commit it permanentlyReducing the thread count is an experiment to see if our more strenuous integration tests pass more consistently when run with less concurrency +Adds license to the test case +Insert M rows in sequential key order then multiget out random blocks of k sequential keys and average and +Well need to either have instructions for building it or start to publish to snapshots repo +issue remains open upon further analysis +In a few rare cases the files size is as large as k +IT added +Also at least the way it currently works it relies on the full URL to know where it is +Given that the causes of the two conditions might well be entirely different +Because when it handles the events in its handle function it only dispatches of kind even for items which have been removed due to a property change +It doesnt work for me +This makes sense Im going to push this change without the increment can you see if this still fixes the issueAlso do you have a test reproducing this that we can add to our suite +committed to SVN trunk and branch +Is this for trunk? No Fixed Version is specified +r provides a temporary workaround for this but Im leaving it open so we can remove the wrokaround code once has been fixed +Reproduced on revision +Currently we scan all +I just committed this +The project has nothing to do with Xerces +Yes you must chose one but Hibernate should not force one on you not sure why that would be needed +Right? If so I will create such a test case and fix it +Ive also upgraded to a newer version of the UI widget in Archiva +Once its roughly finished we can improve code quality or optimize algorithm later +It is tricky +Fixed by applying the by Jonathan Costers in revision . +Patch v rebased against head +I guess Id prefer to add a new method rather than make implement IteratorThen wed make implement Iterable +Ah i missed your comment I tried also and works fine for me thanks a lot as usual Marek +How confident are you about the UTF code? Do we dare commit for? Or leave it only +The file to be imported +We should test if this still happens with. +I seeRunning Tests run Failures Errors Skipped Time elapsed secTry my project again with mvn clean test X and attach the log file +rename endpoint methods in ARS +wiki access isnt enabled by default for these pages +Getting Started manual translated into Brazilian Portuguese DITA files tar zipped +In current implementation reuses async RPC call back function +the patch just adds one new test +Can you submit a patch to improve the error messageLin +Added to portlet container component +We need to come up with a way of handling that +buildivylibcommongriddevbin +Adding as it will complement backport of +With this change I was able to get test to pass +on our systems happens after a while nobody has logged in continuum +Looks great I think its ready to commit +whenif for eap to fix Release Notes fields +Catalina provided a patch to me to check out +Yes the timer is calculated when entered but the variable used to define the expiration date is resolved when the process starts +This was verified using a simple isolated unit test +Are you saying that none of the poms are valid +The web plugin is restricted to static analysis and has no capabilities for dynamic analysis +there is no requirements about name just suggestions +I wasnt aware that the Eclipse plugin also required maven installed on the local machine +The same works pretty well on Linux Leonid Im sorry but I cant help you too much on this subject without more information +Jason could you please explain in some more detail your use caseThanks Jason +Comments says queries which has windowing in those cases we can do this +Seems these tests just havent been run for a whileI thought this would make reviewingrefreshing patches easier as well +I suggest you revert the patch if you want until further discussed as Ill be in vacation for the next weeks and wont be able to actually do anything nor discuss that furtherAnother option might be to duplicate the loop after having invalidated the registration to get rid of possible new references acquired while processing the or just before the service registration has been invalidated +Committed +The problem did not occur in my testing as I was using the default STS VM server which was not showing the you are saying you only support Tomcat then that seems to be fine +We need to improve this for CR. +I added a tooltip to the browser +I looked at the sample code however the JPL WMS always returns from here +There we talk about performance and then we do this regardsMartin +Verified in +Ill go through the list once more on Monday and probably make some minor edits but Im attaching it now so that it is available for review +Why did you add that the class should not be used outside of Lucene? Also I dont think we need to reference this issue in the javadocs +Of course useless here +Probably broken by +orgjbossweld +Tried a few variations of mapping but cant reproduce the failures +Any further improvement in debugging would be tracked with separate close for the release + looks good please commit once out qa is submitted +What do you want to do with this Jira? As its not implemented its not really a bug +We find this problem when using version It is serious to usIf not go into I could accpet +In addition to removing I changed the main function of and to throw Exception instead of +I was working on issues and so that the network of pure masterslave pair topology can be reliable for use in a production environment +Thanks Stack for the reviewThe test failure is not due to this patch which will be fixed with +Some refactoring is required in writing unit tests but reserving that for another patch as we would like to get this into Hadoop +You seem to be reading a lot into my simple Hello World! example +I am loadng the complete test case +marked deviation from standard work on the issue needs to figure out what the rightbehavior is and if it is indeed a standard issue +I also have to figure out how it maps into the other reloading strategy we are looking into +I take this comment as to mean that the javadoc should explicitly define what kind of object needs to be transmitted as part of this +Sounds good thanksNow is there a safe way to get the current source encoding at runtime? Can I access the instance that was used to compileexecute my script +Im a little behind to get this into as is already out +Patch against Spring to fix Hibernate typedef for spotting these fixed for Spring which is scheduled for release +For example instead of +The stack message goes away but we staill cant find the emory backport class + +Since the behavior looked the same I would be in favor of working it under this one +Fortunately Ive avoided most of the vagaries of WFS so far +I dont think it is a configuration problem but it could be +Committed in trunk +im not able to reproduce you please take a screenshot which shows the context menu the view editor and perspective activate at the time this is occurring and optimally describe how to reproduce +I have reviewed the patchI have verified that there is no regression introduced in and suits +If the configuration is not in an artifact released builds can start failing because of a rules change +Done +Hi im trying to get started with reading the on Rails +Diff of change required to make the controls test infrastructure more modular +introduces a new property that defines the used version of JettyAll dependencies were using while was using +I suspect we need unique patches for trunk and branches +Please attach a quickstart and we will investigateThanks +Attached an initial patch +binnutch webtable +Synchronizing the creation of PERVM aspects solved the problem. +Thanks Dmytro and Patrick! +If a snap is taken of a file which has append pipeline setup inode is of type underconstruction in the NN +Im definitely if default tries to send a response +use the Rhino engine to interpret the filter rule +Here is the updated patch +Derby receives LOB information sent via Layer B streaming +Devaraj this is close thanksHere are some comments I dont think we want decomunhealthy nodes in just yet +If you want something more I think you could just have a shell script or background java thread that scps the kafka files without needing to inject it into the actual log roll logic +Problem seems to be in constructor of Console where we check for the existence of the on trunk and branch +We need to look at this within the next week and half to resolve +There is no permission check for fsck currently +D T and F are now all implemented. +Now we have parallel keyword defaultparallel and to determine the number of reducers +This allows you to enter these states more than once if anything goes wrong during init and start the system is left in an indeterminate state you cant stop without starting yet init can acquire resources that need to be explicitly released and which you may want to release without starting the service. +Will need something like this in TRUNK +I missed the last line of course this controller will always throw an exception +Mahadev is going to take the stack size on the Listener threads down to k which should take the pressure offReducing stack size may relieve the problem temporally but will not solve the problem completelyIt seems to me that the problem is due to the fact that a thread is created per RPC connection +Fixed in for the false alarm and thank you for the information +Issue closed with release. +Therefore the question whether the parser has a good regex feature is irrelevantCan you give an example use case that cant be implemented using the existing functions +Closing. +If the failing tests were time +I like the idea +This change seems to have happened between and of Geoserver +Preliminary patch that caps the maximum number of incomplete map tasks for each job in cluster overload calculation +Noted as incompatible in +I thought you want to make changes in the apache logging library +Target on treeItem then can override the value on a node basisThis is what Ill do with this bug +But others may disagreeI was questioning the use of because there isnt any delay +New patch added a unit test case as suggested +on the patch +Can you also please check on the mode and user permissions of the keytab file before initiating the app authentication? It would be a key intermediate step to read and impose the necessary permissions on the keytab before the startupinitialization of the kerberos methodsI would avoid setting permissions because keytab file could be owned by root and readable by a user in a specific group in some setups +Thats why it was added +Small update to html of shortcut list +But since the result is a double and some later code that does a jvm cast expects a float +So its not only me Ive been looking for a bigger logo for codehaus but havent found a nice one yet +has the same problem and fixes part of it +txt files in the full snowball distribution so I think it would be an easy improvement to the snowball pkg to make these available somehow +So I put the classes back the way they were +I cant see any thing not proper in the code +I can fix it and upload new attachements +If you dont intend to keep working as you develop HDFS then I think it would be best to remove Avro RPC support from Hadoop altogether +Please reopen if required or let me know if docs work is required after allRebecca +The solution is to define another queue for the tasks considered as low priority so that they dont fill the execution queueThen depending on the executors load poll this queue for additional work itemsThe secondary queue will only be used as needed and the load is configurable via the system property This property is meant to be used as a percent +Verified in CR +Looks like this is committed in some branches but still open +Maybe another question before we dive too deep into the discussion about the solution Do you general agree that its worthwhile to have those REST available in the web console +For instance the is great but all it gives you is a catalog no way to run requests against it etc +Parking the patch for here +Ive proposed another patch for +Code is completedSetcommand and is modified to accept valid attributes +No logs also help +Fixes the test +Removed the old zip file +Same for compactionsSplit still has to happen at the region level because it is the region that embodies the concept of row range +These procedures just for use by the network server and client rightSo theres no requirement to test them on JME since the client and server are not supported thereSo would a simpler approch just be not running these testsThe patch worries me a little because private methods that fetched connections specific to that class have now been moved intoa public method meaning anyone can call it +Actually I thought we had even already applied it and I was wondering why my console wasnt working that way alreadySo its just that we havent come to applying it still looks good to me with all the updates Ted did its just we need to apply for not having done so so far hopefully well get to it for the reminder +Committed on trunk r +I moved the supplied tests into and +Finally it is still somewhat unclear as to what is going on +Eugene Ive already attached two sample projects to demonstrate this bug +Alejandro Owen thanks for taking a look +Hi JakeI think we can close this bug now as you have found the problem in your schema +We can resolve this +Please tell me more or share a URL related to tomee eclipselink javaagent +The basic idea is to use and as the abstractions for interaction between an and Server +The name parameter takes an relative path to the bundle root and not an absolute file pathThey assume that the framework unpacks the bundle and puts it at some location in the filesystem +Thank you Mayank. +Can you mark this as In Progress +erm ok then I suggest you try +This is because our incremental publisher only copies changed files +bulk close of resolved issues from releases prior to +Yes I do want most entries in the filter cache to be usable for autowarming +remove etclibvertstorage when reinstall cs. +To manage a database the user must provided the needed credentials for the DB Owner before the for the given database becomes available in the server +Or maybe if the new jBPM project wizard etc +But the variables sizand size used in this function are integers and size is written into the PDFoutput stream +The main reason for moving these to Shell was to reduce the number of places where OS specific forks happen in code and limit all such behavior to the Shell object that mainly performs OS dependent tasks that cannot be done in Java +However it does not support return values of type +If youre going to resolve this in a new issue I think this one is ready to go +Updated Affected Release version to as is not release the project under ASL v +Ill pass this bug along to the JVM guys +Yep on the same page +Im sure it can be used just not if we should use it hereIm saying that the genserver stores exactly what you were storing in the application environment and that config change trigger the genserver to do the reprocessing of the configuration settings +Ive submitted the patch so precommit should pick up any other issues +No fault found +I still want to know the number of first first level spills in addition to all other numbers +NET applicationsUsing reflections could work here +Should this be reconsidered please contact the EAP PM team. +Heres a patch which fixes this +Thanks. +I will check with Owen +thank you Julien for your efforts and quick reactivity +But contacting the ftpserver folks is surely a good idea +We do but theres a glitch in it somehow where often the stuff that ends up in the TP! the stuff in the composite reqs mirrorRun thiscd trunkbuildmvn clean install get and updatedant jbds get updated +The last block original has bytes +The queues have no size limit in my case and yes Ill end up with lets say to maybeSure I can use timestamps for those things and maybe thats the solutionHowever give me the first ten entries is usually listing the first added nodes whereas I would have to do a query with a timestamp propertyI just realized that this is about explicitly specify the order so in my use cases I dont have to explicitly specify the order but use insert order +If there is a receiving sequence specified in the send mediator in the inSequence the functionality violates the above semanticI am OK with the idea of receiving sequence in general but for applying this patch +Please review and update as you see fit cheers +Let me do that +Which perhaps is what youre plan is anywayNote that EJB HA! load balanced EJBRemoting calls +building on the work done by +Adding the patch having LRU Cache support +patch looks good to on and trunk +ChristianI think the soap action should be put into the message as it is per messageWe need to map the soap action from the jms properties and http headers to camel message header just like mapping a protocol header into a camel common header +Should I try to change this or will you + failure on SLA is unrelated +I changed the file which is properly encoded in but with the preprocessor of XML has the encoding attribute set to ve attached the patch in DIFF format for evaluation +IMO apache policy requires distributing a source archive containing all the source and the means to build the project as the primary component of a release +is a cleaner patch for the latest SVN trunk +the task will be split into task for particular components review +Im also working on a regression test to verify that shutdown handles the missing permission gracefully +In single node cluster many times +and can both lead to starvation +Use a SQL query for closing stale resolved issues +Got it! Its a change in where a call to the deprecated bean method was replaced with bean but the latter initializes to check their type which in turn leads to creation of too many beans too soon during startup +that was an oversight +Note Ive used queues +Beside this class has been LOG less so far so did not want to do thatIn any case would do if you insistbq +But it may not be hard +No SLD cannot affect the scale line +Eventually +bat file and did not see the when clicking on an exampleI didnt either when I checked the bat files a few weeks ago when I improved them before the CR releaseI think the CR claim is bogus that it happened on +EriksIf you have a chance to look at this issue +But the big problem with NLP software is not the code but the models do you have a good source of free models +It contains compiled class and extracted from embeddable directory +Just updated pagerank +Sorry to have wasted your time here Rob +it fixed in tagsJBOSSWEBGACP +Fix for problem when no was supplied but a bunch of cores were specified as loadfalse +tests included +Thanks Owen! +I agree that the should return an initialized pool based on the properties set and I dont understand why it was not implemented to do that +Id like to remove hql from ll remove it after edward + committed +bin karafroot featuresaddUrl mvnprufeaturesxmlfeatures karafroot featuresinstall karafroot osgilist karafroot log karafroot featuresinstall karafroot osgilist karafroot log customer is expecting their code to work with the M release scheduled for Friday so please give this issue a high priority +Ported from trunk to branch at subversion revision +I see two issues that need to be addressed in this patch if you use git use the current patch is rooted in b directory when applied most of the tests fail +I think John made a few fixes related to this recently +this is not going to be fixed if nobody provides the correct pom +Please ignore the last comment +Yes I know that the SQL for updating does not do anything to update +What triggers it is the refreshPackages call in the Activator sample code above +to fixingAFAICT from reading the support for ab +Since Lucene is now also Java we can just change the code in and the in is no longer needed right +Thanks Dave for the directions. +Thanks for the patch Kevin! +Patch application failed because this JIRA was filed under HDFS though all the changes are in Common +I dont see any consensus on for a solution but if it was thru nested tags then theres still benefit having this since its far simpler to use an attribute than nested tags. +Committed changes to branch on revision +I think the issue has been solved when refining the classloader used when running integration testsPlease reopen if needed. +It involved upping the portlet api sources to draft rev of the specification +I myself was genuinely surprised to discover that they werent grateful I didnt log them as separate defects! +However my workaround works okay +Posted as case to reproduce is attached to this page scroll page down then click relative text abc input jumps upper by the reproduces because of Opera bug moving to future version as no solution exists +Bouncing to Bob to close issue if hes satisfied w this fix in M +controller chanel invalidation in JNDI tree when shutting down the host application. +It seems that the codecompletion couldnt propose Interfaces +binhadoop namenode format command only remove the dfsname directory no the dfsdata directory +Bulk close for Solr +Modified patch hopefully QA will accept ChrisIve removed the author statement and have subsequently found the so I am in the process of testing it locally to confirm the patch will be accepted before I move this JIRA into Patch Available again +Unable to Finish importing mavenized project example +I try virtual users the same resultI have upload my processors in the attachments +This is the jira requesting the Hibernate fix to EAP +Simple patch to at least add basic checking to it only checks for missing description in +patch available would you please try it? Thanks a lot +Thanks for the suggestion and tests I will commit your changes for readBytesFixed soon +gtltetc +Since the is not available externally yet + +Yup Im fully expecting that this will be quite intrusive change but nevertheless very necessary +But we need a way to ensure that the bits that the Store depends upon are separated somehow from the bits that are needed solely for the higher SQL layer +The final merge sort could be merged into normal compaction operations for optimal performance +Any updates on this ticket? Has this already been resolved +For most of these new methods would just sendrequest the same information tofromShould I change my approach? I havent written any code yet +As the class is called how about or had a look +patch for we call jbossasconfigfile jbossconfigfile to not be as specific +bq +Im attaching a patch to thatshows one way to add the styles to the Documents +You correctly describe the behavior is there a concrete problem that it solves? In general you dont want to write code with fixed seeds +I think this will affect users of Hadoop +patch includes this in HHM HHM never deletes the row from Application tables there is no way to distinguish that from hinted tombstones! instead rely on cleanup compactions to remove data that doesnt belong on this nodeHow would you suggest clarifying that +Hi +Double checked the required fix was already included in the SP release so no further merge required. +Heres the spec I added it should handle version string directly do foo artifactns do ns abc end end +I have this problem on with a dynamic service +StackI dont really know exactly where the big improvement comes from compared to the old code but I dont think that it is that weird that we see a big improvement even though the x factor might not be totally true +This problem ought never to arise +Please feel free to browse the code I have provided +The headline changes are Observers now commit proposals to disk Observers can issue proposals Clients can connect to observers as though they were followersI have moved around a lot of code as follows Observers are now in their own class and along with Followers are a subtype of a new class Peer +Not sure which version I should put +Committed revision +He thinks this might be a weld RI bug +Havent addressed any of the comments yet so I could get this out firstI added a unit test which tests migration on a tarballed rootdir +It make hbck tool and Merge tool easy to use +Now that work is well underway with AS all previous community releases are +Nope I missed that one +Now federated features are just pure components which can be plugged along with other regular components like etcThis way it would be very easy to override the core federated functionalityRenamed the Federated components to Doing url encoding for the request params in +do you think this is ok with you +closing since fixes are in released code +I took the approach of extending the instead of the +Hi Jacques AdrianIm near to complete the migration of all the backend application to the new CSS styles and I would like to know ideasopinion to extend this job to the ecommerce application but seems to me that its not possibleIn this case I will leave like it is nowThanksIts up to you I guess its no a and JacquesUpdating the eCommerce style sheet is a worthwhile endeavor +For now I think we can implement a that sends objects to the destination peer specifiedAlso in the spilling queue implementation the combiners are one of the that works on the data in the spilled +We also discussed about splitting that up but it would create too much dependenciesIf EE control is fundamentally different then we should introduce a new backend for which just throws for the boot stuff +AndreasIm thinking of contrib as stuff that people want to contribute but the core team isnt interested in or able to support on the same level as the main codebase +This path can include any third party library not only a jndi implementation +Yeah string would be fine too +It looks fine for me +See for further details. +Rebased +Status +Turned out to be easier than I had thought. +There are several facet settings that the runtime does not handle +Ive included a patch file which works around the bug +Great that worked and I can confirm the error is fixed +Although the thumbnail generated by JIRA in Image Attachments seems not working it becomes all black the actual file seems fine +Closing all issues. +thats dependent on the archiva project features. +Thanks for looking into that comment yes its fine to go too . +the code to do this move is there and it works in a unit test but it does not seem to be working on the roller +committed on trunk and branch +What happens when you try to run the service? Do you get an errorAre you using the sc command to create start and stop the service? Or are you working with the Services control panelWhen I look closely at the command line that you provided I can see that there is an extra space characterin the path names that is you entered capachederby and you probably do NOT want that space afterthe word derbyIf you do actually have that space character in your paths then you will need to enclose the paths in quotes so thatthe argument parser doesnt break the argument at the space characterHave you also checked all the other path names very carefully on your system? That is do dir CDBDBDIMBSoftwareJavajrebinserverand dir CSoftwaresbinand dir CApacheliband let us know what the results are +Thanks Carl +Fix merged to branch at revision I am not setting this to resolve yet since I think a test case for this should be added to the test suite +would be interested in opinion on what would be best approach here and maybe a scoping from someone who understands thecompile layer + Weve tested this on EC clusters and it works +I was distracted by something elseFor CN image checksum validate was done when loading the image from disk to memory +However if you take the distro you wont have this problem +I think you can swizzle it to XHR the content instead bypassing the need for an actual page reloadEven in the case of the content instead of traversing to a new page if youre updating location at all not hard to imagine that might trigger the page bits of Cordova +slip to needs a higher is the only good place we can use to guarantee validating what users are actually running +bq +I assumed that since code affected from does some parsing from the PDF the opposite could achieve what is requested here +So master could select regions as size or other conditions to merge automaticallyIts nice +Therere two text nodes because when you disable a menubar item it gets a white drop shadow +Should fix JIRA +jar from dist folder but this workaround is working on windows +Looks good to me and the tests run cleanly +Please reopen if you still experience this with the recent builds +Actually this does not work on systems where memory is protected such as OS +Applied patch after hacking in fix for tests and looping stopped +Fixed in +I am still looking into why the resource is not found for created job +Passed options into method so that correct is sure to be used +Thanks for the pickup on this issue +Attaching a revised +Unit tests confirmed this is already working +Another option that we could think about is to use Hadoop Metrics for this sort of detailed information rather than introducing a bunch of public apis +Changed to file based solution +samples +Attachment has been added with description font autodetection fix +this limits debugging fails to pmc members +log files for the e computation +Thanks +Your profiling helps a lot + annotation on pointed to rather than +In our case vpc id should have Netscaler as the provider for LB since we have created VPC with Default Offering with NS +Bulk move of M to to m since no updates done by because of no update +em which causes exception +However a client can make an admin connection similar to the way JOPR connects to verify readinessAs for in SOA in ER the example querymodeshapeusingteiid describes how to execute a query which in turn verifies the accessibility of +Its good idea to avoid saving recoverable states without storing +Html tags and attribute names and values were changed into lowercase null value check was added to uiparams value replacement +I happen to like the way works +key issue not addressed +I also changed sstablecompression from to +OK I think all the stuff in this issue has been resolved. +Updated test to avoid deprecated version of cluster method +Hi Alfonso can you please review and comment on whither the issues are the same or linked? My initial oberservation is that they are the same issue +So I think the fix needs to be that while were doing the tailing we continue to treat everything as if it were still in standby state from the block management perspective +Ah interesting +Thanks +Do you mind working on an unit test as well +Thanks +Defer to defer open issues to defer to not essential to address for +may also like the class I just committed for use in updating tests +Im using Fedora JBDS +Also includes a socket stop just to demonstrate that the use case works elsewhere +Still no access to the right +This also makes it so that the process group id is different +Closing this issue because the test was disabled on the offending platform and has expressed any interest in this issue in years. +I always get the following as the response in request channel +Removed the redundant that was being appended before to correct this +Can you attach wsdl and soap response and sample code to recreate the problem +Hi NicoThanks for the suggestion +I dont think anything is being changed in how things are being stored in ZK +This means I cannot use the CVS diff or Tortoise +I dont think IDL belongs in resources because theres no reason for it to end up in the compiled jars +Natalya please confirm that this resolves the issue +I dont know how this happened since I never touched the preferencesRestoring the defaults solved the issue +Thats consistent with the archetypegenerate command +have you tried release +myDirmyFileFor now Ill mark this invalid please close the bug if this works and you agree +Proper javadoc on set mentioning PAGE scope issue would do in my opinion +I understand what you mean +mistake as an array is a reference type variable it doesnt require the use of the ref modifier. +images +Nick can you please give me some more details on this? Im not sure what declarative models are +The initial spaces issue is fixed +r +Im testing this + results in the desired traversal by not continuing to the nextSibling of the root of the original subtree + is the first revision of the classThere is no code showing how it will be used yet but I plan to pass it to so that it can properly configure itselfPatch ready for reviewWhat you people think of the builder pattern usedThe reasons I tried it are to avoid a constructor with a load of arguments avoid having a set of many constructors and a wish to keep the descriptor object immutable +is it possible to get a fix that would allow the all type +Now instead of knowing that unflusheddrained data is lost with durablewrites off I have to understand that I lose my data if my node crashes and keep my data if I manually shut it down +I retested output and it now passes so Im verifying the fix. +The returned environment is a Map type so its not Serializable. +The tests are commented to indicate how they should be modified to behave once this is fixed +A few people have independently invented mechanisms to make this thread local that gives massive speedups in their clientsMultiplexing multiple on the same socket is in other words not a clear gain when the RPC request andor response may be significantly large +Can you try the latest? I just pushed out a fix that is making your project compile for me +Let me fix that +have solved +Flagging the test as flaky may be premature +Provided fixed OVF +If I convert the groovy class to java I dont see the glitchy behavior +Hi Vlado I try to have a look to your problem +Yes at the scene of this issue we can do as you said make the default package as definitionstargetNamespace +Replacement for the originally submitted patch + the patch looks good to me +IE removes adjacent spaces on AJAX update so even if response contains them they are trimmed +Maybe that is only in trunk +We had added the transaction to deal with repeated restarts and also repeated rolls +BTW I have updated the attached exe files +Must be some kind of spontaneous self healing cant reproduce on upstream master +Udai if you have checked the BNF against Castor source +revision all old Resolved issues +Im thinking along the lines of what was done for Snappy and fuse in +We just ALWAYS need to remember that something in hibernate does not work +Provided a trivial patch to fix the problem +In this case I vote for accepting Brians patch for the branch and inclusion in the releaseFor trunk the should be added to +The parser is generated by Modello +batch transition to closed remaining resolved bugs +I think we should set the default value for to be seconds in the itself so that it serves as a source of documentation for the default valueAnother point that may be the content of another JIRA The issue though not directly related to this patch but that gets a bit amplified with this patch is the fact that we have a separate configuration for +Now rolling out to customers siteIs the improvement simple so I could patch the RC or +and the fact that they are being individually deployed whereas a real plugin would usually only have one CAR file to deploy and it would automatically pull in the prereqs it needsI think this contribution needs major rework for Geronimo to make it a true Geronimo Plugin which is built with the use the existing dojo CAR in the Tomcat and Jetty JEE assemblies +With this fix the ant and maven builds of the travel sample complete successfully on both JDK and JDK Im still seeing some runtime failures from the travel sample on JDK caused by missing dependencies + +Let me know if you need help +No mention of on test page +The README file contains a Prerequisites section that states the followingIMPORTANT This quickstart depends on the deployment of the quickstart for its test +One host is so complicated to crawl due to the many excludes we have configured so we decided to isolate it from the rest +but a stale object exception on itself doesnt mean it is always a you have more information please reopen this dont know that I am ready to reopen this as I cant really provide additional information +What is going on with this issue? Can we get a fix version set please +Popup with input suggest closing the popup input suggest list remains ve added an event listener to close the input on a click outside of the control +just fix by the easiest way +We would be able to reuse the ip in the fetcher also we can try catch the parts in the fetcher and in case the ip is not available we can re lookup the ip +Here is the description for Used when target file already exists for any operation and is not configured to be overwritten +Personally I dont like filters interface alteration +So people can switch gradually +This is pretty old +I will not be checking email or phone messages during this timePlease contact Bryan Kearney or for issues related to Subscription Management tooling +Just kidding but we want to map that as Timestamp and not as date and in the above clause we cast to date for a today filter +This is a very annoying bug +my system is Linux openSUSE milestone x +We used to do this looks like this was a regression introduced in +scary stuff +Why not just get locus out without and we can start looking at moving to yet another tycho version in less busy times? this way we start getting something out of the nexus approach which can save us even more time in the long run. +This updates the release notes transformer to use the new version of the eclipse doc plugin rather than the old version number +BlockerCloudsatck UCSDB Discovered UCS Blades have multiple entries in Database Critical UCS Session cookie refresh must be supported +Eclipse issue so i changed fix version on latter. +This is already a plexus objject +AXISPRINTFFORMATINT +AppliedSending srcizpackTransmitting file dataCommitted revision +Hudson is a Heisenbug generator +This isnt strictly necessary for since limits can be set on the components themselves +Ill take a look at this patch today night +Moved a new test wsdl resource away form the pathupdated and to ensure that this new wsdl is indeed picked up and thus policyfeature instances do indeed contribute their policies to individual applied! Thanks! +I manually ran sudo usrlibcloudcommonscriptsvmsystemvm varlibcloudmanagement +That is even though recordreaders might not have the logic to determine whether a particular record is corrupt the framework can do it with some help from the reader +resolving this +Assume closed as resolved and released +aslak assigning to you +Attaching a patch which includes code change and a test for this issue +A few more details on our setup Running on Java Backend Storage is Percona you tried to create a test case to reproduce +Also theres no template job for stable only for trunk. +See +I dont think it is a big deal to add all inserts to the deferred check list when you are in deferred mode but recognize others may not agreeI consider any code that eliminates the inserts from the list early just an optimization but any problem withthe optimization can easily lead to a corrupt table if it misses just one case it should have checked later +So we can add a unit test to verify the integrity of the committedLog even though it should behave correctly at the moment +Similarly compare the clients with +please check again in fixes the behaviour but it looks like the tests from the first patch for the three argument case and from the second patch are still missing +Hi Guysabout I dont like the idea to add an abstraction library it would add useless complexity and compromise performancesFurthermore we would be conditioned to wait the update of the abstraction library to access any evolution of the underlying librariesAbout patches Ive improved the logic as suggested by the submitted patch +Moving it out of release +Happy WeekendThanks RegardsAnkit Jain +Implemented the comments Ran the fix on node cluster +I think it is best to keep them separate after all and handle just setting the CHKCONFIG +Im ready for migration and can give this a to see things goingRegardsPrzemek +In the comment for setPreserve Id recommend explicitly saying that the only attributes which the option will attempt to preserver are modtime and atime +The patch for the issue and test could you please mark this issue as Patch available +Please help us since our functionality is completely dependent on you post the exact version of Spring and the Jdk you are using +I also converted the tabs to spaces +Linking to a related issue where cgen is getting confused about inheritance +They should be fixed in +Yes but the validation code didnt work because return false when comparing the same group if one of the two instance has changed some property +Dont hesitate to ping for review much earlier than thatSorry for the problem I had been focused on a few other things anywayCommitted. +We werent applying the recursively in our migration script. +Upping priority +Hi danielshwould it be possible to deploy the site via scp towwwanyThis is something we already do also in the amber communityTIA +How come this isnt fixed in the unreleased? At least according to JIRA +this will be available for the next release will be supplying a patch for trunk shortly with the changes to the James I appreciate you trying to help but we can not have a plugin published like this + +Case adjust the classpaths and ensure that the classes are found +Marking as wontfix in accordance with Brians comment. +something +verified as fixed +It seems like embedded maven used when updating source folders have different classloader structure and it does not see xerces required by jaxb +Has this been committed +Both are coming from other threads +Untested with Chrome +Thanks Giri! +Hi ChrisIf you are just moving code I think its fine at max we will need to rename the paths in the patch to match the new file structure +I know some of them are already included inside jboss but normally i need a seperate classloader for each application +Run attached to see caching in action. +Not it is mentioned in comment above. +The code is so broken in the C backend that we need to deal with that first +Hi ForrestThat definitely fixed the issues +Youll have to provide a more detailed explanation of what exactly is happening +While using the data base throwing new message may give wrong information to the user claiming that he was using primary key when actually he was using unique constraint +Metrics have changed there in a way that is not similar to TRUNK +I have created a separate jira for this and have attached a patch +The main benefit is cachingCaching has nothing to do with whether files are immutable +Thanks +One cool suggestion that I believe I got from someone else is the idea of using a simple vs +Issue is fixed in revision Setters were marked as deprecated. +I got the same message if i try to configure one server config or add a new one +I can no longer detect any circular dependency +Rather than committing megabytes of licensed binary code we could also use an embedded ant task that gets downloads them into the weaverBundles directory if they arent there yetHope this makes senseI think that only ASF committers are allowed to upload files to +bq +Fixed problem by using a in instead of a for gathering super types +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Was about time to close this issue +The problem is introduced by merging from trunk which pulls in guava dependency in hdfs and forgot to update ivy since trunk has proper dependency in which would suffice for trunk mapreduce +Closing all issues fixed previously after a brief review of each. +Thank you for your answer +Rayees is this due to config issue +Committed revision Renamed the private method to normalizePath instead of normalizeSeparators in the patchThanks Koji and Hoss +Replace the old BPEL WSDL and files with the current example files +I added serialization compatibility test and also fix a minor bug of which failed on RI before +to if the problem does not exist in +bq +This patch needs some modifications since it give compilation error and some other build problems with the modified patch +Im resolving this issue now. +Failed test cases from Ari +reopening see last s fix it in CDI Tools model +Patch to allow configuration of bind address +Hi Kamal Im currently working on a design for this +I tried the fix in a test build and its good GA msgssec CPCR msgssec CPCR msgssec +I noticed you marked this as Affects +Looks good to me +Is there anyway around this? We cant have this in the distribution +description is correct because the error message has typo +bulk close of all resolved issues. +I think we can call this a dupe of +Verified in GA to fix version. +Mostly looking fine +OK +Ok very curious to see how you fixed it. +Lets just do it on with the formatoptions +As explained above +Externalizable and Serializable work interchangeably a nice feature of Java +Many thanks for the patches. +Hi KarlI tried using the dechromed content and description but the simple history reports for the RSS feed links each time +Currently in we force a replication every time no matter what when we call replicate now that std replication has had some bugs fixed and has better tests we may not have to force that anymore and so the next full replication would not actually move any files +The underlying blocking calls the append and appendBatch calls in the Thrift client call the sendBase method in thrift which increments the seq id of the message +yes we use maven and generate eclipse projects using eclipse maven do not use eclipse like meclipse +Thanks for reopening +Yeah I considered the tx reaper but what about if the timeout is +Fix committed to SVN trunk and branch +Lets not back it out again +Thanks for the report Ive fixed this by the systemId before turning it into a URL +I would have thought a Unit test would work here checking that we can put more than the prefetch limit of messages in to the Consumers Synchronous queue +I think with the latest code this pop message should only show up if an AVT is used with an incorrect value +DONE assume that doesnt need to be documented are documented. +The action for a node is already configurable in the properties view in but is not yet visible in the content outline +If we can get basic support for EPP for JBDS that is great but we should mark it as a preview or something +Problem found and solved +defer all X issues to +Can you try it out +Imports are processed before any beans get instantiated even before such as +Rewriting using Socket factories +Heres a script to do the appropriate moves and removes in SVN +While the descriptions in the patch are strictly accurate I dont know that we should be mentioning the option of using for the NN HTTPS port +With IE too +It works best against +Notice the from the hadoop ve encountered this issue myself +I think this should wrap it up so closing issue. +MarkIf your concern is about installing LCF read the Quick Start part of the builddeploy page +I will check this and confirm tonight +I think I caught all the socket creation places +That forces the new logic to be excercisedWith the new logic I have successfully run and on Mac OSX and on Linux on top of on top of Mac OSXTouches the following filesM javaengineorgapachederbyjdbcM javatestingorgapachederbyTestingfunctionTeststestsstore +minor nit can you add the test case to the patch as well +If I used to store typestockSingle I can not find the document using query typestockSingle AND test If I use to store typestockSingle I can find the documentThe Lucene document says the Keyword is index why can not I search it +Examples dont run without this +resetting P open bugs to P pending further any progress on finalizing this patch? +But will be using it today and tomorrow so will know pretty quicklyve refactored the code to use Shiros object +Added a patched so that you can add as artifact type +occurs when building Harmony with free tools. +Could you please give it a spin? If you could possibly come up with a patch containing a test for all of this Id be exceedingly gratefulI dont know whether to think in terms of deprecating the plain JAXB factory leaving it alone or giving it more interceptorsbenson +But not all accesses to blocksMap is under lock +Will come back around for that +Please confirm that the problem exists against the nightly buildIf the problem persists please also provide a patch that implements theproposed solution +Marking as NR. +The concepts apply regardless of language +Perfect! Ill try with m deploying on AS you know anyMatas. +I dont have a Mac OS to try this but I believe the dependency on should now just be there for Suns implementationsFeel free to reopen if you still see problems with dependency +Please pay attention in the future +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first. +Have a working version of it passes the basic slow consumer tests +Will commit it if test pass +I will check and report backHowever due Eclipse dropping the UML tools featureplugin I still need to use I will check the latest Eclipse Windows x release with the latest code and see if the problem still occurs there alsoNot sure when Ill get a chance to do this bit busy at the moment but will try soon +Anyway a fix in the Hibernate would be much betterBest regards Mattin +adds checks for broken entries upon put this is a simple timing issue? The cache gets invalidated based on an observation event right? That would explain why theres a small time window during which the code that initiated the ACE change may not see it +Seems this has been applied so Im closing this bug report +After the build move the release and debug pdb files into and +I wasnt aware of any fix being available to test +Also looking back at branch weve changed behavior and are flushing when we compact +Tested! +DilukaI applied the patch and it seems ok +The top part of documentation pages are being overlaid by the Gora black navigation bar +Is this consistently a problem? Galder could you add our stack trace to the bug on the Oracle site and vote for it? Michal could you upgrade your JDK to the latest JDK? I believe that is +Patch to undeprecate mapred apis +Updated formatting as well +Let me look at bringing over the fileinfo block to pb +Anything against backporting it to +If you yourself instantiate another one inside controller code then it shares the same request as the one instantiated by grails +Fixed in +The failures reported by hudson does not seem to be related to the patch +sorry if this is a blindingly obvious question +To be more precise about the version in question a currently nightly build from a few days ago was used for the WSDLC utility generating the stubs +Will keep posting here until all the kinks are ironed out +Oh sure +This version now has a unit testand generates a and file inside the targetsite dirI will now create the site documentation and then it should be finished +Bulk move to +committed +Ok. +More failures +Proxy server Proxy server port Proxy server user name this is used for HTTP basic authentication Proxy server user password this is used for HTTP basic authenticationlist of host address exclueded from going through the proxy list is comma seperatedlocalhost +I think we could reuse the unit tests Im writing for branch and fix if necessary +I have committed this + +Excellent that would be greatly appreciated! Feel free to assign to yourself and by all means shout if there are any questions that come up as you tackle it +patch for adding these commits toeaadedfeadccf fix clang warnings and errors especially with make checkceacaabeeeeeadbbc fix clang Toell does the work for you +I believe there is an open ticket against Spring framework to fix this in committed with rev dabaccaeeaeeef which explicitly sets the appropriate header via +lets see if i can provide a patch this week +Moving to to be included with an SSH transport +Thanks for looking into this Tom +One rationale is simply that it would be an incompatible change +tar thing +We might seek more guidance from experts on this but I would think that we dont want to automatically forget anything associated with a tx that is completed heuristically +At least thats my interpretation of the javadocthat seems right so far +Aaron this is patch +As you noted there are probably two issues at least one on the derby side and possibly one on the Hibernate side +Fix is committed and will be available in the next build +Patch looks innocuous +So I wont be pressing for the feature if nobody else sees a needGary +If you declare the JPA dependency first it might still pull in the dependency +Maybe Owen has some historical perspective on thisDo we need to put some javadocs to state that if there are multiple groups in the ugi then the first one takes efffect as far as file permissions are concernedAlso this looks like an incompatible change especially because files will now have the appropriate group information +This make the use of sequences pretty useless and slow if the appserver cannot cache themAndreas Beckers What did you unterstand Why is it your fault? It would be helpful for others if you explained a bit more +Re point I tested both Gnu and Posix parsers the relavant behavior is in a shared Parser superclass where the parsing code is littered with inline string literals like and Re Point +no more than x connections per client no more than x connection constructionsper time period +Will be naming this rather than in case we change to some other rss lib in the future +Patch has been slightly updatedIve implemented the support for failproperty in +Moving to +Ive just committed this to trunk andThanks a lot for the contribution Colin and Andrew. +This fix to this problem has been incorporated in svn revision +adjusted patch for review comments +I should clarify in it requires you to individually list each directory +Will make a new JIRA +We tried to workaround this issue Than failed and Subsequently we had to this at th hour to unblock a release +Need to regenerate this patch due to recent changes to trunk +Marking as resolved +Since I have recieved no futher info Im closing the bugFeel free to reopen if you want to provide info. +yes all items are there just the numbers are incorrect hence its a minor issue I guess +I can probably provide a test case if needed though +If youre building from source can you svn up and try againI didnt backport the fix to the branch +is meant to test one specific part of the code the deadlock detection algorithm +I should be able to have a go at this next week +this will be resolved in Railo when we switch to +Not applied to is it is formally an API break +However the map is filled with instances which are created in the same method as where they are put as keys in the map +I will try to make a smaller test case than nstest +upload a patch that reverse the order of unregister and clean staging dirWill do manual test +But there is such a flag as you wrote on the original issue +Heres an example multimodule project using properties as requested by ve tested it and it produces the same package phase works fine so install and deploy should work as the site phase the site for the main project is generated but the links to the modules are broken as told hope this helpsRegards +FrancisCan you work on the trunk patch firstIf it gets ready before RC it can go into +As of r reports bind address and port to the log file using the info level +But it was not deleting these temporary files when finishedThis change removes the file in the existing finally block of the storeValue method +Thank You very much + should also be done for this +It could perhaps be clearer but the docs on say You can enable more than one type of annotation in the same application but you should avoid mixing annotations types in the same interface or class to avoid confusion +Marking as blocker since distcp and other apps that launch jobs cant work under oozie +This includes all modified files Ive submitted for geronimo integration and possibly other jira issues updated to work with j rev +Also includes tests +I ran some profiles that were inconclusive +It should work fine in any thread +Im sorry I havent had the cycles lately to properly review the patches + +My gut feel is that it sounds more complicated than vanilla XP and therefore should be avoided but I fully accept Im probably wrong as youve thought about this a lot more than performance of project level metrics what source repository tag would I need to look at in order see this? Im amazed that hibernate isnt already optimising sensibly and Im happy to help tune lazy loading or caching if that helps. +Please feel free to provide ideas for improvement or even submit an improved patch +Youll need to look through the existing AS documentation to see if this is presently covered anywhere and whether it is covered there related to this is the recent email that Laura sent about the upcoming of the logging subsystem +This bug has been marked as a duplicate of +Committed to trunk +Issue closed +should be applied on top of the original should be put in srctestresources +With these two changes the JUNIT tests now completesI had to hack a bit in the to make it work with multiple good samples testsets +My main point in submitting the report was to point out that in its current state isnt usable +If you wish to provide an example please feel free to this issue. +I didnt commit it myself +Lets close this baby + +This is exactly what we need +Thanks Jack this is a great addition! It is in CVS now. +thanks for the feed back +Here is the patch for reference +Well in any case the wording of the spec taken exactly as it is written leads to this behaviour here even though the intent is very clear and the fix is of course correctThanks to Michael Kurz for sorting this outregardsMartin +The recommended solution is to upgrade to the Sun VM or +Reopening as there are some more I have to add to this issue +No I doubt thatThe issue is probably to do with the special id property +I added deprecated versions ofsome methods to make backwards compatible +Fixed in revision test case is working well now. +Charitha you can close the bug now +Vijay I think this is something youre currently working on +Ive removed the dark gradient which was centered on the desktop that was used to make the white icon pop +what about? Not sure about gradients +My personal preference would be to keep it next to the code where it is used for now until there is a need to refactor it so it can be used elsewhere +What is the problem with looping over the tasks and then getting the variables +Once theres a consensus on how this should be implemented we can move on to patches. +javadocs warnings seem unrelated +Then you can apply +Sure destroy sounds like a better name +Is anyone looking into this +are we actually doing oauth anywhere in shindig yet +Also has this been tried in production +Closed with. +runs all the tests in and reports the increasedecrease in memory footprint for each fixture +Sorry I was not more specific +As always please feel free to make any suggestions or things I need to do which are missing +As expected build fails +Shirt design accepted and attached to the JIRA. +I agree that v is less risky than the selector approach +Thanks for the project Gerhard +I didnt think about the get that you are referring to +It could sort the mutations in memory flushing them out to files as needed +AndreasSorry for the confusion +dont have this problem so the problem is from the native launcher +Closed. +We arent reliant on any of the other arguments to the REFLECTIVECALL being next matter to make this work we need to get method object constants as arguments to +I tested versions going back to the alphas +Preliminary patch mostly done couple of minor loopholes +Yes PREFIXvarlib is meant for HDFS storage +Found the issue +Committed +Other examples of machine learning algorithms for text are Bayesian classifiers and classificationPlease use a name that is scoped to the class of problem these algorithms solve +What do we gain in restricting deployment options as long as its simple enough to keep Solr a cleanWe can do more around out of the box goodness like process management and good logging and have tools and things that know where to find those +Its the right thing to do but then it means I cant apply to and it breaks a talking to a which currently is possible +If the client is currently I can file a JIRA with some ideas +I removed classes Meldware and. +there is no notion of a hibernate runtimeis provided by server runtime Or users own library dependencies +This is pretty major +Yes +Closing jBPM issues resolved in a released version and reported by me. +Any idea on how this could be handled +Sure Im just sorting out the Maven release process under Google Code and aim to release tomorrow +Actually the file permissions restrictions apparently apply to the Network Server but not to the embedded server +Thats a regression introduced by issue +My theory has been confirmed +Ive updated the static css on to hide the authors as well as the repository. +If not for history Id have argued against the Ganglia sinks too fwiw Though worth noting that Ganglia is vastly more commonly used than Chukwa +Table is now and schema SQL should be portable across most databases +Seems like the earlier commit has to reverted in order to apply the patch +Closing a duplicate. +This is going to be documented in +updated to ER in revision +The weird difference in syntax is what made me think this had not been implemented +Thanks for reviewing Suresh +Mixing the two levels of read operations on the same input source is not allowed +putting this back in the general unassigned population so we can start talking about pending refactorings to make this possible +I filed one new JIRA for updating tokens in the RMI started to file a JIRA for the AM to be informed of the location of its already running containers but as I was writing it I realized that it will not give us enough information to be able to reattach to the containers +Knut thanks for taking the time to explain how is converted to problem was with weme jdk +As a consequence the constructor will start a thread on each call while the method will stop the thread and remove the local reference to itIncidentally only two locations inside the Framework project are touched by this change The and methods have to use the mdispatcher field instead of calling the static methodI do not know of any other users of the class but I assume there arent any outside the FrameworkIf nobody opposes I will apply this patch +When composition or component are not facelets tags the whole page content will be shown +revision +Very weird +can you please check? Thanks +This was actually fixed in +Thanks Kim +Fixed Ruebens names jesse this one is so your problem +Ted I hope you dont mind my committing this to branch too +Could you please a try with a recent jbossws version b post the full stacktrace? Native SOAP stack doesnt support JAXWS any more + that works with isolation +Would have saved me timeI see there is still an unresolved issue with timeouts in +I just committed this +To keep English version to use English font while making Japanese version use Japanese font we can use different namespace for English and Japanese and modify forrest stylesheet to select fonts based on the namespace +Its a great start and Im finding it really useful right away +Ive committed this +Awesome troubleshooting! This would make sense in explaining why the problem has been so elusive +This should resolve most issues with dev issue seems to be that the JDKJVM isnt as flexible about openingdealing with devnull as normal libc functions update with comments based on trunk +Fix versions field doesnt include Ode release will this patch included into +Thanks Snjezana Ill give this a go today +RobThanks for the update +For Scott had fixed it by work around we are working on new JCA impl so the old deployer will this ticket the customer need to ensure before +Part of the reason for my ranting is also because the buildr build process is currently not working +Thanks Vinod. +Thank you Knut for working on this issue +Ok sorry about the delayed response so have been swamped with work ran a few examples with the reuters data set and couldnt repro the OOM issue +lol no this was opened nearly two years ago just after I started at Red Hat +Attaching a simple patch for now which addresses the immediate bug related to symlinks with special chars +Ive committed Marks fix on branchx revision Id like to leave this ticket open until we solve it for trunk. +Mike has volunteered to take this. +Okay this is fixed but the adoption of the plugin is needed +The links return +I hadnt seen the with Java only Java +for a class written in a few evenings as a favor knowing that those things are trivial to change +if the type matching of strikes an ambiguity then apply them in the order declared +Oleg your patch has been applied. +Changes look good to me +Thats useful +Ive made a modification to show this property only when Debug mode is activated +I just think its very late in the game to make it for this EAP release +Having spoken to members of the community they have been supportive of being able to test GC ideas within the context of JNODE +Handles shear as well as non affine transforms and uses tolerance in transform type detection +This looks like a duplicate of +The classpath and datapth may be in the pear package but it must be could you please explain what do you mean hereActually I dont believe its a good idea to specify absolute classpathdtapath in PEAR installation descriptor +Committed to TRUNK +This issue is caused by a containing a in Container mode +I wanted to have the same output as in test method in its own line under its suiteMy primary concern is for this to be usable for TDDBut with chunked output it is lot less comprehensible. +Are you checking through eclipse or the standalone shell? I dont think their behavior is the same. +Right +Raghu are you OK with this change it is OK from my +How would the language be defined? Do pass the header? +Make sure to remove the +as r +Also code is in pull to Vitalii +I have an existing database structure created with Kodo JDO +A draft pdf has been attached for QA review +Hey JG take it away +Failed tests are unrelated +The v patch adds inclusion of classes that are required on the client side from the main C build +Updated with the attached html +Checked and understand that Roo cant perform the checking for all the elements that handles in the creation of a like classes that implements or extends for any of its components because it increases the complexity +Fixed in. +The main tar and zip files all have signature files +Is there a workaround for the Command not found loadClass issue any other way to load a Java class in? +I vaguely remember looking at the patch and thought I saw exceptions being thrown etc +Petr could you please list ported tests so I can find them more easily +Patch using findbugs exclude tests addededited this patch fixes excludes findbugs warningsfindbugs warning not relatedrelease audit warning not avail cannot evaluatetestcases failures not related +The default implementation might return for capacity and free and zero for used +Hi JamieCan I have a mirror maintainers contact email address for this domain thanksIf possible can you add to your in your file thanksEverything else looks good so will add it as soon as I get the email addy +Moving out of the queue as these could not make the feature freeze date +Uploading +Thanks +Please verify. +Ive send an email to the list to try to find out if the th value means anything and if its required by the java version Ill await the response on that before either start assuming parts or remove the th entry from the sample ve changed the code to expect parts now so it should work with the samplecontainer again +Ted Youre right +Reopening to resolve as a duplicate. +A patch is available and will be committed soon. +Patch committed to branch and trunk. +It might be good to have the cleanups but leave the little crazy parts out +This first cut has gone into. +Patch to add this functionality to the body text serializer and also better document the two existing stock patch to trunk and! Thank you for the patch Mike! +Connecting objects that were never written to work with Spring or that are written to work not only with Spring should be supported as well +fixed verify on JBT. +Ill get you the revised patches as incrementals as you requestedI dont want to work against the old Thrift +I tested the misconfigured store hypothesis by deleting the store altogether and recreating it +I have extended summary to check that also confirmation by enter key is working properly +Two different way +Patch checked inOleg +I would advise considering this step as your patches are getting biggerSending javaclientorgapachederbyclientamSending javaclientorgapachederbyclientamDeleting javatestingorgapachederbyTestingfunctionTestsmasterAdding javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestssuitesDeleting javatestingorgapachederbyTestingfunctionTeststestsjdbcapiDeleting javatestingorgapachederbyTestingfunctionTeststestsjdbcapiSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiAdding javatestingorgapachederbyTestingfunctionTeststestsjdbcapiAdding javatestingorgapachederbyTestingfunctionTeststestsjdbcapiTransmitting file dataCommitted revision +How can we handle this? +should be very easy to domeanwhile i will review the thriftserver code +The deployment completes successfully if I deploy manually from the admin consoleI am going to close this defect because the issue for which it was created is resolved +From your description it seems less flexible to use compared to +thanks for the reminder Bart! +This was broken again with the new model +It is located in varlibjbossas and in usrsharejbossas dirsThe shortest path to conf directories of server profiles is etcjbossasPROFILEAnyway there is an inconsistency in the instructions from Section +Thanks! +Lucene jars were used by Solr trunk temporarily during the vote of Lucene which is now official and the jars have been switchedDealing with any similar future instances would probably need to be dealt with by filing a Lucene issue to get the release candiate jars published to the maven repository +Elliot how did you test this? This is on distributed setup right? Do you have steps to repro +I am running the patch against scenarios similar to the ones present in the unit tests only +Theres no case where an old client could have been working correctly against an old server but will stop working against a new serverCan you please show a scenario in which this could create an incompatibility between and +If its relavent to the API of the application yes if you want debug output or general tracing then this falls outside the monitor +Can you reproduce from a fresh install? If not I wonder if youve dropped other and are running into +controller automatic name +Why we need changes in +and facing the same issue +I think Ill write a reproducer so it can be debugged +Ive added a patch for this issue +Simply said yes +With that patch in there is one less reason why you cant run Hadoop under a security manager if you so desire +Renamed RSZKREGIONMERGE to RSZKREGIONMERGEDAs to Stacks questions if RS crashes merging regions are indeed assigned by SSH if meta is not updated yet otherwise the new region is assigned and the two merging regions are skipped +Its and make coding more complicated +Tim +I would like to see this for the purpose of generating documentation +No special treatment for the bits is needed in Linux currently because a is created only when the configuration is specified to be used so we should not meet such a +Is there somewhere with more detail +if I understand the point of the method and the way we have used in our env it is to take a string like a user quoted string and encode it into a valid javascript encoded string +bq +This is done in the Not sure if there is a test that will do this check for messages being sentIf I discover that this is required then I will fix it at that time +problem is the session type defined switch to cfml sessions and it will for input +Thanks Carlos! +When some error happens we need to force a close of the connection no matter what +That WARN is some leftover noting actually if you request some page does access log file gets created then +fwiw Im trying some stuff around this currently +Not sure if this is the problem but I think resolving has improved in the latest Felix releases so upgrading to might already make a as part of +Here is an updated patch that removes the read ahead and fixes the same problem of line endings in the get method of +Test caseThe problem is Hibernate didnt report the error if strategy! The repeated problem column is already fixed but it was very hard to find it because of this issue +Moved a couple of methods to a higher heirarchy for common inheritance use +Patch against Wicket that implements the feature +There is now an example for this in the site docs in svn +Waiting for review to Mar +I dont care much what bugs Spring has + to this idea without it cant track topologies that change over timeThis should include failure domain topology time last updatedHaving a timestamp will let the decide whether or not to rebuild their local datastructures when querying the rackActually having a notification that the topology has changed could be even more useful for some applications but that is very much feature creep +Please verify and close. +do you wave the entities? i think glassfish maybe does itabout TIKA stuff you can avoid serializing it using a transient field i close this issue +Ive verified mvn clean install for default and profiles +Will learn more about how to add content to the hbase book and will create a patch and JIRA +Could you specify which private methods in e you needed to copypaste? Perhaps we can look at how to make it more extension friendly +Raul it looks like I am not able to reproduce this error on mySQL despite me executing the test attached several times in sequence +Documentation is the key if we need to clarify anything +The exception occurs also with the portal and authentication examples +has been marked as a duplicate of this bug +And its those tenants databases which are really used +Includes one file and convenient wrapping class +Ive just committed this to trunk andThanks a lot for the contribution Colin. +Simple distributable output including the stacks. +This issue duplicates +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Hi JacquesWaiting for comments from devs and waiting for Nicolas to make the tests passCheers +The column schema with the changes from uses simple data types which are supported across +these were set to CR but it was they got built into thus fixing the fix version appropriately +Thanks RandyIve also added the examples to the test makefile. +Even if had the in it the test WILL work +Thanks for the confirmation on the patch Heinrich +If youre brave enough to push there it would be cool +Im not sure there is a problem I think its correct now and was broken before +should be it is fixed what is the fix version? Please resolve with the proper one +Im not sure about the additional exports but I assume since these classes are listed in they need to be exported so that the Felix jaas bridge bundle that processes this bundle header can find themI havent actually used that in my work so feel free to ignore the additional imports for now until we can confirm that the jaas bridge needs the export I think it does but havent verified that in practice yet +Xmsm Xmxm XXUnfortunately Im still unable to exactly reproduce your failure +be careful +I have a couple more points of feedback but I will file those in a JIRA +thanks for Frank Ive committed this +trunk Committed revision +Can you please test it and make sure it does what you expect +Bug still present using commit cbebcdbcabdddedbfbcfa +Fixed broken test added for x and trunk +I have added a README to the sample will add a document on how the support done with pojo deployer +Thanks issue is closed now + to trunk +In a very large cluster requests would rarely span across shards because the data would be partitioned in such a way that the queriesupdates would be contained within the shard itself +confirmed by Steve on IRC +That is the reason for command failure +This is not setup correctly +Ill be working on improving the quality of integration test cases and will try to eliminate unnecessary noise on the console +Patch that cleans up the tests submodule of +Assign to Ales for some future release +Applying the patch would but I completely agree that the current code does not work correctly as the stream were reading from is closed too early +Sounds good thanks Areek! +Ok so I rolled it in and then we did a sanity check on popen before releasing today +Also Id be interested in that code +new system tests are added +It would be better to take a look at that first +Right now the behavior is that we stop consuming its output but the task still succeeds so long as the exit code is +Added file Simple regression test +Knocked down priority from blocker to major +I would be ok with a different value if there is a need but using a growing delay would mean that after some time the delay for the keep alive could be one day which does not make sense i think. +Please choose a number of prepared transactions equal to or greater the maximum number of connections specified in your so that there is a transaction available per connection +Thanks for your input Eugene +This is fixed and can be closed +Guram Savinov a workaround put composite id properties into the entity and mark them as readonly +Artifacts displayed are probably due to the same cause of the linked issue +Fixed in r. +it seems a geometry is never encoded! This patch is pretty incomplete in my opinion +As I filed it building the cocoon docs worked with it now they dont build any more +wait you tested it locally first right? Its not difficult to set up a streaming environment locally +I went back through these exceptions looking at their call hierarchy Im happy with the selection as they are +trunk Work in Added operations to JMX to suspendresume a route should enforce a route must be stopped before you can shutdown and remove it as a fail safe +Because on my setup I cannot reproduce the issue using only stop etc +Here are the trivial changes +Simply wrapping sasl in protobuf headers would be sufficient to address our near future needsI meant would not +I believe addresses this +I will create the other one too +Merged to the branch in revision +I think we would need to embellish this patch to save the setting in the system table before committing this patch +probably in the mailing list +Are you saying inconsistent data is something we should allow +The user should use the admin console +These are outer bounds that a sever might impose but if the connector requires it should be able to manage tighter expiration than those imposed by the server based on the connector binding properties et idea here is centrally managed caching service under single node which would give us better picture of what our memory needsfoot print across the system manageability clusterability +That is if the complete request is larger than sizeMax regardless of how many files or fields are being uploaded within it it is rejected because that is exactly what sizeMax is defined to doThe semantic of file is the maximum size permitted for a single uploaded file which allows the processing of the entire request but rejecting any individual files that exceed the specified maximum size for a single fileMost of the issues being raised in this report could be resolved by simply using file instead of sizeMax +applied this patch the + +The problem is detected in and given the status of warning +Here is new Chapter on Upgrading that I added that makes mention of this issue and the fixup script +Changed the name to +Missed a checkin of a simple change to the original code The configuration uses framework properties instead of system properties +Because of its low status +I would like to hear Guillaume on this +Is the bug fixed? Why there arent any further reports sent to me +fixed error on add portlet screen in displaying the list of portlet JyotsnaI assigned it too early before looking at the patchSomeone with framework commit rights would commit this patchcan you have a look at this problem +and on without intrinsic instructions the implementation handcrafted by Yonik was actually faster than the one in the standard libraryAnd the so called hand crafted method is identical in and Yoniks code +Let me know if you want it as a patch or if I should go ahead and check in the change or you want further discussions before I do anything else +Unfortunately I realized towards the end that the patch requires the common part of to be committed first +Thx for the test and patch Bill +Closing this issue an old friend because no long valid +Wanted to keep it that way as the that are in memstore keepthe timestamp separateI agree about the fact that we are exposing the internals to the writerreaders +Im also happy to assist if I can in any way that will promote thisOf course if this can be done using other means Ill be happy to useitCheersWould you like to comment on this proposalCheers there been an update? Im interested in finding out more about this feature +Patch to change the log level of these statements +Simonis this still a bug for you? If not could you consider closing it so that we have a clean JIRA status forThanks +I am currently testing the fix in this ticket to see if it works for that issue +Similarly we can also have a check with of patch to fix the review comments +I can check with Leonard Rosenthal form Adobefor a better answer if you would likeBR EricIt seems like the control is useless on the since the Device GrayICC based color space seems not to require ahard coded +OK +When the hashmap falls out of scope the memory will be reclaimedSecond any maps that do remain in memory are required and used by iBATIS so we cant go clearing them out +It begins The following two examples from the sample database +This we can address in a separate jira +Will do on the weekend sorry for the delay +Is it really good to make a record reader if there is no output dir configuredSame with the input? You know there is this realtime processing which does not have no inputFAIL nosql databases have no output pathThanks for the fix. +Yes indeed +Maybe its better to remove that view at all? I dont think anybody use +The new version works on the basis of installing an into the application so if someone needs different behaviour he can simply modify the installed copy. +Add some more system properties so that the server comes up with BUILTIN authentication and some default system administrators username and passwordI would appreciate the communitys guidance here +It seems strange to me to use the package target in order to get compilation errors +Is this designed to follow the pattern we use for JMS sending +Do we have the source to the PDF +Working in latest SVN HEAD +The compromise by Johan Oskarsson on seems like a reasonable solution IMHO. +What version was this patched on? I applied the patch over svn and it breaks at least the excludes +Vikram what is the use case for incrementing in double? Just trying to understand since we are thinking of possibly redoing the Increment API +Note that the archives also need hashesHashes are required so this should be fixedIts trivial to create the hashes however only members of the lucene unix group can update the dist directory. +waiting on the name change +The page waits for all builder and validation jobs to finish +Found several issues with how were being handled as well as how the code handles these nodes that exist only because theyre ancestors of projected content. +Checked that there is no javadoc warningsI have committed this. +Integrated into trunk and. +Version of the GPD is not compatible with Eclipse +Job well done +maxThreads is present looks like it is just the mapper that is missing +Should the built docs be included with diffs or will that all shake out in Hudson? Thanks! update added explicit mention of need to set electionAlg +Will open another JIRA if needed +Lowering priority for now +Bulk defer all frontend issues to +is preferred and the old will be removed as from version +So Ill be fine with the patch once the rest of the code comments are incorporated along with a knob for disabling preemption altogether +Doing no harm is my first priority so my inclination is to submit a patch for the time being +Please verify and close this Jira or provide steps how to verify it +Tested on a full build seems to run fine + provide solution for update a solution on JIRA project is for the Enterprise Application Platform product and as such it is limited to fixes requested as a part of the product development and support life cycleTo file issues for the community Application Server project please create an issue in the JBAS or AS JIRA projectsPlease open a support case referencing this JIRA ID if you need to discuss this issue further as it relates to our products. +The JCA change is in +This behaviour is retained by this patch +target this for CR will include the fix. +ok I moved over eclipse to use put it in svnignore and nuked bin from svnignore +BULK EDIT Moving the unassigned features and improvements out of +This is a candidate for SOA +The user here just doesnt want to click times +DohPossible other s in this code path but both of these must be fixed before progress can be made on the issue +Thank you too. +Just think how to solve this +Hi MarkPatch committed +It would also be nice to ease deployment of applications using Search versions not necessarily compatible with the Hibernate Core eventually embedded in the app serverStill this means one more module to pull for developers this case seems worth it but it I hope we wont fragment it too much for example could we the Infinispan module? it was separated as it would depend on Java +VlastislavThanks so much for the infoIve added descriptions for those features +About removing static references in Craigs case I suppose Version could be loaded at runtime by reflection but the logger has to implement the interface and it has to use the class both supplied by +I addedremoved a space and resaved until now +will do the incrementOtherwise the expected value should be rc +Personally I have no idea why anyone would want to use the but since it was part of Logj I didnt feel I could drop itFeel free to run with this +Pushing issues to the next CR +fix failed test cases + +I have a couple tests I will be adding to that should be fixed by this though +thanksthanksdims +The filter is in the same package to access the protected local fieldThe actual implementation searches a Spring that must be available at filter init time +Committed revision . +Assume closed as resolved and released +svn ci m Applying Alexander Kjalls patch from along with a unit test made from his example +Min did you stop working on this? amr +Committed revision +To fix this I had to completely rework the CAS heap +Recap from the ff discussion We will try to implement a rebase operation purely based on the journal +yep thought about that too but the default collection interval is minutes so this would mean that after the first ten minutes since Ive added the to the inventory messages has been sent +I have reviewedapprovedand committed the patch to the latest development code +If you finish it before we release beta please correct the version +Nithin Did you get a chance to run the test mentioned in this ticket? test +I would suggest you to commit first before other changes are made +All complete +Fixed in in runtime build H and JBDS update site build H +Sounds like we should detect strings and convert them before concatenating +Verified in EAP +The statment failing is as indicated this will be failed but this select does not contains errors the same sql run correctly on versions before select from where test The exception isij select from where test ERROR Y Schema ADMIN does not existThis seems to be the correct exception the column name in the where clause is which is not qualifiedby a schema name so it will resolve to an object in the current schema which is ADMIN matching the user nameThis scheme does not exist Derbys default schema at login matches the username and Derby allows the user to bein a schema which does not existThus this seems to be not a bug +Thats because i have links row in my repeater with itemsMaybe a classNameoutcome cache is possible +It does appear to be the static final field thing probably because that causes it to add things to the singleton class and so the included module inserts some extra fun into the hierarchy +A more general comment perhaps philosophical perhaps religiousThe purpose of hadoop is to allow work to be divided and then executed with a belief of correctness that is the same as would have if I ran it as a single job on a single machine +Sorry that was a stupid typo on my part +orgjboss +Committed public api at subversion revision +Patch for trunk +In this case I think we should not do assert +This is jvm problem and is not relevant to fop in any way +Just so that it is obvious that these are likely to change in the futureThe rest of the patch looks good to me but I have not dug into it in great detail +Let me write it +No less than different bugs in the same test case +No Release Note Required JIRA pertains to documentation not +So we can let the start script start multiple slave instancesAre you meaning to do this on production machines? why this can not be done on production machines +In this situation the risk assessment process is essentially just a web service and is invoked successfully from the processThis has been tested on the +Faces config and exact +committed the patch toThank you Gera for the patch +yet another version which has a cleaner interface design. +If a subclass does not explicitly override equals it will be broken +I committed the patch to both and +The serializer unit test wont compile without the patch attached in applied thanks for contributingSome comments +We would need to resolve how to localize labels in the JSF page though +Realistically though if we can just support Timestamp and before the Signature then that should be enoughColm +Should have +It felt rather strange having to patch because of something that is only needed in one Separate the scm reports from the xdoc plugin and move them to the scm is something that Im not sure I know how to do yet +Needs to add into api component +Adding patch +Would be nice to get just a warning similar to unresolved glyphs instead of an exception + had a word with the Hypersonic people and now have a patch for Hypersonic which will bckport the changes made to to allow Hypersonic to accept downloaded applied the patch rebuilt Hypersonic and tested in a basic way can now start Hypersonic using an address and connect to it using addresses embedded in the JDBC URL started to use this in testing on the experimental branch of JBPAPPGACPIPV at some later stage we can run the Hypersonic testsuite against this patched version to be safe +In the tail end of development cycle and wont have time to look at it till end of monthPatch is a year old at this point +i think i see how to make this reproduce more reliable +pushing off remaining issues in order to release version in JIRA and create release notes +In my demo environment this morning i went back and switched out the f for aj +Rb not accepting it yet since has probably not synced with the last commit yet +Can you try max with max +Shouldnt be that hardperhaps tomorrow. + By default the resource filter should be empty displaying the following lable What? The html component Project should be renamed Component Of There is currently no way to remove the selection of the Component Of tested +Release in progress +Fixed as +Hoss Yeah its scary isnt it? But then there is no piece of software that is bug free and anybody running a production server will be running migration tests first before running on a new infrastructure +Hi Richard Patch committed in revision +Now that work is well underway with AS all previous community releases are +Can we change the logging settings for Velocity? I see two news files created after running the unit tests serviceschef and serviceshadoop +and would be betterupdating bundles is +This is not yet marked resolved. +Checked this inTransmitting file dataCommitted revision +i still have a few questions will appreciate your response +Mike lets leave this around for the next release +Got it thanks +Cant reproduce thisAre you sure you are using the correct xwork setup +Finally got my auto format set up right and it found a couple more issues my eyes missed +You cant change how works without changing the standard +Will upload patches for hdfs and mapreduce +Thoughts about that approach +Make every mode like bitmap instead of levels + +Reopen for assigning to myself while backporting to to +The docs speak otherwise +Claus granted wiki access now I start to work on the Karaf commands doc. +markrmiller +Checkpoint stuff does need to go in first though +I forgot to mention this beforethis patch doesnt apply to +Thanks Mithun for reworking this interface +My bad +Because it does this the system doesnt compile normally on Mac OS X +There has not been any activity on this issue for years +bharath what is the average size of an object currently +We have an implementation we use Goog and could perhaps port it over. + +Please comment +Committed this to trunk and +So maybe the jury is still out on this +it returns trueI meant that done is set to be true +Ive attached a small patch that explicitly checks the versionRange for the latest this patch applied I do get the expected result +Patch that adds everything that was described +looks good to me after skim +rrrrrrCommitted fixes to the woden branch +A simple workaround deploy togather with the web app in the lib dir +bq +ill leave the jar option to keep it backwards compatible +Public now since fixed. +Reading empty file compressed with will no longer fail with +Maybe it is a misunderstanding +I isolated this stuff to separate project it look further for any is closed. +This looks like a very specific usecase Andy +It happens that that edit log is distributed and underneath but the NN would see it as a single required journal and crash if it fails to syncGot it +Minor doc changes neededThe new template id in restore virtual machine API is optional parameter if not specified vm restores to the same templateAdditional thing we need to add is we support base image update from vm created from template to another template and VM created from ISO to another ISOCross functionality is not there +I had ten minutes spare and came up with this +GWT addon now excluding from DN enhancement in commit adbecfcccdbbaff +comments on the test case Can we refactor and its subclass for by introducing an API like canRun in the base class +Will fix the test +Try changing the script location to use the component syntax +Allocating to release ve checked in a fix for this turns out requests for shared data handled by the class were not adequately handling null responses +would require a patch to eclipse upstream i is all the server actions only work on servers not modules +Daniel please verify +I just committed a quick fix into trunk +looks weird without Installers with says Never mind it just looks weird without installers with EAP +This is an improvement that would be nice +I think we can close this bug +Will require some work in the coprocessor framework to actually support unloading in a reasonable way +Then we can close the issue +Eh never mind not normally serialized I just meant to say I realize that the hosts are handled correctly by +The issue here is that the for the war has the ear classloader as a it undeploys the ear it should remove this domain +I have commits to branch as well if you need +Would that fix the problem for you +AlexanderYou have a completely different NPE that occurs in +To turn it on set the following property in GERONIMOHOMEetc filetrue +The patch has been applied +also need to add this jar file into systestssrctestresourceswsdldirectory since patch text file cant include binary file attach for this jira to add resolver interface and implinclude unit and system testfix typothanks SergeyIve gone ahead an committed the patchThat said I did make one changeYour patch only provided the resolver opportunity for clients +Surefire stack into the branch and into the branch. +NiallThe problem is that there were a lot of changes between and the trunk and among the others there are changes that allow Struts to work with Tiles Not to mention that this particular issue is connected to a large change in the API +Looks good +I have committed the patch trunk and +Its a relatively straightforward fix actually +Sending srcmainjavaorgapacheservicemixsaxonSending srctestjavaorgapacheservicemixsaxonTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixsaxonSending srctestjavaorgapacheservicemixsaxonTransmitting file dataCommitted revision . +backported the fix to Steve Is this fix there in +Thanks TimThe fix looks goodPlease close this issue +The question is do we want this? This is probably a question for the mailing I tested this patch with a different machine that I own patch p was unable to remove the old +No longer getting the error either. +bq +Hi AllIs it enough to change the title only or are we going to implement a new message which is not an error? Then I will do itThanks Content SaveI think that you should and new messages to all languages xml files and use labelid to get localization messageThanksVasiliy +Built +Fortunately the thrift calls change so infrequently that we just deal with it manually +Discussing this on the dev forum for feedback if the plan is accepted +seems to hold up better +Im just confirming that my SWT application works fine with Spring Framework +This issue is therefore our showstopper in M to M migration +Committed changes to svn revision These will probably be in post ER build +Please just keep the final analysis per year +im trying to find the issue where that patch were defined +Marking closed. +incidently in r I changed this from using live node count as N to using total known cluster members as N +Thanks for pointing it out. +fixed lookup code so it ow correctly identifies overriding implementations +The new functionality will be run as part of when goes in +It will take a little work to fix so it is possible that I might hold off until the next release is out the doorAnd Remy this bug is now properly assigned to so please dont close it again since I will simply reopen +checking failed tests +My original point was that this aspect of the spatial search is still heavily in development and I was advocating trying to reduce the visibility of the Spatial Tile implementation so that we are freer to do that development +A possible workaround may be to install your own security manager instead of using the default one +Patch with fixed testUnless I hear objections Ill commit this tonight or tomorrow morning +The attached is one of sample validators that fail after the upgrade +HiI applied your patch and did some modifications on rCan you try if it works on your endThanks +Michaelthanks for providing further information on thisJacopo +Edward are you saying that that makes this UNLIKELY to be completed? Or that the three things should be merged together to communally get over the hump +Maybe Im missing something +With this could also come queriesfooas a shortcut for foo +Weve deleted the DOCTYPE declaration because we need to parse the document as a XML document in our application +Patch addresses and. +For last iterations I have seen improvement on terasort and other custom example jobs +Please review and the fixes make the tests run successfully! Thanks Xavier for the help! +Noting that the patch applies cleanly and all tests pass +Ha also inside the file +We used to link to them +Since the original issue creator never responded to Pauls inquiry on where to start Im going to close this as Wont Fix +Should be good +Bulk closing stale resolved issues +Set up the projects as follows child gchild srcscriptsIf I want to have the parent configured to invoke both child and gchild then how to I point to the script in a common way? I mean I know I can have two separate executions and the config both times but note that this is a simplified view and my real example has around or so different projects to exec from a common place +Fixed in revision +How about another ticket for that as a feature +It was released with the fix in +Solr has already committed and had to workaround it not being in Lucene with classes like etc +Only the components owned by that instance are able to modify those fields so a concurrent modification means the session is being access by multiple threadsAre there multiple threads accessing a session at the same time? Perhaps the observers are doing work on the session with which it was registered? Per the JCR specification the Session object is not and must be used only by a single thread +I will mark it as invalidMark before raising bugs it is generally a good idea to ask questions on the user mailing list +looking good just committed this thanks Terence. +This can be pushed to next release +This works as advertisedIn the View menu and on the popup menu when you on a node in the view youll see a new menu Freeze Node on +A small nit is the E in PAE means Exception so the var expands to +on +the SHOULDNOTBEDEPLOYED fault scensrios I change disable code that it is not call bulkAssign method so the region is deployed but the table is do not know how to write this test case +We might be able to just query the all table but the feature type definition should carry around also the schema in which it was on and all tables should only be read if the schema is specified +Thanks for the clarification +And what? First request will be form database and second from FSAnalogical http browser must allways to read local file when it got relative link form html pageVictor +LGTM +As per previous comment the updated test fails against SVN +lgtm +I will try the whole thing with the java and will let you know +Lets revert to yesterday +Reproduction code + +I can make the attached repro into a junit test that can be extended by later +Do you have an estimate for this +No news on that +See also! Committed a somewhat modified version in revision + +As you have noticed quoting isnt done consistently across our messages +Will fix +Bottom line dependency plugin does not seem to do anything for this moduleIt also has a izpackversion property that doesnt seem to be used anywhere +I was also able to reproduce this just now on an iPad but not on my Samsung Galaxy S +Unfortunately the custom JIRA plugin where you had to sign the ASF contribution by clicking the checkbox is no longer working with JIRA now used by this server +Is that Jira spam +julien I think this issue has been fixed in Tika right? If not feel free to reopen or better yet the issue against a post Nutch release +We now use versionned jar files you may find examples of how jars are named inI uploaded libraries of a shark in the archiveI have renamed files of a jars with a format +Committed at subversion revision +Manually testedFabrice and Evgeny Obeo guys dont need anything the request comes from the support team +Hopefully this is more general not just for +One important issue about updating needs to be answered soon as we update to BC Do we want to keep supporting version method on RequestXCRL? As far as I can tell this method is meaningless because as best as I can understand the version number is defined by the RFC and is the version of the specification not the version of the X object +In revision I refactored et +Well since I filed it myself I cant vote +Though this fixes the test case it doesnt deal with the case where the new key is being set as a default in one of the filesThis patch includes another test case that shows the issue +Im scheduling this for so that it gets prioritized accordingly +on this patch +it actually does have a connection just no mapping for EAP etc +PaulYou have made your test with CS? or another version +Hi KimI agree that the syntax you prefer fits better with other examples in the Reference Guide +master pull PR with transformers for legacy PR request +Please use dev for the JIRAThank you +Ok heres an alternate technique that might help us find the guilty thread as this object in is designed to be single threaded +Unfortunately the task umbilical protocol is still on Writables so there could be compatibility issues. +CMS builds thru staging are available but you really need to fix your pomto support target directories on the command line +Logging and debugging is not well described or implemented in Drools for the Seam permission system usecase a developer typically needs more information which Fact was matched in a Rule and which didnt match +The book is specifically for version +All of this was done some time ago. +Nicholas you are free to email a copy of this poster to whoever you want +Just a small comment on review board +In my opinion this should be only in +Fix committed to both SVN trunk and branch +proposed patch at revision to the branch in revision . +Also lets remove from get since its used in code paths +Updated patch with throwing aThanks for testing the patch Bobby +i thnk it make sense to support this as did not update before testing +The published javadoc is the API that were promising to keep and we should expand it only cautiously +Close this issue and start on the next oneThe package rename would be a good candidate for the next redesign effort and we will merge those changes to the trunk when they are completed +Fixed since revision In a multi module project the creation of an archetype and the following generation of the project now ensures if the artifactId of the initial project is present in the modules directories it is properly propagated at generation time +But I thought that once I add ESB config file into any type of project ESB facet will be added +trunk +A solution would be to remove the links lazily +Of course sorry for not providing the tests right away but i came up with this page in a couple of of peak hours of work so i didnt have much time for much moreAs soon as i can ill attach a new version with tests and message catalog +Im not building anythingYou reported a compilation error which would be seen while buildingcompiling the module with Java +Do you refer to the order of the arguments inhandle? If so fixedPatch vb passed all tests +Here is the Teiid patch for cleaning up the messages +This first patch turns Cocoon modules as bundlesI will submit a set of new patches including new bundles definition for Cocoon modules Karaf features descriptor some fix around the some checkstyle fix in the POM better usage of using services +The ones that seem empty actually are used to verify other jars and do not have classes in them +If possible could you please also add a unit test that shows that the current implementation is broken but your version works +Ive added some more debugging stuff into the trunk code so if you can try against trunk and give me the new stacktrace that will also be helpful +Good point my mind is now at ease +Well quite a lot of cleanup in Jitrino is done beside fixing TLS access normally this should go to a separate issue otherwie the work is great thanks to all contributors! +The patch is updated +to Olivier with null columns in map is ignored is a bug of the hibernate? +The old style createstartstopdestroy methods are now supported when for annotation overrides before the annotation style is supported +We should consider defining an and lifecycle mapping for rar packaging +Nothing on his blog for years +OK the problem has started to occur again +I checked build and the tests have passed +If thats not that case then please let me know or raise an issue with the EBR Spring DM temp repository has been deprecated in favour of EBR +Therefore there is no deprecation action required to implement these changes in the branch +committed to trunk + was getting pulled in do to pom having it as required changed to test +That undoes the exceptions in the logs are unsettling but though its not showing we are still using the exception mechanism to deal with normal operationRegards region moved message yes region servers have no remembrance of regions past but I was thinking the commit return message nothing more than the boolean succeededfailed that you suggested or a region aint here +please try different variants of load commands +for Q If no common agreement arises could this be configured via a propertyBy default but you can configure it to be via a property? +Please have a look there at a patch Ive posted which forces benchmark tests to run in sequential mode +Hi AdrianIts great to have such patches so people that need to use it and cannot wait for Cactus could apply the patch and generate a private Cactus versionBut at this point we better not apply the patch on Cactus branch as we wouldnt be able to properly test if it works and this branch is temprary we need to roll cactus ASAP Felipe +Waiting for an for the a PROCESSINITIALIZE event type now this event is broadcasted as the first event +Committed toThanks for the review Stack and to env issues on the ubuntu jenkins machine +this was done on trunk just not in looking into it +A common patch is submitted on +Thanks Devaraj! +Trunkjobjsf +So it would be good to have this backported to +The tool can be written in java if thats more convenient +Setting those properties before server start worksIf those properties are only to be used in startup and not with the deployer then this jira is resolvedThanks Forrest +bq +Thx LeonI removed the GL account info from the ecommerce fileRegardsAshish for your help Ashish! +Then our instructions need only say drop your Recommender jar in the lib directory set the and then do mvn packageIll try it out this afternoon +And Im sure JAXB could probably be used on one end still but that seems like its be extra work and I dont have any reason toNot out of the woods yet though! If I try using with Aegis I get an error which I am attaching the stack trace for +I just mentioned we still have this issue openAnyway i can work on it soonso im taking it. +I can confirm that this fails on nonnix systems as Im encountering the same error on Windows +Care to give feedback if this is what you expected +FlipThanks for the link +Depending on the number and configuration of the comments occur or comment boundaries get mismatched resulting in cropped literal text and ommitted logic +committing. +Not remove them +I can confirm that the listener is loaded in tomcat without the listener definition +Recreate +But its during an integration test with junit +I have committed this +Generically serializing the form will result in everything being strings which is why I decided against adding it to the library +I would even add that it will also be a problem in the hypothesis of +Could you please open a review request +Correct KVM Hypervisor +Hi KevinI thought is what you want me to use +Ive tried reverting back to the previous code where we are setting the JNDI system properties but Ive not be able to reproduce this issue +You can use Eastern Time Mountain Time Pacific Time or whichever timezone you preferAny thoughts +Slightly modified version of your new patch applied at the revision +Will do +This may well be related to +New yarn client +This was brought up in another ticket as well. +copying comments from email to jiraHi FlavioI got the same error messages +Once this happens this patch can be applied +applied patch in svn rthanks +Ok I look forward to seeing the DSL you come up with +com +neither for breakpoint or stepover +Checked by dev +trying hudson +What can the community do to help out on this? It would be great to get a version of this extension out to Maven Central. +RonDoing it that way is exactly what we DONT want to do in CXF as it completely breaks streaming + +The EAR project does have the Seam facet installed +Anton could you please try and reproduce this +Merged +It would be nice if you can merge the test code into the patch as regression +Ah +None of three suggested solutions worked so we took different approachThere was issue in jbossxacml module which prevented to deployIssue was fixed in +Here is an attempt for a fix +logs for addPages operation under DB Page Manager +This patch is for but I think you will get the idea +Well except for attachment urls surely? dbnamedocidattachmentname +I am going to commit this after usual validation unless I hear otherwise +Thanks for the patch. +But we may need to add this to facilitate merging of clusters +Committed revision +I generated class using wsdlJava using the snaphost from it STILL doesnt workThe stub is still being created with method that takes no parameters + +to which? the patch modifies of I missed that the st change was the site +Should we add these things in thereAs for the bug I think its because you end up loading the same twice The CTOR reads the config by default +Patch to complete conversion to C +Im going to commit this first iteration shortlyPart of this change allows you to setup collection to config links before solr even starts up in cloud mode if on startup a collection node with a given config already set is seen it is simply used as isThis should provide the flexibility of one of the more complicated cases that came up on the mailing list a couple months back +Ill have this fixed once is resolved which will be in the next days or so +It would be fun to POKE at it while we PEEK at the source code +The second patch has been checked in. +Whoops! reopened because I forgot to merge to branch +This means users tasks CAN potentially create unwarranted filesdirs in the jobdir Set drwxrws if the is same as the ttuser set otherwise +Committed +Verified on. +This sounds like some maven ide tool not maven +I have also attached a patch for the SNAPSHOT code in SVN +So I would like to work on this issue first +The actual underlying methods are unmodified +Ill definitely come back to you onthis you please explain to me why you do not want to frame each query with anindividual transaction +I suspect the bug is that is using the recursive rather than execing the du commandTrueIt seems that all the jar files for each job and their expansions are placed under the archive dirThey are not shared at allAnd the jar expansions explode the directory depth and the number of files dramaticallyAll those make the getDU call with the archive base dir really expansive +Attaching logs and DB you also attach SSVM log for the failure Unable to unpack snapshot OVA file at mntvolumes +The changes to the pom look goodBut why are the changes to and needed +I have committedmerged them to +Could somebody please set Patch Available on this issue as I dont seem to have the right +Thanks Mayank! +Proposed builds on CI are Qpid Java Build Qpid Cpp Build Qpid Java System would be nice if the Hudson setup could also publish the Maven artifacts to a snapshot repository +My understanding was that an application is Pending if no AM has yet been allocated to itYes thats correct +An test case to prevent this tests are committed in both and trunkThanks Pedro! +Hmm the deploy a JMX facade used by Tools +But the class must ignore the system property +indeed fixed in +Thanks Sumit for the contribution Done at rAshish +I am considering making the default +Seems unrelated +With this latest patch I have run derbyall with no failuresI will until next week days before I commit this patch in case anyone would like to give some review comments +PatchThanks for the info +I just verified in trunk +Its possible for a commiter to create a branch? Its really difficult to progress with big patch and this issue is really interesting for refactoring mail systemthanks by advanced branch has been created at rev the latest patch in this issue has been committed at rev The task has tasks in errorWaiting for your updates +Trivial fix for a trivial problem +Docs changed in fix for on your pull request Ive improved the documentation +Please feel free to make the commit +Can not reproduce this problemIf this still is an issue for you feel free to reopen it. +I have learned something from you +Thanks Lars +Unfortunately the patch that went in made the KSSL configurable instead of removing it and thus this patch needs to follow suit +Well evaluate the patch in the coming days +Previous commit in wrong branch reverted and committed to trunk in . +define an interface for processing requests and responses turn into make processing methods public move default implementation to an implementation package and can no longer derive from since it is in an implementation package Makes sense +As it impacts multiple files there could be some conflict but this jira is quite self containedFor the test part reading the code you mentioned yeah it could be +Attached simple patchThe patch also fixes a couple other places where we hold onto memory for too long +Please test +Given that is for test only agreed with the changeUpdated patch +In Java this will be a single recursive method plus a single call to this method in just before the is createdMoreover this can be an optional feature +I think maybe some tests depend on the sleeps done here as augg +Ive hijacked it in this particular error case +Username rmaher +This is verified +These changes were committed into SVN trunk with r and r and into the branch with r and +buildivylibcommongriddevbin +I have no problemusing this as the name for the suggested enhancementI agree with the patterns you suggest +detects markers that indicate that a field is a checkbox such fields if no corresponding parameter value found +Issue will be opening until getting the proper solution using a Custom URI ResolverThanks Andreas Veithen I have applied your patch + +per raons poltiques I think needs Lindas addition i n a viure en un pas considerat enemic because Americans moving to Canada arent said to be defecting or not as a rule +Ive just committed this +Christian There is also a patch for you to apply to db +for the patch +And because of this I made wrong commits +just Resource +After that Symantec will arrange a where we can explore the details of how this would workThe documentation is in the private foundation repo under CorrespondenceSymantecIf anyone doesnt have access to that area ping me directly and Ill send you the filesId like to get at least one representative from Logging Tomcat and Infra to review these docs and to attend the +I guess the only question is do we want OR +Another possibility to solve this is to use the TCPMPING combination +Really appreciate someone looking at it +Thanks +We should remove the cwiki site to avoid confusion +If not I might be able to reach out directly once Im done with the things +The never gets called in the simple process of the Admin user going to admin page getting redirected and then continued after authenticationWhat is the mechanism by which the clear method gets called anyway? A static is created but I cant see where its ever referenced or used anywhere else in the code +As only one of them is installed only one upgrade will be applicable. +That issue is marked as depending on this one but is it the other way around +So what I did was changed the code so that those attributes are not included at all in the Attributes object +PigHCATCOOP is addressed in View is a typo in test case I upload to fix itI am still looking at Authorize and HiveCheckin +I committed my patch to branch and trunk +Ill upload a new patch once finishs +I tried to apply the patch after committing the streaming milestone and unfortunately that patch does not apply +temporary workaround in the sonar installation directory replace warlib by the attached file +Note the existing roo logging command +its not a big deal in this context and probably not worth creating a ticket aboutThanks +Which application server is in use there? this for the time being since theres still no standard way to detect stale JNDI proxies +I have two use cases at runtime Start a standalone Tuscany runtime with embedded Jetty to provide HTTP communication +Not a blocker for +I was mistaken in my earlier comment that it still occurs on trunk and managed to pick up the jars instead of trunk +I just wanted to know if there are any guidelines set up by the entire project and agreed on some refactoring standards? If not I can definitely come up with a proposal +I think we could leverage some of the ideas in Spring but does have one interesting feature service models which makes implementing this stuff MUCH easier than the way they do it with Spring +I suspect though this is going to need to be solved in BTD +I tried to apply the patch but got some build errors +Thanks Namit +Part of a bulk update all resolutions changed to done please review history to original resolution type +Accepting +Hmmm +You can assign back to me ifwhen this is changed on the back end +a list of the signed copies of the sameRPM signing and JAR signing are not relatedThe conversation in this thread appears to be confusing the two +The reference manual does not have refreshRow listed in the tableof supported methods from the interface so theembedded behavior is probably what we want +Is this where the confusion is possibly stemming fromWhen we make code commits we are committing to the trunk development rather than the development +Latest patch from RB +And no the test suite probably does not have perfect coverage of this. +Will isolate the problem change and resubmit +If youre building the itests projects while being behind a proxy get will fail +Updated the pull requests is now inherited rather than hardcoded in parent +setting affects to and also +Updates should be installed from jbds update sites +Here it is sorry it took too long +At this point I dont think is related to this issue but I may be wrong of discussed dont think this is connected to abort +To alleviate this issue fault messages and response messages have now been wrapped in a intance of the class which fixes the problem. +Resolved +In the event that the socket drop is not detected then the tcp keepalive wont kick in for up to two hours which is why you would see an error like this +Weve tested this with the WSO ESB and it seems to be working fine +MaxAssigning to you because I cant update the archetype +Woops last patch was missing added files this one should fix it. +dubious practice I agree +This allows for example a query based on previous results +Thanks for the forward only clarification text was removed in JDBC +Perhaps it was in an email but I seem to remember you reported improvement in favor of BS + +While we had some discussion offline and decided to do the logic in object store level so I need to make some more changes +Tom what do you think? Can we go ahead and make this change now? I suggest we move the python contrib to a branch in the subversion repository +Committed patch a to trunk with revision The test has failed more than once in the daily tests +It seems there are more issues that could be closed being obsolete or old or not reproducibleMaybe a list could be collected on the forum to cleanup those old issue? +I dont know enough about replication there The put or delete gets written to the WAL with one UUID but internally contains another +so while the validation will state that username is required I dont feel we strictly need a visual identifier saying as for the other page I think its fine to have the required on it even if the administrator decides to not put a password in there +Add the missing TestAll tests are green under eclipse +Wish is more designed for extension +I tested on and then rebuilt the patch against Apache git trunk just now +So instead of a wildcard you can simply list all possible names used in your cluster and only those that actually exist on a host will be used +v fixes the yaml indentation so it parse correctly adds a checkbreak out of the inner loop over the page size and finally logs that hints are paused at the end so its clear that they may not all be delivered +in addRecord and get its an error if lastModified fails since we cant tell whether to call set or not +Ill go ahead and commit this +Both patches should be applied to the root directory of a classlibtrunk checkoutAdditionally the following svnignore properties need to be setAdd +With this my extended test case that was crashing with the same repeated reference problem is now will merge into and the fix in MSP with r +I dont know +This work is a temporal resolution +ASF KiranYour patch is intrunk r r r +I just committed this. +expand on what cant bind means +I will let Permaine know as soon as I know +Also the status attribute doesnt seem to affect the of the job or the step +This patch does the following +No it is notI ll make a updated one against current code baseWhy the change to bestFit in this patch +This looks good! Thanks Steven +Even tests run parallelly it will get unused port onlyUpdated same in latest patch +to new patch +if Application is still running then we may not want to delete the logs and wait till the application finishes or we may just aggregate logs during restart +I will continue to former work on if others have no time to work on it. +i believe Rich is working on this one +bq +Thanks! +This is not necessary and is a risk for introducing errors +The system cannot find the file specified +Problem related to indexed parameters in the column definition and where the underlying data contained empty cells. +I checked and it appears that Im running the version available from IBMSo I switched JVMs +I think this bug does not exist in the code anymore +Afterwards will be accessible via the transitive beans mechanism +This file is kept for adding the drop and create new database statement like the other database Oracle sybase postgresql think we can remove this file now and add these statements after CP release. +At first read this makes perfect sense +bq +When I browsed it earlier it looked like youd written an entirely new method but I can see that its now just refactoring +Said this I wont push further we should have a testcase and decide if the on the fly is to be kept in the public version or not +Please let me know if there is something I havent done or if you have any concerns with any part +Update of JVMs is not working properly for me I need to check with the console snapshot bundled with AS server itself +Ive added a comment and link to for your method use case and a link to for your MVC strategy requirements +The code cleanup went in as revision into codeline + resolved issues + patch looks good +so it seems my suspicions may be correct +META +is maintenance branch we do not release it you can build it yourself or wait for the next release +This issue is fixed and released as part of release +They can then be opened without having to specify the password using async on +Filed for follow up. +Patch looks goodCommitted revision +If there are no stack traces I would be surprised +a maxpacketsize error +When the TT is stopped normally there are better options to stop the health check script +Vinod this is looking good could you please regenerate the patch it no longer applies? Thanks +You can work around this by going backwards +I think splitting like this will help get more codes and ideas in and ultimately Solr will benefit +Looks like this was addressed in +patch looks good +from me +To clarify is a direct copy from most likely for Java compatibility in NettySince the website claims that runs on any platform with a Java or later runtime and is in abundant use throughout the source I assume the use of is a simple error +Done for you +But its so convenient Id almost rather try to improve and make a new than change thisand this still doesnt explain why it works for ant +this was just a first idea but youre right +Please verify +In any case its a totally different issue than what this issue is about here the issue is that even when an inventory item is pending transfer it will still be reserved against orders +This could potentially affect other use cases but up to this point this only affected the integration +does not seem to be an issue with Springs but rather with the underlying web service provider +Yep thats the ideaThe class will have to be Writable and should live at the same place as +Giri has been released so technically this new patch wont affect it + +good to me +Johan for the quick commit. +ClausI decided to normalize the payload first to a standard pattern and then apply the match finder +This could mean encoding the level or the size of the slice into the slice which would elongate slices in general I suppose though that the level index would only add one byte and that would be okay +I like modularity but we will have to carefully benchmark if we make that change +What is actually looking things up in the jca context? Why cant we just drop it completely +This is also fixed +Refer to docoI am closing this issue dont like the idea. +I would think path for both and this bug need to go on +A new parameter excludedParams has been added to table tag +Performance indistinguishable from baseline +For example we have some tasks that use +Michael Frhler submitted the following patch to do WKB with +Uploading the correct latest patch +Hi Scott could you clarify for my what you mean by follow the complete hierarchy? Thanks +You might loose some functionality by not putting it in some frameworks however in felix you will not loose any +Im attaching a patch to switch to opening the output file for overwrite + to trunk +I agree I think I need to see integration tests that act as examples +Makes logicalclock volatile Creates accessor methods for currentVote and state and enforce their use Accessor methods are synchronized +Maybe we can append the currenttime with a random number to generate the storageid +Im sure this is fixed by in some update +Committed to the trunk +This revision does not change any groupIds in case people hate this and we want to revert it +Thanks Kev but that wasnt the question I asked did anybody know what was the intention of capturing the ftp setting? Has anyone an idea why is it being set on the since I cant see it ever being used which is what I think Gearge means by ignores local FTP location? If its not being used for anything then its just a case of removing that code perhaps +One approach would be to have a property which could be set to any combination like telnetconsoleTo disable it have no value +This sounds very promising! Thanks for diving into this +I closed out a similar ticket recently +But tomorrow now +ignore last comment I was looking at old workspace . +I cant do it myself +Nevertheless the new ACL module is demonstrated in the Contacts Sample and has been tested successfully so it is definitely recommended for all new projects over the old ACL module. +That said I dont see any evidence here of it using! HEAD +Attaching and which incorporate the latest comments and also fix a couple of typos +batch transition to closed remaining resolved bugs +Tested on a small cluster seems to workOne limitation is that we dont automatically resolve deadlocks +We have to make sure traffic redirect to Apache site after migration +This one doest not contain the home page and only contain the process and instance pages +Is this still a problem post? It ensures that all filesystems returned paths in objects +on this patch however even with the backwards compatibility now provided Id like to see the HDFS counterpart and commit them both at once +HelloIm trying to use JAXBJackson in order to have one model but being able to choose between XMLJSON formatI encountered the same issue as described aboveSerialization works well but not the deserializationI suppose that the deserializer should use the unmarshal method of the to transform the String into the adapted type +I have the same problem I use the embedded with Jetty to publish a webservice in my application +As Im stressing tests to track down other failures Ive learned of another subtlety in this in some cases you need to decrement twice on error +Attached wsdl file generated by Axis +The ant sane ant all then ant insane etc +Hi MalcolmPerhaps but how will one handle the security check in an Ajax context? Generally one would do the security check and perform a redirect to another page +It is not a bug +ve unprotected the extract method and added a create one which should satisfy both requirements. +Resolved both in and. +Based on my investigation I would say the application needs to store the table fields that go with any views during CREATE VIEWCurrently the application keeps track of which tables a view uses or which views depend on a particular table +Moved to +Will include in +patch is excessively simple so no review board +String contains only single quote marks +In that case you want add +Thanks Kevin +The patch looks good +Nice feature BTW we need it in spring xml as well adding a TRACE log where the looping takes place would be great And it should be added at the release notes as well +The fixing pull request is allready up +No not yet +bq + for the reply when I validated the XSD using the XML SPY I dont see any error can you advice +Hi HiramThanks four your help +In order to better understand what is going on here I thought I would try to develop a repro without the user database +We have a second plugin which seems to work correctly and where the plugin has the correct am looking further into this +AndyThanks for the reminderDid you mean that would cover coprocessor handling +Ill look at implementing a destroyreset type function in clusterMetrics +I will commit this if there are no objections by this time tomorrow +Can you please give more detailsThanks +It works good +This will effectively act as a valid lower bound +Applies locally and if a change to hbase happens in meantime should still work +I cannot duplicate this error with current CVS +Hi Suresh Can you also backport this into? Helps avoid backport pains since this touches and adds a lot of lines +Would this work ErikHow will Result work? Since thats a return out of maybe you want to wrap the Result in a? Then have all method invocations on delegated to the wrapped Result? Maybe KV needs to work this way too +Hi MalcolmI agree +Or simply read through when you close it ensure no cache file left on disk and correctly close the connectionAppend as +Verified on test is Success. +All volumes are upgraded as a unitBalancer Should we clarify that balancer doesnt move the namespace? Only the data is moved +An alternative to cut down on space would be to see if the licenses allow us to extract only the relevant functions +Youll need to delete the gbean if you dont want it started after a server restart +We use standard grid editor class as you can see there are no a single line of code from our project in that stackYes all traces are the same +Closing the jira based on this +Patch applied +Attaching a patch v that adds some clearifying comments and makes the code change described above +Fixed in dfffdeddfbfddbTodor thanks for your Patch! I went with an extended solution to the problem but I used the enhanced test case that you provided +Next tryr in santuarior in wssjHi MarcThanks fix confirmed +It will really help in alleviating the Task failed to report status for X seconds +some code feedback on RB +Needed but we wont have time for RC +I would not consider it +I am trying to test it in more scenarioes currently it is a preview version thanks for any comment +Moving out of +I removed that code +Alessandro Im working on duplicating this and providing a fix +So having a in the method sigs is not a good idea +Fixed in revision to the branch in revision . +We think we could improve ourthroughput if we didnt timeout so often +Thanks for the comments KonstantinYou are right about +Ben Can you point out where in your MRI VMS port this is handled if indeed its logic? We can probably make the same change in once we know how MRI handles this particular I actually meant was that to mimic MRI behavior tries to look for an executable target +Scheduling for in svn +Please use the file to store the Message Constants that you want to use in the exception +Looks good to me +Reverted r +Patches against trunk revision currentCompared to previous patchesfiles there are only javadoc updatesand the javadocs of Searchable are also patched + +But you forgot about second case and srcmainjavaorgapacheodedaojpa +However it would prevent using POJOs in the future +has been reported on single core machine with journaledJdbc so primary factor may be machine digging a bit more it looks like the spring consumer endpoint is getting itself into a strange state +I just realized that wipeHosts will knock out the system table as well so the downed node might try to bootstrap when it returns +Remark previous patch was specific for Attached patch for Apache Oozie I dont remember what made me think it was were also seeing it now in CDH and we dont have +just checked in extra validation code that should catch this situation +I dont think we should use across the board since I dont think relative paths would work correctly that wayHeres an updated patch that includes my fix with your tests +Made the averages dynamic rather than trendsup +Hibernate is just using plain java collection semantics here and if your equalshashcode implementation is weird then the putget operations hash based collections will also be weird +Lets add nodeprobe support in +Then we could also commit it to +I think its probably best to open a new JIRA for the bugfix since this one has been closed for a while +just committed this +Fixed in SVN version . +Finally the two most methods you mention need the same data structures so synchronization wont help them will itFor these reasons my instinct would be to instead try to optimize these methods both so they use less CPU and also if possible minimizing the time they need to keep things synchronized +will be a nightmare +addressed Todd and Coss review comments +Committed a fix for this +This issue has been fixed and released as part of release +Were the tests run using the same set of queries they were warmed for? If so an interesting benchmark might be to +and Decoder scan the input source to detect if there is any characters +Pulling from the branch +Please close this bug when you have a chance to verify on a Windows machine. +Thanks for testing attached patch should fix that +I just committed the update. +Summary changedregs applied +thats good for me +Once was resolved I did not get this error with GF and the Weld Snapshot from th +Done at you Jacques. +Also what happens when the user refreshes the page? Typically you dont expect that to be counted as navigating away from the page. +I put it into trunk and +we can tackle the todos on think we might want to beast the tests on this before merging +Waiting on feedback from the reporter of this issue +if the content length metadata is available the stream render servlet sets the response content length +Canceling patch to address Sureshs comments so far +The main problem though not impossible is syncing the segments +Triaged for +Patch attached +I just committed this to and trunk thanks Rayma and Matt +The ASF does not do any development on closed lists +Thanks Todd +Ah I see what you meanIs there a use case for user home that dont contain the username +Id rather spend my time on something more sensible than duplicating code for something that will never perform properly +Because a session scoped lock will expire as soon as the session logs out there is not need to provide a lock tokenIm not sure about the case where you want to transfer a lock to another session i guess this is not valid +Also affects for being confusing andor confused +That way existing applications should not be surprised by the lock timeouts and setup to handle them +In this case it is an integer that its why its failing +Dependency so that correct constructor is used to create Exception. +patches contain improvements for both generate and validate goals in backwards compatible manner without new noticed that the codebase has had some grooming but the patches still apply with trivial adjustments. +Updated patch with change from reverted +Typo correction the Hibernate version is not +An issue was found and fixed with +Thanks for the code change Hairong +Wherever possible you should create separate tickets for changes that span the +I missed the last Saschas changes and Erwan noticed that they do not begin by Accounting +You will need to use the JDBC driver version or newer and append responseBufferingadaptive to the database URL +As soon as the dispatcher is stable we should remove the code in targets has the result of the was an easy fix +I have attached a patch which adds the new method test to to check this bug fixes the updated patch addresses all mentioned problems and the tests no longer throws a NPE +Will merge it in to master after more test at tomorrowRohit if you could run marvin test on for this fix please let me know if there is any regression caused by this fixThe fix should greatly reduce the memory level and also improve the startup time of it after dont face this problem anymoreHence marking it closed +It whould be nice if the other committers can give some featback +I think this is different +Thanks KristianIt makes my day to know that Hudson does not find any problems with my backport +Committed but move outside of the block in case this happens elsewhere since zero cores is always invalid +I am planning to attach a patch for this in the next few days +Then the JVM uptime is not the same as your app +Yes this is indeed a bug in and +It also includes a shell wrapper to make starting the client more +also it would be nice if the variables returned by this method would be sorted declared as an event type get delivered as an +Thanks Shai +Resolving this as wont fix because users of would never see the fix +So this is a failure only when hsql is used +Thanks Sam. +Im not sure how that helps with all of the other persistence providers as this would be a Hibernate only solution. +Additional testing This is not just about the automatic prompt to accept unknown certificates this error is the same if the CLI is configured to use a trust store containing the servers certificate +Oops +Could you please try a recent snapshot and see if it works? at your in the forum I think the problem lies in the way the servlet is configured in and the way returns path info +Alsothe with respect to errors tag the delimiter was probably meant to be an attribute but not defined in the tld or no getter forI reopened and did some more work on that when it became apparent where the true problem lay +The current logstream implementation is relatively decent +Thanks Tucu! I agree I have created. +After resync everything works just fine. +Running warinplace as part of gwtrun is not a bad idea anyway. +I didnt do that at the time as it seems like its own JIRA +There was an error in the entitymodel definition for theFix is in Trunk Rev +Marking the two features as optional +Patch committed +Fixed in ffefc +This moved to in a post split world +Its related to +Verified with +Attached a more informative Intel Thread Checker output file obtained with Intel C +the issue depends on queryexporter and hibernate plugin execution doesnt match up well +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Testing with some links provided in search engines were successful. + +I have fixed some versions of in ivy files for corresponding with current maven dependencies +Same as but I use replsync with no FC +I dont think that will work well +Thanks Trustin the fix in the branch works fine +directly send leader without the purgatory when all partition in the request has replication factor oneJoel Thanks for comment new patch uploadedJoelThanks for comment this part is changes in patch v potentially some other places also need be changedRebase from Jays patchSeems some issue with rebase +Thanks Yongqiang +Test failure is unrelated to this patch +Patch attached +But due to the cost of an allocation and the specific use of this buffering filter on my configuration it seems accurate to use this special mapNeed other testers to confirm it +The only ejb file included in the standalone is this file will have to be added to the exclude list when building soa esb name serverdefaultlibclient +Issue resolved as the requested changes have been made and the modified document. +Although I am excited about the cache solution we feel that the proposed solution for is less controversy straightforward and also exposes more information to for users to write more efficient code +IIRC the windows loopback adapter does not support muticast +I decided to make the factories stateless then it works much better. +Cleanup of unclosed issues. + +If this is a oft repeated it might make sense to put that in the bookkeeper client code itself +The workaround is relatively clean and I dont know of a better fix without changes in itself +Seems awfully limiting and its not really checking for validity of anything possibly used later in replicator code +ToddDo you need more details from ChunhuiThanks +Maybe add links to where the sw is located +Ok thanks +As such I have resolved it as WONTFIX to try and clean up the bug DB a bit +Fix committed in revision +well we can do that as well +execute view collect privilege to access view but do not collect privilege requirements for objects accessed by actual view uqery execute select collect privilege requirements for objects accessed by select statement create view collect privileges for select statement the select statement for create view falls under category above +Assigning to Arnaud +The text cant be extracted as the pdf doesnt provide any readable character mapping +Can you please give some specifics on what exact action you are taking in the gui to get this to occur? Are you running a scenario? I ask because Ive sucessfully run scenarios which logs a user in +Now if thats the case as well hmmm +Create new resource with predefined H connector +It seems like something strange is happening with SVN +Id advise rolling that back +An optimized index is a single segment reads from multiple segments and so I would argue its not optimized and should return falseIf will work when all of the segments refs are optimized somehow I believe the workaround belongs inCounterpoints +Closing released issues. +just committed this to trunk and +Lets see what Hudson thinks of this patch +Also adds comparator metadata for column names +Additions in this patch Fixed results summary NPE Redirect page at Done to the index screen +Part of a bulk update all resolutions changed to done please review history to original resolution type +I will make those changes and check them in as a separate commitI wanted to check on the exception handling for the change +I think the second version of the patch is broken +populates the data table and in addition Ingest populates the directory and index tables +Part of a bulk update all resolutions changed to done please review history to original resolution type +Confirmed that this is only an issue when resolving snapshots +Sorry about the mess and thanks for trying hard and being a good sport about the patches +The counter needs to be given to and never the merger on the map side + Username anorbert +I think done in the last day or two by Igor +The time of works for this issue was days +The only remaining problem is that even our covering this issue pass there are some edge cases when I observe that Cocoon puts my proxy class into illegal state +got a screenshot of what you describe? hard to follow for me +We use eventfd to wake up the event thread on an unloaded system but it would be best to avoid using it when the system becomes loaded as it is expensive and tends to cause spinning on moderately loaded systems +Kathey as for not being granted the server will normally work and still restrict permissions +Guess Ive got an eye for stuff like that +that is for sure a big motivation in spending spare time on fixing bugs. +What interface do admins have that make this obvious? If a cluster has it isnt going to be obvious in a web UI that any given TT is sick +Ive filed for making the two code paths the sameIve also incorporated all your other comments +HiYour patch file doesnt provide any informations on which class has been modified +Could you write up a technical description of the method suitable for inclusion in the specification and think about writing test casesWhen does a class appear in the list? In JDO classes and metadata are loaded lazily so we need to decide when a class is required to be in the list and when a class is allowed to be in the list +Please add your additional comments thereThanks. +Youre right +Fix applied in trunk and the NMS fixes branch +passes all tests in my local +Just trying to clarify is the idea to just remove the interface classes or remove the interface classes and rename the Impl classes +bulk defer of unresolved bugs to +It is ok for to be missing +Perfect +The DLA is an address not a queue but a routing address this may have n queues bound to it including the default +Could we solve this issue with Bryans suggestion from Store Mapfile indexes in soft references + +BULK EDIT These issues are open to be picked up +Resolved problem with TC +The view is made up of delegates +See my last sentence I think your solution is fineIn my comment I wanted to point out that assuming and documenting immutability may still be dangerous especially when all the millions of Pair implementations out there do it differently +Does that mean we will advise users against using it with? If were not going to test it that means were not taking responsibility for it thus I conclude that we must officially drop support for the feature however thats done +to coredumps +Lincoln I didnt directly answer your question +Is there a reason for the definition by programming by contract +Hmm I dont think so Ben +Closing. +I think the smart resolution thing is only worthwhile if we really want to keep backwards compatibility +Heres a script that does the svn mv commands and a patch file that should be applied afterwards +Xenserver support can be merged at any moment for KVM we have to wait for new libvirt java binding releaseUI has been implemented just as the spec described +Now that work is well underway with AS all previous community releases are +to be already working nowI think you can close this IssueThanksSebastian +Todd Suresh et al thanks for trying to keep the quality high +Verified for EAP with httplocalhostjbosswsserviceshttpslocalhostPBwsdl +I enjoy squashing bugs or adding new features giving back reports is not especially amusing +I ran all of the HDFS tests on and they all pass +The fix is in SVNThanks is a compiler bug since the extra CV qualifier should not be part of the signature of the function +We have a column family for historical reasons we run staggered major compactions on +Doing anything else is a break with the Thrift specification for the binary protocolIf prior to this patch you were encoding and encoding strings yourself and storing them in string fields then yes you will have to change your code if you want the field to remain a string +if the form was somehow a GET what happens then? Do you still get the nullpointer? Because then it is the method is the same +Committed a to trunk with revision +Done. +Ill attach a maybe im missing something +With an upgrade to Tomcat this issue no longer appears +It is Drools team which is doing integration of DroolsJbpm with CDI +Well now that is fixed +Created. +Confirmed and exclude lists updated +Marking bugs with invalid fixed for release as fixed in. +Also as another note +Ok thanks! +bulk defer of issues to +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +Thanks Carlos! +If somebody knows how to filter this out +Hi TomI wait for your input at the weekend +Removed protocol package and this patch uses protocolPB to be consistent with other protocol names +A patch handling it with the following strategy if the exists nothing else try in TCCL if found use it else try in compiler artifacts else try in runtime artifactsWhen found the is copied in target to ensure it is a File +Currently with both tx and non tx all owners flush the remote L caches +This method of deployment is not working for the AS as Embedded has version differences in the common libraries that of +this +Not convinced that Collections will make it toNon decorator code in the various util +Please see and the user guide for details of how to set the table prefix for the extended daos used there +We tested some other XSLTs with our applicationin the zip file i sent you testdata +I am not setting the fix version yet +error logs +perhaps this should be a feature request +Nonetheless all preceding HADOOP tests are passing and this particular test in no way reflects on the functionality and compatibility of within HADOOP +Im not sure when Ill next find some free time +Seems to fail with +definitely feel free to check the source and create issues if you find any remaning issues +So could be checked in as it has passed all tests +Unlinking from. +Thanks Niranjan! +Thanks Arun! +Can then see in morning if it actually fixes build +The biggest problem here is that we need to raise Thrift exceptionsAlso included is some minimal system tests +Turning off ipv entirely doesnt help +v attached with those missing parts + +I dont like the patch so much though +Uploading +Same thing goes for the setting of static variables present in the compiler required for some customizationsIt would be nice to have a unique and common interface that could be used for all the possible options +There is no community version of Apache HTTP you think for the sunfreeware binaries those are compiled with GCC but I doubt they free to check +Rescheduled for major refactorings regarding these issues is in progress. +Thanks Karthik +Switching tabs in the tabPanel component is of critical importance +Thanks Chris +Doubtful but I am checking the build now +Just pinging a reminder no reply onlist to dev from days agoI currently have time get all this done but busy period coming upGav +Rev +I reverted this patch from TRUNK +Ive committed this and merged to. +I will file a follow up jira to do this for both the sinks since the constants are valid for both sinks so it is better to make these changes in one shot +It does not hang around with +Im just rethinking this should we really throw a deployment failure for this? Shouldnt we treat it just like any other disabled subsystem +Patch applied +A patch that disables these tests +The victim rechecks that there still is a deadlock so that it doesnt have to give up if its no longer part of a deadlock +I am able to test the patch and it worked well +Duplicate of resolved issues for already released versions +Its not as trivial as I thought +Pretty straightforward patch +I skipped the column split and its runing now +That might be the reason why there is a getter but no setter +Its a json bug on serializer generic that I know that you use the and Jettison is not involved I had the chance to take a closer look +Please think long and hard before it is applied +About the merge +also committed to +Sigh +crapa jarjar is working to upgrade cglib to asm should ship a new jarjar soon +Maybe it would make sense to start a new project somewhere at sourceforge or google +Daryoush Please find the files in the attached zip file +Internal snapshotvirsh Name Creation Time Statevirsh Domain snapshot createdvirsh Name Creation Time State running +One interesting thing I noticed is that if after bouncing servicemix if you wait for lines of Warning The encoding is not supported by the Java runtime +The X is changed just once a yearRichard and Maxim do you agree that Sonar must absolutely fail when the provided source directory doesnt exist but your main issue here is that Sonar also fails when the source directory exists but is empty +When flexible indexing goes in users will be able to put datainto the index that allow scorers to calculate a cheap scorecollect then go through and calculate a presumably moreexpensive score +Ok lets discuss that on the call on tuesday +The branch passes make distcheck for me on and RB +I just committed this! The failed tests are not relate to this patch. +Works in +I believe this is addressed by. +This patch is done with Please check it inThanks a lot +I agree that this information usually only needs to be stored once at the node level +limitationThanks +things are looking good guys +but i still believe this represents a bug in axis +In revision I committed your latest patch to the branch and in revision I broke the Java dependency +Not sure under what other conditions someone might run into this so leaving open but decreasing priority +LTQCLQ are hindered by having to honor FIFO with +Attaching trunk patch +Amareshwaris proposal looks good to me +Im trying to publish a snapshot of Commons CLI under the groupId and I get an access denied message +Agreed youre not the first to bring this up and its an easy thing to do +Not resolving for easier review +Otherwise the information is lost when the model is persisted back to xml +This bug does not exist in +With the resolution of there should no longer be a need for a Controller +as per phone meeting thanks Jeff +In this case fsync call will end up persisting empty blocks details and there wont be any block collection in new Namenode related to file +I dont see any remark in our aCC D builds anymore. +after looking around more afterwards there are lines that say the same thing and a few comments saying something to the effect of that it appears to work that way +Verified on. +This update is on the table with BLOB columnupdate t set status where id The attached is subset of the repro provided by Army for +Yes this would be really helpful +Verified showcase and core demo on IE and tomcat. +Erick please take a look and let us know whether youd want to provide more tests on following the template that Im attaching +correct the log stats and add a variable for ssd +Committed to trunk. +bulk defer to +Wondering if I should create a new JIRA under common for this to work +This OCD is only available in english committed changed according to your remark. +Patch needs some more work as loading failing on temp table creation +That should have read asClosing this as INVALID since weve reverted. +fixed in +Omkar can you please address the failure? Tx +Attaching two files new test case to demonstrate the problems with the remove method Patch for which fixes the +Well it wasnt the OS disk that went read only in this caseOh right I can readAnyway its an extremely nice warningerror message to have +Brian bug has been fixed in Xalan Java +This is a minor issue which cant affect default JET execution mode +Thank you Hairong. +Was this fixed +not a priority for Batik and SVG implementation supportI will do that when I have time +And so what about maintaining backwards compatibility with the old constructor +incutatorpublictrunkThe patch includes a breakdown of the new committer process as practised in the Jackrabbit project and described by Roy in the recent mailing list discussions +Integrated to and as wellThanks for the reminder Lars +What are you using to call the testDao +Complex processeslike what you describe are much much easier and straightforward +bq +Should we commit this to branch as well? It really helps debug problems and it is a unlikely to cause stability issues +Ill follow the existing design as close as possibleI think is not a good point to start test +fixed in not yet testedapplied in +Im fine with doing and have committed that in r. +Anyway that second build is an incremental one finding nothing changed and seems to resolve in a few millisecs. +Thoughts +confusion is only due to us not knowing where to find the full product version info It seems obvious to me now that we may have several versions depending on the layers and this will make things a bit less clean +Current patch but still at least another concurrency issue +Great! I committed the patch in revision +Nevermind. +Thanks for the feedbackJust want to know your thoughts about this alternative +Patch submitted to make changes discussed above +Bumping the priority of this bug +Patch uploaded + is part of but it extends from Suns internal copy of Xerces! This should explain the problem you see on IBMs JDK + +bulk defer of featurestaskswishes from to +The exceedingly speedy script is key to the issue +can you post a sample test case that reproduces the problem +What Im afraid of is issues where some function or feature is not detected as present because GNUSOURCE was not defined +Iam testing on svn revision +I enabled security manager by adding the needed lines to the policy file and fixed a preexisting problem with the output when the launched method fails so that it prints the expected output strings and not just the array referenceI also added a decorator so that the schema objects get cleaned up +Logged In YES useridYou are right +Remember that we check for ability to abort compaction every row if were compacting a wide row it could easily take over s w throttlingOh yes sorry +Sorry need to change interfaceName instead of resolved issues for already released versions +Last series of commits have addressed all outstanding issues in this ticket +I did end up rebasing on the hashmap based code. +removed SOA referencesadded extras and extras to Installation Guide +I recommend that anyone who is having this issue try out of nokogiri and test it +Copy OSOA license text from binary distribution LICENSE file to the LICENSE files in and toolseclipsepluginscore so that the same text appears in both places +I went with the first one +A bit annoying as it requires manual persist cascading but thats probably a fair priceAs Hardy pointed out it is still worth investigate the diff between core and EM semantics +Can you try this with the SNAPSHOT build? It should be fixed there +changed. +i know FOP support ttc and ttf +Looks like it works now +At high level the test is doing something likeget timestamp of password filedo an action expecting timestamp of password file to changeget new timestamp of password fileassert that new timestamp is bigger than old timestampThe waits are there to make sure that time between getting the first timestamp and doing the action time changes +I encourage our JDBC spec experts to have a look +I confirmed that it is in the file +Problem fixed. +has been marked as a duplicate of this bug +Also Context has as return types +I wonder whether there is some articlepaper that describes this particular approach of implementing Naive Bayes? A colleague of mine with a much deeper statistics background and me took a look at the details of the computation today and we were left with some open questions +making this against trunk +This patch also deletes the straggler left over from bookkeeper +It takes a while to do a thorough audit of the legal stuff so if we do it too early we may end up missing something thats a release blocker +bulk close of all resolved issues as part of closing items +For trunk the third party jar files are duplicated times across common hdfs mapred +Its true that jobs usually takes longer time but given it seems a better idea to treat all tasks the same when reporting progressI agree and on the job tracker may seems bad but user should see on the command line all tasks starting and finishingWe can print out task startingfinishing information on the command line for all tasksFor example every time a task startsfinishes we can print out started +Adds batching of meta scans so should be faster figuring if table enableddisabled etc +Is this maybe about Spring and its or Spring and its? I cant think of a reason why the behavior would ever have changed so drastically between versions of the framework + This file should be removed +Closing all resolved tickets from or older +Are mojoA and mojoB in different artifacts? This should affect the release plugin and many others as well if theyre in the same artifact +That would also mean less work maintaining a list of reserved words +Implemented and kept the legacy name support as well +Somebody have plans to restore functionality provided by ? Feature still doesnt work +Node save node A to DB +And here is my file that I used to compile the schema +For what its worth I believe that this is simply not possible +Right now I believe it should never trigger a Load Error in the location bar if its a navigation to a url are these Updated Unit tests that keep getting posted with a link to a file called? Are these fixes or just tests +Closing this issue. +This patch sets the value of the shipping estimate field to the ship group estimate +Fixed in CVS. +Im pretty sure the Stop operation already uses the jnpUrl principal credentials etc +The patch passes all tests +Youre exactly right on both cases +Im not entirely sure what youre trying to do +changelist is a for allows full JDBC in timestamps across the network +findbugs warning is +Thats basically the some progress approach outlined by Andrej above. +Like some kind of automatic import of data +The array has to stay around for required error messages but does not need to be used by client converters and validators +why do you need relative paths for anchor links +looks good +Oh OK +This is the exception generated under +Attaching +It is part of Alpha release testing . +I this i have the fix +This test is excluded on and so I think the patch can simply be an updated output +OK at home Vista Home Premium and FF shows the same behavior than with +Hi Dan Im a little concerned about just copying the string across in case the prefix is different unless elsewhere in the copy cope its guaranteed to be the same? +And by wikipedia I meant +Need to investigate if we can do a compatible way in as this would be a good improvement +Assign to Scott to look at +The fix ought to be somewhere in +RegardsGonzalo Abollado. +Filed a related issue +Disabled the test +please attach screenshots and not word documents. +Writing multiple k strings with a prepended flag for the count on the initial string might work but seems like a nasty hack +I agree searching with API is a better way to handle this and will try it tomorrow +Hi CelineI dont see there is anything wrong in the fragment of your code and I cant reproduce the error from commandline by using java +Java does not support that pattern in one file because Foo is unknown at compile time +Great to hear that the latest patch looks goodbq +The problem with background color in Note sections was local in one place of GS was fixed +Multiple implementations can be supported +It provides better clarity to how the file system capacity is represented on Web UI +So I get this issue link from the result of search engine +I am not sure why this is a pertinent reply +sigh This was introduced between v of the patch and v +found +Its ok for me +The New Maven Project wizard is a me wizard +The original issue is fixed +Pushed to and will look at covering all forms of logging to set some standards about what is INFO DEBUG TRACE ERROR and WARN +Closing because it is not JBDS or JBT issue login throught Openshift Express Application works fine. +I will run testsTouches the following fileM javaengineorgapachederbyimpljdbc + +As of this submission you need build of in order to run the jdbc tests cleanly +The patch has been applied +Regarding escaping Im not sure how that would help us here +Ill change the install in to move this to +Spring doesnt even ship which Hibernate is using there so we cant do anything about this +Lets keep this to getting the core functionality working and then we can deal with operations separately +Closing all resolved tickets from or older +for the patch +In this case the was setting an invoker on the web service es since it was the last builder which caused the infitite loopIve verified this with a testcase in coreextensions +Along the way I also updated the canons for dblooktest and dblooktestterritory since this patch changes the results slightly +Digging deeper into the file I can see something going wrong earlier in the file for column When I compare the locations in the index to the locations in the data they are aligned up to row but by row the locations are wrong +Great thanks. +The Message API already supports attachmentsSee addAttachment getAttachments etc +But after they kick in its near linear for quite some time +Committed. +bulk assign all unscheduled issues to next release +Improvement of the with First action is added in the Chain History API to view the flow Parameter allowing the chaining to a previous as rev +Provided some hints in the tool back +So calling an through a core admin command cannot work +So you should be able to update your Sonar instance within a couple of days or so. +Samuel if you have a test case please attach itupdate the subject to include the project issue number and add EAP as affect use this to verfiy you test casefor eap +I see + +Jarek got it +now provides a property called force +This will work perfectly fine without breaking any specs +Fixed on the release branch and the HiranyaNot sure I understand the requirement correctly +I do not want to work on plugins since I want to use Grails as The search is over and not as The search has began +fixed checksum will be in central in Juven! +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Should we exclude them since default has no juddi or what +Did some tests with curl and watching the postgres processes on my machine and i dont see any leaking +start up your eclipse and then from another terminal navigate to your bin folder and type What I mean to say is start up your jboss from the terminal and then run the twiddle from another terminal +I tested with the jetty deployed not with the one +This is just another result of the problem described in +I think we incorporated some patches of yours for this in timeframe yes? Optimistically marking this fixed. +Was actually caused my incorrect JVM arguments i used to run +IT added +Lets not change too much until we get the mockups from Cobb +import statements rearranged as well as final markersPlease replace with a minimal patch +More generally we need to review the design since it will be part of the supported API and I think it still needs some work +Merged to doc branch at revision +Website has been updated with new version of javadoc and svnpubsub is working as expected. +Changed some timing settings for +So this is an expected behavior with hadoop while its fixed with hadoop and onwards +merged to +Fix merged into trunk. +To the committer who picks this up would you please the patch to both me and Arpit Agarwal? We both wrote code for this patch +And I ran it with only one thread +This does not look good at allInstead of changing the behavior of which would change the behavior of perhaps too much how about creating a new type of that adds itself to the body of the table? This would be optional and not change anything to the current if not used +Sounds like a maybe bug somewhere else but I guess we can try and prevent the issue by detecting duplicates in the. +Ill attach a patch to skip those tests +Cleaned up more Consolidated the test classes into one Formatting in for seemed off fixed it I didnt see how we could use ll add resuableToken support to this patch +We may end up porting the logic from the C code to our sprintf but if this is working it can be our interim solutionFor the record the place this would go for mode is in srcbuiltin which includes a bunch of pieces of stuff weve implemented over the past few months + for the patch +However the visible length is constant for the same even if it fails over to other datanode since only uses the one obtained in open +Dhruba would it be better to discuss lease recovery at? This jira seems to be about client close order +Has there been any progress on this issue since it was reported? +The rest of the compiler now loses the information about the loop +bq +That is more than DIN A. +Please verify +By the way this isnt specific to but is an issue with the itself and in theory can affect other such services too +This has been in place for ages. +Zhenxiao Please see my comments on phabricator +OK if I commit this? Ted +toThanks for the review Stack. +Otherwise the parser might prematurely stop reading the image data which leads to a corrupted stream. +Hi it would be very nice if this were fixed +Apache Karaf and is using features to install all what you need +What you feelOkbq +I did not really touch any of the scripts +adding similar tests near there would be sufficient +snjezana fixed some of these at least it works much faster on her windows and my mac installation so its usable +Nor does Tomcat or any other container I have seenbq +Verified at +I remember the Hadoop API is in hadoop as well not sure about +Thanks to Paul Nicolucci for provide this patch +Also I might add some logging on the JBMAR side as well +Thanks in advance +Merged onto the master and branches. +I think this is already documented +looks good to me please submit patch so that Hudson can check it out +the recent changes Alberto checked into cvs plus other changes made earlier I think the problems reported in this bug are resolved +Building more tests I noticed I had made the ALTER TABLE ALTER CONSTRAINT syntax too lenient Ill fix that in the next version of the patch +Otherwise this needs to be implemented and then its probably not going to be an easy one +This issue was introduced with the fix for which throws an error when lobs are accessed after commit +Largely it moves from assuming unicode to using encoding for everything +Unsure if well see a second roll. +Thanks +If this is the case then the default behavior of the aggregator should be to replace nulls with empty strings +While we have talked about it we dont currently try to populate the transaction log after a replicationSo the second core replica is replicating its got docs but no versions then it tries to become the leader but just like with the empty index it cannot successfully sync with no versions as a frame of reference +This way we can identify if this is a bug of seek or a bug of hflush +If I specify I get no error +Also make sure the bridge is up ifconfig clou up +checked in changes so that the GC at least can be built on AIX +Its been modified in SVN so the next update wont rollback +Which metric is measures via JMX? It would be bad if that metric would be double reported now via JMXThis is last jira open against +Hi ClausJust sending a long vale of will not cause the to kick in +introduces a interface and two implementations of it and +Please suggest if there is a better place to achieve this +Need tests +Committed +They all work when actually building the project youll notice that that tests all ran and the jars were fully built +Whats astonishing is that the new GWT plugins now have a dependency to the appengine plugin +There is now some cookie handling in +Also provide generic concurrency numbers instead of LR specific numbers because I will not able to use LR to reproduce this issue +Attaching patch for issue +Now to get working samples with this +Done closing. + recommended +Records support other representations +Can we add a nodetool command for modify rpctime +Should start on that next week +anticompacting data for a newcomer +Committed revision +Or am I missing something againStevo can you comment on this +And the next question is whether we want to spend one additional pass just to find the mean +no tests for point yet +Will we be able to do almost the same thing with blocksI think the short answer is that HBCK might come to look a lot like a fsck for a filesystemHowever going below the filesystem to the block pool directly presents difficulties +Thanks for opening the ticket +Ah we have fixed this in the httpservletjetty component recentlyThough ideally the should check for OUT vs IN currently looking into that +Ive written a maven plugin that parses all maven project definitions in DC workspace and spits out a simple text file that states the inverse dependenciesI also wrote a perl script that reads that file and updates all DC project configurations using perl YAML could be much improved by writing the plugins logic in Java not Jelly it and using a Java YAML library to update the configs in one go but I needed to hack up something fast +you do have in mind that it depends on the locale dont you +There is still work left to do on determining whether the server is out of sync following a reconnect. +I also check the option on each override rather than once by adding a parameter to the method call we could check less frequently +Base has been part of codec since release shouldnt we close this issue and put a note in the changes +By the way The Cache of instances in is a so unused instances for no longer used readersparsers are GCed +I believe Jesse is working on something as well +Help in resolution is appreciated +Well mostly retire it in anyway so Im not so sure what to do with this one +Thanks +I dont see efforts to clean out the remaining usages of the old Hadoop +I think youve implemented this right? If so can you please close this puppy +I believe this has been resolved no and checked into trunk and the because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +whats in the name hadoop? Because had oops +It explicitly says that classes with final methods can not be proxied and that this should be treated as a deployment errorMy gut feel is we should leave the spec behavior as the default and then provide some way to when it makes sense +PR for divert PR for remaining messaging resources conversionconvert JMS topic resourceconvert core queue and JMS queue resourcesconvert regular JMS connection factory resourceconvert JMS pooled connection factory requests have all been merged +Coding is done +You can not develop any integration test with Spring annotations +Marking task resolved +Carl you dont have to worry +I dont think this is an accurate problem report since the assembliesfeaturesframework kar builds fine for me +Jopr r adds defaultValue attributes to all primitive Datasource and Connection Factory config props +Changes committed please test +Verified in JBT B +Actually it would be sweet if we made that a constant in master so that all constants are defined there instead of at the actual configuration get sites +After upgrading from pdfbox to but also updating the application server and changing the from iso to we are facing similar problems with probably unicode characters +Patch containing both new Filters Factories and test casesFeedback would be appreciated but im not in a big rush to commit +My mistaked +Ive committed this and merged toNice work Colin! +And when speaking about maven repository where jars change its really dangerous but should we remove the support for this huge public repo? I dont think so +I tested against CDH Hadoop in this case but if you use apache Hadoop it will say thatSent from my iPhone +It should be fixed in trunk and a release vote is currently underway. +Sorry well that happens +Answered on the forum post +Now it does have a check if the webinf was added from the war plugin before +Reduced priority of non critical issues which have current issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be now +Attached files are based on revision Starting Solr branch +I guess the difference is that I start my ant under RalfI used to have this issue too so I will try to reproduce it this afternoon +updated and reinstated to build. +oops i had some stuff in solrexamplework that caused a lot of noise in the patch +That actually gives better loading semantics +Yes you are probably correct +xBut im leaving this issue openOn IRC rmuir suggested that the correct fix for this would be to use the ivy conf option with filename patterns so that the wouldnt need any special logic and the could rever directly to the files in +Is there an additional cost to adding the bundle id to CNFE? It would save heaps of debugging time where the usual practice is to run the app in debug mode then put a break point in Felix to examine who is throwing what +Done. +runs distributed shell with containers +Done see API on +Would need to check that +Closed. +This has also broken see. +Quince are these new compilation errors a result of fixing or have you never been able to get this far +Committed revision Thanks Henri! +Hopefully other people will be able to find some time to this featureThanksRick +Julien just pointed out the same thing to me +is resolved +I thought Ive already done that +Fixed in +committed w MB fixes +Gavin WDYT of allowing that in the case +It is definitely more tricky which is why adding callMethod is ok for now +The changes made in this patch are very similar to the changes madein Approach except that instead of the streams we throwanFor now the error string is in the +This issue though focuses on the metadata +Trying again with reps +Thanks Noble +The Dictionary class never supported switching of the case after it was created +Im used to working in the httpd world where theres an eagle eye cast on such things +We should also fix this in +Thank you I will try the snapshot again tomorrow + +I only want to import a single quickstartsubdirectory within the zip +Based on Carlos latest comment Im assigning to Emanuel who is looking at cleaning up similar things. +Cant we use that here instead? Imho it is a bit wrong since the is for resolving dependencies that have nothing to do with the transform but it allows our eclipse guys to do whatever they want to do and we just get back a +Thanks Lars for taking a look +Fixed +Please reopen if necessary +Closed upon release of +Can someone please comment on the status of this bug? I just did a small experiment with the default helloX +Ah that makes sense +It works now +Also when we delete some gadgets in an browser and then move maximize minimize and delete these gadgets on another browser concurrently show unknown errorive commited to to fix these to add Fix versions +It is also I believe a fact that the utility of custom attributes cannot be denied +Great! I will try it as soon as possible +Id say set isPhysical to N and see what happensWe havent really done a lot to finish up the actually fairly extensive effort to support tracking and sales of services +Instead of using fileurl which causes FTP handler to be used these tests could use fileurl +I will run a real example before uploading for review +drop contains the passive and active STS CBS Implementation of +This probably cant be fixed on our end but well keep this open until we find a +To be honest the plugin itself is more of a wrapper theres only a couple of classes of original code +bq +bq +Implementation of my proposed changes +Please try this patch on the cluster +The thruDate of the second record should be more than that of the previous record bcoz I guess no point in having same thruDateI couldnt look for more scenarios because of time constraint but I would definitely do it when time permitsPlease let me know if this makes sense or if you have some other scenarios or ideas I would be happy to makes changes to the patch and resubmit + +Hey nice polishingsNo it does what I require +Agreed resolved as Wont Fix. +Not sure where this fell off the map +It would certainly ease navigation and make understanding the individual classes +In addition i dont think that hacking in support for EJB mapping syntax only will spot a good light on addition to the above arguments I am quite sure that you will create a lot of additional work that we can prevent by first refactoring current castor this schedule from me a absolutly no go +Looks like a issue +Yes C would be greatPlease look at this +Also on having generated in the package name +Attached some minor proposed changes to the prior patch +Tested with concurrency levelWill attach the patch with some test cases as wellThanksHasini. +But this is more of a change in the class +We can probably get rid of Namespace if we do this and just use an still waiting for the jBPM release +I looked at it as it is one of the few open issues for +Pulling back into as blocker +for sure looks like something that should be there +Moving to + +This is a problem +To check migration Guvnor Drools guides are checked links are fixed +Committed to branch. +We should be careful adding new API calls in BSP Peer +Good question +Please take a look at the PR so it can be in JBT build. +prepend context to image are you suggesting this as a workaround to your own bug report? I agree this should work +the current patch does not contain the change +For the issues with test launch configurations you might need to refresh the project or restart eclipse +I do have a script that can change this in bulk if you need it. +Jon could you provide a full patch of your modifications? I know youve created a derived class but I guess you could make the modification on GCL itself +Especially as they cost nothing like the ones here +This is the complete patch for this ticket +Using the same build for multiple products is a valid requirement +I assume you are using directly +Waiting for patch or more detailed use case +This is a related issue in SOLR that I plan to work on this week Robert +I think in the original ticket I suggested that we get separate build running on Hudson that tests all of the unit tests against all of the supported Hadoop versions Im not sure whos in charge of the Hive Hudson but if Cloudera can help get that set up wed be happy tobq +Reassigning axis order issues to in gt +However test print Test failed for output values and +somethingmeaningful should be treated as somethingmeaningful so I suggest using while instead of ifCode that manipulates strings has always been one of my nightmares so feel free to improve it as you find better. +If so please provide some details on how I can reproduce the problem. +You are right that CREATE TABLE has to do work in ICTRAN as well as the user tran +But I have no idea what or when anything was releases +Looks like an environment problem though since we cant reproduce on the same sstables +I will give a patch. +Makes no sense how it works today +Not sure on the current status +The suggested alternative compaction strategies dont sound very generally useful +Sorry Shawn I will notice it next time +Good point +Patch to resolve issue +Erm so what +contains the changes to datacache package in support of this feature contains the changes to kernel package in support of this featurePlease provide any comments on this feature implementationThe implementation to handle relationship of Map type will be provided later +Where the target resource is not included in the document possibly allow the template to provide a referenceSo a basic format for the overall document generated for the TIA Project Descriptor isTitle page with information taken from the TIA fileTable of contents this will be autogenerated by docbookIntroduction chapter providing information in the descriptor file describing the projectChapter per Phase with a section per Resource. +Ideally it wouldnt be as course grained as dropping the response from logging but would instead strip the passwordX and passwordX portions of the string +Max should this be documented as a known bug in NN and on downloadupdate pages as I asked above +Passes a string of Java options from through to task invocationAt this time the only flexibility is setting +So are you saying that there are no consumers at all on this queue? You do realize then that at some point the broker will run out of memory +Better to use DIGEST I m moving over damagecontrol to that file anyhows +Yeah thanks Kev +Thanks Ram +This should address the problem +The VDB and models are attached +if it is a class loading issue then the reason for this to be failing is not a fixed bug in but the fact that and do not use the same libraries +See test in release +Wrote several test cases exposing this and then coded solution +These blockers and critical issues are resolved but not verified +This way it will be bundled into the and can be run like +So having Noggit distributed to Maven is as close as asking your colleague to publish itI would rather state that most Java libraries do care about Maven +xhtml page method get in class returns bug was fixed in trunk but not in this be done for beta or wait? please adjust the jira ready +Naw probably not we downgraded it +Any objections +Thank you ChrisI was not able to do it with eclipse and I dont want to use Ultraedit hex view help me to check that your properties files use UTF. +Im for fixing split META in +then there should be no problem with removing the previous library in the build path +Anyway I think this jira is Teiid or EAP related and not JBDS +bq +Thanks Michael Close for all issues before incubation +I cant attach the font itself because it is not a freeware or pd font file. +This will be addressed when with the joint compiler removal branch I am working on +Why do you think that doesnt mimic a clean workspace +Mikolaj do you happen to have the full logs for both nodes when this occurred +Hia just a short comment a repackaging into the myfaces namespace is preferrable to avoid future namespace clashes +Ive been going through the project and it looks like the following are duplicates understand there are probably resource constraints on what can be accomplished +Im seeing what looks like the same error every time I try and add a new node to my cluster +Right good point +Hopefully this and the new scripts can be done together in +Im sorry its taking me so long + +In progress but not yet finished moving down +Max will you look at this please? in H now uses paired Lists instead of a MapH does not have the same issue since it already uses a List as its reprenstation. +I believe Vijay fixed this in r +upload the same patch again so that hudson will pick up the correct file +Shall I commit +tel does not appear by default in one page checkout though he has one in Same for admin +Workaround is to call goals clean install from +This forces a factorycontext pattern up the interface layers into the or its implementations +IE Win Server R x +I noticed that the issue is cause when the client the created the temporary destination and a subscription to that temporary destination is killed +I attached an exported version of the Beans Graph showing your spring beans configuration +I believe Ive fixed the test +I sent an to internal mail list so the steering committee can think about this issue +committed thanks Bryan +For a client call ID is unique enough to map the response to a corresponding request +Externalized dataadapter reformatted code modified styles on the dashboard page +than there is good reason why he is doing release should run the release cycle in the same environment as the builds +There wont be a release +Suggestion have this setable from a java property such as +I dont have the db driver needed to compile it however can you point me to how to get it +If you could explain how to reproduce then it would helpI dropped a sample webapp with a into tomcat webapp services page was displayed correctly +I will remove it +Do we currently support this? The fetch requests always check that the leader is on the broker handling the request before actually reading from the log so Im not sure how a read on a follower succeeds? All valid +looks good Thanks Eric +Also +Has anyone been doing any work on this lately +Documentation committed will commit sample application on Monday. +that was committed in and that broke compatibility with That was dumb +Looks like this was previously reported as discussion with the dev team we have decided that filters should run after a forward +I have rebuild the trunk reinstalled a fresh accessed it from somewhere elseWorks perfect thanks Emmanuel for fixing this so fast! + +The other platforms have had about a month to implement them the bugs for the other platforms will remain outstanding and fixable +Files changedM srcref +Reviewed +IT +In the context of the rest of the console output this might be misleading +Great work Josh thanks! I was looking forward to trying this out but I hit on our version of Hadoop +Please add the new Features of this fix to the. +Great job +I agree with your patches on const strings and the function prototype fix +Please verify the updates from stage during the CR test cycle +Yea Id make sure all component are rebuilt and +Im also interested in the status of Polygon search +With we now set an even handler with JAXB to tell it to not ignore parsing errors +Patch applied created regarding changes +Forgot to grant license to am having problems applying the patch files +Doesnt reading involve computing CRC as well? If so why would there be asignificant discrepancy between read and write performance +The single root commit was from Bob who owns Codehaus +If this class is Serializable and has a readObjectwriteObject methods two things may happena any direct access to a field will reference the element in the proxied instanceb any indirect access will be delegated by the proxyThis may leed to the situation that the values written into the stream might be from different the above case the proxied writes a size of into the stream and writes afterwards the elements of the delegate +HiTo avoid this side effect we could use css class associated with AJAX indicators span +Patch for trunk +Ive confirmed that this fixes test +for +not a bug technically but it make sense to distribute the source nevertheless +Yes +Gernot do you have an option to click on Edit on that page? If so may be quicker to apply the change rather than report it in a Jira +Thanks a lotBest regards since this was resolved some time ago. +This is a WSDL schema generation bug +So first of all thanks for the contributionI have to question about your code +I think the conclusion was that although making the opt compiler less of a memory pig would be a good thing it wasnt the problem with jython BURS cost. +Remember to use a new number for each test +A trivial patch +Although I am not sure if that is necessary +Notice I met this exception when sending the first mail and second one in this morning the third mail is OK +and committed! +Not sure if a namespace prefix is the best solution or Struts should find a way to provide alternate locations +Not just provide suggestions +Ive tested it on a cluster +We dont currently use the murmur hash anywhere AFAIK +Alexey can you do that +any news on this? +For note during course of about a minute we write k lines to the log +Fixed by applying the workaround +is now sort of a misnomer since youre using it to send SIGQUIT +Documentation patch new docbkx sources and image source for image +Another thing I noticed is reloaded this should not happen according to the codeI will investigate into that as well +So what are we going to do with thisCan we put this into trunk along with +In the next few days i will commit the patches so we can move a step forwardRegardsYes that is exactly what I thought about the mapId field +Updated version of patch as applied to build +As per SQL standard if the operands of the aggregate methods are string operands and they do not all have the same collation derivaiton and type on them then the resultant string operand will have collation derivaiton of NONE +defer to +This problem is capture as part of is now fixed by using the latest gogo snapshot +Hi NicolasDid you get a chance to look at this issue +Hence I must ask Is this still relevant in JSF or could it simply use the saved instance instead +The root of this jira is the line +Then we dont need any pinging or tricky heartbeating +Some comments on the patch change dfs to private and +I just committed this thanks Cheng. +I still think this is low priority since for datasets there will be multiple splits per vnode anyway +This issue seems to be fixed in the release +EdCan you upload the patch on jira granting perms +All tests succeed though so I think we simply miss a few tests +Added Affects Documentation +I agree that two tests are similar +in +Was hoping to keep exceptions generated by YARN separate from RPC errors +Eli! Committed revision to +Ive experienced issues like you described and it was by the following reasons now works like a charm +Attached is an initial take at exposing the from the C example as a web service +OlivierYou made my day as your recommendation did the trickThanks for your help +After i switched to a plain new workspace the wizard works +Looks reasonable has this been +The encoding of the input may not be the same as the OS default which may cause display issues and possibly issues with encoding decoding of characters from the stdio pathsWill post to the development list on this +fix test patch +passes but now i see that is awesomeso theres still another bug +Closing this since there doesnt seem to be any more discussion about it. +Spoke too early +Integrated to TRUNKThanks for the patch Scott +Dejan would you know what the problem isI get this again even if I have configured the as told +This patch removes the debug logging of the temporary SSH keys +No unit tests +I will see how bad getting old would be +did a commit in rev +It has not decreased any current CXF fault processing functionThanks Best RegardsIris Ding +Only a rough one at this point +I manually removed the invalid page +Close as +Reducing priority and moving to unknown fix release none of the active devs seem to have an immediate problem with this although patches for this issue are most welcome. +Verified on JBDS +Had offline discussion with Nigel +Cool +This issue can thus probably be viewed as resolved +Trivial patch skipping rb +This fix is already checked in +Its been running in parallel for some time now on my build server and it seems to have the same failure coverage and at the same time is a start to clean up and associated test codeI hope youll also like the new infrastructure will elaborate about this a bit once merged +Closing old issues. +This is broken in trunk +I had a few problems with this patch which I solved by manually merging anyway thanks for your contribution! +If later we add code that need those styles we have only one place to look for them and one place to modify themIf in the future someone want to change the default CSS of the pager elements i will be very easy +With the new registry implementation we are no longer having that issueRegardsLahiru +Complete source path displayed if a user clicks on the source column +PS Shall we actually enforce topic list as required parameters also besides this fix +Same patch with some misc +Yes LDAP connections are never closed +Changes in the early version of the patch Moved all the exceptions from under package under separate files +However in both cases the XML parser itself chokes about invalid XML +Verified with rev +grailsplugins and the CUsersWeipeng +Since its fixed in M version cant it be fixed fairly easily in the M version too +This mixing up of validation messages can be confusing. +HiIll work around this there is any concerns or suggestions please replythank youCharith +Correction I meant instead in above comment +If you can come up with a reduced case we can certainly reopen this bug or open a new one thats more specific. +not sure what could be causing the the netbeans platform app doesnt have the nbm plugin as extension tus missing on the artifact handler that translates the nbm type to +At that point it seemed things could go astray but I guess if the connection is getting reused then the characteristics are fixed anyway +Its important that the parent directory is or not +David please open a dedicated thread on the Sonar User mailing list because Im pretty sure thats your issue doesnt relate to this ticket +Manual testing showed that I accidentally lost my previous bugfix when I fixed the javadoc +better check the source tab to see whats going on +Hi DanCould you explain the problem that you are seeing when running the ant junit tests on jdk or JME? I have started up the tests via ant in a shell window whose JAVAHOME is set to jdk +You should assign more permgen the defaults are probably not enoughXXm XX I am running into exact same issue i increased the to M but still getting out of memory error which goes away on page refresh +Diff applied Fixed in svn r. +Thanks +Argh +Good but as does not support JNDI anymore the registration there has to be done by some extra piece of code +Should I make a backport jira for so that we can discuss the merits of this change and considerations separately +The issues are now closed. +Arg +I dont see this as a duplicate of +Closing for now please reopen if you can create a situation where this behavior is still observed in a recent version of Groovy. +Could you provide a Quickstart regarding this issue please +Feature implemented as part of +Hey Edward +Maxim any updates +They should be able to diverge going forward +The key field is absent in the new schema and is ignored in the data read and the rank field is not present so its default value is used +Thanks Freddy! Got it. +Finally the bug has been resolved invalidbq +Try cleaning Ivys cache of old Avro stuff +Hmm +We could encode the db name as a string in the for now and just make ql use this api instead of the new one +Ill update the documentation and close this issue now that it looks like were ramping up for a release +Service properties can only be attached to fieldsThe annotation can now target only fields +This patch will free the subscription processor and remove the memory leak created due to not freeing it +Is there an ETA on when the bug will be fixed +This can also be done in a separate jira +bq +LATER. +bq +Dhruba would you prefer a percentage or an integer number to wait for? Or perhaps either depending on the string formatIm also worried a little bit about orthogonality between this and the of blocks to wait for reports from feature +fixed replaced with embedded test in is using central removed +Shelly no newer GA version of is available right now +Leaving issue open until typechecking subtask is resolved +This patch has additional test scripts that do not use the nodefault lzo compression +Ill improve it in +This matches what I was seeing as keep hoping to have a couple solid hours to try to nail the root bug but didnt make it today +For the highlighter Christopher why are you going through the trouble of the custom collector to check that there are no hits? Why not just do a standard search +Effectively Hibernate shouldnt require any extra hibernate jars there The main contains all the modules now including the former Hibernate Annotations and Hibernate would appear that if you are building with Maven it is pulling in the within your ear and this is what is calling the error am currently experiencing this on tomcat dont think it is related to Spring directly though +I dont really think it is just a naming issue +The current approach to removing the dependency on is to switch references to Object and hard code some constant valuesWhile this approach is not unreasonable I do wonder whether this is the best approach +New patch incorporate the ideas outlined in my previous comments +The row returned by the server is actually serialized using the +Rebased +I did try enclosing the path in double quotes as well as single quotes in my test case and got the same resultDo you have more information on the suspected bug in? Ive done some googling with respect to and whitespace in the path but didnt find anything conclusive +Thanks Zhijie +Committed to trunk in revision +So if you need a different content length you will need to set it yourself +Also supposing that exceptions ARE indeed thrown correctly +The patch clearly wont apply to but maybe it should be marked as available for +Looks to be out of date. +Schedule fr rmi to branch for release in Rev +also note XX but this may only be relevant for route with n input queues and priority exclusive consumers +And those numbers are configurable +Ive just committed this +You need a jbossweb +I tried this with openJPA as well but it gave me a slew of different errors +For convenience I rename the base bundles to +But maybe we can just password protect it for internal builds +Another option is not to have a master every region server can do the work a master currently does +Well have to reassign if hes not able to spare the time +we use ranges and effective versioning processes to keep a consistent build hierarchy +To Arpits comment The delegation token is retrieved for the user who is using the WEBUI using hisher credentials +Im not convinced that the testResources addition is necessary here +Marking as blocker since this affects the public external api +Looks nice Hoss +These are all part of the testing +I was thinking about doing authentication that way +As I mentioned the example url above that can be access by anyone which is bad +When I make a jQuery ajax request the deferred fails with an error about callee not being available in strict modeFor now I have a part of the build script that changes the compiled js code from use strict to do not use strict though you could change it to any other string +that all depends on your transaction level +maximum time of getting connection is msregardless of each threads holds connection for msI put graphs that showed response time of borrowObject +Ive applied it + +I checked the correct string with all whitespaces and newlines and whatever you put on it is correctly parsed as a literal expression and the mixed syntax parsed as a compound expression giving the desired string out of a mixture of literals and property names up to the point when it needs to be on +but the first line checks for the wfs schema and imports it +Closing the bug. +Creating a test would be creating a big file about MB and testing itI have ran some tests in similar fashion +from me +fixed in svn a result of +Could a comitter review this? I doubt this breaks a core or contrib test the Hudson output doesnt give details +afaik this was done +example for luni lists for auth awt bean instrument niochar prefs print rmi sound swing and text modulesNote these lists are for windows and linux on x for drl and ibmShould be applied after luni module +Marking closed. +How would I go about isolating that further outside of writing stress tests for zookeeperIf this is indeed a zookeeper bug and you have some idea of the possible issues involved then perhaps you can suggest some additional instrumentation of zookeeper and I could run against a version with more instrumentation which might reveal somethingThanksbryan +The only part thats is that partialTriggers only fires when an event fires and value only fire when the value changes as perceived on the server not on the client +The wrapper class would expose the getter methods and internally call the legacy object it wraps to obtain event properties +Postponed +bq +I think the group id should remain unless you have ideas to turn this into something bigger like a of Plexus. +Can you commit please +More brittle than we would like but workable for now +Verified at +For example the digest part of the search results is generated according to the query of somebodys +I am infavour of Indexable being an indexing only construct staying in the index part of the core space +Thanks Aaron for the reviewNew version of the patch changes +Ive tried AJ to and they all exposed the same however did only reliable fix I can suggest for now thus is to use AJ it might be that other versions such as +An comment that it code can be used in eclipse has been added to eclipse bugzilla +As name implied it is designed to store large data specifically Object and other xml format file like EPR faultData believe this workaround is safe +Will need to take a closer look how this might have occurred with the latest +Ive seen this test fail in some other JIRA recently +Is that what we want to do with the Perl bindings? Like I said in my comment for Id rather distribute separate sources for the Perl bindings separately +If can cache lastseq in the header and provide it in the DB response that would be fantastic +Dont worry about checkstyle Ill take care of it +How about in libexcluded like the checkstyle jar? This directory is not on the normal classpath and isnt in subversion +Is that correct +Yes all my messages are in one file +I agree +That should work +looks good +created for the improvements to and ssl support +looking +xml formatIs this a change to Hudson? For developers running locally this seems unnecessary +Committed in r +Checked in a fix for this +reading and writing can then be done using ledgerhandle apiI havent yet finished my streaming api +If Andrey doesnt beat me on it I can do that in about a week +done we can fail for now +Thanks. +I will modifythe scripts to place the derby jars in either testbin or testtools I am awaiting feedbackfrom SarahTo apply this patch we would also need to remove the following fileinstantDB directoryPlease review the contents of the patch +Feel free to reopen otherwise. +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Maven does not support multiple activated profiles at the same time +And oh there is no test case for this class +Hmm sure +PLEASE upload your complete wsdlxsd +The trailing file extensions should be rewritten to use a +The fact that the test works for is probably because the encoding of the key differs +Just to make this clear The exception does not occur on the client side when sending the REQUEST +This results in URLs like the following paramampparamIs this something to do with this encoding fix or should I raise it as a separate JIRA +The cost of trying to micro optimize this setting is outweighed by the trouble it causes imho +Im working on this one +What if its a bad flag? Exit ? Or default to debug? +I am refactoring the Macro Screen Renderer so it doesnt require empty do nothing templates +We are considering adding something to the UI that would allow for this but nothing has been determined yet. +Test case and possible fix +A pom has been created +All files are ready to publish +Should I create a separate issue for the generated catch blocks that are swallowing exceptionsThanks for checkingtestingThe does do the hashcode thing in the decorate method +Hi Sebastianthat is indeed less than ideal +Fix was applied to the jdbc impl. + +Id still prefer to do have everything go through a single code path first and then we can evaluate optimization afterwards + for fixing my +Definitely agree that we should support Drools as a possible CBR option but I want to make sure we do this in the context of the existing CBR support in Camel routes +Tedsset up a sourceforge site for other Strutsy things that are not to be held bythe main treem it may have the perfect home there with visibility and CVSIt would also mean that others that have specific solutions to various littleproblems can have theirs along side also +Yes as the URL indicates a new TLD under the existing private repoSince updates to foundationofficers go to that is fine as well +Need something more elegant here +on the for setting max +More of the same error here pom now declares usage of Tycho +The nodes were never bounced or restarted and Ive seen the same behavior on and +Youre right to continue to accept the PI from parsers that dont report an error for it +I think the only problems created by the lack of error recovery will be less descriptive error messages in the Editor in some specific cases where the error recovery was reporting more descriptive errors before +I agree about the hardcoding but IMHO it would have been better to make the commit as a part of so that all changes related to that feature are nicely contained in one issue +Tests OK +I think for these two we just should document that it is ill advised +See +Did I miss something +if we went with implementing it we would also have to do the RPCNo that RPC has already been done the one that returned the. +Save KB of disk space? Dom classes do not have any special external dependencies one might want to get rid of +hung in build on Jenkins +Things check out after edits +Mockitos doing well no need to dig into this. +Thanks for the pointer SureshSeems like is already in +new patch looks good +Closed on user request as this is no longer an this fixed or should it be a wont fix or cannot reproduce +Awesome Thanks a lot! +Fixed in current JPOX CVS +It took me a while to figure out +Can you show your file +Please reschedule if necessary +now checks for null values before removing from its context map avoiding at runtime. +Im not sure it will copy the binding though +Thank you guys! I promise next time it will go smoother +This bug has been fixed in Xalan Java +In order to eliminate this need daemons based on Journal protocol that receives streaming edits from active namenode could be used +Please post the and test results so I can commit it together with +This is the main problem +Looks like some anomalous tests +Anyway now fixed +I am investigating it +So I think we can conclude that this is not thread safe and is a bug in Derby +This isolation level state is set correctly when isolation level is set using JDBC api +That was a import its the wrong +We need to add additional notification underneath +Sticking into Tomcat and calling httplocalhostTest gives me the above stack trace +out files so it will probably go stale quickly +Closed as duplicate of problem with running this test on JDK seems to be different than and occurs even when the test is run as the first test so reopenning this issue. +The feature enhancement issue is old +The instances can be container injected to keep the implementation to use configurable +Edison is this taken care of on latest cant reproduce this issue any more +You need to remove the repositories and pluginRepositories entries and make sure your project still builds when your local repository cache is empty +the netbeans integration is using maven embedder +In particular this will mess up the testWhy cant we change? If fact we did change it once already for checksum reasonsAlso this should not be a requirement of class +Hi KasunCould you please provide a bit more information on the exact issue? Steps to reproduce observations etcThanks apply the fix +Committed to trunk +All serialisation class resolving should be done using Paul do you have time to fix this before the release +Ray I think you meant to say that is the version where it is fixed rather than that is the version where it is broken? Like this. +Thanks Arpit +Regressions passed with the patch applied +Hi SidThanks for your comments +I could migrate my project from wssj to this and I made a quick test with this fixes it worksBut other tests are still to be doneCan this issue be marked as resolvedColm +I expect to have your patch in shortly after that +is a private detail +Ok +I have following directory for java source code lion com test contains and java source code with package test contains and java source code with pakage when compiler plugin compiles it does directory scanning and compiles only java code that are modified from previous build +There are two new configuration parameters +so one comment instead of seqDirectory vs seqFile why not just support glob paths? vectordumper s pathtomydirpartDo we need separate flags for if its a directory vs a file? Makes the usage messier IMO + +Seems to work fine +Thanks Thejas +So fix isnt as trivial as it was done in +Asserts will of course make things even clearer +I suspect that the evictor thread is simply not up and running quickly enough new threads sometimes take quite a while to get on the road at least on Windows and with JDK +Closing this +Almost done +If the fix is a simple one to prevent the NPE we should look at including this in the stream as well asBasically this is a error of web service runtime from +This was a critical race condition bug. +Theres a strong possibility wont exist at all in JSE +NicolasIts true that Solr does not support spatial tier search but this was due to the code in spatial lucene being so buggy +Thanks! This issue made my day +There are some fairly significant pieces of functionality missing from the old one as we are not actively developing that metastore +So CHANGES is ok +The client likes some of those new behaviors but doesnt like others +I think its okay to have these additional checks around but I dont quite understand the values of making these checks generic and configurableDo you think the new pattern will be good enough for both everyday uses and migration? If this is the case we can just tweak the pattern and claim victory +I cut my teeth on M migrating this plugin +The problem is that we may not always control what variable exists or not on the particular environment that will run the plugin +Specifically whats causing the mixed mode is that message headers that were originally sent w CRLF preserve the CRLF but new headers which the broker adds to the message only have LF +Outstanding issues raised by Pinaki will be addressed in future JIRAAlbert Lee. +Regarding the JAAS Context in the the plugin uses the connection provided by the plugin +Hi IainI cant reproduce this error it works fine for me +binhadoop job submit +Alex assigning it for your Please take a look at this and see if you are the right person to fix this +Weve got to get it to us files included included in our is still true as of rev m going to mark this as verified +I cant duplicate this problem on Derby it only seems to be a problem on +Should I be using Trunk instead +Id consider using the existing ignore for ignoring what is known to be a really optional policy which just has not been declared properly in WSDL as if it is strictly required to be understood +Id commit the fix to trunk and and decide once others have commented thoughts? commit it so our clover committed the quick fix for now Committed revision +This time before starting the balancer waiting for the replication set for the file to complete +Verified on ER +Maven doesnt do anything special to resolve a conflict between provided and test scope it just picks the scope from the nearest dependency. +It says that was added and lists the entire file with plus lines in the diff +I committed the patch to trunk +Yes i could not find in +Waiting on SME confirmation that this belongs in the documentation +This should be applied to both and master +Thanks +Where efficacious borrows functionality from JDT +committed to branch +An optimization later could be to keep this data on the sideIf we do put it in ZK shouldnt be an issue to do it synchronously it only needs to be done after a flush and would be an insignificant overhead compared to typical flush times +Proposed patch +Please review this patch that resolves this issue +I wouldnt want to have to implement the logic in each index implementation separately +As soon as the embedder patch is in the current codebase the SNAPSHOT nature of the jar can go away +Its missingacceptacceptnonblockbindconnectconnectnonblockgetaddrinfogethostbynamegetservbynamelistenrecvfromrecvfromnonblocksysacceptFeel free to mark this ticket as a dupe +Latest patch addressing some feedbackFew other things that were not mentioned here but I got in an email from Luke If for a metric no slope is configured default the slope to positive for Counters +will commit when tests pass +I reopened the SOA issue as it shouldnt have been closed +For Ive revised to cover primitives vs wrappers in both directions Juergen. +Is there a link we can point to for supported Operating SystemDatabases etc +Closing this bug down. +Let me make it perfectly clear that compatibility is a at the moment +Note that the patch is to trunk Im sort of assuming that you all do patch the trunk and then backport as appropriate +But I dont mind much whichever +I like the last option the best +I assume though if both downstream agents configured in the failover sink are unavailable youll have the same problem +Thanks Navis +I switched to using hselect where I do not have two columns and they look great +Im using and have the problem +Can you try removing the saaj +found +Have you looked at how much stuff uses this static reference and how difficult it would be to fix? Would be interesting to knowbq +See document for an example +I am using the file path deployment +Thanks Andy +This is a bit of a showstopper for development https proof of concepts +patch can be simplified in the future +Cool tip Koji! Ill have to remember that +The small job is logging to a file called medium +Just to keep the version harmonized +Release audit is due to +I have written some tests for this module but I cant find the new issue +Tested on Windows with rev this issue appears to be fixed it should be closed +Regardless I can figure this one out later because I do know for sure via ethereal and the server logs that the encoder and decoder are working +thanks +in fact this is a bug in the oracle driver the interface allows every object from type Clob but they expect a specific Clob +sry Si forte in alienas manus oberraverit hec peregrina epistola incertis ventis dimissa sed Deo commendata precamur ut ei reddatur cui soli destinata nec preripiat quisquam non sibi parata +So in that case how can it be a TCK violation to respect a JAXB +Could we come up with a list of precise steps to check to make sure this project is sane? The flow as described in the page about the simple project is no longer happening the quick fix dialog doesnt come up at all anymore +JBDS contains more things which are not under Tools the word on its own is not allowed anymore so that would be wrong too +Ill commit to and trunk since this wont affect stability in any way +is a string parameter that if set to will result in the URL path changing but all evidence of QT and QTPATHONLY being removed from the final parameters +is not needed it is a special case and is there because we did not have a separate build libs directory +bq +Core PSML documentation updated and new Declarative Menu document added for. +Once the problem with not working in HEAD is solved I think Im ready to OK this patch Ive got the framework running under the patched version of Castor without so far any difficulties k objects loaded and counting +Does Marketplace give the? Or will it M +I just committed this +Committed in revision +I wouldnt be opposed to inlining the content but I dont think its something that we should recommend +Patch f +very possible +I dont have much experience using Lists with GORM so I may not be doing things correctly +Would be happy to see other tests use the class it was very handy for this one. +Test for this issue +Just briefly looking at this one good thing to include would be some jmx metrics along with the bitmap index +Thanks +Thank you Boris + +Close it as a dupe +I just committed this +BRMS is unable to store two files with the same name since it uses the file name as resource name and file extension as resource type so the file name is the key in the repository +Thank you Lars +Running quickstart with mvn jettyrun I was able to reproduce the problem +Im only wondering why all the tests that use internally always worked +resolving pathexpression will be more difficult +Maybe? The code in renames before putting in metadata and only puts it in metadata if rename is sucessful. +is this a bug or enhancement +Good point Andrey +Thank you for your improvement + stack AndrewThanks for the patchBut I think the root cause is that you shouldnt use fromI just check the code it doesnt export package from the metadata header so I believe your customer bundle embed the right? If so you shouldnt use the embed way you should install all necessary bundles in container and import necessary package in your bundle otherwise you will encounter lots of issues which is caused by different classloaderFreeman +I also added a enum to to control how the scores from the same group across multiple shards should be merged +Ok I can write a plugin for this +Will check it for as wellAndrei +Because of the complicated layout of it was a little bit difficult to check the +duplicate of +Thanks for the review Jason I am fixing my fat fingered mistakes +then we end up with bunch life cycles to support mutliple packaking types +Removing from this issues fix versions since its now focused only on improving this is directed at new issues should be opened as needed in the Apollo Jira database. +Thanks! +As comments says will be resolved by bouncy castle. +It says it cant find benny which is the name of my remote machine +Fix Version NoneThe ticket is still open +Change application kill flow to kill the first and let notify that the app is killed instead of kill before killing +Please let me know if you are blocked on a fix for this issue +Thanks Ove for such a quick updateIm not sure what is the cause that hornetq resource adapter lost connection to acceptor +I dont have a persistence on it +pass configuration from getConfiguration method to modified to trunk in Thank you Jim! +Its kind of silly of a programmer to have forgotten that Here the version detailseclipse Eclipse Java EE IDE for Web Developers Indigo Service Release Build id groovy eclipse plugin Version Groovy version feature Im going to add a screenshot of the error Im gettingCheerss the screenshot +fix committed to trunk r +Should we set the workaround variable to true in the test settingsAnother option is to add a bit of code to apply some regexes to etc and enable by default if running RHEL or +Version of further development +incorporated the fix so closing the jira. +Checkin with following comments was using current compilation schemas collation if there is a parameter involved +So I have clean deploy javadocaggregate site sitedeploy +Modified the javadoc to mention both and changed oos to obj and ois to obj in both the receiver and the transmitter +we can then decide on where to house the changes +Right similar although Id like more something similar to access logs generated in logs +Committed revision in Nutchgora branch +So basically out of the box SAX is broken in both Java and JavaAs far as this issue is concerned we could Modify the build dependencies for the module that effected by this to explicitly use a SAX parser that doesnt exhibit this bug +I copied the changes to Tika in revision . +This is fixed now in SVN +suppose to match only? No itll match anything for the domain which would be legal +Current reducer case with map being retried +It seems that the versions of Remoting is JBM is compiled against It is run against In the wrong JAR file there is is this all coming from EAP FP then as J mentioned +Correction the shutdown tests are +So all issues not directly assigned to an AS release are being closed +So IMHO this goes more into Mocking and Stubbing +bq +We might be keeping things separate from dependency standpoint but we continue pushing requirements down on it +Merged into the master and branches. +Maybe its better to time insertion of a few million into a instead +We could do that but lets leave it out of the scope of the current jira +However if it was just an adhoc file posted to the list then the file would need to be updated and reposted +Fix for projects rpc spark and you Chema +Added a Semaphore to replace and extend synchronized block +I dont think this needs to be hooked to a particular version. +The now enforces that you dont overwrite dataWhich this test is doingIll provide a patch. +With new editlog changes could editlog size be! but still it has no file system operation entries? While loading editlogs wait for numEdits to go to and then throw an error +Could you the patch without the prefix a +However the point I want to make is that there is no advantage we gain by trying to put the extracted classes of patches into classes since for any patches to be applied later on that will anyway require manual intervention +Thank you +Attaching updated patch which includes the new license headers +Set true to enable the change +If you still have a problem on this issue please let me know thatThanksIas +Screen shots attached to show the issue +Fixed problem under r r and r. +If the vm being queried happens to be destroyed during the query +I tried to fix the NPE but could not come up with a fix +Jody just said on IRC hes not sure +I tried two version of SOAP SOAP is not invoked SOAP is invoked properlyTo summarize you have to run JBDSJBT with java and invoke SOAP to face this issue +We havent been able to narrow this one down since we started +I add here snippet to fix this cygwin path translation if cygwin then WORKDIRcygpath wp WORKDIR fi +Wont fix as naming collision can be prevented through the use of a binding file. +Meaning they will verify the dependencies are already installed but they will not go fetch them and wouldnt know how toIn order to solve your issue you can either Specify all the dependencies on the command line +We should guard with the is check +I was implementing a mount table file system in a package called viewFs +I will need to prototype this to confirm the suggested solution works as GAE JPA is not exactly predictable. +and replace it with the ThanksBoaz +Actually this happens when issuing the following commandmvn assemblyassembly DdescriptorsrcmainassemblyBut in this case the release profile is not activated and Maven uses the latest version of the assembly plugin not the one configured in +The guide versions updated according the release +I disagree strongly with having isolated toplevel components we should not enforce this as a constraint otherwise we will duplicate efforts +I dont care that the javadoc for one item says how it works and the other doesnt a change in functionality is a change in functionality +Doesnt this imply that the program should either be single threaded or it should use its own synchronization of operations between threads? If it is using its own synchronization then the above problem will not happenIf the client really was and desired that each thread should independently operate on HDFS it should turn off the object caching and use a different object for each thread should it not? Isnt two creates logged in an audit log instead of a single create not a bugIf the operation is defined as being idempotent then having two entries in the edits log is harmless is it not +Patch that adds the puppet executor to the agent +No response so changing version from to as I could never reproduce the bug report in the first place and the reporter indicated the problem was not with the ACL package +This adds an additional check for null on a files block list +Hello Please Can you solicit more information about this problem? Does it still a bug in? Best regards Younes of +Added a unit test for regression. +I will take screen shots of the error report today and add them as attachments +Although our branch is quite off the trunk so its not really mappingThere is still some tests failing although they seem to be older +Changing the packaging type to javascript as specified fix the proble. +Folks why dont we simply not catch the exception +It does make the code more readable +I have have updated the plugin to use Findbugs as wanted in a vote which requires Java all is fine I will call a vote later this +I was wondering why you chose to only have it in the root +Refer svn for the location to appy this patch in file +When Im logged in a llucifer I lose the view permission +However theres no API to be able to know where a +This requires a significantcomplex enhancement to javassist +Ive updated the grey version to create more visual separation +Perhaps this could be adjusted in your repository as wellecho offsetlocalset PROJECTDIRECTORYcdrem dp is location of current script under REALPATHdpset ROOHOMEREALPATH +The method that does not remove the annotation is called by the other method and should not be called by any other class +There are some known bugs in the latterIf it still fails well have to check our DOMDTM input stage it may beinsisting on a DOM with all the declarations +I have confirmed that all modified tests that now use are running successfully with the v patch +Thanks for opening this Andrew +Please confirm the current status of this issue and if its incorrect change it as required +We saw the on get but also found a much more severe effect of this bug objects placed in a variable in one session can be returned when evaluating from another session +Im thinking of the following adding two buttonssettings for the Sling console automatically activate on event automatically activate on error this is similar to what the classic console part does and also similar to the SVN console these should be both on by default for now to aid discoverabilityDo you think this would be enough for now +Got someone that can help trace whyif this an errorhow to fix +Thanks Hans fixed in rev +The test now runs cleanly for me on Java I plan to do more testing and put more effort into improving the buildtest experience on Java +Stanislaw can you try now and see if you can assign jira issues +on the implementation level +provided in the policy password configuration +Thanks Jason for the change I put this into trunk and. +The class was not tag with IO and included in release +It looks like a good thing that I didnt try to polish this into oblivion when the basic premise looks unacceptable. +Doh jira trying to be clever converted the comment char to a number +This also normalizes the path separators in the jardirectory locations so the duplicates are properly removed for +no problem +This patch contains fixes to the following tests that are failing whenrunning derbyall with jdk derbynetclientmats tools updated master file for jdk derbynetclientmats jdbcapi REMOVED jdk specific master file derbynetclientmats jdbcapi updated and to produce output that is the same when running with jdk and jdk updated master files to reflect changes in output derbynetmats tools updated master file for jdksvn status reportsM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTestsmasterjdkM javatestingorgapachederbyTestingfunctionTestsmasterD javatestingorgapachederbyTestingfunctionTestsmasterjdkM javatestingorgapachederbyTestingfunctionTestsmasterjdkM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterNote one master file has been removedI have run derbyall with jdk andThe patch is ready for review and commit +Contributing a fix for thisSome style sheet changes fixed padding pxadded to xapButtonand padding px px px pxchanged in xapThe buttons looked different because padding worked differently between the browsers +Thanks Kihwal +ThanksCan you please fix while you are at this we can do it in this patch itself +Patch for are currently evaluating the to be our stack for services +is this typical? They all pass in my client +added description of on added warning when user specified param is now supportedThanks Andzrej for your reviewing it +I attached a modified patch which includes a junit test +Thanks for the patch Harsh. +i also encounter this bug when i make check +Its imposible to the dialog if you check dont show this dialog +Honestly I would close itI guess almost noone is interested in PNM. +Attached patch to fully resolve the patch resyncedRemoves caja from example +Lets say when we add the entry we have a ratio of and when the entry is removed we will have the ratio of then we will leak some space for cache utilization +WDYT +What I thought was a long file name problem must have been another problem and does not occur any more +Resolved Conflict and upload updated patch +Otherwise what are we going to doWhat should we do for exchanges with more complicated routing algorithms like Headers or XML exchange where the subject is not a factor at all? It gets complicated real soon when you look beyond the simple casesIf you look at it from a high level pov we send messages to either a Queue or an Exchange +I found a similiar issue with references to holding on to objects and causes out of memory condition +this patch closes the streams on a shutdown of syncrequestprocessor +Hi Im currently in the process of testing changing minOccurs to for primitive schema types in axis whats the process for submitting a patch for this issue +thanks Amareshwari +Is this still an issue +My guess is that they are just talking about compatibility in terms of reproducing the same stream from the same seed +This should be part of seperate code jira +A task that can load scan tables +Also I forgot to mention there is over qfile tests starting with udf +Revision . +Merged in daee +zip top level directory naming should remain constant like IR are for our understanding and not actual releases +In the interim attaching a patch that should handle those two warningsThe test failures with the previous patch are unrelated +Maybe someone can answer this question before I have to create an example and find the answer by tryingLets assume we have a AAE with twoAAE AE with type system T AE with type system TThe type system of AAE is of course TT and the type system of the CAS within AE and AE is also TT right? What is the type system specified in the meta data of AE +Duplicate of +thanks +Patch with above delta included +In fact the TCP protocol has the same issue if using textline codec +I made the doc file +This will be a big help +Only plain text password can be transfered +Added missing +to the delta good catch Andy +Improvements and suggested config +One comment to the patchIf a MTQ subclass does something special during rewrite or in its and calls any other method of it throws NPE +We need to the external schema location properties in reset to rebuild such list regardless of whether the values of the properties changedVerified that the attached test case passes +I cant find this code +Do you still have access to the Jira? The xml export would be much easier to work with +Please let me know if i can help fix this issue +The node positioning problem is intermittent and has only happened a few times +It appears as though the server cannot operate in IPV mode any longers not so much as specifically using a raw address in the request URL host field +zip to discover jars in a specified directoryI was able to with an ANT script in add Jars to the maven classpath so that each Jar would not need to be referenced manually in the plans to support that kind of flexibilityIve been using Maven for years and I have to say I believe that while some things are easier to do now others have become more inflexible +could support an exclusive start key using the same binary zero trick that makes the endkey inclusive +However doing it this way would require another parameter to specify where to put taken pictures +but this has to be checked +This second commit would change the new paramter to be true by default +keeps the code that waits for termination of the old writer in stopWriter method of class +was already done and was not entirely applicable +common profile contains only to be included in HADOOPHOMElib +Thanks +Passed to Bill to confirm this fixes the problem he was seeing +Right you are +In fact Jonathan Fuerth and myself have been discussing implementing a full lifecycle API to allow the application to react to more fine grained framework states up to and including what youre asking for. +in +Any progress or plan for fixing this issuePackage level adapters are useful to implement default behaviorsThanks in advanceDavide +One instance of per class +I can pick up this Jira + +Things worked so it looks like this only affects remote after looking at an instance that was broken my conclusionassumption is that this is simply caused by JMX not being used to safely stop the scanning while deployment is simple restart of the war triggered the app to work thus most likely the deployment was incomplete on first scan +Thanks Nicholas for the review! Update the patch to address the comments +Hmmh +Context is reloaded every time some class or jar from dependecies is to the current trunk version of plugin +Did only a minor update to the patch to note some subtle differences between setExecutable on Windows and Unix +Patch that removes get calls from adds createApplication that returns adds getset to +Hany I am assigning both to you. +Closed as duplicate although arguably the duplicate is actually. +Thanks. +Bulk Close for all issues before incubation +removing the offending holding entries +Patch applied + script for script for you attach the changes in diff format against trunk? Thanks +in my opinion the issue states one of my biggest gripes with analysis this whole tokenstreamreusabletokenstream thingwe go to all this trouble to have a reusable api only for this analyzer problem to trip up usersmaybe its best to give s a chance and see if it clears up the confusion for usersbut if it doesnt in my opinion we should do a hard backwards break and make tokenstream reusable by default + patch coming looks good +Otherwise it gives an impression that reader should read this jira to understand what is happening which is not necessary +Can you attach a patch? Thanks. +This fixes the vti so that it can handle the new and the old timestamp formats +investigating was a issue works correctly with updated one rejecting as this not a bug +Fixed +The test failures are unrelated +Example of normal Windows usage CtrlAlt is an example with just F +I think this rules outExample of the lsof outputjava user r REG usrlocalhadoopvarloghistoryjava user r REG usrlocalhadoopvarloghistoryjava user r REG usrlocalhadoopvarloghistoryjava user r REG usrlocalhadoopvarloghistoryjava user r REG usrlocalhadoopvarloghistoryjava user r REG usrlocalhadoopvarloghistoryjava user r REG usrlocalhadoopvarloghistory +Do you think a broken build is the highest priority for the community to fixDefinitely yes especially for the one who has contributed to the broken build +Hhypersonic should not be relevant +This patch makes the release notes generator remove the generated jirasoap machinery after the release notes are successfully built +So if you see a INTEGERINSHORT read INTEGERINSHORT into an integer so you can compare with other integer type correctly +checksum mismatch is not even a reliable indicator of unclean shutdownPlease advice what can be a reliable indicator as I really think this is a critical scenario to detect +If it is intended to be included in we cant just push it to master +Also the leaderCount gauge in needs to use the lock too +Thats the power of namespaces +To try to clarify +I was just wondering how many flags would end up into it +including changes for the feedback above +Users guide for installing EDS +Apparently it was a mistake +what is the status on these TLDs +I hope it helps ciao Fabio +Your please check revision there you can create a any workeffortype within the calendar portletRegards fixed in rev . +The entry is now inaccessible as the index file points to a bad locationWe cant use the stock BookieaddEntry though as this adds to the index before the journal so the same issue could occur +do we have a jira to fix the unmanaged AM after this one yet? we should and it should be a blocker for as well +Fix applied to as of r +So what exactly would you like? Would you like us to remove our file copy over scpsftp and replace it with over management port? Or are you requesting UI changes? Or are you requesting full ability to use arbitrary management commands inside eclipse +Id also like to see added is in there or it should be +Added CHUKWALOGDIR as a build parameter +u need a test where src is a directory +This does beg the question as to what was done though +added classes and xsl file for conversion to html and pdf license header please copy it from other files in the Apache Syncope SVN repository +Closing all resolved tickets from or older +Thanks Jacques downloading GavinIt seems that these spaces have some kind of tries +Sim the query layer plugin is a single file that does not modify the contents of the WMS class +Due to the LRU cache replacement strategy of block cache the more RPC requests processed the more block cache size will be occupied so we make this implementation decisionStack would you please send me the issue number which is about the previous discussion of this function +After having a look at them I dont think the root of you problem is in the Release you get the same error if you run the command mvn deploy +tests litmus test suite and some manual tests using +Though I belief KB should be more than sufficient to generate exception and cleanly exit +This is a minor edge case +This also applies to Spring Security +Id be willing to revise Reimers patch for the current version of the plugin and resubmit but I cannot get a response from the current committers on whether they would incorporate it or even want fact that this patch has been sitting here for months is not a good omen that work put into it would be well is happening with this patch +Just sayGood stuff Andrew +So initially I think I might favor being more consistent I think it would be more awkward for the shard and collection aliasing to be so different than with the new schema api not being consistent depending on how long this takes to get inHowever I would like to change the current admin APIs to all work closer to what your suggesting now that you have gotten restlet inI was thinking to deal with back compat I could just make new apis and keep the old ones around but deprecate and remove for or somethingI know its a big deal to change the APIs but it would just be admin side not search and it would be a big step forward I think +This abortion is caused by the jenkine job timeout +An EAR is not web applicationSo the should be used for the web application and Object should be used elsewhere +Bulk close for the release +It is a tricky one because the lexer is created using +Perhaps if we ask every store file what it thinks the best candidate is then we sort the results by row and then scan all the store files from the first candidate to the last candidate noting any deletes as we go and then ultimately choosing the maximum row remaining in that list +This diagram incorporates all of Stacks suggestions to dateI think it is much cleaner but it is up to the community to decide +Ive just committed this +The zip file is megabytes +Scheduling for + +could you tell me the class in which the list gets filtered and returned? i would like to try to edit the said analog class of filter in combobox would be helpful. +Ive been pushing it to see if I understand the issues around concurrent merges +sh files? all new files need apache headers +Standardized the look of thanks for the feedback fixing it up question for reviewingis the best thing after this to post a patch of the complete diff or justdiff since your commentsBen +Shortly thereafter the requests start returning things correctly +If i wanted it slow when could it be available you thinkCheers least some weeks I think +But I guess this is the same change +Hmmm +I installed and have the same problem with EPSG +contains the Document its Validity the url the resolver and the expression cache +Reducing priority as you can always replace logj in with thislogj cdevCdevlogsCdev +Pete my writing English is slow +This patch fixes the problem with incorrect size calculation +Add a change on javatestingorgapachederbyTestingfunctionTeststestslang +Im resolving this issue +Verified. +Move bulk of performance items to target +This patch applies only to the trunk +Im happy to cover Tomcat and Infra although the more the merrier +Theres nothing wrong with not flushing if there is nothing to flush this is independent of context behavior +Recreated patch from r +However I have added better error messaging for when a EnumClass is not imported +on patchThanks Jon! Sorry Ive been a bad contributor lately +Since the Japnese situation is not a primary use case Im leaving this targeted to beta as fixed +Sorry about that Ill commit this one +ex +How do we handle this +Alex will confirm that I think +to both patches. +So now that you pass a reference to your Comparator implementation there everything is okayIf so I will close the Mark! +I meant the ones in HDFS +Can you try the one attachedThanksSarvesh +v interns for operations too +where are the docs +a fully working JDO configuration file etc +Try again since it passes the test in my machine +This bug seems to occure in again +NOTE I am going to submit a fix for this issue +Thanks Todd! +So the total superstep was times of algorithm iteration +I strongly suspect this is present in too looks like the same stacktrace +Then geoRSS can be rendered by +I really need this for some tests mainly because jetty andor the test framework can interrupt threads this can cause interruption exceptions and channel interruption exceptions that kill the close +Anyway the L and event listeners is already covered in a separate JIRA +Why old? Did Java deprecate inheritance? I dont think so +And please let me know if you signed the CLA otherwise it takes several days for me to receive notification for it after all the process +The patch seems not too big +So above patch is enough +It was taken out a while ago but I forgot to update the +No response +Glad I could help. +bulk defer of issues to +throw anIf the need is there we can still support this feature later. +confuses the debugger as well as +Fixed +Since that Ive tried to run offline sstablescrub it didnt help +Tested patch and Released +Yet to run the attached WSDLs and verify the functionality +we could do what NT I think it was used to do +Installing Required Packages which only mentions this might not even be required any more +Create table syntax is parsed and executed fine with Timestamp data type where Timestamp is internally converted into string type +Update to fix it +Thanks for your cooperation +This issue was migrated to +My query is concerning the and not the +For should be applied +Marking closed. +Please reviewcommit this patch +From the looks of it these lines might have been mistakenly commented out during revision +In this patch included fix for alsoPlease provide your commentssuggestions +Whether to update the HTML docu in the webapp is being discussed in the forum link. +Ive been trying to fix this issues but it seems Im running in circles +We discuss right now going back to +I just committed this thanks William +Committed revision Thanks for reviewing Yonik. +Thereafter I decided to complete assertions support for Simple Types first +Im closing this ticket with resolution Not a Bug +Issue in to print out results for LIST commands as thanks. +Any results Van +I think this will be one less thing we have to worry about for. +I will fix this problem +Logged In YES useridUpdate to and see if the problem persists +Well you get the point +We have REST clients with various requirements as wellA get together at Hadoop Summit sounds like a great idea! Ill see what Ican arrange +Also your reopen time will vary greatly depending on how many segments need to be reopened and what size they are +Back Encoding maybe difficult to implement now. +Perhaps having targeted separate files could make it even easier and I think youve got a good breakdown there but Im strongly opposed to having a separate file per Java source package as some suggest +I hope the history threads now makes sense ve also corrected the indentation to use will sign the committer agreement shortly and hope that youll get notified as soon as I submit the form +I added the license header to each of the three files changed all the tabs to spaces and moved the tests to the package under test +some lines are going beyond characters + Username rodelor +from me pending metastore testing by Paul +However a workaround or possible solution is to create an MDB for each group +so all the bodies are running im thinking background is not so bad after all +Here is the mapping file as wellAs the test shows I do not use the XML mapping but rather the annotations +Yep works +Found an issue with this change +I just made a patch +I think some generic interface and API for this is in order +Patch was reusing a class which unfortunately got moved to different package in the last few hours causing my previous patch to fail +David you are right +Timezones are now in GMT but they are just indicative as they depend on the local settings. +Once the cleanup task completes job is marked SUCCEDEDFAILED +This introduces a small delay before the file is actually removed but I dont consider this a big problemfixed in rev trunk rev previous comment for svn rev ids +One thing I noted is that the ant build throws an error because it cant find Morfologik libraries +Ok if nobody objects I will do that +Seems more like an improvement than a bug fix +is not required for this patch and will need modifications after this goes in +Here is the patch +Thank you. +What might cause sstables with unordered keys to be written in +Corrected the fix version to. +Beyond that I think Adams patch is the way forward for fixing andI still think we should consider rewriting this algorithm +Bulk move issues to and work! The spatial prefix trees will have even more awesome performance with all terms in RAM +Should have been resolved by commit dffadebcdfed made on the master not observed in the latest build hence closing the issue +does three things Makes a random subsample of a given vector set Estimated initial TT numbers from the maximum and minimum distances between vectors in the subsample Iteratively hunts for TT values will create a list of canopies within a given minimum and maximum sizeThe algorithm for the latter is lame but it does something useful in small and tests +Final patch for commit +Thanks JM +Actually fixing the normalizing based of ve values +Thanks James for the cue on the zip ant targetJim Im not building with Maven +I had sent the string without being nullterminated after being memset +Patched will be part of the logging cleanup in some time ago. +I think we can track the possible console feature as another JIRA as a nice to have +Committed patch to documentation trunk at revision +A couple of off the bat observations The Transactional as you have demonstrated is completely redundant in this case +Corrected patch added controller and views for experiment builder and can you please do a svn update and retake the patch? It conflicts after I committed Subhos patch +I would rather avoid magic in the method youd have to assume some conventions like empty constructors and then declare it in the javadoc to avoid issues with custom implemetations +Which mbean are you looking at where you were seeing STATE DEPLOYED +was added +Its very easy to implement by implementor. + +Would you care to post up a patch? Not that its hard for one of us to fix just that Im always looking for opportunities to lure in new contributors +Must decide where should be registered and how to invoke then +It also removes both and scripts from trunktest directory + if the Util class is intended to be used by multi packages including nio and io maybe it should not be in a internal package? I suppose is better place to go +the patch to try again +I could actually just inject my own bean via the as needed through any subclass of I define which would contain any Spring dependencies andor transactional advice I needed +I will fix this its similar to the code example in the book +Apparently I renamed the wrong file for upload +I dont think it would have the impact just curious why the made it into the a number of files just the title of the page changed but not the text in the page +Closing issue as version has been released on. +Fixes the problem. +Thanks Phani! +I dont see a clean way to fix this in the current implementation we could just hack it to do a deeper inquiry into the content of the files when the number of entries in the directory is small +I am going to fix this logging issue as part of a separate JIRA issue +One point raised by an earlier discussion is that we can have too many documents which would get out of date +I know windows is bad but i dont think is a problem of the terminal +I meant it would be good to create also a handler adapter wrapped around a configurable strategy some WSDL definition provider with default definition provider sending instances of beans from the context +Please note that only the test case test fails +Suggested fix +Revision +Id say we dont allow that now +So if you could check this point then it would become safer although I can not see any situations that may give a wrong String for this caseBest Regards Tsuyoshi +Updated files in zip for applying the patch +classpath and +thanks for your prompt replyregards +Thanks RamaI view this as a compatible change since there are no legal hostnames with hash. +otherwise it might end up trying creating ledger manager from to max ledger id due to exceptions +Anyone have a problem with that? all transaction wrapping code +Sorry that I didnt comment here +Updated patch to the current ve just committed this +The code looks like the Boo samples Rodrigo writes +Dont need to ship w this dependency +Thanks for the example! It turned out that its not the stack trace that was the problem but the display name of the test description +email sent directly to rossmason me +deliberately does not share connections between threads so the problem does not arise but in general of course that needs to be fixed +where did you get the download file for hibernate +Thus if a user knows that a given method signature can only accept simple multipart parameters and use Multipart as opposed to Form then this MBR would be registered and it would reliably convert the values. +Added patch for patch for patch from Daniel to all projects +Heres a patch that encodes both zero and one lengths +Theres still a lot of tests we need to get working but this should give you a good idea of the shape of the changes +Can you phrase it more like that in the documentationSince we moved to Sphinx the relevant file is in the git repository docreferenceconfiguration +Hi Craigwhat do you say to my reasoning for cases where required is either true or false depending on the value of another componentregardsThis change in code have break the disabled atribute behavior +The confusion was sown by me my original message on this topic contained the offending blooper +Writing test cases will require some refactoring of HOD code which is not currently in the scope +Shall we also limit the number of concurret writes on a data node +We decided to downgrade to to Committed will upgrade to Spring WS with +added testcase for the new attribute +I wont get to this until tomorrow or sat so feel free to augment the existing test for the code and attach that patch as wellThanks for the contribution +is that close to what you want +Perhaps we can close this issue for now I can reopen it if I can reproduce the issue +fixes a few bugs +Ahumm +I will add them later. +Attached an updated version of the patch +Hi Enrico there will be as much as Configuration objects +We absolutely need to do this +this happens in GK +I think they have a good structure now they just should not be at the core level but at the system levelbq +ant and I also observed the same someone fix this and make it the default version of the plugin? Because of this bug the surefire reports are rather unreadable when building multiple cannot beleive that this bug is still unfixed it basically makes surefire unusable with maven +I will be committing this patch to the branch please feel free to or file a new JIRA if you have review comments +What would happen if I change default font size in settings +patch +Fixed in rev +I think it is a very bad idea to make Avro depend on +Proper tests mean running a couple of instances and using key generators backed by the same db and table +Patch removes the credit card printouts in the log and masks the card number printouts on the receipts +worsened it for every moduleThe attached patch should workThe alternative is to set the location of in each module which wont be maintainable +but without an id sometimes people might have many similar tasks in their task list +needs to be ported to merged! +But the fact that the callers keep the old address in their immutable object can cause problems +I am closing this as wont fix since there seems not to be further discussion +I see what you mean about patch management +about that +I dont want us to have to call in our tests after we get a temp dirWe dont need to do that +In any cases last word is belonging to development community +Its set through the +Is there Fix Version for this issue? +The Incubator stuff currently says a grant is needed and the way of thinking seems to be that it shouldntAm talking to legal about getting this sorted +And the two line of comments are added manually +I think I figure why I had some many troubles with adding and +Linking these two issues. +Line shows up on Windows with Java depends on a fix in the Forge Java Parser to ensure that information about array types is not lost +Patch to fix this issue +And I removed because it was check this patch +and are ok when applied to the bean class via the spi they are only ignored on a interceptor +Please discuss with Jozef and configure all RF related selenium tests under Richfaces tab in Hudson +This is IMO NOT a bug in Wicket but something you didnt setup right +no worries we dont expect first time contributors to know everything +I worked with to configure specj for runs +I think this one is actually a dup of fix has been committed by to put correct version +The appended Z identifies this date as UTC but you claim here it is local time +Can you clarify? First of all the Keys Id value is null so that should indicate that its new and also if since the object hasnt been been retrieved from db then also its new right? Are you considering this a bug and providing me with a workaround or is this the final solution +The cause was comparison for tow +It worked without issuesEdision had implemented this with the above commit mentioned +Tested hsqldb mysql oracle and postgresql +Uwe I agree with your comments I just didnt pay lots of attention to it since I was rather interested in feedback for the ideaI actually think that we should move the get into the termsEnum though rather than using an attribute but for now that was easier to implement thoughbq +That sounds reasonable to me +Thanks +This is not a good idea until the webdav wagon is working properly +File attached +The attached patch fixes the issue for me though i strongly suspect that its not that simple +Ive uploaded an updated srcjavajdbcadapter for which can fix and +I thought Rajith was responsible for the broker component changes +I will start to get this right after a whileIs it alright if I put it under the groupId +bulk close of resolved issues as part of JIRA housekeeping for release. +Patch attached +Thereason is we can maintain a doc id mappinggenealogy as segmentsare merged +build build test runs on win and +Ted what is this issues progress now? Can we commit itIn deed when create table the hlog is no use completely +Jim once cxf jms transport is unlocked on integration we should have at least a testcase and some documentation +Thanks +Thanks Xuan! +Problem still happens in +We unfortunately did not build yet +The syntax is followingboldred Some textfgredbold Some text on red background +Isnt an absolute blocker supposed to use Blocker priority not Critical? o +What revision did you use to create + +Scanning from the beginning of the log segment defeats the purpose of incremental recovery +I think its some forrest quirk. +Hence Im demoting this issue to minor +This Jira issue is solely for the purpose of tracking inclusion of the fix the CP release +Ill take a look at this for all resolvedfixed issues of already released versions of Roller. +Im confused here +Youll need to run buildr package deploy the war to a web container and attach via JPDA to it +I just committed the patch please close if you are satisfied +So Im closing this issue +Is it ok to remove? +Applied +The current solution to ensure the end user specify the correlation expression aggregation strategy completion conditionsWorks the best as there are no surprises with defaults and what not +This should be closed as a duplicate not fixed. +Resolved in but the branch will make it in to the master in the future +The timeout that was reported is not related +A simpler strategy has been adopted +Fixed in current Axis trunkThanksIsuru +Its then much easier for us to propose improvements or alternatives. +Did so nowCommitted revision . +This is the expected behaviour. +In that case we should directly change itself but have to think about compatibility +In fact not having this breaks existing clients like Sling and eventually CQ +I am requesting that someone please review this patch +if you can make the latest snapshot to use i will give it spin to confirm if it just pushed out new snaps +This is now fixed in pdfboxs trunk +Thanks for the review Suresh! Ive incorporated your suggested change + +We only tested writing the values and forgot testing the read side of things +I thought that was for? I didnt even think was in use in the code +This may already be fixedThere were some changes in the FOP that improvedthe way it generated its output. +And so on for deeper nestings making things even more d definitely recommend a declaration with generics in the form of a parameter or field in any such +The fix in jffi is needed mostly for jaffl it uses the same info to do parameterreturn type I caught this jira issue update or it would have been missed +What about Collection? +The patch is on RB now +Ill keep you posted +So you would change the title of the page but create a new page with the old title which redirects to the new title +Ted? On your side +This will be done as part of instead. +Could you give an example of the processing you do on the output of the? With the current API you should be able to achieve the same result with a inside a +I change my adding style +Final patch including docs. +leave as is but add Javadoc convert to zero for consistency with parameters and the existing implementations +Committed to the CVS HEADOleg +Attaching new Y! fix patch with review comments incorporated +AS now says No message context found +Well see. +I changed the case of to throw a and I Modified to handle this change +Adding origin merged original patch and added new ones that are merged with the current origin branch +As discussed above. +It also requires a much higher memory usage which then can cause more garbage collections and such +In other code generation methods each of the in out argument types are generated unique for each message +This PR updates the labels mentioned as well as fixes one validation issue found with the JPA binding +Please my previous commentFeatures like this that allow a user to modify a servers files end up being used as exploits +Patch addressing Owens comments +As far as preventing duplicate tokens an iteration though the token queue with equals should do the trick no +I can take a look at adding this +according to heikos recent comments on it seems that implementing this might not fix outright +The shell script and the filesytem store are the two pieces that come into my mind immediately if I try to name some potential breaks. +A lot of improvements has been already done in but they are not ported yet to because since these two versions are in maintenance stage care must be taken to ensure do not break existing applications and most of the interest is in JSF +When the client code includes a header file does not select the right compiler include file so it ends up in the error section +As I mentioned in the previous comment there should be two lines but Acrobat Reader shows only one +It is what postinstall script leaves it +I plan to commit this in a day or two +Patch containing the fix for this bug +Alternatively we can force them to use different ports +I hope youll come to the same conclusion after your investigation +gradle dirPlus you should specifiy the properties values on command line so something like +addressed comments +website is done and live +It would be nice if someone could verify if the problem is gone +Users may overrideextend the SFOF if they want to utilize s new metadata methodsfeatures +Please commit it +I dont think it makes sense to keep this open any longer +get method should also be changed accordingly +defer all issues to. +One pops up you fix it but you just cant be sure there one be another one popping up somewhere elseThe patch is ok if it addresses the issue +Im now working on it and will upload patch for that +Again this is one Jesse might want to have a look at +Committed. +The CC machines stated the svn version was if that issue out as part of JIRA cleanup +I would be nice if that was fixed +Hopefully we can get this in +closed as dup +it is definitely nice to have client code in the server jar so that you always have a client to do problem determination +So Ill definitely need some help +bq +Merged into with revision missed +moving to for impementation time +e FROM bqt +Id love to get some help specifying what all its actually supposed to do and we can try to get it working then. +Thanks Claus +Those elaborations are both already present in the text of the +As for Seam do you have any forum references or other links for me? We might reach out to to fix this in Seam +and nothing is found +Fixed r +There is packaging issue for vmware +Committed a while back. +This eliminates a possible deadlock in the code +We can leave the ticket open and see how things develop +Attachment has been added with description Patch that should fix this agree with your patch and I have applied it +Committed +Committed at subversion revision +Committed thanks! +I ran into a few glitches with the tool which I patched up manually but I will try to get a smaller repro case for those and maybe see if I can fix the tool +It asks for the password +If this is removed there should no longer be a problem +Patch attached again to kick jenkins +Karsten you are completely misunderstanding the issue +Similar to how Serializations are pickedMaking an abstract is always advantageous +I am not planning to run hudson validation +jsfhtmlimageswebprojects +Thomas could you please list the test classes involved so I can be sure were talking about the same thing +Marking it forDoes this still apply given is replaced by a? move Weisss uppercase character entities are included in the new implementation committed in. +Can you update it to the latest revision? One general questions based on your description currently handles query DML and DDL +It may be the case as Im using solrcloud which is about weeks old +After recompiling everything running things again it appears like example case now works +Others want to try it out? We need documentation for it too obviously but see how it runs on other jobs? It should work on Hadoop too as this ticket comment thread indicates +Here it is common to have the coordinator participants and client on the same VM and this issue occurs frequently causing unexpected rollbacks +Any idea of when this will be implemented or might it be possible to have it already in +It might make sense to ask contributors to upload their patches when they include white space changesSeems reasonablebq +Thanks to all for the contributions +Not desirable as directly depends on List features. +When you say cases of interest what are you looking for? A list of the ESB elements that should be tracked as references +These could go in at any time but the sooner the better +if you consider any of the bugs important just let me know can try to help get them fixedLikely I wont be able to judge the severity of these bugs +New patch uses static as shared uses classnames as keys has a test the verifies that sharedData maps are in fact identical +This because creates the with a null ruleBase +Hi NicolasI tried it and looked at the Type de rle in httpslocalhostpartymgrcontrolmain it was not sorted +the +done in simpler to switch to CLQ since we dont actually need List API attached +ThanksI actually mean second patch looks good +Hey Kannan +Attached fix +Wait for queue to fall below messages and stop the producer +I dont remember the thing about the +Logged In YES useridI am using jboss in production +Thanks! +So we dont need the weak map anymore right +The patches seem to be somehow mixed up +Moving this out should we close this or can we get more details +I need to complete before rewriting this stuff +Checked in a small improvement dependencies are listed by scope and optional dependencies are flagged + AS is now handling all the web desciptors including and all files starting from version where this capability was removed +for feature requests +The key format is thread so the beginning of the array always matches between keys +No need to rescue this exception and this is very likely a design error by the application developer. +The test is no longer timing dependent at least I hope not +I do think we will use logj though because the async loggers sound very promising to me +libev isnt required for any supported platforms I dont think the extra workof following the instructions in a README would be burdensome if someonewants to use it for some reason +I can attach a patch later today +Tx for the review and the core changes Daryn! Tx for the test OmkarI just committed this to trunk and so as to unblock builds. +Which container type remote or embedded? Which versionUsing Container? Using Arq Core +I forgot to mention it although I still have it open in a browser tab +HelloWould you please try this patch? Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Ill set the breakpoint on Mac and see what I can see +Compared to enabling multiple pc resource enlistment this option ensures rollbacks will leave jBPM and EJB timer data in a consistent workaround has to be applied on the server instance started by Hudson +I simply tried the new in a cluster and looks like it works +This same issue affects the sample too +Looks like querying COLUMNBYTESDOWNLOADEDSOFAR in should help determine the progress +I have never seen this issue before +Patch with documentation update. + removed +to trunk and to. +can you remove the extends from the two objects? or is it there for a reasonIt is definitely possible to do away with the partition member but we can take care of that later +If those jars are in the lib directory they should be seen +The day just got away from me pushed to master wMicahs suggested fix. +Can it be turned on optionally Elliott? Backport w it offWe need the bulk load fixAny testing on a cluster? If not whod be up for trying it +Release Notes for Apache KarafThese notes correspond to Release Candidate will need updating in the event a rebuild is required +I do think we should break this apart into a separate exceptionI still would like to do this in an ideal world but I think our hands might be tied by the client community +As far as stuffing everything into the same index weve looked into that and determined that it isnt a real possibility because the size of the indexes theres quite a few ranging from a few MB to a few GB of data would make the merge process relatively expensive and coupled with the fact that the indexes themselves are built and maintained separately wed be needing to run the merging process too frequently for it to be feasible +Almost +Works fine with and without CDATA against! +Updated patch against trunk +Tested +Fixed I cloned the list in the parser it seemed better than cloning the list in the getter. +Can someone review this? Once thats done I can commit itthanks +I just check the default user pattern of HTTPFS it doesnt support numeric usernames either +Right now there is no type information available for keys and validationclass is optional for column values +See issue found patch written by to too. +As per Basic Profile Allowed Operations and Notification operations are not well defined by WSDL furthermore WSDL does not define bindings for themR A DESCRIPTION MUST NOT use and Notification type operations in a wsdlportType definitionMany Webservice stack like axis cxf etc doesnt support thisWe definitely need to add a check for the NPE the question is what should be the behaviour of ODE +tests includedThis is covered by existing tests +Committed the patch to trunk and branch +I cant reproduce this issue on booking example without groovy +perhaps i didnt actually commit it +This could greatly reduce the edit log size +Should the Compression class in wal packageI only see under wal package +Changes have been in and codelines for over a month now so Im marking the issue as closed. +Wait a minute Id argue that the default here should be false and we should change the codeAn example is that in Norwegian language the word and means duck the word or means alder and not means seine +It is now working as expected in regards to nil integer for your work Werner +any updates on this? if you are talking about this on the developer mail list can you provide us the link +How does the parser decide what order to swap? I would assume it defaults to left join when the left side is significantly larger but Im sure it wont cover all casesDo you think we should leave right join in there for now +The creation of metadata from classespackages is done +I say this because this feature is a complex solution to a not so complex problem +Change looks good should possibly unexclude the test against this JIRA too +Committed patch to the second commit on this JIRA +Hence the application would then solely rely on direct JTA synchronization which will work for Hibernate but only with proper JTA setup +Ah I didnt realize that the comments are from Alans code not from LoadStore Func +The default limit for form submissions is bytes with Jetty +No for your fist question +Ive made sure that all open bugs referenced by this bug have the flag set. +batch transition resolvedfixed bugs to closedfixed +This seems like one of those to meIf it happens again go ahead and this error with the username and install directory values that you used during the manual install process +Report errors as errors not the patch works fine +SaketAt this point there isnt a know bug in for this +this is a legitimate issue hence it +Thanks Devaraj +Any objections +Basically the first log file that should not be deleted is the lowest of first log needed for crash recovery or the log file that is yet to be copied to the backup +The airline sample which uses XWSS seems to run fine on JDK nowadays +As part of this fix we should also expand the default list to include piggybank we include the piggybank functions in the default import list we need to make sure that they are compiled and tested in the default build and that the releases will be blocked due to them not compiling etc +I have it working well with Java on a complex project +I have corrected this issues Release Notes issue and moved it from the Known Issues chapter to the Resolved Issues chapterThe amended Release Notes book revision has been committed and it will be available on docstage within a day. +Sounds good to me +I did try doing the bounds checks as the URL was parsed but it was even harder to match behaviour +I think that code was never committed +Bulk move of M to this not fixed +Can this jira be closed nowNo the patch uploaded by Sharad on this ticket still needs to be updated to the latest code and tested +We can resolve this and tackle more recent test failures. +OlivierThis driver could probably be split into two modules each of which supports webwork and struts want me to make this so? not keep both with the two optional dependencies? The Webwork bit is already optional + contains test source code and class bug was fixed somehowit is not reproducible now. +geir and brett added as moderators +Seems simple enough +All situations above seem to be solved +The client code is unchanged +this will be verified probably with EAP ER which will contain console +lack of consensus for committing patch reconsider this addition in the transition resolvedlater to closedlater if you have new information or a patch that addresses the issue you may reopen +This patch solve parallel execution issue and Clean up the yarn resource allocation + +Therefore I think the best approach is a simple for the reader +We should not allow itrename is a package private method that is called when we unalias a core from its name asking it to pick a new one from its aliasesHowever is there is no alias to that core the core has no way to name itself anymore but it may still be serving requests thus the null to indicate that caseId expect this to be rare but the case exists neverthelessPaulThe core always has a name if you use that name to point to another core well that name will point to something else +is already in mapreduce so not doing anything with it +ThanksI wont close that until having a similar patch for tomcat mojo +I just closed this and still offering some patches. +Unless we subclass Resolution internally and have each subclass implement just the code section of these that it needs +the compression codec is configurable on gzip or lzo plus some minor the review comments code rebasing on the latest trunk +XSLT is +Committed +No +If a maintainer will surface it can be easily switched back on later +Thanks for the comments Gianmarco +Im not sure that this is something that we can or should provide +Committed changes to trunk and. +Verified in JBDS B +Since another patch is needed here are the review comments I mentioned above +Im not sure if this should be discussed as another JIRA or can it be sorted out while working on this issue itself +Looks fine to me +set specifies a finalizer target that is called after the run of a Gant script no matter whether it completed successfully or not and even if an exception was thrown +I started seeing a hang in derbynet test +When I was going through the SQL Server document I didnt know why there was a link that takes you down linesYour call anyway now I understand why it is there +Updated to M +However the still fail on Felix since we are relying on require bundle functionality which is not yet mandates this option we hope to get this functionality in a future version of Felix. +Merged in relevant changes from branches into branches with revision. +We can check to see if the type of Node is Element before casting else throw +The changes like that of the output dir should be written somewhere so we dont forget to pass that information to those that will be updating their apps New shell compilation +OK this bug report is we also include this little fix in the upcoming RCI think this is quite safe and there should be no unexpected side effects +Icon obtained via from package attached icons look the same for me bad but caching may be going on on my why dont we just copy the favicon from the default config which is the one with colored balls to the prod and all config +Full it looks like we might want to change that method to a loop instead of our intervaltree or will it degenerate if nodes are added in the wrong order + +Error occurs also for complexTypes +Thanks James. +Reopening since needs to be applied to M as per our discussion on IRC +Attached namenode log +fair enough +Youre right about the default height I created the ticket on SQALE side +It also updates the tests to specify the expected exit value +Appliances are building now in F +Wayne thanks for the suggestions +There is value in running it all in one go IMHO +Thanks for catching that Daryn +Sorry looks like I forgot to hit enter with my comment last night +On the RM UI the state always represents the state of the application +The alert when trying to remove a server from the autodiscovery checked in to remove Javascript warning and fixed platform removal bug +Yes this might be a red herring +Updated patch based on review feedback +Look at the join package for instance its serving as a good example but the point was also that folks would use it but I dont see that happening frequently as people are preferring to write their own equivalents since its not highly reusable in all scenariosThe problem of a broadcast is simple send it to all partitions easy to define and perform on flagged keys +If you could test with trunk and let me know if your issue is resolved that would be helpful +Was this test behavior intentionalMost likely not +Currently in is translated into a full mapreduce job which include which is not the case in nested rank at least no +Sounds good +Its a full blown bean +Fixed +I believe this issue has been resolved in +Attaching a sample jaxws service which returns a String Array +Is there any way to get around this other than knowing to use the value attribute on the annotation when extending controllers from interfacesThanks! +Logged In YES useridThe cmp layer is not usable until the container reaches itsstart state +The physical operators do not have the equivalent of +Bulk defer all frontend issues to +Removed the references to the file extensions +So unless you can attach a complete failing test case this will be closed +This is probably a bug in that command +Ive seen this before +I cannot point you to the code base branch +Patch for ant to add support for WRITING entries GB per GNU on the ant patch by David Wartell I created aI didnt commit it till I got it tested if somebody has gig files around please help with testing +Is the the primary issue hereWould it help to have a configuration option for the where it could be explicitlypointed at a specific database and told use the credentials in THIS database for overall serverauthenticationThen the administrator of that could choose whether they want to designatesome database solely for Network Server credentials or whether they wanted to use one oftheir existing databases or whether they didnt want to authorize those actions at all +will not see a difference to the current dumps you only see the difference when you compare them one by one +Nothing found as of yet +Resolving as incomplete +Fixed Renamed SC database name from sfoundation to stratosfoundation Attached patch file applied the patch with revision afefbefdadedbThanks +This would be the complete patchmvn clean install wpswfsv is happy with good +I know that you can emit multiple keys per map but then you run the risk of your updates being done individually not in batches and hurting performance +Courier New is an ugly font +I think in this case the api was flawed and the only real way tofix is to extend +additional codes for benchmark and memory footprint measurement +Are the tests failing for you or is it just the hudson bot +tested +needs to read in the new +From Jiri via email the problem is lack of documentation +Classes and beans config file for spring +I opted for the lesser of two evils either a trivial fix to the file names with no changes to file contents or a full voting round for new releases simply to fix a typo in the version nameTheres precedent in Apache Commons for retroactively adding new artifacts to the Maven repository directory of an already published version +As does not allow a server with the same ipport number to join the cluster until the previous instance has been recovered this is not an issue for +I didnt know why I wrotebq +Made respective changes in for the same +Yes that works I had to manually add the dist project to Intellij because its not a module of the parent pom though but I guess thats ok if we provide some getting started documentation later on +Stack In my tree it is applied after +we have done if for the hibernate launch configs wiht the irenameparticipants etc + +Van could you findset up on one of our testing boxes a Schema that will return multiple tablesindexes where the PK is unnamed so I can use it to testfix this? fix treats unnamed Primary Keys by naming them PRIMARYKEY +Updated to add renamecolumnI cannot get testsmbmapjoin to repro so im hoping the failure was transient + +Thanks Stu +HiIve moved your patch to +yeah We couldshould default to enable current location instead of the hardwired project name +When is available uses it as the preferred transport so Id expect it to work and thatd be useful to know + Shinichi Kousuke and Haohui! for the patch +Good luck +VFS uses commons configuration +Can you provide some information? What version of Solr is this? What application server are you using? What version of the application server is thisThanks +I am closing the issue +so we can leave them there +applied to +Applied +This is why Im changing the fix version from to. +Thanks for picking this up from the discussion on the list! Its really better to have this here in the tracker +This last issue is reported in but Im not surethat it is related to this one +Looks like a regression introduced by +thrift file says without having to worry about different capitalising in the code generated for the different languages +BrentI agree with all of your points +Moving to please speak up if you want it earlierIf nobody shows up in support in Ill close + theres no way you tested this as I just found and fixed the problem +test case to reproduce the test started seeing the behavior today as well +David can either update that patch or provide a comparable one for comparisonHaving a good discussion is ideal for getting a really good implementation +In any case is it possible to take a look on this file? For example could you send it directly to me? +Will have to work on that on my sideI hope you do and great features and improvements to come out in the future +Patch added +Fixes issues in in revision +still requires a static the root of all evil It doesnt require one though? It supports a cache per segment reader just like this +Bug is also present in I am not sure that the patch will be taken into account I just reopened itSorry for my you please create the patch with the current trunk and using svn diff command +I have not tried a quick update this Exception doesnt appear for me now that I have switched my Maven dependency to version +I dont know I tried Long and It worked fine with Teleatlas huge +If no objections Ill commit this patch soon. +Patch added +If you agree perhaps you could convert it +The repro there fails with an NPE whether or not is applied in my codeline +fajax is a JSF tag not a Trinidad tag and we should not have to support it with respect to client validation in JSF for the following reasonsI would say that we should wait for JSF where hopefully the EG puts in API hooks to allow users to perform client side validation during a JSF ajax request +Linking related issues that I found +How about this patch which eliminates that ugly macro altogether +Id like set +No idea why the complexity was set to low +Attaching the +Attaching a patch with online help for cli expanded and included in the file srcresourcesorgapachecassandra Ive tried to incorporate the help available on the wiki and the existing conf +but i had also searched for spaces or wikispaces last year an got hits strange + The to access the localized areJapanesehttplocalhostlocalejaor httplocalhostlocale new section localization and added a note to end of installation about doc looks good can we get the added to the new section? THAT will finish the book. +This code could be cleaned up considerably if it were not important to try to clean up the xml streamWe could change this behavior if we wish to +The code change related to this one was checked in as part of. +fixed as per previous commit. +All Done thanks +I verified that external jars do work +Lets ask devtomcat +Moving out of patch is still under discussion +We decided to call it mvccVersion in the Cell interface and deprecate the memstoreTS name +Small test case to recreate this problem using only the javamail classes +For converters and validators there should be a dependency to myfaces commons instead to +Let me know what other tests you think are appropriate +Attached a patch containing patches with some work in progress +Hopefully will get updated to support both plain user and a principal nameAllen I talked to Devaraj about this as he has deeper context +what do you mean by crash? Is there a stack trace or exception or something? It would make it much easier to fix +One issue unrelated to this patch is that we have been just loosely following the code conventions we have set for the project mainly because they are too strict +It wouldnt take ages if I actually had some help will get help if we dont leave it to you +It is rather painful to use doesnt really feel like its been supported and might not even have the correct constructs +But be prepared that looking at the patch may cause headache because diffing a diff result is somewhat confusing +Final patch for review +The for each task was kept in ram until the Job completed but since we were running a lot of tasks the ran out of ram before the job could finish +If the other project isnt correct make the PR battle should be about that If the other project is correct and still faster we shouldnt compete by cheating we should optimize the correct path +Werner I have having this issue in the latest version of castor +Ive tested enough on my rack Hama clusterFixed! + Patch looks good to about right. +It appears the projects were not updated +Reopen if needed +Need to take a look at it +It wouldnt have helped with instance methods anyway. +I was looking at another TLP when I created the request and just copied they had an issues +It is posponed for now because doesnt implement unavailable Porltet Tag Library chapter wont be implemented for now as its imposible with current testing Testabletrue Section not testable as after propagating to Portlet Container code is return so there is no possibility to end test +Thanks allr +I can investigate this further once thats behind us +Removed from the release notes. +Also I am looking for example cases where best performance of single client sequential read off of a large cluster is very important + +The library does give me a WARN about missing file +Assigned Charlie for review +Put this on the need to verify stack for +Hi MathieuYes it sounds better like that +Worked around the issue by providing a simple Barrier class for the time beingWe need to refactor the clock implementation to support various special case clock with more efficient implementations +Ill fix that via +Hence you will no longer need to set the security context on the invocation you create in the ws layer +to make a run at this for jclouds is working on improving request limiting but if thats not ready in time Ill try to get some sort of staggering of instance creation. +It sort of complicates the code somewhat in my opinion +IBM has the feature pack which we installed on top of our ND version and worked very wellJoseph +Looks good documentation correction +Go figure +Hey folks just a status update Im waiting on a more substantive review before rolling up the javac warning and Brocks nit +bulk close for the release +Knut thanks for noticing that extra hyphen +Sounds reasonable to me and I agree that it is simpler than having the user create a type +I dont know about but Alejandro we made the changes to solve the issues for the upcoming release AFAIK +Factory is lazily loaded as a member variable of the +Abnormally some message appears at step +Work done for this issue Change condition operator with control statement Used getMessage method of class to show labelThanks Rishi +it now ironhand +htaccess file with a redirect +Compiler will just remove it if still not usedJust let me know whats the final choice and I will update the patch if required +Can you provide and patches as well +Ok now fixed +Do we need two classes and? Could we instead have just and then for each run method there would beI went back and forth between these two approaches +Even if using SUBST the classpath is still too long when is used for app development +I was out of town due to my personal work +Patch +bq +Patch available +Updated to trunk. +No longer needed as per earlier comments +You can do your methodOne and methodTwo examples since Spring even +Oh oops forgot about the blocker +All other uses of it like this work except when the custom tag is called page +You must encode the String if you want a byte stream +It will be nice +looks good to me +So it would be a normal behaviour that the consumer is waiting for the DB to recover +Good suggestion +Committed revision Thanks Ryan +In both the cases the tasktracker is informed about the failure +Sample error messageHTTP ERROR cannot be cast continuumsecurity think this has been fixed already +I have a complex multiproject build which isnt supported by Maven without this fix. +It looks good to me + Used constants everywhere to avoid passing wrong parameters +We can apply my work on top of the release for our distribution so it shouldnt block it but I do think it would be nice if this feature were upstream in the Apache releaseIve got some time blocked off to work on this if I get something working this week do you think it might be able to go into? thats fine +Ha for in in frameworkimageswebappimagesjqueryuicss I guess we need them for the calendar +Committed revision . +Wouldnt the straightforward impl be to put the suffixes of the suggestions into the FSTso for this is a test you also add is a test a testI feel like this could be done with just a tokenfilter used only at analyzingsuggester and would be more performant +passed with patch against hadoop on patch +Thats correct the documentation says root entity now +Ians patch sounds good to me +Thanks Boaz +As far as I know the patch provided added it into Marvin +my test is same build fail with but works withPreviously I was using fork compiler +Ill commit this patch to and and upgrade Jackson in trunk. +Confirmed for result is the same. +Would it make sense to use one for too? What do you think Simon +I close it if anyone has some exception please raise your hand. +Hm stupid jira +a task writes spills each with records totaling M recordsWith mergeing factor being the case does not second level merge while the case doesThe case requires x readwrites of the case +I suggest +We should not go down the path of supporting multiple callbacks or making them until we have clear requirements for other that would require such functionality +Duplicate of Release Note flags. +The patch does not apply any more most likely due to changes in +I believe it has the same functionalityHows that looking for an initial patch +Mircea is bean only used programmatically? Does it need properties +fixed now +What theme are you using? and what did you set the encoding to? The encoding must be for proper operation +Unfortunately if we dont start on one Im not sure if theres a way to detect that were in a wide row without making an extra rpc against the last row seen every time +The you describe does explicitly call remove so I wonder how this is related to the previous statementThis class in general should be used with a lot of care and only if you know exactly what you are doing which is probably not very convincing either +You are welcome to try to fix those http tests in itest so we can add them back later +the name alone is not enough because this way you dont get the compile time checks working +Not sure if this needs to be applied to trunk since the whole test mechanism is fix into BranchSPCP +Please review changes coming in RC currently in CVS head +Its got my ok to check in to trunk and +Consequently I propose to close this issue as WONTFIX + +Heres a preliminary patch +Ill do that today or tomorrow + +Tried applying this patch on current and all QJM tests pass +Verified +Closing this issue it is not exactly resolved but this is a duplicate of which is now reopened so well keep the track there. +classpath even though the dependencies are inherited from ssbaseapplet and ssbaseshared +Heres the patch +I can dupe it now +Nope if you give an absolute URL then it wont change that to relative URLThank yous try to fix this for next version I do not consider this a critical issue since we have a work around +SQL logging is turned on in the to show that only one INSERT is run against table B even though three objects should be persisted +Changes are as follows Reused the existing Replaced string configuration keys with constants Corrected the indentation Removed deprecated api usage and passed on my box +To verify I created a and with syntax errors +Removing the extra type resolved the problem +Again I see it as not core but instead optional for those that want to use it +Here the relevant piece from the IRC logkocolosk i believe all one needs to do is remove the spawnlink couchviewupdaterupdate line from couchviewgroupinitI just wanted to see if kocolosk was right and yes he wasThe attached patch makes sure that a GET request to dbdesigndocinfo never triggers an update of the view group + +and remotepicasa +Well use for further discussion of rd party bundles. + +For details about ALTER TABLE DROP COLUMN see. + +I assume the tez client submit layer does a dag verify +Would require some major changes in the feed generation code feed entries are always generated for directory feeds even comments +You can mitigate this as well by running your couchdb instance behind a reverse proxy + was the original issue filed +I added some logic to to ignore these system created threads and the tests seem to pass +Also it looks like does something fairly similar to this so its probably at least worth looking at +Apache Sling for example has the Apache prefix for all their bundles +Verified in Developer StudioVersion Build id Build date to fix bug in release notes +Unless the gremlins have come out again and changed the bits on me +I believe one difference between Jboss AS and EAP is that the EAP i tighter on security +Can you let me know what it is? Thanks +Could you please mark it as such and set the version fields appropriately? Thanks a lot. +Perhaps for its as easy as adding the last few methods? New methods will not get added in as thats an API break +I think they were something that Eclipse created when I imported the project +When right record doesnt match any left record exclude right recordA right is the same as a left with the inputs swapped +It now maintains a list of delegation tokens as opposed to a map from to the tokens +Ok Donenotifications no longer go to the commits list but the dev list instead + +the IE its not compatible with seam conversational state after exceptions this is fact. +Patch contains the write stuff for cUsigma if enabled +Ill ask the contributor on to add it +Can you explain why the annotated method is called for every requestIt seems to me like overhead +has fix for this issue? And did you intend to attach a patch hereThanks +There should be file in target +Closing this since the reported confirmed the fix. +Which brings me to the next point Im wondering if a layer is marked as non queriable what is going to happen if someone calls on it anyways +Improvments on crs mean that wkt rewriting is no longer needed. +Please let me know if there is anything information that I can assist with +Bulk update to fix version +The patch only touches capacity scheduler and nothing in core +HDFS seems to be doing the right thing for us +Jaroslav if you used some Deltaspike annotations in this example then you might see something like this +I applied some parts of this patch Im not sure about the changes to get in the class +which might be confusing +to look at it +Attached is now a mini project for you to experience this bug yourself +The attached test case should provide you with everything you need +Im aware of this behaviour with did you test this with the latest snapshots mojoThis should do exactly what you need! Please report back if this worked for is a dup of +Only expiry timeout should be considered regardless of reference count if I understand that setting correctly the doco is a bit vague +On Flush failure of entry log Keep all the entries in a list till the flush is called +Why is a final class? It contains some functionality I would liketo extend. +Both problem has been solved! +I will investigate +This attachment contains a new version with a fix for two bugs Plugin was attempting to generate ndoc for projects of packaging type for test code wasnt being provided the assembly of the main src code + +Youre not supposed to call a business method on any EJB singletons included before its method has completed +Question on the +Merged to. +If there are large nodes but for some reason they will never be available the RM now has to predict what never really means +Seems like there must be a way we can improve the check so that it really tests that the versions of the code is the same without requiring that the bits be built on exactly the same minute +Putting a breakpoint on line while debugging this test reveals weird behaviour +This implies however that you are using the script from within the wicket ui +I have resolved this issue as the requested change has been made to the guide and the revised document made available via. +As I am not using mylyn it might as well be +That could lead to a lot of useless communication +Ill push and create a PR asap +Thanks scheduling this for +Ive committed this to trunk and branch +Verified in build r. +Yes Dmitriy you said it +HiNote that our bug tracking has been moved to JIRA +ran just the times against trunk ibm sane classes and it failed in of the runsIt failed times in test and times in testIt failed times in network server testing the rest were in embedded modeThere is nothing from this error in I assume because it is a jdbc side error +touch +Great newsDo you know estimated date when this patch will be applied? +Per JAXB spec the schemaLocation is required so not accepting it is OKKind of a regression as this feature was specifically added for +patch to use perl regular expression to improve accuracy in finding latest snapshot +Also of course we must always set a start time before starting a taskSounds like things are very messy +Marking this blocker since Autoscale configuration fails completely from the UI inside a VPC +The problem was that wasnt used everywhere where an underscore could cause problems in Squeak or Pharo +Yes I had another little play on Saturday after it played up again with the different errorI wanted to keep an eye on it to see what was happening and to make sure it did not before marking this issue as doneIm glad it hasnt fell over since but theres no harm in leaving this open until the end of the week when Ill be happy to close it if no more recurrences +Is this the same as +Arup All the tests passed +reverted the non jms spec compliant part of the patch in found that the version of trunk is +Yuck +Unset HADOOPSECUREDNUSER from for deployment +WDYT +in rev . +If you have any logs please add them here +This change doesnt need tests because of only adding logging +its not a matter of whether or not clients should depend on this bug its a matter of whether or not they do +Stuart sorry reattached with correct dependencies page is opened by testappwarfaces +userservices are higher level calls that get info a user would need from ebay +I moved over to and committed it +The patch switches this so now the root queue defaults to everybody and all other queues default to nobody +Hi DimitrisI have recently been tasked to work on the JBOSS AS you assign this one to the relevant person or advise who to assign it toThanks as Im now on JBDS + +Extremely small sample app demonstrating the issue +Yes that works in the source setting the design mode on that works in the visual editor as well +the authentication check in was still in effectThat was intentional on my part the patch contained what I believed to be working code related to network server authentication so I committed it +Active pages are in any case kept in memory +update the parser to fail with an error if a user is specified without a realmCarl. +Fixed in heap objects during GC is a tricky business +war which you can get from seamAppeartarget +Ok thats fine +The source file is the downloaded groovy zip file +What happens if CLASSPATH is pushed intoIll add some test code here soon but I tried this early on and it didnt work for me +Patch applied +Patch to correct getSuffix +Do you mean declaring create protected instead of private +There we should also set the scope of the dependencies to provided +can you give me an example where this fails? Then I can add a test case for itThanks +will revert Hibernate back to Javassist +Planning to commit tomorrow +The problem is when running or other test runs involving prototype builds seconds many times is a large increase in overall test run time +Our problem is we do not control the clients so when they do not send a service version we have no control with to disable a re not ready to dip our toes into the confusing axis order issues with yet and weve found quite a few clients who actually break with +I believe it still is +Either way let me know and I get it started +Allowing the user to change the initial value globally would introduce a synchronization problems b a performance penalty if we have to guard the variable with a mutex c defeat compiler optimizations if the variable is no more a const But in any case if you really think that your project must use a different approach you are free to change the code and recompile Xerces in the way it suits you +Id like to get it included into MINA and then launch a vote for a release +If it wont be included then from a customers perspective this needs to be documented in Release Notes as a Known Issue not a Resolved Issue +So I closedI have a patch that implements closest in both dom and modify zone +M out of curiosity why would you want to use something from repository if there is a project in the workspace +output hudson +Also all the optionsheaders that is set after the processing should be updated on the revision +Was the configuration option added or not? Its not in and in the documentation +i think the data should go into only and not actually enter or the storesThat was my strategy put it in only and do it one key at a timeIll go back and try it again +Are you using +aka Jason failure still happens but its because the jbossweb context does not throw an exception it just marks the servlet as unavailable +Flume already uses the slfj APIWe include the jars in the distribution so maybe the answer to your question is to remove those jars and replace them with the logback jars or whatever it is you would prefer to useMore details about what you are trying to do and why you are trying to do it would help us help you +Im such a newbie I was actually unaware of the trunk checkstyle config Nice! Will use going forward +I guess we can resolve this issue if you can confirm that you dont have it on any more +The user who always was having problems is not having problems since upgrade to +By configuring only Local cases run will be good rightnow +However I dont think Vysper needs to support RSM everywhere to support it in the Retrieve Items +svn folders from ratcheck +Fixed in rStop generating init method and use its super bridge which has container class prefix as its replacement to avoid error for unintentional override of init without throws clause with init with throws clause +to update the summary with the actual version currently used +could somebody please verify that i did not break anything? thanks +We could look at the resolved property to determine this +tried before something similar to back port but thing is that the real back porting of the issue was rolled back in branch due to different logic +Not a problem Ive reverted the original commit. +Marlon applied this patch for Venkat +jar for includes +We have to fix these problems together with which is addressed for version +Certainly could be a training issue +LiyinDo you happen to have some performance data when this feature is used +here is the patch to fix with a testcase that will reproduce the bug in +The only spec I saw mentioned in that defect is +Thanks RupertIve just realised that the comment above the return is now out of place so that should probably be updated too +Existing behavior would remain unchanged +Tests dont need to be run +I should finish it on my way to this coming week +Im looking at this stuff too +Thanks OlivierIll try it. +Thanks +bq +Verified on SVN Revision +Agree already talked about this to Michael +I think the problem would arise hereLets say you have an order for to be paid by billing account by credit cardWhen you create an invoice for you wil have these records created Invoice for Payment payment for paymentId billing invoiceId amount This is all good but when you create a Payment to capture to the credit card you may end up with this Payment paymentCredit Card for paymentId invoiceId amount This would be right but if the also had billing then that would be counted as a use of the billing account +If buttons and links came from a factory and had the same signature you would only need one and then any extensions via inheritance wouldnt need duplicating +Gabriel yes you are right that the approach outside MR context will be faster but in MR we have few things that come into play like eg when we are using a reducer after group then MR will put sorting in place if we use it or not thats secondary but on reducer the output will be sorted alwaysI have created a version for min function that tries to use things from MR and following the same principle +bq +Otherwise Ill regenerate. +This issue is only observed fairly recently when testing theWe have been using the same tool the same configuration and the same set of machines for and it has never shown any sign of this issue. +In other words the result is exactly the same as in Java +Can you submit a test case or at least detail the configuration you usedIf is enabled we now also check the structure of the soap message. +It needs to be updated for Apache +Please review the same +I will give it a try +So you must have your module set up to use a JPA facet in order for the plugin to run on the classes in the module +And this issue will be resolved as soon as Mustang is released + +One work around would be to turn off validation not acceptable +Brian Minchau has applied the patch because although Ive just become a committer I dont have my SVN access working yet +fixed for version updated correctly +Patch for +Couple of questions on RB +I would also vote to convert this issue into a more general change preview behavior issue and move it over to Designer +You may need to reformat some of the content please do so as needed +I added push to doc writer Deletes flushed is gone only deletes exists In the apply merge deletes case wont we want to add deletes inthe get method? The test still fails so there is still somethingincorrect +Model project expiration was throwing just a blocked exception which was not picked up appropriately by the requestworkitem. +we need and core verify in ER build +This might solve the hanging problem not sure since I havent had this bug while running mvn test +Opened for that +Includes +Marking these as fixed against rather than against. +No IT +This is done +I dont want to run tests that all they do is test that java does what it says it is going to do or have the test call +Looks to me like this issue can now be closed +resolving this issue as the failures are related to the ongoing work on the attachment persistence. +This fix may break peoples script in cases when they were using and unwittingly violating the documented constraint +These listeners need to know about blueprint changes right then and there not at some arbitrary time after the fact +Verified fixedThere is now support for root node images and it appears to be working correctly. +Fixed in ARCHIVE module at repo revision rMisinterpreting return code from create new cache function. +Do you have some examples showing such a behaviour that we could have a look at +This patch doesnt compile after check into trunk + does not disconnect properly on undeploy only when using Apache APR +We can provide a null tuner to remove this as a factor when getting to the bottom of excessive thoughthanks Andrew i definitely like this feature and still if one production cluster has a lower enough th latency requirement considering gc factor a null tuner probably must be provided +Go into bin folder +Fixing javadoc warning +In particular it would be very nice to be able to have Seam manage application and session binding listeners +Resolved with up the test to make it a little more stable. +Thanks Scott! +RobI like your suggestion about transportsendpoints +I think this model scales as well +Connie yes I have found the same thing +Also note that full Undertow default charset support will be added to Wildfly soon so you will be able to configure the default charset +When the file was recovered the last few mutations were not found in the flushed WALExamining the NN logs I see the files block allocated followed by two fsyncs and seven minutes later a lease recoveryAfter that the sort file is createdHowever the commit on the file finishes some seconds after the log sort is completeTheres something Im not understanding about how the HDFS file recovery is supposed to work +You gotta check you route +Strange I fixed this a long time ago and I thought I resolved this bug +lib was removed from this archiveYou need the following depencies to make this not workIm using spring and jsp api from tomcan all required spring +not fixable in current architecture as stated in the previous comment +Username kosulin +Thanks Alejandro! +I am not clear if we have fixed the root cause or the symptomIt looks like checks for native impl being present and if not it falls back to Shell +Additional tests to show that was addressed in another fix +I support the general ideaYet I suggest we use the same encoding for string values as we do the configuration properties files supported by JCR InstallClarification JCR typing and Configuration Admin typing are overlapping +Because termseg drops log it was incredibly wasteful +Will fix the tests and description soon +transferring Bobs issues to Mandana +Duplicate use for doc as duplicate tracking comments in +Thank you Ted +Closing against where it NodeCell done. +is something that the MRAM will use when produing the container launch context that it sends to the NMSo we either need to update the AM strip out from and add it to LDLIBRARYPATH or just put in a documentation change +Do you mean you want the main PB class to beNaming is always tricky but I thought it best to stay away from inner class with same name as outer class. +The sorting wont be the issue though surely? The bug seems to be in the bounding box generation that Yonik pointed out +Thank you DennisBTW when will be release +Making a class that both implementations subclass makes more sense than subclassing leveledmanifest +I thought I saw a fix for this go flying by in the commit emails +Are you OK with visiting all potentially unsafe Application Master? I suppose I could have the cookie do an LRU thing and kick out the oldest appid if there is not enough room for a new oneIt is a bit more complex but I think I can make it work +This is wrong IMO hence the reason why I wanted to treat maven differently +Why it can not be openedI assume why we are handling internally with streams which are locking the object +Closing resolved issues. +Move to +Then the task should be reassigned on Y +Taste is probably the only thing that need worry about deprecating at this point +Renamed all processor to writer and removed +Bulk closing stale resolved issues +Fails how? This is completely surprising to me +Any objections to committing this? or was this uploaded as a patch and forgotten about +I realized that both were going in and I am glad that this is ready and has gone in +Can you try these queries with the pig branch and the property set to true +What is the quick workaround to avoid the error being loggedAn empty listeners property file placed somewhere +Looking further +Should we add a test case? A minor comment is that it would be nice if the patch were named after the jira +Click on More Actions button at top of this page then click on Add Files and then add your patch +oops sorry about the bogus subject on the initial submitcreate of the bug +is that as it should be or do we need to account for those in the ibiblio repo and this POM? I wanted to check before uploading incorrect information + issues resolved in released versions + just committed this +So bug or featureIm willing to invest a bit of time in finding out +I changed the priority +Thanks Suresh +We cant do anything with this until WTP gets back to me +It seems simple enough that its not worth reading from a file +This is a security flaw and we shouldnt release until its resolved +You need to find out if Eclipse can support the requirements in our checkstyle +Sure we just need the final decision regarding before going ahead with release. +Edward can you please check whether your problem still occurs with attached patch? I have to admit that I have not been trying to get things going against your test case but I am pretty sure that default behaviour has been restored with regards to code generator properties loading +Pull it back in if you think different +Yes there is traffic between nodes and its not benchmarkWhat about to gossip via UDP +Waiting for feedback +This patch removes SPECIFIC from the CREATE FUNCTION and CREATE PROCEDURE elements +Ok lacking better ideas I just added which can be used to change wrapper name +That works +Retain the format of the exception message +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Hi Matthiew I fixed this +If we end up putting this stuff in contribindex together with other tools then perhaps we should create for related functionality otherwise it would look messy +after this patch or before eacefdcdcdedce the issue didnt appear +oh +Somehow not compiling the entire source tree from scratch had passed despite the compilation error in +To answer the questions +Thank you for fixing this. + +v attached with this approach +Really cool +This reproduces the problem just run the ant file be sure to run the target after doing changes +I just committed this +Thanks for the patches Jacob +You can change this through specifying a session for your in you want to have your applying to multiple or even all of your definitions you need to define one per +Fixed that comparable +Hey Suresh For it is a public website start with numeric that can be resolved by DNS +Committed revision +Thanks Siyng +Commit the next step changes to trunk at +JackI dont completely understand the xbean jndi code however I think the reason for some of the things like remove are to support federated jndi contexts +We definitely dont want if row is live nothing to do here behavior otherwise well never purge tombstones without a full row deletion +Irony has it that the encoding of the attached tests was incorrect +A workaround for this issue is to map the dispatch filter to +I agree this would be a very nice feature +I took a wild guess based on the apparent structure of the messages being sent by the and tried wrapping the message in a second and voila it appeared to be logged correctlyMaybe I just got lucky but if somebody can verify that compressed messages do in fact have to be wrapped twice for some reason were good to goThere is still a bug in that not adding the second layer doesnt produce any errors +for the changes + +Hi It seems this has been fixed as part of +Change code to let exceptions from audit loggers propagate to the caller methods +Because this may involve collaboration the threads must syncronize here +ThanksI made a snapshot build of branch and now it doesnt fail. +I am not sure whether this makes sense +Not sure if by delimited they mean quoted but coincidentally upcoming supports optional quoting of identifiers +Ok It seems to be working ok without the JVM internal error +Could you please attach the file to reproduce +Fixed on revisions. +Nice work thanks PatrickIve committed this. +applied. +Its working in my case but It surely need some more check to be sure it works in all cases. +However I think we should add a test case to the ee test suite +Just finished committing the patch with revision The derbyall suite ran fine +Thanks a lot David +Ive done that because this has a fewadvantages it avoids a bit of code duplication there is no needs for anychange in the few parts of the code where there is an explicit cast of an to a Column and in principle it doesnt seems completely crazy tohave inherit from ColumnBut it has the drawback that inherits of the isfield of Column but never uses itDid you look into extracting the common parts into that both C and EC could inherit from +for the updated patch +assuming we dont allow a crazy combination of and or suchHeh thats what I was trying to validate in v +Remove the debug statements Add the memory threshold to group by desc +Upgrades the JPA provider to version and updates the JPA configuration to make it compatible with this +Anyways not sure keep native is the proper issue is keep native in vector land does not change one bit the CRS definition and you get the same WKT as the native definition +Thanks Bilung! +Thanks for reporting this errors have been tightened +I dont see a problem here +A workaround for your project in continuum is to add the correct scm url in your webapp so its a maven problem i move this issue to MNG appending of the artifact ID to the SCM URL is by is to explictly declare the correct configuration in the child is another open bug for improving this in +Temporarily assign yourself and add a comment that you are working on it +The fix for activemq is in trunk and will go into problem with deleting the temp queue in can probably be fixed by the team +Any idea when this might be looked at in any detail +Wow how did I miss that is attached to Gadget all this time +HeyI know Im reviving old thread butI have the same problem but using server stubs generated by CXFI have no +Could interested people please test instead of and let me know if any errors still occur +The instance configuration then receives an empty dictionary for this property +But it will still be an absolute internal value with its own problems very hard to tune it so that it works well for general purposeLong term and should be solved allowing us to compute a more reliable measure of how certain a classification is +do you consider the optimizations necessary for release? Or are there other things +That looks like a good approach to meThough should we make the new getDirectory that takes the Class directoryImpl arg public so that external subclasses of could also create their own instances +Any updates on this +I shall commit this command renamed to for still need to be updated +The thread sleeps seconds by default +Ill check ASAP +This change was made to facilitate locking of the namenode data structures +But please note that an ICLA is not an assignment of copyright to Apache +patch includes test and fix for this +This should be fixed in tonights RC snapshot + +What is the status of this issue please? +This is not a major issue as we have control over the but its a little nastyI will do some brainstorming and see if I can come up with any new ideas +Installing by tarball should be a fireable offense in +Yes +You can of course subclass +services and classloader corrections were made in both master and. +This is a very good thing to fix but it is neither a bug or an incompatibility with +Hi CamilleThis patch is updated to trunkDo you know whether the RB issues were solved +Last check with rev +thanks +Change to correctly calculate the number of pages using ceiling rather than you provide an example to reproduce the issue? +Ill look them up and comment after +Anyway I think that messaging system should not loose messages in any case +bulk assign all unscheduled issues to next release +Ill take a look to see if I can find the old dont support a scenario in where we dont have the same set of on every node +Patch supplied by the patch +Thanks for the patch! +Same fix should be applied to +Are you for sure using DR have you tried doing a fresh install? For me it works I guess this is not a problem with webbrowser I tried with Firefox and Chrome +We dont suppress any exceptions in +Using or XML for all the documentation allows for putting the material in the branch +OR that adequate warning in the javadoc to inform the developer that the implementation does NOT tolerate bad behaviour +If you want to integrate that with the current classes all you need to do is to implement a partial which resolves ordinals to using the global ord map? Or actually make that TR the entity thats responsible to manage to global ordinal map so that opens the new segments and updates the global mapThat sounds greatbq + +AlefWere working toward M right now and while this issue probably wont make it in it would be nice to get a testcase so we can repro and triage appropriately +do you after hit the back button +Ill update on the jira in some time +So we can use different objects for different shuffle inputs when Tez supports input tagging +r on trunk +we went back and forth on this for a long time and eventually compromised on this since this was the most often used version +When the modified version of this class is included in the app the no longer appear +Test again +Very soon Ill post further details here +This was actually fixed in CP and CP but due to incorrect Fix Version labeling was missed +Thanks for reporting crocket! +Thanks! A long awaited change +hudson +bulk move from M to I have resolved this issue in version already +I will file a new JIRA for the lack of heeding close in our internal streams and add the corrected repro from to the test as well and spin a new rev of my patch +So whats the problem? Just implement Comparable interface or even better dont use such logic inside JSP which is bad and erroneous +Thanks for patch v +Oh my bad +Sorry I had the wrong version in my Maven pom +A grep through the directory shows grep r junit +Note that the additional interface is not required to be implemented by a handler it is only for a handler that has some complex information to keep track of +many thanks +The easiest solution IMO would be to mark collector as transient and add error messages for remote callers or create custom error classes that are detached from the compiler I dont care much about detailed information on the remote end as long as meaningful error messages can be generated and I dont have to wrap the exception +In ONE CONTEXT etc are used as methods of payments +it would be better to consider how to consolidate the changes already made in +Does anyone anticipate having time to review this patch proposal +We will be publishing the Apache Thrift version during our next release +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +This is part of Target Platforms and that are used in order to build JBT and JBDS +Alex can you review this change please +how do you estimate which index is likely to be the most selectiveWell we can compute the range of index partitions wed need to scan and we have stats on average so I think that gives us the we need for highest +Please let me know if I am not supposed to close tickets I open in error +You can do that from gdb with the command thread apply all bt +This patch also includes support for the Map interface and not just Iterable +Latest version of patch looks goodNext we need to pass all unit tests +Unit tests fail for me +Remaining issues has been fixed for CP and is as such reported in the CP release notes +We need to go through all of these areas and make sure they cleanup properly and do not leave unsuspecting hard links +Tested it manually and it works +There should only be the lazy cores issue other this this and custom hashing +the crazy code is funny to read understand +Sending Sending srcmainjavaorgapacheservicemixwsnSending srcmainjavaorgapacheservicemixwsnclientSending srcmainjavaorgapacheservicemixwsnclientSending srcmainjavaorgapacheservicemixwsncomponentSending srcmainjavaorgapacheservicemixwsnjmsSending srcmainjavaorgapacheservicemixwsnspringAdding srcmainresourcesorgapacheservicemixwsnAdding srcmainresourcesorgapacheservicemixwsnSending srcmainresourcesorgapacheservicemixwsnSending srctestjavaorgapacheservicemixwsncomponentTransmitting file dataCommitted revision Partial support for on the service +Fix verified in ER build. +Ive just committed this +Should be better nowIf you try the test WITHOUT the patch on the test is supposed to fail since the CXF default Fault generator does not parse the annotations the thrown Fault is a generic fault and the CXF client does not recognize it +Refactored assign +Any suggests what should be done in +Can we put discussions in the forums and reserve JIRA for conclusions +ant +practically all of the impls extend today anyway +I guess youre working on some other trunk +SiYou have a good point +Ive committed this +This would allow the flag to propagate transparently to subprocessesAnyone have thoughts on why this would be bad? I can think of one adding flags to subprocesses without users explicitly doing so may be unexpected they might be surprised if subprocess runs with flags they only passed to parent process +Use this in and further add other methods required for fencing and syncing over +I think this is now resolved right? Please reopen if not +Is it possible to backport this commit too SMX or is it too big andor risky to do so +Weve the CLI now please report new issues for bugs in that CLIFixed in commit on the account that the feature request to create an interactive autocompletion cli cloudmonkey was already achieved +There is no ETA for the GA release and it seems not many people are working on itI hope this situation will change soon +But it will not be part of my current implementationregd MB default size It need not be tailored so that it takes one hour to reach the limit +If you miss an update it wont ever get added to the cached row but the update itself will be flushed at some point and thus not be in any memtable anymoreVery trueHow about adopting the strategy we apply with Writers dont update the cache at all Readers merge cache with memtables Upon flush merge memtables with had a look at it and maybe I got it wrong but returns null for a sentinel and calls thisIsnt update supposed to add changes to the sentinel so that cacheRow can detect the race +Please see that +Cancelling old patch +I can take it if you want Chris unless you want to do the honors +I have uploaded a new test case with new instructions +I think I got everything in there +Thanks Carsten for your help. +One link has the type applicationatomxml the other one has the type applicationcmistreexml +Jim Thanks for investigating this +Weve provided a workaround for expressing void but there is no workaround for expressing the action on empty messages. +It is in the old design of cxf which means you cannot called continuation suspend method before you setup the callback when continuation is readyBefore I find a way to enhance the feature from CXF side we should not support to use anymore +The response has to be unambgious if we have rejected the enhancement a reason needs to be provided if it is accepted then a release version has to be commited to +It is a simple matter +I think Id prefer to deprecate it +be fixed +The problem is due to component bindings and session scope backing beans +Moved to please move bugs back to which you will work on in the next weeks +When I was looking for the components on the project page was not able to get them think they are available now so going to close this. +Follow the steps in how to and in step choose +hashCode and equals of which are immutable +I updated my entire sandbox treeyou have to update jackrabbit trunk as wellfirst we got rid of some unused methods second the fix for this issue affects the project and not the spidav sandbox project +Happy to reapply if we can make the test pass +Sorry mistake in attachment +steps should be smoothen for all moviesMax suggestedThe one I did with Seam has additional seconds put in +Thanks Ari. +Does the same thing apply for too +I have incorporated changes suggested by you +Per our discussion a while back we decided to some unit tests directly against +Nonetheless thanks againJames Mitchell +Missing license headers were added to components +All pages will be included in and will be filtered by the list of registered portlets +hadoophive +I added the userctx as well for the replicaiton document +Hi OliverI agree with you this was the main question on the forum +run java with key attached to run java on sorry run use +James You seem to know how +It uses cvs update instead of cvs statusCould be a side effect of +Sorry forgot the most important bit what version of UIMA + +If you arent using the tranql wrappers but some other wrapper this might not have been implemented correctly but with tranql Im pretty sure its working well +richprogressBar does now show any progress +META +Now I have richfaces version +I reattached the patch should work with eclipse committed the patch and found some other typos on the waythese are still in the code Applicationget object for working getting storing session model will be called detach on when the component is detached when the wrap model javadoc seems to be copied from another flag ComponentFLAGINHERITABLEMODEL Flag for escaping HTML in model strings +Fixed svn rev +Duplicates dont believe this is strictly a duplicate of as this JIRA refers to the execution of the unit and interop test currently done using ant and hudson whereas refers to automation of the crash recovery test currently only executable by hand +at the level of CI job +Brian and David Thanks for the informationA couple of comments +Assign to Stan to look there hasnt been any response to my last comment Ill consider it resolved as Out of Date +bulk defer minor issues to +Sorry for the additional workThanks in advance as fixed +They will be much easier to configure and fetch from environment. +OhIn also changed to +Done r +This solution has the severe drawback that it only works if the is used in mode +Yes Bruno +Improved existing deserializer as suggested +Adds a default value to params in lookup and authorize methods in and is checked in r on trunk +jar extension and thus is not included in the list of jars +The side of the equation needs to be clarified then +I dont think synchronizing the method will really work either +I think this is important enough to fix for if we are going to release it +Here is a new version of the patch +Fixed in rRajika +v takes a different approach that Sylvain kindly suggested and changes the bootstrap flag from a boolean to a value so that we can detect previous attempts that failed and try again +Yes but couldnt it be like that we catch the xerces exception and if its this kind of error parse it to find the name of missing bean and look for the references for this bean in the file and add markers on these lines? +Fixed +assigning to Michael for now since he owns the parent will be fixed as part of. +Operation ported as of revision +Ok I can confirmed this behaviourInvestigating why this happend +good to me +Patch question on the patch +Support for richcolumns is covered by +This can easily be changed if there is a special reason for using +Thanks +Please have a look +Fixed version of above in attachment +Yet test fails +If it did please say so on this issue and alert me +version must have been renamed to after it was pushed back or something +Hiro Asari Thank you for clarificationMy fault I do not know my way around the debugger and did not understand that it would break on every exception and I simply had to continue until the program is run +trivial patch +After restart of Tomcat it works again +So all issues not directly assigned to an AS release are being closed +I would be very happy to provide it +Disabled formatting in Description and set Severity. +I have tried it in the function works. +Source and configuration using debug is real a bug I also get it +The fix would filter out content written by the JSP engine until XML document is started +For the Java application a Java is used for the createdTimestamp and the eventTimestamp which is a wrapper class for which encompasses timezone information +Not exactly clustered approach +We should probably make hasToken private since really that is what the Gossiper uses internally to determine if it needs to evict a dead fat client +The job that was consistently failing before under this setup now runs to completion +Jason Java does not allow Override on interfaces +If will not be in will do the following change this issues Fix release to update root pom in branch to not include update root pom in branch to not include remove resources from additional add an exclude line for to However if going to include in need the following remove the exclude line for from uncomment the include line for from add an include line for to add a fileSet element for to add a fileSet element for to +xalan on the cocoon include the fix to this bug +Thanks for the suggestion but there are two problems with that approach +okay in that case closing. +Ups sorry wrong conclusion current trunk seems to be OK +Thanks +Fixes for both +Although even after this patch we still have mentions of cloudera in testsite and a couple of other places +clean up testeepigharness so that it doesnt make those assumptions +Committed +Upgraded to Hibernate and removed exclusions +Hi Sylvainthanks for spotting this +it is pretty clear in the spec as to what the name of the Exception is to beAgreed no reason to complicate things further. +i am undecided +Editing the summary to say as were successfully using it with +Theres no need to share the cache between converter instances and therefore we do not need a static using weak references for keys and values +The logj dependency was added at the end of the classpath after the JBL changes hence the file was ignored +Feel free to create a patch or contact sales +That said if you want a stateful endpoint +And then the support of classpath container in the java launcher is very poor +I have reviewed the schema and its true that the definition is +Updated to include methods that were missing +push out to +I dont see why youd want to not just ignore the idle ones and only focus on consumers that have not ackd outstanding messages in the allocated time +Jeremy could you possibly describe the nature of the tests that are failing in more detail so that I can try to reproduce the problem? The tests I have now seem to work but that probably just means theyre incomplete +The problem was slightly works inlocalhostde you dump the path you will see somthing likedegem you will seede XX are corrupted umlautsBrowser was not changed server was simply updated to latest written before i cannot reproduce this issue i have tested with different browsers with no problems +Originally I wanted to simply extend Member with the additional fields specified and also support full CRUD making it easily recreatable via ToolsJBDS jQuery Mobile palette browsersim +Attached showing commitlog replay +Of course these numbers differ from cluster to cluster but I believe it shows the possibility of some savings +In and projects we would like to use a separate Eclipse output directory but at the moment its not possible due to this very issue as we have test resources that do overwrite main resources with the same name +Thanks +msec buffer +The source files of the +Resolved see r +It does not handle the case where additional options are passed +It is ok to include it in the beta +I see this problem in release also +The patch looks good to me now however you might have to update the patch against the latest trunk +We think everyone can use their common sense +Committed to trunk and branch +That is wierd! I hope you didnt spend too much time on itThanksPeter +But prior to this i can check and tell you the behaviour +This requires changes at wiki as well +Committed Bryans improvements +Closing it. +I cant reproduce this +v retains the code to make HH more robust against causing coordinator +I second this notion +Next time remember to name the jar file inside the bundle +I know in logj added more operations to deal with argument so I can post a patch to include the feature ofIm open to any other opinions +I think this was solved as part of +I have a patch that is partial +Yeah i resolved it yesterday +Patch which will apply for both as well as trunk +If we only run the MR code as a local hadoop job this might be the case because its the natural order in which the weightedOccurrences are mapped out +May be resurrected at a later date with a more rich grammar driven NL. +The patch tries to cast the Java Principal to and if successful it extracts the username which is the short principalIve tested this in a deployed setup with Kerberos and it works fine +However the next openofficerelease will be published in distopenofficeAny idea yet when the next openoffice release will take placeHenk no next release this month +Would the HDFS integration introduce a dependency on Hadoop Commons +But test before commit with modules such as Rampart and Sandesha engaged +Would anyone want to try this +Committed at subversion revision +Moving Sandesha from CVS to SVN was successfully done by the infrastructure +I attach the patch which I showed in the image +Create a zone with a host of hypervisor A +It might be a good idea to revisit code conventions and have it in such a way that it makes sense to enforce +patch against trunk rev patch is against trunk rev +First commit of findbugs changes in revision this for now to deal with juddi issues +Most LuceneSolr developers do not use Maven so youre making it hard for people to help you when you provide test cases in Maven format +rev branch x +Ive tried this on CR and still am not able to reproduce +I attached a fix in theThis adds a new goal called +Same as description +Let me know if theres any other GC info that would be useful +Thanks +This reports PDF as plain textDid some Googling and investigated some PDF parsing utilities +appliedHad troubles with which seems to be in both places and should only be in srcjavaAfter my patch I still have to disable the same testsSorry am not too productive here! problem is that of cannot be based on since is not published codeSo really should in srcjavaIs this the thing that affects all the disabled tests? Could be +Fixed and +Ive deleted the from the jboss and it works fine now +Cloned issue is to handle any further work to potentially automatically apply XML escaping given that part one now provides a utility method to make it easier +Agreed +The mock driver works to the point that our app currently supports I believe environments can be created environments can be started but remain in the pending stateI havent tried running the app against ec so Im not sure how much more functionality has been implemented +The slink works fine if a page where slink is included is rendered there is a method binding recognition for seam ui actions and recognized actions are added to +Sajid RazaCan you please paste the code for your workaround +Should we close this issue then +Can you please check my source code what could be wrong? source code contains several references to your custom code which make it hard to check +Crash at the same place so very likely the same bug. +Yeah I guess there will be situations like that +It seems you are using an old version +As Marek said this is not yet done in EAP CP CR. +Ive no clue if that can result in some performance issue but I think we have done JSF upgrades in this version +Fixed in dba +I think it would be better if the test was in the classYep I just did that + +rebased patch against trunk rev +This would require a release note to alert users of a behavioral change and the workaround +We know +I have not confirmed but think this may be related to to install a security manager by default +Combined with the tilesinsertDefinition the header and footer were returned twice +Filed for Alejandros hints proposal +The sampler should be generic enough that folks wont have to modify it to find it useful it should work for different keyvalue types and for sequencefile and text data +Did you try it out +Cant remember exactly what the exceptionproblems where I believe it was something about not finding the Registry on localhostHowever with all that said it would be much nicer if this worked out of the box +Fixed locally will be fine in Trunk too +This work is significant and we intend do this in multiple phases +This is cygwins diff command u patcholdnewIf not Ill be happy to recreate it if you let me know how you want it done +Ive added a test for this scenario in where the MDB instance fails with a bogus jndi injection as the customer did +Then recompile your app against the final versions of these jars +Attached patch with corrections for +Observed the router state changes as mentioned in the commentsBut when in step when router becomes master from unknown there is no issue in egress rulesEgress rules are working as expected + old configurations to benefit from connection poolingwhats the configuration diff between those two? or more specifically are there configuration parameters that only apply to +Verified in ER +Do you know if it is included here? If its not where can I download SNAPSHOT? And where can I download Rampart +Can you please use the mailing list for initial investigations +The fix was tested and allowed larger results to be fetched through an older salesforce version +It sounds like a fix has been applied +You might get around this by creating a DSL for your DRL +This should be the simplest data source use case we have from a usability standpoint +Im not sure what I was thinking about when I created +I will give it a shot though it might take a few days at least to see if I can recreate the problem first. +I have change to times +SorryIn addition as a nice thing to havePerhaps the literal should be extracted out as a constant so it appears injavabuildjavadocsSuggest some good places might be or maybe justRegardsPatrick +I was trying to learn about and how it was used +Test URL httplocalhostgeoserverwmsformatoptionsregionatebyautolayerstoppAbuildingswidthstylesnycbuildingssrsEPSGAformatapplicationBxmltiledFALSEversionexceptionsapplicationheightbboxrequestserviceWMStransparent try to get this in for I guess RC at this point as it is a bug. +I can see two potential problems with it If the ledger is large then the process reading the ledger might end up consuming too much memory It does not overlap reading the ledger with processing +This must be moved to +Also I forgot to mention that it does not appear to attempt to connect to the database until you click on the database dropdown +Does you seewish for another behavior +Oleg +In particular does a of possible Locales to see what bundles actually can be loaded via the classloader +The default target was changed to gmetrics in you for the change +just reopening to remember to review the doc as requested +batch transition resolvedfixed bugs to closedfixed +I personally feel that a separate POM for each would not be correct as they are logically the same artifact and fulfill the same role with only minor differences to take advantage of features in different versions of the do you think should happen here? Should I the bundle with only these files? just use the javadoc and sources from the jdk variant for all of them +Is there any reason we have two of these? If not I say we jettison the optional copy +Is this an issue using current master? This sounds like an issue I fixed a couple of weeks ago +Thanks +The rename patch has been applied to trunk +Complete see attached file and instructions in comments +There are still a lot of remaining issues with Stanbol if security is enabled! However the main reason for opening this is the deactivating the by parsing the option to the launcher does not work because of an bug! +Scott verified this one. +I think there are quite a few people who use this feature and those usecases will break by fixing this +Patch applied in rev +Struts does this all the time +Fixed the broken search method in in Metamer. +Hi RobertYeah looks like you nailed it something missing for the coverage case +with its guts exposed on mavenhmm by guts exposed you mean the Solr Maven artifacts document their dependencies +Closing issues solved in released versions + resolved issues for already released versions +removed fix version since we only add enhancementsextensions to upstream not backporting it +I was going a bit back and forth about this +Could you please add some javadoc to the state enum values explaining the purpose of each state and what the transitions are between them? Or augment the design doc for with this state machine and reference it from the code +Could you please post a short reproducible test case that would greatly speed up the fixevaluation +Maybe we can make a UI display profile that has that information +Thanks so much Simon for the enormous amount of work getting the code cleaned up like thisAs for moving forward esp +Already was not clear enough I updated it +Verified in Developer StudioVersion Build id Build date +did you mean to say showcase is not using the convention plugin please update it +Humphrey wrote Anything your lawyer doesnt need to know about belongs elsewhereWhere elsewhere does it belong? And who are you to decide what your lawyer needs to knowLarry +test codes to reproduce this is an interim patch for this bug that solves the problems exhibited by tsulins test on my system +Fixed in Rev +testcase have added the attribute but the attribute is completely ignored whatever the value of the attribute is +Thanks for the feedback looking forward to getting some more +Thanks Lars +For every file that is successfully validated the client would delete the +As of all of the examples have been refactored to use the +Looks good in +The build problems pointed out related to locatorwork have been fixed +BTW are you sure you are using the latest revision? I think I didnt deploy the snapshot JAR after fixing this issue +This has been pushed to master +Why yes yes you canD +Hilal doesnt solve the issue because maven is not actually executing the javac you specified but the internal compiler of the jdk that is set in your javahome +I just committed this +Not sure if this issue is the best place to discuss but Im curiousconcerned about the locking stuff +Therefore I do not see why you really need to expose the JVM instance and overload its best method since this is only used as fallback if no such instance is given to initialize the +Hmmm +The SMB test case still has a minor issue the tables was created as buckets but there is only file in the tables +bulk close for the release +Any reason to not just put in some blank files for the example app? With maybe a HTML comment telling folks to put in their own custom links etc? Why leave it to be magic? And when folks are browsing the conf directory they will see these files and immediately grasp what they are for! That would be much simpler then having complex logic in the that decides when a missing a file is a INFO level error and when it is a +patch for an earlier release not for commit in revised +Lets wait fixing this one until that one is resolved +and built in brewUnsigned jars from uploaded toorgjbossnettyCommitted revision orgjbossnettynettyCommitted revision +Can you test master +So should be considered to apply to the property and not just the getter +But thats a topic for or +Are there tests in our test suite for +It visualizes the dependencies between all the ant targetsIm not actively working on the Mavenization but spending an hour here and there on it when I have the time +no? Am I missing somethingPlease someone say something +Fixed with svn revision was a regression but the fix need anther pair of eyes +Thanks for fixing the exception type generation and applying the patchThere is a problem in the wsdl generated for the corbastruct elementI will describe the problem in as it only affects support for struct types. +Here is a patch that incorporates all the proposed fixesI verified that all the mentioned tests are fixed and no regression is introducedThanks +Whats the equivalent here? I see other bits of code paying attention toAre we handling boolean preferences efficiently? Before it would avoid the step when the pref was known to be and I dont see that nowFinally there is a feature in vectors that will save space causing it to write float values instead of doubles since we dont need bits of precision +Do you expect me to write long sequence of nested? operatorsWhat if attribute doesnt allow EL expression at allThis issue should not be dont reopen rejected issues +please review this patch +Local tests are ok +AndyThe patch is not final +removed old patch with updated testcase it another go now +That I would take the stuff from which predates the multiproject plugin and modify it to use the multiproject plugin +I will leave this up for more day in case anyone has objections then commit it +According to Tug Jersey uses the JAXB annotations to mange the JSON serialization so it could make sense to implement it in our implementation WDYT Andrey? +i spend the whole afternoon and night to locate the problem +This was a deliberate change by the productisation team to protect the servlet access using need to modify the ant task to support the provision of security credentials and modify the QS scripts to automatically choose the correct deployment URL +Bulk closing of issues that are old and have been released or marked invalid. +Rev +Thoughts? The unit test doesnt really belong to +README didnt help a bit to understand how examples were organized and such +Committed to trunk in r +When you have time can you document which validations are remaining from the list above +Eli I am adding this functionality as a part of +Thanks for answering my post Any chance youve ran into an infinate loop using modjk and ve resolved this issue please reopen with more details if needed. +gem install tbconsole deploy tbconsole connect Or connect at httplocalhost the moment this does not work with rvm installations so its been yanked. +Updated test cases +and resolved +This jira tracks a change that originated in +Please feel free to provide feedbackcriticism +Patch added +The generated sql is now not runnable at all as it returns the error Cannot use an aggregate or a subquery in an expression used for the group by list of a GROUP BY clause +While the possibility might be good you must watch out that this does not endanger the interchangeability +This is not a bugYoure probably using Blocking on Page ModeAnd youre not acknowledging for some reasonIf youre using DUPSOK look at the ACK batch sizeIf you are using Transactions make sure you are committing your message receives +Apache Directory studio version has been released. +we need a solution for that as well +Patch attached +That can be easily used for testing +That is why I suggested removing the initialization of the from the onStartup method +I discovered which describes the same problem in slightly different circumstances +ok +attachment On MS ubuntu ASF output of cd find +Agree +I think this is the same bug as +The new and upcomming version will support this +Now that work is underway with AS all previous community releases are +So they can set the headers directory on the exchange and then return the body to use +I think this is done right? +the old stuff is deprecatedunmaintained +bravo! The fix is trivial +Let me try one more timeToday by default operates in writers wait for sync mode +What about? That shouldnt source as well me thinks +Updated in trunk thanks! +A few things jump out immediately +Within a team or you are going to have snapshots and you will most certainly have these checked out into your workspace so you dont need the source jars +The backport is trivial but were only applying bugfixes to at this point +See file have reviewed the patch +This issue is not an RC blocker +Not so obvious from the errors but seems the cyphersuits name was wrong and also they neeed to be RSA +I anyone gets a chance to look at this the first thing I would track down is if this is a bulk fetch issue +Attached is the patch for to use the onError value properly and handle thrown when the actual data does not match the column type +The simply creates one split for each zip entry in an input zip file +Sorry I missed the feedback +Just tried reproducing the hanging blocking on cluster here +pushed fix for integration tests to and master to correct +My ultimate goal is to get a patch for which will allow triggers on tables with blob columns if none of the blob columns are referenced +I see that now +Were not using the in the right now it just copies the contents of the directory over from the temp directory to the final location +resolved +Ok I can do that +Ive uploaded a new patch to phabricator but it got a new revision number +Hi attaching patch for issueThanks +If the purpose of this change is to provide a workaround for then my vote is definitely +As this issue is actually a problem with the javac compiler there isnt anything which could be done in HV to address it +should apply cleanly to trunk +This is an intermittent failure probably a timing problem +The srcmainjava tree is all set with and the imports as well +Thanks Pavol I missed that step +do not throw exceptions when is missing +We should make sure the examples work so adding this to +On RS side it pauses and then retries foreverA more involved soln +Once we have the capacity set for each of the teams it will be easy for the PM team to go into the TBD and decide which ones should be in the next release +etc +Upgraded +Document deletion now requires the index write lock and hence cannot be performed concurrently with document addition +Selecting the map enables the shortcuts again. +When opening multiple tabs the same cookie is used +Moving to release +the biggest reason this patch wont work is that with the does not keep track of pending updates to get this to work again it will need to maintain a list somewhereAlso we need to make sure the API lets you grab stored fields from somewhere else as is it forces you to store all fields for all documents +Every time some change on metastore takesplace we have some logic which needs to be run +If ignoring text when rendering is added it should be an OPTION and NOT the defaultPlenty of documents such as electronic brochures lay text over images and all of the content IS to be seen +Thats the first part of the patch i just attached +applied patch and added test casesthanks Andreas! +Thats only possible if you haveown the source code for the annotator that youre deploying +New snapshot deployed +I only had in the classpathIm not sure if its a JVM bug or if its a problem in Derbys engine boot code that gets triggered by the way that version of the JVM invokes it +I think populate in needs to filter the properties it caches based on the get value of either get or get of each so we can be sure we only have the properties belonging to this there could be other side effects to this +Playing the devils advocate +Fixed the findbugs issue +Connection sharing removed from daytrader in rev +Keith and I were talking about himintegrating Adaptx into Castor for use in Castor XML +Thanks. +Kathey I took a look at the patch and it looks good +Ive created a test which makes multiple native method invocations through JNI and it currently crashes drlvm in the same way as does tomcat shutdown +again hopefully we can get this committed soon +Patch v reads the cached data from the controller context +Closed as WONT FIX. +Additionally clientIds are new used for the names of these components rather than the d like to see this resolved for Columns as well before this issue is closed +Oh! +If you can rebalance the disks then you can just use roundrobin or roulette placement +As already said using revision my webservices are published as they shouldBut if I add this extends to my implementing classes as you proposed me to do because of the spec requirement the services are no longer publishedThis is probably due to the piece of code you found in the module +Filters on Appenders would require too much configuration would be fragile and would not perform well at all +Instead I created new ones that replicate the behavior of the previous release as closely as possibleThe source assembly includes everything from a clean subversion export but packaged in a directory with src added at the endThe binary assembly includes all jar files and the generated site +Thanks so much! +This first pull request fixes the issue only partially +here is an updated version of the patch which removes the call +Granted it will add many keywords however that seems a small price to pay to get items to show up during a searchPerhaps a keyword weight should also be assigned so the description keywords will not overwhelm more appropriate product keywords +Can this please be integrated into +Ill have a closer look at the logs and try to come up with a proper solutionCan you please create a separate ticket for the postgress issue? Bart Ive raised issue for the problemBR LucaIve committed a slightly different version which should have the same effect +We are facing the same issue with our setup and are just not able to get rid of this issue the time taken to execute our tests is close to minutes and once this is completed the communication failure occurs +Since hudson can test only for trunk +Here is a bit of a start +If I have a process servicing multiple I doesnt necessarily want that process to have hdfs root privileges +All set. +bq +Has anyone made progress since the previous update years ago? +Karl in your application you store nothing in Lucene isnt itDoes it cost so much to just store an field id in Lucene +heres a patch which just updates the excludes for rat +WOW nice mike! do you have benchmark numbers here by any chance? After all those improvements FST etc +Question what is the need for adding this license +After this verification a good error handling should be enough +I think this should be reverted and we should perhaps supply two options one that logs and one that throws the exception +x +I also have this issue +rejecting this since I cant reproduce it against other common dbs and I cant see how I can guard against faulty jdbc driversLet me know if Derby guys have suggestions or possibly can explain why they freeze the machine. +patch and test should put in applied with thanks +Interestingly the commit where these were added is marked as a rollback of the maven release pluginThese two should be removed +This is the code additions to make this work +Hi BrunoIIRW there is currently a reflexion about Themes +This is connected to a tck issue so moving it back. +Sorry for cluttering up the comments here +Thanks for the example for earlier version not to be committed +Thanks for the patch +Done via +Fixed as per tried the nightly extension works a treat +My only concern is that the reference count seems not being updated in undoAlso why add verify to add? Is it for quota check +just to get others to double check my handy work +Thanks BilginYour patch is in trunk revision Thanks to Jonathon for review and test! +How would the test settings be causing this +If there are no further suggestions or objections Id like to move forward on this patch +Please reopen if you can provide requested info +Does this also work if the blockcache was disabled through the +this is done and documented? I know we discussed a tiny bit about integration with AS +Theres no reason to spin our own implementation for only a small speed up +An initial crack at thisI tried to have changes to existing code kept at a minimum but I did have to Make the configuration key strings for public should be fine because theyre final Strings Swapped some method calls Made the actually use the table parameter passed to its constructorInstead of giving clients a KeyValue pair this uses a which is just a key paired with a table name represented by a Text object +For now hasnt any support for as encoding when adding text +Are you wanting to require this somehow +Thanks for all the great work Narayanan +No tests +For instance we could move the +This code change has been reviewied +Do you think it is possible to get me the filter that fails +Verified by Paulex +txt files in EAPbuildjavadoc src dir +Shall we open new tickets to discuss them +NPE to added as discussed +I will move the inner classes out in a subtask of +Affects versions and can this be moved to release +Closed as per state in Bugzilla. +Also I am able to run the mvn scmstatus command as well +The attached file is your patch plus my changes + +Its not as if synchronization within is capable of guarding access to an instance variableSecondly for compatibility with existing which are capable of appending to a which is the majority these days +i dont see a test for all NULL args for COALESCE +This is because the shared Maven repository where were getting our Eclipse jars from doesnt have source +There are a few more places where we still have double escaping see attached screenshots. +AHH i didnt realize JenkinsHudson was using the redployer +Today if you build a JSF portlet in english and french you need to define the accepted language in and the portlet bridge could easily and magically add those languages to that would be one less trouble to worry about +Ive committed my implementation in rev +Both keys and values must be Serializable +Ive changed the implementation regarding the events on my side +reopen to set fix version field +Beautify the mailing list see the screenshot and the diff subscribe unsubscribe and an archive links by using Thoughts +Here is the fixed patch +points to a possible problem with the fix for this issue +Please could you provide me more info when was this problem seen? +Any chance this will get some attention +UI now shows the prompt with special instructions to the end user +We will also need to add this to out documentation when we get it goingEdell +Dan Maybe checking the java types would lead to the same answer but logically its a check ofDan SQL types only +Being done in tandem with +Satheesh checked in the fix for this in codeline with revision r. +so maybe decrease of sshmanagement from to will be a workaround with respecttrue case need more test +Please apply this patch +This passes the same unit test as beforeSiddharth please open another JIRA for the case you described above +Attaching patch for it +t absuse the ticket t abuse the ticket system +ifwhen comes out it should only have bug fixes and i dont think can be justifiably called a bug fixIn particular im worried about people who have applications that are relying on the eccentricities of pre HTTP response and that upgrading from might not be trivial for them which would either mean sacraficing other fixes or changing their own code +committed to both trunk and branch. +Fired up a debugger and it seems to be having issues when resolving a class called up the call stack when were resolving the type for the field message line of resolves the class name to +The validation doesnt need to deal with that we should deal with it in trimming +Addressed reviewboard comments +Ok +This patch contains an intial design proposal +Pl reopen if you see this problem againthanks is still an issue in my environment +see the have improved the exception message in case the wrong format is given. +This id dont mean for a wich page navigation performed with new viewId or to a same page +Seems Chris cant assign it to himself or submit the patch to hudson so Ill let someone with greater jira permissions help him out +Its only an implementation of which uses the to keep pages +I have updated the assembly +I run osgiupdate on the bundle times one time it appears to work the other two times print out this gave the newly released pax web a test still happens in this version too +Doesnt halt on pom error since build forgot to commit two file last night +Can you explain a bit more about consistency issue? If we taking the snapshot correctly the uncommitted txn should never be in the snapshot when a participant is elected as a leader so we can truncate uncommitted txns from its txnlog +Please review and if looks good please commit +Worked fine with IE +This bug has been fixed in Xalan Java +This should be a larger issue. +Also optimizes copyBytes in +When setting up the ignored paths filter corretcly it works with legacy servlets on too. +one possible solution is using reference counting mechanism each deleted ledger should have a list of bookie who owed it +Should the output directories be empty or should they produce part files +We are using an embedded db and in the past property files because we cannot assume Windows is the platform Geoserver is running onto neither that a certain RDBMS is around +please review thanks +Duplicate of +The was about putting in place the wizened its seen it all so catches and retries the litany of possible exceptions +You may need read the javadoc of +This is similar to andPerhaps all three could even be a single tool that supports a variety of formats +Can you open another issue for the hide and show improvement +In our Lucene work we took the approach of indexing all fields into a single field FULLTEXT which is the default field for queries +Jake that is reasonable but I think I would just attack it the other way round by moving Cluster now to the writable jar and then move it back later as possible +Hudson seems stuck recycling patch queue +I did a weep through them yesterday +updated das and distribution assemblies files need to be picked up other than LICENSE and NOTICE? I hesitate to put in a wildcard in case some build artifact with a +Patch applied many thanks! +Server log Len For the file update do you mean add the steps refer to installing opensso with Tomcat sentence in the file +Ok I found a solution but its a bit ugly as whole idea about VFSP Need to refactor and should be ok +Im a bit puzzled about thisI always thought that async is great but only usable in some situations depending on what the application is doing and what kind of consistency is requiredSo if async is an option you should use it for both local and wan replication but if its not an option how could this be used in wan onlyCould anybody explain a use case in which this might make senseOne case occurs to me which is having a local cluster on which changes are performed and a remote cluster which is just a readonly view to the same data but I assume this use case could be solved by setting up an async store and storing on the remote cluster +Will reopen if xml security still does this + to trunk. + +If you can please try contacting Pin Xue again to get his confirmation on contributing this under the Apache licenseHowwhy did you work together on this? If you share an employer and wrote this patch at work then an OK from your manager is typically enough and often necessary depending on the copyright terms of the employment contractHow much of the code is written by Pin Xue? We can include the code under the BSD license but that requires special mentioning in the LICENSE and NOTICE files that we and all the downstream projects need to worry about +clarifying that new features go to not promised Jan that Id work on this today +Would you mind trying to replicate the problem with +committed fix that causes refresh of the tree when console configuration gets reset this might happen more often than we think so should keep an eye on this to see if it causes bad things to happen +Done +I made some errors while transcribing tests outcomes here is an updated summary tableVALUE IN BLOCK VALUE IN development ENVIRONMENT BLOCK OUTCOME undef tables are updated created when update tables are updated created when undef tables are updated created when update tables are updated created when create undef tables are dropped and create update tables are dropped and undef create tables are dropped and update create tables are dropped and undef update hbmddl not runupdate undef tables are updated created when +Fixed indentation and added a relevant tests +All tests pass +This is a bug in the unit test +We have to try is ri is doing right +verified on +I tested the following Random monitor logj master tserver procs and updates to other procs when those are restarted Logj forwarding works after restart lock advertisement for master is updated after new master is start accumulo info is updated after new master is started +fixed commit it on latest build from branch and works fine. +This has fixed the Gump build of +This problem is solved by feeding the correct base URI to the +I feel pretty stupid stumbling around this project as its setup quite a bit differently than what I am used to but I am trying to get the hang of it +CGLIB seems to make those methods final by default and there is no obvious option to keep them do you have any insights to share here? would as a minimum need to hook into the CGLIB ASM usage to change this +you have to tweak your servlet container to also extend the timeout. +Look into this for either postpone or resolve it before releasing like the idea but theres no patch available and the syntax change is +Nicolas I think it may make sense to edit and renameredescribe this issue now if you are going to make this more generic +done the wiki page is updated +Previously we thought its optional to call the endTransaction +What problem are you trying to solve +NET build of the testsUpdate I ran the tests again on +Hrm merge always seems to do these prop changes +Reopening to move to as was issues after move to in +let me know! I think we can handle this +What others think about this issue +Also in MR land where are splits being calculated? this requires HDFS accessI probably do not just have any depth of understanding here but will access to HDFS not be an issue for the client in the kind of scenario this is attempting to address since what is blocking is the port ACL not being opened up +FYI the waiting on dispatch message means that there is a message loaded in memory ready to be sent to the consumer and the consumer should be able to accept it +I assume you want to be able to overwriteextend the default list of plugins rightWhy would you want to do that? Do you have a problem you cannot solve by just registering plugins and thus need to place your plugins in the same bundle as the web console and thus extend the web console servletThanks for clarification +Eirik Check your plugins version and take +Specifically it allows for a schema package but also covers Config Inst Arg Method and Class names + +Have a datepicker between two and its rendered useless in IE +for backport +Actually I take that back +With Mojarra panel menu works after clicking on the menu item the request is not sent after clicking on the group header the items in this group arent unrolled +Committed changes to the and to support start level assignment based on bundle location in the package in Rev +Introduces logging to note which JPA Identity Store Identity Entity classes were determined for each respective entityIve tested it and found that it fixes my problems +Once I become more knowledgable about thestruts internals I hope to be able to suggest some interface methods that aremore abstract and supportive of assembling the configutation information withoutthe need to use each independent piece of configuration data +I committed the patch to trunk +attach a new patch rebased on lastest trunk +We should probably talk about this on the user list to see if thered be interest in such a thingMy +bq +Did you update the jenkins jobs +LGTM +Unscheduling since this depends which in turn depends on a Metro issue still to be solved +Some examples have +Here it happens because of KEYCLOAKSAASIDENTITY cookie which also shouldnt be secured IMO +For all entities Y referenced by a relationship from X if the relationship to Y has been annotated with the cascade element value cascadePERSIST or cascadeALL the persist operation is applied to YHow do you think should behave in such situation? Should Entity B be managed by two persistence contextsBtw +added testcase srctestcomopensymphonyxworkconfigproviders +Attaching patch which provides this patch which passes unit tests +SOA Platform has been added and tested. +A Patch fixing this issue +Closing as a stale issue +The reliability of the test runs is a priority +incorporated Konstantins comments +Committed to trunk and +Or are LIMIT and OFFSET already being tested and passing +Note to self Chapter table alignment formatting issues +hey weijin youre saying that volatile is not required because were doing the CAS anyway? The only issue with that is when you reload couldnt you possibly get stuck in an infinite loop of compare and swaps if the load was optimized away because there wasnt the volatile keywordYou guys definitely know the freelist stuff better than me so if you think its safe to take out Ill definitely do it +No worries there was only about or so +I would assume its a bug +I was going to have a go at removing the field in unless someone else beats me to it +The easy way out would be to the juddiv prefix to each table name maybe the next JPA spec will support it +PabloWould it be possible to undo the whitespace reformatting in your branch? Because of the whitespace reformatting at the moment its very hard to see what your actual changes are +thanks Alejandro +See EJB FR and EJB FR +In fact it sounds like it should +I just committed this +I should have the opportunity to do that either today or tomorrow + +Verified by Anton. +But I will have a look if I can rename the version for the RCP only +Committed to trunk. +We have already used some code directly from Jama +Will create this in a second +I cant see the changes made here in this patch in any version of struts and xwork +So we need to implement option before the freeze +Composition diagram before refactoringDubious dependencies in red +is now does not need to be public +Because I just learned about it I was going to check the precommit target but it failed because I dont have a new enough python +Token getrenewcancel no longer get a token before performing the operation +I think wed probably want to keep the percentile list limited to start with +Expect a beta version in the next day or two +This patch contains the fixes for the issue and the integration test. +Refactoring required in the code so leaving open for a little longer +It depends on what kind of snapshot it is +Here is it possible that we can keep different values for a pair of deprecated property and property +realized that this isnt quite fixed yet anyway I didnt update +New layout of features on rd party extras site will look like thisNested feature structure looks like this +criteria is not EJBQL queries. +Fixed in FOP transition resolvedfixed bugs to closedfixed +For instance the serverid might be used to provide the trusted SSO server for SSO or identity tokensAgain its a SASL requirement that the same protoserverid must be used to instantiate the SASL client start +Ok one more time! We were playing with this some today and found that the Struts Core was recently updated to include a system scope dependency to from the Sun JDK +Another perhaps better solutions would simply be to store the temporary task output somewhere else NOT beneath the output directory code in Nutch uses which doesnt specify a file format path filter +Create a new Gadget Procedure +BTW the implementation of this is relatively simple the put set method would Check if a Map were the latest object pushed to the stack +If anyone can think of places that meet this criteria please let me know +Jasper if you want to reopen with the requested information go for it. +Ok dokiI committed the patch in Rev Rev RevTrunk RevSo we could finally close the issue +The deprecated reflector is still in place on but has been removed in +Screen shows default type as Seam for EAP AP in following wizard that JBDS knows this is EAP to JBT project since this issue is a bug of Tools +all unnecessary modules removed from both for did you fix yourRob +The patch contains both modified and a diff of files generated by forrest +Create tenantI think you will have the similar exception if you reconfigure in configuration of JCR in an issue that occurs only with eXo Cloud +Strange enough the protocol has changed to the https I configured but the hostname hasnt +I guess you are right about the resource leaks +So Ill tackle that task today stay tuned +Thanks Eric +Without the method we run the risk of having someone clean up code that is not needed and have these things go away +Postponing because of dependency on unresolved Metro issue +Includes code change and associated documentation updates + for the to trunk. +Im busy with other projects at the moment but Ill try to carve out some time to test it later this week. +custom schema SCHEMA SAMPLEPERSONdn cnsamplePerson ouschemaobjectclass metaSchemaobjectclass topcn samplePerson inetorgperson sampleschemadn ouattributetypes cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou attributetypesdn ouattributetypes cn samplePerson ouschemaobjectclass metaobjectclass metaTopobjectclass top date case case dn ouattributetypes cn samplePerson ouschemaobjectclass metaobjectclass metaTopobjectclass top gender case case dn oucomparators cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou comparatorsdn ouditcontentrules cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou ditcontentrulesdn ouditstructurerules cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou ditstructurerulesdn oumatchingrules cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou matchingrulesdn oumatchingruleuse cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou matchingruleusedn ounameforms cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou nameformsdn ounormalizers cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou normalizersdn ouobjectclasses cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou objectClassesdn ouobjectclasses cns amplePerson ouschemaobjectclass metaobjectclass metaTopobjectclass top samplePerson samplePerson inet date genderdn ousyntaxcheckers cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou syntaxcheckersdn ousyntaxes cnsamplePerson ouschemaobjectclass organizationalUnitobjectclass topou syntaxes +esb check to be invoked asap returning false before you do any additional checks invocations +Yes much better +Hey Andrew +If there are no further concerns I do not plan on optimizing memory further at the expense of code complexity +Its hard to reason about cluster behavior that is what I meant by confusingbq +Currently the input format uses map task per arc file +bulk fixing the version info for and all affected issues have in is no indication that anyone is actively working on this issue so removing from the fixVersion +Hi KojiI assume there is an infinite loop +For this reason filtering is hard +it is fine for a no? this is not the only change which is no backward actually already shipped with in +Zhenxiao Please let me know if you plan to continue working on this ticket +I think also that is a duplicate of so I will close the issue +Im not sure that we could make arbitrary topology pluggable without scheduler change +Closing issues resolved in released versions +Ideally in the hook what ever the patch does that is what one will be interested to do +For testing multiple setup tasks I tested on it single node suspended the and submitted jobs simultaneously sothat both of will get initialized and will wait for setup tasks to be launched +See revision Please close and when if this one can be closed. +If you could still provide some unit tests thatd be great. + +Even if only for licensing considerations Id go for option a dedicated module which contains the infrastructure +One of reasons to use maven is to get a common way to perform basic tasks across projects +It makes sense to mention about this happening in the next release and we should have this in the release notes for our next releaseIf everyone agrees please change this to TBD after verification to be included in the next release notes. +I have added your Javadoc contribution thank youAs far as logging goes yes I think this should be discussed onOne thing I really like about Lucene is that it has no dependenciesand only depends on for generating some classes at build timeI will close this bug for now and if choose to add logging we can thisbug or open a new one. +That is why I left AAA out of the affected components list +Closing old issues +If you finish it before we release beta please correct the version +Bug with date inserted into Oracle database with no timstamp defined produces an error in rails layer +Updated patch fixes null issues +Reported and fixed the related issue +Jaco would it be possible for you to provide a patch? That way itll be easier to verify we covered all the issues +Oops missed to added the integration test from the patch + +So I see no reason to suspect a bug here +I have tried to backport the class from JDK to JDK and found no problem +Is this version suposed to solve the timestamp bugI also tried the driver which works fineDoes this mean I cannnot use Hibernate for Oracle unless Im willing to pay for the driverIs there no chance for a workaroungTIAErnst +Committed a chapter section on this +The patch has been applied but there are still problems with large images +Will try this +Thanks Dag for picking this up! I was puzzling it over but you knew what to doThe patch looks great a couple of minor items Latin abbreviations like cf +Consequently XSLTC assigns it the special type reference +No problem about that I was just curious I know I was not upset doesnt seem to work for me SystemLibraryJavaContentsHomebinjava SystemLibraryJavaContentsHomelib Usersmsacks Usersmsacksconf classpath Usersmsackslib conf Usersmsacksconf main classpath UsersmsacksDocumentsworkspaceoutproductionMainUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslibUsersmsackslib encoding UsersmsacksDocumentsworkspacesrcmainjavaorgapachekittykitty connect localhost connect localhost Successfully connected to hostkitty domainsdomainsDomains Detected Users Catalina kitty setdomain Catalinasetdomain CatalinaSetting the domain to CatalinaThe domain is set to Catalinakitty lslslisting files and directoriesobjectName CatalinaobjectName CatalinaInvalid character in key part of propertyAlso looks like gets removed +for adding a trycatch block and making the method faster for those with drivers that support it while letting those with drivers that dont pay a slight performance penalty with the generated exception +I am ok with throwing exceptions in methods instead of empty strings +Bilgin I have committed the patch to trunk +Thanks for pointing this out +Dear a lot +Its functionlity has been implemented and not once already. +I dont think you can get whet you want from any change to either of the jobs that Generate is composed of +Can not reproduce a bug +Patch looks good Dustin +Incorporated review comments +From the JIRA comments above it sounds like the DLQ needs to be disabled on the MDB because Messaging will be performing that delivery instead +Fixing comments from Sergey +Waiting on another committer to before committing +This means +Closing issues after Configuration Admin release +Thanks for the patch! Applied in revision PS +See for some background first +Patch which includes Yasins improvements and also passes all but tests in +I have not used it so far so just wanted to play with it before adding it. +I fixed the iterators in Axiom to behave properly as defined in api +Ideally we could get them a snapshot so they could start their integration testing +Canceling the timer doesnt affect the timer task if its already started +This was fixed in the latest draft of the interceptors spec +Fixed in rev +So ideally if you can upgrade to RC and let us know that would be great +Does it sound good +Id dynamic fields and Id leave uniqueKey as I bet of users need it +Write a new test to make sure that an app which wasnt deleted from by previous RM but has successfully completed is removed after RM restart +We do have the supplemental package installed on our Xenserver hosts +Ah hem anyone +war it is guess its a duplicate of +We should have a group called something like that most users do not belong to but that the user running the mapred daemons does belong to +Also updated the contributors list and copied the LICENSE files into the directory which is recommended for projects +Michael Offner Any word on this? I also have this problem +With this modification it works again +Attaching patch which also contains default configs set to +So is there active development or support occurring for these older releases? I assume no or not at the ASFIf the licensing of the artifacts is well advertised and an explicit action is required by users to downloadinstall Id be in favor of allowing this +This test sometimes succeeds sometimes it fails depending on how quickly the region is recovered +I was concerned about keeping around connections that have closed for ever +I reinstalled the eclipse and groovy the result remaining the same +I believe this JIRA should be closed because already adds zookeeper as a reserved node name when creating the hierarchy for quotas +Adding buddy replication could expose a DB bottleneck that was previously masked by a total replication bottleneck +is just used for user regions +Also renames a few methods to better reflect the meaning +There will be no more twix or snickers implementation +I would be interested to hear from users whether this a workable option +This seems to be fixed since the upgrade to ICU +Paul can you refactor the d patch so that it will apply cleanly to the reorganized source tree as the surround parser is now under modules +I think this patch should work just fine against trunk since we have noit deviated much +doesnt leave holes when it drops punctuationI think holes should only be real words for the most think kuromoji has the problem here it creates fake intermediate tokens and then deletes them and this somehow screws up the decompounding graphit should never create these tokens in the first place! I think its well accepted that words carry the information content of a doc punctuation has no information content really here it doesnt tell me what the doc is about and I dont think this is controversial I just think your view on this is extreme +Thanks Marcel! +for the clarification about initialize cleanup +Really old issue +If its a slave config problem then I can fire an email at and they can investigate +This appears to be fixed in based on a quick check. +This is a very strange behaviour +The above site has been updated to indicate that catalogLocation is available! Added catalogLocation and included only the schemas used in maven build +Rini suggested that it might be something to do with reprojection in your srs +I actually plan to check into tomorrow morning after checking the nightly runs. +One could update to pass a Seekable to track bytes consumed from the underlying stream which would fit the contract for fairly well +It also replaces an occurance of valigncenter by valignmiddle +Are you sure? Its committed in +The installation is therefore by definition unordered +Please with further detail if problems still exist +But i would prefer that take the same approach and simply cast to the Impl classRegarding the changes to seems there are a lot of them +Odi What I meant was that we couldnt just blindly skip the number of bytes given in the header +Airavata API is updated to to do the same +The duration is calculated wrong even in trunk because the endTime is not set since afterCall is not called yet at that moment which is not a big deal +I think it is not likely to show up as a regression in a user application as I dont think there are too many cases where an app would have functionality depending on of the result +Ive found some additional issues is missing activemq jars Ill provide a patch fails to deploy Ive filed for this +Actually the does not do exactly what Ken wants hereThe parameter is for the to tell it what to use +I already tried different executions but that didnt work because I cant pull the dependencies to set the providers into the executions +The code must fit in the Commons Codec styleframeworkinterfaces +Tim the patch was applied as expected +However caches in numRetries and hence if ZK is not running reset will retry for a whileThe simplest fix would be for reset to ignore he cached setting and to retrieve the value again from the setting +In that book was a short story called The Purloined Letter +If we share the cache between multiple threads maybe we should raise the max size +I will see if i have a default comportament and if i post +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +I have not yet upgraded toAs a side note Im not sure this is a problem other than the weird log message the warnings occur before creating the file in S not at the end as I previously believed +I prefer to keep the hash verification mandatory +I got it +please reopen with more details +Please add more links if I have forgotten any or new related issues are added +and versions in turn invoke version +So a default name would be etcIf there is an analogue for query that would be a Java method so I guess query query is appropriate +What version of Roo are you using? Attached is the script and a backup of what your project might look like using the +Appreciate if you can provide your test files thanks in advance +I added the missing license header to in revision This is a good candidate for the release so I merged the fix to the branch in revision +fixed in in not retrofitting to. +bq +There is no obvious way to write a unit test for it +to commit as soon as we upgrade to jclouds +I guess that as is allowed in classname there shouldnt be any is not supposed to thing +Must be completed +The old polling provider is considered to be obsolete and unsupported as far as XNIO is concerned +Thanks d rather see a new filterQuery type like ofq than being stuck with the current options +Makes sense +Patch resolved issues for already released versions +Will comment over there. +I can not observe this behaviour in version +If a user knows its fine they can configure it right? Why do you have an issue with letting a user configure things to work as they wish +Verified again and it is ok +Superseeded by the move to Gradle. +As the Application Registry process content definition of a portlet portlet preference is created and stored as mopworkspacemopcustomizations s child node +By adding some explicit return types by a handful of methods of I can make the scheduling problem go awayI am now trying to find an incantation of casts that will cause the code to typecheck with explicit return types +Hi Alessandro I am working actively on it and hope to have a patch ready in the next day or so +Hi SebastianActually im from Kufstein AustriaI Would really love to join your presentation if i had some more free timeYou do such a great job please keep it upGreeting from Austria all old Resolved issues +Committed this to trunk and +Should be resolved in ER +Is the low priority because there is a workaround? If so what is the current way of supporting different configurations? +I just committed this +hdfsproxy and raid contrib projects were found to be specifying alternate versions for and junit than the rest of the project +What was the performance difference you saw after the patchThere actually is a problem with head contention in the common case that happens with my patch +If one of the original developers could review the patch +No idea +This patch resolves a compile issue in the last patch and has the latest work that was done for the bjoin +See if patch v addresses Jons comment +trunk +Please comment if you feel this is a mistake +to trunk +rev +Note to the committer please add the change description to appropriate section +But now I wonder why you are hitting Illegal seek +Could we merge this patch into trunk firstly? Then after we merge this patch into we can merge corresponding update of to +But this is even just part of the issue i dont think we should do this +On others its +Thanks RichardPatch applied to NIOCHAR module at repo revision Please check that the patch was applied as you expected +Requires the new parent to next to next this from the schedule as this is not really required for a new to parent pom and adapted legal files +Toby FYI is already out +My experience is DRLVM can assert after this patch +It also forces thread context switches for each invokation which is also not something we want the overhead of of the time +Thanks +at alpha renaming is happening appropriately and these test cases compile correctly. +foo +Failing that force seqdirectory to run in local mode and put the result on to hdfs before executing seqsparse +esb archive is deployed in exploded form then even changes to are ignored +skin has been applied and made default +I hadnt thought about this in a while +What are the simple workarounds you mentionSetting on +Looks good + +Note in any case this issue should be resolved and the bugfeature about looking in classpath and not just jars should be opened against b and linked to this one +Is there any method by which the files are consolidated and not held up by a single unconsumed message? Expiring the messages doesnt really work since there needs to be a consumer for messages to expire +Merged into +I plan to get to it soon and will let you know if I have any questions +only partly fixedI removed the Observer and a few other classes which are not in the spec anymore but there are still a few +My profiling shows that the sorting barely registers as part of the time taken to load a model do you have a case where it seems to make a measurable differenceThese arrays are so small that any sort isnt a problem that I can seeStill small wins are wins +Sometime back we investigated how this approach works with and realized some important implications +These are the only slfj we should have there +Should we still support escapes? Would you like to make a patch for this? Passing tests on Windows will be the hard part +Put all per table config in a single node all system config in a single node and all namepace config in a single node +tests hang +If you think your patch is correct you need to click on submit patch to trigger the QA system +I currently have other priorities but as soon as I see an available slot of time if no one beats me to it Ill go into it in more detail +The question is then why did it get put in in the first place +Perhaps this could go in itself if improvements are still accepted +I think there is a conceptual difference between a client knowing how to contact a service and a client knowing about details of the services internals +Dependency injection +it sounds like were not handling fractions at all either +So I can probably address that +actually www directory is not configurable on anything I have seen including our tools +Peter Looks like your issue is because of the client timeout when you didnt receive a response for sec +See issue for instructions on how to add stuff to the GUI to allow this to be editable from the Administration GUI webapp page +branch already merged to trunk in r +After thinking about it I no longer think it is a good idea to rename fields just to be compatible with older Hibernate versions unless there is a very compelling reason to do so. +So you wont get trailing whitespace between the expression end and the next token considered as part of the Adjusted the computation in +I have no idea if they havent left behind an email address +A workflow manager like oozie will resubmit duplicate jobs +Regards release notesCan I have a clarification on how handles this now +fixed in the the resolved issue. +I have added openssl headers and lib to no issues on Debian SqueezeCould you verify that patch +using your patch i was able to start is there anything i need to do now or is the database good again? +Is Cheng Che Chen willing to file an ICLA +may not have my dependencies properly in order +Modified patch committed +Thanks. +Hi SaschaMany thanks for test it +seems helpful at best and harmless at worst +In I asked some questions about the properties in the specI need some further clarification on these properties and how we refer to them Questions +Exactly how is this going to be usable by users with a large number of instances +Just ran the test suite and this is all working +Completed. +So this should come from the Maven central repo +Bulk close for Solr +JSP compilation errors should be fixed before deployment exception handling is for runtime exceptions +indicated a. +Simple patch to fix the problem +This will provide SSL with certificate authentication +Backing out the change and reopen until those can be resolved +The component name here is Maxent names are a bit confusing Maxent is the old thing we currently ship and ML is the planned replacement for it +Thanks. +This issue target is onlyBecause and console each are not fully sameness and it is impossibe to just apply Jans posted patches to branchesSo we need to rewrite the patches for and this work is progressing now +The issue has been worked around for the time being by disabling the jms metrics by default +WFM +Eclipse project to reproduce the can you please have a look at this +great work. +Then operations like removing an attribute or adding an attribute to a DOM tree cannot be programmed if they imply to modify the root node +Java code to view the two fonts +Also experiencing problem but within a different applicationservice that needs to use the as supportsIt well sucks to be using applicationsservices that rely on hope this gets fixed QUICKLY and RELEASED +Committed revision! interesting! +So as you see an attached version of the directory assembly plugin is the only thing that works for it would make more sense to make directory one of the output formats for the assemblyattached plugins rather than have seperate directory and plugins +would describering call fall into this ticket too +We really should add a comment in that generally constants shouldnt go there unless referenced by a lot of different thingsOtherwise ltgm +Hi JacopoI have merge your last commit with my local changes and then I have created this new patch that is the remaining changes to be commitThanks a lotIve applied your last patch however something went wrongPlease let me know if Ive left something outTHANKS +Hi WernerSorry about my answer I just realized that out of the context its meaning was not exactly what I was looking for +Ive attached another +defer all issues to +Uploaded the final version of the file that passes all unit and integration tests and that I will commit to SVN +I had to decrease the workers to and set true +This morning I was working on an alternate form of Schema because I need a Schema structured data structure for another project and require manipulating and composing items associated with schema nodes +a couple things after reviewing this patch +Committed revision +Example containing bundles A B C D where B reexports A and C reexports B but D cannot see A directly +Thanks for doing this Dhaval +revert some changes +attach a patch to delete sub process instance after end +The problem is that is bundled into our bundle and bnd computes imports based on our bundle contents ignoring the restricted imports actually needed for +attachments on that mail a re missing waiting for the user to would expect the parse to throw an exception of invalid xml in this case +I see no reason why the standalone server does not behave in the same way +Great thank you +Implementedverify still dont see for +If it isnt an Ivy problem then Grapes just needs all in all the Grapes code needs work +All other doc related issues should probably spawn a different issueforum topic +Added formatting +I believe this one was seen as a good idea +Can we get the pom that triggered it? Thanks +I agree with Justin Operation lookup is just that nothing else +Tests pass etc +If you want us to progress this issue can you please supply us with the data Ulrich asked you for? CheersTony +for formula using within set collection +Even after changing the API from to +However I dont know if this design of the should impose a design on the Hadoop core +Its so big I dont want to have to extract it every time I extract the core +It also has the fix to the code in Aegis that would use the key type twice instead of looking at the actual value type +I updated the canons for those tests +Looks like its gone stale +I think the core issue is lack of documentation for the localized formats +Porting to HEAD complete +corrected patch name +Anyone care to provide a patch +Still this is duplicate of +Due to this it does not mean anything to add another value of it as the derivation takes precedence +Id first recommend that you try with the latest released code +Closing the issue +The root cause is inside does not work +This actually changes the functionality though +Slightly better patchNo need to wait wait in ZK longer than triesWith recovery is clean +Can you please commit it +discussing on irc Im not sure about the changes being made to the module +Checks if output path ends in a part dir and if not adds it +So I think we can close this issueI will keep the issue open some days and if there are no objections I will close it. +does essentially the same as what did before and it skips the automatic classpath configuration so that can set the classpath the way it wants +Generally because we like to keep the apis as consistent as possible across the different platforms +Note that any active Connection threads spawned by the listener are also shut down causing the socket on the client end to be closed. +This is done in at line +be that they changed the Spamassassin rules +These attributes are not implemented neither in M nor in M +fixed in +Simple patch attached +Support for attachment streaming will be included in Axiom and no changes are required in Axis. +Have you guys thought about what version should this publishing begin with +the server side client module is a very silly thing that does nothing useful but is pretty much required for jsr compliance +initial patch doesnt throw the exception for leading? when matchVersion contribs queryparser doesnt use Version so i simply change the backwards tests will fail because uses LUCENECURRENT when it should really be using some constant that resolves to LUCENE we can fix the test to use LUCENE instead of description to be less move issues to and +just committed this. +I committed the proposed patch in revision and extended it to cover Andreas idea in revision . +I just committed this +Cant reproduce maybe you should zip up your FF binaries and attach them +HiCan we have a patch so that we can give you creditThanks +The Abouut application is not necessary +Marking as fixed. +Moving to CR +However I see different results when MAPJOIN is used +Ahmed Thanks! I think this jira has the longest set of patches +Anyway we will incorporate the changes you have suggested because it is convenient for usersThe var arow goes out of scope immedietly because the method terminates after this +It is included as an eclipse project file for anyone who wants to open it in Eclipse +No need to force users to upgrade to VSDIGY +The failed tests are not related +Documented as a known issue in the EWS release notes +Created to revamp the tab in s the PR +no comments so far +thanks Galder for noticing +Tried to replicate but could not unless someone has the time to look into this one I would like to push it off until full java support is acheived. +There are now two v patches which is confusing +it should be really fixed now +An underyling issue was identified as the broker failing to validate the exchange declare arguments and close the session because they arent supported this has been raised as with a feature JIRA also raised seperately as to add support for arguments. +Patch to fix problem +Thanks Xuefu +sUsing default theme +Thanks RobertPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +If those duplications were not on the same file Sonar behavior would be exactly the same +editsThink I finally see what happens its perhaps mainly a test issue +Thanks Arvind! +Just upload +The class is now public additionally a null value may be supplied. +Closing issues up to jBPM BPEL +Done in trunk at revision r +This issue can be closed now I think as the dependencies are now listed correctly +Packaging things this way makes it much easier to update your code without restarting Hadoop daemons +Sorry +Its written as default which is no versions +I just committed this. +However the problem went back after some time fresh workspace is not a solution +NOTICE files were updated in M to new wording which allows releasesOnce the spec is finalized and released we just need to remove the JCP references in the NOTICE files +I should say that the originating URL isjarfile! was ignoring the path of the URL and just returning a handler to the jar file. +My mindI only test against null not against an exceptionSo let it return null +how do you like a logtail command? This can wrap the tail f datalog command and color the output +It would help if it caught and returned the same exit code in any situation where Nailgun itself isnt crashing +Thanks Luca! +Hi Paulex The fix is fine thanksBest Regards by Jimmy +PavelIve just run your simple test on latest harmony snapshot and it passed both on RI DRLVMAre there any special conditions for its running +do edismax parser supports operators? +I have added this property to the table for the book. +Im waiting to hear back about a specific fix from a build before its released. +Native component can be loaded only once per VMb +Patch of Martin looks good +Be aware that this event is deprecated and will be removed in a later version of Seam in preference of the events +I haveor will be soon commiting something that will break it from building against +The patch is for so youll have to change a couple minor things but not much +relate to test the work around provided in pom file is already configured with the workaround for the latest Jasper version and this error comes up when maven tries to compile reports that use groovy instead of the java default +Not your fault I missed it too apparently! I will take a closer look this weekend to see if I can figure it out +Greg mentioned we need to make revisit this before about +HIFreemanThanks a lot for your comments +Patch adding missing Charlesdone at r +Guillaume This patch contains for modules and configs and +I have changed the code to also send down an invalidation command after it is wrapped which fixes the issue in my test. +Nothing in itself requires a particular order +Ive applied a slightly modified patch of Cyrille +Pls make the shuttingDown volatile +Changing the existing hierarchy from Exception to shouldnt break existing code +This syntax allows a column constraint to be placed on the new column within the ALTER TABLE ADD COLUMN statement +Im attaching this file as it has a partial fix but it is not complete +I will implement it later on top of your patch +Reopened to set the Fix Versions field to JDO beta. +Hi EverybodyAnother observation +That is the best we can do unless we can get a better test case that can actually demonstrate the problem on trunk or +Does this mean we can remove the file from the tlds directory +pushed +Just a noteIn the row attribute is using the JSF default value Im not sure that we should change this default value to in +Vlad can you share this code somewhere +the test cases you are referring to I suppose should be unit tests right +Even though we scheduled this for Beta we ran out of time +Maybe it would be better to make gettruetype a method of ttype and make the tprogram version just be a wrapper +just use ant It will skip the tests +A bit more detail on my problem I have the menu rendered in a border page and each page defined in is a subcomponent wrappered by the bordre page +This is similar to my patch for but it is against HEAD +Henry Im also working on which will fix a number of the intermittent failures on hudson +well but logging out the contextPath would be nice nonetheless wdyt +Fix version was fixed. +PaulIm not sure because there is no surefire output to check +But unfortunatelly another related bug appeared instead +It does not use any of our stream or buffer logic and as a result our logic in to see data does not work +Could boil that into there if it ever gets done +We will go with the optional interface option thenRolandThis is another case where we have to agree to disagree I personally see ability to bind to an arbitrary socket as a common function shared by all socket based connections which should be captured in a public interface of some sort +and not the Codec as was suggested in the beginning +Also the very same checkout that was failing to build before for me after a reboot it built fine +We also own the +ssp rendering is now only to times slower than java +shasumcecbbafbcdfdbdfcda The class did not change since creation of this issue so I generated the patchAll looks ok for me +Lets discuss this on the list +Resolving as obsoleteresolved by infrastructure. +Our Hadoop team is looking at the issue +Edit Browse +By default it will map M chunks and up to numcores on a system so you can expect G of virtual memory usage in any highly threaded app +This part is not portable and we will also not fix this +Committed at subversion revision +Ive encountered this problem as well +It may contain some additional dataIn your patch you override the endpoint setting by the header because Camel will return false in case the header doesnt existYou need to only fallback on the endpoint setting if the header is missing +new version with impooved logging and excetpion for nutchbase +Closing. +Weve seen this issue only for some documents in a large corpus due to filter +Apologies for that +jrelibiclient +Lets look into this once we have such a requirement +In that case would be an option? That is the current domain name +Thanks for the patch but htmlparser is LGPL +needs a plugin parameter +Forgot to tell you that +All the Exceptions are now added with a Event ID and their respective iN messages are also keyed off of this Event ID +Because any of following workarounds helps changing SMPC to +Ah yes thats what I meant to say if you have ideas on how to process with this one Id gladly accept helpI dont yet have a plan on how to tackle this one just a general idea of how when implemented it could play nicely with the other parts +When I posted same file twice on versioned form the page expired +I have one comment on the patch +Im attaching those changes as patch here for referenceIme leaving this ticket open on purpose so that it stays the main ticket for all the to get this to a final +Attaching +I will update the patch and submit for review soon +Teiid went further to refine planning and removed the associated flag from the +Wow that was painful +closingthanks. +Triaged for. +ThanksTo clarify this patch allows the template files to be located anywhere yes? But not the? Or what exactly do you mean by pages in this case +I ran tests with the revision that added +Per review meeting beyond work pushing to future +This is because the process launching in the JDK explicitly prevents them from getting the workaround is to use FFI or the spoon gem to call the posixspawn function which will launch a subprocess that does inherit the tty +Let me look at +Committed to trunk and. +Ability to use on a producer method would be welcome +So I am fine and dont need at this momentThank you. +Thanks +Hi Robert can you take a look at the bindings? ESB leaves the service definitions IN but will unregister the bindings +Please add a testcase with miniDFS if there isnt already one +Thank you Issei and Alejandro +So maybe you did not notice that bug there because of it. +doest need this patch there would be no reason to hold it for it +Fix a findbug warning and kick off an unwanted change in and resubmitChecked the report and locally prior to patch submission and didnt see new items on account of this patch +Tuple comparison should be changed to compare TINYTUPLESMALLTUPLETUPLE +Bulk closing of issues that are old and have been released or marked invalid. +I would like to continue the discussion of to +Perhaps it would be better to do this with python and lxml or something +How can I test that? There is no such parameter inthe descriptor +We think it is not necessary for the deprecation map approach to handle every single deprecation in configuration +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Id think srctest +Alexis this still required? If using Spring this feature is already available and xworks ioc has been deprecated +I hope to finish this in a matter of weeks at the moment Im getting to know the internals of Giraph a little better +Will be in +Jason can you briefly look in to this? It appears that the is producing the string +Fixed in +pls check again. +Fixed for for the Servlet as well as for the Portlet will be available in tonights snapshot +Here is the unified patch + is a rejected is an invalid test that should be removed +Hi KnutCan you tell me whether these properties are dynamic or staticThanks KimThese properties are dynamic +Patch applied inI only changed the DTD version to to apply this patch patching the xwork dtd remember that Struts has its own dtd that needs to be patched as well. +Let me know if it does not. +By the way Where should a property for crawl be located in the? In the fetchersection? In that case please somebody move it updown or rename the property before including it in theBut could somebody please review it quickly? Im not sure its correct +Unescape would look the same +when there is a new plugin release we can use this that is fine +And really there should be no quotes on anything anymore at this pointBasically it should work like BASH does eventually +Paging out to disk only happens when of the memoryLimit are used +I have two pretty big projects +Yonik If you check the code you will see that the code added to the new class is almost identical to the one removed at several places only some cosmetic changes +broke this patch slightly +Yeah that would work as well +check on of a bulk update all resolutions changed to done please review history to original resolution type +Lets see if Ian can squeeze this into was found in but I expect it will be an issue in Jopr was unable to reproduce this on either Embedded or Enterprise +Do you have a complete example? I cant reproduce with the snippet youve posted +The old file is still supported but its use is discouraged +Ive just pushed a new snapshot that fixes this issue +Thanks EliAttaching patch addressing your second NITThe first NIT is not a NIT the property name is there is name aggregation of serverservice +Fixed the nits and committed to branchUnfortunately Mockito doesnt provide that Answer implementation but I based that implementation on a suggestion I found on the Mockito bugtracker. +As for the counters block I can try and see what I can do for it but I am not a web expert I know enough to ask questions and break things +I am having a hard time understanding option and it may be because of the assumption of the current derby base tableindex architecture +Im sorry for the confusion +URL in browser is httplocalhost +Verified upgrade in ER build +Gary could you specific in a bit more details which elementsattributes I need to add to make it work like you suggestId be happy to perform a batch of stress tests in our environment based on your suggestions +Any hint anybody +Can you review the commit too many commits on the JIRA for review in this lifetime! +for an optional setting to strip hostname and timestamp or not +Also Option that Suresh proposed above does scan in a different thread and it does it correctly +In order to make the CXF solution simple I think youll need to have a fix to this that doesnt involve the CORS filter or extending +Sorry but that wont do because you only have the responseMessage in the +This is due to a problem in the DAO implementation +Either we add some generic form of properties to the Netty subsystem or we can create a subsystem that is specific to the Server +Please feel free to this issue if the documentation is insufficient. +if we want to provide mechanisms to test arbitrary user extensions it should be supported and documented in srcjava likegood point I guess you are rightanyway I think this is ready to be committed though I will hold off a little and commit if nobody objects +ChristianIve updated the Libraries page and moved my notes about to the Wiki spaceAbout the manufacturing reports Id really dont mind about them nowThanks for your great help +Make several classesmethods public +If there is a particular use case that requires running gwt setup more than once please let me know otherwise I will just close this +The abstract issue API to use would be Mylars since a lot of work has been put into that and its various implementations check the Mylar CVS +Sreekanth can you confirm running reliability test with the retired jobs feature on is working now +I thought imports were OK for things like junit and Mockito in tests? Looking at the current tests we actually use Assert +in jbds +No suggestionscomments so far +Canceling patch to incorporate offline review comments from Sharad +This is a unit test to verify that the network server installs a security manager with the Basic policy as expected + doesnt exhibit this behavior +In fact its better if the test does have its own version of the constant as then a incorrect change in the code would be caugth be the test +Since this will have to depend on all api modules I dont know if it should go in or +If there is a problem please open a new jira and link this one with that. + +Ivan can you please post patch to jira +Amir have you tried it +The new javaee module takes over ejbx. +on commit + +not a reopened to change Release Notes field. +After has been integrated I will prepare a patch and do some further investigation on the problems I faced +Once the protocol if fully implemented and documented When will that be? Since Apache is a volunteer organization its hard to say when it will be completed +Thoughts? +Now that work is well underway with AS all previous community releases are +Range works fine in this version +Got your point has sense +Hi JarekI have already written few guide for client API +Patch with a couple more tests and a entryCommitting to trunk shortly +Connector category is for jca connectors +The error still occurs in the exact same manner stating that the principal is anonymous and not foobar as also saw the error occuring once with just Thread running +Applications should use an instance representing an empty value in list data rather than an actual null value. +Actually the dependency should just behdfs depends on coremapred depends on core +Thanks for the contribution +And about clipArtAgain we need an which will get a list of available picture urls +Theres still a problem with old index directories not being cleaned up and strange replication on start up +I need your view on following points +group Hibernate group +So we do that on our own somewhere +I dont think that would fix it hanging but it well have a chance to look at your script in more depth a bit swamped right now +lets get out +this is a cut at it +RPC server runs using a single threadI think we are better to assume multiple threads +See for further details. +Hi Vincent Can you provide a patch for trunk please? This patch looks to be for VFS +fix some logging so it has more detail +Is this still a problem for you +In in the failing case we dont find a faultWanted +This should be fixed as part of other fixes +Fixed in SVN +But at least now we have a Maven workaround + +Thanks +This is a bug in the patch ofAlso I agree that we only need to register leader once which we can do in startup before the elect statement +Yes! The only way I found to resolve client invocation problem is to set apache bind port under +Patch removes author is applied +Im linking this issue to +Anyway i will try to debug the code by the way i will sent you the DEBUG level log as well +Sergey Im curious whats the reasoning behind flushing memstore to a single L file rather than splitting the memstore into the stripes during each flush? Keep flushes faster less files etc +Currently there are three tests in the Repository with more planned for the near future +If we excluded known bugs entirely they would simply disappear from the report +Whats the recommended workaround for this usage in Spring Security? Thanks +if we add a new package it should have a with at least a description of the package +Re options and X I dont have any that I can think of immediately +Thanks Thejas +Xircles does because its non standard +When a container fails the only way to diagnose it is to look at the logsFor containers why isnt this enough along with diagnosticsbq +If we do want to do an comprehansive index please create a separate JIRA +This issue is currently blocking the release +I would like review and suggestion for this connectorThanks in advance +I filed a separate bug for JAXB so that its even anyway thanks a lot for a cool piece of software +Variable state on was not initialized so the exceptions Pat observed +Heres some example threads from a dump +just noticed that svn dirs has conficts +Ted Stack ChunhuiI think we may have to combine the change in as Chunhui suggestedJan +Fix is two liner +I think you might need to report this to Sun +isolated the moment it is unlikely this will get fixed +If you move to it will already work +Is this still an issue? If it is please attach something to provide an example of the problem +Lars is not in So in spirit this Jira should be able to go into What do you think + This bug has been marked as a duplicate of +I saw this sentence in the newest revision +This is not how the code was previously +The has now been fixed as part of the work for in revision . +User hasnt reported back since April on status and whether a newer broker versions resolves things or not +The result are instant for me +I looked at the nightly build and this is corrected fromMy apologiesEdgar +Sorry to give you the run around on this Milos but Im confused +which gave me the region server mapping +Thats an essential feature for using sonar in big companies +Actually this has to compile with hadoop otherwise we will not be able to deploy this internally at FB +We can always reopen later when we have some sample code that illustrates this problem. +Integrated to trunkThanks for the patch Todd +The remaining time is spent scanning the Seam jars as well as a whole bunch of other jars from the wars lib which definitely do not contain +is not in believe this should be fixed by +Stale dirs cleanup can be done as part of a issue +The patch looks good +Partially duplicated by. +I perfectly agree we dont expose enough details when things go wrong +Keep in mind you have the ability to specify java scripts that can be run in the footer +It has been applied under revision +This is quality work +It should call the showcase action and display +Can you try ant injectfaults looks like that is the target to trigger compilation of fault injection which sets up all the build dependencies +thanx erwin +fixed some warnings +A test case is attached to the related issue +Okay fair enough +Stefan you cannot update a key +NET using RPC style services instead of being stuck with docliteral as our only option +Applied in r thanks for the update Dhaval +It would be nice to have this available as a background process with an ME buttonmojo call to start process would check availability of javadocssources for all the cached artifacts +Well do this so just keep it unchanged for few days +I have implemented sorting by frequency +Integrated to TRUNKThanks for the patch N +Mapper class needs to only be in the job jar not also on the classpath built by binhadoop +I will post the code for the uber codecserde in a day or two which will enable selection of compression mechanisms +I hope this is what you want but if it is not what you want please let me know +Extend to verify the order the maps are executed +Any chance to stop building after cmake missing message is thrown if hama pipes is necessary? So user may know that cmake is missing +Tests pass running +The attached patch does this +What I dont get is why we are marking a previously successful task as obsolete and invalid upon the completion of a new task without first checking if the new task was itself successful or not +Yes it would fit our needs +All items addressed. +Should we make this one or should it keep it for +I can confirm that this issue is fixed in +It is hard to tell form this output what is really going wrong +I am not sure why there is a difference in how and when the row is fetched from TAB but may be explore that difference further more +The test failures are being caused by extra spaces and character ordering +Added test +However I cannot get a working example running +Closed after release. +I cannot reproduce the am I missingIll give it a day or two then mark it CANNOT the SVN revId for this is place and on the same directory and run mvn specified dependencies files are mapped into one file +Attaching patch generated by git +replace the first line of activemqbat echo off with echo on and post the output of an execution to see which part of the script has a problem +Is it possible to specify the classpath load order of existing dependencies? I have a case where maven installs deploys work find but does not +There are findbugs warnings of which are existent in current trunk +using source I disable the new important features of the jdk I think that for jdk this bug is realy a Blocker +Eclipse log was definitely removed from upstream +Nice +Yes thats right +This patch implements the feature +Merged and updated trunk +We should do this provide only warning the JMS is not available yet check if debug is enabled in logger then show the that it is OK now. +Dont see that failure in my test suite because the test doesnt exist yet +Ok +Latest patch +This is fixed in head. +Ive updated the Javadoc with your text. +Committed to issue as it has been released as part of release +not in a position to test at the moment +The and now use this more efficient old transforming behavior can still be obtained by setting a alwaysTransform boolean property to true +Thanks for the patch Ronak I modified a bit before committing but the end result should be the same +Im also taking all of the above to the Trinidad list +Thanks for the quick fix +The original patch looks good +i know thats far to be module are not in the module so javamodule doesnt work +Hope it gets promoted out of the sandbox. +linking to jdf in JBDS B +Thanks +In all components are generated +I run withxc O sourcepath +Running a full Spec JVM for iterations since closing now completes. +Fixed Renamed to dbscripts and updated referencing files +First change the file to log a DEBUG level so you see the exception and then change the file to have the desired max +I committed the reflection patch but am not yet closing thisHairong will you provide a patch replacing with +I ran the webhdfs tests and as cited above tested that a does not cause a NPE +for testing related to the issue recently closed not to rewrite UTF to be working in +I will perform a major refactoring with release to benefit from enhancement of new sonar plugin apiJust one thing please next time use also the user mailing list when you encounter difficulties while using sonar dotnet +It leaves far too much wiggle roomHowever AWS is also PCI certified +etcaliases mightve been brked on hogshead so I did a newaliases to +Added in validation expression +It just seems significantly more disruptive to implement DAC via metacolumns than just through family meta dataWould be both implementing DAC via metacolumns and via family meta data so the metacolumn can be sparse as possible empty for the normal case at least providing this to the designerbq +fixed javadocs several javadoc warnings renamed open to create fixed file extensions in test CSF LOL! copy now deletes files that are separately written once copied into the CFS +Verified in JBDS H Fedora +maybe decodeencode? normbyte? getget? Something else +Just tried it on a different computer +Kos et al +Due to changes made for where the was removed as a parameter of the this issue has been fixed since all have been removed +commit master issues that are still marked as resolved but were part of the release. +Committed the changes to master +If I see no reply on this by beginning of next week I will move it to TBD +The modification for would fix the test failure mentioned in this issue +RegardsAndrea +Ideally it should have been more of a sanity check because we should only try to replay txnlog once +Closing issues in RC. +everything looks to be working fine +Just a little investigation might expose it as an important blocker issue for the release +Thats exactly why I asked carlo to leave this as optional +It contains some more issue types but Support Patch and some others are not here compared to default one +I just modify the previous code to remove some testing stuff and provide a better way to support the race effect inside google apps engine +then. +Initial implementation of json message object for sending messages +As far as I can tell this is no longer an issue as of Struts +i wasnt using openjdk but gcjwhen using openjdk hadoop dies after a couple of seconds quietly +Ok lets leave it as is then +Attachment has been added with description chinese example includes fo pdf before and after has been added with description new had a look at the patch and think it would be really great to get a proper Unicode international line breaking algorithm into FOPUnfortunately there seems to be a bug in the code which breaks the current layout engine test suite +For a total of images in the rotation +So this JIRA was not fixed as part of my recent changes to but because of some other changes that were made on trunk the issue is indeed not present anymore +I do not know how much that interferes with your release plans thoughCan I see your patch? I found the culprit to be a bug in our code again +Apache Vysper had references to smack xmpp library so i assumed it is under apache license +So even if the writing of the cfs file fails for some reason and there is a corrupted cfs file on disk it is not referenced by the segments file and thus an writer would not try to open it. +Purpose is to use this more for cross node group dependencies as cross rack dependencies should be very unlikely +With version Ive replaced the with a. +I am suggesting just fixing the error messages +for the backport +See info for docs +Right thanks for the clarification that makes sense +With the switch to the lucene lock files are now created in temp directory +Still dont have a final patch +Would it be possible to release it as a patch before? We really need this feature for deserialization of JAXB annotated classes +So dont expect coordinates to if the AS needs a fix it needs to switch to a Xalan release that contains this fix +Yeah thats a great idea +It looks like these improvements were made in +Tests passed cleanly for me on except for known heisenbugs in testPing and +So the plugin would probably live as part of the Servlet Resolver implementation and as such is perfectly able to use internal API It may even be created and registered from the activator method to have such internal accessAnother extension would be to not only list the selected script but also the list of script candidates +Thanks reviewersSmall final patch applied rev +to fixing things one piece at a time +Aaron I am not sure what you are trying to address here +For the record its the other way around I saw what you were doing to deal with char references in your code and since it was much better than the broken method we were using at the time I ported your aproach over +Need a new JIRA for that as its all together a bigger issue +Pending review +Ill let you know when I have a fix and as you to verify the fix for me on yourJesse + +I will check in the corrected import +Pavel I have received your instruction in private email thank You very muchIll try to resolve the issue in some short time and send the result for your approval if Im able +Lets try to get the models off the sourceforge page +See ticket +I just committed this. +So we cannot show per user data + +packages to boot delegation is probably a good idea +Workaround exists add daemontrue to the URI as commented in the issue. +Closed due to the release of Tiles. +Marlon you might wantneed to upgrade these sandbox modules for the Rave TLP statusmove now as wellAs reference look at the fixes forFurthermore it might be good to if or which of these modules still should be in the sandbox or not +signededsteiidexamplessimpleclient jcoleman soa Dec +This has been implemented and passes the tests +Issue is fixed and it will be resolved after test will be Test is added for the test is added +But no parameters to get a clear understanding of what has happened so far +But there is no way for me to do it currently +Jason Im watching and will try again when it is fixed because I feel the two issues are related +After that you can verify the connection is gone when looking at the topic page or by hitting Enter on the Pitcher againThere are two +I just have it on the same branch as the patch for . +Right now i use ignore it carries no impls +Yeah I think Felix approach is better +Missed moving the list element tag which Ive fixed in this patch +In fact reopen is completely unrelated to this issue or the cache here at all +Simply said debugging +Jason I have provided patches to this issue +Ondrejgood point thanksJared would you please update the release note? that passes across all databases with EAP ER +Whether to fold it into some other guides is a question for a future release +David could you pls verify +was just added to the end of the list of imports +Its a simple change included no test and it sure doesnt look simple to testI have manually verified it on a secure cluster +Does it need more testing +What will be harder is to do it on a basis maybe we can do that with a annotation on the method but this needs more is your use case here? instance wide TTL or on a basis? +As a suggestion why not alter the accept method on so that it just expects the uri to be a regex already instead of turning it into one? Then theres no chance of issues in the code and it can just be documented as taking a regex job done +I just recieved a POR EULA and will check it in +I am working on it +As is now committed this issue can be resolved. +Updated the patch so that it cleanly applies to the trunk +The title property would be difficult to implement in a vertical orientation +Tests pass cleanly for me +Committed at subversion revision +Added checks for java reserved words for paramter names if a java reserved word is used encode it by prefixing an character +optional means not needed for compilation +Resolving +Attached a new patch with a completely different joining implementation +committed to and trunk +Yea thats on me sorry +In our testing there were some slight property name differences and we also had to add the delegation tokens to the UGI in +It would be simple to add a XML parse test for the expatlibxml parser +Attaching a patch for demonstration of yet another approach of adding the Credentials to +Thanks to whoever fixed it +by nature if an operation is a fails there will not be any transaction there at all so we should process the transactions resulting from in exactly the same way as normal transactions +The has been updated to use an that does honour Transactional annotations on documentation remains unchanged from Spring and is correct +More tests are needed though +Giri committed the fix +passes also for hadoopI have a failure for on hadoop but it seems unrelated as it fails also in trunk +where is the Geronimo TCK Jira +Of course optimizing for this would need code + +Ah when I wrote that Id forgotten about the translation done by is a good mature product +Isnt it a duplicate of? +It supports INSERT OVERWRITE INTO LOCATION pathsubpath +So from me after addressing the nits Jakob and I mentioned above +I could reproduce it consistently +Looking into this nowThe build area is owned by the head unit it should be owned by the build agents +Will commit in a sec. +But default args shouldnt be allowed in interfaces +Thanks +Above comment is really for +commontests +So now my DDL has a bunch of errors in itThanksChris +I dont have a strong understanding of when you would want to use the provided scope vs +Ive a mandate now to get cascaded WMS layers working with basic authenticationDo you mind if I take ownership of this issueA cursory look at the involved classes seem to suggest overloading the geotools constructor to receive the auth credentials and storing them as part of would do the trick +Done +New Seam Web Project is vanilla one +Im getting a failure from tests pass now +yes I tested ithowever there is the note about whether to use a common ui label or notIf there are no real objection to my reasoning then it can be submitted +I just didnt want somebody else to start working on it if you already were! Great that you are working on this +Attached patch is a near complete implementation of the xsdoverride support +the removeAll is calling instead of +Includes new tests +Actually now that I look at it MRI does appear to scan all args +For now your actions were appropriate +You are using Commit Push which basically means Commit everything to my local git repository Push my changes to the remote git repositoryThe dialog is completely correct for nothing has changed hereFor we cant do the check without actually doing the push so you would get the error later on hereThus im rejecting this since it is working as expected since you are operating with git commands where this is the correct thing to say +Closing as the Commit errors are expected +Also you might want to test it locally before submitting it +Correct me if Im wrong Gary but if we allow any random name to be used for the deployment descriptor the runtime wouldnt know what to look for it needs to be a predefined name +Documents bodies are written to contiguous regions one after another the attachments are stored in separate locations in the fileDamien this is what I meant with my previous comment +New version of the patch which includes changes to keep +I can really go either way but given that this has been here for a long time and I have given time for folks to dispute my Id rather not revert it unless you argue that this will really break things for our users +Thanks Pavel! +The performance advantage of over outweigh any possible extra IO look up regardless of a size of a Scan operation +I did dropcreate CF on node with hints stored +There are still a lot of places where the deprecated box CSS styles are used +patch applied and checked in. +Basically the lookup in mysql for srs is just grab the first geometry in the table and look at its srid +committed created for feature. + on option +Ill create neat and clean dedicated code example +I would rather have a working version that works for most of the people than a documenation that says something but doesnt work at all +Try Hudson +jpeach please review and make sure whether the aio unit test passed or not +Thanks Vinod. +Missed one overloaded call to should in +width is fixed +If not then we can close it. +I committed your last patch you need to update your workspace before you create the next patch otherwise it contains all the already committed changes +If for any reason if anyone has to modify the testcase and change the input data set its difficult to change the static data +It really does not have to load the state of other shards +Attaching a script which demonstrates how this bug can corrupt data +PaulexI thought the decision was to implement the methods to do the obvious thing? Ive put in the versions anyway to get the signature checkPlease can you provide regression tests? that I think I missed some discussion in the mailing list +Please review the patch +Asking the queue for detailed information on every message to get these stats takes a fair amount of time and bandwidth which is why it is more efficient to have the system compute them and just transmit the summary +Fixed by r +Id appreciate any feedback +RESOLVED srcorgjbosstoolsvpemessages has been patch is attached +for closing this issue as well +Verified in Tools. +This presumes you wont be changing styles from the parent to try to affect the childI attached a +If you have deleted the output dont rerun the tests. +lets commit this before reorganize structures +the list of versions that manual tests are run against by QE is a separate concern from jbosstools automated tests IMO +In the fix explained above I used the same timeout value that is usedto timeout a connection attempt to the running slave +Thanks +So it must be the mentioned bug . +Thanks a lot for the review Todd +My thought is that it may be we want another suspend mechanism that can suspend and post an event whilst holding the lock on the queue where the thread is being suspended +I began doing the move and refactorings discussed here but this turned into a very large change +bulk close of all resolved issues as part of closing items. +Private method have javadocs as per needRevisited failing tests which were happening due to change +The results should be cleanerI should have realized this earlier but CC instances dont support instance store volumes only EBS +Patch with fixes to address the values of +I agree that this is not a bug +Atttaching the readlocks diff with more context to hopefully help in reviewing +Eric I just put this in trunk and +This is likely to be inefficient +This has been fixed by. +It does not check if the typed value is valid +Done +Reverted pending fix +This is done now Steve its up to you to get the Merlin wrappers working +HiCan you post the complete stacktrace for this exceptionThanks +I checked with table names which seemed to be the problem +Deferred until we have access to the platform. +I have noticed that it is always different from even when requests and rollingRequests are the same +So there is a lot more work to be done +I am attaching the example that should prove if transaction sending works or not +Screenshot of problem +so fix version is not correct +Unfortunately that would involve changing the queries in your appThanks we cannot use the foreign views as we are writing our own adapter to greenplum +Maybe there is another logic failure somewhere here? Sounds like it should get set when it discovers a literal. +With the version this seems to work +Ive committed an additional fix so that scores are returned if requested in the field list. +To further clarify the resolved link status would still have the isLink bit set in additional to the bit for whatever it resolved to +The following patches have more comments and a link to a wikipedia article +test rel +Please put some light on this +Applied the at rThanks to all! +Can you create either or phabricator entry for this +Cosmetic +No test case included since it is already covered in +Different kind of errors +I think we might want to consider moving this also to the code as described in to make it work independently of the situation a clean is Ill give it a shot if we can fix together with how this feature could be i would really want a confirm request hereScenarioAdmin types restart thinking refresh a very different system state will result from these commands +Any idea why this would be different? Just trying to ascertain if there were any major differences in the way we tested this +A PDF document might contain multiple EOF lines +Hi LonDid you already made some measures with a profiler to see how it impacts performance +Test fixed in +BULK EDIT These issues are open to be picked up +committed thanks andrew. +Thanks +This in Firefox IE and Opera +I have tested all my applications that had this error with cscs by using a customized method and the error never occured again. +It would help it you can include the error message and context of the failedA SQL dump of DB without any data is attached + scroll bar added to delimited string list control +how can this issue be marked as closed the bundle has not been uploaded to the maven repository +Should be easy +I was not sure if I should put instamo in examples or contrib +OK fair enough. +I know its obvious but being obvious always helps +Ugh obviously without the above just mistyped an angle bracket instead of a curly one +Syed how does CREATE TABLE AS SELECT manage to do it? Doesnt it need to prepare the equivalent of an INSERT statement against a table which doesnt exist yet +Hi ChristianI can understand your situation +Its on me to review this patch and then Ill commit itThanks for your patience! made a few small changes to your patch to make the logic a little easier to follow + +The patch is appliedThanks +I will let this issue open until others agree this is the final solution or bring up a betterfaster one +Pushed upstream +Cool +the main problem here is correctly merge changes +Thie current one is just slightly different and this causes confusion +Marking patch available again with same patch +should there be a timeout setup in wait in case a region is stuck and never been assignedSeems reasonable will addbq +one thing i can suggest try to recreate your workspace cause as i think you are using the old one may be prev version of eclipse +net is too old +the wfs spec pretty much mandates us to use the convention that omitting an attribute means null instead of treating empty as null +This has been heavily discussed internally and we just wanted to keep you up to date on our findings +Previous patch did not match up +Confirmed fixed +this task should be never be resolved +It appeared that our old had more importance that I supposed +User documentation is available qpiddocbook +Close this since it duplicates. + +So we never even bother loading the tablets for the clone +There is an existing Maven project that already uses Spring and has the needed dependencies +Will a patch be welcomed? It would be easy for me to write +ftl files this sound like you didnt do an svn add before the svn diff so that the files are not part of the diffbeware +These tests have been fixed in both CR branch and trunk +Already done by Guillaume inIncluded in revision +via Thrift +just installed and now it works without the +Also the zip is not really well named its indeed a set and not a list +Any chance of a patch Brian? Thanks +That is why I just logged this improvement requestThanks Arun Thank you for telling more of backgroundCould you write down a few specific examples that cause a failureif you could write down a few typical and specific cases that users could make mistakes we can repro and see if it can be caught to show further output +please refer this other than done should not have a fix version set. +When the xmlbeans were regenerated with the changed path Wsgen was able to find the new generated beans +This is not a facelets issue my issue here is with Seam because Seam is using an unsupported beta version of why is that allowed? shouldnt it only use supported stable versions? Is there no such policy for Seam GA releases +Create a server runtime with the name is the same as specified in the project in workspace then the AS runtime display all jars but ESB runtime CP cant +also added a bit more logging +Developers please pay more attention to the problem description +So far the run looks good +I would like confirmation of this approach prior to precedingI have added a few checks so event notification code is not generated if there are no add Issue the collected on commit and up commit and rollback events had to patch the into the in order to figure it out +If you give me a jar with the change I will be happy to test it on Zos +ClausYeah I just merged the check code back to branchWill commit the patch after running the test +It turns out that shutting down the mini cluster solve the tests automation problemEverything seems be working now +has a bug in their substring method when the clob length is zero +Move out to +Closing as the CND editor is now available and functioning +This was my fault +So I am not sure what you mean by your comment +Buildspasses tests +While I can appreciate that every community has their own preferred bag of tools to use the factis that AFAICT twig is yet another implementation of django templates with a textile filter +The fix looks good +So in general it seems like a good thing to set the heap size close to whats required +DainIm confused +For the TODO you have in the test can you please create another jira? Other than that this change looks good +I am currently experimenting with updateLog turned on full time and using autoCommit to keep the size of the tlog directory under control +Curator has been open source on Github since mid and has never received any legal notices regarding the name +Ive got the settings model modified to support this I think +Unsubscribed according to ezmlm let us know if theres any problems. +At this point in the execution the particular category method has been determined and the lookup result is being the first parameter of the category method being returned is a type parameter then we can use the value of object to resolve it +In case is not clear I favor a patch that adds a setQueue method to allowing a scheduler to change the name of an application +I am not the copyright holder so I dont believe I can grant a license +During the first invocation Eclipse starts Tomcat and deploys help files there +In most cases the exceptions outside of dont matter much because they will bring downbq +Thats what I am expecting +Thanks BJThis should help +Ill attach the lists to this jira +Dont want added to +made some changes in commit +Will test and commit +Leaving open while I work on the patch +Thanks +Fixed branch at subversion revision +the page with chart finished added favicon unnecessary pictures and classes removed javascript file moved fixed a style of the my mail on what needs to be done for you to move the app to RF codebase +please see the attached log file +Updated and fixed version. +I checked the usages in xbean and geronimo +What exactly did you include in the exported server? Did you use the console or the gshell command? Can you produce this problem with a gshell command lineThere are a couple of scripts to build and export a server starting with the framework assembly in pluginsrollertrunk and these might help to figure out a command to reproduce the problemThese lines should be in varconfig not +Patch with fix and test +Ill play a bit more with your war in as expected in CXF please open a here is how you can confirm on your own it works in a standalone Tomcat use the attached and copy CXF libs to libOnly temporarily update application path to a till CXF gets I confirm running the sample in a plain tomcat I connot reproduce the issue +as I saw it work with Aruns new +camel mvnrun works fine +Migrating this project from the sandbox to the trunk is on the map +Silly me +This has to affect master as well +Is everything in place for the release +Please review the patch and give me your comments +Done with. +But it seems to me that a lot of developers will generate such queries +to trunk +Please try M Vladimir we made substantial advances there around CDIEJB integration +I can be hooked up so that the allow the saving of temporary layers +Thanks sam. +Keep up the good work +trunk +Corrected thanks for pointing this out. +Revise to have its own implementation without any external dependency +I suspect that to make the logic easier we just implemented things so that the move was always done since it didnt seem to cost that much most of the timeObviously this patch is just a bandaid to fix one of the more pernicious problems with this approach and we should still see about tackling so that we can eliminate the move entirely when its not necessary +Then a user code that does not explictly control Broker may get a hook during broker initializationclosureExposing list of listeners does raise the concern about one user removing others listeners but isnt one of the prevailing themes of architecture thrieves upon the users having deep visibility and customization of its myriad internal features? + +The moniter timeout is minutes There are enough time to close a region +used mem +The extra work done by compress table isnt something thats generally useful during compression is it? And there still is a manual step needed to get old databases to pick up the fix for +Thanks +You beat me on the answer +Let me fix it in too +Do you have to be root +I think I remember some issue with the order that tables are looked at being related to hash codes which might vary from ddl to ddl +You can indeed ignore the messages you find in the logs +The id should be removed from the UI and the ID should be made to be unique by using a fully qualified domain name +I can do this but itll take a fewdays Im on vacation +always is linear single threaded as before. +I suppose I ought to this bug for the +there are potentially different map implementations for example im personally intensively using the to UDAF which in my case would be much faster than using an order by clause in my query +Peter Id rather we create a separate attribute for that then to use something like path which would not be very intuitive to other users +I am beginning to suspect that this project is no longer being maintained as there has been no evidence to suggest otherwiseIt is odd that the patch has been available for a long time has been verified by several folks and the project team has not commented on this seemingly critical issues +Filessrcmainorgjbossjmsmessage srcorgjbosstestmessagingjmsmessage +Verified fixed in ER build +Need to create a sample application or extend an existing one to demonstrate the use of the Valang validator +The exact match return URL check of Google is not suitable for current implementation of oauth feature +I just committed this. +As long as it is before gossip it does not make so big differenceHowever I dont know what this check would achieve in a move operation +On the other hand pro and anti may be adjectives in English +Works now +You guys are right nice detective work Roshan +Gavin I can send the merge notice if you want +We need to mention this somewhere in the readme could reproduce itIve committed a fix to not use you update your fork and give a try please +I appreciate the effort of this patch but just providing new code is actually less useful in most cases than providing a real patch that can be applied by an IDE +The default used by JPOX will be mixed case quoted but you can easily set that property to +I looked at I still see it +I attached the result of my the repo in my last attachment wasnt right but I wasnt able to find a valid SNAPSHOT repo for wicket either +All the regression tests ran cleanlyCommitted revision +So the behaviour is undefined. +Now I want to thank you for all what you have done! I have finished several project based on struts in the past monthsand I found richfaces is much better than that uglyframeworkyour response is so again +BJ is interesting how to you reproduce? Could you please give the steps used +Thanks AdrianI tried running the Hadoop test with and without this patch and the execution times were comparable so I dont think that theres an issue with. +Providing a fix +Output of the test patch solving this namespaces xalan issue. +I agree with Glens suggestions for new methods and his implementation direction +Its been a month with no reports of further problems so Im closing this. +Thats true +Executing with an incorrect category name would be a configuration error and I think its important to signal this to the user +Rolled back in revision +A simple patch to add obvious additional metadata fields +Julien any sight on progress with ? issue is quite a peculiar one +Captured throwable Trace logging retry attempt Additional doc on Circular iterator I thought it would be a convenient pattern for selection and in fact makes the code simplerclearer +more replaced +None of the options to thejavac compiler change the language in spec in any way +The issue is that the test assumes unix path separator for a local path +I cant figure out what triggers this bug but Ive now seen it times in the past weeks +Making a will break exisiting applications because of the changes to theclientIds of contained componentsWe should not change this in releases +the error message from eclipse logs The hibernate configuration file entity class and which is the wasnt sure if the enum should be included in the mappings so I did it but I have the error message with and without mapping record for was not shown cause Hibernate throws not as expected +HelloWould you please try my patchThanksBest regards Andrew Patch committed in revision +working on this one yes want to close this issue for be in console +We dont depend on so we dont mind if it is deprecated. +Attached is for review +let me check whats going on +But it was generated by the mapserver shptree +A very long time ago for a very short time there was an effort to get the DB ODBC client working with Derby but that has not been supported for years +I had to tweak the test though +We should do something about that issue or at least put a notice on the website +It is receiving quantity of the Since there is no any serialized product available in the current database there fore for now the code is commented +In my opinion patch is applied successfully +Not a bug + requires script also reveal that we print null alias for error message +Please use the forums for user questions JIRA is for bugsfeature requestsIn this case you should not be packaging jacorb but should use the profile that includes the jacorb subsystem +Making the tests will be straightforward + got the wtp integration now +I included the new icon in Tools. + +Tested +Im sorry I copypaste env from firefox since it was less complicated than copy it over the RDP from lab +There is also a launcher named for launching Felix with all dependencies and bundles +This fixes active expanded from robert applied with modifications +Since thats the case theres no prefix on the name of the node so we look for element nodes with that node name but without any namespace +Karthik maybe we can put the above description in and I can update the + +Also you can try to run mvn with the X command line option to run in debug mode +Here is the patch for this issue +Reattaching first draft of a functional spec for this task this time granting license to ASF +State transfer marshallingunmarshalling logic has now been refactored out of. +Thanks and sorry for reply so late +I noticed something changed between and +Then I ran an evaluation algorithm on the entire data set and this smaller file both times using the date range of the smaller files +could not make them working on hudson as a moduleejb runs are pretty clean with mysql except for failure which is a test issue in a some test side changes ive got it working with pointbase and bug this test revealed much earlier was already fixed in. +This would preserve tons of trouble and headaches I had supporting other maven users with this issue +Source files of the example used +Indeed we probably want to be able to automate deployments and some administration tasks which is not possible through a web interface +Please have a look at the version in RC and let me know whether this is sufficient for your purposesIn general I would prefer to stick with protected template methods for the time being +Manually tested +This class has been in use in our production system for over a year no problems reported yet +If that works then it might just be using htat flag to put your jars in front of the others jars +Derbyall passed on Java SE +is not thrilling but not bad either possibly the best of the lot +Current patch cancelled because it doesnt apply cleanly +so I am assuming that RFE would then need to be created to investigate this issueWhat do you think? +I have converted Java services into minilangThanks to Brajesh Patel for discussion on the issue +Fixed in rev +Thanks +so depending on how much backwards compatibility youre after this might not be needed +is to do with modules not being loaded when using CordovaThis issue is to do with different versions of Android irrespective of the Cordova version + +currently only handles names as strings but as can be seen in the error report names can also be +As the Quartz bundle is already bundle we dont need to provide the Quartz bundle in Servicemix any more +I leave the cache for now Weld doesnt provide any caching at this time +Do we have someone representing the ASF on this JSR in the EG? Yes Kevin and Pinaki are both on the EG +I am on the Aug patch. +All tests are passingThanks Doug that suggestion worked + to trunk +emailed maintainer resolved this as requested. +StackAs part of this fix also needs to be applied to +the patch looks good +But security handlers cannot be invoked until we resolves policies +information have been corrected methods have been described +I added the original case as a test and committed the lot in bebe +Thank you for looking into this +In particular checkNot done considered for this JIRA +Please add a test case that verifies that this is +copy of the generated dita file but with the leading spaces removed manually this to make comparisons with the previous version easier +Add a Forge keyword reference so the wizards can be searched for using that keyword +Reresolving as wont fix +Cannot in +So the num rows printed was wrong and I missed a decimal point whenI was adding the time in my head +name to +Could you make a patch or pack those files with directory structure? So assignee need not arrange the files in relevant folder +Ive also simplified the code +trunk +Radek has been approved via the CLA and has licensed his source as GPL + for the to trunk +Currently it seems only jar dependencies are packaged into the war. +Provided it doesnt balloon the source API I like Bens solution +Cant reproduce and no further feedback. +If not a rowvalue level there is an existing facility that you could useabuse for this purpose +Thanks a lotBest regards by Richard. +but eachDir is a convenient method true just some rare uses cases and not much difference from eachFile some lines for the additional still worth an inclusion +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Alexey it is still there +note If I switch jstl to provided with the current version of ROO ROO will add jstl again every time I add another controller such that I now have two copies in my pom one provided and one with no scope +Created a Multimap interface and several implementations to replace the corresponding classes in the Google Collections library +This problem started when the Ant jar was removed from the lib directory was fixed by fixing the patch to. +I will backport tmrw +Youre essentially suggesting that we shouldnt rely upon TCP to implement even its basic functionality +but there is another problem always lets the cookie version to default value +patch looks good +has been marked as a duplicate of this bug +No tests included because its a doc update +Please vote for this issue to increase its chances of getting into +Unfortunately the result then fails for Hadoop but only in EC environments +A couple weeks is probably not feasible +The reason for this turns out to be ls itself and not fuse or the kernel +As it is now are commented out in the testsuite +So weare back in the same situation of not knowing if the query is a bug or if itsjust that no documents with foo indexed currently exist in the index but mayin the future or may have in the pastSince Lucene is a search engine rather than a search server we should strive tokeep it as flexible as possible to enable that many more applications to bebuilt on top of it +Done by Alex long ago. +If this connection is cached can we arrange to use the connection obtained for transaction recovery for inbound message deliveryI have back ported the fix to for use with Geronimo in my test environment +Now I cannot use this workaround anymore because If I dont use the srcmainapplication dir then I get other issues with the maven eclipse plugin and the WTP features +classes in the first placeClasses in sun +Refactoring of Tests and Comparison operator bug +Cannot fix assembly transitive filtering where there is an overlap of transitive dependencies until this is fixed in maven itself +However I think tasks whose execution has ended should be deleteable to deal with situations like yours where they have become up for your work Ronald +The issue is broader than the statement above and not really solvable without changing the pool contract +attributes passwordnow support. +I had the same experience as Rene today using Grails RC in Eclipse +Yes we set the should to be true +Attached is a screenshot of IE running in IE Standards modeThe tooltip is correctly positioned + Bug has been marked as a duplicate of this bug +NET service when using +No probs +Should we not close this? +The should be able to live completely seperate from the jBPM context in a pure detached mode +Please verify and close +Even better would be remove by regex +These changes look good +Easy pick a project any project in fact but the more dependencies it has the better +So I can make the change to to revert back to +This assertion needs a lot more justification before I would agree with it +as Koen says its not easy to do this without thisyep but the fact is that depdending on not exported classes from the JRE is like a bug +Thanks Vinod! +eg +Done in the bugfix branch +So we chownd +So that would explain why you dont see a performance difference +Marek is CR affected as well? Can you update the broken link? I am wondering why I have never seen this +More info added +It would surely be useful if all params objects were cloneable +And leases rarely actually expire so the cost is in addingremoving themThe only concern then is when some client forgets to close scanners although I think wed have other problems then +java +Thank you for detail information +Horia do you still have the I sent you about a week ago for testing the failing import bugIf yes then import this one and then try to get the brixglobalContainer node and you see it failing +Thanks Regis. +Especially since it works in deployed applications. +locationid +Bulk closing for +The file attached to this issue contains example implementations submitted by another user. +Cant we check the number of versions for the entries and make sure they have the same number into the same bucket? That way even if we a missing row we should still be able to detect that it was missed previously +Uploading which removes the char limit for VARCHAR default strings +I believe this will work for you +I use in production so Id be happy to help trace this issue if you can provide some sample queries or some test cases +Verified in JBT L +is field needs to be updated in the shutdown method of the class of the +I think assignment operator isincorrect too +ChrisThe last date for feature proposal for was and that needs to be done on dev mailing list +I will try to verify it on qa machine +Triaged July Unticked Release Note Needed +Are there any hints how to improve the performance on rendering or sorting action within the IE ? Using Firefox is not an option for us as the customer requires the IE +Lets say my development machine is called duck and my production machine is called Server +AFAI can recall it was and wehadnt tried duplicating this problem in tomcat or another servlet containerFeel free to close this out for now if Im the only one who has experienced itI can always reopen the bug when I get back to the project +now takes a Writer and now takes a that if the ReaderWriter passed to the constructor is not Buffered it will be wrapped +Please link the new jira to this so that we can have a track +Added selective database exclusions for XA datasources for and extensively against the open source databases +Hmm I cant download the attachment +Subsequently upon learning the average size of maps the number of maps to request is determined using the maximum size +Thanks! +This is now fixed and the sun transport works +Also blowing away the plugins directory does not solve the problem +Also add tests for all the cases aboveYou should be able to compare the plan via a explain Paul Butlers account name from the JIRA change history +For example Ive seen application installers that set CLASSPATH globally +It also supports versioning. +This means we can +There is also a Teiid Deployer which creates a directory under deployers +Added request to upload new version of tools +Committed some unit tests to reproduceThanks StefanAs you mentioned it in a previous mail I agree that the should be rewrite +A nice solution +Oops I didnt notice the sequence of events it was reopened nick found a fix and resolved it +jarExtracting systemorgapachefelix systemorgapachefelix systemorgapachefelix +Remove old related +What about a compromise and have both the plain Java object and a builder? That way if someone wants the simple clear way you describe they have the ctor they can go to +Thanks +Let me post my patch as well +For the record has also been upgraded to this version and has been rebuilt to pick up the latest +However the use case for this JIRA was to be able to run a normal test in FI environment when such situation reveals bugs in the code +If you need to buffer anywayuse get and copy the stuff into a byte array or so +New patch incorporating review comments +I then selected from the same table accessed by the first client which provided the same results as from the first network clients select and then used the Disconnect function from the File menu to disconnect the second client that I opened +This allows parallel deletion of the contents inside to +Thats a workaround but its certainly not intuitive +This is due to a misunderstanding of how RELAY works LON and NYC are completely autonomous clusters and dont know each other +Build is stable since a couple of days. +I can imagine some mobile +This would remove the dependency and would mean your team does not have to continue to maintain the stylesFeel free to port this issue over to JDOCBOOKSTYLE and comment what other items you would like to see improved +I am leaving the issue open as we might come with specific optimizations but its not a major issue +and one more advantage to using the static inline function that way you can get away with defining things only in a header file instead of also having to create an extra +How is this +defer to +It seems that agregation does not take the classpath of the child when building the documentation +I cant imagine that normal people are going to be interested in seeing a JSON blob duplicating all of the status information thats already formatted for displayI guess its possible that the definition of progress changes in the future +I dont think it is really a bug either +I can confirm that this happens exactly as described to me as well almost identical environment +Why is Fix Version set to if I can already use in? Is it just because it is not available in Designer? +Ill try to check it out and build it +Hitesh and VinodIt is not a big deal +Contains CTF test case as well. +Please reopen if this is still a problem. +Thanks for the SVN link Jiri +When closing a modal window its close callback at the main page is invoked targeting the original main page version +gridchukwabin +Attached patch implements a connector based on the map repository framework that uses one Infinispan cache per workspace and uses each Infinispan cache as a map of to +Thanks for the patch! +Come to think of it maybe should be moved as well but that I dont feel as strongly about that as I do about sleep and failThe test dependency on the example jar was unfortunate +The Struts beanwrite tag has a nice attribute filter which defaults to true but can be explicitly set to false disabling any url encoding of the string +Closing as the version in EAP Beta is the latest currently available +For we only need to implement shims for metrics sources so interface for registry and Mutable classes would suffice +patch solves the problem also takes care of problem of superstate incoming Linux +Thanks Navis for tracking this down +Can you attach your script and give an idea of the data sizes you are trying to join? Of particular interest is the cardinality of the keysAlso if your inputs of very different sizes you can try reducing the join order +I understand your initial reaction this isnt the sort of thing that I would ever want to put into the +It doesnt have similar code any more +Hi GertThis solution doesnt work for me Im on linux box but my local repo isnt in default pathRegardsFreeman +I just committed this +Scratch that need to work out why this is happening. +Thanks +we should have two separate classes to do the action creation parts +Fixed in subversion repository as of rThanks for the report and for the patch +Apply in the directory suggested by the filename +I am using +The current code does not handle dynamic fields +Dobris answer is correct +It is just a normal now no longer a subclass of +Nice +Delivered to trunk in and in +Ah OK +Also I dont think state needs to be volatile +Fixing failures found during tests +I tried the installation on a clean machine this afternoon +Git diff text will add binary diff to the patch +or never +Merge to revision +And after all this it will cut down the listIf we apply the max earlier we cant really be able to sort the entire list etc +That exception mapper was lost after transition from Jersey to CXF so we had instead of Right +In my experience with the changes between versions of files that we patchcopy have been minor and relegated to things like classes changing from internal to public API +Then the calls the via the local transport +If there are any knifes thrown at me Ill change back otherwise Ill do it for also +What is the use you have in mind +verified in CR +Hi Colmthe patch is working cool +Patch is applied at +Hopefully this one has everything resolved +I can send you application exe with DB which is around MB as sample is not able to corrupt the same +This will be less precise than if the solution would have done it but its more protable +Tested on RHEL and Solaris and works fine. +As a quick workaround drlvm could probably switch to version of have tried using the contained in downloaded as part of the sources target of the sources but the workaround does not work for me +passes mostly due to removing assertions inreader pool release which presumed deletions would besynchronized with closing a reader +I really this it does not make sense from a relational model POVHow would you map that +This should apply cleanly to trunk and I think it also encompasses the fixes for +False alarm deleting and rebuilding the index fixed it +but still critical we are moving to CR +A better workaround might be to set true so that even multicasts on will be received by processes +Please there are only some trivial changes required I try to generate a rebased patch which also addresses Umas comments +When I press cancel then I getHTTP Status User anonymous is not permitted to deploy orgspringframeworkbatch into orgspringframeworkbatch. +This patch fixes the issue for too have experienced this error in its very confusingOikonomopoulos Spyros patch did indeed resolve the problem for me as well +will be made and the case sensitive flag final +all Ok I will provide an updated patch with the comments taken into account +file contains fix for branch +It wont hurt anyone and it might help someone downstream to avoid infringing on possible necessary claims owned by Adobe +The changes look good does it work for combinehiveinputsplit also +Id be more in favor of making this configurable behavior somehow maybe with the options of refusing to overwrite writing new files next to existing ones and totally wiping out the output directory first if it exists +Similarly the should be aware paragraph posits a mistaken impression thatis actually impossible if you are enforcing scheme names +jar like its supposed toShould bejarfileUsersjsichiopenbuildbuiltinsBut isjarfileUsersjsichiopenbuildDo you know how to tell it to load from the jar instead + +There are a ton of Connection objects still open that arent closed even though the associated socket has gone away +Hi Aaron Arpit had written a test which reliably reproduces this on a node cluster +If EAP is created correctly this issue is fixed +to activate the pelt theme in your project set the following properties pelt +I like the idea here +jar on the classpath when executing the job from oozie +This helps prevents transient problems from becoming permanent problems +Verified in CR +I may be wrong but I thought it actually did support a CDATA section in the properties but it just didnt make it through to the binding classes +As the implementation of incoming messages delay would be more complicated I think that implementing outcoming delay could be sufficient +Chris you are ambitious +Think that this should be fixed now. +please check all is good and then mark this issue as Closed thanks and enjoy! +Thanks +Built in brew and revision +Fix for +So as long as you support properly theres no reason why wouldnt work +Committed in revision +Id be fine putting this in +as component build is still not ready it is still on stage site but it was changed to build is fine waiting for aggregation release note status as it is not required for this will be fixed when we drop to the Stable site and Eclipse Marketplace sees the new bits +Fixed in rev +pretty sure I have the right to revert my own commitI can declare the licensing of asl as a mistake and instead full gpl if we want to press the point +Adding an updated version of this unit test has been updated in this v patch and now does the flush with an assertion on tables which also now appear in the block cacheLooked for added in added some additional Javadoc +Heres the working page with the ability to browse for a callback handler and now edit properties +A better option might be to remove the hooks rather than deprecate it would avoid collecting cruft in the list of hooks but would break any existing coprocessors that overrode this hook from +which indicates that the problem is recognised but is only rated a WARNING rather than an ERRORIt should be rated as an ERROR and this will prevent execution of the application +OTOH maybe it would be better to discuss the contentpurpose of individual dirs because it often changes +Would you like me to look into the error? Or have you found my dirty mess and cleaned it upPIm not sure how the patch file got in it +Hmm having a further look I found that uses the same patternand exhibits the same feature in the event loopWhilst my patch appeared to work for me I also discovered the locator wasreturning null values when I recorded the output withWill add another two patches for both and that initialize the off the s locator +The patch from Kyrill does indeed fix the SQL bad grammar problem for DB too +batch transition resolvedfixed bugs to closedfixed +But at home I started learning Camel in a more detail way and I got this error So I reported it +Can someone inform me if the RPin is the same code in or do I need patches +Obviously protecting against the exceptions is good but that doesnt necessarilly mean we need to actually serialize any real data +Attaching new patchPlease review +I do agree +Thanks Edwin! +site is looking good once again. +Let me know if there is anything I should have done differently + + +Hmm +Exactly +Targeting +In some cased the actions taken are not taken at all if youre bootstrapping a previously bootstrapped source +pndisposition was only called for messages and because those were the only messages receivedThe investigation continues +In that case we get and cant count the line number +Youll have to use the version to get the fix +Heres a really hackish solution +I view as a service to apps that cannot do new not a necessity for IW to function properly +and getting a good model in ZK seems really important +Maybe we can at least provide a hook into the element for even if thats not quite a general solution to the +Looks good +looks good luca +for cfc based custom tags this would be no big can do this for Railo i think i will look into it +BYTEA used instead of closed as resolved and released + for patch +I think adding this capability is a very good idea +is a separate issue +Attaching a fix were do not return from the source resultset only from the hashtable +I did not seen any outof +bq +This class should also deal with state validation and the default implementation should implement the best practices for CSRF I am not a lot familiar with Spring codebase its difficult to comment on implementation aspects +Anyone can help me on this +Based of aslak branch so it probably cant me merged directly to master. + Bilginyour two patches are in rev +I also dont like copying code that is almost identical it leads to a maintenance nightmareUpdating only the javadoc to explain the two exclusive operating modes for this constructor and hence to strongly advise this constructor remains private should be sufficient +Does also looks like this with Firefox on Windows +should not be in patch +Actually if it were an Object then devs could use either a String or a Class and it would be more compatible with the Codebehind plugin +verified at r +yes thats right +Harsh Id like to deal with this ticket however I dont have a contributor privilegeCould you give it to me? +. + already make use of the serviceMaybe we should deprecate the former and rename toWe should also change the default to make an example of this new VUT configuration +From an user point of view I think that the features plugin usage is quite disturbingI propose to provide a new maven plugin looking like distributioncreate to create a custom distribution distributionbrand to create a branding bundleWDYT +Commit the changes to branch at and trunk at +That is to say as we start to approach the blocking limit number of store files we add some friction to writes going into that region +Thanks Todd + +I wrote you a rather longish email about the patch sometime in MayAs IVY has not been released officially I consider to still be the official version this is why I didnt check against although I will give IVY a spin on monday and see whether this and have been fixed +How about adding a error or stopped boolean or some indication to AM in the Allocation struct +This passes all tests and specs and will fix this error plus many others as yet unreported +Can you try it with https instead of http? +Fixed +We should probably have a doc +Well need some user input hence the contextRoot suggestion above which should possibly be called fully which we should check starts with http or https probably +These anonymous classes were piling up because they were all being generated in the main classloaderThe before and after can be seen in the attached image +Ron Thank you +So there is only one Vector of instance that is reused for each element in +Unfortunately I dont have time today to track down when started showing the exception but someone might want to do that investigationI dont understand the reason for original call to handleException which ends up closing the session and popping +It just occurred to me this might be our Juha Lindfors so try bug him +The v patch is to trigger pre post events +Yes need to be more patient now it went on +Ill try this out. +If nobody has a problem I am going to set this as Not a Problem +We have unit tests for this already and they pass +the real fix is located hereBut since we use an older plugin version this problem is goneOnce a new Trinidad plugin version is released we will replace the plugin by the new oneBut till than we keep the see the same error log +Can that not be done within the existing file instead of creating a new one? Just helps to keep all related tests in one place +I have taken this work over from Semih since he is very busy at the moment and I need it for my master thesis +Using POST as a workaround means that query strings will not show in a servlet engine log +If these are approved then the department managers are informed and accounting staff enter the numbers in the accounting systemIf a budget proposal is rejected by higher management then the department manager has to adjust his numbers and submit these againRegards is the progress on this +Are you working on fixing this or we should reassign it? Thanks +The new classes are public so they should have audience annotationIf the audience is Private stability can be omitted +Can now view latest working copy from the build agent in revision +I also agree about extensibility for user scripts +As soon as he has done that I can request a transfer +More patches needed +Michael in my previous comment Ive also said that theres a workaround +ok got it working debugging the patch now +Very nice work +Leaving this story open till he confirms validation aspect +MS Log for CS KVM host on the logs it looks like a configuration issue +I have not ported it to trunk + +Ideally both +Gave issue a medium ranking based on the fact it may require a procedure and fact checking +but typically this is not necessary only when clients are also communicating via some method other than just ZKre adding zoosync well then it would be inconsistent with the java client +Patch for +hi againOk your explanation of the leak is mostly what I understood of the problem +Now working as expected +Or a Google service or a Yahoo service etc etc +Im looking at the build process now +Hi Kaz +bq +Use +Agreed! Closing +stack wasnt a moderator for dev so Ive added them +yes go for it! patch in revision . +Packaging may be improved more but at least there are no cyclic dependencies now. +This patch proposal merges the db and column routines in the parser into a single routine to handle all the ALTER TABLE ALTER COLUMN variants +Long term it would help to add a schema or better documentation that clearly identifies the mandatory and optional parts of the format +All we need now is the fix for the newly demonstrated bug +Committed +Unassigned issues move issues to and +Michael we could increase this timeout value but Im not sure this would be the best thing to do +Part of to port +Let me know if it is what you have thinkedThanks ll do it in next days +I would like to help improve the testing of this but it may take a couple of weeks before I am useful +I can provide more details if requiredreproduction +shouldnt this be left open and assigned to the identity module and jbpm ? Now it will get lost again? +This at least allows some high availability features to be built into applicationsDo you think it would be worthwhile to post a test case for the load balancing issue +This is actually a limitation of this plugin due to the fact that we are not using Sonar core to get the duplication data but running cpd instead +Filed for the multiple RMs check +Set and and committed revision in HEAD +I just committed this +As this test case update depends on changes to the test case that were also made to support only tests in test should be considered relevant for this issue +Please make sure this edge condition gets documented +Release plugin shouldnt break their versionlessnessIt breaks therefore it should not be said behaving correctly +Otw if we need to first add more testing etc +Deployed fixed SNAPSHOT with build timestamp +Btw I dont think annotations are that helpful people dont really read them if its public its public seems to be the prevalent mode of thinking That is exactly why we need to cull these and enforce these going forward +However the client does know the block access token and a properly authenticated DN will be able to recompute the block access token based on the block key it shares with the NN + +I have made some more substantial changes to and now sets the parser to be namespace aware now throws a meaningful exception when a difference is encountered rather than returning false Special handling for xsitype now compares the type name but ignores the namespace could you attach the source for please +Actually we did put this into the spec in section where it explicitly states that the packages are available for future resolves +Thanks Dennis +since the local host work on my machine and the trunk has run for a month without failing on my server we need to think of the OS or activity as the triggers for this +If there still is not enough info you will need to pickup the current Branch version of which has some addition debugging logging. +Committed to java repository. +This patch is complete and ready for regression testing +Wouldnt we just concat along the block boundaries? How would we experience different behaviors in each place +Paul per Ramesh if we enclose any column name containing special characters or Teiid reserved words in double quotes the generated transformation will parse +this work is now in progress hereThanks but its quite difficult to follow the patch as there are a lot of irrelevant changes reordering of imports field renames etcBTW we never use imports like import +Thanks +only problem is that we are growing out of space here if every feature will have its own branding installed. +For a user using xhtml though who doesnt want Dojo but does want this flag would still be useful +Mumble +I would like to also use in our company project +Just try to build the jar and look at your manifest +timestampcomparisonudfudfbitmapemptyudfcoalesceudfeltudfhashudfifudfinstrudflocateudfpercentileudfwhen +that a object would give me the correct practice this seems to work even if it is a little ugly +In the a check is done to see if the is a and if so transferTo is used to send data over the socketIve tested this with files as large MB and everything appears to work correctlyMy biggest concern is that Im skipping over the filterChain to do this and Im worried that may cause writeRequests to get queued out of order +Otherwise I think its ready +I dunno Id rather emphasize that passing MAXVALUE is a Very Bad Idea +Duplicate of +do we just have archetypes or add the libs too? not sure here yet +Will throw on deployment if TCCL is wrong should deploy successfully log if TCCL points to EAR +bulk move of all unresolved issues from to +The code in has been a delegator pattern for some time now +Keep in mind we may have to abstract this when we change the hosting provider for On that note Does anyone have a perfered logging provider should we just roll our own? It would need to support Console logging and potentially SQL or file depending on what we go with for. +My example with the calendar and the ajax event is working perfectly nowJust FYI Tomcat logs the following error times on console duringafter each ajax event but as I said before its working anyway +In the meantime we will keep testing these independent daemons on our clusters +For +Closing tickets +Do not think this is GWT related but more a bug in AS concerning exploded ve seen it in other the bug occur if you enable compressed archives in the server editor +And the user will be conscious about the existence ofThey will be aware because we will include it with examplebq +Subject to an update all is well +Seems like a pretty big oversight on our part +Need rework +Couldnt the master repository be on Codehaus with a mirror for working on +It doesnt appear likely that it is related to Groovy +would be done as code loaded as coprocessors +Attaching patch for +My initial approach was actually what you suggest but changed it to the current form after thinking about its possible use cases +Thanks Dag! The b patch looks good +Id think that if you want something verbose then you will prefer the recursive behavior +Try this one is much better thanks Dmitry +Done +Patch applied. +Resolving as a duplicate of +Ive removed it in this patchI ran the integration tests and they passed +to trunk +Ill try to add the missing methods tonight +Unit test case with mapping file and test entity that shows the exception +Could be a bit picky but I thought I should mention it +if this is the first test running in the VM it will likeley fail but if another test runs first and causes an server to be created then the tests works fineand recently committed cahnges seem to have shifted the test ordering in the VM slightlyeither way this just started failing in jenkins as well suggesting that its not my imagination or something that changed on my machine something fundemental changed today on both trunk and x causing this to fail regardless of seed +Sorry Jasper I didnt see where writeCompleted was set +So if you lost two races then it could be caused by someone running mapreduce job kill +breaking up the text to display the jira more conveniently in my browser +Didnt have much time to take jstack and other info +This updated reproducible program does not reference code inside think fixing will fix this jira entry as well +Done with please create a new issue for this +The goal of this task is to use Logging same as we did for Common and didnt find the time to work on that module yet if you have a patch to contribute that would be welcome +The wrapper serves as a sort of modified value storage +That should be a constructor argument if we want the sink to control itIm open to adding an option for UUID as wellEdit by constructor argument I mean constructor a nd look at the code it seems fine to me +Current does not have it because was not merged to at that timeWould you check whether the attached patch is what you intended to post +I tend to agree is the only feasible solution in short termIs it possible to only use filepathfooabc format for local path? So we could effectively get rid of forward slash on Windows +Reopening this +httpd it looks much better thanks! +Also this throws the error earlier in the process +The +OlivierI am closing this issue because it opens a security hole +I dont think it is important to distinguish between null and empty fields in the Token class +If it is claimed that this is the caseI would like to see pointers to man pages detailing the problemsThe username is set from etcdefaultcouchdb +Have not tested beyond running CLI and getting usage s plenty of credit to go around for this one Erik thanks for the help! Michi can you take a look at the changes? I will also have Dheeraj check them out on Monday and make sure they work on our stack +I didnt realize that it doesnt work also with so I have no other choice but to add it in the current sprint multidb valueStorage anabled non optimised JDBC connection tck one test issue fixed that causes this problem. +Then taking snapshots of user can be done by taking snapshots for all user dirsFor root I think we should not encourage taking snapshots since all the tmp files will sit in a snapshot forever +Sorry for the uptake on this the new job is using up all of my limited brain resources +Changing the OS spec because of a java language aspect doesnt make any sense at all though admittedly the bit of the OS spec that defines Enum is bad design in and of itself +before I have a new patch ready +Hmm +The bundles involved in the deadlock are not mine but using the stacktraces I think I could try to reproduce this issue +Looks like I should run mvn gwtrun to make in work +patch files resolves this issue +Implements Felixs suggestionsThis has integration test failures for me that I havent solved yet +So I dunno +Thanks Suresh! +a person has a name and age and their sons and daughters have names and ages too +Yes but has its own list of comparators that never get updatedUgh yes +To summarizewe fixed the issues by wrapping two different calls in EMS in doPrivileged blocks +I want validation on an action class so I place validation interceptor in stack +But I can try it for you +Changed description to reflect name change +Andrew if you want examples you can check out the diffs between v and v patches for and +Please find the attached patch will contains the changed request for the particular case +Last I checked it also causes all output to be buffered before its comverted to XML in one shot in the past this has lead to errors when running junit based stress tests from the command line +Patch applied. +I honestly think this is just a simple from a walk to a should be almost trivial no +dfs lsr hbase on PHX cluster +The path to verify successful localization is already covered +Structure actually is needed in this case due to the nature of the plugin mechanism +Have you tried putting backticks around the column name like +Does it have to be to make it to the hadoop trunkI could easily adapt my patch so that could read the version protocol but that doesnt solve the unit test problem +Any progress on this +Interesting +Theyre a lot more flexible than Files +Hermmm +Any numbers Nicolas you can share? +Changes look good to meStill unable to get the system tests to run but if one of the committers could do that quickly that would be good +Indeed it was the latter no VM crash as far as I can tell +Also commented on the bug number so others can find this working fork. +Ha ha +The Cache can be managed by JMX and it can meet out needs of cache in different camel component +just sending this your way to keep it on the radar +Yes I can confirm the problem +This is a blocker +You can use exclamation or minus to deactivate a profile +Hey BrandonThe entry Why does nodetool ring only show one entry even though my nodes logged that they see each other joining the ring? sounds like it could be the case hereIve gone through the steps of deleting data and commitlog directories and then restartingI still run into the same issueIs there somewhere else where that token might be saved that I can clear it outThanks +So it seems I have nothing to update for these plugins regardlessAre the JMX plugins malformed and not in the same format as the rest of the projects? I have no idea +Note that although changes have been commited to make writable +Ive just committed this to trunk and +Will fire up a windows vm and try there +running tests and also put this into PA in negative tests failed +We have one +qian in that case I suggest you ifdef it something likeifdef aiaddrconfig foo aiaddrconfigelse foo that sound reasonable +If you come up with a design that we can review and implement please feel free to open a new ticket with details +My mistake is still not reproducible at will +Presumably wed have a too or would that start off in For conf I think we could a few example configurations and Im assuming one conf RPM for everything instead of a and +ah the issue is already resolved +Here it is Grant +Note this patch causes RPC incompatibility but I dont think this is a problem since is RPC incompatible with earlier releases anyway +Thanks all! +fwiw Ive observed this on the zone too +Thanks Alex! +When I step through the constructor in my debugger I run into a class called a superclass of +Rebuilding a jar for each job submitted seemed heavyweight to me +The following observation should not get lostIn his original mail Jonathan also mentioned that currently similaritymay be below +Basically we want to move this in but everything should be considering ESB has been moving forward over the last months since this new scripting functionality was introduced +Mark here wanted to provide a much more generalized way of stopping any other activity not just searchTo be fair I think the use case for is weaker +This is a BAD bit of code so it should not be applied to Tuscany but it shows that cloning the solves the problemThe file contains an update to the itestconversation tests with an example of this problem +This does work for fmerendatest a user service only not work for fmerenda a legacy user +Here I added some code to the function which generates and applies our manifests so before generating the manifest we check javahomebinjava version to return before startstopsmoke + for review the xsd might be cleaned a the implementation you can now specify just marshaller or marshaller and unmarshaller. +I tried this solution myself many months ago and found no improvement +I think it is has something to do with transitive dependencies because when the same projects that were using artifacts built with maven would build and the tests would run +Some simple logging could then indicate that dedup was only attempted and subsequently executed on the first server +Feel free to it if you find something that needs fixingOleg +I found the last comment from surprising I am not expecting the tool to process the imported XSD +See for more +Is it still a problem +Ill apply it later today +Looks good to me may be the bot +Thats realy my faul +The weird thing about this bug for me is that a little googling indicates that java should have built conversion for CIFDIB imagesWhat version of Java are you running Arindam +Committing +Committed revision slight revision to patch to gate on isSlave to reflect wakeup inline call to iterate +new patch file made using diff dNaur referenceappassembler patchappassembler +Changes to implement this functionality were added to both the and codebases. +Ron we do need a Download page dont you agreeJust the standard kind of Apache project download page that directs them to the mirrors +But We all know that load will throw in the case that the entity is not there +Current GIT Master +Its fully compatible with the +Please give some advice thanks very much +Ill add the patch I have in a few hours but unfortunately I wasnt able to get a successful deployment +Sure you can close it for nowThanks again +The file is a patch to the other files +Completed At revision +Committed revision +I think so +Remove packaging of duplicated third party jar patch to add hdfs ivy configuration +Try to install this snapshot twice in the same directory +Iteration using Iterator is not very different +Hopefully its reproducible so if it goes away with the upgrade we can confirm that and understand why +Checking it in without enabling it sounds like a good idea +quick version for people to play with +Nice job Todd +thanks for all this cleanup work chris! this makes our codebase much cleaner and test real examples! Very very welcome work +heres theAre you sure of your patch did you test it? Because I can see liatIt for the name of the list and not listIt +Probably a duplicate of +The table functions registered by this program could be useful examples in the Reference Guide +The default FS is file and hence you see thisIts always a better idea to use the Apache Hadoop HDFS mailing lists in case you are not sure its a bug before reportingMarking as resolved as am able to run the security branch NN just fine with proper configs on the classpathRegards notOk could you exactly tell me the steps youre followingHeres what I do seems to work just fine Download a tarball +As someone new to Lucene with a specific problem to solve it is difficult to identify the appropriate Lucene feature to use +Committed +Reporters of these issues please verify the fixes and help close these issues +to doing it with nonce WAL change temp window for nonces to the region in trunk +Your thoughts there pleaseIf so we should consider bringing the capability to wrap sockets with SASL into the +How do I rebase? When I try git rebase remotesorigintrunk it asks about deleting everything that I changed +Sorry +Ive created to rename the design doc and add the license header +I assumed JS is good as far as security is concerned as the context can be limited to just what is needed +Any other comments please +Steps to reproduce pick a non default workspace edit it enable workspace specific settings and wms specific settings +Code that used to handle null now will have to be updated to handle the constant +Not for commit on trunk only affects +See earlier comments +well its not working +The original NascifThanks for sharing this with Ivy community! I like the power this tool gives to dependency analysis +Don please read my comment +It includes work for because it has a similar problem that I also fixed. +ps the problem also occurs when I use intellisense to add an object definition for a type contained in a dll that is external to the solution +the rewrite rules in htaccess things the conversion seems fine so Ive gone ahead and touched the jar files on the maven side and well see if they convert over happily +Weve voluntarily removed the PMD rules that are redundant with Checkstyle +I hadnt gotten around to researching this problem yet so youve saved me some work +POST servlet modified in revision to escape HTML api class modified in revision that class is used by the among others to test it I created a simple test in revision escaping is already available in the class we should use that instead +if its this easy to recreate on it should be easy to prove it fixed for +Trying hudson after +Adding link to to reference need for better overlay controls +Not to mention that most other databases are able to do more than what the JDBC spec says they have to do at the least in order to be compliant. +I just committed this thanks Cheng. +Invalid forgot to link +Is this really a JBPM issue and not an ESB issue +A duplicate of. +Updated example usingUse the Ant build file for enhancementIf you just import the project into Eclipse you might want to modify +Currently its documented in two places the usage statement and +Still need to Modify the unit tests Fix the PB migration code Write the migration for this jira Finish fixing v has some code cleanup done HBCKs test is passing and I tested it on a cluster where I moved +Linking to so we can pick the module name +There are articles on how to configure to avoid this. +I havent been able to figure out the problem and I may not be able to do any more work on it for the next few days +Thanks for the fixCheersJason +This very likely caused IO and buffering effects youd see when piping to bundle exec and friends +In the meantime according the the bean Introspector you can not have property name stating with upper case +added screens for subTable updated screen for panelBar corrected the references to screens for two components dataList and screens for panel panel toolBar updated screen for is done. +If we can somewhat agree on this core API then I would like move along to fixing tests as that is quite a lot of work +It breaks xercesc for any Mac OS other than Mac OSX but with some preprocessor foo it could be adapted to something that will work +Sample certificate for the unit test +Yes we are well aware +Ill do this later after is released +actually this happens in the svn trunk as well revision ll take a look at this today +I didnt think we were supporting for the examples on. +Getting an appropriately configured and operational server is sort of difficult to achieve for automated unit tests +Hi DaveI reviewed your patch and it looks fine feel free to commit the patch +Resolved with revision VV features plugins and runtime with revision +Ill attach the failure report files +New version of the patch this time it passes all the tests +I have no problem with the patch since it is the exact behavior used for checked in the patch to Ill wait for Bruces ok on the since its shared code with JDO +This is still an issue in as well +I will commit shortly +Current implementation of the Distributed Lucene index on May +This issue has been resolved for over a year with no further movement +Thanks for the heads up +Running Tests run Failures Errors Skipped Time elapsed to correct race condition in test case. +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian +just did it and works for me. +I was thinking about what my proto definition looks like when I have a list with maps and so on +Why do datanodes need to checkpoint? Whats the value of storing out the mapping vs +Should be fixed for +This is a showstopper for the upcoming release +Patch submitted by Ramya +Great Im closing the issue then +is not supported with release. +This is still an example patch much more work is needed to ensure implementations work as expected and can safely be +Committed modified patchrThanks for the tickets +Yes this is a neat way for it +Andrew I would propose the following to address your points Make it clear whether the list is complete or there might be other corrupt files not listed +I was suggesting a wrong workaround +With latest build I have not seen this issue. +At the end of the day JNDI is primarily an AS naming mechanism +bq +There are of course areas where trunk is more advanced than Branchx was but I dont believe those affect the JBM integration +Looks like test is the test that could have caught the issue +I just committed this +A boolean parameter isCleanup is added to the methods in and and old methods are deprecated +change priority to dynamic scaffolding now limits delete update and save actions to POST requests. +We have project related to issue +The patch does have one null check in it +I just tried again and cannot reproduce itBut issue remains it doesnt respect which paths should be checked every time and which shouldntNote that Im now testing with can reproduce the problemCommitted to the trunkAttached is a patch +When I made this patch there was no tycho at least not in version which was working for me +Thanks to Martin and Dan for the crisp summary of the existing compatibility issues +fixed in revision for and revision for +has now been Initial commit +Latest trunk works for me +Ill check it later +Ill do it +Fixed in version of plugin + it would be quite handy to read the source code many other Apache projects are doing this +Bug fix in committed as svn +Done +for reviewing Jakob +Please close this issue +Dave I think the problem is that the is completely arbitrary and setting one just seems unnatural in certain my current use case I want to skip bad call records encountered by my item reader and let my job execution complete successfully +And if so is there more work if not then close this ticket +Cause in the past I was told to add request to add new features to Len +I have the impression that the discussiondecision about a new log will take some time +So I at first concluded that the was simply misleading as it was truly an invalid request context +Carry on +Johan what version of maven are you using? Have you tried experimenting with different versions +Added a test that Marker is valid prior to delta serialization and do a complete serialization if it is invalid +pingIs there anything else I need to do to get this patch submitted +Thanks RyanI will work on finalizing the theme this weekend and get it committedI understand an artist would have a problem with seeing scrollbars when font sizes are increased to a large size but in the end the users of have a job to do and work to get done +Please send a mail that include the output of mvn dotnetfxcop X along with your file +I dont have that problem on the command line +Reopening as the problem is not related to only it will affect all setups upgrading to with in + +I could instead do that computation back in the num elements space +fixed in r +Ill put something together +Nathan I see now that I do not have a comparable Union type defined in the test thrift files +the patch looks good to me +Hion the end of stack theres message Caused by No code EPSG Have you added EPSG projection to user projections +AndrewCan we do something on lines of Role based security Like if a User is member of Auto Mechanic group then heshe can edit a and related dataLike We can create a Party Group Auto Mechanic Then make Mechanics member of this +This one is +so I think Ive got the tools to make this happen finally +Guys Ive updated the the description of the JIRA to be better reflect the latest findings +Before that RIT is removed which allows atleast the balancer to runSo we may have to come up with a better one +this could go on forever +Stack yes definitely easy to add +This patch includes Antlr grammar to generate lexer and parser for A wrapper class around the generated Antlr parser and lexer A bean class for A syntax checker for Two test cases to test the parser and syntax checkerI will also implement and add parsers for the other schema elements +Makes sense +on next merge from trunk well pull fix for over. +Yeah this has been done inWe can always open new JIRA if there is any bug +the other issue mentioned is ll add that if your database is committing the work done on the branch and returning RBCOMMFAIL then it is not compliant with the XA specification +Committed first part after testing on Windows +not quite sure why you are working on if as discussed once fixed it will just suggest completions to users to create code that wont compile? To me this specific issue is more important because this bug describes the groovygroovy case and not the javagroovy case but I do prioritize based on actual users requesting the feature especially anything related to Ast dont want to rush supporting these transforms Id rather do some proper design and see if support at the compiler level means you dont have to do anything at all differently for code completion +jar file +All the test cases ran fine so shouldnt matter +patch to remove the javac and findbugs warnings +This is the intended behavior if you unpack a module directly to the deploy directory instead of as a subdirectory it wont workIve added an error message that occurs whenever the hot deployer is asked to deploy a subdirectory of or which is a pretty good indication that this has occuredRev must have an old build the currently I cant tell the difference message does not exist any more +attaching a project which demonstrates that this works with Grails +Could you please explain if you are in fact seeing a different behavior in or are you just trying out the wsdl as dependency feature and not liking the results +Make sure that does not conflict with currently in is not a valid +Actually does not hash the password when received in clear text by defaultBut if the Hash interceptor is enabled then the password will be hashed and salted if required +for patch +increase the log level for the fault massage so that it will log it as an error +Resolving as Wont Fix as the use case for the proposed changes has since been addressed in a different way with the class. +If nobody objects Ill commit it this week probably not before Wednesday though +New date picker widget fixes this +Arun AlejadroI attached new patch for that addresses all your commentsI am still passing to the plugin according to my explanation in the previous commentCheersAvner +This is unfortunately the way that inner beans are being defined at the moment They cannot be directly accessed through the neither for instance lookup nor for metadata inspection nor for destruction +pending Jenkins +Heres the updated patch +Having the same method with exclusive parameters is confusing +if you have time to check this out that would be great +Agree that it would be nicer to abstract things out a little to allow other storage options +A couple of things The point of the management API is to avoid hard coding object names The log helper should to be a which is injected or looked up to help with the previous point to allow remote access to make it appear in the space etc +jar +I check version of pdfbox used +Verified with trunk +Maybe we could still call it skip +Id be willing to create a specialized cluster for counter column families if the risks could be mitigated through configuration various write consistency levels etcSo at this point were looking at using redis sets or cassandra counters just for speed and summarizing transactional data to cassandra integer columns periodically for durability and historical accuracy +Duplicate of. +The fix works well I think +Also anything that improves our io profile is ultra super importantSo I suggest that we move the compaction review out of this issue into a new blocker against +So patch lgtm +I submitted that patch +The message counter history data used for display is obtained from the queuetopic +Thanks again. +The other thing ive noticed is that if youre in a situation where the name can be resolved but the host cant be found it takes quite a while for it to timeout before moving on +tested the new script on a cluster and it replaced the with a new one and moved the old one to +Verified it looks good. +I am closing this bug as invalid +See the screen capture +It is not supposed to transformexisting contents of the map +Regression tests passed cleanly for me +Ill check on them later +Reopening to backport to based on conversation with Adrien on IRC +Tested against and master branch +Then launching cli command cordova only outputs No such file or directory on a new line so the r does not appear to meIt isnt the terminal itself outputting the line meaning the cordova file is found in the bin +ran a test on different servers +We get mysite pathexpiresTue GMT cata pathASPSESSIONIDGGGQQXEUIOFDDAGAHNAFLJNLBNMHJKCN pathWe sendCookie Version mysite null nullnull null Path cata Path ASPSo it looks like we are spliting the Cookie value based on and finding boguscookie pairs +The proposal of using the timestamp instead of a monotonically increasing number solves the issue +branchesSending plansSending plansSending plansAdding plansSending plansAdding plansTransmitting file dataCommitted revision trunkSending trunkplansSending trunkplansSending trunkplansSending trunkplansTransmitting file dataCommitted revision +Please try this patch I also fixed some exception compatiblity bugs of found during the test refactory which is a demonstration why the refactory is necessary +Andrea please review WMS part of documentation changes thanks +What is the virtue of this approach? Why wouldnt one simply create protecting the format in contexts where multiple threads may access it +Not fixing it for now since it needs more work +It now a using the same shared timer as the other cache background processes. +Any word on this patch getting accepted? I ran into this problem with as wellthanks totally agree to you Charles +Ill add a comment there +will +It depends on Hbase minicluster and maybe failing due to that taking too long for some reason +its a one line change in addition +Thanks for the patch Guozhang +now i understand +I tried to put as much as possible everything in one table per graph and also being able to do update the graph information updating maximum two rows +Done. +I just committed to trunk and branch +Here almost means that code was not wrote from scratch but was wrote as reform +I would like to generate only m records of pages +From run mvn test Dtest TRACE fix version for release. +Thanks +So if the license is an issue for I can leave that one file for a while but Id prefer to totally nuke that directory if possible so if you have a pointer to license info so I can clarify that would be coolIll open another task for additional refactoring +The class is not used in the new mechanism +Committed at subversion revision +Patch applied +Thanks Ivan! in to trunkMichi +At that point Bin apparently was still running an earlier nightly buildOlegPS PatchThanks that is goog new for us +However simply stopping returning entries in all cases leaves the ESB project for example with only references to their own ESB classpath container but not to the generic server entries which they may fix for this is to make the ESB wizard add the server classpath container manually +In these cases the order is usually the insert orderAnd in those cases there is no need for a tree structure of course its possible to define some artifical tree structure but in that case were back to the same workaround we have to use in Jackrabbit today +Thanks suggested the change primarily because it took me a bit of stumbling about to realize the relationship between containsBean contains etc +Thanks! +A downgrade would still be possible with the loss of containers of course by simply removing the state store data and restartingIn summary we would need to be cognizant of changes that affect state recovery upon startup so a restart can be used to support rolling upgrades +first rough draft added Steves intro and fixed a mistake or two as well as a couple updates also changed the page names from sandbox to contrib +Provide a POM we can reproduce the problem with +I went ahead and applied this patch we can always provide a flag to disable validation if theres enough outcry for it but I tend to like having validation on from the beginning since you can never enable it by default after thatPatch applied in SVN +The lesser issue of making assembly is still a problem but Im not sure whether it is worth addressing +When we merge to trunk well promote all the open snapshot issues to full issues. +And I am not able to reproduce this error locally on jdk +But after the tag there was created a new revision that applies similar patch to the situation is confusion please verify that repository versions are in sync +Thanks Lukasz for that may be the case but the patch is IMHO wrong if CDI does not support the mixed resolution which is that the API on is all about +Hi MikeIn my opinion the most important reason for committing the patch is that the current implementation breaks the field abstraction +Igor it might be worth to explore to maintain external mappings between changed resources and required Maven plugins +Update The UI reports that the password has been changed as it has remembered the setting but the changes have not been persisted to the DB +Ill have it done on Windows and update the comments accordingly +Please verify in CR cycle and resolve before release setting fix for EAP and and adding Release Notes flag +Patch coming is just test failures from and not having the correct command for the hadoop version +is ready for review +e +Fixed in revision JoseYour patch was late I already fixed it but thanks for catching the duplicate key in. +In the logging configuration you can add filters to look for them +True +And just as a point of process I think adding a new public facing utility should really get a bit more public review and consensus + Removed setter for jobIdentifier in +Running the quickstart with trunk and produces the same sequences And according to this is correct +Completed. +So I am going to fix the samples +Also to get this into the release you need to find somebody who would commit to do the work on this soon +For instance it may be possible to cook up elaborate queries which still fail because the VTI joins to the output of joined tables or subqueries in the FROM listOn balance I think this is a better approach because it fixes a code generator problem in the code generator +There is already a unit test for this +So if the collapsed document set changes your ordering will likely change +We need the web subsystem and infinispan subsystem patches and then a few minor cleanups +I suspect the problem is not fixed yetI used revision +It was an error in our own implementation +at least that worked for me +A mailing list conversation We have permission to use the LGPL driver so I dont see any reason to remove it yet +I wouldnt let it affect your mine this seems like a lifecycle executor thing +Adding couple of testcases and incorporating Marks comments. +return value is different because it can be done consistentlyThe idea is that it is always the proxy that is exposed to application code +Let me count the ways. +Attached file includes the modificationsYou are right +Please review this patch carefully +Since this is superceded by I will let the later more comprehensive Jira take care of the fix +The same happens using Axis as a client +Marking these as fixed against rather than against. +A working HDFS HA mode has been implemented via +Eli you are right on buildversion matching and version not and so on +Let me know if you want me to create a new umbrella jira for all pending issues or if you have some other preference +RTFM +Tucu +This allows additional rules that require comparison between the context project and the suggest import type project as well as the rules mentioned above +I applied the and it allows comments feeds to be properly created when HTML is allowed in the commentsThis gets my vote +Also closed in r. +Committed to branch and trunk + +We can revisit this is the community asks for it. +The install was doing perfect and theserver working fine until this morning +Heres the code on the s the s the commands executed by netbeans for compilationgcc c g I +Please name it resolve again once you are done with hudson job. +Need to remove the contribution from Westport +However it is a lot more complicated solution to develop takes much longer time and involves a lot of changes to the system +If just two test failures Im inclined to think it flakey tests rather than junit version and that we should move to junit and fix the for or more info on before going further in Surefire + +Reproduced in JBT alpha build +Search is not enabled when not selecting an element does it only search the selected directorySince your mockup is showing that Refresh and Search call the same dialog I understood that both of them search the selectededited directory +I will commit it +yetCould you put together some code example how would you imagine using Mackito for writing tests? Just to give an idea what youre thinking about +Marking this blocker +My bad null and! null is already thereOnly think left I kinda recall is better syntax error detection +OK thank you for verifyingResolving this issue as duplicate. +BTW the problem reported in what is the exact concern? that there are HTTPS endpoints? or is there anything broken? +Patches are always welcome +Help documentation is done +Yes your xml example looks good to me +bq +However I think we should add required mappings to templates of new JSF projects +Fixed along with a test case +Otherwise I think we should close. +Fixed +If this happened on multiple regions there was bigger problem +on the patch +I will commit it if there is no failure on on the prebuild envThanks for the review Stack +I confirmed same exception on +This would be low priority for the project +Fixed with commit +You can move forward in the mean time +Is there a +Ive been away and was busy with other things +Has to be fixed in +Could I fix that and run test +DoneAll issues transferred to Public StatusAny person in the PMC or Administrator roles the reporter can createconvert toview security issues +I agree with the general approach but maybe it should create return the array of values rather than have it passed by reference in +Any ideas +this is a feature + for the patch +Committed revision . +Had not noticed till just now +Thanks Garrett! +Thanks Rick +See +The problem has been investigated misconfiguration at demo app found +Equivalent command line to Alains which doesnt require any special classpathjava jar xml xsl pdf +Well there is a thread pool for processing replies +objections to a commit today +Approach is fatally flawed +from me +I have applied the patch without any further review as the fix was prettyOleg +New patch uploaded +Reopening this bug as the commit made to fix this issue has been reverted +The fix for this problem is inCurrent trunk works OK with the provided quickstart +The attached patch fixes the issue and add a test case for reproducing it +checked +While doing so I noticed that appservertemplatesweblogic still refers to it but its not an issue +There are some DEBUG reported as INFO +I was able to validate that the correct code paths were covered +I was having the same issue with hibernate +Can you explain in more detail how the bean archive definition relies on such things best is if you can can quote sentences you think need cleaning up +Fixed in revision +Id like to help but the line numbers dont seem to match anythingCan you either use current CVS or batik for line numbers +have made the synchronization block more fine grained and removed from the synchronized block +Feel free to close the issue. +I have tried to accommodate all the recommendations from Ashutosh +The one I really dont want to change is retarget because it is a difficult algorithm and have it as is makes easier read it and update it +Not sure how practical in terms of us as programmers having good log reports +The existing tests cover this functionality +Fixed +patch to general the jar scheme handling apply to zip tooPatch applied! Thanks! +So we might deprecate this in HV and remove the suppport in WDYT +Require that hfiles already be on the same filesystem as in order for bulkloading to succeed +It should not be tagged as Windows XP +Just because it wasnt accidental doesnt mean it was a good ideaThe problem here is that now every single one of the applications that does this needs to be changed and its not just my apps that depend on this behaviourWhy is it a problem to omit it in the assembly descriptor if by specifying append false it gets left off anyway? And what am I supposed to put in the id in that case just some random string? That seems like you took a relatively natural way to cause the classifier to be omitted and made it needlessly verbose. +In the Teiid Designer case theres a property tester for checking which perspective youre in so that should limit popup menus and toolbar buttons etc +I understand that in you cant do a sync point +v isnt perfect but its a start if the batch starts on a wide row we reuse the token and iterate until were done +Anton please try to reproduce it now +If this is not simple to fix I think it would be find if it was at least documented in the main documentation that filtering is not supported for pages under issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Changelist + to Thanks for the review Lars. +The patch is good +Thanks DustinI will review and commit this later this morning +JSF component should use the latest Common to compileNot sure about the error in +I dont think the new patch solves the problem +if there are any objections I hope people raise them before thenBest wishes and thanks again for being so friendly and responsive about this +Well address this in in the documentation of the mvc namespace +Can someone clarify how the build system should work +HiI can create a patch an make a pull request if you wishBest regards would be the best way to go but keep in mind all other platforms aka darvin and will do +I think I have seen something similar with Avro in the past when using the Specific APIDoes this stop if the Datum that is being written is not recycled? I think it is related to recycling an object with a Union and having the underlying type of the union differ from one use to another +Patch against trunk +As long as +You even provided that test case that DEMONSTRATES the memory leak issue is not a problem unless a is usedAs for what is a valid heap size this depends completely on the other code loaded into the JVM and what memory it usesI use a max heap of with mb index and it works fine +I have been reviewing the code +revision removed the exception that previously was generated when this happened. +HQL is not SQL and I really dont see any need to support this which can only be used in the editor not in the actual hql used in java and mapping filesAdding were only done as a quick hack to allow easy exclusion in the editor while prototyping +Thanks Myrna +Blows up with during app server startup because it attempts to access the ORB before it has been initialized +Let me know if there are any other colors sizes or icons you needCheers Cheyenne added icons and they look great +Looks like a mock FTP server uses some backport jars that isnt there anymore +Please verify +Attached a new version for like this change it seems more elegant to me +settings is really simple way of describing I understand that Maven people want to avoid having multiple configuration locations and want to describe everything in the pom I think that restricting external properties is really going to far +Shortly Ill upload a new patch that changes around the dependencies between modules a bit +I made a few very minor changes to the patch that you had submitted +Patches welcome +Some methods like rmDir can exist on the base test classes however the class also contains large arrays blockStartsEnds and it will just clatter the base test class I think +Apparently the shape of the default plan changes depending on whether you run this test as part of the whole Derby regression test suite +What is being proposed is that we host a Fitnesse instance that can be used by other codehaus projects? Ben thats what I intended anyway yes +Thanks Fintan +Checked into SVN on and trunk +Also it creates faults in the system to prevent acks from being delivered and such +The other alternative would be finding a way to support compiler annotations to structs and making this a option which would certainly be cool and useful for other reasons +I plan on merging changes from this jira and related jiras to by the end of the day per my previous comment +not enough information for us to reproduce or think that its an general issue +sorry for spamhope it will avoid ubuntu this to confirm locally ran failed test case and its finecommitted to trunk thanks for review. +Fixes issue with absolute instanceDir +batch transition resolvedfixed bugs to closedfixed +Does that mean that this specific bug makes it impossible for you to use XSLTC +an project example +The page is cdevsecuregrantrevoke +I remain weary of calling it since there is already Query rewriting support through Queryrewrite but I take your point +fixed it. +Sorry for not being clear my example was running from a console and isnt setup to run as a scriptThe thing to pay attention to is the result of the first call +C +I tried it on another mac and it works very well with RJBI think the simplest fix would be to try upgrading to then +you are constructing the regex replacements themselves with templates +Tested on Compliance Level on newly created ESB project is still set to even when only JDK Java Runtime is File is generated properly without Override Runtime is added properly to patch is enable the modify button for change project facets on the first page of ESB project you review the patch to see if it works? thanks +I think your diagnosis is correct Mamta +Reopen if you can provide a patch. +looking good trying it script is in because it is owned by the side panel and it should not be used by other developers +Applied in master and betaThe feature needs to be enabled in the workspace preferences +Same here I ran concurrent a number of times as it is and also a number of times with more threads and it works with your change +there are no deps left +Setting in doesnt seem to have an effect counts are off +and it has an integration test we need to adjust the way plugins specify lifecycle mappings we can continue to address that in +What is the behaviour for other hypervisors like Xen +How can i get THIEN PHAMs ticket reverenceIs there a way to modify few points of code to write directly the js or css instead of the url of the path served by extensionsListener +Trunkjobjsf +In trunk code getting logic has been moced to constructor +Attached patch for +The patch provided let two different doxia versions on the classpath so I fixed the relevant code to include the newest one +now it seems to be working lets see +So I would keep it the way it isLet me know if you plan to commit so I know when to submit the JCA and other patches +The Release but Ive got the Nightly built so Ill give that a go tomorrow +The lib folderLater on we can create a module that grabs all libraries and wars puts them in the right directory structure +Doesnt seem to affect us much at the moment I guess +Ive found that the plugin breaks grails console +Thanks is there wiki documentation available for this feature +That isnt to say we couldnt have an alternate API as well +You mixing fd and confused me +Marking to fix for final +I have tried with both +Great! Thanks for helping out. +It must be a bug in the new parser +Hmm does that mean we should close this? If fixes this as well I mean +by now writing LZMA BZIP and Deflate is also supported. +Hi viola Now I put the test case into the because and represents the new features in javaee +Thanks Anil this is in SVN rev +This is a bit weird +Is anybody listening about his build problem to apply the patch? Its a simple one +throttles of location results from each split to store in its znode at or less +There are a lot of logging changes +Not urgent for bumped to +Did you have a look at the fallback and use properties in? I think the later should give the behaviour you want +The test provided to this issue is attached is a fix to the attached is a patch to the patches in r +Thanks we get this into as well? Thanks! +This was discussed with Jonathan on the process of users should make sure that all of the nodes are updated to before running any schema changes because its impossible to apply old migrations even if we accept them and users will be getting exceptions from your anyway +When tested against trunk the test passes +Ill see about what to do about your contribution soon. +Sounds like you didnt have your hbase configs on the classpath +This may require a designer change if they dont allow the user to select schema level permissions +Got it +it is processed normally via. +Patch v commited as revision +Ping me when youre back from vacation +will be in ibiblio in few now its gone +Trace level it on commit? at all +Thanks Alexey! +And thanks for the confirmation! +Any news on this issue +Yes I can confirm that there are two failures in this test when I run on phoneME advanced and they look pretty much like the stack youve pasted +I see what you are saying about using the reflection but using it is a double edged sword +Applied at r please verify I added all files +Patch looks good +I dont have access to Windows +By the way we used solation level +Have more tests to run in morning +See updated patch +Give to idle hudson +OK with your app I see the problem at least in jetty +Slightly tweaked patch its possible to have only one IP on the instance so we should make sure PRIVATEIP isnt empty +Thanks for including this. +Ill create another patch to address your commentsWrt +Updating the patch +it would! It seems like youve found a bug in which I accidently broke the case and a sample would be just the thingIf you really want to be a hero create a patch with an integration test in systestswsdlmavencodegensrcit +alan to the patch is a relict +fixed in +So users should be screwedFirst the biggest in maven Ive seen is to accept an artifact with wrong checksums by default +avoiding the +Requirements for EC Larger storage space more GB In RPM repo mc Oracle JDK is about to put the EC specific AS config files into upstreamWhen thats done I can put a switch to the testsuite harness to use them instead of the regular +Phil I really wanted to get this in Milestone which is suppose to release this Monday +Incorporated Toms review comments +ROTFL you fisherman you +Another patch which can be apply after the main patch +The fix will be in CR +but Im not able to apply that patch and I dont have any of those tools available on my platform +if it does we can also change the HH code to use DDI dont know about that having this as a check to make sure this doesnt happen again is a far easier thing to catch than other gossip is out of date problems we might encounter +Really this is a holdover from which can leave orphand locks +Helps a lot +I havent run full regression tests against it yet but I think its pretty solidIll commit in a day or two unless there are objections +Thanks JarekSome Japanese contributors are reviewing the applied translations nowIf mistranslation and are found I will reopen this issue and send nd patch +projectpluginslibmanaged +Outside safemode it could just report since will eventually do the job +I think Richard is already doing this as part of his stats implemented a callback mechanism that allows users to retrieve stats after each job +Here is what I suggestproperty default which specify an URI +when events leave a data window +My test runs through the loopback interface of course +We are still awaiting the outstanding information for the Release Notes on this one +The patch has been appliedVladimir please verify +Ah ok +A B wrote But that said I dont understand why Im not seeing the failure on my local machine and I dont understand why Knut Anders was seeing it with jdkI rechecked my script which runs all the test suites and then Iremembered that I changed it to use jdk for the tests some timeago +I have added Steves proposed flag +Configuring procsysvmswappiness down to may also help +OK so it is clear what HQL to write select from Account join relationships where memberOf +added examples images for the new behavior to fix selectOne in IE +The button already exists as can be seen here in the servers view +There is already for this +My primary concern was the os check +That was first aim +I working on a fix now +WorkaroundUse a different query Cache for example instead of +In the bin kit the path to is antlib and in a development environment it is lib +Setting to final for now will see if anyone works on it +reviewed +When this JAR is in the deploy folder of the same error occured as when it is inside my EAR second file includes the +Implement layer B streaming at +Im trying to determine how order item BOM component price info should be stored +If not we throw an +I cant understand what you are trying to say +download latest version of webwork +So if you pass in the you can then close it at the same time you close the scanner +I will handle only the order of closing the connection and removing from batch mode in this issue +We have also built these files successfully on WindowsLinux which confirms the loaded files are correct again +The forrest doc comments need to be updated for the new functionality +Thanks! +one small request can you change the assertion in the abandon block case in updateBlocks to a proper if statement that throws an IOE? I think in this section of the code wed rather be safe and throw an exception than risk replaying the edits incorrectly +Now add another vmware cluster into the same zone +package will be added to dist attaching the sh which pushes the release out upon successful vote +Ill get Brian Fox to check with his build as well +Hi Jacquesi have an implementation for this issue +Same failure with +Could you against trunk and let me know if you still see the same thing +Filed for deprecating the SLOTSMILLIS counters and for adding in MEMMILLIS. +Ive written the failing test but its a bit more complicated +This is working per spec +I know you guys have a binary release this is what Im deploying +From client side it does seem like a breach of spec +I reviewed this patch and my only comment is that I think that the ij change should strip comments for JCC or for Derby client drivers whose version is less than +But my opinion is that is the correct datatype and is incorrectIm guessing that the overall query tree looks something like It seems to me that the data being sorted is the SELECT dataand that data uses the constant and so that data should havebeen sorted as simple integersThen the should have generateda CAST operation to convert the integer to a biginit when thedata is retrieved from the sorterAnd the used by the shouldhave pointed to the values returned by theso it seems correct that the VCN instance that the references should be working with aSo Id suggest investigating the andwhether it is generating the proper buffering CAST operations +Committed to the on +Committed updated INSTALL instructions to by Alan Robbie and Steve +Bulk close for release +Changed the summary to reflect this new plan. +backport this back as a known issue in the EAP release notes +Final tallyOld minutesNew minutesQuite a difference actually and not in the direction I expected +Fixed in . +Btw if using an adaptive fetch schedule you will can see it being slightly off for gone records +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Thanks ClausIll fix those checkstyle issues and will update the patch accordingly +id tag is redundant please tag has now been removed. +Thanks for the contribution Santhosh +In section optimization for transactions in Hibernate is discussed but theres not mention that this optimization means the session will not flush +Im assuming you only mean certain classes from impl right? We dont want to expose all the implementation classes +Nitpick I thought we were only going to start services on and runlevels +Darrin I cannot find this reference in the docs can you confirm that we have this listed in one of the docs? The Getting Started Guide or the Services Guide +I push forward my thought by other ticket +I have no issue with allowing an optional stamp but feel strongly that we should not be requiring or depending on the client setting of stamps +Changed the classes to use the logger +Item in this list is due to the use of UTC as the timezone in applications for representing date and time values +Patch applied and meclipse is working well with this in place +The script can only process Ant build files where target tags are the only ones that appear at the top level +Its used by when no other configuration is given and I recently modified the Jackrabbit webapp to also use that file as the default +If a datanode does not receive a response from a downstream datanode in minute it declares the downstream datanode as dead +That has changed and formal parameters are usually stored in a nowIn your class would have been enhanced and a concrete subclass would have been generatedI think a better error message would be nice but not a for +The other components listed are still broken despite commenting out the offending line +Tests pass cleanly for me on this second revBecause I had to make changes to I suspect that corresponding changes may need to be made to jfoundation by someone who runs tests on jTouches the following additional filesM javatestingorgapachederbyTestingfunctionTeststeststoolsM javatestingorgapachederbyTestingfunctionTeststestsderbynetM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmaster + +Yarn has decided on REST as the RPC protocol so why introduce another one +It required me at add a method to the interface to retrieve the namespace URI ThanksThe solution proposed in your patch is invalid +And in which cases does this happen +Are there any means to configure Maven not to print this warning? In my case a project uses activated profile only in a single module +sn attribute is mandatory for the person object class +Please note that Ive filed a bug against the smslib library however still have not received any answer so I believe we can use pathced jar file with the fix for now +The idea is to deprecate the feature cause it wont be added to the new API and the concept is to be unsupported in favor of logic +Yes this can be resolved as the original concern was fixed somewhere in between the linesClosing. +cache sizing is something that is easily scriptable w nodetool or other solutions should the need arise +Ill be interested in what you find +Thanks to Neeraj and Michael +we dont manage the jboss repo just central +Thanks Will! +I dont think we have a problem here. +Fixed in revision +Yes this patch is for the load store redesign branch +All tests passing +Has to be applied to the branch as well +Wondering if you can take a look and if they are enough to solve this issue +Marking as WONTFIX since this will not be done on the codebase +Verified on and. +It seems I will have some time in the next weeks also. +FYI is introducing a classloader for MR jobs +HiThere are no nor OOTB in +Marking for I did autocompletion for my project simple java had superior memory characteristics and almost the same performance as tries +Mahadev thats wonderful news! We would really like to get this into the release as well +There is a trailing space +lgtm +Otherwise its just that needle in the haystack +the performance effect of is clear to me nowhot rod client gets only last hash id for each server this means hash ids are likely to end up very closely to each other on the hash wheel which means picking the first two ones from whatever server countthis means hotrod client contacts only two of the four servers ending up with very inefficient unnecessary network hopping +All queries are case sensitive as is indexing +We had to work around it in the replicator for docs with lots of revisions +The testing we all did Aug showed that eliminated the problem +More infoThe root of the issue is is committed during rendering coupled with the fact an exception is thrown after rendering has been initiated +My apologies we are both yahoos so I figured we could internally discuss and then summarize for approval +Its just that it will take a while to go through the xml file and fix the cases that are failing +this seems a little strange +I was using public render parameters but now its too complex to approach the problem in that wayWhats the current status of this issue? Do you think trunk has events implemented? Thank you +Bulk close for to the issue that actually got committed +Sounds good +Hm just stumbled upon +Modified to take a single dict of parameters rather than using kwargs which my reading of and inspection of other database drivers indicates is requiredCommitted w that change. +Is it the problem that you havePlease wait till the patch is finished so we can use the handlers in more generic form for all the droids +First coverage run in progress +Thanks a lot for the review Todd +Forgot the new and reformatted the text to use +Its not the maven settings that are my problem +In the guide facet multy is changed to facet multiple +That sounds good to me +I was also going to ask for a unit test but wanted to improve my understanding of the situation anyway so adapted the posted problem code +lets fix this in +BTW Its impossible for us to support the community on previous version by doing patches and such +There is no ane NN for CDI for Beta +Moving to ER as it remains an OPEN issue for ER +Attachment has been added with description sample PDF result showing a remaining has been added with description Yet another has been added with description updated patchAdded this attachment to demonstrate the one remaining issue I was looking into +Thread dump only adaptor and are running before kill I check out the trunk package and run bin without any initial adaptors then issue kill I see the thread +I have committed this. +No it works since your fix +are you saying that the datanode will watch over the root disk at even though the datanode does not store any hdfs data blocks on the root disk? Or will the datanode watch over the root disk only if there is a data directory configured inside it +Try adding one +Thanks for reporting this and helping get to the bottom of it BerndCommitted revision +in POJO date is mapped to dateTime rather than date +This flag was previously notnecessary because older professional SCSI disks did not use a cache +Closing issue fixed in released version. + +No response assuming ok. +This kind of glitch will be recorded in the log file so some monitor process will notice that +Fix applied in r to clarify how to a privilege and fix the sample ACE edit script from the bundle. +Should provide some ui to let users say if they want the related optional features installBest Alternative provide mylyn integration as seperate features to install +I will mark this issue as resolved for now and close it in a few days. +The solution also uses throttling to prevent the requirement to expand buffers and shows constant memory usage and zero garbage now +Perhaps change the fail to a println +Also having this issue Resin and mysql databasejdk +bq + +Should we also check permissions for owner and groupSimilar code existed in +This is the reason why I did so many +Will inform here of any status changes +As outlined in my comment in the spec I would prefer an implicit mechanism rather than using paths as I dont think use of paths will look right +Resolving in tags should happen in two situations Translation during serialization before leaving an XMAP +Juca if this is covered in a different issue then please close this issue as a duplicate and link this issue to the correct issue +Id say it should be at least as long as the default transaction d prefer if there was an option for a keep alive ping so it was able to differentiate anetwork split from a long response in the alternative the timeout is long enough that it doesnt suprise somebody when it happensI believe RMI uses hours +But I agree that JRE classes must come first in any case +Small typo in +Checked into branch on by Pinaki as r +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Should add a drool for this +OK +It appears that our substring matching code had a couple of bugs in it which were causing false positives +This patch use a different object for wait and notifyso one thread invoking addCall method wont be blocked by another thread calling setupConnection method +HariWith this new format we wouldnt have to read the whole log file if we stopped storing the offsets +Thanks +This issue has fix version and is marked with either Release note needed or Existing application impact but does not have a attached to it +The heap filled up quickly as well until they ran out of memory +Fixed +The attached patch fixes the issue with the user id I removed the user id claim check in the business service implementation so other stacks can send the user id as argument in the operation +Updated the to check for an overridable method use and set the to return true +They should hold off on compactions and splits for some configurable interval and then slowly ramp up compactionssplits with randomized hold intervals to avoid thundering herd behavior +No need to apply my patches +Archetype simpleapp works fine. +Thank you Tobias +So maybe if we link in the job +can you attach the right patch +I also cant tell exactly where the exception is occuring because line of is nothing but a null check there cant be a NPE thrown there +Quick question how do you invoke serialization? It does sound like a bug I just want to know how to reproduce the issue +Attaching updated jaxb provider jar +I think as we clean this up more with the other ongoing refactors itll become clear which of the two is better and we can make the change thenWe can continue to improve on this but since these refactors cause lots of conflicts I think its best to treat this as final for this issue and then do some more improvements in followup commits +will commit if tests pass +Test failures observed in node manager and job client +Any objections if Iapply tonight? from me +POI dependency bumped and patch applied in r +So maybe there is no bugEdit the problem returned again + to trunk. +I submit it first to see if it broke any existing unit test first +Thanks for the update and confirmation +we can increase the roll interval from minutes +Commits for and have been committed to product branch and will be in ER +I reopen this issue based on input from Vasily +Server started with fle set to true the server listens on all +As Christian pointed out there is one more step after the svn updatechmod so that other committers can also do the svn update in the future. +It would also be good to document why the lock timeouts need to be lowered +This approach was used to minimize the impact of the escaping algorithm on existing property definitions +Ill try to implement your advices and reply with tests result +Just need a review then +charactersThis could be an alternative to fixing the? operator in core +I dont have time to do that testing now however if anyone can confirm it works Ill apply the patch +Sorry I missed removing some imports from should be fixed I fixed those issues in updated documentation fixed an incorrect test made output clearer about timezone stuff +jsfplugins +Youd have to build FOP yourselfThats not a big deal though +trivial patch to solve the issue +So in this case you may have a corrupted file +thanks again +Hi EricIm not sure we want a checked default payment +Does this fix the problem? Is it something others running should pull in +The keys can be arbitrary and on multiple nodes the api would split the query into respective nodes and then assimilate the results +If thats where the problem is coming from the fix will be easy +The is updated by the client every time a reply arrives +I think it is ok to do it in every plugin +marking this as bugs marked as resolved before and tagged as version as fixed in RC +Ill give it a try on Linux. +is shipped by MR +I was hoping to keep this issue open to attach any further patches for this +bq +list merged with as dublicated +defer all issues to +Now that work is underway with AS all previous community releases are +We may want to consider marking this an incompatible or making it a config option because the new K buffer causes the tests to hang on OS X whereas its fine up to K +The text was revised +Data is shown abovePlz note i am not selling this patch I am trying to find a better performance for our use case which needs compression +In my local repository I have only +classes now implement Comparable allowing their use directly in ordered collections. +Btw it seems there s still a problem as the has not been included in the patch afaik +Also I think its fine to expose the avro +Exactly the same issue here with JAXP Tomcat +I havent ran your test code Enrico nor experienced your problem before but it would seem that this issue is worth investigating +Im talking about fixing the code so it doesnt generate those unchecked exceptions in the first place +Yes that is a good point not sure what we could do about that however +Thank you very much for the patch +Use the if you need a solution now with +So we need some coordination +Ill test this on Linux ASAP +There will be more plugin related goodness coming in the next few weeks and some maven related work as well +Thus the region is considered assigned to RSB even though +Assuming that goes well Ill commit to and soon +Using pixels in the theme is evil! Otherwise it depends on the theme and how the graphic artist styled it +I agree with Minor that the default mode should be silent even if the runtime is making default guesses +Done deployed to both repos mvn and legacy +bug +It greatly simplifies the integration +I now solved the problem in my special case +The problem with module exclusion is in fact described in another issue +This patch seems to hang the logloader jitrino smoke test on Ubuntu did you run full tests with it +Thanks Francisco +Will start testing under real load by early next week +error messageError while scanning folders in IMAP server Server unexpectedly discounted Success +Verified with AS which was the base for EAP +Wow to commitI wish we used Python as our build tool +Im going to fix this +All the test have been changed to make them by default +it is a major performance blocker for our application +The M reference guide refers to it as proxy as well +Attaching backport patch +The problem with and is the definition of priorities whether ignoral is to occur before joining gettersetter pair renaming or afterwards +Updated patch to fix the tests +Its probably better to do that in a separate patch if we think those statements need to be tidied up +Question How do you plan for to authenticate the AM tokensI thought I covered it but missed stating that RM will share the underlying secret key corresponding to AM tokens as part of just like the one corresponding to +Merged test and changes in one patch +This is meant to demonstrate the problem only its not ready to be commited since it disables other tests in the class and forces all others to use +this patch is not applying to the branch +But I am not opposed to +Thanks +Right this is right now mutual exclusive and Im wondering if this is a real use case In your scenario you could simply set the number of steps to and increase by two for the second and third part of your three part process +Awesome idea Robert I was staring at the Solr code a little bewildered about how to integrate the optimization +bulk defer to defer to +Gabriel one more thing +Thats all really no problem +Theres still an underlying bug somewhere right +Assigning to Scott who has the required testing env where I dont +Bruce Ralf as long as you dont introduce instability so shortly before the release I am fine +davidharkness Yeah Im impartial to syntax +If this is difficult to follow or creates any hardships for anyone give me a holler +messy +Sorry +HiIs there an ETA for this fix? we are definitely interested in the fix +Im concerned this might be a security hole +RM currently expects something to start in a container within a timeout after allocation +Thanks James +Committed revision Im not planning to backport the fix since its such a minor issue +I reduce the stub size to in the updated patch +If doesnt slip too much Ill inquire about getting a freeze exception if necessary +its a slippery slope. +Works well Linda but the description is not the good one +s proposed solution sounds good to me +Can you try this patch and post the results? It doesnt fix the problem but Im having a hard time reproducing it and it adds some more output to the true optionThus you will need to add extendedResults to your flags +Okay I think Ive managed to answer my ve attached an armoured GPG public key in a jar file that has been signed with the BC code signing Key ID is BDDBFCBABIve generated detached signatures for the jar files and the me know if I need to do anythingDavid +Fixed common page used +I see many items that are still relevant +Fixed inWill be closed after adding the regression test for the issue +The current code streams data from datanode to client +closing issue. + thanks david I didnt get to open the issue +Patch to clear queues at the end of a perftest run +this task so that all the reported issue are not overlooked over time and can be fixed +Thanks Sandy! I committed this to trunk and +I wonder why you did not check that? By overseeing this we could have run into a desaster. +Im looking into the failure now +The attached screen shot indicates the correct locationAlso you will not need to add the attribute +We may want to fix this WMS behavior if we post a style that does refer to missing attribute it will complain so we should make the behavior consistent +Patch file containing the changes +Included implementation of which does chmod +Thomasbq +Alexey dont we have a restore default palette function or +What are your opinion about adding Ehcache as I am trying to use in my to bsp work +Something like activemq etcId be ok with something along these lines +Removing from +No planned on that task which is purely technical. +I still think throwing an exception for this is a mistake and makes the API harder to wield +I attach a patch to +Please note the behavior of will remain the same thoughPlease review testOleg +All tests passed +Test case in bugfix as +Lets revisit this issue in the BETA development phase +currently fails the patch fixes the failures and adds several new test cases +So I cant use that for printing out the Basic policy +Jacquesjust to clarify that I am not judging your work +Perhaps we should add a marker interface or something similar that an that does need this functionality must implement +The proposed release may be assigned Alpha Beta or General Availability classifications by majority vote +This is fundamental to the way the multipage editor works you can edit in either the diagram or the BPMN view but in both views your are bound by the functionality we actually support +I think this type of warning cannot be sent within the protocol specification +Preparation for Gora release candidate +The wiki is really a better place +and should be in the ER build +Please submit a patch with everything fixed if possible +Thought of getting everything fixed in one go instead of creating multiple small jiras +Hoss +understood but may you how can i proceed in this case to fire an alert in particular moments in the page? It would be helpful +just committed this +API is and should not be exposed publicly. +Anyway the reason for this is that git reset is blocked by a git hook +I implemented equals and hashCode for and and created test cases +Please provide a way to pass the ant home to make +With this patch the old implementation code path has been removed which makes the code much easier to readIn addition to applying the changes in this patch the following files that are part of the old implementation must be removed before rebuildingI have verified this patch on a checkout from this morning my time +The following functions in need modification initializeuser initializetask initializejob initializedistributedcachefileI will try and add unit tests in appropriate related test classes that call the taskcontroller with relative paths in arguments and verify the right error occurs +Revision seems to fix things when I use in the test project. +Yep get simply assumes the incoming strings to be encoded as which is probably not the case in alle environments +Well i am using M right now but i did not test the behavior again +But i dont see anywhere in the spec related to Fileters they will have the same rule as servelet inits +Indeed weird +Will start working on this +New patch with state machine for stored fields. +You are what? Scheduled for Castor work? Does this mean that you can officially spendbook time on Castor? Me too please +Technically youre right but we do not consider this scenario as valid +How are you enhancing your Entities? Please post the contents of your file +I dont know when could finished and whether it will affect this issue. +Yes this is a bug +Here is v +Committed revision Thanks Johan! +Tomorrow around this time is good +What do you mean by Unit test? +I am looking at the ovidius standalone release as per forum comment above +but I still think the message is a little too aggressive +Bulk closing all of these resolved issues. +So I was trying to cover all the bases in case we need a release from one of those branches +Okay thanks +Ill wait on releasing it until theres an actual release of VFS +This will be done by setting pointer to stub object in corresponding serializer and deserializer +The ejb and ejb schema expect different case +Fixed on +The inner class in Fuzzy query was never intended to be public +Verified JBDS +This is the patch that Ryan wrote and that fixes the issue +The proposed three new have been implemented in the trunk +However the same issue happens on IE and Opera so not sure if this can be a bug in all three it is defintely an issue with HTML code +It wont be a complete showstopper but that functionality is easy to replicate +A comparison table in is an awesome ideabq +I thought the missing perforce plugin could be an issue also when I attempted to resolve this earlier +Im not trying to discourage you and gladly welcome the help but some smaller patches and dialog will get you much further +The dictionary can also be determined in some cases by looking use the FIX session using the session ID in the Exchange headerThe QFJ configuration has been simplified for the component +Kinda like the same discussion we had with the proposed osgi component +No problem +bq +Ive reverted out my local changes and the test seems to succeed consistently now +The powertoys shell locks it but other most likely do not. +This jira deals more with job automation as part of validation tasks in release this as this is directly not related to any issue with EAP builds. +However I could see a need for this in some situations +Like the Flume project +Ill change my new additions to make them the same as unless I hear back from you again +solution is found by Alex now needs to be implemented by Gleb +I will comment more on CMS search +Attach +handle SNAPSHOT dependencies as wildcard versions +for the patch +I just reverted this since it was causing things to hang +fix this so I dont think this is an issue anymore. +to verify i believe this has been included +We can go ahead and commit this for now +just move journal related operations into a +This has already been done see and also. +It asks for my usernamepassword +Linking to which heavily uses ZK +PrabhatJulien is there an update on this issue +No one is forcing you to upgrade Maven nor the plugin version +I am ok with this change but want to voice concern that we should not give up on running on small devices justbecause we have given up on support for the small CDC configuration +Rajesh I believe you are certifying EAP for so I am assigning this to you so you can keep track of the problems + diff to use newer for the ant build +Its interesting this functionality is really only in trunk Ruby nothing released at all +to commitThanks Kim +Back then type forwarding might have been the better option but unfortunately nobobdy thought of itRight now renaming wont help in your case would it? would still be bound to the name lognetFrom the top of my head and without having a +I will investigate how to fix projects though +perform some restricted operation inside a connection to authorizationDB after setting the role used in the URL? Which would then be checked by the create db code in Derby? Or something else +looks to and trunk +Thanks +If cacheRequestuncacheRequest is mostly used by client would fs shell be a better place than for themLike snapshot createdelete operations are fs shell commands and allow are dfsadmin commands since only administrator should use them +Stack Patch that matches reviewboard was posted and passed Haddop QA +Use explicit and unique id on components like form or subview +But yeah I doubt users really appreciate that distinction muchbq +btw the recent patch applies with no major issues some import jazz but intellij fixes thatIm working on perf testing this but Im having problems with my test env right now +Does this overlap with Dons fix in +My bad I forgot to update it earlier now it should be upgraded. +Can you download beta and ensure you can still reproduce the issueIf so think you could post a sample application on or elsewhere that I can use to track things down +i cant seem to replicate this +mostly clean installation of JBDS +Dup of Tim thanks a bunch for the patch +Hi David thanks for this +If I put it in manually everything works great +bq edit sorry simon charset issues on my side versus thing happensbq +thanks bruno! +Please let me know if there are still problems with this +When I initially looked at the code that was my first thought as well +Any objections +Please confirm the current status of this issue and if its incorrect change it as required +I am on Reviewed the patch and it looks good to meOne thing missing is the tcpsyslog source test thoughI have added a test for that source on so will just need to add in the source counter test there as well +Looked into it a little more +But there are a couple of guys wanted this one to be implemented +with respect to generation then that is an eclipse feature for java projects not for the dynamic web projects +Applied at r. +If you use Maven itll automatically select just a single version of a dependency +Can someone please close it for meThanks in advance at Rakeshs request. +this is my patch for this bug +System memory management functions are well documented stable has well defined behavior and clean interface +Its not in use by the framework that I can see either +Reduced priority of non critical issues which have current there is already a option that I have seen used in the Grails functional tests +configuration should be addedmerged with the default namespace configuration which wont apply to the system tables +Close for all issues before incubation +to what Roman submitted +Committed v with default of seconds as suggested by and specifiable duration as suggested by stack +In keeping with the design diagrams introduced in I have modified the design +Ok now I have the basics working for supporting multiple framework instances at the same time +Works for me agreed caching is less than ideal tried the patch out and it works great thanks Andrea +Karaf already ships with pax loggingShouldnt this issue be closed +See final screenshots +Thanks for explaining that Runping backwards compatibility is indeed important +I create the original because class has a problem that the cache file is deleted +Hmm disregard last comment different issue +It was yes +My understanding is that a prerequisite to the creation of the objects is your object model +This is an issue on the platform +Ive been using iBatis for years now and never needed to especify jdbcType why we should do it now its lots of work +So there is at least another code path not involving table compression which sets up this condition + is correct +All changes in so patch is rooted in that file +Affects only under FF when overflow at +In order to use an unparented loader we need to find out how to obtain the URL to the jdo api jar file in the users local repository +Ill investigate +Fixed +It would be great if there was a design description to initiate a discussion +The changes in will resolve this issue as well whether security is enabled or not. +Wrapping would be the logical choice +This is the stackrtrace generated +No simple matter to parse the ZZ and then send the code on to +We cannot make th proposed changes to the signature because they would be believe that the problem is to change the return type for the get method and declare the principalChanged method try to do something intelligent as I wrote in my note Dec AM you solution does not hold and you have not fixed the bug after my opinion +Weve talked to John about this issue +Updated the documentation +What do you see in +Wed need as well no +Bulk move I think the idea of this issue was to publish a jdk classified jar in a maven repository so that the Solr project doesnt need to publish its own version +Further more it will simplify the grammarThe only tricky thing to control is that a name is used only once +resolve as out of date because of month no activity +This may relate to an unresolved bug when launching an editor for JSPHTML enough info to reproduce. +Can you give some information about what happened and what was supposed to happen? And a description of how it was fixed in Dramatis? That will help us decide if theres anything we could be doing better here +This commit comes with the same comment that I noted for the branch +Glad that you agreed on it +This test case is still failing to me +Any subsequent failure will result in the exception handling stopping the bridge +Commit +rename file +OK if we do that I propose that we lower the priority on it since thats also misleading +Lemme see if I can rebase the patch +Modified code to accommodate changes in API and changesAdded jboss jars required by Teiid +Can you submit a new patch that also implements the nested element? +Please review if you have time +Hence there is only the Save Cancel buttons +As I mentioned above Cloudera has a done a lot with moving this issue forward +I do see that all regions were assigned but noticing a disable table request after that +this is the patch to improve +Ive just committed this to trunkThanks a lot for the contribution Andrew. +Allows to talk to +Thanks for addressing this Andrey! Ill commit this after Jenkins responds with for the latest version +To be more precise for some reason the integration tests are marked as successful but the exception is still thereEdit Ive traced the reason why the tests now pass the sequencers have been added only to the second repository while the tests use the first one +Do you possibly have an additional ini lying around somewhere in your classpathThe exception should be harmless though its just a notification. +Rohini since it is an additional field I dont see an issue with backward compatibility +Are we going to fix this in trunk as well? +pushed +Specifying into the JDT Junit creation wizard will cause a status line error if the jars are not part of the project +I am reopening this issue for some additional work +After further investigation the problem lies in that +To be specific in custom HTML pages respectively to be used in form simplification macros. +we can provide extra verification for the syntax of select statement to address the issue raised +Verified on +Can you provide a code sample that can be used to reproduce this issue +Thanks +Thanks to Gabriele too for the help and nudge +Attaching another patch based on Jitendras latest comments +Would something like this be acceptable by the upstream +Thanks Ulrich will check out +Hi AndreiI take it this issue can be resolvedColm +This patch and associated unit test does fix the problem +Will do my best to get something up tonight + and live +Then I see nothing more to do +CR becomes GA so for now no issues to be put there +In enum shouldnt the first constant should start with zero because the corresponding java enum doesnt assign any values +x still wont work because the compile doesnt expand this it relies on the shell to do XDT will let you specify args that dont start with now at least +Thanks for Ted +When the call returns the sequence is either established or the base http timeout has been reached +Hudson seems to have gotten confused and wandered off +should be pretty simple its intent is simply to load from zk if possible before falling back to how would load essentially +Thanks you NamitSent from my iPhone +Thus host port and framed are assumed to be consistent names always validated the same way +Getting the sync marker doesnt seem a huge burden on top of that unless youre figuring youd skip to the next metadata flush before you try to make sense of the stream? How critical is this use case? If it becomes an important use case we might define a container or use RTSP or somesuch rather than using the existing container file format at allNot that this isnt an interesting area but Id much more interested in +to. +You patch was committed at the revision +Brian is this supported or can I remove the book entirely +Do we want to update the repository with my patched version of? The patched version should not affect existing have had a look at and as Dominik notes the API is very different from and will require rewriting a lot of client side to be undertaken lightly +jar file with the unusable classes stripped out +I appologies for the just ran this with the release and there was no crash +Please upload WSDLXSD for us to recreate the problem +This is documented in the javadoc for +Thanks B. +Makes sense +restructuring +If you make changes to any command interface say dfsadmin then you will have to change this document and also the main document for dfsadmin which seems redundant +By the time the has loaded the jar it already knows about allthe files in the current jar +I have the same problem both on Firefox and Chrome +HelloWould you please try this patch? Thank youBest regards in r +Im not sure we would want to embed it in the name node +the import list of the has already been cleaned up earlierPlease close this issue if you are satisfied +Part of a bulk update all resolutions changed to done please review history to original resolution type +Please assign back to me when you are done code freeze is Thu release date is probabaly not be available in time for doesnt work with unit test jaxws +Think it can be closed now +Ive got a As GA installed Teiid and Designer running and my works for meRan simple Parts example deployed VDB queried +This is a couple of small scripts to help you replicate the issue +Its unnecessary as non of the assignment are allocating any memory and several of them can just be removed since they are setting things to what they already would be anyway +I will commit after I have edit the +However there is no way to create a couchapp that interacts with a secure database as theres no way to authenticate the user if they hit up the applicationI dont want anonymous users to be able to access the information in my database through REST +added popup actions on Dependency +Testing this issue Ive confronted a problem related to taking some time to tack + +From pig or hive there might be more to it +Unresolved issues to CR. +It has cut a single process executions JVM time by for us as indicated by that Ive used a which requires Java +Therefore this check makes sure that an error is returned when there is a mismatch or issue with the content type header and the actual soap fact is that not everybody is using HTTP at transport layer but for example a legacy transport protocol which is not soapweb services awareIn this case if an endpoint accept both soap and soap the only way to know is to check for the soap namespace declaration in the soap envRegards agree with Sebastien +btw the current patch requires +Eva has been assigned to this ticket as the writer and had a question for Tim +Removing them require to unify the foreach variable name. +Nice of you to look for inconsistencies in what Ive said though +Then if posted value differs from stored one it is assumed password changed and it is obfuscated and overwrites the one stored in config +The JAXR setup should be part of tsant steps are not desirable and will prevent QA from automating CTS update the build and not the doc if possible +Thanks for taking care of this Gavin +Note that I copied these to local temporary copies out of the repository first because maven fails if you try to do it within the repositoryCtempmvn deploy commonj Dfile Durlscphomekelvingoodsonhomepublichtmlrepo Dpackaging moving the maven artifacts from the staging repo to the true repo the version stanzas of any artifacts which had previous versions had to be merged and the md and sha digests for the files regeneratedfor ref I used fciv wp sha add sed n p dd ibs count andfciv wp md add sed n p dd ibs count The directory for putting the distributions is wwwdistincubator and the maven artifacts go atwwwrepo +a fleshed out sampletest java classes that recreate the problem would be a good startthanksdims +This patch addresses the comments +Attaching patch +Id be interested in your comments about markers. +This does not use a value mapping but rather sends pure HTML params +The only question to me is if this is something so basic it should be handled by the normal parser along witht he regular meta tag +I believe the NPE could be thrown when saving the message after having called when stripSignature strippedMessage! null because it calls setMessage with the same message +The one at is the default and I would rather it be empty or have a way to do that when creating this +logj has a upward compatible replacement of the called and it may be best just to drop the existing and port the +has been marked as a duplicate of this bug +I see the collation property the same way meaning although it makes sense at the database level code it applies to all the databases in the system and hence should be saved in the file + +svn commitSending productrosettasrcorgjbosssoaesbactionsroutinghttp productrosettasrcorgjbosssoaesblistenersgateway productrosettasrcorgjbosssoaesblistenersgateway productrosettasrcorgjbosssoaesblistenersgateway productrosettasrcorgjbosssoaesbmessage productservicessoapsrcmainjavaorgjbosssoaesbactionssoap productservicessoapsrcmainjavaorgjbosssoaesbactionssoap productservicessoapsrcmainjavaorgjbosssoaesbactionssoap productservicessoapsrctestjavaorgjbosssoaesbactionssoap file data +acaaaaf on master branch +This issue has been resolved for over a year with no further movement +Trying out the new iframeless bridge on iOS might be a worthwhile experiment +I agree that actual request response messages should take precedence and be added to the context when available +I do not recommend using a hierarchy +Verified in but wont be closed before JBDS build with release +finderName find finder class +Thanks for reporting this EricI think youre right +To avoid confusion I removed the topic that was modified for the first patch +Thanks for all comments +Thanks for reviewing and applying this patch so quickly + +This bug can now be closed. +Stepping back it would be good to know if there is any specific requirement to have these configurations files inside otherwise I can come up with a new patch that would move these files as regular resources outside of the +Changes committed in revision . +Great! +If you osgi uninstall a bundle listed in it will still get started the next time the server starts +Please note that it is always easy to create a tree from the stream if desired but not vice versa +looks good to me +Ok a few emails on the list that didnt get in hereFirst the new JAI release candidate supports GIF so users could just use that +Unfortunately the new snapshot just verifies the signing of my instead of actually signing them but thats a different problem +Rajesh can you make those tests conditionally excluded from the testsuite +A fix is in SVN +there are a working sample over the Tasmania shapefiles from Cameron Shorter built with Mapbuilder so it should be integrated as a demo +Also affects compiler was a hack anyway +Will this change go out with the next release? Are you fine with agreeing that this ticket will be a blocker for HiveYes and yes +The optimization of sending the leader and isr request only if the leader is alive is flawed +I posted that in response to the below as it seemed slightly apt for itbq +Thanks for pointing this out Rafal. +This is largely complete but in flux +Hi EricDo we have some kind of IRC channel for James Server Mailbox Project discussion +Providing an implementation of the proposed solution +Thanks Harsh I revised the patch +Dyres snapshot patch contains a simple fix for something that clearly looks like a bug in +Can you also file similar issues for the workflow manager subsystem as well? In fact there are a few places I think this occurs in the docs maybe you can grep for the same issue in the other docs and submit some new JIRA issues and patchesThanksCheersChris +You should be able to go back to and log any comments +Thos two issues are related +Heres the pdf causing white image after conversion +Committing +This is about a liability for the use of Microsoft IP in commercial products not for writing code +is there anything I can do +The client gets the list of masters and sends the messages in a fashion +Im of a split mind towards one of these rules the one that strips and similar +older issues for Apache since were no longer actively working on these at the moment. +I need to be able to reproduce in order to help you +Setting prefetch to is not ideal because it can impact performance +Well done Adrian easy to follow o +Targeting compatibility with +Personally I would like to see the short form header replace the current longform header entirely for all ASF content. +Can you assign this to somebody +Mixing a couple of things I missed a couple of tests that I broke +Ill check in the patch then +I dont understand why youd say Seam and any other thing that is not the AS itself must use the Apache original version for the Platform +Can you please rebase this patch against trunk and resubmit? Thanks +This replicates if you save the Child and the Parent also has a reference to it and not when saving the ParentSee my previous example when I save sr you get rows if instead I save r I only get oneI hope this example helps +Attaching a small maven project which contains the before mentioned test case +Coalesce duplicate services into a single aggregate service +Heres a trivial patch +Final version of patch +Hi Enis I took the liberty of writing some code to do asynchronous RPC responses from the server +Thanks for testingSorry for my misunderstanding you are right moving Writables from to wont create any dependencyI thought it would be nice to have one package which contains all Writables +It also works with the latest version on Karaf andbin cmueller +I could a file image here Under each file could be a a match NUMBER +I also merged the back because the bogus values setter was missing in +Can anybody tell me any work around to remove the openshift server from the servers paneluntil you publish web tools beta +Documented in the EAP release notes +Replace diff file because the previous one was missing an ASF header in for a new Java file +It would help build confidence if you had a unit test to prove that the was indeed +What about comment ? Making the constants invisible +Temporarily assign yourself and add a comment that you are working on it +present in jboss eap +Eugene thanks for reporting thisShould be fixed at revision please verifyBTW property is set by federated script only at the moment to be fixed yet +Does that include these fixes +Well take a look at this +Doesnt the hash case you demonstrate just store the object directly? The signature only takes Object so theres nothing to coerce to +Ive just committed this. +Id just realised that my cursory investigation from before was a little off as the actual time is read from the lastUpdated element of the Metadata and wont be resolved until it is requested +This is filed as +Thanks to Marius Petoi for provide this patch +We made some changes on how JPA is detected and the is instantiated +are immediately followed by ampnbsp if used as the first sentence of javadoc +is enhanced to provide various storage directories +Your test case doesnt set the prefetch limit on the so the consumer never reaches the of prefetch limit since the default is so the messages arent ackd in your test +Joe or Tony any ETA on when you will get to this +The missing parent directory fieldscbwrepoppfctp is not in the image and no other entry in edits contains this directoryIn this case although avoids the crash it does not help get the missing directory back +Looks good to me +I dont know what happened my guess is that was using a cache that was causing the problems I had +can you please add some more details on this? like repo stepsthe description is empty +update resolved issues. +This is primarily to debug what the autodiscovery actually discovered and if you are running multiple entitymanagers logs are not enough if I cant link a specific instance to which of the many autodiscovered andor manuallyexplicit configured im currently holding on to +Attaching the patch with approved review Prasad! +You can find them attached as and aswell as and The were generated using the following commandsopenssl genrsa des out openssl req new x key out days openssl pkcs export out inkey in keytool importkeystore srckeystore srcstoretype PKCS destkeystore keytool import v trustcacerts file keystore openssl genrsa des out openssl req new x key out days openssl pkcs export out inkey in keytool importkeystore srckeystore srcstoretype PKCS destkeystore keytool import v trustcacerts file keystore The keytool commands for the truststores have been interchanged so the public keys end up on the respective others sideWith this you should be able to reproduce my scenario +this is for bug reporting please go to our forum to ask quesitons do you have a valid test case that can help us to identify this bugyou know from the attached i cant see any helpful closing stale resolved issues +fixed unknown +tmp aggregate file for the application on that host +If thats the case then it can be present in trunk too +I was originally liked them but they mostly end up creating noise +Im having the same problem and the only solution Ive found is the one mentioned above +Closing. +fixed +Does this make sense +Marking critical against +But this is related to other issues which should wait until after this release so I think we should patch and look to ship +Producer used sync mode +BTW I am using a version of that I built today from the code trunk +Thanks! +Update the patch +A slightly better patch +Typo correction I believe Ted meant commits +For instance we can treat individual elements in a array are comma separated and space separates each argument +The new patch addressed your comments and Thanks Jing +Ah okay thanks for clarifying So best to use two distinguished settings then +include Iqaserviceshudsonhudsonworkspaceworkspacelabelrheaphudsonwdroottmpoptnativeincludehttpd I +Moving code between and will not affect Derbys static footprint on JES +Screenshot of the mbeans not being removed +this looks good to me +Im not getting any exception + +Patch is applied at +Checked in missing view contribution into dqp ui +sreekant Applying the patch creates orig it looks like a conflictCould you resubmit the patchtnx +Invalid case. +Now it passes inthe whole page of detailed error message +Ok +you Ashish Vijaywargiya and Pranay PandeyThanks and RegardsAkash Jain +I just committed this +Since this uses the ajax theme this will need to be addressed in the dojo plugin in must be a problem on the application +Id be happy to help show the steps needed to reproduce this if needed +Why do you pass peak expiration time as current time? Would it be good to pass the close to the end of current hour for example based on calendar minutes? Could be imprecise in some freak cases but much less object creation +This is to emulate the extended stack trace Logback provides +dupe of +IMO its a bad idea to check out sources or unzip src tarballs for Maven projects on windows in your desktop directory +Somebody please add these files to rampartinclude directory and remove the rampartincludeomxmlsec +Are you sure is really thatAnd you dont need the marshalunmarshal as sending over JMS is object message by default +Can I take ti that you have tried to run the XML CTF suite and that it finishes without any problems +Todd thanks for the infoI am kinda suspicious about the memory requirements for the Thrift server +Fix warnings from Jetst to do with not closing streams +This version integrates the last comments the trunk +fixed together with the setter in this second stale resolved issues +Thomas suggested that we should modify conf to include and as values for conf key since this is common case for hcatalog +Will address in +Any additional issues should be filed separately. +and are minkill and maxkill respThe patch is fully backwards compatible in input and output +I have not used this change at runtime as I dont have the samples to test it +Alternatively make sure the startup order is set so the smx specs bundles get loaded long before anything else +Should be fine but not tested as per Alexs suggestion. +Matthias a few clarificationsa Youll need to leave the as is in +although I guess that is a feasible scenario I would never do this myself +DDL command failed Havent seen this before +I think the difference is that in the keys a ordered by a Listwhile in the values are ordered +Interestingly as soon as you start a operation this gets right +seems you are considering the same problem +Hi UmaThanks for the Review and Suggestions for the optimized and good looking code +oops forgot the! character +Just for the record a separate issue for this JSP compilation problem has been filed shall we change this to rename packages to considering +WDYT +The feather Licensed for inclusion in ASF works is missing from nd patch +Misunderstood the situation sorry reverting back toI think I could do the change its the testing which is hard +This cannot be done by reading configuration parameters from the metastore since it is per job and application defined +done in r +But its related to +Ill fix the issue re +We dont know yet why it is GB for you and not +Rob any ideas on this one +full stacktrace in in attachmentI use version of richfaces +Fixed in r +Yes these errors will appear in Problems and Markers views +Thanks Robert! +Could use review +Oops forgot to remove the debug message +Id like to commit this tomorrow +wms packages isolates their services and operations code compiles and unit tests run +Thank you Tim +to be clear do you mean if a project is selected in the servers view or in some other viewI doubt Id get WTP approval to have those actions listen to selections outside of the servers view +Thanks for taking care of it so quickly +Yeah that seems important for to as well +If I change the fooFault element in the previously attached wsdl file to have a named complex type of the same structure the generated code compiles without errors +Committed to. +No longer applicable. +See comment toI have verified in the loanappconversation sample that when there is a conversation id available it does get injected into the member annotated by +There was no real need to actually check for the existence of table at that point +So requesting states tigerRoads would result in +Could you please attach a patch with the failing testcase? I will try to have it fixed as soon as possible! TIA +Moved to the REMJMX project +A version field has to be think you misunderstood the problem +I continued to pursue the issue at that time but got no reply to my last one or two messages +The case that masterKey is null and the index file is not existed which means we dont have the master key to create index file so we have to throw an exception +Id suggest changing the name too +Looks like Galders fixed so unless theres something really tricky I dont see why he or I couldnt port it to an EAP branch this week +This check is currently only enforced in UI +I opened and will resolve with a fix there. +jasper is the best reporting engine and it integrating with tapestry developer can make creating dynamic robust highly scalable web applications +Will list here the dependencies with bad poms +created in revision +This patch brings two Hive admin docs into xdocs The menu is changed in +This one will take a fair amount of work and testing and Im not sure we can make it in time for as wont fix reported against which is long dead version and superceed by +Applied patch and added autoregistration of chunk and skip listeners +patch has been applied when xulrunner is not supported source tab will be loaded by default +If the incorrect order of start expressions bothers anyone then well need a clear test case and we can reopen this bug or open a new bug to look at fixing that in the time frame +Excuse my ignorance but whats a bit Windows OS? XP Vista? Which one do you useI dont know how I could test this unless I purchased a license or used a image +Yes we should be smart enough to figure out that its a Ruby object in the hash and serialize and deserialize that appropriately +Groovy Script and then see NPE written to +bq +No patch burt sorry for setting time as i am a newbie to grails as well as to jiraokay i have done the examples from the doc provided for pluginfor which i get compilation error which i have pasted on the previous posti have downloaded latest and put up in Grails lib directory but then also i am faced with same compilation error +This was for the feedback from you guys +I am going to leave the existing class and create a plugin for file uploads with a set of new classes and much more configurability +added test case to to demonstrate the issue fix it +Fixed on trunk +Thanks! The lack of emits is not necessarily significant the dupes are detected one level up as it were and we record information about each doc at that level regardless of the number of emitsOn the other hand the results from alldocs and changes are interesting +sar file perhaps a simple message listener that logs to STDOUT all expired messages on the server would be helpful +This needs to be there to demonstratethat you intend this to be included as part of an ASF license productTIA +Moving the classes to the tools jar or some other jar would be fine +I will start working on a Avro Specific and see how it goes +Do we still need this? +Do you agree Mahadev + +As a short term solution Id suggest that you set the next job id to be one higher that the last recovered jobs id. + to trunk. +Yes I wanted to check the actual grammar +But then again why do we have a sync create with that signature? I vote for both or none +If this is only happening on MAC we will document it in the release notes +Could you please fix the findbugs warning? Thanks +The only exceptions I see thrown in the regular course of execution are those +Patch to allow specifying the RF at creation time and adjust the CL during operations +Attaching patch with testsThis is well tested so Ill probably commit relatively soon +Attaching patch +Honestly I dont think that option existed when these jobs were writtenI will close these issues once I open an issue +Patch testing +When found impossible situation around class for LOB information the program throws runtime exception +Not for commit here. +Im sorry I dont see any additional specifics in +Please test. +Theres no event which can be used to check for leaks in that case +Please install zip then right click on and then click on extract files +In this case Ive confirmed that the patch is correct +The has been modified to remove the Undefined nodes if and has at least one Undefined node it resolves to Undefined if an has all its node Undefined it resolves to Undefined if a has a it resolves to UndefinedSome more tests are to be added to check that the fix is covering all the has been releasedMoving all related issues to the next version. +I plan to commit it to branch +Well make the change to Nexus to support the repository named +CXF includes by default but also works fine with and +ant test and all of the tests Ive made for Velocity also worked for it +So added that +Left over from some debug statements +Sheng reassigning the bug to you as its an issue in redundant router could you provide the log or setup info when report the bug +Am kicking off the trunk and builds +I think this would be much better idea to automatically create the onclick attribute +is no longer included in so the jar doesnt get scanned. +Also the attached javadoc doesnt represent what it does +Also different branches could have the same protocols but isolated from each example two applications can share most of the Channel services but still use a COUNTER or CENTRALLOCK protocol without needing unique counter or lock identifiers. +Unit test passes on cruisecontrol +I guess its based on something like jakarta common langs Enum right +I applied the patch + Extract the abdera extension complete it with respect to find a merge strategy with the other missing from the specs Definition of type Type I dont intend to keep using the Abdera extension in the Chemistry client and server code in the future as I want to get rid of AbderaIts probably useful to contribute one in the Chemistry project for people that want to use Abdera but I dont want to make Chemistry rely on it +Thanks +ChrisThank you for the clarification +Code is not too hot anyway Lets say freeze in days +Do what you said first +Lucene is not a database +Do you have an updated stack trace for? +Applied in committed revision +It looks like the JVMI believe the thread dump was initiated by us when we detected that the test had hung +The problem is at line the program is trying to assigin an to an array of other type +So the number of groups resulted are always smaller than the number of maxdocs +Finally got a chance to look at this +Committed +not committing this yet for some reason the test does not pass on my setup +please add more explain statements to cover all cases where a indexupdater task may addedotherwise it looks good + has been changed to use an Atomicinteger instead of synchronizing on the +Okay this makes sense +Ive enabled the Java Editor Syntax Coloring Java Method Declarations preference and set the colour to red +Reviewed all these issues and marking them as closed. +Please dont change the indentation in +It does seem to work fine with bit VM here as well +Would you have exact steps to reproduce on a clean database +Patch with test case and with tika parser configurable via parser attribute for entity tag +Please attach the test that is still failing +Its just for other usages of Hibernate that you need to add +Closing +I have applied in Rev +Perhaps this can be simply done by adding the system property +Result is the overall deploy is not atomic +fails occasionally + for to trunk. +Hans fixed it this thanks Gavin I noted about infrabot on Twitter +Seems like this one is fixed. +Attaching proposed patch +Should this issue be marked out of date? It looks like it was resolved a while marking out of date Jaikiran investigated and reported on issues in the underlying test rather than it being an issue with the security layer. + well spotted + to trunk +Plus the sentence in release note is incomplete +Verified on JBT. +bulk close of resolved issues for release. +I just committed this +These are the facts which the Release Manager must supply to the pamphlet generator including Release Details on the build environment Release overview Description of new featuresHopefully the comments in the template will give the Release Manager enough guidanceI like Lauras suggestion that the summary should be placed under so that the community can the overview and the description of new features +Will be addressing in +I suggest we wait for my patch +Also surprisingly enough the most enticing area for me to use Script Beans is the web tier so I can alter my business logic without bouncing my app you for considering this +Thanks for picking this one up +Thanks stack. +Ask users if anyone actually has such a use case in production and go with or based on the responsesThis makes sense to me +Can we commit this? Or should this jira move to +Committed revision +H dependency has been updated to version +The source of the problem is int he Gradle build so I can now look to fixing this properly +It wouldnt break without these changes but it would create an ordinary instead of +The parameter list is the abuse I mentioned + +What I didnt do is to add true support for multiple labels because the simple solution I had in mind didnt work +Cant you send the query as a String and let the coordinator parses it for you? if parsing a query is not too resource consuming it could make from JBC +I amend what I said they break +I propose a new section titled Running as a Service on Red Hat Enterprise Linux +Looks good +The container should be removed since otherwise there would be duplicatesbtw +thank you for your feedback +The poms in question werejavasampleshelloworldhelloworldjavasampleshelloworldhelloworldmcRemoving them caused no noticable problems in building +If not I will get a + +Verified with JBT +Please use escapeHtml properly +Why not just use the entity engine cache? Then at least it would automatically clear when data is changed etcThe running time to assemble the list per type based on a cached result from the database should be minimal so I wouldnt worry about thatOn a side note the visualTheme package name would be the first like that ie with multiple words combined in +this ticket did not mention any request for importing any initial version just setup of the repository +Yes its compatible with the branch +fixed in trunk by r +default is true for HEAD left it false for resolved issues. +So this is partially fixed in that s works on the issue still remains in how rewrites the shebang line +Thanks for the patchI applied the patch in Rev +Thanks for the precise steps +bulk close of all resolved issues prior to release. +Here is my patch +patch looks good +Hi ShameeraI tested you patch on trunk with integration tests min tutorial importing workflows using xbaya viewing workflow templates from registry viewer in xbayaAll tests passed commited the code to trunkCommitted revision I also noticed that the patch related to gsocappscripts is already commited thus I didnt do any commit changes to the sandbox locationThank you for the patch +The targetsource of tofrom String is not strictly defined +I wonder if this could be related to pulling a second copy of ant down to buildlibjars +fixed +I do think that taking on the responsibility to convert all file paths into valid is a maintenance problem +Test pdf file +Any experiences of running the build to report yet +The problem is fixed in the latest pax logging code and should be part of release +Now that is implemented I think the best fix for this is to use a namespace for the classifier attribute +You should be able to commit using mcr I am in the game againCommitted the db stuff on +Should be fixed now +Andy Sorry I was thinking along simpler lines pull the cmd construction code into a utility function and just test the string for absence of c +Also the serde parameters are wrong as well +great +The issue there is that we dont have the opportunity of telling about the callbacks after has loaded at least it doesnt appear obvious re +Evgeny the skip analysis setting is on +Closing this issue the reporter commented that it is fixed. +So that we can have clickable links etc +Jian seems like the default active menu is now the newly added RM nav section can you change it to be the section +Could be written to use runtime resolution but I would prefer to use deploy time resolution +Dont we already have those scripts +These sources are deleted on clean describe your use case +Updated patch is attached +it should also be broken into multiple paragraphs +So in this case both parts would need generation +This way it is controlled per connectionsession which is desirable +If anyone has a box with ie its possible that nothing used actually needs the html shim +Bulk closing of issues that are old and have been released or marked invalid. +Verified by Vladimir. +I think Ive followed the coding standards much closer this time +The second RC vote is ongoing right now +HiIve modified the code adding Bryans suggestions xmldocscreenshot is taken from the generated XML document I think now I can move to create tests for the XML document generation codeThanks +I think we will move this jira to the common project and update the tile eventually. +Will commit this patch against this issue so the build gets a bit more sane than it currently is +Well to see how this works in the code I had earlier implementation +We can go over its support in but first Id like confirmation that thats the intent +Verified by Oleg. +bq +contains one test class and one hibernate config xml +Link fixed +This is a problem for those who wish to use the ruby library with +Thanks Mahadev. +HiSingle and double quote are valid since JCR know but the provided test case shouldnt fail +We have assumed that all of the ref calls would get inlined because they are small and +Can I apply the attached patch? Nothing much there but have still tested it on +This patch does makes sense by fixing this problemOf course multiple things can be done with the design and the clustering approach +I tested successfully on Mac and Windows +I applied +Thanks for catching this +I think this is the same as +Doesnt sound like a bug + +Fixed in changeset Thanks for the nice patch! +I cant remember what the precise reason but I think the change had unexpected failover related side effects on the client +Verified +There wont be a release +Done and committed to trunk. +The original intent was to put all structured data int and structured data is really down sampled hash tables optimized for display only +Further simplified how special initializations for reserved layout version is done +Can we deal with it in Smooks? Or if it requires an update to freemarker then we need to punt +No not yet +Any name overrides must be done with schema annotations so the input to the generator must contain those annotations such as sdonameBogus in the input schema +On the web site look in the OASIS Standards section of the documentation +AlexCan you submit your changes as a diff? Patches are preferred over individual filesI took a quick glance and it looks like your is very similar to thrifttransport +Grant never heard back here but I think its right enough to fix this one way at least so am committing. +The erroris definitely happening because Im trying to add the real email as anattachment in the debug emailPlease confirm and helpThanksDave like a clash somewhere with the embedded emailA test case would be good taken from the code belowIs there an ETA for the fix? My client deliverable is fast approaching Ineed to know if I need to write aNo rush I just need to know the situation +committed the addendum patch to the branch +Yes but I believe that should be a separate issue and could be pretty involved +Regressions passed +Project source code tests and some notes +Ill add documentation to the website in the section under this same Jira ticket +Applied to all components in trunk +This issue has been fixed and released as part of release +Quick question since the patch increases granularity of the lock would the lock be held for an arbitrarily long time if there are may replication works +Ive committed the model repo service and model permission security code +Fixed with commit ddbf +to the patch up on reviewboard +Bug has been marked as a duplicate of this bug +Done thanks! +Oops meant to add the above comment to +comments from earlier still stand +corrects the numerous style issues +I will wait for it to attach and IO Format patches +Attached patch fixes problems with storing to in and sets up ivy to pull down hbase We still have to manually check in as its published in a way that doesnt lend itself to ivifycation easilyI also rolled in the critical changes required to make work +The transformation incorrectly captures x by value +You can run it in Axis +Committed revision +I updated to use for nowwhen the jars are ready all that needs changed is to update the jars in +Further information related to the nature of the bug +Committed to trunk and branch +This issue has been fixed and released as part of release +I have created and will try to fix it +Patch applied to trunk +Filed to figure it out +There is often boilerplate code like or global transitions that are repeated in flows +Please attach an example that reproduces the works fine for me in head +The reason is this looks just like replicate the same work done in apache server so if you really need something like that do the logic in apache serverAnyway if there is no objections I can commit the code and include it in the next release +Committed revision Tim please try this out and let me know if you are getting the expected performance boost +but it definitely isnt usable as it stands +No response assuming ok. +Theres currently test cases failing within the test suite so this is expected outcome +Petr thanks for fix +Tell me if another issue is apposite or send me off to make a patch butHow about if defined some protected methods so that a class that extended it could perform operations at index merge time +Aparently Yonik and i wre both hallucinating because both of us saw bad explain output coming form my application but now i cant reproduce it in an issolated test case and I cant even reproduce it in my app +Objections? Id like to resolve this issue +Committed to and +James server already can use email filters to detect spam so take a look to how james handle email filtering +OK After the forum post I understand what caused this problemIf we use hibernate or hibernate we will get this problem +Hi Bruno I hope that its all right nowThanksLus +Mind testing if it works now +First proposed patched does not works when I have more than levels of modules +Commited in rev +Lets get this in +changes the key into cacheArchives map to use both uri and of the file +Instead of generating a warning I guess it would be getter to fail +Brian Olivier +Thanks. +Shawn perhaps you can provide an patch for your favorite OS? On the we could add an additional check for showing this only if its provided by the backend +IIUC the needs a scheduler +Alright I added a very simple Logger interface which helps abstracting the logging for the and the without compromising the unit tests +StackI have deprecated the existing assign api and added a new api without force flaUpdate in ruby file also +See +Will be adding the updated laterPlease review and commit +I think after click ok button on popup message text verification is changed but isnt changed on UI so it isnt the text validation should be renewed after each request ve just found a way to refresh captcha image on UI change the imageurl by adding a random parameter +It works fine. +The overly strictness of the zip bomb detection logic was already addressed in so theres not much urgency here +Assign to during tidy up prior to the issue now that it has been resolved. +Once I make this change my jelly code wont function under Maven do you have any recommendations on how to code this so it can function under both Maven andThanks it out +Since Im using my new computer I didnt install old Eclipse versions and hence didnt setup older target environments +Maybe +We have to introduce the new method initDispatcher to the +how to avoid ClassforName? I need at least initially to get the loader somehow and for that I already need a class +Thanks Colin. +Thanks for doing all the legwork +Before I added the appropriate user via my client was denied and then after I added the appropriate user my client was able to connect and send a message without any problem +See that issue for commits +Its not possible to write a unit test for this since it relies on triggering the delete hook +Angela notes that this is not specific to the intermediate path signature +okTest project activation after suspension +Changed the description to more accurately describe what needs to be done +So all issues not directly assigned to an AS release are being closed +We can use the HAPI osgi bundles from central the unit tests works now. +NET differ in how a the float parsing handles an overflow +Hod documentation still has references to +Is project clean not sufficient for you? Do you find the performance hit too much if you are building all the groovy and all the java together +Deleted should not duplicate protocol docs +Possibly a duplicate of now sotres only primary names as dependency +Change flag from protected to private as requested +Patch didnt have +I am pretty sure this is the same basic problem that has frustrated my attempts to write an M plugin to support Eclipse PDE artifacts +Thanks to Konstantin for spending time and giving comments +I suspect we can remove this test entirely and replace it with your new one +Any way you could rewrite this +If a dies both NN and BN will start replicating its blocks unless you control it +Sapan Ive added you to our contributors list in JIRA +Some very large projects are running successfully with +I think we should take care to amke more tests and save any information that may help to track the problem +Because it is always possible that a client will have been connected to a server with a valid zxid gotten disconnected and reconnected to a server that is lagging behind +Patches applied thanks for the fix Greg. +I found a number of additional broken items during APT site generation and fixed them as well in r. +Bump to. +any ideas? Thats really twisting my nervs +I am marking this as resolved as I have completed the additional attributions that Ate noted as well as added attributions for everything in found by dependencytreeThe source is all clean new code so it shouldnt be an issue +Patch applied with thanks +Basically we just need to copy the logic from Validator but I am wondering if there is a decent way to share the logic given one path must throw exceptions the other return a boolean +Thanks Brian +Hi AhmedI would like to address this issue in the following release +Two patches pending reviewcommit for trunk for These patches will resolve this issue fully +Not test to help reproduce the issue +Closing old bugs +surefire report of unit test fixed in Rev The issue was that we were matching on the primitive type REALLOCATED even though it wasnt set as part of the matching criteria. +Code looks good but you need to drop the changes +If you want please feel free to open a now bug entry for the rest +bq +Since it is a problem during the clinint of the class logj is now broken internally and will not give good errors again any future use gets incomprehensible errors +Im still pro removing some stuff from contrib so I dont want to be responsible for more cruft going inPerhaps anyone using it should just copy out the code to their project +I enjoyed the experience +The patch avoids the adding of to the nodes +The proper way to post code is to create an attachmentThat said I think you will find this bug fixed in theCVS version of batikFeel free to reopen if you find this not to be the case +Defer all unresolved issues from to +Test case demonstrating problem +Maybe thats why the collectionsfolks havent fixed it yet +the same however is true for accessing adding and modifying properties +Would suggest first a survey of thrift and thrift +Can the tests be pullled into or do they have to be run separately +I will open new bugs forc to e +So Ill wait until gets committed before I commit this one +In such cases I generally resort to cut Im happy to see it improved +If a method appears there and is not overridden it is an error +Hi JustinIn my first example if I use a known operation on a resource then the servlet will return a Its true that different operations may have different specific access requirements +If the user does go ahead and change the key the application is not guaranteed to workIn terms of apis as I previously mentioned I stronly recommend you start using the Hadoop Pipes apis and enhance it this will ensure compatibility between Hadoop Pipes and HCE again please consider moving the sortshufflemerge to Hadoop Pipes as I recommended previously +Also we may want to consider doing this for open and write +I believe one check can been removed +Reassigning to Permaine to decide how to continue to resolution + +bq +bq +ah I meant WTP sorry +Test cases includedWow +Bumped release id to via subversion revision +I have committed this +Changes look fine +I was able to reproduce the bug on my machine using several +perhaps at this moment would be a better default the difference between and is not that big +OK +Fixed as part of. +AdrianPlease go ahead and submit patches for all the issues that you loggedthanksdims +Im still not able to replicate it so whatever I did fixed it apparently +It works very nicely out of the box in GAEJ +It will usage more stable tab before +We would like to be able to use a different resource bundle based on the method being validated having access to the methodbean being validated in the message interpolator or the SPI would help us solve thisSo from us. +After a few tries it can you please clarify in steps to reproduce if indexing in step is completed when you perform steps and ? Also can you please attach your Eclipse configuration details from Help About Eclipse SDK Configuration Details? Thanks +Users can wrap those in the monitoring library of their choice +Ack! The words instantly broke every existing +navis appodictic would one of you guys be able to give this one a look if I rebase this with trunk +updated release notes +Well I would say it is at best d dare saying its better to warn the user on the difference between the workspace in the URI and the one in the POST content andor ignore the latter in case of conflict. +Which JVM are you using +Found quite some bugs in SQL scripts +closed for +DIGY +Arnaud I guess Ill need the two Schema files references within your schema as well +This is a common problem too right +I even got the newest one to make sure everything was up to date +This was fixed in CR +This may uncover problems with the override logic but difficult to tell at the moment +Tests pass +I created a dynamic web project and exported it as a war file then imported it into eclipse it worked well. +Hi ColmI am on vacation so didnt test this bug fix till nowI will test this fix next weekSorry for delayThanksNitin +User was not added correctly through the RPM as per SOC +to fix the doc +Indeed when I run the test in a loop now I see the assertion fail that Ram mentionedI increased the timeout and am running the test in a loop again +This code was let as is to allow run test configurations easily but in this case I think it is more important code performace +This last one really threw me as I didnt do anything to cause the change +Thanks MarkI will test it as soon as hard to make a test that simulates kill +Instead please run these commands in your clientcd javaserversrctestresourcesendtoendcp +Committed to both trunk and branch +Actual testing to be carried out by Rupert Smith +Explicitly adding the +Thanks a lot for the contribution Todd. +Can anyone actually show us how to reproduce a problem with jsvcIf not lets put this ticket out of its misery +Spring IDE should ship RC of Spring DM and make sure that containing namespaces are supported +Constant was addedResolved +i have added testcase with different logicThe purpose is to make sure that the number of th cells must match the td cells in a rowThanks Regards was trying to write a more generic testcase to validate the generated table format +also the messages to should be landing at commits in the mean time anyway +the semantics were broken sometime after What is +Committed +I dont have time to do that in a near future. +Please check that it was applied as expected +Nevertheless If I apply it I propose to use the same font and bold as for virtual product code +Also attaching a tarball of generated html outputTouches the following filesM srcrefM srcrefA srcrefNew topic for the new propertyM srcrefM srcrefPoints the and CREATE SEQUENCE topics at the new property +Kanna I responded to your comments +You can find many existing junit tests in the srctest directory of the svn repository +Applied at r thanks! +In this classspecify any jvm specific property settings requiredNow I found that this class was needed to be able to execute the testsuite at all or I got a +Anyone digging into smartcns internals should be and whippedoh no I think I am in trouble then +Also had to change to accept Mutations instead of WritablesI also changed the generics signature of and +Rebased and merged into the branch and into the master branch. +Hi Ivanthank you for looking over the patch +Updated patch with tasks and finished +I also ran the zkfc manually through jcarder while doing these tests and verified no lock cycles + +Thanks AndyBy the way I had a look at the scripted tests ORDER BY LIMIT queries are present in various places +I wont start merging until wednesdayHow about you create a new task add remote pinghealth RPC methods to all services that depends on this +Thanks Matt committed to and merged to upon release of. +superseded by +I wish I could provide patch for this issue but I cant reproduce the situation you described +Ive copied in the original docs hoping we can just use html sources but if its a big deal and theres a recommended htmlmarkdown tool Im willing to convert it too +EXECUTE Start Eclipse so that active is Java perspective and nothing is openedEXECUTE Open Central and create Java EE ProjectASSERT Project is created without velocity related exceptionsEXECUTE Create FileNewProjectJSF ProjectASSERT Project is created without velocity related exceptions +Reopening to discuss on Forum the validity of this issue +WSDL would be happy to provide a fix for this issue if someone could give me some pointers to how to fix the code. +thanks johannes for contributing! +Attached you will find full repro instruction code and scripts +We consider it is a new feature and is already supported in +i havent yet learned the git voodoo to do merges +With this patch a warning is printed that the build versions do not match but the DN starts up just fine +I see no way that could fail because of this change +unzip the patch to NIO module +Just committed +If we resolve this issue can be fixed too +Yes the correct link to the wiki page is the one you pointed out +If you arent going to have time let us know here. +Here is the manual for implementationThere doesnt seem to be any interest in this so closing this issue. +typo Full auditing to can be enabled which logs the information sought by this JIRA +Will do +I think we should take this opportunity to address and correct that problem since this will be an change anywayWe should definitely clean up the exposure of Jackson in our API +Unfortunately not at least not with my setup as isnt compatible with +Maybe? To be honest I dont even know what does +On AM Rick Hillegas wrote On AM Mike Matrigali wrote If it is not too hard I wonder if we should disable derby from booting if the jvm is not or higher with hopefully a clear message saying why the boot is failing +And another that does the actual migration +I also stated the list of jars that need to be added to the Synapse classpath +bulk defer of featurestaskswishes from to +Generics warnings are only +Could you please provide the configuration options since you said it was a configsetup errorThanks you please explain this. +Does anyone know? A question was raised however as to why weve not run into this issue before since we rely on identifying localhost in other parts of the code +The link you provide is a demonstration that it is fixedRegarding the bug you describe David I guess it is our faulty usage of the tab +This appears to have been fixed already. +It uses to store a reference to the class it is given but then it has no handling of what should happen when that is garbage collected +Jeff feel free to submit patch we can include in the later release +Thanks! +Hopefully we get that fixed with the next bigger release +yes that xpath is real +The original patch from should be applied Committed revision +There is a test checking for roshan extension but what is missing is the test ensuring that the roshan extension is not usedI dont think this is ready for commit will the code the way you are asking +ERROR Not Found +Thanks. +Dont see it happening in time +Please review and approved on Tools +Thanks. +Fixed by Bob Fuhrer which solved several other problems too +It seems right to meMath should also not depend on core +These blockers and critical issues are resolved but not verified +This is more palatable than the anonymous inner class blob to init SPNEGO +If its not too difficult go ahead and do itId say go ahead and get the marshalrandom changes in now and resolve this bug +Not sure I understand what the problem is here +Restart the down nodeNode will log that Finished hinted handoff of rows to endpoint +Changing Affects Versions and Target Versions to target trunk and +I added a note in the documentation that Apple dev tools are required prior to installation. +Its really annoying to include the everytime just for the groovyc Ant task +Thanks for the review Bobby! Ive committed this to trunk and +Still for a I am happy for the to be checked +I was expecting all pages to be marked as free that were associcated with the clobblob that were aborted as part of the pk violation +new controller automatic name +I have fixed the branch in case +This has been run against the release +With all options dont have AJDT and Subclipse installed so it is not a surprise that you get errors when those are checked +Yeah smaller chunks would be fine and would help in balancing +This is a harmless fix which doesnt affect anything +push to +Is it accepted to change the Log format of record only for one particular operation by not compressing the numbers or did you mean to not compress the numbers for all the operationsI can extend the upgradetests based on the approach we agree upon +Other thing i am struggling with in ROO Framework is that ROO does not have an answer for entities +Before the patch the default of verifyChecksum in is falseThis is a surprise +The mailing list seems kind of but Ill try to write a post that details some further aspects of the code +Stress test confirmation will patch with required change in patch up for review on +Ok this should be completed in revs +Same really for score actually too but I digress +Refactoring of arraycopyOptimizable is done now it can be read without problems +both tests are the same +The biggest issue right now is that if we change the indexing format we have to change our indexer at Etsy which is a very scary proposition +Solved javaDoc compliant documentation is generated +Updated patch for +Great this one is fixed +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +The failure in is unrelated +When will it ever end? Fix a couple hard coded versions in the javamail configsSending assembliessrcvarconfigSending assembliessrcvarconfigTransmitting file dataCommitted revision +Finding out if master and slave are in synch should be part of the replication network initialization +solved as all changed from where just merged into trunk +It was a combination of the JIT now trying to emit yield calls and not properly handling cases of more than value coming out of each during methods +Thanks +Broke out the one huge test function into a bunch of junit style individual tests and added more tests from the test plan +Committed revision +Knut Anders +Small think this patch looks great HolstadHere are some initial commentsDeprecate the old means of batching in favor of your new method? You can have the old methods call into your new methodThe check a list is sorted probably takes as long as actual sortIf only one item in list avoid sort? Its an optimization we have elsehwereMulti should be renamed +Fixed in closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +as a layer on top of HDFS but we do not impose the knowledge of splits upon the file system +Remove Scheduler OOB instead if a message from sender S recursively calls another on synchronously +Will look into this issue when I get a chance +I have attached the wsdl I am using +Also +How can we get the transfer arranged? Then we will take care of DNS zonesredirects +fyi you can attach a file after youve submitted the issue +No problemAdded private variables and moved reading conf to init +If there is some subtle problem in the Yoko implementation that results in this problem thats really where this should be fixed +Hi Frank Raymond JeanThis patch must supersedes the previous one +If problem persists on please reopen +Feel free to reopen it if you still have a problem on but it seems to work for me +Thanks for reviewing and applying this patch for meThe intention of this patch is to support JDK as a RUNTIME and not to support the JDK compiler itself +Okay +bulk close of all resolved issues. +I will modify the patch to support only session counts +Would appreciate some eyeballs before committing +Thatll only leave us with the open question of whether we should still add a or type +Not following the standards defined by the original application simply cause confusion for users who are familiar with the application and wish to incorporate it into their Maven build process. +However we use that term in Getting Started and in the Tools Guide mainly in reference to sysinfo and dblook +Fixed on revision . +In CVS so fixed +Im pretty definite the wget command is correct if I put a bad username in I get a error back from this URL type still do not work but the jdobies user in the examples above actually doesnt have the correct permissions to initial the download so the Proxy Error messages are a red herring according to fmerenda that should be able to auth are getting Forbidden errors. +Just FYI Fernando I rolled back my Netty upgrade commit in Native EAP branch as you requested +I just got the latest grails and Im still having a problem with this issue +Ill try to write a proper test to verify the behavior before committing it on trunk +Thanks CarstenThere is one quirck in your patch +I am having same problem but with motorcycle anyone has solution? Please do the needful +Updated per Namits comments +read only clients was just an exmaple +Sometimes its really required to show context menu by key board event in that case the conetext monu is not opening in correct position +In the future Ill run precommit tests for just make sure we dont cause checkstyle or something to go bad +Fixed in closing all issues that have been in resolved state for months or more without any feedback or update +As part of reviewing the factory code I found that some confusion was occurring between create used to create a datastore instance to interact with the inidicated file should FAIL if the file does not exist create used to create a new file should FAIL if the file already existsI also noted that the factory is maintaining an internal liveStores map of created instances this map is keyed by Parameters but perhaps URI would be a better choice +Ted You are right +Thats not surprising as most TCP sockets wont break when a wire is pulled or a switch goes away +What is the version of the exported data +I think we need a svn move for committing this +It would be nice that it gets computed when the image is saved to disk +and getting the special back +Fixed by special handling of arrays since proposed solution by Sun prevents class loader delegation +Thank you very much for adding the links +I am closing this ticket given that the fix for it will be included in the upcoming Spring framework release +Trunkjobvpe +I have a tentative fix for this on the branch +The RP helper was there and assigned the mimetype a lot Marcel for your fast replyCould verify all +We can skip the test if we detect that we are on Hadoop by using Lets do that so we dont have to come back and fix this later +I maybe missing something but to me carrying index and event name extracted already from the event itself is a more complex than just passing the event itself to the method +then put a programmable proxy in front of Solr like Varnish or somethingUhhh I think that is too much to require for doing faily simple authorization +I will create a patch with option and lots of documentation and changed backwards tests +LiCan I assign this issue to Flavio +committed +xml file +Hence cancelling this LuThanks a lot for reviewing +prove the issue is fixed +This patch removes that commented out codeIf no one objects I will commit and close this issue +Best regards George +Thanks Owen! +A very silly questionare you supporting bothSHOW TABLE STATUS IN databasename LIKE tableNameandSHOW TABLE STATUS FROM databasename LIKE tableNameWhat is the difference in the above commands +Hey Scott I uploaded what I have take a look and lets discuss +fixed and put on the autoupdate note that the updated module will throw exceptions on install a restart of the IDE should help +How is this deployers issueClosedIf you still feel comfortable this is a bug under different component +I made a quick update to the javadoc so its a bit less misleading but still needs to be resolved in a stronger manner al la this issue +Tiles works withAdded a test case to the test webapp +I dont want diagnostics to inject HTML +I agree that just supporting multiple DCs in the AD authority is probably the way to go +can you please check again what version of wicket youre usingbecause the version in the generated is definitely for because this shouldnt be marked as fixed but rather as because this was a real bug so Im going to mark it fixed for the right version +i have copies running +Heres the eclipse configuration +But it is quite rare but still possible +With my current patch that information is lost +Assigning back to Jiri +In it attempts to create the trace directory if it doesnt exist +Perhaps this is our solution +Thanks for the detailed report Neil well see if we can get this addressed in the release timeframe +Hi Daryn Yes I believe that an effort like this necessarily overlaps with tasks like that and I fully expect many questions +Slavas fix seems to work but Ron could you verify it too since we are not sure that you faced the same problem that Slava has fixed +works fine but hangsAny progress in the past months on this issue +If you want to fix it in you must remove the class usage +The test failures are not related to this patch +So what i understood is query remote hosts metric over nodetool will not be applicable right? Because Mbeans will be deprecated +Patch with test which fails for high memory +Martin can you do a jprofiler run for both of these states so we can compare snapshots +When I recompiled all code using only the BUG report went away +Consolidating similar issues +I added a flag in the handler class so that it could be inherited by the boolean handler +Has this been tried on a recent build +can you please post what inside the in this case you can also see this in error additional info here? Otherwise we will close this ticket within the next days +Patch looks good +Thats a toString bug that needs to be addressed +A random list of integers is probably not helped by delta compression but is helped by zero compression +Reopening to ensure that this build is included +Thanks very much for the quick help. +Submitting to Hudson +In the run method it creates a connection after accpeting the socket +We just need to address it when we do the JSF implementation in +Thanks for the prescriptive direction here +Yes thats definitely a good idea since the original code is no longer there +Additional noteThe XA configuration uses CACHECONNECTION so if the is used in a Camel route that consumes from JMS a new JMS consumer and session will be recreated for every message +If Ill make it by the release Ill put it back +Thanks Steven I understand now + +Sample code showing how to scan for and find an instance of +Blocks will always get converted into args and passed to java methods +Committed the last patch +With JOINED Inheritance the solution has the added benefit that you can determine the classname manually in your DB while looking at the table of the highest superclass +Hi RickIve now addressed that quoted ? businessIm adding the client functionality nowthe only real policy question now is how to handle quotes a string substitution should be quoted while a null substitution should perhaps not beI need to look into the public API aspect Im just picking up the feel for the way the code is arranged the suggestions sound good thoughIll post the patch when its virtually complete rather than as a regularthinking more carefully I suppose it needs the Interface to advertise the public API it now compiles Im happy that its handling the strings correctly Ill see what junit has to say then report backBest regardsIve had to make some modifications to the tests aside from then was all clearthe new modifications alter so it gives the all clear tooso unless you need to see more test results a repeat of the then Id say the patch is ready from my point of viewBest regards needs the parameter increment in the client implementation that seems to have hiddeno +Could we check if this is still the case? If it is Chukwa Agent Controller needs an update for this jira as well +sorry about that +Craig can you provide amvn X log? Can you try the U flag as well +I think the patch resolved that issue and I also committed it into the trunkNow it is a port of camel +Note this does not hide the statusbar +I guess this is a separate issue so Ill file a new one for this +bulk defer to +CVS patch to make the configuration serializable +Marking patch available for Daisuke so that runs +The error code was changed as part of made some changes in the +IweinIts not eitheror +Please use Axis and also add g to generate all the classes +the check makes no sense if were only looking into the cached apps +Is that too confusing +In the meantime relying on session expiration should be sufficient for most cases +Is there any further progess on this issue +I agree with Henry thats what it looks like at least I couldnt find any evidence on the logs of anything unusual +Yes it was asked by Max +The impact could be like the region is removed from the regions map in master and until we restart the master becomes unaware of the regionKindly correct me if am wrong +needs assumptions removed from hcatalog with javadoc a patch for a different to name changes in hbaseSince this is purely in the hbase directory Ill defer to done by David and Vandana if I have their s I will commit thisSushanth +The test runs without failures now +Thanks for the patch James +You are right +But to use HBASEHOME and ZOOKEEPERHOME must be set by the userNote that I am adding to because otherwise it has to be explicitly added to PIGCLASSPATH to use which is not very intuitive +This issues has the Affects Versions of but it still exists in branch +Denis if we cant do anything then this cant be resolved as fixed can itWas anything fixed for this then at best resolve as partially completed with info on what is to done what is nothing fixed then resolve as Wont Fix +On this machine the default limit of K will be equivalent to about of write lock duration +Thanks in advance for clarifications +This introduces a build dependence on PythonMatt Foley helped write most of the python code +This bug also affects the user deletion +But considering we wont be returning an item from it I think it does closing. +I am going to close and as duplicates is it OK with you +One of the things we need to address is how we set up +Patch with variable pig jar path option and a couple of comments on the way this test execution works +patch apply correctly but build failedBUILD FAILED Warning Could not find file toolsspec to copy +Connection management is an implementation detail of that layer +Both of those issues seem like they could be addressed easily with standard HTTP proxies +Thanks for raising this! Should be fixed in trunk now as part of a general Hessian let me take back the above comment +Im disappointed in you HARegarding putting snapshots into the maven repository +Cannot reproduce +Updated the patch +I think the problem Im having is that the timescore implementation seems of nearly the same value as the current implementation which is just score +You should monitor garbage collection +Attachment has been added with description Unified diff against transition resolvedfixed bugs to closedfixed +Patch that fixes the issue and adds a unit test +Please let me know when you expect to complete this jira? Thanks +Fix committed. +More details about the system test we run in the cluster +I committed this +This doesnt sound too hard so Im going to throw it back into the queue +Thats cleverMy one last concern before committing would be what happens when we are able to compute a new name and put it up in ZK but then fail to create the and the RS dies +this was working in and FOP will properly crash on a lot of documents with long titles +Thank you for the quick response +Please verify then Ill mark this resolved +Ill also test against oracle and postgres myself need help from persons that can run tests against mssql sapdb and sybase +Will commit this face flogging if any is to be had +Im going to ask others to give it a try also +seems to work now with a JSF API fix in the AS upstream +No longer relevant. +Sorry the patch was incomplete please try instead + +I get the error around jar plus or minus +Which still leaves the question of failing test in the is the inconsistency that is introduced by +Verified that ant clean package works +Closing tickets +Proof of concept implementation +Its not absolutely critical in this case since we dont expect many users to implement this interface but +Sorry if that wasnt very clear from my description +Lucene is out now +Is the solution really that difficult? tables already break fairly well at page boundaries +The mail cassandra build has gone through significant changes since my last patch so Ive adjusted my code and submitted a new patch +Without an action the error message is correct +Basically the archetypes wouldnt build and other projects were having the same issue +This is huge +fixed again +Is part of our public API i +Could we use the moderators of the dev list as the moderators for the new notifications list +Ill post something nice and refactored soon +Thanks for reopening the ticket +wdyt +Could someone remove it or tell me how? Thanks +is a patch which works on any shipping method so if no condition is specified on the promo the whole shipping amount is discountedChecking on promos is pretty difficult so comments appreciatedRegards will now work when no shipping method is providedcommitted rHans +Im not sure about the location where I did the method +Im not sure but basically the regionserver will not stop until the split is finished +Is that the only change I need to makeJust to be sure we are talking about the in sa of Cserverserverjbpmdeploy rightAny help would be much can I ask if is coming soon? +Thank you +This issue can be closed if it is not in general interest +A better patch with the findings from hadrian and have reviewed and committed the just wanted to say that I plugged in the latest build into my app and everything works as expected +We should decide about changing in M +version bumped to. +It seems it calls make with null +Attached athanksdims +Any updates on this ticket from infra? Thanks +The patch is attachedThe solution is to load the tree in the callback method dynamically to make sure the tree data has been transferred back from the server +Patch looks good +Committed to and to trunk +Posted in jsr + push to +xslt files +Thanks Mukesh and othersMukesh your patch with slight changes is in trunk rev +The problem was fixed in the linked JIRA so provided wildfly is upgraded to use undertow the current issue will be resolved +Ill take another look and see if I can figure anything else out +Scott asked me to open this issue as Id run into this hitch while performing snapshot updates to metadata +apache r should nail this +Theres absolutely NO reason you should be using a primitive +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +we now replace every non bit char with a ? when transfer the message via POP +And then hope the original arq container that was launched at test start can deal with the fact that during test shutdown it cant authenticate against the server +In any case if we do this we should probably change the interface as well so on only accepts a single item +the bean that is lazy is the instance of +Hey Dave should this JIRA perhaps be moved to Common instead of HDFS +If PMD is known to be broken then we dont use it for parameter counting either. +Reverted Script Expression previous commits +Since configuration changes per release we should probably link relative rightAlso I see lots of other files files referred this way including streaming distcp HAR etc. +It uses a context to store thread local effective user for impersonation per RPC call instead of maintaining connections per context +The class uses a with a length of +Im sure there are many bits that could be improved +I actually think this kind of forwarding is a bit problematic for us +But I think it makes sense to leave the function as static method in case we find use for it as public method in future +Attaching a patch +To make this work we need to write our own exportimport and ask the user on import how he wants to handle conflicts +Actually its the other way around +This might not be related Strutswebwork have a interface that allows a to be injected such that information regarding role and its related information could be retrieved +class files is major version +Well this is a tricky one +A test case is also added where a bean is generated and populated in the runtime +We have to revert the full patch or change all callers so that they are ready to accept null +Alexei I wonder what is the intention of such issue? I has neither clear problem statement nor steps to reproduce is it just a personal memo +but when creating the payload from your self you must create it accoding to the schemaThe class you have given is not relavent to the wsdl you have given here +add the update on this feature +If an uncaught Throwable is caught it is handled by the respective handler script or the as a fallback +The default setting should produce the behavior as it is now in the alternate setting it should assume that after setting a formobject the formmodel should never be dirty +Attached is updated in response to the various suggestions +any idea when a fixpatch will be released +I can squeeze a little more out of the implementation to make it a little fasterI think its a positive change and will put it in after the code freeze +need to use displaySync or async to open it there +Issue number Velocity Engine uses excessive amount of memory when a large number of directives and macros are usedWhen a macro or directive is used they are parsed at run time and the same macro will be parsed every time it is invoked from another macro +using mvn clean install +problem caused by modjk +Attaching patch files submitted by Daniel Ellis. +Thanks Justin +Not sure I agree! means something about place types null is considered to be at all place +Not sure if you ever found anything more about why wwform was a problem Mathias but I just benchmarked the UI tags and while they arent super fast they arent very slow either +I need this information for the Debian package of Zookeeper +Its working +PS I dont like the idea of as a naming convention because we are swapping names around already from dev to development in was just looking for how to do this +Wizard is replaced by invoking Properties dialog with Project Facets page selected. +t quasi sempre nerves com a complement directe tamb susa intransitivament amb nerves com a subjecte +Andre So JBT will keep on creating a new keys on the one specific domain? Wouldnt it be better to just update the one called jbosstools +Throttling would help in this case +Also by the time I created the patch there wasnt that much static text leftI hope this answers your questionRegards +I recommend iBATIS +The is hardcoded in the +The New Entity wizard is not included anymore +Is that asking for too much wdyt +Teaches me to stop working before its too late +Resolving. +The change would be more major than this patch presumes +Committed to trunk +HOD cannot have unittests till is resolvedCore tests failed are not due to this patchPatch is committable +I like the renaming idea +Sounds like two issues here why does template code not handle broken userinput gracefully why do we have exceptions that subclass errorboth should be fixed +Fixed by r . +Fixed in revision +Is it possible to interact with LDAP or Identity plugin to get the correct fullname +Back into business +For me this test passed times and still continued to run when I killed it +Thank you Yonik! Here is the complete version +Moving to fixed by can disable JMX when usingJMX has been slater for a total overhaul and rewrite in Camel +Are you sure its safe to remove themThanks Chris I agree that should probably be a top level class +Oof this was a grueling one but I think Ive figured out the problem +Run mvn verify to kick off the integration tests +The code only uses the API currently +Dropping the priority of this issue +Please refer to commit aefdfddeafda +Attaching a patch +i think the problem im having is related to this oneIm using grails with plugin and my project fails on audit onDelete handlerAfter some tests i noted that it only fails when my class implements hasMany and the issue seems to be in the variableI hope this info can help you guys solve this issue +I forgot to mention that the problem is only with cocoa +ADB supports both for boolean values +This came up in a review for +I normally use gdaladdo on the command line to do confirm the bug also testing the layer sfsfdem and requiring a bit PNG with extended colormap attribute value setted to true +That user was trying to draw a continuous black line and a continuous blue line that happened to be offseted towards the outer side of the polygon +Otherwise when a script is run it launches a new shell for that script +Fixed in rev +Please review this GeraSee +Having a single control which renders multiple html elements seems a little odd but use usage pattern looks very simpleWill the control do type conversion return a list of integersIs the intention that this be in extras I dont think it belongs in coreregards Malcolm Edgar +Patch looks great Andrew +Applied the patch but sadly not until after the deadline +Thanks. +The patch applied cleanly for and as well compilation passedI just committed this to trunk and +I didnt get it to verify each existing license text for changes but you could always do a licenseremove and licenseformat to have them refreshed +Barrie should this be resolved if it is applied +Convert back to seconds internally +Ill be working on this ones just a small step toward improving the HTTP protocol versioning in +Comments are not associated with any sibling elements so removing any of the sibling elements will no remove the commentYou need to explicitly delete the comment +will do it manually for the has been working on this +Notes Now we have the capability to create annotated test clients +As per I know there is something in config regarding supported fields +jar in your local m repo +Merci +Do you think it will be possible to get a patch ready for review before then +My point about not able to throw Error is not valid since an Error can always be thrown without declaring as such +For comment I took a quick look it seems not trivial to remove those variables +Thanks for the review Aaron +Done +Can you please send a pull request to AS upstream containing your tests for this +bq +What is the problem with the current workflow +Added a unit test for the new state transition +Might we instead add a metadata field like +Bryan Yeah the BPA ABP change seems minor and aesthetic and would break code +Couldnt imagine that would make anything worse +I Ignored again as the failure is happening fairly frequently +the patch looks good +Can we encapsulate these Client conditions in a class +Changes +This patch resolves problemNo response in a couple weeks after asking for a more complete test case +Null vale is totally meaningful value which means absence of some value +The default size of the window is packets each of size K +The fix will be included in that release which I think we will go ahead and call +Running run Failures Errors Skipped Time elapsed secResultsTests run Failures Errors Skipped resolved issues for already released versions +Sorry I didnt include the copyright header in the and it had some unused headers and no javadoc +Documentation for Resequencing Msg +This fixes the issue with distributed for documents with special Lucene characters +Im still not sure how the test is causing this problemIt would be good to get bigtop results for this to understand what causes this +In case for some reason code is dependent on this behaviour will provide a which can be set to enable the old type of lookup rules +tests need to be tested its ok. +I have not looked at TMB code recently but the deployer reports to the console The application was successfully deployed and console shows the app running! The behavior in Jetty is different +Great work! +In test you were testing PYCC profile among others +But problem seem to be with the lik the changes done in Kirans setup for permission issue are causing lot of of other issues down the execution +I have updated the remoting config to use a callbackTimeout of ms by default +In the second case Tuscany will not apply confidentiality to the Interface element since it conflicts with Joe +So all other properties of the block are preserved +Lets change it its very patch has been committed this patch contains diff on top of HEAD +if you configure bookkeeper to be the only logging device wouldnt it read from bookkeeperim also wondering about having to create a new Configuration object to get config parameters +Also a patch against bookkeeper trunk has been attached +Merged into release +Hope to be useful in a future +Moves from to +This parameter is now optional for githg and s fixed now +This JIRA is now not part of the release defining criteria for M +Ill review the patch and get back to you +I thought we can move Agreements Financial Accounts transactions etc as tabs on Party Accounts ScreenIts this is ok Ill clean up the navigation and put up the patchRegards thought we can move Agreements Financial Accounts transactions etc as tabs on Party Accounts ScreenIts this is ok Ill clean up the navigation and put up the patchRegards Tab bar menu for tabs that have direct relation to Company we are doing Accounts forIf this approch looks good then Ill move forward and complete taskPlease give feedback Its important to complete first cut of improvements to this component +Can you check this against and +Will commit this once the git servers come back +Ive just committed this. + +According to the one in there is supposed to be a lib and there isnt one + and throw statements should also be accepted in addition to the break statement else too many false positives are be are only false positives remaining +See and the referenced email thread +This version moves the default setting of a minimum of row returned up in the routine so that all other calculations are based on at least row being returnedWill run full tests on this version +Yes that assumes using AJPI am migrating from that since it seems the AJP gets clogged upwe can close this and I will put it on my site +Yes the noncom is also a BSD style license but old like you say +extends existing class to optionally contain a mapping of property names to values and modifies to dynamically set these properties on the component with these values after instantiation +Developpers cannot run this tests locally when not having a copy of requirements folder and dont have a clue on how to change the version of the runtime they want to use for their testsIf we move it to the pom people will first look at it when theyll see an issue during buildtests and theyll be happy to see what they want to customizeSoa people always ask us when they want to change the version of the runtime for your tests +Verified in ER +Hibernate Tools version is you just typed this one in rightso the error is actually that you are allowed to launch it even though there is an errorand yes it is a feature of the plugin that its output has to be workspace path +just found that out today +RSE cannot connect to the instance +Committed to branch and trunk +I just changed the method to just protected so subclasses can have access to it +I suspect doing so will probably yield a lot ofclues for a better design also. +Am I correct that the old example spacetable syntax will still execute but is just needs to be discouraged +Please let me know if something is wrong +release destroys the mutexAlso put a comment in dfsgetattr and a soft assertion that dfsfs global filesystem handle is already set in dfsinitNo new tests as this is a concurrency fix but all the existing tests pass +assume slip +Misspelled member in the summary +Sounds like a wont fixWith the release there will be the jsr compiler. +We have the repository populated +Its not a truefalse statement to have three states + +should have been was because of ioLock introduced recently +Id love to be able to playtest out the distributed functionality of this patch +The cause of this problem is that the Maven SCM returns an empty list of changed files +I dont remember for sure but I think that this change caused some exsisting tests to fail also +Patches are of course welcome if not a general overhaul of the Javadocs is planned before. +Nascif I agree with your arguments about why dynamic revisions should be avoided in repository +Hadoop QA only builds insecureHow would we know that the fix actually works for secure +only has some basic classes needed to manage and convince eclipse and to load jdbc drivers +This is already very easy to do with iptables +Sorry let me try a less snide replyIt doesnt make sense to me to have a situation where you are working with Thrift and Avro and you want to not convert from Thrift but you do want to convert from Avro +The expected behaviour is that the console does not show a management screen for a functionality that is not there +Wouldnt be hard to have regionservers volunteer local time when they check in and if very different from the masters it could at least flag it +Committed revision +Close out the span if necessaryespeciallyoutput the summary as the value of the title attributeNow what detail or summary +Maybe we should just stick with that +on patch for +So we have more work to do +But I have not got any response from a long time +Thanks Barrie +I modified the snoop client to sent a CONNECT request and then a GET request +Quick updateI suspect a java process to cause the issue probably a remaining process from previous testsAfter killing it I had no more timeout on my first hardwareThe process was SystemLibraryJavaContentsHomebinjava enableassertions Xmxm jar varfoldersibTITmp varfoldersibTITmpsurefiretmp varfoldersibTITmpsurefiretmp +Perhaps you could check that to see if it meets your needs +canceling patch while investigating the test failure +Thanks Earl. +Changes to handle this are included in the patch for +Iwanted the class to have both the words List and Set in it +with the problem that seems to come up intermittentlyAre you seeing this on your machine? If so please file or add to a JIRA issue for that test with the details +but I dont see how to avoid thatIf you find time to give it a try in real life to confirm its working properly then Ill close the issue. +Thats why we used the header data which represents the length of the serialized objectWhat do you think about providing object inputoutput stream classes that can readwrite the stream that MINA wrote +commited +submitting patch to hudson +Ported from trunk to branch at subversion revision +I never had any code that specifically handled SSL either in the code using or the new code using +Fixed with revision +Method is of does not consider the URLs port comparing the original url and the new one hence it considers httplocalhost an internal url +right +The added which is referenced in many places in the RM holds a mapping between application and +This fix has already been comitted to the CP branch by in +This includes some javadoc fixes as well +Ximo I am not sure I understand this fully +Add test to attempt to retrieve column twice and ensure reasonable error +Committed patch to code trunk at revision +This is now clarified +Its a very unscientific analysis but that seems to be the pattern +Ive modified the patch according to your advicePls check and try it +Ive committed the latest patch + +Are you going to fix it +In the case of an infinite loop guess its possible that weve computed an unusually high weight for the basic block containing the spill and that could inflate the spill seem awfully high though. +Agreed deferred to Future +Apart from that theres no problem if the counter overflows youll just get negative ids +If it is fixed in the trunk then I guess you can close this issueBy the way I wouldnt assume that overrides command line options if it does then it is wrong +I ran trunk vs patch perf testThis is w on index but index isnt large enough to useJava command is java server Xmxg Xmsg +On deletion of promotion deleted record from entity +Im pretty sure this patch will fix the problem but I cant test it because I only have the demo shipping data which doesnt exhibit this behaviour +Should I try with a client api call with to try to do that faster? I can do threads createdisabledrop too +Updated patch Check for existing UUIDI have to duplicate some methods to add a new parameter with the UUID +We should expose this via low level implementation and maybe not via IWC +Verified in JBDS Build id +I am also reviewing now +Closed +Patch for hadoop patch with latest branch +How about implementing the interface at the mentioned panel and return false when it is not visible +Committed revision . +Patch to fix the aforementioned issue +commit including in your commit message +Triaged for checking repro attached +The contrib changes are OK for now but in the future If there were a limit on even the number of elements in the cache checking that condition is less expensive than iterating over all the users in the cache for each insertion +I dont think you understand the problemlogj has different log levels for different severities +I also added some more constantsThanks for reporting! +Not really +Speaking of licensing that should probably be cleaned up +If you can do any more footwork on this it would help us make some progress +I dont know + +Instead of moving the method I would like to duplicate the definition in the protocol and only make changes to the Balancer code to use a single nameNode protocol to talk to the namenode +Am I missing something? +Deprecate +I will contact Bruno to ask for his help on this bitThanks EveryoneThis is a small patch for the Docbook documentation tool +Metrics api can be used for thisJust curious how do you intend to use it? Currently scheduler is very tightly coupled with a single cluster +Russel any chance you can try out the patch? See if it solves your problem? I think it is about the best we can do without continuing to push Ivy to have such functionality +Append the patch based on Pinakis suggestion restructure the to extend +Thanks for the patch Vlastimil +maybe it should become a parameter I dont know +There was some nice utility code in the old for dealing with security manager aspects of system properties and TCCL manipulation +There is a hope last change in fix some database transaction issues +This should fix the reported issue so Im resolving the JIRA issue +So I will make its value changeable using std woodstox mechanism as well +You can see my changes in svn and make some comments +This change looks good +AM can verify if a file by name exists and use it +But until then you will have to rebuild axiscpp there are many many fixes in SVN that are not in existing binaries +Indeed FFT implementation will probably evolve in future versions of especially for multidimensional transforms +Should be fixed in to ER +also says no attachments WARN there are no attachments thats the issue +patch applied tested +What is the thrift type in the IDL that the UUID class was being dropped in for? Is it string or binary? I dont know the details of how Cassandra works under the coversIf the type is binary then its related to +Ran the tests it passes +As stated Removed Spring dependency unsing in +This looks fine though I notice some other problems with the table I think some of the rows for the float and double values are mislabeled +Building sorces with x compiler on and with java produced the error see +The consequence is that transaction object is not clean up and lives in memory after ending client requestAd +Committed +Can you provide configuration file used +patch upload is still plain text +fixes problems and adds documentation +h option to the +I can not agree was stuck by a problem and I have to debug the mrunit source to fix it this is not a good idea. +Is to the same as firstReferred? We can check it by checking whether the referenceMap contains the inode id as a keyThen we dont need to add dirMap +Its unexplained in the code its just introduced and its very odd +ear on the end of the name +Just wanted to say that you are awesome and thanks for all your help on this +Edell +The new windows results will be reported in a separate comment +If that happens I can host my own repo make the changes there and then just submit a pull request if that would be simpler +The reframed issue has been filed as +This will break with Spring but if you downgrade it to it in you mind trying out the latest? This is also scheduled for backporting. +Merci if you want I can apply it +Eugene afaik EBR does provide the indexes you mentioned +verified by Vasily +I kind of figured that out once I looked more into the patch and this case +It is possible to have multiple scalar variables in the limit expression this will result in limit having multiple soft predecessors +If you will be able to reproduce it after Freds patch please open another JIRA. +Please apply this patch instead +Hi Myrna that output is unusual it implies to me that get is returning null for most of the info properties files +Schedule for a future version +Confirmed you can now start the server start a target and click on it without this message ever showing up. +Still happened on IBM JDK using the Gump build from will revalidate again today. +The ripple run of your app should still work even without the json file +Implementation is now complete. +This patch restores the missing var declaration in +Please review the pull req +Then I examined each node individually +The latter does not convert to the former so the entire thing blows up +Thankfully solved by Danno manually in groovyConsole and groovysh. +Working as designed the message will be in order if both are allow to get to the queue before the consumer receive +Rajesh Have a look at +Thanks Joe this is fixed in rev along with the Jython class so we remember. +Thanks again for taking the time +Youre right Mark +Maybe we need to include this for the upcoming release that has append support +For example add a field value to every existing document +Please check my latest comment for more details +has been fixed and closed +Move fix version to for release. +Ok ok Ill help +The NN holds less state so that was surprising + need for IT +I personally am not entirely sure that its a bug in +Make sure that the HTTP response code is set on the out message of the exchange after the target has been invoked +Need some time to correct the situation without causing regressions for people that already have kml time template setup in their data to hear that Andrea +Almost obvious as the bug comes from me but I vote also for change +We are currently improving itRegardsHans +But for now you have a workaround and not a bad one at that +Asking the we only can get the list of currently valid sourcepaths so we have also to keep the list of previously valid sourcepaths to compare them. +Oystein askedHowever It would be even better ifone where able to communicate the type of each value dynamically butI were not able to determine how to do this with DRDA +New patch all tests pass +necessary as the Server manager still holds the reference of dead regionserver whereas the cluster is downThis was hanging the test +the url identification we couldkeep as a separate tokenizer for people that want that +code looks good +This way the state of active Jiras is up to date under + +Is it done now +Fixed with revision +seems to show one such d agree +Chris take your time on this +I keep going back and forth on your question +It is best if we give the most descriptive error code but if we feel it takes significant amount of work we can start with a simple solution +All the methods in are synchronized +Added new class to run test from the distribution +Ok I wanted to be sure +Preserve and HADOOPHOME locations for developers +Fantastic Gert +It doesnt need oracle so Ive taken it out +I have committed a fix for you to try outCommitted revision There might still be a gremlin in there but I would like to see how it goes on your XP with SFTP +Enhancements to this will be addressed in a version of Struts after +Putting it into Lucene makes it more likely that it will be addressed as part of this patch and thus committed +There are a couple of examples where it is indeed not prefixed but then it is naturally in the namespace of the element in which it is declared +add fails when there is no component but succeeds without problem and successfully updates when there is an existing authenticationclassic component +One small nit here is not formatted with whitespaces +wont change +I could also put it into and but I do not think that is needed +Thanks for bringing up hbase Dmitriy +Committed to trunk. +This is because there is another baseblock encoded into this block that also changes +settingswhich seems complete to me +Added patch that implements above behavior +Thanks for review Ted SergeyPatch updated for the javadoc warningsbtw +having a more diverse caching infrastructure is certainly something solr lucene could make use of +We need to add these JVM options to and expose the compaction thread priority in the yaml config +This event is more useful where for example some resources are allocated and should be freed anyway +LarsThe current way is fine too +My patch didnt broke anything +Test for reproducing this bug is attached +I see what happened I added the fix for the submenu in the orange bar +Patch looks decent +The fact that redispatch of prefetched messages on subscription close increments the redelivery count is reasonable the spec +Just noticed all the s got messed up with the last batch of the logosPlease use the file names when referring to logos +Try resynching your code base and creating the patch again +Hudson to find out +But if it is easy to support pipes that may retain compatibility +And at the end I will have version of examplepageWell i am looking forward to +Also this behaviour occurs only if one runs binhadoop dfs savenamespace forcesanjaykonstantin +Andreas NPE will go away in +I ran in to something similar building nmr trunk need upgrade the version used in karaf to keep it same as it used in nmr +For example if all values from a block are between and the first patch would require bits whereas the nd one encoding a level above would require bits per value so I think I should rather commit the first patch +I would prefer to just use the REST client classes along with for a couple of reasons So i dont maintain a fork of the entire package +Sometimes there might be a problem int he old sources we have to fix +Early evening ESTOn the questions people can still check out and update from CVS they just cant commit +Im getting some test failures locally but that could be unrelated to the patchJustin Yes go for itcoding with red wine in hand can be error prone +Updated patch incorporating the feedback from Flavio +This patch invokes the lease recovery code from the dfs client +Having just switched to maven for builds its ridiculous that I now have to come up with a set of naming patterns for tests that have never needed them previously + the test case with ASF grant +see TODOs in the file +I have a patch ready but I want to watch MDL for one day before submitting my on a test cluster for h processing System metrics data from more than nodes + +updating components and versions +Does not affect class was not set when the editor +Cannot reproduce this anymore +Also added similar stuffs for the report in r snapshot deployed +I like because these are both single row mutations so its more explicitGet and Scan could also implement or extend something but its more like a Get is a special kind of Scan now so it could just extend Scan? That wouldnt necessarily help for cases like this where we want to pass out either type to the same thing but Im not sure its needed for reads +My guess is that we should only include the source including the build and instructions for installing and not even package jars at all since we cant include the LGPL ones necessary for Carrot +And sorry about the Clojure sample code +Will adding Inherited to annotation suffice? Thanks remember to make testcases for thislook at in the source for instructions +The branch is more or less in maintenance mode +reattached patch +The indention of the patch seems not the same as the requirement +I moved setting the leader to to the remove API +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +Hi RobinThings are coming along +this removes row locks which are no longer necessary to ensure atomic can you explain more about removal of row locks? seems like your patch just touches the simple get case that takes a row lock +This is because test server doesnt shutdown properly after running we can resolve this overwriting maven surefire in json +I will debug this a little to see what is index key being used for ie m or m +Hence the for pointing me in the right direction! +Russ was right not setting the ivarsWaiting variable back to false caused the issue +Also see issue +Ill make it more robust +But thats very low priority only if it is easy to implement +BTW youve forgotten to change the function name in unix export list +User node in the Express console view is created before the credentials are actually verified +Addressing second set of review comments +Hi this issue still a problem or can I resolve with a wont fix +Koen I know youre off this project but do you have any suggestions +It takes care of key referring to multiple files also +I did a very extensive audit of argument processing for and Im pretty sure its about as minimal as I can get it with the current design. +Please modify it such that it reproduces your issue under Tomcat and well then try to reproduce it on our side and track down the error +Inside that code we access conf three times once to get the local file system so I dont see that as being an issue +As far as Im concerned the only way the average wouldb e better than NA is if it took the number of lines of code into what you really want is a scriptable aggrator that could take the results of other aggrators and perform simple math on them +Added Apache license text into in dir and conformant text into +How about Cassandra +svn diff of changes need to correct the in Rev +It should be fixed now +Added comments to the +Moving to EAP we do now. +yes Martin the technical foundation for the p update site changed between Beta and CR it was supposed to happen between Beta and Beta but life got in the way So yes we believe it will work from CR and going forward +Im not very used to JIRA +Thanks for your answer +If I push an invalid cid a is thrown as expected +Ive asked the Servlet spec leads to review it and provide their comments +Patch ready to go +Currently one of the least obvious points of integration is how to get data both onto and back off of an HDFSDistcp is the best tool today for this +For this we need to make the rename optional and by default disabled +Hi Francois I set up a env on top of tree and wasnt able to reproduce this +Thank you for the investigation +The only suggestion I have is that rather than always making the hostport the first argument to the fence script it would be better to use a substitution token +xml +Verified on. +Came to the opinion that it might be desirable in devMode but when we are in production mode we should consider it a DEBUG information therefore logging level should be lowered accordinglyApplied the said change to +But I think this nutch feature would be useful anyway +Nice catch Pat +I will add another jira for this feature to be used by pipes +Anyway your commit is a safe one +I would never never discover this on my own +Sorry to say that but after recompiling grails the problem recompiled with git pull ant clean ant ant installed binary releaseon the app side grails clean grails upgrade grails you must have an older cached version somewhere because your application works fine for me +Thanks guys! +I havent noticed this for a long time. +I also didnt find an answer yet on how to include other files inside the jar +Attached wsdl you dont mind me attaching another example for this bugThe attached zip contains a test wsdl calling with the options p and nsp tns and all the files created by wjStub Skeleton and are placed in srcdefooservice like I pleased wj to doThe schema related classes you find in srcorgexampletest wich corresponds to the default namespace package mapping not to the package I asked for +increasing scope +One comment to my patchI could not find constants for adfjsfLibs Configuration seems to be outdated +This looks like its a network issue +The patch checks if the method is called in thread +Makes sense +Coincidentally I remembered about this and checked it last night then proceeded to forget to close this sorry about that +I will fix it on commit Ram +Patch to Default Cayenne logger. +New patch adds test and checks values of awaitTermination +However I keep it assigned to me for now as Id like to test this feature on projects like or the Symphony framework +I think this is mainly due to wiquery putting the dialogs content outside the form breaking the form uploadprocessing logicIMHO you might have better luck handling the modal dialog outside the main form +Its not intended to be a replacement for tpstats cfhistograms et al +Seems these changes are not needed at all as yarn tests are not supposed to be run in parallel +Hopefully it is easy to set those defaults +Mr +And what about throwing an exception on errorThats a different discussion please open a new jira issue +Can you submit the change as a patch +surprising its not implemented yet. +Okay looking a bit closer I think I know what the issue is here +Just keep us in the loop folks this is the first migration in the queue +Did you create the distro after running the maven build? The is generated by the maven build +This work is now complete +in slingcore while fixing this issueCan you please verify with an up to date core bundle? Thanks +Hi AdrianThank a lot for your explanation but can you describe me more? How the notify CCM and what CCM do when notify an inner transactional? reports are not for you want help use the forums as I said above. +Will a separate ant target be ok for this? Or does it need to be part of the target and be generated automatically with the binary jars +Thanks for the comment Vlado though this doesnt sound like a documentation issue but a component bug +Im not sure that we want to try to reproduce this ourselves as wed really like to get Cassandra back up and running +The patch attached to will resolve the original issue mentioned of Queue configuration missing child queue names for root +There are some advantages to using Katta but what we found is that this patch is very NOT ready for use +Theyre in seam GA +attached +Doesnt have this problem once switched to JPA Hibernate. +test fails due to? thanks +Thats for the service which connects channels +This is a sketch of what I had in mind +Attached which is created against latest trunk code +ZK integration is post +The test should finish in both cases but the number of puts written should be quite different +I just noticed that Netbeans had introduced a number of property changes that ended up in the v patch +Have added a testcase to check this +Fixed on and trunk +I have not received or build ids to test for maintenance which is currently a priority +Thank you Hari +Because theres more than work remaining for the closures Im marking this bug resolved for and opening a new one for that additional work. +It also adds the jdbcapi test to the derbynetclientmats test suite +So they have probably solved the problem differently +Is this still a valid issue? A comment on the ICU bug suggests that we may have an invalid versionVladimir Can you please test this again using the latest revision? Thanks +Thanks Sean +Jeff thanks for reviewing it +SimonIs this a blockerthanksdims +Are you trying to deploy the same app twice at the same time or are you attempting to do a redeploy of the app +needs which resides in the codegen module +Thanks curt in our environment it cause every notification fail in a period time for example in minutes all notification are ignored for they all late than right time Its normal once a notification is ignored further notification will be ignored in a big possibilityOur model is there is a master juddi that is used to register and there are many local juddi those are used to be lookup so the process is the services will register in master juddi first and then will be distributed to local juddi by notification +Access time updates should turn off by default +Since this is just the documentation and the +Closing as we initially thought this one reappeared but actually it is slightly different +Committed at subversion revision +Fix that will not be +Thanks Harsh for opening this issue +I have committed it to the latest development code +Heres my latest patch +Fil Its to showhide the splashscreen to get rid of the white flash of the at startup +wicketstatelessidasdfgh but this way the URL for all stateless pages gets uglier +The translators have handed back message localizations for the message changes in and +It would be easy to modify at one place instead of +Call it Add path +I hope that I successfully sent the PR. +There seems to be a bug in where the last word of a non common word query is removed +from the db sequence from the sequence from the second database +If needed I can attach the log file generated from the failed runsIn terms of build it should also have the race condition +This issue appears in deb package as well +moving to in preparation of the rc cut +Which of course doesnt sound very efficient! + +I just found myself thinking that we should introduce an interface for the compiler to use as Id like to get rid of the Ant Javac task dependency as soon as possible +Odd I was having some renames rewrite paths to nest into the target dir +I find those suggestions good Rob any of these we can do easily to remove the worst waittimessurprises for GA or rather wait +I just committed this +Do you think this could explain the lost messages youve been seeing +Just skip this one +will simply be done using waitFor on so it should work the same way on Windows +my mistake on the transport +What happens if a login failsThis is actually described on the wiki page A is thrown +My login name is an digit numbers that my LDAP username +Works well Evgeny could you just change the violation message of this rule Insert an if statement at the beginning of this loop to filter +Ive disabled hte debug option when server is in RSE mode +Committed! Thanks Namit! +Committed at subversion revision Touches the following filesM javaengineorgapachederbyimplsqlcompileAdds a method for adding dependencies onM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileWires that method into nodes which views rely onM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangBasic tests for dependencies of views on +So just trying to install the loader repository wont work as the xmbean invoker has dependencies on our providing info like the correct class loader +Also there will not be any replication as there is no live replicasThanksUma +orWhat kind of response do you want to be returned when there has been a failure? I will check if I can find the other related tickets on this issue +This is dynamite +I am closing this for now +Committed patch in revision and renamed classes in revision +Both issues +for your patch +not really a blocker as all API changes are backwards compatible +See the patch to fix this problem +This is a synchronization issue +The right solution is to fix is especially critical since a major focus of the EAP is on backwards compatibility +The checkout was just else is running on this runningmvn false clean install to avoid the clirr plugin failing the deleted everythign in my repositoryI installed relative jee dependenciesI out your tree againmvn clean succeededmvn install failsand still fails with U flagattached buildplease that workednothing like a melt down +My experience is that it is better If the appliction uses delete rather than depend on the behaviour of the shutdown hook +see resolved issues to so we can generate release notes +HiI have created patches for the same +Ross you are right The Maven Plugin provides this feature by parsing the XML result +Updating patch as Teds commentThanks for review again +Calling +Could one of your team review this patchThank you that should be adapted because this test case is used to make sure that invalid attribute node have been be removed +I should throw an exception if a mismatch occuredAnd the new patch added in two more testcases +We could also booleanize the various methods to return a failpass state if expected and an exception for other assorted issues during the code flow +Update for features requirements to WTP +I guess if you had a very disjoint set of values for a given enumeration doing a binary search might be kind of nice +Well get another RC out soon so its a good way to m definitely looking forward to hearing from Nick and Marco +This deals with authorization not authorization and authentication my code wasnt in my machine +We can leave this open for a bit and see if someone else chimes in +The fix looks fine. +The findbugs warnings and the failed tests are not related +This is only a script that youre protecting +I removed the export statement because it is not needed +Ill need to rework the replication fix +The credentials were being verified during the very first call on the admin api +I think this is part of a larger issues +This was because this syntax was not yet supported in MVEL +Thanks Nigel! +adds the need for to be told of NM failuredecommission as causes for container completion +Thanks +if you intend to add the directory later +orgapachehadoophbaseclient +With this path test passed +Both are really a hackI guess the questions is what should the end state be +If so please close the JIRA +Ive committed it to SVN. +Proposing to +In this case the company does not want price fields to appear for the raw materials of the components only for the manufactured components themselves +Ive committed this patch +Also I need to implement the lock interface in writelockcomments are welcome +since a managed transaction even with standard EJB CMT almost always surrounds the entire lifespan of a Session without obvious you can quite easily have a Session opened right after or even right before a JTA transaction starts with Springs plus and a Spring +But couldnt get any reply for the same +So message should be generated outside of and in case of warning this is not trivialSo solution to show directories included into instead of warning is less and simpler from a technical point of view +Ill contact the UIMA project for an update on this issue. +Running it now +Follow the link try to update this project with no changes +Resolved as we have now upgraded to +Thanks for review Ted and Andy. +What do the jnlp files changes do? file change introduces a new property +Are you sure theres no previous error for this app in before the null classloader error? If not there may be an error getting swallowed somewhere +This was fixed in +sigh +The Im using see attachment is slightly altered from the version Xerces ships to use a parser initialization sequence that more closely tracks the way Xalan is invoking the parser +Eventually Id like to depend on and get dependencies that way but that does not happen at this timeSimply set some paths and use the tasks +The first column is the number of nodes +Hi Marcelwe think the problem is in or in libraries +Its in the svn nowWhen documentation is in place we can resolve this ticket +Change to API. +I will work on this now but I noticed a workaround load your rules file before your flow file into the builder +Im on this all i except one are in i are in now. +Please reassign yourself if you are still interested in pursuing the issue +For the super pathological case you turned up above the will mean we only do two files at a time compressing way but at least we wouldnt be stuck +Closing as fixed in separate issue created for trunk +What package are you building? There are quite a few in the repository +Definitely agree that doing this still makes sense +Fixed by properly returning null as expected by the class +Have not committed yet though +Getting the hang of the audit and agenda views is important as it shows you have a strong understanding of how the engine works and ability to debug the engine +wget relies upon a challenge before supplying authentication so wget will get redirected unless you use the syntax is explained in the reference manual +So you have to search every directory to find a given block +However there are still SQL errors which I suppose should not be here see +Can not reproduce anymore. +Hey JunpingHow about we create a feature branch for and commit the patches there? Seems better to process feedback on the overall design before pushing on individual changesThanks is not really a big feature +v Looks good to me + +I just committed this +I know this might be a bit too much detail for now but we should track bloom filter capabilities of the underlying engine +Also for the community release we have an image in this error page to show the users the process to add a user this image was added as this error page is quite likely to be the first page a user sees when accessing the console if they have not the user so the image shows how to add the user +As Janko says this is easily reproduceable by doing a GET request on a forms action URL +thanks +Existing integration tests updated +It removes all the fancy scanner counting we used doIn testing serverside logs seem fine when outstanding scanner +From buildsHelloBuilding project on Jenkins we must wait a lot of time to have aslot availableLooking at Jenkins GUI it shows the following messageNote currently only builds on ubuntu Builds on ubuntu fail withUncaught Exceptionhomehudsontoolsjavajrelibixawt cannot open shared object file No such file or directorySo it seems only machine out of is workingHope you can do something about itThanks installed across all build slaves +Is it ok +updated with trunk +has changed a lot and some opensource tools have not got ready to it for example Twitter Storm support Kafka in its currentlyI think we should make flume sourcesink available for kafka and then start get kafka ready +There are always admins who disregard these warningsI think they deserve what they get +I am on GMT and missed the message +Is this still an issue +Do you have some time to make a servlet that does the same thing as throw it in the war and invoke on that? If that fails its a more significant issue since as you say resource injection in JSP pages is not supported +The fix in is essentially the same change as that in my patch to this issue +Im going to tentatively resolve as since I agree with Sebastians guess that the real underlying issue is not pruning enough cooccurrence +Also please consider uploading logs and snapshots separately instead of posting them on the description of the jira just so that we dont get long descriptions with every update to the jira +Tested +Attaching updated versions of and +Actually its not about additional information +I was going to recommend we take this to the devaccumulo list +Is this issue fixed now Lorenz? The patch was applied at this for +I am actively working on this +This will be in the next update. +From adminsIve enabled htaccess override for the maven directory +If the original issue is the same as mine the problem is actually that jBPM creates new classloader instances each time +Thanks! +Sorry I mean the problem was with Eclipse + will be also inherited from close for +Additionally Im using the snapshot from and let me know if you need further details +HTTPS NIO is not within the Axis project anymore +Latest patch looks good +However accessing httplocalhostGforbidden without accessing httplocalhostGsecure does not show error or challenge for idpwd or load anything +Deployment +Please refrain from using Blocker +By implementing Serializable and setting the embedded key in the value via the Id annotation works +Doesnt actual now so I havent commit it +This should be documented in the javadoc that any specified for an ATTACH will be be ignored in favour of those specified in the first CREATEWhat is the usecase? Why is this necessary if there are filters on the serverstartDelivery should throw an exception if messageFilter or messageHandler is nullOtherwise the patch looks good +This is not productive code that advances the applications requirements if you know what I you +bulk fixing the version info for and all affected issues have in +Replacement httpblocks containing correct type declkarations +I think as the next phase we should consider about moving part of the job to cloud based on the increasing computing resource requirements +I randomly submitted a name above but it would be better if there was a form to submit a name so we dont have to eventually read through comments to find the different names and the comments are strictly comments and not the options themselves +Attachment has been added with description Patch for Layout Mgr to reduce IPD for padding has been added with description PDF with implemented AND IPD for patches we need the diff u option with its wonderful s and s +Manually tests were updated. +q tests +As long as nobody votes this down Ill put in the convenience methods +So I guess displaying the bind address as well as the server name in the title is the best option right now as it helps distinguish between instances running on the same patch for Branch +The repro shows that the bug occurs when an extra column is set by the UPDATE statement +Same as issue +diff file containing the package structure and a simple version to start with +PYTHONPATH was being assigned out of the scope of the python scriptalso I still think ant test should be dependent onalso there is another problem the tests wont work with ant test being called from the srccontrib directory +This seems to overlap with new contextroutebox components +The client could read an incorrect version in this case +Trivial patch should fix this +Thereafter for the remaining life of the classloader has will return false +This documentation has not been updated and a CR is required +Sorry Claus completely forgot about this +Letting me know who is going to admin the zone and having that person place a copy of their +Basically if the references it and they do not understand it they just crap themselves +I think there are further cases +states this is fixed in but there is no explanation how it was fixed +and are already available subclasses of +Anyone using these suddenly has a millstone on their foot without noticing + for the to trunk. +This bug has been marked as a duplicate of +Would it be possible to make a version that doesnt change but instead just catches the useless Exception? is just a bug fix so Id rather not change anything unnecessary +Closing RC issues. +The scale module had README that was used for index page +Attaching is a combined patch for and as both are dependent on each other +An equivalent thing needs ot happen by default for faultsI will see if I can fix this since this seems to be a simple fix but I need to see where this processing happens +Updated patch +Tree and treeTable not working properly +My bad +What do you suggest instead? do proper justice to the new capabilities of we will probably have to rewrite some of the transport logic in +now released +Ok no probs +I dont think we should do it. +Somethings up and we should not just write it off at this being a flapperYes I agree thats why I clarified my comment with an edit +If you start favoring a word that occurs times heavily over words that occur times I think it works in the favor of bad suggestions +Please post your request for support to the struts user mailing listThe error described is a problem +Any chance someone can reproduce +Done +Please disregard the comment ve ran the test by the means of Ant build and it hands after minutes +Thanks +The tool is in hdfs package therefore it is redundant to refer to fs again +Bec assigning this to you as part of your rewrite I believe I have completed everything this JIRA required but you may want to revise the content +I wrote a test and did some digging in the code adn found that the returnstatement is quite correct as castor loads configurations today +thanks thanks Daniel +Verified on JBT to remove the newandnoteworthy tag so this issue doesnt appear in the release new and noteworthy label +You need to supply anexample if you want use to look at it or profile theredeployment yourself +Closing the issue. +Is it worth pursuing bringing them back to life +ScottMy suggestion is based on recent discussions about the widget models +Thanks for catching this Vijay +Problem was in the database configuration +I am mother tongue German speaker +Im working on Kryten update to latest common headers to patch this problem +There is very little to gain in going to fetch a sequence value times in a row vs fetching a sequence value only when you need it +Fixes said problems +Thanks for reviewing it +LCDS provides support for Data Management capabilities in addition to the RPC Services and Messaging capabilities of +To see that changesetgit p da svn diff c r +The new test executes many SQL statements but it doesnt check the results from any of them +Yeah I agree with Robert +I would have expected this to have been resolved with +the patch does not solve a synchronization problem at the TT +For Eclipse and Tools And its even not a Dynamic Web ProjectThats weird since Tools maven intagration plugin should have added the JSF facetnature to the projectBut you can do it yourself +Many thanks to all of you. +I did not apply to TRUNK +See fisheye for changes +if the node is a comment or processing instruction instead of a text +Also some in inAll these references to Features and Properties should also be removed +The jobtracker would chgrp the logs to that group +I believe it doesI did put some debugging statements in calculate in before and that is when I realized that the problem is in the way endpoints are storedSo here is my usecaseI started up a new cluster no data has been inserted yet +m not sure that I like Wicket plays like a mashup and combines resources from all over the globeThe idea of CDN is to deliver commonly used libs from the fastest mirror for the user +I like this approach than single server wide property +Keith I just skimmed through the patch and I think I need to look at this patch very closely before I comment on itI would also want to test the GSSAPI mechanism to ensure it works as beforeI am tied up this week could you please give me time till next week +QEC works in the prepare phase thus I believe it should do the right thing downstream regardless of its just replace the base query rather than adding onto it +Verified. +Yes I think so +Neither repair or index rebuild helpsThis is on cass Ubuntu LTS u +I think this is a minor issue +As argued +I looked through the code changes and they look good to me +Patch applied at +I wrap the eventList in a thread and add the items a few at a time on a background thread +Will endeavour to make some time during the week to update my setup and try this out again +It turned out the actual cause is throwopt optimization +This is a duplicate of +This should fix the issue +srcqlsrctestqueriesclientpositiveOne gotcha is the comments can only go before queries and not before setadd file +Eventually this will exhaust available memory leading to an OOM +is related but not quite the same +The test case would read the web content and check the conversion was printed +Patch adds documentation about the distcp action +You should use to show suggestions list in modal at +when master start to assign rootmeta region it should finish split hlogs +Test webapp to demonstrate the problem +Patch applied at +I expect we wont do it this way and will want to store the call convention in the same way we store the result of the ffilib call +Last time we used at least different Jira issues +Latest instructions are in attached file +Id pretend we could bump BRANCH too but afaik theres no need +This is obviously causing the dialogue to +Looking back to the closed worls years ago where we would follow the worlds of Borland now we are at the frontier using the internet +After that try to create another +By the way is there a change to fix this issue in or versions +Test results comparing and using defaults +new patch I added random to updateDocument and randomly enable disable docValues entirely on optimize commit getReader so we get segments that dont have docValues at all etc +takes care of putting the two together if subcolumn filtering needs to be done based on the querypath that it knows +Please provide test project that will be sufficient to reproduce this issue +I guess we could add some wits +MarkCould you please provide us with an example of your report and DDL for creating tables +I will not change the temporary attribute +This was never true in the previous environments that Ive worked on batch +Would you mind sharing with us the final sqoop command that your script generates +done. +I was able to spin up a Windows R x VM with JDK x and run the tests and they all pass without tests arent by change running in parallel or something thats the only reason I can think of how there could be another Message getting dispatched to the client since the test case itself only places one Message on the Queue +Our service endpoint is bundled as a +fixed. +Some of the letters are not being filled +Having to run very clean takes a lot of bandwidth and a long time +Review Apply +I have found also a small bug +No I think we leave as it is otherwise it may end up messing with AS . +Mahadev has applied fix for I think only is needed for branch at the moment +Ill try to blame this one on the new JIRA +Assume closed as resolved and released +Ill address this in the timeframe of . +Thanks Lahiru for the patch +Set Fix version to +Hi TsingFirst patch is for another one is for trunk +Proposed patch +Good quick win JulienSuccessfully tested. +Download the suite and replace serverserverjbpmdeploy with deploy +In weve now totally disabled tracing unless you specify debug so its a perfect time to add tracing logic into the generatedcompiled code +The patch for patch without checkSession. +eg +It appears that priority is of reduction of proliferation +Perhaps this is better assigned to someone with a better sense of what is common across the tools? Also I dont feel comfortable refactoring others code so maybe this should be split into issues one per component such that everyone refactors their own base to pull out common testsfunctionalityclasseslibsstringsUI pieces +In a complete run of the regression suite there were a handful of failures +Thanks! +I had an exception EVEN THOUGH I had a dataSoiurce present in my +Guilin I opened as a for +Similar to a hadoop or hbase model of master and slaves where you could add and remove search servers change out indexes etc + +This improvement should be in release version of as soon as possible +priliminary patch regarding comment I think we can leave the conversion as is since we wont have int number of in a tasktracker +Hoss the latest patch fixes the issue I noted +This looks like a dirty hack to me +To follow on Edwards comment I dont understand why beeline is in the patch +Samuel i am now in attending a meeting +Bulk closing stale resolved issues +Turns out I did have a relative checkout that was stale +If optimized is true skip all the regular expressions string replacements etc + +Ill probably submit the whole thing at once so I dont create any interim failures + is get from hbmjava +Im looking into why it needs a forked lifecycle +If you require these fixed at the branch level then please contact sales for subscription based support. +Fixed we use the field name conditioncode all ResolvedFixed issues lacking a Fixed Version + +for the patch update NamitAlso Ashutosh and Namit again for all the reviews +Closed upon release of to. +Patch to resolve this issue +It holds up fine seemingly +The former patch puts the into default instead of into the allblocks profileNot having enough karma to remove the existing patch for the patch this should be fixed now! +I have committed this +Apache Directory Studio has now been releasedLets clean Jira and close this issue. +Thats a good idea + +Justin I think this one was dealt with already +Seems like clients should normally be +I guess we can use there but Id rather wait for blocks +I even tried building in offline mode and its still pulling them downOnce these other changes are pushed and built out to Nexus Ill test the releas module +The cannot be shared between multiple threads as you got a shared countdown latch +The linked pull request fails with dynamic panels as it results in a nested call to set +The patch works for me thanks EgorAll concurrent tests pass ok +At the end of processing pig will attemp to remove the temp files +I will update the patch have updated the patch that applies cleanly to r +another vote for to using Node Ids as me see if i can get done today +So I would lean towards option +Could you please help to verify it fulfills your requirement? Thanks +Closing resolved issues for already released versions +Patch committed to trunk +Will look at this when we work on the next generation of the editor. +WillemI am on an engagement this week +Committed to trunk. +If it is needed for branches you can upload an optional patch for that branch +Closing all resolved tickets from or older +it looks good to me +After we store dead gossip states for days so that any other nodes that were down at the time of removal can know later not to repopulate the ring with the removed node but this isnt persisted anywhere so since you did a full ring restart the only candidate left is the persisted endpoints though all nodes should have removed it from there after the decommissionremovetokenIs there a way I can get a list of endpoints to see how this node showed back upAlso any thoughts on why this node only on a single nodeThanks! from the node which the dead node back to the ringThis is from after the decommission of the phantom node but before the restart which resulted in the node the phantom nodeThe phantom nodes token was +Karthik you were cgists mentor for his internship +Ran derbylang and all lang junit tests +Maybe it could just return the redirect value +For example if machine A is in domain instead of just A you may need to specify +Verified on revision . +Are you sure that you are supposed to run this test under the Derby clientThe dropcrash test is one of the Store tests which is never run standalone +Here is the patch for mentioned featureThanks to Rishi Awdesh for their support and Mridul and Sumit for discussing processRegards Ratnesh screen shots of resultant screensRegards Ratnesh the correct oneRegards Ratnesh patch has been commited at r should we not close +Attached modified patch +Attach backport for junit and junit so it will print out test name as we run test on is the output log file and database for mailjdbc after running full day of mailjdbc for embedded server with version with Sun JDK +Yes I can now reproduce the issue and can also see where it is coming from +The code is not correct +Ran gridmix to check performance impact +Probably best to split it into core and mapreduce piecesAgreed Tom +Thanks for explaining the problem +If nobody object I will commit in a view days +This was attempting to add the security module code to the complete jar +only being able to replicate parts of the database for which the user is authorized +Please check that it was applied as expected +waiting for steven to update the jira +it might even work for +These are all done manually +Im not sure if that would really matter for cqlsh but just figured Id mention itMaybe we could still have the option of setting whatever version we want but then have cql and cql shortcuts that would refer to the more recent known version for instance +see the unit test +Patch for replacing Avalon Logger with think Rana has a larger patch to eliminate Avalon due soon +Thanks HariIm closing this ticketThanks Yongkun for your contribution! do agree with you Mike +Thanks for the review +Technically su is a command we run after we are done with SSH and has nothing to do with the transport +stackcan you look back to this new feature again? I think its very useful either for hot deployment of Filter or dynamic coprocessor +I confirmed on the latest trunk build and filed. + an old patch from my git repo +It is probably a good idea to retain Block and not change to Dir +See issues resolved in released versions +I have absolutely no idea how to do that +I am just wondering why you want static access to these values +Please confirm that this fully resolves your issue +This is a similar issue to and some of the ideas proposed within are close to our desired resolution but they have not yet solved this problem +Job waits for all tasks and taskAttempts to finish not just killed +And of course they will probably not I have to say I never used any of those archetypes +A quick test in my application showed that this indeed fixes the problem +I will also be reviewing this work today and in particular considering how it may be leveraged for a REST that I have for validating incoming JWT access tokens +Ill check it and file the new issue if I find a bug of +From what I have read IE and Firefox have a cookie limit per domain Opera has and Webkit supports enough cookies that if you put in too many web servers start to error out +I believe the best way we could solve this is by using the first is it possible to implement the solution check the state in the execution context? It would be just perfect +hifirstly let me say that i am by no means an oracle expert by any stretch +Thanks Junping for reviewing the patch +With my code adjusted to deal with this correctly the problems with mime types set too late go away. +I am getting the same error +We could up the heap for the parallel maven but if its OOMEing because cant create native thread thats more about too many threads running +t key WFM on OS X +Mickael if we are only on we will still generate tons of install errors which is the problemOk +A possible answer could it have to do with the persistence of properties of and the way its being updated +Changes in bugfix release +Once can be made from the content ofrootpom The xxx group names are internally inconsistent and require a The deploy module referenced from the top pom is not present +The existing unit tests all pass with this but need to add tests that deploy an app on one cluster node add some sessions and then deploy the app on another node thus causing the state transfer +Thanks! Ive committed to branch trunk +If resolved before is released please assign to +Applied the patch +Nope this modification is supposed to be resolve classes whose are different from those Tapestry is able to resolve +Modified methods to add a file to classpath and cache +This is complete now at the pull request location +Complicated subtle or wrong increment in +Hence the test was successfulIf all Level tests on a Level repository throw the TCK would finally declare the Repository Level compliant which is not the caseFailing Level test cases of a Level repository do not have an effect on its Level compliance +linking JanYour patch does not apply clean against the trunkI think the file names include part of your pathPlease check and resubmit Ill try again +Well I cant see anything wrong with your account try attaching something here +question Is it possible for transactions to be operating at the same time on a deferrable constraint and in one case the checking is deferred and in the nd case the checking is immediate +its a little tricky to have latency tests as a part of unit testing +Wont the compilers default classpath be good enough now that were requiring a Java or higher compiler? Thanks +the patch is built for the and only display the replicated tables on master cluster +Here is the patch with a test case +Thanks Omkar +Ill add general retry on server start and stop actions as well see my last comment on +TODO +This is closely related to the artifact attachement bug adding the cliOnly feature or similar will account for that we should prevent infinite loops which was the other part of this bug +Jun and Victoris there really no better way to do this than adding a massive case statement to handle each request type? +For example if you have a jar in your bundle and the deployment environment also provides one you could have two different classes loaded into the runtime and we would not be able to cast from one to the otherRequest for a testcase or more information unanswered for more than a month +Im downloading the dataset but it would be good to at least document what the test is +microcontainer +Updated patch with changes for just +This error occurs constantly when you do some action within and verified +Hmm +Now I need to check what the so called Cryptolog does that confuse the pdfbox +bulk defer to away from IMCMight still be worth doing in the context of native array but will open new issue for that ifwhen it arises +Thanks Ashutosh I can take care of it +Cheers +nope +Need to be verified +hsa Why dont you apply yourself +The implementation in derby is alright +As mentioned above it would be a nice to have +jdo and +Patch intended to supplement +sits entirely obviousits not entirely obvious +Both the stack trace and the process definition to the ESB action the standard scheduler service in use here? It would not fire a timer before the previous transaction commits saves the variable value and the timer record +Ive checked a fix into trunk +Mental note should provide a method with the same functionality than avoid the use of inside the I think the new method could take a as a parameter +This method did not work correctly and was removed +Should this use +There has been some talk about logging configuration on almost exactly one year ago that you could check out +Order of start is A B B I think might be there is in step Bs GMS will think A is the coordinator and try to send JOIN msgs to A which doesnt specific scenario could be recoverable +BTW all that is already explained in the link I gave you This saids its exactly what I expect and it will help me to quicly review and hopefully commit +Try different source locations for the replication +IgorCan you run your rails app with true and see if you see any error message during startup +pages created with +Committed +If the environment variable is available its value will be picked up overriding the property value in another property location +The latest release of our production is based on and we will not switch to AS for this release. +Hello Everybodyhere is the patch that fixes the issueprevious code used to append calenders div at the end of the body tag which was a cause for the layered lookup to closethis patch modifies the code such that now the calenders div tag is appended just below the input field it is been called fromthere are some minor css changes tooThanks to Atul Vani for helping me in developing this patch +In worst case I will use our local build of hudsdon containing this No let me know on issue and I will proceed accordingly +duplicate of +The release is scheduled for tomorrow evening it would be great if you had a chance to test this before the release! grabbed and dropped in the +Will do probably today +I can find out whatever you need though +Michal what was the outcome of this? Seems like were going with so Infinispan would need to upgraded to REST right +Hibernate still hasnt gotten to this + +You mean in backend also +Now might be the best time to stabilize the svn source urlfor the site currently its in sitebranchescms +With the core reviews for each pull request this is happening gradually even if nobody is referencing this ticket +Im giving this back to dont have bandwitdh for this +this much better +Thats the key. +The problem is that every time if we are loading the for a nodes cluster it will put some extra load on the which we obviously want to avoid it +You are right this would be a static configuration but the client in fact would not expect the servers behavior to change at runtimeI attached a patch for this issueRegardsMartin +Looks like it probably is +patch looks good +If it is still an issue pPlease reopen and provide test project that would allow us to reproduce this. +excellent +This is just a snippet of my code that uses to connect ssl secured smtp it helps +Also I left default font that seems to be fix +Thanks +See next exception for detailsCaused by Exception from Cryptography provider +The pattern validator will need be enhanced to support multiple this valid XML +I wonder what is happening with spring modules +If error continues to happen past the threshold then report an error instead of a warnLet me know what you think +you realise that Hibernate is using SQL right +This makes it easier for a user to swap in a new app server he has to edit only one file which controls the app server classpath and doesnt have to search through other irrelevant targets +Configuration property values should match in and +The problem occurs with all gems files +Moving to did you have this in +If youre going to discuss this on irc can you ping me? Adding a comment to this bug would do Ill pop over ifwhen the mail comes in +This could be checked byAnother approach might be to add counters for and local task placements and io +Due to the popularity of this issue we will be making a release sometime next week +The patch has been tested for synclogs issue +Thank you! +Looks pretty good +Ah thanks Kevin I wouldnt mind betting thats it +Fixed on the svn trunk +This should be done in the data source viewer in Kakadu +Thus the eldest process is the system start time based upon configuration +Applied to trunk +I also verified the same issue is present in and +You may use this as a starting point +Unfortunately you cannot customise the name of this column as yet +please dont cache classes +I cant reproduce iam using revision +It would throw an error if you try to write a really large key +If you can think of a way to simulate packet loss in a unit test let me know but otherwise I dont think its going to be possible to come up with a good unit test for this +a littler demo project to regenerate themvn changes can confirm this bug +You should update to at least and try again is quite old +Future work on truly compiled stylesheets is being done in the xsltc module. +Please if it still is in is not supported +Verified by Mark +If the GACs no option then maybe a version wildcard is +This results in more simplified functionality for the in that only one version can be specified +Will do +For now theres a doc writer per thread +As an optimization we should convert such kind of joins where other side is an output of a function always into a +I wouldnt want to make it harder on the translators or the community to contributemerge messages +Also the final images to be uploaded in the HTML +File which generates infinite which generates the cant replicate this with the supplied file so it may be JVM specific +Ill fix that +I noticed while looking at the classworlds config files that the statement which would load all jars found in +Ill stop working on the temporary destination advisory flow problem this morning to see if this test is working +There should be a location section that states where the samples because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +We should be testing this against PBR on AS +Are you sure this is done? I see no commits +The noconfig switch is required for distributions +Verified fixed using expressions in args for messages in and add within a JPF +After clicking no in a couple of Javascript error windows I finally get the homepage +To me this seems like a case where DRY is coming into conflict with least surprise +I think that this must be DB related because i dont get this error for any other db +Can anybody attach some test case to this issue? Otherwise Ill have to close it. +Thanks +Can you be more specific what version you are requesting +Now that we have dropped support for Java maybe it would make sense to make Lucene use the JDK impl of ntz? replace with replace with replace the pop methods with simple loops using All lucene solr tests passed except some zk tests whose failure seem unrelated +Will file a separate bug with logs +pushed upstream and in JBDS Build id +thank you for your time +For disable logging you may change the values set in +On the other hand since it is just one api object maybe the cost of a JSON parser jar and with the other apis isnt worth it +Submit the applied in r +Not required after mavenization. +Please feel free to check it in with your changesThanksRamin +Patch updated with trunkPatch also adds JIP and TIP update statuses under lock +Eugene Want to provide a patch +Looks good to me seems to be building Marc please reopen if its not +Im worried about taking it in this direction though for reasons of future compatibility +srcPath is useful so that can possibly use different policies for different files +Rather than client name Id recommend a workflow ID that can be propagated across multiple Oozie pig hive queries helping you track it down not to a user but to a specific jobThis would aid not just auditing of client use but cost auditing of entire believe modifying the RPC layer and expecting clients to somehow reliably propagate a client id will be extremely challenging +Logic is returned +no reports about issues punting to since VJET build not a release nor tested +Another approach would be to wait in the responder until both the downstream datanode responded and the sync has finished +Adam can you double check this with your scenario? + +Thanks for the suggestion Andrew I overlooked the policy file option +If we treat this as a documentation issue then we need to update the documentation of the so the documentation of the index option warns against switching it on when using java jar +Supplemental patch committed +But the part still concerns me as it is changing default behaviour +This is a generically useful feature as well independent of views +thats an old you try a recent version please +bq +Committed patch to resolve as closed. +I do not expect the JBM would internally open and close TCP connection per message instead of per session +Fixed as part of +Patch has been comiited +Not sure about affects version +If you observed an error did it get corrected? What Im saying is that if you use language objects then quoting is taken care of +trying to run hudson again +Im working with Connolly to determine the best way to convince them to move +Have you had any luck in your production environment finding what the difference is? Things to consider are you crawling using a similar proxy setup in both environments? is the web connection identical in both environments? same seed list in both? are there other jobs running at the same time against the same servers on the same instance? Do you see any differences in the interaction with SolrObviously the goal should be to make the test environment mimic the server environment as exactly as possible +is a patch with fix for this issue +done in cadcabcfcfddcbdd +I can put up another patch if you want +But adding the features in the patch to the filter actually means that this use is endorsed by the community and Im not sure thats a good idea +Joe +A compromise between deleting entirely and bringing back the vmapconfig agree with your analysis +Please assign it to the person who is in charge of that code +Thanks Carsten +Nice feature! +No this doesnt have anything at all to do with the XML configuration +Looks greatCould you also update the organization issueManagement ciManagement and scm tags in tajoI think that this patch is a great change to update them +The Tree Expand test can be used to run this manually with Firefox making sure that you can tab to the expandcollapse images of the tree nodes as well as the selection links +Patch as David! Im applying it think the patch was missing a few things which Ive fixed modified the code in the same manner for jetty and and verified they were still working fine with and modified the m build to use the latest version of jetty and verified it works with still need to update the class with latest dependencies for Jetty and +My weak vote would be to leave it as is +MohammadThis is great +Ondrej I think you are missing the point +So if we wrap these two policies how to know which exchange must be removed from the inflight repository +It has been created from the project root I had no idea rake tasks could take arguments like that +groovy and adding a resolver +Thanks Raghu. +working on it putting status to InChecking back if this is done to close itthanks it just nowmaster feaeebdfabdebdadcad fbfdbabdfcdcabececlosing the the documentation needs to addfix content on apidiscovery tabular output etcAssigning to myself once I addfix the content on wiki Ill assign it to Sebastien to take look and help addfix the content on cloudmonkey in the apidocs +I disabled session loadsave weeks ago but everyone is still on beta +The mapping between Source tab and Outline is much more useful and works both ways +Not critical for because we build zip archive with sources anywayMoved to just had a quick look and it seems goodIn case something is wrong or missing this will be a JiraI removed target since I judge its too much riskeffort for the gain it produces +I am pretty new to and havent explored deeply into the API +Issue was resolved on another was resolved with. +Like this +refactored patch containing the suggestions from tobi and felix +as well +project Peter +Ive found the submitted patch +redhat script from usrsbin Renamed package from to Renamed dependency from hadoop to just committed this +Please check the plugin page for the compatibility matrix +BNF of EJB QL from EJB specification of persistence part given at the section will try to first separate out the QL which can be easily implemented and later we will work for full support +Thanks Harsh! +Yup that was the fix. +bug +patch committed +Could not reproduce the issue with will if the problem reappears in rel +I have much angst +Replaced the patch with the proposed fix with a slightly improved one that doesnt change the semantics of the added in trunk in revision with trivial adjustments to the patch +In Rev + +Problem DELETE this issue or CLOSE is a bug in my JSP +Fine by me +The name of the property is of course open to debate +Handling of has been moved to the last catch block. +Patch with separate options for evictDecay and warmDecay +I will move it to the Spring JIRA +I think we are moving to guava +These tests are going to change when the new integration code is added +Gav +For now not +Thanks for the bug report Daniel +If for some reason you need customized values for the noresize attribute subclass the tag or modify the source and build a custom version. +I havent been able to reproduce M inserts followed by another M inserts allowed minor compactions forced major compaction when done w nd M wipe restart K inserts major compact K inserts major compact +Could we define a conversion or coercion from Array to Array that would take care of this case +Ive written an adapter that uses the derby xa datasource +I made a mistake +Will a comment with more details later +Ignore then +We could do this by looking into the jar and if there is a or such then we could remove the jar from the path +Wondering Alex if you had anything to say before I do? Otherwise will commit in next day or so +Mustve missed this +patch file for easier apply +I have not followed this but some time ago Ive had a problem with cloning and seeSearching jira for gives hits currentlyRegardsPaul I committed this fix for nowFor the buffer is allocated lazily so the extra clone doesnt use much space +I attached a screenshot example of a virtual procedure that has become nearly impossible to debug due to this defect +This patch uses read locks for all three methods +I think Jackrabbit could well do something similar +Can you elaborate pleaseThank you I see you mean should not be specific or tied to Java because Java requires a surrogate pair and not just one value. +It worked fine thenPlease take file +Issue is Callout Mediator default client repo path is not available while running the integration test suiteIn the attached patch I have altered the sample configuration to point to the integration testrepo client repoModified sample configuration is placed inside the integration module itselfPlease review and apply the patch if there is no other better solutions +Can you review and close this issue? +Thanks +But this does come into play w commitlog retention +Id rather expect to have aseparate Tags Documentation section right on the first page +If it does turn out that Solr is the right place for these dependencies it doesnt require Jetty to compile so at a minimum I would argue that the default compile scope youve used is inappropriate +One other point that needs to be discussed is line lengthwrapping +Now that work is well underway with AS all previous community releases are +I dont think the above failures because of your patch N +Curently we can Reload a region Clear caches for the serverShould we be doing that fix in a separate JIRA or this JIRAThanks can expose clearing the caches for a region should not be much different from how forcing reload of a region is now got why this exception happens in your scenario +Shouldnt a concept have a URI and preflabel rather than a when its about adding +I checked the lock mechanism and it looks fine +Same patch just the MR part will do the common change in a separate jira +The general process works but I couldnt get attachments and it broke about halfway through the list of issues +bin rmr hbase From the latest comment seems like the command has been renamed in future releases though +I will commit sometime between today and monday if there are no objections +All the pressing work is doneso I am moving this forward so it can continue to catch ongoing cleanup efforts +I think that this would be good incremental improvement +With embjopr r performing the operations noted in the description is now File updated +I will follow the conversation of this bug report +This set of changes seems to fix the problems on my laptop while just running the single test alone +Great workAs the release is just about to get cut I would like to schedule this patch for +By looking at Tika and see value in metadataI think Solr sould set it in metadata if the charset value is presented by user +Let me massage it a little and commit tomorrow +Whats the open topic on this issue? Which classes have the problem +Fixed r +Invalid since authorization mechanisms changed to be rule status to invalid +It seems to me like this is not a problem +I am marking this jira as a wont fixThanks for your comment Jarcec! +Thanks +No cigar Nexus still does not auth me +Is the patch considered high risk? Would it be appropriate to mark this newcomer if someone wants to try out Armys patch run tests and drive this issue to resolution +I was working on a patch but got pulled into other things +Anyone else? to fine +This request hangs here for days and is even not assigned to anyone +Should that be spun off as a separate issue +bulk move of all unresolved issues from to +Fred is this functionality even relevant now when in Kepler will have this +Im going to mark the bug as solvedThanks Luke +in this case all hot regions will move to this new one +Thanks again for the contribution! +Closing +Also has a lot of overhead this approach reads through stale versions of data that havent been compacted away yet and reads RF replicas of each row +Earlier I tried on a single machine and it workedDid you restart the cluster running the patched code? That may be requiredDid it fail with the same error +The SCA Assembly spec says that a composite MUST have a target declaredThe spec does not say that the runtime should raise an error in the case of the target being absent but that would be my expectationDoes anyone think that the spec is wrong in this area? The problem is a general one for XML artifacts if they dont belong to a namespace how can you senssibly search for them +heterogenous security clustersservices supporting servers which indirectly allows HA to use IP failover etcGiven all the discussions involving more radical changes to the security framework Im very keen to providing the modularity required to implement these systems but in a manner that will not destabilize the existing security implementation else Yahoos deployments may be delayed +So is the plan then to run some benchmarks with the patch enabled and disabled and post the results +Move the fix to +I believe that the underlying cause of this problem may be the same as +The exception doesnt always occur even with use of same data + test close if verified fixed. +This patch has been manually verified +I dont see how this could happen +It would practically be treated as new feature then +Fixed by +If you can provide the Serialization logic for the LB endpoint as well that would be great +Implementing both classes is not an issue but make me afraid to be used somewhere else that I will miss changing at future in any version upgrades +Sounds like a good plan to me +What to do with this issue? Left it open forever or close it +Add a at the end of the patch file in order for it to be applied verified that the patch fixes the problem and also fixes the second part of the rio gem problem reported in this patch and with patch from rio gem seems to be working for me +latest one should be patch has been applied +Theonly difference between the two files is the order of the first two elementsinside the Model element +fix to core pushed upstream +For html we have the for pdf the +we are currently just committing the stub code generated one way but that is obviously not idealIm going to resolve this as Wont Fix but its really a Cannot Fix +Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +The wikipedia article covers several different algorithms one of which is similar to what we have +Ive created it because I thought it was missing +Now user cant change her password or login +The slide out palette looks silly when two maps are open side by side switching to the is preferable in this caseThe Palette is updated to the current tool when the editor changes focus. +Changes in to executes the compatibility tests like Mike described here +Just run ant deploy for springaop only way around this appears to be replacing the cglib jars with the unsigned versions it does not look like cglib functions correctly when also appears to be an issue for the EAP platform see Toms comments in the ESB issue. +v fixed the crash +Im sure its possible but the users mapper or configure method code would need to be written in a way which allows the mock objects to be injected in the right places and would require a lot of test fixture setup code that I doubt anyone would want to write +It moves the major section on The Service after the major section on Distributed Caching with Cache +No worries +introduced bug will be fixed in +Reviewing them might be a good idea +You mean from the master and the regionservers? Or just from Stargate +But it looks like it is backwards compatible and a fairly small feature at that so if you really need it for I think it should be OK but I do want to have more eyes look at it and it would be nice to get some test results on a secure cluster before putting it in +Defer all unresolved issues from to +That should be fixed now +bardtodoIf this hole in the type system stays we have to add it to the spec in the section +Fix long lines +Fixed all regex statements to handle both uppercase and lower for MandatoryOptional +Guvnor version isAttached repo contains rule which demonstrates how the DRL is being generated in rather odd fashion +Everyone +The changes should be reviewed with this submitted patch and the best options selected from both +Start level changes removed in revision a quick and dirty retry if the bundle cant be started the first time its tried +I think that this is a very useful feature for people who want to write documentation but not html jsp or anything else +This issue may have something to do with +Typically this user would already be part of the group configured in the ACL +Not sure why it failed to apply +the patch for Hudson results +Then I prefer towait for approvation of the gurus before posting another attachWhen will be separated the andMy proposal is Separate the Helpers +Verified in ER +He will be committing the patch once he is unit test issue is really straightforward +Gopinathan go for it +AFAIK is only used to set the pages culture on the new thread in case such a thread switch occurs +relying on schedule job and preventing updates to seam model while validator runs sounds like a bad thing since the validators tend to run for a long time users wouldnt be able to do much else than wait for the validators to stop would they +And what do you get with the patch +Resolving as invalid. +I need also consider the case where combiner is usedThanks new patch deals with the case where combiner is invoked +Ah never mind it seems the hibernate config module fails the build regardless of my patch its just that the nightly build does not run tests so it does not get noticed +which then counts for all loggers +The failed test is not related to the patch +Canceling patch so that Harshs comments can be addressed +Any idea when itll be ready to go as open source? Im currently trying to figure out how to deal with our hadoop logs +sorry +perhaps rename the name of the hint to somethign like NAMESPACECONTEXT +I do confirm +After adding check replication policy its conf cannot be null in construct stage so here we remove the default constructor +I see it resolves now +Seam doesnt use the XSD files at runtime +Attaching +delete doesnt seem to actually delete hints now +Lucene has and Output for doing block and encodings +Not sure what the best solution there is +As an additional information requires an instance of in one of its this is part of the automatically generated data on demand test code I have no idea where to begin debugging +What he means is that if you set on the server but some other ack mode on the client then the client setting overrides the server +implementation of drawing borders around elements was used the same as for osx and open jdk +In SVN +I think maybe some wrong with nutchs jsp +iof the request is s POST or if the request was a GET +The modification is when the mixin node types are modified Instead of first removing all mixin node types and then readding some now only node types which are not specified are actually removed other mixin node types remain +Is that a problem that need to be solved too +committed +This should be moved over to mr stuff but for now can we leave it so it builds the mr jar until that happens +So I did not expect to see element class product but I do expect to see type class that you mention it should I expect to see both element class product and type class Product even if I chose type generation +This appears to be a duplicate of which has a patch with a test. +Looks great to me +This is the same problem that in the slice filter was shared among multiple which is not correct in the current state of things +Offers Manage keys that allows to manage the docs with new key management features and instance launching options. +Juergen where you at on this one buddyI have found at least articles about this very issue and Spring needs an official implementation to address it IMO +I suspect proper Big handling as in that other PR is the right thing to do to support these types properly. +Should I remove attachements and reattachPlease attach new version of trunk patch +In this case wed like to filter annotations by the tag values +On Seam both of the two situations works fine +Attachment has been added with description Patch to provide static attribute in svg branch +I think this patch is broken +Images not can I say It is a tough Job but someone has to do it +Ah makes sense Ryan +I dont have any way to confirm since I dont have the point rotation stuff but the issue upon which this depends is fixed in +Im going to commit this if therere no to closed. +Application with embedded Jenkins successfully created even on staging env so closing +We have some fancy jdk searching feature that added extra sourcing of which is culprit +Jake is the class weve been referring to in the above posts and it was around +Martin Olegs patch does contain + +Vivek I believe this was a bug in Weld integration? Please confirm. +Is there +Alexey please verify +They had a simultaneous update and select which were accessing different rows but still they saw this problem occur +Matt I merged the changes from branch to +It all needs to tie together +See manual. +Fixed in transition resolvedfixed bugs to closedfixed +Seems Lars is having such usecase +Thanks for the problem report! Fixed with IT in r +Reopening and marking for +This situation works fine if isnt packed into +changed from hours to hours +This is a draft version of implementation +committed to trunk and. +I have not ignored +by Eugene SchavaThis patch breaks wwaction tag within sitemeshthis seems to work now correctly +The unit test used to pass +committed. +So all issues not directly assigned to an AS release are being closed +It doesnt appear again +The changes have been pretty extensive in the meantime wondering with the gt gs getting into stable state do we still need to do a backport +Fix ll need to rebase this once all of the changes due to the second round of protobufing goes in +Add flag that blocks compactions +Hi Aidan can you take a look at this commit please +It might be easier to detect and delete just the required columns at the write path +When is global false? That is how do we end up w a resultset containing data from more than one CF + updated +Great news! Thank you +With the fix the two testcases take mins +Ahmed Thanks for the patch +Its from Arthur van Hoff +if you reaaaally need all that then move it to helper method that both the test and ui uses so things are not intermingled please +I did some experiments and it looks like they dispatch correctly to the different concrete handlers when copied inPlease give it a try or suggest otherwise. +This is not a valid name for a database +All tests passed committed as svn Resolving +I will commit a fix for it now assuming it doesnt break anything +thanks +Did you get a chacne to take a look at it? did you find something +You can then use this layout for JAVAHOME +As agreed with Jorge he will have to fix this manually. +Can you please file a jira? Better yet a patch +I like your explanatory text +Im looking for better layout mechanisms +Adding a srctestgroovy folder to the test project fixed everythingI think thats a bug in the code here +Also users of Camel would have to upgrade to Quartz to use the quartz competent which is not necessarily something they want to do if they use Quartz in other parts of their application +Just to make sure with my fix you would also have remained backward compatible +Hi Danushka SteveThanks for the reply but I need to get some dlls to test as soon as possible +Thanks Pat! +What I applied to +bq +Compare that code to yours +Also consider using a provided scope for would be a really cool feature to to follow +In the other rule it stops but I cant step through at all it just think there is some dodgy caching going on +Attached a test case +Unfortunately I can no longer use your patch as path is not available anymoreI no longer use the VM hashCode for comparison thanks toI believe the ordering is now predictableLet me know if its not the case. +I applied a slightly modified version of your patch +Patch contains the following +Maybe in the last months someone updated the labels +Whith the current finder query generation you only get record +Thanks for applying the patch so quickly. +fixed in + +If so why the method method is as fast as beta according to the same report tps with beta against tps with rev perfomance increased from tps to news what did you do to get those results +because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. + resolved issues for already released versions +Updated patch with small performance improvements in. +I think that perhaps it didnt work for me because I had an int as listKey in the formtest example auth is a String +Thanks Harsh +Since almost the entire patch in should be committed to I will instead of fixing here +Luke I just committed the new files +Thanks Kay Kay the patch applies cleanly and all tests pass +but what else changes +one possible idea to keep backward compatibility is detecting exception and build old version package +patch committed into svni am fortunate that i received the patch but still confused that how to install this help as i am new to maven and cargo +I will double check if things work ok with that in +I dont think that the is affectedMaybe the service is not affected directly but consider a path somepathHere the would check the extension levelpath which is not correct +Throwing instead should be fine. +I think should get the block length from not from disk because guarantees that crc file is also updated when updating the replicas bytes +Im working with the MVEL guy +this time with all updates from Tims ZIP file +I do have ANTHOME and JAVAHOME set on the server +Hi Alfonso this sounds like areasonable feature request however there are some issues as well and i therefore propose the followingHadoop the upgrade should be addressed +Work in progress +Leonardo I dont see a discussion about this on the list +This is dhrubas rebased on trunk +It fails in +Patch that allows persisting the memstoreTS as Vlong +and thanks for applying the patches +See log +Changing description to closer match the matter of this issue +We could also link to it from the documentation +When correlated the planner leaves the projection in place above the access to DB and does not consider pushing the order byThe fix for could be expanded quite easily to handle this case as well +Rick Im assigning this to you since you are working on the biggest part of this effort at the moment +Ill fix them +Avoiding bad slaves seems to be a better idea. +Sorry about that +The original problem with node has been fixed +Then we dont get any interned strings or timer threads but we keep all the nice stats codeMetrics is apache licensed so I think Im OK to just copy classes over +use getschema rather than getfields +Separating the updating task time from the client submission time is a good idea regardless +bulk close of all resolved issues prior to release. +The check if +Just committed a first version +Sad that it wont be fixed +Surprised we do not it +If current pages url is the same as its before doesnt do page this has been can not reproduce this issue +A more complete fix to the port rolling problems is needed. +From Balaji in an exchangebq +Sure +Not very pluginable ? webapp some handy archiva specific utils creates a war filedesign contains a rsync scripts to sync repositories +I had resolved it by adding it in the +Thanks Anurag for the contribution committed at trunk rAshish Vijaywargiya +In response I removed the test +roy can you clarify the intention of segment parameters +Committed +upgrade version to in rev +the edit log toleration default is also disabled +This is caused because the error thrown by the parser does not contain enough parser contains a fromDescr which should have a IDE will then use this lineNumber and also create a marker +I see you beat me to patching the harness with the pathOf +Also added few more checks in test for secondary namenode +The console shows the following information in the header Product name and product version Release codename and release version as fallback +All zebra nightly tests pass. +Can you attach the project? Is it big? What type of project is it? Thanks +Did I miss some +so maybe were all set with this one +Tested with +This is a heavily altered file for one transaction +Russell none of your tests are included in patch +It looks very strange +This could possibly be related to hang itself was caused by an assert in a callback in the test +Matthias this boils down to filtering within resource files but lets first talk through this + +I spent a bit of time to understand that the reason +committed to trunk +verified atVersion id for set correct fix change for resolved issues that pointed to instaed of concrete the ones that were closed previously for +Tests were passing but have since been broken by +It includes the following changes +bulk defer of issues to +Fix still has not hit the streets +Thanks +Uwe I will checkcomment under your issue there shouldnt be too many issues not much was deprecated in contribI will commit this one soon +It seems this issue should allow us to test that without the wait in a way that we can be fairly confident +I do mine also and some kind people around help me too +Are you able to connect to that are created using the web interface? +Also can you please confirm whether this is a regression +I recommend you make it explicit which graceful request this is referring to +Now that has been committed it should be quite to have the remaining bits and pieces implemented +Can you provide the actual? That would give moreinformation +As you know by now weve been doing some major refactoring behind the Aggregator during the milestones +Ok got it +still an issue but not fixable for critical +upgraded to Thrift released issues. +I appreciate if JIT gurus review the patch +looks good +I am sorry Doug +Im working on sample application to test my solution on and also I will download and test it is setting up on the Struts zone get in touch with him +But a process could be defined +for scmperforce the does not know about changing the delimiter +updated patch incorporating Namits comments +Add unit tests trunk patch forthcoming +Please review my change +Released. +Thats all right I see the point of doing this and I remember that I added it at some point so it should be easy to weed outbq +Thanks for the quick responseBesti in SVN trunk and branch +This bug means thatin a projectone must mention schemas in ones JPA annotations when doing identity generation XML overriding will not work +why is something trying to sysread a negative number of bytes +Thanks +Bulk closing stale resolved issues +Of course having reported the issue I get a breakthroughIt seems only GIF images are supported and PNG images are rejected +Once you have an error in the algorithm Jitrino will failSo I propose to close this bug as fixed and discuss this issue with community by emails before opening a new RFE +bq +Patch adds a return statement after printing the message about skipping jar signing so that signing will actually not happen +released closing. +Thanks! +Ive gone ahead and committed this to SVN on the trunk branch +Subsuming by incorporating it +BillI dont think introduced any problem as it was fairly trivial +Closed after release. +You could obtain avalid FO file by not propagating id attributes from your XML file to the FOfile +Thanks for the patch +yes I think we should separate forked arguments so they can be properly these settings specifically will make it easier to use rather than needing to know X +See hg logchangeset eebuser DeveloperSvante SchubertReviewerweihuaWangdate Thu Feb summary bug Allow load save of documents with attribute values that may be string OR primitive was integratedchangeset aauser DeveloperDevinsvanteschubertReviewersvanteschubertDevindate Mon Jul summary Color data type improvements and usage in Table Cellclosing issue. +In which case is that uniqueness scoped to the database or is it universal? This affects my first point +Need to do more investigations +Thank you +This issue takes the change a step farther +Im still seeing it +In fact does not support a debug property So you caught mistakes +Thats a good point +Your explanation is right +Does that solve the issue +All the rest looks like its the sameOtherwise we end up having to have them in sync and that will never work +I can help with this one if you want +Committed. +Combine the new test case with existing ones +Ill still listen in +Thanks Laura +An class generated for this element +verified in Developer Studio +There should probably be a note in the guide that mentions that addon uninstall will not work properly on Windows when you have multiple Roo shells open as they will keep the files locked +The domain was live hosted by Oracle until recently +Added name support for +Fixed in r +Thanks Ramkrishna +Its now yours since youre looking into it +Hibernate has absorbed the annotations component +The Lucene indexer bug has already been found to be a consequence of this bug +The contents of the represent the actual configuration of the cache +I found out that some Linux distros use lib for librariesso would live in usrlocalinclude +If extends the base implementation type from sca then it should be fine +Since the key design challenge here is the interfaces I think we should coordinate our efforts to make this happen for both HDFS and +The same test case configured for Thanks for the assessment +Final patch for review +When adding the repository that I mentioned in my April nd comment were you able to reproduce this problem + as long as the test fail is not related +Thanks Ben +You are right +But since it is so clients can use it in unit tests this is a side effect rather the intended purpose +I updated the example and also added deprecation comments to the classes that are no longer supported +Patch adding another field to holding an array of instances that the caller can concatenateback together to make the highlighted string +In or tests they were setting and now are setting the correct its probably the correct change but worth being aware of that were changing behavior of those tests +Scheduling on the devel path +Its mostly complete but needs to see if there are gaps in instance management and allow the ability to fill any holes +They all fail because a process cant start due to a socket bind problem a runaway RM process on the build machine perhaps? I ran the RM unit tests locally with this patch and they all passI also manually tested the patch with a cluster running sleep and wordcount jobs +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +The problem here is that to avoid calculating a correct relative path from the desired destination of the drp to wherever the parent is the code was written to write it out temporarily right in the basedir where collisions are possibleThe fix Ive chosen is to calculate the relative path +Marking closed. +We probably want to stub mlock for too but I feel like we should leave at least one test calling mlock +Ill take a look and see whether we can fix this in or whether it needs to be pushed out to may not be a solution at all though +Jozef what to do with this issue +As a temporary workaround Ive implemented custom and +Ah i see now +I think after the kernel and life cycle refactor we definitely should this functionality as David pointed to make it more +Does anyone have got an idea how to make it cleanThe output of the memory checks is in appropriate reports +Indeed Remove again at Revision fix is not in CR and will not be in CR for in CR build. +Issue is closed. +extends Column +Since Konstantin asked for is he can confirm if its ok +can you provide a patch to add proper site updates for your new goal +Great + +I suggest creating a that does a on exceptionMinor point to match the rest of the code use spaces for indentation not tabs +bumping to since we shouldnt mess with tool apis during stable cant be empty on create but nodetool just generates a timestamp one if you are using it +Attaching an initial patch to unblock agree Daryn +Im gonna resolve the linked ISPN jira now so that it can be incorporated in which should be released later today +Feel free to update the wiki +They use that have safe iterators although java docs claim that the iterator itself can be accessed on thread +Since weve done this a number of times for other apps it shouldnt be too challenging +Further tests and careful review of the effect of XA on ac evaluation still missingPlease review +is updated from as follows Fixes a bug of RECORDs ARRAY field at Cleanup of some code +Dont know if thats helping you but Ive the same problem with tomcat and eclipseIts OK with wicket and below and doesnt work with wicket and above +Can we return the exit status code when the application doesnt existWell return exit code as is fine +Do you want to train the model based on Tiger Corpus with that File +Youre right Ill apply your patch which is more complete +cuSascha +So all issues not directly assigned to an AS release are being closed +I just think that if the version of is going to require version of the exported commons logging packages then there needs to be an apache release of commons logging that has been bundlefied to export of commons logging OR http client needs to make commons logging optional in its osgi flavor +Verified with JBDS L +Ive just committed this +We also played around with a global term cache so if the same term is returned by different index segments the same String object is used for the +This is a less pressing issue these days since you can pass a size to the local dir allocator and itll do the right thingDo people think this is still worth fixing? It should be straightforward to do something in for writes + +Lowering the priority to Major and targeting to fix in +The patch is missing. +Yes I tend to agree I will try to check that +I do realize that this is the only quickstart that compilesruns groovy scripts outside of the ESB runtime but I do think this is now intuitive enough for newbies to use +Weve been using the v of this patch in production for a while now and it seems to perform quite well +As Pete stated separating the EJB services from the component model is already on the agenda for the EJB EG +If you want documentation of eg a specific release then you can check the manual +And i turns to all right after i follow qilless instruction +Also you could really help me out by breaking this apart into smaller chunks I can reviewcommit fast in an iterative manner +Patch causes some problems when updating security information in the file but assume this is a work in process and will be corrected with a later patch +What I also meant but didnt say was that it should accept a Ruby object upon which to set the Ruby now setvalue breaks when a Ruby object is passed as the object param which is why you have the code in there +Seeing the problem in I think revision broke the fix +Sent by email and attached here +is a tech preview in +From a recent post not only me but other ppl expect the D could be put at the end + +This patch applies the MSVC variable and works properly on my Windows box using Visual studio +We would still need case by case examples of how to +I should be able to try this out sometime next week +it passes the tests +Re other issues wouldnt it make sense to do first before you improve the +The ACF docs specifically say does not modify delimiters in the list +By design +Sending srcmainjavaorgapachexbeanspringcontextvcTransmitting file dataCommitted revision . +I know the problem its because i have in my +I am on this for one reason +at r +I was thinking exactly that its not necessarily a good practice to hide the actual root cause but didnt run tests elsewhere than core itself +It looks to me like the cached requires synchronization +Applied the patch into trunk and branch +If we try to make Comparable fixes here then we will loose what hadoop could do for us out of the box +Tested with the erb script shown above +Here is a quickstart with simple behavior that implements this functionalityPlease try it and give feedback whether it helps for your use case +attached patch for nice and uncontroversial +This is probably a rejection +is using a function from standard taglib jar +I think the discussion to use phabricatorreview boardpatch should be done on the dev mailing list instead of this jira +The improvement is committed by Sven and so far there are no problems +Give us days +created a workspace called Foo +Does this still make sense? We should probably ensure our tooling is consistent with AS as much as anything +Note that this issue is blocking publishing the asf zk site +Sushanth You have two patches up one via phabricator another directly on jira and they have quite bit of differences among them +Initially this happened during upgrade from to and we spend a lot of time trying to figure out why we cannot access migrated tables +I think it would be great to have this issue resolved and Id be willing to work with you to add tests in case you want to pursue this +if clause added to avoid possible are speaking here about test code and I do not agree that returning null instead of throwing a NPE improves the situationI searched the test code for invocations of the method in question and at many places the returned cas was accessed just afterit was created to get a reference to its Type System +This is a bugChanges will be available in new snapshot version +Import type for me is important to know if the import is optional or not +Closing as per Paulexs request. +Assigned to Lewis PFA the updated patches +Adding a new patch to the bug named since the old patch has gotten out of syncsvn stat has remain unchanged since the last patch and the changes are also the same as the last patch +Alternatively consider your own aggregation function +Add warning about WSRP issuePortlet rendering over WSRP may randomly fails ending up with the following error that can be found in the log FATAL Parent is nullA refresh of the page would temporarly fix the issue +Logged In YES useridThe current CVS Head and CVS Branch do not show this Username schaefera +Bulk closing issues resolved in JEXL. +Just removing the package wouldnt solve the problem though because then you wouldnt have a provider of +JS now uses injected rootScope. +The issue comes from the symbol table builder parsing alone takes much less than a second +According to the JSF spec only those jar files which contains a in its folder will be scanned +do not store server passwords in plain if we are only on we will still generate tons of install errors which is the problem +fixed in r +My comments are mostly the same ones that I had for the previous patchesI dont really like having the resource comparator class configuration be specific to the scheduler +Hi Marco +install does not create the metadata required for Maven to find the you using deploy +Andrey Denis I am still able to reproduce the screenshot that looks very similar to +Assigning to Shelly to coordinate +I confirm that installing a bundle that uses +Waiting the artifact to be released +Nice thats a much better approach +Reuploaded again with unsigned jars from +Going to write a new ticket to download the jars. +Actually there is no real reason why a different thread should wait to process a completely unrelated type +updated with trunk and added a unit test +May be we should add a comment that target should not be called directly +The problem is that setting the property in the jobRepository does not work in all think changing the class as I have done is not the correct way to do this +users of openshift will want to be able to choose which config they run with +most important method calls all got migrated over to the new shipped in +Putting a newline at the end of that was accidentally omitted +contains the documentation changes for this a patch file created with updated svn +When build docs were updated this issue disappeared. +This is only draft now +xstreamjournal files are just read in and deserialized whereas +Meldware is not supported for use with Enterprise Application Platform and all references to it should be ignored +Doids is a robot framework so I think that the classes you mentioned should be moved but not in core project +a database import or export +The one in the release uses an logo already +That would be a blocking issue for any release in which it occurs +Duplicate of +Hi VincentIve reviewed your previous post and updated the docs and the staging +The deep method is useless and never used even by the unit tests +Changing the description to reflect the change +Its running as part of the normal test suite now +I did have felix configured from previous tests +Other opinionsThanks change seems a good way forward +This is part of the strength of the implementation that the Phonetic filter doesnt leverageAn additional change that I made that could be done in the Phonetic filter is blocking empty tokens that are generated on strings +I have verified +Those that did have been implemented in the meantime and tracked by different issuesSo this issue is to be closed. +Good eye! Im sure this is true +What about validators having parametersShould they be grouped per type or per type param valuesIt seems grouping doesnt really make sense except for m on this probably + unit tests do not cover this because they use the legacy data directory format and the legacy loader correctly escapes the data directory URL preventing the failure +While set of new values check if deprecated keys are present if yes then simply set deprecated valuesWe would like to go ahead with option +The bug could bereproduced with using the pattern matchfoo in any template +I switched back and forth between Properties view and Documentation view for several nodes on the canvas and things worked as expected +Good points see if this suits +I needed to add a new option STATICPROPERTIES and add a boolean to a couple method sigatures +Here is my latest patch and a very simple testcaseLinks on making good testcases for me or other suggestions as always welcome +Im moving to a future release as to be investigate +Removed parallelism sectionLets add later +Thanks RegisPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +Showcase doesnt have so tried with maintenance on IE which also worked +Good stuff from me too. +this +Attaching new version rebased onto current trunk +double are fixed +Your regexp looks quite logical +Can a committer please commit this? It looks like there are no objections to the latest patch +If so do you plan to address the test failure mentioned Apr Eventually Id like to turn it on by default but I was mostly interested in making replication work firstSo I took a look at test and it was a simple matter of changing the reader to use the one that provides +Something creeped in my last patch v fixes that and adds the missing javadoc +Duplicate of +If you remove the kerning elements from the font metric XML the problem should go awayGruss aus Luzern +Verified the RN text update. +Tomorrow Im going to reduce the dataset to an adjacency list and write a bsp parser for thatIve decided to use as key and value value is a semicolon seperated list of hosts +That should make them available while not confusing folks who have no idea what an MD is +Okay Ive made a new move sh and a patch which you can apply afterwardsThis was the worst work Ive ever made +Sent Roman an email re the CLA formality +or did something change +I will continue providing patches under this issue after which we can start looking and documenting clear code conventions + +I will verify and file a separate bug for those +Close down as these are moving to Attic statusMerge into the users list as well +SolutionThere are two solutions for this +For now Ill put it back to unknown. +Its getting an NPE when trying to call +Hi SureshSure We can use Apache SIS in the project for the following purposes +Now running the tooling from cmd line has been made easy +Yes this should definitely be supported +Then Ill send itback to you for testing and well work collaboratively +Attached the trivial patch which renames the as Node Address and Host as Node HTTP address +This bug has been fixed in release +Will try and take a look what it would take when I have time +Youll have another round of patches by this Friday Ill also move to eating my own dogfood on this and ensure this ends up as a part of our staging service before the next submission +Fails before the patch and passes with. +We could either add all the dependencies right now or add as needed +A release is a release +bq +Hi Adrian so if I have understood correctly now its ok can be now review andor commit so I can complete this enhanced about Google Base and so I can start the new enhanced about the new eBay upcoming applicationThanksafter a very quick review your last patch seems goodYou should soon have all the commit privileges to commit it so Im assigning it to you +another code path question during the repro do you back up the database +The weakhashmap is fully synchronized now + test Works for me test unrelated +I think this is resolved by the NPE fix and removing hibernate from your EARlib folder +patch +Dont know what they are waiting for +Changing to feature request and targeting to Westport +Sorry for the bit about M object graph resulting in M of memory use during serialization +Ideally when creating a new connection I would just update the entries on a static instance of the environment +Thanks MarkFixed in BEANS module at repo revision Please check that this fully resolves your problem +what is the undesired behavior being exhibited +cool i downt know why but from console test the methode in dont tests from IDE both the methode in and in hintsAs workaround i move the initialize of the from setup method into every test reproduce here +I cant see any way for the connection to be added twice I think its more likely that the virtualHost is null which I guess is possible if unlikely +fixed in trunk and branch +Per my previous comment +Ive attached the latest revision of docs for review this is now very nearly complete with just the character encoding guide to be added +Fixed with unit test in rWe now support propertiespicUrl as well as for the email address + +Likewise +Without knowing the code this may be difficult but it gets what you want +AFAIU normally Java class bound with JAXB annotations will have to define two different properties for each bar element in the choice. +Per user context is still used to cache hconnections +Sanne where are we at on this one? is it commited +The other document analysis components like UIMA should be able to identify parts of documents and then you use this on one part at a time +For now it only exposes Lifecycle methods +Thanks Amar! +The old code matches the problem that Knut identified and the new code addresses the issues and does what is described in the previous comment +Ill take a look at the Sorter today +As I said I designed it specifically to solve that +I see them all in resources are statically coded against the JDK API they are not dynamically generated based on the whatever the JMX mbeans exposeThey theoretically could be but not in the next weeks. +Went ahead and implemented a patch +Also another very scary thing +Thanks will do that later on todayCheers the memory as it started to shoot up +Thanks. +Well not really because I still want to have coverage information for the whole project +We rely on it to be configured in +the search filter is case sensitive and there is no way to turn it off or on +Removing the osgi subsystem from the DCs the deployment works as expected +the Data Injector service is moved off from code base +Will close the issue after the build is succesful in Hudson. +I wont be able to test the fix since I dont see the problem so can you test it after I check it in +Formatted Object files are created at same time as html file +thanks +Component to prevent double submission would be better description for think this should be covered with simple support +Xpath localNames may contains numeric startingcharacter to allow tree creation and than may apply ISO conversionthanks for paying attention to my requestFilippo +. +Update Sources action in uses maven settings to configure source folders and output locations +Moving to +should be pretty safe now +James isnt going to be the only one encountering these issuesBen Ill assign to you can you document this on the ASF wiki then close? Thanks +it is a PNG +Resolved +java source +I remember we used to send the one liner output when socket connection is established and Chukwa Agent Controller doesnt handle the output correctly +Only that they are defined in a class or a method +all issues assigned to +Removed temp workaround. +to using least in my current case isnt a secured resource +an indirect option +Marking normal urgency as it seems to impact a test suite running against Derby +Post your question in the forum then we will answer in the forum +Hi Joseph when you attached the patch you did not click on the option that states you grant the ASF a ASL license on your patch +Im going to check in the basic directory structure and libs and then provide a patch with the source that we can iterate on +I prefer but stated it poorly +Smoke test fails with gcgen and passed with gccc on Linux +Thanks Jitendra and Jacob. +Verified by Denis. +Yup it is now possible to control this behaviour for wfs output +Why would a fox property be preferable? I think it would actually be harmful as it would change the semantics of the FO element it is used on +I think there would be an argument to put the example code in a namespace too but for the present it is left. +Add binary files required to run the webpage Subho I committed it now +against version version +Why not first delete the empty dst directory and then rename the src directory? Or moving the children from src to dst +tested on linux and os works for all weblogic versions +Its easier to view +When this scenario happens it should now the login BehaviorIt will redirect to the login screen ask for credentials and once credentials are provided the user should then be logged into the with and also with CAS and JOSSO +we need to cleanup the persistent stats and take a look at what the snap do +I wonder how many times we are going to see this bug coming back +Thanks! +There is no patched +The attached patch uses scope provided for all the dependencies which basically makes them +Trunkjob +Added to the list of affected versions and estimated effort. +No errors and I get the following tag specific reports createdTag library validation Tag library validation reportTag reference Tag reference documentationTaglibdoc documentation Taglibdoc documentationThe former has a null on it presumably theres a missing configuration for a title somewhere +I also took this opportunity to move the definition of the fencing key out of common into HDFS since its a +Hi sure attached is the original process definition the exception and then the fixed definition btw I am running this whole thing on tomcat and once this exception occurrs the tasks view seems to be broken for good even when I end the process I can provide more info if needed broken example will not include this different naming strategy for the upcoming release +after has been applied and appropriate tests have been added to srctests +would be welcome. +What we do here is we put messages to INPUTQUEUE and the consumers this time configured to use spring consume them putting a receipt to RECEIPTQUEUE +Can you please update the patch? +Thanks +Ready for review +Ideally both a thrift gem and a deb should be availablebq +Thanks for the patch! Could you please upload it to one of those instead? This JIRA was tracking the basic test framework and we have already checked in the patch for this +Patch lgtm +Yeah I can do this +migration to ivy may have resolved the problem +This patch doesnt apply cleanly to trunk anymore +Had to import inside SVN some parts of the ME plugin +contains my current rework of build procedures for MS Windows and UNIX +UliXvfb and firefox are now installed on vesta + Upgrade antlr version to has already been filed not linking as a blocker until understanding the general version compatibility strategy +bq +Thanks for the list Carl! This does help a lotI talked with a bunch of team members +I fixed on trunk and +Also rather than expiry I think it would be better to have limits on number of queued jobs per user and the max queued jobs overall +This particular test doesnt fail anymore but we have the same issue for other tests from time to time +But the error goes away if I change the channel to memory and reappears when I switch back to file +Clustering Known Issues in the Release Notes this was fixed in CP +Apparently adding a space at the beginning is not a complete solution I then get an exception when its the standard lucene parserA complete solution to what? You should get an error since its not valid lucene syntax +I see it included +From the build log Plugin version is. +bulk close of all resolved issues prior to release. +update version of the artifacts +Cleaning the data environment is already in the file called You can start this process by following command +If you are ok I will commit it. +Uploaded with additional changes on top of v +is enhanced with get API +Applied the fix from Lukasz +This should be fixed but please comment if you test out the upcoming and it is not. +Please find attached file which fixes the above issues +The basic idea is that different newline recognition algorithms can be injected into various classes using an abstract interface +Absolutely not a feature may not be ready forThat does not mean that the code for that feature will be deletedIf links is not available in let us document it clearly when all the jiras are ready links will be available in that release +After apply the patch the problems seems to be gone +Fixed. +Seems like a great time to clean this up +And they all pass locally repeatedly +We wont delete directories just filesOne sanity check this may be an orthoginal issue but is there anything stoping a codec from using subdirectories? what if i have a codec that creates mycodecfoo and mycodecbar +Thanks Army for the clear and detailed description of the subquery processingTo confirm it I introduced some crude print statements into the code and without the patch I can see it back and forth between and with the patch I can see that it simply uses the row over and over as you suggestBased on that evidence and on a close read of Armys descriptionIm comfortable updating the fileAttached is which I believe is ready for reviewI intend to commit this patch to the trunk tonight or tomorrow morningand Im also intending to merge it to assuming no additionalconcerns are raised via review or testing +trunk Thanks for the patch Ben. +The previous comment contains a patch +It will create junk files in datastore and generate jcr events +Better conceptual integrity to delete w correct system time anyway fix HH of tombstonesthis should fix the problem wo having to manually clear out old data +Closing this one and leaving as it was referenced from multiple emails on the. +I didnt thoroughly test the code with aliases +Yes please improve an error messageBecause we cant find a position of error on template code +I did a small test with the examples but forgot that click was mapped for static resources +Thanks +Thanks for taking time to investigate with us +Added a feature request in the EJB module see the linked JIRA and have been implemented defined inpluginsjbossetc +Switched the xercesImpl dependency to the version used in the Ant build +Also is the session still open when initialize is called +You can instantiate instead +ech +Everything is fine +All Done. +So when a new shard has to be created the command can lookup and see if there are unutilized live nodes and it can be assigned +Introduces an abstraction layer that detects the VFS version on the CP +The role of the Shark in +If you have suggestions please let me know +It is impossible in current UI and indeed in current API +However the new flag sounds like its exactly what we need +looks good +Ok I will work on a patch for the proposal +Agree too +only log the errors but still add the new userb +what is the actual issue you have seen here? Spending time searching for this seems a little wasted in my opinion considering it works +I have attached a patch which fixes the optimization bug by checking that a subquery has the same columns as the query before pushing the duplicate elimination into the subquery +I have no idea how they came to be like tried deleting them but then I got an other error +Query plan for comment made by Kevin Hore +Assigned now +bq +For instance if we implement a separate for redirecting HTTP traffic instead of incorporating in the webapp becomes a Standby service which runs only when the RM is in Standby stateFor now we can start without the and add it when we need +Patch applied +Another option instead of implementing this on the message converter add a list of exceptions to requeue and reject on a +just committed this +For these browsers info icon in errror windows doesnt work at all now +Yep that makes a lot of sense +Hey Elliotte! Too bad youre not still a couple desks over from me or we could walk through the process together +Since these are components I think we can justify some sort of synchronization around new instance creation +This patch include Java UNITTESTJava and UIJava patches in a single bundle because I was able to parametrization on service provider interface +Task Configure ISAPI to serve a Basic Cluster Step +In line with your comment that it seems to be working I am closing this issue as Fixed for now +Can you check that youve got recent changes from svn and also make sure that all your projects are in the same folder +For safety the application has to implement a mechanism to guarantee that it will only read committed entries a la +Move the fix to +That basically disables the orderby followed by groupby optimization which was the original motivation for the jira +All locales were set to observed on fc with open jdk doesnt build with JDK at all we have put in about patches to make it happen so dont bother with the whole stack being buildable with JDKCould you please propose the patch to the Hadoop community and link this JIRA to it +The basic issue is that clients that want a default implementation need to develop their own in every case +classifier is now configurable. +is this same bug and thus I am fixing patch from Andreas with more modifications to support both envelope and body +Thanks +So I dont think there will be an official fix for this from a Xerces point of view. +PaulexThanks for the numerous patches and improvements I believe that this issue can be marked resolved nowPatches applied to the NIO module at repo revision Please check that the code was applied as you expected +Try the latest build and let me know if its fitting all your use cases +Here the svn d recommend starting a discussion thread on the as it can catch the attention of a broader audience there +I will commit this in the next day or two if there are no other objections. +patch fixing the bugs in the and +But the rest of the comment should have some substanceIts time to get home +Saturday? updating both the doc wiki and +JIRA cleanup Closing old issuesJBIDE no longer supported +fbsd doneTodo Fedora OSX DoneWindows next +No I havent run the restconfig unit tests with this patch +The command I ran wasjava jar +Well manual testing of my changes seems promising +We have reverted the upgrade pending the resolution of issues under investigation that were caused by the upgrade +fixed in r +This issue has been fixed and released as part of release +Feel free to give feedback in that issue this was my first experience looking at easymockI added a comment on but basically you did a fine jobNow that is closed would you mind continuing with Mark +Allen Some people just dont want to install gccautotoolsetc on their Macs +Because I changed filter in dojo js files requests which matches the pattern strutsdojo will not go through struts +As an aside is there a reason to allow a single component composite Yes we want to be able to query with strict inequality on PRODUCT they are ordered by name +Issues raised above should be properly addressed +Cant be done as this parameter already existed +SVN integration in tools is still not as good as for CVS and also SVN metdata may almost double the project size comparing to the same project that is using CVS metadata. +It would be great if you can name the attachments with some incremented number in the back like or so +Would it be worth adding some text to show when at least some null attributes have been omittedCould otherwise include the count of omitted attributes but I expect thats not necessary +supports random provides random access too because it is accessed via HTTP protocol and you can use Range header to tell www server which file parts you want to retrieve. +Here is a patch with a that reproduce the issue and a fix for OracleOnce the patch is applied everything works fine on Oracle +I agree on the fact that JDBC Connection implementations will fail at class load time on a JRE because the JRE libraries do not contain certain I cant understand why a JDBC implementation wouldnt RUN on JRE exactly like a JDBC implementation does +Per review meeting pushing to Westport +Thanks for your review and time Jacques +Moving into CP for now until we can discuss with productization team +Accept to document this as a limitation of GEP server adapter +Moving the refreshing of the list and related refactorings +Resize doesnt help +to trunkThanks for the review Jimmy. +WFM +Added code to allow the alias name to the ssl key store configuration so that if the keystore contains multiple certificates it can choose key by its alias name +This is now fixed +Ok I did not see that it catches the exception after calling the cleanDirectory +this is the core businessplan component +Version Released +commandc The code is already present when the archetype is created +Maybe we should weaken it and say something like in order to fully shut down the system instead +thats different +Actually Im not sure whether will be a good base to build this on +Fix for deadlock +Closing for release notes. +Currently the client andor server goes into a separate mode to handle SASL +patch changes the default value of from All to null which means that by default users will not have the ability to access tablesthat they have created rightIf thats the case then I think you should really fix the underlying problem in insteadof working around it by creating a default setting that virtually everyone will end up havingto reset in +Please help isolate this error +the patch to this to Gerald Mllan for thisregardsMartin +Integrated to and trunkThanks for the reviews and +This is integrated +And one thing need to note is itll have a conflict with seems +Ive committed to trunk. +Part of the problem relates back to the jar files +Oh +the vdb property should just be a single comma separated property +Added icons folder to +There is a patch attached on that make it work in portlets +Hi JacquesI am pretty much stuck with my work these days Ill reply back with the reasons asap. +We get the error about times every day in our environmentMy guess is that the tomcat pool have connections with timeouts that will trigger and then the connections will be closed or unavailable forI dont have any example code yet because of other failures to hunt down +is the text file produced by Tika +I scanned the pages I cant find anything obvious about how to get permission +In the trunk r the boilerplate is differentfixed +This will be very similar to whats being added to +Wire level logging can be enabled by uncommenting the logger config in samples +I thought maybe I could set a breakpoint in the code to see what was causing the but I neverget to the breakpoint so I have not been able to determine the real cause for the problem +Duplicate of +It ought to be possible to systematically run a system like that on a fixed set of programs to check that theyre behaving +BTW all the tests are passed with this patch file +Also does anyone know what font is used for the word hadoop in the project logo? I havent found a good way to figure that out +I am going to open another issue to bring in to the those who are just casually following this issue is there a good summary of current input options and example output +Until now I understand that the value has a weak reference too +Replacement for the originally submitted patch +Aaaaaah I was trying with an imported keypair +I dont feel too strongly about this as there is fine line one side caused it and the other side detected it +Thanks Hitesh. +If you are in a hurry and will work with trunk Ill make a snapshot + +and how about release date +batch transition resolvedfixed bugs to closedfixed +The attribute typejavaprop of the parsable element does the work. +Missing before variable +any thought on having this as part of +Removed references to EE from the source +Inline elements get added to the related part while attachments get added to the mixed should be available in public CVS tonight and also in the next nightly build +I see the need for getting the remote clients IP address in Hive in some places other than just logging +found this issue while trying to research this error on my own +The commit has been applied to the branch and master +Fixed by Charlie in rev +Przemyslaw could you please grant ASF license to include this patch? I will let it in for now but wed have to back it out if this is not addressed +We should concentrate on QA proper. +Ive updated the patch +Hey DenisCan you try to replicate this? I committed a patch +Patch looks good to me +Fixed in r and r +All these are commented out and flagged with FIXME +And also added a testcase to parse values with special characters and patternsbq +updating affectsfix for taking the time to provide a patch and test case although I decided to use a slightly different approach leave the interface checking alone and provide and additional check through the parent class hierarchy if all else failed. +Heres a new patch with updated tests +Other deploys Im aware of use Chef or bcfgIts not that I think the feature is risky really +That would require configuration of multiple reports with many different datasources that need to be matched to the appropriate for M to allow original poster more time to comment +gurlee you have tested intermediate version of UI it is just changedSebastian Click here to share your screen is actually Click Start to publish your screenI think the more compact is dialog the better it will fits various screensIll add description for the new tab and rename the tab butI dont think Advanced is a good label since it has only Publish controls I think we might add one more tab so its going to be Advanced +Solved with minor compaction and works via scans +There is one particular issue about URL formatting that will be in dev momentarily +Thanks Todd! +Switching all issues that have been in resolved state for more than one month without further comments to closed status +For whatever reason it ended up doing the former where the latter does make more sense +Hi Andrea and SimoneWe have just updated the the RC released today and found that it did not fix the problem +I went into my project folder did several ls commands and even an ls R +I was just a bit curious I think it might be possible but that we should not do it for soft upgrade +See reviewboard for more color +John Please go ahead and commit this to trunk +But if you only have a GB machine theres not a whole lot we can do + +As what is done in we should also disable region splitting at the beginning of the operation and splitting at the end +Im doing some refactoring and cleanup of the record handling code so I expect things to be a bit shaky over the next few days +Thanks Navis! +just applied it you can pull it out of branch. +I think it is an API issue but not sure +batch transition resolvedfixed bugs to closedfixed +Will the release of be coming soon +I also didnt see anything but whitespace +You may try to apply his fixes +The nextDouble is inlined also +This can be fixed if it imports and exports the packages in the same manner jsr does. +Not something the ESB should be doing. +Bug has been marked as a duplicate of this bug +Its an eclipse project just add a lib directory containing the following asserts ten User objects then hangs while asserting the third object +Should be fixed. +Thanks for explaining +the release candidate is attached +sense to me +So not on our company domain in order to rule out my Windows macihne at work +Added two unit test cases and noThanks +Seam and Seam verify the fix in the latest forgotten line in for declaring exceptionHandler variable +Although this isnt a single watchdog it should have the same end result +see +if configurabletrue pushed configuration will be propagated to the service registrationIt allows the enabling disabling the propagation easily +Great suggestion Billie +The logj configuration defaults to CONSOLE which outputs the logs to stdout which in turn goes to yarnYARNIDENTSTRINGcommandThe fix is to either update the logj conf to use the yarn variants of these variables or to set the hadoop equivalents of these variables instead + is somewhat related to this and this issue might actually entirely encompassAdded different quickstarts together with a distribution module is the porting of the module from the examples repository is a quickstart that shows how to use from different types of Stateful Singleton Stateless CMT and Stateless BMT +Added class which is used to make direct java calls into a UDDI +I really appreciateSanjay Patil +Thats my plan in and in fact thats the current behavior of the SBN even when operating on NFSHari updated me offline that standby NN currently tails only the finalized edits +Reopen for committing to +This is the test application unit testYou can start the application by executing class ApplicationYou can run the unit test by running jUnit on +Hi MarekThanks for working on this +Plz let me know if you think otherwise +I do not see why this code should be added to +Standardize use of content vs item +Since it would be a lot of work to flush the commit log prior to a rename we need away to get from the old cfnames to the current id +It wouldnt be any more robust that using a single it would justuse more resources as youve witnessed. +Is the solution not just to write a constraint validator for and add the required xml mapping? What else do you want? The only thing which would imo be helpful is a automatic custom constraint validator discovery to safe the effort with the xml configuration +Why are you creating aliases with invalid characters to begin with +Fixed in svn revision the fixversion to as discussed +Sure +What is DSL +Just now we finished the process of reverting those changes back to use and now were not getting this error anymoreCould someone explain why this might be? Do I need to have any specific code for dual ssl auth with? +Yes very strange +for not creating +I no longer see this issue w EAP +As I had some questions regarding I will add a sample for this too +Proposed fix while I continue testing +Since the time to rebalance regions can easily be the same as rebalancing regions the disruption ends up being almost the sameTo optimize that situation we want to keep running the balancer often but first check against some sloppiness +Done +Patch contains updated scripts tested on Win and Linux +All Metro Axis and CXF does support this +I have been given a cut of ER and it does not include Hibernate cut of ER passed onto me does not include Hibernate +Surely we dont need to call it more than once for each connection? Also I think it would be good if we could test connections to the internet because if we dont do that were just testing that the client and server match up if you see what I mean and were not testing the code with a real http server +Look at new screenshot +Remove the FIXME from the testsuites to enable the associated test +That split method was mostly rewritten +Oops I just patched in the IDE and didnt think that through when I noticed compile errors +Hi NethanelCan you answer me the below questionsWhat is your axis jar versionwhat is the xml size u send for parsingWhat is the application server you are usingWhat is your Heap memory size +Same issue with Trinidad on a that has a Mojarra +That code looks a lot like the tail ofjgray said we should add a test for deferred log flush will consider that +Sorry for the inconvenienceSee for the original issue that triggered this JuergenWe can work around that! +With r the markup contributes before the +should I replace the kitchen sink sample app in the webOS repo with the Hello Cordova app? The last set of changes work on the HP Touchpad running webOS +Only two bugs are actually fixed between RC and +Patch integrated to trunk. +It seems like a reasonable fixIm working on packaging right now which moves the init scripts around a bitIll note this to look at it later applying this patch right now to master could cause some merge issues later. +Updated the wiki with some simpler examples. +We should add a Wikipage +Jiannan Ivan Thanks for your commentsYes that make sense we should removing the dependency on construction also I prefer jiannans new implementation of Scan and compare gc algorithm which removed the ordered scan limit +Is this a blocker for or can it be incorporated into +Another use case is providing atomrss feeds at specified urls which should be are many feed readers which could not authenticate with such urls since they get a redirect to the +What do you mean by organise? Alphabetize +Should be ok now +to Christians planAnd yes copy the logic for the AMQ we used to have it already in Camel before the Google App Engine forced us to change it +With improvements we take about seconds to add items +is this issue opened against. +Spent a whole day chasing down this bug is there a maintainer of this code who reads these tickets +change the version information in the I will try doing release of GEP +Applied the patch and it looks like it fixes the problem +No longer using independent logging servers +The manager launches a Felix runtime when the broker starts up and that prevents Qpid from running inside another application +scripts to reference the new XML JAR addition I had to change the class as well to use JAXP for obtaining a parser and change the mechanism to using reflection when instantiating Xerces feedback appreciated +Whenever the Derby docs are talking about data types they should specify data types and not just types +Please verify +Agree with the idea of a filter that can be filter should be enabled so that we expose the standard http behavior on default +Id vote for just removing rdoc completely but wed need to make sure no other libraries we ship depend on it +wrong target versions +Reattaching the patch from Bhallamudi +Issue may already be resolved may be invalid or may never be fixed +Please provide a POM we can reproduce the problem with +If YARN needs to hack a workaround to fix an Oozie problem I would also like to see what Oozie is doing on its part of the bargain +Patch originally done by Dhruba +Isnt that what we want +Looks good +Committed at subversion revision +This is the gs trunk followed the Justins comment and cleanedThe patch is Catalog break when usigi have prepared a patch to patch introduce an interface that the and implementing +Fixed another issue in where output file may not get closed properly +Updated +To run quickstart click on Show results and then on a link below +I will commit to branch +This occurs because scanning for attachment annotations is based on the getter signatureThanks Toshiya +Verified. +for Daniels is correct in that the vast majority of +I suspect that you are trying to escape some char having something like in your DN value +On order for the compaction to make progress we could assign scanners to epochs where everytime we change the readers for a store we go to a new epoch +And we also have no error recovery ast classAnother change to your code is that our unicdoe reader seems to allow multiple u in a escape sequence +This bug showed up during the porting of the Sun BPEL Blue Prints projects +Committed to JBESBGAFP with revision to trunk with . +E +thxi have solved the ln problem by change language settingbut when I customize the portlet still cannot see the portlet correctly. +Will do +Must have been fixed with another is probably a duplicate of +The correct solution in this case is to put where it belongs in lib +This patch failed in release a new patch with Apache license and SVN Trunk Revision update adds three new warning as it uses classes which have been of comments +You can migrate them whenever you are ready +I would love this feature +still fails with the same exception with current trunk rev using the launchpad war +Switched to version of Commons Jexl +I got the same error as Anita when building from the src zip from a clean local repo +Thanks man +And I realize it can be deployed to the server by right clicking on the ds file but shouldnt this be automatic +Its always good thing when two independent source come to similar conclusion +As it is theres little chance we can reproduce this or even know where to startNote that its usually a good idea to chat about this a bit on the user or dev list before opening a JIRA just in case its a known issue +Logged In YES useridThe fact that something doesnt appear in the spec doesnt mean it isnt implied +This was missing some tests which tony mentioned he would fix +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +How we map this into any Client Server interaction or a query language should be the responsibility of those classes not the StrategiesTrue +Got it +Until we can get something more fix for +adds a percent complete test for Hive +That way we dont need special checks at Hudsons +fixed up the javadocs +How about? John +did these get out of sync when moving to hibernate + +This issue was created for the first version of theThe new implementation is tracked in. +Is there a complete list of possible http methods somewhere? Is the list recognized by sun expanded in the servlet spec +for the patch +I want to resolve Drop support Smooks but I have change this issues status +Sorry saw the patches and assumed it was about version changes too. +Maybe as separate issue? I think Cosmin is off around the wilds of the at moment on holidays +How does it look now +OCM isnt an issue. +The main points I like here that Ive been hoping for are The rename of example to server Integrating or at least organizing the various configurations +Lets check +JeanI appreciate the fact that we can work around this issue but instead of forcing all users of Karaf who wish to host Camel to update their file wouldnt it make more sense to fix the bundle to rely on the packages already exported in Karaf in bundle and not install another activation library? Or could we be more specific in our import to use the package from the Activation bundle +Martin the code is updated +In the reference documentation the section Interaction with the dependency system contains an example which refers to this particular error X with the old stringI have not looked at the entire documentation to see if such minor errors exist elsewhere +Its understood I think that it is experimental and not guaranteedI think perhaps the easiest way to test these things is to create a small test case manually verify the output is correct and then save it as golden output +My suggestion would be to mimic how rexx or perl or java handle it +Kirthi I think your build also is missing some dependant jars +If this is so the test should pass so I added an exit condition if the data dir is not available +Thanks for the patch +Added a doc issue +I am still adding a junit test for this patch +Let me know if you want me to commit it +OK so there was a bug here +Ive committed it to the master branch. +I just realized that the docs are only available in source form until is done +Updated name and description in JIRA. +Spoke offline with Todd about this he is also open to other suggestions but concurs that split is an appropriate term for this behavior +patch would be nice +request placed at wiki and linked to this issue +Thanks a lot for taking care of that Colin +Modification compared to the patch is as discussed replacement of the final modifier by indicating that extension and instantiation is not intended +Ignore my comment since symlinks should allow creation of dangling links +Just because it shows how to change to a JMS provider we dont run tests against is irrelevant +It basically redirect stdout to +sbt seems to work for me on a fresh checkout of +There is no way Arquillian itself can have the knowledge needed to clean up the data dir or the tmp dir +Thanks Ashish! +What happens when you run v? Im not able to reproduce this issue on my end +Is there a workaround for this +At any rate I am always happy to help someone doing documentation and tiding up +So who broke this? What needs to be done to fix it? Galder you reported this but the error looks like JMS is missing fro the needs to happen here +on the rest of the change +I will prepare the artifacts for the RC tonight +if you are unable to allocate a MB buffer to send a rpc response it might not actually need to kill the RS because if we are truly out of memory different threads will catch thatThis sounds reasonable to meAndor throw a or similar if the BAOS size exceeds some reasonable value with exception text that points to doc about scanning +Also tested the result type with logging statements and it is only called once +EAR that reproduces the logs with you try any other older version of Struts And it would be better to attach a Maven based project instead of ear file not needed found one +Seems to work fine with the patch +As you can see we expose methods like get to allow you to drill in and get as much detail as you like but appending stack trace content onto an exception message would be a pretty unexpected move for most Java developers +bq +Also got all the unit tests ported over from AMQCPP things are shaping up nicely +Hi lch hstDo you think James does something wrongWhat scenario should we better coverThx know I havent seen this with the IMAP of my ISP and also havent seen this with Gmails IMAPI think theres something going wrong with JAMES As you could see on the mailing list more users are experiencing this problemIll do some more testing today +The extends making a mockery of things at an implementation levelAt least the tutorial did not make that mistake +For somehow reason seems to work +I am doing something very similar to the person that entered this bug +Update the project versions to I have added you as a committer to theCan you see if you can commit the patch yourself +Heres the patch so far +Its still needed if you put orrename files on the serverPlease check and mark this bug as verified if it works else reopen itThanks +Feedback confluence page currently serves as documentation and current status +I think this would be a good solutionAlasdair Nottingham +This leaks over into the runtime dependencies +It is true that these are no longer maintained by the as fabricated bundles +I am running derbyall test suite now +There was a version in the public mvn repo for a while compiled with JDK by mistake +Ive also run all the tests and everything seems to work fine +Here is v patch +Thanks for pointing me to the solution I was too much focused on seeing only a filter declaration together with a think that definitely solves my worries glad you got it workingClosing the issue as wont fix +Ok thanks got time I checked Netty did not have support for UDP but there is a JIRA for that +Hint streaming is just normal writes as far as the recipient is concerned Im not sure how that could cause a lockup +But Im still a little bit puzzled if the name application really fits to all that different view page and docking implementations +push to +Ive been testing CRX Content Explorer and looks great But now I have a question because Ive seen the Editor how do you implement the ACL? There is no default ACL managemente in from the Jackrabbit roadmap due to lack of progress. +OKI see +Thanks Rodrigo! +Well need to define what happens if a has an observer for said event on it +rss so some of the feeds got indexed like the way current nutch trunk do as htmlJust some early feedback +Could you please rebase? I will take a look at it as soon as you do +Duplicate of +I also created the corresponding unit tests for itThe patch is based on revision of the amber trunk +Pablo +lemme see if i can rework it for raw thrift +Ive removed caching from the project builder +Of course if a broker is not up the consumer may not be able to consume data from it +Someone with Edit right could you please change the issues summary +jelly and ant have their own issues now has a separate issue leaving just +Sounds good! +Perhaps you should update to the latest revision +I added a fake string to the plugins Message class and not to its and was able to reproduce it on the new string but not on the two in this jira +Rgression tets are successful +It may be that similar sealing violations will plague them +This isnt going to get done because its really only a nice to have time would be better spent on other issues than gleaming the cube of the uberbuilder +Therefore it become part of the public API possibly impacting the ability to make drastic changes to a datastore +Committed to trunk for. +They are now +MarkCan you split your big patch? It seems it do more than this issue +ah makes sense being ci and allyes lets create an area and keep builds for nowthanks much +xsd files +Lets follow up on for the rest +I dont know if users will need to add items to that we cannot foresee +So from the debugger point of view requiring debug is quite OK +The jps requirement only occurs in the bash and python scripts so it shouldnt break functionality but it will cause the test to fail +You can modify the default using the property process +If we only had the remote case it would not be necessary but the embedded version can get stale referencesAs I try to justify it now I can think of some other ways to do that now that the core is required to be selected in the path +Will make available a snapshot release as time permits. +There are a lot of comments in this issue and from my perspective this appears to be a very intelligent conversation +Preparing smi +pls adjust up the priority +Thanks for the patch Sishen +applied to to +There is a dependency to ajaxjsf and the configuration for using ajaxjsf in the file +a patch for +Mamta I also thought about alternative solutions before this patch +I dont think that will happen in current master tree and stable releasesplease check the result oftrafficline r it must be set to to enable this out to please move back to if this will be worked on soon +See +Reviewed patch +Thanks Mark +If case makes sense +moreover any proposed changes here should also include the changes to and so on before even being consideredOtherwise the apis just get out of wackMaybe we should just seriously consider just switching to + +There could be hundreds of UI labels in hundreds of languages sent out per HTML output +Patch for fixes small issue with arguments and adds crawldb output and status filter +Ah I see I misunderstood the description of your issue +Any help +After reading the WC docs I have seen that we can calculate the Etags in a much simpler way +Thanks for the patch Ryan! +Reopening this to propose a branch a patch for based on this fix +mvn archetypeVERSION +When is released Im planning on writing an article onincluding code from the tags +Ill merge this into the trunk after some optimization on object serializationThanks again! checked the code again +Exceptions and Errors will output in red +Just as a side node what surprises me though is the fact that the code generator does not use the internally to parse a set of XML schema files +So what I am asking is whether you are willing to take this ownershipresponsibility for the +changed to use a different class already in the file so we wont have the duplicate class +Ive just checked out a fresh trunk applied the patch and successfully run the tests +And after all the effort I see this bug and the plugin owners disgusting behaviour +But I added expert ctors to each to force deletions to be respected at a perf hit +Accepted for think that upstream accepted the patch +That means we can now enable for all bit platforms? Solaris windows Linux any others except? needs to be checked but I assume its also faster there +Makes perfect sense +the xbean root pom has a junit test scope dependency +Were not getting jira spams and I was are very close to bringing writable SVN back online and restoring web out now works and is already a good step forward +I guess we can close this and I can ask in the forums. +Fixes the case mentioned on the mailing list but hasnt been subjected to the same scrutiny as the get in the singletons make this excessively difficult +got this using the plexus runtime +Phabricator seems to be down +I will let Andrea comment to that +Editorial pass over all release notes prior to publication of +Fix tiny issues on comments in v patch +Patch for service engine wsn +Test projects you need to import as Maven projects +Disconnected the slave by pulling the network cable and after sometime the host was marked as downCan you try out with XS and see if this is something specific to XS +Willemthere is no unit test except an example class packaged under srcmainjava +Same problem +Im going to leave this up to Cassie or other people working on the OS java bits for now since theres still a lot of work going on in those parts and I dont want to step on any toes +Updated to handle most of the tests +Randy has addressed memory usage by putting a session listener in place and pulling cache entries out on session expiration +If you dont want the map altered make a copy first +Attaching the same patch file same file again to get a test run +Thanks Andy +you mean the current selection not the project right +Could you please submit few test cases for thisThanks for the effort +And it went back to work this morningPossibly there is something wrong with the in the repository +Any chance you want to contribute a patch Here are some thoughts from my side +Id be happy to try this out for your use case if you can share it or provide a simple example +Committed +action except for the patch and started to fix up remaining problems +Issues are mark issues relations use the Link operation of JIRA. +Now that we have the baseline we should make sure that we dont introduce any new failures +Release of the application is round the corner and this is a blocker for the thread dump report +Sorry I was confused by the picture this is still missing in deployment tab has been secured by Administrator Role +I was not able to reproduce though but Ill try harder tomorrow +not sure if maven has the equivalentunfortunately not but the similar functionality is built using the property +Id still like to have transitive dependency checking but that can be added as a separate change. +Can you test and report back if it works for you +Happy to update the FAQ however I came across another windows issue I will raise in a separate JIRA + +Currently the livereaload features name is just Feature +Thanks folks +Attaching as repro this version only adds the fixture in question to show the issue +Submitting the patch +this is a perl script i found attached on a separate forum in a thread containing a similar issue +The patch adds some overhead by creating a string buffer containing the defined operations but only when the deployment would fail with an exception +In each left propagation for retract or modify would throw away the right matchines and redo them again there was no state managemetn +released +Can somebody please review this patch +I have no idea what is going on nowWait until we release and then we can see how it behaves on that +we now corrupt all NN directories we now test upgrade from which fixes some unexplained test failures clean up the test drop the datanodes and use mkdir instead of writefile for quicker test startup +Single integrated patch for stricter checks on filesystems +Thanks Aaron committed to trunk. +No after I add the option sort +bq +I looped this for an hour or so locally and it passed every time +My only thought is whether it wouldnt be better as part ofAnyway if you or someone else doesnt beat me to writing it i might take a stab at it one of these days +Ah ok are you able to share an example app now or are you still running into the deployment issues on beta +Checking this in +To be able to enhance JDBC cache store in the way suggested here its necessary that the right infrastructure is in place in order to allow DB vendor specific extensions. +Caused by +is for the Sqoop branch dont think it will apply directly on trunk +The first attempt to look up the value does so without locking it +a patch I use in my Project to have the name duedate and execution data available in the history task please note this adds column to the database table so it is probably not a candidate for the think this should be postponed to version since shall be released now and it does not bring any data base changes +The new patch makes a minor modification to the error message that is printed when HOD does not automatically deallocate the cluster +Thanks Ill consider this as committed +Merge updates from classlib trunk since rTest status TestsFailuresErrorsSuccess rateTime failures test new failure test Regression test new failure testI new failure testloadLjavaioReader new failure testgetlangutillang the same as trunk testentrySetcontains related testimpliesLjavasecurityPermission the same as trunk errors testHarmony new failure test the same as trunk test the same as trunk testValues related testViews relatedAction To Do +This would prevent maven to pull jtype transitively +Although this does not solve the problem once and for all most common scenarios should work much better now +It is possible to share existing Foxpro tables with +How I fix thisIm using with help me +I discovered the same issue last night still digging in the code but this just smells bad +For what its worth I checked the mentioned methods ntzpop the same implementation is included from Java upward +While I appreciate what you say I need a license to be displayed somewhere on the front page of the website and was really hoping that you could oblige by adding the license link to the main Maven website that is unless there is a good reason not to. +I will have a look at the code in question but Ill be throwing darts without a reproducible case +I tested this patch in windows please verify it in others box +Thanks Dan +Thanks for the patch I really do like your emulation of full disc and network problems of your test case +Adding pluginsrcmainjavaorgcodehausmojowebstart pluginsrcmainjavaorgcodehausmojowebstart pluginsrcmainjavaorgcodehausmojowebstart pluginsrcmainjavaorgcodehausmojowebstart pluginsrcmainjavaorgcodehausmojowebstart pluginsrcmainjavaorgcodehausmojowebstartgenerator pluginsrcmainjavaorgcodehausmojowebstartgenerator pluginsrcsiteaptexamples file data +Dont put it as part of the projects reports +I will write a unit test to see if the existing API supports NULL keys +The namenode that was intended to be part of the same cluster now is notGiven this we decided not to automatically generate cluster ID +I will file a separate issue about the second failure +Let the usersysadmininstaller worry about capturing that into a file and running it as a user with write permissions +I just committed this +This is now mostly done but need to resolve directory listings to if they exist instead of directory redirections are in place I think this issue is done +Ok +Personally Im allergic to static fields and can only tolerate them in small doses +Ive found that in very large SF instances just getting the table cardinality can cause an axis +Attaching a patch for this issue based on the branch this could also be added to the trunk though I havent tested need this patch submitted by the end user into the Apache issue tracker so that proper IP management can occur +Most of them were not useful and thereby removed +But its not just the form fields thats automatically populated +I think you should commit this so other related issues can start to benefit +Still we can discuss about the optimization as we discussed in above comment like we can process only the current genstamp blkID and postpone remaining stuff +Interface is the most important part +Ill probably make a different issue for that less frequent case +xmlcodegen plugin configuration +It looks like the Writer is replaced with a where it whould be the opposit way about the earlier patch +Commited on trunk as of revision +Though it would be good to get bulkload security covered in the current stable release anyway +Classes in still need to be individually marked as deprecated +Thanks a lot! I didnt apply your patch as is because I thought it would be better to avoid the inner class and the formatting method would actually be better placed inPlease realize that the maintenance branch is frozen and your change will most likely not be applied there +Will not be fixed since it is not in jspwiki code. +Just trying to catch up with pogress in my absence +This is entirely my fault for not reading your instructions but something I think people will hit when they download hcat and try running tests +Oops my bad it does indeed work +bq +HDFS uses encoding +Here is a patch which adds a two second delay between changelog comparisons for the case of faster +Just meant that the cannot use the because they depend on validation which the composers do not have +Ive fixed some code in the in and so that you dont get exceptionsBut CSS and doesnt workIve added a simple example and testet it in +Fix committed at r +This issue has been fixed and released as part of release +Example of broken app by the fix +We can use the bundle id as the version property +There is a handy Eclipse compile time check for Serializable class without serialversionUID which would catch them all at compile time but for now heres a couple I found with a runtime test bringing up bin onThe class does not declare an SUIDTo retain compatibility with existing persistent form it should declare static final long serial LThere is a serializable class hierarchy rooted at that does not declare a SUIDBringing up the bin exposed which to retain compatibility with existing persistent forms should declare static final long serial L +Probable Location for this information is the Admin and Config Guide for stream +Yes the branch is in a kind of weird intermediary state right now +HiJust a very quick glimpse +did you have a look at how the python dbapi driver does thisbasically will change as the columns in the row does +or maybe making the locking and an option via +Lets keep the scope of this ticket to splitting the rpc timeout +Fixed in rev . +is default java type to map xsddateTime +Well crap I pulled and still get it +srcjavaorgapachecommonslangexception +If you think it would be better then fine +This behavior is also noticeable on mobile showcase in demo for autocomplete the last input which has selectFirst set to falseIt seems that there is only loaded results and then it is possible only to choose with these other results are loaded when there is BLUR on that input but again only other results are there +Which refactoringthe method renaming and constructor to use factory functions rather than constructors +The same can be done with easily without any neglectible overhead +defer all issues to. +See for the REST API dependencies of the Web UI. +Thanks Andrew +Sorry BrunoBut this is still Jacquesthis is finally fixed in trunk rev +Hi Evgeny what does it require to have this functionality working? If its not too complex that would be awesome to use Sonar CPD as Im not really happy with PHPCPD +Easy wrote the patch on and made a pull request +Moving to later +I think we should then add to the future list to throw an for such cases instead of terminating the connection +Please review attached patch +I think that there is a th option +However Im wary of diverging from how its been configured in the past which is in terms of absolute amounts +Thanks Pete! +So i would like to provide a patch for this +work branch the latest commit concur with you we should close it with Wont Fix because performance of this query can give users negative impression. +OK this problem is fixed +took out too much should have kept the login idea which keeps most calls using a simple session id +What needs to happen here is that any valid API doc bugs from the old bugbase should be migrated to this bugbase with a common label like apidoc and no assignee or fix target +It would only have been affected if it used for WFS GML + +Sample XML that should be processed by a description and components +It is hard to believe its slower but hotspot is surprising +Except there might not be enough type information to construct an object +attaching patch for wcs needs more is an updated patch with some testsThe RBG test works but if uncomment the row to make the tiff dump there is a black row stuck in the middle of the result raster the DEM test does not work at all as far as I can see. +Mark Thoughts on Kevins last postThanks we have someone representing the ASF on this JSR in the EG? If not I could ping a few guys on the EG or even ask on the eclipselink listId rather like to get this confirmed before we implement it the wrong way +See Joseph I agree that it would be best to have an option for automatic handling of in names +reassign it to me once reproduced +Havent heard back from Weston so rescheduling for a next release +I think requiring AL for project websites is appropriateI doubt that will be controversial +was adding metadata constant so figured the moved was relatively the same ball park +Added Bard as watcher this should be reflected in the lang docDowngraded to Major this is not a critical item +This patch includes the MAXIMIZEDARGUMENT constant and is also saved in the user preferences like other arguments +It is quite possible that it was just a error and it will be ok in CR thats why I wasnt sure whether to report this or not +Ok I will do that +Hi GavinThanks that helped after a little bit of fance foot work it now works just fineIt has also synchronized just fine for quite some time so it is ready to goRegardsPer Per added thanks +It would be good to reflect this in documentation +Already done. +I am also using to do SOAP attachments +patch available would you please have a try thanks +Unittest are expected to run fast and be part of the day to day developer workflow +Just delivered the first wave of changes +With this patch linger will be disabled by default +Huh interesting point lets ask on the dev it seems that there is supposed to be an Apache feather next to your file if it was contributed correctly +testexecutesign fails with in generate not all mandatory fields set in V generator because of fixture certificate has as a serial number +Web Support Usage +Thats true +STEP srcqpidamqp Apply the patch inside QPIDCPPROOTsrcqpidamqp +committed to trunk and branch +Also package needs to be changed to +I tried to address some of Dans questions from earlier and also changed the spec to reflect Dans suggestion about adding new property +Output from running git status +But possibly for just enabling JAAS and meeting min requirement it should help +Of course any bugs should be fixed but remembering the rather than mathematical focus +Did the upgrade help +right +problem is a outdated update this jar and it will this solved your problem +Tested with Build but still doesnt set log as static final +Linking this issue to +Now we invoke dependencysources to have a complete downloadupdate of sources that works also without indexes and is an atomic operation +utilise the jline lib for password entry to have first time entry on the command lineWe should probably recommend that deployment only be done from well secured servers by restricted users and have this mostly as a download option +Thanks for reviewing and verifying JunpingThe fix passed Jenkins on I have committed it. +Since Im still having trouble producing this could you run with trace enabled on this plugin and rerun the entire scenario? Thanks +Verified with latest runs +Sorry class exists but doesnt and due to this should be used instead of them +HiThanks a lot for the information +Trying anyway +Havent looked closer but it needs some attention +The problem is when I create a new document will return id and rev +I found a problem in the code that could cause this condition and committed a fix for it. +If it fails to install a jar it does not try to install it again until the jar has been modified +This can be done in I am not too eager to use this feature +we are also using but the problem comes into play when the redirect happens it sends the directly to a page so there is nothign in between to reapply the editors +generic client for services based on from using this as a model how would you get the HTML header information that we need +Closing all issues. +Nice logoMaybe an additional pixel row below Velocity avoids it to be stuck on top of the D +Attaching possible patch plus test +Stefanif Im not wrong now it seems that DR is using +for some reason latest base build was not piblished rebuild with helped to fix issue +Replaced by +What do you mean The No identifier specified for entity was not self explainatory +I have a patch +Im just waiting for the recent code reorganization to get back to xIm working on it! Hopefully finished today or tomorrow +looks like the message of this commit and the merge commit got mixed up +As such I will be going through this tutorial to make sure it is still relevant in the latest versions +As far as the test results I think it is a good idea to store them in a centrolized location +But both might get served by the same OWB jar +Applied to branch and trunk +Are you saying this should change in the case equals amd and it should be set to x instead? Or do we just need to build these on a different machine +All the tests pass except for which I then fixed +Resubmitting to Mr +also only supports the last two versions with new bugfixes limiting how far back to port fixesThis naturally forces people to upgrade to a recent version of Hadoop to get the latest Hive version but I dont think thats much of a drawbackPerhaps doing it differently in the early days of the project is ok but it also sets a precedent that people are likely to follow further down the lineFrom the wiki on the version numbering Going forward new releases of will have the same version number as the Hadoop release so there will be less confusion over which release of works with which version of Hadoop + +there should be an error entry either after or before this that contains the message +Because the ringmaster code never uses the option +svn rev that was fast +I have some time avaibable throughthis weekend +The version number as per my original report is have another machine which is running and it has the same issue from time to can replicate this issue on both these versions by following the steps provided in the original feel free to contact me if you require any further is that when a timeout occur while sending a mail the thread sending the mail was not stoped and the mail was stored in the task pool for issue was solved alongside with this ticket +Committed to branch tooCan someone who has the required permissions create a fix version for +When the statistics are not being collected the starttime shows Start Time Wed Dec EST +The sliding window approach is also good +This is under the all profile correct with distributed cache turned onIs the totalEntries ever +Thus at least for Oracle JDK we have a coverage for this issue +Could someone please merge in that given revision? It looks like this bug breaks both java and php shindig +I respectfully disagree +Closing this issue as a duplicate of +also check your Eclipse directory if there is files like hserrpid +Have a look at I think it does something similar. +Sorry I attached the wrong file +However Hugos comment need to be reviwed so do not close it +Just remove the include from assembly +Yes you are right +If your having trouble due to the scheduler the latest head version of the sources for the following directory fixed my problemsrcmainjavaorgapacheactivemqbrokerschedulerNote that we still see failed to reset batching errors with this update when shutting down the brokers if there are messages still pending delivery in the scheduler but thats not a problem for us. +This patch illustrates how the Maven version information could be used to provide it via the Java API of Mahout +For all upgrade bugs dumps from both should be attached to the bug +Ill have another go at the HR patch for Jacques nextThanks Sharanit looks like the file is missing in this patch regards Sharanyour patch is committed in revision +The missing info is not supported in so just remove this columns + +This issue will be closed. +bulk defer to defer to. +This will also delay the allocation of the free space to others +I will create another jira to implement it +Thanks for the tip +check for the case of a complex type with simple content +Actually we can probably just copy as is as Ive updated it with the suggested changes +fixed cast to +If it errors out with a timeout then it has detected a hang +I am uploading a diff file illustrating my thinking on it and couple of open questions for it +Patch is ok +It is set to false by defaultPlease review +I am not able to reproduce this in any of the normal environments I have tried +Thanks AndrewPatch applied to NIO module at repo revision rPlease check that the patch was applied as you expected +Fixes JNDIClassloading issues on endpoint activation +The cluster expiry codeexpects a correspondence between messages and which areassigned per message not +Quick patch to add fetchdt command +bq +Bug has been marked as a duplicate of this bug +Thanks +I think this because because maven does SNAPSHOT dependency update using proxyAll next builds at the same day will run fineNext day the same problem +Committed revision . +for the file based state could easily be moved to the WIKI as an examplestate manager plugin and removed from the have no idea why the original author wrote their own xml writer +Good idea +Mikhail I assume it fails for you +In IRC discussions this was tentatively slated to remove the JCR dependency as well +the remaining issues have been covered in other patches +No regressions were found after applying the patch. +Alexeyyour patch can not be applied to the current source since exclude lists are now removed from build xmlCan you please modify it? +Added to +Again want to add changes to the cwiki to fix the development doc to include the changes for creating a user +When generating a wrapped service style wsdl for the same service usingjavawsdl y WRAPPED u LITERAL lhttpfoo o one gets this wsdl generated which contains the correct return types +class file +If corruption is found the admin could change the toleration length accordingly +I need to know what URL will be available and then force wtp to squeeze it in +A checkbox for the poster of the comment doesnt do that + for the to trunk. +Thats why created a patch for itHTH + +Fixed +very interesting to read +I think its done now +Fixing the problems pointed by findbugs +Users ask me what is a filter query they dontknow exists and end up writing clauses that should be filtersin the query section +Thanks Paul Ill add some more testing in a bit +My fault +I dont know if that answers your question or not +It looks like the get was overloaded in +committed to trunk and +Make sure relative symlinks are resolvable at creation time +I suppose +cancelling patch to optimize for for trunk and just committed this +I dont know of any reason to expect there to be any movement on it now + +I misunderstood what you tried to accomplish +So we add new methods when those are requested from a userDo you agree with closing this issue? Or do you still have something that you want us to work on +Jason why dont you separate issue for the is useful for Solr even wo great idea! I opened an issue at +No response assuming ok. +APT has a hidden option encoding which is not used +I dont know why anybody can help with this issue +Fixed on and trunk +Verified in IR default config contains ESB +So I guess its up to the uDig developers to decide wheter to act on this one or not. +passes for me locally +made the supportAppends variable to private in datanode and added javadoc for process api +If we want to add new features they should first be added to the C messaging API and then reflected through this bindingAlso with regard to the location of this code +Will upload a fix soon +tomorrow I will this to branch as well +This has been completed +I just committed it +Refreshed the patch to latest trunk +Thanks Martin! I just did exactly that and it worksLets close this issue and hope nobody else ever encounters it +Maybe you could make it in line +committed +the workaround with disabling internal browser doesnt work unfortunately +Thanks for the report. +committed w cleaned up javadoc +is the correct size but I dont know yet where the virtualsize comes fromThis is all happening inside the SSVM so it could be a issue in there +Ill try to wrap this call in a composite operation and retest but I have been executing the add and operations atomically and it consistently fails as described only for add +Looks like an upmerge issue caused by another checkin caused the patch application to not succeed +finally +Thus some sort of SPI should be provided +If any different thought feel free to reopen it +no i didnt compile it myself +Screenshot about the domain creation to to improve the JBDS QE integration tests +Just build the war with the ant file the war is generated in the war directory +Would anyone care to elaborate on what this magical fix is +From your earlier description it didnt sound like you had a valid use case because you introduced too many faults +Also checked that there were no stale hodrings running after deallocation. +Looks like steps perform a fetchZip in +Jetty server is used internallyCan you please experiment a bit with different plugin versions? Ill try a default one tomorrow +Maybe someone with some knowledge of the like Sashin should handle this +Modularization decoupling interfaces etc +will commit after the tests +Previous patch committed at subversion revision . +or create the collection again dueSeems you can actually create it again we check existence vs the rather than the collections node we should still remove the remnants though +Please reopen a new one +What version of java are you using +pushed to master +The modifications made as part of this JIRA resulted in the loss of precision +Nit we dont need serial in do we? Other classes dont have this field +It ends up that the local DRDA experts are working on a preliminary ACR to do this +attached patch for on Tools +I am running regression tests nowTouches the following filesM javaengineorgapachederbyiapitypesRemoves the disabling lieM javatestingorgapachederbyTestingfunctionTeststestslangAdds a test case for this bug +Theres been a thread on the incubator mailing list about whether initially importing into a folder called private is correct or not +thanks! +This patch should fix it. +can we PLEASE get a fixHaving the same problem on on MS SQL +With this patch when executing the target a JAR file is built for the Castor examples +Thanks Robert for the fix +Google Chrome is an alpha product +Good work Tiago! Thank you for taking this on +There is a separate patch in the queue to turn this warning off until migration to the new is finished +Marking for we should try to to also +We need to be able to provide a model to the constructor of the component +Eg next time newer add it to svn +Does anyone object to making this changeI think if we are going to break it then we should make it an abstract class so we will have more flexibility in the future +My above suggestion would do away w the need of the flag and it would make the flag and it would make the check more particular being done per +Thus I figured that should be done as a separate JIRA given that its an actual behavior change +patch takes care of the problem +It would be nice if the mirror responds to Henks request before marking this as FIXED with any particular resolution +Hi Gil would you please be able to supply us with an SVN diff against the current code base? Id like to have a look at your patch but Id need something that shows the changes in a more formalized way. +Hovewer as I understand this is not the actuall thing the was on Your mind? So how would You like to remove the parent references and replace them with some other stack structureNote the attached patch already uses the Assert class attached here +Closing. +Hongtucan you take a look to see if you can resolve master mergethanks merged the code into please upate the review id with commit +I am this and will attempt to implement it and submit a pull request +Let me see if I can get some time to update with a more correct patch +I dont recall seeing any confusion about this on Solrs mailing lists over the years +The Felix SCR upgrade broke the extension the deployed bundle no longer shows up with implementations of and and therefore becomes unavailable to Felix Web Consoles Factory Configurations +I reworked the +thats great camille +I added test cases for sentence boundary and line boundary +workaround at the moment would be to move the relevant code into a service for the workaround +I tried to run the test case you provided after the patch and I dont get any exception +I should mention that unfortunately for the supported properties after ls l is not yet supported +Just a clarification on the issue title when I mistakenly used always +Query execution will proceed in a background thread +Moved to upon release of +If you can provoke it open a new report and tell us how +this patch starts the sequences at the sever start up +If this is not done by next release it may be pushed to the next one +Its fixed on even if its bad to skip tests especially when its inherited from parent pom oCode coverage and tests count values are in such a case. +It is not insane +matej you said we might want to put this in? even though it doesnt always work in mozilla browsers as pointed out on the list +In fact it was en error because the return value should have been declared with an exported interface implemented by the wrongfully declared class. +Recomendation One applied to please lets not close this issue until the two long term options are +Committed to and trunk +It just confuses people +Submitting to hudson +Committed in branchx +Sounds reasonable could you create separate issues for this? as. +i ran the test times +Fixed +Juraj is going to create an automated build for showcase in https post CR release. +Good stuff +To be included in +Im curious if this is still an issue +Thanks for fixing that one +existing commituserdev lists moved +This is still only manual support +Attaching revised funcspec +You can attach a patch to the plugin here and well review it +Patch applied in r +Michal could you please look at why resource optimiazation step fails +So if this functionality were to be put back in at least those who arent running Tomcat or some other platform Seam doesnt provide a specific hook for will have another option for installing the Seam EL. +However if we can have a short term fix that is easy to implement I would prefer that routMain concern is that if users will get an impression that new parser is not any better or even worse while also dealing with inevitable new bugs release will not be very popular +Retrying patch looks good +what other exceptions are you seeingwe dont want to trap OOME for exampleThe code already catches OOME logs it and attempts to continue +There is no indication in the regionserver log why RPC becomes unresponsive +But for reads I think the AL one should be the fastest +I have committed the patch including the changed memory calculation in revision Lets see if the test works now if not the calculation can be changed or the test can be disabled +I did download the latest snapshot and the error did not appear so feel freeto erase my comments if you like. +Should I resolve this issue +Closing tickets +Thanks applied patches with a few minor changes +Patch attached +Question answered +New files for patch for functionality +Update in kit Release Notes +reopened for fixVersion +add patchv for suggested release release note to issue +revision . +please do in future always a description in the ticket and a short subject +Not sure I follow the MAVENREPO thingy +Hi Andreas I have not tried the changes but from the stack trace the second issue should be fixedI have a question for it I might not understand the Axis logic clearly is there a scenario that the clients which uses Axis API directly would wish to read something from the connection if any problem happens? With current changes the connection is automatically closed for those unknown errors +Not sure if it is still an issue in applying Teiid to EAP for JBDS havent used that configuration for a while +If LOCALONE work for all types of strategy manual change the default CL is not needed +LarsOk for not going into +Conf files and trying newer broker releases and reopening if the issue still persists. +Hi StephenHow to override the default mapping with one in our own mapping file for the String classPlease help us to overcome from this StephenThanks for your quick tried using your code in the xml mapping still i am getting the above stated have to use nodeelement rather than attribute because some of the objects are not primitive type suggest the proper solution for the Just did a little testing and unfortunately it is not possible to override the default mapping for primitives and strings +Invalid wont fix + contains some code refactoring and it should sove this one +While I do agree that a lookup table blows the cache it still seems to be more efficient than running manipulations on a processor at least on those machines I had available +resolving this issue since the fix for this has been committed to SVN +topics of June is looked through +The +Still need to update the user manual +Committed this +Identified as a setup issue +Ah my load was broken +Looks a lot better! I did it the way you just suggested and it worked like a charm +Perhaps a warning would be nice in this case +please attach the generated by the plugin in target +Patch adding the feature +java jar cli httplocalhostserver +An added benefit to this optimization is that it would forgo any internal schema object creation or access of the global schemas +NicholasThe patch looks fine to me +Your issue is another one dependent of your particular case would be great to have more inputs. +Karel could you investigateCaused by Unable to lookup component cannot establish Aether dependency resolver +Thanks for checking! +No +Not yet +It might be necessary to check my findings +I think we need to fix this in all of the samples and not just Blog but of course it is good to have an example to follow so Blog might as well be the first +If I revert the part of this patch convertNumber works just fineRevision Date +It looks like it is a problem on hypervisor host either the VNC listening interface is incorrectly appointed or the firewall on hypervisor host is not properly configured + for the to trunk. +I ran and and both of them passed +Thanks Stu for your review comments it was a learning experience for me +I need someone from the pmc who will be listed as responsible and who will be the one handing out accounts to the committers who need it +When I look in the I can find most of those files hereserverdefaultdeploy additional jars can be found hereserverdefaultdeploy I look in the I can find most of those files hereserverdefaultdeploy additional jars can be found hereserverdefaultdeploy are the ones you need I could you try and enable autoconfiguration of a drools runtime called Drools in JBDS? If not assign it back to me +Thank you for clarifying +As noted above we would need to be able to determine who entered the OGNL expression but because of the way JSP works where we cant get the value before it is processed by JSP EL we cant +Rerunning through hudson +Out of curiosity when is planned to be released +This module would pack srcs which would required dependencies from manifest rather than generating a manifest +For testing purposes +No problem Ive learnt something new as well +This issue seems to have a far more profound impact than the others +The patch doesnt appear to include any new or modified testsThis is a build change hence no new have just committed this to trunk and +r +Since AJAXJSF relies on requestMap entries while parsing the outgoing HTML all expected entries are missing +This one slipped through the cracks apparently +maybe that is the reason +then created and deployed a simple process that sends a mail +This file is testexternalsruby reported via any other means punting to +Ill make that change +Zhihong JinchaoBased on v changes waiting forever to waiting for a time out and add the guarantee with caseRS is dead but zk node exists when master start so it is not considered a known server and its region will be assigned with out splitting logThanks for the comment +Assigning to Shelly to accordingly set the fix version and make sure the upstream fixes are pulled in +Task can be closedanonymous quick checkout works in IE without problemsCheers Sascha! +Well as I pointed out on the pull request the javadocs for even state its intended use +JochenWill you have time to look at this during the next couple of days +Thanks for help +This doesnt fix the hanging problem with is actually a bigger problem than I thought in the beginning +The problem is that before AM actually exists it sends a which gets processed by job which closes the and copy the jobhistoryfile inside staging directory to doneintermediate dir +No option to ignore anything like version or time stamp changes. +To be honest Bertrand +When can i expect that it will be released? and will it be released for? thing that i miss is that the username should still be autofilled when logging does not provide any kind of functionality you have to implement that yourself +Looks like theres no easily available DTD parser for us to use so resolving this as fixed since we now already treat as plain text. +So one possible workaround is probably to use gf instead of g +I noted this issue is marked in for but it was only committed to trunk +Does anyone know why JAXB is failing to generate the client stubs with right targetnamepsace instead of empty string +Thanks to Chris and Colin for the reviews. +Please ask questions on the list +They are more concerned with the queue operations +The current semantics of offset are such that there is really no point in specifying it ever as one can just supply a smaller bytes value to get the same effectIf offset is simply substracted from then offset must atleast default to bytes or otherwise the result is always eof +Seems weird that IE handles +Anyway lets postpone package renaming till later +Im worried though that we abuse the semantics of the API and it will be more difficult to fit both functions in a single API as the functionality evolves +Closing this issue as a duplicate. +However you point is well taken that we probably dont need to maintain client cache if we can set CXFs in the Request Context to override the address +Thanks for the request duplicates the request in +Thanks James +But the Groovy resolver need it to compile the groovy script inside the project actually it creates compiled classes from the groovy scripts but they are not persistent +Just curious if there is an easy additionchange to that would make it fail prior to the fixNone that I could think of unfortunately +jbpmplugins + +As the error in the build console isERROR Processing failed due to a bug in the codem going to restart the build to see if this fixes the issues +Integrated change to always check that the annotations passed in are qualifiers +Any update on this issue? I have a situation where I need my application code to have access to raised when flushing SQL to the DB using a transaction synchronization +Seems like you are making case that it stay public +would you mind running some of the examples using the new schema version? after your confirmation we are good to go +Thats just odd +If the query does not have a distinct use aggregation with job +Easy enough to make the change +Can you provide a more complete example of what you are doing? A unit test would be ideal but just a small example of how the pool and its connections are being used would suffice. +Im marking this as resolved since were cleaning today +So just expected that everytime a skippable exception is encountered the skip callback jumps in +This patch adds a new command line parameter to the binpig shell script and includes the appropriate jars for pigs consumption +Ron could you please attach a screenshot of what you see when you open this file +This allows BOF to use the API +Ive been thinking docid instead of id since its further from id what we normally use as the unique key field for documents +I recommend sending an email to user list with the details as the branch is obviously compiling for everyone and JIRA isnt a support forum +Attached patch v that addressed review comments from Andy and Stack +This file tests the implementation of in a number of cases +Thanks Eli. +From NateI think parameter inference works for methods but not was because Pair without type arguments is not a type and so theconstructor lookup wont work +Patch for trunk +a JIRA in Karaf for this you dont setup the gpg keys in your home and key password in your maven setting file mvn gpgsign will be hanged for ever +Good catch +as the test class requires JDK and thus does not compile in and when the proper JDK is used +triaged for Without more information as requested not much else can be done +Committed to trunk and +this is also easy and necessary to fix issue due small architect error will you delay the fix for years +The same goes for +I will make a suggestion +I have put fully qualified name to etchosts but I still get the same error but in error message the host is not fully qualified +Just in case anyone is following this thread this change did indeed allow MINA to process HTTP requests nicely wo a content lengthThanks again Trustin! +Can you please take a look +Verified hat it works with build +Ill be posting an updated copy of that for trunk this evening and the version will follow shortly after that +Just being pointed to this jira while I was working with anirban on +please move back if they will be worked on asap for +Yes looks prett good to me resolving this issue +Attaching a patch here +Before committing I would be interested by a feedback from Suraj +looks good to me + renamed and to and respectively +Changes look good +You may closethis bug report or implement it as a extra feature +The runtime plugin isnt started because it depends on the plugin +Im not quit sure which version use but it seems that use and clouder cdh use make sure you can find correct version jar in your hadoopsrchomelib +btw what are the plans for the capacity scheduler +I will probably ask for the review on the staging site rather than reviewboard. +Ive also noticed over the past few days that it may also be doable with existing functionality +At this point since this issue is closed any new failures seen in these tests in versions that have the fix should be logged in a new issue and canbe linked back to this one +Unfortunately the XML file takes precedence over the +No response assuming ok. +Actually we are using Application in one server +RegardsUlhas Bhole +I updated the patch in the attachment changes include sesskeyfilename ticketkeyname release the context associated data when ctxs reference is add a little bit of parameter description in +NET compatibility and removing this limitation if the DLL is used in a +Reviewing its code tells me it works the other way directoryscanner scans every file and see if it matches with one of the +I revised my patch + +I have reviewed the patch and I do not see any problem +nothing good will happen in that scenario +As I suspected there were still some failing tests +You could avoid this by naming each Timer differenty example Reminder Reminder +This area could also do with a bit of updating since its all a bit inefficient creating a new array on very request and synchronizing every request on an effectively immutable object +This is exactly the same patch as +QE please errors in the java script consoleBrowser Firefox and Chrome both at the latest Fedora JDK openJDK Server AS thank you for information +first series when this issue was reported second series with improvements implemented so far third series same as second series but cacheSize for was set to more performance optimizations +srctestjavaorgapachehadoopyarnserver +The exception messages was introduced by for which the patch contains several of theKnown JCC Bug FAIL Caught exception I agree with John that we should avoid in the master file +Pushing off to release +Attaching a version of the file that fixes some of the introduced compiler warnings +Patch applied at BEAN module at r. +additional cleanup + +Found out that I caused a regression with BMT +After some more testing I think that this patch is a reasonable fixAttached is which changes the compiler sothat it creates a dependency from a statement with a FOREIGN KEYclause in it to the table which is referenced by that clause +Please attach a runnable test caseDoes the same happen using or earlier? Im wondering if this is related to +Thanks Devaraj! +Anyhow we have the additional changes to fix the tests Ill update the pull request tomorrow or sooner if I get a chance +Fixed in commit ff on master +Need to replace existing operation detachCopy with active tx with two operations detachCopy with active Optimistic tx and detachCopy with active Datastore tx +for the code +Actually we might want to do a release to fix the problem for servicemix. +It doesnt have to be completely synchronized with probably its not even purpose of ESB Configuration Diagram should to give user some overview about this configuration and maybe it should be used just for graphical interpretation and not for editing configuration itself but with added value that it will at least generate Icons for particular nodes and user will arrange them and connect together dependent on processing issue with a graphical ESB editor at the moment is +Since the class is not found I assume you forgot to add the jar +Let us know if you run into any issues with it +Thanks +Is it absolutely necessary that these links be added? Would it be sufficient to supply the jar files in the request as I have done in the pastThank you in advance for your response +Updated in trunk r with the following changes add artifacts file types so it gets processed by the consumer added unit test +Thanks Tanping +Thanks Nicholas for review +I committed your patch +I updated the generated source +This just isnt a high priority for us +Milind thanks for fixing this issue. +This is a good thing to cover in the reference documentation for each individual provider module +Fix included in the release + +Duplicates still get error in snapshot build for Tomcat not for Jetty though +Fixed getProcessor is generic to during tidy up prior to +The array ordering is preserved +My little fix did it closing this issue. +I think that in addition the documentation should be expanded to make things a bit clearer +Closing the jira as per the clarification +Patch file to fix the compilation issues +Also it seems like the dependent task you mention is actually progressing well +Defer all unresolved issues from to +that causes one test fail +Cant use rwsetlocale object because of destructor +a gt one and a gs one +I couldnt find an existing example so I could know how to fully create the test case so it would automatically run when the test suites are run +Ok I think I know whats going on are you restarting Apollo without restarting the client? +If you have time could you run the repro of few moretimes with nio disabled and post the logs maybe give them different names to make it clear which is which +Committed to trunk in revision +Tim please run the example! If you look inside the build file you will see that the ra is installed! I have run the example both on the same server and on two servers and there is nothing missing +Rebalance happens during consumer initialization + Ive got a psychic appreciation for which project you wanted. +Please follow the at the same location +Arent some of these fixes in +Other small It wont require anymore changes +When the connection is closed I think its more informative on the client side where it tells you that the connection is closed rather than just saying the statement is closed +Issue closed. +Closing. +Closing all tickets +Perhaps you did it and forgot to close the JIRA +Log a warning message instead of throw an exception when cannot mkdir +Im putting this in as a new feature +With we should be able to quantitatively demonstrate advantages of Trevni over and +passed locally +Maybe We could remove it when AS is released +Can someone else confirmIm using Seam on Weblogic but I think the rendered tag was working fine with Seam as well +I want to support this on webconnector but Im thinking of better solution for a whileI agree especially where there are browsers that accept the bad +Related I filed to make a class so the types are more readable and the code is less error prone + +But thats a problem if you want to mix DI and normal dont use SI at all yet I think that an Exception for an unstatisfiable Setter is wrong +I dont see the preference +the actual problem is not that the test should pass but that it does pass +I could not find an easy way to dig up this information in the existing context available to the code generator +fixes one last mutable static but is dependent on johan fixes a broken locking idiom +Changed version to CP fix will be verified when a new version of JON is available +Cant we simply add a property to choose whether plugins get installed in or? +As I understand it the vacuous wrapper and query object methods will be filled in when you submit patches for and +However at present my belief is that of what needs to be said is independent of whether you are using AS AS or indeed any other bit of Java middleware +Modified to prevent double increment of the count that keeps track of how many delegates responded in a parallel step. +I do not think removing Android is correct here +marcus checks whether there is at least one method with no unchecked permission +The reads would not count as cache misses +Have the similar problem with and on on Windows XP +The source directory should contain the already unpacked archive regardless of its typeHowever in my case I am building a war file rather than a jar file perhaps that matters +The following is a list that seems to be skipping the new changesibatorgeneratedinsertibatorgeneratedinsertSelectiveibatorgeneratedupdate the my last comment I see the setting I was missing +Retaining the Page Design Options dialog for external files will demand appreciable time period and efforts +Ive just committed small changes to make it easier to use the build target now looking at how to actually write the test +Given this implementation context probably OK for firstintermediate request and CREATED probably make more sense +Tempted to close as not being a Tika issue but thought there might be something interesting for you to look at firstThanks Ken +Then maybe it would be nice to change their names to +Attached tested fix and if confirmed to work and not break existing configurations +Hi Steve I imagine things were busy these past days with the release +I just committed this +code looks good +I have to change aThanks for the report. +Everything looks fineThanks by Boris. +This should be in the next BRMS build and the SOA ER release +The Guava dependency in apparently came in with +Do we know what we need to do for this JIRA? I can see in we need to be able to remove from failedDirs and be able to recognize this fact in service +issues + for the patch +Also this is is apparently used frequently and quite critical as for full blown stack +I have committed this. +Committed to branchx in rev +I hope to finish testing today +Even then just a matter of taking those out tooWe dont necessarily need source and Id favor not incorporating as source and not modifying the package names +Searchable does not mirror GORM changes +Ok this guy is updated to trunk again. +Still getting the same errors + +Tech part done now the docs +Our load of the library is acceptable. +Closing issues for +Thats the question because there is no such behavior for all other resources chunked lists attachments shows views etc +Perhaps had the patch been uploaded for review as part of the RTC process thats been agreed upon this would have been caught before the commit +I committed this to trunk and. +Looks like DIH should instantiate separate per thread +I think the problem is less likely to happen on a fast machine since it has to do with speed of ajax are scenarios Client ajax event Server Server ajax response Client Client submit Server Server insert +I think this is not necessary as the authsupplier can still decide if caching is needed +the patch looks good +Patch v from review board. +has been marked as a duplicate of this bug +So what you are suggesting would work fine in most project setups but in the occasional case where the package names of your generated files does not match the source location of the +Those methods do a lot of type conversion that are not possible using just generics +the factory is successfully bound to JNDI. +even though i like the fact i could quickly search through all the relevant tablescolumns +Issue already resolved in trunk by Jason! See revision datetime marshalling of issues solved in released versions +This patch addresses the previous iterations Hadoop QA findings except for the use of sun classes +This was chased up as part of the investigation +Verified in ER +Patch attached + +Due to the number of options and differences in the way each platform is configured its quite verbose and complicated to provide an abstraction on top of the existing implementations +It looks like was the last correct version +Patch by Gary Dusbabek +Since the constructor could have done this since I missed it because extraInformation as used here is not documented with its purpose +The conduits were already checkingusing the objectFixed in repository revision . +Please verify in trunk. +With the introduction of XML based configurations using in we no longer have a serialization issue +We will continue adding tests for the release as well +fix the help message for df h option +Updated patch to fix a compile error in sorry about that +is belonging to the ADF +This was introduced in +Nice one +trunk Cott can you test with latest code on your system +srcmainjavaorgapachehadoophbasethriftgenerated +If by design then OK close as resolved wontfix but it does violate the Principle of Least Surprise +Id say we only do it once because unlikely well succeed if we just did sockettimeout twice because it is too big +However that filtering couldnt be used due to some CXF requirements +Ive reproduced it twice at loops and +Of course he rocks +Since I have no possibility to test on a windows environment this week can you please apply the patch and test the patch does not work I will remove the test and dig deeper in the next the for ease of reading I did not wrap the lines +If thats the case then you should throw some type of error +Please let me know if not +Is it possible to add a check if the pde is true to set exportedThe workaround right now is +Bugzilla was not fully for Xindice +Ferdy is right custom counters are more transparentPatch for +failed unit tests locally they all passed +It also fixed a minor error in +sorry it seems my method copypastes got garbled +trunk works fine thanks marcel! Anyhow on mac things fail easily when things like +Thank makes sense and I agree it is is a real and likely risk +A boolean is not a number + +I like the idea of including the snippetstemplates somehow to handle this +Patch committed +This is required to support more guided flows like Create a Java EE application which would walk the user through +latest patch looks great +That said perhaps we can say that you dont get traceability when you use shell in those modes +Reproducer attached +I had missed a plugin to add the jsp source +Code reformatted to Lucene style +Patch applied in r +Closing with the release. +Committed to and to trunk +Micha have you tried the example I have given? Im using cfoutput for both loops not cfloopAlso imo you shouldnt say an issue is resolved if you cant reproduce it +I agree that breaking the interface is a problem +Thanks much Ted Kirby for this patch! +Defer engine config until first proxy connection is established. +initial entry of and is complete +on trunk patch +bq +Do you get this problem when you run it under a JDK +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Oleg Agreed +Arun any chance to one this one? I was trying to use MR via maven and they failed misserably +Describes new format format and the default comparatorordering of within a +Im not sure if cloning the issue was the right thing to do but I guess if it wasnt I bungled it up +Would storageID and cTime be preserved? BTW do you know why cTime in my test case above +I am attaching patch for conversion of lang to junit test +does this take into account circular directory graphs with symbolic links? is that scenario tested +on +the last problems may require the parent product id to be stored the orderitem entitythanxFred +No assignee +We have the stuff in order now so we can refactor the build as an independent issueIm on interrupts next week and wont have full control of my schedule but will try to prioritize this as time allowsFor automatic snapshot publishing what should be our goal? Every commit nightlies etc? Im thinking we could just have our regular trunk CI job publish a snapshot as its last step +JacopoThank you for committing this patchI was thinking exactly the same thing about the widgets and I am in the process of removing the styling attributes now +Verified with JBT. +For us this is not a wish but a bug because we created a form with input controls and we want to give a user the ability to navigate through the form using the keyboard using the +Did you file an issue with FindbugsIts a good compromise for the new snapshot +Please include an example JSP and what was extracted in and what you now get with and a list of the problemsThanks +Thanks for catching this Dan +With the plugin doesnt check for inheritence anymore which causes the regression +Fix this up and deploy on next bug fix release +It is +I think it makes sense to do the same also for repository implementations +However I am on vacation for days after today and I will be out of town for the next couple of days +CoolJust a question did you handle the case where the EJB itself is secured in the JEE server? If so howI think this component only solves the case where the jsr endpoint is secured by servicemix right +I think reveals the cause for this issue +We should allow builder pattern by default when using the set property IMHO +I think it only happens when openning a t do this often +Not tested on Windows +That said the targetNamespace really would not affect whether the message looks like one or the other +Regardless I already contribute quite a bit to incubating Apache CXF todayfreeing up time for Dan Kulp to help out more on Maven so it is not to have me studying Maven code at a cost of contribution time to the CXF project +looks great Sijie +Bulk close for +Asked Paul to start using the form in + +See +I think at the moment we just run the package and test targets +This is something I think it should be improved sometime laterThis patch allows a peer to get a list of all other peers +But it did catch syntax problems when there were some before going to OOM so I assume its mostly validated +so unless anyone obejcts ill commit in a few days +Seam Ref Guide fop updated to version screenshots for pdf version are fixed +The issue of it just hanging was on my end +It could be improved +I dont understand what you wantTo create dynamically the defined in the application +When I untarred the file I saw no README or other indications how I should install or configure +They will have to readjust the column ordering of their queries +I agree +Switched from a vdb property to a model property +Patches is welcome and it would be great with unit tests +works nowstill dont have optimized fully so there are still some full package names. +Generated or publicated WSDL file with chars are publicated in native encoding. +By uploading this I was not expecting it to be committed +Thanks Ron +I also wrote an extended that is able to reload itself +Incorporates all the review comments +I looked at the skipBytes code +Committed the changes in the file into SVN trunk with r and into the branch with r +Some kind of test that completes is better than running something that is always going to timeout +This code has already been reviewed as part of +That fixes my testcase +I changed the title of the JIRA issue given that the exception thrown when the friend request link was clicked on was fixed already in revision +I could see that a small wordcount program took seconds without the patch +Closing off +Can you have a look at this? Maybe this is already reported upstream +Find attached my little attempt to implement queuing +Check the permissions for the project please +Oh how annoyingIs there a possibility of starting to use the src directories for these kind of patchesJust because some people are stuck in the dark ages doesnt mean we cant shine a light +will reopen if and when it can be reproduced +Thanks for your response and adviceIplan to use job and job to instead of and use the cooccurence arithmetic to calculate recommendation for items had ratingcould you give me some advice +And we have passing integration tests +I tried to cover the maven plugin tests the compiler tests as well as the readerwriter tests +Thanks Aki for getting it resolved +Well revisit the topic for Spring but very likely exclusively with as the foundation for modularization is a great technology but we are a long way from using it here and i think we are not alone in the Spring community +Reassign to a release that you feel comfortable resolving this issue in or leave as is +If an interface is specified it must provide an interface which is the same or which is a compatible superset of the interface declared by the promoted component reference +Just be sure to also check that the tokens are in the current users UGI +Mass closing all issues that have been in resolved state for months or more without any feedback or update +quick clarification were you installing the version from the grails SVN repos? or from what is Githubi installed by grails pdf +Attaching patch generated against hcatalog inside hive +Server emconf has invalid JIT naming so monenter will not be processed +Patch to and accompanying state model to test this problem +While I was in there I changed Derbys copyright claim in srcdocumentation +Its much easier to talk through that than reams of code lets avoid a situation like the one we have in +That said its also by definition so we need to document this that documentation should go straight into the annotations javadoc +Yongqiang Please hold off on committing this for a day +Thanks Owen and Johannes. +Ill update the unit test guidelines when the patch is committed +Attaching a simple patch +As we dont ship the examples of Apache we should excludes those features from servicemix feature file +Im in favour of accepting that will not support parsing email headers with CR line endings if the cost of support is too high +I would like to have this warning for value types too +After applying the patch doant If we want to keep the ivy support we should add a dependency to the compile target +All current of the pass +Sample pngpgwprj same file in geotiff format works just fine so its a format issue +Should be this one +To load a toMany relationship there is normally a foreign key corresponing to the primary key in the owner class in the where clause +cultivated landArable land means land fit for cultivation not cultivated +I modified the pom files to only use source and target for the axis related stuff everything else uses by default +RequiredBranch from trunkTrunk becomes line +Here is an updated patch +JDK on Windows NT and JDK on Solaris UNIX. +Thanks Vinod. +finderName find field reference jpa class +Hi Jean can you take a look at it reproduce it? There is no need for any special load just have the web page open on for lets say minutes or an hour or so +I think this is client side issue not the server +Fixing log message and uploading new patch +No regression found when running the test suite tested on simple java single or multimodule projects without issues +On second glance these look to be valid if the incoming hash is empty we should not raise an error +unbundled modeshape and teiid runtime zips and also changed the used for installation to now copy directories +fixed as of r and tested on postgres and with uDig closing all issues that have been in resolved state for months or more without any feedback or update +If there is anything that needs testing please let me knowOnce again thanks for making this happenRegard feature freeze date is November th +Example of the +proposed patch fixes the immediate problem. +At some point we modified Time to support tojava to the various Java date types and to s. +it would be displayed as a custom type or +Thanks all for the review. +is notorious for its very very slow insertion speed so I couldnt make it faster while keeping the reasonably big anything smaller like kb the output is instant. +To ease our job what exactly is missing +I use it since Thanks guys +Karthik other developers had a handful of design concerns about +This is a good idea but it has few complications buffer length should be stored in order to be used by reader reads should be aligned by that buffer length so we always read a whole checksummed chunk of the data which implies that we will potentially always need to read more data on each requestThis seems to be a clear tradeoff between using additional space to store checksum for index columns for each row +What is correct +Awdesh Singh Parihar for your patchIts now in trunk rev +RobinI just saw your commentThe reason is that this is an online learner +Having an HDFS or Cassandra cluster to backup a flume pipeline seems like overkillOverall I dont think adding a channel that writes to a remote cluster is a good idea and is unlikely to be used +The attched patch implements the assertions above +Ive applied it and extended the test test case to cover it +Here is a piece of code that solve the problem mentionned above +Please use the nd attachment not the st one +moved from to CR +Patch applied +Unfortunately they havent provided any ETA +Patch incorporates Hemanths suggestions +srcmainjavaorgapachehadoopyarnapiimplpbservice +This code is an exact copy from Its about letting the model know that the collection has changed modelchangingchanged doesnt help here +I added check for fsnull in get and it throws if it is nullRetries on failures has sleep for secs already +Here is patch +I did some testing on and that version has the same are right +Strong are you sure this was resolved in EAP +Additionally I feel its just incorrect from an organizational standpoint a the is not a sample its a new featureb the Samples package is considered a production level package its part of the official release it should not be a de facto beta packageIm going to register a new issue about this +I dont know anything about BKJM so not going to work on this. +Vasily this afternoon will do itAlvaro +AdamNot really a convention +Now that work is well underway with AS all previous community releases are +I dont have real payflow account to test for you +Corrects the argument swapinversion and also adds a call into the during Specific decodes that allow the resolved parse stack to advance for the contributionIm not familiar with the C implementation so will leave detailed review of your patch to others +Move to Beta +Thanks +Not very convinced about the description but I seen it used like that for other services address properties +absurd +Maybe the solution is to make good use of it solves some of the versioning issues allows you to easily register more services +in branch +It seems like it would be relatively quick to implement even from scratch so someone might beat me to it +Sounds almost like someone added a keypress event listener as a joke to fool with us +any update on this? looks ready to commit to and +Lets leave it open until next release if nothing comes up againI think we can safely close this one +Bob a blocker issue means that all Sonar users are impacted which is of course not the case so Im decreasing the priority of this bug to major +That error normally occurs when your tests did not run properly due to a syntax error +second file test same time +Perhaps it would be even better if we left the arrays and built the bit masks from them at initialization time +Implemented changes described above in Rev +A first release attempt failed because I was missing the LICENSE and NOTICE files for the source distro +Straightforward patch +It broke the js unit tests +It works only with which cannot wrap m still open for suggestion +But when I am using Struts in Firebug the result is too and the handle of my Ajax call is error not success +Id like to have bundled features when we add this portlet +Added the build directories so it will work from a dev environment as well and committed. +The broker does not seem to get the request to close the connection which gives the broker the impression that the connection is still active on the client side when in fact it was closedYou can see the problem by just doing the following create a connection interrupt the thread close the connection repeat over and overopen the jconsole and check the thread count on the broker +Yes everything is on drive D +Patch applied at r thanks for your contribution please verify +commited the fix on trunk +fixed the issue +Rather than writing this project I would write the name of the project again so that the user gets remindedthat he clicked on the correct link +This still hasnt happened after was fixed so Im closing it as a duplicate +I think it is a good idea to disable this test when Clover is enabled in patch +The test was trying to additionally deploy when the the archive that is part of this test contained this +Then user could look at this out put and parse it to get the info +Thanks a lot FlavioIMO this is not required in as we did not decide to release any point release on that yetIf any one expecting this to go into I would be happy to create a patch on as wellOtherwise we can mark it as closedThanksUma +Thank you Po. +Id like to remove from unit tests +I am thinking the best long term solution is to write delete markers for the METADATA table into the root tablet +Based on this refactoring I am now marking this issue as resolved. +Any insightsthanks did you try to run tests for longer? will memory usage keep increasing and even with wil it eventually cause OME? or the memory usage stabilizes after initial ramp up and stays the same irregardless of how long tests are rundid you try to do thisthanks memory was not being released even after the runs ended in my tests which i thought is the issue +Is it possible that the wrong zip file was attachedEddie +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +concrete data points +Attached screenshot of MAT leak suspects +patch applied with open version range how can I get list of plugins which has to have updated plugins have dependency on plugin and doesnt have version restriction for plugin is dependent on usage plugin but is not dependent on runtime plugin so probably manifest file is correct +The followers are taking to long to send an ack back because they are taking an additional snapshot +Punting +Please check that the patch was applied as you expected +believe this issue should be resolved now that the patch has been committed. +Here is a source file demonstrating the problem with comments describing the behaviour I observe +all works as expected +Therefore I closed this issue. +What code in Roo needs to be changed? Version is handled by the you can provide me with some fix I will close this +Thanks Oyvind for removing +Hi Thiago I take your points but just for the record while can be overridden currently the same service is injected in all cases +Patch applied thanks for the update Rashmi +Sent from my mobile device +I think you should be only adding the libs as a dependency to the pom files and building some tests +Has this been fixed? has been released as of today +Attached patch borrows the code from so that the timeout exception contains all the thread stacks as well as potential deadlock infoNo new tests included because this is itself test code +Hopefully it works with your state of source code +Automatically via the container startup event +I have an alternate implementation that uses the package +Will do +I think I go for +The loaded dump looks fine +I am closing this one because this does not happen anymore. +Try explicitly setting it toapplicationOops I see the trace output is already setting that header so ignore my like the pull failed due to +Forcing to use TIMESTAMP instead ABSTIME for properties for Postgres seems to solve the problem +this is useful for things like empty stopword sets so we do not hash every input token unnecessarily +In our case we are exposing an EJB Session Facade as a web serviceWe have limited control over the interface we are exposing so renaming the methods is an annoying solutionI dont see the problem with being able to tell axis exactly which method to expose rather than which method name +Using environmanr variables is a not very good practice +back from vacation +Can I submit my patch? Or something else is needed +Now using protected instead of private whereever appropirate +The validate method has been added on the deployer to allow custom validations in addition to the basic endpoint validationMaybe we should not return a boolean and force the deployer to throw an exception +Moving out of +Im not sure if its the best way to go +Simplifies the Service Level Authorization to check directly rather than going through permissions +Thanks Hari +Update to the previous submission +reopen to add derbybackportreject label +removed notes about supported vs available removed common as it is not a skin removed because it has evolved to form the basis of the views plugin. +Your reader opens files once and then never ever does it againBelieve me this is so pointless +Unsetting Fix Version for unassigned issues +You missed fixing the bigger problem this benign exception took down the entire application +The management server starts up without any issues nowThanks for the quick fix +ZK as the truth and assign nodes accordingly as and when they join the clusterdecrease replicas automatically using the replication factor and max propertiesWhy dont we have explicit names if that is the objective +Heres a patch that documents the JSON encoding +HiWe are using sonar +Instead of adding a new we can just use the that is being worked on the other JIRA +Making the changes that I requested to push it over the line +This is the origin of the issue +However since most of the +I think the problem is in looking at Subversion this was changed to use classes after the release +UNLIMITEDFIELDLENGTHI deliberately named it like that +When unique constraint was backed by unique index distinct scan was used but after making it non unique constraint this was not the case so the results are not ordered +Our engineer who saw that behavior is out this week so I dont know all the details but it was why we tried removing the flush +Any possible fix for this issue? Thanks +Now that work is underway with AS all previous community releases are +Petr could you add similar paragraph how to get EAP module like upstream documentation has gotI mean we support that delivery in our Maven repository zip info about the maven repo zip and the modules zip and what to do with themMight be overlapping with Rakeshs. +I am yet to do this +New log file running against EAP build from hudson +The only suggestion Id have is to call it or some such in case someone wants to calculate actual distance +Can you include your changes as a patch +thing immediately strikes me as wrong is that read operations on half closed connection keep on returning read bytes instead of expected +Thanks +The patch wasnt applied to due to whitespace indent issues +I think creating a small index with a couple of docs would be the way to go +Hi DavidIm working with Bryan and he delegated to me some of the stuff related to Thrift +The REST representation predates when was introduced +resolve then to QA +Thanks Remus! +Ah very cool thanks for the hintSo the best solution would be to use Nutchs plugin and use that language in Solr to remap the fields? I suppose this is better than plain Solr language identification as Nutch has access to and to identify the language +I realized saminda removed the and added to Axis Configuration +Hence closing the bug + +Issue was resolved even though theres still a remaining issue with printing out the unicode char as its typed +If you try to create an that combines reflect and specific schemas we check the initializer and then throw a fail fast exception so that your job wont fail on the cluster +I have contacted them and obtained a copy and evaluation license for version of their have integrated the server into Spring testing suite and managed to bootstrap the platform but there are some issues with the manifest parsing which unfortunately is a blocker and is thrown early am also in discussion regarding the distribution of the license and the mocks required for compiling the platform +Also cleaned up some of the language feel free to change anything you dont like back +I understand if you want to fix this a different way but Ive attached a patch that corrects the a bad In message and added a test case to verify that the behavior was broken and is now fixed +The pdbs are now generated and put in the zip file with the sdk. +And also a full stacktraceI cannot reproduce the issue on either osx or windows +Factory is surely a place to implement such mechanism with filtering facilityAnd i also think should be registered as OSGI service for every component created like does +configure +Ill close this shortly unless people object and want to use something from here +If only ive put setDebug sooner +Sorry I haven any further details on this issue +I compared the previously attached patch with a regenerated patch and there were no differences +However that can be decided later on and I have removed the configkey update with this patch +Richard can you confirm whether this is an issue with the nightly build andprovide another patch if so +This is to avoid the fact that the listStatus implementation differs between Local and +Youre right it may be weird to allow CVS action on a node containing files which should not be in CVS +The solution does not work for jsp and sun ri was also a fixed problem while restore in the JSF variant +Thanks +Thanks! I was able to fix the template +A combined patch would be nice +I will address the unit test failures +Verifying this issue +why user should see which el resolved by editor not by runtime? Absolutely useless information +Thank you Roshan +Javascript error due to reorganisation +I think the trunk build has the same problem Ill verify it soon +Ive also tested on a secure cluster to ensure it fixes the corner case +Thanks Mike I added additional tests and ensured the missing stuff in solr is fully functional +No problem +Thanks +I will be dealing with newer versions of Xerces through a issue +By reading I found that a random number between snapCount and is used +gitignore file in it because the executables are being generated in the same dir as the source +Do you have the in the lib folderDo you have an activemq schema in your dbIt works in my Windows XP Java environment +Anyway this java doc seems not very helpful +Made it critical +Please verify that is resolves the issue +Should be lifted into its own subsystem that doesnt depend on EJB +Having said that perhaps it would be nice if the code allowed the alias to be omitted if there is only a single entry in the jar +Patch committed to trunk and +Closed. +Cant reproduce +so on a release nowMonday if people dont mind a new release popping up a few weeks laterThe fixes in question do make some config changes so I do like the idea of doing a release now then committing those changes and letting them sit for a couple of weeks before putting them out in a releaseAnd we agreed to make a now without the inspire changes and make another at the end of the month +After glancing at the code I dont think this is specific to +Need the url of an svn dump oftheir sourceforge repo +Even if it worked this would have been a bit delicate and probably a terribly hackish solution since implementations wouldnt be able to properly use the object as they might expect so I guess it does need a rethink +Changing affects from to so version can be removed +Which version of Mozilla are we talking about +quickstart modified to test EJB with one method requiring auhentication policy mysecuritydomain used in the given solution works nicely with AS but with AS it does not +We can leave room for more opinions on this before having a final go +Please try to catch such issues yourselves the next time on +Committed actually causes a bunch of problems for customers that have used longer the character keyspace and cf names +Attached is a patch that removes the use of the classes +Is there anything architectually preventing us from doing that? If not Id like to see it and recommend a change to the spec +Is messaging pulling in another version that it depends on? Your jar is marked as +works as which talks about adding PCS. +Do we know which tests will behave differently with or without native library +Reopening Ivan did a review of this patch and suggested that it might be split up into three smaller patches that would be easier to understand +Now it uses the default replication factor +Patch committed to SVN trunk +Unless there are specific problems that where not addressed within the original JIRA I see no reason for keeping them in this state when most of them where fixed and resolved almost a year ago. +However I was given no context requirements or use cases to justify the problem or any details on what he was trying to achieve and why and whether his approach was appropriate as my previous comments suggest +I will addbq +need to look at it down in ipc findbugs findbugs make accessible from other packages move to and into io package +If not we can provide an autoConnect option but only for connections that dont require credentials +I want to make sure that the way your doing schemas in the parser and the way Pi is working with them in the type checker line up +But this is not related to this exception anymore therefore resolving this issue. +FYI I am seeing this problem with Seam and +I think its better if we can make JPA layer vendor independent so that we can plug Eclipse + +Thanks for raising this! Ive updated the manifest like still looks correct +Can you please suggest a workaround and sample code to work with? +Thanx +Ok in fact I did not know that it was intentional to log this dump stack in debug and I have no problem with this +Seems you found another issue considering the position of the enum andor missing are moving the enum to the bottom of the class or adding a after JSPIve added a test for this issue in strategy has been improved +Well I couldnt reproduce the problem on Geronimo +Proposed asynchronous mode I prefer the JMS Queue approach +Superb may Iwe commit and then continue with the release? Not trying to rush anything delicate here but also wanting to make sure I push through to finish the process as soon as were able +done +Thanks for looking at that +Thx a lot +If its an overkill I am fine with documenting only +I have one week old code +emacssym +Remove it? Why not just disable it for if that is the immediate problemRollback seems like a useful and traditional feature for me so I dont quite grasp what exactly you mean when you ask Do we really need it +Well I put the work in to allow +I would concur with proposal aboveImplicitly that has been our assumption +Looks great thanks! +All tests in now pass and the testflip is enabled in the index. +Now there is a test but its been disabled +I will handle this for and will also check some of the other new adapter modules +See the for the configuration files +will always be used +Jochen +Otherwise any data access has to be syncronized or +It looks like that not enough test coverage is present in this area +That would be a great Jira to fixIm open to fixing this next +A similar problem occurs with describecoverageoperationsdescribecoveragerequestpostxml where a version is sent see also +subdirectories are common here in Not sure if this is supposed to simply be warned against or simply not allowed in any case the ui should handle the error or simply just work +If does not solve the issue well reopen it +please commitI also ran the and unit tests on my machine they pass +Is it planned to fix this issue +Ive just committed this. +I checked the SVN history to see if there was any changes in that can it looks like this section of code has existed for a long time +Some failures seems that the output misses some lines when using keyword explain +I added my code and It is a combination of the work done by the team +To go further I think we should improve the to store more structured information like inclusion levelsAlso some Sling components do not provide info to the this needs to be improved as well +Fixed in the nightly build +superceedes this. +I wont be able to execute it over requested debug info trace file and the sample created a sample project for you and this has been created using the sample the sample project I removed everything but two modules besides the ear module +Suspecting that package has some kind of affect +The page is onAs a test I copied the collapsible list example pasted it before my list and the problem shows up there as well +To fix the issue with your sample will not be easy +The included patches adds the jar as runtime dependency. +Attaching +Thanks Doron that works! +Invalid URI missing host port looks patch allows hdfs uri and uses the default filesystem for this +Of course as was already mentioned this could cause some other problems +And there is also the possibility to ask the DLTK community for adviceA hotfix could be that the tries to close all CAS Editors just before Eclipse is closed +Also the CTS makes use of namespace imports which is a blocker for question remains why has this worked up to now +Im currently not sure how to achieve that we start with the oldest state to remove if the memory shortensJust a hint the amount of states to hold doesnt depend on the jsp file count +w changed to. +I never use sourcegen +Tyler can you verify against branch +So split the success and the failure tests +Closing this issue as a duplicate then. +If you could modify my attached file so that it use and certify that that configuration is going to work on Oracle gR Id like to change this ticket to improvementsorry i have neither the time nor an oracle install at hand. +Lets summarize it this way EXIT is a special signal MRI handles when the system is shutting down +But the setter case for xwebwork should be fine now +Tested with ER and closed. +Im fine with it +This has been fixed long ago it seems +Found some more +duplicate of of +The others will wait until a later release. +bq +Typo in fixed +Interesting +I think this would meet requirement two +patch available +You can try and increase the timeout for requests using the t option +Patch will be in release. +I will try reverting both +Nevermind +Implemented in revision to with rev . + +Committed at revision in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I also added a class which manages the filter list and also creates a filter chain for the request processing on behalf of the. +ResolvedIf there is something to update be free to reopen this issue or write here. +Velocity is no longer receiving fixes +set is known not to work on Pluto because of an encoding issueevent seems to be getting an exception in Pluto but dont recall seeing it before +is this a regression or a problem with the new feature in +Thanks very much Dyre and Stan +Wish youd said something sooner on this topic +Thanks for the patch Inder +Both jars are LGPL +Though this can be relaxed if there is a good reasonUsing stack here might not be needed in most logics I see more of its advantages on debugging side +will be used by EAP in lib and plugin are signed againThere is only one plugin now in the plugins console has been rebuilt with changes andI did not fix the in repolib we can do that after release +There are some kinks to be worked out but is intended to be a simple solution +Ok I see so its not the module loading that is failing but the generation of meta methods +He YongqiangHave you started working on this oneIf not I was interested in taking a look at itPatch link hive is not working can you share latest patch here +I tried to wrap with now I need to implement all the abstract methods +That sounds like an excellent solution! Still wonder about the bounded queue sizes though +IBM and does not intend to change this +My reading of the code was that it could be but Ill work on an actual test case +Line in trunk already represents the recommended change +on the compatibilityCan we bring this discussion to the mailing list +If we allowed multiple requests then the socket must remain registered with some selector and you have to send the response back to that selector +The job run time is almost the same +Marking closed. +I am commenting only sorry for confusing +Updated the wiki page also. +keys +per comment. +Yes that would be a better test +I certainly agree with a more conservative approach on this given the proximity to +BTW do you know where I can get +Verified with revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +comment on to address Gunthers comments +Please reopen if you think otherwise with a description of the specific tests you think need to be added +If you cant edit your repository location after that it will be separate issue. +Can we change the kscf definitions to use kwargs as much as possible to avoid any future problems +Its in the file. +Im unclear as to why you are adding a patch here if you check the log youll see that this bug was fixed four days ago +Addresses a couple of warnings which are valid +I think maybe the best solution is to catch the exception and then rethrow it as a grails exception with an error message stating that there are null entries in your relationships and these need to be instantiated or managed explicity as Grails cannot do this for is possible yeh shall I implement it then +I want to let to go ahead of this change to make sure that it is easier to integrate into the code +Ill see if I can come up with anything for this +Does it affect Lucene too? My guess is that yes and therefore we should issue a warning to our users +Ive committed this and merged to +this patch fix the problem +But I dont see that as a priority for test code +Pulled up this documentation as part of completion. +When you say Not Scalable What do you mean? May be I am doing something wrong here and your comments may helpRegardsLike Adrian said One disadvantage to using Dojo the visual elements are not scalable +Closing issue out as part of JIRA cleanup +you please at least add this to the release notes? This sort of change between release candidates is unexpected and rather bad form +I just committed this +Committed to trunk andThanks for the review Ted Stack and Lars +Applied. +This isnt an issue with the tests +OK +I am reducing this to major as it only affects portal users and in fact is rather something that should be solved for JSF in general and not onlyregards the issue itself how would prepending the viewid solve the issueWhat if you have two portlets of the same on the same page and they are on the same in the momentYour solution sounds reasonable but doesnt go far enough cause the client wont know anything about this and will send the request attributes in just as they were set originallyregards MichaelI understand what you are saying +So all issues not directly assigned to an AS release are being closed +is a method for hadoop +ics attachment allows the +Ill push a fix for that shortly +JVM crash with EAP crash on crash with workaround the release of EAP pending all relevant JIRA issues which are open are being reviewed +bulk defer to +Hi Patrick +If I reverse the patch they all succeed +Heres the combined patch that amends that +You should be jdk to build EAP I think +I will run it more during the weekend +Everything seems to work as expected +fixes the test failure +So to me things loook the way they should +I had the same yesterday evening with in +I will try to investigate this deeper +Closing issue. + +new patch is OK to me +The file icon has an Apache feather if the license has been granted +In the virtualhost you use to receive messages from the nodesWell it must going the main server +Another way to solve this might be to use an nio mmapped Directory implementation +I am fine doing it in the application code for my application +Normally Java is case sensitive so although it could work with some compilers and classloaders on certain I dont think this is a valid behavior to expect. + +After that we can switch back to x if desired. +I know its calling update as I have hibernate sql logging turned on +I need a break from this +Fixes javadoc comments with improper param usage in +I removed some junk code etc +Does Karaf run on GAE +Any related logs in the RM and one of the will help +Cancelling patch +added patch for on rtrunk closing all issues that have been in resolved state for months or more without any feedback or update +binlijin yes it is safe to use the old style for this one +Committed revision Throws when the input array needs an output array bigger than +This is already fixed in the web console with as a after release +The user can still manually deploy the file if they m willing to defer to you if thats the behavior you want +As promised here is the patch +All tests passed on my machine either +Some changes required to properly initialize prior +input as invalid as suggestion of the reporter +We should remove that from the general interface and only handle it internally in +Forgot I already opened for hit ratio out of +Todd I reviewed this with Praveen before he created this jira and his point is that the error reporting seems buggy when trying to increment a cell that doesnt contain a long not that a string or int cannot be incremented +Ok I will attach a patch fixing this in a few +Besides this feature is going to become deprecated in favour of the new natural conversation ids feature. +in trunk rev +lgtm committed +I tried with the latest changes but now instead of hanging I get the error +Committed to trunk and +Thanks martin I will confirm the fix is visible to uDig when the snapshot is nexted deployed. +So it at leastseems like the new code does no harm as far as is measured by the regression testsIll look into each of the test failures in more detail +How about a global flag +Original JIRA to track what needs now contains a candidate patch based on my m currently trying to get confirmation from Chiba on whether the patch is the correct one and hence we can patch the javassist versionto go in the EAP for July +Sure heres a simple oneWe have a table User which contains a list of all the usersWe have a table Activity which records user activities which references the User tableAn administrator can mark a user as removed which removes them from search and makes that user inactivated unable to log in or perform any actionHowever the Activity logs for that user are still preserved and the users row remains in the database as a foreign key for later use to display that users nameemailThe administrator may want to search the list of users by name email or other fields on the userThe activity logs also record which object was operated on +I just committed this +We could not replicate this issue +Is in anyway related +Yes that is correct +Sample to illustrate SOAP header usage +Do you mind uploading one that will apply on + +you do not manually edit +So I have attached the modified version of +bulk close of all resolved issues. +I should have a deeper look to find out if it could be fixed +Ive committed this to trunk. +Assigning it to Rick who checked in a fix. +DFS does the same for close +Release pull updates dependency of camel exchange bus to include dependency on +to wait for official support for on the trunk +Looks great thanks! +This allows system portability and simplifies the classpath management +Queda aix destil un cop imprs el comentari i tamb al mail el punt ha quedat substitut per un gui. +Yeah I think we should only put formats in the profiles that can build it need it for now +patch for mock nice addition for mock driver pushed to master +verified thanks +Thanks for updating the patch +Yep its your fault for letting me loose on the code Andrea I did go into a bit of a frenzy removing the crs field in so that it could not be out of synch with the field area +FixedThe spec is not clear however and I implement the usage of Entity name for EJB queries not to reference entities in associations +Thanks for providing the patch +move to unscheduled for time being to aid release planning +If you could give it a spin to verify would be much appreciated +Cool +Patch v addresses Jons +Mark do you fancy taking this issue should be pretty straightforward but we need to make sure we clear any references to this in the spec up +So while I will look into making this change it seems incorrect to categorize it as a blocker. +There still are some issues with queries at the JCR level but those are being addressed as part of +Added produce satisfied stats +SamisaI am using gcc version on a linux kernel +verified maxconnections with update and able to see updated value after reboot routerUpdated values persist after network restart +Throw some additional logging in and just run it in the debugger up until that point you should be able to see the state of the current object and the value of the field its attempting to set will likely be the dead giveaway Ill bet the code it fails in is dynamically generated code from your intercept +Added rectangular design and a custom cutshape design +Jonthanks a lot for the patch Ive applied it in SVN trunkIm leaving this issue open so you can attach your documentation patch Attached is the doc patch +Committed to branch and trunk +Ill get committed to so that Pig doesnt need to own this +Let me explainThat method is a restrictive method but a required method when you manipulate the OM structure knowing what you are doing +That is an excellent workaround but it isnt a suitable fix for closing this ticket +Thanks +Do you have any doing hudson jobs that would show the behavior +Ill open another ticket for the doc update. +Ill have to adjust how our compiler generates programs a little bit but thats no big deal +This now also works on Maven. +Again I will have to confirm this +Ill commit this soon unless there are objections +If the preferred object is immutable static cannot be altered thenmy code should work OK by eliminating the matching of the on the mirror URI string found in the list of mirrorsIt looks like your patch should workSteven J +Thanks Ive fixed this in SVN and the patch will be included in the soon to be released am having a similar issue +on committing the patch +Commited slightly modified version of v patch with revisions Revision +That makes sense good solution +This is still open has been fixed by adding a new call that does contain the revision at which a feature has been created and the one at which is was last updated +Finally every works fine. +finebut can you elaborate why it is done the way it is now? I did not find the answer in your book +hasArraygetArray +Patch to support array types in MethodParameter Expressions +All of the issues outlined here have been fixed either in Sling itself or in libraries. +Verified in Version +it will be like the server will always NOT write back the cell tagsCurrently it is based on the mapping that we provide and no context is available with the client +Retargeting to and +The refers to the common jbossws framework while to the backend ws stack +SPRINGCLEANING JIRAS This is changing with core discovery. +I think it looks good +Joergis there any update on this? Im serializing object between a servlet and an applet and I cannot guarantee for the same version of Java on both sides +Great +The framework wont delete these bundles but it doesnt give them certificates either +Things are a bit hectic right now +Ill take a look +I tried a svn delete but unfortunately I dont have permission to do itI reassign this bug to Guillaume as he should have the good karma +Agreed it would be nice +Any voters? + Perhaps a JIRA against MR is in order +Thanks much Stu for pointing the issues +We need to make sure that we are reporting the Document node as the when no other node is available but thats another issue +But we have production users who are married to the old api and would require them to upgrade their code to use the new override apis as opposed to a table name change if we recognized FQTN strings in the existing api +Fixed in r by applying patch submitted by SalikhNow a local variable initialized in a nested async block compiles correctlyThanks Salikh +In my app I see this for getting the Spanish language jshttplocalhostlangtDGXLSo Im fairly certain that the problem is in the ckeditor component +through the SPI. +Could you try to test this again with version +Patch to mask password in connectionURL in log files +class files are present in the default Eclipse project output directory +Thanks for reporting. +No it just means that if your renderer defines a baseline the actual checkbox square or radio button circle wont be to the renderer within the button +Maybe in a month +user indicated they cant reproduce the problem. +I think its good to just import extdigestlib for compatibilityThis commit changes the name of feature from digest to +Note this patch relies on the patch committed in as well +updated patch for and trunk +This passes integration tests for Cassandra Hadoop and for me so Ill commit it soon unless there are further comments +patch file for this to ensure this goes upstream into EAP +I used the escape character and the period character +BULK EDIT These issues are open to be picked up +Ive still got issues with getting it to run properly in Eclipse but it does run successfully from the command line using mvn gwt once up and running I noticed that the generated GWT screens do not appear to have catered for the entity associations +Fix included in commit fbafabdcdbdff +Lognet writes debug information to when something bad happens +verified by Alexei +The problem is that the is missing the reference to the called service +I started a new Designer session clicked on a bunch of properties and everything looked good +Any ways data file remains same in both of the approaches +Retry +Looks like this test failed again in trunk build failed in againEdit Same exception as in description +fmui added as PMC Chair +If you dont like the poolingsyncronizing of the you can always override the createAction Method and always return a new instanceIf you are using Tiles youll have to override instead. +Thanks for the review of the patch ysteinAttaching a patch vc addressing the buffer size issue +Will post the patch with the update cannon alsothanx and sorry once again for the redundant posts patch once again with the Upgrade cannon +I agree with Brd LangyIt is exciting thing to have meaningful names for segments and fieldsIt was something terrible to parse Stylus Studio mappingMay be for EDI guys it is easy to remember thousands components and fields codesI think that very few people when take a brief look at the proposed XML will say you what it meansBTW how will be named properties in corresponding java beans e very informative +Robert unfortunately doesnt do in this case +Understood I guess we can just leave this up if people want to take the responsibility for applying it was brought to my attention that the patch I posted before had the wrong unit test attached found a few spare minutes and cleaned it up in case any passersby want this capability +as my comments above the only solution is limiting hbase concurrent open mapfiles or it will cause OOME as data size increasingposting my patch here +Alternately you can use the jars and run against +This is an example of the launch file that is produced by the JBOSSAS server and is used to start the JBOSS server inside Eclispe +Isnt this more an issue about the used date components than the binderbinding +I agree that this contradicts the spec but it looks like everyone else has standardized on this variationId suggest either changing the default for the client to with an option to override for the older namespace or keep the default the same and give an option to use +Committed in revision +Ive used the Asynchronous workaround which is why I began advocating this position +Is there any solution for this issueIm actually running into the same problems and might thus need to create an external ant deployment script in order to read and merge external there any chance of getting an update on this +We can do this now only by actually retrieving the block which is too costlyThis is the first version of the tool any contributions are of course welcome +Thanks Senaka Alastair and Lahiru for the input. +If it is not clear Ill be glad to more component can be kept by +I think we have this one now resolving +First patch +Applied reviewed by Paul +My only suggestion would be to try and customize the tag so that it works better for your architecture and report back +There are some more specific cases that might be interesting +Log spaces in the path will give the a from windows machine +Paul will you take a look +For all the watchers its time to DL and play with this code it is designed to work against too +committed thanks +thanks for fixing just committed this. + +Im not real sure how to use xmlbeans straight out ofthe box +A relocation of all would imply that the default file would need to change +introduced to work with i did a little more now it is not needed to change to the older one should also works with both EAP and Hibernate did have a transaction manager look up class that support EAP but the s JNDI name changed in EAP thats where this issue came from and it makes this function not usable when using hibernate in fix is just trying the two JDNI names and do the look this only an issue in EAP then? Its been flagged as affecting EAP and needing a release note + +There wont be a release. +the token queue was growing correctly but the operations mapmissed copying the last element when it needed to grow +I assume step one is to do this for the file or is it multiple files? If it is not a file it is editing what is it then? Which way does one communicate with the repository? Which version of ESB has this informationIf not a file how is versioning and temporary edits going to be managed? Having only respository edits makes things waay more fragile so not sure how that is better than a good editor for files +Not easy to implement +This was done automatically by Eclipse and can be removed if desired +I think cloning should be done by before writing +Attached is the screenshot of the jBPM runtimes page unable to click Remove +One more thing Im trying to reproduce this with the current codebase notwith as specified in the bug report +Sorry you have every right to be frustrated +The file can be used to set different levels of security to be considered at deployment level +Sure I would like to work on this let me know how when can we discuss I can also start off if there is any reference implementation for similar for missing +The policy about tests does not call for unit tests for complete trivia +Ill send this one in unless there are more improvements or objections +I cant agree with your more on the standard of giving the full power to the end user Lets call the aggregate method when the aggregator gets the exchange each timeBTW We also need to update the comments and change log for it + +Ill leave it to the mckoi team to fix this. +I have updated the patch to return for regardless of its specification in the configuration file +I would like to ensure that the test case is working without setting prefetch value to and even with messages blocked at one consumer for some limited or unlimited time we have the next thousands of messages flowing to other consumers +From Vitalis emailA lot of time was spent to learn debug and understand architecture +I also patched the same issue with in a similar wayThanks for the contribution! +Thanks this is a timely request Ill take care of it +i intentionally avoided passing a Session or similar in thean must IMO not interfere with the core of jackrabbit +fixes the issue +Tasks task attempt status column to the history server page for task attempts as well. +It is related to the bug the solution will be to create for layers +At this point the code necessary to use the multiplexer is in place but the build process does not add the sar to the distribution and the config files for the clustering services do not use it +Thinking about this some more I dont think that any special treatment of implementation objects is needed for because the service registry will only be calling the service factory once for each bundle and hence once for each component +attaching current patch +However please note that we do not yet test on JDK before public releases but welcome anyone who uses Synapse on to let us know if things work well have added a warning if the Java version is not Java and let it continueThis is only on Unix I guess there is no issue with bat file isnt it +What I am proposing is that once a block is replicated it will be deleted from decommissioned datanode +There is a mismatch between the infinispan logic and the infinispan configuration logic in the container +Do you also have the stack trace for the exception +I have tested and add the new test for the test frame work to check +updated phabricator created the new review +Would prefer to get this in and then add follow in jiras to improve this +I know youre busy +as this is a released issue. +I added a note to the docs about rendering the xmlns attribute +Yes I created a review +However now that thats done every method which calls doForward has the same test which Bobby originally proposed putting in doForward +New patch fixes and removes patch fixes the test failure in which was due to a logj config conflict with +perhaps one of its own tests will be triggered +Patch for the bug +Im beginning to think we should rip out the custom intermediate and replace it with Kryo but until such a time I think this is appropriate +If I cant find a simple way to demonstrate the behavior in this timeframe Ill request the bug be closed as user error +depth name pow +This means that in mode we should always represent strings in the same encoding and always decode them as that encoding +After some more thought it looks like the simplest route to supporting this is to always just rewrite arraytable as the individual arrayget calls +Ill track the eclipsebug too. +Attaching patch from +If you override the URL in the child it is set fixes this issue +I fixed the problems encountered in the and the modules +Add code markers to stop reformatting of code section and conversion of to +We hope to get this sorted out soonThings move a bit slowly here I will look at the patch sometime soon is still a few weeks away +closed since I cannot reproduceSiriam your issue is not related to this please open a new jira with info on how to reproduceThanks +Actually Im just going to resolve this since your commit is in and it does work +Thanks Gert! +Maybe a warning in javadoc may note this canbe used to corrupt struts tags and make them produce invalid HTMLThe new htmlattribute may be integrated in projects +I raised this with the Spring guys and the concensus is that you must have a circular file references in your tomcat directory somewhere +This also proposes a mechanism where the primary sends the entries directly to the secondaries +Screenshot attached +Alejandro there is +AntonCould you please tell me what build and JDK you are using +patch Failed to include jsp jsp page in had to be reformatted +I have reviewed Brians patch and I believe that it correctly resolves the problem. +It does work for services now just not for referencesSee to during tidy up prior to +Sorry the patch submitted for JIRA duplicated some code I had also written so I didnt include that as part of the submitted patche +Please provide more details otherwise well have to close this as theres not enough information to go on at this point +xml extension beneath +I changed the patch +There is another issue with empty ranges +use to hold the container info +The upstream issue has been resolved +When someone suggests that you post your question to the users list it is wise for you to do so before reopening a bug +As far as Im aware no scheduler avoids giving out reduce work fearing that it could starve itself of map slots +perhaps should be to? for consistency sakeI also couldnt get your patch to apply from either commandline or due to inconsistencies in the class level comments in +Still that part is ugly +Change management UI to use POST requests for resource deletion +As it seems to me a bit overengineered solution without clear benefits with a message factory custom messages command processor and so on I dont want to touch +Sounds reasonable a few questions though is this meant for trunk or both? what do you have in mind for Maven ant integration +This sqlex function has some quite complicated logic and checks a lot of DB internals so its behavior may depend on how DB is set up +Shouldnt the usage messages indicate what the possible values are for the name conflict strategy +Separate JIRA to follow +Ive committed this +Added code to advance the parser past that section. +Ive changed that in the TCK configuration too +copyright laws the work may be protected under the copyright laws of other jurisdictions when used in these jurisdictions +To support etc +Good work by Sebastiano in finding this +Closing as duplicate of. +Snapshots are really a subject to change +If the maintenance overhead is not too bad then keeping the duplication around may be worth the effort to preserve some backwards compatThanks for taking this from hereAndiYep Im leaning that way too +Any chance to take the rules of the default profiler? Or i have to know the profilers name +Thanks for reviewing I agree some more cleanup and removal of code could be done +It has to be optimized to run more quickly +The message can be and should be improved +when a point release is made will that be a time to close a and create a new one? Or will a supertask remain open as long as it is valid with subtasks beneath it that are resolved in different releases +Hi Edward thanks for highlighting this issue +As you mentioned we will most likely always want to set IPVOPTS to something or not at all +It probably does not affect the correctly configured users +Ill open another JIRA for this +That isnt what we are doing +I realized I dont need to push this position checking down into the implementations +Also does the bug also occur in the ADL or only on the Android device +Snjezana could you look at this? We need to get the runtimes configured for the servers installed for EAP and and in a way so the names arent colliding and not too long either +Its almost there but adds all the dependencies to the rar which it shouldnt +In this is a usage of as well +running entire test suite for ode now +I think projects bundling artifacts should take the NL files at face valueAh that clarifies our disagreementIve been considering the recursion mandatory +Now fixed will be released in and +Added unit tests +Thanks Pavel the patch was applied to BTI branch at rPlease check that the patch was applied as you expected. +I believe this is the same issue that Andy fixed some time back but I had to revert due to some other side effects of the fix. +Marko Im waiting for a sample generated class as I can not reproduce the failure you are seeing with the test class Ive coded myselfCan you attach it please +I plan to commit in a day or so. +Might be necessary to wrap invocations in the in a try block +Workaround made in commitsd workaround for thsi should be changed once the bug is fixedbdc Kludge for not deleting maps take out the option from the UI +Ive been using for a while +Please note there is no need for a anymore an existing should be configured with a instance +Am I right we think this is not a problem then +Its not a big deal +META +The simplest patch I can think of is to check the status of the next field within ve implemented this workaround and Im going to commit it but Im not sure its the best way to deal with it feels +thanks +I dont think references JSR in any way +It seems that on thisparticular +Ho sorry MarcoI was really tired when I wrote such stupidity of course its the right solution! I have commited your proposed change in revision +At least thats what Im expecting +mistake +I did not see any changes to thrift tests +Thank you very much Edward +double click +If the replay fails then the local transaction should be rolled back and an exception should be thrown to the remote site +From the information on this jira An OASIS issue has been raised to make the specs consistent +The operator level hooks were introduced to allow a user to profile their query and see which operator was taking the most time in order to determine whether or not there are query improvements to be made not for measuring the performance of the overall system +Updated reviewboard with Chunkless patch +Just making sure that this is really targetted at the hibernate plugin and not hibernate +This may be resolved on trunk +Thanks for the patch. +Could there be other parts in Camel that have forgotten this +in to track backporting this to or similar for GA +obsolete test case +please provide a sample project in order to reproduce the +This patch contains a fix to return appropriate error codes while trying to delete or cat a file +This is not fixed our projects should work out of the box +The reason I opted to use a second set of constants was because Kihwal had asked me to split the code into two distinct parts one that does the JNI wrapping and one that is a straight C library +Also the title and portlet controls are hidden +The first issue to look at is to use the non minimized version of +Ill continue to work on this +months since last comment but in case its still relevant we are running off trunk and this test PASSES using Pig +Old JIRA cleanup +It should be trivial to provide that same plugin for folks to install on the Nexus instance and all we would need is credentials to push our data over on a monthly schedule +That was never the goal here +Bug fixed for Packed and Packe +svn ci m Applying improved messages for unknown properties in set and get to resolve issue Sending srcjavaorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsTransmitting file dataCommitted revision . +Committed as revision +I will review it once you let me in r thanks for the patch +This is neat but I dont think we should advertise itDefinitely not +srccontribhbasebinhbase Add svnexecutable property +Resolving please if Steves fix is insufficient +The problem appears when compiling and running the testsuite with JDK but does not with JDK +You can override this on the entity command using the identifierType unfortunately not once its created +is the default +applied thanks! +Correction above read asThis seems to have been caused by the descriptor in being +marked it as duplicated since has already in. +I will do this +Ive just fixed it already in! You are one step ahead of me ve heard that has multiple servers for anonymous CVS some of them being more current but Ive never actually checked those for the fixes Juergen +these are the contact mech types necessary to support the import +This issue needs test coverage as described in the linked report +Maybe this is related to for some? does not affect +Still waiting on Facelets and the JSF portlet bridge release +Oops incorrect resolution +In progress but currently blocked by a bug in +Before that it was impossible to use the tags from Sitemesh templates and there were other inconsistent issues +For +OK thanks +A user defined variable pointing to the run time location instead of an absolute path brings a lot of flexibility +as much as I dislike winblows the reality is that their async IO facilities are far superior to anything youll find in Linux +mike what is the status of this +Moving out of +How about something like this then +The normalizing visitor tries to rearrange the AST for the filter so terms in the same nesting level are in the same cannonical order when printed out this is done to be able to compare filter expressions +You have to try and understand what hbmddl does +Modified task description +A simple test caseAdd two in the set after merge only one entry returned in the set +Thanks +Currently they have just duplicated the code from the defaultsYou can for example just use a boolean to check if the custom factory method was invokedAnd then in the unit test method check if the boolean is now trueThis is a easy trick +Alessandro are you able to resolve this issue till next week? I want to release ASAP. +But my evidence so far has been pretty bad although i did not care collecting it for some time now so it related to less modern jvms +Now it iseasy to disable automatic copy and to add my own +Pick the least datanode as the primary datanode Do not calling itself with RPC Increase default from to do we want to increase the number of handlers in this patch? Keeping it at would give us better confidence that problem is really fixed +If you add the dependency to your classpath does the problem go away? +Thanks for the patch I agree those look like the right header bytes to check for +in added transitions for it possible to create a patch from your changes? It is pretty hard to see now which lines you exactly changed andor added +Updated patch with cleaner structure extracted classes and added testcase for it +For now if people really want to use Forge and Eclipse for something at the same time they can just type blah setup which will add the blank activator that Forge expects +The has been closed as it was dependening on closed too +Closing issue. +Added a link to the original issue this jira stemmed from +Looks like this accidentally got incorporated in the commit for back in September +patch to switch from to milestone elsewhere but worth recording here we are currently using these zips as input to JBTJBDS +we are fine IMHO +Great analysisbenchmarking +If you find further issues please create a new jira and link it to this one. +But if one were to use viewfs I am not aware of any additional failover issuesViewfs has bindings like user hdfsnn The failover issues are the same as if hdfsnnofUser was your default file system +Code pushed +revision Doc issues in release versions. +It looks like the issue is going to be resolved by +Committed revision . +Uglyinterface public interface This is smelly internals and here you are giving everyone a whiffIn general its funky +Derby completely compiles from the source tarball today easier than ever thanks to all your recent build work +Here is the minicard for thank you +needs move regions to the new machine +format function has been changed to get +Were doing a after all taking some risk is allowedJuergen +Arvind are you looking into this? If not Im going to have a dig around see if I can figure out what is causing the lock contention +Where is app coming from? The token? If its coming from the client request object itself then can client send different apps attempt id and matching container ids? sounds better to me +srcmainjavaorgapachewicketexamplesforminput +I dont believe this is a good thing at all +We okay with resolving this +Sure! Let me prepare that +Id suggest that this is an important line of development and I would like to see the development continue within the project +Additionally I think we should limit mapdbs use for now +I will commit it once tests pass +This patch works for me +closing issue. +After installing and starting the server the mentioned exception is thrown when deploying +It looks like that JDK is causing this failure +Committed trunk revision revisions revision I streamlined the code a little bit more +This patch modifies the class to add variable entries relative to ECLIPSEHOME as described in the comments above so that eclipse projects with Derby Nature can be committed to an SCM without causing build path errors +ThanksExperimenting with use of timeout values when buildingtesting w Tycho +One of the things I intend to do in Forge is to investigate whether scaffold providers can acquire more control over the parameters they are supplied via wizards that are tailored towards the providers +Finally I think we can apply this patch +Any update on this issue? Were using Hibernate where this boolean property is not yet available +Maybe do that for everything anyway? Whatever we do we should be consistent though +It looks to me that you have mixed context information and the configuration information both on this new context I guess it is good to separate out configurations from the Context information +hbasesnapshot if false we use +Oh right I understand what you mean by currently now +If this ok then I can add a here so that we can work this out separately before fixing this issue +Thanks for fixing this issue Jeff. +French error message should be reviewed by a native French speaker +NitayThe config in should be expressed as numbers and not operations like should be and not +Test case +Obsoleted by. +You could subclass and get this and lots of other tests for free +just something to be aware off +create command and create a new directory toput the eclipse phonegap project into and this worked +cleaned up broken references and enabled javadoc checks in eclipse project settings to prevent future is supposed to happen in Eclipse? I cant see any difference +This error seems to only come up when the interrupt flag is set on a thread that is doing a commit that which causes a page to be written to disk +Thanks for catching this +Same problem on line of dont think this is a blocker pls feel free to correct me +Bulk move issues to and also happens when you simply visit httplocalhostsolr and then try to to hit back +AS name change is not finalized we will blog this when anything is decided +When I set the action mappingsinputfoo and forced method validate to fail the resultwas a message saying that resource foo could not be found +The trunk code has been hardened quite a bit over the last couple weeks +This has been verified +Attaching so that I can get feedbackNote if you are behind a proxy you will need to have ANT configured for the proxy or use like this approach because it doesnt require maven to be installed it removed and and it makes it easier to generate for maven distribution +Im not so confident that youll find a lot of people on the list wholl feel the same need you do for this kind of tag but then since its open source you can change it to suit your needs. +patch unit test +M released +Thanks for the investigation work you did on this +Oh +For sysadmins of clusters of any size providing this in a web page is interesting but not useful for other scripts to consume +Whether thats a separate application written in anything or some Javascript running in a browser served from a couchdb attachment +Have manually tested on a single node cluster +Fixed in of a bulk update all resolutions changed to done please review history to original resolution type +Ive added a row to the table on page in the +To enable this feature specify includeFaultstrue on a sign encrypt element +Passing to the EAP Maintenance lead +is abandoned +Can you provide a patch for the docs as wellWhat will happen if you set eg true but false +Use if you need to add it back +Larger discussions aside I see no reason to not apply this patch at this time +verified in JBDS +A true unit test may be difficult but if it is an integration test I can include it in the Integration test suite +Everything seems fine with the latest CVS version of ll close the bug +Uploaded +Deleted and readded the repo as remote repository and everything is fine +Heres the patch +to trunk +DO you have time to fix itTxRalf +However current logic worked as having queue up not sure if anything relies on thatI will submit another patch then run some integration tests on real cluster +I am also uncomfortable making this change across all jvmOS without some better contract from the java spec on whatexactly the options mean +Ill remove these runners from shortly +Oops +Thank you +must fix for +Then I want to create a subtask to design the rest we need +do you know which jars are needed +fix for IE broken checked in to branch and trunk +Happens for me on trunk as well +You are right we dont know yet if it will work with Apache for example though what is on the plus side is that it is an optional feature disabled by default so if people do not want it or work with Hibernate then they wont even worry about enabling it +We have a fix but no test coverage + this looks great +there are plenty of other ones to pick off in the report +Deferring +this is regarding the directory layout consistency changes made on actually cover more than just that +cannot reproduce it +is only used during prepare +Any progress please +Youd have to build up distributions in a way that can be mergedScripting is a very cool thing +i think it would be better to score based on positions and run benchmarks and so on firsthere we also get a lot more index option flags +It envisages delinking container lifecycle from process lifecycle +The same goes for element which is the same as above except that it is testing that a instance will transition to +Can you please rerun this with the latest felix as I think this issue should be fixed already +A developer can generate jars of the schemasand then add them to his classpath +rather than just cap the filter size you need to reduce buckets too or you will get suboptimal calculation since you havent told it it cant have as big a filter as it wanted +The patch works for me thanks Pete. +Committed revision +Closing +That seems to maybe fix telnet but not nc interactive +The current approach allows this already by extending the files in the gen moduleIn general it would be easy to implement this but I am a little bit afraid that users are expecting an tool that is able to process any xsd schemas which is not the case currently +Thanks HariRafael let us move the discussion on syslog source dropping events to which is filed to track this problem +Gregorz You are right I made a mistake +Thank you very much Gavin! +Start with use Ants filterset on instances of localhost in srcresources +The channel no longer maintains deliveries so this is now defunct. +Failure to set the true property when a deprecated property is set will break support of deprecated propertiesbq +Which version of JBC do you use +This currently forces the need to to avoid deploying while doing the stage a default staging directory doesnt work in multi project a default staging URL create an other will out deploy goal +Closing issues +In any case a shouldnt happen if the Digester namespaces goes according to plan +This is a huge expense for a very rare case +Basically a good idea but we should not use a request parameter for this +If you are using the JDBC disk cache for instance you can get all item on disk use the memory cache for performance and not worry aobut lossing data on a system crash or improper shutdown +This second patch seems to work better +Were already tracking this issue in Atlassians JIRA +there are outstanding eclipse bug reports and lacks documentation +I generally avoid it entirely these days +Patch applied thanks! +This issue seems to have gone away +If it just need to be documented or if an other aproach should be taken +Paul if you can submit a patch containing the fix you found Ill test and apply it +Its definitely something to do with +I think that is still open +Do not share the list of producers amongst all the consumer threads +Also finished review of function literal and parameter tests where some minor todos are left +After implementing this feature we will get another problem for process issues with long commands +I knew about your proposal of new way of testing udfs but didnt realize that you have completed that already +This is fact an error or an exception +I doubt anyone is using it in productionJust in case it is used somewhere in experimental environment it will keep working +Can you reproduce +Also the method to create the old format strings could be moved to but the comment in actually says that the class will not be removed +Fixed by Willem Claus +can you do a example to reproduce the issueExample above has a syntax issue and does not show a you post the NPE java stacktrace you are speaking of +Methods with this attribute will be processed by a new set of handlersb Adds custom a new set of handlers in Regionserverreplicate uses this attribute +Updates the maven screenshot +This patch fixes the described issue +Please review this +Would you please let me know how to implement your solution? I do have Acrobat versions in both PC and Macintosh platforms +Closing resolved issues. +Apparently the bug has somehow disappeared by itself in one of the newer versions +hmm +Thanks for pointing this out! This is a bit of strange arrangement with IDEA even flagging the leaf subclass as invalid because of no setter being declared for the specific generic type that the subclass declares +So far it seems to work or do you know of any issues that might make production use not a good ideaRegardsRoger +I also think that we should be asking plugin developers to namespace their happens if two installed plugins offer the same command? Dont we end up with both requiring the user to type grails fooIt would be better to have grails grails t know if this can be enforced howeverMy AUD two plugins provide the same command say grails fooThen the Grails command line infrastructure will prompt you which one you would like to execute +Please elaborate if you think this is important +The request message contains the invocation context whereas the response message does not +To verify this fix I signed in clicked my account then clicked list orders. +The same comment try to avoid two inner for loops using data structure such as vector instead of array will remove one inner for loopI see lot of commonality between this test case and other two distributed cache test case the code duplication can be avoid if thecommon functionality is pushed into a function and reused +Linking to +Test order is dependent on filesystem order +I dont think it was intentional in work to change the content of jar files +I certainly didnt intend to imply that the current work is not considering this problem hence why I referenced in my commentbq +Fixed pushed into repository and tested on development instance +Committed to trunk +At Dags suggestion I have attached a patch that addresses only +Voted and released. +Those changes look good to me +Even then we might still log it if running is not false since thats the only case where we expect it +Improve bench tool so that user can benchmark a Hama cluster easily. +Just a slight correction to the test case the namespace declaration on the xsl element should be xmlnsxhtml rather than xmlnsxsl +For now the actual host list is updated when switching the views +Version of the disregard +See +This test cases was passing till Aug +Have a finalize on it that can clean up the contained Selector when itI think the hard referencing might be the most logical option here since finalization is never a reliable operation on the JVM +I therefore close the issue as being invalid. +defer all issues to +makes constants act truly constant with nearly zero cost and full inlining +bq +It definately would be the simplest option +Am pushing off to beta +Edison Can we take snapshot of volume of running on KVM +Bumping to +Only change is to make the maxOffset parameter in consistent +Half the number of trycatch blocks would be required if you instead wrapped the outStream operations and use a badOutput boolean no +just to clean up this renames the project name to to match and fixes references +I added a flag named XNOWRITEBUFFER to the sockets code +Its added my application classpathand now it generates valid this will help to for opening new issue +This doesnt need test cases either for the same reasonThe patch can be passed in +Looked over the changes seem reasonable to meResolving oldest open JIRA +I believe both patches are useful at the moment please commit bothOn a side note I have unpleasant feeling that some real problems in TM are just being hacked around +removed changes to which caused the test failure +However the flush and sync is done outside the lock +Any webserver needs this to allow access to a file +I verified that SSL works on tomcat but not tomcat +There are few column names with characters in length +if changed to what you suggested we had to use it to just start a bookie serverto keep consistent style I provided similar startstop script as bookkeeper for hedwig +Knowing when individual maps start and stop would require more work but what we can do easily as the short term measure is to annotate each progress message with a timestamp indicated above +But I still need to throw ilegall on the consumer +Delete the AMQ data directory and restart the broker +The attached file further demonstrates the simple case of binding a variable twice in one template. + +Could you also post the stack trace of the +Modifed source file that can be used in lieu of +Also if you just go to the deploy directory and touch the war file then that will cause a redeploy +It looks good to me +HeySorry if this is a dumb question but what is the Herriot FrameworkThanks thanks for quick response I agree with all of your comments expect for I dont see checking in this code will break any existing test cases since these are new methods +Actually i guess we dont know it for DISKDOC +I just got that error from above +I didnt commit to because the RC is already on vote and I dont think we want to respin another RC for this +Back and forward browser buttons are also working +A very basic init m going to leave this one for Eric to review +list of tests run until failure for safekeeping after this theres the stack trace from the description +Regenerated patch to reflect changes in sitesAdded instructions for running code coverage +Ive asked for confirmation of this on the forum thread but my belief from the initial quite complete report is only MPING is generating anomalous logging +Michael Ive been meaning to look at this but havent had the time +Have the loaders implement state machines +Id like to commit this today gauging whether there are still any objections to bits of it or not +Contains fixes for bridgeEndpoint when combined with match and associated unit test +Cool Ill try no that patch does not solve the issue the leak is still there just slow because Im using just seems that Ive found a solution but it might not be the best thing see the attached on top of the schema disposal I had to add a synch on top of that prevents the other errors even with there are various places in that use that call so I guess it should be made synchronized in do you thinkI also see that the builder is used in other places so eventually Ill go for the thread local dispatcher merged into the applied on and transition all resolved issue that did not see any further comment in the last month to closed status +I also dont seem to be able to replcate with any wcs requests +it looks there is some pressure on someone I will work on merging the branch in IMHO we have to work on this plugin the last release is very old +The build is failing because port is in use and that port used by a lot of tests in the tuscany build +I suggest the following fix +Thanks +Otherwise Im at a standstill on this one +The Aries JPA container is not actually an implementation of the JPA Service specification +Burr could you after you have launched with the settings you think should work go to Debug Perspective find the Process view and right click it select properties and then copy the actually generated launch command? Then we should be able to see what actually gets passed on to the launch and if it somehow doesnt match with what you specified +However as any client code doing this is actually working around a library bug it seems to me like its something worth fixing and in any case should be a simplification +This couldnt be done async because the transaction boundary was criticalI stripped down to the process so that the problem gets obviousUnfortunatelly i dont have the original process anymore we switched to another solution +We recently had this problem with class loading in so we had to use an aware class loader +Fixes so that hbase shell works again. +You didnt ask for jira or confluence was that intentional +HiCould you precise what are the files you are talking about +I think a utils module remains a good idea or else core starts to depend on a whole bunch of stuff merely because of some tool code sitting around +Can you please try to explain your points in more detail any of them at leastI dont see whats so special about this point on the earth youve drawn attention toN W geohash bbfqvuffNo point or location in geohash is special or problematic +The first couple of screenshots have been fixed up +Looks reasonableOther atrocities have occurred inside but they arent from this patchFor maps it is nicer to import and then refer to rather thanI think the question is why do we need to hang onto the object at all? We are doing work to null it out but why cant we just take the one or two fields we need from that in the delayed produce? If we do that then wont the producerequest be out of scope after handleProduce and get gcd? Is the root cause of this the fact that we moved deserialization into the network thread and shoved the api object into the request +Ive been working on a new version of hbck that solves a whole bunch of potential problems in tables +bulk move of all unresolved issues from to +Another problem with only supporting Origin information is that cannot be propagated if the event is +This is my suggested patch to fix the ve committed my fix +This implementation lacks support for other architectures +What that lastelse and catch blcok do is allow Enum instance to work even though they mightcome from different class loaders +Till we figure that out I have provided a quick patch to fix your problem +merged a test for it +Regenerated patch after was committed +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +You must use named parametersThats why criteria queries do workThanks +Thanks +I dont think we should show controls at all that arent applicable to the result set +For performance with a socket buffer and a fetch size of MB I was able to improve the cross DC mirroring throughput btw brokers from about MBsec to MBsec using this patchAs for the decoder shallowIterator only works with the +Were already trying but its entirely dependent on engineering hardware budgets +Heres a patch that doubles the TOCan also quadruple +Submitting for hudson +So now the color changes for stroking ops +I also added thread factories in and so that threads have meaningful namesTheres no unit test but I manually verified this change by running the same Hive query against it and observing that there was no second pause +OK Ill go again to documentation thank you very much +Might that be acceptable +Fixed in the trunk and branch +I think that should continue to discuss the issue which Olav has analyzed in +Thanks for catching that stack +These changes have been made for to Laura as she is the new lead for EAP docs +My understanding is hes comparing a proposed mode to just disabling the table +Didnt help at all +Catch is a way of handling this +for example Ill check this out +Patch +This applies to trunk as well as branch +Thanks +I dont think we should go RC without this though should not get worse with any major release +If there is a measurable performance impact then were doing something wrong +in rupdated +Influencing dependent as they are being read with your profiles makes the outcome fairly random. +Done +This would be absolutely great to have +See comment on Rev + push out to +Patch against SVN head to fix these warnings +Many thanks Albert Great Job! I will wait a bit before cutting a new CMS release to make sure weve got everything in there for shared lib supportLooks like we still need to add support for shared libs on nixgcc +Modification to authorization and once the Ant directory is there +Using a jointable ist default behaviour if you not use but the JPA spec also supports unidirectional relationships without the need of jointablesSee section Annotation of the JPA specThe RI supports this kind of relationships +Will be in next tagged release to platform +One of the main goals of the marker file is to speed up scanning performance +A number of these procedures use numeric arguments to encode BOOLEAN values +Seems a bit tedious but I guess its not too bad +Yes Todd please commit to trunk and +Please install grails support for STS. +At this time the passivity will remain for this feature through +Sven please commit your suggestion so we can test it before is released +Answered query. +canceling patch due to trunk drift +Fixed in rev +I loaded some data into a table +stagingjobDo you have the client trace? What staging dir is shown on the client +But I dont want to invest in their development +Although as far as I see it the bug here is that we allow the underscore after design not that we cant use filters from it +you get full bit values no matter what you have encoded +I looked at the code and it can pick up the nesting in one of two ways +Btw thispatch throws axceptions if identity fields are declared as transient asdiscussed still need to verify whether I need to add code to guarantee that at least onefield is NOT transient andor +The automatic revision reflect tools got it right. +Will be working at that direction +But still it is an annoying side effect +So if you think an upgrade to EMF will help I would be all for it +I dont think such an approach makes senseThe whole reason why the OBR location string is meaningless is because there cannot be a relationship between what the location is and what is installed there +However they are not showing up in the UI for certain +how it look correct Fix Version to one without x like depending on where it was in JBT this what is this supposed to doit is importing internal jars that are not supposed to be use by users projects! The jars in there are not recent and includes hibernate tools jar which is most likely not compatible with what they want to is exactly what we said should not happen +For each release note we need to document the fixes from the individual components as well as the platform level fixes +Ill try narrow downing the issue from my side as well +Thanks Arun! +Stu it turned out there were two problem here the close race but also the client connection close was not closing the selector which was causing resources to leak +So the check of the directory is unnecessary +Added file to the asconfig directory and modified run scripts to use it +great that explains why only you where having the CNFE still this was marked fixed and included in the release since it was causing problems with the release as well +claims that SLJ is not needed for Jetty +Until there is a more robust backup tool like what has been suggested for this would certainly be a reasonable stopgapSince for backup purposes the tool is likely to be deployed and used by systems administrator the README should probably remain separate for now it makes it easier to get it in their hands +The core client code is unaffected +actually this was resolved by fixing it was a problem of a class appearing in two different classloaders +Created for that more general usecase +But since those things are always that rarely get built shipping with a few sample s could be a good solutionPart of the initial setup in the docs would be to decide which to use and rename it toEven still there will need to be defaults so can we still fight about that? +Pushing through Hudson +Im not entirely sure +What should be happening is that when we create the distribution artifacts we should make sure to remove the WX flags in the win build scripts +However I wasnt able to tests for the same reason as the current Hudson failures on head so I guess we need to wait for that to get fixed +This is really my fault +Pushing to for now +Thanks +is this a duplicate +So while we dont know whether the checksum is corrupt we certainly dont assume that its notComputing is cheap every job that we run computes them on input and on output and it hasnt been a burden so far +Thanks Jim +Andrea could you review this patch please or refer to who needs to review itChanges are in geotoolsmain and in geoserverwms once the above are addressed go ahead and thank you for your reviewI have considered the complex structure of complex features and I even tried to make a template that goes in to the nested properties +if you want everything to reload wouldnt it be the most simple version to just use a completely new +Patch available for you will just name you patches uniformly like JIRA will automatically take care about the sequence +Classpath was just set wrongly +Thanks Mike! +The doesnt move all rows up to the top of their enclosing element +Things still left to do fix and use profile data for compiler DNA extract the DNA from the report modify potentially troublesome from statistics like bcbms as it can cause confusion in report generated in my development world with SPEC JVM configured for iterations. +So only BIRT update site has to be mentioned as needed third party plugin on JBT update site? Or we should enable BIRT update site during JBT installation? +The benefit of using this closure extension is that it has no impact on the existing API and can be used in conjunction with all the existing functorsIf others agree with this approach I can commit it along with some test cases +Hi Phil I reviewed you patch and had no comments +Thanks Aaaron +If you are using the complete class path why arent you finding the fileThis is not about exotic use cases its about being consistent in how resource lookups are done so if someone decides to unzip classes or import src are not seeing different behavior +I am attaching a patch that implements Bills suggestion +Dan you might wanna link the JIRA issue if you can find it so that we have it on recordThanks for the quick reply! Ill build a SNAPSHOT and try again later today. +So in the patch It seems I modified more placesIf it cant be accept and I also make another patch with less modify of logsSo theres two patches +Forgot to add over weekend GM run finished with this patch slightly faster s on nodes with jobs +As you said would be good to avoid this issue by resolving in some way. +This has annoyed me as well so if there are no landmines associated with enabling both builders Id be in favor of it +This causes problems when collection table primary and foreign keys are created +without constantly copying buffers to userland +out file +to trunk +New type getters are introduced which return a single type rather than the whole array +the patch to the is is still unresolved and its not clear which of the proposed solutions should be selected Im postponing this issue. +Im attaching a new patch specific to trunk since the number of lines changed +OK will create a separate one for the counter limit exceeded issueBTW I did open for a related issue we are unable to access job counters for the period AM is possibly shutting down as well it may be a good idea to consider that issue in the final fix +I am sure there are more formatting improvements that could be made though +I think we have a policy against the word wonky around here. +Please indicate what tycho build you are tycho does not currently support +move fix version to and +Minor change fixed. +Alexei please verify if the issue is resolved as you expected +test failure wasnt related to the patchLocally it passed for meAre there further review comments about Todds patch +If you want this committed Im quite happy to do the last mile +CommittedThis was not an issue in any released version of solr +Nice work +When a process gets to the else block above it correctly gets the right last but it fails to update the value so it will trample on any ids that may have been used by another process. +good +We should document this in Hadoop and Pig speculative execution seems to be leaving task directories about +Reviewed and tested +We dont have to match the RI if the RI behaviour is unhelpful to developers +Thanks for the question Tony Ive forgot to mention it explicitly +Adds a utility method to to check if a affine transform is actually a translation and if so uses the more efficient translation operation to build the for the record here are the jmeter results for the various benchmark runs +Ok Ill try to add an attribute to datasource element and change the Generic code accordingly +The code compiles with r +Hi Xavier youve obfuscated the filename property values but this is what I wanted to check +Makes sense to me +for the information Arun I think we should extract the needed from ebay to ebaystore and then from ebay hopefully its not too much workBut its not the only dependency entity is also set in and used inYour patch is in trunk at r +none of these are critical their cluster changes +We need two versions of these for the we support now +The latest patches areDevelopers Guide Reference Manual +I made a tiny adjustment to the unit test also +It doesnt need to make test +I entered this bug but have been unable to convince someone to fix it +Tests Apachings against Suns for consitency +Re open it if you have trouble because I cant reproduce your issue +release bulk close +I dont understand the theory in which this is supposed to work in CXF at all +Also present in and + +Anton can you please look at this as part of migration of the new edit model +Thanks Daryn! +Patch filepatch appliedapplied +Ill try figuring it out on my own Im glad to see its something idiotic Ive done +Also no test done on actual clusters +This is quite normal and there is nothing we can do about it +mvn test passes for me but running individual tests with +Are we likely to run into similar issues with other encodings besides +Eli I just added the missing file +Using an archetype catalog file might be a workaround if it is able to tell the archetype plugins generate goal the correct extension to look for +See. +back ported this to Camel and as wellrevrev +And because there is basically zero copy Kafkas memory and cpu overhead is incredibly low for what it doesThese two main factors its incredible performance with very low overhead using commodity components are Kafkas strengths +Fixed the nits +is for This is patch for TS please use it with your own risk +Attached patch with constructor unit based on code in m adding this patch and test +I finally decided to take time to move to SVN +Perhaps not the correct way to solve this but I patch to add a finalizing reference to the old sstables before calling markCompacted and that seems to work +Designer project archive and data source +Child could not be addedSo IMHO the bug should be reopened +Done +Nothing I can do about the jbpm one though +for patch on upstreammaster and upstream in Developer Studio +event is fired after when form is submitted +OK Freddy +Thanks Omkar! +Note sure well do this for M but its just a matter of appending some data to a message in the queue +same problem here I observe it only on multimodule projects and not strictly always which seems to be confirming the explanation far as I can tell this is the same as the already fixed issue as mentioned earlier +If you for instance change and save server host name in the Server Overview editor you will see the same change in the Server Properties page +Thanks fixed in SVN as well +Ive looked at it +Id rather wait to commit this until we have some unit tests for it +It got pretty hairy at times so I would appreciate a review +Fixing generic target on any attributes will remain for but critical cases covered by Programmers Guide will be covered without that +This wasnt a good fix +Closed after release. +I added some debug code in to print the stacktraces +Resolving as duplicate as this has been fixed as part of. +Im the separator is alwaysa forward slash +Hard drive failure core dumpVM crash controller failure thermal shutdown process crash etc are all non sequitor +It just does nothing no new project is imported +Thanks for the patch David. +Thats just a fact of life for now since we are depending on a generated file +Oh the project is to aid in acceptence test where we should log the action execution timesUsing both computer numbers and a manual stopwatchWith this enhanced interceptor I can setup a log category and have logj log it into my own fileAnd I would like to specify my own category instead ofThis is useable for a period of time in pre production and production so we are able to monitor execution timesThe times will be greped and used for inspection in Excel +basically if your query numberequals to either the number of success or the sample then you got +Thanks +I had it declared outside my trycatch block so I didnt see it +So I guess the bug is still here but Ill try to confirm as soon as I can. +I didnt find any slow queries logged in the DB or any exceptionserrors +BTW which constructor are you talking about + to trunk +Ok so I tooked the second patch and tried to apply some comments is this the one that fixed the issue? whats going on with the rd patch? why are some methods set as final? that will prevent extensibility dont chang formatting in your patches as they are gonna cause more differences than really are and will make it harder to apply +jnlp generated for modules contain this attribute nor does as created by the Ant build harness +Good catch +Patch checked in +In a production RPM wed probably rebuild and not include the but this might not be the case for othersInstead I think injecting jars into what would be the obvious classpath should only be done when a developer takes an action like setting a config or creating a target dir. +There are currently one or two disadvantages to using docvalues in solr currently +scv files need all the properties set correct? Currently I only set CHKCONFIG but I think DESC DAEMON etc also need to be set +Much prefer a service account with nominated youd end up with owned by +Verified with JBT. +As a matter of introduction Im a colleague of Chad Michael and Jim over at Powerset and Ive been working on various Hadoop projects that use Thrift +Marking as unassigned since there is no more work to do on this issue unless it after the patch +AndrewThis ticket just provides a pluggable aggregation framework +I reproduced it times out of times +committed +I also replaced some occurrences of HADOOPPREFIX with HADOOPHDFSHOME which I am not sureJust to help debugging I added a binhdfs classpathPlease review thanks +Something really wired +rebased the patch for +I tried those two calls but I get an invalid receive size +I actually just changed a few things out of collectset +This is how it is expected +With fixes to tests and with null dropped +Its even been a while since Ive run maven from the command line because Ive been able to do my work using meclipse +Now that work is underway with AS all previous community releases are +Please use a separate fileSet with appropriate includesexcludes to handle binary files as opposed to those you wish to filter. +Ive taken a quick look at the patch +Oh pretty good +Patch that implements this behavior +perhaps there is some other step I need to takethanksAndrew +Needs to support single digit M and D fields may of been fixed with recent changes +This is my patch for maven release manager +OBS I tried to both downgrade and update the version of jruby rack in the war file as suggested in various old postings I found on google +The changes are only in common base class not in the two specialized implementations +fixed javadoc comment +Requires no changes to existing code to work +See previous comment about how to apply custom type conversion when the Action class is using Model driven +search local repository currently no detail information or popup actions include remote to as the maven repote repositories cannot be browsed yet +Can we access the source code of lwjgl +What about Jim Firbys original suggestion Clients should be able to say Ive asked for a regions location times now and Mr +I just committed this +same question do you have the ER repo configured in your +in what way did the BPEL process have errors? did it throw exceptions other than the one you have includedWould it be possible to attach the example that demonstrates the problem +Will try other environment using then to see if that solves the issueRegards +push out to this at JPL and never closed the issue +and looking for in the location you pointed me but didnt find itsimonetripodiminotaurwwwrepo ls javaxorgxaaSo I performed a silly search but the stuff I found are owned by yousimonetripodiminotaurwwwrepo find +We do need to allow users to tweak preview behavior however as long as the import logic is performed by Designer code its a moot point about setting translator override values at import time since no translator is involved +added much more javadocs and even some more test cases to the testIll stick up the HDFS services tomorrow I will synchronise and run all the tests +Better still feels very crowded though compared to the original thing I dont like is that I cant click directly on a sample I have to click on Web Applications first +move reporter from Steve to to +Nothing right now. +The patch is for reference and paper trail on the changes +Works for me at r +Patch for ant bug of version applied to BTI branch at rPlease check that the patch was applied as you expected +Attachment has been added with description patch for the re right I hadnt had the need to pass in renderer specific parameters so I had missed thatHeres a patch +If you allow the blocked content it will display all right +Hi Andrzej The patch has already been updated to incorporate the comments above +Closing all resolved tickets from or older +As you have not provided a test case I can not verify is it fix +Well it has been almost months since I reported on this issue and apparently nothing has been done yet to integrate the small change that I suggest nor to reject it +The same fundamental doesnt play with others will not have been fixed +Please ask this in the mailing list +Using umbrella tasks hides the actual issues from me because they are +Awesome +With this patch agent attempts to get bootclasspath value from the following system properties sequentially Now affected JDWP test orgapacheharmonyjpdatestsjdwp works well for all DRLVM IBM VME and RI +Is this required on all management server nodes? If so need to add this as a step in docs as well as docs +What does the community think about this? Your failures and others should be connected to the above point I guess +I reran tests same three failed +If the is stored as an it should be available to the and if the doesnt provide it it is a bug right in this bridge +Though I might rename Problem to Error to keep consistency in the source +Ive seen numerous cases such as this in fact its why frameworks like Maven are extensible in the first place +Updated in light of work +This task is about throwing an exception instead of printing a warning if no suitable advice method has been found. +Once we know how to handle this in well support it in Spring +java root root +I think it might also help to add in a timeout feature as well so that if you do nothing at all the dropdown automatically blinds back up after a set period of time +So while I sympathize that it is inconvenient to not be able to change the schema in this case this probably a wont is the underscore issue actually considered a bug? Ie have they released a patched version of the schemas? If so I would like to just update both the internal copy and the published copy because this is indeed a pain that names can not have underscores in them +Patch attached +What were you doing other than just starting up the console? I was unable to reproduce it and based on the exception there looks to be some other interaction with some other scripts being run +please move back if they will be worked on asap for +fixing component with EAP +Dup of +We will do it in a few weeks when we prepare the next CP have ported this to the documentation +Test Groovy project that triggers NPE on my laptop +Keith might have fixed it with his changes in dont have this error +This code is quite complicated +that sounds like a plan +We are going to need to coordinate this commit with openejb +Adding the ivy dependency on and removing that jar from the lib directoryAlso removed need for as a transitive dependency in ivy since we are including itself nowRan unit tests and tests +Patch for adding in a simple simple test and added support into Writables +Reassingning to Heiko because the only dependency left is assigned to him +filter condition is to test the case of D filter C by t the filter will evaluate to null when t is null +So close it +I think this should also be done in the +It was cocoon but I am not able to reproduce this error anymore +If youd like to open up a separate issue for contexts we can track that work there +HiOk I fixed it +It would be very nice to have this onePaul Snyder + +I will only try to use it as much as possible +was attached +Ok +This is not vital I can always do stuff manually +All tests especially pass +Still working on the complete patch but this is the next version worth sharing +then +bq +Patch for trunk +No it was not really the same issues +Check for presence of curl +But I see the older version of and JettyShould we leave the libs in trunk for release +Generally speaking normal Java thread interrupt is not enough because we implement Ruby thread semantics on top of Javas thread operations +It should print to standard error +Fixed on trunk revision +accidentally hit enter on the wrong focus +I resolved this by changing the builtin name from to to avoid the naming conflicts +It might heart performance +Can you include the contents of the variable filename where you say the index of ! occurs at ? I think we are assuming that this string is something very specific +You are correct that my original report wasnt technically accurate but then I didnt have time to look into the details +Also encounter some tests failed with error Sort check failed +New patch removes unnecessary import +The same fix is committed at therefore I am closing this jira +IAC this wont fix the original issue the NM doesnt have the right for setting up compression libs correctlyArguably we can fix this but Id rather do it the right way +just question what do you think to save such secondary information in xml comments? why you do not select this wayif such way absolutely not suitable I think it possible to fix it just modify the file and when the user save it save also the secondary information with the project +Yes unfortunately in Weld we did fire PAT for extension added annotated types which is a clear violation of the spec +I meant add createXXX methods that accept an Object as user object +I updated Components +Do you think it is appropriate solution +You should remove my gem repo as a gem is need a new gem repo for experiments +If it is also missing component please set both so that the issue can be properly triaged +separated the related diff to from patch +the timestamp is time before which we can throw away deleted columns +hold a mapping from queryId to its +Negative +That would also make this issue go away +Rado can you suggest a better wording for this error message +Note that the unitintegration tests all work fine with this exact +What changed in? It looks like its deserialization of the CONSTANTSTATE column in that fails +Closed this by is this still an open bug targeted for? Surely there is not much we can do about it except upgrade to JDK or use the workaround as detailed +Id like to get this into can you put the patch in reviewboard for reviewThanks +The current web flow command only works with MVC and the command is not available after JSF has been set up in a project module +what was the resolution? +Looks like this is fixed +ve just committed this +I vote for Android and HTML +Can you try withRecently we fixed something similar +Was not quite the plan but we can fix the would like to set up a new uDig repository uDig Extras to hold that project and any other non locationtech work +can I see the request too +Similarly in the constructor of +troubleshooted a users query just now discovering that the problem was that he had no q parameter and thus his fq was irrelevant +I filed for the main issue +There is no hudson report after Mahadev updated the patch so Im trying to it +Im sure Its CSS thats doing the damage +Can you pull install locally and verify that its working properlyBefore this code graduates from the sandbox we will add support for configuring the actual instance +Thanks a lot for posting this Ericbq +thanks Arnaud I assume you mean CVS at version not the tip? Ill double check my I suppose he means the tip of the CVS the very latest +You might be missing the phrase such an application must not continue to bundle the Facelets jar file along with the application +Thanks a patch applied at r on branch please verify it thank +It seemed to me that the problem maybe somewhat further down the stack in that it when creating the new filename it was using the baseVersion and that was returning the timestamped version and not the actual base version +May I interrupt your for a moment +heres the new version it works with both ajax and no ajax +This issue is technically a duplicate of which is now fixed. +even Maven does not go to this extend to add all this legal fluff to their deployed artifacts +Here it is +I call the purge link from the browser through a ssh tunnel on the machine where the activemq is +Will include the test in the upcoming unit test suite for Perl. +Please reopen if anyone sees this again. +patch +My machine is a Intel Core i +tests now run but all the velocity tests are still broken on AS filed as is now resolved as well +I have attached two schema files that show the problem +I moved the bytes storage up to +Closing issue it has been resolved for a long time. +Reopening to remove the need for a release note +for is pushed to b +Another option would be use the dependency plugin to round up all module dependencies and then also include them in the binary dist +Bulk assign all unscheduled issues with priority major or less to target +I would have done it that way but I didnt know that Wicket uses Guava +Im investigating how to add annotation support and validation to our tooling now + +so i need to watch whats the diff between bucket cache and l cache under heavy read scenario +created PMD upload bundle and upload request to address this issues PMD version can you update your patch to the latest revision so I can apply it and get it into the next release +Actually I was thinking to remove the center configuration manager in the beginning and all the required actions could be done in the so we just need to start the bundles list in the but that might be the next step action +sorry I didnt realize that went from to +if autocompletion is invoked with Abstrac typed it will be enabled while if A is typed it will be disabled +can be set with hive using the set command from the hive shell +Looks like trunk does not have this issue and it does not seem to be doing the same in checkDest in +How I must define orgapachecxftrace function for this? +The improvement is applied. +as improvement +It wasnt added to the Solr connector because it wasnt clear whether the mime type filter would be adequate for peoples needs and the Solr connector had grown an unconfortable number of tabs alreadySo where things were left is that the infrastructure was written to support filtering by url but the Solr connector only had mime type and length filtering support added +Now it works as documented. +Can RobertThomas pitch in +Can you check out the last camel to verify the fix +Do you by any chance run the tests with java +doesnt compile without implementing this methodBut this is unrelated to the javadocs so I will not include in this change + +Applied the patch into trunk and branch with thanks to WilliamAlso updated the wiki page +Using does not have an issue with the backslashes +Same problemAny guesses on which class might be keeping those first two tiffs open? A reader in the store or catalog +Please reopen this issue so that it can be tracked +It doesnt matter much which +GregoryDid you run all tests in one VM or started new VM for each test? In second case you could not catch the leakage problem +Ive committed this! +Resolved in rev +Client sends this with request +Even though that bug has not been closed enough changes have been checked in for it so that it shouldnt block + +Seems to me you are missing theon inside https s a little fifferent setup for Tomcat and needs on in lifecycle listener m not receiving this error although I can confirm it doesnt work with ll check if it works with stock Tomcat +Im almost done with the import data files +This patch fixes the missing help files in locale NL +In Groovy if you assign a value you have to do a groovy style cast to the target type +Sonar IDEA plugin is no more supported so I prefer closing those tickets. +Is there is another plugin which can be used insteadIt would be great if someone can take a look or simple fix it +Thisissue is fixed in Modeshape which is contained in but the patch wasnt applied to branch +Here is a patch that synchronizes correctly the access to synchronisation and optimises correct synchronisation in Philippe just to clarify is this issue and patch for the release branch or for the trunk +I think we should stick to antlr only let us not check in javacc + +This patch just uses proc to see how many are open and uses junits Assume support so it wont run on systems where thats not available +The approach looks good +Does it implementedIf you are not working I can take this ticket and work or Suggest me some ticket related to this so that I can take it up and work. +Inside AS if youre using slfj you have two choices Provide both your own API and impl JAR which logs to your own framework Use the AS API and impl JARI recommend the latter of course +Easier figuring this stuff in a unit test than up on a cluster yadda yadda you know what Im at +Also Im not sure of the best way to include the mandatory abbot jars in the build but Tim says he has done this kind of thing before so Ill leave that to him to clean this part up. +The following quickstarts have been updatedhelloworldhibernateactionhelloworldsqlactionhelloworldtxsqlactionjmstransactedCommitted in revision +Patch attached +NadyaThank you for being so quick in handling this bugLet me address your concern if the conventions are yet a proposal +xml filesThe code is still rough and we will provide a patch to integrate it into webtools later +We have too many customers that are going to mutiny if they have to maintain two different binary versions of the same code to run on different grids +When I think hCatalog I harken back to the s +Zipped diff and stat files for all directories underAdded licences to two files catalog andCompiles cleanly +Here the fix for the problem +The attached patch adds some more unit tests with a header provided in the requestThe tests also show that the described problem does not exist as always a is used regardless if the is provided or notSo I would mark this as invalid +I was also wondering about the topic on the FLOAT data type I am guessing the precisions are the same so no changes are needed hereThanks very much +taking issue +Pull request linked. +fixes this issueTristan fell free to try the fixed by all tickets +PBR and are completely different code bases so there are bound to be things that behave differently between them +Applied in +SQL is an example +In that Jira is a link to a discussion about this that we had in the Im not currently using Web Flow in any active projects and dont know if I will in the near future +Thanks! +Will commit this later today +Thanks Ivan +Attached diff for Link component adding extra Javadoc +current workaround is to store current orders and filters in bean. +Comments +Why not make the config parser trim all parameter names to be resistent against such typos? I cannot imagine a parameter name to include leading or trailing spaces that would be bad usability +The link to the support case can be found in the original but for greater clarity Iveadded the support case link +Thank you for commenting +to httphostNeither of these work with +OK so if you remove the property from all the then the default value of Derby set in the persistence deployer plan gets usedIm going to put the default property settings in + thanks. +Right I agree it overlaps that thread but is compression support a thick or thin attribute? Maybe we should move this to the mailing list +We will need to refactor when is finished since currently we handle the probelm in this layer +I have a partially complete port of moving the file readwrite to completely async IO as well as using +server +I tried to rebase master just to see how far its moved its not too bad couchdb had some conflicts mostly relating to a new field Given Damiens abandonment of the project Im not sure whether we should push on this or not +I dont need another patch to know that this one is complicated and could be less complicated +Im almost thinking with the lack of having the necessary dependencies on the classpath to run Accumulo without HadoopZookeeper installed on the system maybe it just makes sense to provide another tarball with the cluster packaged up with a script to get it up and running quickly +Thanks Sandy as a follow up JIRA we should make sure the tree expansion remains the same on browser refreshes we could accomplish that encoding the current expansion in a cookie thus not keeping any state in the RM +If there is a way that the template can check its existence before generating it thatd be good +That means if I get by the same Id twice the same sql will be triggered to db twice by hibernate +Users were reporting intermittent NPEs when invoking the update action +One purpose is where token should be used and other is to store how token was issued +It turns out most of the time in the test is taken in executing the selects +See also and to a future release. +HTH. + works think I found where the problem is was set tothen set to if existsand set to againwhy +on latest patch +Closing resolved issues. +Not sure if we need anymore +Now the question is Should we remove anything but from the new API in +This issue has been triaged and decided to not blockpreventhold the EAP release +The behavior should be to put brandsNokia and get back brandsNokia for a dynamic field brands +sorry the title is misleading it should say similar to instead of +I think that there are certain times when including both white space and code changes makes more senseSure I am always in favor of common sense over hard and fast rules. +Heres the first email from theconversationSylvain what shall we doWant to bring this up on the listIMO it needs a vote before applying +java everything starts to work again. +Or do you have suggestions on should be done instead +libs +Thanks KondradIn the trunk the is under the module and it can be instantiated with a simple new +Going to bother about soon! +I would like to compare the results with openjdk but there the code base does not even compile +The best documentation is the javacc source of course but thats a bit unapproachable for a lot of folks that might actually use this I think +Having thought about this Im not really sure this quickstarts adds anything to the project +Im curious is DRLVM or Classlib using with values? Perhaps we shouldnt be doing that either +applies on trunk and +Implementation appears to be working +looks to be a super set of this +You should either select an existing folder that already contains a Drools runtime or you can the Drools plugin to generate a runtime for you +My bad I thought that one of the jBPM schema changes was between the versions from GA CP and so would need to be applied at upgrade for the standard install +overriding the persistence service is the easiest way to get it fixed +Seems like that some of rfaces javascript gets lost in multiple nestings via facelets +Committed and trunk +the sources of the project to start forge its gone +I saw this NPE too +in addition we need support modules for the supported cdi in +Brian do we need to this in Branch and tag fixed for instead of +Christian +you might have delete foo in the transaction log and the snapshot already has the node deleted +Can you resubmit this patch in unified format? just add the u option whengenerating itthanks patch can be ignored as it is dependent on the now omitted +Picks up the dead store in fixed in +It was an issue raised by our testing department stating the date in the column is confusing +bq +Please add information about JSF version +Id propose that any hints would be ignored if not recognized just as we do in other parts of the spec +It uses Rule Flow to orchestrate groups of rules should be included in the docs +Yeah looks like were doing at least one more RC +This provides deploynode for basic zones +The amount of working space to be left on each device is addressed by I think that if the disk is full we should just try again +Patch attached +I propose we patch for starters +If this continues to be an issue we should talk to Maven guys and volunteer to publish the released apis ourselves +THIS BREAKS EXISTING CODE however the current state of affairs is very sad Haskell systems understand byte values between +Reverted at +If you wish I can look into moving the compiler away from the Ant task and toward JCI +Actually this might be an artifact of the test +attachment had some problems +Cleanup of unclosed issues. +There is different code for that scriptsin almost all EAP versions which means its problematic +Not having this was a pretty major setback in migrating a major application from CF to Railo +thank you for reporting and taking this JIRA Liyin and Jason +I never got around to tracing who put the clean there +Like Dims mentioned we have both namespaces mapped to the same package +I agree +I am making this a sub task of for now because the implementation of inner classes will affect this we decided that the version here is the only one that is supported for now +Hello GiannyThanks for your review! Please apply this to g branchThanks be resolved +maybe adding some logging in all the spots that a flush would be called and why its being called would help also spot the problem +Agreed that its not obvious this is a good trade to make +Attached patch +Patch for View level features patch to fix some code formatting +I guess if we go with your solution well still be able to provide an id value from Cayenne ignoring th default +The first one to create Db create a table and one record in that tableThe second script to query the database created and select the row in the table USERCREDENTIALS +Still not fixed +This patch also applies to branch +However still open pending getting any confirmation back from authors clarifying which version of BSD license they intend this to be licensed under. +default I set my Major Compaction to once a week but on topic we need major compaction to run sometime even if no updates so we will able to remove expired ttl data in a timely fashion +Temporarily reopening to update release notes. +The mvn site target now generates the correct dependency information and the site was deployed for WSSJ at the release. +To verify if all storage pools are created we can use the DB dumps of simulator and cloud DB +But that really doesnt fit well to the general observerobservable pattern +It provides an option open when used will print the open files +xlsx files via the same api which seems more useful in the context +What about say connectionerrors? A database can fail at any time and thats something a application should handle gracefully +I check myself TS runs OK and configs are generated OK besides and are the same by default could this a problem Richard +this is a dupe of there was an issue with Jira and it seems that it did post the issue even though it gave me an errorThanks +Im committing those to catch any future regressions. +We are still facing this problem +I reopen this issue +Totally makes senseSubmitting an updated patch +Thomas whats the status of this? Please address this ASAP or let me know what options we have to address this +Check for null was added +Worked fineJake would you mind reviewing the later patch +from my pov this feature is obsolete now as we needed it to implement an workaround for jcr noted in the comments this feature is no longer needed by the reporter + +cd varlogcloudstacktar clf +Ill also be happy to provide any further input that you find useful for a quick resolution to this issue +lets see if this patch can be applied in propose that we add an option to the class to set the way should be handled +clashes slightly with this but I changed the code to use yum not retryyum and it works fine +updated in rev +Didnt change any other example settings this time +Currently you can leverage JCA inbound connector to expose EJB as webservice +Ok I have now tested the reported issue on Jetty as well +I have already studied the effect of the encryption over performance and I can say it is proportional to the size of the state +It is pretty predictable but actually unclear as to what repercussions is having atm +Pls verify itGood luckLeo +defer all open issues to close of resolved issues as part of JIRA housekeeping for release. +Remove all the tabs +is most definitely caused by the use of JDK +for preview not done yet +This has already been reported +In any case we should break the test repositories out into a separate configuration file as part of this issue +Hmm +perfsettings dictates which to use filemedline wikiwikipediaAlso You should use the branch BRDELETEOPTIt has the optimization you suggested on handling deleted docs +Actually attached update version that make the JDBC metadata functions throw +Ill update the patch +This version describes the default behavior by default the istat daemon does not run automatically in +Schlierf wrote +Im seeing a trend here John you also use Mac OS X right? James can you make it unanimous +You can have a look at the and see if you can create a case that matches the trouble you are seeing +There is a known error in the msdn +Copying ivy download library to lib is likely to mix up the source and temp files +The problem is fixed though for the branch and the subsequent jbpm releases that are part of this branch +this bug was reproduced only in version and was not reproduced in versions and on +Is this still valid +Apologies for the typo +Attachment has been added with description A unified patch against HEAD +I am going to commit this one soon +Reopen if this is needed. +on commit +I just tried the problem with Mozilla Firefox and Safari browsers and the problem didnt happen +Yeah that appears to be the problem +Seems like this JIRA is logged under trunk section in instead of being in was initially checked into trunk then later merged to +Running patch through hudson +Yeah that should be fine. +updated the path to fix edit list and view apparently the new patch is not being applied +Its up to the calling application to log errors at a higher levelColm +This difference is constant so the numbers get ut of sync very fast +The issue for me error is because I have TLD plugin enabled in Nutch and not included within the This patch should hopefully work out of the box +I get an exception each time I put hibernate in DEBUG mode +I useXloggctmpon region serverand when I see these messages there is always a Full GC happening at the same time +Clicked the wrong license committed to trunkThanks Mike +and there are quite some other possibilities for illegalargumentexception +David thanks for spotting this +If I add to the build path but remove the following three files the number of +Christian keep an eye on +Also this issue has appeared on some occasions in my project for the list method that takes no arguments +If it doesnt bother her it doesnt bother me +Attaching patch and corresponding html output for all the documents +I think the throttling will not work as desired if we use the decoding entities as provided by because the is built around the input stream provided by the entity which is doing the decompression for us +did that as a part of componentizing the impl given we are evaluating using this code inhouse +Comments? We have this running in production and it does works very good +IMHO this will break binary compatibility +Why do you think stack trace at the client would helpStack traces for this from the client actually hurts us in in that it puts a lot of noise when we are hunting down some other exception +The patch attached to this issue works well with without issuesIf this issue is up for voting I would gladly put in my vote +Why would those need to be changed? Each has its own CSS class you can change their appearance with the style sheetThe whole idea of Visual Themes is to change the appearance of the content not change the content itself +Lance can we get any more info on this? The exception in question to startThe first comment has all the details and a stack trace +Simple input text updated by ajsupport dont lost conversations on ajax requests +Out of order with respect to whatCan you explain in more detail +Lets leave this attribute as is +From this information we will write the final release notes for all issues +May I suggest that we drop waitingtasks? It has only been committed to in March +Patch improvements thanks to little improvement avoiding checking ruby the fix in! +The test failure does not seem to be related to this patch +The exception to this rule is when an indicates a minimum time of unavailability +Thanks for the bug report +Closing as duplicate please reopen if this is not the case +Fix is now live +So it may not be necessary +Let me know if something is wrong +closed +which may or may not happen even for graceful shutdown as we kill NM after some time +No test looks like no needs to set the participant count actually +too complicated going to use the dynamic draw into splash method +Yes I think Jareds point is good and it will also help improving code cleanness +Apache Synapse and Axis related projects are the projects that I was involved in +From what Ive seen Eclipse does its job right +I wouldnt even support the old behavior not even with an option +Please find the attached patch to create blobs and clobs dynamically +You do the converter registration directly and in the same thread before you call an operation which uses the converters right? Nevertheless the issue shows the typical symptoms of a concurrency bugI think we can assume that basic converter registration works in otherwise we would have received numerous bug reports +Hope tomorrow have things running +Now there should be only the stuff from Rubert missing +Have a feeling its somehow related to using +Lets make this the master redesign issue +So in general we favor abstract classes +Since then the RRD config files have been fixed so this is not really an issue anymore +Just to make sure I understand is this with serializationdeserialization? Aside from automatic use of base encoding decoding it seems dangerous to assume transformations between the two most likely should throw an exception with meaningful description +Do you want to remove it +Todd Yes. +Patch +Introduce and fix any compile errors this removes a lot of the casts +Can you attach a patch for trunk? We need to fix this in trunk as well right? We can commit the patch at the same time +Yes it works as long as we also fix +You cannot create the table because your warehouse directory is writable only by hdfs user +I believe this has been handled better already by linking the Smooks editor to the file for easier opening +Slowly started implementing MSV now parses and retains default values may be tricky to pipe them through as no support whatsoever exists +issue is too generic +But I am not willing to change the output format for operations +This patch contains fixes only for the api source +Thanks very much for the comments and suggestions Dan and Christian! I believe that I have the interpretation of NULLS FIRSTLAST wrong in the prototype code and that Derbys current behavior is actually NULLS LAST is the default for ascending order and NULLS FIRST is the default for descending orderSection of the SQL standard is the relevant sectionIll update the wiki to incorporate this revised understanding then Ill have a fresh look at the code and in the process Ill incorporate the other suggestions +MINGW takes the nix approach +Just reduces the burden of writing applications on yarn +Attaching the second rev of the release notes +I have cloned the Git repository and so have Grails assume things on this mode are tracked via the Grails JIRA +Im not a CVS expert but Im pretty sure Im doing everything correctly. +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Everything easily doable or critical has been done moving the rest of these tasks out to move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +This would not be conservative locking now +But its an other discussion Im not totally sure of what the right setting should be +NET based client of the Axis Version files of Synapse versions based on marked revisions of file +Ill take a look at the JCA code and see what it is doing +I think M to M are a lot closer to the current sweet spot +I can set it to something else on the command line and it will change. +In constructor should we check for blockId? Earlier we checked for blockIds array +In my opinion the title of this issue is in collision with +When we call the system call exit this is triggering the atexit routine that is calling JNI methods that can never enter as the system call doesnt transition the virtual processor status from INJAVA to INNATIVE +Closing all tickets +Our entity resolver expected public entity HibernateHibernate Mapping DTD EN while example has HibernateHibernate Mapping DTDEN +I know there was an email about internal and external API docs that I didnt quite Permainemy mistake it wasnt an email it was this JIRA IsaacWhen I have the docs zip ready I can point you to it and the javadocs will be in ll probably need to contact the dev team to find out which external thought the decision was to only have the Guides online +We are now raising our expectation for the fix +Ive verified this we have tests for this issue and they pass. +There are some logic flaws in this file too. +Therefore resolving the issue +reopen to be able to target correct fix version. +Oh BTW this goes for a Map being returned as well +Otherwise I think it would be nice to implement reflection through a separate abstraction where you allow clients to plug in their own implementation in Avro +this was caused to the bug issue +Need to port from +Does it not +Duplicates +Will be done before weekend +no we dont at least not in this particular casebq +we can probably find lots of example to support both nomenclatures +Just invoke a different port when starting it from the command line with a separate SOLR homeMarkusJust to make sure I understand this sends ALL the documents to ALL the SOLR instances specified right +Duplication of of +Patch for trunk +I hope that no code has to scan the directories at all with this API +thoughts on this guysThis affects the letterhead that is on every document in our software +Thanks for the updateThe patch for seems to include some Chinese characters +It needs to add the ecj compiler to the classpath before the Sandbox module is loadedBy itself the spec runs but it wont work if you run rake spec +I guess that it is the same issue +This works now by changing the output folder for groovy separate from m still having this problem even when using the separate output folder +There is nothing unfortunate about that its the API +But now we have a WAL which contains some good mutations which we need to recover and a checksum error near the end of the file +There is no way to fix this until I can finish the +Release NoteOracle users can now specify the database property for the ESB Management console as oraclei or oracleg instead of oracle to avoid warnings being displayed at startup +wsil +It can be applied and work together with the patch in this jira +I think its just as likely that someone wants to use bundleutils without the finder as finder without osgi +to use an inner classPerformed Checkstyle checks +Im going to keep there caches in a topic branch for the time being +If you get a chance to test out trunk please comment here and let us know whether or not it is working for you +The questionis if theres any good reason for isOpen to call isStaleRolandI see one problem with this approach +as r + code just committed this +we should continue to test with a not allowed user and add a new test to test with the other allowed userThe current patch does this its just a little misleading since Eli changed the comment above the disallowed user case which made it look like he removed it +Already fixed +Ok I had the purpose of the class wrong yesterday +I see the same failure in +Can you please try this with that version to see if your problem is still present +This is a serious project and I need support to get these fixed +Hope this finally fits +But these are all on secondary indexes +Assume closed as resolved and released +Issue may already be resolved may be invalid or may never be fixed +I mean SVN +Yes that is correct +I run some fairly complex objects through and didnt see issues +At this step should have been created +Works for me +Thanks for reviewingSince both offset and length are non final variables JIT might not optimize and compute this for every iteration +My guess is that this problem affects any container using XA transactions with the resource adapterDo you know if anyone is actually using the for XAThanks have committed your suggestion on a working branch and I will merge it to trunk as soon as Andy Taylor finishes the maven integration +simplecbr handled in +This looks like the right thing to do +Im happy to commit unless you would like to Ted +committed to shows all till Halting VM +removed parameter for attempt as suggested by Todd +Killed tasks do not complain when they have exit codes from their process + scenarios can now call out to any other code so can reuse fit fixtures if you want +Its more likely to be shipped by them rather than as referred the guy working on it there to this issue so hes seen the contribution +Wow thanks for the fast resolution here! +Id like to see the message improved for EXCA property named must be specifiedCan you add to this message? EXCA property named must be specified or the jar file with the default services entry must be in the classpath +I have added code to create the link to the RSS feed but I have commented it out +Well review it later and will describe what we want to do in a new issue +Thanks + +Resolved in r. +Please close this issue if it works for you +I will provide the patch for when this one will be + +If so lets just close this one +Committed to and trunk +Thanks for the patch Narayanan +thank you for your review Lewis +Martin is there any way that you can enable the base RHEL channel while running the test? if so you wont run into the unresolved dep issue +Ive already bumped to blowing the version. +Im going to leave this as unresolved until we get clean regression runs across all the platforms with +underlying AS issue has been fixed in master and as resolved see issue is not fixed in EAP +added a test removed some duplication in the code replaces last this patch be merged into svn +Grant your response confuses me +Some would argue that having runtime exceptions declared in the throws clause is confusing becauseit is not guaranteed that only those exceptions listed there will be thrown and not even that these exceptions will be thrown at all +Manually testedTo do this test I set a timeout threshold of millisecond and so get a Timeout exception +Thanks for applying this +We didnt find much interest for it back in or later +If you run you should no longer see a stack trace in the ant console it should be in the log +bq +Since you have a trailing slash in your URL the relativeness of it is one level deeper than is needed thus you get connecttwittertwitter +I wont necessarily link this to long running YARN though other than the bit about the token +Im sure its not unsolvable since there are plenty of log analysis tools out there +service name can only be +feature members not marked with a required can you close this issue after verifying my tests above? If you set yourself up with a Jira account I could assign this to you for your feedback +I dont see why we need this if you can already do it in provided that throws the source exception out to the container +I checked in an automated test for this closing. +Now that work is well underway with AS all previous community releases are +Ill be happy to rebuild site today and incorporate these changes +Mr +Why do you want to put the logic of generating a message for all sorts of different objects in new? Always put logic as near as possible to the object that is responsible for what youre generating +Committed to branch and trunk +to trunk and. +Attaching patch that fixes the issuePlease review and provide your comments +Mice can you take a look at the latest failure reported by Prashant +Ive to upgrade in templates plugin otherwiseanyway +Its also based on +What do you think about leaving out the changes to for compatibility +This version updates to say that has an example of a mv script patch +RainerThe dependency for is only for unit testingI have set my own category in srctestAnd the unit test tests that this worksI see possible solutions if should be removedWe could rely on logging but then I need to mess with how you set logging parameters using loggerAnd I think the code in is still on at least I think I saw that in the ant build filesChange the unit test to skip silently ignore this testCan easily be done with a! null check +This was just okayed +Can you confirm? Are there any tests done to compare vanilla and the default plugin +We have various developers working on trunk each of us has many data directories we use to check backwards compatibility in all of those is defined the old way +in the inefficient flushing of as described by +Hi Wei Hua can you check this again? I think the code is writing out values in mm which is of cm not of a cm +Plugin is now installed. +logs from enough information to reproduce this +Thanks Rajesh I updated the issue +Writing a custom Filter or using are two possibilities +I hope someone will get inspired by this script and actually will write the perfect script +I do still think the issue needs fixing since Im sure some people use solr to index large web pages which do need html stripping +when expanding images node everything works like reproduce anymore +Thanks Xieguiming for the patchThere are few of my observations about the patch +Good progress Brock +OK I see +Sounds like the patch failed +Just not all of them + +Well add documentation shortly +We could easily package this to be available outside of the distribution if that is deemed necessary +You probably dont understand the specification very well +Attach the +The NOTICE file was already updated to include Iconic websiteauthor in the patch for +It seems like this patch cleans up a lot for the cql use case but it is adding to the overall mess that we have from supporting all these different implementationsThe storage engine is too tied to the public and CQL seems to just be pushing itself along side it nowThings like is obviously a problem and it makes me worry about the maintainability of this kind of change +We could use a pre compiled pattern instead of replace from String +So what is the problem +attempt ssh login to from kvm host +I already increased it to secs +This is in way of a release +Building twice seems to me to be necessary +Committed in rev +There are overlapping entries in table in version +HiThe aries transaction manager both exports and imports at +now showing startfinish times in separate columns +reopen to fix version info +Could you verify in your environmentThanksJim + for the patch +PaddyI have commited a new fix to trunk about the move bug +Im not sure we can do anything to fix this as the documentation youre referring to is generated automatically by doxygen +Today I started the server once again and missing RF have appeared +Reopen if one is available or this issue is still seen on newer releases. +But folks should still be able to scavenge their data by setting a config parameter that permits them to access the data even if it doesnt have a checksum +But they are affected by inliner +Jobs cant implement callback interfaces like directly because they are not managed by a but rather by Quartz itself +I wonder about IPF architecture +Yeah +I have verified that it fixes my problem +Add Generic support for ThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Now that work is underway with AS all previous community releases are +Field table updated but main issue of JIRA Missing DT and F fields not implemented however JMS support has been Smith to complete this and Timestamp were added at revision rString and Int were also moved to to be consistent +Committed at subversion revision +Fixed in revision +The aciton prints message This should be second On the first node there is a event that sleeps for seconds and then prints message This should be first The expected output of the test is seconds no action then messages This should be first This should be second If transactions are not working then the messages comes in opposite order +Tested on branch +The was a sample esb archive created by Aaron to help show the problem with the steps he outlined in the last comment +Its something I could potentially look into during the winter break +Thanks for the update Kihwal +bq +Thanks Stack for your review +Mailets that depend on server beans still reside in serverThis was the case for the mailing list ones. +Fixed +The former actually tests a PI outside the document element though only for HTML +Dont worry about increasing the time beyond d +how does one create a diffpatch without manually going through each line code? Is there a tool that provides this? Assignee Igor Vaynbergplease create a diffpatch +tags. +Ok then Ill proceed w +Committed revision +Please see initial version +lets try to keep patches in the same JIRA issue so they dont get misplaced. +To see if its actually and not ASCII that is being used you could add a character to the message temporarily and see if it is detected as then +I think we fixed this in trunk? Can you check +Verified in JBT is no way to disable this particular check +This would allow you to write a file utility for processing compressed or uncompressed data +With new method introduced in readers and writers this should not be an issue +As discussed on the list JSF will not revert the change +I was on vacation since mid last week +Can you check the log for this test also? and attach if it contains errors +It surely not logic because all of them define or implement the same interface so documentation should come from same place but its an other issue +Most of tests are failed because the console navigation tree change we need to update test code accordingly +The recursion is handled incorrectly in some cases +Looks good to me +This patch sets is to false when moving to the next row and adds tests for it in langI have run successfully run derbyallCan someone please review this patch +If it does in the end need to change I am curious the behaviour for soft upgrade and downgrade and soft upgrade again +With the above solution described in Rossen Stoyanchev comment isnt this still a problem +Mark since Jenkins wont be able to test the patch could you run ant and ant and then post the results +Im moving all bugs out to and we can move some bugs out to to get some release action going +Ive used the same JDBC driver and mysql version for both tests +I think this can be closed as it is outdated +Thanks BrianEugene +Would you provide a patch +Thanks for the contribution +So it does not appear to be related to JBDS bit or the JDK versionAlso the bit JBDS was patched for the new teiid plugins and the JBDS bit was not patched +is very close to merging into trunk at this point so these issues should be resolved. +Other people will be easier to review itThanksNicholas +Thanks for following up! FWIW theres also a lot of String and fixes in that may repair the original +has been marked as a duplicate of this bug +I will commit as soon as I can +CP fixed in revision fixed in revision +Would be good if someone else also checks this +Will run tests and commit +Doing batch processing +Sorry for the meager response +Now the CMIS Repository Connector uses the GENERAL code for exception handling related to a +Hi guysIm in favor of this move +This seems to be missing documentation noShouldnt the key at least be in +Default value will be switched to true in release. +Router iptables rules will not help much because router is not acting as firewall hereFrom SRX cli run show configuration display set and attach the configuration here +doesnt make sense +In my opinion this patch should be alle poms and added the scm item where they were missing +This high CPU load was caused by not enough heap size for the jvm +It appears that the usual way you do this in C is via a configure checkI think this is too far outside my expertise now since I have no experience at all with configure scripts and the typical C build toolchain +Patch applied at +If trunk is safe then need to fix only on branch +I think I succeed to take account the changes you asked me +This patch fixes that think that we will want to change the extends definition of the generated component definitions to point to baseComponent +Closing released +when parsing a the endpoints arent runthrough the analyzer so they would end up not matching the terms in the indexthat were analyzed +need more of a bulk update all resolutions changed to done please review history to original resolution type +Trying to understand what Tommie was doing I got to the following conclusions He uses which offers modes of operation + +Committed after testing +Hi! has a typoencryptionCypto and notencryptionCrypto +Thanks CamilleWithout proper unit tests Ted Dunning requested I didnt regard the patch to be available for reviewPardon my laziness +I committed the patch to branch +Caused to OOME on trunk up on Jenkins +Postponed +Verified fixed at rev +Its possible for Archiva to not use rsync +Perhaps we should move this section though +Thanks ArunYour patch is in trunk rev +Document as a KI +Were there any code changes that need to be released before a fix is availableOr is the user supposed to set LCMESSAGESC in their environment +The output is exactly whats expected a and b c +Alexey as you suggested method is not thread safe +Maybe we should clean up the core so that it is easy to separate out different plugins and associate these deps with the plugis but we have yet to get to that stage. +In the current implementation of the runtime this test will require two worker threads to run successfully +I did not know it was out there +It allows us to rationalize the dependencies in our may think this is a bad practice after all everybody is entitled an opinion +Please delete the previous attachment upload was interrupted +LuizA fix to is probably going to mean a so best to wait a bit on this +Try Hudson +We can cover the simple case with the normal BSP and as well but can be much wider usableAn interesting point is as well that this system equals to Dryad +Resolved +It uses the JS for and therefore supports +Patch Committed revision Thanks Viet +Perhaps there could be a flag in the files themselves +I click the link and get a KB file instead of a new page of mirrors +mail bounced +Just leave a comment on how to get those if needed +For example if I were to implement I would always expect inputoutput to be Double +but instead of silently swallowing the exception maybe a WARN would be more appropriate +At revision it is look OK +Using a boolean to avoid a null check seems to easily cause trouble when the two variables get out of sync +I have not checked the server log in production we do not have access right away to the log files it is why it will be more suitable to have An standard error of conversion displayed at the user screen The screen shot does not show the problem because in fact we have no message +Release note looks great! Thanks +No problems here to please rename the branch as you see appropriate +Gail here is the CDI part we discussed during the meeting +Patch for lines lost in broken and push +Reviews and applied the patch thanks to Corey Scott again. +Even if i push up my retries to I stillget the Unable to get jmx connection jmx not bound +Fixed on I see no risk in including this in providing there is still time to do so +Hi SebbOn this we agreeHowever the current script is a and I was hoping to stop the bleeding as it were and then figure out correct placement +Add shell help for these new tools +Im attaching the two test runners I have written and granting licence to ASF in case someone wants to play with them +The easiest workaround Ive found is to explicitly name the version of the archetype plugin like this mvncreateNote that youll also have to explicitly specify the archetypeVersion was just informed that a total rewrite of the archetype plugin has taken place and it is available as +Unfortunately I had a nd erroneous binding that was in the UI thread +But that isnt relicensing of the componentsNor is it necessarily sublicensing which is why I suggest that sublicensing is often unnecessary as long as you have appropriate license rights directly from the original copyright owners as Apache does with its. +Im told this enhancement is imminentWith regard to exposing powerful methods to remote actors the broker provides an ACL capability to selectively allow or deny access to methods for specific usersTed +I should have cycles to get it into the trunkYou like Override +Reopening to fix for for cygwin +Yes I think that would be great +the code above compiles with no errors +Still needs to handle cases such at telelportation and rank sinks +Does anybody know about some other library that has a Apache compatible license and that is maintained? +Grzegorz this is a tacos issue +Oops! Even i missed that +After thinking about this for a while Im fine with this change +to look similar to the other method names +import This is very useful code for us +Attention The file tomahawkcoresrcmainresourcesorgapachemyfacescustomdojoextensionsresourcewidget gets one more time to Stefan for fixing a related issue. +They perform seperate functions +the custom is copied to in the exploded webapp which is what the code you intended to change is referring have enhanced the test to ensure the bundled is correct +Go ahead saminda +All tests pass for me on the most recent version of DRLVM on Linux +Could be specific to browser version and the one mentioned in comment above is a bit old in fact +Disabled nodes should be excluded from KB descendands are excluded as well +moving all unscheduled items to so we dont miss them in initial release triage +Test worked properly after common was excluded from lib +This is also the case for the ASYNCTASKS in the class +The xref is trying to refer to one of those missing sections +Perhaps we could simply maintain a wiki page which listed current dependencies or have a maven goal which checks different versions? s Im the one who will Im NOT the one who will +Works +yeah it is fixed on the tag but i did not test it properly ping +I thought that there was a OasisCMIS Jira issue where the claim was made that were not versions but I cant seem to find it now +Would it be possible to create it relative to the trunk directory +I will add all the bug s and the master in the FS +It is not clear to me why the component includes Network ClientIf it was not just a click too much do you mind explaining how this affects the network client +I understand your point about developer sophistication +q files for different Hadoop versions +Thanks David for your fast commitsI have now update my local build and now everything is okThanks also for your suggestionsSee you soon on JIRAMarco Risaliti +I would also really need this functionnality +Failures are unrelated let me run the test locally once again +Could you provide examples of what is currently generated and what youd like to have generated? Which file is this inAs a side note from your description it looks like the console is adding the security realm gbeans directly to +Added support for this under in try it out and give me your m not happy with the name so any suggestions would be Ive moved this to the sandbox for the time being +That being said I do think it would be cleaner to have the kill method check for the DONE state and do nothing if it is in the DONE stateI also agree that once RM restart is in place and it can pick up where it left off we dont need this any more +Attached as v is a patch that does not wedge under load +setting fixVersion to for this nice enhancement since there are items with higher prio +There are other places where the setRGB method is still used but those are probably best treated as a separate tracker itemThe performance enhancements described under tracker item are a separate issue. +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +We will see +Corresponding changes for coresrctestresourcesorgapachestrutsviewsjspui to pass build tests +The testcase must be compiled withjava jar +closing as cleanup +This is important to note for several reasons +This is fixed in bc. +They arent wired up to the installer or tested by anyone other than me so the work is not quite they are usable the problem was they were not added to cvs w the kb to fix for the next drop if want to try em out in the meantime I can email you or attach a +Its solvable but I think we should open a separate issue for that as its related to a lot of other things that need to be fixed with containersI think youre right about the isset branching +bulk defer of unresolved bugs to +Ive reproduced crashes on CPU machine with +The +site passed +seamtests +Added +Codes look good +Just load up the project and go to the show page for the author +That avoids duplicating code too +I use for our spatial backend and I know that they have support for transforming tofrom +Thanks Kelvin +In the file when a mirror is defined the child element mirrorOf seems to be is this not a defect +Please confirm that they have been applied as expected +Dense and structured sparse matrices Algebraic operations on mapreduce IO operations with matrices and vectors MapReduce Input Unstructured sparse matrices +A patch has been submitted for this issue +This works for me now +I can switch off most of logger messages by running mvn q +we dont have a test runner in Griffon right now like Grails has +hod contrib was removed +Committed revision at trunk. +Quickstart example demonstrating the bug +At some point the tests will also be split between core hdfs and mapred +Does this help? Is there any weird behaviour left? +Reported fixed +Another useful Element would be for header meta data tags which are used by search enginesThis stuff should probably go into its own package MalcolmI like the idea of a more generic approach to HTML headers and housing these classes in their own packageHow do you see Page and Control imports integrate though? For example will get return the Control imports as well? Or will Control imports be specified separately with the cssIncludes and jsIncludes variableskind regards also thought about leveraging onRender however users can skip this phase by returning false from an action listenerAt least approaches exist to add elements to the List +Yes I think Scott has a point +This would be really nice to have for upcoming have some initial patch that is likely to be contributed shortly +Domain transfer requested +The root cause was eventually tracked down to the disappearing documents not falling into the hash ranges of theDid you investigate how that was possible? It sounds like a bug in the hashing code either in the routing side or the splitting side +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +nit the JIRA number is wrong in patch filenameOops Indeedbq +Verified in both standalone and embedded servers in CR. +A patch to do this +Trust me people will forget to do a separate build for rat +So here as well we shall move this code to +By change here I mean a complete gt connection pooling rewrite not something specific to Oracle but something that spans all JDBC data stores and can work with all connection pools you can throw at fix attached to this issue is effective but works only with JBOSS connection pools and the Oracle data store +No response +I updated documentation yesterday and it now reference git usage and mvn clean verify. +Attachment has been added with description the original has been added with description the FOP has been added with description the FOP disregard I have to test this with the latest changes in the coderepository which I thought I had but in fact didntSet to invalid for transition to closed remaining resolved bugs +We already have a class to test Tomcat specifc stuff +I will again look into it. +fixed in r +These are the jars that have been deployed into our livetribe maven longer needed since jars did not survive the disk crash +commit fbfceaeccfcceaabed +In the meantime this bug in should be fixed +The more I consider this the more I realize its a blocker + Patch for for crs in mysql on branch for crs in mysql for applied to as r +Alternatively if you have a reproductible test case that illustrates one of the issues describe above then please reopen and attach any necessary files +Dont know exactly why +Alex could you review this change pleaseThanks changes look good for me +Thanks Uma I am working on breakdown patches and creating Jiras +Patch for trunk +It will skip the second group of radio buttons +Ryan Yes +Namely the Python command line tools dont report the ACL errors correctly yetRajith I am assigning this to you as you are doing more testing on the ACL module and are probably ina better position than me to know when it can be closedCarl +I think that a message should be logged saying that certain API actions like SWAP and RENAME will not be available these should not become unavailable because of something like thatbq +changes are made on should be can view monitoring result on the machine that runs the server by going to httplocalhost +Patch is available +From left field the name Grouped Faceting or some variation of it comes to mind when I think of this functionality +As I said already this seems to have been fixed already. +Thanks +Note that sets this property based on an environment variable which is itself derived from and therefore most of the other references are really coming from what was specified in and not what was configured by an admin +Tony I think you can delete your private data as we actually did repro the problem just did not know it +The factory can be passed down as a query closing all issues that have been in resolved state for months or more without any feedback or update +Bulk close for Solr +audit has been setup as per Jims request +Committed as svn Thanks for the good work James +Thanks for the patch everything seems to work +Not sure if this one is still valid and cant even remember what exactly the problem was but since youre working on WCS these day you should at least an idea of whether this one still makes sense or for absence of workfeedback. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Heres a version of the test we can use to to as this is a gotcha that we could usefully fix as we already have the change +Can you rule out the indexer and see what you get when you dump the three segments and the merged segment instead +Maven core has been updated in rev +Please verify. +Inside the tag you will evaluate the expression using the choosen languages +Could you elaborate? Im not sure to understand your needs +Any chance these fixes will make it into the release +So far I just made sure the unit tests are ok +Thats supposed to be +Hopefully jenkins wont choke on it +this is only an option that turns on the path only on the URI if and when needed +NET Beta prints No symbols are available for this location when stepping onto anything different than a for placed onto anything else than a for loop disappear when starting the program +Thanks in advance for that +Can you please attach the whole test log to the jira +Please verify +This has gone stale +If I work on a patch for this can it drop the isBook flag from the and move that logic to a subclass like the Xhtml case +Yes +Note Lucene s aka column stride fields will be of benefit here in efficient implementations +The order of arguments is significant there +Just corrected some misspelled words in +Hi Ted I totally understand your concern and appreciate your feedbackIt would be nice to fault tolerant all kinds of DNS server failures which could be transient failures loss of PTR or DNS service crash +The patch included a form sample +Lets take the opportunity to nuke it then. +Committed to trunk. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Uploading the patch to resolve this issue +Im trying to fix it +again a cumulative patch following the werners previous comment +closing again +Checked in with revision +bq +It is a bit more work but would make us completely independent of blueprint and spring dmSo I think if we can assume blueprint is always present like +Id like to see the exception tests added +With slated for we need to fix this since crcs are going away +And the reason I keep raising the resource issues is thatI think each time we hurt a feature of Derby that is unique and does draw its use +The patch also has changes for completion +If you dont purge logs then eventually the partition fills up and your cluster fails +The probability of it happening by chance are low enough that having it happen once in a while is more confusing than useful +We currently use the Java Properties class to load the properties file and the provide the properties as is +Ive committed this to trunk andThanks everybody for the reviews. +We are using Groovy as main language in are web framework jPALIO +It contains a type system for simple and globus application deployment descriptionPlease apply it to airavatatrunkThank patch from Patanachai +Strange I focused on the pluginsystem test case that by the way only works in case the inlude pattern allows all plugins +In fact the kee of a should always contain the kee of its parent to prevent such kind of issue +Linux kernel bundles compatible drivers in the source tarball +This is essentially option +Thanks for the patch Sylvain +Punting issues from RC to final +Is it okay +I agree that getting fairness to work is a big plus +It appears to me that the patch has been ported to +I think choosing default and going ahead would be this is a transport level error May be we should return Unsupported Media Type HTTP error code. +I will update the patch with the modified condition +Verified on revison +If so Ill commit a new to for previous comments. +Sorry null is not represented by an empty tag +All unit tests succeed +Its going to contain library upgrades only +The SIGN function actually returns an INTEGER or +uploading a interim patch +This will help in testing distributed cache patch. +Do you want to throw this one in that bucket as well? Might be more efficient if one person just bangs them out +Just tried upgrading AS trunk to of to see if this issue is fixed + has been marked as a duplicate of this bug +I am missing anything +for option updating the user guide +Hi George DIGYYes this is a direct port from java +Id hate to see this get put on the back burner +Yay great! Thanks! +Issue is out of date in WFK no need to release note it in WFK +Shahar I have corrected the issues that you mentioned I believeDefault sorting is now count if a facet limit exists otherwise indexI fixed the stuff and I went with the sorting convention that the rest of solr seems to have which is that null values always go to the bottom of the listTry the new patch +This looks like a dupe of but you didnt a attach a test project so I cant check +Accepted for you check the current installation instructions +Closing as such. +Please keep the feedback see I was doing it wrong +The problem is the intention of doing this what youre really trying to achieve is isolation from the public repositories right? Otherwise why redefine the central repository? So if youre doing that you really want your overridden repository to be used by everything in your transitive dependency graph +Please note that bug contains an additional error case not described inthis bug that is mishandled in as well +It also changes startup order in hrs so we have zk before we startup up rpcNeed to test this patch still +The simple way to fix this was to ensure that all classes produced header files +If you had time to edit the and files accordingly it would be superb +It sounds good to me +I already have the code ready for this unless youve already done it +I was wondering if there was any activity regarding this issue +OK I see your point +has been replaced by in the JPA spec so that should be preferably be used going forward +Another try +Sorry about the long delay on this issue Kim +Thanks Nicholas! +The only modifications to my BG installation are the change to the guestfs helper and a change to the rhel plugin to stop it from automatically addding core and curl to the package list +This preserves the user security information thereby optimizing overall processing time +Thanks issue is closed now +Note that the SVN Trunk code which will soon be the release we dont do things quite the same as we are using APR for a lot of the platform independent code +Im not sure about those task dependencies + +for the patch +There was a mess with suspend enable and disable statements in VM on interpreter code path +Also gtkgnome do this and do it on windows as well so Im sure its possible +Fixed in trunk and branchOleg +it fails on trunk too +and all parameterless annotations in chapter EWP documentation is different with only fixed items in and +Version Final is released +Minor TomYour patch is intrunk r r r +Out of date for all Tasks and Feature Requests created before +Since there are quite a few deprecated methods and some of them are quite involved lets create for removing them +Let is evolve naturally for some time and then we can see if a refactor makes sense +One of them is java jar indexer another could be Seam builder or +Hi GaryForgot to ask this other questionCan both user and program gain access to data in the current access control implementation +Unless you are a Chinese PINYIN speaker it shouldnt matter +It should place warnings in text editor for tags that will not be rendered during +Hive always works in side effect files and then does moves as the last step +Thanks for the patch +If you generate the Javadoc on trunk youll see in the jdbc package javadoc some more info too +But please make sure a committer s the change +Eddie this is the patch to update the controlhaus controls +OSX dmg also installs files with no execute permsls l bintotal r dvb dvb Nov griffonr dvb dvb Nov r dvb dvb Nov r dvb dvb Nov r dvb dvb Nov startGriffonr dvb dvb Nov +However there may be some need for this to move from to I vote we move this issue to +Ive upgraded to but I havent made any configuration changes +Id like to add an annotation that would be used on a transient property accessor or in conjunction with PersistHowever the order in which method annotations are applied is not defined and would have to occur AFTER Persist if any +Unless all referencesto the previous deployments classes are free to be garbagecollected there is the potential for the NPE whenadditional classes need to be loaded when access theprevious version classes +your test case is trying to test if something does compile or not +This should be fixed see you lock the javadoc plugin version in your pom +Run in Jetty go to the root page then click on the Brands page link and the NPE occurs +workload e went from mins to mins +As a rule please try to avoid elementattribute names that have a Choice suffix +The xsddoc do not bring much value so I am looking into removing it as well +cant release anything until we fix this issue +This is a bug I guess +anything relevant in the +Group created +Is this still an issue? Im going to close this for now since its not been fixed nor reported in quite a while. +Thanks for the patch! Well spotted! +still an issue in actually resolved. +I think this issue should be reopened until properly solved + +Patch that adds all the globals +To digress a little bit I would also like to suggest a implementation of the ratio test as described in Teds paper +also inlined the comparator replaced the last synchronized methods in w volatile fields and removed get +and set my JAVAHOME to that dir +I met the same problem +protected Object list protected String listKey protected String listValueI know its not strictbut its satisfied in current versionmaybe it needs refactor when adding new subclass of contains field type is not protected String in the future +For what its still worth I was running on with Sun JDK for applying the patch +One idea would be to create a special comparator that looked at row and column only and ignored timestampStack it still seems pretty clunky having different KV comparators that stores must be aware of +I completely removed the dependencies on Synch and Val and needless to say all tests pass +is changed within JBDS initializes EAP AS correctly +findbugs is because of which is unavoidableCould you suppress this warning in srctest +Patch to set the right just committed this +Did you wait long enough to be sure it passedYes Ive waited a couple of minutes now and still nothing +Your stacktrace indicates that you did specify web and went down that codepath as a consequence? Otherwise how else did you arrive at that stacktraceIn general setting the web property does require at runtime +Assigning to myself since I started writing something +Change looks good to me +Anyone know if this is still an issueYes +Sanity check Are you walking the tree for purposes of reporting diagnostics about Xalan or are you just leveraging our existing code for other purposesIf the latter I have trouble counting it as a blocker for Xalan +Chad I upgraded your account so you can assign tickets +I vote for too +Indeed +I think creating tables would be more than sufficient in this case and bring it down to +Maybe using the when no realm has been set will still be neededI am not closing this issue until some feedback on basic authentication is providedThanks for accepting my contribution to your great projectMathieu +Here is a second patch that brings a few new methods to the Avro API +Punting minor new stale patch. +On commit add class comment because is awfully generic name +in this case it makes the administrators life really hard because there are far to many options in the sling authenticatorand an admin fiddling around with it to finally get hisher webapp run will not be aware that the very config option thatwas fine for the regular webpage just broke the webdav supportand probably i wouldnt even care about if it wasnt me that had to run after webdav problem that turns out tobe introduced in sling by people stating +Possibly changing the method to something more generic like set +I think this can be can be marked as wont fix +Thanks Jason and John. +This is still an issue in for Android +Brice again! Ive merged your change into and it will be in a release +If no one objects I will commit +Hi WilliamIts not about the name of the option but its potential semanticsthat seems to have an inherent dependency to the other two parameterswsdlURL and serviceClass +however memory should be in the clobber list please try the patch +I dont know if we can do anything about it +svn statM javaengineorgapachederbyimplservicesmonitorM javaengineorgapachederbyimplstorerawM javaengineorgapachederbyiapireferenceM javaengineorgapachederbylocM javatestingorgapachederbyTestingfunctionTeststestsstoretestsM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTestsutil +Thank you Joe! +Will have al look +Excellent thanks Josh this is just what I was after +We need to fix it +If possible stick to the test and then rely on DFS behavior as guidance if there are differences. +minor fix +So I added namespace based filtering in +Mark is right about the Affects TX semantics were removed from the message aware listener but I am just about to put them back raised an issue with the general SQL handling just before the CP which suggests we were lucky that it worked in the first place +Will run all the unitsystem tests and post the results laterWhat are your thoughts on the timeline for this? Ive been assuming that wed wait at least a few weeks for the snapshots work to soak in trunk a bit before to as weve done in the past with other large features like this +is used with an indexed collection +read etcgroup rather than exec a commandIn general its best to design things so that we do not require native code for decent performance +I should have looked at the list of targets more closely +Good to know +No child dependencies though +I created instead of this +updated patch for modules beans concurrent imageio instrument jndi +Getting Started Guide for JSF with Hibernate chapter is commented out +I close the issue. +Any patch will be accelerate the schedule +The first returns all the databases the second all tables for a particular database and the third all the columns for a particular table +Yes following are supported +create a patch to unwrap the classloader from temp classloader +If the checksum algorithm for were to change it would likely be something like Adler faster than CRC but insufficient for small input like +Due to snapshot chain issue even though user issued deleteSnapshot command we may not delete it from secondary storage due to its dependency +it looks like it was a problem in the maven snapshot I was using when I published this last +bulk defer minortrivial issues to +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +you are great +maybe a reusable but properly implemented shutdown component would be handy +I will create extra tickets for think you can close this issue finally +So I wanted to separate out that caseRegards dont agree to me the first occurrence should be part of the function out put of re also this name is ambiguous the name of the method depicts that it recreates the template but it is not it is just a conditional checkOne more thing is that according to the flow if we are to keep the above two cases as two separate cases we should be having the second case firstJust a few comments but this patch is very useful +Thorsten any further feedback? OK to close assuming the bug is fixed in +I have parts and working now +Can someone else too review this who is more familiar with hfileblock part +Confirmed with engineering that this bug has been backported to and branches of Fuse Camel +Hive happened to expose the issue with one of their filesystems but the issue may break other filesystems +Hi againIve tested now with the new version from the repository and it works for me +As a user I just want to point out by a functional point of view that I had the same workaround has been to ensure that all empty lists were actually empty lists rather than null pointers +Christian thanks for the patch +For example if I want to pass the token as a URL parameter instead of the HTTP support this scenario i suggest resolution A suggested by Jon Osborn remove the final +It might be CRLF issue +Gregory spelling corrections verified at r +Patch against for those who get those +yes my mosquito brain missed it +This fixes the issueHasan ignore previous patch and use this one +Fix checked in. +Leaving this open because we havent yet done an official build +test cases in make this jira a duplicate +looks great +You may check the recent builds +Reach is a nice statistic to have but I think it can be had more simply than thisIn my experience quality of recommendations depends very strongly on the number of items in the history +usially we trigger an owb tomee deploy immediately after such Bugfixes. +Also includes changes and tests forWaiting for before submitting +is present in the trunk but the files modifiedadded by above patch are not modifiedadded +If a column group is already marked deleted and the corresponding physical directory still exists try to remove the the column group data again +DavidThank you for continuing to participate in the discussion +So far I believe that things are not working on because the gwt project depends on +The current behavior of rrn is incorrect since it would be reparsed as nn +Now that work is well underway with AS all previous community releases are +dependent on wtp in build +I dont know if its a problem to require one of the most recent maven installations +The names and such have changed a bit from when the issue was first added and its two separate handler interfaces and matching filter interfacesFor this pipeline was necessary so that the request encoding could be properly set. +Yes I have yet to implement the command for the other SCM providers +also added divide to Java notation styleAlso added divide method for sparse vector +Does the issue imply there should be more handling addedStill what I found seems strange +Are you sure that usrucbwhoami exists in all versions of Solaris? How come Urko said that Solaris enviroment has no whoami command +The BOM should indeed be skipped and the reader shouldnt return any odd this sounds like a bug to patch would be most welcome +I then moved modulessystemlayersbaseorgjbossasproducteap to modulessystemlayersbaseorgjbossasproductwonkaNaturally since this is not the correct way to create your own layered product the server started with some errors But the key for this jira is to ensure it is properly IDENTIFIED +Committed thanks Jean. +Feel free to resolve as invalid if nobody else can reproduce this. +It happened with IE browser when Firefox browser is used Save button works well +was only in just to set release notes in +I thought I tried it before but I guess I wasnt using it properly +Will look at it if occur thats a typo +I just committed this +Too old +How is that different from the query cache I waved my hands about +we decided to wait for a fresh build and hope itll be solved +And can sometimes change the intended ANDOR semantics right? At least CTQ can be used when a phrase query is not used and the ANDOR issue is maybe not a big deal +Hi Siqi I test this in SUSE and there is no this error +do you mean that its always the first output line with error that doesnt work or the first hyperlink you click on doesnt work +This patch updates use of settings defaultCache to cache default in the documentation and examples +Implemented the change +But we should fix anyway. +If we will find a way how to test it on staging we will definitely do thatThanks again and hope your fix will solve the issueBest regards. +That works! Thanks Sergey +Version Released +Closing since all the are done. +Looks good to me otherwise +Pull request has been accepted. +Please reviewPatch checked in +Compaction would create new sstables at the highest generation number of its source sstables +Patch looks good +This allows jBPM database initialisation to with other projects that also use this sequence +Removed the logo tags and also the package May take up to hours for the fix to reach central repo. +In Linux generally if root creates something only root can modify itHence when you run the installer via sudo java jar +hence my masterCompute did not make sense and i thought it was the infrastructure +Am I following? I would like to help out somewhere if you think you could use it +bq +Ok lets call this resolved +Passes +It is already in the docs that you need to use u smx p smx as args to the client script. +But the Block object can be reused +I have it setup in my guessing It might not be mapped to the specific request in question +Fixed in +I found that when adding a prefix to a child table along an association path an exception was thrown stating that it could not find the associated property +Seems like we havent really reached a consensus that this would be a good thing to do +Ive upgraded BSF in Branch +by applying patch +A patch that solves the issue +Project complete +If the need be we probably can revisit it later +This is the attachment to fix this issue +Thanks Lukasz! + please close this of a bulk update all resolutions changed to done please review history to original resolution type +The work is not finished +in case it wasnt clear I am on this +The current UI doesnt work in IE +reopening so I can modify fix target to +Thanks for testing this out Jie! Ill set up an ubuntu VM tomorrow to work through the warnings +Verified +and +VinodVenu You dont have to run the full test suite +None of the Roller developers use or test with MS SQL Server but there may be some users who have experience getting it to work with RollerAnd if possible consider using one of the more commonly used databases or Derby +we need this setting in other MR unit tests + +And then why dont you use version +Same issue here OS ubuntuactiveMq version can confirm the issueOS Windows AMQ It occurs when I try to browse a queue using a tools such as which indicates that it is a general issue +However these changes are not merged to the local states +patch to the xdoc plugin root modifies dep wrapper and test case and applied patch. +may be neededPlease feel free to propose a patchnew implemantationDIGY +Kinda big though Id prefer smaller incremental changes +generates the as part of +Yes Jacques its working i dont know about this that it will disappear when there is no states of the countryThanks Ankit Jain +You could also avoid FLEXHOME and use a relative path +Fixed at r regtest added. +Thanks Tim it works for me Ill apply a regression testcase for this +They have since moved on to a newer framework +I didnt read your comment carefully enough +Also objects are long lived and placed in tenure generation and as these objects grow in number can put more pressure on Garbage Collector +The bug still exists +EricThe issue of Servlet request methods has been separated out into +Could we get a real description +Fixed thanks. +Looks like the last issue that I was talking about is fixed for hibernate +The last patch started from the wrong root +Why? Im finding it very useful to abuse the column names a little bit to be able to group rows and be able to operate on them in bulk +hit a NPE in hadoop +seems like trunk and yes I agree that if the job fails then the job should return a proper error code +I am sorry about my dumb questions +I dont know if all samples are fixed maybe some samples source codescant be seen at the momentJoerg +I am offended regarding your comment +we are doing a release tomorrow is there a specific test i can use to double check it all works ok +Thanks. +Ok +Using p l also has this patch applied to +I also think that the default should be to overwrite not to skip +above I now see that open will only do what we want on kernels which I was not using when running the above test +I added those unsigned files on +etc +This should be part of plugin in already. +Is this correctI also see two references to in one in the class reference in the license and one in the class describing which class to runIf its a simple replace I can do it before I commit +We now have command objects and Validateable support +Anybody up for writing a patch for this +If the CLEANUP arrives while that is happening the internal state of the NM does not know that the launch is in progress and as such it fails +s and Alexey I +describe column +Ok I took a stab and pulling these classes into Cayenne itself +I think the best way to do this is as a static map of in the class +Thus accessing contentsymlink would actually return you the contenttarget resource which makes perfect sense +Thus targeting a specific Teiid Server version is vital for the development of valid models and their correct deployment +Attached file is for so that we match the and filesdeleted metric count +ThanksI am new to open source how do I get access to propose my own patches? Any guidance to help here Id be interested in helping +issue as it has been released as part of release +We can modify it more later +I actually think in my case though you are right that I can now remove my references to as I had also removed the file reference +I was HOPING that wed moved beyond that by nowYou do NOT have to change your code in any way +There are two things to investigate was there anything in the logs or anything additional in the URL when the login appeared to work but didnt? what browser is in use? Could there be a cookie issue hereThanks +For +Jira is for reporting bugs. + +I think we have to investigate using this in OBR rather than the standard implementationI edited this issue to move it to the OBR component +Just are some small issues +Confusing maybe bug Atlassian if you want it clarifying. +just the isolation for jars causes major issues esp +I also just added a simple test to and also manually verified that the resulting POST is notGreat! I am also going to file a new jira to add some more junit tests because obviously there was lack of coverage in this area +Yes I can take a look at it +has methods to demonstrate that this works +Im just saying in my case it means optional +Instead of changing the binary format could you update code to either read fsimage or whichever is available? Obviously one could start compressing at any point but theres significant value to being able to use existing tools to decompress if anything goes awry +various tests +Please against current CVS +rev A WAR file is generated containing the airavata services +Committed in revision with some minor changes to the unit test +this task is unfortunatelly JDBC does not specify the behavior for this +Without the patch the first AMs history files will be written out by and eventually history server picks up the first AMs info +These attributes add the dynamic that is missing in Doxia and seems on the paper to solve several issues +Thanks DagThe patch looks fine to me +Could you send a more detailed description to the user mailing listThanks in advance +Ill be uploading a patch momentarily which fixes this default behavior of +Run the failed tests locally without and with the patch and they pass +Thanks. +Use the property scheme we use in our Hudson grid as a best practice +As expected the problem was caused by the file pointing to the wrong place in the log +Sigar is so we cant redistribute it +Let me know if there are any questionsThis way the +findbugs +looks like a separate bug +It was me that organised it +See patch for more details +screenshot showing incorrect name when the installtion directory location is provided in preference page +Im currently the owner of domain for exactly this purpose to be able to guarantee the name for the project if we go with +Each compilation run makes a completely new compiler atm + for the patch +I can try running the tests on some other platforms to see if I run into any problems there +After resovled this is the current patch +As soon as we detect a manual modification then no modification is done by the me connector +Say the default is disk +I just committed this +the code so that now faults will be sent according to the version of the SOAP message recd +I approve +patch which is based on changes already committed +I just committed this +it is not likely that it can occur as non for m not going to verify grammar as Im not an Erlang expert. +So when committing this be sure to also fix the branch +Thanks Ivan and Chris. +Hi Christiancould you give me a hint how to combine with launchj? depends wether the exe performs an extraction to a temporary file or not +Hi +Verified in JBT nightly +The service is used to determine whether a token is already acquiredI can see if I can use Jitendras feature to enable security for a unit test if youd like +Presumably linked as deals with removing the last character from generated links +But I dont really know +It turned out to be a Hive ivy configuration issue. +Second Part of Party Brazilian Portuguese Translation +You guess right +I have modified the to take care of ignoring newer KVs +See also bug bug bug +Patch to treat website handle as keyword and not tokenized for the bug report but the patch you sent causes one of our unit tests to fail +If you have a patch that improves things and doesnt regress commit it +Using maven you should be able to issue the commandmvn clean compile warwar and it will generate a war file in the target directory +If its not set you get a which is probably not the best implementation for Camel as it definitely breaks streaming and loads everything into memory and such +Im trying again with fork true in the jaxws for the to see if that fixes it +Fixed we are now using Srping +Stack are allowing the to continue its retriesYes +Unfortunately I cant reopen +the one minor thing you messed up the indentation on a couple of lines +Closing issues for +Im committing this to unblock the rest +fixed in trunk not resolved trunk issues +All http parameter names now start with p and are much shorter +Bulk close after release of +I dont have an issue if you come to the conclusion that this scenario violates the spec +With this added cache we will go to the file system once per tableopenRegion +in Brew +Will let it stew some in case others want to take a looksee +This process is called binding to LDAP server and you need to provide valid user and password that will be used to connect to server and query it +I will upload the patch by end of day today +Ah ha yeah cool +Hi Yonik These are indeed useful features +Moving to +based on earlier comment on legacy script I added the legacy legacy netuinames and get javaScript are still not being output for a named anchor +I suppose someone need to do the upgrade +Please check how it works now +Hi Daisycan you think of an easy way to test your feature? It comes without regression test and history told us that everything might break Thanks in advance types of code changes functional changes and changesFunctional changes will change the results of functions +DO NOT use quotes when set an variable +Of course that occurred to me immediately after I submitted the comment +The attached page demonstrates two different setups that both lead to the stack trace +Please review and commit +I would suggest catching any exception in your authenticator method and wrapping it in an so you receive the exception +Yes it does +I am also on Mac and can reproduce the issue with Adobe Reader X +defer to +We probably want to stub mlock for too but I feel like we should leave at least one test calling mlock +I cant reproduce this problem in the beta releasehowever I did find a bug that happens when theres two CRs andone LF so Ill go ahead and try to fix that +Can you please create a review request +I am surprised that is at least times faster than in writing +Im not sure exactly the full impact of that parameter so even if the default behavior is switched to true I think the setter should remain in order to let deployers override if need be +No further changes in version of the spec. +smoketesting found a bug in my patch +that the other contribs continue to compile and test +verified +moved find from to SM changed it to private since it is unlikely to be used out of SM +Arun There are two s from you onIt also clarifies about the port rolling and etc +The idea is to lock the version manager early in the prepare phase and release the lock at the end of the transaction +This JIRA is a request for the support for reading invalidbroken elements and writing them back to XML files +Looks like Joydeep got busy +The requirement for the PLUGIN is still though but forking with now works again so the documentation update is correct. +This sounds more like a functionnality to me +thanks for the comments +Please review it and let me know your comments +javaxxmlparsers like describes the same Issue. +Marking closed. +Unassigning myself from this issue +Our of sheer convenience it would have saved me writing a wrapper class the contain them but it was no more than that +I think there is still value to a simple fast implementation of small character strings vs +This step is success using source and but fails using source +Reassigning to Arnaud for investigation +This wrapper is intended to mimic the servlet API and as Ive explained in other issues there is no concept of an anonymously authenticated user in standard servlet security +I believe the documentation is wrong here +Changed references to package to and to +Ill think a little further about such methods further up the inheritance hierarchy +Do you mind sending me your samples of that before I go the a forum discussiongroup? It may help me to track the problem +Its not supported by all shells you please file a bug with a specific thing that wont work rather than the solution? Likewise for the +I dont know of anything in SAX DOM etc +Example PDF to demonstrate this effect +Comments on RB +Patches are welcome +Assuming a combination of autoStartupfalse and a valid application context will make this issue invalid. +Resolved in. +It should not failI think we can call the module migrated +Updated to the version on the resolved issues +Also w some minor cleanup Made the address variable naming consistent across and Improve InfoDescriptor header and primary fields into the other two failures are unrelated filed and +Im running tests against this patch now +This one was closed to early the uima code is still missing +Thanks Dean +The idea is that you dont need to add your own web markup container for this +Ive so far been unable to get it to crash on any of my test devices +What we are doing now is bad and we should NOT be encouraging people to do it like that +Can you give me a clue? is it a problem with the isNull function +verified at subversion revision +No we need a WSDL to implement a regression test. +But until then I can see no reason to copy these methods +if an app s and refCount becomes but doesnt trip the assertFollowing assert we call ensureOpen so I dont think refCount will ever be less than +on different filesystem than destination store moving to this filesystemDhruba can you please provide more details +closed as per christophers request. +Add readme +I raised a lot of questions +When a referenceable nodesis added the comparison now triggers diff calls for each of the index entries eventhough only one of them was added +that was intentional since I want to only be explicitly worried about whether I have been interrupted because I need to stop cleanly +Maybe theres a better way to make this happen perhaps a fragment bundle or moving the yoko class +wrong +TedI has amend the patch again Please check +ShiftTab works correctly but Tab doesnt work at all +I fn dont care on which end +Almost there +So I am not sure if it is necessaryIts a super class for all expressions that represent a single value projection constants and eventually tuple and map dereferences +Please see for all progress related to this issue +As a behavioural change we cannot implement this until despite the simplicity of the change +Would be great if we could get a mechanism for this of some sorts in +My email address is changed to lfrohman +New patch that fixed all but findbugs warnings +So it can be removed without a problemBut has some dynamic resizing abilities optimized for performance and memory reduction +BTW I would close this issue but JIRA doesnt give me the option Let me seek an admin. +Attempt to retry dependencies that could not be obtained originally when they do show up +Because this is my first submission to JIRA I am not exactly sure how to create one +We need to find a work around to distinguish between real and ivy related build failures +More thought called for +Now expand always invalidates the xml string +But has another problem so Ill file new issue. +This means we report less information in backtraces but it also means that calls do no checking at all which is part of the reason performs so well for those callsMaking the fileline available to for the backtrace would mean wed have to move the down into the target method and perform the checks much more frequently +If there are no objections Id like to backport this to +r pulling up modified contribution to trunk +There are quite a few extensions already that died because the owner lost interest +Change to version +I just committed this +Deferring to sequence diagrams to illustrate the issue +Just to be sure are you running off trunk? +this sounds like being relevant when we add support for defining the values of expressions +Fixed in r for Java backendThe problem still remains in C backend +Is it really necessary +Work around is included in the fix committed for. +From a java class that has associated metadata through the wsdljava process i get the following +Im not asking to remove the link Evgeny just to simplify the text +bq +Id recommend building it from source instead +I think that does not work in and trunk +The Modules menu is handled by the Site PluginCan you attach a small project that shows this issue in action +What does the fb tag denote +Coming back to this issue after a few months its clear that the approach I was taking with and was much too aggressive so Ive marked them invalid +There are no errors reported in the submit issue dialog +Can be punted for next release +Forgot to update the test +It basically adds a few methods in RM that were present in Row eliminates createRow and replaces all refs to Rows with +Dan I dont get any compilation failure with I still cannot run my tests on XP +Great job. +Same problem in JBDS GA build removal phenomenon is fixed +Thanks DarynAlso thanks Tanping for reviewing it. +The patch was just to demo the problem in the description +Trivial logging here is useless because when the is thrown itll cause the server to fail to start printing the exact same thing +Right I forgot about version complianceGenerating dynamic assembly in previous version of +I was looking at this and I think it would be more efficient and simpler +Cool I will work on a patch to add Set support +Will be commit in two days if no objection +vinod thanks for trying to defeat findbugs +Thanks for doing this Vinod +Argh of course +Im leaving this issue open for with likely some more related changes coming soon +I attach this modified Example to show that only the workaround does work but the actual case of the public bean member in the wrapper class still does not work. +Will go over them and submit a modified patch soon +Ive added a Maven profile to enable quick testing under H as well as fixing up the code so that all tests now pass. +So the replay logic does not need to change +for a normal db this would be fixed for the entire resultset but its not much of a stretch to make it +I have not applied your potential fix Chris yet +Yea but that hflush improvement is still reasonably experimental I think +Patrick. +Have you tried ensuring there is an import for Number in the package import configurationCan you also please provide a repository export illustrating the problem rule? This will help debug if applicable +Committed to branches Sending archetypesSending archetypessrcmainresourcesSending archetypesSending archetypessrcmainresourcesSending archetypesSending archetypessrcmainresourcesSending archetypesSending archetypessrcmainresourcesTransmitting file dataCommitted revision Committed to trunk Sending archetypesSending archetypessrcmainresourcesSending archetypesSending archetypessrcmainresourcesSending archetypesSending archetypessrcmainresourcesSending archetypesSending archetypessrcmainresourcesTransmitting file dataCommitted revision +Still need to modify the CHANGES filesIll commit these tomorrow sometime if nobody objects +Ive tried the configuration in that ticket and it does work +If you look in there are a bunch of things in there which we can only do in the Java DSL +I dont see any code on myfaces that could cause the mentioned effect +Close and move on +Ok this is looking good I will check it in +So we still using id in bloom filter +Not sure transfering blob over FTP and then a claim check message over JMS is such a good idea To much can go wrong etc eg if FTP fails and whatnot. +The last status was the test case test failed +Date value does not follow the format ? yyyy mm dd It expects digits for the year +We should bring the requirement of the usage and need of that to HDFS +Added patch Fix for the issueTurns off TI events while invoking jlThread constructor +I am having some problems applying it +js files involving var constOpen up separate issues for these if you feel they must be made and let me know if lurking in all these +I have put a copy of the logo back at that URI +It doesnt matter if you use streaming or not +ICUJ uses letter in pattern for rounding support +Backed out of and trunkWill attach patch v soon +Please ignore the last paragraph +If you are able to reproduce the problem with feel free to submit a patch or a test case +If you are not too much annoyed by the limitations of you might try it as a persistence provider it does not have this bugOtherwise you can add a persistent attribute to your classes that takes the classname upon construction +If possible please build trunk and verify +And all the tests passed on my machine +Should be fixed as of rev test for this is running and passing. +Im against this +committed including. +Shall we close this as now invalid? Or rather we need to reproduce it on top of Jimmys new refactor +expert users trying to build things on topI think its this mix of use cases that makes this API such a bikeshedBut i totally think something like this patch is the right way to go so that users canstore things! Perhaps the issue title should be renamed i could care less if is there or not thats the expert API? I think it should take a backseat to the API +Committed the fixes to Rev +Issue indicates this should work but at least in these cases listed it does not +The fact that the page renders fine in every other webkit browser I have tested leads me to believe it is something with +See for further details. +It related to but i dont know where i will find and remove this package? are there in Sonar sever Maven setting or Sun JDK folders +The attributes are corrected due to Func Class attr +This has been fixed at r +Apologies I shouldve mentioned +is updated is updated is removed and we only maintainRegarding backwards compatibility +I dont see much of a problem with caching negatives but Im not sure we need to anything here at all +moved sitecontentriver to sitecontent +or even better just apply it and commit but dont upload it yet +This also means that chaning all static methods to normal if they have been using logging heavily +I will check in the code tonight +Would that be possible +This is only needed if after the async submit you need to redirect toa different page. +thanks +And guard for short column identifiers following longer prefix delete markers +HiIve that too but it was the only font I could find for free +And it is no wonder that it is not supported since decodes to space as part of a URL. +However there is still some pretty huge contention in +The reason for this is contained in +just read the long thread on +You should now have commit privileges + +Caching option is more powerful if we have an option for enablingdisabling the reload +A comment in docs sounds sufficiant to me too +I like the separate output files as well so I would vote for downgrading +I will then post that contentThanksIt is Firefox version +No sweat +However I added the FILENAME to limit the interactions between the three runs of each test +Hi MarekI think it will be better to just move node from stormgrindprojectsboxgrinder to the root +the points above on the doxia site. +Closing it. +In some cases that is splitting them into multiple tokens but in others it is doing things like to and remove duplicates +thanks a lot +Jiri Tom will update this with details later today but Im resolving it for now so it doesnt appear on Maxs radar +I updated the wiki +It looks like that fixes the issue on a single thread where it ends up reading from the same repeatedly which is why I overrode the close method to reset theIt does not look like it will fix the issue +I didnt have any compile errors after patching against the current source isit line after patchingThe exception is it idI had a look and a try and it seems its no use for inid id toJava and ll submit a new patch +DavidCan you please check on this onethanks GeeMy reading of the submission spec has always been a bit different to yours but nevertheless I think you have found a problemWhat the handlers are implemented to do is to equate the absence of a for to have the same effect as a of the None URI for +If there is any inherant problem with passwords of length then this patch is no good otherwise it has potential +Mind that the web console does not auto discover all the Camel applications in the containerIts currently tied to the which is embedded inside itHaving it being able to list and work on all the Camel applications in the container would be a nice new feature +Attaching release note. +all regions will acknowledge the move message by publishing a move ack message to T +Could you provide a formal patch against +In production people might actually use mySQL or something else instead of Derby +And is executed twices because of log info messagesIf the upload if executed twice and should be thrown and it is working fine +Committed to trunk in revision backported to in revision +Yes you can determine that there was a problem regarding users there but the problem in this case was that a did not exist +So we should wait before reviewing +Yes the component vendors should fix their TLDs but we should at least provide a workaround for end users if its practical +Reverted to using mvn eclipseeclipse instead +this would also be served by the build pushing getting cleaned up a +BillyThis was really useful input +create template from snapshot issue should be fixed I cant reproduce it on is another bug to track this issue +I think as positions become first class citizens and as this stuff gets faster were going to see more and more use of positional information in apps so it will likely become more common +Thanks a lot. +For the future please use attachments for such large issue will be fixed in the next release +The problem here is indeed that the page attribute applies to the table that is being used in the current list view but is also used in the menu +Thanks for the patch Andrei. +Alejandro If the is to be removed it should be possible to extend the frameworks by an external implementation +Simple that compiles methods that arent compiled using the baseline compiler committed in r +This patch corrects all dependencies between projects and third party either apply it or resolve it thanks +rev . +Its Ryans patch +Retargetting from to +Two things in this test case one is the code that doesnt work +Initial draft of patch adding patch with some is for so the results above are NA +Why does vnode change the rational of why this feature should be allowedThe rationale is in the description +Patch with with regression patches in r +Elected to make a patch for that makes that work on +Id appreciate a couple folks double checking it doesnt break things and actually works. +PermaineA different fix is required here +verified to build on alpha +This is fixed in trunk with. +I will fix the escape problem once is resolved +Manually tested +Seem to be fixed by +Sergey thank you very much for your work and spending your time to test it +Created branch and imported Xis m attaching a patch that makes the build of Jikes RVM with work on Mac OS XThe build process successfully creates the RVM boot imageIt was tested both on Mac and LinuxI renamed all occurrences of Openjdk to for consistency with the standard naming +See the above testcase on how this is resolved issues for already released versions +distributed log splitting has anotherP On quick glance this resembles the split code +I would apply a patch in this area +So Hibernate would be the Cachestore if possible and wed use constructs to query and persist objectsThis would then enable a distributed data grid using JPA objectsGood stuff. +New patch addresses the review comments +The behaviour I am seeing with git head is that with transactional true no events are received by the listener but with transactional false events are receivedThis seems to be because with transactional true there is a proxy and Spring doesnt recognise the object as an with the proxy in front of itIm not seeing the duplicate event firing Serge describes and it would be good if you could attach an example that reproduces that particular transactional false it is working as expected +So did the versions ofAny solution whould be most the additional information Ive been able to testit with both and with my own patch +Does that mean itll be in or? Thanks +JCC was only needed when using on the client side in clientserver compatibility testing since didnt have its own client driver +Definitely theres a couple files specific to Hibernate that are probably still being used +I was not able to actually start the server to make sure that everything was workingNow I can and there is a problem with at least one of the javascript source filesI will review the patch and resubmit + Bug has been marked as a duplicate of this bug +As such a template will be passed through defining the custom CSS as well as skipping the step of resolving file referenceJoseph contrast icons implemented for Viewer JNDI Viewer Dependency Viewer Patch posted pending that i am still verifying LDAP Viewer for high contrast icons once i get a local DS setup +If they wanted default behaviours with no dialogs they would use server publishing +The reasons are +on the other hand uses a more flexible factory approach +I am sorry should not be listed +I have attached a patch to that adds the appropriate ejb refs to all entity beans both session and entity in the file +Its more and more obvious we need to provide a way for users to override these timeouts thus when they use a slow system there is a way for them to at least get work suggest we use the same timeout as rhc uses and provide at least a way to set these via system properties and eventually allow for preference control over this in eclipse workspace +Also repeats reliably across AWS instances +Rodrigo any constructor that converts an empty path to is a change in behavior that were not going to introduce +Is it even worth creating the list lazily? Its going to be used for every commit so every connection will have to create one +I think you can probably close this issue now +Compactor is in tests inI changed policy logic as described in the docThanks for feedback on r! Ill set some tests in motion and get to it soon +means building the projects war not the plugins war +This is done on purpose so that web things actually work +some tests seems to fail all the time locally so its unlikely to be good + it be possible to have nice error reporting that displays the possible parents for the macro +This is not a public API +Thanks again for your help reviewregardsrecently I started learning Lucene API and I along the way created a few snippets showing different Lucene features +This would effectively be a cordova version of the TODO app +Run with and notice the extra row +Weve resorted to using our own Maven build where weve literally applied the patch that this issue started with +for the changes +If you dont install or remove it then the Tomcat works shutdown properlyPlease note that my old build from works fine +Seems like should set based on the value of JAVAHOME and then make thisavailable to the launch configurations by defining JAVAHOME in the targetAlso the directions in for importing the project into Eclipse areincorrect and need to be updated +im working through this point by pointusrsharejavausrsharejavausrsharejavausrsharejavausrsharejavausrsharejavaThese are not redundant +I cant find the offending snippet it in the current anywhere +Please close +JulienI took into account your comments +Cant we just commit it as is because it will cost me good time of injecting them off if not where can I post a patch with those refactorings +Changing issue type from Bug to Wish +Let me know how it goes + +But indeed if we get rid of the Builder then forget about that +This direction came out of discussions about the best way to persist master state over inIm generally very in favor of moving the master away from the basescannerheartbeat way of doing things towards zk and rpc +bq +Thanks guys for the patchesI modified the logic to use doStartdoStop +Sorry about that +Verified by Tony. +can be calculated independently by any processThis may not be trivial +Ill go through and check what else is missing +Hi Ben I cannot access to bamboo today does the workaround has been implemented +To explain a little bit theres an deployment for and another for +I will try to create a unit test but wanted to get the patch up since this was a critical bug +Javadoc warnings arent this patch +This patch also integratesthe epolldata into the to get rid of an unnecessary mallocon the critical path and as a first step in encapsulating the various polling mechanisms +And regarding print currently there is no way to print a stack trace to the log +Did this work before? Looking at the code I doubt it +If thats the case then lets just say that we dont support a of runtimes +This issue should be resolved once we have implemented single for +Workaround for focusing on components inside components +Committed test case to trunk +It looks like the eclipseeclipse issue is spurious +This allowed me to first replicate the issue referenced in this JIRA and then I did the prescribed uncomment in the that resulted in the fix +This strategy would also allow us to delete from the zip package +Im reopening this issue because it wasnt really fixed +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the codebase +I dont really understand this but think the new lookAt code should go far in helping this problem. +I agree that storing DOM trees is a bad idea +Without this fix cant comply with basic Java standards +Passing to Rich +In addition to warning of two page definitions in Seam should warn if there is one definition in and another in file +Fixed in r +As for the jetspeed prefix changing the name made everything work as expected including the listener +Thanks that makes sense +Perhaps the nightly build is missing that module +And at that time we can add capability to process commands like dfs count tmp +Has there been any movement or further discussion of this elsewhere? Ive looked through the mailing list and cant find any further references and as far as I can tell no one has found a solution that works +bq +The slave will actually do a reload if conf files are being replicated so its sure to open a new searcher currently +Add to that the number of changes and I think were looking at weeks of an unstable trunk while the kinks are worked out which will hamper other featuresI realize that this issue is just a part of but I think this issue should be broken into several pieces as well decide on the API we want to end up with meld it in or have it replace what we have alter readwrite alter compaction process +Its embarrassing this was not committed earlier +bulk close of all resolved issues prior to release. +if no one cares about these id tags then when i get back from vacation i will gladly volunteer to submit a patch to remove themA quick workaround is to change them back to Id in your local then apply the patch +Yep works fine with rev +and +We will be resolving this one as incomplete if no testcase is provided +Will think about it +Ill not put on rb because mostly just removing stuffDepends on being committed first +Is there a writeup of the supported syntax for +includeLocation not specifieda sync default true meaning lazy initialization like current trunkb async default false emphasize performance like and logback +Should be able to get it done in the next couple days +Frankly I dont remember +Diff file to add test cases which demonstrate this failure +Yes agree +Updated to use pending a release of +I like the intent behind this change but Im concerned that it may be a lot for the Thrift guys to digest at onceIm running conservative because it seems that moving to ASIO for the project Im working on may be too much to take on just right nowTo constrain the scope of the changes it might be an idea to limit the changes to only the and any changes needed inside the concurrency part of the library +How is this reproduceable +To activate Springs own expression caching the cache in needs to be set to true +Does that happen still +One question you added a comment implying that JCC is still supported with Derby +I dont know exactly how it works but from the javadoc it looks like this is trying to find beans in Springs application context +verified +Fix most of the line issues +The object is an interceptor so I think its good to go ahead and leave all the flexibility possible in thereThe reason I proposed making the log levels configurable is because the runs at INFO level +I like the patch IMHO it would worth you checkin it in order to let users check if it fixes the Stream ended unexpectedly issuesTIA +Im scheduling this for M since in its current form the doesnt add much functionality. +Ill check this in and I will be sure to update the revision before checking it into +Ill check it in +update the jca samples to include +It is because Forrest does not have that has the class orgapachefopmessagingYou may check them out for example in my Forrestjongwooklocalhostapachelibcore ls fopIn my Forrestjongwooklocalhostapachelibcore ls fop +Id suggest a test which starts Tomcat using HTTPS and then use the class to ping a URL +Ive just committed this +Isnt this a problem or am I missing something? +I didnt spend time investigating why +i hope not too late for a release +Patches for and trunk +The poms seem to be deployed +I will talk to the Tomcat team to resolve this +Wow I cant believe is deferred +For this casewe had to invoke mvn releaseprepare DusernamesvnusernameIs this the same case? +Is there any chance that this issue will get fixed? If there is a way to set the configuration in maven so that the port x will get set Id like to know +Old file lost replaced from Cocoondev archives +Reopen if the issue is still there +As mentioned on the comments I havent actually tested it on builds +Assigning Jeff cos he did work. +Xircles is great +Paddy a new fix committed could you test it againCommitted revision Both and should be rebuild +This issue is being reopened as the changes were backed out due to Sanjivas +How did you structure it +Actually no +Attaching +So the problem seems to be that the didnt work properly in the first deployment as the second deployment gets this dirty instance +the same code on the same machine booted to linux however works fine +Committed at subversion revision . +As dreiss has pointed out this is my patch +please bear with methanks updated patch sources should be up to date now +I did upload that driver first to deployment as a and want deploy that later into a profile +Please refer to the latest patch on +What kind of advices do you need please? Sorry I have not a lot of time to put in this. +Von meinem iPhone gesendet +Is a sequence standardized anywayBut indeed thats my opinion +Sorry I think this is not a bug as the exception is not handled in the script method +Im adding additional unit tests and more info logging about the major events +Weird it cant download junit from central +Looking at it now I think most if not all of the readers follow. +There was an application context in srctestresources that was overriding the main one +I evaluated DRLVM crash on Launcher test and found that two main bug sources are in zip support native codeClasslib native code tries to allocate memory for zip entry +Its really easy to mess with JBI exchanges states and suchYou may want to open another issue for that but I would suspect either the or your own code +I am making no other claims +Suresh yes I very much agree that this should be merged to +Good point +It doesnt sound like its worth it to me +Net will interoperate +Set the property in the file in Rev. +can you rebaseAlso Yongqiang do you have any more comments on this one +I think Pig itself still doesnt have a good way to display execution errors +After the patch if a next request is always timeout because too large we will retry forever now is it soWithout the patch we do the retry in but now we do the retry in next which is no retry number Not really +I have fixed most of the obvious indentation issues and Russel has committed them to the anyone finds any more just post a JIRA assigned to the Emacs Mode component stating the version of emacs being used +What list do you want it onBamboo is not managed by Xircles I have to configure everything manually +Now the filter rules are as follows You can specify a whole module or class using the moduleclass name You can specify a specific method on a given module using Modulemethod or Classmethod You can specify a specific method name globally by just using the method nameI think this should be sufficient. +Now I have permission to resolve and close this issue. +For the counters the RPC is optimized +Is there any update on this? I am getting the same problem with +Thanks Uma +I could not reproduce this on trunk +I remember being slightly anoyed by this at one point and thinking it would be really easy to refactor the code into two seperate servlets bound to the seperate paths but then i realized the init method for hte servlet is what makes the core and I wasnt motivated enough to make a choice of how to deal with thatThe simplest approach would probably be to have a bound to select that is in charge of initializing the core and an bound to update +In my first comment I pasted the doc for if you would like to use absolute path it should be something like or fileIf no other issue I will close this jira +Didnt we concluded that requires that column timestamp is a real timestampEven if not every time we find a workaround for something like that I feel like we dodge a bullet +Assigning it to Murali +Sorry yes it is happening on. +Deferring to +The patch is rather dirty and was tested only on Windowsia +Was committed by Ted a day or so ago +Checking that the exit code was is absolutely essential or well mask real errors +Attachment has been added with description has been added with description problem appears to be in the Oracle XSLT processor +I can take this up if you want +Verified in Tools +Alex can you please take a quick look at this issue and see if there is anything minorlow risk that can be done +Changed the description that it is a jvm specific issue +You are right the place where the messages are registered with the monitor is simply wrong +Chris can you update to trunk and commit? +This is why I suggested to use W functions in windows and use the UTF chars for that reasonThe exception status stuff is also just an optimization in the error case? So why are you aggressive +Ive the same problem with jboss and jgroups +All flex +Hi AndreasCould you post a simple demonstrating this problem? deploying before starting is not possible but implicitely happens with container start I think this issue is invalid +Attaching screenshot +Is this setting on the client side +The so called LD explorer is just an html serialization of LD resources +As Greg mentioned tests will fail because it expects to have a default collectorFactory defined +Thanks Sanjay for the review +Please verify and close +Dont we have one or at least the info is there in the docs +A form is now added to the start event +Applied the patch and verified the patch does not change the behavior on so Ive committed the Hudson continuous integration builds run on Linux so they should verify that the patch also works on Linux +Crash is the exact Ubuntu version you are using +Production setups may not want that +Pruning is accomplished by setting a dependency to provided scope +This is simpler but would require modification if we started doing dynamic across collectors +Patch to in sandbox +Still reproducible in is the planned release for this fixAnd meanwhile is there any workaroundWe are facing a CPU usage due to this issue in Production +So in Maven world this is the ultimate source of version information +change in release +unfortunately work with Maven Beta ? Is the issue filed for fixing? +this doesnt change the plugin +We could easily end up with stuck jobs that never finish and never fail +Any update on this +bulk close of resolved issues. +We set sleep state just before going into a blocking call like the selector logic for acceptIt might be possible to fix this with a smarter locking structure or by checking on the server socket IO if another thread is already waiting on it +Ill start putting it in +Attaching updated patch and the two changed html pages and +length actual bytes The question is not how to encode binary values in Avro but rather how to encode default values for binary fields in schemas which has no support for binary values but only stringsIt is possible to encode arbitrary binary values in by encoding each byte as a code point +Hudson just committed this +We could probably use a debug statement showing this scope being managed to runtime but the patch works +I reopne the issue as the two suggested filters dont work eitherprobably a bug in the way we parse the filter +are put in right place when they get compacted so maybe thats why you only see when bootstrapping +jar +Marking this patch as ready to go address the other issue of the seen during the run +Can anyone have a look into this issuepatch +Fixed in r +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +rtested on ubuntu smoke and ok +I will upload a modified example war will make this as clear as possible +Thanks +fixed in rev! +did you detect where the wrong setting ispatches are welcome if you are able to fix that issue and provide a patch I am available to review and check it in many thanks in advance +Can you reupload the patch +The symptoms are different but see also for another problem with the order of and in the classpath +Any plans to patch and release this? Sounds like a fix has been available for months +Srikanth could you give this patch a try and let us know? Thanks +Nothing to test here +What should the adaptor send +Simple change to fix failure +I was leaning that way over exceptions +Should be reported at the line the FIXME tested! +I had hoped the new regression would be a quickeasy fix to fold into this jira but. +The code still needs a bt of refactroing but highlights teh approach taken to overcome any compile time dependencies +The NN has an option for an external logger daemonSo this implies some sort of IPC either to another local process or possibly even a remote one +Mail class used by the issue I presented +The failed QA Comment was from a mistaken patch that was uploaded to this JIRA +What exception should we throw when the output format is missing in the configuration +I am testing on TRUNK which is currently and the issue was reported with so it might have been fixed between now and thenAttached patch is the test case I wrote +hello Ben thanks for reporting this +Im using the version of hibernate which contains no package + +we tried with various configurations and startup options with no luck +bq +Steve have you got time to take a look at this +Whats the ticket for fixing the dependency +as svn and resolving it. +Ive ran into a few problems between Task and Link so Im curious as to how this affects those +Nope that +The fix attached +If you are not interested in bug reports without patches please let me knowThats what bug reports are for +is probably the only implemetation used by Accumulo a tablet fails to update theMETADATA table with the bulk entry due to RPC timeouts the update might execute outside of the locktransaction counts +This is just for h but I suppose is the same for for reporting +Sveta English need to be verified +added ability to nest dynamic tags in iterate +Trunk and branch +Changing fix version to +No longer relevant +Ah theres an easy fix +files Hibernate scans the +Corrected patch to cover corner cases +I thought the global variable cacheconfigthreadsperdisk may have no use if we seems I made a mistake +its unfortunate that such a new API in x is already incompatible with +Are you still planning to do it? looks like Ive got it see the example at the end of my last comment +However theres only one unload thread in the tserverIm thinking we could put a decommission yourself in the heartbeat message +Fingers crossed +I have applied it to trunk +Yet another version of the patch with Apache headers added for the two files that were missing them +User may be confused about that +Sure Ill do that +We need to close these issues for releasethanks is different with +HiI confirm that it worksBut the custom appenders start working just after all the server has started loosing all the startup logs entriesIn jboss they start immediatly and I think that its correctIt would be also good to place the in docschemas +newInstance to add a locale parameter as part of Ill check that in in a few days +Closing it. +Fixed +Thanks to Hasan for providing the patch +thanks +Username domic +So it is only a problem with IBM JVM +Thanks for your patch it looks good! contains your changes new tests does not modify of its model character between hour and minute text fields removed when using aThe tests validate ampm conversion and does not modify of its model +It should not have any effects on the issue but it has + ended up issuing kill to the master +This should been fixed on the latest Beta +The actual commit includes files which Forrest generates from the source +grails directory just in case its anything in have any ideas on any other way around this issue +This looks ok now can you confirm +While there are ways to force it to use the index these require adding a dummy clause to the WHERE which mentions the ORDER BY column +If it is addressed in a later version then will have to wait for an Ant upgrade post keep in mind for any jar code in m. +I wonder if it is an issue? I am on Windows +We do not use the resolved status. +The task description should answer the previous and group modules are whole module has been renamed to role instead of group +So far I have introduced an interface refactored the old cacheclasses to implement this cache added a way to add cacheimplementations etc +You will see that all japanese characters are converted to metasymbols +Can anyone test this on generic Mono +The attached patch resolves the issue for me +I think this might be a big win for large field valuesThis could only be done if the highlighter doesnt need to perform any text manipulation itself so its not really appropriate for Highlighter as Robert said but in the case of FVH it might be possible to implement +applied thanks much +all three patches have been merged the last one was svn revision . +Im not discussing the internal logic of the algorithm but whether it has its place in CM because it makes an assumption that it is OK to catch exceptionsHowever if you thinks that your suggestion is safe enough Ill implement it +Need to check to see these are working with Sigar r fedora core This must have been fixed at some point Im able to autodiscover both my network adapters and file systems +Thanks WesDefinitely better ideas above +Looks good +Could you try again with trunk? We had problems with general rake packaging that has been fixed on trunk +Will commit this afternoon unless objection +I wonder how since I applied the whole patch with Subclipse +Could we somehow hybrid some of our existing UDFS to work from both engines? We will surely have to support the hybrid approach as you are suggesting for that users have implemented even though we will recommend users to their in vectorized fashion +As long as you include in the interface all of the functionality needed if you want to add another kind of logger you just need to update the factory and write a small adapter class +At one point I was creating a application that launched a lot of different types of programs +I ran into this while working on and on Mac OS X +Logged In YES useridSee my comment for RFE Username pilhuhn +Fix version No Release means not sure when but not soon +EdisonCan you please provide me the reproduction steps on a fresh environmentThanksCan you please provide the reproduction steps on a fresh environment and mark it as resolved againThanks possible steps to reproduce +Fixed in with r. +We started including rdoc recently. +Performance wise most people w dimension of or will see little if any difference between this and specifically calling out a latlon field type +When I met with real EDI files my first simple algorithm failI rewrote split method and add unescape methodAlso Ive patched to use unescape for fields and componentsAt least it works with my UNEdifact messages +If maven final release is to be stable with a feature set comparable to Maven as we can read in announcements dont you think those functionnalities should really be part of the release? My development teams got used to dashboards of m and I guess they would be quite upset not to have it the final release of perfectly understand that you have lots of things to do for final m though +Im still having some random duplicates too +no clear info available but definitely blocker for release if this is not the case +I will use them and see how it goes +I got the attached result when using the following commandlinejava cp + +with JBT. +I fixed this issue and sent a pull request +Still invalid +The issue about ampersands has already been adressed +This is for? +Its nice to have the ability to write it in any of these three methodologies but Im not sure any of them are better than the other +If you can answer this question then we can fix the issue quite fast +This could be something that may have caused this problem +the good news is itll be easy to proposition for patch use existing code implementing retries existing hardcode limit as NUMRETRIES +Also feel free to add an assert its one of the oldest code I wrote when I started the type checker For test your suggestion is ok for me +I think the jobtracker should save the old jobs to DFS and read them back when they are needed +Great! Sorry about the wait. +The unit test I wrote actually has a test case that shows what package sealing does +in case that is what led to this +How exactly does that call at line result in a call to? Are you by any chance using or the like underneath with your linked to itJuergen +So the Idea is to remove that category from using xsl transformation to avoid selecting Xulrunner features by pressing Select all button on categorized view +there seems to a race condition where karafs management bundle seems not able to detect the keystores created by the new proposed system bundle +Perhaps the patch is as simple as this? Mirrors behaviour of addMessage +This should resolve a host of unexplained issues by providing problem source this a dup of? +As you can see with the current JIRA I have scheduled a fix for this issue to be included in the release which will be released around Q this year +Attachment has been added with description AFP Renderer Block orientation bug resolved +It is interesting that the same authorization id can have different credentials depending on the connected databaseI dont know what the here are but for the rest of this discussion Im going to use the following nomenclaturesystem This is a user name that is authenticated with databaseName nulldatabase This is a user name that is authenticated with a databaseNameIt is interesting that we authenticate the user twice when creating a database +if you want to order in memory on how is it supposed to work? In other words there is no natural ordering sequence for an arbitrary entity +oAshish Vijaywargiya +No worries Jacques +Could you please also provide a complete wire log of the session? Feel free to obfuscate security sensitive information prior to attaching it to this report +I havent tried applying this to but it shouldnt be much work to adapt it +On Chrome I always get unexpected server error message when Im trying to save data into the db + +OliverThank you all is correct +Subsequent investigation was +An expiration handler on the messaging service queue and throttled local hint writing should work very well +Following are the added samples Sample Introduction to Synapse Callout Block function template Sample Introduction to Synapse Splitter and Aggregator eip function templates Sample Introduction to Synapse combined function template Sample Introduction to Synapse eip function template Sample Introduction to Synapse Wire Tap eip function template Sample Introduction to Synapse Content Based Router eip function template Sample Introduction to Synapse Dynamic Router eip function template Sample Introduction to Synapse Recipient List eip function templateThanksMalith +We have already pointed you to the correct place if you want to discuss this further +Id call no end time a serious ERROR +Fixed the precommit builds and resubmitted all of the patches that had failed due to the problem +test project using looks like I should have refreshed this page before I posted my comment +Lets make critical so it gets in +In their implementation they might use that path for some purpose +It attempts to iterativley increment a sum but does not get the desired with same code when run outside the container using a works as expected +I have the same concern here as was expressed on another JIRA recently regards htrace +Verified against a build post +to be clear at some point we were just adding features into one big root feature making updatesinstalls which if there are too limiting constraints could be problematic as I recall it +The d indicates that the rolling log file should be in the directory pointed to by system property +Added description for ajaxKeysajaxKeys timeout onbeforedomupdate +Greg Im gonna look at this again and try to prepare a first candidate for committing toWhat we have is already quite good I thinkDo you have time to do the planned last changes? Do you need any help +Updated patch to latest +Committed. +Buts its breaking many tests +if you can come up with a that still reproduces the NPE please attach it +no I lied it seems JDBC datastore does not supports native LIKE filters +This could be an issue if we are copying files written in earlier versions +faileddeploy is present that means the operation was rolled back +If you do have any follow up issues for this we can do them on a separate JIRA +a new JIRA will be opened +I assume its still in the configs is it also mentioned in the documentation +Patch to update Thrift lib from to latest release sense +We did not see the regression issue that you mentioned +This will also give us some time to release this fix in a community version as a Known Issue in the EAP Release Notes +I just committed this +This happens for commons logging refer Bug +Responses from Tom with regards to schema review +that is right it depends on server but i use all default settings and get the error as i think a lot of users get the same problem at least they should get a message about it may be this is not the bug this is may be a feature request but reject this is not a solution +Mithun if you are readingwriting a large file you are still using the FS instance through the corresponding INOUT stream youll have to wrap those as well to check for FS activity +Its been a while since I looked at this issue so Im not sure from the description what the problem was +Emmanuel do you think this feature is worth adding Hibernate Validator specific versions of these constraints for now? I tend to say no +The patch is ready for review +fix this only is the new one +Restarted HDFS with the namenode port +OK thanks Zhihong I was confusing which JIRA this was +running tests +add code be fixed now +JLR do you have any objections +Doing such a think in a middle of the build is harmful +I will cut the branch from master on that day +Ill commit soon +At our company we had the same problem with merging of multiple partial we decided to try out the above mentioned solution +contract on interface +IMHO this should go to synapse release as well +They are mostly suggestions about code comments and style +Please file a cleanup JIRA for and utility function changes mentioned above and a separate one for multiple statements per sessionOnce that is done we can proceed to commit this +pertest etc still work needs to be may have been fixed at some point but I cant reproduce this problem with the latest trunk +AS was released we should use it +Current patch matches the bookkeeping behavior for TATTEMPTKILLED or TATTEMPTFAILED when were effectively ignoring the event +Sorry if you are also working on it +Closing per Daves comment. +I have the problem locally with wssj version Axis and Rampart version CXF version +This issue seems to affect all versions of Mozilla Thunderbird like till latest nightly build When trying to send SMIME encrypted it send usercertificatebinary requestbut get usercertificate +All of the browsers I am trying are failing +Bookmarks satisfies this feature for me. +Closing out the issus for the release. +From previous conversations it was suggested modernizr is the defacto standard to switch between minified vs javascriptIs that still the case +Ive hit this problem on pure tests as well +added Private and annotations +A jar is not recognised as a sub deployment unless it has EJBs or an or it is listed in as an +Logo issue fixed by Rev +It is likely? or something transient that got fixed. +I had to remove the line number part of the message because its not available for +OK I attached a new patch this time including the simple fix which is for to mark itself as having pending changes whenever the head commit point gets deleted by the deletion policy on initI will commit shortly +And that fileset copy is broken and wrong since it thus duplicates the styles +Too many linksBut thanks +Hoss have you tried? Im not sure where the state of that project is but I used to do a lot of work with it +Closing old resolved issues +If this assumption is correct this ticket should be marked as Not a problem +It currently waits for half the lease period then starts trying to renew leases once per second until it succeeds +As claudio pointed out the is not the employee object instance but instead is a wrapper around the bytes along the various message bits needed to transmit that data over AMQ +Please reviewSergey could you please your application against the trunk with this patch applied and let me know if that fixes the problem for you? patch works fine thank you +This is a very differnt exception then the one you posted on the tried this with and the current branch and it worked without problems. +I tend to agree with Steve I see this as the same question about contribs +I didnt want to force developers to create a record for every preference setting +Please give me several days +I discovered that the successful download was from the maven repo and when I downloaded from the main repo some dependencies in my caused trouble +and performance degradation on the attached micro benchmark on my machine +I sort of have it down but am still learning +Moved to next milestone release +But if youre needing guaranteed delivery check with Kurt +This skips the +So let me rephrase it If the view doesnt get renamed it will still be duplicated in the show view dialogsui thus naming it Human Task sounds much better to me to at least separate it from the generic task view and mylyn task view +restart of cassandra service doesnt fix this +Strange +This issue has been fixed for CR +It isintermittent because the mgmt server is probably loaded when you hitit and possibly producing an invalidempty response +Closing old issues +On some OS temp files use a differ file system +Mahadev Exciting! Any chance you could post a pdf or html version? +PS Thanks for reporting this and providing such a clear diagnosis +bq +The patch addresses the case I mentioned before since it each tree fold has changessent set to false for initial accumulator and the changesenumerator callback sets it to true once a row is accepted by the filter +I searched for a way to minimize the use of the +This appears to have been a transient issue and isnt currently manifesting itself +I have set the Xss to its default which is +Did you mean +I dont see any problem mandating that overridescustomizations adhere to a sorted order +I should mention that I adjusted the tests to be able to run cleanly in parallel in the branch +Patch addresses comments from Todd +Added the ability setget maxmin compaction threshold in interfaceUpdated nodeprobe with these new commands +Attaching and a new topic +Marking as closed. +bq +s open another ticket for the bug created instead of hijacking that oneSee +I think this should fix the problem but its completely untested +These should work from any timezone +You dont have a password UNIXPASS contains your new password for subversion +Thanks Steven +Even then the protocol is to move the statements to in the root directory when other people start touching the codethanks patch with copyright notice moved to +Tested and works correctly. +Thanks Ted +Feel free to reopen it if you disagree +The fix has been checked into the soa build +Hi AlexDoes your build fail with this? It seems to work ok on my machine as well as on the build machines +The value of is also used by the balancer if it is an HDFS URI +Basically this spawned out of a request of mine to support Ys custom authentication and access control system for the until Hadoop gets a security system in place +add +The geronimo commit also includes partial support for virtual hostsopenejbChecking in modulescoresrcjavaorgopenejbserveraxisnew revision previous revision Checking in modulescoresrcjavaorgopenejbserveraxisnew revision previous revision Checking in modulescoresrcjavaorgopenejbserversoapnew revision previous revision Checking in modulessrcjavaorgopenejbdeploymentnew revision previous revision Checking in modulessrcschemanew revision previous revision geronimoSending trunkmodulesjettysrcjavaorgapachegeronimojettySending trunkmodulesjettysrcjavaorgapachegeronimojettySending trunkmodulesjettysrctestorgapachegeronimojettySending trunkmodulestomcatsrcjavaorgapachegeronimotomcatSending trunkmodulestomcatsrctestorgapachegeronimotomcatSending trunkmoduleswebservicessrcjavaorgapachegeronimowebservicesTransmitting file dataCommitted revision +The dependency to javassist is picked up transitively through the dependency on +file is available +after restart snapshot taken +The patch that I committed for this issue isI discovered that there was already a section for SQL Expressions and merged the two sets of topics into section. +These three files cannot contain a license banner because they are an input and two reference outputs for a unit casedk +I prefer approach since it seems more consistent with the scenario folks seem more likely to use +Why is referencing the XSD wrong? I think that in all cases we should referencethe XSD instead of the DTD +I just commit a fix on the feature as the dependes on. +Is it possible that this was never backported to? +port made to trunk. +that was a bug +Ahmed and Alejandro convinced me since this endpoint is and admins already have the ability to get into the cluster. +for the reasons for not doing the full symlinks It will miss andor it is too complex +I have no idea if this has real penalty on namenode performance +Ok my employer told me he correctly just sent a CLA for Kemen at Apache Foundation including myself as employee so you know have my ICLA for the implementation +added a verifier +Actually in my example execute should appear last which would imply the same limitation +Also thanks to Daryn for his review +applied patch more or less in r +Thanks for the fix Li! +Resolved as suggested in revision . +Assuming this has been verified and resolved +The first time the server is run the default properties are loaded and saved in the server root as and Subsequent loads will use the local +Comment on from newJeff +Fixed with patch from Dan Gornstein +Cool +Excuse my ignorance about the build process for but when will the change appear in +Moving all open items for XDT to XDT since has been released +The new interface and class does not contains the ASF license headers +Committed to branch and trunk +we know +The patch been tested manually for branch as pointed by Devraj. +rc file needs hand edits for all releases +Attachment has been added with description SVG zip file that leads to font issues after conversion to has been added with description The PNG file after conversion with Batik +i think you can use IN as well +Enables setting of values from lognet node and code +We probably want to readdress the issue of how to vary the chunk size dynamically however I think it can be pushed back to post release +Consequently it may not be possible to open a decision table authored with or with. +This should be the final version of the patchThis includes Brian Eatons new signing interface and some preliminary hooks for doing proper signing +Committed to trunk +Proposition of implementation for this request +Anyway we can filter these symbols when we want to create a comment node in the visual part +will be committing now to trunk and branch +defer all X issues to +Could you put the patch in review board? I have a few comments which would be easier to make there +Any idea if the setting needs to be applied to both server and client? or just client +Fixed in CVS. +Done +JSP used in the class for the backing bean bean used by the subject +Perhaps make a separate decaf target or something +This is the upstream issue does not separate these two scripts with commaWe would need to simpler sample for having this fixed +Finally fixed in SVN which changed two more throws clauses which werent changed initially. +Please let me know of any difficulties you have. +Moving over to lives there. +In the name field enter JAVAHOME and the variable enter the installation of your java SDK +This bug still appears to be present in JDK +Heres a v patch +It just means that some exceptions from the shell arent going to necessarily contain the original path given on the cmdline +Well it looks like there are two issues here +usernames not real names t the idea that it searches by both now? coreyhulen but usernames are guaranteed to be unique whereas real names are not and with k users there are quite a few dups ambiguities +Thanks for the info Jon +Ill fix that to be the actual file nameThanks for all the quick response +The tests under still fail +Another feature is selected! +Server Manager Reference Guide The only place specific for JBT is Figures in section +Should be addressed in trunk now +Sorry I forgot to make the paths relative on the last patch +Also we need a profile for integration testing similar to hadoop where all the dependencies required for integration testing should be definedCurrently is defined as dependency with type war +I had noticed a related problem where the Configuration was not passed to the when I was trying to give it a Configuration with added +I guess you do a JMX port scan and query every candidate you find +Ping +and users can always just change it at if users uninstall they also remove their workspace +Duplicate messages are called by the fact that when richgraphValidator is used RF is added together with JSF bean validator +As per notion of owner region will be this issue since there is a design change as suggested in which will invalidate this use case. +to Paul smiths proposal that we create a artifact +Who are you waiting for approval from +I wrote this test because I knew it would fail and that on reconnecting clients dont jump back into the leader election processSo I also added to the client reconnection impl on reconnect all are +No reply infrom markmail +I will have to do a trunk build from home as my company firewall does not allow my build script through +I had no possibility to put the issue to resolve status on December due to power supply problems in Moscow Branch. +Howard is it your view that this is no longer an issue and that should be used instead of in both XHR and requests? +Fixed in rev +It seems that no one else is having issues and I was told that it isnt even a supported way to run trafficserver anyway +Once this patch is committed can we make a release +Go ahead if youd like +converted screen and and +If you still have problem post it on mailing list +Resolved by new module importer for but for release looks like the new importer will have to be enabled explicitly as will not be the defaultNote though that in the new module importer when modules are reloaded they are loaded into an empty module object +if you would like it to be released earlier start a thread on the mailing list +Hopefully the only remaining work is to update golden files +Please find the attached files for more problem occurs when the thrown exception is in a different package from the service interface and the exception is not in +We built Grails trunk +So assigning this to him +Cool that a test caught itThanks Ted +this is not easy to do that with resource plugin or something to maintain file on resourcedirectory is another solution but limited because we have to be careful at line size and so +Changes look fine to me. +Moved wizard implementation from to provide steps how to verify this with JBT. +Also since requires to be set for a fully distributedso you should modify the srcjava file +Makes the new binary format a vint length followed by N Texts +Im working on expanding the changepatch to allow configuration parameters to choose asynchronous execution and set a wait timeout with acting as wait indefinitely +A test to demonstrate the problem and a proposed patch +Sorry. +Thanks +We could also run current tests against old versions of the daemons +Plus we are working to get a GWT archetype into Centrals Create Projects section right now +i took a quite peek at making a patch for this for the lucenetrunk branch +This causes it to throw an error if the checksum doesnt match +This patch addresses the synchronization issues and also replaces the copies of Doug Leas concurrent classes with a dependency on the classes +Code review and rewrite done in geotidy +I believe that the same piece of code is responsible for as well needs to be reviewed quite carefully +headers are mutually exclusive +Let me know +To fix the problem check and it must be greater than or throws exceptionInvalid value +Thanks. +Another thing thatd be great to fix about is its intermittent checking of the some docs had more than one token in the field errorAdd a that only allows one value per document +Sorry for the multiple updates +On the producer side however it could easily be adapted to become an out only endpoint that would write keyvalue pairs to memcachedFor that matter a consumer could be written so that it periodically polls memcached for a given key being present +I have just fixed a related issue to this +Do you want me to make tests standalone and run them with Java and code compiled via JDK targetingYep if you can build using JDK the target will be set to in the parent +The stuff that posts the link to the users wall should rely on this new facade instead of doing it for itself +Hi JamesThanks for contacting usWe might be able to find someone for you to interview +Sounds like a good approach +Setting up the ensemble the way we want it in the first place saves about s on my machine +since for Eclipse we do mvn cppy dependencies I think that this LIB is copied anyway right +intermittent failures in test introduced by the issue lang in this changeSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestslangSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision Deleting javatestingorgapachederbyTestingfunctionTeststestslangCommitted revision +Dont want to surprise anyone with new orderings in a minor release. +Embretsens commentIn my opinion in the sentence The Derby Network Server can run as a server with Derby as an embedded part of the application +verified on +Awesome +And Ill also look into the SWC not being generated if no apps are defined + +Attaching source patch +nd thought wouldnt it make sense to do this by reverting the Preparing for release change? Looks like the and needs to be edited to +patch to move logic outside the candidates loop +Workaround use +I have tried IBM JVM SR SR and SR +Try git clone git +So worst case the String test seems preferable +In case you can see what is making rat angry by looking at the created in the target directory of the module +LGTM pending jenkins +thanks daniel and shravan! +Is there a workaround +I think inner classes ordering on imports is different its usually placed below +Theres no point in having each plugin copy paste all the same code to do the same actions +Could you try out the trunk version and see it it works for you +Block finalization is delayed until its files lease is expired +The patch has been applied. + correspondence I included in the link in the original message refers toI filled to cover this issue +Ill leave this open to make sure we get it resolve with the guys before. +Currently has one Default and is otherwise changed from outside by user controlled code +good point will consider what kind of problems +What about IBM Java on SLES? Just asking. +Looks pretty good thanks for figuring it out +Bulk close for release +No +Run the tests again +Knew that since I recently added some helper methods to. +It is still brokenJun PM parseSEVERE Class not found! Class not addedJun PM parseSEVERE Class not found! Class not added. +Vikas heres the updated patch with my perf tests the easiest way Ive found to run isbq +A separate patch for branch is required +Committed to trunk. +Cool Ill change the javadoc and resolve it to a documentation issue +and for in rev +Talk about it +This is because requires usage of from SAAJ whereas geronimo JAR contributes a SOAP Factory as wellThis code has been tested on Tomcat with the Sun RI. +i think the directory and space must be handled with care and it has to been discussed with Ersin prior in detail what to do +committed on on trunk awaiting qa available via superoverlaytrue appened to a kml reflector on +The error only occurs on one of the servers but it happens to both other two nodes work fine +It would be nice to get this in for the final release +Hey everybody let me know if there are any objections to this +About the build This should be a Lucene module + +To go further there should be no deprecated code in Lang +I remember well there was a discussion about this and the conclusion was that it would be difficult to release a version based on Axis because there is already some code in Synapse that relies on featuresfixeschanges in Axis +We can get this supported with the new Camel properties +startstop service yes however I mean startstop Sonar +This issue is closed now +Hence it feels like it should set up and clean it +Thanks +So now all runs faster? Its just that the disparity between add and decrement is greater nowWhats taking the time now you thinkWhat would you like to do +Krzysztof who reported this bug says that the fix works for him +Asked him to retry +thanks. +v fixes the to the snapshots closed. +Changing the issue type back to improvementI did see omitted onActivate calls but was unable to reproduce that so it was likely a logging problemCheersNick +The correct behaviour should be not invoking the method inwhich case derives a package from the target namespace. +It is this version of Ingres that I am working withTake care version of changes for using Ingres or newer withSupercedes previous diffs uploaded +I wasnt able to reproduce this +Tested PR with client no API changes. +targetfind name syslog xargs is used as well and also lives in main +Ran the test case on trunk rev and it passed +agree we never allow the server to auto deploy +Thanks +What happened to the patch? Was something wrong with it +Folks it would be great if we have a consolidated document that describes the design and some details +Committed to trunk with revision and to the branch with revision Im closing the issue +If Im wrong please. +Reopening issue because in constructor are still happening +And one thing strange is that if I myproxy name it passes anyway +Once and are done the Git patch of needs to be updated. +Patch applied under r +This information really helps +No test case yet but this seems to fix the problem +As discussed in this change while necessary to move forward on will cause a regression when a new client tries to connect to an old server with a database name that cant be converted +VOTE accepted Proceeding with IP Clearance process +When set to it will refresh the cache for every operation +Since there is no cleaner way to address this at this point we decided to move this out of to make a release. +This has been addressed in r by getting rid of the iframe structure which means you can simply hit the browsers reload button now +Scaled back the aggressiveness of the fix for +I took Bretts db and tried upgrading it with latest codeline and I see the problem there too +I quickly tested synchronizing the openclose functions but the writer still ends up being null +I was thinking of replacing type and element to target or targetType in both but type would be at least worth considering +This repairs problems with some of the settings and adds a template for new +Let it keep and did some code cleanups +On the cluster made sure there were no regressions and the jobs ran fine +DaveI agree the interface should be extended to support the channel name +If no error occurs things goes fine but if error occurs things were not cleaned up. +This updated patch provides and to be used in instead of and when wishing to activate the reloading of classes +Any ideas? to issue has been fixed by I thought that there is something behind this but it is really simple +to and this an issue in as well? +Theres no sane way to do this with getpagedslice as it currently is +Please find attached patch for the fixThe fix will go in both the release as well in the trunk +When testing a customer issue on this was observed +There is one issue though +Refactoring support now includes a participant to the Change Method Signature refactoring for the case the method name is changed. +One solution is to have just one copy of jobconf with the jobtracker and delete it upon restart or job expiry +CB Jira project changed to go to issues list. +Ill commit this shortly +Okay I think Ive narrowed it down +Could you try to set forceipv and rerun it? My working machines are all Linux boxes +Committed to trunk +Patch to use detaching of values before holding on references to them in reservoir sampling +Moved Simons changes to force the use of the sorted variants to use fixed size packed ints to + +Thank you PaulexTim could you please close the issue +It is good to me +ah sorry I mixed this up with a user forum posting about pageflow +Thanks for that! Adding will comment on the issue itself when complete +Fix is included in the Synapse branch and now the throttling related samples and integration tests are workingA future release of WSO Throttling will properly address this issue by implementing proper support for Neethi upgraded to new throttle version Test +This problem gets fixed with patch provided in JIRA JIRA and JIRA Unassigning so that it can be closed with dependent jiras +file due to +nit why the increase in number of workersIts far from nit +Actually Ive tested on master branch before change the status to fixed +Seems Existing Jira Accounts in the list have had group removed +Other files are new +I have committed this first slug of work at subversion revision I have some suggestions for further improvements I found that I had to create a subdirectory named system in order for the tests to run +Just deploying the attached from the AS testsuite will cause the same error +seem to fail because the method canWrite in returns false on both calls for +It seems to stop fine for me +Im still seeing bad degredations in solr I think its because the default way for solr to sort strings is with which isnt specialized +for a modification at the decoder level but it will be a lot of work +I would like to go with the version of the patch uploaded on Feb rd +Just a quick note +Umm yes actually this is somehow irritating +Well +Also porting a webapp from JSP to facelets on JSF is not really a big problemJust a thought We also could use which we add automatically to the tomahawk components to render the needed javascript +The default protection after upgrade should continue to be and that folks should need to explicitly tighten permissions rather than explicitly loosen them +Patch committed at level because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. + +It was in the version after the one Allen tried above I think thats why he may not have seen it? Please reopen if not +The best I can think of is with a model that is creating the view or procedure that way all the other models use the previously cached metadata and only that particular model gets reloaded with your new viewprocedure +remember this would depend on the file system used for be the case insensitive to catch the bug here. +Im facing the same problem here with surefire andA successfull workaround is to downgrade surefire plugin to +The script needs a minor update as it complains withsvn srcsiteapt is not under version controlNow there are related files +committed close for release +Thank youDoes the lead replica has to iterate all and get the latest value of th counter before applying the decrincr mutation? If so the read path can be a performance bottleneck +applied test and close issue if ok +A small value could have an impact on performance due to faster flushes +Access to connection capabilities has been added toAdding access to properties for the pure Java implementation would be trivial but this is not yet present in the C API. +The model uses the Maven GAV to identify the project where the Forge plugin resides +But is it a good idea to always do soCant you come up with another testcase where the opposite will holdThe reducer has a highly selective filter at the beginning followed by a selectIn binarysortableserde all the columns are read whereas in lazysimleserde most of the columns are not even materialized +Its okay we can keep the patch here +This happens in wars as well +It was to enable something for the fisheye integration +Minor formatting changes are included as well as the code base doesnt seem to fully follow the Maven formatting conventions +Let me look into where it got stuck +This is the generated by Ivy unfortunately the report xslt runs into a similar endless loop +And we consider using a new pattern for Camel that James suggested is a better header nameBTW We also have a throwFault DSL +As long as that all returns success Ill commit it +Restructured Region constructors to workaround constraint system limitations +Please note that this patch includes all changes from the previous ones +but Ill shoot a note to the branch owner to see what he thinks +The internal conversion is not the developers problem as it would have to be taken care off by the user themselves + +jchochol asked me to approve on this so here it isYes mickaelistria user should be allowed to push stuff to nexus forjboss tools releases +Thank you for the great work +Thank +this is asking for outdated and wrong documentation because once we will produce a final release nobody will have time to carefully verify that the docu is updated +move to normal priority pending further there is still no new release I cant go into an evaluation phase +I got this same error with JDK but it went away when I removed thefollowing jars from libextFor me this was a error +I think my plan is to postpone the large refactor until after this gets through +Seeing two issues there actually +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved +Other than that there are no apparent problems. +Looks like I can only reproduce problems with primitive arrays so fixing those properties seem to work ok +I can confirm only that most of the JSP tags that are expected not to be visual are rendered as green horizontal lines in VPE +Matt are you still working on this? If not do you mind if I take the task +I suspect you might have grabbed a bad build since running from CLI fails for you +You can see in +ill create a patch once is applied +Just to clarify Im seeing the error when both FLATTENEDTEST and FLATTENEDTEST are in a separate schema from FLATTENEDTEST +Updating fix version +Committed at subversion revision Touches the following filesM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsD javatestingorgapachederbyTestingfunctionTeststestsupgradeTests +we should make sure the admin user is there the attached patch is a good proposal. +The two extra test queries specify distinct for the results +So we cant go ahead using Spring jar because it is not officially released +and if it was an intended feature then whoever did it should fix the unit tests +It has been tested with trinidad +Is that something that should concern me here +Say what one might about Oracle that report in was the intersection of DBA and developer reporting +So I guess we need both modes +Closing issues +This is a bug in +has to be patched too +I believe Alex is correct +ok sounds the bean properties to system properties and updated the admin guide. +There are some issues regarding CMP which is described in the forum post and requires more input +This duplicates my comments in variable scope changed in sample. +This is not a Derby bug but a problem with the JIT in the mentioned JVMI hope the reporter closes the issue if the resolutions seems ok +As I mentioned the Sesame and parsers and writers are limited to as they follow the and specifications which still explicitly say +need to narrow down where the problem lies +confirmed this fixes the build for me +We have not experience this problem when running the JDBC driver with our products against our development Hive serverI will browse Jira for a case on this +Patch to apply ASAP +I think we schould do it in the same way +IMHO theres nothing but some wording that can get improved but we tell the user what operation he triggered went wrong unveiling the internal details in the details? +Committed trunk revision Now backporting and adding sophisticated backwards +I created issue to track this. +And since thats what people decided on it probably behooves us to respect it +Havent forgotten about this its just more difficult than we originally imagined +This is fixed uses always to ensure that the +It seems not necessary that convert is public as wellDonebq +I understand that it is a bit more expensive but I thought it should not matter much given that the code is a bit more clearer +Sending out hinted data needs correct support +Shinobu do you have time to regenerate this patch with the latest revision ofthe docs? If so Ill apply it ASAP +on test worked locally for me and is unrelatedCommiting to trunk +If a were an admin I would be able to resolve it as a duplicate of right? But I cant the context menu does not give that option +This patch depends on the class found in in trunk and closed with release. +line esp +Need to add compression support to your webapp? Register a compression filter on the chain +I am not in your country since I live in Italy +Assuming that the NPE is because of the scenario I have mentioned above uploading a straight forward fix +remove the static method updateReplica and change update to +Implementing WritableEvery class returned by must implement Writable +Fine with me. +Would that be an option for EJB as well +Because this is what the gadget rendering code does I think the social side should probably do the same +Jonathan do you fancy submitting a patch? Then the new feature will be added in both java and spring DSL in Camel We have plenty of gaps in the spring DSL and we need to close this gap +Events onslide and onchange added +Ive committed it fixing it as resolved. +Closed without integration test. +Resolving platform issue as the project peer has been resolved +week attribute produces week macrodef on clientweek attribute produces no macrodef on verify and close shure its working under +etctrafficserver This issue has been pulled out into Jira Ticket As a recap you need to set in proxyconfig to a valid interface like ethYou can bring up Traffic manager on linuxsudo PREFIXbintrafficmanagerTraffic Manger will come up with the Web interface disabled but you can execute the sudo PREFIXbintrafficshell command and see some statsWhen this ticket gets closed a new ticket should be opened up regarding the Traffic Manager +please try again +currently waiting for root to create setup and SVN access granted. +If the file does not exist no message is reportedWhy should there be a message? Maven can try several repositories and there is nothing wrong if the artifact is missing +I tried that alread but I am facing with another problem underAfter I put +The fix introduces the new property createdAt for strict comparison of dates +It affects all deploy goals as well +This sounds like a great improvement +Thanks for the detailed analysis +We are no longer shipping rcp for the same issue occur with the plugins +in release +This should probably be optional +Ok i will drop them +Thanks stackIll fix them +was just closed as invalid because the data is stored in a blob not a varchar +When is this fix and the related deployment fix scheduled to be released +So dumb question +This is not quite complete +I would be great if you can go through it and provide some feedback +Closing per Bryants approval to do so. +this patch to master +Ill get it done tonight +Ok +Please ifwhen reopen when updated +Also generally review that all historic files are being managed as expected +settingexport REPOcdhuand then running whirr was insufficient at least for Debian flavor +Im pretty sure it will make it. +Committed revision Thanks Tang Yin Qi for providing this patch. +Is there any news of this proposal? It was close to be committed years ago +Checked +As agreed with Stphane Nicoll Im attaching this patch to this issue for review +Im guessing because of this patch +The logic to do the optimal palette computation is in the custom map producers and we need to move it up to the this is some mediumbig code restructuring not a monkey see monkey do thing +when the nd nodes web app is hit it might destroying or corrupting the session of the st node +Moving these out for for now +rebased for as Let me know if you want it for update on this? Its been patch available for a while and would be nice to have committed +I filed a new one +That means the compiler has no information on dependent files +This is an issue in patch has been sent to the mailing list +We are using hadoop this issue will be applicable to that version alsocan you give more deatils on thjis points +If your stream from a socket has some custom stuff then you could write a custom consumer on the search side that builds whatever search construct you want +I have it and cannot reproduce the problem +Sorry for missing the queries earlier missed them in my emails somehowI have a fix for this issue +Apologies I realized that the generated protubufs should not contribute additional findbugs warnings because they should be in +Ill commit code to write the replication state through the directory in a moment +It looks like this issue is duplicate of +Please see There is one bit that needs clarification the last item in New features says Filename extensions do not change +So I guess I need to describe this way +However now the problem Im seeing is that the rule is activating twice! Continuing to investigate +This would require pulling the incoming HTTP headers and feeding them to the XSL as parametersAre there other request handlers that could use parameters? For example the Velocity handler? How can it localize Velocity applications without the HTTP local header +gzip and lzf compression codecs for Avros file format or Avro and s for mapreduce or perhaps a version of Dumbo that uses the Pipes protocol to more efficiently get complex Avro data in and out of Python programs etc +I wont have time in frame for so I prefer to postpone it to the next major version +Feel free to open additional issues specific to DB and Oracle +Attach new patch rebased on trunk version +Please reopen the report if the issue is reproduceable and do attach more information or ideally a small sample that exhibits the problem. +bulk defer of unresolved bugs to +I can operate on your trunk to add the streaming +Possible fix which converts the relative error to an absolute error +To be verified. +This patch should be applied to plugin +I continue to disagree +Thank you +pushed in Developer StudioVersion Build id Build date +Thanks MridulYour patch is in rev Sorry somehow I missed to close this issue when I committed the codeAshish +But right now Im focused on other projects +What container are you using? Jetty? I had problems with Jetty loading resources in Eclipselink twice as well +Indika Thanks very much for the fix + +After restarting the situation was better but it went wrong again +it seems like this is an attempt to make a bit less mini +Patch for mvn curly a new patch for branch to remove mvn references and added curly braces to if statementadded a new patch for trunk to add curly braces to if statement +But in this case it might be easier to handle with puppet +Hi JayI am going to change it to the way recommended by the doc you have posted aboveThanksThanks for the detailed update +I checked and and applied the necessary changes on themLeo can you think of any other places where this needs to be changed +This time in format which could be parsed easily +I put in another fix for trunk r +Rather than assert use to check for a fencer? Im not completely sold on the idea that we always have to configure a fencer but we can address that later as an improvement +Manually deleting the file should work +Thank you AndreaImproved it with r. + +In r. +should probably included as well to hint that the line can start with the address too. +In particular the interface is very specific to the changes in those particular versions +the attachment is still useless for most people so if you could replace it with a normal zip that would be great +Reopening +I will work with this some more to see if I can implement your suggestions +Ive tested the sample email and this is now working with trunk +net bobcert then I get a message into my that states The clause was not wrapped with the required encryption token Googling seems to think that this is related to a bad cert +Attaching a dont add notice for other folks meetups but making an exception in this case +Given this has been reviewed I will commit it once I get from Jenkins +Two has been the stated goal since the projects inceptionI prefer two because it permits more code per line while staying with columns and using descriptive identifiers +Upgrading to newer version as detailed in will be done as part of +Added dependency to surefire to generate XML report +I really dont mind rerunning the tasks +Any ideas for more scenerios to be tested + Username sshastry +Even if you pass in same each time +Anyways if someone wants to pick this patch up and is willing to do a bit of testing please do +Dropping from +Test resultTests run Failures Errors Skipped The above tests includes the new test for load +I appreciate all your help and the prompt commits to and. +Hi guysWe need this capability so that at runtime we can clear the cache when certain application level events are triggered +Attaching patch to fix this issue +reopen to keep track of this problem +Releases happen when someone takes the time to push one through +This has to do with what the compiler thinks the source folders are +Maybe we can hack around an thing to use here +However I kept it in +Check that the application server and JIRA have permissions to write to usrlocaltomcatattachmentsTOMAHAWK patch replaces and also includes the changes made for +Do you want to attempt to provide a patch with a fixIt seems you may have to check the length of the key is to avoid the NPEAnd if possible an unit test would be really good as well +Ill be removing the examples from from the rpm and zip for RHEL +This duplicates +Rossen I just switched to the latest SNAPSHOT published in the maven repository and it indeed seems to work as expected. +The documentation does not contain an information that only one database for all components is supported +If you use blueprint we have a component that can debug easily. +we can commit the fix and have Adjoa Poku to test out the snapshot +The same correction has been made to that edition of the guide but for technical reasons I cant make it available on docstage for review +I have committed this +diff against qpidcppsrctests test script includes my test which causes the crash +Postpone to because JNLP links should be in that release. +the patch provides version d of this change for branch +I fixed one more potential place with race condition in statics in this file +Re shingles see lives in contrib with analyzers +Patch adds a new option to the fsshell rm and rmr commands skipTrash which performs as expected +revert the changes made by TB use TBs Modal instead of one +Is the max optimal DWPT size related to the size of the terms hash or is it likely something else +Nice to have +Can I take over this task instead +Attachment patch has been added with description Proposed you please provide the exception or describe the corner caseIts not exactly obvious why your patch fixes this problem +similar try to unify conversationPropagation type to NESTED instead of have fixed NEST occurrences in documentation and in Seam UI component but I cant see why it shouldnt create nested conversation if you try to use slink or sbutton with propagationnestedAre you sure that it doesnt work? +Not reproducible. +Sure thats fine ill fix tonite +As stated in the problem report I dont know +Thanks for the patch. +In the interests of keeping the bug database clean Ill mark this WONTFIX since it doesnt look like anything we in can do anything about. +The fix causes test failures on and +I will make necessary changes on our side to be in sync with your latest revisionmany thanks checked in +Sorry! missed that again + addresses this for oozie +Indeed I guess we need more details on why this is not automatically translated in the given scenario then +Nice patch Harsh +Simply commit as is +There is no product in the session for an user of a back application and is not available +pushed +I have created an account on as username headius +Any update on this +Committeed to branch and trunk. +Dont forget to apply before trying the CA portlet +On long term we should replace the logic by a faster alternative like Jolokia +Alejandro We can keep my suggestion and remove the translation code from now +Much like in user module +This fix needs to be applied +Given that we cant reproduce problems with the Remember Me feature and it works as designed for other people we are closing this should be reopened if a way is found to reproduce the issue. +what was the problem that the patch fixes? BTW the logging level shall be error not info I can see Project uses classes from transitive modulequoridorjar which will not be accessible at runtime +The was an older mechanism of determining what the destination type was +alternatives or +Testcase follow steps in +Thats why I thought Jira would be more preferable than the forum in this I can continue on the forum as youve pointed for reply modification to JAVAOPTS would break script on LinuxNamely the block that checks if java is is executing java JAVAOPTS version grep i Adding would cause exception because the class is in which is not loaded at that commandAs result would point to wrong locationThe fix is to remove JAVAOPTS from that checking command and specifically check for d or dvr +Ive attached my patch +Making the method abstract would also be incompatible and break subclasses but in a way that theyd easily detectPerhaps the javadoc should just be clarified to better document this default +This is the modified source the previous this one is the correct one +Yes thanks for clarifying Adrian +Since we can create an outputFile which we could use to determine if some sources have been modified since the last run +Thanks your precursory workIm sorry to find your plugin so late +The issue seems to be with in the have a fix for it for console but in version at the moment will take a look at soon as well +covers the +Im sure there is a way to bundle this up into a single file but I honestly dont have enough time to watch the several hours of training videos it would take in order to explain how to do all this from git +Closing the JIRA. +I am using the open session in view filter when using spring MVC to wrap a single transaction around everything which helps reduce the impact of this bug somewhat during MVC use +I have not come across any this morning +Affects version should also have added to it +fine close the bug +let me confirm thats the problem if thats the case yes a ticket should be opened +Section added to security chapter. +Patches welcome +bq +Please download from there +Could you please provide an example +We do need to update so we can address this then +Changing the replication strategy or changing the snitch to report different racks wont cause data to be streamed to any new owners +WSDL file of Microsoft SQL Reporting Services requestresponse for Render Report request to Microsoft SQL Reporting services +Just committed a new fix and some test bxml for multiple is preferable to leave things stable in SVN whenever possible +if you are going to do this you have to make sure the javascript is executed in the correct order so that modal is opened by a link inside modal and modal by a link inside modal +Fixed with SVN +Itll probably be early Jan + +Sorry I have no idea +defer all X issues to +to add more tests +It would be an adequate solution +Can you try this on trunk? There have been a number of Java integration fixes for and I suspect this is probably fixed +Assume closed as resolved and released +HiWhat is the status of this task? Do you need any supporthelp? I am willing to volunteer if you need any help for this taskThanks biggest thing we need to do now is merge in changes from trunk and make the corresponding updates to the model in the branch +Please verify in your environment +log from doing forrest +I can reproduce it here Im good +submitting patch only so that if the fix is ok with Doug Nicholas and others it can be committed +. +Otherwise today may wait whilethe new SR is being created +This is good as I think the current behavior is correct +I think Xerces XSD validation is not exposed via DOM +Then a new recovery thread starts +Test script takes care of adding +Note that it has been fixed as a side effect already so nothing was changed. +Thanks SianPatch applied at repo revision rPlease check it was applied as you expected +Working on a better version +would have included it in the previous patch +Changed to Major +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Like you said using nested unions seems likea bad idea tooAny ideas on how to unoptimize without loosing performance gain somequeries might be getting now? It seems the optimization was done for aspecific customer query that is supposedly improved performance by times +I like orThe patch mostly looks good and I agree with Vinods commnents on a get and setI am assuming that a container will show up repeatedly in until it is either preempted or removed from the preemption list +It cant seem to find +The earlier we can get it in the more testing can be done +I think it is going to be at least minutes as that is the scan interval +I got some feedback on Pig I think it is safe to include it! Included and sent for a vote +Good call +ve committed the patch +I could reverse the option so that it is disabled by default rather thanenabled + patch looks good +schedule new features for the next issues to next release +The called routines can handle manydifferent datatypes +Dave +The only way I can work around this is to narrow the search space which isnt a viable solution since it means that I can only login a subset of my users +Thanks wido i guess the same as file format are differnt as you specifed +Current design solves consistency but not consistency +IMHO it makes no sense to resize the combo to the size of the current entry +Hey Gavin I just drafted Rainer and Jrg as moderatorsPlease add reiher and doebele as moderators for the mailing lists +Just saw the intent behind this JIRA +Its yours now Joe +Thanks +I once tried to change it but ran into all sorts of problems hence I dont touch it m able to run T +Found a NPE when the Cas being copied is passed as a base view +since the issue doesnt exist in close it. +It is used as the image in the revised banner page +Oh and to be clear My plan is to officially support in Buildr and provide backward compatibility for it in Buildr versions beyond that +bq +Hi RobbiePlease could you review this patchcheers Keith +If you do not mind can you please name your files +They should be in the JAR file as well +It is recommended to create separate format instances for each thread +I run the commande on the specific module and I have the error with the code while pushing +OK I moved over the matrix module but there still needs to be some refactoring done there as there are currently two layers of matrix packages +This unit test is still +Ill open a separate improvement JIRA for this +testcase attached +And yes these are realistic figuresThe spec allows to dispatch events coming from different threads in parallel +worked for me Thanks +Well spotted +The test is further complicated by the fact that there are in use which will be populated with both uppercase and lowercase key values by design +Avery how does this one look +Yes. +The error happens when the destination removed arrives before the consumer removed at the network bridge +I have experience building RCP applications so if this feature is not too difficult and its just a matter of lack of time to add it I could try to do it +Much obliged to the quick suggestion +Please review this carefully before applying +Can you check if youve defined in which is what is read and passed to the databasecreator +svn ci m Applying patch in that avoids the test from failing sometimes on Windows srctestorgvapachecommonsclivalidationSending srctestorgapachecommonsclivalidationTransmitting file dataCommitted revision . +good Kay Kay +Can you reproduce? If not Id say close this issue and lets be wary going forward +Attaching diff from the recent TCK bundle from Sun +I might still have a pseudo cluster setup on my home machine I can check +Since the translations are ready we should wait until they are checked in +Fixed. +Hi RobbynA test case is a complete but small application that we can download or cut and paste that compiles and runs +Lets fix it properly for +Rahul do you want to give it a try? Ive added you as a contributor on JIRA feel free to assign it to yourself +Looks like this is similar to to the problemThe same code works fine from one machine and doesnt work from another +Hasnt been a recurrence in a while +Thanks to Hazem Saleh for provide us this componentI have fixed some problems on the examplefixed at revision +The problem is that the lock table statement is missing the logic to collect the required privilege at compilation phase thus it fails to enforce the required privilege needed by the statement at execution time +merge with branch +Any changes to configuration class semantics would have to be a feature only + +Is this fixed with? +recycles commit log segments after recovery +Align with its this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +If you think thats ok want to close this Jimmy +Your app must prevent such situationsIs it ok to close this issue with a WONT FIX +RobOne problem it breaks the convenient three columns screen we used to have +Revised fix applied. +If yes then you can post it in this issueDIGY +So there are a number of examples mentioned in the Beta documentation which dont appear in the Beta distribution +We cant actually extend he CGLIB proxy class because it is final but I can grab the interfaces from that class and merge them into the supplied interfaces I assume +It is not an impossible task +We hit a regression when server which has configured resource adapter to another server cannot be gracefully shutdowned when the remote server is not started +This isnt really a JMX issue the requested data isnt available through the core management API +unless theres a real issue id rather not touch itThats why I submitted the patch I think having the put and remove operation mirroring each other is important as it relates to the cache efficiencyNot wanting to touch the code is perfectly understandable +Close all issues. +My bad +Hi ThilinaI fixed this packaging issue other than that I can see some other improvements so I will create another Jira for code refactoring and give another patch with all the refactoring workThanks by Lahirus contribution +Change will depend on the resolution to that issueIve fixed the second oneHavent looked at the third one yet +Forgot to follow up here +Committing client jar +The patch adds new properties to enable configuration of user searching to make it more dynamic and capable +looks goodCan you add a comment about this fixAlso the variable names seem a little wonky +Would shutting Derby internally when a bad authentication class is found would be an acceptable solution? I have no idea what that solution will involve but I was just curious what we think would be some ideas on handling such a case +Viva resolved issues for already released versions +Both of those test failures were failing in trunk before I created the patch +It should also translate well to a variety of operating system environments and not require detailed kernel tuningknowledge to make happen. +Based on conversations in IRC with Gary Jonathan and Stu well genericize the arguments to the constructor into a Map so that we can add the configuration options for the there +To nuke this reference and only use the core cache key! Please dont make things in unfinal! This is bad designIt took some time until Shai and I understood why showed so much RAM +From there follow the descriptions for deploying portlets from the pluto documentation +Patch for coming please review when you get a sec +Applies Java standard to +So I am closing this issue. +It all seems to run more regularly +We would like to get this one in first because everything else will build on top of this +I think we should have not special option but rather show them by default as done for MISSING blocks +ve just committed this +Yes since code completion and validotor use the same EL resolver +Will this now break or other projects which also use Jersey? appears to use jersey +What exactly is getting fixed here? Also since url is only used to get the mime type not find it the file its perfectly fine to convert it to lowercase since thats what they did in Android + +Updated patch merged with subversion HEAD changes +Currently Im working with Sun Java Beta +ftpserver artifacts are moved now +Im attaching a patch to fix patch iterates through all the configured mirrors in case there are more than one mirror for a certain repository id and when all the mirrors fails the original repository is then checked +I havent yet taken a detailed look at the patch and I hope to do so shortly +However I wasntsure on how to reproduce this issue on using tables in +Assigned to Julien for review +committed to trunk +Here is a new patch with the required updates to Image model class +I reproduced the problemFixed the problem in r +I think it makes sense to increase the default minimum size to avoid repeated array growing and shrinking with small number of blocks +Could you just prevent the result page from being scrolled down when opening the modal window +So I fixed the againI tested and I got a proper resolve report +name concat issue was resolved. +Assign to during tidy up prior to +In retrospect the fix for was incorrect +Resolved after fix version change +During migration snapshots are migrated to move their descriptor into their known subdir +Awaiting in Jenkins +Andy has sent in his CLA +I just committed this +Tests are running fine. +This patch fixes the issue in my environment +No problem with the new version +The main challenge here is to deal with the fact that the various components take different arguments in their constructor +Could change the scope of to provided for the hdfs project so that jar is not included in the hdfs projects lib directory +java should be to avoid this duplicationMinor the default constructor can be left out +This contains the patch I just submitted +HiI tried applying Wills patch to the code base and it did not solve this problem for me however if I run my tests with Junit this class loader bug goes away +looks good thomas! thanxcamille does it look good to you +i really dont think its worth the hack for those precious few highly capable developers convenience +Reopenreclose to deal with migration bug. +I have been doing Jazz although I have been trying to do a bit of electronic stuff +Please if you think there still is a the same error as the thread starter of of of of Running Windows bits Java bits Roo to an improvement +Cheers! Only had to recommit the CHANGELOG entry for which was committed a few houts ago +When the header is being sent out it is nearly too late to fix anything +Fixed in nightly build which will be included inThe same problem existed with so I fixed it there as well +Will this make it into +I also tired my test with Beans and the result is millisecondsPlease find my as attachmentThanks with sample +Current plan would be to implement our own provider rather than wrapping the system one and not munging the password but Ive been unable to find an ASL licensced implementation to repurpouse +can and should be advised to have separate conf dirs for agents and collectors if they need different hadoop versions +Spring context initialization is moved inside spring module +I just committed this +It was an internal list and I failed to cc to you +Changing the dependency in the pom from hibernate to causes the tests to pass +I looked over the patches and It seems like at least the second one is completely dedicated to new users and has nothing to do with page layouts +I think it is resolved because tatletale report was reviewed and nobody filed the jira against differences between these jars in AS and Seam distributionDo you agree +The failures are due to start calling delete +It is pretty clever +This imports the local io instead of the build in io moduleIMO revert the fix and well address that after +Verified in Tools. +I found some errors in the earlier patch where the javascript submits were not correctThanksTom +ran integration tests pushed to master +Had to do some modifications to match latest logical plan syntax updates +Brian solid research thanks for the patch +The patch v add a check a reset for the I will see what are the results +We take one or more uncompressed messages and compress them using gzip and store the compressed bytes in the payload of a single message +Removing the previous destroy call is an oversight +I used java version +Thanks for your help and your great concern ChristianI visited your website +Seems fixed to me + +Resolved in +Well we could declare that duplicate symbols are allowed and that references should resolve to the first instance or somesuch +will it be easy for you t add that particular test to the unit test? The reason I am asking is because this patch actually removes an unit test if it could instead replace it by a relevant unit test that will be great +Thanks for the patch +I think the thing to do is to create a custom that will only load classes from and then make this the parent of old so that classes from are loaded from the classes directory and all other classes are loaded from the old jars when running the phases of the test with old +Applied to branch and trunk +We are so international +Lance can you send the patch in +Attachment has been added with description pink stroke is due to unexpected alpha values from arithmetic feComposite filter just read in the SVG draft that alpha compositing usingblending operations will be required to be clamped as followsThe opacity value should be clamped between zero and one inclusiveand the color value should be clamped between zero andthe opacity value inclusive +I cant find any differences in between EAP and +what is the use case that has prompted your creating this issue? +christopher validation runs on replication time +Thanks for the patch Dan! +we should look at getting this in soon as it will then allow us to cutover to sort order +kk thx for explanation +the test many This may show that we can handle say medium size map files during a scan with out much impact on speed +Save the attached to frameworkentityconfig and overwrite the existing one +The block is marked as corrupt in this scenario +Viva Espana! Arriba La RojaWhen could I get an official version of JBC with this patch +Ive just hit this one as well +Changed log level +mayank im referring to the file keeping the restart count +Workaround till fixed cd to uimaj and do mvn install N before doing the main mvn clean install in uimaj. +Can you provide an example +IF we decide to go this routeStill I think this may just be an issue of not using a when we should be and fixing that will helpfix everything +Also worked on a second style implementation that uses metadata in but I care for it even less than the approachIm going to give another go w and utilize versioning in HRI so were properly backwards compatible +looks goodOf course everyone should start using at some point +First pass at updating build +Why dont you run everything on Linux and check results? I ran all three on my laptop and everything was perfect +Nicholas I dont care about windows customers +well a replacement patch is ok +OK Ive updated the documentation to remove most of the sourceRoot overrides that were falsely implying the need to keep overriding this value and also added a recommendation to stick with the default +Patch v fixes test failures and adds new test cases +With the recent fix in this also must have fixed by now +I guess we can probably add this as an extra attribute to http without breaking anything +Andrew I think you should mark this issue as assigned to youDavid the docs you wrote look great +I gotta find some fellow nerds so I can show thanks for your work on TB it rocks +That would make most http clients work correctly +the old schemeThe test I was suggesting I think could be done with updatable cursor andpositioned updates as long as the properties of the cursor are such thateach next goes to the database +ipr with my IDEA and Eclipse project files seem to be out of date +I attached the files here +Therefore I close the issue. +There is a difference between keeping a static reference pointing to a live object and setting it to null +bq +I cannot use the attachment you providedCould you create the diff using this cvs diff u Thanks +perhaps my checkout is somehow out of date +Id be fine with a X that gives a different filename to load from the classpath +Alexey please verify +Sounds great +corrected code for aj for all remaining components +The dialog pops right up no items soldCan you get other dialogs to pop up +And it adopts the maven convention of itest +META +Lewis it worked I have attached the patch +Maven use NSIS but dont use dist to create the m happy for this to be a user driven thing rather than the NSIS plugin forcing you to run k Close this issue then there a particular reason that Maven doesnt just override the dist goal to make its binary distribution +Hey just so Im clear if the benchmarks run fine now should I not worry about the patch or did you need it to make work +Yes I updated byteman to +Karaf doesnt do anything like this and since most stuff will probably be starting asynchronously by start level soon I dont think theres any way to make something like this work +Debug logging when adding layer from IE with geoserver on a linux server. +The change is for the default schema only + +Ack well thats brokenRichardis this a reasonable thing to doJunany concern with this going inA meta question is how to handle the poor state of maintenance of this Hadoop code +Any idea when this will automatically show up +Thanks Arun! +re i think the rollover should specify what the type icon stands for +Ill also generate the for review as well +Doesnt seem like a good idea to? +In one can replace the standard DNS resolution code with a custom routine by using a customOleg +to trunk +We tracked the number of sessions assigned to each server using a load balancer +Any chance this could be changed to an INFO +Just had to reopen so I could record the release note in the correct place in the ticket. +I dissected all unnecessary information from themI assure you that our entire project is much biggerI even proved that the mappings are correct and work fine with Hibernate Now Im wondering why this bug report has been rejected. +You dont want to group the dependencies and thus change the order of the libraries in the classpath +I would appreciate if someone can look at this patch +Planned for release likely with upgrade to ANTLR to release as this would either way impact EQL and must thus be in a major release probably addressed with ANTLR upgrade in will not be fixed +Making description read more like a bug and be more descriptive for the release notes and updating Affects Version to say affects not. +to public +Hello PatrickIll improve the doc +trunk +here there is a quick n dirty workaround fix for as included with hardcoding the respective buffer and stream to +This is why the patch is failing +We may be able to code around this by rewriting the code doing the Introspection in that case invoking methods through reflection upon the javaassist proxy will still function correctly rightIm a little fuzzy on the particulars when doing reflection on proxies +didnt we go back and forward on this on the list? I thought the conclusion was to cache s +Ive tried to catch this but its not elegant or generalTests are included +For many Javascript framework there is a zipped version +My suggestion is to ADD a example not to replace as stated above and are very different products +Jenkins crashed during build so uploading same patch to kick it again +Fixed in dom +The write continues to occur to the DN and DN +Could you try this patch and let us know if it works for you +It works for tried In private Principal userPrincipalAre you sure you have the right Principal class +I want to understand the doc requirement for Derby +ve just committed this +As such I will not remove but simply the code to reflect the fact that we are on Java and higher +Also I changed package names added the license verbage as well as added maven and ant for contrib packaging +which situationOutch! I wanted to say solution +Looking more closely transient is persisted when its value is other than +initial drop for new logical plan the operator plan that returned has the same java objects as the original plan +Note the times I added the comment at the same time you did before you committed +WHERE AND EXISTSWHERE IN +Shouldnt calmslowDeterministicunbalance etc be named policies instead of monkeys +No response in weeks +We have a file consisting of a sequence of records separated by LF nRECnRECnAnd it is possible that some records may contain r +Never mind it works for Strings not for ints yet +Attention issue regarding the site seems to be that I have recently made a change to the EAP edition of the HTTP Connectors Guide +Thanks Martin for the backport +DanielI see in you commented that the patch there did not fix your problem +On the other hand the parser is probably created right after a so the exception has to be handled already +I plan to commit shortly +I think we can modify the testcase first and wait for upgrading to +Thank you +Closing this. +Apply relative to geotoolstrunkmodulespluginshapefile +I dont think it was there when i created the patch +If the property is not set the patch checks the system property to determine the JDK version +Theres currently only a way to create a batch scannerSeems like we would need the ability to create a scanner +Verified for as well +I didnt mean to suggest that this issue shouldnt be fixed I was merely pointing out a workaround that might solve your problem in the meantime while the filesystem semantics are discussed in more detail +Prabhat can you assign this patch to somebody on Richfaces QE for testing +I am having a similar problem with association in Embeddable +Attachment has been added with description XMP XMP Specification from Adobe +Exactly like thatEveryone else can we please quickly push a update to JDO to include the manifest fix thats in this patch? Long term it would be nice to come up with some tests to confirm proper manifests in all JDO artifacts +But I am confuse about your comment +Verified in GA build. +I commit this to trunk +Attaching +I have a patched version of AS with which Ive been testing iPhones iPads Android Tablets Safari etc +For case can we detect that jar file contains multiple coprocessors? Otherwise the class loader should be removed from cache +It lists snapshot schedules for a given volume +This is the addendum for +Thanks for your inputKind regardsMircea Corban +So unfortunately I cant figure out what JTA methods EJB CMT delegates too! Youre the genius! The error simply disappears after I implement your suggestion +Different people can pickup jiras and contribute +For example if we have a set of separate incoming buffers and were going to be merging them it would be best to balance memory consumption rather than letting the first stream fill up most of the available buffering memory and spooling nearly all of the other streams +This sort of thing often requires quite a lot of work which would probably need the Thrift community to drive it forwardStrawman the following is all that needs to be in the header of a file Licensed under the license +Creating regression tests +released! +Initial patch with refactoringsnew Windows classtests and native windows code to get resource information +Close as invalid +Rev . +apk files generated in the bin folder +It also does not work to switch over to the tab and make a change in the XML +The newer SSL certificate handling features in the catalina container arent supported yet but persistent sessions are along with new NIO listeners +classify the component +Will have to be some discussion whether its the right news! Thanks! +Theres a few lines of code which tell NOT to scan deployments and changes while Tools is still publishing +New looks good + +Fixed as a result ofPlease verify. +An FTP server could refer to the same file as an HTTP server etcI would support switching things from instance equality to URI equality as an improvement but I dont think it is worth our while to invest in a more elaborate equality mechanism. +What strikes me is that I get a replication started successfully even when it does not +I can see hte appeal of this but we should remember that there are applications where merge is a significant part of the reduce runtime +Addressed Johns comments on items +unit test and code added as described above +Perhaps unfairly giving this to Nick but I think he can find this faster than I can +On the forum they say to be working on a REST api +One of the caches verifies ok and the other three error with the null pointer exception +Yes Im using windows +Patch committed to trunk +Committed +The main JIRA will receive it the next time it is bounced +Koji Id stick it in contrib +Its possible I have something else on my classpath or somethingThanks! error I see is Pig does not ship +The backport is not complete because of package name problems. +Dennis Did you get a chance to work on this after your last comment? We would love to have this available to our users in +The web console is deprecated and to be removed in Camel + cancel attribute shows the same icon as save listHeght and listWidth attributes dont work done +The first includes all of the jars from client and the other contains only the jars from client which also appear in the Manifest of client +The once both run doesnt quite sound right to my ears +As per Katheys comment Ive removed the said method from the patch and will be running the tests again to catch the failures +KVM host addition works fine on the latest builds +Committed in rev +the example one would discuss new example items the main readme would discuss high level updates +The current solution is as simple as add +This issue should not be assigned to him +Favor to ask when this is ready to commit can you give a few days notice so that the rest of us can look at it before committing? Ive been keeping up with the comments but not the patches +i simply start another bound to virtual network interface +Thanks Scott +The few +Hi Eva we have a few more adjustments to the install instructions +I think this is as done as it will get. +If one confmaster is down itll just pull from another +we set internal status for consistency inside of IT +Can you tell me what you discovered? Iwent into JIRA and it looks like you updated the fix to come in butinquiring minds need to know the rest of the story +This can be cleaned up later part of phase II above +Promis jurca +We will have server corresponding to each shard +It must not be showed anywhere in the UI but always exist in the classic +the whole Multipart mechanism has a default ENCODING of and that what brings most of the probelms +Yes al though Im questioning the need to cover these complex cases since if we dont do it correctly there is a much bigger impact and the performance gain will not be worth it +In the patch from where is the orgosgiee artifact coming? You specify a dependency on but this is not the same artifact +Send me everything you want uploading along with a suitable html contextdescription there isnt anything set up for this +This is the code of the classWe do use query parameters but apparently the query is not cached in theThe performance improved drasticly when we eliminated the class loading +It would be nice to be able to instruct the compiler to throw errors when compiling modulesclassesmethods that we explicitly tell it dont use dynamic stuff +Fixed in revision that commit went to the branch but should have gone to trunkReverted previous commit in revision and committed the changed to trunk in revision +Oliver added concurrent textI think the doc is basically done but I do notice that the user guide needs to lose some content so its talking about +Test trivial fix +Feel free to test and add things to this module +r +Also confirmed on further investigation my best educated guess is that the problem is related to how bytes are encoded into Strings in the call from The byte to char cast was producing different characters on different platforms +As long as at least one link is optional and the binding is dynamic we create all the components and then later as we unwind and create the implementation objects add the optional reference when it is availableIf you still have issues with the current behavior please explain the scenario or add a test that demonstrates it. +Also i am redirected to the main page when i try to register an existing user id +Great that the patch is applied but while reviewing the changes myself I realized something the creates two jar files +TedI have updated the trunk patch and created patch of the other versionsplease reviewthanksAll patch verified in real cluster by my reproduce steps +If this is so the plugin should probably just have a private static final Object lock added to the plugin and a synchronized on this lock added to the execute method +I need a way to hook in my own implementation +The example works fine now and I will contribute an APT file once I get this sorted +apparently theres a price for the direct updates of the nodebuilderSounds like something is wrong +It cannot be applied as a patch +Thanks Steve +some error message would be patch fixes both problems +Victor your patch is ok as a first approximation please apply itI will create new feature request for other cases mentioned in my comments +I am thinking we need to build cpp client and run its tests on precommit build otherwise we could not guarantee the quality of cpp client +Affermative +Committed to TRUNK +This version silences the DEBUG messages +Any time I handle a bookie failure I block add completions until zk acknowledges the write +Should we have itself implement? I cant see any other way of exposing this bit +Please validate if this make sense so that I can proceed with the good to me too +The log files are deleted only after a successful backup +Fixed in CVS +Thanks Jason +this patch removes the zookeeper facade and makes it work with the current trunk +Ive been following along in parallel and have been able to get past the +We decided that a customer request needs a name of the request when it is not in the draft statusI try to find the create customer request option in but i cannot find itAttached a diff file which could solve the problem but could not test it +Looks like this change went in but not trunk or +Thanks Bikas +I had to do a few adjustments to keep the current behavior for the other read lock optionsWe may want to improve and look into if its possible to gather file modification file size details without doing a list dir operationI set the default interval to sec and timeout for sec for FTP +Henry Im not seeing the applications on the gsoc site +I think there are two parts to this i packaging Avro tools and ii harmonizing the version of Avro used by components in Bigtop +Here you go +release bulk close +Then each component simply needs to extend those classes for their own use +It looks like this has something to do with from +batch transition to closed remaining resolved bugs +Patrick Jarec Venkateshwhat your thoughts about the above configuration to combine file and memory channelHaving the flush frequency in file channel configurable could achieve both tight relaxed semantics for scenariosThis kind of setup would be a shared hosted setup in any org and they may want to combine different streams on the same instance wherein depending on data importance different semantics would be required for different streams +released +Daniel also brought up the issue that JMS is overwriting the ESB which would cause problems for correlating messages in monitoring +Prepare time decreased from seconds to Finish phase threads are notified to stop all together but not one by one with joins between stop signals +Nigel +Blocking other remove message into the pending space list on a send does not make sense for network producers +As far as Im concerned just aborting early would be enough + Tools JSF Pages Templates View +Fixed and checked to set fixin +Checking this with Hudson +good point should always have tests +I dont know about tests Ill hope for the best but prepare for the worst +Ill make a trunk patch later today +Fixed in r +Thanks Will Ill try and reproduce this +Any date set for the release +If client is passing timeout less than that there is no need to keep the read request running for longer +is a bit odd to me +Also if you insist that we sort ourselves on the outside you are saying that we should abandon the use of ORDER BY in the query syntax which is unacceptable given that our use of Jena is motivated by the fact that it implements SPARQLAnd it is not because SPARQL does not define what happens when a query is cancelled that an implementation cannot be pragmatic and do something useful +I moved property from to +that should remove the regression will work +Fixed +End timeslice derbynetmatsderbynetmats Server Process did not complete in time +Im not sure Im happy about having a public setParent method that ignores the input argument and returns +with +Sorry there is some confusion +published +Oops +the issue since the patch is applied to trunk +Can you come up with a patch for this? Its probably not hard but were pretty slammed with preparations right now +Hi MartinCan you please check this is only happening in our distribution or is this something that is in upstreamThanks just ran into a similar problem with removing of a User Database +I dont think theres going to be room in the world for two entirely distinct ways of running the example and I want to avoid that for documentation reasons if for none otherThe other reason for wanting the example setup to be identical is because the example organizational assumptions are partly baked into the code at this time +Hi KarelIm a bit confused +I dont know how I missed this but the class name is still wrong should be not as it is now +A tablet will have files in multiple directories including directories under different tables as is the case just after a table copy +If you run ant youll see failures in + on nice doc +what kind of retrybackoff should go in the client? Or should it just bail out and let the client write their ownOn thing that may make sense in the client is a bit of jitter to handle the boxes reboot simultaneously problem exponential backoff jitter may be even better +They are perfectly fine in some schema layer over Lucene but here in lowlevel guts Id prefer names that clearly state what the hell do they do without forcing me to consult javadocscode +patch looks good +Ouch it would have been easier if you had simply moved the repository with a single svn move +Committed to trunk +Again kudos to our Mark Butler for the backport +This is the main use case for the annotation and is less ambiguous especially considering that other annotations already provide default values with expressions where needed. +a JMX User may be restricted into which it can access or even see +As discussed not migrating at present. +Documented. +Ah one more thing what data source is used for these imagesIs it that oracle georaster provider you were talking about? What if a plain geotiff is used instead +Any news on this issue? I tend to close it as duplicate +Contains implementation for using Delta CAS serialization when calling a remote service +Bulk move disagree that this is a fundamentally different feature requring its own plugin +Instead we could just retry +Thanks +I could file a ticket against +Ill file a new issue ifpeople agree +out to be an error in our configs +Fix and unit tests to check correct detection of circular dependencies +Otto what do you think +However in light of seeing nothing in the traces it supports the theory that the cluster isnt hung but rather Im somehow not even getting that far in the Hadoop cluster +Resolving as invalid. +Jeff any objection to committing this +failed because the patch prints Fsck end time and total time as the last line +This is one of the only improvements in Ma though so I would wait for now +Considering its internal for test purpose only I think we need only one resource file because we will have only one language per project I think +Colm are you still on this? I can probably trying providing a patch on my own unless youre already working on thisHeres a better attempt at solving this problem +I need these just to see them on Projects tab +I will try to create and attach a sample project with profiles for the issue and also the new as requested +bq +Vikas +Im not sure if the cache is really usefull the char array for a negated range can be huge +Tests pass for me +This is ok? Could I import it directly +There are some slight changes around local filesystem temp file name generation etc +But the issue will still opened +After some tuning Ive got the following results with baselines to compare against r repeat summarize runs stp rtp lmin lmax lavg repeat summarize runs stp rtp lmin lmax lavg repeat summarize runs stp rtp lmin lmax lavg repeat summarize runs stp rtp lmin lmax lavg repeat summarize runs stp rtp lmin lmax lavg repeat summarize runs stp rtp lmin lmax lavg DURABLE trunk Revision repeat summarize durable runs stp rtp lmin lmax lavg repeat summarize durable runs stp rtp lmin lmax lavg repeat summarize durable runs stp rtp lmin lmax lavg repeat summarize durable runs stp rtp lmin lmax lavg repeat summarize durable runs stp rtp lmin lmax lavg repeat summarize durable runs stp rtp lmin lmax lavg +Patch applied +The unit test steps is the same as my manual testing +The plugins affected were Dialogs Camera Contacts Network Info VibrationOnce Fil chimes in maybe this related issue should be filed as a new issue +a few things +IIUC the issue is about the naming of the parameters +I tried both import and Update project configuration +I will open a separate issue against pool +For blacklist and release there are no dependencies so handling them individually is fine +Jesse mentioned to me this morning that Indexed shapefile support is in +Thanks a lot PaulexBest regards by Richard +Mind attaching a fully working test case incl +Happens for me with both and but only if client side validation is not disabledAny other idea what I could check +Changing the semantics may break their use cases +acquire be lock free +Right +Is this happening in the GUI or in the REST configuration +Test passes now but it should not +Sorry man you know how you load up the issue in your browser and then in the mean time while Im writing up a comment someone else changes the ticket behind the scenesYeah that +I dont believe there is a design issue that we have to wait for checkpoint to switch +Why are we adding null columns in the first place +Thanks Xuan! +I have committed this +You would need to run the in the Maven build to get the correct set +Why did you divide the issue about into separate like supports SELECT statement and supports DROP TABLE statement? I think that does not work until all the are submitted +Suggest using in the filesFix should include these featurestrunkvpefeaturestrunkxulrunnerfeaturesas well as the relevant plugins themselves +The svn logs for the ticket are for much earlier in the year +I have tried it just now +If its faster than it would be a good replacement for and +Also tried with without success +was itThe jar didnt have any classes in it +Looking more at streamy logs the issue is more why would first time through a code path trigger major compaction but then when we go to run it the same code path says that the store doesnt need major compactionPatch rectifies what was missing from first pass through major compaction check and fixes up loggingThe timestamp issue remains but is minor in scheme of things +Ahem conversation are already globally unique because by definition they are combined with the unique session ID +Timeouts occurred correctly after secs. +Thanks Amareshwari! +Patch for javadoc for could you please review before Daniel commits the text and before Corinne updates the docs +Sorry Jeff can you cut a product branch from the tag and add this change to that branch +JIRA is not for support. +Martin some very old issues I wonder whether such a feature us really required +sct values which would then override anything set the first pass +This file is the input to a configurefile cmake command +Well still want a memory for override since issues like what other code you run aon a node will dictate your preferred RAM allocated to +Army I believe that I have addressed all of your corrections with the exception of +Are you sure that fix version is M? I fought that we have released it yet +Still have to get my head around as to why +what extra should we add +DONE +newer version as of July at am ll need to check with Trev about the different bsh version we may not be allowed to replace this in the platform +That would cause NPE only if deployment passed in would be can reproduce this always +Now it should be easy to add something similar for other things also input and output formats combiners etc +Merged with the documentation +This will not be included in the release +Please reopen if you find it still to be a problem +I will try to construct a clearer test case for demonstration. +All the unit tests for are not passing failures out of tests +Let me know if you need help +Please let me know if there are improvements to be made to the patch or comments +Revisit issues that have of date for all Tasks and Feature Requests created before +Extracted out all messages to properties file +The execution of mvn as a fixed string is not appropriate for our system +Sounds messy. +You can register a null implementation of under the name application and simply do nothing in the code +Sounds like soft NFS mounts should be avoided as the integrity is the priority for edits and image +Manfred said this was legacy code from the early days and was ok to remove +GEFs still not supports +This is class is defined in the test part of the source tree +Is the above test case good enough? I just ran into this same problem and am willing to help get you what you need in order to get this bug scheduled +SureOoops wrong JARAThis patch adds abacus code to packageIt includes one unit test for the new codeAfter a release with this patch the user should be guided to use this package instead of using contribabacus +The idea is similar to height balancing as used in AVL trees +Do you have any benchmark results on this yetI am running some performance tests I will publish numbers in a day or two +Except for adding to the same test the rest of this patch is independent of the first patch for datetime builtins +I am writing a READMEUBUNTU now +Note that I discovered a performance issue that I will handle under +Jonathan I tried your patch and it fixes the problem +especially for updating to current versions this will avoid confusion about missing coresolder versions +But since the packages differ +bq +you should at least mention it +bq We should probably try and allow some sort of sanity check perhaps by loading up a tmp core or something +SvetlanaNathanFixed in ARCHIVE module at repo revision Please check that this fully resolves your problem +Any idea when this might happen? Id love to provide a multi module maven archetype for our projects and at this stage am unable to +The patch checks the state manager for the current state method and renders the appropriate javascript +The explanation is not very clearbq +HiYeah I committed fixHowever somehow the changed content in svn not get picked up by site +Are you using embeddedjsp plugin +Attached herewith is the patch that takes care of the problem reported here +bq +StackWhat do you think of my proposal above? last flushed sequence Id another option is to embed it inThis way there is no need to modify +were not using parallel reader but we are using direct core access instead of going over http + +Youre welcome +Wondering if we did anything specific in like post filtering in memory + +Now will compile with an off the shelf JDK +Not reproduced ever since +The name for my function may not be very helpful +FYI I ran into this because I am experimenting with different ways to index spatial data and want to index the same data with multiple representations +Connecting a Profiler like to the JVM on startup indicates that there are no blocked threads and no deadlocks are detectedAdding a mbean monitor thread to analyse the threads within the JVM indicates that there are no blocked threads when the node stops respondingIt does not look like this is a deadlock condition +The Release Notes and Getting Started Guide have been updated according to the meeting I had with Max and the information Nick has providedThanks guys added to guides as detailed in the locations provided by Isaac +There you will see the error +Both the targets you are invoking using antcall are run under the target +Thanks for the info +on the code change +i think it is very helpful me follow the correct waythanksbest regards apache velocity tool is simple powerful template engine that renders data form java objects to text XML email SQL Post Script HTML etc +that is probably because of another bug fixed earliertry to use the latest cvs version +Starting a thread per seek would still be costly +I think the best solution would be to introduce a new attribute in the allowing the define descriptor names to any suitable name. +We may should address separately +First time I have seen the Transformer class used and wow +I had the same problem not able to obtain the name of the logged out user in the listener for the event +I am not very sure whether the bug you are pointing here exists any longer or not +Initial patch +I was submitting the above on behalf of a customer if the flags are no longer needed this week then we can close this issue +Pulling it into buildlib nowWell need to update our maven artifactsbuilds as well but otherwise everythings been pulled in +I just added your fix to your live project and now everything works fine withThank you once again for your support. +Developers specifically download it from the release managers URL verify its mdsum to be sure that they have the untampered package investigate it until they are personally satisfied and then vote +Definitely agree there is alot more we could do with the layout this was a case of a light touchproof of concept +And as you say if there are problems that are too difficult to fix we can always back it out +Can we apply the attached patch to simply ignore this exception since it is supposedly fixed in a later version of Suns? As long as we are dependent on Suns I am not sure if there is any other option +I reverted the whole class as it is very sensitive to binary encoding so please leave it as it is +Pass it to Bob since it requires test code changes +I just committed this branch and trunk +It appears to pick up micro libc on your machine +It can be a hardcoded check as no one will want to allow such a search +Hi Eric attached which provides some details +But the invalidation fails because this is the only copy for this block +This big results in the following log statements The creation time is ahead of the current time +You can also try it yourself +At reduces you already have k g of buffer +Please reopen if youll see this after release. +Sure that seems reasonable +The full benchmark run completed with no errors +this is indeed not working as you said when stressing the the whole method is also resulting in performances loss +Thanks! then could you close this issue? Thanks in advance +In my previous comment replace all uses of declared with defined +Please provide a few entry points on how to do I believe youre working on this if not pls assign back to I am working on this +If it is not it would definitely help in fully understanding the kind of chargebacks you are referring to and what they are intended for +Its referring to the payloads of decorated messages +Im using the profile but ozzie is picking up ve verified that changed to works +Updated patch locally to use otherwise applying +Ive already worked around it locally but I wanted to contribute it back for othersThe big point of having interfaces is to not relying on Impl +Closed upon release of. +is still open since Jeremy plan to remove the hard dependency of xalanRex +Patch submitted +in the this directory +I am down to two errors which are related tothe aforementioned issue +Works in revision the demo rules are provided asThe log is provided as a List of String where each line starts with the level like DEBUG something happenedWe might want a more structured format +The javadoctest failures seem unrelated is there a jira for those? javadoc warnings areI filed a JIRA for the tests +NamitGot your meaning +Thanks Rohini +Attached patch documents the provider +trunk Committed revision x Committed revision Thanks Tommaso +was kind of wrong but it is not clear as to why oncomplete on hiding the popup has such an effect +Whether these need to be rewritten or not is up to the Messaging refactored tests do have failures which someone will need to look at at some point +Because it is one less hoop for a user to jump through +This needs fixing +Anyway if nobody is willing Ill do it +Committed +We are still awaiting the outstanding information for the Release Notes on this one +Doh! I forgot you can just use the C is operator +It works perfectlyThanks you Clement +Submitting patch for misspelling in secondarysort ve committed this and merged to +Ryan if you want I can try to create a solution for this and attach it here for further discussion and input +The new patch resolves all the problems +how to delete this empty comments +However as I have not yet fully tested my application with the enhancement I will be holding off my changes for a few days +This is not a commit candidate but for reviews and testing +This is because user can not add modules to Transport In Pre Dispatch Post Dispatch and Dispatch phases +Looks excellent to me +Committed. +It is a premature optimizations +Thanks Hari! +can you post the test case you createdMikhail can you post a patch with your? I dont understand why you needed to modify to break the infinite loop +Can you test with the latest trunkMany problems have been fixed sinceIf you cant build trunk just push a comment there I can provide a jar on my home page +If you believe this issues should be fixed please reopen and add reason for reopening +A couple of points after this runs the region it inserts appears as the last region for the table +jdk support has been support has now been completed as well. +If these are going to be tested as part of release then they should be tagged and released with the other examples +Im testing a fix Jetty before deleting the directory fixes the problem in Tomas I committed this. +OK but do you mean to take nightly now and test it or to take it in two days and test it then? be sure you are testing my fixes I would grab tomorrows nightly. +Patch makes sense +Some of my asterisks were interpreted as bold formatting but you get the idea +One line catch Erik fixs been committed. +ambitious +Committed revision +If we move BSP to API module there is no class which is needed for the parameter +I have set the component appropriately +Ive just commited the changes. +failed due to +Launchpad contains so I consider this fixed +Stay tuned +Changing the issue to be about all execution properties +Added a unit test to exercise the cacheadmin help commandAdded the path specifier to the help usage of the removeDirectives command +will be fixed in and or whenever we post this doc. +Fixed in trunk and fixes that was quick! Kudos to you for fixing it so quickly. +Fixed typo +This is definitely a problem as it also breaks protected methods + +Committed to trunk and. +Works now +This patch makes sure that hostname from the request is used when printing a wsdl +We probably need something like startindex to mean the same as startindex but without sorting +It should work with rev +Removed the pure master slave unit tests that were failing to build due to these code changes +Ive now updated the links on the site +Please retest and if this is still a problem +I have a reason for proxying I have a logging aspect +Agreed that upgrading makes sense +Checked with r closing as fixed +There was some talk in the old comments about scale and precision for output and inout parameters but as far as I can tell the scale set with register is not saved anywhere so cannot be used here +Done +But I dont see that as a priority for test codeMy primary objection to making tests dependent on contrib would be that it might substantially slow the compile test loop to the degree that folks would run tests less frequently +Fixed in r +It may also be useful to provide implementations of the conformance tests in the package if they are not already run against MarmottaIt may also be useful to look further at the module to see if there are any SailRepository API tests there such as that could improve the test coverage for Marmotta +In addition to the one here there were a few other class name changes that were not reflected in the message file +Im more OK with putting minor features in I doubt a KB patch represents a minor featurebq +The thread safety fix for has been merged to AS master +If you decide to switch on DEBUG traces for all of its only fair that you suffer the consequences +Resolving as a duplicate of. +Unit test is in +I fear that would be dangerous +If that worked it would be a bit easier. +Ive performed some manual tests and I havent been able to break anything +But for now we need to put it in Pigs source tree in order to compile +Oh Glade doesnt require embedded resources so my vote was not necessary +So youre saying that this works against but breaks against the snapshot? similar scenario including a works fine for me against the current codebase +This issue is down to the upgrade to Spring DM rc from milestone There was a change put in from m to rc where now the namespace handlers are now registered when a bundle is started instead of resolved so this is causing the issues have decided to use m for the release and this will be fixed post the release when we upgrade +Increasing priority to Critical as this is undermining trust in the reliability of the version control system +Summary of changes Prefixes prefixed get fix state stores renamed to state store test failures are unrelated +Anyway thats a Hibernate issue. +You can generate the second message with a schema with a targetNamespace just fine +wont reach consensus on this +So for example setting my current theme to basic and then previewing or changing it to x or sotto did not give these symptoms +thats not a bad ideaThough Id be tempted to turn it the other way since were starting from a clean slate +The old style Range Query is highlightable +Hopefully I wont need a full understanding of Xalan to fix this bug if the identity transformer is rather isolated +Committed +Error related to was because there were some left overs from identiyt tests which should have been cleaned up +Oops something went wrong probably OOME +bq +Here is the patch +With your modification the test is passing +Needs to be verified in Hudson +Updated to the latest rev available +Waiting for review to Mar +But that doesnt happen +Do not add multiple concerns into this one requirement jira. +Difering this as it doesnt seem like weve time to get it done for please shout if you think its important for because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I do a build of and do jar and verify its in the local repo +Bulk close of old resolved issues. +Devaraj can you rebase the patch and modify it to use now that is in? Tx +Thanks for catching it Claus +Test case failure is committed this to the branch. +Ahh my underlying issue was that I was adding the nature to my Eclipse project +A patch that adds coordinate sequence and geometry factory hints on trunk +I agree but the issue is that we have made releases which automatically check for the available memory size fields +Originally only one node had a corrupt file but i could only reproduce by corrupting the file on both nodes and starting Solr +I guess its nothing to do with file but with file +We need to Apply the again Update the Hudson job to not pass the python argument Im doing Nigel is doing Apologies for the disruption guys +Its too important to be postponed +Fix some typos in comments +removed unused imports for +This missing Affects Version +Sorry but issue against a database which has been EOLd for over years cannot be critical +think the problem exists in the thrift server mode also it is just more difficult to reproduceCan you run concurrent clients and see if the problem appears +Francois This is a separate issue from what the others are reporting +I dont think we should apply this patch if its duplicating info +root cause of the bug is in in and added an integration test in the assembly plugin to verify the fix. +try and fixing a few places that werent handling the IE correctly +With statistics we can keep track of which portlets are slow and move those into thread groups based on their historical performanceThe key use case for parallel rendering is when one or more portlets are taking a lot longer than other portlets to render then the other portlets will not have to wait on them as they do in the sequential strategy +DonePS +Bulk move of M to M +Changes made as requested +Hi Arkathe problem is still there and arenot displayed in directory listings please follow the recommendedconfiguration as shown in the RegardsHenk for pointing it out gavinI have fixed the issue +Looks good cancelling patch for a minor update to the log message +Started trying to do this last night +I have just seen this issue on the branch +Therefore if you attach the underlying pdf to the clusteredPoints arent you actually turning kmeans into fkmeans at least by some loose definition? or is the weight calculated differently for fkmeanssee +Added output from the latest attempts with the patched patch does a better job of handling references to PyNone let me know if this helps +We create both and Session beans if service implementation is declared in which is not correct since this extension should create only a bean for and another extension should create a bean for Session +im seeing a similar issue +rhotrod p cacheconfig + +Verified by Richard +Postponed +Ported and tested +Moving to +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Once the project has been established this issue should be moved across +michael gets anything he wants +Its standard behaviour to select first option +I still dont understand what exactly is the hurting part hereThe examples are not meant to be preloaded until they are actually executed correct +I think this can easily be removed +I think i was really concerned that there was a loophole that somehow released an invalid connection but i guess we just simply have to do a similar connection check we do in FTP and again +Misty handing over to you +Marking critical for based on parent feel free to unmark +Quickk fix exclude rich from testsuite for dropTarget is not solution just workaround +Reviewed all these issues and marking them as closed. +srcjavaorgapachecommonslangexception +For me the max hosts I have in my cluster are so using consistency levels is ok but you might want to change that if youre adding more than nodes into the list +Yes this is not supposed to fix the real issue yet +close issue prior to next release +Hi PavelHave you been able to reproduce this? I am wanting to use these servers for production traffic and am wondering if this issue is a general bug or due to a corruption in my clusterRegards I can reproduce this one that happens because when you keyspace it wouldnt change the UUID version from the original so when migration is send to the node originally it wouldnt correctly merge it into the system table +a will only recognise the values it had in the enum so no CLOSESUBSCRIPTIONIf CLOSESUBSCRIPTION is the last value of the enum then a server wont recognise +I remember a failed attempt at reformatting some years ago +Real exiting feature that deserves more promotionIs this possible in blueprintspring xml too where you can mix eg +Yes the issue is that there is a rebuild flag in the that is never set to true so it will never rebuild the proto +I applied this patch and the test fails with an NPE +This is because of the latest dojo upgrade +I thought you were talking +Thanks for the quick response +Add one there +still seems to be mapped to +More details next weekThanks for pointing out the problemHave a nice status of this issue sure I am not sure about this issue maybe still need to be opened +While youre at it can you add an implements Serializable as well? I dont know what it would be good for but maybe someone someday wants do deserialize params objects in anAgreed +This will allow us to safely refactor the code and significantly improve the tests +The original bug number being used to track this duplicate issue can be found in the Related Problem section of your bug reports Problem Detail viewShazron was the bug id for the one you filed? Theyre taking a look at it + +Hmm ok +The XML folder I copy into the release is generated by the create script +Please check that the bug is fixed for you +I was doing more testing with sftp and the session is down issue and I found a possible clue +Hi Donald initially this is only for the testsuite to try to flush out some of the problems that others are already experiencing having with SR +The funny thing is that I was not able to reach wave by using IP address in url as well +I have improved the wiki documentation for the all resolved tickets from or older +You are right +I committed this patch plus a test for it +thanks Mike for the evaluation +Im not too happy with breaking compatibility in that special case but I understand that the current behavior is somewhat unintuitive Thus there is an argument that it should be lets change this for final then +Also added durable and recoverable flags to the +Attached here as to trunk and branches. +As in the two instances are not the same +Check whether all screenshots are available +HTH +Will it be supported? I am getting the same problem on beta for my web literal web services are required for basic profile +I believe this can be closed the problem is due to a misconfiguration +Tried it with latest build with following commit id +Looks good to me +Hi Gregorythanks! Where can I get the patch from? You can either send a pull request or attach a patch here +Missed it the first time on the BPM side +The bin rpms are for customers running multiple versions of on a single box when migrating from one version of to the next +This negates the whole purpose of EXSL to provide a commonly agreed upon library of XSLT functions +Sorry this slipped through the cracks in November I wish it hadntI still think the oneway caching is pointless and obscurs the code even the simplest optimizers will safely remove the function call and caching it risks getting it wrong now or in the futureThe rest is a nice cleanup. +It will be nice the failover can work +Reminder. +wont fix. +Note that Derby will only detect a deadlock if the thread holding the exclusive lock is also waiting for a lock itselfOf course +To do this we would need to introduce the name into the Menu control to follow this convention +Final patch. +Attaching +Thanks to Hans for providing the patchThis has been applied as part of the commits on +How can I edit this issue so that the component is changelog +Improve pattern matching and make it faster by preprocessing rulesLet me some days to extract it from my codebase +I work with Guillaume with the same just noticed that the trigger activeruleparamchanges is used when disabling one rule of a the release creates correctly these triggers for Oracle? +ok it will be integrated this ll do some refactoring and corrections. +Closing this record since is now released. +So different languages can define different profiles with the same nameI thought we would remove this constraint but its not planned anymore +Its called a race which obviously many of us who have hit this bug have encountered +If you could write a for it it would certainly speed up the process +Sorry I missed those failures +Thanks +Work in progress patch which compiles most Canopy changes needed for Hadoop +Its necessary to restore article on JBDS on wikipedia +Not approved for EAP +h modifies the lines in question. + fixing IO retry logic to throw an IO exception ratherthan loop forever on IO read errors +However if we want to write an empty value we have to modify the null to an empty value because of the that happen if we dont change itFor our purposes we want to skip them if the values are null +Thank you very much! The fix works fine for meBest regardsMarkus +Its working perfectStill have the same problem if i put the document mode manually to IE but its not really important for nowNice work! +Perhaps just let the be handled by the handler? Actually I think it would be fine to handle the the same way as too since that would indicate that the slave is notI didnt quite understand the comment about a clear failure at the time of the error rather than a subsequent NPE +Is this a new feature or regression +Id love to do that but once the issue has been created I cannot set the Patch submitted to yes in Jira +With revision checked in a fix for this. +this issue is closed work to be done should go with. +I can also confirm that when upgraded to rails it now runs perfectly +Dave I backported patch to branch +Ive added the possibility to have options to the mvn command in +Verified closing +See we just need to have specs on both sides What terminators may a parser report to a sink and what line terminator should a sink issue to the output document? bq +Your patch is in CVS +Im not sure how the master came to be in that bad stateIf our cluster enters this state again I will restart with DEBUG logging enabled and try to catch it again. +Ugh +Fixed for transition resolvedfixed bugs to closedfixed +I have been unable to work on this issue and leaving it as assigned to me potentially prevents another developer from picking it up +Is this still active? The last entry seems to be on Sept th +Firstly pigExec was always null and never assigned a value so it resulted in NPE in certain code path +Will be in Atmosphere GA which is targetted to be released next week. +Is there a chance you could also try with a different JDK update version than update +This issue is broader than the Import XSD action in both the Status View actions +Johannes thanks for the patch +adds an that might be able to use that might help with this +Shahar I think Ive corrected the issues you reported +Duplicate of +stack +I can see the value in this though I worry that tuning may cause some people to submit tasks that trigger scalability issues someone else will have to comment on that +Ill test somewhere else until we solve that problem +Bulk closing all resolved issue that have not gotten any more feedback or comment in the last month +fixed in xCommitted revision +As soon as you indicated there was I tried it +Its hard for me to say how smart Groovy pasting needs to be for Java developers not to stumble over it +In my implementation of there isnt a need to return the watcher +I suspect that all of the SASL code could be pulled into a and and that would minimize the coupling +and can probably wait a little till some system testing is done +See for further details. +Patches are welcome +Adding on both sides of the relation does not help +Hi Sannesorry I wasnt clear in my description +Bringing into +wont current users s be parsed as one big line of sql instead of multiple lines with one sql statement +Resolving as fixed +r for acls +modified to use them +Thanks for sharing your findings from ALTER TABLE DROP COLUMN Bryan +Thanks though its mostly Jinghais work +embeddedejb +Closing. +Looks like Prasthants already got a patch on +Uploading new patch +that are all considered works in progressIts likely even if someone is starting from scratch theyll look over your approach and say Wow! I didnt think of that Ill have to do something to take care of that case +New patch +But I noticed a few small issues +Thanks! The job works fineStephan + for this idea +It appeared that due to some magic feature it made maven do not work even when it was set to OFF! After disabling it all problems disappeared +So you are saying that portal DB configuration doesnt handle properly user names that contain backslashes +I will resolve this issue as being superseded by +looks goodwhat would the jobconf be used for +I got another SEGV from as same as the first error dump on Ubuntu bit GA +Great idea +Is there a workaround +This latest seems to be working a bit better +Again your feedback is appreciated. +We really need to catch a I believe +Maybe freebuffer is getting called twice on the same structure changed patch to null out the buffer field +Is there something like a directory or file watcher inside the karaf universe which makes my approach obsolete +Build environment has been modified Maven support +Change should not have been backed out +mistake I misread the guide +Excellent! I will close this issue since you have verified that it works +Id be on supporting an call in the API +It works on unix because you can remove a directory if it is not empty +It supports metadata which we now have to do ourselvesSee +I think jira should be more about how to deal with exception in no? In we just issue a warn message upon an exception and perhaps this is ok as long as the socket is still open +the latter concern also seems to preserve old behavior +Could you pls run and post the results of test and here +attaching commit file removes the returns statement changes to commit file removes the returns statement changes to on. +ok go ahead I dont have time to look into this further right now +created for validation of ttl old patches and adding up to date backported version of what was included in trunk for +for the patch +Thanks for your help and again sorry for the confusion. +But if all you want to do is read the patch to get the gist of the changes I think this one is OKI also included for completeness the patch for the tests +LGTM +Ive emailed the grails mailing list with the stack trace hoping they might have some suggestions +I think it is best to leave the ESB Admin Guide as close as possible to the upstream +Will check why could not be merged +crc files are included in the source file list which causes the collision during rename +Patch for the should provide adequate infrastructure to makeinteractive authentication possible +Requirementschange to EAP +the name i listed was for the server +I thought it was a bug that we were the name incorrectly +Ill try reworking the patch to drop leadership +Backported fix to with revision For further discussion of the issues mentioned under this issue see. +eventually we could barf out a release with the fix +So our course of action would have to be to review each mapreduce operation in our pipeline and consider whether it might incur a high number of threadsConversely itd be much easier for us to consume a version of Hadoop where threads doesnt grow with write operations +I do not only want to store the changed values but the whole entityThis functionality must be available and also possible to achieve in a supported audit log +Never mind +to should not be a duplicate of as this one covers process +The test case +so im adding that +Remaining problems if any can be new jiras. +All test cases passed +I saw thanks very much for that +These are generally new code so they shouldnt have anyimpact on users and all tests should still pass. +This is the prototype of the end aligned forward reference page numbercitation bugFOP reserves some space for forward reference page number citations whichis filled on being resolved +Sorry Jochen but this only a idea +bulk close of resolved issues whose fix target is either unspecified or from a prior release. +OK that sounds very reasonable +We will also have to look at ant resource feature I think there may be interesting things to do with it +Ill have to think through this some more. +Ill check the problem and inform you how to solve it within today +Updated description there is a bug wrong property for version better description of the steps to use maven cargo needed described above +Adding a shutdown hook that flushes the appender. +With JEE compliant web service only stateless session beans or pojos are allowed +May be worth adding another option besteffortwithrepairLets save this for a followup after we see how well the blacklisting actually works in productionAnd since this is the main reason wed want to persist the blacklist Im okay with punting that down the road too. +Ben or Ramesh can you take a look to see if the revisions make sense +Thanks again +Verified that the unit tests work withCommitted to the as version rI have not done analysis on whether replacing CXF jars with Jersey jars has performancefootprint benefits +Micha Can you elaborate? Im not sure what this means +Hi againFinally weve decided to fix bug on the assumption that null values are allowed to pass throughTo solve all our issues with null values we also had to fix thats why weve attached a patch forBoth patches have been generated against the latest sources in the branchWe have thoroughly tested this solution with our use cases and it works fine so farI hope it also works for othersBTW Ill post another comment on to link hereCheers David +My vote would go to newlogo +Ill go with that in the short term and we should definitely contemplate the impact of interface typedefinition on service name going forward +Hopefully you were happy with the way refactored code interacted with each other +Better to across all subclassesI dont want to add a statehowever as that just complicates things that last comment pasted into wrong issue +No error but it asks me for user and works fine +cool! thanks for the effort Yonik +Note also that this exception does not happen with the Seam community release only with the Seam release in the EAP download +Attaching new patch fixing the release audit warning +Committed revision Thanks Akshay +Geir + +Michael Thank for taking this to your hands Your absolutely correct the parameter name on slide should be realmBest regards Michael +We have a workaround for whereby we use the newer resolver for the main work and for the forrest task we use the older resolverSomeone on Windows needs to investigate and develop a test case and report the issue to Ant +Based on code and comments input I think the requirements are There should be a class created with methods with parameters type or +Straight forward patchAlso modified so that setuptearDown is called only once +So there are with the same component name +Took a whack at returning from +Threads from camel type converter deadlock. +I really hope to get a working version for C +I was thinking of including it in if we can converge +AwesomeI think then we shuld focus on making the cut to DV thenFirst though we should have a migration plan +Run ant test with Doverwritetrue +The is to not use but to specify the server typeIm not sure its a good idea to default to any particular OS if the command fails +Any way for trunk we haveTedWhat is your opinion on including this in next RC +attaching patch to fix it for maven you need to reopen the issue to bring the attention if you just comment it can be missed +I just tried it and everything seems to be working so Im going to rename the file back to suspect its an Eclipse issue rather than a Spring Rich problem. +OK then +We simply tack on the model name to the cached file +New patch with the following changesFixing Amars micro review comment corrected a couple of Java docs and modified the taskTrackers to use a instead of for efficiency +Note that adding in some dependencies works around this problem +is only from the remote repository when a +Kathey was so nice to point out I did not do diff from the top +You are a good man +Thanks again addition to this issue Ive also noticed that some exceptions are thrown when using the scriptaculous tag +What I have works for what it does which is let you specify if you should dig down deeper into an object to look for more Field annotationsNot really sure how we would flatten or relationships +Jared this may have been dealt with in the Installation Guide rewrite +This log shows that the RS we kill was archiving logs when the master started splitting them +MacroI have reviewed your patch although I didnt test itI really liked your efforts to migrate old contentNow I am going to test your patch on my development boxThanks for such a nice contributionRegardsAshish Vijaywargiya +deployed reaches stateA failure to start the bundle does not mean that the bundle failed to get installed +Im pretty familiar with +Correction of sorry to do this but I just found out that will remain in FLEXHOMEideflashbuilder rather than FLEXHOME +I like this proposal +This is a very simple XML document that shows the bug +Need to use JBDS or patch the installed. +Verified in Section of Installation guide. +Fix appears in trunk docs on Derby site. +This patch removes Changes merging it into Changes +atl +Hi back for review +They just disconnect suddenly +Is there a specific version that every developer should be using +Taking it out of although the last fix did go into it +Every committer already has access please check your svn command andor try again +BrianI have added the snapshots deploy to Hudson over the weekend and that seems to be working OK +thanks for resolving this you for catching that something was amiss in the first place Thats the hardest part. +This issue is made obsolete with the resolution of +Thanks Stuart +When i commented about this in i did not mention the usecase where we are trying to add such hooksHere we are ideally not doing the forwarding that is done by the Scans whereas it is something we leave it to the CP to decide on what to reseek rather than we reseeking on the kernel like in +What happens is is Storing all the Factories with respect to the Class Loader +We wont do this directly just now but will likely include it in future rework of the management layer so resolving this JIRA. +Checked the three methods bellow +We havent had other users reporting issues like this for a while +This seems already being fixed +Just another short comment as I have amended the JIRA notification scheme in use for Castor +The file though is an awful monstrosity of lines +Ive updated the fix version to updates on these Glassfish jars and should no longer be pulled in +Bertrand Ill leave for this Gavin to comment on if thats ok +Please note this issue blocks +the old patch I did the null check but forgot to break +bq +Time to add some upgrade index testsIn fact Lucene already has +Thanks Oliver the patch is fine +All the version of hbase weve relased up to now had this property +This issue will require a wider and breakage of backwards compatibility +Moving to as we still havent finalised a solution here. +Feel free to commit this to for the comments Rob +In many local discussions I didnt hear one final conclusion regarding this +I dont know the process for submitting code changes but if youd like me to make the actual modifications I can do so +I am very busy at the moment +initial commit done now adding to go Brett +Added two more examples and +We could create an alternative more codec provider impl as you describe but still keep the old one +if you want to have an unauthenticated user then dont include +Does anyone know how to workaround this issue +Actually delete and readd didnt work when I click in build now goes back to in queue and nothing the server fixed the experienced it again after a failed build due to missing dependency from a clean continuum with only this casei add a project click on build now it buildsafter finishing and fail I go back and click build now againthe project is In queue but no activity in continuum logif I restart continuum the project is with no state and I can build it again but if I try a second time same problem reproduce with snapshot +the properties and +Resolved +Latest upgrade went smooth +Revision from March +Thanks Devaraj for the early patch +Bloom filters on a specific node would go down a great deal after a major compaction only to increase back to near their original level over a few daysWhat I believe is happening is we had a staggered repair schedule along with a staggered major compaction schedule +What system properties should I set to override the Sun implementation with the Axis one +OK this was a little harder to fix than I first thought +Anyway I agree that if we need to support building geronimo with Harmony we cant apply the patch of this JIRA for nowis it really neccessary to add additional build complexityAs Kan said Its really necessary for in translation contributors +Whats about attached version? I also figured that batchCallback had the same issue +Id suggest you have a look at there are examples online that demonstrate reasonable usage +Alternative to exception approach would be to use method + +Thanks for taking your time. + +We cannot reproduce what we did for Beta otherwise +It requires additional external dependencies and therefore is not included in Cayenne core +No additional tests because theres no real functional change +It will force anyone with custom code to recompile their code if they upgrade the version of Flume in their build environment +The plugins may not exist +Lets see if the flag the patch has any regressions +No worries it wasnt critical but would make it easier for people to migrate from One of the few things I found that was an issue along with a few more zookeeper exceptions being pushed though +Hi BryanIve seen the same thing post rewrite but for the record would you mind tossing up your benchmark and the times youre seeing pre and post patch +Closed as per state in Bugzilla. +Sorry but it does happen to me at least once a day +was not being properly escaped so that SQL could be used in the password in such a way to gain access as a user without knowing the users password +Yet there is a reference to a file +Fixed nowThanks for pointing it out +Yes this issue can be closed along with the other issues you listed +So it is a problem with wrapper? I indeed prefer to use the same connection across multiple DMLCs +Place image file in modulesrestservicesrcmainwebappimages fixed Thanks amila for the contributionRegardsLahiru +This bug was indeed addressed though not by but rather by +I dont think our admins would ever change the target of that symlink while applications are running +yeah this was a big mistake on my side +Ted Yu I ran no failures +No its the same with demo data +added conf +What is the use case for putting text in front of a figure block? Isnt a figure caption enoughThe can and should be solved by using css +the same is being repeated in srcmainbinhdfs and binmapred scripts +Hmmm the is in the wrong directory orgapachewebbeanstestannotationbindingbut the package is The package and directory should be the sameChanging the package makes it not longer compilable thats the reason why Ive put it in the other package +notification works fine and i successfully checked in some code btw +The verification of NNs block report works now +Attached a V patch incorporating the above comments +I am using Sun JDK on Windows XP +Ill go ahead and set it. +This change is also transparent to existing implementations as the culture settings will continue to use the default system settings if no configuration or programmatic changes are applied +Instead of having the CF available under another JNDI name it creates a wholly different CF with its own pool threads etc +Just adds an inventory service this module uses +When not running inside theres a host of javax +Not an issue +I think we need to document that Derby doesnt support indexing group by order by on long varchar for bit data anymore in release notes in case some one is already using this incorrect behaviorI think it would be good to port this fix to so we can prevent more incorrect dependence on this behavior +eclipseplugins +Adding Justin to the mix he was the last to do heavy modifications to good to me +Bulk assign all unscheduled issues with priority major or less to target +This issue is caused by fix +we can report a bug against eclipse specific types sounds correct to me +Updated Branch to and that allowed the of the solution as mentioned aboveThe fix should also indicate a potential release but this has be removed from JIRA. +Ok is done +In addition the link citesImplementors should be aware that this specification is not stable +a few minor tweaks belongs in the thrift patch seems to have gone missinglogindone needs to be threadlocalIMO addressing these in your pachset with the working auth backend is fine +BTW I have a patch too to do the remapping of the columns +correctwe set it on creation of cases +Nicholas may be we just suppress warning? Any better idea for resolve this +I just committed this +I have just been spending some time on the website and came across the following +Sandro maybe you can take a look at these +Ill commit shortly. +Committed to trunk +I will add only a null check for the session +Why start mucking around and destabilize it againbq +That probably explains a lot of unexpected whitespace Im getting in some of my other apps +Closing old resolved issues +The fix is good to get us moving forward. +After reading the trailer last Should the two numbers be the same +The project has been tested on GCC v and MSVC compilers +Patches added did not really target trunk + reported the issue and posted the patch +Connectionconnect Connectionfastconnect and Connectionbindconnect were replaced by two new methodsopen Create the socket and set all options that are set before connectionconnect Connect the socket to the target +Come back to us you were doing cool stuff +New jenkins run is clean +Small edits to Simons CHANGES entry +The following patch fixes the problem for both double and float data types +The schema is now generated as part of build process and has the new option +Jon addressed your feedback +couldnt we run the storemats suite with the system configured to use this store module +Yes you are right +Patch is now in CVS +Thanks NicolasYour last patch is in trunk at r +If re establish connection was not better then make it recover then node will not been marked down by FD and then immediately upI do not want faster connection reestablishment i do not want node to be marked down for fraction of secondIts over WAN network yes +if yes WIIN! if not looose +Will wait for hudson before checkin it in +I thought you were just talking about the sample JSONYoure probably aware that theres a link to the parent data type in the documentation right? Youd like to see that documentation duplicated in the documentation for the subclassing data type? +I think that it shouldnt be difficult +My REST service extension also stopped working with the same reason where i am forced to add mail jar and activation jar to fix the problem +Id be completely happy if the preliminary code made it into trunk with the expectation that I will continue to submit further patches to improve CircusThanks +There will be a few values where this causes us to truncate a trifle early and lose a trailing bit or two as in my previous solution +Added thanks Tom. +So it is a very easy fix I suspect just change the component type when calling createComponentregards if it works with the RI feel free to change it +Can you attach your schema file +Thanks Brock +Noted as incompatible in +In fact it is possible to move the code into a separate independent filter +Good to see this +Resolved in svn rThe license text was moved to start at line Also corrected the list of generated sources referenced by the projects and added +adds the missing reference acquisitions for new index creation and load +Need to reopen this issue because it was closed with affects version of wishes which makes no sense +Thanks for the pointer to +Committed revision +The reason for this is that there is no opportunity for theframework to analyze the request before the JSP page receives it and store thecorrect into a request attributeA redirect causes your browser to issue a separate request to the destinationpage and therefore violates the above principle +We change it to to do the select Java raises an error for select against blocking I suppose the proper fix here is to just return nil if receive returns null +Jonathon since this is a concern +It works. +I updated the patch +Thanks for fixing the secondary namenode issue +on v +Thus regions state in AM is still PENDINGOPENT +I believe is one of those modules that is really in no mans land +However I am just worried about losing this advantage users can use all methods of while currently only offers a subset We could modify to provide all the methods but does it worth? Assuming all users are going to use eventually we might be just OK to remove +set scanner caching to and then deal with mb rowsThen the scanner caching was set too high +There is certainly the same issue with the new code colorizer +Over in the map of attributes was added to the available in thrift +Each server listens on various ZK paths registers the same listeners and follows the same code path to apply state changes to itself +Consulted with Remy +Looks good thanks Andrzej +Can you provide a patch with what you suggest to be changed? Or attach a file with the updated as currently marked as resolved. +This fix has also been posted to the juddi project +incorporated Zhengs commentsHadoop does not like file in the aux lib latest patch is an svn stat output instead of svn diff output +Why not use an action redirect +wondering if something changed in recent Tomcat versions that prevents us from unwrapping the connection? Did you try with slightly older Tomcat versions? Im positive that with Tomcat it was working fine +Im still not clear how we were able to start a JSR block in the belief that local variables in the jsr block were NOTSET could this bug repeat for blocks? Well done to Daniel for catching this hard to find bug +An implicit recursion is a bad thing I think +Solr doesnt need to knowTrue but what store would you propose using in unit tests? I suppose you could develop a Mock RDBMS which could be even simpler than Solr so unit tests dont need a solr running +Our projects are hosted on sourceforge +I like it a bit more because you keep track of the merge by going through +appears to be a regression of on Windows +I am not getting any cassandra crashes or exceptions any more. +This way the user will not need to touch the filesystem of directly +The difference being that the latter will explicitly retrieve all matching resources from all class path roots even if the relative class path names are identical +I tried to search for existing bugs but missed that one. +But now if selection is not on css class or css field is empty and we call CSS dialog we have no any messages at all +as name seems less than congruous perhaps would be betterOpinions? +To build it locally its to much work for me right now I do not have time for I suppose it will be still broken because your PR didnt fix anything with layout just displayed text +Still Work in Progress news later on +I just thought that would be the right place for them +Thanks Jaimin. +Patch applied with some minor modificationsThanks Stefan +The older version does not have the attribute +Should have realized that when minimizing and maximizing a root refreshes it +Otherwise runs the risk of being labelled immature and people will turn their backs on it +using lang yes but should we avoid an available convenience in cases like this +Youre right I thought I had to change because of well faulty reasoning at an earlier stage and did not backtrack all my changesAttaching a new patch that only changesIf you see this failure with classes too and the change to is reasonable I will commit this fix +what exactly is the usecase for overriding this +bq +The existing code reads to bytes more depending on startOffset + why does text overlap in certain places +Didnt fix anything though surprisingly +Committed revision +David is looking for a test project however Im confident the shading changes in more recent versions of Maven would have already addressed it +The results can be found in but in summary Memcached suffers same issue as Hot Rod and so requires each operation to send lifespanmaxIdle +Start Geronimo server +Good fix +has been released closing issues. +i dont know much about the standards +If all the other bundles are present and resolved but is not finding the classes there may some additional logic required in the Tuscany activator when running the bundles under Eclipse +OK thanks + +Thanks Eli! +There are a few ways to do it +We need tracking of performance results over time but is not going to be the way it is done +For example requesting a project in browse where the parent POM is not already in the repository should resolve it correctly +Error says that The type cannot be resolved +oh I found the unit test in the patch +I also added some clarifying notes to the deprecation note inI will post a separate patch for trunk where this setting will be removed altogether +Lets figure out why Im not OOMEing and you are Holger +starts for me now +Itshould be randomly generated on each request +rv I prefer the font from rv for this onervb I think this is my favorite however I think it might look better if cluster were given emphasis instead of modrv This looks nice big but kinda crumby when smallrvb I prefer rv over this variantrv This is ok but I think I the roundel looks better separate rather than interacting with the textrv I like the concept of the logo style matching the font however Im not a fan of this font +or if I should close it as Incomplete since there is no way to reproduce the original issueAnyway +I fix the definition of the annotation to only apply to types and put it in the right place +This issue is fixed and released as part of release +I get some sort of configuration exception when I tried to run them +Thanks for the help Ive set this to resolved but will open it if anyone has any further comments +Included in release. +srcmainjavaorgapachehadoopyarnapirecords +Currently is excluded +Im still not sure I totally understand what happened? The realm shouldnt be required unless you specified a realm in your config +thanks! +I can use download for now +This one should go to the EG meeting for review +Were you still wanting this feature? I personally have some concerns about it but havent had a chance to look at the patch or comment +the issue is not clear to me +Duplicated +This patch updates to throw an and updates to give more helpful and specific error messages for the following error conditions Multiple matches found no mapping in geronimo deployment plan Multiple matches found ambiguous mapping in geronimo deployment plan No match found no mapping in geronimo deployment plan No match found incorrect mapping in geronimo deployment thanks for the patch +On the negative side malformed multipart entities mess the rest of the message up completely +I didnt know that modules can be versioned +AgreeIll change code cause warnings like findbugs and license and upload a new patch later +The issue can be still have tested it in the current trunk that includes revision +Dont unterstand why no one of you took it on +Applied to trunk and branch! I bet theres more like this lurking around. +Its just that the ones I am aware of here are actually smart enough to not require the not null since it is redundant to require it + +Does anybody a complete example application of integration between JSFJSF into a PortletBecause I try to integrate this framework in a portlet application but I dont achieved my try to use the but I dont obtain any result +Hi MichaelRegarding I took the parameter modes IN OUT and INOUT from CSDL specification assuming they can be used for transforming stored procedure definitions into +I have been so focused on finding the cause of my problem that I forgot to think about any possible work around before i opened this JIRA +Resolved this issue as has been resolved +Even if we dont throw back to the client it is still better if they are handled in some central place instead of all over in low level IMO +Oh now thats interesting both views and are not being releasedCan you share the script +period for template +Marking Invalid as it appears to be an APR issue +Removed two unused imports that were recently introduced in +Could I ask You to check whether the result returned by to returns in that case ? If so I could eliminate that class as the source of error and look somewhere else +We need to worry about that as basic use case will failThere are at least two ways to solve the problem do not include those http headers in the camel exchange do not use these headers when building the http request Im thinking we need to implement the second one as if we dont requests will be unusableBut the first option may be interesting to save bandwith and could be controlled by a flag in the uri maybeAlso I suppose the same patch should be applied to the component which uses asynchronous nio +Of course the leak would show less effect then +partner web service I want to invoke A copy of hello only Ive added everything necessary to invoke helloService +Could this be streamlined somehow? Perhaps we could use Subversion commit hooks? +Thanks +Or at least still contains the bug anyway +Tests pass locally +Issue which added aggregator like this issue was fixed with version +just thinking that all these tests up on are passing because its we and ec is quiet +The assumption in the code is unprotectedXXX is called for both regular operation and operation while replaying edits +Im not committing this bec I dont know if the of the apache plugins should be setup as its parent tooThis needs some work also as its a bit outdated against the plugin documentation standards +for the patch +Are there other benefits? +That helps in creating a single raid out of them +Ok Ill do that ASAP +Closing the issue now that it has been resolved. +I regenerated the patch +Applied on master and +Applied the patch into trunk +Thanks Namit! +Shouldnt be documented in +But user often just know about the boundaries between the application he is developing and the outside of that application +add brocknoland as a contributor for Sqoop project +This ticket is obsolete since have instead implemented JDO objectstore +I did not fix exactly like you suggested instead I did remove the surplus methods before adding them to the +It relies on simple work day schema that is identified as start and end hour and by that applies to all working days same way +and added a test case +Appreciate your helpthanks you show exact way you are trying to connect againDo you have same problem if you dont encrypt? am wondering if maybe you are mismatching encryption somehow +Carl I just tried both compatibility cases +and I think should be fixed by this +Is very large because code is not formatted with udig formatter template +marathon could be named something different +I went ahead and implemented this since so many people have requested it +Should be fine after tests component is built +John to discuss with Kurt +Sure this is not intended as a committable patch +root root May ls ld r +assigned to release this change may break existing scripts so we want to get it done early +As a workaround I have found that deleting the Lucene index and having it rebuild the on the next start fixes these stuck waves for meThough some seem to end up doing this again after a few weeksAs for why I have no idea +Will look at this for RC +Claus I havent forgotten about this +the bug is also in version build number Any news about the resolution +If we are saving the first key of each block we should probably also save the last key of the file tooI would think this is just a new class that people migrate too +has been replaced with in the log method +let us try to fix for to +I think the patch should be reverted +The patch applied fine. +Maybe someone else out there can test and resolve +Sachin can you check if reordering the moduleFile and planFile arguments in the rd method has any impact on the DEVTOOLS? If you can identify any calls to this rd deploy method we can change those accordingly and conclude this JIRA +But the corresponding code is not there in tomcat I dont see that option in comment in the sysconfig file for tomcat +Hi +resolved and verified in IR removed from blocker list +I dont need jenkins to have push access but I dont see how that matters in this case +Doesnt fix the warnings which need to be fixed +Apologies +Thanks for the jstack traces +same issue w webflow attached is a patch which simply implements Serializableappears to work fine for me now +Thanks Christophe! +If anyone else does they should speak up too +Ive never seen this happen +Committed thx Dew +Reopening for handling in +If we ignore guards then this problem disappears +on both yaml and Chriss was wrong does support java enums +John assigning to us since Ryan is out +Thank you Plamen +You may also want to consider the case where the do not fit in memory +How does this resolve in felixequinox +The hostname option was added to the shutdown command a while ago +id +fix committed in r +Hit by a bug in +Checked this out and I can fairly easily make the patch with a caveat +Fix expected seen vs +mention the current memory limit beyond which we dump the data +For example and +The resolution here was that the dev ought to implement a custom OC for such needs +name would be javacompenvejb and jndi name would be in my case ejbsomeproject +Not sure what you meant by thisI meant that this will play nicely with which already expect a and offsetlengthIt is clear that there is a large room for improvement on the merge code +By the way introducing as we did when coming up is reasonable since the scope is just limited to test +Realized that I posted the incorrect html fileI am posting the correct file now +The only problem with the GET is that this bean is stateless atm +bq +If you think they are critical for CR please them +Why do you mention an incompatibility with if the errorcode is coming from Search HSEARCHAlso why are you using a custom producerAssuming you use a custom producer why is sending the value default not validIve read the reference you sent thank you but I cant see what is invalid compared to what we are doing +Fix version updated Ive seen that several time +s config will be set autolly +The file in question has been added days later in commit ecbaedecafcbee. +Fixed at revision Please check and close if youre satisfied +The new patch picks up hadoop from the path +Updated key documentation. +on the JIRA +Setting resolution +Suggestions for less creep welcomeThats more on the shoulders of those that want to get it in its not currently a feature Im after just one Im very concered about going in and changing the game on Solr security support without support of the developer communitybq +We have tried both old style hibernate annotation and the new JPA orphanRemoval attribute but both are just as broken in this regard +merged to fuse trunk in rev to fuse branch in rev . +Let me know if you want me to submit a review request +So this has to go into. +Added note about the test improvements being independent of the refactorings +slipping +I can update +Thats the way works +Patch to resolve test failures when building with ibm jdk on aix for customer certification +Proposal fix +Please your applicationwith the most recent CVS snapshot +Committed with revision +This is a real problem sometimes +In terms of verifying the fix Ive looked at the new revision and it looks OK but unfortunately I dont have a way of verifying the solution in practice for the following reasonsSo far we have not actually managed to reproduce this problem in our own office on our development machines which are not the same as the customers set upIn the deployment at our customer site Xerces is being used which we incorporate as the pulled down via maven dependencies +The patch file reflecting whole changes on content hub component +At least the compiler isnt complaining +You should add capacity or throttle your client +isnt it confusing? Need to queury the core developers for not hesitate to change the name +Seems to duplicate which is already fixed in and branches +people may have written things above CFOF so it might be nice to keep it around but deprecate itare there any benchmarks or is there anything anecdotal about performance +If you want to label every interface then we need a way to enforce that for consistencyBut which every way we go only interfaces will appear in the I dont believe there was any disagreement on this at all in this or the other jiras +Yes sirSent from my iPhone +Weird in the I hope this doesnt get a lot more complicated and cause us grief in the future kind of way +Can Apache projects rely on components under prohibited licensesApache products cannot distribute any such componentsAs with the previous question on platforms the component can be relied on if the components licence terms do not affect the Apache products licensingFor example using a GPLed tool during the build is OKQ +In my case I selectedCProgram Files solved my problem setting GRAILSHOME to CProgram FilesgrailsRegards +HiJust one question the activity name is available in so is it possible to just add it in the interfaceThat is all my needToday I have done a derivative and it directly use to populate activity name and it works without problem +fixed withWhat is the attachment aboutThis issue is closed +These changes look good to me +This is a different issue than. +What about a special menubar bean that acts as a beanpostprocessor to search for menubar entries? We then would need only a marker interface like for the commands that should be plugged into the menubar +Thanks for trying it out Stack. +Replaced by standard Servlet Filter and added simplifying filtering a bit in Rev +This is about he and Committed +XML configuration for simple webbeans is finished nearly +committed in r +You will see that count goes to zero only after a block report from the datanode +The change has been m leaving this issue open until all the testing has been doneand issues found resolved +The numbers just doubled confirming my suspicion that instanceof is a heavy weight +Yes the type converter will return null if there is no right convert method to be called +We can add additional information to the file when the build is tagged if necessary. +Sorry I came off as harsh +This is a regression from and thus needs to be fixed before can be released +javadocs patch looks good all the Lucene warnings are fixed +Also there might be no script file +Also via we get Pig for free +transport +I do not try +Fixed in nightly build Thanks for the patch Brian! +schemehostportb +Edited according to please review +Then this part was fixed on trunk and needs to be backported to +Removed in metamodel and all tests pass +The document should probably call out specifically the case of generic names saying they are OK +IT added +Fixed in commit +Hi AlanNow I have the fixes for expenses problem but I found an error in I finish to solve this last one Ill prepare a new pullChema +Then youd have all of the existing Avro tooling for doing things like structured grep etc plus the advantage that you could easily write mapreducehive queries over the converted logs +Linux specific so scope is limited +Well Problem is wicket do not work in mode because wicket namespace is removed +About Rest integration tests are failing after rerun +Just Committed to Trunk +Attaching patch for allowing eclipse to process the dependencies properly +But on I havent the problem +bulk defer of feature items from to +Full master log should be available soon +This will be resolved by updating SWR to issue is fixed by updating Arquillian Core to which brings SWR in + unscheduled issues to the next release +is not to say that there are no holes. +Infinispan is version +Beautiful +hmm so far I havent tried that ispn inside of AS Im on it +Devdeep Sanjay what is the latest on this? Please update the the changes +So I introduced a tiny object and changed the Content construtor to use the object instead of the object +Fixed in subversion repository as of rThanks for the report +ill check the guessing this can be WONTFIXd? +Trivial patch +So yes I like your precondition idea but theres only one and its invariant do we at least have? If not return a +Committed at subversion revision +See for further details. +Attached error log +With this job profile fetch failure processing appears to be well over x faster +can you create me a login? Ive tried running it on my machine and just get OOM Andy +Please reopen or create a new JIRA if you get a chance to reproduce this against a latest released version +MladenCan you please take a look at this one as wellThanks dont have NEWS file anywhere. + +It already exists +Please review and commit if ok +This should be fixed as Dave put LIKE in to JDBC native code so it shouldnt do post processing of the like filter which I believe was the problem here. +the screen seems to be working fine however the description is not saved + +This option will give us a handle to list the unhealthy nodes and also be backward compatible. + will track verification if the native gfs fix addresses the issue seen with CR +Can we close this jira as a duplicate +Updated patch with +I actually implemented +But should i only change this variable? For maximum flexibility i would recommend to change all the variablesShort description of my issueA feew times per day i get a timeout +Jan can you provide some more details here? Perhaps a mini code example or better steps to reproduce +Tim thanks! +This is the script I used to start up on Windows +here the system property keys public static final String PROPERTIESFILE public static final String PROPSCHEDINSTANCENAME public static final String PROPSCHEDINSTANCEID public static final String PROPSCHEDINSTANCEIDGENERATORPREFIX public static final String PROPSCHEDINSTANCEIDGENERATORCLASS PROPSCHEDINSTANCEIDGENERATORPREFIX +Verified +John this looks to be a problem in +But the server in Ubuntu can monitor the server in windowsThe server in windows also can the server in windows +Am not able to see the logs for this test failure +Hibernate bugs go under MOJO HIBERNATE. +OK found a workaround Adding a encodeAsHTML to every gmessage tag fixes the problem decide yourself if this behaviour wished or unwished +Hi JacquesLatest patch for this issue contains lots of file with duplicate entry some file are repeating times and some files are repeating times I have reviewed a file and found no difference but couldnt look to all file so is these file added to verify with Jinghai purposely or its added mistakenly +I would prefer to see this implemented by using a generic that would be used to aggregate other appenders and deliver events to them asynchronouslyThis would be generic and could be used in conjunction with any appender +Jims modified files require either the amqjqueryadapter or the amqprototypeadapter which in turn require jquery or prototype respectively +Like Simone Ive found it easy to use with Guice +David assigning to you to review the patch i can commit if you think it is ok +Revision +This fixes the issue raised in +Committed to all branches +Although archetypes are not really required to build we probably want to release those as well to keep the version number in sync with the componentsTo keep youll posted I am running the components release and should be completed soon +If someone doesnt know some complex math problem but sees that somewhere in it is the equation then it is wrongI do not pretend to know everything which is exactly why I asked the quesion in the first placeIf the people you are refering to as being insulted includes yourself then I am sorry +Correct a typo in the code +Vincent I get this too but its a bug in the MPIR plugin so should be filed there +Im about to rework this part of the code for +Thanks Ben I can now see that I am a member of the I still cannot administer that project in JIRA +because of the plus some version changes +But it looks like you are using it from an EJB can you confirm? In this case you may hit a known issue that we are currently fixing regarding the class loader configuration +Graeme cant reproduce the problem with normal second reload sometimes works +Yeah I see what you are saying about that case but as the second matching message would begin with the terminator it would be filtered out +I can understand the concern for splitting the interfaces +mmm jason don give me words I didnt say +Also +Your solution is good thanksWe could probably filter the pom and include the upstream versions if we wanted to. +Moving to for now can always pull back +Could you please try it as lib +Needs more thought and a decision on scoping as part of +Assigning to you Norman as you know this code best +instantiated arent serialized and passed to the run time so theres no advantage to stashing info in the returned object +I tested it manually by setting the NN HTTP addresses to be wildcard in a secure HA setup and confirming that checkpoints still proceed as expectedThis patch also takes the liberty of fixing two typos I noticed in javadocs in the course of researching this bug +Could you pls provide few steps for verification +It may have to wait past a few alpha releases +Also used to be in SDK is gone as well so we need to change cloudstack code to be able to work with SDK and also construct our own apputils wrapper +From what I found was that the throughput when writing to disk was as good as the throughput with no persistence so I didnt bother getting this in +Id like to add default values for the new config knobs so that if the user already has webhdfs configured they dont need to do anything extra +Kristoff it is fair enough to say that WTP deployment is slow +Wouldnt S receive a FOLLOWING notification from S not a LOOKING notification if S ended up following S initially? Isnt FOLLOWING the last notification that it sends? In which caseT S starts FLE from e xid sees a FOLLOWING notification has no quorum and goes to round +This is a proposed fix for that will read a plugins property as a String and parse it after release +Hi Matej andI can confirm that this appears to work correctly inCharlie. +Alternatively a service assembly could work too. +Ive just gotten a chance to look at this +the performance is still quit good for the interactive client +At least thats the way that I read Myrnas last comment in +If thats the case we could remove the call to flushBuffer and the response would not be committed after rendering fragments unless the response buffer has overflown +Alexey try and come up with a suggestion for this and we can look at the UI not sure the current method scales well +I dont think this is the way to go about fixing this bug +Youre seriously making me regret using APR The pools thing is extremely annoyingI dont have time today to look into this to much +and +push out to Max I dont have objections to moving forward the module +Ill map the users based on email address +You need to bring me a long +What do you need me to do for the Porting? or shall I close this jira +Attaching +Sadly this approach also breaks others queries they all seem to be characterized bya mix of inner and outer joins + +I suggest that we should open another JIRA for accepting DSA key pairs if needed +Synchronization was removed from the perpareStatement methods per last comment in r +This runs in local mode and fails as reported above +I know that it is currently controlled by percent and it can be configured per queue +It cant conditionally include content no +I am debugging both of them on a KVM host now +TYPESYSTEM loads a type system descriptor with the dkpro types which is located in the descriptor folder +Changing removing the discovery mechanism willhave no effect if you dont use clusteringIf you need clustering you could use a different discovery mechanismbut this has no impact on servicemix itself. +not bug in HBX but in HHH +dat files with unix style newlines +Hi Robbie would you mind reviewing please +Should we rename to a more generic name and purpose like +Did not happen in the need to do this binary compat was restored. +Its a normal behavior and the same in KarafThe PID are different because its two different configurationIf you take a look in Karaf you have the same using configlistkarafroot configlistgrep i fileinstallPid Pid and you find exactly the same in cellarkarafroot cluster defaultgrep i fileinstallSo its the same outputHowever it makes sense that cluster output looks like configlist one +It would be nice to add a link to that project in the documentation so people find it +Downgrading the urgency of this issue +Since both CVS and SVN commits are arriving on the new list I am closing this item. +Gail Of course +Thank you every one for testing +Ill have a look if there is an appropriate ticket already or open a new one +src xc dynamically checked calls or field accesses run with VERBOSECALLS for more detailscpsiriusxamplemmXbin mpiexec np +Reopen these issues or open new issues in then release process. +Actually its not a failure as the BUILD SUCCESSFUL message appears but it seems like the package does not exist should fail the build as well +Ignorefilter newlines in baseline comparisons to make it easier to develop across platforms with different OS newlines without affecting saved golden output +If there is a need for a defined list of domain names I would suggest that the library supported both and open +Is that in WF now? +You can always use the hsha server if thats something you want to avoid. +Jean ANDRE has this issue been resolved with +We have a task to take care of this at some point certainly within ESB +Could HOD take out the branches that were added to support? Since after the current patch the configuration will work again +For this the jar has to be created before those tests are run +Yes it was Joe +The mutually exclusive change still needs to be investigated. +The saved document is missing the document as expected +think Dave needs to step in and sort this whole thing out +so its kinda like a fileset +The issue can be closed. +committed to trunk in rev +The attached patch seems to work but could use a bit of I think +now looks like a race in the tablename caching +Abhinav please verify and move this to resolved if you dont see the issue +I am going to mark this as resolved since it was an issue and others may come across it +The secondtry also seems to be progressing so this patch seems to be fixing the hang +I have checked the patch but unfortunately it contains an serious bug +If we are going to go for the suggested solution of changing API between AM and RM perhaps it would be better to open a new jira with a more descriptive title +I should add that this problem affects as well +Hi ShivaThis is nothing complicated by any stretch and me just nitpicking to make the little things a little betterCorrect me if im wrong my apologies as im still learning about dojo but on pluginsplancreatorsrcmainwebappviewconfigcreatorThere is a list of of widgets +wicket in r at least for filter usage. +Looks like its not good enough yet +hmm I used to get that particular warning +I have this assigned to me but I have no time this week to work on this feel free to assign this to yourself and submit a patch if you have a fix. +This is actually on purpose as initially designedThe idea is that if you put a message explicitly thats because thats what you want to get instead of the power assertIf you want the power assert dont use the message +Clarified the issue with the team now initiates when the container has successfully started. +Here is the patch +Hi TyronneCan you provide the relevant information? need to have following libraries in your classpath when sending a message to the provider from a remote clientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEserverYOURSERVERPROFILEdeploy as Im now on JBDS +Im logging their being added to the queue and am then dumping the queut but the flush thread is not waking up to clear the queue under load +the same affects windows users that do not have administrative access which is the case for CI +The patch cannot be backported cleanly to and I suggest not to hand merge to get the fix into the branchIf anyone disagrees I think a manual merge is relatively easy +Please let me know if the upgrade is not included +This causes the configuration to be provided to Jetty asynchronously which in turn causes Jetty to stop and be restarted with the new configurationQuasi at the same time your tries to connect which of course has a high potential for failureThe second exception I cannot readily explainYet I could not reproduce your issue with a current SNAPSHOT reopen if this is still an issue with the latest snapshot of. +Unit tests really help a lot +r also +Thanks +Attachment has been added with description A unified diff against HEAD +Any other credit card number from a different bank is considered to be invalidAdditionally a unit test can be found as an attachement +Moving out of +I guess it should be fixed in Shindig +If everything worked ok for you then I may have missed something +Thanks for the patch Hemanth +Thanks Sanne +I guess the only differences is that fallback happens in the anonymous type case as well? However i am a bit confused how the simple content case ever engages +I am closing the other issues since basically it is a duplicate of this +Even now CS and fifoScheduler will have the same semantics for this config +Should be renamed to +Shelly not really +Why not just load FILENAME or run FILENAME +Ive refactored and into various template methods for easier overriding +Tika can integrate with version of the library +instead of +Also what browser is being used when this problem occurs? Have you tried another browser? If so same result +I like this better but Im still a bit hesitant on a couple issuesFirst why do we need a second path to put plugins in? Whats the benefit of separating them out of PREFIXlibcouchdberlanglibIm not super thrilled about the manipulations for handling deps +At some point well either find the solution it will go away on its own or well decide we dont care anymore +This will be Dupped to a new JIRA +Ok I can confirm it complied and ran +Would need a testcase for this as all of our https related unitintegration tests are passingRequest for testcase unanswered for several months. +Will take a look. +Hey ShaiThese fixes are great and I will incorporate them allI worked this up very quickly based on other less general code I am using +I think in future we should move all the lib dependencies in the eclipse files to come from builddistlib as that will help us migrate more stuff over to ivyWill run tests and commit once the tests pass +Thanks for the review +I cant reproduce with the same code sample and with a similar on the devices that you listed +Ive just tested this in and it does work. +Hi LarsWhen running in these kinds of concurrent session usages performance issues it might be better to just use a session per thread +Consider a scenario like this Users Connected to the Whiteboard Draws a ElementThat update is send to the server and the element collection at server side get update and undo snapshot is created Draws another element which depends on his first element and parallel click the Undo ButtonIf server received s drawing update first then it will be added to the element list and Undo snapshot is created +Addressed last set of comments +Is this still an issue? If so please reopen +OK that makes sense +Passes tests locally +Im just curious if any developers can confirm this +Resolve with both versions to our QA +Out of curiosity why not upgrade to SLFJ while you are at it? +when a null parameter is sent to the modelStatus event. +Finally I managed to add the two missing lines to the patch +If you want to backport the fix for to the branch the fix for must be included as well +I am having the same issue also withmvn releaseprepare clean install +Tom can this be closed? Looks to me that the patch was applied +Removed excessive synchronization from the UIMA AS reports synchronization on in the project +However testing that image with the section commented out still works so I guess only doing that check when the strict flag is set is good enough +Is anyone aware of how to do that +Testing updated patch + is currently hard coded to CHUKWAHOMEconf as class path for locating configurationThis is a duplicate of code should be deprecated because its incompatible going forward with newer version of hadoop +The fix works for meI agree that the really exotic are going to be too much to try and support I think where we are now is just fine +The negotiating transport needs to preview a first few bytes of the content that the client sends to the server in order to detect the transport negotiation header +Emmanuels suggestion seems pretty good to me but I dont think this is something to dive into for +Well need to fix the receive no wait issue first +Im against this given we are going to make the ajaxjsf CDK a dependency of Seam which does this and a whole lot more +Logged In YES useridBeside the fact that the given examples are both not correct examples I could not reproduce the problem +I grabbed the project this morning to fix the generics issue +But I have some postprocessing tool in palce to make final output as we need it so I try to patch this problem in this postprocessor +What is the status of this issue +Tested +Fixed rev +Patch attached +Yes I think this makes sense +moved to CR since this is a enhancement +Excludes the new test from running with client driver jcc and j because it gives out of memory error +Please create another issue and move your sample files to it +what about systems with versioned packages in their doc dir +Thanks Myrna thats a good idea +Ive added some catching Exceptions and checks for null and diagram works but the problem is cant findload my class +Am I missing something? Also is it possible to get the change I proposed into the code baseThanks have redo the type resolver to support all generic typeYou case should be handle by the current code even if you dont see your fixSee unit testGilles +The fix will interpolate version expressions for parent project dependencies dependencyManagement plugins pluginManagement dependencies and reporting plugins +pass unit tests and addresses Thejass review comment +After deleting hints I get an NPE in count and list still shows the host and also a new entry that appears to be garbage +Please if you still have problems with the test files provided +It can provide the utility methods as protected plus implement the general framework for index replication using abstract callbacks that are implemented by the concrete handlers +For the Tools M milestone this is dependent on the upcoming ER build +Attached the trivial patch to fix this issue +Just FYI the fair scheduler does not use the cluster status to find the total running maps reduces +Scott if the object has a metaClass already then using the property on the class wont change that +Restoring the handle method in the anonymous inner class in does not fix the failure +Tried restarting eclipse but the lag still occurred after restarting and without the Controller annotation +One note on this I also inadvertently upgraded the dep that this code originally had from the API to version +All but one of the subtasksdependent items are already closed those things have been removed from AS trunk and will not be in EAP +For some reason I had the impression it takes an IR which is obviously wrong since it wont be allowed to close itbq +Max Can you give this one to the Seam TeamThanks can if you could put in info on what versions of Seam and what you are actually running into the jira +Patch to not check the generation when gossip is not initialized +Another update SMX root pom has been branded for FUSE and set up in +Jetty trunk has merged in the jaspi work +we should probably be using the vm transport for the connectorsIt would be great if we could put in some substitutions into the spring config file +I think it is simpler to leave it to when you want to handle the response that way + patch that removed the initialization during the creation time +I would like to have similar control +Im going to commit this momentarily +This needs to be reset +Ill commit this to the branch tonight or tomorrow unless someone else beats me to it in the meantime +Attached mgmt server log file agent log from and cloud DB +Unassigned so someone else can pick it up +Updated patch with hdfs only in this patch +I will but the indications in that bug were that doing that fails for builds which I took to mean Xerces as a whole and I only build shared + +Example of project whos classpath is too long for the exec plugin +Thank you Markus I have committed your patch in rev +reopen if im misunderstanding it +Currently MRQL code includes both spaces and tabs +It does look related +Hi Plamen thanks for offering to helpThere was concurrent work to commit in the two branches and I filed this Jira to reconcile the conflictsIf you can review the merge patch for the change to commit that would be great +I changed it to be ubuntu +One minor nit the debug log should be within an if debug enabled checkNot necessarily related to this patch yarn has enabled by default MR has them disabled by default +I just committed this +Linking +being caught No I dont think we can guarantee that +No doubt someone who understands all of this can sort it out +Closing this as invalid as one of the other tickets put in a in module essentially fixing these tests. +Pretty funny tooIf you post a patch Ill review it promptly +I added your sample project as an IT run +We could use reflection +The fix is documented in the Description field +Subsequent failures in fixed by Satheesh +It should be possible to have the former without the latter +We should rather optimize the query structure somehow instead of caching a rewrite method +Even Digest would be unsuitable in your case given the desire to avoid and the requirement to use them with I guess some users might like to operate with a BASIC authentication mechanism but still free their users from unnecessary logins +David is this still valid +Please provide provide instructions to reproduce +Changelog request for Japanese translation of new messages in LDAP connector +Virtually everywheremlexHandler is checked against null before being used but not in this locationI found one other place where mlexHandler was not checked against null before useand added a check there against null so Kens patch and this one other location isrolled into patch +Closing as invalid feel free to if you have more diagnostic dataevidence. +This patch version now startup in jetty and loads all the servlets +It should +I assumed eclipses xml wtp editor would handle this automagically +applied thanks! I made a few modifications to clean up following this. +I think the concerns that Martin and I have both expressed revolve around the patch designed to meet that enhancement request +its urgent to get the solution to get the fixed version of the continuum jars +Currently I am doing an actual byte increment +I would be happy to give a try if you send me a link to a build snapshot +Thank you +Could not reproduce +Note that the stacktrace above uses which is identical to the staged release of when this bug was found. +Thats itHopefully this will be small enough change to be accepted without much debate +Patch has been applied +One potential solution would be to use instead of bug must be fixed with caution since it could create a security solution that was to use internalLogin would mean that we need to make this method public which would create a security hole so it is not a solution +I forgot to say that previously. +I have not run the perf test to compare gzip vs snappy yet +Committed to trunk +I can reproduce in the see my previous comment +Thanks for providing this patch AndreasEven though I would like to commit it I must admit I dont fully understand whats the problem and how this patch fixes it as I dont know guts of Batik +No problem I will change it +Also both tests failed might be worth checking +What is the problem +Havent seen this and not reports in years. +This has been fixed as of in Felix and later +looks good +Tasks in Hadoop run with the uid of the tasktrackers so restricting processes by usernamegroup has no effect +I think I am done here +They are not set by default on the You have mentioned this in your earlier commentsThanks Viraj +The codebase as of revision to be donated along with matching MD checksum and signatures +opened to track +In the tomcat deploy model it does rewrite the address +Which sort of makes sense from a ui perspective do you really want to show things to a user? instead just show the first few and a count of how many will be deleted +Resolving. +Hi MariosWas the patch ever pushed to masterIf so Id like to close this JIRA out. +Documentation is now included as PDF in the binary distribution. +Did you have any other configuration changes you had to make for the text extraction to work? I am concerned that aside from the Tika issue that there is new configuration needed from that has not been updated on the Wiki +It works as expected +But it would be nice anyway +I went ahead and added it to the Gradle build so if anyone else builds it they wont have the issue but this does not impact the deployed javadoc since the fix is integrated with the build +This coupled with a dispatch mechanism caused your construction to route to the private Stringintfloat version of the Font superclasss constructor +I look forward to the release +Resolving this issue for based on the positive response in the linked forum post. +I postpone it to the next minor version of eXo JCR + +Tomas thanks for working on this. +JPP is correctly detected as JPP +There was some earlier sample code in related issue but I suspect its no longer valid +Thanks for the review Brandon! Ive committed this to trunk and. +this looks clean to me +I think my issues should be how to convert the byte array to the objects in client such as ruby python etc but not the hbase +For RC we now include a bit installer +Can you provide a more detailed example? I think this is related to other reported bugs about how to treat absolute path on Windows what with drive letters and all +Once fixed for we can consider a backport if necessary +keystore in homeclerzzainstance launched clerezza with nohup java XXM Xmxg true jar +class files from hitting the diskWhat was your environment? Tomcat Jetty otherNeed details to reproduce this or it will be marked unresolved +There doesnt seem to be a way to inject error handling unless Im missing something +Otherwise looks good +I have created the bundle for LuceneIm preparing the patch for Camel +. +The change is simple it makes the code more robust Ill check it in +No way to serialize the jobs and resume the progress at a later time +If thruDate specified then I guess the overlapping should be limited to a limited time what I mean is the next overlapping records fromDate should be validate to be a month or so before the thruDate of the previous recordFor exampleI have added PROJECTUSER group to a specific user the fromDate thruDate of first record was respectively and then I added one more record with fromDate thruDate to be respectivelyIn this case the validation should have been that The fromDate of second record can be maximum of a month before the thruDate of the previous record +adding that would make it evident that the translation between CSS and SLD is not exactly what one might like rules are out of order and more rules are created than in the SLD which was taken as an attached you noted on the pull request the titles in the generated SLD leave a lot to be desired +Agree I think rolling on restart is sufficient and we should tackle this feature more generally as you suggest. +We perform the check for conflicts between and when the user finally submits the offendingbq +Security policy implemented in Rampart issue +Checking for null before the test can not be compiled after your patch +We will add those +The fix made it in for final +But now it clear what these issues are definitely the same +Which cases do you mean +There is a known issue with M +done +I applied the patch before submitting this bug and it resolves the problem +Not a blocker for Axis release +I have fixedthis +Resolved in r +In my case it takes about seconds to finish the snap +Verified on +But perhaps I should have been paying attention to when you say I have no idea what causes this instead. +Im unable to reproduce this issue +Marking as resolved. +Quite likely I only use project archives no webtools modules +Your patch is good ive tested it and its working fine +I did a of the patch +Anyone who is interested in knowing the Producer logic inside out can give this a try +Could you attach stacktrace +Sorry new eclipse and project settings +Moving to as it is too big to go in the line now. +Linking issue with proper solution +We welcome patches with associated test cases if you have a fix in mind +We are setting the content type but as Bernard points out is ignoring it +Seemingly small changes like this requires lots of testingIf you need something fast I recommend externalizing your configuration reference the configuration in your POM and have your plugin use decentxml parse the XML to preserve everything +If you do so it is natural that this handler get invoked for all requests on the server +Lets see if we can get this in Camel if not postpone it to need an version of the +Anyway that was different since we were creating new artefacts rather than renaming existing ones so youre probably correct on reflection from an ASF policy +So would there be just a single for the entire address or would we get different counts for each durable subscription +If you can still repro this feel free to reopen providing steps to repro. +It seems to me like this should be fixed in relation to +Did you tweak the buildResults and other metadata artifacts which might still refer to carbon instead of cocoa +strongly agree that encoding order should be explicitly defined +Hence boostrapping will fail if they are not present within their respective registriesAlso note that we do no need to implement all of these things +This is the TODO list of the Hibernate developers not a support portal +Jackson is up to +Ditch the interface and move on +The code that worked in Axis was changed in the cycle and now I am just not sure we do the right thing here in all cases +Yes often a capital D is appended to the name. +applied. +This would make the relevant question to be is Mavens jira managed by or on behalf of the ASF +Assigning to Marshall to coordinate any necessary documentation updates with Yosi and Sivian +Hello AllThis version of the project successfully exposes all the Configuration and Instrumentation Content Messages and methods of all the objects of C QPID BROKER on the JMX agent +This works fine for me +unit adding a summary +Encoded and encoding of document we could see that the purpose of Encoded is to handle the encodedecode for usSo if you dont annotate an parameter with Encoded you need to handle it all by yourself +I also want to use it on a portion of the site that is restricted to shopper +This a parameterized test of with Geohash Quad gridsI deleted which didnt test anything it was someones work in progress with no progress +But still need additional info to try and track down whats going on + +Ive updated these filesjbpmpluginspluginsfeaturespluginstests I have to ask how some jbpm plugins are at when we havent had a release of +A new patch to match changes in Common +Though as this is a synchronous communication which uses dynamic response channels to send the response back to the client where the responses are themself jaxb objects in need for marshalling again it is impossible for me to add the conversion to the dynamic channels +Thoughts +jar files in studio plugins directory +I wonder if we could do some estimation though by wrapping the input stream during and counting the number of bytes read + +There is a question mark that we should send the author some money +Could you please advise me in specific topics to studyI have saw that some functions are already implemented +Hello AshishThanks for your comments and renaming the the recruitment process that we have followed for implementation of recruitment sub module is followed in our organization and the concepts have been referred from book named Human Resource Management by +Verified this functionality on the latest build and everything works fine without any issues. +I added the timezone lines and then committed +Id first try to get in touch with Dennis or Nigel but I think ultimately thats what we need to do +I assume you tested it and it works for youCould you possibly resubmit the patch as an attachment and click on the grant to ASF check box? Thanks +see previous sorry for being a bad listenerPI believe it to be fixed now. +This is actually a regression and even on Windows +Do you see any other potential problemsNope this patch looks great please check this puppy in +And btw this can lead to long EL expressions for the parent group if you have many subgroups. +I moved it up so that it can be reused for logging also +Integrated to branch as well +Ping for review +Also I have observed that when deploying an artifact on the repository the statistics does not get updated unless the Scan Repository Now button is clicked +Hi Aniketwith reference to your point I think it is the same issue I raised in I also think that it would be good to solve it for good even though it will require quite some rework of the grammarBut I think it will pay off in the long run by making the grammar cleaner and easier to modify +this patch should resolve the problem it also enhance performance of VPE update +They are doing this in application level +Publicprivate keys signing encryption can be used as well via the interface +Added unit test in for this the patch now to include unit test for this bug please review again +You probably forgot to change the portlet API to +Without something concrete to reproduce your issue from its hard to figure out why All works fine for Tont and others but fails for you +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +May prove exceedingly tricky without being able to change directly +Hi Guillaume it should be released this week likely tomorrow morning +What you have done in page management is to delete the page not the node this should be expected behaviour from the error message +I originally had that to and did not change anything +Are you using a bit jdk with a bit download maybe +Jakob suggests to make use of a config file +overwrite flag was already passed as true from +again maven netbeans projects +I think Ian might be able to do it +In this patch closed the div and render costsForm inside the instead of +So I am ignoring the problem above +Can we use JAVA specifics for the trunk +Um +IllBilly if you are still willing to contribute a patch I will be happy to review it for inclusion +This is really steamcannon appliance +Once we get the hadoop tests passing I think we have two precommit jobs +Fay is the attached patch still validneeded for +Thats what I meant +but it wouldnt hurt to clarify that +Attaching a new patch for branch merging the patches +So by simply using as your type instead of should be save +Tested +I will write a fucking man page. +Zipped simple project that demonstrates +I would be interested in hearing if the patch works for you +It can generate different types of data and different sizes +committing now +Which one is correct +Looks like the test is invalid and needs to be reworked +So apps using the mapred API and compiled with Hadoop seem to be compatible with Hadoop for job submissionmonitoringstats +Based on Oracle data generates bad schemas +Could you check all these cases and either close the bug if theyre all fixed or attach a more complete test case if something seems still to be brokenThanks +This is possible when a client is stopped while application are sending requests +Ill try with latest version I dont want the parameters to be matched +Deployed to snapshot repo +Is there no manner in which to leave the old methods in and simply deprecate themThinking about this +Part fixes the workflow update SQL syntax for Postgres version and Part fixes the query syntax for Postgres version. +Plugins could help if there are better detailed ways of getting resource statistics +Fixed. +I think there areseven in the fileThe fofootnotes in are not strictly legal but they are notstrictly illegal either +Hi Saschawhen displaying a calendar I still have the month label in english +Would it be possible to give an example of a connection to an LDAP serverThanks! +To follow this principle velocity should be configurable in the same way as velocityCount is +Thanks Dmitry and Sergey +From what the Description field it appears that the issue was that specific configuration changes could be made via the JMX console but not via the configuration files +I just stumbled across the message trying to track down some other problem +Nicolas this issue is a critical for +So crawl script would generate a random id and pass to the generate command + so we have the code in branch dont resole the issue yet though since might not make it into GA +verified on CR +Changes just need to be applied in prod now. + for the patch +Thanks for your comments +Attached patch addresses the problem at least partially +If there is interest in a release I would be willing to backport to +Patch which addresses the issue Ravi pointed out +Can you verify and close +if you have any specific ideas regarding jasypt integration in the future Ill be happy to discuss them with you +Not at all +Anyone reproduced this recently +Please find the attached file +Commented on the pull request +I reviewed the changes in version +Maybe the above isnt true anymore +Thanks for trying out and posting your results Dave! Klaus maybe you have time to look into thatDProbably not before the weekend +was introduced when we fixed a similar issue so that has really been an issue for a few releases +I am just offering an improvement over the current dialect +Hence we create out of the parity files to reduce the number of new inodes +Version attribute. +Alan +I have also attached the plugin +I suspect I could log a warning but other parts of the runtime would blow chunks with a similar NPE due to the missing TM +A fix to the code would be to make ant test dependent on package +Delaying a fix for +JohannTheres something wrong with the attached patch +I think the better fix would be in method which checks if the class is specified and when it is not found logs it at info level and throws an exception +What could I be doing wrong +Will commit now +So this is probably coming from some image service +So it is +Hi Taciano TresI think that this shouldcan be done not in Sonar core but by providing separate plugin +Are you interested in a new patch to add nested instantiation using select new to Hibernate +I have pasted all related codes in the forum post +I created a separate issue rather than modifying as it describe a different use case +Thanks foryour support sofarRegardsHenk m very sorry for any problems caused by my rsync +Ive checked with a clean checkout and on OSX and it works for me +here we go +Otherwise WONTFIX. +Bulk update to fix version +It is already possible that multiple URL encoders are mounted for the same page type +The DMR model doesnt allow modifications atm +So I shall wait for a fix. +all issues to the CR +Especially in the latter ones its easyto build Cocoon yourself +I will try to find an instance of windows to work it outDid you know if the problem appears with other windows versionAt final even with this warning stuff is done as it should be +I am pretty sure and I have run my codes at both and +Note this only works on tomahawk for JSF in facelets mode. +Applied proposed patch +ColmI fixed the thing so the test now passes +Thanks to Robert. +Ive looked at method execute +Reopening +I am attaching patches and to respond to the comments from Andy Purtell +This patch also updates some stale information on the derbysource page +If a level is mentioned in the listCapacity API the top consumed capacities for different capacity types and hierarchy levels under that level are listed +I just committed this +Ramin please reassign yourself if you are still interested in pursuing this issue +For stage verification we test against stage Test Plan +This patch is the piece from that commit that addresses the local filesystem problem a change to only +I will turn this into a Groovy unit test and add it in +Test has been running for hrs and memory and threads are stable +Fixed in subversion repository as of r for branch and as of r for issue as it was included in version which has been released +Nice! Thanks +Have fun +If you set the options to be save truethen everything will work as expected +patch the patch I think this solves this issue +Sounds good to me Ill go ahead with itI thought I may as well tackle internalCode while Im at it but I cant find much information about it +It may be that you have not called start on your connections +This feature should be marked as experimental +Mahadev the in trunk has a line for this JIRA in the trunk section and not the section +Ping me on irc if there is another problem please +issue exists for this reduce task page also +This was generated with git diff HEAD from the directory +As I said before I think this is a bug in Pax +Ill try to nix the findbugs +Hadoop uses the logj implementation of slfj which both common and mapreduces explicitly require but perhaps not in all the right contextsI think Avro marking it as a dependency masked MRHDFSCommons dependence on it which isnt explicit? Its printed whenever I try to start a MR Tracker from a binary tarball +There was a problem with RENAME on the mailing list today where a user had all their cores sharing the same instanceDir but dataDir was missing on all of them +Thanks for the excellent patch +the APIs are not different that is the problem the implementation isFor now and is using the same implementationLast time we added these on the same classpath of there were classloading conflicts when we had different jars +I also found which is exactly the thing that addresses my use caseThanks. +Changing works around the problem +Actually usrlib is a common place on x architectures to keep the libraries +You are right Junjie Im sorry I misread the code. +I agree +If any objection please reopen it with comments +Assigning as per related issues +Merged +The person having the issue has worked around by using a server installed elsewhere +Ill commit later if there are no objections +on Jenkins. +I forgot to grant license to ASF for inclusion of code in the previous patchAdds field name to exception message for missingincorrect typeIf a field is a destination for multiple copyFields and it is not defined as multiValuedtrue then it shows a Warning during schema parsing. +This represents a performance decrease +Yes you should not have two versions of thrift in your cp or stuff like this can happen even though the data may be wire compatible the libs usually are not +If youre using finding the is not a problem and if youre using a then you can instead use a +Yes! Ill add them and update the tests +I plan to remove it and fix corrupt replica processing which should remember corrupt replicas with wrong GS reported byWill address the issue in a separate issue +Pavel done is it ok +I would like to work on this issue +Looks like the dummyRow flag is not needed at all +That way we can better evaluate converting it to use Avro +My last comment was a red herring +Uses now as initial value for codecIDalright that looks goodbq +Thanks to CamilleFlavioAlexPat on this! A true community patch +No user feedback +It turns out that that page is not correct +You shouldnt place in the ear only in the sub appears that the automatic ear isolation isnt working +Patch applied on transition all resolved issue that did not see any further comment in the last month to closed in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +pushed +This patch just gets rid of class and uses everywhere + has same problem +Awesome! Thank you +So far nobody seems against itThe downside is that it MIGHT break some existing code +thats why i scheduled it forbut after all the implementation is still work in progress and Issue has not been closed up to now +Bulk close for Solr +The patch does generate valid mapping files just more verbose than is required +from me +This one is from our staging servers where you would observe that the Job Scheduler shuts itself precisely at hours +Patch applied at r thanks +Any news +This is a core issue as it shows up in multiple as a duplicate of see that issue for further discussion and a patch. +I think it would be nice to update javadoc according to your answer because all it says isbq +Ah I see +Thanks Jing! +It depends on what FULL means +Committed at subversion revision +We know there are some serious issues re building Xalan on Sun JDK with the old version possibly preventing the new one from compiling properly +Additional patch for more functionality +Until it is and the brew build is redone from the QE approved src the current integration is equivalent to a snapshot +Works fine for me thanks for the patch +Indeed the problem is visible also under Linux and is in progress +So you have to pick the framework and convince why you did so + updated patch per Uwes comments and unit tests +Thanks for your input. +qdox does process artifacts in definition order +My mistake +If someone wants to implement a newlock manager based on the ideas in this patch please feel free to doso +gets saved asClosing this issue +Hop on asfinfra and talk to Joe or David Blevins this was setup the other day so you can just symlink but Im not sure what the actual setup was +Thanks +is out now at codehaus maven repo should get to central repo in an hour or two. +I forgot to mention there is a sql script to create the sample table and data +There were several issues related to failure ofIve just linked them together +Checking the updates now +I think that this default value is enough long as a RPC request. +This is not an bug s not fixed but a wontt fix +Apparently an alternative RPC can give a performance gainfinagle is very mature in compared to my own +See if you like it betterThe downside of this approach is that classes would need to be changed or deprecatedCheers OlegI think the patch file did not pickup the new file but thats fine as this solution looks better +Thanks for the patch Binu +Negative an association might be +Retry testing with the attached file and tuned file to detect failure of coordinator quickly +Minor delegationToken variable name should be renamed to tokenFile because it actually represents the file name that contains tokens and could also contain tokens other than delegation tokens +I cannot reproduce the failure +It would be handy to be able to specify a custom entity resolver as well +The issue was that sample autocompletes were using ajax mode which doesnt filter suggestions on autocomplete to client mode have resolved the issue +It is now a case of ensuring that we are using the best regex to catch all the idiots +This problem is handled by following the instructions for importing a project set which will be described in the migration document +The changes were integrated to trunk via and its related jiras. +Thank you Paulex +It uses a readerswriters lock instead of synchronized so we allow several readers at once +Does the exception go away if you change holdability to CLOSECURSORSATCOMMIT? c What happens if you close all open result sets but not the statements before point Thanks for the report +bq +Updated patch to address failing? unit had time to look at this again +yea agree +Part of a bulk update all resolutions changed to done please review history to original resolution type +They now pass on Windows and Linux +Agreed this would qualify as a tool to register in whatever comes out of +Will let it stew a while +add RINGDELAY sleep to +Thanks Keren for picking this up! I think there is some interest in exploring this idea and there is certainly a value in prototyping this +Redirects are handled by the graduating project WITHIN the incubator websitePlease see their graduation docs +Then provide a tested patch that implements your changes and reopen this report +Oracle ultimately owns the copyright available under the license specified +Running tests +He will submit a few optimizations that will take it down a bit +look all the tests fail +It might be an option to alter table but great thing to use +Closing. +Applied patch modulo fixing one forgotten is a completely unnecessary patch you can already specify the name of the lib directory thats the entire purpose of the LIBDIR setting I dont understand what this extra code addsUnless I get a cinvincing response Ill revert the this patch +Normal patch without shenanigans +Chapter Performance Tuning in the guide should replace chapter in the guide +bq +All dealt with records the need to patch the rpc plugin. +Nice meeting you at last week +use a local variable to keep the prefix for nill case as well. +Is the link still wrong? +In any case it works fine again in. +Im looking into checking for null inputs of addType and add +Regarding these aspects it works properly only persisting wont work +Thanks Earl for you contribution. +Tested and cannot reproduce +arirang is the name of traditional korean song +Seems to be a issue to Gary because he owns the depended issue +This exception most probably is not expected to be thrown to the user +should be this file +However with rollback close connection and reopen connection the mailjdbc embedded test still has deadlock error after hours +bulk close of all resolved issues prior to release. +moving to mojonetbeans where the freeform generation bugs are filed +Tried HBASEHOMEbin no change +There seams to be some work left as not all possible tests are executed yet +However I would like to understand why the IO and SSL sessions got out of sync in the first place + Why there are new FindbugsCreated +No lets reopen this agree the committed fix does more harm than it helps so I reverted itIm trying some other approach so will hopefully attach another version soon +Yes Freddy +I am getting this connection object and I want to use this connection object with the JPA entity classes which I have created +container provided jars leak into the run Failures Errors Skipped Time elapsed secThis tests verifies the jars that get installed in the target container +distrvm Xaosinitialcompileropt XircO Xircverbosetrue Xirchightrue Xircfinalhirtrue Xircmethodtoprint XvmmeasureCompilationtrue s m M a jack initial HIR has basic blocks +Implemented as documented in the svn commit +to branch and trunk +Trunk Committed revision +On local machine I had less errors update of IBM JDK to version SR requested on Hudson +This is related to clustering as well +Because if more than one forms have the same format name then the generated might use the same format name to refer to the form fields which are in different HTML forms how can we handle this situation? +Bug has been marked as a duplicate of this bug +gcc is way out of date +Fixed +yes I think what Ning is saying is correct +bulk close of all resolved issues prior to release. +Ill update the patch +Committed revision +Updated patch with the ininput rename +Thanks for reporting +get all open statuses and get all closed statuses now reads from the description column +I have cygwin installed in Ccygwin and Java in CProgrammeJAVAHOME is set to cygdrivecProgramme +Just uprated the gwt version to this version has only minor functional changes +I think this issue can be resolved as we removed avro related code in trunk +It is probably more robust so I like the idea although its slightly more code to perform for the Posix case +Fixed in revision +has reference to Configuration object +now has a property schemaResolver which can be initialized with constructor or the setter + + has been marked as a duplicate of this bug + moved into ipc package renamed to removed +Binglin cool +A workaround is to remove lazy loading on your relationship +I think we have to improve for this which is +I realise but one of the classes listed above is an Infinispan class +People interested in this patch should take a look to that one +Perhaps that approach could be taken here. +If the testsbenchmarks need something cutting edge in mapreduce then the corresponding mapreduce jar can be committed to HDFS at that point of time +We will wait few more days to be your updateMy understanding was that a fix for the issue was added to SP +Moving out of +Maybe we can open another jira for this to implement the exists check on grant and verify in all pre if theres nothing leftThis is a good idea since its a different problem scope than this jira +probably this should be done irrespective of scheduler +Attached images for documentation +We add a if statement to prevent the corruption just like we would do in any other case +Updating affected versions based on whats reported in +Yeah this can be committed. +See for the version of this +Hi Alessio JBDS is callingI just tried to call on Vista machine from command line and I cant get error results as JBDS has +Work complete +configuration file for embedded more information about java environment activemq is running on +I have verified as mentioned +Ive cleaned it out for now and will investigate why these arent redirected properly. +will handle as please refer to my commentBasically this means recognising v and ignoring them with a warning +I think Ill convert to using artificial prod events rather than external timing +I may be wrong but in your patch are you assuming a datanode cant be in more than one pipeline simultaneously +Yes because I already pushed everything into master last week +The mechanism in itself is fine +So what you really would like to happen is that the Release Plugin should update the version of the skin in the of module during releaseprepare + +So all issues not directly assigned to an AS release are being closed +Bryan I have got this working for my teamproject last week +Yes the protobuf code is careful to convert endianness where necessary +See for jetty access for adding the new issues to JIRA! Do you need because you are exposing the cluster on the internet +Looks good thanks for test case it will be good to have a function implemented with no use of the base classes +IPC threads can do the job +I use partially my own program based on the verifier from the asm lib or the thatverifier from the asm lib directly +Ill fix it for sure this week +Added a test to demonstrate that splitting by a specific route key is already possible at the level +Thanks! +Minor fixup ensure the pos returned from is not remember this case +First patch for Watermarking. +Why dont we directly go from SQL DDL to? +what are the basic statistics +Im back from vacation Ill retry tests with the latest patch +If you want to use the colored logo it also fits better in the green scheme because the green is again represented in one of the three arcs +Allocated cluster +Im satisfied this is unlikely enough to happen we can resolve the bug retroactively +Fix should show up in the next snapshot build +Triaged for +Ill see if I can find a better approach +Its variables are substituted to create the actual file for a given project +This was all on the same system but on different startup sessions +We can really use this in the ASN codecs +read the wrong file name its +EAP CP got released with the wrong serial version uid +This introduces an optional dependency to to +All bundles have appropriateIf the complexity of this is low why not just do it? Does someone have a patch +B it indicates a changeregression in the installer or JBDS it self that can be tricky to restorefix in an update C it might indicate problems in the runtime detection +I only wrote down the question that came up my head when thinking what will happen with or without this patch and how this feature can be tested +I like to run whatever test runner the systems that are running regression tests against the code line are +Applied patches and delivered. +Do you have the command line in question? There might be a way to pass parameters to using a file +My preference for serializable classes is to use the number which would be generated by serialver +Cherry picked commit to with commit verified it on the latest build with the commit id cabcefdcffecaa +nd attempt at Patch incorporating feedback from Runping Qi +See the latest +Unfortunately Im developing on and only classes got into the codebase +Changed owner to submitter. +The attached patch resolves this issue +The bugfix from an administrators perspective is the logging behaviour +It sounds like this has been committed +Patch for patch incorporated inThe patch is also applied in Hudson builds +Let me comments on the patches I just attached is for +The Java EE implementation module is not exposed to external plugins +Please fix there are some bugs in Seam validator +If not then we error out or at least output an error before continuing to bootThat would take care of any random new parameters ours and any other use case +With java in my locale these two string parse just fine in java and jabjhoijgbThey will both return the same date +Alex please rewrite the paragraph that describes showArrows. +I think an append will seek to the end of the committed data and then overwrite the spurious bytes +Considering this issue resolved. +Robbie had asked me to write the test cases +This jar is intended to be extracted from the Tuscanyjava directoryThe content is a test case for demonstrating the problem or verifying its correction +Cheng we narrowed the scope of this ticket to actually get the rewrite in do you want to open another jira with this feature request? +This is dependent on changes in the maven site wont fix after fixed +Committed to trunk +Moving this to RF because Im assuming you are referring to the source code tests and not the QE tests +Change the title to reflect its a UI bug +Could you please make a try with it +Upgrading automake to made +Every part of a document is stored in a different class extending the abstract class +Please close this one +With JMS the classes were and so needed to contain internal modules and as dependencies to get around this issue and app back to work +Is it the same for youYoure on Win are you using Cygwin to run Forrest? +They are just name spaces +Yes thats exactly what Im talking about Im not looking for a more elegant solution at this point +With this patch this issue does not depend on any more +This implementation seems reasonable +applied thanks again! + +I agree with Mahadev +In it was mentioned that the calendar object does not so getting time of day is actually going to get you time of day of the calendar was first asked +These are the versions of the libraries Im using +This patch makes package fully signature compatible with Java This patch can be applied to core and modules +Updated buscamel +This kind of enhancement has to be performed only a trunk +Bernt is busy elsewhere these days +See am deferring this +My bad I seemed to have missed them during update now +I attached a screenshot of the output +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content issue only to change Release Notes Docs Status field +versions of +Sorry for causing confusion +So there are two options + as Improvement since it doesnt look like a bug to me + install with journal to replicate ve added some warns and messages that have no queues get ignored now. +Ive not seen this error anymore +I have applied this patch +The hard way is just to support generic typing which I can do +and are the two standard ones please show how any user is supposed to change that? An just resolve from name to a value no prefix setup is possible +If you find further issues please create a new jira and link it to this one. +Thanks! + was a bug with permission check related to dashboards and was not working properly with any generic implementation other then portal one + to allAs a side note How about requiring Maven +Maven is a significant step backwards in terms of being able to dynamically add Jars to the classpath at runtime +Reopen to move into our new testing state +Aspects are relative to the config file that specifies them +IMO these test failures blocks any other patches from getting in so Im motivated by expedienceIs this good enough? do you want me to follow on to make this test more robust +According to the MSDN this seems to be the caseWill be fixed with in the trunk by making sure the option is passes when setting usernamepassword +that is generally what hadoop itself does by finding things relative to itselfAnd definitely agree with Joy +Uploaded patch for both and trunk +Most projects prefix commit messages with the relevant Jira id +Unit tests have been updated and all run +Thank you! +But here the problem will be that NN has already removed the backupnode it is not sending any requests to BN +All my changes are in the last patch +features +However I am not certain as to why you would want to remove it and create one large file that stores all logs such a file would be really hard to index or run mapreduce on even if you create splits +Superseded by +explainning how to integrate with google check out +removed old pdfs added with do you have a copy of the file to review? Thx +I needed to update the Scan instance properlyI have also made the changes as suggested on Review Board where I will post the line item detailsThanks for the feedbackcommentsant test Dtestcaseant test Dtestcaseant test Dtestcaseant test Dtestcaserun successfully +Instead letting this bubble up to couchhttpdmischandlers should generate an appropriate error response to the client and avoid the incomprehensible badmatch nonsenseThis patch should be applied but does NOT close the issueIm continuing to investigate +So wouldnt cassandra validate the dataI have not changed my schema eitherThis is happening when attempting to add another node along with one existing with a replication factor of +Ive just deployed new snapshots +docview XSLT docview input used to create example committed the XSLT file to sandbox in revision Im not sure if well find more use for it somewhere in Jackrabbit but perhaps someone willResolving this issue as Fixed. +Sorry I got no reply from a PMC member about this issue nor did I see any discussion on either the dev or user lists for Juddi +One more small change make the message type id bits and the version id bits +Apache License headers method flushes indexes changes from the reopens and returns the latest Searchable +Well add support for persistence of timers in the next minor release +I wanted to try this mostly just to get myself more familiar with thenative maps +I create a patch for branch +package private methods are added to to read access token for testing purposes +Tests are added for basic and digest authentication modulestestsuitesrctestjavaorgjbosstestwsjaxwscxfhttp and modulestestsuitesrctestjavaorgjbosstestwsjaxwscxfhttp +Ive already updated English version and will update Russian language file soon +Ted Is that added by Subbus patch +Add unit Thank you Andrey +VitalyDima lets apply that workaround for now but keep this issue open for introspection +Jacob are you willing to provide test project that would allow to reproduce and further debug this issue +Just to make sure you have tried this against what ? or? And migrating towards JAXP sounds confusing to me + +patch for lunisecurity and cripto patch has been appliedVladimir please verify +This would probably require some package protected helper methods to give access to the private data +For some reason when I call the webservice on the second installed instance of the application it goes into a loop when loading the messagefactoryI am now using tomcat and the saaj version supplied with the JDK and do not have the same issues +Found a simpler implementation which acheives the same end +I agree that the test covers the reported issue +RobbieCould you please review and commit the patch attached? +Checked into CVS +DNS updated unix group created +attached depends on from from with rThanks! + fsck move should bail on a file if it cant create a block fileNote This is based on the patch I uploaded for +Patch for this improvement is attachedThe implementation is as below +We should not force our users to hunt through every last nook and cranny in order to discover the complete licensing for an Apache distribution +Ill add a unit test +Another quick fix is to start up the server with the bind address set to localhost +Hopefully the patch will help resolve our issues so we can get these plugins working really well asap +Ill link that Jira for future reference +results should be the same in either case +. +It is public to provide an easy way to update tests where we have only one key and multiple values +Attached patch to address this issue +Updated the functions with the appropriate arg mappings added testsAlso tried on current build of trunk to make sure all the up and down casting magic works and it doesNote the new methods in and pretty handy +Committed the patch on and trunk. +Yeah Jason +By fix I hope you mean that this declaration is acceptable and that it will only generate a single FK in the DDLAs a workaround I can get it to generate the correct DDL by only including the reference in the subclass for DDL generationThat being said I havent yet seen what Hibernate will do with both Collection references present in both class definitions +See plugin +Ive escaped deploy error by ignoring +Hi Tim do you run the testcase on a CPU? What exactly is your change to the testcaseI am looking into this and will let know about my findings next week +I can look at it next week and see if I can figure how it would integrate withSearchable and a +Other than patch looks good + really fixed +Committed to the branch thanks Aaron. +Michal is the issue still present once an Infinispan version containing a fix for is used? You wanna wait to test CR and then we see if the issue is still present +Sounds like it will be a longer term project to make into a more flexible interface so in the mean time there will be places it has to cut a full by copying bytes +This is an issue on windows because I forgot to close the file when quitting the server +Submitting the patch +Attached is a smaller reproduction with just a single thread calling the tests method in a loop and checking tables after each insert +In general channels are expected to preserve the order of events +This was fixed as part of Ive added the unit test anyway to catch future regressions. +There are more classes then those above please ignore them in context of this jira + +Closing as Wont fixThis tools will not be committed to repository +That causes a cascade of teardowns which can take minutes especially if the line comes back and the remote brokers try to login againWe do not have problems with a high number of messages to be sent in either direction or one logoutloginRegardsRon +Do you have a specific reason for this request +Without the patch monitors get added to and not +The Geronimo are broken at a minimum because they make use of the now unsupported use of XML entities to specify dependency versions +I had to wrap with a trycatch +Now abstract nested page flow classes are not required to have returnActions +I have verified this fix by rebuilding all the stubs with the latest nightly build and I see no instance of this problem +So the complete module was removed +for committing now +OK +since we do not document them and they are not in any header file in the include directory removing them from the set of exported functions by the shared library would not change the API +Patch looks good to me for what it does Should it also handle GBYDo you want to file a separate JIRA for the select case? I can try to make repro next to trunk +It would be really nice if the pid of the instance would take the name of the configuration +host the one with the host attribute to bind to an IP address +Today I tried it on a newlycreated empty project +Then we avoid adding that extra step in the route model which gets visible for tooling and whatnot at runtime +I had added these jars seperately in the lib folder of the warThe pluto portal has successfully deployed after implementing your suggestion +Thanks David! +Resolving the issue +ajaxRendered true by at +a look at the fileIm actually not sure what you mean by this +Ill have a look at your recent changes and commit to trunk shortly if everything seems fine +Hopefully well have time to look at this soon +charater at the end of +Should we clarify that too +And with that option it should be easy enough for someone familiar with ODBC to clean up later +rejecting this issue cause it was a usage problem +. +I can double check a little later though +I still need to get the CTF to complete a successful run but as I am going on a three day workshop right now this wont be possible before Monday morning I am afraid +So maybe may give you the best of both worlds change the file locally and commit then sync the changes to production where they take effect without when considering to fix this issue have a look at that has been closed because considered as a duplicateActually is different in that it does not require to reload GSP on live BUT need an option for creating a WAR without GSP +For either way works but makes more sense to use now that we have the decoder in the consumer api +I just want to make sure thatrequiredif will be left in the code base till the enhancement to validwhenis made to support mapped I agree we need to make sure validwhen is enhanced to support mapped properties before requiredif is removedIve changed the title of this bug from Problem with validWhen while using map backed beans so that it flags up more specifically whats neededNiall +Ill commit it once Jason comments on the above question +I didnt see anything obvious I read the code to make sure it was not going to reintroduce db corruption ofI would suggest more testing and at least making this feature high for buddy testing before releasing it +A diff file to correct the problem +Ive added a solution with a proxy +I think the fix to will also fix this problem +Gone in todays build from HEAD. +But there is another problem came up with this and I created a thread in mailing list for thatI think we have to wait a bit till we get a good reply for that +More progress +Attached patch fixes the issue and adds a unit +this one is fixed! +It has fds open +Juergen +This test case demonstrates the problem +UpdatedCan you please file a separate Jira for the link issueThanksArpit +Ulrich thank you for your support on this problem +Do you think applying the updates made to have some effect and resolves this issue indirectly +The javadoc of those methods state that they should be implemented when creating a virtual directory +Lars This issue can be resolved because you are looking into it elsewhere +gtr invalid backslash escape at end of stringecho i tr sed s +Currently for Sling we have two options fall back to or step up to +Relased earlier in. +Running through Hudson again +One thing to note +Committed revision Congratulations Ramin on your first fix. +Reason of performance falling have regression need additional investigation +Thanks Rob. +Version Released +Trying to connect to jdbcderbynativeusermarycreatetrueConnected to database jdbcderbynativeusermarycreatetrueStoring some sample users in the databaseTurning on NATIVE authenticationValue of requireAuthentication is nullValue of sqlAuthorization is nullValue of default is noAccessValue of full is sqlsammaryValue of read is guestClosed connectionDatabase shut down normallyTrying to connect to jdbcderbynative without username or passwordCorrect behavior Database connection refusedTrying to connect to jdbcderbynativeusernoaccpasswordajaxjxCorrect behavior Database connection refusedTrying to connect to jdbcderbynativeuserguestpasswordjavawxConnected to database native with accessCorrect behavior DDL is not permitted for a connection user or databaseTrying to connect to jdbcderbynativeusermarypasswordlittlexylambConnected to database nativeCreated table accessibletblValue of accessibletbltextcol is helloGranted selectinsert privileges to sqlsamTrying to connect to jdbcderbynativeusersqlsampasswordlightqbulbConnected to database nativeValue of accessibletbltextcol is helloInserted string into tableValue of accessibletbltextcol is helloValue of accessibletbltextcol is samCorrect behavior User SQLSAM does not have DELETE permission on table MARY +I consider this one resolved for now +As I understand this is restricted to client applications web applications are not affected +I also tested this with snappy +Will fix +Just to summaries we have multiple options to solve this problem now After detects a bad block from one data node DN it reads from the next data node DN +Fixed in the branch and trunk +In that case Ill try to get this committed tonight +Ill take care of it +Thats the reason I propose this +I would like to introduce you a small feature to GC heap verification infrastructure +bulk defer of unresolved bugs to +Super thanks Thomas! I just committed this +Example Type class autogenerated by +Should this have a Release Notes flag on it +I guess that means Im in favor of over this fix but either could work +This bug also happens on deployed artifacts to EAP which are neither exploded content nor zipped content as seen by EAP and have always used deployment markersAs far as I can tell the bug is a general issue with anything deployed using deployment markers vs +this needs the patch +I think this bug might be a bit misleading +Proposed changes to the Balancer +Fixedbmaster noting this is a feature +Dhruba will look into the DFS code and working on that +ok I personally wasnt super fond of that change either but it fixed the findbugs for now +I will play with it a bit and see if I can get something just realized that I have not switched this computer to the new repo could be that that is the problem +Patch applied on and closing all issues that have been in resolved state for more than one month without further comments +Fixed +Since youre a Yahoo! have you confirmed that clustering works in YTS or? I was under the impression that it was broken in and ATS is basically the same as +And at least for some the spaces are not importantAlso added a test to know if it breaks againTest checks if all cells in a table are in correct order and have spaces between them +As best as I can tell there are no conflicts between this filters +Not able to reproduce it so closing able to reproduce it +It looks good +attach a patch to run filter on cpp client +I will take it up with the Classpath folksThanks for your help in any case +All tests pass +Currently the is propagated to higher level code without checking for +This wouldnt happen if checkstyle was not invoked on every incremental build +Well if Falko could give some pointers already that would speed up the implementation +Thanks Patrick! +Thank youTimVerifiedI will catch up with itLeo + for the patch +If you do this you wont have to deal with changing the offline deployer list either +But would you please let me know how to recreate the next issue? When it was thrown +This is something Id also like to do but its not source compatible and thus something that requires some attention from a migration standpoint + +You can configure different number of slots for Maps and Reduces per TT +I believe this is fixed in SVN. +I was told the reason why once but Ive since forgotten + +Attaching patches for trunk and +bad UI to change stuff based on UI IMO it was called Explore before +Other than findbugs exclude rest of the changes do not seem necessary or adds value +The QA team have created tests +Warning jetty has not been released as a production version and still contains some bugs +If you want I can muck with it if you write outline of what to do +Hiya probably got lost in the previous comment sorry about that Each image runs a local instance of AS purely for JMS and requires access to an FTP server +Given that implementing is very trivial that could savemany ESB users some headaches Tom + +Fixed in commit +Please check that the patch applied as expected +btw changing your log level to a higher level like ERROR instead of WARN would fix your fine for me in a app. +Without further information I am not sure how we could improve the error message so I am going to mark this as wont fix for now. +I will sign the Apache SLA and send it back in order for me to have access to the wiki +i put a small file appender to plugin and i got these logs +This is an ongoing work item but the most significant docs are fixed for +JacquesIndeed it is working +Yes this is definitely important for Phoenix +I am sure he will have more inputs on how these adapters should fit with his structure +Removed the leftover references to Queryscores +You may also look at the code of the script its straightforward Python code +Begone! +Here is the patch merged into trunk +Two data files used by the test is also included +The same applies to scoping We recommend scoping over scoping in particular with respect to the default singleton scope in Spring that differs from the default prototype scope in I see your point +to make tests pass +I am attaching a patch that takes content of changelog file and creates a rpm spec changelog implemented changelogFile property and not inline text + test case to demonstrate uses JNDI and its LDAP provider to connect to the server +Looks good to me Id say commit it +trivial aganst its the same issue +But as a first step one can forget about writing I guess there is overlap with the complex feature work as part of what they do is mapping +It says nested comments are not allowed +has a for all libs so no need for every component on this task +The attached document collates all of them +Any feedbackcomments would be appreciated +Create the patch based on latest trunk the deletionService to verify the deletion +Do I need to do the upload procedure again +After nuking xdoclet stuff from local m repo things went well +Jars are now located in the following locationsAS serverdefaultdeploylibdefaultdeploylib serverdefaultdeployerslibdefaultdeployerslib in revision . +Ive added code to validate and print errorusage messages for invalid inputThere is a small catch here didnt have any validation code forcheckingwhile trunk had one +Disadvantage is the addition of an external dependency +Maybe the is not using at all +However that means that if Custom was selected and you went to the page Green would automatically select as Custom wasnt in the list +I just committed on the stream +Thanks Henningyou seems to found first issues in installerVasiliy can you please address this +Patch applied with trivial modifications see rThanks for contributing! +Master should probably move up to built +GreatIve already assigned this issue to myself and looking forward to a new patch +This is basically what javadoc saysThis glitch is checked in the assertion here in the jclouds codebaseThe above ensures that the hostname command corresponds with whats returned from the api +Seems like a best practice is use and if you really insist on NOT using it then this JIRA will keep you from deadlock +These conflictscan most resolved by priorities and associativitiesIn your case I rearrange the priorities of the terminals +Also I think get in TMD should return a set instead of a list +To have a configurable timeout while the Felix patch is not merge into trunk checkout Felix trunk apply patch in mvn build install the httpjetty module or all the Felix trunk update to the dependency in your launcher build your stanbol configure the http timeout property in the http bundle configurationEnjoy +This is resolved on trunk + Updates to to allow us to do our own parsing and running of Updates to to work with the new jar +Please note that the issue as described in the original description does not exists in the current branch +For example certain huge shapefiles could not be indexed they were simply too large +I cant see how this patch can fix the issue since all it appears to do is add a temporary variable raster and looking at the patch it is clear that it doesnt in fact touch the line of the test at all +Adding the capability for a user to set hisher own directory as snapshottable is a good enhancement for the future +Perhaps as part of orgapacheavrospecific +They will never be able to preview any of the response tables +I just tried to deploy a simple application the data could be found in the file of the deployed applications +that was intended for ticket +Get the and looking into entriesAdd unit test with a ear fileNeed to complete with new unit tests +Looks like the bitmasking stuff has been split off into this for reference purposes for those who get here via google like I did. +Bulk close for Solr +All I found in the partition package is the putting in value for GRIDSIZE so I can make it +No IT added there is probably some work to be done around integration testing of widgets since the current testing harness and widget API does not give us easy access to widget configuration +After remove doc warning I commtted this. +I will likely leave a config option that retains the current behaviorRather than proliferating config options I suggest we provide a which terminates the job on and use it in streaming as default +This seems to be a consequence of +for patch v +The name gets resolved by the dnsmasq to the primary ip of the eth interface +ftl files and not four +Charles perhaps +One more thing about this Jira issue I need this committed so I can fix the problems in the Workeffort calendar feature +Ill reopen it if the problem occurs again +or +Marking this as invalid. +That element would be usable at OR within a chain +to apply +I can work on modifying the JIRA plugin + Editor allows code completion for seam beans +This patch is not ready for committing yet there are tests failing +It would also be a good thing if the quickstart was deployed and tested as part of the release test suite. +Sanjeev can you have a clean setup and retry? From your db dump I feel that there are some leftover items on your db from previous S testing +This was open for Release Notes review +some times store rules in a standalone database will be more appreciated by some peoples that wants a better security and access control we can easily modify class to use both and for mysql mssql db +Changes includes following Created new search form for eBay and Restricted product and prod field to search +And I have some ideas on how to create a more simplified wizard that handles a similar set of tasks more straightforwardly +Whats the new? I guess I should just give up and output user item preference but it just seems like such a waste when I already have everything for the user vector matrix +None of the URI stuff is there yet I guess I ran out of brain power and could not find anything suitable to copy and rework +Why do you add server codes again +New Patch fixessimplifies the two queries Virag pointed out on thanks +Look forward to getting my hands on when its released. +This can be done with the following Hadoop optionsskipD D The time out mechanism could be interesting but is not really straightforward to implement so probably best to forget about it for now +I think the semantics is actually the best that we can do with the current protocols +I fixed the problem in the attached patchCiao +Would it be acceptable for these to be Windows file system paths since I am working on this issue on a machine running Windows? But using strictly Linux file paths would not pose a problem for me either as I have a machine running Linux as wellAlso I have updated the tutorial to say that it was created using JDK +It is not committed with the revision number . +thanks for checking committed to as commit ddbdfafefcdb +Thank you for your contribution +please looks to solve it with the encoding +Id also like to see something like this eventually +Else we need to chase all properties to set +The mentioned releases have already been deleted by now but I deleted the following additional releases o o Closing issue +reopened for beta issue no longer applies to JBT and JBDS +FYI I cleaned up the code tests synced with trunk removed enabled assertions again and documented the expensiveness of complex query analyzers +please post a sample application that reproduces the fixed guys! I got solution for my problemThanksSergio close this messageBecause it was really fixed in new versionThanks GraemerSergio Teixeira. +Improves status implementations for S and local filesystems +Take two +All failed with failed as last test in result and nothing else in result indicating test hung and then eventually harness killed there a fail directory with an file that could tell why the test failed +They may be covered after mapping unsigned types Java primitives +Thanks Alejandro. +The similar issue like am closing the issue +We may not actually need to add anything +So has this been fixed in by fixing? Not sure why its still in Open catch thanks + +Triggering Hudson on updated patch +NA works correctlyAndreAs to your patch I dont see any significant changes but havent tested it +The reason is to allow easy extensions of IW such as +warnings are still possible during shutdown but should be greatly reduced +I realize that the open source community relies on contributors and I should contribute more than this jira +seems very similar to +That ways we can check if the dependent app is loaded on all the nodes +Settling on the general approach now will help us avoid pouring efforts into an unacceptable direction +Id suggest that jruby community might want to spend a bit more verbiage making your license more easily interpretable by the +Hiya default logj configuration is found in the AS archive +Closing and making it solve by you. +And cygwin is explicitly required in the documentationRequirements +Sorry about that +Here is the patch for given description +Actually is more relevent to your situation +I havent looked at JCR yet if it provides something in this areaTherefore I would be very happy if you could live with the explicit hint for now +Am I missing the motivation for the change +I checked in an initial patch to the CVS +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +One thing we could for the the fault injection is to move all of it to a module by doing that we wont have to publish the instrumented artifacts +The issue has nothing to do with clustering per se +This really does diminish usefulness of Grunt significantlyThere isnt an easy way to apply the existing preprocessor here which is a shame but its an issue that comes up often +to improvingreplacing the existing class +The form id could be made available as an attribute andor scripting variable while in the scope of the form itself +Committed a fix to master +Dupe of. +So it seems that we just need to create a new instance per instance +from my side +Fantastic +Fixed or not an issue now or record the work item somewhere out of jira +To be applied to vectorization branch +Hey Ashish Good work +The issue is in find +So we have to keep them independent of each otherIf possible suggest to close this JIRAThanks +If the regions are not located on the host serving server on the first compaction they would be stored local first and stay local from there on out unless the current servers hard drive is full and it has to store it on a different server + +Thank you all for feedback +Moving this to TBD until it is clear what is requiredTom do you want this fix in EAP as well? The code changes for this issue show commits only on the EAP branch +I have no idea why theres a custom scheme nor whats the difference to the supported Codehaus schemes +why would this go in +I will mark it for so well hopefully get as much of our patching into upstream as we can before the final release. +Heres an updated patch which should address your feedback +the Patch file name to stick to convention +I believe this satisfies the request made in this issue +The pull request has been handled for a long time now so marking it resolved. +Stuart had fixed the xsd issue in commit dec and Alexey had fixed the. +The file is the case which you described in email that the entity resolver doesnt work to get the local copies of so Ant goes to the network to retrieve the +fix +Ok +Both are represented by the new Query a side change became +With these changes it would be possible for people to change configuration properties underneath the Felix instance or the activator list etc +Added the fix for the redelivery test case its true that it would fail in the case where the for loop got a random zero and didnt trigger the redelivery flag to be set +my doesnt work any more due to that enum package inside that jar +Ive taken it and made it work against +To accomodate a more rapidly changing network topology we can expire items from the cache after every hour or so +I have the same problem as Paul Chans +The forum thread that i referred in my previous update is a slightly different issue where the user is specifying in the xml unlike in this case +She will be providing the details +It appears to me you are talking about a core issue and not one in the plugin but without a sample I cant be sure +ok looks like broken DDL generation for this case +FWIW Im in favor of reverting the original change and providing an alternate API for those that want to automagically handle the case +The right way would be create a new bundle again with everything you sure I would need to create a new bundle with the sources? This is an apache project similar to and +The patch looks good if we could get some tests in that would be great +A new patch replacing introducing a sperate with its own parameters +I should clarify that this is following the CDI spec which doesnt allow Weld to look at the type of the event injection but only at the type of the event object +Fixed typos +Closing tickets +As far as your error I do belive youre seeing a problem I just dont know what can be done to reproduce it +There if a column is nullable and unique a unique index must be used in place of a constraint +Its new facility in +Just because a constraint doesnt mean a property index cant be used +Linking to as this would be a good way to calculate the bounds +Applied r. +we should expose options to set these values when creating a coreBut they should probably only be visible if in cloud mode +So the question is whether you can reverse engineer parts of it with map think this is a good solution +patch committed thanks Ben! +So I can double check my end +Thanks for the patch Jimmy +It is not clear what the term Bean used in meansb +cordovalib directory and adding the platform again +Closing for now please if there is an issue. +Implemented and added test cases +It would be nice if this patch could get in +The java dependencies are removed +And sorry for the delay +Working now on modifying arg line to accept slash escaped characters +Thanks on behalf of users for all the hard work you put into this project +A +I changed as suggest can you please test it and then close this issue +on JBT. +Best not to log too many bugs against these two features in + +can be used as an example +Could you provide a patch +UI still says Customize where it should say Configure +The message is discarded eventually +Avros primarily intended for user data +Its attached so we can get some early feedback on approach we have taken to implement the requirementsCore method have been tested +doh forgot it is a quarterly objective +Your sample project works fine now. +All capacity scheduler tests passed except which is tracked at +Do you think this circular reference creation is the design of Apache AxisWhat about your service call? whether synchronous or asynchronousIf asynchronous call then you may store all the objects in to callbacktarget rightSometimes i get the object request for allocation between MB to a big request comes to the server for can see good amount of free memory whenever we check the GC logs when out of memory problem will be the heap fragmentation because it wont have contiguous memory to allocate the requestSo can we conclude like that the two big objects are accumulating memory always whenever request comes to server and not been cleared properlyHave you created any sample service for testing purpose and tried the same to call thru apache axis engine? if yes then did u see the behavior of axis? whether creating circular reference or notAndreasCould you please reply +We had the same error +its better to start discussion in users mailing list prior creation of issue +If you could import this asap wed really appreciate it +tar should not be treated like rpm +This is only a stop gap +Will update the patch with unit test timeouts if folks think the change looks ok +The Seam detector doesnt work when the AS or ESB detector is Drools detector doesnt work when the AS detector is have tested and EAP +Do you need any further information to resolve this task +I replied to Peters email with thanksNext step wait on the results of Peters of the upgrade procedure then complete changes as agreed with Peter and Rajesh +Open new issue with any other more specific neoj problems you encounter. +I do not know how I missed that +I am running the full hive unit test suite on this patch +Now without trailing whitespaces +Not going there eitherMaybe the right action is to wait for someone to volunteer to be the guinea pig for the various filters I suppose we could advertise for volunteers +Thanks +VIOLATIONOFUNIQUEINDEX is combined with VIOLATIONOFUNIQUTECONTRAINT in +Set client state saving in +I was able to verify even on Developer StudioVersion Build id Build date +The smooks editor is no longer being developerd so moving to LATER as agreed with +For example jline made its way into theHadoopdistribution until it was excluded explicitlyAs for upgrading jlline thats a good idea as it would make features suchas reverse history search available in the ZK CLI +It will go according to the first four guidelines in description +Default value is false +I missed that +More goodies coming +I need to figure why first +Hello Santiagowould you mind also providing unit tests for the whitelist usage +The following files are automatically recognized as a welcome page and +Sure cant agree more +Could it be that there are multiple places in which converters are registeredI do not have a clear understanding about which threads are running in your application +Deployed new snapshot containing the fix. +is not really scheduled yet but we could make a release soon if its really urgent for some people who cant upgrade to right away for some reason +After thinking about it I guess my impl is not special +Simply search for and change the value reload the configuration and that mime type should be you report back if this works for you? TIA. +No comments at all +So how you are showing page etcWill be good if you attach the code +Webtools integration seems to be installed as part of Tools but Project Archives view let me create only JAR archive +nvm I missed the point this was about the branch +Please please +Added the provided zip file to the coding style guide so people can easily find and download it +Just noticed that the jira was raised for but not so here it is now +The API module has them as that module uses the latest version of JAXB which is not found in some of the older JDK distributions +It works properly in the last snapshot +If there is no significant overhead due to checksum validation we might be better off doing it in the Servlet as well to help prevent transmission of the corrupted output +Im dont have enough time to play with it but may be this is better way to get argc and argv parameters right in function isnt it +Chukwa patch for defaultbuild properties file and README fileNote No new test code changes to documentation only +I havent seen this using scpexe +Shouldnt documentation submissions be as a patch in dita format? I would like to close this Wont fix +Agreed this is not a good solution +creates the app with the added to it +containers at RM itselfBut what i meant is not just container metrics information but any other metric information +I just opened Bugzilla Bug for applicationand committed fix just for appletsCould you verify with latest CVS that everything is OKeven if you are not setting systemID explicitly +Updated patch with above changes +Add these to integration the porting of these and added them to the. +Those footers were added to the in r by Simone +modified JDBM files to add support to disable cache and accepting File as the input parameter to record manager +I removed all the printStacktraces from the code but still there are some in the Test classes +If this is the intention then a hint would helpcqlsh use UPPERKSBad Request Keyspace upperks does not exist +I am sorry but I am still not finished performance testing the patch +fixed in the latest snapshot +Yes this is for only +That solution would be perfect thanks. +Moved Hbase config constants to a different class and using them in both sinks +I updated the build script and properties to get the WC test cases and schemas +Committed to. +Thanks for your feedback this is what I needed +Id like to commit this soon if there are no objections +How are you implementing +The new indexing framework including new field indexer field extension point and field filter plugins +Reopening to edit release notes text for release notes +If not what is the error that you are seeing? patch +Sure but certain features like for instance PARALLEL keywords are ignored in local mode so certain things you can only test in parallel invironment and slicer seems to be like one of these features +I agree with MikeYou could also move derbynet to the end of the run and see if it still happens thenAnother theory is that due to the interrupt the subprocess doesnt terminate and interferes with the tests +Thanks +I would love to create a unit test but I cant determine the cause +Now that we will do a we have more time to rework the FAULT stuff in Camel to future as we do not have time to get over this in Camel committed awaiting wiki updates +classpath directories and files +Mass transition all resolved issue that did not see any further comment in the last month to closed status +tested w nightly +The issue was that until now you had to manually run roo json setup in order to add the dependency to the project classpath +Verified with JBDS +In order to use these engine right now I need to be in packageThe original idea was that youd use this by configuring protocols to use particular engines +Patrick as a matter of fact Castors support for JAXB annotations has not been finished +Well see if we can get a fix for this in that +Fixed in wrong location +The same problem apply for wars from maven repository +Ok so the Andrew File System from which Ceph draws a lot of inspiration had extensive caching and snapshots and mirrors +Maybe you could use the existing workeffort calendar as a guide +Please be sure to check any changes one more that just the showcase application to make sure the color scheme remains good +Changed the priority to blocker because the plugin tries to create a bundle activator line for every bundle now +But I just took a look at the Groovy download page and this doesnt appear to be the case +Id like this to go into trunk and not into unless its really a bug +Patch applied +See quickStart attached +Simple patch to use the task attempt id along with a unique id +Thank you +Yet to seen a case where it is broken may be after +Forgot to mention this is using +Impl not updated +Looks like is not added in but is added to srctest +I corrected the jira so the SOAP fault always send in a soap message +bulk defer to +Assigning all documentation related issues to final since they dont represent a change to the seems old and out of date closing +Please review and comment +I added two unit tests +Thanks for raising this! point +Many moons ago I needs something similar in Lotus Notes to randomly display a document from the database +probably limited to shapefile renderer? Use streaming renderer for rendering selection as an easy fix +Darrin do you have time to fit this one in +Would updating the JAXB implementation fix the issue +Looks like the subversion changes vanished from Jira they used to be in the tabs +see my comment on +seems like the wrong name for this parameter since is also used by other places +datascroller has ids generated on controls +Just noticed that it seems to be enough just deleting lib +Removed from the qaquickstarts build in revision +Is because of deprecation warnings +for the review +Various other cleanups and add test showing that Hftp is also audited +has been introduced for that purpose +the system would parse this out as table bar in namespace foo +this basically means I have to ditch that patch and start over as most of the work was actually the casts and I guess replace it with null checks instead +Issue already postet in +Here is a pass at implementing Freemarker as a form rendering engine courtesy of Active Data Technologies +Itd be great to have a debug trace of this as well +which is good all resolved issues with a status. +Sure I probably could review it today or tomorrow +Andrzej what do you think about this one in light of Emmanuels last comment? I am still uneasy about running twice but I think Emmanuel is right that semantics dont change +Here is a dataformat archetype based on component archetype from the committedMany thanks to Bilgin for the patch +Patch applied at +this patch fixes the releaseaudit warnings link +Still need to update the Connectors Developers Guide +Kevin to find out if it is possible to use the provided by portal for this +Just getting the rev onto the attachment metadata is step one +this makes sure is called after quorumpeer shuts down sync processor +I think its ready +Most concern here is to deal with hdfs down and housekeeping operationsFor I think the main problem is there is no administration daemon process or helper script for it +Javadoc warnings are unrelated +Now a certain subset of this space has been given a new semantic namely an encrypted password +How are you building your source tree? If you arent checking for errors then thats a problem +The bvt test for this has been passing. +So with java you should see little or no difference in performance from +Could you please provide a diff and ideally a few unit tests to show how this fix addresses the issue +Ive committed it +I am registering the binder as well which is the for the products class +Roy +Did the workaround they use in by casting as a chararray for now +It certainly sounds reasonable not to do something special for just the Accumulo project +Here is a simple procedure to reproduce the problem download and unzip from the Red Hat customer portal +for the patch +This issue has been fixed and released as part of release +This one should to branch +bq +Can you paste in the URL you checked so I can look into why its not updating according to youre testShane +I think that is ok +Thanks Alejandro! +Owen This ticket is ready for review +Ill check in a problem in where the IDKEY wasnt present in the Components attribute map when the id storage is in the +Dont know how were going to fix yet +Cool! Thanks Claus +BTW we will have a New Message and message editor at some then the work around is to drop a file onto an endpoint to send a message or drag a message on an endpoint to a folder to create a message file you can then use to send to other endpoints +OK so we will see our testing in RHEL and add the patch to need a thread dump or stack trace of one of those or a easy way to reproduce +jpg matches the actual extension +Rebasing and creating separate patches for MR and YARN +Jonathan thanks for testing this + changed and to package private +We have this practice to putthe mapper and reducer class used in herriot in testjar so it can be used from there +Still on track for the GA +Thanks for pointing it out +Backported to with revision +Specifically I think stacking the iterators is producing a search when we need to be using a search instead +The attached JAR works with no +Thanks for the work! Much appreciated. +Im running tests to make sure it still works as its supposed +Following up on the web services +Creates symlinks for the runs to be same when doing live status checks +Im willing to mentor or proxy +Just as an update a full compaction for me in is now taking to days vs where compaction completes in hours +Thanks for applying the patchConfirming that the fix solves my problem too closing this issue. +Can we declare this issue resolved +Initial chukwa release +Oh right +I would not think this is a bug but an expected behaviorsince the data is being updated so it would be safer to hit DB instead of getting stealdirty data from cacheto avoid this try commit tx after the agree for the select following the update but what about the next onesThe idea would begetBylabel put in cache version updateget put in cache version get get version from seems awkward to me is that the timeout is used as a bottom limit for cache useUntil you reach the timeout the cache is not used then when you reached it the cache is used! This is the opposite of a timeout behaviourIf you change the test to use a timeout of secs and then you introduce a pause of sec between two get then the cache is used for the next call. +in Revision . +If you need the whole project then let me does the stack trace ring a bell to you +The patch was applied in XBRANCH but does not appear in trunk +Please check that the patch applied as expected +For this particular application I dont think the patch numbers matter +Thanks for catching this +on latest for TRUNK +The reason I was using axis was to just give an example of using an external web service client +Thanks +Code modified as per Generics Junit test source code for logging sql this modified code for generic and logging sql generic support and logging of sql statement +The sendMessage method for binary data payload knows to remove the node from the ring however all JSON related admin methods do not execute the same logic to remove unaccessible node thus in hconsole when data and admin clients were separately created the admin client keeps hitting the off line node even when it has previously failed connecting to the further investigation I found the problem is actually in the algorithm used to pick a random node for admin requests since the initial seed server list was never updated for the client it ends up picking the offline node +should option will be removed and replaced with autoStartup in Camel not having also autoStartup at the context level since it has the startstop methods? It is somewhat convenient to be able to manage the lifecycle of whole the whole context at once +So this is being marked a duplicate of that issue +We just experienced this behavior in production and it took us hours to determine that the join was being performed incorrectly and even more time to discover the issue with our annotationsI dont think an error is necessary but a warning in the log would be helpful for diagnostics +and +However the data is the expected length +Patch for +Pushed to trunk and branches +not blocking M anymore purely sanity cleanup for title to reflect the new name of the jar +How do I set up extra Tomcat logging so it says what JSP is being compiledI am using an unmodified official release of Roller +ResendCorrects DLL importexport settings for new qpidmessaging DLLBased at qpidsrc +If Tommaso can provide a key and upload to +Only the should do this +This looks like a configuration issue and not a client problem +list is similarly broken +Issue relates to Groovy compiler and may already be fixed if not will be corrected when is noformat it works in and then shouldnt we close this issue +Looks like which is fixed in +havent tried applying this on a clean checkout but it gets the idea across +But what if i dont need the roundrobin assignment itself to be called like how we have in master startup? I just meant a similar configuration like the for eg +Trivial patch that adds for the following targets findbugs + it possible to fix it at +Sorry I didnt go through the code rightly I should double check the code by running the tests. +I guess the minimum overhead for a component is bytes header length value inclusion flagIm not seeing anything else that wouldnt let me use this as a functional replacement for the original so Im on it +Should be fixed now in SVN +attached a patch where if the env HIVECONFDIR is set its not set +I am unable to reproduce the failure using the attached test neither on Windows nor on Linux +Passing in a way to make a custom weight object makes +The patch looks good to me +moving to M +Hi JervisGuvnor works fine with SSL enabled if you use a different client like Nautilos in Fedora you can connect browse and upload files normally +Can you review and the patch it its all good to go +The request created with was wrong +hide focus in trunk +Using Before is a good idea +Fixed benchmark code +Oh and I did manually test on a secure cluster using all combinations of secureinsecure server client using kerberossimpletokens +Thanks for patch Erik +into trunk and +If at least there was no natural order of a involved then we might be able to sort our dependencies in a way where nothing gets lost but currently it is pure coincidence if it works +committed change to the trunk +Thanks for the guide will take a look and try to reproduce the problem and post it here +Lukasz any update on this? Is there still an issue? +Well try loading that and seeing if that fixes your be done over the weekend sometime +how does this compare to etc + +What browser do you useThe example works OK here with Google Chrome and Firefox +But the rationale for this decision Something is better than nothing +Verified using JBT nightly update site with build +fix checked into with the release of +I think watchingParent can get set to false when it should not. +It means when the performance is bad on my local machine due to more than consumers the performance is good on remote machine till there is also more than consumers +Covered by an IT. +Do you tell it to target a maximum number of segments in the index? If so once the index is large enough it seems like thatd force the small segments to be merged +I agree that better names should have been chosen but we are where we are and they are what they are +Committed revision srctestjavaorgapachecommonsjexlsrcmainjavaorgapachecommonsjexlinternalsrcmainjavaorgapachecommonsjexlinternalintrospection +I will incorporate your changes and resubmit the patch +Cleberts fix added to remoting. +Seems to me it would be more sensible to write your own version of that just does an FTP upload or thats what Ive ended up having to do +The actions name has been changed from touch xml files to Restart Module + for the patch +When I was working on I noticed that it would much simpler to use instead of so I prepared proper patch. +It would be nice if you will use one code style for all over the code Eclipse code style is a good thing only possible to use more long lines for example symbols +So Id recommend you stick to a single file +Bulk Close for all issues before incubation +I thought it got committed for +There are some piggybank and zebra test failures attach for the fix +Table CP are loaded by as many classloaders as there are different jar files to load them from rightWhen goes in we can build on it as needed +Owen the consensus is far from being reached +Duplicate of. +In fact we are thinking of this issue in our project +I should have fixed these when i originally added the yarn CLI but may be late now +Yonik if I instead use a doDebug boolean flag initialized in the constructor would that sufficiently satisfy you to commit this +Not that adding this before will cause any harm but it may cause some confusion + to this idea Ive seen this problem a number of times for new users +The same for the its a long chain from the browser until the request arives at Sling +The easiest way to sort out the confusing branches is probably to rearrange them in svn to match the standard layout +an example doc with non latin characters was added since the last time the tutorial was updated so the count was off +Trunk is +Hi Etienne thanks for sharing your improvements and documenting them this well +There seem to be more problems +Well have to keep around for at least another release for but Id like to deprecate it +I just committed this +Ive verified it so close this +Until then all of this is work in progress and users need to be aware of limitations that are known and being fixed +Implemented using Bugzillalogin and Bugzillapassword parameters so less RPC calls will be out to the production server +Thanks Sven I did that and it worked +Scheduled for release +Move to +If I submit a linux patch and Linus takes and it to the kernel tree and push it upstream then you will find Linus as committer and me as author +You should uses the mail list for general supportTry using the current version +The changes to make these longs were inPulling a GB segment into memory is +for this feature +Then well need to bump the version number on the branch so that the upgrade code is triggered and invalidates the old incorrect query in applications that already use head of the branch +Slava what is going on here? Why are our editor removing data +But an alternative here is to compare the number of incoming transitions with the number of executions positioned in the join these two replacements i would think it is better to remove the previous will ask Miguel if the previous properties are really needed in XPDL or wait with doing the updates till Im done with the jPDL control flow nodes +If you are interested you may follow the activities on this branch once Ive got a better handling on the impacted scope for this change Ill update the rough estimateKelven +But we try and maintain index back compatibility with bugs tooMark you are right +Long term none of this seems ideal +As you can see error message wont look like an exception but if you add debug option it will print you stacktrace +Resolving as invalid +Does anybody else want to voice an opinion here +Is there a issue related to creating a pipeline in Manifold +a test case for tagexclude and also revamped all the tests to better exercise sorting following the collapse +This version records changes made into designated table so we can undo at later date +I am using Eclipse as well but I dont get that error +Bulk moved to +bz itself on the fly +Opened to fix +Added job set to build hourly +I think what Im saying is correct +just meant generated code +We should probably comment out the documentation for now then +When will this be rolled into a release +It is returning a to your application through which you can check the status yourself +is failing due to +Thanks Ramya! +Applied patch in revision would like to change the audit logs in also to debug like the other ones or is it needed as info for someone +All issues has to be filed against from now +Thanks +Issue simulated in Metamer +I confirm this issueDetelin any idea where in the code this happens +looks good to me +A bug in automaton that only hapoens in? So its the cache there +Rebase with latest +Since this issue is resolved upstream in AS marking this resolved as ER should be rebased to on ER +it doesnt because the request may redirect to a different page +is Apache Maven version I is Apache Archiva version +FOP invokes your servlet while rendering the imageusing regular http connection and the only one who can create new session isyour +For example UbuntuUnity desabled it by defualt +Ran multiple times locally without being able to reproduce the issue +Error logs +This is a setup issue as the MS ip is not correctly blockers and critical issues are resolved but not verified +It looks like the patch for CASSANDRA might work for this issue also +Is there any possibility of this being implemented in the near future +Retargeted the issue to since is being released very shortly +Resolving as Wont Fix based on the above reasoning. +Ah and less threads also +if the repository is corrupt it stays corrupt +Reserving a name though is a breaking change +Thanks for the patch Jon. +Thanks Radim for the small change +Updated agent with j support +Thank your for applying the patch +and also in no I dont believe you can control this from Solr today maybe open a Solr issueLikewise for +Gregory still whipping + reports the same with some more transition all resolved issue that did not see any further comment in the last month to closed status +I will hopefully be doing the same with S +This also still affects the existing QMF C API which is generated fromboostiteratorqpidframingqpidmanagement +Setting fix version to +The only one thats left in the main code is the unsupported widget. +Karthik Sorry for taking so long getting to this issue +Youre never irritating me dont worryMy answers were probably not clearly explained sorry if I confused you in any way +Updated patch created against servertrunkPatch includes additional logging and fixed duplicated code by using refeshUsersGroups throughoutEncryption of user properties file is option by including the optional attribute of encryptStandard which has been added to the plan so the default file will now be encrypted on the first server startup and will encrypt Console added users as they are added or hand added entries on the next server restart +Yep fixed in +Looks like I got the fix committed to the Mina code line faster than I expected +Thank you for letting us know about this but this isnt really something we should file a bug report on +I hope you keep testing +Is it possible to rebase i them into a smaller number of cohesive patches? Id be fine reviewing a single JIRA of Miscellaneous bug fixes for python but I dont think multiple new features should be conflated into the same JIRA +Opening it to change the resolution since it was not fixed +Missed the in the above commitsvn commit m Add DISTINCT to query in srcjavaorgapachejdotckqueryapiSending srcjavaorgapachejdotckqueryapiTransmitting file dataCommitted revision +If you want to be able to do this at least until the last two property replacement tasks are complete then enable the service binding the service binding manager is enabled then it will handle the property replacement instead of the ESB service classes +So you shouldnt need to exclude it anymore as of HV +Can you verify everything works as expectedIf thats successful you can close this issue +Using the database instead of an structure for document priority assignment with seems to markedly slow down crawling at first glance as I feared +Changed to now support capturing of information including markers errors and warnings that can be later retrieved via the +I cant reproduce this on Windows Ive tried a number of variations and it allworks fine +Uploaded a new patch +Committed to trunk. +It sounds like your design may support this already? Granted its way more expensive than it needs to be +Its also difficult to implement since has tied chdirpwd to which is at odds with! +This requires a centralized controller +I could successfully deploy on snapshot Nexus repo +Still work +Agree +Applied the patch in CVS Head with little changesThanks a lot +This will work as long as classes from C that are used in this plugin stay binary compatible +here is the zip of the files modified +This patch simply changes the path of the images reference by the css +If that fails please attach the exact WAR file that doesnt run for according to jsf spec chapter a managed property needs a setter method to be initialized +It would be easy for the Spring maintainers to just bury this problem and declare its not a significant issue so I appreciate you taking it as seriously as you have been +Even that portfolio binding took me some google search to figure out +Why would we want two different when one works for both use casesSee requirements in linked bugs dev want sourcesIf we compare it to Maven I feel there is a concept of scope that is useful for dev and missed by Eclipse +ok Ill try that +Yes I forgot about that +New TCK tests for Event Conversation and added +Yes we will probably alter the file format and this will probably be able to handle those changes but it might not +In addition I added more unit testsThis is now ready I think +Then its possible to have Java DSL and use Spring XML for dynamic values where they are easier to spot and change +Claus when you say trunk now works again do you mean with? I am crashing scalac with my patches applied +My bad fixed in trunk. +will fix in both +its so you will work seamlessly with an existing hibernate session were introduced to allow users to still use the Hibernate normal ways of doing queries etcI dont think skipping that can become relevant before the classic Session interface is exposed in the interface +I think SMX have a similar but let me try to locate where it is as the project structure is a bit different than with CXF I guess +Currently it allows the relative in the attributeThanks +Right above each table of this section the line The following table contains limitations +it will be very reminiscent of the issues we have with Common and HDFS +fix seems ok to me +Pushed to +Did not notice the did not catch them +the version should be set as you wish +Sorry +For me that its not resolved cause the current way make no sense +bq +The abstract class has a protected constructor thus it doesnt seem feasible for an application to provide its own implementation of the abstract are referring to the to carry an actual instance of +I noticed that too +I have tried to follow the guidelines the test can be unpacked to srcbugsxml +Thanks to Hoss +Implemented and then secondary problems fixed in the following svn versions and then secondary problems fixed in the following svn revs +I never had the time to set up my linux box at home to redo the patch. +In the dfs do we allow different values of bytes for blocks in a file? Do we allow different block size for blocks in a file? If data are checksumed at the level there is another complexity in the dfs +Is there a reason why you do not use in? The CRC additionvadlidation code is very similar to that in Output Streams +I will have a look Enum keys obviously should be supported +duplicate of +Chris +Given the use of base Operator etc from Hive if at all there is interest and commitment we should do this ASAP when we only have a handful of operators +But it doesnt include the bpel editor +Moved this to JBAS + needs extra filter based on process instance +Works for me but still think its odd that put and delete return void while get head and post return objects +But Version is compatible with Tika +Anyway sorry for the confusion this bug report is wrong. +I onlyrealized my mistake the very last moment before posting a comment thatwould have been really embarrassing for me And someone else mightget the idea of calling the setters on that one tooLooks good to mecheers Oleg RolandThanks for taking a look at this one +Committed to and also +So I feel good about this patch againIve been following the discussion around delegation token +Somehow you are not running the Cassandra version you think you are +The only thought of change that came to mind for me in this exchange is that perhaps the double negative might be confusing +thanks for all your work on this one +Please close if satisfied +This also means I can either implement the transactional acid part of the upgrade or start working on dbcreator already +Please the case if you find the problem still exists +Will give honor of blessing it +They do effect +Sure Simone +Please verify in CR cycle and. +Will do this in hours now +Rolled over to to get a stable release m seeing this in due to an error in +log extension that was human readableIf a problem still persists please create a new bug report do not reopen this one though you can reference back to this one +See Source tab for commit that this is fixed in immutant LATEST +Editorial pass over all release notes prior to publication of. +Did you try deleting the keyspace with before running the test with +prepared for next +Modified pull request for Embedded Arquillian implementation with CDI test +The problem is that the job launcher needs to get the last job instance so that it can get the next value out of the incrementer defined for that job +Im going to close this and open other jiras for additional enhancements. +HiI am trying to send a mail to following mailing listBut mail is bouncing +and could break existing applications +For null values it just uses scale and precision +Unit tests pass +When a commit fails the method throws an exception +Are you sure that this feature was added in last night build +I have started using the head build this fix +This patch fixes this issue +bulk defer of open issues to +Patch to which solves problems with implementations in GSP pages +Alright committed thanks well followup on. +Can someone help out and commit the update to NOTICE +and unit tests in The fix is to consider the actual servlet path together with the containers context path instead of just the containers context path +Ill commit this patch and we can discuss native pipes building etc +Launch instance give it a name +its ok the author controls +Ive fixed this in accordance with Kens last suggestion turning off the undefined symbol checking for versions of cmake prior toFixed in I have confirmed that this resolved my problem +It is fine with me +Thanks +This looks like its done to me. +cvs looks okCan you please fix the complete src tarball +Issues resolved in +isdeploying gets removed and a +So should we just report another more representative metric on the region server consoleVarun +Yes I know it was written as a demo +Committed to trunk and +I think it looks better in this way +Attachment has been added with description Indent inheritance test PDF generated by fop has been added with description Indent inheritance test PDF generated by fop P open bugs to P pending further review +I installed the jar and I cant reproduce this so far +Its now available as val requestHeaders httpHeaders has been renamed to responseHeaders +Initialadaptors should only be used if there isnt a checkpoint +Grant or George do you know how long it might take for the infrastructure group to set up ASF CMS space +additional patch added +Its really a pita to get this fixed +I was reading the code wrong. +If you turn off wrapping the tabs will expand beyond the screen limits +Does this make sense? Might something like this work +Any one of them could have resolved the issueI posted the additional findings above for future readers that might have this problem so they know which releases this is working for and which it is not +I have looked at the patch and think it is good enough to commit +Is the semantics cleanup targeted for a particular release or milestone? If so it would be great if this JIRA issue could either be included in that milestone marked as depending on an upstream issue or closed as a duplicate +Thanksxiaofeng +Logged In YES useridToo bad it didnt make it into odd that thetestcases did but not the typo patch. +But the problem would still persist with rd party dialogs +btw +Bumping from unless we determine this is a critical need for the current release +thomas please post as much information as you can +N Trying to reproduce but testing the NPE fix at same time +We should put in vectorudflegacy We should put in vectorudfgenericYou can chose to do this in patch or in this one +Wow I was really surprised that just leaving off a single line will do the trick +Its not source or binary incompatible but it will break users that rely on values of SLOTSMILLISMAPS or SLOTSMILLISREDUCES +Gail can you work on these +In particular I would consider broken tests a liability +The new chapter Context Menu Preferences and Options is created in the Visual Web Tools guide describing the possibility to add Custom capabilities to JBDS JSF Struts or Seam the JSFStruts Tools Ref guides the links are added to Visual Web Tools guide. +See for further details. +The bundle currently embeds and exports the Jackrabbit API extensions +I have never tried +Stuart can this JIRA be marked resolved? Was the fix done upstream +Nor force them to jump hurdles in order to do get their job done +I should have thought of thatBryan Not inherit but maybe it could have a but then thrift is all over our application not just at the porch +See table of the which defines the allowed methods in a given context. +This is a big issue which has been bought up before and will be worked upon in Westport too late for in my allowing the user to have a per VDB this can be solved +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The issue is gone after upgrading to +ok maybe this isnt critical +Fixed HTML page has been uploaded +I added the patch in revision as proposed by Timo BoehmeThanks for the contribution to Timo and thanks to Lars for the pointer +Since I dont have an army of programmers building a metrics system like Simon Ill likely just continue doing what Im doing now using a perl script to find the log files using a regex over the directory structure and manipulate them that way +That check is on the broker side +Client starts update session Tablet A is unloaded Client sends mutation batch to tablet A which fails Tablet A is loaded Client sends mutation batch tablet A which succeeds Tablet A is unloaded Client sends mutation batch to tablet A which failsIn the above sequence the failure of batch is forgotten by the current code +I think that if we get the list of content types correctly admin should not have much need to configure that +I checked this across ext ext and ntfs +The only benefit might be if it helped to fix as well +Weston could you take a look at the merge conflict? patch was not supposed to go into the mainline build +In the proposed system is there a separate database that has to be accessed for every connection to a user database +Still getting used to this merge from trunk to branch stuff +The problem is not fixed by this patchThe expected result is the corresponding security realm shoud be removed from repository after deleting its dependent db poolBut via testing the realm folder is still there +new clients will not be able to talk to old clusters as reported by Mark Jelsma in +the plugin for archetypes is not released. +It should also be noted that if I remote the element containing the xmlencoding attribute the build will succeed +Remove conf if present? Otherwise looks great +Hello TimThe fix looks good many thanksBest regards by Andrew. +I borked the unit test will address that too +Thanks for the patch Chris! I committed both to trunk and the branch +Closing all issues created in and before which are marked resolved + +We only need to make the change specifically in the InputOutput definition to remove the KV dependency +The class youre looking for is probably srcmainjavaorgapachedirectorystudioldapbrowsercorejobsHope this helps +Do you define the environment variables MAVENHOME and MAVENHOMELOCAL also when running netbeans? or only in file? we dont scan the file and when processing dependencies in netbeans we dont have the bat env +Your mirror has been added to the rotation it should appear within an hour or twoThanks for your help +anyone can fix this bug +I need to check images in ve reproduced this bug on Hudson in nightly build of Dev +No worries +Committed to trunk and +Let me a take a jab at this +This patch is for +Me and Dushshantha has gone through the full code base some time before and found some little problems +In the status valid changes are purposely maintained in separate entity to facilitate the business owner to decide the changes in status for custom requirementsTherefore it is advised to not to validate the status changes at code level by hard coding let to decide the valid changes in status of any object +The settings arent used generally very well as reported by Brian Fox so Ill take that one. +for this awesome patch Peter! I will take a look +Openwire strings are marshalled with a string type header that indicates whether the string is a Big String type + rearranged the codes in the all tests locally and manually tested the script +Problem solved moving to Sun JDK. +Once deployed I understand that the JVM would take precedence in the classpath so now that I understand that this is an optional dependency on JVM it all makes sense +Need to migrate the proxy to use standard SAAJ steps will ensure that combo will be Axis free +Looks fine otherwise +about changes Nick suggestedalso happens as designed +The only thing thats not done at deployment or runtime is scanning for +Thanks Jacques +It should be possible to use a test just like the one in the description but adding a new broker to the cluster just after creating the pauper and then consuming from the new broker. +new MD SHA ip clearance document already committed to incubator SVN but attaching it for the record +Thanks Jingkei +the same is no guarantee that data has been written to disk +I think we should close this issue and target the workspace import of protected items for another release +These changes have already been checked in +The solution is to add additional folder containing file to src directory +Reopened to set the Fix Versions field to JDO beta. +This keeps the focus on Cluster Summary which is probably more useful than Startup Progress during normal operation +how can i do this Jay Balunas? can you make one simple example +other than the one in the description +Creating a schema that wont work isnt what I can except +Attaching screenshot +I tried to fix this by forcing the database to be grown but this has to be done explicitly +Hi NickI just checked the code of there is noCan you double check the version of camelBTW a patch could be more precise to discrib the question +Thanks! +we hit this issue today as wellAttached was a thread dump during hung ocurredIve filed to resolve to qtp was not a daemon thread our hbase shutdownhook was not be Liang +done +or not allowed to view +Good points Jan and Benoit +Those keywords should be flagged for inclusion in an index in case we ever get around to the index +Ignore the Hudson comment due to my messup of Jira numbers in the commit of +I just committed this +A patch for version of the patch for AlexeiWell doneI believe it fixes many potential defects than this JIRA exposedJust for you information that RI is not compatible backward since the patch fails on RI but passes on RI +This would lead to problems in the following case A app has an that never changes and is never closed but creates lots of for caches +Personally I am on the fix if you run a lot of deletes and cant cache your index files in ram then dont disable bloom filters +Committed to trunk thanks Marek +Load balancing +Teiid accepts any column name that doesnt contain the +Uploaded a rebased patch after a patch that adds the requested tests code changes and documentationbq +Serializable MarkusIm not understanding this issue +Then please ask on the ml and well give you the Ah well also need your osgeo user id +Im also happy to provide the patch if youd like to assign the ticket to me +Next Id link to the new repo containing the artifacts once served by the old repo +Thanks Liyin +Fixed. +Tim Ive applied the patch +It matches incase of nd and rd attempts +Max no just skip that for now for the general case +Should I check it into the tag as well +It works +Not sure if it still does this +Ive just committed this to the branch +wont there be conditions w this version where out will be hit in the while loop so get returns true so it will be called again doing another seek? so where i was going with this is if we hit out in the while loop we really need to return false so get doesnt get called again +Close after release +Fixed security branch to use absolute path for setup scripts +So we can copy or rewrite as appropriate +I think we should probably provide minimal refactoring support and forego the more complex stuffI dont think this should be a priority for us since we are encouraging use of Forge IDE of choice +This is a top priority one for us as well +Committed revision Thanks Mck and Noble +So for all the stats that we get for the same URI we will be adding the stats across the stats and exposing it up +HiPlease find the passthrough get processor enhanced to generate wsdlxsdCheers applied. +Agree with Matteo +As it is now a second thread calling into originating from call to can yield a deadlock +Committed revision x Committed revision +It should create the appropriate directories even if no localization is going to happen +We just need to be able to load a different comparatorYes thats it we permit a different comparator to break the stream of sorted keys into calls to reduce +The same exception occurs when try to any page in the same way +Sure no probs Ill apply tomorrow need to complete some Atom logging workcheers Cyrille yes please go ahead with this update sorry it was my mistakeSergey +Committed merging the trunk into the branch from to +brilliant jayI tried updating the by following your links but without adding SET NOCOUNT ON I thought it did not fix the problemNow it works on my side too +So mechanism for wont work on +Again thanks for the quick answer +Closing. +Latest Axis supports OASIS catalogs and the Geronimo integration work will be done as part of. +Its definitely not ideal but worked with a production based public cloud Overall the driver needs a lot of loveSebastien +Anyone have any new thoughts +If could build Ambari then it could use it as a test framework easily as wellWe have this on our roadmap to do next quarter unless its done elsewhere first +It checks the BCHM behaviour matches the example in academic paper +This patch removed the implementation of file level locks from HDFS +I just realized that the p and the a are joining to form an infinity sign +Were going to try to add the good parts of this patchset to the approach already committed. +Help in resolution is appreciated. +Because I dont think that a fix for will be available +Same issue with +I would say it is safe to exclude the generated man pages from the +Mmm reading fast it seems to me that I just changed the type of the issue from improvement to bug so no should not be close until working fine for me if its not please write steps to reproduce +I removed the fix version +Btw OT but I could use some help with +Also will upload a patch with zero return for standby so that this patch can be committed +Im off on vacation for a week so wont be able to take a look at this till I get back +Thanks for the May take up to hours for the fix to reach central repo +Let groovyc simply set that from outside information can be put on the for informationconsistency sake if that is what you are asking for but the main pieces will remain as is for the reasons explainedSo no open objections remaining right +Removed the from the gem spec file and tried gem list +to report passes on local +Ok I started a job with these parameters +Thank you Doug + +Looks good +Added test resolving for now. + +I will update the patch with this change +What do you guys have in the way of regression testing this sort of stuff? I realize you have unit tests in your code but Id look for more end to end testing to verify this does not break the various corners of Axis +Ive modified the test case with a proper locale refactored the Decoders a little so that can also be specified with a locale +Adding patch files that fix issue +Anyway examples are using RF and also should do that too +It also includes java sources +Thanks for the pointer to Jesse +svn statusM javaengineorgapachederbyimplsqlcatalogM javaengineorgapachederbyimplsqlcatalogM javaengineorgapachederbyimpldbM javaengineorgapachederbyimplstoreaccessM javaengineorgapachederbyimplstorerawxactM javaengineorgapachederbyimplstorerawM javaengineorgapachederbyiapistoreaccessM javaengineorgapachederbyiapistorerawM javaengineorgapachederbyiapireferenceM javaengineorgapachederbydatabaseM javaengineorgapachederbycatalogM javaengineorgapachederbylocM javatestingorgapachederbyTestingfunctionTeststestsstoretestsA javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreA javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTestsmasterA javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestssuitesM javatestingorgapachederbyTestingfunctionTestsutil +both build and discovery is released +May somebody pls try this +Thats good to know +Very helpful for folks like myself who are familiar with load balancing in modjk and the plugin but are new to modproxy and information in the attached screen shot is similar to what is displayed on the page +I dont think this would be viewed by the PPC as is higher in the hierarchy than the module context +Merged patch with latest trunk +ChrisThese changes look great +Im not able to reproduce the issue with the srcmainjavahudsonmodel Jenkins file. +for the new patch +If its only returning top N is it based on score? What is considered high score? A row with most number of hits in the box? Cloest to the center +To verify the bugCheck if ip rule is consistent across the after boot upAssign multiple public ip ranges to VR check if ip rule shows the same resultRemove some additional public nics from VR check ip rule again +Please add comments on top of the patch once Ive checked it inI think that your comments will be more useful than mine +Ok it seems that we can do a javaxml mapping but not a javacsvAnd while testing I discovered that Im unable to map in th Apply Template task from a Java Mapping task if I do a xmljavacsv transformation +solution is very easy +Committed +I also thought about this before i did the changes +Stefan Do you have a test case that doesnt rely on JDOMAlso have you tried your test against +Sorry about this +HiI am very interesting by this featureI prefer the native namespace of PHP because a use a modern framework which have an function based on the namespaceI am ready to help for that +has been marked as a duplicate of this bug +I dont think it does +is a much more invasive solution since it requires code changes +iBiblio hosts the older iBatis jars although they have changed location +Hi AllSome enhancements I am considering at the moment include Support for XSL based route dispatch support Support for Drools based route dispatch supportI will add these in due courseCheersAshwin +Thanks +Unless you want to take a look at the IT we whipped up and see if you can. +Unless you force them to use something like Hoop +Xuting could you please take a look s fine for me +Ah yep +Tests passed cleanly for me +I think people are starting to want implementations due to the fact that more standards are being released which use geometry as a building agree totally about both naming issues +bulk move of all unresolved issues from to +I rerun both and nserverdemo example +I sent of a mail to the apache directory project +Tool is MapReduce applicationOk we can write standard Java application with main method and run it from bin can do this work but only after June sounds right +I have cloned the Bug because it still doesnt work for me even though my file is configured correctlyCan someone please helpThaks in advance +Can you put the second code snippet into a test and provide a patch? added a bunch of tests +Yes that is the case +Hence reopening it +The third argument specifies the pad character that is to be removed +i just want to make sure it fails reliably on the hudson machine +You are welcome should have added a bit more explanations right from the beginning +How to reproduce the Scrolling bug of Input Configuration Panel +again with the latest patch to and +Its great to see a patch that fixes metrics +Added forum link. +this rings the bell did you search the mailing list archives? I am almost sure this has came up before +Committed to trunk thanks Carl! Ive marked this as only for now +Justin could you please verify? Thanks +an experimental s patch fixes a rd kind of bug itself can return duplicate rows even for a single node and range +Catching OOM is the wrong solution +Oh forget the dev thing +Yes it works with is strange as in the changes was around are you sure it used work in +Please could you verify that your patch has been applied as expected +Committed revision +Correct behavior works. +Figure out any obvious bottlenecks in the scheduling code +I am also on using native calls over CRT functions +Where unable to reproduce the error so i have to assume its on my end. +Right you are +Its a valid case +Dont print a ton of debug it looks like both climt and does this +I just pushed this to security and mapred trunk +In fact the reason is after the st AM is killed and before the nd AM attempt is created keeps querying RM for the AM hostport and RM responds back st AMs hostport +If the default heap allocated is larged than the allocated container the memory check would kick in and fail the test +comparing the and it seemed the main difference was that the +May be fixed but marking incomplete since its old weve rewritten Tempfile and there have been no or. +Whatever solution you come up with the js client will have to decide where to put the response based on the response it gets from the server +It has been committed so just SVN up and test outfix the shell +The offending methods aregetIteratorgetComparatorgetget +Still same in EPP after fix from +hurry though as we are about to release +Is this still an issue with new CL Shutdown policy GC +Yoram and I discussed over IM and here is current approach +Verified on EAP ER. +The three html files are for review +Fixed to be set in the mapper +It seems the entire is disabled in trunk +Hm +Make sense to put it there +Thanks to Pavel Stetsuk. +r +Is this going to marked as a bug? It is definetly not working for me +Please test it if it works +Its that the error youd get if you tried to make two method calls with the same was concurrency related rather than state related +Patch applied to r +Need to get this out to shell or into bash scripts +I guess the most tricky thing is to avoid messing up parameters in the mojo configuration when combining both tools in a single execution +Added a PR that changes the repository explorers POM to use a version property for the GWT plugin +Aaron could you please take a look +We should make a patch for trunk too +Posting a patch now while I think about how to do so +Actually changing the method name did not solve my original problem so the thing may not actually be an issue +Now I think of it unassignTask does not need to throw an exception when it already is unassigned but for consistency it might. +OK dont worry jansi is ASL licensed jline BSD as RenYour patch is in trunk at r +Your code has N threads all listening for connect events on the same server socket +Using hfile tool I confirmed that the keys in the two files are ordered +Thanks Dag +Hi Luciano my connection lost when i try upload file I got just now +Trunk patch same as patch? Let me check that assertion failure +reopening for verification of this fromThis is an issue again because the annotation was moved from to in Seam Betathis specific class rename isnt in your list of problems so just reopening this to see if it was missed or just not listed +bulk assign all unscheduled issues to next release. +Thanks Ill have a look +Logs from starting persistent unit from AS and M +Note that Shale is retired and probably should be excluded from the deleting the data seems bad +Ive searching for another file but didnt fine one yet +Im quite keen on this idea +Moving this bug out of this fix would mean refactoring code +But the Entity component doesnt care about that its just using the Entity annotated class and processing the Id annotation to find which getter or field represents the id +Ah thats a pretty stupid mistake on my part +Looking a little further I found that the member failovertrue so if it is explicitly set off then the multicast discovery transport works more like I expectedBut then that still leave to question what would the purpose of failover mode be with used of multicast discovery +reopened to correct the on Tools +Closing all resolved issues for. +I have no way to do so +The local variable bytesTransferred in record the same as reduce +Maybe it had been fixed +Now it uses the user provided pk if it is not present it falls back to the solr schema guess the latest patch can be committed and this issue can be guess the fix was not you have a failing test case +Yes currently the main issue with Jacoco is that it lacks integration in Jenkins +I added the additional tests in +The counterpart of for mdb +According to the specification the constructor should throw for invalid patterns +I would vote for adding the txid to each record +The would let me delete the last row and the on the rows would insert a blank row after the current row +This also conflicts when using because then you dont even have a separate maven install +Fixed for next release +Hi Carsten could you take a quick look at Retos patch for the scr plugin +Commited to both trunk and branch +As you said problem was with our code only +One of the fields is a cumulative which makes sense for a single threaded scenario but not so much when you aggregate multiple m going to to show my current preference for a summary per thread using the same structures output format that it currently uses and then an overall process summary which sums various fields and drop others that dont make sense to me in that context then see what think +Check the plugin documentation. +This does not include the remaining inessential change from +Was already fixed on on trunk I just needed to flip the line width optimization did so now +Can you allow ? Our mirror monitoring happens from several locations so it would be difficult to give you a consistent set of IP addresses +All that is required is the email address +Fix javadoc issue in v patch +project content using Eclipse API before project is imported into workspace and +Yes OP meant physically appending to the block +Thanks Todd! +Now the table tag will release references to the collection of object IF this has been set using the name attribute at the end of the that this is not done when using the deprecated list attribute since as per jsp specs we cant remove anything has been set via a tag attribute +I implemented the missing methods and unit tests for them +So we should then close this +Simple logo with Cup File Console and Apache feather Font used AsapPlease share your commentsThanks really like especially the square on the left +I updated the dependency on ecj to and the test still fails +Please review +the patch was hunting this bug? after applying both fixes i still get a big memory leak when doing custom sorting but i cant find the cause of this leak +Our tests doesnt fail with any exceptions even when the drl file is empty +Hi MamtaThanks for looking at my tests and your comments +Thanks +This seems more of an api deficiency and we should investigate it in a new bug +Added patch Fixes the issue +We need lease recovery as a +Any sample to recreate this issue +And thanks to you Bryan for taking this one to completion +Example database URL jdbcsqlserverlocalhosttestresponseBufferingadaptive +If anyone cares down the line we can make it me know if you still see them after picking up a dev build in a day or so +If we abort RS when state is in state during transition we could abort RS too easily +How does introducing make it harder to use? Can you give an example +I am changing the summary to reflect the actual issue +Ill try your other suggestions later today +v patch +Indeed +It is also not just Chrome but safari shows the same +I didnt notice when I created this issue. +The method will throw an with the message bar cannot be null +We may need to what questions are asked that do not have nice defaults there are some many steps questions in the current wizard that it is relatively easy for the newbie to fall off the trail +Please check that it was applied as expected +Hey Ram Thanks for reading it throughbq +patches +The PR applied to the and branch +Minor change +JiriI checked the following problemMODIFY Reduce Input methods Processing Task section Input methods Simple X input Method +Could we make this an easier experience for new users by having the MR AM double check that the is configured at application submission time? This is a hack irrespective of whether we do this in the AM or the clientUnfortunately even that is insufficient +It will be great fun to replace these classesimport import import import import import import import import +The background of the link is changed to when the debug window is not openedAdditionally I changed the color of the prefix in the debug window ERROR is red now and INFO is blue. +Committed revision merged to for +Burt please attach your changes to jeebooking example in patch format to allow me to be able to run the example on geronimo +As operations only holds a key with the operation name +good point! +Uploading a fix +You might see unmatched message acks on the broker with the older ones +Decision for this is made in +I think a JIRA issue should be represented by a single commitThis issue pertains to making properly handle surrogates when lowercasing +Will Trinidad work with JSF RI since in future Seam will embrace JSF RI is the advantage of Trinidad compares to ajaxjsf +I have implemented this functionality in Rev +Thanks Jarek Im also thinking about how to make it right +What do you guys want to do regarding the above two options +FYI The that is published by blueprint as an service can be used to parse file +The User Task exposes properties to set the actorId description etc +I added the methods but took the liberty to change the implementations a bit +Yes we can just do the ifelse statements for those +That golden file is up to date +Review for release + old date issues. +David assigning to you as per last weeks WFK meeting +Please excuse my brevity +I tried to look at this issue on the latest AS code base but blocked me +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content Services +and idealy provide a way for someone to write a custome auth module. +Note this is assigned to Tuscany now as the OASIS xsd does have the required extensibility it just nedds to be added properly in Tuscany for the Tuscany JMS wire +They are both running Solaris and kernel patch +It is clear simple and the separation between distance and point is really an interesting featureThe only remarks I have concern the naming of the Clusterable interface methods +Attachment has been added with description The for doing this +ThoughtsEither this yes or we add tabs to the relevant views +Committed to master. +Unit test cases for and the translation from to Note +Thanks Amareshwari +nor does it know the table so it cant cheat and look them up globally +modifing the filter chain +it is a very serious bug it would be fixed in? why not +This has not been fixed in the CP branch + +I havent yet seen demand for this and IMO better reported as an error condition +Attaching a quickstart for +You need to use p to use the patches supplied +Just committed this to trunk +Building the roadmap for am not against writing specifc deployment descriptor but we need to think a bit about a sar with a connector is valid +But the number you have given is too large +I check in the code and when it is updated the build will fail +Our scenario of usage is as followsWe delete specific columns in a specific row +whenever it wakes up and there is positive credit on the link and no messages to send +I have verified that the plans generated for this query with the existing patch on look fine +the point is you dont have to name your JSP pages with a + +As far as I can tell PHP has no support for connection pooling or reuse +Forgot to mention I think the interface stability should be set to Unstable for now +Identified as blocker for testing integration +Fixed svn rev +It was determined that the problem is not that the exception is thrown after upgrade but rather that it the exception is not thrown before upgrade +I wont commit anything until i hear back from you +Moved back to to in trunk +I think that plugins should work independent and should work on both windows and unix systems +Well thats not good communication is the key +We also need to be careful about version dev when installing from a local directory instead of downloading. +Hopefully this clears it upll try to get those numbers for you they should be in our logs and if not its easy enough to put them +Removed all compilation warnings Improved readability of code Corrected spelling mistakes in class names and methodsIm certain there is scope to continue these improvements further but it has reached the point where these would be better carried out while defect fixing etc +deploy the sample war file to geronimo on sample +what specifically does that mean? Do you have an example of code that you are trying to execute with batch that you cant? +terminate or destroy is a better word choice than kill +I like the call to unassign if FAILEDOPEN before moving on +Something like +Already tested in +I would expectthat errors are important enough that they should be offered to thelogging system hence I changed to default toLOGERROR +probably because patch missed a hunk +If we look at the jdo file we have persistent properties and in the orm file we have persistent properties +My experience with this is a little lacklusteras could any developer willing to help improve this plugin +fixed +In particular the code in is effectively duplicated +Cancelling patch +Could you please verify that it is o Dtest testRunning run Failures Errors Skipped Time elapsed could you please have a look at the test case that I created +pushed to both for setting proper fix proper fix version +Attachment has been added with description PDF result after applying the updated patch takes a different approach first let the default loop do its work check if the gathered exceeds the available IPD and if wrapping should be forced process each character as a separate to let the sort it outThis effectively limits the effects only to those words that would not fit in the line as a whole +Example expected message +Try one more thing for me +Will try the workaround for may be of use to some of youMy plan was to use branch to denote a specific release stream +Committed revision to close for release +This is a know error in Domino httpsupportuid could certainly provide more information in the error message +you can add Jboss Web Service facet to your project and then you will have a chance to change the runtime for generating Web Service in other words you dont have to keep the consistent between project target server runtime and required runtime for Web Service project so you can still use the dependence of and use another runtime or higher Server location +the patch i sent to the MLBut we should decide where we want the output to go in the solr src dist and also if we should run it on the lucene CHANGES too and consolidate it in some way +worked the best out of the box as it supported it without any code changes neededThis was fixed between the original release attempt and the redo so its part of +Access granted +next idea as patch +Ive updated karaf to use this latest version +Would you please try the patchBest regardsspark patch applied at revision r thanks a lot for this enhancement please verify it is as your expected and any plan about the actual implementation of +I will keep adding markers when I find something that could be fixed or improved +Seems this UT has been modified a lotThanks +Good idea! Ill update the patch +I am currently running some more tests on it so I did not yet commit this to svn +Thank you very much for these patches +I will debug and tell you the resultThanks +This patch does not apply cleanly to the current implementation +Looks like the new parameter introduced is and not Comments The is not updated to include this parameter +Alright that sounds like a then what if we could do it in a way which was backwards compatible but required changing the way the scripts are called if a user want to use new tools that are introduced +Committed trunk and branch +is actually being used in the new module and working. +I just deleted the mentioned project and it +I am also curious about what modifications will be required to other so that they can interact with both the WF AM and also the RM directly +Attached warnings +Closing resolved issues. +Currently we are able to detect and skip the entries but not to extract them properly Ill change the issues subject to reflect that. +I had set the Commands directory to be JAVAHOME +But I wanted to know if this same bug exists in Seam +I updated this to properly reflect how it worked previously +Im not sure if they are effected by dynamic metaprogramming +This patch restores the behavior of by checking to see if custom insert SQL has been defined and if so using it instead of the generated SQL +Addressed Jitendras comments +This patch handles the len case +Im going to merge this to the branch as well +mrepositoryorgslfjUserstucu +The patch doesnt seem to apply anymore but it is still an important issue to resolve some day +A fix for the trunk will follow +In worst case will Ignore again in case of a problem by which we would not be able to handlefix +What do you propose I do specifically? Change the netty version in AS across the board? Or just pretend the jar is a +Lukas so far the mobile compatibility targets Webkit based browsers +Initial patch for review that includes the following changesa include as new dependenyb set scope of dependency to xerces to test onlyc change build +Juergen +So the wiki itself doesnt know the permissions +attached is oneI think it looks kinda dumb but if its useful Ill commit it +Patch to fix the issueThe fix employs a technique similar to that of the pipeline processor to preserve the exchange type +Committed to. + on the changes +Blah +Inactive pools can only be seen via virsh all but they will still block you from creating new pools that conflict +Those tests cant be used as integration tests theyll fail with Messaging as the implementation differs +I think this was a dud +This has been raised to the OSOA spec collaboration +A fix is in progress +Yeah sorry +The code does not exist in ASF repo +Not a bug in EAP +Id suggest to leave the targets inside the distro +All other things being equal the simpler patch seems better to meso I think is therefore preferable toI dont believe any of the properties in the properties file areneeded by the test so I think we can move ahead with this approach +Attached the patch please check and apply. +I have tested it on Indigo and JBT trunk and everything also seems to be goodWhat do you think + +In the mean time the contextual menu in the Express Console provides a Web browser action which uses the application url displayed in the properties view. +Files are now using geronimo package names by default +I got to think about that +it looks good to me +It wont make it into the build I am currently creating but it will for the next reopen if you continue to see problems after this makes it into a build. +net version for compatibility with Java using different data types +Works great +Voted up! +Well fix this through overriding putAll for a guarantee that delegation to put will happen independent from the implementation strategy in re right I am on IBM sorry for the sloppy report +So if the scanner sees a +From an end user perrspective it is still unexpected behaviour +One last comment for people who run into this later +Looks like PR was not applyed +Its probably a premature optimization +bq +I play with a few settings +but thats probably more than i can go after at this point for the +svn commit r syncopebranchesXconsolesrcmainresources close for +is the source for a bundle that stores the encrypted auth credentials in a cookie +Now the get methodes return an empty enumeration instead of null +The tests test the and as well +If measuring indexing what kind of analyzer token filter chain is used? If measuring queries what kind of query types are in the mix with which relative frequencies? You may want to experiment with modifyingcommentinguncommenting various parts of the driver setup for any given target scenario +I updated my scripts to handle these situations +I think we should wait until after is in and the tests to be sure +mvnsar support not intended to be in JBDS M reevaluate for per my fix above it should be in JBDS in trunk +It would be nice if we could get a full stack dump on failure here +Never mind +definitely deceiving +Moved to please move bugs back to which you will work on in the next weeks +So its ok with me to remove the extras. +I will the tests info after our discussion +I agree with Eli that it is better to add an Admin API then adding them to +I was just taking a look at the two issues that block this issue +With test it and let you know +Thanks John One minor note the patch will have to be rebased against latest codebase +Cant reproduce this bug +The reflect API already suffers from known performance problems +community archetypes should be encouraged to not inherit from right +ill then do a search to see if there is any example codetests etc that can be simplified +Owen! Ive committed it +a simple patch for the issue +Changes look good to me + +Dominik can you take a look at these tests on windows +Added extra patch which also supports IE +What is the manual step you took to fix this issue? I thought the should have mapped it to the files thats checked in as part of the spring jar files +We dont have any distributed environmentresources for it so its preferable the single PC test of the task to be enough for its completionBest regardsMiroslav s when passing in it still displays after that probably a thread issue +What if we just added a new wrapper class that acted as a proxy around another OK I like this solution +Here I am waiting all week to get and then you go ahead and get it +release audit failures are in html +These properties are no longer there in after. +You can ensure that for and raise an ExceptionI dont like the hibernateid thing +Thanks! +Id like to have some getters on the factory for manual exploration of the settings +Thanks for the review Vnay and Daryn +pass for you? If so lets commit and figure it out later? You want me to commit the fourth draft +Good point +This is very odd +I hope Im on the right lines +Submit my patch +Bug has been marked as a duplicate of this bug +Some news of including this patch in the next release? the have a corresponding issue like this one? +mapbuilder will consider the unique but the wms will ignore the extra paramaters in the URL string +The security team has made zero changes to the JACC stack or anything in +Worse a server restart is required after fixing the path if redeploying to the same context +Create a database with and disconnect +ScottYes that should be the way +Works fine for me as well +Patrik Hunt been waiting from someone to have a look at the patch for a little bit +However with this issue has been fixedMoreover it seems Simon had several issues with his prototype which let me think thats probably not a good idea to spend more time on it just to refactor the code +These patches let the mock objects handle cookies +configure localstatedirvar prefixusrlocal mandirusrlocalman infodirusrlocalinfo buildand its really not compiled in +Use the mailing list to make sure people get at least an occasion to see your the has been created by a developer that is no more there and it may well be no one of the current developers know about the class +JornYes I can provide some test data +We do not have many special usecases where it is different to get the localized message +Execution log of the attached script showing the full list of unrecognized public for the assignement of this issue it looks like creating it by cloning was not the best way to do it +I also noticed the namespace problem +I cant actually load the video to test this +Either in a follow on jira or in here +I tested with and all the tests in contrib passed +HDFS test JAR was not including the webapps directory +Netty jar that is causing problems +Looks like this was introduced by +Of course this fails on the second try since the file is not checked out patch simply adds a check to see if the POM is in the list already +We will consider the request +the property of the manifest sounds like a good first step and shouldnt have any +For example a udf could collect several bags +In the case of a down bookie this will cause s of log messages +I could not reproduce with the application from the archetype +Hi This is a clean setup with zero data to start with +Clarification changes are not yet committed +batch transition to closed remaining resolved bugs +Maybe this is another example +The added profile works for me We might want to run tck against JDK to find other potential issues. +So template creation from this snapshot would fail with NPE since the snapshot entry with storerole as Image is not present in snapshotstoreref table +Submitted this patch +DaniThis looks really cool! A few questionsIs the goal here to fold this into a contribplugin dir? If so would committing this require the inclusion the codelibs or can we use ivymaven +It also failed in other and appears to be flaky +Not a blocker for +Ive tried to get this error but its not reproducing for meIf I see it again I can reopen and attempt further research +Just add a comment saying that youre wanting to look into it insteadThanksVincent +When set inherittrue it inherits the jsps that referenced in the parent controller class and working in the context of the inherting controller class. +What was the reason for replacing the +Thoughts +Document in Fusion Sections +clustering doesnt work +We could iterate uncompressed messages +it seems +Committed to TRUNK +Xerces handles include files well +same with released version +Please could anyone review this +errorDir defaults to input dir so we need to ignore error files too +Trunk At revision The proposed solution is not what is done +Im on X +Fixed by the addition of new transformers. +thanks for reporting +Couldnt reproduce +The problem scenario described by this issue is now fixed so Im marking this issue as resolved in and +I recommended that Bradley file this bug report after a bit of conversation on IRC +Commit the workaround codes At revision So it seems that it is better to do this in Jettys codesDo we need to open a JIRA to Jetty any commentThanks +As it uses a background thread to update the load +snapshot is actually not a directory and just used to represent a snapshot path +I have the same problem +So that we are backwards compatibleBut it would be good if changing the hardcoded UID to some other string is possible via the configuration fileWe are happy to receive patches for the sameThanks! think this will be backward compatible since in the past only openldap directories using the uid attribute to store the user login were working +Ah I see I didnt fully grok the new preRollpostRoll dance in my first review +This appears adequate for practical usebut it should be noted that this performance is from CPU cache to CPU cacheThe implementation still needs quite a bit of work +I am not using lazynoproxy because I am looking for the performance benefits associated with using proxies +But Im pretty sure its a problem in something +An example is attached +got the same problemFedora Core attached patch should fix this issue +Why should folks be tied to Xcode any more than theyre tied to Visual StudioIf youre writing the server in node it might as well be an HTTP server so we can extend the function easily and then brogrammers can use it from code via XHR as well +Once we tighten things up this bug showed up +Is the markup ready for this? If not is there a jira for it +Fixed in revision Added property +This patch implements the assertions above +Committed initial use of the plugin as revision in branch +This has been causing conditions for users who pick such long volume pathsI doubt that the directory scanner is the cause of OOM error +RegardsRobertPs +It looks like the for each of these files is already around in memory but the Groovy Resolver is trying to reparse each of these rather than use what is already around +OK +JacopoIve been using this patch all day without problems +Sorry but this is the way its supposed to work as per the SpecificationJava Bean specification Capitalization of inferred namesWhen we use design patterns to infer a property or event name we need to decide what rulesto follow for capitalizing the inferred name +Attaching and with the change to the Using topic +So we are a goI should now untar the attachment and import to incubatorlucytrunkimport +We shouldnt just repackage without deprecating it first since it is a public interfaceMakes sense Tom +Already added ARM Java Language nt added yet because of internal server errors still checking up the problem +Thanks Raghu +See the wiki or code for to +checked skin parameters for panel fixed some errors +Probably want to put a! null check around the writeLock +Sounds interesting +The fix looks reasonable to me +This allows the text to be indexed as a text field and all queries including phrases will work as normal +Making the Derby JMX automatically available in the server will make it impossible for the user to make some application with an embedded Derby db manageable thorugh JMX without also making Derby manageable thorugh JMX +The concept of interceptors might fit into that nicely +Have another think about it it isnt hard to change an issue from Resolved to Closed if that is the final transition you aim for +Ill bring up something on the dev list +The default config didnt work for me last I tried it but admittedly that was long ago and Ive been copying this config over ever since +If the problem still exists in the new release please reopen the issue. +Im on to it +The is exported by JDK the package has no version range if we install other JAXWS API bundle Im not sure the platform can help us to resolve it +Is the patch relevant for trunk as well or is this because the backport didnt include the tests +I viewed peak memory usage of around MB so set Xmx to MB +reopening temporariliy so i change the resolution to Wont Fix since we didnt actually fix this +Tested with branch +So Id recommend that once you switch from one log to another that you unlink the previous one when you get the chance before using it again just to make sure you dont get any ghost writes showing up later +From the comments I see you guys are working on this bug +This should be fixed in the current versionI get the same results from batik and fop for the svg rendering +Moving to Beta first version which community should start usingtesting is in my opinion beta +If I change the request message to use lower case element names the service works fine +And I would call this instance on updating the metadata every time a bug not a next attempt to work around this problem will be to update all the to use something else than the default skin. +I was the original reporter so Ill go ahead and CLOSE it. +Anna can you try when is completed +Attaching the files I mentioned in the description of this issue +As you can see at the end of my previous post even this jira site work incorrectly with UTF chars tooInstead of chars ? there was russian chars Seems that many developers use only ASCII chars and do not warry about peoples in other half of the World +Thats correct +Thanks Brock! +Then we can keep the issue open and discuss what additionalthings should be moved to the module LICENSENOTICE? I know i have been polluting the lucene one heavily from analyzer poaching +Maybe we can leave a comment there and fix it later on +Then it will also get the context mapper and message composer should be completed first +Why was thatIn any case the correct way to set up our default permissions for a newproject seems to be right now create a group for the projects developers if it doesnt exist yet add developers to that group create a permission scheme by cloning Old Default Permissions +I just had a chat with the guys behind and they would love to help us reuse the Puppet recipes they wrote +Thanks Klaus I had to upgrade but now it seems to be working wellPerhaps we could change this task to make this the main official wiki for +That is tracked in. +Sure but if is better than then surely is better than Basically why and not some configurable option +A few nits to fix DEFAULTAPPLICATIONTYPELENGTH should simply be called APPLICATIONTYPELENGTH? All the changes in lack standard formatting It can use the default constant instead of it +A patch will be provided FYI the way you are using urlencode wont work with all the supported Python version +Dont give attention to my last comment +if you want a beta testeur for the fix Im point the patch fixes treemaps and timemachine matrix but we didnt detect that the problem also occurs in the chart +And is a good start point. +I think we should disable Transforms from the list is this JIRA subsumed by +Ted I will change the category on commit +I will wait a couple of days to see if anyone encounters a problem and then mark this as resolved +Thanks +Since we try to avoid working with AST because it would critically slow down CDI tools I implement comparing of simple names only and add test that will fail because of coincidence of simple names in different interfaces +an images checksum is stored in memory and persistent on disk in VERSION +New cumulative patch with rework from review attached +Sounds good +test removed files attached previously because of errors in the patch +Done +cannot see why we do the unassign +If you disable these you risk things might not work as you expect +I will have a look at it this evening +Mikhail I think it fair in cases like this where a bunch of the code base is touched that us frontier folk more familiar w trunk pitch in +BTW I just tried to resolve this issue and assigned myself to resolve as Cannot ReproduceWell +has been renamed the +I attached a possible patch for this as a comment to +Its just not as noisy +for the patch for branch +Along the lines of zfs rollback or netapp snap restore functionality +I committed this to branch. +Letting transaction proxy definitions check their transaction attributes is not a bad idea +fixed in +We could try falling back to the id if the name is +Maybe if Curtis or someone else could this resolution the issue can be closed out +Assume B WLOG +Ive raised a ticket for all graphics in the Security Guide +Correcting +Doesnt make too much sense to keep a Jira open for releases +I am working on this. +Ill double check but I think this code goes away once we move to commonslogging +Thanks Ivan +so first I set the level to minor since there is a work around and the work around is the intended I set it to wont fix as we all agreed on that. +Load a corrupted avro file withoutignore +Im not sure the jaccard index makes sense in the context of spelling correction since order isnt captured +Am I missing somethingMuch of this SHOULD be fixed for and I think a little more for SNAPSHOT +Tested on Eclipse on Linux +Hi Guillaume for me this bug duplicates and relates to +on patch +The war should also contain the sources +artifact is now available on the revision +Hi Florian we recently added a fix for this issue +point into another issue +Hence later versions should also use +the rules are there to provide a healthy repository for everybody to use if you dont want to contribute fine but we cant do the work for all the open source projects out there +I guess over the next few days I will have to get the same working for unmarshalling +Actually this should be an easy enough fix to make + +I will confirm +Here is an idea +Some comments not directly related to the patch Can the errorMessage in be improved? When the application paths are not specified correctly it just says invalid parameter which is confusing +If this should instead be considered a bug and that the messages should be a global then that can be done relatively is definitely a design choice see this is clearly an issue currently hooks must be nullary functions +I used as this is the encoding dictated by the JRE for +I just committed this +Thanks +Olivier Can you please take a look at and let me know if you see any problems with it +bq +Closing resolved issues. +Cfoomaster and Cfoo attached patch fixes this problem +Sending email to both lists doesnt decreases response time +Comments are inlinedThere are two major ones +Ben I suspect all of these are obsolete but please take a look and keep openretarget any of them that are still useful to youthanksdave +Once ESB supports this functionality we can look at adding it in the editor. +So far for a couple of years I havent come across a situation where with Wicket I had to use a session token +I sure hope this is the correct way to submit a patch +I looked at Ruby Gemsbut I could not find new what means can I experience the reflection of your effort? +I suppose we might as well go to at this point which this patch should be able to handle pretty easily +Youre right the regex is better but probably the chars are ok as this mimics what will split onWorking on this made me wonder if perhaps itself could be made more useful for languages if it was configurable to breakcombine withon characters other than the space +Username sjcpete +Sorry I shouldnt have mixed to issues here lets discuss the digests thing inMy concern about an explanations page about binaries hasnt been addressed so far +Thanks IlyaPatch applied to ANNOTATION module at repo revision rPlease check that the patch was applied as you expected +Commited a change in rev so that the VC facet macros and put method are used only with VC since the standard appears to be correctly implemented in VC aka aka +I will need to get a resource very familiar with the deployer and classloading to lay out the groundwork for this refactoring +All these properties can now be specified via the connection URL and some of them are used in various test cases. +I just verified this again with Roo master as of Git commit faadfdebcdbecbdbbfa and it correctly handles spaces when using tab completion. +You may need to remove the part where the props are changedI think that only the classpath and the are has some unprintable characters in it +anyone tried this? Should I go ahead and commit? the patch looks good +only. +Ted Addressed your comments in the new patch +Attached patch is in synchrony with the latest sources +OK is would be even better +for Wont Fix +this also broke the cassandra kitting by forcing the managed version of! I was looking into fixing the deployment and was not sure why it was occurring +thanx yanick! +I took the idea and rebuilt the patch against latest code +I find this incredibly hard to believe +Marking this bug as resolved +Oh I should also mention that Im running on AMD +Im going to keep this issue open until I can set the correct Fix Version. +I think we could theoretically release sometime soon + rebuild of Lucene indexes is needed +Im looking for instances or classes that come from Groovy or that are around but probably shouldnt be + +Please dont let the many lines of new code distract you +to verify for Beta +Put line breaks to long argument list for readability +Can you attach a simple test case? If so ill fix the problem for next release +Make it default value for page size was set to to adopt for parameter for disabling rdbms idnexing +In which version is this issue fixed +bq +SvenPlease attach the project files to this JIRA ticket +Verified on Hudson. +Please confirm the current status of this issue and if its incorrect change it as required +Quote from my sandbox commitPatch changes touse a main memory database for some tests for increased speed +trunk +Do we need a separate guide possibly a common tasks guide of some sort +Attached MS logs for all management servers during on the exception looks like some setup issue +Hey James This seems like a nice change +Maybe we need a +But I cant find any pdf having an issue with it +If wsdl changed you should regenerate code. +It that required for an existing generic udf? but regardless whatever ill do it +dims +Henry you do need an instance of QCM to get the listener but you dont need the listener because youre using LE +Thanks again for the quick feedback DanYou are right I modelled SYSFILEPERMS on SYSROUTINEPERMS +so heres the thing +Or as Ali mentions you could patch it for this featureAlso note Jetty differs from and for cargo another patch was requiredId recommend using jetty or higher with cargos jettyx config as the wonky stuff with JSP config was resolvedkind SVN rev +Im not sure if this is due to major compaction but you can see that major compaction took place several hundred times +It does not appear to be the major issue though +I believe hudson is failing to apply the patch because its a MAPREDUCE ticket which doesnt touch the directory +Looks like an easy tweak when you get back Bob +It is normally for passing variables to the task +I can see this is done properly in filter so I guess I just did not pay the proper attention to the code so thanks for fixing itFYI a redirect uri parameter is of no use at the moment Ive been planning to build some support for the authentication may be around itThis was only committed on +null pointer exception when no measurements are being done in some cases +will send the patches together with other fixes +Included in release. +However P is a client not a participant maybe P is not removed then? We need a unit test for this +It only creates certain events that will be processed by the IW once the DW action is done +part of release. +On the patch I will look more closely but why cant we blindly try to split if the cant be found in ZK +With the and you are already using the to dynamically render the content +modifying documentation only no code changesrun site and verified correct generation +We need to backport these changes to as the test is failing after went in +QA should try to find out what xenserver patch is faulty and open a ticket with Xen +Updated in the ESB codebase will be in next merge +Committed revision +This will be fixed as part of +Heres an updated patch which takes care of that +The SOAP message itself may not contain all fields given in the request uri if they do not follow the schema +This must be done for every segment +Changing the schema is a big deal +Looks good +However I dont like forcing people to do the transformation outside either +No adverse reaction post release of so closing. +think current updated patch can solve +Patch looks good +I added code to to strip underscored properties other than rev and that fixed the problem +Here are the email addresses for themtbeerboweryusakuYou now have moderators for each list +This changes the iteration to use entrySet instead of keySet +Not an INFO level issue +In order to use the persistence for continuations you have to make sure that everything your want to suspendcontinue is serializableIf not you can still use Continuations but you cant use the persistence for the stackIf I misunderstood your issue can you provide a test case for it +So any unannotated fields and setters will not result in references +I was hacking on this against +I just committed this + Perhaps it would be nice to use it in future official versions? From a technical POV there is IMHO no reason to switch +Thanks for reviews guys +Introducing a third option might increase the level of confusion +Yuri please open a new issue +The vhosts show up as mbeans in a pattern ofhostxxxtypeHostand this has an alias property +If its an unrelated column then the name on the order by item will not be set and the metadata form of the name will be used +Another solution might be to check the Items returned by get first whenever a child item is needed +Finally I found how to add attachedI have removed all dll get from conversion along with pdfbox +I just need to rebase and retest on my jenkins setup +Thanks Jarcec! +As we are week away from the release of do you think that you can resolve the issue on timeIt is absolutly not a problem if you cant just let me know as we can move the issue to a lot +Merged attached pull request in master +With the patch applied this is no longer the caseI have attached the performance test results +The dispatcher will deliver the event to an idle thread in the thread pool where is called do the real work +Hi allas far as I could see all old Tomcat links have been corrected nowHowever the main issue here are references to +Regarding our previous communications you have replied regarding changing cardinality but you have not yet replied to this query +Fix is in SVN +Switched of corresponding warnings in eclipse as it also moans about catch blocks of which need to be catched +So it can overwrite only closed files +logging lgtm +PeteThere arent any new releases planned for AS +Will commit when tests pass +Arup Malakar recently checked this in to enable Federation capability +It looks fine and I agree that use doesnt make any sense for aCheers worries David your concern was valid per the original comments +IMO its nice for people who are not familiar with EDI to see the descriptive names however later on one get to used to segment names and their usage in XML is more natural and conciseAnother problemi is group names right now they are called or what would be more meaningful names for them +fix it +If we take the as a canonical ejb server these files indeed should be deleted +Since currentbuffer is null this wont add a buffer to the dirty buffer listAround line switch is called because someone tries to get a chunk of log in a case where there are no buffers in the dirty list +Im not too familiar with the underbelly of Cordova +Would that be better? Then youd have one job to maintain instead of two +I am running regression tests nowChanges are mostly of these types Correcting capitalization and punctuation +The NPE appeared inThe problem is fixed in +This is definitely a problem does not occur with a version of that I downloaded from the Grails Hudson server on November + +Currently I have no local changes in the according packages that would need to be mergedthx in advance in trunk +Well this message is an access token +Would you mind reassigning this JIRA to me and Ill put a patch? Alejandro +When will we know the verdict of this task? Is it based on a certain number of votes by a cut off date +If everything runs fine then I will go ahead and commit your patch alongwith my changes tomorrow +Thanks Ben +The answer from httplinuxgeowebcacheservicewmsSERVICEWMSVERSIONREQUESTgetcapabilitiesTILED I found that I used not the latest version of udig my link to the program was wrongWith Version it works as expectedThanks for your effords +Ill look into this issue +Try Hudson again +Forgot about this +Good Tool. +Since the test passes when running alone It may also be related with some garbage that couldnt be cleaned up after +To run use mvn initializeThe second project uses the telnet task but since the first project is executed first by default from the modules ordering the telnet task hasnt been loaded +VPE can recognize templates in current project or could be directed to folder where user stores how hard would this be? I could definitly use this for the eclipsecon talk +Ant do you think you could investigate this problem? Thanks. +a patch against is more then welcome +Im looking into this at the momentWe cannot use static initialisation because in general there is no way to know when such a variable will be initialised +Reupload to see if it triggers applies almost cleanup to +As far as I know we just need TLD and SCI scanning for comsunmain +Update patch after +Would you mind filing a separate JIRA with a fix version of for creating a validator pool to improve the performance +ThorstenId suggest you provide a patch for this first and attach it to this issue +Thats what I thought too +But if i use enctypemultipart request type portlet not working in provide the alternate for this and Regards +Hi again Chrisfinally got some time to start working on this and I would appreciate some advice from you on this +I think if we have a nice sort of chain of processing then you could slot in your authentication and authorization initially and not have to wait for us to handle the problem +existing +And can you retry from command line +It will entail changes in and +Verified on ER +I tried your suggestion this time i am using AS Final +Please reopen if you disagree. +thats a very weird bug +If you clone the array its all good although it does put some load on the garbage collector but not that much since the arrays just end up in short term storage and almost certainly wont have reference cycles +Real clients may or may not want to keep the connection alive +The patch looks good +Ivan the changes look good to me + +Cant apply +First shot is bold visible enough or do we need some other style of marker +on jbdevstudio. +oh ok +If frozen the node will stay in place when the layout is refreshed or changed +Proper handling of the service is very important +Trivial patch I went with List since we seemed to be going from List Array only to go back to List in a lot of cases +Renames the to DecoderEncoder to be inline with java API +Here is an updated patch which is completeIts already too large and since it doesnt modify existing behaviour I think it can go in to unblock other patches while I add more unit tests via an +This issue is related to issue. +I had some singleton service holding a collection of allowed assets in mind +Resubmit the patch to ask Hadoop QA run through it +See the attached patch file +Thanks Himanshu for the review +Could you run ant test and ant manually +Attachment has been added with description First set of tests for the has been added with description Second set of tests for the has been added with description Third set of tests for the set of tests currently fail with a Class Cast Applied +Provide the unit test which reports the bug +We need to send that back as a header and check that on the consumer side so we know its a fault +bq +You should be able to run it by just running java jar +Annotation properties can only be of type String Enumeration Class Annotation primitives or arrays of these types +If the table is not in use anymore it should be deleted from the database. +What do you think +The base implementation does not open files only examines their lengths +can you speed up the rest of the skipping tooWell the code I posted is slower so Im afraid that if history is any guide here I can only promise pessimization. +I meant Running the test with clients applying load I get the Too many open files exceptions on the the server after a few minutes +Minor comment Would it be better to throw an wrapped with your message so that we can avoid the subsequent +Its called +Lets park this until +DER is the wrong format +Im for the file system shutdown hooksIm also for +I would highly appreciate it if somebody could look into this +This is a known limitation with how implements the data model +Ill commit this momentarily +Im having a little bit a hard time when exactly that is triggered +Keeping references to may not be enough +That said the issue description really is wrong +I do not think it needs any link to the bug which solves padding of hexadecimal entities and processing of uppercase exceptions in +Soaris instructions for running product are still missing +As an aside for some reason I cant add Jaeho as a contributor in JIRA so that I can assign the issue to him +moves the configs to solrconfigscollection in the tests which config to load is discovered from solrcollectionscollectionconfigcollection system property for the name of the collection to work with consolidated zookeeper host and solr path sys properties into one ie localhostsolrI still expect everything in this patch to be very fluid and change as we move forward but its something to give us a base to play withWe should probably start a branch since this issue is likely to get quite large and hopefully have many contributors that model has worked quite well with the flexible indexing issue in Lucene and I have gotten quite handy at quick merging from my practice there +patch attachedTry to help users who specified a command without the mandatory leading dash +I dont think we want a reference to BEA at all +Can you parse the file succesfully with Tika stand alone? Is the file larger than the value set in +Why that is I cannot tell yet +Since we dont have async in Java the unit test have to be in unit test +Ive been thinking about introducing some new API allowing rd party adopters to define their own rules for enabling Java EE facets +I think my impl was simpler +diff was partially applied already by Mickael Istria but did not include updating the +Will post it soon +Changing object space configuration at runtime is discouraged because it wasnt supposed to work that way reliably and because that actually change the objects id calculation which is really unsafe thing to I dont understand is why you really need to enable objectspace for Nokogiri? Is there something in there that requires to be on? +Fixed all I could find +At this point I have ported the patch to all branches from forward +Need to remove from all the example code if that is the case though. +By moving the escaped quote after r the plugin works +Still cannot reproduce this bug on my machine +Ive improved the patch +I created simple JTA test with global on and it works well +This could be a significant performance savings for distributed deployments where we are currently wasting time serializing and echoing the entire CAS contents +break backward compatibility is too heavy +How could both project A and project B build concurrently? That is the sign of the that Project A depends on Version of B but while B is building it removes all local artifacts trying to remove too but A is using it +Is it because we can have a race and the output of wait is not reliable? Check the latest patch I have uploaded to see the approach Im thinking aboutOther than this issue the patch you submitted looks good +Maybe I miss some you attach screenshots with expaneded source menu items +Ive removed the outer loop thx for that +We should discuss if the refactoring should provide an optional rename of the MikeI personally like it that my files have the same name of the classes +patch tests for points and +I can confirm that both approaches work for me on Chromium using Linux +Is this issue resolved +Also tests cj and af fail due to this issue +Looks like this is working after going from RC to +Vinod we have +I think this is quite likely to be sufficient to fix the bug +Is it possible to use several with? As I understand the problem is that the same job is loaded by several and one job is executed several time +The usual workflow is for the developer to run Roo and check the generatedupdated files into version control at which point the project is indistinguishable from any other Java project. +Bryan thanks for your sample applicationI am adding unit tests based upon your work +Diveshthanks for your detailed explanation of the bugIn made a fix in rBilginPS the reason for the initial change in was to prevent the overriding of the service results from each iteration +Exception Log. +Thats very interesting +Fixed for and +Another use case for Equinox Transforms +I did migrate the security provider to felix as of revision +Done many moons ago. +egIf we have a pool of generators that could contain generators of a particulartype and threads it the theoretical numbe of open file handles or If it is it makes sense that this is where our problem is and the ops guyshave got too many components in the pool andor too many threads +But Im not satisfied that weve covered all current use cases we should +Applies and runs tests cleanly in my environment +We didnt update the sample database in the content directory and so it wasnt omitted from the patch +So keep that in mind if we are tempted to multithread streaming +We use +It can be declared now as completedThanksMarco +they both point to the same function but theyre not actually aliases +igore me +Feel free to submit a documentation patch Ricahrd in the meantime Im marking this issue as resolved +Yes that is weird + Move clientFinalizer definition to the top Delayed file initializationcreation until first log event do a chmod on the file in order to give read access to Chukwa +Thanks for taking a look Grant +Isnt that the same for a column delete with many versions that span many blocks though? In that case the would need to hold on to all prefix markers +Applied thank you! +added restart link to console +IMO here Where ODMG is missing a feature that castor has its worth keeping the feature in the castor way but where ODMG has a spec and a set of capabilities which cover the same syntax in Castor OQL IMO +Are we wanting to do compression for a struct schema or do we want to do something? For example are you suggesting something like Megastore or are suggesting something like JSON format where we can do optimizations like dictionary hashing +Not approved for EAP +with Colms suggestionsCheck it please +Additional patch to applied after to allow reflect data to be used with mapreduceI havent got the mapreduce tests for avro working in my development environment so they still needs to be verified and extended to test reflect data as well +Although this has been marked as a duplicate was actually a little different and Id recommend considering the discussion in it. +It still doesnt work +Still mulling over this problem unfortunately its not as easy as just removing a jar from our install +Fix in CVS +At least thats what Im reading from the comments +Scottjust curious +Thanks committed as fbcedacfdbecfa. +Patch integrated to trunk. +This is definitely one of those quircks we can do without +Hi Ravi can you reproduce this reliably? Can you write a test to exercise this case consistently +Reading via works nicely in this case +Thanks for the review +Hi Michel we really cant manage to reproduce this issue + +Next week a colleague will execute more tests on other documents with complex layout and will check some corner cases +some refactoring to remove warnings added a download file to local directory action added an edit server location action +TODO Add in the wiki +The reserved slot should dissapear +Let me know what your take is +I totally agree that we need pack for trunk and branchIm working on a fix that would exclude +Linking to because this bug was discovered while investigating that other issue +Would be nice if it could be fixed for Derekthank you for your starting from version XT Ajax do and do calls have the concept of request parameters they could be the right candidate for adding a form do you thinkCheersSergio B +Applied Amys patch to trunk at +Just let me know +Resolving patch is attached and review request submitted +Rampart now generates Encrypted headers as defined in WSS SOAP Message Security specification +This should fix the issue +Any progress update +User must use browse button now +I pushed a couple minor changes and dont think there are any more test warnings +Seems to be fixed in +this was an issue when the root most entity had a rootEntityfalsecommited close for Solr +Thats what Ive committed now the info passed to the exception will be examined for with password and passwords will be masked automatically +Integrated with console issues for release +Adding a change that was missed in previous version of the patch +Well provide this in Spring unless there is a this problem only seems to occur with Hibernate beta snapshots +Its not a partial graph its the circular dependency where the first module in the loop is put as +Applied patch by MckThanks Mck! +In the client you can use the logjPath constant while setting up the localResourceChanged this JIRA to and reuseWe can the looks good +Were you using that test or did you try the unit test I providedDo you see any differences between the configuration of my brokers vs +I have not have time to Write Tests Figure out how to check it into the trunk I would like to work on this asap it would help with just a few dummy steps how and from where to check out the file check it back in +JasonI just put this into trunk and. +Do you have a test for this as I would probably refactor some bits slightly and would want to make sure everything is still good + +Good stuffIn general the aggregator needs to be followed by an action that pieces the aggregated messages back together in some meaningful way before sending it on +Please review and if looks fine please revision for this issue with patch for the commit Kathey +Reassigned because Noel currently its not active in the project +Yes everything is fine +Ill expand on that in a new commentbq +I think we can fix this failure the same way +I dont agree that this pattern causes confusionThis could be a simple enhancement but both of you are trying to make it unnecessarily complicated +I will that +OK makes sense will do in this for CXF AndreiI applied your patch and also rebuild exception handling on some further +Fixed rev +Thanks for fixing. +is open to figure out what to do about has. +use to get the layout you need +ChandanPl do verify this onethanks please test with the latest build +String is not yet solved for core TRQ +Then you plugin came into the picture +The test is still failing expecting result but getting issue out as part of JIRA cleanup +Actually on manual application the whitespace errors are easy enough to fix +I think the missing unit test was another brain fart of mine when using the subclipse eclipse plugin +Attachment has been added with description Patch for this bug +We should also make sure we dont unnecessarily serialize return values if were running in replicated mode +Updating the patch to fix the test by substituting a different policy +Thanks Uwe +The following are not fixed +Patch that applies on latest code base +So I added system property then mode test passed. +Of course index must go into the new methods as parameter too +Yes Hiberante now drops the schema first and then create so it will throw these errors +Thanks Nuwan for the contribution. +I have a potential fix for the WSRP issue but I cannot test it completely because the deployment fails anyway +This is only on the latest master +If absolute standards compliance is required then Id say that theotherAttributes attribute is necessary +not even clustered +Sweet thanks! +and +hung in the above build +All but two errors now resolved +Removed link to as that JIRA is closed +It makes sense to move this to builtin rather as Cheolsoo proposed +There must be a bug somewhere +Committed to trunk and +Thanks John! +Did you find a solution to this issueThanks have not a solution +We are therefore currently unable to publish artifacts to Maven Central +Nope youll need to wait for is getting released? +s inv +Remember the tests are essentially in a random order +This looks great +I wanted to bring this up because it hasnt been mentioned yet and its currently a topic of discussion on the list for query results of composite columns are you going to deserialize the column name or leave it as an opaque blob? For the current implementation of composite columns in Hector the type information for dynamic composites is encoded in the name but that information is lacking for the static variety +Perhaps you can check if that issues was fixed already otherwise please this issue with more details +Mail example also throws the above exception after sending an email +Agree it is somewhat strange and probably should be in +This is fixed since call sites are used in almost every case now to do invocation and they obey visibility. +Thanks Jeremy committed in rev +Hmmm I still cant +Max sounds fine to me honestlyOk +Thanks karthiga +Thank you for your comment +Looks like it just happened again in the latest build +And having an option to select default categories for variants. +The directory is not even created +The +You give it a Recommender implementation and it wraps it up in a deployable JEE web app +is the first test that fails it works fine when run stand alone + nuke default charset shit +Andy This used to be done at r but it was removed after your later changes +Patch to fix the pyunit tests run via maven +contains patches that provides an alternate implementation that uses servlets bound to the usergroup resource types instead of to perform the usermanager actions +Latest CVS prints out an Invalid request error +our model is more elaborate then the portal model +All unit tests pass +Its still open and not resolved so I suspect that we still have to look at it in the next release +This issue has been fixed and released as part of release +state the class is in javadocIf you like replace instance variable with instance variable in issue description but I still dont see any justification for using anything transaction related +They have been taking the approach that the code was not licensed for this use but also giving IP as the reason for their concern +Reopenreclose to deal with migration bug. +Forgot to ask the favor of having a look also at generated findbugscheckstyle reports and fix potential errorsviolations in classes in the package you addedTIA +Another task may be to fix failures at current tags +Hi Srinivas Could you attach a code sample for this bug? I havent been able to reproduce itThe original description for this bug seems a bit different than the comment you added was the behavior you described in the comment causing this issue? Thanks Chad +Maybe try those changes locally and see if it makes a difference +Building packages and things like that should be easy +I finally had occasion to look at this in a multicore setting and the extra field caused the schema browser problems and more importantly caused an Exception to be thrown in the when trying to output schema information for that extra multicore field +Please close this issue if it is ok for you +Also if you have Google installed on your phone and select to auto upload the photos they dont get saved in the Gallery +Committed to as well. +In fact I dare say that n should not be for the EOL character in any of Groovys Text Processing +Issue done over to you +I will remove it +Maybe platform and environment specific? I am running Fedora xb java version desktop environment m installing Java to see if thats the issue then Ill try and reproduce on a Linux VM +As a mail from heiko said now we only support jboss AS I change the error message to select AS because of the reason now we can not support tomcat +buddhaThe release for Nutch has some of the features for Nutch but it is not a complete Nutch Architecture +include new files +I should have attached the deployment archives in the first placeAs far as how to exercise the example first send the then wait seconds and send the +We can modify a little bit if necessary and check in the template +for the explanation Marco +do you have an example? I dont think Ive ever excluded a goal before +Browser firefox +The currenltly used cglib version is +Ive just added the patch to solve that problem only in case if parameter is set to true +Vadim and I apparently do not speak about the same thingI speak about the whitespace around the XML tags and theXSP XML tags that will be part of the produced documentVadim speaks about the whitespace around the Java code textI would like to get the first type removed +Im working on this but Im not smart enough to assign this issue to myself +Ill keep everyoneposted if I find the time to do this +Im no longer quite so sure that my problem description is correct +I submit it for review +Hey VishalLooking at the patch at a glance it looks good to me +To be honest nobody agreed either there were no responses +Thanks for reviewing and testing +It didnt dawn on me at the time but there are already configuration properties for changing the packages to be searched for default handlers for both stream and content handlers +Seems like theres nothing we can do about this hence Im closing this issue for the time being +Between M and M the auth code was refactored and theres no Windows support there +Fixing Affects and Fix in Version +This is the modified file +Still remaining is a fix to to update the captured environment when transforming locals to fields +The attached jms notification spec shows the type of information available in part of JMS payload and JMS header +So we either didnt support this yet or what it nowcalls first doesnt take care of this so I think this problem is inthe code ported to cxfEdell +Whatever is left over at the end is considered the last token +DearsPlease recheck againI have fixed the issue as you requestedBR Zareh +Haunt me if I got it wrong again Roman +It seems that we only added this check to make some old TCK happy +The fix was to close the Solr instances at task completion. +for in rev +Forgot to set the Fix Version to new Jetspeed security model no longer loads user credentials automatically. +I guess this was lost in translationSIll update the patch +LeonThank you for submitting the patch but there are more problems that need to be fixed +Running tests and committing +Too late to fix for so slip to an alternative implementation we could think about having the job first mirroring locally the repositories they depend on before they start up and use a to use the local mirror instead of the remote benefit is that once stuff is mirrored locally it wont be removed later +It replaces the navigation xdoc file +Thanks Knut +Covers all properties notimplementing +! Thanks! I also ran with mvn clean install and it looks good +change file for this +Step These image was then opened in the COTS toolStep The Image in the Native Raster was then exported to Tif using the same COTS tools MSS Image was Imported to a native raster format of a COTS tool + +So this is a Galaxy Tab that hasnt been upgraded to Gingerbread +additional renames in commit +Thanks Jason +Also point to a lib directory +Looks like when I reran the command it started working all fine +do you have a sample project to reproduce to problem +closing due to lack of response. +I think this is finally nailed. +You could also simply set the size in the constructor +The can probably be substituted with minor modification +Still I dont like nulls let me try something elseRe We should open another JIRA for that TODO +Thanks Ning +Then we can add this import command and just close this issue +Checked in by James +Hi SamisaI believe this was a problem when throwing exceptions between libraries and the error string goes out of bounds causing another different error to be thrown +Patch committed to trunk and branch. +I also rename the references inCommited revisions r ll keep that in mind for the following patches thanks +I doubt Ill have time to tackle this in near term. +Well open new issue when we get a more complaint. +Theres no need for a separate locking mechanism on the filesystem we already have a lock in zookeeper and a lock on the port +ASF guidelines are that patches may only be accepted via JIRA not via Git +Investigation turned out that xalan musthave gotten very confused about its internal datastructures somewherein the process +for the patch. +Applied patch with thanks to Edge +and stubs complete too asking for testing resourcesLast connector we need to figure out is Meridio +Also would it be useful to set a main class for the JAR +I reverted thechange to Cocoon SVNThese icons were in use at Cocoon in the days and Forrest adopted them +Setting Fix Version to future for issues without a set fix version +Well AFAIR the button worked back in +patched after latest rollback. +Well I thought it was better to fix the wrong results and look at optimizations later on as a separate bug +This fixes the issue +I have checked this into trunk +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +However need to add it to repository on what exists for future are tests as part of BVT +Committed the attached patch in svn . +If theres no workaround we should at the very least catch the exception so that the job does not die +the Parser uses perhaps the worlds most mind numbing technology +Would be nice to addressthis in a future release if the issue is confirmed since we went ahead anddeprecated Template +For adding a pause in put after we put and before we cycle on the gets waiting five seconds makes it pass + +Tejas and HenryYes you are right +Now as to the fat jar issue +Please verify +Sorry about that bit of a case I didnt think of +Jim Im sure that will be a bigger issue for you with XA datasources than this issue +fixed with using which can use +Let me know if something is wrong +Everything after the comment characters isignored +Will wait a few hours if nack on before committing there too +for the ported patch +It wasnt needed for the manual delimiting that I supported in the first phase +Fixing to match PHP version +for being an impl of +Burr im not following how this is different fromAnd which perspective are you in since I dont get the same list as you do in perspective +will investigate the problem more +Until then manually adding a trimrtrim function in your teiid sql or using the trimstrings setting would be the suggested workaround +ran into an unresolvable implementation issue concerning the JSR +We need to make this explicit in the files +What you want is to diagnose intersect as invalid +Will add a unit test later + should discuss this on the ML +Roo is built using in order to provide extra fields and methods that are required but are not core to the business logic of your application +It makes the a lot easier +Shelly any reason why dont we shipenable async logging for production env +The user workflow to deploy DI enriched process definitions currently is ridiculously painfulSo big for Sebastians point and suggested minimal fix +Im also willing to perform tutorials to determine their usability and help ensure that tutorial directions are appropriately descriptive +Im betting that whatever calls to get access to that scratch file is probably buggy and maintaining references where the called has fallen out of scope +Just noting that running with an insane build dosent result in any error neither in the insert nor in a subsequent select +Pushed to CR I will try to replicate and make sure that its fixed when CR ships +Lukas is this issue superseded by the inIf not does the placeholder component play nice with the calendar defaultLabel and validation +As we are not waiting for the root location to come up we get NPE now. +issue is resolved as fixed as the user said he modified both and +instances may not be shared by multiple threads regardless whether the implementation itself is thread safe or not as they may contain safe objects such as andEach thread of execution should have a dedicated instanceOleg + is SQL State Code for syntax error or access rule violation +Do when applying patches in future N. +This would pull in trunk level dependencies or higher +the checkbox should be on the other does the fix on fix the problem? If so we can close this issue as duplicated +Navis +Committed revision . +indeed it I suppose that is easier than what I thought was needed but I think its still more awkward than it could be +CVS works for meexport CVSROOTpserveranonymousscmcvs co it thanks +The glob rewrite makes it claim the entire path doesnt exist which it didnt do in and earlier +Ill start by extracting a class out of I want to use the functionality there with generics +Username sreich +Are you sure it is still requiredIf so please let us know where in the current Admin and Config Guide it should go +Why do you want those classes serializableOleg +New version +Moving remaining bullets to separate issues +Ive linked the issues +I have no idea how this is possible since all of my changes are in +This should be removed when a new version of drools is new issue has been raised to update drools and check this issue is still fixed. +Will commit +updated README with proposed changes on r +Right after submitting Ive noticed I wasnt looking at the real patch only Christians you +is open as a general deal better with running out of space ticket but we can open more specific ones +It is very easy to do +Thanks Nick +That work has not been done and there are other bugs cataloging it +For instance do you need to disable all of the triggers on a table? Do you need to disable individual triggersYou might want to discuss this issue on the list + go for itLooks like there are a few different types of things fixed so commits per action would be good if possible +They are discussing it +Seems to me we need to do the following in your casesa return the object to the queue and decrement the internal processing countb decrement internal processing countc remove the latch from the queue +I think having a json writeput interface might be good for a wizard to build a things with buttons +One note on the patch Like the jar plugin it expects to build ears with classifiers as attachments to the ear wout the classifier +Awesome! Thank you very much. +The part of the patch had typos in it +The patch fixes the bugs found in the following JirasThanksKeith +I am not resolving this yet since your original problem has not been solved +maybe Ive misunderstood what you are saying +I need to check into some things first +So the following call with the dash as parameter becomes incorrect +Leaving the JIRA In Progress to reflect the fact a further test case will be added as protection against future breakage +NOTE this patch was generated for the release +Attached the client and the master thread dumps +This will be fixed in Riftsaw which I think will not make it into SOA +i will fix it +Our tests didnt catch this since all of them use package statements and this problem was hidden +Does this issue make the QE testing extra difficultIf both of these are answered No then I agree with Jason and we should move it out since it is not a regression and we are in the end cycles of is expected +If you add a url mapping which maps to a controller action your filters should workI think the uri pattern does currently support ant style paths like and demo +It would be great to get this in + +we have column for info like app start time? If so we can build filter on it so that we can list jobs running within hour day etcWe can do that in the web service +The svn client on brutus is too old to work with the checkout though first time Ive seen that happen. +I think that this patch needs to hold off until the Web service Control builds successfully +This was an issue when using the DEBUG flag to compile an X program that contains some forms of objects +Using a filter like mime type for deciding whether or not to store the attachments in the gziped form seems a good idea to me +Galder I was also getting this exception very often +in the set are these cyclic dependencies an issue? Commons Collections is not modularized anyway and the bundle contains all the packages +so you say the bug is still there +Is there a good EPL statement use case for this +I implemented a simple chat server with MINA for experimentingImagine a user shouts and you want that to write output to every other users session but you dont want it to write output to users that are executing a character creation workflow and are not actually in the game yet +That chunk of code actually appears multiple time in the file +One of them uses an the other one uses +No idea why this is happening + +I applied the patch to +It can be a little bit dangerous to kill threads but I think it is possible. +Replied to Mark Littles email with comments about the community supplied issue. +Max according your suggestions I look at it is suitable to fix the issueI remember you told what it has some drawbacks with performance for DB with many tables as I see this problem could also be in xml code complition somay be it will be right to use here and create new issue to fixworkaround performance problem +option just disappears too subtle for rv option s colours good contrast on the orange and for rv option s shadowhalo to give some depth make it pop a little for rv option s metallic shinesmudge makes the wrench more realisticId still prefer the to be larger +Daniel can you please attach a quickstart that demonstrates the problem +For now Id suggest using working sets as a workaround +bulk close of wont fix unreproduced and duplicate issues +The needs all the versions tags that were in the original patch to compile correctly +Chuan thank you for the patch. +With Forge tooling the runtime is is not external its bundled inside as an osgi bundled +This is still susceptible for spoofing unless cryptographically signed IMHO the solution is to encrypt the whole as soon as you need to authenticate to be able to send the first request I assume its quite unlikely to have somebody spoofing a cookie +will commit +This change is covered by the change in. +Suggestion taken uploading new patch +Yes the strings arent but the behaviour is predictable and most importantly safe for the DB +Broken patch obsolete once generics is available. +Just tried out of curiosity +gotta run off to a meeting Uwe I wonder if you could test the attached patch +Trace would be more suitable IMO +Sorry +Verified in latest alpha manuals +Is anyone working at this bug +Niclas I asked the author of these sketches to register with JIRA and add his content directly but once he registered he doesnt appear to have permission to comment on this issue +Can you let me know which version of xerces you may or may not have on the classpath +Any idea how quickly this will be rolled into a new version +you use git your name will appear as the author of the commit. +Please provide stack traces and if at all possible a standalone test case +Also note that I am not married to this approach +At the moment has a dependency on JCR API +Changes applied to breanch HEADWhat a terrible way to spend the weekend +We need to test this extension wide +Attached is a sample application illustrating the problem +patch closing the input stream in the Suggestor and adds additional safety to in the case of an error we close the stream internally +Well I want to develop maven plugin with inside in order to add compare process to the maven build +A new patch with modified is attached +Handlers can use at will + +i think alex is right +Any news related to this problem +Looking into this issue +Agreed this has burned me on more than one occasion the default url should be localhost as that almost always where my broker runs when Im unit testing the code +Thanks guys +If the second argument is not specified we use the encoding. +can we make the datanode to use a config that specifies the classname that will implement the above interface +Ive put in a stronger set of code that deals with empty and quoted character sets +Good enough Ill go for a single file and a note +I found that the code David is referring to needed work because the commenting was done by the model classes not the HTML rendering classes +markmodifyreset can cause problems if another thread also tries to use the buffer in between. +I think it was introduced in a snapshot though so I doubt if this precise issue occurred before that +Without changing the actual code to put in a sleep to stop the synchronization and try and make it break its not possible +Hrm +Done please wait few hours for sync +Is there any benefit to storing key input as BB? I dont see any we only convert it back to String later +No users outside of have modified any files or folders in the underlying file store in a manner that would affect this commit +Perhaps finding out why those cannot be rendered in the RI might be one path to take +But feel free to do it especially if you have the logs or a scenario +Epic patch thanks! Ive finally had a chance to review it committed with a few small tweaks in r +Ok I added everything +Now if there are no accounts setup on the phone the contact API defaults to using the phones native contacts +Sorry I meant the whole grails app +I tried out XML and generated a schema configuration with the free version +It duplicates what mvc provides and in most cases it doesnt make sense to have both +Assigning to Howard as Tim wont be able to work on this +Closing the fix to original patch for YS +A bit refactored patch committed in SVN revision +also and if there arent more than pages the pagelist construct is suppressed too +Apparently auto activating safemode is bad idea because it will stop any Name node activities and halt all job submissions +attached images of agent and collector with batching and compression metrics displayed +Thanks +Fixed in SVN +Patch looks good +Make sure the error references this issue +This also conveniently reduces the resource requirements on the serverHow does this code handle the case where the user starts an upload from more than one multifile upload in the same page or separate multifile uploads on multiple pages +like you truly believed in your heart it did +Attached the groovy script and output +Thanks TomThis patch ensures that the following code paths conform to the configured block placement policy +this is patch against the against the trunk +Moving issues that arent going to make it into to votes to make a blocker +If you dont use it in your configuration then Jexl would not be required +Or I should insert absolute path when selected cfg file not from source. +I kept safe because it was used by a class which extended and relied on that method to be called + +Fixed in CR +See for further details. +Cannot reproduce on GA with the given test page +Concur with blocker status +Please ignore the wget results +IMHO the size of the AMT should be increased to handle the proper size based upon business +I understand theres some difficulty to write the test hopefully we can find some other way to write the unit test for LDAP provider +This was fixed a while back. +Moved to upon release of +Can we get this added for the ivy file as well? Due to Eclipse constraints for project definition many workspaces contain separate projects for source and test code +Patch added thanks Calvin! +This is a vital missing feature +It seems that JAR verification is spontaneously failing to occur for certain files +JBTIS is for issues specific to JBTIS buildconfiguration issues and should not be used to report functional defects in specific said it is perfectly acceptable to link the SWITCHYARD JIRA back to this issue +I am not sure if it is complete with respect to functionality but it worked on my local examplesI did not follow the steps in my previous comment precisely since I just wanted to get it the functionality in place before worrying about performance +Though the portlet demonstrates limited functionaily right now I can see its usefulness if it provides full function and I would like to see this portlet in full + +bringing the discussion back to the original issue in looking at the new i dont see any filtering of child nodes +There was a change made for that addresses this +Id like to commit what I currently have and then make changes from there +This is the WSDL file he suggests +AMQ can install in Karaf +Pete let me know if you also want the version of the HSQLDamian +You prefer another design approach which is fine +This example does not exactly reproduce the described setup but illustrates the same effect +Yesm that would certainly have allowed for more powerful logging but alsowould have caused a tighter coupling to firebugs logging at the interfacelevel +Albert can you please try again explaining what is the problem you think yousee in the existing code that you have pasted in here? You evidently seesomething here that doesnt look right but its not at all clear to me what thatmight be. +I think the root cause of this was me using values that were too small for and +Why do you have to import the source into VAJ in order to use CommonsCollections? Cant you just use the binary version as a jar file on your classpath +That doc is hella outdated but the real issue here is that the interactive plugin has been deleted +Alpar your question relates to the support of Cucumber see. +Where as Karaf is using Felix and AriesAlso I think gnodet found some issues with those SMX bundles for XML resolvers and what else is there they didnt work to well with +Let me know if that is not ok for you or feel free to if you encounter any problem with the suggested approach +Im not sure how useful this specific encoding is in thegeneral case +If I reload the full page and click the Home link things work as expected +Thanks Hairong! +For the same reason should we do a in the second part of the loop? Otherwise an thrown within the loop leads to exiting the method +Fixing the stroke drawing code leads to the result +done +builds against trunk but now gets NPE at runtime so needs to be updated working on it +Still tests failing in smtp and pop server projects +In any case this is an SSL perf issue not an auth issue yourpassword isnt checked unless you are pushing somethingI will have another look at the SSL config on the server tosee if I can change something on our end to improve the situation +This is the new patch not just cut out ofI removed references to the Base class and fixed the edge bug +Show code snippets +However I would like to see your proposed implementation first if you can come up with an solution that doesnt need a new xsd and that satisfies everyone I will happily accept it +Latest nightly installed using the apache flex sdk installer from about hours ago +I have only recently gotten my emulation of this force behavior working so I havent discovered all the wrinkles yet +If it wont succeed Ill use the workaround suggested above and access the DNS before so that its failed resolution will be cachedThanksOz +The following are the specs of the UNIX box +I think its not just the problem of sync +Attaching missing module +Anurag would you like to review the modified files for this issue? It would be good to know if they are okay to commit so we can get them into the releaseThanks very much +running +Attaching with this text update +It would be nice if a jira issue were created though so that this problem isnt forgotten +This file will not contain that apache header +I dont think it should be removed +The problem exists even today and is much worse with this change the serdes have tomake the change only once +update it to what +The servlet objects are not available in a portlet environment so the is expected not to work there +Ill get this reverted +The packages now contain the right files and are the right type. +how are you assisting? I did not see any patch from CSIROOr youre saying that youre going to work on it +Michal in terms of uneven request balancing do you know if the request unbalancing happens for all clients? Or if only some of those suffer +Very hard to find the real issue here +I wonder if the commit didnt get pushed correctly +for use of dainty in a JIRA subject tho. +I put CP as the Fix Version as that was the only alternative but dont know if Ill be able to do it in the next few days +Instead of downloading the plugins when finding an empty project +If the application was installed in a directory that does not need escalation when uninstallation is tried escalation is not executed +Closing all tickets +the error might only occur if you have an older httpd running modproxy in front as selenium test to reproduce this issue. +Documented as a known issue for EAPResources from the resource servlet such as CAPTCHA images do not correctly because they are cached by the browser +Tested with both and and problem no longer happens +After updating the patch I can test again with the final version + The pictures shows that Label in does not work with Harmony +Yeh a in the file as suggested will be very visible to anyone thats using I dont love it I do think its still the best client library for and the one poeple should be pointed at I most definitely aim to make the best client side HTTP library for Couch but it isnt there yet similiarly with nano its still more aimed with node in mindI just dont believe its canonical development location lives inside core similiarly with fauxton itself and many other things inside they can be used developed and tested independently the Reductio ad absurdum of keeping anything couch related inside the same repository gets silly quickly and I think this is one of the fairly obvious and uncontroversial things that can be movedIts worth mentioning when I maintained a forked version of due to difficulties getting my patches into apache I started receiving numerous pull requests to that repo I suggested they be sent to apache but I know that many didnt make it +yes go ahead well use +Does something else have a handle on the file? Are you sure that you arent accidentally running two instance that are using the same directory +Lists are created people are subscribed +Earlier this used to work without any issues during host reconnect +See for further details. +This is addressed in the latest version of the patch. +Added try catch blocks around code that will fail when popups are opened in tabs and fixed it so the browser chrome isnt blocked when modal dialogs are openedPerhaps a future enhancement could be considered to replace with was an issue with IE too +Removing the gwt ext caused this +This issue has been resolved for over a year with no further movement +This would require a notion of a transaction however +Everything runs fine with the axis server on port but i dont use thatI have a normal service running on the server machine this service outputs in plain xml +now advises against mixing this with deleteColumn for the same qualifiersI added more test and propose this for commit +Patch for Example Shai created the patch for +Fair enough +i committed patch under rev +Patch against branch thanks for the patchI think that would probably work since the only code path that could write extra bytes is there but I want to confirm by writing unit test for this +Build was broken +This was due to the fact that the View User has been marked as an ACTION +it has lots of nocommits still but we can iterate on the branch +My last round of lexing changes broke this +tomcat if a patch is not immediately fortchoming would it be reasonable to markthis as an enhancement for the milestone +Tests passed with Asserting +tests doesnt is correct just clean rebuild all. +If you already got JBT why would you install devstudiod same as ce not following what this means the updatesite will need to be there no matter what and its what we would start with before pushing it into marketplace +In any case code is simpler +I think thats a mistake Ive made with the RT branch so far +and +Rather than adding an api I went with a behavior changeduring a partial encoding visit we no longer call setup before invoking the visitCallbackIf the is ACCEPT we do continue to call setup before setup +increased the test wait time from seconds to seconds since the close might take up to seconds +Anyone else using this combination will also face this sure where CMTBMT come into it though this is all pure jBPM in a very simple configuration +Jesse DougI think Jesses comment Oct should be put into hbase book +Thanks larsh +on the patch +We can wait until for AS supportHowever it would be good to follow up with the AS team on this before is released in case they need to make any changes +I added a warning on the component wiki we can downgrade to to have the validate work +Yes +bq +Can you file a bug more details will be great to haveIll run the test case again and see if I can get more details on what is causing the connect timeouts for tasks when they are launched +Can someone please reopen this bug and fix it correctly +Well thank you Aaron for doing all the hard parts and persisting! Sorry this took so longThis will be an enormous improvement for +Still needs the confs + +There is no ambiguty atSELECT o FROM Order AS o WHERE Smithas both OQL languages access fields in the same way +Uploaded the latest patch for upcoming Smooks +Anything else has less priority +Plexus has solid mechanisms to process these t know if this should be done automaticly cant oversee the impact yet considering backwards comment would be welcome +It was and that I was unsure of +To add to the previous discussion probably the most sensible way of Fuseki would be to split the project into two This would be a WAR project that incorporates all the servlets and refactors the existing configuration mechanism into a +Werner the patch works well for my case + +You can reach me by email or mobile +rather than writing a timestamp file why not use the that javadoc generates +This one does not apply +serverdefaultdeploy isserverdefaultdeploy is wrongcserverdefaultdeploy is wrong +Need an icon set for +Im not able to replicate this issue with clean JBTEclipse installation. +I filed +Without this it the PENDINGOPEN can be set before OFFLINE has finished and well get ourselves into an unwanted state + at r on and r on trunk. +The MXJ problem seems to be because JMX is now standard in +The revert just makes more work +Does anybody have a strong opinion either for or against reject all cookies policy as a standard feature? dont have a strong opinion on the matter but it sounds like a good candidatefor contribs certainly a feature that should be provided by the stock implementationEvery web browser has this feature so I think we should provide it too +Is this still valid? If so can you rebase the patch and well take a look otherwise lets close this issue +Just tried and the error is still thereThe output looks exactly the same as in my comment of Oct on Jan Eriks last comment I suggest we reopen this issue for further investigation +Two comments on the patch Sijie present is read only in an assert statement so I dont think it is really needed +Some more work towards full SAAJ implementation +I didnt study the closely but is that something that will trip us up come release time? These files need to go into the bin dir in the bin distributions +OK +Im using version +What is the precise issue? You cannot set the version attribute? Is it not defined at the API level +has that changed or is it a different jar or? +Another JIRA sounds like a good idea +Ive committed the patch and asked Juergen to take a look at. +updated version applies to trunk fixes a few bugs in prev version +So Im selecting PlatformInventoryView Results resulted in a the first time I tried it +First pass of review comments added on Review Board +Checkout ui from tag +Hi Justin I need a wiki account to be able to create the GSIP +Ill just create a small test project and see what that gives us +Thank you +The file exists in the distribution required to install ESB but the comments are out of them in the project will be in the next merge +Hello GeorgeThe fix looks fine many thanksBest regards by Andrew. +Otherwise I would try with a basic auth filter on the web filter chain +add links From RF dev guide to JBDS and Tools to the corresponding guides sections add link to Max Katz book on RF +this was with JDK Jozef will test with JDK does not appear on have founded it similar related issue with jboss web service Shellys point on the release notes this is not an issue +Removed test programs wrong parameter +Is this related to +Works for me +As you can see hes browsing to servercustom and selecting as the configuration +After creating a new server and clicking open launch configuration you can clearly see there is no b flag at all +is v? Its whats up on rb +Well Id like to have the possibility to select an account youd like to have the possibility to enter new credentials +Apologies +But this patch gone stale a little bit +Use of for Web plugin and Sonar works perfectly +Tracked in. +Sorry here is the correct patch +after the EJB remote method invocation asynchronous method invocation timeout or message delivery completes orThis pretty much implies what you wroteWhats imo missing is TimerBtw there is atm a HUGE difference between remote and local EJB invocationsAn example Imagine a User CDI bean which gets injected into a Stateless EJB +I just voted +The zipped file is bigger than mb thus I send it Dims one +For upgrading we may throw an exception if +hadoop archive works +Lukas yes Id like to be able to enforce focus on particular component but declaratively from facelets +Ill post a new bug report again when I have this sorted any update +Will this be a one time check and store state +Added a Before which resets UGII ran all the unit tests with this patch and they passed +there are more coming! Not all of the tests complete successfully they reveal other issues +The is applied at revision If VM memory is confirmed I suggest we close this issue and create a separate one for VMDo you think we need to add this test to regression suite +Teiid is mulling over ways to accomplish this perhaps setting a max buffer size property to have a frame of reference for disk usage and free space +Santiago does Nialls patch provide a workable framework for you +Just unjar thencd JAMESBLACKPROBLEMsrcant build warThen after deploying the war file to build and test the client doOops you need to change the file to use beehivehome sorry about that +My patch of also fixes this issue +Added synchronization to prevent multiple assignment of a regionModified to requeue itself on the delay queue if all the meta regions are not online +Thanks Mark for the fix. +Correct +But if not I dont have time to add explicit default dependencies int he EJB deployer before scheduled Oct release of RC +I was making another build on the th after capturing the db and before making the screenshots which added +Thanks. +JDK introduced a class +If anyone thinks otherwise please submit a testcase. +Added scripts to generate example ssl stores in r r +Patch to to review +I will be pushing out snapshots that are known to be stable on a regular basis good! please let us know and one of us will apply the patch and fix the etcthanks note in addition to the consolidated patch the jar needs to be installed per the instructions in earlier comments +Bikas makes sense thanks for summarizingOn the decreasing +Thanks! +Not sure youre going to get much from this s the problem this is trying to solve +Bumping this back to In Progress because I just thought of a corner case I didnt cover when cleaning up the marker node will fix it up soon +In method there is a check for existing file before verifying valid extension +Ive played with renaming users and so on and it seems that the behaviour is no longer exhibited if you do either of these two things ALTER TABLE itemusage DROP COLUMN value DELETE FROM users WHERE username LIKE USERThe new attachment contains the DB +Take care to use same code formatting style as existing code +Estimated remaining effort. +Marking as closed. +Applications that require the binary format can use the Writable methods +Also made logging more patch to modify perf test frameworks queue management to allow JMS implementations to be tests were inadvertently using persistent messages causing misleading poor performance test results +Added and committed thanks for the patch. +If namespace is explicitly given or implied it will have a fully qualified name +Please review it consider commit message for more info why I did it so. +I have no clue what you are talking about regarding proprietary libs but if your build generates themso whatThe point is not that the build generates them its that it requires them in order to generate complete javadoc +I dont know if Ahmed ran benchmarks on the current iteration of the patch but I saw improvement on terasort and some other customer workloads with this patch on back when I originally proposed it +What are you suggesting converting the columns written to blank values +Bulk move of M to things needs to get into place to have p work have done the first round of this Added to of plugins that did not have it Use Execution Environment in classpath instead of specific named to next version +Seems like you have the issue resolved. + +What prevents the task tracker from overflowing memory itself? Considering the memory leaks weve already seen in the name node I dont trust the task trackers to be leak free either +Passing it through hudson +Please report it to wtp team +Patch committed. +I just committed this. +If I have split panes next to each other like thisx x xAnd I drag the first divider towards the right the last pane in the series loses all of its space +These rules are per the WC Contact spec that we implemented +Verified. +How can i verify this? Person data is displayed fine for Friends Gadget +Thanks a lot Ellen! +This is not the case and only store input stream or reader passed via constructor and have no additional processing +Ill check with infra why its the contents but dont have permissions to fix the directory or the archives +Dependencies are updated to the AS latest +Uploading which uses for testing purposes +So we should probably skip the ftp unit tests when the ftpserver jars are not present +Lol +Not sure why this one was cloned. +This is activating the quick assist rename rather than the refactor refactor rename we are executing the groovy rename functionality but quick assist rename is executing the jdt rename which will not work properly in groovy to figure out how to usurp quick assists so that the groovy rename is executed instead +Would remove the old stuff and talk about how we have a listener in surefire that dumps a line at end of each unit test noting thread count etc +zookeeper itself should not deliver with logj dependencies anymore +I missed the modification to the grep +So this is not strictly required for +Note that the problem also occurs when the replication sync parameter is set to quorumdisk and quorummem +If there is a gap in sharestorage should we find a way to repair it if the gap exists in localThis will be a big revision +and in order to avoid regressions in the code it is always interesting to have a wide variety of test messages to ensure the changes one came up with didnt break anything +There is a MB ticket to track the MB fix for that dependency issue. +Yes its there +Patch adds a null check on s outputStream in the flush method +Opened as for Standalone Tiles since thats where active development is taking place +I kinda like that character to mean number +patch merged please confirm if it works +I have mostly figured out how to extend to do a test but I dont know how to copy or use the solrj with initCore rather than the example configschema +problem should be addressed by waiting for resolution on that before determining if there is anything else that needs to be done +Oops looks like we have an unrelated bug the error message is missing +This change was implemented because of the sharing of the organization between HRM and accounting not all departments need accounting +Ill post my findings in can never know how widely they are used thats the rub +After cleaning up the testcases from here they should be transfering to bug +Ive just tried a Maven build and it works fine even if I delete the lib directory +Going to resolve this as wont fixI agree that there isnt a compelling reason to support destructing syntax for field declarations. +Attach to resync with trunk +Thanks +Or you could attach your application code so we can try it +Maybe I am missing the point here but surely a real fix works of the time not just of the time +If we are doing backport it seems that we should have it in as well +I find that and depend on as well as other unclear dependenciesI think we should make sure v absolutely depends on bundle +Thanks +Reason for change is that during an automated test runs with an external broker the broker will overwrite its previous configuration as the APPEND value is set to falseAs this change will configure the Qpid broker default to create an unbounded log file +Florian I cannot see that JIRA distinguishes between ticket types in the notification scheme +rebased +Am I missing something? Or is there some other callsite that needs to use the integer version of skip? If its the latter where is it +But done from the roo directory its the main directory where you fine the LICENSE and APACHEHEADER files + to trunk. +Rick later the test and made it run cleanly on platforms that didnt have the required XML libraries so I believe it should be running fine now if you do an svn update first +Logged In YES useridI am using In YES useridI think it was fixed in +I dropped it from to segs +I see these two as separate means to perform the same task +Kicking off Jenkins to see if it complains of anything +briefly looked at some of changes on the client side and let me start with a positive commentIt seems a lot of unused code has been deleted +Time to ask what is the best way to handle vhosts and location to the community I guess +Will check the examples and commit a new patch +applied. + already describes this issue +Verify that when you patch your local Hibernate nothing currently regresses +Work on still pending should we do a different JIRAThat seems fine to me +Closing +you also need to have it per datadir since those could be on different volumes and hard links wont workWith landing soon I too dont see how this can provide any value. +Submit just committed this. +It seems that was caused by this it took some time to trace it back here +reproduced only in Seam Ear ejbModule on custom macros in palette +Thanks Robert! Mostly I was making sure I didnt lose anything +Ok please try again now I have removed the view restrictions from that page and given you full write access to the wikiPlease confirm +Added a new client version +Vincent timestamp filtering at read time is being implemented as part of FYI +There are only two things missing I was not able to do +Well need a small unit test with EXPLAIN for this case +The code segments have been removed from the manual +Let me know if you need any other size orCheers heres a banner for use in MagnoliaCheersCheyenne +If it is null the user is not only way you can avoid this is to use synchronized on each request which would drastically reduce the performance of your application +Our page size specified in the configuration was set to causing data corruption. +It should include the kernel features but it contains the obr repository instead +Please enable logging to see what modules are loaded +Fixed in branch and because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I test your patch Pedro and thats work perfectlyBest is a small problem with the patch +Is it really critical? Especially for I set it to Major +Also why should this go only in +It also contains the skeleton of a JPA integration test set which can be completed when support for has been implemented +Well isnt included in the Groovy distribution so I should even unschedule this theres no rush for closing it actually +Confirmed in Couchbase single on OS X clearly not an issue just on windows +Gregory thanks for listing thisIm going to make some improvements in crash handling printing modules mapping was in my initial plansNow Ill take other wishes into accountIm hoping to finish this by the end of December +Its kind of tough to verify the patch since it shows the renameedit as a delete and add +Yes it is still an issue in +Chris no problem I created this confusion until the patch is OKedonce this happens i can include some additional testcases that I had problems withi have all revisions i made of this filter locally so i can see which scenarios fail on each previous iteration I think these are good tests +The failed tests seem related although I cannot get them fail in my machine +If you want to upload a new patch thats okay tooIll also try to run the test tomorrowthanks +good point +Right now different entities are using different data types for durations so there is no consistencyI would recommend entering durations as yyyymmddhhmmssmm +The and etc are all currently based on discovering classes +It should work whether is to be true or not as it doesnt relate on the size of the thread pool work thought? agree that the throttler should possible not use a scheduled thread pool but instead have a internal FIFO queue with delayed exchanges use an internal background thread to periodically move exchanges from the FIFO queue to a worker thread pool use a regular worker thread pool for processing delayed exchangesThis allows us to use a regular thread pool for processing the delayed exchanges which allows the thread pool to growshrink it avoids scheduling many tasks on a scheduled thread pool which seems to consume more memory feature of Dynamically changing maximum requests per period may make this a bit more complicated as the time slot size can vary over time +Can you give us an example percentage of bad records that you see in the jobs at Y +I have zero experience with Maven so I dont think Im the right person to take care of the Maven uploadI might be able to handle the Parser although itll probably have to wait until the library author makes a new relicensed release available +asking for all matching docs in sorted order isnt that scalableIf we get the interface right though more efficient implementations can followFor that reason it might be good for implementatin details like collapseCache to be private +Modified version of to enable durable messages for testmanagement +Bringing into is a patch that before it adds new daughter region to unassigned it checks that region has not already got a server and startcode in the +We can either manually update the output or much with the xmlcodegen script +I THINK Ive gotten them allPlease feel free to reopen if you find specific areas left in codebase +So do you think this would sill apply to JDK as well +Sure +Ill commit this soon unless there are objections +I confirm this issue too using +When having a sheet in the tab you got an error message in the log file no height in parent container setting to bug is for both dimesions for width and heightThe reason is that the doesnt call the mainProcessing recursively in the case of autolayout. +Was fixed in +I dont think it is necessary for the user to choose the format +Revision made JCRSPIs referenceable in case if the underlying instance itself is referenceableIll leave the issue open for now because this may not be the final approach and for now none of the Jackrabbit SPI contribs currently support this +I am not sure how to handle Documents with Fields that have a as the value as I believe these cannot be serialized +Thanks Pete! The test failure is a known issue. +Resolving as duplicate being implemented by +Ignore the above comment +yes sorry my fault +Committed to x and trunk +I need this to occur so I many use it in a webstart app but webstart dont support files apparently +Committed to trunk and. +Reports should go to the style editor. +It only checks the WAR file + +I currently work for customers that moved away from spatial due to its early stage and flux warnings which is quite understandable though +I verified the property is getting picked up by the test in the description and will do a full test run and check in if there are no comments +solutions +Also added additional tests to test new exceptions. +Applied. +I dont have clear answer for this but some ORM can give a specific answer +Edited +Using the latest VC compiler from the windows sdk I do not see these warnings even at the highest warning level +My upload must have failed earlier trying again +Added a section in the configuration chapter on using alternate JPA providers and what must be done to support them +bq +Im a little iffy about this because there is nothing else like it in VTL +I still have to verify but if that is a possibility than that could be very much the case. +Thanks +I tried by putting in debug messages and was able to reproduce the same behaviour +Updated in SVN. +Try the newer versions of the plugin +Good points Finn +If the user defines Djspc the maven attempts to compile the existing jsp code into targetweb folder +My previous eye but Im actually fixing the close issue on a separate jira +Thanks Knut +Patch checked in +Hi GaryThat sounds good but I cant see the patch +Committed to branch and trunk +So the comment is going to be removed from the codeFor comment I thought about it too +As for the error when you try to use a zero index it should probably be handled the way other out of bounds indices are handled by throwing an +not blocking JIRA type from feature to bug to ensure it makes it into the Beta release notes +Closing all resolved tickets from or older +I am assuming this is restricted to only Cloudstack project and when I select any of these options the emails will be sent ou to corresponding mailing list +To me it makes sense to include the bitwise operators as part of the syntax rather than as functions because then the correct operator precidence is available. +When tested endorsing token with message level securitysymmetric binding and it works +Patch applied with some modifications to handle date and time being parsed by the handler +Someone reformated my code it now looks really ugly The problem is that the implementation doesnt match the semantics specified so you go and change all the code to the new semantics and you find you cant have multiple options and if you try it eats all the parameters whether they are options or not +Attaching + +This patch contains the filter index documentation +OK I committed the part +Thanks for holding my hand Lance try that +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be need to add to your dependencies in to make this work +I suspect it is a weird interaction that only happens on Linux +Please let me know if you have any it works please close the am interested to have coverage support for DB but have had no time to think about this +I just committed this. +I applied the session sql scripts patch on trunk and it worked +is already accessed without synchronization from +Could this be because I moved the repositories to another pathIn such case what do I need to do? you resolved this issue +We should upgrade for the supported version in EAP is already at level this JIRA must have been moved here from some old list +Please feel free to reopen if needed. +Also have you asked about this on the user forum or the Seam forum? Do you have a test case that reproduces the problem? Do you have a proposed fix? Any or all of these things will help you get your problem resolved much fasterI doubt that anyone has time to do the research for you on this but if all else fails you can open a new jira task for itRegards to the issue at hand Ingo has reported to me that all Seam examples are working with the latest fix +Hi Simon I have already applied this patch locally and just waiting for a complete successful build before I check this in +Committed revision to trunk +Thanks to Paul Rivera for provide this patch +Ping I am still waiting for clarification on this issue +Please check that the bug is fixed for you +It can be fixed easily by changing the MEP toAs this feature break the backward compatibility I dont plan to merge the patch back to other branch +Wellto be sincere I am not well used to the content application +And in firefox all works well +done +Committed dated PM for Dianne under revision . +lgtm +Patch. +inczip type will be unpacked and included as system include path at compile interface has a new flag called deployableGive it a try I will need to update the doco and add srcitlinkageswin example now uses the inczip scheme +Otherwise wed be trying to figure out if a handler was set and concatenate the checkbox hander to the users handler and setting that instead +This is the same technique that powermock uses to access field which has been a disalllowed testing technique in the hadoop stack +Committed revision +See +del file of different sizeNevertheless all data is loaded without loss and an optimize will clean up everythingObviously another bad side effect of and NOT seen with versionWill be hard to find out what happens +Merged to the branch in revision . +push out to +You must upgrade these file because I have the chaperon completlyrewritten +I think we should leave an ONOFF switch for the reasons Ahmed mentioned +Fixed in the trunk under revision r. +Hi Cristianthe method name is still passed to the +IIRC we used configuration elements designated for older version of Arquillian than we were actually using. +This affects trunk and +Thanks Juca +Is that still the one to review +Patch for trunk +Without those things I couldnt get the client module to compile so probably not possibleI think what Jesse means is to do these changes before the module split and moving the files +Attaching simplification patch +code review says this is fixed on and trunk +LarsThanks for reviewing +Much of this is due to our command packets using variable sized data +Should we ask to include this in +Which could be done for classes created by Groovy but not for classes from Java +I cant reproduce +Excellent way to address the version problematikRegardsMichael +Two possible patches to solve the problem +Its not the wicket way +The go producer and the scala producer should send the same bytes for the same message right +Possibly a java Problem? I think the issue should be reopened until we know and then closed invalid once confirmed and as a java problem +Assigning to charles so he can confirm whether this property is ever pulled from the DB or not +Wed probably want to cache that andor do it lazily +We can now be sure that it is fully private not even the attributes are touched but I see no real sense in this requirement People should never use old and the brand new API in one consumer +Am sorry for not providing the complete info +Ahhh +Ifwe arent doing so thats worth considering +I havent decided how we are going to handle this yet but we may need to load these jars in some jdk directory or is true that it is about and the Java version and probably a only class To make sure I tested the following combinationsjava no with okjava no ok +Problem is invalid +As we have a version ourCVS taking again a would be not that problem +Ill follow up with MODE and see if we can find a real test case to demonstrate our problem. +but filter out all the table with any system schema specified in the set in + hintThis issue looks like low hanging fruit for anyone wishing to participate in the contest or for any eager QA person looking for a point +Sanjay had an offline concern with the negotiate response for kerberos but as documented its necessary for IP failover support +I can send you my test offline if that helps +So we had to patch Capacity Scheduler there as well +Im going to review what we can do there + +You can find here the varve build the mevenide from cvs sources +Lets be optimistic and affect this issue to back to is no longer valid cause we dont rely on JNDI in the code anymore +Hi ZhengOh! Its workingI add a line export HADOOPCLASSPATHCLASSPATHHADOOPCLASSPATH to binThe old version export HADOOPCLASSPATHJAVAHOMElibHADOOPHOMEconfHADOOPHOMElibHBASEHOMEbuildtestHBASEHOMEbuildHBASEHOMEconfHBASEHOMEHADOOPHOMEThe new version export HADOOPCLASSPATHJAVAHOMElibHADOOPHOMEconfHADOOPHOMElibHBASEHOMEbuildtestHBASEHOMEbuildHBASEHOMEconfHBASEHOMEHADOOPHOME export HADOOPCLASSPATHCLASSPATHHADOOPCLASSPATHThank you very muchThe problems are many days +Thanks for the fix +Updated patch to address latest comments +forgot to add configurations changes to add pthread library +Ive also added an option to export to Atom using Abdera so we get the threading extensions easily for comments +In the context of style mapping works well with the existing implementation +Somehow it doesnt like the files you moved to a different directory +When all are clear the localization is based on the browser settingCan you please provide more detail on seems to work only for parts +fixed in current this since has been released + +JoeWhat format did you upload your attachment in? It seems that it is in binary +I am setting the issue to unassigned in case someone else wants to finish the WITH DATA option +I used it for example to init redis connections +this test shows the problem +Business calendar now delivers the expected results from business time business calendar properties to official U +I just committed this +Fixed issue summary +The next patch will be containing the code for viewing the status of the nodes in Geronimo ACE +if i have added axis kernel jar to the same lib directory then its workingBut I need to get it done by Axis kernel +We need to announce the change to the console developers and ask them to use the dir or keep their pace and apply their changes. +We then have to extend the Java API to add a new functionWe may add a extending the JCR one with new methods. +Any comments or reviews are welcome +Just a quick update I am really close to posting a patch +That was quick +Should the installer put the modules in the soa layer if installing into a SY EAP +I just committed this to trunkThanks a lot for the quick Alejandro +Thanks LeoApplied to LUNI module at repo revision rPlease check that the patch was applied as you expected +Sounds good to me +for some reason I etchosts +Hi for this let me have a look and Ill get back to you +Tomcat is available at gitTomcat trunk is available at git Jukka +There is no stricted schema as long as parser must be loaded ahead of time and parsers must match the column family in table +I also had no idea people used repository as a stand alone library +all of these jars must come from the BRMS version used +Resolved for a while +Keith do you have thoughts on which the lexicoders will liveI was thinking somewhere under but do not have a specific sub package in mind +seems to have all the answers. +The default behavior is to spread all tablets in the system evenly across tablet servers + +log WARNING JSF Unable to find or serve resource from library +Done closing. +make it so +But we should fail before submitting the job. +Thank you! +docsupdated +Refactored! +Hudson seems to have stuck again please run ant test to make sure everything is fine and that the docs are building properlyThe changes are committable but you need to make minor changes to the patch directory structures so as to reflect the latest project split up +Released to track the EAP up issue +Please restart container and retry once or twiceit does not fail every time for me but nearly every time +Thanks +I backported them in at rJeremy could you please check if all the changes you did are in rThanks +After editing SQL WorkbenchJs driver configuration to allow connecting via the I opened up a connection with the first window from SQL WorkbenchJ and then performed a create insert and then a select on a new table +IMO if were backporting improving just change the ticket name +There are EJB containers which implement JSR and use CXF as the Web Service stack like Apache Geronimo and Pramati Server +would you please try itThanks a lot +Yep +Committed +Verified by Svetlana +and +Please check againI added your test case to our test harness as well +Yes youre right +Ramgopal do you have retries enabled? Bump to say or and retry with a fresh cluster +So I guess we can resolve this one? Thoughts pleaseSupun +Thanks release note for publication. +I assume people know what they are doing +Everything works great now +The interactive bit is so that developers can get images asynchronously +Ill answer to most anythingAs I told Arun in IM +I cant apply cleanly to trunk +I get the following output when I insert echos into the script +No response assuming ok. +Presently we do not spend any effort in differentiating Windows versions and are unlikely we ever willI am inclined to mark this as WONTFIX +No it was just a comment about how I did it +it was removed because it doesnt work across changes to the file formats +Ill commit this shortly +Also I think it needs to check for the sync already having been added so you dont get more than one +At the moment Im focusing on getting the scripts working smoothly +RamSure give me couple days to spin a new VPX in our lab and retry +Adding VERSION to fix this issue + Trial is a known restriction of XSP +A improvement for replication metrics +It sounds as if you may have fixed some those as well so Ill just wait and see what the overlap is +As a bit of backround we are migrating from an Ant based build to Maven +A reference to cellar in the list of should be enough for now +Fixed the stop issue +As you can see you can have more than obe value for the attribute which shows that you can have more than one value for an attribute +The hurries of attaching didnt find anything related at myfaces wikix Mark this as WONTFIX pleaseregards is already Cagatay thanks on your fast response +Updated the to dynamically generate the address if the host is undefined +can not reproduce the bug +Defaults and namedescription on all artifacts for +what should this do +Thanks a lot Ram +I havent looked at why the test fails on IBMs JVM yet but this patch makes the test print more info when it fails +it will provided you are overlaying two symbols or something similar +If we do want different behavior we should add in a new API and deprecate the old one +system jdk for default selection is now available on all build slaves +Patch available +We should remove this block from and add new Block +Although LDAP support is available for Redback I still do have problems unless I use the latest SNAPSHOT of Redback ant put it into archiva WEBINFlib +on patch vThe test failures reported by werent related to the patch +You need to download the zip and point your p update manager to install +Just to put things in perspective I ran few more tests on our setup +bq + on PR it works great +Splendid thanks for such a quick fix! +trivial to make with a reasonable ide +Grouping task for scalability hotspots +Created +for the patch +Any version that is lower than does not need to add LIMIT +Merged on the branch as of revision +Theres nothing sacrosanct about Tapestrys built in components and is nearly all cases they dont use any internal either +Or maybe Im just worrying too much about it and its not really that big of an issue. +So Ant isnt one of the module dependencies +Why choose? Wouldnt encourage a major compaction and get s out of the picture sooner +The value is used by when retrieving connections and not directly by the connection managers +Ive taken a crack at fixing this inside the code Ive attached the new jar file this should drop in as a replacement for +closing +Fixed it on revision is fixed in RC +I observed a similar issue once there the tagdict restricted the outcome to outcomes which are not in the model +Im using the primitive it appears only in one of my domains where I cant use boolean when this class is used alone the problem disappears +To reproduce this +This inevitably leads to situations where you have to catch exception even though you dont want to +If the project packaging is jar and the jar file contains java classes there must be a for main artifactCan you remind me +bq +Also applies to the AS verified you can now use as a bind address and it functions as expected. +The lack of it in hadoop trunk is forcing us to ship a hadoop jar just for a line fixPlease commit already +A feedback on how it integrates with Solr would be welcome +Committed to trunk and +Nitay +Rest are looks great +Input only some space dont input tab name if has space only +If we keep this option perhaps we should find a better name +I think an even better solution would be to instead define a new conf property called and make it a comma separated list of acceptable URI schemes for import and export +I havent fully worked this out yet but I thought Id bring it up for discussionMike +I alsoupload a bundle with the changed HTML pages for ease of review +Looking through these they seem to be fixed. +Instead this belongs in the validator so use so closed. +The identity hash code as a component of an object name does not satisfy that reason this is needed is that the cmp engine cant create the tables fks etc until it knows about all the ejbs it has to deal with +Committed to and +Ok I think I understand the problem hereThe tool correctly determines your t is a simple type due to its being a derivate of the simple type xsdintegerHowever it incorrectly determines tAge as a complex typeLooking at the code this is because the WSDL parsing portion hasnt added these to the table of simple types +Puts back the for the delegation token renewal registration +Reset fix version to +For months Ive been busy due to preparing the +I left out the failover lock stuff for simplicity of the explanation +But it isnt in the current code +Please test +Thanks for the patch +Easy enough to check +Unfortunately there were other errors in other files when i used +Before that Ive removed luni modul +Committed the patch rev Will leave open as the second scenario still needs work. +How did you make it thread safe? Perhaps an example of the compiler options please +I tried it but it did not help +Please close this ticket for now +Committed the patch to trunk +So there may be transient pot holes whereby ognl fails even with caching disabled +hellocan you let us know what excatly you did how the input source looksThanks +I think the issue is that get fails since there are some rows in +For the most part it is really not much code to create a session and ensure it is logged out in a finally block +Could you make name space saving to be done only once +have merged the HDFS Snapshot feature branch to trunk +release bulk close +This keeps the door open to potentially later add things like a Hadoop +We could close this after that gets +Could you provide another patch for trunk since this one fails +messaggi tradotti messaggi non dom feb CET Total number of messages not yet translatedBye fot trnasltion statusSTART ven apr CEST +In my test the command was forwarded to the new owner but I need to dig further still +This time using to generate patch and ignore issues in my private tree +Im sorry I couldnt help much in the last few weeks +Yes same patch applies to trunk without modifications +Will check in a change shortly +Thanks +Updated patch +This is what the attached patch address +Closing old issues +yes i am working on dev you could use ssh to connect +this patch will make aeua filter working with not set as the native good boy +It also includes the test case that Tom great! Pushed in d. +Committed this to branch +Hi KatheyAttachment please find the v patch that addresses your commentsI tried this on and the error messages you reported are resolved now +I wouldnt oppose the use of a base class but I dont think well suffer much without it in this case +I dont disagree Joe but I feel that the mapping should happen in the config file +Hard to believe this did work before +Committed at subversion revision +If they want to allow READUNCOMMITTED isolation why not just enable this on the connection +Committed a fix into the trunk in r +Could you please point me to it +Please see my comments on review board +Here is the change that I made to support a type of float in the statements +Ill take a look +It would be better to have a second activated versionthat includes the semicolon +Perhaps interesting issues can be moved to be subtasks of this issue and the ones made to either full issues or marked as no fix intended +Now that work is well underway with AS all previous community releases are +Applied patch in trunk r +added license informationfixed failing tests +does not occur when sending requests after changing the maxSize +With our current design we do not support conversion of bytearray from UDF to Pig types +Tweaked how where one stores the answer to the question about adding the Groovy Nature +How it sounds to youSounds good to me assuming youll add get as well +I should have mentioned the name of the project +Include a dependencies missing from the maven build +Would be really nice to know someone with access to the main Maven repository so that they could nuke the first broken version. +its rather late to bump this to a new major release a few days before GA +libs +This fix will be in CR. +Failing in VM only not can you provide the requested info I would like to get all older issues to be resolved ASAP +Sorry for the spam my browser confused me badly +we dont have any timeline for thiswe will work on this as time permits and according to our needs +Whats the problem here again? Not sure I follow? What is the fundamental reason for any change here +Add test size I get a review for this blocker issue? Thanks +bulk cleanup of Jira versioning +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Its not a Sonatype decision its a Maven decision +Verified in +I can help with Forrest if you need it +But Adam would know that aspect better +Similar with targetNamespaces meant to link to the other as a clone of not a blocker +I agree with Curtis Paris +I committed the patch to trunk andThank you Tsuyoshi! Thank you Chris for the review. +As Result? Later we can optimize through a facility for registering stored scripts and referencing them by name in +Wont be able to introduce this for but this is a probably a priority for final +Committed revision Thanks Henri! +I would be very happy if it included some scoping so that configuration could be provided on a intra cluster inter cluster hostuser or job level +web +B is probably easier on everyone but I would say we simply cant do anything that might violate the consistency guarantee without explicit permission from the user +Mikhail said that upping to helped w but wasnt sure about trunk +That makes sense but the behavior of the window caught my eye as a bit awkward compared to similar dialogs in Eclipse +this was fixed in CR when we finally enabled the equionox transformers to filter out the legacy AS adapter properly +Disabling the plugin by configuration is IMHO not required because the Web Console is targeted at admins only +It takes the trunk work and ports it to back to Mx +this doesnt apply to since includes is branch ready to be tagged? We need WS build integrated into EAP by +It would be best if the new methods accepted a Date argument instead of a Timestamp +for fix this +Theres also another problem Ive just noticed +I agreeWe can use reflection to detect whether the new API is supported +This sounds like it is basically the same issue as covered by the same fix +The tag is not correct and I will change it. +the only way it can be fixed is to provide installer with +Thanks Ning +Thats how I noticed the problem in the first place our CI server began to fail for no apparent reason during an mvn install buildI think how it works is the same regardless of which maven phase the unpack goal is bound to +Of course it is not necessary to output a namespace declaration if the namespace is never used on the other hand I dont think it is an error eitherPlease provide a sample XML and XSL pair that show this problem so we can see why the output HTML is not valid +The whole optimization package has been refactored so a few things have changed but the principle is the same of courseAlso this change will only be available from the release or you can test with the latest trunk of course +If it is intended behavior can it be turned off +Will try to extract the ddl and provide the report +It allows one to pass in an externally managed in the most generic constructor for the class +Verified at +Shouldnt we raise issues on MRUNIT to tackle all issues? I believe thats where it ought to goFor developers of YARN framework who wish to debug at the container levels the way is to do it via the IDE +Verified the fix on and trunk and closed the issueAlso tested that the test program indeed did hang on andOutput from test program on java classpath +Ah ok +Marking as fixed following Andreis commit. +html and +all modules generating a download file should now have a label attribute that can be applied to customize the label +and employees paid by commission and piece work workersThe rest of it sounds fine to me +Would you please help to try it +committed to trunk. +good +Nice to Have because customers ask hey what version of Eclipse is in this JBDSMax can you provide a screenshot of how a would fish out the Eclipse versionI think it is buried underInstallation then somewhere in the several thousand lines of text in that Configuration tab I think copy to clipboard paste into a text editor and then searching the text is the workaround right? +Being I prefer jobcontrol +ok tried again in a clean eclipse environment and that worked better +Yes and my point was that they shouldnt use different CSS classes +to trunk. +please aways put it on a valid release version before closing a task. +It just means that a xmlns default namespace is specified and so not every element needs a prefix +build seems fine tell me your doesnt work +I just committed this +This is the change I propose as a patch to the released distribution +V +In the More Actions drop down above there should be an Attach Files option +Introducing the export in progress marker looks like a good idea to me +This version takes into account whether a new region server has just been discovered by master +Thank you Raja. +Fixed in rev on trunkFixed in rev on branchI have uncommented the code that prints the password in clear test only if the log level is debug +The patch is attached along with a test case +Restore previous usage of setting in to top level directory of findbugs binary distribution +committed to trunk and closed with release. +Actually adding the it does work ok +First problem was already fixed +Also would be easier to plug when we move off MR +Yes its completely broken in cygwin apparently +I have made a simple reproducer with ROO where I have the same prolem +Merged this patch with +Thanks Daryn! +is pulled in under the lib folder but the still references +The testcases for the particular platformjdk combination that we monitor in CC runs you want them opened please use the JBPAPP project using the particlar platformjdk combination under which the tests fail +code used in Axis to call to ll give it a look this afternoon andor try to repro this issue using Axisthx AmilaJorge +I pulled the relevent changes from added the unittest from the attached patch here in this jira and its all good +HMS has the potential to provide the ability for cloud service providers to provide a clone of Elastic Map Reduce +Change Readme to reflect new Maven and JDK requirements as of Karaf +properties and what file values are you concerned about here? which use caseall depends on what context you +This would mean executing an EXISTS LIKE for every LIST returned +Maven is not maven The most voted on issue for spring is m support and they ll do it for spring while several of their subprojects already doPlease reconsider your point on m or leave an issue like this open so people can at least vote for it and give you an idea of how many users want it +It works best regards. +if its something you feel strongly about please open a separate issue for it. +You dont need to install a database because the whole point is that it cannot connect to a database +There was a discussion on the mailing list regarding this page and someone suggested that this should be include as both a page and a componentWhat do you guys think? +In were incorrectly creating a wrapper for a outer class +Sorry for the delay on this Ashutosh +The admin console depends on openejb so if you shut down openejb that will also shut down the admin console and its not exactly surprising it stops working +Patch looks good to me +I propose that we accept this patch into and trunk so that the immediate performance bottleneck on the namenode is fixedHow much harder would it be to add the new method to the protocol? It seems to me the patch wouldnt be much larger +Even if we do decide to use as the default behavior it doesnt belong inside the Strategys +I definitely believe both of you are on the right track +I did this quickly and forgot to make sure I created that test in the Integration directory +Still need to understand the test failureSince migrates from I wonder if the test itself should be maintained in TRUNKMigration from to isnt supported +Why do you think this is a bugNone of the inputselect form tags do have tis attributePlease provide more infos why you would need thistia its defined in the TLD and Weblogic pukes if theres an attribute in the TLD that isnt in the class and my app wont start up +time to fix it +I checked with maven built installer and it looks like help is working fine +left the close statements bare in CLH SSTW and so they can unroll the stack on exception +We cannot afford already launched long runnings jobs failing because the timestamp of jar changed +The timeouts are somewhat related to in the sense that when the AM heap starts to fill up from buffering all those responses it will spend more time garbage collecting and enough garbage collecting leads to unresponsiveness and ultimately timeouts on some of the clients +Wow good on your for finding this! +Rather I think what we would like to do is change the strict mock default of to the relaxed version so that additional calls that are made the dont impact the over all compliance can be added without impacting hundreds of unit testsThoughts +This change leads to doubled submits in the following scenario user enters something in a textfield user clicks ajax button change ajax event on textfield sends Ajax request click ajax event on button is queued no longer returning false browser submits form via queued click ajax event is send via Ajax resulting in second form this change should be reverted and the issue be revisited in Wicket think that the solution suggested in is not good anymore + applies only to the applies to the branch and +documentation updation will be done in a separate issue +Will close once verified in JBDS +the original patch misses pushingpopping component in invoke method +If we change it then it needs to be on all those platforms which we dont have access toIm happy to hear comments or opinions about this but I would close the bug as invalidThanks P open bugs to P pending further priority for bugs with a patch +Thank you! +When I was configuring FOP I mistakenly caused my own problem by setting the Java max heap argument +Dave can you check if the builder is any faster with +Closing this record since is now released. +Step I think we have to figure out what our process is for doing candidates and votes and stuffI havent looked at the instructions page here but i think it provides some integration with thisWe should figure out how we want it to work +But Id like to avoid having multiple as we still have a chance of getting this into +So in our case seperate implementation from API modules +Please someone commit this Im away from a laptop for a while +Full complement of files for distribution +I fixed the calculation and added a config knob to adjust the error rateThis time it passes the tests and gets s from ant + +Patch committed +It looks fine from my point of view +But callout mediator is designed to support blocking type of invocations +Or too much work +not sure why assigned to me +Added the described additional check which is always performed irrespective of any configuration option +Maybe we should provide more clear error msg to users +I couldt provide a separate patch until you applied the other one +I hope that the final binary release will be JDK compatible. +The config in the patch is named It basically does a nice n using the config before launching the containerAttaching the patch which Ive tested +We need to get users migrating off of it as quickly as possible +Benjamin Yes if Tomcat is managing the credentials it is impossible for to change anything +And yes other systems will not be able to decrypt it at least not after this task is completed +It is a feature of the new compact protocol this issue is discussing +Upload is parsed to an graph first so store is irrelevantWill look for other examples of unclosed sinks +I am not sure this is the right fix for this problem how can the root classloader fix a? I checked that going back to the configuration of the release of the jalopy plugin everything works for me without having to specify the root class loader +WW Update fix version for +Part of a whole bunch of JSF related issues that need to be tackled for +I suggest we can write something in release note like thisTwitter has deprecated Basic Auth and all applications must now use +JimmyReg the testing of this patch? Are you planning to test this on or before we take it into those branches +You could email me directly tsurdilo at redhat dot you couple emails with explanation on how to increasedecrease the canvas size +The patch looks good to me +Pig Patch . +Attached updated patch addressing review comments +Adds return true in the Iterable clause and a test case +Heres an example when wrapping version the Legal folder included in the original jar gets discarded +However there is one unit test failure after changing those default thresholds + +next step is to create a patch for amq trunk at apache with the fix and a simple jnit test in it so that the fix can get committed back to fuse +But if ICUs string transformation does the trick for all fonts we can use itMost of all let us go with a good solution +That being said it would be nice to extend the packaging concept to views to be able to organize folders under views in the same structure as controller packaging +Thanks for the insight +Heres the latest version +Ouch +So unless there are more people that will want this site to be huge the description will be added as the tooltip. +Implemented this guard in Rev +Please with latest + SergeyI just submit a patch to the +In one can currently try to embed switchyard into an application +Committed to trunk as r and as r. +Tim dont have a unit test yet but presumably can create one +patch looks good +Moreover the component manages calls with the deprecated keyRuby components do not manage deprecated keys yet +I cannot reproduce this +Im just building Pig with default options on my mac +At least that is what is done for styles and layer groups +files in tmp +i didnt realize that patch was not yet applied +I think we dont need to hold this jira for that +Ive been testing it under heavy an heavy upload workload that used to fail before that for hours and its still working going to commit +Merged into master +asking Chris to fix this +Hello ColinThank you for the reply and the link +in Eclipse with the Subversion plugin you in the Package view and select TeamCreate Patch +I suspect that we actually have a higher level vulnerability here if the method isnt protected eitherCan you provide an example of such a system? I can say with a high degree of confidence that the authorization semantics of the current patch are the same as that for the vast majority of the systems that Hadoop is currently deployed onMy inclination is to implement it like this and if a user of this hypothetical system which you refer to wants it at some point in the future we can make it a configuration option at the NN +Discussed with Nick prior to posting +This patches fixes this issue +Nothing can remove thisIn Problems View for each project Target runtime Server Adapter Runtime is not defined +All right lets assume its fixed we havent seen this one in a loong close all issues in resolved state that have not received further comments in the last month +Many thanks for a fast response! +Sample WSDL file that works using this import WSDL file that works using this import tested the referenced WSDL in Designer with no also tested the attached and using this import option in and it worked without error +Please reopen if this is still a problem. +Fixed on and closing all issues that have been in resolved state for months or more without any feedback or update +Regardless it seems like real possibilityShould we just ensure unique RS timestamps in master? Would that be easy +JohanPlease describe your usecase when you were resuming continuations in differentsitemap +Just to widen the problem I am facing the same problem with JOINED inheritanceCan anybody change the issues titletopic +Actually CDI puts exclusions into the spec +This greatly reduces the complexity as the probability of an user having a type called intArray or floatArray is very low +removed redundant null check +GETMBRSRSP could track the number of GETMBRSREQ so in the exemple D would know A B C already sent GETMBRSREQ and are perhaps trying to elect themselves +JimmyJust curious to know +Build log for jee example under Seam there seems to be a lot of different references to versions and and report from the build +I will help you reproduce the bug clearly +The control from the original binding is placed into a scroll pane and then that scroll pane is also bound to the same property as the original binding this is the expected you can not provide a failing test case for this code I will close this issue +The libraryitself is not distributed with FOP you have to get it yourself and perhapsrebuild FOP from the source refer to the documentation delivered with FOPThere is no guarantee that Jimi supports the TIFF subformat you are usingrefer to the Jimi docsAlternatively FOP can take advantage of any JAI conformant library you cantry and get such a library capable of processing TIFF imagesApart from this not working is a bit slim for a bug report +Anyone fancy doing thisregardsMartin +Ill take a look tomorrow +X can easily support HbaseWhat are the other methods that requiresNicholas Let me get the list of Habse using currently reflectionsI just seen the comment for solving one issue in Habse +The thread dump shown in JBAS is exactly the issue that led to the creation of +Thanks for the patch Benjamin +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +The version move was part of a general move of lots of for housekeeping reasons +I thought there was a separate jira for this but i just did a quick search and found one +Thanks for the patch Max. +adds sslsocketConnector accessor along with simple default sslPassword and ssl that will solicit the default jetty behaviour of looking for jetty system properties or prompting the userIf your config needs are simple the accessors will work if they are more complex the ability to set the sslsocketconnector should be used +Thanks Gregory Ill try out your example shortly and see if I can figure out whats going on +Also adds the default support +It may be better to continue this discussion on the mailing list if you didnt mind with a link to the JIRA then the user community will find it easier to benefit from this discussion +Right so Ill create a interface and I can look it up based on the configured versionPolicy parameter +Looks good +I fixed this in revision by adding a check for an unexpected property offset value +Patch and regression test were applied at +Cant reproduce it locally in tycho build but it seems to be a problem with the test +Verified on EAP +available in the public downloads at +If it cant be assumed to be provided transitively then shouldnt it be treated as if its compile scope for the transitive case +instead of buildtest its buildtestA or buildtestKworks like a charm no +So this should be able to be called from the other samples +Functionality added and tested +Not being able to get coverage results reduces the utility of +Thanks +So a surprising compile error will occur if the methods arent all there + +on makes the two tests pass mt Dtest mt Dtest I noticed addendum changes s dependency on from compile to test +Running retroweaver is trivial and the batch file for doing so is included in the distro. +This patch will suppress the two flags which differ +Verified at rev +more test passes +The patch appears to fix the problems Ive been seeing where errors take down a regionserver but the master gets confused and fails to reassign the affected regions +The code only be functional if computer is using ccmp module so it shouldnt influence other parts of VCL +Doesnt hadoop offer skip failed row options? +suggested to come in late the patch has gone stale +The fragment exports I needed to create this fragment and export these packages in order to resolve dependencies on these packages as I kept getting a missing contraint error on themfeaturesaddUrl mvnmxlfeaturesosgiinstall s mvnfeaturesinstall odeIt when I start step that I am now getting the error on missing import +The project need to be built for hibernate to actually be able to load the classes including the file. +Looks good +Committed to missed adding and removingFixed in addendum patch. +Taking a few bugs to work on. +Thanks. +I have not reached to any clue why the finalize method is triggered on the +Thanks Yin! +I dont understand what builds is about here? builds would still use the SVN repos correct +Marking this as an enhancement since the original intended usage of theseutilities were to format Strings rather than binary objects like Dates +A null array an empty array a bad set of indices are all the same in most places in the stats package all violate the fully specified API contract and lead to informative +Patch that allows improper integration. +What would be an improvement is to add a new option to ignore unresolveable properties like the ignore which Spring has +This is a much better solution for my case +I found a bug in the command in and am testing whether fixing that bug fixes our ability to exec +Its fixed now +I knew that unique constraints were defined on them yet I did not envisage that they were used for relationshipsI do not know if a lot of people are using this mechanism yet it seems fair to enhance our code to support this case +Hi DaveIf youre unable to apply the patch that I attached to this issue I would be happy to mail you my patched jar file that weve been using for a year or so now +this one seems to get lost +Ive updated the patch with a SSL test case just in case +Ive discovered two real problems over the last couple of day mentioned in other bug reports here but I would like to know if there are design limitations in using mockDomain +If you find further issues please create a new jira and link it to this one. +This is a big patch Vibhor congrats +Thanks for the patch Jev! I already applied it to trunk r. +The othertypos were more like missing commas parenthesis etcI know that these are not complete examples but when doing a cutpaste ofthe code fragments from usagepatterns into Java code which I then filled inthe details I found these other typos +Im going to be stubborn and try to write a test for this even though you ed Hopefully will do so this weekThese issues are so rare to reproduce in the wild that I think its especially important to have targeted testsOtherwise its quite likely well regress them down the road +Fixed in r for +Patch applied in revision +We can use SCM activity Jazz RTC now +also can we print a warning message if the main class does nto implment Tools? that way we can get rid of the static jobconf in +I will try this out after work and write here my results +Thanks for the patches! +back ported fix +I spent a few minutes on it before my morning commute and concluded that I wasnt going to be able to do it quickly and without a lot of research I wont know if Im doing it correctly +It seems more memory is being taken up because we use a global document to create things in the DT +Yeah I think this ship has sailed for. +Many thanks Andrejus +Ach sorry your name has been suggested to me by issue wizard +bulk defer to ve removed the changes for some issues that are also part of the x patch set those already have separate issues. +mainsrcmainjavaorgapachefelixkarafmainAdding mainsrcmainjavaorgapachefelixkarafmainSending mainsrcmainjavaorgapachefelixkarafmainSending mainsrcmainjavaorgapachefelixkarafmainSending mainsrcmainjavaorgapachefelixkarafmainSending mainsrcmainjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainAdding mainsrctestjavaorgapachefelixkarafmainTransmitting file dataCommitted revision +The MD is Just to let you know I have not forgotten about this +It seems that better handling of super calls eliminate the need for circular dependencies +It passed hadoop tests +new test to cover the edge the change myself wo review. + has been modified +LincolnGoing to work on this now so please assign it to me when you have a chance +fixed warnings on win version slotsoccupied variable removed +From what I understand this is now fixed reopen if I misinterpreted +adds support for passing other value types by reference such as double short decimal etc +Deleting a row would add the special value to each columnfamily for the designated rowNot pretty +My case is pretty easy and Im interested to listen about your case. +I am reluctant about changing it now since it is usually not a problem also not in batch processes +reopening to change the component to Module API +The is considered to supersede the. +Moreover it has no direct dependencies on xalan implementation t make this to uncompleted issues are now moved to the new To be determined for future releases category +Does it seem a reasonable approachI did find an elegant solution to this issue in the scala script that launches the scala interpreter +SijieJust fixing the computation wont adress the metadata where all closed rangesI understand that you have worked on it but I am not sure if your patch is closer to the solution rather the other way aroundTo be clear I started with adding start in the switch is work +Well doesnt Hadoop MR do just big huge chunks in its shuffle? Were not doing any worse than that are we +Looks like information on grails website was not the source to M seems groovy and grails M seems groovy based +This should also take care of the logging output issue +Gert could you review and apply patch +It ultimately only impacts that specific null element case +Suppose we have projects A B and depends on BB depends on project A declares a service X in the source folder in the project A then this service is not available for project C +Then the datanode decides to send a block report and will report that block to the NN +Simply commented out all code in the method and the error goes away +Fixed for Crowd and Ldap by updating to +Ah right I assumed that is using +which makes parsing really really hard if there are dots in a table name +yeah NETTY javadocs really kick ass +The problem is that I dont have a test case +Attached updated patchWhat done +on the server side or in Eclipse some where +This is a new patch I have solved the broken link from the keywordsearch to the pageThanks in rev +To run this run Chat on each node and then run Probe +Can you attach a trace and possibly some additional relevant code? have attached my test case +We can open a new issue to cover Nonblocking servers +The backtrace is always the same though I havent confirmed whether or not it happens with the same data or after the same number of rows have been processed +Were there many related fixes in? I would prefer not to the code base at this pointIf these are simple pom fixes and some reflection here and there I am all for it +Also does not crash withfirefox syncfirefox +As does in the +I think youre right reading the code agrees with what youre saying +Gabe can this be closed +Greg any findings on your side +Please provide more this bug is already resolved my last comment has no I meant is that you could expend a lot of time debugging a when testing but the actual problem is not the lack of that class but the extra whitespace in the line within the for resolving. +jbpmplugins +Hi VinodCould you describe your fixes to help me understand what you did? And howto check the fixes in the src filesThanksDu +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I would go a step further and change the warning to recommend the use of as a possible solution +Its a moment I haven not looked at that +This should not have been closed until it was proven to work at least once +Thanks Andrzej +Reopen the patch as we would like to add regression tests based on code coverageDevin would you like to come up next month with a suggestionLike aim otherwise explain why this can not be reachedThanks take care of this +While this test uses a Java client were also using STOMP clients for other languages in production so useCompression would not help thereApart from that its just the test that uses text messages +dont close issues in the future until they are released +cannot recover missing data if the data disk blocks become corrupt after recovery those records will be gone +no Leif I think you can mark as invalid +So if you add any dependencies to the core module in my test case they will also be unpacked +Regarding perhaps rather than having the magical streaming directory we should just leave files that are being streamed in temporary status and kill the subdirectory +EXECUTE Create bean as in +I missed one part in +Our continuum builds are fine as well and I assume they run from a clean target directory + patch looks good +when you want to delete a user from the usertable you have to clik on the delete link in the usertable and then you have to confirm it with submit button +Thanks wasnt sure myself whether the annotation was required in but I figured it cant harm to have it there +The string conversions to and from timestamp can all be done in the query layers itself +That said we still should not count on the or behaviour because underlying committers can be custom committers that have different behaviours +Committed to +You can then assign carriers to the suppliers FACILITY +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Since zookeeper is kafkas source of truth we probably want to see an exception if the zookeeper data gets corrupted due to some external reason +Ill apply the patch and run the tests as a first checkThe potential difference introduced by the patch is whether a nested transaction is used or not when accessing the data dictionary +Venu Is passing? Looks like in the above comments you mentioned that is failing +Lets keep moving and fix the issue first +Patch committed to trunk +The application is waiting for all threads to complete before deleting the threads +exactly what wording does require an entry in the NOTICE file +Igal thanks for looking into this +This was fixed together with the other SSH bug +Can you provide simple applicationexample that demonstrates this problem this way it will be much easier to fix it +on both patches +Regular and persistent datasets were not affected +Im ready to check in as soon as JPOX is ready with the new methods + to trunk +Tim I made issues to cover your suggestions above made at the end of your Tim Sell Aug AM comment +simone any idea on this file? maybe you have it locally +So should be changed and written back in this case +Here is complete log file from the instance foo +Building is SFSB and its bean types are and +The real question here is why is in the first place? Something is clearly up with that +so what happens if ann john and peter are rendered +Im continuing to add new capabilities +is a command line utility that can read and write SPARQL results translating between formats +Please note that this item is related to the listing of the service in the Axis management went tru all the and could not find any so can you please point out the which have the problem +Thanks Igor for your immediate responseAm I the only one facing this issue? Let me try that again and come back with a quick start +Looks like it is a general problem with cyclic dependencies between projects and bean config sets +ready to go in +Ive factored out a corresponding create for JuergenNow I wont have any more serialization of Hibernate collections problems again. +weeks and one day after the initial bug report a new snapshot has been how many weeks have passed since Dennis asked you to provide a to nail the issue down? I guess you had something better to do just like the committers +Hi Can I assign this issue to myselfI use pdfbox and have extended itWhere can I join the developers group so I can commit +As you said I still get the markup filter errorI really enjoy working with +added the necessary help messages for encoding and iso patch from Tommaso Teofili is applied now thanks. +I just committed this +Basically I just removed the from the soapaddress in the and added a copy statement to assign the correct EPR address for the in the BPEL file +Hi Richardcan you please attach a screenshot so we can see what would be wrong +Could someone who knows the optimizer comment? It would be good to know if a select that looks like the delete query has a problem or not +This patch fixes the issue +Theres no way to avoid it that I know of +Could you try with another maven version? For eg with +Gregory patch verified at r +Please synch up and confirm it works for you too +The volume chooser is pluggable +Applied Hirams patch. +And Can let the two blocking job go parallely? Oh +iterates using deleted docs instead ofcalling +Same for trunk + +I have committed this to trunk and +This looks to be the correct fix forCan you please check this patch introduced one findbugs warning? is unused now +dct file and how to convert from text file to dct file? +Hi Junping the posted patch is quite different from the ones in and HDFS +Please review +Sorry +Maybe someone just guessed that the typical enabledisablecreatealter table operation would fall within this timeframe I dont know +I will commit this patch soon if nobody has any objections +that the fix works with the latest. +Ive attached the code generated by version +Ill try it against your version after and then we can go from there +Also I changed to job level configuration in the test +Try Hudson +and are added +Verified +revised! Thanks Roshan! +just committed this. +Custom colors and breaks are still not persisted so that wont make it +Probably no progress but heres some inspiration +Right and Im wondering wether karaf should reuse the provided feature or continue using its own +I dont get how one would do renames in this solution without either specify something like a migration in the file or having seperate unique ids in addition to names that are shared on the cluster +This information belongs in the Component Matrix available to the public +Yes +There is a spec now and the impl mostly reflects the spec +If you access a JSP page directly theres no action behind it IN works via the backing actions implementationIn other words the tags depend on their being someone on the value stack that can do localized text lookup +I cant upload our project here and you dont want that either I guess Perparing a test case would take some time perhaps in Nevertheless I tested the patch on our project and the particular Exception went away +It seems with this change we will begin to serialize and in our plan and will send it to backend which is not required since fetch task is executed from client itself +A patch based on the JDT Java editor +It may not all be relevant anymore +It came for free when we moved to later hadoop +Once this is fixed the package still does not build due to real errors in the rules but nothing as previously shown as far as errors goesHope this helps +committed on and all issues that have been in resolved state for more than one month without further comments to closed status +The patch is attachedMay I commit it +We need to make sure we document the WSDLUDDI and WADLUDDI and the link to these from the GUI +A currency converter sample to convert RMB to USD HKD JPY EUR GBP which uses AJAX and JSF techniques +This is expected since we removed parts of the ESB in order to slim the configuration +I think it might be misunderstood functionalityThis search panel requires Text in the Search Text field if the Where toggle is checkedEnter some text and see if it enables +Closing. +Works for me +Ok +Testcase reproduce +I think I need a test data set +Ill try to add some tests in this area in +does it possibly stop others from creating other tables with names similar to this one +This issue should not be back ported as it could cause a change in performance for some applications that expect the zero estimate. +I have applied itPlease open a new ticket for the missing multi thread processor otherwise we forget about I think the Config object could be a shared object for all kind that uses the resequencer aggregatorSo we should consider having a shared config object for this in Camel +We need to document the upgrade scenarioAlso does this need to be into master +Fourth patch for the new style to the Party forms +trunk When we have a new async API in place then we can cleanup the last pieces in Pipeline andI have added TODO in codes where we need to re look later +I just committed this +Are you sure that you arent just seeing the same screen output as the previous command which appears as if nothing happened? For instance if you type osgilist the listing is so long that the prompt ends up at the bottom of the screen +TPS without this fix threads Rename Delete +The new doc is much clearer on how it should be +closing for release +Thank you Arpit +Fixed at the revision +I committed the patch +You can go on +Alexey please verify +Anyway Ill make the changes needed to pass all the tests it might be interesting to keep it in an experimental approach section maybe in the wiki +In other words it is similar to a FIFO scheduling +However after I tested the rollback methed cant destroy the messageit worked not as JDBC rollback +I cant reproduce it either +Global improved defaults for +same issue +First cut committed in r +Yeah but that will actually break things more +Well this patch add the append option to the write functions +it is indeed helpful but well need to do it all over again for osgi. +updated patch instead of looking for output using the previousOutput output previousOutput via the comparatorrandom test passes its possible to speed things up thats great +I think is the right solution for performance and forking as were doing now is fine for people who dont care about performance +so just add a line to your start script if you do things that way or add it to the install service script to make sure that that directory is createdas soon as you have created that directory things are then perfectly is an issue of the script not being able to find or access the temp dir you need to ensure that its properly configured and accessibe for the logged in user. +This patch is made against zookeeper which could support IBM JDK +Committing. +If its a reserved word in Mac then the error message should indicate that +If we had to use a blocking queue we would need to add those items into the queue here the semaphore acts as a lockcount tracker +Need to be applied to tomcat source +Clicking on question mark the label showshidesLayout is still working progress + for the patch +Its working on Firefox but not in Chrome +This ticket has now been completed and the jira notification scheme updated +bq +Based on the overall lack of interest and the fact that weve still got issues bundling infinispan query lets mark this as Wont Fix +Since Javadoc is not in class file it is unavailable to Enunciate +screenshot showing Giraph logo +I guess that it is because my simple test does not pressure the JVM garbage collector enough to call the finalize method of +I made a change to the configuration in a running server and still saw the same issue of creating multiple directories when only file was published +This has been asked before +Im going to look into this shortly +BTW Asanka arent you a committer now +we cant do stuff like this + +Ive updated the qdox version to the latest release some quick tests with my projects didnt reveal any problemsCould you please check if your problem is solved and then close this bug? your fix corrected the problem +Empty platform log files removed in revision of quickstartscommonwsmqrouter quickstartscommonxsdvalidation quickstartscommonJAASaction quickstartscommonswiftxmlaction quickstartscommonxmlswiftaction quickstartsembeddedhelloworldbridge message +Any idea when you think youll have time to review the patch +Looks like a problem in Axis as the is not able to find services in the WSDL +Might be +Weve been broken for days now and thats too long +Attaching the patch with test cases +Set Affected Version to login has been added. +Patch for the integration between these two projects +No will still be on +When a new topology is submitted we want to request enough resources to to run that topology +The new feature in should eliminate the need to correct the HTTP we wont fix the JBR HTTP support to address this issue. +FixedPlease open new ticket if theres other suggestion. +At least not that I can tell +somepackage and have that html file get inlined into the tags +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Weve recently upgraded the markdown implementation toIs this still an issue +Do you have an example? Keep in mind that JSON Objects and Arrays are not typed so contents can have any types +HiThanks for your response +Thanks for noticingbq +Add time update of issues with fix instead of the correct +Updated the patch +protected field and its setter is protected there should be no injection +A B C D E A propose v B C accept it then A B died +Bug has been marked as a duplicate of this bug +It is a simple test case that check the vertex degree. +Added more documentationI think this collector is ready to be committed +im not yet familiar with svni use a new field because the definitions field is subject to be null initially and in that case synchronizing on it will cause an exceptionat least i think it will wont it +As such I will mark this as Wont fix +Attachment contains thw same minimal project in two steps +Excuse me if missed something but such a decision can only make sense if modeshape repo can work without the problematic DDL script +Thanks +Forget git add get sad +Done +Unfortunatally while i was testign i have found another bug that i have report in item +Attached patch is the fix. +Used a flag to mark calls to as open a new issue with an example to reproduce +The attached patch renames the parameter both in code and config +Correct +It might be nice to have a keepStubstrue flag on groovyc to assist with debugging. +Although the negative retry number is misleading it seems to me retrying forever when allocating a new block is a policy when a previous block in a file being created does not meet the minimum replication factor +Than we need to preserve an order of the migrations that we accept from remote nodes otherwise we dont have sufficient information to apply modifications or am I missing something? Can you please brigly describe the process how you see it +Also will be movedRest will be handled during refactor the Matrix structure objects +I dont believe the bundling project should be responsible for checking further +Vladimir special thanks for fixing the bug in +problem is on our side our GWT Integration references preferences categories by ID but doesnt declare dependency to the GWT bundle that has category definition +They are used in as well but this makes sense +You should not call cloveron twice +Committed to trunk +Targeting for the next version +So I believe it wasnt caused by the patch +I can then swoop in and update the documentation and commit a few other tweaks I had in my patch +Is there anything in particular I can do to reproduce this issue Kev +Created for backport. +Infinispan is our POC for Git support in FE +Bulk move issues to and +In theory can work on any so I guess it really should be named +Straightfoward patch +Does this look okayComparable can be left to a subsequent patch +Includes RF but Im also seeing it in RF can you please try agin to in metamer with the source +patch for for chance to make the latest build with the patch applied available through the jboss snapshots maven repositoryShould i create a new issue for this? +I committed the change +Fixed in CVS HEAD +Maybe I should host it on codehaus. +yep there are fixmes in there that are all you +Running pig example verifies that the change is fine +I thought we only ran this for creating final distributions +In a previous version it was definitely possible to have Aegis on the server side and then use the dynamic clientSomething has been broken thereDo you want to open a different JIRA for thisCheersTog +Of course there is a hook +I believe my understanding of moduleTree is OK I wanted to make packed only inside war as contrary to ear deployment where its fine to make it unpacked +When there is no that handler runs before my gadget handler and the templates are available +Doing Open With isnt supported currently but well work on that. +seems kind of weird +Im aware of this limitation but it is hard to solve so Im not really willing to solve the issueWhat you can do is use the to define the columns on the Entity and not the id class +Was that the question +And as we need the anway we can also keep to maintain an ant build for them +This is resolved for our initial release +I would avoid assuming the files are encoded +I ran the test manually and verified it fixes the test failure +This is going to be a know issuerisk to we will take up a better solution in with soft links +I attached the file necessary to reproduce this problem +screenshots s not a bug calendar out of preload rage renders dafault markup +So in this case when the application is undeployed we should ensure we remove the key and after that all the values will be +It is not my code +This issue occurs with have added a fix for and users to be able to create the default repository now +see comments at this task +Im adding a comment to the properties file to highlight that it works fine with SBM and Im also adding other properties with default values and information about them. +We do not need to bloat hive with not invented here type code + That is some nice pedagogical doc Doug +code delivered +I have been developing a dictionary for Ingres and have just been waiting for the JDBC driver to make it into the central maven repository before I submit itIt should just be a week or so now +I can submit for trunk also +Attaching the patch file +Thanks issue is closed now +As far as I can tell the fixes for also fixed this problem. +Thanks closing this issue now +He has enough work on his plate as it is we dont need to add more +New Component struct Class with controllers methods Template with one html text to show the selected option and one hidden with the id of the object to implement the comunication ajax and show the dinamic contentPrincipal Parameters Encoder ti recover the Object to show all the options +and were implemented in +attaching the updated patch +Fixed as of revision should only be available on +Ill post updates to the patches on here when thats more or less done and I have a chance to port them forward +Steve in your sproperty tag you can specify escapefalse That would fix the ampamp problem +Full information on the error log would be useful +Attached document outlines some requirements for the accounts part of CRM +Could you provide some test cases for thatThe easiest way to test this might be a test bundle that is loaded by the launchpadtesting module and provides some additional services that can be tested by the integration tests +This patch should change the generator to cglib and moves the libraries to cglib +Fixed as part of closed as resolved and released +Now that weve started working on some other projects that use maven artifacts we are all rolling back meclipse to use which is very sad because fixed some other headaches we were having +The test issue with is being fixed at +I thought they probably should but didnt think Id end up so luckyOnly one cloud test is failing when I dont force a local fs based dir for it Ill try and look into why soon +looks like the is generated from someone with old SVN client Should we truncate the message to only show the version number +It comes from a KAME import years +Im reopening this because it seems like the patch only works for the development build but I still see a long list of log messages on the binary releaseMy guess is this is a classpath ordering issue generated by the fact that we have many files in the project +this was unassigned assigning to you +Perhaps this should be filed on the bug tracker at +Same patch with Increased clusterhadoopmapred aggregation duration from past minutes to past minutes to past minutes to past minutes +Patch it is just the other way round According to the wrong version was downloaded +FYI the future version of the that will include the binding file will include better error reports +Committed thanks Namit +This happened during the client testI tried to debug it quickly but besides from noticing lots of rather small byte arrays I couldnt find anything +Im adding this to the milestone +Just your patch and the believe this to be fixed now +Remember to set the component field when creating issues +reopened just to set the resolution to fixed +everythingThe problem I see with that is that youre opening pandoras box and you allow for weird things to be made easily while the solution is of course to straighten out the repositoryI dont see a repository as a non versioned thing because I know that metadata descriptors are never truely done and requires polishing over time +OK saw this exception now will investigate it now +Hi StephanNo its not correct +Note that Unified Push Server should support paging in the API in order to be possible for UI to use it +Thanks. +Thanks Ed Kohlwey and Charles Reiss for contributing this patch! +txs +It seems like an easy fix but we have to make sure there are no configs that actually needs a trailing space in the value +The recent multiple inputs changes in seem to have fixed this bug as a side effect +Granted there could be compilation failures due to headers and stuff but we could sort that out +The change from to mavenshared has fixed this issue. +will need a lot of unit tests to accept all this +I wrote the code +This caused a performance issue as the jbpm saves all instances in the auto when it closes +is about converting Get and for an good to me +Should we keepEven when someone writes their code to work not all implementations may be able to provide readers +Patch that includes Bertrands suggested change as well as an initial approach towards building a CTF regression test for this +I just committed this +The version number of is located in the fileWe require all libs in Lucene to be named with version number +Im just cleaning up and Im closing what we dont plan to implement as opposed to leave it open forever +Its kinda obvious this stuff is in the localization file so obviously it appears on the display as well +Will commit as soon as is released or attach it to this issue if that takes too long +Maybe there are some stale jars or some other garbage +We might want to revisit it in +Reopen with updated stack trace if the test fails in subsequent runs +To avoid the above scenario i have tried to use the znode version that comes along when we get the callback after transiting the node to OPENED state +Thanks got the idea but still Im not able to get this work Ive tried adding the exceptions to and to but Im still getting the warnings +The default in the project has been deleted +See and try using Tiles +Will commit shortly +I end up using as a multiplier for heartbeatInterval +defer all issues to +Created and for that. +The corresponding lists the files touched +The official logging of this issue was also logged based on trying to control the directory structure for how the target is writing and hopefully that can be simplified and fixed as well +and here are many file types that we can not filter all of them and maybe there are some bugs with tika parser with some file typesso we can filter them in from valid file types. +Fixed at revision +Could you please that youre actually having the JPA API on your classpath and no old copy of JPA around anymore? have double check +The reason i did not change it in this patch because it is not directly related to this issue and could conflict with another patch that I have outstanding +I forgot that we need crawlparse to collect new +We need to explore further post when we start to look at the revised parser longer a priority +This would not be fully backwards compatible since existing in strings would need to be escaped +The patch has been appliedPlease verify +Not returning these exceptions to the client would cause the client to violate the JMS spec +And we will continue to research this as we move forward +made it local +Thanks Patrick! +Since the appclient jar type was created we need instruct how to specify the also to appclient type +Fixed in Branch in both SFC and FC +Im going to commit this momentarily +so that developers know exactly which datatypes that they can rely on Roo to handle by code generation and which datatypes they are responsible for handling them self +karafroot featuresinstall Error executing command Manifest not present in the first entry of the zip mvn hbase we need osgi wrapped library as its not osgi compatible out of the box +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Do I need to do something else to see this issue happen +Other than that this patch looks good +Priorty was inadvertently changed to Blocker +I will go ahead and apply this approach since it seems like everyone approves +Since this isnt limited only to spectral clustering perhaps we should put this somewhere more general in the wiki +fix verified will be included in for release +If its missing optional channels this is the defined behavior +I am going to try using an older version of xerces and see if it resolves the problem +I will put back a fix for that in the next release. +I believe this to be a broker side issue captured in +I have modified as you suggest but the bad behaviour seems to persist +It would be time consuming to iterate through all metrics and see if they are for column families that are still being served +Using Solr commons connector is not an option for us. +Fixed +These files add support for class declarations to Boileplater +If you still get ERROR orgapacheregexp Could not compile stylesheetwe have another problem +Will verify +Calling setFields on a forces the cached hash code to reset to +pushing to next release +i believe that this bug has been fixed so i am marking it as fixed +So for now I will move it to instead +Patch script reads in a language xml file and sorts the translations by the language key +JBC is integrated +I might be wrong thoughIve seen that asynchronous client but never used it +Since the available reference mechanisms in the specification all rely on a token that is not embedded as part of the actual XML digital signature the token can always be protected using the STR Dereference Transform or directly referenced from a dsReference when a wsse is embedded in the header of the messageAs such the original use case for this issue is handled by existing capabilities now that is resolvedIm resolving the issue as Not A Problem. +approval for is present for +Sorry for being late on this +Fixed online snapshots as well +groovy into +all further developments should happen on trunk +I wasnt able to find any so we are good at that if I didnt missed anyThanksMridul Pathak +Youre encouraged to look at the test failures but Im pretty confident theyre unrelated to this patch +Initial line and header encoding needs to be configurable. +The memcache makes a snapshot when it moves current memcache aside replacing it with a new one +Zip of a maven project reproducing the error +It is quirky since the behavior change is primarily the absence of a warning in the log +Common code moved from RNCN to WFCN +I also noticed that you have marked as optional but it really isnt because it is used in import statements that can easily get accessed via the public API ofThe last part that needs to be completed would be to update the interface to use generics and varargs +Do we really want to have it back +The underlying file system has sufficient space available to commit the changes +This can be used to allow multiple copies of the same application to log to separate log filesThere is also an example appender in the that allows event specific patterns to be used in the file name +But maybe an infinite density should be considered a fringe case and removed from the average +Ilya Has someone committed this code into the repo +Only Sales app has to be reviewed I will do soon +Im using Maven +Im moving all bugs out to and we can move some bugs out to to get some release action going +Would you please tell me which version of Thrift youve applied you patch on? Trunk orCheers patches are against trunk +uses a backend and wont have the bit FFI problems +I will open a new issue to remove all write methods from bytesrefthis is a ref class not a stringbuilder +directly returned by filters +Before for a case if query failed it was possible that some partitions gets inserted while others dont inspite of acquiring table level locks +These are optional settings in the schedulers config file +To reproduceAdd a column of type boolean to an existing classtable +Committed revision . +Attached updated patchThank youVasanth kumar +Manually tested +Tested this out in the admin UI and it works. +We will try to replicate +b will be cut out of looks fine to me please take a look and double check +The approach used is to add a separate to distinguish from the such that when the job is shutdown it will not delete the staging dir if its on its not the last think we are on the same page +Thanks! I did not notice that you had uploaded the edit log and image as well +Code looks good +Patch committedOleg +consider it closedAt any rate there is some thought in bringing Spring Social closer to various libraries +can sometimes bring the whole thing crashing we cant find some magic bullet by end of day Wednesday this will be punted to post +Assign this back to Javier +The fields are disabled as expected but when submitting the form the disabled fields are marked with validation errors but no validation error messages are shown +Since theres a it might not be necessary to fix right away +Moved the altered code in to +Test source files in a zip archive for considerationI think that expirePage must fail with a bookmarkable because clicking the link should the page +Well see what they can come up with +Revised outline Introduction to FUSE Services Framework Developing Web Services Overview development development Data bindings message bindings transports Developing Services Overview Data bindings message bindings transports Developing Services in Overview Data bindings message bindings transports Running FUSE Services Framework Samples Prerequisites for Running the Samples Sample Sample Next Steps the final version of the book on after receiving and incorporating technical review and copy edit feedback +Min has an idea about the root cause of this issue reassigning it to her to check +This is inconsistent +Is everything checked in? can i do it anytime +This has been fixed and tested against and the test now passes +The patch from had no effect but I did notice that to trigger the issue all I need to do is telnet to the ATS port and disconnect without sending a request +Ok that is the behavior I would expect +is already available in tableDecorators as explained in comments closing +Tests run cleanly for me +That class assumes the use of +It would be an easy matter to rebuild the jar file with a compiler +But I will not block it +I tested again on geoserver producing using Type Band Number is not valid as you can see in the stack I tried producing with Band Number is not valid Producing with and a color model produces an image but the image is quite green I think there is a channel the junit tests my test pictures have and I can produce any types I like +If youd like to port it to thats fine but I wouldnt consider it a major issue +As mentioned in the latest comments this was a problem with the users DTD definition not with Struts + version to Trunk since this is an improvement not a bug fix +The attached patch should solve the problem +Bulk closing stale resolved issues +I was not able to reproduce the problem with the enhancer and the Taks class +Is that a separate system and do I need to open a separate issue to get it to show up there? +Issue cannot be reproduced build a multi module project according to the POM inheritence youve mentionned A project module project projectproject has a description and is correctly taken does not have a description +I have added functionality to this code in the past and the patch looks OK to me but I dont think I in that much of a better position than you as I have not touched this particular piece of it +FabianI forwarded the issue to HenryHe implemented the current stuffbut I am still aware of it just waiting for an answerI come back to you later +I am attaching a patch which I believe fixes this issue +Where the code can be obtainedCan you please share information on where it was committed to and how folks can try it out +We dont check for the interrupted flag anywhere and I dont really see any methods that throw either +This looks like a dupe of and version of the site plugin are you using? You need +The region was not transitioned after the CLOSED transition got a call back for assigning it +Now I have changed the code considering +Tedi was doing some tests on table balancer disable and then enable ithowerver i mistake using drop table instead of enable Ctrl c is immediately sent and enable table is also closed behind then i found table is deleted but i couldnt create it anymore +HiI have similar problem I have deleted EJB Specific methods also also i dont have method description in suggest some AllAfter some digging I got the solution posting here if it help Problem was use of bad practices Resource annotation with EJB and I came to know that dont work with I used EJB problem was resolved. +Ive implemented handling for includes and excludes. +Recommended fix is to allow overwriting null header +Commit changes to At revision and trunk At revision it. +Correct the value map this bug after rethinking the whole thing it seems more convenient to support getting deep properties than preventing this +Just small change imho +I dont think it ever happens during execution but rather when the server is brought down mid buildOnce fixed the button can probably be noting I suggested this should be in the admin interface only +Thanks Max for processing the patch +Attached producer testcase + Great catch Sam +Jim thats exactly what I was suggesting +The patch +Patch was applied at +Patch available +Demonstration of issue using CXF snapshot obtained today from official siteAll libraries used added to lib inside Zip to ease attempt to reproduceThis isnt really a bug in CXF +I had Encoder extend COS for a while +The mere presence of the Grails jars in the classpath causes the problem however as I dont see what we are doing wrong from the AST side of things I regard this as a Groovy bug +Replacing jar to new one solve this problem +spointport +Avoid infinite loop in if out null by removing entries with null value from pendingCreates map +Removed singletons +Thank you test for me +Let me take a look at it again from the protocol perspective +Attaching v rebased and that remove the use of flushaftermins +Yes I am sure with correct classesI do not see the same with EAP these are ESB was sent to Ron see attached file +This is dojo without any tests and demos which reduces the size of the zip file by +Closing issue that was resolved several months ago. +The daemon flag is not sufficient as the connector can be shut down without shutting down the container to do it right the thread pool has to be shut down after the connector is shut down +Attaching complete log for from the console output +As per Brancos comment. +For very long classpaths this problem still appears on Windows +Security +Cool thank you! Seems like a fix +Ive svn rmd the relevant directories on trunk and. +Another option is to simply have a custom website after allIm thinking about it as well +Duplicates +Updated patch attached +Yes +After a long investigation i finally found why these kind of jobs where retired by the jobtracker +This document is ready for release +Of course its possible to have attribute explicitly defined if its in exclusions array +I am just trying to tidy thing up as much as possible if we do a shortly +committed to as ebdbfaaabffdbdd +The problem was really with the names of the proxy classes and in particular it was uncovered by specific ordering of events with producers and their products +. +Instead of fixing this bug I had just removed this useless optimization +I dont think it should be like this +To prevent anybody from deriving from Utils and overloading these methods? No its just an utility class with static methods and I made the constructorprivate to prevent incorrect the bug thanks Werner! +The topology is hard coded in the because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I dont know the rationale but I assume that this was done as as way to facilitate the RPMDEB generation. +How it works nowThe Server control has WTP server icon by defaultWhen you choose any default server there then the icon will be changed to the icon of the serverFor example the icon for TomcatSo if you choose AS as default server then the icon of AS will be usedIf its not ok and we dont want to be confused even if AS is a default server then we have to create a new icon for AS +Waiting for some feedback out of before s working great for me on trunk +Its not fixedLaunched within Eclipse the warning does not appear anymore but it still appears when the plugin is launched within Eclipse or the Apache Directory Studio RCP application +Ericaccording jira information almost all webdav patches are marked as Licences for inclusion in ASF works +And if you agree with me Ill fix it +this is the patch as file +All were implemented +ThanksAfter migrating I now have the test running again and no longer see the issue. +has the diff that was put up and has the thrift changes +Ok thanks for verifying +Unless we plan to have the log statements redirected somewhere I propose we remove the log statements and the LOG file altogether +i meant versions. +Hi RickI created and ran an ij script against my database but that does not reproduce the problem +Screenshots of the quality profiles and violation drilldown pages +OKI have modified ALL lookups to use theNow there is less code for every lookup and there are no more container tags usedBTW I find that the bug that makes the collapsed to expand again during pagination is still there dont you +fixed verify work on binary files +It is working now +And updated the wiki page as well +bq +patch for trunk +The the elements of the portal were not altogether on a sign in version when it is an application filled this phenomenon came in with the versions +Hi NicholasYoure correct it was discovered yesterday that the master branch of phpcassa was setting a default timeout of millisecond +Well see if we can do something about it +Perhaps we should keep replicas as a list and make sure that we dont add duplicates when changing it +Yes in my reducer +Added a new member just for locking now to avoid the findbugs error +Can you upgrade the database on another platform? If so does it work when you transfer it back to AS +Just so that I make sure we are on the same page when you say my diff youre referring to the patch in when you say that it is ok to remove it youre referring to clearing the database in loadData +It seems to me that it is supposed to fail in this case +So a user could clone a table and then tranistion it to FROZEN instead of OFFLINE +create a new issue I think and link it to this one. +Reason failed to create connection +Guesstimating hours +Done the github mirror will catch up shortly + +I confirmed by using objdump T on the +This patch was done on the latest further questions please ask on really Preloader is not +Also updated the code trunk as these changes came in after the trunk was made. +Servlet in turn has a references to servlet context which in turn can have reference to entire container +lgtm will apply today +with JBT. + on stacks comment was going to same about the same myselfWe should update the wiki if its out of date really it should just point to the +From Ted DunningActually jmockit uses byte code patching so you may suffer less reflection overhead than expected +As expected find size orgwwww fixed the path in the earlier comments it is not +Closing. +Hello AgainI have solved my problem here this was the solution +We already have plans to move PROJECTwww to PROJECT which more or less addresses your As to Im not sure were planning to do that +In the future can we get tests for new features to make sure things keep working. +I looked at patch again and see how I misread +Our solution has been to turn on the timestamp that will override the producers timestamp and adjust the messages timestamps to be with the broker machines time +Heres all the files used and the main mxml separetely +Test failure unrelated +Removing release version since theres more to do +which container +This patch simply alters to ignore parsing links if they are more than levels deep in nesting +Added another int test to validate this +I guess that needs to be added +If Im reading the MAT thing correctly its saying of the ram is used by the executor queue meaning that commitlog recovery was reading mutations shoving them in the queue faster than they could be consumed and appliedApplying the stage limits from the first patch of should fix this although the rest of will have to wait for +When a binding is used or can only have Columns as children +Dan Ill try to test this out shortly +Thanks for the quick turnaround! Ive tested the snapshot plugin and now the projects show up with issues on the dashboard +Attaching a patch for proposalPlease review and let me know improvements required +Please reopen if you still get the issue. +The second patch is a proposal of a modified implementation to support spaces in paths +Could you update to the Seam codebase? +Created testcase project after testing with just on the classpath and then just on the classpath +use incubator groupWe have started a vote to give karma to all incubators +Closed issues related to Roller release. +So this is a mere bug fix not an enhancement towards the out only MEP +Ive had this fix in production for one of our affected customers for a few days now +wasnt thread safe +First stab needs some cleanup and to port over to hive codebase but this establishes intent and is independent of all classes +This is now changed to be determined dynamically +Of course I am also still missing a test case as requested weeks ago +I wil commit this in a day or two unless somebody has any other comments +sbuild onbuilt on +bulk close of all resolved issues prior to release. +Sorry sure my fault. +Part of a bulk update all resolutions changed to done please review history to original resolution type +The bug is caused by an incorrect implementation of loopback +is fixed at thanks for committing the patch +According to the documentation for this class it can be used for a new or an existing property +In general jclouds needs to be better about retrys and especially API throttle issues +Can you please tell how to reproduce thisRegards you try to test adding a local provider in loop? This was while back and i think generic provider was not able to call local provider service +patch applied +Can you please try completely removing and Forge? This looks like an installation problem +Startup times for forrest run SVN rdate forrest run +Idea of adding was withdrawn so close off issue. +I just committed this +I did indeed notice the needless use of instead of and I think it that was in my patchRE wow yet another issue about the same problem that makes +Width attribute work fine +ready to commit +Fix is in CVS +committed +I now fixed the branchs and all remaining about the APINow the horrible stupid to port all test starts +We can afford the bytes IMOThose statements about reasonably synchronized clocks worry me +This ensures that whats returned is the same as whats returned by the Java client +Can those of you seeing this problem try it with trunk? We recently reverted some code that tried to access devnull directly and replaced it with code that simulated devnull by accepting byte writes and not doing anything +I also prefer the current report format where I can see at a glance the failures rather than having to go into each module +My environment was setup incorrectlyPlease close this artifact +Ledger slice alone does not capture the idea that there is a sequence of ensembles over time +This has been explicitly tested in Logj +Attached logs with the +Attaching +The vti can now be invoked with a constructor +It can definitely be enhanced as you wishHope it helps +call can be extended to call instrumentation in about the same places where logging is done there currently +Hi Kevin could you re test this issue against final? The failover transport should bre reaisnt those interupted and resumed events now +Attached the patch file that excludes as many of the Security dependencies as possible +For references sake also the patch against the branch sans the formattingThe two attached files areSee previous comment from me for details. +The problem is in generating select plan +The attached patch file got qpid trunk to compile in VS +Merged this to +This is just a remark but its so noisy and so easy to silence that we should clean it up +Ill review the patch more carefully probably make a few minor in the process and commit it in the coming daysOleg +Updated patch based on similar changes in is now a service Moved filesystem startup to the startStorage issue is a known problem unrelated to this patch see +Let me clarify Im not suggesting this be committed +We released M so I guess we fixed +This has been fixed by r which changed how the link is added +OGNL released now so no more worries about +Then the client side will retry the same NN if it gets a +Applied in r +It would be interesting to get some memory snapshots to be able to determinate which objects are being lost +update project to use common publisherNote project should be updated after suite integration +patch based on stop decommission is called it checks to make sure that the excess replicas are identified and put in the right queue for deletion +Fixes the DOCTYPE syntax errors +Ill check what goes wrong +Thanks +You can get this fix only by checking out code from svnThe fix will be in and when they are released +Could you explain more what you mean byHence when dependency manager works with it only knows about UUIDWhy does the the dependency manager is working with only have a UUID +Usually slfj happily functions with only a warning when you have multiple jars but I havent tried mixing andWe definitely should get rid of the junk that is pulling into toolsjar its shell scripts and other stuff +The problem is probably because server does not send a Vary header when it should for cacheable response +A bidirectional Order and Customer example seems to be a quite common example So why does this alread break? Shouldnt it be the responsibility of the JPA provider to optimize eager fetching? Is this such complex matter that this is technically very hard or impossible to do? Or is this simply low priority for JPA providers +AndyI cant reproduce this +Good point! I missedShould we just rename it to or some such? Apis like killTask killJob etc +Thanks Rich +I figured it was committed once you closed the m in no rush. +Please close +fixed in close for the release +So the problem would randomly happen depending on the state of the heap andor stack I just dont understand why this hasnt happened to anybody else +thanks for the rapid turnaround. +Ill try with latest version this is effectively a duplicate of and was fixed in the release. +Logged In YES useridI can deploy exploded EARs with but I cantredeploy them +Reopen JIRA for backport to a patch for which should be applied after +Committed but its a feature so I put it under . +Also please mention any edits to +st quick review Please create your patch from root folder and not specialpurposepos folder It seems a lof ot changes are not intended +It is harmless and only happens in the tags we create for release +Maybe we just need a default agreement ID per product store +No idea when it will be implemented +Given the hdfs has rigid notions of block sizes compression could leave the filesystem with varied hdfs block sizes and NN has to be aware of the varied block sizes +One solution is to make trycatch in getter method in patch for the issue +At least some of these fields have accessors already is there some reason why the accessors are not sufficient? More importantly when making private fields more visible its important to do a careful analysis of what the impact of external mutation to those fields would be +I think you need to check why unit tests fail +Moving to Server component. +Looks good +The plugin looks great save for the issue with +Should this fix be ported to the branch for the upcoming maintenance release +Im going to make a few cosmetic changes and commit this on behalf of binchen. +verified. +Actually am not adding any new HCD only changing the existing attribute +Attaching the test class +Added initial clean up +Is that right? If so is there a way to fix itThanks +Is it being solved +The patch deletes from +Id actually like to reject this one further research indicates that SCTP might not be the best choice +Heres a fixed version of the v patch +Maybe not identical but definitely a related issue +Object model scaffolding created and to follow agreed upon issue disposal to resolved +I encountered the looping once before was reverted but also before this was committed +True I missed it altogether +Sigggh Ill take them out +And if it turns out we cant do the first then lets not bother with the second +At any rate I dont see when this can happen a will move to FINISHED state only after the remote container is clearly doneexited +Kicking hudson +Ive applied a patch with version The patch ensures that the clipping path is never null +Are you going to add the IS changes here tooYup Im just working through the best way to expose the API in the IS while supporting per segment walking +Please contact mailing list regarding OM usage +The RHEL command should be correct already +Andrew Where is this currently checked in? Is it already in our master or is it in a branch that you have? Do you have a link that I can look at +Interested folks please take a look at the document and provide your comments +Opened as. +You shouldnt mess with them anywhere in your codeThe server returns valid JSON responses and is compatible with all JSON parsers I have tested so far +sqlex Now this one is interesting +The following components have been createdcoredocumentationexamplesThese closely align with the different modules within Click +Defer all unresolved issues from to +Let me know any problems with CR +Can you please explain what you mean or at least include a sample set of inputs and an output that demonstrate what you think the problem is +Please attach patch file rather pasting patch in the description +Ive attached a patch without an author tag +Fixed by +And I think there is bug with this in Hibernate Entity Manager +What part of the docs did you follow? If this is a documentation bug then we should fix it. +Adding to MAVENOPTS solves my for the this to the doc would be very helpfulthanks +So all issues not directly assigned to an AS release are being closed +Will there be workaround fix in YARN to get around this problem? I would imagine more and more users will try to run YARN on ubuntu and. +End users dont use the s parse methods directly and two are enough for engine implementors +I will link the issues so we can keep track of the correlation between the jira entries. +Fixed r +David I looked at but after applying the patch to trunk the offsets that are returned span the full length of the field the highlight tags any idea why? +See I updated the dependent validator patch for the current CVS head and attached it to that bug report instead +The most recent code is +My opinion is that we shouldnt delay the release for adding in both and that we can add in in +Thanks +Any ideas on how to proceed +Changed the previous patch to put SMX components in the deploy dir instead of hotdeploy +Here is a patch to upgrade OCM to JCR with running testsIt contains also fixes for all deprecated API usage but for XPATH and SQL +Ah I missed some stuff Invalid POM License information missing Developer information missingThe same problem exists for xmltooling and openws btwColmAs a sanity test I suggested deleting your maven repo and doing a mvn clean install again and checking that none of the Shibboleth jars get downloadedColm +the xml rule drl a bug anymore + Patch Available to get Hive QA run +only added one more testcase which has a dummy transform in a subquery +to a limitation in the implementation will only be loaded from the system level +Need to investigate more +Im more interested at first for what people think is the best general strategy for dealing with thisPutting tabledesc in a subdir only solves the problem for region opening +Oops just realised that this issue and proposed patch belongs with Sis financials package +No problems +Please try nightly of a bulk update all resolutions changed to done please review history to original resolution type +Yup +I had to turn off Auto Import in intellij as that was overriding the output of the formatter plugin +In Im not seeing any exception message +They still work for and below +Test to check performance of JNI +see +Lets test this on Nemo +This jar file can be imported into Guvnor where we can build rules using the properties from +No I think you are right +There are ASF headers now too +I added some new test cases and committed a fix +Would it be possible to add a unit test +Ive committed this to trunk and +EG in this case the reverse is not true rightExactly +the patch is good to me +Patch updated to trunkChanges Router and range for the current core is used during split Change replication to check only sub shard range instead of state Fixed replica allocation code such that replicas are not created on the same node always Splitting a works Slice state and range kept in is used only +I thankful to you and Nandana for applying the patches and all the folks who helped for this patch. +The patch for that issue includes updated versions of the code appearing in this issueLet me know if you have questions +Is there any full thread dump + I dont suppose you still have access to this db to see if this is still happening +dims +the patch looks good to me too +Follow described work around. +We could cap based on size where we yield the network thread after n are written on the wire giving a chance for the rest of the smaller responses to get written on the socket +Many of the changes we find will apply to those docs too +Theres a problem with the current patch it usually reports ps at the start of the thread +On million keys these are the resultsFST bytes bytesA difference of about megabytes +Attaching additional blocked on +Whats the status of this Alef? Will it make RC? unfortunately it wont +In a lot of cases the local installation of Hadoop Pig Hive etc have no corresponding upstream artifacts +Ive done some builds for x Linux +Sample works for me too! +Verified fixed at svn revision +Fixed in trunk +Thanks Uwe new patch attached +nightly so I will have to wait for the next stable release +bulk close of all resolved issues prior to release. +ThanksD +Anyone like to comment or give me any suggestion +This should fix the a committer could commit this that would be great +I implemented my suggestion in r and simply add a virtual task to the head of the workflow called name and seed it with the workflow global conditions +That looks pretty good committing it. +Recent fixes have been made to be sensitive to classifier its possible this would solve the problem also. +Changes in Remoting to and allow the socket transport to be configured to communicate with older versions +Improved from previous patch +Ill reopen it if necessary. +We add a python script which relies on these ideas to validate that cluster members really do have the same state +Actually the calls can be made to work with the attached patch to fix sync calls +That is not what any implementation is trying to doSorting isnt needed but it is probably the easiest way to split the data into training and test data +New translations available for Trinidad trunk +See for swapping and. +But they are mixed in with all kinds of other libraries like FEST reflection +Issue has been fixed with r. +That doesnt work for things like Birt etc +Well do manual tests for the main use cases +For example I had some concernsabout the which can be addressed much easier this way +If you have one it would be great +Targeting this for +Im not sure why you cant build +is this patch using +Thanks Hans +Porting the metrics change from to trunk +Hi ChristopheI never used this plugin but looks like it can check validity of xml file against an returning to original issue validity of xml isnt the same as compliance with POM Code Convention + +If this is the right way to go +Not there yet but no fault of your patch +Has to be revisited when we implement CMIS. +Update based on Mikes feedback on to use loopback IP +This is the second +Sounds good DarynI put this into trunk and. +Reply from Janne Jalkanenguitests runs the same tests as tests but it uses a different test runner which shows up a pretty gui +Completed At revision +The one that work depends on which filter is on topIll close this issue as re not bound to support only the spec and if Mark finds a way to support multipart without breaking the rest maybe its a good idea to have a look at itBest regardsGanesh +I had a dependency to in version in another project +The patch is simply to clear that latch on exit in all cases +we also need to encode using Text i thinkalso metadatatypedserde i guess we need to fix that as well +Bulk closing stale resolved issues +Ive applied it to the source package and it fails nicely and clearly +I agree with your path to fix but i disagree that not killing which have no issues is an improvement or a new featureIf you killdestroystop something thats working normally and you should not have its definitely a bug +The default value is +a is getting there but it lost some of the whimsy of juliens sketch and is a little boxy +I ran HDFS and tests using this jar they all ran fine except for the known test failures and you are right the ownership of the internal buffer gets muddled with this patch +I will be happy to fix or a to try to fix The test had the real work done out side the Timing The other thing I like about is is that there is less data saved in cache +for the to trunk. +If there is no more input on this before the codefreeze for I will resolve it as Cannot reproduce issue +I kicked the Pre Commit job to to make sure this works +Its not reproduced any more after I updated to the latest Jboss Openshift Tools +I get the same error message using the COPY command which doesnt even support the use of RETURNING +Puppet agent related alerts are now being exposed by the fix for +The findbugs warning and the test failure are unrelated to this patch +Pull request have testet the latest release the fix works and I am now able to see the JNDI tree +Ill take this +I didnt realize the GWT generation wasnt part of any of the maven plugins +There is no way you could be having JBM unless you base the build on an EAP build that is more than months old. +Ugly but combines reasonable reporting with reasonable security +HelloWould you please try the patch? Thanks a lotPS +My conclusion from that is that doesnt work correctly +Thanks Liu +I think this was fixed in +Fixed. +Close all resolved issues for. +how does you strategy looks like +Verified on. +Attachment has been added with description Image test has been added with description PDF that one is the pdf output +I havent used either but its part of the spec and would be simple to add +Thomas I did a quick review of this and Im a little bit confused +Although it is different from changing it is potentially from that there isnt anything else to fix here so marking it Not A Bug +Try adding the coverage profile +The file looks reasonable and our own as well as the CT tests all pass so I assume this properly works and I resolve this issue +Hope this works +I was using a old version of ADSAfter having installed the one I get from directory web site I tested again with Goran ldif file and configuration and everything went fineGoran could you just check that you didnt do the same mistake I did I mean using an old version of ADS +Ah I just noticed implements and +This is done by upgrading to JSF Deployer +If you had more in mind please let me know +Probably the solution here is to have an option in the groovy preferences page that allows you to set where you want the default directory to then in each individual launch delegates Arguments page in the Woring directory section we can force the default to be whatever the global preferences is +Last minute edit of the test case messed it up +just resolved this +You can see that in the command line I used +Thanks +First svn remove dependslibsthen apply the above patch +And this morning without changing anything it is working fine +I wrote a small unit test and I encountered quite the opposite behavior as described by Hui +mvm site generate the site document and mvn to deploy the site +Closing as wont fix +There should be big changes +You do want to close X because it has been removed from the. +Could anyone to run console testsuite test result after this patch applied? Actually to run console testsuite before submitting the console related patch would be recommended. +COOKIENODE was introduced in but before that we had BOOKIECOOKIEPATH so Im not entirely sure what the history of duplication is +OK I finally solved the problem +please review before somebody made big changes to the again +There are some known issues with config not being persisted until restart does it still fail after a restartDoes it also fail with remote reposArchiva is using the standard +I am reusing the code from which adds the same feature to snapinstaller +This jira does not break usual deployment people can still use combined deployment plan +this patch enables jdiff for zookeeperTnx +Freddy I added description to the rule could you please check it +Ill try to get him to put the patch here and see if we can get it pushed upstream +For now i am going to move this back to since its not a critical issue +append a patch basic idea in this patch is that when cant create jmxConnector from default serviceUrl then print out a message and use the local jmx server as applied with svn revision +This by the way is already the case in ASI also change the subject line to reflect this general requirement +Committed to docs trunk at subversion revision and ported to docs branch at revision . +A lot of things in this direction have happened on the branch that I hope to propose for trunk after everyone has had a chance to review it +It only adds to hfile meta info not to each key +jdbcderbyCUserspaulDesktopwidgetDatabasewidgetDBcreatetrue +deferring to of the images with the version from the numbers are as per the docFigure +patched to support attaching an authenticed Principal and a Group of roles to the mock servlet to support creation of a servlet request that represents an authenticate user with some roles +having independent properties is easier to understandIt could be more complicated because of this feature If this is empty only the defaulthosts are allowed +Forward this to Simon for review +Just wanted to note I just ran into the same issue with loadConfiguration trying to load a source file +Thanks for the quick review Dag +Thank youRev +POM with WSDL used for generation and imported schema file +Quickstart that reproduces +wonder if a flag indicating not to dispose the imagechain at this stage could be passed to format as part of its argument or something like that +I have integrated your patch modified regarding commentsBefore letting this into the branch it would be nice if you can confirm it solves your issues +LB rules are also created with port +One of the concerns in was the version of jline +The issue is gone in Undertow. +Attachment has been added with description The enclosed contains the and the pdf file generated from our environment +Once you set it to background though you only see percentages on the quick progress bar or if you go to progress view you can see file size information too +Isnt this issue out of we dont need a dependency on we have solved this IIRC. + +Adding debug to shows that WINWINNTx at the point of the error +Is there a work around for this issue +Hi Dave I upgraded to and checked results +cp dist izpacklib +You should ask bug or normal on the email lists before creating a bug report +Hirsync repo is now and the update interval has been changed to hrs +The total number is very useful and it would be nice to add the requestsec on the web UI as well +It also runs on a couple of CC boxes +There are now mechanisms for services to clean up when they restart +Verified at +Thanks a lotArjun +Please let me know when its ok to check in +r build was used to verify this issue +Any thoughts on thatIf we keep null then Ill fix to check whether did not return null +Sure the more hands the better. +Stepping back just bytes is often not enough for disassembler to pick up correct instructions beginnings so it can show bad instructions +The fix for bug also has no effect on this bug +Picked up for the first update to the release notes. +moving to next version as its not a blocking flow and would get some more time to give best way of handling the case +Indeed the changes are pretty localized and can be found in two places the data setter the save methodNote that I created two static helper methods getProperty and setProperty +A severe memory leak is caused by the use of in fixed by need a patch to only fix the memory leak issue without improving the design for maintenance version. +I hope this will help resolve wether this is a bug or the stylesheet is incorrect or the applet is incorrect. +Slip needs EJB fixes +When using on a EJB endpoint implementation class to point to a local WSDL file with schema imports with relative schemaLocation paths the schemaLocation paths are still not replaced by the absolute path in the published WSDL +provides a solution +I would then belive something was wrong but it is not +Works for me +The unit tests are passing on cruisecontrol. +Could you please submit a testcase so I can resolve this quicker? Otherwise Im going to have to queue this with other work I am engaged on +Based on our discussion Ive expanded the patch +Thanks to Owen +Patch attached +It could be shared between developers +any problems should now get their own issue +Nevermind I misread the title +It could be extended by returning the default value for that type but that might be too much magic +Can i assume that this is going forward and populate the delete delete fields when ever a document is to be deleted +Please fix this again! Thanks +Ill look at the other issues you mentioned and then post another patch +batch transition resolvedfixed bugs to closedfixed +I didnt see a way to edit the original submission +Should this be part of the MAPREDUCEThe is based on the framework defined for as everything else in subproject +Maybe the enhancement can be discussed in later +Tests. +Unzip geronimo s build to geronimo +I will analyze the thread dumps but any help on the topic will be appreciated +I see the same results +Reopened to add NN label +Sorry for missing the two class bit +Will open new jiras if anything new crops up. +Form and has similar code which we should cleanup as well +Were down to one again +It should be added under the section in the +Jochen agreed that the test case looked like what we want +Im not sure I like the packaging yet but the cycles are gone +Would be good to see both of these issues taking the same strand of through +Ive not looked at your trunk patch yet but I expect it to be okay will do a quick verify before I Basically looks good and tests succeed +As you can see the current pathc has issue +You have to change the content of the whole file anyway +This issue was making jetty start at less than a snails pace due to its search for tlds and fixing it considerably speeds up jettys startup. +Initial implementation committed +I must fix a duplicated id error before to commit it +i will try with a new project when i get a free minute +This update needs to be upgraded in M or pushed to +Done for chapter +Please add a valid jPDL for this. +Question here is after completing the operations successfully also they are showing? If they are growing and not coming down also problem +s commit this and file a separate jira for mapred +Committed to trunk +It still works for me when I split it up into three processes +Thats the only way I can pull from the descriptor properly +Doh! cp +cant you just implement a compression option that will utilize what youre doing planning to do for compression in terms of block structure and block level checksumsGood question +Can you elaborate a bit more what you think is a problem +It can be committed alread but it isnt meant to be finished +I copied it out +As the previous request for clarification was not responded to and I believe the URL detection was addressed in there does not appear to be any requirement to make further modifications to the +However does not support interceptors so we are not able to implement the flow in a popup as we are in the connect adding the display parameter as hidden form field also doesnt work because the library deliberately only extracts the scope parameter from the form and ignore everything else +shipped in +Fixed on and trunk +Neglected to take hits into Justin +the datasource stuff +The fix Jesse described has been applied at an early phase in the uDig lifespan. +To my mind the reason for focus lost is that some other applications are being executed simultaneously on hudson +So I think groovy should take the same approach +By reflecting I meant splitting the patch and track a common JIRA for the common changes +bq +I could not replicate lately +Ive reviewed the change. +You going to add these JG +LiyinThere is also srcmainjavaorgapachehadoophbasemapredI guess similar change is applicable there +Uses method notify to have the deployed deployable list have to decide if it is necessary to create further Jira issues for this small you were right the notification thing was a small thing +We also had such a regression in our application but changed our appI guess with imagelistimagesTEtours you expect +At can we just focus on converting what is absolutely necessary instead of changing everything? I say this as this patch seems to add a lot of boiler plate code JUST to make them services +Thanks for the comments Dhruba +TedI verified by running the testcases using my IDE they passed +Could somebody from committers integrate the patch + +Either the name of the attribute should be something like node or it should behave as the reporter expects +to Alejandros suggestions +You can open another bug if your problem is not resolved. +Next would be to add new copy fields for existing fields +Testing indicates that using a single external DFS for all tests instead of spinning up a new for each test would save us about a minute overall +Marking it as resolved. +And we are facing many issues beacuse feature is not available no support when posted on jira. +is not in class path and so we cant know if it is implements one of the required interfaces +Details can be found in the updated users guideMarking this as fixed now. +Im not sure of the performance impact of doing this but it seems like having the retrieve the data in a single consistent fashion is a good thing in the long run anyway +If you alter the code according to my recommendations and still see the update fail please go ahead and create a new issue with a test case showing the failure +Sorry for keeping you waiting +If we are in that loop then the user has specified that they were looking for the java components to be builtThis is not necessarily the case +Switch back to Home page +Ill see +Thanks for looking at this patch RickCommitted as svn closing again +This never got into Indigo pusihing for it to go into Juno +New patch addresses some of the comment +bulk move of all unresolved issues from to +Fixed by allowing defining ordering when starting module activators +The use of force and having to manually move files around particularly on an agent is failure to properly package and handle updates +There is no template for rich to in. +Initial version DEVELOPER USE ONLY +Konstantin are you considering adding some way for Camel to provide a classloader for theThats whats needed to have it working with or in fact other platforms +I have patches on that use nanoTime to avoid these problems but dont include the limit on number of expirations per tickI suggest that we mark this issue as a duplicate of that issue and apply that patch since it is up to date +But liveupgrades are not the blocker atm +Attaching a new patch to include Aruns comments +Do you have any idea when that might come out? I need to let my boss know when he might get this functionality fixed. +Thats the reason state transfer commands to not carry any m not saying that we absolutely shouldnt use a flag for the server startup issue Im just saying it wont be in any way related to the way the RSVP flag is added for the state transfer I am saying that I dont see the reason for adding a new flag when this can be handled much easier by adjusting the replTimeout and desiredavggossip settings +May be we should separate into different guides or in the minimum different sections for administrators for users and a general configuration glossary in the same vein as HODs guides +A couple of things about this patch It doesnt apply cleanly to +Can you please elaborate on what exactly you want to achieve? Thanks +If they want to share data between webviews they can use statics but usually they want to have data specific to a webview +I see no logic behind disabling it +you can use svnexecutable property on SVN +bq +What happens if they dont follow the procedure by mistakeAlso this special work around needs to be followed in even in future when some one is upgrading from to +Committed in trunk at r +Cannot reproduce with r +Oops +for the patch +closing since fixes are in released code +Patch looks good but its missing tests +It may have been a dirty build or something transitory +applied the patch. +Can you expect any code generator to follow your own coding conventions +This issue is closed now +Which version of Spring? The symptoms sound similar to +Make sense +I will integrate this into Sling tonight or tomorrow morning +Basic results are as follows DocumentsSmall query msLarge query ms DocumentsSmall query msLarge query ms DocumentsSmall query msLarge query ms DocumentsSmall query msLarge query msAs I expected the large query which basically is a hit on all the documents increases dramatically when the documents in the index increase +I have pushed a branch already called post that contains updated stdlib plus a number of other tweaks Im not comfortable putting in +Not a directly supported platform waiting on patches +The patch contains the update + +You have to register with Jira at The Codehaus and once you have logged in you will see an option Attach files under More Actions +i edit this patch to make this changesuse a class named that manage a xml file named for maintaining hosts permanent and temporal informations in a multi thread environment +merged patch with latest trunk +Another interesting idea Itd be interesting for error messages to have parameters sent back to the client +Work completed. +I expect we will have a bunch of different implementations based on the requirements of the application +Thanks for the patch +But I think youre suggesting that its really something that should be specifiable on the command line somehow as a common arg that overrides the default? Sounds easy and good to meThe broader point youre making I think is that there should be a more consistent structure to Drivers and that structure ought to be X Y or Z +Hence I coded it the way I coded it +We will talk about logistics on the dev forumI would like to pool test cases results in central location so that they would be tracked appropriatelythanks are all committed as part of the code in the pluginIll see if i can write something down about the functional tests im running +JPA SQL file wizards belongs to EclipseWTPData Tools +not about the hbase client to server Interface +I have changed parselog +Updated WSDL has been committedThanks Ann. +Theres actually another problem with this +Please review to verify it works as you would expect +Which means the formatter will never be called and the wrong timezone will be usedSo attaching the ugly patch multiple locations where Dates are logged are converted to Oozie timezone +Updated fixver and updated affectsver to when we added configurable bffpchance +Addressing comments from on the javadoc one log levelHope its the last patch +I added some extra pointers to the javadocs of +Hi christian +I have scenarios to consider +I dont want yet another variant on the parameterized tests +Also this code is not because there is no atomicity between the exists and the rename +Assign static IP to that VM from reserved range +But youre planning anyway to have one licence per version in the future +It fails with the code removed and succeeds with the code in placeThank you Knut Anders! +I will clean upThanks for the review Ted +This is failing due to no in the that contains the ejb +Your patch mistakenly removed the in fieldThanks Ben +fixed one specific instance of this issue +removed the last forgotten file with revision most references had been removed with the checkin forNot sure whom to assign this to as a number of people helped on different tests +Its like my session gets corrupted somehow +Too old +The code patch probably also needs to update which checks the exact message text of this error +Attached patch for the Unit tests in applied many thanks for the fixes another patch for the unit tests in see svn commits +Yes the is exactly what we were planning to write too + is when we first tripped on the bug +about uuid for dbs i dont see the point as well +Committed revision +Patch committed +The version without the filter should only be used if the filter is null +Kytt correct got a funeral in family and as a result did not manage to commit the sorry to hear that +Shelly Im passing this one to you as youre knowledgable on whats included in these Seam example deployments +As said I still think using CQL table from thrift has other way to be confusing but why not +I think its fixed in rev +pretty rare +So the test youre looking for is starting and then have the write permanently fail at any of them and in all cases have the hsync fail on the client rightYes that should provide basic coverage for the failure cases. +Map Resource and cleanup +I mentioned that this bug does reduce the robustness of the proxy +Looks like someone committed some conflicts before you applied the patch +The attached doesnt work for me either when using +Oops I used the blocks link to by error +Closed after release. +Well have to discuss this with the juddi project +Are these simply static HTML pages? If so Im happy to provide you the patch or updated page with corrected links +bulk close of all resolved issues. +Agreed also whenever you have some time even after the project completes consider also minor update to Atom it should have a configuration property oneway and if it is set to true then should just add a HTTP header adding WSSE username token support this is less critical +For deployment errors we might used the Problems panel were build errors are reported so we keep it consistent across build and deployment operations wdyt +Thanks I was waiting for review before committing itDid you also test a build without cmake installed +Any hope of getting this patch commited? Its a simple fix for a potentially big problem +must be able to run on Java so this patch makes sense only under jdk unless you also make a patch for +This has been hanging there for more than years +Actually using the is not very good +As an optional new feature that doesnt break existing codeapis I think its valuable +I wonder if there is some type of list that either the style bundles or j itself uses +next JDK to work +Why does the order of the json matter +Attached changes +i am looking at backporting this issue to +I have attached an updated version of the audit previously attached +We can reopen it if necessary +And Signing an AM would be done only after an user defined auditing +Also with the monitor option we should probably setup a timeout to ensure that the we get the actual message back in some amount of time +However do feel free to submit a patch for the branchOleg +The following checks were added +the clue is in the stack trace +Hi DanielI tested with my application and it seems working +How works for you? +I think that the order in which Eclipse reads extension points causes this Eclipse reads the ext +Does that mean I can create the wiki site for the smpp component or should I wait until the Camel release was official announcedRegards get started on it after the release eg start of next week +This is a situation where we are off of JDT +I think JMS hornetQ spawns a separate thread but it is joining the same XA transaction +this issue cant be solved by Maven the fileset references files in the local repository with complete directory a problem with the Ants war task if there is no feature to flatten the fileset when referencing it +defer all X issues to +All necessary configuration steps should automatically be performed during tsant incorporate the relevant information from into the is a reproducable CTS run for jaxr without manual intervention except what is described in +WIP +not a method it can be a simple protected boolean by subclasses +Michael I think you mentioned the other day that you figured out the reason the JVM was seg faulting right? Can we document that here +Updated in ESB will be in next merge + checked in. +This is happening for us only during releaseperform is there any workaround? Its confusing why the javadoc for one project gets invoked from another project +I think we need to keep a clear separation between the BUILTIN authentication provider and other functionality such as roles which should be independent of authentication mechanism +I think Ive got a patch ready so Im going to file another jira for the proxy ugi changes +Closing resolved issues. +Bernd will look at +generally fixed we just need a better build system to all our Fuse archetypes +Fixed +Cumulative invalid GC testingThe reason why GC alone is tested in first two test is because if an invalid GC is configured the call to the method would be two times +Patch to fix pattern Stevensvn ci m Applying Steven Wyatts patch from to fix the referenced pattern language not Sending srcmainjavaorgapachecommonsbeanutilsconvertersTransmitting file dataCommitted revision . +The class that it needs to load is in another bundle so I would assume that I need to add an import statement to wire the missing package +I totally forgot myself that the original still exists in too +I understand just let me know if there is anything I can do to help further +add a safety check before going into the loop I still havent been able to explain how this case arises namely that we have a lot of log files accumulated but we cant find a single regionmemstore that contains some edits present in the oldest log +Sorry. +Then I unzipped in modulessystemlayersbase and the problem went away +Eh even so it would be good to catch on our own rather than find out when were doing release testing or a user reports that its broken +Further examination reveals that the patch for this bug is for a different file than +bq +We still need to test this +Marking bugs with invalid fixed for release as fixed in. +Thanks +I have fixed it by copying and into war lib directory +Having repeated similar errors in tests +Hi David the site patch is committed revision in trunkThank you +The patch fixes that +I do not have to option to custom build Grails +Committed +Nahi I did a short review of this and did not notice anything wrong +Partially complete the following items still need to be done build the plugin documentation upload the plugin documentation to its web space +With this test case we can increase buffering later without regressing +Ive finally tracked down the difference to the nodelay socket option in config +More a feature request rather than a bug not assigning to any particular release +Will upload a patch once is committed +Anyone use this? I think its convenient myself but these patches have never seemed very popular in terms of making it into Lucene +I only expect up to a x max use in threads but it could be off and allow for more than that if there is some small bug Im missing +And in case of now the tests do all pass by me no matter with or without the provided patch by this ticketHave got no idea why it used to fail at the time I attached the patch for this ticket with my first comments on itMaybe it was some Maven magic behind it after the Midnight +I think that it would be a good idea to produce the Employee Metadata and data beans using this new patch and put them in the trunk and distribute them in the new release +Applied patch to trunk +Unassign issues assigned to Nate +Hey Pedro thanks for the update Ill look at this soonI also have some updates on the IPC layer that I need to pushI dont know if we want to drop integration with ant I believe it is necessary for whoever runs the releasescutting can you confirm +we need to update the wiki page to show the installation of camelIt is a wiki page please feel free to modify itWillem +What are the issues +Should we do the same thing in? Yes I think sobq +closing the issue +Changes were committed to the Teiid product branch and should be in the CR release +Thanks using custom working directories with custom directories generation does not work because it generates into default output directory and cant find it during package +Ok given that these properties basically exist to preserve backwards compatibility it seems easiest to just convert them to system properties +But even advanced nor normal users should have to specify regular expressions for something as comon as having old libraries duplicated +hopefully should be enough for JBM sure you change the logic that checks for counter overlap since right now its assumes millisecond you are right about million a second for a single ll be careful about the logic +This could be a bug in PDF Box +Resolved on the branch +Ive confirmed this is fixed for us with +To make the result of this JIRA clearer in r comment Jeremy saidBTW I decided to use add instead of create methods as addremove is betterthan createremove or createdestroy IMO So we know the convention we are working with. +I tried with Chrome locally and it didnt work +Hi AntonIts a good suggestion +I think the fix is quite simple +Saves me making that extra patch for just the two filesThere are some occurrences of Derby here and there instead of the productshortname conref +lets do it over in new issue workOne questionFrom the doc I see last flushed sequence id of the region is stored in ZKI think we should store the last flushed sequence id for each store of region otherwise would cause the problem of data correctness when replaying logs +Simple! dumb +Would you be able to give any details on the percise details of how this bug was fixed so I can document it correctly in the Release NotesThanks in order to only change the fixed versions to exclude the FP release as this update will already appear in the CP release for customers of the FP +Patch solves the issue +Section of the WSEE spec provides the schema for the mapping document +User may enter a value +Issue should be investigated further on Windows +I dont think that will be a problem as nothing that changes the actual structures should be grabbing the lock without also grabbing the FSNS lock +it makes sense to provide an extension implicit boolean conversion operator for string +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Thanks Hiram! +You might want to check JIRA for that +I just committed this +Thanks +Just now i checked with branch with avro jar +Try Struts to see if the same problem exists +diff of added review patch manually cleaned correct file name +Michael I think I confused you by not giving you enough background information +I put the new event type in iocoreeventsystem where the other events are +It forces the installed plugin to get the new information +in path when managing roles +Checked with the release and got a better error message +Attachment has been added with description The patch as +That or somehow specify on the fly not to do lazy loading +Its a known issue on Jenkins build it pollutes the generated files +batch transition resolvedfixed bugs to closedfixed +Yes it has been tested with Weblogic +Close all resolved issues for Engine release. +Thanks Jarek. +If you want to have a new version of hadoop you have to create the set of submodules in hadooplibs +Remove RVM out of the equation +was fixed in but not in into develop branch +Sorry The notfound control simply renders a notfound view should be corrected as The notfound action simply renders to a notfound view +This probabaly duplicates +Can you try now +Georg I assume you are working on a patch for this +Your link results in a +Exists as the action on the node and sends a request to an ESB service +Junit tests that shows the problem and a possible solution +Released +Shall I mark this defunct then? It hadnt had activity in months +Reproducible test projects build and reproducemvn clean install in clean install in inFirst do mvn +Committed at revision Ive added the Ignore for now +An easy introductory issue for new the two classes appeared so similar Id removed one +Thank you all +a SHORTER name will fail? THANKS a lot Frank for the input I will test this tomorrow +And read up on how mahout converts text into vectors +Thank you guys for all the effort +For documents its better to load them into the Tuscany wiki site +Throwing an exception when none of the guarded transitions are viable informs them that the process isnt correct +Makes sense +Same as in the case of this patch is the diff between ydist and patch for branch +Will someone please try ibm jvm mac and any other jvm to see +Without a reproducible script or access to the database in question this is going to be hard to debug +It works for me now try getting the latest from SVN and having another go +Go for am working on it +What about that test looks like it would be an edge case +So there may be a build issue with the current scriptsYes examples should have been built alongside everything else so strictly speaking the second mvn install should not be necessary +Committed revision +Full backtrace of further investigation this looks to be an issue with how the json gem determines which handler to use when outputting objects Ill raise the issue over there + +This is because the version number has not been changedBy default the samples all use the version number +Is there a unit test already for temp directories +The changes look good to me +Since the nodes in the middle of write pipeline does not perform checksum verification and writes data to disk before getting ack back from downstream the unacked portion can contain corrupt data +EdwardI cant seem to figure out why the tests are failingAnd using your formatting wouldnt help those changes +Wow +. +yeah same as +A Customer has a requirement to have roles with the ability able to create a user that will only have access in read to that means to allow executing some methods depending on the rolefor reader role getAttribute getAttributes get for writer role setAttribute setAttributesfor invoker role isRegistered is get +do you want it committed to the branch mahadev +Quite right of course +with your assert passing +for vfor v itll break the method like get and get +The aar contains also the sources for the service +I have exactly the same configuration and Im getting exactly the same stack trace +Dan are you looking at this? Or do you want me to take it +rest patch more time to remove the reflection getMethod warning +Checks if the date format has time in it and then adds the time option. +Note that the excludes are not used if you use mvn ratcheck +BQ the patch into and trunk +thanks flavio. +Ill continue to work on this and will also give some thought to the other suggestionsyou made and hopefully Ill be back before too long with a more complete and working proposal +This will quickly grow hairy for more complex type hierarchy checks I think +waiting on more data from Gary +Added a test case for checking that the gzip filter does not interfere with expected return codes. +IOW it can NOT generate bit binaries on anything but a bit OS +Thanks Christian it looks good to me +If so no attempt will be made for trying to parse only a warning message without exception is shown instead of the throwable that you normally get when trying to do it anyway +you are right +I gonna close the issuebtw we have released ADS this week and we successfully obatianed the Open Group certification +Is the issue that when you have anywhere on your JSP it will print out the items in the session or some other issueCertainly you can block setting session attributes easily using the +This looks fine to commit +I got all my tests passing +We are looking at specifying the security constraints in the work currently being done for the platform release +But to access an Object or its field there must be a variable binding +could you please try if something changes if you add a to your scm URL +bulk close for +Also that error message needs to be cleaned up +If there are multiple possible methods then the same logic as we support now should kick in throwing an exception due to ambiguous methodsI guess we could repurpose the Consumes method but just let folks omit any URI ref +its different every time but Ill grab one for you. +Talked to Stack and Gregory C +Yes I also encounter the same problem mail node and mail action can only specify actors or to? that means I have to change the process definition each time a new user need to be notified or an existing user resigned +This issue has been fixed and released as part of release +section about java files is added +Addendum I committed to trunk and to fix build breakage. +Have run the testcase times locally and all of them passed +This has introduced an additional mandatory dependency on backport concurrency utils which could be removed by abstracting it if necessary later or just by moving to for a later Java release +Can you try this patch and see how close it gets you? We may have to go back and forth once or twice +Martinin Losanne you suggested to move this package somewehere else when I asked you about the meaning of this package so this jira task was just a reminder of something you proposed +But if it is really meant so then OK and sorry for false alarm +I verified that HDFS tests compile and run +Committed to trunk +What more tags should have the same behavior proposed above +bq + to or +Without tracking the big in Rome we have little chance of sorting this out in Nutch +The static cache is now not threadsafeAnd original had nice diagnostics for ambigous dispatches +I assume this will come with BV this issue for implementing the API from BV +I think it will really simplify things +Migrating to tests from tests resolved the issue. +Ive send out mails if the old server still exists and can be brought back online again or if someone has a backup of bugzilla but did not get any response +You have to explain the difference between a reopen and a getReader somewhereCan dobq +Thanks +Thanks Kapil! +Fixes +we have no pressing reason to update right now +This is my first flume patch so it probably needs lots of review and changes +removing from roadmap + +Users belonging to this role will be able to access the Status link on the main tomcat index page but will receive a Access Denied when attempting to access the Tomcat Manager +you add tests for consumers too +It should now be fixed in SVN rev . +Im guessing the help shows up because any C error also sets the usageError flag +This is an optimization that avoids some unnecessary code execution +DNHR now uses thread context class loader if no classloader is supplied will I believe solve your issue. +Others can do that. +I believe that pombreda has this working now for consequences is there a way to leverage his code for this? Any chance this can be done for or shall I mark it as FUTURE +this is equivalent to a core dump in C +Thanks Bryan for the patch I am running some tests with it +Depending on the type the MSB might have more information for you +James I think I can work with that +Done +It was +Thats great +Thanks for the review +I have no preferenceBut we fix the highlighter bug in lucene trunk too +This is correct behavior +but I notice your config specifies an avro source +Slipped a digit +No user comment on this issue seems like issue is fixedThanks experimented a similar problem fetching a list af components using POJO webservice using the +Ill go ahead and commit this second patch unless I hear otherwise +well the button component was what i was trying to use as buttons are generally used for non form submission task +has been marked as a duplicate of this bug +I can rebase a latest internal patch agaist branch not trunk is that OK +Thanks +For example does string comparisons on the authority part +fails with latest patch +The patch is missing file +If the task type is map task basePort will be the debug port for map task +just checked and I also put the configuration in the parent POM +Applied. +Doc team does not have cycle to do it + and the XML Plan are not written to safely handle exceptions +Also Id expect that new resources are normally added only in dev mode when caching is disabled +to the trunk. +I fixed if you want to test feel free +needs to be fixed its tracked under +We should do the right thing but I think its lowerpriority +Client doesnt currently have an SSL transport so the failure to connect is not unexpected +dont want to wast your timeThanks +need also the fix for the udp codec +WSSJ does a lot of things that require very strong encryption +What was actually producing the exception was bad math when integer math overflowed int we were trying to do negative offsets +Werner No I just ran my own tests not the Castors tests +Ok part of the reason theres a little bit of a dance going on here is that it wasnt clear to me that the ticket name Add thrift transport factory didnt map with the existing interface +Agree +Will let it sit there and make sure it passes tests across the nightly platforms and look at backporting it next week +Please verify when you get a chance. +Patch integrated to trunk +Which client did you use against which driver +You had agreed that I should file a jira to specify how a home dir isspecified for the different implementations of file system on various platforms +Looks great to me +Wont fix makes it sound like this is not a valid bug which it is. +Alright i would need to check more what can be wrong for the work! +This preserves the original WSDL documents and any other files contained in the process archive which can be used for a variety of purposes including web service deployment and status reports. +Sure +Thanks +ups +You did not understand me The above works with the existing API only by changing your code +Patch from review boardWill take a close look tomorrow +Ill keep this around for CR just in case. +Ok now I also understand why you added the comment to this issue +However it should work for common class of topologies +If this is the case we should be able to remove most Serializable tags +I almost forgot this JIRA +To test Configure nodes in a cluster but start only single node Node Populate the Node by querying a materialized view Start Node wait until it is completely started has time to sync with other node +I was wondering about that +The patches for the launchers and additional patches for runnign ofbiz under the Eclipse jcr core patched Francis keep the issues where they belong it will be easier to manage themAlso who patched the jar? Is it the jackrabbit team or you? Maybe it would be better updating jackrabbit to if the bugs are corrected instead of patching a jarRegards +I find the resource jar needed by the test didnt commit to repository would you please check it again +I was almost certain that this thing will bite me back after a while apparently it did but still theres a bit sisu left in me because the fix was basically a one liner +Is it possible the occasion of get on context aware of is not reasonable +Committed to branch and trunk. +IE is reporting a Tapestry is undefined javascript error but otherwise the buttons work fine +Modified the way the AM finds out knownNodes based on feedback from Hitesh and VinodknownNodes are now reported by the RM on each allocate callOne known issue with this approach +You might want to try running the output through an html validator or pasting it into an editor which can show errorsNote that the information is only approximate it actually refers to the page before it was modified by the to insert resources such as scripts and css +Pushed to with JBDS +Stack for SOA applied a fix to spaces issue and should be in ER release +Maxcould you check it on Mac OS and close it? +to keep the property I will just then remove the comment. +Will look more into this after I am back but I did want to point out that setting the set to infinite ie after a user has made a connection will only affect the future connections +Attached management server log file +Fix committed +release bulk close +The demo webapp for which creates k files in the tmp directory +This book has been completed for EAP please this issue with information if errors are found before to Laura as she is the new lead for EAP docs +This will probably take extra time but it will save you the bytes +Im abandoning attached patch +uses jakarta has been partially implemented and should be finished prior to the release +This was an assert that my earlier patch added to the unit test +I could see the changes in UI when I tried to add second zone +resetting P open bugs to P pending further includes two OTF files that use the SIL OFL which I believe can live side by side with Apache License. +Ive changed the modules to exclude the mail and activation jars as theyre not necessary for policy support but the rest axiom xalan etc are need right now as the scdl is now being validated and weve support +Will commit this in some time +Does not provide any testboolean to check whether the native libraries are available +Steps to reproduce Invoke welcome page +The only way this can happen is if your replace string has group references that are incompatible with the match regexp +Actually I use the hbase in my client side and hbase in my server sideAfter I update the server side to the +If that changes Ill post it ve worked around the issue for now by copying the test class I want to reuse +I apologize for marking this issue with too high of a priority +Closing all resolved tickets from or older +I dont know if the build will work in all operating systems without it though +aj files are compiled with the aspectj compiler? For that is no problem but if that is the case you should document it properly I guess. +By skipping the keyvalue pair we eliminate null pointer exception problem for the down stream application +Dupe of +Final names of attributes and properties need to be in CR +I believe that this case even though is not so common would be of use by some users and would be not a good idea to tightly couple a queue with the component and therefore leave this functionality optionalD +We just need to figure out how to build some test suites for this and rock on +Thanks Stack Francis +to calling it truncateNicelyMakes it more obvious why one might want to append a string to the truncated version +Its extremely likely that other implementation packages should be added to the current exclusion list +I doubt that but anyway lets research it in Eclipse +what does auto meanI do like hybrid though makes more sense to me than background +I havent seen any issues similar to the reported since last month neither I heard from anyone about this +required by the context editor demo +Either you save the namespace when you stop the namenode or when you start the namenode +I also verified that with in place you can run tests individually and get them to pass +Hi just wondering whats the state of this ticket? At the moment both git and svn are +I was thinking about checking if we write or localhost as the server address in ZKThen alert somehow that the cluster isnt reachable remotely? Sounds good to me +Thanks Sebastian +The thing is if I set the too low my only recourse when the job fails is to increase it and rerun the job +We were setting all the variables instead of choosing one +Fixed in the new website. +Bug in the JDK bug in Lucene not a bug its all the same +Oh and in the case of it makes sense that wildcards would be rejected since enums are implicitly final +pretty sure I have the right to revert my own commitOnce something is in the code base it doesnt matter who committed it all the same rules apply +A nitpick should be more appropriate for here +Heres the simple fix as suggested +Thank you Steve +I understand your point +Users should explicitly state that they want to ignore a potential error condition +First crack at a patch +So it looks like the way Lucene does it is w an ant task +Ok Ive applied the changes along with an property to view it and documentation for properties +Track any upgrades to fix regressions in individual +Followed the steps above problem has RC installation was put in directory CUDIG +Do you have an example where the current functionality breaks? I am currently able to update form components just fine using the semantics +However the API is massive and not many people have an AS at home to use for developing and testing But a great feature to have for integration to these huge systems running on the AS +Can you describe why this is more impactful for your scenario? I assumed it would be a minor inconvenience to use instead of in most cases +itself is a SPI +Because in our application after we use the property afterstatement we find the performance to be bit slow when we land in create ExampleWe have connections we do a create call the performance was good before using the afterstatement property +tabled +The purpose of that decoding step is to handle URI characters and non both of which we really needI think the choice will be between putting all such keys at the end of the URL or them +Requesters would include an optional parameter that told remote shards what time to use as NOW and which TZ to use for date faceting +Patch added +I need to figure out a better way to include the sun tools jar which is used for the free to ask qeustions here or scott or scryan hope to get this to the sandbox soon so releases are easier +Thanks for the link +Compiler errors fix in r. +I have applied it to trunk +Fixed in Rev +At least now this will not block Hudson buildBut we need to fix related tests +I committed the test. +It will probably be easier to check for that on our end rather than waiting for a JRE fix +Submitting patch I am not entirely sure how to get jenkins to run but I think this might be it +I think also that we could do with enforcing much more verbose logging but dropping it to DEBUG level +This will be an issue in case of a target server restart and the redistribution didnt arrive the target node yet +Committed as svn resolving +In my opinion yes +Among the files already updated there are the following files that should be fixedby Richardsrcmainjavaorgosgiserviceobrby StephansrcmainjavaorgapachefelixmosgiconsolecomponentsrcmainjavaorgapachefelixmosgijmxagentmxjutilsrcmainjavaorgapachefelixmosgijmxremoteloggerThe rest of projects never updated are +Thanks +in the Metadata Extractor library + +Are you saying I should import both projects +ok thanks very much +I am still exploring if it can be achieved +As already mentioned in a former comment the isnt welformed +appears to be doing a local read in the client thread but youre right +Uploading the look good some minor comments for null in addDeprecation method +Sorry I have not been able to allot a good chunk of time to get this done +I think that will make more sense to usersI hope these changes are ok and that they will make the filter even more useful +Please make sure that you havent specified any path to m repository in plugin properties +For me it is related to bugs and improvementsThese reflexions born as I am trying to be precise in reporting and testcasesI know they require time to get a good answer +Hi Using with as suggested by your I find the following problem +We either have to rely on or we have to introduce something like can now take inclusionsexclusions +Maven project will be better +fix bug in doing an unchecked read when it was already complete. +I think characters should be good enough normally +I hope that Antoine will provide you with the test code you need +New failure indicates problem is likely is there a testcase that you are running to see this problem or did you just manually set something up +Reopening to write release note issue now release note text has been written +I wasnt logged in too when I tried the to search browse but instead of the NPE I got the Unable to find project model +Thats a hopeful sign that this patch may very well do the right thing +this is also happening for IE on still exists in SBS Tested in and IE tested this issue on and IE after release which comes with fluid design and its OK now. +In the release the ContextLength response header is not set. +Sorry for the noise +Lets just drop this unsupported legacy sar and remove the +The updates its meatadata with tis means that a new get call will automatically retrieve the last good size of the last block +Wiki page updated +Checked in both the test case and fix to +If are faster for sorting too its getting very hard to see the downside of that representationSince relies heavily on arays of sorted indexes into ordinal arrays Ill make sure to performance test both and straight arrays for sorted search +IMHO I dont think that this approach is the best way to go +Done +I dont think it is a good idea to refactor tests that dont initially pass +David I think is good to go +Need someone with access to start a run thoughThe LCE appears to have the same issue with a restrictive umask +The decision we made for an FP will be different than what we decide for a CP +The question is if we should really commit this without any specific usecase or other code using itare you going to address earwins issue still +Hi Yoram could you give a try with Maven? Thanks Freddy +fixed in in r moved logging to the controllers +I posted such issuer as you pointed out and also hope someone could fix it as soon as possible +Another difference is that the dynamic wsdl had the following names for wsdlservicewsdlportname http http and where as the output from has just two with the following names +fixed in destination package exist in files system generator stop generationg classes with overwrite warning +Also FYI Ive created to track upgrading to work with Avro and to track a separate issue with the unit tests +bug fix +Ill then try to make the split tomorrow +That leavesMonday December noon PacificTuesday December pm PacificWednesday December noon PacificFriday December am PacificFriday December pm PacificOf those times I can do the followingTuesday December pm PacificFriday December am PacificFriday December pm Pacific +Please check that the patch was applied as you expected +Thank you! +I still see the stax dependency in the Mead build +Code implemented and documentation updated. +This works for me +rev +Fixes release audit warnings and a few style nits +Hmm we still have to support Java +Wow so much great stuff happening in the Solr systemId love to have this tool for my Solr workshop coming up +we can remove it +The bug is that it shouldnt be trying to open a file with tmp in the nameI wonder if this is another one of those bugs where we have assumptions about opened files that dont hold under Windows +Avner just sorted out the JIRA glitch and assigned the JIRA to you +Issue in sample code not Apache Flex SDK. +Patch addresses Dorons comments +I decided to implement this properly by using namespaces +batch transition resolvedfixed bugs to closedfixed +OK good to know +Thanks for the reply! I downloaded and installed this specific version +Hi SamThats another kind of issue and Im not suficiently eBay Store aware to fix it quickly + +The code didnt create a single task for to but went through the to directory and create one task per eachThe reason for that is +Thanks a lot for cleaning these up +and in Commons Email call the setContent methods +Hi SylvainI am rejecting this bug as you did not pass the Enum but its string representation +Wait until it connects tothe server shutdown the server pause execution and look at the Threads +Ive used version but the phase is still running twice. +If we want support this query search implementation need serious refactoring so it may take more than week. +and javax +What other possibility would be there to solve this efficiently +bq + +Thanks Binglin +On a tangent there shouldnt be an explicit check because the remote server might have security enabled but thats a separate issueIve got a patch Im almost done testing that will reduce principals to simple names if security is off +new version +A junit testcase would be appreciated +We could also the String and Blob implementations from Orderly if its a critical feature +First cut at making pluggable +Thanks Kamal for spotting ll close this ticket. +add task options to property file locations with default property files in the jar +This should be working fine now that has changed considerably +I would love to see this feature resolved +But Ill commit to the branch shortly since it doesnt get in the way of anything +Created for itProbably the usage of default extension was not as well reviewed +Committed branch and trunk +the jar in your bundle doesnt match the official one +If you are still getting problems please paste the exact code snippet that causes the problem +Fixed in ESB will be merged next in beta +I dont have strong opinions about it either way +Maybe then it will be a good time for someone to try groovy with as per your previous comment If the AST transforms that implement the groovy behaviour are packaged separately to groovy and put on a projects classpath in it should work +Resolving issue since the original reporter says its not there anymore. +Thanks +None of us are running on Suse so unfortunately youre going to have to do most of the leg work on this one +Also with security coming in deletion of the distributed cache should not be in Tasks space +We thought our design would work out of the box and it didnt because of this behaviorWould be nice to see the expected behavior implemented +Recently i have debugged many issues due to OOME in my clusters +and yeah we need to test and likely modify the binary dist patterns as you hinted at + tag removed from party component into rev +cannot handle serialized message by design. +It got something to do with the changes made in ASF release of creating a cloud user and corresponding cloudstack named paths instead of cloud named paths +This is a dupe I believe but for the life of me I cannot find the original to link them +Adding initial version +Either the tests will be automated there has to be done work to automate them and also to maintain them +That is it still will upload the entire file +Could anyone please review this issue? I would like to commit it in time for the release +This should be against trunk right +HiAttaching the patch which has addressedthe following removing all trycatch blocks and instead throwing exceptions +o thank u Martin commited to trying to explain the hack after show the options list the text field wasnt losing the focus but the text cursor disappeared or goes to the wrong position +not sure this is a big deal +Cleanup the patch by removing accidental changesIf there are no further objections Id like to commit this +Sometimes these dumb tasks make you feel good about getting stuff done when you arent in the mood for clever stuff +Soctt Harris the last commit still not solve the problem completely and the master still clear cache when aio enabled + figured out what you meant I put a CDATA around a bold tag thus disabling the bold ll fix it ASAP. +For the case when unit test are in individual project we attach PDB files to project so could copy all all dependencies required for test using dependency resolver +Wouldnt it be sufficient to run the WS tests against JBAPP +patch LGTM +Added KiranYour patch is in trunk r r rBut please dont use tabs in you patches +Logged In YES useridyes times slower +Checking it in +Note that the enivronment for me is Windows with JDK +I get +For info This is not reproducible as I mentioned sometimes randomly it comes +Submitted fix for satheeshSVN LOG have tested it and it works ok so closing the bug +Thanks Henri and Gary for the comments +I just committed this +It is a kind of secret knowledge that developers keep to themself +It appears that seems to care about the protection of the class in which the method was declared at least in JDK which is the target JDK for Velocity. +When this happens the portlet container must wait for the specified time to pass before creating and initializing a new portlet it doesnt seem to be much testable using current testsuite. +bq +Rayees could you tell which version of libvirt you are using on kvm hostYou can use yum info libvirt or something like that to get detailed package informationThe storage pool not found issue wont happened in the last week after I fixed +Considering this test hasnt been running on I think its fair to simply update the output for +Great news +implement logs and metrics separately in RM and NM +Thats why I put it in! It can be used to extract things like HREF into other fields or not +Sveta please check this and check if this exist in guide +Example RW C unit tests using cxxTest Framework +Make note of lack of Trevni Boolean support got bitten by while doing some performance testing with this ticket +Why remove the code in srctestjavaorgapachehadoophbaseiohfile too +Its been working great for us +I dont know of any cure but eliminating iframes or accelerated scrolling +Fixed +Well done Ilya thanks for improving this area! A few observations what about renaming portfilepathshortname portfilepathbasename which is more intuitive and less confusing +Ive made the changes but I would like to test it before commiting +one last thing can we do number instead of no? then I think we can commit this as a step should make things a lot easier for experimentation if you are new to lucene it will make life much easier +Also targettype and sendid should be evaluated as well in line with the latest Working Draft +I am wondering why you have moved the class from a separate class file to an inner class of +But I dont found this in the documentation and for me its not clear how to do this +Will come up with a patch for review soon +sorry for confusion +and providing the support for old group names in Counters class +It also happens in Grails +Speed seems good in JDK and happily this isnt an issue with the underlying source just the test +If we end up having to update manually in order to ship a full binary I prefer to go back to keeping the dependencies in svn +Should also mention that without this patch the servlet is not tested at all +An application can not access any web resource without using the proxyWhat I do not understand is under which circumstances the inbound configuration needs to be changed +Just committed a fix in that area +patch against trunk r +Based on some observations it looks like occasionally a column bundled with others in a Put is broadcast into all column families +Removed methods and replaced all usages with which now contains the same logic sans in stuff +this the functionality most of the mail clients today provideI have triggered my interest here +It seems like the right solution is just to catch Throwable and make fault a Throwable then +Minimize database documentation for Chukwa deployment +I have not verified on yet between snapshots +Did you mean that the title of smooks editor doesnt change when you renamed the smooks config file +However we would need to know the format +Works great Pete +Console Application demoing SVN +Hi can you help us to solve this issue? Thanks can you switch to the user mailing +Can you please share the exception details? Also you will have to make sure that all types and depend on are also listed in +There are some limitations on how close we can comebut I think that Really Is what the customers are going to want and expect asour default mode of operation +Fixed with since is now released +Attaching a patch that solves the issue +changed the methods and added a method to find the fixed issues. +good catch +Often when this would happen the had gained focus with a border and carat but when the keyboard lowered the proxy image would use the image of the focused field +Stefano can you verify this is still relevant +If you cant see it in the metadata displayed by you definitly wont get it in +So examining this further weve had this behavior forever +If you have any questions regarding the fix please dont hesistate to askThanks I wish our talk went in different tone +Should update the recommended HDFS configuration in the book then? I think losing a region of data after a compaction and power failure should be prevented by default +Ill fix that upIn the mean time I think you can probably get your urgent work done simply by doing run as only from the editor window +Thoughts +RunningEclipse IDE version Juno Service Release build id +Attaching which makes some minor improvements to the tests +Great and no problem +For example the new ing of the ideally Id like to use the same instance that deployed the application but at the time locateService is being called a lot of these types of artifacts seem to be not conveniently available from what I could tell +Attached patch a which removes a set of unused methodsCommitted to trunk with revision +Zebra and use the code in to read this data off disk +Thanks Eugene +I am working on Windows and I dont diff or other tools to create the patach +For example by default Java installation sets JAVAHOMEcProgram FilesJava +place as dependency plugin declaration would not work? see the it does not could you provide an example? I am little concern about back ward compatibility after to copy the example I still get a submit a pom and jrxml file that can produce this project illustrating reported parameter class is already compiled to insure we are not running into is a modified pom that works on my evnmaven javalet me know how it goesAfter looking further of the code the example we provided is wrong you must provide your own jasperreport in normal dependencies not in plugins dependencies +The need to exclude not wanted dependencies breaks portability of deployable items +Thanks +I think this is fixed so I am moving it off the until Andrew has time to take a look at it +Ive been trying to get this working with the BBB task running into issues and limitations of the bbb server task +Logical modelCube A cube is a set of dimensions and measures in a particular subject + +Regression tests that appear to fail due to this bug include v +The aggregation tree is a great idea because we can piggyback aggregation connections on existing worker connections for many levels up the tree and hopefully only workers will actually ever need to add extra connections to talk to the master +thanks for have a look +It should be fixed too +Duplicated issue +Updates completed. +or not and which is ahead and also the timezones on the two systems. +Ive questions +Ive attached an logfile +I would like to use mina to build gateway applications that expose various serial networks used in industrial automation to TCPIP networks +Where do you anticipate wanting to use this new hash function +Bulk close of resolved issues of +QA Verified for server and WAR files. +Please verify +I dont know how get the output of shuffle phase in hadoop or +Would you mind reviewing rev +Thats what AS has been doing all alongBtw Ian Springer from the RHQ team is looking into the possibility of reducing the retention policy of this annotation taking in account how the RHQ XML file is generated in order to see if the runtime dependency can be removed +One or more additional Jiras will be written to address building the update site web page uploading the update site archive and instructions on how to build the archive +This patch looks great! Thanks for taking care of thisIm sorry I couldnt review it earlier I had a power surge and had some connection difficulties the last few days +I will try to do it this week and get it into +Im fixing minor stuffs on the and adding the file +We will upgrade to a newer version when we migrate to Axis has already been fixed most probably with WSSJ release as Colm mentionedI verified it in the latest trunk using the basic sample. +Relaxing the security rules on the groups page fixed this +The recent changes to pool mean that this patch wont apply cleanly +Jason needs to answer this but AS is the upstream for EAP so I dont expect were going to be replacing EE tech with EE +Hi Wancan you please check whether it is working now +We have some implementation of editing for many components in vpe +thanks for Mr +I applied the latest patch from Jan in r +Attaching untested patch implementing logic above +Patch applied +Will take a look +Thanks David. +Patch looks good +Testing Jenkins is green +This new patch additionally changed check to check if a block is complete instead of checking if it meets the min replication requirement +have applied the patch +Much better and should avoid the last of the timeouts while setting a hardcoded seconds between calls +Thanks +Closing rejected issue. +Clearing fixVersion. +or we should remove the custom deployment locations since it was confusing to the user and as had a clean deployments folder thus removing the necessity of us having metadata or custom options +It has been specifically designed not to fetch the object until explicitly requested through get. +wont fix for unless someone chips in with a patchthanks this with Axis and worked fine +Help About screen in JBDS showing icon and divergent provider ok for now +Changed logic that resolves references to existing entities to ignore case +About to cut RC +Yeah actually it looks like the fix is just to spell replicateonwrite correctly +I thought that it may be something on the Amazon side but havent figured it out +defer to +May be Evan would be able to provide more information +I still dont think this is correct +But Im left with the question of whether design is part of the document name or if design is more of a namespace for design documentsPersonally I could see it going either way but I havent groked the internals enough to know if theres a reason somewhere else that requires the design as an actual part of the name or not +Sorry to be a PITA but I need a diff u format +I installed that version today and have been using the version until today the zip archive name of that dist is +Bulk of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues patch that deals with a minor change post we have taken this issue as a starting point and extended and polished it quite a bit +Thank you! Im on vacation +When its time to open a new writer we stop giving the writer out wait for all refs to come in then open the new writer then start giving refs to the new writer outSome things that complicate thisAn NRT reader takes the writer as an argument +It is definitely a test config issue +would you please try the first patch for java +bad +Please review the patch and let me know your feedback +This is likely to be a problem with the Hadoop configuration or machine setup +Ok +Dan can you please move thisPS And add me to the seam dev permission in JBSEAM +In both cases they seem to be treated as if they are new nodes that are completely separate from any nodes that the processor might previously have processed +currently because of this in amlauncher we end up converting bytebuffer back to credentials +I just havent been able to get to them due to some other projects going on +Another dummy patch just to run MR tests without the actual fix to +though we need to work on a sampletutorial +Can you please change the patch so that lines wrap at characters? Otherwise it looks good +Javascript call was missing in template. +Tatyana did I understand correctly that comments in exclude list like the test fails on RI means that the test is invalid so it is added to the exclude list +Can you verify that switching the Maven versions does not result in usage of different versions of the Release Plugin? Otherwise this looks more like an issue with the Maven Release Plugin than with the Maven core itself +I commented these lines out because by this point in the code we have already guaranteed the project IS faceted +The contrib version now uses the as a comparitor and in places there were unnecessary reverse operations on the paths into the Entries far faster just to iterate in the right direction rather than always expecting the list to be in the preferred order +Please use spaces rather than tabs and align with spaces per indent +Here is a picture of the dialog in question +Sigh +xml file the xml file at least to me is much better readable +Im going to find out the solution +No QE action needed duplicate +matching the exact minmax values of your data range +just that the Zookeeper team need to listen to the community +I have removed from the drl and since as it turned out have nothing to do with this only thing that matters is to run the BRP in stateful mode and that was available in +fix at revision +Open new on the new UI if necessary. +Patch looks good +Youre talking about synchronous and asynchronous mode right? For synchronous mode outputBytes really seems to be a good candidate For asynchronous mode how about we also monitor outputBytes but with much higher timeout defaultSo basically we do the same thing but with different parameter +yes +Plus itll help as Compact View isnt always visible thus lessening confusion if buttons moved +Patch for verifierAdded support CONSTANTClass for ldc instruction +These extra files were actually introduced in earlier releases but not noticed until the release +This was reverted in febbadcebdcdc as part of +JAXRPC tools issues may be fixed by community contribution +NoHow effective would JIT help eliminate the overhead of objects created and forgotten in a tight loop? This should be the kind of things that compilers are good at. +DaveIll take a look tonight and get feedback to you ASAP +Minor patch probably could even skip review +If an application fails because of framework issue for a MR task fails four times and thus the job fails will be FINISHED reflecting no platform errors but AM will send as FAILED If an application finishes properly and the framework is happy too for +Thank all for your inputs +In it was found when partial state saving is not used use transient listener prevents it working on this case and it should be +Itamar in case its helpful for you my code is not in production yet but close and I am not using latest patch but the original one +Ive asked Infra to enable editing of the site export via the CMS in All being well well probably be ready to cut over in a week or so +bq +This really needs to be addressed as none of the current versions will compile any of our apps and it forces us to use a version that is five months old +Sorry Nick +Ok +bulk close of resolved issues. +Love it +Small changes in a couple of tests needed integrating new ZK +I did a little more testing it turned out that this change increased the training time on my test set from minutes to minutesI believe its better to have this fix because then the stoping criteria works correctly and avoids over training of the model rightAnyway I think it should not change the accuracy of the trained model I did in both cases iterations and the difference during evaluation was percent +IIRC this is harmless +Verified on +I think that may work +Changing the serialization is a very trivial thing once this patch is on the branch I can switch it pretty easily all of the logic for doing this is currently in the object so very simple to do +Works perfectlyhere the first to lines of the prodices file +Cute but Id hope we can do something quicker +bits for a major version and for a minor +Regressed with OSS build and build +I just committed this. +I changed this to issue type there a reason that my fix hasnt been applied to SVN yet +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Theres some useful system info that would make a good addition in the uihttplocalhostsolrcollectionadminsystemwtjson +Glad to hear it! Good to know that about the Eclipse WS wizard. +However I included a patch for this issue +The problem is perfectly reproducible +In the future if needed we can add synchronized keyword to s readwrite methodsThese methods will be called once per row so the cost is much smaller than what we had once per row per column +However this is a general classifier problemWhat appears to be happening is that the +Attached patch lets ATS examples compile on but this isnt usable state yet because ATS fails to register kqueue eventsI most likely wont have time to investigate this further during the next few weeks so if anyone wants to take a shot at it it would be great +Reresolving. +This would be a standard Seam component needing accessto the portlet configurationoptionproperties and provide complete decoupling +Other classes dont use refaddr for object keys they use self +I have no idea what patches need to be provided anymore so I just uploaded a whole bunch of them +Is the software grant now listed for this contribution +srcorgjbosscacheaopcollection +Perhaps a feature could be enabled to support an expires header with +NadavI see what youre saying +Workaround when creating new ESB projects reference the runtime regardless of the installed runtime +Apache would timeout and return an error message to the browser +im not sure what this error report is about +Please close the issue +Verified by Sean at repo revision rThanks Tim +Things To Do are all fixed. +In particlular I am on changing the and classes to eliminate argument guards and jiust allow to be generated and thrown with no context info when one of the arguments in null +Now even on the fastest machines the timestamp must change +This patch cannot be applied to trunk and is relevant only for branch +Is it not possible to revert back to a prior version of JIRA that works correctly? was created back in July has no listed workaround and has not been commented on for over a month +So Im not thinking its growing is the issue +Here is a +The spec says that only primitive boolean types can have an isXxx +A use case for fields with different languages in different columns +New patch with files missed in the previous patch +to what Todd said +Found the issue committed the rest of the fixr +these fixes look great +It replaces NPE with a friendlier message and also augments the original stream out session message to clarify that streams may still be going in the background +It might be not that easy to find our in a few months why a patch with was committed and why no tests were delivered +fixes this +I was just caught off guard by the added dependency in the nightly build +Ahmedthank you for spacing time to review this patchBut when sort plugin is finished this patch will need be modified +I believe the fix should be done in the getReference method of the to set the TCCL to point to the view class classloader before trying to create a session for stateful beans +Yes I can see how this would be really useful +I thought wed set this up once before already but I just looked at TRUNK and its +Havent had a chance to look close but cool +Why is this the natural place for this info +The state is triggered if a TAKILL is received before the result of the CONTAINERLAUNCH is known +thanks there are no objections Id like to backport this to +Looks like this issue is fixed in the latest code +I have already broken that line into three lines and it has less width that lines above it +Look into this for RC +Uploading patch includes changes to and resolved issues. +Therefore if two none OSGI jars export classes for the same package one need to embed them in the same OSGI bundle to have all classes provided by the different jars +However the problem appears to be something else the server sends a FIN before it could ever know the clients opinion of the reply +Pentium M CPUs throttle their speed which causes SIGAR to detect that the speed has changed and define a new CPU +IanThat was a bug in PS? if so what AS revision should have this problem fixed +issues +We will be safe if its passing. +Try it out +We are still awaiting the outstanding information for the Release Notes on this one +I forgot to drop the table I added in tearDownThank you for your advice tamtam +I will file another issue for that +What cant xsd specify for jaxb? I was under the impression that was a significant point behind the technology +Looks like you have been approved for commit access so go ahead and commit this patch +Turns out stock oozie doesnt have and libtools bits +Dirk you mean the example above does not work in for you if you use and it does work if you use the version of +Did you open an issue against Spring +Im planning to add a bunch of dry run tests +Fixed +This should be a blocker +Slightly tricky code but not monstrous since the reduce gets chunks in sorted order +I just committed this +I was incorrect those files are actually so go for it +The changes look good to meI assume the overhead of decorating the keys in the comparator in the wont be too bad if its just a few entries being compared +Just need to do changes in POC. +do you exactly mean with the schema can be specified in configuration but the value is currently not propagated to +Looks good to me +In particular it would be good if there was some decent error reporting about data store initialisation +Is this a concern with this changeIs there any risk of statements silently being rolled back without any exception with this change? Or a sequence of operations not committing atomicallyThanks so much for taking the time to look at this +Lowering priority due to it being only one single model of phone +I see no reason to separate the two in the random read mode +Attached is an initial patch for this issue +This patch consists of only documentation changes +ASMYou can always roll your own and replace the default one +Here is a revised version of Jonathans patch which ensures that read is called in the exceptional case as well +Thanks for the patch Daisuke. +I am writing a testcase for it but then I am not sure when this scenario is going to ever occur +Of course if you a failure early on from not having enough memory this is likely to only prolong your agony +Is that intended +So it seems these arent available in time for the service strangely enough only if I define more than one serviceThe service itself doesnt actually use these classes it does use some other parts of the Rails app but neither User or the Api +In my case I only need to do a few run to trigger a case that can bypass checksumAlso attached a sample log file from this unit test run +Trying Hudson again +Corrected mistake in arg handling +Makes sense to scout around before implementing one +Uploading a patch that addresses Bikass commentsbq +I have test the patch of as the I attachedbut I find it become slower than the original implementation in mahout +Im moving this to as I really cant stand to release with thisIf this becomes too long we can move it again to +thanks for for the quick fix. +I cannot reproduce that problem +Changing name to +I dont think that we need to have a separate category for docs. +Ugh sorry Gav that you had to do this +Verified thank you +This is a Java backend to +I think it could show a warning if user is deleting a SYSTEM the warning popup message has been showed when user attempts to delete SYSTEM node +The assertion is in a jetty thread so this is not a blocker for isnt it? That means it might never get into release +Did you update your workspace copy? What OS are you on +WTP supports JSF +Ive submitted a patch to the HDFS services under which is in sync with this to show it worksNow how do you find out javac warnings that dont appear to get printed to the hudson console +Magnus +I believe it isnt suppose to be this way +i see what is happening in the code but i want to make sure i understand the semantics if i pass a list of multi ops and there is an error does it all fail? what is the use case for the check operationYesImagine a graph where each node is a znode +Hi Spencer for a good jobdevs Cant we integrate this FAQ list into our existing documentationRegards we can +Hi Chackara Thanks a lot for working on this issueSeems to be valid scenario +FYI I tried this out first without the patch just to get a baseline and I was able to reproduce the problem without the patch +Yes cut and paste error +Leaving open but resolved until we merge that to trunk and get it running in the jikesrvm installation of cattrack +While working on the implementation I have seen cases where this assumption was not correct +I believe that the correct behavior is to generate a concrete impl on the fly + +Created for follow up +Shall we create a branch and commit this? The downside of course is that wed have to commit patches to trunk and this branch until is out +I will try to put it up this week +Just was about to ask the same A simple test case would be great +Amend documentation +Will do a big effort before first production release things are going to change before it so just wait a little more. +The object in SLFJ has an event timestamp +I think that it could be implemented with some break lines +There is similar page in ecommerce but easier to useThe new CustomerEmployeeVender etc pages in Party manager have lots of details I think if we use ftl for doing such page they can be made much more friendly +Id resolve this as Wont FixThe CLI and GUI classes add no extra dependencies and cost only a few extra kilobytes in the Tika jar file +The default flush uses the regionserver sequenceid +Yep it would be nice to completely mavenize and I strongly believe it should be our goalMaybe we can do it in stages have a hybrid one on day one when we merge into trunk and then do the whole nine yards? That way each can proceed independently +Thanks +So dfs package is still called dfs +Checked in +Mainly it doesnt run sonar as a post action so doesnt help +Thanks for narrowing this down Dominic +This can be further reduced by turning off map to +committed +Was not able to reproduce on Oracle +In particular I have been looking at for absolute addresses and for relative addressesThe way I have been testing this is by exploding the war and running a testcase which runs against the wsdl file from the war +Ive been waiting for any official release of use the with the Apache cannot release my own projects with any SNAPSHOT dependencies including deploy a first official release of this +Affected teststest their versions +ran the failed test locally a few times w patch in place +The ANTLR parser rule is srcmainantlrtajoengineparser +Is the call to serialize returning a long instead of a string +Thanks Peter and Mark +If that happens Ill be able to do the same inspection I did against Ryans dump +is not a working copysvn Cant open file +from what I can see could be mistaken +Generate formula +I have the Scope on the impl and not the interface +We should commit this patch soon +Affects the module. +Until some other tickets I reported were closed I have been unable to proceed with RailoI cant add an attachment +The current attached patch for makes a ton of changes inThis stuff should all be pulled out into a separate package that can be put into a Maven module like done with security inWhat changes to core code remain after that +Attached patch adds support for and via OWB EJB Interceptor for CDI managed beans +Updated to reflect recent changes to +I have attached that file too +Misconfiguration +Hi MarcId be glad to implement it and had sent some proposals but we couldnt agree on the API this should have +Tx +Maybe an issue with theKonni +to commit +key should be instead ofRegardsAmit is the updated patch at revision Thanks ArunRegardsAmit Sharma +Merged to doc branch at revision . +Hmm may have been looking at the wrong patch +Error occurs on and latest nightly +bq +It would be better to keep these separate +I have modified the installation notes to remove the configuration of the interposition interceptor +Haha wait that is some artifact I forgot to remove while doing the test case +Are you building with any of the maven like faeAlso look for failures in the module in questionThe reason for this is that we did change some of the failure behaviour for but in theory it should only affect behaviour for fae or similar +Not sure how I missed this +tinderbox ran cleanlyClosing the issue. +NTLM is not suported +I will incorporate Eniss comments into patch this weekend +Please verify that the fix works for you +The same errors are being generated even after removing the space +Sounds like a bug afterall +Sorry for inconvinience. +Thanks Jeff +Spatial uses error correction if you use decode +Caching with field collapsing is an optional featureChad +As far as submitting a patch for this it would be huge as I assume that all of the code would need to be revised I dont understand the architecture +A few thoughts do we worry about security? is anyone allowed to read anyone elses logs? shouldnt LAs be allocated according to cluster topology +post the the cleanup actions are queued and are done offline which the isnt aware of which I suspect leads to a timing issue where the is rightly idle and the cleanup thread hasnt kicked in yetShould we add a sleep in or perhaps another api to check if the job has finished and has been? Thoughts +Because we use symlinks could you please update your rsync config to use L rather than l so that the content will be accessible +Let me know if you need help reproing this +rd section of your comment yes sorry I mixed it up with the th section and thought it was related to security as find was existing in the controller +Hm I cant duplicate the behaviour you describeI have an index with a few entriesWhen I search for travel I get match containing Land a hot travel deal +I was able to successfully apply the patch to the source in svn so it seems to have arrived fine +Sorry my previous comment was meant for +That would make it easier for me to apply itFor the log shipper issue either of your first two suggestions is OK with me +Why would you do this? This should be handled by some service not by the session +Weird +Perhaps this behavior with CR is related to that was recently fixed? Can you try again with the latest SNAPSHOT? +Starting the conversion to resources +This patch passes mvn clean install +Ill close the issue then +Hugo Do you have ETA for this issue +Im working on some improved tests in this area but Id appreciate any help or additional test work +Its up to you to decide what you prefer here. +per comment Brianpatch looks good on formatting but you missed the parts about updating in the srcmainresources directory and adding the entry to? Can you add thoseCheersChris +thanks Steve yes the session are getting closed up and it happens afters say exactly min for test for users for users it works fine +I do not delete because it could be usefull for further improvements. +Punting to +For future reference I think it is good in general to keep the old ones there to see whats changed +If someone needs a similar issue please create a new one + +Its not exactly the same behaviour as when I run against my stub but there is something not running as expected +Please dont add retry to +Current axis trunk supports multi dimensional arrays. +final patch with some checkstyle issues resolved. +Dunno +Can you please add some description on this mechanism +Stuck locally as well even the timeout does not work +Merging to can be done but Im not ready to introduce it in synchronization changes make me nervous +Im not sure about the side ofPlease just validate this through the wiki pages where we have the final requirements +In most cases they throw a column not found which brings me to the conclusion that there most still be something wrong in that needs to be resolved before commit +Min is working on to update base library to use vSphere jars +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved +this patch takes care of preserving the architecture specific lib directories in case of and also takes care of retaining the symlinks for the so files in case of the rpm address the issues mentioned the right this to blocker because it did in fact block the RC for several days while we straightened it out +The attached patch makes only developerConnection configurable +User Guide configuration uses channel for sink channel configurationlike cFor R it was +myfaces jars are already in my classpathI have just updated my code source with now we only load from libraries +Yes you could just switch Jetty to jetty I was just attempting to make the code more tolerant of changing product names in the future by lowercasing everything before comparing strings +The old patch was not handling the case when the only forward mappings were of the mapwithrecvport type +I can either upload it here now for review or in a couple weeks when our QA testing should be complete +Nice to committing the new version as isAgreed with your point of avoiding too big design changes until someone volunteers to lead a more substantial designbranding update. +Backported to both and. +Overseer can have a pool of threads to process the queues instead of a single thread as it has would like to know the state of the other shards as it often has to query against all the shards in a collectionYou are missing the point that its very unlikely for anyone to query across all shards in a VERY LARGE cluster +this one is fixed on the CONFIGREWRITE branch +This might be true for some of the Sun JCE implementations but Im not were only passing the if the cipher isECB mode +Closing per Linda. +May have to move a ton of class from groovy to java just to work around this unless anyone has another idea? Like can I slip a step into the gradle build that would let me hack the stub and change the to a +So the issue seems to be localised to using compatibility mode in jruby +Bulk closing stale resolved issues +Test runtime exception used in +Hi Bryan if you have some time please take a look at the patch I attached to the patch +Any chance you could try a snapshot jar from the master branch build? Thanks Paul +See +Need more time to confirm this behavior +Log a warning instead of throwing an +Committed to trunk. +At least it works and doesnt make any other tests to fail +Ive attached a patch that will throw an exception with that information hopefully that will help with debugging the issue +But thanks for your eclipse settings i think i can extract your formatter from there +Applied the patch into trunk with thanks to Aki. +Wrapper method +That said I ran it under duress and no noticable degradation and all unit tests pass +My comment is about the test error on offline discussion with Chris I filed as a separate issue to track failure in test. +OK here is the latest version with javadoc warnings cleared +derbylang passes successfully +Keep up the great work! you have the chance to give the fix an early try please let me know whether it works for you! news it works perfectly nowThanks againAndrew +I could reproduce this on Microsoft Windows Server with reduced users rightsI prepared patch for this issue +I will finish testing it and check it in tomorrow +The pull request is still not merged reopening so we dont forget about it. +Unit tests pass and I am committing this +Hmm yes but we are currently trying to get a big bang release from Sling out of the doors +This issue doesnt show up in stateless and singleton beans because a lookup doesnt require a EJB client context for those beans whereas the stateful bean needs a context for session creation +Kyle committed this a while ago +So I sum up in telling that things work as expected +Ok I am changing this to Resolved +thanks. +keep cause of sense +see attached screenshot of the wizard page +address comments added tests +Yes the updated patch on phabricator is ready for review +verified in all these to which is in line with BRMS +AndyJust to understand +If people are happy with the then I can also do a variant and doc updates +just some remarksare you sure that the classes all come from one place? So they can only be loaded by one and only one classloaderAre there more the one created at one timejohan +Its OK if thereve been changes to this shared data structure since a reader was opened that reader knows its max doc id and so it can skip a term if the first doc id in that term is that max +The addresses that problem + +Moving items unlikely to be completed for out to sign of current activity for some time so moving out of issue out as part of JIRA cleanup +Okay sounds good +marked as WONTFIX +like a class path issue to me +screenshot shows total cpu usage overcommited cpu resource at cluster level +The slowdown of the overall cache is a side effect of the locks from the prepare not being released +So the problem is connected topermissions in some wayI can run the jsvc fine though if I just choose to ignore that warning +Im perfectly comfortable with requiring a hint to indicate that a cache should be used +But fixing is not I seem to have not realized all the things that it fixed +It seems I have got a jenkins git build stuck in the way Im working on a fix atm +Thanks +However a good case could be made for defaulting this to something like hr so its one less config knob the user has to explicitly setfind +Could you check it on your environment +It occures if you closed project which is parent for project and rebuild this ejb project +An element of an array being null does not affect its length +Just a lil bit of improvement +The one thing I would like to advocate against is that Cassandra should not fail to start if there is a corrupt +A timeout on response could happen for many reasons so the answer is to set it at such a level that you are able to detect an error versus just a slow connection or machineif doesnt make it stable enough try +Bulk move issues to and +It would be nice to control the map parallelism via the PARALLEL keyword if the default slicer is in use +Thanks Ed! +I then narrowed the failure that youre seeing down to the optimize property if its not specified +It looks like this is a bug in pluto per of the portlet spec which statesThe following methods of the must be equivalent to the methods of the of similar name getScheme get get getAttribute get setAttribute removeAttribute getLocale getLocales isSecure get get get get get is in trunk and by delegating the get call to just realized the version of the WAR I put up depends on a local HSQL db running +Removed unused imports with revision +Will the standbycheckpointing node have to have twice the memory? I thought the main reason for running a secondary namenode on a different machine was because checkpointing needed just as much memory as the namenode needed to maintain metadata +We might as well leave it for the Java Java requirement +While this may not be a complete fix but certainly one of fixesPlease see for more details +Fixed in trunk +and should now passI should have also mentioned this patch should be applied on top of the latest patch posted on +Unfortunately I had to pull Spring from our platform due tobind issue but looking forward to the final release +This would require changing where in the pipeline we parse read requests +does not remove a block from its namespace +Depending on how its implemented I dont expect that it would cause the of and to diverge +Committed +mvn profile hadoopnonsecure passesmvn Phadoopnonsecure clean verify +Fixed some minor jira number dyslexia +Nigel could you please explain the need for locking if you are doing this? Id think that each of the builds of A B and C writes to separate directories so there is no need for locking +Hi Glen Mazza Thank you for your quick reply +Patch committed +Ive created my account on jira +HTTP always sends a response so with that transport there will always be a handshake response and thats the only transport specified todayFor the Java implementation of this I thus intend to refactor handshaking +If someone has delete access on JIRA feel free to delete the above lengthy comment +Updating the summarydescription to expand the forgot to ever link this issue to the one that seems to have caused havent had any more time to try and make sense of this and dont anticipate doing so in the near futuregiving to ryan since he worked on in the hopes that its something he understands and can help bang out a fix for easily +Although this issue can be fixedcorrect on its own it really needs JTAJTS and XA Resource support in the connectors to get the transactions to exhibit transactional behavior in most environments +Test failures are unrelated and I think should be fixed now that is in +setting a fix version setting a correct description of the a dump of my workspace so Uwe can review the new attribute +In a sense yes but this patch has a much smaller scope the only thing it does is to allow compiler clients to optionally provide a custom class loader for loading AST transformations just as they have always been able to provide a custom class loader for loading compile dependencies +Good point +trunk already has the capability to set the JSVCHOME environment variable to point at the operators preferred version of jsvc +Since has many patch files I created ant scriptPlease use the following steps to apply them +Are you going to port this fix to the X branch since the other portlet fixes were ported + checked with AWT renderer works thereSo the problem is probably in the PDF problem only occurs in combination with font sizes andis a consequence of +Best regards George +How can i get the plugin patched for Eclipse +Thanks for the comments Andy +Since it is a requirement to read the data from disk there is always a potential for failures +I also think some of the code does not follow coding guidelines +patch looks nice +The ack of a heartbeat packet is the same as a regular data packet as well +OperationsTypeMeanSDththththththPSBBh +Ive just sent a developer access request to so Ill try to assign the issue to myself right after I get the access rightsThanks +For example New UID Prev UID Next UID For each message these values are different +Defer all open issues to +Vladimir thanks for the test and suggested ifxIve fixed the bug in slightly different way than youve suggested please review my fix and let meknow either it fully resolves your +Added check for partial state saving on facelets stuff necessary to make it work +I was not sure what to do then I am forcing the locale to ENGLISH nowI also took a long time to figure out how to implement the numeric configuration it seemed to me the best approach was to copy the way is configured +Ive done some testing on the branch +Patch applied on and trunk +So someone should to synchronize work with this object. +there are no patch attached to reviewive added label assuming youll add one +Why groupId +The fix is completely contained in Polyglot and would require a new Polyglot until +Should I try upgrading hibernate for this issue perhaps +What does jps l show +this patch will always highlight correctly as it can be switched to boolean query rewrite mode +All shell tests pass +Supporting client code provided by CXF project +summary of discussions +Patch looks good +Looking + I think thats a good approach +Seems like this would be better in the module +Patch committed revision +The WAB extender could deploy the WAB which could ask for taglibs before the bundle with taglibs was scanned +We should figure out a way to do this for svn and do it in a separate jira +It adds an extra lookup in LDAP when the user DN is resolving +I plan to provide a patch for the trunk that modifies the config classes directly after the work on is done +That truly reads to me as the real portable way for a PU to apply for a EE platform default there is nothing in the specs saying it has to be the same default datasource as javacomp for an EJB I guess its fine to keep the feature in JPA config to define a specific one but I think it would be nice to have these in sync by default which is what my patch at starts +Can you explain what you mean by dead gossip list and how this prevents truncateThe decommissioned node is showing up in the UNREACHABLE list when calling describe cluster +Does that work with Camel +sorry I misread your comment about dramatic changes +Hi Amos a few observations are you using the native libs or the plain old gdal that comes from the the ubuntu repos? setting GDALDATA like this DGDALDATAGDALDATA wont make any differenceActually It is entirely clear to me what steps you are taking +The upside is that its more configurable and will hopefully do more of the heavy lifting up front +The document source is the Wikipedia English XML used by contribbenchmark +we have final patch for available but have been trying to get the one for trunk into shape +Updated description since the buggy case is the opposite of what was described the problem appears when there is a main module configuration declared in the configuration mapping and not on dependency configuration +Yeah +Right but if we want to bind decorators we need to give the MDB a bean types and qualifiers in order to resolve the decorators +Hi Marcono problemI understand that by using my class is stored into a database sorry for the misunderstandingThe real issue here is containing IDs of objects stored into a databaseWhy Im storing my objects into a database at the first place? To be able to pull it back after server restarts unexpected application crashes etc +Committed revision +Verified on +Any live demoexample for change name via resources bundle? +The fix seems fine thanks AlexeyClosing. +When requesting a wsdl for a service that has user true I always receive an ip address for the hostSeems to me this JIRA should be reopened +This doesnt appear to be fixed I still have the issueWhats stragne is that generation was working last time I ran it I added a new and not all pojo members of a particular type are set as the hibernate type not the class they should beIm using +Patched long ago closing. +It is a way to reduce a possible longtail +Please reopen this issue if thats not the case. +LGTM +Note that this patch also renames to because it doesnt run automaticallyThe change to and is the same comment out a test and add a comment to say why it cant currently be fixed +I have Sonar in my laptop and want to connect it to the maven build with the sonar plugin in the Solaris machine +Essentially my patch merely prevents a crash it does not fix the underlying problem of references to scratch files being made in error +See linked issue +delete oldlogfile once it has been successfully processed +Confirmed that this error is reproducible with the change to trunk +We need to clarify the definition of first +Updated patch for the patch for this patch will fail until has been committed + +I actually resolved what was wrong with this issue +I can take of that if you want +Logged In YES useridThe httpclient has been updated to the latest version. +Well accept this an enhancement +Will create a jira forNicholas thanks for corrections +This is a patch to resolve submitted for review +Committed to trunk and +I have rewritten the testLE case +Use cases would be helpful to ensure were all on the same page +Thanks Mike! +Thanks! +But till such a point in time where becomes deprecated of sorts we need to support it and this patch could prove useful no? And as an aside needs porting to the new mapreduce package as well +Thanks Colin! +If you want to open a new issue for that please paste in the link here and Ill give you my thoughts over on that issue. +The issue will still appear in the version of Citations Plugin +I know this works for the so my first guess would be that the is misinterpreted +should solve the problem with maps at least for i and stringsCheersps I removed the old foo struct test this is already tested by! just committed this fix +I tried to reuse as many existing highlighter parameters as possible and only added ones where something didnt exist or would be +Uuurrrr +Binding to to and to would fix the issue Im facing not sure itll open others thought +Sorry I can only follow all the Castor development sproadically +There is now a Jetty release that fixes an NPE issue that we were seeing on AMQ in the Web Console we have moved to that version as fixes the security issue in +But we got this problem +We could +Documented support for annotations in the framework within the reference manual. +Attached is the patch with test +Applied by Evan in r +Its not as good as having the script itself test each profile but itll make things easier. +The decoder and encoder do not handle recursive schema +Copying Fang Edison for more commentsAnimesh +workspace +It was verified quite some time ago I just forgot to close it closing +Attached the service in rivision . +Quick question what was the cause for the nested attribute stuff? You said my patch affected that? If so how? Just wondering cause Im not sure what would have done that +This situation would have been a bit more graceful if we could have alerted on the IPC threads getting exhausted +lets look into it after GA + was setting the final exitStatus only after stepExecutions last repository update +java files so that it is found by javadoc and included +Thanks Chris. +Im running into issues as well and though I havent pinpointed them they sound similar to those was experiencingThanks +Once I get to a Unix machine Ill test it and build it into the documentation properly +s in the select part of the queries +Here is the patchDont forget to add filebuildmakeexcludesto subversion repository +Hey RoshanSounds interesting +Not sure how to do that. +then SEAM could buid a extends as far as i can see the approach you describe is approx the same +As stated in the comments this issue was invalid so closing the issue. +Attaching a patch that replaces NBHM with an that is appended to and counted atomically +Feedbacks from users are always welcomeThanks +The required file should now be included in the patch +Id really like this to be there any workaround for having session scope beans serializable? I tried switching my Scope annotation between and but it doesnt help. +I hope this gets closer to what is needed +Harumph +Verified in +bq +thanks! +As example refer to +This patch fixes the ui and adds getters for the values that used to be thereThe totals are computed in from the totals of just added the extra info since we have them +I tried the resolver in the beginning and it works perfectly in the embedded Jetty when generating a war and deploying it to Tomcat no resources are found +Awsome! I will add an issue for providing a karaf feature then +Just doesnt seem like it would be very fast +More could surely be added +As I understand it to view collections in a standalone iReport you would need to perform the extra work of creating a implementation as a source for a subreport to view Collection properties of aA report for Struts does not need to do this as the does the lifting for you by creating a for any Iterable propertySo any report with a subreport on a collection that works for standalone iReport will also work for Struts the doesnt recognise the property and just passes it through unaffected +I linked to jodatime documentation in the JavadocI verified that jodatime exceptions throwappear well in grunt +That problem should be completely unrelated to these patches. +Productive +Your improvement is much nicer to read and understand as wellGiven that is a API class and Tool was the real class for users I think we should not suggest GOP at all and rather show the message just with Tool +There are at least two types of repositories a general library repository and plugin repository +I just committed this +This is remotely related to the cadvise warnings issue +name should be +It was a problem that I just committed a fix for +Please open a new jira for any new duplicates that are introduced. +HaraldWould it be a big deal for you attach wire context logs generated with this works and this also works examples? Olegno big deal except I have to relay this to my colleagues tomorrow +Thats why it is disabled by defaultThis isnt to say we should start features to users +Any news about this last issue with duplicate rows +Im uploading a patch for castor +Comments from triage April Henry Z +Right now you are forcing me to put the below classes in my project which introduces a maintenance problem for usPlease kindly address this issueThanksGarry +disable CDI settings CDI facet should be unchanged right click the project and call MavenUpdate Project ConfigurationBefore fixing this issue the CDI settings hasnt been enabled +Above patch is an itial patch for preview and comments +is a little better +True + Using a Matcher as a static final member +it is not if you want the right one hopefully by next week Ill have time to work on it and give it back to Spring supposing theyre still the so dont think this is anymore needed +We need a fix for this ASAP so that testing can have at least one execution this week +This patch removes most jar files from Avro retrieving them instead with Ivy +cd AVROlangjava +Closing old issues +I think this is duplicate with +Also we can do a replay log +When this functionality was originally added to it was a deliberate design change +Illustrates the first problem with an image called the second is a workaround at least for the dynamic siteCredit Juan Jose Pablosthe idea is to write the full URLFor part changexlinkhreftoxlinkhrefhttplocalhostimagesFor part replace xlinkhreftestwithxlinkhrefhttplocalhostimagestest +Looks like UIMA has XML in mind when serializing +Alessio I still need your input on this +I have attached a patch which works with the version of the plugin +I just committed this +Does the comment However If the derby jars are updated to a new version mean that the bug is fixed in the new version and so thiis bug is already fixedIe +Minimized the jobtracker init changes +Thats why Ivy publishes your file as a pom then later tries to parse it as a pom and fails because it isnt a pomSo I think we can close this issue since its due to a problem in your settings +Nice catch BTW +Problems were fixed by changes to buildtarget platform +I dont want to start down the road of customizing the UI for this sort of stuff due to the maintenance overhead it creates. +A couple of notes +I just checked in an test suite class +Shouldnt this thing needs to be tested under root as the instructions in the header file explicitly state that +If I understand Jesses proposal the idea is to have a WAL interface and essentially bring all hdfs dependent WAL code under? We would then implement an equivalent for a backend? I didnt understand the point about the class being relatively well encapsulated Jonathan +Patch was created on Windows XP in the Geronimo root the problem record to reflect change in symptoms since originally opened +URL should have suffix +Removing the name from the join table makes hibernate creates separate join tables for each subclass and envers creates one for eachThis can be marked as invalid +Should rename +Hi AndresAny way you provide a new patchI can apply it due to some other changes +Hi Graeme thanks for fixing this +Here comes the zip with the two to addThe problems are already there when I request so as Ross suggested it is likely an issue +Must have been a corrupt jar +still does as RFC does not restrict the message to a specific character setPlease verify and close. +Marking bugs with invalid fixed for release as fixed in. +We can even move the test to but I think it has to be named properly +no? So one single file for the test run should work no +Patch seems to be good and the little Mockito extension is neat +This is a good start but would likely need to be bigger in scope than what Im seeing in the second draft. +master cdfb febbfaccbaacbaefdaf +can you give this to me in patch form +Dan can you look at this issue +The Seam Javascript in has changed significantly and this patch cant be applied to it +bq +Hm probably not a bad idea. +Lets release without it and then set the requirement for IO to Java +FreddyNew version is compatible with eclipse +Committed said prior patch had added it. + +Typo this is an there is an +One question is this causing any acute problems? Any scenarios where youve actually scene this issue surface +Bulk update to fix version +Attaching a patch to fix this bug as well as provide additionalerror condition checking and logging +M released +This is the patch for the exception I dont have time right now to do the complete Carlos Ive applied your patch to fix the exception and will look into supporting the includesexcludes settings +the patch looks good to me +Hi Kevin DoneFWIW to install this plugin I +The fix is simply to add an order by and update the +The problem is fixed in latest versions of Eclipse +fix for test cases in faulty introspection via property with introspection via andFixes the following test cases patch and delivered. +Made the following updates refactored renderTitle field to hideChrome in and hideChrome now also hides the border around the widget in addition to hiding the titlebar updated Person Profile Page template to demo this feature using the groups and experiences widgets +patch applied +An easy fix is to use a different split policy for this testNot sure if it is good to increase the flush size +not a jbpm problem +Thanks for the great work on this Slava +The problem is gone now +Should we close this +I could take a deep look about the test failures for the root causeIMO many test cases are not very stable because of itself or potential dont know whether this caused the test failures +Exception Service failed to open socket Address already in use JVMBind Service failed to open socket Address already in use JVMBindDo we need to disable ejbds in geronimo +Roy Rich Will this be fixed for? I dont see a patchthanks am working with Roy to get the patch +we have removed support for on from cache storage this reduce this network traffic to MichaI am running on my live server which has websites with files +Reporters of these issues please verify the fixes and help close these released and verified as part of release +You cannot use jps if you only need to know if the process is running? Am not sure of your need +The fix will be released as I hope not just added to the trunk +I agree with Robert +This issue would make are the bean types and qualifiers of the MDB +Karthik +I will log a separate bug to track those errors +Added JAVAOPTS in the check committed into JBPAPP revision +This will only work with the change to from the previous patch that allows deployment of local wars as regular wars +Initial implementation is done but there is place for lots of improvements. +The basic idea as explained in the encolsed HOWTO file is to enable the creation of webapps quickly and easily with minimal additional work and only a very simple +Hi Alexander the fix version field should only be set once the change has been committed and the JIRA resolved +Whats the latest Seth +Closing. +Yes it has been merged upstream and I didnt see the Address Already in Use exception anymore. +During the recovery the update would create dummy objects with mapreduce slots for the trackers +It seems like this bug report requires a bit more context +looks good will merge if the tests can you add a test for CTAS for for the hints NamitAttached a new file +Same problem for +Comments for the spec file Add an empty after the license header Add an empty line before the if suseversion The requires need to be updated for the alternative command +Personally I really like the idea of being a shop for these +If is set to false then the ipc Client does not send periodic pings to IPC server +How can we proceed +Would you please attach your patch file? tks +I werent able to get around to it as well +This will essentially be what you describe but will probably use absolute paths so that it can be put in the users PATH variable if they desire +I think that will work fine +I agree with you +When you set the date on Feb and then on Jan it jumps to Jan +It is still not working in beta +simple fix committed + +Thanks for the patches Prakash and Ramakrishna +Looks very clean great work +Im closing this until we get more info. +Great that fixes it the source build runs fine nowNow onto testing the maven artifacts +Here is the trunk version + Bug has been marked as a duplicate of this bug +is a typo in the previous patch +The zipfile doesnt contain the binaries but they will be built when you run mvn clean package from the root of the project +Then I have another questionIn harportdirfile what is the behavior if is indeed a directory and is an archive +the view of an instance will still be fixed +how big is the timeout for git cli? It seems to wait forever or + +svn ci m Applying Vincent Behars second patch for fixing a bug in abbreviate such that lower limits greater than the length of the string werent working correctly srcSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +I am guessing your changes would go into primarily +See his comment to +Hi Todd +This makes it possible to integrate many different tools because they only need to understand the pom +To use addresses with Infinispan you need to use the new Configuration builder introduced in via properties and the old deprecated constructors will be dropped in be related to +creating multiple simultaneously is not supported prior to +The way we should do this is via has been looking at On and the current version has problems +This should do us as a workaround for now +Sorry but I no longer have the config in question +I would rather have another jira for that since Stack is cutting a RC tonight and will need a imports cleanup +These really should be final propertiesfields +This is the same problem as +Any progress towards a fix or a workaround +Released to the production fair enough +Does it workAbout changing the protocol in my experience changing messages is a pain because there are many subtle cases and it is quite easy to get it wrong +We certainly could include the type parameter but it is not a bug that we are not including it currently +Previous hudson build failed because of +Need the isExposed method so that checking if a portlet is accessible from a registration does not become overkill. +Pierre he said to me on IRC that he changed the RUNASUSER to root in the etc script +I have removed it from +Im not convinced we should do this now but theres no doubt it would make our lives dont see why not its an easy migration for people +I tested the new patch +I mentioned in the other thread it is by design that the channels do not block +I think this was solved +The problem is not related to object doesnt support file transferring so Im working at an Ajax implementation using calls which is the only available hack for allowing file uploading in forms +Any chance of this making it into since it is fairly trivial and the fact is in +Thanks Tom! +My bad +I have quals comingup +This is exactly the same semantics as native X +Bulk update to fix version +Ive tried the latest build of and still the same results +Hey BenThe previously posted patch works for the hints but uses a manual config based hint approach instead of an API based one +ThanksI am rebuilding the VM now as it was destroyed when we moved to a new host systemWill let you know when its back JacquesSorry please try logging in againIgnore the place where you put your key for now infra has changed their minds on that usage for now +Jeff yes when you use more than one bean in a page this will render the errors from all beans. +Patch applied +Were have a bunch of conflicting patches nowLets take Enis one as the authoritative one +A quickstart showing the both on and onDetach are called just once per component +This code is not in use for and Timestamp functionalities +I just checked in the beginnings of a startup shell script in the scripts directory +Compared to thespring xml file it does look much nicer and is super easy to understand and specify +Merged fix for and into branch with revision +Sorry this patch doesnt apply cleanly +Thanks +Hi AndreasAny contributions here would certainly be helpful and very welcome +I wonder if this is more general for example if we provide joinpoints for properties that would typically be jited away at runtime +I think we need some strategy to deal with such kind of failures +rated up to be a blocker because proxy handlers dont get called in events which will almost ever lead to +I didnt notice ognl to be shaded I excluded jbossjavasist +fix path + and Photoalbum verified on FF and Photoalbum runned on IE +Not sure if that is right +That is what I thought also I remember it does not use connection profiles it has its own way to define the connection properties rightI am just trying understand we do not create any duplicate ways to do same thing +I did not found another is quiet critical +Attach a complete patch can you review the patch +The same model is used to generate Axis description hierarchy +If you have more than one schema you end up with a corrupt schema file +I regenerated the metadata for all the files under orgjbossweld +Is it caused by bad disks or buggy ext software? Any ideas on whether XFS on linux avoids this problem of disks goingI think the mapreduce framework has to handling similar problemsIf a drive of a machine goes bad the tasks on that machine tend to become stragglersThe overall performance will be impactedOverall Hadoop is much better at handling total failure than partial failure of nodes I think it is better to decommission a bad node at a drive failureThe admin may later choose to remove the drive from the configuration file and restart the node if he does not want to take away the node for repair +The new output is as follows java ij version ij connect jdbcderbybrydbcreatetrueij showERROR X Syntax error Encountered show at line column Issue the help command for general information on IJ command syntax +The patch for solving this problem is attached +No longer resolvedwont fix issues. +Ill generate them from the project directory from now onNiall +But usually an test is encouraged for a new feature to make sure the whole process works +All but the first will be ignored for now +I just did some work to find out we should set right for theNow I just create for it +The testing page does indicate The percentage reported is based on the lines of code your test cases manage to test please limitthis to real tests although we demand test coverage for supported modules we would muchrather this is produced honestly +Currently there is no way for the persistent map to know whether the value is modified or not +Patch to compile with JDK +Attachment Source code for Java and JSF resp XHTML you could try to experiment more for example try removing out of the picture and see whether this makes a you could try starting a new web app with literally just the part that is failing since there might be something else in your stack that hides the actual this helps +No Dave it was in +Reverted the change for now +It is building the package for suggestion completions +patch to restore use of mmapindexonly option +I checked out the sources and I am building it at the moment +OperaBuildSystemx +Also moved the new code into a abstract class so it doesnt pollute so much +we are currently debating adding features here which Apache Hadoop will never useMany if not most of Hadoops features are primarily used by downstream projects +can you elaborate perhaps with a test case that is broken +Closed after release. +The crux of the issue is the public List getElements on Model +This should stay open until jbossws cxf is released upstream and updated in AS +but it still feels like out of order iteration and Imtop scorer or not are orthogonal concepts +Bulk close for +this ticket is duplicated of +Is there is a download link for the build which has this fix? Like to test this fix with my application +Thanks for reporting +This got truncated from the last comment inadvertently +C codegen problem fixed in +What was the version you were using before and what is the version now +If you want us to consider the cache issue please open a separate JIRA under the Cache project not the Application Server project +I wish JIRA had a priority between major and minor +Daniel could you give the latest patch a try to see if it addresses the timeouts in the pig tests +which is a default servlet handling all requests for which there is no more specific servlet +the early version put all these jars in lib and foldes and JBDS adds all runtime jars there to ESB project CLASSPATH but in most of jars such as are moved to deployerslib and current JBDS can not set up ESB project classpath correctly +Since latest eclipse has adopted a new update manager this problem doesnt we have decalre we didnt support GEP update from GEP +im on the coding rule patch on the branch revision +Fixed with. +i plan on just applying to trunk since the change is rather deep +AFAICT there is only one portlet for setting up jms rars and this has been the case for a very long time +If itdoes it might lead to existing mechanism to save off a stream so that it can be reopened +Thanks for the hint with +I think I understand your issue a bit better now +Resolved is used by the person who actually fixes the problem +Also I have included a patch for +Now the Groovy launch config looks for a config with the same project and main type name +If you think they are critical for CR please them +The ideas explored here are now a part of the project. +I think its safe to take this one +The fundamental note of the simplification is basically rather than having a byteblob which encapsulates a principal and whatever else is needed we switch to a style which still has the principal on the outside and a byteblob which gets passed to the Authenticator to figure out how to handle it +I think they could also pass in a composite to the one that would delegate to other depending on the record type probably using a map similar to how its done in the sample applications now +Once you install the Xcode command line tools however the install fails with the above error +batch transition to closed remaining resolved bugs +Committed to trunk and. +Its about time to ask about this again +And that would be the default timeout in +If you finish it before we release beta please correct the versionThanks +We should think about solving this sort of issue once by maintain such lists in a plugable source of configuration and supporting the ability to hup the service +Extends generic class in +Railo noticed in your commit that you added an explicit throw when using the cftrace tag without debugging enabled +any estimate on when the patch will be checked in? I am using my own patched version but it would be nice to have those fixes checked in? thanks for all the hard work +Committed r . +The remaining classes that do not have coverage will be removed once the graphing support is available in the RHQ plugin +Ill need to validate against a real url. +In order to get ejb to work I had to also set up aspects aop and ejbx in the same fashion. +I wish we could figure out why it is incorrect for youAnyway Spark needs to use a different IME logic because it is intended to support text +If anyone still has concerns with this patch speak up before it gets committed +My is +Cannot reproduce would need a test server +Yes the div really is necessary +Ive increased the page content size to avoid thisNew version is placed in attached file +Marking closed. +Im confused +You might have a better solution +patch to +Please check svn diff before committing next time as this causes the whole file to be overwrote your changes added new entries on my own and merged to all branches +I marked it as wont fix because we did not include performance measurement in +So this shouldnt require any release + +The actual pom is copying an applet jar and its dependencies into a directory +patch for ivy for ivy +As the message will arrive on a completed transaction already +packages +Until this is fixed I added a workaround for the test thats failing in revision patch with some minor things fixed Better test for the close child case Fix Solr now hopefully the last time by implementing the fake reader correctly as subclass without passing null to All tests pass +As we dont hear anything from Lucas and can not reproduce the problem Ill close the issue. +Both subtask are resolved +The should reference the disruptor jar as a dependency +Reviewed and ran unit tests +the field have a Party look up that shows only the Person +I started upgrading my slave shards today from to and ran into a bit of a showstopperSEVERE Master at httpHOSTsolrlivereplication is not available +Could you clarify +The Spring way of doing things these days seems to be not custom but it is easier to understand how to do it with the so Im doing it that way for now +Ive uploaded the second patch +Thanks Zhijie +Check Mac Dpropertyvalue fails every time with the weird quoting this may be fixed by the newer +From this information we will write the final release notes for all issues +Ive verified in testing of the branch that the above is true +Marking closed. +True but the issue here is the other way roundThe browser has a cookie for jakarta but none for When logging in to it sends the jakarta cookieFor some reason Firefox does not do this but both IE and Opera seem to +Viknes I applied this patch +Can you attach it to this issue instead +Just realizedMaybe it doesnt has to do anything with transparent resolvers at all but only with nested repeaters and adding cells from the innermost +I just committed this to trunk and +I was thinking the wrong way around +Id merge it into masterCan we get this into please +Issue closed with release. +bq +Thanks Pierre fixed in at thanks that fixe the issuePierre +Removing your redirect rule from the page constructor is not an workaround +zip +Sorry I think were not talking about the same problemThe problem being it does not find the EAR if it has not ever been built before? But if we do mvn install then mvn it runs ok +The problem with is that the same style paths which work in and below do not work in and so this is a regression of behaviour +Ill go through these tomorrow and talk with the team to see whats definitely going +Thanks Nicholas +Once its applied no more a copying constructor added to so that the files are not checkedcreated one more time +Andor a JSON interface to create update retrieve node types +now tried to remove data center from zone +Merged from fuse rev in rev +Lets schedule AMQ where we will split the core into smaller pieces and create a client JAR as well +Good idea the same principle can be used for where theres an sql error functionality added +bulk defer to +If it receives a LASTZXID message it generates a unique id number that does not correspond to the server id but ti doesnt matter if we are using majority because for majority we only need to count votes +I only change the value on the Backing bean which is the property in my caseI thought the input component gets populated by the model again +There is no other formating than the paragraphs +Patch for this issue +Could you please elaborate +Fixed in and +Yes it make sense +Ah yes that makes sense +This issue duplicates. +revision in trunkRemoved dependency as we prefer to use slfj over logjThanks for the contribution dependencies are +This is the annotation for the staticmethod that return a list of parameters not for a method that is an actualtestThe thing that causes the framework to gather the list and run test for eachelement on the list is the annotation on the class AFAIKOr Im misreading itErick +Chose add patch to wrong issue please ignore +. +I cant reproduce this issue +Ive just tried again +Cancelling the patch since this work is not complete yet and maybe requires further discussions +git diff agains signed CLA. +The cache issue in IE is something we cannot program around +This function is very in need +Aleksey could you please close the issue +Both have their merits in their respective situations + +Is this issue closable +Attaching a file with discp over viewfs tests +Thanks patch committed after +When Hbase Shell concept document in wiki and uml design document are done you can better start +Dan can you tell me when you have the generic invoke ready +Pending resolution +Where the old buffer manager cleans a page synchronously in the user thread and thereby updates the statistics the new buffer manager may have given the page to the background thread and therefore doesnt update the statistics until a moment laterCompressing the tables sounds like a good idea but it does indeed look strange with the extra page +Libor can we strip the p tags +The propoesd fix is not to use key but to concatenate the field name with key +On a view change all members in the current view are removed +The subproject hosts the Accumulo and abstract inputoutput formats and any future derived implementations +Closed after release. +Apologies The work around to this problem is to include a on the should be The work around to this problem is toinclude a on the +In the RC previous to this one there were some problems with and the enum type +the change I can think of is that the is no longer loaded by default through Groovy +Thanks to Val Blant. +Released into PROD +Im continuing along the path Chris started with his first patch mine is coming soon +But who knows +When resolving transitive features the version is not taken into account. +No need for +In my experience there is little difference in variability run to run on the same instances as opposed to two parallel runs started around the same time +Kathey noticed the issue with conversions of identical values earlier +Passed everything else. +Property cleaningOrder marked as deprecated. +If they are active they can be very disruptive during their formation +We might need to wait for the next Redback release for the fixes +I had the same issue where my mirror was not being used downloads were always attemted from central +The browser is caching SWF or howto blockcache event or data inside SWFThe SWFFlash is separated runtime engine +The process definition contains three nodes +But do we really need a own logging facade? I dont like to copy the classes from Trinidad +we have a tab for Receiving products +Alan can you run this query without desc and see if it still produces wrong results + to trunk. +bulk move of all unresolved issues from to +My main point is if we can fix it in the source then it would require no tweaking in the user side +Hi kalle I am using +Thanks Kim and Micky for the quick reviewAttached is and an updated version of containing the correction to noted by Kim +The thing is that when I reopened this JIRA Len asked me to make sure it gets in the release notes +Sonar seems to be ignoring that the is set to in my build +Not everyone is familiar with maven andwill look under Project InfoDependenciescheers Roland +For the cpp broker released messages are requeued at the head of the queue so the order that messages are released determines the order in which they are placed back on the queue +If there is a specific problem that can be identified with this code please or enter a new JIRA for that +Thanks SianPatch applied to LUNI module at repo revision Please check that the patch was applied as you expected +Well as long as the patch passes kernel tests it s OK to commit +Ill put together another test case and then add your suggested fix +This is going to be a low priority bug +So all issues not directly assigned to an AS release are being closed +I just checked MRVJT code and it doesnt look like we kill any running tasks +Then create a table that has actual data in it that is bigger than k +Tested running from via a symlink works and that setting HADOOPCOREHOME from HDFSMR root can start the daemons +If yes log the closing socket info and close the key +integrated in revision Anton please check that it was integrated as is done as expected thanks Mikhail. +But internally here also we have not started working on this yet +If there are better ways of doing it then the code can be replaced in the future. +Ah +They are expensive but given the result confirming it with those could be justified +Did not bother to post to review board for this very small simple patch +Updated in jBPM +New patch now attached as please ignore previous file +Thanks Doug for the clarificationIf that is the case we can move it to closure because your expectation has been handled in trunk codeI can see the references only in file +check all tables of attribute before a week till the release in order to be sure that all attributes descriptions are present +Seems good now +I will cut the branch from master on that day +All those issues are problem of CDI not a problem of JSF +The fact that the document is missing prevents me from knowing it has changedSo in order for this idea to work I would need to touch the document for every update conflict I catch in my app and hope no one screws up things by changing the database from the outside of the appThat doesnt sound good to me +So Ive added such a method +With that in mind is this patch reasonable +This long witing issue has been fixed +Exclusive flag no longer treated as implying +Since the Development Phase is over please move the Jira to Resolved status +Hi JashaFailing tests are indeed blocking +Much better! I hope that users will notice +Ill reworking some of the tests to address Owens concernsThe patch also needs to be updated for the latest trunkIll create a target to run of the new tests which take a while to run +Im on board with whatever you think is best +Much appreciated +how much RAM do you have? I would think GB would kind of work +flavio Yeah I think and this jira are related +This class makes use of now and is much more simple and robust +What are the advantages to adding this to Avro instead ofPrimarily that it can be used with older Hadoop releasesThe Avro version of the patch is at +I dont know if this show up as a bug though columns shadowed by a row tombstone are removed by and column tombstones are removed before being returned to the client +Fixed for Designer +Note that at this time there is no register of permitted functions and so any remote user can execute any function as the database user + unscheduled issues to the next release +Ive also attached the resulting HTML files +When you do patches please dont reorder the code as the patch changes too muchAnd the template components such as freemarker velocity stringtemplate scalete are similar and consistent +Juergen could you also specify the JNDI name of s runtime JMX server as specified in Robs post on the forumThanks point! Ive added the corresponding JNDI name and a configuration +Thank you much Randall whenever I tried those with P with a interactive shell it wouldnt actually modify the process limit +the server manager +from my remote systems +I cannot reproduce this +This bug has already been fixed and can be closed +Latest patch looks great +Problem is that in that method an array of fixed size is created +apply patches +There wont be a release +The class is NOT a test its name doesnt even start with a properly Test keyword +Anton attach please the example project +Added first parts of solution unit test to point out simplest problematic case resolution of mapping from formal type variables into declared type but bigger change is to pipe through this context to use when resolving field and method types +Records are inserted in the DB but Railo never returns control +With maven does your repo path contains spaces +Change to Failed to save secure ldap password +Also No such file or directory is clearly wrong +Ill try to put one together for you +solrupdate as would not modify the defaultsOrB Assume that all support a GET parameter committrueIn that case we could append committrue to the given URLI know for a fact that solrupdatecsvcommittrue will work +Would a patch for this come in handy? It seems like a rather simple change but perhaps it would save some time for me to provide a tested patch +no new tests required +Ack not so easy +So its no problem with complexity and concurrency +only remove dead thread instead of all threads +Committed at subversion revision +moving to M + +bq +should just create the schema in the ctor rather than lazily later +for some of the protocols such where clientPrincipal is needed uses the second part as host name for validation +I will be trying this patch on Lucene to see if it fixes my production memory leak with Tomcat Redhat Java +If I get a on this test then I can work on adding a second test that performs the same sequence of operations but for a quorum configuration instead +The is the time it took for AFTER to complete compared to BEFORE +note I can see that there are files like and in their projects but that is runtime specific config not editor related +Moving these all to a Doc release version. +I have also been playing around with abdto set the application dir but I had no luck with the deployment items +I alway use mvn clean install Pfastinstall to kick off the building process after I update the cxf codes to avoid some weird error + +For now Ive just disabled the main +The build for still fails +Any updates +Start an runtime to run Tuscany with it needs to resolve bundle dependencies on the ServletJSP +I guess youll have to wait for a couple of hours public CVS is typically a while behind +My main reason for extracting get and get out of was that they look like utility methods and I was hoping to reuse them in +Can you maybe change the background to a smooth one and center the forge logo? Are there other optionsThanks +This patch adds to the list of versions on Derbys rdf descriptor +I am not sure if it should be session or statement levelI am leaning to it being session level but would like to see a discussion +Turning on and off Airplane mode breaks it? Yeah this is a duplicate of which is why youre experiencing timeouts. +Ill attach two small patches to make Spring IDE compile against the current Mylyn CVS head +Provide link under Job Details page Time Init column for each entry to individualized for initialization breakdown +Ill have a look at it for the RC +Very very huge +Fixed already in EmmanuelI first noticed this w a build from the beginning of sep +what is the use case for this +Are there any plans to address this issue +Good call +The issue was caused by the block indexing changes +I dont think all of this matters as the problems seem to be on a receiving side but still thought I would mention this +I see that was added to the features +Could please tell me how can I retrieve the current web in theThank you +Original issue fixed possibly open up a new issue if we deem the other issue important enough to implement. +OK new patch attached Ported to flex and cutover to +v of patch including javadoc +Since we need this fix fairly soon and it would be a shame to have to reinvent the wheel Id really appreciate it if you could share your solution to this bug +it makes sense that empty strings are converted to nulls this is what of users expectI have to disagree on this generalized statementwhat if the holds an Integer? we cant push an empty string into the model then but we can a null +Since the Logging library is compiled to Java Im not sure this would work anyway +I guess this one is long solved +Avro and Trevni are specifically designed for Hadoop workloads and other tools like Pig are including Trevni immediately +jar and newer core +A reminder to all that this isnt part of our release process +putting for M but I assume this needs fixing ASAP +I added a mention of the default connection mode to the comment on the turn methodI also updated the descriptions of the programs to explain why I am creating the database as mary in addition to specifying her as the user +Can we just add one +TedIll apply this patch against DTD +The tests failed because of the test run by Hudson could not find some of the core class +Heiko how are you invoking the source generator? I have just tried to run the source generator against the XSD provided and after providing a binding file to fix some name resolution conflicts the source generator creates successfully the Java classes from the XML Schema +Is there any chance it will be addressed +Ill take a look +Breaks on FF and above tooNew issue created after release. +Thus proposed workaround does not work for me +Closing all issues created in and before which are marked resolved +Requirements defined in this Jira is complete +My hope is that our new bridge logic will just work and that set doesnt have to become a supported API +Considering the timing I propose that we not fix this for JDO +Do we still need to unwrap the exception +I had to preserve pnendpointt as the first field in the struct because the implementation needs to be able to cast pnconnectiont to pnendpointt. +Resolving then. +changes some of the log levels to DEBUG in tasks +Investigating +I should be more specificI followed all the steps in the user guide +Workspace is CProjects and the project is located in CProjects so the project is within the eclipse path to is CProjectsresources very important +And you should be able to generate the idea project files for the examplecd examplescd mvn idea package of the exampleNeed to find all the jars needed on the CP yourself +See error log for more details +So you can either rework it as suggested above or remove it completelyWe can make the log INFO or even DEBUG but then more people wont see it at all and will file tickets here when they realize that their code doesnt work for IE at allYou can always set ERROR level as minimum for this class in your logging configuration. +Two classes replaced and some +What is the default runtime library used in Mac OS X +The main problem is that the patch is not aviable anymore +committed to master. +Committed to trunk +Yeah my worry is every zk node listening for changes on every other zk node +It seems what were requesting is to push the remote back into the server and make it editable +The code changes are H I and J are now +What are you using this for +I agree that PE should be updated to call first when required or at least this should be specified in the documentationI never looked at +Related to needs to be of type File in the mojo instead of String I suspect +Im shifting this issue to M but I think well probably end up with this as a Wont Fix +Ive tested the built from svn and it does work correctly now +If someone would like to create a patch for this feel free otherwise it will take me a couple of weeks to get one +I can pass to set +A new version of JG will be incoroporated before tomorrow I hope +Closing all resolved issues due to a successful release +Thanks Edward My initial observation is It took approximately seconds extra time to complete the full test run on my unix test box +Fixed. +This is a bug in in both and +Jimmy is this change important enough to get done in? Or can it wait till +Need more use cases to bring this into the language +It was on my mental TODO list but I trust JIRA more as a TODO list. +I suppose the confusion comes from the different signatures for the methods that create exceptions in the client and in the engine +Tests look good +Added for code and the for code as their names imply +largeWhile we were chasing down the bug we were seeing related to we had to make changes to functions that were implemented in headers causing large parts of the library that didnt really change to be rebuilt. + +Can you add your findings for that to? +I dont get goals mentioned do parse the and files to generate the is it you want them to do that they dont +Tried but could not able to create a reproducible project but the error i can still replicate in my actual project +I verified that the site builds and looks ok +I reopened so a release note can be added descibing the user impact and describe symptoms users might encounter +Looks good +Running tests nowTouches the following fileM javaengineorgapachederbyloc +So maybe this already would be good enough +hopefully not too optimistic +The underlying Antlr exceptions are now never nested as the cause of +classes wont generate an import +Attached new version of the patch +Need to make it clear that the new config para key is in percentage of of keys +The Jira fixes ant update does not exist! problem for stress testsIve add this message to allow search of this fix by key words does not exist +Applied thanks JamesI dropped the deprecated Leniant methods as theres not been a release +Its a minor issue the NPE is not entirely inappropriate but it is certainly unhelpful +Userstcunningesb testStep Make sure your server is set up with jbossws and ESB +v. +Patch Convert initialiser strings to a byte array usingA helper class is introduced as getBytes with an encoding my throw which is not allowed for initialiser +Thanks Gunther +Ok but what about the upgrade path for older installations of! We have a rather large wiki at the moment which is based on Latin and we would not be able to upgrade if this is not supported in future versions +Thanks Aaron! +bulk move of all unresolved issues from to +Do you have a working simple test case reproducing the bug this would make finding the problem easier +Merged the upstream fix to EAP +The workload to fix this issue and the advantages for the user bear no proportion +bq +I suspect a more subtle issue in camel related to calling commmitrollback w nested transacted routes +Superceded by its hard to say when will be released but is available and Im under a bit of pressure to remove the SNAPSHOT artifact from our repos +Attaching a catalog which addresses the issue of remote access for the schemas +hi! I think I know where the problem is but I have a conference next week and Im overloadedWill fix it end of next week +The example from Pepsi is for illustration purposes not to actually be scanned +This is to protect against bad code since the security intent is clear even if the application code is not doing everything it can and should + +Blasin rev I removed double ampersands from auto because it was breaking the lookup autocompleterWhy do we need double ampersands there in case that auto is not a real url but part of javscript codeBilgin +Please check that it was applied as expected +js files +The test failure is in HDFS which I didnt change at all +Fix for typo in type and merged into the master branch. +Verified that the patch fixes and +Users who pass a null HBC will soon learn that this will not work why do they care about unit testing +fixed patch to not change +Any code that uses Hibernate has been removed from Struts due to license incompatibilities. +As far as Guvnor module is concerned module is just a repository that stores backend data +Its almost there but therere some xmlbeans processing and additional resources specified which all need to be checked out carefully before we announce its final migration step +Comments in the jiras will be addressed +just committed this +Ive committed this and merged to . +See the matrix +Thanks Carl! +the attached patch merges the change over +Thanks Sharad +I was in the process of updating this sample application to Spring so my working copy is currently a bit ll solve the reported problems as soon as well decide if upgrading SM to Spring or staying with +Wait a minute + are added to the component tree not until rendering of markup this is too late for header contributions +and committed to +I think it make sense to support this so would accept a patch here +Check where was written ls Applicationsconf Applicationsconf No such file or directory ls + +Could I get a patch? Its a showstopper issue for me as I posted in the duplicate bug + to trunk and +was a permissions problem +Since a single shard wont split up the documentsI doubt this is the root of what youre seeingBut yeah a stack trace would tell us for certainAnd Mark committed which uses a differentmechanism anywayTo recap I doubt this issue is a problem in setups. +Even if is backed out of we still need to solve the compilation error reported Dec Actually it is trivial to use reflection to check for in so that can work on hadoop The tricky part is the Override in Konstantin is going to integrate into hadoop but it would take more than a month for to be released +Even if we wanted to change this we need to couple this to a model version update or with existing POM is at danger. +I tested it by running some class which use the changed method +Covered with selenium test +What kind of urls do you need to set the schema for? to external servlets? Within a portlet Im not sure if this is needed +Patch for branch +A patch that solves the problemI think the real fix should be in but I cannot make it working without making an endless redirect loop. +There is support for binary encoding also but not multipart operations with application encoding are limited to single +This patch seems to fix the problem +A test revealing the bug +Have you tried make k check +I am uploading a patch for trunk as well +An example client and server in the tutorial folder would be helpful as! looking forward to see tutorial +There seems to be some consensus that moving libraries out of CommonIm sure Id call it consensus yet we need a discussion before we reach consensus +Finally the packages should probably be in rather than +this is the patch for +So you are saying that the method should take a and then cast to a? What is the benefit of that since it must be an instance of the latter anywayAgain please explain what the actual problem is that this causes +Patch looks good +Hi Dave +If you want I can fix that +This is not a bug in but in the match logic thats reusing it across multiple m going to try an experiment that avoids reusing at all and only sets it but does not read it during matching this should isolate match results +It might also be worth analyzing the dependencies of the plugin +But CDI Tools based on Eclipse JDT and its design time model +Unresolved issues to CR +The recurrent connections and log messages both are gone from the client +ah thanks +Should this have been closed as Invalid instead +It works with JBT on the same Linux approved and merged on master and with JBDS +Thanks for your comments + is of type which is part of the public API for Java SE +If its going to be global itll need to get unregistered during webapp unload whereas if its local its nice and for whether or not one would ever want to register more than one simultaneously who knows but going local avoids the question +httpslocalhostecommercecontrolsetprofiledefaultviewprofileproductdefaultpartyId JacquesPlease find the patch in attachment I hope that this should find the issueRegardsSumit Panit +Thats the behavior Ive seen +my deployment was still failing with a error on +Thanks Vikas for giving valuable suggestions no changes in the functionality changed the implementation logic used java script +Implemented in plugin v +Hey this works great now! Thanks +If we go with plan it would be good to minimize public methods easier for refactoring +In the latter case we can interrupt the thread that executes the system command if it takes too long or user interrupts the job +For example in case of hadoop properties this is what happens +the latest patch looks good to meThanks Andrew +In the Derby code tabs are normally four characters wide not eight +If it is and youre so inclined please try the patch attached to which rewrites branch instructions to avoid branches that exceed the KB relative offset limit and which also attempts to break up methods that would have been larger than KB in length into smaller methodsAs I dont have access to your stylesheet I am unable to try it myself. +My particular use case now was in the generated runtime but I guess these two things should become one soon +Is there any update on this issue? I know we have workaround for thisBut it seems this issue is blocking the release of new version havent had time to finish this yetThere is no rush to get out the door +is returning true is incorrectIt uses to delegate to which does not implement is +It surely doesnt depend on class implementationTheres some code in +Just realized I used afpanel instead of afpanel +so I think we should start a separate bug about themThanks for helping us tidy these cases up +Its probably more efficient to create new instances in that case Garbage collection is really good at handling instances these days +and rev +If this is a CREATE or ALTER TABLE statement then is not a column in the target tableI also see this error when running against +This should be done now +wsdl and the file +Need to reflect this in plugin +Closing +Im using the default navigation generated from the multiproject +check component to work in work ok with current of a bulk update all resolutions changed to done please review history to original resolution type +No it looks like it is not related to the minOccurs attribute as the error occurs within the and not the. +I create bug on it +For big regions you will be limited by the network anyway +Thank you Brandon for the review +Makes sense and could place that profile outside AS too for even cleaner operation +This is a rendering optimization so you get to consider it Jesse +pass +Average is not robust for this if you have a couple of dead datanodes theyll skew the mean upDead nodes dont contribute to avg +Theres nothing we can do about it +Need to add handlers to important configuration in the book though +Im in the process of releasing today and the fixes for this issue will be included in that release +It does makes some sense for this code to live in an apache namespace +Patch proposal attached which allows to set autoCommit mode and eliminate related overheadBy default autoCommit mode is still set to true ensuring that no changes are required on existing repositories +Also backported it to and notice may have more problem about optional dependenciesWe should test again carefully +Will open separate issue to deal with this failure +If its inside a data table thats another matter entirely +I never understood exactly how this was supposed to work or what it would look like in the API +. +If someone can point me in the direction of the fix I am happy to submit a m also happy to pay a bounty if someone out there can fix it quick +Or simply have a startup check in init scripts or a java main. +As far as I can see you should be able to override doHead without any issues +Lets use the new method +Removed old from and in r +Ill run perf tests of Kuromoji and post backIll the JAR size impact I think it was only a bit larger +I believe this is caused by code that was fixed in r +So lets leave the scheduler instance inbq +Everything works fine for me and even Hudson is fine now +We can ask them for feedback and thoughts on can you please do this +Final commit is fdcbfbdccccc +Nice work +I did some deeper investigation in this and found the disturbing problemIssue is between ivy and its use ofBasically when encounters a it asks credentials store for credentials again +LGTM. +visual web tools is corrected +We were able to do the same with a different regex which was as fast as for the feedback +so Im not sure its easy to establish a rock solid rule to know when to do and when to put them in a single ticket +Fix committed to trunk and +If I regenerate the jar with the same code and copy in the jar I did not see a problem +the in the Jackrabbit performance test suite measures per logins +Concurrent changes to entities of different classes are not affected. +Last call +ended up fixing this. +So we dont need addrubymethods but can use Hash methods against Javas Map object. +I committed a lot more changes that among other things dont drop any service removed or modified events +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +It was just a case where the file itself needed to be flexible enough to support output under Struts and +the application server +I think that this issue really belongs in but first let us see if it hasnt been fixed upgrade to version of and try again +Hmm an administration question why does this bug cloned from +I agree that being able to configure the default partitioner like we currently do with the default mapperreducer would be best +Actually I consider that a feature The Spring IDE Extension to the WTP XML Editor is not tied to the Spring nature +I also tried this command below provided by Dave Grove followed by the g command in the description and it creates a +I vote to commit +As far as I can see breaking the path down isnt the main issue +Attaching new patch for current trunk because just got merged to trunk +Please reviewSumit ok +I do not want to deploy it +I have attached which includes stax +However that is outside the scope of this patchM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceSome new error messages +This patch removes the part of killing a reduce task if another speculative instance is running due to issue reported in +With having to have a semicolon as line ending the parser can know the expression does not end till the semi is reached and therefore put everything in the expression even if there a re new lines +Temporary workaround Extend and override supports set it to true +sleepBookie +close for both cases +In a post I made to the list a while back from my investigation it appeared that the culprit was in the code +Currently hbase doesnt have a separate client jar that can be obtained through maven +Resolving issue +The code should now check all type nodes for missing parameters so Im resolving this as fixed +WAR demonstrating theThanks for your responseIm afraid the wire log isnt very helpful it shows the presenceabsence of the Expect header and shows the body being pulled either duringbefore the request processing pretty much what youd expect +Im still seeing some diffs in store tests but they dont look related to this issue they look like a timingharness issue might just be on my machine +I asked in the forums but not reply there either +Also a major rewrite of the parser would allow for improvements like preprocessor directives macros and make it easier to different syntax or languages +thanks Angela +It is only IE which has the layout facelets file contains a simple set of tabs which exhibit the problem +Looks correct +Same patch as v but add all missing apache license headers to +I dont think its feasible to replace all implementations when any one of them changes and it would be nigh impossible to know whether its intended to replace one of them to reuse that name or to provide a new implementation +Changed the create table statement to upper case and removed the semicolon from the select statement. +just to the loose end with the branch could you please back out the change you made to the branch +Please make these new expressions serializable +This patch intended to demonstrate how what constants on OS X should look like +Thank you +Thread dump that demonstrates problem a lot of threads are blocked on +Is there any chance I could see the HTTP packets generated by any of the browsers? ticket can now be closed because it is possible with to override the cookiespec definitions without changing the core httpcomponents library +with correct path DFSWRITERETRIES as its not really necessary with the new kernels +You can setup a timeout in milli seconds and run so that failures would be minimal +SEGMENTls CRAWLPATHsegments sort n tail n is not a good option as it wont work in deploy mode only in local whereas using hadoop fs ls works in both cases +Attaching the UI my setup I was able to successfully resize my default disk from gb and the UI updated the displayed parameters accordingly +Verified onFAILURE All components are not recognized on the Visual Source panes and on the Preview tab +Before starting the upgrade log into a Datanode and delete a particular file +I checked in fix to master and only +Devs Can we get this done for release Or does it take some more time +Actually if connection is null this means uses a cached session and the underlying connection is already started + +bulk close of wont fix etc resolved issues. +Yes I will try the attached patch and report back +This is currently marked as a bug but I dont think it is a bug its an improvement +It is a good idea to completely solve the problem by either adding NN verification logic or let client notify DN and let DN verify it +Yeah this is working completely incorrectly for my projects +Patch to set pos length when appropriate +Ill leave them on list for a day push them and update this ticket +Why should I disagree? I am just not a big fan of leaving issues rot without a resolution of some sort even if that resolution is to do nothing +running +Rebuild the vdoclet plugin and installed it +Back a few months ago the cleanup hasnt thrown any exception +Some more context please +Thanks for the info +Thanks +name svn diff srcmainjavaorgapachehadoophbaseclient TedIll add a test +please ignore +I updated to qpid svn and msgstore svn rebuilt everything and it stopped croaking now errors with exactly the right message +This patch has the rpc server handlers discard any call that is older than of the +Thanks Yohei! +I just downloaded it but if I check in hibernate code there is no trace of a patch allowing to assign value into variables +All the other public classes are marked privateTagged as private and evolving since it is used in +Not catching general failure now reporting it in an error dialog as before +I agree it would be nice to control this behaviour. +can be reopened to hunt down the root cause +Configuring deployment dependencies would probably not work in the case of two independent servers +Please check that the patch was applied as you expected. +Pushing this to since I cant reproduce it +I wonder whether a specific subclass might be more better +Apologies for delay old mirror deleted. +Just encountered this ourselves will try patch applied thanks! +You dont have any edits to the Avro version in the POM right +Ill deploy the artifacts later today. +Updated patch with generated artifact and service name changed to test failures are unrelated to this patch +I am working on applying this and merging it forward +ramYes I forget to consider TTLSo we should create empty file if no kvs after minor or major compaction +Like it is done when you log in for the first time and need to create a category +Not a big deal +no more RPC +Confirmed +did a hard reset and the changes to try and get a cleaner patch in Aarons comments regarding the formatting and coding guideline for the info Aaron those are all definitely reasonable requirements and I should have caught those on the first round +HelloWhen I execute the goal sitestage using the it produces the following error The site does not exist please run sitesite firstI also tried mvn sitesite sitestagethat gives the error below Missing distribution management in project parentI have attached a little project to reproduce the original errorwithThe ZIP file contains the strange directory structure alsoIt seems to me it is somehow related to theIf the reference to the is removed from thePOM the error does not occur +Thus weve been compiling this way nearly the whole time on our project as a work around for the problems with groovyc tasks +Expanded URI variable values are encoded as well much like query parameter values appended to the URL are +Add cache clear command to hadoop add the clear method to is enough since only cache class need to be clear rightIll give a quick patch per Shinichis patch had been verified by manual the unread field per findbugs Steve RESET operation should be meaningful for the cache class only right? only the cached information need to be reset +That is probably another reason why there is no JBPAPP JIRA for that +But for the applications folks have mentioned here and the applications were using it for at Twitter realtime analytics system performance monitoring etc were completely happy to sacrifice a few counts in a failure scenario in order to get the perfscalability that offers +Maybe you have a compiler check turned on that I have turned off but in general I would say the most specific method is first to be found without generics which means the map method +On my end I generally always use the latest release or a SNAPSHOT build after latest if I need a specific fix +This does actually work +Thanks +Not sure +source is insde the warJust call +Most are transitively included because of hadoop dependency +Its dumb we dont respect locale +Also doesnt use tabs in source code +however Im still not sure assuming that problem could be fixed that would be the best approach to take since these renderers overriding the implementations renderers may still have other unforseen impacts on other implementations +This file apparently should be in the local directory +Tom can you take a look when you get a chance +I suggest you not make this style change in this patch but propose it separately if at all +imo your other priorities are more important i think +Doesnt the client always know the username of the current userProbably not if the user name is transferred to the server as HTTP Header authentication +Fixed in trunk by Yonik give it a spin and see if you can make it OOM now +In that case axis is in full control and host isIf you need to have a differant server user the not I purpose to close I agree let us set the server in r +Excellent +I refactored KB +Im confused what part of malformed request causes error is a bug? Or are you just wanting better error messages +Tried it few times in JBT +Andri nice spot! Thanks! +Will this futz with the liveSize information in +eg if the storage dirs are dataname and dataname it will ascribe an error in dataname +You want to match the root just specify the root with an asterisk and be done with it +Thanks Amareshwari! +is a enhancement of it extracts comparison statement to separate method and improves +Attached fix it want accounting for buffer offsets +bulk defer of unresolved bugs to +Would be great if anyone could send me thejava file + adds the new freemarker template and the hbmtemplate task to note that the is now required in the lib folder! hbmtemplate attempts to reformat XML files after processing them and will fail if it cannot find +Therefor I expect to have commited some mistake +SELECT statements +Given that counter are a specific case anyway as far as failed write is concerned I think just allowing BEGIN BATCH would probably not be a big deal though I do kind of like the idea of a BEGIN COUNTER BATCH syntax as it highlight the fact that counter and regular column cannot be mixed +So this is no longer needed. +I used a fresh build to generate the PDF and I passed the attached to FOP like this fop c have to generate the PDF from the IFfop c if applicationpdf fop c ifin can see the problem now +Sounds reasonable? I can help you to implement that workaround +ah yes maybe i broke it when i added the handling +comment fix is my bug fixed improved safe spring compatible version on steroids +Missing UML diagram of the ESB page +Fix merged into trunk. +Either the patch needs to accomodate the proposal in or needs to be updated with a new proposal +Plan is to include in JBDS +Here is an updated version for current trunk +If yes what kind of tasks it can acceptMany reasons may cause task failing on a nodeMost common one is resource limitA task tracker may be ok to run one task but may fail to run more tasks simutaneouslyA task tracker may be ok to run concurrent tasks of one job but may fail to run concurrent tasks of other jobsThe task tracker should accumulate some stats and make decision based on the statsA simple heuristics may run like this When a task fail the task tracker records the concurrent number of tasks running on the tracker at that timeIf multiple tasks fail at the same concurrence level the TT should stop asking for new tasks until the concurrence level drop lowerAfter running for a while without task failure it can bump up the concurrence threahold level againcertainly the TT can remember the jobs of failed tasks and decide whether to refuse tasks of those jobs only or refuse tasks of all jobs +Its useless because Sonar has its own code parser and does not use results at allDoes the build fail or is it only an error log +Is the problem specific to? Im using and have not encountered the problem + should fix this bug +You also avoid the dependency which is in some use cases a bit annoyingI still really like the idea with the java script api +to doc should also point to Customer Portal rather than +Updated patch to reflect changes in the CVS repository +In version this sentence disappeared +Unfortunately I have made several independant fixes in the same file so cant really do SebasienI cant find this snippet of code in the trunk version of Any +we hope this patch will check into trunk +changingblocks size hwsectorsizetoblocks size STOREBLOCKSIZEseems to resolve the issue +Since adopted threads get added to the default thread group this would leak instances which reference wrappers which hold +Which must be deferred so the route is fully started so when the job is triggered it has a fully started route to process +Can we add +The is forsrctestjavaorggeotoolsrenderershape closing all issues in resolved state that have not been reopened nor commented over in the last month +Right now if anything goes wrong you get into untested code pathsWhat do you think Himanshu +I am using Facelets +Maybe that is the issue here because the cross validator code can only split the training data on a per sample basis +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Hi Jan is there something else that I can do to keep this patch moving on +I have the same problem when I try to rebuild project even when I use fresh working copy from repository +Tomcat JDK POI +A whole new filter that also does input text normalization such as removing double spaces and what not +Bulk close for release +Lets close that one and leave it undefined in the specification +The that fails was created in the same threadby the constructor +Then there must be a content assistant associated with that content type +core tests passed +Also there was a rpcSinkSource added that uses a parameter in the xml properties file +So the only allows one but other implementations may allow morebq +A quick look at and shows that they may also result in this same exception +Attaching new version addressing comments and now with more testing +See +Initial screen now displays version number as in IR +tr is fine for me too +includes a cf param now + should also take in a Locale to perform the same sort of comparisons +All plugin languages must now feed those two metrics complexityinclasses and complexityinfunctions +Can you try this patchOr provide a quickstart to test it myselfThanks +I dont see them +If youre exposing Cassandra directly to attackers there are easier ways to DOS it +Thanks for the info +close after for comments and it make sense. +Ok lets reopen this issue +the snapshot I am using is label like +Its testing an area of the code unrelated to the changes and the test passes for me when I run it locally +But the source is commented properly and I do not see what else can be done + This is a trivial patch it adds to +I think Russels idea of a javagroovy directory is a good one +I created portlet using JDK version and then deployed it in liferay portal using JRE version error Bad version number in +Thanks for your contribution Kousuke +Will file a new jira if Sid can reproduce on his mac +The plugin should throw a mojo or mojo to stop the build +The issue is that with nothing in your local repo Maven cant find the snapshot of in order to inherit the repository definitions +Any plan for this? metrics especially statistics is not very ganglia friendly +I did not review in detail +This patch fixes the problem and adds several unit tests for various cases happened in various clauses +Its not that bad since the previous method of enabling zookeeper with security still works +They may not receive even write loadWe should set configuration parameter which governs the maximum number of concurrent on each region server +Lets remove this outdated code +committed to trunk and +Patch looks good to me +Patch looks great +And again please initiate discussion in Users Mailing List by explaining what you want to achieve and which problems do you have in order to get help from communityThanks. +USE WITH CAREIn this patch all occurences of stricmp in the xalan sources are located + +A couple of long nights with the debugger helped but then the resolve semantics seemed to change when I upgraded to ivy +Jonathan this is an umbrella JIRA +Attached patch that does the following +The way it is currently setup the are used by a lot of modules during compile time +So potentially all the Participants in a shared data context may be ownershosts +OK +However I have never tried creating an assembly jar with the same name as the actual artifact jar +Thanks! +Thanks for the fix Julie and for the detailed README text +This time it is caused by exiting prematurely +Not a blocker for SOA +not after adding a serialization patch. +Also think about how to mark Messages that are scheduled as being required to fire on restart even if scheduled time has passed +Im currently on the go will have a look at it later when back at home +This tab would not be associated with any particular repository or output connector +The changes are slight +Even if we had in now youd still need this support to test against or test against a HDFS cluster +Weve generally tried to avoid exporting any information on just why authentication failed +Perhaps this is related to +Let me know if this issue is not resolved +explains GBY aggregation etc. +after paying attention to the big fat startup warning about using JDK instead of JRE the issue seems you think there is anything left here please reopen and provide complete test project that would allow us to reproduce this issue. +It should be a patch for as soon as but this still should be acceptable since isolation levels are a minimum level of isolation that your app can deal with +Fixed +what would be the right place for such a fixWhat is the thing that is assumed too muchSee also also +I cant seem to get problem even if I boost the validity period and compare the expiry time in the test method with the one in the token they match OK +This patch merges Kans patch with the trunk with some slight changes +It didnt completely work From forum comments Im not the only person who tried to create objects this way so adding the function would save others from the missteps I made +I think this issue will be ok +pending to trunk and issue as it has been released as part of release +Adds more integration tests with embedded derby apart from mysql which is disabled by default +Thanks Gary! +Im not too familiar with xdoclet plugin development but I will have a try +These config files will be incorporated after the EAP GA are releasedNote that after incorporating the config files into the upstream repo or the prod branch well need to make some changes to the RPM +Windows is enabled by defaultOther potential sync issues should be documented as possible causes of corruption Running Derby on a NFS mounted disk Using runtime compression which can cause sync to fail on disk fullIn general Derby needs to know that the sync has absolutely written the data to disk in order to ensure recovery +bq +Ill file a JIRA about fixing +if theres other testing related code that you think would be useful for other people we should refactor it +In revision I committed the change so that the Validator lookup code reuses the same for the same application +It would be worth working out in details what the answer to that should be in due course in fact +Committed revision . + Should there be a new method or parameter forcing not to cache the terms Digy Yes I think there should be +Native Installation Notesg Missing a section for CXF alternative installation from +committe dnow. +I believe that currently is the only thing that needed patching in our entire for ranges to work +Sorry for the delay on various different things +I just committed this. +Sure I can improve the error messageFor HADOOPCLASSPATH when we run pig binpig script will put into HADOOPCLASSPATH which contains +In that way if you started with a small amount of shards and both of those documents wound up on the same shard the search would be restricted to just the shard where they reside and the filter would be applied to only include the docs you want +Investigating +merged down a slow impl is better than an exception +Thanks for reporting +It gives a poor initial impression of this plugin that you have to delete and your whole Eclipse project just because you clicked Convert to Drools Project by accidentTake a look at how other plugins solve the issue for instance some add a Remove +I should note that Im also using the Run Groovy option to run my test cases frequently from within eclipse and I notice that eclipses processes increases a lot each time I Run it +That helps especially if your link should not work for some reason +Currently Im running the version but the earlier work on this issue was done on +I still see killed tasks is runningTasks at Groomserver +Yes the scope is +HelloThank you for your analysis +We need a DN that has checked in and then failed to check in if I remember the repro conditions correctly +Duplicate of. +for the patch +I defined constants put and delete +Thanks Andrew! +This patch also includes the fix to +This has been implemented and checked in into the branch. +DoneGreetings the patch with thanks to Piotr after fixing the missing file issue of +sorry dont have it anymore Ive blown these resources away already +management I am assuming this will be fixed in +Wont Fix. +Sorry for the inconvenience. +I added a so that we can can have a some control over whatthe style is +Applied to +Hi BenjaminI am afraid bit its less likely that I am using Maven as is the only version available within my company and I have that thats the only version I have on my can you suggestSanjay log you produced was that by running Maven via the command line or from inside an IDE +There was a bug in my patch I added another unit test for thisI think its ready +Java Stack trace for the can queue DFS writes but still might wish to have disk writes inline for the sake of immediate availability of for +I fixed the code to check for the value range +Sorry for the inconvenience +I might not be too familiar with Lucenes ngram tokenizer but I have been working with and Lucene before so I believe that I would be able to learn quickly +The big patch I do not think is a good idea to apply on the versions +Even this commandCLASSPATHlibbuild ant clean resourcegen +threads are not returned from so connection are freed even if they are in use causing deadlocks and other very nasty stuffIt would be good if any Java thread adopted by was returned fromWill it solve the problem if we call at the entry point and when the method returnsWe are currently avoiding the problem by increasing the connection pool size so it will never reach the upper limit but when usage increases we will have to find a better wayAny suggestions +Closing Jira. +Put in POM support +After no feedback Im closing this one +Shubham it would be good to open new issues for bugsenhancements rather than attaching patches to old issues especially of they are already resolved +Thanks +I think I see the problem its the partitionValue +Please note the situation has evolved since this this issue has been created +I have a guess now and will post soon +Verified on JBDS. +I wasnt no I only just started looking into cassandra let alone the source code so I assumed my rather crude patch would be taken over by someone with more knowledge of the internalsId be happy to put some time in though although Id probably need a bit of help guidance. +Sorry for my curiosity but could you give me an estimation when the issue will be resolved please? In fact its a show stopper for me and I would be glad to provide you any information you need +Id be curious to see how thats implemented if you feel up to the task +Eclipse Equinox Managing in the installation step process it is written about restartinglaunching Eclipse IDE not JBDS +It sucks too cause this project is the one I want to be working on but it isnt as high a priority to the biz so much and I will test it out as soon as I +Thanks for your effortsm going to resolve this as the client api is done +You can assign that issue to me +But I suspect the foo coming from outside and being loaded by a different Groovy runtime which of course then has a different +The issue tracker wont let me attach a file now that this is closed so please get these in your local copy or just reopen this one real quick +Attaching basic examples for patch +Closed after release. +mispelled function name +Similarly for get could be get since the AM is setting this should this be removed from the +Hope this is now better resolved. +Since password substitution can only be used with an authentication scheme which we think is unsafe we are sending mixed signals by continuing to document password substitution +Thanks Jitendra +The patch didnt apply to the branch as is if we want this in well have to prepare a separate patch. +I supplied a patch in that writes the archivename with the jar test and comment on it +If not we can revert thisSure lets revert this +Java is not worth supporting any more. +Going for them +Oh I love that idea +Thisll do on this issue for now +This feature makes the database creation scripts obsolete +Please review +Leading are not allowed and are stripped out other dollar signs are converted to underscores. +Move to post since this changes the really would like to see this fixed soon +It is just a backport of the trunk patch + +No test included because this only affects logging and there isnt a clear way of testing anyway as the logs were generated due to delays in waking up by OS +Thanks and RegardsVishal +the dbids have been added to servers as a prt of +Thanks Jonathan! I wont mention it in the docs then +is a tricky beast because it uses s Parameter mechanism +Works this doesnt work +Decided this was the simplest approach with the smallest amount of code +tried our performance test with a bounded queue thread pool as the executor for the web sockets and the issue does not occur +Here is the patch for ulimit n suppose you are talking about OSX you add the line ulimit n toIt will set the maximum number of open error message will disappear +complies with the letter of the spec but not its spirit +Patch that only processes schema if derby is the selected good +Hopefully no one is using anyway though +Regarding the documentation if you can provide a patch for the part that derailed you it would be great +Maybe Ive updated dojo or done something differently since you originally filed this issue. +tested and the new FT works with http headers passed as params +debugtraceshanging log file entries during hangdebugtracesstartup debug traces during startup with persisted messagesthreaddumpduringhang thread stack dump during else that can be tried? Can i provide any other information +mocking the property +Need to collect community feedback here +Bulk close for Solr +Thanks Eric! +Committed. +We are backlogged so we have some performance work todo. +need to run w right version of Eclipse +I just took a look at the MR code +I have merged this to. +Looks like Steps +Why should a page that isnt participating in the session rewrite urls that do participate in a session? What is the use case for some pages in an app not using sessions and some that do +Theres code at various places where we unconditionally create unchecksummed filesystemsI wonder whether it is possible to revert the entire change and start again +Yuki can you review +if objects will move this issue to +Enumerated works with as shown in the unit test case +for the patch if the above comments are taken care of +Instead of using th wrappers from you should use your own wrappers +The priority is Highlighter looks if there is and if it exists uses it +The issue has led to some productive experiments and discussion including the good suggestion that we should measure concurrency periodically at least for every release +Thanks JunjieThe patch looks good to me +Jostling around the release goals and numbers +Use in when schema is known +I understand from your threads elsewhere that you do not have the tests working with the scenario however this is not a huge problem +That appears to have fixed this problem in the internal lock and unlock methods I have added an effective lock if and only if the event channel is notified of the lock or unlock which would cause a lock and sync after the lock manager jvm lock had been aquired +Reopening to patch and +There is no code changes required support this fix +Passing to hudson +Let me go build an hadoop that has and applied +If you can get me that sample doc Ill include the patch +It is indeed not thread safe but that was kind of on purpose +It was some conflicts with my profiles +Ok +If framework pom defines some dependency that is used somewhere in framework module the property definition goes into framework pom and so onThanks for the patch btw! +There shouldnt be a requirement for a matching registration +The only thing it will give us is enough info to be able to go shoot the containers +Will somebody try the patchGood luckLeo + idea why you werent listed in bamboo as a griffon despot ldap says you and it came good +Client to test against latest builds of +I got it now and hence this issue can be closed. +Reopening this was a Bugzilla LATER. +Sorry the build server is down so youll have to build from CVS yourself +Tim any progress on this one + +HiI added all entries in Xplain tables to the s XML file +Searched for fails and could not find your jira +Hmmm we have two actuallyedExceptionWe need to support both cos they refer to different scenarios +therefore our fix should be removed after upgrade to +The test failure lines are most useful when a test fails and youre looking through the output in the test directory or on Hudson and may not have the context of the failure so Id prefer more than less but if youre ok with the current level I guess its fine +Lets rename it to +Does a data store play nicely in a clustered environmentYes +As a workaround define an alias for at least one of the identical column names in your query +New implementation works. +not to close it +Use should be able to move values from Second contact to first in a single click +Please close this issue if you are satisfied. +Column Families with wide rows need some love on the latency side too +As soon as I said that something odd happened +I remembr justing double checking the catalog supported workspace changes for stores what kind of error did you see? feel free to report and assign to me since the major databases are taken care of +I have moved the changes to the with new constructors for the line delimiter +Will revisit in Iain for keeping working on this + for using +I have no profile in my setting which makes complain +It would be better to simply published best practices stating that events should almost always be immutable but there may be a use case you havent thought of +Thanks m resolving the issue +Dont deploy jta +Over the weekend I had a query if I had found a work around +Try it in a new project +Later Ill submit another patch to maintain the existing semantics of pre +Patch is coming shortly +This WSDL was generated from Savara so a bug needs to be raised in the Savara jira for this. +which worked perfectlyMany also had the very same issue +is a java method we have to support java +Attachment has been added with description FO document with +Ok got it +resolving again so qa can is now limited and user friendly +Sandbox projects wouldnt create traffic on the core lists and wouldnt hold up core patches or releases +There are no outstanding issues with the tab example AFAIK +It is better to pursue solution in +Everything is the latestAPR +Dont copy the return value of aync method for now because is not serializable. +Wouldnt that solve the issue +Ok i changed the jar name +Arvind thanks for the reviewAttached is an updated patch to make the interval test more deterministic +There was a problem with the original patch that caused a null pointer exception +Please downgrade version of from to RC appears strange error Attempt to use a script that has been garbage collected +There is no test case added +wont be depending on Commons lang in Roo so and are not options +I think the per table config route would be best +are encouraged to migrate to SOLR for indexing as this will be maintained in future versions of NutchYour patch should be useful for users who have to use or older versions though thanks for sharing it +support is working well on CARGO beta +changed the log message +c +I can understand your concerns but it seems that if an interceptor method throws a that nulling out generated primary keys is a reasonable step to take before forwarding the exception on to the caller +Attribute size doesnt work at allstyle and styleClass attributes arent applied value is was fixed in patch related to +Do not checkin these patches +Patch commited as is. +The currentname can lead you to believe you should use it because you happen to have your content as a Java String +But some people will know in the newspaper biz we use pre processing filters to map anything outside of latin to latin else remove it +Attaching +Tom if you want to attach your latest patch as is I can see what I can do to add the ordinal named parameters bit +I just committed this +project this time this issue seems to be solved. +Farah can you please have a go with this? Its important to have a fix asap +here is an upstream for a design document please see varlib rather than a varrun path since varrun is a tmpfs on many modern linux distributions +Could this be a configuration issue with your or other file? The error is stating is cannot find the wsdl at wsdl +add another maven project that builds with multiprojectinstall build successful +DainFix applied to LUNI module and at repo revision rPlease check that the fix resolves the issue for you +The maven import is using the name that is in the maven pom not the name of the application +Done the general is now not the first one in the list +Typically just simple types primitives like dates Strings longs etc +This patch should apply up on on hudson +lenient? how does useonewayoperationsThe matrix could be split up in server and client side takes SEI as input and generates wsdl schema +Committed as r +b classfile syntax work for both Windows and Linux and alike +Attaching the file that goes to xdocsdocumentation and to add the conventions link to the sitemap +change in enhancements release +I was wondering whether +This will probably need to expose some API from Master +that pom s wrong anyway with snapshots missing license +Unmarking regression Ive gone back to saved history and found some earlier occurrences +Closing +minor version means that would start with minor version of +I have a fix which I am testing now +I wonder if this is a regression of by any chance +Attached example to reproduce +Made super configurable added tests and documentation +Rejecting and closing this issue as the reported tests no longer fail with the latest CI runs +Fixed by is still open but resolving that issue will also resolve this report +Patch v aligns with logic of +It seems to work for me with the netbeans daily builds mevenide daily build on macosx jdk think this issue is fixed in +One bug with this which I missed mentioning earlier +Indeed this is meant for synchronous use and without transactions at all since the message needs to be published immediately in order to synchronously wait for a +So now were effectively adding GWT and Spring into this Solr contrib right? Not against it just want it to know for sure what we are getting into +There is already something I found in the IT for that might need to be extended +Thanks Trev for discovering this and providing a patch +Applied with and to trunkSending assembliessrcizpackSending assembliessrcizpackSending modulessrcjavaorgapachegeronimoinstallerprocessingSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsTransmitting file dataCommitted revision +Okay done. +Ok +With this patch code in kernel module compiles and all the tests run +Konstantin At the end of a decommission process a datanode actually shuts down doesnt it? If so then using decommissioning might not work in the current scenarioAllen HDFS currently assumes that the entire disk space on the cluster is readablewritable +Create new issues for more features. +Closing resolved issues for already released versions +Restart Eclipse with a new workspace +bq +It would be great if we can make this patch into is a patch for +Aaron please feel free to commit on BranchStable and TRUNK + +However I am wondering why you have placed only the Interface into commons and put the implementations into FOP? Wouldnt it more make sense to put the implementations into commons too +Teiid Designer Branched back in August just for JBDS work and our project trunk has not changed and wont until were much closer to JBDS JBT +This may be helped by to denote this is only for as and the updates we made in to resolve this. +Thanks for the patch +Im not super happy with the verbiage in so please feel free to improveNote this is also a problem in and but any fix would be much more difficult since deleted docs are screened out at a very low level int he. +I think it makes sense to have the logging page first show events and only show the levels as a second stepIn the logging events are returned as a so that we can have a unified UI if the events come from a search or the simple RAM how about this oneThe Handling of right now is a bit poor bit ill have a look at the that is really smooth +Then could you please handle this case in the trunk the same as well? Thanks again Sam +Yes once we drop CR Ill update the staging site +IBM httpsupportrsdcDqwebsphereqwebsphereqjdkuidswglocenUScslangenThank Daryl Brian +IMO this should be the default url convention style if not for this reason the supplied patch aim to change the default or is it some how intended to be configurable? the attach patch is useful its a breaking change and something configurable is is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Ill keep my eye on it +Merged into the master branch and into the branch. +when you will resolve them could even resolve issue without additional review +the jcrsystem subtree doesnt get removed since its protected +Users would have to create their own module import the engine expose the same service and add the override properties +The only thing that needs work is caching +And all io threads will shutdown one after the other +I hope to get to this in the next couple of days +Oh sorry +Do you want also try to provide a patch for this improvementAs you probably know we love contributions and the probability increases that this issue will be resolved in the next Camel release +Thoughts +Added licence files and reformatted imports in my IDE +Support said file a ticket and I guess did not reproduce +Looks good to me +Rebase +Yes submitting for +Check mvn dependencytree or run with X to see where its coming from +Regarding the warning note that you added i think that you misinterpreted Ross comments in the email thread +There are three options +for patch +bulk assign all unscheduled issues to next release +Could someone please try this patchThanks +Trivial patch +The test is not valid as you never pass null into the iterator +When it is a timestamp based delete same TS to be used for all the columns where as in normal delete diff TS can be used +Heres the script that Ive been using +Reverting the patch +surefire report for the with MINA bith Junit test passTest set Tests run Failures Errors Skipped Time elapsed sec +I generated the on this index folderPlease let me know your thoughts on these output files +Nicolas any progress on this? Im having the same problem as Phil +Thanks. +Put a fix for this in CVSRegression testing in pass +Looked if we can avoid calling velocity functions that require in queries but got right on velocity runtime instance initialization +Patch for and we point to one commercial monitoring product we should probably point to all which is kind of a slippery slope +I believe this feature is and will therefore not modify +reopening issue because of failing junit testcasesTestsuite Tests run Failures Errors Time elapsed secenvironment wk sun jdk assigning to jukka since felix is on vacation +Just to be clear Im willing to consider providing a different deployment node selector implementation from within the EJB client API project if there are clear requirements on its algorithm for node selection and if its useful enough for client applications +The changes were made in and are in both HEAD and in branch. +Filed for improving the high watermark maintenanceFiled to cover the refactoring suggestions +Sorry my mistakeI intended this to be an enhancement request for Spock not GebThis request can be deleted. +I just raised related to this affecting Derby +so lets target that to M +Assuming it is okay. +patch has been committed to the traffic server trunk and dev branches +All produced from this input CAS will not be allowed to continue and instead will be intercepted and dropped +I dont understand why it doesnt grab the +Plamen I also just added you to the contributors group so you can be marked as assignee +Should be fixed with. +Another problem is when the model is null for example and the user key in some valid value what is to happen? If the method above is implemented since the model is null a proxy will be created that just returns null if any property is access and the error will still existsPat Jason what do you guys think +reviewing pull request merged also changed sqlite alias to point to the new sqlite entry +story of my lifeI got it building it took updating a lot of little things here and thereand some of the random js libs are still broken but progress is progressThanks for the reply. +Might as well Id hate to break the tradition of having the assembly plugin require a new release of after all +parent Sorry if this is a lot of loading with only works with constrainedtrue as documented +bq +bq +Not doing so would be a regression +The new certs have made there way to the branch and so will be in the next tag of that branch. +Just updated pluginsWhat Im seeing is that the save button now leaves the window dirtyPresumably before it was just failingSome interesting stack traces are being printedSee +I dont know how to say that any simplerI agree that the user login code is monolithic +Maven and ivy seem to be handling the transitive dependency resolution just fine already so I dont see a big reason to make the change +Committed thanks +and then try to open it in a browser +Committed patch a to trunk with revision Awaiting further confirmation from the nightly tests before backporting to the branch +The space between the parenthesis and the keywords however remain +Please see latest source code to see if this suits everyones needs +Thanks Patched applied +Thanks. +Adding the exception to the log as per review request +also fails on occasion for meNice work stackYoure thinking in w ROOT removal perhaps +As a workaround just remove from modules section inYves After quick review seems that your patch simply hides a real problem instead of solving +A few small comments in I think Santhosh has gone back and changed all the schema getSchema calls to just check m removing the check whether mSchema is null +I am also seeing this issue with AXIS +And we dont add +is the final change for the file name rename +The CF in the log entry is the one +Also made the other changes we talked about now uses and does not have a hash or compare function +This is basically the same issue +Yes Jetty causes a NPE on startup +Yes I did Thought nobody will noticeSorry about that I will add a test case that I meant to add very soon +Reopening to remove the from affected issue has been fixed in HEAD as well as the tck branch +The premature end of file originates in the codebase before it gets back to wise +Thanks for all +configured in commons logging explicitly orB +Its tricky because the optimizer overrides are specified on the query before its been flattened but I want to force a particular join order after flattening +I want to think about this a little more before closing the issue in case I missed more cases +Issue was caused by a wrong assumption in the GC that all branches are uniquely identifiable by their branch root id which is wrong as one can create two branches of the same commit. +definitly +The patch makes sense to me +Maybe its a now +fix which fails in build like out build system is stable now. +Ill be looking into this post release +Which library generates this HTML code? As soon as this is found out the lib can be replaced by a fixed version or a bug report can be filed in this project +Monday and Tuesday were a national holiday in Hungary that is why I did not read your answer until todayBy the way is there any plan to release OGNL and Struts depending on OGNL? I have seen some activity in but this year not so much. +I hadnt picked up that patch +Our CMIS server will internally be talking to repository and more propertiery repository +This looks +What exactly was the fix for this issueAnd is this issue the same as? If so can be closed as a duplicate +I cant think of any way to get rid of these files short of relying on ANTRUN to delete them after the test phase which feels like a realize that if I had a brand new project that I could modify I could find better ways of doing this but based on the comments Ive read here and elsewhere there seems to be a hundred different reasons why someone would need to exclude files from the patch should fix and integration tests for the includeexclude functionality +I believe that when you ask to sort you dont need scores +Hi The fix is fine thanks PaulexBest Regards by Jimmy. +We need to figure out where we can keep such properties that are useful only during core creation +Good idea! Effective with the nightly build any change made to theLocale at session attribute will be copied to the of the component tree so it will also affect JSF component rendering. +Ups wrong version selected +Im assuming that this update would go in the tuning guide as a separate topic in the Working with section +attribute appeared for tree butnot for to dev team the attribute should be only in tree table of attribute appear in dev guide in the corresponding section +Waiting for your review to good thanks for the patch +See SVN +This is really by design +agileIs this the problem caused by +Closing all resolvedfixed issues of already released versions of Roller. +I used linux form my workstation so it might be a reason why It is not working on windows Ill verify it +take this issue Assign to me at the top of this page +Pending resolution + +When this happened I got some Exceptions in the startup of Solr +Committed +In Maven a core fundamental concept is that the pom should describe everything needed to build your project +Passwordobfuscation does not help very much to achieve that +I encounterred the same problem +The behavior of the method is exactly as it should be for performance reasons +bq +Too risky for and fixed in +Im going to work on this today +Ill keep the unified patch up to date as that seems the easiest to review +For me it happened only the second time the page was rendered +Consigned to the great issue repository in the sky +fixed in r. +Correct me if ImI will try. +Fixed a typo +The next two runs took only seconds each +I have merged this to. +bq + easy shoudl extend andnot Exception +If you flush it then all regionservers all flush at same time killing any other server going on concurrently because of the io load on hdfs +Either way Ill apply today +seems to be refactored out of existence and Im too lazy to figure out where it went to +Thanks for the patch +Anyone has got a patch? Many made changes that should fix this issue and checked them into subversion +Perhaps there is a way to do this in Beta today? but if there is its not readily there is what classpath would it use to deduce the name +The simple patch +This patch checks if the disk is whenever an occurs +Android should not be under Web Development how about a Mobile Development category +Heres a possible way to organize the work fix leaving just the simple assertion on vertex ids add a more detailed test on checkpointing with different test cases to check for vertex idsvalues edges and messages +Can you post some of the diffs that you get failures on? I had a problem with running the tests on nfs mounted directories +Committed this to trunk and +And I cant reproduce it +Thanks all for your help +all tests pass under every locale with this patch +Kirill you are absolutely right that issue should have been another JIRA but I have fixed this already +And done +Example user A will have access to data in database db as well as db +Patch committedOleg +Jars have been published to the Apache repository which central proxies +Either this did not work for or something other than wise is likely to have changed +This dialog may create even more new images at start +Im pretty sure thats just cruft left behind when the child was named. +Im ready to commit this soon Id like to get it in for +Looks like the method went from being an instance method to a class method +Great work +EricGet the latest trunk +So you should be all set +nit Few more occurrence of white spaces in +add testcase for s scenairo. +I did experiment with the at property in the configuration file but I was not very successful +Seems reasonable Ill try to modify ant build script to generate proper pom files for both asl lgpl distributions +Also since we are going from a smaller to a larger value there shouldnt be any problems with existing values +from and something else outside our build is referring this plugin directly we want to see it fail now and not after the release and have dead codeplugin to maintain on updatesites etc +Tried with M and found no issues +OK looks done +Attached a repro for the issue +Dumb user disregard. +Junittest is failing unless alternative is activated in +Ported to trunk +I also renamed to a more descriptive and probably originally intended name +Virtualization is added as an optional section at the end of the guide +Patches are of course nice but not a necessity +So I think we just need to remove the and let source editor open the plan for user to edit +Thanks for reporting this bugIm going to put in double precision as that is the primary name float is an alias +Will the java api be supported? We want to use these api to do some automatic monitoringIf cant do it +This puppy has been sitting around for a while and still Ive not clue what to do about you still want to see something related to this fix please with some concrete and preferably and example project. +As an example with shuffle we dont need at all since we can just do network merges directly +Can we concentrate on the actual issue here without dragging in those extraneous things? Maven artifacts not those other things have been provided by Lucene since the release +cellpadding doesnt work becouse of style class that force px padding for all cells in data tabledata future it can be replaced with inline style padding attribute depending data table cellpadding attribute +I guess then I prefer to include the sources under and manage any updownstream changes manually +I committed this +In that setting Im not sure what number of decimals would end up meaning +The that was used to verify the bug was not on truck is incorrect as it doesnt make use of Queueremove +Maybe the thread leak is normal because the test used the coprocessorExec and there are regions for this table so it creates threads to get the result +I dont think the referenced forum post is relevant in this case +I think that the cause of the new Insufficient data exception that is appearing with has to do with a specific part of the change to the restart processingI changed the restart processing so that instead of just closing the Session objects on thethe restart code closes all the Session objects in the sessionTable +Marking as closed. +and thus it made life easier +We should probably change from to that as well +there could be a race issue w autocommitI wouldnt be surprised the tracker has never been thread safe code +I have committedAdding javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestssuitesSending javatestingorgapachederbyTestingfunctionTestssuitesAdding javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +OK I can see this is due to some quotes or funny chars in the rule names upsetting it +We dont currently support setting parameters or output settings on the transformer instance so I dont think that will be a problem +Uploading the branch patch +In general commands should require superuser privilegesWhen I worked on I added checks for superuser for all commands +It is so great to know this +It would be terrific if you could bisect further and reduce the number of revisions that introduced such dramatic performance regression + on changing their names to dfs +Spec implementation comitted rev Ill leave this open until its used in g and we are sure about the nameversion +How is the synchronization preventing the system to achieve better performance? Are there many other threads blocked by this synchronizationTheres also the fact that these classes are not in Ode itself they are in rd party libraries +You certainly could not wait until the service is invoked to have the session created +Closing per Bryants approval to do so. +is not sub packages and and as you ca see there are c items under the b node +Accidentally removed this patch. +Can you add a release note field for this issue with brief description of the new feature and a pointer to the docs that describe it +Not yet but Ill provide patches as soon as I get around to work on this +The attached patch attempts to solve the problem by adding the jars from builddistlib into the test classpath before the Ivy jars +I committed this patch. +Not found this issue recent runs +hithank you for the fixi want to test your modification +jiri as far as I can see from your screenshot we already disable the ok button +Closing as as were most probably not going to move using Maven but moving directly to in the future. +Im on Windows XP with default locale you try again at your end Kenney? If its not working for you please specify your default locale and OS. +using merge introduces a new use of the using +Well commit this soon +Yes we can override with a configurable setting add another value to the constructorWe are pretty close to releasing here and I think it may be better to waitI really dont like this constructor with values I think there must be a better way like propertiesEverytime I want to add a property I have to find all places where the constructor is used IF you really want to this change to go in send an email directly to the mailing list and call for a vote on itI can do it I just want to make sure I dont destabilize things and we are all busy trying to get the release out +I have now fixed this problem as well as a leak of an arraylist structure allocated as part of the axislibcurl structure +Can we close this issue? Looks like David answered the questions. +Fixed on trunk with provided patch. +Sounds reasonable but I guess it wont be implemented +Patch fixing the issue +Feel free to give it an early ve committed a revised version of this searching any level of cause +Attached is a quick and dirty patch for authorizationUsage Apply patch to sonarlib Restart the serverThe idea behind is to have a sort of with a set of Authenticator configured for each link in a chain until the default one is usedNo configuration available with the patchPS Please dont judge too bad because that was my first in your patch dont relates to LDAP plugin this issue moved to Sonar core +Sorry +But instead the followers have offsets +Thanks +Duplicate of. +set ROOHOMEcdrem echo Resolved ROOHOME ROOHOMEjava ROOHOMElibROOHOMEdist classpath +How about the result ofgrep APP to be JBAS one. +I see a reference to nestedPath in the comments +the older versions are coming from CXF so Ive opened +I think this can be greatly simplified to just removing the class static cacheThe har AFS delegates to the standard har FS which uses a cached fs +Thanks for your understanding. +It currently does not deal with synonyms as yet just wanted to get the base correct and equal to the +However while running in a container the parameters need the same workaround as above +This patch file IS related to are a number of issues with this that I am reworking the same ones highlighted in incorrect assumptions about the location of the POM didnt test m relocations at all exception handling in the relocation was not ideal applyRelocation didnt need to return artifact it modifies the passed in applied in to test this still works in as it was applied for +One remaining concern is that we should provide a compatible method so that code doesnt break when users upgrade from to +And there is also no documentation aboutIf there is a really interest to see this component within Cocoon then reopenthis report +This is fine with me though I have completed of the work for the existing schema +oops +How do thesetwo bugs fit +it should work fine with too right? So there ought to be a better place to put it +Specifically Im having quite a time with and I suspect well have issues with +Actually the test failed even in the hudson build of that patch +all bulk edited issues have in a comment +This is because the Stomp spec didnt require that you assign unique Ids to subscriptions so the correlation is made per connection using the matching clientId and subscription name +This is just one reason that action chaining is a bad idea +Is it possible for any of the Priveledged person to take a look at this patch +Thanks Earl. +I dont know if this was a conscious decision on our part or if we simply neglected to apply a patch from upstream +BTW since were trying to get out the door and changing node types is a bigger deal it is better to do this sooner than later and thus before the release +This issue is fixed and released as part of release +generics patch for smartcn +I think is enoughRegardsDave +I realized the rationale behind immutability of and when I thought more about it +on the new patch +Kev I tested this under jdk as and jdk as going through all the webserviceproxy quickstarts +Do you see the that full query text on the server side? If not and you are for example issuing through squirrel then the query tool is stripping the comment +The interface currently only exposes methods getLogger hasLogger and get +In I fixed the commit message but already pushed the changes without modifying the branch so it only shows up with the correct JIRA id in the master +Done +Closing xml and console issues +I think your issue is that the aggregate report is enabled by default which is a change compared to previous versions +sMax +Must be using the wrong any hints +The ref object is reclaimed because all its methods are inlined and it cannot be enqueued +OK we will see if someone comes with a better solution o +Yes I still have it on on Jonas +for a linear chain of tasks +Running tests will commit once done +Confirmed this problem no longer occurs. +It just the API is not good +Thanks for the review +It keeps you from making silly mistakes and banging your head against a wall due to a typo which is something I think the compiler should readily support +Thanks Ravi and Hong + +Very much so +new and tests code +No it is not fixed in +David wrote I think its confusing that when someone does try to use connectionAttributes to set the database name the error they get is Required property databaseName is not set +Thank you! This new view is very useful for my current project. +fixed +The counters of running job are obtained by locking up the job which also does not help us in terms of performance +See my examples above +I am currently testing latest patch if I dont run into any problems I am going to commit it in a few days +Each mapper calling dfs get timesWithout the fix namenode was showing getblocklocations per sec and threads blockedWith the fix getblocklocations per sec and almost no blocked threads +Thats why I scheduled it for for your help! is an initial patch to demonstrate the concept +Not going to make it into +Updated patch with tests +think we should answer the question that was asked not a myriad of potential questionsThe answer is that there is no harm in reading the spec or the license +I just committed a patch for it I will close the JIRA when we release twitterj bundle. +Then we could revisit the issue with selecting the working group of locking strategies +Lets focus on getting the patch backlog cleaned out and getting the release ready +One note I moved the test to package +An optimisation based on the fact that nodes contain refs to locks which may not be necessary in opt locking node subclasses dont have this so this refactoring may not be necessary will chk up on this +I hope therell be someone with time interest to tackle this +Note that I am using maven are you on the same version of maven +pushed +Is there anything I can improve so one of the developers commit this patch into the svnThanks in case one of the people with svn write access can commit this +Sorry for the delay Im not sure about this change +Fixed by +Any update of this +Thanks +useTpl relies on the javascript side to make the substitutions +Now you can find a war file with its source code +Thanks for clarifying +I agree the test should be renamed +Perhaps there needs to be a programmatic API as well reducing the need for people to read directoriesIn fact I worry we are building something to make a faster UI but making it impossible to use in any other manner +Olivier it is a good proposal +Could be a serious bug raising priority +BlockerMajor Fix Version move issues to and +I have the patch using svn +FWIW Id say stuff that converts text etc +I have a hope that once its in could be solved by moving the code that generates the from to +As noted in the init method will never be called by the Axis runtime +Committed revision +Is this jira for modifying or using it with META? The name doesnt match the description +Revisit Jira assignments +Need to approach differently. +I think the problem is more inside the static initializer +When using a factory method a default constructor should not be was a temporary problem caused by the reinstantation of classes for type conversion +Fixed it by trimming the header +This would be the preferred solution + revision . +Im going to do another pass that does deprecations and tries to preserve compatibility +That does hardly qualify as performance problem +Patch implementing proposed change. +They need to be as otherwise no user will understand what this means and how it solves an issue they have +Pointing to a commit made years ago doesnt imply that theres an issue today nor does it imply that other codepaths have an issue +can you please make changes to the ant task as wellthankst forget the documentation too +For reference the performance report has also been attached with thisMy sincere thanks to Anoop who was pairing up and helped in getting the design coding to be done +I didnt see how this could have worked in the past because the task never had a keep attribute as far as I could see +This is not a Xalan bug +Right +bug of a bulk update all resolutions changed to done please review history to original resolution type +This patch allows users to provide the algorithm implementation class using the algorithm attribute of the loadbalance element +I actually committed the fix outlined above in and I think this was the cause +Validate. +We might if Simone or Alessio added them in +I confirm it now works well in both cases +closing per reporter as the spirit of the issue has already been captured in +I dont doubt that I said this but Im not sure that I agree Many people do delete and just not removing the delete markers would be unexpected +Transmitting file dataCommitted revision . +batch transition resolvedfixed bugs to closedfixed +Should be fixed in SBS reopen this ticket if it happenLibor +I am not sure how this would affect the push downinterface i have not looked at it +True enough Eclipse shows it too as a dead code +Im hesitant to include the whole conversation as Im not sure how to obfuscate the info +This issue is related to the problem described by as in the given Example a rather long noun phrase Strache is not correctly linked because it does trigger several entity lookups with different parts of the investigation have shown that the using the POS tag to improve the selection of Tokens used to lookup Entities within the vocabulary is not feasible +I dont see a matching command for them in the current task servers handlers +This has changed a lot of assumptions in HDFS +Therefore the big synchronization block has been moved now into the loop of processed types and is based on the currently processed type +This patch takes care of the new features mentioned in the issue +Each of the part can be done independently without rendering trunk in a state +wroteal meu entendre la gota que fa vessar el got i la que faltava per al duro sn expressions amb significats diferents +We have to remember that the properties view always tries to show the properties of the selected element + +The problem is thatif I keep developing in a separate branch i will diverge very quickly and be never able to converge +on latest versionSome of the test failures are strange though +verifies under Adobe Acrobat as PDFA compliant but under pbdfbox preflight give the exceptionthat we are talking about +Seems to work so it will stay +Not sure if thats the right fix however +This new patch takes that into account +Any suggestions on how to solve this? Having a unit test tied to a specific timezone doesnt sound like a good idea +But unfortunatly not on forms using sorry +can you confirm that the change Ive suggested above is correct then update the ESB Programmers Guide with whatever the correct change is +symlinks are always created no matter what +sharewww +I appreciate your efforts at cleaning them upBuffing up the code to production level will also be on my listThanks for your improvements if you do any more work keep me posted +Basically the same root cause different the patches are very close tooI guess this issue can be safely closed now +Looks great +Wdyt? +Dont see the border one of the commits talked about but dont think its needed to have a border +Otherwise Ill create two subtasks and start working on the first + +Please use for applyingIn the updated patch I fixed another native finalizer and ref enqueue related bug there are no class loaders assigned to native finalizer and ref enqueue threads +That needs to be addressed +There may be potential for using info about region server load that is more realtimecritical than metrics channel in master for balancingetc +As for I could do the same thing +It seems we have everything commented out and ready to be set free One suggestion would be to add some documentation in the wiki or site showing how to use nutchgora with other stores if we dont have it already +Thanks Vinod +lgtm will commit to branch +Then the Holder object will become Serializable +When we have federated namespaces we will likely need a way to test whether a given directory structure exists on a single file system + +New patches +What was the reasoning behind and which substantially increased the probability of violating the discriminator column length? I assume this was the outcome of an EJB discussion in which case does the spec describe what should happen in such circumstancesObvious solutions that I can see to this problem are Increase the default discriminator column length to +What documentation did you have in mind +Updating add blackberry +has been marked as a duplicate of this bug +Committed revision +Good idea filed +If there is no new news here I like to close this one too +The whole thing needs to be as there shouldnt be any case when exceptions thrown are ok +I downloaded and installed it from sourceIncidently Comments on installing from source +Make the test better +Ill see if I find some time to investigate on the issue this but I cant be sure +It uses two tables one that is used to define a unique id for each instance and the other is used to determine who the master isThe tables are created by a single instance using a transaction that processes more than one sql statementsThen each instance generates a unique id by executing the following sql statements +I added code to catch the in revision + +bat files +Closed upon release of Hadoop. +Please ask questions like this on the user list firstAt the end of the day whether you use jsp directly or tiles what matters is what html is generated +Hi NestorIm still waiting for your reply to comment Are you still interested in seeing this patch applied + +Im ok with the priority but we have already released it and tagged +Heres my take personally i dont see any problem with how Shai works here as long as its still a proper mergeI dont think we should allow limitations of our current SCMS to dictate how we develop +Yes I have undone changes for skipping the timestamp printing +Can anyone spent some time reviewing this so that I can commit it +So I wonder if itd be worth it to port back as well? I dont feel strongly about it I just figured Id mention it in case you were interested +i think you can setup your to point the resource back to srcmainjava with files in srcmainjava are included in the final jar without the need for any new mojo +Ah good I missed that I need to take a closer look at thisNp let me know what you think +But if a job producing temporary output fails the output status need to be added otherwise the job status check wont find the failed job +This one looks to now be complete for? +the K buffer should handle this indeed +It is used as a default value for the staging directory tree +Youll need to convert all to amp otherwise it wont be valid XML. +This issue is resolved in and can be closed now +This has been reported and fixed already an unfortunate side effect of Ms new merged bean definition caching for the case of parentchild bean definitions with placeholders in fix should be available in the latest nightly M snapshot already in the worst case it will be in tonights Juergen I appreciate you quick response. +to change the fix for version field. +Do you have a requirement to use Maven? We could consider moving up to Maven on trunk but doing so on a service branch may introduce some unexpected build dependencies +Committed to master. +I think we have something like that for the that can change frequently +One thing to note is that the warning is only a warning +Thanks Eric +SGTM the goal is to run with the JVM minimum stack size how about attempting to query it directly? Going to k on ARM systems with Java is way too muchOn amd with u java XsskThe stack size specified is too small Specify at least kError Could not create the Java Virtual MachineError A fatal exception has occurred +Thank you +the listenaddress and rpcaddress are still +While I like the feature there must be a better way than to bring dependency to openjpa kerne +Heres a revised version of the original patch with the proposed change to included +And what I get is empty values for the function names +for previous version of hadoop +Works for me in trunk +I thought the Resteasy project was going to own the API codejar? This is important IMO +routers with options each vs routers with or whatever +HelloThe Karaf build is currently going on and will post the details once is ready for release as well Ive updated the components file to refer to the released versions +I could now reproduce on +The problem is caused by the fact that many Jackrabbit components cache and reuse +java and all can have EL too where they can be relevant in code completing expressions that refer to other resources +Jrn I tried it and now I see what you meanThe constructor needs the serializers loadedWe need another way to do itWhat if we make the artifact serializers map static +It is simply sending a fault message using deliverAsync +committed to TRUNK +I just committed this +Apr connector has unlockAcceptmethod that should break the accept call bymaking a dummy request to +Better yet rather than using static use an unnamed namespace +old chukwa files have byte counts +If not then we can go the rd party route +Thanks +New patch corresponds to latest patch for +right? Note ESB Console installation instructions in the Monitoring and Management chapter of the Administration Guide has been updated with details of new options specific to Oracle implementations +Updated the patch with unit test +Amin Tests are failing with your patch changeDo you want to take a lookAlso please incorporate the comments given for this patch in your next version of the patchThanks Nicholas for taking a the patch as the comments need to be addressed. +After that AM just need simply check the exit code to find out whether there is any auxService failAttached is the sample code for this is adding new events to so this patch may have impact +It will try a new map task only if the earlier one succeedsfails +Looks good +it doesnt affect production +issues +This should be sucked into the next or RC or whatever +Or in the case of source file the sourceses +I spent an hour trying to reproduce this in both CVS and HB and was unable to. +Closing per reporters comment. +Sure it could be resolved since all the ejb related to same IOR could be started correctly now +verified on JBT Beta H +Thanks +The list has not yet been updated as it hasnt been migrated yet +This test is useful though flakyWe can detect the hadoop version in test util and return early from this test if the version is beyond it is fine with me +Hi Rajeshcould you help with assigning this jira please? Thank you EAP is still not ready for testing with Ipv +Considering this done for now +Hi Petefinally here is the ve tested it with the head branch ll still need it to get Seam applications working on regards Peteheres a bundle of patches needed to get the jee booking example build for my change descriptionsbuild Added dependency to which is needed by weblogic server to run Same as for jeebooking Added support for jeebookingresources add to solve a problem with versions Changed configuration for weblogicexamplesjeebookingresources Added missing listener these patches help you to build an officially working example regards a forum reference with more information on the BEA patch and what is needed +Andy got logging working with the new Plexus IRC stuff +Ive just committed this. +Even if its a little weird I think we should go with indexing threads by default instead of +ping with being cut I am still curious on this last question +This violates immutability but the reseed methods currently provided arguably do this already with essentially the same practical consequencesWould it not be better to allow passing in a instance via the constructors? That way instead of constructing a default with default random generator the user can provide their own which they constructed using their own random generator +My first bug +OK for the last the packages are wrong +Guice is now in place feel free to contribute a patch for this +Primary key index helps to create primary key lookup +Using +Bulk moved to +Username vincentaumont +Fixed on master Edison can you please check if the fix is already or if you want to assign a default value to the imagedatastore id column during upgradeAlso as this will make into I move it in file +The region mentioned in the desc was open on server A and was moving to B +It could be due to how the test is formed but I dont perceive this as being a blocker for runtimeSo at the moment is required to build but Shiro will work in environments againIm moving this issue to to address after the release as this no longer constitutes a blocker +bulk move of all unresolved issues from to +Made the fields package private in r. +Not sure on best approach i was thinking of just documenting lots of ways for now and seeing which of those seem best +But I am you have time to work on this this week? Otherwise lets punt to out for now +There is no need to grab the read lock +It could be replaced but it is there since EAP +Thanks Sean! +Have you install with sudo? Im beginning to think it must be some problem with my local vm +Simply closing this issue with the lack of patches as a reason doesnt seem sensible to me +Updated the patch +I didnt see interests by lucene to work in a OSGI environment so the binary release is not OSGI compliant +Postponing this issue +This may have side effects because the Authenticator will be used for all in the JVMWe could implement the second workaround in but Im a bit reluctant because of the side effectBest solution Dont protect the WSDL +Ok Ill rebuild +Will file a bug for +Let me know how you get on +Adds back get methodIncidentally fixed trailing whitespace in both files since my vim setup does that automatically +Still open to feedback on the things we were discussing +And it looks like an issue because instead of getting declared as constructor in both the above examples Foo gets defined as an enum constant +Tony you are good at keeping us on our toes The content is in a good enough state that we think we can move forward however we found we have a few things we need to fix so we are more inline with the Apache website guidelines +Thanks KevinPatch applied to repo at r with minor changesplease verify it was applied as you expected +fixedThanks for reporting +Closing all resolved tickets from or older +Fixed git clone issue in cartridge agent and added. +Otherwise entities in the child component remain passivated + +i think i fixed this problem earlier today +This amounts to the previously discovered super hard cast of weening our test cases off the old filters I started doing some work to handle this for the libraryxml modules + +The one failing now was not failing before +Applied Nandanas patchFixed in revision +Then if you want to communicate with an older server then you would instansiate with that and the code would have support for whatever format changes for the various versions +Patch to ran with a patch on Linux and Windows and both worked as they should +Will commit if tests to trunk +Please find the list of missing pictures attached +the patch this time it grant the license to the patch with grant license +Commited at revision at last in +This one doesnt apply to the stream +The Java install on newer Debian systems seems to die in the install script at a call to to set the default Java JDK +if its its included in a span it will think the span starts ends at without payload it looks and with payloads or something to that effectReally anything that counts on the position of the term is going to be screwed I think +To enable it recompile everything with RWSTDUSESTRINGATOMICOPS defined. +release should be in final block otherwise it would cause leaking +it would be better to at least have the +Thank you Peter! Your suggestion to use pluginB rather than seems to work for meHow about creating a feature request Declare plugins inThis would simplify the usage of plugins +The storing the result back in the metastore can be tackled after that +Good catch Ill look at that one tomorrow +Committed to SVN +to be included in a map reduce job +Reopening because I closed the wrong ticket +Patches were applied and passed the suitePlease and close if possible +Can you test v +I committed this to since that introduced this bug got committed to. +This something I missed when I ported to trunk branchesIll commit this if an reviews this or if first thing Monday unless there are any concerns +trivial fix attached fixes the test case for me +I believe this patch could also incorporate by creating new logins +Its targeted forIf no hflush is called dfs client pushes data to datanodes when a packet is filled up +See previous comment +The copyright statement in the pdf file is now identical to the one generated by the xdoc plugin +Attaching which bumps the copyright year and release id used by the docs build +Plz do not close bugs before there verification by QA +I checked also its all right. +Documentation change was committed closing the issue as jvm properties are now documented. +If someone spends the time to write a similar function against Windows I dont have any problem with checking it in +Hi Derekthe implementation looks fine from my side and your modifications met what I had in mindHowever Ive recognized that most of the files have a copyright notice +In an I used to simply merge s using an XSLT +Attaching a release note for this issue +Later on saveNamespace save the namespace again +Closing resolved issues. +didnt finish once +Try Hudson +Add my modifications for current phase anyway +No I Havent and indeed it was the problem +Ive committed this +Progress bar works ok on demo site +Even the samples of batik are withoutthe unitsIs there a workarround how th get FOP to resize such images +I have fixed the circular redirect by changing some setting in the and will upload a patch shortly +could you your commits from to so they are in sync? Thanks +as well +The nativeascii tool can be used to convert property files to and from other character encodings +Still fails in case fails in because the job queue is not bound +Dimitris Im assigning this to you just to you take a look on what to do next +All is well after commit Thanks Anita. +Moved version declaration into in thanks +I would also be interested to see how this interacts with which can have wait times in the range +Nothing major and nothing that I am very tied to so if you want to push back on them I would be fine checking the code in as is +It is thus more awkward to efficiently access the files of a cluster over FTP +Sorry Stefan Seelmann I have missed to import the template files to new Workspace +Attached patch to create non existing directories during the initial cwd +Linking the issues for future reference +Committed to trunk in r and rolled back in r +How can we use during the karafassembly build the features xml file generated by this karaffeaturesassembly module +Commit +Benedikt Yes will upload the patch +Wild guessCan url whitelist be the culprit +But my manually tests today failsSo the last patch is just for the trash +The attached patch activates the stopmaster +That should work +for this issueTajo can support various types of storages by abstracting the storage layer. +Please if required. +Previous commit bbdaeaacece is working fineFixing this again due to some merges after this commit +Perhaps distill out the good stuff into new more pointed issues +Looks good Daniel +The section on the notifier is duplicated on pages and +Note ellipsis on To Recipients field. +This issue has been pushed back +Works fine in JBT JBDS there are references to mouse cursor and mouse buttons in the summary and in the description but nothing about mouse in the rd step +Maybe its just the limitation of +Now that work is well underway with AS all previous community releases are +closed per above comments +committed on and trunk +The tests have been fixed a lot of time ago. +I wonder whether there are some requirements to use the Java serialization system +in the flux of package upgrades there was a conflict in my maven with antlr +Note that it took hours to recreate +Can we fail a message to the regionserver? or are we going to have to accept SPLIT messages when META is offline then queue them until META comes up +Perhaps this is a similar issue to +FTR is an experimental remotely injected persistence unit +Ill disable the old acct in one day from now or sooner if you give the earlier +class public static final String PROPSCHEDTHREADNAME public static final String PROPSCHEDSKIPUPDATECHECK public static final String PROPSCHEDBATCHTIMEWINDOW public static final String PROPSCHEDMAXBATCHSIZE public static final String PROPSCHEDJMXEXPORT public static final String PROPSCHEDJMXOBJECTNAME public static final String PROPSCHEDJMXPROXY public static final String PROPSCHEDJMXPROXYCLASS public static final String PROPSCHEDRMIEXPORT public static final String PROPSCHEDRMIPROXY public static final String PROPSCHEDRMIHOST public static final String PROPSCHEDRMIPORT public static final String PROPSCHEDRMISERVERPORT public static final String PROPSCHEDRMICREATEREGISTRY public static final String PROPSCHEDRMIBINDNAME public static final String PROPSCHEDWRAPJOBIUSERTX public static final String PROPSCHEDUSERTXURL public static final String PROPSCHEDIDLEWAITTIME public static final String PROPSCHEDDBFAILURERETRYINTERVAL public static final String PROPSCHEDMAKESCHEDULERTHREADDAEMON public static final String PROPSCHEDSCHEDULERTHREADSINHERITCONTEXTCLASSLOADEROFINITIALIZINGTHREAD public static final String PROPSCHEDCLASSLOADHELPERCLASS public static final String PROPSCHEDJOBFACTORYCLASS public static final String PROPSCHEDJOBFACTORYPREFIX public static final String PROPSCHEDINTERRUPTJOBSONSHUTDOWN public static final String PROPSCHEDINTERRUPTJOBSONSHUTDOWNWITHWAIT public static final String PROPSCHEDCONTEXTPREFIX public static final String PROPTHREADPOOLPREFIX public static final String PROPTHREADPOOLCLASS public static final String PROPJOBSTOREPREFIX public static final String PROPJOBSTORELOCKHANDLERPREFIX PROPJOBSTOREPREFIX +Should I close the bug or is that something left for you +Ive amended your patch to permit this and augmented the test you added +When I remove the dependency on and I cant even build my project +MANY apologies for wasting your time on this one my can just close this one out +Fixing the uDig side of the fence to produce specific wrapper for specific filter interfaces ie +In some cases the is still needed for example the tck +committed thanks +A connection is established to the master broker +As this was a dependency being analyzed there is no source present just the class files that had been copied to the targetclasses directory and should have been exluded from cobertura analysis +If something doesnt work the user should get a nice error that makes it clear what the problem is not something that implies is broken and ends up wasting everyones time when it gets reported to JIRAOne possible solution might be to use to count the length of the submitted form +So this would actually mean that after your patch the behaviour would be the one you mentioned in your comment of Jan AM +Committed to trunk at revision and ported to branch at revision +Please do not attach a patch without signing the CLA +bq +forgot to clean that up sorry about that +This diff just adds some more sanity manager logging to the spots where if the retrys fail we will lose space until a full compress table is done +This has been corrected and will be in the next build + +Closing as duplicate of +So there is thanks +do a close flush in its own thread and not wait +I also tried importing an existing Grails project from a + +Personally Id make it a requirement to use the new with JDK +Added my patch for code review before I commit +If there is no problem now we could see which seems to be another issue for cache evacuate +Lets add a new test for this issue +It still does a delegation rather than inheritance tho but the code has been simplified and a bug eliminated +the patch looks good +Applied Andrews patch svn rev . +Thanks Knut I assigned my self to this issue +in fact the project is created as a maven project +I also added a subview to the shale usecases validator exampleI appreciate the help Jack. +I just committed this thanks Jerome. + +Nice job Aaron De Morgan MyersWhat was the impact of this if any +Changed to code to check for that +Maybe what we need to do is log a related JIRA so that a developer can add the diagnostic VTI javadoc to the public api +Ouch! Good catchFixed in r +They become unreadable instead +Please close if satisfied thanks. +Please see another screen +The idea is to provide events at certain points of query execution and then support various listeners +I would imagine others with a similar use case +However no decent build has reached the updates site yet due to failing CI jobs. +all issues not touched in a month to update to work with Solr Will make comments about some of the issues +I dont have neither hang nor assertion failure +Ah I see what youre saying now that I actually look at the Title +It certainly shouldnt error +The patch doesnt work if you have TSUSEHWLOC defined +Cant reproduce under this bug is reproducible under Linux too +I have renamed the file to be more user friendly especially to readers + +Issue filed due to confusion encountered by someone I corresponded with +just committed this +Thanks for the comments Neha +Is there any chance we could get this patch on the branch +just committed this +Fixed thanks +I have to disagree on this generalized statement +Shouldnt be too hard to fix but I dont think the fix will make it into the release +It sounds like youre not using the latest script +Can you please file the doc jira Mike +Ive gone ahead and added this mime info in r hopefully thatll help with detecting such Nick +I am based on the beta tag and have the dec patch and have appended the detailed test output into the attachment filethanksNow that release is out we can start working on getting your patch into the official code line +Start small and iterate. +However I share the concern that this encourages users to operate in a way that creates a single point of failure. +Flavio thanks for reviewing and modifying the typos +HelloThe patch looks good and has a test but why did you use regular expressions? Isnt it possible using a simple? revision . +Heres the patch +eg +Ive asked Len because it seems he is currently the only one who is able to reproduce it +It expects that a type system file is placed in a certainlocation if not a dialog asks the user to provide the location to a type system fileThe styling information is written into a styling file for the type system and a viewis used to edit the styling +TYPO I slept on it and now I wonder if we should not actually change the calls to the deprecated methods but keep them to avoid the issue David reported +It would be really nice to get the same answer +Please can you attach a patch for branch +Unified Patch +You will need to install cvsspam or similar +Can you try using thrift gen csharp if and see how that goes +It is on my todo list to address this soon +Patch that contains the first reverted changes. +The details are already on the JIRA +But in my project I do need to use EJB tx because the ejb client needs to be distributed to other divisions that are not ready to adopt Spring +The attached file contains a patch for the +Thanks Brock +Like Mubarak already mentioned there are start stop and status scripts in the packages provided by. +Complete patch for +Updated the latest patch to include a check for the temp file before calling +oh and the other funny thing is that with this patch applied the oauth and cookieauth tests break because they try to upload a secret that is invalidP +sudo yum install +I have created a small project that is supposed to reproduce this issue but it doesnt +There is a manual workaround explained in and +I tried the same with with the same me it makes sense to use another projection easy to use with UDIG and this public server only delivers EPSG I use +IMPORTANT To apply this patch correctly you must first do svn mv srccouchdb srccouchdbAfter renaming the file the patch should apply cleanlyAfter some discussions on IRC about the incompatibilities of Xapian and ASF Ill be opening a google code project for the indexer scripts used +patch fixes the above described problem by checking the return value of the original button onclick handlerand only calling the ajaxSubmit function if it is not false +But I could not simply change it to produces errors. +My test cases work with the patch but the passed onclick handler may already contain a return statement +launches does not exist +Such duplicate messages cause what you have observed but they are supposed to be harmless in the sense thatI do not expect them to lead to any inconsistency in the service +Returns only those binding services which have the same key concepts in the query with tModel instances binding services in +Assume resolved and leave for Dan to review if he wants to. +It is important that we dont do that +Then I cut things out till it stopped breaking and that was around when I pulled out to the Inject on the home controller totally get the point about a lean install of Social into Roo +binbbb as well +Code look good +Sorry +Tests passed +Attaching +The initialization of the endpoint servlet is just the transport part of that +All tests that were functional at the time of this patch still work and a test is included with the patch +How to commit renamed files? Just also commiting it here and yes sure if you want I can do sharing for this mojos as well and also I see can use then URL as additional need to create a patch +Same issue here +m configuration file in your home if you run from command promptbut another may live at another place for the user in charge of the serviceIf your allowed to do so I would try a full search on drive to see and check file modification time +i can never trust myself +overlay icons for Database node now possible states connected was a try to connect something going wrong not conected no overlay icon +We could move the and the Schema instances but let me first check whether theres any dependencies we dont know yet +This solves the closing core too many times issue you can have by just simply checking if a core is closed before closing it in shutdownBy the way why not allow removing the DEFAULTCORE? Shouldnt it be treated like any other core other than the special property of getting a alias? edit We still need to solve listing alias as primary cores in as well +if we can do this differently I think we could remove +This patch subsumes which should be closed if this patch is committed +Greetings + +Sounds good +sounds like it would never get set +updates are atomic so a crash doesnt result in duplicates deletes are applied on every new segment flush so they dont build up too much in memory applying deletes doesnt force a writer closereader openclose writer open +ATM I have no idea how we can handle these situations +Wed better do it sooner than later or not do it at all +Mv mimics the unix shell +The consumer was throwing an exception which I didnt see in the logs so the message was rejected +Its also strange to see some leaf having no childrenSorry I dont catch what you mean hereThat would also accelarate the navigation to a low level leaf as no intermediate page will have to be loaded +I pulled the newest changes and now the old way are right the new way looks much more intuitive and I changed my scripts +There is a lot of stuff in the patch that does not obviously belong here +This seems like a good change +michael yes the stemsfor I am scratching my head trying to figure out what they were originally intended to do +I put more thought into this but couldnt figure out the need for check against legacy splits that is failingCould change the legacySplits implementation but then the test doesnt really make any senseNot finding an alternative removed the check against legacy splits uploading the corresponding patch +OK I commit soon +This on top of patch which is not committed yet +bq +good point Evgeny +We can do this by having the get the non compacting sstables from the and work with those to generate the buckets +I logged for the missing test in Thanks for the reminder Enis +I wonder if instead we should add a console action to download all the schemas and dtds from sun +Need to check that the tests have been applied +Its a if there is no payload +I think the unit test should be in common as well +With the this is now easily possible via the use of sequentialparallel +Hi RaymondThanks much for the patchIf I am correct the attached diif file is from git +Lets try a run with this +Always logging to the category from the concrete subclass follows that for the time being +No time to fix this in not the content of a mail sent by AlexWas wondering about revert operation do we properly revert so operational attributes are reverted +I cant reproduce it and unit tests exist for this case +The source code remains the same as the unchanged one +Arvind for adding Nemon to the list +Also for the common +there were some bugs in issues with the old patch +ThanksChatra +The Groovy NIO module has a comprehensive set of Path methods +Im marking it resolved since theres another bug for the backticksystemexec shell redirect stuff +Looks like the code you commented is not really related to this change +Patch with Javadocs improvements and trunk revision +Sounds good to me +I looked at this one and found some pretty strange results that make me think it is not a bug but a problem with the server itself +I will implement connection pool for tajo rpc +I think not with the API which is normally used for embedded server scenarios but perhaps for the command line start up where we also start a security manager and try to be more secure by default +The vtest could be changed to work like to use to start individual composites but theres also work on going to use monitors and trapping exceptions so a validation error doesnt stop the domain startup so maybe thats the better separated the mixed test out from the other tests an enabled it +If you find further issues please create a new jira and link it to this one. +If it detects any new installations uninstallations or updations it tries to start all the managed bundle unless it has been configured to only install bundles +I agree that caching may not be best approach I just felt it important to log it as I was backing it out so if anyone wanted to look closer they could at least know what had been tried and not to make the same mistake as existed before +I have read all of the documentation and was not clear what the default is supposed to be +Done +Thanks Ye for the contribution +Im wondering if we have document somewhere to address this monitoring attributes difference betweenJetty connector and Tomcat connector +Since no one else is following it lets close it. +pushed +It worked for me +Im on removing the permalink displayIts handy because the end URL is important and I want to see what it looks like before and after I hit the publish button +I need some way to report on the exception it just seems to go away without any logging +Ill look into +Committed Jukkas patch with modifications as explained in revision +I dedicated this week to solving your problems So my business hours are yours +It would be great to have this issue fixed +It should really be community and productThe files which currently determine the installation are bin and modules although we do need to introduce a simpler way for customers to identify what is installedThe current expected combinations are as follows although we expect to add at least two more for BRMSBPMSPlatformsEAPsloteapEAP SOAslotsoalayerssoaEAP EDSsloteaplayersedsEAP BRMSslotbrmslayersbrmsEAP PortalsloteapEAP BPMSslotbpmslayersbpmsEAP SOA EDSslotsoalayerssoaedsEAP SOA BRMSslotsoalayerssoabrmsEAP EDS BRMSslotbrmslayersbrmsedsEAP SOA EDS BRMSslotsoalayerssoabrmsedsEAP SOA BPMSslotsoalayerssoabpms +Also Yay for breaking changes after codefeature freeze! sigh +Apply applied patch +corrected in beta line +Then lets close the issue +The Turbine jars are still available in the downloads section of the Jakarta site. +Thx for the patch +Marking as minor this is so avoidable please use a namespace instead +How does one make sure that in all of the hadoop stacks components no two files are named the same if they are part of say bin sbin etcThis is the responsibility of the community to ensure that there is no conflict in the file name +See + +Nice catch +can someone say if this is still an issue with jbpm or and if so if we want to support a construction like open bugs older than a year as Out of Date +The resultwriter method delegates to the scheme just like the regular struct writer methods +metadata +I like the idea +Patch containing a library for allowing writes of Reflect data to files +Could you check it please? If you still can not reproduce it then Ill keep diggingYes when there is any output then double input works fine +Though Id hope jboss would give me a relevant error thanproduce bad sql or is that not its place? Tnx for pointing me in the right dierction. +patch committed +Carl I ran and in eclipse debug mode with my patch and both work fine +The patch looks good +The information on apache and tomcat is quite nice indeed but IMHO its not much about the Cocoon installationI will mark it as invalid because only slightly related to Cocoon installation issuesDid you send this info to the Tomcat guysAnyway thank you for the effort of notifying usIll see where to place this link +I have proposed the RN text +Planning to cut a release +Its been a while Kan do you still need this change +Is there a workaround for this problemI have a service that works on version but does not work on version anymore. +How can I change issue status from Awaiting Test Case to something else? The code submitted in pull request includes a test case. +The idea is to provide a standard serialization format so that clients in different languages can easily parse the query results +What would be better would be to explicitly include the source files we want +Just simply adding the property wouldnt break the compatibility in my opinion +All Axis versions using including trunk can be affected +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content as rejected as as of today there seems to be nothing to fix +Investigated build log and files on not determine why the zip has strange contents +docx support for anchors and bookmarks added in r +Thanks for pointing this one out +Maybe we can do this like this elsewhere too +btw this is not win specific just that on windows people tend to create directories with a space in them and on unix they dont +The first is why do you differentiate between a job failing because it hits its and a job which passed but has X number of skipped records? The second is if the number is arbitrary why force the at allWhile I cant speak for Dave my answer would be because most people want to know when the data can no longer be considered good +Marko has an example of this to review +I couldnt reproduce it on my dev box +Please add all comments to this issue. +this is the expected resultUse an explicit left outer join +The flickering is due to the current notification concept +That ZK log entry is ok +If you wanna give it a whirl be my guest +The max evaluations in the would not be used in finite differencesHere is what I intend to do replace Nabla by Math adapt Nabla to the change above keep Nabla as is adapt to have an additional max evaluation parameter in the Solvers spirit +The question I was asking is can this happen? I dont fully understand how the fetching restarts in the leader change so I am not sure +Thanks for your patch Ahmet I also added the files so the UI wouldnt complainCommitted revision Committed revision +This issue states its only about multicast and file consumerHave you check the components such as as well +It works +I dont think we are ready for that +It is so simple so if I respin it can be inI didnt want to have commits in because if I respin a release I would not be able to only take some of the fixes into +Fixed the stopping of the modeshape garbage collection cron thread when the engine shuts into the master branch. +These are for people checking the sources from svn or using a source release +Thanks Brock! +Thanks TimPatch applied as expected +Should we not change them all +Good call to trunkThanks! +All have been addressed +I think contains is better for complicated fields but the results can be far more confusing an unintuitiveThanks +byType causes the problem +The responsibility to keep track of all active elements may not belong to the group but perhaps a separate actortask registry +No improvement with intrinsic locking according to the docs it is enabled in JDK anyway was just making sure +Im sure James Page would also be interested +after such first failure no more updates until cache cleaning +to commit +Thanks Vandana +Can you perhaps include the steps you used to reproduce the error andor attach your sample app that demonstrates the error +One more minor issue fixed in euteut minor error logging issue on Windows in euteut +Committed to trunk but not to branch I think it was a bit invasive. +Is there a reason you chose K in particular? I see no real problem with it it just seems arbitrary +Applied patch provided by Antonio Andrade +The Programmers Guide already containsDeclaration of the contract schemas will automatically enable the exposure of the ESB service through a webservice endpoint the contract for which can be located through the contract web application +I havent tried to actually produce this bug yet but it may be reasonably important since this is also a synchronized method and a full block copy could take several seconds at least during which the DN cant do a whole lot +These are now part of openejb +I am open to add something along these lines +I corrected the typos changed the direct attached language to shared network +Sounds good to me +My first guess is the cluster code updates are not synchronized so there are incomplete memory writes +Very cool! It would be nice to have a unit test for this case +i do not understand the problem to produce a unique to is doesnt use MD then there is a eh problemWe get wrong cached query result from time to time it comes normal after timeoutThe problem occurs pretty rare so I think it was caused by the cacheKey collisionDont know how to debug it +In the old days there was a bias towards not committingI have to give uwe a bit of a backup for this +I think there are two ways here +can you confirm it is still broken in +I thought I reported this before +bq +Need a good use case for replicating the problem and a better solution for keeping the GEP and the server in sync +Thanks for tracking it down! That seems fine to me +Can you help with that +Can you please attach all required logs in this is test cases failing from regression suite making this as blocker +r removed erroneous code that propagated a function guard into the type constraintFixes +Logged In YES useridFixed in and cvs with at least the JRMP invoker +Thanks Tim. +Sure Will do it +I am also using IE emulated mode so something to keep in mind there tooMy test was as follows Use IE w developer tools Log into portal Switch to IE browser modeBottom line jQuery isnt there when the page loads +Ok or notAssigning to Carlo +Im hoping that fixes the problem because I cant reproduce it locally +This seems like a revolutionary build change before the releasebq +Closing all resolvedfixed issues of already released versions of Roller. +Should be fixed in latest CVSthanksdims +Please add this to will be to late for us and we do not want to rely on a nightly build +No Would it affect if it was open +had missed one local commit +This will only work though if the additional dependencies from Kay Kays latest patch are added tooShould hopefully be okay because Thrift is released under the Apache License as well +Youre right however this is something that needs to be supported at the driver level +Committed +Ive just tested this in a development environment so I would love to hear any comments or suggestions +Thanks for pointing this out +Looks OK to me +cancelling patch for just hit this bug +The patch works for too +Hi Douglasdid you manage to find a work around? Im facing the same issue +if you have enough greens w this patch enabled to give you confidence this is ok now +Changes made to trunk have been ported to +one line fix pushed to master +Hi Myrna thanks for catching this +Im happy to try the build snapshot if someone can get it into the maven snapshot repository +Not quite write once run anywhere +Thanks for investigating this so much well get a look in please confirm if this is in AS +The will have similar concepts as +hehe +Thanks Jan I think we are on the way to fix it now +Eclipse Guvnor Tools Reference Guide +So we can close this bug for now if new problems occur please open a new one +I believe it wont introduce regressions +Cant reproduce with last meclipse and given myfaces svn + +Any objections to that +The only thing that is left is that we need documentation of the new arguments and their meanings in srcdocmanual +One more note +This broke when was integrated +That looks very good thanks for fixing this. +I was able to reproduce the bug on version through the +Ok here is the patch for the tests +It is actually that I found it against +FreemanWhich bundle? Isnt it shaded in this wrapped drools bundleBenjamin +And if theyre set up correctly the plugins will declare their dependence on the plugins which means we can safely remove from the Central connector and just let p resolve the dependent plugins instead +I would be in favor of removing the options +Thanks Ashish will update the page with for Glassfish when the information is available +Lets do it +The javadoc of the api says that the history file may or may not be present depending on the job state. +Do we have a similar issue in trunk? for this patch going to and +Laura please update Release Notes +started writing code +So there should be no longer incompatibility issues when using with different versions of +Ah I see you have committed part of the patch again +the whole AJ works +Yes its fine you Kabir. +Probably due to manifest change in +After I saw Ted create the first sub task I created all the rest +i understand that +Hi Mateusz RenIm almost done with the work on I still need some polishing and some checks on the build system and need to review and to see if it is possible to include themI plan to cut a RC as soon as possible will upload it so a nexus staging repository and will then share a link with you so that you can download it and run your tests +it works with recent glassfish nothing we can do about it +I think well need to revert and when this is committed to +Thanks for the patience. +Im pretty sure this is fixed but am targeting for the next release until thats confirmed. +If I do so could someone commit to review and apply it? Also getting a new release out would be necessary +Fundamentally disablegossip disables gossip +Sorry +committed r +Thanks. +If not is it a good idea to move this to CDH related jiras +Then will someone to fix it? +Lets reopen it forAmichai thanks for getting your attention on this issue +It appears that always returnes the correct page state safely from the page mapWith this jira issue I was hoping to address the double submit which I know Wicket handles correctly in other cases but I dont know about the details +Check to see if this has been resolved +I checked the build and it hasnt run since June which is strangethe job was pinned to hadoop which now seems to be gone +This issue can be closed now +Ive renamed the issue as per your suggestionAs Kay Kay has done a wonderful job on I think its best that we wait until that is resolved or at least until Kay Kays latest comment is answered +We are planning a sprint for an upcoming milestone release +Just to add to the above description I get the error when I try to access the WSDL using httplocalhostservices? looks like files werent commited correctly +Closed all tickets +Same for files +patch adds XMP specificaiton part one namespaces and! Patch committed in revision +This will not result in a removal from the database +Verified in ER pwd +Douglas you cant experience this issue since it has testcase that was added to wicket testsuite and passes correctly +One more question is there a way the PPMC can be notified when a ticket is created and marked? +Both options make sense from my POV +Having the Done button bring you back to index page of plugin installer instead of viewplugindownload page +the continuum options dialog exception issue was fix by ehartmann lately. +Is there an ETA for getting this issue resolved? Thanks +Thanks very much KnutCommitted patch to documentation trunk at revision +Hi Mike It dont think that has been checked in yet +Attached is a patch that fixes the issue for me +Please close if satisfied +Cheers Mike +Does this sound reasonable? Shall we annotate as privateI think these are intended to be public no? is deprecated in favour ofIve also added licenses to the new files +In addition to Jodys comment my own feeling is that an interface for is not warranted at the moment +thanks to Clement! Still everything can be improvedRegarding Provides yes you are correct you must be explicit to say what if any services a component provides +Peculiar +Ran out of time to get this in. +But I see that there hasnt been a successful build yet so the jar you are retrieving does not contain these methods yet +Tx +Vladimir could you please provide with more details on this issue? How do you run the tests +See commit ebaceeaecdabbdecfcefbThanks Dinesh! +Agree with Luke that putting service level in a fixed RPC head can shorten the latency for service level info being handled +Alex I cant add you as the Assignee on +Incorporated all of Konstantins comments except and I like the serialization code in especially because it writes the serialized information to the transaction log +this is really a defect against either EMF or the Sun for those platforms. +Please make sure it is done by August +How come I feel a certain desire to see a patch from yourself +Issue reopened to correct Release Notes field. +It can definitely be a public attribute but it is provisionally setup in the init using the object id to follow the same conventions as other Spring like a plan in terms of continuing ahead and concentrating on the testing for confidence in our refactorings +Will commit a patch latercommit a patch? or upload a patch? I still think we should wait for official Jflex release unless a jflex developer comments on this issue otherwise +Thank you Brad for identifying this oneliner fix +It was due to the names of the docs changing between releases +Yes that would be good +If I patch out the changes all tests pass +First cut at a patch +One day all bug reports will be like this one I only made one change to your patch which was to break out of the loop as soon as we know we have been allocated an objectAgain many thanks. +Hi this issues is related to maven and the pluginremove it from your pom and the problem will go away +Thanks Tim the test patch is ok. +Attached a dump taken by the report on another dump file +Patch to fix the default constructor of per Scotts suggestionGiven the nature of the langjavaavro its to add a specific data test +It might be better to keep a map of renderingspecifications to rendelter registrations so that nevers unregister renderlets register by other means with the current version if a bundle tries to override a the overriding service disappears when the bundle provideing the ssp it wants to override gets restarted +Another possibility is adding cleanup or setting the properties in the beginning of each test or waiting the socket availability +Meant warning severity +Checking in the changes. +Hello first thank you for your comments +Code has been present in the trunk for sometime +This patch does not influence thatThe patch is specific for the g fileThis patch fixes the bug as described +How about this? Let me know what you think +Apologies for not checking this earlierAlso I believe the last paragraph of contribclustering does not hold any more as all are now distributed with Solr +fixed +Ifwhen we need we can simply make it support a list of files +Ill add one more item to this change tomorrow +Closing the bug as I couldnt reproduce it on the latest build +So I put some bubbles that remind bubbles from and Exadel logos +If theres a way to have access to smart chinese at compilation time let me know I can remove the reflection stuff so that the refactoring is more reliable +It is not possible to change to easilyCan you please elaborate whats the issue +provides a until the issues can be sorted out in the WAGON project itself. +Any mapping shouldnt attemt a conversion if the runtime types are already compatible. +Thanks +there was a bigger discussion on the scm list about things like this +Thank you JeremyIt works fine in this special case +Calling commit here ends the transaction before the message has a chance to be added to the transaction so the last message that you commit in the onMessage method will always be left out of the commit phase +Appreciate all your help +Did you try netty to see if you see this issue there too +Why should it matter +The default is ON +Or is it possible to build an application that references a class without JDBC +The attached patch fixes both the incorrect day and incorrect minutes bugs +I already know servertools team hates making big changes so Im not even sure if itd be worth my time to try to code it up +Committed revision +Nits the indentation should be two spaces and star imports should be avoidedOk +A good example isIn Hives case we do have quite a big code base and at some point we may have so many UDFs and s that we should draw a line on criticalvs extra +Committed revision +Any news on this Hans +The release audit warning is due to the inclusion of two json files that cannot have license heading included +More tests +A strong this seems to be a change that is easy to remove once we no longer need the workaround it doesnt change the default behavior and it was tested on Hadoop Hadoop and Hadoop Now if I only to make one more request this would be to RELNOTE this workaround in the Oozie docs for the upcoming release +I think that Marcelos investigation has probably addressed all of the concerns youve raised previously but it will be good to hear your thoughtsNo problem +I understood what youve done I didnt mean that anything that youve done wasnt working but that the original port of the code was based on something that would never had worked in +No Im sorry not checked yetIm busy in another job but this is important issue in my project and Ill check soonThanks again +why on earth would it ever make sense to have a in a property nameuntil you provide a sensible use case i suggest you patch your own version of hibernate +Good point about Hive server +Here is the list By the way is there a place in the web application to see all the installed plugin +committed +I guess so if the plan is to keep the current API stable +Ive fixed this through introducing ignored dependency interfaces in addition to the existing ignored dependency types +Hi Christian! Thanks for the research +This was long overdue and kudos for doing soOne thing I wasnt very happy about was the removal of All though it wasnt perfect this class at least abstracted out the building of sender and receiver pipelines +Mind adding a test to the suite posted above so thatone could reliably work on this feature +Using and Pig the bug is still here. +ChadanDo you know who has this infothanks to bring this up in Dev Mailing list Wido or Hugo may be able to help with this +Lets file issues for further improvementGood stuff Lars +Closing as we dont have enough info to resolve. +is for the leak in +Moving to EAPEWP +If in the context of your particular application entity enclosing redirects are OK you probably should be using a custom redirect strategyOleg +If wed been redoing it Id probably make it work the other way +Folks on the user list who need this are coming up with homegrown solutions and not everybody gets it right everytime +Committed to branch and trunk. +Attaching the new memoryUsage report with reduced pool sizes for JCA and Remoting +I suspect that we will make logically managed at some point as well. +This should be solvablefixed with +Cannot commit until branch is created +The old create method has been deprecated +BULK EDIT These issues are open to be picked up +Weve seen this in builds +Milos I applied the patch can you make a quick retest so I can plan a release +It makes it all a bit more brittle but there is only the one code path so its grokable +created a new JIRA for the docs noted addressed in other. +thanks for the appreciation and for having resolved the issue. +The downside is that performance will go down +However Im curious about the implementation +Okay that worked +The class loader could be using the variant for code sources. +If we can get it done in the next couple of days we should be able to get it in +Hi CharlesSorry for the late update things got busy +I have changed the issue type to the more appropriate improvement type given it was working as designed given that users would seldom use Tiles and Sitemesh in conjunction +Can you send for one of the module +If i remember both regioninfo cause problem with empty +Konstantin it would be great if you can share your plans and timelines +I agree that OSGI for our connector packages is ideal but it still will not address the issue of getting these drivers into the class loader +that doesnt work +I modified the attached wsdl slightly and now I can reproduce the problem Chris reported +It works in my test cases yet testing with more would be ideal +Thanks Vladimir Raghu! +Fix seems simple enough andsShould be fixed in next weeks point release +Thanks Todd but if you are busy we may let someone else pick up this +Matthias could you detail build demo instructions a bit more +We should not tolerate +Is this vs cache additions from concurrent reads +Part of a bulk update all resolutions changed to done please review history to original resolution type +we didnt see users asking for it +Completed +patch applied on behalf of Peter with thanks +of course as devices become more powerful this will all likely become moot +The site is back up again +I think this issue is now complete +in all other cases eitehr failure failures or failure and an exceptionZIP file containing the results is attached. + +If the axisService is not null we get the binding that the request came to and add it as a property to the got this straight from an Axis developer +I still believe the best way to index the data is to use the munged superinterfaces name as key and store the details of all proxies with that superinterface combination under this key in a +The issue David is referring to is +Hi Mirko and JonathanI cannot reproduce this with the current version in subversion Ive committed unit tests for this case in the but they ru nwith success +Two more thoughts +usrbin +And thirdly if the ADF BC Wizards dont recognize the reverse filter then that is a bug in the ADF BC Wizards not in the Plugin +Scott Ramki a similar requirement has been captured in already +The search component framework merges standard and dismax into a single handler that just has a different query moves common search operation into reusable components +It doesnt work against Weld for me either +On top of that splits are happening much more frequently as often I would sayThere may still be other issues lurking around this area of functionality but I would commit this issue +Idea to have configuration files in zookeeper is something that is sitting in my queue for quite some time so thank you very much for picking this issue upWould you mind putting together some prototype patch for your suggested class hierarchy and functionality? Christopher Well I would like to choose unified Zookeeper file to store the whole properties +Lets fix this +Thanks this certainly resolves nicely +It is also not an issue for the current trunk anymore because we switched to m not seeing this issue using jdk on rhel +I also did not integrate with the current because of my other bug which found the listener being removed after +Attaching v that handle the token syntax +a default build definition is the definition used when you clcik on build button in summary page +Thanks for the quick verification Ole and Dag +And it isblush +Yes I am going to deliver unit tests for the new imageio classes later in a separate jira +Its complementary to the docs on and the code on githubIn short its distributed commit where the prepare phase blocks all writes to all enabled via a barrier node in zookeeper +updated patch looks good +We could also just copy Solrs and change it to suite our needs im in favor if that anyway +does not need JVMIt just needs mapper tasks can work with multiple splitsEverything else reminds the same +marking this as resolved +Would a Solrj client be able to intrinsically handle a distributed shard request? It could make separate requests for each shard but you wouldnt have the nice advantage of distributed searches with aggregated facets ranges etc +Fixed with in abfadadeedebeeca +The fact that we log the full stack trace makes it big but even if it was a single line of text the logs would have filed up your entire disk if it were to be ignoredI dont think that ignoring this error in the code would be a good idea +I will create the release note once I am back from vacationImplement Krisitians suggestion to synchronize the setter and getter for variable text in in trunk and + +AlbertoI have cinfirmed that this the feature of C runtime that we are using on Tandem platformThis is not the bug with xercesc parserMany thanks for your detailed analysis and replyBest RegardsRinil +I fixed the test. +i didnt realize we had to use a special API to read back a improperly closed file! What happens when you use the normal API to read a file that was not properly closed? You dont get all the data unless you use this APIReally +That will fix the build issue +done and are optimization of the eviction method +Latest patch for the issue +I committed the patch to Federation branch. +merged +for the patches. +Whoops looks like this ones already been fixed + +Verified the test passes locally with the patch so at least this is not a regression +Freddy I let you work on the update of the SQALE Quality Model +project file +More or less working +submitted for scout +A stack trace appears in when you access that page with the failure +Feel free to punt +Plan to commit soon if theres no objections +Thanks Randall QE will verify that all these scenarios happen the way you describeWe should also dicument this for the users so that they can understand which file are changed when they perform an install +I would second this +Found the button this is the complete patch for the feature +please organise replacementsYou are also mod for their users and commits and dev lists you are happy to remain mod for those +To enable new graphics please apply patch first +That said I still dont think theres any reason the processor wouldnt be able to see the annotations if thats all youre worried about +I think we should remove the walking tour as it is completely inadequate and basically wrong for s +bug brought forward to anybody is working to solve this I could begin the debugging +You may have to wait up to hours though unless the key is synced to the accounts +Triage for +Right there is more fixing needed for the other collectors and other situationsBut I think solr should still be fixed for the common caseI dont like the duplicate calls to score +I appended call to rollback method to the exception handler but in doesent help +Being dealt with over in +They are the Driver +Patch attached +Fixing the test +Based on the comments changing to major +I can also view the WMS without problems inIm affraid I cant help you with a test case since I encountered the problem with a WMS hosted on an server that is only visible in a private network. +Im trying on. +for changing the default version used to build shim from to lets see what others say +seems one such instance +What is that youre seeing broken AndreaCan anyone think of changes that would have caused this? Like in the last month maybe two +That code is used only in assertions which are meant to be enabled only while running tests during a build how is it that youre seeing themA more accurate computation without the spherical approximation is available in +If we pass in a context node set object then since we dont make a defensive copy that can be changed externally later which could change the size and position +I have prefered to check it in the to be backward compatible with Commited in Rev to the branch in revision . +Please correct me if Im wrongSo you are saying we wont get any info till application finishes? I thought we will get this information once container starts but it may or may not have finished or application may or may not have finished +Can we get some more detail on what the app isdoing when this error occursIt would also help if the app were tested against RC +Moving out of +Clearly the method should be corrected to immediately return null if parameter is null to avoid NPE and extra checkings +Except when using the base ll update the is being released today +Thanks Ryan +Thank you Chris +Does that sound like enough or should we use a multiple of this +I assume NGP wont be implemented before. +Also affects and. +One more unit test and a performance related TODO +It did work fine but those are really plain jane +Scratch last comment +It will be in the version released for +This bug is way out of date +try guess you didnt update your local from SVN +The latter one is only partially shown and there is a third field right of it which is COMPLETELY INVISIBLE +Exclude lists updated +If running in reverse order the classloader contains of even when working over the flat module in +Ran all and everything still works +seems overly generic but I agree that is straight ugly +Hi Thomas I understand the frustration +A new JIRA will be created for the new design and implementation. +Example application that can be used to verify this issue see change was introduced with in Struts have migrated from you should experience the same in really strange Or maybe this is side effect of though +If you finish it before we release beta please correct the version +sorry for the error on type +There were two as of +Yeah it is different +Fixed +Thanks +If so hell probably have a before and after result for this patch +The transform quickstarts should demonstrate how to use this. +Hello Kojiyour patch seems fine to me from the functional point of viewJust I dont think the should be emptied I wouldnt remove it preferring to assign to it the simple responsibility of reading args without the previous explicit Node traversing but as you did using the Solr wayI also made some fixes to remove warning while getting objects from the +Looking further the does not even attempt to download and unpack it +thanks of my latest tests links are working fine nowregardsMartin +I found the problem and fixed it in revision +Patch checked in +I agree that the dag client should be able to talk to the dag while the dag is alive +Attached a simple fix +Enis Soztutar our idea was close to what youve mentioned +Can you provide a test that shows the issue? Also what OS are you running on +Patch applied in r +This is a patch to +for this +Id say change the ms to s sleep at least since unlikely compaction will complete in this time +Sure Ill add in tests covering the Java changes and repost the patch later today +I am really sorry for bothering you with that I opened this report too early +Can we do such an optimization for the query mentionedAdditionally can we use some optimizations that Hadoop has such as running in an Uberized task rather than launch MR jobs? will kick in automatically in no need to do anything on the Pig side afaik +Im looking into issues and provide find attached +Added a new patch that moves the soap header processing from to +However one issue found when I did this attempt is that a fake container for AM need to created before submitting the application +If that is true then please comment on the cases to reflect this +Thanks Suresh for the detailed comments +I have got it to free on only once +this in and trunk +Fixed on and closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Rebased +Fine with me! +However the test code violates the Dont mix concurrent transactional and writes to a single workspace guidelineHmm I understand but the above mentioned test works just fine in the branchPrzemo does you application use XA transactions mixed with regular writes +I also ran into a different problem butwill file a different bug report. +You should use not +had the wrong component cannot exactly tell when but it looks like this has already been fixed and works for me for the current trunk +Should be fixed with as it wont be included in the d like to somehow have the option to tell the plugin to make it alphabetical as I dont want to have to go through all of the poms and alphabetize the reporting section +Also included a new unit testThank you Jim for reporting +What really worked out for me was to insert my project into Apaches workspace +Bulk moved to +The issues are now closed. +This bug is difficult to solve since it only occurs on IBM JDK but not on Sun Jdk or Openjdk +Thanks for the review TedCorrected the grammar as suggested +Working on remaining items +DaisyI have extended our patch a little bit please review again and push it to the server +Patch broken due to +the problem is the same as before I get an using settings that without the patchThats worrisome though +These artifacts were never deployed to the repository because they went directly to Maven central +Well Ive installed the httpd on RHEL and which dbmmanage no dbmmanage in +We could also try to check if we still hold the lock and update it if neccessary +Honestly Rob will hopefully be able to point out some solutions as he and Denny or Grid did the work originally +Will commit as soon as JB provides a SNAPSHOT deploy of pax url aether. +There is already a job history analyzer contributed to hadoop called hadoop vaidya +I have done some modifications in the code to accept multiple but those multiple changes still have to be Rename operation has to be reviewed and simplified there are many common cases which can be handled by helper methods +Thanks for the patch Elliott +ant has passed +Pull request merged. +return was +So you can just provide a NOT reg exp +LANG PERF JDK PERF Length replacing length with length times +In this case the property request is of abstract type requestType so we must specify what the real type of the property isI have tested this and it appears not to break anything else so please can you try this out in php and let me know +Thanks +After the M release we should look at all the extension API stuff to enable as simple as possble container impls +Using Jdbc persistence without jornaling seems to ahve done the trick I know its BAD for performace but better then the whole meesage queue getting frozenAny better suggestionsCheersV +It is fixed by the patch for. +Nit can we make buildBound return Collection to avoid the extra copy? +Yep confirmed that the same registry key can be used if the version is substituted +New bundle with fixed pom +Can you perhaps nail it down to a particular test +so this is a recent error then? Do you know the last version it did work with +It always seems to happen on record XXX even with no jit runPerhaps the Btree has to get to a certain height? That is maybe the involve a Btree of height or or +The callback works differently than the callback +I have a static lock for protecting the index closing all issues that have been in resolved state for months or more without any feedback or update +Thanks Hairong +in fact it was done back in July +I dont find the propertie in axis source code +Im asking Arun to commit this +sorry steven i didnt notice that you had commented +then lets extract and make it reusable for idea as well +to Steves suggestions Id say go with both falling back to the localhost if the URL cant be reached and also adding an option to to bypass the remote check in favor of the local option if given. +The weird thing if I remember correctly is that the available element has been around since like ant or something +this issue relates to myself as it looks like some work is being done and I dont want to appear to be blocking it or such +The patch fixed this issue +for the committed this change +All should be run in one AS server current tests startupshutdown the AS server for each and that is time consuming and unnecessary +The process takes megabytes of memory in Windows XP +Committed +Applied the patch. +Done in trunk +Thanks team for contribution +The links may not work though +Indeed + +This fixes this issue as a result. +Fixed in revision +Thanks for the very prompt attention to this. +Patch open for business +Ive now put in a bunch of fixes for this such that were probably of the way there +They want to have a global configuration for all of the slavesIt is important to report this total to the JT so that it can be displayed andor used by the scheduler +Defer all unresolved issues from to +So sorry for thisproblem solved +I just committed this +and watched with developers tools +unfortunately Oracle versioning is something we do not support + +We may do some additional tweaking as far as to reduce the size footprint of all the reader configurations but that will be discussed there +Updates the description of problem an title of issue because this is specific to Oracle drivers +Oh you asked for a pull request not a patch +Attachment has been added with description AFP has been added with description PDF transition resolvedfixed bugs to closedfixed +As Rainer pointed out Eugenes problem looks more like described in to work now. +The banner provider your provide should only implement the methods not annotate with +I will check the new code o +we need em see +Im digging around +Cant reproduce in Apache Flex or Apache Flex develop branch. +Best regards George +We can then also move our to the proper location +Hi ToddAny clues on how this may be reproduced +Need to check this with language team +merged to trunk thanks for the closed. +Sorry for breaking the HDFS build will check thisTodd after you last review I merely fixed the bug when running the test on hudson the homes is NOT a Yahoo thing but Linux default for Hudson machines +Move the mouse over More +We should fix the to run tests only +When AXISARCHIVEENABLED is false zlib is not needed to be included in util +It places this new file which is titled right before the property so that it appears in alphabetical order in the Properties section of the Tuning Guide +I fail to see what value this change adds +I realized that my patch broke +Summary at this stage the NPE happens every time in our own application setup the trigger seems to be to do with the uninstall of bundles started as transient from previous runs of the framework the exception itself seems to come out of a package refresh which is triggered by the same bundle as we do the uninstall from it doesnt happen on a simple setup with just a few bundles our application takes to s to startup and has around bundles implying maybe timing or complexity of bundle inter wiring could be a s a shame +Works great thanks Brian! +The RI might not be our solution but it is very likely to be the one we use for at least. +Ill get to it tonight +Did you already have a chance to try the new version with the root context being? I dont actually know why it complained for me +Anil Saldhana says do not upgrade to Scout introducing xmlbeans dependency into AS code needs to rerun the TCK jaxr test suite using old scout to ensure such combination will work +patch is applied into trunk +Attachment has been added with description Example Java App and SVG to demonstrate the memory further testing it has been found that the problem was outside of Batik after all and even though the GC doesnt seem to want to clear up the canvases until the last minute its not an actual memory leak. +By the way now we store inner drag info into static field of class and use empty string as data in drag object +Im marking this as FIXED +No what ever has been checked in is correct +I did that two days ago and again we have the same annoying problem +I am using the most recent trunk version of PDFBOXI am using and i still get the same was released please give it a try. +If we are going to do this it should never inject and fail at runtime if injection is attempted +Then please let me know what you think +Hi guys were you able to reproduce it? can you build and test against trunk? I tried testing against my local workspace with but my changes to make Drools Repository use arent recent +In the meantime we need to think of how to link lucene analyzers to columnmetadataJason This currently works by executing the query locally if that does not have enough results it moves on to the next node +redhatewsshareconf +as a workaroundChange your job configurationGit Advanced section checkbox Skip internal tag +done commit the into branch ssd +I realize this issue came first but they are similar +Attached Java code that shows the is not an issues +Ive added a simple test in and that is passing. +clean up the committed to trunk at revision . +See +if the container hides your exceptions this becomes hard to debug and do you think it is a problem to log a stack traceit is after all an exception that is not caught by the delegation class itself and it is thrown into the jbpm engine + +Fixed in revision +What did I add to the wiki that was more specific than that +OK +I was just a bit too late! Nice work +This seems a kernel class issue whats your test VM? IBM VME or DRLVM +And create cloudbr on bond +I am waiting for this to be attached +Or at least the test attached doesnt exercise this scenario +Maybe it wold be a nice feature to be able to specify the column name too Si it wolld be easier to integrate it in own projects +on the patch +In this way components from both Supplier A and Supplier B can be included in an application without conflict +Committed revision +Anyway its usage is I suppose the best to verify it +Commit changes to at +applied with modifications +Mark as Newcomer +And yes behavior of the plugin should be the same no matter if filtering is turned onoff. +add pages for viewing switchyard message +Did you search for public static final class +My regression test run finished +I just forgot also upgraded to DS +Ill also to trunk which has the same problemAck trunk does not have the same problem the jar in trunk is exactly the same as the artifact already available from Maven central repository +Sending srcjavaorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsTransmitting file dataCommitted revision . +I changed the documentation and the source code so we use the class attribute when trying to locate the correct method to apply the properties to +It was supposed to be dismantledRemoved we really need the prefix APPATTEMPT inYes we have to +Can not reproduce need some help with this + +I changed the to use the hbase classes +As I have said in the description this will also enable backing up of only one copy of data +Updated to track patch to current trunk +Shoot I will fix this and after lunch +Where is the problem +There shouldnt be hanging anymore +Assign to during tidy up prior to +For instance it can cause a lot of problems if used in the same container with jsr +This is a mistake +It would still be better not to encrypt and simple backup the files +Is there javadocs were talking about or where can i find some documentation before coding the interfacesAlso i have one question +Sounds good to me +push out +make sure that the values you added dont change the behavior of the component see value for calendar data and datascroller attributes has been values have been added for almost all components +exclude list for your patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The same behaviour has uirepeat for begin attributeAlso hdataTable with first throws exception but it shows all elements of model when first againstThe question is why data model accepts but not +Attachment has been added with description intermediate patch containing partial severity from major to normal pending further P open bugs to P pending further of +Closing per above comment. +Jacquesif its OK with you Id like to take care of this issueCheers +Please reopen the issue if more is desired. +The other two according to Prasanna are used for cloudbridge +Groovy makes some fundamental assumptions that type names start with an letter +paritally is really overdue +Regressions ran with the a patch OK on Solaris xJDK and with one error which I believe is unrelated on DebianJDK both with insane jars +Committed to trunk + +this is just how it was done initially +Here are a few ways that Developer Studio makes the work of a developer easierCommentVery small point end with a not a +This is not fixed for Point layers +I would think that this should make using it easier Vs harder +seems to pull up a boilerplate Apache landing page despite a valid in www +for DRLVM revision GC gen mode also works +Yes lets make this change +Please attach a complete log no matter how big it is +He seems to be bit busy lately +Id rather restrict the user now until Im positive its safe +This already exists as a plugin on Android and should be made a plugin on other platforms +i was able to add a product both in created and approved status. +Sorry Its my mistake Both works fine thanks Greg please close this issue. +If there is a leak it needs to process a ton of data basically java objects that describe every object in the heap +Forgot that the java code had a seperate jira issue +this is not how +Before trying to get an attribute value it is checked whether an instance of has been registeredThis way the will not be thrownAdded a test in the test webapp that inserts a definition without usingtypedefinition +Excluding it from autoproxying is therefore not an option +Same patch with did you get a chance to see the latest patch +Joe I cant upload our test because it requires an account on our server and synchronization as well as setup steps that cannot be performed outside the application +mongrel can show the image +The workflow you describe is exactly the one I use for would help to know the path to your non default location and the path of the default location +Resolving as invalid +Corrected s kit for installing into AS removing unusedunnecessary dependencies cleaning up the assemblydistribution adding support for storing indexes in Infinispan caches and corrected a few test casesAll unit and integration tests pass with these changes +created let me know once a release or KEYS file is present in distrelease Im going to assume that I need to change the status to let you know that the directory contains the release +or K is back +Screenshot how it looks on I just realized that I failed to see that the classes for X and have been generated indeed + +I just attached the memory graphs Visual VM snapshot and log file from the latest test run described above +Modifying Javassist to do so is not a simple task give me some time. +I dont see any other reason +Committed. +I committed it to trunk and and made the commit message reflect this factThanks Jolly! +Please take issues to the per above comment. +But if this is in fact an improvement it seems reasonable to fix the test +When the arguments call was complete and popped off the stack methodname didnt return back to the previous value but instead stayed the same +this patch just brings in configurable parameter thats already committed in latest nicolas compaction changes on trunk +This is causing the reports to be placed in trunktargetmungedIt would be nice though to be able to have Hudson run multiple profiles as Jakob mentioned above +A single patch makes release make more sense because with DDL but no DML you cant experience list bucketingWe have to pick up one +For the record the Jenkins build area has been cleaned up and we now only maintain builds trunk Nutchgora and a maven trunk +Rev +And update the JAI one too +I will try the simplification you recommendIn the network case it was easy to pass the error from the getter to the factory method which makes the +Please speak out if I get it wrong +Good show case for measure first and optimize based on measurements +But now that you mention it I indeed could not release because it was a git subfolderHowever my point still applies +Closing all issues. + unit test since this mostly a document change no behavior will be changed for now just to be consistent as the code evolves +Ok makes sense +I would prefer that no duration is allowed in the script file +uh +I think covers this +Logged In YES useridthanks for the report will fix Username lqd +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Implemented for. +then please commit +Yes i know +Update of all files +I just wanted to share with you experiences and wishes of lots of Maven users of my company +Thanks. +Laurynas thanks for the patchGood idea about the new option names to keep it similar to the body +If not do you have a description of a manual procedure for verifying that cross talk does not occur? +Embedding the app in a servlet container would bring overhead and constraints that only get in the way +thanks. +The patch that adds the clone method to the had to change the variable to in order to change will provide a junit test test to test the change made to the +Verified to be fixed and closing issue. +On my machinewith a Sun jvm the decompression performance for bitsframe size varies between about M intssec unoptimized and M intssecoptimized as reported by the test +But because these issues we can not refactor the portype name change in the bpel and wsdl file +Oh yeah and an addition to the docs +Running nightly on the branch now too. +It was easy to code the methods because they are mostly copypaste work but can you lend me your help with the documentation? Im posting the patch now +Simple tests which require GC seem to be working now +Removed the last comment but cannot be left blank +Move is done and compilesHowever it introduces some unlogic dependencies project now depends on projects because uses the FilePersisentSo repository hierarchy must stay in core projectsWill revert and try to only move and Store hierarchy +If you really insist on it you could expose a scoped proxy through a and inject that one I really wouldnt recommend that things considered I believe that the present solution injecting the raw objects is the preferable solution in the general this feature mentioned somewhere in the documentation? +Sorry for my bad english you saying that you see an unexpected flicker effect in the window when you resize +On the basis of inputs it shows the resultant of minor changes are required other then that patch looks good +Thanks Dag +Patch it fixed the current counting also modified the unit test a little to check if the prefetched result size counter is not always in case current is not calculated at all +This would not change the behavior of sonar but would improve usability +However if we want to host it as a plugin here it will need to be ASF licensedOkay so I hereby state that the code is dual licensed under BSD and ASFIt would be really great to see the plugin hosted at +If everyone thinks it is fine let us push this into patch also includes changes for plan to commit this shortly +Committed changes submitted by Stephan using revision into trunk +Ian can check this in and mark resolved +I never use ant myself anymore so Id totally overlooked this issue +Attaching the latest patch addressing both Uma and Ivans comments +I updated my for +I think I might be attempting to pack the contents of an empty dir the dir where the jars will be but not where they are at the time Im trying to pack +Patch applied at +avoiding the issue by removing comments. +Thanks Flavio. +it has max parameter but no offset +The JAXB annotations are lost +can we play any tricks with the dirty flag instead? news on this issue +We generally dont put things into the Patch Available state until the contributor believes that the patch is in a final form +all issues to the CR +Thanks +so maybe we dont need to also log the actual element count +Yes I realized my memory was not that accurate when I opened the dictionaries distributed with LT +Sorry +Patch a fixes the typo in the error message +I think these constraints should be honored like regular form validation controls though violations may have to be displayed in a window of some sorts to not interfere with the regular Grails table +Thanks +If we no longer support bit timet should we add some codes to detect the violation of constraints during preprocessing +Need to merge to +Im not sure if this is conclusive but it seems the larger shapefiles are the ones gags on +dupe with out resolved ticket +We can document it but this is not exactly the behavior I would expect +I am using clean install site and for some reason the surefire test is executed four times +Its only happen in and +Sounds good +Im pretty sure its fine though since I was able to process the chapter in a different context +for example feel before closing the streams it would be good to have null checks to avoid unwanted exceptions +The files dont seem to be equal +This needs persistence support. +I have done a fresh install of Eclipse and all the and cant reproduce the bug myselfS +Closing this JIRA as EPP core docs development is stopping for the minor issues please open a JIRA against EPP this JIRA for blocker or critical issues +Some of the inherited methods loop back through the JCR moved the protection check to delegates in revision . +aab Fix RFC plugin buildab Plugin to implement the and features of RFCThanks for contributing this plugin! +It is apparently +who knows +That is why we need to rev up the schema version +This change will be released in Spring I have not included explicit GZIP support at this time + +Probably something should be done about it +Committed this to trunk and +You say an infinite loop happens +Committed to both trunk and branch +This will not be so efficient to support inner classes +effects as to Permaine +This issue really does duplicate so Im linking them as duplicates +So there is a chance of other property prevent this wed have to prefix all system properties +Julien if there is something that you expect from the jBPM team please create a linked jBPM issue and explain a bit more what exactly is expected +oh I didnt know is there any replacment +Im not sure how the tests are passing since all the peers are going to bind to the same portI would propose removing the electionPort property and adding a port to the server configuration line addrquorumportleportso becomes for exampleIf the leport isnt specified should we assume to leport is one more than the quorum port? is the same as The first order of business should probably be to figure out why the test cases are not broken +TO Camille Fournier what you said is right but it does not solve this problem +ZavierCan you take care of the line length warning? the long line for readability +bulk defer to defer to +It can be closed at the same time. +The test case works with DB but likely fails with DB +Can you attach the code used to make the call +NMs sync with the RM every second +If its hard then I would propose to prefer MTOM if both are set just to be consistent with Axis behaviorthanks is the patch for this JIRA +Committed +So if you supply it with block locations it will leave safe mode and start play role in block replicationdeletion lease recovery etc along with the +I think when a connection is destroied it should remove itself as a listener to the related +So the purpose of this jira is to fix the following situation +It seems that this bug is still in maven +Eventually it will fail +I recreated the workspace and I still get the same problem so you should be able to reproduce with the project I have already uploaded +By unexpected I guess I mean that introducing a new dependency as the default was a surprize to me +I know as I said I wanted to be sure these tests were what you had in mind before investing time to get them running in the new implementation +Fixed with the patch applied for Issue +It is now the responsibility of each project to keep the latest subproject of metadata. +This has been resolved in commit acaffdeaccbfbbeca +Excellent thanks Richard +taking this off the list at least for +Attaching +Attached you find a patch implementing what was proposed in the email discussion on the alias Loosen the restrictions on the placement of extension elements meaning that the jdo implementation needs to do more checking +Pushing back to +This patch combines the URLParser and Patch from Benjamin Gould +Closed upon release of Hadoop. +Ive done the bare minimum to make it work the same as it did before +I think the more common thing for projects that generate multiple jars is for them to be disjoint so it might be less confusing if theyre disjoint +If you find further issues please create a new jira and link it to this one. +Eclipse project contains Test Case that run successfully in and fails in +The Spring Integration team has requested this feature so that they may use to meet their Twitter integration goals +Patch applied at + of unclosed issues. +Ive attached a patchPlease review +Verified +Attaching initial patch in case review board is having issues +Thanks for your prompt response! However the PNG image now has an odd version committed so could you please replace it with the attached correct file +I did not do the part because +please move back if they will be worked on asap for +Anwe are not speaking of the same thing + attached Brian any idea why this is happening? The normally work fine +apparently i forgot to remove some that was related to my os x machinesall should be good now revision +Thanks stack +Almost forgot Thanks Arpit +Verified in jbt +Went with a slight variation of Cocos textFocus on your application not on software development infrastructure +At that point the thread pool still has a factory assigned to it +Commited patch +Here it is and I put a more explicit wording. +Keep in mind that you can warm the reader yourself by issuing a sorted search before putting the reader into duty of course you dont get to warm from RAM like with what you suggestKeep that feedback coming +how do you like it to get changed +all ResolvedFixed issues lacking a Fixed Version +Since the spec was deficient the branch for test case challenges should be updated to skip this test and the trunk should be updated to check for for datastore transactions and for optimisticThe spec was updated to reflect that with an optimistic transaction makePersistent transitions detached instances toThis change now makes JPOX fail the test +bq +I resolved the dependency issue on the branch with the caveat that the dependency is now on because nobody has pushed to Maven yet +Uwe are you working out a patch for this +orm which are in my workspace +ALLOWMACHINENAMES with a list of allowed machine names is a good start anyway and itll cover most possible cases +The shell could check to see if it was time to dump the trash each time you run a command or somesuch +In the replication use case were adding ramDirs so we dont need to immediately mergecopy them over +A workaround for now is to temporarily pause the scanner jobs and use the force compaction feature of the master UI or API +Thank you very much Sergey. +Yep +Im not sure we want it in branch at the moment since code that implements wouldnt compile anymore. +is only used in and +also the payloads pointer +There wont be a release +We have tried to apply your patch from but this didnt change anything +Hi JeffI tried this with and it seems to work fine +Thomas +Can you attach your and? It would be nice to have the rest as attachments as well +I just committed this +Chris if you have any direct pointers do let me knowThanks guys looking forward to making a success! I think you can still apply to if you want Im not sure how the ranking will work out so the docs you posted may be right but youll never know if you dont try +This may not be the case already everywhere +Cant did make some fixes on other bugs recently which may have fixed this +was fixed a short while ago Central should be fixed in an hour or you! Now I can Grab the latest Selenium. +This adds a barrier to entry +One of the first things incubator projects do is to set up a website so I assume not +Simon hard to kinda split the issues when there is nothing in openrelevance svnThough this might look largeoverkill for one collection maybe I should have done two to illustrate better? the patch is mostly basic stuff you speak of License files s +Allowing the creation of the DBF when set throws an exception is easy enough +Specific tags may have their own support for encoding +If all your products depend on XML Beans you want to be able to declare it as a base dependency and have all your modules inherit that dependency +Ill mention the backwards compatibility issue in the release notes + to change the default of toplinkfalse to toplink patch updates the default value of toplink from false to issue is dependency of +Task is of a bulk update all resolutions changed to done please review history to original resolution type +Im attaching additional patch please apply itThank you for this notice +Committed to trunk +It is no priority for us at this moment +Igor why comment the comma out and not simply remove it? Seems wed want to remove unused or unnecessary code to keep things clean +smoke tests itPart always uses an svn checkout not local copy of stuff +Could you please write some short instruction how to configure Nutch to serve results in XML or JSON using this pathIt would be very useful for newbies like me thanks in completely reworked framework with extension point for serving search results in different format +Some of them still need to look into additionally probbly with Nick to add some details carefully +We can think about it in the followup +The failures as a text file +patch to upgrade +The patch with the formatting changes looks good to me +Yes unfortunately thats going to be a of doing this +When testing the +Looking at the Java EE latest draft it appears that the spec calls for JPA compliance +r +Thanks Ill take a with a thanks again Bela +Hi JodyThis is a first draft on updating the doc for releasing geotools +side comment i think we should rename all of these new params instead of +Can you please provide more information? Line numbers for the stack trace the code used to setup +The page has to be updated for the I therefore do not close the issue +I cant watch the RSS all day and frequently miss replies to posts that I have tried to help with + +Once that is done I am happy if you want to merge to trunkThanks +Ive got it so that I can reproduce the fact that couchjs processes arent going away +Ill if it turns out youre lying about facets +Wheres the validate profile defined? I cant enable it in trunk +Perhaps I could add this to the repo +Fine by me did you disable the test on branch +The patch uses a system property to control the behavior +Enunciate could name them appropriately but then it wouldnt match what CXF is doing and youd have a broken WSDL document +Ive attached third patch that returns if input is null +Merged in Horias which changes the test case only but does not fix the issue +Thanks I tested stuff out I got a boost from doing buffering that cant be replicated with just the grouping of atomic increments however if were just going to go with the simple version this patch is as good as it getsA similar patch is needed for input streams +Marking as resolved since this new feature has been implemented and verified with new integration tests. +bq +Thanks Jie Huang! Ive commited to trunk slipped into here +However we will be producing one or more developer preview drops from the branch soon +As an example we do not know the datatype of an inputparameter at parse time so we wont be able to decide if it will be a numericexpression stringexpression booleanexpression or something else +blockerviolations and compute the aggregation at the plugin side +No Axis changes needed. +committed rThanks +attached log +we need to update the package names to be in the apache space +What I figured out is that somehow maven sitestage is not called on the modules first but starts on the toplevel project +Thanks Chris! Ive updated the with the details +See Andy needs a little more time in deciding whether is good enough replacement for +Fixed in trunk +Lets take care of the connection based authentication state in first +WARN Error starting tracker Permission denied userB accessWRITE inodemapredsystemhadoopsupergroupDoesnt this mean that the Jobtracker failed to write to the mapredsystem directory? And that it failed because it did not have permission to write into mapredsystem? Im not sure that it has anything to do with the jobtracker not cleaning up after itself +I verified on IE also and issue occurs on IE as well +This makes sense +OK +Correct duplicates. +Looks good +after release +Under what circumstances is this showing upAlso if you could please repost your patch as an attachment and in diff u formatThanks below is a new solution to this problem along with a test case +The transformation +Play the game with the trunk +Verified this bug is not present in the MR container executor so marked as resolved +Ensure the unit tests are made generic and that they run in both JTA and JTAX modes +PONGMy super comment +Part of the resolution for this issue was checked in with +You may want to open an enhancement request in Eclipse Bugzilla for that +some of the resultset numbers are now missing in the middle RCL and this cause code generation to complain for the last query in test +Also cleaning up temporary directory at the end of the job needs to be staticbq +In particular I had to change the access modifier of a variable from a arguably I could have just created a duplicate variable within the class but I expect that other subclasses will have similar issues so I opted for changing the parent classThe other ugly change I had to make was move pos into but still require inheritors to update the variable + +verified +Thanks Harish! +What about Ted +Scott IIUC youre using the system property to define your JNDI connection URLIf thats the case please provide an example value for that we can use for testing +Chris is this still valid? What do you want to do with this issue +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +One solution would be to provide a similar method on the Manipulator interface and to implement the arithmetic operations for all DM implementations based on the Duration values +Cannot reproduce in trunk +I just committed this +For now my recommendation is that we dont fix it +of course it would help to get comments from real usability experts because usability is not my area of expertise so this is jut my personal opinioni have no doubt that you find the feature useful and so will some other users but my feeling is that our decisions should be intended to make things as simple and streamlined for most users as possible and in this case i think that putting the url back in there is ugly confusing and ultimately of little or no value to the majority of users +How can I stop Master? Stopping working master means that my business process will crashed +The changes included +I resubmit it because my laptop keeps on crashing on me today so just to be on the safe side I dont want to lose the diffs +Could you make a try using KarafThanks +That is what Maven uses. +And it has nothing to do with a probable bug in the query api +Link to the jacorb patched update +Neat thanks +Afaik theres a default forum at the Codehaus right +Resolved for see for +To enable better tracking of capacity utilization by teams sharing the warehouse we want the ability to group users by teams and let them operate within a database as their logical containerI dont see how you accomplish that by adding this configuration property +The error message has been changed to the followingfailed to create file xx on client yyy because there were not enough datanodes available +restore the snapshot is easier +Already attached a release note for it +Can we gen an exception saying Incomplete transport sender missing sender! or something of that nature? +All work is done on develop and merged into master when we release a new version +Yes I checked that and I dont get any calls on the CS mgt server +I saw some questions about the holes in the approach that were mentioned in the JIRA much earlier and someone couldnt find them but I cannot find the question nowI put the details in comments in the will be going off the grid next week +Support for dynamic dimensions added on all issues that have been in resolved state for more than one month without further comments to closed status +Sorry +Thanks +Works also with the WMS indicated in the bug description. +initializers are there to synchronize certain things between application nodes +I hope we fix it as part of this jira +I think the work to pick a trigger master is really awesome +interesting +Defer until next release +Lets see what dblasby says +Even after upgrading to Spring final +So the null is retuned for the privateipaddress table +Works pretty well but The title of emails for those three types of notifications must be Project XXXXXX change on issue YYYYYYYYY +Now I work on which also use jettyI think I can provide the patch next week +I have added the page to the wiki with only minor edits +bq +Please apply the patch from site location +after chatting with Vincent the use of the deployer needs to be refactored and there needs to be a +Sorry about that . +Ive committed this to trunk will leave open for since QJM has not yet been merged +So there could be a bug +Enabled last patch passes tests on four platforms and is ready to be committed +Wow that was fastI committed the changes and saw the results almost instantly +MaratCould you provide a unit test for the new class also +Dont bother to attach yet another patch +Committed the appropriate versions of this patch to and trunk. +Added a check to see that the child is an instance of before casting it to same +Is it to capture operands? It doesnt seem to have any special features +Actually given that the exception is No enum const class search field +incorporated all comments from Chris +Er thats RFC that obsoletes RFC +Ill ping Pete and see what he thinks about it +This bug is not fixed in IE and of version +Add a masters file similar to regionservers with pool of servers for master election +Maybe you can check first that JIRA is for handle error message in GEPThanks +Patch verified +Thanks Owen. +We could just use the strictTck flag for this +LGTM +The conf file doesnt exist on a clean checkoutGood catch +There have been no changes in the build from to so those jars should be completely equivalent in terms of their you find any signficant differences between your jar and? Differences in content or manifest entries? further insights? If not could you please submit a test case for this one that work with jar but fails with jar? this as not reproducible for the time being +It looks like logj added a trace level so just one change to the should allow this to work as expected +Please let me know if you think there is still a problem here. +Added this functionality tests in branchThanks +Unless there is a jboss client that is exposing this info the bug is in the unamed client application +Thats however a different issue if anyone wants to pick it upI also logged a small fix to the web site +amended the docs but havent added links to the containers at this to both guides +Oops some renames of need to be done + is included in EAP +projects for reasons stated previously +Patches were applied at the revision +Resubmit patch for automatic qa testing +Please keep reporting the issues big or small and well see what we can do +We do not want to encode them in the component uri stringIf we use the to do that we cant get the parameters from the encoded uri because there is no any or Willem +could you help putting them in conf tooIve found only three parameters missing +Same issue here does not happen on +Users who subclass and also borrow s impl are very advanced and can change their code +Patch soon +Eg +Nonetheless all preceding HADOOP tests are passing and this particular test in no way reflects on the functionality and compatibility of within HADOOP +What is the opinion of othersBest RegardsAnish +Moving to +Overall I think this is a good start on a shell +but would be good to get this in regardless +Hi LilanthaIm just wondering why you get this kind of an error +sortBy should be of its just coerced to now due to wrong type in config file +I hope to get this worked out soon with my employer but for now Im reassigning the JIRA to you +When log rotation happens the new filename does not change the creation time of the rotated log file +yes for twill you are listed as an you as a Committers and as an Administrator for I still cannot perform admin actions +t look to apply any more +Hmm Im using source which should also set target automaticallyIll have a look +Closed due to a lack of feedback. +CommittedDIGY +I noticed some stack traces in the tests when the scheduled disposal occurs during the bundle release phase could we handle that? Maybe stopping the disposal when the release begins? +HiThanks for clarifying a similar note Im not sure what others think but would it be better to set dropDupsfalse as the default value? I find it a bit unnerving that the default value might make me lose a lot of data. +No further feedback assuming fixed +The idea now is to add test case after test case to check various scenarios and to validate the taken approach +I didnt do any refactoring as per Andys comments +Looking at the test and the code this is a EJB client API implementation issue +It has been doing for openejb release and release +Manually tested +Thanks Daiji for the patch +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian this issue. +Thank you! As you say it is a little hacky but it works for now +to trunk +war I will still get the CCE +If resources are free then it means RM does not have work for them +It would be quite nice to have some examples for conversions +Check if facet is present and has proper state +Thanks for the suggestion! Added for you for the fix it was really +shouldnt be burdened with coding for a specific schedulerAnyways Sandy can you please update it? Id like to get this in for thatll happen sometime soon +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content can answer this from a technical point of viewWe have the compilation instructions for each platformTechnically it should be possible to compile libaio for these platforms +After all you paid real money for this stuff +Will post a patch a patch +I tried restarting NM several times in secured mode +I managed to mess the simple patch +EricI took a quick glance at the code +Lot of changes have been made to trunk since then +I made a few minor changes and committed itI assume we no longer need the first submitted patch so I didnt commit that +Just FYI Oliver Tomahawk also uses and if something is changed in shared and tomahawk you also need to build locally in order to build tomahawk without problems +Failure is in an unrelated hbase test +Is there any indication when is going to be released +wont fix this in release +A new bug will be opened for entref which is a different issue +I think I was confusing storing the whole field with storing the term vectors which can doI still think at the very least being able to store binary values via a stream is a necessary addition to +Actually only line numbers changed there was no conflict +Awkward names may in fact be beneficial as they will discourage the direct use of the implementation classes +No need for IT +Fixed +The behaviour should match the JDK classes whatever that is +Committed to and trunk thanks for reviewing. +Did you found any solutions for thisI have the same cannot work! The Java VM is still a Windows app and as such it does not understand the Cygwin path mappings +This issue ties into the ideas expressed here +Im going to study the Hadoop RPC code to determine how to apply this +Thanks a lot +If you already did the work why dont you open a bug at and submit it as a patch +this is a scenario that has had regressions in the past +Maybe its our setup +We need to think about what to do with it +Patch for the issueBumps version and starts using in +Im moving this issue to since it is actually just tracking a change that will be available in Spring +I assume the timings include the percentile calculation right? This could be related to the changes in the Percentile implementation in +Dfs needs conf to get started +no I think its better for JPA to read the validator annotations +is is included in last JBT update site contains further objections or lipstick changes this is fixed. +I have added Case These changes reduce the number oferror in opening zip file? exceptions +Created a test case but it does not reproduce the problem need to see what is happening on the real site +Submitting the latest patch to Hudson +Verified the dHTE scenario patch to remove dAH as describe and expose dHTE in jmx via +Incorporated Zhengs comments +OK I think its starting to make more sense to me now +But try running the identifiers side by side you will see there is a considerable difference + +In addition there is currently no good place in the Felix project for this code +The whole lot is configured in a Spring Beans XML file using a custom +There has been no decision to do this and in any case it cant be removed from them all as some require it +Even just general design direction if its not currently where you want it to go +Draft complete +When I try to look at the attachment you uploaded I get the messageWhen working with LZH files the Extract operation requires the LHA external program +Ive also just tested out this option with and confirmed that it works similarly to the Sun JDKIm going to go ahead and commit this shortly +Committed to and trunk +If somebody wants to use it on maclinux its better to copy it manually than IMOThanks Andreas! +Nice hack Ben! Doesnt have to be pretty to work +I got the file from dropbox +done +So this will only work for +I approve and have applied the following parts of Dave Brosius patch to the latest development codeI did not apply the fix to because this is a generated partI did not apply the fix to because this last store to a variablethat is never used is dead code but then so is a bunch of code leading up to that storeId rather just leave it alond for now rather than upset the apple cartI did not apply the fix to because it is wrong +We could flush the buffer every minute or so so that when the namenode crashes wed lose only the last minute of access time updates +For your convenience I am attaching which has the with extra logging but not the fix for so we can see if there is any more information in theThanks for your continued helpll be happy to test the attached additional logging +Have you been able to reproduce this issue on your deviceI can run some tests on my Xoom if you have something for me to test +History works fine in FF but breaks in Safari +Not yet +If I change the test to not use fast diff then it passes +The patch was updated after the later commits +Must be used with Jettison and the NPE error happens before the actual request is sent so the endpoint is never called +The bug is solved in the current Timepicker version +Sure +I am convinced that is the way it used to work maybe something went wrong in the mavenisation of the project +mind you creating a quickstart with junit tests for the different use cases +Why do I need to bother with transactional scope if Im not even using any database resource and my services areRegards +bulk close of resolved issues. +I update the issue header +this +Any progress on this one? Theres been some discussion on dev about the importance of having the comments get copied into JIRA +The only point that should be improved is that once the association wizard is closed theres currently no feedback to the user if the association was successful or not +Oops mine wasnt running rake so I suppose theres a number of other methods that get pulled in there +Missing Apache license as comment at top of new unit test anyone tell why hudson couldnt apply the patch? I generated it according to the hadoop wiki docs and can apply it just fine to the trunk myself +I have only finished adding big data test and complex column data test +Thanks Dyre +Patch applied +Thanks +Close after release +haha yeah I think soStack Im going to file a couple of issues during this week related to Cascading and have no problem moving them out of if everything else is done +Fixed with +I think this is a candidate for a blocker unless we can explain why it is occuring +Maurice no one is more senior than anyone else here and all checkins are commit then review +I am fairly confident that they should be removed as they as inconsistent with all of the other quickstarts which dont include eclipse project files +Since i believe the status is correct i assume the modifiedTime value can be reset here as wellPlease comment +Thanks this makes it much clearer. +Added a lock to prevent concurrent running of and paintComponent and drawLayers methods +Created a little test case to demonstrate the issue +However I think the semicolon should only be treated as comment character if there is at least one whitespace before it +Deployed to JBOSSHOMEserverSERVERCONFlib +One of our customers is encountering a very similar issue +Will work on patch and provide asap +Thanks Todd +If you just include their names it should be enough since they are standard artifacts available everywhere +Verified fixed in CR +so as a result i have no feedbacksuggestions on the patchusage +Warning Could not find file Gcvsmavensrcxsd to time minutes secondsFinished at Fri Jul CEST The file is called in the HEAD the attached patch to fix the problem in +is this still a problem in the nightly builds? if so can you submit a patch using cvs diff u +A regular DES algorithm is not considered to be secure enough +no pom in +Im sorry but the only way to solve this problem currently is to uninstall the plugin and then reinstall will be looking into this installation problem next week +Fixed with +Thanks Becky +Attaching the full with the NPE +batch transition resolvedfixed bugs to closedfixed +The patch removes those constructorsThe patch also cleans up the callers of the constructors so that they no longer cast the instances toAll tests ran cleanly with the patch. +What compiler are you using +I have updated the axis engine such that all XSD types are supported for serializationOver the next few days I shall put in place the remaining deserialization support and then updating to fully support these additional XSD types +batch transition to closed remaining resolved bugs +I agree with you! One note I believe that tasksrunning field is not necessary +Merged to in Ricks +Thanks Michele appreciate it dude +debugged through the code and tested the new deployment command in the in the instance +I have created an archetype for ESB projectIn order to add it to your local maven repository the following should be done +untested +Both logs contains thread dumps on different stage of server Cluster just started Test users connecting to cluster Users stay connected to server Users disconnected cluster hangs up Few minutes later Even more minutes workaround would be to avoid TCP and use workaround might beusesendqueuestruesendqueuesize +Section Agenda Groups states Agenda groups are a way to create a flow between grouped rules +Unit test that can reproduce this issue +I havent been able to reproduce any OOM issues with in STS +Let me rework my patch then +Thanks Areek! +Still the test compile seems to fail. +to trunkWe may want to consider this for a if such a thing comes about as in a sense its addressing a regressionI tagged this issue with so we can find it later if we decide a is warranted +Swithcing to the indeed solve the problem +This is the main feedback Server has for the client +Will these have the same UUID as the retries in above +sorry to waste your time with this +Not to start a religious war or anything but I am kinda wondering why not to use a standard Java webapp framework such as Grails? Theres a huge community working on it and theres a lot of people with expertise which will help to ease the development of user applications on top of +Thanks for the detailed explanation Kihwal +nodetool is displaying wrong IP in the ring commandI see that you reverted back to use instead of and I believe that is where the problem isI had no issues with the last svn revision writesreads all were working great with that revision tooReopening the issue +Pull request appliedThanks Larry +Apparently the hide method is not properly called when coming from a XHR requestL +Do you want me to write something up in another Wiki +Editorial and committed minor editorial changes +What confluencejira key would you likeXPXPLANXPLANNERLeft to my own devices I go verbose and would do XPLANNER +Fixed in a moment +Fixed through +After deploying the war file with sample posted above the application is not accessible on the virtual host +per the issue meeting on May +Please verify and fix if its needed +I will check +Youll need to then explain your setup better then because I dont understand what it is you are doing +I do not know if others have to install JAI but I am running my prgorams in Linux so my platform requires this +Rebuilds the in the when readers have changed +Navis you can use my rebased patch to update review board or if you dont have interest in this any longer no worries Id be willing to take it up +Fixed the files with an extra licensing header due to a bug in my RAT tool patchThe fix is committed at rev in the release branch +bq +just double checked +Fixed a bug in the aggregate that caused a Free CAS request to be sent to the client instead of the outer aggregate CM +Hmm +Would you please try my patchThanksBest regards patch applied at revision r with minor changes thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Heres an updated patch that does as you suggest +I reviewed this patch and have no comments at this time +I have to dash of to Germany now and will be online again mid next week +The gettruetype is not for the current list its for the element type of the list which is why the cast appears inside +for trunkcomponents +Id rather have constructors overloaded +this was actually done a while ago. +This makes more sense to me Id rather use a Label for this and simply make it clear to people what the string should be since this is how the Label feature is intended to be used +Heres patch for review +A new class added to and build successfully +I have tried the mentioned metamer page for simulating this issue. +Sorry for the rookie question but can I apply the patch or does a committer have to do it? If I can do it whats the process for that? Just a simple svn commit +Duplicate of +Does even support comments? I cant find any reference to it +We cant ship code that would have no good coverage now can we? +Thanks +Do we care about thread safetyIf yes then I think wed either need our own version of the library or get some fixes rolled into the upstream project +Documented as a fixed issue in the EAP and release notesThe password could not be changed in the Seam Booking example because the transaction was aborted due to invalid validation +it lives on as indexExists now +If you need any help feedback or just want to bounce some ideas off of someone please let me know +The user may have reasons to create encoded XML files with no XML declaration with providing the encoding thought the external media +Do we look for yarn +Kind of +If the user is using an enterprise maven repo then the plugin would simple let them choose the versions out of that repository +Thank you very much. +Closing resolved issue no response from reporter so assuming ok. +Ill try to install a version of the plugin Ive used before +Tried it once got some wierd property type exceptions just doing the rename mandatory tweaks +Thanks. +Thanks +I dont think we should rename the directory analyzers to analysis I would rather go for analyzerscommon and analyzerssmartcn or a similar scheme +Im assigning this issue to myself +Figure out whether this is still broken on branch +moving to This wont make it for think we did allow users to delegate AJAX calls to their behavior as they want with the new event dispatcher API +serverdefaultdeploy is the correct result +Done +If the report is run after a clean the problem occurs if the classes are compiled in your target directory then checkstyle can find the custom class may be a little hassle until a fix is completed to do this rather than the post report generation goal mentioned in the previous comment +In general if you provide bad input you cannot expect it to behave properly and hopefully it should throw an exceptionOK you can close the bug as this is not a bug with such a data interpreation policy +So chosen to name the pre import query as delete +Having such a request shutdown the database might actually increasethe potential of a denial of service attack +With this patch all of specrubylibraryzlib tests clean with +Thanks Xuan! +Rare +An point to remember is that we might want to perform the mapping after the pipeline immediately before delivery which is what the current scheme doesAlso keep in mind that we need to recognize whether a recpient is valid which requires an account an entry in a VUT or a specific entry in. +Thoughts appreciated +Closing. +Why not just give them an example and let them come up with their own? to remember the examples are tied to this usernamepassword so we are using these +Merged to fuse branch +No errors +While I agree with not supporting DCT specifically in cqlsh it wouldnt be crazy to have a default strategy in cqlsh to handle types it doesnt know +Great after applying the patch for the PDF was created perfectly will test it again after you submit your updated pathtch for but for now my test case is working +Extra comment on Jonathans commentAbout deletesSo when you enter a new file and walking down the lists doing the sorted merge between the data in the store the gets and the previous deletes +What I did was adding a caching for the meta methods involved especially Closurecall and asBoolean +Marking as wont fix as indicated above +Hopefully most of the problems share the same problem +The same is true for +The patch works for me +Committed to trunk +Committed to trunk and +Another thing worth mentioning is size +please apply to applications is the latest in branches +Thanks a million for taking care of this Nick! + +When I changed to use the same method as the problem went away +am reviewing this as part of review Patch Available backlog +When it is a timestamp based delete same TS to be used for all the columns where as in normal delete diff TS can be used +All tests passed please commit the patch +Here is a first cut +I just have one call in my test codeThanks! in current SVN +So detecting the abrupt exit might not be the issue maybe that version of the loader should be recreating any existing index filesFrom what youre saying if a load breaks I need to manually delete any existing files as the indexer will whats there even if its broken? +This patch sucks significantly lessit only takes about seconds to reverse an enumerable object of items which is likely the most anyone will ever use it wipe my hands of it +There is nothing special being done here +looks into trunk +images was for Beta to review review and assign on on +Aliasing is not needed actually in the code snipped +definitely for this +Actually this wasnt solved as a side effect of rather had to be addressed separately +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Were already working on an container implementation as shown in +Done as part of the types work which is now in trunk. +DennisI object to this issue be downgraded from a BUG since the desired behavior worked in a prior version of the have installed unzip on my box to get around the problem of the issue I continue to have an issues with the chmod command +docbook exists and was complete for NOF but still need to go through and update from NOFIsis +The assumption is that the region is broken somehow +Sorry I forgot the fact that the thread is dead when the uncaught exception handling is called and thus was thinking that thread would continue to run until the would stop all the threads. +Maybe its a duplicate of +mvn N versionsset or mvn versions mvn N versions all help updating the parent problem is not so much that updating the poms is difficult but rather that changing the poms has some quite annoying consequences +Now that we arent looking at release deadlines yet I think we should spend some time trying to create a test for thisAt the base it looks like we can create a test for this by writing a junit unit test that manages mulitple hive client instances each of which resets the conf at the every accessFor I was considering something like writing out a hbase table directly using hbasehfiles and then trying to read it as a hbase table from hive +I added an deleted the singleton for the and user the init method of the annotation processor +For debugging possible errors or for presenting the it is more comfortable to have a with only one blank and not or more blanks between two wordsGreetings from Hannover is it really that easyWhat about double spaces in selection textselect from mytable twhere USER where there should be spaces in the USER text as its a char fixed column +Lifting priority as Ramkrishna suggested +we have a training today I will update a new patch for related jiras. +Awesome! Ill look at the patchReuse is important w s +Please check that it was applied as expected +Believe me I speak from experience here +A simple way to emulate that is to test using a sliding window Days training day testing then training testing etc +Fixed in think we should be using is were investigating if a type implements a certain interfacetrue if the c parameter and the current Type represent the same type or if the current Type is in the inheritance hierarchy of c or if the current Type is an interface that c supports +Please close this issue thanks +Actually we can track this version check while creating ledger itself +I think it is the way to sure what you mean about the item buffer we dont always have all the items in memory for the duration of the chunk so it would add some extra footprint possibly +OK I think this is a good approach in the emulate old on flex layer and then in the docs for call out that the incoming String cannot contain unpaired surrogatesJust so you know its not perfect back compat though +In the test case I would recommend we write a deprecated key with multiple values like the key X and make sure that loadResource actually loads all the keys both old and new +As a sanity check for I built Hadoop locally with the patch from applied and changed the version of in the POM to +Sorry about that +Patch looks good +i will be happy to move the labels anywhere +Quick start example +Attaching a patch for trunk and +It seems the behavior you see is due to a corner case dealing with the repository subsystem creating a duplicate index +The fix is to explicitly reset the arrays in the objects when we need to change the object stateTouches the following additional filesM javaengineorgapachederbyimplsqlcompileExplicitly reset the array of types in a when we bind the original cells in the arrayM javaengineorgapachederbyimplsqlcompileExplicitly reset the parameter types of a routine when we bind the individual cells in the parameter arrayM javaengineorgapachederbyimplsqlexecuteExplicitly reset the array of in an index descriptor if we end up changing one of them while processing a DROP COLUMN action +They are read only +New chapter Creating Maven ready Seam project has been added to the Seam Tools Ref guide +And also include the changes for hungarian noticed that two files in your diffs included no right diff because they were wrongly marked in subversionAfter you svn update themthey should be right +And we will only call this method once now +I plan on committing in a couple of days +A sample +why is that? This wont work across any milestonestrunkrelease builds +There have been many MINA source code structure changes and want to try integrating your codeThanksMark +Perhaps there should also be a show unsolved on the threads would be useful if people would get into the habit of marking their threads perhaps there could also be moderators that could toggle the flag on the obvious cases +Cool! I like this version! Thanks Ted +Fixed. +The issue can be in +It is called +Making it patch just committed this +I agree that comment is a bit confusing +This works for me +Theres a meaningful difference in checking closing before we do the wait and attempted connection logic and waiting then checking close then trying to connectSo if we want to wait and then check the close before trying to connect vs the original way we can do that but lets make it a conscious decision rather than something that actually got pulled inThe second part of the change is actually more of a because Im pretty sure you cant change the state anywhere but the send thread so checking it after a sleep on that thread is pointless +Thanks MahadevI think adding hostname and processID to the set data works +Please take a look at the patchAlso Ive tested mvn clean installandmvn X B e clean install package rpmrpm python both work fine rpm is genereted +cool more code to get in mind XD thank methods have their respective specs in rubyspec and they pass fine. + following issue was found and fixed locking of temp file if uploading has been stopped by user + +Comments from JIRA bug Triage on Febrary Was discussed in another JIRA meeting and the problem with this patch is that it makes the interpretive processor depend on XSLTC +I looked into this and found no reason why the artifacts are not signed +I was able to run the new test both with and without your code change and got the expected results +On the good side in the first patch the get did an RPC at each validation which was somewhat bad so having it in HRS is more directReview please +Attaching patch for Robert +Closing since is released. +Ravi any progress on this? Thanks +needs to be sorted +Later we can still use those settings instead of the current indexing configuration +Yes I think that would be a good manual test +Youre right +I think this is a valid use case. +Messages continue to be sent when a peer leaves but they never get consumed +The design is simpler in terms of the model but the code sounds more complex +and spatialschema could just disappear +hold an array of visited nodes and check each new node against the visitedlist to prevent cycles +VladimirCould you please let me know why you assigned the bug to a verifier category? Thank you in advance +So does passes as well. +Changes merged to and branches. +Fixed in trunk +The issue is caused by there are bunches of type converter which can turn a byte into a Source +Fixed in nightly build nice catch and fix +Please ignore +List added and tested migration seems to have messed up this closing again +This fix will go into +Tested and closed! +this pull request Ill create an updated one tonight that adds the required methods into the as opposed to creating a new as discussed last works great! +tests includedThis is a modification to the shell wrapper scripts and thus isnt under the domain of unit testsManual testing is slightly tricky as on a well confgured cluster this wont have any particular impact +why would you need these resources anyways? I can only think that they could be needed to run unit tests +until this last line of data +This no longer the case other changes have probably taken care of it the row doesnt line up properly with the the columns and the text is wrappable but theres no error these residual issues should go into another bug. +Essentially the fix is this if a reference is the name of a method call then search for methods first before searching for properties and then fields of the same name +I wrote a note to the EG about this issue and will hopefully have an answer soon +In addition I have created and tested a subclass which implements an updated SIMPLEFLAVOR based on the latest recommendations +On the other hand if you just want to participate in resource handling mainly for operations then use PROPAGATIONSUPPORTS but without the Exception part +Tristan suggested opening the JIRA so I assume hes interested in some meaningful error message but could be quite a PITA to start covering all invalid configuration options for Hot Rod +My mistake I missed adding the link earlier and added only after seeing this JIRA +I think this issue has merits +I am running Railo final released Sep +Looks good to me thanks Uwe +If you still intend to work on this please reassign it to you +IOE is superseded by FSWE and RTE can stay the same +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues looks good Tomas I like this solution better than copying individual headers to the map +Verified +Nice patches! I applied them and was happy to find out that the site looks just fine! Great that we can get rid of redundant files now +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +This issue is still actualPlease see attached WSDL file +It should not be marked critical +this way we shall reflect that the new page Working with harmony code in Eclipse is not only for class libraries +Thats a really weird error +Do you have sample code showing Pig embedded in andor Pig written inThanks +Looks like misses test category +The new patch addresses Toms comment +The sample corba wsdl is not using a stander soap binding we can ignore that part of errorI committed the patch to fix the validation error +Patch for adding exportimport +Im still not sure that will resolve your problem though +There must be some combination of how many deploymentsredeployments I do or the type of application I am deploying for the AS to get stuck and not wish to shutdown or restart +Though embedded contains a wrong version of the jar so it is not passing closing this issue as jar version is covered in bugs me is that it was a hibernate issue but we test with jBPM +thanks! +Current problem is lack of isolation +this looks a bit odd to me that this one also takes care about creating the root node the uuid index and the setup of the version storage. +We can take the opportunity of having a Range class in the same package than for simplifying the implementation of the later. +Can we please enable the REST API for SBS? There are tons of use cases where we could use this data to help streamline processes in our projects +Just verified this works locally +Please disregard above comment +Similarly fs configs but only if we are not breaking any existing configs that are shipped for Write a simple test which tests down when RM inits +no reports of trouble in a long time +Addressing Vinays comments +Moving to CLI since it has been implemented there. +Im closing this ticket for now +I have some data mining background and although there is some reference to statistics I would not directly relate clustering to it imho +Closed +Trying to reproduce this scenario would be a good exercise +Hmmm +Fixed the problem with generic super types +Assigning to Davide since he developed the freemarker template REST sure this one might have been fixed already +I created a much nicer version of the example which actually works +Update to include of data to ConradAttachment is my change if it works for your case I will commit itThanks very muchFreeman +As I understand it though this will only work if DEBUG is enabled at compilation time as method parameter names are only available as debug symbols +Moving to EAPEWP +I can make the change you suggest and it is probably a better idea to make it more general just in case but my thinking is that this code for backward compatibility should be gone soon +Quickstart with in Wicket reset my quickstart project to use to try it out +But I was not able to reproduce this possibility using a Sonar analysis +Fixed in bdae in master +Thanks Amareshwari! +So what are we waiting for +It is really a major issue which can lead to downfall of Hibernate technology because it is big problem which not taken care of while imlementing HibernateThe behaviour of Hibernate is very crazy how can it replace my alias +I ran a simple test that shows this feature works +Is the problem link to adding data to a cluster with less nodes than RFBecause in my second test I could add new nodes despite having only nodes when I added the dataAnd I had the crash after calling nodetool loadbalance I had active nodes +True but the thanks was more for the hand holding +Unless the directory in question is actually a URL in which case we probably have a bug here +then formatting like this would be a lot easier +It would be very convenient if maven would be able to handle this without a for that handles the next version more carefully to avoid the described in the the next version information can not be determined the proposal for interactive mode is set to empty string +Fixed documentation that says semicolon instead comma +Not sure if description would be of much use in those scenarios +Maybe Ill do it Phase I think you have everything you need for now +Dimitris and talked and we dont recall exactly why it was reopened +Read the stack trace +Martin Snjezana had the following to say about testing thisI had to change the AS Runtime detector because of changes in +This removes the functionality from the admin UIThis patch appears to apply to both x and trunk +Hudson did more than test runs and did not fail so I hope this is really fixed by. +Adjusted tense in the final sentence of the release note +We hope to take this on at some point this year starting from the patch provided by are starting to see Rave be used as a modular application development strategy for smaller targeted apps where the need for low gadget load time outweighs the benefits of the iFrame. +Please move all files and folders from this directory direct to the archive root +Its a bit disconcerting having to wait in the order of minutes for new instances to be registeredChanging this JIRAs title to reflect the actual issue +Thanks for the review Stack and TedIs it time to rewrite the client yet +If some feature names are ambiguous we will change it +Ah OK I thought maybe the +We should also add debug options and ensure that they play nicely to +Thanks +Our clients schemas are more complex and they still do not validate properly but I will add another bug if I find a simple example. +It could be a timing issue and I will look into this +I will check the patch you posted at forum +A test to get the reported for the bug report! I am able to reproduce this using your test program +Thanks for noticingDo you have a patch +You can completely customize the way your artifact is build and replace the compiler with aspectj compiler +The FFI route would also help Rubinius and any other impls that present an FFI compatible interface +Trying to reproduce the first case +Now Im confused +meaning when i read it in i cant send it anywhere else as the subsequent serialization will output the type property as null making the next deserialization of that serialization failIf that is possible in I guess its resolved already. +As you can see from the FO the only difference between the two red square SVG drawings are that one uses the and height and obeys the and the other does not and therefore uses a default There is also a separate small issue with XGC which I will post in a related patch relating to it scaling images on that side +Is it documented? +I also moved the native protocol version constant from the relatively obscure location of to +This should be better addressed in RF development cycle +some more review is inline of pTLSvmthread and tmnsuspentenabledisableIts accumulative patch and include previous +Fixed in the SVNRS +Oops forgot that deals with the fix for trunk +Ill it if the fix lies within geotools and not jaitools +Added a test in c as well +Made the user tables page link to the table detail page +Bulk close of old resolved issues. +Not sure it solves being able to gets offsets from the query terms and still mask for positions thoughCan you explain that more +and application error log attachment using Derby debug jars +Can u suggest where I will find JAR file +If you have a patch for the build we could fix that as wellI dont yet but I plan on working on it later today +Clearly it would be very helpful if either Radim or Lenoid can reproduce with Jonathans patch but in the meantime we should probably start looking that way +sorry i did not notice Elis review above +This was one of the main reasons I was proposing the type modifier extensions to the IDL passing the buck along to the application writer +In will have a full fix +Patch applied thanks heshan for the contributionRegardsLahiru +Instead it is responsibility of the implementation to minimize the associated overhead +fyi i recently added a test to ensure that these properties are available. +If so it must be something else +Im sure he has some feedback on this as well +I guess what you are after is a way to add to the list rather than replace it +Why are you doing this +Could you confirm thats the case +There doesnt seem to be a good way of detecting this at from the data though +For performance reasons alone I would not like to send notification for each and every CAS update +Closed +Fixed it submitting +Then i will leave the class as it is and you can implement the required exception handling in the new class +Maybe this was intentional for a perf reason +bulk close of all resolved issues. +Committed to trunk +Added a test to the TCK and fixed in core +The hibernate plugin probably needs to focus on documenting what it can do at this moment not on providing a hook for an interface that exposes a Session +resetting severity from major to normal pending further P open bugs to P pending further document is clipped when rotated because IPD and BPD values are being set incorrectly in the class without taking into account the reference orientation value +I dont like the idea of supporting different rendering behavior based on properties +Patch updated accordingly +Fixed in latest bzr + for fixing content height and forcing scrollbar +fixed at +Neon encapsulates the whole request in one transaction while serf does gazillion of requests +I agree that scheduleReplication is a better name than compute +Whats the latest on this? Im in favor of fixing the existing one since there doesnt seem to be any reason for having the extra slash +Thanks Tom! +Tested with versions and +to trunk an branch. +Why the fix version change given that it was committed to +Does your test confirm the fix though? Maybe Im not reading it right +Changing it to major since there is a workaround +Resolving since I have fixed similar issue +I once tried different strategies to add support to the maxent package doing this just in different branches on a platform like github would have been niceNext thing where I will use it for are hash features instead of our string features +A regression test for this issue will be added in. +the issue is fixed by the commit +Assigned toThese items have all been ported into the WIKI +Switching traffic to the US would also this issue so Ill start that process now +Sorry for the late response +Yes if I just revert the fix of your test starts going through +what about python? assume the user to have it handy on local diskdo you think this will work assume we have python ready +Pritam I think that line an old one and its purpose is yes a check the table can be reached +Thanks SaschaYour patch is in trunk at r +It is unlikely an authorized user will DDOS the job tracker +I think similar change need to be made to +If you are not observing the same effect then please submit a reproducible test case for us to follow up this issue +Hi BryanAll are perfectly ran on my environmentThanks Bryan thanks for everythingI am closing this issue. +I fixed error by skipping metadata generation for unreachable plugins but it means coverage report shows results higher then they really are. +Id gotten two clean hadoopqa builds on v +Distributed search uses for internal communication so something must be fishyIt is not clear to me how the tests are failing just that they are +update and accordingly +great work! Ive verified mvn clean install +That means the XML file parser should be changed to fix this bug in the first place not only the XML files. +Looks fine to me please feel free to commit it +Patches going into branch need review +I have now linked the issues. +I believe Rene made some changes lately to deal with this exact kind of issues but Im not sure if they are already releasedCan you try a nightly and see if that solves your problem +I just committed this +planning to work on that +Thanks Carl for updating the patch Lets check this in to unblock RC so folks can start playing with RC we can always if need be later +But we need JDT model which is provided by Eclipse Java builder +This part is tricky +we keep this resource load stuffwont Mikhail GrushinskiyI really appreciate your efforts for the trinidad podling but issue like this should be discussed in the DEV list first +I believe were initially done with what we want to support for +I was wrongIm seems good to me +The older clientgen task was removed because it is not compatible with +See +I am running on OS X Eclipse +Its up to the app how to respond to an interrupt +So Ill try to get this as RAC running +dont waste time yet It appears to be my fault +We also plan to make test the default behavior which seems to fix the problem +bulk move of all unresolved issues from to +StackCould you see this issue with latest patch againThanks +We can resolve first +Let me know +at best by a then committed thanks +I also modified a code to reuse builder to avoid code duplication +if the namenode is configured as somehost and I have somehost mapped to myalias in my etchosts file I wont be allowed to connect thru myaliasYou might be affected HDFS issue the Doug mentioned above +Nested sub processes are supported now +If user want to use an external hadoop heshe need to override the default hadoop version explicitly eg Dhadoopversion in command line +Important for this bug to occur is the indirect access to the in order the specify the type of the params +if not add investigate if default labels have their own rf classes +There are no plans to fix the RF photoalbum demo project to allow it to be imported into the IDE +Assuming it is okay +For instance you have set it to true and then set it to false afterwards but we already checked the boolean and wrapped the Directory in a which might be confusing +Committed to CVS will be included in release in release +but there you title +None of that or any of the rest of the text in the policy draft relates to where category B is placed +Tried the and it works as expected +Forgot to mention the same happens when using capybara instead of watir +Martha please try and build this one +Resolving the issue since the patch is already applied +it seems to me that it is difficult to create a UDP transport for Axis +I havent run into this particular issue on Android on the Galaxy Nexus nor on the Nexus S so far +addition to the tests I dont use schema and commented some tests with in +Thank you. +was fixed in trunk +What I am not yet sure about is if we need to have an SMTP host per sender +I seeThanks a lot. +I really dont like this I like fluid resizing layouts +The reference project now has dependency to + Erics +Will update on afterwards +The problems are due to the hacky implementation of handleA proper implementation of this function cant just skip the +This feature has been implemented inPete if its still doesnt work then please reopen it as a bug with step to reproduce the issue +I was renaming segments quite often so I would vote for reading the date from the segment instead of using dir name +Oops heres the correct patch +Assigning issue to Nadav +Is also seeing this error and have two possible solutions +Should have read the JIRA more closely +results in an exception if the variable is not available +I tested this with creating a web application in eclipse without geronimo runtime and found a similar behaviour as has been reported in this JIRA +dead code removedoptional code all in osgi but still currently started by ActivatorIll move this optional completely out of the runtime bundle later +in order which scheduler looks at it +Test application +I agree +One would not have to update an mbean for new metrics and also the metrics would be published automatically to themetrics context +Please reopen if you continue to have problems. +Yes I think order is quite important for this command so Ill be sorting the expanded dirs to ensure deterministic behavior +this patch modify all Attributes created by provider to ignore caseand move convert method from to +I think what you are seeing in correct behaviourA duplicated key results in the current statement failing not the complete transaction +code update for compiling against the new API upgrade the descriptor parser to handle xsd writing more test to validate features +r reverted the build back to using a downloaded ecj +feel free to let me know if there any feedbacks or things to be taken care in this regardThanks +The screenshot demonstrating leaked resources for deleted temporary queues +This is the quickstart project that I have modified to show this bug +No information. +It is very unlikely to be a thread safety issue +For example one random client renames one of the top directories +All schedulers are supposed to enforce the admin value and not determine it for themselves +Frequent Pattern mining has been deprecated in Mahout and removed from Mahout hence resolving this issue as Wont Fix. +Updated with a minor fix needed as you need to use the updated attached and components +Hi Ben sorry for the wait for reviewing this patch +Using a snippet like for would be much betterIf all is ok change bug status to VERIFIED else sumbit your commentsThank you this is a very important step for CocoonD +Assume closed as resolved and released + +Bulk closing stale resolved issues +Fix confirmed to be ok +Some tweaks will be needed afterwards such as discarding some Externalizer implementations for Externalizable implementations and enabling externalizing of configuration of magic numbers externalizers +Yes if you want to predefine a collection you do it in the same way that collection is done +Same issue manifests on Linuxx +is not part of the testsuiteI think that situation when we will need to support whole build on machine not supporting is far far away from now +Found it going to test on cluster now and upload a new patch +Resolving the issue with Amilas patchThanks Amila for the contribution +bq +GabeIsnt the method that assigns synchronizedHow about a unit test that demonstrates thisCheersChrisSent from my iPhone +Well I was working on this around midnight +Class can load the test dataPlease tell me if you still need an isolated testcase I can prepare one tomorrow +Test failure is related +Specifically if you think its +The bug is actually in interpreter + +Ran local build without errors +A users profile wont be considered complete until they have chosen a default availability zone. +this is rather complicated because the entire pagination seems to rely on JavascriptI dont like the search at all as it currently is it is full of complicated Java JSP scriptlets tags and Javascript with nonobvious interdependencies +fixed +Thanks for the patch Dave +However its not an entirely satisfactory behaviour you probably requested for a reason if you didnt care you would have specified orI think the reason for the current behaviour is that the serializer is only able to handle a single contiguous range of Unicode characters when it decides which characters can be represented in the requested encoding +Should we also check permissions for owner and group? Comments in create should be dirsubDir dirfile Do we need to give rwx access to group? Setting the permissions for user should be sufficient I think +We are planning to put it on latest in the next couple of days +Once I do Ill commit +Thanks Raja! +Fun stuff +Closing resolved issues. +files +The patch is not very different from the first patch +Exposed the repository statistics framework in our public API so now the enumerations and interfaces for accessing the statistics are in the module +xml files after running the generator +Lukas why is this issue duplicating the linked RFPL issue +Im submitting this patch with my latest changes for the UTF support on the client +thx for clarifying +In interpreted mode is just walking an AST made up of objects loaded from the same classloader as proper +Patch to for initial problemThis has been tested using Adrians WSDL with response and real server but not for any other RPC services. +Applied the patch into trunk and branches. +The key attribute used to be initialized as you suggested but this causes the hierarchy of objects to become +Indeed for marker type of annotations what I tend to do is only call the marker once +I just noticed that the comparator in the patch is bad! Sorry +Tested and committed to all major branches. +I will attach that anyways but not sure if that should be tied to this JIRA issue +Closed upon successful release of. +At a high level I like the idea of compiling the trigger action along with the firing statement if only because it gets us out of the business of asking people to drop and recreate their triggers when we correct problems like this +Generic language codes cs ja da ar vi need significant work but there are many strings already there so they probably should stay to ease the translation work if anyone ever wants to do itCheersPaul Attached +Contrib test failed is because ofAll other tests passed +called the correct method +I agree that infinite recursions are evil but if you want to ask forrecursive traversals +But rather than take the performance hit of synchronising thischunk of code it would be simpler and faster to just use a local variableinside of the load method until all the parameter classes havebeen loaded and then assign that to +Resolving the issue. +Fixed in r. +It sounds good +It appears that this is happening hereI think changes other than that should be another issue a or a linked issue? As it stands Roberts patch having the same name as Simons makes it appear that it supersedes the prior with the same name +I hope that helpsThanks for simplifying the to make our lives easier +Change the issue so we remove from the distro rather than add in trunk as well. +They should be able to put get delete etc +Verified. +Oh then fixing them is really easyIll create a patch +being able to read the logs knowing the location +This is not a difference I it should be moved to DRLVM component +Thank you Steve. +We dont test on so we need someone from the crowd to help out here +Other than that patch looks good +sorry too late need reverts? or go ahead since i clobbered it already +To be fixed for +Fix in place +resolved as discussed with Greg and others +the XID would have this integer in its bqual therefore are there possibly tools that would want to read the bqual and need to read meaningful information regarding the EIS without the benefit of our utility classes +It is not necessary to add some special SELECT processing +So it was expected a but what type of object did it actually get? Also would you be able to track down the actual section of the PDF which is causing this? It may be a parsing issue which caused the data to be loaded incorrectly +This causes workspace problems after importing blueprint projects into workspace +Instead of creating the bundles with the jar plugin ive switched to use an ant task for this this avoids adding the artifact to the project artifactsChanged in revision ve done similar changes to the osgi installer project which also created a dummy test bundle +I agree its a good idea to make the assert more common +Attached an updated patchI decided to use which is supposed to be more stable thanThe patch does not yet include the unit tests from the other patch but if this patch is accepted I will merge them together +Ive also been able to successfully deploy both EJB and EJB jars +Julians suggestion workaround helps indeed +Otherwise its up to the application to decide if sufficient benefit is derived to offset the risk of the added heap pressure +Move open issues to CR +Verified +I agree I just committed an entry to +debianinit script to be able to run it from the shell against the cluster started by CCM +Since there were no XA transactions there was not the SYNCCTL with possible missing SYNCCRD exchanges that I was looking forYour logic on the above change sounds reasonable to me but I dont think I am the best person to review that change +Updated to include the proper Apache license header +Hopefully it doesnt break this is ONLY at FUSE +Updated see Phabricator +This is a blocker bug for World Singles +This allows users to navigate around the project pages +Im curious about how you gathered the coverage data? That would a valuable contribution on its own +New patch removes tabs +LOB Streaming +Thanks for clarification +Slight improvement to your designIf a new server knows that it is compatible with an older client then it should accept the clients versionIf the client is new then the decision can only be made on the client side +Committed revision +Moving this to RC we may even have to push back farther +The DWPT that happens to exceed the first tier is flushed out +I cycled k times +Moved issue from JBCLUSTER to JBAS +Lets get this in being able to do this is important for the first user flow +Committed thanks +Hoa if youre using warbler this is probably because your +Only solution I know of is to log into a file +Attached is the patch rebased on trunk with some minor modifications +As part of that Id also like to changethe deploy story over to using capistrano +Currently only map join operators will need them +Cool +Yes this is the latest version of this patch +We seem to be running +Well Its not the best decision +The fact that it has votes after years and no action is very concerning +Please reopen if necessary +Since one can store anything in the job conf we should be careful with the references to this object we should not hold for long durationAt most these references will be held for the lifetime of a task attempt right? So not so long +Im pretty sure the merge I just did will fix the problem you ran in toWould you be able to build either trunk or branches and see if the fix works +committed thanks! +It now inserts a row with the rows default valuesCan someone please review this patchI have not updated the canon for under j andI have successfully run derbyall +Also if a user uses heshe shouldnt even know about? In a way should explicitly compose and redirect accordingly +Fixed log messages +Ideally wehadoop should be moving towards a client being able to access with multiple clusters with a single configHas a meetup time at the summit been organized? Ive scanned my mountain of email but didnt see one +thanks laxman + to trunk. +You can also attach the patch for Thank you in revision +Based upon my work to add support for polymorphism I think that this code really is not used at all and hence shouldcould be removed +See if you patch fails second or third time if it does add debug to your patch to help figure whats going on? Thanks +applied with modifications thanks +You could just as easily do an eachByte on a File or URL in which case no encoding would be involved +Testing this patch presently +Persistent +thulasiramThis is a common problem that you need to debug +Right now the translations were only committed on trunk +looks good +Its possible that the use cases you mention are not passing in the attribute +This bug is not abolsute critical but needs to be fixed +You can first go with a contrib folder or a rd party lib and merge to core hive later if it proves success +Done. +Committed trunk and branch +Thanks for the question +To avoid more svn work I think committing soon is better then later +The test code ran perfect for me as did everything else in the patch +Easy examples UFS ZFS WAFL +Moved to +Adding patch +Its just a warning message everything should work +No tests submitted +Fixed inWill integrate into if everyone agrees +Also we could add those IPC threads to the thread leak filters if theyre harmless instead of doing +Simple enough patch +The remaining two test failures look unrelated to the patch +If the current unit testing strategy is felt to be insufficient a new issue can be opened for M +Sorry this has been languishing Andrew +im fine with making all the actual properties protected +Datanode adds block pool and corresponding to when registrationfor corresponding block pool completes + +The similarity with the API is dont think pooling should need to be explicitly turned off for this particular use case because the Java LDAP provider keeps separate pools based on their principals +For example the bounds check is inherently less efficient than using the proper unsigned comparision like we did in the generated inline sequence +Documented as a known issue in the EAP release notes +Ive filed to track that. +The issues are now closed. +We should make sure final still works +Its less about typing and more about readability +Thanks for the quick response Army +Linking to a request to control log levels on debug page how many possibilities there are here Im seeing the need for a separate component in JIRA to track debug information screens. +Activity should probably mention only the embedded example and the utility class unless you want to modify Activity tooActivity tells people to copy the embedded program to and make the two changes that turn it into a client example +It does not alter anything +Please report it to Eclipse bugzilla +Look fine to me! Its indeed quite simple +Same problem for me +Currently only Gridmix uses Rumen but in future Mumak will also be using it +Current problem is leaking of resources from appcl down to deployment so +My argument is on making transfer of headers etc mandatory +Jesse is working on this not sure if hes done yet or not + +It worked +For commitersPlease note Erikss remarks about files names +Any word on a fix? Im feeling less than motivated to rewrite using HQL string processing +This issue appears to be fixed since at least the version of the archetype reopen with fresher data if this issue is still a problem. +Im seeing that a script that flushes all tables on the cluster does not bring on a log cleanup when expectation is that it should +However we dont need to add a get method in that patch itself +Im part way through a review but have to leave +Java sources and DDL I think this should be in a new maven component and preferably with unittests using something like Apache Derby which we can easily include with the unittestsA couple of questions Is there anything specific in there? I note a couple of WSO copyrights but the package structure is currently +released. +What we should do is to prevent forever caching and just use a soft hash map instead +might be good to keep in mind for this +Ive implement this feature about two weeks ago +No feedback so closing it as incomplete +Youre right moved to +I cleaned up servers data log tmp and work folders and started fresh +Also attaching repro scriptWithout patch you should see an error trying to register an twice and the first index will no longer workAfter patch you should see an IRE. +very much Knut +Ill rustle up a patch to fix this one +My original intent was to subclass from to avoid the pushMetric hack but the implementation on used introspection and would require a complete reimplementation of the class +Used for testing surefire junit private version +Anyhoo it works fine on so I dont think the triggersp is doing anything too strange +Fix confirmed thanksColm +Ive send a mail to about this +I couldnt find the classes in the attachment +Upstream indicates this has been solved in RC as have other similar issues +The javadocs say cal the Calendar object the driver will use to construct the timestamp +Perhaps that is part of why its commented out as well I dont rememberThe above tries to just make sure the writer has been initd rather than forcing a new one +Galder Im taking this since it is a consequence of my configuration overhaul and there are still improvements to I just imported your test into my working branch for and all methods pass +Very good question +Applied the Patch +Will commit if tests pass +Verify key signatureis still missing actual keystoreThe line should read keytool list keystore JAVAHOMEjrelibsecurity the keystore path in the procedure +sure let me commit today or trunk +Mybe this brales things internally and a systematic check has to be added on each roo startup? changing the name of the entity in my real project didnt solve the problem but nothing garantees theres no other forbidden keyword in the sources +See discussions on interfaces vs abstract classes in and +Sounds reasonable Allen can you add tests +Verified +Please fix it +Thanks Brett for working on this problem +Add in in Update to extend abstract Add in in Rename to to updateIts taking a longer then I hoped but I just wanted to let you know Ive got the code compiling and passing all the existing unit tests +Configurationsetup needs to be updated for parallel deployment of +I will email the project is still pinned to and so it will remain until this bug is fixed +Assigning to but Ill have it done this week. +this patch fixes the issue +Move test case into the test codeThis looks like it might work for allowing reuse of the test cases but is slightly weird +Updated title to be more clear this is an improvement request +I did not know that I can convert to DocumentIs there any page already for the convert? I did not find one +Attaching fix as far as checking if we should finish an inProgress round we only need to keep the most recent inProgress commit that has a value +I tested Hadoop examples on the fully distributed secure cluster with +And the test vcase from the second JAR is actually included but named +Thanks for the code review +As far as I can tell the root issue is that the errors with the ivy configuration should be raised by the Ivy resolve job not the code that creates that job +Yup +QA Verified. +I threw in the simliar handling for reset just for completeness I totally forgot about this bug So after I create an example for ExamplesCookbook this issue will have to be closed +Well gotcha! but you took the wrong file +Polite to see if anything else is required from our side +Committed patch to trunk revision +Issue with adding another method for this +I cannot dedicate the time to fix this issue right now +Ok this patch ensures that the output directory is set and does not existIf the application wants to clobber old data they need to delete the files themselvesI added the check for the output directory being set because otherwise the job doesnt fail untilthe reduces try to run +for the review Stack committed to trunk. +No answer +Triaging + for the patch +Renamed blockArray +and since m is ruling the world right now we are forced to work with their repos +Passing the row will be present twice both as the passed in key and as a field of +Im having some problems running the unit tests on trunk as well +John Ashish Please tell me how you want me to change this patch +Could you please take the burden of internationalizeTIAJacques +Need to insure that the is configured before we start accessing or creating databases +stack can we close this issue +now adb supports Choice +This is using in Seam with the fix you recommended above. +Case has affected measurements on +Hi Tiago I think a static variable and single initialization for basePort would be good but it should be in +Im just suggesting that the part that comes before the qualified name is unnecessaryI understand that changing this would necessitate adding a new key type and transform logic but Im not convinced that the translation would actually cost that much more time +This is fixed withNow Woden class is package as a separate jar file this will enable people to choose among Woden or Sun Java SDK implementation +packaging should be optional since in my case I would to upload zip files my internal remote has no official packaging can create a dummy deploy project for upload purpose and therefore groupId artifactId id and url become left are Dversion whole idea is to minimize number arguments i need to enter on command line +Patch applied and this will go out with +I think this is a blocker because for an unsecured environment anyone can register as application master and start requesting resources +fixed in per moment I closed this I realized that the patch may be useful for the ROP superclass +The wrapper shutdown seems to work great now in all the tests ive done +Fixed +The new section should appear in the table of contents so that it will be easy to locate +It should be possible to flag a manual issue as False Positive +with some minor changesI moved some files around because it seems doing an ant test under srcplugin returned an error before +Closing. +Max please review the ve merged into hibernate tools core +I definitely think these are some things that the interface should expose but unfortunately does not +the guidelines is that usually the patch is geenrated from the trunk so the reviewer does not have to worry abt p or p +If your entire cluster is dead then yes you wont be able to access the data inside of it but thats only one extreme case under which its not useful +Releases in. +It is fairly simple and I have done this type of thing in the past +Because I realised that I was wrong and this is already supported +updated to handle can you explain the use for this over normal? I dont see why pretty printing xdoc is important +Yes agreed. +Happy +Shades have gone now what you call regular features still exist at the moment but theyre not used by this +This Patch will ensure that Buyer should be able to specify PO otherwise it should be generated automatically while placing a purchase orderTo test this patch follow these steps +On another note can you add DB descriptor parts for other databases? I think they were present in old installation guide that I pointed on my previous comment +Another gem which demonstrates this problem is rspec +Patch Will commit soon +Removed rvm and reinstalled from scratch +No server logs for this cause I am using the default logj configI cant restart the server now because another bugMy guess is it may be some concurrent problems when deleting all no prove eitherThe error node is still in my datahow can I provide more information +Updated patch should fix findbugs warning +More investigation needed +The request must be discarded when the result has been sent to the client not when the session diesNow is it necesary to store this request? What is this good for? If we have a very long request like a search through all the base then we will have three cases there is a time limit we returns the data we found at this point +I plan to commit it soon +I ran derbyall against insane jars with ibm on Red Hat Linux and there were no new failures +I just downloaded Xeres for Mac OS and got the same error during compilation +It can be invoked by any other mojo in the plugin too +Thanks Tom! I will prepare the SGA today and send to secretary +Ive added a test for which adds more that one feature within a and works fine so it seems that the problem has been solved. +located in. +I will file jira on that +I recommend everyone to take a look into his code +Upping priority to critical as its a regression compared to patch adds a catalog reload button to the status page +The bugzilla reference has been updated about the failure in WSRP which caused the fileUpload component to fail +I was just thinking that all these rules could be configured at the same place and that the text should be justified accordingly +bulk close of all resolved issues. +This issue was introduced in the process of getting a stress test to pass +Suggested changes +Weakens my case based on consistency but is marked final which Id also prefer not to be +I would appreciate it if it could be scheduled so that we have more or less an idea of how long it is going to remain a! +So the first task should be checking that turning it off actually undoes all of its effectsSeems like I think should for this +Thank you MyrnaIve added a patch that makes the apporipriate modifications to javatesting documenting that jdk has been added +bin unload regionserver but that gets messy in bash +We could add a portable plugin sure! Are there sufficient functions available as part of the JVM itself for this +Also the doc id is also guaranteed to be the second item +Could you give a little more information than this? What are the errors? Do you have a branch I can look at +Thanks MariettaChatra +Fixed for tonights ve confirmed that is fixed against +still in tooll need to login in order to see the Accept or Agree button +Could you just attach your +Looking at your screencast it looks like the clone happened so Im guessing that the upcoming phase with the ahngsfails +By the way couldnt reproduce this problem at my end +Please check out the trunk and test if my fix works +My understanding is that uses native libraries via JNA and it shouldnt be dependent on PATH +Ryan Was it easy to repro? And w this change it goes away +As I said in my previous comment please create another jira if further changes are needed +That being said it wont be out in time for M so pushing to got released should we try putting it in for m ill try some custom builds to see how small we can get would rather wait for M since M is so close +Provide any easy to use tool to merge contents from two Contacts in system +Thanks Steve for chiming in +I would need to change lots in the build system +Script for renaming INKMD to for consistency with the remaining of API forward as this isnt no longer exists +After a restart of JBDS this context menu item is back to Mark as Deployable +When you run gwt setup Roo scaffolds a complete web application for you and STS recognises this and enables that option +Instead cant we make interface as Abstract and remove these duplicated classes in and? If we remove the duplicated classes then we could put file specific features such to to a directory +However we need to fix our docs which presently state that objects can be shared +fine to have the default in thrift if that simplifies things imo +This bug has been marked as a duplicate of +More Text is not fully right aligned +the solr node may die if the election time over the solr retry time +If this issue is same as that please resolve this JIRA as duplicate of +I will try to play with a bit of a spike seeing how it might work with easyb which typically has extensions like +The dependency plugin only does what you ask of it and analyzing the contents of the jars is outside the scope +I can reproduce this issue on +Tho wrapping including the handler would be a more conservative approach +Let us know if you need some help or consultation about plugin development +push out to +Checking this in +As when we useListfalse we do not know if the file is there or not +Those dont work at all yet +This BPEL works +Since work has been done on this please resolve before and open a new issue if there is anything further you want done +Thanks Jukka! Thats a good idea to move the class out Ill do thatIll mull how to assert the sax startend elements are valid +Committed to trunk. +Im justguessing that this probably shouldnt be larger than about k but this isimplementation specific +Also the team calendar indicates that will be in December and has no date attached to it so I assume that is for next year +i plan to backport once I verify that tests passed across all nightlys on trunk and I run tests on +Actually I am now able to replicate your me some time to figure that I forgot to enable an option in the builder configuration +dig noall answer t txt +this patch updates the version in the template pom in case some one else wants to commit it +Ive implemented a patch for this issue but it requires my patch from to be applied firstRegards on r well done! + for immediate inclusion IMOI agree in principle that theres no need to iterate over things that arent loaded as they clearly cant have changed +The issue is reproducible now on Windows and on DRLVM. +This issue is reported against a version which is no longer maintainedwhich means that it will not receive a fix +bq +Jacopo I was thinking that is neededused only in order creation routinnes and once the order is created the production run gets the configuration and issues the needed components for this configuration from inventory +Ill change the build config and see if anything else is neededThanks +seq is optional to preserve compatibility +This bug was opened against +Moved to CR +Most likely you just need to make adjustments to your URL so it fits your specific Orcale URL is wrong when generated using persistence +Ill have a look at this later +Guys whats the timeline for the next release? +as part of revision closing stale resolved issues +Can you try remapping the type from SMALLINT to INTEGER +Added a buffer. +Maybe in this situation the conflict manager should resolve the conflict as if they were not forced +Somehow or other the namespace snuck in there as the package name +It should work now but it of cause needs to be tested on other browsers to make sure that we havent introduced some new weirdness. +Whats this about +failure is tracked atI am pushing this in +The Hudson specific directory should not be in the code +backported to +It helped. +Is that on purpose? I suppose I cant just emit a key in my view for deleted docs +We never got more information and were not able to reproduce this so Im going to mark it as such +What we really should do however is give the option to also include the fidprimary key as an attribute right now there is no control over that +If this doesnt have everything required please reopen +A fix is in SVN. +and maybe we should merge and +A ton of other factors interfere with this kind of predictability +Also added a new test case to exercise this code path +Bernd did you change something in the calculationI think the margin is missing on layoutcalculation in the layout managerAssign this to me if i should take a deeper look into +Although the spec makes no direct claims to either support for refute the challenge there is strong secondary evidence that the EG intended it to work in the way the TCK enforces +can you try to do this with the classes provided in? The +Committed patch as svn Will hold off resolving till I hear the results fro the IBM +Apache AxisCApache RampartCApache SandeshaCApache Savan Apache SavanCI left some of them blank and some of them with there Java version logo for the time being +I had to look up the implicit field usage +Theres nothing stopping a show from returning JSON and for a single doc running it on every request wont be significant performance impact +BTW we normally using the JIRA to track the issues of codeyou can add some comment on the wiki page instead of fill a JIRA for the doc +I eventually went with a new acceptExisting flag allows for sharing per cache on including controlled shutdown by the that actually created the instanceThe shared mode still exists for compatibility with code that uses s old access style +Moving to if you need it sooner let us know. +Hi Aaron I choose xdoc since apt has some limitation + is not thread safeits not cacheable +Thanks Christophe! +Patch fixing the above errors +So the display would be first ordered by the class name with section headings then each section could be the list of counter values from that enumeration class +I would like to know the same thing +Hope this answers your question about Dag +Verified by Mark. +Marking this as fixed +As for complexity I dont see how adding one field to the response that may be ignored is adding much in the way of complexity but then Im not doing the work soAll the rest is supported OOB so the are realizable even now without adding the name of the component +Same problem with in trunk. +Then leave it as it is. +Ill have a look this evening to see if I can get the fix in +Creating separate issue Transaction integrity checks which would deal with the issue that current transaction creation services are not checking whether accounts balance +Thanks Naryanan +It is sort of on the borderline +Patch to add new z option for understand this is old code and low priority but if anyone else happens to need this patch it is available . +Still waiting on PR review of trivial fix +Thanks file contains a patch for second scenario +Committed almost as is. +committed this. +Im afraid we cant use this patch as it is licensed under which I believe to be incompatible with Hibernates LGPL licenseAs long is this is not clarified I wont look any further into it +META +it has the fixes from meLooks like your changes work fine! Thanks! +on +closing since is now released +fix verified in my app +Committed +bq +This is a simple improvement but Im taking advantage of this JIRA to do some code refactoring to improve the structure of the +I can fix the issue at +I think thats by design +wed act on it without waiting to see if any options for port or address +Ugh +Sites updates in SVN and +Tangring confirmed on th Jan in IRC conversation on that Documentation should be the No +We should get approval andor inform interested parties that some additional tests may be advisable +made it so that doesnt always set the number of map tasks to +Patch that should correct the issues noted in this bug +Also added units to some of the headers +I am working on the fix +Manish thanks for your timeAnswers to your commentsI like your name suggestion because it is shorter but one thing to note is that the existing do have collation associated with them its just that it happens to be the default collation +Actually for mark occurrences there was a bit more of an issue since the editor was considering the to be part of the variable name which was messing things up and I fixed this as wellFixed with regression tests +The pdf uses type fonts which are not supported fine after resolvingThanks for the report! +Am bearing with you +Anymore to be done on this ticket +So the fix is actually to avoid throwing no method exception when initialize the for those Classes without no argument constructor such as primitive wrapper classes +If youre using or later never face this issue +So you dont have to duplicate implementations for all the value types +JarekI just pulled down the latest build of snapshot +Andrian please try it on next build +I have unblocked those addressesBe sure please to use dump svnsync for CI services at your end or if there are lots of going on use mirrors instead please +Thanks Olegs comments +Just committed this to trunk and +Thanks again for all your help and sorry for the trouble Ive caused. +NET WCF s a specification for the configuration service as I envisioned it +Dirichlet and minhash take forever so havent tested those yet +For windows please retest with +Maybe also include the pointerreminder about Window Preferences Spring Roo Support in the STS Dashboard Updates or upgrades ill probably remember but for newbies to Eclipse it is not obvious + for removing it if it is not used +Note that these are scoped only to the AM level and have no bearing on the YARN platform level access controls +Freddy this happens because marker for Violation wasnt refreshed so in editor we dont know that review was created +Let me know if theres any issues +Hi ScottDo you have more concern on this or this patch looks fine to you so that it can be part of trunk +The second one allows you to specify a setting for an individual workspace +It does not notify when FIDS are updated +All comments are the same as those on patch v +Im having the same problem that Luis Arias wasis having using AWS libraries +QA Verified. +Thanks for your advice +The order of the messages doesnt really matter just that an error is encountered when trying this particular operation +Trivial patch +the patch works for me txs +Test passed on all tested platforms +Hi Shameera I am sorry I missed to apply this patch and proceed with others and it now conflicts +reopening for some polishing by in IP module. +Please review the upstream for QE to verfiy in the +Patch including changes to CHANGELOG and but still without testcase +Great thanks Willem +You might want to use packet sniffer such as Wireshark to see hat kind of packets get exchanged between the site and the browser +Seems related +I seem to be missing something here +Good thought Tom +Publicevolving should documented but must be as evolving +Ken Wilson provided the UI design for the portal. +Adding a remote repository now IMHO from a UI perspective it is not obvious that the Add repository link is intended to add both managed and remote repositories for it is located on the same line as Managed repositories label +Is is happening on the main build server? Ive never seen it ran into the same failure and it is intermittent +Steps are not reproducable anymoreVerified on Developer StudioVersion Build id Build date +Go for it +this corner case is not possible with our current client implementation +This is blocked by right +We can close this issue +A few related issues arent closed but the basic functionality is there and working +Added doc +I dont know why you are seeing this behavior +Patch is the same as the below +Thats very good news in that case Im fine with the Gora API as it is now we should change Nutch to make use of this functionality +Final patch for review +Travis can you check out Marks patch and see if it is doable for +The issue with this patch is that if any method is marked as flaky the whole test class is ignored in run although run works fine +Compiled ant docs and verified HTML docs +We can fix the infinite loop as part of that as well +I do not see much of its benefitI am thinking to allow users to specify replica placement policy at runtime when it sets the replication factor of a file +So I downloaded the newer version and installed it +Why did we need to narrow the return value from toWe have no Bits interface in +the timeout of iogetevents is so it is never blocked +Thanks MahadevIve incorporated your feedback and also finished up the whole doc +Can you describe what it is you think would be valuable not only in terms of content but also in terms of layout? That would be really helpful +The problem is that there is no way for the callback javascript code to know where the use wants to add the action errors +Anybody wants to comment on that +Sorry it took me a while +Thanks for pointing this out! Were setting the content type now and also fully support Hessian as of Spring +The update patch for the hibernate feature +I need a mergedI am using localsolr andAny pointers +Verified in +This have been fixed in EAP branchAlessio will decide what is the EAPrelease number that will incorporate this bugfix +Verified fixed at svn revision +How exactly does that error arise? shouldnt require a so passing in ay should be fine +EduardoIve moved the final banner into the DMS system and changed the link on your site +Patch that implements the fix +Submitted patch for If this is already merged can you change status to resolved +Implemented by +We can and should run them more frequently +This template is garbled under jruby but comes out fine under and p BigeI debugged it to the handling but seeing as though behaves differently than perhaps there is something else at work here as wellIf it would be helpful I can package up a sample Rails app with a spec that demonstrates the issue +I guess the engine thingy was broken by the patch +Restored default prefix mappings tests break otherwise looking for a prefixIt seems there are still tests of DAML stuff in Jena in +I just committed a fix in trunk +New patch to implement those suggestions + +Also we properly validate the isXYZ returning Boolean +I also think we do not use the author tag +The portlet contains a portletState interceptor which is supposed to be the very first in the stack so you should consider either moving your custom interceptor to the bottom or create your own stack based on the portlet +Patch applied in r. +You just need to open it with tenant identifier if has been set +in the last patch the file is missing +With nodes if even get bogged down simultaneously then the jobtracker would run out of handler threads +verified on latest build +but I really need to force them to do it +Please provide a list of Confluence usernames for all those needing to be added to a group and another list out of those who will also want rights for the spaceNote that those without a confluence acct will need to create one +Actually new test shows none of the membership rules is enforcedTestunitsorgapachebeehivecontrolstestjavapropertyRepro run ant from controlstest +committed to branch +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +my only concern with this is that it should be done in a secure manner or the user should be able to make it secure think it should be fine as long as we make sure its well documented that its not being sent in a secure manner +Fabrice right now all this does not make any sens to meI need you to explain me what is it all about +Its against trunk +Thanks thats solved the problemThere are some dead build results that cannot be removed but we can live with that. +Currently there is no way to dictate the firing of a scheduled task based on if a transitive bean is initialized or not. +That constructor and those methods are there as an escape hatch to allow client code to access that Thrift hasnt abstracted yetIs there a different part of the patch that deals with anonymous unix pipes that didnt make it here? I could see how supporting anonymous pipes in the future would motivate the ptrdifft changes +NB that there are other ways to get out of sync during an update operation most notably by having timers fire at an inopportune time +Passes full set of tests on windows and linux +As I saw in test cases testcase calling createSnapshot method with argument testsnapshot +Loading a partitioned table without specifying partitions was being validated only if OVERWRITE was specified +Thanks for the patch Harlan +Push againAttempting for perfection leads to never releasingPThat said I also understand the frustration of breaking code for bureaucratic reasons and it doesnt feel good +The UID you store should NOT be based on some list order solution it needs to be based on an ID stored in a map because otherwise if the backing list order changes the converter fails +Thanks Dhruba. +No longer detects for Authentication +What I do for testing is creating some virtual machines and run TT DN inside +It needs some annotation and javadoc +There is more work so that we can come up with good design for auto compactionIf my patch passes review I plan to modify the subject of this JIRA and open new one to continue with auto compaction +Please test it and tell me if it works for you +Because I dont know where the new debug messages were added Im not always sure what they represents and what to make of those logs +May not need to do prepared statements but should be sure that one cant do sql injections +That is the diff file showing the differences between the old version and the one i patched +Yup Tomcat follows the spec +Committed fix in trunk in revision +Service deployment support for most options available on +Close the ticket if you dont have more planned changes. +All unexpected enum values turn into null +like you can see on line of that the lock open on line is closed in any caseso the only explanation that a thread fails to acquire the a lock on line is that a other thread is still between line and we have a lot of testcases for exactly this s of threads calling the same page where no application context is definedperhaps it has something to do with the file upload you are writing about i will run some test with thisdo you have any other exceptions in the log timeouts exception in the +Just wondering because I didnt see it in the docs and not much help on the forums. +I am committed to doing it but that doesnt mean that I can commit to a timeframe +Need to report progress for new one added the is a problem in the checked in results seem OK but I am getting a diffCan you investigate? There are problems +I think it conveys more dynamic energy +But database context is not available at this point in the recovery +This will allow the plugin dependencies to be separate from your project dependencies and fix any related issues +BTW even though I added Vans new unit tests Ive marked the test as Ignore since is logging the error regardless +Thanks Amareshwari +Mass closing all issues that have been in resolved state for more than a month without further comments + +Spring IDE should support Eclipse and WTP is with the update site because WTP dont have xerces feature anymore +Peter Thanks for creating this test +UI bug +Theres a fair bit of work involved but it would give us validation for other Java constructs as wellFinally we could punt and just do some simple regex validation +log without sql with sql to it is just the example that is not completely example should try an operation and retry in case a SOSE comes out of the transaction reporting an optimistic locking exception +I reapplied the pieces of the patch squashed by new master commit a while back +Thanks. +tests pass. +default search field foobar not defined or not indexed when raising the +Hmm I guess this happens when nothing has been deleted +PFA for the Error update here +You can configure the memcache flush size by setting the config parameter the default is MWhen a reaches this threshold it will call for a cache flushIf the cache is flushed a request is queued to determine if a compaction is necessaryIf a compaction is done then a request is queued to determine if the region needs to be split +to see if it was still an issue and its not +I think its router issue not KVM specific +I have refactored the integration so that the flow now defines the intial view state similar to how works +I will take care of and next week +The table was created by SDE which doesnt seem to like using only unique it be possible to add this functionality + +if target namespace differs only in caps on Windows the generated resources will collide and there is no workaround for this at this point +I cannot imagine how such simple thing can take that long time +A table function returning variable length values has the potential to allow values out of range of the types definition to be stored in tables or returned to the client +ArjenIt is +Thanks for your help Harsh +Once you provide the updated patch well get this into a SNAPSHOT +Perhaps this fix should be rolled back +As for step Deleting a project does not delete the module from the server +WDYT +From Michael Watzek has the same issue +Assigned Andrew since it seems like a pretty big if simple change needed +I didnt touch tformdata etc +We should check this from cross platform inconsistencies +I can take a look next week if it is not urgent +We might need to do some adjustment of the classloading environment for digester if we dont export these classes +Move the majority of and blocker issues to the next point release +Jan to commit this in not too patch with in and someone upload the war to Maven as Jan Hydahl mentioned +Delivered to +Sorry to come in lateSome clarifications MR JT kills all running tasks on a TT when its deemed lost +Thats why I commented on Phils commit on github +What are the implications of this patch with respect to backward compatibility for exiting instances +Hi GeorgeThe fix looks fine Thank you very muchIve raised a new JIRA for submitting new tests +At least one patch is outstanding so I will address any comments you have on patch c in a separate patch +Fixed by using a in place of a for jobName. + +is it possible to do validation during castor unmarshalling? We could find property in but is there anything while unmarshallingFor setting up in classpath how should we achieve this? as ours is a stand alone application and using XML mapping file to unmarshall the elements into objects +As Keith mentioned if we have other options then Id love to explore them +The server stopped responding to any deploymentsBut the seems to be running it is in state TIMEDWAITINGRob what now? I will save the state of the VM so hopefully I will preserve this behavior and can investigate more later +this addresses the crux of the issue and I believe that the profiles will just add additional flexibility to this +Thanks for pointing this! +This is ANT doing something to stack traces dont want to know what it does or why +In that case the host portion of the URI changesSimilarly one could imagine that the hbase root Dir is moved +catch statement only the Thread that received the exception will be impacted +apply changes for branch at revision and will verify it at GEP release +Got it +I wont have time to test it right away but will reopen this bug if I find a problem when I get to it. +Patch committed +They should appear automatically if you wait seconds +I know you dont but I do +Ive reviewed and approve patch +I just committed this +Updated patch and removed the need for an explicit Guice Cache module if using the default +Submit to hudson +its already possible but still quite some work to map i support on top of Search currently I think we should explore best practices and provide some way to make it easy. +zip file and seem to get the same problem +bq +I will try to look at it more closelytest it but if I cant make it by thursday or so its likely to not happen until a week or two later +So there were several issues here +Same here we have limited time to perform library updates and solve any integration issues that may come up +I cant access the annotations to find it myself or to give it to the default serializer when I ask it to serialize an object +Fill the test method and run test cases I could execute both sync Asyn test cases +And if ownership stays with you Ralf please dont forget to accept the bug +If the caller can recover from a badly formatted cluster this exception can be caught +Adding the name seems useful though +By the very definition of a single repository uniqueness is enforced and as long as those group ids are reasonable there is no reason whatsoever to enforce a link between group ids and domain might I say that you dont let me do anything when it come to my project +the browser was the problem +im still trying to make leader election mod in a backward compatible waythe code code is pretty messy now and im torn between cleaning it up or leaving until later +Andy I suspect that the problem is in a metadata access that is not properly protected by doPrivileged throwing a that is not properly caught possibly indicating that a field or property doesnt exist +verified by Oliver +Once that gets done lets add a test for this one +I have tried the official solution but it does not solve the problem +Can you describe how you are trying to specify the servicePete R we should reuse the stuff when doing reflections in Weld can you describe why you want to pass a CL to the Weld class +calling open on the rpc which adds region to queue but since all openers are stuck waiting on meta the queues dont get processed +Specific patch for orgspringframeworkaopframework +Currently get plays a role that should be in Compilerconfiguration insteadAnoher thing is what we read from the file +Initial suggested patch was missing a line to set the new field based on init param +just tested with the same error +Test is passing for mysql but not for derby which is blocking this issue +This is not in my estimation a Major or Minor request +By mistake I selected tapestry version of +srcmainjavaorgapacheservicemixsoapmarshalersTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixjmsSending srcmainjavaorgapacheservicemixjmsSending srctestjavaorgapacheservicemixjmsAdding srctestresourcesorgapacheservicemixjmsTransmitting file dataCommitted revision . +And it makes senseWe should not let users wait for this code for several releases just to stick to the RC definitionBut if you dont like the idea I have no problem with putting it into contrib and proper eithercheers Roland +KeithDo you mind to review and commit the patch attached +What would be helpful is to attach a tcpdump of the data being exchanged between the browser and the server +This is better than to create a new entity +Attachment has been added with description Some minor cleanup and two bigger revamps of two has been added with description svn script to rename the you attach the you referenced in your patch too? Thanks +Sorry I just retested +This PR handles all of the tweaks you suggested David +Thanks! Patch committed in revision +Retry always fail +Thanks Bryan Army for investigating this provides a very useful starting point +Im not agree with this implementation we have for all scm communications and we must use it in maven plugins +Unsure why this was assigned to Frank initially +Everything is fine Ive only changed file +It should work alright as long as we are not getting an NPE +The Felix whiteboard only registers Filters it can deal with +I checked with Jimmy and he said the change looks good +attach my patch +Well try running again with my patch +I do have a couple of questions on how it should be implemented +found the issues one has to do with my Mac OS X filesystem being case insensitive and the other with the groups root belongs to in a Mac +Closing due to release of. +Replaces v +The bridge is now working. +If you have pointer to existing information such as wikis Ill bring it into one of the more general AS documents +to signify the normal case of tokens being adjacent to each other and to separate groups of tokens +I will give a thought but it will not fix the problem mentioned +Following the changes from data created using CQL is not visible via the thrift interface +as always. +Kinda depends on perspectiveBut either way nowhere in any java specs is it gonna say that a serialized object must deserialize into the same instance from which it was serialized +we should discuss this in +This looks good but if were making it a feature then it seems like we should have a way to control it from cql too +Thanks for reviewing the patch +was s good as we will be use the api in other containers +Funnily thats the hardest part about it +applied closed as resolved and released +Maven submodules introduce build dependencies into the main parent build + +This can only be called before any data is written to this write sounds like this should be a constructor parameter to meIt does but given that we dont want to compute a digest for say the commit log or other uses of pushing this in the constructor required creating a bunch of new constructors andor have the creation takes one more boolean flag +I also saw Segmentation fault when running in interpreter mode Co +Hi Ross AntonCan you send a patch for other usersDIGY +Either way is fine with me though and I appreciate and thank you for being willing to do the work regardless +batch transition to closed remaining resolved bugs +And there the old contextual instance is storedIve commited the unit test which shows the problem +I dont see where you reset cur back to if you wrap around +Koji pointed out that the logical transformation in above case is actually correct +This is a duplicate of check it to see a workaround +Why dont we flip the default used to be non inheritable? Rather than relying on a system property being set at the moment I have not seen a valid reason for using an and this is not the first time it has caused us problems +now that the infinispan connector is up to we can implement this +Could you please attach that separately? Thanks +Could you separate formatting issues from functionality as the formatting changes make it hard to Management integration tests +The crux of the issue was that children expressions were not being validated and that was masking some other issues as well +That JIRA uses the term Abstract Data Types in its early comments +Lowering priority for now +Thanks for confirmingI added a new class to bypass the problematic area +The searcher is already capable of this. +Patch added to trunk +Version that Im planning on committing +Deferring all post soap service xml and console issues. +So the issue is fixed +Verified with still throws the error in opening zip file. +in ICU contrib and other autogenerated code +This also affects version think our best bet for this may be to modify to not rely on EXDEV +here is a more optimized patchAlso did not generate a windows patch because the windows start scripts do not seem to be setting jvm flags +May I ask for regenerating the patch in the top directory of Apache Geronimo sources +Thanks Robert +It is better to merge hythr code into harmony than integrate these assembly sources into classlib tree. +Does thing think of it +Oh thanks +out of hunks FAILED saving rejects to file commonconfigOther than those hunks everything else is in rev +Can you also add some unit tests for this +Tested +for +It seems like these would be outweighed by the ability to group up update operations but I could always optimize more + +So you may want to check with Arun C Murthy if the intent of that jira matches what you think it isOn a comment here on April Arun mentioned that he would be happy to repurpose it +I may actually have this in my branch +Lets get this uploaded please +Verified that it was no longer needed +Changes applied together with changes to +You should upgrade to the SunOracle JVM as it suggests +The real issue here is you want to be able to override templates but that isnt in the current design of the theme engine +Sample application +Now it asserts asserts instead of returning errors in vmattach function +Unless of course you want to do this. +Added to release plan +I dont see an easy approach to test this +At least once weve seen a node cleaning up the mess after we forced it to replicate after restart +The section has been updated so that it is easier to understand +fixed with rev +Using the transformer in the simplest pipeline +Thanks I too out the guide yesterday as it was not adding any value +Please reopen if it occurs again. +PhilipeAs you have already noticed this is not a bug +If this condition fails return error +Thank you Nicholas for the review +Diego if you still see this then please +Close after release +Im guessing that when you remove the version you get due to some old metadata in your environment +Again Im closing this issue JIRA is not the place to deal with what is currently believed to be a config problem affecting a single user +uses therefore now +But if the user wants to use other ledger id generation we need more works on it +Write description of the component Creating the Component with a Page Tag Creating the Component Dynamically Using Java Details of Usage API Definition of Custom Style API section Details of simple Details of Usage plz update information +Also no need to do the string round trip in the reverse method right? See the in Lucene contribanalysis +How about any structure containing a map +Thanks tried to deploy jar to G via deploy command which behaves same as in its not a GEP report it in G server not in GEP + +With of the editor this issue does not appear anymore. +Hmm +Youre right NickThe warnings came from this +revision +Good points +Thanks Zhijie! +v is lovely +You will see that one of the threads starts to use CPU then another then yet another +It compiles and runs under solr +Andy please triage as time allows +This problem is solved in Gant because of the change in the way that the Ivy jar is loaded +Actually looking at it now it seems it might be a issue +txt example files to trunk +Two years back I had to dump iBatis because it couldnt do reflection correctly with generified classes +Not sure its related to the ear plugin +In its almost implemented I put breakpoints in inboung and outbound relay but the dont seems to get called on message delivery +I just tried a artifact and got a +Please reschedule if necessary +Not sure why it works without this in but blows up in Tomcat +This would result in the failed trigger firing but the write eventually succeeding +log here if you canThanks +Removes erlang client lib build from for the patch Justin committed +There we had to use the dynamic import feature for the weblogic jars to make it work +When the attribute selectFirst is set to true and something is typed into the input element the only thing which happens is that the first item in the autocomplete list is focused +I just committed this +Basically I am all for having support for yet I wonder wheher it is our task to do this +The pull request should be merged in and in master. +In theory we could use it for indexing much more than just product data as well +As this was predominantly a coding task and not a bug there is nothing to verify here other than that at no point do you see multiple shells for the same server + looks good +Not in terms of creating a topic on +I am not sure we want to spend thetime to find it even if it exists +Hi RohitWe are seeing this issue on not sure of master branch +Fixed in . +Line of could add the additional jar files +This probably has nothing to do with upgrading +Attached is an updated patch proposal which incorporates Knut Anderss suggestions The text describing the examples uses italics to refer to the table and column names in the examples rather than putting them in quotationmarks which I agree looks superior +Remove all internal code refs to DEFAULT as well +Note that I can successfully build the branch using Java on Mac OSX +and need to be replaced or provided +If the test has taken no destructive actions it is potentially masking other system issues from the operator +It would be strange also to issue a POST request and to get back a with the Allow response header listing GET and POST +I can wait for release +Fixed on my end please test +Fixes appeared in documentation so closing. +Moving all issues to the CR +Okay everything passed except the lack of a test case +After comparing request and response data on tomcat and weblogic the onlydifference I could see was the response content typeOn tomcat which works its textxmlcharset but on weblogicwhich gives the error it is texthtmlcharsetI will try and configure or write a hack that forces the response content type to textxmlcharset onweblogic and see if that is the cause of the problemRegards the content type for partial submits to textxml fixes the problemI will open a case with BEA to fix of bug +I think this is the best way to go since it is the base class and another project may have a very different +Note the weld manual could be clearer about Alternative vs SpecializesResolving this issue as Deferred because will fix the Alternative case mostly +It also doesnt yet have any inputformats so it cannot be used from mapreduce +I hope this includes relevant changes onlyM srctuningM srcadminguideM srcadminguideM srcdevguide +for your changes +Will be available in works great +Verified it works. +see +Maybe there is a patch in my futureThanks again +This also appears to affect +They are very useful for me to investigate the issue +Update Tried again with the standalone versionIm seeing the same thing +Attaching a patch with a simple fix for the NPEA general question I have is whether it is ok for certain name node hosts to be unresolved during startup +Attached is the patch for the gwt file +Yes there is probably no good reason for caching the session. +Eclipse +Sorry about that! I see it can cause performance issues where there are large number of partitions + + +please ignore my previous comment +on top of a binary message exchange protocol than to intermix that logic with the client and server message exchange implementationThe implementation should still be layered +Tested fix on you for this +This is great start for CF algorithm +Username +No issues +Now it does not handle exceptions +bq +Isnt the plan to make libwebhdfs compatible with libhdfs and then can work with either libhdfs or libwebhdfs? I think we should keep the default to libhdfs since its more stable and weve got a lot of users on it +This area is overhauled in the openjpa code base + is mapped to accounttype There was no mapping for the Company from accounttype to a GL accountHence the exceptionPlease review the AswathIs this using seed data? Because it seems that using demo data I have any problems +There is likely a core in the wrong profile or something but thats where to look +So downgrade this to fix in a future a year later still not fixedDoes anyone have any ideas how to fix it or work aroundI need to fix it and have some time to do that but im pretty new with Axis code +Im becoming a little allergic to class names because they dontreally speak to the purpose +I fixed this in jbossxb api impl +Note that you can get around this with to prevent adding our own endpoint to saved endpoints and if detected remove it + for the patch +Tracking down class loader leaks is just too painful +This is fine for a small number of tenants but with a larger number a single Manager is required +Otherwise shutting down is dont think retrying silently from BKJM layer is correct +we are not delaying alpha for its current state anyone who install Alpha on Linux will have his workbench crashing within a few seconds +So rejecting +Marking this as major because removal of killed jobs in Capacity scheduler depends on correctness of theCan you explain? How exactly is the Capacity Scheduler being affected +I added the +But I think it isnt a good solution +Test to set the Fix Versions field to JDO beta. +Henning you have previously assigned this issue to yourselfI was going to mark this wontfix but i think you should have the final word since it is technically your issue +Fixed rev +Thus it is still necessary to know the exact number of spills instead of inferring it from the sizes +I think its good idea +This error seems to occurs once a day and aparently randomily + +NeilI guess there is a missunderstanding +Not moderated +Nobody submitted a review of the proposed patch +So that on the server side you know it +Then the other two replicas of this block will be also reside on the th node of any randomly selected remote racks +rd build +If I get some s or no comments after Monday Ill update it +Grant can we resolve this as WONTFIX +Ill have a stab at it this evening as we need this at work too +something to align for our next release +batch transition to closed remaining resolved bugs +I just committed this thanks Ahmed. +Memory appears to be sufficient alsoThe environment is very similar to the work machine I was running my test on to try and replicate the problem +Wont fix because of +As I think VPE community ask for reformat decision +How about following the specification instead +All tests passed +Im going to try a fresh co and build on my mac well see if we see it there +attached patches forsandboxsrcjavaorgapachemyfacescustominputsuggestajaxresourceand tomahawksrcjavaorgapachemyfacescustomprototyperesource Volkeris this the final patch that is currently appendedOr do you plan to work more on this I didnt find out from your last statementregards i have no idea how to fix this for opera so thats currently all i can do for thisThe problem is to calculate the needed size for the popup +This is a problem of the JVM +Better late than never +Its in a build that I ran last night +Ill be at the week starting June th +Feel free to reopen. +There were a couple other issues to sort out namely the SEs implementation class needed to be changed so that it exported a custom service name as otherwise both the SE and BC used the same endpoint on the NMR +This could be confusing however since a path that looks like an hdfs path would really be using some other protocol +If the connectionFactory leaves so should the +sure lots of folks uses it +Thanks for looking at this Jason +Looks good to me except the new set method should be named just setField no +Arguably one thing that is missing is Javadoc in the default implementation describing the ordering issue with Bytes +Sorry I have overseen that the type parameter of the root part is also set to textxmlBut in this case the The parameter of the packages media type MUST contain the content type associated with the contents XML serialization is broken when SOAP will be used +Do you know the status +It seems like theres a bug in nutchs where the classes of the keys are used for keys rather than the actual keys +Or am I not thinking clear? I just got into the office and on the first cup of coffee +Doesnt matter which LDAP Im using the user record is properly synchrinized with the for the info we are not using LDAP directly we are using a custom CAS server which sends back the roles data +defer issues to +trunk Is now JMX managed as a bit +Components field was changedAssigned to Alexey Kazakov +Committed to hivesite +dtd by +Thanks for seeing that Tom +That should have been the previous patch of changes requested by geronimo with unneeded changes removed +I suppose youre right +I will mark this an enhancement request +Thanks for getting me started Dmitry! +Ive found out that the problem is fixed by replacing with the latest version +patch looks good +Could you ask on the user group firstSecond what version of Struts are you using? In such NPE isnt possible at that line so your version is different +Thanks Arvind +Ill try to get one written soon +Patch as git diff output +Check this with Hudson +Mike I did go quite tough on your code and changed a lot so please do nothesitate to point out if you find anything disagreeable +Fixed in and +With the Forge integration around the corner and because the phasing out of the Forge integration is near I will resolve this bug as Wont Fix +Vikas +Bulk closing stale resolved issues +Yeah for now lets avoid this change in the timeline as it causes problems. +Im going to commit this bugfix patch tomorrow unless there is an objection +No problem +It looks good to me +reopen to set proper fix proper fix in master and branch +Oh +Created test Spring configuration for the portal in + +Committed for in revision . +I request for v to be considered for a commit +When I say it works well it works but it does still push up to origin +I updated several of the missing jars in the wiki +Let me know your comments here +attaching patch for trunk and branch Giri and Arpit could you please update the title of this JIRA to say something like should put proxy user info +And by the way a mistake the JSR expert group repeated with the introduction of in JPA +Ok so you want it the same as Hadoop rather than thrift +Lets do this soon and integrate Hama w bigtop TRUNK. +dont use text boxes +Thanks Knut +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Maybe its too early to decide this +I understand job execution creation needs to be atomic but I dont think it implies the whole logic around creating job execution must be handled by repository +Leaving it open in hopes that James or someone can get to it today +debugging and finding the transaction you need is very easy + +fix this and add some unit tests +Ive just committed this +this seems to still be an issue for regular +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +As I know in first FF we had in Colombo we decided to introduce the concept of service isolation +Doing it makes sense to me +I think it is interesting to have a sonar report for a short term solution +However Im not sure that folks would appreciate the increase in the number of files that parallel files would createNote that this codec could be implemented using the existing compression codec API it could accept a buffer of serialized records then parse the records using the files schema splitting their fields into separate buffers and finally appending all of these buffers with an index at the front +There are still a lot of issues with the synchronization of the xml page and the graphical page +the code but left a readme file pointing at +Updated the fix patch +So youre saying that the is not releasing the even setting the property correctly? No need to another JIRA this one seems to have a perfectly fitting title +If not it uses comma as a separater for multiple files +Can you try a recent I think this is resolved now +I didnt know if its related to the code Ive changed but evidently not +I already fixed it on my computer but I dont committed it refactored the plugin to not use ant for fop and xslt to avoid xml parsers problems between different java it works withMaven I have a last problem to fix for svg files with batik +On the mailing list we discussed the and the plan is to migrate it to HiveCan you update this patch to remove the stuff about using features from Hive? Then we can check this in +I am removing from the Fix Version field +What are the current steps? Is there any way to simplify the process +There is also an exception when saving the page at second time after adding a new exception Trong referred to is due to the lack of synchronizing storageId between model and UI object once the model object has been persisted +Changes looks good to meThanks for the fix I verified the defect and closed it. +That way we could have a version which calls the super constructor in the Servlet version with the exception and a version which doesntHaving said that it would be simpler to leave as it is and just do it in using the Servlet version of directlyIts not a great burning issue for me at the moment so I probably wouldnt be willing to give it any time maybe later I might in +CR works! I also installedftppubredhatjbdevstudiobetaJBDSEAP as fixed in that in the missing featuresplugins are newer than these repository found containing repository found containing Why? No idea +This would be an api enhancement for stacks component +The best thing would be to simply accept the encoding we nave in the source file +Some test methods have dependencies +Well we did not change any of the native code +Great analysis indeed MattCan we make NN just ignore the second BR while it is in the start up modeTrying to avoid introducing more configuration parameters +Ill collaspeall the CHANGLELOG entries related to bug into one entry once we are done +Can we break out a separate issue for work on trunk and resolve this for + for the patch +Investigating further +I might significant corrections and notes to both the and sections +may be worth a look too +LGTM +the will be deprecated in once we introduce the new expression language stuff +So the end user knows it ahead of time +Investigate the problem for inclusion into Final. +Commented on the RB +If its only or days Im in favor of going for it +As it turns out the Non Default iterator on the was looking up hashmap for values +Yes its something like that +Set v patch All the stuffs previously mentioned Incorporated Santhoshs comments Removed unused code Support null schema unit tests for all operators Support and you refer to the not being in the source tree are you referring to and? Those are in the source tree +now encodes cluster name as UTF before it encoded as system encoding decoded as UTFIs that accurate? +By the way do you have additional requirements +However Amareshwari and I discussed this offline and we are not sure if we want to build this rack before get method as it would mean calling get for all the blocks twice once to build the rack and once in get to build the blockInfo maps +Please apply the attached to fix the issue +Thanks +Could you please upload your wsdlthanks is a copy of my WSDL in +Seems a lot of the overhead is simply due to Hadoop too +For most recommendation applications I would be surprised if iterations are importantOk I have no first hand knowledge myself Nathan mentioned it did appear quite important in his previous work +Could you check with the latest Portlet plugin +I should add Service Tracker will be moved to the core JAR in the next spec release to avoid such an issue in the future. +An updated patch that replace the previous oneThanksMarco +rev I added a condition that check if the resource is already started before trying to start Verified this now works as expected +Patch for apologize I am not sure how to write a unit test for +Dropping to minor stale issue not enough information to reproduce +It should never be null because other places will throw too so it would be best just to check for null in the constructor +So benchmark is the first place to start m running with an empty table but we will also want to get results with different table sizes as well. +That works for most things but JSF requires an entry into afaik we cant do anything that +Nit seems like SNNparseArgs should for parse exceptions does this log make it to the CLI ouput? If not then users just see the usage info and not the actual error +Can we change the unit test to not run on windows for am not sure why is blocked under cygwin +I just had a look at how logj is using jalopy to specify coding guidelines and format its code +Assiging to Henry Zongaro who agreed to look at this one atthe JIRA triage on December a patch that I believe resolves the problem +Updated patch attached +Anyway either that or please give me back my that was working ok for me +This did not actually go in until fix version +are there any navigators that depend on these methods being implemented in the superclass? And of course more documentation is always appreciated +Another info if I dont filter the WRITE events only it works just fine +Thanks Stack for the review +Not that it would make any sense but any chance there is some sigar stuff mixed in there +Thanks Scott your modified patch is in rev +Sorry about that +Need to fail if systemPath is not absolute +I am working on patch with full test case +Good catch Jorgen! Thank you for raising this JIRAI am confused about having to use different monitors for log shipmentand for waiting between log shipmentI am more confused because the javadoc for notify says thisThis method should only be called by a thread that is the owner of this objects monitor +ram Thanks for pointing this one out +The patch which creates streaming directory under srcexamplesstreamingAlso adding the bigrams streming add the appache apache headerthe test was failing due to timeout and nothing to do with this the patch as per previous commetns +backported to resolving as fixed there and resetting original owner. +So since we dont access the data synchronously with the checksum a client trying to read the last several bytes of a file under construction will get checksum errorsSolving this is likely to be very tricky +See the filesA testwebappsdrttestRecordertestsA testwebappsdrtcoreWebdatabindingdatagridbasicXhtmlA testwebappsdrtcoreWebdatabindingdatagridbasicXhtmlA testwebappsdrtcoreWebdatabindingdatagridbasicXhtmlTheres also an update to the test to include an anchor with multiple parameters +At the same time all Strings of length should be seen as char too +However my thinking is that the can be used not only for links but for other items tooBut arent all those other cases where the return value can be a link? Ill try to review the patch ASAP +I see +The difference is the ability to assign a function to a property rather than a string number array or nested JSON object +guess we can mark this as resolved based on revision . +Should we have another jira to keep track of this +divorce the rate limiting entirely from Dir +While we are participating via Twitter we should also post ideas on the Apache wiki once they launch their GSOC ideas page +It is my contention however that the should be implemented in such a way that prevents these mishaps from happening + needs to be merged to tag as well +Thanks Rohini. +Is that expected behavior +Here is the patch and unit testFix description there was a bug in Handler as it did not try to canonicalize the path so ive added this action there +AS far as I can tell at the moment refactoring of that area should be finished +Can you confirm if we are changing this to I think thats more generalWe can name it as description or maybe +defer to defer of open issues to +Believe this is the same as +Creating constructors only for required fields gives the nice property that adding an optional field will never cause the code to breakDavid if constructors only contain required fields does that resolve your issue +Alexey contributed this and applied the patch so I think waiting for verification is unnecessary +So I updated the javadoc both for the class and for this method to be correct +The ci server does not close issue only comments on them I closed this issue +The class was removed since it wasnt used by any of the code and had some limitations as well as a few bugs that werent so great +There are no new features +Merge to revision on branch +Were closing this for now as it appears based on the information provided that the tooling was working as it should +I am working on the patch for trunk Matt +Postponed to has been releasedMoving all related issues to the next version +All effort goes into jBPM +I have updated the Component and Fix Versions fields +I cannot find article but multiprocessing architecture of Google Chrome uses some internal. +Thanks +Obviously this didnt happen for final but it could go into k +They are identical except for the addition of some annotations in the +Basically the same question as above +Works great now in trunk for Beta thank Snjeza +Thanks RatneshThere is a session issue actually there is getting object from session when session is expired then object will be null and there is no check for this object and call method ofI have uploaded patch guarding object for that exception +Integrated JSF based on Mojarra milestone +To clarify I was expecting field error to contain Description is required but the field error for contains nullDescription is required +one workaround to try is putting the dependency on the site plugin in the build section +Also checked into cvs docs feature has been moved to the version by agreement of the development team. +Will this fix make it into Seam? I noticed is in beta currently. +Just committed failing tests in this point I cant see a need for property nodes generated from setters but this could be necessary in the future +There is already a public setter if you want to add more but feel free to make a suggestion if that isnt enough for your use case +Made constructor public not perfect but probably OK as it is an internal API. +Closed for. +Ill clarify with the team today +enjoy +WDYTBTW I have not tried the fix in Struts and since I cannot replicate this issue +during first start time you could generate the identifier based on hostnameport +Documents the file +Will commit when tests pass +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I wont have time to review and apply until the next week if I forget please ping me back +The problem is it should also be possible to call user defined +Even if we could switch groovyc to mode some apps wouldnt compile or run correctly anymore +I think thats a nice solution to the problem and also reusable in other situations +If this looks ok we should probably add a test to too +has reference to which is not available in Crimson +There are no definitely lost or possibly lost memory for most of the samples on linuxOn Windows there seeems to be problems in case of Exceptions and also with slls +Will show up in the next snapshot build. +Thanks it seems we posted at the same time + +for the patch +looks good to me +Richard not sure about the reason of using a properties file instead of embedding the properties in the file +Writing a test to tell these two cases apart would be quite difficultI think the best way would involve turning off all entropy generation on linux and then consuming all the entropy then running to make sure it doesnt block +remove redundant endpoint publish in patch which follow servicemix checkstyle and pmd ve modified and applied the patchThanks +All unit tests pass with +I hope we do not hit this once is done +isnt that how it should be? otherwise Ive misread the code before changing it +for the latest patch +As it is a widelly used component needs some comment from downstream projects get into if the change is approved for AS this is a requirement this JIRA needs to be moved to builds are schedule to anymore uless QE finds Blockers and only builds to fix those Blockers will e allowed. +HiFirst of all please use Sonar Users Mailing List before creation of any ticket in JiraSecond Seems that one of installed plugins is not compatible with Java +memlock gives us about improvement on our test bed we were maxing out our cores +looks good patch approved and lets get a bug reported against dali +The component works properly the problem is related to monitoring i rely on these properties to monitor the processes that are executing inWithout these properties Im not able to recognize the process that is executingFor example the HTTP component has been already patched with this change +ZhijieI doubt if there are going to be use cases where users actually want a string like xyz to be converted to false +Fixed tested both manually in debugger and run ADG mustella tests all pass +on the patch assuming that the all the failures and warning from are addressed +Again there is no thread leak +I should be able to provide a patch in the next week more specifically towards the end of next week +We can registry for example. +Possibly a regression introduced by PLFED +It always looks fine in html but any transform that goes through the fo xsl files doesnt extend the cell to all three rowsHowever in this case the information is meant for all three rows in the table in which case perhaps it doesnt need to go in the table at all? Since it applies to everything what if we were to put the information in the Implementation notes in a paragraph following the table? Would that be sufficient? I know it doesnt solve the problem directly but its probably the proper way of displaying this information anyways and it can be done easilyLet me know Id be happy to create a patch if no one has objections +Logged In YES userid added to and in Username pilhuhn +So maybe isolate this function in module that we can plug in or plug out in production system +bump ICE to critical still a bug at +Review board is also updated +As soon as the Codehaus repository is in place a Codehaus Bamboo build can be put instituted +It seems like one has to fill the buffer before printing the special character in order to get some result in the Web page +A simple caching was added to avoid unnecessary validations +I also backported some of the NPE protection code with revision . +Is it actually missing in the log +Lets create a new issue for that +Thanks a ton Isaac! I will definitely review it and push it to asfmaster if it looks perfect +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +the version of HQ going in EAP is notSince you took the documentation from the version you would need to do a svn diff between the two versions in the doc directory in trunk and manually merge any changes you find into the EAP docsI have had a cursory look at the docs and quite a few things stand out as wrong but I havent had time to do a proper review yet +An updated patch +doc explaining the changes and showing the new look +In case you have a patch for this test you can attach it here +Second part of the fix is in rev +Bump to has example of how to do simple sftp symbolic link massaging +The job is using output format which has the which is set to null +Im wondering if this problem has magically gone away on the trunk +Committed patch to documentation trunk at revision + +Manually tested +Now with +Dont tablets already have fixed length ids? I thought this was what corresponded to the HDFS directoryIsnt the point of hashes that they are and hard to reverse +Any chance to review and apply attached patchesIts important for example when you use to keep information about build number or date which is displayed in the application UI +Updated version for new trunk +Could you please close this issue if it is resolvedThanks now being called +Fascinating +Fixed. +yes +However Im not aware what the equivalent of the Equinox is there +When an request hits this callback it is natural to expect to find initial value when accessing the serialized model this is the price of using static models +The best thing you can do for Solr tests if you are interested them is lend a hand +The fix was committed as part ofElena couls you please verify that the issue is not reproducible any more? +so dont get your hopes up too high yet +for the patch +This may not be correct behaviourI hope this is of some help to you as it certainly is of benefit to us +The difference between your test and my example is that my example is based on a row key prefix and your test is based on full matching +Fixed thanks. +So if the artifacts for them are not available yet we need to wait andor exclude the testsThis said if you want to put some temporary workaround in place locally Im fine of course +This needs some rethinking now +linking this with Pig for Hadoop issue +Im having this problem too and the workaround is simply a workaround +Looks like the same fix was made for instance of String +bulk defer of open issues to +of a type mismatch in the key +Applied patch in revision Thank you for providing the patch. +bq +For we will continue to catch all exceptions including runtime from user code +A is something very different from a explaining it cannot find a reverseengineering strategycould you show me the stacktrace of the NPE? +selenium test for to file and documentation were added in revisions and respectively +Its better to let the minority find out how to enable document order than to let the majority find out how to make jackrabbit perform betterI think the descriptor should just tell that document order is not supported + +It works when you use the property in your file +believe this was fixed in +If we remove it from the viewmap afterwards its method will never be called because invokes only on managed beans which still exist in the viewmap +This patch is using some code fromWith that patch you can finally cut copy text from with embedded TTF CIDfonts generated by FOP +And declaring some stable within the regular class hierarchy is simply a for me +Incorporates proper checking of master operations by implementing +updated patch with comments incorporated +Fixed the typos +What do you mean +Done +I have applied your patch in rev +shutdown is expected to be +Maybe Im missing something obvious here +Actually this should be fixed in already. +I am attaching a small maven project that demonstrates the problem and fails about of the time +Another example of this common problem see +can we fix equals here too +Thanks very much Dyre. +This should solve a lot of problems but there are still some minor issues left +Can we continue discussing there +Heres a patch that should fix the problem +This could be further improved by scanning for files in the data directory before advising to import +I think making aware of each hadoop version set of is too error prone +its more like sometimes it works mostly not +Thanks againBest regards Raymond Gao +Here is my second patch with enums which seems smarter than first +The original reason we forked commons is because of the infamous class loader problem when is embedded in WAS +Committed revision in +Because as it is it seems we have to say anything that uses the update log does not support rollback or notbq +I agree its messy but its part of the spec the acknowledge mode is ignored if the session is created read that wrong dont work without coffee +Attaching and with changes to two topicsM srcrefM srctuningThere is no need I think to add a warning to the documentation of the CHAR data type +well patches would be warmly welcomed +leaving open for other test failure taken care of on get the same thing Sylvain does in is in use by another process on Mac OS X +Hopefully theres a code solution I can put into Smooks or instructions on how to write the template work with the namespaces +The domj version has been updated to and the and the classes have been removed from the. +no our patch for wtp wont solve this because of the nature of the code +to be working here + has been marked as a duplicate of this bug +From the documentation it does seem that the store clears out a log file only if it finds that there is no reference to a single message in the log file +such a mismatch will cause similar issue for +Batik actually embeds a font but a standard Java font one I think +The footer of a column header and footer of the datatable are not for patch mailed to Gary that patch available somewhere? What would be the desired behaviour in case of multiple in a cell? Simply concatenate the values in a single excel cell or create separate excel cells for themAnother thing I use ajcommandLink in the header cells for sorting which does not extend and so it is ignored during export +I linked back to this issue from there. +Fixed in rev in branches branches branches and trunkThanks Donald for providing the patch. +Some comments mostly minor When is changed in for sandboxing the child it affects the instance in the +What is the current status of this issue? I would like to use MasterSlave failover together with a MS SQL installation +Verified in JBDS. +Implemented in revision +Apparently Martin knows how to handle this so Im hoping hell chime in +I wonder if the code and the should throw a not connected exception instead of a +Is still a problem or does reverting that patch fix this issue? +rUnbuntu please test +Id rather err on the side of keeping things out of the core than let it get bloated and unmaintainable +for to branch +Putting this functionality in the would also confer the advantage of saving the same work that was done in for frameworks +I think this is sufficient to show the problem +of course i meant that only will became a seed in the example in the description +thanks sadeep for pointing out the issue and contributingRegardsudayanga +Pending work to upload on demand +Agreed +I just committed this +Thanks issue is fixed and released as part of release +Interesting my local schema is valid too! What JDKOS are you using Sei +Oops +is good to me! Thanks! +The service and reference tools in generic can be dropped on the composite or any component +Sounds like a cool enhancement +Most of parsing for is using woodstox +I can help if needed +I have also asked about the patch I have uploaded to and how it relates to this one +Thanks Ownen verified that job tracker correctly assigns the jobs +Could you please provide one existing OOTB? EricIm not sure there are any uses if this in OOTB +I am closing this jira +and IBMs JVM seem to have this issue with varying severities +Fixed in revision to the branch in revision . +A filter that allows the th doc wont do very well on a bunch of docs searches +OK turning this into a documentation task then +This is a copy of the buggy POM file in examplescamel from the +Please verify the fix worked and then resolve the issue +Maybe doubling thedelimiter would be more correct that means replacing ESCAPEDDELIMITER in thecode by PROPERTYDELIMITER PROPERTYDELIMITER +see my +We had such discussions before +This patch fixes both +Sample which uses SeamHibernateFacelets +Pallavi is out of Office till the end of this month +If something in the job history startup does throw it is calling directly +Thanks Sadeep for the contribution +Adding test case for the backslash inputs. +Theres then no sitemap service manager. +Shouldnt we have this revision on all webservice jars +Committed to trunk after trying it locally +Thanks Gopal +This is a patchfix against trunkIt also includes a that triggers the issueThanksArcadius +If I enable the correlation optimizer by default which of the existing tests are expected to fail +Yes thats possible +I still think that Toolbar has issues but that does not seem to be relevant here +Also can you please link this ticket to the other list bucketing tickets? Thanks +According to the logs you have requires jodconverter failed to convert then all other converters have failedI propose to move this discussion to mailing CelestinoIn your server there was OO +Doh! +The file is a standard form login page while the has been changed to use the form authentication method +Works for me in +Thanks Svein ErikFor future patches please use the extension +Please apply this locally before you upload your next patch +I wrote the test originally to test what was a legal cast and what was not but Ill look at the test some more to see how I might best check the output +Stay tuned +Would it be possible to record the server response? I have a suspicion +Sample java app and WSDLPlease use the W flag in wsdljava W noWrapped turn off support for wrapped documentliteralThanksdims +see srcsiteaptexamples for more information on how to see srcit for tests. +Ive already done that +author +Ok foundThanks a LOT for the perfect test casePatchFix incoming +They fail under with the object graph is definitely toastI looked at all the contrib table columns within the debugger and all are properly allocated and part of the columns collection and each one has an allocated propertyDescriptorOGNL refuses to evaluate ognlMy work is being held up +Need to do something smarter there which may have performance impact +Yes mostly when parsing tags like the THREDDS documentation which contains full sentences including commas +Victorare you still working on this one? Things have changes a bit we have new options now like thinking of using the mailreader now ported to WWSAF and such +This is whats required to get the interface to be available to the on the proxy transactionmanager we provideI particularly dislike the part where we need to the Geronimo packages in every single JBI componentPerhaps we should just bite the bullet and provide bundlesfeatures for installing ODE in the suggested osgiinstall command does not seem to work when ode is configured to use the hibernate DAO +We use DK all over the place so expanding its memory footprint makes me cautious +Thanks Abhijit +That should be faster especially if a custom provider has not been optimized to cache the token which was just retrievedHope you OK with this update Ill keep this JIRA open for a bit just in case +Fixed now +bq +Yes I am referring to and in we have to use the copying var mechanism to create multiple server instances +But we wont be using this builder for +i following folders should not be needed xcuserdata xcuserdata +I just need to alter the build scripts +For collators Id suggest a +Due to change in interface as a part of commit my patch wont apply cleanly on trunk anymore +This is complex enough that Id like some feedback from you before committing I dont want to have to back it outbenson +Bug still present + +I tried out trunk and it works correctly nowThanks so I mark this as resolved now +Thanks +Ill test on some older C version but the test is so damn long that fell free to test it on your side +This workaround works for my current uses but it isnt very robust or +But would rather have them everywhere +Is data skew going to be an issue where some are overloaded vs other? Would this an issue when there is other data stored in hdfs along with hbase + on dependency updated. +A diff from the solution was trivial but also added a unit test for is a good practice to make defensive copyes of the objects you return I also added the defensive copy on the getInterceptor and get Listeners +possible but not likely as it works in which has also benefited from the id introduction. +Ill document this when its doneCheersRory +The implementation is no more +I dont think this touches packaging at allMeaning do not worry about including them in packaging at this point +Which icons precisly? screenshot maybeAnd by components you mean Seam componentsAnd by components you mean tags or components +The usage of trim was apparently not necessary +Some other test which might be creating using default configuration may still flap with the same error +My checkout was outdated +I must say this is a great improvement and this also fixes the shutdown issueThanks a lot Eric +used for component generation didnt include core components attributes +This issue has been fixed and released as part of release +code looks good +Then one you pasted above has instead +zip it up unless I hear for your time +Closing this specific issue is resolved in build CR a new issue less severe is introduced see to BRMS CP release notes as resolved BRMS Web UI provided no way to exit from the Advanced View when editing a package +No and we were never able to reproduce. +Reasons are in the commit message Im not repeating myselfPlease reopen if it rears its ugly head again it may be a matter of modifying the pause +That problem is that metadata queries will not get dropped when reverting to releases before The issue that we will always have to be careful of is that the metadata queries that get accessed with the metadata queries that get loaded with will not be able to reference any new stored procedures because +Hairong I agree probably not worth putting it in +Moving a bug between projectscomponents now updates the to the appropriate dev mailing list for the new projectcomponent if necessary +Indeed it does +The groovy content assist can find things that the Java version can will need to make the groovy content assist mechanism more powerful or figure out a way to filter dups +Will do this Post Final release +I tested this solution and it worked +Verified with JBT. +except for the at the end +The thing that I noticed on both the HTC One X and the Samsung Galaxy S is that the phone slows down and is virtually unusable +Can you attach your test case as a file here and tick the Grant License to Apache box so that the test can be included in the source +The IT still works with existing Maven version but breaks with branch including the security +This is interesting +Agreed Robbie here comes a better patch I hope +Spring is requires as a minimum which should help a bit on this issue +Patch to setParameter methods in and to enable to be set regardless of setting of Naela your patch is in CVS. +I just run a test and dont test the recovery and the others +An alternative implementation has been provided working around the use of but this requires modifications to the CXF configuration +I tried the thing on trunk and it didnt seem to work + for the patch +Hey thanks a lot for checking the other changes in +But for now we are going with a simpler approach of letting the user specify this to us in configurationOn the other hand to communicate the used disk space you are following roughly the same mechanism as what is defined here IIRC which is to compute the free disk space in the heartbeat and communicate it via the +Ill apply it to branch and t commit at moment cos svn is down +Bulk defer all frontend issues to +its not a regression +Currently we rely on the TCP stack to handle flow control but things may be better if it moved to the application level +He actually needed the nd page of the wizard +The patch to fix the problem +As is being deprecated is there alternative pool to replace the old one +Lars Francke sounds like a good tactic to me +Lets go ahead and create the Infra ticketI have migrated using the other option so anytime we need to switch we canThanks for clarifying +Added patch as noted in my above comment +Thanks for the patch Dave Thanks for reviews Stack and Ram. +Thanks AlexeiNow it works as expected +The freezingdefreezing of maps in is incorrect in the patch so that should be left out if portions are committed +Since the underlying issue in RHEL cannot be fixed a work around will be done in JDF +Simply I just removed codes +Ive cleaned that up. +Definitely its working now every mapper works on their own range however i have test in single node cluster with Hadoop Pig and Cassandra +What about services that use a background process to play music or download asynchronously or maintain the network uplink for a game? If there isnt something specific about Cordova that makes this problem worse I dont think we can fix users deliberately crippling their devices +Should i add one job for solaris sparc to ci runs? Maybe there are more such tests +Im not sure theres a lot of value in burning time on this with being right around the bend +Please dont mark fix versions let the dev team decide when something is fixed +Done +It is probably is also a bug +This would allow for other authentication mechanisms like or +The previous comment indicates there were still redeployment issues after this issue was closed +We still have an issue with custom checkstyle rules +Dont you think it is useless to output the info +I just tested a deployment without using Felix and it redeployed successfully +The handler mapping API was written specifically with different mappings in mind +If you find further issues please create a new jira and link it to this one. +dont you get the resulting error code back from the shell and can react on that +When Derby is running in mode in it is best to avoid increasing default max length as it could cause problems after any downgrade +Sounds issue seems to present in also +Added a minor improvement with a description of how to setup for Ubuntu. +I would suspect this this problem does not occur when using a simple Model though I havent verified that Rob and thanks! Let me know if you need anything from me to help +Because we have both ways +Because this is an upstream issue and it should be fixed now we will expose the support also in upcoming bugfix release. +Not a big deal I guess since the plugin uses and theres already the possibility that well use a feature in thats not supported in s not a big deal but should probably be noted in the docs +Heh +Although I dont know which value would make websphere happy and if it uses the soapAction at all +The utility of JCA for this purpose is somewhat suspect anyhow +Thats ok people can set commitWithin in Solr as a workaround +Please review retest +I simply use different methods for the different behaviors +Well I did not try uDig further in previous releases because of this bug and was so annoyed that I did not even care to report it just uninstalled uDig right have to restart the whole data import wizard if you do a data import error +The patch catches the and verifies whether it is related to IP resolving +Maybe its already fixed +As there are a number of other covering this bug I shall reassign this for review of the change to +Transparency is what it is +Seems to be doing like it used to +First the relationship between two parallel arrays is implicit +Thank you for your concern. +the patch to both Rampart trunk and branch +cmislib is simply a client API +How To guides for things underneath the archive section would be links to archiver documentation so that we can work out how to configure it without trawling the mail archives +Added Tim. +Fix with a test to mention the patch is against the trunkBut I guess it should be easy to back port it to. +Thanks for the reminder. +What is the worry about these? none of them should be there +I hope it is using so much mem for a reason +MartinThis works perfectly thanks! I live in the south of The Netherlands so perhaps a beer in Eindhoven +Thanks Todd +In the case of YARN I believe its setting up the file in the prior to any being executed +Ive fixed that. +Tested and closed. +Thanks Ilya Aleksei the patch was applied to BTI branch at r +Moved to we removed the property setting part to be done by the standard object configuration section it would probaby not be too big of a step to conver the whole the aop xml schema into a little language considering the complexity is in evaluating the expression +At installdeploy time I everything to get packaged into a archetype +Im running build through in releng root + +Closing resolved issues. +Contact me please when you will start progress on this could help you with many things here because I did the same things in Systinet too +I agree that its time to iron out kinks +Is there any progress being made on this issue? Is there any way that I can help +OTOH War app list has got this jar in the list +Adds as well as fixes facet to include the example code with the javadocsI think this is ready to commit +In practice its fine to have the clients updated with new brokers but you really dont want to combine priority backup and cluster rebalance as the clients that are on the priority backup will ignore any rebalance requests +Calling from a user code should be forbiddenMay be detected by an existingto be created PMD M has been releasedClosing the issue. +Maybe it was a local network issue +with and you cant use an XA connection not even for select operations without an active transaction context +It holds a connection to the region cache this kind of stuff +Using Teiid Designer and creating a view procedure and adding view model to VDB successfully adds the source model as well +It bit me since our classes were not overriding this new method +It was a result of merging a bad patchCan you please also your compute docstrings patch against the latest trunk? Thanks +Autosave not implemented as described but documentation updated to add additional note about need to save session to avoid timeout occuring. +You can add additional configuration settings there +Great +If youre confident all are ported you can close the issue +Please check that it was applied as you expected +Here is the info notice how the same info is repeated for the namenode and secondary namenode opts Command specific options appended to HADOOPOPTS when specifiedexport HADOOPNAMENODEOPTSINFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT HADOOPJOBTRACKEROPTSINFODRFAS INFOMRAUDIT export HADOOPSECONDARYNAMENODEOPTSINFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT INFODRFAS INFODRFAAUDIT +I left in the alls well message in trunk but removed it in branch +Closing as reasoned above. +Looks like you have some changes in the patch so just be sure to not commit that +rev +Full means full alternative that is it should provide an implementation of the former functionality +because currently is eclipse findbugs plugin broken in while it is fixed in repository its not released yet and its in soon status for about months +Fixed in under svn revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Package Admin in Rev +Dag also had suggestion on alternative implementation for array handling involved with collation element creation +If not I will commit it later this week +I couldnt find a way to modify the description +Thanks Arun! +I believe that if you have a dump statement explain script will cause the script to runIs it not possible for Pig to find out that there is an error with store syntax? unit tests are pass on my local machine the errors reported above seem to be related to the environment +But if the files in this region increase to or its priority will be or the compact on this region will do first +There is a purpose batch replayer does FD check against it for fresh batches +pending Jenkins +on looks good to me +From this information we will write the final release notes for all issues +We privately confirmed it was a broken JAR in the local repository that corrupted the class path of the plugin. +Ive committed this and merged to and +Your patch and test case look good +I reviewed these changes as well and they look good to me +Thanks +Verified by Paulex. +Yes this was in the and releases +Agreed all three are not regressions +I alternate way to do it other than the way my patch does it that I think might be better would be to use only instances rather than having a configuration specifically for packages +Ill check this out as soon as I can +Please tell me the desired space key for the new space and Ill create it for you +The problem was in my implementation of the equals methodThanks Christian. +Perhaps we need to find a way to log a warning or error if comes across a field that has both conditional and unconditional declarations? But checking for that might be a pain because we dont want to have to run through the field list on every addition +assigning to Emmanuel to get his opinioneither we do the change and write a red warning in the documentation we reject this we find a way to implement the new one and deprecate the old way of setting the values so we can give users some time to update configuration files +Yes all reference values come from R +The problem is that most developers of these scripts develop on Linux where sh and bash are the same program +It round robins dispatching messages between consumers even the slow consumers +As I said something is lingering +Charith are we planning to implement this feature for the release +Im attaching a patch which makes the change in both the places along with unit tests +Still I think it would be good to have this fixed +login should be faster now +svn commit m document parameter inconsistencies for java mojoSending srcsiteaptexamples file data +Looks good to me +Thus clarification +Not terrible but not ideal either +Closing issue as it was included in version which has been released +Perhaps it would be possible to detect this condition for integration work in early enough to destroy the object built by Grails andif necessary instantiate your own naming object +The release will come the road map for more info +Moved to Infra which will be able to help out. +Just saw there was a new patch will review and hopefully commit +Ive just run up this test application and it appears to work properlyThis bug is pretty old now +And theres not much discussion about the bug duplicate services bug unpredictable runtime names bug etc +Is there a particular reason +The only difference with is that the wsdlRoot and test defaults to null for but defaults to the targetgeneratedsrcmainjava onCan you take a look to make sure it all looks OK? +The failure seems to be in as it fails with or without Travis new script +Attached patch merges with trunk to resolve with the refactoring +How are you running mvn +once check the attached screen docs check markers in eclipse ide its showing so many errors and problems of the openmeetings did resolve as much warnings as all old Resolved some more are there if possible check xml warnings classpath dependency validator messages jsp problems java task java problems and spring problems too +Alright committed thanks. +But one step at a time based on need +However the already contains plenty of methods and these will clutter the API +You should have nagged me more Please find attached a patch for this issue +It stands pretty much on its own querying +A parallel comparison with Hadoop should help pinpoint full patch which should properly save the bytes advertised +Patch applied +And yes there is a race condition at midnight in the test when I assert current date +Closing. +This one stalled a bit +I should be able to just give someone access to services without access to the parent server right +Tiese can you add this one too? It is a pain to have to add the start form in manually every time the model changesThanks patch which implements formKey support for key is added to start event thanks Hinse! +Apologies for the extreme tardiness of my Greg is correct +patches for trunk andSimply check if the tag exists in the set +Added a patch for the following Remove the old harness test and its files Add the new testderbyall passed passed with just one failure that I believe to not be related with my changes + +It should use the yarn which will read from the instantiated minimrclusterminidfs cluster +In my eyes the result of the webstart goal should become a dependency of the war packaging projects and become part of the projects final artifact +The main reason that contrib exists is to add new featuresprojects which are being tested may take some time tomature and are reasonably so that they dont need many changes in existing code +See for issue details and the fix for it +thirumg I merged latest the latest upstream trunk and reapplied the changes from the last patch +I can reproduce these quirks easily on my local copy of so Ill look into fixing these +Thanks MarkPatch applied to site directory with slight modification to reflect the RMI contributions held in JIRAPlease check that the patch was applied as you expected +I have made the changes you seek in the committed version of the code +I just found it by accident while reading the forums so it is not a show stopper for me just felt it should be fixed +Changing version to +See for further details. +Youre right +bq +Just add the spring and aspectj jars to the lib folder and run +Thanks Bill! +Simple patch adding set to TMP and defaulting it to +Attached are the following files and their decription +Applied patch with thanks to Edward Campbell +Is this desired behaviour? Shouldnt the default be used as fallback if the negotiated schema fails instead forcing default as only scheme +Havent seen this in the logs and Ive been looking +Thanks for response +nulls needs to be removed fromyou need a license headerI think it is best to stick with the same sql as in the script but fix the comment to match what is being tested +The patch is commited +The problem is that the SNAPSHOT is then always considered by Maven as the lower version and under some circumstancers Mavben will select the release alpha or beta even if you intend to use the SNAPSHOT because only you know that it is newer but for Maven it is older +Thanks Gautam! +Will do that after releaseRegards fix will be Post +The old snapshots are being cleaned out now and in the next hour you should be able to get what you need. +I just committed this +Please note that the contract when implementing Interim will have to allow from now onAll the algebraic functions that weve got should be alright +So the stub holds the header blocks in an array and everytime the stub is called it added the header blocks and when stub destructor is called it deletes the header blocks +Or +Owl also handles other storage layer interactions like data pruning +Patch reviewed looks good +Looks like there is a tab in that patch +I works and no blockers any moreIm just not sure if the cache actually does its job +Wondering if the attached change would fly instead of a config change +Given that has only abstract methods itll likely change because of additionaldeletionmodification of abstract methods +Also at the place where the NPE is thrown the document text is not available +My environmentjava versoin maven ubuntu +Committed revision . +Also this one should finally get rid of the javadoc warnings +Let me rerun for you now +A great example of that is the noggit release +As long as I am not attributed in any way you may use the remaining content as you see fit. +Heres a project generated with the patch that grabs the service name and target namespace +We dont have a Sybase source that applies to this change. +According to the feedbacks from Robin we are improving it +Code duplication is uncool. +Is there something going on between UCL and dynamically generated classes via +Please change it to Trivial +patch to new patch from Upulthanks conversionconverts mustUnderstand actorrole converts fault element structuredepending on version +Where is that patch +I have committed this to trunk and +My focus on S is as I said to enhance existing Struts applications perhaps S is a more appropriate place for this +It only increased complexity and made learning Tapestry more complex not simplerYour example seems to ask for logic something that definitely goes against Tapestrys phylosophyPlease post your ideas in the Tapestry dev mailing list before posting here +When I run a simple servlet web app it works fine +Heres a patch to remove the file extension checking +I have added warning to wiki page and notes in javadoc about the fact the adviceWith cannot work when advicing multiple times on the same routeI added a working unit test based on the attached file eg only advice one timeCamel should be able to support this when the routing engine becomes more dynamic +However if something canshould be addressed in feel free +Well maybe start with why computing maxScore at all even for a? We use that to normalize document scores when doing some form of distributed search +Closing this per Yoniks comment. +Thanks Narayanan +If everything looks good let me know and I will make up a patch that will apply to also +This issue affects as well but I am encountering conflicts backporting the fix +Fixed in r nightlies starting Apr +Thanks Juergen resolving this issue leaving the other issue open for that specific issue for M +Documented in the EAP Release NotesThe deployment of a persistence unit into the EAR will fail when the unit is outside of the EAR and a bean within the EAR tries to inject the persistence unit into the EAR +Confirmed +Can you give a case where a Schema isnt sufficient but a would beFurther converting to a is pretty easy but Id be very curious to know of a use case +I will work on that +Writing and testing the new patch I found manipulating log writer which exists in seems to be insufficientI could find error was printed to +To JervisIts too badIm backing office and fight for my windowsxp to compileAs you saidI try to change commands character encoding with chcp before running mavenBut the same wrong message disappearSo I change commands character encoding to english with chcp finallyIts not a different way +Closed since this does not cause an issue when the phase is long enough for the emailserver to start +Im having trouble applying this patch to FB branch branch +Hi AndrewWill that require a migration step to migrate existing data on the disk? Or will that apply only on the blocks in memory +incorporates the changes writeExport methoddocumentation for is have tested with properties and setProperty all worked hope this can be incorporated into first one now incorporates changes to the this means that the filter is aware of the change. +Sorry for the delayI have created a patch please let me know if there is something missing +unless Roshan wants to take care of this +Below are some of the files that were giving me errorsIndex srctestjavaorgapachetuscanyscabindinggdataconsumerproviderIndex srctestjavaorgapachetuscanyscabindinggdataconsumerproviderIndex srctestjavaorgapachetuscanyscabindinggdataconsumerprovider file has all the changes done under module +Looks good to meve slightly revised my plan for the next step +It looks like the IO functions from are not compatible with the current version of LCMS +Lets commit it +There is still only one behaviour model so there is a singleton lookup of the entry point +No changeIf it was rhetorical question Ill let it go byElse adding a whole new set of apis we need to maintain is a massive api change +I got that wrong +But check first to make sure you are running at least +Thanks closing issue. +Test failure was caused by updated messages for the es locale +I separated build files for Tomcat and Tomcat +Mike can have this one as well +stack Mikhail is working on the open source port +I think there should be an associated issue to make this work as expected. +Left some comments at RB +It would be great to get it going again +reopened +I understand the reason for that part now +wizards dialog The XML importer launched from the guides view after the class workaround The XML importer launched from the Import +Verified +bq +It should make it into ESB currently in karaf mainline not sure if the current main line is the one being integrated into ESB NMR features project right now. +Tested + +A client should be able to close a file even if its last block is not complete +This way were growing basedon demand and availability +unit as requested by Vladimir. +resolve as out of date because of month no activity +BLE is fixed +First I do a layout for adocumentation in +One should not have to hack a declarative component with a custom attribute +Patch to convert to use CLI for argument parsing +Having committed this to the branch and subsequent RC tag the and artifacts seems to be fine however and are not and still fail to produce the top level folder within the generated artifacts +So if you have structural changes you want to make make them there +VERSIONSWITCH is a system property that can be used to fix that but what you propose is not a good default +patch checked in +on the other hand TTF can be either subset or full embedded +the same problem occurs with all othersLet me attach screenshot +Another patch with pread fixes +Can we conclcde in the issue description what we should do here? I think the artifacts in Lucene and Solr are exectly the way we want it? It works out of the box and starts a working Solr? What should be changed +Some outer library rescues and does retrywraps exception +Jody and I brought most of these up to the spec but it still needs another pass before we can stamp it complete +Logged In YES useridSorryit was my fault +In my opinion we should add a pre processing utility that makes a graph sequencefile out of a text adjacency listBecause it seems easier for the user to let him operate on text files which he could modify all the time to see the change in the output +Im going to run all the tests just to be safe this time +I guess you can close the issue now. +Each supports different use cases and scenarios +i will do the necessary changes to the test case and upload them here +Yes verified +We can measure an approx size of conf by the file in the job history location +Ill document these steps +I tried setting it to with single quotes and I didnt get a syntax error like I had without the quotes +close issues for. +Done for GSG +Heres a patch showing the differences between branch and trunkI think its ready thanks to Mike for helping with a lot of the cleanup here +if i just change the value of some property at some the new value does not trickle down if i try to execute mvn from the child level +Revise patch after license change changed line numbers +Most lookups are done shortly after the first insert and thus in the cache +Reply Channel? why do you need the Channel there in the first placeRegardless I agree that if the Reply Channel is set on the message the should just use it instead of complaining +I think it was thrown when trying to send a message after the connection had been closed + +Manually tested +Yes I use Continuum to build Continuum itself + +I have added one class and modified the existing class +Please remove it before commit it +Committed +Allowing counter mutations in regular batches and not allowing them in atomic batches still might be a problemStill think the right thing to do is disallow counters from batches entirely at least for CQL +and up work in latest trunk. +I see +If there is a volunter it would be nice to make some experiments with the EPSG database +Sorry I meant this bug is triggered through out use of not some fictional Xml appender +now generates the code for all the services and How can I get the fix for this issue +These are both fixed in master and Ive requested the fix be added to if theres time +I like the optional dependencies alternative better add optional jetty dependencies to Solr and change the javadoc for to note this requirementThinking about this more the rationale I would ordinarily use to justify making a dependency optional involves minimizing runtime dependencies +Great description and I understand now +I think what is on the site is sufficient actually +Closing as Nialls fix is committed. +Please review +Fine with meFYI I didnt receive email notification on Aug th +Just a though Im guessing whether theres a way to make your life easier +Are we talking about a problem using geotools with db or another applications without geotools having this db problem +This happens when the artifact searched for is found +Thanks Namit! +Also in the final patch there is utility methods for startstop cluster this a duplicate of what is there in the other patch please clean up this issues +Take +Hi AlexKeith and I looked over the patch and noticed a couple of things it would be good to fix before commitThe consumers are set to evaluate latency which we probably dont want for this test +We tried and Cordova rc +They need to be looked closely +If you like me to use approach for this patch I can modify the patch +One more thingIf using netty ssl the connector needs a and acceptor needs a and aWe need to take care of this also +Attached patch shold address the useJmx issue +Earlier you complained about having to deploy too many bundles to get the desired functionality so this comment is more directed at thatStill I am more thinking about how to handle the generic issue +sorry about missing this typo +We need to try to reproduce it w the latest bits +Without that facade no operations can be invoked during on the resource during the installation implement control action facade in AS plugin improve error handling when plugin is passed a with a null metadata field +Then breakcontinuereturn would essentially become throws of a singleton object of the corresponding class when transitively in a finally context and the outer block would need to catch all three of those and perform the corresponding action +Java Lucene has a release candidate for at this time. +Once we decide where to move it we can always create a deprecated proxy class in its current location so that therell be a time where people can use the old or new name but the old one will be deprecated and can print a useful warning +Because works outside of that environment we needed a way to provide with the fileThanks have applied your patch a little bit modified the exceptions are now logged I put the property file configuration close to the ivysettings path as both are quite relativethanks for your contribution Will! +Question Im concerned a bit over allowing different service ports for each of the HA hosts +Hows this proposal then? keep the sortendpointsbyproximity signature as is and pass the client endpoint along with the list of data endpoints and add the following logic sort the endpoints using the endpointsnitch +I have compared runs using Surefire and Failsafe with current upstreammasterSee the attached +On the other hand if no one is using decimal then it might not matter +attaching unit we have a case where this would require changes in existing client code? If yes I think it should be easy enough to change the implementation to store date values as strings instead of longs +Minor change to allow for progress indicator since that is valid +They rely on a lifecycle that WSRP would need to be made aware of to be properly supported. +I ended up running the code in a standalone Tomcat that had the JAXB jars in its endorsed directory. +The test failure is also unrelated with the patch +Anyone not happy with the missing poms in could also be encouraged to move to +We probably need a map from disabled which should be checked everywhere of was not fixed in the branch +That makes it choose instead ofWe could have Falcon load its defaults from and then load +This patch covers the case +Fred we still cant reproduce the problem +should be fixed now with James latest patch. +Thats a good point +The attached patch for adds support for relative paths in the order by clause of a JCR statement +Also I would suggest we fail noisily if the last job we get is not in the PREP or RUNNING state so that we wouldnt have false positive runs of the test +I have not run regression tests on this yet Im just posting a quick fix that resolved for the problem for me while I was looking at +Patch for method applied +Closed as we should open new for issues similar to this one. +apps make proper test deployed app should be at least mb often deploy much bigger apps and deleting that big deployments might take lots of extra post the details of exactly whats in the tmp dir +me facepalmSorry I should have tracked the timestamps on the patch submissions +Duplicated links and the advertising text have been deleted +Incorporated Devarajs commentAttaching combined patch of and as there are some clashing code changesPlease review and provide your comments +For your first example we may set snapshottable for all users +We expect far fewer cache directives than other namesystem objectsIm attaching version of the patch with the simpler approach +A will only be created when its Step is actually started +I cannot reproduce the problem +It has the right logic to fallback to the Java implementation if the native one isnt available +Sorry Todd my fault I just fixed it and updated the patch +At least with regards the aforementioned test case +Updated in ESB codebase will be in next in ER tree jbosslocalERJansamplesquickstartsudpgatewayjbosslocalERJansamplesquickstartsudpgateway lib scripts src org jboss soa esb samples quickstart udpgateway test directories filestree jbosslocalERsamplesquickstartsudpgatewayjbosslocalERsamplesquickstartsudpgateway lib src org jboss soa esb samples quickstart udpgateway test +Currently there are tests for we need the same forSo who ever wants to tackle this I wont be able to getto it for a while +But its abiggiebq +It is used only if you have custom disregard this the mean time I am able to use toachieve what I +good +Needs some deeper investigation +The basic gist Im thinking is that we need a region rebalancing threshold +What did you have in mindMaybe should be somehow combined? Dunno +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Is it already decided where the code is going to be located during development +OK +Patch committed revision +Thank you for submitting! Scheduling for +Some feedback is needed as to wether the current state of Gant represents a resolution of this issue +NPEs are bugs +i have deleted my whole repository to be sure i was in the same configuration as lambda have used the same snapshot of dashboard and the versions of no problem all works finei used under windows and you +Any workaround for this issueIs this by design or will it get fixed? +Ok this might be a problem in file install then ill try to have a look later today or tomorrow +Drop the from fix version +Submitted this patch +Developer documentation in Eclipse help +s the last platform for this +I think this was broken in +Seems to fix the timeout problem mentioned above +I havent seen it fails recently +I have added so that both profiles are activated the Launchpad Testing Fragment Bundle is now correctly built +While the previous patch left a single whitespace character around every JS token in the file this patch restricts this to unary operators + +I havent noticed but example uses and there is a reference to instead of +Try to add the next parameter to your wget commands clientcert resp +comment from is still isnt addressed +Its possible to use and IMG tag instead of an INPUT tag for the updown arrows that would cause the pressing of the tab to ignore the arrow buttons. +Thanks Amar! +? Using the last item in the glob expansion or search in reverse order should generally get the latest versionLinux Same as +However the reported bug by is still showing up when I have a tableFirst export page table will be missing the captions +I verified that the second does exit the method just via an exception rather than via the return +Perhaps we should whenLOCKDIR is null explicitly use the index directory as the lock dirDoes that sound right +Id like to commit the fix for this but I am getting some strange and unrelated problems +Hellowe are using Myfaces w Tomahawk same versions at Deutsche Bank for an intranet application to render large tables and are observing simillar memory issues during high workloadThose may grow up to rows w about colsWe are still examing this issues and are trying to optimize our code but we would highly appreciate this issue to be observedWith kind regardsJens Ludwig +See Issue Links. +Resolving I may tweak more but its got what it needs overall. +resolution status status +this is the patch purely for the not as per Adrians comment + +Ill take a deeper look monday +Unfortunately Vienna is not part of my holiday travel plans +Resolving as fixed +I fixed a few things but there are still problems with it +will +In r check to see if this is a Java typeTODO we still need to generate mangled version +According to the datanucleus website is scheduled for so lets wait for it +This patch does better failure handling +Thanks Nicholas for the review! I will commit it shortly +Chris this bug has been in critical state for many months +Tarball containing overlay that reproduces described issue +Can you think of some scenario where these would be needed in our current operator model +Im sure you can create a small test case and attach it here +Logged In YES useridShould be fixed in CVS. +Patch generated using +I think phrasing this in terms of is overcomplicating things its sufficient to imagine one large ring and realize that a recovering node is going to received throttleinkbN since all other nodes are going to replay hints to it +I have a few comments in review board also +OK +If that is the case we can close this as duplicate with right +Ive created to improve our handling of this +remove the dialog +It looks like the jars do not get downloaded again if the only changes in the jar are on theHas anyone faced this issue and know a workaround for it? Does anyon know a why of forcing the jws cache to be refreshed? +Previous file did not handle null to and +FYI patch no longer applies +Committed at revision +is used when every token has a positionIncrement ie when several falseWhat does happen is that the that is constructed is limited to +Im not sure there would be significant countervailing advantages to the other policy +Do you have a fix for it yet? Thanks + is currently only used by +Im interested in using in some code that is generating emails on the backend +If your intent is to use Logj then remove +Just to clarify +Committed to trunk thanks for review Stack. +fixed in svn +Thinking more about this one we can exit safemode faster if we can compute mis even before we have one replica of all blocksStep the namenode waits to ensure that there is at least one replica of all known blocksStep Then it invokes process to update neededReplicationWhen the cluster restarts the namenode starts in Step and starts processing a storm of block reports from all datanodes +If the test would check for then ant would turn offOracle support in my case and ant compiles without errors +Moreover if I create new formaction with running server new pages and classes are not deployed even with server restart +Erikcan you give us a testcase please +Looks like again +C JDBC +Sorry thought this had been addressed as part of some other work but this is not the case +cant reproduce bug under current version +The data is stored in the model using semantics insertion order rather than order +Can you provide the code of the service that causes this exception +Argh renamed the files that the findbugs exclusions were in but didnt update the findbugs xml file +Files included are +agree +to branch and trunk +Once this change is made there is no need to copy the in ASC as done in +This is required to easily differentiate between the statistics objects for a indexCommitted to trunk with revision +Thanks a lotBest regards by Richard. +I cant duplicate this with Roller +Is this only problem with domain or standalone as well +Ilya what is the status of this? Isnt this already completed? +Is it committed + +passed all reproduce fails and seemed not related to this. +BTW will we need to have the concept of containers for the input side as well for uniformity +Talked with Andy Miller and he suggested removing the XXm from the server +adds retries for assigning meta if regionserver comes up after exhausting the retries the problem will still be there +Ted can you run just the groovy script against a server to see if you get the template exception +You guys seem to approve and Im only on the fence +The website has not made the announcement but SVN license has been updated for a while +First renaming all of the type references and then renaming of the itself +My comments are with respect to your earlier question I am really not sure at the time of the sort whether both template and row data should be or +Commited additional tests with revision which checks the collation type of the database +Derbyall and passed on Java and with the patch +Yep theyve postponed the release till then +Dont think the patch can affect the test in any way +So other placement policy can override the default deletion policy +here is a catch Simon looks good +Attaching the testcase report +I agree that none of these issues should block +But to me configuring twice should not be an issue but processing the configured property without realizing that the same configuration has been processed already could be avoidedWill it be ok if instead of maintaining a List we maintain a Set of keyspecs in the and provide an equals method that returns true if its params match exactly as the passed in +I ran test commit and I think it should be sufficient for this patch since it is not touching any real code +Added the changes to good to me commit after tests +Remove unnecessary you post a small unit test to recreate this problem +I just gonna try it and we will see +The workaround of using a separate database to store larger text and binary stuff has never really sat with me terribly well. +for this patch +Well done Evgeny manually tested +This issue is marked as fixed in but I m working with a and I dont see this patch included in my version +done +We did our performance testing with a app +Verified that exception is now properly thrown at . +Thanks for the heads up. +all issues to the CR +The fix has passed the null local variable test +Investigated the test and it happens randomly +as is just enough for this purpose will create another method that will use producer threads and consumer threads +Yes this patch affects namespaces declared in external files too +Hi Stefanin revision Ive committed a slightly modified patch which checks for a trailing after the target path +These commits are wrongly assigned to this issue +Addressed the above comments and committed to +Please test and close this issueRegards Upayavira +Tested +May be David can give some guidanceI think that the get should give a that a entity was not found and return a nullad the comments suggestor mabe use a different method to check for the entity +Hehe yes I saw you there when I was looking for things to do +Applied the patch and did some minor polishing of the unit test. +H Improving the modularity of the is on the map +In other cases it may have been put on an error queue +no the is maintained by the domain controller only and all changes propagated to each host within the domain +For example deploy a datasource and use this driver as a dependency +We might have to remove them +There is also a in junit package as well in +Are you sureIt was true for AS but that was a mistake bugAnd what do you mean by affect +For some initial testing I ran ant clean caused code generation to happen as expected +urlhttpslocalhostcatalogcontrolshowCATALOGtrying to delete categories +Uploading the modified patch +Would appreciate it if you could test more with tonights snapshots. +Alright well thats fine with me as long as the limit is not so silly low Ill at least be happier +This was checked with version +Anyway thanks +In general I just took a conservative approach and made anything that might possibly change something use the write lock even without being fancy it completely dropped this class off my jstacks +It is better to have the different encodings in different classes that use the standard apis +outside the project you have loaded in netbeans +trunk +That change updates the to also hold a jms as well as the jmsSession removes the setJmssession method and change the getget methods to lazily create the session +Hi Samuel could you please explain the issue a little more I dont understand +Should be removed from the patch +Unfortunately it is not possible for me to test this since all my Linux boxes has all +classes implements nothing implements the of runtime could be removed +Please have a look +But I cant get it +Assigning to Arnaud since hes in the process of making changes +Thank you +Then the UI can just do a GET with some query params and then the server can spit back the zip +I will review this patch this week +Still apache is depending on +looks improved over your previous resultsShould we commit this to bulk branch? I think we should also remove the version of Simple we have there +I think its an improvement even if it doesnt fix the master issue +I tried it +Having trouble duplicating this under test conditions +teststateSaving like some of the settings in the instructions are needed and it was forgotten to upload a patched attached a patched that allows you to actually reproduce the issue +Well n isnt ascii and t isnt ascii +My username is lresende AT apache DOT org and Ive added myself as admin +I cannot reproduce these failures outside of the unit testsI call for help +Worked with Bhavani to implement this change as with xmi serialization delta binary cas is requested by UIMA AS clients +I created it using cvs diff u in the Dialect directory +Run test suite without changes to extraneous changes for to trunk +click on Details +verified with revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Alexey I ran it times in Windows couldnt produce the failure +Nice one +Tested on my cluster works as expected + +bulk defer of unresolved bugs to +We got a NPE in get when we try to use uirepeat or ajrepeat. +the feed timed out causing feeds to loose them. +The patch still keeps the which currently is an that always has snapshot feature +All these are nice to haves +May somebody pls try this +For this reason I suggest to increase the timeout to one or two minutes and to drop the property +It fails when trying to get to see what to install and what is not found +By making the jira work with the MPI standard and making the portion configurable you allow users to pick a secure MPI implementation such as +Its very probably that all the modern compilers remove unused temporary objects so the assert wouldnt affect the performance in the release mode +Loading jars by an env variable can be problematic +Will this issue be fixed in +Now the method use pkField of realEntityand retrieve for each pkField the correct colAlias Im waiting for your commentsCheers +Hi JacquesThanks for the response +I cannot figure out a good way to run tests against rebuffed jars currently +If you declare a range you are saying that you project WILL work with any version that matches this range so why people say that the snapshot should be resolved? why would you choose a snapshot if you can choose a release as both match? If you need something that is only on the latest snapshot you should go to you pom and change the range boundary to say thatHave such a bug open for this long time without a workaround is a shame! If the community dont agree about a solution so lets use the most voted the one the project leader decides or whatever +Is it possible to color code the attribute names as suggested in the code +gaurav rayees is this still an issue +One more issue appeared after commit incorrect values of and by defaultFixed at . +It got on Tools +it can not be tested that way without powermock which was rejected +If you beleive that should be please and contact the program team. +Thanks. +As i remember Fast Infoset works fine with +Is there anything about it thats different than the other configs that use job +So we have a separate meta log and a different pool for opening meta regions +Sorry Ive been AWOL on this caught up in worklife +Thanks Karthik +But now I can rest easier +Same patch just different paths also works for +This patch does squelch the compiler error but it also breaks the tracking of minimum values +Closed due to the release of Tiles. +A month after the bug report master has changed a lot and I cannot reproduce this +And thanks again for your patience +Is this important +Thanks! +template should be automatically downloaded when you restart SSVM +Patch was applied correctly +Each checkuser call frees the values for bannedusers read from the configuration +Invalid for oss build +Could be related to he webstart issue as well +Attached patch +I should have had dug more in jdk sources +Thanks liutengfei! +I committed this +Lets create a separate ticket for warmup +It still didnt work so it is more than just a jar problem +testmanagement DDURATIONin a loop +everything fine +I also think that this is a great idea and that finally well get a consistent and common way to compare different approaches to do things +Patch correcting the issues with the JSP pages for Struts first patch I attached contained an error in the cancel buttons javascript was setting the value to save rather than list this patch corrects that + +Config sets are kept in solrhomeconfigsets by default but this can be configured in The actual schema and config objects are not shared between cores unless the share schema flag is switched on +A new jar will be created and published +Does this issue only occur if its a war project in Tomcat? Or does it also happen in +The issue this bug is about has been fixed +Thanks a lot for the patch +So I am going to work on the module using nova and I will post it when it done +M still isnt released +Confused myself again +Noticed the commends and wondering if we need some more managements on the Feature substitution group +If we change the to just show the majorminorchange numbers it looks like things work ok +Use services +Hey Steve Im sure Eli will be happy to address your feedback in a JIRA +Thank you +But it still may be convenient to have a log configuration in place while running the unit tests to get extended debug information if a test fails +slaws reports on IRC seeing this yesterday when building against the trunk +Updated version of patch that will also push the client address down into the bind requests +Committed patch. +Patch +Does the same thing happen for manual index usage +As far as the item +thx. +Bulk close for Solr +Applications still need to be copied to +Fix by r and +While +Heres a new patch incorporating the feedback +Fixed in close of resolved issues. +The config need updating as well +There are a lot of things that will go wrong if zookeeper is not able to create or expire ephemeral nodes +dodeploy have not always been removed by a +Is this gonna make it into surefire one day? I eventually need this applied if I ever get the chance to Mavenize the Groovy build + +We tried to use standard Eclipse extension points to point it to the same menu as JDT does but without found out that JDT hard codes items for that menu and it seems that there is no way to add our item there +Could you please provide more informationWhat have you tried to do? Can you provide a stack traceThanks not to be an issue anymore since Snapshot where new bundles were introducedPlease if issue still exists. +Is this only or a issue as well +I went back and the RFC and I see I interpreted it incorrectly the spec says that the server must accept at least octets but does not seem to place an upper limit +I dont think much of umbrella tickets +That isnt the Apache repo that syncs to ibiblio +If you think both the commits are reasonable then please close this JIRA +Also the Fix versions flag is intended to mark where the patch has been integrated please dont set this field +This has been fixed +Thanks Max +Tests passed cleanly on +This was committed long ago +This is the real patch +The password variable can contain a hash version of the cbh is why global variables can be dangerous. +Hi PhilippThe change youre seeing is by design and is indeed a result of +OK I understand why they all keep going to h because there is no way to request anything but h so it requests with a +Will do +Looks as I expected +I did not find problems and regressionsIf I find any problems on other container Ill file a bug and look into it. +I have just added code to Groovy Eclipse the ability to specify where the class files generated by the groovy compiler go +Im pretty sure you two wont actually hold a grudge but try to be more civil +Tests are updated +In cluster mode if timeout the in stat will be increase more and more. +Committed earlier by Jitendra. +Is this not in thrift? If not our bad because we had something like this at the place where I used to work +Jaxen already defines the previous dependencies as transitive dependencies so this made just declare one additional dependency. +moved to the whole interceptors are completely different +Adds some missing javadocs Rewrote some loops that used indices to use the new for syntax easier to follow +This jira issue deals with the not the testIf you look at the core pom this test is still disabled as are several other tests that seems to have timing issues mainly +Jsp file demonstrating error +and added the in Rev +Closing this JIRA out will log any new issues test cleanup in a new JIRA record. +You can see the exception at the end of node and node log + seems to be modified +Fixed in trunk also added checks for the case +Added the resource map back as a parameter to the vdbrepository add method and hooked in the load from the initial indexmetadatastore. +Patch to fix Getting Started link is Rubezhny I would be inclined to delete any reference to Viewlets and Tutorials and news as those categories dont really relate to the enterprise Develoepr Studio +the function needs to be renamed to create or some such since it isnt creating the equivalently we can move the creation of there too +A patch will be posted soon +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Patch which fixies exception in branch and trunk +We should do Benots first suggestion +Thanks Harsh +Jclouds is already using the secure tempurlsig mechanism when a timeout is passed as an argument to signsign +But I get the same result +Could you propose a patch since you seem to know a lot about this feature thanks +I regend all our pb code using the protoc +Which is inline with current thinking anyway +Found out that the commits made in r +It makes one variable transient to match the others removes one comment I had added for myself only and adds a couple tests to +running tests and will update a review board + +So how about a depending on the existing test for setupconfigetc? Please do soFrom my point of view there shouldnt be any setupconfig stuff in the +srctestjavaorgapachehadoopyarnserverresourcemanagersrctestjavaorgapachehadoopyarnserverresourcemanagersrctestjavaorgapachehadoopyarnserverresourcemanager the bug +Putting my Ant team hat on here we dont think people should run builds designed to take days +closed by mistake during bulk operations +In fact I dont think Seam would ever set the params of docDisplay on render +Flume shuts downHave you met any of these conditions in the files that are remaining openBased on rollInterval you will not close the file automatically until hours has passed after the first event is written to the file +Although it might make more sense to go with everywhere +Whats the circumstance? Maybe a somewhat ugly workaround could be added that reads the value first if that fails recheck if its possible to get everything via +Once I fixed that they connected just in another one of my applications I have only one type of member and they all use +As discussed the patch is in a form that can be applied from the root directoryThe problem lay in how are dealt with in BPE +The patch is applied to so this JIRA could be closed. +Ignore the previous patch +Committed to trunk +See comment by Sylvain Leroux on Nov PM +Here is the code that I updated to support this feature +Committed into with revision +Checked this revision looks fine +The release note is clear and +For the record the problem is not related to the the problem also occurs when no caching takes place at all +isnt it possible that by the time filelist is invoked the indexcommit of the version is gone? In that case no files would be we dont reserve the commit after an indexversion command +So looks like the original test detects the known restriction of the implementation support and I suggest to close the issue because it isnt a bug +panel +Running mvn clean verify fails to build with Maven that it is working with Maven alpha +Yes of course dont hesitate to use it +I propose to refactor the subproject only if others implementation are supported +Issue caused by tested patched jar provided by Kevin and it solves the problem +Hi Artem Can you submit a failing test case? I will look into it +Patches are welcome for treating column names as column labels via a connection property +My projects uses I cant find the string anywhere in the logs +Verified with JBDS +JES is supposed to run on devices and it contains both and +Hi Brianworkaround you suggested seems for me better solution than special instructions +It seems like the build is picking up the library but maven is being run with a different JVM +The patch is for the x branchTwo things should be noted The commit is fired even if the delete doesnt really delete any document +theres probably other stuff from SMACK that isnt used either +Misconfiguration +Did not see that was already open for this issueThanks Bryan for pointing this out +I think the question is answered +Heres a patch to ensure that we dont get false failuresfrom test BThe patch also includes minor cleanup of the in the urlTemplates test app +This is the configuration to run on must choose as the when running the need to get this working using meclipse +Thanks for the review! Committed this. + +This is a and will not fix any of the underlying problems +Shall we delete these references off the apps home page +I agree that my patch will make the class equivalent to extends with nothing extraMaps can have null keys values and null mapped to null +We have to change many files +Do you know why the other file wouldnt work with hdump? Do you think its related to Tika not parsing it too? Tika uses the Java library so Im wondering if they are related. +There is no corresponding logic since when local it is assumed that you are not starting any server daemonsWhat do you expect to happen here +unfortunately the netbeans dont allow for such layout of projects +In fact theres a load of other tests that fail when TRACE is enabled on and they all seem to point in the same direction the Fix Versions to TBD as were delaying this one until the next CP. +For there should be no difference +bulk close of resolved issues. +removing dav stops using the webdav wagon and uses http +Dart please attach the patch and binaries for this work to this JIRA and go ahead and deliver them +Guys check the code in CR +Issue closed. +GoogleIf you give up and instead want to submit a patch that downloads this version of jcifs from svn then youll have to include a patch to and so that people can push the version to their local repository easily after running ant . +to commit the latest patch and finally we end up with the original assert from the original test fail +I changed this to improvement +We are building software that includes the capability to pool LDAP connections +Nope still get that error +Ill take a look at the possibility of making the consumer asynchronous as soon as I finish the documentation +If not it will still replace the directory by the fileCould you please include a test for that +I fixed the transaction feature +I was incorrect fragsize works with the default highlighter +Patch that contains the updated files based on comments from Kim and Stan +The problem is that the build did not finish yesterdayCan you try and checkout the trunk source and just domvn Pfastinstall then go to distributionbundlesjaxrs and build and cp it into your webapp lib? Youd need to remove a soap import inthe though +Patch with addition to Lucenes also the bundled was removed +Your note in the test says the parser should catch this problem +should I try calling again +This is todays behavior +Dont want to put too much effort in here just need to address the leak +Never the less Ive tried it and it gives me the same results +Fix ported to and +Doesnt seem like it should be correct +Punting to for now +Working file from the web +Needs the queryparser piece still +Attaching patch for Hadoop + test for you see we have rejected this change on many past occasions. +You may be right +I am closing this JIRA +deleted them out of the xml file +Even if regions it could add a huge overhead to instantiating an a user may only use for a single GetAs for the I think it makes sense if there is a cached value that gets evicted to do a of an extra row so you have a good chance of picking up the other side of a split +change in enhancements as +Yeah just tried killing server that was carrying ROOT +I kind of figured the system would have to be in some pretty dire straits for this to ever matter anyway +I also think we could do with some more words around the other pictures in the overview page and perhaps splitting that page up so that the UML is linked +This is a dup of +I havent tried other databases yet but will +Committed revision trunk Committed revision . +Will somebody try this patch +I ran the two failed tests locally and they passedApplied to and to trunk. +address aarons I wasnt clear +Applied patch ages thanks +I think we need somehow to move to let statement be driven from the actual rscursor being closed as you tried initially +If the reduced output is less than a quarter of the allowable RAMkeys keep it and recombine it assuming you get at least a quarter buffer full of new keys +Patch applied on Branch Branch trunk +Ive file jiras for the two categories bug in Hadoop +Seems to be the CREAT flag doing it +Thanks Marco +Please use the users list to describe your use case +I just committed this to the vectorization branch +Improved and final patch removing dependency on source generator from XML data binding classes +Tried this +Sent the data but did not verify the return correctly +This error occurs while building with gcc you need to upgrade your version of GCC +Please reopen if Ive misunderstood. +Thats correct patches only work with Im afraid +If needed please intert a separate issue +me apt is disabled by default +When SVN properties are set properly Subclipse works correctlyI think proper SVN settings as displayed on screenshot below should be mentioned in release notes and some How to document if such a document existsIsaac if you think it should be not mentioned in our documentation just assign this Jira to me and I will close it +In my experimental branch I provide the methods such as entry exit etc that JUL supports +Bulk closing stale resolved issues +its used in the bootstrap process just not from the entry point +Recommend closing this issue +Only registering endpoint for all the https +Q Do you want them to be ASLed? then I will add themYou are supporting Result Set Management in retrieve +It was the solutionSorry for forgetting about the separated view +Please let me know if this makes any difference in the various use cases +At least the reports dont fail with an exception now but they also list all time intervals on Derby instead of collapsing and reporting just the maximum which will make these reports far less useful +Anyone +Ill investigate on my VM +I see what you are you write this up for the CDI EG as I dont believe the spec is clear here +Didnt completely read through everything but is idempotent createappend dependent on the client generating unique names? It is an assumption that should be very explicitly communicated to clients but even then will something break for someone else in the file system if some rogue user doesnt generate uniqueAlso even if YARNMR gives a service like that what about all other applications which directly only deal with the DFS +Closing resolved Release +an improvement cant be a blocker +Hi JosefI am unable to reproduce this issue in trunk +How would I submit this for review +Thanks Maysam for great feedback! Sorry my response was delayed +WONTFIXING +Compiled files only? Source too? What about documentation manual source of manual javadoc +Committed v +Corrected previously attached manifest file +the patch +What was the scenario in which you ran into the above issue? I need the views for all the nodes plus the relevant section of the logs +Marking closed. +Its both there the landing pages and the console updates +derbyall and ran fine except for some strange network server timeouts that dont seem to be related to this change as I get them without my changes as well +Yes I overlooked that the type had attributes +Bug has been marked as a duplicate of this bug +No it shouldnt at least the parser is not able to process delcarations like that at this have not yet verified if the import directive produces source because I am currently working on a Polish machine and dont know what the Polish is for view source +The next drop would use numbers generated by the revision manager module thus providing snapshot isolation +does the skip test still cause issues? Ive never had trouble with that one. +Just seen on one of our dev machines briefly but it didnt hang around and has vanished again +Turns out BSD basename can do without the s +I suggest the index metadata should be stored at jcrsystemindexes +after discussing this in team we came to the decision that this is a issue in ACF and we will not do a bug compatibility fix here +I wonder if could be extended to allow recursive ie +Interesting idea above + push to +Now set is set to true +Marking as wont fix +Please test this on and if it is still an issue +We need time to look into this issue and review if when and how we support the specification +scaveatissue +Thanks Jukka +cancelling patch to address the above mentioned issue +Make title more illegal state is simply reset to the correct value now. +It is loaded by model as unrecognized xml +Yes its +bq +I would like to thank my employer Sonalysts Inc +Close. +Here it is +compute max in construction or configuration reloaded +Sorry about the confusion +See following attachment for a Test add into the version control tuscanyjavasdoimplsrctestjavaorgapachetuscanysdo full path version of the in revision +Having it in our hand we canfollow the track and see where the problem could bethanks AllThanks for your effort +Shouldnt the unescape also handle percent sign prefixed hex values like +Downgrading to Major +DanSince has been broken out from contrib Ive moved the bug over under the new mrunit project +Really exciting that its finally fixed +the performance improvement should be measured and demonstrated to be significantThe initial concern was the incompatibility of serialized objects between Lucene versions +Thanks for the feedback +In this case the service invoker is which ideally should rollback the transaction +Here is a screen shot to be specific about the issue +on patch looks good and tests pass for Scala on applying this to the trunk now and to the branch too for next throws errors while compiling +Hi OtisIt is a lucene bug +The attached patch fixes all warnings that I got on Mac OS X apart from unused parameter warnings +Classlib knows nothing about vmmagic package +Committed revision . +Updated release notes +do you have a PR I can look at +It has to be noted here that class Abc comes from another Maven module and no source is available for Abc during build of Xyz +Just saw patch +Committed to User all issues that have been in resolved state for more than one month without further comments to closed status +JakeI was wondering if you still need my help for this this is on my list to do this week Ive been busy with other things +In the test the child class uses a protected method to overwrite this + update storage +i checked it against RC works as expected now. +Is it possible to do it with the history though? It currently appears that all the history was lost +helps when there are complex dependencies +I tried a version of this with Click and jiratesting +Patch for attached +Proposing for release +Integrated to after went in +Thats a unit test failing +Patch to provide additional information about the are you still interested in working on this? Im thinking of enhancing the a bit more cleanly debating whether to work on your patch or start from scratch +However the slew of tests added in were not carried over +Committed thx +This looks good then +Thanks for the additional info +Prasad I left some more comments on reviewboard +It works for polygons +Can any one review and commit these changes +Thanks alot +Applied the patch at revision Thanks regards revision +Thanks Mamta and Yip for the review! I will try to figure out why the NPE is related to transaction handling +I put it under mandatory +Hi Robert Sonar should be used along with a scheduler like Jenkins Hudson Bamboo +Ive already begun building the RPM and have asked Stan to look at the code to determine if EAP should be updated to use this +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Investigating. +I may be missing something obviousNah but I probably need to see a doctor for that dyslexia +Will have to look into it though +Also while Im changing this part of the system anyway Nigel suggested it would be useful to attach the session id to DFS metrics too +See now resolved +Patch for ignoring unexpected enumeration available on r +If it has been set and their are no artifacts then an empty Collection would be appropriate +Thanks Sergey! +I was misleaded by the issue title +Yes with separate fields we wouldnt have to parse the keys which worked for Text but not for binary row keysFor METAHSK I would have table startkey endkey +Ill look at it first thing in the morning +Committed revision I have been going through and applying languageXML or Java where it is clear as wellFor those tables with alignment issues or text overflow I have had to remove the full classname from each table cell and put it as the header of each table +Its only the +I am not sure if the in list optimization is coded to do that +Committed. +Ill start working on the testsuite +This issue does not affect the community project it was only due to a missing style dependency in the product branch +Logged In YES useridI had the same problem with tried and its fixed +Reducing priority since no example application sample so closing +I have committed the and patchesI believe these are the only outstanding patches for this issue +This is some awesome synergy +It should be targeted for s not clearTry clicking on the linked task and you will see it was resolved in the JBM SPCP release +Ive looked over this patch +Sorry I felt that applying the both patches should resolve the problem +This was also not documented and added to the release process and any future releases would not be deployed to the npm registry as a result making this an ongoing issue which is not resolved +It works fine now. +With larger commitInterval this should be only an edge case +Same feeling than Evgeny +Logged In YES useridfixed by jae +Posting a patch +Also the ones that are already developed by others +Should I cancel the previous patch and delete the old attachment +Verified in JBDS H Fedora +At this revision the attached patch from Stuart is still vaild it got things working for me +I have doubt whether some other problems in the related area +Looks like Nagles anyways. +The leading slash i needed because these images path are absolute from the site applied +Marking because though new facility its too good to pass upAlso marked it so lars would take a look and consider it for +guess I posted the bug right before your checkin I saw your bug first +All new classes have been documented +The WSDL importing relatively on XSD that in turn imports another XSD +I love pull requests if you have it done already +My final merges may not have been quite right +Changes have been made to the build and in the code so that when now instantiate a service client stub we pass the constructor the service endpoint url which is stored as a config value in the database +definitely agree +version but I didnt find this jar out gems my warbler just noticesones such as etc +Are there any other paths we can consider +Sorry duplicate post. +Patch v modified test since we could create empty store file now +RM scalability now matches +In an Insertion ordered get makes no structural modification and if we synchronize putremove we should be fine +Also added Jody to the would say it looks good but I would move the first check for if the state already exists out of the synchronized block to avoid having to sycnhronize for every access +Yes we can hack around and make the TCK work +Will +This is resolved althought I would probably remove the factory in favor of session property +to commit thanks for working this! Lets get this in asap +To achieve that I removed the EMF dependency from and moved the ecore classes to the project since thats the only place where they are used +Seems to be working fine. +Ad David has pointed out the root problem is in yoniks patch just avoids this bug by eliminating the generics trask the underlying s the patch with a test added that normally manages to trip the exception. +The test is not really HA specific since the dups can occur in the config itself right? unit test tests what the function does +a second interaction would register new subcribers added to the old ones pending to shutdown but waiting for the messages +I will right now thanks Bruno +Atached patch which fixes the naming of attributes en test patch applied Revision revision on revision on for committing BerntI intend to close this issue within the end of this week unless I hear objections +Since there are many more users than implementors I dont think it is a good trade off to makeI would like to propose instead that we do a class that extends that we can use for testing +It says producerData now but should be messages Minor nitpick Rest of the metrics classes have name as their constructor argument except which says tuple and its not a tuple +I see the size on my problematic matches RC when I transferred the new files over I must have neglected to cd over to the test directory and transfer over +so not sure how it will be less useful +We log all exceptions this way not just genuine thrown onesetc +This service might only be provided by the secondary namenode to reduce the load on the namenode +Uploading src and test patch files +Yes with Roshans published workaround I would recommend closing the issue if Johann agrees +Clinton would this patch address your needs or would you also like the change to the interface +The reason that I chose the other way is because the confluence export isnt very good +Changes committed to trunk and manually tested by running broker and ensuring correct files were created +That IS really weird +and it breaks no testsIncluding the patch now +I am understanding the patch correctly the changes allow you to target repair to specific data center +The patch will definitely change the numbering of the existing constants which means there would be some clientserver semantic inconsistency if the two sides use different versions of Thrift compilerAnother solution more compatible with the original method is to maintain an occupation list +Thanks in rThanks Jyoti for the reminder about only concerned +Hi RaghuRather than attaching cant you attach the SVN diffThanks in version using address like this Servicewsdl gives local ip not public Servicewsdl works fine we can see original wsdl file +Ive used this jira to test some precommit env settings +Thanks Cheng I just committed this. +You can post to the user list to see what other peoples experiences are +New patch addressing the ran successfully +The issues of bug are independet of the ones Tim recognized so I removeddependency +Your suggestion is even better Joshua +pushed +It works most of the time though +dims +bq +youre my hero +Ill take a look at what Ivan suggested to come up with unit tests +I had simply misunderstood the prerequisites +Thanks for bringing this to my attention Mike +cfh file was included +Will close as soon as its verified +Patch fixing the problem +As a result its possible to write a protocol without a modifying the Dynamic code +Passing into Generator implementations now so they can have access to project pluginDescriptor encoding etc +Can we not make that a prereq? Having Python is a very reasonable thing +This is important for the use case like avatar standby +Latest remaining incorporated in version. +Defer all Harmony issues to +It has nothing to do with reflection I just needed it and added support in a way similar to how short class is handled +I have verified that this is a bug with the webapp +The ticket was not fully covered as the else part of the statement did not have to be followed either by another statement a compound the previous check has been updated instead of creating a new one +ready for review +The default value is MB +At least one problem is that clone doesnt take into account relocating tokens only moving tokens +Sorting also accounts for a large portion of the number of times data is written to disk so the corruption could have happened there +Second try +Sorry about that +Running patch though +I have now though upgraded to using in the central masterI have given up using it shouldnt now be mentioned anywhere associated with +This entry can be either an IP Address or a hostname +Lets push this to for now and most likely close it once were closer to finishing with the series +Maybe this isnt the best way to setup my normal tests +In the logs second election also happend but by that time loop also ends and failed +Is this issue done? Can it be closed +Which will certainly break a lot of stuff +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Source code in eclipse projects structure for client and in uses the policy used in this issue +Why not Netty? Bill Burke had some interest in running on top of Netty and that makes sense in my opinion +Setting the components +Not yet this is why I havent set a fixFor version for the moment +Is there a workaround for this issueUpdate you can manually exclude from the dependencies as appropriate allowing the server to start up just fine. +I applied your patch thanks +Cannot reproduce the problem with SVN HEAD r +Whats the latest on this? Can we close? Do the new artifacts introduced in resolve your issues +YOU decide when its GA +Marking the work done in as resolved. +Ill need asf availid list of any that will need access to the machine +The is prepend should only fire if a previous is was satisfied and generated sql +I checked with RB and Bobs TZ patch still needs to be applied +Could you review the attached patch +I just tried this and the old pint type works as a unique key +Duplicate of. +I actually think there is a valid use case for this +Working correctly now thanks! +I just committed this to trunk. +The trunk patch works for as well +Rremoved deprecated classes with no error could you verify and close this JIRA? Thanks. +will reopen after requested info provided. +I am now able to use to run some of the samples +Since some replies just came in today and no one day works for everyone lets do Monday pm PST +DN fixes a number of visible and bugs so it will allow us to improve metastore in more ways +Without checking the new entry I couldnt know which entry was the right one to removeThis change reverts it to a remove then add when the hitCounts are different in the right direction +Also I think we never should delete user files on database creation +Fixed. +Patch applied to CVS +pagepageIndex pagesVarpages are dataScroller attributes not dataTableCheck and resolve +I also agree with the issue of naming and would argue that Spring does the same thing +Default seems to be derby and postgres +So you can safely update from toCan I see the stack trace without the use ofWhat version of Xerces are you using + +Did not see this when build was made against alone by running the followingant clean package tar +when the execute method of the statement is called the connection object will be flaged with executionMode and once it is done with the operation it can come back to the idleMode +Will work on the fix +Weve decided to keep them around for comparison with the new impl but it should be clear that no work will be done to improve things on that side +Very old and deferred multiple times please if this is a must fix +not shure if this is correctLet me know if you have any problems +Since Im new could you explain how to migrate and how we can work with the Wiki +So Im closing as Wont fix +Thank you for your help that would be think that going forward there will be need for more support and sent an email to dev mailing list to probe amount of interest +I think the tests had RPC timeouts that prevented it from passing nothing to do with the patch +And rename to format and name issues +bulk close for the release +Do the the also support caching +as the feature file is still in the old format committed +I dont see a inAbout the component the idea is that clients can be represented as normal components leveraging the normal wiring policy etc +Thanks Navis +Works fine thanks. +The problem is user dont have privilege in default the workaround is to give a with privilege +Hmm +I think I can work around the placeholder problem for now +confirmed this is working +The patch file from running svn diff on folder +Closing as per Henk comment. +Patch looks ok to me +I also needed to add the to so that the object would have access to the +We validated this by setting the block report interval to minutes and sure enough showed up in minutesWe think the checksum patch caused this because it enforced a blocksize on +if the real question were debating is whether we are still considering switching Cassandra to Avro before then we should probably discuss that elsewhere +thats fine i only suggested it because thats how does stuff +Patches are ready +Committed patches a and a to trunk with revision Backported to with revision I dont expect more work on this issue +Thanks Emil +Thank you for your finding +I think it will be better if we allow duplication on group ID and not on group NAME +Maybe we shouldtry to handle this in and a larger buffer in cases likethis? I will not have the time to look into this until next week though +Test cases for the getset and behavior are checked inBoth and are checked inThe spec dated has the changessvn commit m Add property to PMF +With that I will sync up the fuse branch with the apache branch from where apache was released +What would the impact of this be on resume scenarios? Would there be any circumstances where the API choice would end up preventing resume because it forces the use of separate domains +There are several other lines in the code as is so once the discussion settles down about style we can adjust all of them as necessaryThe code style differs from the other projects I work on so Im fighting s autoformat +Users may be less confused of which format to use +I need to do a review of in general which I dont have time for in was fixed with the latest batch of documentation updates. +Upping this to MB would mean that the default configuration would require as much as GBThe defaults should aim for the low side of average so that things work for most folks without alteration +Wo this fix the scheduler page is not showing enough information for queues +What about people that use Castor in mode iow people who do nboithave access to a container that supports JMX +What about adding to cfstats thenSeems reasonable. +Please verify +Even I would love to see this feature present in Maven +That shouldnt normally be a problem so if clearing it causes new problems we shouldnt be clearing it +Just copied the similar fix in +Regression test added in rThanks for reporting the problem. +Looks good to me other than a small suggestionYou may want to have an overloaded handleResponse function that also accepts TIMEOUT and let all the default not bother to pass it +Fixing it myself +I think as a safeguard we should probably also disallow +Galder is working on thatThe fix I committed for only ensured that the initial failure isnt fatal and the cache view installation and succeed the second time +Maybe a sample demonstrating the issuesolution is good enough instead of providing a testcase +thanks +I noticed an indentation issue in the patch for +Great catch Jason! ThanksIt seems like we are missing a hook in to the condition and trigger another merge at the end of the merge itselfHere is an illustrative patchThoughts +new patch much stronger test a couple fixes refactor most of the leader election code into its own class +Introducing the thread fixed +I can confirm that Java and Java work correctly for this example using the query and configuration you providedIn Java platform limitations forced Linux Java to use a large fixed stack size +How does this relate to? Seems like that is a similar thing trying to define common IO or am I misinterpreting +Could you try a new snapshot? +Those components rely heavily on Unix concepts such as seteuidsetegid supplementary group list group execution permissions for binaries signals etc +Thank you for the complete exampleI have used your test case and have confirmed the problem +On the one hand its good to have another implementation if its providing value it would be significantly better the more it can be refactored to integrate with existing implementations +Attached +I dont think thats a bug so I switched the status to Task and move it toif someone can do some analysis on this with some thread dump we can go father +Ive attached the patch and will commit it unless you have any problems with it +Its a pretty far stretch to say its going to impact performance let alone kill it +We are still awaiting the outstanding information for the Release Notes on this one +MaybeI have seen use cases where several classpath containers are configured for one project +Geronimo has improved its java handling by using the aries jndi code and now the remaining java prefixes are causing problems +Net only class +the basic issue here is that the property view is only update when the selection in the workbench changes +check out and is now fixed in trunk +I will try next week to work directly on my c drive to see if it solve the problem +It now seems to be fixed +It might be better to use multiple parameters rather than a comma separated listdirectmyQueuebindingkeycarbindingKeyvanAlso Im not sure the last form is necessary and it might be confusing to permit it since it is really two logically distinct types of munged togetherWould it be sufficient for backwards compatibility to simply permit consuming from the first form +This isnt going to make it into CR + discussed Ill only implement the characters length exclusion for ll see on Dory how goodbad this will be +It can support correlated subqueries and selection from multiple tables without any join operator now +Webkit seems to not implement the wc spec entirely correctly and does not call the onloadend event handler +that is not relevant for the error reported in this issue +Patch lgtm Enis +in this case should be copied into tmp directory and then deleted +release bulk close +Jozef could you attach the stacktrace +Code looks good +Sounds good to me +Recipes would use the new cache API to allocate watchers instead of manually allocating themhint hint whomever implements this would make a very good candidate to be a Curator committer +No its not the geowebcacheThe server isnt a high traffic server and doesnt show that many different views +Like because the space navigator portlet uses session cache for spaces but the breadcrumb portlet just depends on the portal site menu this inconsistent behavior look happeningI think wed better postpone this to the next release for the same reason on +Thanks Kevin +Nice work +Heres a patch to expose get from jobProfile through +Fixed in the nightly build + +bq +I can confirm it works in Maven but not . +I used the configuration verbatim and it definitely has +Ive just committed this to the branch. +Nice work again thanks! I committed the latest changes in revision It looks like were getting close to marking this as resolved +Verified with JBT. +Fortunately the template and document are similar in form we can use document as a substitutes +Yeah the heap has all the scanners except the current one +That isnt a error +So I opened to implement some kind of caching in +I will review the opened issues and see about getting out soonish +The issue is the name of the enum not the name of the class used to store constants +Sebastian We wont just import these into JIRA as they may contain username space clashes +To solve the problem with the I introduced something similar to that +With the build this defaulted to Applicationsjbdevstudio and this does not have the permission issues that the build used to have +I clearly saw the errors but not sure how that happened +This includes not only Filter at the class level but also at the collection level for and associations +Committed revision +Ive just committed this +As for the rest of potential tooling youd want to see something that gets people going with the executionfactory class and then makes the corresponding translator jar ready for deployment this would include the jboss beans xml file. +I hit the same problem +Would it be possible to make this test according to conventions +Seems like some DPI unitsconversion problem +Can you send us some code that would allow to reproduce the issue? +Fixed! Works fine tested with MAC OSX. +Submit the unit test result for patchThere are some err in result after patchedBut those err also be found if no patched +This template method is explicitly exposed for deciding whether some request should be considered as form submissions or on detecting and comparing the source URL of a request might be quite fragile in some cases +Like you pointed out the schema import is caused by the use of classes such as and in the JSSE XML related factory bean hierarchy +Looks good! makes it even better +Hi AritzIve looked at the document and the github repo but I cant see how I can reproduce this problemCould you add an instruction on how to reproduce itIn addition I see a lot of Spanish in the logging messages +I will refactor the code again so that the history is preserved +I added the notion of +This breaks the build on osx +Also what seems like the timeline for the C client since like libhdfs my guess is will provide c wrappers using JNI from what the discussion here looks likeIf thrift is the way to go then we are looking at creating a tool that takes an ODBC data source and loads all the data from one table to an table +Whitespaceonly diff results from Jitendras run +Thanks for any comments! connetor builder absolutely need revise when the improvement is done that is eliminate all the switchCase +Committed revision Thanks Paul! +My concern was whether you wanted to keep both versions of the jar in the backend too as my patch just ships one version to the backend in case of packages inside +But since tagsoup now uses Apache anyway I am not sure if a separate LICENSE file is still necessary +on the patch in order to prevent the situation +Kevin The new file is missing an ASF header +Corrected now +This increases my suspicion that the attached repro may turn out to be another bug in table compressionOn the related issue table compression was not used +and pass on my machine +I think that is good idea +Ill see further what the error could be +rev contains an adjustment to the parser and a test +Hi CedricFraid not +I came across this when working on and then realized that this particular change could be separated as its own issue so filed this jira +This definitely fits the bill for being dynamic and requires zero configurationThe bean xpath regex etc +Which mojo did you useCould you point me to the updated? the Jira and svn urls you provided didnt work for have a trouble to see how the plugin could affect your build in that way +Lets verify that these jars will be included when we package Tomahawk +Else let us know which code base should change and we will make a patch ourselves +I would still vote for keeping Tomcat as the default but having Jetty in there is still great +patch for needed in too now. +waiting for a wasnt clear if your last comment was intended that way +So Im not sure what happened here with copypaste but I can still reproduce in the exact same way still +It should be fixed on trunk now +Up to folks who are more actively contributing code than I +I agree with Julien Nutch should be described as a +The file changed to load by default +Just to ensure were on the same page in those places where compensation actions are taken but the exception should still make it to the routing engine I will need to catch compensate and rethrow right +Once goes in these absolutely will have atomic behavior +Thanks MateiBefore applying this patch svn mv is needed +I deployed the axis WAR file in Tomcat and I still ran into the same issue +Saw same testing locally +How does this patch work for you +If that so I can commit the code on all myfaces branches +Otherwise we should exit with an error message indicating that you should run rollback using the old version of the software instead of the new version +But it is not possible to reproduce this with the current codegen plugin +svn diff is back +Ive implemented the column aliases as previously mentioned and made fixes for the issues Namit had pointed outOne change is that the patch only requires parenthesis for specifying multiple column aliases for a UDTF in the select expression +Changes are in trunk rev +Since came back clean after implementing Todds suggestion Im going to commit this momentarily based on Todds +With consensus create for each of the platforms +Let me upload the patch with a test case +Konstantin Is there a way to schedule a patch against? According to hudson it runs twice a day +Ive added consistent connectTimeout and readTimeout properties to Simple for both and I wonder why you would configure arbitrary request headers through a bean property +Solution committed to branch in r and merged into trunk by Thomas in r. +created a new task for all unit tests +Fixed in the ESB codebase will be in the next merge +I tested this and it does not seem to be the case my events are being correctly initd with the oldStateNote that if an object is being reattached no snapshot is available in or +I heard that someone says that RXTX will be +Keith worked around this in the AccumuloIts a little more complicated than just padding +Additionally the Forge core plugin registers a shutdownHook with the Java runtime when it starts up +Seems like a flag would be more appropriate +Sorry you are right it seems I have merged those patches too +I assume youre proposing to move something like Utilfile and file into another module? That sounds reasonable +Its in the CVS so why shouldnt it be in the M? +JBDS issue moved +See my links in the services plan for how to obtain and install that +So I fixed that +I just did chance this fix will include in release? Im using Grails and it seems that upcoming Grails will use branch +I have created for discussing improvements to the availability of NN. +I just did another cli install and noticed the followingSUCCESS Minimum requirements for android metBOOTSTRAPPING androidERROR! Could not create a native android project test fixture +Patch is committed in trunk rThis issure remain open until documentation is updated for the new configurable property of Query policy +backported change from trunk to +the scripts dont normally use prepared statements +Still awaiting response moving to next another Hibernate +Committed revision Committed in Rev +commit to address issue that was witnessed on the frontend +We have a few sample apps already but if there is a concrete web stack example that anyone is thinking of that should be clearly supported as a sample application please bring it up on the user andor dev list +At a later point we can choose to do something wise with that exceptionIf the argument is set to false it will effectively try to send the RDBNAM to the server and it will still throw an exception but we dont catch it +Does this patch complainwarn about the problem records + applied thanksClose this issue if youre satisfied +Hand testing of UTF data works +I am not sure about timeline +Ive tried to work out a way round this but am failing so far +Im justdumping the data sent to vmdata into a file as json copying that upto the router and processing it there +gitignore list this has been reverted +No Im talking of the java one +Ah marging instead of margin? Yep with also go beyond docs +The problem is that your action URL has a trailing slash +It would be nice if get in namenode also throw a +This doesnt have to be inline w +fix Committed revision +Applied the patch with thanks to David +autodetect simply chooses byType as default in the present this issue is hard to general advice to solve it is Dont use autowiring by type at all +Would be great if you could give it an early try in your environmentsJuergen +The history of this tool is that before the release the binactivemq script was only able to start the broker in the foreground +for userThis patch seems like something that could do that +Maybe it can be handled with strict configuration both on and Camel sideStill the configuration that clientID above has to be set on connection factory level is an issue +Might even be able to start later today if my final buzzwords slides start coming together +skh putting into the bucket since John was looking at this type of issue will be fixed in general when we migrate the xml relational logic into the engine and model document retrieval as a procedure +our tests are producing OOME like crazy +Thanks for looking at this Carl! +See for backward compatibility with content persisted with DNA +Thanks a ton Pi for working on this +I just ran into this problem +Manually tested it and things work fineI believe the patch is ready for review +not the next one for news concerning this issueIs there a hope to see a solution a day? Is there a workaroundHelp will be really appreciatedThanks +Thanks very much for this patchBest regards GeorgeLooks fine +Closing since fixes have appeared in latest alpha manuals. +No patches to date so moving out of the release plan +Hence it is best to recover all segments on this reformatted JN +I have verified that the problem does not occur before the fix for and does occur after the fix for in and so I am marking the regression flag +There are no automated performance tests +For compression we used zip software +A full test run completed with the following resultsTests Fail Error Error analysis in progress +Fixed in geotools. +This part contains code are not being relicensed but are compatible with the ASL +Check if the model path exists in modified versions of in order to check if the model dir this be a patch? Its not clear what is changed by you and what diffs are drift since HEAD versions have changed +Ill add retrying there and give my people a retrying version of the other code +It works in and it does not work in +And jbdcapi suite runs fineIs the test output file already corrected +Should be fixed as of commit thats the quickest fix Ive ever seen! Thanks +Refactored the main class to several task classes and used reflection to run each task +Should I reopen this issue once enough test exist? +Closed. +Standardized usage of enable +Its probably something we should discuss separately then +I dont like using FI for fault injection +Closing all resolved tickets from or older +Something for you to play with later on +The API? The main content and Struts site dont change very much and the Struts docs already autodeploy +Committed +Bulk closing of issues that are old and have been released or marked invalid. +You are right I confused the twoI dont have an opinion on +trunk already has scout +The last release of was on Jun st which was a few months before this +ThanksSuresh +See my comment Feb +Attachment has been added with description patch file avalon loggingcommons loggingthe Avalon logger for the containerUtil that you mention +This issue is now documented in the EAP release is documented in +The persistence tests needs some TLC +Resolving only in the fetcher is too late to implement these features +Its my understanding is bundled in Oracles distribution out of conveniencemarketing +We have a which returns the scores of a Query +I guess I would expect that it would be pushing it more +Has anyone created a patch for this against or could give me some advice about creating one +Can you add some logging before that line in and get the values of address and port? My guess is that port is a service name and were not properly translating that to a numeric port number but I dont know how to trigger this +We dont need to make a release to make the site changes go out do we? These things are decoupledNot that I know of +In the future try to create the patches from the project level it makes it easier to locate where the patch belongs. +Id hate to miss an opportunity to reduce our style warnings for +Can we get the comments from there instead of putting it toAnd can we modify the desc table commands to show comments from instead of putting a from deserializer +That is a fair amount of overhead on tb +not user +This is important for the Maven reorganization that is going on +Sorry dont know it was few monts agoDoes it work with latest IBM x JDK? If so then its fine +Akin to how shuffle works with creating a schedule for transfers upon failure we can insert the range that failed into a system table on a replica that succeeded via thrift and when the node recovers we can repair the range +Id invest some time to provide a patch for this issue +Attached patch fixes compilation error in JPDA module appeared on Windowsx platform +We dont have debugging enabled in our logs but Ill do and post the output when it happens next time around +Like we do it in browse and finally edit file in editorBut this is up to you +If it didnt produce a DLL of the same name then it wouldnt work +Thanks BobBoth are mineI will fix the st but will keep the d as I still hope to find some time to work on fine can we just comment out the request in like you have with the entries? My walker finds the and then tries to find the associated and reports an error +If you dont have an internet connection on your machine this might cause the portlet not being registered although everything looks like it is deployed correctly +All containers I know of are inherently request when it comes to attaching a session +Thanks +More consice patch that handles the numbers in the properties +Looks like at least one of those test issues is legit +Or p +Things look ok now. +Thanks for the patch Kai +I have attached an example my recommended changes +Should be fixed +Please revert them +What if the src directory was renamed to lang or impl +Added to allow clobs and blobs to be comparable +For example we could prefix it with LONG inorder to preclude any ambiguity with the old persistent format +Ill refer back to ATMs comment that we expect people who are writing audit loggers to know what theyre doingOn a different note I played with the let exceptions flow approach and it seems that the NN catches the exception at some point and morphs it into a + +If the historical information wasnt saved as stated how can this new graph appear to be partially correct? +Closing duplicated issue. +Hows working these days? Anyway +But the problem is is looking for in emptyFile +Knut Anders Hatlen commented on Could we also make mf final? Then we would have caught the error at compile timeGood proposal +So for example has a constructor that sets to nullRight like I said thats a copy paste from the class it can likely gobq +Submitting the updated patch +Need to verify against latest JPOX release +So maybe add a maxThreads cound then sort the by maxDoc and then only spawn threads for the bigger readers and then one additional thread for the rest +docs look good to me +a bit confused here +stuff from the local repository but maven downloaded them again +To get it fixed we need make changes in facet and richdatatable template +but at the moment I cant see you think you could have a go at reproducing this +Heres another typo +SushanthCan you upload the patch here +Try passed +Ive got files marked as needed to be added to version control but there already in version control +Some minor renaming in patch v comes out please submit to Hadoop QA for test suite run +One more change that I have made locally that allows concurrent reads with blockages now appearing in the is to eliminate the synchronization inside the protecting the +Having used a lot in the past I never liked its syntax that allows column positions +Bulk closing stale resolved issues +So if we can do that then filepath and cmdname should be in code font +All patches and the documentation have been committed. +After discussing this with Alan Im going to change the label to Find Pets. +on integrating the patch and launching another secure buildThanks for working over the weekend Devaraj +Just assigning a fix version so the issue can be verified by Eng QE. +Should have something to add soon +The problem is only with the package caused because tools is not in the package +What jaikiran suggests +ReviewedPatch looks fine +patch applied and snapshot deployed thanks much! +Fixed in r. +HTML Syntax error +I seeI will modify the error message +This will be solved with a fix to fixed +I doubt anyone will convince me this is a actually a good idea +In my scenario if I either make the view not matierialized or make the type conversion explicit then the error goes away + implements Comparable so we can order them as we want +If the original payload is an empty list the current thread will be hanging behaviour is unexpected and as mentioned above dangerous +i remove javassist and ongl package manually from then works +Thanks very much for the patch for this +The attached patch now has correct indentation +one merges segments but it is very simple +Ive attached a patch that works +I attached diagrams that reflect my idea of how service life cycle could be managedIt is just a draft but enough to begin a discussion +It does look like those archiveswere merged with the current devhc so I dont thinkwed be losing any history +But it would be nice to skip this in the future releases +Upgraded to Spring Batch toSpring Framework and Integration +This patch fixes a few unit tests which were failing in the absence of mapping rules +I also agree the default for compile should be true! Unfortunatelly the current implementation does not allow overriding the default for hbmjava since it is directly set inWhat about changing the default directly in? If not set within the configuration it defaults to true thenI am not sure about the other at least the run good with this patch +Thank you John +Did you want to do a Nexus migration? +Jonathan I think getting the signature exactly right would be hard with the extra issue that every change to improve the signature instantly invalidates any cache based on the signature +Also I have not yet investigated the implications of using the Eclipse IDE for Java EE Developers artifact rather than the current eclipse SDK +Thanks Brian! +strange +If you absolutely refuse to allow null values to be overwritten by Hibernate you could at least log loudly when you do skip over a null value in the resultset to warn that trying to change that particular key is likely to fail +includes patch will make sure it gets in next release +Since the exception is likely to be the result of a corrupted log I interpreted that to be within the realm of the flag setting a generalization of the intent +Many product stores could all be part of the same organization +Fixed a bug where adds disposed object check to static methods. +Finished all implementation besides notification handling which needs to be implemented throughout +The intention is to throw the original exception when you access the from the requestregards Malcolm in changes for approach as described by Malcolm +Abhijit Meshram +If you see any problems with that let me know +jar files for Seam wont be in sync with EAP jars +Ive made the changes to should work OK now with trunk +Create a an empty file + Username tpeuss +Attached compiled page corresponding to the patch for the convenience of reviewers +Thanks! +We may need update the title of this issue now thx Freeman +The winconsole gem is needed to handle ansi escape sequences on windows but this gem is not explicitly required in the gem spec +Its very weird this hasnt got caught in our testsuite which is why Im slightly worried about rushing a fix for this in BETAIm probably gonna release BETA tomorrow morning CEST time anyway +No response assuming ok. +To run test case unzip attachment and place the app folder under a Web root then browse to appduplicate for input +Thanks everyone! +Of course we still need runtimelocalbinnutch +at last for replacement and overloading +Hi Ivan Could you please go through my above comments and would like to know your opinion on thisThanks quite +This JIRA does not yet talk about the engineering aspects of implementing a +OK lets assume you can get the body tags +so we need a way to be JTS? Use the SAX parser and then punt control over to GTXML when getting to geometry? panicSee IRC logs of March th for a discussion +Also seeing this issue +Thanks Rafal +I think mostly the theoretically methods left in the code are ok +Bumping +In such cases there is no information about the speculative tasks logged in the job historyThis can surely be fixed in trunk +Can you attach an example of your OQL and how you are binding to it? I haveusers from Postgresql to mySQL to HSQLDB and all have somewhat different +For now we follow the same pattern that the other server adapters follow +Make sure to resolve this if you find the latest SNAPSHOT fixes problems you were seeing Need to know when to start thinking about a new release +Thanks for correcting it +And I also test it at gant environment the result is same +This seems like a very odd behavior and I just got surprised by it +Roger exactly what issue are you reporting here +Wouldnt this stuff be a nice contribution to the project +Of course a pool size of zero isnt possible with +Right now we checksum data as it exits the java output stream buffer and enters the filesystems buffer cache roughly speaking + for fixVersion +Can you attach your config? You must be using TCPPING with a long initialhosts list and a large portrange +Couple others like activation etc +Updated patch with refactored interface in and a test caseI thought it would be good to have the old get interface explicitly take the IP address and xfer port arguments +It doesnt seem to have any issues associated with its fixI dont know enough about to comment +I will run the tests with replication factors of and to fill in the gaps on the table and report back +Ive modified the to pass with the new code the corresponding would presumably need a similar change +Though should have added the two cns that do not have a language option +Your suggestion is logical +Actually there seems to be a pretty straightforward solution Ive added a check whether there is a bound to the current thread only invoking unbindResource if found +commited the s not possible to reproduce the issue mentioned in the last comment +When we could get this patch or how long do you think you will need to develop this patch? ve just checked in a new version in which I try to solve the problem following recommendation by Barrett +We should also decrement the value of retries in case we get +I would be ok to look into that later +Well the problem is that if you do true the is not included in the making the from the version unusuable if you want to use tomahawkI close this issue as wont fix because we cannot change the release but I open a new one with the issueFor those who wanted to use the from the release use the three separate jars and tomahawk +I can actually delete this if I want to +Done +While hang occurred in the Xint mode too this issue was divided to this one and for the assert in the opt mode +Issue closed with release. +However CCg isnt a viable solution various things fail to compile with g +Please feel free to this issue if you think that I am misunderstanding anything +This needs a test case +Lars I can work on it if you would not mind +One caution with this approach is that as it stands it will generate Avro data files that can only be read by the Java implementation +the references to mailing lists in README +For now do you think it would be possible to commit the patch without the unit test and come back for the unit test later +I cannot agree with closed status +Go to your WebsiteSettings page set the Default Category to root and report back +not including version ranges at this minimal resolution reporting fixed some bugs local dependency management used for transitive versions to force a versionscope adheres to design to later allow version resolution +Manually tested +Im not sure about the effects of setting the innodbsupportxa to false +FromTwo new clustered profiles ec and are available for environments such as cloud providers that do not support multicast trafficNext ActionDetermine whether the information belongs in the AdminConfig guide in the section that describes the profiles or the Installation Guide when configuring the system +Bulk moved to +Please make sure your environment is clean +Close issue in preparation for release. +Fixed in rev +I cloned this issue since I could not reopen the previous issueI have this issue inSee CXF for example from Kaj Kandler +Here is an updated patch +Ken will we have this done soon +As has also been raised before that quickly gets messy as new data comes in that doesnt exactly distribute the way the first set did across the partition point +The advantage of this is that I can interoperate cleanly with bundles +I think if we use here we should store the nonce and also the incremented valueIf the RS goes down after adding this we should be able to just replay the edit and use the value from this Edit +Martin as a stop gap measure do you mind if I change the relativePriority to ? There is no due date for geotidy to be ready to be used and moreover I need the fix on please do +Attached is a patch for this issue +Determined that client should not be closing the result set in either case +Unfortunately the Package Explorer is not open for extension as the Project Explorer +btw could we then have a deployment sorter that respects mbean dependenciesRegards points out that compex dependencies like this is really to be solved in so Im rescheduling +Let me reopen this and downgrade the importance + have gone somewhere +Yes thanks +Yes totally needed +Your continuum was installed on a fresh db +Perhaps we should punt this back to ripl since they would fail under anyway +committed the corresponding testcase in dependencymanagertestsrctestjavaorgapachefelixdmtest +Im not sure I understand the problem +IT test done +Thanks! +Thank you +I wonder what is next +What i did was i detach theactual deployundeployredeploy using so the actual deploymenthappens in a +Determining if its a valid table is left to caller depending on needDid you mean something else +Also the HTTP headers are not properly set in the was due to the RESPONSEWRITTENTRUE in wrong context +Please reopen if the problem still exists after implementing Darrans suggesting configuration. +I just tested the attached test case with hibernate +Patch made against similar issue also exists in branch and already reported here thanks +A small improvement might be if the databaseName property is null in the and databaseName is set in connectionAttributes to raise the exception databaseName attribute cannot be set using the method set rather than required property databaseName not set +I thought I would try in a non mobile browser to diagnose the issue +PaulI just wonder why we need in the tests of and currently +Vicky can you take a look to see if this is related to the commits inThis is a regression from the previous release expected TCK behavior +This is still happening with for webdav traffic + to +Also please add a description of the parameter used by the plugin in to understand do you think anything more on this one +Thus its not clear which is a better choice here +Thanks Sebb for pointing that out! +Resturning this because theres nothing in the SDO project that provides specific support or restrictions for certain WSDL patterns +I compared the fileformat docs with the code and verified that they are in sync now +Link to the issue that will fix those failures has a dependency on which is resolved in which is not yet released +May I ask if this code still works with Nutch? It dont work with nutch version He still use the Lucene Document and not new ApisIt easy to correctIf you want to use it Take care with this codea fast read you can see was open and never close Exception cath to NullThe idear is good Nutch distribution plugin dont permit to customize easly Index dataThey are something to do +Committed revision . +UI just reads the output of listHypervisors response +Fix checked in +wbuildrsvn ci lib CHANGELOG m fix for Sending CHANGELOGSending libbuildrcoreSending libbuildrpackagingTransmitting file dataCommitted revision . +You can add settings in MAVENOPTS +Im sorry this has taken such an awful lot of time Ive been very busy with my day job this summer and I havent really had the time to sit down and do any proper work on this +Fixed formatting set Severity scheduled for and took a conservative guess and the effort required in fixing this +Could write an scripts with a event that set t this correct behaviour for unit tests eg including real config integration testing surelyShouldnt all unit tests populate their own settings here? Perhaps config should be an empty rather than null? +No followup marking as incomplete +However I still somehow am not permitted to assign JIRA items +However I put all the other stuff in bm because it is specific to the bmpm method and is worth having publicly visible as you can do some custom things with it that are not reasonable to expose through the codec +Romain can you tell me where I can download s test case for cluster and session replication +Test is added for the issue +It no longer occurs. +I will comment when Im done to see if there was success inside of transactions +What if my application is working fine with an JAXB converter and then I add Jackson to the classpath? Suddenly it breaks with no warning + contains fix +Simple project which demonstrates the behavior +A am reviewing this is the right one and doesnt need to regenerate +Attaching image of where I have got in trying to understand how it hangs together +my understanding is that closewait is different from timewaitmost likely it should be a bug that the Thrift server side does not properly send FIN to the client sideI guess some fine tuning of TCP keepalive parameters could reduce the count of dead connections at the server side but that does not fix the application layer bugshould we reopen it? +Committed a fix in rPlease make a check and let me know if the exception is gone in your case. + and Default still used internally but Step and Job now know how to launch themselves +Im glad its working for you now Erin +This time it is the server that can not join ensembleThe first time it succeeds but the rest it fails +Heres another stab at enabling an appSettings property to confingure lognet +You okay with resolving +Well need a something like a where we flush by rewriting an existing store file interleaving the contents of memory or some such to slow down the flush but also to make for less compaction to do +Patch for fixed this issue +OK +i saw that i had checked one item in the java classes and resources +Proposed patch +okobviously I miss this detail +Mark I agree re setter injectionAs Martin said the big issue is around producer methods as initializer methods are clearly intended for injection purposes +yes it can. +move to next cp wait for the community issue be fixed first +It was decided that AS will not include a UDDI server +Looks fine will do closer review tomorrow +Confirmed and exclude list to EJB jar in branch +I have written a patch which I am using locally in order to circumvent the need to run SDOAGGRMBR in order to generate a doc +OK I agreeSo can you help to review and commit it if there is no problemThank you! RegardsRay Chen +It is definitely worth spending a bit of time on this +Therefore Id add a comment to state the fact that the result depends on the implementation details of that is vendorversion specific +Yes regions will stay in this state until intervention from shell most likelyLooking into the unit test issue +thanx for clarifying +The documentation for both cursor and driving query item readers is complete +So am not using envers +I dont think we need to have for now +If a datanode has like K blocks this is going to be a amount of time +has been marked as a duplicate of this bug +I want to understand the scenario a little bit more +not marking this as a regression assuming it is yet another existing intermittent error in the replication tests. +Which means that the root of the problem is Hibernate detecting an actual optimistic locking failure +Note that this only affects our own plugins +It wasnt really necessaryThanks +Ummm my mistake. +This issue has been superseeded by which has an even wider scope of rewriting the whole configurator subpackage. +Updated documentation in trunk r. +Ted is it the same issue you are facingThis failure is on my windows box +Ill have to think on this + +Merged to the branch in revision as fixed based on comments above. +The documentation changes will be made with. +Also please update list I think tellison is no longer a PMC chair +initialize method did not return a view name +Incorporated Vinods offline comments regarding refactoring of code and added more testcases in Attaching new patch for trunk +Hi KimI could see from the temp files that space was being preserved in the merged XML file and then lost in the conversion to FO +Is src attachement to classpath the only feature this is for? I thought you had some templates or something you needed from the distribution +Pushed to master +no longer seems to fail perhaps in the intervening weeks since this issue was filed a bug in the has been fixed +If anyone else is going to review let me know +Ran apps fine on a secure cluster +Thanks Gunther! +Hi Sagara When committing the patch to the branch you have missed to svn add a file +All the changes seem to be doing is enabling warning people if the authentication is set to anything that does not evaluate to a boolean false so it would be enabled for both kerberos and simple iff simple is explicitly set +I tried your and it works fine for me +We probably should +Fix is simple +At the same time app is not ready for production and it is not safe to have the app modify the data in the production database +Then we need to take care of the attachments later +I didnt see any tests checking +Where can I find the enhanced class +A snapshot mechanism would be very valuable for those of us relying on HDFS to hold critical information +Tests passed cleanly for me except for the network ping problem +Patch applies to the branch +this variable holds the data parsed from package scope is very much are getters and access but not to the variable look at the references to the variable all of those get methods can be overridden by plugins +GangTo resolve this issue for is it necessary to port both and patches at the same time +the incorrect usage of the default namespace +Committed to trunk +sync with current Trunk +Not knows that if storage exists the storagedirectories have been set +Then I would need to adjust each there also a way to use classpath scanning to automate the remote lookup of the same interfaces? This would be very useful for us to avoid boilerplate XML code for DAOs at the client side +Opened identical issue there +Merged into master +Im concerned about throwing in a due to the cost and its not trivial to detect whether a repository has already been scanned at that point of initialisation +This seems to still be an issue in by +Whats the status on this? Trying to get ready for release. +We might need a solution that saves that information +thanks for tracking it down +Problem had already been fixed in the +if using datastore txns then this may be immediately and if using optimistic txns then this will be at flushcommit +Added synchronization around access to keys +We will commit it or something like it +Do all the tests in the comments and final patch from pass +There is no violation of the throry that WSA to is the target EPR +In Mercury ordering behavior is defined by a list of Comparators passed to the tree resolver +As a workaround exists and Forge is around the corner this issue will not be fixed +I understand that you need that class but creating all these issues and rushing ahead is rather counter productiveRobert this class is standalone in this patch and doesnt know about the analysis chain +if you wouldnt mind submitting a patch when you fix it for your app well be able to incorporate it for future releases +done using mvdbapache +Running by hadoop qa first +Exactly what I changed on the page the page that got it to start working is beyond me +Please use instead for the fix on the release execution condition +Note there are changes proposed in to relocating the functionality which I also intend to action +Thanks a lot in advance +We should follow the cluster setup +use build i found one more small issue where the strip version got messed up +r fixes this from what I can see +Thanks KristianI already attached a patch as a fix +I think failing fast instead of trying to recover seems like a reasonable choiceUmabq +Hi Andreasas we are a bank we cant provide you with the normal documents as they are internal documents +Im convinced that this particular issue is resolved +IT was committed in and trunk +My fault here +May I ask why there is no progress on this ticket? I have sent a fix via github and I cant understand that nothing happens for nearly half a I can help in any way to bring this forward please say t get me wrong if you believe the proposed solution is inappropriate or whatever then close this ticket and reject the pull request +ok thanks +No need to create a patch the change is really trivialThanks for the report though +Thanks MarkFixed in BEANS module at repo revision Please check that this fully resolves your problem +Just an idea +Patch Attached +Thanks Stefan +Committed to trunk! + applied +Second rev that makes SFNS a singleton +Workaround provided in the issue comments. +Testcases are attached +Thanks +Fixed README and scripts updated +the componentType file that looks for is tuscanyosgi +Thanks most of records recoverdBut why index got corrupted? mybe compress or writer has bug. +Actually nevermind I had a bogus repo in my boxgrind app file and thought the issue something else those changes I had listed are not needed. +The messagings transport param attribute model does not declare that it allow expressions even though it canThis tasks refactors the code to have a consistent model which properly declares allowing expressions and resolving for master merged in master branch +This mode detects structural changes to have source files and triggers a only if required +Thus it would be better if that method were abstract no? Or have I missed something + Thread dumps done after the analysis the lock seems to come from very slow response time in and particularily inFurthermore looking at the class seems to have some synchronisation problems calls are not synchronized findCache synchronizes access to util while only getting an element which has big performance impact since util is static clear and clear does not synchronize while clear does why +But for all those bits that arent tightly specified thats all we can say +Sorry should have said above that this is fixed in JoprGACP not the RHQ branch +So all issues not directly assigned to an AS release are being closed +I was able to pull down some of the Derby artifacts and use them in a trivial test project so they seem to be liveLets wait and see what our users have to reportI wont be surprised if we have to improve the later but let us deal with issues as they are reportedHopefully people will choose the version over since the latter artifacts are invalid +This surprises me a bit because the order shouldnt matter but at least its a workaround +Sending srcjavaorgapachecommonslangTransmitting file dataCommitted revision +I think you guys are doing a great job +Otherwise users will get confused when they download and are unable to find these classes +Can you attach trunk patch +Need to run again now that we are done other changes to ensure nothing else broken +Please review so that I can submit +Could this be a regression? +letting the ops team decide when a data node goes down? This seems like a no brainer from a practicality perspective +the findbugs are also not related to this patch +I made a minor modification and changed assertSame back to assertEquals for null to branch +What performance test is that? On my system the tests take about times as longBut that is using Java on an old XP laptopCan you attach the code it would help to ensure we use the same tests + +Its too complicated for me +Jon But I think it should be in the start of the pdf so new users read the simple walk examples and doesnt endeavor into some more complex examples +ah I missed you comment sorryWe can move this discussion over to if youd like. + Username letiemble +If you repeatedly encounter such exceptions whether your destroy methods are implemented properly +ArunI also suspect the assumption that the AM has a basic idea about resource requirements +It should be possible to this code with little change in pig or other systems +Update to MANIFEST so we export the package +closing +Ill keep this open but were going to have to test other bridges to see if we can mitigate this +Erik Thanks +For example from the simple question to more complicated ones +This issue has been resolved for over a year with no further movement +Added support for single dimensional arrays +Committed to with a minor change using on a list instead of. +just to add more info i tested size of Mb and lower without any problem +Im not disagreeing that it would be good but Im wondering if thats something we really need to get done in the branch +Hi Franco just attach the patch here +Committed to trunk +would like to update my copy of railo to benefit from this fix +I need to track a weird behaviour I see with this patch on looks like this patch causes a bug where we hand out more tasks than we should I need to look closer +This includes the WAR EJB and EAR deployment scenarios +Committed! Thanks Thiruvel! +have you managed to move forward on this issueI noticed that the patch has not been commited and it no longer applyies cleanly against latest SVN +is difference is probably OK as the values for and are fixed +OK applied your PR to and master +I guess weve fixed it then +Sorry +All the better! +This issue is out of date and will not be fixed +Thanks Zheng +Text Filters have been removed +The comments tell the translators that the strings are complete even though they look like the kind of string fragments which might cause concatenation problems when localized +Rather than these various policies it would be very nice if the compaction dropping policy were pluggable +applied +Fixed by +The current plan is to include only the functionality that is currently in CVS with some slight enhancements +With this fix they now doIve run the whole suite of tests and all of the tests that passed prior to my modifications continue to pass +I dont think this list should be configurable we should be able to put in our documentation what is legal and have that always be true and not worry about what happens if someone changes the definition of legal characters +Reassign to me if there is something I should add to the home page specifically in regards to +I am using Sonar and running sonar on my projects from Hudson +Should still output a blankCheersChris +it looks good +Maybe this matters maybe not +Ah oki but that was the one we used to get handed over to our Extension Contexts as well with +does all the required work. +From your code it is apparent that you are using XSLTCs native API which is not recommended +Visual Web Tools that the guide is fully universal for both JBT and +Resolving this as Duplicate and fixing it as part of +I have verified this patch fixes my issue +Feel free to revert it if youd like +especially if no one noticed before now +Committed change but leaving defect open for a few days for comment. +Can we increase the priority on this or at least get this to work with cforEach? Im still getting No Active Tab! errors when I remove tabs via well remove exception and add diagnostic warning for the case instead +Thanks Raghu +Moved to where Mitosis will be to we wont release a is no longer present in the server a new replication system is has been developed based on RC +But if the maven team decides to change this behavior I have attached a patch that can be used +Related to +Id like to just include in release but we dont have dblasby approval but the datastore extras should advertise it +Thanks Vivek! +Enclosing a patch which removes the unused second argumentRunning regression tests +This is great stuff Claudio +About the class +Nope its on my mental TODO but go for it +I have committed this. +Make a longer timeout +Allen if you read properly my previous comment Im not saying we are going to create this anyway Im using a conditional tense we could move +Im not sure if this is okay or just overlooked +Steps for testing Create a sales order Then Go to Verify Pick tab in Facility Manager Enter or Pick list bin Id Then enter the Product Id and depress on Verify item Or select the items which needs to be verified and depress on Verify items Then depress on complete button Shipment should be created with SHIPMENTPICKED status in Shipment entity Invoice should be generated and its status should be INVOICEINPROCESS in Invoice entity +It may not be valid for but in he needs those fields becausehe has to put different xpath +When you have done it can you send me the link to the new versionIs there a place like jenkins or somethingIm on AS in any case so +It should be done automatically i suppose +If you have problems relating to Any you are encouraged to move over to and report it over there +After trying some things i found out that the problems also occur on that ubuntu machine + +What is IMO different to is the fact that the type information is not really needed since it the instance is already there and only needs to be looked up using its ID +And if I need to wait another week our project plans need to be plan to make releases for this module rather often so it would be helpful if I know the time that is required to upload it to ibiblio after the release is done +See also and +Tested on nodes cluster +Resolved in trunk +If I give it an infinite loop there are times when I then cant interrupt but not with the script you gave +I will add comment in that issue to correct in as wellThanks is for branch +Folks can look at their binary data in hfiles easy enough +Great thanks for taking this on +All unit tests pass +Adding a NPE protection and throws a normal exception would force the tss link gbean get started again after the ejb deployment info is initiated +Hi Marnie assigning to you for review feedback Changes completed and merged to trunk +includes showing multiple choices hold on I have not tested on editing yet +Resolving this as a duplicate of. +oh right I see that didnt change +Darran this is now to be reviewed considering Ive committed the changes on trunk for using Netty instead of Remoting . +uh +Is this possibly a duplication of? If so can we close this? +Excellent point I completely forgot about +Net to follow +New translations for attached + are done +Works fine with as well +Fixed an incorrect override that used the wrong signature +This seems to work the best and be the cleanest solution +It has one additional feature in that you will always get the latest version of the subprocess for each process you start +Tickets to have been actioned as at to have been actioned as at Tue May thRT was submitted to update a gif image found in the source +I will spend some time looking further into that solution +Markus what about modifying the default provided with and add our fields definitions instead of trying to upgrade the old schemaThis way it would be easier to compare with the default wed have all the field types up to Ive committed +I see there is an initiative to removeI override this service to insert logic into etcPlease undeprecate dont remove itThanksbulk closing issues that have been resolved for more than a year. +Changes look good to me. +The main merge thread should just keep tracks the mergeable files and start actual merge thread whenever a merge is warranted up to a predefined limit of merge threadsThe optimization discussed in earlier comments still applicable +Updated patch now that YARN is no longer under +FYI I dont have more than what I uploaded in the patch +I would want additional people to test the shell interface through the monitor to verify it works reasonably and safely if were going to include it in +I got your point +Im not sure if the test is obsolete or we really need sstables to be sorted all the time +After all as time goes by and companies complete migration to and this functionality will be less and less used +Im going to try one more commit this evening to try to get things mostly working and try to find some time later this weekend to get things properly cleaned upSorry for anyone that isnt able to build due to this little fiasco +Thanks. +This also changes to add additional whitespace to generate properly formatted HTML +Thanks Suresh + +Obviously I cant speak for everybody but I dont think the most common problem is really people doing a select with parameters +Ah we dont have generic expand and collapse CSS styles +Thanks for attaching the war and source code +Heres a working version of the patch +And if were not supported wed never reach this code +Thanks for reporting Ive committed a potential fix could you please try the latest from svn trunk +So all issues not directly assigned to an AS release are being closed +so they are not available when we go to do the compile of test classes +I am not sure when we started put them in svn path +Please answer its urgent for meThanksRajesh +to AnimeshCan you please commit bbeadcdaffbeafcbcecfa from branch to +There is no way that I can test this program unless you bundle all the dependencies and attach them +Attaching a gutted project that shows the error +Closing issue as it has been released as part of release +Using the default settings I cannot reproduce +This is another reason to have unit tests for this as well +If it is returning DEFAULTID then it is picking up a classloader which is not associated with an ESB deploymentSecond issue yes they should both work so that would be a bugThird issue yes there should be no +bulk close of resolved issues. +Both trunk and Nutchgora branch are both called Nutch +As for I really dont know what to do with it +This quick patch should increase the lookahead value to and also make it configurable via system servlet context and enviroment looks good +Viacheslav please provide steps how to reproduce or close it by your own thanks +Open a fresh Jira issue on any remaining issues for this. +pushing to as it would change the current behavior +closing bug as fixed +Can anyone download that attachment? JIRA is being weird about it +Reinstated Validatable in +committed a small addendum for for only as both were part of. +Will looking into the issue +And different routing policy +I tried the case again with a new server built on and got the following errorgeronimo deployconnect p Connecting to Geronimo server localhostUsername systemPassword Error executing command Unable to connect to server at deployergeronimojmxlocalhost Failed to retrieve stub xception argument must be an RMI URL String or an array of them +Returning the resolved paths is not transparent +Corrected + for the patch +As you read form the JIRA title its really about mitigating a breaking change that this will result in +Time to commit +No i thought you solved it but I read wrong +Is this a duplicate bug of or is it another one +Here are some potential addons using dependencyManagement for project dependencies like +Does the following process seem reasonablePhase Send an email to the dev list soliciting objections feedbackPhase Add an option that defaults to falsePhase At this point developers can and should begin using Phase Default to but still leave the option in order to allow folks to turn this off if neededPhase Remove leaving no ability to turn it off +bq +log session startup removal of discovered that it was incomplete so was not able to start correctlyrootnotebookusrlibeclipse find +Any changes to check with classloaders etc will just make the code harder to follow and the aspectj runtime jar is only about k in size in any case +Do you plan to merge it soon +Thanks very much Anurag for that information +Close after release +Patch for Yahoo! distribution security branch +Captured results of my building +testcae works as was a change with lazytrue to version perhaps this has solved this problem as well +Martin can you please close this bug once you have had an opportunity to assessthe changes +out of date +Fictitious Duplications in Andril if you set the new CPD option Ignore literals to false as specified in the last screenshot your fictitious duplications should disappear +see that youre saying Spring in the title of this issue but this doesnt seem to match the line numbers in the stacktrace +This patch fixes the problem and adds a unit test that will fail when the problem remains +I suspect that the volume is zero +Closing all resolved tickets from or older +If and when Xerces moves to SAX this patch will satisfy the behaviour of SAX and SAX +I am afraid we have the problem with the patch applied to trunk +But under the hut we may use other techniquesWhat do you think +All unit tests pass +This works for me with the latest kit on Ubuntu +This adds a test case fixes switches the src and dst for clone moves the to test and renames the example toIt now passes all unit tests +Thanks for the review Colin. +Also with the CP release the documentation is packaged in a separate +it looks like this mode is never tested +Close issues after release +Please disregard the first attachment patch +I was intentionally synchronizing on that array as if the attemptRestore runs multiple times in sequence a directory may be readded multiple times and I didnt want to lock the whole object +Okey can you push this into the branch too +However since individual retries cannot be larger than the global one needs to be changed to a larger number as wellAnyway the proposed method sounds interesting +Thanks Toms +Java makes all this redundant so lets wait to then +Looking forward +Did anyone upgrade the code to support glassfish v +Attachment has been added with description Patch is applied against revision +it is also not possible to not bind the jrmp invoker to any port +Vikas +Is going to be released anytime soon? It looks like it has been about months since was released +The change is within Smooks the Smooks class has introduced an explicit dependency on jaxen +contrib testsTest failure is due tobq +Fixed at some point on trunk. +looks good to me thanks +The test reprodices the issue when run in clientserver modeThe patches are ready for review. +Your comment indicates you fixed this on trunk marking the issue closed. +Daniel youre on the roll with the wave of bug reportsWill fix this one +is that a? is there still something there that needs fixed +I found the issue we are failing to drop poold readers in +Why would it make the things worse for your project than withI would like to separate migration to hadoop from changes to the library structure as I think we need to have a discussion before hand on how we see integration with hadoop and other libraries +Bulk closing for +Committed to trunk without the fix +Patch with +I will get back now to this issue and hopefully we will be able to solve itThanks again! Brianyour response is much appreciatedSiegfried Goeschl +Dont comment them out +I believe no one just used debug build to run this test before me so I was the first to encounter this problem +better a failing testIll commit this shortly +I will write the information I have collected over last couple days on the reason behind the bug and apossible fix for it but first would like to comment on Mikes proposal about backing out some of changesI agree with the partial backout of +Will do it soonHappy New Year +Committed to trunk and to +around the corner etc +This is related to right? does it mean that when using instead of mahout wont work on hadoop anymoreNot sure if it will work we didnt actually change over anything just yet +Oops +From TomMy suggestion is to remove the section on providing RMI through theWAR +One way would be to use and not would search the application context for beans of type and let them register the my attached patch for details +This is the RPM were running if anyone needs it for testing +As for this patch I updated two files orgapacheaxiswsdlfromJava and orgapacheaxisin +Updated application to Rails using rake railsupdate +Yep fixed +Ideally i prefer we work on the test case to make it more accurate as per the workingYesits needed +Tomcats is trying to do a call while loading a resource +another one with license granted +I got to about hours CPU for million users M ratings and Im afraid that it can easily go up by orders of magnitude if some of the optimizations arent here +I can do it as my opensource effort during free time +Preferably a testcaseAt a minimum Id need to see the wsdlschema and the soap message to make sure the message that is being receivedgenerated is proper compared to the schema +properties or +If there was a nightly build I could test but I could find none +Here they are +Setting to future pending a patch +to branch. +Will somebody try the patch +What is tychoIs it a projectPermissionsNotifications +and extend from it +I updated this issues leading description info box to point to my repo for this code and its evolution +I did not see anything with regard to adding the profile flags on that jira +There can be many reasons for not accepting a job like missing files blacklisted users blocked queue etc +The plugin as plain sources and binary for using the plugin without patching latest codeJust copy the binary to the plugin folder and add the plugin in the nutch configuration fileThe source can be copied to srcplugins please note that you need to change the srcplugin to add this plugin to the building script +When some issue has been dealt with it is first marked as resolved it is closed post release. +Long term I think it will end up simplifying a lot when that happens +We havent switched to UNICAST yet but I dont think we have a good reason not to +moving off CR since described as +Thanks Ill take a look at that +To a certain degree we have to acknowledge that the existing code already allows incorrect timestamps in the face of clock skew +I have applied Stacks patch and also but still have to add the test it supplies and run it as well as all other tests +Perhaps most Lucene developers are not using dynamic class loadingDynamic classloading is popular and accepted in JEE servers +Roller allows an admin to set the sites absoulute URL to be used in newsfeeds +Bumping to so as to remove version +Lets retry to be sure +Verified in Rules Guide and Release Notes. +A lot of our existing AOP changes intercept invocations within a method why is this not possible +Just a guess though. +Please test it +what does rpm qlv hive server show +The first should be the simplest no ability to take snapshots if data loss is possible no ability to loss data from existing snapshots and no fancy policies until we get the basic default and definitely correct implementation in first +Resolved in SVN version to build a Pluto dist on Tomcat +Jason can you please look at thisThe mvn command actually fail on the atached project from the command line when local repository dont have ear and war artifacts installed saying that it cant find ear and war project from the reactor those artifacts are installed to local repo running in embedder fail with errorFailed to configure plugin parameters for committed new embedder to the trunk and it is working with the attached test project +Done decorator added +I have created and fixed a subtask that deals with fixing the issue on the build side +that you fix on commit +Patch applied thanks Kristen! +key validation class would actually be the varchar in KEY varchar PRIMARY KEY +Use a dedicated server for formatting the job history +The tests are now passing +Thanks to all who contributed to these windows scripts +Closing as a duplicate of. +Of course we want validation for all the data that will be inserted +Sorry for that +This was fixed in Xalan. +bulk close of all resolved issues. +If a bean is declared as implementing or but is not loaded as part of an then the setter methods simply wont be called +That bug is also solved by this patch but in a different way +For eg if I have a exploded and within +Ive changed the issue type to improvement +liang Thatd be great +I think the consensus that were coming to is that we would like to do thisbq +Thanks Bikas +The normal Avro reader and writer can read files produced by +What are the advantages of having separate components nowa clean project structure and separation of concerns as reflected by the mk and spi abstractions +In current code we use for this purpose +Transactions are heavily revisited in not only enhancements bug fixes fixes as wellPerformance should be looked at post stuff +this has been fixed now +Will want some more unit testing before resolving this for good and finalizing release +I also tried by creating a very simple EAR project but couldnt reproduce it +Committed revision + +Nick or AlexCan you guys link the you created to this issue? +Patch which fixes the names +I forgot to note that this affects both archs i and x +Thanks Amareshwari +Thank you. +data browser works well. +Update docs +Its still quite a bit but until we have an alternative way to extract text from OOXML files were stuck with the large jar +While I wish we could just be optimistic its important for us to be conservative especially when we dont have comprehensive testing across a grid of OSErlang combinations regular etc +Yeah this could be the homepage overview but when you are in a path that you need it it should show that too +Please attach a sample project that includes a failing test + +Not sure if this is correctable through SOA productization or whether this needs to be fixed in EAP +Findbugs warnings are unrelated see +Im on Mac OS X and I saw some tests fail from a fresh build that should have passedPlease see Alan this looks good! from me +Agree +over an Mapping and a count over a set +If thats the case then I think we should move this functionality to the class +The latest changes in the firstrun plugin are the cause of the issue with an incorrect path for external EAP or AS runtimes for external EAP arent added at adds correctly AS runtime for ASEAP Seam runtimes for EAP and JBPM runtime for +I just tried it again but found out that actually what happens when you hit Cancel in +Im going to generalize this by adding the ability to use iptables instead of jclouds security group calls for +ran all the failed tests locally with the patch and they all passed +Added chooses one of files which can be indexed from resources directory and puts it to repository using webdav connection. +I hope there will be some one who is interested in mentoring to this project idea +Tests pass in hudson. +Lets keep it tagged for for now then that way we have testing plans in place if you are able to complete it in time +On IRC we discussed this and came up with robust options is to have couchdb d configure the server to delayedcommitsfalse for a few seconds before shutdown is to call an os level sync command after shutdown to ensure that anything written but not flushed is flushedI dont know enough about but if it is actually robust it might be simpler than +Details in commit message for r +If we throw at the beginning the issue will comes up again when iterating the iterator while NN fails over in the middle +Part of the requirements for will be to selectively copy descriptions associated with selectedcopied model objects +The spec will be changed to resolve this issue +It makes sense to me given the background +Lets get rid of those ennoying warnings forreal +Hence autowiring by type will only hit the scoped +Please set the Fix Version +This patch enables users to see the optimizer overrides specified in the sql as part of runtime statistics info +I had to openclose the session in between operations to replicate this fully +Thanks for the patch Jimmy +fixed in the latest snapshot. +the attached patch solves the skinny war problemthe patch deletes all +Heres a simple data file that I am using + example showing the issue with global depage with pure JMS paging has now been removed +I had set the target release to because that would be ideal +Committed at subversion revision Touches the following fileM srcref +Close if looks good +Apparently Michals using snapshots so that might explain why he does not have that issue maybe its been fixed already +Nothing new other than a few minor bug fixes and the newimproved BPEL reference manual +I like this trial +make similar changes +I just have one small comment +That should be it. +If user wants to edit this file he should open it in the I rightIf so we can close this bug +Can you the patch? It is not applying cleanly +Yes using to generate the docs works quite the problem are the missing libraries in the CLASSPATH +I think I have resolved all of the issues in Katheys commentEverything runs fine on my machine at least +Fixed the code to take service level policy from the policy attribute of the enableSec element where as if one need to configure the endpoint with message level policy then he can use the attributes inboundPolicy and outboundPolicy to specify the two message level policiesBut this fix is not yes complete because it needs a fix from has fixed the this issue and I am going to change the rampart version to SNAPSHOT so that with some changes this will be fixed +Hive shouldnt know about Haivvreo +Added patch that adds logging of MEP properties and full message body +Will be completed for +its the thrown by Previous outgoing and new incoming message +So solution for installing under root is skipping eclpse configuration generation from installer +Wernerdoes it make sense to add setName and getName methods to the Cache interface given that we need to name the caches to wrap the various distributed cachesOr is the name simply an artifact of implementing a distributed cache which is hidden in the implementationWe could possibly define a interface which extends Cache and declares the setName and getName agree that using the class name from the mapping file is a good idea as it will be unique within a +I agree with methods deprecation in release +Oh and one other thing to mention the code also uses which is ONLY in JRE +Is this related to +Didnt notice the problem with the patchResubmitting +Test case and fix added to trunk in revision . +As of r you can now specify multiple entries in the to ensure that specified accounts are never locked by the security example updated to reflect this change. + +Hmm maybe I grabbed some changes that shouldnt have been there +Patch added +It should be better with openid but I cant guarantee it +Do you have a simple way to reproduce this issue +ok fixed and released new version of the plugin +continuum +Maybe we could do it only when the file is renamed anyway such as when it is moved to +MatthiasThanks for the patch +runTest runs the test +testvpcnetworklbrules has failures +Should the request be automatically posted to the jira +If the previous value is not present locally without the remote get it would not be brought and hence you wouldnt be able to guarantee that the operation will work as expected +I may be in trouble with something that uses only instanceof I guess a service may have custom logic about the encoding that goes beyond the simple return this means the services would have to work against an output stream instead of pure java method calls like now or not +Having test it in your production system for months is great +But there is no other way to find out whether two sessions are the same +I would expect the just happen once would that be a performance problemActually we are thinking if we should switch to integer key to save spaceI wouldnt post strong against to this rollback but I dont see a significant reason for dong that +Integrated to trunkThanks for the idea and reviewThanks for the patch Jie +Sorry I should have mentioned this in my last comment before you made a new patch but can you also add a unit test +Im going to add a core plugin to handle this issue +It was backported to the branch as r +Ive reworded the comment to remove the absolute path +Also committed into branch under you Raymond Feng! +In our case were using a remote broker in another JVM so we cant the broker instance +Thanks Amar +Nicholas any comments on if this applies to old pipeline vs new pipeline +Go ahead and patch it +invalidCDATA or xhtml namespace must be above +Yes I did +Aha +The improvement comes when we release the artifacts to Maven Central +Patch committed +Awesome +Also please feel free to tidy up the code it took a while to find this and I didnt spend much time making it prettythanksThe code looks as tidy as the rest of the class and it works +Fixed in r +thats not acceptable its explained in the instructions +Thanks Aaron! +I need this functionality in a current project that I am working on +It looks like it will be for the release +for the patch +properties file inside +Have you confirmed that the checksums are correct on your newly downloaded jar? As in have you actually calculated the checksum locally and verified that it matches the checksums published on the Tika website for the jarAlso you seem to be using gcj once you have verified the checksum could you try running with openjdk the oracle jvm +Patrick I missed your comment in my rushing on Friday +Pete was it intentional to break validation with the new XSD +I am investigating it +In many places Im careful to document expectations about changes the position marker or the backing array but I can review for that +Attached the correct to Testcases +Henning I will definitely move the repository +Yes Im working on this now +The whole point of was that you would not have to rely on the default etc +Of course I was the one that added that parameter The history is patrick added support for validation so that the background process wouldnt happen right away but unfortunately it forced you to always use validation +Here is the patch while there are a lot of lines its mostly configuration changes +This fixes the to go + Bug has been marked as a duplicate of this bug + +Furthermore the new pattern is more generic than the default value of I doubt that someone will ever hit this issue during migration in practice +I am actually my string and include the password as wellThe only thing Im unfamilar with is the parameter from the manual Column of the table used to split work units is that similar to bucketingpartioning? Can I skip that the test table is not very largeThanks for the suggestions +Adding to the number of errors should happen only in the case of unexpected events say the disk is full or a network port is already used +Do you suppose you could try to run master and see if my fix resolved the issue +If syntax highlighter recognizes over languages and I assume XML syntax is included in this list why syntax highlighter isnt used for XML snipets across documentation as well? +Moving out to move back to if this will be work on soon +Is this expected to be fixed in any future releases? Environment Process Server and Axis +Nice idea Ill start from that +But Ive checked it in with the functionality in the optimizer for mainly to preserve the optimizer framework and give an example of how to use it +It doesnt look like this is testing serialization capabilities so why not construct the table from Java code in the test? If it is testing serialization I cant see how you have created the +How is the status of this bug +This jira corresponds to the changes in common +have reviewed Yashs patch and I believe it correctly resolves the problem +Help in resolution is appreciated +You must have enabled git support within Roo +Looks good to me + +These are at least very small and they dont share any fields with the parent +Additionally Ive added an cast to show sorting by the boolean value column +Workspace dependencies are now resolved based on exact version match so you can either increment version in the workspace or close those I was referring to disabling workspace dependencies +Ive realized theres a hole in my current implementation +Changed the package structure for gridmix system tests specific classes +This avoids multiple roundtrips from dfs client to the namenode +I believe this is a duplicateany automatiic publishing should not go to the main example site since it will overwrite examples GA users already depend on +Please dont use tabs to indent we always indent with two spaces +One more thought on this none of the filesystem will implement creation date so I propose that we remove this feature from the API supporting only modification dateCreation date falls in the same category as things like block size and replication which are also mostly unsupported so maybe we could treat it the same way +Maybe something likeOther than that I think it looks fine to go ahead and convert the other tests that execute pig scripts to this so they run in local mode as well +weve added track statements idle timeout flush strategythanks for the comments +Bugzilla has been closed as not a bug. +Is there an option to pass a command line argument in the configuration? Maybe that way the can be specified in the s plugin configuration? +apologies for my ignorance +The patch creates a new page called Updatable resultsets in the admin guide under Differences between running Derby in embedded mode and using the network server +reading the thread dump again it seems i have only one thread deleting and a lot of adding ones +Im closing this for nowUri Please reopen this if you remember a valid usecase for this feature +I have prefixed the +I certainly intend none I just want to understand the security implications of this change before it is included in an official release +What about having a selectbox in the table that can be changed +Thanks Nicholas +to the site and use the generated xml to aggregate have uploaded a zip file with example of my Dashboard your comments will be its a maven plugin and for Maven a dashboard plugin have been already made by Vincent bad +Please add release notes mentioning new passed based on patch v +This patch is contributed to the ASF with the permission of my employer + +I think the key lies in the Right Ive mentioned a spacial base class a few times and this is why +Minimal test case which illustrates the problem +one solution would be to compact the files in the bg ala bigtable or mapreduce sort +Please reopen +Currently there is no way to list the information of since we dont have anything similar to ls ld +Thanks Gary +Its not there any more +it is the release candidate before the final that is supposed to be released next week +Committed to trunk. +this happend in as well +Otherwise +I make use of attributes for support and +Any news +If its possible could please create these patches for current Axis trunk instead of tag version +I ended up restarting the server and it all worked fine +That might involve merging the currently separate long arrays for high bits and low bits into a single array + +WDYM it has passed the validation phase? Did you run changes +Have taken out the legacy sections about datasources collections etc +Hmm that may have been a touch unclear now that I read it I meant that I added it to my project and built it but I didnt add it to the apache codebase +This patch should be used not the other attachement +Let any compactions etc +The determination of FORRESTHOME works correctly if the symlink to forrest is a relative pathname +Really weird +Code never needing this package is not affectedSo at the end of the day it is not an implementation dependencyCreating a VM specific fragment bundle would be another option +Correcting a few more errors I missed the first Lars +I would therefor apreciate if you not only duplicate mysql scripts for all other databases but instead use their availbale of fixing the scripts I improved classes quite a bit +Or push to +Issue Borders issue in login Lookup forms are too small and the new theme does not fit well in Back link does not work at httpslocalhostpartymgrcontroleditcontactmechpartyId Ashishs comment on user MLOn the search result form sometime it shows the link for ID fields on mouse hover but sometime dont show +will commit later today by patch the same changes +Please dont start this yet +the attached forums thread to discuss your ideas +closing as clean up +Sushanth did this Jira get fixed with main reason for this error is that does not get set +agree +Pat Nigel agreed that we should be able to produce Maven artifacts +I also added a root file that meta redirects to docs. +Thank you for help! +So seems that the problems remains +Will do in a new issue. +Joost no it really is a bugJeremy could you have a look +BruceDoes the uploaded sample show the problem in RC? dims +All the changes are correct. +In JBT Version it works is BASIC but default in SOA is set to FORM +Sorry! +But due to testing before and after Chrome had cached the buggy version of from and did not detect that it changed with the update +This should work even when the values would not all fit in memory +Thanks Lester. +Moved to CR +Changes look good +View with less R +If this option is set to true an implementation will choose performance over RAM +Thanks for reviewing! Ivan adding timeout annotations for each new or changed test case is a requirement in trunk now +thanks +All the interceptors that put a onto an exchange bail when presented with a GET +Can you add this +Recently on with the default heap size we get while running +It should instead add them to the context class +Simple patch to fix the to out the issus for the release. +I wonder if this patch will improve your profiling results? thanks! Ill try to create a custom build and gather some data +Thanks for the reminder +Applied. +Warren encountered this same problem running the Staging Tables test so it wouldnt appear to be related to recursion +Bulk closing issues resolved in JEXL. +Please let us know if you are still having problems and we can reopen this issue target fix version so that it shows up on the roadmap right +Also I agree with your statement that we need to know how to use JMX in a Karaf context +Unfortunately I need the XML output so I can pipe that into +But without it theres no chance I could help anybody +WellIf you dont think the difference of is critically harmfulwhich was introduced in before I think it is OK for you to commit +r +There is a HRS running alongside every TT +I have not yet tested on Java +into trunkCommitted revision +You can manage config or view installed features or plugins +With this it is working fine +Sorry +Can you set trace probability to and get a trace +Forgot I also have the right to resolve issues here + to branch +Cant say I really understand the behaviour of but this patch worked as expected with all tested Maven versions +Now Fixed in the CVS +Any results so far +Hi Jonathan the logs unfortunately do not report anything directly related to the two failures might have to add some logging to the code and then wait to stumble into this one +Both plans work for me nowrev as this needs to be applied to trunk as well +Once I compiled from the I found a compile error +Nope I disagree on this development rather put these methods in CRS or JTS helper classes +These look okay within the ESB server +Committed as part of +But I didnt actually end up using it in the test case because I wanted to throw the actual exception encountered so that the stack trace was printed outIm running derbyall again so hoping to check in tomorrow morning unless there are issues with the Kristian Ill fix this + +Increasing the priority since this makes very useful code completion functionality very problematic +Changing the test code so that lookup searches for something is not there reproduces the issue easily +Thanks Janko and David I learned a lot about JPA through this issue +Yea that looks like me sorry +I havent worked on this part of the codebase before +Hope this one succeeds +We detected the problem when Oozie submitted a job with a hive token with a job +This fix would also render moot the need to have the + to trunk +The problem with the current design is that we try to piggyback things on the timestamp one for versioning and the other for ordering +Seems like a reasonable request will support additional mode for +I uploaded the new website on my personal space for the guys to checkout before committing the patch +I believe these tests unlike all other db tests are oblivious to the presence of an overriding file +list currently will not return anything since code is commented out +Im to commit the bandaid patch until we can get smarter +Do you have a testcase to contribute? if so please attach a patch to the issue +Next question +new patch lgtm r +Lets wait until we move them to do that +I seem to have been mislead by one particular issue which claims to have attachments but there are none thereHowever what still cant see is any indication of attachments in query results +Yup The changes are done yet to test them +. +The python script does this on linuxGreat! Ill add this to my sh script +Setting even a bogus agent name is an insignificant effort compared to the further complication of the code and configuration options +Any plans for that ticket +This went in as an Incompatible Change in +Already covered by +good pointPerhaps the better approach would be to call get in the constructor +Easy fix just a missing dependency +uses this to tokenize its initializer string. +I also made the print curious string on exception from BTSTC moreI think its ready +Committed to trunk +data graphs etc are all examples of Java Spring apps producing images that would require +Thank you for the fix. +works on staging +Are you saying that we should let the monitor run even there is a bug in it? we should check other monitors to do the same thing better create a base class for monitors implementing this logicTook a quick look is not doing the thing you described +I have not analised all the changes but as far as it seams to me most of them are about valueable improvements to castor +Patch attached here +Hi Colmyour patch looks fineAbout the problem you highlight above Ive experienced the same only when ext is dropdown and also played a bit unfortunately with no luckFinally I was expecting that when checking the mandatory condition text box was automatically set to true am I wrong +Hope that is fine with you +but our plugin snapshot repos are separate so users wont get them unless they configure it so I dont see any harm in switching to cpl as long as that doesnt cause more frequent update checks +in any case the script is in Python patch committed in trunk rev +Heikothe merge of the instance cleanup has been completed to the trunkCheck it out again and let me know if you experience any other problems. +Added two new methods in revision Ill leave this issue open for some days for potential feedback dont see any tests in revision but maybe I missed something +Tagged failures. +I attached third patch +Uncommented the appropriate tests in and which now run fine +Please have a look at it and let me know whether it works for you +Patch was applied in revision . +Hm +sorry i think i missed this would the uri be harportdirfileinhar or harhdfshostportdirfileinharthe first one is not opaque but will only work with HDFS +That way if the underlying stream is a the parser can still find out and grab the file +Again I intend to commit tomorrow if there are no objections +So while the behavior is not really a bug per se the change certainly does impact a lot of software out there and unfortunately the onus is on software providers like not Sun to rectify the problems this causes +IT to rerun the analysis in verbose mode to get the list of available languages is counterproductive +The parser chooses to represent this conditional as a negated exists rather than a not exists which are semantically different +Add a patch +Sounds reasonable +I am going to generate new patches myself and start reviewing them +Im using grails and jQuery Plugin version guyssince jQuery is coming out within the next weeks Im working on the new version and will fix this for guys I fixed the more bugfix to go a few tests and the release will come tomorrow +Submitting +Also the manager and node name can now be set as properties and can be passed down into the wrapper classes +Please see my attachment. +Thats a good suggestion and would work +The img and frame tags are fixed +Also made compliant with your patch I think +if an address starts to get full then the producer can be automatically throttled to prevent it sending more messages +I didnt feel I understood it well enough to manualy apply itI have added the null graph store transaction additionsI wonder if the best approach is to focus on the SPARQL Graph Store Protocol for bulk addition of data +I think its high time work started on this +Though I think having output collector reference in operator class is bad design in first place since they dont belong there just a bad implementation +Now jar and test already had a descriptionI have added one to cleanemma should not have a description +But if its only a single entry in logging config Id prefer to have it fixed it would help testing the AS console by keeping the log readable +bq +This issue involves the time out portion only +The rat tool is happy now +Thanks Stack for review +Do BodydoBody is used inside +Improved patch from +HTTPS filter added. +Agreed we dont want to cause those that dont need to to reindex just to try and be safe +time to close this one now? +the log file +for the original patch +What do you mean by a remote thread? We should be using interposition in all distributed cases which means the Synchronisation will be registered with a subordinate coordinator +so you can resolve it! +And the sql scripts does not contain SQL statemnts to create identity related fixed in as it was incorrectly closed by Thomas as it should be fixed in the next CP candidate which will be based on a newer version of jBPM +The real type can be gotten from the instance of the variable +This isnt quite a duplicate of +Its a fairly straightforward workaround and would lessen the problems that people run into like +If the multiple submit buttons are leading to a that need todifferent settings for validation you can now use the insteadAlternively if the submit buttons are forwarding to another action as Marcusnotes the initial gateway action can have validatefalse and then otheractions can use validatetrue as appropriateTed. +Please use mailing lists before opening issue especially in case of issue with documentation +If you are around today Yasuhiko and can run a patch againstthe TLD they are kept as XML documents hidden in the userGuide directory +The Page link from attached application will non works +I think sdecorate is a neater approach to field layout +Resolving +I absolutely understand your point of view as well as I understand your change requestWhat I wanted to say if the case of enums is to be changed this should be done for the whole API +Somewhere somehow there is a reference to an object of this class in your application +Its probably going to be a while until we can migrate to and having all these available without having to drag our own components around would be really niceOf course since I wrote this Im biased +By installer you mean Update siteThe metadata of the various plugins requires xulrunner bundle to be thereIt wont neither install nor startup without those plugins being in place right nowThe only situation we handle now is if xulrunner libraries for some reason cant loadThere still need to be an actual package for the matching osarch combo which we dont have for all thus my suggestion is to make them optional +ds file is generated and is updated +Should not indicate in any way that someone other than the PMC or committers have a say in how the project is run +Im fairly certain this is a regression as well but I cant verify that at the moment +i therefore tried toaddress this at revision and removed the ignored test from the at revision please let me know or reopen this issue if the problem pops up again +This patch fixes this particular issue except for the following problem apparently does not register the DEL key so if a user mashes DEL instead things break but what are the odds of that +May be we can use pattern to load the classes from some configuration of on Teiid or VDB module it self +I need a description of the fix for this issue for the EAP release notes +Committed revision Will also to +In the absence of further information the assumption is that it is a duplicate of. +Ultimately there is no end to the features you could add +AFAIR it was local absolute path in the child instances file instead of relative generic path +RLR allows you to specify all possible paths and that would guard that path + +For linux or solaris users this should make sure that no block files end up being dropped by the application. +Resolved +Closing all resolved tickets from or older +Thanks +DavidThanks for the feedback! I am reopening this issue until we address a more flexible approach for handling any Object type +I have hit this issue also +for the to trunk +There is no need for testing here +Ok following your suggestions and +While a SIOOBE is not really helpful a seems to me mkore appropriate +Patch implementing bus wiring by a +this bug requires has ability to modify template you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +I tried copy an external jar into the EAR thinking Id just choose itself but I cannot do that since it has a +I then ran it again with a path with +the patch works +resolve until actually update fixversion to +The last issue has been fixed but unhappily a javascript error is thrown with IE IE and IE +Ive attached an example +Is this example broken? ivyInstall task seems to ignore the from attribute +A manual workaround is pretty ugly. +Do you usually look up your beans by name in Java code? Ideally you shouldnt have many such lookups but rather rely on bean references for most of your wiring +Hmm now I see an attribute which is what the forum discussion was about +All are successfulThanks to everyone participating with this issue +Thanks Dan for pointing this out +Test no longer exists +nobody is maintaining that project though +Chris Actually in your case couldnt get the remote portlet info and throw an exception +Attaching and with changes to the following filesM srcdevguideM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefThis is just a first draft with questions here and there +Will address these forms of issue over separate smaller and more divided in futureClosing out parent JIRA and hence closing out this. +Thank you for your contribution +Can you send to us a test case +The only real user of the compressors apart from streams is the codec pool which seems like kind of a hack to me +Good to know it +test code I used for investigating this bug +uses a and a +Automatic checking for updates is turned on by default. +Makes sense +Anyone interested in tackling this or shall I assign this to me +Thanks a lot for this enhancements please verify it +i committed the bugfix but i will write a good test before resolving the issuethe test i uploaded here is no good its huge and undebuggable and dependent on the scoring system +some fails and some succeed +There is no such there are more problems like this +This is something that will be done with the implementation of the +I will be back in the office on th january Kind regardsOliver Strer +Attaching patch to fix issue by bumping to +If Jochen was going to weigh in I assume he would have done so by are all against and I was really only and mainly wanted to have folks make sure we have the right behavior. +Here is the updated patch +changes look good +felix what about this issuedid you have time to look at itor shall we resolve it as wont fix +Ill add this one later today unless there are objections +styleClasses +checked in converted test. +Sorry duplicate +it looks that cool thanks +Patch that destroys the server process if the shutdown command failed +Thanks for the patch Gregory +method down a line +Hmm +Currently awaiting changes in the integration code to allow for async calls from jBPM into the you link the other JIRA as a dependency for this issue +Am assuming your approach would achieve this. + seems to have methods setLastname +The correctness always comes firstI feel this is something important so we should do it right in the first place +Is the only thing needed to make a config xts the addition of the extension and subsystem a la the patch aboveIf so I question the point of at all its just a file that will inevitably fall out of sync or require a lot of work to keep in sync all for the sake of something that could be trivially explained in some docs +Im currently running tests on this +Ive updated the docs in svn pending a push of the site. +But I may be wrong +Closing all ResolvedFixed issues lacking a Fixed Version +Added Updated junit testsStill missing junit tests for Add login to some components like previous uploaded file is not correct +I agree with your easiest solution the meaning of t option is changed to how many trees should be made per mapperBecuase this problem can be vanished if we can get the collect num of mappers from I think this will be or is fixed on Hadoop fw +sounds good committed + +added tests for closed as resolved and released + +I fear the changes from und are missing so the problem is still therePlease look at the patch file for the missing codeIsnt it possible to reopen a bug? +Juergen +Thanks again +I did try this and almost got it to work +Yeah a little ambitious but we have at least months this summer i confirm that we can really do some thing useful +Move it back in if you think differently +The section on home vs +A tidy fix +We may need to create references for this but I am curious how the reference is used by the Referencable objects +I believe we should handle both of them +bq +Thus I edited the part with the example data in my bug description +Alejandro with the methods you are proposing an array or Objects created to pass the varargs +batch transition to closed remaining resolved bugs +I was more looking at testing a use case of this new components is that possible out of the boxOtherwise as you say the test can just check that the components are present but if the test is also an usage example thats much adding tests as part of makes perfect sense thanks for the explanation! Marking this issue resolved again sorry for the noise. +Adding in another even though it predates some of the ones currently used just seems like it would slow down testing and not improve the overall quality of the unit tests +for this third should I expect you commit +I believe this is fixed in by +All works fine task and to the workflowWeve now had over issues go through the workflow. +Doesnt look like the DST files are used anymore so wondering how internationalization was added +So all issues not directly assigned to an AS release are being closed +I should have realized by the file sizes +I have still not understood the issue properly can you direct me to some pointers to start reading from +Any progress on this issue? Or any pointers where to start +This should not be done until Groovy no longer supports JDK +I confirmed that this is a Java codegen bug related to generic interface method +Still happening but tested on a slightly different platform +Im not sure how to address DNS with your approachAlso I think the best place to resolve and record the is in the fetcher because it has to do it anyway +Updated patch that applies atop +includes git branch functionality is in +Im not completely sure this will fix the same problem you see but it definitely fixes A problem If you could include a snippet of your varlogmessages from the VPC router when you see it get misconfigured it would help as well since it has all of the info on what was passed in to the scripts +Actually the IRC snippet is a lot changed between b and b in the area of pinging and netty suspect the problem has occurred because of different versions of clients and servers being used it seems that the JBM version number was not changed between b and b releases making it harder to track down errors like this +Yes even if we have enough memory I dont think there is any special advantage of including authentication server functionality in NamenodeWhat we are actually interested is examples of embedding kerberos client functionality as a service provider +If someone needs to fix this for they can reopen. +This is not a bug +Ill take care of this issue once we merge the and related branches to master else well have a nightmare on our hands w N active development branches +However the exists inside the resources folder so the has to be changed to include the version instead of +So I cannot simply add a optional classifier argument to each group +lock per heartbeat could be expensiveWith my answer to it is just read lock and it is already to be granted in heartbeat behavior +Very odd that it fixes this problem +Cargo block I am using just incase I am doing something wrong +Assigning to the maintenance lead +Regression tests passed cleanly for me +Sending srcjavaorgapachecommonslangbuilderSending srcjavaorgapachecommonslangbuilderSending srctestorgapachecommonslangbuilderTransmitting file dataCommitted revision Opinions very desired I moved Maartens register in appendStart into the append and append and changed appendInternal as I mentioned above +I am going to commit this change shortly +I can still SSH into +All quickstarts verified using Sun JDK bundled AS as well as Sun JDK downloaded AS. +Regarding the syncasync thing we have two options here +Please review and apply the patch +NOTE need to verify whether inserting duplicates can cause a memory leak can plug if it can +Please let me know whether we should change so that null will be considered as itMy only concern is that if the fix version is or higher we might be confusing the users with the default behavior across the versions +I cant assign to this by myself +pity? Wouldnt it be better to have them on all the time +This is very much a wagon issue since it does not download byte side via http +But I think leaving them behind like you did is useful for the sake of +Closing resolved issues. +Added option to specify the deployment of a webapp in wci with respect to the initialization of a servlet. +Should this be reconsidered please contact the EAP PM team. +It seems the geometry is not parsed and the reprojection code breaks on it note to say that EPSG being xy is our internal convention +Would be nice to get the same permission groups as for ODE since its a subproject and has the same committers +is there any way we can bundle demos with the data directory itself rather than backing them directly into the webapp? So that when you change data dir any demos that depend on that data dir go bye way i can think of is if we had a separate module for the demos that rely on the release config +Just give me a couple of days +Thanks Sharad! +Im OK with the separate files approach as long as we have clear documentation about what the different combinations mean +Patch against the point the trunk was branched would replace orignal patch +Jesse for the sake of clarity Im reading your as a on a new postXXX method that accepts an additional Throwable parameter plus a deprecated postXXX with that Throwable in the context for compatibility as that is what was proposed +It seems that we cannot perform a reload of the connector instance simply reloading the framework keep instantiated connectors into a connection pool and doesnt provide a way to retrieve a specific connectorMaybe the last possibility to solve this issue is to have a feature to dispose and reload all the connectors pooled by the framework the core can expose the rest method to dispose and reload all the connectors the console could provide a button onto the resource management page to invoke this methodWhat do you think +Thanks Holger! +A in this case +Server is expecting an encrypted entity +Verified on work properly. +Attaching a repro for this issue using generated by the testAlso attaching jdk output +Rules are clarified on what root means and how to do relative links for a resource in Section Make sure we implement them also section An WAR application client jar or EAR can define a persistence unitWhen referencing a persistence unit using the unitName annotation element or persis deployment descriptor element the visibility scope of the persistence unit is determined by its point of definitionA persistence unit that is defined at the level of an WAR or application client jar is scoped to that WAR or application jar respectively and is visible to the components defined in that jar or warA persistence unit that is defined at the level of the EAR is generally visible to all components in the application +Base on changes it makes more sense to take the profile approach + +patch looks have committed this to and above. +He is talking about nightly build not the old alpha release +Hi AndrewOK great thank you for letting me knowKind regards +The patch looks good to me thanks David +Closing. +ha thanks guys! Its nice to have appreciative users + +New patch addressing comments +before exploring the possiblity of parallelizing some code can we explain the problem statementIn the entire codebase there are very few instances of closeHook usages +Alex I didnt see your comment sorry +During the Oscars the HTTP cache in front of our Solr farm had a hit rate +Server will not SASL respond to a client until it sends either NEGOTIATE or INITIATEOK use client NEGOTIATE to replace my empty INITIATE +On latest myfaces core impl code we have this package as importedresolutionoptionalthis solves the problem you describe so Ill close this one as duplicate +Thank you Daniel! +For now there no possibility to contribute support for tagx templates +New patch which addresses the review comments +Fact is though that the use of causes metadata fields that contain a comma to be treated as multiple values when instead they should be kept intact +Karel of course not +A little more research is needed into but this is what Im going with at first +Theres a couple of changes in the latest patch +Consequently I dont see how the deadlock situation you describe can happen +So there should not be NN combinations +What I worry about with this fix is what happens when the line crosses the entire display +See for the VM +Good to see some in these net testsPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +DWR validation is working fine for me with FF but the submit action for lesson is still not workingIE still throws some really disturbing error messages which I could not yet resolve +I think precision is incorrect +Dan Your compaction files wont go away until anything using them has also gone away +See +PallaviDoes this patch account for Teds concerns? Id like to get it committed +To add a new core on a new machine just use the core admin api or preconfigure the core in to join as new replica +Alex Stefan I want to draw your attention to the fact that this issue has Resolved status now but only case is actually fixed +ack issue as fixed in Deltacloud commit version aadfeaecfcaaeabdTested the deltacloud client using Vsphere provider with two datastores +Merged to the branch in revision . +No Current state not STARTELEMENT ENDELEMENT or ENTITYREFERENCE exception tooSo this exception exists only in Camel versionOne more request can you also try with the latest CXF snaphost? I ran into similar errors two weeks ago with some Camel routes I was working with and ended up fixing a bunch of things in CXF along with some issues in Camel +Checked into CVS. +Verified in released beta build. +Inorder for that to happen make sure to set the taxPromotions flag in the table to Y +If this is something that would make it into the main code Ill make it production worthy and submit it for inclusion +Covered by +This seems very much like the necessary bits arent present in the complete jar +Throwing an exception is odd +Id need more information to work out why theyre having issues +HiYou get an Error assigning to ! from your program right? +Should I come up with another patch +Thanks +New mechanizm is using in +We need find the split having the partition keys and only submit that split to the task otherwise there will be duplicate results +Only issues that are actually fixed should have a fix release. +A best effort is made to allow additions and changes that are not critical +Committed to trunk +Moved spurious Release Note to commentbq +PSEmail for Hi ShaunI looked this ticket +But up to this point we were able to make our own annotations to the +Much harder than you might think +Id also be happy to commit the change as well if youd rather +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +orgapacheuimaconceptMappersupportdictionaryResourcefor repository +This is a critical piece of the companys platform offering that designed to increase the developers base around CDH +Extra credit if the remapping logic could be factored out and shared +Just thought you should knowThanks Mark +The same for Chinese +Otherwise the patch looks good to me +You should either support a community schema in which case the schema defines the attribute name and you get some level of useful interoperability through a contract with the consumer or just spit out any old rubbish and forget about what its supposed to mean +It works for me on Tomcat and on TomcatIll file a bug report on Jetty cant get Jetty Beta to runI find the behavioural change implemented by your patch rather aggressive andit makes the transformer hard to use and almost worthlessAgreed the change is very aggressiveI dont agree that it makes the EUT useless because nearly all site links areones without the hostname in them otherwise the webapp could not move toanother domainname or have a site be accessible under different domainnamesits the containersresponsibility to determine if the specified URL could be part of one of thecontainers sessions +Would it be good to implement this patch with this sort of moregeneral framework in mind? It seems like this could affect the API as wedwant a more generic way of representing scores than theprimitive float we assume now +Then I can proceed with a tentative implementation. +Load in Squiggle +but I dont know where to share it +By the way does that work from the command lineIve been working on fixing issues with dependencies and still have to fix some issues at startup time +I verified that authkey passed up by the later versions of was correct and that it could be decrypted properly +you choose to wire users input directly to a property evaluator in which case what you see is the expected result +Thank you +Also added all the required constants +See Bob for details +I did include the basic flow of it though +The ant target succeeded +Can someone assign this to him +Could we perhaps pick up the version from If we supported versions attached to the package names then we could add some initial spec versions to our default configuration properties and users could override the defaults with alternative versions asjavax packages dont always vary with JRE +Will provide additional patches as I uncover more sensitive code blocks +Closing JIRA. +Here the patch +Thanks Pranay for pointing this out +I just committed this. +googleinto cannot do it its against the gmaps license +applied to is outright strange +convenience +Backported Remys fix of removing synchronization block and using a instead of for the runCommitted the fix on metadatas trunk so we need a new release of metadata and modify EAP component matrix to use this new version +NET its a Java app +Created a screen which shows the list of payments +This can be simplified to on drop we dont do anything with the actual files +With the above changes for the patch +Added in the base contrib directory to prompt for local m of the parent should any changes occur at least until we host maven artifacts +Attaching original patch files as unified patch +Verified issue is resolved with fix applied in +Added comments about potential data loss if thrift server crashes with coalescing increment turned on +This type of fix needs to go into a or be postponed +Checks to ensure compatibility of user specified schemas with the generatedinferred schemas +for fixes hadoop version to that was over due for quite a long time now I will restart the vote for a release using as its base +Stripping down embedded zip to produce standalone to ensure common base +version spec for should be removed once new scr plugin has been released and incorported into after release +This is how far I arrived Scott just in case it saves some time thereIt was working for the css generation +actual problem is in plugin which mixes several platforms support in one plugin and tight coupled with plugin +Probably insignia is what was meant +be careful about different licenses behind the drivers +Resolving the issue since code is committed and there are no legal issues with it. +Thanks Suresh and Arpit for the review and commit! +We should probably try and just bring over original RPC from hadoop anyways wholesale +Dont you think we should correct it and add some kind of consistency before adding new features? What do you think about adding some kind of validation for field names? Shall we open a new issue for that +It still has only +Im not sure that nested includes within nested tags is required behavior forthe release so Im marking it for the family +From a functional point of view the patch is almost complete and it only needs cleanup +Adding a test for multiple ZK quorums +good catch +The patch for was merged to branch also +Granting is performed when the patch is submitted by selecting a radio button +Could you help me understand why this is usefulbq +Fixed for metadata and referencing modules +inQuestion on how local cache works here +What I see in the screencast doesnt guarantee no bugs but Max has directed me to replicate using the simplest interpretation of it and in that situation I see no bug and am unable to replicate anything like this +Thanks for spotting this though +Added entry to for Solr +Ill discuss the HEM issue sometimes this week with Steve and all +fixed +Can you confirm such a connection Todd +Bulk cleanse of EMB issues moved to legacy version +Have made the changes required for affinity tests +Can you also try to run your sample on browser if it is possible? It will help us to find out what its origin is +Can you attach a small sample application context file +If it passes Ill check it in tomorrow barring any review commentssvn status outputM javaengineorgapachederbyimpljdbcM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmaster +Need to implement new toFloattoBytes and toDoubletoBytes withoutIf anyone wants to do it say so +If the workbench is restarted the problem goes away permanently +Fixed in df. +Either way Im willing to go in whatever direction the group decides +Attach patch file for and prepare for Metamer portlet and Showcase with and PBR extension included and error on attribute change still occurs +Arnaud can you look into this pleaseThanks is no longer supported by the XML Schema propose to deprecate all the types that are not supported anymore and this bug I suggest using Duration instead +If use creates a mysql table through createSchema there is no way they will ever get srid information set + +Fixed as part of the new class +Im on it since diogo gas ciommit rights on it was on repost this on Jira some days ago and you set the permissions + +This means that cooperating processes may use locks to coordinate access to a file between themselves but programs are also free to ignore locks and access the file in any way they choose towould you please try them? Thanks a lot +Committed +Sorry +My build and test runs were clean and Icommitted the change to the branch as revision +Fixed in +Assigning Christine L +Attaching the patch +Did we get this solved +Assume closed as resolved and released +Spaces vs tabs is one thing everyone agrees on and where theres never any question of aesthetics So lets start here and maybe some day ease in some checkstyle +I will take a look to escape the character in URL in Jline +I wasnt able to fix it quickly and didnt want to miss the boat with the next RC +I do contend that the act of validating SCXML documents is better placed with XML editors code generators other authoring tools etc +To simulate rejoining of nodemanagers I did not restart the nodemanager but used kill SIGCONT to resume execution +well has been fixed so can we assume this issue is fixed too +We should look at this for +Ivo Not really since theres no code to actually load +A is now part of +Too many workspaces +The time is taken outside the test method the est method works with an empty String I will not reuse the buffer by setting it to +Environment updated news on this one? Its pretty annoying for me +push out to +apply Thanks a lot! Best regards Andrew +But then realized hudson hasnt come back with results yet +Does this work for you +Thanks Ill plan on doing the actual import Tuesday morning GMT Gav +It would need to be nested +Thanks Ian I will look into this +Ok when I feel the code is in decent enough shape to be submitted as a grant Ill follow the directions you supplied +I will close this issue soon if nobody else expresses interest in tackling itThis is part of an effort to close old issues that I have reported +Have started to make use of the activemq resource adapter +Thisfix adds the validation to our. +Im working on it +I very dislike that path btw +interesting do you have a test case +Yes perhaps I could fix this problem +If such node doesnt exist it creates one and copy prefs from node package +As client depends on both of these artifacts mvn dependency shows the vmware jars as dependencies for client as well +When service called through the client the expected exception thrown or no error +Resolved previously in +I wouldnt worry about this case +But you dont want the users to be able to guess the directory names and grab each others tickets by looking at their +The problem is an infinite loop on info +Im using the when submitting the issue I couldnt see any other versions of acegi plugin in the drop down +LorinCan you point me to where camera calls out to the EXIF implementation? I didnt see it and may have missed it +It happened on the client side while writingclosing a file +The other replicas will send with an offset that doesnt exist on the leaderEffectively it seems that we will end up losing some successfully acknowledged produce requests +Will be released with version +Ok we got lots of comments from many people and everybody seems to agree that should go into contrib +Moving out of +Theres also no cron script to automatically refresh the file +applied. +Ill take this +The reason to keep all tests in file is we make sure all the ptf tests run before every commit +JaikiranHow much is left on this +project files etc should not be contained + for the patch +The best place to start is to remove the Hello records from and create the war with a containing the root element with nothing inside of it and then use the Admin portlet app to deploy your repackaged war +I prefer the latter +A nit needs to be updated as test case failure is +I am super excited about this tooThe biggest problem with coffeescript is no debugging +I saw you created an extra issue for the last part +This worked for me +Simple fix for review +Does a data store play nicely in a clustered environment? In my example in the initial description I have multiple each with its own repository but pointing at the same set of tables in the databaseThe suggested feature is simply meant to give jackrabbit a bit more robustness in situations where it is controlling the connections with the database +If no one objects I will commit it tomorrow +ooo does docbook or something +seems like a bug to me +Patch to remove unnecessary attribute +Can you run your test case on the latest sources to verify? +The behavior of the is still the old one +I think the goal is to avoid making the allocate memory for tasks for a job that causes the limit to exceed a certain quantity + javadoc seems to be existing issues from trunk +Has the correct bits for been released and will STS automatically find it nowLOL serves me right for living on the cutting edge but hey thats a developers life +The patch has been appliedPlease verify +Not fixed +Not able to add primary storage to zonecluster +Danny I dont understand if you do think it is good to have the workaround moved to the or your instead prefer to leave it as is and have each mailet altering the content to workaround Javamail +committed the patch without dependency check with revision and made two further minor cleanup with revision and +We will have to do something more to make it available to hdpdefault +In SOA triage we decided that Kevin will close this issue. +Claus Ill pull this list together soon +do we know why we set to false all the time in REST doesnt know if it should add cache pressure to the RS block cache or notbq +Next add functions to stop the daemons +This is quite common in large enterprises +The mirror still uses nginx +In the trunk branch it should be fine to reassign the same region concurrently +This fits in nicely with every log segment belongs to a single epoch +Some quick checking suggests that these numbers will be difficult to obtain within Java +could you supply a test casein other words some code which you dont mind making public that demonstrates this bug +John is there anything pending on your side? if only QA is pending you can resolve the issue so that it gets back into Sanjeevs to go to testing +Same patch just the common side +Thanks John +To default the list to include the status can make this call quite slow particularly when you have few hundred feeds processed distributed across colos +Thanks Hong! +it looks like this patch is doing something w describekeyspace and super columns but not doing anything to prevent us from adding inconsistent metadata to the we be using subcomparator for columns of the column family marked as Super instead of using comparator +srcitjni builds fine on you that that JNI example with your environments +to the patch and committed to +Closing per Bryants approval to do so. +Changed from! null to StephenYour patch is in trunk at revision +Thanks +bulk move of all unresolved issues from to +Since bbw are only added to targets the premature completeFile should not happen due to add +Id prefer making capabilities changes but it would be better to make them in +Fixes to applied DBC and committed revision . +I do not understand why you are changing semantics for Combiner this is incompatible with Java +Some further issuesOne of the directories and the packages it contains is misnamed of the filter stuff necessary? Garnett added the code into ext and cleaned up a little +Changed my mind wont be included in unless specifically requested by users +We have done much to improve the situation but there may be places in the code where are known to be thrown and caught by the generic catch Exception block to be translated as a failure +The current implementation is lookup hadoopjars then environment variable +something which as far as I can tell should work but it obviously doesnt +If you have better idea on how to test the project please let us know +Ill take a look at this for +the patch looks goodMerged this to +We only run it by the way to help insure that the +It throws the previous stack trace with the new export classes for RTF Excel and the RTF issueMy setup from before works for RTF when I dont use my Decorator +It looks like this issue is caused by the change in +Rik unfortunately we have other issues with higher priority +OK so we have two ways to solve it both have their own failings +I agree that simplicity predictability and consistency are our most important requirements +Uploading new patch using +Thanks +Failed in also +Oh you did it +Personally I think it is a for +AndreasWanted to check if you can confirm the feature by validating the use case +We already mention this in the reference manual and I dont think its worthwhile spinning a new RC just to copy that information into the release notes +For HA requiring an external trigger for editlog rolling is going back to previous inflexibility +patch applied +Closing resolved issues for already released versions +Reopen because the fix does not work if Stanbol is running within an embedded OSGI environment and the Context Classloader is leaking a solr +Great +Ive just committed this to trunk and. +After the addition is flagged you can create the patch file that will include the new files +Emma XML report should point to trunkservertargetsiteemma thx +Note in Java EE the right option is normally mvn package not mvn install where the goal is to produce a +mergedThanks! +Merged my changes with Bills patch +Setting values is something i dont yet see how to do it +Betaa was delivered +bq +It may be that the user is happy with world readwrite permissions depends on the installation +Create to track the ssh server issue +I think the stacktrace i was getting was due to an invalid mapping file +This is in source control as oftoday get the latest with the subversion client or wait for tonights dailysnapshot +already takes in a Locale though it was originally intended to be used for date comparisons +On the other hand something like a metadata error should cause the deployment to fully fail +Thanks. +The final circle would be good with livenodes but have a state of recoveryfailed so it would be redAt a glance you could roughly see what your cluster looked like and what state it was in +I would classify this as an iOS bug since set never occurs even once in our code +No condition +maybe thats to compile the source on my will fix it +Confirmed fixed Gabriel +mark I was thinking more along the lines of groups of messages +I added the module pom to master +I believe that this is by design +Its also spooky that text fieldType has different indextime vs query time analyzers? Ie WDF is configured differently +Added httpgateway to Description +I cant say that I see anything suspicious during startup +My second comment originated from our practice of tables + on your comment Jacques +Per the JIRA triage meeting on Oct though Henry Z +This patch could be a part of that bulk or we can simply do it now +Where do you think Im doing the wrong thing that the test arent catching +The patch includes Fixes at agent to ensure the upgrade command is correctly interpreted Support for python script execution Fixes at the server end for upgrade report Additional logs Stub upgrade folder for upgrade Misc test merging with the latest changes +I think there should be another jira to propose this +So as there is nothing to fix in the VPE source code this issue could be resolvedResolved +Lets get rid of this issue in +Tomcat was not changed and hence this is why it continued to work +The only question I have is shouldnt the configuration variables start with hive? as opposed to stream + I too think you can get what is required by using other methods in for your suggestion +Thanks a lot for guiding me +sounds patch was only for source script files +Patch with comments incorporated +Well need this changed in the console for ESB archives +It would be great if you could review the patch +Your best bet may be to drop spring JAR into your Grails distribution and apply the attached patch to your copy of the remoting plugin +to the patchOpened to run the patch thro hudson on the MAPREDUCE branch +The pk attribute is ignored +Im going to commit then we can iterate +second patch is provided +Its generated by the webform at the bootstrap site on github +Thanks for the links I understand it now I wonder how often this is used does libraries like richfacesicefaces uses this or do they do everything programmatically +Also some larger context Over the next few releases we will certainly be see a shift from forms to forms which bypasses the question of whats in the URL + file contains test source code and class files +We need to be fixed first +this jira +Comment deleted +Verified on. +Thats not to say that we have the clearest conventions +though to tell the truth it got difficult already when we switched to ivy +works if true +Please close issue as Im unable to do so +I dont think I tried it out +also case failed in vmware failing in regression to master and +file as well? And I think you might have to set the root class as well. +It is a significant work but thats almost unrelated to codebase +getPattern is optional +I just committed this! +There is no new information on this bug and I could not reproduce the memory leak so I am closing the bug. +for the patch +Ok ill do it +Attaching test +This change reduces the period of time where lock on instance is heldNow you are thinking that there should be no locking problem? But you end withbq +GavinI am able to attach files now +Ill buy thatPatch applies cleanly but mvn test from root doesnt pass after the patch +I added a note about this problem in the release notes +Ok +This fix sets the session of a proxy to NULL when evicting +There are some parts of the edge cases that are a bit weird still mostly relating to cases when a ref update causes a large number of emails to be generated +It may turn out that this solution is sufficient +Not making +or at least missing functionalityThe endpoint does not support the fileExistsAppend have attached a patch that fixes the append bug I reported before +I just tested this +The proxy should have sent the Connection close if it was not going to HTTP +Thanks for your references in Vasily! +so please put changes made to documentation back. +Harmony doesnt seem to have it although it claims complete SASL class coverage + shows this issue to be resolved but thats not what my tests show +If not IE or less than version displays warning message with links to Website Setting page and Microsofts IE page. +Tim Can you please add the configuration property to disable this DDL? Thanks +Committed in r +Reopened as the property names have changed in the specification +Agreed. +True or no return value specifies that the event has been processed +Thanks you folksAlso Thanks to Cristi Prodan for making progress on this +Is XA wrapper in the roadmap +with JBT. +Thanks +Hi Simone whats the status on this one +Theres a regression and this doesnt work but fixes it for. +Now the directory is just mb if it is still too much I can make it smaller by cropping the images but I would like to keep them this way since they are very useful for testing purposes since they are easily for comments to close the +The failing may be caused by +above situation both refer to zookeeper leader solr service cant work before this election procedure over +Ted and which other class did you mean +on for the view +A test case +I dont know what you expect to be done for this +Overrides any other file found in the source dont see the metainf property discussed in your comment +exact same db certainly +Good job Alex your fix sounds good it is exactly what I had in problem was fixedNow repository configurations from extension was merged with configuration form persister. +It looks like the MRI issue has been assigned but there doesnt appear to be any action yet +I did not add a test case because it is a little awkward to parse the log file +Closing all tickets +Jeremy as for performing formatting changes I noticed two little things in the patch in from what I have seen we usually dont use this in front of variable unless necessary probably there are some unused Localizers +Preventing the querier from making mapreduce assumptions just seems sad +This could perhaps explain why the row count statistics are different +Fix attached +I got it but the CMS does not allow creating anchors in websites +Thanks for tracking this down. +Delay is to make sure the VM are in the correct state which is already checked in the code +If the parameter interceptor took a list of exclusions you could achieve the same goal +The problem is that the JNDI location for the JTA transaction manager is +Whats our committing policy? Can I commit issues nodoby objected to for some time? Especially if they are trivial like this one +When we need access a simple lookup with iterator creation is suffientNow keeping this structure in synch with registrations is the key +Attaching a minor patch for and servlets which do not share out anymore +I agree with Marco +I have set socket timeout to but when I increased it to I did not the exception +Its clear whats going on theres a huge statically allocated variableSetting the stack size to k means Cassandra can start up with javaagent +I reproduced the issue even locally +Let us know if this works +fixed by R +How are you getting an m POM in the mix for an m project +If all the nodes in the relative path exist except the last node this method already correctly throws a PNFE +we should not corrupt that prone? I dont buy that +The classes have duplicated +flavio can you see if you notice anything +Logged In YES useridI am not familiar with the SQL spec +Heres a patch to the source file causing the issue +fixed to handle nominal attributes +Do not forget to add timing information to events +Thanks for the review Tom +Its not ideal but if we need we can look outside ASF for resources if that helps +Attaching +Pull requests sent +Yep looks perfect +Since then ECS has talked about this internally and this will be out of date now. +Thanks for the analysis its very welcome +It looks like their are two instances of this problem in the ds m release +For eg distcp the input file of the job is small but the data it works on it is huge and it will be a disaster to run it in uber mode +Could you check with the latest snapshot +Well I propose to get rid of this patch altogether but my previous comment is still validSome places of the code base look pretty bad +I should probably add that there is no way to calculate the timeout value from the code because the decisions take place before the test code is started +Thanks for the patch! Fixs been committed +This was fixed as part of and +still need to add more samples +So long as I include each individual camel jar except things are OKAdd to s and the run mvn test for +I dont think making people learn Groovy is worthwhile +run +Thanks for the details Justin +not fully integrated in maven build for the project +The tests ran cleanly for me with these changes +Glad to hear that this is working for you as wellThanks for the feedbackCheersMartin +Max I think somethings wrong for you +Please review and commit +Sorry +Thank you Jackie Chang. +Really looking forward to moving all of our CF apps over to your engine! +Joel Ill deploy it through NexusThanks for your help. +fixed in upstream master and branch +At startup Synapse would register a as default +Eric does this look ok now +shipped with +Sorry for the stupidity +Nice thanks! Hope youll also remove that nasty require and change private access to configurable properties to protected +Given that Ted and Nick d and the condition for stacks conditional appears to be satisifed I will commit if there are no objections +to commit +Minat version +Oops nevermind that was probably the result of my attempt to backport it hastily +Applied the documentation patch and sample patch +I personally have no idea who should be the best person for it +I have the same problem even though there is no associated with the parent pom +Thanks Arpit! +Id like to commit this +Can we pass masterRestart to either function or constructor +Applied thanks +The PDF Parser for also fails on Korean textSteve +Duplicate of +Did you mean to close this issue? Which issue does it duplicate +Ran the unit tests and they pass +Sebastien any updateopinion on this one +We should still fix the bug though +Assigning to you because I agree +Thanks for the updated patch Dart +As I recall we ended up with because some folks argued that it was easier to access than pulling up the right wiki page +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Ive now created a dedicated service that will not paginated results +Shalin yes I saw it I still feel the underlying issues are just as important if not moreso +Thanks for updating the patch +This will catch problems soonerfaster and avoid the above snafu +The problem still exists in issue also exists in the +AteI never had the issue with the not getting infused +I think it is because the matchtest pattern does not match the namespaced test node in the xml +Still applies +committed +I will keep you informed as this progresses +I see no reason why tocomplicate the code just to allow non alphanumeric chars for service name. +I think there will be some autoComplete config that might help this and if so the user will be able to pass it into the tag +Issues resolved in +Actually Id rather have the unallocated disk space in the heartbeat because when is implemented the should be given fresh information to decide which tasks should be launched there +Sorry I ment in validcredentials? Also one question did you investigated if savon support VCR recordings? It would be awesome to contribute also some minitest tests into testsdeltaclouddrivers +The importance of fixing this issue was raised again by a BSC user + +core tests due to contrib tests due to +Applied change in rThanks for the report and patch +Target this for since some work has been already done as part of this bug since its my fault Removing the LLONG crud and fixing the areas where the is not properly safe +The +Resloving as Fixed because the code was committed to CVS + +It looks like I missed that +Committed revision Ill update the wiki with more details on the rollback command +To trunk +Attached the patch so that its easier to figure out what to change next time +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +However the sitemap gives a nice looking page +md +we dont have tests for letter words please review like conflicted with this patch +Even though your workspace may be clean the server may be running with a project with that ID +Why is this needed? We already have upper limits on resources per node already +Im still not convinced of throwing an exception +Is this working with latest? +Two more files needed stdint +A number of the old cases are from the times when PDFBOX hasnt been an Apache project +Thanks VargaI have commited your changes +Thx for fixing Julien +This was created by slightly modifying the default Forrest skin +getting rid of files of Mb each is better than getting rid of files of Mb eachYes the ratio proposed when this started has a tendency to get too stuck on small files +Unified patch provides new interface for default implementations test cases and documentation +At least I think that show in remote system explorer should also be available in the cloud view +Please join me on IRC if you experience another problem so we can discuss while working through the screens together +What we can do is to hold off committing other conflicting patches for now and have this patch more thoroughly reviewed +There is a lot difference in in trunk and in this patch? I have concerns about replacing existing distcp with new distcp +We need to print out warnings in +As youve already suggested we can easily set the deployment class loader as the TCCL to resolve this issue +It also removes the provider before adding it againColmThe fixes commited to SVN work fine for CXF +In the package summary it lists classes but it should take the first sentence from the comment and use it as a summary in the right column of that table +Done +I guess we would have to evaluate all the resources and compare them for entries in the grails application and remove them if they are no longer there +One commentcdevcsecureI would remove everything from Other security holes onward +Rather it would be up to users to make sure they dont corrupt the registry +eddieyoonThis one is the same as how to delete it +Thanks! Ive took a quick look and it looks great! Im now running the integration tests +I think we actually need both the workqueue and the thread pool itself +I can give you all the information you want dont hesitate to ask meAnne Noseda +Then I did mvn install from the windows cmd shell +Im not sure theres a huge use case generally for this in Tika so Im happy to move this code back to Nutch as a dependency on Tika depending on what others think +sorry +Closing this +Please verify +bulk close of resolved issues. +Thanks Sid! +All versions starting from causes greater Permanent Generation space consumption +Last patch had misplaced text outside a target +Commit in JBDS in trunk appears to be causing compilation error +Also affects the test sybase issues to next as wont fix if someone run into this issue feel free reopen as Hibernate Known Issues in Release Notes +Please attach your in ER +This is basically a duplicate of or at least the conclusion of is being retargeted as a feature request for the next version +How do you keep the separate libs importexport separate? They all seem to use QPIDEXPORT so for example when building the client lib shouldnt the client symbols be declspec export and the common symbols be declspec importThanks more question for Danushka + on. +I either dont have the privileges or cannot find out how to change the status of this issue but for me it is verified +should be in to be treated as part of deploymentnot separateIf you want to have it as I can look into it a bit more +Closing since Alexei mentions in a later comment on that Confluence now runs which effectively verifies this issue. +HiIs seems that one must leave out BBOX WIDTH and HEIGHT all together like they were missing in the request of the original report +disregard the first description +themes +I would also want to fix the following These names are too long instead we could simpler names like +Anyway we will need to see where it is being used and if it has any potential use beyond the scope of axisc or its subprojects we will have to fix this at a perhaps +I felt like removing altogether but figured it was there for a purpose +Mostly looks good but I do not find the server side code for suspending and resuming task with this patch +It is somehow possible on tomcat if someone knows how to do it please share +Linking this to an already resolved bug +the patch also organized the imports a bit +Can you attach the WSDL and perhaps the trace message from +We should keep these styles but change the names to be AF and make them public +Updating dependencies page to reflect target release candidate versionssvn ciSending Transmitting file dataCommitted revision +Fixed and checked in. +Patch is against +Quick patch to reference correct JIRA +issues +Since most work will be done on the CXF side I think it will not take much time to port it from Camel to Camel branch +I tried to something like that effect without success +Vladimir this looks to be related to your work in the lifecycle module? If not please feel free to assign it to someone if its nor present a WARN should not be thrown particularly when no query module is in use like in the Infinispan demo +Ive just committed this +I agree this is something we need to clean up +Looks like the retry cache works while we still need +Attached the patch for +Many can be divided into +sorry for duplicate +Added to version and version . +an test case preventing the bug plus a missing else clause +Next versions will +Michael Ive been on holiday in Europe for the last weeks +Patch for +How about addendum I prefer the sorting to be done by the client +There are a couple of issues that require review by someone who is out of the office until Thursday +Discussion thread added and useless discussion moved to people are just incapable of following basic instructions +But I think its best to go ahead and close this and if docs for on this are needed for another one to be opened +but now i have containsKeycontainsProperty +Im completely sure about the groovy version im using +The uses an optimized path for the which creates values but attaches to it the Standard input object inspector +I think this issue is invalid one needs to have two separate declarations to get specific endpoints createdIs there any specific reason you cant do itHowever perhaps it should be actually possible to get a single endpoint representing multiple interfaces +go it in times r creating module examples r removing examples from to remove filesFixed in r +Today I found the same error when I use enable a table +Iwill look into this +Committed to trunk +One pattern I noticed in the code generator is that if a class foo is generated and a corresponding descriptor class fooddescriptor is generated the foodescriptor class only imports but foodescriptor is actually refering to other classes like bar from package +You can let me know if there is anything I can do to help +Hey Jukka and NickThanks +The GA suffix is reserved for product not for stuff produced by a particular build system. +httpshudsonviewviewjob +What happens when you run a web app? Do you get the same error? Also I need more information so please attach a backup of the project as well as the Alansorry for reports pass but I think thats because the time inserted in is unique by problem is however general you have unique constraint on two fields together not on each one +Agreed we will depend on Widget Factorys cleanData hook +Not planning to deal with that until +class method has been updated to support expect OlegI would have to agree Im not sure its worth the effort +The is only available during JSP rendering. +If none of these other chars are present the routine exits at the first return before getting to the code which does the escaping +when memory usage is above the high limit the write will be suspended waiting for the flusher to push down the memory usage by flushing some regionsafter flusher attempting to flush any region the write will be notify to check whether it can go onif memory usage is above the high water mark it will go into sleep again otherwise the write continueHopefully im not wrong +Please also remove obsolete and published +This patch adds the missing call +Sure no doubt well implement this very soon +Thanks Dick! +AnilYou patch is in rev +Is this actually fixed? I still do have the same problem with as available from Jasons home on with a date of +was marked minor so noone saw it marking as sure we havent bumped where needs to be bumped +our code should not attempt to copy the driver if it is not a jboss appserver +Is this not the case? Were there any test cases from the original code which we can adapt to the new names? Perhaps we can adapt some test cases directly from JTS? For testing we probably should be thinking at two levels does the code wrap JTS correctly and does the code perform geometric operations correctly? The latter suite of test cases could be with a different implementation of up the good work +Sorry I will in the future but it cant. +Use xjc to generate the boilerplate code then modify it +Attached patch contains previous patch and unit previously attached patch. +Ok its fixed now +Since the Connection does get released on Session close this isnt really a bug in the regular case +I just removed the button to minimize now it is not that easy cause the video itself is played in the SWF app +I attach patch file for quoteident functionConfirmed with mvnPlz review my patch +Attached the file which is used to from maven to run the sonarsonar file contains thesonar setting which work currently for me +Ill post it +The above exception was caused by action +Is this a blocker Gary on +I agree with the intent Ive done this myself from time to time +The results should be cleaner +The problem actually occurs during the compile and I was being lazy and using perform tests to get the compile to happen +JIRA interpreted some of my comments as the start of a list +So we havent users org unit +Removing fix version as this fix is not related to core but rather its plugin specific. +I think this was already solved in the last snapshot +Can you please explain what is wrong with the current code? Are you seeing an issue with this +I will work on that patch if noone does +I expect this to get fixed as part of think this we are now throwing more generic errors that dont mention the target type. +We still need to decide how to handle woolfels changes +Added version parameter to the VDB Deploy Via URL operation as well as file deployment +When setting up a project that doesnt require proxying it works like expected +Please turn the device or simulator into landscape orientation start the app open and close the and you will see that the views width still remains in portraits widthBTW In my actual app was never the height +Added param as suggested on the list +Verified on +This is similar to Johns approach though the code is no longer relevant to the APImodulesetc simply start a monitoring process on files and then use the regular site generation on files as they changeI think the first is more beneficial as we can introduce the location of other modules in a reactorUSD to also solve the problems presented in the other JIRA issue about being able to see the entire hierachy in its desired layout +I am finding it hard to review this diff because it has lots of diffs that are not inherently connected with the attempted fixThat was exactly the point of asking for another +Will commit this shortly +As described in the linked mail message you need to trylocalhostremoteDemowhich should show you the Forrest home page with the xml source being retrieved direct from the svn server +One additional patch to change the api doc site link from absolute path to relative path +any comments on this +The attached shell script removes +Fixed in revision to specify component +This patch was not created from the root of the project directory +does support floating windows this this issue be closed then +Theres a couple of other with patches we need to apply and the issue thats just be raised on the dev list so once theyre all done lets do a new Abdera release +However this is something that is required to be able claim a maintainable wire compatibility +Please help +I improved the javadoc and added a trigamma function with associated testsThis patch is relative to current trunk which already contains the digamma function that Phil just committed +Closer examination shows that it is NOT the case that tiles are being missed + +Resolved for a while +moving to and date time support has been substantially improved in geotools geoserver pushing off to far as I know this has been fixed +Applied with some modifications +works with firefox and not with IEI will try and use ethreal or other tcp analyser and see the difference in the content exchanged on the wire between browser and server for firefox and ie +Doron were you able to check on the patent situation? If theres a chance of solving this in a positive way how long do you think this could take +Thanks for taking a look at it +This is the original issue where the fixes has been reported and by parent issue +great comments im gonna get on top of this tomorrowthe view editing was the last thing I did so Im not that surprised that its not as refined as I would like +But in scenario you seem to be talking about a stronger guarantee assumed by the client +Network server reports it correctly +AllIs there a way to restructure the code so that naive implementors dont run into this situation? Essentially the code as it stands is and it would be nice to make it be in the presence of new opsOr is the addition of new ops rare enough that this doesnt matter +the square bracket names are the actual property key names +Moving off as it needs more feature is almost done but Im moving it to as it shouldnt hold up a GA release +More general means more better! BG Im quitting to go work at Facebook. +I dont know if its worth trying to fix the +Spring Modules will be released at the end of the week +Two it does not work with the package or tar targets +Thanks. +I just committed this +Marked for pre +Needs new Lucene jars per the earlier comments +fix resolved that error. +Patch committed in revision +applied +Now they use the same +Yes +Thanks RobertAnd a very good catch Sid +who would migrate files in matter of day more than one + patch looks good +Uploaded a new patch +users should upgrade as the fix uses updated java memory model semantics in java and java concurrency features +thanks for your comments chris +Moved to a later release once issue has been fixed. +Reverting version until we agree +This is something thats currently being worked upon for +That issue has fix version set to Betaf +But will consider fixing in the future if there is enough support for now use final modifier is ignored for groovy files +Part of the power of Maven is that it catches these problems early and gives you tools for resolving them +Never mind about my previous comment +This is a first basic implementation of the cactifyear ant task +Note that this error only occurs when you first open the Deployments tab once you change the radio button selection the fields will be disabled as they should be +Committed to trunk +Ive pushed to the branch +yes maryam bug is fixed so lets close this +Attaching another paper from Ted that makes it clearer why the method currently used has problems for small text chunks +Hey MaciejGood suggestion +I also ended up fixing the Accept header parsing in the JAXRS code to actually work with the default sent from a +Patch removes trailing spaces +Attaching the patch mentioned in the description +could also be used as the super class of an Image element +There were very few if not zero use cases of manually adding jars to the initial startup classpath +The testcases for the particular platformjdk combination that we monitor in CC runs you want them opened please use the JBPAPP project using the particlar platformjdk combination under which the tests fail. +Replaced with in javadoc and made some similar changes +This is the second version of the release notes for community review +Will try Java later on +Perhaps if the moved to Ant then it would also barf up the same I should revert to Ant +this is an awesome step forward +If this is still an issue please reopen the issue. +There are a couple of considerations is this would not be a designer issue but rather Teiid next is that we allow file system access through a translator as to abstract how files are obtained full pathing and file system dependent details etc +on patch v +Permissions issue or me overlooking? +Currently I use a mixture of Jettison and Jackson to handle collections in Scala runtimes it would be nice if this would all work with Jackson +Bulk move of issues into after vote that we merge branch into TRUNK up on list +Digester is also used by a simple SAX parser would be good enough there +Ill leave the task open and reassign to Steve to complete the Windows part +Fixed +This looks good +Adding changes to be committed. +The rat plugin is pretty crappy in my experience too +Ive removed from Fix Versions in order to avoid confusion +Limitation is that stats gathering logic is currently written to JDBC layer rather than to connector level +PingShouldnt be too time consuming to resolve this issue +Should they be careful not to drop a user permanently until after they have removed the database objects owned by that user because otherwise no one will be able to access the objectsI gather were not actually changing any tableprocedure names or any Derby behavior at this pointThanks +Benjamin has already done and committed most of the work +A srctest folder you can use to reproduce the problem drop it in a fresh project run ant test +I try to create patches for theseRegarding +Updated the code to not perform the connection test on startup but started a polling repeated task in the connector wrapper with delay to test the source connections +can this be the cause of failing to stop the due to concurrency issues +But what happens when they do? I suppose a clever thing to do would be to merge the overlapping ranges but that doesnt strike me as particularly relaxed +I was under the impression that we were working towards a because that way we can remove at least some of the code we have currently in place for backward compatibility +This is no longer relevant as the unstable suite has been eliminated +Should exclude all of jaxens transitive dependencies only the main lib is needed +The same lock is also used for adding tasks to be monitored before their launch +Build is now working on UbuntuIt would be great if you could install it on the other ubuntu slaves as well +Andy please note that its preferable to NOT change the groupId for this release because these artifact are already used with aspectj as groupId in the Spring release! id was changed after discussion with the spring team about what it should be moving forward when the process is automated +Lets try a patch that actually on +I get successfull result. +I committed to trunk +Fixed in r +No longer relevant +failed because of think this is the right one I fiddled for a while and checked a bunch of the test cases +anything else you need from infra +I did discover blatant reference to standalone in Configuring the adapter to use a standalone Server +Andrea this patch does no look too bad but I think it would only be applicable to the stream of code? If you are interested you can apply the patch. +Yes all you said makes sense +There can be room like Andun Room which is video only by hiding the white board +It is probably twice as slow compared to the classic BIO and is times more complex +I will submit a new patch based on reviews +Everything passed without any changes +Should I put request on SEAM JIRA? Is there workaround beside dont call pageflow? +after surefire tries to parse a MB logfile full of +Not entirely sure how you got this but closing as fixed since it doesnt happen in +I find things becoming too brittle for my liking +Didnt notice that yeah looks like it is +After running cmake in your build directory make packagesource will produce the source tarballThe only caveat is that doesnt try to be smart about what to include in the tarball basically anything in the source directory is included +I managed to get the plugin to run the code the error message is now fixed and the bug cannot be reproduced as its already been reference the version bundled with the eclipse plugin that i tried is +This seems to be OK now were using Xvnc metacity. +Our primary network does not make use of the virtual router however we do add a secondary network with it to aid in automation during VM creation +If the cloud actually changes the node resources then the NMs should get to know that something has changed and report it +Did you test wo checksums enabled +Pavel I didnt understand the point of the bug earlier +Thanks AlexandreI think this solution will fix our issue +I believe so yes +If someones wants to get accessadmin rights just send me an email +Im moving all bugs out to and we can move some bugs out to to get some release action going +The main changes impacted configuration and changing some tests to explicitly choose the socket based method. +If I dont set the size of the receive buffer this never happens +In my version it is possible to save a project setting which overrides the general eclipse settingsSo this seems to be fixed +Unzip the attached jar in the server directory +This lets people put interesting formatting tags inside the summary paragraphs of release notes +regenerated patch against latest a few unused lines and formatted at some placesFergus I am tempted to make it go in? It looks largely OK to me +I have a potential fix that I am working on +After you open WindowPreferences ToolsEditorsVPECode Templates to check that temlates are defined and then after returning back to the page CA will show templates proposals +I dont think I would have ever got to that solution on my ownReally appreciate your time on this and the speedy +Do you plan to use it later? for the patch +Thanks Sami and Bruno! +Critical priority is correct here I think +In addition get is also not implemented last mod time block size replication are not available through libhdfs +Dont include t as per Ryan suggestion +Closing all issues +bulk defer of unresolved bugs to +Thanks for thisI dont think the permissions are correct though +This approach doesnt fix that assumption but is consistent with it +which version did you test that +Closing old issue fixed well enough focusing on improving rather than having Wiki being complete. +This has been fixed by Aaron +We only need to send the login html page as response to unauthorized clients if they give higher preference to texthtml +Updated controller with default package Matts suggestion to the code +Patch applies cleanly and passes tests for me +Attaching a patch for with some bug fixesThis is an example patch not to be committed +I thought this was understood about modules anyway that we move the code there all committershave access and its shared code to support both projectsCool +closed for +The phase to bind it is it will break multi modules build if you want only install all of them +No other configure change +Hi PaulexThe fix looks good +I can give a patch for too Lars +Can somebody with committer access help look into this before the patch becomes out of sync with the trunk +Alright I removed force and set TTL long enough to not get tombstoned so we get consistent test result +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +What do you mean by a missing tag? The file sizes for in CP and CP are the same +It wont fail if I have also remove the from the right? I was under the impression that tdocument was optionalIm using tomcat +If anyone likes to have adiscussion about that he should also use another bug changing the name of JDO to or which I prefer at themoment is also indipendend of the discussion if and possibly when we want tosupport JDO or said that I will also accept as it looks like more of you tendto prefer that +metadata +Thanks OliverPatch applied at repo revision rPlease check that the patch was applied as you expected +This sounds fine with me +Ive found a couple of minor items that could probably be cleaned up a bit more but that will be a good test of the CMS bookmarklet +but Im going to say no to that idea +I confirmed the software grant has been recorded at Apache +Maybe this gives you some idea about itBye use Camel todo the JMS spooling +Modified method is private so not an incompatible change +I wouldnt mind if the patch made it conditional when using the intel compiler but when not Id rather see it be simple and clean so we dont accidentally hide some problem with artificial changes to ve attached another patch for this issue instead of the previous +I am still waiting for it to be rebased for master +Now includes entity migration service +Im not sure about the type of exception thrown +I have to look for what might produce that error in my production code. +What Fernando mentioned is correct the patch for the pom is provided by Andrew so that we can build with JDK +Many organizations simply do not allow anonymous bind +Lets keep testing the new web UI and open new jiras if we find new issues. +Why not parse the XML header and retrieve the encoding from there? If the encoding is missing use as default +Ok as youre the person with that use case youre probably the best placed to decide if the hassle of creating a list of instead of using varargs is worth it to not have to deal with the compiler warningsActually come to think of it if we add a version that accepts a List or will probably be the most common way of constructing that list and theyll just result in the same compiler warning so I guess wed best leave the varargs version alone +Marking as fixed in +msg file only contained one such entry so we never saw this issue but apparently its possible and even likely for Outlook files to contain multiple such entriesI fixed the issue in revision simply by introducing a special marker flag that prevents the Outlook extractor from being fired more than once per document being parsed +Sorry missed updating the metastore client with the kerberos prefix in the principal name references +Can we commit this patch if the review and the explanation are fineThanks edge ordering in the implementation is the order in which the connections were made +However if this is based on the then this is an issue because this is whats relied upon when importing via Teiid Designer +Dropping the priority the patch as it is not safe and invalid. +There is no more IDL order because fields always has tag orderIve missed the order of fields in the constructor +Fixed in r +Originally I had problem installing rails with my jrubytrunk and I have the same problem with just tried it again with a more recent trunk and it works for me +QE ACK on this one! So cute congrats! +Fixed in trunk thanks for the great test case +Attaching a patch that starts only one tracker so that the ordering is maintained +ah i see i dont have the areEqual check +from me just committed this to and trunkThank you Plamen +Those Hadoop release have all the that Sqoop expects +This was fixed in vm setting fix version correctly +The only problem with marking this issue as invalid is that there would be no acknowledgement that there is a semantic problem in the API +Committed to trunk +But that would have been quite easy for anttask I cleaned up all checkstyle warnings and infos of anttask module exepted the usage of deprecated classesmethods +However the workaround is so simple so why bother with JVM bugthanks for the fixAmir +Here is an updated patchHari the location information is a hint to the namenode and the namenode does honoring +The previous change to was fixed already by someone else +Thanks guys +can you retry with the current development version +The tradeoff is a pretty close semantic mapping though definitely not to the client API on the one hand and loss of the cursor upon process failure on the other +msec transferTo +finnally this is getting fixed +Not really sure what you mean in that first sentence +Ill blow it away once Ive finished remapping the names and do a fresh import so you can confirm it looks ok +was reverted +Ted whats the reason for the tag? This is currently developed for this branch but probably should be ported to trunk as well +I was thinking the output is a Jar file with a in the MANIFEST so you can invoke it as java jar or double clicking it in Windows +If you turn the to be true + and is completeWell build a new before we devops that one +Pull in tomcat source code into external directory r +Gracefully handle cases when throws exceptions during an alter +The upgrade code needs to be made conditional +No tomcat needs to be updated to run off of the +Here it is attached +Being able to set response codes for failed updates would be very useful +Look at Liquibase or one of the db migration plugins. +Closing we released version. +full +Ill see if I can eyeball a fix for now +For the memory values MB instead of M +refactored the codes and updated with trunk +Looks good +Thanks for the uber quick fix +Finally. +I was going to ask you if there are any other details about this problem in bugzilla issue +Andrew is there a plan to move forward with it +I left out some java files in the last patch +But this was the case that was not yet accounted for I believeI cant think of a use case where the second scenario would happen thoughThe first scenario though is critical for migrating so its great that you committed thisThanks for your work on this issueJoel +Deserialization works since the parser will provide a String anyway and will look up the type from the field that has to be written +Please apply the attached patch to EWP branch +Can you please rebase this branch against the current trunk +Ok +At revision I added a test case which looks at how the exactlyOnce profile intent resolves to at and at in the endpoint +Dave has already made significant progress towards this +Only two bugs are actually fixed between RC and +Another approach would be to leave the Palette as it is and then change the Event Definitions to only allow a single definition to be added when in BRMS profile +Having the profile available for later when somebody actually cares is a nice idea +Thx for your help and work +A scrap of servlet code which runs in a Tomcat server +fix committed to Metamer + +This also implies that the web developer also has control over this setting in environments other than production +Its not clear to me why you are loading an Equinox framework class while running on the Felix framework but ignoring that fact for nowAre you saying you are trying to load a class from a bundle in your class that was loaded by the application class loader? If so there is no way that is going to work since the application class loader does not have access to bundle classesPerhaps if your code is intended to run on Equinox then perhaps that is why it is trying to load an class not sureYou could also try a newer version of Felix since that would make it easier for us to debug +Can you please post the entire exception stacktrace that you see +Patch applied at + +It appears to have been a dependency conflict when Maven is used +Myrna analyzed the issues and determined The windows run failed to start due to a configuration issue +Once I create the database with that attribute the very useful program provided by you gives the same exception that Dan is running into +moved code from rstryker workspace to on. +I will try to validate that there can be no duplicates +Possible solution based on which is automatically registered when is usedTODO add JMX attribute for the max value +Adding a test case is better +Could you attach patch files in diff u format +As of eca embedding API had improved runtime initialization methods +Reopening because changes are still required for source files in thetestjava directory +i dont think we should in a minor release especially since we cannot use random access Bits +HDFS tests pass fine for me +Has been fixed months ago +I am using version +Good point + on proposed change +Including a patch via svn diff +That part of the code needs to be changed anyway so Ill move it to an extra method when the time comes +I am taking this from Eric because while working on I started fixing the tests anyways just so I would know that I didnt break anything +Committed patch b to trunk with revision +Added the support for derived numeric values and these functions are now complete. +Can you please redirect me to the correct repository? Thank you +Amareshwari can you please provide one for? Thanks +to FIXED +OK it should be simple enough to remove all sorting however do you have a view on where rows inserted whilst sorted should be listed? At the end of the table in the sequence they were added seems a good default but thought Id check if you had a preference. +This may change once we have an important use case that needs better cost estimates. +I see no point being a more fervent Catholic than the Pope himself +Thanks for taking care of it! for the very helpful reviews FrancoisCommitted patch to documentation trunk at revision +Should be included with the cargo integration being discussed in someone is interested in working on this jetty was just released and if you have issues Ill try and help you iron them version no longer exists do changing to unknown fix version +Manually tested +But I do not like it and personally would prefer to see the behavior either fixed in a performant manner or backed out especially in what we are claiming is a maintenance branch +Pending resolution +In is it checks the wrong class +need to check with logging +Pradeep could you please look at the first issue +I thought Upgrade for this issue should not be part of this JIRA +Committed revision +The saaj test suite passes on without failures +User can turn it onoff in server editor section named Refresh bundle within application +Patch with all lists added to POM +We decided at that time that its not worth making further improvements to farm deployment since we are making cluster deployment in a part of cluster administration +Thats a fairly common scenario +Committed on my workspace a fix for +a test case for targeted for in Rev +It would be great to be able to find a way to apply this to Counters in the future though perhaps by strongly enforcing +Michael Stack I can appreciate how forbidding can look with all these libraries and the rather large COPYING document +Actually we will attempt to connect with the subject first and if that fails connect again without it +Here is a patch +Why would you want to exclude isolated methods? Isnt that exactly what the metric is trying to measure +Definitely needs to be looked at before the next release +The patch is also tested for regressions in skin creation by comparing the resulting css before and after the change + looks good +Patch that adds descriptions and tests and to be applied on trunk +Fixed in as were still in alpha stage the parameter was simply renamed from targetDirectory to as this is a parameter most people shouldnt specify this should even less be an issue +Empty perspective wo contributed viewsactions +address last comments +Thanks Alexthis is a review I was waiting forIm going to update the patch tomorrow +I contacted support to figure out how they think this should be done +Ill close the issue when we get some clean regression tests +I just committed this +Well have QA look at this as with other inreasing the heap size solves the problem +Theres a limit to think thinking at some point obscuring code by hand isnt worth it +I am not sure whether we could specify a static location but that might be a good idea will try it later +If user click on Edit button on an image page Seam Debug Page is opened +i tested it quickly on streamand i had to change the so that the servlet path was not but app and then signin did go fineIs that also the case at your place +I appreciate that it would be nice to be able to do this and we will look at it in the code base +I also attached a log file with all the hibernate queries executed to load only the index pages once logged in with remember me option +Applied Richard Wheeldons patch to the latest trunk code and fixed some checkstyle issues +Another patch has been uploaded with the following changes +This change should only cause an error to be logged to the NN log and cause the metric not to be output +It would be an error in the input file wouldnt it? So you might prefer a more friendly exception but it is nonetheless an exception +Also the +Thanks for the review Mike + is part of my thesis work +Concise no surprise its not +Yes this looks like a normal thing a stale lock file +Sample application that will serve a PDF file to demonstrate problemFile is served by testshow action +is blocked on line +Now you need only to specify adapter to view generated SQL or save it to file +The synch to Central has not been done at pm or pm last night +Thanks a lot suggested this issue but you actually investigated it to the very endDont even know that much Bnd syntaxes +it was the leader. +If someone wants to provide a more complete fix then we can relook at the problem but I think otherwise we have higher priority issues to address in and that need attention +Its working now I be able to test the new build more throughly next week and Ill report any problems I might findSorry about the delay +let me know if you still run into works now +Cleanup looks good +Somehow nested moves do not work well and prevent me from getting good svn diff +If you have completion jobs queued with the executor they will be lost on server shutdown even with persistence +Compiles fine locally +Unfortunately we dont yet have an official release of Apache +Something along these lines maybeAnd we are back to +I think should be as small and simple as possible to make discussions and reviews easier +I have also added extra logs for flow control but they are at trace level since debug would be too intrusive for a production system on this case +might be interesting to dig into +The approach to a mock connection object seems to be the right way to go +I did not add some files via SVN and the patch did not contain all of the files +A name for the zone is too short +Ill commit this tomorrow unless I hear otherwise +Christian same effect + the issue label and description I realize that I wrote cargo instead of in several places making this issue barely label should read Document Jonas container and description should read the official cargo site doesnt provide any documentation on Jonas support + really should be moved out of the endorsed lib as there should be no reason this is needed at the bootstrap class loader +Alternately the test build could generate the svn number into the test properties file as well for checking +I had it on my list of things to do eventually +Fixed in M sounds like the bpel one is still broken just commented out of the build be good to try to find a real fix +I just entered Bugzilla Bug for Minchau to reveiw the submitted patch +Hi Koen NickI havent used Windows for awhile so I guess this issue could be closed for now + sounds like a good idea any attention youre willing to pay to improving flaky tests would be great +This will have to wait for some more time till someone has some free time to integrate this new version in AS trunk +Attached patch is submitted for subtask + has been marked as a duplicate of this bug +The comment in talks about but it should be Done +Please confirm +If you prefer get to type thatd be fine with me the change hasnt been part of a release yet and thus wouldnt break compat +We used to compute the line number of an error marker every time we mark some EL as invalid +Fixed in r. +you dont own domain afaik so you cant use +Sorry that patch was against an older version of ddlutils +Change a constant do a clean build +If the client is writing to d d is forwarding it to d and d is forwarding it to d then the primary is d isnt itAlso this bug should already be solved in the branch. +for changing the description +I have merged this to +Are you in danger of creating a dependency loop here? What might be interesting is to do a downstream project deployment or some witty name that handles this and the other packagereleasefunctional testing stuff +Without deleting the database this bug will reappear because the wrong index values for objectClass are stored +Take it over + +Trivial patch to shut thrift down just before gossipMS +Am I missing something +The question now is how to make it NOT read the requests for struts actions +The goal of this change was to disturb as little as possibleI believe I might have a different solution to this problem +Yeah lets make that a new JIRA thats going to be a bigger change +Mistakenly resolved +I attached it now as. +Closing as requested. +thatd be good though I think its only necessary to list each org once would need to we currently have developed by ASF at the top as a blanket and then list a whole lot of ASF software that doesnt need to be +Thanks Dag! Committed revision . +please move them toWe should keep those close to where they are used +For example if we provide a default CSS file for selectDate the resource path for this might beseamresourcewebselectdate orseamresourcewebcss physical resource file can be stored anywhere in the jar file +Closing issue which has been resolved for some time +Looks like a sane approach +I fixed this for chris + +Eric I am not sure what issue you have with this +verified in reopening to update release note info. +May have been before our hudson was hooked up with comments now you just need to include the issue number in the commit message and itll make a comment here when it builds successfully with it +I fixed a few bugs in regarding the patch +Is the provenance concern any greater with WTFPL than with any other alternative licenseNo +Reassigning all XJava debugger issues +I think I may have found a workaround to provide the behavior that true should provide I posted it in the reference forum thread + push to +Applied the patch +It could also be exacerbated by a startup issue were a lot of outstanding DNS requests for the same host are done unnecessarily +We should not be breaking Java refactoring +So i am going ahead and closing the issue. +Bug still present ajax is hanging +Zip with chance I can ask for icons for mergingunmerging on the new decision table too? Screen shots of existing icons attached +But it may give us a different simple solved things for me +Closing all resolved tickets from or older +bq +This is the same code present in the initial comment +attach my patch +Given that keys are hashed to a partition and each partition has its own write positionit makes more sense to minimize the number of partiitonsIn order to reach current and projected disk sizes we will need to increase the current directoryentry size from bytes to bytes +These are forthcoming +Logged In YES useridIve also attached my database population script and +code angthanks for the tests Ill use them however Im closing this issue as its a duplicate of. +The patch looks ok +The is for it removes some redundant codes is for Yi XiaoCould you please dont do code format before creating the patch? the format code mixed with real fix will bring troubles to review the patch +Fixed in r +sorry for the confusion +string and bytes should be able to specify a max length because one size does not fit all string and byte max length should be +The issue mentioned above about the is caused by configuration for the in +There are many values that users can read but not update +modify and to compatible with Harmony code applied at r thanks for your contribution please verify +Oops I forgot to remove a shortcut I added for testing +extthriftprotocol +How many heritrix instances? If one how many Writers? is default IIRC? A byte array of MB is kinda crazy +Attaching a new patch that makes the testcases memory loader static so that once the map class gets loaded the loader stays in the jvms heap without getting +Running the current trunk on the Enron email set revealed one weakness of the mimej default configuration +I copied the behavior of setCodec which throws a NPE although you are right that most methods seem to set the default value +Chiahung got us to the point where is aware of failure and updates its status +I think I cannot implement fieldtype version of this issue because when Solr is initializing Solr is in initialization step +Version have just applied patches that could help solve this issueCan you please try and see if i works for you +I think if ancestors are still alive assume compaction didnt finish and delete the descendants is good enoughyeah agreed +In trunk a partition only exists in broker +bq +Hello Andrea was fixed I think you can resolve this issue +I was using the global endpoint +Will commit afterwards assuming no negative reactions to this approachUnit test code also included in the patch +Heres a patch that improves the error message as I proposed above +Just an idea +Backported to for backport consideration +Ill have it integrated soonIve made a few minor tweeks as I integrated +thanks for the test and the patch +The output for is consistently rn and itd be interesting to know if svn on Windows accepts this file. +This stuff plays with the fundamentals of setting up a service and symbol resolution requires several basic services to operate. +Thanks +This patch adds a test case to verify a spec clarification an exception should be raised if a or is used to compile SQL text + LGTM +got a little sidetracked at the office but Ill try to post the updated version on RB tonighttomorrow am +In latest SVN Please try the option wv when running wsdljava and at runtime try wsdl instead of the usual wsdlthanks dims +Thanks +By the way this behavior is not documented +Left a comment on RB +It would also help to have your updated note its not complete I just wanted to get a basic iframe after a bunch of digging it turns out that they just used the reference implementation and added a few lines that should not impact the operation of the hub +FaronThank you very much for fixing that +Weve been bitten by clock skew issues a few times now +trunk depends on +Fix the findbugs warnings +GC root paths from runtimes to +So yes I would probably suggest to change next and previous for char byte and short to return char byte short and not int while still keeping c being an think the consensus is that this isnt a bug +Thank you Alexey! I agree with your changes +Im ok with fixing bugs here provided thats in a separate patch +This fails if such a is created during server shutdown +Improved version of the +Not going to happen in since there is no good way to make both specs happy +Committed as r +Is this still an issue with +Changes to plus new patch that removes support for gradient formatter and uses parameter instead of to control how the is created +Please retest +I just committed this +XALANJ +there are extensions points enough for doing custom things like this +Petclinic has been moved to Maven and outside the Spring trunk so this issue no longer exists. +I have the similar behavior with injectiong prototype into prototype using Resource even in main application not to mention tests +The failures can be reproduced by repeatedly running a few times +This is another unrelated bug +But there must be a lot of words that cannot be resolved unambiguously and I think that we should not support any kind of resolution algorithms neither of them being recommended by the Unicode StandardWe read file content with standard Java readers and parse properties with algorithm implemented in +Just to see for manually setting the +Thanks Jing +fixed in revision +Id suggest thread per core at most or maybe cap at threads +testing minor compaction is major compaction going to be at the speed I seen aboveif so we still going to have major problems when you start talking about mb regionskbsec hours per regionthat means running only major compactions we will only be able to major compact regions a day not good for someone with or so regions per serverBilly +Worse yet I cant even figure out if I were going to file a new bug on it where that bug should someone update this with a plan of action to get this fixed? Do we wait for the next version? Do we have to use a snapshot? This has been mentioned in the mailing list over the last few days and continues to be a problem but it doesnt seem to be a part of the standard test cases that are re frustrated here because were not even sure if people are admitting that theres a problem at the moment are we waiting for a release of the plugin or is it this components responsibility? We cant even see if anythings going on and despite the fact that there was a comment on this bug saying that it was still broken on it didnt get tested on snapshot or bits of the system just havent been pushed live +Fixed in +Ive added the output format to the Ant task +Here comes this brandnew patch +Fixed as described +The namespace did not have any file named useruser + is adding an async to simplify writing apps so it might be good to incorporate that when its finished + for the to trunk and branch +Patch looks good to me +I found a locution that made this work while back so its easy to install plugins you built using maven into the local repo using the catalog maintained by the. +Deferred to since resolved +I need to see a heap dump for this problem and I need to avoid running the test on my current plan is to capture the heapdump on hudson +I just tested Elmers patch and seqsparse with custom analyzer works no + happens when you dont set HADOOPHOME but hadoop binary is in PATH +jeesomething even if I set the execute permission on it and the file command recognized it as an linux i something +Completely disabling the software suite immediately resolved all of my problems +Thanks Daryn! +I committed this. +Committed to trunk +This has been a pretty annoying issue with people are running into repeatedly we need to fix it +Can you verify this on RHEL +Verified in +how could i fix this m using. +Changing this from jaxbri to adb produces worable client +There you cannot afford the operations producing a copy of that structure containing almost all of the elements once again +they are not closed only linked as duplicates +That last option sounds like a good route unless anyone has objections +The GWT commands have been tidied up with commit changes made to rectify this issue caused GAE support to break +classjavac +Unfortunately I did that with SVN rather than Git and things went downhill from thereI also did a couple of fixes to this to keep working. +The test failed because append was not enabled +Pushing out to issue isnt obsolete there are still in the code to revisit this. +Caused by the linked compiler bug +I have to admit that this error is really happing on Eclipse will have to look into Spring IDE in order to make it compatible with the upcoming Eclipse versions +Thanks for review +Thomas Would it be possible to add a unit test that includes unit tests to make sure all the application finish timestart timediagnostics and other are set? We can do it in a separate jira as well +Patch for http binding component +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Hey arent assertions supposed to be disabled in optimized builds +Do I understand that you are wishing a negative version of antavailableWhy not in ant taglib? Or +hence cannot use cobertura! please solve this +without having tested it it improves an existing significant functionality and since were using a CM we can always revert so IMO go ahead and commit since its been out there for quite a while. +bq +It looks like the jbossws suffers from the same issue on +I will attach current patchRegd checksum calculation at the source as we discussed on it should ideally be solved by not buffering data at the higher level +This bug persists forAdditional relevant platforms Linux Cygwin +here is a start to the new approach i describedit needs some cleanup and optimization but mostly just to add a cache to like i did to +More to come +I applied the patch in cocoon with minor changes +I did some initial testing in the code bases +I also have an additional wish on the list Can this remain public API so I can use it from within my own app? Sorry but youve lost me here +libs in the root of your couchdb source +With the code change in the patch it would end up crashing with an NPE right +A deployment manifest should be taken as is +Uploading exactly the same trunk patch again so that I can kick off Submit Patch and JasonThanks for your review +It does not seem to make sense at all and aside from theres virtually nothing I can do +The only time where it really becomes a problem is with ints and thats easily resolvable +So wed have to pass a dummyvalue that we do implement INFOFORMAT its not confusion on our partits the ahead and the bug and attach my comments +done however it will work much better once we productize the TCK and add client docs +However Ill investigate the best way to add another config option and repost the patchThe solution to simple with or without token and existing secureinsecure access issues is for hadoop to properly use SASL +sorry the tests puke for me at the AOP tests regardless of this patch I figured that by then this had run +Hi Roger yes Im sure that the resize problem is not related to your fix so it will be fixed in a dedicated issue but later. +The present implementation of extends from +Koen wasnt this already done +What JDK were you testing on the Win +committed +As I see your requirements perfectly match to a Session level and Request level cache +I would propose fixing it as Raghu suggests by locking the for open +This is the list of issues that we want resolved for image bb code is not yet implemented +Utility to indexMETADATA table mutations into Accumulo tables Utility to query would add ability to consume the logs into an accumulo instance so we can query and search G of logs with greater parallelism and speed statistical analysis of traces which will probably require the accumulo instance to be be up and running +Why wasnt this committed to? This seems like something that is definitely needed for future patches to work +Cant lock bottomUp as other thread needs to be able to get in there and get so end up with deadlock +My solution is that when history is turned on the compactor will only delete old revs if they have delete nodes which wont occur normally +for patch and for does not appear to be in at all +That sounds good to me too +Its OK for meThanks +available on the autoupdate site +Patch for patch to marking as fixed +Patch looks good minor comments below +I love it! +No point of keeping this as a blocker since that is how Axis has designed to work +Otherwise there is a probability that one long mapper will delay the job completion to a large extent +Lars and TedWill submit a patch comments from Lars and Ted +Thanks Nigel +the patchlog a warning and return +Assign to Franscisco to take a look +looks like the streams are not to jbossas component +Ideally wed like to have to paired and +Will be addressed in +something like this perhaps +And making text available for search engines is a use case then we should handle thatAlso the URL in hyperlinks should be in the extracted text as well even though that is not visible typically unless you hover your mouse over the linkSo Id say the criterion should be all the text that the user can typically edit in a document +Hopefully you can add this to your regression suite because its a pretty tricky thing for us to test in our suite +The patch with the cache. +I specified all constants needed +I cant reproduce it with the latest sources +Very very odd when I try to build again it all works this time +This was fixed by r on the trunk and r on the branch. +Just updated the jars +AdrianOops sorry I did not see that this issue is marked as being a part of Tasks To Be Completed for Release +If you are using it in a multimodule project you can tell findbugs to skip the modules you do not want to run on by using the skip option +TODO after committed document warming tips due to change compare minimum count currently needed to the term df and avoid unnecessary intersection count +To configure declare a bean and use it in turn to configure and. +The default implementation will still log a warning and continue to execute but you can override this to perform the call and handle exceptions option should be useful in general and is hopefully useful in your scenario as as covered for the time being +Patch looks good some minor comments Do we need different interfaces and? Couldnt we collapse both on? If keeping different interfaces make an outer interface class should be annotated as publicunstable line has a false space diff should be annotated as publicunstable +HMS +Code reviewed +Actually since a standard holds one shared we could simply cache the in an instance variable there +I did and its fine + +I introduced a test to verify that the exception is really about the action state itself if not it will verified the behaviour with the sample you provided and it now seems to work correctly. +ear +Are you aware of the DSN support added in JAMES for RFC ? See and +i can not recognize your solutionthats probably because you havent really explained what the problem is +I think this has been fixed +Since we have a jira at this point please make a tar of your modified version of the sample and attach it +Ill add toggling to the +bulk move of all unresolved issues from to +Were probably not going to be fixing this in +Abstract class was introduced which sets egressType +Thats the use case when Setup API makes sense +This patch will contain Endpoint serialization for onFault attribute Address Endpoint serialization Test Load Balance Endpoint serialization in the patch with some minor changes +If we need we should add it to our dependencies not include it in our source code +Fixed min +The specification dated has been updated to split the row in the state transitions table so that the returned state in datastore transaction is but in optimistic transaction isThe test case needs to be updated now +I will provide a cursor processing was never tunring on the doneExecution flag thus it never sending the ready call that is needed at the end of FETCH call thus seems like hangAlso fixed FERCH command complete to send correct rows terminatetion message handle error situation in cursor execution +IMHO misleading metrics are worse than no metrics so I could see this in as well +This bug has been marked as a duplicate of +Marking these as fixed against rather than against. +Note that mvn verify fails due to style errors in but they are not related to this patch +The fixtures to reproduce the described behavior are added on the patchIll be running with both this and the other patch enabled +That would be great! Upload a patch when you have something working and we will review +Just like to add that Im running into this as well +Without having gone too deeply into the details I agree with the interpretation youre making of RFC and the consequences thereof +They have existing copyrightlicenses +The patch appears to include new or modified tests +Ran all unit tests +This has been far as I am aware this has been have created a updated version of the JBM examples +It should be possible to to disable create function as well. +select from ab inner join c on inner join d on SQLN An ON clause associated with a JOIN operator or in a MERGE statementis not valid +And theres an important difference between bothThe first checks the label followed by the name while the one first checks the nameNormally names would be more unique then labels so Im not really surprised the second implementation is faster +It will be nice to have this test case however if we ever do write our own profiler +On a side note checking the ledgers after a rolling upgrade is a good idea but I get the impression it wasnt intentional here +fields +The API offers the ability to set these options on the surface but underneath the options arent actually applied +to be clear on my earlier request c I would like to be able to pass in the actual rule text to the API not just a file name +Tomorrow I will test on another build box +Will commit after test +Is this okay? Should I install the lib edit module +The SASL exchange was amended on another jira to send a final ack during SASL exchange +replacing patch file for minor code style patch file rebases on most recent trunk +HTTP Request and Response gotten from tcpmon + +Assuming it passes what exists on Central will be valid +Report for test in Eclipse too so should be able to debug this +to trunk xAnd my thought on adding isIndexed isStored in light of. +shipped in +Resolved in branch and on +May file separate +Added an instance level flag that is used in a protected method named prepareCollection +Closing for now Karthik will reopen if he gets a repro w stack trace. +He can essentially do what he needs without built in support and personally I think thats the way to go +this patch just means you give the JAVAHOME instead of having to supply the full path to the runtime jar right? another name for the property? It was just my first idea. +Thanks Uma committed this to trunk +Hello Q gives one more usage for +Passes all regresssion tests and PE randomWrite with or region servers +I think you should make necessary arrangements to make sure that Tycho issues and posts in the mailing lists are answered +to ignore work referring to null primary keys and a test showing the problem +It didnt have any problems pulling them off there and then displaying them +First you need to mvn install dependency +If a new test is needed then create a new ticket as we need this ticket to be in the changelog for upcoming releases +I think the proposed change is correctit will require two lookups in the allObjects map for the nominal case but otoh will only need synchronization on the pooled object itself rather than the map +Can you try using Spring instead? sure is addressed in Spring and revisit this for SGF +This is my contribution for supporting tapestry runs in portlet mode + fixed bug extended for fixVersion + Modules may be upgraded to as versions resp +Cant we at least fix the jar and it +Granted we havent even slightly configured regular Solr for high speed insertion with regard to segment configs but this was good for us to get us quickly off the groundend quoteI think has gotten better with the release +Can you check if this issue still affects you? And if possible provide a new patch? +ivycache from the classpath is asking for trouble Poking around the build file I noticed that the plugin is already there though nothing seems to be using it +I will look at backporting this to with revision to with revision to with revision +Of course youre right +It looks like a bug +fixed +Added the javadoc comment +Initial patch with a new unit test +use there are some behavior differences in unit test cases I have changed them according to in +Its not happening with client driver after several testing but we can reproduce the same using embedded mode in just one or two attempts +I tried it about five times and the result was the same each timeI only see entry for vmreservationid I only see entry for in vmreservationThe part where I get lost is that even without triggering this bug it looks like it goes into then advanceDeploy then back into +So yeah the topic very well deserves an insightful evaluation and a proposal +What do you think +itd be useful to add this into the Project Navigation paragraph ve added Twitter link to project nav +Push forward at all costs including forcing someone with to upgrade just for hive +Ill try and look at it some more tomorrow +Olga please apply additional static styles to tabs and pmenu navigation please also propose some styling for new marks +sstablejson should convert values to be +All tests are passing +Here is path file +Note this is the hdataTable not Tomahawks tdataTable +Related fixes in r r r r and r +It is related to an old code snipped which i forgott to removeCheers at for the correction on calendar +Fully working +ah k +Good to go. +Reopned and left a comment +But if anything comes up we can resolve it in the next release +cezar what is your wiki username for that wikiCurrently fx is the only reliable username that I will add to the and +The includes only so you will have to rename it by appending Case to the class +since I have that patch already implemented do you mind assigning it to me? I will take that piece of code and apply there for review +Existing applications could have been using the SESSION schema +Parameter types needed to be added to the tree +I believe we have an CR ami candidate +This patch addresses the issues listed in the description +Committed backport to +I lack the general picture of this great pluginThanks! the patch with a little modification to look for only jar artifact type +Configure a limited list of SRS and then enable the Output bounding box for every supported CRS flag right below itIts all in the WMS service page +OK +AIX should now be as good as it can get +Ill make sure that the changes you proposed will be included in the next patchOn exceptions when do we use runtime exceptions? Im trying to use runtime exceptions to indicate programming errors such as precondition violations or internal state errors +That would give me something to work with +Why would that be better? The current way is how http params are supposed to be if multiple values are not ordered + +see the picure +Will somebody try itThanksSpark +AmareshwariThe reason we made this change is because we saw that even with the task which performs the deletion can still get second timeoutI agree with your point +Its work when slop +My current idea is to simply have the writer counting the number of bytes written and store it to Does this mean that every addEntry call to the ledger will result in an additional zk call to update the ledger length +Its also a plugin loader for dynamic frontendsdatabinding which the jaxb and jaxws things fall under +fixed in r +This is the patch for the and files +The shows that was loaded by the class loader associated with while was loaded by the class loader associated with +Done +I havent seen this issue outside of +I did not touch anything its more likely that the Oracle server configuration or the network between and Oracle has been fixed +Patch applied. +Thanks for the review comment Ant +Hi LarsIm interested to contribute with this improvement so I would like to know what do you think about to create an inner class inside and inside create specific parser line class? Thanks + +Im starting to suspect there is a race condition with the Sonar Views plugin so that multiple agents are stepping on each other when running sonar analysis concurrently +The JPA Provider should not assume that its internals are available on the persistence bundles classloader only on its ownIf is a package that gets added by enhancement then this is a different story +I know Ive written this code at least once +Justin what about the db tests + tested by playing with the java property +Looks like this is still open so we can confirm a change made it on to trunk +People just love struts for its convenience +Can you enable DEBUG log level it should give you more info about what is happening +The first negative test that proves there is a threading issue fails in under a second +Gzipped log file from todays run of committersrelicensesrcperl in practice mode +However I dont understand your medium wish could you elaborate pleaseAnother option would be to copypaste searches as LDAP +Ive an improved version +Patch that improves it? The patch uses simple but maybe I missed something +I do not think that timeout is issue +hi longjing geI am a little confusedWhat is the synchronized in Reader suppose to do original? If it is to protect the readSelector then the patch cant protect it any more + +Looks good to me +Ok will fix it up +Any update on when this fix is going into a final release? Im upgrading to RF from and its causing quite a few issues +I ran the file successfully with both the compiler and the for the patch Alex! I applied it in the BUGNMAVEN branch tested it and have merged to the trunk +We can this issue if more information becomes available +There seems to be a requirement for such things as qurom and fencing then +Were currently looking into the tr issue +Would you add this functionality to the new wizard +Moving it to since the branch is already cut +Attachment bug has been added with description shows size problems with bmp and rendering differences with the same attached test file does not contain the referenced external graphics or referenced file please provide in order to keep this bug open otherwise it will be moved to resolved P open bugs to P pending further review +Fix punctuation in some log fix +Right now this is happening because we attempt to do reverse DNS resolution and as expected it fails and it returns the raw IP address +The authentication is working now. +He Read above a little bit + committed to trunk. +Empty partition keys are not +The page asks for confirmation to kill the job +Sorry you are not writing it to VERSION file +Patch to add local name matches to bug is directly related to +NET installed +Yes I see the issue now +Should this be optional +the url for the Sentry website in Jira has now been Jake +from meOf course review from Mikhail et +Its now working for me on svn with Thank to have corrected itself or been a bug in instead +Attached all feedback from Reference Guide +I feel that it makes more sense to make the key to be the remote server address +issue reproducible when tooltip hovered by mouse in moment when component mouseout with offsets partially +I dont see why a work in progress should not be reloadable +Are all of the clients from? I thought some were from external pull requests +why not simply using groovy here as all components do +IT added +if dragIndicator attribute of dragSource component is not define indicator for the component evaluates as cloned object of dragSource parentThe rest is correct +committed +Weve had a long discussion about the documentation platform +Verified at +Diff again trunk revision +TF itsself is again very old API without reuse and so onBut this one is new a reusable TF is calling another TS mixing the but the changes also affect the other variants +Not sure its worth it +Can you stop posting these rhetorical questions on our JIRA please +It was never usedBy which i mean it was incremented and decremented but the resulting value was never used +Linking them so that we can track them +Ok this is actually due to a change in the stateful session handling of the security state that is producing an invalid based on an old and invalid change to the where the principal will be the identity if one exists +You are correct Filipe sorry I missed that in the output +closing will open new ticket if they really want to have warning in log file +I will try changing the dependencies for Tika as you suggest and see if I get different results +fix for finding libraries on freebsd +I agree from you guysI will revert those changes soon from release branchAshish +Are the different rows different test runs +The patch is out of date. +Just moving a print statement from one file to another +But not always +Is one of the projects tasks to improve themSincerelySamuel Student samuelqueiroz Mentor project is approved by +added a class for handling for the commands rm and a private class +Regardless of whether mutations have deferred sync we might always want to flush periodically +But actually they are parameters +I had to the patch to exclude some other code +Related to bug fix for . +Testing not easy +Mark can you post your patch +Log out from +Thanks +How likely is it really that a job is killed with kill on the process or by killing the session on the database? To my mind this is by far the least likely way that a job could fail so I would argue that we can postpone this beyond +I will apply it and close this bug +Is the problem here the punycode in the TLD +yes +We dont need to change the existing policy can add a new policy +The filter parser works so it must be a problem in the way we create the filter when we decode the incoming PDU +for removing unused methods and classes +I can NOT believe something simple as save fails +Hi DejanThanks I think my setup might not be correct can you send me on the and that you useI can logon to the console using the user and password Ive defined insystemevrstram amqAdminHowever when I try to create I get an exception and also when I hit purge the same Possible CSRF attackSean +So like Eli I would like to ask if making it a separate project and building a community around it is something you have consideredOr other choice is to contribute the style rules and static validation features to an existing popular open source templating engineBut I think the approach described by Robert Evans makes the most sense +Im sure we dont still want SANSELAN but Ill askHowever we do want the new issues can you move them to IMAGINGIm normally able to Admin commons issues but edit and bulk move are not available to me +Refreshing patch +I havent checked it on trunk yetWhere should this go? or +Also unless you are quite sure you found an bug I would suggest first writing to the mailing list so that we can help you with the triageClosing this jira now. +Patch attached to be reviewed applied for +It shows how to reproduce the problem even with the latest in our internal tests we disconnect and delete complete index directories which on reconnect get recreated +Otherwise plugins bypass the logging configuration of Maven +Thanks for the patch. +Updating Fix For to Unknown on issues not targeted for Robbie Could you review this change? Thanks Keith +User should solve this with common UIMA functionality and include +Fix attached +If it still exists with please reopen this bug. +see comments on the forum topic +d jdbcderbydblooktest usertest password as duplicate of. +And again another duplicate of the same issue +Will test it and attach the patchthanksNandana +Patch for the patches were made on top of uncommitted changes and cause some issues when applying +Thanks for testing it Bogdan! Im glad it solved your problems. +Attaching a script which populates the database and runs the query +I agree it is out of scope to go as far as a builder API +Adding new classes to a package has no change on jetty +Modified in order to correctly cast expression inNow the visitor takes into account that has an expressionThe return type of the expression must be Integer or Long +This a minimal patch to allow use of the noop setting without an inprogress repository +This patch adds a test case but the test works as the exceptions are not fatal +Hi MattThat should be ok +Attaching jdk specifc master file for to temporarily resolve the diff issue +Hello Asankha OlegI dont see how we could address this in the API +I believe Gunther asked someone to look at it +Patch that extract the relevant part from +log if there are any stack traces +If theres any specific reason please apply alternative patchBut this patch will remove postData recreation part to fix the issue +I edited out the params for readability also note that the two profiles were run on different computers one slightly faster than the other but that doesnt affect the number of calls or relative percentages of time spent in each call +Committed with comments clarified. +Applied +revision Thank you Bilung for your contributionJarcec +Hi Theresa Sorry that you so much trouble +Ive emailed Mr +xml file in srcmainjava even if that goes against conventional maven principles +Modified the patch to address your comments +Uwe if you can attach the jar here I can make the necessary code changes and run some tests again +you wanna have yourself at the first place and me at the last that does not work +Already existing enhanced to test the described wrong behavior +The ESB build script has been modified to deploy the version in the ear should now be a copy within the tools directory and another in each server profile +Closed after release. +My application toolkit only makes response input streams visible to users not the objects so Ive worked around my problem by wrapping the response input streams in my own class which aborts the method and releases the connection +I doubt if use of UTF outside HDFS will be removed anytime soonI filed to move the new class to hdfs. +Thanks +Yes we need help +Yes please upgrade to if you need to use Spring and Spring Security +Changing the title shouldnt have any effect on the database hence no migration is Jacqueseven inwhy naming a relation with the name of the related entitywhy naming a relation with a space in itAnd usually naming a relation is used when there is more than one relation to another entity like fromto or parentchild so when making a get we can have the data requested +Its not really the role of this interface +Here is the correct patch +The issue is not reproduced in seems like your case is different than endless scrolling +Will look at these fails in separate tickets in morning +Rick thanks for reviewing the patch +Moved first code over to execution project as test Subclipse is miserable at this stuff +No time for this work at the moment. +There are still some of the form callable that are made +Thank you Ravi +The ant script for this sample has not been updated and still reference dependencies +Report back soon and hopefully we can get this in + +Id suggest you add a simple way for us to pick the best compressor for our node +If patch v still contains empty line you can attach the two new files to this JIRA +Here is the patch with the getEnumget changes instead of rewrite but with reverted which was stupid +Second I tried to enable +The biggest problem is the potentially incompatible jar dependenciesHowever am I correct that basically all you have done is adapt the Solr connector to use the library and add additional parameters and dependencies as required by that library? If that is true then we need to decide what the right approach should be in dealing with all variants of Solr +Resolving issue +Probably via +This issue is still open +class propertycontextmyComp namepropA propertycontextmyComp columnpropa namepropB propertycontextmyComp columnpropbprivate comp +Christian did you get the defaults fixed +Please check the link in the description you will see it there +patch looks goodI have committed this +This difference wasresolved and now the dtd only existsin the commons validatorSo get the most recient nightly buildof struts and recompile your applicationagainst this JAR deploying all the new +It seems likely that this one has been fixed since the test case is there and there was no feedback from Hilda +and gives the same hashes for both +raising the +It is absolutely meaningless to put it in the session scope but there wont be any error message +I committed this patch into SVN +Incorporated patch in revision +Hope it works now. +The column named are already in Klingon Ill just check the data for personalpricing information and attach the cleaned log later todayAs mentioned changing the timeout solved the issue so its somehow involved even if it wouldnt be the +the patch looks good +He explained this very well at a couple weeks back when discussing and explaining some changes to which are now inI would like to commit this patch tomorrow +We can fix this issue to make the help info and the code implementation consistent here +The usage text has been updated with revision +Ill take a look right awayThanksBest regards GeorgePlease try +Also looks like the funcbr quickstart in Beta was updated to have examples of new and Regex CBR like well just need to modify the existing New Content Based Router +The startstop behaviour should be consistent +passes locally +to be able to patch I can only provide to release branch some of it may be considered new featureI will leave others to make it work with trunkthis removes a lot labor from providing patches to we can try that +I like the patch if it solves this issue and we are all happy again +The first solution is good for tiny number of users for many many requests per user +Ok thanks Ill check ASAPOne thing do I get it right the both project XW and XW will be merged? The case is that the XW was already imported from some time ago when project was moved under ASF but some issues are missing +Im using version so if you just did that Id be happy +for the patch +Installing from the RC I suppose running +Since already took care of his problem I am going to mark this patch available +I havent heard anything from our committers but Iwould imagine that somebody would be interested in mentoring students +If we can back out the other changes Id say its ready for commit +A task tracker will not start if all its local directories are not readable or writable +Ill take care of it +Please continue to follow the guidelines +Otherwise let me know what needs to be improvedreworked some time ago. +Thanks Patch for committed +we now need a way to determine that a codec does not store positionsThinking more about this +Whilst sorting that out is a very small change we should be looking to migrate our config to a more manageable format +The idea of preallocation is to somehow make the underlying filessytem allocate disk blocks for the file without initializing them +Looks like changing extern inline to static inline introduced some other compilation error +Actually reading the definition in lib shows that code reloading is explicitly disabled by threadsafe! Enable threaded mode +On the other hand its too late to get into +For the change in I should have done it in the way as inHowever makes the SELECT query a job so I have to do it as in the +I wonder is this still being worked on +I just did my best to backport this to the branch +Patch commited without amending the Database interface though +Verified with EAP +Where the child would override the parent +Thats fine Ill be able to make a switch. +Youre getting this because most likely the request has been recycled already +Fair enough +I also think that it should be a warning +Ive verified that the tests pass on Mac and WindowsPlease credit this patch to the original authors Chuan Liu and Bikas Saha +Thanks Gavin I modified the accordingly and its behaving as expected now +The problem seems to be located in line where my String array is converted to a simple StringI attach a test program reproducing the bug +committing fix on for testing purposes +Merged in the branch in revision . +Unfortunately its the last little bit thats going to stretch this out a bit +The test should only test binding to java also create an FAQ about this and how only binding to java namespaces is supported +It may actually fix this issue as well +We cant quite support that just yet but give us a few days and well have it sorted for you +Perhaps adding some more logging would help with debugging +Agreed Toms +As far as I can see the only issue is how to get the version number updated and tags put in place in the mainline CVS archive +I think the fonts look current and its simple +the way is there any good reason of validating the requests withinI would presume that its for ensuring that requests are valid before using them to generate the caches keys +New patch with unit tests +was also updated to set synhing both on the stream and the descriptor not only the descriptor as was the case error showed up because we didnt support Ruby IO on Channels before and now that we do support IO writes and read will sometimes not do anything depending on network speed and so on +sigh +GiriIf another jira is being opened for this can we make sure an link the issues for tracking? I think it is agreed that this is not a issue but rather a hudson issueThanksGary +i believe that only the content of the is parsed if such a task controller class is you dont find it in a reasonable time ping me for it and ill sort it tom +I like the patch +Maybe someone with some serious mvn foo can help +Take a look at in the source code for testsAnd see if you can get that working with your patch +How we apply and enforce the terms is a matter of policyAs I said I dont see this as a useful capability +settings folder create a new fileWhen user create some project run configuration its not stored into project +Committed as discussed above +look awesome thanks Rob and Adrien +EAP Running Tests run Failures Errors Skipped Time elapsed for EAP verified in I please get a description of this issue for the release notes? I need to know what was causing the issue what effect that had and how it was fixed +I cannot reproduce this in the latest client +However Anthonys comments do show an issue in this code +Another JIRA should be created asking for better LOG messages. +Full url for image now processed correctly +A similar hack is already present in the code in the if branch that applies to a relative path so I just fixed things up so it would also produce the appropriate results in the absolute path case +Some are but not these three AFAIKYes they are there by defaultbq +It did +This was about vpc offerings +Fixed in DR +Please reopen the issue if that is not the case. +Unless we want to write tests for both cases and maintain this functionality in perpetuity +Thanks for the response +So maybe try with eclipse plugin or or whatever an older version that used to work +When I receive the quote Ill post a few different color combinations on the design site for you to choose from as well as the correlating Cheyenne W +Thanks for verifying +Applied patch. +I may be able to find some time within the next two weeks. +I have spent quite an good amount of time on MINA since Norval pointed out the problem in order to be able to get a grip on MINA +And the keeps reference to all exchanges which means we end up with a lot of Exchange at once in memory which eats up memoryWill have to come up with some way of discarding not needed exchanges during processing +Antonio Souza wroteThis happens because theres no endpoint declared +Fixed on master +You would have to supply a that is using PKCSIf you wanted to update the tomcat https stuff to use a as well everyone would really appreciate it +One of the methods was actually are updated and one extra test added +How can I submit a changeset for review +for the patch +Sorry I cannot help because I could only debug the code on a Windows system +Patch for trunk after merge reintegrate +results using specs and openejb SNAPSHOT and hande installed activemq jars +David I dont see and in zip tomcat list above but these are present in rpm list +I have thought this issue for a long time and I think preventing processing of SSH is a clear and simple solution otherwise we should consider many cases where meta server died in different time during master initializingWould we do the above as part of another issue Chunhui +Updated the usage as per comment +If we set androiddebuggablefalse the get blocked in the main web view but the solution i posted above works for the main web view only +subclasses and custom qvalue factoriesmay provide a more sophisticated approach +Closing the issue now that it has been resolved. +Patch applied. +Verified in JBDS Linux +Fixed in r + +Sounds good to me +It now works greatThank you Henrik +Kurt can you add tests for the Aggregator and CBR actions in the package? Thanks +Logged In YES useridFixed as of release. +So not a huge deal but would certainly be nice to fix for the windows folks +We gonna reopen it in anyway so just go ahead and close for now +My patch does not wrap long lines +I tested bitsI still working on an artifact that will contain the most used boxesI think should be the good one +bq +The class is not delete in patch file +A Configuration already gets passed in when a schedulers reinitialize method is called +I propose here a solution to resolve to have a look on it and give your diff a treat +Fixed other minor items like the rough count at head of master ui so now its more likely close to actual region count +New patch fixed the scenario when its last also return FAILED the patch again to return ERROR state in case of last retry +bq +Updated patch files received from Martin Elzen these appear to work with the current trunk +Fix the failing test presumes hardcoded HBASERETRY but that is going to change +json or +Which perhaps should be optimized to avoid synchronized call +Support for using at the was added in usage of has now been extended with support for class modes which can specified via the new classMode attribute of the annotation +Ive committed this +Fixed in r by adding explicit return type to methods with added Native +Resolved by Daniel +I realized that doesnt make any sense having an VarcharChar field with default value and that isnt an problem that Hibernate should resolve +Im not sure what good that does to put a latlon in a single String in georsspoint format +I guess it has something to do with the way plexus build the archive +This adds some more permissions to the default test policy +Apologies for the JIRA chatter +This issue is closed as out of date as per the middleware process rules for aging of bug reports +The Directors view is that Despite concerns from some developers that it is not sufficiently liberal satisfies the key requirement of permitting the creation of derivative works +I have found out the problem it is +if a document is added a new is created for that segment with a new creation tick +you Ronihi for the review +Thought I got all of those +By not yet ready do you mean early in normal startup or started up with explicitly not joining ring +Please open another JIRA issue I will provide a patch +Here is a simple patch that does this +Revision Author ritchiemDate January Message Added exec to Modified incubatorqpidbranchesMjavabrokerbinlowercase Qpid seems to have failed to register with JIRA as a commit +Putting up or shutting upThis splits apart whether to count terms +See the unit test for it in +Ive merged a number of patches +Move fix version to in preparation for release. +See Effective Java nd ed +Im attaching an updated patch that applies to the latest sources +Patched in version of out to the prod +I split the Patch file and made a seperate one for the upgrade +Well since no additional info was provided I guess this is fixed +Evgeniya please verify +I had couple of questions that I posted on the rb can you please answer them? Thanks +Fixed in revision +Your right shouldnt do that +AttachedThis issue also applies to exists +I Waiting for a fixAnd thank you to modify the title of ticket +I know it currently IS a declaration +But they are indeed for the implementation of each file systems IO streams +It works yes but the addition of random methodName! null into leads to unreadable code in the endSome reshuffling would do the code good +Does anyone remember whether export controls were discussed during incubationThanks anyone remember how Jean determined that is the correct classification for Derby? Thanks +Committed patch c to trunk with revision This concludes the initial work on adding support for +to the suggestion to cease flattening NOT EXISTS although if you think you can tellthe difference between queries that are safe to flatten and those that arent that would be fine too +Making null string configurable is probably over engineering at this point +It is the first cut of implementation +except that Navis for review +Rearrange indent and code styleBut there still remain many parts required to be refactored and improved +Updated documentation have a couple comments on this +Seems that the servlet goes into an endless loop +Not required closing. +I have an idea how to fix this but I dont know about the implementation as it currently stands +Agreed specifies this +now points to the +ive replicated it on a simple web project using jboss but i have no idea what can be causing it +Let me try changing all of them +Ive filed it would be great if someone who worked on this code added documentation at an to the ref guide +But for a future framework havt information about styles priority +I think it should be to the patch +please stop submitting nonsense issues +No further comment from reporter +So may a better idea is to wait for nutch meta data support +Let me know how this looks +BJNope only if the desired role or role is NA +What you mean by wrong? has a new groupId and artifactId instead must beAnyway Ill commit the for testing +Committed thanks! +Now I remember +gridchukwabin +is imported by but only for javadocHowever as I mentioned earlier is a potential unifying api class +Please disregard my last commentRegards +The patch follows +Marshall please review +where this might also be useful +tools dir using ant junit +The patch was checked in but the build was still broken +The other question is the conversion cost on performance. +This seems to be more a database setup problem where multiple schemas are the schema visibility in the database would fix this +there are still places returning zookeeper return code as bookkeeper return code +AntonioCan you please resend this patch with the latest trunk? Sorry for the delay +But under high load you can have a significant amount of inserts done while flushes are queued up in sort write executors and this is where you run into trouble on replay +See the new patch +Set a flag on Scan and Get? Hard part would be adding the info unless we did something like add it as meta on Result returned? Should really be in envelope of the response rpc. +Any suggestions are welcome +I cant reproduce this error +Leaving Jira open pending code review and commit of fix for trunk and +I think this is a misunderstanding probablyThe status code error handling scripts are only triggered by the calls + it should work but I think something went wrong maybe an unclean state caused this? I dont have a setup and may not have free time this week Kishan or any other dbcreator gurus can take a look to help fix the blocker +but if you get thisresourcesmarkup how on earth do we know what to getBecause we lost everything that is related how to get itWe lost the classname that we want to search for +will still fail as it needs the new editsStored which is a binary file and cannot be added to the patch +Makes me think this may be a pervasive issue in all Cloudstack debs +Time for the reevaluationsWith the previous numbers I would have said Id it +Make clear that the information relates to to CP +So now everything says +New is out +So users with appropriate privileges should be able to edit the project add releases assign roles to users edit issues etc +OK after thinking about the fix there should really be no issue with them being switched unless speculative execution is enabled and something very odd happens +I know it took quite some time and nerves to locale and obscured other issues really sorry about it hard to predict all the in a codebase as complex as this one +Thanks Chris and Nicholas! +and if a project has a runtime that does not exist put a warning marker on the project dont fall back to some default +Each one of these tests have a query without an order by so the order is different between versions +I just committed this but without the class +Please verify whether the problem is resolved as you expected +added the whirr deployer role to the user +The other is an NPE at line in +Implemented in SVN trunkOleg +Comments added in Review Board +You mentioned that the unix touch works as expected so the permissions part is not a problem we can strike it off +Its more like settings based on the default formatter +Sounds good to me +for this thanks for the comment +Im still not sure whats going on here +On second thoughts that may actually be right should arguably make an explicit call to if it gets read only from the doPrepare +The userpassword files are still there even if not in use. +Please verify the patches for imap mailbox +Thanks Ning +this is a duplicate of isnt it +Marking this as resolved as many of the annotations have now been javadoced +The point of that is simply to indicate to the user of the API if expansion will happen or not +Where the conf was being used for settings involving only or user properties they were shifted entirely to the with getterssettersThis was a lot of work +This is probably related to +Be default all outgoing JMS msgs will include expiration time equal to the timeout value found in the deployment descriptor +Patch applied +ReaderverifyTerminator allow more than of padding at the end of the edit log +Attached is a trunk patch to inline the doPrivileged blocks and remove +Cool good information to have +This is a screen shot of the confirmation page +Patch applied in +We realized that such a profile would be helpful and recreated it as part ofThe current trunk build already provides this fill profile a release is imminentThus I would think this issue could be resolved as a duplicate +Quick question while it compiles +slip should be fixed before GA though +Resolved thanks for reporting this issue! +I thought Ant was case insensitive in which case we have the former +Please add a comment about the switch as its not obvious and would raise more questions in later maintenance +Archive not createdCan you explain how you created the source jarCopied the camel solution +This allows me to look at the method calls for both configuration elements +on the checkstyle fix +I try to write a patch that reuse the only single update requestSorry this patch is based on branchx and I write a simple test onlyMaybe this patch that improved a little +Schedule for +Now when you start the server again with the new deployment settings the deployed project is not in both locations but only in the original server deploy folder not in metadata +It looks like the capacity scheduler already rejects the app in this situation so Ill just fix the fair scheduler to do the same +Converted servlet to context listener +very expensive? Wondering if something changed recently I did profiles where the whole request took like ms and havent seen this lookup show up prominently in the profiles since I made cache the bean you have a benchmark before and after the change showing a significant difference +changes do not change the results of functions +If its run as root fusedfs can get access to kerberos ticket cache file of the user performing a FUSE operation +Silly me Ill update the rest in a moment +Could you please provide some more details on how this feature would be useful? If the file didnt change at all isnt this as good as a periodic wakeup something could do? Thanks +What purpose does this serve? I would rather the methods be called only once + Username triathlon +by cschneider Reordering packages in fixing bug in checking if an Endpoint is already importedM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerAnd more +BruceI have commited your appender into the trunk +Generally extremely stable used extensively within google well tested good test suite etcOne small aspect I dislike about Thrift RPC is the inability to return null objects +resetting P open bugs to P pending further happens with Adobe X on the Mac does not happen with Macs own Preview +So this package depends on only a few parts of the io package +Appreciate the help +Hopefully the HOWTO describes how to run the CTF as I already have applied the patch locally +Thanks for reporting this! +Recommend moving them to debug level +This should be done +Confirmed that this issue still exists with Firefox Chrome seems to be working with the correct behavior +Could not log in +No joy +Testing in progress +I vote for Option +Fixed by not instantiating all plugins up front in class transformer. +Steve has added hibernate to the repository now Emmanuel needs to verify its compatible with HEM and alter their accordingly +put the query out of its misery earlier rather than later +Somewhere I do have a patch that needs rebasing onto trunk +Apperantly yes ECJ is compatible with javac in the matter of debuglevelSure stack traces need source info addingWill fix tooIll look into reusing but not in this JIRA I suppose +Problem was that it was storing suspended deletions as an instance var on the +batch update of issues with fix instead of the correct in +any logs +I quite like all thisGood luck reading the patch +Heres v of the patch +The fix looks good +Trivial bug +It looks like the first dao is reinjected +Sending srcmainjavaorgapachegeronimogshellwisdomregistrySending srcmainjavaorgapachegeronimogshellwisdomregistryTransmitting file dataCommitted revision +Its better to know about a problem than to just hide it and not being able to contact zookeeper for two timeout periods is a problem. +If I follow correctly youre saying that the BR contains the block with GS but the NN has already seen the synchronization at GS so it marks blk as corruptShoudlnt this be OK? the NN will then issue deletion of the replica with GS the DN will ignore it because the GS doesnt match and the corrupt block entry will go away? +Added back the based on our findings with EPP CP release testing +Thanks for the patch Suraj Varma +Should be should exclude numbers that are defined in be better! +This causes the stream initiate done message to never be received because doesnt actually retry it only buffers +Two first patches one for DRLVM index patch rearranges links on the page adds more info the other for the file to bring it in line with formatting in other pages and enable correct display +rename add toutctimestamp fractional timestamp precision siyings comments from review board +we can move code sections implemented as reflection into version specific implementations guarded by Maven profilesBig on thisbq +If I send a message I cant expect it to be available right away with if I ask the queue browser to see if a message is available and then try to receive it with receive I would expect the message +Anyway here is the unit test maybe you want to add case committed with svn revision +The patch doesnt solve the problem entirely +It works like the new svn patch in svn trunk It uses the revision numbers in the patchs file headers and fetches automatically the requested version patches it and then updates it to the version of your checkout +In src assembly the prefixes for the +Could be similar to Dan +Committed. +sorry about the partially incorrect comment early for the patch feedback Paul. +For me contribs are an for various reasons weve discussed before on the dev listNow that we are a TLP we could consider hosting subprojects for things like this with their own SVN trees and release cycles +the Container option from all the screen definition of projectmgr component wherever possible and put the screenlet thereThanks to Ratnesh Upadhayay for helpingThanks and RegardsAwdesh Parihar +Ive spun up a separate issue for future all tickets +When it is a fairly obvious that bug report is not a problem in struts we move the discussion to the user mailing list where some might know the answer +Added a warning if session timeout is less than connection timeout +Hmmm Mircea I dont think youre understanding the meaning of first available +changed the query in Also fixed the get so that it is deterministic now +Closing resolved issues +JBT work in this complete +A flag on IW makes things simpler from the user perspective though cause then there is no need to customize the built in merge policies +I was trying with eclipse and it was giving me a conflict +Committed we know the github revisiondate corresponding to the zkclient release +As of today I also need to do this check when parsing firstNumPlease note that my quick fix prevents the group listing from throwing abut at the same time makes it impossible to retrieve articles from the group +This patch fixes issues you mentioned +I think that the class requires a better explanation of this case when it applies and how it is handled +out of hunk FAILED saving rejects to file coresrcmainscalakafkaserverpatching file coresrcmainscalakafkaserver +Ill try manually insteadJust so that Ive got it right You renamed the srcit to srcit and modified the verify script by adding a check on the attribute + +But we didnt complete that work +Aha a directory share was created on my C a lot of experimenting I get it anything between the first and is assumed to be localhost and thus ignored even if it contains results in the following behaviorurlfilenonexistenthostsharerepo error as expectedurlfilenonexistenthostsharerepo a directory Crepo is createdI wonder if that is desirable +with only HDFSetc +set to LATER since there was never ever anybody complaining about it. +Is that something you are prepared to do? Alternatively you could quite easily script something to upload that directory into Archiva with default but then you have two copies and you probably only want to maintain one going forward +stagingjobidQuestions How and when to cleanup the staging area +Closing +Ajith is still working on itthanks in current works +verified on +What is the point to print only launcher version after all +Yes both zk and hdfs should be up and running +Patch committed to trunk and yes applied the NCHAR and NVARCHAR mappings directly in base class +patch v addresses findbugs adds some comments on pooling and switches some junit integrate later in the evening if there is no further review comment +I also added a test case under rThanksRaymond +Marking down for and +If I choose sv or any of its variations and have the sv bundle in the repository it resolves that fine from just the header + +Also stumble upon that problem from time to time +just cheking s! null and added a few unit tests +Fixed in . + tested! +Use Hibernate was refering to the AJAX layer and I did not close the reportOn a side note HV does not need to do selects since it can apply constraints on the hibernate metamodel end then alter the DDL statement +Revision . +This issue is fixed and released as part of release +Patches for the webbeansri and projects which adds a new service bean to bind the Manager to JNDI and removes all the naming context support +Fix for since this is a customer resolved issues for already released versions +Otherwise looks great +branchx version of config files including codestyle addition +Works fine tested on NB +bq +Thanks Sarvesh! +Patch applied thanksMSending modulessrcjavaorgapachegeronimotomcatdeploymentTransmitting file dataCommitted revision headCommitted revision . +The new patch based on the new patch has been uploaded +Junit test was written +The attached patch contains the refactoring of the abstract class to an interface +This has already been fixed. +I do have some questions though Im guessing that the test runtime should dynamically figure out what sort of classesjvm version it is working with to execute the correct testing framework? I cant imagine any form of property setting logic being fun for anyone +We can reopen if needed in the future. +with latest build +The major roadblock is when the node is not yet exist then there is no way to lock it except creating a fictitious node +Does the directory contain a s entry? What is the exact cmdline and the exact contents of the directory +I will work on the unit test code as what you said +I committed the patch to +IMO we should just stick to compilemaster +As far as I know this is how it is done for thrift dependencies in other projects +Source code for webapps. +Make sure you dont have another orfop classes or hyp files somewhere in the classpath +Minor delta to earlier patch which includes the minor changes from and also a small change in to include a static member for jobConf for use by recreated patch with latest svn committed +It will break some internal things and likely some user applications that depend on the documented behavior +Ill post a comment whenI have a cleaned up patch ready +Eh rusty perl find +Tested on version +I do not think that was the intention +Working as designed +But if you want to try to works on a patch for this issue that would be great +Ive uploaded the patch for the issuePlease review it +Ive hide use extern process option till the time of resolution. +Heh I went to all the effort of writing tests for those cases just to not include them on the svn diff command +Thanks Brian! +If you want to have a full discussion about your proposal please let us take up the conversation at dev. +After digging through this issue +Thanks for the feedback Marco. +I use GA +runtime or tooling problem +v is the patch that was up on RB +We cant commit any patches that increase the count for findbugs or javac warnings +Excellent +Funny a colleague just reported me the same issue with the very same value from eDirectory +However no transaction manager standard or implementation since the s has been able to live in that world since strict PC is a blocking protocol +Ill commit soon +The xerces classes are not in the java +Im assuming that did not offer the Velocity? Or was that part of the velocity contrib and it was move to coreWould prefer if these libs where not required unless you used the functionality myself as well have not look to see how easy that is to do +hadoop package? Im going to rearrange the bits already in there into subpackages as they no longer share one related purposeFrom there if I have any other comments we can look at those after its committed +Any objections +Great Ill fix that now. +Discussions on led to a and a vote for LATER +I suspected it might be a big taskMINA object stream classes would be quite helpful especially if they could be added to existing code with few other changes +Still appears on HPUX with to a next release +Should I open another Jira for the EAP fix +Nopes this doesnt exist on Kryo serialization and it doesnt repro with javaXML if you have applied +I shut down that old one pointing to svnI created repository in reviewboard pointing to your git repoEnjoy +Fix comitted and deployed to and master branch +It was the easiest to do and matches RPM conventions best. +Closing. +Hi folksCould be treated as a solution on? Or any other features are requiredFor now could stop service of failed disks and could add disks or disks after repaired them by means of reconfigure based onHi Wang Xu I have some comments regarding +No good reason that I can think of +Any opinion if we should this so someone can create an port of? +eclark reported some success with this patch on IRC but intended to do more testingI intend to commit this and include it in a new RC on Monday +Committed to trunk. +In this case any EJB would be injected +My assumption is that the problem is either the log record written out is bad or the recovery process on the log record is bad +Thats when I noticed that somethings wrong +Renamed to +Instead of implementing the interface at its better to add a new class which is implementing +We cannot redistribute but the code can certainly download it +Start by consolidating multiple calls for data into a single objects +Reformatting original description +Ill reactivate it and add a patchLets create a separate Jira to address this +updated version of patch attached +I dont follow the optimization of the starttime Ignore +Hope that clarifies things! Go forth and propose improvements in this area Im a fan +HiI have sorted the problem that I reportedIt is just needed something to be added to the disconnect method +Without artefacts to go into the Maven repository then the Codehaus repository holds little real appeal compared to housing the master at +with only one change in the name of js file +This should help with reproducing +passed regression and DABCE for the target loop works +The shell can use the client side interface +Jeremy Ive assigned to you as you opened this issue +Adrian I tested sounds better but not rev Jacques thank you for your assistance +It also includes one that tests this against the failover protocolNote that I have copied the keystores that part of another one of the activemq modules and have not created any new ones +fixed at rev +JacquesPlease request a vote before a final commit +we just tested the same on ACF and got exact the same result only difference is the function that is not existing in ACF so we get a other exception in that case +Adding to identify missing CR issues +This patch is for post unless there is a take of patch as I missed one +Thanks a lot for the explanation I think I get it nowThat sounds like a good plan +This way the ordering is preserved even if you retrieve the dependencies from the XML fileYou can revert the changes made to and or you could keep them since it saves you from an additional parsing when you do the resolve and the cachepath task in the samecould you reopen this issuethe problem still occurs when you resolve a configuration which extends another configurationIll upload a patch but I have to test it a bit any news on this issue? Did you checked in your your modifications fixing the problem +I have a test case that reproduce this scenario and the custom manifest is anyone close this issue +bq +Fixed on branch a long time ago +I think its something like of +BTW I dont think that it will take more than about one or two hours to realize it and it will disturb nobody so whats the problem with it? The alternative for me woud be to look for a way not to download the scm plugin automatically and instead make private changes to the plugin +add code tag for easier reading of looks like cglib isnt in the classpath +Reassigning so that it gets reviwed +Then its a java issue +For future patches please be sure to work on a directory that has been checked out from the apache SVN repository +correct? +Ah that makes sense +I believe I have identified the problem +block notation support for nested style block syntax in the fromto keyframe definitions support for aliases and variables in those nested it turns out the was implemented as a pseudo class not a pseudo element +ooh i actually do liek that +A lot has changed in jetty and xml munging since this report and I believe both jee and earlier web apps now deploy without problems +sure +The message that are in the core itself I moved to debug +Fixed in SVN rev . +If I do a generic replace those quotes would be duplicated too which is not what we want +Short of redefining that rule detection of would arguably lead to messy semantics +Appearing to have the same cause as +Added custom Thread factory to the that creates threads which Initialize in a primitive controller +Updated with feedback from reviewboard +For storage on local disk we probably should support writing a second copy to NFS so that a different node could bring up theIn any case the extra completed task state should be on disk rather than ram +Thanks for the reviews Bobby and Eric. +The current runtime depends system allows for dependency testing at arbitrary points in the sequence of action in a target definition +Let me know if you think something is missing +patch checked in + looks goodIt would be nice to add a test so we dont get any regressions in the future +Committed to branch and trunk +I was meaning for the scenarios where you know what the path needs to be in the URLFor example you may have your app deployed in Tomcat as but you want the proxy to use the root so a request for httpproxycmisatom gets routed to httpbackendcmisatom +Committed to trunk +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +is supposed to be released soon and be backwards compatible thats why I changed the target +My suspicion is we need to have an actual Hive set up where the metstore is populated with metadata +Committed this patch to subversion as revision and cleared the flag +This patch fixes the issue +Im working on as well as +Commit eefccfaafbdbfBranch this record since is now released. +My biggest feature type at the moment has records I cannot image that this will possibility is to implement a cache in the dialect +the policy files in places +Tomcat has not changed in my test configurationin common +It does not require that the option appeared in command line +On further investigation it seems that as you suggested we were using the sjsxp event reader when we intended to use Woodstox for everything +I see a few problems in xppxpp The directory doesnt have a jar in There is a confusing version in the directory +Kits artifacts are copied over to Apache Karaf distRepo is promoted to releaseRelease notes are updated on wikiNews is updatedDownload links are updated on wiki. +Ill have to get back to this one laterThe project is still in a state of getting properly configured and the build procedure may change as a result +we use that heavily for test setup and tear down in Hector +Another thing we should consider is to do the new API methods which will receive a Context object rather than a bunch of individual args +Will update by tomorrow +This should have been corrected with the fix for +Format processor that I mean is the regular Camel processor +updated review board as well +I will attach a new one +What thread model are you using? An ordered or an +Do those contributions need source code +worth another look since some of these sources may have been collapsed doesnt really look the same anymore +Id be happy to include an implementation of a with this work +Thanks for providing the patch. +It will be repaired to generate the same class output as before +EricThere is a little problem when attaching the file you forgot to check the ASF grante licence box +Fixed on revision + +Currently does not build +please signed regression test this patch with EAP Customer and Eng have confirmed the patch works +Ill do that while Im working on the thing if you dont mind +Verified. +verifed on chat by snjezana the old id is back restoring harmony +issue as this still exists in trunk and the fix needs to be merged. +Is this also needed in trunk or can we resolve it? Thanks +The autoStartup flag is checked upon refresh and if true it will be started. +If you plan to compare logical plans this belongs in unit tests using junitI would encourage you to also write tests with the new sugared syntax using old standard syntax as the verification script +Im guessing thisll be included in the next ESB +What youre saying is the same confusion that seems to be the reason for this setting to begin withIt has nothing to do with CPU cores +Looks like support for JDK collection types has been added through other issues. +I was waiting for Skye to update the patch to include the missing test +This is irrespective of the issue of fullabbreviated class names With a full class name there is no ambiguity +Misbehaving loggers should be removed +hence the patch provided used the approach discussed above +Patch applies fine for memvn Dstagebootstrap mvn the bootstrap works but the mvn fails due to the using plugin in same build problemThe first few applications builds fail with messages like Error assembling WAR Deployment descriptor Usersdavidgeronimosvnapplicationstarget does not existI gave up on applications after of these +Do you guys have any idea when you are going to add this patch to the code? +However instead of having a and multiple monkey implementations we should have one monkey implementation and a policy factory I think +Enable it again when we have committed when it makes sense +Its been running in a loop for minutes now and doesnt seem to be hanging +Patch cleaned up tested on all browsers +Patch applied and loaded to newest snapshot +This will imply incorrect behavior when currentslot becomes +Not sure how the code allows it right now though +So was my first impression to be honest +Discussed with Lars too +The changes were adapted from Adrian Taraus patches for +Please review the attached proposal +QE should verify that head requests stop in this case +bulk close for +But for one class inheriting from another case you could check that the super class also had a field definition if you had Override on a field in the child class +So I +There is no problem with using XML editors +This will be addressed at some pointYour form does seem to make sense assuming that is for the EBWS service and I believe you should replicate this for busproviderlistener +I just committed this +perhaps the project does not load jQuery Mobile because of this? Also I have set to YES + +Try running plugman prepare and see if this updates correctlyOn the other hand the prepare behavior is supposed to be run after every install and uninstall so you still shouldnt see this +Now that is using Ldap it should be fully synced +lets get this asap if possible Id rather not have multiple apis +In the case of Android the world market contains many many Android devices still it is important that our tools be able to create apps that are testable and deployable to Android the case of well iOS does not have the upgrade problem most folks move to the latest iOS quickly causing developers to also have to upgrade their s fairly quickly +When you do FilelistFiles on lostfound directory it returns null +Will fix the scripts with this issue +How about thisThe idea here is that either you the system do the right thing or you take over both and management +fixed in SNV transition resolvedfixed bugs to closedfixed +I tried the failing tests locally and they pass +Attached a version that can handle the cases when another field is added in the future. +Can you tell me what settings related to setup are required for testing scale vm scenarios? I have enabled the option for zone still in KVM its failing saying This operation not permitted for this hypervisor of the vm and on it says Dynamic memory control is not enabled for the +Patch looks pretty good to me +Marking issue as resolved for we can close it when the is updated on trunk +left recursive though they may be +Hopefully this is not one of those +Restored entry point in any changes related to this that dont make RC will not be happening +I think theres also a of the CRC stuff floating around in the tree +protocol +Immediate experiments no going well so bumping this to +Hi Martin! Sure! In the new runtime wizard when creating an as eap runtime there is a combo for JRE where you can select the JRE you want hte server to run with +Fixed at close of resolved issues. +suggestions on cap +GPG signing is activated when the assembly profile is activated in both and trunk. +I dont see a problem with this but Im happy to iterate the patch in JIRA as well +In order to see this issue you need to force to always use MTOM +I agree if cosmetic changes happen to the script all bets are off and youll get a different signatureAlso agree about the items out of scope here +Thanks Konstantin! +Jimmy Once we agreed to option above lets just add all useful methods to the interface that would include the startend key methodsJon Excellent point +Those have already been taken care of +the current implementation works and basically offers the requested functions +Looks like an issue with the dont have much WS expertise on this team +Also combined two AMQ related fileSets into previous change to combine AMQ fileSets into +This may be a duplicate issue as I believe weve talked about this issue before +I should not have to add a path to the it is the framework for this project +Partly done +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Worst case just need to have one group per table and so we can define rights per tables working with the related groupSo if I understand correctly this as been started but is not fully developed? Do we already have another JIRA to follow that? +Aligning this area of work across all interested parties is critical +argh in fact I changed that from logging on the interface to logging on each if you have a big interface with many methods and just constant it will be better to point directly on the constant +Using JAR tools Ican see and extract that very same class file with out any difficulty +Have checked the possibility of using Asynchronously index update +thanks for Ivans replies +I agree with removing the automatic initialization of the but youre right I dont agree with adding it to +I miss the NPE handling in my last patch but the question still I would finally suggest mean the word locale and timeZone are kind of reserved commited the patch in trunk r but tests dont pass +Could you perhaps rewrite this to be a little more clear +Here comes another patch with a proposed generic solution for all header delegates +We are removing user from container as it is redundant because container already contains user name +Im not sure how to delete Items from JIRA but the date should identify the new the small triangle next to the to get to the Manage Attachments menu +Once a project indicates that it is producing compatible artifacts only compatible are usableenabled by default +Unfortunately this was not successful leading to the creation of this sufficient background for your purposeKev +Richard you are the only person around over the holidays involved in the Map Layer wizards +I now have a much better understanding of s purpose and scope +Great! I will take a look at this today +NickCould you please review +Good suggestions Dag +These are all important factors +LarsDoes Jimmys comment above answer your question +I just tried and my tests passed as well +we wont be able to fix this +Im a prospective student for GSOC Yun Lee +Funny +Please can you try this on As and if the issue remains please feel free to +make more robust +After more discussion with Marc we have decide to yank support for the encoding attribute and stick with encodeAs. +right +I just check in JDK and there also it work you using Windows XP? +This is implemented +Attached herewith is a sample schema document with type some minor changes to the way the default type definition is set for an element declaration +It also doesnt support an important property that we desire where checksums are verified as late as possible so that we catch more memory errors not just disk errors +Just one thought we should be careful with adding these workarounds they make things even harder to analyze and they may negatively affect performance when the checks are run for each and every serialization +Thanks all. +Why cant the starting of the server trigger a Show In Console selection for me? As a new user I am just watching the Starting text for a long time in the Server tab otherwise +If you look at the class you will seen an example of a handler that loads a delegate in this way +Adds a Unit test and Changes Moti! +I did some more messing around and I was wrong about the FS stuff +Shouldnt have been resolved as fixed. +Fixed code for which does proper URL decoding +I tested and together +Can you provide some info how you are planning to optimize it +Thanks Areek +It would be very helpful for eclipse plugin developers to build eclipse plugins using Maven if this improvement issue is resolve as soon as possible becuase I do not find any other way to build plugins using Maven except after this improvement is +A best fix would be a new lock manager interfaces that allowed it to wait for as long as needed while insuring it was not waiting on parent transaction +I think it would be a text field because Id rather have a blank field than a zero if the issue was not migrated +Moved to where Mitosis will be to we wont release a is no longer present in the server a new replication system is has been developed based on RC +archivesfeatures +Once is fixed this shouldnt be a problem anymore as access to Account application or wont be allowed through HTTP protocolFor now I am leaving this open because buggy behaviour still exists. +You can take a look at the screenshots to see what I meanRegards +Committed to trunk. +Thanks you please close the bug +bq Is there an easy way to know which Req +I have been struggling with this issue and the few discussions that I could find on the internet suggest using a fake password field in the database which is hacky and I would like to avoid +Everything is fine +Any update on the release? +Also there would have to be quite a lot of changes made to production and work effort if we wanted to introduce a duration type there +Igor this patch fixes the issue +The underlying code that uses a range slice since a ndary index is involved was using an for the end key of the range which outside of being useless in that case is actually wrong with random partitioner +Good idea +in branchAny issues please let me know +I would fix it myself if the fix wasnt as simple as just adding a boolean parameter +Created in Eclipse manually removed +Configuration for test configuration for Application without private usage +In the inner loop it builds strings for each node to test if the nodes are related +Marked as related to +Please reopen when you have a patch + all manual cleaning added to the build +Please show us the test that proofs that this is done +I have several errors in the other tests as well +Committed change. +Really Id rather see Nightly and Stable since we only support the two types atm +Defer all unresolved issues from to +Heres a patch for this +Resolving +Committed revision +I did manage to see this with MB of RAM and a heap space of MB + SOA TP Release Note Process method now asserts that parameter payload is a Map +Let me try your unit test and see if theres an obvious solution +The idea here is not to cache the end user result sets but cache the intermediate results that are based on the source results based on design time hints +I have it dumping the times taken by individual test threads so that they can be for request latency analysis. +I just committed this +The Create Projects section contains only Eclipse wizards +Thanks +This is the point of a modular toolkit you combine the tools to solve advanced problems +His presence simply changes the logic of recording volatile index from temporary document buffer to recording each changes immediately to the disk after each save +Does that really remove the confusion? Not sure I agree with that +Do we have to attach this mojo to phaseIHMO comments are welcomeThanks +I am not sure that changing the order of startingjoining the threads actually helps I think it may just be a distraction for the reader +Sounds very reasonable +Im very hesitant to inject Hibernate specific workarounds when using JPA +Resolved by upgrading from warning to error when this condition is recognised in committed in resolved issues +batch transition resolvedfixed bugs to closedfixed +Profiling branch checking if key is of root region is consuming loads of CPU +Attaching patch with a test caseIt occurred to me that we may want to file a followup JIRA to clean up a few things +So we have to revert the changes and we can do the changes after +They should now appear above standard proposals +Sounds good to me +You can handle the new API +If would be great if you can list out which one is used in each operation +I have resolved this by copying ant jars to my hivelib +Portuguese Translations for rev +Mass closing all issues that have been in resolved state for months or more without any feedback or update +OK sorry +If the first job is not empty we may merge st job pipleline into ndth job however we need to make sure we sample the input after the pipeline and also partition the input after the pipeline +I dont know why I keep on thinking about distributing JS in the future as a war file +I dont see how I can make this work in all cases without forcing the user to provide a parameter or use something like the file that the release plugin uses +Change Avro reflect RPC so that thrown exceptions modify their message string to contain their name +Would you please verify this scenario as well in verifying this bug +Fixed in revision +Thanks for putting up an issue +Will this include Hadoop too? If so do Hadoop jars need to be published into the apache repoI suspect we may want to tackle the Hadoop dependency in another Jira as it could take some time +I removed Serializable from the implements clause +thank goodness not everyone lives on the west coast it would have tipped off into the ocean by nowthanks for your patch +Does workDoes any thing need to improve +the serialized file generated by RI is attached +You should read the sequence files holding the results of the clustering. +Minimal source code to test the case +Created a patch that fixes it on my desktop +Verified in CR build +where is the real bottleneck that we are trying to avoid via this proposed mechanism? It appears that we are trying to avoid copying lots of data via the network interface +I reactivated the marshalling test in JDO example using referencetruealso changed xml to seems to me that the user is responsible of avoiding circular references whenmarshalling an object graph to XML and the relation is acircular will close this bug except if Arnaud can provide a better fix +Use the correct variable for hiveuser +Eugene no I cant upgrade Eclipse because Im using some company plugins that require eclipse +I cant reproduce with my current version of plugman and CLI +Thanks for providing the URLs for the documentation by the way +That would make sense since and areSerializable +This issue is closed now +Ha got a test case that reproduces it will attach it in a moment +Different keyspaces can not use different property filesGiven that I dont think we need a snitch in this case +Cartridges are loaded lazily +Ive sent an email to Roman asking what his UID preferences are +srcmainjavaorgapachewicketextensionswizard +gcc is out since June so the workaround is no longer neededI have not tested this using GCC so it would be nice if anyone could confirm that this works with gcc or later +It doesnt fail in my end +One concern is looks spookily heavy weight eg I dont know the typical cost of reflection +Oracle perf db is blocked by ESB testing so I am doing some preliminary tests to get a least some results now +Thoughts +AIUI the big problem comes in when you have really dense areas that are used by high traffic sites such as Manhattan or somewhere similar and could have a million latlon pairs all in a mile radius +If thats the case this might indeed by a problem or maybe +Split implementations out from and move out of classesEither the get methods of an will be called or filter but not both on the same object +corrected spelling error +starting to verify +Be my guest We can do that for both and +Denis thats +In I tested this patch with block synchronization turned on also works fineCan somebody please review this + on +Should we close this or do we support release candidates when final versions are available +Ah I see this works if you have the tarball extracted with in toolslib +Youre right but I didnt expect to find the classes there and I suppose other people would have the same problem +Added a zip dist to AS config to AS +I would say there is not need for thisOn the queue is a static variable statically initialized thus is always not nullI suggest closing this issue as invalid +Not an issue +Changing from bug to new feature +What do you mean here +I guessed at what to say about authentication and authorization so that should be checked too +Just taking a peek at the source now +I tested with a different sample and it seems this should work too +Will try on ASAP. +I will move this to minor and I dont think its a bug either but good that you found itll commit this soon Thanks +Marking as close since the fix was verified by the reporter Marek. +This includes filters load on startup servlets and context listeners +Adding a few versions it should be fixed for its a nasty bug +Did you guys get it +I have sent an email to the developers on this issue of license hope to get a reply soon +it will take hours +I saw that you merged this into AS on the th Sep and I have fresh trunk and the problem still happens +I wont add it in the domain model +Fixed at the same time as! +Havent tried with the latest codebase +It is really unnatural that the domain names of the Apache Directory Project and mailing lists differ and also will confuse our users +Thanks Alexei! Everything is fine +Just to be a bit more boring is that OK youre checking for a annotation for static fields but the spec is talking about static final fields? +Logged In YES useridThe invokers should not have to deal with class loadingknowledge as this ties them to features that are outside ofthe scope of handling the transport +I am not sure about this argument +on patchNice job +Hopefully it will +Apply the patch for before applying patches for this task +This makes it easy to understand which members are results of writing vs results of reading +A patch? What for +Thanks +That by itself though does not address this issue given that clients are assuming is the like escape character +Hi Wolfgangthanks for your supportYour mirror shows a test pagePlease make sure your mirror works properlybefore asking to be listed as an official mirrorRegardsHenk Wolfgangok thanks looks good now your mirror willbe added to the list in an hour or soRegardsHenk Penning +It seems specifically related to durable subs +BTW I tried the dfsadmin command upgradeProgress during that time reporting progress +Yes It is a interesting Roadmap however it is required to my mind that plugins PMD for continuous integration server allow to manage violation in the same use Hudson and you can choose to manager high low or all priority levels +Committed the docsv patch to the docs trunk as revision . +classpath files and +Well end up removing it when we require changes for transactions but in the meantime it provides a backup in case of DBCP problems +I created a test casefor this bug regardless just so in the future I doesnt pop up eitherItll be committed some time over the next day or few +Move open issues to CR +patch for the file of the Resume reviewed and applied +GMT actually doesnt observe DST also the default time zone in Java is GMT so using UTC would require many additional changes +Committed revision . +Thanks JamesI get a couple of unit tests failures +Reopening to discuss the validity of my change +Here is a simple fix for branch +How about we change scheme and authority to lower cases in the Key constructor? That should work +Sorry about that +Implementation is ready waiting for your comments before commit +prj files +Both and properties were changed to +Here is a patch to I wrote to fix the Im not sure but the problem may also occur in plugins but I havent hit one yet +This then influenced the other test depending on unit test orderI refactored the code so it can be tested without changing system properties +Also changed the name to +looks like to me +I didnt have time to figure out how to get iBatis to log the sqlThanks +I have committed this to +this looks serious enoughIve ran all the tests with DBCP and everything passesfixed in revision +Make sendMessage blocking +The wording can be changed a bit +I can look in to the tapestry side for repositories its been a while since I checked but its possible there is a better place for defining the repos +Patch applied at r thanks for your contribution please verify +Logged In YES useridThis has been fixed for. +Can we include this for? Solrs replication depends on this featureOK Ill commit backport. +Regression test is underway +The time difference between AM finishing and NM starting can be as low as sec +If this was enforced at the CF level we would require a different CF for each index +Im already at run and itis still the same performance as at the beginning while before it was sloweddown massively at +Really nice patch RichWish I had done it myself +It seems like this happens for me when web requests come in while my app is deploying +Itd be helpful if you could attach the xmlxsd files and possibly the java file if you think its related +Lets revisit this in if only to verify whether its still a problem and decide if we can do something about it or not +Im really not sure which way to go +Same for me attached youll find my pom files +So yes documentation for string features is certainly a good idea but we need to mention the document text as wellyes thanksFollowing through with this would it be a good idea to offer our users methods to a check strings for characters that we cant currently serialize to XML and b to replace bad characters +Please confirm the patch was applied as expected after the change becomeslive +OK +changes looks good except for one minor error in doc +I fixed to include this jar +Id assign it to Lucas but I cant find him as a JIRA user + With the patch applied the time to transiently create nodes with properties each drops from ms to msexcellent! for the the patch in revision +Patch for +Can you please coordinate +Linking to since discovered there and relying on that schema +thanks for reporting this +with pleasure +Arun I meant +Add some javadoc too +do you really find it easy to run tests when you have to commentuncomment this all the timeIs it that slow to import the project per test +Tests passed cleanly for me on +Hey Arup can you post the code to github or something else thats public +K bytes taken in namenodeWith HAR HDFS block size of M All the M blocks will be written to block in a HAR part file K taken in namenodeAs seen the amount of memory consumption goes down considerablySo in this use case if fixed performance overhead is acceptable to application then HAR is good choice for LONG RUNNING JobsHowever for files M HAR does not have siginificant memory savings +I was outside and thats why the error came up +yes we should add a doc page as we have for hivesqoop action extensionsit would be great if explains how to run distcp between secure clusters between secureunsecure etc +Manhattan distance city block distances L Chebyshev distance Linf +Hope to see more contributions from youbq +Because then you get into hairy cyclical situations where you cant read the manifest until you replay the commitlog but replaying the commitlog requires writing new sstables and thus knowing the +Looks like attached patches do not kill the system +Updated the patch so that it appies to current trunkPlease review and provide your comments +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I dont think thats Brew +Derby cant tell the difference between an application crash and a machine crash +But maybe if someone wants that they could add a filter to decompose? It would be impossible to do composition to get back the original URL however so Im leaning toward adding URL tokenization +Doesnt a lot more than this fail in Java like Override on implemented interface methods? or are you compiling with and running on ? If so is this really the only incompatible difference at runtime? Hmm maybe soI do echo Teds comment that switching to Java is a good idea but understand that is sometimes easier said than done +Fixed +the file name in has the schemaWhen we create a new file we should remove the schema +Thanks Anil I think it is fine now +You can either manually backport it to SMX or upgrade to SMX +patch for it +The test changes are trivial and good +Mike HariI have attached a patch for that addresses just Haris original metricsI attached a separate patch for on that JIRA issueThe patch contains logic that dumps all the metrics available from as well as the start and stop times registered in the class +add snapshot version to fixin +Thanks for the review Vinay and Nicholas! Ive committed this to trunk and +Actually configurationsources works for both though to use separate main test sources you have to configure the goals in separate executions +for me i already had a solution using older release +Since thats the reason for maintaining the indicesit doesnt feel like a representative benchmarkThat said the benchmark does show that iterating the full keyset has itscostTheres only one other caller so Id prefer to just fix thatCool and I agree that compartmentalized changes are good +May be a bit time consuming but I agree that this needs to be done a little bit at a time +spacesfilecompression snappy spacesfilecompression snappy +m not sure that is advisable +Also WAL reading and writing is spread all over and presumption is its encapsulate inside a wal subpackage Id suggest +we decided that the DAS should insist that datagraph root is passed to the command +If an exception well come up out of the split method but the parent is not reopened +What logs are you looking for? or is there one Im not aware ofThanks Solr logs defaults to std out I think so probably just tried a bigger sanity test than our unit tests do by hand indexing M docs from a CSV file +Unfortunately this wont make the GA release so itll have to be pushed to maintenance +It would be interesting to see if the caching added in helps here +backported fix from trunk to branch +This means it is not possible to query the collection in the component +OK seens good to profile was generated by hornetq script for JBOSS AS mate +items +The handles null whereas the same method in does not +Resolving this issue +Applied in de. +Snjezana could you add some details so we can propetly verifiy this issue? Thanks you open WindowPreferences Tools Runtimes you will see that the ExportImport buttons dont exist +Furthermore I like to thank Dogacan for helping me on the way +testing +This test now seems to produce the correct bug marked as fixed +If you add extra conditionals there then you can probably point apachethriftstdcxx at boostfunction for msvc . +Cheers +ok I believe I know how to fix that +Patches attached +Attached screenshot of the test app in the wrong state +was also converted with this issue. +Verified it passes on latest trunk with the patch appliedAdded a one millisecond sleep in the test to make sure the assert doesnt fail because renewDate is not the same before and after renewal +would new code be able to read old data written with the above schemaYes I think so +ditto for field descriptors +Just exploring other solutions out loud +Will try to get a patch out for this today +this may be a false alarm the test I uploaded is wrong but I am going to check to see if I can reproduce in the project where I am having the problem. +Please review this file +Hong javadoc for chooseTarget contains an unused parameter I wiill fix +Previous patch the unit test does not compile +it looks like the Strategy here puts each replica in another DCId rather extend to allow configuring the number of +Thats certainly useful but its a little different than the view that statspack provides which does the summary for you +Merged fix to and committed revision . +committed +I have yet to figure out what to do with which is using the editlog data streams as control streams also +Two more afterthoughts +Brice to the rescue! Ive merged your change into and it will be in a release +Ive attached a little patch in the other issue which allows me to load latest French dictionaries of +The patch to fix this removed the old imageIO stuff for a good reason and we wont reimplement it for the same reason see for further detailsCan you be more specific what your problem exactly is +If I had followed your changes then it would have resulted in the same because you changed the default constructor +fix ported to branches rev + +Mike are you unsure if you want this approach included in Teiid? I agree with the issue discription that other approaches for the multivalued attributes would be better such as additional tables with an explicit join or a userdn table that is already joined with its parent select value groupdn from userdn where value like abc +Release Notes marked Not Required as this issue relates to Red Hat internal testing +Then is it necessary to maintain the original implementation? I may take it as a burden +Are you saying you think it should be something else +r +apologies if you didnt want it reopened. +Tests still need asserts and some of the QS need their messages put into JMX but they are now running whatever is executed by the runtest target +Rob what does AS client jar updates have to have with EAP +With the latest patches there is no longer a memory leak with the sample have been running for min see attached screenshot +Im wondering if it is a good idea to select JAVACLASSNAME as REMARKS +I guess closing all issues that have been in resolved state for months or more without any feedback or update +Please ignore the in each connector in the patch Ive just submitted and just keep the andThe utf properties are source to the nativeascii +We dont have feeds +Kevin no technically UTC! GMT GMT transitions to BST at start of summer time where as UTC doesnt +The derived objects are not in qpidcppsrcwindowsresources +My Jini for Maven uses them to launch the services through goals and this also allows users to not have to download the starter kit separately and configure it before using Maven and the Jini would be very helpful if these were uploaded +Solr should take care of this stuff itself +Fix including in rev +its a pity that it was not included in the fresh +On the other hand it shows that felix already handles the writing of config files +I just rechecked the log files and compared them when using both versions of ognl +I have committed this. +. +patch looks good +Sorry It was my mistake when I was posting a bug +Im evaluating Thrift for inclusion in a new product and much prefer the version with your patches +Attached a patch that splist the class into separate REST handlers for each URI pattern +Removed assignment on me because I will be out of project for weeks +Why are you calling this a stream handler in various places when it is in fact an interceptor? Could you please change everything to just be interceptor +the normally catches exceptions and returns normally so a normal query exception wouldnt have +However we already have tests to test if access modes are correctly set in the block tokens +v rebased for +Is there any related taskalg in contribbenchmark or somewhere in JIRA +I installed the requisite packages to build the ssl support verified the problem and fixed it +Minor diff from v to make the patch go in +i tested with both and serverHowever I am still not able to deploy my mdb jar file which I suspect is a WTP prob +Relocated transitive dependencies can be either missing or wrong during tree resolution +Attaching and a revised +Its a public wiki go ahead and update it. +This jira should be just about committed +Similar to I wonder if this should have a home in maybe +Will also try JVM upgrade to see if that helps +Set Fix version to as this is a fix we want to get in for release +Thats not a call I would have made but they didnt ask. +Patch for v +passing in the name of the process as a prefix for the lookup +Feng When you dont explicitly show me what you did and then something goes wrong its hard for me to guess the problem +packaging type orbit should be considered as jar for all dependencies +bulk defer minor issues to +Devinis this still a problem? this issue +Id love to see this in final! +Please find patch good +Lack of reply from reporter so closing. +Attaching a patch to update the website +If it was going to be it should have taken a String as an argument +I did the change on varcatalina while the server was down and it workedIs there others attributes wich cannot be changed while the connector is running? +Dont we want to support fragment names Archive and archive as different names? Currently No +Awesome work! Nice tests too +So lets close this ticket +This didnt get addressed in the milestones and thats understandable +But I guess if they can use it for this purpose then other people can tooAt any rate well keep this issue around since it represents a real need +for this for the future +i didnt mean to resolve it +I have also removed the rsetbeforequit being set in the mail configuration +regards aki +we cannot just bump the settings as this may well have a detrimental affect on a number of services +And so the project name should be removed from the body of the email +Note the exception in +Do you think thats betterAlso what do you mean by because the rest of the transaction related settings are on the service definition it would be nice if this went there as well? That we should move the to the services xml? I dont think that would work because this is a setting on whether to wrap multiple calls to a service in a request in a big transaction or not +I just realized this is a build so I assume youre doing some nightly buildsregressions +Thanks +Dag your obesvation about the nature of the parser sound correctI am spending some time going through the grammar rules of versus rules implemented by Derby to see where the discrepency happens +Now the question why I did not apply the file system patch +These are based on feedback from Vivek +Done for name change +This build will be w the removed flush above +Thanks for the suggestionI spent some time looking at this originally and dont recall now why I didntcouldnt implement it this way +I applied the volatile variationCommitted revision . +thanks for your super express supportcheersCharles. +Can you try to hunt that down and perhaps clean up the patch there? Last time I checked it solved many issues but broke a number of tests too +Will consider this and either modify or close for M +It is related to the aegis binding component not recognising a maps values type correctly +Which case? Dependent fields or producer fieldmethod +was a major refactoring in policy after so Sandesha is not compatible with Axis current codebase +It is expected that this will be the solution for and +I was running it on Java +Yip For CREATE TRIGGER is this statement accurateYou can create a trigger in any schema where you are the schema owner + +Reopened for backport to fixed in +I used that one to avoid disrupting hudson activities on other s enough to confirm my belief that this is just a test issue so Im pushing it to. +Updated patch turns off compression by default and adds the clob property back +thanks Robert. +This is only worth doing IMO if doesnt go GA +But if we do output subclusters we add only the documents assigned specifically to the cluster because the subclusters with their documents will be included in the response tooS +Any update +Bulk close for release +I do not know if spec is incorrect or the runtimeP +I dont think the QA here is right +In this case however the intent is to interoperate the scope of the code affected is contained and a reasonable attempt was made to get a express license +The APPID of a couple of those should be either CONTAINERID OR APPATTEMPTID +HelloIm having the exact same problem +OK rev +Fixed in revision This is an important one for Asian users +Got started on the feature in rev However for some odd reason the in does not register the component and thus it failsI cant see any errors in the feature file +verified in JBDS +hellobut when i encouter this problemthere is no jvm memory errorand there was no stdoutstderr files in the folder +Scott this is correct +Also removes one unnecessary minSegment null assignment +It would be better to print bookies in different two sections one is rw bookies while the other one is ro bookies which make the output clear rather than mixing them together +If thefields are values it will emit tuples with values that do not match +could you please provide a working mapping that it fails on +Nope hes always said to me +Arun this is then a catch +This will affect both Grunt and all unit testsIn fact this kind of breaking changes are supposed to be highlighted in the changelog but I dont even see itRegarding the tests I dont have a windows machine to test +Sending textetcTransmitting file dataCommitted revision Revert previous commit on that file + to failing to startup in this serious misconfiguration +Committed +Problem is now resolved in a new branch. +Updated patch +Can you please post more details about the packaging of the application? Or maybe attach the application to reproduce this +please give me hr to look at something +But to produce a truly pristine source tarball you should make sure to keep the source directory itself pristine +Initial patch for Json related changes +But the source plugin overwrites my normal sources jar because it dosnt support classifiers +Closing resolved issues for already released versions +answering my own question going to long long so we can write GB on bit machines +Here is the patch for +This has been fixed by Raymond in r by annotating the component implementation allow pass by reference +So perfect timing I will discuss this with richard +I think users will adapt to the different version numbers for the model vs +Hi all sorry for the break +. +Im still working on this +I cant tell if this new functionality is ok with that +Please review and close. +Ill be posting a fix shortly that Ive been working on +Patch v limited the number of retries +Hi RobertWe would love it if you decide to own it is all yours +NickThanks for getting back to me so quickly +Please take a look at patch I have posted +I pushed new logo on Behalf of Utkarsh +ThanksMirror Added +Uh +Though on some systems with there is no making it impossible to have a include in +i just finished merging in the IT +Also I keep a cache file around to only instrument files that are changed as instrumentation isnt was able to modify wicket classes on the fly with this change +Current thinking is that we may expose HFI through API methods on +The attached gif file works +This issue doesnt make sense Ill create separate ones for different AOP todos. +fixed in revision +Patch committed as is. +It seems sane to me to keep the property that client library shouldnt have to parse queries +I did actually do this a few days ago is it working as expected +The patch now catches such invalid options so I updated the depending test cases as well +This bug was fixed in Xalan Java. +Rebased v after issue also affects +Actually the JBPAPP branch is frozen for another week or so +This particular issue is fixed in r +JIRA is not the best place for such discussion should we move it to dev list instead? Simon +I dont know how else to print a stack trace and I dont think we can afford to give these up from a debugging standpoint +Youll the find inside of javacommon javaclient and javabroker patch files which should be applied from their respective directories +Ruby can use that as well so were good + has forced this cleanup since the Server injection stuff does not work with Jetty anymore weve cleaned things up +This story needs to be defined more +This is fixed in the JBPAPP branch +The latest code doesnt have this problem anymore. +Tests and docs included +the type on seems redundant in one sense we can look that up from the engine but I guess it kinda belongs in themodel +slip + +Indeed I couldnt reproduce it because I fixed it earlier Could you please commit your changes since I see this has been fixed by you too +This seems kind of pointless for such a simple issue but I will do it for practicePI moved the described code to a new method in class called initialize which returns the configuration if it passes the sanity checksCheers have no objections +This kind of bad feedback gets only worse as the cluster size increases. +What is the concern having them static +An undeploy plugin maybe +Marking closed. +We dont hold a hard reference to the Thread itself but we hold hard references to the others references which arent getting cleaned think the simplest way we might fix this would be to wire in a smarter bit of weakref logic that periodically cleans up references AND removes the associations that exist +committed! +Can watchers of this issue please switch to watching the new one? If no one has objections I will close it tomorrow +This is a new problem and requires investigation +Having a single procedure allows us to have simpler procedure names but I agree that its not so clear with the three parametersAs to the naming of the different procedures could we pick one of them that the users are more likely to run and give a simpler name? For instance SYSCSUPDATESTATISTICS update one named index SYSCSUPDATESTATISTICSALL update all indexes on a tableOne more thing about procedure naming +And the DLC should also have this exception handled check in its processing Reminder to me +Thanks. +Thanks Owen! +The constructor is removed because it isnt used anywhere in the code + +Well if were going to get into complimenting each other here I can say the same thing And in your case its probably even true +Results of patch +as far as I can tell +Looks good to me +bq +If the dies then the watcher restarts it +This is unusual for HTTP but not forbidden +Heres a strange observation +it load all modello plugin in classpath and the store plugin require the possible solution load in classpath modello plugin only when users add it in plugin configuration move exception that throw in metadata classes to an other formating +Uploading a trivial patch that converts to use junitant test Dtestcase doesnt run any tests now +Install cloudstack agent on each kvm host b +If you have a period of seconds where each idle time never gets to seconds go ahead and commit +Same question have been posted to hadoop user mailing list with no answer +I noticed that changes in the relevant css files were not made +bulk defer of features from to +Seems like the sanctioned way forward is to leave these spurious entries where they are but we should really fix the poms to have a matching artifactId +I think that a more likely scenario is that app client is not doing an undeploy on shutdown +A simple patch for let the trunk compile with an empty m maven guess we didnt see this duplicate earlier +Seems related to where we do resolve on both master and regionserver and the two can disagree +Kim I assume you meant to assign to yourself? this issue to Gunnar + +Marius could you please provide a or even a +yes JAVAHOME is set +Thoughts +Approved for. +In which case it is trying match the null xhtml template name against the predefined pattern to see if a pages should be parsed by the particular parser and hence the NPE. +new patch as described before +Thanks I just built maven from the latest source and tested it out on a proper commons components looks good +This patch implements the fieldext approach +NOWAs told Ill let the test running tonight + +Thanks! +I am not sure if I understand this correctly +Fixed as part of. +To follow validation of metadata checksum +bintrafficmanager version it gives me version +Managed to repro but ONLY on the iOS Simulator and not on the device +Ill run test with larger numbers and update the resultThanksShameera +Its fixable but not trivial +I just committed this +Reverted changes to ctor introduced via +for the latest patchbq +Will use system default temp dir instead of hardcoding path +An for has been added to the Cluster project +Committing is easy but you will have to wait until the nightly build is doneI will also include the version bug fix if you can wait more minutes +has been marked as a duplicate of this bug +get getObject and get removed from the Resource interface Added and microsling to the modified API + +I am not sure why a change from must go into trunk or +Ive sent a mail to them on this +Just because it has no similarities for an item doesnt mean it doesnt existIt could add a check purely to implement this behavior but I think thats overkillinstead I think this is best solved with documentation +Commits and patch are attached to on. +Checked with nightly build and still have same problem +Done. +I think that at some point the UI plugin must return a null value instead of an empty array in a viewerAssigning to Xavier for investigation +thnx has tested it on his project +Caching can be separate issue. +In fact we already found issues in the QP outputstream +Partially resolved +Gaawwwww! I never thought Id fall into that trapThanks Aaron + Fixing Fix the test only +Related but not exactly +Because when i manually overlay the same AIR SDK the issue is not seen. +patch looks goodI have committed this +patch adds another testcase to for this particular bugthis new test patch is ready for review +Or can I +For JON download is the name that will be used +Would like to cut an today +In this case a table would be owned by a specific user +Running run Failures Errors Skipped Time elapsed resolved issues for already released versions +and Francois are coming up with the email notifications and using Eloqua to send out the messages +cppsrcqpidagentSeeing as this is the only place that is used why the change in interface and in fact whats the external distinction all about +However this was such unexpected behavior for an update that it definitely seems to be a bug +Note also that doesnt work for RTF output in +Tested this patch with relative path and absolute path for and + LGTM +Please go ahead and close. +yep +No issues found running this patch on a secure cluster using and +Yeah I misunderstood that piece +Ill chase this up tomorrow and then we can work out the best way to reconcile these versions +I modified the cleanup code to call it hopefully someone will discover stray files if the clean method failsPatch attached +Could you please clarify where and why this change is required? I dont recall trunk using any REST +Trygve can you verify this is all good now +Modified Eclipse +Having played with my patched version for a while now Ive discovered another bug +bytecode manipulation turned out to be too tricky because theres no standard jee way of instrumenting the application classloader + +Attached the request for change to be submitted to OGC +Closing out and retroactively assigning to Prasad +This issue has been resolved for one month or more without complaints +This is better than HTTP BASIC authentication since in the case of HTTP DIGEST the secret key is never sent over to the server in the clear and it allows for mutual authentication +has decided to make dependency on hadoop optional +I am about to integrate trunk patch v should I proceed +Add threads to receive client callbacks +should we add a to remove the potential for that code getting out of sync w the register +Thanks Colin +Thanks a lot Mingjie + +I dont really have an object to this improvement either but I agree with Graig we should be careful with adding enhancements to the Portal Driver beyond its purpose of being a testbed only for the containerAs it seems this is only neededasked for the version of Pluto Im removing the fix target +Update to match TRUNK +So it is not ideal to update the state on each node all the timebq +In doing so I am first trying to wrap my head around the documentationOne thing that jumps out at me early on is this in order to pilot this tool I need to have user manuals open all at once The Tools guide +The issue is that the compiled mode since it emits bytecode gets caught by the JVM trying to violate the access rules the interpreted mode is able to simply ignore and bypass those rules +My changes are attached to this record +Awesome I will give it a spin +Just one more bit of info please for the release notes Was the actual fix here to add the? +Takes s without this patch and s with this patchA question I had Why do you need to do this skipping via a coprocessor? Would a filter not work +I need to log a JIRA for it and then check my code in +this issue is resloved in revision +I will commit this to trunk and branch +Heres a first attempt +By default maven will create javadocs during a release but the recent jboss parent poms turn this off. +Logged In YES useridThis appears to be a duplicate of ID which has more details +This is a broad enough interface +Did that happen? if extending the old doesnt work provide a complete implemetation for the former functionality +Thanks DarylThis is fixed with with revision +provide GAV +My guess is is not relevant +Applied the patch in Rev +Attached first draft of functional spec based on wiki page +Instead of returning null Integer we could return given callId always starts with +Integration tests now pass again this is resolved +Its to make it easy to configure endpoints! It has been link this years +Verified in Teiid Designer +Tested it on my local machine it work +Attached patch +Now at least you get a nice short error message from ofBiz that you are working with an invalid +Fixed. +A significant gain was also observed when using an EBS instead of an array of in the policyIm surprised by how much of a gain occurs so Ill have to experiment further to understand if its factual +In this way the access information is kept close to the code where you could use enums and constants to ensure type safetyThe general strategy of Struts is avoid XML configuration so weve opted to not use to store a lot of configuration information that existing in Struts +Ivan Chuan and Steve can you guys also please review this patch for the parts you had reviewed in +Still dupe code +However due to design of the code at present message that are delivered while the connection has not been started will be dropped and not delivered even when the connection is started +To be precise I want a reference where by I could configure a broker destinations connection factory by spring beans using s XML inside it +comment put in wrong to mark with correct Hoss for this quick response +But the performance was going up by maybe in my test only with a very big range of times so that this might be below the border an effect gets really measurable +Yes very few of those accounts actually still needs sudo +Id appreciate it if someone would give tomorrows snapshots a try to see if that works +Also I added support forAlso I added a dependency for apache for the implementation +I assume that git would have a similar configuration option but Ive not used git before now +is there any estimation for a or version which fixes this? +Build issue and deployment issueSteps to produce the bug First add the dependency to file as mentioned above mvn grails works well +My implementation uses the JTS directly +Verified by Sian. +These are dependent upon the yoko ORB being in the me repository +Interesting +Max any chance you can provide a test pom or better an integration test +I think the patch is ok to be committed to svn +HiI have attached an Eclipse project of a composite diagram generator prototype using Apache Batik +Then it can invoke rollback method of the participant if coordinator has returned . +Thank you for you information +During config we dont add dynamic methods to abstract parent classes and also dont map them at the Hibernate level +For setting up in classpath how should we achieve this? as ours is a stand alone application and using XML mapping file to unmarshall the elements into objects +Yeah this patch is subset of +Should appear in the next build +Changing the order of those two beans starts the AS instance perfectly and the test suite has no issues +bq +I think we should wait for release and then add this to trunkThanks completed authenticator framework +We were unable to reproduce this with current versions +The later is also now deprecated and is planned to be removed in versionPHP settings page has also been updated and now shows option +Now I am running all the tests with your fix and going to close jiras as soon as I verify that they are fixedShoot +lets fix this thing +Hopefully that should fix the problem with +a standalone example that show the problem whoud be helpfull +INFO nntpserver Connection closed INFO nntpserver Connection closed INFO nntpserver Connection closed INFO nntpserver Connection closedif thats any one last peice of information +Aaron do you think youll be getting back to this? If not Im happy to find someone to work on it +can someone just add a comment to mail some list if changing the file in future? A little warning would go a longway. +Id greatly prefer one well contained side effect instead of uncontrollable proliferation of duplicated low level code +The latest patch includes the removal of the configuration stuff thedocs for that and the javadoc clean up of the NTLM classes +Filed for this +Yes This is old bug +ve committed this +Using the CAMELCASETOLOWERCASEWITHUNDERSCORES naming strategy everything works fine but not with mine +The proposed implementation will not offer the user to compile individual Groovy classesscripts +Please reopen if needed. +Ill take a look at this after the release is done +Hi RandySorry got your email lost in the email jungle +Finally got a unit test working with loading via property files +Bulk closing stale resolved issues +comitted rev +deferred to with the very excellenthelpful Phillip Thurman on this +It seems that the top speed is ks which happens at the very beginning and it can go down to ksFor those wondering the dips are caused because we max out our links when the flushes happen all at the same time +hod contrib was removed +I also add serialization tests + +Applied +Changes that work for me +We probably still want to kick it after deletes and cache pool addmoddels thoughActually we could probably pull the same trick and update bytesNeeded and filesAffected before returning from a delete operation +Verified on JBT +But is still there +Should this go into? Seems like problem w the api +From the advisory Im happy isnt affected +But if I list the folder using using I can read the first children and then encounter a null reference +Havent seen this fail in a good while now. +Branch has been merged +Alternatively most JCE api calls allow an extra argument which takes the name of the provider to use this way BC doesnt need to be added at all +Lucas knows he can give the customer the untested patch from this JIRA if they need to use it in their environment. +We tried to minimize the SunOracle JVM lock in but JAAS doesnt give us enough information +Thanks a lot guys +Closed it because of rejection. +Should this go in or ? I figured Id put it in either way +So the number or RPC level threads blows up and you know the rest of the storyAttaching patch +Rebased and cleaned up the patch +Nothis one persists and does seem to be related to the finalizer +From the perspective of the build the code that is seen and the fact that it is unmodified from the repository is equivalent +I guess it also affects +edit binding properties for example name +In lets see if this fixes the test issues +I thought I had updated the master in the patch +Yeah sorry +Im still able to reproduce this problem in branch +If thats no longer required then I suppose they could be in a +I will fix +Ive added the asterisks for required fields in there so hopefully that addresses what you were seeing +thanks looks like the issue is fixed m not with the solution but fixing it in the Codehaus renderer takes a bit of effortI will have a look at doing that as it removes the javascript requirement. +Can you elaborate on what problem you encounter in trying to apply the patch to the source code for +As well already have some API changes to accommodate integrationhence this would be another one. +With the processing of external truetype fonts shouldnt be a problem anymoreSet to resolved +Aleksander Mladen But is it ONLY an issue in upstream code or does it filter into our productIf it is an issue that also exists in our product it should be documented as a known let me know ASAP +Thanks Dimitris +i cant reproduce it in trunk right just happend twice when using it in JBDS +Will leave the Jira open until we verify the fix on hadoop side works +The abuse is clearly there and we should probably try more to address it than just docMy opinion is coming around to leave it for change it to an expert option for that works the same is understated and is called forceMerge or whateverBig to making it a no op +This should work with otherdatabases +here is the samples should now work because the file prefix was added to the resource path for all outbound now Ive changed the wording of the display to read Default inputoutput directories +Implemented the change as described above +But I would think before that can happen all of these threaded bugs need to be worked throughok +The attached patch redirects these four pages to their corresponding history pagesNote that there are no unit tests for JSP pages +committed a fix that calculates a checksum hash over argument types and throws away cached method reference in case the checksums dont still needs some performance testing +Thanks for your comments Vlado +It seems to be a common configuration error +Thx. +The fix is not correctly working for messages that use the PRINTF format +There have been many changes in and trunk since then +I only tried within my own usage patterns and all seemed fine +Merged to +When it should be used? +In the original statement ORDER BY seems clear andunambiguous +Added the patches +Please post you comments if anyThanksSandhya Turaga +This does not appear to be an issue. +thank you very much Nicolat for your investigation I will check it and give you feedback my comment was for the JCR +Exactly so changing the kit to conform to an invalid connector binding is not the right path +In this scenerio wed remove the assembly and create a module called say which would do nothing but depend on the other shims with a scope of runtime +Thanks Ben +See my comment on +Committed to trunk +Hello executeScript doesnt appear to be in +Cos thanks for catching this +Ok great I opened for the specific issue +The filename of previous file would not work with the pattern expected by the pre commit test framework +Roshan I am assigning this issue to you because I think you recently resolved something compareable +Id held this open hoping Id be able to do before closing but since thats not in the cards any time soon and the fundamental task for this JIRA is done Im closing it. +Applied to trunk +open +I propose that we take the target name and have it +Very nice to fold pulsing into the default PF +Here is the updated pom +bq +I put the content of Exception into we can reproduce the issue? and how you get property with null value +Note that should not have any other changes +Thus Ive only modified the JCR XML exports to use the new serializer and kept other XML serializations based on the existing class +I didnt think the DST rendered anything that needed in? +i believe this can be closed since we now allow overriding the default values for constraints +Thanks to Dyre for the workflow reminder today. +Approved for +moved to +No tests +The har is interleaved inside they each instantiate and pick up currently without a way to insertinject parameters or overrides +As Stack mentioned it is not good to overwrite a WAL file +We shouldnt support local anymore only for back compat +I tried running the tests locally and fails for me with and without the patch and the others all pass +In fact there is no regression in version compared to according to option is mandatory in version to be able to import coverage data and when it specified reuseReports mode works exactly as before. +Same goes if gradlew is used +What the customer is asking for is to return internal details in case of an internal server like asking websites to return stacktraces in the public if some error occurs while end users browse their web site +It seems like it would be nice if we printed the original exceptions stack tracesomewhere prior to discarding itOtherwise when this happens all well know is that we ran out of memory but wewill lose some potentially useful information about exactly which line of code wewere on when we ran out of memory +I should open an enhancement for thatThanks for the issueGlen +Typically in an api module which shouldnt be bound with GWTDespite the fact is in the GWT module async interfaces generated from it will lose its generics. +sure! If we dont mind breaking any Script implementaitons that may be out thereBut I know I can make it work this way because I saw the unit tests pass +Same issue as so has been fixed +Fixed in r +No With the database schema will be dropped when the is closed explicitly +See linked issue +DISTINCT is the default if you specify neither ALL nor DISTINCT then duplicates will be eliminatedThe next section in is titled Precedence of UNION INTERSECT and EXCEPT and starts with There are rules of precedence in how many duplicates will be contained +There is also a component +This is not always true +We want to get away from using a repo so Id like to see thrift get into apache repo before we upgrade +Patch updated with tests to verify patch to add one more feature log the exceptions in the test run +if desired I can split out the changes to loadaverage matching into a different patchticket but steffkes said on IRC itd be ok to include +I will look into the issue +Veirfied in CR +Did this happen w a config? Or is it just consequence of splitcompaction refactoring and we should do as Andrew suggests +Thanks Aaron and sorry this process has taken so long! This is a great improvement +Vikram +The point of separating the structure from the code is so that if java script is not available then the page will still render +Also its not obvious what the schema violation in the file is so i wonder if that also needs updating to have an error +Dont worrier I will take care of itAnyway welcome to be a camel riderWillem +I assumed it would be easily reproducible and then the problem would be fix at this point is to eliminate nested trees but this is not ideal +correct the components the location of from tcpserversimpletcpserver to transporttcpreceiver solved the in revision +As I mentioned the downside is divergence from upstream +want this for +Based on the discussion I dont see a reason to spill to DFS +Indexer was changed to handle path handling changes for local files within Hadoop +I will check in my changes today +This was resolved actually in ER with the HQ tag provided +I think we can this once weve go our code cleaned up and primary JMX CLI features intact +I have finally figured it out +Now it is possible setting many values for a key but still getting the last set value using get or getProperty methodsSince meta data can have now multiple values for a key it was necessary to change the content serialization methods as well +Of course I did not test all existing versions of AR but I think that this example is satisfactionaryTo comment more versions you did test we are not using it and so I did not test it CE regional czech version produces the error +autocomplete in current directory +I commented out the Product relation in and that got me past that errorWhen I tried to update the party relation the service engine attempted to invoke a simple method +added completion assistant and other helpful this actually done now then +The amount of effort necessary to write documentation for this would be pretty much identical to the amount of time to write the logic to handle it +The code must be modified to get the first child from the soap bodyOutput from the console is poorly formatted +For the you can use logproperties and pass in a which has a hardcoded DEBUG level for the root logger +Ive updated the syntax as suggested in my last comment updated the unit tests accordingly and updated the documentation. +Adds and by appending to CP in binjruby +Maybe we should create separate methods for thoseI also changed an exception message in from No such accessible constructor on objecttoNo such accessible constructor on classbecause I think that this is more convenientRegardsBenedikt +jstack for the hanging am removing this from +So all issues not directly assigned to an AS release are being closed +I have copied the code related to the caching downloaded files from the project examples plugin and removed the project examples plugin as a dependency +There is a test verification problem in the previous that does not create a single split correctly for sorted rows verification +Sending docsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsAdding docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsprojectsSending docsAdding xdocsprojectsSending xdocsstylesheetsTransmitting file dataCommitted revision I believe it will autodeploy +This should hopefully resolve everyones issues if it does not open a new bug please. +The format of your new code looks fineThanks JeanneThis is the patch with the changes you required +to include in +Could you please improve the DT documentation just a little for and trunk +Closing this JIRA for to show that was not included in EAP. +Resolving as Cannot Reproduce for now +I just want to confirm that this really helped to me particular adding the explicit dependency made the must be some magic because the used by itself depends on but I didnt bother to explore it deeper. +Its odd to create a Person without firstName and lastName +for the rm patch from. +It will be great if you can directly commit files but for this we have to gain some momentum +The patch is a definite improvement over the current situation +Added a featureContainer extension point that allows the runtime plugin more control over the creation of +I guess the problem is that Avery got my but didnt commit yet +If you are running an internal maven repository it is a web server and can be used in this process but you must deploy to thatserver before calling jboss is the fully qualified name to the web server where you deployed your install builds the artifactmvn deploy pushes the artifact to the repository in your distributionManagement section of the jbossdeploy w an http url jboss tries to download the artifact from the location where you pushed the artifact w mvn deploy +We should probably approach this by having a lists of explicitly installed plugins +How can it be disabled from server side? What configuration file can be changed in or to do that +I have to change the include in a few files for itto compile successfully +I have a solution and I am committing it now +According to JSR it is fine for a repository implementation to defer node type checks to save +Ive also changed the implementation so that an EOF in the middle of a transaction gets propogated as an error +This is especially true now that jbpmflow are SOA Tooling components which are built downstream from the Core components of which Seam is cant speak to the RUNTIMES however I just know the dependency order for building the various tooling components +Its not working when you have Link with Editor button enabled +This should display the error at the proper time and in teh proper location +I will add them to the jdbc test suite +defer all X issues to +The interesting thing about this issue is that is has switched VM in +This is a blocker for our upgrade to SI we are using Spring Batch Admin to display information about our job executions and so on +This brings up a good point which is that we dont cache ZK results anywhere +Max the Common component or +Its due to db upgrade issue +you! +Patch seem to work but Im new to cassandra codebase so there may be hidden problems. +am sorry can somebody point me to the Axis site? are from? Dont we have an axis site under? I guess I found itRuwan +As requested. +Hope can be ignited by a spark of encouragement. +An improvement is added at rev in Qpid trunk to avoid creating new strings when doing comparisons. +NULL is very tricky as in the SQL world +This patch adds PEM support to the Factory and corresponding test case to its unit testThanksAlexander +In this code you can see how I implemented my own +I dont understand why we need the chmod what are the files created as by default +the nil problems were the result of File not existing +bq +If Im missing something obvious please reopen. +All DEBUG messages from go to as expected +Alexey should I merge it to +Sorry to be a pain but that is the often issues turn out to be already reported in JIRA or already fixed in later versions so we dont need to add a JIRA reportWe need to go through the triage process to ascertain whether we need to add a new JIRA report +Absent a patch moving to think this issue is in fact part of issue which was recently solvedThe new implementation of still ask for a starting interval that does bracket the root but it may then slip out of this initial interval as per traditional secant solvers doCan we consider this issue is solved. +I found a quick fix that will output of hits +I agree refactoring the code to a plugin is a better solution +Kind regardsStefan + passes +Documentation on the logical replication scheme added to Apache and changed source are you planning to submit an ICLA to ApacheIf not there is the significant chance that this patch cannot be accepted into the Derby codeline +What if we had readrepairchance and readrepairdatacenters? Not quite as flexible but it covers the known use cases without being confusingCan you be more specific on what you mean by that? I dont think youre just proposing a naming change right +Fixed at revision as far as I can tell. +Fixed in Trinidad in revision +I debugged through the problems yonik is agnostic of the request params +This way we give people a three month window in which it will work and also allow them to update after that +User chooses a and the corresponding algo is executedI have tested the scenarios failure and success scenarios from my end +Done fields are protected now. +Tried to grab the bundle but got this dont have permission to access on this server +I dont understand what inconsistency you refer toThe semantics of hbernate and annotations are different in places deliberatly because they annotations is about configuration by exception where is less you are referring to the annotation exporter doing something inconsistent then that is something to lok into +Im going to revert it +Hmm +In fact some users requested an alternative way of sharing variables +Which branch are you usingThanksHoward +Thanks Kan +The same change should be performed in SMX +Patch applied at . + +I have committed the patch for it in facelets project +Explicitly specifying XNTHREADS can solve the issue +I replace my previous patch with this much larger patchsome of it is autogen but I reviewed Is there anything else in contrib to convert? If no I would close this issue for nowIf we find further Java conversions after release of we can open new issues +Thanks for the patchIt also works for version just the line numbers differ attaching the adopted patch for just in case +How to replicate the bug Create a java project JP Create a folder inside outputs Create a project archives for project JP where the destination is outputs Create a fileset under the root archive matching everything Ensure no copy error dialogs appear and no OOM errors appear +Can you please verify +Seems like its delivered in this featureeclipsefeatures some compilation problems caused by API changes +Was given the to include Hibernate into so these changes are included in that release. +This patch has a problem as we removed from the parent pom therefore the needs to be rewritten to use hectors embeddedserverhelper plus this dependency needs to be written into the parent and s This is not really a headache as the class is very well written an very easy to use but I wanted to get opinions on what is best to do +I have moved all my lists to label value pairs and with that switch to the optionscollection +Pushed to master +Thanks Panos +Thanks +Looks good to me robert want me to take the issue? thats great if you would like to only thing is i would like to try to do contrib ones tooif you dont have the time to do this part also ill get around to it eventually but in general i think its best to try to keep all the qps in sync +for your few notes I dont grasp the difference between replacement being binary code or just exact jar combination +AS has no true notion of stopping a deployment +Added JMECDCFoundation to the list of interfaces in the WORKAROUND section of the release note +what benefit do you get out of being lazy by default? Do you actually have so many beans that may or may not be used in a given runtime +It seems to have been fixed + +This could be a parser bug so I will start there and assign it to me +We should do that in Beta releases once it is clear what is strategy for resource packaging in +It is widget issue under moving to be good to fix this upstream in Eclipse Luna. +Increased ulimit it passes now +Marking it Blocker +I dont recall this having been a major issue in the last year so closing this issue as a WONTFIX +stack +I think the component in Tika lost some of its accuracy when it got moved from Nutch to Tika but I think it makes most sense to build on top of that and improve the one in Tika instead of having something special in Nutch +Thanks Nicholas +Im sure it cannot be reproduced with this annotation never relies on properties with matching names because you do not have Id properties in your entity +First the MTA that you are using isscanning spam on both outgoing and incoming email and blowing away theheader fields that would tell us what the Apache MTA thinks +Uploaded the updated patch with a and some minor error handling added +Key idea is to not associate a container with a single process launch and exit +Never mind re Jenkins I forgot that the patch wasPlease run the HDFS test suite or just the HA tests before committing this +I am currently using and text extractionsearch seem to work fine +Jack your JIRA account is not a developer account so you cant close this issueYour comments are sufficient Im closing it. +There is an overflow in set +Now valid patch is +Need to change sort order of limit operator the same as the sort operator +There are likely are examples where having and or or wrongly treated as a operator is catastrophic therefore the onus should be on the client to choose the correct lowercaseOperator option for their +Looped times on Linux and all passedThere was no hanging test +The real problem is somewhere else and a correct bug report will follow +No fancy associative arrays that are looped through to create the appropriate HDFS directories and the permissions like the puppet code does +This is probably the same issue as +After I receive their response Ill share their comments hereThanks +Some small changes in should do the rest after is done +Ive tried it and it works and it is a realy easy way to plaint my Proxies +to the trunk and the branch. +I renamed the last of the examples core namespaces according to +There is not regression +Im just looking forward to using this +We look like a real ruby library +Yeah tried a factory reset and update already to no avail +though there are docs and tests to consider +add patch for UT and all integration tests pass +Comparison between the dependencies +Vinod the protocol I was following was to close the as soon as they are in trunk and to add them to for backporting +Ill go ahead and close it then +I tried this didnt help +It does look like the force call should be removed to make the test do what was intended +This can become useful at some later time +Maybe just defaultRenderer? Do we plan to have any other renderers apart from object rendererdsLooking at the naming I would prefer that the renderer interface be called instead of +I also think that this error occurs on x code base +Committed as r +Only +OtisYou can query across multiple collections with the collection parameterhttplocalhostsolrcollectionselectcollectioncollectionNYcollectionNJcollectionCTYou can also setup an alias for multiple collections and then search the aliasThis is a nice approach because you can simply change the alias to point to a different set of collections and it will be transparent to youre application +Thanks Jarcec! Patch committed. +I was considering this change for backport to but as it changes the message file for English only I prefer to because of the completed translations for so I only added the label +Here is a new patch +Will this be fixed? +Fixed in head revision +Koji could you point me to how to use that thanks +Verified on +I have not delved into the issues with the previous attempt but started out anew +This was a silly bug that occurs when you do not define a field collapse cache in the +The security change in Http server will be addressed in a different JIRA +OK guys thanx a lot +While accepting connections from other quorum servers I feel we should validate the client quorum server against configured quorum servers +Fixing Affects and Fix in Version +I dont expect any problems there from the last increment to the patchBefore checking patch available I want to add an abstract superclass to +Updated wiki page as well +Is there some way to patch the geotools +Guys! Youve already done a huge amount of work with ve used Spring from version Preview and now I have a problem with my customer coz he has to placed his site in a shaded hosting and I seem that unfortunately I cant use Spring until you will implement this feature +I just checked on the following platforms Windows Revision Windows Revision Linux Revision Linux Revision VM compiled in release mode +Initial draft of the patch +That being said I would like to join effort in case I get green lights there +ships with a feature patch for JDT that augments the feature with our compiler extensions +Check Release Notes for all the great info for the Release Notes +Let me fix +lgtmNit you can remove the isDeadtrue in line now that you unconditionally set it later +I do not have access to some box with automake +Attached patch for trunk and X thanks for the patch! +for each widget in the tree and will be sent to each widget +Removed remaining master files with revision . +The results are in very simple format so that they could be processed by Excel of R system +This patch also corrects the copyright notices as Jean recommended today +screenshot of the feature +Ant version has also been feels brittle what about all the other dependency versions? +Attachment has been added with description due to lack of further input from transition resolvedwontfix to closedwontfix +This issue draw up an unbelievably long discussion +Sure I could test it +Restart and wipe the cache directory to be sure GWC creates the layer based on this t really think of a permanent fix until GWC can pick up GS configuration changes +Resolving as rejected since it looks like it was an issue with the json gem itself +Anyways we all know that URI class will require quite a bit of attention in thefuture +I like that +LGTM to and trunk. +We can reopen if a problem is found. +This hack only works when destroying the whole container and specifically does not work when releasing a classLoader or single instance. +Checked out the changes and all seem to be present +Fixed on and trunk +The PM is responsible for reading and loading the ones that have onload set to true right away +java X in default mode +I created test +the patch looks good to me +With svn revision youll get the name of the entry that is causing the problem +Hi GaryThis code is a port of some PHP code +Actually this will happen for the initial start maybe for a few seconds +BTW sorry for raising all these supposed shortfalls when the more serious ones have turned out to be invalid +Looks good + +I think we are just closing issues once they are fixed +Make more generic based on the test cases +Do you want clients to have logic that decodes the protobufbq +Patch is now in CVS +Maybe could exist on an +And If you have any good reference of connecting a grails app to hadoop +I agree that replica placement should not throw an exception when there are no registered datanodes +Patch is in rev +Can a committer please commit this if it looks good +After hours of testing Im on patch v +The option you mention was discarded because in HTTP new connections are open for each requestresponse pair +At the moment it looks like it is needed for the which is confusing for a new Cordova developer +Steve I can take this if you have not already started +I will provide a patch and the procedure how to apply +See Ramkrishnas comment Sep +Initial fix for review +This patch fixes the issue by moving the parameters into the body of a POST request +Committed trunk addendum to and trunk +looks good +There is no attachment +if the inside the war contains the path then the war in the deployment should have the name and not +Normal anchor tags cant submit forms so the anchor tag has to use javascript to click a submit button. +Quickstart is meant to be run from within Eclipse with the Start class +I believe its time to close this JIRA +Modified documentation to mention both capacity and are of type floatDidnt test +Something like! for fact I was thinking about an example of metric whose value is a percentage and which doesnt bring any quality information +push out to +Also it is more consistent because that way all your factory objects will look the same and it will be obvious from the type name that it is indeed a factory should avoid to create custom config schemas for simple things and use them only where they significantly simplify configuration +I have spoken with representatives of the Windows Store and the MS interop team and they have confirmed that the branch is properly namedThey do not want to see this branch named however we need to adequately document that these apps are distributed via the windows store. +NET compatible compilers and +When an option is added to a command line its parents are also added +Well I started but the duplication was so wide spread I thought I was not understanding your design +Another workaround would be to implement a solution like we have for EJB which will avoid the usage of remote JNDI +Fixed in the new version of the patch +I fixed the format +Also makes minor changes to Schema and the walkers to address a few oversights +its awesome! +Thanks Arun! Ill put a trunk patch up today +It seems like patch is applied successfully +NicholasHairong Are you folks saying that the approach adopted by this patch is not sufficient and it needs more changes to make it efficient +The issue was solved by upgrade derby toThank you very much +Thank you for the quick replyso Java in my local machine is configured asusrlibjvmbinjava Srikanth KommineniGraduate AssistantDept of Computer ScienceRochester Institute of Technology. +Go ahead and attach it thatll make it easier for us to test +But this bug is reproducible with facelets +Leaving Fix Version as is +patch to fix null pointer in fetcher for filtered actually there were two places which needed fixing I also somewhat simplified the logic Thank you +yes +this seems fixed svn revision +Added an evictObject method to handle eviction properly +I tried this a while back and reverted it I dont remember why +Although for packaging purposes you could have a build dependency on the appropriate devel package and not include the actual filesThe extra work I think this needs is to make the Cmake tooling now work both standalone and as part of the full build Im not sure how good cmake is at that +Although Im not sure this would work given that the lifecycle of the class is managed by +Marking closed. +While NPE occurred a Dialog Frame is showed +Fixed in time for +Once we can get these fixed we will be ready to get other datastores working +Reopening since hasnt been released yet +If not though I will not be porting changes done to tests across to envers anymore +txt +The problem here is that you are getting this critical which already brings down your site +seems like issue is fixedAbout the release We are planing on this weekThanks +no servlet mapping for your jsp so it was getting served by the default jsp servlet which doesnt have a role +This is already a subtask so I cant add a subtask +This is just an option and it only matches usersdoing OR BQ over TQ against one field with default codecsim etcI certainly dont think we should switch Lucene to Cbq +thanks +Moving all issues to the CR +They are the ones who set the and the to the message context so IMO its really their responsibility to set the along with them +Looks lime my commit broke the build +Can you have plan for this demand +Lets wait for Guillaume to say if he was done with this issue. +One class of tests cannot be written today the explicit casting of other data types to BOOLEAN +Thanks for the response +Fine + +Cannot reproduce due to lack of feedback form the all issues that have been in resolved state for more than one month without further comments to closed status +Can some committer review the attached patch and commit to trunk if its considered OK +Better than what we had +bulk defer open issues to defer to not essential to address for +We could write an alternate CFOF like that can be used when the TT is running on the same node as CassandraThe reducer would write files to then when the reducer is closed it will contact the local cassandra instance via JMX with the output dir to be loaded into via streaming +I spent some time in comparing the code path for the working case and case in a debugger +Patch patch status until Prashants comments are addressed +I dont think local hints need to be put on their own queue +We plan to do once we merge the code to trunk +Looks good Jim +Attachment has been added with description Sample FO file that triggers the have used Fop Trunk version to get that above has been added with description A cut down P open bugs to P pending further review +Thanks Sandy. +Typically we dont delete issue tickets +I ran the test case alone without the patch from both Maven and and I get the same behavior it passes once then fails on every successive attempt unless clean is performed firstWith the patch applied the test passes under both and Maven regardless of whether clean is performed first +Thanks. +Added logic to strip out the and created our own version of the javascript that strips out the same colon character when resolving the form name +I dont know if anyone has ever done strongly typed forms in an action method +here is the patch +Hi DeyanThanks to care I guess Adam is in holidays I have not seen any messages from him for a while +Christian Thank you for your comments and give us details about your experimentsI think the length restriction is an acceptable option for splitting the Katakana compound token +IIRC the touching logic in prevents to AJAX callbacks in page from get dirty +cscs does not because its also applying the ellipsoid transformation its compensating for the fact that source and target ellipsoids do not have the major axis and flatteningI cannot be sure but I have the impression your expectations are wrong if you want just the projection to happen then the datum should be exactly the same including ellipsoids axis and flattenging +Sorry +I had a few environment variables pointing to the older version of qpid I was using +Was there any patch applied for this? Just checking to see if I have missed out somethingThanks for your patience Venkat +Apologies seems to be rvm related +In the discussion of introduction of a new state was not considered the best choice +These types are expected to be unique +Lets make a new version to add major new features such as this oneAlso on a format converter +Im just being lazy I ll test again later this week +not currently but a patch is welcome +closing the issue on the assumptiont that a current checkout will resovle your issues including the docs +thanks to Dan this issue is now resolved. +see alejandros work +forget the above comment +That goes back into the guts of the tool I remember correctly theres an assumption about the wsprovide tasks takes additional classpath references whichdont work when running in ant mode +Verified with JBDS and MS Widnows +It strange that nobody else get the bug could you try with an updated DB +Running on STS Eclipse I still get a classnotfound when running srctestjavacomxxx which uses srcmainjavacomxxx class +Closing as not a bug as a description was provided on how native queries should incorporate custom beforeafterIf incorrectly closed please reopen and comment. +Apply at workingvm +Does this actually crash the app or is this a bug about a noisy logcat +Each substitution into a URL would need to be escaped in some fashion to do this properly right? Sounds tricky +Note that this is NOT subsystem support the subsystem spec is not all that stable yet and support for it is very much more complicated than what I came up withIf by generated classes into version control you are referring to the jaxb model classes for the xml I now always put jaxb classes into version control after initial generation +If you start strict and learn the problem space you can identify optimizations such as these and introduce them without fear of destroying anything +No DBA was found willing to take a look at the schema +Jetspeed has the same problem +Fixed get to get +The requirement is that should have exact snapshot of blocks +Verified +sorry about the duplicate comment +Hibernate Core allows you do define the table and column names +will be handled with +Trying to fix it +One word fix +I agree with Carsten this change is of high priority and i believe a simple flag is the best way to accomplish it +No worries most issues there go through here first +Rebased patch to current HEAD +Heres the merge patch for that I used for checkin +But probably that is not what was intended +A test in this case would prevent relapsing to the old behavior +Moving out of +Thanks Jeff have you tried to run the console testsuite to see if this patch breaks anything +CXF is available in MEAD and as svn tag. +This fixes a bug indeed that was revealed by the vector test in certain situations +A to the SSO facility to allow other SSO users to authenticate against Portal Identity now these two are explicitly dependent on each other +minor typo in summary +FWIW initial steps towards adding a uf param to let users specify what field names can be specified explicitly in the query string with optional default boosts to apply to those clauses +my compiler is gcc +Patch looks good gaojinchao but it doesnt seem to work for me on TRUNK at least +sorry for long wait this is a table script with Spatial correct data and featuretype a will send today i found relation with jdbc driver +I dont think this is a critcal bug it is more a limitation of the use of and there is a limitation on whats possible with should be possible for your application to authenticate the user before you decide to use can you comment on the possible changes we could make with CAS to allow CAS to supply the redirect URL +Thanks for setting this up Ben +There have been some recent additions in click extras No rows found +Will give this some thought +Forgot to note issue in svn commit but it was r +So far I didnt find block issues but still need to verify it thoroughly. +Attached patch is not for committing +This patch is over for earlier version of Hadoop +So I after looking into the I agree it needs to be added +All tests done with same Sun theres only one thread accessing the Groovy meta system then it works fine but as soon as there a two threads it comes up with the error +Umm you did mean review this with her didnt you? +DagIve also committed the patch with revision +Try it I agree about removing multiple redundant fields a facade is better +Perhaps its related +AntonPlease review the documentation for technical accuracy +Patch doesnt apply cleanly +Please ensure this is true and I didnt miss somethingAlso note that I fixed a problem with the C cli code +Applied to trunk +Committed to master +Adding them to +To reproduce it youd have to write a custom Derby network client that closes a connection in the middle of a conversation +This worked like a champ thanks +Applied the patch +Can you check the creation order of your beans for example according to the INFO level log? created a simple test case that exhibits this issue +Looks good +Related Spring JIRA was resolved for Spring IDE +It just came to me +Thanks for the patch Benot +Maybe inverted and on right side would help and perhaps smaller +Tried to make a more focused patch based on rjurneys public github commits + to both! WINDOW is a reserved keyword in both the standard and in Derby Thanks Kim +MuraliPlease look into this bug +Disable the keydownup events or have it call some mco that doesnt add a significant amount to the DOM every time and youll see theres much less lag +JanThank you for the information provided! Well check the problem +Closing +Dave please help Rebecca with the release notes +Patch applied under r +I made you a contributor and assigned you the issue. +For tables with no primary keys an identifier is created using all the columns of the table +Uploaded a patch that adds libhdfs and gets it building +sort is used internally in to sort and group the map outputs before giving them to the reducer +I tried to pull down your tests but I evidently dont have the right dependencies +Was bitten by this today trying to upgrade hibernate versions to fromIs it possible to merge this bugfix to branch? The patch should work with codebase as wellIf theres no volunteer I can take care of that +So whats the verdict on this one? Seems like most concerns were taken care of +Would you please let me know +Committed at subversion revision +You could store this very efficiently or on disk +Please find the patch attachedThanks regards +It appears that I have forgotten to merge the fix from master toJust did it +Let me know if you guys need any further tweaking of this patch +reopen to set fix version to again with correct fix version +After inserting second delay it finished in seconds +Now I see the delay +For more information please enable DEBUG level logging on the logger +Attaching new patch that applies to trunk since there are some conflicts with the older version +Consequently all services and other required app will already be loaded. +The patch does not change the code logic +We really want to get this in can we get some feedback +Doing exactly as the use case states does not produce said error +It is easy to fix once you know what directory to look for +let me look into this further +Status update Aug Progress Due Date N Have tracked down Xharness bg process issue with to Terminal initialization code +Btw if it is deemed that these cannot be out of order and cannot be repeated we may want to actually go back and treat these as positional parameters +So if its not considerably more work that sounds like theway to go +Ill do some extra tests this week to make sure that this issue is solved +Glad it is sorted +But why the discrepancy? How come the and both provide an while the only provides an Executor +In the tall table case the row skipping is ineffective and gets into the way. +I tested this +and working correctly. +Thanks for your as patch against branch is implementation that can be included into is attached +I will get right away +We now have the nodeps jar +Opinions +But that is just a warning +Works as designed + +rogue master processes stopstart master seems to have cured it. +If a streaming jobs reads from Input but does not write to output for a long time is that considered progress +FYI I committed the here to trunk to get it going in jenkinsI will merge back to the branch +Changed Fix Versions to +Folks who use the MR API in Avro retrieving it from Maven central need to have an upgrade path that lets them still get the code from Maven central +Ive tested that this makes it much nicer to enable HTTPSSSL in Oozie and that this will also make for a more dynamic solution to as well +A typical use for this kind of message producer is to send replies to requests using the requests destination +So if you passed a filetmp to libjarsfilesarchives will copy it into users job +No deadlock but is blocked into a shutdown hook +Thank you +This error also apply to skewed with trunk +Without the current patch all tests fail +The truncated edit is which located in httplocalhostimagest know why it is an issue +succImho it should fail undef MRI but gives x even more surprisingly coding x +Verified +Made behavior consistent with other arg handling of put by throwing FNF if parent dir of a copy doesnt exists +It is deployable and usable in its untouched form +Is that expected? to add the session bean definition to the plan files +jvm version? And a jvm dump could help. +In this case then this issue would not be relevant anymoreWe will reconsider as soon as the API discussion has been settled +I checked and neither the checkstyle or pmd checks give html output +Closing this bug down. +Forgot to metion +Checking on trunk now +Ill give it a shot here and make sure it doesnt break anything you also provide some sort of test case? Perhaps something that loads a jar with some files in it that attempt relative requires +well look at it for or maybe note about the fix +If someone has a easy good idea fine but not worth much effort + +Attaching a new patch which I fixed Roberts comment +Waiting for bugs to show up could mask the potential for performance improvement. +You can see that still the old version is deployed inside Apache JAR really has the correct Metadata but still the problem persists +Den do you have any hints on how to catch it +Will commit once test pass +It is are we doing this or not +Please always close issues when done if you are the reporter the usual way is to close the issue directly without resolving +Ill done +Simplify the way to collect the information + +Looks like the link to the above test case is broken +to columns and click selectthe page make just refresh and the value of layout is not i wasnt able to simulate this issue on my local machine but it seems that it is a problem for many hudson machines +looks good to me +Yeah I think its a bad design as its much more common to make the mistake you did than really want an option value with a? in the text +Ive done some testing today using numeric preference keys will cause DWR parse errorModifying to use works fine but now DWR is causing problemsUsed modified weather widget to testchanged preference key city to numeric value failchanged it to a for example worksProduced errorsSafari parse errorChrome unexpected number IE Firefox Unexpected +not necessarily an optimized oneSo the solution should be if replicateAftercommit is absent then pickup the latest optimized commitpoint +Can you answer Edwards question? and also refresh +If the user wants to count without nulls then the user should use COUNTSTAR +Good points +The problem cannot really be solved by us because it depends on how the database system handles concurrent transactions The temporary solution is we now throw a in case this situation occors on H database +Attached updates to the Tools and Utilities GuideOne file was updatedPlease review and comment if these updates can be committed +Im not really sure but since we noticed such an improvement in performance I dont see any other explanationI will further investigate this +Is this still relevant since I see being used +Could you try importing again with the latest revision from Trunk or branchx? Based on my limited understanding of this problem I think it might have been solved withIn either case we probably should write a good unit test for this so that it doesnt break again +Patches applied. +Good catch there was a potential concurrency issue there with lazy resolution of class names to Class references +Its not too surprising then that a very long and arduous bisect points toward +Please change the lib setting in file +RayeesCould you please paste the latest failure +Oh just noticed that theres a in bsp package +I have uploaded a new one +User guide slide deck is incomplete +The changes look pretty contained +I just those tests but the fixes were superficial ensuring a mock call was recognized +If Derby is shut down right after the driver is loaded the thread will continue to run in most cases +Sorry some of the data included is invalid +So you get the sstables in L +I just committed this +Thanks Arun! +bq +We should remove the type and other deprecated attributes from the DTD before releasing +Changes were committed to the branch that will provide the soa build the ability to specify the statement delimiters when generating the ddl thats right +Will defer just because this is best integrated when adding new modular registration mechanism +This is a bulk cleanup operation to remove Version tags. +done implementation of a is committed and the method also is in place now! +Im concerned if upgrading breaks us this means hadoop broke compatibilityOn the other hand I see no problem setting up builds on Apache Jenkins or the EC Jenkins which redefines for the build +When we update for Ill rename the old version to +Its major because a lot of people have the Nexus and ordered the but they arent the majority of devices +It hardly depends on how many apps where started before if there is some heavy stuff loaded before we just need one picture to provoke it in other cases we have to shoot pictures more frequently +Patch from off I agree that backwards compatibility is very important +The warn patch for adds the suppression and redirects it to stderrThe patch removes the warning +Does it mean that the felix bundleRepositorys is wrong? Am I right to say the bundle needs version or above? that is what I am saying +Fixed and a test has been added with all the different cases +Then you can drag on the nd route +Not too bad +Also removes two unused variables and an unused import +Committed issue +Ive mailed the Codehaus support regarding this +Committed revision +xml instead of +Did you reproduce the problem in other JEE Server? I mean I agree it should be fixed despite this workaround were just trying to provide a workaround while this isnt properly fixed +These do indeed seem to be valid addresses +The copied database can then be cracked and inspectedRight now authentication is performed AFTER the database is restored +Disabling UP and DOWN keys doesnt resolve this problem +if you want you can write your own js for this you are not really meant to use functions directly. +Any update? +Thanks Dan! +This problem is comming with Could you provide a small war with everything needed and precise steps for reproducing the issueI make this in the next daysThank a workaroundI saved problem +This matrix is very cool +Ah but as per recent discussion about Extras site vs +Committed patch to documentation trunk at revision +Thanks Arun! +I think so will make backports easier +So I see the advantage just not tempted to add this at the moment but it can be added +And we are back to Yes But maybe the fresh perspective will get us through it! +Also the assigning to lanes is garbled +Id just like to know if theres anything Roo could configure differently to avoid this problem? Or if it is completely related to the plugin? +I dont think we care about bundles not managed by fileinstall +marking this as and so it shows up +The downside of a big number is that values are lost when the system shuts down +You file is missing a dependency on which means that the JVM is silently ignoring the annotations +Marking this as resolved +else +Reassigned because Noel currently its not active in the project +The integration is very ugly at this stage and the attached code is definitely not meant for inclusionFor very rough performance testing I created a simple test application that creates a versionable folder with files in it each containing about kB of application data +Mike resolve if this fix works for youWell +log was null because you did not override doExecute instead you overrode to figure out a simple way to make integration tests for these puppies too +Changed Installer to use LF for in CR +bulk close of resolved issues. +So its better to consume more memory for better performanceJust to summarize current default option would be used for smallembedded parsers and new option would be for servers +Brian are you sure about this? I cant duplicate the problem using the latest code for +that was ignored +The issue has been resolved as a part of the fix for +Furthermore AOP defaults to use anyway thus I consider this class a cornercase +I cannot reproduce this problem +tests included for both and paths. +v patch adds that +So +parent adding to change the name of default maven repo to central +Also could you break up the patches by component please +Hi TomAccording to the BPMN specification lanes do not have any execution semantic they are basically just in the picture +I dont mind seeing it grow to accommodate other Revision types in the future +Verified. +I m afraid this mailinglistJira issue reporting mechanism is not threadsafe! This issue duplicates +No one else seems to have turned this up so I question whether it existed in the first place +Bug has been marked as a duplicate of this bug +Carl agree +All the warnings are about fields accessed in the service methods which are not synchronized on the objects but should be fine as they are just read and not modified in any of the cases + for the to trunk. +Bulk move to does it still happen +Thats why I did not bother about the +This was incorrectly reopened +Working on a patch +I also have a valid license for Maven use that I got from Cenqua +Minor comment One line in and a couple in chars +The algebraic funcs have a return type but they are of the type +I used the tasklet sample job that Robert created as an example so it should be a lot better +As has not shipped yet maybe we still have time and could do this in EAP wonder if the developer tools is affected as well +Hadrian any update on this +If you are a RAD customer or even just use Eclipse with the Dali plugin you can use the metamodel code generation facility within these tools to process the file +In the previously uploaded patch file only whitespace waschanged +Thats why I say the easiest way to fix this is to change the look of the forge project detail page +This patch is only for this particularI dont think this patch is in sync with Nobles latest proposal +Which means that due to a mistake during reconfiguration the agent can enter this bad state and would then require a complete shutdown in order to fix itWill does this interpretation match what you have observed +Ok makes sense +Thanks +To which implementation described by Lukas are you referring toEDIT OK I got it + +has been added with description Slashed down version ofthis error was caused by the fact that the rtf renderer did not set an rtf style to the list +If in the future we decide that we need to add the custom styles in the XML there will be very little to change in the code and no modifications to make anywhere else in the frameworkFor the same reason i kept the public getter instead of retrieving the constant directly +I think that the boot failure may be merely correlated with this event not caused by it +I have no budget for automated testing but I hear there are some nice folks with Hudson servers somewhere that occasionally donate build space for projects like oursThanks. + +no longer longer relevant +I agree with Saeid the minimum time can also be defined by a value that you need and you can even implement your own job executor +Hi all I have run into this problem and I think that after a fey days on it I found a quite easy workaround +I have no problem with defaulting to long but dont count on any display like this being around much longer +i have confirmed this with as have others on yahoos groupim sure this is a well known issue but im wondering if there any plans toimplement this +If not we should reopen this issue right +The only cost would be iterating the TermdocID iterators a second timeActually thats major otherwise I would have already just done that and stored it in a separate cache for Solrs needs +One layer would contain only horizontal characters while others would contain vertical and diagonal ones +For whatever reason setting the busCfg on the provider in the just wouldnt work +that would not be very friendly for client side debugging +No it happens when theres almost no load +Is there any progress on this? +Just add the docs to this issue +Forget importing ejb as a binary dependency for now + Here I prepared a patch for the compilation warnings which include the pointer casting and the signedness issues +common nuisance error wuld be good to to so we can then rename to +Should we go ahead and close it until more information is available? +The JMX issue in is not serious in that sense it logs WARN messages +mdo files to get everything working +Please the patch by selecting the radio button +Taking a custom Executor and wasnt thought through +This I think would make it simpler to load up an existing xml document and then manipulate it to do what we have an example of a Maven MOJO plugin working in Maven +In theory the same is true in streaming but the details are more subtle +that makes sense +We can try that for +Whats the status here is this done in another bug? If so please close as a duplicate +Concerinig item use instead of for the tip I have the relationships working fine using +Commited in RC +My concern was that if I used my own policy file it would fall over trying to set this property but I tried it and that does not seem to be the case +Duplicate of. +Please verify and close +if that does not fix it either then our original diagnosis is incorrect +Without impersonation under it can be impossible to access files at network locations +Are they accidentally createdYes bad Internet connection +Natalya thanks for the! Is there a simple way to reproduce the assertion +set by user Job requirement for map tasks +I checked this jira in latest build +Feel free to submit a patch and reopen if its something you really want +because our internal Jenkins revealed that I missed symlinks in a test +Released in. +not to the HTTP content negotiation algorithmsIll be out of town for the next couple weeks so forgive me if my responses are a tad slow +TODO update the release note and wiki page +Rob could you review this please +right now users cna configure which registry they want tot connect by configuring file inside bin directory and start xbaya again +Im also not a big fan of mixing up performance and correctness but the discussions in this jira are referencing other changes so I cant really say what the best way is with all these changes in flight +Thanks Jing for reviewing the patchI have committed this. +Would not be better to have timer on the server side? I mean server will give response to the client in every case and it will not hang on unavailable JDBC connection? Message will not be acked on server and client will receive exception on ack this should lead for missing message and client will not be blocked +If I alter the test case so that there are several items in column that naturally flows item onto column then the fox extension is demod more readily +I think the findbugs count baseline needs an update +The default version is apparently determined by mavens resolution logic but there is still a default versionIf you are going to say that Ive not followed a maven best practice documented in an obscure document that that is your final position I suppose there is nothing I can do but spend all day tomorrow modifying poms however I still assert this is a regression bug +Verified that issue is fixed with sources as of revision +Instead provide a maven which can create the lib directory on demand. +Committed to trunk +Any case where the CP framework misses a scanner close is a bug +I used also ant with java +I have run into what seems to be another bugInitially I thought this could simply be a normal different behaviour between embedded and client drivers +Please verify this issue is fixed as you expectedBest regards at r thanks Richard +thanks this helped me to resolve the same issue for me too. +Again better i proceed with that tomorrow +It does everything it is supposed to +could also be committed with this one +Works fine in my works for me now as wellThank youbulk closing issues that have been resolved for more than a year. +Please verify and close +Committed in trunk rev +Committed thanks. +In the meantime Im unchecking the patch available box +But of course there are cons +See screenshot. +Reopening to make this change +There is no rush +lets try for next some resizing feature of iframe portlet in portal mode is already fixed so Id like to try to find a solution for this too +Trunk build was on July th +income statement +Thanks issue is fixed and released as part of release +m having some trouble finding the error again since it was tripped by a bug I have since fixed +Binary file +I heard that pmd is going to release a soon +GreatIt is a bit out of scope but not that much +Thats right +Addressing the dag plan part of things to abort early for invalid dags ones with vertices or having vertices with tasks +Please verify in ER Im watching this EAP release note looks OK but I wanted to wait with closing until I see EWP docs +Hi Tonythanks for having a look at thisIm reopening this issue because I can hardly accept the current situation as given +Heres patch for branchBuilds and Integration tests are all passed +I am positive that this patch solves the issue +Ive implemented polygonAlign options using enumeration hopefully it will save you some work +Verified in CR +This patch depends on for unit testing purpose +we should wait for the rpc client to be created before doing UI +I solved in a quick way which returns an empty range when its given length is zero +Hi Gavin r proves it we have a success +This ebadcdbceaadddc removed the previously ignore rules due to recent bvt merge dont worry alls fine +You should see link for Delete after Display NamesClick on Delete and you get the NPE +Please be kindly informed that this patch depends on +This is less flexible but introducing a newnode type would have an impact on other areas property lookupsI wonder how far our support for comments should go +Smooks editor is no longer being developed +Works for meThank you Alexey. +Would you be interested in helping to extend that Vagrantfile to add the integration steps? Then wed have a one step approach to see this integration working +FYI I assume there should have made multiple operations on a broken properties file + +Try to upgrade Hibernate Validator to version or later +Get to an agreement already +Closing the issue as the bug reported here is fixed in Wagon +I have looked at your fix and the test code and it looks good to me +Framework needs to run on R Core in +Commited fix in trunk +It complains about the base diff path not matching the repoIve tried empty trunk trunkgiraph and other creative combinations +Im not sure if those are a results of the terminal I was using or something else +I retract my earlier statement about injection not working even after the is found +bulk move of all unresolved issues from to +rev openejb rev stage of minimization seems complete +Attachments are usually part of the message for this component and you can access them using mtomIm not sure if it would be a good idea to allow passing attachments untouched +certainly AltShoft on the platform i refer to the source context menu which can be shown via hotkeys and not just the normal right click context menu. +Rebased patch. +Revised the docs and applied the comments from above +It looks like there are some questions included with the problem statement and I have no idea what IRISH means. +I have a feeling that the issue may have been worsened though as Seam JMS already takes into account hornetqs TCCL issues and works around them +I have applied the patch I attached at +Praveen Yeah +The change went in as into trunk +Once we require Java or above we have the option of using wildcards in the classpath and that will simplify matters and avoid the classpath length issue +Attached the new patch +In revision I committed a test case demonstrating a failing assertion when accessing the builder for a node when a intermediate node in the hierarchy does not exist +Found the problemThis is NOT a bugI am an IdiotI had a stale in the stack which had old Segment Inforamation hence it tried to access segments that had already been mergedSorry please close this bug +just committed this +Hairong I guess you mean provides the visible length and under construction informationI will open a JIRA for the under construction information for HDFS and HFTP. +Https security is not required for our application but must be supported if request comes by https because we serve pages that can be included into another sites through iframe. +Due to the use of generics the get method that was used to fetch values from the map by upper case key wasnt being called + +Anything else to fix in the tools topics perhaps +My recommendation is to have an additional option lowercase which defaults to false +Reassigning to Marc. +I like option b which gives us some secondary control of pushing out stable versionsmilestones targeted for JBT milestonesOur branch is dedicatedfocused on the JBDS product release so there should be no interference with our unstableproject checkins +Attachment has been added with description Possible Type fonts appear to specify a default encoding that doesnt match the actual encoding of the glyphs in the AFM file +Hi Sean It sounds like a good idea to upgrade +If the first one happens to be down all the clients creating files will fail +Could you check what went wrong? +i cant quite remember +bq +I will close the issue +If the base id is different every time you start the tm theres no way to determine if a branch that has been prepared on an but not recorded in the tm log came from this tmI think the problem here is that our documentation is not sufficiently insistent that you NEED to configure multiple geronimo servers using the same RM with DIFFERENT tmidsAs long as the above is followed we could randomize the most significant bytes of the long counter +Logs from a failure where was expected but was returned and the pycassa script to reproduce +I tried creating ACTIVEMQMSGS manually as well +However its over the limit on file upload sizes +What version of Guvnor are you using? The template editor has been for and support for enums included +committed code for validation and model mgmt not yet error rendering +Currently all it is doing is succeeding if the manifest parse does not fail +Some comments on the second iteration +He didnt do the bug either +As I understand it and my experience has been that once a directory is listed in it is ignored by the CMS +Not sure if you are tackling this as part of +Here is a update i created a new patch against the latest trunkThanks Erwanhm strange +Ill commit to +thanks +Ill keep an eye out for a passing build +My issue was more with the ease at which a user can break a syncd connection +Cancelling patch for the extraDo we know how many tests are reporting such exceptions but passing successfully +Removed the jetty feature from. +We will need a solution to support the use case of arbitrary user supplied security credentials +Changed an uppercase tag to lowercase so that it would match the corresponding lowercase tag and pass the lint tool for release notes KatheyKnut was wondering on whether the release note correctly describes the behavior on and +If the password escape mechanism is broken how is it not a bug? There is really nothing improper about the password used and it is currently the only way to centrally enforce security and to have zero password generation or clear text keys on the client +Do you have ESRI ESRIs ARCGIS Desktop running thereIf you dont have ESRI software there maybe thereis direct loading way I will ask tomorrowif thats possible +Disabling Automatically Detect is only to have consistent behavior in M +from me +ok will move to then +Fixed in trunk and branchxThanks for taking care of the extra sysout line Erick. +As far as I know we have no special treatment of numbers in the labelling code we just use own layout engine +My problem is I tried to Bind a mixin parameter to the textfield translate parameter but when I try and set it from the mixin its read only +I used the following project to test the transaction support in RMI +Tested the patch and exception was not thrown +Should have +Well if there was a URL encoding in Java script yes +He has a few emails to the geoserver and geotools list is building a wcs based on geoserver +Tested with latest build fromThis issue is still reproducible using this build +Right now a datanode that fails to serve a block is blacklisted on the stream until there are no replicas available for some block when the list is cleared +The install is working for me +I am puzzled +Not sure about the linked bug though. +Thanks Anurag that explains itSo the correct fix is not to assert on XJ but on XJOr better change XJ to match the error code used for embedded +This attachment includes a sample bpel process to demonstrate the issue +Is a in a wars classes dir supposed to indicate Seam should scan only the classes in classes or does it mean it should scan every single jar in lib even ones without? I was assuming the former +I should have added to the previous comment that additionally to the change should have an option to keep a connection alive for think this issue can be duped to +I like the way your patch tries to retrieve the from the Servlet Context before falling back on the +Although not completely implemented I think the current level of support justifies resolving this issue +More patches are attached + using username will make the search much more efficient Why is that? Are most operations that touch the job history? I would have guessed that most were rather that the most frequent operation would be to browse through the job history by time +Marking blocker +Sure +Wim Bervoets did you test the bundles that you patched? Do they work for you? If yes what environment are you using because I have some issues using them with Virgo +Sorry I was out towards the end of last week and so couldnt respond sooner +Thanks Jaikiran +Applied patch into trunk camel and camel branches +New images committed to trunk and pushed to wallace +Thank you +Defer to defer open issues to defer to not essential to address for +Link Corrected +Reviewed patch committed making the code a bit more safe +Please ignore this patch its another test to see if the bug is related to verified that is happening in the failing test +deploy helloworld +v +yes definitely +Cache using JPA Cache Store can only store mapped JPA entity as value JPA entity is limited to one ID field marked by Id w primitive String or Number type without Cache key must equal to the entity ID +If you were for example running the API under Tomcat while Derby was serving the UI youd definitely run into issues though +Whyhow would they be out of order +Sorry for this jira +the problem is caused by get where add and add are usedthese criteria do not require translation +Thanks Alex +Bard you can decide if there is an appropriate point in the lang spec to emphasize that clocks can be used to coordinate activities across multiple places +Fixed the build +Will do +It seems to me that this case is revealing a deeper issue with the groovy compiler as demonstrated by the behavior differences with the Maven vs +As for the second return it is dead code it wont have any reasonable effect +have a look at this +not sure its related to the project type +I think this is ready to be committed +We should not introduce an external dependency if it doesnt make things significantly easier for us or our users +Nick I dont want to kick that job since I assume that will trigger everything else to build and not sure you are ready for thatThus assigning to you +If you go to Targetted runtimes on a project and click on us the Runtime composition shows something very specific to us as if we do it on the default adapter the JDK is shown +What was the thinking behind using a pointer in this case +I was doing a search for something related to the problem Im having and stumbled upon this issue +NYoure rightPatch v passesLets see what Stack says about it +Attaching my patch please use it if it looks better +Attached patchIve tested it on my system for during the last month and it works fineThe problem happens when an error occurs and the operation is that is theres no message receiver +Why is this not sufficient +Its working now +Mika thanks for your patch! One of the Avro C developers should review it soon +First off does this happen in the client or the server +Patch attached for the documentation update +change ACINIT macros project name to zookeeper from +things has changed so not relevant anymore. +A very first preview of a plugin that helps to simulate crawls +We really need to work on the error messages here +Please restart the namenode with the verify flag to search for a different imageThis new flag would then cause the NN to read through the image compute its md and compare against the stored md +as discussed with Michael offline the latest version of the test is lacking a save call after the move +It seems that the current patch removes this validation but doesnt added anywhere else +If users want to use some db as a catalog server we just have to implements interface +has both and native versions of the various methods weve bound into it +I revisited this just now and updated my patch per Alan commentI do agree that the original code is very confusing but after looking at it closely I think it works as expectedMatching does have an effect if the rule matches clientenabled is set based on allowflag while if it doesnt match clientenabled is set based on the invert of allowflagOn the other hand youre right that the loop should stop as soon as clientenabled is false as theres no point in trying to match the remaining rules if it continues to deny given that a previous rule has deniedI have refactored that bit of code slightly to hopefully make the logic clearer +Accepting +Have also set VPE set to block when BOTH upstream and downstream are building and JSF to block when upstream is building +I looked our java code there is no place to delete primary storage pool after its been created +show the error message only if there are no project in r of branchr of trunk +Adding Permaine as a watcher not sure whos responsible here +I just wanted to mention the error handling issue for consideration and Steve made some good points to justify the behavior +Thanks for review Ted. +The code may have to be changed elsewhere to get this fixed but I guess it would be the very same issue and it should be treated together with this one doesnt itI mean + has been marked as a duplicate of this bug +As part of this would a mocked Server be appropriate? Or should that be a separate ticket +cssvjava jar part of the reason it works the way it does now is that when committrue it a single commit at the end of multiple file if we use the param based commit it would either need to specify commit on all of them or keep track of the last one only add the param therei dont object to changing to use a commit request param instead of sending the XML form but this isnt a bug its working as it was intended +If there is a problem please open a new jira and link this one with that. + +Also patches were rejected by patchMake sure to update frequently and perform a full Maven or Ant build before submitting code +the prev patches breaks a testcase and it is fixed +Ill check the job later in the evening and hopefully give you some useful info +Jody please review this very simple patch +The reason it is a problem for us is that we have idle connection cleanup turned on +The other work around is to add a dependency on the plugin for the version of jasperreports you are using +Otherwise Im to revive my Linux +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +The reading side has the same problem as noted by Rohan +I guess this is how mkdir m is implemented in Unix +Since this is allowed now I guess at most it should show a warning +Same here with on Win also the error message is very confusing in my case I see the help message of the java command +I have fixed another bug with the Guththila hope everything will work fine this time +I will post my findings hereRegarding your point about iOS the issue comes on Android too +User assigned to mapbuilder +Btw the svn repo seems to be down at the moment +This is a C problem +Im overhauling the cache somewhat because the cache needs to consider whether statements were prepared with specific auto concurrency or cursor holdability requirements +This particular case appears to work well now +Your right they are not apache compatible +TedIf ok i can commit this today +Reopening the Jira as per Aleks comments +I will create a patch for it +Yes sure Thank you for correcting me +Looks like a weird serialization problem +it would be very convenient for me having them in we do included this in that would be in a release instead of +Patch revises with current information from the wiki and from Thejas Nair and Sushanth Sowmyan plus some editorial changes +FYI I think Ive got one more version of this patch coming probably later today +I need to try to reproduce this error in a system +Probably for this issue we need support directly from some guys of MicrosoftI think that we need a shared server for implementing and testing the protocolI have some contacts in Microsoft and Ill try to understand if we can follow this wayI hope to let you know good news soon +good on ya. +JBDS will include the Central zip so resolving. +Other tests exhibiting problems with special characters are numberformat and numberformat and output +checked Seam +uid play a vital role in +All is in order AFAICS +Any preference on the variable use heresplits splitcount subshards numsubshards +Fixed in Kepler SR +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content issue is more than a year old with a work around that is accepted by the customer +rather than etc +OK thanks +Looks good to me +ops diff too much +Hey Rob can you answer my questions here? +again bugs with invalid fixed for release as fixed in. +When I get time I will test this with and check if is the real cause for error +Is it really true thatcygpath has such a small upper limit to the path length? If so can you testwhether the attached script solves the problem? can confirm the problem using cygpath on the long path value in the reportabove +But this is only a temporary situation for the internal installs right +Committed to and +Thank you for your assistance +thanks hong +fixed and added test +Keys are string +Thanks Martin +Im reviewing this patch +and create a new bug report as well +Any further updates on this? Does it still occur +Incorporated in refactoring +Maybe it is out of date +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Rather than semantics being that the NM will launch multiple different commands +Please detail your environment and how you disabled caching as well where your template resides +At we used a preset max amount of allowable skipped records +FYI cant reproduce this issue on with Derby due to the Hibernate known issue +Please help to review +Not following how this would workA parent pom doesnt get used unless other modules refer to this parent which modules would refer to this parent +Requested id for subtasks resolved. +update looks VPE screen? Does here printed some error message and link to FAQ +The issue has something to do with the version vs the version +The trouble is that snapshots need to be deployed manually perhaps I did not deploy them after the changeCan you check out the CARGO source code and compile pleaseIf not I can also redeploy +It starts a Chore inside regionserver to update the znode every secondsIm not going to finish this patch +Same problem with and +Ok so an NPE isnt good but it should throw an errorIt is intentional that the does not work unlessthe document has a viewBoxPatches are still welcome +palette had a popup menu with settings where you can select smalllarge icons for palette items now there is no menu at all +Committed revision +Committed this to trunk and +I have committed a patch that should fix this issue +does filtering out test resources qualify as a reasonable approach +Hey LewisThis tool is already present in Nutchgora +With the holidays this week I should be able to work on this Thurs +a sample webapp that cant reproduce the issue duplicates so we can close it as fixed. +Ive rolled back the change to the test +Verified. +We decompress because we need to build secondary indexes compute the sstable stats clean counters delta etc +This will ensure that the forwarded request will not prematurely time out +Deferred to the priority at Vijays request +Should be writable now +Tested manually +Closing as is already released +mixins and subdirectories of those +Also the documentation will be fixed in release for Apache Ambari. +I added the patch as proposed in revision +THANKS NEILNow Xerces can get on with using the Xalan serializer +Sorry that this took so +Closing all issues fixed previously after a brief review of each. +Closing all issues +It would be an interesting idea to see a hive job just turn into an oozie DAG +Let me know +Moved to future versions due to existingPlease repost question to users forum +Do you plan to include those as further patch revisions on this ticket or open followup ticketsbq +The question is to myself or anyone willing to dig in to JAAS +We could include that in the design spec attached to +Returning the to the client is necessaryWhen is thrown by the broker the connection between the broker and the client is still established but you can not use that to create sessions consumers or producersNow we get a useless connection and it seems like a connectionn leak problem in this caseFor failover transport in the client if the client detects and then close the current transport and retry to connect to the broker or when the broker detects the it just close the connection so that the client will do the reconnection logicNot sure it is reasonable or not but seems it will be more transparent to the end user +I did a lot of work in the cookie code to bring in line with the RFC +doesnt use the client so that logic is not going to kick inIts a direct test of the client its not suitable for JME +another patch with a simple testcase +The patch described above +Can we separate the the edges role into a separate class that can be tested and add a unit test around it +You can read about it in Java API or my javadocThe patch also adds strings to be internationalizedThe patch do not contains the unit tests for public synth classes +packages file that references a third than that I cant imagine this failing in this fashion +These changes affects Seam integration testsuite with Embedded as container for integration test +I think this is actually not what we want because it affects the constructor too +NETs naming and doesnt start with a double I +If URL syntax is going to keep throwing surprises and cross platform issues at us maybe we need to create a syntax that doesnt pull in a lot of unwanted baggage +There is another api getcolumncount that currently depends on getslice +the eclipse plugin in is not the reference +Unfortunately its at the cost of readability and maintainability +Showcase works fine for me with Tomcat and JDK after Dons fix for the missing xml file +I dont know whether the synchronized is necessary +scanners in etc +updated patch I plan to commit in revision +Configured JBT to point at that folder and it seemed OK +IMO there are two optionsA Fix the specB Fix you sure that Application action is related to? I dont see any usage of create in myfaces codebase +But that only works for the cases where we do not want inhritance specially when polymorphic calls are not required +Id suggest you invoke ant on TRUNK rootMaintaining two build tools is and we cannot assure that maven works for everyone all the timeSorry for causing this inconvenience +Tested in +Assigning to Dirk +Whoops I probably shouldve titled this with milliseconds instead of nanoseconds even though in the Timeiso I pass in for the scale but I specify the milliseconds in the constructor +Basically I mean that the confluence wiki page on Codehaus is great but thats not the place where to this kind of documentation in my opinionIts great to list the available features explain how to get started but I should not have to leave Sonar open the Wiki page to understand how a rule works +I suppose it is difficult to make the meeting room layout customizable +depends on but the package is not exportedSo I propose to also export the package +Ill take a pass at figuring out what is up with this tomorrow +I guess the problem was that the was not escaped +AteCould you suggest a way to fix this issueThanks your portlet would display an article or a banner an wont be able to use more than one instance per page you would more important than an improuvement +Also of note is the use setting in +Close this defect as not a bug +Secure permissions on this file are really really important and are validated by the binary itself anywaysYep the issue here is that if the admin has messed up and has an incorrectly configured floating around the user can evade those checks and then use it for ill purposes +The user needs to extend this class to add just business logic such as implementing validation process the uploaded file etc +Support for maven repo was contributed by Thomas Hiram I have limitedno experience with this to provide good answers to your questions +Hii am not very familiar with configuring datasources in but i would suggest not to restrict the configuration of views to datasources +Ive been studying Forrest and Cocoon a lot lately and would like to lend a hand +Patch addresses +So we can Open the file and read it if not then you have todo so not sure if always doing will not be a better option Open the file Create a new and copy data from one to the other then close the file and rename +Closing this since it was committed. +Is there any reason not to support an expression for the ssl configs? The other attributes that point to external resources support expressions. +If we can make a dummy private key and store it in the certs folder then we will never have a blank key in signing fetcher file Also this way we can throw the exception when one tries to sign a request without having an actual key +If nothing shows up initially retest with trace logging +Merged to the branch in revision . +Patch with test case and potential fix +Erik I admit that your handlers Velocity template presented in was remarkably succinct and readable but it does force the use of Velocity which isnt in core just as arent for similar reasons +Committed to master +I used long type as the userid and itemidit can not parse the input format when run the job +By default the config value for this could be null and this would mean that we would use the operating system specific class +Versioning We were going with a previous suggestion to add a version parameter when we go to the next version +compaction triggers it +We probably should Do you know who might be able to do that? Since the account was with a phony email we would need an administrator help to access it +It is probably a good idea +Please feel free to any of those when required. +Modified for unit test +For now Nexus needs to have a stable ip for inbound deploys it uses a combination of user agent ip and user to divine a single stream of artifacts from a buildWe are rewriting staging planned for in a month or so that will allow a different mode of staging where this wont be an issue +Start sonarcontained before sonarcontainer and they both fail +Yes +Attaching and containing a small change to the topic that describes the contents of the manualsNo changes need to be made to the tools guide so this is the last patch +Current patch passes on but fails on ubuntu +According to Koen theres nothing NN since M +removed from fix version since it is already a as a jar file is not consistent with the specification +Ill push this up. +I could look into the changes required +But thats my point I have to rework my configuration to work around a bug in an existing class +Committed to branch +Im wondering if we could implement that while implementing the ability to add new entities dynamically to the search factory +Would you mind trying it out? Thanks +So wont fix please. +can you try this one on windows +now that it is included in MHOME it will need to wait until a Maven release +Updated patch looked great +The version control is gone from the IDEA project this a single or multi module Maven project +Revised patch +After I created a link file which points to command line dfs command works +So shall we have committed it to the trunk and +Committed this +Bulk close resolved issues +Committed +Not a TP problem +patch looks good +Created tag for code at subversion revision Created tag for docs at subversion revision +Fix for is merged to AS community master git branch +Thanks Bertrand latest patch looks good +I am this issue because of tracking +AttachingIt contains some minor javadoc comment and formatting fixes and additionsThanks Ted for the review +The nest big work item is fix the hcatalog tests and the remaining qfile tests +uses abdera +Fixed by addendum to +I guess we should pass proxy to the ftp connectionSo i add this code based on your patchFurthermore i add some tests for this modificationCould you please try this new patch? Thanks very much +Fixed in BranchSPCP Branch Branch and trunk +I think we should pull in immediately after thisThanks Bobby +has anyone tried applying to +Todd once is done we can get started +Hi LilanthaThis link is working properly for me +Thanks Ive attached a new version of the patch with the fixes you suggest +submitting patch to address Konstantins request +The order in which properties are defined in the configuration file is important. +Adjusted patch to include fixes from +Rob please make the review request public +Added headers +Fixed revision by introducing both variables daemonproperties being deprecated and daemonProperties being the preferred one +It was only executed after I did a restart +allows to be added to a ugi so that they can be made available for authorization checking in the RPC method +Please verify +Patch attached with unit testThanks to Todd for getting us a reproducible test case +Changed to use the new generic Event class with dynamic attributes to avoid dependencies +This has two advantages the user knows how to construct a valid query and we can clean up the patch work inside edismax +Keith when are you proposing doing the binary search? On the block after the block has been read into memory or before? I would advocate keeping block sizes small enough to be read into memory and if a binary search is done do it on the block in memory +this problem has no effect on the clusters health +Otherwise the admin process is to onerous to constantly be adding new people as assignees +Ok got it +Thanks +Fixed +Pete did you try to build the broker with these changes in? Im seeing locally defined symbol also imported link diagnostics forSingleton methods +I dont want to draw a dependency from EJB to the JAXWS API because we bring this in via +Oh it would be lovely to have being upgraded to a more recent release +Do you mean I should use instead of? Is there any example I could follow to start a job history server? +shipped in +The seems to be missing from the +Why not leave the archived data in DFS with its replication level set to infinite and make a change to the DFS client so that it will replicate blocks of such files locally when they are accessed and ensure that blocks that are local are accessed through the local file system instead of through DFSWouldnt that be simpler than having a whole new mechanism +I have committed this +Please confirmRegardsWernerCan this issue be closedColm +Setting to a large number for theMETADATA table should disable merging minor compactions +Pointing trunk to Beta +Unfortunately I have not been able to test it in mode as I cannot make to run trunk on hadoop +As soon as I get disconnected I start getting a flood of thisI can easily cause this to happen by having more than one zookeeper client in a single process where both zookeeper clients are using a chroot path that is multiple levels deep +Fixed findbugs warnings and add missing Configurable interface from +Revision +the type of clientresponsebodybytes is intt d can not represent it in system +Turns out this is now a race between a compaction or flush that finishes and a partially created object +Thanks for adding the test +Feel free to commit this +Once is in this feature will be updated to address this +The problem is that the plugin doesnt take into account the package directory of test source files when they are copied +I agree with Stephen on this assuming the intent is to stop adding to and move new development to the new components while continuing to support with bug fixes etc +Theres still some weirdness in the karaf stuff but lets be honest theres always some weirdness in the karaf stuff +One approach that Alan proposed was to make bags when they are created to ask for memory and have a central broker in charge of the memory pool +The suggestion works pretty well after I extendedMy next question is whether this can be done outside of Java? In other words is it possible to add user defined functions in the hibernate configuration file? Looking at hibernate source I dont think its currently possibleExtending the dialet classes is okay if working with a couple of databases +this functionality works correctly in the latest whole M +Hicould you please provide a patch for this bugas it makes life so much Eclipse all you need to do is on the project and select Team Create Patch +Are you still working on a patch for this Terje? +This is the initial version of the code I wrote seems to work pretty well but I didnt test it extensively yet +And we would have to add an API toperform the same action without an exception being thrown +The provides multiplexing and dynamic appender creation +Yes that should help +UIMA Status page updated as well. +If maven is supposed to be about the benefits of convention over configuration configuration should not be required in for the most basic task of compiling Java source code +Just a thought if you apply this patch I believe you will have to run autoreconf i to recreate the etc +studioeclipseplugins +I am sure means empty and you are creating empty vectorsSimilarly you are doing max df percent of +The diffs in the master file are expected +Dont have enough data to guarantee that it is the correct fix but if someone else can verify that it works for them I think a new version with the fix applied would be nice. +ing issue as part of JIRA cleanup +based on your ideas +I fixed the above bug and fixed other problems as follows CREATE TABLE and INSERT OVERWRITE will not print the result in CLI +Attached a patch that removes the duplicate year that is within the otherwise clause +Glad it works for you. +This is now working as expected +Committed at subversion revision +I dont think the window resize problems are related to this fix at all +I do not have code to support what I propose at this juncture but will come up with it in the upgrade JIRA issue +Thanks Yonik! +I tested up on a cluster and it seems more regular and live now than previous but I may have missed cases where we used to rely on the rpc retry +Does it happen very often +The attached patch fixes the mkdir errors in these tests +Ill see if I can get this done this weekMichael +sorryits my have make a mistake of spelling the name of method setCommand +Thanks works fine! +We can likely add a lot more tests to cover other areas +You can change the properties in in the plugin test to prevent the failure +Attaching the as Patch Available for review +Added the copyright and am adding a testWe may eventually want to store arbitrary bytes too but no need to nowI changed the property to be which means to store the bytes of the body as a fieldThe testing has already paid off +thanks Chrisminor fixup here the rescan thread now removes objects from the pending uncached list for a DN if the nodes are no longer cached on that DN +Uploaded a patch that takes away blacklisting and changes the name of the flag accordingly to relax locality +AdrienI am using the suggested approach of using only key and not and then mapping a separate property and it is working for me on Oracle. +looks good +Downloaded and added the ezsetup library which is supposed to be the bootstrap tool for setuptools which should allow programmatic usage of nosexunit +Maybe try manually removing buildlibjars and run ant test +Ah I see +Im sure youll find something there! +Not vital enough to block current impl disables the link when theres no application available that could be imported +note for myself brokenBarrier must probably be volatile else I will run after race conditions for hours again. +In a sample file uncompressed text MB compress orc MB row index in orc of course Trevni could be modified arbitrarily to match the needs of Hive +Committed to branch and trunk +No this patch allow property traversal for subobjects fields in property paths +Whyd you take out the class check in +Some preview +thanks for the patch +I havent played with any of its settings but in theory it should use a fixed size connection pool so there shouldnt be unbounded growth in the number of connections like we seem to be experiencing +I have add a check to verifiy the current import not already try it out and close this issue if it works +However I do believe that there are a few community members working in this area so it would be nice if we could fine something to unify around +So in the DSL its not even possible to use onException after routes +since most of the standard properties already got featured as named parameters this is now barely used anyway. +Theres no consistent style across files however +However it was committed only to trunk after the branch point +New version of the patch based on trunk integrating Elis feedback and unit test +Part of a bulk update all resolutions changed to done please review history to original resolution type +Also there was a comment about using to sort +Ahaaa +First composite +I have unit tested the operation with S but it could still use integration testing +I am not what you can do is to create your own github account clone Owens GPL compression library push you changes and file a pull request +Uploaded the rebased patch +The test case applies better to the modification +I believe this is currently actually done on purpose ie by design +ok looks good to me +Committed. +Lads signaling using plan types seems perverse +Thanks ATM +Calling if method is HEAD every time doesnt seem like a right this to do +But thank you very much for your feedback and well see what we can do +This is the configuration introduced to the file for achieving this +Because the web container has TLS support we just change endpoint to SSL without modifying policy and test fails due to endorsing checking +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Parallel processing with file input will be on option in m but is covered by and children so this one can be closed. +Tx for the reiteration BikasIvan to clarify this is how it works is given a URI it could be a local file URI or one on hdfs sees if the path is on the same FS as s and if not it copies it +just a matter on what to expect by default isnt it +The list we will be active after great thanks! +also as maxim mentions we should doucment our tests running on some given ports +When I suggested a savepoint name that was unique within a transaction instead of a globally unique name it was primarily to avoid the need for synchronization +Whoops test failure that I should have noticed but missed because I added the test but hadnt actually built after adding the test +Looks like there was some problem with my build area +Small to trunk. +If you were in my shoes you would not have used the word simply +One goes to hadoophdfssrcdocsThe other to hadoopsite +But one thing though +for reviewing it +It seems this only provide atomic write of acceptedEpoch and currentEpoch files +Hi I built a local version with your sources and they seem to work ok for our use +feeling motivated I guess another patch with a bunch of patch I suppose rename existing tests to and a new that just tests the class itself +At a minimum somewhere there needs to be a noticewarning that use and ownership of uncommitted patches is a potentially questionable and risky activity and that permisssion can be revoked at any moment +What is +If you can send me the shapefile or attach it to this bugfix I can have a look at it +I think the structure layout grammar and location in the source tree of this document and the related images deserves some discussion since this may be the first point of contact for new Derby users +Will do +no its not in the select fragment generation this is relevantthis is something that should be done when building the named queriesclosing this since jira is for concrete bugsrequest not devel talk +So if some project is still on Magnolia and would like to have this feature it does not need to migrate to Awestruct just because of would just need to discuss the protocol of the data transfer that would meet all the requirements +Fixed the ones that were obvious to me +well I guess I didnt fix it +HICould you attach to the case your schema located inhttplocalhosthelloworldxsdThanks with cs + now calls skip on the reader only if exception was thrown while reading the item +Wait for someone to have a use case to need themShould it go insideIs it the same algorithm as the current code? Looks different +Testers want these tests to all be part of the same framework so a suite of regression tests can be run with a single command +Committed revision . +I will check +But we think it is the better approach for the lead qualification process +has been marked as a duplicate of this bug +This is the code I mentioned beforeTo build the project youll need Javamail activation ant +I am ok w the code being ASL licensedOption is also fine w me +Issues resolved in +ignore last comment. +Added information on how to build for Hadoop in +I would suggest issuing a warning instead +Thanks! +What happens if the +I have verified the problem by using this patch and adding the following tofalseThis tells the plugin to not report ip addresses for any interface other than the loopback the network cable from my desktop doesnt reproduce the problem because I have desktop installed which adds two interfaces with their own ip problem is that certain finders on the server side try to use an ip address but filter out fake an ip address use the property in like so should be valid format but does not need to be you have applied the patch to your branch just update in the agent pdklib directory edit and restart the agent +Can you please add a pointer to the RHQ defect? +Im also getting roughly the same speed I was getting before +Init should be calledBTW set all the fault properties to the in message context when you are throwing the fault +Im sorry for confusing +This is more than annoying +bq +Brahma Reddy I have added you as a contributor to HDFS and Common +Best reject this fix it properly +IIRC implementing was a need in order to configure James with and still use the same bean also for the server domain list +We have agreed this is a must fix for +This will throw no exception +in JBDS +on and committed to and trunk. +It looks like its a derivative +This JIRA adds a generation stamp to the Block object and is a blocker for all succeedingThen Nicholass should be committed followed by that I am currently working on +I youre smart youll listen to Petes suggestionsBe aware that you have to configure for the jtsjta quickstarts +I dont have any idea how to give Jenkins credentials that can be used to commit to svn +Build without patch +Mostly irrelevant now + Username dsundstrom +am i missing something here + +Maybe thats related +It tells you nothing useful just restates the name of the field +i dont think there is a good place in javadoc for something like this +I will test archetype next +Else any subsequent transactions might accidentally flush inconsistent state from the previously is a general Hibernate recommendation to essentially close the Session and throw it away after a rollback +The attached testclient uses the Axis samplesamplessrcuserguideclientsas basis modified with global initialization of a loopThe corresponding service is also from Axis samplesRunning the client in Eclipse with M for the session the client gives up after about invocations +but indexer cfs deleted +But in any case I wont be able to look at this again until tonight +Also removed the version from fixed versions +KevinIm still seeing failures in that look related +is there anyway we can reproduce or verify this? if you have any test sample code to verify this it would be really helpful +I think this patch is ready to goso I committed it to the trunk as revision +line youre calling the build with the parameter map which is +I till dont understand why we had cases as the last comment +Closing all resolved issues for. +removed wrong fix version +Workspace log for +I am sure that will help +As far as I am concerned this is not an issue since I cannot repro +sar going in the configurationHave you tried having the output of the archive aka the completed and packaged +Hi cheyenne that looks there any chance that each image can have a link associated with them so if you click it we can link them to a page outline the feature it is showing offIn any case this is a great start so ill try and work with ASAP +There isnt anything inherently about Hub pubsub restricting it to just that set of widgets seems like additional work to reduce its value rather than enhance it +Thanks a lot +Very nice! I will run some more data through it now +Im marking this issue as incompleteThe description says the bugissue is intermittent which makes it harder to write a reproIf a runnable repro is attached please reopen this issue. +However the second patch provided seems reasonable to meDo you agree on applying it +Thank you for sticking with it +Marking as closed. +building the vm added for access however it will fail until you add an ssh pub key to your ldap acct +jason re Should we be generating? Using explicit artifact references +No writing of new store files +This sounds to be applicable to unpackOptions as well +Updated Scott Fleckensteins partial classes patch against latest sources added test cases updated syntax files for new partial keyword +Is this still an issue with +Patch looks great Shai to commitYes that is very sneaky about the private fields in innerouter classes its good you added a comment explaining it +Here the code was trying to interact with SWT in a different thread and that is what caused the nasty message +My concern was it might throw this error after partially upgrading to and might require rollback to go back to previous release to save namespace +Closing it now +seperate from trunkI just updated after your last updaterev not sure what to tell yousorry for all the trouble +Or someone clever finds some way to change an interface without braking backwards compatibility +Fixed in release +Steve Im not sure that this issue belongs in HHH JIRA +Could you tell me where I can get information about this one slipped thru the cracks and wasnt scheduled for +Williedo you have any news on this issue? +As the proposed fix is to change documentation to state that you shouldnt change a value from a default to something that wont work I am going to mark this as out of dateTo the reported please note that the configuration option to change would actually be in or not JBTM +Daniel I love you dudeThe label should probably just read Launch Developer Studio +Just confirming that I have tested this on and it is still a problem +This pattern can be provided via xml with the new attribute of the element +So you have that and its still not working +did this query work before svn revision and fail after this commitYes that is what I observed when I ran the test yesterday +Regardless I appreciate the quick response +Ok but just one more question When is a component marked as pooled but actually does not get pooled? If I configure it to have a +Lets attack this in two different modes Remove support for both type and nameThis will be addressed by issue Add feature of caching preparersThis issues will be addressed by this bug +Thanks Arpit +Resolving. +bump +bq +Committed to trunk and branch. +This patch works in the general structure of and as such should be significantly faster +We should land this initially as a compile time option in the devbranch to get wider production time on it before moving it to defaultThe main reason is that it is invasive and complicated particularly in theway it will interact with the VM system and it would be nice to see how itresponds in a variety of environmentsIf it is much better than then perhaps we should package it up ina more general form as wellWas the design based on another allocatorpaper? Any referencesjohn +When I follow my steps I am still getting an issue not in error log but instead of that a dialog with title Error building project archives node appears +if M is enough to start Eclipse +the method you invoke on the formatter Reproducing the problem can help us find the solution faster thanksRgds is a testcase that shows that xerces includes the character xe in the XML although the default setting for is to fail on unrepresentable characters +After testing this again it may be related toIm going to close this as a duplicate +ve just committed this +Added trace log and a stack dump +It would be really nice if this could be fixed in time for the upcoming AS release +I tried the test using javadocaggregate and the problem is resolved +Stacktraces +I made some progress on this issue +Im about to mark this as a resolved +Bulk update to fix version +Attaching patch that is adding he protoc dependency on a version that is driven by the Hadoop profile +I guess this will make it into then +The PR was we say that the maximalrecommanded platform is now and update parent pom and CI jobs accordingly +Use a new singleton class to wrapper class +what led to this stack trace +I will adjust this JIRA issue to reflect that +For example testing that an array access thats out of bounds throws an exception +On the other hand we have the source anyway +Could use some rd party data structure butI dont get the compile error thats a java change tooFor the time being Im inclined to just leave this as a java patch and wait until java adoption to apply it +There are two left over files in tmp directory causing this failure +this should already be fixed in +Patch to River +Fixed in rev there a release date planned for +See Deprecate multiplex transport +Guess I spoke to soon the xml one is there too +Glad to hear this +And the Sign In button is misplaced +Please test at earliest convenience +Jaimin could you file an additional jira for creating unit tests for all the config validationPatch looks good will commit shortly +I tested it by installing it and successfully running a job +qfwhonamewhoThis is probably due to an error and the solution is probably throw an exception but right now this is not considered and will lead to a stack overflow error +push to +That should be set from the build but how? System property? The reason is that while this pattern can be useful for many cases it wont be suitable for all +Looks like this patch was included in the fix for +says configure as node wiht specified source and sinks says dont start status serverThis is being documented in the cookbook +Upon entering incubation was our stable branch and we were just about to releaseThe code should be pretty similar to the data store +How are you parsing the xml? Did you register an error handler with your application? Take a look at either or to see how to use oneDavid downloaded the latest source code and it appears that the bug got fixed. +were you using the same JVM before the upgrade +Attaching patch for Branch committed this to trunk and branch +Some confusion for me which patch file is the correct one did you alter the tests as described in my coment Where should it work or or handle as a separate issue duplication of encoding logic is a separate problem +Thanks BillIve added checked logging at both the places otherwise we pay the cost of for nothingIll commit this shortly +Here the patch +HiI have spent some time fixing the Resource Adapters serialization issue in order to make it work in glassfish v +At least now you know how the problem would have occurred +Bernard if you can confirm that this works for you I will close it +Not sure about how to make it useful to folks who just installed to deal with a path deep into a gem seems we expose it nicely via HTTP on or somehow? Versioning is still important +done thanks. +Looks reasonable +Thanks Sanjay! +This implies that all the blocks that remained under replicated have only one replica and only on this specific datanode +DevsPlease reviewThanks revision +The problem with rendering templates that are within plugins could be seen from both the render tag in a plugin GSP and the render method in a plugin controller +would it make more sense to deal with this in subclasses of Hits +release bulk close +It resolved my problem +Where is the being generated? A lot of the DOM Elements in WSSJ are explicitly constructed with the wsse namespace +Thanks +Thanks Njal +This is of course Optional Feature T +Alan you can close this task. +committed at revision +Committed to branch and trunk will continue working on making compression better in other jiras +But i find it somewhat ugly +Ive moved the discussions to closing stale resolved issues +war and deploys it throuh the +Section of the JSP specification states Resource injection is not supported for JSP pages or tag files +Opened forum thread +We could check partitioner on bootstrap the way we do cluster name +Passed Unit Test and System Test One drawback is that for if the number of partitions for certain topic in the source cluster is less than the number of partitions in the target cluster then some partitions would not get any data +Created a new PR which fixes the regression and adds a for it +ok reread and spotted the subtle whitespace diff please apply to branch and set fix version accordingly thanks +where and how is the folder definedTheres another similar issueIn general Im not happy about our goals generating a war file at all +patch submitted runs on like Jeff fixed it. +A naive approach will break any existing configsI think I would add this to the bucket of things that get fixed in with a full of how configurationinitialization happens +But there is a small issue in issue +Its due to extension name of temp package +Indeed this is a classloading issue +I filed an appropriate issue in JBIDEopenshift +Attaching a repro and a draft patch for the problem +some data dir changes snuck inb +No fix yet +oops patched from my wrong branch +it would seem the first fallback would engage before it and since the binding for is already around it would take over? Perhaps I am missing the case but I think there is still some overlap +All tests in pass in my dev env +This version of the patch enables the torture test +like log entries exception stack fine with jboss GA and postgresql +I remade a patch for trunk resolving conflictsUnfortunately after the patch is applied xc will abort by an assertion error at +Do I understand correctly that there is currently no functionality to upload except if it is an image +cmpilato screen a great program to protect processes in the same way that nohup does but better as it has virtual terminals +Committed revision merged to for revision merged to for +Two little suggestions Recommend you use the two argument versions of assertTrue and assertFalse which take error messages to display in the event the assertion fails +Overrided some functions related to marking +I am using hadoop job kill jobid to kill the job +patch to trunk +Is this still happening? Do you have any plugins in your project? Are you able to attach a project that reproduces the problem? I know that the fix is simple but I think it is sensible to find out why this is happening for you since it shouldnt be +yeah I think with this fix stupid GSE will finally work as it applied the do you want me to leave this JIRA open till you check it on Jetty and Google App Engine and see if there are any issues related to this fix with them +Entered new ticket. +Built a version of the patch for for +If more data is in the pipeline invokes the task executor to start assembling the next message while the current thread sends the message to the configured channel +Heres the simple way I thought about takes no requests has two ctors one for SSDV and another for and initializes the proper FA underneath +Ok done separate thrift source patches attachedI double checked the versions of thrift were right for branch and for branch +The titles is a bit general and there has indeed been testing improvements that made it into +and classpathcom +I will try to reproduce with different versions beetween and and post the logs here +Given the tradeoff between the two situations Im opting for having an automagic sync that the need to remember to do cleanups manually every timeCurrently we ant resolve on ant test anyway so its to full syncing to meA middle ground would be to have ant cleaneclipseidea resync everything and ant test and ant resolve not do anything special but like I said I think it should be consistent with ivy files +Ill start going through these soon are ofbiz decided and ready for the move +Great +And I think what you were saying is that SQL spec has different associativity for INTERSECT compared to EXCEPT and UNION and you are wondering if Derbys implementation is enforcing that difference in associativity +Any additional checks would have to be added to the which is created behind the scenes +Thanks Tim! The fix looks good +Docs fixed thanks +Will try to have something in the next day or two +If we dont release a new bugfix version somebody has the chance to patch his checkout +I dont see any way to support multiple clientspecs without explicit support in Continuum and SCM +Can I expect some resolution of this issue +I also tried a artifact and got a pom +Attach a +This is done +This is caused by an file +However the build process does not distinguish between application identity and datastore identity wrt to JDO files +is probably fine but did it expose the original problem? I never saw it failing +Tested +Created a RC to be voted for +The patch was getting ugly like you mentioned so I also abandoned going that wayAs I see it your patch improves the situation from what we have today but also incorrectly fails on a much smaller number of querries +Since currently user would be looking for resource only when json format dump is needed storage of mapping is done only when user wants to get dump in json format +for CR since it only affects RSE +Ive missed this so far +The simplest way is to move dataScroller info table footer facet +Hopefully this resolves the issue +Thanks for making this change +I cannot speak to the patches correctness but I manually applied and saw great improvement in terms memory +Either they are configured or hardcoded IMHO we should expose the information to users +All the tests ran cleanly for me by the way +Thanks for the reminder! Ive added a corresponding warning to s +I created for what I believe will be the various parts of the statement pooling feature implementation +I suspect this is because the Axiom support relies on Declarative Services which may not be being included by the userAncoron please can you try this with Declarative Services enabled and let us know if you can recreateThanks totally agree with your argument that the mentioned statement should not be reached within an environment where declarative services are available but I have an appropriate environment and I see the axiom services but nonetheless Axis does not make use of themAdditionally I issued a grep Hnr +But if you think something like JOOQ makes sense I am totally up for it +Seems to have been fixed for some time +Bjrn to allow us to better understarstand this issue can you please create test project that would have some regular classes and aspects and corresponding configuration for compiler and aspectj plugins +Im just thinking that is not a parameter +This patch checks to see if the project has been deployed and grabs the domain port from the server +Ive been very busy lately so I did not manage to compile the patch + +I think it should be resolved as WONTFIX +Thanks for starting the release note on this Myrna +I had someone else here try it and they also got +committed s leave this closed since is already released with the patch +The plugin needs checkstyle which currently is in beta to use new language think the plugin wont be released with that support until goes final +To implement this Logging needs to be integrated +Feel free to open new issue if you are still having problems. +if grouping just needs the string representation of hte unique key field then thats what toExternal is forI agree +committed +Verified on work properly. +all issues assigned to +Verified present +Thanks Edward! +Ive just tried it out and it does not work at all what works for me is mvn clean install release +Right now it works but without being able to get the Servlet registered for a mapping means there can only be a single entryPoint in the web app +That sounds good +Just configure the plugin according your needs +all issues to the CR +The in the metastore directory includes so it should be getting +Committed at subversion revision +For example youve switched to using a in the parser which might be a good idea but which makes it necessary to review not only the changes to relax the date parsing but the entire new parsing mechanism +Unit test to followCheers on the value for inclusion +It ends up coming out as a due to a trycatch wrap somewhere in the call stack not sure if thats a significant problem or notPatch is done against branch because I cant install Java on my home computer but if its acceptable I can do a Wicket one at work. +Hi Gary SQL Server has the same problem as +Cced also Ben +make a backup of eclipseconfiguration edit the original and delete the entry for the a restart the compiler should no longer be around +Patch to have toString use good is there a unit test that shows this error before the patch but works after? If not we should add it to this patch +updated version of the previous patch +I committed it to trunk +Small exampleorignial request httplocalhostconnectfacebookfieldtesttargethttpAFFlocalhostFdonecoderesult of calculate httplocalhostconnectfacebooktargethttpAFFlocalhostFdonefield catch thanks +Not sure we need to push the responsible of knowing about EE packaging down to all but we need a alla for Extensions like that need to bundle them selves via descriptors etc +Im going to commit to TRUNK. +This patch fixes the dependencies upward +open separate bug for thatif you go and look in the deployments folder of your deploy project it should contain the modules you have added additionally +Verified. +Should we close this issue +I suppose its possible that that some no MOPTS cause and my current ANTOPTS do notbq +Committed issue getsingleton is broken it doesnt return the right factory when istrue. +Hi MarkusHave you tried your example against the latest CVS version I believe this issue was addressed already +From the download pageHowever this is NOT recommended in Maven builds as it does not retain correct dependency metadata which can lead to Maven working incorrectly +Im sorry for inactive contribution + +Could you please resolve the conflicts and upload a new patch? Thanks +Fix applied for not yet done +I forgot to add in that this patch feels like a hack to me with loadResource returning the properties to be added in but I could not think of a cleaner way to do it without completely refactoring a lot of that code +New API httplocalhostrestpackagesdefaultPackageOld API localhostapipackagesdefaultPackage +Maybe its not needed I havent tried +Fix the No tabbar bug +reopening to fix s fixed not invalid +No tests +Attaching new patch incorporating Hemanths offline comment +The continuation should be independent of the selector +Thanks Jean +Switching the static state to instance state would help a lot +If you are extending can you send me a copy of your zhCN so I can reproduce the problem? Thanks +Does the host field do that +Matt I have taken up the review of the latest patch +I guess I agree that its importantThis is probably not an issue for installs on a smaller scale but gets annoying when youre in for moar +However the is still created when the runtime target is executed +Fixed +Thanks. +Thanks Todd +Attachment has been added with description Example using CMYK colors with a CMYK color Peter I didnt want to imply that your patch was of low quality +I am wondering if it makes sense to use the PR just to encourage contributions +Needless to say I do not have easy access to any zOS systems. +Any reason that the clients would shuttle arnd the servers so frequently +What is the content of symoutgoingbatch on the root for batchid +Committed to branch +Unzip in fixed now +no problem +Just to be sure when you said just keeping the most expensive things in the genserveryou meant keeping them out of the gen server right? Pretty sure its just a typo but wanted to double check were talking about the same genserver +Also fixing the failing test cases +Pushed upstream. +Could you please update this report with a URL +The methods can be easily implemented through the callback and serve IMO a corner case see the simplified part of hibernate support in Spring a lot of methods were removed since they were not widely used +Thanks for everyone involved for this wonderful contribution +make a API and put that in use artifact repository factory instead of new remove uses of other than a runtime dep in still the things from might also include common artifact naming artifact itself should know how to generate a file name at the moment this is duplicated across many classes and many assumptions are made about the format +The template change is made specific for and rest of classes remain unaffected with the change +Even if we allow an app to ask for multiple paths and they do so its more likely to spread the load around for apps with unsophisticated load balancing algorithms. +A new patch integrates comments from John and Ashish +patch applied with thanks! +Thanks +The only thing that makes me wonder about this event listener is that the task will be deleted after this event listener is executed +Grey edges on the screenshots should be deleted in Migration guide and GSG guide +looks goodWill integrate later today if there is no further review comments +Fixed by fixing is now out so closing all resolved issues +Committed to trunk and branch. +sounds like it is doing some continuous repairs Why not justPS After the rename I actually looked at all the changes to see whether you are doing something different and might be running repairs actively +including the +Add the unit test slightly modified to reproduce this issue +Patch against the Junit test +I already noticed the in the stack trace +its going to have to be scrubbed in order for the release note generator to work correctly +Yes its no longer +Tests ran cleanly for me +Performance overall looks much better but Im intrigued about node async repl perf with JBMAR +But I question whether the initial implementation should contain either optimization +Ive assigned this issue to you + +Works great +This is a problem in AS which has broken api when it comes to handling spaces in deployment pathsThus marking as wont fix as long as AS isnt a priority nor have the needed fix for us to fix it. +Please make me Oozie contributor +But I dont think Hadoop expose this construction +Had a typo within the file +The issue helps preventing bugs like +Dupe of +JacopoAshish +The best would be to create a new issue about this specific problem Vinay +Ah to explain all labels from all layers are lumped togheter and painted later layer ordering has some effect on the label painting order but not that much as grouping and priority both require some out of order processing. +Perhaps you could try adding some code there that prints out the filename? It does look suspicious +anyway it would be nice if you could test that +committed as dbbdaeedfb +Save the users ugi and reuse it in the streamer object +Ok there I agree thats over specified I had one thing in mind but its not the only wayEdit Fix to clarify like This will have the two codecs that would be needed by the client and server +Althoguh the good thing about Spring is that it can be completly standalone run simply in JVM without any other requirements +Thanks a lot for the review Eli +Could you please attach the binary +Ive merged the tests and added new target to that copies test resources to the directory +Since we are going to use the provided by the companion code this fix will no longer be included in Felix so I am removing it +It seems that removing the DESC keyword from the original query is enough to make the filesort disappear +I ran the failed tests a couple times locally and they passed and earlier run on this jira were fine +it shows up only with Arial font of that size +Can you post the contents of your orm mapping files and the jpql xml file +ATM reasonable to lower the min id to for the testsSure fine by me +After this fix it renders now to which is not as its supposed to be +Thanks for the clarificatoin JacopoI suppose theres a UI issue here rather than a problem with performFindWe can close this issue. +The service which is just the example with a minute delay +A problem remain with pig testing cant run against YARN server although one can run any MR against it +That sounds like good solutions to me and yes its better to never show an empty classpath container +Anyway I close this issue will not open anywhere anyway +I have added it back in please keep it up to date and in good conditionThanks. + +the same with Drag and Panel +Please apply to both branch and trunk +Revision +Old dog learning new stuff +Thanks for the explanation +WDYT +Changed is only for the action is imagesxstudioactionstomcat is referenced by +Ok and heres the branch commitAdding containersrcmainjavaorgapacheopenejbSending containersrcmainjavaorgapacheopenejbconfigAdding containersrcmainjavaorgapacheopenejbutilTransmitting file dataCommitted revision . +That is certainly one issue the message does exist twice in the body +I have a feeling this is related to +Thanks a lot +Whats the best practice then? Reopen until decisionOr rely on rescanning all changes in case its gonna be +Btw I think this patch is pretty lonely sitting in reviewboard +FYI The stress test involves running a number of clients against a single broker using persistent storage +Committed to trunk +Resolving as duplicate. +Completed. +Waiter is now deleted at appropriate time +Closing +Closing old resolved issues +oasis file committed +shows the new Naming Form Page with only two section starting in expanded state +All relevant Pig tests pass +There must be some ways to implement this +Committed at subversion revision +I too agree +We were seeing these on jars from Maven Central not just our own ones +Right now we are testing our fix and results match our requirement +Ive pushed a fix in caafb but I need to write a test for it +Fix by settings loading improvment +Ive created and linked it to this issue for tracking purposesThe fix for both and has been attached above as +Resolved Adrian if you want further features can you raise another JIRA +Since for differs from the version for use with I will open a separate issue against +to removing too +Doc +the same works for a node +the +Any ETA on when we might be getting the Git repo for curator available +The only way to declare usable proceduresfunctions is to use the SQL Standard syntax including jar ids USAGE privilege and classpathsThats a very clever technique you discovered for granting privilege to set properties on a basis +Resolved as Wont Fix acording to Andrzej Bialecki. +Second patch +Hence can be considered the successor to as well +Keith applied a patch to use more concrete return type than in the generated unmarshal possible please fix this for the next release as it makes code generated by the source generator uncompilableThanks Matt Im looking into the errant patch +Normally they are requestreply over WS +Patch applied +Verified. +An effective that Im using is to also create a file and shunt the bean definition and are both merged so they can reference between themselves +Making a change to the default value beyond that introduces compatibility issues +The patch looks great for me +Bill did this patch work? If there were no problems with it in production Ill commit it +I think I misinterpreted this +In addition there are the beginnings of instruments for data cache query cache and prepared query cache and a JMX platform provider +What do you think +Thanks Tahir! +This was discussed some on the mailing list +If I just provide you with a collection of classes then you have to scan the classes for the annotations you are interested in +how did you prevent automatic splitting +Im going to upload J All artifact to the M Repo +Thanks Mark the patch is fine +If not feel free to reopen. +Committed thanks for all the work on this Kenjiro and Jens +right +Cleanup changes +Something else I found is that it appears the database is being setup twice +In addition there may be more Korean comments +For this revisionwith sanetrue I get same exception +Sounds more like a JDBC driver problem +Yuki mentions that it may be caused by too +digging in +Attached you find a new patch +Going to leave that one for now +We cannot throw if a heuristic rollback occurs otherwise heuristic resolution will not kick in either during normal execution or recovery +I want all of the user interface programmers on my project to use the bean validator + +Duplicate of +Dan had mentioned in a thread on a different topicIn addition most of the packages in are now sealed whichmeans that classes in a sealed package must be loaded from the same jar +It removes the line with JDBC information from sysinfo and also removes calls from the xahelper class which lives in +How about adding a test config dir with values that make sense when running ant test? Something like buildtestcassandra patch should fix the broken tests +You most likely dont have permissions +If property gets changed or removed in subflow it is also propagated back to original flow +Attached example to reproduce for +This change also worked when we moved over to activiti +It was intended for another issue. + added get and get +I dont see how this could be done in a backwards compatible way +wsdlfirstxmlbeanswsdl +Lets please stop adding random commands like this +Can somebody elaborate +Use the Roo backup command and attach a zip of your project +Attaching a bug for the backported patch +For you would need to turn off the spring schema validation. +Thanks a lot for your contribution! +Arun Do you have the input generator for the patch +This time with support. +Indeed guys this would make sens +I tried to find other reasons but I couldnt +This was simply a bug due to ignoring the argument but calling a similar named method +duplicate so not keeping this. +Wow I cant believe we missed this patch for months +But due to discussions about the usage of and decided to go with some new approach for the trunk version +Thats a good question +The reason I suggest that is because I can foresee two help systems with contradictory information or one system that is kept updated while the other isntI global file can be loaded in the +Completed +Thanks for the update? any idea when its going to be fixed +File exclusionsexclusions will only be considered for Java resources +Any objections to committing this now? Id like to start on and want to avoid conflicts +Without the fix the Measures view is not working +A separate module that builds along side the main giraph jar for extrafunctionality +Few points Weighted methods could be removed from and should also be synchronized Should we rename add to add to be consistent with the other method names +bulk assign all unscheduled issues to next release +with no methods now compilesBuild synced to SVN M +I really need this +Please find attached the patch with the comments addressedthanx! Committed the patch into trunk with revision +Patch looks great what a sweet cleanup! I love all the code removed from IW DRCan we remove ? People can write these if they wantAlso you moved +We use the plugin in a lot of projects that are locked to maven +I noticed a subversion commit message and considered it committedApologies again +See the screenshot +Which application server are you using +This seemed more logical and fits better with the existing framework +Thank tou Jukka Id be already working with the current revision but I dont because Id like to have sources available when I use tika as a maven dependency +I left some more comments on the hive patch phabricator review +Attached patch file for this parameters for proxy host and proxy standard maven codestyle to +Please comment +Duplicate of is already fixed on trunk +Appenders are just for appending log messages +This works well on our environments so we would like to get the change into Trunk codeI will prepare a patch against trunk shortly which will solve the problem +To provide my use case for context recently i have been working on modules that make use of h for storage namely dbconfig and monitoring +Generally you DONT want the user to close the app anyway +Thanks KimIll keep that in mind but will hold off until were sure we want to forbid those methods on LOB columns +You should be able to specify the wsdl port you need generated +This issue needs so that the fix can be applied to the correct Also when I tried this with on Windows there was some problem and would completely hand when executing the query upgrading to solved the issue so I am assuming it was some problem with but have not had time to check it will add some comments to the Wiki for anyone using so they can fix the file themselves. +Unused imports While looking at the patch I realized that there are a couple of other things that we will need to follow upa +I will do a of the baseline and post the results today +The getindexedslices method doesnt offer the parameter for a keyrange that makes this useful for a map reduce job +Ah that makes more sense +config generation disabled for installation under root user to fix the issue +I have not investigated further dependencies so I do not know if depends on other files in +Im definitely not opposed to but I dont know that it was ever a success I think it might have withered out of commission +Once a project starts to have more than iterations we would be forced to navigate to find the iterations we are most likely interested in +If your or is very high you might be flooding your myself server +See the simple joined patchNourredine +The issue is solved for the maven validation task by redefining the snapshot reposity to be local +Whoops forgot to grant a license +Jeremy The jira is usually assigned to the person who uploads the patch and has done work on it +What I have a problem with is that the demos are not using the application name +After I made the change the test worked +runs by default over Felix +I wonder if it is OK for the patch +Please give me the OK to make it public Rajesh +What exactly would you expect to see in terms of integration +not logging it +This property has to be set in and is optional +Verified this works as expected on issue is fixed and released as part of release + +I should have kept the other process pid changes separate from this bug fix +lets just use Usage logic and pass it to the openshift library as a string + +Fixed in rev +What breaks in exactlyAs for an update to EMF yes this is something i am very interested in for a number of reasons just never got around to completing it +Have a look at srctestjavaorgpermacodeatomicHopefully its self explanatory and the logging output as it is should show the SQL output +will also need change +Bulk close Solr issues + +A more detailed dependency tracking should be done in the context of another jira +So it solves that part of this issue +Added full deprecation new patch incorporating Simons suggestions +bq +is the actual changefixUsage fixes places where the set api was being usedChanhes creation to be via a static method in itselfWill upload another patch to run through jenkinsThis is all WIP +Thought this would be more complicated +Some graph store implementations might be able to answer that question without iterating over all the results +yes and noBecause this trick doesnt with surefireSo I didnt check which change has fixed that +InOut handlers are dedicated only for handler logic +A cluster is a graph so use structures instead of custom ones +I am not sure that it will matter but here are some of our project statsWe are using the following plugins in that application +I think doing lengths is probably +Bulk close of old resolved issues. +done +bug brought forward to message to be less have updated the text for the error message again to remove the for data integrity the last edit command was removed as this was not actually happening. +Erkki did you try the method suggested in my post from nov +please Affects Version and Fix Version are +lgtm +It looks like Javajavalangregister symbol is not mentioned in file +Assigning to marcel for further processing. +Split stuff is whether we should generate one or multiple joins for multiple expressions that go over the same path +As I think the problems of and should be separate I opened for the fomer issue +go does not work +same fails on DB multiDB VS enabled not optimised JDBC connection +Woops I meant to report this against GROOVY sorry. +Closing old issues +bq +It should be noted that running and passing the consistency checker does not guarantee that this database is consistentand recovered fully +bq +I wrote it in a way that should be robust against future changes like adding continue to the loop +If we remove the assignment then value of running reduces would not fall back to zero in case a job which is terminated when it has maps in total and has running maps when it is terminated +Running As +Also can you provide your values for PATH CLASSPATH and JAVAHOME? I suspect you may have extra quotes in one or more of them +Moving flow elements within the same container should always be allowed +Resolving as per users comment +on v for commit if test is green +The whole site hands or only the adminNow that Railo is officially out you should really update to the stable version as it fixed many issues with us know if this happens after you your emails report a stack trace? Can you post itSee Michas comment above perhaps you should increase the timeout on your cflock +Found the source of problem +latest patch from giri +This is by design. +Attached patch +Even if you swap both changes as we check against the schema each single change you will get an errorI think we should do this check after having applied all the modifications and only then +Snjezana you looked at spring ide are these issues fixed in the updated version +There are other bugs for missing behaviors and enhancements not yet implemented. +At least for the simple searches I tried I found the ranking needs to be improved +issues after release +Was done alongside refactoring in. +Thanks Koji! +Ill run the script attached to and see what happens +This is really a bad bug +AlsoPublic mutable ListProtected mutable arrays +Committed in revision +looks good for me as r +Agree but would be great to deliver its description directly in the as there is already a fileTo have an all in one consistent packageThats the only missing part in my opinion +Implemented a sample to test the +Ill work on improving documentation as wellThanks +Regards applied. +This will be in +the menu item should just be put in under the Run menu where it usually is +We have exactly the same problem here +Ha thanks! I am in bangkok and that was the problemI went into System Preferences Date Time Timezone turned off Auto and clicked somewhere in eastern europe +Heres the patch for this issue +GEP build finally fixed this problem via upgrading to release. +Ive committed this to trunk and. +Closing the bug after verifying the fix with the latest master build +Where do you want to do that +I think there are too many different approaches to dealing with attacks and people should generally plug in what suits their requirements +Thanks for applying. +Committed revision Thanks Patrick +Please add at least one screen short illustrating the problem +set affects version and fix version +I added the weakreference stuff +Is there a canonical dataset that we ought to use +need to make one more change +Thanks +Linking the opened infra issues +v includes a test that passes locally for me +Some concerns It looks like doesnt have many adopters despite being around for a while +Sorry about not responding sooner I have been on vacation +added imported statements of system test specific classes +Thats easy +This is now done +I only tested it on OS X +bq +However I dont understand why this method shall only support creating of new and not updating existing onesChecking if the resource has any property isnt an appropriate way to check for existence of the resource if you want to make sure nobody has ever talked about this resource you must also make sure its not the object of any statement +to trunk andThanks Chunhui for the review +Hmm cleaning and rebuilding the job fixes that issue here +Tests pass locally again +Patch generated on a slightly outdated trunk which caused a conflictUploaded the refreshed patch +Shane bear in mind when doing this that is going to change as were porting it all to AJ CDK +I am sorry Knut I misunderstood the comment about merging to so I thought It was also committed +But as we are supposed to remain compatible between and we have no other choices but to postpone it to since any interface change will brake the compatibility. +Reverted changes from rev and in rev My next idea to fix this is to give the ejb web service deployers sets of abstract name queries so they can find either the jetty or tomcat web container +I just encountered this bug on a recent project +Should I do the same in my implementation +Hi DagCould we also remove the messages for J J and J from messages +Tomohisa Im assigning this to you per our conversation on iRClet us know if you cant do this for any reason +looks good to me +Fixed +If I remember correctly an AndOr filter with a single term is invalid and will throw an exception if you try to evaluate itI can do a patch though Id want to make sure were agreed before doing so +Custom Goals on project dependency in the Goals down key to get into the popup window +We also need to spill earlier if memory is constrainedUsing freeMemory may cause us to spill before we need to but the important thing is that we make sure to spill when memory is constrained +Perhaps Recycleable and Disposable? Or maybe just +can you provide simple deployment to reproduce thisotherwise contribution in form of git pull request is always welcome +This needs definitely to be an attribute +Thanks Daniel +Marking as critical since it is causing a functional fix for the empty problem +the release still appears to have this issue with the unsupported cast +nick please answer the questions instead of just closing itThe pull request is not closed and the pull request got the details about why its a bad fix +It seems it is better described there and this is just a particular use case of that one so I close this as this one it seems it was fixed on so has the fix. +Attached patches allow the url to be normalized to allow continuum to retrieve pom files in a flat if this simple patch is still I dont think is relevant but does describe the issue of restructuring the working copies +I reproduced the same problem on my Ubuntu +See object P open bugs to P pending further review +I prefer box is implied +I can address this in. + + is guides are checked versions are pages and guides are check versions are corrected where necessary +Lets put this into trunk +Thanks Ben. +Latest should apply cleaner on trunk now +Committed revision Setting the flag Release Note Needed since we need a release note for the changedThe patch removed one error message but it didnt remove the translations +assumes that views are located one directory above +Committed revision thanks +Lenny what you put in the block parameters of will still probably require different markup when you change your layout +wait +It is currently not possible however to define these in the User Task +Hello Im working with Kevin Stewarts code and environment and can verify that this is indeed fixed in the latest trunk. +I need one of the development team to talk me through the example that ships with the Message Broker so that I can in turn point out important parts of the code to readers of the Getting Started guide +FYI the reorganization of our build has been postponed until Max is her ein NE next week +in fact the issue is that in creating a page we have some buttons under the popup +to be able to predict the class of an instance one must classify it using all the trees of the forest and because each mapper has access to a subset of the trees a second job is needed +I have fixed to resolve the that caused many of the test failure +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +Hi TimThe fix looks goodPlease close this issue +At the moment this requires me setting all profiles not operator as Erwin suggested would be very helpful +When adding a new widget and selecting WC Widget I see a browse button and when I click on the browse button the window accordingly +Well it simplifies work considerably +I had started a discussion about a source earlier that was configurable +Could I have a status please on this issue from either Tony or Giamma +I too would like to see this bug fixed for but the reality is I am not going to get to it soon +If you have a testcase with a particular setup that suffers from this problem please share so I can test my solution with that. +Resolved the wrong issue +Thanks Vinod! +This patch contains changes to Solrs core from the last attached patch +Also apache build machines are not hitting this problem either +Thank you! That made things much clearerCould a note be added to the docs in case others make the same mistake I did +Added in shouldnt this be closed as a duplicate of the original +Can you a bit more explicit on how to set within Eclipse forI can find no way to do it within Eclipse with Tools server editorScreen shots maybe +Patch attached as requestedCan you tell me a rough timeframe for the +I dont use these bindings +unicode sounds good to meThe matchVersion could be used to ensure that back compat always works +So all issues not directly assigned to an AS release are being closed +It should really just be removed but concern with removing it is compatibility and whether anyone actually uses it +If using an xwork snapshot the continuation subsystem will now be initialized properly +veriefied +I reopened this issue because there is a minor bug which I found during my upgrade test +What if we add a new bool attribute to the component itself which allows the users to specify whether they want to use titles? something like displayfalse as default should be ok +A with patch for the issue +Verified by Gregory. +dims +nope +With further work though it should be rare as there should be other options for performing large dependent joins +The else branch of Rev get looks wrong contextPath pathInfo viewIdIt looks ok to me in Rev +the deployers should be in the else the will always be started beforeI move the deployers into the +I dont know if its easily reproduceable I can check but I cannot switch into a nightly builds for long I add an internal debug log option to be more sure whats happening +Can we close this issue? +Wed have to do both or we are applying a special meaning to s Named here When doing component scanning Named is more than just a qualifier annotation its suddenly a kind of stereotype then suggesting a name value analogous to Springs own Component +Good luck +But nobody has to build it themselves +r represents the merge into trunk +It would have the advantage that if a node has a stale copy of a block it could incrementally update the block +I will look at this problem in next weekAshish +Must I install Fuse in an approved location? If so what is it +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsernodes +Craig what do you think? The only thing I removed from the patch were the changes since they are already committed +But we may not be able to easily eliminate themYoure welcome to try to try the high road yourself +Thats the theory +java extension maps to the incorrect MIME type +on Swarnims comment +Thanks Ted! +Any hints when will be released? MirkoThere are no immediate plans to release +Tests passed cleanly for me on on Java and Java too +However you can treat this comment as an LOI +Patch for trunk +I am not familiar with the price to pay for calling the same server right away but if its not really high Im tempted to apply the pattern every time +I briffly mentioned the issueof SQL authorization in that jira +Closing just a Confluence change so not tied to a source code release +I am trying to see why hundreds of datanode threads are waiting on socket reads +Most probably one is included in eclipse and the other one is provide by the next week or so I will be releasing meclipse integration with that will solve the problem you are describing +Have removed thoseMuch better nowGot another failureAttaching the log file checked again for updatesPulled +committed to are ok +I have currently added unit tests in the class +pid file to check for process aliveness +has version in new as while trunk patch has it as +Thanks Vikram. +Awesome +Since a human cannot tell them both apart you cannot expect a parser to do it like that as well +Extracts and into separate files +Id it for backwards compatibility with ACS implementations +Added the interface but did not implement to get by what I needed to do for this is still an issue we need more info on how to reproduce this +This also applies currently to these listscommitscommitscommits +Moving to this is not a bug and would change the behavior in a notable s been a considerable ammount of noise around this issue so Im assigning to myself and Ive given an initial estimate of days +You just have to use the binary type where its required instead of relying on a bug +Looks good +Replace Lieferant with Verkufer for Vendor and with Grohndler for think too that VendorVerkufer makes sense in the context patch committed +The above stomp code doesnt set the on the unsubscribe so it wont actually unsub anything +Oh +Same patch for trunk +The patch updated tabs replaced by spaces removed redundant code in at r please verifyThank you. +Yes patch fixes the issue. +Here is a first shot of a compiler for avro schemas in ScalaThis is incomplete in many ways but I hope and believe this can be used as a starting point +There needs to be an advance zone for the add guest network button to be there +Work as expected with +for this kind of behavior we have Check components +That was the technical glitch +Why the rename of classes? This seems to depart for the current naming convention? the seems it should be only HDFS YARN not correctHave you builddeployed and tested it works as expected +We are running in a multi dc environmentThe changes in prevent the thread from dieing in case of a +Attached the readme file for the archetype +This made the compiler emit some unchecked warnings so the patch also changes some variable declarations in the build code to silence the warningsAll the regression tests ran cleanly with the patch +Also some globals that are sent to globals can be retrieved from site properties but agent uses globals like kerberosdomain hbaseprimaryname and hbaseregionserverprincipal +New patch handling another scenario of an update relating to previous update in the same segment +A quick try would be to set CCg before running +Attaching patch implementing Jonathans idea to record the CL replay position along with the sstable +With a customer and product store relationship in place it would be possible to create a view where a store owner sees only customer of her store +I would like to implement these with single selector if there are no better alternatives +renamed the test directory name in the junit test +BTW no new tests are needed for this help message change. +But this method is not called because the last parameter in the constructLink method is set to falseAs a result cookies turned on i have access to the session data +Field renamed +similar fix apply here too +First tests need to land in +I missed seeing your update +BTW the change will be available in the in a few hours. +So basically we need to do polling for branched CVS projects? +A common use case for us to to have directories organized by time yyyyMMddhhmm +looked at the stylesheet above it already had the code to generate anchors +Tobias jukka youre rightIt was Angela credit where credit is dueI think youre right about questioning the excessive passing of parameters around + +Thanks +Hi Alex it was already fixed in by the new hflush implementation +Thanks Michael! +They went directly into the bin directory before Ricks changes but now they go into the generated directory too +No it hasnt happened since so this JIRA can be close as you suggest. +To be reviewed by Simon but the filter mechanism works +The revision includes a new test case +Amareshwari what are your thoughts on how the user can specify which is a fact table and which is a dimension table? Or are you using storage based statistics to infer that informationIf the query involves join of one big table with multiple small tables and join is on different keys we can always generate single job +It should not be that much more expensive than reloading a searcher in the larger scheme of replication +In my environment the test does not run on C drive +launchtemplate to add some variables reflecting the changes in and + +patch looks ve just committed this +Updated to apply cleanly after trunk update +Thanks Milinda for the patch. +I cant this bug on my local box +Closing all tickets +Cant apply to trunk +Theres usually a logical explanation for the way things are configured if one just asks first +Thats a fair concern +It is also possible to use generics though plexus doesnt support it yet. +Maybe in our smoketestThis issue was triaged on July +Ive added some comments to the code to look for after the release +Ive attached which is a table dump that should reproduce the disjunctive delete table scan +Thats much more than just a loggingBest Im correct this is the issue in class with mrecursiveGuardTo avoid data lose as well as prevent not sure need to use a message queue +Fixes widget gallery display for widgets wo defined icons +Marking Patch Available +Can we extend this patch to solve the issue for as well +Looks excellent +Could you please provide me details? I can not reproduce this error on my local environment +ThanksShould we close this as wont fix +Go ahead and commit this +please make sure can work well since it is a public class +committed +on having a JIRA for on Hongs idea of fine tuning the policy +bq +for +The failure is happening right when we are doing the very first connection for the create of this database so hardly anything special has happened yet other than this property +looks good +On or only revision is published every time +Thanks +Patch v from review board +So now there is no need to stop meta information can be updated by executing command +Probably best to use a separate interface +This would probably work best for our hypothetical Depth iterations +If you look into the source code extends +This will be standardized in a future release +Beanshell script was changed to as a test case for conversion simple method performs marginally better on the large amount of data in the entityMaint screen +Rough patch +lets do or only +So I think we need to pull up the fix to +Moving to java Integration was little discussion and everyone wants this so Im marking it as resolved +I agreeBut we should do this in another issue +I will look into this more closely tomorrow evening +Im still having the same issues +well what would the deprecation have suggested as an alternative +Hi RayeesCan you please give additional detailsI could not reproduce this issueThe account got cleaned up after execution +The same comment is applicable for wait and waitPlease add cos and vinay as watchers to this jira +This will be different for each signal +I am fine with the original patch with no tests +Yep I get the same was using windows vista with finnish dates and locales but not sure if that is causing the M has the problem not just the standalone +New patch incorporating Vinods comments +I agree this is a problem +Fixed typo in enum +CVS doesnt natively support triggers +Ok I looked at a lot of my ppt files and the parser was functioning normally for ppts from as early as And I also discovered my crawler was corruptedSo I rebuilt it and ran the crawl over night and thankfully there were no Tika errors! In fact there were no errors at allPlease close the ticket +Ah thanks Yonik +Decision on how it will be resolved still needs to be taken and may not be for +here are my benchmark results with that patchMultisegment index segments M medium wiki documents no deletesTask trunkQPS patch patchPct diffRespell Fuzzy Wildcard Fuzzy Phrase Prefix Term Multisegment index segments M medium wiki documents with deletesTask trunkQPS patch patchPct diff Respell Wildcard Prefix Fuzzy Phrase Fuzzy Term executed on Linux with Java Version cmd java XX Xmsg Xmxg server +it would then roll back to and overwrite the files and Im assuming this would be a bad thingCurrently the mcur is stored and processed as an int giving a max value of +replaced patch with correct Daniel for the very good patch +trunk +Darn this happens on as well both login and logout are broken +Attached is a fix for this issueThe problem is detected by the jdbcapi test provided in when running in embedded mode +The patch is of course illegible +Invalid JIRA shouldnt have a target because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Again there are a number of solutions to this problem which very much depend on specifics of your application +Class that can be used to simulate the error +Unscheduling from as we can do this also for a later release. +a step by step on how to reproduce it would also be appreciated +Pull it back in if you think different. +Resolving as fixed. +This feature has already been checked into HEAD and will appear in Tool +I just committed some of the generic testing capabilities separately so that they may be more easily merged back to x +Rest can be moved to next release +Some methods already refer to State while others refer to so this change would help make this more regular +Ahh +I will post the final merge later +Reader? Hows that +The only use case I could see for this is having the bind tag in one JSP and the status access in another JSP that gets included is this really useful? remains as is for now but and all the form tags are updated to use +Committed +jar file with sources configs and logs +works for me +Dhruba were planning on a release very soon based on +This has been added to only. +I am just wondering whether users really need to override the importtsv mapper +Therefore will be closing the issue +and somehow use the same code to initialise the +And I really appretiate your efforts in building. +There should be tags with separate purposes. +See. +I was going to report this today +Newly created files get set to the users primary group on UNIXTY added that check back inbq +Committed to master. +Although Mono can be registered to handle execution of +So not all test cases will run on this database +Patch to update frameworks +Ill be happy to address them however the comments related to string parsing should probably wait until you get committed +would you please verify this problem is fixed as you expected +Looking at the documentation changes +Marking as resolved. +Thanks +Please close this Jira in case its done +Intent is to keep most of the API and the underlying block mechanism with its attendant block finding mechanism as well as all the metadata facility and end but in the guts of tfile thered be the and blocks of byte arrays only +i like the diagram on the Wiki version of the Web site a lot better +The report will document the various conversations Renato and myself have had regarding the restructuring to the module as well as the semantics for +Thanks for your help +Best RegardsRKOCS L Support . +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +Had to revert since it causes a regression where the errors view rendered a blank +Thanks Eric +test Debugger verify it can roundtrip all expressions in tests +The bug is triggered when the last hint is zero +Note that it is not possible to create nodes this way since nodes have to be created in a specific locationpath which is not possible with the approachThis does not change any of the existing methods but instead only adds functionality +We removed bulk API in. +I incorporated a modified version of your spec as well as sending a PR upstream to rubyspecrubyspec +Added a fix in javasrcorgapacheaxisencodingser +Applied the patch of into trunk and with thanks to Carlos. +Since we already have error code at the level should we get rid of the error code at the level? There is no obvious use case for the latter +Closing the issue as a duplicate of. +Found another deadlock +But my personal opinion ddmenu should not be used for that cases +I had the same issue but upgrading Xerces in tomcat endorsed fixed things for me +Sorry this was my fault. +The collections have already been loaded at that pointBesides that is hardly an optimized way to do it +point +I checked the server logs and it was evident that the cluster thought there was a node down somewhere so it couldnt come to agreement + st patch +issues +u admin p admin query +Patch for branch +Thanks +anyways Jesse I already sent you the patch for on trunk at revision on too at +Hello the problem here is that different quality snapshots on the same project have the same version which is not supported by Sonar +srcjavaorgapachecommonslang +Would this be possible? I think you also should consider whether all verification of acks could be done in a single method +patch to trigger the build and see how many files need to be updatedAlso I probably do not have time to update the patch until next Tuesday +It would certainly be the preferable behavior in our environment +For example opening help from the Catalog and Facility main page opens the same help file since the file is keyed on the component product not the web appI will have to replace the files in the attached and +Why not try to add the desired lower level metadata to the existing service as a patch to Hive since its already got pluggable backends and a server implementation already defined? Also could you better define what close to HDFS means? Theres a lot of HDFS metadata stored in the +In the latest SVN revision this has been merged into so that theres no duplication at least for endpointsHowever this doesnt solve the problem with +Chetan The annotations have now retention class and we use ASM to extract them from the class filesRobert Im open for using annotation processors +Reopen the issue because Jira sets the state to UNRESOLVED +Somehow my environment changed and Im not seeing any merge views although my configuration is the same. +It would actually make our XML processing a bit simpler too +Much of the code surrounding this is not new +Patch was created against an older revision but added the Eclipse files to with the release. +Unit tests are good +Simply encrypting the password is a false sense of security if you really have sensitive data you should instead be using https to encrypt the whole transfer +This is my first time using JIRA +but the version change didnt apply because it was already used +Jonathan is it possible to attach an updated patch based on your changes to branch? Not sure if that would be simple to extract +In the page ajax is setup so that when a is entered in the textbox the combo options would be to when b is entered in the textbox the combo options would be to reproduce enter a in the textbox leave the textbox to let populate the combo then change the textbox value to b and leave the textbox to click the combo here due to ajax delay the dropdown still shows s selecting one of the option would cause the error +In server side mode it is possible that some components on the not selected tabs may throw exceptions during decode because processDecodes is called on them even if they where not rendered beforeI provide a second patch for this issue +authorization is done at the outset when a client connection is first set up + +Thanks Joe +Forgot to say that this is about the module +I would go with this solutionIs everybody in with me so I can go and realize itCheers SolutionNow adding Alternative to the Producer and the producing class and the class is listed in the as an Alternative +I have attached a c patch that supersedes the b patch +This appears to be resoved with version release of fine in latest version +Patch covering and it looks like the nd patch lost the necessary mods to can you post a new patch that includes it? Thanks +Here is the patch +Sample Model Project Set for is a problem with the setup of the SQL Server instance. +type dand send the output on the box in questionalso find +On a node cluster with slots per node I run the testhadoop jar usrlibhadoop sleep m mt your patch divvies up the tasks +Committed to trunk and +can be specified in any of the following locations by default when working with a web application system environment variables system properties JNDI servlet servlet config you can contribute additional property sources at however environment variables and system properties are really good for this sort of thing +venkat +I modified the patch to check for null +It wasnt until the end of the testing cycle just before that I switched to using the +Links to earlier this to the undertow subsystem +If the config KV we want to delete is in another file were going to have trouble finding it +The patch looks good +Only few number of really useful images should be in plugin storage others should be dispoused after it use this is formulation which I got from these recommendations +This copy by the tserver is not something that we could rely on to happen it does not help fix this problem in anyway +reopening so I can modify fix target to +adding I +I will attach a patch soon +Committed trunk Revision Now backporting +it should be removed + revision +Enjoy +Thanks JoergWould you please advise when Maintenance version would be availableThanksAbhijit soon as possible and it will be announced +Its a separate wiki because its cleaner that way otherwise the docs get awfully complicated with This feature is only available from to and This document is obsolete for onwards please check +enter in google ledger segmentation and you will what i meanI will create a branch so you can see how it is implementedRegardsHans +We could address this if we restrict the encoding to which may be an implicit requirement of the current implementation anyway +Started looking at this unfortunately since does not implement +The patch was great +Sure due to phrasing I didnt come across that issue resolution +I have some troubles to process it further +The two jars should be places in lib +We just completed vote for Sandesha release and thats the end of Axis release cycle +Thanks David! I will look into this I was unaware the previous problem had been resolved +The root problem is that you havent set a source List in that case +References to jBPM were removed from JBDS Welcome Screen +Not sure hell help but it cant hurt I supposeIt is definitively weird that the cli and cqlsh would differ here as they should end up with the same call internally basically. +Reformat the previous patch +the response headers can be set till the first byte has been received by wrhapi from the body. +This ticket should be closed as fixed +This is the wrong architectural direction for derby we should always be looking to spread this work across many threads as does current implementation where work is done in the user threadIts true that everything happens in a single transaction controller +Im dropping the priority proof of concept doesnt get a major priority. +Uwe I can supply updated patch to yours if you want since I am already staring at itIn order to support reuse it will need to be changed a little bit but for now we can simply resolve the clearAttributes issue +Not sure it will get done but the latest version will be JenniferI agree that the SCA Spring Specs does not provide a clear picture on whether the propertiescan deal with primitive types +Yeah fair point +ok I see what you mean +An after the fact note I have heard from a few users that they could only stop a cluster using shutdown +So the implemented way is a much simpler solutionOn the otherhand we can parse an int to serialize method and an use it to generate prefixes for each serialization +Should be fixed in off release +To me it is not clear how the system will behave when the pool is busy with index merges and then additional tasks like these in the are added +Moving to +Anyway thanks for bumping this issue +I would close the problem but Im not sure if I canThanks Robin Taylor +A solution is to control the scripting engine to use by explicitly adding it as test dependency eg Groovy. +Integrated to TRUNKThanks for the patch Anoop +This should be tested with the existing not sure whether Hudson runs that by default but it passes on my machine +Do you have any dependencies that need to use a different version of ASMPlease send a copy of your POM file +I am an ass +the plugin should actually just be deleted from head. +Ill wait to hear from you before doing anything +This file has some examples of +Theyre not really +Nothing fancy for someone not deeply familiar with the codebaseIf you still wanna work on the ALS code we should discuss this on the mailinglist +Denis can you take a look at this one +Therefore pending further details Im closing this issue as Cannot Reproduce +added new method to return the that are defined in the configuation for the whole system. +Duplicated issue +Apache Felix Web Console Memory Usage Plugin has been released +I can take a stab at doing the work and post something tomorrow +Patch to fix the javadoc and findbugs issues +defer to +Here is a patch +do you have a copy of your journal dataI dont think this is a blocking issue +Is this second patch applied after I made the changes? Meaning did you update then fix the rearrangement problem then generate the patch +Wont apply cleanly until after goes in +Hi Sateesh Attached the files of yeaterday and today whcih might help you in s log file might contain the details you are looking for +A common mistake will be to attempt to process a data +Tested +We had to make our own implode algorithm from scratch for other purposes so we already have the code to decompress EXPLODE +Committed to trunk and +Thanks a lot +dont understand virtual read only schema +Id agree that people doing releases are probably advanced enough to know what theyre getting into +Merged into the master branch. +Obviously there are cases where you can change the logj adaptor and cant add a deamon howeverIm sort of skeptical that a DAG of collectors really helps you that much +Sorry I dont know what I was thinking about the unit test +In CVS +Please verify if possible +Trivial patch +Resolving. +is unresolved so I dont think this duplicate should have any fix version +LevelINFO et al removed in rev . +Do we need to update the metadataversion numbersthe change didnt change the format of metadata and the original code works with ip or host +You yourself state that a newer version of the JDBC driver solves the proves that this is an issue with the SAPDB JDBC driver and not I am rejecting this bug +Left the code that sets up the network server etc +Ill need to spend some time on trying selenium +defer all issues to +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Hi Samisa Thanks for committing code I looked on ant scripts of Axis C main build +Attached patch that fixes this issue for the english localization only +This seems to be resolved completely. +Running this test on trunk shows that this feature correctly works +When you have a chance do take a look +Unfortunately both of the patches suffer from a deadlock since the invalidation and fill are wrapped up in s locksT acquires cache read lockT acquires write lockT acquires cache write lock on missT is blocked on cache write lock trying to invalidateT is blocked on read lock trying to clone to fill the cacheTrying to work on a fix +KojiDo you need this in or? Per it is no longer available. +So either the instances created by Introspector have the attribute xmlName set and are found +I have seen the validation happening in the latest trunk code +It prevents concurrency +Here the patch again with the missing abstractI could also commit it if you like +if done right one should be able to support any archive format noThats a different goal +I also started with Twitters algorithm and at the time shared Twitters desire to fit ids into a long that affected my choice of suffix +If you can give me a way to reproduce then we can determine what the problem is +ok +What is very interesting is that the duration depends directly on the size of the array. +The only problem is the schema violationHowerver IMO the schema check should only be done at the end of the modification when all modification items are appliedTwo similar examplesst I want to add an object class which contains must attributes that dont yet exist in the entry +I just want to see the stacktrace of the original fault in the server log somewhereIve never run into a case in which I cared what the processing exception was +Agreed Brian + +Similar problem and possibly solution as linking +Fixed on trunk revision +bulk defer of unresolved bugs to +New patch and tests +Patch upgrading apache httpcomponents use patch for upgrading httpcore GA and httpclient GA and switching to +Patch has been committed to did claim to fix a bug so I owe Jody a unit test +Thanks For Junping I have corrected these could you also run all the unit tests for your patch +I think moving it into a public package in the post servlet is the better jcrsrcmainjavaorgapacheslingjackrabbitusermanagerimplpostIt inlines in the It also inlines and Perhaps they should also be made available +Marking closed. +What is the RWR code? It seems like the problem is happening at time +We should probably add two specs to rubyspec for these new tof discoveries +Thanks +I committed this in r. +I assume the fix is a simple Username asomayaj +Can we leave a for the total size at the start and go back and fill the hole at the end of compaction +a lot for the precise problem isolation +Its already well established that generate completes the string +Update patch to address recent changes on in v patch +For now I am just checking it as an alternative to avoid modifying the ANTLR AST to groovy AST conversion code +There was a module dependency that didnt have the latest version in the other words the problem was so this issue should really be about improving the error handling +Defer all unresolved issues from to +Targeted to +If subsequent attempts see the final job status marker then they should skip straight to the client notification process instead of running tasks +Requesting you to please review this patch +We decided to rename project and re implement it from scratch as much as possible but based on top of Pashas work +Lets discuss it orally +To clarify Peer One sends the message to a separate Topic for the purpose of sending Requests from which Peer Two consumes messages +Now that work is underway with AS all previous community releases are +It introduces an incompatible change in a key interface +Hi Richard can you attach that sample HDF file to JIRA here? Or point me to a URL where I can get it? let me know what version of Tika you are using looks like can you confirm thatIll take the above and then investigate what were seeing and get right back to you +If something like the class wo target host is needed it can be with the same or a better name +This issue can be closed +This issue has been open for quite some time and I really need a release +is what it is +The same bug appear again in the +Bugs in udfs can put additional load on the server making it almost inaccessible +Bulk close for Solr +Thanks Robert +This makes it sound like there some other behavioral change or bug +Work apparently related to the issues being discussed here were also discussed in and +We have reproduced the issue again with and threads doing unicast requests at the time of the leaveThe problem could be that first runs the suspect check on all requests and only then updates the view reference +IMHO we should notify we community about this +Now youre complaining about it in November? Didnt realise the explicit flush would cause you problems +patch looks good +Derby is not expected to work correctly if two versions of derby jar files are present in the classpath +BUT +Thanks. +Also fwiw this is not a issue +updated fix version to since patch is provided +The other test cases are for reference only they try comparable situations in which Apache DS works just fineThe test case has been checked with Sun Java System Directory Server and IBM Tivoli Directory Server in both cases it passes completely +Reclosing had to be reopened instead. +can I say b They should be able to use alternative packages sources for some of these packagesHow to achieve this if the alternative sources also dont specify any package versions and user does not know the bundle symbolic names of the alternative bundlesAre these valid requirements? If yes does the current solution meet these requirements? If not does it make any sense to have an attribute in addition to or instead of package version qualifier so that user has more control over wiring process +In that case this might be resolved as Invalid unless you think it merits an exception documentation fix etc +Im also getting this problem on RAD Error on adding indexing context central read past EOF No index update available for central Error on adding indexing context central read past EOF Error on removing indexing context local Cannot delete Ctempworkspace +If you have time and are able it would be interesting to know if you see this error when you upgrade from to +I am submitting the patch +A default value it true +Client sends data using ssl to mina acceptor and gets the response back +Where do you see the fileWhich package +sorry maybe not clear i am not discouraging changing the code i am just saying the risk worths the benefits +If the original author of this ticket doesnt respond in a day or two Id like to suggest we close this ticket and we can reopen it later if necessary +Thanks Alex! +This issue proposes to do the same thing Wolf this looks like a promising approachJason John would this sort of thing meet your needsIm not sure we can remove everything from trunk immediately +Chad can you please attach the source files rather than pasting them into the bug report? Pasting them in can introduce bogus line breaks that makes working the the files more difficultThanks +I am pushing this in +Ive fixed a broken test in the pull request and merged it into trunkThanks +now also defaults to omitNormstrueSo now starting from people will not need to remember to always say omitNormstrue for their own fields primitive types defined without omitNorms specified will still get norms as default if schema version patch with tests +OK I have checked in some revisions to this stuff +fixed in the trunk +I can not reproduce this bug +Checked in the closed with release. +Committed + +bulk defer of unresolved bugs to +Even though it works perfectly fine in IE Interestingly if you navigate to a different page then back to the portal it works in IE +I think its best to keep build infrastructure within the project itself +Let me do it before appling this patch +Closing issues up to jBPM BPEL +Thanks to Daryn +Thanks +Ted You want to check this in? Where? Into srctestbin +Good questions and a cursory scan of the code appears that hftp is not handled correctly +Will look into it this weekend +Initial patch missing test and test fixes +Your test refers to the persistence unit wharf only persistence unit available is access Person Stay is EAGER in your testDid you actually have those logs with this test +Updating Fix For to Unknown on issues not targeted for ve had a look at this and it seems low risk for inclusion in +Kindly regardsEirik +Not really if a cluster is defined by the UDP groupport how else the AS supposed to know that these foreign nodes dont belong to the same cluster? +Thanks for the patchDo you mind adding a patch for an unit test as well +Agreed that this is for all subprojectsI just created this issue to not forget about it rushing into the scr release +? That might be a better default since it does not change the current behavior if the user does not know about this update +this fix should also address the unique attribute for natural identifiers as well +What problems did you run into? AFAICTthe perl code looks fine +I really liked what you tried to do and want to make sure we get there +Attached a reproducer quickstart +Administrator can change the property values at runtime by the property values in +I could not find the specfiedinterface in the released source or the latest inhead therefore I added the interface myself and made sureit compiles +Thanks for all the help +Where did you get from? +Please attach just the diff of the changes you made +Attaching the release note for the for backport analysis +I didnt even say anything mean about rb +patch look failed tests do not seem related to the patch see +Adding a patch with a basic loadstorecaster wiring into the loadstorefunc and some adjustments to to help with the loadcaster methods +If you finish it before we release beta please correct the versionThanks +However I cannot assign this to myself as I dont have permission to do so in by adding a null check in svn revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +in previous patch merge with skipping conflicts method is not correct and here i have fixed itThanks +Marking this as resolved +But what about the other issues Ive highlighted in my previous comment? was just thinking about it if cactus builds on why do we need the other version for? Since this was supposed to be just a build problem that were trying to resolve right +Typos corrected +Check the Release Notes dependency is still open can this bug still be resolved? If yes then should still be documented as fixed in this update +tables and there is no correlation between the ingest performance drops and table splitting +My mistake +Why has this not been addressed? +Old JIRA cleanup +Oh your commit broke the build in the WPS module +Mattias thank you for the patch with the tests +FWIW theres also nothing in the JEE spec that requires JNDI reinitialization for every accessing thread +issue as it has been released as part of release +neither Andre nor I can reproduce in betaresolving this specific issue Stefan if your issue pops up again please open separate issue +In terms of its affect on Maven this issue affects anyone who compiles and runs many Jelly scripts over the course of their application +Mark as improvement because in standard cases this works good but currently its not overridable +I will wait for clean nightly results and backport to +The attached patch to appears to work correctly on a cluster +We dont generally increment everything just because a new release is out +Paths are always normalized this way +It fails to find the schema file if there is a space in the URL +I will try some stuff now +Gentlemen no copies here please +Therefore I intend to close it as INVALID +Checked in as submitted +I know you are used to throwing them around rather casually but generally are used to indicate votes of direction at ApacheThis test in particular is actually many tests compiled together to save the time of setting up and tearing down lots of jettiesIf you were to Nightly it in the name of so called progress I would break it up into many tests each under a minute and increase the total test run time +Ill be pushing the newer javadocs out to the web site this evening. +Bulk closing stale resolved issues +Ive updated my test case I split it into two modules one using Jackson and one using +Create any rule with literals and theyre all enclosed in quotes on IIRC all editors that generate their DRL from +This gets a little invasive because we assume pending files per manager instance in several places +Any suggestions +Im on the verge of finalizing this and committed some code few minutes ago +This one is generational spill big objects firstTheoretically should be very suitable for +thanks +Sorry I meant to say This is useful in cases where you want to define your preGoal in your SUPER projects and subprojects dont have any tests m attaching the updated that has the fixes I mention above +I am thinking along the same lines with other pieces that you already mentioned in +is this still a valid documentor is that ok if I reslove the issue as wont fix +It could well make sense to use YARN to partition resources for or other components +Linked bug report with related patch +Jacques your sounds more like gett better themes for the front end than allow the customer to change the front end on a whim +These are from the Firewire drive on Brutus dated July +Am I right in thinking that setting the retry count is new behaviour in the standard jBPM? If so then this now differs from the JCA inflow +Added a setter and some documentation covering the thread poo case +Thanks Brandon updated as suggested +Just make a patch that would encompass the change to all generators and I will put it in +Bug needs to be confirmed preferably with test code +Any work on alternatives should be limited to improving what we already have in place +Patch doesnt apply cleanly to trunk +please make this Junit testcase +I was only suggesting a nod be given to it and that crud pages not be written with reckless abandon +To unblock I think we should proceed with forward port +unused to all the ramifications of explicitly setting JBOSSHOME +added licenses add license to other files properties xml +Solution committed in r +So synchronization has been removed +In fact I find it puzzling that it looks under lib at allWell we certainly dont want to puzzle anyone +A good idea is to also add a unit test that emphasizes the problem and the fix but I will take care of it +All seems okay. +But it seems that it is not about our code problem but WST surrounded the get and modify calls for style sheet adapter to prevent the adapters modification while it is acquired in the cycle within the EMF model +I have Windows environment +My guess is that there is still a problem with correctlyidentifying and processing BILEVELGRAYSCALE images in +Patch applied at r +bq +The hibernate libraries will only appear when enabling the maven and JPA facets on a project AND the maven hibernate integration is installed +Ok the unit tests are only for standalone caches not as services +Jira is for reporting bugs and not for asking bugs +I created which will add and update the documentation +I just wanted to point it at because the patch forthat bug covers bothThanksMark Mark Thanks for pointing me to and letting me know about your fixes! First of all I just wanted to reconfirm that is not a duplicate of because the leak occurs on the very first call to Terminate +attach an example that reproduces the response and cannot reproduce +I do not know how this happened scary +The new test case proofs that +Max please with Andre +Bug appears to be in the test case and not the server code +The distribution of the JON plug in will be handled by the JON team +Closed after release. +Added some checks to test for valid object +It should be fixed by related issue +This means that you have to work at one task after the other +verified +also could you just check that you have the keystore file in the same directory as your applicationMaybe also try setting the full path name for the keystore likefileCProjectsmycerstskeystore +Committed revision thanks +this took quite some applyingI fully componentised it and shuffled a few things around +I cant believe I missed that +This issue no longer reproduces it must have been fixed by subsequent Collation work but not sure what change. +See related bug +When the user clicked Next items to were displayed instead of to +See also discussion at where we suggest the same approach but add some new features to the + this as wont fix. +So one has to use build script plain ant isnt enough +kbsec from machines easily met the GBsec data Chukwa is designed to process data in parallel with stateless transport +see my previous comments +I need to hit someone else. +This is fixed in SVN +patch applied thanks +Scott please skip the tests like this mvn clean package was able to recreate with the latest AS sources +Exposing may not be too much change +Needs some testing with the new code base +Can we please fix the sample for +If you send me the correct xml I also can try it against GWM +Well it boils down to how the client implements the specifications +bq +Attaching the patch on Chriss behalf +The IBM XML toolkit does use the ICU for transcoding +Please do change the HEADERSIZE into an assert +Therefore an RPC sent to all which only expect responses from all members local to the site. +I assume other code coverage tools work the same way in this is an example on how to use this in the file +QE can you please try to reproduce in RF this issue that was previously resolved in RF +Defer all static analysis issues until someone can be found to work on it +patch to restructure the single deployable factory and how it stores info +The patch is correct +I suppose you could have a configurable maxAge which defaults to a couple weeks +Thats definitely an option and probably the one to go with for the time being +But when running the test we get a about a hibernate class no longer exists +User retracted claim +I didnt follow that you could figure that out by knowing the impl name +These patches are almost the same as the original trunk patches except for minor changes to resolve a couple of conflicts onSo the steps now for applying the patch are Apply updated patch +Gah +we could eventually supported retrying the unmanaged am also +Sorry no I dont know enough of all the internals of Web to comment on why that error wasnt logged +Tests passed cleanly for me on +Change from counterGroup to was two things that I changed here +Increment the Counter in after applying the combiner which is currently impossible because the RPC Connection does not know the contex +I just want to set publication time to the future and entry should appear in the log only after that time +I would make the behavior of accumulo on this configurable +I will stay there. +I tried to standardize on one spelling of excludedgroups unsure why there needs to be two different propertiesTo make a testcase you can adaptmodify the test project in srctestresources and add test cases to +Patch to remove excess synchronization in +Derby doesnt have the capability to add a default using ALTER TABLE +Questions Create a separate artifact for binaries? Ship the transitive closure of our dependencies in a lib dir? Include Javadoc too? Any legal issues or Apache customs to considerI think we should find an Apache project to emulate +Thanks for pushing that issueWe are actually working on that point +like Nutch perhaps +It was added to the core confidence test package in MVEL trunk and passes fine +Committed to and +for the latest patch +Just to set resolution to wont fix +Attachment has been added with description Bug demonstration square should change colour when clicked +There is no proposed design so this is a Wish +Made changes based on handling which should resolve problem +Note that I have started a few builds manually in the meantime but the commit trigger still doesnt. +Thanks guys +bzip compression format also supports concatenation of individual bzip compressed files into a single filebzcat has absolutely no problem reading all of the data in one of these concatenated filesUnfortunately both Hadoop Streaming and Pig only see about of the data from the original file in my case +logic to determine the implicit type of query expressions especially Constant expressions +I did an inital commit of the source files to svn +Details and steps to repro can we foundin attached that was reviewed by Nick Bailey +Mmm +What it does when it reads the property file header it detects a geometry column and pass on the CRS as Hints to the converter if applicableThe then extracts the CRS from the Hints when converting String to Geometry object +Works on Tomcat and +It doesnt need to be an event variable +Attachment has been added with description Patch v +Would be great to have participation also +Your assertion isnt being executed in or +Above patch includes a random collection of thoughts on how the section onmailing list could be structured +Since leaves this wrinkle as a surprise to the developer might we want to document this responsibility on the part of the developer? Im just not sure where the best place for that might be +Moving items unlikely to be completed for out to back to this is quite important +buildclassesgriddevbin +I myself am interested in this issue since I have encountered exactly the same problem on AS CRBRegards +api needs to be dealt with asap +Could you doublecheck your driver? My test driver My test DB Postgres Mac OSAny word on this +Patch applied +Thanks +Only need to submit a patch once its ready to be tested for committing +Shai can you post your latest patch where itself is passed down to the collector +We created and whether or not it should be a separate or if we should rename the project is unclear +It is based on the Hibernate Component documentation since usage of both is very similarCould somebody add it to the wiki +It would be good to mention in a release note +That repo has all the extra dependencies required by IS above and beyond the core platform +Ive committed this. +Theyre pretty close +Closed upon release of. +It seems unreasonable to keep this open on the Harmony side +With this patch we will show the new job runtime as indicating that the job runtime is unknown +but I dont know exactly that is it +What Patrick describes in is not a transformer use case and copying a file doesnt feel like a transformation +Committed the change +Patch looks good to me but needs to be rebased a bit since we reverted +Ive fixed the in trunkkind regards of +Is this still happening with the latest release +Im sorry its unlikely Ill get back to this issue in the near termHow do I compress the tables? The schema is pretty straightforward some basic bean info ints varchars etc on each of the classes listed with join columns on the id fields of each +In order to do that we have to have it be released +Im not sure to follow each line is a valid json +Arent you using the scriptaculous autocompleter now? Anyway dojos seems indeed broken with that regardBest thing that can be done is allow setting their autoComplete parameter to false so that the textfield doesnt get we use the scriptaculous autocompleter exclusively now +Added srctestprojects +This is one year old jira +The RI does not do this +Patch attached. +Anything we specify here would be specific to Groovy and only Groovy +Do you mean to say is should throw an exception? I wanted to keep the semantic same as before it tells weather the name argument appears to be a region name or not +Binary jars are a bit of an anomaly in the clojure world +I think perhaps you and I were testing around the same time as I got these this afternoon as well for a whileLost connection means SSH did not succeed probably because it is not fullyproperly configured yet you do not have a keypair named root andor have not configured the path to the ssh private key file in +Fix is committed with regression tests +Cloves if foo instanceof fails even though foo is such an instance then this is clearly a case of class duplication +Which means the hunks may not have applied completely correctly either +Resolving. +Committed branch and trunk +Sorry that was optimized versionCorrect numbers M and M respectivelyThe additional optimization I have done to get up toI replaced with when checking for stopRow +Im also selfishly interestedin the inevitable stream of emails to from folks wonderingwhy Hive wont let them insert data into the table they just createdDo you disagree with this approach +This is only quick and dirty +My document file format is Microsoft word fileI converted Microsoft word file into sxw file formatI want to delete some attached file and attach files againHow can i delete attached file +It is not needed and I for one will not be developing the patch? Comments +Thanks +First test is for splitfalse second is for splittrue +Mahadev Yes the propsplits part is a little cryptic +Cannot complete this change now as doing so will modify the constructor parameters + please deal with that +That repo has to be reachable thru http the url has to be static it cannot be a trunk URL +Will get things working for maven +Otherwise its all good +I guess its not true that Rails depends on unless you run your tests with profiling enabled +I think some determination of is the node reachable so that can get good information from it is what needs to be factored into +furthermore since we are fixing this anyway the default one should be named EAP +Ship it +For some reason outputs the exact same text for different time zone +For DB you can use mysql instance on perfdb in qa lab +did you change anything on the environmentI can not reproduce after logging in +reassigning to greg since hes building the installers +I think what Ive written so far illustrates a feature request rather than a bug +I guess you meant git log oneline grep! +john assigning to you since ryan is on vaca this is a fairly annoying bug we were hoping you guys could help us figure out +jan set executable not for your quick answer BrettReports back first thing in the morning +This hasnt been fixed in CVS yet and the patch appears to be correct +For lock contention youre right +That is why online schema alter is disabled by defaultbq +Hope this works for you +Since the Cookie guide has already beendeployed at the site we should not wait too long with its revisionand updateChanges to the cookie spec look good to me +I have the exact same problem withWe use Sybase cache no query cache + on one of the injected members should break the cyclic dependency +Wont fix since JDK is no longer supported +As a workaround you could create a wrapped that catches and ignores +We have moved the debug options out of the template and add a independent page called Settings +There are going to be slightly different call sequences more or less overhead frames in between calls and sometimes even native frames might show up +Tested with build fromipgateway and ipcidr parameters are now displayed in the network detail page. +Reporter information too stale +Im on this +I dont see much point in this +My fault I think I was leaving off the install goal +This sort of makes our problem tracking easier +Does the null value show up in the admin console? Im seeing defaultPartition Cluster Partition Name +Patch rebased on latest trunk +Attached is the inputThere really isnt enough information here to try and reproduce this +Replace type calls for Patch contains replacement of type calls for wocompat and projects +Added the join page as well +Thank you +AsokanPS +Reopening for at revision +You may also just want to specify other java options and not have to copy all the default admin ones again too +Verified by Paulex. +Entry serializer adds CDATA elements to the serialized local entry configurations even if the source XML does not have any CDATA +Its aim is to create a new specification instead. + +I changed the to already in the updated patch +Forgot about the other drivers +Regarding the stability of the write pipeline the only concern is with concurrent failure write pipeline recovery +The redirect page now also contains a link to the target page. +Applied in r +Can you consistently reproduce it every time +Cannot reproduce it on after several weeks of trying monitoring +Patch checked in +Thanksbq +However I had some kind of weirdness in my build env that made it work sometimes and not others depending on whether eclipse was refreshed the module was installed in my repo etcCan someone please pull this fix and appt in their env? I would appreciate it +Was writing test for this +Maybe its better not to use the EU mirror then +any progress on this +Laurasection +This introduces a property which when set to rollingFile will trigger the rolling file implementation to invokedThis patch also removes the swallowing of any or thrown by the +I agree with Ivan +User will take this for granted and force loses its ability to safe guard user from making mistakes +Ive deliberately left the API for enabling automatic intropection rather obtuse as of yet so that the community can decide how best to make it accessible from outside the library +So marking it as cant reproduce +This one has gone stale over the versions. +Once the patch there is merged into the trunk Ill upload an updated patch for this issue +LOL +Done +I attached which fixes the test. +No need for immediate fix +bulk close for the release +Thank youThe Opera is a good choice +please be explicit in the comment that the patterns are prohibited +Hey Chris sorry for the late reply +Thanks dims it worked +In every new workspace we create a copy of the default user template from ECLIPSEINSTDIRplugins +Thanks a lot for the quick solution +Thank you for the review I will make both changes tonight or over the weekend +The new entry point for most options would be server properties +Have fun +As such it depends on which in turn depends on. +Doesnt crash for me on OS X either +we should then also align the names of the +Jon ready for your review +I still need to test JBIDE on my end and I need to include that Juno is required for it to work in the documentation +The test +Unfortunately this means it will also not be active by default for the Activiti Explorer +bq + +All web browsers are able to remember passwords and match them to user ids so remembering the user id is that hashing the password doesnt help since its easy to just find the hashed value in the cookies and send that to the server +I think dims currently working on correcting the build break +Per infrastructure thread domain will be good point +I will work on this but contributions including junit tests are welcome as always +external in corewebpageflowcore covers this. +Can I get an quick outline of the steps required to configure this where to find what config files sample config a blocker for SOA think this needs to be an update of the kbase article that was created for Can this be confirmed +The bulk load test does about the same Test Load and Verify as well +It is better to make the name changes +thanks Henri +Thanks for the patch +There wont be a release +are now using embedded version +Start AccelerometerStep +First drop of implementation of most recent design +Thanks very much Peter +Yes it works well now +Perhaps work similar to needs to be done on the TTJT +Committed in revision thanks varun +patch added +Need a patch for branch for branch +That makes sense +Closing released issues. +Great job HowardThank you +The question therefore is should we do this now for Jackrabbit or postpone this for? I for +In Teiid we added a import property for handling this gracefully +Ill create a new JIRA +Adding it to the security filters has no effect since you are bypassing them by using filtersnone. +Thanks have verified on Vista and if we comment init method on its works +Jonas we discussed this subject and decided that rather than a reverse feedback of the view on the model the preferred way was for widgets not to reset their values if the corresponding request parameter is not present +Thanks for the review really usefull! Regarding I believe it can be supported as well let me add a projection test to make sure +fixed in +I believe there are now not anymore files so I close this issue. +See for this client +Is this going to be done by next Wed? If not please reschedule to a later release +Could you just keep it open until I get more details +Russ where would be a good place to discuss this? Is it time to use a different tool for transferring data from Hadoop to elastic search? Is it something that is fixable with some decent java dev time +Thanks Ravi and Dmytro! +Post +Thanks for your path Xilai but please dont resolve the Jira while the patch has not been reviewed and appliedThanksRegards on trunk revision +Ive thought about some more +This is quite a simple patch +This is still a problem and I am not sure how to proceed. +Do we want to make both changes back to back? OR skip this one for the other +doc updated in each resolver +Huh what a quick support +Attaching the rebased patchPlease someone review and give your feedback +When an exception was thrown due to an error on my test IMAP server it tried to raise in the client thread +This issue was never marked resolved +Tareq has stepped forward to do document this interceptor +Weird behavior of the transaction annotation though +When they complete I will report back so that this can be checked in +Leaving open for now in case someone wants to take a crack at providing a patch for this +OK I reworked your patch to fit in with the new test suite + +Closing issue +Uploaded patch +Resolved in Git ID aaebebddccedddee. +Added a separate component and refactored clientserver to as discussed. +Moving issue to a next issues to next release. +But I think that option with using Memory with token is ok if we can encapsulate it properly. +In my mind there are still some issues w dev that are being worked on +So I think the implementation from Weblogic side is correct at least for themselves +The application programmer can do this if required either by writing a trigger or extracting the information from the JDBC driver in the applicationI guess what Im really asking is if storing the properties in the connection object is okayI have never used this feature before so please correct me if Im wrong +It would still be good to get some comments around specific use cases where this practically matters. +Changing this from a bug report to a request for a new feature +Yes Im sure for and it should be reproducible with test example attached here +tnx for inputa other workaround is using a UDF like in the example or BIF preservesinglequote +This has also translated to some astonishing performance gains in the overall performance of our app +this is a good test +expected HDFS file does not have x permission +Can we please fix this forthanks in the current SVN +has gone when array library was redesigned +However I am not sure I understand the benefits of implementing ZAB are +I agree with your arguments plus the fact that contrib simply adds unnecessary complexity +This could be related +konstantin I really liked the simplicity of this design +Great thanks for the heads up Bertrand looks like should be good from an XSD perspective +This functionality is addressed both when deployed in local mode and via the Hadoop Map output record counters. +Metadata has been fixed and replicated. +Hi GaryThat makes sense +I simply couldnt make sense of this issues have now amended this issues Release Notes entry and am waiting for it to become available on the site +Im glad youre looking at this nowFor more justification see comments at the tail of where Stack asks if coprocessor load failure would expose hidden data and the responseReferring again to the discussion on it can be ok to tolerate failure but only if that is explicitly configured +I probably wont have time to look at the final patch but it all sounds great based on previous looks and the changes you describe +Joe Serge seems not to have the time to test it yet +Will commit in a while +Resolving as no longer pertinent +Well there is simply no JNDI resource bound at javacompenvjdbc according to that exception +Therefore it will need some restructuring in hive code base as well to generalize these operators for in other projects +Any clue +Yes +bulk fixing the version info for and all affected issues have in is interesting +Closed as duplicate +FTL in common directories +Good luck +We could then avoid also changes to version of other classes that rely on Metadata such as Content et al +Dont know what to do about the two shells seems a shame and unnecessary to have two different ones +I committed it and clover in builds on +Submitting +Ive slightly worked on it yesterday and added code so you could run test coverage on only one +Permaine problem exists on RHEL x as well +Updates to forrest documentation needs to be done +In response to We are running which is also affected +mmm yes and my workaround is far from being one it cannot work at all +I meant +Marking closed. +yes shouldve been barbaz +managementicons +Patch for not for commitIve enhanced which is already used by etc +it can be standardized through the dialect just like Gavin wrote +Fix included in release. +Thanks for spotting these +Good point Max in earlier versions we were doing connect in constructor but we moved connect to a separate method so we can remove the exception +Meng thats good finding +Arrghh +all issues to the CR +One was a really ugly definition which is now corrected for Java EE +Willem this hasnt been committed to trunk +They need to be conditionalized based on whether kssl or spnego is enabled +Moving all old targets to for now +I suggest we let the star where it is now +Remaining TODO is. +The same thing happens with the C STOMP client +The top level application class loader delegates to the lower class loaders as needed +is the most important component to test if we want to get warned of API breaks in Deltacloud +i will compare dist maven profile with normal profile +Thanks man +However both of those mechanisms have limitationsIn one use case see a customer using a client that is able to set the when putting a message on the queue and then expecting my client application to obtain it when getting the message off of the queue +ThanksBest regards patch applied at revision r with minor changes thanks a lot for this enhancementVladimir please verify that the problem is fully fixed as you expected +Perhaps this issue should be changed to a documentation feature request? Here is how currently works When a producer sends a message to a topic using stomp and a consumer is not present the message is silently discarded +test results look encouraging +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Hi JonathanYes RANK is now an operator and thus a reserved keywordWe can add it to the release notesThe parser is definitely a bit rough and could use some reworking especially in the error messages so I am all in for it +Please try with a build snapshot +No. +When we finally abandoned our old search software and reclaimed its hardware we decided to use and a distributedreplicated architecture with a load balancer +Added a patch to include the missing Apache licenses in some source files and +In the current trunk the first datanode in the pipeline sets a timeout of minutes +How often is the collection modified +Fixed in must not contain the why do you close the issue so quickly? Its not applied to branch and its not even fixed properly in trunk +The bigger compactionflushsplit improvements should happen in that version +What minimum revision should I be using +Every hour we delete the excess of days +Hi Paul I renamed apache to few weeks agobecause theres coming from Sun JAXWS RI and I wanted toavoid confusion with the jars renamed to in AS distro. +We want to keep relying on it +may try the other one when I feel more comfortabletwo quick operations +I found out that it has worked in wicket im going to bet it is was broken by +Have a look at Shims implementation does this look rightI will submit for tests right away +However I had to install the bit support for the gcc which was not installed by default on my system to make this this raises a few issues Since PGAS currently is bit the total addressable space of x code will be restricted to GB is that correct? Is PGAS and similar? Source for the later seems to be available +But they should be able to coexist on the same slaves. +Are you using some kind of proxy +Back porting shouldnt be a problem +This would keep the scope of the changes as narrow as possible +That might tell you if theres some problem loading the direct buffer class +Given recent webhdfs jiras I think its safe to say its not yet hardened enough to be suitable for production environments +Setting the Patch available flag +Ive got it +I havent looked at the other democlients +substring fixed +AndrewSorry dont have any experience with that +I have attached a document detailing the solution I have devised as well as the source code +Jasha you recently added the links +Applied patch in revision Thanks Nandana +use the attachment of to was a bug in the AST visitor that the formatter relies on for finding list expressions +We found root cause of this issue PR sent. + problem with Acrobat +In this way in order to run Atlassian JIRA on Harmony it is enough just to edit the jdkjrelibboot file and uncomment the bootclasspath entry for +You might want to say which strings I should put inI spotted an inconsistency in serverquorum +Thanks for reporting this +Patch attached +It depends on what you value more lined up fields or standard UI +Ive been able to track this down and its also related to it seems like the region server shutdowns if its unable to connect to HDFS +The previous list wereandreeabmitarumshiskybetharddimdymjamesmasanzleebeckerpeistationspmurphystephenwutmillvkaggalweitechen +thanks +Manually tested +For future reference heres the command I used svn propedit svnlog revprop r very much Knut Ive made a note of the edit instructions +When we switched to and to hadoop all of the things standing in the way of jmx were removed and everyone should have it +Tanks a lot Levi +For the dtd I will try to convert that to an xsd +Committed revision . +Thanks for test running the two patches Lily +is committed +Start MS +Verified in ER in the admin console and JON +Anyway I will see if this really need to be in release when reviewing +Could you please test and then close this bug +Agreed +mysql has been installed according to the steps in along with the most recent snapshot +Correction Std deviation is high with patch +The patch as an attachment +Reattach key store +I can change if necessary +bq +So it sadly is not a put it in and forget it tool +Close out issues. +It just looks like a regression to me +Both failed for different reasons +Hi VaibhavThis seems like a specific +I will continue with this patch as described but give the community some days to respond to whether or not it can be included in this release +Added topicpartition info in a few places where this is useful +Complete output of jstack F mForgot to mention that the process does not respond to kill at all +After these changes all commands fail if JAVAHOME contains spacesCctargetbin runUsing GERONIMOBASE CtargetUsing GERONIMOHOME CtargetUsing GERONIMOTMPDIR vartempUsing JREHOME jreThe system cannot find the path specified +get from +No response assuming ok. +Maybe you need to add UTF also to some other Connectors +Right the table is Disabled +The class itself +Adding further validation to check the length etc would change the behaviour of the method which is currently clearly documented to only check for valid characters in the Base alphabetNot sure its worth creating other methods to add this extra validation. +Which may or may not be the case +That would be a helpful feature for doing something like this +Fixed for in revision . +We are doing no such thing +It is better if you can upload the on you build the source code in Windows +Hi ClausWhat do you mean with grant Apache license? We have the licence header in every file +The region flush is optional and we are not currently considering coordinating with the master while performing the flush over the entire table +all issues to the CR +I hope there isnt any issues with the Apache feather +I see many advantages not only for RC testing +For my understanding are you adding a new function collectarray or are you enhancing collectset to have a deduptruefalse option +Yes I was off by a factor of two +fixed in r +Enrols a Participant Enrolls a Participant +Unless we have tests that check whether we update serial every time we incompatibly alter the class then I dont see that specifying it does much good +the problem with using is that you would need a different config to run the same thing on unix vs windows +optimizes the top row seek if the ROWCOL Bloom filter is enabledAnd will optimize the top row seek if ROW or NONE Bloom filter is enabledSo will optimize all the cases +Ive looked through the Javascript code for the date picker Norman do you mind if I rewrite it from scratch? Ill probably reuse bits from whats already there and the code I attached earlier +do you try to search via main menu Search File +Attaching +This requires no development effort but doesnt really make it very easy to just connect and have it work +This is the of a lease recovery and is internal to the +This is basically the same solution we also have in Jackrabbit for handling readwrite concurrency +Thats why I named it +it would be nice if Ron could verify this is still a problem in CVS noone says this is still a problem I will close this agree this issue is resolved +This affects and trunk +to Chris +In fact the test was not always reliable for me in its original formThe patch works for me +Sorry it took me so long to get back to itFor compatibility should be off by default so the option should be pretty to enable itOther than that +Cumulative migration issues +Resolution already set to FIXED. +Thoughts +It seems to me from reading the spec and the definition of that this information is intentionally left out because it is a matter of binding +just writing this because this is the third time i have this kind of ticket and in all prior cases it was just a misunderstanding in how has works but of course we will check +Changing to CRITICAL since the issue that depends on it was bumped to CRITICAL as well +Patch file for trunkThere is still an issue with the reversemapping example not building which was there before the changes in +Unlink from +This issue is fixed and released as part of release +Im fiddling around with it locally +Thanks +As it happens I was also annoyed that gives to downstream transform no chance to deal with exceptions in its partsI fixed up the for our site to catch exceptions and forward them as SAX events +ClosingAnd the scenario with broken recognition of JPP runtime has been verified as fixed in +So here are the files that should be removed as they are no longer used resimgs all except resui +I will continue in the nd week of January +Using the latter method will not enable statement cachingBTW is in the package whereas Connection is inThanks +The following properties can be overridden via syspropsprivate static final String URLPROPERTYSTACKS private static final String URLPROPERTYPRESTACKS The previous property is now deprecated for following incorrect naming conventions though it is still respectedfor now deprecated private static final String STACKSURLPROPERTY +I just think is too specificWe need something more general to reuse it in the futureThoughtsLiyin +Surely this is a problem every app writer will face +Thanks +Like Stack said its a nice carrot to help convince users to try out +IE are small patches released every couple weeks or every couple I would appreciate if anyone knows any that dont involve modification of the schema +Fixed in r +This is true until we move to zookeeper that allows us to use the which will reduce the latency of fetching metadata from zookeeper from large number of zk paths +I agree that we shouldnt overall limit the modeling elements +Patch is not being built by hudson + to forbid using DEFAULT in UNION +This puts you in a very good position to understand why the dynamicallyscalable column is not being added to the table but it is being referenced by DL +The patch addresses both and +looks good +Tomas can you please try replace minimized jquery mobile with full version instead of removing it completely +Thanks for pointing it out +Changed Target Version to upon release of +for the patches +Mike has confirmed this fix works on the test boxes too. +We spent years trying to get out from under an untagged version of Avalon that no one was quite sure what was in it +My mistake should hv written in JIRA +Yes youre right +IMHO we shouldnt be using the EDSCH mechanism +But still i dont see any potential situation when you must send many headers with the same name +This issue has not been picked up please review and see if you can help fix the is a UI bug +This is a maven project for linuxThe is a shortcut command for compilingmvn clean assemblyassemblyThe script is a shortcut to typing the execution commandjava jar target srcmainconfig srcmainconfig that bad data is provided by a raw sax parser for this specific input file +If you do this you also need to provide a way to turn it off +Is this updated in the snapshot? I am still getting the failures after downloading fresh copies this morningI am trying a source build to see if things start working +Sorry the description of the problem is not quite correctThe unbind method does be called after the deactivate method is calledThe problem is the component is deactivated after the mandatory dependency component is deactivatedThat is to say a bound service for a reference having a static policy becomes unavailable before the component is deactivatedFor example component A is dependent on component B +If the plugin just adds its config to the end of now you have two environments blocks and trouble ensues +Bulk close for Solr +Triaged for +any progress here +The interop tests are moved into trunk and being run on the hudson CI environment +Regarding SVN should we hold off committing to the incubator SVN repository until the migration is completeKind regardsBob +resolved by SVN +This was marked for but the fix for this is blocked by +since it was working perfectly in older version hopefully something went wrong +Changing this bug to major because it appears that there has been a regression and a basic java hello now consumes x more memory than it used to +I am trying to do it in Eclipse from the SVN plugin +Do we need to say Refactoring of Castor JDOXMLIsnt it just a part of development activitiesAlso change progress of the product to progress of the projectAnd why not use for acronyms like +Throw so that tasks can fail for invalid fix seems important for a test ve just committed this +Closing resolved issues for already released versions +Locking down plugin versions is not limited to just the plugins explicitly declared in the POM it should be done for each plugin your build uses including the implicitly used plugins for the default lifecycle bindings +Thanks Leonardo for your fast response +Im sorry chris snow I post not enough information +This looks to be partially fixed now in Dev snapshotSite URL portalclassictestGroup URL portalgroupsplatformadministratorstestDashboard URL portalusersroottestBUT there is nothing stopping you from creating a portal with name users and then have a page on that portal called root which means you can then have a site URL of portalusersroot which conflicts with the dashboard URL +duplicate +GeronimoOpenejb M does not use bouncycastleYou may need to check out and build openejb yourself although I thought the jars were published +I will also post the problem to the users list and the solution for it +Upon rebuilding from trunk everything works +This is resolved for +bq +It captures it normally In job level blacklisting scenario tasks are continuign to be received +Rather it puts the into the same classloader that serves as the classloader for the behaviours to be verified +Thanks Gwen +You should add pydoc to init createnew and createforappend since its not clear which one one is supposed to useSure can do that +Seems to fix ij repro in this JIRA going to run tests over night to see if it has any obvious problems +screenshot? any entries in error log +just ignore classes with exact same fully qualified name and be done with it +Gary agree with Snjezana dont understand what facet settings you refer to since there arent anyIts just about having a runtime selected for the project that has the portal stuff in it +part of the the indent I realized that it is because you are mixing tabs and spaces +attach my patch allows you to run the task on separate JVM as below but the code is very dirty nowedward usrlibjvmjrebinjava Xmxm classpath homeedwardworkspacebin +Feel free to provide a patch with unit testsGary +Patch applied thanks for the fix Greg +If you have the BV but not a provider then youll get a warning in the log +I am also out tomorrow so will review this the day after tomorrow +This is a real case on kryptonite red name node + to trunk. +The patch looks good +Patch applied but I was having problems generating the commons site +I have changed path to richfaces xcss stylesheet in +Otherwise this becomes a monster issue +Please if it happens again +I just disabled it in trunk +you are talking about using EL syntax for substituting preference variables in template right +There isnt and inserting nulls via both INSERT and UPDATE works fine for me prepared and not andIs your Cassandra cluster at ? If yes then it must be a issue +Patch to check if the length of array from doesnt exceed the people are corrupting their key cache files with a length between maxint and? Isnt that ridiculously unlikely +I am going to close this bug report +I had my test harness configured incorrectly +Committed revision +I need the confluence user names of the people who will have admin rights for that space +Ill check it out as soon as possible +See attached +Thanks Alexey the patch is fine. +We have started sending pull requests for a few of the minor changes but our hooks for Maven support need a bit more time and effort +Good catch! Patch committed in revision +The remaining issue here is the Naming of Graphs which is an important issue in itself but should not be tied to the web proxy bundle +nd patch applied with r +Converted to subtask of this as Sriram asked to +From a functional perspective I think matters could still improve +Can you give me some instructions on how to reproduce? It kind of sounds like a problem with http chunking where Undertow is starting the next request without having read the final from the chunked stream so it interprets the zero as a start of a next request +applied in master Daniel I have applied the patch on the master +Most of filters I saw in the patch ignore this flag +That in combination with the trunk patch allows the test case to passThe reason the clob case was passing without the patch on is that the before and after columns were pointing to the same +Hi Would you please try my patchThanks! the patch is OK but out of curious the patch looks different with the jira subject? the subject said two methods are removed but in the patch two test cases are added while only one method removed did I miss something? I asked this for confirmation that the patch is right thing to commit +Fixed +We have configured Nexus to also use the M repo contained within our Sonar deployment +but lots of tests are still failing +I think that the issue is solved with the trick. +Thanks +Maybe next year at +Also Maven with version doesnt work correctly +This requires a testcase we are counting on the s reports to rebuild the state we should have a equivalent where we wait for minutes before launching new tasks otherwise we will trash the cluster as each new reports back +Also if the default separator for specifying table name and then column name is in pig then shouldnt the documentation highlight that as normally +is used for the old Uberbuilder whereas the other feature is for the Tycho build +Thanks for the suggestionJuergen +So I think my patch is right but it doesnt fix a test thats not passing without it +Does this have to do with git +VitaliThank you for the explanations the exceptions make sense +This removes the dependency on sort order for return from list status +I just committed this +applied +skip +on Lucene its natural Solr takes advantage of what Lucene has to offer just like any other project that uses LuceneLike I said though it may make sense for analysis to be separate +So all issues not directly assigned to an AS release are being closed +No IT +The problem this issue covers is case where a regionserver has say k regions and it so happens that one of these is over the store file upper limit +Wont make it for +Thanks for a review Brock! +no there is also fieldfoo in that case the compiler usually generates code that will directly set the field in which case the conversion has to be done by the compiler looking at this again my last comment actually makes no sense +Added unit test +The formating part was taken care by +I will open a separate track for and thanks a lot for your tests +Finally this seems to be a memory issue +Ill update a patch based on the idea +Thank you Chuan. +and retry +Another update to common up the logic moved to the AndySorry I missed a part of this +AndrzejI forgot aboutOf course I agree with OtisInnovations code didnt survive against threads I tried to use it because suspected performance problems +Logged In YES useridThis works fine with from today +Ok in I changed for user +It seems this same issue pop ups with anything looked up with jee lookups in spring context with anything fetched directly from JNDI +great job mahadev +A counter for number of records spilled to disk in maps and reduces is added with this patch + +Adding tests and fixes for this +I agree that it might cause some ambiguities +EAP has been promoted +to and TRUNK +Sorry about that +all wiki incorporated into new site +all issues to the CR +I will publish the short list at the end of the weekTom will you have time to add that last API to sometime in the next week? Thanks +Id like to see this coming back again too we liked this in the old +Guvnor Tools Tasks Adding new Guvnor connection Importing Guvnor Repository Resources Getting Local Copies of Guvnor Repository Resources Managing Guvnor Resources +Committed +Forgot to assign a fix version +I have a patch almost completeThe component however has a dependency on the component +I agree +Is this patch ready for review +On second thought i like the event handler support for null meaning dont render +Thanks +Do you really use Spring Integration and which version of Spring Framework do you use +Test show an performance increase compared to issues solved in released versions +Hi NicolasI wanted to apply your patch and close this issue being another task +Thanks! just commented this to trunk and branchBut Sid told me offline about a disabled test +There needsto be a more sophisticated solution to handle SVG +The state machine will be maintained and changed only by the leader for a partition +needs investigation +Patch contains full implementation of negotiated content types +done +The discussion is similar to changing norms through IR and removing that capability did bring up some hard discussions + has been marked as a duplicate of this bug +Scott I have spent some time on it but its not far along + +The uses a loose bbox parameter for the same reasons +As expected the Online accumulator does much better than the at computing std +Seems it was due to an old Node version. +Assigning this JIRA to for investigationtracking of whether the patch should be applied there as well +Im not saying that kind of thing would be common +After updating from sonar to sonar it is can close the case +Thanks Denis for the feedback +That way it will be easy to tell what was changed +Do you have a complete test case that demonstrates this bug +The classpath length is a restriction I hadnt considered +Patch looks reasonable +Also requires some fastutils classes +Again Im not clear which version +Integrated to and as wellThanks for the patch Enis +Both axes and bubble size are configurable +Attached are the trace logs when the slow down occurred +Daniel added Mina support +We should only do that if its an implicit block in the current callI have a fix and spec in the works +We have previously marked similar issues as wontFix +thats part of s already fixed +Is anybody currently working on this +The fix I found that MSVC compiler treats in source code as positive zero +It should walk you through most of your setup steps including setting up HadoopId come up with a mental model of how youll set up the computation Hadoop +One bonus doing such might make user easier in test mode where ZK is running on its defaults your opinion +I just kind of threw that together as an example +Suggested Patch verifies by checking through all children commit has been made along with a test case +This might result in a race condition +What the general thought here? Do this? Or increment the versions of the Put and Delete classes and add a new byte member for the cluster idOne one hand if its an attribute it means theres a possibility we can ask the cluster if it supports replication and in the case it doesnt then you dont have to add it +Thnx for the tip +Thanks for the feedback not sure how I missed the JIRA to contribute so will fork and go from there perhaps even this week +we would have an abstract choice renderer class to extend from instead of implementing the interface directly in case you do not care so this change would be pretty much transparent in a lot of casesi patched listmultiplechoice and added convert +By Beta you mean Teiid designer Beta +They are in the same name space in Java and if javac sees such name it is considered as the first component of package name +Id need confirmation from one of the mentors please +Fixed on Cellar trunk revision +This fails since there are no order items +As I mentioned any mechanism that allows whitelisting also allows blacklisting by definition as it is always possible to whitelist all the nodes but the one that one doesnt want +QE already have an setup just use thatJonathan +The problem is that the current namespace support is already a hack working for one parser only +I think the test may work only on either one of the osgi containers back in the days +Yes correct +Replaced by +If the region had been in one place for a while most clients will have a valid cache value so moving it will invalidate every client +Ill check it out +Canceling patch until we decide how to control throttling on the client +The permissions on the SVN repository have been updated to permit HTTPS commits to the repositoryThanks! Committing through HTTPS now works again. +Thanks for the revised patch +Convert bad file URL which derived from windows directory to unix path +I agree the UGI methods should be annotated public since per Harsh we provide documentation on how to use them +The git connector module does have the assembly profile but we still need to add it to the distribution module and check the actual output +This problem got already fixed in +Verified in JBDS +Adam I leave to you to decide if this is worthwhile +I think that we should have a version up and running at the end of next week then we need to put the old client side methods back in there that uses the new api but that should be too bad +in +It also removes the prefixes from the mentioned property names and use numShards as shard count property name +force unique table names per test +Closing issues before next release. +Understandable +The apr ones on RHEL and are links to the installed apr on the system and need not be in SHASUM so Ill remove them from also checked that the Hornetq ones they match on both RHEL and +I was able to get past this by recreating the dependent project as a Java EE Utility Project +Sure no problem +This patch fixes a potential problem in exception handling in Ill test this further once I have access to a bigger cluster +Hi AnilJust a quick review for now +Patch reworking. +Ill document this all shortly +Attachment has been added with description patch for LuisThanks for submitting the patch +Nope Im not working on this at the moment +It would better to add more +thanks works great! +I will report that as a potential enhancement for the Apache project that we might depend upon some day +We found the the order of http methods in hdrtokenstrs is not match the order in so I changed the methods order in hdrtokenstrs as the same order of the config comments +Failure in in is due to the order of the newly added testcaseIn trunk its not a problem as the master restart related testcases are at the beginning +The setting set to true user will see snapshot option for running instance volume +It would be entirely awesome if you could also contribute a unit test for this condition case to trigger this problem can only be caused by an incorrectly delineated or garbled HTTP message + removes dependency + this sounds like a good direction to me +This will fix this problem. +If you identify that it is actually a bug in Struts or get no joy from the user list then you can this later +Committed patch v with a minor improvement by testing multiple topics in +Im waiting for clarification from PMC what should we do +for both the trunk and y version of the patch +Im not sure about the Linux Java version +update database configuration in +this will also be following suit with how the Red Hat Enterprise Linux Release Notes are handled +New icons for CDI elements Bean class Bean field Bean method Injection Point Event ! Annotation +Are you able to reproduce this with some test code without a dependency to your application? This would help to analyze and fix the issue +Patch looks good to me +This issue did not result in any change to so should not be closed as done. +Currently there is no method for this +dims +Perhaps that isnt creating the patch ve attached the file +It probably isnt a good idea to introduce functionality that we know will soon be obsolete + just committed this +Similar issue was fixed earlier a screen shot of the stems from where the HTML columns were updated but the JSON columns were can be easily verified by running a single job looking at the RM UI apps list page using HTML clusterappsrowlimit and JSON clusterappsrowlimit fixed a small typo in HTML test failure is addressed by just committed this +Might come into consideration when Git tools mature in the Java space +h files Clownfish writes to perlautogen duringthe build process now get the proper licensing info too +Until now I only got a response from Arnaud that he had forwarded my mail but nothing more. +Thanks ve added Kris and John as watchers so we can move the discussion we have had in offline email to here would the suggested patch be ok for you or maybe something like that could be done as an extension point +Hm i never searched there +Will update and open issue in bugzilla +Can you please give me a day or two and I will put my patch on the relevabt issue? I think it is +push out to +Thanks for the feedback +Can you please attach the wsdl also +I am not sure how to get a thread dump of the spawned process though +can you add me by hand? please add devlistsA note to SF would be nice i cant belive its an issue only affecting myself i tried it from several providers every time the email bouncesThx for adding to user maillist +Attached screenshot of client +Moreover I have seen in the compatibility matrix that is compatible with but I guess that this is just a SP is also compatible +Otherwise Ill carry on as Id rather produce a tactical solution that works on Monday even if it gets superseded than wait another year whilst you try and get a change made to Eclipse +Can you remove it and see if your tests work fine? If they dont please attach any stacktrace you get +I will look into this bugHi Stefan From the stack trace it appears that a base table has become corrupted +Committed to trunk for now +Cool thanks a lot for your effort Daniel Christian and everybody involvedI should be able to try it in the afternoon Ill let you know ASAP +Well I may just as well try to show it another wayI have pesimistic lockingselect v from t for updateupdate t set vnewVNow I want to replace it with optimistic lockingselect v from tupdate t set vnewV where voldVthe problem is that this can failUsing delta updates I make it so that it cant fail because it doesnt depend on the current valueselect v from tupdate t set vvdeltaVI hope this make it clear +Just my two cents worth +is that something working with mvn and not with? +I even out the code and it still fails +It looks like we need to add specific base version support to all of our connector template names +Do you wish me to assign the issue to you? Are you able to port this to the APIRegarding the patch I think the tests you had added are certainly valid and we can add to them if and when they are required +This is most probably a corner case being hit upon since this case of can only be thrown if the bundleChanged event is called after the method has already terminated when the Configuration Admin bundleThis may be the case when the framework is stopping the Configuration Admin bundle while the framework event dispatcher still has an event pending to be delivered to the of the stopped Configuration Admin bundleThe fix in fact is rather simple I add a flag indicating whether the method should actually be handling events received or not +Im going to commit this since its currently an inconsistency even if its not the bug +Im will close this ticket and open a new one against Archiva +Language removed from spec +Its completed in and one reached the sec +Unsetting Fix Version for unassigned issues +complex columns are broken in the query because the table alias is optionalPlease file separate for those +Move open issues to CR. +Thanks Sreekanth! +but this would be a low process checking all jars in all plugins +and esp a that describes what this is and how to use it +Receive of into +Could you send us a test case +With that said Im open to suggestions since I dont think this is going to do what I want it to do +And we need to look at this JIRA also in the light of saving all state possible to the to work beyond RM restarts +Thanks a lot for the contribution Colin. +see and that seems like a nice solution +When we catch a block we throw it intoAt one point does stuff go into the offheap cachebq +Well sadly we have no continuations +yup it makes sense having it reopened until addressesIll try to take care of it with Simone weve found and fixed the bug +Patch that should fix this issue +Ill get in touch with Jeff when I get back from vacation and discuss this with him +bq +Patch that fixes this issue +If we can improve the docs then thats great +I committed the above note to trunk branchx and lucenesolr +Nice work Maja! +I ran a few jobs on a cluster +It is a bit surprising restriction while building a sample war file +My October th svn commit should have been against not +Move the majority of and blocker issues to the next point release +Perhaps we could trim some stuff from them to make that easierIt would be better to use CDATA sections so that we dont need to escape all the angle brackets and also make it easier to compare with the core sitemaps +This IS a bug +Fixed r r where shortcut looks for condition so that Platform Documentation shortcut is not made if docs pack is not chosen for in EAP CP CR and EAP CP CR. +Thanks Ravi Teja! +Im repoening this because I want to kindly ask you to not use this type of aggregation +yongqiangNingDo you think we can do these things in jiras conditional task output in explain and union struct object inspectorIf we want to merge Johns patch also by Tuesday we should get this in +I dont know what the original intention of this value was so I left it in for now +r r r rClosing the issue again. +IDE then in Fuse IDE useWindows Show View Error then you may have entries withFailed to connect to JMX No could be found in the has that stacktrace pasted above +Following are the configurations of my machineMemory GBSwap GB Dual core processorsUbuntu We are testing in low resources machines and will let you the results +Closed after release. +I was able to clone and added a simple readme which I was able to push upThank you so much +do you know why it conditionally decrements the position and then increments it later unconditionally? This was done in to prevent position increment from becoming negative which messes up positions if the field has a payload +I am removing the Patch Available status since there is an open request from a reviewer +Add unit tests and extract String stuff into aShould I once again split that into two patchesThe first one would be only about + now returns the entries that correspond to the modules specified in the first stack +For example what if I want a view of design docs with a view called bytags +Added a patch file +But what about wrong mappingThanks in anvancewith respect this in fact still a problem? Lets fix it for if so if not lets close it +Jake has offered to help with blogs admin stuff +From the file The classloader seems to have problems with inner classes because the the inner classes is not transformed by the which caused a +The next available number on Ubuntu RHEL Debian is +SVN R +Hi MilindFor your first comment yes users class have to be +Did we fix that part already? mingzym can you make sure we dont have duplicated bugs here +Thanks for the suggestion and codetest contributions +on patch for and note this may change even prior to because of changes coming out of the header. +There are tests and all tests are geen locally +Visual inspection of build output is complete and has been completed. +tries local tests seems to run well but get stuck at a moment in the middle of the tests +s HTTP is always seeing the correct encoding +Simply adding useStreamtrue in srcorgapachewikiutil for Tomcat fixes the issue for me +It is very critical for us so I was wondering who is looking into this? How can it be fixed? ThanksAmy transition resolvedfixed bugs to closedfixed +As Freddy asked me in attachments somes content of several elements +Lapse of attention to detail from myself +Closing this out +These dont exist in either the trunk or the release poms +Though I uncovered a problem +The key issue is not to introduce a tuple with less items in it + +Patch with a testI couldnt exactly reproduce the error in the test case but without the code fix the test does fail albeit with a different exception which still indicates the same underlying problem of missingThe test passes after the +Thank you for your contribution John Kew +Committed revision Thanks Koji and Noble +Ive looked at this a little bit and maybe we should just drop the entirely +Thank you for the reminder +looks +Verified that transitive deps are resolved correctly using the same small smoke tests and one large build that were tested for the the patch +This will not work because while logj supports TRACE apache commons logging does not +We are running a lot of test and now we cant get the same error +Tomcat JPA example deploys to plain using the ant jboss build +Besides that I cant think of a clean convention over configuration approachAlso and I think the current allow you to specify a property to set the default for everything that they setup +As I recognized days ago there seams to be a bug with returning sequences +TrustinDo you have any preferences if this should only get added to master? I think we could add it to branch too as it is an isolated feature and so a non risk changeWDYT +was closed as done must have been an error since no code was committed for it nor does the issue say something that indicates something actually fixed. +I committed this + +Could the footer tag be extended to allow column tags to go in its body? These column tags could output any total data +So my pull request is fine in that it fixes the symptom +Well a similar problem exists with XML configuration but it isnt quite the same +For example see the comments in +just pushed this. +Reopened to address the issue with the event brought up by Bernard this does not solve much +I think so +I also had to deal with a lot of cases when people just forgot to upgrade their default configuration +Is there any other library that you support for advanced tables view with paging sorting export? thats what I was looking for +Alan good point +on commit +Also methods comments are converted to to provide a guide for aspects test developers +I didnt mean to change the structure just the sort of instantiation of the from lazy to eager so no need for the method get to be declared as synchronized +Connection handling is not thread safe therefore no connection in constructor. +Thx +Accidentally hit enter during creation of bug please close this one +I cant reproduce it with latest sources +and I will fix +I cant see how this could affect anything unless threaded execution were involved +With the current architecture this cannot be interceptor must be attached before the Tomcat deployer enters the create state for a +hamahomeedwardworkspacebin +net webservice asapYou could switch to using WSSJColm +Thanks Claus +This is the last plugin our project needs to enable parallel builds +Can some kind committer please look at this one once again? The same unit tests are failing for most other JIRA submissions too +AmilaThe last change you proposed preserves the existing functionality and provides the functionality I requested +I forget what happens with chunked encoding but it would not surprise me if that also caused problems. +This may be helpful to others who want to test the problem on their platformsI can verify that the hang occurs on the Mac version of Java and Java +I think it looks ok relative to the trunk patch +These are the custom classes I used to solve the restrictive nature of +now i found the cause of this trouble there are two version of spring and +The test in the patch fails +This has been discussed many times on other and forum threadsPaged messages have not been routed yet +Im just pointing out that these components behaves differently I agree this has low impact and requires only simple workaround do not use no but facet only +Closing. +One workaround is to use HQL +At a guess the click event is being sent to different targets ie the area and the square that is being drawn +I updated my plugin to work on cordova +Thanks Sian patch applied at repo revision r +Committed +Im not sure how this should be handled on AS whats evident is that service mbean does not exist +Documentation patch that explains how searching worksCovers both Google and Lucene searchPlease review feedback appreciated +This would mean that they are actually compatible and need not extend + +If you get a chance could you take a look through the attached patch? Your feedback would be great +This patch adds missingValue to the lucene and then passes that on to the +please take me off the mailing listThank you +Upload the revision is the latest patch committed. +PrashanthNow that is committed to we can commit your patch +We certainly dont want them to start forming bad habits of relying on Xerces internals and have been careful to keep these samples such that they can always compile with the JAXP API alone +Submitted last missing patch for so hope it is now covered +Other than that Im not sure about others +Yes it would be really helpful if this could be it would be really helpful if this could be fixed otherwise it will using a different javascript library. +This is much more clear +The locking is very obviously broken here +facelets uicompositions files hierarchy was added +I think a copy of this source file is ok +Updated to patch latest trunk +Well how can you pass in branchId to all the time? In merge you can do that because you know youre merging a branch but in regular commits to a branch the only way to find out whether a commit belongs to a branch or trunk is to check its base commit and see if thats part of a branch right? Thats why we need that read +done at rev Thanks Wai for for release rev for release +patch applied in r +Look at uses of for a similar change +thanks henry and flavio! +This is not true if the pool has a function which automatically puts a connection back into the poolThis is what this pool call an abandoned connection if an application do not pass back the connection back within an specified timeSince not all programmers are perferct and not all programms are perfect this is a nice feature to prevent an db connection leakBut if the connection is passed back to the pool the object passed to the application should never ever be validWhat if the wrongly programmed thread which connection is passed back to the pool issues a commit while another thread is in the middle of an transactionTo the performance problem If one uses abandoned connections heshe has to live with this performance gain +Im testing against C service and this is working for me +Except Notepad +Code looks good to me +Damien noted it was just missed when setting up the website changed to an appropriate phrase. +Ram I dont see any mentions of in this JIRA except for your comment +also what we have found out is if the server is recycled after the remove then it works fine +Thanks. +I dont end up with instances of the framework however my mvn dependencylist will show me both dependencies +Moving to +Instead of catch Throwable and set the flag after catch block can we simply set the flag in finally block without catching throwable +Closing as duplicate of which has been closed as well. +At best I think we will wait and see on +sure go for it +Should be cannot over load getters in a sub class +My index rate is high and some times my boxes are under some pressureMy problem is that I dont know how deal with the situation +So this is really an openJDK issue +How large is each block +It is resolved not in but in by Peter Royal. +At some point the amount of data presented to the user gets overwhelming +bq +Nice! Since Lukasz just resolved do you want to take care about this issueMany thanks in advance +Roger was going to tell me how to get GF and install it but it will be important that use the latest Weld version with GF to test this +updated to use the to get teh +I agree renaming would make things clearer +I will take a look at it +It passes the two tests of and +I dont know if can be made to work with Glassfish since that level of server is Java EE and is uses the JPA API which is Java EE technology +Once you mvn clean installed multiple you can give a try to it to build locus using mvn clean verify P multiple at the root of the it works its worth a you code committed to your repo so I can give it a try +Patch to move to latest and use its +Done already now has +SIMPLE is still supported as it currently is today +My first question is why do we need another send mediator? We already have two mediators that enable sending messages out +The user wanted to limit the number of those tasks running to less than the number of cores +reformatted and committed thanks +Filter definition definition parameter for declaring multiple for to allow for annotations on a persistant class +Others may appreciate the option to be extra fast + +Its in the directory but dataimporthandler might make more sense +Or you could just throw exception warning user they need to add at least to make it work on +Let me know when youre done and Ill do a review +A subclass in a different package can see protected methods but not methods +John do you have some time to update the guide to reflect the changes that have been made to Woden +Ive created and committed all the tests but now I really have leave able to reproduce with you still cant solve it I will spend some time creating a standalone test to reproduce +Seeing as the is such a bottleneck during startup I would really like to see this as a subclass to and then use the property value to determine which class to use +WernerIs there any information you need for this issue? just a bit more time +on thanks for the review +Then I am closing one again +Maybe we could close this issue and suggest using arquillian? +logos +Happens when an interface adapts to one of its impl classes +Not a bug but user error. +Ok here is another patch with Mahadevs suggestionsAppreciate any feedback Ive tested this pretty good +We are now packaging in the jar file and when they do require pigudf the class will pull in all of the required filesI have no problem running the tests without your patch so Id like to get to the bottom of that +looks goodwill merge if the tests you remove the new table nzhangTmp in at the endSome of the tests are failing because of that +simple fix +I think you will find it easier to get stuff in if your refactoring is related to the issue +Updated patch with the null check +Do you think allownodescriptor would be better? Or do you really think that using descriptorrequiredoptional is really that betterThe more I see it the more I like your suggestion but until now we have often used boolean values +Im attempting to resolve the issues Jitendra cited by breaking the patch into multiple parts for easier reviewdiscussion +What exactly is the purpose of a tidy serializer againIf there are any problems with the current HTML serializers Id rather see themfixed then hidden behind an additional layer +Committed the new test to the trunk as revision Thank you for your contribution to Derby +Committed to TRUNK +Testing shows that the code executes the correct destructor calls using the workarounds already in place in the Pointer class. +IMO having entries in a filter is not usable and the first problem is here not on the pagination itself +In v patch I included following changes make sure the min number of required RS is in par with other integration tests +The drivers require or above for the changes to work. +This bug is potentially similar to bugs and but would rather not mark as duplicate as I am not familar with the underlying code +Hi AndreasIn it talks about a Stream Closed error +Verified in +HiWe did some analysis of the data loss probability in HDFS under different block placement schemes and replication factors +I had to switch to the sun implementation to get things to work +I believe this change isnt controversial compared to others IMO +Accepting a configuration object will bring in a lot of junk +Because Im using FVH to highlight terms in fields and these are not ideal for +testcase is set in onrequeststart it exists in see attached ll attach a new test MichaDo you need any more info on this issueA +Yes +I changed the behaviour of Nestable but left alone +This means that you dont just need to check every store file for what it thinks is the closest matching row +I think this should also be in the procedure code should be the same +RonnyAs you resolved the issue I believed it worked for youDid you given the correct rights to the apacheds user or did you changed the user in the script to rootIm curious to know +Somehow the problem has gone away +this looks good guys +The is a new addition of mine just recently +The other attachments are now obsolete +This is a copy of an ammendment to bug in the database +Fix with test +Patch applied +Daily at a minimumQ It would be a lot simpler for us if we dont use the sonic NS service we now use who provide a far better NS service than we had when this ticket was openedA The hidden master should let us just list their NS and allow transfers in bind so this can be a debate for a later date +At this point I have a unit test that replicates this scenario and fails at deployment time +Yes it can consume a memory +Are you happy with the name or do you prefer I prefer bq +The attached file contains the fix and can be used for verifying the solution its a replacement for the same file in the system folder on a FUSE ESB installation +So using a view would simplify the code in at least but possibly more placespierre better not complicate this subject with other subjects +Addmodify tests to verify the above +Initial patchThe approach is nice and simple All tests pass but I put a bunch of nocommits in there and I stopped short of the stuff Ill have to redo after committing ooo merges +Using PMD Maven reports to inspect Sonar codebase is quite unusual! We already eat our own dog food by using sonar on sonar but feel free to attach your patch to this ticket +I noticed when looking at the update that I provided that I had neglected to remove the log messages +Will report on the results +Otherwise this is a major open area in AM in these branches +To implement this we need to create a impl and inject it into the ssh client +Isnt it amazing howmuch trouble it is to retain ancient features of a language such as the abilityto reference columns by ordinal position? I wonder if the SQL standard defines a process fordeprecating unwanted features over time +binary search sort +The filename should be in lower case +fixes it but breaks some IT in the +Very elegant fix to the problem +The only thing we can do is to close the issueThanks by submitter. +If it can calculate the cost based on measures you need some real examples that show it +We need to upload icuj too +Going to test again now and confirm +Yes the wourkaround is to use the static scaffolding using the generate voters already for this issue so I have the feeling a lot of uses think this feature isimportant. +Resolved via svn revision . + for the patch +Altered hiearchy of component to enable EL in all common attributesCleaned code get rid of all inherited propertiesAlso altered example to prove correctnessRegards Zdenek +Again the cause is positioning +gdbinit fileMaybe we should also provide solutions for perl and python debuggers too +In essence we would prohibit users from registering more than one user ID per address +Fix for closing insertTemplate tag in the description + +Thanks +Looks like trunk is OK just gets the server location from the which is the custom +I gave noticeI know +Thank you Patrick. +It does not seem to be a duplicate as we were able to reproduce the problem with the trunk that already contained the fix for +m not sure whats the right way to fix this since we cant have dynamic metricsWe can +Ive created a pull request for modellogithub with a fix +patched against trunk +is also available on windows FYI +Thanks Chinna! +Patch attachedThe problem was that the common view was always assuming there will be a filter textbox and filter columns +I think the outstanding repos are whiteboard and external +Should be fixed in development mode as well now. +Patches applied +It doesnt sound like we have agreement on this patch yet so cancelling for the moment. +This needs to be discussed in the forums before hate being presented with solutions before the full requirements are discussed +commit iddadeeddecaecffbdbeafI still see the issue +The underlying issue was that when JNA is not enabled we end up shelling out to execute the readlink command +publishtrue by default +This patch fixes the incorrect use of the condition variable in librecords +Still awaiting migration of your svn tree to top level +Above patch carries the from the original patch +HelloPlease use the user mailing list before opening any ticket +I removed my surefire report zip since Elliot also had added it +contains the features to handle the BPEL debugging +Oh +But still end up getting HTTP ERROR Any solution? Please helpThanks in advance +This is a nice improvementAfter this baked some in trunk we should probably port over to +Nope +Yes this needs to be done before +Run another analysis +All tests pass with this patch +I just committed this +Can you please also attach a patch with the actual driver code and testsThanks +I wonder if the pending improvements to will handle preventing this FC insanity +The changes are +Patch applied +Can you attached the +Unit tests are welcome +The patch contains updates to the and a new file +Ive it to let it not get lost under the radar +ok used the attached +There doesnt seem to be a way in JIRA to change the project that a bug is assigned to so I am taking ownership of this bug closing it opening a SOA issue and will link that issue to this one. +You can see how it calls rejoin recursively +bq +Test case working for +We can edit the refguide to remove explicit javadoc building as part of assembly after this goes inRelease note dev vs user api +Good point +Thanks for the patch Marica! + updated +can not test patch will not apply to trunkwe should get this in to f its showing good results like stack reported above +Closed no change needed in CP +I havent started yet +Recompiled vts with gcc problem is fixed thanks. +so it should be finite +Wont fix sounds good to me + +The change I made should work but could you please test itAlso as a side note was there a particular reason that you were using the code method especially with a single character? +Heres a patch that implements paging of both groups and docs in each groupstartrows for groups for docs within a I just checked in a patch for this +This patch renames the example requesthandler to dismaxexample updates the outdated comment with more proper reference to and switches to edismax as default +Doing so will also serve to break users who use wildcard addresses of which there are many +A first step is done +Most of the credit here belongs to Elliot since I just tweaked his code from a bitTested by running the unit tests and also starting up a Namenode and jstacking it to check that theres only a single thread +The patch solves the problem for me +I tried in MS SQL and this is not supported there attached test case reproduces the bug in but the bug does not show inThe test case is almost the same as that I submitted for so there may be a linkI note that previous posters noticed the bug in releases before so I may have found a different infestation that happens to show the same symptoms + and on what Lars says +Page from Encana There are other pages with comparable problems and these problems are undoubtably related to similar problems with ChineseJapaneseKorean Type fonts +Thanks DanAgree with changes have not concerned about these things especially concurrentAndrei. +can we use the build time generated jsdl beans? I dont see any issue with that right now +My only concern with doing that is that should be up to the developer calling it to clear thatWould it be advisable to have a key password in addition to the store password +I will look into this tomorrow and see if i can find something +This patch fixes this bug and specs wont be fixed until theyre updated the the latest version from the rubyspec +Now I understand your explanation +Thank you! +This patch doesnt affect the serialized format at all +everything works fine with version +Josh can you add this to the online docs for GA and then pull it into the docs for the CP release +bq +I thought of making PVM still experimental +sig files +I noticed that the issue is quite old and decided to implement this new option +only changed some message no additional tests just committed this +I would be okay marking this not valuable enough to justify the potental +Any idea +I executed times and did not see the error +Question why service is need in the methods ofMakes it convenient to know which listener is sending the call back +But do you think the information we are sending could be useful for the client +Prefetching javalang looks more like a workaround to me though +tested! +Could you post your struts +As we get more of the subtasks we will add more details to the document in +Both XML Spy and Visual Studio agree with that the unique constraint is legal and that the document is invalid +However because so much would be changed anyway moving to a different testing framework might make sense if there is something better out there +Bulk update to fix version +for target +I noticed theres a serializationhelper +cql documentation needs to be in r +This feature also includes some UI refactoring which aims at complying with the new design rules of createeditdelete operations +OpinionsMAXBLKEXT I think would only be relevant if you took the alternate approach of trying to push the remaining data out to the client when splitQRYDTA is called and not save it for later +um +Adjusted the Fixed in field +But that is another story that Im working on +At YARN side does not unregister +Fixed +All tests pass on AS snapshot! Yeah! +Send me that via and Ill try to parse it in staging. +HiI was expecting this fix to be available with release +Essentially this makes sure you new document does not start with a value for skuThat wouldnt work for fields though rightIf we keep this option perhaps we should find a better name +Verified on JBT. +uses and target as the target for executing the junit test for the hadoop core and contrib respPlease change the ant target names in this patch accordinglyTnx +has the format of consumer current randomUUID +Untargetting +but it all comes down to when is deemed better than for distribution +If you are able to run them they finish before i get back on results please feel free to commit +Since this bug is already fixed I just want to give my thoughtschange LDAP grouprole to organizationalRole which has a multivalue field roleOccupant which is a MAY field and not MUST like in group +However there is no CR listed in the Affects Version so I set toIve also upped the severity of this issue because it would be a critical issue if push can not work with AS +The patch file +I thought was not used in MR mode so Im not sure if we can push the Foreach into +Good point about Hadoop +we should push for it at eclipse +Copying and pasting that into Pivot results in +Suggestions are welcome +I finally chose to remove it so the bug can be closed +anyway +Committed to trunk +for use initCause on the new to include the stack trace after visual check and compile on my client space +Added that binds to listenaddress with defaultconfigurable default RMI server and registry ports and respectively +Thanks for the clarification +I think the best and simplest solution would be to treat them like normal footnotes that are printed once +Asked Jive Support to take a look on it + +Its fixed in right +I dont know why it WOULD have to do this +This issue benefits from the patch to the extent that to be able to interrupt RPC operations properly you need the patch appliedI propose merging as this patch provides the test +A dev build later today should contain the change Ill let you know when +Jobs in both xx and mastertrunk should now be able to force a publish even if nothing has changed since the last published SHA +not sure why affect version were be sure that you dont break something when fixing this in +You can make load thoses properties file this feature has been recently added +Well +Bulk defer all frontend issues to +I think the block is taking into consideration the line spacing even though there is no text +bulk defer of open issues to +The problem is how to do it +Ashish In order to test the patch I wanted to reproduce the problem first +Ok the artifacts are imported from central to the repo and the project is all setup +Patch attached +It contains links both to Maven documentation and external sites +The unit test is a little bit contrived but does demonstrate the issue and act as a regression test +also to do not like like may be +This patch does not address the issue +Im using STS and GPE M +Thanks for this Eli +In this testcase we validate if the whole line matches the expected regexMissed mentioning in earlier comment +Did something change since the call that has not been communicated to all participants on the call +based on a conversation I had in IRC on this it looks like you could put that component jar on the classpath of the enforcer and then try to load it +Instead of framework level integration a rd party Struts plugin has been created. +In case of repeating terms in the query this might be slower than previous patch but it is supposedly correct in all cases while the previous one was not guaranteed to be always correct +it still has some issues with repeated pks I filed to keep track of it. +I tried to respond earlier this morning but this site was down for agree I am surprised by the lack of response to this +Ill contact them about that +Im going to add the class diagram on the Graffito site. +Did you check the patch on EMT? It seems not to work for me there +I will fix those issues ASAP +Its nice for the development process +However stores that retain this information typically do so for query optimization therefore it would usually just be an estimate for costing purposes and not necessarily accurate +Just a thought +As you told different directories can have different capacityBut If I use threshold as constant such as MB then this will be the minimum free space bookie expects in all configured directories +AntonInvestigate the problem +Yes ODP and other files work since even basic formatting and tables are extracted and written to the XHTML SAX stream +Should I make a new folder in the samples directory? How about directory of +Awesome looking forward to it + ve committed a patch in SVN trunk that hopefully will fix the problemCould you give it a try and post your feedback herethanksve check the latest SVN with the original file and it works flawlessly +The PR is ok to me +If the user is unmapped they have no rights to do anything related to managing the there is some specific way we can improve the user experience when this occurs lets look at question about an unsecured management interface +You would have to supply your own grammar to get your expected behavior implenting the and adding it to theI will provide such a liberal grammar with no links to the tracker but it is a little hair stretching as Remy +I think it is a good practice that make sure those are closed and therefore related resources are released +The only problem is with providers which get registered with their proxy note that proxied providers are still. +Tested and merged into trunk. +Thoughts anyone +Confirmed by a user as fixed in. +classpath updated in revision . +You are right is much betterPlease do the change if you wantThanks a in r +Fixed by taking out the delete user feature +So is this done +When I put in excludes they are completely ignored +we also need to upgrade jackrabbit since the old client relies on introduced a with the new codeAll unit tests are working again but I think this is because we lack a lot testsThus serious testing is needed the next weeksIf anyone reviews the code Id be pretty happy too +If you need any further help from me on this please let me know sooner rather than later as I may not be able to down theLeigh +Once CVS comes back online will check in the fix. +Draft complete and reviewed. +For jline specifically this should not be neede din mode +Ok I understand nowthe error is on the Apache Flex site when you dont have flash installed not on the Flex SDKBut it was difficult to guess from your description +Who knows when will be ready so its hard to say +So far people havent reported random failures anymore +I tried installing it but that did not change anything +Then you could iterate over the set of named components and set their visible flag as needed +This patch makes the following changes to mapreduce +We have a constant EPSILON available in all tests that is pretty good for this purpose +This does not mean any new fature automatically will work with all current old builds +The file didnt change since beta but I redid the patch anyway +Thanks +Also just noted that while loops have the same limitation +Contains the of my client that produces the errorand three showing what happens on the line +I got some spare cycles and had another stab at this +Thanks +will fix it +if the submitter could confirm that it is fixed with the latest nightly snapshot then the bug will be closed +Thanks +The has access to the maybe even higher than that the client of the might be the right place to start +For the split Ill create a new issue no problemSo lets close this issue. +Ive tested updated scripts on Windwsx and Linuxxxia and they appeared working well +Added in unit tests and corrected spacing issues +Not sure why but on callback was called in a separate loop after all items have been processed +thx! + modifies to run with a connection +bat files also works before i can close this +Thanks Karthik +I hope Maven moves forwards and decides to leave Javadoc Tags +Wait till tomorrow so the snapshot build does include the fixAnd you can enable DEBUG logging for which is the componentIt should log something about NOOP and reconnecting etc +After gets fixed Id like to update it to pointing the API of obtaining the container log because IMHO its enough for the DAO object to just hold the log content which is independent of renderingThoughts please +It seems like this is resolved in latest nightlies +One more to go +Please let me know the branch or revision so I can sync and again +Moreover to be merged your patch should be in is the same plugin but rewritten for JDK have updated the code +Thanks Warren! Committed to. +I have just got an answer from the shindig dev mailing list that this has been fixed by. +Added +Any updates on this issueI guess we can just automatically ignore this test when detecting that it is running from outside home +Seems to me that lets keep it simple for now +Closing this as a duplicate of a newer change. +fix heisenbug in where a compaction could add back a new sstable to the cleared +If you are an app server administrator who loves the the role is for you +revision tried on on and I still hit the issue +Please do a submit patch when you put up a patch for a JIRAThanks I have added you as a contributor so you should be able to assign to yourselfThanks it thanks +Im not sure we want to support system views within the configuration itself +Periodically cleaning that up would be possible but unnecessarily complex I thinkAnother option would be to remove the previous after you roll rwcc forward and release the row lock before dropping the lock +So we are moving it to the planned contrib folder now in case of any future development about it +Theres some discussion on the dev list about putting this in but will hold off on that until there is some consensusThanks to all who helped review +Amita could you please take a look at this and see if there is any remaining issue that still need to be applied to the wiki or close the jira issue +I have made only those changes necessary to get and to compile and all unit test to pass +But need verify the history of AAR file deploy is it always take long time for an AAR file to deploy or not +Committed to trunk and +In revision the implements most of whats needed to make the javascript client workThere are a few left in there and the integrations test coverage is very poor at the moment +The test in question runs the following commandjava server load srselect loadoptsIm curious about what results you see when running with the insane jarsDo you still have the rig up so you can take Derby for another spin +Oh I applied this patch to branch and trunk +FgI checked the file in the tag I checked the trunk and I also checked the file that got shipped with the jar file +Thanks I just moved from Brisbane otherwise I would of bought you a drink requested some feedback from chrome ended up taking mins in total +If we throw an exception then region will never open +However there seems to be a relevant JIRA is by no means an actual solution +If so he also outputs a deprecation warning to the log +I downloaded just fine and verified via Firefox +It looks like I need to configure the with extra hints that are used for constructing a fully inized message? Cheers you! +please review and commit +Thanks for the kick in the pants to get going on it +Is this actually resolved for EAP +I committed the patch to trunk and +Having some issues with the nexus server not sure what is happening investigating +Ok after reading the mysql forum setting fetch size to int min seems the only way until server supports fetch sizes +Updated patch based on feedback above +Oops uploaded a wrong file +Fix committed for today +Can we do this as part of +changed the API such that administrators define a list of in conf +implementation creates always new instance ignoring the fact the instance of the filter is already created in I have attached the source code and war file +resolve as out of date because of month no activity +both queues and have a operation that can be used to flush their the filter parameter is null all their messages are removed. +Add ec to drivers list and provider types +Pushed to master +thankyou for the information +In any case the output of ps is formatted differently and wont work with the rest of the script as is +pretty sure this will be fixed now +Can someone move this issue from to. +However this is limited for and there is no code generation support for +thanks shravan for contributing +Ok I think Ive found the way to make the problem happen using the GUI may provide a clue +I have added another configuration setting to dictate how logical works +duplicates Issue shouldnt set a fixed versoin +so some clarification would be helpfulThanks +I can fax this tomorrow +fixed +Do you plan to add any test cases for this issue to the replication test suiteYes but a security manager is required for that +In addition the internal page reference was being used instead of the XSLT engine generated idThis was too late to make tonights doc build but I hope to verify the fix in the next doc build and close this issue +Thanks to Rafael Luque for reporting the bug and creating the patch +I added a couple methods to Helper to simplify and clear up the single Class case which generates a bunch of warningsThe remaining warning are all related to the deprecation of handled on and +OK +No it is different from In this case the NPE is caused by the fact that tagHandler object is nullAnyway by a static first analysis it seems that the NPE is caused by something similar that happens in though I am not sure if this issue is a duplicate of that oneIt needs some dynamic investigation +That proviso limits the average seriousness of the bug but file descriptor leaks are never good +Len wanted me to close this one the plugin is now installed as part of the build. +It also makes the DSL a bit more readable +That is dont throw an error +Please do not set the Priority field its for use by commiters in organizing their work +The service implementation beans is used if is present on both +This suggest the issue can be somehow considered a duplicate of issue or at least tightly related to that one +Attaching a patch with unit test +Trivial update to merge cleanly against r +Hopefully this is something that can go into or a maintenance release +James +I want to ask for some more ideas for this new featureAs search base I think the currently selected entry should be used +Icommitted a patch that I believe fixes the problem that you encountered +Looks like we still need to add some code here for our new plugin types but I wanted to clear that up lest I confuse people +It now uses to decorate the Poller if transaction attributes were defined +I attached a patch to increase the thread join timeout time +Is this resolved +My rationale is that if we commit this then there wont be any motivation to refactor later +I guess we need this for trunk too but we can do it laterAttaching patch that should fix this +Petes response wasThey are as necessaryunnecessary as a serial in any other serializable class +If it is not on the wiki in the FAQ we should add it or maybe add a pitfalls section to the wiki. +It would be nice if this was fixed as part of the patch +Should we expose the metadataKey as a configuration attribute in the namespace? Should we even require the user to provide a metadataKey in order to prevent ambiguity? +Pretty minor issue though and Im not sure whether we can or need to fix it +what would be the objects sids and object identities involved +However if they are run as scripts they just work +The is obviously not blocked by them since its started +buildivylibHadoopcommongriddevbin +Ive emailed the developer list to seek feedback on this issue +Can you please run your app and watch the following object in the profiler? BTW what is your consumer app memory settings +Thank you Raja for the review. +bq +Committed to trunk and +Sometime it works other times it does not as i can see be easy identifiable by looking at java version and show a warning if gcj is being used +Attached is the first version patch which is on top of the patch for +There were just things modified +Also this commit fcaebfdbdbbecece missed by ASF bot due to typo in cloudstack bug semantic +This testcase is part of our shouldnt be hard to create one into +Ive taken a slightly different approach and provide an option to cache keys go unresolved + At revision Thanks Maja! +Lets punt this to these as fixed against rather than against +Thank you. +Committed to. +META +This could happen if returns by Session Factory path +AshutoshIve verified the query is working when both and are applied +Username thetooth +On Linux messages resolved on Windows there are we should check Mac also +bq +Thanks JunpingThanks Jing also for reviewing the patch +I went with x since that was the size of the moocard design for and worked great for a sticker +please check in Chrome also +Among them are also seen failing with other just posted patchesI run them manually and they are passing +Bulk move issues to and +My project was initially written in java and it works fine +Patch committedThank you for your contribution DavidJarcec +Try running set VERBOSE true before you run the command +missing license have uploaded new bundles to the same +Uses double memory or some other side effectYep which can cause previously working systems OOM errors +Please verify +create new patch based on the comments on +Resolving as fixed as per Gaos comment that he could not repro. +correct a typo + Bug has been marked as a duplicate of this bug +Ill commit this soon unless there are objections +There are so many places in the code that presume a regioninfo in +Can you tell where is this fix? Cannot find it in trunk yet +That is the problem +no such report with versionsDuplicate content can occur at XSLT stage before FOP stageIf anybody encounters such issues feel free to reopen this bugIn that case please provide a FO file reproducing this issue +Test still has to be Ronalds patch has out of date +thanks for addressing my comments Ahmed +The issue was a corner case +OK on the balance of things Ive decided to make this change +Ill fix with the other patch since theyre newer +So all issues not directly assigned to an AS release are being closed +Default setting for RPM builds is LANGC +up to date in RC +With a set subscriptionName it is working as expected +This bug is related to however the solution is different due to refactoring in TilesAris please attach a patch using svn diff if you want it to be applied +This is the right way of doing itIt particularly makes sense since the main file and the checksum filedo not necessarily need to share the same buffer sizeWith the file buffer large the checksum buffer doesnt need to be large at all. +Reported through automation +Im going to commit this +Patch looks good +i mean why do we need ivy to push our release jars into the repository +Ill try to get that done within a weekIm also not sure about the right course of action at getting it used in +There are some JBT related issues in jira +Reporters of these issues please verify the fixes and help close these issues +committers please tell if youre ok with renaming renaming if included tests works deletion otherwise +In addition by pushing the up the hierarchy it will only run twice when you have two script containers without a common parentcoreWeb and coreWeb demonstrate both conditions +The issue simply is browser limitations on literally of all browsers and limitations in the encoding on the javascript sideAs far as it seems the implementations handle the non utf pages and mixed utf ajax requests quite fine +Otherwise we can get it committed next release +Upgraded to xbean +The demonstration can be found at the +Ive just moved up to an and will try the patch now +Fixed in revision +This follows the Postel Law as you mentioned previously. +issues +HelloWould you please try my patch for this issue +Further to Marcus comment I can now also see that the fact that this example fails is documented in the server configuration guideThe first call should succeed while the second should fail due to the fact that Four is a word +I agree that user code MUST NOT run in the Hadoop servicesJust to make it clear my suggestion was to have an service interface like JT has the Scheduler interface that can be use to augment server behavior +Thanks again for taking care of this I think this could be a good candidate for? +Patch added +I start by the step be able to define the rmi in the etc file +This patch creates the copyBytes methods for Text and +Verified in JBDS. +The problem is that the HTTP specification doesnt really explain how encodings should be handled +after you fix the javadoc warning +works good thanks jukkaI reenabled the failing test in rev marking this issue as fixed. +just added your changesty worksbq +And its resolved. +However the documentation is still wrong and highly confusing +Please tell us when hibernate will support Comment on Column Entity +Or may be its just that Im getting back to Java after a whileFrom the current usage I see that we should honor the following formatsa +Ported from mainline docs to branch at subversion revision . +Can you change the test to add is not null in the second join it is painful to verify the results +One more thing the current code fails to honor are the crs forcing and crs reprojection fields of Query + for the patch +I say that for the person who put fiscal year setting in webtools while it should be in accouting +I tested on Xoom LTE running and saw no problems when rotating device around multiple times +of collection mappings is prior to specified in CriteriaPlease see +The patch is now committed as svn free to close. +It would be great to have another logging message around how long it took the server to be ready to display applications +What happens to established SSL or sessions when calling reload? Are they killed or do they continue to use the old certificateAtm the connections are not blocked from reading and writing +Its not even working with simple key likeidgeneratedvalueprivate Long idin class +Go ahead and close issue. + +Here is a fixed patch for current for inconvenience +Considering this resolved +Do you want me to do the following move the plugin to the feature and make it functional remove the pluginI can also add a builder for the new +However one might doubt whether the solution is really that easy +Ive updated the patch accordingly +Martin has been redefined +For the moment moving on to to round out the ASM bits +Thanks Dims +Zookeeper Quorum localhost Addresses of all registered ZK servers +after discussions with Peter these values can only come from the XML so well fix it in the +Thanks Skye! +yesbut i dont tink its necessaryso you can close this issue +Mostly a rebase of s code for the said tool +I am sorry +In my head at least I think the concurrency issues are workedout in this patch +if objects will move this issue to +Checked componentsajajaxListenerajkeepAlive aj components ajhtmljsFunctionajincludeajloadBundleajloadScriptaj components ajlogajmediaOutputajoutputPanelajpageajPollaj components richajaxValidatorrichgraphValidatorrich component richcomboBoxrichcomponentControlrich componentsrichdatadata componentsrichdataTablerichsubTablerich componentsrichdragIndicatorrichdragSupportrich componentsrichfileUploadrichgmaprichvirtualEarthrichhotKeyrich componentsrichpaintdrichpanelrichpanelBarrichpanel bug is +etcI see some benefit from having docs in schema and thats the fact that IDE or other tools can piggyback on that to offer information to the user while typing the XML without having to go to the config doc +Then delete it once it has been copied + +I see tags is not in SVN anymore and there is a release of on the general siteI will download this and try it out +so when the ledger is opened to read we just could access entries until +In the future browsers will implement the validation logic for these types making the javascript solutions unnecessary +in +Added a comment on review board +Looking into it more +Just applied to TRUNK too. +For instance I found the recursive invocation of particularly questionable +Any feedback for this +Actually perhaps I will probably be able to do what I need with a simple +Ill also work on backporting this to the other release branches +Does this issue duplicate or relate to +bq +patch for compilation like juno now ships with +Thanks +This patch fixes all the javadoc warnings +set fix version +Which approach is better +Its also supported by iOS +I dont think I realized it was using and notThanks stack +benwhat do you think about holding off on this until I get the new CAS clients out and then the complete CAS client for Acegi to allow for sending a ticket to an arbitrary url? given its an improvement as opposed to a bug fix and you lead the CAS integration I am satisfied with whatever you feel is best +Looks good passes unit and checkin tests +I propose using instead of as it is already extensively used +that last patch included some unrelated code +any other update on that page +Patch integrated to trunk and. +In order to establish the wiring so that callbacks can later be made the observable MUST call a method any method on the observer +This has been merged from trunk to branch so closing. +For now can we just have a default that allows any endpoint or channel beans to be exposed via the +patch applied thanks. +It may be so but IMO the design of the Job Object and the submit API also should be discussed during consider about overall architecture +rOK I applied these changes since they are less likely to conflict with other changes that we make and patch conflict always happens anyway +We are going to integrate Apache CXF when it becomes available +Thanks Jeff. +This issue is no longer relevant +Anyone can still enable it for whichever workspaces they like +ping should not be in since it should not be visible to the user +Ive updated this ticket to get the final release +Martin do you have any kind of a test case to prove this problem? If so pleasepost it +AxisC has changed a lot since +I built it for But I didnt have trouble patching it to +Looks good Frank +So can we close this + appears that this fat jar contains the dependencies of the test artifactsIm not sure how useful this isThe dependencies of tests may be different from the dependencies +Im sorry for that inconvenience +David the admins are aware of the problem and currently try to fix it +Modified the patch posted on in the following ways Changed disk amount configuration from percentage used to bytes remaining +Making blocker on +Patch for hadoop is added. +bulk defer of unresolved bugs to +Committed to branch r +I just committed this +In the former were marshalled sending deltas around and in the latter instances +Please see the screen shot attached +Here is the updated patch +Looking at your stack trace you are not passing in a this means it will instantiate a default package which uses jdt if you arent passing in a you can use a properties file in the right location + +This bug is urgentIf the application is running and the log deleted the log is not regenerated and all following log entries are lostIs the fix planned to go into the next version? +Cheers Stefan thats done the trick! +Need to have different destionation based on the server version +I plan to commit this until if nobody commit this tomorrow if nobody if you want I will take care of contribcollation tomorrow +Jerome I think we should remove chmod code completely from Chukwa Appender +Owen the way this JIRA is going really confused me +If you finish it before we release beta please correct the version +Hal was correct about the tag class not being available in the extensions jar +Need to test it before committing +Which means it is save to simply put the String in QE because no will be inside the string +Can you please consider submitting a patch against javasrcorgapacheaxiswsdltoJava to corrects the problemthanks in. +OK +done + +This issue is very similar to +Hi Im querying the server by asking class to send me list of dependencies and on my local host Im able to get the whole list within a fraction of second +Thanks Bruno! +See domain and controllersLet me know if you have any more information that may help reproduce the problemThanks for the help +Will help with us with pushing incompatibility feedback upstream +as far as I understand this warning me tells you that me will not build the war in its default build cycle within eclipse +there is nothing else special about this class as it really has no other fields + has been added to productize and make this script +If we want a memory sensitive cache then we should be using hard keys and soft valuesI vaguely prefer a hard cache that drops entries for classes that are redeployed since I care more about speed than memory footprint +Marking as +Voted for this too +I reviewed javadoc but right or wrong I didnt come across this +Im unable to duplicate your problem +joy we dont need to move to +Thank you Raja for the review. +I have released which includes the fix and also updated the branch +Ultimately we should make sure we do not have two asm version at the same time +You do not want to use d twice + +Changing them a this point would be a large undertaking +that is what hornet does just dont know if it is the best default way to go +What do you think about this version that doesnt change the code formatting as much +Can you please provide a separate patch for +Ive attached a patch for camel this should allow AIX users to successfully build from featurestrunkNote I used a standalone copy of IBM JDK for my buildtesting. +AFAIK They are being used in branchx for back compat +There is a small typo in the first part of the patch the private method should be named get not get +JianhongCan you try to build the latest camel trunk in you boxI already committed a quick fix for it +So we should leave it as it isTo the buffer problem For easy external use we could also provide a expert API that works like the current public fold method which is memory efficient +Nevermind about +Moving out to move back to if this will be work on soon +The now subclasses which was part of a refactor of event handling as per new task has been create to add more consistency to the authentication mechanisms which was the second half of the request +r turns on encoding for these headers +There is nothing in the spec which says that the definition cannot be widened +Im really sorry about this but my checkin for revision totally invalidates all of your test results +Committed patch with revision into the trunk +Verified on JBDS GA. +Please verify if it fixes as you expected. +Cause INFO balance hriaaaaaaa +Did you update Cordova as well? What Cordova version do you use +Committed to the trunk +I am working on this issue and I will be working on whole functionality of this issue +Right? But we can not change the bytecode for for example that if it uses a that uses our files that not the default encoding for Java is used but the encoding you did set on the file +Sun has relicensed from SBCL to CDDL so the switch may be avoided +Screenshots +Error occurs when using either firefox or that I couldnt see the error because I had caching activated which almost entirely removed the concurrency issue +gsh works after including the configuration +is good for trunk for additional bake time +If you import them both into your eclipse workspace you should be able to run the test in project B that calls the service in project A +Ok I will prepare a pull request for them +Good change to have +see test output in comments aboveAdditional issues mention in this JIRA are tracked in the openstack gem as per marioss comments. +Whats the problem with a static CLIENT? What problems does this cause? The client has the connection pool so one potential problem is that a large request or response will delay other requests or responses tofrom the same host +I checked in a fix for this problem +Please use the patch +Thanks Knut AndersI committed patch b to trunk with revision Depending on how things play out I might consider adding a reference to the stream in the descriptor objectIll leave the issue open until I have a clearer view of how and where the descriptor object will be usedIf anyone has opinions on keeping a reference to the described stream in the descriptor please add a comment +If he needs a target he does not need to execute it immediately the user wants to define a graph of targets first +This patch fixes Jo! and also +then the problem occursin Xerces supported clone on the without but in it doesnt support it +This is a database that is empty but is corrupted somehow +Added configuration parameter to disable is for srcmainorgjbossseammail is for srcmainorgjbossseam thanks for the patch +Its a slightly larger download for mobiles but if we keep it as compressedsmall as possible it should be ok until a good responsive image solution comes out of the crazy discussions going on right now +I removed all these and tested fine in my local copy +is a classpath issue as commented above +Committed in revision . +on +Will get back to you about this +Looking more yeah class should be called JVM + UTC +Thanks I have a presentation today so I will be quite late back at home +Thank you for your sugguestion it is very useful +It sounds reasonable and smart to do so no matter how short or long your branch lives +Verified. +Sometimes all tests pass sometimes only one of them fails sometimes all of them fail togheter +schedule and off to release +Do you have any idea on that? also mentions that it is not thread safe although we seem to be updating it from all readerwriter threads +Thanks. +which base version is it? I guess it is from branch or not +Site is now moved into drillsite as requested +bulk move of all unresolved issues from to +Thanks Jrgen! +See comments +Uses executors for Auditor and +Whats the use case for this feature +Thank you! I see glassfish artifact in central repo. +So part of it is my ignorance of UI +Do you know why you got the permission denied error? It was working for me earlier so Im a bit surprised +Instead would it be better to use watchersThanks for starting work on this +Resolved on on +Patch for and +This was fixed in by adding synchronized versions of the unsynchronized methods +Thawan Out of curiosity why a variant and not this patch +This branch was voted in on the mailing lists and Ive merged it to the trunk +Please reopen if you experience this issue again +You are right thats correct! I misunderstood your answer +attach a new patch based on +I cant tell for sure since the results from this JIRAs run have been deleted by Jenkins but I suspect that this wasnt noticed since doesnt report on tests that time out rather than explicitly failThis test failure is tracked by +I went with just a minmaxaverage on these +We attempt to monitor +If nobody has objections Ill run tests and commit +catch Mike +All tests pass but because tests had to be refactored a bit Im not sure Ive got exactly the coverage that was there beforeIll leave it up here for a bit but Im going to want to commit this pretty soon because this patch and the one for are pretty conflictory +That sounds like a very reasonable explanation +committed to trunk and appears this is no longer being considered for inclusion in so removing from the Target Versions field +Thanks for the patch Jeffrey. +I looked at you patch and decided a variation on it might be a little easier to useIm guessing that the +Thanks Naela +fixed in rev . +There is no option how to deploy it from JBT. +Please reduce priority of this ticket from Major +Retrying +sorry comment added to wrong issue ill try again. +Works correctly. +the patch is rooted in the javasdoimpl I hope this is better +Generally weve been trying to avoid new time functionality as the underlying system in the JDK is too painful and the code ends up with buggy boundary cases and painful codeIf anyone would like to submit patches however I could see this change going in Lang at some point if the code to implement it looks simple and clear enough + +issues has been solvedi will add a description of the changes as soon as changes are available. +Ive committed this to as well to let this logging feature continue to work as in. + +which seems reasonable but at the same time is the reason for the slowdown Ive commented out the code that sets up the tiling hints and boom got back the it would really seem most of the speedup was really due to the translation path not using those hints in origin +fixed in r +John What version of are you using? Would you please show me in your workspace +Thanks! +bq +Also notable is that if I manually add a file any mappings I put in it are ignored +Added at apache trunk +I noticed that was not present in this build +coughSorry my fault mvn clean did the trickHowever connecting using still doesnt work since the schema is not served +This is an integration bug +Also thanks to everybody who has made helped and made this happen +Inthe added test case this is triggered by off line commit which effectivelydrops and recreates the base table and all of its indexes after loading thedata into themSometimes these Observers may execute work which adds to the Observer queueand that queue can miss them in the first pass throughA previous fix for this problem added a recursive call to notifyObservers inthe place that could cause this addition of observers +Looks like the tests hit metrics when not in the region server +Do you know what issue fixed this? Also can we can get the test committed back in the test suite +The option has been there for years now +patch for for your comments BJAny clue how to go about it +Only meant for trunk thats all people wanted +If you think there could be hard to diagnose issues however then we can just leave it with you +But I think some of the changes that Sanjay is proposing is worth considering if it simplifies the design +Deliberate +And I have posted it on mailing list for further discussion +Havent heard anything back from the project so Im pressing ahead +Development will continue and the CI builds must pick up changes in any of the child modules +The issue with Maven Grails plugin was that it is still at version and there is no new updates for +Going forward lets create a ticket for each separate patchbatch of changes +Theres unfortunately nothing can do about +Submitting a small patch +I have tried having the server side with plain ruby but that does not help +It avoids a slow and has a fail fast check before switching each char against a rather large list +The test depends on having a single mapred local directory so I did not change that +Applied patch in r. + +I think that the demo code will continue to be copied into the source distribution along with the rest of the source tree + the patch looks good to me +Thanks for taking care of javadoc warnings in trunk and Knut +Hi CostinUnfortunately your hint did not work +Yeah sorry about that I knew this problem sounded familiar but couldnt nail it until I checked the source to apply your patch +Saying that you have to build it on your own using the current trunk +Could you prepare a small demo app? Then I will be able to debug and look into it +The major change in this patch is to convert a bunch of Stat objects to classes and instantiating one instance of each such class at the high level Producer and level +Funny I never used it but still was under the assumption it would work +simple as maxlength +I now remember that I was going to fix the problem with properties bug forgot toI removed the warning because all the other processors I tested werent issuing a warning either +What is bad each call of get makes new class instanceMaybe you have own vision how it might work +Soory nevermind +Gregory? In addition Id like to have something now something that works +Thanks Alison +VRW only requires another +BTW if you agree with this approach we could use some help implementing these createCopy methods on each of the models! Were also fighting a few more fires in other JIRA open for so any assistance you can provide here in terms of a patch would be very helpful and help the fix get done faster +This seems not nearly as good as the implementation that is part of Dmitriys random projection SVDIn particular there is no need to cache the results +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Now I can in fact confirm this problem is solved usingthis new driver even using the thin connect mode +Is this possible to replicate +Should an exception be thrown deployment time if is added and no Seam Security libs are deployed +This is not a crash so should not generate a javacore so users and the tests should not have to work around generating javacores in this situation by setting is currently a session level error which seems correct to meSo I think the problem here is that our determination of whether this situation is a a crash is a bit offAs I recall the determination of a crash was when we got a session level error on a booted database +Resolved with +Have you tried this against? Im a little swamped at the moment but please give it another shot +you get the jsessionid part appended to your link +Final patch against trunk +Applied +Trivial patch +Removed myself from assigneeSory lack of time those days to support same patch updated for SVN +So the current global planner couldnt deal with new join plans +This caused three failures in the nightlies +just committed this. +Agreed Struts doesnt support proxies for interceptors and actions and the workaround seems to fix your issue +I will write unit tests but Im having trouble getting this to work in my actual error case +One thing Ive noticed is that under IE the component receiving focus is scrolled to visible but under Firefox the component only receives focus without being scrolled to visible. +Every doc is now in that test +Painful experience +Interesting! Can you post the message source of the message received from James so I can take a look? Alternatively if you can view the message source how depends on your email client is the missing header anywhere to be seen? If so where relative to the initial set of headers and the first blank line prior to the message body? Steve +Changes to trunk Added swithed to as needed latest transformers +Committed revision +Allhidden nodes need to be excluded from being versioned regardless of the OPVSee. +Still closing this issue as proper synchronization of localObject is committed and tested +Maybe the Azure Blob storage driver does not leverage features of the base class correctly but I am not very familiar with the class hierarchy so I cannot really tell +I will say it one last time I will veto any patch that uses exceptions to to communicate links within the implentation +However I also never use log collation in production +I think your issue is related to using Jdk +This huge difference in performance cannot be caused by suboptimal implementation +for oh I forgot to mention die die die +Btw also looked at the geoserver patch seems reasonable as well +Wont fix +Wed presumably want to incorporate that data structure into whatever we did for pooling +The patch from Egor Pasko in fixed the issue however still fails on RHEL update +Hi KenneyWhere do we stand on this? As far as i can tell the whole site inheritance stuff is unusable due to all the bugs and strange regards to the multiple parent items appearing in the final leaf site as a quick hack I have removed the populate and populate calls from mojo that resulted in the deployed being modified before attachment and moved the populate from within the recursive get to its caller so we now only ever get one parent item in each projects site +well or bits is also good as it can greatly influence the vm vendor and os name are now included in the output +on +I am going to revert the patch until we can fix all the tests +Upgrade is disallowed from toRelease changes Change the layout version to to disallow upgrades from to +Can you tell me what else I need to do for JIRA to show my commits related to this ticket +Heres a patch that should fix thisPlease let me know if this works for you +Remaining fixes need to be done on branch and compiler +Can you assign proper statusthankssudha +should i put the integer in the new error code rather than added a new deprecated constanti cant find any failed tests in the test results +I vote for easy in this case if users get so low level that they care what this table is named then they have no right to be offended +bertrand pls ping me when you start working on this so we can add the continuum server as well +This patch is still WIPI have tested this with fully secure hadoop hbase +I think it will be way too heavy to parse all this in the browser side +Just minor comments on RB stack +It seems to me like all you are doing is adding a poll method to a repository connector +What an ignorant thing to sayWhen disabled Benoits patch leaves Couch unchangedWhen enabled every site on the Web can query your couch with your users credentials +also we would need some build files for building these recipes +The log you show is an asynchronous event not in the thread of the main application +I am not sure that any work is going to be done on this for now +If you set in intellij it will do the right thing eclipse has similar settings +I will send patches for these in the near futurePlease review and commitThank youDumindu +flag +Furthermore the class doesnt add any new fields just adds on another form of serialization for the preexisting class +In the case I agree about properties but Im not sure about attachments +models and model resolution is already complex enough to warrant only adding new features that add complexity when they completely make sense +close as the current logging is working perfect without any crashing issue +Any Final fix for this problem +Code tags copied the sample from the docs for for the issue and pull request. +See +Im not sure whatd be involved but I figured I should make a note of it for reference +Its fine for light use see the example code in but it depends on what you expect from a rich text editor +fixed in trunk and backported to and +This patch fixes an issue with spaces in the SQL query that is causing column filters to fail when a global search term isnt provided +Thanks! But for my opinion you can lower the for package again +Please reopen if not. +Thanks Ben Tom and MahadevAs part of the commit I created a new section and moved issues committed to from trunk and into that section. +Some comments the update method esp UNHEALTHY stuff +We run out of memory in the reducer when data is skewed +Matt what actions would you like to be taken as a result of this? I can imagine that we should +I used Ckeckstyle Suppress Warnings criteria to detect usage of annotation making PMD failed during analysisA screenshot is attached +Right your test code is do the same thing as mine +It uses the serde property bbbbbssbsbbfor specifying column level storage options where stands for the table default b for binary and s for stringUTF storageThe table property binarycan be used to specify a table level default. +arpit Id rather use configdir but if you want to use config Im fine with it +It may be that the ajax update is not able to locate your component in the DOM if it is attached outside the form. +Labeled derbybackportreject +Currently as mentioned at we cannot do much on this and I will be updating a little so that it can report an issue if other validator than is used on initialization ofI am planning to put it to +Thus were definitely getting closerLatest patch attemptColm With this latest patch all the tests now pass +Not a problem +The problem was that the repair did not stop when it encountered the root nodeFixed in svn revision depending issues and have been fixed Im now quite confident that this issue can be resolvedIf there should arise new inconsistent situations with the search index a new issue should be created. +BooWhat about adding to cfstats then +This is a branding issue so should be fixed by is done +With that hint I just pushed the updateRegarding moving the site to drillsite I can definitely do that but we are using git for the code so there shouldnt be any conflictWant me to do the move anyway +I am stealing this issue because I have written a patch +Well its OK if it is introduced in jackrabbit but as a user I would not expect such an issue while upgrading toProbably we should write some external program checking if given repository is compatible with next release and maybe escaping invalid characters in the repository itself +in is not used butit is not pruned right away +Heres a patch +I corrected a mistake in my call to delete +Marat Yes I did manage to get all of the tests to pass before I restarted this thread but as you said my use case is now quite cumbersome and I think it really does not have to beDid you try the option I suggested? Please take a look at the patch attached +Bulk move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Thanks Martin +If you could please tell me how to get the current axis trunk +Closing tickets +Adding nested property support to would be very nice feature! +v adds support for crappy old jna has no effect for me regardless of jna version. +Another update to the patch +An exception is thrown from Catalina saying path is null because of the reasons explained in the description +Can you checkout the project from the using CVS then in JBDS use File Import Maven Existing Maven Projects +Personally Im glad you bring up issues like this but our mailing list is quite active and while my comment is please consider before steps instead of after steps next in order to get bugs fixed +Moving this out to please move back to if this will be worked on soon +New files +There are only two things that might be missing javascriptcontainer appidsyndurlIm not sure if this synd is just a string that can be renamed or has a match somewhere but I think it can be renamed too +Ive attached two etc folders +Ill pull it all together today +sounds fine Jacques and thanks taking care of of thisRegards at revision in trunkI have asked for opinions about backporting on dev ML so I resolve as incomplete for now +implemented +I think most developers would be surprised to find that turning off debugging actually breaks an app +defer all issues to +If I add manuel web app librairies to the webproject build path it works +ftl files +OKdo you have any suggestion related to my requirementthanks very much. +bq +Why do we need to wait? I dont see anything in that article about +Is there an easy way to map btw apache user name and jira user nameObv the Administrator should be me in JIRA rather than me as Apache user name +The problem with the is that it has not been well tested to be considered as production ready and it has not been designed to get the best performances +Played with this a bit tonight to see how it worked +I found this bug one month ago I have lost the core file yeahMaybe we can test it again I had described the steps above +I just committed this +Thanks a lot Todd +Either that or have some back end data store for the proxy to hold that information +perhaps should be done when the Fluent Sender API is in and it resolved it +Thanks for the quick review Knut +Thanks +After further thought this is harder than I imagined +Its helpful to have this test in the hopefully minor fix +I also have no experience with the Spring tests +After reviewing the test I think there are quite a few create tables that you probably just need to keep in your test fixtures since we are testing default at create table time for some of those tests +This is odd +Also testing wise do we now risk losing tests for the other highlighterYeah before the phrase highlighter was undertested and the default had more and nowits the reverse +Setting JAVAOPTIONS truefixes the issue +The recent patch that disables the coordination in generated queries and the test cases for such generated queries nicely exercise the attached from which the coordination has been removed +not a jira has been closed as part of a review and archiving of the aj project +Jochen you are absolutely correct about the use of and yes I did try to get sun implementation working after not being successful with +Thanks Ted +It is still failing on the ibm nightly result runs on with the same error +Please attach the wsdlfor duplicate element problem Ill put a warning message +And also for to handle HTML text and so on +Removed from +Attaching small demonstration project +Perhaps we can read the to discover where the REST resouces and SOAP endpoints are mounted in order to reduce the configuration overhead +And committed +Please review and report back +HII upload example SHP file that contain cyrillic data in attribute table +We are preparing a including some others fixesAnyway I close this issue as it sounds fix to me +Brings us this fartherSuggestion work around? +is supposed to be exported by the system bundleWhat is the exact error message you are getting +It is still open +lgtm +Stepan spelling corrections verified at rThank you for applying this patch +I agree with Alex with IDE tasksDownloading goes ok with both ways adding either new task or modifying artifact to download +A patch is attached for review +I think the installer determines which one to install based on whether you have or bit jvm +I attached a screenshot of EclipseME and filtered the type hierarchy for the artifact string +Do I have to enable read repair for the failover to workThat may be the problem but Im waiting for Brandon to comment since Im not surebq +I think we should move the inboundoutbound modules out of the main samples distribution +Thanks Yongqiang +Generate a CSR from keystore and process the server certificate request using Issue New Certificate link in CA portletStep +Mass closing all issues that have been in resolved state for months or more without any feedback or update +bulk close of all resolved issues prior to release. +Im still struggling to see the value of this change +Continuing to investigate +If it uses the ppr wont work because wont be loaded unless you do it manually +Can you see the problemDavid +No the toplevel is the only file in which we specify an explicit target +resolved issues. +I also agree with Manfred +The current navigateNext and navigatePrevious are changed from their recursive form to call these new that this patch is not against the latest version in the trunk as that has changed a bit since I pulled out the code +Unlike for example which was our serviceIf this only affects sf well likely push to +duplicating it for purposes +tests that will print the whole stack trace in case of failure to reach client s some extraneous stuff in but if anyone can repro this bug locally and run it with this stack tracing on that would be useful +Just do whatever you see fit do it in stages +Ive submitted a pull request on github +Id prefer this issue one sub task for other than not crashing is there for? Just reading the descriptions I still feel that these two bugs are identical. +It would be nice to know which field is being processed when this error occurs +For now I have marked as deprecated +Thanks Khee! I slightly changed the patch to use mincount and maxcount per Yoniks suggestion to overlap w faceting +I will put this JIRA in patch available once it is integrated in +Added code that limits the number of login attempts to one per API requestSo I think the rework of this code is complete +But can you tell me how will i build this codeunder your HADOOPHOMEmvn true package +Sorry folks I notice there were two composite files that I neglected to add to svn +same is due to dynamic partitioning +Our dataset compressed something like with binary protocol and its probably less than that with the compact protocol +Heres a patch that should fix this +assert compactedCf! null +Tim thanks! This solution will work +The SYSIBM schema is used internally by Network Client and Network Server for JDBC Metadata methodsIt has no impact externally except that the user cannot create a schema by this name and it will show up in queries of the system tables +Then I will begin the work on this issueGreetings ideas how to best write a test for the remote wsdl feature +Hmmmm +If we can move the function to the correct place or even identify where the socket connection is happening then I would be happy to amend the patch + +Dmitry you may be right that my patch doesnt fix all instances of this problem +Oops havent rebased with trunk in a while will fix in next documentation to the Admins Guide rebased on trunk small fixesimprovements from comments +Devaraj can you please take a look +Okay progress update I found the overlay code in the KML raster transformer was just repeating the request parameters from the original request so a little tweak to override the format was in order +show page that handled embedded properties +Thanks Steve +sync thread uses entrylogger so should only be used after entrylogger is initialized +However to avoid deadlock problems the application needs to have access to the same readwrite locks as uses +The latest GAE is +defer all issues to +no sure for the exact reason but as soon as I have changed this it should work issue should be fixed now +I started an integration test for actually launching jobs but the testing framework does not behave properly when jars are added to the distributed cache +Looks like most of the problems are related to package protected tests are in the same package as the classmethod being tested +Hi PaulexThe fix looks fine thanks a lotBest regardsAndrew +Vinay agree In this case need some changes in +Yes the exported version needs to be changed +how about adding it to the jscriptI am still following it down to see what may need to be done for any lookupsand why it works once but not after this particular situation +Heres a new version of the patch that has more tests +Im not sure but would this work +Subversion dump has been successfully loaded +Logged In YES useridIs your configuration a standard configurationAlso what JVM are you running withI cannot reproduce this problem on the version of Jetty that isincluded in Username amanpreet +Looks like a case of RTFM unless we actually need to change some hardcoded values? +Thanks a lot! +Well see other forms of this issue Id say +The chosen impl would delegate to the factory. +Yep thats the basic idea +Can you run the derbyall suite and provide a patch for the other master files that need to change + +Why we should generate surprise behavior for users +I can check this today +Theres a handful of non classes tests that still fail though so I still have to fix those +However it is not possible to verify that they are working correctly until is resolved +Attaching the patch is Cannot ReproduceSorry I was confused when creating this issue +ok thanks for your feedbackIll mark this as duplicate of. +Patch changes null check in to throw instead of and changes existing testcases to expect that exception +I though of usecase when the tree state could be +Commit for the branchsvn commit m Sending tcksrcjavaorgapachejdotckapipersistencemanagerTransmitting file dataCommitted revision . +Correct is +Bumping to future releases +So we can close this one. + for to trunk. +Looking forward to read your blog about itIve merged the missing bit of so this should now be working for +I can acknowledge the problem is fixed in final +Im using for Solr because its easier to manage the patches Im maintaining in my local copy that way +Thanks for the patch Peter +stausIdstatus on onbeforedomupdate +Please justify +In rare cases this may require loading a +I agree Trustin re changes to encoder +Andre could you do that +apologies for not finding the previous discussions +FYI this is already configurable in the build +The numbers being output are rounded to decimal places +The HTTP spec has items that pertain to both the structure and completeness of a Request and a Response You could Infer that a request was good enough for the server to make an interpretation and supply a possible answer to the a request but it would hard than to do something reasonable with the cache beyond saying that it should be allowed to cache any xx responses that make senseI have to admit that I am starting to lean in the direction of easing this item out of the caching code and tests rather then tying it into the Protocol processor +Ive pushed the release will upgrade Lucene soon +Proper way to use in Solr tests for the new Varun and Areek +References to should be changed toThe When is the used? section is outdated +Can you please copypaste the generated HTML fragment for the link? There is a good chance that a +See attached report for details +If it doesnt make our initial release it will definitely make one almost immediately after +The goal while deleting excess replicas should be to maximize the number of unique racks on which replicas will remainIf this is for fault tolerance is there any reasons to adopt different policies for allocation and deletionIs it because this event happens less often so were willing to pay the extra overhead +Attaching and with further changes to filesM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideI hope Ive incorporated your suggestions +Hi AnuragI applied this patch and ran derbyall +Do you want to provide a patch asap? Ill commit to thanks +Unit tests passed successfully except a few intermittent ones unrelated to the bug +If you have questions please post them to the user list +It doesnt matter whether those threads are managed by the application server as they do not need any JEE security or transaction context +One more way to fix this issue its disable navigation for up arrow and down arrow leave only left and right arrows +I will commit this to trunk and by the end of this week +Feedback is appreciated +on release +registermailsubject changed to Your account details +Then perhaps just drop a comment in the ctorOK Ill put a comment where I append w the delim explaining why I cant useThanks Shai! Ill commit soon +In this case the query complains as it outright fails +We can add more message like Go to varlog for details +That said I have done some side by side testing using directories of very large documents and the performance improvement is great +Patch checked in +They are statically defined tied to how views are organized how machines are labeled and requires multiple search and replace to actually are actually not a problem as its just following the layout of cache folder +Marking as a parser issue its not handling those characters as a valid method identifier +Simple fix +SvetlanaFixed in LUNI module at repo revision Please check that this fully resolves your problem +Could be a weird side effect of the for the detailed debugging Cameron +classpath when filtering is applied +For example +look for a version of the plugin by July +Patch looks good so Ive applied itThanks for the update. +Ill check this in +defer all issues to +So Ill go for the minimum change of changing the website location from to and leave the rest alone +batch transition resolvedfixed bugs to closedfixed +Its a partial fix for now then +The deprecated attribute is now being added correctly +bulk close of resolved issues from releases prior to +So there should be a flag set on the data passed to the JSP indicating whether a configuration was a child or not and if so no actions should be displayed for it +Slightly modified version of the unit test. +Committed at subversion revision Touches the following filesM javaengineorgapachederbyimplsqlcatalogThe actual changeM javatestingorgapachederbyTestingfunctionTeststestslangBoosting the preallocation size caused a combinatorial explosion in the number of cases stressed in the boundary test for sequences +Thanks for explanations Jason +Hi Eli thanks for the patchI dont think this is the approach we want to take +for leave them for make them part of instead of SS yes having this part of seems like the Right Way to do this +Here is the patch file. +Anyway please send me the environment details through email +Ill add an example soon +Ive been unable to reproduce this +is the platform related issue +I recall doing that exact change when Alex and I were looking at the failover stuff last year and we ended up reverting it +Everything is ok +If you make this simple change then you shouldnt encounter these problems +the utility class and its unit test +Right I am not disagreeing with you here +The incomplete procedure was the second in a task to set an encrypted in MessagingThe same task has been included in the EPP Installation Guide as a single procedureThe incomplete procedure has been filled out with steps based on the content in the EPP docs +I see some fields in s javadoc +I will leave out the stuff and just commit the changes to the component framework that make it possible to configure +it may be better to change preExecutor flushwriter postExecuter chain with something different that correctly handles errors thrownIll just drop this for now. +The product is portalAs discussed previously check the portal documentation to solve this particular issue. +One of the two should be resolved as duplicate +Looks reasonable I like the moving up of yarn and its separation from mapreduce +is shown in the inheritance heirarchy filter it out just as we do forJorgeYoull need to add getset method to exception to get the necessary elements in the type definetion of the generated by using the a lotCould you tell me when will this patch be available in a official releaseAnd how can I apply it to my axis +Builds for now +The SQL which is produce have mutiple column aliases this breaks the paginated query +Attaching version of the patch had to resolve some minor conflicts +assignee +It can be reopened and revisited if a case is made that the current model is lacking. +Status client service generated by the Maven plugin +Thanks for the feedback Keith! I incorporated your suggestions inLet me know if you have any other suggestions +The patch is about as minimal as I can make it right now +OK very good I restored this part +it was on there at one point +The recursive part and filling the list happen a few lines below this code snippet +Im not an expert for databases and they are also not under control of our department +The new code is working properly about Japanese +I am running full regression tests againTouches the following additional filesM javatestingorgapachederbyTestingfunctionTeststestsstoreA javatestingorgapachederbyTestingfunctionTeststestsstore +Although there are some disagreements between parser and POS Taggers of overall integration is completed +Here is this fixed code +For No the capability does not exist +Thanks DhrubaThis was marked for but it looks benign I only committed to trunk +ear I removed all the +This was resolved by giving the child jvm a special logger that captures it as part of the job and provides it via http. +should now finally be fixed +Once theyll finish Ill commit the fix +But Im good with making this a back burner issue for now +Weve also heard that it can cause issues with other which dont handle unknown service contexts very well +I tried it before committing the patch +The patch adds the features section changed the section that talks about IKS the was removed any reference to IKS +has been around for yearsThats far too long +I did have to create a symlink from targetnativetargetusrlocallib to targetnativetargetusrlocallib +The new build is available now +Blindly removing the primary from the recovery pipeline seems too cruel +I have been prototyping an approach that launches the container with a low integrity level and an access token with restricted DACLS +Patch commonsrcmainjavaorgapacheservicemixcomponentsvalidationTransmitting file dataCommitted revision Sending commonsrcmainjavaorgapacheservicemixcomponentsvalidationTransmitting file dataCommitted revision . +Moving this to as I wont be accepting any new issues for anymore +There is also a package with jars built with line number tables to aid in debugging and reporting problems +Code format and java doc comments are little out of order which I will correct later +Marcel I think documenting a preferred case is a good idea but Im not convinced of the value of enforcing it +Schemas renamed in rev +So files are copied from to Core then they should ideally then be removed from since relies on CoreAs for the classes shouldnt we import these as a jar rather than as source +Might it be safer to assume errortrue close the connection in a finally clause and only set errorfalse for normal returns? status would better be named more +But hard to say how these fields are used so hard to disable this feature not to break workflow for some me this seems not as major problem because number of these fields is relatively limited and when you type first letter then completions list is short for all letters. +This is the current code +sha files included in the binary distribution that dont need to be therelicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenseslicenses +Sorry meant to provide the source trace as requested +And with a large ontology SDB is very slow in a database. +Meanwhile you can add transformer in between the inbound and outbound and extract the required values +Uploaded correct versions +Thanks again +If acks offsets in the response are treated as if acks meaning only the leader has acked +Since its taking a long time for a single patch file Ive broken it up into patch filesPaul License has been applied +Interesting +Thanks! I will take a look this afternoon +With such big input files the default logic should split things into dfs splits +Ok awesome +Unless you provide extension points for users then this is just going to create a messI dont recall seeing it in the spec but is there some way a user can insure that their servlet container initializer runs first? If we are going to keep this I would like to see examples of how users can perform customizations +bq +Is there anywhere else we need to document the fact there are database changes +Odd +Patch applied at r please check that it was applied as you expected +As a developer of Hive I am definitely in favor of keeping code simpler and interfaces cleaner +The use of volatile is really only going to be effective if the code is guaranteed to run on Java +attaching test am not able to reproduce this issue under +Related to which is marked fixed for +Thanks looks good to me! +attach a new patch rebased to latest trunk +Are you sure? Tested against latest SVN head and this works on Mac OS X which is nix is it not? Ill get someone else to test it on can you test this one on one of your Linux boxes cant reproduce on Mac OS X or Windows +What should I do better in the future? Single +Closing this ticket as no further issues have come up. +dont be panic +You can see them using the aboutplugins URL +You can apply the patch and retry. +The attached patch adds the bind method to +Attaching the patch for the Authorization Request builder of the client +One of them might be removed +So extra testing necessary in this area +Thanks for picking this up +Sorry still not working +I agree Florian +on commit of addendum +reopened just to set the resolution to with resolution fixed +Please create a new issue in Hadoop JIRA and copy the desciption most folks from Hadoop team are not involved in the Nutch development so they are probably unaware that something is wrong +Finding a way to link the click event trigger with its button +Fixes javadoc errors inRemaining javadoc errors are incontribantcontribluclicontribregex +Sunitha since this seems to add new functionality should this entry be listed in Wiki page? I think keeping that page accurate while optional will help ensure documentation andor release note additions are done correctly for +Please change to if you intend to submit a fix for +DanThe ordering will not fail the tests the assert already take that into account the only thing will fail the tests is the anonymous types non anonymous typesI agree with you though we should use the xpath for the latter testsThe reason I replace assert with assert is that i dont know how many xpath ill need to write for the existing tests as you saw i didnt change any testing resourcesMarking this as resolved as we now have everything working with ad and IBM versions of both and such so any remaining places are not really and issue. +I have not found another editable table with similar behaviour +An would be great but wont be as effective as one that does NOT stop dead in its tracks at the HDFS API +Merged to revisionsThere are still memory issues with a large number of unions in but marking this fixed in since all of the linkage error fixes have gone in. +Note the in the case of the thrown already contains similar message so it would look like the message is duplicated in the same line +I defer to your and Jerrys testing +make more exceptions skippable rename to to better reflect what it does +Closing issue as version has been released on. +OK thanks thats what I needed to know +I have added an option compress +This seems to be causing the stragglers that we observed +for default size tiered strategy instead of compacting larger into one even larger per CF we would be compacting numtokens smaller files into numtokens larger ones per CFAm I missing something here +Yes we fixed itIf you experience some deadlocks try generating a full thread dump +This tool also needs to be updated for for the broker failure and other system tests + +Updated with new code to normalize the hostname +Fixed in branch MP revision . +Jan which is your position? +Also wondering about how it relates to the ordered member and the first and lastThanks far its easier than that when they match they all match so you only need to keep the input Spans around in List or whatever +Jenkins logBUILD ant clean tar findbugs xml true true FINDBUGSHOME FORRESTHOME CLOVERHOME ECLIPSEHOMEBuildfilecleanBUILD FAILEDhomejenkinsworkspacetrunk The following error occurred while executing this linehomejenkinsworkspacetrunksrccontrib The following error occurred while executing this linehomejenkinsworkspacetrunksrccontrib The following error occurred while executing this linehomejenkinsworkspacetrunksrccontrib Source resource does not exist homejenkinsworkspacetrunksrccontribivyTotal time basedir explicitly to the contrib calls should override its local settingThis is a bug in ant and but not in ant + +were committed to cr. +Unless we take a big change on native code design I dont think the code is dangerous +The new patch use for the +No I am not a Java developer and I have not yet started looking under the hood of the Ofbiz framework +Please verifyThanks +Thanks for raising this issue +A few more modifications in some tests +Ill try and support both formats as they are not mutually exclusive +These optionsenv etc are not quite in that context +we dont use jira issues for tracking this +On this page is nothing new I already know how to change the web sites +Youll have to try to know whether its fixed in Groovy +since we have a good build now I attached the same patch to get a clean run +this bug. +I only have or machines +Yes as shown in the attached diagram intermediate events could be attached in the same waySo if that is your understanding as well then I think we can close this issue +In fact I said that above I feel we need to have TWO plugins here +Indeed JEE would require Java and we just updated to Java +I only changed a single line to make an inner class static that should be static +any news about the shade release +Closing all issues +I had a look +Nick whats the status on this patch? Should we commit it +On the trunk the maven build now places the needed axis jars in the lib directory. +We can reopen if it is reported again +The exception is gone but files are not movedWorking with windows client and ftp server still works as expected +Thanks. +The namenodejobtracker would know the correct value and treat as rogue any nodes which try to register but dont know the current correct value +verified +Are you also seeing the problem in other clients +I just committed this to trunk and branch +Doh! Youre right +Ill close this after it has processed its first email +I think everything looks good thanks for working through all of the comments +Will upload a patch soon + patch looks good +Looking at the code of it has trace in place I will try to run with log level set to trace tomorrow and post results here maybe that will give more clarity. +Hi MikeTHANKS for raising this I keep forgetting! Can we put the output in testoutput that way we can have a git ignore on just testoutput rather than testoutput TEST +Ive just committed this +sorry for not commenting on this earlierRick +I have optionsa separate locale setting for the server componentb local setting passed from the portlet to the serverI like b better as it doesnt add an extra config a consequence I need to encode the locale to the filename of the image file otherwise a English and a German porlet will display the same chart +Attachment has been added with description An SVG document that demonstrated how arabic vowels are I dont really know anything about arabicBut it looks like the diacritics are just normal glyphs I dont havea clue what tries to position them over the previous charactersSome of them seem to be done as ligaturesEspecially for SVG fonts we just do what the font says so I wouldsuspect that the flaw is in the font generation if there is oneWhich may point at ttfsvg being the problem but I dont really know +Id rather just have a way to export the entire setup of the portal which would include the pages the pages portlet instances security attributes object placement preferences etc +The RM will stop renewing them and wont issue duplicate renews +This issue has now been fixed for and the fixed documentation will be included with that +Moved to Hadoop common jira bucket since it touches the top level project. +In a wsdl first scenario the current code is likely needed as the normal setup is for arrays to be in the unwrapped form +Would that be possible to fix it in the same time +I just want to make the solution completeIt would be great to see my module becoming integrated directly in the Tapestry project therefore I want to provide as good integration as possible +However I think specifying a restriction of character will take some added work so Im schedulign this for not +Anyone try this with? is pretty old +Works well FabriceJust in case would it be possible to remove the sign when the value is +Following open issues in my opinion if I am not missing something Client instantiations +Listeners shouldnt work in AS +Note that this can also happen if we submit a task to a timer that has already been dont need to fix this in and +It would be really nice if we could have a japanese example too +Its just the original patch plus the required changes. +changed fix not fix this using the unsafe keyword +The patch does not work for me on MINGW however Im about to fix it +Bug fixed in latest Axis svn trunk +I dont know if Im the only one to misunderstand it +Caught a comment change or two that v didnt but otherwise the substance is the same +The class should not be made public. +s right +Load a file to hdfs +Should we expect to see Avro bindings for Lua anytime soonBruce Mitchener is doing some performance work now on that should dramatic improve performance +Just committed this to branch +I know results of would be fixed by other issues +Duplicate of. +It isunique that it is written in java and is architected with modules so it has the ability to be different things to different users +I am for deprecating Thrift and fix Thrift up +Wonderful! I had already written one to return streams but it was not as comprehesive as yours +Thank you for applying it +groupId mygrp will install the jar at repomygrpmylib +note there might be some info in the Console view for Hibernate console +Please verify the fix in the latest assigning back to Farah. +Also moves all of the subclasses bloom half and hbase out into io package +Not a relevant backport +will you commit your changesNo because I found a problem with may approach It causes the plugins to be eagerly accessed as soon as the top navigation is to be rendered thus preventing the delayed plugin loading +Dont have time for this one right now will do in a critical issue pushed to issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +revision +This adds a first method to +I think it would be a great addition and am marking it High Value Fix in case anyone finds that inspiring +Fixed in CVS test case added +Removing the user password digesting is now +We are moving this to future because the workaround using ajloadBundle is easy and was partially implemented to resolve this issue. +Tests in can cover all encoding typesThe downside to building up redundant tests in is that we will always be afraid to remove them or modify them creating a maintenance struggle +But having the build is already modular we have to test it +If you alternatively specify everything in Hibernate properties the standard rules as stated in the Hibernate documentation +Roman says the issue is closed and so do I +I hope to use Jenkins parameterized builds in terms of the end user interface +Username lafr +The problem is not related to pending but the send option which altered the original message instead of sending a copy of it +Actually as a closedresolved issue can be at any time I am marking this as resolved now. +First to the Forrest documentation +for RF smart clients are hopefully becoming the norm with the Java Driver now knocking around so its not a small use case is common but RF is not +Archive file with new files not under version control +Yeah mostly +bq +Adding a cache to a bad idea is a well established software engineering practice but Id favor either fixing our use of ivy or replacing it if middling performance requires this +patch for ran the unit tests and saw issues with most of the python oriented tests +No missing deps error seen this time +Hi EoghanI might be wrong but I dont find you expose osgi service as interface from your bundle which is mandatory osgi service reference for apache nmr +to Koen. +Fixed on master. +on my linux box etccouchdbafaik this patch has not been added to couch yet +Manual testing done +Per Henris suggestion marking this for the core issues will be taken care of in in may not be able to look at this for a week or two +This patch fixes the four shareable node tests. +The class is Ive attached the source +Let me take a quick look at this +So it can not be a reason of radical performance operation ITERATIONS test millistest millistest millisHeavy operations ITERATIONS test millistest millistest contains benchmark configuration and profiler exported views +Attaching screenshot of trunk +I will work to submit a patch for this tonight +Hi CameronIm OK with the general principle of your issue however Id favor either approach +Youre right the ddl example is confusing +In revision updated Tomcat and Jetty containers to register into service registry for the WAB as mandated by the spec +implementing today a way to set a deploy folder for your server other than the default will not be on a module by module basis but rather for all modules deployed to that server object +To ensure database tables are free from corruption Derby uses a locking mechanismto prevent multiple from accessing the same database simultaneously +Closing resolved issues for already released versions +rebase the patch to latest trunk after is in +Looks like the fix is already available upstream +Committed to trunk +Something likeMaybe include the fact that this is for IPC timeouts not say httpI am thinking to start with a large number like hour or day +If the bookies all have same issue the client simply give up recovering the ledgerAll entries get lost including those already acknowledged +interface maybe +it is clearly dependency is no longer considered as not passivation capable as of This issue may be closed. +In either case this is just defensive programming as that enum is never actually serialized in the code +Hi SaurabhI just want to confirm whether your latest bug report was against a table which had been newly created with +Test service and client using the ADB binding generated for the attached WSDL file +Bulk closing stale resolved issues +Theres good material here but it does need some work to make it suitable for the product docsIm actually thinking the content might have to be split between the Smooks Guide and the Getting Started Guide possibly even the ESB Programmers Services Guides too. +just one line change +Shipped +Bulk close for release +tests have been thanks Namit +Also when I ran my client accessing blocks I saw +Done now. +Done when labs went live. +yes we should add a test +I have created a new JIRA +David can you provide a patch +Please verify and close +Im also moving the clearing of the cache to an explicit method so the renderer can control it +Is this the feature that you are looking for +Yes its still on trunk +When the new TLP version of Karaf is released we can at updating the Felix stuff +The CI tests for jdbc use an embedded derby there are not external process dependencies in the CI runs at the momentWonder if adding the tests with an exclusion via a property in surefire will work and have a profile that redefines that property such that the tests are run when it is enabled would work +No so much for the testing I suspect +The rest looks too involved for a simple backport +smile already replaced using the checksum of the source rather than the date of compilation because that was Runpings use case + +Then will clean up the patch +I do not need this fix anymore +Move it back in if you think differently. +It is expected that the second step prints out all items except the th one because this one should be skipped during the second processing of the chunk which previously failed +Heres a patch which fixes the issue +Resolved with revision of buildersriftsaw message Remove temporary build directory +Committed patch which updates the Tuning Guide +Introduced as a superclass of all test classes +Good work Jiannan +patch applied with revision +Please let me know +Attachment has been added with description applied the patch to the and it solves the compatibility issue with Intarsys PDFA Live! validatorThank you +Looks like it is self +Norm +Committed in r on trunk. +consider it is done +Running through hudson +would be great if this bug would be resolved before the release +The correct behaviour would beSee +Then well see where we need to add the support +Backporting the fix for might be a good start +This issue was resolved for tck with svn commit r +oh whoops will resubmit patch + +See patch attached for your pom to actually make some sources to combine in shade +website updated +IP Clearance updated with references to the vote on general +Hi Tony thanks +I would like to emphasize this one more time after failing the workspace cannot be opened anymore and JBDS fails on start with this errorIf you wish I can try to create a workspace which always fails on start and attach it to the issue +Raised to think I fixed this one pretty much when doing cleanups +Ill take a look at your patch over the next few days +Done Thanks +Ha now I know what the issue is +The same goes for the example. +fixed the fix version +document to QE for with to QE for verification and closure +I definitely would prefer NOT setting the ISCOALESCING to true as a default +I would prefer to get all tasks finished before commiting as only checkstyle is missing +Verified that it is ok now. +I intend to submit a new patch for this issue when the versioning discussion is resolved. +Sorry but I didnt get the point yet +postponed to CR since we went out of time for this and Beta is already branched +Patch applied at r +Thanks for looking through it Ive added comments to explain why cannot encrypt data transferred using transferFrom +attach patch for and +Feed f was served by process p in the past but now f is served by p +Been put on another project so I dont get to dedicate any of my billable time to Surf +We dont have that yet but thats the goal. +There is an XSD that can be generated but it isnt published anywhere yet +Verified on JBT. +Keep us posted +I opened to clean up the dependencies +It would be nice if we could provide a JDBC connection and input a SQL query which would be ran on a configurable timeline +Which jira is Daryn making this change in +Here is the quickstart +Matei we recently built a based cluster for the capacity scheduler +patch that fixes this applied snapshot deployed +Okey so this issue can be resolved and more specific ones opened for it if need be +Thanks for your analysis Costin +So I guess were dealing with a potential buggy test +igor done for and works for our use cases thanks Peter +This patch incorrectly changes synchronization constructs +The patch contains changes to scan a table using GET table with scan specification as query parameters +new features are feature list from the guide +When we did the last release of the framework we had to delete this test case since it was not compatible with the new version of mocking library which were changed to make things simpler +this would require another to Heiko to close when the release is done +This happens without any inner classes in the project. +Closing as CR contains global transaction handling changes and I havent been able to reproduce the race condition with the new code +Committed to and trunk +I will test and commit once this done +May need to be if others hit this. +It seems that autocommittrue on DBCP was conflicting with and preventing validation query to works as expected +Ive started the regression tests +This resulted in major refactoring +Very works fine on you is it ok without continuum +thanks Vinod! Ill commit this shortly +I will backport to +The product has its own version of the modeshape configuration file + +version? Its the latest available dated end of September . +Maintaining this in memory for the time could be problematic +Is there also a way to specify a field to have a TEXTAREA instead of just a? Maybe if the Entity has a Max annotation greater than then it creates a +AFAICT this now works so please close this issue +Ive tested the client and server counters with the CXF support in the Sonic mgmt plugin and all appear to be working ok. +A connection can timeout immediately after being leased from the pool before any io on the underlying channel resets the last used value +And these are actually quite useful to start with as examples +export MAVENOPTSXmxg and specify mvn +I dont know what Christian did but heres my patch if it can help +Deal breaker for my current project +Type Font files extracted from +Deleting them from the Eclipse workspace and importing them again would be the way to go. +I added the fix to explicitly close the streamRevision Author dantranDate PM Wednesday December MessageClose explicitlyModified mavenwagontrunksrcmainjavaorgapachemavenwagonprovidersftp can build wagons trunk and drop the to your mavens footprint +I think its not an important issue now because as you pointed developers can use a newer version without breaking things + for patch +addresses the same +I can confirm this annoying bug +bin Now we start opening inputfile + Username rabegg +Theres no need to spend the work adding those segments just so they are dropped on the next merge +I cannot run this test since a lot of NIO tests including crash for me on DRLVM even on +Yes the problem is fully resolved +Done +This is used for quick one off builds or so on where you dont want to have to modify the POM to enable it +Matching patch for to branchesdain was applied to in revision +Changed that +VFS Integration +yeh I can take this one +This error also occurs in version receives only security fixesPlease upgrade to. +This looks great to me +Youre probably thinking of simple member types in apps where the developer also writes the templatesAnyway thanks for contributing the Uberspect +Ill go ahead and resolve this ticket given the is now generally available via the Roo addon install command +late. +In my opinion a new JIRA shall be filled to deal with this issue +Please ignore all commentary about the conflicts now the error is the only unusual thing demonstrated by this patch +This error will only occur on the first time that annotation class is required +as we need info to communicate with the server authentication +problem here as I indicated above that throwing a separate client exception with the server exception as root cause will effectively hide the type of the server exception +Anyway if you are really sure that its not going to break anything I would be happy to accept itIs Batik block in affected by this problem +That was just the patch +Im still concerned about the one multiplier but real world clusters will be the best judge +The patch was basically extension of classwith attach which was used for adding user roles to repositoriesUnfortunatelly and I guess because of the archiva api changes in user management area our servicefor adding roles for user is not working anymore in versionWe are trying to find other solution so I tried to apply exactly this patch to that codebut I was not successful with running the application thenMy guesses is that this patch cannot be applied but maybe you would have some advice how to do itAny information about ability to do user mgmt thru xmlrpc in Archiva will be helpfullBest regardsLukas +Rejecting +Committed revision in trunkGood work Apos. +In order to avoid mistakes like we had for this release I would need to open jiras NOW targeted to final to revisit the version ranges +For each term we would be specifying the same column compare term and value +Im glad to see you working on this! Well take it for a spinKlaas I got your points now +c production in the console +Robert I tried this against ER and cant reproduce are there other steps to reproduce or configuration that might be missing here +Theres no way for that logic to easily see that the data buffer has data in m going to play with a couple possible hacks but the long term fix is to get to enlist in our wrapper logic +Scott we rolled back our QA and PROD systems yesterday to and +Verified in CP +So the port was missing the path was wrong +Which version is this +Eclipse doesnt seem to handle this very well it creates error markers in the Java source even though the +Is the code still failing in various cases +I dont follow the need for a to launch the scripts and attendant complexity + for the patch +Do you prefer? v would run tests using hadoop This patch should use hadoop it local +Also like groovlets +test patch! Applied in SVN revision and +Please submit the patch +You are correct +test patch doesn make any decision on the no of tests run +Removed hardcoded strings and replaced them with literals from a resource bundle +can you do a testcase for this +Fixed on trunk revision +Let me upload a patch for deprecating the property also +That doesnt sound right +Im surprised by this behavior but it is what it is +Form this point forward please open new as its much easier to trackSending Sending srcmainjavaorgapachegeronimocacheSending srcmainjavaorgapachegeronimocachecommAdding srcmainjavaorgapachegeronimocachecommSending srcmainjavaorgapachegeronimocachecommSending srcmainjavaorgapachegeronimocachecommSending srcmainjavaorgapachegeronimocachecommSending srcmainjavaorgapachegeronimocachecommSending srcmainjavaorgapachegeronimocachecommaioimplSending srcmainjavaorgapachegeronimocachecommaioimplAdding srcmainjavaorgapachegeronimocachecommamqimplAdding srcmainjavaorgapachegeronimocachecommamqimplAdding srcmainjavaorgapachegeronimocachecommamqimplAdding srcmainjavaorgapachegeronimocachecommamqimplDeleting srcmainjavaorgapachegeronimocachecommimplAdding srcmainjavaorgapachegeronimocachecommudpimplAdding srcmainjavaorgapachegeronimocachecommudpimplAdding srcmainjavaorgapachegeronimocachecommudpimplAdding srcmainjavaorgapachegeronimocachecommudpimplSending srcmainjavaorgapachegeronimocacheimplSending srcmainjavaorgapachegeronimocacheimplSending srctestjavaorgapachegeronimocachecommSending srctestjavaorgapachegeronimocachecommAdding srctestjavaorgapachegeronimocachecommamqimplAdding srctestjavaorgapachegeronimocachecommamqimplTransmitting file dataCommitted revision . +Well open other issues for fixing concurrency and authorization around dropping and creating databases +I used for loops instead of iterators. +Just needed to do some more extensive cleaning and +Looks good +Great I was missing this feature for years! At least now I can see which document were not indexed +Will patch vm files appropriately +Lets mark this fixed +Pardon +Can you please rebase? ThanksGiven this is not an issue with should we still commit this? Im happy to but not sure its useful +create +Thanks +In a second step it could be made configurable by the to print out more less nothing +and Response has been refactored to keep metadata is not complete yet +Verified by Mark. +captcha and email is not part of cms workflow +Testing in progress +So I fear its a issue +Not sure thats the same because is released with warnings it may eat data +and after that periodically logs that Finished hinted handoff of rows to endpoint +IMHO its hard to get the point if you get warned in the nd page only since its on the first that you selected the application The nd issue is that the limitation to letters is not valid +I guess I could create a configuration that changes the view is one of the places where there is actually room to have these names. +Ruyue +Improvement nearly +If so please close as Out of Date +as I understand this wasnt fixed but instead not fixed due to not using the verifier any more +testcase was not able to reproduce the issue i that there are missing some parts of the puzzle please feel free to reopen ticket with addional info of what im doing wrng with my testcase or even better extendmodifie the testcase +xcoulon your updated wizard brings this back right +We should either allow Facelets and other rd party dependencies into Tomahawk or allow none +Is this the same as +This issue is minor since it shouldnt actually crash anything +The difference is F related to rasters and related to images that are used as paramsI think it is better to use fix that Ive suggested +Also if its really easy to use it may be used for more than development +committed a fix for into trunk and M branch +this patch change the regionserveripcmapreduce code to enable the bulkload remove the old was saying that if you take write lock there is no need to take read lock at the same time +Its not committed to trunk +note came across this looking at the case a little more carefully to examine the contention on destinationMap +Should be assigned to me after fix without resolving +Fixed on commit There was a conception issue on how configuration is instanciated in the php plugin Configuration was instanciated even if the project was not phpThis has been fixed and now the should method of the sensors does not require to be instanciated +This is a known ant problem +It worked well for me +Should I attach a small postgis table of polygons? +Fixed package name creation based on schema names in Git ID eccfbdeecbefaacc +All those interested in support in may also want to be watching this issue KonstantinWould you be able to send me a copy ofI know you uploaded it a few times and I know I should learn how to work out but the fact is that it will take me weeks to get to that pointYou could really make the day for a poor guy in Holland who spent a wholeweekend getting AXIS work against NTLM only to find out that he needsThanksKind regardsMarco BergmanHi here is the full snap shot of really working project. +There will be a socket connection open per slave +Could you attach the whole example +Did anyone have any comments to add into this one? If not Im going to the remaining s on the page and possibly work out a way to get the main site updated +Hopefully a chapter at a time might be more workable for all +Thanks fixed. +Will clean up quite a bitThat being said is too lofty a goal to land this +However your issue seems different as the classes in the client module should be in JDK Can you try the following and post the outputmvn vThanksJim +The SNMP module build process uses the prefix covalent in the names of the SNMP DSOs +fixed +Post the question to +Fixed in Branch at in +Dont mind found you be a bit more specific on this issue? Nobody has to build the docs as builds are on the website in both htmlpdf formats that are built nightly +If that API is too tied to DA then we dont need to use it but its concepts are still valid so can they be reused? Would it be worthwhile? +Still your optimization may be useful +Thanks Anurag Knut Narayanan and ystein for the additional clarification +fixed in trunk and in jbds +But thats fine semantics +Another improvement based on perf tests config changes and for compactions +will need to mock all of the dynamic controller methods provided in to be viable +Ive seen this done for projects moving from Bugzilla to JIRA and I did it at my day job when we switched to JIRA +I wonder if that will happen here? Perhaps log it if we get a null out the linkedlist? Do it here and in since its not supposed to happenCould it be a synchronization issue? I havent looked? Maybe the linked list needs synchronizing? All access to the list +My observations are Doesnt work with +Sent email to Chris on th October to request more information +Simple test for single step +No response + seems quite related +If there are no objections Id like to commit this by tomorrow +The cygpath thingy is done in the and files after HIVELIB is used +btw by making an attempt to read all the regions of this table and then use the start keys of these regions I mean get the start rows for all the existing regions of the table and then use them for job configuration + you investigation the javadoc for ui only contains uicore javadoc items +EJBs have a dependency on weld and they need weld to be up when they are being shut down so that CDI injection will work +Thanks for notifying! Ill check it to see how to fix this problem. +Ive fixed this in SVN +I tried to lookup for codes which construct parameters map for calling and fortunately I found itSee attached patch for detailThanks you all +order x GSTPROMO Shipping +Read the migration guide! Proxies are now enabled by default +This was fixed with +It basically only looks at types of methodfield even though serializers use annotation overridesIll see if and how this can be fixed. +Of course I do not think soIf you think this can be improved and Seam Team can improve it and make it available like the Seam keep this issued opened till the solution is availableI only hope I can process the manytomany relation as Seam way no extra codes for it like the approach in my codes +its not a console issue. +Got some help from gnodet +Changed Target Version to upon release of +All tests and precommit passed +I played with as many compression related features as I could think of and couldnt come across any differencesPlease correct me if Im missing something but I think we really can just drop this dependency entirely especially since miscdepends and shlibsdepends are in use +There will soon be little demand for this feature since Dreamweaver MX allows you to easily add tag libraries by importing the TLD file. +Testing with Designer CR and problem is still there +This method should be called in the method to make sure we really cleanup everything and do not have to rely on the objects being gced properly when the context is destroyed. +Somehow CM is ignoring Hive Client Configuration Safety Values for +Set fix version to unknown +In particular I plan to make a new class extends and throw that +Fixed a few spelling and committed +We faced some issues related to length file in Hlogs +conversion now happens within the and this bug should no longer exist. +I guess Id hope that the code would be reduced in size overall but for specific features +everything looks fine except when running three tests the JVM crashes +Wendy see my above comments +Why do you suspect a GC bug? I didnt see anything to suggest that +Here is my recommended patch to fix that issue +Patch updated with new mared directory structure +I want to create an that uses clA for loading the module what would the code look likeAlso is there ever a case where more than module needs to be visibile by the classloader? I mean maybe the module and one or more additional modules? If this is ever the case it would be worth considering allowing a collection of classloaders to be specified +I see +It is better than putting those versions in the +Both Tomcat and Jetty build can display the navigation tree correctly in Chrome windows XP +A patch to Hibernate has been submitted +This bug is easy to fixBut if it stop release or something you can change it to new feature. +Thus at the current time theyd like to keep the Serde in Parquet +About the hurricane one more reason to switch to git The socket issue is interesting looks like NIO is not respecting the timeout value +Ive been looking around a lot lately at the unjaring and the code for and this would explain some of the difficulty Ive been havingI havent seen any tests that check this feature either +i tried it and with a date field and it didnt returned any pivot results for that date field +The problematic places are highlighted with red at +Yet for all of the other protocols the situation is different +on patch v +Fixed with rThanks +But Ill check it once you and Jason will be done with the Embedder +Alan I got your point +If something else already touches the your interceptor will probably end up with the bad sequencing of stax eventsOther than that I dont know how this can happenCould you change the position and compare the resultthanksaki +Yes it will be part of +This is draft +Still reproducible +So very illogical it changes but not really so we change it enough to make it look like a change +Jboss AS works fine. +I am interested in working on this problem +Heres a screenshot of the jruby classes that grow per each request and stay even after the first thing I notice is that your war file contains jars +They add little value to the document because the elements are listed below in the varlists +needs to be applied into current git tree +reopened just to set the resolution to fixed +I found the note in the manual about Resource and that fixes it +bulk close of resolved issues. +Should be fixed in shows that the Choose devices windows displays aboveon top of any Video windows +Oh sorry I have found the encoding IS CORRECT +We need proposals in the wiki and at least some discussion before new features are just tossed in +Id like to move those methods into module and make and modules to be ORB providers so theyll have only createORB methodNow method createORB has arguments String host int port boolean server +The minimum bar is for people to deploy an ear containing a war and ejb with the ability for the webapp to lookup and execute an ejb +Integrated to Thanks for the patch HimanshuThanks for the review Enis. +Could we have an optional parameter to specify the jar suffix instead of hardcoding it to sources for the sourcejar goal and for the source goalThis way we avoid any backward compatibility issues +Hi Tejas +Our XHTML validator should not varify plain HTML files +Added Equator interface into functors in svn r +Children are not expandedcollpsed if their parent is expandedcollapsed nowThat is why out of date. +Thanks Ludovic Ill take a look +to commit +As such Im changing this bugs severity from Major to EnhancementOne of the committers may wish to change the target milestone for this bug as this new feature is not planned for the release +Additional jars can be registered and can be qualified from hereAdditionally and are loaded if present in Pigs classpath +Local test passed +Given the decreasing average skill level of users and the defaults on common distributions this issue is becoming both more common and confusing more people +Thanks for the review Sidd +fixed rThanks! +Well try and get it done for +I am sure youll have issues has been additions to the file +Should now be working +Ill make this a required field +Thanks Oliver +Core should start up and work fine with out it the Active Container Profile should rather append to the current Context like the SuiteClass in SPI work +Yes Im using attached testcase castor may try the trick as well but you said earlier that it applies to M is that also true forAlso as I said I managed to get it working wo any change by reversing my patch and applying your patch to the class +right now this issue is pretty much a shot in the dark. +Rob is this another bug caused by the regression of settings not being fully persistedsynched +The latest patch has the lines wrapped +Thanks for the time to commit +YESSSS +I think there is no problem to get it in later +My machine HW specModel Name Pro Model Identifier Processor NameIntel Core Duo Processor Speed Number Of Processors Total Number Of Cores L Cache MB Memory GB Bus Speed John did you pull from a wiki dump or use the random indexeditNM that explains your shortened table no wiki results I go it +A plain node reference is supposed to be a that redirects to another one and contains itself no generics information +I ended up using the text representation SHA in Avrobase for my schema keys +Clearly it does and should default to false so it wont harm anyone if they leave this alone +So we created a new one +Sorry for my bad Comments in the xjb File please delete the something more weve renamed to due to a given from xjc so references to! +js filesTo add a fallbackStyle property into Session would add complexity and would be hard to implement this as there are places where style is used in a different way like loading in inside property files +For the findbug warning it is due to a dead variable +I attach an improved version with property sheets for inheriting compile finally got VS installed and could perform the upgrade myself +Committed for in rev +Feel free to switch reschedule if you feel appropriate +Ive addedTODO mwringe in the doc where some inputdetails are needed +Note we scan annotations from dependencies and projects in reactor +It just worked please close this ticket I will reopen it when it happens again +Graphene is now responsible for creating Drone object more precisely sending events to let Drone do the work +Jerry the closest I could find to this jira was which hasnt been worked on since it was created in August +cleaned this up a bit and applied it I mostly changed the naming to since it seemed a little less redundant but I had originally done that naming convention on the build definitions so if this works we can go back and fix that up later +Log format now displays +Applied +Patch moves all the information from the page to the forIt does not make any of the corrrections that Dag indicated are needed just wanted a first step to be a faithful copy of the information from the web page +Thanks Jing! +You may not see the original exception because it was being logged at a detail level +Patch available +Can you please see if you can duplicate this issue on an Apache release +Your patch is merged +you would extend it to implement a concrete client +Isnt BOSH all about recovering from the errors you are describing? I dont think we will ultimatively have an implementation where such effects no longer showa bit problematic does not give a good feeling of how often this situation is occuring +The create component method doesnt assing to them new unique ids and the unique ids are broken if the amount of components changes +pdb files included he can explicitly specify internal property tested updated build on Windows and Linux xx in both debug and release modes and it works as expected +backup now being worked on in +stack Hadoop QAi will improve and do another patch which included test ut +I guess you can index directly +What happens when you use rn +If a path resolves to a Bundle Based Folder and a repository item exists for the same path the repository based resource should be returned instead of a +We cant afford to neglect trunk +Verified. +updating patch after commit +Removed some old patch with many thanks +If the same happens sooner it could be a memory buffer that gets full +Patch attached +when they get logged at the WARNING level their full stacktrace is still shown +This issue is fixed and released as part of release +Hi allThanks for the suggestionIm sorry but ive not read the faqSorry again. +I think there may be a Java dependency lurking there +Attached sample of the the proposed how Ohio is no more cut after the on trunk closing all issues that have been in resolved state for months or more without any feedback or update +Id recommend to review whether union or intersection semantics are expected before using the flag +Clearly the killing of containers on shutdown was done in the wrong place we can see if that can be moved too as part of this JIRA +To some extent yes +Since this is the same root cause as you filed I checked in a fix aebecbdcbbbcfbdffc for it +I guess there are not many users using or anyway even less of them using rse +Phase When new regions come or when regions are reassigned from a server that has failed the master will balance the load across the active region servers +Also not sure which version of you are using so either upload that jar here or just let me know which one you are using +the NPE is still a bugIve got a patch for that +This updates patch uses factory methods to create instances of the proxy classes and uses in these factory methods to selectively enabledisable proxy creation +And we dont allow to change the schema forShouldnt we be fixing in that case instead of restricting combining of partitions of different schemas in one split +looks good +I disagree with the assumption that errors can becleared up before you release the application +The language specific packages have dependencies on their languages as well as the rpm which has the thrift definition +I think these should be explicitly separated so its clear where this data lives +Is the Appender you implemented something you think others might find useful? If so please consider opening a new Jira and attaching it to it +Yes indeed we can close this +done +Thanks +apache tika patch includes many changes +reopening to change resolution type to Wont resolution to Wont fix because its a known issue that wont be fixed +Apply that if it is appropriate +This is a patch which adds the xmlspace attribute for source to the DTD +The second and third messages hit the hello and return a of size +What version is targeted to be in tag? There are other project dependencies on netty so going to this late may be a riskComponent code freeze is today +Does anyone have any opinions on this proposal? If not Id like to begin working on this soon so we dont have this blocker sitting out there for +This is incorrect +The changes are working well for the user who originally reported the repainting failures. +committed +Julian what are the exact classes that you have in mind +Created a new test Alex We tried to make the computation of containsQuorum as simple as possible to be efficient but it is right that it has the drawback you point out +launch by converting to the launch framework in eclipse instead of a pure wizard +This is slightly more of a nuisance if you developing in a manner +is committed and I cannot rebase on that +Using mvn clean install eclipseclean eclipseeclipse fixes the problem +Thanks for this additional test project +Please use the attached jar in deploy directory of your server configuration like defaultall etc +I think this needs to be changed to be stored on the and pulled from there +Committed to. +Buildr this but itll still be available in the jira db for those who want it +Attaching a patch that fixes the issue +I cannot reproduce the error +Chris This may not be the appropriate solution for that component it may require JettyBased on our current design for the servlet container shouldnt matter as the servlet interface is well defined +what do you mean by second process +Id like to find some time to add the benchmarking codeIve started to look into it but stopped because of other duties +Fixed in trunk rev +The reporter should Close +Fixed in Rev +Thanks for the pointer +But you have to remove the opportunity for the user to change the values for max connections and max connections per host +Maybe my machines is just too fast +Make the averages dynamic rather than trendsup +Unscheduling +we had this discussion already and agreed to use american spelling +in previous comment point should have been duplicate referring to this completed then +For what its worth the content of the corrupt index file is consist of s for the entire file +Sending content CTrinidadTrunksrcsitexdocdevguide Completed At revision +The key to the map was the element which was being overwritten on reuse +If we do a it should probably copy everything success or failure just like and suchIt seems to be working but of course feedback is always helpful +Closing issue as version has been released on. +A while back I saw someone who had proposed a compressed representation of KV that had natural prefix compression +Resolving as Wont fix +I will work on creating one +It would be a little work to remove the interfaces later but I do not see a reason to remove them nowOn the view functions I can see their value but not yet their explicit need in any submitted algorithms +May be we should forcefully set the state to OFFLINE or we may have to abort the serverBut for doing it we may have to read the data of the znodeSo can we introduce a config to say if it is ok to check the data and then asynchronously create znode in OFFLINE state +Based on the approach weve agreed upon in that allows separate values to be set for processing power and parallelism closing this as wont fix. +However that will affect other valid deduplications also so the user has to be educated enough to turn it off selectively or we turn it off globally in but give up performance in other casesSo using this config is only a short term workaround +Closed with released. +in +Bug has been marked as a duplicate of this bug +Our RPC system should be atop a API not a API +Agree I will add that +Also is there any advantage to this over the new methods which take lists +Yes it throw the same exception +This task is a forward looking task for the next version of the C standard +As I said on the MINA List it did the trickI have not understand before the real difference between sessionCreated and sessionOpened I supposeIn my business logic I need to stop early a session if the server is in shutdown progressso I close the session in sessionCreated +Sigh +Thanks Jacques that really helped! I thought was supposed to be an element in the file +If coprocessor bypasses the default action just wait thereIf we really need this we can modifycreateTable to pass a flag indicating if should wait or not +This is a great idea but I wonder whats the relationship of the new storage manager and vectorized engineIt doesnt look like involving any columnar operations +Thanks Dobri +It seems like the former would be a step backwards +We can run hbck againFor the second issue Rajesh mentioned about check the current patch is fine right? It handles both meta is not assigned and it is double assigned +Ill review the APIdocumentation and see if theres for a clarification. +Patrick I did not try to coordinate the client and server logs but rather drew representative samples from each +This looks good indeed +So in your case you could retrieve the Text of the message and then call which will reset the message bodies state to Read Write and then set your updated text on the message +The patch to fix the path to resolve the graph not displaying problem +This is how the generated screen looks like after adding file and applying the patch +I guess this means that it is working as expected +Claus it is a normal the object is created during broker creation but the run method is only executed when the shutdown is triggered +This should scale to support result sets into the millions +The error is because of the CDPATH setting on the target machine +Thus we just have no idea whether we do have a session boundary at all etc +And anyways maven would take care of the dependency resolutions +Thus if javah produces something else the build has been able to explain so far why this strange f appears so I have gone with the version produced by the newer javah that omits this +I think that we should support the version of a spec throught the product whenever possible +Let me know if you need more details on anything +I asked him about it and he said s logic was still evolving at that pointSo in theory if we can confirm that our Twister logic is sound its a matter of porting the logic intoNahi Since you filed this Im putting the ball in your court to decide what we should do +Is it possible to create it? Could you add it +Max WDYT +Sorry for the long wait +The first one is a bundle which loads content in the repository +Marking closed. +Ive looked at those conflicts and my patches can be skipped +The newest patch is of that form +Please confirm +This patch fixes the issue inAs I already explained to Ankit at is well and good but not enough +That is more important +Passes tests and nosetests +Thanks +You could use nightly builds to test. +So the question is what weapon do we choose so that we solve one problem and hopefully get the most collateral benefitConverting into an interface and subclassing them as you say seems as well as a major overhaul of the code base +For exampleFinding the maximum value within a group etc +The DATETIME isnt to the trunk andI will try to create the Maven BIRT repository and update the project +To get familiar with external interfaces of Castor and also the codeing conventions I suggested Udai to meigrate soem test cases from old to new test suite +But now I am starting to wonder +Will add more javadoc and unit tests shortly +I filed for this. +Ive noticed the same behaviour at bloglines for the plexus space and other spaces hunch is that it can have something to do with confluence beeing restarted +Ron this got +For now the components which could be defaulted to Symbol values are in others need some changes in the framework which are being discussed at the moment +Hi JakeI fully agree with your proposal! Just created to extend the Test FrameworkRegards Roger +Keeping only +If is configured than all origins are allowed and will be returned in CORS enabled responses +Im going to commit what I have +This will be set to true if the element is element qualified +Best regards by Andrew. +Talked with Tejas he get more observations than listed above +so it is possible that it is leave this open for bit more if it happens also on some other i will close it +reduce logging verbosity make shutdown more like other hbase test cases +however actually its more confusing right now +but like you said outside the scope of this patchTo add a datapoint about testing the user who was experiencing this issue inside tomcat tested the patch and reports the problem is indeed fixedIll commit to and trunk shortly +However +looks good will commit if the tests in trunk running tests for in trunk and +The ajax theme requires that the onSubmit returns false so the entire page does not load +The actual patchNote that this patch contains bugs and is not the actual committed version +Integrated into and trunk +If you happen to get this in before we release beta just update the version thanks +I didnt remove the comment because its actually in a block of comment containing notes on why the call was removed +Yes it is and its very hard to reproduce +Fasial has tested the current code on Linux and besides one minor fix itseems to work the same as before so at least nothing seems to have brokenduring the reworking +Ok fixed +Fixed under svn revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Is that the caseOr is it just that the test case is invalidIll check if the test case is correct or not. +We dont need couchfilerename or any of that stuff +But then I thought its better to fixadd that when needed +If the caching is turned off why is the template null check and if not found is it still added to the missing template cache +close this. +Attached are the files that reproduce the OOM that proves that OOM is related to Advisory on present on AMQ and by SVN revision on and on regression test try harder to avoid interfering with other test cases +Better patch added that fixes highlighting of the Orders tab too +Its just the package we picked for objenesis + on patch +Not enough Committers have enough experience with it to promise back compat at this point I think +yeschecking whether to enable TLS extension support +Going to commit this in as wellRan a few times +implemented leaving open until I add added +A line change +Just an addendum to Konstantin converting to junit would be tremendously useful but it wont be trivial since the tests use a custom framework to execute the contents of xml files +picture of +I have tested with a snapshot from July and it seems to be fine nowBut youll want to avoid crashes as a Kaha recovery on a message store after messages have been queued and then removed takes about aI guess all of the in the Kaha log files takes a while to scan +Couldnt reproduce +Thanks Amar! +But recently the st part of this was already backported and has many changes already. +Review as part of EAP QE +However for local backend just the logical to physical translation should be able to build an executable plan +The fix was to escape the funky characters explicitly in the testThe line I changed had the following comment should we the cedilla here for cleanliness +I just committed this. +An exception is being thrown +No this is for transactions over Hot Rod since it will involve the client participating in an XA transaction but the server side just acting as a recipient for such commands +to what Bruno has said thats the way I personally always interpreted that particular chapter and verse of the everyone for the input +Committed revision +But its confusing that can be used to write postings yet not just because the postings format hasnt changed +They show up as binary files +I have used the and there are a number of tests and samples that use it as well. +In that case we should fall back on a solution like the one Aleksandar proposes +Utimatley I would want to do this augmentation in the layout could be the following features +Called as thats the issue I was looking at when I saw this related but separate issue +Please that is present in buildlib directory + Username emberson +Diff is synchronized with the current SVN trunkLinux EMT build warning and build test run OK on windows and linux +Test failure looks unrelated port in use error +You want to file separate issue for +Thanks Xuan! +Any comments on my changes to your suggestion Thomas +Fix and add unit completion entry of to prevent memory leak +Have you tried this using the failover transport? A raw TCP connection wont reconnect on its own +My dev environment is not currently in a state where I can easily do anything with this and the problem is not giving me any pain at the moment +The same spec lack for such a core seems not to support it. +Ive renamed the meta table to become bq +Merged r from branch into branch jBPM issues resolved in a released version and reported by me. +Yes this can be worked around by our and yes it should be possible to change the build scripts to publish via Nexus +We cant do too much here from myfaces side +Fixed in SVNAnother nice catch +Ill have to think about it +I would not have a problem with you taking a couple issues +I restructured the directories slightly and also modified the sqlline launch script to not need maven for classpath +Added patch that projects node types onto jcrsystemjcrnodeTypes at workspace creation time + needs to be taken into account when doing date math +faces +Also thanks a lot to Daryn Owen and Eli for the reviews +Switched to jbossxb +Id prefer to only introduce a dependency to it when it is out of incubation +I would really like to see this fixed +although I can only imagine the potential deadlocks +Patch looks good +fieldName bruttogewicht type field date jpa class +Ill post a patch that removes the translations +it is possible to close this issue cause there is a feature which is in description but create a new feature request for Run As for +HiI have got the latest from DOTCMIS and tried creating a user but I am unable to do it +Thank you very much for the refactorings and support on the first patchHere is another one with the pending algorithms +I dont think your patch will cover all cases but it probably cant hurt so can probably be easily added +Yes Ill do Submit Patch after making one key change that Carl pointed out +It is of course not an official Apache release with all of the associated extensive testing and voting but because we only typically backport low risk fixes I think it is a generally safe thing to do and may be advisable in your case given the seriousness of this issue especially if you think your users are likely to stop the TCPIP Stack while network server is running. +Thanks for the deep details Ill study the Spring MVC Portlet code that rely on Servlet then Portlet loadingFor your proposal I let you see for now the priority and need for resolutionI consider it as a wont fix for and a may be for a later versionIll post in the forum an information if I got the Spring MVC Portlet to work correctly. +Tests running +Doing a cvs diff u makes applying a patch much easier +A graph showing performance of hit collection using andIn essence there is no great win in pure search time when there are more than documents +This is because in only the message ID is taken into account +Thanks Sean! +bq +There is still one remaining mismatch +Your sourcewas not as easy to compare +Derbyall runs cleanly modulo wisconsin noise +So all issues not directly assigned to an AS release are being closed +One thing I do not quite understand about it is not appropriate to special case individual request types currently the just receive a Request object and attached that to the corresponding key +Thanks Navis +Added a JAR dependency to Jettison JSON library +for the thingsWhat about using instead of Apache Directory Studio? We used it a long time ago and when searching for it on google all the links are pointing to Apache Directory personally like Apache Directory Studio betterAnd Im not sure users would understand such a moveApache Directory Studio is a pretty established branding now +Issue does not reproduce with +pulling this back to we need to merge jmspool with activemq pool get the fixes that are in both and ensure the result is generic +defer all issues to +Brand recognition is important +uploaded with minor changes +Closing the preference page works in alpha build and no error is logged when JBT starts for the first time +Reopened issue after comments from Christoph. +Ill try and reproduce it later tonight +I guess while the ones in SCCT test whether or not it can collate properly TSCR checks that the response it sends back is properIn any case this is just another one of my brittle tests! Because were using a different comparator results with tied scores dont come back exactly the same as before +Rebased on trunk and patch revised for resolving conflicts +Custom slicer is a pretty heavyweight thing for users +Hi AntonI cross checked on the same and found no exception for have already attached the logs which I am getting with this and Regards pls jars list +Manually tested! +Moving this to next release +This is fixed in SVN keeps track of unacknowledged and delivered +Closing all resolved tickets from or older +Please cross checkTIA for the contributionGreetsGerhard +I will deal with this error when back from holidays +Looks good to me +Please provide a full stacktrace for this problem to allow us to track down the root for the time being due to lack of details. +You are right +of the tenantMany of these changes are in our forked version of the code +Not sure whether thinking about it as a transient sink makes it any better? We were contemplating replacing an existing system with flume +The rationale for this is since the index does not support duplicates the only way we couldRemember that getting a lock on a deleted row does not mean it is a committeddeleted row it may have been deleted by the current transactionProbably the most straightforward is to follow what we do currently forunique indexes but apply it to this new index which allows duplicate nullsbut non duplicateI dont know the exact answer on locking it is complicated as we have never lockedrows to the right of a row during an insert before +Seeing unnecessary messages in the logs on the monitor can hide relevant messages and leading to other serious problems down the line. +This might not be a bad patch but I think ungetc needs to actually unread the full character and save it +Check it again +Craig thanks for those clarifications +Straightforward patch +This patch fixed the memory leakShold be applied on applied in revision +Fixed +This issue seems to be very old and no longer valid so Ill close this one as cannot reproduce. +Committed to trunk. +Works fine with Tools build H and EAP v +Added extra test for intermediate +pushing this out past pending critical issues with for the patch +This patch contains a fix to the problem +Going over the list to see what should be made final will add an empty static initializer to the rest. +Committed in revision . +After applying the patch my code generation works +I havent yet noticed any configurations for commas +It is trivial for the user to create the +We know that their puts are no longer useful but the takes might be +Currently table is only created if it doesnt exist +Stuff can be at many different places and it depends on the final resolution what gets actually used +I seemed like there was an error downloading one of the schemas +here is a test case for the named query suggested by ralf added it under srcbugsjdo is also a cumulative patch +There is a lot of work that still needs to go into this +Im not really a user +r + +It needs to go into trunk +This patch fixes the egress firewall functionality in the Palo Alto pluginThis patch is for the branch +SQALE model updated +r. +You cant inject a Seam logger into an object +This was on SUNs JDK on Windows XP +Thanks Hitesh! +Unpack and run mvn test with a local AS instance running +create was borrowed and modified so that the expected values would easy to demonstrate atomicity failures +Im fairly certain that this is a duplication of +It would be nice when the user selects install JUDDI during Geronimo installation if a Derby database named jUDDI would be created in varderby and populated with the db script which is included in the jUDDI distribution +can I pleadItll get fixed over the weekend +Seems to me like that is what most people want w this +WW Update fix version for +The output from mvn v there is Maven version version name linux version arch i Family unixI then ran the test in cygwin on Windows +Thank you GeorgeI fixed the wording as you suggested +Hallo Igorthe file handles are left open on each rendering cycle of the Wicket page +Cool +I mentioned it as an emergency workaround I cant use it here but your situation might be differentThe zombie thing is more an annoyance than a problem for us but our ops people dont like it when it happens and users seem to forget often during testing and debugging to kill their jobs +Feel free to close this issue and sorry for the a related note it might be nice to be able to track down for which class this instrumentation fails but its not critical. +I believe these issues have been fixed with other fixes +It deals with sorting all of that stuff out when the is returned +Yes sorry +looks a bit hacky doesnt it? i would prefer to have but a single save during the loginanyway +Will commit if tests pass +Just wondering if anyone from the Geronimo team can react on this +This feature seems really useful +Arggh +Corinne could we remove the option from the documentation +Jeff can you please clarify what exactly is missing meant +It blocks because theres only connection available and I dont think commons http sender implements pipelining +Paulany news on this +If you disable checksum how much improvement do you get +Agreed on all that Mike Ill try to do that forI also wanted to deprecate for a while but I agree with Yonik that its kind of a feature and we shouldnt toss it +Thanks Jarkko +bq +HTML default encoding xhtml +The patch looks good +Hi Thomas +This might be a silly question but how do I know which contrib tests failed + +Can you provide the code of your and classes +core tests +Did you ever try Tika AFAIK it uses the current version of pdfbox +Patch v addresses Matteos comment +Rajesh Yes when I changed the order in the assertion it passed then +Having both the top and sidebar is confusing anyway IMO +Done in trunk +Should be investigated to know how big an issue it is +Hi AndrewIt worked just fine on this new URL + +sounds good on thatthanks +HTTP POST does not have a length limit +When checking the failure this has been addressed in trunk +That worked +Moved logging back to the constructor Replaced unintended for the updates NarayananPatch committed at revision +The SNS consumer sits at the receiving end of the event driven consumer eip and handles messages coming to an SNS endpoint +Final ETA is still unknown +revision +Cool +So I have reconfigured the box websites DNS and rDNS so is on the default IP address and other names are on different IP addresses +The stack trace indicates an NPE in the PluginShall I move this issue over there +Applied patch with thanks to Patrick also merge the change into +So we are not able to using the latest hive releaseThis bug afftects only since uses a different pruning strategyIm wondering if still need bug fixing since has been released +testcase for input +Other widgets like Weather work as expected +The bug is fixed +Shouldnt we initially support only the charsets specified as must in and not let the user add any charset + push to +I think the code change is ok +What we could do in the future would be to build a generic commandline tool into core then allow plugins to plug in their specific command +Under current trunk version we can not even run the applications under the Uber mode +Knut warned me once about not closing in the patch I submitted for and suggested a there +Added a unit test that verifies that MALLOCARENAMAX gets passed through to the container +Ive modified the test according to my comment above and removed it from the excludes list at repo revision r +Rebased onto trunk and resolved some conflicts +If this is a temporary problem please let us know when it is fixed +Same problem for me with iBatis +Moving this to for now +Jessican can you check on this issue +Attaching the patch +Committed this to trunk and +Im not super keen on using ZK for that seems overly complicated +is our standard default password +I will also verify that it fixes the issue manually once the other tests finish +I checked the issue and totally agree to your argumentation +I still must check it +Tested on Liferay +If youre still having this problem could you download the distribution and try itMarking bug as resolved + release must have solved these problems please check with has solved these problems +Attached a patch fixing theCould you please review and commit the patch +jsf is java ee which itself is requiring +Aright +please find the final patch in attached file +Current version of installer just detects Mac OS X and use d without checking if it is actually supported +I am hitting a REST endpoint +This is just waiting on legal paperwork +The following command works with maven and sonarmvnsonarNext step is to upgrade the Codehaus mojo in order to use the shortcut mvn sonar +Ah that makes total senseGoing to put this into the if anyone actually needs it they can code it category and close for now. + +All the tests passed locally +Why? In RC this is still handled by the +The bridge is a good option for people who want to mix and match with SLFJ +I think this is a duplicate of +As Martin mentioned will solve your problem +modules are not used anymore in version so this issue is out of date +Rebased patch addressed comments from Nicholas + Remove duped code that was left behind when generic utilities were boosted into Updates Column construction for optional timestamp and thanks! +Merged to. +Since a new SPI interface called was added so it is possible to override the current parsing algorithm and if it is necessary do the additional steps to resolve other dtds or schemas used. +Attached latest patch could you please review the test cases +Thx btw for this great plugin. +It would be good to have a pluggable interface here that you could specify ensemble in the config +Update web site svn ciSending Sending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexcommunitySending srcmainwebappSending srcmainwebappsiteTransmitting file dataCommitted revision +Modified Summary Environment and Description so that they reflect the situation more accurately +too +Such a guide would be more than the upgrading notes in CHANGES +The throttler was fixed recently to use a which checks if the throttler is full eg you can set a thread pool thread pool profile on the throttler and it uses the maximum option +I added a test to ensure there is no regression now +Heres the patch for both the +It would be better to host the translation on so it is clear that it is the official translation +I have added a patch with a testIn test I added a assert in the catch block because if a assert is added to check blockStream as null then in failure condition the throwable catches the assertion failed error and testcase passedIn catch block I am checking no NPE should be thrown +Id prefer to keep the Groovy code separate from the Java code +Rustem please help to provide the patch as Ilya suggested then I will commit itThanksPrevious patch was corrected as Ilya suggested +fixed with svn revision seem to have removed the settings for svnignore with this patch +might be null when not configured so the above suggestion does not work +I dont think an seems worth it but I guess I dont know either +Daniele havent you just committed a patch fixing this old report +This could possibly achieved with the invalidate command carrying the puts address +As an enhancement to Seans suggestion Ive added the code to all getter of simple datatypes like int long string float and boolean. +Tested under Windows and this works. +Vadimcan you provide a status update? The CVS has no tag but the mentions it for th of March? I guess I underestimated my workload so there is no release before marchstStatus page mentions not thats why there is no tag yetWhen Cocoon is due? I think we should make xindice before Cocoon have now and there is an option to set in +cached but other places are hung up waiting on a ROOT +Instead filter file is added to child. +I dont use and IDE and I dont use hard tabs +Screenshot showing various reports that only contains charts +Pavel please verify this fix on the CR docs bundle and close the issue +I could not find a release for so I used the latest snapshot +It occurred with and doe snot in. +To verify that patch fixed the issue in EUT one needs to run jdtcorecompiler suit at least times. +Changed to new component names in the +project may not exist before importing project +Thanks for the review Tom +So I havent run the test on jdkPlease take a look at this patch and commit if okay +passes on my machine +It would certainly be great to iron out those differences but we ran into problems involving backwards compatibility when we worked on it previously +However both Debug and Release build dlls of the same name +Will need to coordinate with Burr to see what should be added from SOA +Rajika +See the file concerning the points implemented +Issues detected +removed because this is a valid query now +Ahhh +ahh +Understood about the performance benefit Im not saying that there wont be a huge performance I still dont feel that theres a reason to do this the way youve suggested and now that Ive done this I have to say it still isnt fixing some of the ugly s an improved patch +Ok cool +looks good to me +Are you saying that the request was filtered out? Oleg +Note API and IMPL are not released separatelyIll close this issue as invalid +Integrated to trunkThanks for the patch JesseThanks for the review Andy +I might be coming from a different angle but here is my take In general I think that dynamic revisions in the repository should be avoided as much as possible +Per Please create a new ticket and provide a simple test project that reproduces your issue +Ive committed this and merged to. +Attached is a patch that updates hsqldb version toNote that I see several test failures after applying this patch and am investigating them +Attached patch seems to work well +Old versions are no longer supported +tmplogsapplicationcontainerstdout tmplogsapplicationcontainerstderr Obviously my java +buildivylibcommongriddevbin +got it thanks +Should we push this until after the GA +Root is an access control idea and not a permissions idea +This rule just compute the density of comment lines whatever is the type of those comments +Hopefully we can come up with something thatll allow us to weed out the stubs more precisely +Updated files in zip because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +To give an extra element of closure to this item Ive just build along with the source patch patched the binary on my development machine and the problematic jobs have now been clearedThanks again KarlBTW Is that a record? Seven minutes after my bug report you had a patch available to fix the problem +Just wanted to bring it to your notice +Which is well suited to our use case which is mapreduce analytics of logs +is used in both client side and server side +Anyway I will complete the fix as soon as I can Thanks +Youll have to find a new scapegoat to blame this on Sorry +Noticed another way this could mess noticed another way this could fail and attached a patch +Hi NicolasFor some users the ability to use a partial or full checkstylepmd configuration defined in the pom file is a feature so we cant simply ignore what is defined in this pom file +I have a script to run the test repeatedly but have not gotten it to fail after over attempts +I had planned to start the release yesterday but my day job got in the way +We wont do any explicit facility here for Spring +If you use them we might get a better view of the state of things when the error occursI also made one little change in the Session Executer but I doubt that will fix anythingAre you running optimized code +Except i guess that some do use generate standalone But if you have a nice patch that would result in the same thing please attach it +Trying patch and the later both failed in hbase test log rolling +Would you for instance expect Hibernate to close your connection pool when you shut down yourI dont want to sound harsh but Im not going to change what I consider sane design just for pleasing one user +Would anyone want to try this +The Common API would have to change for a single platformThe best place to start would be the existing plugin and looking up what the intent is for the Panorama View since we use intents to access the camera. +Looks like two files that are part of the patch didnt get added +crc file correctly +Tested locally when in a secure cluster +If you want to just comment here with a strategy feel free to to me +Forgot that the filter finds all issues +Looping through dependencies was changed from first iterating through a list of all pom dependencies and removing up redundant ones to iterating through the redundant ones first and then finding the one to remove in the pom using a targeted expression +Duplicates about the dup closing. + provides the initial framework for readingwriting the new collaiton metadata in store +Fixed to M +I just committed this +Fixed in trunk +Test application usedsee comments on how to manipulate urlspeed +Julien could you reorganize the Content of the Eclipse update site as follows Sonar Java Sonar Java analyzer Sonar CC Sonar CC analyzer Sonar optional connectors Sonar me Connector Sonar Mylyn Connector +The inclusion of changes to was to show that the problem has been solvedSince restoring would be done in how about dropping the changes to in patch vPlease comment +You can use the file last modification date in that case +will be part of next out +I could see adding another hook to the filter to get the metacolumn data at the start of each row as wellThe combination of persisted meta data and coprocessors is a pretty awesome oneI have a few other items I need to finish up but Im hoping to get my hands dirty with coprocessors soon +Is anyone willing to look at the patch any time soon? Itd be nice to increase our confidence it is the right thing and that it will show up in future versions +Have raised queries a couple of times on the dev list and am now closing +HimanshuI have some questions about your design docbq +Forgot to say the OOM errors were on iseries I ran and derbyall with the same build on Linux without any failures +Ill take a look at the codebase in both of these +Integrated to TRUNKThanks for the patch MingThanks for the review Ramkrishna. +Putting into +How is it we dont have support for that stuff in the new +Closed. +so +Done +Hi BryanThanks for committing the testI am closing this issue. +pull request for master request dont know what the decision was on the transformation stuff but I dont think we should try and transform expressions +Lets fix it together with +Postponing for the next version +NOTE was added in the end of Details of usage section +This has not been fixed in Branch Branch or trunk +Anyway the patch was applied as expected +this has already been fixed in the trunk +But the security bean that binds the javajaas context is coming later in deployment +Get rid of enumerations unify logging drop Java EE generify all classes etc +Oops code falls over when fid filter contains a multiple of this is the change I made to resolve the problem +I didnt notice Claus had opened this when I opened +Right now it is very easy +Done +I converted the main app bar to a simple unordered list and styled it as a plain tab bar +bump to bump to +Id vote for one big report for all modules +All of these enum bugs you have raised recently are likely related +It would be good to integrate the enhanced tests of that patch prior or similar to yoursFinally I am am yet not fully understanding why the signature and encryption need new templatesMy hope with this schematemplate project was to decrease the dependencies to ODF over the time leaving them only as an example so other XML dialects could use this librarytool to generate their typed DOM tree or documentation +Sorry for taking your time with it +has been marked as a duplicate of this bug +Which version of Felix are you using or trunkThat particular stack trace doesnt make sense if youre using trunk because line is appending a string constant to the passed in buffer which must be according to the code path +We also dont have a transposeView which is very handy some daysyeah +A part of this has been committed now works inside of +Not sure why they failed on hudson +I have a strong interest in and fair amount of bandwidth for getting this issue resolved so feel free to contact me or whatever +Also we should make this a function and call it in debug too +Applied the patch in revision Thanks Nandana +Net and got the same result +Heres a patch for the first point David made I used an ftl and it might not be suitable as it doesnt really follow best practises but it should make reviews easier to work withRegardsScott +Wed like to reopen this issue +We may need to keep both NMPORT and NMHTTPPORT for the it need both of them to do the initialization +IIRC an optional value is consistent with even the MTOM poliy stuff +I was wondering the same thing as Bill when I asked how youre doing this experiment +Attaching changes which point at the release notes license and notice files +Id go for servercore +Correct we need something like this +I is also not possible to let the script access a local variable outside of evaluate unless it has mad available through the binding +closed shipped in +Im not able to run that nd app you installed +BTW fix does not modify anything but comment +I think this is a generic message which is not related to Unsafe specifically +and this patches should be applied together to get proper results +MichaelI plan to add an additional properties file called where people can make all the changes needed for the iut without even touchingI will try to come up with a uniform convention for all the classpath names +Tested with different locales it makes no difference +Fixed at revision According to CC regression tests finally passed on Linux x. +If not I would suggest make the comment changes and rerun +patch looks good to me tests passed as thanks for reviewing Patrick +Ive committed to the snapshots branch +Heres a simpler to use more pythonic version +We create a private static repository under speicifc domain name and make all the calls create a log from that domain name and repo +I do this in a hand built app for SOA purposes +Yes I also noticed things like that in the diff sent to the commit mailing list for instance +Here you go +thanks +Ill take care of applying it +disabled fixed disabled +This was done +Committed to trunk! +I had set them to seconds but thats not enough on some slow nodes +Thanks for fixing this Knut Anders. +I committed this trivial but very important fix. +Because of this I am unable to run the Flex apps created in FB using Apache SDK +The patch was not installed by the quickStart option of mavenIs there a or do I need to install +guessing at the fix version +Im currently testing it with hprof and heapsites +Yes that was a typo on my partWe should be able to compile against either Xerces and then run against either +Im marking as Wont Fix +hereby grants the right to freely use the information supplied in this file in the creation of products supporting the Unicode Standard and to make copies of this file in any form for internal or external distribution as long as this notice remains attached +This patch should fix the test jar to be able to work at it used to +I assumed its correct since it passed the unit tests correctlyhmm +Issue closed with release. +thanks for the extra info. +Same happens under you please download the standard Spring Framework distribution and try whether it works with the binary jars included there? The XSD files are definitely included there maybe its just the Ibiblio jars being incomplete +Patch has been executed from the srcmainorgjbossseaminterceptors directory cvs f diff u N is against very very old code. +Deployment to jBPM can be done in a lot of different ways the GPD deployment page just being one of them +So not the VM +Thanks +Thank you for the alarmNow we are trying to embed the config in update processor instead of loading it from out of the problematic snippet are gone +But drop database would remove tables only from new location +ok just making sure there is nothing lost with to commit +Well it is fixed so Im closing it +jawher can you provide a osgi setup that can replicate this problem? something that is easy to run and debug preferably using something like our Start class +Fixed for. + +more to deal with doc issuesif LVQ commit increases in size in my diff I will break it outCarlAll patches committed for thisAll three items have been completed +Shay you did enable bean B in right +Raised priority to blocker priority +Yes I agree with this worst scenario is not the most way would be to get a svn lock and then release lock when the tag is done +This should have been solved by simply adding dependencies on the message queues +RTC work item now have a quick fix to add imports +For an extreme example the escapeCsv wasnt intended to protect against security or weird issues in MS Excel +Messaging is the new enterprisean enterprise and please add and Changed to and since and would add to the perception that the had cutoffs like and +merged to trunk thanks Ted and Stack for the reviews. +Excellent job Doug Im going to refer people to this document a lot +Its written that its fixed in Do you know if a bugfix corrected it +If you still see problems please create a new issue +It looks like the patch that was tested was the first one out of the two which had an issueLet me update to trunk and the changes compared to last more tests +Yes I also think that these changes are also necessary +Additionally theres another problem Im experiencing with the +committed +Patch is ready for review +I also thought I remember shawn saying in irc that he confirmed that no code was reading this setting in solr anymore +To fix a failing test case I actually did end up implementing the proper resolution rules fro the codebase classid and data attributes of the object element +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Upgrading to the latest helped but did not fix itBut we are still seeing those problems for inserts through THRIFT +Plus I think the deprecated section can just be yanked out since this is for exampleeducational purposes and the comments leads to confusion +Lets target should be out for the vote in a day or so +to +Fair enough but given how often Struts users have problems with messed up PUBLIC for Struts I think getting the DTD on the site at the right location should be a pretty high priority +Blackberry support has currently been removed from the cli until further notice should be good by +Ignore my comment about contrib the patch is on contentloader Id be happy to have this in there +My to commit this patch +verified +forgot to set fix version +This patch creates a new method in for formating percentages and uses it in all of the progress fields in the webui +Bug may only be relevant to older Rails. +I think it is ok to avoid closing the PMF until the end of the configuration +The order in which constraints are evaluated is still undefined +Btw +The problems Ive seen in the past with this also solely were with Linux +Will test when we build our installers +Yes +Add a few unit tests +Thanks +Ill need to using an alpha build to fully verify +! Sorry folks +Its probably not ok since lucene build script will probably fail because of that +The exception will be marshalled and sent to client you dont need to wrap the +See comment from Jonathan. +I guess my position is that protocol changes should be backwards compatible in the same way that code changes are deprecate the old usage but maintain it for one version back and then be free to break with two versions ago +FWIW also doesnt work for me in m +Clearly spec does not provide that decorator must be injected into Decorators delegate injection points +Please attach a copy of to this issue to help us narrow down potential causes +New patch after rebasing to TOT +Actually in this case compile is what we want +Fixed on trunk at r will backport to when the freeze on as well +Test cases +Thanks Yongqiang +Thanks +I am glad you like it +Patch reduces the number of clients created +Verified at +Not sure if you missed a test fileI havent managed to my OS X build going yet +The problem Im running in to is that the JSON representation of a resource doesnt have the path in it +It is unreasonable to ask for access to all build boxes just so you can debug one build. +would be nice to have handy always +Unfortunately JNDI is not supported in all environments like GAE for example +Hi DanushkaI have done this improvement You can find patch with this +The odd part was that the code internally wanted to know the diameter of a circle based search in miles +I want to contribute to I have developed a patch of this task. + Bug has been marked as a duplicate of this bug +Moving out of +committed +A box could potentially include everything a toolbox is always tools centric +Once a release is classified by the PMC Members it may be distributed to the general public on behalf of the Foundation +This should do. +I have checked the in trunk and there is no such alert I will analyze and merge the trunk to FP branch +Tested on Linux without problems +I notice has some code in common with but I assume they test different thingsWould it be beneficialpossible to run the new tests for the other encryption algorithms as well? +But this one was simple. +It is caused by both and +Hoss Mark thank you for your comments +this patch adds clearAttributes to chinese fixes the issue +Great work Scott +I also share the worry that we need to solve both the and the inlining of TLS access simultaneously +doc generator +Well either Base encoding the data and sending it all together takes longer than requesting each column individually or its the other way around +This problem occurred after the port to AS so its possible something in the port broke it or it could be a further incompatibility with AS +Ill try to reproduce it and will tell you the details +Fixed by +looking into using nullability annotations +If we maybe yanked the being used in is and replaced it with one of Geronimos creation that might work betterAssuming I understand the issue correctly +This feature is not supported by the frameworkThe has been some discussions on the developer newsgroup around the issue of persistent page properties which probably relates to this issue +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsernodes +Checked in by James +It would seem to have been missed when updated the product doc +Can the command output be along the lines of what we see on linuxRHEL throws thisls doesnotexist No such file or if patch is approved I will make these changes while adding testcase +Bulk close for Solr +karaf cleankarafroot featuresaddurl mvnxmlfeatureskarafroot featuresinstall +Hope that helps +I can reproduce the bug tooIn my case this results in broken encoding stringist keine gu Zahlwill be transformed toist keine g bug happens for us sample project. +I attached the file with DEBUG lines around the assertion ERRORThis is just an excerpt from which is much larger +defer all issues to +You should try the to examine pdfs its easy to use and will automatically decrypt encrypted docs After overhauling the font rendering stuff I assume that some of the encoding code has to be refactored as well +patch file +Thanks +So +This issue is reproduced by in master +Please check that it was applied as you expected +Push to a later fix version +sounds like a JXR bug +Heshe simply hovers the mouse over the component and presses and releases the left mouse button +All the images are generated correctly for me +In particular we should try not to change it incompatibly more than once +classic usage fails +Sorry for the delayed response +Upgrading to solves this problem. +The scripts verify that invoking the trigger after the invalidation causes it to be marked as valid and updates its lastcompiled timestamp +However it cannot obtain a lock since it is held by an application thread which has been in turned blocked in +Do you know if Twisted is comparable to for ruby? Im curious if you have any thoughts on whether or not the approach youve taken would be suitable for +ve just committed this to the branch. +Okay +I propose doing a hacky fix for in which we just instruct the use of mvn clean package mvn assemblysingle possibly as well +Yes your expectations were correct +Attach patch contains Synapse configuration for the sample Documentation on the sample +That would have been ideal +Hmmm I agree that we should change it to void +Added info rfom +for y +Updated patch to account for recent truck changes +safari does not seem to have at all +It is actually their top feature request +Actually you can compile gwt with ant in this mode it will display stack trace with real classes but wont display the log panel +Please update the wiki with the command to build the release packages and also the website you will need to update the section that describes moving the API docs as wellThanks Mike! +This patch was tested with jmeter and a profiler and proved to increase the server throughput +Ive set the queue size to unlimited in Revision +Committed edits to and to TRUNK +In this solution users will see the types of complex columns instead of a java class which may not be wanted +Committed at r. +Committers can you please review +Could be an issue with permissions +CTF runs fine +Downgrading the urgency of this issue + for Robs comment +probably can be fixed together with related bug +One more file is included in this patch +But there is definitely a bug +RebaseSeems good locally but let me run it through hadoopqa +Thanks for looking at the patch DagCommitted revision +it could use some better error handling and maybe replacing with ant. +I cant really read the file so Im not at all sure if this also needs to be removed from there somehow +For the example flusher can we see it +This is not optimal as nothing else in the REST controller uses anything +Attached patch for and trunk +Debugged more +Committing both these changes together will keep the window for which HDFS test fails small +Back port? I think thats fine +I factored large string constants out into resource files +Attached patch moves back into the main test tree and fixes up the mocks etc to compile and pass against current trunk +Susan I will make available a snapshot release for within hours for you to use for testing +started impl in implementation is a good enough there already examples of how this works or could you provide one +Fixed on trunk revision +This patches fixes this I think +Thanks for the reviews Suresh +Could also pass it along to Daryl +Marking closed. +Thanks +Closing as duplicate holler if disagree. +disabled the connection pooling tests if we are running in a environmentAn alternative fix would be to move the if clause to the suite method and simply not add any empty suitesPatch ready for reviewTested with Advanced +Ill add a check if instancof andinstanceof or something like dont know if I fixed the problem but I changed the IF statament +Username bvpaj +Move some member variables and methods in to +Then he would continued with the second wizard page as it is now +i think it can we did find that very occasionally we could get a successful render but at least of the time it would fail completely +Thats one part of the issue +There is very little to the classes but Im guessing youll want to mod the createfind user bit +Thanks Kris +The attached patch fix the issue but possibly more changes are required to resolve the issue completely +I was able to reproduce this using speculative execution and killing the job at the right point when it has speculative tasks running +I think itd be best to open a new issue +It is not supported to navigate a path expression in HQL +And then we can all give our suggestions and vote +I fixed lineage +Or if the broker sends ANONYMOUS to the Java client it should just skip SASL negotiationNot sure which option is better +committed in trunk too +Agreed +Both have a bunch of metadata fields defined +We need to fix the namespace issue +I dont agree with the statement that its confusing for the user +the variable search must be correctly set in +For example when we solved the problem lets say it turned out that one of two cases could be truex and x x and x In this ambiguous case I was setting x and x to +If its ok Ill commit the fix to x and the to trunk +Attaching the updated patch which implements what was discussed above +I have seen the SEVERE errors when the JSTL jar files were missing from the web app +Fixed a typo in the original patch that used the wrong property using bottom as default value to ensure proper backward behavior +The notice looks right to me +Push this to since there is some issue with Cache itself dealing with pojo that is not serializable +A patch for review +And then i will attend to remaining outstanding issues with separate commitsissuesWhats next +and your alternative proposal is +Ben Can you take a look at this? Do you think this should be a blocker for +Related to this issue and the fix for the page flow overview now uses a simple for the form +David do you mean to simply comment the example request handler definitions? Or something more formalI can see having request handler level control over whether content streams are allowed but thats of course a bigger change than simply adding comments +Closed after release. +It should replace the one in srcgrailstemplatesve attached a patched that resolves this issue. +We accidentally opened the file in rw mode +I rewrote this code and the core dump is gone. +With the problem is not reproducible +This lets you preflight check the operations +Just to double check and all jars are what we want +Thanks for the patch Michael +Since there seems to be no more feedback on this issue and I cant reproduce it Im closing it. +But I think the patch goes into a wrong direction +Dont think there is any code for views and constraints to get recompiled +If you dont get to it I will try to next week +If the error still exists it will be helpful to get a s Geronimo the work onBut the problem arose on within a test without geronimo I only mentioned Geronimo to prove the are fine +Now dynamic validation is used as default +Updated patch to replace with in for me +Bulk closing stale resolved this patch ever included in one of hibernate releases? I still have this this error and Ive tried versions of hibernate core up to +After investigation I found this scenario will trigger version number change event of app client facet in WTP +Resubmitting patch because seemed to be broken +Max Can you confirm that me no longer needs this on disk to work +forgot to say I checked out axis with axis tag before developing the patch +Thanks Knut +Attached is a patch which sets the system properties and on the test process and the networkserver process when started from the test harnessWith this patch I could successfully run derbynet with norwegian localeThis patch may not have any effect on all +We need to have something in the docs so that people can see we support BPEL +To avoid this I would need to specifically test to see if the resposne I recieved is a which feels rather tightly coupled to a specific implementation +Thank you for the patch Colin. +Also needs to be in +If you dont do FOO but for example PI then it works +Going through all the methods every time may not be efficient +If there were to be an error between those two events that could screw up the index +Ive just tried to run the same tests using and everything appears to work fine +Nested comments are cool but we should really follow C C and Java on this one and not support nesting + +This jira and documentation talks about using substitution variables in the value of workingDir but the patch did not have that functionality it only exposed the ability to set the value of workingDir +Most are visible within STS under srctest packages +Can we merge to as well +Create N java files with seam components in the project and refresh +Thanks +Committed +Shows the output in json +It has both a general overview and a walkthrough of implementation +Can we remove all the comparators from and leave the ROOTMETAREGION ones in the server moduleLooks like that would be a bigger refactor but maybe a necessary one +The other changes are castingenumerator movement issues originating from the conversion +FYI +I think you must have the wrong poms because it works for me and the nightly build is clean +Assigning to Flavio to document this recipe +Will post RB link in the Description of the JIRA momentarily +Some facebook related stuff found in the patch and got into trunk +Try mvn clean test Dtest +Verified +Thanks for your report. +showsql reports columns placeholders +On laptop Firefox is good enoughI am just trying to report the problems I encounter so that other people will not have the sameOtherwise I survive the development phase problems more or less +A fix is in SVN + +HelloIt is an additional question related for this problemWe have some concern as below If we upgrade to latest version for the Struts +I already fixed the errors on Drools trunk so they will be included in the next release. +Here is the patch i ran tests on all databases except for db +Please see exiting pattern +I will try this +I also adjusted to use an int instead of byte and moved all that code to and cleaned it up a bit +Patch looks good +bulk close of resolved issues. +The resolutionrules use a search path and thus can handle both the use of aliases as wellas the use of underlying column names and distinguish ambiguous casesdepending on whether the column reference was qualified or unqualifiedIt seems like something similar could work effectively for column referencesin GROUP BY and HAVING clausesIll have a look at the work we did in and the queries inand see if the same or similar techniques could be used here +Fix appliedI will submit to a vote. +Should make a copy of the Scan object passed to and remove the HEADSUP +Can you attach a screenshot please +On Windows NM can use s to make this happen +I dont get it +Sounds good to me. +That patch lgtm +Thanks a lot for the contribution Andy. +Are you able to run +Yes it is the same +component component is a word from the XMPP domain +logj is an alternative but it has some issues too +Will drop from patch for now +Stack hmm +See attached +Ok I committed a fix for the resolver verification aspect of this issue too +When an issue is resolved with resolution the issue must be automatically unassigned +Committed at subversion revision +just confirmed that M supports BEARER which is default behavior of so i think this issue can safely be marked as resolved. +For fixcommit for and releases you will need to work with IBM service channel to get this fix in these releasesAlbert LeeCan you provide a more concrete test case and the conditions reproducing the failure +One cannot design useful features without use cases +Thank you for the provided patchApplied the patch in revision I also adapted srcmainjavaccxpath which generates +Is anyone using this target in the autotools build? I havent missed it in the builds +I havent had time to look at the patch but the original intention was for there to be a lock managerusing multiple +ar +Applies from src with patch p +jar as before I am ok but keep in mind as we add more contribs we may want to separate the CP namespaces for them to make thing easy for debugging +r adds the parsing +Victor did this validation get fixed via other issues or is it still missing +I agree with changing this item to an actually ran across this problem while creating a project where one of the will use the Spring Mail integration but not contain persistence code +Yeah i agree with you that we have this problem +Resolving +I committed this +I know this is frowned upon for changes to existing code but was the only place using a style of brace positioning and variable names +The patch attached +Ill backport to +Julie this is fixed but I just pointed to the API JAR that is available on ibiblio as part of +I think it is save to at least change the priority of this issue +Im completely not aware of it +this is just a reminder that I think if we go this route we will need caching similar to what Uwe added for back compat so that construction performance is not bad +This patch allows the deployer to decide how much he cares about predictability and server origin so I dont see a reason not to add it +Your wish is my command! pw is now PROTECTED +actually it is used +The patch doesnt fix the issue +Now the Queue class is a simple data class with accessors +Fixed in r +From the explanation of Mike this seems an invalid issue +libdeltaclouddriversecin eachusrsharebin +Probably I didnt understand something here but I wonder why it should be proxies and not simply the objects themselves that are returned by these new methods +Resolved with incomplete description without any further informationWould you please at least tell us what information you need to complete the description +So if others want it you probably want to hand if off to someone else +Since this issue has been Resolved I presume it will appear in the EAP release? If so please set the Fix Versions field to EAP so that its Release Notes item appears in that releases Release Notes document +Added a revised patch following the review +Is there overlap between this issue and what Vishal is working on over in ability to support includesexcludes list in Hbase +But that would look strange and the user wouldnt understand why those operations are not available +Depends on and m working on the C backend piece of this +This patch is to get a value covariances and errors in estimated parametersThis patch is based on the file in because I cannot access subversion repositoryThanks +Proposed fix +A more uptodate version disregard the last file posted this one is cleaner + +Given Chriss comment I agree we can use either SHUFFLEBYTES or REDUCESHUFFLEBYTES to avoid confusionI also agree lets postpone the new counters before we fix the problem inWe still need to figure the story about compresseduncompressed bytes +Nice +I see +I understand this is an exceptional case but it is pretty odd that we are doing a trademark search what about the rest of the udfs +crawler operation met productPath stagingarea met met actionNames Unique +I noted Craigs comment about the case where one might want to actually serve xml documents +This is actually an incompatible change +This issue is fixed and released as part of release +Mithun Any reason the cleaning up of m repo fixes it? No other way around it? This will cause all builds to break if we check this is as it is +Please update the README file and include also in installation steps +Thanks Ben for scheduling this issue +I would suggest to backport the patch to branch +Have at it Sir +bq +usually relates to PUT and POST requests so you can attach it to +is improved and it works with both the leading slash and without it. +Then use the same generated code to generate the schema using JAXB schema gen tool +Im sure someone would have noticed by now if it was failing with just threads +Calling back to the failure function if provided would also be a plus +The issue has been addressed using the suggested fix updated in revision . +The bug doesnt appear any more. +Marking as wont fix since it appears this is a Ruby stdlib bug fixed in s stdlib +The example that generates those numbers which illustrate the problemHowever this is not a blocking bug just cosmetic +I see this file is in trunk but not included release will commit this after new XSD diff +I committed the better logging for next time +Hive Unit Tests passed +Im uploading a dumpfile with modified usernames to eris now +if thats not the case feel free to reopen. +Ive reproduced this issue using +Anything it has missed or any issues in the implementation? Review comments patch looks the required check so it fails in compilation if for an annotation attribute which does not have a default value the value has not been provided. +Simply commented out the line and the problem is fixed +Hi MedhiNow that you are a commiter and mockito is used is there a reason for not applying this patch +Logged In YES useridCalling deploy is not a workaround as this deploys the jarthrough a seperate class loader and can lead to and +Filed for the NPE issue and linking +What version of HDFS were you running? Id like to reproduce it +Ok I would try to make one then +Ive almost completed the dialog +verified that it is correct now +newbie error +Well have to look at your system to figure out why this might not be working for you +yeah I like suggestions or +The regression tests passed for me after applying this patchTouches the following filesM javaengineorgapachederbyimplsqlcompileSmall edit to enable the literalsM javatestingorgapachederbyTestingfunctionTeststestslangNew test case to verify that the literals work againM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterTweaks to other regression tests to reflect the new behavior +HXTT DNS settings for mx with the hash code we see that we have different query object if the second query would come from the query cache it woud be the same hash codeand second in the dump we see that the query is not from the query cache +This current patch has an error related to the get metadata request for inlined gadgets which I have not had a chance to track down yet but should give you a start on refactoring +The mat view refresh is given at transformation query using ttlXXXXX Some tables upon initialization get Assertion Failed however if we explicitly hit these tables first it wil l be loaded alright +next time please create only one issue listing all bundles +I will replace it willIf a BADVERSION and we are not retrying we should throw the exception in case other threadcaller set the same data the same version alreadyI think we should check the data only if it is a retry after a connection issueI think the patch doesnt break rolling upgrade because in retrieving data we use the dynamical id length from the data not the static copy +Thanks Enis! +Editorial pass over all release notes prior to publication of. +Bring back the node to healthy reservation status should reflect accordingly d Decomission that tasktracker +Just until the policeman says final patch +Could you check out as well? Its simple just removes uses of isFile isDirectory and exists from HDFS which needs to be checked in before this patch +Looking at this further there are some strange names for some of these headings that dont appear in the original doc +Really why are we having to vote for this? It should be a standard part of your release cycle + +Wont fix in JBT JBDS . +bq +Attached fix and test patches +Anyway this is a minor technicalityIts something that we will have to come to consensus on +I would prefer to keep the existing version becauseit is simpler and therefore easier to understand and maintainIf you can demonstrate how this would make a difference in a practicalsituation please this ticket but for now I wouldnt want to make thechange +So how would you know that a presumably working war file doesnt work out without a logger telling you whats wrong +No other attempt will be made to execute the failed action +Hi Christian please replace spec bundle with spec bundle during the by cschneider Upgrade to CXF fix for problems when service could not be generatedM cxfdosgitrunkdistributionM cxfdosgitrunkdistributionsrcmainresourcesM cxfdosgitrunkdistributionM cxfdosgitrunkdistributionM cxfdosgitrunkdswM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgidswhandlersM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgidswhandlersM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgidswhandlersM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgidswhandlersM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgidswhandlersAnd more +This is a bug tracker +What needs to be done here? Any additional pointers +It also proves that works. +I am not planning to run hudson validation for this +Thanks Ravi for testing the patch +Perhaps some committer can push this patch and update the target releasesThanks +However I couldnt find any split implementation while a file is being compressed into bzip file format +Two basic solutions come to mind +ASFAIK that function works with both MSSQL and +We should follow the Hadoop practice of requiring all new features or common bugfixes to go through trunk first +Introduced by our POM to pull that? Our tests will pass again wont they if this is in place +Run it using ant you need to specify a property see for four files the build script a bean with which to test a main class with example error output and a junit test case +I plan on trying to get these fixed today +to remove entire block if anything goes wrong and for patchThanks Simon. +This one is tested for sure +Doug answered my concern +Seems dangerous to pass a default CFM that claims +updated patch in the case when we are blasting through useless positions we dont care about the return value from readVint just want to skip over themthis seems to help a lot none of the shifting etcsetuprunrunrunrunrunrun +Id argue this is an implementation detail that goes outside specable behavior and I think most of the Rubinius guys agreed with me when I said we shouldnt be testing this +looks like its related to bug had the same problem described by this issue using Castor +Thanks. +Patch applied in rev to the branch in revision . +Patch applied at +Thanks! +So youre workaround that is targeted to just your join condition may be more appropriateTeiid would need to add a true char type to support comparison and other methods that ignores padding +Merged fix to the branch and committed revision . +Committed revision +If a classloader has been given on the then I think it should be used everywhere so this does look like a bug with Should be a fairly easy fix but the unit test will be harder guess well want a wrapper around the normal classloader that logs calls which we then check later +On the other hand the fact that almost no one knows it an additional alias listAll sounds ok +Are you ensuring that only a single thread uses each Thrift connection at a time +Hi brett you closed this issue but i dont understand whether the behavior is changed completely or configurable in the next plugin release +Yes I fixed that NPE in the compiler just two days ago +attached file regarding the error thats being seen when the large ddl file is being published +Another note if you dont have the sys admin handlers in solrconfig the old admin pages work fine but the new page will simply show a spinner +Hi Suresh could you please take a look on this for review +has been released by buildhttpteamcitybuildIdtabbuildbuild is done now +In addition to I think we had a user report of a hard to diagnose issue after this change +Spelling error in the method name +Thanks mac! +attach a patch to avoid garbage collecting when zookeeper handle is null +Your point +However each configuration has its own Appender with a and +Minor update messed up diff previously +Sounds good to me Clint +I started a refactor and then the plugin got moved to maven plugins and is called +Geronimo Realm is now available to Tomcat applications through a conglomeration of the and security configuration parameters in the file +A runtime is not the complete answer +The barrier to this is the use of resolved artifacts instead of the dependencies list so if they were passed in separately and the Ant tasks were used to resolve them there shouldnt really be a need for a project at all +We could even put the check if null around the method calls themselves in case the Java compiler isnt inlining it for some reasonIf after that they still introduce a substantial amount of overhead theres not much more we can do and Id be ok with removing operator hooks +This issue made the Wisconsin test stop using to populate the test tables so and are not in use anymore +After some help of Jay I have done some refactoring and explicitly set scope when outjecting then I cannot reproduce this error anymore +First upgrade in the branch and upload to +Attached a patch that will fix the lock contention and improves the performance by flushing multiple entries to all interested listeners +closing stale resolved issues +Getting a patch in would be a big usability improvement for Hive +Sure please do looks good to me +Resolving. +I will experiment with making tomorrow +You can revert to old behavior using false in you Mr +Done +Yes I mentioned that on the mailinglist +I made some fixes related to fault handling hope them fix your issue too +Closing as invalid +It still logs number of splits but instead of hanging at it completes the map then fails with a Shuffle Error Exceeded MAXFAILEDUNIQUEFETCHES +I propose to limit the capacity of the not by the number of but by the predicted size of +We want to have the same functionality for as well +I first set sanetrue in my did ant sane ant all ant buildjars then set sanefalse in and did the snapshot stepI hope that is correct +Bulk close for +This is probably related to +Sorry for the delay +Please review looks ok but could you please do one without reformatting +. +I modified tag and updated the repository for that onat toolsetcbuildmagic in Branch +I checked the data set linked above by only running the evaluators on it +Committed revision . +Move the fix to +When I released I had to manually copy information from the parent pom into the JDO API pom +looks fine thank you George +This works now as of revision +The cpp broker was using an area beneath buildscratchdata rather than the buildwork +Hi NaokiNice idea +the characters should not be in the markup file because they are encoded differently on different operating systems +Project has been retired so wont fix +Igot says to ignore this one +Committed with unused import removed and CHANGES moved to +Checking in +We can define upfront that a value indicates disablingmissingunset and that any other ve value indicates setenabling +Aint going to happen for this release +Closed with. +Proposed Solution +Thanks Nicholas for the mods the test is so much clearer now +arch extensions work and how our EWS tomcat packages should be designed to handle them +Thanks +Removing TODO where is thrown +Maybe I get Friday or Saturday some time to look at your issueIt would be nice if you can provide the example as a maven project please +oh +Ive tried created an archetype that has a variable archetype name so I suppose this ought to be fixed on the maven archetype part +If someone did want stricter handling of nulls they could do it through handlersI assume that the above behavior is only for the root variable so givenif a is not in the context then nothing is rendered +I did this modification also because rebuilding tags is a disaster without is taking care of this plugin or it moved to maven apache +Everything looks fine +This might not actually reduce the amount of computation much but it might greatly decrease the amount of allocated objects and object allocations are considerably more expensive than arithmetic +So the file is right + +Affected to we wont release a has been releasedMoving all related issues to the next version +Whirr cannot be used reproducibly unless the launched bits are known from prior testing +a particularly old version +Patch +Include any binary as a resource and it is filtered thereby rendering it useless +Whoops the previous diff missed a file +batch transition resolvedfixed bugs to closedfixed +Ive checked the sample app +Possible solution to the problem would be to split them off as etc +It hardly justifies the release of a jar file for each framework +Uploading here and will check back on RB +Shouldnt they stay in the module +We need to make this better +Since I unfortunately can not reproduce this issue currently help would be appreciated +bulk close of all resolved issues. +Please review +Fixed in r +Verified in Developer StudioVersion Build id Build date +Looks good +Removed default version from +Is there some sample to reproduce the error + +Im uploading a new version of the program I used to convert the files to escaped ASCII files +You could also include your reference to the Wicket solution or maybe even attach a patch to that ticket so it can be added to Spring DzmitryFirst your research on this issue is greatly appreciated as you have narrowed down where the actual problem is +I used the following test setup create a tree structure with a depth of each node having child nodes +What happens when the ownerless stuff breaks? Who maintains it? Collective ownership? I think that stuff didnt work so well in LucenelandBut its for sandbox so go for itBertil want to create a patch +test and pass overall +So please test again with a fresh build apologies if it still doesnt work! And if so please also ask the user to send some db dump of the names of his db table namesCheers rechecked against the latest version of trunk and the bug is still thereI have attached an sql dump of the database +Steven I am sorry but I didnt really get your last reply +It would be nice if the crypto data format had a table which listed the options the data format supports +This patch looks ok to me too +Seems like a different project has been added +Add also bundle mvn think it would warrant a specific feature for war supportI will close this JIRA and create two new ones to split the update from war support. +updated to add requested comment +How the heck do we fix this from a compatibility standpoint? Just make the change and stand back or whatProbably yes +Patch will follow later today +Can you put the patch on to make it easer to review? Please post a link to the review here +But its not complete +to trunk +The fix is now fully tested +Looks like this is fixed. +If it just partially brings the webapp back through the deployment process that will be a harder problem +still need to get the wcf configuration for WSASPHP BS DOTNET OPS +Cant we have it in some other base class or utility class that already exists? To be honest I looked into it and couldnt find a better place so if anyone else has a suggestion please comment +On the last thing to do I meant Try to secure the root node Access just created +The fix will be avaliable in the Xalan Java release +For usage or development questions please use the appropriate mailing listsThanks +These names seems to me not that uglybq +I just committed this +Not necessary I merge squashed and committed +done +So is it commited to SVN and is ready to be included in new releases +Cleaned up the javadocs and added a unit test based on the original testLet me know if there is anything I missed +Uploaded new patch with no compilation errors +This way you can remove the Windows specific comment in code and everything implicitly works +Its possible to have two copies of the same integer in a set with this patch +The statesDetermines if the class or interface represented by this Class object is either the same as or is a superclass or superinterface of the class or interface represented by the specified Class parameter +also there does not seem to be a best place to report issues about jira + +Looks good overall +Has no performance impact vs not limiting it +Patch attachedThe start offset is printed incorrectly before the inner loop +Patch for applied +Looking to the correct one resolves the problem as invalid because xalan doesnt behave conformant to the spec inthe case of a which is not declared meaning it throws no error +Fixed in Rev +I hadnt intended to but it can be done at the same time +Funny enough HDFS always worked fine +Aaarrggh how stupid of mine to have a again! I am sorry Ill check check double check in the futureI am not in the office today so will create a patch with your suggestions tomorrow and edit the wiki +I also agree with Brunos comments +attached using a reverse engineering approach for archetype generation so this gets fixed automatically. +Refactored code to reduce impact of synchronized code + +in method defcreatefromThe file is in librubyrexml can definitely be fixed in it takes minute to fix it + +Verified with +for the ve committed to trunk branches for and +This nd one is different due to which is an easy fix +Tests ok +some extra shotoku features +Filed for the poor naming of isAlive +Will release new plugin version shortly with this fix as +believe this was fixed in a previous release with the restructuring of the roles matrix handling +It implements owens second proposal ie +It would be good if Pat could run it to see if the problem still manifests +Everything was working just fineIve tested with debian and windows acting as server +It doesnt seem to apply for me +i built etchtrunk etchreleases etchbranches +Download the Nighlty Build after a sucessfull with this new feature +I will +Added an isolated change to dynamically add partitions for messages without key +This behavior would have been introduced with the J argument processing then +updating looks good except I agree with you about the rd party jarsLets be consistent on that I think it should be in the with LICENSENOTICE in lib It would then get pulled into the classpath because its in lib +The aar contains also the sources of the service +Probably git is out of sync with svn again +Glad to hear it is already fixed in trunk +Ive committed to trunk with revision +Forget the previous comment it fails with too this is because I didnt do a clean before +Shall we just do that now instead of proliferating YARN with more specific conceptsIm fine with the plan +Actually using becomejava! I am seeing the issue again. +In this case i cannot use the field annotations and have to use the method annotations because I cannot know if someone using my code with a bytecode enhancerHowever i think the current behavior is a bit surprising +Specifics should be filed as a separate Jira. +Patch attached +Im very sorry I planned to work on this last week but a ooziehftp blocker came up +i doubt its a real fix but if anyone needs a workaround this might help +Does this issue still need to be openI think with the javasamples project we are covered +Hi Michaelany news about the new version +Its not so easy to provide steps to reproduce the issue +However I was also able to recreate the issue +Allows moduleFile to be treated as a file if it exists rather than assuming patch added integration test to verify the. +Reopening to set release note now release note status has been set +It was something that happened while loading a file with but unfortunately Im no longer able to repeat itAnyway you can see where the cast is happening in the code and its an invalid cast. +And this is in fact the format added to to allow dev builds that have letters in the version name +It would be nice if a documentation patch accompanied this +Closing all issues +I would prefer to keep stuff out of though is the followup patch that brings down to the response factory +the unlabeled on applies to and master +Javadoc comments are not from this patch! let me check +Right after it starts close welcome screen and focus on Servers I think it is timing issue too +Hadrian was this fixed with that other JCR patch? Can you follow up on this and ensure we got all these JCR patches in +We need to see a diff that shows exactly what you change +contains some comments about threading problems with reload operations in general +Trunk was already fixed. +The Seam NPE was caused by a bogus entry in +I only made changes to the followinghibernatetasksNot sure what to do about thesehibernate The Hibernate artifacts are not in +passesRunning test suite +This patch is also forward port from the metrics related infrastructure change +This also happens when we edit an existing role and update the display name to be blank +bq + patch is not applying cleanly +I suspect we need to supply Path match Without no metadata indexed +Have to look into how to convert nanotime to something like a unix timestamp +Starting to look into this one +In looking at the changes it looks like this can be changed to use generic dialect methods and not include any reference to +Oops +Verified by Aleksei. +WITH defaulttimetolive +I created this JIRA to update the readme text is broken +need to investigate these problem reports +I have improved the test +Rich could you please give this patch a reviewLet me know if you have any thoughts about the schema changes +Deferring +So be patient +Resolved by Steve for pressgang although some parts have been split up into separate issues +My taglib is not set by library class have attached three screens +Patch committed in r. +Thanks +That is why I propose to give that control back to the caller to make the precision lose explicitly +Thanks +Ran GB terasort with input t test splitting +I changed this to enhancement because the current functionality is not broken +It comes from the fact that if the class resource to be searched is already in the cycleCheck set the method should return null and not throw an exceptionThe attached patched is a modified version of yours which fix the problem +NicolasAs Nileema indicated she was no longer working on this issuePlease assign this to the next person so that progress can be made +bulk defer to +lists moved +Just one more comment as for Swapnils currently partition id is also used for +So we broke them apart so that we can merge them all again +However it would just kill the immediate child +This happened again +This would have to be hints or repair + +i havent considered this casethe base class solution is the way to solve the problem but the validation should do a helper class +Committed to trunk. +Thank you for your contribution! +Fixed this for +I modified the in DRLVM smoke test so that it can easily make this problem appear +Thanks SadeepThanksSenaka. + to add to updated +If not exit the test with a meaningful error message +It is fixed by this as resolved as it is resolved for and were already at +Also thanks for the speed of the replyIt still worries me that the is releasing the connection in the first place +These lines look like errors in. +Thanks a lot anyway +Had an issue on some slow AIX HP boxes. +Attached updated upload a small deployment which reproduces this a sample application with code har filedb schema and data source please try to deploy and +it does not exist in +Forgot to resolve though +This JIRA is marked as bugThe reason behind not choosing approach is to facilitate backporting to and I think and should carry this fix +Just about done aside from integrating those once they get done +In my opinion the first two goals are more important than the third and we could provide an ant script to run these jars as we do for the samples in the Tuscany SCA distribution +Even if you move the Desc class to which is loaded you dont get the command mapping and end up with an error undefined method +Copy it to the directory where the other release artifacts live +I have merged the changes to master! I removed the use of ivy since using it for only one of our dependencies felt rather peculiar +I had missed that comment +Fixed for greenpepper and +batch transition resolvedfixed bugs to closedfixed +Hi Chris AllI just had a look at it and yes indeed there seems to be a problem there +on patch. +Thanks +bq + +This patch caused ant javacc to also run in contribqueryparser and automates the steps in in the ant script +Thanks Christian +depends on something marked to e resolved in also my recent comment on is no longer a supported platform +I will work on improving my implementation and will attach a patch to this issue +Plz provide us a testcase to verify this issue +they work fine on my dev box +changed all occurences in the code where a null value might have been I suppose Oliver changed this originally because there was a failing due to thisI committed a fix if the TCK compatibility breaks due to this well need a fix to theregardsMartin +Perhaps this is something we should add to the list of things to do for the release + fixes test failures and uses the new methods +small quickstart that simulates the problem +profiler screenshots +It should be an option and one that is off by default +Bulk close for +commit fdabbebfeceddb +KentAre you just requesting the WSDL for the web service via something like wsdl? I am not too sure what you are trying to do here +Does that mean you have found the problem and are working on a fix or at least a speculative fix? That is excitingRichard +Code readability would suffer and from that it would become a source of new errors +More NPE checks have been added in Git ID babeadaedfbe +Can you fix it +I know ant doesnt and Im saying ant doesnt work here either +looks good to me +nice +Im not sure if just the uncompressed file can be patched and then a script is run to compress it or how that works +As I detailed in the remainder of my comment the problem is reproducible with only code and appears to not be dependent on the aggregator component +Try to include more information +You can consider this closed. +Im closing this JIRA because installation speed is back to normal +This difference seems to be caused by not sending the truncate message to its peers in that flow the test case is following +Now I see a boost in the performance for both versions of server but I still see performance difference of messages per sec vs messagessec between CR and beta +issue fixed in the current SVN code base +Pipelining has been committed to the Java code base as of r +Looks simple enough Please go ahead and merge +for ESB issue is a duplicate of and has already been fixed +It is not solved in JSF spec +Ive applied your patch I made some modifications to the code that finds the parent pomThanks for your contribution! +Agree wrt the fix for Windows +Its value determines the maximum number of member properties until additional intermediate nodes are inserted +Benson can you please add the EOF exception mapper to the branch? It is not there at the moment +Unsurprisingly failed +Once the initial version of the statistics publisher is ready we could create a sample cartridge image in EC +Nice work +This patch addresses the rmic issue by adding the from the SDK as a dependency to the antrun pluginUnfortunately this breaks the build process on Mac +reopening for improper upstream issue is marked as wontfixed +There is also a which should really be called +Hi CarlosIve fixed the groupId +If the logging stream is not there the would write to the raw stream and nothing would get loggedNote it could be possible that the needs to be moved earlier in the chain +It would imply to use another kind of rewriting based on regexp or such +Probably we should patch so that it propagates the exception back to the caller +Thanks Brock +if not close that one and open this one +File that we tried to open +It will however illustrate the problemRunning this takes s per test on my machineRunning it will a small set of changes to the code and it takes only +I was just surprised I saw no mention of back compat in the discussion other than Mike mentioning that the change should be made through deprecation early on in the attached email thread +Youll need to dig into the logs to diagnose the problemI noticed this in the end of the last +Done +It now makes an alert when the ruleflow diagram can not be created +This way we maintain the API and users dont need to care about the form component type set or notThis way the set javadoc needs to have a good explanation saying that this property will be respected only if the user do not override the converInput method +Please validate it before I release +Thank you so much Roland +Im no longer using so if anyone else cares theyll have to update the patchSorryChris +Files to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This issue is fixed and released as part of release +It does not work but fails with an NPE visible to the +I am not sure if this is the right place to ask this questionCan you please give me some pointer regarding how do people usually solve problems in this communityGenerally when this kind of problem comes up I ask for steps to this problem +It actually fails because of theInteresting thing is that when I print the attached file to another PDF file and then pass that file instead your code works fineI wonder if it could be the document Format the one you attached has its version as and the one that I created by printing your file into a PDF has it as +If something is modifying the map you dont know how the implementation handles the insert of a new value +This appears to be resolved. +FranksGood news +I still think this doesnt belong in Camel but in Karaf +Patch committed to SVN trunk +move to jboss tools since no relaiton to JBDS specfically as I can see +Fixed +I think his fix was slightly different but achieved the same effect +The question is if pooling is so much faster than creating a new instance especially compared to time spent on parsing the documentEven with pooling you might end up with memory problems for example if you have some instances in the pool still referencing memory and these instances are never used again +I think having the tool drop and recreate a table in preparation for a test is a good think +Restore and and s use of it for cross thread locking in place ofThis issue does not show on hudson build nor on most of the machines hereabouts I have to run it on a limp old linux box to generate the +So the fix should cover both tools +Closing issues +plugins +I just committed this +FYI heres a list of the modified and added files included in the patchA +This was a config issue +I dont know what happened but I agree this change should be reverted +Alternately we might just generate a single interface and not use different method names rather distinguishing by method signature +Yes it does +htaccess for redirectingAssuming codehaus site is under apachePatch intended for +But all other icons look nicer than it had been before +Saves you time to code a trivial patch and Ill learn too +The CMD scripts attached in the patch have been tested reasonably well through scenarios +We think it could take a few days to implement such a wizardWe can report a feature request to bugzilla +Commit procedure to Swift data locality API is in testing for Testing Swift REST client against as many Swift public clouds is in progress +If a different page is activatedall tapestrys ajax support can do is inform the client to request the new pageSo indeed your parameters will not be taken into account ill check if this is you suspected youd have to use session or flash or cookie persistance to pass the values to the new page +Failed Tests are +Im pretty sure you guys have more important tasks to work on so I guess we can close this +Thanks Jacques +Just need the ICLA statement or CCLA paperworkJon +Bob Honestly I didnt spend any time looking at the test with but nothing comes to mind that would have changed +It is an easy fix +Oh right Zebra +We can use the defined in the examples as they were and we can do runtime partitioning of graph in the +Reloading on every exchange is good for development mode +Take uploading right one +Patch applied to trunk and TILESX branch thanks! +Patches should be submitted so that only the actual code changes show up +Will pull the latest master and going to check it again +It passed times +out of scope use the plugin to manage licensenoticedependencies files. +files or directories starting with are ignored now +Here is a patch for review +thanks +coho r android and found was missing a header +Marcos can we have some more details on this +Try +I do want to say that once it is clear what the functionality is being provided it would be great to then write up the implementation details on how we are going to achieve that functionality +a patch to honor to use the javac in it +Yup Ajax hasnt been serving the live sites since late December +I think its more a temporary glitch +He said the comments wont pass strict HTML validation +Might it then make sense to wait until the search component thing is done? everything would be easy +Note that in VPF the feature ID tends to be the first attribute called id +Unfortunately this is one of those tests where I cant reproduce the failure but Ive seen it happen a few times on Hudson +Thanks for thisI just applied this patch to revision orgjruby had an incompatible change for the patch think this is still useful +I think that approach makes the most sense +Could do the same thing for tables in trash a running master keeps track of when each trash table should be deleted +Would it be possible for the applicationName to default to the? That would remove the extra configuration requirement for +Sorry for the confusion that I might have createdYes you are correct this is probably a problem as it appears from the error logs + +Thanks again +I am confused this morning +Manually tested and it passed +This would improve a number of things but breaks the backreference checks for example when deleting versions +The problem was a buggy implementation of the with a predictor value I fixed that in revision . +Added for and +Just deploy it invoke swe in a webbrowser and youll get the NPEIll upload the sources for the EAR in the next step +A test would be nice eventually +What I meant was that the and should have more elements and or attributes in them so that they support more interesting variables and assign copy statements +Both options are possible +Reviewed by Barry L +This isnt the case for all majorminor dependency upgrades. +Demonstration of the attack +Is there a possibility for me to activate some logging or using a debug version of derby to shed some light on the problem? The error also occurs with one database only +Look under the subsersion commits tab this will give you the files and let you see the diffs +Here is a patch +But this work should be done for +binapachectl gracefulSLAPDusrlocalopenldaplibexecslapd AlsoNeed to add monitoring for controlhaus to nagios +Tomas seems like the config that you provided works fine as storage you create a separate issue to follow your issue? Dont wanna mix with node kill if you can replicate the issue again and provide JDG version information TRACE logs +Fix is in the latest nightly build +Patch to check number of reduce tasks and set it to in case it is set to +I dont think we need to worry about shading dependencies and such in this case +I believe they should not be as it is a vendor specific interface and encouraging applications to use them will result in applications +yup tim is doing it up right now +Thanks for the info +Ron could you please attach a sample XML document that in your view should successfully unmarshall against the classes generated from your XML schema? I need to build a test case to avoid future regression issues in this code area +Proposed patch for this issue +The files are reserved for replication for X amount of time +Oh and thanks a bunch Musachy for the doing the bulk of the work +Seems to be fixed +I am actually planning to add an implicit single embedded zookeeper node when nothing is specified bit also adding a role that you can use in the template definition to explicitly specify which node is to spin up a node as well +Should go in at the same changes only in this patch so no tests cases addedmanual test was performed to verify new features are working on standard tests as well as classes with Before After and annotationsverified +Thanks Jitendra! +Where can I get the snapshot +Added more unit tests +I still it may have something to do with finalizers preventing GC when it needs it +In my case the phase takes several seconds so I dont want to repeat this just to the tests +Thanks Elliot +Maybe you are referring to service definitions +Fixed in revision +Before this patch the bug was not present +But this would require a lot more smarts in the system and of the user +Agreed on the attributes +I have another patch that Ive been working on that gets rid of those because they dont work in Visual Studio which will now conflict with this +I guessed we need to modify precommit hook to check Test to ensure there is a timeout provided +Could you please try to turn your test case into a specs for specjavaintegration? You should be able to just run rake specji to run very much for the patch Committed in r. +Our qualifiers tend to be long so them helps quite a bit +Can you please attach it here and click the box that gives license rights to the ASF? Thanks +rAfter applying and the patch applied cleanly and all tests are passing +Thanks Janapriaya + for patch +Build seems to work now can you please check and if necessary upload a patch for? I am OK to commit for you +Thanks for testing out Lucia! I will commit the changes then +Howard if the resource adapter doesnt work when you change it then its probably because the jars are in the wrong place +It would be extremely helpful if you could create a diff file against the latest SVN and attach it to this issue as well +we should probably not generate the method stub if the name is not valid in Java +Modified to remove Jetty just committed this +Some of them are set up as recurring calendar items so the message can repeat at intervalsOne suggestion on your work so far maybe put some of those screens on other tabs +Im marking this open to community. +With latest buildWe have the ability to configure External DNS IPV addresses at the Zone levelIPV Router and guest Vms are programmed with IPV address for external DNS entry for IPV network. +to proudce just this patched class and update the project page says version depends on +ACC solves the latter but not the former +Unscheduling from +Code pages are now automatically obtained from the JVM +the px file is the smallest logo that should be used +configdisable if this file is present xgl mode of the linux X server is disabled for that user +Please review it +If anybody could give me an indication of where to look in the source code Ill probably be able to fix this +Would be great to include in esp if someone is willing to create the patch? +Then you need to build the coreWeb and deploy it +Steven why would you want to add elements from the tools namespace in an beans XML file? The tools namespace only defines meta elements to be used within an XSD +Marking closed. +I changed it to only cache when there are no methods +That makes more sense then I thought maybe you were sitting around on the weekend trying to find ways to make more work for meWhat are the pyactivemq unit tests doing when they crash? I assume its not creating connections in an infinite loop anyway Let us know about the static lib +the patch is in zip format +LanceYes I fixed all files that were not in Tomahawk +These are the latest versions of the test programs +all issues to the CR +I still think that a solution where special characters are escapable is somewhat nicer but the character sequence is probably not used very often +The script attached actually runs fine +There is no logic in that follows your recommendation +I get this stacktrace error when deploying EJB MDB +Ok for both svnpubsub Ill update our release documentation +I have committed this +Perhaps we should change the logic to undo allocation in this case +Ok thanks Jacques David I will try with percentages and then I will upload a new patchIm sorry for the problems that can be caused by this commitI will let you know as I will be ready for a new patchThanks againt worry Marco apart David and me I think that nobody else has suffer from this patch o +Looks good overall +This section provides a brief description of Derby followed by information on the skills and software required to perform the activities presented in this document and ends with a brief description of what is presented by each of the Working With Derby activities +It sounds reasonable and does address the concerns that I raised earlierI think its worth it for you to highlight the potential drawbacks +I added aclrightsdefault Allread To wwwconfig on Eos and it now appears to beSo that should be done +So the coversation scope doesnt start to live after the restore view phase +Could be completely unrelated some seem unrelated and I havent run the testsuite in a long time so have no baseline +I used the recursionCounter that we keep in the request to detect a forward +The code is not more difficult to understand IMO +Improve this for believe stack trace sanitization has helped here and this is not really fixable +As the unit tests are integrated with the build if any part of this fails I will get an email with the details +Is it case? If yes then it is actually data unavailability but not data loss +Changing releases on the fly may be possible but I dont think it is a good idea +all I know is that I have SNAPSHOT dependencies sitting in a release repository +When in strict mode and a reference method was called on a null value like then would throw a +Reordering the occurrence of some methods in file does not always cause a new order in the class +Where in the docs did you see XERCESROOT in place of XERCESCROOT? This we should certainly fix but I dont know if it would be practical to have a FAQ entry giving all the incorrect spellings of XERCESCROOT +I believe you mean queue not togglePanel given the link and the queue has an apply button and the togglePanel does not This is not just a issue +hex is dec hex is dec so is hex lets just handle them. +tmp +These tags would have to be changed to catch a and throw a. +Note the tests all pass when run against a RI JRE but not when run against the submitted classesAlso the implementation of the tests doesnt appropriate use the assertion methods +That is ps will show both the build process and the scp process for days +this is not a blocker issue IMO this look like HQLDB specific or your env +I forgot to point to the dependent project as part of project references +Thanks Thomas +Maybe we can close this issue just by just committing these files +Its the first issue I found right after start with CXF +The problem with seems to have been fixed in so now all exceptions with that message id are passed one argument and linked to the underlyingThe fix committed here seems to have broken this in +Patch for javalangStringI guess your patch may be void by the issue please look if this issue is applicable to the classlibs version of +Tested with a classic Java project by replacing sourcessrcmainjava by srcmainjava +Zhihong Jan if the versions looks good to you I will commit +Is the change in sufficient or do you want to keep this open until has a more modular setup +I believe your interpretation is correct +Not an actual problem. +Werner I just read today and I can imagine that your patch from Monday would solve also my will try this as soon as possible not today because of lack of time but hopefully this week +Should be done let me know if its not done the right way +This one includes whitespace changes +I dont see this bug on Axis or +skipping wiki pages whats new section as they need to be summarized and inserted into the corresponding guide section +Im just getting started on the client side features so its useful to have some concrete targets +I will have to verify that this has changed +yes that is why I submitted a fix to thrift and linked the thrift ticket two comments above patch for was committed in Thrift SVN r +Please provide a quickstart and we will reopen the ticket +Hence the error when the tries load it +But such an implementation would lie wouldnt itThe number of bytes you can read after running through decompression is most probably different from the original countGiven that the bzip format works on blocks it is quite possible that trying to read a single byte may block on the compressed stream because it needs more than a single byte from the original stream to complete a block of input +tested the code completion in several situations and got non of the exception +Unit tests pass +Thanks Tom it sounds as if this should be documented somewhere in the platform release notes +lgtm now +Ill reopen this until Im sure about the results +If the sort by could scan both lat long values for the one closest this would be solvedGrant can you look at this +i added the logj dependency in svn thanks for pointing that outregarding your problems in jetty i deployed the war succesfully in tomcat and jettywhat problems did you have in standalone jetty +The patch has been applied to master branch. +Proposed changes +The right way to support queries is to increase RF +Attached patch for workDisplayeda running total as records are selected on the screen together with the number of records selectedb an opening balance +Changing the configuration works on top of changing the hostname as it doesnt prevent you doing that but you can try other things like bringing up a new JT on a different port on the same box new machine etc +Do you have any use case why it needs to be back ported to already released branches +Thanks for the patch + +I will put something into place and then see if the community wants to tweak them in any way +MahadevPlease let me know if any changes or improvements needs to be made in the patch +no idea if this is still an to testing +In today we have such eta for both detection and detection + is changed to an interface rather than an abstract class for better decouplinginterfaces are generally discouraged in LuceneSolr because it can make the API hard to modify +This issue will be closed and two new issues will be opened for these remaining issues. +Hello Tommy just wanted to quickly check with you if you are still looking at this feature requestIf there are any issues you are unsure of please feel free to raise your questions in the developer forum and we will try and point you in the right direction +With this patch applied I think this issue is completePlease close if this works as you expect +AmilaWell need to unwrap the return elements and we need to handle the client stubs as well +Dependencies are added in +adds length information to as rev. +Switched to the resource resolver factory api in revision issue after bundle release. +Patch attached +Check out the antlr wiki for more details +Basically fix all the issues raised in the comments +Bumping +Hi can you give it a try removing dot from path please +Just launch the ant task +Please change to if you intend to submit a fix for. +Ive added a junit test fixed the bug and review code to avoid such problems +Since the is changed to have File type to get the renderer mapping for File in created by builder plugin xhtml parsing finds no renderer for while inspecting +If you still want the old behaviour you have to copy what does +Indexes are for ordinary values not blob or clob +Maybe as SCA widget or as Http binding service +Thanks ThomasWill resolve once get this committed to +Transcription of prod type metadata info to in r. +Create a Hello World feature that just prints something to the console +Thats OK then +What you mean not everyone is on the West Coast +Otherwise they will be lost +Regression Test Update The patch has passed the regression test +Thanks! +Closing resolved issues for already released versions +I am having some issue to import it though +all issues to the CR +It is much more clear and simple to me to check for null in this case +It should not really matter but I included it just in case +The r change just broke the trunk build Michael I assume you will fix it +But since the test scenario is so similar to the one in its probably fine just to remove this test +Apart from the main issue this patch ensures max min in case of incorrect configuration +Darran any update on this? We need WS CP drop to get it integrated into EAP CP +This looks promising indeedThanks Bruno for your continued effort in this area +Its not related to this problem +It seems that can manually redo the commits from a git repo +While this is great I still think its better if projects not requiring expression capabilities can simply not include the JAR in their the consistent failure issue would it be possible to detect expression usage and then immediately throw an exception if the JAR is missing? I dont think you need to fallback to standard evaluation just detect its missing and abort application context startup +Commit the addon patches to At revision trunk At revision + +Do you expect more work on this one Rick +Looks good to me +I fixed the on both dont know what you mean with the red circle since nothing should show up here until you actually added something to the problem of column not showing in the title I could not reproduce I assume it is an font size issue +Patch to try to cover all scenarios from this and +Looking at the code though I dont know why its only catching and not any error +See same bug recently fixed by pull I believe. +thats basically what i wanted to do +Crashes due to JVM bugs are not our fault +The web app is upgraded to Servlet also introduced JSF and facelets into the app and started on our basic components in facelets +It sounds like that +Could the fact that I havent changed my since be affecting this? If yes theres a bug somewhere +It looks for an increase in the number of files that dont have property license headers +Bulk closing stale resolved issues +One thing I just noticed is that the error message has changed from to. +Applied and snapshot deployed +I think we can make the prepost more integrated and generalized by extending sharpen +Like the new comments and test modifications +priority for this issue is not high enough +native core tests execution time drops from min to resolved issues for already released versions +Documented as a fixed issue in the WFK release notes Calendar and popups had fixed positions within scrollable containers and would not appear if a user had scrolled too far +Hi Rafi is there any more work required +This is a duplicate of +Please file a ticket if you think we need one +META +Hi PulasthiYes this project is still available for Thanks +You can use the without breaking it apart here as well if you need toI think its a slippery slope we start having to support both the segment ids plus the ids +Patch available +Screenshot is are addedPlease review +Working closing thanks a bunch. +Oops +Will discuss on the mailing list +I just committed this to trunk and branch +We not be able to claim it as Compliant but otherwise should be able to freely discuss +I will try to upload a project demonstrating the issue later this week +for the patch +Thanks Jorge Rodrguez Pedrianes +Ok I will add there +I cant believe Matt was really serious about that long name +Fine by me however I will not be able to testverify this until the next release is officially out +i like core +Hope this helps + +were stuck without the correct setup just now +verified by Richard +Thanks. +Example which provided in a first comment for leads to exception which mentioned in this ticket +Im working on the left part and subsequent ones will be coming +Sorry to report this late but a few weeks back I was getting many Server Failed to respond issues while working with Hive Wiki +defer all X issues to +Is that purposeful? Also notice I used a more compact notation for passing variables to make +If you get it working report back and I can modify the dbre code +push to +Work for this issue has been completed and its in production. +he is using the wrong address +Im marking this as resolved. +Commit message was better footer rendering shows Apache addresses +However I can download the DT and it is the same I have uploaded before +So I dont think that results are different +Not a bug. +binstress and fixed bugs which was stopping read +Uploading which fixes the issue with the added test +Does it make sense to prevent the manipulation of if they have configured the managed schema option? My personal take is that barring having a hard time making this happen in the code that the two options are not mutually exclusive and we shouldnt worry about it +No its already fixedIt might be already fixed in Beta or Beta. +patch provided as Siddharth suggested +Items can be added and removed. +Please make the extensions as mxml the default ones +Rescheduled for +See Cache Containers +This was a great ticket to start diving into the read path thoughAttached revised patch +I do believe it needs to be applied to both indexed shapefile and shapefile since it uses the factory +Wanting to make some more testing I look in Axisc tests but there are not many tests they are not automated like with in Java often test says SUCCESS with env status set to FAILURE So Its seems we need a test framework which allows us to write more tests more easily +Here is a patch to fix issue of refund for orders placed through Express checkout +If LDAP server fails everything failsI am not arguing against centralized configurators in general +it takes some time to reach ibilbio is still wrong +bq +The build with this fix is not available +Josh i figured you would be the right guy to answer this +Maybe you can save the time you spend inthis contribution for the documentation of your own applications in short you need help on +Hi HenkCan you explain to me how clicking on beehive will go to that page? I dontsee how the directory command you pasted will do thatThanks +Introduced the notion of +bq + +Set Fix Version to unknown this patch wont be ready in time for trying to test the interaction between DROP COLUMN and GRANTREVOKE I encountered bug +This is more consistent with the workflow popup panel and it never would have worked in its current implementation without the job web service being changedI think we should change to add pagination for all popup panels which will encompass the web service changes too +More handoff required between threads which would impact latency but better throughput in high pressure cases because of fewer threads contenting for locksAny chance to measure context switches during the test withwithout the patch? I wonder if those increase due to thread handoffs until the ctx switches from contended locks outweigh those +I fixed test in trunk verify on Jenkins with all templates. +uploaded with all dependent files +Ive applied to trunk and identified some more places where this should be used instead +where is the patch for review for Beta +See for a problem the implementation of the resolution of the LWG issue causes. +Thanks! +This is usually the testtest due to unable to allocate thread error +Jukka youll need access to server so let me know when you are ready or if you cant do this +Net done. +The bug is fixed in so this will be verified and marked fixed in Spring release. +If you think test is alright we can commit it asa patchThanks +Thanks Stein I have just done this myself at about the same time +I cant decide about that +Instead of helping a user pack data via encoding we should encourage the use of compression +Shweta please use the latest build to test + may split +Patch with changes Update class Update test cases that test Update where filters are initialized in http serverNo new tests as very difficult to test these changes +Uploaded test class containing enum and the parsing error we got +With the two patches I get a failure in lang +Knut I was wondering if you know why this might be happeneing +Yes it can be set either way +Added to trunk after some minor of comment Im not sure what the purpose of the contextTypes Map is +Unit test testnativebasic failed with the patch looking into it now +IMHO it should only filter when the id and body is equal +Debasish Duttaroycan you please check with the nightly build as I know dims has fixed that +The source datanode of a replication request sends all the data to the destination datanode but does not require any response +Closing this one to mark at least the progress so far +Fixed in r +I just committed this +Should I make permute be an option for with a default of on? It would have to be an option since all commands use now and some like test cannot tolerate permute +TimLast I checked bugs dont magically fix themselves Someone needs to confirm they can reproduce the bug before but not after it +However theres a problem with safari and the resizable drag handle on that bar preventing the select from being clicked +doesnt apply cleanly to trunk anymore can you upmerge +Jerry Any work happening regarding this +Please provide a sample class that exhibits that compilation problem otherwise itll be difficult for us to find the cause +But selecting multiple dates issue with have still problemI am using struts version +Amareshwari I got the same error while running ant +sCFIPCFIF sFIPFIF +Comments on rb +Just to be sure this is what I will do Lucene release with and without complements With the default With a NOOP empty Sampler implementation that I will return by overriding the get method in the Sampler that I will provide +But the following projects do not have logos for them +Please note that the diff may not apply cleanly because the patches in has not been applied yet +The next thing to do is to add json calls +Still broken can we test this next was fixed just that the error messages made it look like it wasnt +I will continue updating thereEugene Sorry this one has sat around a bit but wed appreciate your help now reproducing it so we can fix it for. +Comments please +I verified that this works in the latestTorsten I dont have rights to comment on support cases at RH anymore since my subscription expired +When we tackle those changes it may be interesting to know that the top level devguide artifact is created by the pipeline task in the debug target in Touches the following fileM +Agreed +Or you know you could wait until the next nightly build of shared is out +We have plenty of demos working in Talend distro with CXF but that does not mean there has not been a regression which your bundle may be exposing + +I created an event on Node after move it the UI appear events the logs +I still dont get it +Fixed in in +Ill check if new Service proxies are responsible for thatThanks fix confirmedColm +I just noticed the update so and conn timeouts being set for tests were reversed and fixed +At that point it becomes a potential replacement for +Ill either attach a patch to this issue or commit my changes to a branch this evening +I cannot think of any +HADOOPHDFSHOME is for developer only to work on commonhdfsmapreduce code in separated source tree +I have two separate machines running into this issue +I dont see how the database could be corrupted +Created the doc bundle with com +If you find further issues please create a new jira and link it to this one. +Will open new issues as needed as described in the previous comment. +How should this be accessed? I wrote a little bit of code for the repository portlet that checks whether an artifact you are adding is a bundle and if not runs it through the wrap url handler +Sending gshellsrcmainresourcesspringTransmitting file dataCommitted revision +Tested against CVS and and could not reproduce +It is just missing documentation +Today the way an expression interacts with the Bindings is not with invokeVirtual its just another valuesource coming in to its array at runtime +I forgot to say that the error appears at the starting of the nodes while is trying to construct its beansI didnt even started the applicative test with its insertions updates reads and removes +This patch does the improvement by replace the with operations +Ive also reverted the buggy assertion so this should not be a problem anymore. +I will have a look at that soon +bq +This Request should use https +Good! I thought it seemed so that it should but theres always a small chance so thanks for verifying it does work. +if you change the dest dir you need to change the classpath to add s your use case aspects that you only want to use when testing? +It updates the root pom to pull in modules for and but those dont seem to be present in the patch +Were not big users of that field but AFAIK replication is the right one as its the name of the package +Since the first performance enhancement suggestion is clear and easy to implement maybe we can resolve it seperately +Triage for +So if theres a project the server cannot have two copies of it deployed +I wouldnt worry about breaking the current thrift api completely if its a big improvement contrib is always optional +the generated documentation on the site is wrong +The compiler phases are still relying on magic constants whereas the patch was to drive if off of enums instead of numbered constants +the only thing that needs to be fixed is the import +I Am Not Able To solve this problem Please Help me discuss this on the user mailinglist. +Resolves it for me +Stan You might want to send this over to Jay I spoke with him today and he may come across this in testing this week. +Can I get this ticket or an updateIt looks like theres a disagreement between what is or is not supposed to doEither way Id like the ability to specify redelivery policies from an configuration. +Lets hold off on that and let others offer their inputWe already discussed on the dev mailing list that user preferences would not be used on the eCommerce side +Linked that JIRA here as related. +yeah my fault the patch restores the former code for node MendoncaI try your patch but all map data change to nullThey really have data in client side but all null in server side +The test output is different from ibm jvm or sun jvm +Is it possible to send an empty modification to the server? What would the server do? Just update the modifyTimestamp like a touch on unix +Thanks Sergey! +Done use wars from tsurdilodesignermaster until is released +Would be good to update that patch on trunk and get it in +Hi Lane you may want to reattach a new patch and click the submit patch to kick off the automatic hadoop QA process +HI EliI have updated both places +Ditto for with +also break over time as was found after +Ok then Ill have a look at whats needed to make it work with EAP +It simply doesnt work +I prefer to have a valid json line by line rather than having to remove ending coma each time i want to use this line +For example a query where qfood probably should not return a suggestion of food +Most people poke around for values they care about and ignore the rest +Not till Im afraid +verified in JBT +done +This a very preliminary patch +Patch looks good +ok ill take a look the i misunderstood so this works for you in T but fails in T +By default the relative path is relative to your local file system directory +Emmanuel The behavior you demonstrate is exactly the behavior that has deemed unspecified +Can you print the error you are getting? Logs would be useful +It took us a long time to track the issue we had in our code it is not fixed within the jaxen code we will have to fix it externally +Is there any chance that you running with class verification? If so can you turn it off? Since we introduced a dependency on ASM to correct these errors with Java is there any chance you could buildenhance with Java +For the skew join if the partition table turns out to be small then we can convert it to a normal join which doesnt need the sampler either +upload form is in Contactform without upload is in Aboutclick Contact first upload a file k then click about to submit something +At least one milestone or release candidate will also be published such that the community can validate the metadata prior to going GANote that any future releases in the Spring Framework line will continue to contain metadata and will be published immediately to the EBR as per usualInterested users may want to place a watch on to be notified of further updates +It seems there we throw an exception to tell the client to back off and thus this no longer easily collectible at the server so its probably OK not to have this metric in +We suspect it goes in deployteiid build script is missing to include the missing JAR to the kit +Im not sure how a Git patch plays with SVN +I from trunk and made a new patch and there was no difference between the two +Aliaksey Im not sure that Slavas fix was commited to this build +new one please. +It might be a coincidence but maybe not. +Fixed thanks to pathc from Paul King +Almost there +Other than that no bug. +I will make a mail on the dev list to introduce it. +Glad you got it sorted out. +Rev has the changes for trunk +Committed patch at revision +the logic now attempts to authenticate the request without doing a HTTP if there is info avail in the request to perform the authentication +I just committed this. + +Ah that figures +Looks like Amos could be right +Sorry Bruno I had it committed alreadyOf course I should have changed the assignment first my faultThanks Boris for your patch test it and tell me if it works as expectedregards have downloaded +This is not an issue +Reattaching patch as I had forgotten to svn up before creating the patch +Ill take a lookHoward +Thank you for this enhancement to the tests +We now have a merge conflict +I dont see these changes +Are you saying the defaults are wrong +its not that trivial +Five months turnaround is better than never right? Right! +This test passes all Backward Compatibility Tests removes the dependency of and uses windows directory structure +OrgApacheOodtBalanceCore patch with new hookbeforeall plus documentationI agree that the existing hookbeforesend and hookaftersend hooks could also be used for both types of requests +Just some more data when I use an the word count works fine +It works! I gave it another try with and and this time everything went ok +So Im setting Fixin unchecking Patch Available and resolving this issue +Ivan are you going to address Salikhs comment? Is a fix needed +Yes these tests do not work against the current code base +Uploaded +Another twist is that we have to support the constant prefix as wellGenerally if you have any kind of length to the prefix linear wouldnt be so bad either though right? Guess it depends on your terms though +Marking as resolved. +Its just that the states sld we ship with defines labels as black explicitly +stopping backporting at +To avoid historical version explosion we apply the normal policy of only supporting upgrade over a single major release +It handled the case that a filter clause returned null different than the case if the clause returned an empty bitset +Ill work on the test if no one else is +Added +Fixed by someone else some time in the past +The does assume that the separator is +In the Dispatcher the first thing that it mounts is the file with a relative reference +This partially fixes Yorams problem +Thanks Leviev please can you complete and send in an ACQ so we can use your contributions to the project without any issuesYour patch was applied to the NIOCHAR module at repo revision rPlease check that it was applied as you expected +Alex I assigned some to you +Hi BilginThanks for taking care of this I havent had a chance to test it but the code looks straight forward enough +Im going to move this one to JBDS +patched +Dimitris can +provides a copy operation that is now being used. +Do you want to mock the log property on artifacts or do you want to log from your testsIn fact we havent set up the build system to use a logging framework so you simply end up with the log property provided by which is an instance of as Sven has worked out. +I regenerated the patch +However I think for most iBatis users who want to take advantage of ehcache they would prefer to have their cache configuration within ibatis +Looks like this one just needs a merge to +Thats for file in input folder +Not sure which log level is the best +For me its a little bit strange +Thanks for helping out the update to our docbook is now committed. +Id like to commit this in the next day or so unless there are any objections +Didnt realize that it worked with SQL Explorer +Committed! Thanks Liyin! +Sorry again Im slowing everything up feel free if you dont Ill do it when I commit the Highlighter fix in a bit +the explained closing stale resolved issues +jb esb file using smooks action for transformation as shown in attached image +Guide as well as the factory class +I need a method that proves that the native SQL is used +during application initialization mean observersMarks suggestion contained a very important part +Issue can be closed +Ive got support for these in a world of mine I could get the changes in +And please use the users list next time first +I am still on this patch as it is +Shame on me +No matter how we define the slingkey to be defined natural language or some cryptic code the query must of course escape any special charactersThanks for reporting +post the patch on reviewboard +If you want I can also add that patch to too +What ever they are called static final constants +This is now fixed we can handle many many producers being created closed +No it works well with server scoped library so its something only problem is the keep generated as are held in the and not releasedThe ognl seems to be working fine with the that is a server scoped library +The code is reversed back to what it was before the above changes were appliedChanges are checked in r on trunkNeed to change the examples in the wiki page as wellI also removed the NFE in the useSSL method highlighted by rob +Thanks for the comments Todd +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Are you sure that you diff the correct file since there two files? Let me post the diffSorry I must have compared the wrong filesbq +I thought we used example example example etc +I just committed this. +OTH if the generified Collection goes beyond one level it can get nasty pretty more context on class I believe all strategies are run with the builder set as delegate this means you have access to the current nodes context you can pull a lot of info from may also be another choice note that OGB will still cover this case +here are couple of design discussionsUI Change Action automatically swap launcher job ID to spawned job ID +Its the test that is flaky with its assumptions +Thanks +v +I think its now how it should be +But I prefer if we can wait till +we need this out of the +setting to but if I get close to releasing and there are no patches for this it will get bumped +Uploaded another new patch with minor the system tests with the patch and no issues were found +But this is definitely workable +Forgot to mention the failover state machine would be a little simpler if we always required that a fencer be configured however I left it as is for now +This should fix as well +Its going to be a fairly change and will take some time +The +MadsThis seems a little weird +Update the title now +stackCould you take a look about introducing safemode to delay SSH after master is initializedI think this solution is more easier for the issue +Thanks Yonik for the patch +This Jira issue is resolved and no further discussion will be entered into here if you feel the current behaviour is confusing feel free to submit a new Jira issue. +Should we need two s to commit itIt also introduces a new configuration file +Can you determine if theyre actually the same call? If so we can just tweak the Windows version of our POSIX library to do the same call + +putting back to trunk +Thanks +This is fixed as it is +My understanding is that they do Graham Davis here did the work +should be handled outsided the scope of an aspect +Thanks +Suggest to close it. +Arnaud can you please have a look at this bug at +We dont want anyone getting the wrong idea +I didnt even see eximportkeypair which does what my patch is supposed to do +The status bar seems much more work then gain +I am getting more or less the exact same error and it is making flume unresponsive to shutdown requestsThe interceptor isnt necessary to recreate it +Im not an RMI friend but based on all the back conpat discussions we should try to misplease everybody equally +But in any case I have commits on my git repo I can squash together +I still find some +Good sugestion + +The maps are short lived so its not hard to imagine a situation that all the reduces start bombarding the jobtracker with requests for maps at the same time asking for more +Moving issues that arent going to make it into to am happy to close this if there is any question about its usefulness +Added to group which has this permission +Source manipulations worry me +I meant for this particular situation not in general +We only need to review the documentation +Its not just a matter of not being specific enough +Im sure this is a result of changing how the extension manager worked +I didnt bind the goal to a particular build phase bc executing this goal seems like a rare operation +Mass closing all issues that have been in resolved state for months or more without any feedback or update +trunk +The buffer is nulled directly after unmapping +I mean user could create a custom server name +linked to root am getting the same error when trying to create a Directory of Spatial Files + + +Passed +Any updates on this? It would be a good time to clean this up +Im back on that now +Good luck! Leo +or HTTPSMTP thingsIn this case I dont have a strong feeling either way so Ive changed the keys to be named as things +I dont know enough about how headers work so Im going to fix and comment out and put a TODO there and someone can address it on Patch commit revision . +Paul can you explain to Alexy how to build ESB using your new JBTIS target platformThanks +The patch looks goodI am attaching the patch branch +In this case should not blow up +Reopening issue to correct target release +submitting v of patch to Hudson +has plans for native executable invokers which will solve that problem + +Pull it in again Todd if you see it and have notion of whats up +If you actually type a single character for the username you get a validation error +I could have the wrong returns the table names in both the database explorer and the hibernate any chance you know if postgres have issues under DTP +Consider opening a new issue for follow up work +Thanks a lot for your contribution +IIRC something like that is suggested in the original Kleinberg paper. +Is there a resolution for this? It would be nice to not have to download the source and maintain it anybody require the current implementation +New model versions means newchangedremoved elements +Based on r +Assign to Ales. +you attach a code to this issue +Having a true monitoring solutionwould be really nice though +Adrianit has now been moved under WebtoolsConfigurationTemporal expression +Thanks Suja for verifying the patch +Patch applied +If there is no such the method will get one from the +I have briefly looked at your patch +patch applied +Moving out of +We need to refactor the standard implementation to accomodate new configuration attributes for expressions +m on build machine and it passes on my laptopChecking this in +The SQLXX classes themselves comes to mind but we should also verify that there isnt a mismatch between what the SQLXX classes expectassumes and what the store is capable of +The following patch was applied to the SOA GA release on behalf of a customer this disables the calls +Can you confirm that +I tried both incremental and TB +Looks good +Thank you Ivan +the simply contains the links I found for downloading some Air that I found particularly hard to find +Fixes both this and +do you have a way to replicate the problem +Whereas works fine +Fixed svn rev is a patch to add the compiler test for the expected error message +need unit tests +This is a good question +never mind +The changes have been committed into trunk +You should definitely see if you can improve your python script by looking at a static resolution file instead of dynamically pinging DNS at run time +Its necessary to make minor corrections in guide text and code patterns due to the application update Alex Tsebro will do this basing on information provided by Andrew Markhel +OK +This project was completed. +Sounds good resolving the issue +PS Does it make sense to separate the issue from this one? +Sure but since Grid is already involved in getting these sample projects into shape hes the best one to look into thisGrid assigning to you +I actually did not see any OOM problems running the test suite before without those parametersAll tests run through with the MB standard jvm memory +Thanks Jeff +jconsole esbsmx ant script client app using ibm apiAll of these applications were passed the usernamepassword information to be able to connect to the on server tracing it has become apparent that the usernamepassword information was only received by the server using the jconsole command +You have not provided enough information for me to either understand or recreate your problem +I have tried the code you supplied and whilst the inappbrowser opens I just get a white screen like before +No look and feel jarsclasses getcopied to Geronimo jars +Yes it is something that should be tested in the PFOR codec. +Do you want me to move into the bm package? I put it into the language package because that is where all the other encoders are and I presume having them in that package allows them to be automagically imported by things like the lucene configuration files +jdbc +This issue has been resolved in but not ported to the branch +This patch resolves some conflicts with the svn revisions for the ref guide +This issue is happening on +committed revision backed out the test change and am reopenning this issue untill tinderbox failures can be resolved +Passing over to Richard for inclusion of session passivation content + Niels no time to look into it right now but Ive added Justin and Christian as watchers +From the logs Ive seen the deadlocks all look like the client just didnt receive the expected message from the broker and is stopped waitingWe should probably make them timed receives with long timeouts so that the tests will eventually fail +Hi! The patch contains a bit more code than this issues requested +What Id be happier with would be for a API build on something like that had enough of a stability guarantee that client apps be they IDE plugins command line tools firefox addons could talk the far end through whatever proxies and firewalls got in the way with the API compatibility rules helping the uploaded code to workAs we dont yet have a based API yet we could design in some of this stuff from the outset +test case result for +I was using where I needed +reviewed +Ah this appears to be just a really feature +It is intended that you map Integerserial Long stale resolved issues +Committed trunk revision Committed branch revision close for +Thanks for taking a look Scott +is a core element of the framework and its behavior should not be configurable but the sinkssink processors etc can be +Committed several fixes to the code and it seems to work pretty well now +Makes sense I just wanted to make sure +Your tweet a few minutes ago may mean this is now finished too Arrayshuffle We certainly could change Arrayshuffle to do the logic +Author of jsch repsonded upstream but still nothing significant to add here +Heres an updated patch ready for review +Ok looks like everything is done please test and let me know if you have any queries +It looks like the router that is used is the not the +This is not critical +Bryan Could you take a look at this? Let me know if you want me to rebase it to latest trunk +oh and we definitely dont support Tomcat +Fixed via +This is one of most annoying bugfeature I have ever come across + +Further evidence continues to mount that this is likely a race between lazy initialization on a first request and following requests that come too soon afterwards +That way itll be easier for me to replicate it in a testcase and fix it +But when use cql this problem do not exist +it turns out that all the keys used in the hints writing are the same key at least in the examples I looked at in the debugger if Im correct in this observation this is a serious see the key in hinted table is the ID of the dead box +Im trying to verify it now but it seems like the problem is in the TCK itself and not in our impl +Ability to see docs on the website manually copy a local bundle of docs to the website and add a linkIm attaching an archive to be put to the web siteNote its better to use Linux while working with this archive unzip l grep xceptions drlvmintfdochtml drlvmintfdochtml + +project level administrators are part of the administrators role in the project itself which you have +It creates a table with a point geometry that only has a single record for vehicleid +Do you mean Spring Security? If so could you check with the latest version? If there are still could you possibly supply a test app which reproduces the issue +This sounds very much like This provides an install button on the bundles list page which opens a new page where you can upload multiple bundles using the multifile upload pluginThis has already been released with the Web Console releaseAm I right +Its not different from running cassandra now which listens to all +renamed the local variable +Thanks to all who reviewed! +Hi MartinCould you tell me where I can download the withThanks +Try adding theversions to the Maven POM I dont think the surefire report works with different versions of surefire also in the pom. +With the Synchronizable interface it is possible to get notified about changes to all objects +If the property is not in the pom to where the dependency is linked then the property is not an update for that that support inheritance the comment that supports closing as not a bug is somewhat curious +Ant precommit passes when run from the top level +mrepositoryjunitjunitmvn X deploy Dfile Durlfiletmp ve just noticed that if you use false it wont generate your pom on the server but it wont delete it locally +Code to test the bug +Please reopen if this is still an issue thanks +IGNORE would be the right value for the OPV flag I guess +I ran and which passedIntegrated to and TRUNKThanks for the patch ChunhuiThanks for the review Ramkrishna. +Will the ever compare equal nodes? That is is there any expression that will cause the same node to be passed to this method? Perhaps a union expression? Something like a ? You could then turn this around and see if a givens the same result +Note that Spring doesnt have explicit support for the JDO eitherI see +From a quick look at the description it sounds like some service is not being reliably stopped on reload which would imply it doesnt have dependency chain that leads back to +We got a green CPI last night +will be falsely triggered in any environment which has slfj present which will be most environments +Marking critical and +why is that a bug +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +However I cant make it happen now +It also exposes the list of of services and and also exposes the of the they belong transports we could include the endpoint address in the object name? +Committed patch svn which replaces the system property setupteardown code with +Seems like it at firstlook +Verified in JBT JBDS +The changes look good but I was not able to test the patch since there are a couple of tests that are failing one is and the other is +You can use any JTA runtime +Shame now Ill have to write the method itself +This implements functionality for PLT +The idea of validation classes is just wrong +Reader and Writer just need these two dependencies to work properly +Closing with the release +The JOPR needs to make a admin connection using a password combo +Sorry previous was wrong versionSomeone with admin rights could remove the older because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The earlier patch did not apply to trunk +The fix is to move the append to Partition so that either it is the leader and it finishes the append or it rejects the produce request since it is becoming a follower +Did you +This means that get will return something that works correctly for Class objects +Corrected the frameutil and rulecollapsesource logic responsible for determining the insert pushdown query. +Geir I think I need to fix the patch for build file and remove the testing files from the default build +I keep an eye on the build logs and will update this either way once I know something +So in similar lines we can have a n option to list option defaulting it to hours. +Had to implement my own folding menu version to handle the very long code page menus +Clint Im giving you this +I do not have that information but asked the customer to expand on his use case +Okay Ill try to get it done within the day +Works for me. +Updated patchWas thinking a bit more about the STOPREQUEST +Indeed I wasnt sure about how aggressive my checks should be +There is one related to vmware but that is hypervisor specific +Im moving all bugs out to and we can move some bugs out to to get some release action going +Hes right +Actually this doesnt cover +See user list for many emails related to this problem +on patch + +but if a metastore has support why not use it? if a metastore doenst support we could fall back to use CAS +That is not much more work than installing the Oracle tool and is perhaps a more familiar procedure to Maven users +But once changing the width of the DOS window everything works fine +Attaching zip file with dtdsschemas from jboss trunk build +Renamed it to createMoved this service toRemoved duplicates effort of setting emailAddress in infoString field +In this layout the tool bar will not disturb me more +I added the jar and made it so that when you do +Reason Auth fail +wtpmodules have different requirements +Thanks Andrew reminder +really showstopper apparently not only happens in webinterface also at runtimecontinuum queries db and +That finally triggered the think this is rather a shell problem +I think I nailed it +So I think I can commit it +This is the fix for handling hidden files correctly +Hi DennyDid you get this issue in your clusterThis peice of Code moved to in latest trunkBy seeing the peice of code i did not find the issueBecause after removing the element he is breaking from the loop +For example at the very beginning of the script if you have something like thisnonexistingpathbinperlHowever this should be a negative test case so I will lower the priority of the bug +The same translation unit but only gzipped +Thanks Giri for fixing this. +Ours were in the range of secs +Fixed +Thanks for commentingBut my point was not that it didnt work in servicemix because it does like you describe +Resolved in Trunk deadlock seems to be in the Python Library +I dont think there should be any objections +We could let report returns a flag indicating if we want to retry or not + +I told him that this test will be removed with my web service changes so its only disabled temporarily +resolving as wont fix for now as discussed with thomaswe will revisit this topic once we have a concrete use caseneed for a file system abstraction in +Also what should I do about the testcaseAttach it to +There should be some other tickets which is related to this +If I include the schema delcaration in my then I get a duplicate attribute jaxb attribute in my resulting WSDL +I have assigned this to you +closing it as resolved. +Hadoop was released. +Have to move to since this came in so late while was being released. +Committed fix with regression tests. +I generated the patch from the uimajtrunk so that may be the reason why you cannot apply it +the test seems to succeed on a mac as well as on linux +please link the other JIRA as a blocker of this patch is having a dependency of patch +I will take a look at these patches +Future objects +bulk assign all unscheduled issues to next release +Process the elements of the parents and then of the current typeTanks for patchingWhen will with be available as compiledI am curently using a local artifact of version with the applied patch as I sort of had problems building a version +It would be fairly confusing for anyone looking at this at a later date theyd see refer to the spec only to find it not there +Resolved compile error. +Ondrej can you please attach logs so we can see what the problem really is +to separate out the compaction policy refactoring that bit is nice + +We can open another jira for whatever is causing the timeout once we identify it. +For now will continue to have the submission process be manual +So for instance I would say that we should not boot if we think the db is readonly but we find work to do during restart recovery +For a workaround it looks like pubsub works fine +There is not a known issue that causes this +Mr +Either we add a new site configuration option where a comma separated list of codec classesCurrently using this as configuration +I will see if I can recreate the issue today +Additional unit tests were added to verify the tokenizer and parser behaviorAll unit and integration tests pass with these changes +An exception is thrown if deletion fails +Ill send the material personally mine m dot schipperheyn Ive spotted a problem with this and fixed it on trunk +Would you be able to review this patch + +Ill have a think about how to move the code +the change is only on and remains the sameThanks +Yes that latest release did fix both issues +Ive used implementation taken from jline project that have ability to mask input characters if needed so Im closing this issue for now +By deployment I mean deploying war in tcServer v +committed +This one has been around for awhile I am assuming it is dead +Fixed under r by making consistent with fullapp and updating the README to match +We dont have control over the scripts +The issue is with Hibernate +Committed to trunk and +Seems to be a style problem only remove content pre see in that case add backgroundnone to content +This was specific to the Remove annotation. +They should share common methods in but code such as that for reserving room dates moved into the specific cart itemGood luck and thanks for all the contributions! and All I am working on the refactoring of Order Checkout Process in the Ordermgr for creating different sets of definitions and two different progress bars one for sales orders and one for purchase orders as par the description given in this issue +Tempting to think they should be moved over +Hadoop upgraded from to +Stay tuned +Ill commit this soon unless someone objects +I believe the issue is happening on a multivalued field where the document which contributed that field has been deleted +This will take just mins maxthanks Sergey Currently I have several fronts that keep me busy +I didnt try with so far only with +Needed to include xerces deps to get a sensible parser that supports attributes in case the host JDK uses a lame parser by default. +Hi UweThanks for reviewing the patch again +submitting +Planning on doing a release later today and incorporating this +Martin +Committed to trunk as r +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Closing +Verified in WFK. +do you need any slave access or just the web ui +Thanks Eric +This patch solves the problem seen earlier +I hope to get this fixed in jruby +Fixed with revision +OK so it sounds like this issues isnt resolved yet but then neither is it a block on the release +Excellent thanks for your contributionCommitted revision . +thanks +Please change to if you intend to submit a fix for. +We should specify this for the same reason that we specify that the ints over the wire are +Hi David My account has been created finally +If you feel strongly about this raise a separate JIRA +Added +bulk fixing the version info for and all affected issues have in else have thoughts around thisOne performance concern of mine revolves around commit the vast majority of people used it for visibility of documents not for persistence at a specific timeIm warming to the idea of a flush param instead of softCommit and it seems like perhaps it should default to false for agree we could clean this upI worry about flush since it used to mean something else +This image is my dslr file with Chinese characters and some errors. +ShrutiWhat is your sortingranking criteria? Can you post your full queryOrdering can change when you add the minprice +Just out of interest what happens if you turned on prepared statement pooling +to treating this as a question of validation rather access this makes very much sense +This may be a duplicate ofFixed with commit febaeaaeaeeffb +When will be released to include this fix +Jamie to be homogeneous in all subprojects Im gonna update logj in NMR components etc +Its the same as with HTTP sessions +js files +Patch looks good to me +Ive changed priority because it blocks! +James block currently implements mailServerget is a method of the mailServer service and is used also by the mailetDo you plan to introduce a new Service interface to be implemented by a new component or to split James component moving the implementation to another component +Would you please help to try it +Great example Army +srcmainjavaorgapachehadoophbaseclientThis probably should not be publicstable I cant see how they would be used by someone outside of hbase +This patch needs to go in a completely different direction +Dont know why Hadoop QA said that tests failed for v when all passed +fixes failing tests in launchpad +The problem with the current version is the namespace isnt for a specific language +close it. +Last call for feedback or well close this issue +Would you please have a try and apply the patch on? Thanks a lotTo apply patch simply click TeamApply patch +Unfortunately Tomcat doesnt support setting the character encoding in the Valve +Moving to PA +Is it not possible to fix the website between releasesI thought easy website publication was part of the idea of svnpubsub +Let me know if youd like me to have a look at the combined patch +Thanks! +stack That is what Hadoop does if native compression bits do not exist for the platform or were not installed +not everyone watches jira +We are migrating to jboss cluster and cannot believe this problem +Id be willing to rethink it if someone else wants to chime in +Could we investigate this peculiarity in our implementationAlso I found a race condition in the test code +And it also shows the user the error message got as HTTP responseThis feature will work for all plugins automatically +s suggestions new patch created +I can create a patch for this but i dont know which is the best approach because of some issuesa shutdown the broker gracefully and display a warn sign that the broker is shutting down because of invalid brokernameb trim the starting and ending space and change spaces in between the brokername +I verified your fix your fix is better +Setting fix version to ER build to on ER. +If the feedback on this is OK I will of course continue to cover the rest of the system and provide a second patch +Using a semaphore now to track the active threads +Glad to know Im not crazy thanks Juergen +Thanks Arun and Robert for helping refreshing my memory Note the above list is incomplete +This is the list of mime types recognized by Jetty out of the box +Castor JDO supported this +The issue fixes as part of closing. +Is there an ETA on a release or potential backport in a point release +Need to change Beta branch +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +Updated patch is attachedThanksParimal Gain +and pushed both upstream +with the relevant info and provide a patch please +Thanks for all the contributions +However the bug I ran into is that my scrubs do eventually complete but the information in compactionstats persists +At least I was able to build beta on Solaris with Sun CC and +But like Si Id prefer to be sure that nobody has changed this file between +Or you change this internal URLS to actual URLS that our code understandsbundle file etcThis is the reason why Scanning provided as SPI +We need to set up JBOSSHOME via the envp array and this will resolve the issue. +In Firefox the st columns are too narrow +I agree about not mixing in fact I wrote about it in the first place +note that the above case updating node was refused +it seems we should use is +I tried this on a handful of sql and it does at least a good of a job converting them as the original patchI do think this could be useful in the future as well as for the current test converrsion project so if no one objects Ill check this in shortly +Also I need to double check my comparator understanding bc perhaps the doc ids are off due to getting a top level reader +That failure looks unrelated +We have no interest in the juddi documentation examples nor definitely not an ESB task +False positives when used as default parameter s add as exceptions by default and enable the rule in Sonar way +Another option is to change STRUCTUREDPROPERTYSEPERATOR to something more uncommon +Included unit test I just want to point out that the unit tests that are failing were failing on trunk without this patch +I would like to provide more info on the previous comment files are failing in range of bytes after a first block of successerrorsuccesserrorsuccesserror successerrorI have tried to change the different buffers and windows size with no positive effect +Probably it is a good idea because in theory developers should be able to invent new keywords and overriding make things work +fix offsets for thai analyzer so highlighting etc will work +Pull request sent +This should be simple to add +tell me what modification i have to do inorder to support user authentication session multitenant approach in my application +Hence moved these lines to the parent method +So I wonder why File and URL do not have an encoding would be quite natural in my opinion since Groovy extends the semantics to also cover the file File this could probably be done by implementing extends +Committed revision shutdown Autodeployment File Scanner Thread during Spring bean lifecycle +Also please include the exact code that reproduces the error the in does not compile at all +Test case Start with the following environment variable settrueFor example add the following toJAVAOPTSJAVAOPTS true Deploy attached Run attached test that sends a cookie named mgm with value todayfriday to httplocalhosttestboxservlet which prints out the cookie The output file shows the value as just todayNamemgm Valuetodayfriday Version Apply the patch and run the same test +I have just committed to Committed revision Have not seen any new failure in my env with this pathThanks a lot Nicolas for your contribution in this issue and closer track on itThanks a lot to Nicholas for your reviews +Currently the junit testing framework enforces a timeout for each test +The problem is that the change for brings deployment level logj classes into scope of the commons logging layer but in this case the deployment level classes are not consistent with the newer logj classes the commons logging logj factory is looking for +Thanks NicolasYour patch is in trunk at revision +what is it about? +Ill add a test tonight +C Seems to be also a performance all old Resolved issues +Ping bob +This is actually good news for Gora Ill continue digging regarding +At that point we can consider backporting these changes as needed +If you really need it in unified diff format please tell me how to create regards u OLDFILE NEWFILE or use your IDE +LOL! I didnt type that that is a message from JIRA +rev of release. +Reminder to setup Karaf branch on Hudson once the branch is created +Thanks Paul! +Otherwise check if the block length is the same as the block length sent by NN +Client is writing a file with replication I meant a replication +Then I started with a fresh source tarball and tried to figure out how to just get it to log an error instead of throwing an exception +How about levels of access +bq +Same problem happened so I think it might me something with the dispatcher type forward +Reloaded fresh version +Updated the correct one +I have been testing this feature for a few weeks now +delaying since we need to clarify the busyFinally I think I understand what you mean +Is this still an issue? If I get a chance Ill try the attached project but if its out of date and someone knows that that would save time +Its public to allow visibility and access from the Test code +This design would make it easier for you to manage patches from my local fork +Like that we are able to drop something into the dont support +Im guessing the reason the tests failed is that my original patch doesnt trim all the whitespace +David do you have a sample test which could help us debug that +for packages +Todd any news on this? Pauls suggestion should work if Im not mistaken. +Support multiple version strings in server pluginSolved issue where server never goes to started statusFix bug with method permissions annotation on session beans being added too many binary files to go with common code to modiuleUpdated build script to pick up new downloads from as but with an idea plugin too +Verified in +With NIO in JDK I wouldnt hold out much hope for future file improvementsAlso optional platform specific features really go against the ethos of Javas write once run anywhereAny reason why reopening is a bad idea on other platforms +however the build will still run on +The note pointing to the Security Guide is right at the beginning of the security configuration sectionThe note about configuration of security pointing to the Getting Started Guide which you requested in the previous comment is added right at the beginning of the Security Guide just below the initial paragraphs under Part I +At the moment i cant reproduce this BUG +Basic Java Config completed +The patch was written against XercesThe problem is when the breakLine is called it doesnt properly check ifthe new line is necessary between line and text +Needs some rework +We DO like to have tomcat in there though +Proper patch since the earlier one only reflected the webapp project by mistake +I havent time to debug this issue yet +Committed in branch needs to be done in trunkcoresrcmainjavaorgapachecocooncomponentspipelineimpl pipeline locking has not been merged to trunk I cant apply the modifications in this issue +A test case demonstrating the bug +OK I am going to use Edwins suggestion since most of code already there in handling +this is not about the default behaviour +Ted not sure why this would break rolling restart +This should probably be assigned to the as it seems the issue has not been solved at all +This can be done through passing HADOOPOPS env vars +which is kind of a pain +So clearly those dependencies are required +waiting for ssh keys to be to latest version. +type attribute are going to be reimplemented in scope ofLets reject this one once implemented. +Consensus seems to be just move it into the main cassandra jar because you cant do much without the classes there anyway +Address findbug warnings +Reading the documentation and attempting alternative methods to set parameters didnt work eitherI mention this because I hit both of cases previously presented as reasons to fix this issue +Property changes on CworkapachezookeepertrunksrcjavatestorgapachezookeeperserverName svn nativethere are some changes that are not a patch fo this patch i think and are alos not relative to the src tree! could you get rid of them andrew and upload the patch again +Moving this from since Im trying to cut a release candidate and this is not diagnosed +Shouldnt the fix actually look for compatible server runtimes instead just showing everything +We need someone to volunteer as release manager for this +The algo is too slow because ofGUI the current one is a temporary one but should be replaced by prefused library as suggested by Ted +bulk defer of features from to +Thanks Karthik. +it needs to markCompacted each group w the cfs it came from +need to add int tests +I will test and get in quickly before it goes stale again +dbbcbaffadacbb is changing code of commit bdfeffdceabaecfc +Fixed in in +At the moment I have no ideaI will chat with an eclipse expert in our company +and this would result in inconsistencytherefore im inclined to resolve this invalid +OK so what it is? Anyway if that is expected please close this one +Im concerned about the verbosityWDYT +How frequently does that happenUnfortunately it really depends on how big the files are in the region such that it would decide to include all the files +forrest patch +Ill have a look but theres no easy way to this +I havent seen it +Thanks Bowen +Thanks Vivek! +Ill have another look but I already read pages of issues after a xslist search ando didnt find anything I could match +Committed +zip files +Thank you for your comment! +Patch for gridmix with newapi +bulk defer of open issues to +I completely aped and keeping up with all the stylistic idiosyncrasies of the two until someone is willing to clean up the whole shebang +If someone finds a way to reproduce this please reopen. +Once is committed this can be made a LATER but perhaps tag it so that when someone sits down to do the migrate they will know the problems that will surface +Tomi Thx for the contribution +They have different defaults for example how many default cells whether or not it automatically grows whether or not is had cell borders by default +Some changed made on marketing component impact on common forms here a very small patch to correct itIf a framework committer can review it then I can continue to clean manufacturing componentThanks JacquesI agree with you but I have simply rename the Segment labels to and I didnt see this issueI propose to move those labels at common level and remove it fromSo do not consider this patch I will provide a new one as soon as I canThanks a lot cleaning on Manufacturing and components and some italian translations +Daniel Please check it before using +I thought it might be a descriptor for serving static content or something +That makes Solr easier to secure and is also the biggest use case for this feature +perfect Dinesh could you also prevent the check from logging an issue inside a static method +Kindest regards work Fintan +The projects have been combined but one of the thoughts was to create a better looking examples and clean them up a bit +Ill take a look at adding a for itColm +Sorry the previous patch was for +I am ok with the final patch as r +Sorry that was my bad +Konstantin Ok i will change the size of bufStore as you suggstedNicholas Isnt it better to keep the serialization of inside rather than putting it into? thanks +tentatively set for by upgrading picketlink IDM +This for example happens with Array Inspect in Ruby mode In Ruby Mode all strings should be declared to have the same encoding apparently this is not the case in my project +Unchecking patch available review comments indicate additional work to be done +Fixed +My times a little constrained now and I dont move nearly as quickly as I do in Java but I try to get something written up +I would be surprised if this was missed in testingPlease check! +Need the svn url of your sites source +Closed upon release of Hadoop. +Sorry James my mistake I didnt realize that! +I try to upgrade the conf +I feel rejectedP But I can understand your approach. +Koen delivered the NN trunkdocumentationwhatsnewjbpm Koen +If you come up with some tests just as a separate patch and let me know about it +Removing obsolete variables +Im not sure I agree +Verified +Another wish we had closures one +Hey Nigel +I think both of these are crisp and I would be ok with both of them +Hopefully that is a comprehensive list +Your fix for that part looks good +Though the description is different the indeed requirement is same as +Probably the broken mobile browsers got fixed +Thanks +Needed to confirm that this didnt break any tests and theres currently some instability in some of the network tests +Andrea Justin please use +I just tested on Linux and it works fine +Interesting part is that some of the other are depending on the latest kernel and system +The first on line of the file retrieve the same marker either with or +Is this an issue in an OSGI environment only? Because I cant reproduce it in the regular engine deploy ways +Hello Tom in fact this problem occurs each time a maven extension is defined in the pom file +I need an example to work with before I can schedule this for had the same exception after my first project set up +I use full name including path to save the module packaing mapping info so there is no problemFor the prepended and appended class path element I didnt save their mapping info +Not done yet +samplestore and store tutorial scenario were all moved to use the new binding too +I imagine you cache validated objects to avoid infinite loops +Ive retried and yet to duplicate +Just to follow up +Dont see initialized +Can you attach a thread dump of the Solr JVM when it is deadlocked please +I was unable to reproduce that issue here but I added appropriate null checks to prevent this in future. +This has been fixed in response to +Sorry Ive been away since Thursday due to a family emergency +Feel free to reviewcorrect them +Part of a bulk update all resolutions changed to done please review history to original resolution type +As long as its optional why notFancy providing a patchregards when I get some time +I dont see how to get determinacy without itDeterminacy is pretty critical for most programmers and applications +Its a bit slower at the and there might be problems with highlighting if you use the regex to change the field length +Please let me know if this is acceptable to you +Hi KoushikSince its not a feature addition but a bug fix is there a chance we can roll this intoThank youilya +When I did the refactoring I did not know that this was expected to work so I did not even check this +I will check tonight and tomorrow morning +Patch attached +In addition if you want a pure http proxy you should use and off the shelf HTTP proxy no need for a gatewayRegarding webhdfshttpfs interop the test does exactly thisRegarding config keys I was keeping things as isolated as possible from the rest of HDFS until Webhdfs Hoop share the codebaseRegarding a code sharing JIRA as soon as this JIRA goes inAll the other comments Ill integrate them in a follow up patchThanks again +Unfortunately the to the test files do no longer work which makes verifying the problem difficult +That will also give you access to the log directories via the web server +Patch is should be fixed. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +A patch that deletes the staging directory after the system directoryManual testing showed that with this patch I couldnt get a recovery failure in the scenario in the description +I added This source integrates the changes Matt made for dynamic columns into the version of the source code +If not then what is the default policy configured? And it means that we can never use an object of same Entity bean in inner transaction if it has been used in a transaction starting inner transaction? + +The patch is screen is deprecated +This needs to be fixed for think there is a similar bug in the Python library +I just committed this +OlegI also consider isolated CR a so for option +The previous run is still in progess +I can understand if we dont have time to do another release but in that case we should consider downgrading the recent update +The real change +Thanks TomI put this into trunk and +Thanks DanielApply patch to TRUNK and +Obviously it doesnt +Ran my new test again passes +Rolled back changes from branch for now. +The user group mapping is repeated in at least two places +right? thoughtsbq +to master +patch for jvmti applied at I had to make the following corrections to your changes +Most probably starting slash symbol was mean to say that the starting is missed in orgrichfacesrenderkithtmlscripts +This is not needed with the fix in +I know now atleast that it is a bug +Weeks have passed and after going back and using the released version of version on my project I can no longer reproduce this issue +This should make it easier for testing +This may help narrow down where to look +The WSDL generation code is currently generating theTuscany wrappersI think its important to make the change to using the JAXB wrappers because this should fix a number of WSDL generation problems including andIve been looking into this and I have fixes that seem to produce a clean build with all tests passing except for the problem being tracked underIm therefore reopening this issue and Ill add further updates here about the code changes needed to get the JAXB wrappers fully working +Every time someone addsdeletesrenames a column in my superclass database table my code breaks +No it is a not same bug please open a new one +Thank you! It was easy I just didnt know where to look +adding custom matchers forwhen the detection fails and stuffBut i only did lucene corecontrib so far +Ill amend accordingly +i noticed it when i was trying to improve performance by leveraging Out instead of relying on getters which get called many of these getters rely on an entityManager which is closed on destroy so I was getting far I am handling this by checking if the entityManager is still there is a better way +Ok I will make the change + is accessing conf file on DFS for reading job view +as the javadocs werent published +The issue must be fixed now +Rafal please provide exact steps on how you do thisOur attempts to get anywhere near this is if we do not start from Eclipse JEEI must be missing something could you show the exact things you install from our updatesite +Because disabled controls cannot be successful there is no need to validate them +Baby steps +Please update this task. +I will add here patch +I believe this issue is popping up as a result of catalog theory is tha the old configuration subsystem kept disabled stuff only in the config area but did not move it to the live catalog +If you believe the issue is not resolved please reopen or open a new issue +Bulk closing issues resolved in JEXL. +Hi ThomasdevI dig this and found that when we get the class for above pojo class it always expect setters and getters for all types but boolean primitive +Postponed to +Patch Grzegorz patch applied +The reported error is still valid but the patch here can be ignored as the patch for committed all the test cases together at revision +I just created an issue for both +Ill let you know how we get on with for child contexts. +Updating fix for versions to cellar releases +Part of a bulk update all resolutions changed to done please review history to original resolution type +Ok +Nothing was changed since and that is something what works for me under Windows with user wo administrative privileges and it works when installed into default location +Verified +Or you could rename the group title to something short like EJB Ref Type and then put the instruction inside the group above the radio buttons +Please do not create JIRA issues without asking on the user mailing list in the first place writing a comprehensive description of your problem if it happens that your issue is really a you read the first point of my comment? please ask the user mailing list before creating issues in JIRA! +Bumping the urgency of this issue +I am looking at backporting this fix to so setting ownership to myself +There are a number of issues in this area linked to and they need to be addressed before we release +And if the community feels comfortable with doing so it may change to storing data that is not backed by files in HDFS +Unscheduling due to inactivity +Lets see what are still not covered by first +So my above comment is not applicable +The problem is that tcommandLink is not compatible with the RIso fast fix dont override the rendererproper fix make tcommandLink compatible with the RIregards lets discuss possible solutions to still have a dummyForm functionality even with RI on +Supporting vLan get would break compliance with the bean spec +Changed Target Version to upon release of +except that the test that is failing is one that explicitly tests s ability to parse a that does not have an XSD +deferred in favor of can use this to cover anything from the threads subsystem itself +or by adding jsf into a shared ll check it ve also sent an email to IBM support to see if they are going to incorporate JSF into WAS anytime soon +v looks good as well +I believe this is fixed in since we no longer depend on Ill try again with a snapshot +I added unit tests showing the issues +what I did as a workaround is to put a targetblank on links to download files +So are you thinking maybe instead of +IIRCIf I had to choose between the two id rather disable the stress test on jdk then have a parser that lets uncaught exceptions fall out +At a first review it seams you forgot to model some properties like +To be more general we need to add a notion of admin env so admins can set MALLOCARENAMAX and others automatically for all containersMakes sense +There are reasons for both approaches and each will cause different test behavior +Plz comment there. +I have checked in the changes along with some test cases to the branch and the trunk +file size defaults to mbIt would probably make sense to disable memory mapping for the backup store as it doesnt add much value there +If anyone wants to contribute some unit tests with use cases now is the time will help to ensure its solid in the release +I updated some unit tests as Im not sure that youre properly handling whitespace +verified on Tomcat with Chrome Firefox IE and IE all found bugs on with Chrome Firefox all bugs found reported +Yes lets do that +Polling would be a nice second step for sure especially if it can be set as a low priority thread only active when nothing is hitting geoserver +Implemented in revision Ok its in +Thanks Robert for contributing the patch +Hi Joerg feel free to create a new JIRA ticket to ask for this feature +JT crashes too frequently to get a decent statisticLets check in a few daysYoram +I have to hardcode my return url currently because wicket does not expose an API to generate an absolute URL +Thank you that was quick! I will try the patch and see if it happens again +duplicate +Hi guys Im new to the MJS project +ChristianI walked through that a while back +Verified on EAP and RHEL. +thanks +Affects all released versions +What would be a real problem is if you end up with weaker isolation +Any comments on the fix? I was now planning to go ahead and apply it otherwise +This also affects AS +Building the index makes the listStatus call as it needs info of all files +Kudos to maven team and thanks +Can you verify it occurs when you have ONLY Eclipse Kepler R or SR installed without Tools or JBDS? Secure storage preferences page isnt one we contribute AFAIK +stubbed master doc +think the simplest solution is to move applyAliases invocation to +Does this mean that this patch requires to be applied firstNo +I have found that if I create the full directory path that the references then run the goal again it will work +btw changing this methods means we have to change the public interface and this will not happeing before Railo railo it does not matter if the exception was thrown because the key does not exists or something else +Patches are available +resolved by implementing all our interceptor logic completely in +in +Should be resolved after the patch for was applied +The problem described here is rather tricky and seems to be an AvalonPhoenix issueFor the time being I suggest as workarounds to use JDK to be able to enjoy or switch to another protocol if JDK is neededI will continue trying to understand this +Willem do you mind if I take this issue? I would like to release a new bundles set soonThanksRegardsJB +Ive logged a bug with Sun and hope to hear something back from them soon +The patch looks good +The bug is due to which starts to look at the line read from position instead of +Didnt we drop that already +If I change the SLD of the layer metriaGVATGARDSOMRADE from our custom style to the included style polygon all requests show a scaleline +Uploaded a standalone reproduction of this issue +You were saying that JAXB annotation would not fit to JSON so I deduced that you would prefer to develop annotationDo you know a standard to do similar things with Json +Not understanding the request to include this in the google calendar +Thank you so much Dag for the very thorough review! I will figure out how many files are affected and decide what to do accordingly +How about this +Another workspace example using and additional project to move resources out of classpath +For other it will simply use false so the object refs will be guessed to have bits which is fine as upper memory limitThe code does only use public Java and falls back if anything fails to false +hmmm +and above are now solved +If it does not work we will pull the example +groovylib +No functional changes +not ready patch too bugy +I absolutely like this Idea +Will be fixed together with +Fixing mismerge on the last patch +No tests included since this is just a refactor +Even so I would prefer a fresh patch +Resolving this issue as the patch of solves this issue too and it got committed +However is shared by both and +Without any caching serving resources from an exploded deployment is super slow +Does the th of June output include output from the patch uploaded July th +merely stands to benefit from improvements there when it happens +Same problem and error +We should make NM export it as part of the env +The proxy approach requires two security implementations for outward facing security and internal security implementation +Thanks Kim it looks good to me now +the latest snapshot version of the plugin is now available from the asf nexus +Debugging simple configuration problems falls into that +bq +I would tend towards all the sharing the same JVM settings cant edit this ticket so I will add this comment +It would also be a good ticket for someone new to the code to work on as it would be hard to get it drastically wrongIll leave as blocker for now but Ill say Im on letting it slide until it has a sponsor +Committed to trunk. +Just uploaded a script with tpch queries Q Q Q and Q written usingValidated results against what is in the SpecScript includes ddl at the bottom +push out to +wo +push out to patch I just submitted for should also fix this issue +All three engines appeared to be in the closed state +But thats fine to me to close this issue +Change in testsuiteimports was introduced in commit for +Thats right this is using Java for both client and server +The limit has to do with some of the low level posix apis and how they are implemented I saw a post some time ago about how to make this better but for the time being its generally inefficient in most filesystems to have large numbers of filesdirectories in a single directoryAlso as you point out it makes it next to impossible to easily add additional storage since there is only basically one mount point +Fix verified in revision +Note how the guava code uses a static inner class and then the get code catches any exceptions thrown trying to classload it +Its will take +But this leaves us with admin accidentally formatting additional namenode without specifying a cluster ID and a cluster ID is automatically is generated +Thanks again +that was my first ticket commentfixing attempt that Ive written using Mac OS X +Flume is not exactly designed for that and replicating a channel on multiple machines was never a design principle +Danny seems confident its the right change and Jake felt it was probably right +But I think that until we have a concrete for such a method we should keep it as simple as possible +fix in since is. +Patch Passing Unit test are cutover more to comeThis fifth patch is to monitor progress and see if something is wrong or could be betterCutover along the way looks good Nikola Ill commit it to the branchI removed the nocommits from I think they were leftover from copying from Document +I think what Im saying is that if we can open up the norms computation to custom code that will do what I want right? Maybe we can have a class like which apps can plug in if they want to customize how that length is computed +Im about to do just that +This seems to be common across the ASF Jira deployment +Do we have a use case for more than one out there? Also the prospect of each having their own plugins makes things a bit crazy +The capacity scheduler is the default scheduler +Looks as if updates to include Java would be needed for and the trunk assuming is meant to support Java +Reattaching +xlsx file and a password protected +yep +Feel free to fix the code and test it outThanks +Adds new interface that things like HSK BU and BO implement making estimates of size that is not just count of payload +N Youve written it as a proposal which is good +So that will be seen in the classloader +bq +Thanks for the patch +batch update of issues with fix instead of the correct +In particular the client bails out if it encounters that the datanode is already doing a lease recovery +Hi Darran are you still going to do this one +having three nodes fail at the same I think your guess is right something in the coordinator must be corrupted and once failure happens on one node it will happen to other node that wants to had tried to debug the problem and I got as far as the coordinator had received JOIN request and as part of processing the JOIN request it was trying to do a flush but flush return false as a result it doesnt generate JOIN response and add the member in +Note that I havent used hadoops as the code is not yet moved to and hence there are compatibility issues + copied to EWPEWP updated to includeSVN revision +Attach a patch that the to avoid the file deletion issue +The very first line of the message is error which I think makes it more severe than DEBUG and INFO messages normally are +Ive tried it on Linux but it seems that it only fails on Linux. +this is probably a wont do now +See the duplicate issue for a patch by Julien Nioche + +Its good to hear Now I just need to understand why AS doesnt have that filter since it means everyone using esb in the community will sure get redeploys but also get it when they dont want to +Would you be willing to help us with the implementation? +Added support for +This would be less error prone nd more reliable than other tools +The problem seems to be with Ant +Thanks Bill +However I didnt realise that Chris had originally intended to return a response which upon reflection is the preferred solution as you point out +Kishancan you show me the ls ltr from usrshareconf? I need that to determine why the upgrade path fails for that fileCheers applied to as efffbcfeabaaPending answer around Hugos question above before resolving this bug +Ive committed and pushed this to master +Attached that moved test to increased the timeout time while using less region servers +Is this still needed? I think this has been fixed by the flurry of webhdfs jiras I recently worked on. +In the first instance I think that the code that adds the groovy nature to a project should first see if those libraries already exist and if so use them +Need agent log +bq +Lets do these one at a time or were more likely to introduce bugs +Im asking QA help setup the test for Ingres +Now I am out to walk the dog and get some fresh air +Applying actions scheduling actions and user control relating to groups. +Looks like we have to change that +Still trying to figure it all out but I could use a few pointers at where I should look +Attaching +Clearly in the new hooked approach this is wrong and there should just be a finlaize am thinking or new JIRA issues rather than continue on this one? +Username rickardoberg +Ill update it shortly +Do you want to provide a patch +Thanks Sean for filing this issue +See also for another Java object marshalling issue. +Initial implementation of methods for for creating and attaching volumes +Can you provide a patch to fix this issue in a timely manner +Id like to add that this patch is crucial for the demonstration of Arquillian +Needs to be commited to Branchx too. +My editing skills were not +I recommend a commit +Moving to +Im working on a project that uses CXF and am seeing this same problem +Replicating all of zookeeper would not work well would not want to replicate info related to the root tablet location tablet servers loggers and FATE operations from the master cluster +Here is a failing unit test that demonstrates the issue here +Attached an updated file +This issue is not observed with Patch A path pathHence closing the bug +did check with trunk build deployed on my dev environment and its now replicates properly even with the incremental index +You would only see a difference if you sort all hitsI think we can commit this too +Added comments about options to query and search methods +true but the necessary pde option destroys my manifest +this problem only occures if the state is stored on the server side +And I have an illogical liking of the name but I cant identify why +Upload the same file again +So we should really only need to change the version string if we stop using AvroNote that even if this turns out to be wrong we can always add a version number to the version string later +I kind of got held up by the combination of limited time and apparently being required to engineer a standalone build for the patch +and this Issue are imo the sameCould someone please apply this to svn since it has a lot of impact yet the technical problem is a small oneThanks +Support forgit gems in Gemfile was indeed added in Warbler. +This avoids patching any code in the boot image runner +Jason thank you for all the insight +Fixed in rev . +I would prefer because otherwise it would be impossible to use Thrift without installing Logj or SLFJ +I think it may be better to find or create other +I am a newbie of axisc +Thanks for your patience +sounds reasonable +I have a prototype now working locally +Sebb there is no remedy to this but wrapping the map in another one without patching the input taglib +Iteration this must be the right one +Is this issue related to the code base +in general are pretty uglymaybe just mark and let people subclass at their own risk +the described behavior is reproducible with a simple login of a single client +Amazing! Thanks Harish +this is probably because the as code is ignoring the durable perhaps you can link Jeff to your quickstart to show the issue +I think there should be an explicit exception for an invalid filter rather than +Please compile the file likejavac encoding classpath Usersyokojrubymainlib runjava cpUsersyokojrubymainlib t forget to specify the encoding for both compilation and the patch is applied the test code will output? but correct Japanese characters after the patched is applied. +Merged in master and updated Chemistry to in BOM. +to force text to be quoted +Here is the simplest test to reproduce the problem +If were going to take a real shot at fixing this we need a test case that displays the problem +its only for debug +I will do it once the branch is merged back +Lets handle this problem in +Theres no point in turning those into templated files but the should be able to figure these out correctlySo the aim is to essentially create a single set of JSP files where the View can be customized far more freely than before with the brittle Controller bit moved into and the providing the Model +The actual version of xalan use d by us is XSLTJ Java Rohit +koen is it actually all the sources that is needed or is it like with seamgen in seam where we will extract the template part to use for tooling independent of the runtimes +this is clearly a functional test so it make sense to create a separate directory structure like testsrcfunctional +seems very much worth it +The project has a reference to which is being used as a JPA implementation to bring the tests upI will now start on the service implementations +i am actually adding supports dynamically throught out all my ajax xcomponents and it when components are generated dynamically using cretaeComponent and when added with support i get aj undefined +If Roger is wrong then please +Mapping to JIS also creates problems in some Arabic +I havent actually tried it intomcat so could you do that and if its ok close this bug? Thanks +The new API on executes a Pig scripts and returns a object that contains a graph of execution of MR jobs. +I got bogged down at work +It has been flaky forever but seems to have gotten worse for me recently +I dont want the behavior of the methods to change. +Verified in JBT Version Linux +We can convert to spaces later +No issues appear to have arisen with this change on the trunk in the last hoursand my merge and buildtest results with were very clean so Ive committedthe merged patch to the branch as subversion revision Im a bit concerned about +Committed + +Changes onlineRegions to a CHM and removes the synchronization on the get +I am not sure if too many people use them but I do +Thanks Kan Jitendra! +has very low flushing which if it were to be read correclty by split policy would cause the region in the test to split causing failure +Marking as resolved since no further feedback received all current fixes are on trunk and in the branch. +This appears to be confined to the Lava XOLO x +I dont see what is the problem of having member variables in a +Working on security +Thanks to Matt for finding and fixing the second one. +Didnt know that was required +but Im afraid they will convince you that this is not possible at least not without and even then it might be faster to reload from the database as we do today +I will review Cristis patch in a day to see if we need any changes +For example a RS can enter GC pause have region reassigned then come back to life +The Users in this room list is missed also +Seems to happen only with Mojarra +Interestingly if I remove the java client flag the test passes +Steven heres a fix for that test +I think I missed the because I simply dont use it But if thats the only one from using it its a good signId also like an early to be released soon also with and as added by Eitan +add Java SE in JDBC Public API Consult this javadoc if your application runs on Java SE and Java SE in for thinking about the JDBC level supported by +Im not a ruby user +This was originally reported against XercesJ +Using that fixes my problem +Since the change impacts what a customer might expect to see in the log I think it needs a release note + +In that sense doesnt really belong as an Infra task but is convenient for tracking purposes +For the final version of this I have removed the arity dependencies and computation functionality +This isnt a patch for trunk and will require some work to apply and right now in my opinion the current version of the patch is not really using the Metrics classes correctly. +The was due to stime +Look for usages of EPR and Courier and from that youll be able to see what needs sorting out + +Committed to edit log branch +Leaving this as is for better backward compatibility the new foreach index functionality always begins with . +I tried to find the file in my JBDS install location and override the guvnor tools preference value for testing I only found the file in the directory jbdsinstalllocationeclipseconfigurationbundles +I am actively working on this +Isnt that a good thing though as opposed to nobody ever being able to get a lease on the renamed files until the daemon is restarted +I should add the patch also adresses Knuts comments +and the press the OK button +to trunk and +Added logging to test that failed +If I do not write any comments until than feel free to send a reminder +Maybe a better place to inject the expiration policy would be into the object +And this is a big change from to so we better check this a bit +It might only be working because its used on meta so theres no timestamp magic going on and the updates are not that frequent +Well actually you cant see any whitespace +Two points to make Have erroneously typed MSVC SP in the descriptionI am using SPThe command I used to build the cProjectsWinVC was msdev MAKE all Win Debug USEENV REBUILD OUT! The most of the Test applications you has mentioned are a little obsolete and use the old +Hmm +I guess the only way is to change the interfaceI almost think thats the lessor of two evils +ReallyWas this fixed or is this bad triage +We want to try and support as much as we can without having to have forks of each client for major version changes as seen within python +Its either fixed or a duplicate. +Yes exactly thats the problem and we must find some solution +Norman Cool last jar is working with HQ +You used the right field +We may also then have to partially localized files particularly for distributed cache. +Fixing this at the generated wrapper code level is the easiest but that leavs us with memory leaksDoing this at the Param class level is the cleanest but that means lots of work at this point in timeI propose that we do this at generated wrapper level as proposed and fix the problems at Param level later +Forgot to mention is removed but there are still references to it eg Override +AntonConsider this issue for filtering design in also +I havent find test results on hudson so I have run locally test and all tests been resolved tests adjusted. +Bulk move of issues into after vote that we merge branch into TRUNK up on list +I gave it a try and it works great on Safari Chrome and Firefox on Mac +Agree +batch transition to closed remaining resolved bugs +Sure Ill watch it the next days and if it doesnt occur anymore Ill let you know +Heres a from me +Looks good to me too +Deployed to on +I forgot to mention it the is based on the OLDNow I realize I did not make it similar to the of when moving to +Ok this task appears to be resolved a lot +Attached is the patch for +Patched applied and pushed. +Its EAP scramble at the moment +I saw how some other modules are deferring those by overriding the test methods so I am tooThis latest version implements the NRT search stuff in the latest Solr with the exception of deletes those are still doing hard commits afterward +I had to apply the patch manually so please verify that I got everything right +can you provide a test case for this +It was uploaded on September +The difficulty arises because its unclear where the responsibility of knowing a tokens TTL falls the or the +So there is MRU seems useless +I checked the code and theres nothing dependent on unions in the tutorial so nothing should fail +There are so many case where the url could be wrong that I would prefer not to have this to my previous comment how do you provide the url if the user is starting AS with bin b myhostHow do I get myhostIn case of an AJP connector how to guess the httpd server name +Also it should be noted that Struts is targetting Java so annotations are definitely an option +looks good to me +Hi Jay Can you provide a patch for as well? Im running into similar issue on Windows +Snitches are all configurable at runtime +Looks like the dependency for Tyrex is actually needed as the plain junit tests for the cpa module wont run successfully without this dependency specified +That way all standard JBI components would benefitfrom this featureWould you mind writing a patch that way? Else Ill try to find some time to to itin the coming weeks +Also super helpful for refactoring to add relations in the middle without propagating them all the way down the chainI like the syntax too +All examples seem there +You should be able to specify charset +Which means that this might have to spider into the individual handler functionsBasically I cant think of a good place where we differentiate between attachment requests that are streamed to disk and requests that are buffered +matmul X test program for matrix vector multiplicationmatrix size loop count places axis for parallelization The time is s +please provide more infoRegards solution seems pretty simple to me have project manager use the view and put the work effort application back the way it was +Heartbroken that you used emacs +Just catch the error and bail with a meaningful exception +Code and UT with small config provided by Amita was applied by Adriano because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I started to review it +like highlighting the adjacent position with a unified background color andor wrapping them in a strong encompassing boarder +So if we choose to make it we must deprecate the existing name for at least one release prior to removing it +We just experienced this error +from be verified in gatein master and reopened if still valid +But I believe the optionalfalse is a necessity too +Yep were using at the moment and we received this error just yesterday +Let me know +The split itself will be retried by the Overseer Collection Processor again but the test does not take that into accountOh okay so the fix is really just fixing the testIs it the same thing with the chaos monkey shard split test +Isnt this done +I have now merged it to the branch and committed revision . +A PDF which demonstrates the is a more general issue of incorrect parsing of incrementally updated which I am also experiencing +Rob Excellent! Could you send me some instructions on how to set this up so I can play with it +Closing all resolved tickets from or older +Closing tickets +Committed some work in the client packageThis is a big job +I guess the first one +Please give it a its not working then please supply a sample project for this +bq +Thanks Corinne! +Ross any updates? If not Ill go ahead and close this +Parser change is good +At the very least a manual utility to merge all tablets less than X would be very useful +Patch files are based on +The test failure above is unrelated +Please provide steps to reproduce +thanks for reviewing +Resolving the issue as fixed by. +at this point I suggest that my jira not be linked to this onemy comments removedand let this one go foward with chrises guidance +In the case that the original response is too large to cache it is not closed byI have also included a test case that you should have better luck reproducing this error with in +Yes this class is configurable in the xml file we could use our own in GeronimoI recreated a patch and once it is accepted I will do the change in GeronimoThanks +Do you mean because NNAs ZKFC does not have Cs address or there is a fixed mapping of the ith NN to a server and that mapping is staleThe proposed solution is to make every standby ZKFC restart when it discovers an active leader that cannot be connected to? This would mean all standby NN would be rebooting in the above scenario when C becomes master right +This is the patch that probably solve this issue and now I will commit +Ive committed this +Patch for the issue +Fixed and Verified in and this seems pretty severe as this is probably one of the first use cases people will have if theyve been using jBPM in the past +Rebased post +Bulk assign all unscheduled issues with priority major or less to target +Fixes pushed in e +Attached patch is for tests not for commit +Fixed in r +Its a sporadic issue probably due to a port lingering +Thanks Grant +ours is just spread out all over the world and we have very few dedicated resources +My best fix is to just remove it +Other than that it seemed reasonable +no i havent tested this use case with the new build plan +I have committed this +Would you like me to attach a patch +Verified in WFK +Still gives you shared resources such as scoped Hibernate Sessions etc but no strong rollback seems to me that the warn logging is appropriate since it points at a semantic mismatch in your configuration +I am not aware of a workaround +cancelled as per macs comment +I dont see that happening hereThere is a difference between configuration management system and package management system +When the callback occurs and the bookie adds the entry to the ledger storage then the mark is passed to the sync thread +attachment more of a proof of conceptstrawman to illustrate how i thought it might addition or instead we could make it so that which elements from the responseHeader are set as http response headers is configurable in for each request handler +Moving out of +I think DEBUG priority is more appropriate in this case will apply the patch with a few additions to HEAD and when I get hometonight +Optional dependencies should not be removed +Reclassified to improvement since the behaviour is correct although slower than desired. +Marking as BLOCKER as Datasource test development cannot proceed due to this issue +here is the c program. +The test in question is atve run the test a few times and I got two kinds of errors +for the patch +Brought over the unit tests including modifications for the new API +Correct me if I am wrong but I believe this may be caused when you dont have JAI in the classpath +Perhaps my memory is playing games on me +In SVN rev I updated the eclipse classpath file for the mail jar changeFor the other files in your patch +patch applies clean and unit tests passThe design looks cleaner than before and it looks positive +Hey Mike the patch looks good +version +One solution would be to force the implementation to read the entire message like you suggested but then the effects of would be better solution would be to create a Connection abstraction which is opened before the request message callback is invoked so that the request message can be streamed to the server and closed when a response callback is done with the message so that the response message can be lazy loaded +All issues are either reported expected failures or are environmental issues where the database could not be found +add initCause to exceptions caught from thanks! +It saved me some time +Thank you for the patch +If people dont feel this necessary Ill happily remove themThe tests for the package do not add anything therefore I havent moved themPorts to Queries module +You then have to send it a rebuild again to actually rebuild the index +AnilPlease stop reporting issues against +I performed the change on the hibernate code is the state for this issueThis isnt fixed in the version +A transaction timeout in the DB is the only way to resolve such a situation +This has already been implemented in . +Committed to trunk +I think its ready +bq +Fixed in close of resolved issues. +Given that there a number of unknowns about this feature it cannot be marked as a blocker for release +i left this open to remind me to work on the wiki pages which i did without remembering to resolve this. + +Patch attached. +Please advice if you see some problem withThis feature is very cool +Bob Swift opened to pick up on an implementation issue following on from the solution provided for this issue +And r. +Did you file inspect issue? +We have upgraded to tomcat in marking this as resolved +SCR has been released + +Other related operations are functional +Just curious why do you want to save to a non file URL? To write the configuration on a FTP server +Lets wait for M release and set this preference in JBT +Any ideas when it will be resolvedThanks reason is that the XML signature of the file specifies Cp +The patch doesnt appear to include any new or modified testsThe existing unit test should be able to verify this behavior +Besides if a task reports itself done its doneResolving but reopen if am wrong. +Isnt that too large a name? Sorry I couldnt help but notice such a huge name which I thought was just how you had written it but then I checked to see that such a file name actually exists +Yes NikThe fix works fine +The XML file is and the XSL file is which calls +Attached is an updated version of the patch that refactors things a bit +Please verify +Calling for a vote to the REFRESH zip obsoletes the other zip correct +Stack trace +Thanks Carl +The support for such a loop is only valid with XOR gateway and not OR because it simply not deterministic meaning with OR there is always both paths possible +I meant. +Please someone help me +Yes PatrickIt would be cool to confirm there too +thanks for the effort on this Alejandro +In addition to Hoss great patch this one changes the MANIFEST file in the demo jar and war in a consistent way adds a patternset via refid to the metainf type of jarify +What exactly doesnt work anymore now? Which exact versions of Spring does it work with or not work with? it does work in Spring but doesnt work anymore in Spring? Is this what youre experiencing? that time I tested the version and apparently it worked +Attaching and a repro for this issue +Unless there is something more under the API level that I am not aware of +same with assembly do not add comments on a closed issueIf you have questions please use the user mailing list. +Revision +bulk close of resolved issues. +Latest patch looks good +Can you please confirmRegards the command line I ran bash c ulimit a on OSX and it worked as expected +Yeah it is the property file that I am concerned with +If the classpath contains a directory and that directory has a database as a subdirectory then the database can be accessed via the classpath subprotocol +Please try this patchThanks a lot +It is ok +Ill commit shortly +Paul just duplicated it on windows +The shade plugin supports automatic merging of services files. +This is the project that we use for Fuse Training and it has been created in the past using Eclipse IDE and not +Can you move these tickets out to Future since these do not meet the criterion for inclusion in would also suggest that Domain Admins be able to access Usage Records for their domain +So use inherited member from Andy +as in v +Merged to branch in df since some users are reporting this with Rails +Then would not need a finalizer method +Im very interested in your suggestions +Yeah it could be the horrors but it might not be too bad if we just put the modules in place and gradually migrate code +Thats how I did in a test case I was working onFor the inbound could be worked around with a Type Converter but as far as I can tell the user will have to provide his own version of a for the outbound scenario repeating or overriding all the logic in the do method +Thanks for the patch Jacob +After a quick look at the diff it appears that it is not minimal +The bug is that all openers are blocked waiting on a meta that was not going to open or that was openIm going to close this bug as +Moving this as the target is currently for AS Thanks for letting us know! We will look into this when we resume this task +The configuration switch would be quite a good solution as you can then just put your application into mode? It should probably also be added to the documentation +This test fails because of configuration issue +If you find further issues please create a new jira and link it to this one. +s fires when the shared core is closed +Seeing that is coming up is there a chance to get this fixed? I dont think its resolved and its definitely not fixed despite what Deepals comment saysActually at this point Id be happy if someone even so much as commented on this issue as a sign that its not completely off everyones radar +Are you talking of the site plugin or the webstart plugin? Do you expect the plugin to move the generated webstart into your site? Ive never used it like that +But the general idea that you could implement most of the informational functionality of the File API with a object is in my opinion a nice property +But we should think about more scenes +I personally find this a bit too much of aprice to pay for not using when the entire entity body fitswithin the K buffer +Seems something else is wrong then I think +If this is still an issue please mention specific points where documentation is confusing and reopen. +IMO we should do the warning for Hadoop as well +bulk close of all resolved issues prior to release +Theres no need for full verification process because the effect of this fix appears only when an advice binding is brokenThanks for the review NicholasIve just committed this. +for the patch +svn diff srcdocs is attached +This is now supported +Please bear in mind that you wont be able to use an expression such as at all as we wont be able to access driver specific constants +Hi Charitha if possible could you please the patch and click Grant license to the ASF +That method will throw appropriate exception if the index is closed +sizing critical for +example how one might define this interceptor in spring to supply default exculsions +VPE can not open and display files in correctly +This version takes those out +I think that something needs to be clarified in and before modifying +Stripes wants to start itself first so that it can stash its configuration in the which one of our classes in later needs to find +Bug has been marked as a duplicate of this bug +noting I am not thanking of havening two compaction threads just two options a compaction can take we still que up a compaction check on a memcache flush butrun compaction above if oldest map file is X days else check to see if there are or more new map files less then the max compaction MB set and run if true +I also brought up a small node cluster and ran a few tests on it and they all passed +were backported to as well adding the latter fix version +I dont see a way we can reach out and influence a solution of any issues they will sooner or later introduce into the product +Cloned to Beta +Ive marked this field as final +We can reopen if the demand appears +HiPlease see the patch and commit itI have specified the problem in my above commentThanks for your precious timeRegardsAshish Jacopo for taking interest in my patch. +The problem was not on the broker +Right for sure +This method calls the method send which triggers an ajax request +So unless I can duplicate it this remains a mystery +v lgtm mostly except for the remarks at the beginning of my previous comment +over in +Suggested improvement to capture namespace snapshots and related test case +class psuedo property behavior +Started to work on this going to fix a related issue +Your patch is already in the SVNThank youDeepal +Resolving as duplicate of. +I have also seen the Derby jars go to one directory and the testing jar to the otherant clobber was run before the build ant buildsource buildjars followed by ant testing derbytesting jar +I missed it +Thanks Myrna. +DoneWe are to report EVERY MONTH for February March AprilThere after the months for Pivot to report are January April July missed reporting in February so the reporting schedule is now insteadMarch April Mayand thenFebruary May August November +Looks good +What is the recommended way to return UUID valuesGood question +You dont know if it is also legal to truncatedrop SQLCARD in the reply do you +part takes size of region memcaches at start of flush and then subtracts the size on flush completion rather than set things to zero as soon as the cache starts up + is closed now we good here +Thanks Army +I did not try checked exceptions as I dont have any cacheable methods that throw checked exceptions +Hi LeonardoI have always thought that JSF should add a for such issues +Sorry for the inconvenienceCheers is the ticket. +Even without DBRE Roo will always add the entity manager to the entity ITD as it is generated from anyway +If theres only threads in the pool all messages are consumed +Thanks George +I will do this and amend the patch if necessary +OK I agree +Assigning to Laura from the docs team +Hi Carlosthanks for your contribution and yes tests are always highly welcomeCheersve attached the test I use in my project to make sure the code I attached before works +It doesnt need to unit testing +for the patch +Thanks for pointing this out +can we consider this as a bugWorkaround to turn off subquiries would not be an option when it concerns bulk updates +Moving to backlog +This can be disabled back though if no one objects +Attached is the patch for the that fixes the applied +After examining the behavior with and without the patch I currently have attached to this issue I believe that the current behavior without this patch is desirable +Arquillian is marked as experimental thus not blocking unless serious critical unsuable issueIf you believe this issue must be fixed before GA please set fix version to CR and provide patchreason +The javadocs look good to me +with JBT +Why is the classpath specified in both and +fixed in revision +So I am marking this one as Wont Fix +Eclipse MAT shows thatClass Name Objects Shallow Heap Retained Heap Object of type are using MB of the java heap size +Killing a task tracker and restarting a tasktracker takes off all the information +It deploys fine when manually altered to be compliant. +We have a plan for this but it still needs to be fully needs to be broken out into pieces with some pieces scheduled earlier than CR +attached is the patch from charles comment causes all lines in the patch text to be prepended with spaces causing the patch to be invalid +Thats what Im doing right now but this feature would reduce the amount of configuration to get this working +Predicates dont seem to work directly against an +Im attaching an updated patch that applies cleanly on jBPM +Sven says that the new revpos variable associated with attachments is now a functional replacement for the intent of this ticket. +So how do we get this process started againI apologize if any comments I made about the the were hasty judgemental or offended anyoneMy concern right now if to get this process going againSince the deprecated code will only exist through Struts I would be a on whatever method we chooseShould we call for a consensis voteI do believe we should provide a deprecation path for usersWhatever method we use providng a deprecation path will only cost us about minutes more than a straight rename +Ive set the version to why did you change fix version to? +I think the bulk of the May PM comment should go into class comment or into the hfile javadoc +I had to increase the loop because sometime it faisl after more than iterationsLets the bug tracking begin +Also once his account is created your pmc chair can enable this access for him +This issue has been fixed and released as part of release +So edit logs with same txid but can be differentiated using epochs +Tried it out looks good to me +And I change there drawSequenceflowNow it draw flow as one line with rounded angles like in + Bug has been marked as a duplicate of this bug +Feel free to reopen if theres a problem. +Can you also supply a patch for thisWith pleasure +Thanks for the explanations Todd +Will commit a bit later unless somebody beats me to it +Vitaly could you please take a look +As discussed on the mailing list here is the patch for the functionality with test case +Taken care of in issue is fixed and released as part of release +Closing issue for now test cases pass for my JCRSPI based custom implementation. +Fourth case +Confirmed that the new tuning suggestion is in the Tuning Guide for. +reposting so jenkins will AllenRecovery uses the normal edit log loading code so we support loading slightly older versions just the same way that the normal edit log loading code does +Dont forget to update the online documentation. +sparklr runs on port and the tonr instance tries to bind to the same port not the port defined in the wiki +for review under +The name of the session ID in cookie is sessionid +It probably needs some more work but its fairly straight forward +Its my understanding that based profile is replaced with new profile logic so we dont have to distribute anymore +When is run with the sort argument the following is produced for the second pageThis is mSecond line here +Need to test on distributed +If not please reopen. +Ive just committed this. +I cannot reproduce this bug +Applied patch in rThanks for reporting! +Fixed comments on the latest patch? It should give us all the functionality with maximum future flexibility +I did not try to go all the way and check it on real code but I am confident it will be possible +While using the Export tool I can successfully export the entire ear archive to the Server with out any you need more information please let me know +Hans Ill apply to my local build and test with it this week +I have identified the file handle leak and fixed on my branch +I read the javadoc for a while and I didnt seeany reason to believe that our current implementation shouldchange due to the NULLS FIRSTLAST support +Looks like the cdk plugin may need an option to set encoding to it was necessary to redefine defaut system charset +Im not sure what I think about the additional complexity of an exclude specification +Expect a release in the next week or two that fixes it +Thanks Harsh for seeing that +As a frequent user of named vectors I would love to see this supported generally. +The separation has made the tests impossible to run wihtut moving back into the lib directory +Committed the change and update the Release Note for it +bq +committed +The log statement at present says SASL config path set to xyz regardless of whether the path is valid or not +Retry v against hadoop I have a script that can parse test outputI can contribute that script if there is a proper JIRA +Closing all resolved tickets from or older +with Pojocache reqsec +Patch fixes the issue by allowing to return a doc map that translates doc to their new value so that can commit merged deletes has been modified to make deletions more likely to happen concurrently with a merge +Ive also done the CP change +I believe loading the configs more than once to not require full system restart should have its own ticket and tracked separately +thanks for the quick turn around time. +Applied at revision +That is good advice and thank you kindly for the reply +For a general solution it is a bit too specific and we should discuss all possible solutions to do this +I am analyzing the failure +Any progress with this contains same problem +snapshot folder snapshot feature is disabled if therere snapshot in the +Thank you for your participation +It seems to work initially but not any more later on +It is unclear from the documentation on if there is any optimization outside of the Job scope +ok I saw your checkin just now +I am closing this now because I think there is no much benefit to do thisThis will increase complexity of the code. +Fixed in trunk rev . +classpath and +HTTP server should starts before those components +I dont have a preference other than to see it placed somewhere otherthan in a branch +If no severe regression will be reported will close this issue in a day +Ill be looking forward to it in +Use it at your own risk +This does look like a bug +We get a few hundred boolean clauses in ours +Successives calls inside a conversation are executed serially because of exclusive locking of the conversation +to be clear you can check and change the options at these you for the information +If I were to do it they would implement an interface +There are quite a few conflicts with the latest trunk +Hadrian +bulk close for the release +I have written a batch file for Windows and the tests passed in both Linux and Windows +New patch file correctly generated using SVN against the SVN trunk +that is good news +When I add then tests pass +Order by descending wont work because its not exactly the other way around that orders the nodes +columns returns the expected values +Simple app demonstrating the issue +I am looking into this today btw +Since this is the case I would recommend not adding convenience methods to the builder +Defined constants and implemented support in Rev +It will be a good idea to remove the dependency of this patch on the deprecated patch fixes the issue but only on the previous patch fix it on the filesystem side which would requrioe changes to all the filesysstems +Workaround prerelease works +if tests pass and you tried it up on cluster +If no one is on it we can try and get a patch for itFeng +I incorporated your comments and am resubmitting the patch +why? +Reviewed by Cliff +All deps for all Weld projects now this go in the Weld reference guide? I would really like to have this information in one place to minimize the work to keep it updated +Ran into some troubles with the clientserver decoratorsThis patch is a tangent to the changes required for meaning that it taps into one of the decorator with the port change +Steve sorry I see what you mean by the add operations in the both and +Thanks for this Jeremy +Thanks a lot Jrgen I will give it a try tomorrow +The failure to resize volume is because the virtual disk with IDE adapter is attached to VM as SCSI virtual disk +The main idea is to make reconnection transparent to the client since with the current code a client has to reset its watches manuallyResetting the watches does not make it fully transparent though as there might have been events between the time the client was disconnected and the time its watches have been reset +everything seems to work as expected +Attaching +We have no way to specify the inner schema now +Committed revision Added new property +Yes of course Mike taught me that +removing the distinct fixes the problemill keep the upgrade of the oracle dialect to also updated the oracle driver version from to +Change the wstx jar to versionad +You must be looking at an older version than what is in trunkThe current implementation uses a Jenkins hash rather thanYou are correct that there is no guarantee how implement an array of booleanPerhaps using a would be better +Moving to if you need it sooner let us know +debug log of common Chandra +Will we get a sync storm when we have a lot of column families +Hmaster start ok +Oriol makes a good point I think balder in the literal sense is not just any sort of but has to be too big rather than too small +Thanks Gavin! It seems to be working fine +Thanks Liang. +Pending review comments from Hairong +The eclipse files can be generated by ant Before importing the project into eclipse for now one also has to do ant packageAfter these two steps are completed the project can be imported into eclipse and at least it should provide all the editing goodness +Closing old task. +upgrading the Guava dependency from my mistake +Will commit when passed +Reg timeout monitor the one major change is now the CLSOING state node is created by master itself and it was done by RS as in +thanks to all the developersthat helped me clear this up +Improved patch added request map to +Closing resolved issues for already released versions +Thank you to Chuan for contributing the patch and thank you to Ivan for the code reviews. +org team helped me get this figured out! +Master is our release branch develop is our development branch +yes that is the bundle containing all the dependencies +Closing since there have been no updates to this issue since it was marked resolved. +TEST Test code not criticalEX EX EX EX ACCIDENT ACCIDENT EX EX ACCIDENT ACCIDENT ACCIDENT ACCIDENT ACCIDENT ACCIDENT EX EX ACCIDENT EX EX EX EX EX EX EX EX HAS ANDHAS HAS OREX EX EX EX EX EX ACCIDENT ACCIDENT EX ACCIDENT EX EX EX TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST AccumulatorTEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST EX EX EX EX EX EX +Now that work is underway with AS all previous community releases are +change header level for Search page has to be hChange browser breadcrum to use icon and search is in the same line to save space on toptry also to use icon from same library for browser ul liAs lots of space available when browsing artifact put version and info in the same page doesnt work with this patch +Here is a patch that will abort the resolve job if the resolveQueue is already emptyPatch created on r +I dont believe its perfect but it does work +bq +Changes made to CVS +Works well Linda but Ive the feeling that in naming conventions are not so well standardized than in Java so would be better to remove this rule from the Sonar way quality profile +A strict format can reduce errors but may be inflexible and the exception can be little harsh +Yes +fixed in. +But when referencing that project from another the groupId is not picked up out of the installed POM +Heres an example log from reproducing this along with a thread dump +Just curious any plans for this? +I dont expect well be able to take action on this very quickly but as soon as something is viable lets jump on it +Thats not a legal URI +Hey Arun since you seem to have been adding the typed bytes binary streaming patches to I wanted to suggest including this one as well because then would be fully compatible with Dumbo out of the box +So we may need to skip this version of dojo or look into the problem in the classloader and dependency viewers +A couple thoughts This forces anyone trying to use the to provide a loggerName in solrconfig +this is blocking the implementation of quick fix support and anything that tries to generaterewrite code via bulk close of resolved issues. +Fixed long from Resolved to Closed per above comment. +Set my classpath to include javabuild javabin javabin +The present s behavior is wrong +Just wanted to add more commentsIm using WSAD server patched to The resource being loaded is wsjarfileDJavaWSADWorkspacesDevorgapachestrutsresourcessNote the ws prefix +If you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +It was a transient issue introduced without an affected release so it should have been dummy. +contains follow settingsAxislogsAxisconfCProgram FilesApache GroupApacheAxislogsCProgram FilesApache GroupApacheAxisconfDWorktempbinTransporthttpDWorktempbinChannelHTTPDWorktempbinChannelHTTPSSLDWorktempbinremark I write path with quotes and without it +Added Security link to the sidenav +Was fixed in to mention we are using version +is buried inside at the moment +patch looks good thanks for the fix +I just committed this +Not a blocker +bq +That said Ill bounce it from M to Beta so theres enough time to verify this in the codebase +The old PDF is something with buggy Id better to setup one new doc one this section +If youve ever had to wear a pager in support of a batch system you would feel the pain of someone who has to be paged at in the morning because an exception hit on a record type no one thing would be introduced +Seems it needs more investigation and a separate JIRA +Indeed our s were not setup right +I will be returning to the lab on Monday +Applied. +A restart of the master and slave cleared everything up in both our testing environment and now in production +It seems that we are going to a full CMS integration. +You can specify dependencies between Jiras using the link button in the menuThe semantics are a bit fuzzy but in this case requires looks like a good choice +I also ran a few client commands and verified that client failover continues to work with security enabled which is a lovely surprise +Alibrary that only needs to dynamically load classes provided by the library itself cansafely use the Class method forName +Resolving to LATER based on feedback on the dev list from the Shale dev team +I can now see only first issue +Changes to branches in Rev reduced the build test time for on my Mac from down to minutes with SunStill working through the changes to trunk as the new tests rely upon lock timeouts +Thanks. +Im attaching the test case here for reference +Can now close this. +I just committed this +If you run emulate from the cordova folder how would it be able to find the xcodeproj file without a path to it? I am on any shell information i can provide +yes +FYI after committing I intend to immediately move up one package as there is nothing tier specific about them +The parser now supports enums however it does not make them available in the model +True + the fix version. +It is more reliable to monitor tasks using virtual memory than using physical memory because virtual memory is not as dynamic as RSS and we can stop the task before the physical memory goes highAnd I also agree with your second point +This patch set a default image type if the current image type is TYPECUSTOMso in this case a new buffered image can be spotted thanks EricYour patch is in trunk at revision +I think to have default console configuration selected is good +If nulls are received it indicates a programming error but Derby shouldwill not fail +Yep +Xavier is there anything for the REST side +Problematic file and testcases attached +Thanks for the prompt reply guys +or given up? What issues did you encounter? + +One commentrrefsistabssysperms ISGRANTABLE No other table has an ISGRANTABLE column +Updated Patch minor changes to fix forbidden api checks and documentation lintThanks Erick and Robert for the review +Patch uploaded with simple test +can you review this one +I guess will resubmit? Hudson is becoming useless +However if you are concerned about compatibility then we should go with Owens suggestion +Changes to enable abandoned object tracking and removal +Ill put some flag in to handle this +If people want special events they could subclass andAdditionally is there any need for pre and post change events if Vetoable changes are properly supported? I think most of the logic for this sort of thing is implemented in and doesnt need to be reinvented here. +Specifically in this case the failure should abort and shutdown the HA NNEven if we fix the NN shutdown case this bug still needs to be fixed or else the edit logs will be left behind in an inconsistent state +From this information we will write the final release notes for all issues + is related to this I think +Modified servlet registration such that all possible path are registered in Rev +would have been nice if they have just set it to the first two digits or at least asked +In either case the problem is not in BURS but in whatever part of the opt compiler earlier injected the new operator into the there a path through one of the write barriers that is being inlined in that is raising the? If so that seems like an obvious path to cut with a pragma +Do we need base OOP library also +Which jars is it that is needed +defer all X issues to +That definitely looks suspicious +What do you propose as a fix for this? There is also the stacking tabs issue when there are more tabs than the screen is wide +Attaching patch for +Thanks. +Also need to document from performance point of view this is not preferred implementation since the simulation would make extra function calls + for the patch +Doesnt need the test case since it just gives nicer interface for the existing configuration i need to know number of splits +actually have this problem when I run the default version with it configured to use +Considering there is a ceiling of files should it be sorted +Now also affects to be backported to branch as well +Final patch +Suggested text Chosen Java VM is a Java Runtime only it will run Developer Studio but it is recommended to use a Java SDK +What versions are you running how do you createsetup your projects to have this happen to you? +Thanks for your checking +Also I added a test to check that a is thrown when an entity is modified in another conversation but no exception gets thrown +It can be committed once we agree on where the test should liveIf you wanted to improve it further you could a Use instead of assert +It should look just like an expiration of the Java sessionThings that are not yet implemented The region server hasnt been modified to use the native code at all +Attached are files and +Ill try to find some time for creating a dedicated test case using but it might take some time before Ill be able to do so +Tested with standalone runner and the Ant task +for patch v. +and the D option for selecting a target path needs implementingas well just as it was for the maventest siteBuilds are working tho +By M we must have some movement on this +Attached test case showing what doesdoesnt work +Do your examples output different html? Does this happen in several browsers +Patch supporting using file channel to write +JIRA issues updated to reflect there inclusion with proposed Karaf release +Again i just wanted to mention i think its fine to commitConsider my comments +So i guess we need a different Form when partyId is given to display that listWhat do you think +Didnt we already promise that minor releases would retain compatibility +Okay where and how can I help +I dont think this meets the requirement of being a critical bug fix + + for to trunk. +Lets commit something very close to V soon +Thanks for the comments MateiI see +this is not a valid senario +VickyYes the cases were reassigned based on your profile change +build test +this is being ported up to ESB directly +To use untar then ant f src clean build deploy it assumes CATALINAHOME is set and deploys there +Felipe it would be great if you could test the fix that Herv just committed +Patch is submitted +The patch looks good to the attached patch as revision . +ResolvedAdded iPhone to the devices menu +This is the effective patch Im running on my local James fork +This will help enforce the contract +If compliance is the main goal that focus should be increasing the variety of compliance rules on the source tree rather than including Test sourcesGo for it +Lets wait until I was able to find a solution for +more +Synchronization of fConf is critical since your code is modifying the fConf which was previously +This patch is closer errors and failure on the test +The command line I used to generate the stubs wasjava l c or d adb wv s S unrolled u uri After changing every instance of axisarraylistt to axutilarraylistt the generated stubs compiled successfully +Out of curiosity can you try launching with the stock from the distribution +Adding the configuration file to examples Difficult to maintain +If somebody disagrees please reopen +We use and the issue is thereERROR Error Could not parse predefined CMAP file for +Id rather keep working on this now that it is fresh in my mind +The patch adds the missing constructors which simply call the other constructorsUnit tests arent really possible for this +This way you can see which version of Ivy you useIll leave this issue open a bit longer +for the last tests failures shown above are not related to this patch +I polished it a bit and renamed the option etc. +That is unexpected as UIMA shouldve been deployed with +Currently we use a hacked to provide context sensitive lookups but with this service we could simply plug in our custom versionIm not sure is the right name but cant think of something better at the moment +Thanks John +Updated in revision . +Fixed +Closing old resolved issues +So should I create another issue for you or you will fix it within this one +Cannot reproduce the issue in my environment +If one bookie dies the client continues to write data to the remaining bookies +still unable to duplicate in my environment +It looks like what happens occasionally is the previous test will not terminate but instead spin on one CPU indefinitely causing the following test to return unexpected results +As well I spotted an issue in the disconnect logic not setting the session instance to null. +Once is done then you can expect to have this functionality in the CLI shortly afterwards +Thanks Phil for the contributionCommitted revision +Improvements Should store on disk with three letter extension Should maintain original aspect ratio when resizing images for thumbnails Should maintain original aspect ratio when user sets image width and height When Attaching uploaded file should initially show search results on type image After editing a media file should update title if it changed Insert button doesnt really stand out wasnt clear how this page worked at first Like rest of Rollerc new UI needs some work to make it aesthetically pleasingOne suggested RFE better bandwidth and UI via true thumbnails Store real thumbnails instead of resizing image in browser Create and store small medium and large thumbnails on disk for each media file Small for view media page +Nov st and nd I am out of town +All the loader methods are passed a job object +Ive just committed this to trunk andThanks a lot for the contribution Marcelo and thanks to Todd Andy Daryn and Suresh for the reviews. +Could you possibly give more details about your setup? Do you have a modulewhich has a name matching the first part of the aciton youre redirecting toPlease examine bugid to see if this is a duplicate of that bug +I didnt know which Component this issue relates to please change it into the proper one +Setting normal urgency +Julien I think you forgot to include the new files that need to be created +Patch attached turns out x doesnt equal xF +It seems to me this has been fixed already +turbine and stratum ones are fixed +Gehard due to IP issues its best that you provide the changes as a patch to this jira +Thanks Junping for reviewing the patchI have committed this. +Reducing forkCount would still leave each JVM with the same heap and if its not sufficient it would still failI can however imagine that the current setting is a little tight and when the machine slows due to swapping it makes the GC give up due to excessive time in collection earlier +Thank you Sanjay for the review. +Patch from a baseline for making stream support generic enough to support SSL on a HTTP transport receiver +Nothing to do verify +Looks great +Patch has been Venkats second patch as well to fix invalid security annotation in +Unit tests needed +Maybe another one for you Chris? I wont get to it this week +Jakarta ORO SVN repo has been made readonly +Ok that was just a tought +I find it a bit troubling that must be specified explicit as the map and job output types instead of just about we add a job argument to set whether to use orLooks like we will need an step in Synthetic Control now to convert back to form +Patch for the task on comments on comments +Argh logging examples again +Paulex It is ok +I understand the problem and have fixed it +Phew! I was finally able to completely understand and then solve the underlying fundamental problem and unit test now passes. +I totally forgot about that and sounds perfectly reasonable to removed this capability in +So you are not seeing the Run as script item at all? If you make a very simple script do you see it or is it missingI am running OSX and I am able to run scripts from the package explore and from the editor +Does it? Awesome! Ill get going with this asap +Getting access to your environment helped a lot +Preliminary result of a history collection patch seems to indicate that the container was closed due to an compress prior to the shutdowns checkpointing +You need to setup the Ant proxy properties using the Maven proxy settingsThe attached patch file solves this issue +The best way is to make the plugin available on plugin central so thatpeople who needs the plugin can download it from there +The has been updated to test accessing an ejb requiring a identity from its init method. +I can reproduce this +This is in for +But is there baggage in the current and formats that you dont need or want? Why not improve those rather than duplicate functionality +Lookup the Package API +The linux font code was failing for me after the commit for this jira so I made some minor modifications at r. +Username user +Relates to as we need to set the location via a to as we want to be able to create several projects in the workspace from the same issue depends on and cant be fixed until that issue is fixed +patch looks I dont see any objections I will commit the patch tomorrow moring +Thus inserting and deleting strings wont affect the balancing of the tree +tests included +utils as of r +Yes that sounds much better +Patch v corrected grammar in addressing Teds comments +Cordova will still run on older phones with a high target +So I have modified the patch +Fixing the build as an interim step toward merging RPC is fine +Using bamboo now. +Ive updated with the missing file +You need to recompile the stylesheet when you use the new XSLTC. +Current patch +v w the comments taken into account and rebased on trunk +HiSorry no progress +These two features link up well +Thanks Dhruba. +With the non standard namespace Wicket confuses this BPL for a non wicket tag and so applies that filter which causes the damage to the href when it attempts to make an already relative link relative +Please give it a try +of course sorry I suggest we dont change the grammar and let the compiler fail in APP if something returns a which should not take one +committed as r +simple patch for removing the plugin from being mapped to PDF content type in +Again be sure to see for caching as part of +Mice Have you added this functionality can you look at this ticket +Shouldnt these requests be sent to the queue so the correct instance of the service is stopped +I have not seen any other suggestions how about the following zero admin solution? It is not perfect suggestions welcomeAlong with the statistics storing save how many rows were in the table when exact statistics were calculated +etcI understand this is not the ideal solution but at least a plan to meet the goals of all the interested partiesThanks for listening +Conf also updated to reflect changes +Letting some time pass and rereading the documentation again I understand most of the mistakes I madeAs precaches all appenders it will not read another appender simply because I implemented it according to the tutorial +This is a comment +Tom suggests that we need to review any use of svn externals here to make sure theyre updated correctly +We are giong to have it done on monday Oct have almost completed list of components +Thanks for the patch. +I will try again with a clean workspace and let you know update site should be back again +thats a specific case we have to keep trying until the server is have already improved this +Fixed but not committed waiting for clearance +We could probably add a comment into the section referencing the doc section that talks about the tomcat connector problem locally for future versions +Patch applied with additional exception handling thanks Javier +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +This appears to be a duplicate of which you also entered +Patch to fix Subversion in RC +java +Sorry please ignore this item +btw +Bug fixed at revision for the bug fix worked well +making fields static is OK if they are also final otherwise it makes the class +The sample code in question has been updated significantly since this bug was first filed +Always seems to be the case when sender receiver! Probably a deadlock in the same is also an incorrect receive count at the end of the tet run in +Thank you to Chuan for contributing this fix. +I found what make it to change the behaviourif the is textplain the decoding is not done if the is application the decoding is doneDo you see that too? Im investigating a bit more in the mean time +Same for drop link error +Bulk closing stale resolved issues +If you find the fix working then please provide detailed instructions how I can verify it thanks +this will now produce a compile time error +OK make a call and do it +I cant see how it would matter whether it produced random key X vs random key X for any given record +Patch resyncd with trunk +Thanks for applying! +Committed to trunk Thank you David for the fantastic work here +I just committed this +Yeah I thought about suggesting it as a dev dependency but isnt everything in a dev dependency? I guess the argument could be made to separate the dependencies into two groups those required to build fauxton and those aimed at assisting development +It should be x x Sparc +Notice the shaded circles in the images +but yeah get it in +No testing has been done for it though +Unless anyone objects or proposes something better I will commit this first thing tomorrow morning +There are no immediate plans to change this in the series either hence Im closing this +Okie. +Thanks Vasily +And I would be happy to be able to distinct beans I explicitly obtain from context in Java code just by looking at their id in XML +More like a classloading issue in the security subsystem +test for applying and release? comitted this weeks updated the pom slightly to remove ambiguity of version which was causing trouble but otherwise this looks good + +The alternative to this I guess would be periodically waking up and checking that the nodes we are waiting on are still alive +Otherwise they can both be whatever the user or application chooses +Statistics schema cannot support replication since no valid primary keys can be established for these raw logging tables +Does the mojo project have any way of receiving notification when the current snapshot gets promoted to release status? There doesnt seem to be an annoucements mailing list +The jbpm db is not hidden beneath an api +This is by design +I agree that it should be its own eventuallyIt is a unique feature + +I have just removed the check in commit befacfabeabaaaeceefade +Committed. +I have posted the patch only as a help for researching the cause of the problem +Boom +I dont understand +patch is resulting in ton of failures in my test runs +solutions +This is definitely a bug in +Thanks a lot +Corrected inheritance hierarchy +The patch looks innocuous enough +Hence if there is an XML block size the MR will execute multiple mappers but in all the mappers the loaders will load the entire XML file +Lets see +Attached a patch that checks to see if there are both Interceptors Decorators and if so creates a new Interceptor and associated to add to the Interceptor stack to provide the entry point into the DecoratorsIm attaching a patch for now so this doesnt affect the M release +Please run using grails +Thanks for the patch +The issue was configuration +The only other character that I think you might want to add would be a not and we still have! and to fall back on for thisId like to ask that this be submitted for unless someone else is willing to implement the alternate single quote solution +Your issue apppears to be a issue that I cant reproduce on on the Galaxy S +Besides that I havent investigated it more preciselyI started last week to add the plugin bundles of an Indigo installation to my local maven repository in order to reach a state where at least I can build the project on my local machineI think we will need some special solution or workaround if not for the normal eclipse plugin dependency but then for the DLTK dependency +No fixed date as usualThere have been about three months between and but almost a year between and +This is how policytool looked like on July +When exactly? Do you have an issue +Dont stop for me +ATM thanks for your feedbackfindingsAF version of scriptpatch fixes the issues you run intoWhen testing make sure to do a mvn install from trunk before running mvn from trunk as there are changes in the other modules as well +It does redefine what means in fact it replaces it with and application ID +Screenshot showing no option to add the datasource to the server modules in the jboss server I cant reproduce either +I think this will be relatively painless it just needs to be done +is years old and not maintained anymore. +This is very urgent for my project +Deleted +It is possible today to do what you want but you need to provide a custom fieldbridge +Just one last thingDan moving the server code to the test src tree wont be burying it at all its just attempting to keep the boundaries of the project clearer +There is regular ee test failure introduced by the patch +Same would go for XML +Patch containing update to in srcmainresources first version of patch +the fillMSG method seems to have subjectobject position inverted in filter queries +I have attached the whole project not just the test +Bringing back in give it another review before punting +started work on already so Id prefer to commit that before opening this up so there are no merge conflicts +This shouldnt be exposed in nodetool though because it can easily be dangerous +Thanks for the patch! I dont know why we forgot to include the in our build +committed +Maybe you can educate me in the correct protocol +This issue was already resolved by other various issues. +Wendy I have overlooked the fix for version my mistake +Patch to address reported issues including missing license in sql init script and conditional unit tests that wont fail if theres no jdbc driver +It passed times with the patch +Thanks for verifying the fix +The configuration override was actually a side effect of ratherthan and that has been removed +Closing issues solved in released versions +batch transition to closed remaining resolved bugs +It has simply got default settings and some advanced options that need to be activated ve just noticed you try to specify a placeholder on the itself! This of course does not give the a chance to process the placeholder since it needs to be instantiated first before processing placeholders in other bean definitions +You could also use javadoc to guide in what exception to throw +Candidate for SOA eds kitting step wasnt retaining using the filemode option to retain the file attributes +Run the following in this foldermvn clean installmvn archetypecrawl +Same on +Fixed at the level r + run Failures Errors Skipped Time elapsed +Closed without integration tests. +Why dont you just disable caching on your instance of +I look forward to seeing your patch +This error no longer appears +What a stupid name I chose for that object +I made a copy of the original message rather than modifying it +Committed to trunk. +I am actually not sure whether the problem is with or as Patrice seems to be using lazy loading + test classes that use old stuff for testing +do report the number of failures and it should be straightforward to maintain and report the number of successes as well +public fields in the bean should work as expected +Here is a patch that may work i did not test because i dont have any test case +Yep I also have a few Collections where it works perfectly +Add rendertrue to the tinputCalendar tag. +Im ok with this +investigating +Thanks for the review Stack. +All tests pass +fix test minor cp Im working on that right now actually in +you have to use the option nodetach to get the error code of jsvc is notstarted as a deamon in this case +Filho filhocf at gmail dot com +Verified on CR. +However that is still an issue and really needs to be fixed + +Im rejecting this feature request in favor of a fix for the original bug. +Im guessing this issue no longer valid +Please verify +I ran the tests and it seems theres nothing brokencollections was refactored between and and many classes were moved to new packages +Full log with consecutive thread dumps +There is no error running the original set of two migrations with of attached patch for and Adds a test for creating a new column with a NOT NULL constraint with no default value Fixes the to pass this also ran the test for mysql and they still all pass are changes to these files M test M test M libjdbcadapter is a new file A testmodels failed for hsqldb +So the solution as I see it is that the value must be truncated to characters and a warning provided +Resolve attempts are very very slow especially when they failWith later versions of and we are now caching unresolved classes and this situation should be much slowerIs this still a problem for you? If so there may be something else going on +Best effort has been done all except for deployBamboo gets Authorization failed Not authorized. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +We are not doing local reads to the log files here and Sanjays proposal does not help in this case +The URL to the test case that reproduces this issue doesnt seem to work anymore +This is a latent bug then I dont know didnt write this code +applied in revision Vladimir please confirm that it resolved the thanks +The latest patch looks good to me +The build instructions can be found in the in Roo master. +Added common code to close and shutdown mini dfs +pom kept. +Im going to close the task and as new types are identified each can have a new jira issue created. +I have updated trivial patch to fix this +Marking bugs with invalid fixed for release as fixed in. + didnt catch it in time but you should use the Apache repo +FYI there were no problems on Winx +The other part is in the AS integration that should exclude decorator classes from scanning which I will do once my patch makes it into AS +correct rendered image with image using issue has been resolved for one month or more without complaints +Please see attached files for the fix +This will also allow inclusion into the long as the patch can be done in a form which allows inclusion into the CP Im fine with producing a temporary +Thanks Connor! Yeah I havent tried that +Will likely file a POI bug and a new TIKA bug to accomplish the equivalent for +That is OK with me but the getKernel function has a bug +Mrinal commented comma delimitation allows for a very easy and intuitive way to have Tags with multiple words as follows +rsync rtlzv delete homedownlodpublichtmlapachereceiving file list +not being able to open the fileMaybe you didnt use a supported file system +for clarification what does this issue actually fixed +In trunk as of for a while +Committed to trunk with revision Committed to trunk with revision Marking issue as fixed since I do not plan to make more changesWill close the issue in a few days. +Attaching patch somehow it got skipped earlier +I think i have a immediate need for this +for the updated patch +Web page now display version similar to hadoop version version from by tgraves source checksum beaeeedbddddcbHadoop version from by tgraves source checksum feafeeebfffeed +Just sketching out an idea here +Isnt this inefficient and? has no copy constructor +Working on converting over the table znodes under +This will be resolved when we implement the CSS in to Bugzilla +I added a +pc files are packaged uniformly across all the distros for this solution though +It shows a short summary of what it wants and then you have three buttonsPreview Ok Cancel +hbase turns the tables in to mode +Done for the JDK version +Committed to trunk. +Is this a conflict with the maven artifact plugin? It takes the type of the artifact and deploys to the relative directory type s if the type is source it therefore deploys to srcs +Awaiting merge of pull request +Scott Ill include your change in the website update Im working onColmScotts suggested change has been applied +Does it mean that it will be a Lucene contrib +Also it handles the case where some racks are smaller than the others by giving the smaller rack lesser regions during the round +Not sure how that happened +The reason I was marking this as a blocker is because it exposes a new public API that allows an external process to monitor the memory consumption on the JT +Small patch. +its entirely possible the delta is noise +Arpit thanks for catching that + duplicates the showing that only one of the will pass this is because only one the two opting servlets is active +Patch to OGNL to distinguish situation when property source is null +Updated the patch +Will look into it tonight +Included parser and sequencer tests +It has been observed with both and + you for the contribution Navis! I have committed this to trunk! +Whilst the functionality wont change the API will most likely. +These are great suggestionsI am looking at the possibility of removing and +say db version is and latest release has all fixesAlso to understand the range of your changes do you plan on backporting the downgrade changes or will these just be in software and aboveI understand it helps with the tests +trycath added to to log exceptions and rethrow them +Nick Ive committed the patch as isIf and build arent relevant anymore could we not just remove it thenAbout wsfeatures then this feature have been there for a while +can you add a comment as to what is going onalso adding a method for the test to use would be cleaner than poking through +Please put it on the schedule +Verified on and +This is already part of you could give it a try if you checkout the trunk +Oracle only supports double measures so a single array of ordinates and measures could be usedBut first we need this to work and then refactorI also note that the SDO test should be rescued from the unsupported oracle spatial module +Ah I forgot to add a unit test +Wed consider adding it to or also +The UI aspect also needs to be fixed +The attached patch takes the previous ones adds the following the runtime behavior must be enabled with true is minimally serializable to provide a means for testing the threadlocal runtime logic a unit test is provided for the working serialization cases it must be run with threadlocal runtime enabledIn making this patch and the associated test a few problems with the original patches came to light +Just headers +Oops forgot to say that I will attach a patch for review later on today +It is due to a change to the setup of the at the scanner level causing the specific file protocol handler to no be used and the fact that suns file protocol handler does not implement lasModified correctly +After fixing this asserts still fail +Link to the bug report +great we did no work at all on current snapshot is and next release will be. +Revision . +Patch addressed some review feedback +Isnt there some hook code in our java that isnt necessary w this change +It also includes some regression tests for and mode +Assigning to Jim for commentAlso Ill be on vacation for weeks +For scripts please can you also take a look atThanks +If you can use the snapshot until then this would be great +Ive attached the dev studio solution containing the sample application +I tried searching for a new ticket and looked in the mailing list for any discussion but I havent found anything +Correctionmodule one is a jar artifact module a war artifact +Completed for JBDS and JBT for now. +Closing. +Fixed in the SIXPT branch by adding an for GAC dependencies +The downstream user needs to intentionally enable Logback use +This will allow us to queue up an entire files worth of edits guaranteed +looking at Fuhweis attached test can not apply the attached patch +I am already working on the issue +I will rebase this +This is the same behavior as the reference implementation as well. +Ive posted questions to the mailing list for further discussion +I think I have been able to accomplish that in a number of ways Turning off Flow Control Increasing my prefetch size and Using a pending queue +Looks good to me +Ill work on a better fix shortly +I dont have any issues with that myself +the testcase works t we have a way to disable arbitary replication maybe as a default +Please verify +Doing so from the server in Germany download time was seconds +This was reported against an old version and the reporter did not respond to a request for more information after several weeks. +Cannot reproduce as described on Fedora +sorry for the change of person but I often forget that forums is only a portlet not a portal! Ill try in theese days to correct it with the latest version +This makes analysis and can be checked for correctness by the compiler +Actually the repo is stored in the directory where you start your AS from +Verified that fix is commited in trunk so it will be fixed in next version. +I also do not think the strict mode is useful but at the very least one could have an option of making sure that a lone CR would cause a since this problem happened to be such a major point of contention +Hi ChrisCould you test the latest trunk on this Solaris machine +IMO anybody can store whatever the state data needed through out the life of the service in theAre there any special cases where we cant do that +Youve also reformatted the code which causes lots of unnecessary should be the files. +Yeah options are not yet supportedWe have added a hack to support binary for the certificat attribute but tahts it atmWe know how to implement them but as we were working on a complete rewrite of the internal of ADS we still did had time to implement options handling +Sample and are now functional again +The patch also moves the third invalidation on compresstruncate to a point where the new conglomerate information has been written to the system tables so that the statements get recompiled using the new conglomeratesAll the regression tests ran cleanly with the patch +I also threw in the stuff that requires Java that had been commented out for a while to use the getCause method on Throwable so that may help certain things too +See my previous comment +Could someone please provide more information about the Cause Consequence Fix and Result of this JIRA for the release notes +At least I do not know any other way +Need to wait for with Apr integration +If you specify explicit separate on the fotable then the behavior will at least be identical to what produced +Thanks for cleaning that upsome comments you should still cleanup the output dir After the I took your comments into account +The patch from the initial description +I still see the problem on our version +See for example +I support this change its a useful feature that is clearly documented and no threat to the Comparable contract +Please verify and close. +The patch is worth committing because is a valid variant +Fixed in a better solution be to use the ifdefs for headers and methods like weve done elsewhere in the code? +Thanks Vinay. +perhaps this guy is a good candidate for a +The current complexity is already too big +There doesnt appear to be any alternative to this at this point +Stems from the same usecase as +I should have created something like a dependency on that JIRA +Unfortunately this is all I get in my logger +spec Server Thrift should serve inside a threadrspec +Changes in the addressing and kernel modulesmodulesaddressingAdd a new parameter to allow the to be configured to ignore headersmodulekernelAdd new apis to manipulate WSDL metatdata contained in an EPR +Hi Chris thanks for the hint about Configuration +This one returns Collection instead of List internally which will allow me to use +How class demands creation ofIm interesting do we really need this staff in production +Whatever +This should be made a blocker huh Gary +Im not sure this is really a bug +Please review +Sorry siswas. +All tests pass with +ChristianI have applied the patches from yesterday +I think the code fix isnt necessary. +A patch that adds most of the missing physical operators +I doubt I managed to cover all cases but I got the ones that affect us +Thanks for pointing this out +But since both seems to work fine I guess its a that can be closed +Should work ok on top of many thanks for the response +This seems same asThe latest code passed all my test casesThanks +Also if it gets killed it is likely that itself is the rough task because it uses the highest amount of memoryBut in general I agree with you +I suspect no requests will go to the +This could be more than one machine +Attaching two patches +Do you want to set ALL base urls including? +Manual Installation of Snowdrop Added the to the list +Documentation was actually updated and there is agreement that we will open tickets for new features so Im going to resolve this ticket +Ondrej are you seeing this against EAP or community AS +Committed toThe available patch applies to but not to so marking fixed in +Steven I follow your think that the line number is useful but not mandatory +Will run tests and commit +Finished in r +Ignoring the provider when it returns DIRECT partly defeats the whole purpose of the provider in the first place +I see this should be documented for customersI think you can close this issue documentation should be tracked under based on your todays comment there +Im going to start closing these left open as reminders to include in +Since I am intensively using SI I found another where an Aggregator doesnt get processed at all +For meta parameter we can keep its name since it is for +This issue has been resolved for over a year with no further movement +given that AS does not provide nested properties inside properties this fix should be done to provider migration +looks good +AFAIK inserting a tile from one module into another module is not a supported use case +We should now make it input again this also conforms to s design +But if you do this then you have to actually use the Class instances as keys which implies solving the hashCodeequals problemSolving the hashcodeequals problem? You mean your concern about these operations being slow? At least this caching is faster than repeating the calls over and over againIt also seems that the require at least one HARD reference so maybe your suggestion of WEAKHARD is something to look at +It sounds like this issue needs to be raised with the Servlet Spec to extend the Exception handling functionality to the entire container not just for Servlets Filters and certain Listeners +There should be no difference between remote and managed they both use the same code to communicate with the server +mmm should I also fix the title since we didnt change the timeout? +Complete with rev ffafffeefbafeacffa +I think I even tried with to confirm the problem. +I think this is ready to commit to trunk after the release of course +am attaching the patch without formattingRegardsNuwan +I think this issue is solved +Thanks Rob. +yes i just refred to bucket cache with mode enabled +This was purely a bubbling issue that was resolved. +Could we rename this issue to indicate specifically which classes in collectionsjdkbranch these changes genericise +I dont even have much to complain about of themIm not sure how to implement them as structs except perhaps via a code transformation +Thanks Joe and David for your responseJoe if you look at the Global context after uninstalling the connector you will see the connection factory name is still thereIll digg more into this +Symptoms are more global than first thought so changing the title and description to represent that +jgroupsclustering is only included in the all you want it in the default config then you have to do it work around is the solution its not a bug +since its in the Spring jar file can we rely on containers to pick it up directly from there and avoid this problem in future. +The javadocs zip has been created +Can we upgrade the priority of this issue? It breaks core functionality +This is the associated patch +change the fix version from +Accepting this as a task +Retargeting CR +I agree +Use a debugger +Thanks a lot Chris for your valuable response +Great thanks Wes +Facelets current triggers for any change of the currentDate is th January we change on th January +The trace is thereAbout the phone number remember that +There are several extension functions defined out there and with other processors exsl works fine +Jasha Ive changed the code so that the is now an enum instead of an entity +if you have some time to do it itd be nice if you could build against head and tell me if it really solves the issue +Should do the same thing that was done in which is to add the file from the templates dir to srcmainconf +But what should the semantics be for an application that uses a component that uses that doesnt have a root appender configured? When the component goes to log it will get an unconfigured category and will generate errors +missing import for have line line currently contains all items from that patch +Verified that overridden message bundles now appear correctly with the message bundle declared in the current pageflow being used instead of the one inherited from another controller. +This was done in the past for API consistency +Hi +This is a first go on moving to maven + +you just need to set +Lets see if it cures the failing in. +The will show how you can specify alternate bundles for the validatorin the same format posted by Chris Johnson +Thanks +Now moving from to +Gone stale was probably fixed elsewhere. +deshoputil is in classes +For example users must be told that hdfs file checksum can be different even if the contents of files are identical due to use of different block sizes and checksum parameters +bq I was wondering why none of the other methods which use are not synchronized as well +Thanks for the review Sean! So should I still along the lines of the most recent patch then? Responses below Okay that makes setting up chkconfig even easier then! I figure rest and thrift services can be added as a case and set to the same chkconfig I am pretty confident they can just be brought down at the same time before and Master shutdown + +Hi AdamHave you a suggestion? Maybe we could follow Karim idea but then we would have to create an DB upgrade path +Suppose there are nodes A and B +junit tests did not run corrected in this patch now including segments for quick create acct transaction and showing segments at the transactions detailsImproved and better of codeThis patch still for ofbiz R +Hi Sriharsha can you please attach any relevant error messages and logs +I guess I relied too much on my IDEs support to add static imports which didnt work as expected +retrying hudson +bq +The question is what would an example STOMP frame look like with type info in it +Hi Ryan +have submitted A patch wich must be applied to the file for Xerces +Got plenty of work to do on this ticket still. +Slava please look at it +Is it going to be a common use case to give ORDER BY bindings and MINUS say ? How would you know that even made sense? Separate symbols means a user now has to configure up to different options for all of which he will have very little context or knowledge of what to set it to +Wrong fix for version. +This is reported against Continuum is the report file mentioned above +Feel free to commit it if you havent done so already +Moving to I cannot explain how state has OPENING for the node but the znode content is MZKREGIONOFFLINE without more context +I have to honestly say that Im not recalling this issue so I would have to take a look at the code once again. +It would still be nice if the maven extension has it as a feature . +Here it is against branch +dims +So I was happy with it and thought youd find it handy either way Id just like to see a maven plugin that works since I appreciate your effort very much +oops put in correct PMC! sorry for delay +Rerun test to see if test failure a transient +Can a committer please commit this? Thanks +Producing to the temporary destination to test for its existence is not working +The wizard will also not add your existing key the error will not occurrThe error only appears if you kill your key and then create a new one in the wizard +That makes it easier for committers to review your work + for patch +It should look like a simple recipe for the very common +Ive just committed this into two unreleased branches in trunk +This seems to be fixed in available build for GEP +The fix propesed here for is not right. +I would suggest that a company treats what it buys entirely separately from it sells +And it should be showing whatHave you configured confluenza +One of the issues I am stuck on is that there are some third party jars that we need to use and I would like to be able to use something like like in maven +Moving back to since Ive removed the backup files in the CR docs zip +Its in my previous comments +Um +Upside is potentially greater speed if its implemented in native code and no new Solr dependency but the downside is more config hassle for the user +Your contributions to the Apache community are greatly appreciatedI am running tests on the patch and will commit the changes shortly. +Henk whats the chance to get a few more patches +Can you please check with the latest build and let me know? ThanksRegards as not reproducible +yes now I have. +Then it just stopped +Do you remember how many extra threads were introducedThis question is answered above in my last review +Can this be addressed Can you tell me the repro steps for this? or paste the logs +For the sake of completeness and accurate statistics could you change the status of this to fixed? Thanks +Committed to trunk +Can you elaborate a bit more about your special use case? We had until now no need for the in a since those converters are typically used to convert a dedicated type so theres no need for providing this info again +Brilliant +I just pushed the test fix +it is good to add a conf for this and by default disable it +The constructors of changed with the addition of a new argument +Removing exclusions since they are no longer committed the patch to trunk and +Rahul Looks like a dependency in the featuresassembly pom was missed as part of removing the for + true a timg component with a src parameter that defaults to binding prefix context may be just as effective and less intrusiveI like capturing these provocative ideas and the discussions in JIRA. +Added docs +Thanks +Setting to +Will pull the recent patchs from svn and retry +Given that the responsibility has shifted to Oracle all bets are off +unzip the patch to NIO module +Removed making lengthSquared instance variable to transient +I recommend we think about it filing another JIRA for the same after this goes in +The Jenkins job which compiles our nightly builds was down for a little less than a monthIt explains why you got the same error the nightly build is currently almost one month oldIll fix that +ApplicationsContentsDeveloper +If the user does not check that option and his domain has applications the operation will fail and an error dialog will appear +Werner you do realize you have access to modifying and checking in those files right +Hope there will be no problem with other JDBC drivers +that has the Jack Rabbit fix? +Not sure about the other +Changing the compilation level of the project would require a full publish +Most of these seem better now but Ill leave it for RC so you can a few of them before resolving it +Yeah thats a good idea I will implement the changes and attach a new version +Luke there must be something wrong with this commit +If you want to share your current work progress with others without disturbingthe main code branch you could store it in a separate CVS branch +Ive filed to get it upgraded +It sounds like is going to be the most appropriate and least invasive workaround for AS +Will incorporate their changes instead of the original patch +No I dont think I looked at this yet. +All caching and everything is still done per segment +Please discuss this on the forum +We need a help from your team to investigate whats main reason maybe from product configurationI know its better if there is a testcase or similar but I cant so I think its good if your team can help us inspect the need to know that the JCR Team doesnt know well PLF so we will waste a lot of time to configure the environment in order to debug it that is why you need to be sure that it is JCR relatedDid you check the content of the data given to the JCR? If so doesnt it contain this character +This probably needs to be copied from the main webapp configuration +I get that for a first cut its easier to stick with the existing disk quota system +The Sun does have an async thread that will tests the connection and reconnect if necessary however if it fails to recover the connection the thread will stop and never start again +Basically to be able to run the tests outside the home directory we need to add back the link to the test root directory and not just the home directory +I filed to update the tests +It has a lot of kit for maybe the first of the code to set up and run a job from the command line +An addendum that makes the successfailure message explicit and more informative +Can someone take care of the injection +is there any code for this or the available? I like the idea of just having a log to write and read from +screen Patch submitted and an added GUI functionality for adding a keyfile in in procedures listed are you referring to? How can the original problem be reproduced? +Tested in +StackOne more thing is was backported to as part of for explanation Ram +Tests passed cleanly for me on except for the heisenbug in +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Thanks for patch Kay Kay. +The retries are visible now and dont take place quite so quickly is the only differenceI will take some time maybe this weekend to look through Solr servlet code to find out where this happens and propose a patch +Hi Mikecan we add some additional parameters to the SCM URL? Just the or something elsexyzurlmyNameskylab because the common case is that the SCM URL is in the POM and therefore cant be user or +Please always say what version of JBM you are using how you installed it etc above is not enough to be able to recreate the problem +Normally I have someone do the proof reading since my english is terribleand Yes I can go back days after and see some of the errorsHowever I cant at this timeI guess iF my english is in the way it will not get submittedThis is not irogants just frustrattion on something I have faced most of my lifeI envy people that can write good grammarAnd Courses will not help It is how my brain is wired +I checked the patch of the related MR jira +Lars you missed the root pom +Attaching patch the patch +We need to logstore the bad records some place so we can viewdebug them and make a more robust loadfunc etl process etc +Im attaching the output log +this to trunk and + case checks that page is recovered when the user clicks on an m having a look at this seems to be related to page if I change the code returning false I got the same url when using or +She will working to making appropriate profile +Fixed in revision after release +Would have it run if it was under? If so we should move it +After enumerating the active set of containers the AM could make its own decisions about how many containers it needs and where +Partially done the parent has been donebut the modules still need to inherit should be possible with aggregate with release of javadoc plugin and matthias upgrading to it +Whats the relevance to New Quickstarts +However a new issue needs to be opened because the new javadoc needs to be wired into the documentation. +Thank you +Did any of these patches get committed btw +closedtopicsessiontestclosed also fails because of this problem +The optimization has been isolated in static utility methods +Sounds reasonable go ahead on the after a second look looks like its actually building IT which is simply mvn V clean verify with maven +blackdrag Should this work +Ive started up a little package under named +In some cases the actionId will be null +Its all in your hands now + +Either quiet looses some meaning or strict looses some meaning +Thanks for investigating this Narayanan +Hello PranayI am uploading patch for sorting existing billing account on accounting componentSteps for testing Select a partyId from party component for which a billing account is exist +I think this fixed by part +was caued by using +Too much hassle for too little benefit. +I filed an issue for as well +on patch v +will commit +The information has been updated. +Thanks for the patch +Will fix it fixcrlf task corrupted the +This morning my screen was black and I had to reboot +Once they are in secure storage you may need to connect with unchecked remember password to have password actually removed from secure storage +Fixed in svn +The last reply over months ago says +But beyond that I dont see a need to restrict our actions here. +Closing this issue as not reproducible +Thanks Thomas +modified the location of the upgrade file + to shared DSML parser is now part of shared +You could make your bean lazy as a is a timing issue the GORM metaclass methods are wired up after the application context is initialized so your workaround is the way to implement this. +My application is a application and is invoking the load metadata during a bundle installIf the schema is lax then there wont be too much of a performance hit further this feature should be made optional so client can determine if the hit is accessible in the clients context + +Mamta do you have the test directory of a failing run? If so might there be some clues in there as to the reason for the failure like something in the? Otherwise could we add more debug information in the event of a failure + +Some of the uiLabels are not defined in a standard way +Closed per request +Waiting for RN to verify text +flow control +Yes I can confirm that +Theres no test case since this is just like the other existing catch clauses and any test case would just be testing whether catch clauses indeed work +to be more precise +Closing resolved issues. +Thanks +The bug comes from the same part of the code +The attached proposal is for what I would like to see happen to the AM UI web proxy and job history server +If we have implemented then this wont occur +So obviously when doing refresh operation RDS servlet expects definition in but because of using spring there is no any definition of bean in file +is the patch for +Will commit this momentarily. +Additionally I would recommend allowing the user to choose or create a in the web root folder as a target for the scaffold +Committed patcht for version +Erik can you provide some help to figure this out so we know what needs to be fixed +will fix this +need juddi bundles +I updated the patch and am hoping this operation things JIRA wise ie runs tests and email Doug + tested! +When the broker host goes down and the client is waiting for a response from the server I think the client will have to wait until the timeout +Ill post the fix soon +Just committed this +There are several problems with the schemas and how it is used in the wsdl +Previous patch was missing some changes due to a bad diff +As far as I remember you can always set the frag within the Castor mapping file +It makes more sense to do this in a try +I tried to apply the patch but it was not created with so it wont apply properly +Verified in revised release of June . +I have not looked at the actual test failures yet but are the failures simplified at all away from having to wade through various dependent libraries? If it is possible to reduce the test cases it will be much easier for us to fix the issues +Thanks for noticing them Knut +Does this by looking at modification time +done the property has been added to trunkIve a remaining of question though does the build to every build? we arent getting themthanksjuan pablo +Should we try to use here or we should write our new datastructure +What do you thinkOr lets say we still make this system property available because some users might want it +I am getting enough pushback on this refactoring that I believe we should not commit this patch +However during the admin api based host addition it will not add all the necessary information +We both working on it +Thanks for review +thanks for pointing it out chrisI have created a new jira to fix the problem +This is fixed +Here are the reasons why there is no source packageFor debian software are usually setup Write a control file and prepost installremoveupgrade scripts Patch source code Build source code Create deb package from binary Create source deb package with source file list and binary deb packageFor rpm software are usually setup Write a spec file Prepare source code and patches Build from patched source Discover dependency Generate source rpm and binary rpmIn both systems the meta data packaging file are created outside of the scope of source tarball and patches can be applied to create patch level packages +Thanks Eranda! I will have a look at the patch as soon as possible +Should wait for teamcity to verify +You are right only one parameter would be needed for both functions +as far as i know one should not make a dependency between wars +It is possible but not necessarily desirable to have a separate way of detecting differently from value +A number of issues related to greater descriptor flexibility will be linked shortly and Ill issue a patch that addresses them all +Any update +log skipping bytes message at DEBUG rather than WARN +How about this patch +Agreed I wanted to specify this in terms of qualifiers but couldnt due to how arrays work today with qualifiers +if so cant we just start it at instead and be ok +Oops +tail of log behaviour appears in JBDS CR too +Fixed +Sandy trunk builds are all failing because of this +The first which is less popular among serious devs is to just switch manually when you deploy your changes into production +r +trunk is still an issue. +There wont be a release +Dima commit his changes he set table column names in upper case so I ask the point is to discuss is upper case names right decision? why not lower case names? or camel case names +and to separate the logical pieces +the only change is if the transaction is rolled back not when the schema is created? The Rationale for change is more describing what is implemented rather than the justification for changing it +The structure mirrors that of in that there is a common base class for benchmarks +If we run out of a bag return EOP instead of NULL +Verified on EAP CP CR. +Committed +A few questions Is it always the rd superstep? It looks like the task lost its connection to the service +Thanks Maja +This is the response any ideas on why the array takes up so much spaceIs this something we can actually fix without breaking compatibility +Updated Patch will commit in revision cannot show the node tree as the previous says Fetch zookeeper datawithout showing anything of the solr node +Can you take a look at them +Just tested with the latest snapshot and it was fine. +If we start posting patches on top of the existing patches well have a mess +We may in a future make paging to act more like a database and this would then be moot +Canceling Patch doesnt sound good +Can you please test the feature again and make sure that you have the correct namespace declaration in your file +This patch fixes the build and works with both Maven and beta +No the issue is that the current designer does not match our purpose of was always to fire on a transitionevent and never as part of the main Node +The in this bundle is missing some of the required info +Should be fixed now. +If we create an interface for I think it should be ok +Do you still see this with the latest version +No +Because currently I +Robin I understand the philosophy of mahout +Can you suggest what needs to be done to have is to return true with Lithuanian languageIm not sure Im not super familiar with how Tikas language ID works +So can you please attach the new patchThanksI think you may have been confused by the multiple filesI the patches +I think this patch will do the trick. +You know +As far as I know such nd level cache integration only works through JTA +applied the fix for the file +Javadocs are something that should always be simple to fix before checking in a patch +Simple patch removing bundle fragment +The stylesheet +If there are any Windows devs with Erlang experience out there help would definitely be appreciated on this one +No tests are required since its just changing the visibility of a methodIll commit this shortly unless there are any comments from anyone +Just a one line change that fixes the lastpolltime +I have made couple changes in this patch compared to the previous Both the patches require that user had minimum select privileges on all the tables in the select list +Received the latest schema from Michael Brauer and will update the sources hopefully later this day after the integration of and +bq +Looking at the code I think this exception will happen every time a bundle is uninstalled +config pathtochukwaconfigpath compressedfilelocation decompressedfileloc +Mayank Im not sure what you consider attributes? Are you referring just to the filtering part? The web ui and webservices print almost everything that is a part of the application report + +We have been thinking something along this line in the patch which can throttle HDFS readwrite and Hftp reads +I misread +batch transition resolvedfixed bugs to closedfixed +Fixed the call for with the inclusion ofThanks for the patch Jacky but a note I think this is not your real name but anyway I added it in the SVN commit logI think its better for you to use your real name so your help will be better rewardedThanks again. +trunk +Since the error is based on the value of lock timeout seems reasonableto throw the lock timeout error +Yes sorry I should have been more descriptiveId like to see the form be a POST as I think that is more accurate for operation being done +Updated the Linux build to include the based service overview and setup page needs to be updated with respect to the native C service wrapper. +Please post the logs of am having the same problem in a windows xp sp cvsnt cvs environment +Max to answer your question +rant startBarry I am totally with you on java generics +I think the main problem is in the bug not in the test itself +Submitting patch passes all unit tests locally +That analysis make sense? Can you figure it out with that information or is it necessary for me to build a repro case too +is still used +I also think this should be a configured responsibility of Maven +Is there an RC branch? Otherwise the issues metadata should be fixed in +Attachment has been added with description fo has been added with description good maintenance branch has been added with description failing trunk has been added with description newer fo document which uses foinline instead of fo seen arePage Unresolved idreference INDII foundPage Unresolved idreference FAMF foundPage Unresolved idreference INDII foundThis is on latest trunk code Fixed in FOP trunk +Code in is probably a better basis in any case +used by selfservice users who want users to be able to declare a queue but only get a or less queue sizepackage and class refer to QMF so you can scope ACL to any QMF methodWhat is OWNER used for in the Java brokerCarl +Yes that may be true when you build the skipping the building of the sandbox but not if you do a normal build +Okay I am on it I assume you can build your own local geoapi +Its very unlikely this has any impact on the error but its something to note +Closing all resolvedfixed issues of already released versions of Roller. +I also fixed Anita as wellAs for the warnings as long as they are documented in the Derby exception messages and SQL states section in the Ref Manual I dont think that we need to mention them here +I am ruwan +Need to remove qos and do as elliott says and just let client say what the priority is +Please confirm the current status of this issue and if its incorrect change it as required +Feel free to give it a tryJuergen +Will check it out +Please let me know if theres any case that Netty outperformed MINAAgain thank you for your contribution +Request object can have many parameters so using looks dirty +varName that should not be at all in +And also how patch fixes the issue +Updated the description as I have now established that the issues effects all versions of +So all issues not directly assigned to an AS release are being closed +Are you able to confirm that diagnosis in this case +details sent vis email will check in the morning how you got on +Attaching a new patch that addresses Xuefus review comments +Yes it is +There are queries which are united using UNION ALLThis happened several times with different from that big query and the query i valid +We havent noticed a node being listed in live and dead list in +I think its good to go +Revised summary +Then looking at my stored results again after your note just now I saw that Id managed to blow right past the error in +Thank you +whoops just realized this bug was created before the drop so the patch in is likely do check and could you also change log level to DEBUG in to get more info in just wanted to clarify that this is not an issue related to the servers being on or off + This bug has been marked as a duplicate of +Can the Priority be raised to at least Critical +Thank you. +Fixed in trunk +Thank you for committing +Derby log with the query plans +Thanks all +Sebb is there a requirement to have all the authors from the commit log present in the authors section +Cool +One of those created scripts seems to have that errorThe problem is without the script we cannot fix this error +Igor I think you are going down the right path +Then null means id is not specified +And the test will verify that a instance is returned as and a instance is returned as +Is this similar in principle to storing the content directly in the filesystem and just a location in +With all tests passing we should commit +Ive started a thread in the list since I suspect the answer to be hidden somewhere in the connector spec +I should add that the odbc logic was not yet changed to the simplified callback mechanism since there is a lot of complexity built up around future chaining +If OK Ill commit backport this minor change to the test +Oleg +What do you think +Bulk close for +This appears to still be problematic on trunk but only when running surefire someone can comment on this? there was a lot of flux with surefire recently +If so we have that all over the code and just fixing it here will not workqlsrcjavaorgapachehadoophiveqlsessionjavadocs for these +I think its better to do this and get the rest of this patch in rather than bounce the whole patch since this patch has been stewing for a couple of months now. +Consider for or mark as WONTFIX if looks to be better designed for this +Lets be open about it at least +one affecting core code and one is the full change to be committed which includes the unit test +Fixed method name +Less risky version of the patch for passes the unit tests +We should document the settings needed for processing but I think the default should be style processing +Im marking normal urgency and patch availableKristian are you still working on this +Also It fixes issues with the logic that determines if a scan is interested in a region +Patch integrated to trunk. +I need the version to fix the Schema no longer modifiable bug! when should release? so please make sure does this matter is a bug or not +Looks good to me patch applied +I try to recreate the problem using WSDL you have provided +How do run maintenance on it if necessary +committed renaming do to simply write +RC is scheduled for the end of this month I believe but you can easily build Roo from SVN yourself if you want early access to a release that contains this fix. +This contradicts what joes said on infra last week and he made +although I have seen this mistake come up a lot its better to have consistent semantics and just educate users +Thanks RashmiAs you may know we dont support no more +Good spotting there James +So how can somehow write a Collector which has a hard assumption on that? Maybe only if he used a Query which he knows always scores in order such as TQ but then I dont think this guy will have a problem since TQ returns truebq +Sorry for the delay +You shouldnt be starting a bookie right away after that +Fixed in rThanks for the report +I think some of the problems you would like to see fixed are not involved with sortingbut some of the other datatype interaction with locale which I dont think is part of this JIRA issue +but if you change the sort order any part of lucene sensitive to it might break +Hi Robert this has been checked into CVS! +Thanks Yongqiang +Works fine +But it still can be useful +Applied the second patch into master and branches with thanks to Dhiraj. +Question should interop tests use raw sockets or HTTP as a transport or both +Darran Im not sure how FORM authentication relates to plaintext vs +Also there is text in the license file that allows you to redistribute the software as you like provided you meet some requirements to provide the files as is etc +Committed patch to trunk +Ill open a ticket for it +If you find further issues please create a new jira and link it to this one. +grails directory +Our deployment generator was adding to lib directory for unknown reason +soon +Perhaps we want to add a note and sample configuration to that page +Right ok so youre using a proxy app to shield the Solr http +Current Integration Tests refactoring is complete +SVN patch for using for this seems to be a duplicate issue toIm not super familiar with the current patch in but i believe it is a bit more general purpose then your solution allowing support for any arbitrary type of Directory implementationPlease take a look at that issue and chime in with any comments you have on the approach being taken there. +The fixes that have been made so far are now in trunk +Thats easier for us. +Thanks a lot +How is this looking Yuki +Added the info about enabling the Nexus Repository to the fileI dont think the shading will go away anytime soon unless theres a pull request sent or something similar. +Thank you very much for finding the cause of the problemOleg +Otherwise the are broken hereputting this concrete map on an abstract class that does nothing with it thats not going to happen hereLets fix the real bug that rate limiting isnt just wrapping with a different IO impl! Then the originaldirectory factory apis work fine with it if you want a concurrent map you can manage this all yourselfIf we want a sugar thats fine +The containerprotocol element is intended for configuring the protocol +Patch applied at +This too should be addressed +i guess i will look again. +Ive done the same globally using IDEAs organise imports +remoe duplicate directive in a note Nick Youre allowed to fix this also +Here was my initial test just screwing aroundI ran with ant test Dtestcase true G +Please commit these changes +Supported modes include AFTERCLASS AFTEREACHTESTMETHODI invite you to try out this new feature in an upcoming nightly snapshot or Spring and let us know if it meets yourSam +Actually Im leaning again toward a fast version since it might be nice to have on a GUI display and its not clear how long IW may spend holding the lock on thisAfter further code review using a volatile seems safe +But for Groovy there cannot be a fix for what do you want to do with the issue +Here is patch for UI enhancement for one page checkout processSantosh shot for one page checkout UI enhancement patch +It is a tradeoff between allowing this condition to happen and performance +I will switch these to use the Specs which should resolve the version issues +Also I promoted the default value to +committed to as commit dffbeceefbfce +What google code are you working with +I missed adding Test after Ignore. +Givin the test thing a crack hopefully Ill have one +One thought could be a race that is triggered more often on machines with fewer cores +jitoptsrv test +Stuart any update on this +This is how I fixed it +Sorry forgot to grant license to ASF for the attachment +Im sorry I had two different sets of changes in my environment +here is a in trunk and mark deprecated inregardless of whether its right or wrong if we arent using it i think its good to clean house +That is why I did it that way +Can you reproduce this? Does it crash anything other than your client? If not this should probably be downgraded to a lower priority +for the patch +Its actually the PLAYERS table +This can potentially cause the log corruption described in patch changes the writeTo API of the to use the message sets buffer to write to the +I also made all methods that should not be subclassed final on so hotspot doesnt mess around with it +Hi Per any news? Did you manage to get rotated views working +I just committed this +Ill check whether VPE catches this key stroke or notIf not the issue is related to the JBDS Model +Issue rejected +Can you provide some simple project or unit test which reproduce this error +The patch I submitted fixes the sorting order +I will file the two and post them here +This reflects Pauls suggestions and feedback +Patch applied +Everything works fine with as well so you can use the latest one tooIf there are no objections Ill commit it shortly +Anyways +So I do not think any thing is broken in terms of pluggability of RPC engines +Committed with revision . +You need to build it for as this is the version of Java that we support for Seam +Thanks Dmitry. +I have checked feature under linux and its broken for all XR versions +Reported fixed in JON Sept build Mark Little for fixed in JON Sept build change to JON +However this is a little incorrect because +Ill take a look at where the framework should call it +Revert changes to jBPM console xhtml +Yes of course! +So they can use Camel with Spring +Until we have no more analyzers in core exspect Whitespace +java filebq +seems like the CI URL via Sonar Settings needs to be Hudsonhttpserverhudsonjob +if not then start from scratch +I would like to close this issue as I think that Thrift shouldnt be in the business of allowing other string encodings +You should try to upgrade to Groovy this was fixed a while ago +need to do more researchthe advantage can be minimal and it is better left out +When that happens the JDBC driver will try to the connection but only if the autoReconnect flag is true +Thanks! +Patch applied tests pass checked into SVN. +This methodology also allows for each action to have its own validator or if you choose all have the same validator +Thanks DanielI have no admin rights there probably you need to assign them to me then I can clean up +This is fixed now +The dead lock is between an observation listener setting the session to on the next access and the logout call of the same session which waits for current observation calls to return +If we want the meaning of the operation performed at Dialect level that means the CRUD operations caller has to give info on what it is trying to do does that means to you we should expand the interface with methods to transmit that meaningintention +I will add that over the last week I have built this application from scratch four times and each time Roos behavior is the same +It is SVG images that are not working +Do you still want to take a look +Sure +Bugzilla cleanupProblem does not exist in FOP Trunk due to improved handling of property +default the decoder to the Date Decoder and initalize the decode params with the defaults from the decoder +Could you give me a hint why the ejbClass could be null? I could then check some breakpoints. +Thanks +Attachment has been added with description fo possible way to work around this segfault is to pass the following to the xsltproc engine stringparam hyphenate have something to do with has been added with description proposed P open bugs to P pending further review +passed in jenkins. +Added to trunk thanks for the patch. +Committed +This applies after patches for and are appliedIt has the genric checksum patch but does not actually use the generic checker inside DFS +Are you using a of? Does it work for that +I do not think we should be opening up Jiras for documentation +The jdo patch looks good +Cant replicate. +Would it be simpler to split the stuff into a separate Component +Committed revision +We also have a patch from Robert that will switch to downloading the file at compile time instead of including in the source +The patch looks good +Updated Quick Start section according to the new design and functionality of the +for the information +so I just created the context when creating the ssl channel +just always print this identification string in all messages? It is platform specific but defined to be unique so much better than J J. +these are related bits of functionality +This project is owned and developed by a community not by individuals +I also think thats the best approach +Now from the Return Header select the status Requested Return Accepted and click SubmitSo that a replacement order in held status is created +We can use Mockito for some of the explicit extension of classes for mocking purposes but I believe this code is going to churn a lot with the new UGI stuff thats coming so Im not concerned about using it now +This is a clue to me that groupName isnt doing much internally +or should we always fallback to standard Java on exception +for the feedback Michi its highly appreciated +Default gdb script is run which prints info abt threads thread Id and function in which it was running when task failed +added support for jogj +After retrieving tokens from tokenstore we put them in abstract classs maps pull it back and then delete immediately +I have created this issue for supporting all the set methods including streams so I think its better to close +what are the of computing a separate crc per transaction vs keeping a rolling crc but storing the latest accumulated crc at the end of each transaction +Hi RobFirst of all I didnt get any notification regarding your comment so could react faster +I would like to get the sample working first with Geronimo before applying the patch +git patch against the git trunk that makes the plugin work on you for this ticket and the patch + with some minor tweaks +Should do the stop signalling issue separately +The locality information is in the objects the master works with taking another run at this should be possible to do this without querying the file system from the Giraph side after all +for too if only to prevent our discovering this again when profiling +Thank you Christian! +Upload a patch with test case to reproduce and fix it +We can do something to help the user do this +And deploy an application with version twice and redeploy it againThe result is sensible +this is a problem with YUI code +Thank you for the fat patch and the for seeing it through Anirudh +Hi Julien in fact what you call the virtual authenticated role is what we call the group +bug for the +However if you have two server then I couldnt think of a workaround +I read your article Whats new with Apache Solr +I did commit the change to to disable the query execution when the user just wants the plan +How do the arrays of outstanding txs work? How do you correlate a particular bit to a particular transactionI dont get how this would be different than read point since you cant return to client till all transactions before yours have completed right? Or does the bit array somehow expand our vocabulary beyond binary read point +Using JZOS of course +Implementation attachedIll wait for comments from community +All the unit tests passed +So what you have in mind is probably support for plain Spring beans that happen to get exposed as? What use cases do you have in mind there beyond what you can achieve with standard Spring lifecycle and injection callbacks? no worries Ill leave things thenJuergen +so maybe we leave it like it is but also provide a direct link to the show file request handler for those files like the admin links do +Can you take a pass at removing those extraneous white space changes and then Ill take another lookThank you +There is additional isolation from the hot deployment scanner as well as the ability to suspend hot deployment via the interface now +duplicate to to +the selects arent relevant either Ill hack more on the test case +mmm maven plugins have to go in a different way Ive uploaded now by hand try it but I believe you need the tooIf so you can deploy it to your maven repository and well copy from ve copied the metadata havent tried the plugin though +Code demonstrating the for and fixing the issue +Can somebody give the solution to this issue we are also facing same you check if this also happens with +If a is already in active state splitshard will fail +Will need to look at this more deeply +Test and fix look great +I even tested turning on and off the watch location +Ive put together simple patch with this feature preview +I strongly suspect it was fixed either in or prior to so intentionally unset the fix version so this wont show in release notes +Within a minor upgrade say to overwrite the version info and proceed as normal +Just tested with latest trunk code the fix is working +We need to be careful with this because it is a compatible change +You cant create a new class that extends Vertex outside of the package +Would love to see this in since this will ease up monitoring deployments a lot +If your stylesheet and java code is complicated try to see if you can reproduce the problem by using a reduced testcase +hmm +I was getting around ksec +This one needs only one entity bean to reproduce the bug +Wait a sec I thought we had concluded that we cannot use Trove +This patch adds a new option that allows the user to specify the minimum amount of memory needed +Im not sure this is relevant anymore +view from Design level in Structure +Then I verified that I saw the and the Warning! Can not load requested doc messages +Adding unit test case +Thanks Andrew +Could we change the fix version to +If an additional patch is created it should fix the indentation in the generator +Thanks Stanislav! Leaving this open for the fix +Thanks a lot +Working ipv and failing ipv on Windows is good enough for me +Can you mail me you schema file? If that says VARCHAR for the column I cannot possibly imagine how the generated classes would define the corresponding property as in Int +Thanks +Having said that we can look into allowing the script to be run from anywhere +I resolved and closed the issue +If there are issues as a result of the comments thread please file a new issue. +adds to the to branch +Ill try to work on specs for it but like I said I havent had a whole lot of luck getting them to run even using rspec and an older ruby +Please try to get this in CR. +With window size W the server will only send W bytes to the client before expecting a response and blocking all other sends +I like this idea! In addition to removing the of Struts actions this should cut down the startup time of my app by quite a bit since it will only need to instantiate instead of +If there is a problem please open a new jira and link this one with that. +the fix break completly the URL parsingbecause URL expect a full path and URI support relative path +Also the finder find you mentioned above is not valid +Interesting is that supposed to be represented as null or as an empty string? In any case if you can supply a patch I would be happy to review I noticed you reported this against have you tested? A lot of work has gone into the so this may be one of the issues that was caught +so far cant can you try a recent build and see if you can reproduce it? May this be something OSjava minor version dependent? +I believe ivy can generate the necessary poms for publishing to maven +This restriction is present in Cassandra on linux +Looks like thats probably +In revision I added a simple first draft of a +So if you have some code or ideas lets try to define a common API and merge it +Token service changes erroneously started using the port in the uri +Got verbal approval from Mamta based on the html I attachedHere is the diff file +Also moved class to util package +Another thing to think about is if this should be unified with the very similar I agree this should be easy to obtain +But we do have a solution that we think will work +The zip is located in a Maven repositoryIf you look at the patch youll see that all Ive done is modify the following areas +Hi MichalSorry for the confusion Confluence is not linked to LDAP or acct creation +If not then the copy statement doesnt have to include the location of the repository +If we see this again we can reopen. +Will verify in clustered setup +Is the deprecated class coming from binary or source +Im newer to james development as you +Adding +Yeah I think it will work with the current patch +The along with a patch for there been any movement on this at all? Is it likely to get done by releaseLet me know if I can help at all this would be very useful for us as we could get rid of some embedded SQL which is lurking in our code waiting to bite us +Clausyes in could cause some trouble I will add some test to verify it and polish the patch it is needed +Can we think of any other unit tests we could add +It may be that Pavels idea is better but I dont think were going to make more progress on that wo actually coding it up +I can improve the method comment before I commit unless you have other things youd like me fixThanks for the review +I would appreciate the addition of this feature +It seems just like you began waiting for the dispatch thread to terminate and that allowed some other thread to start and it made a call to +I recommend using shadowing this way for example at the bottom you have the copy pulled out from the DIB and around it you have a wrapper +Geronimo must be applied before this patch is applied +My use case is that I want to be able to start a process definition based on the version +Empty map can means the case that the container is at LOCALIZED but actually theres no localized resources +I know that sadly some people have to still deal with EBCDIC +Dhrubabq +Do you see epollwait taking a lot of systime +mea culpa I didnt wrote a unit test the first time I fixed itThe patch proposed by Oleg seems fine Markus hint about the may be appropriate but the class is for internal use ATM and it should never happen to call that methods when the stream is closed in mimej so we can improve this even in a future release +I just sent a message to the dev list about this +Included in. +Thanks +bulk defer of open issues to +Will this issue be planned for any future release +I think the test should bepassing now can you confirm this in your environment +from me as well to allow the distribution of DBLP database use for with the caveats on attribution cited by Craig +I understand the fileSet in the assembly but Im not clear how to use the dependency plugin +Do you have sample solution code +I eliminated the example component demo code and dataI applied the patch to the framework folder of the most recent SVN version and it worksIm done with this issue +Proposed patch to be applied after s a patch Ill apply on trunk soon unless somone objects. +Only works inside the workspace +Error messages were rewritten for the. +Jody fixed that by the apache logj plugin +Darryl Ted this one should go to trunk first should it not +That makes senseThese files tend to be largish +It was released with the fix in +Change the based website to confluence based + issue working under the current to be reopened add label and resolution. +Waiting for a release. +Man are you quick! Thanks a lot. +Patch for issue +with both the latest trunk and branchThe original issue is fixed but even if not likely to happen its still possible to get the authentication to enter in infinite loop +As we have more contribs we can do thatA few missing things are Javadocs for contribs are not included in the nightly builds Clover integration has not been touched Probably many more that I missed +The WF for this JIRA is broken created and posted patch there to run and do the review +yep this is fixed in rev +Closing see previous comment. +Create default Seam project press Finish +So I did a regexp on all the source to unexpand these down to just Id and then I could apply the patch +Can I take this one +Thanks for pointing this out +anything you know aboutKrasimir you are running with a bit JDK right +Looks like were doing another retry so this is now in. +If its holding on to the objects across call and multiple threads tried to do this they would invariably step on each other right? Now it may be that this is not intended to be threadsafe but that may not be obvious +Work continues when is resolved +I ran test manually and it passedIntegrated to trunkThanks for the patch Ian +With the new running scans feature on the monitor page I noticed one node was slow during verification +I think there is something wrong +I just committed this. +We can add it back in later with an updated example +I fixed the grape occurrences but there is one more in still outstanding that I am still investigating +No patch yet +Was fixed by issue +Thank you Jack +I still think that clearing the session on afterCompletion regardless of the thread that calls it is correct and necessary +Your vti seems to be compatible with the charter of that schema as described in the Reference Guide Derby provides a set of system table expressions which you can use to obtain diagnostic information about the state of the database and about the database sessions +But it is a workaround not a general programming model +From there I can create a simple uber jar with a simple script to call it +Nothing with s works +We need to limit the amount of map output bytes or reduce input bytes as they must fit on a single physical node +But not for any very good reason +Bulk close for Solr +What you are seeing is normal +If the memcache size is not growing over time and causing then I think this the best we can do for +More detailed description of the project +Also since it makes futon completely unsuable I think the patch shouldnt stop it and should also patch utils +Thanks actually have much deeper thoughts on HarmonyRVM integration +if I got any issue I will inform you +This change is now in trunk as part of r. +Then we could get out of the business of having to CM jar files which Ive never been a fan of +Rob which part? the problem or suggestion for solutions +I the to and heres the attachment +Overriding the throttle value to should solve this +How does anybody feel about committing this? Has anybody been able to run acouple of tests against this +using the partialstilesinsertAttribute behaviour implemented too? Its unclear by the wording of the issue. +I turned off clamav for the domainKeep an eye on it and let me know if it helps or hurts +Hi MichaelThanks a lot for getting back to me so fast +looks like is taking a different direction so may never be released. +java truncate same sensitivejava truncate same sensitive commitTruncatorjava truncate same sensitive commitSelectorjava truncate same sensitive commitTruncator commitSelectorjava truncate same insensitivejava truncate same insensitive commitTruncatorjava truncate same insensitive commitSelectorjava truncate same insensitive commitTruncator commitSelector TRUNCATE TABLE command hangs waiting for a lockjava truncate different sensitivejava truncate different sensitive commitTruncatorjava truncate different insensitivejava truncate different insensitive commitTruncator Selector hangs trying to read the second rowjava truncate different sensitive commitSelectorjava truncate different insensitive commitSelector Selector trips over an NPE trying to read the second rowjava truncate different sensitive commitTruncator commitSelectorjava truncate different insensitive commitTruncator commitSelector command is DROP TABLE DROP TABLE raises Operation DROP TABLE cannot be performed on object T because there is an open dependent on that object +There were some other text files added in this bug as well to be verified +Issue may already be resolved may be invalid or may never be fixed +Testcase is attached both sources and executable +Thank you for your update +See increasing default permgen is just a stopgap setting it to any level will fail to m in and serverproduction think that should be a good default. +bq +I committed the fix to cleanup the cache and Ive also switched to using the jar created by a WTP export as opposed to our own home grown export +Added a method to +IMHO getting this working and tested is probably simpler than pushing for a clean shutdown option +Thanks I understand now +focus on tests already exist +cannot vote or change your vote on resolved issues + resolved issues for already released versions +I have committed at subversion revision +Removing from NEEDINFO which means moving to ASSIGNED in Bugzilla +trunk . +fixed as far as I know +Patch applied in commit edc on and on abafc on master + file system cache test +XSLTC is now able to parse expressions such as and and and or and child +Final patch. +this isnt the right approach we are going to make the multithreaded code perform properly +I could not reproduce the unusual ordering of test cases which is turning up in the Java test runs on Solaris +Lars stack some thoughts? Would be great when we can commit the patch soon +Want to submit a patch? Please make sure you add the old config name as deprecated here. +This patch does not have any tests +Im not sure that we need more than a better Upgrading section in NEWS +In order to have a complete regression test that checks that both this issue and are fixed Ive improved the designdocs tests accordinglyThe tests can now really prove that fixes as wellThis patch makes sure that we have a complete regression test for the bugfixAlso the success of the view cleanup test doesnt depend on the view info tests success anymore. +No so on WindowsHowever when we delete directories we might want to use instead of and that should solve the problem + doesnt exist in the codebase so can it be postponed to +Sushanth I havent debugged this +Sorry for the delay this should have been checked in a long time ago! I am currently trying to figure out where it should be placed in svnSee the discussion on the dev list +Any commentsreviews +We could btw +creates an internal set method from the +I am not sure I understand this and why the patch should be reverted in Apachebq +Committed to trunk with revision Committed to trunk with revision +provides karaf commands to check the status of brokers etc + + updated forrest doc +Part of it is time constraints user experience and unclear requirements at this time +Do we have a comment on this issue? +Thanks Vinod. +A patch would be awesome +The dependency of conf on the should be to avoid the warning on the deployment of when is not deployed +Is this latest patch fine +This is a workaround for the perf problem thats valid using TCP but not if using RDMA +I agree with your commit message description to not fill up the svn logs +I have read more carefully this issue and I am pretty sure that is not a duplicated issue of it so I reopen +Ill begin working on a solution immediately and post it here for review as soon as possible +It is still the same style works as expected SVG does not +Thank you RaviAlso changing the Target Version to +Python Perl Tcl look OK in CVDBut the tcl example generates duplicate annotations and gets an overflow error +That said I didnt want to blindly remove it without knowing why it was there to begin with +bulk move of all unresolved issues from to +Added service for save amountPercentage in the entity +converterMessage requiredMessage and validatorMessage attributes added to richorderingList richeditorconverterMessage and validatorMessage attributes added to rich add the ajaxKeys attribute for richtree the corresponding +Some logging has been added to Apache release +Patch to concat indexes ahead of the parent +It wont be necessary to have anything special in pom +I reviewed the patch +I think this is really a documentation problem +Its with the test case provided. +I will have a look at this +Product constraintProductAKNumwhere Num HGNum rows selectedijI will split the DB by files and provide the partial zips and provide them shortly +The attachment is one ZIP File of a Netbeans project with sample Implementations +What happens if the users gives the wrong parameters? As a user I would rather not get confronted with a stack trace even though it is an example +I attached the mentioned file +Deployed SNAPSHOT version of the Commons Auth bundle +Im also not sure whether option or from above would be the best key mechanism for the +My one requirement is to know if data is lost and if it is a small amount of data to be able to identify what was lost +After doing a flush on the table the scans are about x faster +reproducable with ve been playing with this option a bit and rethought + +This is a +After doing cleanup failover works as expected. +Fixed for Ip reassigning the bug to Sheng for fixes + +bulk defer minortrivial issues to +Im unsure of how to best integrate this into the code + another test scenario +Could you tell me which jira is for the above changeI believe it was in the patch for this jirabq +Im not sure how to fix it but making progress +I will fix this as well +Ive just committed this to trunk +When we update Gora to we can commit this +Can you regenerate the patchIt is not applying cleanly +fixed in resolved issue +The target of is to integration the great systems especially for the situation of multi installations with one ll try to finish this work in the end of this Feb +Idea is too add an addition tree node level of web app path such that the combination is unique +This works in SVN head +Ill try and look into the problems you lsited today +Is there still time to instead of just deleting those two lines implement proper filterI can do that tomorrow first thing in the morning +I didnt specified for any at that time +iocoredns +Removing warnings also in the trunk patch +Found a few additional reasons for confusiondifferences across devos you run help you get All profiles incl +looks good to me will commit to tonight if no objection +A new class would be best +I think you can count on a general agreement that this is definitively a step forward for Oltu is there any chance you can submit patches to improve the current design +It seems like it would be worth trying to see if it pops this issue +Back compatibility fails +Attach a quick fix +If it fails to load the image print out an error likeFailed loading image from foobarfsimage +Did this fix work for others or was there more to it? +Julian what are you trying to do exactly? Ad how are you doing it? Please give more details +DEBUG User limitcomputation for userfoo in queue default userLimit user required memory consumed memory limit queueCapacity qconsumed currentCapacity activeUsers clusterCapacity DEBUG User userfooin queue default will exceed limit consumed memory limit memory +applied. +Unfortunately it is not clear what exact change did fix this issue +Should I mark this bug as closed or does that happen on a release? Not sure of the process around here +fixed. +Looks like a Eclipse bug in this case. +About the documentation I think its better to simply encourage users to configure useDLQfalse +This issue is resolved could anyone close resovle thisthanks. +Thanks for the additional feedback Rossen +and was afraid my patch would fail to apply if I did more hadoopqa cycles +Coolio +Jason any more thoughts on this? Why and how +This looks like a class loading issue +I already have developers breaking things while touching thesitemap and would prefer to keep far from their hands and from thenecessity to change it often +As discussed with rnewson on IRC this is the classic Windows vs +I actually tested this patch +When run via ANT they still throw NPE but I think these exceptions occur while using the component API to obtain the model contents +I have been looking through the code and found one place in the where there is an apparent transaction leak +is going to fix it +BADIf we make it a retry operation then it will drop in to the block +the ones that fixed it of course +HiTry my latest patch that is just committed +All in one for trunk version +Please let me knowthe status of this bug +Probably a hole in the exception handling logic which applies to closing the TCP connection before the remoting Connection is up +bq +As soon as theres like two or so pages of code the scrolling is becoming painfulAndrs told me it was related to how the text area redrawing was coming on the event dispatch thread I believeSomehow wed need to move the redrawing away of the thread of find a way to just redraw the current page view +PR Ive removed as per previous comment +Ive just committed this +Thanks for the patch! +Thanks Jon +Ive also found this problem on my projects s with s with not with any the attached project I am seeing a good improvement it possible to get a release of with the improvements so far +Unfortunately I missed updating the golden file for the new skip default partition testcase +Thanks for reporting and providing unit testI used the test and fixed the thread pool to be per camel context +What ant tasks are you using to build the softwareIf you are using mysql you need to runant Ddbmysql +Id rather see implemented first since that is a logical separation while this is developers guessing which configurable variables are more important +I will implement that fix in my version now with one changeI am replacing the check for an oracle driver with a check for an REF output parameterThis will allow the original logic to be called for standard procsThis modification will hold me over until this feature makes it back in the baseline +Indeed Bryan you might also want to run ant as this runs the file where Ive added a couple more testsThat runs fine with no failures on my end +The curl exit codes are well documented +Tuomas said there seems to be no way to find out if its this error +In this way we do not need to change the axissvcclientt for each and every new security requirements +Hi Zoe Changes committed in revision +Possibly a breaking change for anyone that has clients coded to the current a feature +I do not have a working build please testwith. +Also I think we need to addimprove some Test Cases to test this behavior +Fixed in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Attaching new patch which will take care of most of the comments +gonebad gone re all still bad ok to rm the last two are over days old anyway +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +I assumed that bit was to resize segments we recycle at restart after the user redefines their commit segment sizeThat makes sense +Dominik has something changed? I was able to deploy the driver as a module a few months ago without modifying the file +I have a full table scan in isolation benchmark Ive been working on +adjusting fix versions given that it was this bug now with embedding the console in your own applicationNot just using the command line to fire up the console? +Minimal multi module setup demonstrating the problem +Sorry about the link looks like it got truncatedYes the issue still exists with rc +Updated patch with the comment on +Attaching the patch that fixes this issue +Nope sorry +Im not sure that the default behavior should be show all +I dont see that we can get them all to use IE or FirefoxThe choice of the communication mechanism PPR or XHR is very important +bq It should be instead of I actually meant to remove it because there is no logging going on there I added it for my own debugging +Apologies WONTFIX would have been the correct resolution in this case +Please review +I think that maintaining the information on the wiki is working wellAny objections to marking this issue as Resolved Fixed? +Delete the previous one please +However JRUBYOPTS works +Comitted with tests +Found someone on the project with enough karma to disable the job. +And I think it is OK to use the localhost instead of +I am very sorry to say that it is my fault in setting up tomcat to encode jsp pages using +Unknown release vehicle. +Please send me your comments and critics so that I can get better +At some point you will have adjusted it high enough that further upward tweaks wont help your application +No worries +Could you post your configuration here please? It sounds awkward that your leader was hanging there for over an hour without support for a quorum +Task Define ISAPI Virtual DirectoryNot immediately clear why this needs to be done +If you finish it before we release beta please correct the version +Am I mistaking somewhere +ah this is output from the generation +Also added reflection lookup of if this field differs between Oracle JDBC driver versions. +Yes and thats simpler to accomplish its just potentially bewildering for the small number of people who inadvertently reboot between installation and configuration +Hi Keith thanks for reporting it I will fix that tomorrow +Christian +Closing old resolved issues +but we can do this after the larger portion of the warnings is removed +from RC until that is complete. +see the linked task of the linked task as well +Let me try this patch on + +So I have updated them allAlso the unit test changes you wanted are more stubbed out then I expected +Partially working prototype to illustrate a fix for this issue +The implementation cannot be static though since multiple Esper engines per VM can +So heres the problem +From my initial test this works fine +The link between the concrete Java class and the Ruby class is hard and it should be possible to safely associate instance variables in that scenario +It is when you edit it from ServersOverview panel as seen in the screenshot +awaiting testing and verification before closing +I also moved around the code a bit +Of course Spring transactions will the Session and perform transactions on that I see no need to make the associated JDBC Connection available No code should need to access that Connection outside a transaction neither should code access the Hibernate Session outside a transaction +Verified with JBT. +Can this be fixed by +Actually what needs fixing is the remote that produces the map +Does mean this can be resolved +Linking as per John Kaputins comments +Spring Security follows Spring because Spring Security was done before Spring +I am very excited about Tajos capabilities and am interested in contributing to the projectI am one of the main engineers working on deploying Parquet at and have made a number of contributions to the Parquet project such as adding support for the FIXEDLENBYTEARRAY data type and a number of Avro support improvementsI am excited to see that Parquet support is planned for Tajo as well +Ive fixed archivaaliasmonthly +patch starting fontlib on xHe changes the use wchart to unsigned short makes all data structures used in fontlib to separate +I dont think this should exist but we should probably tackle that in the next release to make sure we are fine with the existing changes at first +Currently Hadoop uses +The only changes were made to +doh! Thanks Jason for catching itJacob your problem should be resolved w the one line patch that Jason provided +We need a reusable way to link pages from jsp and tag that use a standalone helper class would be helper class can be used by other tags like the outline tag +So I think the issue is actually caused by increased size of repository +We should make sure the design failure analysis is solid before committing that feature +We have this issue in our application in Chrome however the workaround did not work for me +we can start work on this one ge it reviewed by the community and by the time it is ready if is already comitted then we rafactor this patch to be compatible with +Not really +Thats why the rw node type manager is also in +Maybe its the most stupid thing but were quite dishearten right now +The patch could not apply to trunk after the changes in + +The user should pass the imported XSD to the tool separately but the NPE processing the importing XSD is a serious issue +please ignore. +The test cases for the in representation available in which is attached +Probably every command in the accumulo shell should be replicated in the util class +The usual situation for me is pretty limited to one component at a time +JBDS crash dump showing problematic stack in workaround description +on as a bad default +So I would like a decent amount of bake time on any seemingly good solution +I just committed this +Therefore retargeting to the release +So proposed solution will add unnecessary complexity to plugin +Also note that I use a French system sets as a French time got by this damned cancel button +The broker log isnt terribly helpful +That means that when the console plugin is created it will have a dependency on and pull it along even in the tomcat installation +oops named it wrong the first time i attached +With security off this introduces multiple code paths with added complexity and reduced code coverage +New patch address the comment +Committed +perhaps your patch would be unnecessaryId rather not add something so to stdlib if we can find a better way to patch itself. +I am working on getting the patch out asap +Ive done a bit of investigating on this but dont have a windows machine handy at the moment to dig deeper +We always welcome patches from users who can figure out what the cause of a bug is +Committed revision to +Please update the bundles with this new added module when you apply the patch +Also the seam catch documentations section about integration could be more useful if it details how to integrate with jboss AS +I just committed this +Eh I said Wont Fix not Fixed +bq +Attached patch the optimization and fixes the safe mode unit test accordingly + +closing this issue +Help welcome +Closing all resolved tickets from or older +It simply lays out the groundwork if we find it useful to return that information through this API or even a separate API +Close issues after release +Small fix to a bad problem +The only case this wouldnt be true would be if the tool was only for tests +This patch depends on the patches currently pending for andDoes these scripts take care of issue outlined in +Uploaded a straightforward patch with implementations of and that the tests use instead of their own +I really like your proposals +We need to see what state its in and if its still relevant +A string beginning with a +Please attach a sample pdf if the issue still exists +It seems like a rampart problem +Reproduced locally with run +Renaming this issue to more specifically describe what I am going to doWill not make it into the STS build +I connected to a hung job to debug it and add my comments so it was working in thatIve had a look at this issue and updated the description with what should now work +Nice improvement + +Sounds like a good idea +In the meanwhile I am going to try developing with which is obviously not optimal because that version is outdated +I pinged Mr D +This solution does not cause a deadlock in the case of bookie crashes which happens if we change the pending to a blocking data structure +Attaching patch for fixI introduced new message to ACK upon file receiving +I factored out what I could +It was for +We need to release the firstAfter that Ill release the +This means that the solution of setting the to the build target dir will not work +Could not find your rejectionThere were a couple of issues where CVE was not translated correctly for some databases but this is a different thingJust take a look at the unit test it seems perfectly valid for me +Note that this is NOT a correct fix +logo included as part of fix +bq +Unit test that demonstrates the problem +This was removed per Owens request to make the system more secure without depending on environment variablesThis revision will check for HADOOPPREFIXlibexec exists +Committed +vdb attached +I dont know much about gradlegroovy +fixed by correcting the way property names are build +Thank You Kathey for your ReviewI have modified to include your suggestionscommentsThank You +Perhaps we need a specialized request processor catalog for applications that utilize file uploading +Read +If possible please attach the hereIf those WSDL files are already in the svn please mention the attached patch has fixes in the namespace uri definition +Then you can maintain a single plugin. +An alternative would be to use the as default as it makes it possible to reuse cached threads This task starts N threads and waits until they all finish +So I really have no problem with us folowing only Although it would be real easy to adapt to the underlying framework Im afraid of the potential for creating confusion +Trunk can build even without the exclusions listed in the dependency for +P open bugs to P pending further status from ASSIGNED to NEW for consistency +Added patch to update the docs requirement patch but didnt state the JIRA issue id in the commit message so the subversion commit is not showing in this issue but it is fixed. +That will make them testable with +For Airavata case since the deployment is bundled with all services all of them are collectively issuing one unique id +One of these options is strictly correct and the other has the overhead of a single line of code and is far more convenient to the end user +I will verify once more on the next available build +A user has what attributes measured for similarity +How about filing a jira to implement so we can actually fix the issue so we dont have to work around it in the test +This in turn leads to a different of cubes and colorsThe cubes should be normalized after spliti continue to work on this +No Ive seen this happen during any CommandBut it hasnt reoccurred in quite a while +Hi RohitI glimpsed over your patch and it looks good +Though not ideal this has been a design feature since support was added to Roo in +Yeah you included a snippet on the mailing list +Addressed comments made on the review board +Adjusting release note status. +You guys appear to have necessary magic such that compiling against does not build the compat jar and vice versa +check for the in the finally block before releasing the row lock +Deprecate during remove in +This patch removes all nested else warnings in the test committed changes to test cases. +Suggested SVN patch filePLEASE NOTE! Includes suggested fixes plus test cases for SOAPJMS Spec implementation needs to implement Section Fault details binding via CXFs WC SOAP JMS Implementation should generate mismatched per Section via check +This is actually bug in the Maven core and is fixed in the upcoming release. +Looks good except I dont know the name of the new GIF file +for the patch thanks +The fix has been checked in. +So will get it done todaycheers sent a pull request +The picketlink bindings for EAP have been moved to a separate jar +Moved to next milestone release +Changed to use groups +for the patch +failure is being tracked atIve opened to track streaming +One Comment for Avoiding byte copy when init seems will not save CPU time +Committed to trunk +FYI ant is helpful in easily identifying which files are missing license headers +Also should the fails should the build fail +Alternatively this may simply belong in element of the +I was only able to get this NPE to show when CLOSING the mbean explorer view +Im not aware of any compelling uses for multiple server instances so Im very much against introducing more complexity to make a small aspect of uninstalling plugins slightly less problematical +toSardignia +I have to manually read the poms because I cant get the real info from the models about the plugins +CP updated in revision +Patch generation mistake on resubmit again with RPMDebian packaging to layout files according to proposed file structure layout in +Relationships +Merged into the master branch. +Is it that hard to get the proper poms metadata? Can I help +This was a test of Teiid Designer +What browser are you using +The problem really is with this is not a Jetspeed bugThe unique index cannot be made shorter without a complete object remodelling we need all these columns in the index as it represent a logical model constraintsThe fact youre overrunning the mysql limit of bytes key length indicates you most likely are using an extremely long preference nameI suggest trying to reduce your preference name length as easiest solution here +playing devils advocate here +Please go ahead and create a YARN JIRA to address the YARN part keen to see your write up and contribute. +Thank you Ivan +I am not really sure this is something that can be fixed in VFS +What if anything do we do to verify that this is working? Im also unclear if what Ivan did makes it editable through the webgui or not +well see. +Help in resolution is appreciated +CP fixed in fixed in . +patch file on this JIRA +Thanks Isuru! Commit is ok. +Can you attach your comparator classes to the ticket so I can test with them +We will fix this and the more general problem with the Ref classes +Part of a bulk update all resolutions changed to done please review history to original resolution type +There is one other option +Waiting for snapshot to be available before merging to trunk +The sample file is small +I think we should make an hierarchy for the pages of the Wiki in Confluence. +Looking at the code where the NPE occurred it seems like it could be host OS specific +Lets fix this +Committed to trunk +Hey Ales logo definitely I put it up on the monitorPrompt I lightened it up a bit and perhaps expanding to show the full monitor will make a little more senseCheersCheyenne +Committed revision +Its used almost nowhere in the code +wouldnt allow short circuiting if hashCodes were equal only if theyre unequal +Need to look at some example unit tests first +Sorry Ive been away for some days +Emmanuel I linked the issue that fixes this problem +Forwards compatibility issues with older consoles has been addressed by as the addition of authentication to the RMI connector prevents them connectingThe patch is largely identical to the previous changes with one alteration made to catch which get their key properties from a different method than the other and so were not versioned by the previous changes +latest build appears to resolve the issue +RolandCan you live with these changes to the interface? checked in +Have you done any benchmarks? Perhaps using the tool introduced in +the patch looks good +This is really up to BEA +Makes a lot of sense to me +good thanks! +Thanks this is fixed for me +I agree that it makes sense for validator toinclude a built in way of using its scripts but calling everything javascript israther limiting considering the tag outputs the script data which could be anythingIf other validator folks think its ok to add a JSP tag to the core distro Iwont stand in the way +This patch fixes all test failures in +So the path mapping allong above lines probably is betterWDYT +Thanks for the note Andrzej +We dont currently have an md file on the edits +Incorporated review comments from Rob Chansler +Sorry for the inconvenience. +Im attaching and with your suggested modifications to the topic +And thats why you wont find it anywhere else +Patrick anything you want us to try or should we patch up something w the above two issues? +So this issue will be solved by FE rollout +i think it makes sense to null out the node pointer in the token after the fork has executed +You can have the example point to the same Solr +Could some one test this example classes? +Cool btw i still thinking that some developers find this feature more usefull mainly if we receive just the name of tables that already created by system analysts +This makes no sense at all +Not the most elegant should probably make logging consistent across all native +Very likely there will be errors in the error log +Moving fix version to +No stack dumps +But the incompatibility introduced in the change is unnecessary and is also an important issue +Another unit test that shows up the same issue but this time all in one class. +This has been committed to both trunk and the experimental branch. +r on trunk +Steps to checkuse Console Configuration with many Configuration nodesselect all its nodes and Open Mapping Diagramautolayout or change something then file should be saved without exceptions. +Can somebody review this patch and give some feedback please +Ill try out the tomcatrun later +Changed the version number +I added it and now passes +Eclipse Plugins even in the name of the chapter is implied that we speak about Hibernate Eclipse tools or in order to use the info of the chapter in JBDS guide it should be some of the sentences of the chapter this fact is also underlinedlikeRefer to Key Features section to find all benefits that you can take advantage of while using the tools within Eclipse +mvn version of the site was not caused by the but rather the +There are one problem +Enable filtering of resources +Its much easier to understand and to maintain over the choice of annotations +The patch looks good +If not you know what to do +Any additional constraints would be defined in separate apiimpl jarsThought you were alluding to some grander reorg +Hi Brahma can you tell me exactly what you did to generate the keytab in question? Its unfortunately quite easy to accidentally invalidate a keytab for a given principal if you later export another keytab including entries for the same principalThe fact that servicePrincipal is null is interesting and I would guess that its due to having not properly configured the relevant principal setting in your configs +I changed the targetNamespace to be the same as the other file but that did not seem to help +Vladimir please close this issue when youve fixed it in CVS on HEAD and branch +Will provide a patch which resolve this issue +Patch to upgrade server to derby also if derby gets the release out in commit mupgrade to Derby and clean up console dependenciesSending applicationssrcplanSending etcSending modulesassemblySending modulesassemblysrcplanTransmitting file dataCommitted revision +I think we can consider this for the future releases thenthanks ran the Axis unit tests with a non coalescing parser and discovered yet another issue with CDATA sections in AXIOM +Fixed minumum line after some tests I see that some block of code in this patch now generates some drawing problem but it seems only in labels inside Tables and Lists for example see the screenshot in attach from a run ofWe must fix this now +Hopefully this will be eclipsed by which has unit tests but I always thing unit tests are a fair request +Looks good committed to SVN. +All margins and HTML commands will be transferred to HTML plugin. +Ive seen it of the builds +That would be a different JIRA dont you think +Ive created the sample for this and initiated a pull request +Subtasks complete declaring victory. +Inside if you use the you would likely lose the ability to configure conduits from the configadmin services as that is done through a +My badForgot to include the new fileAdded the new files to the the corrected looks good except you have some spurious whitespace differencesa new patch fixing a few places in comments and extra spaces +Added patch that fixes the problem +Would you be willing to write a more pervasive fix +Another maybe even better idea is to have a normal maven project under +The content type is determined by which belongs to the Java Activation Framework +Ah yes sorry + +close it. +Excellent +No it is no problem In fact is used in httpd logic because is invalid +Well I would argue that a is not supposed to bypass the proxy in the first place It is simply supposed to invoke the specified method on the given Object whatever that object is +The more I think about this the more I agree +Is it possible to put a snapshot with this fix somewhere so I can test it +I was thinking it might be easier to edit in the file here you could supply the services IPcontext path +The variable names are not correct for Struts rc +Thanks for the fast work +If I put a scrollable on the st tab of a tabpanel everything works as expected +It seems to be resolved +Hi PaulexThe patch looks fine +JSF branch version of the with custom +GrantI removed the Patch Available status. +Based on the following points it looks like its complex or unnecessary +Please be sure to make these changes in the javaengineorgapachederbyloc +I think the removal is ok +Rick +In previous versions youd be tweaking argLine on the forked jvm to allocate more memory whereas now the buffer usage is within the surefire plugin itself meaning you have to tweak MAVENOPTS to get more memoryI will try to fix the underlying problem just figured youd like to know this workaround as of now +added Apache header to are few failing tests but all of them are related to and will be fixed correspondingly +Pulling in additional +In FP release will be merged into trunk with the remainder of FP tasks +Thanks for the clarification. +HiCan someone please commit these changesThanksRavi +Added a link to +Can the FST be made concurrent for use in as the terms dict implementation +Runnable version of the new designimplementation +Any reason not to commit this patch? Its been hanging around a long time +Can you please give me a time estimate of when this issue will be resolved so I can use this information to my PD management? Otherwise we cant advertise that our web services will be able to use +I am committing this +This patch adds a trailing TABLEID column to the SPACETABLE vti +Can you please make sure that the code doesnt depend on classes in the package +I tried these changes but persistance still does not seem to be working +Changes were checked into Teiid Designer that will be in the next release for JDBS +Completed At revision +Also the delta container request would have to specify specific locality to place it exactly where the existing container is or else the RM can allocate the delta anywhere +Wait for eclipse bug fix. +I cant change my password given by my admnistrator and this password contains +I think so +Bumping to Critical since it affects Rails +Fix only available on trunk +I used an older version +The problem was that the conf being sent in to the serde was being picked up from the but wed not called setConf on it in the case of a so there was an issue +My local findbugs is happy with this version of the patch +Teds comment was correct my initial scenario had a problem I updated the scenario +or similar separated somewhat from the source projects. +External interfaces REST Thrift and CLI will not require significant changes as tables can be referenced using +Unless I am missing something +These one doesnt crash for me +Scott any progress on this? Let me know if I can help +Ill leave that one open just to review the test output again for CR +We can have the same path there or dump the jobConf in the log file +I have also included some scans of diagrams I drew to complement the documentation text in the patch does not currently link to or otherwise include the diagrams +Seems like some Hudson problem pulling core jars +More diggingyielded the interesting fact that this behavior disappears when I removethe INFOconsole setting from the launch configuration +Yup Jozef is technically right +Attaching a patch on behalf of Ravi +Hi ZeinaHave you had a chance to look at this issue? +thinking on it my change will not work +Closing +Dmitri there is a syntax error in the descriptor +I would probably not name it but something like etc +I am fine to commit it to +have the reflector be smart enough to choose between vector and raster +sorry I didnt read its a windows issue its not fixed +Oh yes I meant I just misspelled it. +Change was reverted on master and due to failures related to to master +Attachment contains Length Date Time Name srcjavaorgapachesolrclient srcjavaorgapachesolrclient srcjavaorgapachesolrclient srcjavaorgapachesolrclient worked I needed to explicitly list the commons codec jar file on the think this issue is now out of date it looks like moved everything to +Committed revision I am used a simple archetype so feel free to remove everything that is too much +I have committed this +CPD show two files with duplicate lines +removed the nested jar since this is a very exotic approach and I doubt we really need it +Patch that implements it +I created another runtime definition using the same EAP install as before and a new server definition based on it +If I change the repro to use a instead of a Statement I dont see the NPE if the server version is but I do see it if the server version is or +It said VNC Proxy +We found the root cause error +This is the main point of us using live code examplesIf we really want to save some time we could set it so the sphinx docs only are build when mvn install Pdocs is called +Committed the log bundle project in Rev +I think Roberts right we should not have shoved this in at the last minute even though it is a pretty big issue for those doing highlighting of larger documents +Applied the patch bcdadaafeebdfffeb +I just ran these tests using the previous transaction framework they passed fine +and pass cleanly +Yes Ram +related to +Assigning this to Alejandro to correct +update on this one generified the patch minor bug fixingBut not working properly tests failingThe ability to have multiple implementations identified by an id attribute in derived classes obviously is not yet implementedNo test cases yet for new is sad so see that the resolution is set to Wont Fix which I understand to mean it wont be implemented +But it seems like doing it once right when ZK has kerberos is better then doing the work twice and in the interim have a solution +Will commit when tests pass +This should be fixed in +Thanks a lot! +Let someone else pick it up +Perhaps so +Id imagine that anyone else who embeds Axis does the same thingI dont see how there can be any harm done by providing this war file +Bug still present +Thank you +One more comment to item for EAP there is Instabilities of test on IBM JDK +VM interprets it as and dont free native resources +Thanks +Broken links are found on wiki RF pages articles needs fixing +is in +Will fix up the diagram ASAP. +Kind of violates the spirit of imho but may be simpler +I dont exactly find the correct source code position but I think we should implement our own metrics package which we can later add to ganglia +I still dont see any local latencies being added with you put in a breakpoint or line or three? Is it going through the weakreadcallable path? If so is being called +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +You have to run uDig with the configuration option to run it as a shared program +Franck if you need this feature urgently I would suggest as a stopgap to use something like to listen as needed and to forward to a local service +If is going to be a bit we really would love to have these issues fixed inWe really do appreciate the additional work on fixing these issues! Many thanks +Which is why we are only fixing realy critical bugs in +This patch changes all of the testtestdata files that use per Michaels comments +Implemented. +Hi Eranda thanks for the updated patch! I will try to have a look at it this weekend +looks good +So thats more evidencethat this flag is useful to reproduce this problem +I am committing this patch without because these are minor changes +Didnt look at the code someone know the code should also take a look +You can switch to the Derby persistence manager in the +That should workRegards DushshanthaThe same behaviour is seen in echononblockingdual sample +I have attached the diff file here +First this patch still need a lot of cleanup but we would love already share the patch in the hope to get more feedback and poeple that test this patch +This patch starts the data receiver thread in run instead of offerservice +Please provide steps to verify this Jira +I will revert the patch for this so at least quick search works again. +Im resolving this as incomplete +Now there are only failures in Seam UI +Could you explain how DOM JDOM etc +Updated patch for todays trunk +The patch is applied to CVS +But as I did the latency testing got better and better +In other words if the chunk size in the is smaller than the attachment size then it works +thanks henry! +Should be reviewed +That looks much cleaner +Committed to trunk +Im not sure why it was not part of the interface +From the stack trace it seams to me that spring trys to lookup but there is only defined when using Java enums +Only onblur event works for the filterEvent richcolumn components attribute +I have this issue +It works for both +Since its just fixing tests Ill commit this into all branches +I had the impression that the patch you uploaded was a development patch of mine that somehow i got it on some older issue and you were kind enough to relocate it for me +This one should be right +The repro worked on v without for the on my comments NarayananI have checked that the reported issue has been fixedPatch v committed as revsion +i will move back the loginmodules in that case +real codepoint orderBut this is a compat problem I think +Any updates on this? Weve landed a bunch of stuff lately did any of it fix your problem? I was never able to reproduce it on a completely fresh installation +Either rejected or a dupe +Resolved +Wanted your comment on the fix +May be CASCADE delete in HQL solved by generating ON DELETE CASCADE ddl during schema generation +It is probably an emulator issue +I put a lot of effort into getting up to snuff and almost all of the tests for that reside in itself not in SolrIll bring the tests from into Solr +Verified on +why do we supposed to include this folder +I think its better to require the timeout value to be positive eg Then you dont need that special if its check etc. +However I don think this is the case any more plus we want to avoid the case of least surprise so this might be changed +So whats the solution +Basically I want to refer the reader to the JVM specification specific to hisher version of the JVM +Is anyone planning on exercising a +Not sure if this was ever done or if we would still want to do it +thanks Brett for the patch +this patch fixes the issue +Updating the code to align with characters per line standard +What should happen is debatable here due to a of we cant do anything aggressive here because things were printing out happily if they were under width print +Not stress tested but heres an example of what I believe should work +It seems to check fast enough to leave in the default build +New version of the patch brought up to date with current trunk +so the cfmail tag uses the username and password of this account +thx for fixing it +Darran ping Tomaz Cerar about this as hes doing some refactoring now in web and subsystems both of which are big SSL config use cases. +Are you sure this bug wasnt just introduced by changes to the composite router after was released +done +Basically my heuristic was if I cant open or get to the file and theres another one thats available Ill dump it +No problem Ivan +I will fix it tonight +well let the folks who are involved figure out what they want to do with this +You want to move it outside of Cocoon or something +I was able to reproduce it and Ill try Tims suggestion +With those all references to classes not part of the client profile will be hiddenThe official build system of lognet is and this automatically sets those defines as needed +Lets get this in for. +moving this bug to erroneous the wrong issue as duplicate reopening it +Can we have confirmation that fixes are in place for +These changes look pretty good +The only way to start the Broker appears to be to restart the ESB +View proposal screensot was attached +I havent had time to thoroughly test it yet due to trouble with the testing environment where I was working on this but I wanted to make sure I didnt lose the patch since I had already put it together +I didnt expect that +Thanks Mahadev for pointing it out +I created a separate clean up task thanks Stan. +michi what do you think +Closing. +Ive attached a nd patch for this issue +And even if thats possible I wonder how I woudl distinguish between scenarios where a Java JVM has the bug and another one does not +We contribute featuresbug fixes to the ODE project +on to trunkThanks for the patch AdityaThanks for the review Stack. +Please verify +With our use case being able to read and write ORC files outside of Hive is very desirable +Daniel over at Mozilla is going to try it out +A more sophisticated fix would lock on the cache name so that distinct caches can be started in paralell but startup requests for same cache are serialized +I cant see any problem with it or any reason that this is a blocker +No I dont +Moving these issues to as I dont think well get them for +The refactoring actually was happened months ago +Makes sense +Ill consider this for Spring news on thisIt would be really nice if Spring could support this as nnotations is very fast clean and flexible way to express your model and its I guess a lot of people would try this approach and expect the Spring framework to support this +Tested! +Im guessing that the deployer does not have access to the repository when processing a class in sharedclasses +Better now then when distributed grouping is released +Would you elaborate on the reason why DEFAULTMAXOPSIZE is increased to MB +ill upload a patch for and branch as soon as gets committed to those branches +We decided the implementation is going to reside entirely in repoMoreover this also encompasses integration with standalone Undertow since that is going to be bundled as minimal instanceSee +The fix has been committed to trunk and so Im closing the issue +Right now I only added the code for jetty +committed +Problem goes away if I make the weddav wagon a dependency of the site plugin instead of an extension +Thanks Dominique +And Data is overloaded and Frame might be used instead +Moreover if there is no printer set up on localhost then there will be user friendly exceptions thrown on the routeIf you have any ideasthoughts on this front or altenative options I would be happy to implement themPlease let me knowCheersAshwin +Im not sure how I missed the updates Ill pass it by the PMC and get back shortly +using immediatetrue is not a solution we want to validate the entered date +Call to have been delegated to method that catch the change the inner exception stack trace and we should use this method in another places in the codeSame can be done with in method stack trace string is now modified to include the information however this is not present in the collection of +So I supose that is a with reservations +there are no local hbase files +Will be partially available in but the definitive fix is due for the next version +patch summary using to setup port in thanks! +Just found the temp override set the system properties have merged this in master and created a PR +The current working dir for each task is the task directory for streaming all the contents of are symlinked into this current working directory for the tasksThis patch depends upon +this makes using properties set at parent levels rather clunky +I will start another Jira issue to work on the xml property file idea +Thanks for the reviewI will run some tests and then commit the patch on the trunk + +Thanks Jeff going back to the subject of the bug should the model property be set when the view is the default? If not then just close this issue but if it should then I guess it still needs fixing though it can be worked around as you suggest +Closed after release. +Changes both the class to adhere to the valid modes and window states and the wsrp test portlet to behave more defensively in regards to supported modes and state +Verified in ER +If you finish it before we release beta please correct the versionThanks +For getcolumnssince string columnPath should be columnFamily +Its due to only have entries by default and ipv would got them all leave no space for ipv +Feel free to attach in this JiraDuring this time I resume my SMPP mock and add the SMPP BC in the SMX distributionThanksRegards has been included in SMXI make the change the include in next SMX release +extra child node not needed agreed. +Instead copy the inside the into your and then create a new property containing all the packages you need for JDK +Hence the resolution is Incomplete Description not done +mapping multiple parts with the same content id to a collection parameter bound to that content id with Multipart annotation is also supported +Thanks for the spotting catching and solving +Manually tested! + +bulk defer to defer to +Closing issues resolved in released versions +Hi JorgDoes your last post indicate that we could now use with say IBMs J VM for JME environmentsCheersAndrew +Thanks again +Someone will take a look +Right now they fail in a predictable way if you dont have bash installed so theres no confusion +It is an issuebug tracking system +That doesnt makes sense +Closing this feature +I am hoping that will preserve the history +I think is tracking this similar issue. +sorry about that + from resolved to closed +Applied patch in revision to the branch in revision . +I guess I was too fast on the checkin there +These are the screen shots for and CXFThe same service fails when using +Would it be possible to forward to platform with the mode parameter in the request URL +The method is public because some non abstractdatastore implementations were making use of this class is there any way a could help us out here? Some way to clean up after the +canceling patch to get Hudsons Hudson +Bulk of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues this as Wont fix since the fix is checked in to MCFs source tree +This is working as designed by IE +So its ok for me that you closed this issue. +The Roo pizzashop sample fails to compile with the version with either JDK or +Thomas and I discussed how to get this working and we decided that rather than sleeping we should retry the hierarchy mounts in the tests +I think you should handle the overall tasks case as well +This is wrecking havoc with our setup as well +I also did testing on and jetty build there are no problem either +Ok so now I resolve this. +Modify the files under managementsrcmainresourcesplatformconfiguration to add entry to +We will return t this issue after August release +Replaced the event delivery with a plain solution. +This also included updating the interop reference guides to match and the changing some of the example build files to remove duplicate just need to port this to trunk +is what unearthed this issue +Hibernate final should resolve this issue it resolved in rc? +ThanksVincent +Thanks for another patch Matt +Actual implementation for the above mentioned methods +I got a query that works correctly now +Please include the link author name and date next time please +Bulk closing stale resolved issues +Changes to the DAS company sample required by with revision +attach a patch to fix this issuethe patch is quite simple just force index creation when the ledger is evicted from open ledgers list +trying hudson out agian after all the changes went in +Nice work Binglin! Some initial comments +bq +Addresses Haris concerns from +I have avoided patching any upstream files +Closed after release. +I actually prefer for the components not to dissappear for consistencys sake +in compliance +so I sympathize. +I guess I can close this issue now +One obstacle to this feature is the configuration file of a service should be packed in the Java project containing service be verified in gatein master and reopened if still valid +The gwt works perfectly with grails and gwtMay be something has changed in the way that grails run? +As of Friday Oct problem seems to be fixed on Chrome +Junit tests will certainly help here +The trivial optimization in this case is thatif bid is an object that I simply transpose the object into the newstream +This issue was migrated to +Please verify and close. + +Fixing +Alright committed thanks +Resolved via r for both and releases. + is used if nativefalse attribute is specified for web subsystem element +But changing them will be part of another task. +Here it is +Passing successfuly +Thanks everybody. +However I have not been able to determine where this object gets defined +In the meantime a brief README would be cool to find out what the current feature set looks like and in how far it compares to other stores integrations +Looks like this is a problem caused by a combination of and enabling Native Compression +Ultimately it should never send the +Given the size of the dependencies is such a small percentage of the current size that we dont want to support the as actually pluggable and that it would be nice that hdfs was supported out of the box just as local filesystem I dont see being a contrib being much of a win +This method currently creates a using the passed typeId +Pushed the to trunk +This was actually just recently handled for inline segment cases in but it does not solve the issue for normal cases +The same paragraph describing render parameters also statesThe must not propagate parameters received in an action resource or event request to subsequent render requests of the Event provider must take care of hiding these parameters before firing the event for processing which is only done currently in the rendering phaseThere is already made wrapper for rendering purposes that hides the query parameters +hadoop dfs mkdir userbar creates a directory userbar and you cant deleted user as rmr expands the globSpecial characters like is really problematic even JIRA cannot display it correctlyAgain we need to escapehadoop dfs mkdir userbar creates a directory userbar and you cant deleted user as rmr expands the dont we declare not to allow those special charactersIt will simplify things a lot and make everybodys life easierI dont think that causes serious limitation +The dependency has moved to the next release and so this issue +So we will introduce a quickShutdown method to use in these cases +Im not sure about this patch +Thanks Sean the patch was applied to BTI branch at rPlease check that the patch was applied as you expected +That way you could have a central meta data text extracting server without needing to install java tika on all the hosts where you might need it in a replicated CMS environmentSo the scenario would be that a CMS trys to extract text meta data from a file but does not have a local tika at hand +isnt a bug +Just tried this out and I couldnt find a way to navigate to a different drive +There is a object member relatedURI to trace the relative path +The only difference is comma after day of weekThank you very much OlegSo it is configurable nothing to fix +the behaviors from the profile are then passed down via normal POM inheritance +all hudson accounts all have privileges to publish any snapshots its only the releases that are secured by project role +The log is completed and follow up tasks is already in jira. + +Less than week to pushing off more items to +I dont think the test failures are related to my changes +Learning more about Hama so that the algorithm can be implemented using Hama graph +backport +Stephen I need this for the SCM system Im working on so Ill probably be willing to contribute with this feature +Sorry I was wrong thats impossible +May have been resolved by +This occurs because the TCCL is wrong when +When would the input files be deleted? As soon as the last entry is consumed? What if the merger is dropped before all inputs are consumedPerhaps the iterator or Sorter should instead have a method that deletes things that can be called in a finally clause around merge iteration +I plan to commit in a day or two. +I will also try and figure out a good way to add some systests around this +for it to be merged in +management server check the minimum memory requirements for server and management server. +There is also the overhead of the proxy itself +Committed at subversion revision +one question why we need to move in both get and flushLedger? I think we only need to move when flushLedgerYou are right Sijie Its needed only in flushledger +Koji we have files with replication factor of +Revision renders the HTML and javascript separately using the mapped to the jst selector and js extension to generate the javascript codeThe page title is the value of the first of these properties thats set on the node being rendered if they are present title name description +Hope to see the as soon as possible +Dave please reopen this bug to attach trunk patch if trunk requires this change as well +Somebody knows some workaround? or somebody knows Which version of this plugin does not have this issue +patch applied thanks Ray +Thanks Gavin +Some naming nits Id prefer to name this just Files or perhaps rather than +This attachment shows an error on the customers ISP this is not a jboss network have them try again with IE +Supercedes the previous patch +Should we mark the log level as ERROR instead for missing agent configuration +For a quick patch this should be enough +I used following procedure +to make the rsync run faster +Attaching +I dont see extend in patch version +Ill work on this issue +The WSDL is contained within the SDK jar files +Check the corresponding Hibernate documentation and the corresponding reports on the Hibernate JIRA for will do +Fixed at apache and in our private felix branch +Seems to work can you see if this matches your expectations? I tested some with this patch some and I like this behaviorbq +If max is then dont ask how many are leftAlso added a junit that tests the case where max is not defined in the config +Would be nice if someone reviewed the patch and provided feedback +I have committed the patch +If the follower sees a commit subsequent to doing the read op but prior to getting the lastzxid in FRP the client will end up getting an invalid lastzxid +Its the tiebreaker in the class which isnt breaking the tie because nanoTime apparently doesnt have a good enough resolution on a MacIve replaced the long with a UUID and it seems to work +Trevor is this still necessary with the update toAlso could you please point me to the hudson job where I can see jbpm running on the SOA platform +Otherwise you wont be able to do a release from the branch. +Thanks Sid +Attachment has been added with description Encryption Level screen in the secured for encryption larger than bits has not been implemented yet +Marcelsounds like a great plan +You canstill commit it I guess +That URL doesnt exist now the website has completely changed and the M tools are still being implemented so Im going to close this +Would you then please comment on this issue with the full exceptionconsole output +Ill watch after VTSVM to understand if this fixes the failure +This was a bit premature by me +Just remove the directory +closing all issues raised by me +According to structure the XS value of the tangle in the screenshot dropped from about to +I can confirm that the test runs successful when the path in the query statement is quotedI think now that we know whats going on we can apply the patch again +The dependencies for jbpm and drools within designer DO NOT MATTER +Doug given that this made change to the protocol I recommend merging this change to +on the proposal +Max please either merge the PR and then resolve or just resolve whichever you prefer. +We build using the maven ipojo plugin +Verified with JBT. +I can integrate them into the main one or leave it as a separate build file +changed component to be solely language design until there is a design to be implemented +Done +Thanks +Fixed by fix for +Clarified in documentation that uses the flat dataset file format You need to provide a dataset for +bulk close of resolved issues. +Well thats fine but we still need more info +Marking LATER pending a course of action +I must disappoint you here morfologik simply compiles a list of triples it has no logic for generating these forms from lexical flags base dictionaries +The patch contains disabled CUD features +Whats a trunk jar from a previous release? Releases are off branches not the trunkAlso why keep any jars around the jars can be from a codeline at the correct revision? In some cases the regression might have been caused by one of three changes included in a set of built jars so the built jars may not always lead to the specific change +It was a while before I noticed the trouble with compactions +Anyone have any issues or feedback on this patch? +The file for new suite is missing in the new combined patch +Fixed the problem for me both on Windows Linux +The UI displayed capacity at usage when this occurred +Does this issue belong in the WAGON project? It looks on the face of it like the jackrabbit DAV wagon provider isnt setting up the host information correctly +thx. +All these ids would be sequential so the max value of the node id would be the same as the total number of nodesFor the new node create after startup we can keep allocating node ids using the same counter +Attached the error report +So is there a plan to get it back or create a new table so its not db intensive +it seens that isnt fixed. +Thanks +See the other places where XercesXalan invoke it for examples +This may be addressed by with branch +Thanks for Clarifying Brian +Instead if a request cant be processed for any partition we just set the same error code for each partition in the response +bulk close of all resolved issues as part of closing items. +I look forward to trying it out once I get some things off my plate +Its highly dynamic +So I dont feel it is that critical +NEVER use dates in primary keys +Patch against SVN Trunk for the improvements to the dropdowns +to edit fix version +There is a new group Summary front page but it could use some polish I think in terms of look and feel we can keep this issue open in case someone has improvement ideas on that front page like perhaps a quick summary box above the particular groups or perhaps a two phased front page one that truely is just a summary of all project groups not rendering individual projects at all and then nesting down to the old summary page which would be restricted to just a single project group +Unfortunately this is preventing us from porting our framework to Spring as its the slowdown in deployment is really unproductive when developing on WLS couple of comments from me as we use WL MP wno issue +Thus thats the reason why they have no shown up in the previous releasesThose files will be in the next release +Some notes This patch implements a for Avro wrapped Trevni which allows it to be implemented as a separate file format for the existing This patch contains contributions from Srinivas Vemuri derived from uncommitted patches to Haivvreo Avro version has been bumped to to bring in trevni and modules of Avro are now included The test has no corresponding +bq +Attachment bz is a patch proposed by Pavel by Committed revision . + for the patch +Fixed as of rev +but Ill check that out +Thanks for that info and your continued work on this BryanAs for the implementation strategy of such an improvement hash lookupis interesting +Removing me from Assignee because I dont have an environment now +Fixed in revision resolved issues to so we can generate release notes +I worked on Eclipse while creating the patch +Yours is not complete and will break some things +Are you sure to use it? We want everything in the is accurate something that is crucial to correct operation +runs cleanly but I havent run any other tests on the patch yet +A separate problem is that permitting a blank breaks the WFS Specification +I committed this. +Attached the test +Since Schema compilation is the most expensive operation the schemas should reside in a cache that can be reused when necessary +I attached a patch file to enable the extended entity to be able to override the field +Certainly Seam cannot mirror every piece of JBPM functionality and at some point if the user needs to do something really tricky they should probably be doing it themselves +Yeah my design will support reading with one key and writing with another +Tested! +This is not reflected in the patch file so I am attaching them separetely +Done +Ah missed that one +I merged this change to. +Let me look into the to make itI dont see the reason to have two different sentinel conditions why not just use NOROW in both casesThis is because the expected key is sorted +bulk move of all unresolved issues from to +Development of the old maintenance code was frozen long ago so this bug will not be fixed there +It needs to be a seperate patch completely because of the noise issue inthe CVS diff log +The test case in question has been moved to a comment of the corresponding issue +It doesnt show what methods would handle the request +So it makes sense not to show any error message in case of blank name field and wait for user to start typing but in our case we have a default file name and we should validate if it already exists +The current patch causes to fail +fixed in for a while +We can improve on the warn message saying that your checksums wont be verified and you might get corrupt data if that setting is enabled +JamesYes we need to share these code in a common http module to avoid updating the code in and modules +No easier to add constants and expand the array +What about creating our own exceptions which extends and wrap throws exceptions +I think the variables must be tied to the process not within seam +I confirm that suexec has been removed from the httpd package in the EWS RHEL channelVerified on +properties files lying around +I thought I had resolved it +Thanks Sharad! +Run mvn clean install jettyrunThen using SOAPUI or another client take the request from and send it to the service at URLhttplocalhost like this was fixed in JAXB do you have a CXF target release numberIts fixed in the JAXB sources but there hasnt yet been a release of it +Thanks for the heads up +No +Verified in CR. +HiYes is a function i wrote however is unrelated to the issue that was being experiencedCheersJames +Hi Brian I followed your instructions and removed the beanb The bean should be removed from deploycluster +Netstat shows that the packet is already at the datanode side where port DD is the datanode and CC is the client +We fell into this trap for a year or two in development and it only came back to bite us hard later +Works for me +I assume yes +for this patch +There are a couple of choices in fixing that which would also solve the +moving to not resolved moving to the Portal team requesting an update on the status of the functionality to have our screencasts up on the PortalI have also asked if they would be able to keep us through this JIRA +I will be out of the office from Wednesday and returning Monday +I have addressed Sanjays nit and committed this to trunk and +this is fairly important for our ongoing implementation can someone with appropriate karma take steps to get this into? voting for an issue is a good way to increase its junit test for preserveOriginal is added to this patch +Fixed by +If users want to experiment they just need to copypaste a file where is the problemThat its much easier to not have to copy past a fileAt the sizes of the files involved your just being a masochist +It doesnt change anything +I am going to commit this unless there are objections +Also the diff is done under core package +Taking a look for it sounds like youve worked out a good test case for this issue +FYI +Then again his comment above might predate that +pending patch to get build going +It was Colt that had a portion which was licensed acceptablyAre you saying these errors occur before you change? I dont see these failures in headThe first error cant tell you why it happens but can explain it more if thats what youre asking +Committed +I had to delete another instance of in utils does that make sense? I admit I still had patch problems here but thought I manually fixed them allI can handle issues like copyright headers +I will close this issue as resolved in hours if there are no objections under lazy consensus rules +closing as duplicate of we bump the other issue from improvement to bug then? The point is that this is something that will confuse folks who would like to write client code against Solr with tooling. +I was speaking collectively of all of your proposed changes to +Attached the patch for review +The test failure is unrelated and already addressed in +Could you tell me if its solved which classes changed for the solutionIf its not solved when do you expect it will beThank you very much +Change committed +I am unable to reproduce the problem +It doesnt apply to +I think Im not the only one who wont immediately realize that set delimits on all isWhitespace characters so it would be good to add that to the javadocCould you please do that and upload the new class its unit test class as a patchThanks +Paul are you going to pull together another patch here +Blocking the undeploy until httpd has received the information +has been marked as a duplicate of this bug +I was perhaps too influence by Python which doesnt have the ErrorException dichotomy as Java does +We have to wait until Eclipse is final or at least stable enough to start developing for that plattform +Hi guysis there a way to avoid this by keeping jaxbri as databinding providerIve tried snapshot but it is not a working solution yetBest regards Nicu +Simple patch capturing above changes +Have any Seam developers looked at this issue yet? Ive very keen to see this feature included in Seam slink +The crawl succeeded with no problemsIf there are no further objections Id like to commit this patch with these changes within a day or two +bq +STRICTSUCCESSIVESPACEENCODINGRIMODE would sound to confusing not that it might confuse me but it allows one to reuse this configuration for other non ri compatible stuff too +fixed in svn r by applying bertrands patchthanks for the patch +Go ahead with checkin Nicolas +But we should flag it as incompatible change so that upon upgrading folks will notice the release note youve provided +Ready to be released for +shale builder has been removed from the VPE templates +Thank you Emmanuel but you didnt close the issue +This behavior is correct see the earlier commentsWith an unwrapped array an array containing one null element and a array that is a null reference are ambiguous +How about splitting method testPositive in two test methods test and test? I propose to put pc instances into the collection used as query candidate collection +Bulk closing stale resolved issues +Taking the bug I have a test case and Im working on a generic solution for allfile based configurations +Getting closerThe issue is caused somewhere in the methods if they are modified so that the dummy indexes that are empty are never created and the Wrap methods return immediately just wrapping the arguments and returning them then everything is goodSo no issue in the codebase just in the conditions used to set up the test +As a result we are closingthis bugIf you can reproduce this bug against a currently maintained version of EJB please feel free to reopen this bug against that versionThank you for reporting this bug and we are sorry it could not be fixed. +I just finished reviewing the patch +According to docs get retrieves the current result as an update count if there are multiple results returned from the query or returns if there is only one or no more results available +sharding only starts to show when sharding is actually done at a very deep path level which I think is quite unlikely +What do you think +Changed asserts as Nicholas proposed +Dont you get the same behavior +Running script to test +class deprecated but still works the same +by the way I assume that the tested version is the one in the sandbox +The content assist processor is completing against an offset corresponding to after the close paren but the source location of the inside of the parens has an end position that is inside of the parensSo the processor should find the AST node inside of the parens but doesnt and so completes against Object or String +That calls the listeners without holding a lock any more + +I was too having this issue using and it was doing what Steve Tekell should be checked +Close. +Bulk close after release of +I have a couple of integrations Im working on that require this feature +Otherwise this appears to be a rather oddity overloading signal processing with a dummy zero signal for EXIT and I dont see a good reason to implement it in submitting rubyspecs for this behavior would make me more likely to consider adding it. +removed forbid from the result sets seen in earlier patch versionsRerunning regressions +Need more details. +Hello JacquesYou can close the issue i will provide another patch if it breaks something for someoneI tagged the issue as a bug because returning obsolete prices was creating bug in display when you present the different prices by quantity for a product +Sorry to chip in so late but when I was looking at this class I noticed a FIXME that states that it would be a good idea to use the same component that is used for the and XSP logicsheetIve started working on this and it seemed rather straightforward but unfortunately something crashed and I lost the modified code before I was able to properly finish and test itSo how about killing two birds with one stone and see if the class can be improved by using the component as well as fixing this bug? Maybe using the component already fixes this bug +Is there anything I can do to detect this case and retry? or work around it + Username chrigri +I do not want to put unnecessary efforts into refactoring a lot code if it is bound to be replaced anyways +At least holes would not go silently +Im getting ready for a week vacation and will not have email but would love to continue this dialog and am very open to your suggestions below +Thanks tom I just commited this to trunk and + +Dhruba whats the implication of a pluggable block placement policy to balancer and blocks? Currently they all assumes there is only one replication policy +Thanks Patrick. +You should use ejbSelect instead +Ill first writ a parser then Ill see how I am able to integrate the extracted data in the site +Updated patch to one that applies cleanly +This should be handled by MR +Unfortunately i dnt have the repro case but it is just that we are executing sql statement to update a row in db and it is failing with the error mentionedWith other DB like oracle i am able to execute the statementThanks for the update VinayIm afraid it might be difficult for others to reproduce this +If youre some developer you might not have the resources to have that dedicated cluster +When I pick the block for replication we will increment that index +bq +However we cannot make the other flush messages OOB as they need to be received in orderVladimir can you go over your design and make sure changing to OOB doesnt cause any bad behavior +Im guessing that the SOA branch replaced the community release with an EAP build which might have been the latest snapshot available at the time the SOA branch was set up +Can you possibly reduce the project down to a failing test case and attach it as a project? At this point Ill need to get to further investigate the issue +PATCH Allow binshell script to process several commands in batch mode using b or batch switch +hold on +Thanks for patch Aravind. +The test be corrected to change their test config +I just committed this. +I fixed the issue on the branch and marked the issue as solved for. +Ill take a look at least to figure out what the problem is +Synced it patch is recommended for those running a stock release +Committed branch and trunk +If you beleive that should be please and contact the program team. +Can you try again with the latest trunk? We fixed some build issues recently +Ok +Other minor cleanup +Here is a patch for +Hey Jesse is this one still in progress on your side? If not I can start on this one pretty soon looks like low hanging fruit +I can add it to other perspectives of course +The patch has multiple issues +also should no longer be republished That was a bug caused by it never being and then having the timestamp changed anyway +Added a unit test I had forgotten +There was an issue with the staging repository on Nexus +Have to see how hard that is +Committed v to +Pete is there any easy way to do what Manik suggests? +push it +Revising patchAlso adds a range facet for popularity appreciate if someone could test out this patch on your own data and test various combinations of and gaps to see if it is water to started to take a look at this but ran into exceptions due to youre right +is a class which represents one interface in a network +Its quite difficult to find out why and how to fix it though +jcarousel is commited at r +I have sent made it shows the pom dependency and added a button to apply the dependecies to the +fixed merge it seems the patch got stale again +Both from a UI and a code perspective +But should we be bothered about the degenerate case of having to write index files on disk when the memory limit for the index files is exhausted +Some other changes sneaked into the last patch attaching corrected version. +Evidence of several established open source projects called Rat ranked highly by search engines +Patch to fix this +Cancellable should hardly do anything other than flipping a flag +I understand your point of view but what I like with the is its extensibility If for some reason one parser doesnt work as we expect we can easily replace it +The patch looks good +I was hoping you would have some +totalDigits suffered from some similar issues and these have been fixed in the same checkinPassing to Rajiv for test. +Hmm +Its nice to see you back using Search Ive seen your name often in the oldest Search sources +I have test cases for these components +I just ran numerous examples and this is more or less the same as so I am closing it +Update the patch to remove public for +Can not reproduce as is. +Fixed in +This issue is fixed and released as part of release +Does this make sense +The history would useful I think for seeing all the changes and associated rationale +Thanks Bartosz I will reassign this back to me for verification with build which I believe will come with fix for Notes item not required as this affects internal Red Hat testing only. +Already addressed in web flow do you mean already addressed in web flow docs? Are there some different docs for which are not yet published? The shortcoming I described relates to the current reference documentation available at Spring site +Added Justin and Mauro as watches to this jira afaik the are the LDAP integration experts +I am looking at doing a retrospective update to the database to fix old values +branch revision +Here is pom change I made. +The manifest in the cached bundlefile now appears to be correct +Requires some thought and possibly an interceptor plugin +Agreed with Snjezana +Example applied in r +The enabledisable capability is removed which means that the stats will always be running in jetty potentially impacting the system performance +Applying the changes from this patch related to the security component Changes to the Addition of Much more elegant way to handle the policy coordination. + +Im also available to help with this task if needed +Hi DaveI just double checked and it seems to me that build should contain the fix +Assigned +Committed documentation to the +The changes to the unit test is included in +The war works under Tomcat and JettyUnder Geronimo it does not work +Applied the patch +You would think but installing rc and then upgrading to works the other way around gives an error +Difference is that in Geronimo they are in a archive obviously and unpacked in the Unit testI cant post the +Actually it is not even that it is not needed at all this step since we just need to tell hibernate to update its schema at deploy time to achieve the same +I take that back for +On the second page there is a validation that prevents user from creating a new project +committed. +All fine now +Having a transaction manager capable of working with multiple Hibernate session factories would keep a lot of users from having to use JTA and its associated container dependancies +Olaf re could you add similar comments to this JIRA in the release notes section? Well need that when doing the release itself +Work effort cascades the calendar style sheet which restyles basic HTML elements + branch no more under active development +Hi Dmitriy I like your idea +Although not exactly the bug described in the fix for will also fix this bug +Committed revision I also plan to backport the fix to +This was due to not having tested on server state test for me +see +Another unresolved in keys +Jep youre right we have to refactor a little bit +It merges the formatters into the converters +resolved on change of shipped and resolved. +This turned out to be a much smaller change then I originally thought +Should just be a pure server side issue +Thanks Vijay. +Would it be possible to get an svn diff format version of the changes +The versions appear to be identical +Is it some kind of an annotation or text box that appears as a descriptive element on the graphical canvas +Thanks guys +I echo Hoss commentproximity searching is important even if it isnt used much directly by usersHmm I agree with you and Hoss especially in applications where proximity is used to rank results of OR queries +We could require that producer container classes require an annotation which would help but now were making it even more complicated for the users +Would it make sense in this case to create it first in the sandbox +Theres no major benefit to add new enhancements to the logic tag library +Thanks Carl committed +bulk close of all resolved issues. +Itll make it due to the respinI committed to trunk +HiCan you please point me the url where I can download the new versionThanksArvind +Set component to NOT use SVN until after the commits mailing list is establishedFor future reference SVN should not be setup for a project prior to the the commits mailing lists being established. +tcunninglocalhostrosetta svn commitSending rosettawarresources file data +Target for so we make sure to talk about it at the hackathon +What version of the surefire plugin are you using + +Thank you for your commentI will close and a patch for removing +Im concluding the dbf file is broken probably can be properly read only with the software that generated it +Latest patch available +Some configurations were missing because of a incomplete example +The patch is ready for review + +I thought Sam said this did not reoccur in beta +Latest patch addresses the comments from Nicholas +This issue has been resolved for over a year with no further movement +I think someone has the wrong jira tagged in their commit or Jenkins has gone insane. +Hey Hairong Dhruba +I use top command to monitor the memory status when the program is running +UpdateNew function if only one label is neededSomeone an idea where to put the java methods any ideas? Im not sure if is the right place +Script I wrote to wrap Alexs program and give the results of different proposals in different cluster settings +Thanks Koji! +Also a unit test would be good to verify which Exception was thrownThanks! think this sounds reasonable +The second model will be garbage collected after the first use +Another benefit would be it allows introducing additional methods in the future without the need to enumerate all of them in +I dont really see how an entity in a stateless scope would workI think you want a dependent scope for this +However a little testing based on the supplied code indicates that it is execution of a closure in the binding that leads to the Gant instance not being reclaimable +So you are sending constant bytes for each method callI am planning to send the ticket only once per connection irrespective of how we implement it +Seems like this has been fixed +Done at +this makes sense +Just before committing I recognized that this patch breaks quite some tests of our XML test suite so we have to search for another solution +Tim thanks for applying my patch it was done as expected +Applied patch into trunk and branches with thanks to Aki. +All guides were updated recently and should be working now +In the future patches should include the path starting from asn +Well the doc says overrule the name of the actual class generatedThats not what I want to do +Dennis yup that is exactly what works +See my log try the latest build should be ill try again +Im not that familiar w JARJAR but if you want to try it and we can compare +Then add the site goal later +The patch must be applied to the class orgapachemavencontinuumwebaction of the component for the improved patch +Apache logcxx released April +hi Jacopo thanks for looking at it it looks like that is was caused by an error you fixed previouslyhans +Kathey I am trying to see if I can rewrite my code using getOffsetsetOffset rather than previous on +Ive committed this to trunk and +I just put a reference to in my simple Rails app and I get an uninitialized constant error in both MRI and when in I put the constant after block it can see the constant in both this error is what it is supposed to be? +Closing tickets +Attached is the release note +In particular it seems likely fixes to get this to work will conflict with the hopefully superior way of doing things in the branch +Likely cause the shp file itself +I wonder if I have unwittingly been testing the default storage factory with the converted test +offering a different default value is possible +Great news thanks for quick m closing this issue as duplicate of then. +Hey Richard the patch process does not seem to be configured properly for so theres no hand over to QA phase +Is it valid for mixed mode too? If not is there other way to get deployment URL in mixed mode +Damn even the base classes of Java dont respect the naming conventions +bulk move from M to be done even for current wizard but it is not save changes assuming we have CR and han to M +The info is deleted. +I hope that commiters will do that if they found it valuable +ping post posted website information awaits the restoration of svn access and this list may be closed now +Have you been looking for this phrase? p +Please look in history for what the actual resolution was. +Are you monitoring that server with JON +Committed a first version we dont have any useful UI yet but at least we can build a war in our build +I didnt put corresponding remove code into the source but I think it would be changes to updatepersistIts really an error if the instance is null on an update or persist and should be treated as s often useful to know where to insert code for pre or especially when some people on the team dont initially know where to put the code +This results in the removal being sent cluster wide on commit time +Thanks Kihwal! For the new patch do we need to declare check as volatile since it can be set and retrieved by different threads +If we flatten all iterators and settings to a single znode we can guarantee everything is in sync but then theres the race condition in the readmodifywrite but that can be handled by stat versions +Remote Server To Local Guide Me on Forums will migrate to JPA since version +I have added the virtual keyboard in Paid In and Paid Out screens as requested initially +Changing to Blocker urgency +Move into post +role stringAgreed nice improvementAttaching new patch +Do you think it makes sense to include the actual above that triggered the bug in the unitWe should verify the patch to make sure it fixes the real bug +Applied the patch with thanks to Aki and I also fixed some similar issues on the other cxf blueprint parser +Was just as I said a wild idea +The new test you added has never succeeded under Maven Thats incorrect the new test has never succeeded under Maven on branchx +bulk move of all unresolved issues from to +Thanks for thepatch +is behaving exactly as expected +As well as a check for an empty name before and after the colon +know how to solve this reference implementation and myfaces cannt contain in lib at the same time. +And there would be always the same problem The implementation must be able to detect by an environment which might not be the same object which was used for the last context wants to connect to a previous used cluster with the same options like encryption username etc +Even whatever is the max varchar now is not going to be a good index with each page of an index tree having as few as nodes +Vaibhav Please change the status of the ticket to patch available when youre ready to have something reviewed +meclipse team does not plan to maintain. +Good call Julien +Please find the new attached patch +we can fix the remaining tests to use this stuff as we get around to them +I can only suspect and not know since Cedric already proofed that the information is there only not picked up by the debugger +I just imported the number guess project into eclipse deployed to a tomcat instance and modified a class making WTP redeploy the app +Moved to upon release ofAlso moved from fixVersion to targetVersion since it isnt fixed yet. +Committed to trunk with revision +Fixed on master in de and on branch in b. +Theres definitely no need to create an other jira issue for this especially as its an internal change not visible to users +I was not able to find a way how to display all downloadable runtimes in GUI +Users can always clean up Event Log when they feel event log is too long +Switchyard components like BPM and Rules require process sessionId or correlationKey context properties to be set by the client in order for the components to function properly in a unit test it is possible to do so as Eduardo points out you have to write quite a bit of low level code to accomplish what on the surface seems to be a simple need to set a property. +We need the ability to changeresolve promotion on +There are a lot of interesting ideas here but I dont see a pressing need for this since the actual functionality is supported with the existing components +I tested with EA but it failed because module is looking for I tried the patched modules but it seems that Grammar or Core module is looking for and does not support in trunk updated the autoupdate server and the zip file +This is because from an SQL perspective the WHERE clause would follow a SELECT clause and so will confuse user if the second format is introduced +See also for a similar issue on +Fixed at revision +We did not see the sender to hang or crash during our test +This issue duplicated so it should be discussed there instead +bq +In case if it removes that directory too then how would you determine whether the logs are in + rc the wire log it says for the version I just grabbed the latest from CVS and tried it and it still does the infinite loop +I have recycled this issue and changed the title +This are diffs in the unit tests +I just realized I got distracted and forgot to verify that things were compiling correctly +However this API is heavily used and would require changes in all user code so I would not do it a few days before a release. +The problem with the accents is fixed andnow we completely ignore tested +If you have an approach I wouldnt mind hearing more about it perhaps we can work towards something of mutual benefit +Ive applied the patch and deployed a new +If this attribute is not specified it defaults to Could you please test this and provide us some feedback? I dont have such a repository to test it myself +Castor should probably display a warning to the user when they use java keywords +Thanks for the update and explanation Galder! I created an upgrade Jira to track what we need to pull in from Infinispan +This issue is fixed and released as part of release +marking as resolved as there for beta should not be any technical reference of any struts plugin or are just leftovers needs to be done too but that are invisible to users +It was working fine just slow +should we close this as good enough for now +I committed this to the branch +The Seam version in EAP is Seam +Integrated the patch to trunk and +Ill see if I can find a Windows platform and try it unless somebody beats me to it +attach a patch to fix this noise error messages +I am attempting to build on a system but the patch sent by Feng does not seem work on a build +You need to run yourself manuallyI looked at the patch and it looks OK for the most part +We can close it if I find conclusive data will reopen this bug. +It doesnt matter if a user has invalid translator properties defined +Patch apologies its not committed yet +Weve decided to do this by allowing actions in navigation rules +Yes you seems to be right +This bug has been marked as a duplicate of +This is due to some new analysis in Findbugs +pending Sandy +I add a patch which can resolve this problem +schedules both block replication and deletion in one iteration +Is there a small corrupt index can reproduce the problem by Thanks +maybe you should consider to allow absolute paths for the pom file +I will start looking at the failing tests +Kien The last patch just removes the ERROR logs which is not a good thingthe fact we do not currently support the apirpc feature in gadget container so what we need to do is just to remove it from the supported feature listI attached a new patch to remove the feature from the support do I need to also apply the rest of the patch? static final String JSONRESPONSEWRAPPERELEMENT data static final String JSONRESPONSEWRAPPERELEMENT result +Ive looked into merging the patch with the current state of the source specifically but some changes notably revision for which make the patch invalid +Seems my last comment was a false alarm +Ive taken out a lot of extraneous detail from the attached DB without changing the behaviour +Here are the attachments mentioned earlierand + +Thanks Binglin +Please and get back to I changed to simple PK and it worked so I suppose the issue was due to composite primary thanks for your time +poms updated to but cannot proceed w job renaming until we know what the next version will be +I dont see why it cant be thrown twiseYour nd point its quite interresting doe that scenario ment the statement was never invalid +hehe lessen learned thanks jacques for commit +Attached patches aboveThe patch basically makes use instead of since it does not support onmouseover and onmouseout according to anyway + +Somehow they assume that the strings are always interned +The problem is the way the url is composed not the JIRA to be working properly in the version currently in svn this sounds like a duplicate of. +I fixed a couple of the critical ones and there seems to be a regression +This patch contains changes to filesThe file is no longer used and should be deleted +That response might have been intended for Cyril though +Moving to +Import autocompletion has been added +we are getting really different behaviors here +Attach my patchCan anyone review my code? Thanks in advance +Guyscould you please provide a unit test +Yes I believe you are correct +Actually I resolved this problem simply extending the +Fixed +RXTX is LGPLd with an exception of linking +When i am creating cookies then i am either not able to retrieve them or the maxAge is set to which is wrong +I see it now thanks Patch updated +Pending to trunk +Thanks Knut for the clarification + +Could someone with the necessary infrajira karma please this issue to something likeAll wikis have useless for users with www browsers that specify a language preference +So I need to download and install AS +This sounds like a classloader issue possibly multiple jBPM jars in deployment +Will confirm shortly +I got that fixed nowIt was a refactor that didnt look in the sftp file +Im a fan of Logj but I love something is simple compact and no more dependencies like JDK loggingBy this it allows end user to choose their desired logging framework at deployment time +Then we turn the tables back into modeI guess somewhere in there we should remember what tables where disabled and where in mode before the snapshot started +Yes +Done +Looks like this one may just complain that Lucene is needed +Ill check those next week +In order to fix this we need a unit test that works against the trunk +Ok ya definitely not getting the deviceready event to fire +this patch is running in the code base as i remember +From Atlassians support reply activity streams in JIRA are not compatible with Confluence +Steve I applied the patch from and all tests ran successfully +to is a situation in which bat uses the Xerces and Xalan jar files that come with FOP when using SDK +Also fstat +Apparently the Site Plugins class path leaks completely into the reports which probably should also be fixed sometime +This bug was fixed in the trunk in January +Second patch attempt +Ive checked whether this problem is present also in jboss and it is althought the package has to be configured slightly differently +In branch we need to wait for whats going to happen with conversation in the OASIS spec. + +We just ran into this in one of our test clusters +Moving new features to the next version +The main console still does some deployment which I plan to move to the new plugin +Manually tested +Werner how am I supposed to build then +I actually forgot that casting was something allowed +Fixed +Attached is the source code found in the branch in the github repo +Thanks +Daniel your proposal looks Not quite sure what you are saying +I simply called it from the JEE Eclipse Web Services Explorer +Actually cannot be considered API stable as its underlying library mimej is not yet API stable +Closing issue it has been resolved for a long time. +And the development and testing of an Netty based DN will probably take on the order of months anyhow +Thanks for the quick turn around Jessebq +Sure +Extract attachment to your webroot and call to see for input +Sure It should not be too difficult +For example perhaps some existing code is depending on the current failure behavior to take a certain course of action. +Jan In this the change is making JNDI provider and Encryption Util pluggablethrough a factory +Thank you think this would be a perfect fit for the sandbox project! +However we waiting on some resources to volunteer and help us maintain modmbox +for the old behavior + merge this to branch sorry +Attachment has been added with description Testcases for status from ASSIGNED to NEW for consistency +The fix was missing in master +Unless there is an updated patch posted soon I wont be able to include this in +whether good or not jdo isnt part of jee so its not really appropriate to include the spec in geronimo +Few days ago I requested the hook to builds but the request has not been handled yet +Yes it should be picked from +With vfszip impl this test passes. +DONE +Incidentally right now the lifecycle has as suggested a RUNNING and a READY state +waf after removing those entries as well +Eg +Add the activemq module in the distribution add a configuration file in conf folder +Verified in JBDS Build id +mapping removed. +I think the doc patch is ready to be committed tooCode patch committed with revision Doc patch committed with revision Thanks to Anders for fixing the code and updating the documentation +what would having tools in a module do for usWondering more what this means +bq +Hey Mahadev Seems like some people are hitting this bug in did you mean not to commit this to? If not Ill go ahead and commit this there +is this issue still valid after Leif fixed on master +Assuming it does will this satisfy s requirements? Besides the Hive do you need access to anything else like configuration files or shell scriptsPS Can you ask the INFRA folks to turn on wiki markup for the JIRA group +thanks am unable to find the dev list discussion on this topic +Fixed in the issue now that it has been resolved. +Assigned to you +usrlocallibbin port Hadoop is installed adding hadoop classpath to falcon classpathFalcon started using hadoop version Hadoop I have now two instances running in the system pts Sl java false usrlocalliblogsdata usrlocallib usrlocallibconf cpusrlocallibconfusrlocallibserverwebappfalconclassesusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfa pts Sl java false usrlocalliblogsdata usrlocallib usrlocallibconf cpusrlocallibconfusrlocallibserverwebappfalconclassesusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebappfalconlibusrlocallibserverwebapp problem exists with script also. +Forrest doesnt accept relnofollow +Can you verify whether you can apply it to r +It should be predictable so I guess there is a bug if you have seen a different order while your dependencies have not changed +But your making the wrong assumption Same Datasources! Same Source Models but wanting same behavior +For better or worse cayennemain is a GB svn tree so each iteration takes several hours and the failures are largely at the end of that +Regarding the collectedHits if there are docs as results and if we want to return only bottom by giving appropriate pageScore and pageDoc the expected result is to return only docs as results +OK? Thanks! +Furthermore were consistently delegating to Springs abstraction now using Java reflection for existing Class references and ASM for plain class would be great if you could give this yet another try in a dm Server environment +v patch with fixes to missing options explanation +Thanks for your time and help here +Patch for yahoo! distribution with a couple of bug fixes in test case +Lets add them all in index page +I had set spellCheck with collate in select requestHandler if I remove it everything works fineMaybe it is the spellCheck that causes the problem with the firstSearcher +i can understand why this is needed for selected mailboxes but wonder whether it might be possible to use something lighter for methods which act on unselected mailboxes +By simply removing the guards on and elsewhere I now have navigation content assist type inferencing and mark occurrences working +Hi TihoWell the customers concern is just about the appearance +Could you submit in patch format +thank you for looking into this but would it be possible to attach a patch to this issue instead +Can you attache the code here or send me via email? Or tell me the location to check out +but got the same errorMay be my was wrongAbove article shows a for Geronimoso I replaced s with of lib folderDo I need any other modification +Closing all issues fixed previously after a brief review of each. +closed for release +on latest patch +Bringing back in +Please review and if satisfied close this issue +Revised in response to code review +Applied +HiFind attached a zip file containing a simple web test application which triggers the error a mysql dump file to create the mySQL databaseDeploy the zip file in a web application under Eclipse and TomcatThe URL to test is httplocalhosttestWhen you click on the edit button you have the errorIf you cancel the sortnatural in the file the error disappearsThank you in GailI tested with the last version of Hibernate and the error is always present +The both start eclipse as Developer Studio at least about Dialog says so +It shd work +I just created a new wave clicked reply and clicked delete +Generalized to work with any given level between context and sub query +Thanks for looking into that Knut and for the quick +I dont like hard coding the name at all +The fact that the methods are still there is actually an oversight +No +Doing another rev +When a new DTM was generated starting from the synthesized declaration of the xml namespace DOM node identity cant be achieved +Adrianthe screenshot looks great and its pretty different from mineThe main issue is that all the forms in the other components seem to have loosen thir stylesIm attaching you a diff patch of what Im running +System tests reviewed by Robbie and approved for. +has a better solution to fix this kind of issue +Then finalizing will go slower but the will be able to proceed with normal activitiesThe jstack you attached I do not see that is doing any file deletes +Marshall says that the PMC chair should be asked in order to get access +Issue closed with release. +dims +Updated in ESB branch will be included in next merge to platform +I made the change as you suggested and checked that the was property closed in the debugger +Since cloning is removed in changes to in would not be necessary +Well look into that but in a mean time you can use maven launch configuration dialog to specify Java for launching Maven +Hi Andera avoid share model between pages +Looking at the code it seems the repository already exists when the test is startedI would like to verify this really is the caseI dont think this will fix the bug so I leave it open for now +I have migrated this patch for ODE +Thanks for your response Alan +Do we really need the class? Wouldnt be enough to have the class use it to define an appended in the config and make sure the right logger category uses it +AFAIK it does not support unwrapping +They show all bugs of the previous implementation + +The generated call sites can get since its a in that keeps the reference of a generated call site +Wait +keystore as binaryupdated with current patch with the help of svn propsetThe current patch has keystore file checked inThanks +Fixed in SVN +Set your hint throttle level appropriately instead of trying to synchronize +also does not contain a bug in that caused an exception in a rare case involving lookahead +I want to add more avoid duplicating code when possibleAbout enforcing backward compatibility +re it wouldnt have issues so it could be added any time no need to link it to this issue or to rush it +Correct is libgothrift Thrift allows for nearly any combination of protocol and transport + +No validation will be performed itll be up to the user to pick a separator supported by the parser. +Previous patch did not have this one does +The lib will be packaged by the servlet in Shindig with all its dependencies included +Ill fix the LICENSE files +I am used to having a committer actually work with me Mike is awesome and we worked on several issues together +Framework patch is in rev +Hi I want to try converting incrementally +Not sure if this applies to other commands too +The patch is in the trunk now +Assigned to me as i have patch ready with me +Instead of Other how about Project not visible or something similarTaking this idea further how about letting all users have read access to the queue but obscure the names of inaccessible projects? Is there any reason why any user should not be able to see the queue? +Reopening to mark fix in +Patch uploaded +But the guys are discussingimplementing some features that would use some Hibernate specific features like +is your problem that the classloader is not released on undeploy? do you have a cache manager per app or is only the cache per app? when you deploy and undeploy multiple times do you see multiple leaked classloaders or just the first one +So once we branch it should be bug fixes not right I just meant about tying it to + +whacked everything this morning and tried it again and it works +Thanks I applied your patch and fixed other modules at revision . +TracyCan you also past the stack traceThe spring feature wasnt changed for a long time Im not sure why the bundle dependencies order can cause this kind error +The schema is in place and handles jms ftp fs and sql configurations. + +Youll see +I updated to use to encode in HTML hrefattributes +Given the information from Brian I am resolving this issue as rejected +This sample demonstrates Session Affinity Load Blancing using SOAP session type +This relates to good +has a mailman archive. +Its fixed in the latest trunk +Committed thanks! +However generating two different versions of classes with the same name offers significant chance for confusionThis a commmon patttern observed in translators +Justin feel free to move back its up to your discretion I just wanted to be sure you didnt lose it in the shuffle +Scheduled for release +Two I dont like the removing of the check in the installer that the driver is in the correct place + +Since that is no longer true we should be able to remove these +The differences between it and the basic class are quite small +Value has been changed display hasnt kept up with it +It seems that when a ledger is flushed the list of pages for that ledger are sent to writeBuffers along with the number of pages to flush +Logging is fundamentally an end user choice +i havent had a real close look at the interfaces and methods in this patch myself so thanks for reviewing +These configuration choices are also missing when prompted for which config to deploy the SEAM apps +Sure here is a sample +It appears that using the Eclipse Team Create Patch command doesnt create unified diffs very well +Propose to set outputdir after the instantiation to add support for with is no particular reason that I recall it was just a convenient place at the time +Patch against trunk +There are four cases we need to handle in +Removed a part of the description that I added by mistake +Fixed failure +I expect it to be officially released in early + +Who knows what sort of crazy shit backdoors he included in it +all bugs were in handling of double consonantsexamplesaangetroffen aangetrof expected aangetroffafvoerbonnen afvoerbon expected afvoerbonnklommen klom expected klomm only passes of snowball tests +this patch only contains a testcase demonstrating the problem +So all issues not directly assigned to an AS release are being closed +When the key functions of dbre are finalized some refactoring of the project structure to support additional jdbc drivers will be added prior to official release +May be it is really a caching issue in the plugin +Aaron and Todd for many reviews and comments on the Jira +I have corrected this for next release. +But from Knoxs perspective the input stream is already consumed and Knox dos not send any dataThe connection simply hangs until time out and then times outThe fix here is to convert the data sent from the browser in and replay the entityTo protect against too much resource usage we do the conversion to only if the of data from the client is less than or equal to MB +Recreated patch with completely new trunkModifies all +additionally when the pom if badly parsed and cannot be read I reuse the old instance better have old than bad data. +Fixed. +However what basically happens is that when you load the page then the other pages are is a bit unrelavant because all it does is check if you push the link Legg til postadresse which will call the it is in there that we have the suggestionbox that introduces the bug +Can anybody give some informations +Please let me know if you still have any problems +There is a difference between CLI options hyphen and the MS Word same as there are between regular ASCII quotes and the curly quotesI dont think that CLI should honor or accept either of those alternatives +Yep fine by meI think it would be far easier to unit test document the final version rather than something that will change notably? really Im talking about formatting style use of libraries etc +Attaching a second patch and with changes to most of the files in the previous patch if only to change to or later +The idea is that during continious integration the cvs sources are fetched as is so that the complete sources can be cleaned and fetched again +If were going to do it right shouldnt we have etc +only every fifth message +Just added another example CP in as well +Now that work is well underway with AS all previous community releases are +The same situation is in +ant passed +Its default centos template and i hope pv drivers will come by default +Ive also not yet figured out how to exclude the generated source from being examined and obviously that has a whole host of problems like unconventional method names s probably also some good stuff in there +Thanks +Fix applied for on trunk r Fix applied for r +Hi Adrian to help me understand the visitor solution do you have an example of how the visitor pattern could be applied in +Easy patch +If you know whos the right contact person please let me again for your help +Committed revision +Sorry for taking so long to fix this. +Looks great for me +I looks like maybe some channels are lost during concurrent processing of by Jacob +this is current as master same problem exist in EAP branch +With the current code a value from the Foo enum has type Foo +Can I get the permission to get bugs assigned? Or how does this work here at Jakarta? I dont suppose I can get commit permission yet but what are the steps to go to get involved? +This is actually working as designed due to the way Aegis doesnt flatten collections down like JAXB does +Does that example help +and too +Hello TimHere is the implementation of network system +Maybe I should have done this testing before I filed this idea going to close as InvalidThis was an interesting issue thoughA negative result is just as interesting and informative as a positive one +Small nits please use curly braces around the if and else blocks in the OPADDBLOCK section per style guidelines in readBlocks move the construction of inside the else clause since its only needed in old the two things from Todds previous comment +Hum I dont know about that +That should do the trick for us +My comment about channels was incorrect +x +Im uploading a new patch address wildcards +ArunIs this patch where youd like it to be? Anything left to do here +Its not in a format that can integrate with the existing documentation +added dependency to in r +That code is not temporary is itNo this is not temporary code +Reducing memory consumption in the Namenode is partially addressed by +The one provided by James so Derby I guess +Added a PR for the tooling so this is ready when we need it +closing as cannot reproduce in recent versions +I still think this is a good idea but I ran into problems with my initial prototype. +Hopefully it can be added to the release +The test case takes a long time to run usually in the order of minutes +Interactions with use public key ssh auth but I dont think theres PGPGPG integration +Updating the release notes fields +Implemented with svn revision additionally with svn revision now a default behavior is that module will check for the timestamps of the users and groups files and reload them if they changedThis is an optimal behavior since it shouldnt affect performances and doesnt require broker to be restarted on file changesAlso reloadfalse is now default value and use reloadtrue only if you want to explicitly reload files on every connection. +Should we just wontfix this +Added javadocs for interface fix version of to to have a complete changelog for +Please do not close this issue but reassign it to me in any case +Id rather just the bypass values together and if any CF wants the CL we throw the whole RM into itthat would make the CL logic since the flag would be respected depending on what gets sent with it +Second patch addressing parsing issues with CDATA +So not a requirement but a strong suggestionYeah we should discuss on dev +Is it actually an svn patchSI think wed better start with version rightAnd eclipse version target wise I am still using Indigo and worked pretty well with it so having a lower prerequisite than Juno sounds reasonable +Hi SeanWill have a look at this one on Monday and post my comments on the same +Having said that I am not sure if that change wouldnt break things at JDO +I can replicate on OSX when changing LF to either nimbus or metal +Moving out to move back to if this will be work on soon +Steven can you upload another patch with Ashutoshs suggestion? I verified that it works already +Erick do you think should be abstract +Only because you said youre doing it +patch for trunk was been changed patch for patch for and for patch for and for patch for and Ive fixed your comments +PMD and would be you really want are a list of patterns that you would specify +Migrating bugs marked as resolved before and tagged as version as fixed in RC +That eliminates the instanceof in the conduit and also allows us to do the same if we implement NTLM auth later on as that will also need request caching. +Excellent +Closed at users request. +Python client broker r. +attaching the patch +introduce a Check property names that is enabled by default to check where you know the customer part +I agree that getting a reproduceable state is key to getting this resolved +While I sort of like this idea it seems to have a couple of disadvantages Any passed into and would have to have these properties set from the beginning +Verified it works with a small sample app +This is a valid point +Moved from to just after merge of old branch into TRUNK +I dont see how this solves the problem of I want to CAS row X based on the value of row Y + +Litmus test passed The Preference portlet is showing all of its values +JSF event listener created in component generator have been override mannually defined check for a existing property and merge property description values +In revision difference from missingset DEFAULTJAVAOPTSserver XmsJAVAMINMEM XmxJAVAMAXMEM XXJAVAPERMMEM XXJAVAMAXPERMMEM KARAFDATAderby true set DEFAULTJAVAOPTSserver XmsJAVAMINMEM XmxJAVAMAXMEM KARAFDATAderby true +well deal with this issue in since that was the ultimate cause of this issue when it was reported. +Hi Jamie demo doesnt ship in the Fuse kits +cvs co r Branch +If it is also missing component please set both so that the issue can be properly triaged +Its included in the framework assembly +we should probably look into madvice direction if yoi want to control over CLS mappingsAlso a good idea +It seems a little odd that what worked in the last version doesnt work now but your suggestion did resolve the issue +The timezone param had not been set +Is there a timetable date for release? If there is insufficient time to get a proper release of the JSR artifacts Ill request to be allowed to release a release candidate into the main repository +A couple responses follow The is supposed to point at the libs +Ive added one more point in +Each language writes one or more data files in buildinteropdatalang containing randomly generated instances of the sharetestschemas +but why did this happenSorry no clue +But apart from that looks good +We should close if we cannot reproduce and the original reporter cannot provide a failing test. +Slava the exception is not caused by seem to miss some jar fileTry to simplify page layout to narrow the problem +This patch adds the verbose implementation in GCGen +Patch attached to resolve +Then use mode +sure I will try to cut this release ASAP +Patch for attached +I took my cluster down for minutes +Dhruba when was this class refactored? I think users are still running so Im assuming that its deprecated in both the classes in hadoop exist now but is deprecated +Rows per secondsResults are quite interesting +This makes much easier to embed and use in downstream test title for to and trunk same patch applies +Here is a patch removing task the patch to address hudson failures +and committed! +I will change this and submit a patch +If it is not reproduced in CR then maybe it was a build problem +As a user of the Tuscany plugin how would I add the Tuscany dependencies without adding the Tuscany library +accidentally made the default with this patchThat would put the practical limits of a table at millionish? Perhaps we should fix at keys or something +Heres a patch that stubs out the mlock for that test class and adds the missing license headerI also fixed up the stubbing in +Test suite passed +Please fix up remove the config change to ssl +The joinrelationship is of course done later on query +This was also fixed for the EAP release. +Reopen for backport +Its not easy to reproduce this consistently but Ive made this change locally in the past and it fixed the problem +Hi Yannick You can find the daytrader ear under daytradersourceassembliesjavaeetarget +it is easier to handle outside of bookkeeperfor example you want to move one ledger journal disk from to ssd +If not then it goes into the next big release which is +Working now. +Eli I was also working on the same area for so I attempted to integrate this patch into that one +But while trying to start another instance of using the netboot URL as the base URL many exceptions are thrown +If he has problems getting it setup based on that Ill add more specific documentation recommendations to this jira +Looks good +Therefore we dont have to specifically filter out the HTTP endpointsThanks please +Just looked at the code starting the thread makes the TSM think its running which disallows token ops from the edit log +I have done this and the resulting email headers look right +Id add ASL to each class and Id change the packaging toIs this you can do so we can get this in +So we are listing one but not the other +It looks good to me +POST method did pause for approximately seconds before the body started arriving +This is because getFiles is used by job client to calculate splits + +new version of Undertow was merged into upstream +Confirmed fixed. +This way there is little distinction you need to make to decide between the Message vs PayloadBarry has a sample procedure code how this can be achieved +This applies on top ofI did not end up incorporating the linked issues afterall +Today I am working in the backportObservation remember that literal string must have simple quotes ZEMLISTE your string +Hi Uwe how can I see those warningsThank you +But thats beyond the scope of this issue +trunk +Just going though open issues have you had a chance to verify whether or not the fix worked +Thanks for raising this localization issue Knut +doc patch for rcUm +The plugin comes with a very simple basic Simulator implementation however this already allows to simulate the by today known nutch scoring problems like pages pointing to itself or link chains +SteveIm sorry but you have reintroduced the performance problemOne of the problems is that alot of instances are created at different times in the creation of theEach instance does its own lookup of the extended transaction class each lookup takesThis basically kills performanceMax and I agreed that caching the extended TX object looked up from websphere in the sorrounding instance was ok +tried to optimize preflex codec removing more conversions and removing scratchbytesref copying +However it is available in trunk +Kelvin KakugawaI pay attention to a cord of githubThanks +When I dont put the MVC annotation driven element everything works +final patch committedOleg +Also added precedence tables in the expressioncriteria sections of the reference +It seems that all the tests are in httpfs +resolves conflicts with checkstyle change +I applied the addendum w some massaging no refactoring +The only removed by this patch is the write stats in which is already covered by the and attributes +This JIRA covers many classes and Ill try to provide patch in steps so that it is easier to fix for me as well as commiters +Mylar doesnt work +Can you point me +Let me rerun them to see +StackIf we presume that the list of servers that joinClusters received containsthe server hosting +I thought we only injected the maven jars if we saw a target directory ie that its a builddev environment instead of an installed one +So the problem is not with pom downloading but with asking passwordCan you please describe in more details where exactly you configured proxy settings +I have showed Yahor how to reproduce the issue +it may not be related to upgrade +tests passed here before the recent rebase I just kicked them off again but they will take a while +what is the structure of the and the class annotated with +The Javassist team think this is resolved as of +Marking thoughI can take this +More infoFrom svn checkout in HADOOPHOMEmapreducesrccontribmumaksrctestorgapachehadoopmapredthere are no and files. +Right but unless we store the parent ID having RDN alone in the Entry wont be helpful +Once again IDE didnt crash +See my comment about that in I think its doable +This would really help organizations using eclipse with a lot of proprietary maven artifactslibraries migrate to maven piecemeal +Thank you KristianIm still not clear about Typo strongger how to typo it in code file? Some trailing whitespace in do we have some common code templates file to format code +The changes have been committed +Can anyone review please? This will fix in trunkThank you +Immediately put a Apache patch with updated code headers and other license related files of trunk to comply with Apache +The WAR file does not embed and anymoreMy main concern is about the mandatory SONARHOME variable +Is this on the roadmap for? Its going to be a change to the agent wire protocol checkpoints and initialadaptors Im afraid maybe more +Really anxious to see this one committed to the trunk +Imagine hundred applications running in the same domain which would be using the same default profile +Documentation has been updated and sample configuration files have been modified to demonstrate how to configure the multiplxer with. +in the end it just increases the state stack in simple to this should be fixed has also fixed a error that we were having trouble creating a test case for +jarExtracting system system system +But since I have also experienced this type of problem within the Eclipse environment I thought I would post the problem to JIRA so that we had it recorded +Am guessing this is running into the same issue. +I give the priority of my language setting to English but the admin console still displays chineseEnvironment windows xp professional sp seems that the locale priority problem still exists when using IBM jdk and this patch can fix the problem when using IBM jdk version +Serge please reassign yourself if you would like to continue work on this issue. +Just assigning a hope this patch will help solve the issue +Max the solution does nothing except the correct importing of projects created into the difference is that the test ear and ejb projects were copied into the workspace instead of importing them as existing projects +Missing the client srcmainjavaorgapacheservicemixsoapinterceptorsjbiSending srcmainjavaorgapacheservicemixsoapinterceptorsjbiSending srcmainjavaorgapacheservicemixsoaputilstaxSending srctestjavaorgapacheservicemixsoapSending srctestresourcesorgapacheservicemixsoapTransmitting file dataCommitted revision . +CP updated in revision updated in revision +Yes you are right +thanks +The tinderbox failure I saw last time was also not there +It seems that the patch introduced some javadoc warnings +Thanks +Ive finally managed to reproduce the bug with the test case you sent by Ive included it as a junit test and it now passesPlease tell me if it works in your environment too +You can only create TCPIPconnections at a certain rate due to the mechanics of TIMEWAITconnection reclamation +Bill or Lukas do you happen to have any other more complicated examples +I am now able to install by adding to the manifest of my fragment bundleHow do I test this example? Still using derby would like to use mysql + +You have a patch for this now Kay Kay +And again sorry for the delay. +I just committed the code this morningIll be doing another snapshot build sometime later today which will have it +and use the for usage forum +Brian Stansberry replied to my email and thinks that setting y is ll mark this as resolved and if we get a better solution. +Darrin do you regularly do a document comparison with the project docs to see if things like this have happened +Added a class to provide ketboard navigation with UP DOWN LEFT RIGHT HOME END keysSimple keypresses are used to jump to the nextprevious button whose rendered button data begins with that characterFocus transfer via the TAB and SHIFTTAB keystrokes can be modified to transfer focus out of the +Thanks! +Thanks Eugene for the contribution +bq +What do logs say Bryan? Is this tests or a standalone cluster +Could you please retry with head +Test now passes +yes it work here is done. +Yes Sun is leading this JSR and will provide the RII started a thread on the mailing list two months ago asking for advice representing this JSR +buildivylibcommongriddevbin +Included in. +I dont see how it can be done in any efficient manner when you have to work off a Searchable and obviously you cant do anything with multiple Readers using the on the client side so it would seem making the method work is the only option +Improves javadocs +bq +Outstanding +We did this with all +Looks good thanks Mark +Same for getApplicationsDone I think returning empty collections is fine to indicate no results +Verified all +After I do mkdir p datatmp Im able to start the child instance and is generated in datatmp +Another use case In Hadoop there are scenarios where Hadoop returns plain strings with the word error in it instead of throwing an exception or an error +Or conversely when you change a file to writable it automatically gets added to your changelist + +We manage to get simon and chukwa running in parallel +Its now working thanks for your quickWhat was your final solution hereJeff +I have a new srcittest directory that subversion says is already under version control and yet it wont let me add it to the patch because of an error statingdiff old homejkuhnertprojects Working copy not locked this is probably a bug please reportsvn Working copy homejkuhnertprojectssrcittest is missing or not lockedIve tried all the svn commands I could think of to fix it but am running out of ideas +The full exception will still be logged when DEBUG level is active +The previous attachment misses a unit test case +Removing incompatible flag +I made a slight tweak to the patch to put the sample files under then committed the fix and test in r. +The Selenium and support in Tapestry can be considered for internal use only in I favor Spock and Geb now and encourage their use. +A default class is created to emulate the current behaviorIncludes a simple unit tests and passes all unitsystem tests on my machine +Thank you ApritThank you Chris for the review +When latest false it will check all the versions and if any of the version value is that row will get selectedBTW how is the Scan object? I guess you have set in Scan to get all the versionsSCVF will not decide how many versions of a cell to be send back to client +browersim! cordovasim +Nope its fine +I can not guarantee that it does not produce error at all because I made only finite number of runs but no errors so far +to fix ASAP this is a huge problem as it breaks logj as Benjamin has previously stated +Which devices browsers and versions are you testing with +will take that into account +Can you confirm the fix +documentation build does not work missing in source distirbution +Update the changes with replacing all the with so that make sure only one single folder is neededCommit the changes at revision +Jim is taking care of the release this time +and a ROOT as long as I supply encoded name only +This is not a problem of SSLR but a problem of consumer of SSLR API in this case problem in PluginThis message simply says that rules for Lexer were constructed in a such way that they doesnt allow to consume char on a given input positionIn case of lexer is unable to handle one special character which is known as BOMAnd in case of your work on this might be a clear indication that lexer is badly implemented however hard to say without code +Declaring it manual should give us a better control of compatibilitySo we vote for the method names solution right +Vikas +They should be ignored because they are not annotated with Provider +Submitting diffs +Maybe we can split these into two as is independent of Ejb Server portlet and is related to it +Both of them should be modified to run JBDS under mac +on the patch +This means that when a new NIC is plugged in it doesnt work for seconds which may mean its not getting its DHCP address in timeBridges created by cloudstack should be getting a proper forward delay so check thatIs there a chance these two ended up on different physical hosts and the issue may lie there or what does your environment look like? Just looking for additional clues +It would be really nice if we could wrap a instead of a instead +What is special about this ce region? is this a split parent or daughter +Could it be done with an escape character instead such asabcd +and ant test passed on my machine +Select whether prompting should occur when adding the groovy nature +Its a beast not the cleanest but has interfaces with essentially purged +skh this should get fixed when or other related UDF is logic to treat a null category as misc. +Further if this was the reason test case itself would only fail but currently it doesnt fail but causes the crash of +This would be useful in the RNG methods Remi proposed where all that was needed was whether or not an operation was valid +Ill revert them +paul Can you test whether numactl exits with a proper exit status on the system where it is not supportedIndeed it does using interleaveall when there is no NUMA policy support causes an exit value of . +The bug was fixed in ead and confirmed. +It is definately possible to write a thread safe yet stillefficiently without the locking +For every backend there needs to be a translator that converts the logical plan into physical plan with the nuances of the backend factored in +YES I did specify EXTERNAL SECURITY DEFINER on the DDL to create the proceduresI will extract the relevant parts from my scripts used to create the data base and its objects and double check try again +How so? In factChanges in NM capacity triggered from outside of the regular scheduling would unbalance existing distribution of allocations potentially triggering preemption +Maybe we can merge the patch to? By the way could you please recommend a committer who is interested in so that we can invite himher to have a look +We got an earlier successful test run +Ill take a look and if various issues above are still unsolved Ill create subtasks to polish them off +Would it be OK with you to not register the medium security policy by default? Hi OlegIn general Im fine with your suggestion +bulk close of resolved issues in. +Hi MajaEverythings on one machine and the master and worker are the same task which seems incorrect since I have set to trueI need to track down why they are being started on the same task +Will get this in ASAP after we get out the door +Thanks Hairong! +That said im taking a good look at this now and hope to get it in today +The tasks using enter a zombie state and wont finalizeI have failed to reproduce this in several tests running different jobs i guess the zombie processes were the leftovers caused by the earlier versions of the patch +bq +Blocker does not start to describe this +You are correct I was using the Control Bus when really all I needed was the Service Activator +Does that make sense +Heres an updated patchOne can run tests using Avro serialization for RPC data by specifying +minimal test case for demonstrating the was committed together with other changes at +As these path elements are fixed we defined Java constants which we wanted to access via state that this isnt a bug but to me its not obvious why EL should be usable in some elements of the XML configuration but not in Java configuration looks promising I will check that out +If there are any concurrent scanners then they should either see all of the new data or none of the new data on a given row +Why I cant generate all picture cache in one timeAnd whats the meaning of remove all GWC related files? Thanks again for your patient and replies +I think its reproducible at and +Oh btw the Geoserver Im using is a nightly of sorts but I dont remember any significant change in shapefile rendering from to now +Eli per your comment I have created a branch +Patch to tell the class about the new anon uri template. +Agree with the issue priority also +It would be much more efficient just to assume that these transactional tables are never so the replication for these tables can be skipped entirelyThis could be achieved with a customized but since its not pluggable then its not possible +Thanks for noticing this stale issue Paul! what about and documentationProbably new issues should be filed if that documentation still needs to be updated. +Foundations Proposal should replace the Name Space proposal +This jar should be placed within the directory +This interface is well part of the project +I am closing this for now +integrating Toms request regarding testThe only concern is the directory structure layout is a regression from +It turns out I had an error in the interceptor I was using +Demo showing that this wrapping a system JAR is possible using a placeholder with. +The reason is that if an image is drawn and the composite is set when drawing the image +a step is disabled after it has been returned as the next in a flow but before it is used +I just committed this to +verified that the DB manager portlet is functioning OK which means the jstlsql tag can load the drivers from using the new approach David implemented. +Ill just keep this report open until I have the tests running successfully in case there is anything else that needs to be done +Replace hardcoded path to thrift executable with usrbinenv thrift +Can you open a bug on the PDF please +Thanks for the patch +Suggestions welcome +Thanks Rick +Thankyou Adrian a massive improvement! this issue be closed now +Attaching patchThanks top of Zhijies comments The expected exception in the tests should be +I committed this +For canceled compactions though the filesbytes is going to be with heap sized changed and the description changedI also changed the descriptionThe counts wont be wrong they are coming from the cr +bq +Ill fix it +class or if change a +Looks good +hi Marcus are Xenserver supports merged into master +To resolve the synchronization at we not only need to care on the perpared statemnt but also on the counter and the resultset +server i s txfooap unset LOGCXXCONFIGexport BLACKTIECONFIGURATIONlinuxexport BLACKTIECONFIGURATIONDIRclivalgrind toolmemcheck yes yes +Either way is fineI agree either way is fine Ill let you decide If you create a new issue please mark it as a regression so that we can make sure it gets fixed before the next release + +This is a suspected bug in Dojo +You will have a fix in CVS tomorrow +Thanks Ivan for the deatils +Were adding reprojection to geotools datastores and once its in place should be able to handle it internally if they have the capabilities and geotools should handle the reprojection if they dont +This doesnt work for me +small update on this one +Marking bugs with invalid fixed for release as fixed in. +Can we do the same thing for and trunk? Im happy to open a new JIRA and submit a patchI see no reason we cant do the same for and trunk +A more complete patch for +Closing all resolved tickets from or older +I dont get it Kannan +Fix to fix has been attached. +submitting patch to patch and adding release note as is incompatible change +They are not for customers just want to run one specific older version of +Marking Patch Available +Matt Birks is working on this but I cant assign to him as hes not listed as a developer +This version locks down the version for the to and shows this issue is fixed in version Jesse be sure to use version of the +Closing all resolved tickets from or older +Serialization and Deserialization of Message Context work perfectly with standard Serialization of java +Also it needs async disk io along with async network io +When supported I can help translating to like the votes are growing up +My bad +will build a jar from class files if a packages jar cannot be located +Looks good +First one is good +Instead work with the portlet requestcontext objects +IIRC using is a comment for C code is defined as a standard in C +Not approved for EAP +fixed problem and tested changes +Thanks Kevin for updating the patch +I suggested to take February but then we were already placed in the MarchJuneSeptemberDecember scheduleLets stay there +thanks guys +Im closing this issue with approval to proceed as there are no known patents and no objections. +Please make sure to build your patches on Java +Can we please keep the Fix Versions EDG +Ill start looking at this today +Do you have a feel for what this may be due to? I dont experience any significant network loss on the client for the late response here Mauro +This seems to be out of date +JesseI checked the two outstanding Hadoop QA jobs around they were not for this JIRA +Oh sorry I just cut and pasted from the issues list The correct version of the above comment is has been fixed for CP and is as such reported in the CP release notes +If anyone finds the comments about or the filtering of duplicate references in still relevant I think opening a new ticket would be best. +Sane and adheres to principal of least surpriseBackdoor provide a means for the user to indicate the specific setter type to use +And Im kind of inclined to call this not a bug until I see some breakage within Hibernate caused by this +I couldnt follow when and what may cause this +I have removed that patch from the bug reports and have added this fixed patch +Just to clarifyPlease ignore any earlier comments about adding a time we have seen this issue it has been due to the user using a non transacted data store with JBM +Adding srcmainjavaorgapacheservicemixbeansupportAdding srcmainjavaorgapacheservicemixbeansupportAdding srcmainjavaorgapacheservicemixbeansupportTransmitting file dataCommitted revision . +Node put A in cache +Fixed in +Backported to branch as svn backporting upto codeline +This is the proposed for the final beta +root root Sep SUCCESSdrwxrwxrwx +it doesnt display the form by default and if you manually create a that should work you get a malformed XML response back +didnt run any testcases because the changes are otherwise it would have failed due to the mismatch between the protoc version in the jenkinks machines and the protobuf JAR version in the +It is not clear from the comments in the issueDigressing a little bit the fundamental confusion seems to be that the key used by the cache needs scheme authority and username but it only requires scheme and authority for look up +Ship it +Since this bug seems to start happening after some libraries added to the +This can be tricky and time intensive to test +Seems you miss +I will not have time to work on this for the next few weeks +I will email Noam and ask him to help you setup an environment to test this +of bytes and records in sorts input outputb Validates the xor of the mds of each keyvalue pairc Ensures same keyvalue is present in both input and outputRun this ashadoop jar sortvalidator sortInput randomwriterinput sortOutput randomwriteroutput +To be more precise about my comment this change introduces a Heap Corruption on Windows when is built with Visual Studio +There are a lot of question to Mac OS X platform +the read case should return regular features +Was this ever supported? This seems like a new feature and Im not sure why its in +Failed again with age obtained from cache must be +I just committed this +I am trying to see in the integration testsuite whether this test uses the handlers from the deployment or there is a setting done in AS +Can we add this to the Maven documenation? This is a deliberate decision to break inheritance. +we decided to keep the current strategy for user and group +It follows the steps of existing compact index +Tried a few things replacing Singleton with Stateless adding adding an interface and using Remote replacing Inject with EJBNone of these changes per spec injection between top level deployments is not currently supported +Ear file used for the pass with +Hmmm +See last comment on that issue. +However i have not looked into what is missing for jre +Please provide some clues so that we can assist +One issue here is that doesnt currently listen for advisories so it doesnt know which destinations have been deleted +Is it a must now to specify cascade type with Lazy fetching +This patch refactors into object +PS your IDE seems to add TABS! Please switch your editor to SPACES ONLYIt is generally a good idea to switch your IDE to show whitespaces for all Apache projects +This is impossible to implement correctly +Caused by compareTo with a null instance +ive been doing some tests latest patch has been working fine for mei think jason cleared up any remaining resetreusable issues with this latest patch +This is a provide full source code of this what do you think is? +Linked to discussion thread and changed to a patch +Mapped controller actions are not invoked in other app servers either +bq +Brian No validating the ability to connect isnt a part of starting the services + has one method called sayGreeting sayGreeting renamed say method say used to be a problem +Thanks Christian I committed this. +Well in the most generic sense already implements range partitioning +I hate to push this back like this but features are already backing up for and we have had little to no interest in this feature +I have to head out for today but will look into it from work tomorrow +However such improvements need also to be applied to all the other modules of Stanbol + thanks Raimund this would be of great interest! Ill be happy to help you shepherd it in +Just comitted the first portion of your patch related to the compilerI would prefer the aproach from alexandre he made good progress on making it more portableThere are just a few ifdefs and it will be much easier to manage the code base with one implementationIm looking forward to bring this into the next release +We probably want to check to see if any other archive formats suffer from this issue too and fix any of those at the same filed for a change in Commons Compress thatll make it easier for us to do this. +note it could be multiple items selectionthe better choice here disable menu item for generation but it could be rather long operation to check all selected classes check this +addresses issue in comments of last commit should have been associated with and not this one. +signin is internal and no application should be able to set it without authentication + +This issue is about master not starting due to failing to clean znode right? Is there a related issue for +Thanks for the patch Marcel +Is the patch working? Has any one tried the patch? +I should add that the current patch puts in logic in to do with handling specially in some cases +Which platform and jdk version did you use? The code for this fix is not related JAXB so it should be working for JAXB +hasPayload and getPayload is already there its just that today it always returns false +is open to fix. +Patch is available for this trivial +Forward port this patch of Michaels taken from at the patch it is a long way from trunk +Sorry I think the above test actually fails because it will sometimes decommission all of the nodes on one of the test racksBut if you bump it up to have nodes in each rack youll see the new code path from get triggered + +Could do it at db level like encryption or at column level withnew ddlThere tends not to be standards in the ddl +Just needed to turn it off +Its more work but itll do +Yeah +Patch for trunk. +Therefore the attack may not be all that easy +I thought the premise of the assembly plugin was to repackage a completed build with its dependencies andor other companion files into the sort of distribution file akin to a war or ear +See for further details. + +You are right I was thinking through the various cases and it does indeed work with the single loopFeel free to commit +I have looked at the v patch and I think it looks good +Its in thanks! +in CVSpaul +Let me know what you think. +devttyS is the first serial port on most systems +sh scripts +Fixed in fix introduced a long delay in VM startup a better fix is on the way from Daniel. +Ah good discovery +Also drop in the deploy dir the sar and the ear +Patch is committed +After thinking about it more we have the component init for a reason and we shouldnt force components to add code to an activator that really should be in component init +Problem is that its not always reproducible only about every third startup +It will be in +As its advanced zone even there is no iptables rule for the new bridge the guest network should still workI am looking at the migration issue now +Take a look at for a simpler alternate proposal without needing to resort to hints etc +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Adding to difference component to aid tracking +Patch committed as attached +Wont Fix implies that there is a problem but it wont be fixed +Version Released +One test failed and needed to be fixed +Investigating +RS has transited the node to OPENED Now the will try to delete the znode and it will succeed thinking the region is in RS +It could make sense that it is the default for thrift structs +Since Im not always getting the last job instance I get execution errors since I end up using the same value as an existing job execution +see entry in Release comment +thats bc they support the fractional sizing as well +Make sense +this as a purely issue to avoid complexity +But as said on the mailing list this doesnt work well with all FTP servers which expect you to traverse the folders using CD in a stepwise manner +now I know what I missed you guys you are awsome! archetype used in the does not create any java classes so I am not sure whatwhy created for you +Oracle DB and JDBC driver +Certainly we could debate and see what the correct solution is +Removed usedefaults to avoid conflict with default property +In AS if the system property is set to true then all WS connections will not verify the hostname +just one line change i havent run any case yet +Also to test the change I built Oozie and ran the script it worked normally +First version of the tool +Is this the same as +This is the output of the from the is the backed up DB after the creation is the DB after is dont know how to read ebcdic format on my windows machine +So do basic processing and then we queue up all log splitting in the new set of executors +Please refer to the reverser example included in +However I think that the real fix would be for SNAPSHOT plugins to be resolved dynamically and be reinstalled if necessary +This still means that when we want to restore a subtree using snapshots we wont be able to retain quotas settings because we cant use move +I might also introduce a property which will disable quering for the filters altogethercheers Sergey +I will post the test case for this scene for your reply with the very clearly explain +For those interested in this issue please take a look at the support coming out shortly in +Thanks Nick. + will commit if tests pass +This issue has been fixed and released as part of release +The workaround for my unitary test is to set LAZY instead of EAGER for all collections +BTW this operation only applies in the Server environment not in Embedded +Unless there are objections Ill commit this tomorrow +starting the HC with but it effectively downgrades it to a standalone mode as is never attempts to contact DC +For all of these custom class use cases I would go with the service loader approach so that users can deploy them in a similar manner to how JDBC drivers are deployed +A great deal of progress has been made we can now safely request the correct information every are now stuck on we need to do this so our D CRS used when we call with is accurate. +So now there are two solutions SaveLoad setting to some file through importexport wizards which is the standard eclipse Create additional file inside the project and updateload it each time user starts of them have advantages and disadvantages In user has full control of importexport operations and knows what to do +I will have to close this cannot reproduce without knowing the reports will attempt to correct the NPE error reporting in though +The headers will contain file information +Someone else could give it a try But Len was able to reproduce this as well strange +Fixed in by not caching the instance of the objecttypedeterminer at instantiation time. +The failed contrib tests are not related to this +Great workWhen will this patch be committed to the trunk? +for the patch +Tried to be pretty thorough but it could probably use clarification in a few places and might have an error +New patch with fix on other filesystems +Yeah that is missing +Funny you should report this now +Ive decided to make a button that can Add JSF Capabilities using existing wizard +the branch is what will become +I am not very much in favor of providing hooks in the applications +I guess its not unfathomable that it would be useful elsewhere +This includes an implementation of that makes approximations optional +The need to add tests for such scenarios is filed as part ofReviewfeedbackcommit would be much appreciated +The code has been in place for sometime now and I have created an issue to support the G collector in. +The initial tests I wrote up only showed me that FF and Curl both have limits and were throwing errors so technically I wasnt even testing +However the JAR has to be included correctly because I would get that error when is put INTO the session +Most likely a duplicate of which is windows specific and has been fixed +Have you investigated with a profiler yet andrea +The scope in the child is not overwritten in your sample project +I went a bit overboard I see that the doesnt have the logger stuff +Pch +Just to clarify the reason for proposing the code change the way it is in the patch is that upon investigation our understanding was that the config was being written to file by the before the replacement with the keys listed in MR deprecated keys are invoked +Kunalkumar already attached the patch and want someone from Maven Org to review it +Please see my commentquestion regarding this in +If you combine and you will a hugh number such as +Same file +Thank you for the ruby foo Christo. +The patch looks fine to me +Closing this for now. +Verified with JBDS Beta +No entries need to be manually imported into the new database +I also had a look at the Servlet SpecSection Welcome Files seems to cover this scenario +Hi Andrew Your patch is rejected in my ofbiz environment +So well have to wait till then its not going to be fixed for with jUDDI is fixed however it depends on the API version used you will need version which ships with AS and up +Yes this is a promotion request +There are changes included in r which could be ported as I think a fix for this issue is provided there +Current behavior of inplace behavior of inplace input with maxWidth attribute +That is depth should include empty grand child objects +However when it happens all jobs that have tasks running on that node or access data from the slow node will suffer +Linux fix in on commit issues from RC to final +I cant reproduce this at present If you can supply an html file that exhibits this I would be very interested +META +FrancisThat makes sense +The binding file is specified via the one question though why isnt this the default behaviour? +I added the getter for the to the +This was caused by a rather pathological situation but we can avoid the deadlock anywayDS delayed service registered under interface ADS service used a greedy reference to bind as A causing to be instantiated and activatedThe activate registered the same service under A +Thanks Michel! +The result may be different and is it make senseYou are right that two different could possibly get two different visible lengths +committed to trunk and cr for. +HiSorry for the late reply but Ive been pretty busy +Bootstrap has a class +Sorry about that +treats all Data Source names the sameSo on exitproject close only the UUIDd data sources are cleaned up +Closing this off +committed in trunk rbranchx close resolved issues +Scheduling for alpha and assigning to myself +Replacement for the originally submitted patch +Many sites are already using this functionality please check your configurationAlso if you are using any web server in front of your servlet container you need to make sure it is properly configured to pass and other data on to the servlet container without damaging the encoding. +The problem is not in the extends but in the way the engine manages missed activations in your case due to the enabledfalseA fix will be on the way in a few daysGreat catch thanks! + +Yes +non of the comments have been related to this + +So the root cause is on the upper level +Happened after restarting the app with jconsole +Actually the easiest way should be to only fork the file and place it in exact same lacation in your other war as it is in +The file is It has been tested with trunk branchx and the releaseI do not know of any outstanding issues +Yep my bad too I didnt look and was just guessing at what the problem was +So all issues not directly assigned to an AS release are being closed +I just import them from git to Eclipse workspace +instead of using a pattern how about just saying java +The aggregator EIP already validate this on all resolved tickets from or older +This will run geronimo from var and this image of geronimo should not be used for running multiple instancesJason could you please review and add the unix counterpart +Thank you for the bug report +In this particular case however clearly doesnt handle the URI in question correctly +Can you please add this if committing. +The logj exclusion on is required since it was forcing some dependencies that were not available in the central repository +Resolved by providing better error messages and debug logging for the actual stack traces. +Duplication of existing pages can be fixed separately if need be +No I didnt see this error +for the refactoring +However Im confused by the class since it uses which I would expect is unavailable in m +Log a deprecation warning when max is found in but honor ita Should we make it possible to go lower than Lucenes default? The current patch wont +If there is no objection I will commit this patch soon +Oppss +User has to add builder manualy or remove support support +Please correct me if Im you can have several profiles active at any one time by using mvn Pprofileprofileyou can also verify the combined pom with help mvn Pprofileprofile helpHTH + +Done +This task could be done in two parts Basically the idea here is it will be possible to configure a to log into twitter as a particular user and receive tweets those tweets would then be sent to a configurable address in as Also we could consider outgoing tweets +names for attributesadd should be defined in application bundle or in for propertiesRICHFILEUPLOADSIZEERRORLABELRICHFILEUPLOADTRANSFERERRORFILEUPLOADENTRYSTOPLABELRICHFILEUPLOADENTRYCLEARLABELRICHFILEUPLOADENTRYCANCEL in attributes stop uploading label i +Yeah I guess mine is a duplicate +No longer an issue in and upwards. +Sebastian please feel free to the issue if you have more details about the expected behaviour +I committed to branch +Trying to move the site src stuff +I ran junit All on v and found no failures +Hmm now that I think of it I am not even sure how this thing works with respect to deletes +I guess the question is whether to wait or not? If the session has not started yet +suggestions for implementation column header METADATA or can be used repeatedly code snippet row should contain metadata name if possible check that metadata name is not used more than once rule rows should contain nothing to omit metadata or a value to insert the metadata entry +imports generation is controlled by the template in myfacesbuilderplugin +Would you happen to have a stack trace you could add to this report +use to find start and end ranges to transferI dont think this works you still need to perform the compaction first which means you still need to dump to diskEDIT Unless you are suggesting streaming each individually? +AdrianI was just reviewing your comment +Im pretty sure I checked all files + complete +Ive committed changes that remove the use of ntohl etc +sorry sorry! You are right +trunk +I was able to run this successfully in against a secure cluster +The SI code now resides there andthe code is commented out +Not intended to be included in the final diffsubmission +Changed this to be an improvement and schedule it for DC +Thanks +Backported fix to with revision Ready for verificationclose +Assign to during tidy up prior to +Tested it with a GB file +New version of code patch fixing testconnection master file +In my caseINFO Exception while initializing process enginejavaxscriptWhich is an issue with java compatibilityDavid can you check if you can reproduce the error and see if you see the logged message above? +If a user has a login they will be able to see the ticket in the CSP +Committed to trunk +Deleted the old jar and a attach a new jar built with patch v +I was trying that with a new pre gem but I dont seem to be able to push it againIll attach the gem to the bug so you can try it +Thanks Ashutosh! +I lowered the severity to minor which is more appropriate +I do not have a system with uClibc installed so I do not know +I just committed this +The general consensus of that discussion was there was little real world use benefit to treating them both as generic Items at least not in line with the amount of effort necessary to refactor everything +So sorry +Cant see why it wouldnt work +you can assume a Teiid Runtime is installed +So to get run by CI this test needs to be added to or similar I believe +The other Map objects are still accessible through a path expression +Committed the latest patch and rewrittenbackported the javadocs for +We dont want to flush all ledgers and all entry logs as in case of flush just to make a clean page available +Add more file to trunk with revision to with and to with . +how do i get to the git repository? Im having a similar issue in +Verified at rThanks for efforts +bq +Youve commented out some code rather than deleted it +Hoss no problem +Can you provide a patch for it +Java is needed by the but this class is entirelyoptional +Hi ChrisCustomer says that they saw the issue even when cacheLevel +I will close this +And here is the improved version sorry for the previous one +Thanks for your quick license files in the attached jars are now named instead of +Committed patch to documentation trunk at revision +Moving fix in version to +I didnt include a test for this since the existing jute code doesnt have a test either +We had some busy loop issues in the past and it was because there is no pause even after there is no new data to stream +These compile with javac from and but not with Eclipse Indigo SR or later +Looks good to meTo reproduce it you could run multiple sleep jobs with s of MapReduce tasks +is also the server trust store +Thanks for posting the fix saves me from tracking down exactly where this might have occurred +My team doesnt use Maven or Ivy were behind a firewall and Spring is already an approved software package +I have not seen this again except for that one time +Hmm +A slightly more refined version of it might be If the database is not shut down gracefully then Derby could check the integrity of all the conglomerates after performing crash recovery +Agree with Steve we should limit the length of a tag and number of tags +fixed the race condition in but still has this problem +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Hello TomIf you confirm this to be a bug I would be happy if you the issueMaybe I will take a look at the code and figure out a solution but I cant promise anythingCheersAre T +Closing as invalid per the previous comment +Only intermittent failure in lang +Un gain de temps et de maintenance normeBonne continuation +Your last patch had some issues Your patchs targeted file was an old obsolete one +Reopening to mark fixin fields properly +In particular includes a test on a input directory that contains two symbolic links +The problem is that superstep after superstep you might have chasing partitions where you keep on switching stuff and have additional IO +Lukas please assess and schedule accordingly +Verified works withAnyway could this be fixed without need to escape +However I did check in some changes that mean we can switch it out for another HTML parser if there are no workarounds for this issue +Type quit or exit to quitcassandra connect localhostssException connecting to localhostss localhostsscassandra connect localhost Exception retrieving cluster name check you have connected to the thrift portcassandra connect localhostConnected to localhost cluster Test Clustercassandra exit markw +Included +Telling a socket to bind to port means pick any open port +It was released with the fix in +The previous implementation was reading splits generated by the client and localized for the AM in the initializer +CamelCommitted revision +Fixed in rev +Patch based on Dhrubas +I think this is the area to look for +BUT in this first iteration of tackling this only focus on the implementation classes only +thanks +patch looks goodThe findbugs warning is not relatedI have committed this +I guess I underestimated the probability of touching the impossible ehm +Thanks Todd +Input and output could be from files named on the command line or standard in and out +bulk defer of open issues to +to trunk and branches +If it tries to send a response immediatelyit will fail with a +Patch applies on top of will make this unnecessary +I think we could have some time to enhance this feature this month +So it should enter the satisfied state firstYeah right +Moving to as we want to look at the MEP stuff the patches into trunk and branch +Hence for the controller it is initialized to zero upon startup +Ran +change the explain plan statement also along with the query in all cases +Hehe OK great +Could somebody please take a look at that? +It also seems like it would play more nicely with the advanced case where the user adds an Authorization header to the replication request instead of going for the protouserpasshostport syntax. +This patch allows people to change to jsch if they youd like to try this wrt changing modules add the jclouds maven dependency and set the following +So all issues not directly assigned to an AS release are being closed +I meant field date type fieldName date notNullinstead offield number type fieldName date are some other types of that are not properly generated like is seriously missing in Roo and will be a limitation to build apps that deal with monetary values +Attachment has been added with description source FO used for embedded PDF has been added with description PDF with has been added with description PDF with P open bugs to P pending further could you try the latest plugin please? Let me know what you think and we can change the way the current if you think it necessary +We should add them to the distro which is another jira +Have the context menu part of the equation working now as well +Special components Space in lt property should be ltproperty instead +There are a number of complexities here for a couple of reasons This is integrated into the runtime compilation and we should be creating or throw exceptions at this level +I will handle this myself when were ready to make the conversion +The issue here is that we are bundling with the Seam distribution +I already had something similar but didnt have the perfj aspects for Aspectj weaved in +What Id love to see is something analogous to thewsdl query arg something liketest that generates an HTML form based on the WSDL description posts the data back to a little generic servlet that takes the arguments invokes the web service using the arguments and then renders the results back as XML back to the browser +has been marked as a duplicate of this bug +This issue was created to but i tink this probblem is about not m sorry about duplicate +Doesnt Java have a utility method to normalize a string for HTML +again since i started to work on it and nothing has been committet yet I will do it today commit will follow later +The quick start name is and should be included into EAP release but may not make it into Alpha. +Thanks for the patch! +We still have the issue that users will lose any changes if they redeploy but the analogy with Tomcat and editing unpacked war packages is a good one +What is backward compatibility in this case? +Now that you mention it I bet youre right +is resolved +Can you elaborate on that and do you have any references for more material that explains the significance of content type for frag ids + +Travis can you see if you can reproduce the problem and if not close the issue +We mitigated that to some degree in like there are no really good options in which case Id rather stick with the status quo until openjdk makes the problem moot +Dig into after released +Can this be ZEROORONE as we are defining an Association End that can take either one or no entity instanceOne more request can you please enable the annotation module as bundle +Mm attaching the updated source file for just the keyboard shortcuts for NextPreviousQuit +Feel free to give this a try with the next snapshot and let me know whether it works for you +The problems with the datepicker whent away +HiYes I can use the regular checkout for simple projects but for I like to use checkout as Maven because this creates all the submodules as me it seems like there must be any kind of support in Eclipse TeamSCM features for using the name like it is defined in the +So it would be better to incorporate a way to manage the exception than can occur during should cover this use case already +I have Cygwin running on Windows Ultimate +Can you try to run the test again with latest JDK Willem +C has featuresnoexcept or noex to disable exception support +Should we be reverting this change until new reworked solution? BTW could this change be causing +Hence this is the entire patch +This seems to still be a problem in jline +The height of the left nav pane is fixed +I meant to say that used to raise a warning. +Committed to trunk. +I added the server side code for this +But again if you dont want to know of failing tests I can stop posting them +I am sure that a few more items will pop out on this wish list +Thanks +do you agree +If you need to see where it happened you can look at the stacktraceOne other thing to keep in mind is that this patch with conflict with so whichever get submitted second will have to be careful not to break the others functionality +All tests passDIGY +Doesnt work in rev on FF for review +Seems to have a test failing +commited to trunk not yet to branch on JBDS to set proper fix to issues that were set to +Assigned to Darrin set documentation component selected documentation affects flag +You can get the eclipse module from the cvs if you want them so that your feedack bug report is for cvs head and not plugin m already working on the CVS version as we needed the bug fixed +Ill have to try out the workflow for myselfIn the meantime though as a workaround for this particular ticket +Thank you Arpit +Luc is too busy ATM to implement support +I dont understand +As you can see again in this case your issue was not a bug nor a missing feature but a lack of search from please respect our guidelines +I have to check that +Then +The three methods i added have been in use a lot in my projects +Reporters of these issues please verify the fixes and help close these with latest build from When there are multiple secondary storages are availablesnapshots get created randomly in one of themClosing this issue +Excellent +Adds toStringfromString methods that generate the metadata in YAML +When I tried the PCL Renderer it didnt work with colour images at all +The default runs over and the alternative using +This is a trivial is a query which should be raised on the devuser list +Linking to since we need to make sure all the keys and values were in sync between source code and before stop using it +The documentation of now reads The time in milliseconds the tasktracker waits for sending a SIGKILL to a process after it has been sent a SIGTERM +The change is trivial but hopefully it helps our users to use this feature without the concern of making the job running much longer +is also a common hyperviser and the data volume may not enough after a long time use +Now it doesnt. +It would be easy to move this to a header and would improve security headers usually arent logged +fixed at revision s necessary to add some tagExcludedtrue +So the problem is on eclipse +This is a pretty useful feature +Well it is a blocker for our usage of +testing against other clients thanks for the patch +note they ALL return s but are custom +Confirmed I can see it has changed in the docs +Manfred +Anything more I need to do to have this included in next release +Closing all issues related to WFSV as WFSV has been dropped quite some time ago +Ive created for the tomahawk patch. +Please read the later comment in as we found GC Threshold problems with the in this as resolved for based on the work that has already been done +I am thinking that this ticket might be best viewed as improving the exception that the web connector throws when it sees an SSL exception to give a hint as to what may be the right resolution to the problem +Yeah I think so +It would need to be configurable though since the session cookie is only sent a creation and I would imagine that some people have sessions that go on for months +Chris JacquesYes I like the OPTIONALLIBRARIES approach this might just rescue the work so far Is there any reason why we cant automate the install of the selenium libs? So if a user tries to run selenium without the libs we have some code to check whether they are present and if not prompt the user to install automatically? What would be required here? would we have to show license headers before they proceededI think this is important as Id like to look into Chandras suggestion of recording the solution to some of the common questions that people ask as a kind of visual how to +reloading of application contexts inside a windows service. +It cant return a Vector because then you lose the original index positions and access to the corresponding terms from the dictionaryI used the existing class from the taste module and moved it to math commons +Changing release. +And we can parameterize if we needed to that was just an example of an alternative solution for the issue you specifiedI looking for a list of requirements not a solutionAnother comment is I dont think the solution you are proposing would work + +A partial record that ends at the file EOF should also be ignored and not considered corruption +Dtest there relevant information found in +Thank you Blasyour patch is in trunk at revision +My patch is attacking that issue which exists with or without injectionAnd actually even this change doesnt guarantee determinism since we cant control randomness in libraries we use though there may be none at the moment +Looking through the code it seems to make sense to move methods that accept a JSP to a different class and have the remaining methods be ignorant of the view layer +debugging shows that the webconsole is doing everything ok +you deactivate yarnmapreduce on bootthese services shouldnt be started unless the HDFS has all needed directories in there +Theres not a lot that we can do with thisDoes this only happen on +I have no idea why or how to fix the ruby based one +Olga please review a design +its ok when i use sqlservernow i changed into and met the same problem +partially reverted changes of svn r New Revision revision I added a mechanism that avoids the big synchronization block but still makes sure that no two threads can concurrently load the same item +When checking the content I think there are going to be pages that use plugins not in use in CWIKI that will cause errors on pages +for the patch +Updates patch bzip codec is implemented in Java and is not required to build or deploy the native libs +Looks like it was already fixed on the site so my POM improvement is minor +someone with hive committer access should try this to see if we are able to publish to the nexus staging repoant Dversionant Dversion +Though I cannot promise you that Ill be able to do so +Feel free to punt from +It took about half the time that it normally takes which is great +The same problem +bq +I just commited the code +Can we add a license header +Building derby not directly but using ant from the command line but from another build file using ant call to the all target of the trunks top levelProperties are not set in but are set in this wrapper build fileI need to go and revamp that build setup theres some cleanup that can be done is necessarySo its okHowever Id not like this change to get backported +Can you please explain why a converter is required inside of a composite component and not outside of it +I dont understand why there is a send before sendDonesend is required before sendDone because sometimes last status update can be missed by communication thread +Mapreduce tests also ran except for some test case failures that are already logged +Please see the Google App Engine developers guide for more details +Added javadoc comments to each of the test methods +Duplicated by as a dup +Hi MartinIm interested in learning more about your new approach +Please reopen if you continue to see the problem. +No I dont but youll have better luck if you ask about errors like this on the users mailing list rather than reopening this bug reportEven if it turns out to be a bug its better to open a new issue for a new problem and you need to use better subjects for your bug reportsIve fixed the subject of this one for you +I really like this idea +I will do some investigation on it +My suspicion is that the python lib wrote a multibyte character string to the wire with a character instead of byte length header +Ruby tests still dont pass +May I please have a reviewcommit of this aging feature +I think the reasoning is that by the time our compaction runs another may have run before us so it is ok if the geography is differentNeed to fix this first +Thanks Thejas! +No problemo if you dont forgot to update theThe commons collection is provided with a lot of Jakarta projects and it is not very easy to see which of them uses the most recent release. +Thanks DeepakDespite your effort I have decided to not backport in releases too much work for me for a so little thingYour patch is in trunk at r +Thanks Todd +Thanks Teddy! +The release note looks good nowIts hard not to regret the old decision to violate the ANSI rules and create schemas implicitly on the fly +patch works for +Every time we update the docs and every time we do a release we run through the steps for updating the website +I have done an initial benchmark and profiling +If the problem still occurs please attach an example file that can be used to reproduce the issue +We need a fix for +Your doesnt have to be a null implementation +This also happens in Oracle +A test program we can run before and after your patches would help confirm that the performance is in fact betterThank you program +Export Restrictions hamper this approach +Milosz You are correct that some of this functionality already exists in the +However is this the expect behavior for the fix? Hope this help +this can be worked around by overriding the respective method from within the webtest. +Patch that fixes the issue along with a test +Leaving the original issue as Fixed in +This issue is fixed and released as part of release +This has fixed the issue for us on Windows Server +hitting F I land on the Map method too +Why one would want to merge it in a local bundle afterwards +Thats where developers will lookAside from ASF policy one important aspect for was to reduce the amount of work that a would need to do to be compliant with +StackYes Stack it is impossible +patch committed +Could you please attach the actual file that causes the problem for you on +Woops I forgot the Assume check in +I cant reproduce the issue +Upgrading fixes TLD Doc but the tag reference still goes unparsed +So with this in place does the editor get focus? I would imagine that would be an outcome of only having this patch in place for the given scenario +The patch addresses each issue listed in the detailed description of this bug +Do we need to switch off failing test as long as WTP team will fix their issue +DavidAs of now the code basically uses the exact same code as the export command for osgi +For now I think Im going to put this in the App plugin since this plugin is basically the Android essentials plugin that deals with Android specifics such as exiting an application and Back Button functionality +Thus if we increment the cores reference during a time when we know that no other core is modifying multicore we know we have a core that is safe from being asynchronously closedTo destroy a core simply remove it from the map and call close on itThere are other issues that are still a bit muddled IMO +Unfortunately this bug exists in the released as well +Please post on wiki +If an issue is found and fixed itd be good to have a regression test +Patch has been merged to master. +HiAny updates on this problem? I got this problem here too and would be very intrested in a solution +Its possible that the web container will refuse to load SCIs that are not contained in the WAR +I will open another JIRA to deal with that +In the meanwhile also asking Mayank to chip in with his views +Dirkjan I agree its confusing +Huh? It needs to be committed to be testedAs everybody is sleeping in SF I commit it +Regarding the I think it would be fine to commit itIs there documentation somewhere like on the Avro website or wiki that would need to be updated to demonstrate the new callback functionalityOn the topic of load testing I think it would be useful to have some sort of utility for testing that goes beyond what the unit tests do +Thank you Jing +I just committed this +Patch with sync and async unit tests trunksrcmainjavaorgapacheservicemixeippatternsSending trunksrctestjavaorgapacheservicemixeipTransmitting file dataCommitted revision Thanks Ron for this backport to SMX branch if possible +domj claims to implement DOM and DOM requires that adjacent text nodes be merged on parse +has some limited support of EE +Thanks for the new look and the icons +Are you OK removing the copyright notice from the source file? If we end up committing something based on this we will include you in our contributors fileIt would also make it a little easier on us in reviewing and applying your contributions if you could create them using svn diff against a checkout of the current source code +Applying this patch should resolve the longdouble issue +Ill upload the patch later today. +Im spinning a new rc anyway +Is it possible to reopen this Issue or do we need to resubmit a new one +Marc Palmer is you say wont fix please elaborate +Sorry typo +for the trunk patch +Just perform mvn package and deploy WAR. +It is not implementing Addrinfo but handling it in Socketbind +FROM DESTquery SELECT DEST +Munging now appended to our main README +Even the most simple of these systems saw a need to tweak adding custom fieldTypes perhaps a default requesthandler etcThats not true like I said dynamic fields out of the box can build many apps! Especially when you are doing things from something like ruby where you have easy control of the field names! The out of the box request handlers are fine as well +Dont forget to VOTE for this issue if you want to keep ANT builds and can see through the Maven smokescreen +applied patch contains the following +But I am not sure this approach of fixing each of the side effects as they are detected is a good oneIt looks very error prone and wrong for datanode to not know which files under tmp are invalid since +I committed a patch to stop appending +Agree? DisagreeNice catch Mike +this is not must fix for so I am resetting the Fix Version field +Can you provide us with a sample pdf to demonstrates the improvement +This issue is just a project announcement not a discussion forum +hmm that was it +patchI added few comments +Shall we add an unstable annotation for the same? Is private or public? We should add the appropriate annotation +Therefore to avoid introduce more lib into the project I the jars in +Jetty has I think a s timeout so increasing the client read timeout above that should be a workaround +We dont want to set a precedent for this +Most of the panels related with assets now expose the delete functionality thats why Im marking this issue as resolved +I shall use only a patch file name +Im really not certain any other changes can be made at this point unless its skipping wakeup when the executor thread is not selecting using a volatile boolean +this is why we have just for such usecases +I removed from project +The same applies for JDBM BDBJE +Someone will need to port over the IGEvent code to make this work +cannot be used from code like as that exception class resides inAlso there is existing code in and that throws instances so this is not a new behavior +Failed test is unrelated +from the top level in current cvs version +I see that there is some initial support for this in trunk however given that this fixes several critical bugs for IE it seems we should make this a high priority and get it out ASAP +Andrea the patch changes method renderImage +Yes similar symptoms However this one was observed on vCenter +I tried this with and and it only happens in it be caused by our struts migration? +Koen sorry I just couldnt find any other way to make it work +Yes youre right +This can be very useful for debugging long message should be forwarding row? This is in the forward function not the process functionFor operators like filtering the numbers can be different +Can you please provide a WSDL that reproduces the problem and that you can contribute under the Apache Software License +This bug is currently blocking our continuous integration cycle +This may not be intuitive but I believe it implements the standard and the behavior is well defined and allows an easier holding implementation than guaranteeing the current row is available +Committed to branch +Scheduling this for CR as this looks like the same issue as it is the same then this should be closed as a duplicate or our bug stats will be misleading +Would it be possible to add this to the Quirks on the docs +Please relook at +MehantWould it be possible to log some output while the diagnostic operator is running? Some info I think would be useful to report on each iteration of upstream batch Number of records typesMy thinking is that this will help diagnose issues which would otherwise be masked by the tool +new generator based on iterator on instead of get and getAllows more than two duplicate mark +I looked at build output and there is nothing in the +This is now supported in the new Vaadin based web UI +I just tried your idea and it works like a charm In fact I think removing the setting of the JAVAOPTS from the groovy scripts and then putting SCRIPTPATH in startGroovy and then the D line in the exec statement provides a solution for all the shall commit this and then see if I embarrass myself again +is the new track for +Hi I wrote at least INFO level not exactly INFO level so according to tableWARNLikely to be a problem or it could be detected a problem it can recover fromOr you would prefer to have another JIRA issue requesting to provide application feedback +Can we please group all security related ticketsissues under one label with Security and track +It shouldnt be that difficult to change that +Thanks! I tested it and it works. +Part of the problem was that the relatedTo field was not being set in the fault response and therefore the replyTo field in the activity event responses was not set so requestresponse correlation was not possibleHowever in this particular example the fault name is not available as part of the on the wire message so need to include this as part of the information processor functionality +There is still the issue of the double definition of for which Ill propose a solution soon +Some images were in our skin I have added them to the was an error with the have updated the example and we provide a erport generated on another project while we are waiting for Codehaus svn urls to be of Mercurial is doneoWhat blocks me from using only Maven scm Api is the fact that I cant get the revision in a +reopening +Attaching a work in progress so we can synchAdded weighted graph supportTested it on a small graph against a reference implementationI didnt manage to test it properly with unit testing +Thanks Mahadev! +Node try to replicate A +Thanks Jitendra! +I mostly run a part of the script simply by from the editor into a Grunt shell then stop to examine the intermediate values +We will need to push this out +Committed revision +Can you confirm the above errors even when removing the YUI jar +Invalid expression hellopetstore +HelloWould you please try this patch? Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Should the examples target be removed from ? +I move to revert the portion of this patch from and push it to trunk given the fallout weve seen thus far +NO. +Give the latest trunk a whirl +Thanks SparkPatch applied to BEANS module at repo revision rPlease check it was applied as you expected +Patch for Yahoo! distribution +was reapplied so I assume this is fixed now. +the maven proxy settings work not just for other projects but also when running seam forge I deployed the hibernate tools plugin and it downloaded the dependancies via the maven proxy settings +This code already exists in trunk and seems not being have reviewed the Round Gizmo as +Reattach patch for branch +It would make the code more readable +The use of +Ill try to grab an fc environment so I can test this out. +Thanks for updating it +PierreYou suggested a link to an employee in the role of account manager +But based on it sounds like we have to add a check to our code +JC do you have a test case for this by chance +If I click on the th June box I would expect the time to be returned as not +DtestcaseXXX will also emit output as soon as possible to the console +Mircea volunteered +Just a question why do we need a queue here +The last two log messages were at +No I defined a user library just containing the JSF jar and put this user library in front of the jars provided by the server adapter +I think this is due to some of the changes in the command line processing +Ive put a question out to the mailing list to see if we can get this answered +I will let you know as soon as I commit for you to give a +Felix Declarative Services has been released +The implementation of the process method would embody the policy that the processor providesDoes that sound reasonable to you +Issue to update Release Notes info +Thanks Phil we will look into it +Attachment has been added with description renedered has been added with description Rendered PaulI am a little unclear what the problem is here +I also made some error handling on +I dont think there are any new options relative to though +Ill try and patch this in soon +I just ran into it and only discovered this bug after extensive searching +Fixed the typo and replaced the isInterrupted checks with shutdown checks +Committed revision . +This is an incorrect relationship aggregate functions have been adjusted to work directly on. +regarding this just today I was trying to get some sys props to tests but couldnt make it work Ill discuss with Aslak tmrw +He is going to try and incorporate this code into MVEL so that you can use reflection or JIT bytecode accessors +This seems pretty much implemented by now isnt it +Since Ive not heard any about throwing Ive opened an issue to do this and am blocking this issue until its resolved +Colin are you able to make distributed MLT work? I cant make it work at all +Great st issue for new contributors +First attempt to download the zip stalled after kb of mb +well kinda +I actually need to get used to the new way of generating documentation +I was able to reproduce thanks for helping narrow this down +This issue is fixed and released as part of release +It should be created every time +I just formatted and applied patch to svn +What you think +I created it so that when we do upgrade drivers to another wtp version any commits related to that can be targeted at this jira +Closing issue fixed in released version. +Adam could you please verify if my guess in my last comment is correct and then close this issue +This one is a work in progress with +Is this still an open issue? Any progress? +Attaching fix to testcases on Ravis on the top of the previous patch +sorry +Default behavior matches and the number of threads per processor is still configurable. +Thanks +Current workaround is to configure the version range for the package to be imported inside plugin configuration or linked bnd file +Kris I dont know of top of my head +Marking as wont fix with known workaround rtfm +Weve had enough fun and games in here please its over years old +fixes the tests and the Hadoop classes +With just my fix the unit test failed but in combo with yours it passes again. +You can now flush the bundle cache the first time the framework is initialized +Missed fix date moving to any update on this? Could this make it to Beta +This gives us to oppotunity to experiment wih CDATA a little and see how we feel about this approach +Each user may add files to this listHow is it supposed to work as multithread? There is many object each have own +The patch is for +done +If people start using this new version with mutable position all around the code base were going to run into nasty bugs where methods disagree on who is responsible for position +setting original owner back. +There is no change in the code +patch to review and decide whether we fix or security reasons we dont store anymore the scm usernamepassword in the you really need it you can configure it the build definition arguments with D args. +SiI think Jacopo has committed my patchAnyway attaching patch for future referenceRegardsAshish Vijaywargiya +It is also not longer possible to cause an infinite loop +fixed +A WADL would really help our customers comfort level and our experience with Spring WS contract first concept +with a Why section that has a little more background on scenariosIn a word Statspack +However we should switch the +cache files. +Hello AshishThanks for your commit and suggestion Next time I will keep this thing in mindThanks RegardsDeepak license to ASF for inclusion in ASF works somehow i missed it in previous upload +It passes +Bug in Groovy preventing fix need to wait for new Groovy it is Drew +Will do more profiling tomorrow +Mass closing all issues that have been in resolved state for months or more without any feedback or update +The Runtime detection plugin doesnt change the deployment location anymore + the tests seem to work now I try to regenerate the site I removed my cached torque libraries as you regression tests now work without any problems + +In your tests the error dont occurIm using a Decorator that I made to format some dates and in the other pages i used aonther ones +The latest patch looks good to me +Closing resolved issues. +Running tests will commit soon +Attaching patchAdding static var and using that to decide if we need to run getconf or not +Will investigate further to find out the way to fix this +If you use an async backend or the JMS backend you might be taking a decision during according to outdated state we could hardly claim this is still having some transactional semanticsPeople might not be happy to waste more space in the index nor extracting a stored field is very efficient in Lucene +Please just to set the resolution to fixed +Heres a version that runs in seconds for me since it only checks out the top bits of the tree that we touch making it easy for folks to evaluate debug +No +Add new jsphtml +If there is interest from people to still have support included as an option in Tika then we could reintroduce it as an optional dependency that can be enabled through explicit configuration +Once something goes into then its public and we should continue to support it in future releasesAn alternative might be to use a field or method provided the unit test is in the same package as the parameter in question +Commit changes to trunk at revision thanks +This bug is on btw +Some have crept in to trunk so they need to be fixed +the single test failure appearts to be a test env issue resubmitting now +That looks like a bug in noBased on the above discussion shall we retain the patch that is submitted? It ignores exception in clean up only when regular code path has already thrown one +Ivo given my comment above does this issue now belong to Misty +What is your thought +so maybe we can clarify somehow but that is exactly what we are trying to show there +For the third time why cant we use the Metrics API here? This is precisely the sort of thing it was designed for noMilind says there is no way to programmatically add new metrics to the class +The empty Map does the trick +Files to enable this feature +So what error is the client getting in your malformed test? +Previous commits on this issue has resolved the places where the events should be published +Continue to look for alternative +Is this the same as +Closing issue as version has been released on. +pluginsrootnotebookusrlibeclipse find +The idea here is that we need to get people writing test cases for their JIRA and provide the means to test specific function but we also need some level of control about what is included in the CTS jar default test casesThe tuscany impl just calls the which includes the adopted test cases +I think this issue should be fixed in a Ubuntu machine I had a similar problem +I was sending up t o messages to a queue with slow consumers +Nice finding Enis +Does anyone not want this in trunk? I applied it and ran the tests +Should be able to attach column names with their values because Hive knows the column names +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Page shows the Loggers that are used by +Doh! You are right +Attached a patch adding HTTPS support to REST Management a patch adding system test HTTPS protocol in REST interfaces +BULK EDIT These issues are open to be picked up +I have just committed this to trunk and +Then when they get the specific revoke actions like what my patch sends the dependent objects will check their privilege requirement against what is being revoked in the and then drop themselves if their required privilege is being revoked +Infinite recursion still occurs under . +Move issues to next release +I also think the whole enabled flag we had is no good +Im not sure this functionality makes sense for the +The problem is the WSDL document which is by Ode when I deploy the testcase as a BPEL process +It should be regarded as a test as IMO has its drawbacks Nanoseconds are lost in timestamp as does not have support for them It only works for widget forms +Currently the Eclipse can do this but it would seem to make more sense to move this functionality closer to the core +Does anyone have the authority to do thisOr was JDO placed here to collect dust +Nevertheless we currently follow the guideline that its usually more appropriate to share a protected logger for the entire class ll close this for the time being +The problem with this is that the archetypegenerate goal seems to assume the artifact has a +Well have to add some property like managed so that users disable MINA from releasing it +But I prefer fixing it as Thomas suggests in the linked forum because deleting makes applying EAP SOA patches very complicated later +Perhaps implement as a static flag to turn it onoffOr simply put the flag in the which is easily accessible. +Perhaps we can update our quickstart with the windows script dir when is released +I dont have any way to reproduce the issue so please report back when youve tried the fix +Thanks! +This might be a issue but I havent had time to look at it yet +That reminds me a bit of? Lynx doesnt support CSS nor Javascript +I would suggest that you write a subclass of s +Ive just committed this. +Would fix this post fix later if required right now the inline script mediator is being updated to not require explicit initialization but still operate correctly +Committed to trunk +ive seen the same in +You could have tried to push the attached files to wguttmn as the jira list is for everybody but Jira itself +Just updated to latest nightly +Updating previous patch +Message delivery lock is acquired inside of close thus we do not need synchronization on delivery lock in Could you please review and commit the patch attached? +The tree store is not constructed since it is constructed in the callback function of DWR +I cant do it tonight + +Im seeing the same thing and it seems like a huge regression +I had this wrong notion in my head that directory names with spaces apply only to Windows and hence not added a regression test +you lost it is the same issue tools lets projects with errors slip through be the error in xhtmljsp or how does it affect documentation +Ive seen this in jstacks too recently hanging in disableTable +I do not have JEE installed in my Eclipse so please confirm +Some comment corrections Altered how the select operator was created for lateral view you regenerate the patch after refreshing? It has some was with a test output file test to revolve the conflict +Bug has been marked as a duplicate of this bug +This was added to the for H however it was not added for the Hibernate will need to port the H entity and collection cache changes over to the old +Patch looks good Ram +In this case I see no entries +Thanks IgorThe patch did not apply directly due to other changes made to Hadoop since you contributed it +Fixedthanksdims +Added extra Unit Test method including implementation and updated unit test +Thanks for the feedback everyone +still uses old code expr +I will see what I can this day August th as +I am surprised that this does not break many more testsThanks for the review +Lets help her do that +Are you using the most up to date onesUpayavira +The strategy handles features like caching +Im marking this issue as Fixed although there is still quite a bit of work that needs to be done on the themselves to work with the new classes +Work completed on revision the existing protocols of davhttp and davhttps are still in place this change is backward merely adds more protocol prefixes that wlil respond should ease the integration points for extra protocols should also reduce the number of incidents of issues and support requests around using webdav with wagon. +I hit that in a test case and it cause a really obtuse NPE +Would you care to finish this discussion in the forum +Can you please give me your entire mvn X output so I can compare +A possible solution would be to implement some sort of event queue in the entity manager were we collect all occurring events at flush time and try to events +Changed it to false by default and explicitly set it in looking more closely at the testcase output there is an issue with this patch +then it will work +Correcting the bug description this is not related to the location attribute usage but only the revision numbers. +If you favor the jars in HBASEHOME then theyll be confused trying to figure why their fix is not showing up though they run post build +Thanks for the reviews +Thanks for your contribution! +is probably one of the issues but i think after that gets fixed we can run into the issue that pat identified +We decided to move the Local keystore to the dialog that allows the user to choose a name for the new addingdeleting keys also in local store on deltacloud plugins +what about the Indexer patch +LGTM + use the Dataset class to get informations about the attributes whether they are numerical or categorical and to get all possible values of categorical attributes +guys is this supposed to even be in M! isnt this the work done for WTP in December +Committed to branch +I believe this is the same than and being tracked there. +They are never checked +Andy I have some code +Looks fine to me feel free to go ahead and in rev +Since this is a new feature I did not touch issue has been resolved for one month or more without complaints +Ill keep this ticket updated at least once a day til its done +Hello Johannes WeberhoferI have tried to build the trunk code with your patch +alertdef add Enable adds an it is not to enable the alert definition. +Ill submit a patch for soon +Thanks a lot for the contribution and discussion all +More complex mapping metadata is available in +There is no need to reopen this report +Reviewed This patch looks good +More updated patch for earlier version of Hadoop +I agree it breaks setup but it is no blocker as there is a clear work around +Hi Jasha just to clarify is this an expansion upon the locked attribute that Region and objects currently have to make the properties more granular +Start with testcases the first a test that confirms a stuck child process via SIGSTOP doesnt take down the regionserver +Right now I believe the only naming requirement we have is that queues must start with queue and topics must start with topic +Applied patch in r +hence resolving it +Do you have any examples for cglibThanks here libcglibtesttest is the cross language test suite important to use for compatibility was committed closing ticket. +Ive attached a patch that fixes this issue as well as adding another unit test to stop it regressing in future +Yes my plan for this was +Sorry! +The temporary hacks have been defeated as part of so wecan resolve this issue. +Fixed on trunk +JohnBut then these tables will always get created even when you dont want them since you are not using the feature that require them +HelloWould you please apply the patch? Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +What is the change? I would like to change my patch to work in the same directoryjar structure +This patch creates a new class and fixes a typo in the rootPretty sure this is all we need but Ill have my colleague at work verify since Im out of the office tomorrow +QE has nothing to verify +I actually duplicated parts of the additions to srcdoc as well. +It only needs the partition data to be in ZK and doesnt require all brokers to be up +Actually the will end up storing instances of specifically inside the class +Sorry to have taken so long to reply +I made a small change to but would appreciate someone reviewing itAlso I added an example to the documentation to indicate you dont need the protocol with a datasource and that it is in fact hurting when youre using a jvm supporting JSR +This has a possibility of being no longer relevant if we incorporate HTML scaffolding in since this is a bug in the existing Faces scaffolding +I agree that the null check can hit before the app starts where redirecting is usefulBut for crashing Shouldnt have already fixed the original tracking url to point to RM web age already in case of crashing? I just checked the patch and seems so +There was also some weird stuff in +The current included svnkit is based on a fork before license changed +For the non implemented functions should we through a non implemented run time exception instead of just returning + +Sorry +Do you have a small demo application that would allow me to reproduce this problem? Do you launch Jetty in Eclipse using srcmainjavawicketquickstart +No test needed as the patch only adds dependencies +Thanks +Duplicate of adding as management plugin to the is still failing the same from with eclipse rc and our latest nightly build Could stop the server and the server was shut down gracefully. +I cant find where the index is scannedtake a look at get and its countTerms booleanby the looks of how its used it seems like the assumption was that if the numTerms param! then youre clearly okay with it taking some time so it might as well compute the total number of terms in the index for you but the default value of numTerms is so in the adminlukeshowschema case it computes the total number of terms even though it isnt finding top terms for any fieldif we just add an explicit countTerms boolean param that should make it possible to speed things up a lot +Random modelstructures can be made now. +Having the user think of running tasks and scheduled tasks separately seems unnecessary as it doesnt buy them much +isa srl Yes PHP needs some memory to do stuff +Lets keep a simple example in our codebase to showcase this +So unfortunately since Seam is currently compiling against JPA I cannot simply implement the missing methods without changing Seam to compile against JPA which we probably dont want to do in a point anyone against me rewriting the code to use a Proxy instead of a concrete class +BULK EDIT These issues are open to be picked up +This should be fixed in +I ran into the same problem when running code on a particular Linux machine +I am providing this for purposes of discussion +Thank you very much +jsf bvt restrore has been passing. +Dependent on the post to rate a session should return the new rating +Fixed. +Strangly enough udig cannot connect to this database either the problem with the error reporting still stands +So this problem is resolved +My SFSB always needto nullify all home references before going to passivate andreinitialize them before activating otherwise exceptionthrown will occurs with activating. +for the review committed. +Oh it seems a problem with our custom classloader +Issue title should beGrails Cannot execute local closure from within init closurerather thenGrails Cannot execute local closure from within init closurealso reference to grails should be and be more careful about version nomenclature +Thanks for the work stack +There must be more to this issue +Fixed on also now. +Applied to +This includes only the change to openstack common and a test that instantiates the base Connection class with the timeout variable set +Committed to trunk +there is no here tomcat adapter +Turns out to be user error I was using the directory instead of +Do you now how to put them in the target directory? Do you have a patch +Andy I have a patch for this +More work on this is needed +Thanks Ashutosh +varcatalinawebapps to run them +Should we make all methods of not synchronized after a readwrite lock is introduced +Remove the parsing of showsdks output and either use iphoneos or iphonesimulator values +bulk close of resolved issues. +Looks good to me +closed +is the main issue for Java +Correct Fix Versions + Worse yet introduces sequential node ids +Looking forward to a green Common buildThanks again for the review Tom +New patch in junit patch while I look into this some more +Patch looks simple enough I dont know about the behavior you mention Ill go see if I can grok the standard on this +It works in CR. +bulk close of all resolved issues. +Committed to trunk +As I said on the mail list to reduce RAM overhead try use optimizeAcknowledgefalse and reduce the prefetch value +I can help +Could this be related? If so I think the VM is to be blamed not Derby +Defer all open issues to +CommittedOn trunk at On at +The issue is not with a corrupted jar +I got a fast failure on head of but then a clean run on the after a failure on the one +If you agree with this contribution I will update the wiki pageRegards looks awesomeCan you attach the +I missed out some infoOn iOS it crashes too whenever the user selects a picasa image link +In general I like to think of all that work as something the computer can do for me Baby steps though Im just pushing in that direction I know there would be a long long path to get there +Updated the patch based on the review commentsFor comments above the one next to the last the map should only contain one entry +enolan Applied patch. +Well release it all within an hour or two +This is deep within Tomcat +Are you using this class already in your code? And if yes can you please attach a valid sample that highlights the use case +Please try this patch including patch file for update in and golden files for serialization +Verified in CR +The x tests consistently failed without it but trunk worked just fine +A patch against +Thank you Ant +Can you verifyThanksAlberto +I checked the class into the trunk +Used rev with no luck +The also has a bug thats causing it to get stuck because of this +Added link to issue for tooltip test using tooltip for different targets +RichCould u please chime inthanks into this now +Please let me know if you want me to post some more information in the test case about why I did what I did +Is it too onerous to adjust the property to allow for jobs with large numbers of ranges +code looks good +I have not been able to reproduce this issue since fixing the I had in my application. +Reverted logging change +removed as in +Moved from to just after merge of old branch into TRUNK +Yet another problem detected by Jian Liao many thanksSending modulessrcjavaorgapachegeronimoschemaSending modulessrctestorgapachegeronimowebdeploymentAdding modulessrcplansTransmitting file dataCommitted revision +Thanks for verify Jaimin +It works when datasource xml is copied manually +Resolving this since the issue is fixed by the committed fix to +Of course only if it has been extensively tested and the chance for breaking stuff or changing existing behaviour is really minimal +Can you attach a gsp file where this has happened to you? Also do you see any entries in the error log about this? Aso after you see this problem can yo right click on the gsp file and select validate? Does the error go awayPlease comment on the STS issue that I have linked to +Tests ran cleanly for me on +What about situation when reopening take place a few days after iteration closing? What about situation when closing took place while iteration is current but reopening took place when iteration is not current any more +Heres an inital pass I took at this based on the OG source +timer start event +I guess Ive replied to the JIRAs notification mailThere isnt source code for itHoward +Well want the C brokers implementation to have the same look and feel as the existing Java functionality +Attached patch to fix the bug +Hi Justin thats much better thanks! Ive just tweaked the matching because it wasnt working for a class derived from Geometry +Attaching patch with mock task changed +I may be wrong though so any debug information would be useful. +EgorThanks for your formatting hints +All tests passed with patch v for branch +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Path applies cleanly +ALPHA. +Then I could build the project and run as local camel context +We got in our local repo thanks to need the FIX API jar from class is used fromimport +We want tokens positioned in begin marks +For embedded the default has always been focused on zero admin and so pretty much wide open by default and not secure +Something is a lot better than nothing here +I definitely think being able to index the key of a map is useful +Would you please try itBest is the conclusion here? Do we want to be exception comaptible? I think so +Submit the patch for hudson review and then I can commit it +So all issues not directly assigned to an AS release are being closed +Fixed with resolved issues. +This again ensures backward compatibility +What I have done is to define a custom that is used by the default +If the implementation is not thread safe lets mark the method in Java as synchronized with corresponding comment +The rule flow file gets open with no issues when i open in any of the browser +I do not see any recommendation in Jira comment +Added for investigation in related to workaround seems to work. +Please review and commit +I have No +Thanks +Reload the same patch to kick off another test run +Introspecting the class graph is expensive +I have started derbyall on with this fix and I intend to check it in on that branch too +the fixer of this JIRA either create a new or incorporate into an existing test a test that would create a for this function so that we do not regress? +So if the variable initialization flag is set to true then instead of asking the user if they want to create an initializer the BPEL Editor would just create it +I messed up the packaging of the binary +Thanks for the quick responseNot sure what should be included in the project? does it mean I have to extract out domain dependent stuff and use mock objects etc? that sounds bit more complicated hope there is a better wayWill it help if I deploy the application on one of my servers and send you the URL +And success the EAR file works perfectly now! Username letiemble +After having reviewed your patch it has definitely a better approach minimising code change and better readabilitySo Im for your thanks +attached failure is different however +Scott can you take a look at this +In the end I found out that this can be fixed by making the database using by default and then ask hibernate to use by default as well in the +Bug is fixed +It is useful because if is called repeatedly and written towith the old readers thrown away the byte arrays will accumulate asgarbage and this could affect the garbage collector +I believe this is a single user but test +The in is a constant +trunk patch with corresponding unit for adding the test +All tests now work +Until I saw the explanation above it was disconcerting as it looked like there was something wrong with the DFS especially since the log message is an ERROR +this is not for alpha since being released! Glad this will be fixed +I pulled again from the top of the x branch +and this bug looks alot like one I already fixed +Committed +Ive discussed this with mic unless a community member contributes solid and working code we have no plans to do this especially as we just arent sure how it will really make sense. +Verify that resource usage at +Getting an object reference to an entity via and performing on it +With tomcat this problem does not occur +take care of the modal window size +Throwing core exceptions is the approved way of indicating you have problems +Closing issue Oozie is released +In the case of javaClass it was in an existing schema that worked fine with Castor +Do you have code reproducing the issue? Can you try with the latest snapshot +RN marked Not required as this issue is not relevant to them +Well this hook was to replace the native cache there are other hooks to get into the native cache. +DONE +I agree with the with test AbhinayCommitted +Oh I missed that +it may be possible to do this declaratively using the +Updated wiki page for it +The attached findbugs results were generated using it might be the difference +but is Experimental may replace in a future release +Streaming is a neat hack but with the UDFUDAF frameworks I have never been convinced it is needed +Should COUNTER BATCH reject regular row mutations? I dont see why it should but need your opinion on this +Merged patch to doc branch at revision Merged to doc branch at revision Merged to doc branch at revision Merged to doc branch at revision +Lets see if we can fix that for has been releasedMoving all related issues to the next version +Ill be sure to develop some flow control tests against clustered brokers and have you review them +Willl clean them them later today +Applied at r. +First row should be tN not t. +I was about to get my hands onto this for the upcoming release but it seems due to some changes in the codebase the patch attached does not apply cleanly there a chance you massage it a bit to be applicable to the current master +rTested on ubuntu needs to be tested on winXP as with windows too +It applies against a clean revision for me +Could you rebase? +Will do +test case and mapping files to reproduce test case passes with Oracle on trunk version of +At the moment Im not sure what the best way is to resolve this issue however one option is to bundle the Forge libs in a separate directory +I just recommitted this +Please review that +no must have been from manual testing with the example ignore any changes +Momentary brain damage on my part +can you attach a sample code which compiles but that fails to parse +The API simply does not allow us to change credentials its essentially a thingoraps It seems that the people who wrote the LDAP authentication instructions are not reading these bug reports +to be release the code freeze for AS CP is Augist th +You can enable Tools Knowledge Base for your project if you want to have code completionopen ons +Verified on CR +Looks good +Attached the patch +The configuration outside works we need to use an other this property exists since +Added alter statements to allow for the update of view and procedure definitions at runtime +I guess this is confusing +The easiest would probably be at the container level ie +is the one receiving attention recentlyDoes this JIRA depend on +I did see some potential problems with that I have fixed upstream however I doubt they are your problem +I also added retry functionality to a few other places where it was missed and added a unit test for the bug you caught above with for addressing my comments Todd +The authentication and encryption should be optional as they are in SASL +The paths in the needed to be changed from to +If this fix is backported please make sure to backport as well to avoid regression test failures +This complicates things a little bit but I could expand the proposed method to pass in the appropriate properties which seems to be name and omitNorms but that doesnt seem quite right bc we may want other later +Once the test cases execution completes will upload the same +same problem on are right it came from dynamicAnalysis Thanks. +This is strange because the fix was commited in october +Corrected compilation and build snafus +Tested again and it passed the testCommitted +The biggest problem turned out to be not but the parameters with their route map of maximum connections allowed +See latest patch +This is related to which unboxed generated fields but not parameters +Never mind +Are you passing withing the headers Map of the AMQP Messages properties or are you passing it as the actual property there +So there is no new command the same command should create a snapshot irrespective of whether a commit optimize has happened or I guess this fix should be good enough for it and it looks good +My understanding was S would use file to configure this +Discussion with Bikas and Vinod offlineThe approach doesnt seem to be much more disruptive than the wrapperextension approaches particularly given the changes in +I am reattaching the patch with another name +Dick Im going to try to break some of this out into starting with agreeing on a design for the API +that makes sense initially that was just information so you could get oom +This is an uggly but quick patch redownloads every time doesnt check for dates and redownloads from every repository installing jar from the last one +Never heard back from Anil in May so sent another today +does seem to fix the timed out tests +attaching patch with the change sugggested by failure is not related to the patch +That caused the map tasks to be reported as having been lost and failedThis happened multiple times on different machines for the same map task until the job tracker aborted the jobThe problem was due to config file update during the job running +However it is really easy to overcome this I created an FAQ entry +I see two problems +and Im using Spring still fails as Xavier described a NPE when the DAO attempts to use its injected +Added a queue argument to the checkPreconditions method +So it will avoid some efforts wasted here +This patch adds unit tests +Its purely for the purpose of fixing the build +Having the mapper stuck at sounds symptomatic of a merge phase that is taking agesif that does the trick we can weave this into the code. +Here is the patch for this issue +Great stuff +Deploy an application without version twice and redeploy it again +Heres an attempt at fixing this but I think it adds too much overhead to gets and seems to freeze up other parts of our test run +Regards Ratnesh Divesh RatneshYour patch is in trunk rev +IfI get some feedback on the reason of the patch in htmlunit I will reopen itSorry for the lost time +Please note that the objects are Serializable through inheritance from the +Ill commit shortly +My build is now working within eclipse +Thats the reason for my suggestion toonly increase the buffer size for input streams that wedont clone like the base stream in the compound readerBut Im just sort of guessing here I think we should runsome performance experiments +The code differences look minor and bug +Since I was started off with rowcounter I ended up changing it some to suit it for the case I was looking at +If the blocksize is zero then return the size of the first block as the blocksize +should be returning a null instead of null +Verified on +Just a small one today TIA for review +ChamilaGo for it! +you consider using lucenes termvectorsSome of the feature extractions would be easier to do with termvectors especially when the index contains many more docs than the ones on which the classifier is builtClassifying a document from its termvector is also quite natural +make sure to reset the counter before calling initThis is ready for review +Have reached out to rubys by for CC details +Backported to at at bcb for master and a at. +had no internet connection last monthsFirstI wrote a dirty fix in my own class +Savage James +Ive just added repository url to on my laptop it was using RF libs from local Maven repository +The reason being that there are known critical bugs with and it will save users who might copy paste the example script as is +You should be able to script a way to sign all the files something likefind +bq +Screenshot attached +What is the best way to determine the IP address the server is bound to? attribute oftype +Uh thats not quite fixed yet theres another race condition thispatch causes after getting rid of oneSomehow somewhere the cache invariant is getting broken +Looks like a regression caused byr Get unbuffered IO streams from process if possible for +GianlucaHave applied the fix for this +Closing +Those that occur in the edits but not in the image +Attached is a patch to fix the test encoding issues withjdbcjdbcapijdbcapitools tests and found some test encoding issues which I corrected. +this was broken by you! +My bad +This patch provides a test case that shows the problem +Whether these are correct needs to be validated +Bring Peters patch forward to Add more logging thrift native added an escape latch property I used s as the min wait instead of s +use w instead of dw wont have error if the path does not exist +Regards my scenario above I dont think its possible +Here is the clean up patchIn this patch Removed the embedded styling +Jukka thanks for the motivation +It creates a single jar with all the other jars flattened into it +Added wiki code tags to the issue would like to close this issue +Thanks Amit good workYour patch is in trunk revision +This should fix both the harmless double processing and the problematic skipping of elementsThe patch does not include a regression test and none of the existsing regression tests have been run yet +daily weekly and monthly graphs posted somewhere publicly +I was struggling to find what could cause the corruption from there was suspecting the copy to the other dir somehow but the writer close makes a lot of senseAnd explains the mysterious segments file when I only expected segments and segments +bq +Well isnt it that you can use this API method to set a Attribute to the SOAP method as the API saysPlease provide your idea +Was that useful in understanding what is causing his crash? Is there more information which Tim can provide? Thanks +If anyone else is interested in collaborating or simply provide their ideascomments do not hesitate to contact meCheers GuysThanks for the effort here +Works. +Tom I dont think that other one was ever created +Ping +UGH +This was made obsolete by another fix +Patch files over Trunk and branch the test failures should now be taken care of in this patch. +The documents Im indexing dont actually have html in them but the schema was set up to use anyway +I havent ever run into problems with the schema version left out +SSL is supported by the JBR gateway as is and the title of the issue was re performing BASIC Auth so I didnt think this is what was being suggested +Thanks with revision Committed with revision +Hey Juraj I have commented in the +Thanks. +I will check to see what is the difference +Others do not +It also does that but adds other functions changes an API and deletes functionalityIts the latter two I am taking issue with reallyAny other comments from the lurkers +This fixes this issue +That is because use on LMP decides not to do that because the size of the segment which is bytes is more than of the total index size which is +WhyBecause we dont need there any more expr are dealing with Inspector in initialize and Object in evaluate so we dont need a wrapper like like before +Released. +I dont think we should be spending time on maintaining or enhancing this + +Why not define LITTLEENDIAN and BIGENDIAN unconditionally since we are comparing against their values? +Hi JesseSorry I didnt respond earlier in my project weve decided to switch to maven because of a number of issues with maven +somehow the connection change didnt make it in the first svn diff +I will have a look at itLars +so why dont you name the znode as hbasecluster +In addition I would also change module and contrib test compilerun to only depend on the to be compiled + committed to trunk. +We copied the source for and noggit? Feels like the cure is worse than the disease here +Attaching a patch for this issue +Initial patch comments to tried good contribution thanks Sergey +and no need retry hereSo we only need to consider scenario +This patch should fix this issue and should fix a couple of other issues +Currently it is not possible to restrict a list type +It looks like that code was from the original checkin and Makund isnt watching the lists any more +Calls to get in get where changed to the more specific get to get a object instead of object +Added to arch +is for Statement timeouts and does not indicate the transaction has rolled back +What happened +Old log files in old location will be ignored. +New patch correcting spelling in javadoc +fixed in +Yes this is standard Gnu Getopt processing where userfoo is equivalent to ufoo or u foo +I didnt see how was related but once again I probably dont understandOne thing I did notice after applying my workaround was that the collections in the loaded class would not allow lazy loading even though the session was open +This code no longer bears resemblence to trunkAdrian +Since this is an invasive change and requires a metadata upgrade its IMO something that would be hard to put in post v GA and since were only planning one v beta it seems like wed need to target that release if we want this inThanksEli +Also this aids to deploy the UDF module in single VDB artifact +But note that the fix is not there in it +Jim ping +Hermes ebXML Message Services MUST use JAXRPC Handler mechanism or if u are writing the handler are you are in turn using the Hermes API then you MUST write your code to work in a JAXRPC Handler +I have committed this +Just to make sure you know you cant use define without a name argument for the block you are definingWhat were you hopingexpecting to happen? A better error message? Thats a fine idea but that hardly qualifies as a Major bug +if you are satisfied just close it +Applies and works as advertised +thanks for reporting this issue and providing a patch! +Could I get a review +Tests in passed locally +Ill have a look to it laterI dont think that we use any webtools plugins +When execute returns false the methodget returns an int +Andrew can you please see if the patch you attached is still necessary and if so commit it and close the issue +The RM should probably exit if the scheduler thread sees exceptions instead of the RM continuing to run without the scheduler thread +These are in CVS now +As you must have noticed by now I care a lot about keeping the code consistent with the Java version and I believe we all need to aim for thisThis said I would like to know what the Java Lucene community has to say about this issue +Just to confirm the bug was fixed in ER the release note should inform users that both the old and new syntax are supported correct? +Lowered priority to minor +Also fixed the problem with Test Decommission failing because it was not progressing quickly +Added a section on using for loading and added missing options +I generated the release package and it looks goodThanks Giri +I have not made any changes to or databinding +If its something only on this phone and only on a version of Android that theres already an existing update for wed recommend telling your users to update their devices since this should already be automatic. +Just he directory structure is enough +Heres a patch that changes it globally for days but Im unsure if its needed I found out that futon does this already for a database during a session +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +The and can be used interchangeably for +As all the work is taking place in issue I am marking this issue as closed. +there is no current way to override those settings and say darn it +Hi MichaelCould you check out the latest snapshot or camel snapshot to verify my fixing? tickets +Patch looks fine +Next question +uses the sample code provided by Piotr +Yes it is a pain to get backslashes through +Issues resolved in +Committed patch a to trunk with revision The next step is probably to rewrite one of the tests to start using the repository and to make it available in the test frameworkI do plan to remove the code that downloads the from Apache because +With some more testing I think we should focus this bug on that problem and file an RFE for a new overridable configuration to solve problem Thoughts +I cannot agree with the closing +I have attached the doc patch please verify the configuration and apply +Maven has a complete rewrite of the dependency resolution code which after encountering the original version when writing dependencytree is definitely a good thing +Selective edit encryption on a per family basis is future work +I was running against M sorry. +Use Case Eighteen Statistical Monitoring Example When a connection tries to authenticate as an unknown user increment Case Nineteen Central control of TLS cyphers +It definitely seems like we could limit it to at most the overlapping sstables for a single L target at a time though which would cut it by a factor of +for the patch +It is only used to pass values +Add Security to the Application chapter updated +Havent checked and Mamta for double checking +Unless theres a fix on hand can fix this in +The patches here have now been merged into the trunk. +Closing as WONTFIX +Thanks for doing the additional testing James and for the updated patchI verified that the new test cases fail without your change to the grammarand succeed with your change to the grammar +I dont think this patch warrants any new unit tests +See is already address in addressed as of Story . +Hi Suresh I have a small question how a client gets correct of a path such as abc in your proposalThanks +Fixed on all branches good catch. +sh files +reopen pending suggested refactoring applied + +Here is a patch for trunk +This is my first contribution to a open source project so i am not sure of any guidelines that i should followThanks and RegardsBhanu Vanteru +Attaching javacores the first one of process id was from the rerun of that hung the second one of process id was from the run of just +verified by the patches author +The only question that remains open according to my opinion is what we should do with the left file +out of hunks ignored saving rejects to file srcmainjavaorgapachehadoophbaseregionserverMind rebasing +Adding another workaround +For starters the fsimage on all the namedirs will use D the edits file on all the namedirs will be on D and the on all namedirs will be on DAlso since edits and are on different disks it is never the case that the Secondary will readwrite to the same disk on which the namenode is currently logging transactions +I think it sounds like a Java bug +Please commit to and +For now I guess well stick with rethrowing the exception of the first candidate under the assumption that this is the common caseIf you want to test this i deployed snapshots otherwise you can build from trunk +WTP doesnt want to deploy it from lib folder so I left it in root folder +If we would set a clear goal to migrate to symbolic names then any step would be good +bq +Maybe the test should be changed to a scenario that still use Layer B streaming +If you change the Persons id also as a String and annotate it with Id it also fails +Reporter doesnt have the problem anymore so unclear that theres an issue +There can be two possibles explanations Either the loginpassword on batch side is not the good one Or those sonar tables are located in a dedicated schema +Both are under library or lesser licenses +xmlspacepreserve is probably not sufficient thoughI guess Im still a bit skeptical about the whole thing because I dont know of any other framework that does this +That was from memory +It then ignores arguments completely. +Why taint them with dangerous hacks for the sake of zero returns +Implemented and documented +Patch against version +the task could be done after linked dev issues is closed +Im all for updates to the code and build system to make couchjs build on as many versions of SM as possible but we should also consider what it would require for someone to build from scratch on Random Platform X +It would be nice to have this setting persist as part of the Eclipse particular project +Have filed for the change +These sorts of issues often come up when the native side is updated but is not. +yes but more generally this means that people do not read HAN reference guide on top of Core ref guide but rather instead of +Fixed in +Attach a patch +It seems that does not return until previous activity used in this method is blocking +And here is the patch for regression test +It works +accountName dsmiley +If it is accepted then I was going to provide a patch but Amith if you want to do this then feel free to do so +Re +Will move the comment upon commit +select same instance in look at properties view alias is not there + +I am terribly sorry I missed that mea maxima may I humbly suggest that perhaps if groovyc cant work reliably with the cpclasspath not the first argument it simply did nothing and reported an error if it happens to be called with anything before that? I cant imagine it to be a problem or would it? And it might perhaps save some confusion with complete dumbasses like yours truly who miss the help information +It saves money and increases security +Tim you probably referring to the XML editor since we havent released the full POM editor +The bug was in the way the class names were being generated for Vector expressions +Fixed by generating empty methods in the extension point class with TODO MODJ compiles ok now. +thanksbq +All worked fine please close the bug as not reproducible +I think this still needs to be fixed +is open for adding a digest to compressed sstables +Add the other tests for DB spatialite and whatnot as you see fit. +Committed +You should read first and then answer +Main reason is that when we are affected by this issue there is no work around +Thats the only remaining issue known to meThere are already a ton of similar todos out there +apply to nio modules +I added a mapping to test so localhosttestingtest gets the jsp +It was attaching the screen shots and hene attaching the document with screen shots insteadThanks +Any suggestions on where to start learning about how to buildedit source for the bigtop contributions +Attachment has been added with description patch file resolvinf bugs and has been added with description first has been added with description add spacepreserve property to text small test case allows to test the spacepreserve property has been added with description small test case to test the spacepreserve seems to work nowThanks Herve and has been added with description small test case to test the spacepreserve property +Also for Enunciate +Thanks Lin +split out a piece which I modified and then attached to +I will try using Jackrabbit rather than the workaround forThanks now its clear +Summary for clarification The BRMS WAR has its own log config that directs all output to STDOUT The default log config directs STDOUT to the console This configuration change is to direct STDOUT to the server logs instead +Is it safe to use the script attached to this bug? Which versions does it work with +Im sure someone said to me recently that every project was supposed to have a status file +Committing small fix. +This is related to +Since Object does not provide an implementation you cannot succeed in your attempt to call the implementation using super +If the collections are interchangeable why there are collectionsThey are not interchangable +It is also possible to create datasources based upon ODBC connection information but OLEDB I would guess presumes the use of windows +The principal programming model is programmatic rather than declarative +The fix shall be committed in the Jeff +What i described under A note about the nd case described in first comment in previous comment is a change that can be done as part of a separate jira + +java files unless you change the configuration +Changelog bugfix the last range in the bfrange entries was ending at translated toFFFFFFFF in the PDF file +So all issues not directly assigned to an AS release are being closed +Already fixed in Branch +Updated to Blocker status +So that new consumers can get their fair share of groups started after they connect. +Aside from that would you mind adding some detail on where this matters in your case? I can imagine a response large enough to take longer than usual over a slow connection in which case knowing the content length helps to show progress +I created this to note the earlier completion of this part of the original task +I dont have an abstract benchmark for this yet + +this wasnt very critical +Please remove the repositories section and check that all dependencies are in central bundle has been updated +As RHEL users are all suppose to use the init script to start tomcat as a service with the init script +Can anyone explain why these docs are in the site instead of a site? I filed to move them would appreciate your comments. +so the hubson job is failed with compilation failure +Jayson Solrs update XML does not define a container tag so we are constrained to only one of adddeletecommitoptimize at a time +There would be a row of circles for each shard and a column for each replica +I use Logback +Information required to proceed +We can even with interface continue to use the constants the same way use now but then interface doesnt buy you anything + is not part of JSR thus I assume the test has no meaning for that environment +Thanks for looking into it so quicklyUpdate Im unable to build the server right now because of is missing for in pluginsWaiting for this to be resolved seems like the pluto stuff was changed lately +box from is a logo on a feature named Developer Studio that should be the JBDS get this Developer Studio feature shouldnt show up in jboss tools though +Thanks Werner for helping me with this issue +Test cases are also fixed now +But thats probably not satisfyingA complete solution would be to underline statically imports and then let users augment this functionality through +Please review and apply the patch +I think we really should fix this for consistency +the current patch is doing that no intermediate znode between app and attempt sorry for the confusion I made +Good one Takuya +It is a combined patch for . +Committed. +I just tried and did not reproduce on trunk demo I continue +Sorry build master was hosed I expected the staging build to be operational when you made your changePlease try again +That way they can be recognized by an Aggregator or Resequencer that is downstream from the Splitter +Hey Jason on my Mac I saw some carriage returns M in the patch +I added some errors and these were detected correctly +When gets a URL string of NONE of none for a keystore or truststore it now creates and initializes the store but it does not attempt to read values from a test +I think is responsible for the change +So all issues not directly assigned to an AS release are being closed +Fix typo in patch +minor nit I would prefer the web services to output empty string rather then NA for the node +We should make sure that showcase components are updated at the latest the milestone after they are implemented +Committed +Thanks for the feedback + +Piotr please do provide a unit test it would be a great help thanks in advance + +I just committed this +Should we add statusbinary as well? Like we have statusthrift +Closing all issues that were resolved for. +Committed with revision +Had to have a deeper look +Spring already has a but no filter +duplicate of + would be great to register some Listener to the Filter which would make it possible to show a progress Bar. +Ive opened branches for supporting CPU scheduling in YARN +thanks fixed. +Initially well only specify http urisAvro should only specify transports that we intend to support +I successfully loaded a MB shapefile in one test I made +yes i agree it would be good to use separate jirasthe patch looks good +Applied Jacobs patch +Nice work Tom +Thanks again for the help +Heres the updated programmers reference that has the latest info about the actions +fileDuserapachetest +Moving on pom so it works with hadoop trunk +Ive attached a patch which will walk up the parent of the Navigation Item and look for the first component it finds +its not my fault +I was going to create a stubbed for testing +If I want to prevent this from happening I can run nodetool drain to flush my memtables and prevent any new ones from being createdThis patch violates the principle of least surprise by triggering a flush as a side effect of shutdown but only in certain cases +I will release bundles tonight and I will update in KarafI will close this Jira after the update to the released version of Jasypt bundle +This is induced in the fix has similar fix +is one example where we use shortnames +a +I believe this is a bug that should be resolved in a manner that doesnt force the developer to abandon the JPA specification and pursue a Hibernate specific solution +You have any new patch Stack? May be we need to work on testcases if the latest patch is fine with you +bq +Heres a patch to let distcp take an input uri via the f option +The attached patches are proposals for removing the last instances of the deprecated keyword in the Commons Collections JSE codebase +attach a patch to provide Checkpoint abstraction to make checkpoint +This looks reasonable +I can test against SVN head on a vmware XP or Vista image and see what happens +I dont have new snapshots uploaded but I might be able to do that later tonight +jukka could you elaborate which jackrabbit setup you are referring toSorry for the confusing language originally +In release could be further are in release +I also would welcome an explanation of why this isnt a bugPerhaps someone could explain it in the HQL documentation expand the Order by section from its hugely comprehensive lines of text on this topic +on commit +thanks. +But is should be able to handle interrupted exceptions as the occur in normal processing and continue if necessary +There was some problem with the previous patchKoji can you please try this one +Do the failures reproduce if you ssh into the hudson machine itself and test from thereI cant currently get into the hudson machine used the wrong username the other day and seemed to get ip banned pretty much right away +OtisSolr config was a standard config +reverted since it wreaks havocsome Maven developers are working with a Mac without problems seems the problem is more subtle how is your machine installed + +Kristian could you add some details about whats causing the bug? I didnt quite get that by reading the comments and the patchThe stack trace appears to happen when the session is being closed +Dont know yet if this is cause or symptom +Once patch is applied bug should be considered resolvedfixed +Thanks AtulYour patch is in trunk at r at r +We would have to call twice in order to trigger a deep gc +Now that we have the json metrics I dont think we need this one +To get Cassandra to run on Windows the following configuration items needed to be commentedremoved from the filepreheatkernelpagecachecascontentiontimeoutinmsBoth items were causing errors that stated class did not have the properties +New site deployed +i should have also mentioned that this patch was done by flavio and utkarsh +The pattern thats used to create a locality group would need to be kept in the +Attached patch changes to static as well so that connections may be shared in cases of multiple cores + +looks good pat +to elaborate I have cross checked all calls to artifact construction and within the core the code paths all go through the project builder and its validation routines. +I cannot reproduce this nor locate why an exceptions is m a afraid that only you can debug and spot what exactly causes the issue be it a problem in your installation descriptors or a bug from our side +we are going to commit the cpptasks jar file into the antlib directory so it comes with the checkout meaning you can remove the line in the file saying that you need to place the jar there +Test fix +Right I wouldnt expect it to be great with a single char prefix especially with the random terms your making +Affected to we wont release a has been releasedMoving all related issues to the next version. +This cannot be safely done because is public +Understood my only concern is that the mucking up with sed and stuff was considered to be iffy for the original patch so Im not sure we should be resurrecting it +Thought +It is isnt possible to use the test suite to tests vhosts +If there are other expressions that we want to rewrite using this mechanism we can file separate issues for those improvements +Rather we could have a facade class with factory methods for creating instances for adding packages entities etc +The reason is that this would work on all platforms in the same way and on of the first type it would avoid spending unnecessary time in awaitTermination +The above problem is fixed in r +Code that interacts with JMX and the dpeloyment service has been committed to CVS +After looking at the tcpdump output I think this problem is actually a server problem with Archiva +Rajith can we consider this as fixed now +The patch works just fine +Any attempt to use store is reverted now. +pranavDue this bugunable to verify this issue so reopening this verify and close this issue onceCS is addressed. + +Ill more work on this issue while we waited for I would like to hear your opinions about shell for mapredfilter classes and hbase mapredfilter packagesMay I touch hbase mapredfilter packages +File a issue for that +Ted How does your patch relate to Sebastians +Anyone to commit it? +This is exactly what my patch fixes +Ill open a subticket +The Tomahawk theme hides the page title so that is a problem in every screen +Committed revision +this currently only works on simple xhtml theme but not ajax are srcjavacomopensymphonywebworkcomponents srcjavacomopensymphonywebworkcomponents srcjavacomopensymphonywebworkstaticoptiontransferselect srcjavatemplatesimple srcjavatemplatesimple srcjavatemplatesimple srcjavatemplatexhtml srctestcomopensymphonywebworkviewsjspui srctestcomopensymphonywebworkviewsjspui srctestcomopensymphonywebworkviewsjspui srctestcomopensymphonywebworkviewsjspui changes to add an example in show case arewebappsshowcasesrcwebapptagsui webappsshowcasesrcwebappclasses webappsshowcasesrcwebapptagsui webappsshowcasesrcwebapptagsuiwebappsshowcasesrcjavacomopensymphonywebworkshowcase +Otherwise it is unlikely to happen in timeframe +Test case java files could be dropped intoejbsrctestorgjbossejb built test cases and my bent version of the file I added a target called xmleminjection to build the jar files and added a fragment in to run the jars. +At the same time doesnt initialize DWPT anymore it only handles and pooling and the actual DWPT creation is done on the consumer level +Im satisfied with the fix from our side for this by Chathura so far +Brad do you consider this one fixedYes this change has passed for EAP and I tested the jars attached in an AS Alpha build with a richfaces demo and all was well +This issue is fixed and released as part of release +Thank you for reporting this bugIve put a fix in CVSPlease let us know if you find any other problems with serializer +Changing the averages back to trendsup +to M from smooks editor is no longer being developerd so moving to LATER as agreed with +Is the upgrade needed for +Some comments +Note that the components pull is on the same branch as the fix for so no need to process that one twice +But dont worry too much about those details +Thanks Charles +Ah +Let me know if you have any problems +As Jochen comments does the package declaration in actually match where it is on the disk? Im sure that is related to this issue but it just wont fail for also said My +bulk close for the release +Very nice +Im resolving this issue the only remaining item will be handled under ChintanaMy previous comment still stands +One proposed extension the parallel processing extension allows us to correctly deal with multiple chronologically independent tasks + jobs services and message processors are tested and verified. +you have no guarantee to be in the same thread between ejb calls +Actually initially I tried to implement my album aggregator in class +why not we could also add a parameter to let people choose as its done for queries +If there are still problems which can be demonstrated to work wth but not please open a separate issue. +Thanks! +Otherwise its very confusingAlso if the entities should continue getting detached when retrieved outside of a transaction I would like to know if there is a known mechanism which automatically wraps a transaction around every request interceptors or seam persistence you name it +Attached +Patch for branch. +LGTM. +Information about structure of the deployment artifacts links to the XSDs and a brief description of the major elements that are part of a document needs to be in the Overview or readily accessible from there +in the contract web page +The goal would be to provide the developer the ability to specify some keyvalue pairs either at Eclipse project or global level +Patch problem +so ignore that warning +Zebra has the problem in a couple of ways for unsorted tables the index is not built on disk +Or on the other hand there are too many synch blocks in IW to make it safe to call into IW from DWPT +Did you have a script for this Emmanuel +I works on tabago +They should know what theyre doing and since they wont be running their eap with java they will know they shouldnt use it in their project either. +Issue may already be resolved may be invalid or may never be fixed +Ok if it worked in and but not in then im fine with this being added to the branch +Ran tests locally all pass except known bad +Please enter a bug into the PMD issue tracker if you want to see this fixed. +Thanks +I have amended the implementation class such that it now checks for a null typeName parameter value prior to performing the lookup in the cache of cannot reproduce the error indicated by the stack trace though +we need to ensure zookeeper handle is connected in +This is for trunk unless there is a new blocker that forces a respin +Let me take a shot at it +Jira to make sure we dont pull in the clover jar +The exception is supposed to be caught and handled internally but that apparently doesnt happen in this case +So this bug will be used to track union not working +do we preserve fs layout compatibility in some way +We have the use pools approach I think thats enough +PLEASE REVIEW AND VOTE NOW +Why we need it? for the same reason we need any of the simple start stop scripts +Can you explain a bit how are compared? Do they define any property on keys? If can faithfully implement OLF then split combination may not be safe +Thanks for taking your time with. +Applied +Note I dont see this behaviour on IE on +jsttests +HiI experienced the fail in parsing response as I write in my previous messageI tried to patch the XSLmodulessrcorgapacheaxisschematemplatecontained in AXIS project used to generate C codeIn attachment the patch starting from I hope this is useful any comment are welcomeCheers DimuthuI checked out the November th Snapshot of AxisJava which contains the WSDLC tool and the code generated with it now solves the issue I mentioned in this threadThanks so much for the great turnaround timeFrank + +yes in m you can open any module in parents modules by double clicking on the modules groups cannot be created automatically as there is no API for that AFAIK. +The deployer has been tested with my standalone test bridge but not yet with the cargo tests +Please add your votes to and watch that issue instead thanks +Regressions ran ok +From the very first look it seems like a memory corruption somewhere but cant handle where it is exactly +But the patch gives different semantics to the constructors of Path +This is a known problem introduced in +My process reads a file does some work on it writes it to output then saves the results to a database +I think the only symlink I set up is putting snappy into libnative +Issue fixed at Apache +Has new unit tests passes existing ones +Having problems getting access to Windows host but will hope to verify later today +Duplicate of +Updated summary and description for accuracy. +Btw if you can format your testcase of as a junit it will be good to have that as well +Since they are not public we do not need to deprecate them first +I verified this in Hive +This header config is request based and only relates to the one HTTP request between the SOAP client and the SMX consumer +How should linkdb updated? Or better should linkdb be updated somehow? I mean if page A has a link to page B and page B redirects to C should we set an incoming link to C from A +As the first step I am attaching a patch for your consideration that removes references to specific classes from the class +Patch added +As long as is the right version and this compiles its fine +So there will be an extra step +That make sense I would assign this as a feature request forbut we could make it part of releases in case you would create a patch +I like the WITH change +Thanks all for the reviews +Add being able to configure mapfile index interval +Deferring to a later Dave Since the fix has not yet made to official codebase I was thinking of attempting a fix locally to be able to move forward with Google integration as a whole +It should use the Type of the element in the Holder. +FYI there were problems with the pom contributor inside sorry +Files attached as requested by +Thanks Lukasz I review that. +Additionally XSOM is unsuitable because it fails to handle the more complex includeimport scenarios +Also I tried the junit test in and it also produced the same plan there so not a regression caused by changed in like the issue fixed in +Attaching a simple patch which changes the hfile referencing to not skip creating family dir in case there are no files to reference +Patch applied on with. +Need to think about how to refactor this +Because there are a lot of things with much higher priority +right +If you want to bundle the site with your artifact then you should use the sitejar goal and bind that to the appropriate life cycle phase +I dont usually have projects in the workspace like this anyway so I moved them out and I dont see the error Eclipse defaults to putting projects under the workspace path perhaps some people would break them out into subdirs in that location like I did in this case however I think more experienced users with Eclipse do keep their projects elsewhere much of the time as you implied Max +Better formatted version of Wednesdays results. +This is known the security rules are only applied when a instance is started but are not checked afterwardsA couple of approaches can be taken here Have the agent flush all the rules every X minutessecondshours and apply all the rules again +bq +Failure is unrelated +All the existing apache CI systems can already deploy any +first part is schema name second part is tableview name +Thanks Sanjay! +Sending trunksrcmainjavaorgapachesshdserversessionTransmitting file dataCommitted revision +Do you reproduce the anomaly with the latest nightly build +Hi PaulNot sure if Im missing something but I think this patch may not work for scenarios other than the simple option of a single filter being used on a searchA Matcher does not have the same utility as a because using a you can iterate across it using multiple threads +It should do null checks on the securityContext. +I havent been able to download the attachment +Im not much of a Codewarrior user any more +Can you explain why it is not a good ideaI think itll be confusing when this gets printed in the log since first the writers stack gets WARN logged then when the exception gets caught and printed well see the writers stack again since its part of the exception msg and then the waiters stack after that +much better relying on a rd part componentcontainerwhatelse that manages it and I dont have to become crazy +I hope this is the right protocol +bq +This will abstract the location of +Testing with the latest patch now +HiMany applications use and no one so far complained about such issue +The reason for this was as a test that older Serializations still work with the new API +Thanks Im not sure how the Strings get externalized so if you could do it Id take a look later and follow it for other contributions +orgapachehadoophbaseclient +Changed Target Version to upon release of +If the problem persists just reopen the bugreport patch did not work +We wrap cassandra in a custom web service so yes and no? I dont think that should affect this +OK Ive changed SVN mailer to use the new list for the new TLP directoryIll get the PMC members subscribed before moving the old jakarta list mailboxes could be redirectedmoved as well? This is the list that was used before the dev lists were merged so is historical only +Not sure whether this can be included in the project lib or I should the patch with plain zookeeper +Amar incorporated your comments +AM can verify if a file by name exists and use itSo the cmd line option logproperties to AM can be removedREMOVEDbq +Agree? You also said By keeping it around all the time on the component we can always lookup the active consumers +So this attribute should pass the JPEG header check +The attached patch implements this improvement for +fixedIll release a snapshot later today +Talked with Tom and he is going to wrap the ant task with the scripts so everything should be as before +I looked at this patch +And since they are a part of a it leaves us with one timestamp for each type +I will also make that change in this patchI take back this +There is no specific file in my Glassfish configurationPlease check the file where sonar startup is describedThanks +This error occurs with multiple users as can be read in the mailing list topic Cannot update Railo to latest release from July to July So I guess this issue should be reopened and looked upon +Type information is consolidated and streamlined +Edited the last paragraph to point toThanks for your contributionSimon Pepping +Either as a temporary workaround or inside a develop profile +I am not sure whethe this is java backend problem if not please let me know +I believe the reason that I was not able to connect at the end of my experiment was this the server was actually brought down + +Are you sure you have logged this against the correct versionI thought you were using +OK +Should be ok now +Defer all unresolved issues from to +Moving all issues to the CR +Andy do you have a small example which currently fails +It is very convoluted but what it ends up doing is print the max and minOccurs on the particle of the type of the element instead of the min and maxOccurs on the element itself +Ive got some serious back problems and was pretty much away from the computer for a few weeks +Thanks +Distribution fixes like the one in this patch are not easy to test anywaysThe code change looks correct overall +Other Relevant Resources on the topic underlines that in the guide we worked with Strunts tools installed on eclipse IDE +Seems to be fixed already +Moving out of +A temporary fix could be seen in the attached file the fix is currently commented out at +Still needs testinglooks better now +Pretty sure this was fixed at the time +The same applies if you use Byteman with an OSGI based runtime +I argue that theres no way to avoid calling it if you want the functionality reason for different is exactly as you say to avoid confusion whilst in the we add performance logging to CTI will be a desicion that Ill make after discussion with Juergen +Introspector caches are flushedBut still cannot be unloaded +Please just checked this on GA and the bug has been fixed +It solves the described situation +I think although the issue is marked as minor it makes Derby easier to use and therefor may have a positive impact on the adoption of Derby with starters +but in solr when designing up front i was just saying we shouldnt limit any abstract portion to when another implementation might support or BCP +It helps our CI servers to pass the tests +The same patch seems to apply to the trunk view on my desktop +Moving issue back to unassigned so it appears on reports of fixes waiting to be applied +That used in combination with whether childinitpool is still zero or not indicating parent or child process may give you enough to work out when to return and not execute the function again +Change the context of the startup metrics from default to dfsUnfortunately I think this would be +Especially with multiple appenders +Fixed managed X in resolved issues to so we can generate release notes +add hwitest and hwisrc along with supporting libs from hadoopcoreversionlib +MR AM already does something like that +This will be in the next update. +I will finish this tomorrowPlease check the patch +Thanks for the review Lars. +Another workaround create a publish area using the View before publishing files. + +It would be handy to add some docs somewhere to explain the different actions +in revision . +Reassigning to tucu after discussing this with him +which must be handled at some unknown layer above +Also if we want java serialization to work it gives us a leg to stand on +Im OK if it is not allowed as namedoutput name or multiname nameThe reason is that there must be a separator character to avoid filename collisionsIf there is not such character the the following named outputs could be configured for a job FOO with BAR produces a file named FOOBAR FOOBAR produces a file named FOOBARAnd that would mean that data written to logical locations end up mixed in the same physical location +SamWhat I can tell you for sure it that it now works with the simple front end in +If I remember right when I got the directory with replication stuff working and more tests using ram dir and what not I started seeing unrelated test failures pop up and has to do some test work to address them +Fixed with rThanks +But as an example it would work +released. +Rename the references from to by Brett in rev +r. +Len as per IRC discussion I have backported parts of the book now from +Geoffrey +Hopefully that will reproduce this and maybe find other issues +This is a fix for +Please correct it +Should we consider pluggable interface to enable thisThat seems reasonable +Im doing a JSONP request and there arent good ways to detect that the call has failed with a response code +I will commit it to as you guys suggested +I had a quick look at the shading plugin +Ive applied to patch to SVN trunkThanks a lot for the contribution! +Thanks Arjuna +Hi Simonit doesnt seem that the suggested patch went already into the current src codeAre there any plans ifwhen to integrate this patch or a similar solutionBye the trunk code we already do a bit more by adding referenced classes from the generic types +The changes inside the block in question dont change anything about its behavior best I can tell +But please not in the way your first patch worked +here is some confusion on whether the testcase is rightgood enough +Take your time to work on the code no need for a hurry +fixing desciption revision +Thanks for the fast responseI added that in the plugin configuration still the same FNFE +If you want to test quickly you can post a message from the stream or public page and the switch back to the user pageCan you verify and then resolve this issue as invalidI think it might make sense to open another issue in referencing this issue that is aimed at getting new messages posted on one of these pages to trigger a refresh of the public timeline +It isnt valid anymore + +Assuming it is okay. +Ha! I now see that this is what Alejandro already suggested +Dynamic storage CF does not have added to eliminates most if not all DESCRIBE issues in no matter what kind of table we are dealing with +Best to follow existing tools like IJ DBLOOK model so it would be easy to use and to enable derbyrun option for this tool later +looks good mike +There wont be any test cases for this as the patch is in the ant build script +Also added a test for the class +Probably by rev +thas is amajor feature of me i want keep using +Attached a patch to fix logging statement +I propose that we have a canonical representation of schema default values that has no external surface area of a third party library as part of the Schema API +Bulk close of resolved wont fix cant reproduce etc +This is not a bug it is an inability to read the documentationAnyway the necessary documentation is in the examplesadvancedThats an example for doing Japanese +I was thinking about that it would require to keep all of the files open do we want thatIs that a big deal? I suppose if you have its starting to be a big number of fds open but even then it doesnt sound like such a big deal especially given its for a very short time +I think t reproduce for current trunk builds +implemented cos review comments +Made some adjustments to reflect the underlying issue +to this issueIve got the same problem +Thanks +Please reopen if anyone has a good test case. + +Otherwise it will be hard to keep track of them +Just for the record +Everything is fine! Make sure to do ant clean or remove the stuff manually +We shouldnt be signaling a process unless it is in response to a jBPM invocation +Hi Ken +Hey Daryn any update here? I just bumped into this myself +thanks +Um are these all flat deps? I would guess not +the screen I dont know what happens +Can you also review the patch? Well need this soon +Issue on Windows markers are not created on the correct by David! +Should raise that number and run interactively to use +Aniruddha a dump of all thread stacks would be very helpful in diagnosing the problem here +is not related to the patch +No +I believe this issue went stale +bq +Those are as much a necessary part of the framework as are the jars +Thanks for the patch Dag +Most database vendors provide versions of their product that can provide some degree of both properties not sure where Postgres is in this dimension but you can be sure they are working on it +Should be fixed by +Youre right +Retry for a from hudson now build has settled +I recommend to stick to session tracking which never shows the session id as part of the for the time being considered a problem with the Servlet APIs session management rather than with +Initial patch will be dependent on hadoop if we need a backport to legacy api thatll take some more time on best effort basis on my part +Hi Matthias in fact we thought this use case was not possible as for us a is fully useless on a test package but perhaps we were wrong +How far I am taking this? Am I just focusing technical content +The definitive fix is now in so that makes this patch obsolete. +Lets open new issue for any specifics needed +special purpose component +Hi Fernando is provided by in my memory both tomcat and jbossweb packages have this jar but tomcat doesnt have repolib and jbosswebs repolib doesnt contain this jar so the solution is adding a repolib for tomcat or adding to jbosswebs give me your comment thanks +Any chance of some coordination as Id like to get some patches accepted for apache at some stage! +Kindly use this patch +We can address for version a way to both use a Array of final fields and a array of bytes to avoid GC operations on the temporary repository. +Still you are not answering my question on why you need to expose fileId on the Web HDFS REST API +definitely a split brain issue +I get the same test case failure on my linux box +If ADF BC Wizards dont support this that is a limitation of those wizardsIf JDEV were to remove the reverse filter THEN wed have to do something about it +The will call with metrics property names +Then user can read the files but not write themLets not reinvent the wheel +this is also how i develop +Also I will commit it to the trunk +It has been doing for openejb release and release +Ill check a fix into the branch +That is not a very good work around because then it will not work with terminals that conform to the standards +Great patch Yea ing on documentation references is bad bad bad thanks for checking updating them! +Thx +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Forgot to add having a get or get method looks fine to me as in after release +I added +I would be very happy to work on once this major milestone has been achieved +bulk close of all resolved issues prior to release. +ack mirror removed from list contacted the owner hartman to owner was removed from the list may Sofar no reaction from the ownerHPP +Do you want the installer also unpacking a file during installation even if it is already on the target station? This would be a rare case IMHO +I agree +Feel free to give it an early try! One step ahead of the game as always Thanks very much for the info just goes to show I should have read the roadmap before posting. +Locks would have to be ephemeral nodes to prevent permanent locking of a region server when the failover region server dies +Since the git repository is not yet set up I am attaching the code here so that the community can benefit +Actually sorry theres a problem with this code +Thread dump of the server under test after this issue occurs +Thomas can you review the Tools document to see if I need to add anything to the feature matrix +Patch with updated all issues that have been in resolved state for more than one month without further comments to closed status +is a related issue +fix patchs problem with fuse wrapper + HenryI agree that the README is explicit about that you need to have the C libraries in your systems library path I just want to get the information out there for people that dont exactly know what options they have like me so that I could take less time to get it running +Hi Lukasz yes all my actions extends and I have just found what was wrongThe Actions that have a Constructor the Container object is null so the class use the instead of my custom one +ThanksIll deploy the for test +These wrappers are no classic wrappers in the sense that other languages are using they are not there to represent values of the runtime and to allow faster invocation paths or something like that +Yes Im sure +sense I like what your patch is doing and will apply it on trunkThanks for the contribution +Also were making this change at a version boundary so we can announce the change cautiouslyAs far as the GWT stuff goes would it make sense to spin that off into a separate generator that inherits from the standard Java generator and therefore can make its own adjustments +I used Google chart tool for in my intern session +bq +Reviewed and tested looks good to me +is modified to log the cleanup task as +In r implemented workaround by always generating fieldInitializers +Seems like the patch is outdated could you please modify that and attach again +Not really for existing sourcesinks +when ready do localhostdate when see Done in browserAverage of four runs eachUsing pelt skin +Christian as you are the original author with nothing changed otherwise can you please import those classes into and update the license header to our ASF standard one? txsA srcmainjavaorgapachemavensharedutilscliA srcmainjavaorgapachemavensharedutilscliA srcmainjavaorgapachemavensharedutilscli all important classes are now either moved from other modulesIP clean imported by their original authors or got rewritten +committted +Hi DanielSeems like a duplicate of +Great fix it was breaking our Hudson build which is now running smoothly again +Lots of ideas but no dev +I found a bug deep into the reflection code +On exampleIf you want to get a attachment from gmail you will only get a content url +Gregory Martin Ralf how does everyone feel about adding a converter to +As discussed in an email thread BPEL is part of the ESB platform as opposed to the AS platform +First patch uploaded +This is currently hardcoded at +Now I am working on few test cases and will attach another patch soonThanksSukhitha +I agree with Luke +Closing the bug for nowOleg +Based on BPM Console +Committed to trunk +when you had find it problem +to trigger auto testing +Well sorry in fact Musachy did not say that he used it in the trunkBut anyway I was right it is used in the trunk inside and of the simple template +Ive switched the apis to use arrays instead to help find all cases where such things were happening +to updating the Github mirror on each svngit commit +Ive fixed a bunch of them and committed the full spanish translation +Regarding the file Yes that is right +Give it another try +Thanks Willem for the checkstyles. +Im just worried that any implementation of this request will make the system less stable and that isnt ok +I dont know what a pull request is +If Spring MVC provided some other means to control how the result is marshaled such as setting a media type property on the I could get the desired result +will give you a myTool variable in the contextLet me know if that is enough for you +Same occurs for the removal of proxy connectors +I have backported it upto so farAlso I will leave the jira open for another week in case if Tony finds something while trying to reproduce the DELETE case +I am gonna close this ticket as Wont fix and well have to wait for or whatever is coming out in a buildable form +This ones not quite done yet +We could gracefully redirect the user but this is not a security issue +Patch for supporting addition of credentials based on a request patch provides one possible approach based on passing the Request object to a +Committed +Also the script do not have x for unix environments +Moved the filtering code into and called it from Table +We have two cold backup servers and both try to become the primary +That works for me +Please take a look at head if you are unclear on this issue +Json neither I believeProbably createcreateString might be better +And I wanted to split the output of different categories into multiple files for the wikipedia example which btw is not necessary for Bayes it reads of all the files anyways +Can you describe what exactly behaves wrong according to you +Retries exhausted it took too long to wait for the table randeyuvg to be disabledIts a hbase bug +Sure I will submit a patch +Changing the component +Once the profile tool the Profile View takes control which there are no shortcuts for +for this release reads one chunk at a time and has only a buffer leaving larger buffers to subclasses +Most improvements seem to come from not converting the second parameter if the return value can be determined only from the first parameter which we cant do in old UDF functions with wrapper +Does not work +It is expected to see unit tests failing with faults in place +Thanks for the updated patch +id thought about having a polling adapter that asked Activiti for all open Tasks +Olexandrplease test setting the log level to info +Hi KarimAny update +Why is this bug getting absolutely no developer attention +renaming basically CLI doesnt support command line arguments like fae +Over time more and more tests will be fixedand an increasing number of test suites will be able to be run in your configuration +Moving to future as it needs more discussion +Personally speaking I dont have a strong preference for the indentation of throws +I believe HDP and both use Hadoop +Will commit removed the auto key generation Committed revision +when the logging config changes it will trigger up a rebuild of the logging objects and the old objects will be disable and removed +Resolved in. +Yeah over time leveldb is intended as the preferred choice over kahadb +testsfunctionalorgjbosscacheaopcollection +Can you put it in ? without it Postgres metastore is broken by default in the setting for Direct SQL needs to be disabled manually +In Netbeans with the trunk mevenide and version of nbm set to it works +Sorry to close again but as time goes along the chances of doing anything to this situation are very small. +cheers Amit +Resolved +IMO the extension approach increases the flexibility of the wrapper approach without adding too much complexity +Is there an implication for the regular +This could have accentuated the condensed looked of the pageThe current screenshots are in normal size +the whole class got ditched problem in +verified at r thanks Paulex +I looked at this and it seems like a very straightforward port of the changes to +I was able to reproduce the problem but it was very random +are the memory nodes in working memory or in production memory +Any help andor suggestions will be highly appreciated +This is possible to fix though if there is demand for it +This JIRA was in a hybrid state with a DAS component and a target release of SDO +I think this may be resolved in the trunk as far as memory consumption goes +or maybe a officermember a senior officermember +If its set to false then the service expects the meta data to be keyed on an ll upload the project with the modified service and test cases again + the fix to at +I believe all the issues identified in this JIRA have now been commit is . +Todays MR Apps is and it has signals as to what happened with a container launchstop succeededfailed +Itll make it into +They are few in number and so having them carry finalizers around would be pretty painlessFor individual objects of shorter lifetime finalizers are more of a problemIn any case what I really want to know is if you are objecting to my patch +I havent looked into it yet but what we need to do here in a nutshell is +The command is meant to do this +URL Handlers has been rewritten and included in the core framework +Duplicates can you have a look at the test case inAccording to my test run is broken +Ive been reviewing this and +I didnt know this bug was here when I filed it. +This issue was filed for the initial version of the +Reporters of these issues please verify the fixes and help close these fine. +Main fix was to patch up there are many other improvements to get the scripts to work in Cygwin MKS bashtcsh etcIm satisfied with these scripts now since the jars built with them pass the JDBC suite +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It has also been updated to sequence the latest Teiid Designer file +As noted this is really a bug in Commons Codec which provides the Base class and static methods being used by +Indeed in the first version of this plugin the exclusion patterns will have to be provided through a project andor global property and according to a dedicated DSL +Forgot to do this in trunk +So wherever it has to write or it writes down whatever is availableCases where the maxScore is not asked for and the value is present in it is writtten downMay be we should ignore it if maxScore is available without being asked forid are also written down in the same fashionShould it be handed over to to write the values so that the context is available +Uncomment it and it should fail +Any idea how to do that +If I have time Ill see what I can bang together around this once thats feature complete +Guys again keep in mind that this log calculation is less than of the time spent in this skipper initYeah just having fun +This issue is closed now +Oh +I committed this to trunk and. + to issue as it has been released as part of release +I contains the version of my local code. +Its not so bad anyways having this filter package pollution since its in client facing code and clients need access to filter stuffWould think pollutionimport Should be pulling in a class into an Interface like this +Would you really want a tag that generates your head and body tags? Even Struts didnt do that +Mark are you in a position to provide us with a minimal set of XML schemas to showcase the problems mentioned? +It avoids null schedInfo +The code has moved to the resource resolver module +Instead a uses a counter to manages the number of concurrent block moves +the updated patches and updated my jqueryAlert branch on githubCheers +I believe the current intention is to make it the same as the SOA for not noticing this previously +has gone in +added transaction support using a new option useTx with default of false and added appropriate tests +I added as a dependency +Committed revision One question What should happen with? Should it be removed or should it be modified +Ah great idea lets hope it works +You might consider downgrading to until I can fix this in. +serverproductiondeploylib +Checked the bug with the nightly build There are no shifts any more Great Job. +To reproduce I have a dependency for which i set the filemode of multiple files to +Closing. +Resolving this issue as the required information is in the Administration and Configuration GuideSee Bug for details +We already have a checksum on the record but were not handling all the cases +I get group names that are nothing to do with the project I have just added +g file into the +bulk defer of featurestaskswishes from to +Thanks Serg this may be something we can let the user customize with the binding file +code review +How are you adding the second binding? Each binding is uniquely identified by its bindingKey +Looking into getting that undonebq +Committed Thanks! +We discuss this question with Tihon Kuprievich +issue has now been fixed and the updated docs with this fix will appear in the next CP to next to the EAP maintenance lead in the event of another release +Id propose adding a new method to that is about explaining why it wasnt foundI need to get back to and move that forward +Im therefore going to leave this ticket open +Similarly if any files were copied use svn copy +The attached patch removes the captured character in unread +I agree but I really think should be solved first +We could certainly use for hashing out the overarching facelets migration strategy since there has already been much active discussion around it +This needs to be investigated +If a zone object in listZones response includes vmwaredcId property UI will show Hide dataceter button insteadReason behing not adding vmwaredcId property to listZones is due to each zone might not have hypervisor OR a zone might have a mix of hypervisors +It seems especially to happen whenever the cursor traverses over a long line of code. +Its specified by the category +We have to leave the deprecation in +good feature is there any way to provide this config parameter in way +Hopefully in such a way that we dont have to code it many times and yet without forcing it on custom converters +It seems Web in AS solves this issue together with Atmosphere +added revision . +it should wrap its code in tryfinallyThe problem with the broken functionality in Wicket will be addressed soon +To this respect I have already opened to review the issue. +So this does not solve that issue +The has intentionally been omitted because it is assumed that the TCK war is deployed in an environment that already contains the +The metric stuff gets obtuse pretty quickly especially when the parallel scanner is closed and all the threads are interruptedcanceled Slightly related I think pushing the metrics via Scan is broken +This policy will allow to process exchange only on a single node of the cluster at a given time +sshould be movedshould be moved to the new ticket +this is the same issue as reported code submitted for this is exactly the same as for did you guys submit this twice under different names? +Wed love you to claim the credit for the new feature from a git commit point of view but if that isnt possible we can progress it ourselves +We can do some housekeeping some other time. +Disabled formatting. +Im wondering why this issue hasnt been fixed for weeks +Icon for split visual source position has been left old +And then I posted in the thread again about the scaling problem I then encountered +Thanks Andreas +Hi Alessandro! Sorry for the delayIm not familiar with LUA however being a committer at Thrift extends my knowledge on multiple languages from day to dayI think a replacement of the Thrift Compiler with a template based approach is a huge effort and a such generator that supportes languages has the be heavily testedAs mentioned I really see a potential for that approach for the following use cases testing purposes prototype new languages generate custom documentation generate proprietary codeThanksRoger +Id love to commit this but without a resolution for the JVM issue mentioned above +is nothing i can to from here because that is clearly a mavenissuePlease raise the issue at the right place with a reference to our discussion here +I also added an initial database creation and boot so as not to skew the resultsWith IBM SR I saw the following measurementsTime for first database bootTime with securityMechanism Time with no securityMechnismWith Sun I sawTime for first database bootTime with securityMechanism Time with no securityMechnismIt seems especially with the IBM JVM this is still an issue +It contains the url to the AM container log +Is it worth raising seperate feature bugs for these points +it currently appears in the of the window rather than relative to the button that triggers it +The has a set taking in seconds hence is unusable for such control +Some more ideas for payloadsI dont like how we double every position in the payloads case to record if there is one there and we shouldnt alsohave a condition to indicate if the length changed +Do you have a test case for this? Have you tried using a +Thanks for the scripts Mark but now we have some direct tools support in the laucherIll close this issue to +I also agree that we can remove similar code from the stored procedures +i would rename it to path thenthe rootPath is the resource that was addressed by the post +It allows to enter customer screen shipping address screen and after asking shipping options it always returns to shipping options +Am I getting lucky or are you using a build process +Triaged for assigned normal urgency +Please try again +I agree that getting the formatter to work reasonable well is the first pragmatic goal +process the command line parameters and strip duplicatesRather than do some post fix like this wouldnt we have a consistent approach? Why introduce JAVAHEAPMAX? Either respect it always or discard it I would think +Tried with JBDS L installation ok no issues found it works great +Just doing some final testing before creating the PR +Doug this sounds like a simple fix that you might have already done so Ill leave it as R +Moving all issues to the CR +we use backslash to escape glob special characters +BTW the classloader isolation issue is +Ive revised this for for the quick resolution. +I forgot that even though I wrote it +I suspect that this is a problem in the decorator or the network server shutdown code and that its just a coincident that it happened in +sorry for my mistakeThere was a dependency in the maven file but the scope was defined as Test +Makes it easier to exclude them when doing releases. +Fix for this bug implemented by subclassing Shiro class +So strip of the + patch which restricts the change to only be within +bq +Thanks for the quick reply +String concatenation is not awesome +okey so we have to go with option thenWe should open a request at eclipse WTP about reading to add to the classpath +Agreed thanks +Does that fit in this project or is that something I should post in UIMA or so +Synchronized Source and Visual Editing +hinder the developer +Also fixed spacingtabs in +Owen Could you help me understand how it will be more difficult than it is today? Also do you have an alternative recommendation +Not in standalone mode anyway +No need this fix. +preparetasklogs method should return PREPARETASKLOGSFAILED if there is any failure currently it returns PREPAREJOBLOGSFAILED +is +h file and as a result sized a static array in the exabyte range on my build machine +when accessing the solr new admin UI httplocalhostsolrcloud the error message window pop up saying error getcloud cannot call method attr of undefined +Thank you for the test case it will help us resolve this more quickly +add merge of results for server modeadd true flagupdate emails subjects Vladimir the patch was applied to SECURITY module at rPlease check that the patch was applied as you expected +This is the new patch for according to Lars suggestion +Once I create the account will upload the patch to review board alsoRegardsAshish works for me +Considering the fact that this is closely related to the series of jiras listed in Im converting this one as a of +Thanks from me too Guillaume! +doesnt get rid of ALL JCR references but I think it is as close as we can get without breaking backwards compat +So thats my fault +Mailing list request submitted in revision +This was related to our overlapping work on the node dependency autoinstallThis issue has been resolved with your contribution to the create script last week I believe +In simple test application status displayed properly for a all provide full page source +My guess is that there might be some duplication in Daniels and Normans work +bq +Such subtle issue worths to be commented in source otherwise this fix may get lost easily after some cleanupsbeautifications +Is this still the issue +Here is a partial patch when the retrieval of the Ethernet MAC address fails it uses instead random data generated using +which simply inherits from at the momentThenThis leaves the default choice uncontracted and open for change later. +Done +We could just drop them from though they have not been deprecated +A simple way to reproduce it is to create a jar project containing the same file in both srcmainresources and srctestresources create a WTP project link the jar project to the WTP project run the WTP I hope that by link the jar project to the WTP project you meant to declare jar project artifact in wars project and then running Maven Update project configuration +See there are more problems than just the this is being handled in another issue +I just committed this +We have had DB getting deadlocked with this case when there are too many updates +Ask for suggestions on wherehow the feature should be added +Im seeing diffs sometimes and other times notDo you want the RSS XML +You are correct that CSLM does not override putAll so there is no benefit to be had with that approach +Please test and vote for on the user list +This has probably been fixed rewritingWill reopen if needed +The methods that calculate the safe number of required chars bytes will be added to more complicated calculation might be required to avoid waste but Im not sure if iterating the data will cost less than allocating more +Isnt that what you meant +Fixed by switching toI am running into this exception still using the latest from head when doing http session repl load test +Like an appropriate constructor etc +Verified during test cycle looks good now closing. +An IT for the is quite a bit more complex since it never has a local archive to test against +Manually tested +please take care of this +Its a contribstreaming fix that is important to us +Duplicate of +Languages affected as c go java javameI removed the set +Heres the feedback +I created a patch which is attached next +Pls check both the patches and provide your comments +I dont really have a preference regarding the legacy parameters +there are no limit we wait for the server to build the full result and we send it back to the clientI any case we dont need the Request to be stored am I totally wrong +When with myeclipse will be invoked twice per project + +Currently you can debug the only with from my point of view +As was pointed out on the summary above is merely a description of what was discussed on the call for preparing for the security session at summit + +Awesome! Looking forward to see more projects from supported by Whirr +patch patches +However the mailing list consensus was that this is a hack and the queuing logic should be fixed to be correct +The encoding of the parameters works now +Resolve this issue since can not with latest kit. +This should be usable for any rcfile data +Having it as a junit test will let it work properly with the rest of testreporting +This patch is not compatible with branch +Please make sure this is handled for the RC build on Friday +Hi Andrzej A patch is available that you can apply quickly to remove the text parser as an option for pdf +Need to investigate this a bit more +Also mention which snapshot this was against thanks +This turned up another problem in the clientserver configuration the numericdecimal tests in raise an overflow condition +This is a request to add a new site goal supporting the use case mentioned above then. +Async HTTP? yes we do support it +The patch has been appliedPlease verify +The right encoding for character is +nrm file with wrong norms mismatchedfor different fields +It uses compression +in an SP for have replaced Vector for members and pingablemembers with a COW list +contrib tests +Early dev builds will be availble in a couple of months +Im also against more tests with sleeps +Clint I also added a README to the top of your contrib package pointing users at the package documentation. +Note that also has this bug make sure and use along with +I ran the tests locally and verified the only failures are the ones due to +I dont know how I was able to get an exception ie have libhdfs call get on a but I cannot replicate itI will just add a unit test to to check that DF works +Adding this class doesnt break the API but it does changeit +Added support for arguments and some regression for patching +Another change Id like to see is to pull the snapshot war out of the maven repository +I am unable to reproduce this issue on the Windows machine +I was hoping to get to this this week +Anyway thank you for your reportingForrest +Any news on this +and this fixed the problem presumably +This appears to be due to a configuration problem +Attached the reviewed file really minor corrections. +because we all know that no one will ever come along and fix that issue +Hey LucaCould you comment on how this patch relates toThanks feature is coolFrom a point of view it would be nice if the namenode can continue to log its edits log to a local file as well as write to +I will attempt to download and recreate as wellMarking this resolved as the trunk and branches both process the wsdl fine. +Thanks for the policework Uwe +not seeing how it could happen yet seed does not reproduce for meLooks like there was a jvm crash as part of this run as well +Reopening this to keep track of progress in WTPs Bug . +Any more update required on the patch? Please let me know +Fixes JIRA +Would it make sense to log a warning after we set the hostname to ? At least in the case where there was an exception maybe we should log the exception +In the callback of the first I issue a cancel and expect that the callback of the second doesnt get called + I was looking at only +Great! Thanks VladimirChanges committed into the JBPAPP revision be part of CR +Updated patch to use the default +Sorry for so late reply +Right so if you look at the names of tokenizers and filters there they are super light almost like the background +Patch submitted against +Duplicate to +Instead of communicating various states the node is in over gossip maybe penalizing hosts with pending reads would work better since it would work universally + +Patch which fix editing problem after visual cases open any jsp html and xhtml file +Reopened due to the revert +HDFS can be compiled now +for good measure key caching is also tured off except in +This wizard would be nice but the new work in gets us closer to an answer for creating a more complete virtual table up front +Im getting bored of repeating this +Now we can only modify the yarnprotos and to add the new state enumAddedbq +Maybe we have to discuss this one more time to come up with the right concept and folder structure for engines +I juts committed this +Attaching new patch with slightly modified log message +Divide by the number of nodes for the limit per tasktracker +This site was added to provide an example how to use the ofbiz content manager for the ofbiz website and to save Erwan some time on his french versionI agree it is not yet ready for prime timeI would like to propose the following the english current website as a language +Regarding your second comment the check for contains is because the classpath is massaged to prefix PWD and to postfix and PWD +setter for it would be clearer if we add one entry in so that we exercise the code path along which the entry is passed +Patch to fix this issue +This gave me a with an empty body +all issues not touched in a month to version of patch which passes tests and precommit +Oof ryan my apologies I was running an older version of this patch +Hi +Im reviewing this patch +One example could involve a loading indicator where prior to the update the indicator is set to the status of loading and after the ajax request the loading indicator is removed to reveal the updated content +With in and rename to in iapierror the build process will build classesorgapachederbyiapierror change is so thread dump information will be in insane build +Thanks! This bug has been marked as a duplicate of +Jacques do you know any free images libraries on Internet or where from are downloaded the images in ofbiz +Isnt there some headers from MQTT you would need to map into Camel Message Headers as well +Maven onlyNeeds a top level Ant build and examples module Ant build extracted from the current core build +I dont think this is an issue anymore as we are just returning a without any exception +was this problem fixed with fixes to the client code in +The code has check and so does +In that case this is working as designed +Woops it should not be protected! Ill fix +Thanks for making this change. +public key used +There are two approaches I can think of +But so does +I will try upload a new patch +dom is attribute of modalPanelDefines the DOM element which stackingcontext will assimilate the values body form parent +Im a bit confused by this issue also +Nils srcmainresources is a legitimate source directory and wtp should be aware of it +Im going to leave this one unassigned for now +could you expand on the description a bit more? all I see in this patch is the moving around of the setting of the password to to the user and then moving around where the is caught +correcting a small typing +the minimal size beyond which the panel doesnt shrink is determined by the label at the ve fixed some issue with vertical list resizing when the panel dont think its worth keeping the issue open +You can store the urls in a list and thus make them all persistent You can introduce a filter to lookup and redirect to the proper url if a bad one was used +I just want to put cents in that this is actually a pretty big deal +Im currently testing the new version but since its not officially released yet Id also vote against closing the ticket now +This will work for us +If anyone wants to pick this up feel free to do so +that cant be the right solutionHowever i dont know how to patch both so everyone got its own actual instance if someone can provide me with some more details how it might be done i might be able to write a patch +Can anybody take a look at above issue and fix? Many Thanks +Im using forward slashes for everything +Wrapping will allow application using for example to use this feature without implementing a new special class for themBest regardsMoustapha Cherri +Right now the client HAS to stay connected +Integration tests should be run on CI separately. +Thanks again for the fixes +As default we dont set package name and use the default behaviour of wsconsume +Sorry I cant figure out how to add an attachment +Thanks Stepan +I think we should keep the cluster and single node setup documentation in Common +It is the defined behavior of blueprint per the specif you need to know if a dependant service is available or not you need to use a and the methods will be called whenever a service come or go. +Will bug you on the mailing list next time +Using the is something I didnt think ofI think that you should use the this enables you to have multiple and hence produce connections to multiple databases +If we choose a name for the thing we are pulling like image then we can use makeimage pullimage etc +this was dealt with in see +That was a mistake +Let me dig into moreReopening it +was mangling the root directory for test data and I needed to change it to properly handle a Windows path that starts with a drive specification +passes with and without P security locally + +Applied thanks Kousuke! +sample page can access them using gettersetters +Yes +This has been fixed and now saves the correct value for the selection. +New dependency to go in srccontribhbaselib +Perhaps theres a problem in the impl on the HP JDK +Thanks again for the patch and for revising it GwenI already applied it in trunk r with some minor changes to the patch applied formatting removed unused methods and imports removed changes to srcmainresourcesorgapachemavenarchivawebactionadminappearance as this looks like its for a different issue + +In addition as Allen points out requiring an fs restart for user additions is unsustainable +Siva please create a corresponding issue in CDI so that we can discuss adding to the spec +I think its going down a very dangerous slope to introduce things that dont make sense to integrate with things that dont make sense +Using the prototype we can avoid specifying every Velocity view in the context but still get the dep +I think we should add a new property something like which will default to false +Finally got a working network connectionI like the refactoring in the builder and the introduction of the geronimo clustering interfaces +I think it is more logical to expect that URI instances get removed from the collection if the remove method is called +But there is a small issue +reopening for further consistent qualification and pushdown naming of get to inject into that schema specific functions have a pushdown check. +Closing +For secondary namenode communication I think it is better to check the user in get and get +looks good +Their framework would be really awesome if it can fix all the bugs. +Hari Any update on this? Should we move this out to +Closed without integration tests. +My patch includes only an update to show the sale discount +Patch that converts to junit +Now complete +Same v patch but fixes heapsize stuff +Attachment has been added with description Revised for the patch! applied to svn. +Revision close for the release +If he wants to close it with wont fix thats fine. +Plain old mvn command line +applied. +The patch has been committed to trunk and +For now weve put the code in validateInit but I consider that a hack +Completed patch for facets rendering was should be executed and updated in order to correspond to the new rendering mechanism +Update Affects to M +I am ready to commit pending one last holdout question +I should have done a little more homework on the distribution before opening the As of Spring which will require Java or above well significantly reduce the number of libraries in the +That doesnt work with this type of aggregation +I was also thinking on the idea of as it solves the problem too +svn revision +Sorry forgot to grant ASF license to doing this nowI grant license to ASF for inclusion in ASF works +Actually this fix has not been oommitted yet +Can you do a review in RB +Ive tried that and Im currently building you a zip that reproduces it +ve just committed this. +By the time the control reaches the soapmonitor handler the soap message is decrypted +Its Property Page +The hod build file needs to be modified +The check for container opened in mode error behaves much better +This is by design +The problem with ASM is that the fact that the promote attribute doesnt actually reference a component is raise as a warning but the assembly builder but then the code that creates a reference to this service in the node doesnt take account of this promoting service not having a valid reference to a promoted service +bq +Hey Juraj could you please provide more detailsI consider this report quite incomplete +The rel points to t and the rel points to t +dims +Anyways I found a way to make it work +Ive attached a new version of the canary following the stack commentsNow the canary is a tool and has a command line with a couple of optionsIn this implementations the canary runs forever and has a pluggable sink interface to collect and output failures and read latenciesAt the moment the only sink implemented is the that allows to use a file or stdout as output device +Updated dependencies for uuid key generator ejb deployer and dynamic classdownload service +Need to include GIT checkin +It is in USERHOME +One static function was being used in another test so copied it into the other test +will fixed though not encouraged. +Lets update the JIRA accordingly +rev put a small sleep for operations that are executed immediately to minimize potential for seeing the record in both tables +Thanks +Marking bugs with invalid fixed for release as fixed in. +I need to think a little more about the best approach +The fact is our ODE package implies really rich algorithms + +Are you seeing a dynamic instance created from the deserialization? If thats the case its probably related to the static type registration +src of uddi configurations has been modified to use geronimogroupId +Next up Surface this in the API +Hey Todd reviewed the patch again +Having groups defined and included in only the ACL file parser was what I was wanting to change +Need debug fixo Creates a request for on each completion need to aggregate theseo Reviews pending for management changeso Reviews pending for queue flow limit changes +Yes the patch matches the need +For example you may have an AJAX handler that is called from several places and each place needs to have different callback behaviour this can be dealt with by virtualising the callback function and providing a different implementation for each place that the handler is called but this feels less clean than your suggestion and would result in lots of duplicated code in various AJAX handlers +Is this correctFrom the above The patch for this problem will follow shortly +Its strange that my Maven installation reported no problems about this +Most of the problems should go away if it gets addressedObjections +Id like to nominate this to +drseuss tcunning svn commit username tcunning m Set the to the empty string if the key is null or the keys ID is null +Here is a ci m Applying patch from +What if you stretch the browser window wider +reopen the issue to restore the old method so that it is backward compatible +The author of the description on the other hand appears to be unaware that needs to be removed if youre going to log to Logj +Then make the following files +I will fix this right away. +a user classname +Fixed a small error forgot to convert the bytes to kilobytes for the memory output +Both PC has almost equal +We decided that we would pursue with HIVE but restore the for backward compatibility +We should try to address as many of these as possible in that release +I synchronized the two methods in question and additionally bundles are now only updated if the new version number is really higher than the installed version numberIll leave the bug open to further watch this problem +I have attached the complete stacktrace you requested +was opened for this test failure seen by other patches +Will commit if tests pass +This is now working +Why all the changes to the classpath stuff in? Looks a little confusing to me +sure Ill do that in a updated for to +moving to + +I dont think thats the case +Looks like the secured jar is also missing so I will create that one too +OK +Where did the min cannot be larger than max checks move to +This code creates a pdf with incorrect positioned elements +Apparently nobody is using security oozie hftp onAs for the patch I think taking and from trunk is almost sufficient +Ill deploy a more recent build today +With clusters running for months we can have multiple datanodes refusing to restart leading to the data loss +This patch contains unit tests for the OM parserSpecifically the following files are introduced with it patches applied +I see them in my copy of the standard draft too +Needs unit testing and integration with context for configuration +Argh! Haste makes waste +Attached patch with missing test filesPatch applied +For example GET HTTPrnrn maybe required to be escaped in serialization format if is not honored and HTTP protocol is chosen to wrap around the serialization protocol +You should be able to grant permission just to and the testing policy file already covers reading these properties +Illustrate for types branch with testcases +without the quotes and placing the file in the directory in thejar proves that facelet tag libraries in make a special case +verified by Svetlana +Need to test +Applied the patch I also changed the base file for the merge info +This appears to be a hibernate issue and howwhat its auto detecting not something that can control +I just shared a dropbox folder with you +See XSD +Committed to as revision a minor typo in the msgEnsure that the dataformat is valid and the associated camel component is present on the classpathEnsure that the dataformat is valid and the associated Camel component is present on the classpath +Please ignore my last deleted commentThe NPE is caused by the is not updated while unloading the applications correctly +Bug also was reproduced with cluster +Thanks a lot +back in a minute +eclipsekeyring arguments os macosx ws cocoa arch x product keyring Usersburr +With the fix for these specific exceptions the brokerbridge are not stopped and fire is never called +Not to mention simplified configuration +You are right that Weld does check the dependency graph for cycles but probably misses this one as a combination of managed beans and producers is used +i think we can close this one and remove the filepagestore all together then because i dont think it brings anything extra anymore +Were using a product that allows us to manage the artifacts in our repository and edit their dependencies +closing +Committed +Or you can also contribute if you are planningThanks +Ive been pulling my hair out all week trying to get our app ready for IOS and your fixes have really helped +And hopefully it will work. +This issue is not going to back port to +Guillaume the new resolver is basically complete and committed +Since execute is not getting called recursively same thing should happen for initialize +Picture of eclipse not including resources +Bulk change of Fix Version for all issues fixed or closed for +This seems be users problem as it is his responsibility to ensure proper shutdown procedure +I can split the patch into generic dist system tables and specific configuration backup rebase and use the common part in wdyt +Opened +Seems like a good fix for +Seems they renamed ehcache to +The build process starts with the all config and modifies it to get default so the standard version of the file should be the all. +With this I can easily tweak a container implementation for my own purposes something we definitely want to be able to accommodate +One more thought +Jon +I think that should be part of the javadocsHeres a proposal if we think our users find these names confusing why dont we do a poll on and get people to vote on whether they find these names confusing? Because it looks to me like its a flavor thing some of us prefer to read javadocs while others prefer methods names that tell us what they do without reading javadocsI dont find that open saves me from reading the javadocs because I still need to read them in order to know what will be returned if the reader hasnt changed +Memstore flusher wont write old cell values to as they are not referenced from the Memstore + +Looks good +Closed after release. +Hmmm +Hope have a standard +Committed to trunk +I added support for some tags in version of a local version but needs to be improved +Are you still working at this problem? Other question How long it will take until it works? Im waiting for this +I just copied the description over from and forgot about this +Hahaha +Check the implementations +How about we just expose a select list or something and let the user explicitly chooseCheersChris +This includes a test +Issue was already logged by other user +Included in. +Scoop is an entirely different matterKevin +Ready for the Cos +I will try to inspect whats causing such a dramatic slowdown though it isnt normal +This patch only applies to so failure is expected +Raising priority since when this occurs it hangs the job +the test case fix version for bpm to expected inclusion +Attached a new dump you think the failover transport is more stable we can try that out +I can not see any wrong in the generated code whats your expectation hereIf you want some wrapper child as a return type then you have to name it as return if you want to the string as a return type you can name it return instead of the string in your schema +totalSize numRows lastmodifiedby etc +Can you verify if this still fails on windowsAside from this it seems we need a dedicated Windows machine that runs our tests though +Could you fix the eclipse warnings +Closing dup. +Sry guys. +is not created anymore forwicketpom module with r. +And good riddance to them! Now on to my next vista +Strange indeed +Attaching the thread dump +I am not sure how speculative can help in this case since these map tasks have to run times longer than other maps in any case +The pdf isnt even adobe reader is complaining about it +They wont be merged into Woody codewise +What version of Cordova are you using so we can try to repro? Update the issue field Affects Versions +Attaching the work done so far +Thanks for your comments +Because some users prefer to support S with Spring it is impossible for them to use postback forms with the custom Spring subclass of +r. +I will remain this issue opened until i will submit something to show +Probably I have to focus on the layout of the cms but I think that it could be an expensive task +I applied patch to current branch but it does not compile anymore +Would you provide a patch +I think the old version will return null in the case of an path that found a file but it didnt pass the specified filter +BTW Jacquesi dont need any data to be load +Could you please give more details on entry log corruption the possible cases +Submiting the patch for the same +Dimitris the docs say true the suggest false +Hey MistyThis change is pretty much under the covers +Would that be sufficient for your needsIm reluctant to add another template method such as onDispatch since we already have so many template methods in those controllers +Backported in r r r +Should we and upload it into sourceforge or should we do it on only +try this patch to improve the channel test cases +Flavio that would be cool though I think code review would be enough +Im not able to ever show a stack trace to a user for that reason +I see only USD in the test but it is not part of this issue +Any suggestion to get it from +file necessary for the junit biggest difference is probably that the is also stored in the Ant project and that the cachepath task uses this report to build a path instead of parsing the XML report produced by the report task +addresses this issue +Hey Dhruba +Can you provide steps to reproduce +take two +Whats in v that is not in v? Does it address comments +So for the client we need to add the configuration right? So will this may need a client restart right +Good point Sebastien thanks +Ok +Thanks Evgeny Manually IT with a +If the key contains the specific mark the partitioner will include the key as well +And if you used any custom builder properties andor a custom binding file please attach those as well +Thanks Brunoyou have done a very great jobApplications labels committed in rev +The tests have compiled successfully on my computer +Items to be fixed in multiple releases should have one issue per release version or we cannot accurately filter for the release notesPlease edit the fix version field to only one version and clone this issue for any other versions that this issue should be fixed for +Attached is a patch for version +Please can you update the status of this Jira to indicate when it will be ready as we intend to release M on MondayThanks does not apply to trunk as it uses maven not ant. +Editing the XML no longer allows you to manually recalculate checksums for the edited opcode + +svn blame shows that this test was added in to check that a different schema may be used +at revision while running mvn site +Committed for in rev +You dont need to provide a patch for releases +Firefox is the server I use all the time and the one the unit test execute againstJust did some manual testing and it is working on Mac Firefox +Not sure what would cause the exception and request for more information about it is unanswered +Attaching query normal response and errorThis response causes exceptions +This is actually not a bug but rather a limitation that originates in Hibernate itself The standard Hibernate implementations only really work within active transactions +applied the patch with the modification discussed. +However youd figure they tested the hell out of it including instance count verification with the garbage collectorIf we assume the code for is correct then the only idea I have is that the value a has a reference chain back to the in someway which would prevent the key in the inner map from being collected +Moving to longer applicable +Changed and to private as indicated in Pradeeps comments +if you move the node within the workspace or clone it to another workspace the uuid of that node stays the same and so does the respective version historyif you copy a node it gets a new uuid and therefore a new version history +Thank youI will try to do this +Hang on I dont understand this +Removing the assert works for me +BTW I think Dani Rayans nd argument for the rowkey is a good suggestionAlso I think there should be an md decorator to get the md of the event body +is passe Tools is the way forward we will settle just by having a release end of summer +Patch for +This needs to be added to the release notes page in svn if it hasnt been done already +Done updated +If it comes up later a new ticket may be opened +Thanks! +If you call an executable from within Hadoop code and it isnt distributed with Hadoop this should fail and that should give appropriate warning +There is some discussion around this issue within CPEG and it is not clear what the answer is +And is there any way to get Jackson version at runtime to be shure that version is +This can be done once by moving it to the constructor of the producer pool +the output for running +rad system properties for this and they are already manageable from DMR +Sample output of the producer consumers latest version of my test cases +Do you guys mind opening a new ticket with the refactoring changes? It gets hard to track whats going on when an issue is committed and closed and a new patch comes in +Yup will give it a whirl here thanks +please try my patch +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +Clients can provide preference for storage type to write to +Testing in real cluster is appreciated +Thanks a lot Ivan for the detailed infoIve attached the patch addressing above comments +Sounds grand +Closing all resolved issues for. +I will make the change you recommend in a patch + to trunk. +Really most apps do not need all positions stored ie they only need to see typically the current tokenI think this is why it got slower with my patch in practice it didnt matter that this thing did backwards linear lookup due to this reason +If PLUGINSDIR is not given an exception with a message is raised +Patch Applied thanksRemember to continue adding links asif more samples become available in the plugin. +Closing resolvedwont fix issues. +The statement in getValue javadoc about encoding may not be true in all cases +Attached a screen shot of the Test Connection UI in Admin is fixed in EAP. +Otherwise my checkin will break common usecases + Sent email to Lars awaiting his thoughts +there are two reasons for the problems Currently in some exported packages from jar files in lib directory cant be seen outside the plugin such as in +As per conversation with jdcryans +Please have a look and let me know your feedback +attaching entire eclipse workspace used to reproduce this Brad I was able to and fix the issue +Im happy that we concluded to the same solution +But totalHits returns all the docs +The metadata is loaded but not respectedFWIIW redistributable isnt respected at all +lets start w sstable writes direct and ignore commitlog for this ticket +And if rogue untrusted code is getting into your app then I think youre screwed +jj parser working with macros +Im not expert in Collection use optimization. + +feel free to reopen or create a new one for specific problems. +You should see the session be reused before the timeout is called then a new session being initialised if I am correctRegards as I remember something about GET not being able to store session information +Comments from Triage on April Brian M +For standalone jar files this uses the patternservicemodulejarnameFor EJB deployment in ear it uses this patternserviceearearnamemodulejarname After the unique module names as per spec EE come into picture i guess we can use the module name to generate these jmx names +You create a snapshot of sd and sd under the same name +antlib the patch compiles +Since we wanted it to be fixed without waiting for the next major release +No worries once your highness is satisfied please close this one +But needed some more changes to make it fully doneCan you confirm if the two are duplictes if so one we can invalidate +Please see he lines with and download +In looking over the intent of range David implicitlyment for it to be an Integer rangeSo if this doesnt break the functionality freeze I propose adding in range depreciate? rangeInt rangeLong rangeShort rangeDouble rangeDate adding in validateRange depreciate this method? validate validate validate validate +I pulled your topic branch and ran the automated globalization tests fine on iPhone iOS and iPad iOS +The test failure is not related will take a look the hbck one +Code in XML file +So we have to upgrade to +how transaction could change something here +Heres a better patch for It unpacks the needed plugins and reassembles I was afraid that Ill run into that long filename problem but I didnt after allIll check into unpacking the various jars inside the main jar. +Installing the Repository now lists both and ZIP files +v adds carrying Action indices index is postion in original submission and then over on server setting the passed index into the response Action +Oops the get issue is really related to and not this issue +I did not have resources to do a node run +We wont really need to access part files and there access times to get filestatus of a file in hadoop archives +Removed hardcoded reference to files with calls to get +Setting to false eliminates the error +Im getting compilation errors after update the RT branch eg is missing +Also in your example the ordinal will not help in any way because those are two different handlers +I am using Tomcat also checked jetspeedlogs cant see any clue +Numbers should be handledslightly differently +I guess about the reasons +Have a version SOCIAL SHOWCASE working within Roo project and am educating self on how plugins work +It could be this is rarely hit in practice on more normal input +I think anyone who wants to revert it is incumbent to demonstrate the patch is actually in error +Yes it appears every issue gets assigned to Matt because hes designated the Project Lead in JIRAIll mark it as closed for you +The previous patch broke some harness test cases +The maximum on ms sql server is this to a note to test the email Any idea as to when the fix would be incorporated into hibernate and made available for the users +The issue could be reproduced see. +Possible solution can be found in the related issue added some logs in redback and Im seeing a lot of no permission for the user admin which should not be the caseI think this might be the root cause of the missing buttons. +Modified to list the ENC for all deployed web applications +when there are heads local filesystem doesnt seem to help +lets keep the noise low a single patch per project should be good enoughBTW thanks for doing this thankless task + +I have just done virsh dumpxml and i dont have any ISO attached and yes the is a fresh install +Include the lib in the pom as jboss as uses the logj by underline +Add an function for check of Oracle? +I apologize that I forgot to mention that +This is a configuration issue +This issue blocks POJO in POJO response I hope this will be solved in is the aar and source files for the shipment status web fixed this few months ago but it has shown up again +Axis Codegen wanted all the importing XSDs in the projects home folderFor example when i had in homewsdl folder +This is a small program which reproduces the problem +Forgot to mention that this query works just fine in Hibernate Im verifying that nowIm starting to think this may be a regression in my app and not related to hibernate +This isnt very important to fix and is not an issue on trunk +Wildcards were added after +NETs implementation is allowed +Security probably wont be in the first version but as a veteran of the Hadoop Security Wars myself Ill be sure it can be supported +sorry should add p as reference into p and add the p into the created server +This is an long standing issue that is affecting production installation +As a matter of fact I have decided to go with after internal voting +Please send me an email if youre interested in testing the new projects files +Ive been very tied with commitments recently and I havent been able to get a solid start to this sadlyIf anyone wants to complete this task then I am okay with thisApologiesAlan +Should be within hours +Separating these concerns would be an improvement +The previous versions were incorrect the pause was not long issue seems to occur also with the LRU Algo but not with the Fifo job Galder! Your patch works pretty well on my usecase +Looks good +mmh wait +Another findbug warning +Fixes the vtest +Thats a fair point +That is the reason i renamed it because the file need to be recreated anywayuomId can mean anything and adding the currency to it makes it more clear +To get a little more context here Is this problem repeatable? Do you see any behaviors not working for you? If you go to Preferences Groovy DSLD and click Recompile Scripts does the problem happen againAlso if this is reproducible any chance I can get a hold of the maven project that causes this +Please t have cancel button still sending requests +Please close this issue +to the idea of incrementing a counterAnd if you need to work with huge numbers where precision matters the patch +moving to the trash can is essentially a deletion operationNo its a rename +So farxbean rev openejb rev geronimo trunk rev branch rev +Also it makes for Groovy not much sense to compile against an alternate JDK +Did not add a geronimo specific message as there is no message handling api in geronimo thru property files +I guess I can change the path in the test to use libraries from downloaded distribution but would still be useful to know how to build the whole lot +There are a number of screens where you can uninstall things I can think of web JMS and ultimately EJB connectors off the top of my head +Accidentally selected wrong license option +to Jyotsna Rathore and Pranay Pandey for your valuable patchesI had applied the recent patch from Pranay Pandey that resolve all the issues from the previous patchYour patch is in rev Ashish Vijaywargiya +This is very good +Node removed from include file was not being shutdown +re try to account for the amount of data being sentreceived by the region server to the clientThats pretty much what Im looking for actually +mrepositoryorgjrubyM Userseclark +I confused and +Replaced the old patch with a new one. +thirdpart version switched over on the SPCP in the test suite +Since its a simple refactor I regenerated the patch +Separate meta log is there in later version of and is disabled by default +HTTP Date Tue Sep textjavascript charset Wed Sep Tue Sep Servlet JSPHTTP Not Tue Sep Servlet JSP responses to compareTomcat HTTP Tue Sep Wed Sep text chunkedDate Tue Sep HTTP Tue Sep Wed Sep text Date Tue Sep GMTWLHTTP Date Tue Sep textjavascript charset Wed Sep Tue Sep Servlet JSPHTTP Date Tue Sep textjavascript charset Wed Sep Tue Sep Servlet JSP responses after refreshTomcat HTTP Not Tue Sep HTTP Not Tue Sep GMTWLHTTP Date Tue Sep textjavascript charset Wed Sep Tue Sep Servlet JSPHTTP Not Tue Sep Servlet JSP think we should check caching under main supported servers and main browsers +Sorry about the lack of null check +Bruno retains the original host name it was resolved with +The problem with adding a dependency from testreport to test is that if a test fails it wont generate the report as mentioned above +This issue is now only seen in believe this is fixed in latest patch. +Should only ignore FAILEDKILLED target states only for the AM ContainerAlso please look at the javac warning +I missed the earlier comment abt the test case being part of +There is only one nasty piece you currently have to add a dependencyManagement section which contains all the tomcat artifacts to update the version +Fixed in revision closing stale resolved issues +This issue is continuing under. +Should be fixed soon. +Rainer You are right overseen what package that was +still makes it necessary to skip them for the client driver caseRunning regressions +What about JCR and +Thanks Koji +Thanks a lot +For the moment Ill mark this as assigned to me and in progress so you can recognize it + This bug has been marked as a duplicate of +For EE containers always runI agree the message is a bit confusing +If you use a relative URL then the full redirect location is built by the servlet container so I dont think this is a Spring Security dont raise issues against releases +The fix should have made it to the nightly +Now analyze command succeeds since we lowered the number of unique counters required +There are some failures in the jdbc suite caused by these changes +What can we do about it? Shall we recommend to use another milestone of Eclipse IDE or milestones are of no use +bq +Yes! Ive got the error now +nice +BTW I forgot to remove from we dont need that anymore +We may stop short of this due to time constrains but it should be investigated +This is related to the issue that documents where in the code you trap additionsremovals to the inventory +The initial fix caused several tests to fail +Assume closed as resolved and released +Done. +I would like to still have the option to disable it +this patch makrs generated files for deletion but svn does not actually delete it + close this +Gert was so kind to add it to the servicemix issues +And I also tested it in real cluster +Implemented SchemaisSuperset Schemaunify and a Tool for generating unified Schemas from avro data filesI have one open question about how unify should work in the case where two Records have the same nametype but different default values +This is causing bugs with apps that already wrote producers for Http artifacts +Hi Jason a couple of other suggestions the second timeout seems to be too short for push replication +But when I then try to run I get the following error jdench java cp exporthomechaasejavadbmorecodetrunktrunkjarsinsane +Sanjay this is being done in a branch +Screenshots are retaken in guide Exadel Migration Guide guide RHDS GSG and Seam guide except figures in seam guide should be done through creation a database +If the reload operation is supposed to work no matter what the server configuration is than there must be an issue in the operation +New patch propagates pig log level to backend +You might want to take a look at the folder in the folder and see if you can create an integration test that reproduces this issue having a platform independent test that we can incorporate into the test suite might help to find the issue and would ensure that the problem doesnt easily get reintroduced at a later time +Note sure its worth going there +By server side I mean for the jobtrackernamenode +Corrected JDBC deprecation link to in both EWP and EAPRemoved Messaging from EWP release notesCorrected double comma in EWP release notesAssigning to you for verification on or live Rajesh +The initial Joey infrastructure is in place +Im changing the ticket name so it more accurately reflects the changesRobert +Ill work on getting the information you requestedMike +Sorry for theconfusion but thanks anyway for the sample fo file which we will be able touse as a test once we fix the problemVincent This bug has been marked as a duplicate of +requires you to bind to a real IP address not for clustering to work properly +Code has been checked in for to move into our new testing state +Code review and suggestionscorrections are welcome +Any ideas on how might do this? I took a looksee and client and port stuff is left behind in rpcserver and not available up in the server when balancer goes to run +the same patch without trailing workspaces +start and immediatly stop and compare the logs +Additional threads showed no significant improvementThe problem is threads colliding on logjs synched method +But like I said making that condition an error seems over the top +Interesting idea! Could you make an example of what you would injectIts not a bug BTW or youd be required to add a failing test +We can create another Jira to port this to +Any help would be appreciated this is a major annoyance when monitoring over any sort of is setting the socket to linger an option here +Released in greatThe main problem was the dependency of +This patch should add in the missing functionality +So it should work even if string is set to null +Please reopen this jira if you want me to commit it to the branch +See +Do you mean? Or maybe some other flavourdistro of Eclipse like +Do we need to provide buddy replcation node in the JSESSIONID? +Yes of course +They do not show up since the first few will make them drop from the because of the reuseI am not sure where and how this errs and if it does at all +The fields all seem to be arranged such that the upper baseline of component B is pixels below the lower baseline of component A +This is another race in the JCA codebase leading to the same deadlock as seen in try and ship an errata for this fixing the GA releaseMoving to continuous +This allows the to display machine names as hostNames rather than IP address +It is an addition to +Will this be backported to ASEAP ? +This patch fixes some unchecked compiler warnings in the compatibility tests +For giggles I tested it out against embedded Open Web Beans using Arquillian and it worked as intended presumably because OWBs interceptor code is not similarly affected +Can you confirm that the current handling meets your requirements? +Well that is exactly why we have joint my understanding is that I do use joint compilation see nbproject the build script was used in a bigger project with mixed code and worked ok +Ive just committed this +I prefer your version of the docs +This is the same bug as +DUPE +Its functionality has been included into the has three jar files +Punting issues from RC to final +Thanks AnneFixed trunk rev rev . +It still seems odd that this could generate thousands of producersJMX entries but not out of the question +May need a separate JIRA for thisVinay please provide the exception seen here +This might be a good area to allow for some sort of plugin interface to be able to add mutators like this +I agree with Max this looks like very useful for debuggingI think the code is trivial I could provide a patch if you tell me on which interface implementation it should be exportedI suppose we cant just add something to right. +If we can implement a Virtual Attribute layer Collective Attributes service can be implemented as a part of that layer and the issues automatically will disappear +So just to repeat the following sequence of steps will result in an experience which is consistent with the symptoms you are seeingsvn updateant clobberIf your environments are still broken the following may help Throw away your subversion client and it from scratch +remove empty values before joining +Like I said I will have a closer look later +Adding unit tests to latest patch +It would be great if we could have some kind of branding around the Rocky Mountains +We should really visit this before releasing since a patch is already m on call over the next hours to get this thing nailed down! Agree that should not go out without this. +Hadoop was released today. + +Brett recommended latest version though +Note it is also mapped to the ESC key when editing the tutorials cover pressing ESC to cancel your current edit. +Now I guess my version doesnt support a Map as model since the method signatures specify but thats an easy change Pauls version is probably significantly different from this anyway +Thanks for your support +I think this bug is no longer applicable that setting no longer appears in in Hive +I cannot see how this is related to Spring component scanning +Are we mostly afraid of GC pauses or disproportionate action caused by small network hiccups +My initial code incorrectly read the script as a normal file +I was testing it last two of questions on forums +Also try out the new automatic index usage support if you get a chance +Lets reduce the severity level for now after I have cleaned up other blockers we can revisit this issueThe memory issue has nothing to do with and I also profiled the time spent on it does not contribute too much either +ThanksInstructions indicate I should be able to svn move the repository myself but this isnt the case I tried +So in our opinion given a total number of partitions is more practical +plan on committing today unless there are any reviewers who dont think it should go in +Confirmed fixed +I think should address this +looks familiar +as i think this is very user friendly behaviou +ok committed. +Committed original patch from all resolved issues for. +Committed +It might be nice to have a stat for when this happens too. +The broken links are fixed in the patch for +Nick released today +The more I think the earlier suggestion on sending events for de registering and then cleanup is probably much cleaner +Hi AlessioI have merged a potential fix for this issue +Jira must also be updated to send mails to dev instead of The project URL in jira is also still pointing to the old location +I had a fix that did chain the exceptions but it broke on jdk so it was backed out and put into a Jira entry which I think has yet to be fixed +This works on static dynamic and run trees. +do you also have the heap dump files itself +Heard that multiplexing channel selector will solve the problem +I feel like we shouldnt be punting in deletememberbracketcompleter since we should be able to determine what our possibilities are there +Changes lgtm + +More information on this issue can be found in +All doneThe Java issue is resolved + is now done as well all tickets +Will have a look at the code thisIf you have a snippet of code that could be use to reproduce the problem this could save me some time +Just an update Ive got dependency issues on the machine I was using for dev +is known to be flaky and passes for me locally +Yes Patrick I am OK for this patch to get committed +Ill take a look at this test to decide if this is test problem +hi guysany update +We need to add functionality into so that it can identify the different parts and extract the SOAP body +More specifically Id like to test the effect of merging it to +If you have some time to test the fix before its integrated in master please grab the branch in the pull req. +To be clarified +In order to have a gig java heap we will need to run on a bit Hudson slave machine that probably has a gig OS memory uncompleted issues are now moved to the new To be determined for future releases category +Wow! Its great! +Heres a patch with sophisticated backwardsId like to commit this and open a followup issue for the URLEmail one that one is more complicated and needs to first be ported to Standards interface +Screenshot showing a trace from a to a GC root in anundeployed Tomcat app +More specifically I used CMS garbage collection which based on previous experience in support works very well avoiding false suspicions +So does this query happen on every hit or does it happen once and then it does not happen again +Not necessarily to not cachethe tree but by making there be more levels in the tree then it will costmore to research the tree from the top +Added wiki all resolved tickets from or older +Thank you Juergen! Jim Kriz + head CVS patch attached to fix eclipse errors on openejb core project +We may be able to test this in a few weeks when updating our you. +Problem is that the message is marked as taken so only one consumer can receive it +Thanks very much Dyre +we need to fix first tooYeah I havent merged with trunk since we rolled back so we should fix it first catch up with trunk and then make deletes work +Marking as regressionThis is only true for the client driver where Blob performance has dropped significantly afterThe reason is that the existing code with the performance issues has been taken into use by the implementation of locators +see subversion commits tab +to the idea +Logged In YES useridA better solution has been added for +Jarek Can you describe how to reproduce this issue step by step? thanks +If you need it in the branch please comment on this to the branch in revision . +While I agree that it should be easier to work with a set of modules allowing Maven to just ignore missing modules is a not a good ideaHow do you check out a sparse tree +Marking for +Yeah I bet thats due to the datastore not catching up the namespace parameter passed by +Patch has been reverted due to problem described at working at I needed to change setting of parameters on datasource fromusing Unmarshaller to java reflections +You are right Sid there still is a race there +Attaching a patch that adds synchronizations in +Logged In YES useridFixed in both and head now. +New patch fixes the integration tests +Still investigating +trivial sneaky part here that fixes the bug is that the dependency is ordered after so we are guaranteed to have all the jars +Patch v compiles +Considering were already past our timebox for we cant afford to delay +eclipseconfiguration +by the way +We just have to add in the dropdown list of the violations viewer the rules created for manual violations +Changes are as follows Testcase changes to run a job using and then test Rumen parsing against that as compared to having a jobhistory file +Every bit of information we can give them helps Id say +Marking this error as error in the header so that it is not lost +Closing all resolvedfixed issues of already released versions of Roller. +Thanks David! +wonky updating +Marking closed. +See my I use the have a look at bug +How can we make this optional? With some UI flag +Thank you for your quick reply +Full compareTo is OTT and looksIve spent some time looking at this issue and I cant figure how it came about +This is a feature so changing version to be consistent +Beforeafter switching Sun Java nodes toThe nodes already at the bottom were accidentally running already +Navigation works to their declarations +Of course as I mentioned allready Im not very deep in cocoon kernel and my patch is just a proof of my thesis and therefore no proper solution +What kind of table should be used to generate that kind of code? with compound primary keyShould we fix template in branch +I added. +the internal map would do and future proofing it if the map does not define a value we use the given name +I have created a task dedicated to this step given the complexity of its design changes will be discussed in the forum if we decide to integrate the graph to trunk +The description on the web site appears to be copied out of the header comment on +The identity capability should be fairly easy to add +ohmy jdk verson isI will try change command line encoding to englishhope it can workbut how to build with chinese +Hi Sagara Thank you very much for your feedback on my work It would help me lot to implement a good json support phase to Axis +it is build issue so when I build the seam it will be solved +Tests that are too low level can also be a maintenance nightmare too because they assert current implementation details you dont really care about +The event was triggered because submittedValue was set to empty stringWith the latest SVN head it works now so one of the changes these days must have fixed the issue. +Dear allEnvironmentTomcat should be Tomcat Could someboday help me to update the Environment infoThanks a lotMichael Xiong +This issue is resolved with the changes forSee comments for more details. +Using ant may also affect the compilation and deployment of samples +I synced up trunk and checked it in my two MAC +If possible can you please submit SVN patches instead of the entire files? That will make it much easier for me to check and commit the source +Is that the wrong stylesheet? Which stylesheet generates the site +So far so good Ive not quite gotten to testing if it fixes the busted rings problem Ive been seeing but at the very least I dont have constant messages about broken MTU size +New cut of the patch Kelvin uploaded earlier +from me on committing patch v +im computer science student +However my broker still sees the original connections as ESTABLISHED +Actually we supported such functionality before by extending compute +Both the buckets and the projection need compression +The reservation status should reflect accordingly +nudge Apache the completion of whitelisting to release +Closing + fix has been checked in +The output will be directed to and you can use a tail command to quickly see the results tail f grep TRACE UsingJust keep running it a few times in a row +m in my user directory that also points to the local repository +Otherwise we create a new directory +has someone else beside Eric seen the plexus plugin + of a bulk update all resolutions changed to done please review history to original resolution type +Anyway these mocks are not realy used in tests so you can use null argument instead if you want +It is fine to make heuristics better and better I have nothing against that +Still have to check the patch but here is a first agree currently this is just messyheres an updatedchecked patch +Could you please further explain when this would be useful? I could understand that it would be useful for method the take generic lists as a parameter but I do not understand where it would be useful for us to return a wildard list? Probably my bad understanding of generics +Patch applied with thanks +my centsOr we could have a more generic wayThe problem is that the mysql Job table does not contains columns for more than one stateWhat we care about in this table is the final Job state +Write javadocSpecifically I think it would be very useful to document the guarantees with respect to which notification events occur when +We really should try to get this fixed for +That sounds good as at least that would be a clear issue +Merged from trunk into codeline with revision +Logged In YES useridShow how the principal is being modified +Werner I am thinking at least for now for maximum compatibility +Not sure how setting the tcp linger time to is helping you? What OS are you using? I find it surprising that setting the linger time to be doesnt delay packets to other sockets +lgtm +Is there a that gets built in Trunk or do I need to build it myself +moving to would also require to update everyones documentation and examples to work + on patch +They allow for multiple event streams to populate the working memory +Care to update this for Clover Michael? Then I can hook it into the site and the nightly buildsThanks Grant I will update the patch soon to use clover instead of EMMA +It doesnt represent any production query +So this will not be a problem anymore in release +bq +I cant edit it +Now they suspect the mother board is believes the ECC error is not impacting our db performance +If you find further issues please create a new jira and link it to this one. +David can you be more specific by in the? I am assuming you are referring to method permissions of EJB? AFAIK does not have method permissions +breaking session constraint on threading? you have the listener for receiving and still using the same session on producingThat could lead to other issues +Its sufficiently small to just live alongside your validator instead too +fixed in rev! +Lets say for the ETA +The reproduction script is quite complicatedbecause it has to fetch a large number of intermediate resultsenough to force the internal buffers to fill and the partial resultsto be written out to temp filesAny suggestions for how to write a simple standalone test casewhich exercises writeExternal +Thanks bud +This should be pretty much issue +so the only fix would be to remove per instance +That doesnt work currently and is related to +Hibernate seems to be smart and in many cases it will quietly add a column if needed to manage a collection +I was wondering if we could get this one committed soon +Im going to shelve this +bq +You might want to set the property to false and try again +Patch to update VME version number on Harmony website +davisp advises we should be using a refcounter on open view files and switching when these expire +Attaching the the schema file +here are two more candidates +I like to plan for success +do not use datasource for connecting to database but use proper setting of connection and then the workaround can be applied to showcase configuration instead of container one +Thats not a compelling reason though +VerifiedAll the Issues mentioned above are resolved after following the steps document for zone migration +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Just so its clear Im willing to work with you on this butI dont really understand the problem can you propose solutions that meet Cocoons needsObviously Batik needs a security controller in the normalcase +groovy as default but you can add more file extensions via WindowsPreferences GeneralContent Types +Hello ThomasSince Spring and Spring Data are the separate projects should I also create another issue for Spring Framework project +I only have one comment on the Reference ManualOne of the lines in the Reference Manual change is as followsstart existing lineIf you include the collation attribute when you create the Derby database the collation is based on the language and country codes that you specify with the territory attributeend existing lineThis line should be modified to say follwoingstart new line suggestionIf you include the collation attribute with value TERRITORYBASED when you create the Derby database the collation is based on the language and country codes that you specify with the territory attributeend new line suggestion +This is one more reason for using it gives you a fair amount of freedom with specifying formats in configuration +Attaching the patch to rename the file +Except then we have to have some way to say do or do +Java testcase to reproduce schema to reproduce in CVS +I think I forgot an anyway +Is the server telling the client that it is being disconnected because of FailoverThats doesnt seen logicalIn production a client will need to failover when a Bad Thing happens on the server the server most likely wont get a chance to send a Failover message to all its clientsFailover is something that the client can try if it knows how to and is configured to but otherwise it may notAlso the Qpid client could be run against say the server what happens thenJohn +Its up to trunk looks like its an issue with id Im so sick of running into that +interface is fine for the time being +Is it with the guest userDo you have a entry in your? Im not sure Archiva was modified to support a virtual guest user +Fix patch apply bogus failure I the patch test +when all replicas are corrupt +I also didnt see any error messages that indicated that there was an error +The bug has already been fixed and the fix will be included in SWF that there is an easy just use classic Spring beans definitions instead of the Spring XSD. +Tweaked the issue title spelling fail Also probably another issue that stems from my commits +Because usually the only gets the checkout URL +If we chose a zip file for example the browser wouldnt be able to handle it so it would prompt you for downloadAs for why would someone want to stream back an image I agree it isnt a very good example +Shauria de corregir la entrada dempeus i afegir tamb up +Im inclined to close this issue unless others have concerns about this implementations correctness completeness etc +Some comments on phabricator + +Thanks again for all your help Knut +The javadocs messages were not from this patch +If you rename one version in the History page this should solve the issue +Resolved. +Allows specification of file character set +Im in agreementUnfortunately from time reasons I cant prepare a test case currentlyBest regards +We cannot prevent the readFields method changing the content of the object +The only concern is that now need to know that requestKey is actually +Spring Summer cleanup for old bugs +fine with Sun JDK v +I feel hive server and other such code would naturally have less concurrency problems if we cleaned up thread local through the code +Updated patch with new signature +Getting rid of allthose trace statements may be a good idea though +This issue has been fixed in the CVS. +Thanks a lot Ivan for the review! +Is the gist of this issue to ensure that all other inactives are explained +No I am saying you incorrectly subscribed the address to the list rather than to the allow list +A is and favorite disabmiguated +I see my work on this issue now as being done with having things working decently if someone wants to patch the fajax tag go ahead +Eli filed +Please let me know if you have made progress and would like to keep working +Maybe manual build is just always Full builds +However if using gem is the most common way to install a ruby library and nobody needs a debian package then we should remove ruby completely from the debian package build +Thanks to Dimuthus patch this is fixed now +The skinny jar is just a part of it and the signature an issue in a particular caseWhats the issue addressing it with mvnorgfoo +OK I merged all commits to x +I do understand that it is an issue +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I have created a patch that checks if the current entity its superclasses belong in the exportedID entities +Should we add a note in the readme or a config step in the QS +will upload a new one shortly +Thanks for the link +Thanks! +Looks like this is missed in the release. +This can be done by mapping the CDI Conversation Filter in as the aforementioned link shows +The issue with update project configuration is a duplicate of and it does not matter if project is module or not because right now update configuration is not using workspace resolution +This is NOT critical +Glad to see that creating a potentially big list was not necessary at allBy the way just for seek of tidiness I attached another patch that removes redundant code in without changing its behaviorKind regardsAnahita +Can you please let us know how we can enabled logging in Axis so that we can capture that requestThis is happening on our production server which is accessed by many clients +Thanks for the quick reply +Attaching patch that makes deprecated and makes ans usePlease review and provide your comments +Any opinion +Is it not a bug if I change and restart but nothing changed? Also the ini had proper bindaddress when i reinstalled it +Ive been using this type of project layout as a workaround for a few weeks but Ive found it actually works better than my original configuration +Closing as complete thanks Joe. +Pretty much same as expect reading from command line instead of script file. +Yeah while some of its settings could possibly be per package most are definitely global and would take considerable work to change them +Do you think we could describe the expected behaviour somehow +Thanks +In all cases sending from Server to Client works fine but client to server fails +corrected at in files +Seam and Seam the component as FP will be now a part of regular to distinguish correct seam +I have merged this to. +Files added with add file then can be access with +I dont see why we should replicate all of the Java decision espeically when even the Java camp would remove it +The callbackhandler strategy is a MUCH better alternative than my original suggestion +I believe I resolved this issue a few days ago with revision +Thanks for review Ted and Stack. +And He said it is trivial so anybody who has interest in this functionality could write the code easily +So youre good with the patch JonLets add a warning to logs and then Im fine with the approach +Thanks Dhruba + unscheduled issues to the next release +Lets discuss these issues on the dev list and come up with a good solution +Well leave it +I pulled the changes and after a few test runs no failures +It must be applied after the patch +jar on the Windows machine +Why does the number of collation elements for an escape character need to be oneThe escape character is there to escape special characters it has nothing to do with collation +LGTM +and Delete selection buttons labelled Clear Selection? +This issue replaces the and changes from to with revision +No failures +If there are better alternatives suggestions for those are certainly welcomeFor future proofing we should also look at including host addresses in the mixA patch attachedNot sure how fare in Java but we could try that here if perf isnt a big issue +Does typechecking of this code complete successfully? If not this is a typechecker issue +This patch was created using sources from nightly build +Sun jdk +Optimization patches are never put into the frozen branches only into trunk +Sorry stringified the sorted array wrong +See +on build +After patch integration all Classlib tests related to this bean passed on Win and Linux x +Currently the classes prepared by the plugin cannot be used anyway but this is not a problem but a problem not supporting this mappingWhat we might do in the plugin is provide a functionality to automatically insert the path mapping for the JSP classes +I fiddled with it for another minutes and still couldnt figure out how were supposed to guess ahead of time to set nowrap +And then we can open a new jira for to make hbck even gooder +We should make sure that the header checks are case insensitive +Ill keep an eye on this issue and test any patches if needed +Since JMS is inherently asynchronous I believe the mechanism for JMS is quite lightweight +How about we open a new issue to modernize the compiler versioning based on source control? I dont really want to take steps back +Please feel free to reopen it if theres still problems on AIX but I believe IBM Java and Java should be working fine with now. +In fact allowing arbitrary ASCII would be +Changed to improvement since throwing more memory on the case makes it work. +My stance on Oracle is the less I know the less it hurtsWould there be any way to automatically detect if a schema is required when the user is setting up the connect? I dont know enough about schemas to even know where to begin looking to find a query to suggest +yes antmaven classpaths are incompatible afaikwhy dont you use templatepath instead of classpaths +Thanks Uma and Aaron for taking a look at the patchUploaded a patch addressing all your comments +They register a transaction listener and flush the session and close the jbpm context on beforeafter completion +Could the fixer of this JIRA either create a new or incorporate into an existing test a test that would create a for this function so that we do not regress? +This is the reason for the error +I can perhaps email you directly with materials needed to build each of the proprietary connectors but I cannot broadly distribute them because of licensing reasonsHowever Id suggest having a look at some missing inlocalization work before getting screen shots +This is not only a problem with filenames +The other memory adjustment whioch should be made is the setting ANTOPTSXmxm in and which is low enough that the testsuite build most always fails to generate reports correctly due to OOME on Java heap space +Not sure if its new I dont see it on my linux I only discovered it today on a windows machine +This does not hold true when I look at the entries but has a rule been set going forward or should one be set? Please let me know your thoughts +These are done as a +Committed +The id of the snapshot is Give it a go and lets see if it makes a difference +Ramon Casha and Grigoras CristinelCan you verify that this issue is fixed for you by the latest nightlies so I can close itThanksMike +Closing. + +Thanks TomReopening for backport to x +What do you guys think +rebase +All are done in pms colors +Are you proposing a new policy that only complete features should be allowed to be in a single patchThat will slow the community down significantly +I filed for limiting the number of times that the web services returns and to fix the locking in the server +I also clarified the jdocs of DSR and made hits final +Im closing this as it appears to be a non issue with the code tree +I forget to say that the bug is in the classsrcjavaorgapachecocooncomponentssourceimpl +Good point! All done please update. +Subscription section to docs with appropriate links are added +Failing such commits is probably the right thing to do +Can someone apply the patches from this ticket to trunk? Or it needs more work in order to be commited? I would like to have this functionality for the upcoming version if possible +Thus the implementation was shifted because the first goal is to ensure bsp task working smoothly +Our JPA module works as designed +By the way +I just dont think its worth anotherRC all by itselfcheers Roland +Hence marked None +This issue duplicates one that Tony is already working on +cluster is the coodirnator cluster is the failing node + had similar issues it is fixed in +bulk move of all unresolved issues from to +This issue has been around for a while and as noted there is no good solution since the spec doesnt guarantee that the generated keys will be made available +If I remove the path attribute from I do not see the NPE +I was worrying too why have not discovered that sooner +Test failures are unrelated +Closing +This also makes it explicit which core is usedI personally think that the default behaviour of randomly picking a core when none is specified in a multicore configuration should be changed to showing an error message or specifying a default core in the multicore configuration +So we need to give two impl both doing sameNot really impls +I just wanted to make sure that the problem Im having is a known issue and Im not doing something wrong. +Copying over Baileys patch from made this +Nothing lasts forever +The camel component config schema can be found heresrcmainresourcesorgswitchyardcomponentcamelconfigmodelvThe config model for the camel file binding can be found heresrcmainjavaorgswitchyardcomponentcamelconfigmodelv released +The release has the fix for has been integrated in +good for me +Ok managed to create a test case that demonstrates the problem +Patch forthcoming on top of that one +Applied to the trunk +Not sure if theres a way to set this somewhere else than in the title +See as a class that might provide the right Rule hooks to wrap the test with +Verified +the management plugin has been committed to release note is required for this issue as server compatibility will be discussed in the overview section of the document. +Im going to get started on this +I prefer programming +A huge patch for such a small change +it will help me ascertain if this is the out this is just a really bad case of logging +I would agree that the memory leak looks like it could be the same as listed in +Resolving this bug. +Only downside is that if two disks have an equal amount of free space it would be better to degenerate to rather than random selection +Committed this to trunk and +One can write the define statements in a file and run it at start up. +project to +added a new test to make sure this feature wont be broken without notice +To release with these outstanding doesnt make for a good first impression +Thank you very much for the patch +If you can try it please let me know if it works for you +I didnt do any major testing on this but Ive attached the patch Steven would you mind verifying this worksTo install just stop Eclipse extract the attached zip into your Eclipse installation directory and Eclipse +I found one possibility in the connection expiration code +Works fine now. +When we use store the exception is not thrown when a job is failedI have checked for the jobs status in registerQuery method of and threw the exceptions accordinglylooking for suggestions +Link to related issue +Right now its PDF +I will try to remove thatBut recently tried that while running a Spring embeded Jr based web app on tomcat it did not work +Attaching a new version of the patch +Here is a patch +EnisCould you check this mismatch between how user name is calculated +Correction the reactor sequence is Parent Impl Ejb +We can discuss those when you are ready to work on this +This patch breaks Thrift down into smaller packages one for each language library +Noticed that copied configuration file is renamed to a hardcoded value +Johan please find attached my patch for +This doesnt look so good as it REQUIRES a JIRA ID for every commit + +I have an open discussion under way with some folks from Microsoft in the meantime continue to use the term Windows In my mind is an awful name but windows and windowsRT both possibly imply the wrong thing as well +So it will be good to shorten counter names when neededI also noted that client update the partition object one by one from client to metastore which is both slow as well as +All rights reserved +If nobody else will do it Ill pick it up again late in October +If you want a servlet jar with the fixed files you can download the latest Tomcat release and use the from there. +I dont mind making the fix though +tnx for m still having an issue with this bug inI have a check login function which I access remotely +The following command line on a image is sufficient to recreate the failure with Xaosenablerecompilationfalse ocO ocinlinefalse method check er main s m M a LIR looks right to me so the bug must be somewhere in the IA code gen andor reg allocation +Since it is a maven bug Im on updating the manual and maybe a short comment on parent +The file may be under copyright +Then I redeploy the vdb M with version through the admin api +I dont expect any specific problems but it will need checking +The goal is clearing all sub nodes of bspRoot when Hama cluster is started +close issue as requested by reporter. +Anyway works fine + +This is quite annoying and should be fixed +Ha I think I see you you use an ear file do not use them +The problem was caused by missing registrations for the Binding Operation Binding Message Reference and Endpoint HTTP extensions +To make the purpose of this issue clearer I changed its summary +jar not path plugins +Yes it seems highly likely to that logj issue +Its easier than it looks IMO +Patch and test case +Picture is made with DPI +I always wonder what select element is supposed to do? Is it just a badly named and defunct menu or does it have a purpose +I was setting the default setting on the AUTOUPDATE configuration variable to true which was why the testcase was screwing up +No followup so bumping off +The index seems to be created by Spring Build through task jar defined in file common +has been deprecated in v +We had a rule in place for a short time last week that allowed all committers to commit to incubator +I commented on it because I know it can help in the design stage to get various viewpoints on syntax so that the end result makes sense to the most people +Looks good +attraccessordependencies +Verified on build. +Thanks Bhushan +Anything better than the current one should be OK for at least short termAs far as we have the right direction and we follow it multiple evolution steps is fine P +Make it self contained +Sorry I didnt provide more details when I raised definitely know that I have never used modello myself so any examples would be within cant find the maven projects I was working on the time to provide any more details but you could open a few with mdos and check them out +I did following to run the test +This was a feature which was never documented or advertised on the mailing list +The messages were not actually lost they were not dispatched +This will be tested in the first build. +This was fixed earlier +Hi SanneThanks for your comments +zip file generated +fixedthough im not sure we should ive left the local definition of errorMarkup inthere for now +Is there any update on your side? If not we will go ahead and close it +ok Ill ask helpdesk to upload projects together with guides +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsernodes +Thanks! I think you can resolve this ticket now. +Yes there is said NOT stands before a boolean test expression not beforesome string or numeric expression +Adds contructors and one mods one test +Ive changed it to correctly use resolve +Its been hanging also on my boxthere seems a test issue +Mischa could you please be more accurate in future during creation of ticket +Makes it easier for you and me +Corrected affectfix version. +I authored the extension for the displaytag library +The attached patch adds special parsing for paths starting with +All of my info is against this version to continue with tomorrow if the upgrade itself does not solve the problem +Thanks George +The test failures are because the directories webappsnode are missing +Martin can you please put your Oracle hat on for this one iima +Have you tried updating the in our plugin and see if that fixes the issueThat would be the only thing we can do shorttermThe longterm solution is to launch the queriescodegeneration etc +Andreas unfortunately the bug appears in my integration tests only + built wo maven rebuild when maven becomes requires +Need more input info on this task. +Specific CI issues should be directed to builds who manage the CI tooling +will not package into distribution +Just thought Id mention it +this patch allows tthe user to set java property at the client +Im hoping will allow importexport of these settings so we can easily distribute +Are you running with a security manager installed +Thanks working now! + new patch file +Here is the thing So far we havent change our version of code and xalan +push to +reopening to change deleting the m repo I didnt hit the issue again +We should generate data only once +Upping priority +Attaching a patch that fixes this bug +Let me know what you thinkcheers Roland +Do you think it would make sense to try to createdelete a symlink with proper permissions? Will it be the same as with a regular file +Lowering the priority as there have been no reports that the problem persists with Gant +Becaue strncpy is severely +A Software Grant has been signed by CS to contribute this implementation to the Apache Software Foundation +. +Sorry my a wrong build should use build instead +Included classname as specified +The patch includes a fix for described problem for review +Resolved +Maybe the restarting status has a lease? Server must become responsive within a configurable time periodbq +did mvn test clean run successfully on the cluster +When this happens again could you get the stack trace on the datanodes also? Right now it is not easy to know the datanodes involved in the write pipeline +Verified. +Minor for that Paul +Ah I only checked the CP branch +If tx rolls back thenthe instance will be thrown away from the cache so there isno need to invalidate the. +Unfortunately Ive stopped using Thrift so writing a patch and a unit test wont be a good use of my time +Thanks Nadav. +Verified on. +The very simple store mechanism we have written outside of lucene has a gain of x yes over lucene stored fieldsJohn can you describe the approach here? +BTW why do you consider it as a bug +This sounds like an in issue rather than a filter operation issue +Well the TLD is shipped in the corresponding jar and also openly in the specific Spring distribution +Modified code from have a look at and the patch attacht there +news on this? issue is duplicate of issue +Ive changed the ticket name appropriately. +I will recommend testing the patch manually with security enabled +makes sure that the size of invalidateBlocks does not go beyond block +Moving out of +I expect this issue will get resolved as part of the installer +we can do this if we need to retag for any other reasonI wouldnt retag just because of this +Attached the output of jar tvf lib Ant task includes all the jars from the lib directory in the classpath so I dont understand why its not finding sortPerhaps try running the shell script instead of Ant that does its own classpath setupJAVAHOMEusrlibjvm +Emmanuel you want this one +It seems best to remove Serialization from Lucene so that users are not confused and create a better solution +Test passes after the commit of +lets commit this +I think the configuration mapping to a list of RM nodes is an artifact of and shouldnt leak into the rest of the RMOther than that security can be the place that takes care of tokens and their translation to the correct RM +I also feel inits may have different signatures for different servicesThanks your additional notes had some good catches +Test methods in was passed +Your changes will be in tag Ill set proper Fix version once it is known +Oh and the xml getter sounds like a good idea +Apply this patch to project +supports JSF do not run the demo with JSF. +fix thanks for pushing on this kinda fell off my radarThe original patch from Todd had code derived from Android is that still the case +Patch attached to be a little bit more thorough in which errors we catch +This is an issue which affects us in relation to Clover and performing coverage tests on Clovered with canoo webtest +I disagree +plugin has been broken out +Bit if a change from the original patch +I quickly redid Anils patch as a differences app so that the two can run side by side and the requests of interest can be pointed out easily +Were about to ship this in so lets close this now +On second thought dropping it silently seemed outright stupid so added a little warning transition resolvedfixed bugs to closedfixed +The video window now flashes when the user is speaking if you do not make recording test before +All changes have been merged into trunk. +if two L files are compacted with a their each L is striped but since the could overlap the stripes could too with other stripes from different L filesA nice property in is that only L files have overlapping keyspaces with other files all level L have no overlapping keys within a level and no file at level L overlaps more than files at LRight now only major compactions can remove delete markers because only by looking at all data you can guarantee that you will see each KV that might be affected by the delete markerIts not clear to me how we get around this unless we introduce a formal notion of levels and know which L files overlap with a file at level LWould love to discuss more at the on the Feb th +Done +The idea is to use in place of +It still hasnt been released yet but it can be built from trunk if youd like to try it out +Ive incorporated the feedback on the code review page +Browsers seems to send spaces in URL encoded like thisGET qwhajusthappened HTTP think this makes sense but yes number two above is important as the problem scales in parallel with respect to crawl job size big problems index maintenance issues +The feature has been merged back from Git to SVN +task is done +Ive fixed this +this issue is intent on capturing aliases solely +It should also be very simple to add it in +Advised that this is already fixed in. +for reviewing the patch DagCommitted revision +So the junit version would need to be added to the POM or it needs to import the BOM. +But Im guessing this will really be in or LATER +ryan I cant remember if you guys answered this but nothing is in the JIRA so I guess not +Thanks GregoryVerified in closing this issue. +Hi FreddyI have setup MAVENOPTS with value Xmxm XXm +Remy this issue was filed a while ago in JBEE category so just moved to JBWEB +Commented again in review board +This seems like something we should fix in the +NET framework and has utility only in a web application the other hand is well delimited and independent and can be used only as a DAO engine without requring access to a web feature +As per recent changes if scaling up failed on a host CS will try to scale vm on another host in cluster so as per my thinking this issue is not dependent on +Steve can you please attach your Eclipse configuration details from Help About Eclipse SDK Configuration details +But often with these types a problems thecorruption happens at an earlier time either as part of the application orpart of a previous shutdown +is replaces with sh UIMAHOMEbinmaybe it should be sh UIMAHOMEbin to not fail if the UIMAHOME containsspacesWhen the dot notation is used to run it seems to get the parameter passed which is passedto and not the params intended for it in the scriptIt then fails because it cannot find a class with the name of the deployment descriptorin my case bin where is not a valid class name +Nathan Patch is cool and with the fixup in it should be solid now +Got passes resolving. +Actually looks like this isnt new in perhaps new in or perhaps been here forever +This way we can fix other issues beforeJust add the Apache snapshots repository to your projects + +Here specifically you say You cant ask who deleted rows from table x? because delete is translated to a WRITE action +The fact that any of these fixes dont have the same effect means there are probably bigger problems to solve to really find a satisfying answer but in the meantime I need long jobs to not time out +The deadlock discussion is now separate from the clock specification per se +Minor commentsNits Rather than pass timeout and allowedRange to wait Id have them be constants in the test function seems like the tests shouldnt require different values for these parameters +see Steve Thank you Its Work fine for static inner class JPA. +I think we need to pick up discussion on this issue again +The jBPM console does not see a transition attribute set and simply takes the default the attribute should fix the problem +what is in +changes are mostly in private to protected +Thehidden mechanism is related to access control since the rationale for such hidden content is to avoid the need for complex access control handling for things like query indexes and other internal aggregate data that might otherwise expose content +Why the addition of the Accessor interface? A comment explaining this would probably be good to add too +Changed Target Version to upon release of +reopening for N Tools doesnt show up in SoftwareUpdates in JBDS +Sure Ill get Tim on itbtw your last question i remember ive replied +We should try and make it such only would require code preprocessing and later hadoop versions work without preprocessing +Should they be implemented +Yes just had a look at the WFS spec its indeed correct you can use resulttypehits a new bug report be filled or will this one suffice +Milind othersPlease review new namenode benchmark +This fixes Maven tests setting the proper Jetty dependency and specifying tests fork mode as pertest +Hi Aidan can you review this change please +I just committed this patch +Regenerated output for more test cases +I checked and it seems ok. +see I havent looked into +This makes the current locking mechanism not work as intended since you will get BlobClob objects without doing explicit getBlobgetClob calls +I fail to see how its an unordered Set issueIt might just be a result of which we had a long discussion on how to fix it wo obvious conclusions +This might get confusing if we have a rename table functionality but at the moment we dont +Knowing this number is useful when you need to make slots for the speculative tasks or when you need to preempt tasks from other jobs. +With the AS different webapps could use different cache configs and thus have separate values so I dont see much benefit in pushing the config for this up to +if its just you feeling this is a too big of a change for the sake then I am not moving sorry +Fixed and merged to fuse branch. +The current class doesnt set the flag for all contained +just find we are using Portlet we could not define +I solved this problem by using a pagecreator instead +Pierre +Dont need to finish this feature for +zip file with the codebase for this component +Closing this +Yes mvn deploy produces the same if was blaming the wrong component +Could this be a bit smarter and if there is no unit assume megabytesIll wait for your weekend testingimprovements to give it a spinThxs +It can make a big difference if its frequently used or the size of the input is large +Next time please ask such questions on mailing list this is not what JIRA is for +Test cases for map values containing close parenthesis and closed curly braces +Thanks Meng for doing this work +Is this what youre referring to with regard to trying to provide a default value? +It might be that we instead need to extend the Tree API with methods that control the ordering of childrenwe already have thatthis issue about hidden items in general see the test case +We will upload other jars soon. +I realize this may be expected behavior but AFC shows them so I want to make sure if this is by design or by mistake. +This problem goes away when I use jdk SR +Thanks Ryan +But when I saw test passing on my machine I left it m going to disable test as it generates spurious failures +Finally I got it working +Committed the delta patch to after fixing the logging. +Thanks for the suggestion! Attaching a new patch +Thanks Suraj! If theres additional ideas Please open new JIRA ticket. +Closing all issues +I may be misremembering what the Loaders currently provide +Committed to svn +It fixes significant problems with the Kerberos support in the branch and trunkThanks for considering it We should write some unit tests for this +bulk defer of open issues to +The centric trouble here comes from the fact that some ops care about the global under replicated count which is fair as there isnt a granular option +Patch applied at r thanks for your contribution please verify +Imjust saying wed learn more from the perf test that way +Integration test added +Added the original exception as constructor argument to the +This patch should be up for review now +I hope to address it in the next couple of days once Ive got a firm understanding of all things involved +Paulex everything looks OK thank you very much. +I could recut it for if that would be better +The compiler stuff is super right now and you guys both have cases we could repurpose into real unitfunctional tests +Changes to fix this issue +Attached script as per but with extra lines as per Stack +Per regla de polze les paraules amb la seqncia our atnica i sovint final sescriu or als EUA colour vapour labour neighbour harbour etc + for patch +Hi Guys Any progress on this issue at allForrest is looking to a new release soon and we have an issue that depends on this one currently scheduled to be fixed before the releaseThanksGav +Hadrian I se no more bug in there in the cleanup +I am glad to hear you got this resolved +I missed the vote period but I would vote for BWill there be different images with different sizes produced for thewebsite? Henry +Will check back now builds DONE +But Id be fine with the second approach also +This issue contains the patch for +Patch against that resolves this issue created from a subset of the changes made on the trunk for this bug report +I believe this will cause most NN failovers to be accompanied by a partial or complete failure of the clusterI agree +Can you tell me how you are creating services? I just run test on one of ours examples and documentation for multiple services works fine +Fix with +A new class is added and registered to +It mustve been solved elsewhere with session munging +The tar file contains a wsdl a test file and a class that implements the web service +The fact that Grails goes out of its way to ensure that the is after the last +Committed. +So all issues not directly assigned to an AS release are being closed +I see +SOLR is all about performance +The issue is that characters with accents are not escaped into the proper HTML entities such as eacute instead of +Isnt it enough to disable auto focus on the form +i removed the required classes because the numbers are validated in a separate method which validates all three input at once +my pleasure +As an example what would be the problem with pulling up +Additionally for obviously simple problems going straight to close is ok in my opinion +Thank you Tiago for picking up this issueWhat seems to be happening is that for logical connections when we close a prepared statement it under some circumstances tries to reprepare the statement +Currently any attempt of running startSlave that does not fail authorization will report that slave mode was started successfully +Good point tmp files are now deleted at startup. +Thanks Aaron! +Ive created a new issue for the problem with the and mark this one as fixed. +regex is not +Mark yeah thats pretty much exactly it +Closed. +But in the long run we should add support for incremental updates and signed documents so that this workaround will become redundant. +Thanks +We spent some time not so long ago to rid Cayenne core of dependency and now it is coming back through other channels +is still using the string when parsing properties from the old harness +The also dont seem to be included in the and that go in the root directory +Provided a patch that adds the value alwaysRecompile and some +Thanks for making the new JIRA NicholasThis latest patch introduces and documentation changes +Part of a bulk update all resolutions changed to done please review history to original resolution type +Hello ChristianIs the mongo appender going to make it to the logphp? I kinda forgot about this issue +I think we better mark this as wont fix it has been a task since geotools +Closing this since we havent gotten input for some time +fix new nativetests execution should be in the native profile not the KDC server profile +Some state could be returned to clients +As long as the according import is there it should work in almost all cases +Here too I was testing the patch and my local trunk was compiling without m new to the process so let me know if there is something that i can do to get this put into the trunk +Attached file for inclusion in the images directory as an original Draw format in case further editing is required +when i plugged in +Weve had some discussion regarding this of course +Added the upper and lower bound to the submitInternal method +Sorry for the nomenclature error. +Resubmitting patch to run through Sorry I am a little unclear on what the patch is doing +So the parser loads it and there even is a features variable but it contains only null for each attribute +Well if you agree with this patch Ill apply it so that we can close this one then open new issues for the new problems we find +If anyone would like to change their visual appearance to make them better that would be great +If need a particular feature or bugfix to be included in you are welcome to get involved and contribute +However its best if the RM or someone with more context takes that call +sorry about the Windows update +Svetlana can you explain why you think the decode behavior is wrongThe test as written in our SVN repository passes on Harmony code and the Sun implementation of JavaAs Vladimir wrote this may be a compatibility issue with the BEA implementation and I agree that our behavior is to spec. +Ill post what I find +svn commit m EIP Splitter and used together cause problems in xpath evaluationSending patternsTransmitting file dataCommitted revision . +That should help with debugging and maintaining the code +The has been modified to use NIO to resolve concurrency issues +bq +The resequencer also had the option set by default as shown in the UI +That is even nicer and makes my life again easierThank you +Will commit this +Committed to trunk and branchxThanks! +I will have a look ASAP. +For most of the array values IM Addresses emails photos set to empty array to remove entire property +Shawn please see and comment there about the suggested work around +Le danger de la paperasse dans lUnion europenne +What led me to this conclusion is the fact that the is never used throughout the chain of methods being called from an event being caught and the connection being destroyed +If youre listening could you take the example in and modify it somehow to get a delay? Or alternatively can you attach another example that has the delayThanksll test and get back with the results +And second because I prefer to have the order clause within the SQL statement and not being externalized in the client +if B finishes before A we need to wait until A is done otherwise the deletes frozen by B are not applied to A and we might miss to deletes documents in AThe Ticket queue simply ensures that we push the frozen deletes and the new created segment in the same order as the corresponding DWPT have started flushing +Im also experiencing this +my bad added if for the package target +Thank you very much! Uwe +Not supporting volunteers in an appropriate way is sub optimal for our project in the worst case they loose their motivation and went away +I mean I dont remember if the spec mentions something about cant think of any +Thanks for all your work +The real issue here is that Continuum isnt using the default goal stated in the pom +Looks like a debug symbol problem to me +set the two parameter only if the operation is two way +HelloWould you please try my patchThanks a lotBest regards Andrew Changes committed in revision + +Ill think about it further and let you know a possible plan of action +You mean Microsoft Visual Studio voluntarily creates pathnames containing forward slashes for separators? That would be a bug in the product that we should report to them but I have a hard time believing they would do such a thing without seeing a test case +Verified in ER +Where possible this is done within the existing method namesDavid GSTVAT patch seems to handle the cart and the creation of the order correctly +We cant reproduce this on our end +I have incorporated them in rev of this patch which Ive committed at subversion revision +Dans the manAfter talking it over with Dan he convinced me that it should be done +That is to say when using QUORUM or ALL were getting duplicate rows but when using ONE we are not +If you want to restore the selectedItem when changing out the dataprovider you should probably be using a data structure that supports fast search to compute the new selectedIndex and then set selectedIndex which is much faster than setting selectedItem +My original change to wasnt quite correct +There was a time when the Namenode was littered with plenty of adhoc data structures each for its own purpose +From consolesrcmainorgjbossconsoletwiddle changes for this patch were the attached files coming from the latest in this code rarely changes I think you could just use those files instead +On it +The samples are built as part of the top level build but not as part of the individual runtime project builds anymore +problem was that state in logical node table not reloaded on master +Work in progress for Schema Repo attached as setup works as follows +the latest patch looks good to me +New ticket +There are lots of places where bundles might get loaded +I thought java exceptions that percolate all the way up cause a exit codeEDIT Hmm +How does that look to you Mikhail +Lets take his advice +I didcvs diff u and then just appended the new files at the end +this is what you probably meant by data of VM structures +Im fine doing this if you are sure they did not change this binary in any form or fashion +Verified with. +Changing the summary now +Attribute is now excluded from replication +Thank you Laura. +Feel free to reopen if this is still causing problems. +are added for the test cases + +We should however add a test case to the unit tests to check that +But thiscode doesnt have a lot of lifetime left hopefully the release after next willbe JCL with significant work done on all parts of the libraryAnyone else got an opinion on this +If you explode the archive you would see that it also contains a war and ejb projects and they have no problem with only the ear project +Perhaps also describe your environment and what commands you are running etcHelp us help you +Maybe is the root problem for this when there will be some patch ill try it +jaikiran +And added the filter by endpoint uri as well like the send totrunk all tickets +Reassigning back to Galder for review +OK I got what you mean +In otherwords this once worked and the bug has been introduced in some more recentversion + We have to take care that the number of modules we provided does not explode +The attached patch fixes this problem +Moving this issue until we decide on a common configuration to run the now it needs the actual server to run to do this test +NET CommonTools too +A small amount of introspection is used here to retrieve the subreaders of a +committed. +jsftests +Moving to for Beta we want things unsecured for and will put a big fat warning somewhere +Is this a fix for onlyDoing a get instead of a get is good because well handle deletes +Are you sure that problem is fixed there +ran ant test +Hi lufeng yes +Attaching patch for Y! with the review comments incorporated +Here is the new patch +Moving out of +Ive recreated the patch so it should be fine nowThanks +Doc into text at your end using POI and then supply the fileAre you sure there isnt a problem with POI failing to parse the file correctly +Hey thanks Luc +Looks like we have to start looking into +For example in UIMA its often common to run multiple instances of a component at once but with different configurationWhat do you think +Sorry for the spam attached wrong file last time +Wow this is a mess +Also it will be great if axis documentation can be updated with this piece of informationregardsCharitha +This causes the web UI to prematurely indicate that blocks are missing when in fact the NN should still just simply be waiting for more to report in order to exit startup safemode +Sorry +Moving these all to a Doc release version +The has been suitably modified to +We still dont have a proper fix right? Moving to the next CP +I made a careful trawl a while back to make sure we were using everywhere instead of literal s but its possible that I missed some or more crept in +Ive sent them all to the list +All is implemented expect for the automatic deletion of plugin being no more embedded in a master plugin when updating this master plugin +thanks for your workthe component is included into CVS +Patched +The implementation could of course reuse code +This seems like it means replacing progress calls with more overhead and perhaps synchronization just to make the code prettier? Giraph is already up against the wall especially during INPUTSUPERSTEPI think the beautiful solution youre looking for here involves a move from Hadoop to YARN +Full documentation with examples will be included in the next s greatthxMartin +Patch AttachedIm new to git coffescript and gradle so Im hoping this is an acceptible was an oversight that this did not exist all along. +It seems to match the given behavior fairly well +What request +Thanks for the test +Guys Id appreciate some comments on this know the long term solution to this problem is a DBMS storage for the config but Im looking for a solution for the present today something that allows to be usable into a multi admin I forget some obvious change +fixed +Move some stuff into methods shortened the comment and did moveMinor nit Should there be a comment on when the point of no return is +I will send you a service to reproduce the bug later today or tomorrow morning I have to create an empty example without our real code +Merged to the branch in revision . +This is not ideal but we can update the tests incrementally then break this dependencyWe now list the artifacts that are published so they can be published to the local cache +If that happens again just ping us by commenting on the issue or asking on the mailing list +Patch committed to trunk +In that nightly build the error did not happen +applied patch thanks Andreas +A release shouldnt depend on manual CLI arguments which would result in a customized build and therefor unknown how to rebuild it in the future +Would that also be fixed by your changeThanks a lot for taking this up immediately +Nasty +HiThere are some other missing scm element in jUDDI modules Two possible fixes adding scm informations add the parentSNAPSHOT for this modulesConcerning karma currently you have Project User Scout Project Administrator Scout Project Developer ScoutWhen I will be able to add jUDDI you will have the same +Earlier the two modes had separate +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +ICLA submitted +pls check +trunk +If implemented in core it will be an API change that should be dealt with before the freeze +The result set is still empty when I try Alphab +We can deprecate the old method name and institute the new one but I dont see this as a major issue +Does this happen with the latest myfaces code +Committed to trunk. +I suspect that the specification is in need of clarification on this pointShould I take it up with the spec list or are they already aware of this problem? +Thanks Rob! I committed this to trunk and. +Thanks Hitesh and Vinod for looking into the patchIm sorry for the late responseI have updated the patch with the suggested change +this wasnt a Continuum issueThere were messages in the devcocoon moderation queue +Thanks AhmedI put this into trunk and +Also be aware that the shuffle was completely a year ago so the version in and trunk is very different from the version you are running +Revision . +Thanks +It would force me to handle the null case as an error when I already have a application in production which considers the null case to be valid +Thanks Mayank! The patch is updated with assertion on mapreduce property names added +deploy the site as well when youre done committing. +Applied a cleaned upimproved version of the patch in revision +Patch with a test case showing the NPEThe question is how to fix it +Thanks Paul! Good catch +We have not been able to trigger this on +Verified fixed in documentation set. + +HiteshI assume that you are aware of the code changes required to fix this JIRA +Request listeners started listening on the queue right when the relation service created them +Hi SvanteI looked into this issue today +Patch submitted +Implemented default deserialization constructor generation in Java backend in r +Thanks +I use activiti message boundary events works fine for user tasks +New API implemented and upgraded trunk. +Here is a patch with test case +If you dont solve the issue of the zk expirations this is no real surprise +If a caller does not act on this information this is not a problem of the interface +Closed. +Thanks for the quick updates guys +Change subject +Patch diff file for adding facet reverse sorting +lgtm +Is this something wed need to tackle or is this purely for Rails to patch +Im just going to close this +Bulk close for Solr +Thanks for your contribution +I just committed this +I think the gshell plugins are not getting used in this build but we can look at the dependencies added to the plans and see if those are reasonable +Patch applied as of for a while +Maybe we should contact him and ask him if he is willing to contribute to in generalThat would be great +I am fine with implementing this for both HS and CLIHS and since we use strings for HS and HS at present +Juergen +HelloMy bad +Yup I see your points +this sounds like the user database is not getting linked up correctly with tomcat or something +In any case I apologize for raising a false alarm +Next patch will contain the fix. +Please reopen it if we get a negative it if we confirm its fixedMarking as resolved for now as Im not doing any work on Topic is an aggregation of queues +The might be reason to create a check and verify the attribute is present during init +Patch for to handle filenames with spaces +Here goes Client name includes the name of the thread where it was created in +The current method looks to be more of an online algorithm that processes module names a section at a time +The user may change them if neededthumbnail! The launch configuration is reused on the second and the subsequent clicks on Run As Run with +Do I need to leave this in Open status for the patch to get picked up +No longer applicable +Fix for this issue is related to collecting information about references to renamed seam component and presenting them in standard refactoring dialog +This means that this issue only needs to be documented in one set of Release Notes and not the FP Release Notes as issue will be closed again straight after the fix version edit has occured. +This japanese code is not application as it is +Rather than do it twice I figured Id just work on HA branchIll do a round of benchmarks on this patch tomorrow +I will fix this and post a final merge patch this evening +First cut for so that others can reviewcollaborate +committed +I will fix that for Karaf +Thanks Devaraj +But we need to assign it back to some legacy application field that unfortunately requires a specific Collection type +bq +Otherwise you would need a kind of failsafe set invocation in the lifecycle that forces all components to be in the correct state which I think isnt the best way to goNonetheless I think the framework is very good and my use cases shouldnt impact your design decisions +Unfortunately not because if any of the URL mappings capture the request URL then Grails will attempt to dispatch it +Hello all Ive made fix forMy build for plugin is attached +Ill install it in my local repo for xfire needs to be included and deployed later to maven repo asxmlschema this is a good time to ask how will multiple POM versions be handled at? ibiblio doesnt support m for option as work! The +The control has been implemented unitary tested and tested in a and confirmed +In addition hard upgrade typically enables new featureso Soft In this kind of upgrade data in the database is NOT changed +So basically there is a new rest endpoint on the server that publishes configuration data upon requestin the client side there should be a new Configurator that issues such a request and applies the result to the local Config Admin service +Snapshots continue to store table descriptors directly in the snapshot directory and is refactored to share code to write sequenced descriptors in any directoryThoughts? Im leaning toward the last option + patch looks good to me +ie +Marking as invalid but can reopen if needed +It does contain multiple functions +Reopened to add fix version. +Checked this time +NamitAny comment on this one +Tooltiptitle would be something like Email webmaster to report broken linkswebsite issuesThoughts? +I have that exact environment set up but cant reproduce it +The cluster example has been added to the kit again in rev +I can commit the patch if you update it with the new output +coretrunkrepository +Its a kb patchObviously not on this patch +But we only want to allow that the keyB is used for the verification +The newly created page will use the parameters which have been send to the expired pagePlease test it harder and let us know if there are any issues +More information required for SOA releases notes should be using in AS environments at present it uses is a classmethod? in ESB? I couldnt find it in the Likewise are methodsclassesenumerations? The AS behaviour was being used even when running on AS +KonstantinPlease review the documentation for technical accuracy +Then it comes down to a sensible default configuration for the userPassword attribute +Previous comment is about of course +And I suppose when using the Id need to provide it with the in order for it to avoid +Stanislav Poslavsky Did you give Pauls suggestions a try +But I would not object to closing this issue as Wont fix. +Hi ColmNo its not possible + should be pick up from on the classpath to avoid having in +Thawan would you be able to add a unit test for this +Regarding CI are you using Cloudbees? Can I access it somehow +If you drop time range how would you get all the versions between two time stamps +So some kind of needs to be used +Youre right I will fix it now about performance commentary was my error when sync two classes +Why do you not just run a mvn compile prior to running Sonar? Can you switch to the mailing to continue this thread +This has been something that everyones been clamoring for for a while so now they can have what they want! +The patch is doing it only for testing and we are not actually running with the setuid bit on for this part of the tests so there is no risk of a security issue +Edited reviewed and committed on and trunk +xml no? This way any framework level changes that handle the frameworks configuration would not be tied up by specific scheduler implementation +LGTM +I reviewed it and posted a note in IRC but that looks to be missed +I see +This was probably added for debugging +I thought that the number of those who already know locate functions signature are fewer than those who dont know +Still failing on Sept the example to more closely match the runtime example +Missing methods +Thanks Kay Kay +It is when I try to invoke the CP +There are serious problems with required fields in JSF and therefore its best just to wait for JSF +IMHO I think the javadoc has to be clarifiedThanks +All current samples work as expected. +In this case we want to ensure an older write does not overwrite a newer one +If you truly cannot do it in JSTL I suggest extending on of those tags or using your struts extension +Ill take care of that thanks for your help +Those two new labels will be more meaningful +Thus it doesnt optional choose or just simply a instance variable +Attached referenced patch is the entire implementation +Will commit if tests pass +so I can make the final RC then Id like to get a few extra days for the voting process +Thiruvel any updates on this we need it urgently in order to deploy +if this gets rejected somehow +My apologies again for putting something through rushed last time +Yes +Yes these are socket timeout exceptions +Thanks Lohit! +I have logged to track those changes +still +Can you validate the solution +The jar name is only appended if there is a name collisionIt seemed more natural to append the jar name instead of prefixing with the jar name but I could be convinced otherwise +should be fixed with using wagon based on asf if still an issue. +Okay Ill agree to bend the rules a bit and remove these public methods that were in a release since we dont think anyone will have used them +bulk defer of featurestaskswishes from to +I just added a document describing the inter cluster communication as implemented in branch + +Test case in the form of Eclipse workspace using hsqldbColum productcode is defined as NUMERICSame exception happens as using Oracle +Couple comments does a good job of cautioning the users to be careful about the correct syntax when using the clause +So could you please create ticket for this here +OK I cannot reproduce it without the original code maybe I can if I ever find it +Correct stuff as they show up +Applied a lightly modified version of the patchThe patch had a bug where the key and key fields were passed in the wrong order on creation of the this was picked up by one of the existing unit tests which failed on inital application of this patchThere were also a couple of deviations from our style guidelines. +Rebase previous path to Thrift HEAD +Setting of min widthheight using JS options should be of a bulk update all resolutions changed to done please review history to original resolution type +This is the latest version of the Apache Wink Developer Guide version + adds support for Thejas looks good +Fix findbugs warning in v didnt see the same UT error in local environment so rename the patch to v and submit to run test again +does concurrentcompactors disable compaction? that would be a lot more intuitive than the current minmax voodoo +If we want to just pass in the actual sqoop command then I need to write a custom sqoop parser for handling both regular and +done. +bq +Sure but it still seems like it is easier to say The first gossip message will always follow this format and keep that in legacy mode if necessary than to add this extra layer +vibhors patch for added an avroSourceSink and thriftSourceSink +Looks very good till now +is also a in XML processing +The code looks good +Sorry I didnt have time to get to this +Kent please open a new issue or discuss on the forum this was working in our tests so we need to see more of your setup. +push to +Committed revision Thank you for your contribution PatrickJarcec +Woops didnt see that attachment for some reason! Sorry serves me right for talking an at the same time. +Thanks +Keeping this issue open as more work need to be done to completely secure all the in these files +The bug is no longer reproducible +I had to modify the code to cause the issue to show up more often +The node does not get a real ID thus exchanging a node cannot be detected by the tree. +Reinstalled Tools tried with both embedded and external graphics +That is right abstract super classes are best candidates by for simple parent by default is not so goodIll add Add for base classes checkbox +running on Redhat +cvsrc is not a reasonable option for eclipse users as this will affect all their projects and not just maven. +Closing it now +Changing to span when the field is already in some inline type element is fine +The which has been ported to the GACP branch works fine for me +what version youre using? +Modify widget detail page for Admin featured widget check box for admin user detail page +Still cant see the functions but manually adding works. +Good morning Jacques i hope you had a nice WE +Added validation for +Just created a branch to test this how do I run it? Like a regular release +Marek did you saved this articles as draft and after some time published it +Hi Norman have following priority I cant check with different prios because doesnt exist! Its just a fake name to disguise ones true and I cant change DNS configuration for it +I reproduced this +As soon as only the automatic properties created by hasMany and belongsTo exist the error goes ll extend the test case with the workaround and try the original version with as soon as the bug if fixed +When I change the hostname to its IP address it works +Orchestrator and Service Manager to employ new Crypto constructor signature +Maven supports a greater breadth of SCM implementations which I believe should be considered here +Im testing it with SMX and it seems good +I suppose those issues are all relatedThe logic for the files navigation into jars is in srcorgjrubyutils +I thought a bit more about this and I agree with shutting down the bookie in the case the sync thread catches a runtime exception +Committed +Due to the way works everything has to be part of the CVS which means that all the jars have to be versioned also +BTW any objection to me removing the sqrt calculation from +I took a chance and contacted the guy that was named in the JPL capabilities and pointed out the error +Note that the activio lib released with AMQ now uses +Thanks for looking in to thisRegardsShould we close or do you want to provide a patch +Thinking about this I realized that the big main based Jar we distribute is not what someone should use for embedding +Heres a simple patch that changes to +I am blocked with this problem +Just tested client with IE still working +bq +We do not support the RMI if a remote connection is needed the protocol should be used +fixed in revision in case of bad credentials cluster code was holding an autoptr to the cnx in its local cnx map thus preventing the broker cnx from ever being destroyed +NOOOO +The best middle ground I can see is instead of deleting the old file move it to a +It is possible to create a new node named aaabbb +Marking as resolved. +And in that case the first argument could also be changed from to +to fix I will add the Dynamic Web project facet and this should fix this issue too +Which version of Java +Please test. +This one succeeded the second timeCommitted thanks for the review. +Were trying to avoid using JNI code in our next step +I have a similar test over in +Bulk defer all frontend issues to +affects versions to see this problem also +It would only check if the cell was deleted in the not the memcache +For latter problems please refer to and instead of the two issues above. +Maybe it doesnt like that I uploaded to files at the same time trying again +I agree with Prescott that it will be easier to manage as part of the project atleast to start +I agree +This issue was still open when I marked as released and moved open issues to +Doesnt affect CQL in either C version +can improve this later but this is the easiest win i think +It seems like it would have the same effect as having multiple classifiers but works with the current implementation +Please suggest where should I look for this issueThanks that is strange +Let me know if any other changes are needed +I can ask him however +This requires to be fixed +I tested the patch on windows for the sample clients +So the best solution is to make add the jars to +I committed this to trunk. +This issue appears to be Microsoft specific +added to backlog for revision of fixed in r +So do I have in the wrong place? in seem valid locations to me? The question Im asking myself ist WHAT makes my pair of example ears pathological +But you provided a patch and so that means this gets to be fixed. +I see just problems a picture is worth a thousand words please see attached Pizza tree issues blocks of options pizza price increases spectacularly when checking options boxes +bq +One change +I added a new feature that will allow timezone offsets to be suppressed and nowby default the timezone offsets are output I didnt use your patch Leif I didnt want to use but Icertainly appreciate you taking the time to submit the patch in the first placeand alerting us to the problem +If that is slave on which the test is running does not have a loopback setup correctly we could see these hangs +Cool thanks! Will get this applied +I also believe the default updateEntry implementation is incorrect as you should not be able to create new entries in the update method +Ill file a JIRAto against the modification may cause failures on other elements also not only forBut as a security mesure we should use this patch for WSSJRegardsWerner +This issue should be fixed because the fix for also works for these two samples +Thanks but can you patch it using svn diff please so itll be easier to apply and well know which part changed +Hi Manfredthe behaviour of is correct the DOM specs require write to use as I see +I dont have time to finish it so Im parking it here for now until I have more time +branchx Committed revision +Thanks Om +Are we willing to sacrifice this functionality in favor of running in managed environments +This is a bug which has been fixed in the latest CVS +into +and Ill archive the code from the example with it so that it doesnt hold back the rest of the showcase +You can expect a new target in from me soon +Patch for correcting the expirationtimestamp formats. +Apologies for the delay please post the patch you mention above wrt using this to fail over between federated to +I tried it +Seems vmware has API to configure cache policy not sure about xenserver +Thanks for the checked it and surefire crashes even without forking +I tested GWT and it appears OK but would appreciate you testing as well +David can you please check the FIXME comments in the and make sure that we need to exclude those deps if not plz remove the comment and close +See for an example of what to do +This has been committed in r. +Latest snapshot should now have that fixed. +Yes you can start Avro migration +Will Hibernate support having the XML fetch profiles override annotations +also the internal representation only changes for resources that are placed in a higher level package than the scope class +At the end of the day its not that big of a deal what we call it the work will still get done +OK looks good +i am still on FFseems everytime I upgrade with ms I have to spend a week dealing with things the ms has messed up +Classes removed from build +Suppose the geology department which requires a bunch of cores and the history department which runs jobs that need a lot of memory decide to get together on a cluster +These exceptions only appear in the transition period between the server shutting down and the plugin realising that is the per Charles suggestion. +I have added an on textarea to show some messages when moving the mouse over the textarea +Stop the server +Can we do that in another issue? I agree that it isnt a good example but its still a good test since it guarantees partition movement between workers +pls use your actual server host name not a name that doesnt exist +Ive not looked into the pros and cons of this yet the current site is generated by Forrest and the generated image is checked into svn when we publish +Anoop Excellent work +It seems like this tests have been failing a lot since gettingAndrew do you have time to take a look? Otherwise I think we should revert this +I just committed this +Could have saved me some hours if I only had found it earlier +You cant really get an HCM legitimately not wo doing some instanceof and casting of an returned when you do a getConnection +Sample with it has already been fixed +caused by incorrect jar name in the issue now that it has been resolved. +Ill ping here once this will be done +Committed in rev +We likely will add one more that combines both a D and the tiling capabilities into a single mostly as a convenience mechanism +Not really must fix but we must resolve the regressions at least and I want to make sure it shows up in our release checklist. +Thanks Fredrik +These stats DO NOT pull anything specific to tomcat in a Jetty image +I am making this dependency because only when gets committed should this one get committed isnt it +We no longer need to validate hosts + a simple fix +I have been bogged down by EDI compliance projects here that I have not had any chance at all to go and demo that functionality +adding a test case to iterate xsd validationsI dont believe itCan you attach a jconsole graph? Can you run it for more than min +Completely rewritten benchmark class utilizing real data +I can add to doc too if you do a two liner. +There is also another todoTODO We dont proxy the container managed PC yet +project +properties cases +Patch looks goodIt would be nice if you can run the patch through test suite by specifyingll run the test suite on this patch with Psecurity locally now +rev corrected the required status of some properties to reflect what information is really needed for binding sets +Please close that one +Correction +The default limit in is now which should be entirely sufficient for typical web data binding correction With autofalse we actually do historically try to grow lists but use null values there failing if null values are not accepted +attach a patch to clear flushing flag before trigger shutdown +You have not bothered to addfunctionality to the application to set a user to activeintending to do it through a sql interface to your datastore in the rare instance where you need to +and are the classes that are visible outside the packageI need to write a few more tests but thought Id give people the chance to have a look at the code over the weekend The patch will enable some tests but none of the new code will be used by Derby yetPatch ready for review +This issue also exists for Hibernate +We wanted to target branch but I am afraid that it might be too late as it will take at least couple of weeks to come up with patch +I will move this issue to MNG +Made changes such that id is now evaluated against the stackIf id attribute is not being specifed webwork SAF will used the name attribute which is mandatory in all ui tags +I modified srcwebapptemplateslayouthtmlcolumns to get the header and footer to function +see the comment in has been deprecated see +first step in think that I understand the problem now +Ill create test cases and attach a full patch later +This patch applies to both trunk and branch +bingo! solved might be useful for the Roo devs to try building with Maven as well +ok +Yesterday there were some questions on this one on the chat IRC channel +me blames Igor Or maybe theres something in the process +Merged into the branch +Ive taken this to mean to allow tcpip lookups +fixed +I will start poking around with this a bit. +hibernatetoolstests +Vinod I see that the ToolsConfiguration is standard throughout so it probably doesnt make sense to remove that +If activemq feature installs correctly with this patch the default broker configuration is missing +Closing as fixed +We are just fixing an assert to be explicitly checked so that IN case something happens we can report properly + +Nobody does +Im seeing the same issue when I set a priority on a background job with the same backtrace as above on +Its possible to create more than million treads for now +Heres a test case of the bug thanks for reporting this Simon +Thanks +Do we really want to enforce subclassingAs for per class vs per testcase if your test case has timed out then it will terminate the whole class run wont it +JessicaHow is this patch coming along +If this this sufficient for the moment please close the issue otherwise clarify what needs to be done +I like the median thickness option! Definitely better than least weight and much better than full weightAs Goldilocks said This porridge is just right +submitting again with javadoc warnings with comments from Sanjay and Tom incorporated +Is the issue still relevant? If so what are the ToolsEclipseWTP versions? +The unit test provided fails under current version of MRI and +hmmm its in branch too +Andreas i tried to download your patch from Bugzilla +also updated due to autoboxing issue netuisrcorgapachebeehivenetuidatabindingdatagridservicespager jira from controls to build since its tracking netui now BryanI have reverted the changes I made in Controls to accomodate earlier versions of Eclipse that have issues with autoboxing +Patch settings are not passed to the store to get stored +Eli given we need a test for checking right exceptions are thrown from for various file systems I have created to capture that effort +moved to get regular create a class dialog in M +stack you can ignore this patchIll give out a more considerate patch +Priority set to minorI agree it looks like a locale issuse what locale are you in +It doesnt bring much else in for dependenciesDoesnt this suggestion just get back to the original problem of a circular dependency +Id rather do something simple that works and then optimize as necessary +Part of a bulk update all resolutions changed to done please review history to original resolution type +Needs forward port to trunk +Pre refactoring we had package cycles +Is there a timing when this issue will be solved worked onIf Im reading this right this is the intended behavior +The Zip contains This contains the patch to be applied on +This offers a flexible solution where you can create a public facing controller which does only allow list and show and a admin controller which allows all actions +Committed a fix for that in revision . +thanks Marcel for the final push to get this doneIve pushed the patchFixed with revision and . +corrected code that was extracting wrapped exception +There have been other times when situations like this have come up and the entity engine was refactored to accommodate itSo lets do what weve done before add another attribute to the datasource to control how the SQL for the function is generated +It works for me now +Im closing this as wont fix because it has already been discussed on the mailing list +New patch include more unit test and bug fixes +Too late for +must be fixed before can be finished +by an external source +If this is fixed you should close this issue. +fixed above +Attached patch for branches +This is pretty straightforward for jetty slightly less so for jetty Im not sure about tomcat and AFAICT impractical for openejb at this timeThe problem with openejb is that authentication happens independently of which app you are dealing with for instance when trying to get the ejb remote initial contextIll probably commit my jettyjetty work and keep thinking about openejb and tomcat +Closing +Document is now present. +Still needs work in the DOM parser though. +Ok if its already doing it then no need to do anything extra +The cancel attribute should be the first thing checked and skip population if it is canceled +Clarification It has been moved to a Jackrabbit Commons subproject and is not part of the Jackrabbit proper project any longer +Hi Jacques I agree totally +Patch +Just applied the patch on Solr and it runs fine for basic scoring and ranking searches using the default +Running the test on my linux machine +This is fixed in trunk +I verified this page +Thanks Owen! +Also prune the jars in lib do +Sorry that I havent had time to review this in the last week +I should probably attach to Jena instead +Attached file is the file that cause the problemBest NPE issue was already fixed in current trunk but the test file still threw an because of an unsupported character encoding +Type is a byte only in KV +just run the unittests with and see what fails I swear that this TABLECATNAME worked with the latest h at the time of committing itI also know that H author changes the structure of H information schema at some point +Im seeing the error but havent figured out whats different as a result of my patch + +In fact in the Apache Sling bundle I made use of the ability to inject a custom to be able to use Slings mime type resolverSo I would actually favor a solution which would provide a Mime Resolver +The patch resolves certain formatting Sveta Please let me know what you think of a still newer version +Wed be polling the NN with reqsAre we really concerned about a s latency after the cluster was idle for a long time? There are many other reason why replication can get behind further behind than s +If we map to Axis policy subjects to Attachment policy subjects I think the mapping would be like Endpoint Policy Subject Operation Policy Subject Message Policy Subject Endpoint Policy Subject Operation Policy Subject Message Policy SubjectSo IMHO there is no policy subject in the Axis description heirachy that maps to Service Policy Subject +Hows this coming along +Please try it out and close the bugThanks again +It doesnt need to be committed to branch because of that +the test coverage was fine only the case where no style is given was not covered but that was neglected for all methods and does not have to be fixed with this patch +This is the branch I worked so thanks +Complete sample demonstrating the has been added with description complete sample fo +agree +Marcus can you do the porting +I also disagree with Wont fix The dependencyManagement element is not for managing plugins or extensionThis is not the point hereIf +The problem was introduced in r with the check for a tables existence prior to retrieving its columns +Im leaving this issue open for now the downloads are still located under lucenenutch +HiThanks for taking the time of creating a bug report +Ok +Running run Failures Errors Skipped Time elapsed all API and Pattern issues +Thanks! +Moving to release beta +Usage in the PR and in the docs of course. +I know that you accepted the fix for this but does it actually fix the bug +Hi Robert thanks for taking a look +Utilizing that makes this a non issue for the WS externalService +The behavior seems correct to me +So whenever a column is inserted to the cassandra the cassandra needs to read before write to make sure the new column is valid +Hello VijayNot sure I understand you want me to have a default value in code just the same as Java default? So basically in the patch I should change public Integer rpclistenbacklog to public Integer rpclistenbacklog If that is it no problem Ill attach another patch +The test failure is unrelated +I would really love to have regression on the crawler unit tests +Verified at +metadata in eclipse and also deleted the project and it out +such as recent whitepapers +Im using and couldnt find javaxjcrBinary there +This is new to me +Yes I meant architecture when I said cpu +At least this only happens when you click the info button in the browse schema for a specific fieldBut in order to construct the histogram it needs to enumerate the terms anyway so this isnt going to change the perf and will give better results +Moreover with we will be able to know exactly when each violation have been created and track newly created violation +Sorry for the confusion +Ive updated to svn revision and it works now even with debug enabled +Duplicate ofPlease reopen if the solution is not correct +Some more javadoc Backport to x +Just to avoid any confusion in the naming +We patched our cluster and startup time didnt decrease much if at all +Close all resolved issues for older releases. +Not to say hasnt been undone but it wasnt this issue that did it as this was resolved before was even created. +Committed in r. +Patch to update the build to use Clover code coverageThis will require to be pointed at the Clover license available to committers +Please update +Added Javadoc toMoved Copyright on changed files to Moved randomize to +Please keep open +Updates of and R passes for finding average User pref and average Item must admit Im now a little uncertain +See for refactoring HRS issue. +Second version of the granted code with Apache copyright headers attached everywhere feasible +Development complete pending icons from Processes page created +Tested OK against old services. +Use new feature from to provide a private security realm for the default subject used to call the MEJB +esteban can you also explain why the configuration has a list of importers? what is the relevance of this pluggability? when would people need it? +This must be a quoting bug in the launch batch dont use Windows so that launch script was untested until now +Ive targeted this to the next release and raised the priorityThanks for catching this Kurt +Well take a look +After we changed way to generate temporary file we dont need generate a number of temporary file for the same wsdl or schema file +Otherwise the patch looks good to me +Bug has been marked as a duplicate of this bug +This bug is in the official release as well as the current builds +try this patch +Im planning to work on it over weekend and at the beginning of the next week so a week until completion is a safe bet +Is this because linear scan isnt considering the exception edges? In both normal and exceptional control flow l is never actually viewed having the value of null +Closing issues in +compile and invoke the testAny thoughts? Ron +However I was primarily focused on a POM that would generate artifacts to provide rather than make separate profiles for end users to generate their own +But still this looks something that should be handled by a schedulerScheduler should be least concerned about jobtrackers memory +Deferring to +Since the increment is relative to the prior token it should make no difference when there is no prior token +Defer all open issues to +Since the instance creation fails it is in where the is created +Please reopen with a new patch if you like. +Simon did it already and its on master in the Apache has had everything renamed. +Ill revert +mand whatever and then run ant i think? tbh i think we should totally mavenize the build when we have time +But soon youll have mvn sitedeploy mvn sitestage mvn siterun if each goal needs an Ant target your buildfile will really grow I really dont see a good use case for the its feasible if its available perhaps someone will find a usage for it +Maybe we can leave the existing code to handle UDDI v requests +Most of the time the default SMTP port is but its better to ask such question to the Sonar User mailing list. +Available in the next hour or let me know if this solves your problem +is a Tapestrys mechanism to avoid the response of a AJAX requestIt means although I can use some way to avoid the bug but some one others would trigger it +There is a RFE for Xalan bug bug was fixed in Xalan CVS on Octuber th +Of should and would Im only concerned with would +Reassigning to since will not be an official OGC release +Assume closed as resolved and released +Currently we sort the entire walog +All the skinning ecss files were adapted to use this methodId appreciate any feedback on the success of this fix at eliminating the reported error +svn folderssee the patches except for the Derby switch since there was no final relsolution on how we should handle that +This is useful in investigating incorrect metadata issues which is what I observe on one of our clusters but not sure at what point it started since the error code is in stuff +The suggested patch applicable to all the main branches trunk +I hope to see it merged this annotation and default value initialization is now fully supported +Confirmed that the spec fails without my change +move resolved issues to so we can generate release notes +Committed revision +Also it would put us in like with the developer experience from other appservers like which do not seem to have problems passivating such beans +Everything is fine Tim +Thanks Benson! +Verified this bug on the latest build from ACS branch +Thats how does it for example +I have attached jstacks taken during the run +I dont see the issue on trunkJust realize that +yes so it is using the projects default classpath +Committed to trunk + +bulk defer to defer to +Purposely breaking the build really isnt necessary to make a point about a known bug +I agree to the original request +But if you want to create a new index you shouldnt know if it exists or not you only have to said I want a new index in this path and I think the writer must do all the workWith this in mind I think that a posible solution can be that the indexwriter find out if the directory exists or create itPlease comments to my idea +But my client and I would like your second option +available inI close this issue after is committed +Please review +Right I agree solr should work with a encoded doc in this waybut the code is still wrong its not a correct way to convert characters in java +Committed. +Ok Ive added the src and pom as a tar +See for changes in Alpha +Closed after release. + +I did some more testing and we are coming in close enough on Memcache sizes +Patch to make catch exceptions on parsing dates and return null instead of dying +Still if other ideas come up it would be great +James after I had reported it on this issue I documented it here Werner is it possible the problem is cropping up again +Adds an option to to specify the number of keys to slice at a time and a rangeslice operation to perform range slices +If so the only resolution to the problem is to load from a shorter path +This should be the same prob as I dont think this is the same as log aThis could be the same as but Im not sure +No JMX mappings in this patch if recommended we can still initialize it to default synchronization check to Hashmap is not synchronized +updated the issue to reflect the problem is wider than just patch for patch for +No longer having a Manage tab +the widthheight limits are applied last. +The issue is occurring because the avoid list for deploy is not getting updated +svn revision Are we sure that the check really is fasterAnd is it worth trading the additional memory usage for the speed upHow much difference does it make overall to reading a fileI would have thought any increase in speed is likely to be dwarfed by the other costs +Bulk close of old resolved issues. +So instead of a failure you would get an error in the failing caseIts not obvious to me that its worth reworking this to change an error to a failure +Welcome comments +Now when creating a container and if there is an ivysettings at the root of the project it get automatically configured +Is someone working on it? Otherwise I would like to far as I know no one is currently working on this +I think lack of unit test is not a gating item for this JIRAWhy not? Lack of unit tests is what caused the bug in the first place +but as Mark points out its really not that helpful at allSince Micahel doesnt even need it anymore it seems we prob shouldnt even keep it as configurable +An example client that shows the problemis available but it is not a unit test +I think a simple Object is enough +up +Others have also mentioned problems with it +If not Ill synchronized reads also +see previous comment for where to save this great with one minor issue is still loaded from ill look into the issue due to several bugs regarding the check for the in check is too weak because Wicket already exists if is loaded first +Once you promoted a slab in old gen it stays there instead of being GCd and replaced with a slab in new gen againThe bookkeeping might be worth it yes +Resolved a long time ago under auspices of a different JIRA +Ted when you added the XSLT bits you it XLSTIm posting a patch to your new file but it may be easier for you to just do aquick SR. +This fails on two distinct environments and passes on one environment I have access toFAILS +Rollback and resolve as wont fix +Goto httpslocalhostcatalogcontrolproductId Update product price one or more times +This patch represents what I will check in +Did you see some obvious errors in the log +Patch seems to work +itself has no such references to JPA classes and it is only the JDO TCK that does due to the RI. +net and maven +Nice to haveMost of this patch has been applied already or superceededOutstanding is an attempt at a compaction simulator that could be used how well a compaction algorithm worked given different sized files + +How do you produce this error? The wsrp admin doesnt exactly work right now with have to activate the fix for +Support for primitives committed in r +Yeah we love I will try to get it complete this weekend +We should also rename the getset methods to not refer to List etcFor +exe on the end of the command or notI dont understand how removing lines of code that doesnt need to be there and causes problems doesnt simplify command line parsingPlease just remove the check it is unnccessary and incorrectthanksMichael +closing self reported issues resolved on versions already released +Ive played with RI and found many differences with Harmony in keymouse modifiers processing +We dont have the capabilities to automatically append things to the abstract but well look in to it +meaning I can reproduce the issue +Can someone please explain? Thanks +heres some fun lsof now tried this a few times +The sonar service is run by the Administrator and the C directory has full access to all users +Were already starting to upgrade and build the components for this CP would it be possible to tag this by the end of this week? Or do we need to get the fixes into this tag as well +The latest PR includes a proper solution for the problem. +it can make the JVM less efficient at garbage collectionhttpdeveloperworksjavalibrary +has been the domain for years and many people are still trying to go there +Symptoms were as previously reported and also included spuriousThe attached patch fully fixes the issue for me please consider for inclusion in the next bugfix releaseSteve Glendinning +Mark Im pretty sure that we have numerous features that make us +Got a program or reduced test case we can use to reproduce this +Attaching a file that has my review comments for the documentation of the XML operators +BULK EDIT These issues are open to be picked up +Mazz will fix this for final because we need it +Hi EnriqueThanks for the feedback +Committed to to branch issue Oozie is released +I will look into it +as Niall states it works fine for him +I do indeed see the bug using nightly builds of versionThe JSF components render correctly while none of the Trinidad components renders +Sorry certainly I intended this for inclusion +The UI changes have been made largely and merged into trunk +So its the minutes are now visible to the world +This wont go into as such +Issues resolved in +fixed in the common codebase according to the suggested corrected the netbeans specific codebase +If it compiles on your local machine its probably due to having a newer version of hadoop in your local repository +Downgrading to Maven seems to have solved those problems for me +The observed problem was due to using aIt remains a strange fact that this is only observed when the end point is an. +Issues resolved in +should we default the verbose checbox for analysis to on? I could be in the minority here am I the only one who clicks verbose every time when using this +Thanks Carl! +A getListing RPC returns the number of remaining entries to be listed instead of a flag to indicate if there are more to be listed +We could switch to blueprint and simply add the intentmap to the context that starts the dsw bundle +bq +Yes Yes No the policy file is accessed directly from the jar file as a we need a templates directory? Will there be anything but this one policy file in it? Why not keep it in demo +Ive applied the patch Ill be putting up a new snapshot for people to get some exposure on it before I close out the issue. +Uploading a patch that makes implement and makes the config a instead of a boolean +A message explaining this would be more appropriate +But now I cant even get the Sonar way built in code quality to run on my project as checkstyle is now giving me a lot of exceptionsI think I will just revert my installation to the base Sonar release and not update anything as that worked fine actually +MarekIm reassigning this jira to you as it turns out that it is that requires s Emmanuels answer to my emailSeam requires Hibernate Annotations can work hand in hand with Hibernate Validator but does not require Sep at Permaine Cheung wrote Hi Gail Emmanuel I have a question about the hibernate components dependencies +It would be useful to also have a look at your individual plugins but only do so if you want to investigate furtherMy guess is that there might really be a bug in here somewhere +I think is the right one too. +Changes have been made that should address this problem but as it is intermittent it is impossible to verify for sure +This will be resolved by the linked issues. +Worked +Unit test passed +I am running regression tests against Derby jars built by this jdk in order to verify that the resulting jar files run correctly on Java I would appreciate it if other developers could this patch on Windows Solaris and Linux +Making the dependency explicit on does not helpSpecifying it directly on the does help +Any more feedback +I still see the original URL in using +However that is really moot since we have decided to no longer release versions of the official JAR files +To resolve the ambiguous productions i had to remove all type checking and do a lot of changes to the BNF +Me too +Where is the chronos pluging searching for the file? In the target folder? Or in the jmeterlibjunit directory? And where I place the dependencies? I see that the chronos plugin copy all the dependencies into the jmeterlibjunit folder +Thanks for the patch! +Yes it was happening from the time the test was added +Liit looks into it next week +Also if possible it would be really nice if the entire soap message could be logged when there is an error unmarshalling it +I built the docs and verified how it looks too +See the comments on the TIKA issue +Should that be closed +If the region is offline based on ZK the region is about to +Fixed in trunkASM packages are now private to iPOJO Core was released. +I just committed this thanks Cheng. +Hi RickThanks for the replies Ill add the STMTIDYeah you got it right! I think we should come to a conclusion about the API of the tool and it may be handle in a different jira linked to this what do you thinkThanks +Size of the required libs are kB +Evaluating expressions is the core feature in XSLT result as it is in Freemarker etc +Duplicate of +Rewrite in supports the user choosing not to do this easily +The sample Folder is missing there while it is present in the +Directly sending the data to biaNext to work on is hermes eosTony +We prefer to put jiras into Resolved state +Thanks for looking at the patch KnutIll look at your suggestions and test them +So it sounds like we just need to remove the c option +Is this suitable for backport to +Reviewing the code +Alexey I implemented loading static libraries +Backported this to branch in r +Thanks Lars +agreed +Updated trunk have committed this to and above +I have not tested it with jenkins nor am I able to unless I get more permissions to edit the jenkins setup +Looks good to me +I do not own the Copyright and hence cannot grant an Apache License however these files downloaded from the site already have Apache License Headers and Copyright noticesSince this material is available from an Oracle Sun Microsystems website and the author was employed by Sun Microsystems at the time we can infer that the License Headers and all files are published by Oracle Sun Microsystems and are genuineIm uploading the files from the site as it is easier to tell these havent been tampered with and contain all relevant license and copyright information +will the stuff as a diff issue. +JVM bug could be possible since Andrzej specified JRE as the execution environmentAndrzej Could you please upgrade to the latest JRE Update and check if the problem persists +I just committed this +I will investigate it +Assign to Charles to remove the link +defer to +Craig this is in ALPHA +I think that I just have to make the layer not adapt to +True that I have made a change to replication trigger to SET in the test so that I can test scenarios thus I spotted this issue +Patch was applied as. +This method has been removed in branch am also Facing the Problem using Httpclient for uploading the file in IPV scenario +I thought this happened based on our release branches changin contents of VERSION file +The class names could then be substantially shortened tooI will submit a patch for that moves outer classes to static inners +Could you attach it? Once its attached the Jenkins bot will notice it and can comment on it and others can review it as wellUsually we iterate on the trunk patch before worrying about other branches because normally changes go through trunk and later versions before earlier versions +Regression in r +My point was mainly that we havent been doing these checks up to this point so I didnt think we should be blocked on it +If the release in immanent we can consider for CR +Is it an option for you to get the latest sources from SVN and build the yourself? If yes you could in addition insert a new LOG statement at line of that to see what it is that is causing the? If not I could ship a new snapshot release of even though I do not really like the idea of adding debug statements on purpose to a release +Would be nice to have it up to date for rc +Makes the killed boolean volatile since it is updated by various threads without locking +Other consideration is that the classes are being picked up in another jar +Backported in r r r +Thanks +Resolving the issue +Fixes both this and +and trunk +Changes to parse the signature in the context of the parameter types of the function or procedureEnsures declared signature Java types are mappable to the parameter types as specified in the SQL spec part Checks for various invalid formats as wellIll commit this in a couple of days addressing any review comments +Ive attached our and a screenshot of a jmx console to a similar broker that is running since dayswe do not have any warnings or errors in the logif we reached the temp store or memory limit on the pending message cursor would we get a warning log entry +SamarthErikHow does this patch look +Attaching patch aIt pulls out the connection logic and also fixes a number of spelling errorsPatch ready for review +See my comment on be more clear I cant see in what circumstance would someone use this method +Seems to work fine in ER +Bingo +Contrariwise dont use last last but last update +But as this JIRA is now committed in so this addendum is neededI think we can ensure such things doesnt happen for otherGood on you Lars +Ok please commit +Have a look at the template in there and hav fun +If the fields are itwill emit any Tuples that are in on of the but not the other +Currently testing the solution +Closed after release. +Completed +bq +The name mapper could avoid all namespace lookups as long as no namespace remappings are in effectI can look at that optimization but the calls might also be caused by calls that use expanded namesThomas do you have a few call stacks +See for +It is related indirectly because the path resolution code is shared +Verified. +doesnt support the default package +Yes the patch should be reviewed before it is applied +The trunk test passes successfully for me +so am i missing something easier +Works fine now +txt or +Verified for at + in branch and resolve to both versions please +Thanks to Paul Rivera for provide this patch +it shouldnt be possible to crash the compiler regardless of code validity +Committed thanks +Attached stack trace for during upload image via richupload on to be specific problemIt doesnt appear on Mojarra only need one jira open to cover myfaces not working with fileupload +A patch has been attached to this bug please indicate if any problem with it +Expected Behavior Upgrade should succeed +This patch will add functionality to configure the port range for Hadoop Services in HOD hodrcComments +Updated patchRemoved static from fc initializationRemoved invalid comments +This issue has been resolved for over a year with no further movement +I am not able to attach a sample project +done on i need access to +Bumping urgency +Changing it to DEBUG seems fine to me +This limitationbug is very dissapointing +Didnt expected it up there +PatchIt has one nocommit which we can remove once is in +Im going to mark this as a wontfix please reopen if you think theres more to do David but it sounds like things are good. +This was fixed in series of the tools. +Improved existing groovy file for fetching fax billing and shipping data +Ben Im just saying that yesterday out of a whim I tried the same modification and single type DFT were missing the schema imports +Seems to be code cleanup for moving forward so marking backport reject +Lundberg notice that the apparently some documentation looks like deplicatedHowever it is necessary because the section Make The Jar Executable called which explains how to specify a is targeted just to jar package type +Heiko can you elaborate a bit on what the issue is and how we should be solving it? thx a look at the its an equivalent of the Axis +batch update of issues with fix instead of the correct +It seems like the only call to is now inWould it make sense to move the to method from toand make it private there +Submitting +the described problem is the expected behaviour. +By digging into the source code I am hoping one of us three could answer whether it is properly implemented +Thus a test case that first sets a binary property and then reads it fails unless the preserves the stream contents +Fixed rev +Committed revision . +Can cmake currently build on Windows and Mac? If so I think this would break that since the Snappy and libz libraries wouldnt be +Found the problem Ill commit a fix later today +Its hard to experiment with obviously but so far the experience has been unsatisfactory +Thanks for the patch Venkat. +But because of the backreflastline change we cant accurately detect when we need that scope +I cant replicate the error with latest code +I think we will keep seeing users asking for more characters +to trunk +Are you working off of present project trunk? +I can confirm it has been filed +Are you sure you tested against revision or greater? The code you pasted in is not or +Thanks for posting it +orgapacheant +Rename srcsitexdocstylesheets to srcsiteresourcescss +It will take a lot of traffic to make it break a sweat +for patch +It is not known whether or not the fix addresses the problem reported by VarmaDoes that accurately describe the current state of the bug report +Tx +duccling is updated and verified correct. +Moving to TBD +Remove FIXME comments from class and open JIRA to track the issue +And it has improved a lot +is committed +All my schema files +Have you been able to repro this on another set of machines other than the ones you first observed it onAlso could you try this with the latest trunk? Some improvements were made around the FLE which may shed some more light + patch +No longer commented out and runs ok now so I assume that this is fixed +So resolving it +It may be resolved laterClosing this issue as fixed. +It does not add any value for local SLSB invokers +I didnt get your concern locking +All good +Then assign to me without resolving but with changes can be harmfull for this three components +nope makes no difference +I dont have any idea how we can resolve this problem +Unit test for serializable bean +Youre saying you want a load function that parsersabcd as a bc dand not as a b c dis that correct? Does it have to be done with quotes +Not for branchSorry +Note that we are expecting id of JIRA priority +Well address the issue between CR and GA since it is a the current with the suggested asking Sande and Pete for ones arent really that great eitherIll dig up a list +John I was wondering the same thing +This patch corrects some issues with small cachesto provide more robust performance +Any chance you could add a comment indicating which JIRA it should have been associated withIve looked at the jar for the release and current maven output for and the only thing it includes that isnt part of our JNDI implementation is the package which just contains interfaces no implementationFrom what you say I suspect the problem might be that you have two implementations of the JNDI specification in your runtime not that our implementation includes the interfaces +Yes in this context it is completely safe but Im all for consistency here so Ive made the change +Has there been any change in the situation? I notice that the AXIOM bug is as yet unassigned is that a sign the issue will not be addressedIs there possibly a work aroundUsing SOAP is not an option for methanks in advance +I also definitely like your idea regarding the set command +looks like a straightforward +While its pretty easy to do but shipping docs as standalone artifact makes us more service neutral +Committed to trunk +yes it remembers and the next pages keeps using whatever first value I put in as existing also have to change the Browse to Browse +New Quickstart added to merged +It was changed to enhancement because Spring was not released at the time +Hey Lars I did svn blame and svn log against that file +Branching from the WCS sounds a bit risky to me especially since this stuff doesnt really affect rasters +Hi Alex How do I reproduce? For starters can you post the entire request +Still waiting for a response from the original reporterAnd we should push this to before +Reattaching for automated tests to pick done +Thanks Mike! +Relies on multPAP method extracted inProposed patchGives to on Harpertown. +The only overhead is an extra pass over the list of containers +Im working on it +Hi Olivier due to which is going to be fixed in the next version of Sonar Eclipse I think this ticket is no more valuable so Im closing it. +I can work on this and port from tag to the trunk of course just want to know if it was not done or was reverted because of some problems discovered later and not mentioned you +Hmm commented that line cannot see the exception during undo after moving away from the query +committed on branch +But it wasnt included because we were not sure we wanted to support this kind of deploymentSo it needs some discussion and agreement +Ops thanks +But the above testcase failed bafore and after this commit +Well see how it goes. +All changes have been applied to and docs will reassign for verification when they are available on +However the value of attribute remains some valueIf you now restart the server the value for attribute correctly shows +This can be implemented via a custom translator as well +Terrible +Uploaded a new patch because short names should be compared +in the output collectors? Could we pass a single Configuration instance through to all of these dependents +For its straightforward to introduce a +I looked at the UDAF evaluator stuff and the UDF changes +Other than that the patch looks good +If you have suggestions to improve it then I can work on that +Includes the tool and the Outlink tool +NOTE my environment has changes since I reported this bug +Merged back into along with r to delete the test file +This allows you to add new brokers to an existing cluster move some partition off of the brokers to be decommissioned and then shutdown the broker by killing it +Note that EAP does not have this problem since the of the distributed session manager in does not match the pattern queries +lets fix that correctly +This component have relative positioned elements. +Verified fix in and trunk +Administration portlet has cache but you get it always after login +Hi Kris +Patch looks good to me +revised hasResolution method to check both X and Y which are both set in setResolution and their own individual methods +Testcase attached +Stefano if using pax runner helps us get closer to having a testing framework then I am all for it +One last thing there has been no code contribution from the therefore Im guessing it was never pursued +target managed component serviceString pid nevertheless gets its configuration and gets activated fine despite the exception +elements that were referencedin expressions in the stylesheet were given the stylesheets default namespaceinstead of the null namespace. +ks to test this with as well. +The are what we need to ensure that the refactoring doesnt overlook existing m hoping that your motivation now to get this in will get an IT test so I dont have to make a subtask that says add test for +keeps a reader open for each store file and thus we end up with a lot of direct memory usedI was actually curious about mb as the default size it seems even as little kb should be OK. +Oracle. +Greg could you add Jira entry for implementing bundling in ODE? Then please add relates to connection from this issue +Instead it looks like there are files already in the srcit folders alreadysrcitresouces srcitresources +Your intuition is correct at a high level of granularity Ive not seen signficant performance differences between an unsynchronized and a +The final status says failure or timeout but there is no direct way to find out which one of the test timeouted. +resolved at revision woudl be nice to be able to keep bz as an option it produces substantially smaller files for those with limited bandwith particulary with source releasesAlso bz would take longer by hour seconds is seriously out of scale to what it should be +When using the trick of setting an id first and then setting you can be sure that the id will be set first letting you look up the person and then set the name +I had this situation when Eclipse was executed with JDK and projects used Ive seen this also with pure JDK environment +Ill take a closer look but I cant get to it right away +logcat theyre loading but its REALLY SLOW! Like slow! I have no idea why SSL is so slow but if your connection sucks I can see why this looks like this wont load +Is the issue here that you didnt import the tables on the other end of the association +This was not a bug but an intentional design decision to discourage the throwing of a without a message. +Im wondering whether Spring works for you here? Have you had a chance to test final in your scenario alreadyJuergen +It appears the server is throwing an error that does not get logged but I can see the stack trace from the server in the network traffic +Attached patch addressed Teds review comments changed the name mismatch found by Stack +Was committed in your tree at the time of your testsThis may have been the problem +Also removed the changes in test test method +So for that we would probably be looking to use custom Endpoint annotation on an endpoint SEI or Impl and then a minimal amount of XML to configure the classpath locations to look for endpoints just like you do for spring Auto Wiring +required resolution for as a precondition +What about or +The metadata file in Maven central now again contains all released versionsThanks for the report. +closing as clean up +Ive verified mvn clean install +batch transition resolvedfixed bugs to closedfixed +investigated no update should be done to guides according to this task +This patch resolves the warnings +As it was the on Mono would say that +Next question do we trash the old ALL plugins zips or keep em around for another milestone +Argh I forgot that and are incompatibleThis proves that we need to register wrapper components for Spring datasources +I merged the patch to. +ve committed this +I added that while trying to fix a contrib dependency on lucene +At revision I made admin sessions eagerly refresh on every access +Your proposed API impact sounds very reasonable and ideal +Fixed based on +If possible can you verify it fixes the problem for you +Daniel you can apply the patch and give it a try +The HTML files are under the +That could help thanks +Maybe as files are added this will become required +I see but I dont know the intention of the author of +Patch applied +Is there any trace I can takeAnyway Ill try to reproduce the exact circumstances in which the error occurs +Hi SnjezanaI found out that actually the scenario is a bit different but I think its still a bugTry this +Whilst the set might be initially I think this will be growing +It will probably make it into but needs some more testing +If you have please go aheadOne possible explanation is that a connection goes stale after being checked for consistency but before the big lock of the connection pool is obtained +Ted Andrew +We will have two versions in CP +I will commit shortly to trunk port to branch +The current code seems to have the desired constants class. +The thread catches all types of exceptions logs them sleep for seconds and then continue from the beginning +I am also attaching a patch diff file containing changes to existing filesThese combined patch changes modify the Web Service binding to add code to implement callbacks correctly +My current thinking is that the fix to this vulnerability involves performing authentication BEFORE the database is restored possibly followed by a second round of authentication AFTER restoration finishes +Welcome Thanks for fixing it! +fixed +I will investigate this +sorry for the is another more convincing attempt to force help our instance to survive the GiraphHadoopGiraph transition intact without altering the current interactions between the frameworks +Moving this out to please move back to if this will be worked on soon +I will commit soon +Separate compilation targetCompilation targets separate intoinput toolsdata collectiondata processinghiccFor input tools data collection and hicc compile withFor data processing target use +Attached patch makes it optional in falling back to if the required parameters are not given to the job configurationWhen merging to I will take out entirely +Si I think this issue should be closed +presume this is required in and up too +Please only use the Apologies for my error +also add dependencytrue for so that OBR resolver could kick in applied thanks. +Is it true that the perf cost is avoidable? or are their implementation details which can be optimized? We are working to make to ready for trunk +Hoping that Jason can have a look +Maybe BND is unable to recognize that pre java class loading logic and assumes everyone uses LDC instructions for this now +Updatedrebased patches include the dangling semicolon fix +So its like days of no emails followed by one email with the last events assuming an ERROR occurs on the th day +Thanks a lot for the contribution Arpit. +JacquesOh I havent the same result +I would rather avoid it and go for additional memory given newer machines are coming with more memory than make the code unclean +As soon as Nexus will be configured to release Struts +How does this change backwards compatibility? The index format is certainly the same +This is because I want to be in control of my own key and if Solr fails as soon as I dont provide one thats going to help me find the bug in my indexing code right away whereas silent success will allow that bug to fester +I do plan to write the test for it but I can continue with it only after next week since I have exams till ll add the test case for it after thatUntil then I too recommend committing the patch +Add a seqID in logwhen chukwa restart this will help chukwa to locate to the right fileFix the bug of cant collect the log when its first tracking files size is +EAP DR can be created without a problem using the new server wizard +Patch to to drive redirects not automatically handled see my previous comment +Could you please explain what you mean +Thanks for patch v +I think that means I understand the issue +When you say search the classpath are you talking about get or are you actually going to try to parse the classpath property? Im thinking its the former but I just wanted to check +Basically I want to take care of housekeeping by myself if I can not to bother the infra team +That was the fix we were looking at +Closed with the release of. +Ex +Thanks for the patch +I was hoping someone else can see it differently or have another solution +Opened + Fix ivy test conf to include dependent jar files Use init target to initialize test jar file download rather than individual test type +The call from is does not have a top level call through a at any point and this will violate the use condition for the entering loadClass +This is the issue implementation patchBesides implementation of aggregation several bugs were fixed in it Fixed irregular failure of cunit testnativesuspend test +Hi Eamonn thanks for the patchTheres one thing which Id like to think more about +This has been addressed in the recent reworking of JDR. +Good point the TT shouldnt if offerService returns DENIED both for this new use of DENIED and the current one +I havent had any other moderate emails that needed approving yet after this one thoughMight as well close it as invalid +patch attached that enables tracing in the InvokerSender threads +vote in support of this ticket +Thanks. +that doesnt happen until raise is actually called anyway. +I do not want to introduce an incompatible change for nowI dont understand the compatibility argument since this is new APIWe could add a parameter to indicate whether or not continue if a change in the has occurred +What are the current thoughts on this matter? Ive been using Buildr with the modification for a while now without any +That was done to make code simpler +Ive checked all the components and updated them +Its still trying to do the xinclude its just looking in cwd instead of my conf dir +This issue couldshould really be split into a bunch of them one for each set of files going inThat is how that tracking and management should be done +Thanks for the clarification John! Ill go ahead and commit this then +It is important to include this JIRA for the release +I believe this was fixed along with. +Ill take a look at it now to see if tree map is just the way to go +jar wc l your comment from I assume you meant Issue +Just leave a comment with what GET param you want to use and Ill hook it up in the php version might be good to add callbackxxxxx param for JSONP. +The styles location has changed for the newer versions +As Jakob pointed out many projects including use that +Thanks. +The propertyDetails are null in this caseI have tried to fix this issue patching and which seams to work +Volunteers arewelcome +Ive just tried with the Python Memcached client with threads and showed no problems whatsoever +Nothing to be done here closing. +Thanks for your reply +Maybe that information helps you with the development because my server is case sensitiveGreetings and are right Harry +Now create a VPC +Ronald found this issue while cleaning up still think this is a good addition +does a refactor to add a utility class you can use for thisI think that can be done in another jira +Anyway its still there its still happening +Basically lazy seek helps to reduce the numbers of hFiles to be seeked right +in revision . +ok looks better now +Yes you are is resolved +qmail parts doneMailing Lists next +Just a question since only the short names are being compared does this mean that same user in two different realms are considered to be equivalent +Committed revision in sandbox. +The plugin is now packaged with Grails which doesnt exhibit this problem. +Fixed on x and ll open a new issue to remove acquire +This is now consistent with the behavior of the commandsThis changes implies that current users of Stomp must now check that the properties all exist before attempting to access them or handle the. +Other than the JIRA merge are there any other factors holding up this upgrade +Slava AFAIK you changed this code recently fixing some bug +hmm seems to not be included for some to bring that back in +It was user error on my part +Descoping from not a blocker +Make sure that the +but what does tomcat actually change in the deployment? what should be changed in jbossthe project +acknowledgeMode is probably listed as a required config property for the activationspec in questionPlease attach and TRACE log from applied to is this still an issue +posting my question via users mailing list +New patch addresses Jitendras comments +This forces the build to exit early if the compiler level is not at Java or higher +Sending srcmainjavaorgapacheservicemixeipSending srcmainjavaorgapacheservicemixeippatternsSending srcmainjavaorgapacheservicemixeippatternsSending srcmainjavaorgapacheservicemixeippatternsSending srcmainjavaorgapacheservicemixeipsupportSending srcmainjavaorgapacheservicemixeipsupportSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestresourcesorgapacheservicemixeipTransmitting file dataCommitted revision . +Bulk close for Solr +MarkI have applied the patch after a review but not actual test +value +Also fixing one report +Ill test this today +Sscoping items for early release of Java Eclipse Mgt Console +okey but why not just create a dummy jar in a small project for this test instead of changing standard jars and still have the overhead of a large project? +Committed revision . +Identical patch +Will close it after it has been merged. +Agree +core machines +Ah saw your spec right after I hit send +I agree that there isnt any good reason to have some fully duplicated java source files in a project +My problem is just in Oneiric +Committed to trunk +check that the ova youve downloaded is not corrupt from the in that folderIm able to start the vms after importing them to Fusion but I dont have any real infra to test againstChandan can you Share the host version Try with the template which has openvmtools? Try to manually deploy the ova on just to verify if it gets imported and startsIf it fails pl +The tricky part is who stores the hint? Since its a doing the streaming +Thanks Alexey +Running tests +Will do that today +done in a series of commits today +Will commit tomorrow except if I have a no goAs a side note I ran locally the tests with nc l nc l nc l to take some commonly used ports it went ok +But I think we can do nothing to binding WTP Faces Config editor +Tell me where you live and I will pray in that general direction am not sure when we will incorporate those changes but we will +The problem still exists if the server component of the web service doesnt access the +Patch looks good Ian +As per the logic in HRS it will return as the processed Puts count +I will upload a patch soon and we can get this in +Good Luck to explain this in the doc. +Fixed after revision +Carl this is the only blocker that doesnt have patch available +Assigning back to me since Im almost done cleaning this up +Completion of the passing of all Hg SCM provider tests relies on a fix to + you are at it could you fix the spelling of the test case name +Moving these to please move any bugs back that will be worked on before the November release of. +thanks Jian! +Added caching for roles look at it thanks for the report and patch! +The matrix and user vector should have nonnegative values so I cant see how it gets a negative element from multiplications and additions +Duplicate it appears of not being able to have the command line values set override anything else +The javac warnings are due to deprecations in the new version of avro + +Thanks Corinne +ExampleThread A writes bytesWrites B writes bytesThread A calls hflush and gets interrupted throwing and throwing away the packet including Bs dataNow B has never seen an exception but its data was also lostOOME seems very unlikely wed need to have many thousands of threads all hitting this very rare condition at the same time and the buffer sizes were talking about are fairly small +The current owlapi version is +The route method is only intended for use by extension code and not for sending events into the engine the sendEvent method should be used by line to trivial well look into if the behavior can be to Major based on newer +Thanks Brett although I found rolling back to version of the plugin was easier +Fair enough +Verified Fixed +Attach the initial patch +I just accepted it today but it had been pending for at least or months as I dont log in that often into xircles +I run the tests under the Valgrind no double free was found +introduced the problem mentioned in +Im not keen on just using a regex to determine what goes in here +Thanks Navis +OK final patch this changes stuff to geofiltgeodist +I added your test to ensure everything is working +was this one of the plugins for mojo? If you could check it in Ill take a look and see if we are missing any other features rather than giving classloader construction tools +The problem seems caused by +thanks Scott will look at a migration per your suggestionfirst stage for refactor is identify all the code that needs to modified and make a task for each +Im not particular interested in rss but if nobody disagree we can add it as an extension but +I would like to see you generate the map with the FIELDNAME constants instead of the ids directly though +Can you tell me where it is there for +Thanks Todd for your comments +Have just tried this with latest svn head and the first app now deploys without issue and I can addeditdelete just fine +CommittedThanks for the review +i also had as the default mail server + +I mean as a user Id expect it to be true by defaultbq +I just tweaked the errant classpath entries so that shouldnt be a problem +in the end i would like to point to the webdavssl support feature Same here +I am still having trouble w the offsetshas anyone else tried the patch +Even in a local build of the book this text is still incorrect +Closing issue as it has been released as part of release +Added at revision +I am closing this problems seem to go away Mladen I am assuming in the next build I will not see these messages coming out in stdout anymoreThe parameter is for failed +Yep I have a patch for that will upload shortly just writing the tests +Please look in history for what the actual resolution was. +Revision +When writing the adapter from scratch I did not compare with other adapters and used my own constants +I and Sambavi will create the corresponding Jiras as well +Bringing into to test. +So I did think this was a bit of a hackThe reason its there is that without it the inclusion did recurse forever +All done closing again. +Bulk moved to this functionality for firebug lite will be rather difficultImplemented for weinre. +Patch includes a fix for this issue as well as an extended test case checking for this issue +Hmm is current release +Perhaps Chukwa was not running +were moving away from dojo integration in struts so you wont likely get much traction on this issue +You also need to update the +Committed revision +This does NOT happen on but DOES happen on trunk +Not for metrics system warning +Trying to synchronize here gave deadlock insteadOne thread owns allocCache and is wait for latch +Reopened and temporary disabled the test +Shouldnt we fix it +But compatibility is a good reason to make this small fix asap. +There is a page property built into the validation that can be used for validations. +This is at least what I planned to do +measures performance of the cluster for reads and writes +not sure if this is possible or not will have to look into it +Sounds like a duplicate of which was fixed could you try a nightly build +It looks like report is no longer called +Ive updated the subject to reflect the point of the issue +failed in build passed +I dont think it is a conflictSince the is overlayed i dont know if it is starvedstuck but for sure the plugin is starved +Patch needs regressions found +Unwanted jars have been removed +you should use the Camel jira for this issue with the camel file Camel +Also can you please respond to the point Prasad raised about changing the default from PLAIN to NONE? Thanks +Is it possible to create a tag including the linked issue to be incorporated +That is if the client uses CL and the service impl uses CL the binding should not rely on the to switch the data into the correct CLSo I take back stating that as a problemIll see if I can rename the JIRA to focus on the need to copy faults +And even there not even Commons DBCPs implements Referenceable +this is a patch generated with git diff on my own issue branchI think you have to manually rename the test folder as mentioned in the issue description +Thanks for the patch Ashish! +So can we have admin notified about what needs to be done to have these update successfully +Its probably fine but Id prefer to integrate with a real HTTP server if I were going to use this in production +While we are doing this should we also add the ability to specify a priority and a for the eprs outgoing messagesIt not specified they could default to made me aware that Persistent TTL and Priority are aspects and should not go into the Ive been poking around and the found that the is created with an instance of a +Which is why I didnt reopen the issueI put the marker here more to ensure we had a proper connection between issues for documentation purposes. +There needs to be a way to import and export data from clobsblobs +On multiple runs I pounded my broker with messages and I got all back +Attached reproducer +Lets all forget this ever happened. +Anon types now appear under the proper method declarations +All Done Enjoy +Finally update the jira with the jira with the patch check commentsEnhancementsWe can also add findbugs rats for release audit at later point to enhance this patch test automation process +I have found that the bridge methodology of caching the and releasing the at the end of the action and then getting the render to use the cached in a new will not work consistently in JSF +Closing jBPM issues resolved in a released version and reported by me. +testing or +I will work with him on this +Please disregard +handles the null +was updated +shouldnt a configuration object be passed to the +I am also willing to help out with Maven support +am I wrong +Related issues and is lacking any unit tests which would prevent the fixes from being protected into the future plus serve as a demonstration of the possible problem +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira +Dustine I meant what directories did you use for Namenode? the directories where namenode puts it namespace +FYI the tests you cite were run with Jaxen +fineCheersJarek +In some cases the functionality in question is still under development for +Sounds reasonable +That affects only command rightTo connect to a specific controller the following should work +This will catch possible future issues with the underlying data structures +The proper method to run forge is to runmvn execjava from the dist module. +The definition requires that a class that implements this method must also provide a public constructor +But that only really matters if someone is doing that and I dont think they areIm going to commit this on the basis that it leaves the spec file in a better state than it is now but Ill leave the ticket open in case there is any discussion that needs to be had over the way versioning is handled +I just committed this. +It appears that a race condition is triggerring between the HDFS sink syncing to hdfs and the shutdown of data node in the test +CLONE of the issue with is still no pom on ibiblio for version +The thanks. +I think we may have to go with the dynamic sql to find the pk the problem is an upgrade from a schema where the pk name was not provided when the constraint was created +stackI think we could introduce initializing state first for +move it to next think we can close this but lets get some confirmation +In the catch block of get of Sqlca changed the code to construct directly without first creating So if there is a failure in fetching the message the resulting exception will not be localized but at the same time it prevents it from going into recursive call +The patch looks good +fixed in rev +Also fails doing TTI +This causes a memory leak +So that is another reason to store the timestamp in the ISO string as you suggested +Created to capture the need for cache policy assignment to work through the configuration graph as well as through the programmatic API +or the bridge +Would the block has two checksum types +Done +Removed from code repository. +defer all issues to +LiborCorrect Im referring to the mouse clipboard +Thanks so much Mike +Users cant include such file in their jar archives to distributeEven though the config file is not the same common format will help users as well as documenting +What I would expect the filter to do is limit the entries being logged +fixed with wtp version from that what will become WTP or WTP +thanks +The change in schema is distinct improvement Id say +Sorry for the delayI mean columns column families are in table schema the ability to match column qualifiers by regex may be more useful in reality +Im currently working on writing a proposal for addressing this issue +We doesnt have to injecta proxy for the Selectwe can rather injectthe Select with a proxy to the +This patch fixes a number of the issues and a run of mvn site package Pdist true with this applied took ms versus ms without it +Thanks TomIn a couple days this patch will be committed to branch +Joes committed mailing list config change that should fix this issuesvn commit r in infrastructureapmailtrunkbin that now works +Missed file extension +If someone can commit it it would be great +There is no such problem in current trunk +If anyone can help me on Java test I will be appreciated +Thanks jdeolive +checking it out + has a deployer that add data to the context params of +Putting this in +Added Aggregation table +As for this is wat I haveunder pluginsorgorgschema like it is there +you for adviceI tried your +Thank you Nicholas +Excellent works great Ted thanks! Patch applied in trunk r +I have closed for now. +Would switching back to a release version be recommended +Im trying to remember why we didnt do this sooner +We cant change the released jars since they are well released But as of the s contain the statement to alter the version in the database rather than relying on the to do this. +But because most of time the set size is we could set the intial array size to be The latest patch changes Datanode container associated with each block to instead of a +I just noticed that you had a comment here now Dhruba +Brandon already has posted a patch on +I set up the build requires condition for ant in brew build script +Could you provide a mockup to show how the feature looks like in UIBTW UI team could give some proposals for this too +Add patch file where also checks if the boundary pattern does not contain leading rn and take this into account when rn is in front of the located boundery +Basically put Im dubious about class with so many constructorsThe essential problem is that it is too tied down to the details of where the properties file livesI will look into an improved solution that divides the code into two parts a factory service that can provide an implementation for any Resource and an implementation based on reading from a ResourceThe factory may have additional methods for common cases such as a file on the classpath +On the other hand using a logger requires using a classpath when running java jar would be new and solution is to try to load the logger class using reflection + The quick fix works and removes the warning +I have noticed through the TCP monitor that some messages sent by simple axis server are incomplete +Some issues that have been cropping up include the anywhere direct field cache access is performed +Did you see this only once? If you have any success reproducing the situation please let us know so we can revisit this +You could just implement another framed transport that discards buffers instead of reuses them +I want to change it to jmxrmicamel and will take a look at your patch +I seem to remember this idea coming up during initial development and being rejected but I dont remember why +Going to commit this today if no objection +For some reason I get s of compilation errors complaining about missing hadoop packages +Yes the problem has been solved thanks for the your supportConclusionI will not build the lib into my because of the potential risk mentioned by RemkoI will put the in my classpath and use to configure the logging. +Jiri +Yes if you use the API you can implement this +After making the modifications that Ivan suggested I noticed while running the tests that was failing intermittently so I investigated it a bit and found the following +Additional POM for the server part of Apache Solr +I feel like this would be best addressed as three unused importsvariables generic types depreciated classesmethodsIf this approach is agreeable I would be happy to own and +This was over a month ago dont remember everything about the systemDoes seem that META was on Ive not seen this since this occur again but we have not had the HDFS issues we were having back then so hasnt been taken downTheres definitely something to this though just difficult to recreate +thanks +Thanks for tracking this one down ShaiAfter fixing this is it then possible to fail lucene tests if we cannot delete the directory or are there still issues with special tests +Lets say this is your view and is one of n possible interpretations of the spec wordingThen its the only correct interpretation +So I think the fuzzy parsing is the best compromise +moving to. +Heres the patch +Cant we however just add a boolean isServer to create and skip the trustStore stuffs in that case? The rest is the same and this keyStore business feels out of place in the Server classThe rest lgtm +I suggest moving the code from core to a newI am already working on this and it seems that the remaining core becomes very smallTaking into account what Carlos wrote There are several parts of the code in the jsr module that could bemade more conciseclean if they were separated from the coresuperclassesdependencies +xhtml file +Ive attached a patch that logs an ERROR message when and security are enabled but the or configurations have not been set +You could choose your Store wrapper when you set up your object +When doing the random test the same analyzer is used by different threads so my first guess would be that the problems are related to this somehow +Worked as advertised +Tommaso is there any reason this cant be backported to branchx? +the patch should be complete now +Checkstyle fixes committed in r r +is now the only patch that should be applied to trunk +Is there a workaround for Continuum or an update for the SCM component +This issue is very simple will be addressed within +That JIRA applies on top of this one +Thanks a lot for the patch +Specs patch for patch has some problems but I think I got all the changes in rev New specs module is named and has initial version +Attaching a new merged patch +I changed this such that the pmf is closed in any case +In the end every attribute named like a transient field did abort deserialization of following fields +This will help with with Java API +Looks good +Anyway when I see anything I will post new findings +I think this approach is preferable to wrapping the in a Dispatcher since it requires less code and is more flexibleThanks for the suggestion to use +Added class to package to be able to resolve variables in JSF expressions as beans in the root application project provides a different flavor of integration reimplementing the JSF managed bean facility on top of a Spring application there anything particularly wrong with the ways is integrating? I really like the way major disadvantage to only having a is there is no way to wire a spring jsf managed bean with another scoped spring jsf managed bean because you cant scope a spring I missed the discussion to not go with the method of integration somewhere so perhaps someone could provide a link to itMike +Thanks for looking at it Ted +best not backported to existing release. +Now I see your point +Thats about all I can think of +Moving to reschedule if for AS and the +Hi Martin can you review this change please thanks +The code in trunk around this is really complex +ok marking this resolved then +see last comment. +Oops! Should just be changed to duplicate not resolved +Im going to put back my original fix at least this enables things to workIf you want to work on something thats better fine go ahead but please dont leave trunk broken. +Forgot to unassign +Verified in JBT +Fixed in +and +In fact my entity unit tests test the validation with the hibernate validator assuming it would work in Seam +Merged to branch as revision . +Oh +If you change Y the monitor should see the change recompile Y and trigger thereload of Y +sorry +What is this values type? String? Its not clear to me the argument type from the documentation at least where I am reading +I updated to latest repository and there was no fix to the issue reported in +name +This is pretty embarassingAhh +So I am marking this one as FixedI will try whether this fix has had any effect on as well and post back +We may want to not run any test when the profile file is invalid +From my perspective go ahead! FWIW the tests that failed in passed on my RHEL machine +thanks! I confirm it is fixed +minimal version of renderingservice for for the patch Daniel Ill have a look at it after the release +Whats your use case for trying to order things in a so fine grained wayStart order should only be used as a way to optimize things but bundles should not rely on it because bundles can be stopped or started at any time and they should handle those use cases correctly +I just managed to commit a fix for but wrote this issue in the svn comment. +Committed revision +Please attach test project that would allow to reproduce this +Lucas knows more about this than I do +Have you encountered any problems with this fix? If not Im planning to finally cut the release later this week +snk out of the box +I believe this happened after I moved the scheduler events to a separate dispatcher as part of the RM refactor +Some fragment of my code +Bulk closing for +Ive activated the two inactive tests and passed the +On further thought lets commit this for and work on refactoringenhancements if we can find timeCommitted revision Thanks Sean +Rebased and committed to trunk as r. +The will only be removed after the releaseperform is successful +Sending trunkassemblySending trunkassemblysrcmaindescriptorsSending trunkassemblysrcmaindescriptorsSending trunkassemblysrcmaindistributiontextetcSending trunkassemblysrcmainetcSending trunkgshellsrcmainorgapachegeronimogshellspringetcAdding trunkgshellAdding trunkgshellAdding trunkgshellsrcAdding trunkgshellsrcmainAdding trunkgshellsrcmainjavaAdding trunkgshellsrcmainjavaorgAdding trunkgshellsrcmainjavaorgapacheAdding trunkgshellsrcmainjavaorgapacheservicemixAdding trunkgshellsrcmainjavaorgapacheservicemixgshellAdding trunkgshellsrcmainjavaorgapacheservicemixgshelllogAdding trunkgshellsrcmainjavaorgapacheservicemixgshelllogAdding trunkgshellsrcmainjavaorgapacheservicemixgshelllogAdding trunkgshellsrcmainjavaorgapacheservicemixgshelllogAdding trunkgshellsrcmainjavaorgapacheservicemixgshelllogAdding trunkgshellsrcmainjavaorgapacheservicemixgshelllogAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainjavaorgapacheservicemixgshellloglayoutAdding trunkgshellsrcmainresourcesAdding trunkgshellsrcmainresourcesAdding trunkgshellsrcmainresourcesspringAdding trunkgshellsrcmainresourcesspringSending trunkgshellSending trunkTransmitting file dataCommitted revision . +There are also other features we need to make developing simpler like launch support for an AE in eclipse +Thanks +Just download and will reviewtest ASAP. +Conditional router mediator and the URL rewrite mediator have been updated to make use of the new feature +The thread is nowcalled the + to trunk +This can be fixed by using the latest tomcat release or applying the attached This will work for all versions +It should be discussed on dev ML IMO minor detail anyway +AlthoughI am not sure if can be used as a log buffer as itis we can reuse much of the code +I am travelling rest of the week but will follow up in detail next week +Need to give priority to this +to run classificationin my opinion the whole code needs refactoring as i really dont understand and agree with some of the design decisions that has been made while writing this patchof course your input for my patches would be really valuablealthough i must be honest that the more im trying to fix up this patch the more i think that it should be rewritten from the scratch +Im not sure I understand your logic +This looks like it reintroduces the NPE you can get with no ulog in Ah thanks I got a merge conflict and then missed your update +I agree with the cure being worse than the disease +Thanks for review +Give precedence and the right XSD will kick in +You dont have to run two nodes +priority I is done +We first need to fix the KB memory leak on each thread create +See the ROOHOME is not determined I still have the ROOHOME set to the correct location I commented the ROOHOME setter in the script so the correct location was used again +I do believe youre right and Im almost positive no functions in this file are intended to be exposed to plain old C code +had a first fix but needed to correct the replacement character strategy +When they bid the messages and updated price information are passing to both the users +disable table newname +shutting up here now +Also committed to +add undefined or simply reuse void as the value of undefined variable references instead of null which should only be used as the value of defined but not initialized variables +ok assign the issue back once youre done with the preparations you want to do +subscribe to the user list and then post your question. +Theres another bug logged torefactor all the tags to make them more extendible to make them easier tooverride functionality with extended classes +The component has been designed to allow access to all of the over methods exposed by the API to work with all Facebook Objects +This is a message distribution issue +Certainly we shouldnt force a rollback record should only be written if the tx was prepared. +resolving this now that it is possible to configure an of type that will enlist in the current XA transaction +In that sense I think the calculates well +Ive had to change some tags I defined to take in the converted collections rather than performing the expression processing inside my tag +Thanks Andy. +All issues has to be filed against from now +Here is the patch and the test that shows the bug and the fixed behaviour +Every time anyone has trouble well have to examine whether or not its due to the quality of our KML +And as I said calling it doesnt really help the confusion matter IMOa bit of history I hope I recall correctly +to implement JSR compatible interface using stuff from here +Ive added this mocked method into the map and reduce context objects +As you said the fix looks pretty simple +When clicked on the tab The first field Email address shows an empty drop down +WellIf I do findbugscheck during the build then I cannot do findbugsfindbugs in the same build because it already failed a build I have to do another +sorry +Hey HowardThank much for correcting my attempt to guess whats under the covers +Was able to recreate the issue in a test added test to trunk +Thanks Christian + patch Ive fixed INSTALL too +If so fire each event to external script +Hi Chris I was going to ask if there was some issue in downloading that dependencyNo +Please dont publish the test caseWhich JDK are you using? JDK but is the from IBM or Sun orSince JDK bundles Xalan inside of itself you must have put Xalan on the boot classpath to over ride getting it from the JRE is that rightWhat are the other problems that prevent you from moving to? Are they problems with Xalan itself? Im thinking that getting a service update on Xalan is unlikely but fixing Xalan problems in is much more likely +Its maybe not the best way but it works +Weve done a lot of work on the output system and can no longer reproduce this bug. +Andy can you take a look and comment +Tim the patch includes update for make files and this update was missed in commit notificationThanksStepan +the patch looks good +Fixed to M +This is in cvs now +I though of fetching the data from the GIS more like of a convenience thing than the actual task by knowing the id it could be done afterwardsThinking a bit more about it leads me to the question if we want to support the case of having results per input SpanYour proposed find method would simply return the best match where a second find method could return the matches +Port still listens but will only serve deployed EJB remote interfaces by default +Thanks for the pointer +I have a question on this if we have two documents in the index Document entrry contains field field field Document entry contains field field field Shall I assume that the total unique fields in the index are ? We have a set of key and value pairs that we add them as fields to the document +Flagging as incomplete because no feedback from reporter. +Here is the equivalent patch for lucene +to trunk +I was hoping to improve the maintainability of the manual by putting in those variables instead of actual dates tags and versions but a they seem to cause confusion and b obviously I missed some +I have already mentioned the example of mysqlThe problems with the existing solutions is that there is no automatic failover +Couldnt we keep it and have it link the download page +all issues that Ive reported and are resolved. +This still adds the String parameter to process that is never referenced in the body of that method Does this test time out often? Is there a contract that this should take less than a minute to run? I completely sympathize with the impulse to retain the information that ant discards and something is probably amiss if this test hits the timeout but if were going to fail tests because they violate limits then those limits should be based on something quantifiable and intrinsic to the model were testingIf the purpose of this issue is to use logj instead of to print the error inclose then this should probably be limited to changing that line and updating the unit test +This patch seems to fix the issue +This is working for meHave you defined a Spring IDE? Do see any beans at all +Perhaps we will also need to add in a command line flag or configuration value to switch the behavior over to the new API instead of the old one if we cannot easily dynamically detect which API is being used +The only difference I saw was the missing androidconfigChangesorientationkeyboardHidden on my main activity +This patch contains a broken system test that uses a bus shutsdown the bus and attempts to reuse the bus with the same server http url +Attaching new patch incorporating review comments +Thanks +this patch defines a system property that you can use to express the preferred list of containers to use +Removing from +Committed at subversion revision +ShazronWhat Im trying to say is that if I hard code all to false in ios outside of it works as if it were true in ios +Just took a look and know what has to be done +I think your fix is right as we should escape special char from committed to master and +This source is very similar to what we need here for our requirements +Patch applied. +is actually allowed in Java resource names +Poorly designed +As usual I found a couple corner cases that my first patch didnt cover +I have two more ideas +A tricky one +smile + +pushing forward +Looks good to me. +Ninjad. +I dont want to make proto too fat +I just need it fixed +On a serious node +I am running on Ubuntu and also have Spring STS installed +I was thinking that rather than sampling and resampling repeatedly maybe what you could do is look at the region start keys figure out what the edit distance between start and end keys is as a proxy for size of the region and then scan the presumed largest and presumed smallest regions +OK +I committed the change to trunk and. +Assertions fail at the moment but this is due to bugssrctestresources New files for better testcases +My testing suggests that this was purely a result of my attempt to patch +commit log has to be recovered before reading from any CFThat cant be true can it? At the least we have to special case schema +Guide +I am not logging another issue since the root cause for egress rule not working is because of router having multiple nics for the public ip +TYPO After testing I cleaned Raghavendras contribution and created this patch +tested sharebli createupgradehelp commands successfully +This eliminates the race where the container can fail because of the preemption and provides a way for the AM to potentially checkpoint the state of the container for faster recovery +Proposed change to taking advantage of + Username mkotsbak +The fix is really simple change last after modifying data treeAn alternative approach would be to set zxid while taking snapshot to last +MarcoThis field makes perfect sense for POS environmentAbout the field name lets add also the word to it as it is a common pattern when using Enumeration entity something like reason or paymentBilgin +I believe this is related to the change the in the way the results are created when projects are built from metadata +Due to comment provided by dmalarevich its eclipse bug so to later +Sounds related to but a bit different + please comment on this talking to James moving this to RC +This might be a documentation bug +Works fine for mebinhod allocate d tmpa larchxorbinhod allocate d tmpa larchxWorth updating documentation +Thanks Michael. +All issues resolved in released versions +You are saying that the newly rolled file the one that is created inside the synchronized block that it got some edits and these edits were not observed because notification of roll happened after the new file was created? If so on your patch +In order to reproduce the bug one needs to download Nightly +IMHO the role comes not from the expression but from the type of the component +Committed with revision by Scott and merged into with revision . +Technically No +Do you mean use a special token type so users cant spoof theSYSTEM principal? Im okay with this too +It reflects a sort of guidance that we might be better off just for the bean definition metadata indeed thats not necessarily picked from the actual method used at runtime +All of the packages referenced in NOTICE now have sections in LICENSE as wellI have also added Apache license headers to a couple of files where it was missingThis patch should be applied to trunk and the branch +tested +Creating the same project in Eclipse Java EE Kepler M doesnt yield any errors +Fixed by. +Very soon it will have it config reloaded whenever ll take a look at Multiflex blue and mule I hadnt finished debugging it in UnityBen +Would you take a minute to verify that this solves your issue completely +Thanks for going back and retrofittingIm letting this patch hang a while because it would be good if Andrew took a lookseeMeantime running tests to make sure avro or rest dont barfGood one Riley +Is the objective to be able to enforce IO throttling for the cluster or for certain applications to be nice? If the former this must be enforced on the cluster side not on the client side +I have the same problem! Apollo cant restart unless I delete all the data +The test was failing because it was checking for file sizes on disk and the sizes were hardcoded in the test +fixed +Example to reproduce the compile error +bq +The ctor for Key already accepts conf but it ignores it +Im keeping this issue open until the are closed. +Sounds reasonable +Sending srcmainjavaorgcodehausmojojspc file data +not sure what should happen here +Thanks for your time Rodde! Once again your patch is simply amazingYes I am in the notification list for that issue as soon as youll attach a new patch I will receive the notificationMany thanks in advance all the best! applied in r thanks a lot Rodde for you effortLooking forward to apply more patches from you! +Andrew can you review please? Thanks +We need to fix this part + This bug has been marked as a duplicate of +Id say to hell with windows minority +Heres a patch to fix this +Hey Brund Hey Jacquesyes i checked the patch in a fresh checkout +Are you including the classpath of the generated xmlbeans on the wsgen classpath +Corrected toUpper to use the patch +Thanks for the patch +Vivekbq +Hi BenYes there is a level of effort duplication that might not be necessary for something that could just be addressed in making sure the documentation is up to dont think its crucial to have a Java version of the sample remove BenI removed the samples example project +Yep did it on cmd lineConsequently now it fails when run from Eclipse the first page is rendered all following pages are inserted with correct size but are blankThanks for saving my project timeKind regardsMichael +Thanks for figuring this one out Brian. +Is there a plan to have the branch use the new DITA open toolkit? Doesnt seem like an appropriate change for +I wouldnt trust the user on that +Fortunately its in there since M thanks +Can anybody provide a patch for? Thanks very much +patch makes changes to to make it CF comparator aware CLI show and describe commands works just as expected but it breaks related migrations with some weird NPE exceptions + contains the changes to port the testlargedata to junit +Adding to the UI is nice sugar but exposing via JMX and the hadoop metrics interface is higher priority IMO +Quickstart is present in ER +Also added a simplified version of Katheys test case to +I just type up a whole long thing for this and then lost it when I clicked add and my session had timed out so Ill be brief +A fair number of tests are failing +Thanks +The main problem being that users would no longer have any short cut for role mapping definitions where there is a mapping between group names and role names +Site patch is applied +It would be great if some can review and commit this patch +Changed Target Version to upon release of +You are right this example will be changed with the next release +Hi thereUpdated a couple of the patches to keep up with changes +list is fixed by commit ececfdedacb +Without an MSC failure the management layer is not going to detect any problem and will not roll back the deploymentIf this assumption is incorrect please assign back to me +Ill start preparing a test case to capture this error using the request generated from the ldapsearch supplied +We are done with our testing and everything worked as expected +If you find a patch please submit it they are always welcome +when you start typing token sug it wont suggest token sugar but instead it will suggest token suggestionsbq +Sure +The question is just really Should we have it be the default on SolarisAlso Once I have a few days time to port ATS to Sparc it will make a difference +Even if the project is not faceted +It seems that the trunk version has already fixed this bug. +handler call logic has changed to take this into issues resolved in released versions +Manually tested +It would be nice to test it +Fixed Pyro issue and updated documentation. +See forum post for workaround +I just downloaded them from probably a day or so before you them +assign svante to good! Make sure that the test is updated as well as when the test passed and the metadata was not updated the test needs improvement as wellThanks Devinseems as you have not added yourself to the ODFDOM project only to the ODFTOOLKIT project +Problem in all geronimo versions +Closing a number of old issues that I raised but which I dont think are still relevant +It adds a maven to api +The parent and the child projects will have a different solution could be to force users to define groupIdartifactId if id isnt defined +It does not affect what is released in EAP +this should be fixed contains version information +This subsystem is still under discussion so moving related to CR +Thanks again for the review Chris! Ive committed this to trunk. +Regardless both are functional and have been tested with their respective versionsFor on run mvn package P debFor on run mvn package mvn N jdebve been discussing backporting the maven changes in trunk to so we may just need the one ScottI patched the branch and was able to build a +Attaching initial patch with some clarifying comments removed for final submission +Created an issue to track this regression +Bug has been marked as a duplicate of this bug +Ganesh can you please verify that this fixes the performance degradation +done +ok looks good to me +Yes that corresponds to what I was seeing as well +Ill patch this for release Thanks very much for spotting the error +Which build of PTP do you have +I have seen a lot of weird crapola over the times but this one beats them allI have not see any other framework have to do tricks like this to just workAnd if its really needed then code comments must be in place stating why this weird code must be in placeThe does not have the need for this weird code +I just committed this +will get it in to that branch thanks for are up after upgrading this build we are testing now +hence preference initializer would be used when you ask for the preferences the first then restore defaults button would basically just mean nuke the preferences and read them again +Thanks Amar + will commit after tests thanks Ning! +Everything looks pretty good to me +Im also not sure why one Linux box would be fine and another wouldnt be lingering ports seems a pretty common issue on Linux boxes +I am using Indigo Service Release Thanks in advance +Nor is stopkey on index scans something we want to optimize for +merged with trunk +Clustering it on anything is just one algorithmic modification where at the end less information remains +Modify master file of result for blobclobBLOB +Ill incorporate it into my patch +Given if you want use Jboss tools dependency interface you can write all the Teiid based CLI commands to do various tasks yourself with no dependency on Teiid classes or if you can provide the to the Teiid Admin API wrapping classes to to expose same functionality in nice Java API +Is there a chance this occurs on the branch as well +Its better to develop new JMX module for Axis +I committed this to trunk and +This patch addresses Dougs commentsUnit tests for recursive listing of symbolic links will be added after is fixed +This leads to a severity from major to normal pending further P open bugs to P pending further review +Indeed Vincent Ive managed to reproduce this error locally when trying to save some Settings on a project +Thank you Jing. +attached patch which fixes it +Enter an existing order number +Please verify +To reproduce +Will have to try tomorrow +Thanks for the patch +I run there is at least the workaround to define the Boolean as described in the above link but maybe adding something to the connectionstring also works jdbcmysqllocalhostdatabasetransformed the following statement throw a light on what is observered exception when inserting new boolean values always returning false for bit values already in databaseEspecially Im curious to the exceptions +Thanks for the feedback +The ordering dependency will only provide a partial solution and we dont like partial In other words we have to think a bit more on this one +Does the restriction above come from the fact that we cant support the alternative behavior? That would be to allow the procedure to be called on a soft upgraded database to remove orphaned statistics and then continue using the older version afterwards +So a couple more notes for these figures the dataset was LUBM with materialized forward chained inferences giving approx triples loaded into a TDB database +Jostling around the release goals and numbers +Genernally no tracking +Girish has fixed this +Thanks Isuru! +Its been a month since issue for raised +Im using windows XP +r adds this mantis feature +Sorry about that +As Scott said this is the drivers issues +The test was committed at r +so resolving it again. +I always thought the intent was to have everything underone context and it was confusing that fallbackURL was different than loginURL and unauthorizedURL +Messages with the same can have different values in and the connections associated with the respective networkhosts would be used to send these messages +However as we currently have a couple of bugs fixed its probably worth a release as soon as is closed. +In the meanwhile there is a simple workaround +when +Thanks everyone for reviews and discussions +This is an incompatible change +Pretty much done all tests pass +Here is a patch to remove the sections numbers and resolve both issues that Kyle identified +It does seem to be an issue specific to the Virgo server +Seems problem is inSee line is +Some uses of the older were replaced with uses of the new Graph API all other uses of the old were replaced with the new Requests. + updated for current state of trunk +Committed to and as well +Maybe its linked with some kinds of documents? I dont know Im gonna test some other possibilities in case it can help +Ill take a shot at fixing this already for the release. +The Spring Web Flow integration for JSF also relies on Spring MVC to be the front controller +Namely we should not add unit tests and then put down Ignore +Automatically synch up the planned and actual? But that isnt what you would expect eitherNot sure what your expectations are with this JIRA +Two authors will take care Dominik Pospisil and Ivo Studensky +They were +After this is done do you think we can come up with a simpler design? I do not have any particular idea so just spitballing here +Keith Im unable to reproduce this +That is supposed to be the append release correct? I worked with Suresh and some other HDFS committers to get this in a couple months ago +Claus should this patch also be applied for the branch? Because its more a bug than a new feature or improvement +fixed but please comment if you think of any issues with provided perfect +I need to go over it a few times +is setup by the parent process to ensure the output are redirected properly to the desired location +Ah ha good to know thanks Martin +Marking this as fixed +I completely agree with your fix and appreciate the on the context between maxTries and maxCollations +Let me just commit this +The only simple way to get this fixed would be to check all the entries to see if they are using the disabled schema and it would take a very long time +I was able to avoid injecting validators and a transactional service into the read only controller +Fixed in rev +Removed unnecessary quotes left behind in the documentation +It just seems wierd to put it in a file that has paths in it when the service objects are not +As a result I noticed that to the unregister method the actual object is passed and no proxy of it +The version is able to hande the entity on Top LevelBest regards Micheal! Ill take a look at it +Consequently an observer has to make sure that it has seen all updates before committing the ensemble update and be ready to ack and commit the ensemble update once the leader of the new ensemble proposes itHenry could you sketch out your thoughts on how your modifications will handle ensemble changes? I wonder if it is best to try to reach agreement on this issue first before submitting a patch +jboss uses Tomcat if that is of any use to you +Patch works fine +I guess you cant deploy other applications either +Resolving as incompleteLast few uploads and internal use here have less open because of changes made in hbase prompted by investigations around this issueRegards datanode using select instead of a socket per seems like folks are more leaning toward timing out idle connections as means of resource conservation +My rails application uses the soapr to connect to a on a Linux guess the bug is still there in +So now I reuse that code and just add support to use this function as function query in. +Fix bugs in to fix the throughput calculation and the logic for having multiple threads +Thanks DarynI put this into trunk and +i just pulled out a lot of hair trying to track down WTF so i need a break +I suggest then that you create a patch that is complete so that I can commit it all at once + +Added class definition in a unit test as suggested +Does this happen on any other device than the Huawei? Can you reproduce the error reliably +Fixed on and closing all resolved issue that have not gotten any more feedback or comment in the last month +The archive should contain the new POMS a couple of I missed on the original upload and MD checksum files for everything +will bring this patch up to speed +Fixed per +Thanks Mike! +Hope pradeep included that +Anyway this can be enhanced of course +so you think I should check if everything is still where its supposed to be +Since we identified the problem it is a cosmetic issue that is caused by invalid input so not worth mentioning as far as I am concerned +outputSchema in outer UDF uses inner schema to infer alias +After adding +The name of the tables for the classes are respectively tbc tbb and tba +We probably just need to adjust the dependencies on this +If its your itch to also support the DB JCC driver go ahead but your first implementation of this could include just support for the Derby JDBC embedded and network drivers +Check what the root cause is for the difference in return values too +Thanks a lot Richard! +Updated patches to add timeouts +If other operators need more compile time information they can extend +I have the same problem with MS SQL +Attaching patch for reviewComments in the code should exemplify sufficiently +Heres a patch that goes back to the META to check if daughters are presentTesting it now +Can you verify that the steps there work +applied because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +New patch that uses from Javas concurrent package +It makes little sense to me that you should be able to annotate an interface then implement that interface and then require yet another path annotation + to close it as wont of scope +Maybe even create separate project with its own mailing lists and JIRA project for it +revision Thanks Johan! +This was a nice thought experiment but at this point is mothballed and Im not particularly interested in pushing richer types into +Attaching management server a part of baremetal that youve done can you please makesure that baremetaldhcpdevices table is being inserted as a part of theDB upgradeIt should be fixed in both and master branches +Whoa good catch explanation of the issue! Sorry about thisI will fix these tests so they resolve the field by name and do their comparisons on that basisI realize that exact file name tests will need to be updated when we change the file format but I really wanted test coverage that verifies the generational logic to derive file names is working correctly for segments del and separate norms filesI will temporarily turn off and until I can resolve this +I dont have anything for such a test the existing eval framework wont work here +Have you tested with? This should be fixed in the current trunk +Good point about the default Ive added that to the description +Marking this as minor as we can come back around in a point release to add the polish +A comes as an for release. +Uploaded patch for upgrade to and Lucene upgrade to +Occasionally it takes awhile and I think interrupting the database upgrade is what causes the issue +Only the first call for eachseparate locale would spend much time in the function +Let me put some thoughts hereMy understanding is that currently we rely upon a safety property from leader election which says that the server that arises as leader has the highest zxid among a quorum of servers +Option is active only if hibernate version is or later because according to DTD it is illegal o prouce configurationwithout mappings +Hi InfraLast we heard on this Jake Farrell said he would look into the problem on the test server +Site rsync complete test passed +If you find further issues please create a new jira and link it to this one. +Do you think this would resolve as well +Reopen as necessary when committing to other branches. +Looks like the attached works and it has been commited to the repo BUT it was after and hence FFI doesnt work in under Linux x +Thanks +This is now fixed in the alpha branch at least +While testing the patch for I printed out all lines written out by this class and the missing lines for the unit tests +I have added a complete HTML dump of a directory listing from ArtifactoryAs you can see Artifactory does not have a trailing slash on directory entries +closed. +Fixes unit tests +Moved issue into legacy component +We still need to solve the case with small texts +I also wanted to mention that the stack traces showed more or less nothing +has been removed +Its great to hear that this could be resolved so fast! I checked the latest snapshot and it works for meA big thank youBest thanks for the feedback +I dont seem to be having luck reproducing this +enough to sink the current rc +Current patch is based on baidus branch Im testing for it +No response assuming ok. +Yup Id rather see classes pulled into their own files than having a +Cannot update from to due to the change in underlying Eclipse base +Just verified +Even if it is broken it might be better if I check in the code sooner than later so others could start reviewing the API +eXo will automatically become available as soon as this task is complete +Good points +I just committed this! +Fixed on all issue has been resolved for one month or more without complaints +Regarding how they coexist and whats the difference may I quote a previous email I responded as below +Committed at subversion revision +various minor fixed for +but link is incorrect as for instance this link is httplocalhostdashboardconfiguredid instead of httplocalhostdashboardconfiguredid +We should pick this fix up automatically +It does not use allowSnapshots at all so SNAPSHOT deps are not considered dependency updates even if they are available in local I filed about this masking bug +Its not an issue of the runtime not starting quickly +There is no dependency on whether the first client is dead or not +Thank you for your patience +Proposed fix for Embed issue +Old style not supported anymoreThis changes the configuration style +When the changes are commited the files are all renamed +Also add following functionality to the same page Enhance the screen as per description +Fixing +My to move forward on this +looks good to me +This is a Java backend to +Im taking it out and trying again +When we send the block reports from DN to standby it may give the command to DN as invalidates as that block information is not avalible in standby +Ive marked this as critical as we need to get the branding right on this +Cool I will do a one pass later today or tomorrow and add our regular contributors into that list +Could there have been a code regression? Does anyone else see this problem +If you think so Ill do the an observation as this method is frequently used by the graph in the joinpoint graph branch I have created a cache for the results of this method which improved the performance with the graph +MarkusThat looks like an excellent solution +To respond to Aruns comment above we will integrate the slicer stuff but we dont have to now +Hi Chris I cant reproduce this and I added a test for it in SVN +but the parameter does also work alone +If you edit this jira youll find the Release Note field +bq +Ive marked this bug as fixed because Ive applied the patch to the source repository but I have not repackaged because I dont personally have the resources at my disposal to do all the testing that would be necessary. +we got a stale resolved issues +I got reports that tap disks where not destroyed and mount points notcleared +Ajax areas will be reRendered in order that this defined at page so if you want rerender one area before another specify this areas one before another on the page thats the right behavior +Please look at it and see if you can see how to solve the problemIll start reproducing the problem and attach another with the first exception we get which is different +is already obsolete +I have had no responses about maping the stat structure through JNA +all tests revision . +It looks good +The current patch does fix the Pig but the MR job would fail to find missing classes in classpath see +Nice catch about the feild nameAs for putting the field names in the file lets see how this would work We need a matching pair and both should expect the same fields in same order In addition has some fixed field names ID Body Name Date Title and a flexible Properties which can include anything else +The configuration files for datasource have been moved to confdatasource +That is what I believe we are looking for a JSF with option for users +Hi SteveI agree that does not handle the stale digest parameter but I do not think this should be a problem in standard use +Never mind trying to rush before leaving the office and the tests fail here +Alternatively we could make classlister skip all classes in the sanity packages when building insane jars +I think it can makes following the discussion in the comments easier if someone is trying to correlate comments with a patch +OK my badVersion dont have dirty check so entity get always indexed correctlyVersion have better dirty check and we have Field on computed field +Mahout shouldnt be doing any sort of configuration internally +Just copy your wsdl file in to this and try the link again please +committed +What you described regarding save is already the case but then we have about too many commit retries +However is not working as it should with Groovy +This can be extended to implement more methods ifrequiredThanks Vincent. +The default installation directory for a JDK under Windows is in Program Files +coretrunkrepository +Correction on patch provided +Handled some corner cases +We could remove the Profile Name from the SQL Results View History by changing the property preferences. +So all issues not directly assigned to an AS release are being closed +attached patch which uses thread synchronization to accomplish the wait until gobbler BrianThanks +Guys please validate the ll close the issue today for a release +Theres a similar in +Probably we will use new features of for DND +Please note is no longer supported +This would be kind of ugly and verbose so I dont really prefer this approachAnother approach would be to always generate the getters and setters whether or not were in beans mode +This patch adds a Kerberos functionality and we currently dont have a framework to unit test it +I understand your use case well see how difficult this is to be done +The changes should be mostly transparent but I did change a couple of as well +If its PUBLIC the resource will be cached and reused among all users and all jobs +Thanks for all the work +EAP contains seam with sources so it is easy to implement we guess we just need attach folder that contains sources to each jar +bumping it up +I agree with IainI have use cases right now where structure is quite important I compile native c components with headers that I do compile against I store artifacts like configuration which I do not compile against but will use in testing at runtime +sorry I added dependencies to IT but it didnt cause any traceIll really need an example as this one as a duplicate of since I can reproduce the error in +I agree that would be more appropriate +OK so I realize now that the java server tests are mostly designed to be integration level tests of each server +Could you zip up the contents of the log directory so that we can poke around a bit? Thanks +fixed in +Thanks Patrick! I will take a look and get back soon +yes it will matter because thats the code that will be executed by you if thats what the dependency ll get that uploaded for you ASAP +I believe that allowing Sqoop to run on more environments will allow the project even brother adoption +Ahh didnt look at Vijays second patch but it more or less does what I suggested however I thought we were already keeping the resettedAddr in the system table but the second patch adds that in +Fixed in revision +the new output is much darker in the letters +Thanks +In addition to the configuration that you have above you also need to add lombok both as a build and a compile dependencyThis patch looks good and Ill accept it +If so this issue is mislabeled and the patch needs redoingI was just suggesting that if you want to actually drop a regions data you could pass a flag to merge and it would not bother copying over the files from old regions +nitay. +You may use the to generate the JSON but recognize the different JSON representation +Thats what the maxRetry is introduced forHoward +mrepositoryorgglassfishgrizzlyUserstucu +Dependencies of the JAR can still be specified with the add option +The name attribute is in fact set dynamically via javascript during file submit so I removed the name attribute from CDK template +as r thanks Sijie +Thanks Lukas +I committed this +This has been resolved specifying in now gives you access to +Ok Ive figured out this is actually a load sequence issue +In this case the should ignore the and still analyze the to check if combiner optimization is possible +So in this case the should wrap the newly found value from the data container and put it in the context so the can properly remove itTo be honest I thought we had this case tested already but we must have a slightly different variant +In m as servers and runtimes had only one wizard page as a potential optimization but this led to severe issues +checked in NN for Teiid Designer +Fix by revision change language must be click refresh to show language which changedExpected output Dont need to click refresh to show language which an ajax request we couldnt reset dir param on template to ltr or rlt so we must refresh to affect. +Rather than removing with r for was removed +We do not any longer place the native libs under a specific arch directory and rather just place it into the native directory itself thereby solving this automatically +The only cache would be one internal to the webapp caching index information and that doesnt yet exist. +Fixed see +Reopening to change fix version from to +Thats why I wanted this bzip thing in the first place Ahh OK +to committing this +Please review but do not commit this patch in trunk +We were including the boot image compilers into the boot image +I want the same type I would get if you substitute val for var +The fix looks fine. +Im more than willing to improve this task if theres functionality that would assist Forrest +I see +No need for IT. +Resovling +It seems like theres something strange about the set up on hOn h and h what does yum list nagios show +Updated patch including unit tests +has been marked as a duplicate of this bug +Committed to branch and trunk +would perhaps be more appropriate. +I guess that the is not a real Java class file +kept experimental items by committers renamed experimental +Looks like this timed out too +close it. +Id also like to change the terminology in the code t you a committer now +Applied thanks. +yes that would be great +Thanks for all the cleanup stuff. +Even there are some unused imports Ill fix them before commit this +patch ve committed this to trunk and branch +The problem with setting any timeout is that we dont know the upper boundLarge commits or optimizes esp could take a very long timeThe example server is configured for up to K threads to avoid running out. +Fixed in +This wasnt a problem either +I use this as workaround at the moment it means much more performance overhead for my use case +Thanks Devaraj! +Ive added a corresponding warning to s +I have a rule configuration file in order to establish the rules which sonar must apply to calculate metric values +The issue is that having two trees both with tagIdtree they must appear in a that creates a unique nameSee coreWebrichTreetreeSC which does this and contains verifying that the tree and treeElements can be accessed +This lead to a few more duplicatesThe above causes are listed in the descending order of the amount of duplicates introduced +Thanks. +Thanks for the heads up +once i did that the error went away +So the old client will fail since the pipleline is completely torn idea is that no writer is allowed +I found this bug report only by accident +Committed a temporary solution to the problem in Rev +Jay In there was a discussion around how Owl is different from Hives metastore +Did you see it failing again +After changing this timeout to milliseconds I was able to run syncope. +I get the exact same that i was getting previously +Would you mind confirming the resolution for your needs by giving a recent a try? These fixes are also scheduled to be backported to +Verified that this issue has been resolved correctly +Weve worked around it by using Toms fixed version in an own component for nowThe fix works as expectedCan someone please assign the issue to tooRobert +Assume closed as resolved and released +Can you attach the file where this exception occurs +As all work related to XML configuration will be done through I am marking this as resolved. +This should solve the described issue +Attached patch fixes the issue in my secure test environment +Move this issue to as there is still some work to do before resolving it. +Patch updated +It creates an issue of type Task and not Bug just in case anyone is interested in knowing +Ill also give it a spin on today +Denis what was the fix +Ivan have you considered private branches for your developers? It is a great help on large projects because large projects cannot tolerate work that isnt ready for integration +Been trying to figure out whats the problem for two months now +I assume this will be a point of some discussion +When I use your code without a server side code I get a Connection reset which is which is obvious +This was done as part of +However we can still run them as a part of the nightly task using +Great testedYou just need to update the text with what the doc team wrote on the Wiki +A leak and an easy fix to throw into +On production systems logging is normally a WARN and up +Ive sent an email to my contact at to request more information +Pushed a first into the master branch +Smooks editor no longer being devd +I have read a bit on the lazy initialiation of plugins inside the Equinox framework +What else would you need +Regression test and the fixJust wrapped the code which accesses information with check for null +Molt probablement tingui un origen se sacseja i no es remena quan els ingredients sn poc consistents i interessa que quedin el ms sencers possible +The second one is more descriptive +And that test verifies proper parameter binding and returns the correct result +you please open a new JIRA from your new issues +This might be a bug of could you test it in pure openejbs container and if it still exists you can raise a bug thereThanks. +Instead of deleting the test I changed it to pass if the correct error code is sent +How about keeping the prerequisites section and then having another section below that called Getting the Code if you dont want to duplicate the checkout information +yes i was using the trunk from +Written groovy for data fetching +Hi my friend I followed your steps but couldnt reproduce the issue here its working for me +This is confirmed to be an issue in by Rostic Sheykhet in +Unit test giving expected exception +Granted this push parser sounds more like a parser +It is called if it doesnt work! +Not all the new redelivery tests are passing yet though +If the appView isnt attached then theres no UI for a spinner dialog to attach eitherWe might just have to wait until is out to actually resolve this since the check is in there +Can I assume that the Codehaus infrastructure has Maven or binaries +Please see Scotts comments in the forum Tests passing and my injection in EJB is now working +Update the fixed version if you happen to resolve them sooner +I will try to change the call from contains to containsBean in and see if it helps as previously mentioned +bq +This is a useful place to document discrepancies between the early access and GA versions of JDBC +If someoneknows a way Id be happy to hear about it. +Somehow in all those testcases on I didnt test the one case that Hadoop uses a wrapping range centered on a member token +bq +Since jBPM now uses BPMN as the process definition language the more generic question would probably be what constructs in BPMN can be used to implement which workflow patterns +As I said in another bug please do not change the default behavior +should we do this inProbably its not worst shooting ourselves in the foot +Patch integrated to trunk +Will merge to and +start stop and error as JS API +At least we can add all examples to the war even if they are not used by the of the organization of the examples into modules we need to keep the in lib directory to not get compilation errors in eclipse +Snjezana thanks for looking into this +Btw +Please checkout SVN trunk to get access to this new code +Now that work is underway with AS all previous community releases are +It might be possible to make the parser pool pluggable and then use the implementation +I added a new class that tracks hardcoded quantiles and rolls over at the specified intervalThe test cases are pretty thorough but Im going to work on hooking this into either HDFS or to make sure the API feels right +I think this should be a good start +I can create some nant script to generate installer for you guys? You will get an MSI installer from that +Thanks for adding the comments! They look good to me +Thus the master saw the RIT timeout and then reassigned the regions which may go back to a weird region server or not +The byteman grammar rules really needs more work on them so this issue is being left open for future releases +Should be assigned to me after fix without resolving +Ben I suspect all of these are obsolete but please take a look and keep openretarget any of them that are still useful to youthanksdave +Sorry for reply again +Sorry for late response +I put together a test project using the Maven Struts Blank Archetype and configured it appropriately for Struts version +Git commit how does this work? How do we apply this without pushing in the roo files +This has been fixed +Related also to Yeah that probably would break other things +This is a duplicate of to trunk and! Thank you Mike! +pushed to in JBDS B no exception when deploying the nested project +Considering the hassle required to override the super POMs plugin management for a plugin that is primarily used from the CLI the super POM is IMHO not the proper place for such plugins +Len can you verify +The definition which define the length of characters +I know this code its on every platform the same +Uploading latest the failed tests on local T E S T SRunning Tests run Failures Errors Skipped Time elapsed secRunning Tests run Failures Errors Skipped Time elapsed secRunning Tests run Failures Errors Skipped Time elapsed secRunning Tests run Failures Errors Skipped Time elapsed sec patchV for HADOOPQAResultsTests run Failures Errors Skipped +It is NOT a workaround that the locator stops searching after finding an AS server and no longer searches the nested paths +Actually thinking about it maybe the message should appear on the line that contains the return statement? I think Ill leave that to the groovy language experts to decide though +Throwing it into +This fix looks good to me +Updated patch against tip of branch +Hi am the CAS maintainer +You going to commit Lars +Hello everyoneIm a undergrad student from Brazil and Id love to participate on this idea +Now dont start fighting on whos going to be assigned +Just noting here that is complete and the API did not change. + with one comment is the exit really needed given that its in an structure anyway? Thatll prevent binhcat from being run with a +ugh +I dont love the syntax right now but I like the functionality +This removed theStefan +If we do not want to implement a url database on an entity level i would propose to build a simple filter that +Even if a warning is printed it will end up deep in some logs and probably go +is slower thanrealusersysrealusersys +addresses the issues raised earlier today +I need a bit more information here or I can reassign the task if someone finds it easier to do it than to explain it +patches from Mohan KR appliedWe now support JSP templates with the tigris decorations onlyIf someone wants to convert over the other demos that would be helpfulThanks to Mohan for a big contribution for his first experience with Jetspeed +setup is currently available for your testing MS host VR MASTER currently stopped in UNKNOWN state VR backup in FAULT cannot reproduce the issue on my setup +I dont think theres any way to resolve this issue +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks for the quick adjustment Kim +The latest changes are looking even better +correct but now everything seems ok +Would that work for your usecaseWe cannot just pass in directly since it can change in the lifetime of the query that should be ok mabye actually move it to some internal package +Also this test has been moved into +Hey Rahul the code changes look good thank youRahulJosh Im confused +Dump log segments was printing the value as the key fixed +oups my bad its really a scm issue +Please apply this patch to the trunk +Im not seeing any issue at the moment +Although there is nothing against to upgrade that version when embedding jbpm in the you tried with? Maybe if you could provide more details about the issue we could look into it as if that is actually an issue when using gmail that would be rather serious +r +its just waiting QA to close it. +Fixed +But we just copy what AS does here and it does it exactly this way +But I just tested it and the app still works so the warning is benign +looks like we can actually incorporate the plugins from the command line +Not a frontend issue until we get a consistent design and spec +Patch containing perclass patch containing both lock and of the last patches are against trunk +go for it +This is NOT critical +The attached patch will be in the code drop +I do not have one that scales from image +changed for HQL nested news support +James I included the revised icon to Tools looks good AFAICTI believe you can set this as resolved +Using still seems to failing in the same way +it only occurs within transactionit is ok to execute the same process without transactionIs version manager able to handle the transaction case? +Patch applied +fixed in not fixed +Err that conversation is both wrong and totally irrelevant +Fuseki does not yet it willI was expecting resettable timeouts if possible ideally both sync and async +Yes Vladimir the transient on Cache field solves the issue +It is being handled by the other tasks under the jira +I understand why it is happening but cant think of a nice way to explain the behavior to new users of Groovy +Ted this patch is just for +Ill revert the commit while we discuss +One thing I noticed is that if you add these lines to the file in +A patch against trunk will be forthcoming shortlyIll take a look at the stress testing Jira and see if its something we can help with +added outofelection small additional change to drop notifications that do not come peers in voting viewTesting donejunitsystestrebooted peers several Vishal Apart from not having a test it is for me +I seems configured all the beans and stuff like that fine +I was able to double check the patch using a sample pdf Dave sent me in private as it is confidentialI added the patch in revision as proposedThanks for the contribution! +Updated in revision . +the description is a little laconic it requires an understanding of how the exif support is implemented +Includes a test case +Patch containing documentation for the integration +Hmm +Didnt make it into tooling +The problem is in various quotation marks regular expression of short report assumes double quotes only +The change bar starts and finishes too early + should be retained as part of the hadoop jar as it is need in backend nodes at runtime +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Please test +Here is a straightforward fix +We talked about this offline but regarding review comment I personally like the renaming the +Im going to push the code back into google code so Mark can review and add the improvements he was thinking +we cant reproduce and no follow upif you are still able to reproduce and can verifyfill in some of hte follow up questions let us know! +As of this point the query does not show any issues any more +Taking a look at how mountable is constructed it looks there is no straightforward way to get given mount point or path +Isnt this yet another reincarnation of the root cause of? TX commands for same tx are not supposed to be executed in parallel +add PREALLOCATIONLENGTH constant rather than +Apologies. +Fixed in revision in the trunk for +I am attaching my test execution and an updated patch +verified on Tools to write release note issue now release note text has been written +We have created desktops for all projects for which weve designed logos +Migration was removed from. +Whats the status on this Ersin? I think you fixed this one already no +Thanks Siying +I looked at the source code of your test class and found that the actual cause for the is the being used in your sample +Downloading Cobertura should be a part of the test process not the build process +PR for maintenance +Both keys in the key map point to the same spot +merged into is marked for but never made it in +With this patch I have built and run the new upgrade test on using Sun JDK +its a bug actually +Found bug in registerstack spilling so values were truncated fix will follow +Committed this together with + +patch closing all issues that have been in resolved state for months or more without any feedback or update +I would like to leave the simple version so maybe we can call that module and you can make a docbook module +Thanks! +I hope the patch attached file will solve this problem +All of the doc review comments on the Ref Guide that could be fixed for were implemented +Do you think this is a bug in Netbeans or a feature +I decided to leave this issue open to grab for now since this issue is mostly in the ANTLR generated code and I am still learning about ANTLRIt does seems that its recursively calling foreachcomplexstatement pretty excessively but I dont have enough knowledge to deduce the grammer so it still worksI did verify that the AST generated in and generated is exactly the same if that helps +For instancea system doing scans a second should likely not trying to be updating this value times +there is no schema change and systemVM code change between and only binary upgrade is neededwill provide upgrade path myself +Raymondcould u please help write a tiny test case against just AXIOM? then we can move it to JIRAthanks +It is now working it was null +I pulled this into +Hoping it can help tracing down the problem +Whether disallowing negative parameters makes sense can be discussed in another issue. +I apologize for the patch spam +Attachment has been added with description another is the current status of this bug patchAny thoughts on applying Peters code from comment Im still running and I havent found any mention of this being looked at or fixed in +This would cause an error without merging +Committed revision +Hi ElenaBug fixed at r please verify if it fixed as you expected +By checking our code I noticed that by convention our query parameter is not not just specific to this parameter +I think the use of the network server in a single but shaared machine is an interesting scenario to understand +The one wasnt supposed to get included +In normal delete one CF delete and other ones column delete and yet others version delete can be combinedJust mentioning the differences +after reading through the summaries Im almost sure that my problem is not mentioned in any of the reported issues +modifies suite adaptor to support changes in +similar error on antidote for this problem has been committed in +Maybe should not keep references to the maybe should use a new for each document ormaybe something completely different +I was using Scala iterators which returns true to contains only if the element is present and its in the same position +Trivial patch attached +There is only one problem with span wrapping minor fix +That would help us very much in moving forward with Google integration +Bigbank still has a stop without shutdown but this should be covered in a separate JIRA issue since the bigbank account JSE client does not seem to work at all anyway +In domain mode you can specify a plan that gives flexibility to have the changes commit for some servers even if it fails for others +I plan to change that all later but currently this works in most for now I will close the issue +This shouldnt be a matter of opinion or guess work it should be based on complete factsThen just ask for more detailed information before you close the issue please +You also know a lot about Axis +I think the issue can be closed +Moving the issue over to next rather than closing +httpXXXsvnprojecttags adding the explicit revision to copy from solves the problem svn copy file tmp r +This seems like fundamentally a flaw in the Flume interfacePerhaps someone familiar with NG can say whether this problem is solved with the new design +Note bogus user ids can also be introduced even with native authenticaion +Thanks for the complete test caseCorrected the termination tuple to the proper width but went further to completely disable lob tracking for continuous since its embedded only we should not need to track +Addressed some minor changes recommended by Francis +rightAnd I assume when the service release is out this will just workAnything we can do to catch the error and fail less abruptly? +Carl I ran that in my fresh checkout after the build so I should have been getting the default Hadoop version for a Hive development environment which needs to be something that works +The NPE regression is fixedAlso this sample has been moved to dassamples and is not yet automated +I thought you just opened a bug report feature requestIf I knew you are working on it I wouldnt touch it +Weve run into this as well +In this jira we should also change shallow and size in from long to int +Voting for this because applications built around libhdfs depend on the binaries being available to end users +Is this security loophole of having the file in deploy directory resolved in EAP or ASIs there a plan to work on a more secure JSR implementation? +It actually is a bug because we do not atomically write to the datafile and the metadata file +Otherwise Ill try to get around to using the info from the rails link provided +sounds likely to be correct behavior use the user are looking for the query cache please read the documentation and use the user forum. +Perhaps then someone else will be willing to try and run the release process? Claude is right that i should not be the only one prepared to do releases +the setWidth methods of and +rttransportshttposgi needs to be refactored I was actually trying to get rid of some duplications during the last couple of days for the changes to make into but I ended up seeing the demos I was working upon failing as a result so wed need a bit more timeId probably consider even removing it altogether and may be adding an OSGI Activator to the http transport code not sure yet how plausible it is +Modified the patch to close only if the out stream is not a standard stream +This is because the portlet mode buttons are inside the draggable div +Is it possible that the new patch causes that code to be exercised for the first timeYes thats exactly what happened +Any thoughts +Ill commit it later today barring objection +JBESBGAFP +Recommend trying against the SNAPSHOT build + lets deal with this in +This is file of +Okay Ill have a look the patch +files once they become provided sample rails app behaves much better and leaves significantly less files in Id say that rackimagemagick need to be fixed in order not to rely on GC and implicit closeunlink +but I cant easily separate the work at this pointThis patch puts all the generated metadata in a separate jar and makes optional at runtime +Checked JSR home page as of July and CR remains the last draft +reopenning for documentation team to get this into Seam reference guide +Ill try out your fix on Monday. +How to fix it on the database level hem not sure how to handle it +Hi Clebert thank you for your comment +Ill remove the other to avoid confusion +Agreed I think we need to clean up our namespace usage here + Added tests for the following scenarios +Whole SE for file from original file from previous modified +Fixed in branch need to determine if trunk needs fix and apply there +Earwin this is looking really nice at a glance +I suspect this restriction is provided for performance reasons +Patch for the issue +Spring IDEs label is truncated in the default view +This seems to be +I think you might just need a longer TIMEOUT for that awaitTermination +im having the same solution or work around +changing component from store to services +The project is private + Username sreich +basing things on S +I am using lucene via solr so it may need some solr extension to reproduce it +XML is removed from specification +dont create a site they are just installed into Maven and their contents are resolved later while consuming them +Lets close this out for now until this verification that this is still an issue or we get a test case that shows the failure. +Once it is time to read how does the reader gets a consistent view of the log? One JD alone may not have all entries so I suppose the reader may need to read from multiple to get a consistent view? Do the transaction identifiers establish the order of entries in the log? One quick note is that I dont see why a majority is required bk does not require a majorityWe decided on majority quorum to keep the design simple though it is strictly not necessary +The engineers are considering enforcing that the admin port not be used for jdbc queries +I have an issue when I use select a from Actor a where ? attached +Added in revision . +I think WSDL URI issues is fixed in Axis please download that and see +Will see that this does not the review commentsReg the ruby code i have changed it after reading some materials in ruby +It is the first time I use it +It is difficult to check diffs for new patches. +Would that work +Thats a big milestone +I dont know what it was in recent versions but it seems like there is no longer an issueIll do some more testing but in half an hour of poking I found no problems with cqlsh combo +Good thing you suggested a look into thatIm attaching a update to the dozer bundle so the nd patch isnt forgotten +Since I switched to RF the file locking issue is gone +Please test and verify the fix +I would like very much to see this feature included even before ships and Im willing to have started some time ago an attempt to integrate acegy with jcifs +You may not have a choice +Not sure why different tests failed in the previous two builds +Dan feel free to start working on this based on my repo if youd like. +As I mentioned before there is not much we can do here. +The lognet build file looks quite reasonable in this regard +else it is a normal eclipse tab not see the error every time when doing undo in the file i attempt to save in my last comment ive just noticed +I believe these items were completed +Simple fix that will fix the issue for this case +after a reboot and filesystem fixthe log informed me of used memory getting fullI do think its a problem that a consumer behaving badly can bring the entire messaging system down. +The reason is because the run method itself is executed in a privileged block +And even if it was randomly determined we store the selection in s cookie +create adds one file and the insert adds another filesorry i thought you are doing an insert overwrite can u do inserts? is needed so that the rcfiles in the target table are compressed with Bzip +I think everything else is complete though it should be reviewed goalsideaprojectideamoduleideaworkspace +To verify that the test fails without the patch I can simply add the test without applying any other modification in the patch file and then run the test +ah empty string how do I love thee +Reducing severity its only one special case in a component of CocoonHas anybody an Oracle DB to investigate on this issue +Thanks AnkitIndeed I did not notice it was not returning the name value when I was st asking for it some comments above +How does your property look like? Is it connected to a valid pathYou need to subscribe by sending an email to at +Thank you +The most likely situation is that the module containing this snippet is not being loaded +These look good to me +The max characters thing is directly from the lucene contrib highlighter +Why run the combiner? Isnt each word coming out of each map unique? It might be necessary to set explicitly to make sure the memory limit stays fixed even for different client configurations +Touches the following filesM javadrdaorgapachederbyimpldrdaM javaclientorgapachederbyclientnetM javaclientorgapachederbyclientnetM javaclientorgapachederbyclientnetM javaclientorgapachederbyclientam +Putting this on the list for this is due to the weird parser magic in respect to escaping and comments +Attached you see the views in the enterprise Signavio +Released which contains pluggable RNG validator. +Doesnt seem to be the exact same problem Im investigating +I guess we could run unixdos on some config files when building in BrewIt will be hard to control what people commit to the SVN source tree I guessBTW the problem is more generic as AS developers not using Linux will encounter the same on their SVN sandboxes +however Ill try to run this on Windows Windows it fails as described +In my case I have an email address being passed as a but the TLD is getting truncated +I am afraid I am not currently able to test it because my test case is based on our AS release +The OP may be OOL +The issue is occurring again +Fixed +So whenever you change it all threads in a given process see that change +This is just my opinion +Injecting a would mean that wed have to pick it from a field based on some naming convention which feels Thanks so much +Or encode the servername in the znode +Needs some bash script hackery or else confirm that the Windows executable we now ship works as expected in Cygwin +Bulk closing all of these resolved issues. +This is consistent with the comment I posted prior and I am sorry to say that I dont know enough about Solaris to know how to block ICMP +When no default dirfile modes are set for an archive and the OS! windows modes will be detected from the filesystem +Can we put this one in? It seems like a fairly straightfoward change +However any client trying to return rows to an application would be required to read process and buffer almost an entire of data before being able to return the first complete rowThe client is able to specify the maximum number of rows that should be returned in each fetch request +I hope Atlassian folks introduce required changes to account for these +Very interesting always keen to see how we can improve Velocities performanceWith regard to the alternative classes I believe supporting JDK +it cant be done in portlet as they cant acess URL +properties to the new unit to CVS thanks +Add a new Resolution column Add a new Resolution search criteria Extend the index on Status to include the security mechanism is now part of the search SQL query +When the message catalog cache grows to capacity it is reallocated and some threads may be holding references to locale objects that are being held in the buffer that was just deallocatedThis patch avoids that problem making the reallocated buffer hold pointers to message catalog data and never moving the actual catalog data around in memory +I took a closer look at this +Perhaps there is some other factor causing your failure +Im attaching a conversation about a potential approach but it doesnt really get into implementation details or what the API for the and functions would look like +jira errors when i press the attach files link +Forgot the css file in the zipped files. +This time I checked the contents and they look benign +Please be sure you run grails in a bit machine and java +Regards FreemanI have added the appropriate dependencies compiled and tested the archetype and generated maven against JDK +Porting the unit test over isnt trivial though + +reopened to correct the resolution this jira as rejected from done. + committed +But the issue has another problem +So violations appear progressively as soon as a source file containing some violation is open +I should have said it may return even if the end of file has not been reached +Neil thanks for coming back to me +Is it too little too late to ask why do we do profile activation this way? Are we using profiles in a way that Maven doesnt intend +Committed to TRUNK +Thanks for the good suggestions in the review Dmitriy! +First patch makes jruby behave more like MRI though on MRI if another thread has put server socket in accept mode gets will pause until client disconnects THEN throws ENOTCONN jruby throws it patch test added to should probably be refactored to remove possibility of future class cast exceptions. +This is the only way to reliably send failures back over the wire without requiring everyone a both ends to have the same set of all in perfect sync. +DennisIm running from what appears to be a clean build here and its all ok +We only run a full GC when the heap is full +Help requests go in the forums not release notes. +Patch adding the documentation for loading the native libraries using distributed cache +I added a new optional mojo which you could use if you wanted to +It waits around for the split but even if it goes past the max loops it doesnt actually ever fail any assertionsIm also fixing the test itself in my patch +Thanks a lot for so quick response I reviewed master branch and it seems that all is ok with there +Lets get this fixed for the next release +Agreed this might not make sense +Two deployable BPEL processes and the request message to trigger the error +Then we are still good with the approach where the client can maintain the current ledger sizeJust a hypothetical question does Bookkeeper make any assumptions that there is only one writer to a ledger? if so then what happens if a second client starts writing to the same ledger will it get an error +Alternatively you could store the delimiter touse in a member of and pass it in the constructor +We found that we have different users use their own versions of distcp and with this change we break their usecase +ill take it as Im in this code right now anyway +I dont understand the relationship between that deployment location and the repositories you just cited which are all on not +Patch that should fix the compilation error +Screen shots of manually created ws procedures +Run with mvn jettyrun and go to httplocalhosttest to trigger the errorAdded an interceptor before and after chain interceptor to demonstrate that this is the interceptor that does the actionError made a change please verify if its what you expected +My preference is to not logSync at all with lock heldNo disagreement there just pointing out that it might be a little difficult to do in the case of entering safemode and pointing out that it seems theres actually relatively few places that need to be addressed in order to get those tests passing +version +rcoli points out that commitlog is less straightforward +Verified on Hudson +I just committed this. +just committed this although I had to increase the heap size used for unit tests or else would get an error +am new to ofBiz what are the files to be upload in not completed +A question of configuration no actual I make this a request for the next version of the specification? I feel like this change would help greatly when using your framework with user interfaces +It is the same problem of the grammar is not switched when an anonymous complex type is being trunk of is working correct nowthx +The issue ERROR XT An internal error was identified by module is actually caused by another exception which can be seen earlier in the diff file ERROR XJ Java exception +Can you give details of the steps required to reproduce including the versions of Grails used to release the plugin and install it? its not although I dont think you can release to that address +Variations of the example above are using an uninitialised variable and leaving out the assignment in one branchThis should be added before in +Unfortunately that loads all the cores and in the case where there are a large number of them this will be unacceptableI propose a new core admin command LIST that will return a simple listing of the cores just the name + +Replaced my previous patch +And it removes more code than it adds! It looks like Dag handled most of the edges for the read uncommitted case in and with your change to ignore tdnull I think it should be safe in the listnull case tooI was able to reproduce the deadlock in on nearly every run with an insane build in my environment +Once your driver is ready just attach it to this issue and check the Grant license to ASF FGCP driver for inclusion added runoninstance operation +Thanks for the patch Himanshu. +This is on Cocoon with Tomcat +That one makes my life easier debugging javascriptIf no objections Ill commit the patch soon +Something like v +my bad I meant mvn grails not mvn + Username tfromm + resolved issues for already released versions +Thanks for applying this patch and closing this issue +Up until now there was no intention to deploy this by default not sure why the root page was changed to can certainly discuss this for +The rest was already taken care of +Yes I think its a good idea to check for too +For example a few places stats are not calculated completely like alter table +This feature has been implemented as part of the related issue and so the patch file names are the way they are. +I did not add new tests since this is reverting a previous change +Of the interface methods here are the ones that need are currently tested findAssociations tested by JAXRfind tested by JAXRJAXRfind tested by JAXRfindConcepts tested by JAXRfindOrganizations tested by JAXRfindServices tested by JAXRit looks like find and find are the two that need tests +waitReplication +Also I think any performance impact was probably more closely related to being able to test performance with statement logging etc turned on than a typical user scenario but I thought I would mention this bit of history in case performance of the new needs to be measured +We could also publish some of the test libraries to the repo benchmarks and the like so that its easier to run them against clusters you are just bringing up. +Probably very difficult to deternine automatically so might need to be provided perhaps with as default? +So it seems we have some issues to cleanup hereAs far as hangs I inspected your dump files and it seems they are from the ant jvm not the forked process that actually runs the tests? It could be the case here that its just in some test such as +It is ready for review + +Also just State instead of so the full form would be +Im going to start working on the fix +good request that is not by the patch +Moving to +Closing all old Resolved issues +The patch to has been moved to so that patch needs to be applied for this patch to work +What is your suggestion for that +to updatehibernatetoolshtmlsingleimagespluginshtmlsingleimagesplugins all images are +This is apparently fixed +I just wanted to ensure that whatever was in there was read and acted upon by someone which is the case so all is again for all your feedbacks! Thats really great. +Collections is already checked in as a dependency of Carrot clusteringThen we need to move it to coreJason +On futher thought support in SWF would be a nice way to avoid this +However there are other issues which I am still tracking down with this issue +Are you setup to debug and catch the NPE at? Would be good to know the npe cause and more details +My bad I had not applied patch which is required for this patch +Thanks! +it should be available and +Changed file appears in latest alpha docs so closing issue. +Please attach a sample AndreasThanks for your attention +A typo corrected +Updated with different locking around the post commitoptimize handlers +Tested it and its all good now. +Hopefully in the future mlockall or Java will error out a bit nicer +Same as +It could then go fetch another batch +Updated with Alejandros comments +Less is moreDo the above add up to a on v of the patch +Closing +This is for both and trunk +Basic is EAGER by default in the spec BTW +Thatd be nice +Cheers +I am reviewing the patch with Ming +Committed. +solving a null pointer issue in +Which right now works but failed like minutes morning Andrea +Upstream issue. +Babak use the mailing list for discussing your build issueThis JIRA ticket is about the mvn eclipseeclipse goal failing due to an Eclipse bug +so probably publicMutationOperation seems internalDitto comes out in API? If not should be private Id say +bq +Reopening as latest comment implies its not fixed +Revision . +Cancelling and resubmitting the patch +Now when choosing to analyze by field typename a drop down with all possible typesnames will be populatedIn order to support all functionality of the the co +Ive pushed them to. +Managed to fix the memory leak is fixed our thread pool is not a real thread pool +It was not mergeable anymore so I merged by hand and +You can fix it by including xercesImpl as a dependency for the project +should we make Mutable a default for Entity +not a b to to trunk and maint +Committed to trunk +Can we this issue please +Also a that does same as +reopen if more clarification is needed. +Thanks. +Then people will be free to choose what they want to use +That bug was fixed in the commits that followed a few seconds after this oneI work in a git clone of the subversion repository so I commit a lot more often than one would do with subversion along +You need to put it in toolsjava +which might lead to some more traction. +Once is resolved we can work together on a implementation of it +BTW you could just use plain Configurable Autowired will work in any case +Test cases added and passed to reflect the new logic. +This mostly answers my question +Patch for problems file for problems file +I think it should be okayAgree +An ivy file describing this dependency? No I didnt but Ill try it now +Missed one more spot to use the new constants +Since thats whats in the index I think it would be less confusingFWIWm guessing there is no way to display the time the log logged? If thats not possible my enterprise voice says UTC and my general user voice says local time +This was with the assumption that it was only printable characters only on the key +BrianI have done testing with the and ran every Scenario related to this issue to find the memory leak +I guess we can do what you described +Thanks to Paul Rivera for provide us this patch +Changed iterator method to return iterator for if no problems exist +Duplicates exceptions have changed due to some weld changes +Firefox will not show xml have to do view sourceI suggest an option to create a zip file and download +Please review and if it looks ok close the issue. +Enhance example configuration to be able to run a mailing list search service revision Thanks Preetam +bq +Closed upon release of version +Username user +James could you look at this one as Stefan is away +I was able to test the fix and it looks like it is working +Thanks +It was a simple fix of not giving the managed dependency map to the artifact resolution +I was therefor very surprised to find out that maven release does not support it +Three days ago I added an example for this behaviour +bulk close of all resolved issues in preparation for release. +Those vm files are used as a default template when they are not overriden by a PAWhat I found that in pom the resources path was customized not using the default maven resources pathSo IMO the resources path customization in pom should be removed and all resource files should be moved into srcmainresources +Changed the fix version and tried this sometime back in December and it worked +I have put the rev number in the comment so you can find the commit log and the unit test etc +Personally I prefer not to install eggs since it complicates viewing the source and renders pydoc unusable +Adding Applied patch +Im assuming that it might be an problem Jay could you please test in AS and let me know if it still happens +Fixed in ddcddffbcdfaabef thanks for the patch and thanks for the iterations! +and it is still the same. +Attaching a the patch +Fixed +I dont see such behavior for visual are you verified on exact version of JBT +Hi AndreasApplied your patch +m sorry +This has been ported to trunk +makes the capacity a float +This code stream is outdated and will not be revived +Hello I work for Sun on the Java VM and tripped across your report and decided to investigate the crash +If we dont publish the servlet on the root we cant add or remove the http consumer dynamically without stopping the server +We can use reflections to get the Codec related class paths too +I made some tweaks to bring this feature inline w the latest Solder API +This commit broke the job because is called when the job gets created thus replaces the ZK addresses from the source cluster with the target cluster +thanks +the exceptions you see in the error log will also happen without specifying additional VM arguments it is due to Forge not correctly removing the previous launch +Hardly critical +for the patchOne observation is that we are not specifically setting the LOG file in the TT via l option and so all the log statements are directed to stdout which are duly ignored +Please verify +The patch fixes the issue +Coded in and work in pages without encoding tooI think language files with appendix should be removed from because cant be used in wwdatapicer tag and are incorrect +Test succeeds if my patch is appliedPavol Ill be testing your patch too +iml files that sets this I think Ill post a patch shortly that puts the project SDK inheritance back and should hopefully address the problem youre seeing +This patch is against revision the patch thx for the patch +Br! The extension has been modified to properly handleWhere can I find the patch or how will this fix be delivered? I can only find fileBr issue is fixed by adding an entity resolver to the extension +yes I can consistently reproduce it +This was fixed by Ivan via the get temp files workaround +So you may understand that I am not eager to add a new is very well possible that due to a programming error a will occur +Is it going to be seam based +But does it crash with any other feature or special char? Perhaps some existing regex stuff that I dont know about +I pushed a release of which removes the workaround +branch is very stable now and is being prepared for a release as +bootstrap targets more or less +It stops quickly and estimates how many might remain +Thanks for this Sebastian see that you had modified the patch and cleaned up my codeThere are unused imports in that I had introduced and should have been taken out could you take care of that. +Thank you TonyVerified using. +So you dontneed to set it manually any more +Thanks for the patch Edward. +User that requested feature also did some testing so I think this can be resolved +Im preparing a pull request today so maybe we can use that as we upgrade +For the time being I just added an extra check for enclosing quote marks in order to avoid of cookies that contain Netscape style expiry attributeOleg +I appreciate your time in working with me +Hello Simonewould you mind including this in versionRegards to me weve got a new problem when both bars are rendered the sidebar is missing the links right +was created for the installer issue +Build log captured using ant verbose +Fixed in r +Unfortunately jxpath doesnt really allow you to plugin your own implementations for specific classes so I think you would have to make your own version of jxpath or thought of a very complex but clever way to solve this issue but while I started working on that I actually found a much simpler solutionBasically jxpath allows you to iterate over pointers rather than values that match anWhat I did is change the jxpath implementation of the complex feature attribute pointer to return values rather than attribute objects per default +Manually tested +Thanks. +Showing missing icons. +Yes I know it can be dangerous in our case it would be a blessing though. +If you modify the eclipse preference without updating the descriptor by hand any attempt to run the app on a server will result in a +Committed in revision +Committed to trunk and. +to commit +Agreed that youll want to drop stand alone since its included in Hive +This would have some overhead but less than the overhead of passing and decrypting a ticket per RPC +in you had said I havent been able to reproduce any sort of deadlock using latest trunk and the provided tests in +Some of the mismatches were due to inconsistent badPackage exclusions between the tested versions +Stephen you hit the nail on the head its about VPN in all our cases +I added a AIOOBE to all the match methods to make the interface happy +We should disable csharp rather than aborting the build if gacutil is not installed +Aether is now at +Fixed by +Corrected in SVN rev . +Note with you need to enable session failover otherwise youll experience something similarThe mechanism used to replicate the state in a cluster rely on servlet spec +Just change the timeout from to minutes +Barry this is just a catch all for the JBDS PRD weve been working on +In that case they could be shared among different modulesc +As commented earlier the snapshot on does fix the problem +Actually thats how I found thisDue to the improper handling of assertion failures this test was passing before +feature request not blocking usage +Steve right +well it doesnt look like these bundles are going to be released in the next two weeks so youre going to be using timestamped snapshots of something from Sling either way +This patch hides AWTSwing issues +We can go ahead and commit this for now until our deployment stuff clears out +I like Vijays proposal +Added Seam Users Forum reference +Jeremiahcan you provide repro steps then? There dont seem to be any steps to follow to cause this described in the bug +I ran the checkIndex on the partially generated index folder + +Alfred did you implement it in Pivot? If so you could submit it as a patch in this ticket. +Closing as a WONTFIX +I will take a closer look when I get some time +Without that change ptest has no way of passing a profile to mvn +This then causes the Spring test framework to the context for each test which is needed since we added a drop of the database before each test runs which in turn causes Spring Data to recreate the indices +This Groovy file manifests the issue the Grab annotation for can be changed between and with the latter throwing the exceptions +A patch for the same +fixed. +If we will decide to make whole messages compressed current changes wont make that hard to do +That could be an issue if your application cant guarantee that it will park its databases gracefully before the VM exits +in the meantime ill get to work putting the final touches on t be the first time How about we agree that we are in bro love and yes we will need to chat in person which is fine bc well bring the decisions back to the list and to Review Board and stuff +My local history shows that thischanged right as I checked stuff in +Would it work if a domain class first does the bootstrapping as done in srcgroovy and then adds its own dynamic stuff? In that way a user would perceive a domain class as something extra instead of something different from a srcgroovy are there obstacles doing it like that +lgors patch looks good to me but it got stale +will be registered automatically only when ruby UDF is used same mechanism as jython UDF +Whoa +So is parsed and the endless loop beginsAn interesting conundrum +And your server doesnt seem to have Internet access to load the schema file from the why are you running against the Oracle XML parser there? I thought you were running on WAS? Please remove any or the like that you may have in your classpath +The class is part of the original code drop all other Utils I wrote +Commit unit tests apparently does not fail and I thus conclude that this is an isolated local problem and not a problem with Tuscany +Bad patch +Its broken down into two now as well +Hey thanks +For better or worse thats the reason we have them to specify dependencies. +After a week to ten days they responded that the issue was resolved in a recent releaseRescanning a recent build reveals no threats and no files moved to quarantine +Purely random as you can see in the trace its some opensubmitsearch sequence +The patch has been appliedPlease verify +Asokan would you please review the documentation for correctness? THX +SaschaCould you update your patch please I cant apply itpatching file frameworkexamplewidgetexamplepatching file frameworkcommonwidgetpatch malformed patch at line Index frameworkwidgettemplates Jacques yea the patch is outdated +Either it forwards a kv to a peer or does not +Thank you +Konstantin pointed out that this will be an incompatible change in that were one specifying localhost as the ndCP Nodes address and relying on that value to be resolved on the node +This issue closed by revision . +this patch finishes the job and uses logj of WARNstderr for bin tools so that doesnt get in the way of redirecting stdout +Ive tested this with an initially empty Maven repository locally and it seems to work but please let me know if you continue to have trouble +Willem could you please not start working on this just yet +Patch updating the ant docs both on trunk and branch with the patch +Getting NPE during retry while trying to create a new listener for a new temp reply queue +AFAIK the soa ones are temporary +Pushed to trunk and branches +Its probably related to the changes in Beta which I fixed using inGlad to hear that your problem went away with the final code +That would be cleaner + for my English +Please Chris validate the patch i think its ok G +The last release that passes this test is +Committed to trunk +Submit patch +I committed this +The servlet unit tests have been run with AS +Moving to since has already been tagged +If you look at how the are actually persisted it gets each type separately +The behavior that Dave is seeing is what we were seeing as well + +Patch with proposed corresponding methods for get are get and for get are get +If theres a snapshot youd like us to try let me know +But sorry to insist its just that Im trying to convert my schemas away from SCF and so Im doing manually what this patch does automatically + + +Shouldnt the fix version be set to for that issue? +I have no authority to do any workflow operations +Evgeny could you also take care to remove this rule from the embedded SQALE mapping located in the? Thanks +fixed for +Thanks Gunther and Navis for this cool optimization! +i am sorry got confused its reported in the correct category +Thanks +No its not yet possible +I encourage you to contribute the code to ODFDOM +revision all old Resolved issues +I believe this problem was fixed in that build +Patch file +Closing tickets +Thanks Remus! +Committed to. +Yes it seems like faceting should be for +Robert I missed your last comment and have committed v +Closing as duplicate superceded by the application of. +Committed a fix for this +Could you just please explain the reason you think this implementation would be wrong by design? +Closing this issue since ds has not responded to verify the fix and it is over a yearsince the fix went out with Xalan +Issue may already be resolved may be invalid or may never be fixed +This is a duplicate ofThere are tests for local and repl mode in the package +Ivy cache is polluted with missing dependency information from grailsHome. +I will check it in when I find a chance to update it to Trails +from my side once the patch is rebased +In +Im a little confused +Shane can we do this +Please attach a project that can reproduce this problem +Thanks for the workaround RonIve looked into this issue a little it appears to be some strange side effect of using ant instead of forrest +done +Duplicate of +Hi Uwe thanks for the vote of confidence and reviews and help +Actually it seems always to be the case +Everywhere else thats Planner code +OK +Part of to trunk port +could be an issue with an older version of form file or any case form IN is working for me in netbeans marking as fixed in +Reopening as the zip file in the Aperture issue is still causing problems for Tika +Ive updated to match and there is no error +I dont know in which issue I therefore set this to Out of Date +However this bug suggests there may be other area where weve missed issues with implicit casts +Thanks Kim +for the patch and for +I think we also want to be able to easily locate select and use near mirrors without manual t that solved in the mean time? +Changes look good +One of my disks failed however it was available on the system as read only +As noted in the comments this is not a defect but is actually allowed by the JCR specifications. +This can be regarded as a bug in the unit test +we can ignore this for this release and make sure they are resolved for the next version of seam +implementations of without parameter types +Yes we still see the issueOK Ill open a separate issue to try to get to the bottom of this +as a feature request +User can opt not to create a secondary staging store if on exists already from migration. +but it looks to be right around the corner now +To do this the process tree class keeps track of an age of the process which is how many time the process tree has seen a process with this PID +Changed Target Version to upon release of +This issue can be closed. +Ill ask to add IBM and let you know to create a new environment on Jenkins. +Ive committed this + +Post questions to list +Knowing what binaries were pulled to assemble the update site has been handy in the past though I wonder who if anyone still uses that +Um sorry +Pull request was accepted +hadoopqa issues are ok rat is complaining about the new jarschecksums which is not a real issue +Sorry it took us so much time to look after that ve just applied a lot for submitting the fix and for reminding us to apply should be available in. +Is this something that can work with the Avro KeyValue Schema +created a dns entry for moved svn from under incubator to the top level +ms seems insignificant as compared to the total runtime of the reducer +Offhand I dont know where to put the package docs for maven now +Thanks for the response makes sense as it could be considered a feature at this point +Fixed on and closing all issues that have been in resolved state for months or more without any feedback or update +These kind of environment setup issues are exactly what the test was meant to catch and it looks like they have +Hi Dhruba +ChunhuiI agreeThats why I think creating a new exception is acceptableMaybe Todd or Stack has better idea +Attaching current output with XSLTC +Need to resolve before we can move forward with this issue +What can happen is that the watcher list for the cold topic wont be cleaned up for a long time +In any case Ive sorted out a similar NPE as part of and I have made the COT code more resilient and more informative in terms of what goes wrong when the reader externalizer returns null +But this is related not just to this dialog box +Ive assigned this issue to you + +I DO like air on the side of caution though +The patch for in has to be applied first +I attach to replace the above removing an unnecessary checked exception +I forgot to update +I dont understand why this is unresolved +For someone who uses RVM for Ruby development or Virtual Environment for Python they would want to be able to install the language binding to their virtual environments location but not the whole Proton codebase there +I would like to see this method add if we can get it +Before the patch join column did not pick up the id column length and defaulted to therefore this message did not happened +I think this might be because the does not support bcc or maybe it does but the server only receives one message +Joda works fine. +Yes doesnt need to apply to you on the latest branch before applying the patch? It needs both r and this one to work +Hello PDid you resolve this issue? If so could you please tell me how to dealt with itThanks +So I decided to just test for Enums in instead +defer all X issues to +Using a consensus brings us into the realm of using a Dynamo like architecture +It should also block root as the user + +bulk defer to defer to +If this attribute is set to true these methods are fired instead at the end of the apply request values phase +blackdrag I assume this can be closed as Wont Fix +Sounds fine +The patch looks goodI am attaching the patch for new hdfs trunk +I have not committed docs in the past so dont have the setup yet +If you could provide an externally usable pom that reproduces the problem then I can take a look +A test would be great! Thanks +put a vanilla in SOURCES +fifth screen after hitting save parts of +Theres a bit more cleanup to be done and this hasnt been through much testing but Im headed on vacation next week and I think Devaraj wanted to take this over +using classes where applicableFound r mpollack supergroup +Not the absense of the Note the absense of the form +Well we hope to be able to release a new version in the next two weeks +Deleting gshellsrcmainresourcesorgapacheservicemixkernelgshelladminetcDeleting testingsupportsrcmainresourcesorgapacheservicemixkerneltestingsupportCommitted revision . +For ongoing development this is less desirable as you would have to run the service after every change that you wanted to see the results +Really good is hard but I think a simple solution like above can get us a lot of benefit for not much work +Committed thanks Namit +Stefan Do you know when Jackrabbit will be released? Or could you provide a patch for the bug in +Were there any specific components you had in mind +The order of the resolved artifacts is predictable? Without this patch I have just did several times some resolve and each time I had a different order +Fixed and code checked into develop branch +very much Rick +Patch to class to fix the session access issue +WSDL for the service in questionWas this resolved on the mailing list +jars and tested it in another application +since jukka said that he will address those concerns in another patch ill wait for the next patch +A null check is broader and might mask other issues. +Patch updated to trunk +However I still believe that this should probably be an abstract class that can be configurable in a common way +Fix committed to trunk as of applied as of r to branches +Tested with both gshell and command line deployer that the confusing message is removedbin linsun +attached a different approach +Will it find all Persons by its UUID +OK greatJuergen +Mark asked me to sort it out +Best yeah I started thinking about it more and realized that there may be some cases where developers may want to flush the child and and not the parent and really the developer just has to make sure not to burn himself by not having child changes dependent on parent changes in that situation +Patch against trunk +For kicks queuing up again +Latest patches for review +I updated the patches +added the as well as modulesxsd with a testmodulestestorgapacheaxisschemacompilethanksdims +I will be happy to have your code review +Thanks a lot Simone i will go through this +unit tests all pass +But there may be problems withSELECT o FROM AS oat Castor QL this refers to the class Order while it is invalid at EJB QL +Daniel is this fine with you too +Especially when the SQL Standard Authorization is not enabled this list is not required at all +Have you tried the parameter +Test m not able to reproduce this issue +to Hudson to get another run of the tests just for completeness since the original run has expired +OuchThanks for reporting and patching this +I have added a link to this issue in +Adrian said that its already in the Framework +keepStrange +Like in this case how is max attempts set? Can it be set per queue? Can it be changed at runtime? Can a unique deadletter queue be specified per delivery queue? +Can you possibly recreate it without the formatting changes +Since the discoverer will have a complete list of all the artifactsmetadata in the repository we can call the indexer and index the artifactmetadata as we iterate through this list +I think the right course of actions here is to +The summary references release but Kamikaze seems to be up to now +This is a two parameter closure which is called only at the end of a command line run of Gant +Ill open a forum posting so we can clear it up there +Jitendra noticed I missed the SNN part of the patch which is actually blocked by several other patches +Ill deploy to the old repo until these Gradle issues can be resolved. +Im raising the priority of this as if affects Wildfly where passivation is used in our default configurations +That will give me a big head startYes it would be very helpful to be consistent with the authiduserrole terminology +Please feel free to brainstorm +Moving to for now as not sure how much additional work we want to do around vs focusing on BPMN might depend how quickly we can get the BPMN tooling up and running +Closing as addresses this issue + +Removed limitation on debug build only. +As Maxim pointed out we merged a different fix for this issue in. +This issue seems to be solved +see The patch there fixes the naming issue +I can now get on with experiments as it is not required to checkout the CLI source and build the jar myself +Atmosphere integration is available for months now and it still has no docs. +Jon since youve been pretty deep in async lately perhaps you have some comments here +there is a ton of Java bits in activeio which is usingI missed this before as I was still compiling with JDK but set the compiler to generate classes +Committed r trunk rr close for +This patch fixes the issue using the approach suggested by Daryn it outputs configure test code that makes use of the library being detected for both zlib and snappy +I think these were as intended as they were +That sounds promising. +Then would be more flexible since you could select more exceptions to reject +Kishan has already started working on this Discussed with Kishan +Im not against of course just would prefer some indirection via a Monitor +It also makes it possible for administrators to make a special continuum account in proxies and configure that without giving access to all deveopers. +Please verify and then resolve and close +Fixed a bug in since it didnt write the correlation id as part of serializing the request +We dont use Enclosure in our app so I cant test nothing more than Wickets tests and examples +this is message ordering in action it is expected behavior +I think the problem here might be that youre not using the +This fixes the problem +PMD rules added to +Please look at new version YC design +Patch removes Cloenable from +Thisappears to only be the Display value and not the effective value since I am still able to start +Patch to Adrian as he wrote the initial just fixed the wrong config assumption +I dropped it in the patch to followAlso on the response structure we certainly could make it optional although it means having to go do a lookup in the real doc list which could be less than funPatch to StanislawIm going to commit soon and I was wondering if Carrot has a handy place where they keep all the licenses and notices so that I can fill out Solrs and +Ill upload the new bundles in a few hours Ill post a comment when Ive done it +However Benson recently added support for flattened arrays through the aegis xml config +Proposed patch +Update patch to trunkI also updated the tests to use the new connection kwargs in the driver args +Maybe we should mark command line tools as special classes in some way in order to give them more freedom +The Hibernate session is closed before the jBPM context +Full disclosure Im not a UUID expert eitherYoure also probably right that we can remove the subtraction and division from the end +This is the cost to pay for the build speed up ant filters out unchanged sources to skip unnecessary dependency recalculation by cpptasks +Patch implements the above mentioned changes +It should be set to whatever size the four images are +META +Take +Verified on revision +For this I am using the CPMThe CPM should be stopped after the first exception occurred to showit to the user +Everything seems to work fine again. +I see newcomers struggle to build the containers every week and I just dont want that to be their first memory of ArquillianTeaching the developer about the features in the build system is something thats more of a goal + +verified that enforce macro only called on controls with set. +Getting logj warnings just because you start using a commons component seems wrong +Rebased on trunk +Thanks for taking a look ram ted stack +Patch applied +Definitely agree on the testing but since this is a rearrangement of code its not clear how to write many more functional tests since we already have some coverage +Actually in your log paste it never starts any +Closing old resolved issues +we will of course release it +bq +Manually tested +I refactored and called this because it really is a helper for the +OK here it is handled the way you suggested +to branch. +We need to simplify this what if the user just want a seam projectIt is related to that is fixed +See comments in +The patch is created against revision +Fixed in head revision + +Can you raise a JBESB issue for this? Ill get this fixed before goes out +They demand some respond from the server to work correctly +committed +Job is till passingPlease look into the patch and provide your comments +Applied to +richcomponentControl is missing attribute event in event was added. +Very old issue +Also converted the into a single class +Joe which Mahout formatter do you use? It the comments and should not +The change to rely on provided infinispan module in EAP was done due issue in ISPN. +Thanks for looking at the patch Knut +Reopening in order to reclassify +Then we get two alternating Sonar reports and its difficult to find out where the wrong report comes from +Oops comments +Its only a blocker on principle +IMO a better patch should bind the xaresource lifecylcle to the tx only if the session was logged out inside the tx +You can just load it up into the Axis WAR and you will see that the operations dont declare any of the Faults though the faults are now being defined as WSDL types +I wish I were able to edit comments +Added some log statements to +is the class that is ambiguous would it be possible to return the cause of the error +I just committed this +I read your comment wrong +but the one would lock up +Its ugly +The code looks goodThe mechanism in specific and I guess will look for factories registered in the Spring contextWorks for meOut of curiosity how do you make it work standalone without Spring +fixes that +I meant of course there is a better way of doing it no doubt about it. +Let us know if this fixes your problem +HelloWould you please try this patch? Thank a lotBest regardsThanks a lot The fix looks good +Oops that is not an the previous two of yet I havent gotten to work with +Please retest your and with a test case if it has not been. +And besides this custom class that I will be making isnt annotation too +Its OK to remove thatI didnt do it on the old JIRA cause it was marked Closed +Im currently using H to implement a matching engine which takes a series of events and matches aggregated information held in an database in real time +Confirmed +or for confirmation that my patch to the cured the problem +Thanks for this patch applied in r. +Looks fine to me +May be i should have named the option use instead of use but this is a different story +I will change the plugin to take only a single file to be in sync with findbugs +This issue has been fixed and released as part of release +Fix verified +The spec does +Hi FolksMy teams tests are setup exactly as Geoffrey describes above +Hi AndreasFrom my point of view provided by you not typical for most users +Hansall this looks really greatIts nice to hear about the captcha featureFor point could we think about dynamic menu widgets +What have you configured the test to useyour test doesnt just take a random directory and uses it as the runtime right? Dont you use some system property or similar to look it up +will not be implemented based on discussion +I would recommend reporting this to the slfj project +patch attached +Sending srcmainjavaorgapacheservicemixbeansupportSending srctestjavaorgapacheservicemixbeanTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixbeansupportSending srctestjavaorgapacheservicemixbeanTransmitting file dataCommitted revision +The specs are more up to date and do more than the TestUnit tests +If you want to leave it open past the beta release let me know and Ill put it in more versions +The versionID in changes from L to L +Ill have to let you do it +I think its a bit too esoteric at this point +Lesser priority now +Ive added a single queue and has not changed any defaluts +Ive added support to the pluginGen tool in the RHQ git master branch +When I unzip epp currently it explodes to +It works fine for you please include this patch into upcoming version +no logj api access in the source code +Was committed to but issue was not resolved. +Note the patch was created at the root level of the sling project +Fixed in release version after merge +We are now integrated with WTP Eclipse in our latest nightlies and the upcoming release +The problem was that the was blocking on a read when there was still SSL data to process in buffers +I think this is no longer relevant since we have more or less have given up on the catalog tree view +Done in Rev +Thanks Sascha and BilginSascha your patch is in trunk at r +And they can push all of this down as a domain policy template +The test creates a dictionary from a corpus and evaluate a created with this dictionary using the same corpus +Yes it is great if it is configurable +If something is wrong with their server configuration whether its the hostname the port the username or the password if they just ignore it they wont be able to deploy properly and wont be able to shut the server down properly +Applied patch in trunk. +I verified that this always happens when writing data larger than KbAlso it comes from the directWrite as I was able to find out from the stack trace +Its like freeing memory on the same pointer multiple times and wondering afterwards why your application crashes +Yes +Heres a patch to update Velocity to +Commenting the method fixes the am thinking of using manual autowiring with as a workaround for now. +Does Chukwa need to be group hadoop? The user hadoop wont exist on most agent nodes I would expect +Yes March should be fine +Ideally Id want both definitions to be used with one overriding the other +Thanks Im marking the issue as resolved. +Please see the screenshot for FF the popups width is just as wide as necessary and not up to the right border as it is when using can you please attach WAR file? I cannot reproduce the bug in FF in both M and CR releases +this sounds like your database was corrupted +Ask for helpIm not familiar with work on document +So the remove older asm +The applied patch seems to have taken care of failures +postponed to valid in +They are checked into the enhancements branch for release +Reviewed and committed +Again we use diffs to represent changes from a local checkout and the svn repository +It has nothing to do with the container though the fix suggested by Stephan works and should be included in the trunk +If locales are messed up it may use a lowercase letter from another alphabet believe we are using which does localization by default +resolving it as invalid as the pig rpms being used were wrong causing the need to change the order +Ive made some progress +That is why I changed it to beWe still need a trivial additional test that positively verifies that a bookmarkable page does NOT expire with but with the same page after session expiry +Problem still occurs in that problem is not Chrome specific +This problem is preventing me from connecting to a serverAs Galo says WSDL is usually protected on the servers and this will not change any time soonI am surprised this is considered low priority +This causes NPE when doing a custom sort in distributed search +Which version of CXF are you usingI think we fixed some issue of Jetty TCCL with few month ago +Thanks again for the patch and the reviews +We have that information about the database owner on other CREATE statement pages +Let me know if I can help in any wayAnother thing I was considering while reading your patch is that it would be nice if the messages went through the same code path regardless of whether the NN is in standby or active mode +Looks good +An issue related to this one now also supports AS +If theyre tied in with the resource manager then thats another reason we need a new resource manager for each pearWe may have to wait for Michael to come back from vacation +Its calling signal not kill +The explorer view of framework I am not kidding +However Im wondering if an example configuration might serve better in the short term +The expected output will never contain a licence header +If a given unencrypted system table is known to begin with a certain pattern of bytes this gives an attacker some useful information in cracking the cipher. +Can we change it to allow identifiers that start with digit as well + +It has to do with how Oracle deals with empty strings being considered null +Delete the extra definition and add +Thanks Can you tell me how can I solve this issue? I am also having the same issue that in eclipse for jdlk I am getting is not visible +After they are recovered coordinator does not know about the transaction so the participant has to be rolled back +Corrupted PNG image file afterwards +Next iteration +Committed to trunk +b from CF where KEY foo and KEY barIts not possible to specify the key in the return +for and should check br and ba on all datanodes in the pipeline +Can you please send a patch for that change as well or a patch which covers both +Kabir Please confirm that the problem is gone and Ill close this issue +Looking through the test do we have test tags for local vs shared storage? If so we can create separate tests for them and define the offerings to match +Here is an idea to handle this problemOn standby for rbw replicas we do nothing and just put then in a queue when received in block report +gives us the ability to cherry pick what we understand and are willing to maintain +ping +This patch modifies gold files and to reflect correct output for corresponding testcases +Yes but that doesnt mean that setenv should not receive the default values for thisAlso consider that setenv really should receive KARAFHOME as this path cannot be overridenIn addition for setenv to do any real work it might need JAVAHOME might want to check for cygwintruefalsemight need to know JVMVENDOR so it can set some XX +the readme was updated some time ago too +For now on one will need to specify the exact address to use which is much more correct in my opinion particularly on production systems that may rely on IP aliasing +So Im scheduling this for for the time being +Please check that it was applied as expected +looks cleaner Sachin. +Is you test case reproducible testable? A unit test would be the bestUsing synchronizedMap is inefficient +Also another sanity check is good to put in so we dont set sort order if sort columns are not set +old and according to our excludes from MRI s tests we are not excluding pack tests from the given file +There is a lot of confusing index flipping logic there do you mind adding some docs to explain that? For example its still unclear whether the second replica is assigned to an adjacent broker compared to first the logic is a bit cryptic +Attached patch contains test case and fix on trunk +Also our sub projects will use the feature +would it be easier to just use svn export to get a clean tree for the source release +I added your great suggestion to the troubleshooting section of the groovysh doco on the wiki. +I think that my doubts are related to the fact that now we can get a valid instance of Directory which points to an invalid filesystem path one which perhaps cannot be created at all and we are going to discover this only when opening with createtrue flag +for the patch +though it does not really improve much +that would be better +Applied to trunk tests added to trunk pending push to +Test case reattached from bugzilla +both files in the svn preferred patch format +Yep its the JDK executor and the typical reason is the Queue is full +This patch has the code change in the new place +Yes you are right +Uhh scrap that last comment +but we faced that issue adding the import scope +was not able to reproduce issue with version please see the attached testcase +Ive added a null check for counters to fix this +I have never looked at this code before +Committed revision +I think that is part of the integration though I dont think it is right to put register a customized the initial part of the summary of the Struts the integration consists of A plugin The refactoring of +Derby will not be able to register and that is not a problem +But I think we should implement the simplest case first and then go for the complex scenariosI would like to know where this component will reside in Airavata architecture? Will this be a activity in a workflow or workflow activity some daemon which handles provisioning of cloud resources and job? Im bias towards the later approach but I would like to know what is the most natural way when compared with the other Airavata components +I found that if I renamed the files in the following projects so they are ignored the build succeededapplicationsdaytraderconfigsconfigsI seem to remember there was some mention of moving these projects out of the Geronimo build + +Moving out of since Andrew marked it trivial +Ill rethink this design and try to resolve the bug +Does that work? Did I understand your comment correctlyYes that solves the problem but then requires that the recovery protocol open up and read the edits out of the logs +I assume from your comment that we need to switch them over to use the hadoop variants +Taking back +It seems that encoding is ignored although Im on a mvn versionMaven version version name linux version arch i Family unix +resurrecting an old patch from February +I will also have version of the patch +refine summary and beautify description +Committed to and trunk +Not a bug report raise a feature request +From OlivierHi KishoreThanks for your email and explanationsOthersAs Im involved in Helix incubator my idea was to use it to a syncmechanism with Archiva for artifactsBasically a primarymaster instance where users deploy their artifactsthen n slave instances for reading artifactsDoes it make sense to you +Hi Heiko +Im going to remove my support for this request since it seems to be more pain then its worth +I think we need a bigger JIRA to figure out memory restrictions when we have these multiple optimizations in playPlease review +Thanks Zheng +now supplies the proper contexts and objects to the +Its not clear to me how that query should behave and so Id say its unspecified or if we want to get nasty we could throw a HTTP back +When is supposed to be released? We shouldnt upgrade to a Beta level when we are within weeks of trying to release +This is due to +You can make it happen when you add something like the class definition to a node already on the screen +I get trying to receive response even after putting up a master first and while trying to rolling restart a to regionserver the offloading works but then the onloading failsWithout this patch I can rolling restart from to smoothly +If the codebase moves to being the core JDK it will surface again. +It looks like the problem is when compress asks for a checkpoint while another checkpoint is in progress +I found the cause for this this process is part of a larger project i found out that some other module in the application has configured a listening to the errorChannel +Fixed with revision +Im not sure how entities like this are usually handled +Can we describe the problem users faced before describing how it happened in detailThanks for adding that +Using fixes the issue +which will do the same as what Run As +Many thanks! +Thanks Christian +As this is does not seem very useful atm but there is a different jira for this is now +And I ran the full test suite successfullySo I committed patch to the trunk as revision Thank you for following up on this to improve the behavior +Committed patch b to trunk with revision +Id like to close this as seems to address this issue +I missed that files can be named +Downgrading this to a critical as this is more of a feature and can go into the release after +I think I can perhaps put them all in a properties file +BUT a file lock must be aquiredWe need to check how the performance impacts are with closing the fileHowever I think I would prefer to close the file at first glance +Cheers! I had already written something which prettifies the output however itwasnt very good +Im going to be cutting a release candidate soon +unit test cases to verify code to match expectations +Can you create the patch from the latest in branch? I already applied your previous patch and the last one has some of the same changes so Im not able to apply it +If nobody objects I will commit this to trunk tomorrow think we should target for this? Ie commit to trunk let it back ship backport +Issue is not reproducible in +I get the same error in IE version series is dead can you try and see if it happens there as well? and trunk are the only maintained branches of m seeing the same error in + +moving inbound HTTP issues to M need to explore the new code along with etc +Thanks for the review Arpit! + +Thanks Brian! +I just went back and double checked the How to Build the Website Wiki page and I must have skipped step I will get the site stuff fixed committed back into the right location and the site soon +Yes however that information is not passed into +Perhaps we can have Get and Update metrics +Are you observing this in the cluster + +AwesomeTwo comments +DB for Linux setup scripts for testing. +Here are the guidelines +Separate this jira into two part +I think this was some fluke race condition +Is this an issue? Or can this be closed +Changing the is bets done in a separate patchYes I plan to create another issue for removing which was introduced in +I dont think Derby supports recursive queries +What happens if you install bubbling manually then install normally +rough code containing ideas on how thread state changes can be an updated file +build hadoop with the above patch applied and was able to successfully make an open call on a secure cluster +Just an idea +gencore jira project setup using standard configuration +Looks good! I think that solved it +From notes of AS meeting +Is the leak fixed? I was already getting OOM errors but I wouldnt think just creating a lot of objects would cause an OOM wondering why doesnt GC clean up these objects? +Please reschedule if necessary +I guess I should go back to bed. +Does java need additional libs from these required by branch +This appears to work with latest Eclipse so targeting Westport release +Nicolas The block size in the DFS usually will be set quite large lets say M +Patch applied +In that case should we just enhance Marketing component instead +Can I actually take it that you are not using the Velocity mode of the XML code generator +Seems to be a problem of testbuidenviroment +Does anybody could give me the patch +So dont expect it in near future +Committed to trunk +A working workaround is to obtain the Subject in a way and invoke repository operations inside a passed to the methodThis is however not portable +I can write a unit test for the dataset +See for a patch that solves both issues +Bug has been marked as a duplicate of this bug +Can you add some unit and QA tests before merging with the trunk? +But Andre Dietisheim tried to reproduce it and he wasnt able to reproduce it either so resolving +This bean should not rely on JNDI or use the JMS API for talking to the JBM in the trunk +The POM file need not include artifacts for and as the existing entry pulls these in as its dependencies +The POM is now fixed and the file system connector is fixed as well +and will write localhost to znode for master location which remote client cant useAFAIK we dont force localhost in there what we do force is looking up localhost for the zookeeper server by default and even then a remote client could just specify the right address and it should work +Closing as cannot reproduce since we dont have enough info to reproduce at this point +I can see that making sense in many cases but not in others depending on who is using the system and how much redundancy they have setup +The size of the idFields and idTypes arrays if specified must be the same +committing now +Looking at it right now +I think we should fail fast instead of trying to find a region server supports that codec +Duplicate of +Review got ship it +That is the correct procedure right? My version now is it would be in which has not been released yet +Moving out of until we see it again +How can I get to the findbug warningserrors generated in this build + +Seems like a pretty big bug yet it is still open after almost months +noway this patch goes into jboss tools source codeCentral code should know nothing about SOA or any other specific site beyond its main siteAny magic property replacement is to happen somewhere else not hereI also do not grok what the patch is actually trying to do it seems to just add a method with a url and nothing else +All unit tests passed +It could be done although on a more simplistic form that whats in the client the failover would need to purge any pending messages and transaction state info for the client and resubscribe for all the active consumers +You should have a look at the available documentation of Woodstox or post this question on their mailing list +Thank you for looking into this most aggravating and elusive issue +This is the only safe way to do it +Which does save memory in practice because a node will be connected by many clients that will usually all prepare the same set of queries +I tested the test under the build but NOT under autotools +Now we are in a different phase +Fair enough regarding HTMLXHTMLThe bug report thing is one of the primary motivators the other is in our own use when supporting clientsRegarding security the only place I can think of where something private would be ni the daemon confs would be S access keys +I will make sure the changes are committed soon and the object mode +A implementation issue +ant test passed on my just committed this +Will work on that trivial +Re Maybe we should not publish to the updatesite if there have been any build problems +I somehow never noticed that the escape characters didnt correspond to their ASCII equivalents. +Now retrying the compaction + +but there are few issues with the scenario editor oh and Travis KILLER user name +Im not sure how to coax it to use them +So the issue is reopened +It is more belong to additional information is needed category +looks good will merge if the tests +Hi ThilinaThanks for your replyi have also one problem i want to change a code of client and servicein samplein such a way so that client can send a message to service so for thispurpose which method should i usedThank YouPrashant. +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +Sounds good to me! +Should I add documentation there tooThanx +The attached prototype has adopted +Okay looks like I had some kind of inconsistency in my project somewhere +Closing old resolved issues +Krasimir is your project big or you have the problem even with one newly created kick start project in workspace +Thanks Ill be checking this ASAP +or just downlevel the for your plugin so it depends on instead ofIve done the latter since its like keystrokes and seconds instead of an hour or twos work +good if the original job name will be displayed as ansounds like an excellent idea +I follow Stacks reasoning on this so we can do the patch +Is it possible to add a unit test for this +Wow respect excellent analysis of the problem! I dont think I would have been able to find itIve committed a fix in SVN trunk +closing issues for released versions +jar? I would have thyought the changed package names are going to break the build +Looking good now so added thanks. +As of Teiid Designer we no longer create XSD to Relational models this code path no longer exists. +And attaching a way to reproduce the problem would be best. +Fixed thanks! +Since you are of a jacobin bent you could remove in and just configure it off in +Also need a flat structure without modeled ID fields +They would be used in addition to the existing line separators +In other words seems not worked +The svn one is color coded and expanded the git one is not color coded and compact to the point that its quite difficult to understand whats changedI thought the nice color coding in svn was due to fisheye +Doesnt look like this is going to make and is probably too late to change this so pushing to I think where we left on irc last was well fix jsvc +Do you have a unit test that demonstrates object retention? If so we can add it to the test suite with a fix +bq +The patch ported to! I mean the patch ported to +If you have the entire log please attach +updated with findbugs warning and the test failures are not related to this +Could you please create a new diff against the latest snapshot off the SVN trunk? against checked in +Great job try with the suggested encoding +Reopening for at revision +Here are the three joined files describing the problem and my investigation I investigated furthermore this problem and I found what was causing the troubleThe mistake was in my code I was generating string content that contained characters +pls +If you have a solution then please send a patch +classpath file for that +Merged both GH and patch. +That could be solved in simple ways instead of adding complexity to this layer +Close as fixed +New method +Patch applies but that lone if line at doesnt build or make sense over here +Could you please explain how I can change the logging in such a way that is no longer logged but all other areWith my limited knowledge of logj I could change the level of this class to FATAL so that ERROR wont be logged anymore +Hi folks sorry for the delay in responding +This could be fixed outside of jUDDI +Should be in the next SNAPSHOT build +Ill do that as soon as I think the task works properly with resolve scopingAfter that Ill have a look at your patch +Eclipse Kepler R releasing this week even if we provide fix it is not going to be accepted until SRI am moving to streams +is there any code i can steal from the renderer which will analyze an SLD get all the filters out of it and use that to query the datastore? I know that its hard to do in the general case but we should be able to hit hte case no +We have been using this patch successfully since reporting the issue +This is something for a future release +Ill wait for the tests to apply this +So my previous comment is not validJust want to add that is testing the message printed out by Network server at startup and shutdown +Attach new patch based on mentioned above the findbugs is from the and unrelated to this change +Thanks that will work just fineKenneth +Not a major issue +Re +Well be checking this in as a separate client module once is in +Looking back at your original post asking about using the sed test harness capability if you go that route +Can you check passing jobconf xy xy throws an exception? Can you also verify if it is easy to split the values from streaming and add them +Is this correctThanks it looks like someone else found this before I did today couldnt figure out why my memory was creeping and creeping during a long batch process even though I was clearing the sessionSanity check Im running on right now and all I did to get around this today was call clear on the hashmaps in when the is cleared +Actual problem not a validation issue +Patches to and trunk adding support for Parascale storage as the DFS +oops new patch without + +Product name Enterprise Application Platform If space is limited use EAP here is a patch with some minor updates to the content based on what we previously had in EAP can you apply this to the ballroom git repo +Im marking this as WONTFIX because in we are no longer going to have a JPA connector. +Will wait a little however to see if someone fixes it +JonIs there you can send me that I can test on my end? Thanks! with progress on this? Thanks may be related to a larger issue with the bridge +httplocalhostcatalogcontrolproductIdThe screen does not work in I was going to say that everything looks good for me latest svn +of the launchpadbundles be possible to update it via a bundle fragment though right +Updated the EAP kit so that the security related files are placed in the servers configuration folder +Youre saying that when I use cordova build android from the command prompt it copies everything but doesnt generate an apk that takes into account the content src in the xml file +is this because you did a dry run and no releasecleanThats a bit weird since it should do a clean checkout on releaseperform +I will take a look at it later today + +Hence closing the bug. +Yeah seconds is a long time +I think Matthew means Heres the codehope it helps! If you have a chance adding options for and macros would be great! TIA +Patch addressing some of Mahadevs comments +I have attached a file for the master without information +The test case is pointing at the numpunct facet caching code +The API documentation should be more specific here maybe even containing an example +Hi ChamikaraIm sorry I guess I was a bit sloppy I folded in some cleanup at the same time +Before DN report the finalized block Clients first completeFile call reached NN and marked Block as COMPLETE but lease not removed since minReplication not satisfied +Please feel to reopen with a test case if its still brokenRobert +But now that youve raised the issue Im concerned too +Are you on linux? Maybe it is a macos thing +This does not seem to happen with recent releases over Windows. +This is one of the reasons why we need to move away of popups +Sorry I have been completely consumed by construction +When initializing throwing if history is present and it is not a directory +Should I just close itThanks Ole +jarfile +Is this something that goes away with the advent of +was trying to implement in a quick dirty way the createdeleterestoreclone rollback on master restart +confirmed that patch is not a fix the objects need to write to other cluster members may still unable to writewe need patch with a real m going to move this out to a target unless someone is interested to work on this right now +htaccessr +The implementation classes are renamed to and +Applied patch to CVS. +Netty is apparently even worse as it hangs the connection and I cant even out of it in telnet +We need to move to the lowest common denominator for the jars like. +Obviously there is a lot of work to do with this effort +Perform VM Manual live migration for one of user Vms in this networkAfter live migration we still see the the rules relating to this Vm being deleted in the old host and reprogrammed in the new host to which it is movedClosing this issue. +No in some time. +I have written a small java program to test that I can load different version of derby classes using a classloader +for the possibility with Validate +I guess its mostly working in bigtop already but just the individual loggers are using the direct calls to the old jcl libraries +Use the forum for more info +I have attached the pmd plugin class which will create and outputpmd data without adding it as a report in the following formatsText XML CSV Emacs Html YAHTML Papari +For jira issues please user Jira project +for the patchbq +This is somewhat manageble if you have separate applications and you can choose your name but it gets really nasty when you have two applications Foo and on the same AS +My concern was the instance would be across multiple sstables therefore should be chosen per buffer +I have not seen this defined anywhere but my gut feeling is that aa should be counted as two characters +Will verify in next build +This patch fixes the failed oiv related unit tests +the patch looks good +was not meant to be the file I intended as it contains some content which needlessly overlaps with a previous patch +Afterall XML is slow in geral +With new api the framework no more calculates this +I have attached patch for enhancing Billing Account have added Find and List forms for Billing AccountsTesting steps are as follows Go to Accounting main screen httpslocalhostaccountingcontrolmain +Let me seek some feedback from MS guys whether this approach is acceptable +in srctest +What permission do I need or how can I do it? +I had to make a few minor modifications to the and renamed the to so please sync any other affected sources +please test again +Please see my earlier comment. +Did you try to rebuild the project? Does it help +Ill try to fit it in to my schedule but I cant promise when. +On further thought you are right that it may be useful to conditionally use on the output stream based on whether an exception has already thrown +Hi VINAYAKUMARThanks for posting a patch +Trunk should support now HTTP and SOCKS proxies with authentication +A missing dependency should not result in a NPE but rather an appropriate error message +One last commentAfter stepping through the method I noticed a wrong user name chaul instead of chaulGoing back through the steps it is there already in the wizardSee Christian Haul Aug AM edited above CVS location ispserverchaulCVSABPRJCDEFGHIJ checkout wizard shows cvs and pserverchaulCVSABPRJCDEFGHIJWhen I manually remove the from the SCM URL it does work fineSo the real problem is the broken URL +Yes it was sorry for the confusionYes the intent is to not package the examples +I had the same problem in +peteI just talked to one of our kernel guys and he isnt sure as he hasnt done that much IO stuff on Linux but thinks the K readahead may just be the maximum +Now we could extend the sitemap source to allow adding removing or changing of parameters +Ill have to check on the conflict resolution mechanism to make sure its youre right the conflict resolution isnt happening because of the groupId change +The problem is the streaming of the file is it correct? You can set the header after the first PDF byte has been sentI dont like the idea of doing this only for PDF +Thanks Allen +Ive only done a few small benchmarks at the Store level and havent benchmarked a whole cluster +There are things to clean up to make the code look a little nicer +from what someone told me on an email its not as bad as you guys though +I will try to update the patch also with this fix +Undo +Committed to the branch as revision Marking the issue as resolved +Fixed by +but that should definitly not affect this caseI was just about to write that you must have written a bad reverseengineeringstrategy impl to make it put the classname there until i just saw that were calling simple instead of get +I guess it was originally intended to only work for binding rightIf so Im not if it will be possible at all to make this work for binding +Should have a more user guide feel than the component reference +Custom adapter is created indirectly via + +The issue can be the bug +My kind apologizes +What JVMplatform are you running on that this is a problem +Fixed on trunk +HiGoogle should not be used as the first tool for you should do is to read the supported outputformats with WFS +Reviewed by Ted +Updated to trunkTicks schema version to defaults to omitTf except for +Hi Paulex Second patch works fine for me +If I only create the tru trigger everything works fine +core tests +QE please check if the issue is still resolved with redesign in +Sorry. +Drop the attached jar into modulessystemlayerssoaorgswitchyardmain +on this idea +I am kicking off the full test suite to run tonight +fixed again! +Ignore that pull request was working in wrong issue +Are you able to check whats in your? Do you have in or homexxxpig +But shouldnt an array of one look different than an individual object? should i report this farther upstream somewhere +Fixed by Move report of split to master OFF the heartbeat channel +It also adds the same warning to the main functions page +Because with a hash map there is no guarantee what order entires will occur in +Hi Jonathan I think the attached patch covers all the previous concerns except the timeout to be in micros +Just one commentThe javadoc comment on test should not refer to embedded but to mutable SCO or Date field +I will look into it after the releaseIn order to print tags or other features next to tokens there must be more spacebetween the lines or between the tokens I am not sure if that is possible with the currenttext editor +You mean is picking it up in the ndnext pass as it thinks it has changed hence it itIf this is the case can you please debug a bit further on why it thinks it has changed +Yes makes sense to deprecate but leave the push functionality for a whileRegarding that was on purpose to avoid using multiple JT threads for this +Any plans to get this fixed +This is also beneficial for someone that is migrating from dbms to another +Thanks for the review Francis Alan! + +Attaching a patch that fixes the problem and a test case +Ive spent time on this in the beginning and I dont think a delta protocol is appropriate particularly since the RM and AM arent in sync everYes this could lead to some waste but the system is eventually consistent +Phil could you review please +Im all for making things look nicer +I think the current one is a hack +patch file for branch attachedThe github didnt work for me +Im not explaining wellWhen Im TYPING the full command the apollo freezes +Im using Firefox and Firefox +So its actually set at a point that understands the various versions +It is the messages for next two blank fields show passed +Closing M issues. +Alternatively smokeTester could look for this +Yes SAX is deprecated but the proposed patch makes backwards incompatible changes to the API +The shell command has been fixed but the should be updated too +this patch moved +I committed the related class in +Thanks Andrei! +Unfortunately I get a lot of rejects off a clean trunk +Wed need a proposed policy for the project to consider in terms of major and minor releases specifying forward andor backward compatibility guarantees +Need to discuss what the expected outcome would be in this case. +Thanks Shelly Im marking this as resolved +RyanId rather have this in the default mojos done in such a way that we have a deployed inside our plugin make the template parameter optional in the mojos +this one is a flaky test which intermittently fails +bq +Otherwise another exception is thrown trying to get a test connection +I am just running a build with this patch +Verified in SOA. +Fixed by way of. +Thats what I was thinking this bug is not really a bug but a misuse of Enclosure? I think we can close this one +Should we have the statements? Checked for localhost and hostnames on my network +Heh +Attached patch a +I also noticed that the value in state is case sensitive +Can you clarify why is not meeting your needs? It supports using locationsWe need the configuration to be from many different places so I need to use as Bruno pointed out +Approved for inclusion in +There is nothing that forces you to use chainingSee also I brought all arguments + +How you tell they hung? You run a script against console output +When this is occuring the time to build my project triples from mins to about mins +The zip distro is from which is built with jdk not jdk +actually and all minorResolved in commits with r r r +not fully consistent with everything else in eclipse but neither were the previous iconset +Hi CharithThank you very much for applying Message Requencing Processor patch +bulk close of all resolved issues. +Heres a fixed that grabs the required jars using the ant get task +The test case I wrote to try to reproduce this issue is +I hope this patch solve the bug +I dont love the idea of bringing the classes in and would rather just use the methods available in if possible +Difficult to catch for me +The patch can detect the reference and serviceIf the reference or the service of composite is not compatible with the promoted component refercnce or servicethe runtime may throw an +I personally prefer not to have IDE specific files in the SVNthanksThilina +Also in we should get rid of those get methods and use the val directly +I have not seen this problem anymore since using Karaf and Servicemix like bugs fixing themselves accidentally Thanks for reviewing this again geert +Heres a patch with the code for the new based +I have the code for the latter approach available on request +heres a patch for the work in definitely struggled with modifying this mojo +Sorry its that big but almost every file had to be touched in order to implement successful mapping +Obviously we have a instance as a in memory configuration but we dont have a file when working with +Wait +Note that this is hardly uncommon +Thanks be solved as part of related issues. +This method is called to fill the combo + +Thanks +Updating description to match the fix +Patch +Caused by the fix for which has been rolled back that issue reopened +Please find the updated log attached +For some reason my web browser always creates jira issues twice cancelling this duplicate +Im not quite clear on the requirement here +Sotretus Dont confuse status with statistics +how long it takes to balance N regions etc +Added component +Backing out for now to see if build starts passing again +to Wesley +Committed revision . +Other than that the table is completely functional +In general we recommend direct use of and over use of for and +Verified on JBDS. +Would avoid quite a bit of a mess in the hem +regarding listreplicatedtabless output I think it would be better to group by CLUSTER else on clusters with hundreds of tables youd have a pretty huge list +I added a fix that will correct this issue in most cases +Ive extracted the common get file from srctestresources via classloader logic and used that all over the place +So I suggest to close this issue +Twitter probably doesnt consider this to be a bug though +The namespace quota is quite specific to HDFSIs it? I dont have any specific counterexamples but Id be surprised if other file systems didnt have a similar concept given that are usually a finite resourcebq +drop it into wicketstuff and share it that way. +expected output IF output actual output resulting from rendering the IF into cannot reproduce this issue with the latest version from trunk +Ive noticed a few more irregularities on but Ill hold off filing issues until your change from today gets released + +Since fixes on these issues would require an update on file Ill just put the fixes on patch to avoid confusion +Patch for trunk +Can you please post the file used to run the test? Im curious what Hadoop settings you used +c files of wsdl with wsdl +srv passed +Forgot to include in previous patch is staleThe long build times are a problem and ivys a big part of that but I agree with your assessment this is a hack +thats fine the TCK will mostly get executed on its ownMaybe we should create a release profile where we can include it +Im on ubuntu +More thought required +Punting this to you david +Links created for ajaxListener calendar dragListener +So back to the data loss issue and the current solution do you still have any comment +Contents uploaded to docsen and is liveWaiting to change the links until were sure that is done +IMO adding this might help some other also +Sure thing Ill create a banner for the blog +Ive tested this out and it appears to be working as expected +I did both features Enter opens Edit dialog Click on a cell activates cell editor +This should update the class to set user dir as root by that the fix matches the SFTP provider. +Is it ok just to patch the existing or should I create a new test class +This is not a trivial change and requires a lot of testing with the core code plus some investigation to make it work with projects that dont want to use ARC +In general sounds to be a valid request the additional characters will need to be checked for compatibility with the SASL and HTTP authentication mechanisms usedHowever based on current priorities this is not one likely to get attention soon so have scheduled to Awaiting Volunteers +It also patches so that INC and DEC properties are protected and settable from the child +No response assuming ok. +I need to look into how this can be either better supported at the tomcat layer or by a custom request +I am passing in the null must be happening INSIDE the toolkit +Here is the command line I use to run ee testsant Usersdaijydevetchadoop Usersdaijydevbinhadoop Usersdaijydev Before that I set PIGCLASSPATH to include tez libraries +Most of the differences are in documentation + +Just to add not sure how valuable it is to run on Windows given that we actually use +Recently I am proposing to introduce some configurations mechanism for the deployment process then we could define a tag based on it the user might only need to set a tag in the deployment plan or configure it on the admin console if they would like to use their own JSF api and impl +How about systemconsoleconfigor systemconsoleconfigThe Web Console core supports this and I think the config plugin does it to IIRC +Im on the Mojarra announcements list +Committed to trunk and +Ping +If you feel like it is the first step would be to log the exception +Again fixing documentation errors and enhancing documentation of existing components is likely to far outweigh the amount of documentation that occurs for new features +One question Is a NULL key allowed in the API? It seems that s close does accepts a zero length keyAhad +Please review this patch +So while we could probably start using ranges we would need to use closed white lists to indicate which versions are supported +IIRC it will take about minutes to finish +Fixes to docs page fair point I was trying to merge and mainly becaues has the stuff for resizing right on smaller screens and doesnt +I suggest a patch and relative Junit Test +However if I am adding a collection it would seem I am appending the value to an already existing setThat sounds like a good approach +In of all usages OWB and are not used in environments +re attaching the same patch but with a different name +hello graemei tried your sample app +Otherwise this is fixed +Looks like this should be close +However the functionality went in for Is there any testing scheduled by anyone +Edgar reported this presumably due to his testing Sling and not being able to get the mapping for the root node +Added a unit test for checking token in last blockUnit test for READ access in get is difficult because this method gets the access mode from the authenticated +Using resource array instead of decl +This feature is long pending and will be great to get it included in +Thanks Nigel. +Only unschedule volatile jobs on shutdown to by default let the jobs be kept in the databasetrunk +to being restrictive +If i resubmit also it will fail +If you can verify that the time delay needs to be longer we can increase it I dont have an FC box let alone a slower one to test that on +I think the tag Expert conveys the fact that users are not supposed to play with it +Thanks! +Im pretty sure this got fixed. +regtriggering hudson on latest patch could not see output from last run +Forgot to grant ASF license +I know often users aredoing the interrupt to stop a thread +Can you regenerate it based on trunk please +Jesse before I close my pull request I was curious to see what the code change was to resolve this +And the benefit that we would get for this is not worth the effort we would need to maintain the patch from here until eternityI would be fine with copying the code into. +FYI this also happens on and EAP because class loading may depend on the way in which we inject the subsystem +So if this is still an issue we need a test case +Hi AlexActually we decided not to use scans for this type of query +patch against Jochen for the patch +Can you post us e the document link +I just committed this +Perhaps cross that bridge when we come to it +Unit tests and passed on my local machine +I understand that for MR we will still not entirely have achieved this ideal because the client still needs to know where the MR jars are +The Roobot server fails to connect +Hi NeroCould you try copying the jar to usrlibeclipseplugins instead? Thanks +I will keep investigating but not sure about this yet +Btw for the consumer clients Im using the sample clients that is shipped with the distro +No takers +I like your solution with different coertions for primitives types and object types and everyone will be happy +So I dont check size at all and just catch EOF when opening the reader +I have the following jars in jettys libext +fixes the patch implemented in rev +Ok your chart shows that write throughput is roughly constant with the number of followers as you increase the number of learners of the system +It doesnt seem like it would decreaseperformance considerably as most of the paths through this function wouldreturn after it found the locale already loaded +Small patch +if my patch is acceptedWhat are the next steps? +In fact Nacho is currently working on the mavenization of itselfAs Nacho is currently enjoying holidays well wait until he returns +This could be related to +Good find lads +Changing this to an enhancement request +Tests from the original problem file updated to fit in Jelly source patch +ok +I just picked the colors that are available readily in Java Color classAs for the expense of generating it it depends on how complex the workflow is +Please try this patch +Committed as r. +something like this +Verified on JBT. +I would expect the producer to give up after a configurable amount of time or number of retriesThis also causes issues if you are running within Fuse ESB as the reconnect thread cannot be killed. +A test case to reproduce your problem will be helpful +Are you okay with relinquishing your claim to the copyright? Why did you decide to get rid of the filters implementing the ordered interface and opt for inserting them in the configuration? Was there a technical reason for this? What was the reason for adding the additional parameters property to the consumer token? Im not opposed to it I was just curious if there was a technical might have more questions but I wanted to field these first +Maybe just keep the first sentence and move the extra details to the symptoms sectionTypo in incompatibilities section can no cannot or perhaps can no longerIn Application Changes Required section to detect when use of numbers outside the DB limits were used +as promised by the Eclipse WTP team we should get these improvements in Eclipse WTPI am providing all affected source files as well in the +The example was as copied from the sf wikiI incorporated bens comments I kept the example but replaced references to yca with references to the foo scheme +I think the space navigator portlet should be hidden because it is designed only for the top header areaI will find a way to hide this portletPlease leave a comment if theres any better idea +Well I already consider that situation +Since we have and way to list all bundles easily now + is committed +As we discussed on the mailing list things can happen in an odd order in an open source project +The X flag to basically just prepends Djruby +Fixed in Git ID bfebeaaebdddbab +Committed revision a small patch to fix the Javadoc failure +We set zhfd to after the close in handleerror and when we create a socket it should get a new fd no +Thanks for the suggestion this did the trick for xalan and coWhat about the geronimo deps? Id rather not include them in the classpath at all as they might conflict with the Sun versions of the same but AFAIK theres no way to tell Maven that say is the same as we use the geronimo jars at least in part because they are of known provenance and license +This is a known maven eclipse plugin issue +JonJimmyThis has got something to do with? We had an internal patch for version where we would try to get the blocklocations from NN based on retriesIf with in the retry i could find the block locations then the will be non emtpy +Thanks Sergey! +I guessed it was a delay but not two weeks +for fixing and applying the patch! The changes look good to me. +Thanks ArpitPS I added a javadoc to the new tt param for during the commit +Robert patch applied with thanks! Snapshot deployed +is the culprit of the problem +That said im fine providing snapshots builds with changes but just send out the email when the suggested TP changes and which changes are coming dont move the used TPC by the build up to this immediately +Thanks for the thorough and detailed explanation it was quite clear and helpful! this statement worked until broke it so the correct fix is to make it work againAgreed +patch for for available +Since the scope component was only a proof of concept and some of its mechanisms will go into the new component done by mario I will pull this one I see no use in having it in there anymoreThe scopeshop demo can be remodeled for saveState in fact it would be an excellent demo on how to use saveStateAnd the save and restore mechanisms explored by this component already are recycled by Mario conceptwise for its own approachThere is no need to keep that component anymore and its associated phase listener +If you have a patch for this issue please attach it and it. +this seems already committed to cvs +wrong custom categories we end up REPLACING existing categories instead of ADDING to them back to the drawing board sadly +Further updates to the plugin manager and activator as well as plugin and pluginfactory to include reference to constants although these still do not work +Applied the patch by Kasun +Why not just allow https as an acceptable url for archetypeCatalog +how do i close this one? thanks +bulk defer of unresolved bugs to +This allows deployment specific packages to install a custom alternative with a higher priority +The error node is only intended to show if the connection has worked and expansion of the node fails +I think we need to port to +That is where the big win will be +Applied to trunk as r. +Sorry I wasnt thinking clearly when I made my previous comment +I agree with you +If I had to set a feature to get the behavior then so be vote would be to set the default behavior to what I mentioned above +Could you please supply unit test code and a documentation patch +I believe this is no longer an issue on latest releases of Hadoop and latest distributions +It is possible to do the same for CXF and other frameworks but nobody has written this code so far +Feel free to switch reschedule if you feel appropriate +No need to submit a new patch +Dima not sure Im following you +Does it happen with the release +The problem is that it tries to check the view server permission using the servers platform ID +Fixed in trunk thanks for reporting. +I guess we probably need to support only additional excludes + +Hi TedCan you send me your complete failure list +Does something need to be pulled before it ends up in the master branch? +The file is missing from the archive +o Hooked edit notifier actions for Project groupo fixed navigationo moved around Notifier form validators to approrpiate for some reason on submitting the AddEdit notifier form multiple times the errors keep on displaying incrementally +i think that would help alot done its in the war which is probably where it should have been +solution could be investigated it would surely be very efficient but it would also be extremely dependent on the fact that pop would be called on the same component instance and Im not sure that its always the caseThere are some weird pushes for which I didnt find the pop yet so I would keep a stack collection for now but we should certainly reevaluate the private variable option once were more comfortable with the spec and can identify pops for every pushes +Actually the problem is that the General tab is missingThe Main Config tab never had an id and name field in the first place +As a continuation of net module would you be so kind to show me resources where I can find out which modules exist nowIt would be great to know that to define list of modules for internationalization and to update resources on time when new modules are about to come +Appreciate your attention to detail re the code guidelines makes it much easier to review +Infact that is alrady an IOE and method also throws IOE +to Davids +fix BK compile add test +Hi I have fixed the issue regarding outbound rules +Postponed to is no added value in using MINA state Machine for decoders. +Changes are +There might be some unforeseen issues with other test providers using the filterclasses method but I doubt it +Tests ran cleanly for me on Java +I dont think its related to the itself +Alright Ive committed this and opened as a potential follow up. +Im going to contact the HSQLDB development team and try to figure out why what Im doing is not working +resolved at r +I believe you are referring to the escape character but not sureCheers it was a biggest issue than that +Also this was the behavior in +Dave Want to try what the lads suggest? Change config and restart master +Ill put the other fixes in it as well +I tried a fix for this myself +one line was still wrong +should be fixed in trunk +sure and Id be happy to lend a hand here +Screenshot with preview of hgraphicImage with attributes name and is this expected to work or new feature +I am using Seam Best regards Anna +Alexey please verify +Backwards incompatible introducing more problems of little value +Im not familiar with running those tests for does someone know whether this is normal and can be ignoredThanks for picking up this work +But I will change it so that the order is unchanged +The problem is that no library awailable for x eclipse platform for osx +The error is handled gracefully the only missing part is that the is not reported in Explorer nor in the list of embedded cartridges when you edit these +Ive applied it to the trunk. +But if the application is killed immediately Ctrl C then shutdown hook can call close +Suresh it looks like the changes in have created the conflict on trunk +Switching all issues that have been in resolved state for more than one month without further comments to closed status +AFAIK you cannot switch off the Knuth algorithm or fallback to a simpler algorithm +since the edit log has a new op layout version should be updated +Complete patch after using the move script and incremental patch +Considering the limitation I would lean towards a Category B classification +Forgot to mention that it was fixed in revision +Help in resolution is appreciated +Thank you so much for fixing this Andrew! I was hoping you would have the knowhow to do that +Perhaps the wording could be make sure that you properly include all configuration settings necessary for a basic faces configuration and check the logging output for any exceptionsAn exception on initialisation was my problem a colleague insisting that was needed in the because the error message was telling him so +I dont see any mention of quiet anywhere in the hibernate plugin code +No other changes +The installer itself uses already kind of a queue its differently implemented though +good to me thanks checked in. +BULK EDIT These issues are open to be picked up +The question I have is why is this considered invalid on Windows +I can also reproduce this at Mac OS x with Groovy Version and JVM m assigning this issue as I think it has enough votes and is serious enough to warrant looking d also have to handle +We are facing the similar issue in our application with axix +I have reviewed and approve Sarahs patch +Thanks for your review Andrea +Forward interface declaration support should now be complete for all types +right thanks for reopening +Also removed Order Ecommerce and Party +Sample EJB Eclise for the example we will investigate the issue and try to find a solution +Thanks for taking a look Mark can you commit the changes for me +Nice piece of work by the way +I have run a couple of performance tests with the test client in +It is working fine but this parameter is included as mandatory feature +rebased corrected typo +Committed to trunk. +No Unit tests till is resolved +Im setting this issue to blocking because the benchmark Ive created fails with an NPE +Moving to because Im pretty sure this is just a problem with the ANU machine not a general breakage of +This is a bit of a problem with Hibernates class It stores context such as the SQL String but doesnt include it in the exception message +Patch is applied at +The test failures were due to Too many open files +Thanks Sarvesh! +Interesting change +These changes look good to me +actually another big change was before that +Still not critical if you hit this again +Perhaps let the programmer specify a flag to the compiler whichis consulted to determine whether to emit a brief or expansive error message +My patch for is waiting for this commitThanks +Will get something in that covers all of this for +Weld which do hence make my stream unavailable for that is a different issue hence closing this one. +It seems unreasonable that javaWSDL cannothandle class if there are no getset methodsI propose this as an enhancement +ThanksBumping up the priority since this is a problem for default installs on single node clusters +This may be well known problem in in in typography +However this JIRA is about adding in a c option not just for META scan but also to run any other series of separated commands without having to echo and pipe it into the shell. +The only full task remaining is implementation of the cut job on the affinity matrix +Do you need me to attach it again +looks good to me +I was talking about a geotools patch that I introduced to make it use the custom property syntax above +It looks like auto preCommit test is still down +TedI understand that the bug report has a lot of information but it does answer those questions +has a sub module depends on another sub module the scope of the dependency is default and many other sub modules depend onAfter the scope of was changed to import when compiling maven failed to resolve dependency for those modules which depend on many packagesclasses not found error were threw outDo you have any idea with this issue +Its not easy fix because need to cut down on config +The entire case of the BPEL situation really messes up any possibility for a uniform UI +Verified and fixed +polling by different hadoop clients can be avoided +This should be fixed now and you can install the feature withfeatureinstall a change that should fix it +I found it in the XML output +Add Graph function +There is no foot violence by having to manage multiple configsThat being said this patch would not affect the current way Cassandra runs +This is the correct patchIgnore latest one WDYT? +Patch moving bundle event handling to the Loader class and fixing the bundle filter for setup in the a modified version of Felixs patch in r +Note that there is a clean when is generated +Yeah sureIve got the conf tests passing tonight +The problem is due to the duplicate columns that exist in the following tablesDuplicate Column jcrpath in table ntaddressDuplicate Column jcrname in table ntchildDuplicate Column jcrname in table ntpropertyDefinitionThese duplicates are being exposed by the jdbc driver when used to import the tables to create the model and +I am opening this issue because there is another patch that needs to be committed +Thank you Ravi +adding a responder delegate interface that is invoked during request handshake but before request processingThat sounds like a separate issue +I verified just by replacing the Richfaces bits in the app thus the Seam version remained the same +Please log an issue but we need a repeatable that will result in the phantom annotation +trunk contains the corrected code +this pathc addesses the issues pouinted out by benben can you take a look again +Hi Matt could you check that your issue doesnt relate to? +Same patch but moved the CHANGES entry under a Changes in Runtime Behavior looks good ready for a commit +EAP CP has not been tested to the the cluster always remains intact +Lets table this sort of thing until later and let some usage dictate the need for these mappings +Yeah I should be able to do that by tomorrowThe reloadConf calls can pass the name up to a bigger class and it can pass back the real filename after its done the lookup and checked modification for the code to useMight be able to unduplicate the in the same way +Should I close this as Wont fix +We are also seeing the same behavior and the issue is resolved on restart + file to ad toThere is already a health check functionality seems like this might be combined? Or instead you might have an external script which does this same check and sets or removes the health check file +Chunhui I am not convinced either honestlySince the failure is so hard to reproduce I can only wait until the next failure now +Bear in mind that the install scripts should produce the configuration that you want +looks to branch thanks for the reviews Eli and for the good ideas Chao. +Here the files +Our hack for utilizing the interface wasnt cutting it and the TSR implementation that Marc provided got us around a couple of the problems +its a squash of the last two plus one from todayI rebased on trunk from yesterday other patchesI corrected docstrings added a few tests including listvolumesI corrected a lot of functionality in the networking Still missing tests for the port forwarding rules but they are a bit more tedious working on it +Ive committed this and merged to. +Thank you all for putting effort to solve this issue. +This needs to get done for the first compliant item does not show up in the up in the menu now as well. +How did you get the IBM JVMPart ofSince the problem would be due to a bug of the IBM JVM I need to get a copy of that you know any way to get that JVM? Otherwise it would be very difficult to investigate this problem any more +This should already be supported +Maybe the test should be applied in +META +Resolved +Checking this in +Seems well need that method hack for Oracle +So the cannot detect any attachment within it +The structure of the g projectfile has been changed a bit so it requires a little bit of work and testing +Id guess that in most cases the recompiling wont actually choose another plan so doing it less frequently shouldnt hurtThe timeout will only happen with queries and statements in triggers though since the execution plans for those queries are stored in the database +So I just ported the example codethere +h to see the usage info +The API bugs out all the time and its really quite separate from the functions that its calling +But I am not a CSS expert so I leave it to those that knows better +I need to make sure they are good +Hi OlegI checked out and got svn version built client and core and we ran the test +Can you confirm that +Used the same workaround and it resolved the problem. +Typical the URL you should access is something like httplocalhostplutoportal +From reading the Drools blog it seems that will beis a major rewrite of Drools that will not be backwards compatible +added factory method to create DOM based instance +Should I upload a new patch that just changes the dependency version or one that has all of the hooks and depends on htraceAlso the htrace maven repo wont be synced to maven central for another hours so I cant upload a patch until then +Fixed for in +There was also a bit of a problem with the sorter +Clean checkout of HEAD does not build for me +Diagram after the formatting problem has occurred +I hate silent failures and that would be one +Assigned to Bojiar Penchev who is working on the DDL sequencer as part of the Google Summer of Code project +closing issues associated with existing releases +If nobody objects Ill commit this later +any objects regarding the changes to be made tosince is omnipresent throughout id like to apply the changessoon in order to avoid a lot of conflicts with ongoing improvements and bug fixing +But specifying is confusing so Ive removed that from the documentCertainly true that theres no difference in encoding but allowing arbitrary mechanism names is +Im not sure what logic is now +A kar file can contain arbitrary content that ends up in the server features bundles configuration for the bundles +I dont have enough spare time for that. +Yeah I fixed the script to kill the serverD +Jim makes the point that the config +Im not sure we should make a vhost rule listening on both ports by default +Fix up resource bundles +EnisTherere files which have conflicts with branchCan you prepare patch forThanks +patch for branch patches rename the clusterName cluster and the default value from CLUSTERNAME DEFAULTFARM DEFAULTCLUSTER DEFAULTWADICLUSTERRex +JMX to only reschedule another compaction check when the active check resulted in a task being think the patch on would also fix this +All set. +Both daughter regions are damaged +The extra bytes are no longer used +Punting to +Theres actually a way to configure it into the startup scripts +For and it would be nice to find a solution +Attachment has been added with description patch has been added with description list of reduced the dictionary object hence the objects package now removed +Rajesh no action required +Furthermore we have to change this also for JSP. +tricky thing to watch out for hereassemblyassembly has an execute phasepackage but what we want is to bind it to package and have it run after the normal packaging +Tries recovering using tmp then old files +Once the execution of test group completes the log is appended to the main log file +No Peter +Therefore if another component wants to reuse that script it has to have a componentmycomponentwebappmywebapporder you want to make stuff reusable definitely refactor it to include from a screen definition rather than leaving it this way +Yes you are right +Ah I remember why this was done +last new modifiedsorry i cant delete old two filedownload last new file is ok +Ive dropped the severity of this issue to Minor as there is a workaround one that the reporter has commented uponIve left the issue open for now pending more supporting evidence that this is a fetchmail issue or an issue requiring resolution elsewhere +Revision . + to trunk +Simply adding a comment and enhancing the XML configuration tests to assert the configuration. +After fixing a couple of bugs in the Sc interop tests the Asynch tests are all passing with the native stack +ok +Are you working on this at the moment? I was thinking of picking this up next week +Only saying this because the patch appears to be a fix forthis instance +Not sure if we want to default to warn level for and trying this out a bit more I decided to incorporate a small formatting improvement +Thanks set to issue is fixed and released as part of release +I have addressed the comments in the latest patch. +I changed the code that i mentioned and build cxf on my local machine new +It may be even better to list only the snapshottable directories that the user has permission to take snapshots +It seems like this needs to be rebased though +Can you add the API call going through where you see the signature mismatch +Im not sold on the interface being one of set conf var foo to value bar +is that just ant testnot familiar with +Not only does the SCM reference both projects that I have added so far it references the entire contents of each project every time a scheduled build happens +Can you review the patch here? I can try to tackle this if you need to +I didnt dig too deeply but I suspect that some piece of the opt compiler maybe was relying on the old behavior on empty iteration +Claus is that fix already included in IDE? If yes could you close the issue +Any updates on this? This issue is quite old and affects every person using a huge group of people +the enum is missing the SEMI here as well just like +Thanks +Answers to Satheesh Youre right Ill remove this class +Patch file +on the pushed and as well as master +You can discuss on the group and update the JIRA with concrete plans +I assume a few had some opinions on the matter of location when they moved the whole mr test directory out of hdfs and into mapreduce +This is the patch for the addition of a mediator +Committed to and trunk +Runningbuildjava jar with the classpath set as before again causes the issue. +Printer control commands cannot be passed through regular content in XSLFOalmost by definition +You can usually cd trunk grep R THINGILOOK FOR +To be clear I like the first option best +i files from other projects and thats good enough for me +Also we will need a that converts the finished computed vertices to the output +I imagine the updates are relatively small but I havent had a chance to look into it just yet +Correct unbalanced paragraph tag in release note. +This demonstrates the CSS styling that appears to create the issue +So all issues not directly assigned to an AS release are being closed +the authentication problem is fixed now +Thanks +this issue is fixed in r together with. +And btw using does not in any way make it +bulk defer minor issues to +Longer term we can look at implementing a dedicated mail server so Ill leave this issue open as a reminder. +I am using and still seeing this +Segment is good +is only included in the sane builds +I think we can close this now +But with more than RS I really think maxthreads should show better improvementsConclusion its all working fine +The analogy with killing the container was one of the reasons for thisbq +Is there a possibility that the order in which properties are set on the will influence the behavior of this fixFor example if the Interfaces property is set before the property then the fix should work as expected +Snjezana did we or I have any reason for not including it in JBDSIf no please assign this to Nick and work with him on getting it included +Check the CVS +avoiding unnecessary shade will produce a clean namespace in bookkeeper which is easy for maintenance +I want to make a patch for this bugbut how to create an review for patch? +Attached is the updated patch +Ill try out the latest patch +After the obvious fixes I have not gotten it to compile anywhere close to cleanly +Patch see issue description +The tool is helpful in finding possibly unused dependencies +Sending srcmainjavaorgapacheservicemixkernelgshellcoreSending srcmainresourcesspringTransmitting file dataCommitted revision +Included in. +Currently the build succeeds but it didnt really work +Ive added detection logic to the to detect a situation where the form object is not the same as the target object wrapped by the Errors instance +patch look goodI have applied it and the casting problem has disappeared +From every indication within eclipse I had thought I had changed to JavaI removed from the machine and then once again scrubbed Eclipsepreferences this time finding an instance of a runtime setting set toI changed that to and retried and so far it is working like acharmThanks for the goodwill +NMS the NMS code so that it sets the max pending message limit in the Consumer Info when a consumer is createdFixed in SVN Rev +I am also not sure who did the commit of the CSS +Will be more changes related to this task for Roo version? +All releases are Java compatible +Thanks for the patch! It looks good and submitted +I have now committed and pushed this patch +A patch for that preserves the form components values when is clickedWicket devs please review +set proper fix version +Located in +Attaching management server api server agent logs and DB dump +I will eagerly await the official change +Two initial skeleton files to give an idea of where Im heading with the initial implementation +Moved to and assigned to Rob +See also +Right but in order to add to the TODO I think controller needs to be added +concatws sorry. +Please reopen if thats not the case. +This patch incorporates suggestions passes tests +If you finish it before we release beta please correct the version +The current webapp examples demonstrate enough of the functionality and pretty soon you get on a level where a more realistic webapp would need some webapp framework to power it +Maybe they should become STARTING and LIVE which make it clearer when a service is just starting up and when it considers itself live +I have been unable to reproduce it on simple setups myself so Im still relying on our QA tests which comes with all this annoying baggage +We now include the jars with the plugin so I dont think addtional docs are required +Creating the Database via JDBC is not supported by most databases and so that will still need to be created by hand +Good catch +Closing tickets +Marking closed. +On the stream declaration you have too many parens +See for more progress on this +I guess I need to abandon my code habits and learn how big guys are doing it nowadays +This seems to fix it but the NPE should require a check to see where its coming from +Resolving +The problem is that async replication is actually sync the caller thread does not return to the user even until replication is performed +I have fixed this issue just testing check srctestgroovybugs on the current CVS HEAD +The version is missing from the update server while some maven dependencies try to download it +And if you can come up with a patch that would be even better +Is this still seen as necessary? Or more of a style exercise? +You would put a comment on every jira mentioned in the change log? wont that pollute jira very fast? Another related interest would be to have jenkins kick of builds against pullrequests and report if it makes the build fail or not +Same patch for trunk +Applied the patch thanks Michael +Yes thats the one Im thinking of +Just for information the +Thanks for the patch and the help. +Committed to the trunk and branch +I unassigned myself you and Dawid definitely deserve the credit +Duplicate of +Screenshot of the bug +Is this Issue a Blocker or Critical priority for EAP GA? If yes +This was committed a while back. +The API methods in standalone that take an IS. +Correctly solving will probably also solve this issue +Has anyone investigated to see if this is a Tomcat bugAt one point I added code that lowercased all usernames but it appears to have been removed somewhere along the way +I can probably do a good job reviewingthe changes that are there but someone else might need to review to think about the changes that arent therethat might be needed +Uwe is your stuff working well with trunk now +Done at rev Thanks Seweryn +Attached test case from bug +Filed a separate issue for the problem in downloading the snapshots from artifact info page see +This is behavior that could easily regress if we dont have a test +Fix rev +I applied the patch correctly but the bug is in the pycassa script itself +Theres all sorts of crazy variable names such as rstack rpos r fid ftype that have no explanation or documentation +Punting from currently if you shut down one RS after the other on a live production system you will be shuffling tons of regions around +It would be nice if the patch could include a new test which demonstrates the modified behaviorthat occurs with the patch +Since goes away with it what endpoint cps will based onAs for this patch I can rewrite the test to use endpoint cps +This patch fixes the erroneous disabling of security for contexts that are not +Committed to trunk +Ill be sure to start at the mailing list next time +I looked at the patch in wich you modify srcmainincludelogcxxBut in my copy there is no such a file +This time with the grant license to ASF option +Patch included more file movement +Method add retrieves the oid of the parameter instance and stores it in a local set +Default directories should by the webroot railo flag for search local like custom tags have as well +And of course I forgot to on the error line to get the stack trace details and now I cant seem to it +on patchIm no maven expert but my understanding is that the release plugin will manage the versions for you hence the upstream frown on use of a variable in that contextAs for why this patch solves this issue Ill leave that for a maven maven to explain +class files +I dont think there is anything new please reopen +Any updates on this +This is usually a best practice but this may need to be different because the engine may have multiple problems and because of the way its part of JNDI and initialization +Updated title and description accordingly +Would love to see a patch for this early on even if its still incomplete or alpha +It is the token position +Committed revision +Rado now that this pull request is merged into master can you try to reproduce this error +Reattaching files with Grant License to ASF +Here is the patch both for the code and +Thats exactly what I mean +Olivier we removed it from both local and central repository so we had fresh hibernate dependencies downloaded but the same error occurred again +It never failed on my machine +On investigation the header appears or disappears based on whether the mobile site is loaded or not +hmm + This is a failing test showing the problem +The Packet Bytes file is just truncated +Since optional in a dependencyManagement dependency previously was ignored this marker can now be used to identify which dependencies came from and which came from optional transitive dependencies +Moving this issue to next release +Probably a good idea to mention the names of the scripts that help with setting classpath set set set +vhosts can only work well with its overkill but I thought it would be interresting for users to have a shott access to apps rather than pointing them to dbnamedesignsomedesignrewrite or smth elseInstead having smth like somedesign would make things easier I think +In response to point of Santhoshs previous comment I agree it is fine for now +But most likely it seems that the issue is what you are pointing out +Weve tried this file with the Apache tool using the option and got the same binary stuff out too +This patch moves the message context management code out of class to the class +A clone of this issue has been created for possible further solution in can you attach your sample project +If we commence with them they might seem out of context for the reader +However my environment is Windows you tried Grails? It might fix your problem +the fix version for the time being since Im not aware of anything that Spring is doing wrong here +The compiler now performs this check and displays an appropriate compiler error message if the methods are not found. +I will post a reference to the information your gathered in that JIRA +LGTM to trunk. +m not following what the issue is +All links in the sitedocs need to be updated to point to new location if not done already site wiki svn mailing list etc +Patch for is in branch and in why did we drop the +I guess I dont have the privileges to reassign tickets +And if I look into the of the targetgenerate directory structure it doesnt look like the one in target what should I do to proceed with your workaround +Thanks to all you who helped get this in +Note this change will need to be ICU since changing void to const void will not workwith older ICU versionsA macro like ICUVERSION may be needed +Tnx But which version will include out of the box +The only way I was able to workaround this is to add the following to allow passwordless sudo for the cloudstack user +Fixed. +David has verified this is only a test issue +Thank you ashok for your tests +So assuming that they implement equals accordingly an AOP proxy equals call would in turn lead to true for pooled objects again +we should note that the sample file was generated using self using the code shipped with fopthe problem arises when writing stream +IMHO this works svn up username test password passwordThis should be fixed with last + +So if the returns columns with wide row the map method gets called times when it should ideally be called just once with columns +This issue is being handled as part of as the test cases include the testing of this issue +Can somebody working on this area review the patchThanks applied +It should be fixed in trunkCould you try it out +ESB does not support CXF via which is why this is a Feature Request You are the log files and ignoring the real cause I already included that exception should also add that there are likely to be other issues that come to light once we tackle the CXF integration +The use case Id like to cover which is probably a good portion of the cases is just doing one layer referencing one remote style and not having that style need to have any knowledge beforehand of what layers will request it +This seems to be a problem with Dojo not Struts +Reverting commit +Thanks for the Varun. +I also found your problemif the response including xml stringthe client cannot receive the result. +Mark as resolved once we see merged into a build on our CI and the absent from the run. +This change is similar to changing +Can you resolve it once review request is closed on review boardthanks in eabfbeaddbbedf but need to fire another bug to update the download url. +commited in rev Thanks Sean! +It does introduce this strange bit of code in all return null which will never be executed +My mistakecheers Roland +Some minor comments can you expand import +We tried the th September build and our issues with task runner threads have been youWhen can we expect the official version of to be releasedAlso could anybody please comment on +Cleaned up the order of the filterfiltermapping to comply with the xsd +Martin The problem with the classes is that its easy to end up running the tests twice or omitting tests that need to be run +The attached patch should fix the problem +You want to read this body +By decorating the Conversation bean. +Apache rules for code disallow author tags +Lets see whether we get past the CI build and resolve this if so +I made a small change in template to correctly propagate isRepeating for the cases where the left or right input is both repeating and all NULL +Ill attache some pictures to demonstrate the final output +Ive had to remove the version of to get rid of a since everything else use the version supplied by the JVM +SVN commit has effectively ported the patch and fixes from to +For the same reasons I do not suspect the code +is simply coded defensively +I wanted to backport to releases but wout any helps its too much an effort for me so simply closing +which then does a Noop operation Failover occurs so it throws the exeception all the way out +Do you know what state the Bundle is in when the adapt is performed? Could you check? It seems like the number of revisions shouldnt be zero unless the bundle has been uninstalled +I think this patch is fine except I think this should be done only if its being run on HUDSON +The patch is committed with r +Removing myself as the assignee so that I dont hold back others +Good point +in at least one case there we are only actually using the value so iterating values makes even more sense than is the third copy of the issue but looks like the best effort yet. +this patch check for wsdl name and call to correct method +Is there a test that could cover this +bq +Thus if open or close block or take a long time the heartbeat thread becomes blocked +Anyway did not do any perf testsLooks good to me +Im not sure what package managers have to do with Apache Cassandra +Sure +Thanks a lot Christian +In other words I think choosing a new session name might simply be masking the problem rather than solving it +I assume we are eating the spans in the +weve gotten rid of the blocking annotation and related checking. +RTF is still behind the other output formats and probably will always be simplybecause of the fact that RTF is not as expressive as it should be for mapping +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Is this the same as? Yes the only difference is that describes the problem in a more general way +Unbroke things for standard packaging Im not certain we can get the usrbinaccumulo sorted out without refactoring the scripts +The change is available as of revision +Here are the steps that I would recommend for the attached file tar xf cd ant f homejjohnsonprojectsbeehivebuilddistant homejjohnsonprojects ant f homejjohnsonprojectsbeehivebuilddistant homejjohnsonprojects build jar cf mv CATALINAHOMEwebapps The point your browser at httplocalhost and click on the links +with the desired license +You are free to change it +Verified by Sian. +Ah better simpler than clever! Youve avoided callables all together yet the code doesnt have original dups +This usage might be little bit strange I would say +You can just try to change the values and youll see that everything is back to the default configuration after one of the applicationI personally implemented a workaround for the issue so I dont really mind about the priority +An actual patch based on Daves patch with one difference Rather than add the new files in the package they are in packages where their related classes liveThis means that is the name of the Guice binding file since it lives in the same package as that it is based upon +I think its correct +And I havent seen this error for a while and I had tried to look through the code logic to hunt it down but to no avail +Its probably very useful for a large subset of needs out there though +patch with testcase +Attached is the sample application that was provided by the community user + +patch ready +Since the AM doesnt know the proper any node other than the node the AM is running on will have no for the app and the corresponding task logs will be recorded without +Thanks. +Will test against work development system on monday +Convention is pretty stable and is the way to go in the future +Committed minus the block of old unused code +As noted we now have preferences for Android SDK in Tools +But we can agree that here is JIRA for it +updated patch merged my javadoc cleaned up formatting added javadoc My to commit this +I had commit problems yesterday server was down +Updated patches to s is to getRack and is to getDatacenter +Thanks for the report keep em coming. +I dont think thats going to happen in the next few months so would probably hang on to this for a version after but valid nonetheless +Use of +needs more tests + +Ive cloned the original issue to reopen it at Bretts suggestion +Heres the relevant phraseIf an SMTP server has an implementation limit on the number of RCPT commands and this limit is exhausted it MUST use a response code of +i modified the structure to this nestingul li div x span li ul li div x span li. +This is a known issue +I want to be able to continue running the tests using one command +Oh I had seen that comment weeks ago +Hi AnkitFYI inIE I dont reproduce the st but the ll have a look later +Hi Tom thanks thats a great finding +Generation of content thats static should be fine but generation based on external content we dont control +looks good +please provide a simple test case which can be run against a standard jackrabbit +The bug was in seam validator +Ill need to run more extensively to get accurate numbers but with the patch Im seeing ranges from minutes vs without the patch +Ok I opened another bug for hang in release mode +Attach a different fix + +The failed test is because the data was missing and the test passes locally if data is in place +I would also suggest using Bintray to distribute Groovy binaries but thats up to you guys +read and write permissions +Have they been taken into account? If so can we close the issue? +This was happening for me using Ivy when a hsqldb transitive dependency was encountered during resolve +Closed for +annoying indeed +It may worth running it with J on Linux to get the first clue if this is DRLVM issue +And have releaserollback only WARN that the tag must get removed manuallyb + +Ram Which assert should be changed? Do you want to include the assert change in your patch? Or are you suggesting a previous test case is broke? If so which? Thanks +If there is any insight in the form of comments that you can add to the deadlock code I think that would be extremely educational and helpful to anyone that has to visit the deadlock code again +To be precise it is descriptionLet it be Create a new Interceptor Java class and Create a new Decorator Java class. +It issnameCXFversion Im not sure how the discussion in that thread would help me +You now have the ability to assign bugs to yourself +That was it. +i havent update my device till now +Especially something like this where any update the the requires updating the serialization version and then breaks compatibilityI suppose I can live with this in and a lets try really really really hard not to break compatibility in the releases + +Yeah Uwe thats true +just had to move a comment around in a merge conflict in the new btw +Heres a sample patch of what I mean +I would also like to see the Override annotation on overridden methodsUnless there are objections I will commit changes that remove unused imports and add the Override annotation in the next few days +an eclipse patch file of the proposed submitted nweeds to be reviewed and for the delay but I am really doing my best to manage my backlogPlease verify the attached patch which is based on yours but drastically minimizes the impact +So the search is immediate +IT added +When I deploy the app it just works fine I mean references to any properties are resolved in a JSF +pushed to in Developer StudioVersion Build id Build date +can be this merged to? is closed +Cool thanks +This will only be addressed in. +Its because of Embedded and testsuite +Thanks Mamta! Committed revision +What do you think? I can generate a new patch that has Zhengs fixes and the test output update +Please feel free to +Im going to triage this by disabling the test +Equinox didnt work with it earlier this week +Do you have a URL for these sections? Which book is this +re Nick I didnt care much about false since this was supposed to be just POCHack +Fixing test +Ive edited the code examples to remove the visible html in both back to you for verification the docs should be available on by the end of the day but I wont be able to edit them for EAP or any more +This is similar issue we had with testsuite +surefire console reproduce the problem +The kernel does not have eventfd did you try with +In the interest of time it might be better to save this for future work +Can this go back to as well +I tried this on nightly and the same problem occurs +on +Ok I will need to have a look custom type id resolvers are not as extensively tested so there definitely may be some flaws in thereThank you for doing the detective work here +Closing all resolved issues for. +Typo of attachments in package patch attached has been updated to work against includes test cases and fixes the bugIt also fixes a related bug where the length of the content id is less than characters long +Doubtful that it will make it in for Beta +The ETA for this issue depends on someone providing a test to reproduce it +Thus setting this as MEDIUM difficulty +However it will be some days before I can really look at this +My scenario is pretty common uploading a file to a webserver and then running a batch with uploaded file +Bulk change of Fix Version for all issues fixed or closed for +Forgot something after The created node should be displayed with the default rendering you have to add the +The only argument I can see is that we could set up some extensive test suite that somehow tests all the bundles that comprise the Service Platform and we test and release a packaged snapshot as some sort of sanctioned service platform release but I dont think we will be doing that for a while +The instance is always null +Im pretty much going to revert at least in my local branch to continue testing +Typically I dont know that libpq changes all the time or lacks hundreds of featuresIts interesting that you bring that up because libpq uses bits for its versioning +Absolutely valid comments and I have implemented them +I think we can just do this as we create new schema versions +Attached metaphone encoder that uses apache to avoid LGPL problem +DONE partial implementation that uses Hadoop TODO convert the partial implementation tests to Hadoop +While I agree with the user principal being extracted from the token I think its best to push any issues of token serialization outside of the client as well as keeping token deserialization in the Authenticator implementer +there should be some official tips to tell us is it a bug or not? if a bug it should be fixed immediately or show some resolutions officially +Unfortunately for the you have to set the encoding at creation time but has no clue about the encoding of the incoming data +I think that means committing this patch or a change to +As part of the service lifecycle patchId rather see a patch for this separate from the lifecycle work +Your reasoning seems good to me +Attached log +The behavior is slightly different now rightNow if theres an invalid version we wrap an inside a +Sounds good +Marking resolved. +I change it from s to None +Triaged for +bq +But the easiest solution is just to omit the parentId +Patch has been files are still present in master should they be? +We could rename it to make it clear that it is the latest status from all executions +Probably the version would make use of optional +s +Rather than these various policies it would be very nice if the compaction dropping policy were pluggable +They are all put in buildlibhdfs +Hi Tobiasare you using the script to modify the project file to use ICU as transcoderI think we are looking at different things hereBy default NONE of the projects shipped with Xerces has ICU enabled so either you are using a VC project that someone has already modified by running or you are building a copy of XercesOnce you double check which case you are in please close this bug +Marking patch is available minor edits needed before commit +Closing all issues that were resolved for. +When NO native type is defined for a column the corresponding runtime type will be used for DDL exportstring bigdecimal biginteger blob boolean byte char clob date double float integer long object short string time timestamp there any reason for Teiid Designer to keep modelling the views using the design time data types? Can we just use runtime types +Jochen +Thanks Chris! +pending jenkins looks ve committed this to trunk +How we achieve this will be interesting as this stuff sounds like a service +Resolved +The fsimage just keeps bloating +thanks integrated! +Leaving it in trunk since James said hell make the improvements +Can you provide a set of sample pom? so ill can check it +The test case failure does not seem to be related +Ill make sure I dont touch the settings +I finally understand this Id prefer to have that depends on cobertura asm etc but this solution will do fine +Regionserver dutifully closes the region +I reversed my changes from this morning because a new SNAPSHOT of Camel was pushed and I had to change the from to in order to match it +I have made the patch to be surethat I dont loose the code +Im bumping this to Later then +i wonder why it was renamed back to now +the hadoop fs use +So could you try on a fresh server and workspace to see how best to reproduce the NPEThanks +Duplicate of +But a workaround exists in Wicket +Were trying to set up our own SVN for now +Committed the fix for both Condition and shutdown to trunk. +Sure +Is the only time when this happens a loss of the ZK connectionAll that a new is doing is a call to setup +Not sure what is the bugsolution you are pointing to in this jira +Thanks for spotting that +for further steps to reproduce +The NN loadGenerator generates load on the NN +Hello everyone +It makes sense to ahve such a feature +This tool is for people who need a defined category structure or want to grab all or part of the dmoz category structure without urls +Approved for +in +Fixed as part of close resolved issues +the localized configs should be marked final +Maybe we should split off the admin tools into different binaries? Exposing analyzer and such in the same shell wrapper as binhbase seems a little odd to me +Didnt realize this was still open +patch is built against trunk +Looks good to go +Good info thanks +No it isnt there is no change in behaviour +the assumption is that it is always a subset of the select columns and therefore we dont need to do anythingyes +yes I now think it should be the immediate parent and what you argue is correct +Can we target a fix for Pig +I dont think Im going to change this unless theres a popular demand for it like I stated previously Investigate is essentially the same as Unanalyzed and as such it wont be shown in the analysis dialog. +Thanks David for following up on this and opening a new jira for the new issue found +I think the proper solution is to provide a soap connector bundle that works +Sid is working on porting this to trunk +This isnt the heap where you give things plenty of roomThis is a memory leak that should be fixed +noticed a bug +Please attach stack trace of exception from log if available +ran and test manually +Trivial patch attached +Feel free to give it a tryJuergen +Jon reviewed my patch over on review board +Resolving. +Attached wrong previous diff this is the one you want +I think if the handler is added to a global phase this is a correct behaviour +That does seem strange +Attention you would prefer a separate cloned JIRA issue please tell me and Ill get that done. +Fixed. +Heres a patch with a unit test showing that is sufficient to allow concurrent execution of in +Thanks Slava! +The first pull request was merged but we still need to upgrade to latest versions +Committed it to. +It brings in also as a dependency but we should add it in explicitly if thats required for another test +Claus I believe the code is not simply doing logging it is actually overriding the being used +the last index entry is always the last columnno it has to include the last column but it wont always just be the last column. +Your version jre jdk should be ignored and not chosen now for launching as +and are tagged in github both along the branch +after logging in a flushing the cash you see the new user on the first page +connection +Still works for us because we run hbase processes inside our with a on classpath +Its just that the formatting changes make the patch seem larger and more complex than it really isDo you have a sample image with that particular palette problem that I can look at and maybe add to the testsDamjanSo far all the samples Ive located have copyrights or are proprietary +Merged to doc branch at revision Merged to doc branch at revision Merged to doc branch at revision Merged to doc branch at revision Merged to doc branch at revision +This has been fixed and merged +Sure just give me a couple of I made a simple project to test this issue in a reduced environment +Hi SagaraHere Im attaching completed works so far please review my works and let me know your comments +Ted. + added javadoc and comments rearranged codes for better readability changed map output key type to removed duplicated file srctestorgapachehadoopmapred rewrote some comments and reverted some incorrect changes in looks good +Raising to critical bc this is need for statically checking rank of points when iterating over points in a region +If you modify the perl script with the Im attaching then all receipts are received just fine +Hey JeffLooks like this patch is breaking couple of thingsIll look into whats breaking and see if this patch is useful +Maybe I am wrong but if thats the case I think the Calendar type should be dropped in Hibernate mapping since user may experience uncertain results + +the notion of truncation and defrag were too intertwined an empty storage file would not be truncated as we wouldnt defrag +I have just tried and am having the same trouble +When there is no consumer topic messages are droppedUse a durable consumer or CACHECONSUMER with a single consumer with default prefetch it makes sense to cache it. +Never seen that transition to closed remaining resolved bugs +Local entity resolution wasnt being used in all cases on +So all issues not directly assigned to an AS release are being closed +Looking at them with sstablejson I see the all the subcolumns are actually there so I think what you may be seeing is could you try to reproduce with that patch appliedJikes that one looks pretty badAnyway with the patches in in combination with the modifications of mentioned earlier Ive so far not been able to reproduce any of the issues Ive observed earlierThanks for your help +Unfortunately Pax Web does not recognize the interface and so all my bundle filters that depend on this service stopped working correctlybq +HeyWhen did you change the groupId from to by the wayIm unable to reproduce this with a locally built appassembler with this new groupId +So if we provide CAS within the region server row lock primitives are redundant and add complexity for no apparent gainThe only downside of CAS compared to a lock is that the client has to handle the retry on failure +I have done a run through with turned on so I should be able to supply an ij script to reproduce the problem +bulk defer to +Marko what you say on producer method agrees completely with +we are good +Next week well heve to indicate what is experimental +I havent dug in completely yet but if the main schema is updated to have the proper attribute it looks likt he one in the is used in this case somehow +bq +I just committed this. +Make the object capable of running multiple fs commands +However it is possible to get past it on the keyboard such as clicking the mouse in certain areas of the screen and then typing +Andrea the solution proposed by Jody would in my view address your concerns +Done. +If am wrong you can let me know +The final implementation is now faster than in Commons Lang and provides htmlEscape html and html methods +fixed javadoc error +Marking fixed issue as closed. +It has optional bindings to libcurl and a simple HTTP client is already made because its used for running the javascript tests from the command line +Removed the extraneouse subdeploymentsubdeployment +HiIs there a solution planned for this issue +This patch checks for admingettableFiles and adds the request handler if necessaryIt also builds the hidden list as the inverse of the gettable list +Attaching updated reference doc +Weve been having some discussion via email about this issue and we thought we should move it to the JIRA so Ive copypasted all the questions and responses below +Thanks for the patch. + FYI + +However if the directory is empty it throws an exception +to commit this +Is this what decidedAs of shipping pack or not shipping pack in archived site Im in favour of keeping pack because if we dont keep it it means some people consuming the zipped site to share across their infrastructure may be tempted to stuff which might generate problems. +Stein I am going to commit this any time soon incl +Fixed in ontonet +Able to create private gateway without any closing the bug +This issue fell of my radar but its back on it now +Updated provider for mysql see issue +I was not able to upload it in my Plesk administered system +But the can not have mutiple columns or a row value +sid Can you please confirm if ant tests are still working with this change? The changes look good to me +After the compression patch such an exception is no longer thrown + +if there is no objection I will commit shortly +there are one empty folder in the two zip file of the sample project but when the sample projects created using the Project Examples wizardthe empty folders are not created the empty folders are output folders of the have tried adding a a empty source folder to the zip files but when the project get created using wizard the empty source folder was lost so far the issue any empty folder put into the zip files of project examples it will be contained in the created example projects +thats not supported by either backend + +Is this pageScope variable something added recently? Because when I opened this bug I first asked on the ML and Graeme told me there was apparently no solution at the time +Internally it already calls the async anyways it is just missing the split computation which could be done from the client side using the public Bytes methods +Any better now that it points back to trunk + In my experiment I cast a to unsigned int to work around +Gant is one of our test cases and we are having no problems compiling it now +No comments anyone? May I commit this in? +EricI think it would help a lot if you would perform the spoken word poeticversion of that comment over a jazzy bass line with a touch of snareAdam +Maybe this is a problem on windows +Kelvencan you take a look at this to see if the correct fix has been submittedthanks can you check if your agent is updated +Committed with nit fix thanks! +We need to have the principle name and the keytab configurable per daemon +renameTo fails on windows if the destination additional documentation +I will try this +I also updated the message on the constructor where it would result in duplicate information +One significant change was the removal of a global change listener from +Reattaching patch to get a new QA report +fixedcommit in master to branch +The first one would do the wrong thing nothing +What else do you want to know +Did not backport this to +Having done that can I consider this approved for +Maybe someone else could take a look at this +looks good to me +as long as you run with maven o option +So this does not seem to be a solution currently +Any idea is welcome I can try to invest a bit further +I repeat step run Seam Generate Entities for dvdstore project one more time now I see the sutuation during minutes Ill inform when this end processing aga it stop after minutes +Yes for security reason we can start with user level cache +You can reopen it if you want and that was what youve you need to provide a pom as stated in the instructions you can made a simple one. +commited to the sandbox +The only possible alternative I could think of would be something like keepregisteredtrue instead of deregisterfalse but maybe it is fine the way it is as that would be kind of wordy and we would need to change it quickly +Reopening for handling in Struts +Assigning back to David for final edits +But this however breaks any java program using Swing + +Would it be possible to update from git again and regenerate the patch +Fixed in a series of patches +I purposely disabled it because the actual demo coding is not easily fixable +Did you consider that solution +Closing all issues fixed previously after a brief review of each. +We need more functionalities on the server side anyway +updated initialization patch from +Moving it out to +If it fails to create it a single warning will be placed in the error log +This is expected right +Most people using it will not care and those minority that do can be the ones to go looking though the documentation to find out how to limit it +This patch fixed the javadoc warning +Cant reproduce this issue anymore looks like the recent changes have fixed it +Proposed fix +It is true that in some cases its not completely as easy as in thrift +Go for it Jimmy +But if a build forks and repeats the same thing well theres not much to be done about that in i think +One other thought on the naming I could very easily see adding thrift for the and for managing them and getting stats and such +If I remember correctly his latest suggestion was something like to unwrap the exception as long as the inner exception class exists on the client +I have committed the patch. +Indeed this works now +looks good please file a new jira for the problem above +Added syncrhronized block to and made file scanning also concurrent +Writing an app that takes advantage of all of the new CDI goodness but I also want to take advantage of the new Async capabilities of Servlet +After looking into this a bit more Im afraid this kind of support will have undesirable repercussions across the project particularly in the generated client code which needs to be able to generate faults +Ubuntu Fedora RHELCentos +Glad to read this ChetFeel free to close this JIRA +Possibly in my mail but I have no time to dig it up now +Modified Ref Manual topics and new ATAN one appear in Latest Alpha Manuals + exposes attributes and both of which are by default set to whatever you set the b startup switch to +Any plan to backport this issue to +Agree with you SiddharthJust a thought wouldnt it be more logical to enforce to accept as a parameter rather then checking existence of all the parameters in since there is no way to know what user will override in Please feel free to correct meThanks simplest would be to check the configuration type which keeps the API stableThe reason I mentioned parameters is that apps that use may have their own configuration type + +as is +This problem only occurs in version and newer +Otherwise it will fail every build in the future if we commit the patch +Only one is needed in the zip while both are provided by the xerces package +Would anyone know how exactly does the endpoints are removed and should be recreated when caching is full? What methods are called on when this happens +When a new node is bootstrapped then heshe will run into issues +But that has since been replaced with passing the cwd parameter to +The part of my program that needs the functionality of will use JDBC calls instead of Hibernate calls +Quote should be allowed +can we remove from the warNo those are the classes with the client code that we use to talk to hdfsThe tests add a variety of other dependencies and test jars and actually start up hdfsSolr and the webapp simply talk to hdfs +we need to close that stream no matter what the current hadoop version is +Tim good catch thank you +With r I introduced which is used by andCurrently the API in doesnt allow to use as +uncomment option cn which is required in I am able to create a user without any error and I see that user in ldap user as Question If I have more attributes on LDAP side will all these need to be mentioned on? Can some of the attributes be marked as optional? I think its more of a LDAP question that portal identity question +Thanks for the detailed report! Ive switched to fix so this gets addressed before release +I suggest to download the binary package and then you can take easily all the dependencies from the lib folderSounds reasonable +In fact if you look at the attached source theres nothing remotely grails dependent in the example +probably I misunderstood but I thought that was ok for the moment as long as we also can find a better solution for Oak which I think would have to deal with the same topic of shared resources probably on a dedicated issueno you got that right +I think pingThread was sending updates without any problembq +hmmm running a test I get Soap message is incompleteThe SOAP serializer has detected errors in the header or envelope +I cant produce the problem anymore +If this is to be added it should definitely be configurable with the default off since this is a major backwards incompatibility +Meh I did not pay attention to what local branch I was in and the commit for this one went out along with something else so while I was at it I also backported it to to avoid forgetting +Benjamin Yup know the line in Hadoop the exception comes from just not how to find the damn stack trace without hacking hadoopMy thoughts on and was that if the prefixes were to persist until the layer then all should be ok as Hadoop can handle both file and shared prefixes +this bug casued by a Axis C web service that embed an axis c client + +Make the unit test cases work with newer versions of hbase and zk +James any news on this? Im suffering from this as well +Patch to update site Update documentation link to change to Update documentation link to change to Change page to point to common +Known issue not our problem. +They are in indication only that redelivery may have occurred or that a transaction rolledback +Not needed + for the patch +Ah the patch should also make the class public +Upgraded to Bamboo +Note that version also allowed HTML markup which was not in a CDATA has also been used +Not gonna happen for anyway +need to update demo after you fix issue is not reproducible with the current code in to be rechecked after implementing the feature described in +It could perhaps be changed +This should probably be +First we should establish a release process for releases with Maven in commons +Emile ha I shouldnt have used the second person there you in you forget me +No aggregation across the cluster. +and another issue when I fixed the second one +if no one objects id like to commit this in a few days +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It was anecdotal but it did appear when the user was running upgradesstables which is different than building the indexes from scratch +Should we deprecate the ability to set these caches using fractionspercentages and go with fixed values which can either be absolute memory usage or absolute rowkey count +I think the advantage of this method it that it is way much faster than the normal way of solving alternating least squares plus the way it it handles implicit data works well +Thanks! Service Properties File is always encoded in because it is written with private static final String SPFENCODING Guess the comment should say too +I removed the std prints in lucenecoresrcjava that I could find on quick greppingIll leave this open so we can somehow automatically catch this +I committed this no problems with any other codecs +Thanks Dave and Kathey +This approach converges much faster for the large x case and fixes the problem experienced by ScottI added the example submitted by Scott as a test case and got it and all existing test cases to pass with my changes. +What version of Cocoon are you using +Hi JustinThank you so much for running the cite tests for me +Implemented these changes in Rev +Can you rename it to say earliest +This upgrade may resolve too +Maybe we can try the nightly build EUT again. +Indeed is the version I testedI will try the CVS version today and see what happens +The minor fix for IBM JDK also resolved this one +The is affected by the same bug +Forgot the to +bq + +Did anyone see the issue yet +Kristian was telling me client driver will ask the server about isolation level if it is unknown +There a lot of edge casesThese cases are hopefully covered by the patch +So if ANY other project you are trying to use leverages commons you are forced to use hives version in some cases +Where the screenshots and log +Nevermind I dont reproduce the issue +For the it does not matter so I leave it there +Pull request +It seems this did not take place +That is odd because we have this option set to true ever since I reported this issue and I am pretty sure it has been reporting duplicate literals in annotations as violations ever since +So I would say it covers this use case +I will take this project wiki is now up to date. +Closing as wont fix +adding link to revision . +I dont really think this should block us as we technically dropped Fedora support for the release due to EOL status +Closing all resolved tickets from or older +Ok thanks for this precious feedback i think thats the re using the version of the plugin but i think that your plugin definition causes some mess since it should re going to investigate the issue in the meantime to fix your problem you can try to upgrade your parent pom to version if you have such cannot reproduce this issue with maven and sonar and a project with plugin defined in its pom and Do Not Extend Java Lang Error rule cant find unfortunatly maven anymore can you please give it a try with maven and and let me know if your problem still occurs? as there was a work around and I suspect this has gone in +Its not my patch that is making it behave inconsistently it already is behaving inconsistentlyAnd yes youre right the same issue would occur in equals +You may simply derive from the +Applied to branch +Fixed at SVN change +But wait theres more +The codebases can also be properties look at the policy file for the test harnessYes I mean a security subdirectory in the release directory but it could be done other ways a example in the documentation that could be cut pasted might be enough +The actual code fix is only one line because the of resync bad edits was already done by Boris viaReview comments welcome +Storage already has Apache license header +I would like to test that it really does resolve the problem +Reopening to add more docs for this feature. +Committed +Also passes the test + of column eacute my good Claus! Young apprentice though I am I must leave the nest and put my CLA to good use +Ill try to reproduce without it but it would be helpful to have +Wait +Hey AllenIf you find the config description both too generic and too specific perhaps you could suggest a better one before vetoing the patchThanksEli +Please let me know if any wording changes are needed here +no negative feedback since putting the change in +So the repository ID of the build repo used for the brew build wouldnt match the ID for the latest repoWe should try using the build repo for generating the Maven repo zip so that we no longer need the latest repo +I have added three new ln properties Implementation Class Component Factory Name Configuration Policy +In this case I think we should put the event in the waiting list again +What are your plans for getting information to Sean on usage and updates to features +This issue can be closed. +Thanks Andreas for detailed explanation now its more clear +I dont encounter the same problem and therefore the refered use case is not valid +In particular the next release uses separate work queues for mapping and reducing documents so couchjs spends much less time idleIm not sure adding extra couchjs instances will be much of a win in general as for many simple views couchjs can more than keep up with what the DB feeds it and JS engines are only getting faster +Sounds good to me +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian issue is fixed in the release! Sorry im new to this bugtracking system +The changes were reverted and deferred to ER +the moment and maybe for ever whether the auditor runs should be configurableI havent included in this patch since there is still confusions whether to treat as a plugin service or not +by Deeply nested if statements are hard to read and so to maintain +Otherwise we are essentially shipping releases with dead code and unused dependenciesFor the dependency I am fine to remove it for the moment +Might be useful to add a Web Service component +Im not sure why this is the case except that you do need to be very careful of what is in the libext directory +Fixed in revision +If the limit is exceeded TT will kill the task consumes the most amount of memoryWe use reserved memory instead of total memory because there may be cases that we have nonuniform clusterMemory limit for one mapper +Verified in +Pull request sent +I just committed this to trunk and +Reopen to wait for Cristians response +At this time the model does not need permission security annotations. +Youll have to download the latest snapshot to try this out +I should be able to verify this within the thanks for the update +Now I found the way to call CA and I can get the Im writing the testing part of testcase +Comitted to and trunk thanks Hiranya +My cents should have a javadoc explaining the basic use case +To keep reference to valid stream have reference to resultset and blobclob also +Hopefully as a later checkin to but I will mark as for now +This jar is located ateclipsepluginslibtools find the reason this is import it is necessary remove this import and replace to please apply my patch or can I commit the changes directly into branchesBranchtools +Im curious about this one its marked as resolved but I checked out Radix yesterday and tried to build and deploy according to the instructions and I get the same error but Ive checked the env script and all seems fine +I had submitted a patch on to disable it on Windows. +This issue has been implemented in +A syntax similar to can be followed + to nextAfter as addition to and for the approach outlined above for resolving this issue +Im unclear on why is called multiple times + +I have one question about the test case though +Thanks +This patch makes the task tracker copy the stdoutstderr of the child process if it has a bad exit +This patch fixes the library path for the different contrib tests +BTW Im thinking of finishing all of INLN by March for the release +This was different in previous versions of MINA when a small number of filter chains were shared between sessions so we might have to think this through +Thanks Michal thats good to knowIf you find the time to run it again a couple of times with CR please do Im curious whether it was a fluke or something you could reproduce reliably +Oops sorry here it is +Thanks guysErwan the only part of your patch is in trunk at revision r. +May someone pls try this +Thank you Dag for the fix +Eirik are you still active committer or I have to fix all this issues myself +Completed the following functions fn fn fnremove fnsubsequence fnunordered fn fn fn fn opto Still need work fn opunion opintersect opexcept fnid fnidref fndoc fn fncollection fn the following functions fn fn fnremove fnsubsequence fnunordered fn fn fn fn opto fndoc fn fncollection Still need work fn opunion opintersect opexcept fnid fnidref fn +Cool +OK sounds like you want to replace more logic but thats not much harder +Ifhandle returns then the flow drops down into the discovery partright? So maybe the code needs a throw new after thelast handle? And then the logj discovery doesnt need to be inan else as that never exits out the bottomWhen testing for the jdk logger Im not sure that the code to check that java is actually available belongs here +I added a couple comments on RB +Please review +I didnt have time to extract it from Cocoa +Thanks for the catch +Thats how its done in Karaf and the paxlogging api has been recently enhanced to also support custom layouts filters and error handlers. +I will go ahead and commit in a few hours if there are no objections +I just dont think because a class moves to another lucene component that it helps the situation any +Ive added the feature to branch in commit +What jars are on your clients classpath +Please if you discover this again. +Im planning to use Dojo popups as tooltips in the near future which will hopefully solve all these issues +Committed revision . +I have no problem with this? Can we close it? +It would be sufficient to put it in the output itself +Sweet it works! Thanks Gavin +Collaboration on this sounds like a good idea +Manually tested +with a slight modification checking for jcrcontent instead of just jarcontent to prevent false positives. +not holding back +I had created for the doc changes for this patch +I am not quite sure why +Please confirm the current status of this issue and if its incorrect change it as required +Ill have to check on that tomorrow +Fragments and Panels are created to provide a break in component replacement +This changes s semantics and may not work for if one is say using the number of outlinks as a factor in scoring +Yes +Patch attached +Links verified closing. +This sounds like a Geronimo problem which needs to be fixed +Once we feed SOAP message in to the model its SOAP after that +I am not sure if the fix in solves your issue regarding having spaces in the path thouBest way is to fetch the source and test it your about the quick judgement i thought you have spaces in your parent directory fixes your issues +Thanks for the patch +Because portals have its own authentication bridge includes Seam Identity component that understand portal authentication +Are you both sure is right +assuming affect was supposed to be fix version +I talked with Randall and a new set of examples need to be created based on Jboss as and removing the dependency on +It adds about lines of code to which is already quite big affecting readablityWe should rename that class or some such thing +Please review the updated patch +Can you please align this JIRA issue with whatever issue that you have open for the merging deployer +Revision in trunk fixes the issue +Patch for yahoo security dist branch +Cordova does not support at all but it shouldnt not work +The release does not depend on it now +nice work Kenneth +out some common code from and into an +bq +As per ML comment +But if you prefer a new JIRA let me know +It only covers the namespace not the other terms +Closing +We like to keep the code as clean as possible since we end up having to clean it all up when we prepare releases and that is a thankless boring task +But dont forget were in an open source project +The diff in the previous patch is missing some additional tests +it is just a documentation fix +No everything in old jboss was bad we need to reinvent the wheel. +which have performance problemsyou would have a better way to do it +Not reproduced +It revises the way that RPC exceptions are handled permitting arbitrary objects to be thrownTests are still required for this patch +The issue has been fixed is not purged anymore and It confirmed that it passes +must have found the problem +you Daniel for the review +No nothing at the moment +The reason why I want to compare against the original impl is exacly for the reasons you mention making sure that our impl compresses well and trying to find bugs in it. +This is likely to be the reason why it no longer associates with the esb deployment +Sanity run ran r with failures +In that sense it may make more sense and be easier to understand whats happening in each state if FAILWAIT was a separate state from FAILABORT +We have the same problem with Spring IDE Extension +in order to have better test data id suggest to add tests that create a lot of big collections and access them randomly to address the concerns expressed by julianresolving incomplete. +Actually I dont think that will throw that exception +so I can incorporate it into thetest case +Does anybody know ifwhen this will be fixed +The example Central sourced +There is no problem with cif component behaviour with visit algorithm also +Slava do you have any time to look into this +This part solves which had a race condition where the proxy start up code depended on a side effect of the plugin init logic +Run Cobertura on mycomponent unit tests with Cobertura this will produce a file in the mycomponenttargetcobertura +Is it worth reverting this change +Thanks for the patch. +The attached is the patch base on branch +This requires that the action of the configure form is determined dynamically +ManjulaI am on version of axisc +Poms in also need to refer to the released version of the Zoe thanks for reporting I have changed the version also in and and now project installs fine on a clean repo +Hi SureshLooks like this wasnt committed? Im going ahead and committing it in +Verified fixed at svn revision . +I will commit a fix for this with the JSR fixes +I like this title +Of course those are all just in the test suite anyway +Event body contained complete data string as specified +The correct patch for the latest contribution to vtests componentcontext +But DB is still bottleneck of the test +This patch is applied on top of Tuscany revision applied +Final code changes And also other files changed are from dassamplesThere are patches for the sample client for The readme under the sample gives all the details +bq +its their prerogative. +It may be difficult but possible +I have added these files to the and EAP repos thanks for your help Marek greatly issue is now resolved +I just confirmed that it works in Firefox here as well. +This is both inconvenient and a performance problem +Another possibility is to look in etc and match the contents against releases +bundle repackaged into are syntax errors in the pom doesnt validate +The incoming Query has a which supports multiple feature ids in an efficient container +LuisTry turning the logging level to debug you will know if the error is happening inside the UDF +Does the bean deployer also require jdk for its base behavior? We cant introduce a jdk dependency at the common module level though +Also it seems weird to have to add a query in the shard casebq +bulk close of resolved issues. +Fixed with blockers and critical issues are resolved but not verified +I dont understand why neither of these options can work in your caseId really like to see the basic not related to your production case Spring configuration emulating the problem showing the basic server client Spring configuration where the interceptors are actually lost +That would be the case if we changed the generated cover classes to be interfaces but that wasnt the intention of this issue +Ive just committed this +Verified. +Please verify in EAP documentation on stage +Now that work is well underway with AS all previous community releases are +Thanks for the hint +JeffCan you provide a test case to reproduce the bug? it works +bulk close of all resolved issues. +patch removes SNAPSHOT references from the DAS M Branch and include the windows batch file used to create the release in the javadasetc directory +file see the target +Looking at the fix I dont see how it can work +Chris is this what you had in mind? Are we good to go? is better but it still makes the a public api which Im not wild aboutWe arent we considering the Merger? It would be simple and would easily solve the case by aborting the merge after N items are seen +txt or name +When saving the context Ive added a check on whether the context attribute is set in the session so it will always be saved if a new session has been created during the request. +The attached patch included some changes that were unrelated to this issue +here is a first cut at a patch for both this and +Laura this new behavior introduced by and the flag which gives the old behavior needs to be documented in the performance tuning guide +Fix in place +Good point +bq + can be simulated with the Ant command to create an adminIt will not show up on any UI except Webtools viewing the Userlogin entity +Moving this out to please move it back to if you plan on getting this fixed for final release +Transitioning to closed. +Unfortunately I dont have any better ideas at the top of my head +This is a spec licensenot a software license so it seems our needs are still adequately met by the legacyWC document licensing scheme +Running now +From my perspective the current implementation of Complex equals which is consistent with what we also do for is natural and convenient +Patch which removes the QP dependency from the queries module and from highlighter +properties file in order to pass info that could be overridden to the tests +I am not sure whether it is the task fo the sendError method to escape the message +Also as as far as I remember CMSG both process all types of references on phase even if refs to underlying objects where lost in young gen they are still promoted to old gen +applied +Need to verify +Ill check it later +Tom C was looking at this yesterday and I believe he has fixed you update your source and see if this is still an issue +We should update this issue to reflect the version +Thanks ScottYeah Lars this is your old stomping ground +updated titles for screenshots should be retaken from blueSky skinned component +Do not reopen issues against resolved versions if a problem create a new issue for the current form of the problem +They are only used inAre these threads daemonized? I dont see it looking over in Threads +It should be up to the library user to decide whether he wants tosend such a meaningless message +nice catch thanks for reviewing Uwe +Assume closed as resolved and released +we can put links to a wiki page directly on the replicationdistribution just as easily as we can put them next to links on the main admin screen but if you only put them on the main admin screen youve still got to worry about people with bookmarks directly to the page +is it an issue with the ordering of the callbacksim also wondering about your patch +Updated the target for this bug. +Please ignore my weekly stupidity. +Reopening for backport to +I will update the test case soon +Since C doesnt support this notion and I guess it is for some legacy VB reason so Im not inclined to add this feature +updating a row if it doesnt exist yet is not supported +I tried to do the steps you advised but have some problems with my eclipse +Im not sure of the need for this we dont need to provide back compat for anything in the codecs module just the default codec +I am already an Apache committer in various other projects +Prasannacan this issue be resolved +applied in revision Alexey please check that it worksDoes it make sense to this issue since the test is in exclude list again +Let us get back to you with ideas +What sort of security policy do you run with +committed w NEWS mention +Per Santiagos advice Ive rebased my repository and this patch should apply with p +for keeping all of the temporary files under +Hi KiranCan you please run following command in management server and let me know the resultmount t nfsexporthomekirans mntsec +This would also fully resolve this particular issue. +Thats what I did using those tools on their own for some time before taking on this issue +Synchronization on the directory of files is only done when writing the files and only when opening files during openm but not on every access +Fix is in CVS now for rgbColor and iccColor +Updating it to make the title reflect the actual issue since it does not happen always +Attachment has been added with description the transition resolvedfixed bugs to closedfixed +fixed in r. +Actually seems that there is a way to get around this by setting minimumResults to and maximumResults to in the join strategy +has been raised to address issues concerned. +So I implemented something I think will work where the reentrant lock is used if available and if not doug leas unfair lock +Logged In YES useridAfter remove the tag on all get methods it workscorectly +no one actively pursuing. +Had to modify the three implementations of Codec in order to allow extend and override of new method signatures +The patch includes the change to and to +Attachment has been added with description Result after applying the think I know enough now to know that this issue is clearly over my head +Ill let you know once is released +Same problem has Ben +the resources should always be on the classpath +So I think this ticked could be rosolved with Wont fix resolution +Chukwa demux has not changed and this works great +I tested this again still the same result +META +There is no work to be done on the JBM side +I think we can be happy with terminal only especially if this is the ANSI standard way of clearing the consoleIf the rest of the community agrees I would commit it once tests pass +I am going to allow context to set extendedKey name value from Airavata API client +Ning working on it now +Neither are last I heard and if it is a its not an Apache +Updated version with documentation provided as initial content +the patch with a correction +This PDF is inverted when testing with +What would we have the deprecated mBytes setter set the value onIf we are ok with abusing this setter for now we need to definitely open a JIRA to completely get rid of the deprecated method in +Please check that the bug is fixed for you +Reopening for more investigation and at least a documentation update +name Edward J +We arent going to get this in for M +Error to annotate a private method with Init +Looking forward for it +Not everything changes daily perhaps only the published update sites in updatesnightly would be worth synching? Also I agree rsync solution would be MUCH better +lets look into it for next you give us an example file +I just committed this to TRUNK +If then you restart the flows the subscribers wouldnt unregister and I think this was due to the fact that the flow was left with messages pending to complete +It seems to be reproducible as simply as restarting the server +OutdatedNew implementation solving many of the issues already raised here is now covered by +Thanks that works great +The attached patch adds the String arg constructor to +Nvm + +I posted a rd revision on rb that incorporates the review comments and fixes + +there are no tests for the commandsThere are many tests for the shell commands but theyre unfortunately in the HDFS even though is implemented in Common +Please dont catch and ignore the exception in +Committed the patch for trunk +Per review meeting does not fit in features +Test will fail when attempting to match to root +When the patch from is also applied all of the streaming tests pass +thanks vinodbq +Not yet sure this is groovycs fault because our stubs are generated by +I have introduced new optimizedThreshold value on the value of that variable will be used to see if we want to inline or optimize the attachments +Mladen has updated Hudson builds will be in EWS CR and EAP as resolved +For now lets stick this under AMQCPP until we are sure that its not a bug on the client side +Moved service cluster filters to Introduced new logic to generate file at load balancer +If the data dir points to an empty directory ie one that doesnt contain the contents of a data directory then those contents will be created +Just see my comment as a help for people running in the same s not nice when you upgrade your app it breaks and you have to go searching upgrade path was issues that arent going to make it into to issue no longer relevant +let me explainthe Add operation was checking that the entry didnt exist before being added +Giovanni you mentioned that you would test with your patch applied +Thanks to Bob for all your effort in providing the content I needed for this to occur +Duplicate of +to commit for and revision in Committed revision in trunk +Andrew I was going to implement the general idea here of spinning up cluster at start of a test suite and keeping it up while tests run using the suggestion made in +Bulk assign all unscheduled issues with priority major or less to target +We know that Hadoop does not guarantee stable sorting of values in reduce method cf +Also the are related to the level compaction and if I remember correctly youve said in a jira that you want to start the stripe one instead of the level oneTry to organize your thoughts about both compactions what is needed to the store file manager what are the extra metadata field and so on and put them in a document + scroll should not move with horizontal in opera is shown on screen shot +Resolved +It seems that this block query operation should do the sameHowever considering that block details like block generation stamps etc +In addition to the patch Ive added since tags to the javadoc srcSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +Can I close this as a duplicate +Ok but keep in mind that you have to use the in the new optim package +Same problem different shapefile or memorydatastore for correct has correct unit tests. +Ian I see that but the data store is generating feature types that do contain the URI instead makes me think the datastore internally its configured to used that namespace instead of the topp one +After deploying the application to glassfish v the application can be called with httplocalhost. + +This patch solves the problem for me +Thanks for the review Gao. +Ive added extra logging for the cause of the error +The patch for includes integration +Otherwise no jar file is created +Let me ask a potentially silly question hereWhy do we just check that the id is the same? This basically says that the version of the data in ZK was set by the current recoverable but what if the data is different? This seems like an error to me +Good point +It would help if someone could explain exactly why the object is needed in the method context +Thanks +Consider using if you dont need synchronization guarantee +Bulk close for believe all are correct +But is it really that cumbersomeEverything else being equal Id prefer to keep the stuff inside +This issue has been resolved for over a year with no further movement +I think its actually going to be the norm that a given language were binding to wont have native support for all AMQP datatypes and so we will be obliged to fill in the blanks at least enough so that we wont barf on messages that use the full range of types +I think you missed the changes +The problem here is that the framework signals successfull installationupdate when it is done +Logged In YES useridI confirmed that the mdsum for is correct so the download is not the had pkware pkzip for windows version and that failed as described earlier +Thanks Corey +This seems to be still an issue in Grails is the corrected template for. +Screenshot of server editor has been added to the build to say Verified with +For now the issue has been mostly fixed but raises a lower issue that we need to fix before truly closing this issue. +Yep your interpretation is the insertAgain taken from the EJB Spec +Ok this patch has some crazy rejects in it + +A few years ago I proposed a directory structure for binary tarball to look more like standard Linux usr directory as prefix of the binary artifacts for shared project dependencies +Will test it as soon as the snapshots have been updated! +ftl pages that I found while testing this that I will update on the other JIRA page when I get home later +New patch adds the get API in and the API in AFS is changed to return a list of tokens +Currently the block map in and link list is only protected by lock +win and mac os x +Please clarify where you think is in my last comment simply leaves it to logj to find its properties file +Moving items not being worked on afaik out of M Fix this was fixed at some point. +I m thinking to change the code manually rebuild the again and test it to work it out in the worst case +does this make sense +I improved some comments in revision to address a and db and c were addressed in revision as part of +Can you please post the entire exception stacktrace and maybe even attach the +Within AS Ill create an implementation that takes care of the lot +Patch as previously described +Do you have an idea how often the method is normally called? To get around this synchronisation issue we were thinking to use a Singleton object and put this into the Map to indicate not annotations have been found +Rohini thanks a lot for your explanations! This patch added a new function is in which return whether AM restart recover is supported +Applied the prefetch policy changes in the branch +I am also bumping its priority to Critical +Issue may already be resolved may be invalid or may never be fixed +Closing as a duplicate of +added patch for +Adding security tokens will go in another ticket. +does these two problems be resolved? If not i will open another jira and resolve it +for the default setting should be for most of or newbie we should not expect tuning per doc always +Snjezana what do you mean by index the maven repository +is currently failing on trunk and I think that the failure was spurious +So I have just committed some debug statements +Only one difference no SCANDESCRIPTORVERSION constant serialisation in Scan +I want to provide a patchPlease guide me how to contribute to Sonar and Sonar pluginsAny links to documentation or other help will be appreciated +this does not mean that jBPMs workflow pattern support is loosened +Note that to reproduce this an application has to be deployed to and you have to bind to +Updated patch for trunk +Changing to instead of EAGER still produces the necessary i can upload another test got your note in relation to s +This means no generation of child sites in Jenkins or I remove all parentchild relationships +Since all of these encodings are guaranteed to be supported by all Java SE implementations the removed error handlers were unused code in the first placeAll the regression tests ran cleanly with the patch +Sorry for taking up your time +Do not paste code + +Heres an update changing that as well +So you are saying that is not reproducable on EAP? If not then this means it has already been fixed +However I just made one change to the Comparator which could be related +Then my queries are still compatible between Hibernate Core and EJB +Im sorry for that +are already overriding implementation for another reason so here is one more reasonI have forced ignoreWhitespace option as indeed current implementation looks wrong +Cleanup of unclosed issues. +committed to trunk in revision +Hi Dave I have modified the New bean definition +push to +Aslak is this still an issue +all the file I mentioned above are in this the file I mentioned above are in this a dependency into the realm in war deployment plan will solve this problemUpdated deployment plans are attached +File ranges webhdfs uses HTTP header hoop uses query string params offset len +The only clean solution to this issue is to use the metadata cache store +Should we address the testcases to use short names too? I am okay either way we can fix the testcases in a separate jira +Patch is against +Not at the moment just little more time A lot of code here +we should have a design discussion on what we need to do and drill down on sub tasks for thisRight now users using using on the cloud are using colocated strategies +then yes we can remove the override +Regarding the fact that the EL impl seems to be correct I tend to agree with you since I doubt that an issue as big as this one would have gone undetected until now +The import code is rather straightforward +A predicate is just a boolean value expression +But it sounds like that patch is a bigger change so not a candidate for this upcoming release whereas we might as well get this simple fix in +All the duplicate issues are closed but none of them have a fix I found the duplicate that resolves it. +Max to remove AS logo +Any plans +So the cache entry is overwritten +Done should appear on the list in an hour or so +I didnt see a reference to any port so is good enough for me +Passing verifyChecksum to patch looks good to me +Changed to minor +Reduces memory usage but breaks +The patches provide the default configuration to automatically activate the Refactor engine at +Often during development your descriptor does not load successfully +I pulled your check out to a private synchronized method +I can work quite happily with your proposed method I have no specific requirement for them to be comment based so long as Im able to annotate themIve tried your suggested approach and that works just great +However my point is the fault that engine generates for this scenario is not correct +Could you take a look at this one and see if theres anything we can do in timeframe +Ive also committed it to the trunk +Groovy is a dynamic language therefore the types of variables are not defined at the moment of compilation +I think were not going to release any further update to the series but Chris may want to prove me ve checked the datastore in the geotools branch that is using and it seems to be ok should return proper you try it out +Unfortunately it no longer applies cleanly to svn HEAD +Whats this one about +Did you see the same thing +Oh we can get rid of the source I guess +updated the patch to use the method inThat method uses in to +maybe through a simple timeout delay before assignment +Patch checked in +Im on changing the name calling things svalidateXXX is a I still want to see it be possible to vaidate two fields are equal and not have a value attribute on the second field so that we dont have this redundant verifyname field on the backing bean +Assigned to release +As soon as you tell me you are through with your work Ill put some final touches on the sample code and tutorial and close the issue +Registering bundle functionality was floadBundle element into fview is finished +The ant target also came back successfully except for a html page change in the release audit warnings which can be ignoredPatch is ready for review + +And for the NOTICE just add that to our existing or start a new one + to add the right Fix Version +INFO Adding task attemptr to tip taskr for tracker xxx INFO Removed completed task attemptr from xxxx INFO Removed completed task attemptr from xxx INFO Removed completed task attemptr from xxxx +Successfully tested on Jenkins stable version +No particular reason +Verified in CR +please +This is currently presenting issues with the multiple versions of yui we need to keep in our source code control for support +Uploading patch which opens up for using deferrable constraints for primary key and unique constraints +Defer to +I think as far as deletes and such are concerned we need to nail down what we dodont want to do during a flush and during minor compactionsWith we now have taking care of the ordering of duplicate versions so within each they will be ordered with the most recent at the top +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Missed the instructions about unzipping into an existing EAP folder +bq +Thanks did you guys take the snapshot of xdoclets CVS +I would reject this +These are not blockers for BRMS issues are not blockers for BRMS moving to future + The reason is that in we append to the WAL a KV with a timestamp that were not reusing for the can some kind soul please explain why we do not use the same timestamp for the WAL and the Memstore? Thanks in advance +Marking this as a duplicate of. +Thanks for the patch Jesse +Were using that for quite some time now. +This is probably a we still see these messages with the current release please open a new JIRA with RA leaking shouldnt be happening any more m only closing this as done as we reworked the RAs connections. +Regarding boostrapping you are right that you need a custom aware +created for earlier comments in this issue the Wiki is a complex example to take as a reference +Moving out to +updated with basic old resolved issues +patch applied on +If you run a rich client and your local server happens to have problems or isnt running your client finds other folks machines +Add support for the bzip codec +Are you able to extract text from the rest of the documentThose logged exceptions look like warnings indicating that the summary information failed to parse +Once we decide what they should contain it might be something we can add to the QS Tools utility +Even though this is hardcoded on trunk as well it is less likely to be a problem due to the project split +bat hardcoded in the config +Might take me a couple of days to get to it but will see if I can figure what sequence of start restart and uninstall are needed to trigger this every time +You are right I withdraw my earlier comments +ThoughtsI think thats prob OK users wont notice when using filters to search but may get different behavior if they use it for other purposesShay as far as and it seems more powerful to just let apps create a new query type themselves? Thats the nice part of lucenes openness to user query types start with the code for and hook up your own caching logic +please attach a complete quickstart project that fails +The necessary dependencies are added with this ve had the same problems I solved it by just adding looks like the rest is solved through its dependenciesI committed it now could you give it a try pleaseThanks! +CDI for is also good +Indeed! This is actually the case in for spotting this Philippll keep this as resolved in the sense of allowing for customizing this behavior in an +Even though the is specified as a compile time dependency in the security module it can be excluded and CXF falls back to using the implementation in WSSJThere are four configuration values available a A boolean switch to cache nonces +Note Its necessary to click the new subscription button to show the configured nodes +Yeah it wont work forIm uploading a new patch taking your comments +Its running now +Ok makes sense +in all my years of using Spring I have never just injected a raw prototype in a raw singleton +released in +These are almost certainly due to the same issueTheres another similar issue as well but I didnt find it with a trivial search but its there +The issue in JIRA arises when the same RS restarts rather than when going to different server for the same region +are broken so you never has the option to store classpath in eclipse +This program reproduces the bug which is causing and to fail when used with scoped addresses +I havent read the new docs yet but Facets need to play a bigger part maybe putting some of these utilities on the facets makes more sense +To be applied to looks good to me. +This patch applies to both and trunksIt also applies to +Removing the maxExtent in OL also fixes the issue +Here is the sample file and the cpe schema used to cause this forgot to mention Im using +patch +Since the original issue of mnt creation has been gotten over and is a setup issue and since Chandan isnt running into it Im closing this bug + +add following options to generate code which conforms to specificationd jaxbri l this is a temporary solution +bumped to +should be placed in the in order for it to be operable which will be placed inThe Ocean Solr code patch will be attached to this issue +But it would be nice to refer to it but that would mean making it public +The current issue is when we are using serialization to clone the query plan so there is no reason we need to use xml encoder +Lets reopen if we have a need for it +Will it matter if I use the newer flash builder +Hi AndreaI noticed you added the test case so Im assigning this to youCheers sorry I left a debug statement uncommented on and trunk +only part of those files are signed correctly and they are reported by nexusnexus of course verify the key with both x and they are invalid because they really are invalid eg you can download orgcodehausizpackorgcodehausizpackfrom nexus and run gpg command to ve set up a new key that shall be +Patch to add the setter and getter method to class +with JBT +I will wait for being fixed to avoid the risk +I apologize if it ends up sending twiceThis issue was originally pursued as a workaround because the technique used in was frowned upon +The string CNBeforeOTestCGB should be CNBeforeOTestCGBWe should be immune to bad strings like this +However is a bit concerning +This bug has been marked as a duplicate of +I will add this issue to the Release NotesIssue appears as a new feature rather than as a release note +i plan to commit this fix at the end of the day if nobody objects to it +Dont we support using both the SBN and NN though? There are issues requiring the SBN be the checkpointer eg no one is doing checkpointing if the primary dies and you or the SBN still checkpoints but doesnt have enough memory +I dont see in the sources +Resolving. +Hi Marnie All development on this task is complete +Working with Arquillian under ASSeamCDI works natively just fineCheers great! Thanks +What version are you using + +and the leading tildesif you have a better suggestion you are welcome to submit a patch Eric said above that he was open to suggestions to improve this but none were forthcoming +Slava this is one similar to +Forgot to add that the deletion should only be considered as an edge case imo +Looks good this is a really good step forwards +Committed revision +Good +I think its OK for there to be a default value of as long as it doesnt NPE +Im attempting to merge the old xdocs into the new xdocs format +Er +Hi yes custom headers are expected to be honored +MarkPlease the issue when you are ready to contribute a patchOleg +We are using right now WTP and I still reproduce this bugLets wait when we start using WTP with JBT. +Im attaching a version of the with edits throughout +I have updated the documentation to explain how this works +I think I have this fixed in revId but without a test case I cannot be sure +Hi Girish I see that the we are trying to add multiple nic to vm on the same network +changed the groupId toWe might need to change the script to reflect these changes +Im assuming though that nGroups will scale better but your comment here makes me wonder if it indeed is worse +Running Tests run Failures Errors Skipped Time elapsed sec +I have restarted several times and the same behaviour is repeated +For me the job ID is passed correctlyhttplocalhostclientapicommandqueryjobIdresponsejsonsessionkeyD blockers and critical issues are resolved but not verified +There is a corresponding JIRA +Any entry point of a thread should be entirely enclosed in a trycatch to avoid taking down the application +Definitely looking for feedback on this and some other aspects of the patch +Thanks ChrisYour patch is in trunk revision +corrected v includes updated I think the sync should probably be pushed into the Periodic shutdown Batch shouldnt need it +Added a Factory on the interface for creating a default stream consumer Added javadoc for to describe when to use each of the two factory methods Added ASL to add new applied long ago +Closing resolved issues +Hi BernhardYou have attached the two files and not the patch +works except the +Using the base URI from the importing schema document for resolving the URI specified in the instanceI believe the st is at the discretion of the schema processor +I know about the delta code that I applied and I can safely apply it on I go ahead and make this change on then +Thanks for reporting this +Issue fixed since the two last versions. +Resolved as duplicate means We acknowledge the problem but we think its closely related the other bug and fixing one will probably fix both +my fault +Dimitry what do You mean with +What would you use this for +Hi Kiran could you please give a link to issue as its related to that changes +This one applies against branch. +git rev adbafbcbcecefda +The support will be removed moving forward +The issue still exists in RF IE also doesnt work +If you could provide a unit test for the new feature too this would be even betterThank you very much +This seems to be true only for javaappappName and javaModule the services which manage these bindings are not included in the deps for the component thus there is no guarantee that these are still up when the components kicks PR where the app and module context processors account the names of special jndi binding services such as the ones above and then the adds these to the module JNDI binding deps +I can confirm that with the private that Devin just sent me that my testcase now works even if I do not do an explicit set to float before calling set +Keeping the jira open until this is fixed +Thanks Luke for your review and comments! I addressed almost all your points in V patch except for putting in topology related factory class for YARN as I think this is the object created within but not YARN like other objects so in different package +just committed this +Thanks Ali patch committed in r +This maybe a little bit too much for I think you can go with other options listed above for simplifications +All AWT SWING tests done +Just a quick addendum initCause itself is not the problem calling initClause multiple times on the same exception is the main problem +Lots of relative paths are constructed that do not contain a scheme +At first sight should be restored to remove TRACE +Even after thinking about it for a night im convinced the current trunk treatment of setting macro args is superior +Not sure whats the process to the code but you have my permission to use my code in whatever way you want +It is a regression introduced as part of fixing a different bug where the token would only protect you from one duplicate submit attempt but not more than one +Fixed in trunk revision +JIRA closed. +Note that this is already possible in you please provide a link or some additional information how this is possible in +The test failure is unrelated and I didnt included unit tests because this is a small logging fix so Ive committed this patch +The groups that results from there are the documents that are presented +I am in favor of adding in priority again but I just want to be sure that we address the reasons why it was removeddisabled in +Only hope is to ask someone whos done it before. +Committed +patch that just fixing the expected string in test +The fix has been checked in +You are concerned about a single CSV line be split in two among different splits right? No that wont happen because I wrote a custom reader that reads N lines at a time +The fixed worksThanks for the fast reactionBart +I can continue on it +If B marked its sstables as repaired itd never repair +Happens also for users which have not been suggest to remove the deleted from the Issue Title as it also happens when a user was not Martin if this error also happens when the user exists that means that you dont manage to see any review in your context +icon for with error login code with error message box and icon +if we fix our local seamgen our wizard wont work for seamgen +Or as an alternative on Solaris try whoami if it fails try usrucbwhoami specificallyIn either case it is reprehensible that this still isnt fixed +Thanks Jing for reviewing the have committed this +Prepares to add +This means we always have the available that tells us what to doI plan to review this again tonight before committing it +Seems to be resolved in +but at some point I will retest on and reopen if necessary +Let me know if should be done there too +It is possible for the close method to set the threaded state to false while the reader is still active +patch updated to resolve conflicts with recent commits +Same patch as the one I attached to the report +Attaching a patch that includes the package changes necessary to get the tests building against and +committed this. +This is code complete but cannot be tested until has been ported to branch +HelloWould you please try my patchThanksBest regards Andrew Patch committed in revision +defer all issues to +If the changes are not in an release you may reopen this issue or create a new issue +Each test is now isolated and the data is rolled back each time +by putting a member variable in the adb structure to identify it typeI think with that approach we can do this +Reassigning to Andew Flavio +Pls check the latest patch +WAS MQ is not setup in the lab yet +Ive added the info in trunk and will probably publish a new snapshot tomorrow. +I will +will refresh and run tests +Hmm this suggests that there is probably something peculiar about my environment +I will commit a bug fix for this just committed a fix +Request review for the patch +My goal is to get a service release out for the BPMN editor by Feb +The spec files and depmaps have been fixed to include correct dependencies of build version of is +Since there has been no response on merging the fix with and we have pushed out the door Im closing this issue without merging it with the branch. +Some annotations are found but others are not +register would be in accord with the beanMapping element as used in WSDD is a nice idea +I am not sure what the semantics should be +Okay making progress apparently I did not need a docs folder the index project is already set up for this purpose jdeolive is going to look at getting the result actually unpacked and usedJustin can you close this issue when the change is made? +fixed r. +Was just making a general observation +See +Patch that includes changes to changelog and updates the release number to +Committed to +My manual run successfully runs for more than a day +Setting it to does replicate the behavior of removing the setting completely +Closing issue fixed in released version. +I confirm this has to do with the may be doing something wrong but cant find how to import a jar fact ll wait for the final can close this issue or leave it open if you think workbench is missing something here +Test case added in revision +Maybe +Duplicate of +Bulk close for +Hi guysAny news on this issuekind regards Bobthanks for asking go ahead and index over the weekend pleaseLet us know when done so we can close this issue thanks +Went through the call stack to ensure the behavior is not changed +As it turns out this was caused by an application problem +Its working on Fedora with Sun Java at least +Added properties to +Thanks Alejandro! +Applied proposed patch +If you find further issues please create a new jira and link it to this one. +Later min replication is changed to same as number of datanodes available and cluster is restartedIf the environment configured like this I run the test case successPls attach the failure case log if it is convenient +Haha +braces on new line +mark would say that with this you should be able to explicit control the naming of aliases and thus handle the level column +see patch attached to bug it implements configurable buffer size andoutput encoding +I will abandon the JAR file creation and setup the plugin to properly provide the ZIP file as default packaging +Fixed in rev +See linked issue +Im wondering about the precise impact to GC that these changes will introduce +. +Bulk assign all unscheduled issues with priority major or less to target +Yes it seems wonky maybe billing metrics is a good idea +I just tried and it works for me +I dont think deleting the sstables is the right thing to do since the loader didnt create those files and they are potentially very expensive to recreate +Fixed for Seam EL Resolver +The way it usually runs is we have a fat umbrella issue like this one a bunch of work gets done under this umbrella in this case a bunch of the above will be addressed by the patch but then subsequent amendments or additions get done in separate issues +However if VM kills these daemon threads during their shutdown they may left monitors lockedThere is alternative idea proposed by Andrey never let agent daemon threads to finish until VM kills them at shutdown +For in release +If you dont want to register a listener how do you propose this be accomplished course we do need to use a JMS listener to implement the requiredfunctionality +of them are in the related to and s new classes + +As mentioned before I am going to improve decomission performance +To decide whats common the POI build script looks at what is used in the unit tests so adding a test that uses a method will cause the appropriate parts to be included in the next you able to share one of the files that triggers this? The easiest way to add the unit test to POI that would have the classes included is with a file that triggers the have the same problem +The patch actually applies fine +lets push it for now +I have a couple comments here We need good reporting on how many records are bad in terms of count and which splitfile they came from +Applied to TRUNK +The main thing I strived for was total separation from the main build +The code should probably be deleted +Highly scalable and distributed structured data services are an essential feature for todays enterprise applications +Tested this with and Im seeing the same behavior +So Oracle JDK is overkill but it does make it absolutely clear to readers who dont know about the reservation of JDK for solely docs will go with Oracle JDK until advised Sun and will be amended to or +Thanks Koji +this is schemavalidator being too strict +Youre right it was a browser cache issue. +Ill leave it as an exercise for the student to figure out what the datanode really does when it needs to look up a blockIm glad youre enthusiastic about improving Hadoop +Sorry about the time it took to get this in +Mavens not stopping you doing anything here it just simply cant support every new feature without getting too complex to use +Im reading up on the Sonatype Forge right now and might be able to get this workingAs for Hadoop is there any requirement that we manage the jar lib for Hadoop in Nutch? Couldnt we simply pull the jar down dynamically via ivy or via some magic in? I think if we could then we could simply remove NUTCHlib whic was my original intentionIn the meanwhile Ill create another issue to track all this but you have my to commit your patch and mark this issue as resolved +Ive opened for that issue +Such a allows for flexible access including different cases for column names +All tests are working now +Thanks. +Mladen as our RHEL build uses apr from base we may not have explicitly included the fix in our httpd RHEL build +bulk defer minor issues to +Ill generate a patch which does the swap +Yes on it +SureshYes I completely agree about the name of the JIRA once Virag agrees that is the real fix for the issue I will probably close this JIRA as a dupe of and then file a separate one for parsing the stack trace as new work +I need to see the traffic between and +The first time itll download all dependencies so the internal authentication provider from Ivy will be installed up and running getting the proper credentials +Initial patch for review +Lets postpone this to since I may not be able to deliver a stable solution before +Or we move the all these to core along with Randomizers and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Moving out for now +This should be to test if the file exists rather than catching the exception +Proposed patch +The following should also be added to svnignore trunktmp trunkutilsMakefilerun trunketccouchdb trunkbin adds a couchjsdev script which launches the JS view server without requiring make install +can you please provide more details to thistnx for input +Thanks Daryn and Kihwal +Done for the concluding parts for strutstoolstutorial jsftoolstutorial and jsftoolsrefguide are added +I agree we should be in there but we should only add the if it can be done without the welcome visual layout of solves that pretty nicely +Im keeping this issue open for documentation purposes but I will link it as superseded by +Something like a Hide Task List link along with the other links +I would even argue that the log level should be lowered to DEBUG if DB exposes warnings for such perfectly acceptable conditions there +I also modified the unit tests to not use the JMX helper class and instead use the long way around to create the proxy +HMS tracks the configuration history and store the state changes in Zookeeper + file container patch and additional jar file for web checked the patch however it does not compile on the current trunk looks like you have to issue the command TWO times in the webservices directory +Great! The build helper would be a great home until the feature is added into core maven +Verified. + +Fixed in revision +Work completed +etc +ouch +We have considered something like this with xincludes and a mode +This issue is a duplicate of the +so handing it over to Dave GYou should still make sure that does not occur in the manual other than in the list of defined operators and in the API section around arraysregiondistpoint +I wouldnt do away with the simple archetype because they are ideal for people starting out +Committed to trunk. +Attaching the patch for version issue was found in all and versions +You are right Jared +Joram My previous comment was indeed not in sync with what we decided before +This bug causes the behavior described by +Try itThanksJayachandra +Heres the output from the repro in my environmentThread failed No suitable driver found for jdbcderbydbcreatetrueThread succeededExpected output isThread succeededThread does not appear to be a problem on Java +We use database strings for the in our HA datasource +have you finished the build and test of this plugin? I am interested in this plugin for ms tfs feature hasnt been implemented yet +correctly maintains the file size of a k text fileand a k gif +Fixed in rev +asavu added to! +This settings forces default realm to be set to which is in environments with different realm namesA better way is to set this property directly in the functional tests requiring this config file +Commit the changes to trunk At revision +This patch adds a compaction request when an starts up if there are too many files or if the store has references +Changed the patch to handle changes to work with ydist to patch to apply on trunk and added comments by the new patch moved metering along with other counters +I dont think weve ever really had a situation like this +I also think opening the file seems to +There are shuffle methods inThanks! that works very nicely +NB in the container runtime to update to slfj should be upgraded also +Please provide some information as to which version of Android and which device you were using and we can try to reproduce it again here +This patch remove the dependancy between and +Applied +You forgot to mention that you also changed the name of the thread +Please rethrow another NCDFE this is an Error +The in are a segway to full datetime support +please choose a Fix Version for this JIRA +Bug still exists +Both truncatefalse and truncatetrue work as expected +We can do this for version with the submitted May take up to hours for the fix to reach central repo. +In this particular case I feel compelled to rename field to uid in order to allow jBPM to run with Hibernate versions prior to +I am not very au fait with Buildbot +Adding this shouldnt be too difficult +Then once is fixed you can close this ticketThe problem is that if another user wants to report the same issue he will probably have a look the open tickets +My guess is its something to do w conflicting names and I wonder whats the difference since its JSR is included in Java and Ive not seen reports from others. +Hopefully soon. +Have you an idea on how you want to do that already? Have you committed anything in the CVS repository or should I start from scratch? Cheers +Simple and useful changes +How about a name like or? That would indicate a bit larger scope for the component +Thanks work fine +Thanks. +you need to enable a remove JMX port if the in process mbean server is not visiblesee the SUNJMX env entry in the activemq script +Hi Im using SCM activity Sonar plugin not Maven SCM Plugin nor Maven Release PluginSure thats a maven based project and the quality control blocks when running mvn sonarsonar specially in SCM retrieve +Closing old resolved issues +Yepthats why I added the bit about proper split policiesNothing stops us from setting up +If you perfrom about steps logs are not writing to and api log these files are always blank Also you can see the owner of these files are root it should suppose to be cloud This need to be handled in packaging +I am not sure how he is migrating users over but doing the change password will help him will know what they are using to hash passwords +I see same failure on another patch as well +NP and yes a singleton map is better than +Deep in EAP GA panic at the moment +Also because repairs are based on cfs the gossiped object needs to be on a cf basis not on a node basis so application state isnt granular enough to capture this additional stateI think the possibilities are My example is wrong and Im missing a component The new value needs to be gossiped The new value can be incorporated into application state somehow Coordinator tells all nodes about the state of the repair +It looks like some other test is reducing the lock timeouts without resetting them so is seconds not the expected seconds when the test is run as part of +eclipseplugins +Splendid! When I saw the issue yesterday I was a bit afraid to have to spend a few hours understanding what could have been the problem +The retro build can create a compatible version for so the usage of that class is no problem +Do you have special consideration here? Maybe we should mark all as Unstable for the initial AHSbq +Ive just updated the tika bundles to version for +This bug could be related to reliability test crash described in a link to to the test +Pipes still needs to be handled +Yes thats right +The new patch fixes this +Just to expand a little on the use cases where a custom messages map is useful Plug in a that conditionally caches properties Use an alternative algorithm to load property files +Closing the issue now that it has been resolved. +OK cool +Please check that it was applied as expected +As cant know whether a method or a sequence of methods are idempotent it should probably assume they are not +You can close this issue +Please review the patch +when jbm is done ill assign it to Martha +Verified by Alexander. +so why not just build this into the ujax post servlet +Ive resolved a small version property issue in interceptors interceptorsexchangeIf there are no other issues Ill proceed to commit this demo +Thanks Joy +I see +For example the enterprise workspace is and the ownerid version is +It should add segment in path before creating runtime and server +Saving password OK but when I uncheck Save password click Next then Cancel close the wizard and then Restart Eclipse and try to import application the password is filled in the password field +StefanThe stars didnt align for me porting over to during my last release but Im hoping to look at porting over to in the next month or two +buildhamahomeedwardworkspacebin +Good enough for now +MikhailThanks for sharing your experienceCan you disclose the number of nodes in the cluster where the test runner executesI wonder if this model can be applied to Apache Jenkins where build machines are shared among many projectsI dont see +Screenshots posted on cwiki showed minor grammatical errors in the paragraph below How will you be authenticated for the hybrid version +Many thanks JuliusOleg +fix unit test failure caused by the patch. +How to reproduce configure topictestTopic on your instance start instance deploy in attached zip shutdown of is a kind of duplicate but the cause seems to be different. +It shows only classes from sources +Thanks Uma! +Thanks Kan! The new patch reflects your comments +Thanks Shwetha. +Im also unchecking the Regression box because this behavior change is a bug fix which brings us into compliance with the JDBC spec +I remember why this is so +How to +Now fixed +This will be great contribution +They are related to JMX but not documentation +Okay guys I think I updated the to have the old methods and marked them as deprecated +Dont have the bandwidth to work on this at the moment +I believe the specific situations are starting to be addressed by other Jira tasks such There may be an adjustment needed to the lookup code so that it picks randomly from the list of returned +with thanks +I personally feel uncomfortable with code that catches Errors or works with Throwables instead of Exceptions +Any Grails snapshots built after the date of this comment should work. +See for trace +Now after a few days the heap usage is still low so I can confirm that the only cause of the problem is missing recovery plugin in the configuration of xa datasource +It always uses the cached version what is really annoying developing a appplication and a used plugin parallel +Need to add a test Im just stashing this preliminary patch here so I dont lose it +This is too complicated to do generically so marking as WONTFIX. +Decoupling policy impl from policy api so the validator only needs to reference the policy enforcer interfaces +Committed revision order to move forward with and other improvements we need to fixBut this is difficult when the Collation support exists mixed with the +In the future I will take care of thatShould I update this patch based on your review or is that okay for now +este fichero anexado es un documento se aade un comentario despues de haber creado la tarea y haber aadido un fichero word +Marking as resolved under the assumption that the are properly assigned as tested elsewhere +However if a plugin does not have native bits this entry is not needed +This is significantly large of a task that will likely changeIve made the issue type epic + +snapshot from trunk with patch applied +I dont see it scheduled for a fix version +This patch close the reader at the end of the start method however working on this issue Ive found a bigger one +Hi Miha Poberaj +Wont be working on these and theyre old +These names are probably a bit more intuitive to people unversed in IR theory +I dont see how +Would you like to try it +Hi could someone review this? Thanks +Just gave it one last go and the patch applied cleanly +Im working on a fix for that issue +Can you provide more detail on what you are trying to accomplish? What would be a use case for more variables +In my experience the ejb interfaces cant be found from a war even if they are elsewhere in an ear such as in the ejb jar file +If we are going to change that then a lot of places will have to change +Verified works again w +Network acls are not cleared even after network garbage collector has run +Since is the one thing that cannot be modularized this code base needs to be kept as small and minimal as possible +This is infrequent enough that the multiple logs uses should be sufficient protectionThis is not a quick fix +Im working on RC today so Ill add this in for the next RC +contains more description and the patch file +what was the SVN commit number? thx chris +Apache Felix Declarative Services has been released +Fixed before rev +There have been several fixes applied to woden in prep for the November Interop +In this case we can lost log you checked this usecase with on Tomcat +Unsetting Fix Version on Unassigned issues +Extract of server log +Log data source definitions and VDB are attached +Yeah that should definitely be separated +bulk defer of featurestaskswishes from to +The new binding framework along with current discussions on buffering are enough to solve this issue going forward. +no one else has mixed logs +Build the new tree +ThanksIll resolve this issue without Fix Versions since its the issue for unreleased branch. +Passing to you andre +I have added the and it does not generate the package correctly +Merged previously +bq +It would be great to get this committed at least as a starting point so we can do more extensive testing +You need the attached Patch reconnecting for Queues +I have tried to install and uninstall other plugins with P the plugins are also kept in plugins directory +I believe Jersey already supports this? vote for this implemented tooAnd wonder about WSDL generation +A quick look at the patch seems good but this is undoing some of the work done by +In this case we want to choose not slfjThis really has nothing to do with utility libraries +Relevant patch from to and master. +Ok I fixed this and hopefully any other edge I nuked bracelists in groovy mode entirely this seems to work quite nicely and the few cases where there is actually a bracelist it still indents ok +Stuarts above idea or something like it could work +New patch addresses all the comments +I am sorryI have overlooked that option in the server settings +The SQL for copying updatetime to pubtime would be update weblogentry set pubtimeupdate time +Oh I had no idea +Committed revision . +Hmm +will commit in a day or if no one objectsThanks for reporting this revision Thanks Frank +Hmm the webView dimensions ARE x which are correctIf I remove height it conforms to the right dimensions +New version patch in attachments +Updated fix applied the settings only if keep is false the policy is only applied if durable subs go inactive otherwise they retain their original configuration +I am confused there is no class in geotools graph +Kos the ACL proposed in actually used a space character as the list separator + objections to this +The reason of using is that the interface reworkMapred can also be used by other formats in future like some other file format also want to change the mapred work depends on the inputwhat do you think +The question here is that upon receiving a should we really call resign or not +Has to fix some glitches +Do I need to specify it in the config within the archive too +Jaikiran Any idea when this problem could be resolved? Thanks +Responded to comments on RB and made needed changes +If I were to extend the framework Id like for it to be done in a way that provided a general solution not one specific to Terracotta +Thanks Shubham for your patience on this one +Sorry will port conversation over there +However that can be decided later on and I have removed the configkey update with this patchI challenge you to find a benchmark that produces meaningfully improved results with only running the check every seconds instead of every +This is a duplicate of. +Proposed code patch on behalf of Tore Eide Andersen with thanks +but Id rather just log the start +So You can investigate is test running and why it doesnt catch this case +Closing all the edits mentioned previously in this JIRA are in the docs. +To do it the functionality should be separated +fixed this problemThere is no need of ieurl in modal window code +Master would not get notification that a region server is opening a region only its taking longer than usualAny other suggestions +A patch +For us making the fields protected is good enoughBtw I noticed the fix version is now for this issue +Verified on JBT. +I just committed this +skh pushing to this valid any more +This is now fixed in the latest EAP branch +Fix Versions since this should probably not appear in Release Notes +Any comment +Ill be reviewing this patch shortly +file +The other option is to pass the Rtf straight to the normal Rtf parser have that processed inline then handle the attachments as embedded entries +If no one else objects Ill mark this as fixed +It preserves the critical notion of timezone and daylight savings time and of course lets you express time in UTC as well if thats what you want +This will ensure that the Preview are all synchronized +Just filed to address thisI will review your new patch soon +I have the core tests working +Could you please reopen and fix this isseregards upgrading to Seam Im experiencing this issue again +srcmainjavaorgapachehadoopyarnapplicationsunmanagedamlauncher +You can summarize that at the end of the test run +It seams this is a bug of WTP +Yea actually an int would probably do we shouldnt ever have B rows in a region +Patch to class populating the wrapped attribute of the class while the bindings are being populated +Can you regenerate the patchI am getting a lot of return false if number of rows foundElse it looks some problem with last modified piece of codes +Bounce to Doug +I can become maintainer and push this to extension level if youre ok with this +worked out pretty nicely i think +Unfortunately it still doesnt fix the M build for me but hopefully it will for you. +I have passed him the use case and datasets for reproducing the issueIs there a way we can assign it to him! Assigned to Thiruvel +The mbean has been moved back to the conf with a velocity template that addsremoves the commented out mbean based on whether the pack was selected. +patch that mark race by special API for prevention of further alarms should be verified +The question is how these components should handle blanksnullsWorkaround dont use Booleans create a enum with values ON OFF UNKNOWN +Frank could you please resend this patch againts the latest CVS? I cant get it applied +Ive tried this test several times but it works +orgapachehadoophbase +I can definitely envision the use case where debug logging would want to be asynchronous while audit logging would not +What operating system webserver is this failing on? You mention XP IIS Resin for your local development but you dont mention anything about the server that its having problems on +if you look at the menu on left youl see that it is kept as an online example to let users discover changes report renderingthe content of the could probably be enhanced to better show that it is only an example and not the effective s changes report if the sentences werent about a plugin for example it would be clearer definitely the report is useful as it confused at least one user on the user mailing list +Do you still want ADF Faces added +Since the +LGTM +I think this is caused by not unplugging the Socket in when the master replication functionality is turned off +Cool +Usage problem +Specifically a little nervous about how components are handled but I did add some specific tests for them and those worked fine +Closing this as wont fix +I see now that the solution is agreed with Nathan and support it +ok +George can you file it as a separate bug? This ones been beaten to you too +what do you thinkMoreover at the moment theres no check that the path you want to go to is present or not for examplepwdtypeServercd foopwdtypeServerfooeven though the foo dir doesnt exist. +Eclipse project for reproducing the testing application attaching +strange +ear file +Unless there is further input Im going to close this ticket. +Please review why we have added the below in getent passwd cloud grep q varlibcloud sed i svarlibcloudmanagementvarcloudstackmanagementg etcpasswdThat is the reason for this issue +Rave already has a content negotiating view resolver that can serialize to XML or JSON +on the revert patch +It should show up in the next nightly snapshot +I think it happens when you click on your link but you havent configured it as allowed + +This JIRA can be closed as this feature will no longer be available from GEP onwards +Some thoughts we should start work on merging this because I dont think it will get sufficient review outside of trunk unless there is a performance cost it should always be compiled in but default to disabled at runtime when merging break it down into smaller uncontroversial pieces for example add the configuration options first then some infrastructure then more functional pieces try to avoid the unnecessary whitespace and general fix changes do we really need another linked list? it would be great to have some unit tests +Hi Aidan can you review this change for me please +It checks if the implementation support and if not it throw an once Blob is given as +I dont think it would be a good idea to integrate it as is and later have to change a lotfrom what i have seen at a first glance i dont think that there is need for lots of changesin any case i will update this issue as soon as im through and list all things that would from mypoint of view conflict with the jsr access control sectionbe asserted that i dont want to do the integration work twice and incomplete list of notes and comments based on my recent work on JSR security features +The current code does use the schema if it has been closing all issues that have been in resolved state for months or more without any feedback or update +The afforementioned tick is is the revised patch +Under if the element with attributesis wrapped in a the attributes are returned +Graeme it looks as though you fixed this in Septembers! for verifying +So it simply comes down to a matter of preference and to add additional ways to configure the source generator +I am attaching a patch for junit test +I just committed this one too +It has been successfully tested here in HamburgThanks in advance SvanteYour patch review pass you can push it +bq +bulk defer of featurestaskswishes from to +But it will probably require increasing the file descriptor limit +Update docs +Agreed with the aboveAction item add to. +Previous versions remain until specifically removed +Is this a regression +Patch removes the volatile variable +Frankcan you create a patch that brings the ADF Faces selector to Trinidad +This patch fixes REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILEDWill be filing the following remaining failures into separate tickets REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILEDNote that for the following SDK API regression tests to pass SDKAPI SDKAPI variable must be set to +Thanks Cheng +At least I dont see any in my local copy and Ive been using maven to build it +Recommend you supply some patches and even better tests for the issues you are seeing +Note that I had to replace with to use the mail notification feature in jbpm +attach my ve just committed this. +This isnt the include tag or the action tag not cleaning up the problem is that the parameter interceptor is being called when you include the action or call the wwaction tag +In most cases this preference page is not needed at all since KB will be enabled automatically or by Quick FixesVPE +And I bet there are a lot more +Since the classes have no functionality at all I dont see much point in leaving them in the codeline +On a side note the class is a part of the package +Basically the problem is solved but I suggest should +But whatever you think +I was intending to post back but you beat me to itRegards Jim +I will close this based on the resolution of that issue +Looking good +But they should be deletable with a different rm option +Shelly I tried with yesterdays snapshot +Fix of previous attachment +I would like to understand why this is happening with only this host which is why I am replying to you +When VDB is undeployed the prepareplan cache and resultset cache are flushed +I will take a look +Hi AnkitI tried that i get the same error +Confirmed issue with NB Java final latest released mevenide getting Broken contract class when trying to do autocompletion +Sorry I had missed that other issue. +Can be scriptedautomated +having developers to either change the version or using a new local repo per branch is nearly is is a suicidal to ask developer to do way is a problem but the fix will create more problem +This is here for people to throw darts at +i will get round to it at some point but i dont know was just wondering if this patch is likely to make it into a releaseEntity interfaces would help me avoid some nasty workarounds +Ok +Duplicate of +If we can fix the extra job before release I am fine to commit it +so it will appear in the release notes. +Playing with the synchronization here is dangerous +Ill let you know if I have any trouble with that +This should fix the bug +defer all issues to +I can change it back and get the exact error if youd like but its the standard Cocoon file not found error with the path of the redirect +Attachment has been added with description Trivial patch for to fix image scaling +Im using Win Enterprise x SP +though shouldnt be depending on Sorry my mistake I thought returned anbq +Should now load in the full deps and classpath +I think this patch is solid and have committed it to the trunk asrevision +WFS? On the ml I thought you were talking about WMS? I dont think we have any filter limit in or were you talking about instead? Many people confuse the concepts and think its the same as WFS but its notI can see the limit of filters hard coded in the I think you are right and it is +was the original issue is that we wanted to set the encoding for the request before sitemesh got a hold of it? If so what about having the filter dispatcher set a servletcontext attribute with the encoding which the cleanup filter will use to set the encoding +Now the default is jdbc compliance + from the test report for this of solved in +The latest upload include a few fairly minor cosmetic put the hthreadt pointer back in struct VMthread and my patch +This will prevent for example two layers of encryption from being enabled +I personally disagree Perl is a better alternative to Cygwin +See revision . +Flavio If set is called after the object is constructed then it has no effect on the number of threads in the poolyes ur right +I will submit it later today +Joao how would you see this functionality working? Would you simply think a Properties menu on the repository connection would show the properties of the repository +Scheduled for this weekend +Added a forum post in case anybody has input on this +no it should be passed in as a parameter so you can control what importcontext gets usedwe should also consider moving dao related stuff to a daohelper or something +Something being done twice is easier to overlook than something not being done +SiI dont believe the problem is resolved +What do you or others think about the possible performance hit? Not a big deal? Or worth keeping the old code around for a while +I will commit it after running tests +ArunMy feeling is that JIRA should be duplicated to which fixes this problem more systematically +Having those methods in and may cause misuses +Rebased +Raising the issue priority to critical as it breaks the promised backwards compatibility for AS +The number of threads went from under to almost in each of the instances in less than one minute +I think you dont need two Estimators just overload one to serve both purposes +I think the way is to just add a I looked at doing this as part of general purpose codecs and it gets really hairy and the use case is very atypical +Patch to remove th method patch patch which also removes the overriding method +So what does actually do? Can you reproduce this without using Sencha +It deploys in my limited testing but its not very useful by itself +Patch v integrated to trunkThanks for the patch HimanshuThanks for the reviews Lars and +but i would like us to discuss already today if there was really no other way and how we are going to address this in OAK before we ship that featuretherefore i was somehow waiting for a confirmation that my concern has been noticedsure I agree and will be happy to discuss a proper way to handle that in Oak if we end up with releasing it with that entry in the i would like to see us having at least a FIXME comment associated with the as this is not secure at all +Please check +That said I think it would be much less for the users and easier to understand and debug if you followed the Hadoop API much closer +had a thought since and I think its easy enough to fix it directly inside the +This issue is related to SMX Components +Reopen we still see data loss +AhmadNeed some details for this task pleaseregarding browse the FTP dir part which FTP service are we talking about here? feature will not be able to make to +patch +Its also possible that older versions of the mapping contained more indexes in which case the byte limit holds +still requieres JSF stack trace on error feature is working as expected +Trunks seems to die at threads. +APL license is effective as ofI will look into local and xnio support in Netty +It works through all four patches posted earlier for +Patch looks good +This patch is a little simpler than what I have proposed +I wasnt intending to checkin the draft patch just threw it up here for people to take a look and give feedback +Makes since +I thought so I just had to askMaybe I could fake my own version of hibernate for linux +Also I merged your fix for the error listener into my patch which has been checked in at revision Please grab that revision and let me know if it works for you or not +for this patch for now all good comments theres plenty of improvements we can make so lets line them up as separate issues +in head +Lets name it as Web perspective already exists +i would never store passwords as plain openejb we support encryption +It always adds the column as nullable +sergey v +Am I right that Sebastians latest changes kind of incorporate this? The last patch here deals with but Sebastian moved this out to +Here goes the third patch for Pipes implementation and test +This looks fine but also depends on wc so Im unsure of the net gain since its just trading bash for wc +Hi JimI have added your mirror and should appear soonOur current max image size is x yours is xIf you can provide an image that size Ill add it inThanksGav +Closing this based on the exchange above +nm +The date in the checkpoint reflects data SENT not data WRITTEN +Wouldnt it make more sense to just make the update do all updates as a single transaction +Let me run the tests +There were two problems with the solution I was pursuing It required JDK features It didnt work in the Network Server case only in the embedded caseMike suggested some other implementations that might have resolved these problems but Im not currently able to spend enough time to pursue these solutions so Im unassigning myself from the issue in the hopes that somebody else might be interestedIf after a while nobody else appears to be interested Ill close this as +I will provide an updated patch when finishedBTW I love good testcases +As a temporary workaround I commented out line inI think this should be promoted to blocker since it grails doc fails and consequently grails too +do you have a unit test as wellthen I would integrate the patches to +Instead wait until the enum is full constructed before populating the map +IMPORTANT Patch does not compile +We unfortunately do not have an supported controls hence your search is not returning anything +It now produces the following compiler errorD +Simon thats great if you would like to only thing is i would like to try to do contrib ones toowill do no problem +I really would prefer to see something with the server where we could fix some of this in itself so that we dont get the automatic content type for web services +Will post another patch that describes these configs in +We somehow have to different types of input for the content provider once the and the which the view should show +This is a serious issue +Not present in branch +Thanks Gianmarco that much cleaner +I assume you mean real distribution classes and the base class is +Marking Patch Available +Sorry for response so lateIvan I have modified the code and the issues is gone now +Another idea I can configure Policeman Jenkins to sometimes run test without compressed oops +So the invocation looks the same but you can turn the encoding on or off as needed +It doesnt make your contribution any less important +But Im pretty sure the buffer for writing it unique to the +But i am not able to figure out how to know the exact delimiter directly from the database +Using the two attached xsd files with the case does not cause any dependancy issues that I could see only the compile issues +Patch is committed at I am not closing this bug because I dont yet know how an agent has to react to the error returned by in Java +In this specific instance I defer to your judgement. +nice to have not priority +Install scripts for he following projects use LIBDIR with included PREFIX datafu tomcat whirr hive jsvc pig +Thank you Benoy +after resolving add fix version to bug and resolve it to in branches revision +cannot use set with value from Also fixes the test to fail if we do not get an exception for negative test cases +Ruby encoding pragmas would probably need to be part of this discussion as well one could argue that the ruby side of things works unintentionally +From the recent discussion we concluded we dont need any core API modification +for exmaple it shows up with clean was simply a missing mkdir on the writing of probably happened was that you edited cygwins +Patched anew +is undefined +Hi RobI am reopening because I had another use case in mindImagine network of brokers where I dynamically add brokers for scalability +Attached class extends the payloadmapping to take the jaxb annotations into consideration +Also the static API should be overloaded to provide the timeout parameter +If someone else could try and just let me know if it works for you or not forget actually applying the patches for real that would be great +This is a planned feature for seeEven with pipelining the two requests have to be properly serialized +I appreciate the reminder +Sorry I will provide it if I have time +and moved it to common +Will fix shortly +Committed the patch to trunk with revision +The use case that prompted the request is the code that registers listeners do not initialize Persistence +Verified on upgraded setup +Are you trying to register the reader for a new service? Or an extension of one that already exists in the core +Ive attached an updated patch +Lets try again +Its not properly tested but it works with the current test without the and whichever way round the formats are specified +The change looks good to me +Not easy to fix since the import rewriting mechanism is borrowed from JDT and all are replaced with +Fixes a small bug in serialization Adds the ability to either get the status of a remove operation taking place or force a remove operation to finish immediately Tests for removing tokens Move shared code for creating a ring to Util classRemoval Process Normal Case Coordinator sets status of failed node to REMOVING Coordinator blocks on confirmation from other nodes Any newly responsible nodes stream data Newly responsible nodes send confirmation once all data has streamed Coordinator updates status of failed node to LEFT Done Failure Cases Coordinator failure If the coordinator fails the remove operation will need to be retried This can be done on any node in the cluster +Emmanuel Ive used the snapshot lib you suggested but that didnt change anything +These are working in the latest patch +So was a change made to the Groovy plugin or was a change made to the plugin that fixed this issue? I have tried getting the latest plugin but I still get the same error +And then creates a watcher on that +Darrin DavidPlease make sure these things are clearly documented for users in the installation notesFor instance someone using on EWS how they configure users etc +r should fix this problem +Why is this bug supposed to affect? As I posted it affects the current version +with Scott missed that also +disappeared from table of docs will be deteled from info about the attribute is deleted from the guide +trying hudson again +So if for a entry you actually fake it saying its indexed but there is not index entry this is equivalent to saying there is no data in that row. +Merged into master +I think I know what the problem is +Revision srcsrccommunitysrctestjavaorggeoservertestsrccommunitysrctestjavaorggeoservertest +It has been verified that when the night test changed to use a temp directory it had access to the test has not failed anymore +The patch committed to trunk and branch. +But I dont know how to implement the change +This can be avoided by applying the proposed patchRegardsSenaka +does not allow specifying an output path that is different from the default output path +even if you would be right you should introduce a parsing mechanism into Roo shell to reject improperly spelled fields support for composite keys is limited +This is the implementation class with some changes to make the output more readable +For Blobs in particular I dont think theres a way that Cordova could fix it up that would not be horribly inefficient. +on that approach +I think the invariant that we want here is do not delete log files with takes with a sequence number higher than the sequence number of any other active put +What is your umask? Did you set some special access right on the directory where you publish? Did it worked differently with ivy +ok refined the test +Just remember to remove the usernamepassword from the wireshark file after you copture it +I see absolutely no reason why we should still support +Thank you Benjamin +Heres a patch that approaches the problem from the RS side +Unfortunately what you did was clone the issue twice so I was cleaning up the mess unfortunately I cleaned up the one you were working on +patch appears to have done the trick +Fixed per submitted patch +Changed the indentation and also using a in revision Thanks for the hint. +This patch includes a fix for +Thanks for that feedback Dag +We never know what new and related errors could arise from the Create a PO process +It should apply to current trunk do you experience any problems +A summary of the changes Setting and updating MR app progress as a function of progress of setup map reduce and cleanup tasks +This was against version pgSQL version +Verified in ER +Parser Changes a work in progress +Add Mark and Josh as they are generally aligned except for order due to those tricky However as Im looking at the signature now Im wondering if people might see Class as an argument type and assume its for dealing with Configuration classes? +k +At the risk of throwing a red herring in here but we can accomplish all of this using a site compiler tool like Awestruct +so the only problem would come if the annotation classes were changed during the moveI cant see why we would need to do that +This issue is no longer relevant for it is fixed inChanging the affected version to. + +Ive committed to trunk and +Not issue in JBT +I think my comments can be addressed later on maybe and shouldnt stop these improvements from going forward so the way obviously since you have been doing all the work here i dont want you to read this as me questioningobjecting to the change just trying to maybe help save you some sanity +Next create a blob store aware cluster state store +This issue is still present in Continuum with Maven +trunk Added max redeliveries and dead letter Update wiki documentation about max redelivery attempts and dead letter wiki all resolved tickets from or older +Thanks Ahmed for reviewing the patch +you can run this test in any version of maven to check for this have not been able to replicate the issue anywhere +Changes have appeared in Latest Alpha Manuals so closing. +It just illustrates the condition by means you think of entering into a loop no +kind of decoupling can seriously help in workaround for the problem +required small changes in now is a part of public and as a resultrequired a lot of changes cause there are a lot of changes in +You can use a custom for this +This is covered in the migration notes regarding how import projects and the steps to resolve compile errors. +Here is an updated patch incorporating your comments +change the are right YonikHits currently tries to hide this by normalizing the scores to a maximum of simply by dividing the raw scores by the maximum score returnedThis is why the scores from Hits are currently not comparable to each other +The branch had diverged a bit more so I resolved one more conflict +I just committed this +DavidI would like to submit a patch for thisIf it is the same way as KVM should be added to SPEC of RPM +I honestly dont understand the difference but I agree the code is really simple + +I created for removing +Simple implementation without nb successfailureerror per project +There have been years of bug fixes and improvements since the temp build +Or am I missing a piece of complexityCheers. +The build was fine and derbyall ran cleanly +Can you provide more detailed description on the change? code handled the case whennext called with limit! +Have you tried to do cache clean +Ill follow whatever convention that you guys want to introduce +Moving all issues to the CR +Even ths of seconds do +Please change the parent type of application to application skipped one which looked to be incorrectly switching an OLE filetype to an OOXML one +Jose mentioned theres a toolkit for arquillian testing coverage jacoco +Bumping this to as there are a few additional possibilities I want to discuss with you Tomo +A component in what sense +Heres a cumulative patch that wraps the methods and the method in a retry loopI tested this using the unit tests with a remote MSSQL database +in +a regular empty packet and be handled exactly the same as a regular data packet or +Checked in the implementation code +Dave any plans to come back to this issue +You are right though that having compression block not aligned with data boundaries could mean that sometimes well have to decompress one more block that would be strictly necessary but thats more a drawback of having compression blocks not being aligned to any data boundaries than anything related to this actual patch +Nicolas For your trace Id recommend running with the actual Java concurrency libraries rather than the backport +This is a duplicate of please see that issue for the cause and workaround +Here is my console logCtoolseclipseeclipsec consoleosgiCtoolseclipseeclipsec consoleosgi ss Framework is launchedid State Bundle INSTALLED osgi start Cannot find bundle osgi ss Framework is launchedid State Bundle INSTALLED osgi +I just checked in the for and there is no longer an openTemplate attribute defined for the Autocompleter tagIf you are still having an issue with this using Struts please this ticket with specific details for the issue. +The bug is not there in +Also to simplify things I have moved all the required jar files from release and the one built by me for client into one directory and the classpath now looks as followsexport CLASSPATHCpclientsmaintoolsjavacpclientsmainibmlibCLASSPATHI am running the tests right now with the text interface to see if the problem will reproduce +Just looking i see another bug in + +Hm I might have added that in a different ticket +Xavier any ideas +Reviewed and edited both documents to bring everything up to the current state of the code +Hi had download the SQL Client and configured the hsqldb in memory use the default username sa and password without any other configuration and the client could think if you do not change the default password of the user sa you may not get the problem is I created one hsqldb database for each of my project if we do not allow password using empty string I have to configure the database properties file by hand at the first want to use the default configuration to skip this step make my work you want to have a look at my test project I just want to let it work without any configuration for the +This test has consistently failed in the tinderbox environment when run in the lang suite under Java +The DCL extension impliesThat filename is not valid on the VMS file system +But this will work only for as long as mathjax service is up? if it is down or moved it will get unreadable +WDYT +Looks good +this is different to would be possible under the current profiles this functionality is added +As of revision has a API for attachments that doesnt require base encoding of data +Thanks guys +patch looks good +Ill try and do some more debugging when I get a moment +Cheolsoo had a question as to whether we should commit this to trunk and branch after that +Hibernate uses in rev as part of +Does this apply to released versions too or just a +The build failure is caused by including of JDK +Functionality is working fine without changes of file +jar files solution doesnt solve the original issue like Tony stated +Hi Todd can you pl take a peek at the latest patch that I uploaded +I will put up a patch for review +Is there any way I can reproduce this? There are now builds of the new invokedynamic logic in and we could help the Hotspot team by testing your stuff out and reporting performance numbers +We should just regognize that the property value is a url and handle it as a url if its the case +Do you mean what is loggedyes +Committed in rev +Please cross check and thanks for the patch +upgrading severity on this since its a rather big usability issueLets reproduce against plain egit without our codeusage and open bug at bugzilla ASAP and then work on patch if needed +I guess the only reason would be that the actually artifact of interest after a Job has been run is the Aggregator value rather than the state of the Vertices +This issue is mitigated by updates to how subimages are handled +Fixed in trunk +I was trying to commonize code between web ui and web services +Updated affected versions of this issue +There is also an intermittent exception on J although in another place +In case of the the name will be reported +They could then investigate further and possible remove the failing disk +I think that all of these logs are already at the trace level on the teiid side but that our logj integration does not distinguish between debug and trace +That has to build the full stack trace of the current invocation it should be slow +r on trunk +rememberAction shouldnt be called with a null action parameter +Theres no reason why you cant have Java source that takes as arguments the request response and and does the work inside its own code +I have updated this patch to address style issues highlighted by you +Is this expected +If thats the log conf file that is used without f and not with f yes +Let me know if you determine that there is something missing beyond the cases I outlined above +The REVOKE topic has a lot of verbiage about limitations and system tables I wonder if anything is needed for sequence generators +Remove HADOOPCONFDIR from srccontribchukwabin and add it to good and cleaner +is a very small test case +Was wondering whether we should have a single that does both the protocol and the admin parts +Ive committed this. +I guess you mean the Activiti Designer? I move the issue to this JAR export didnt work because the folder paths were not created correctly on windows +You changed the order of the imports is this on purpose or some automatic IDEA thingMine import import import Repo import import import +It aborts the whole datanodeI think is correct hereScenario user configures DN to point to two different which are on different clusters and starts upPossible results DN connects to cluster and joins the cluster +Like I said earlier this is acceptable to me +Merged into piper commit fadabaefaba Thanks Daniel for reviewing! +Committed to branch and trunk +The following files are missing headers package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package +Give me a configuration that fails to work with the +The default behavior will have to remain consistent throughout the series so the default wont be changing either +Please review + +I think that should do the trick thanks +Mohit Please file a new JIRA ticket against this issue +Can you post a simple executable code that stress the problemIt will help me a lot in fixing the bug +I guess youre after a setting to put yes? When you raise an issue please be clearwhether you want to have different settings for and +This seems to be by design done as part of +Thanks Pamela +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. + pending failed test is unrelated +Thanks SebastianYes is totally unrelated and already assigned to the culprit If you get a chance please submit a patch for the case you spotted +Closing as Later. +Hibernate Tools Reference Guide was checked and updated. +Attachment has been added with description RTF output before has been added with description RTF output after +batch transition resolvedfixed bugs to closedfixed +Thanks for the for Owen Ill the change and run some test +It is not clear exactly what this test app is supposed to demonstrate +Jason +fix wtest +Longer term c would be great or if IW has an ES then itd send multiple flush jobs to the ESLost in abbreviations Can you remind me what ES isbq +This is probably the same as Vinods last patch but I upmerged to latest and retested to be sure +CarlosCould you make the temporary fix? My client currently has me buried in tasks and I just had my first baby girl so Im highly sleep deprived at the moment +As a side effect this bug is resolved because apt is no longer called programmatucally +jboss which would be the right +seems like the buildtools directory is just redundant of the Maven directory +Approved for +The s arent used for site generation yet +Is this intended for or? Im getting lots of conflicts on both +This will be a sign that we need to stop running the trunk tests on those platformsTouches the following filesM javaengineorgapachederbyvtiSimplified the build scriptM javaengineorgapachederbyvtiParameterized a usageD javaengineorgapachederbyvtiM javaengineorgapachederbyvtiMerged into +Resolving as fixed +Fixed in revision to the branch in revision . +Once i get something committed Ill start to think about a framework release +bq +This is way too old and as Ken pointed out this should be dealt with upstream in Tika. +The attached file is a propsed patch for this applied +It is in the works of being tested +I missed them +I do not want to see that again +Now that work is well underway with AS all previous community releases are +Will keep this Jira until it is fixed +For me this looked crazy because refactoring like this should not change perfThis explains the change back after I thought oops responsible for the speedup. +I used script on from windows commandline +I can add in a separate page for an attempts counters but I think that is beyond the scope of this JIRA and I would file a separate JIRA for that if you like +Still Ill doublecheck in the next few days and will close the issue after good now! have other issue but before reporting them I need to first check if theyre realThank you for quick fix +done in to M +Closing all resolved tickets from or older +Added to by SVN revision by SVN revision +I volunteer to do right away if not throwing the ISE is the right thing to do +Personally my clusters always have a synced etchosts so a lookup isnt a problem and would be terribly convenient compared to using a script or memorizing the +I have few queries +verified by Vladimir. +I use normally only my text editor for developing but for such a thing Eclipse is goodRight mouse on srcjava folder Source Cleanup Use Custom Profile remove everything except add missing Override +The patch is not though because I commented out two lines calling fix +Worked on my continuous integration setupthen tested with setting CONTINUOUSCONFDIR and removing the host list and files from testsystemcontinuous +Created a graphviz diagram of the dependencies and thought I would share it with the rest of you +The current version of is poorlyfactored and would require the Rule implementations to havemore knowledge of the internals of how does introspection than isreally appropriateTherefore Im marking this as RESOLVEDLATER so that its not forgotten + a patch that changes format of default describe table command as hive just noticed that the required changes to are missing in nd patch +Closing issues before next release. +A patch to come for Annotations to use this new management itself + the latest release is being used for reliable messaging and is obsolete +Any more messages that exceed the fullSize will be stored to DB and later will be loaded as the messages are being delivered off +Looks to me a clustering configuration issue please use the user or dev list to get the issue resolved +By keeping the values around my code is confused and thinks it needs to a message to the user +Bulk close of old resolved issues. +Id like to attach an example but Im not able to +The patch will fix the broken build processAlso an exclude list is specified to avoid servletapi and other unwanted files to be packaged in the war s +Correct jar and archive files being produced +I think youre probably right that the bug is in Xerces but I logged itagainst Xalan as this problem does NOT seem to occur when you do astandard parse using the SAX or DOM Xerces parsers directlyIt only seems to occur when you set the INCREMENTALFEATURE when usingXalan I didnt download the source release of Xalan so I cant seewhat unusual Xerces parameters or options Xalan is setting that causesthis error to appear +I fixed it post +Delete own attachments now enabled forThere has been no change to the INFRA permissions. +to can just use LUCENE to match his old behaviour +This project will likely see a lot more than just phone and tablet devices in the near future +Resolving this issue. +So this issue is not specific to Web Services but due to WST bugs? Lets close it? +This behaviour is really annoying +Super! It may well be that this is the fix I mentioned coming in later due to CVS sync + +Andrew opinions +This cluster has disappeared for now deferring to we get a hold of a cluster big reopen once we have a cluster +Would you please help to try it +I did not see the shell start and then immediately fixed this problem and committed it +This patch applies to branch +What about just storing it as a string +But if you are planning to add some feature for this then everything is OK and probably this is not a fault. +But should be straight forward to update +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I just want to make sure that it is case with this patch +Hi MaryI dont think this is a issue since the forums feature of the forge has been totally removed for now +Because of this I think the README should cover both the normal case of a source release being used directly and the alternative possibility of a direct svn checkoutThe README doesnt need to describe how to download and use the binary distribution as this will be removed by the fix forThe README has been updated under r to describe how to set up JUNITJARI agree that details of the ant script for running the domain manager need to be added to the README +Since the changes are minor I will commit it without waiting for Jenkins again +Should we wait for to be released or will Spring work with the Apache NMS? Is it just a matter of dropping in the lib to replace the lib that ships with If not when then idea idea when will be likely to ship +Patch for the branch of SMX +Oleg said that this bug was fixed in +I have a patch for this one but it requires more testing because it supposed to do some critical changes +Commit changes to branch at revision branch at revision and trunk at revision it +Heres what users will see when the above are appliedJBDS! Tools +Attachment has been added with description updated test showing additional has been added with description updated test caseCorrected the last case to reflect the situation described in the comment +I guess I should add and I hinted at this in the defect submission Even if you do make that package optional the use of will be dependent upon a new release of Jexl that includes the headers to export packagesIf the Jexl project does not plan to update its distribution perhaps someone might the distribution and make it available via the Bundle RepositoryAt the very least though making it optional will allow anyone not using Jexl to leverage the new version of in an environment +how Am I supposed to debug the issue if u do not even generate a core dump? or a log message +Bulk close resolved issues. +Thanks for the review +Is that possible? That way wed stick with the stable ZK branch for people who arent using security +Jeremie my previous hot fix did not help so I immersed into this problemAnd Ive managed to find root of the problem method find and projectThis map persists until Eclipse restart and configuration wouldnt be removed if project removed from workspace and from disk +Thanks for catching this Nicholas and sorry for not checking this beforeThe code changes look good to me +Closing this issue. +Replace function just return a new string any attributions try to innerHTMLthrow an exception still studding an solution +Added a section on exception handling in flat file reading. + to trunk. +Thank you for your answer +Logged In YES useridDo you have the same behaviour when using the shutdown script with the same arguments from a command line +Down side is will only run in the the AS now we can support for Tomcat later +Someone should create working example config which we can add to distgitItll be used both for RHEL and multiplatform builds. +The unit test depends on the +Bumping this up to major because of this message from the mailing listI am finding that groovy is duplicated whenever I enter anything into the groovy shell console view +now generates file to facilitate chart png browsing from CI serverThis patch is intended to be applied on top of the one that Keith already attached to this JIRA +Georgeregression test H failed when I tested your patch on EMT I was wrongI rechecked compressed mode on OPT assert occurs incompressType passed type is +One small point in rrefattribderegister Both the second and third bullets are instances of autoloading +I found a bug in the code so Ill be releasing another patch for the sameIll keep this patch in the JIRA until I replace it with a new one so everyone can review it +I tested it on a distributed cluster and it worked fine +I will update the javadoc +We will do so once the patch of OCM for Jackrabbit is accepted and I have commit rights on SVNAlso if possible close this issue. +Very good find! I wonder whats that gibberish at the end of the file +This patch conflicted with a checkin that was made by namit so that reference was added into and my patch which was generated from a previous version of the tree did not have those changes +Especially so late last nightIm not sure if turning on schema validation by default would be a good idea its often best to have a product performing as well as possible out of the think were agreed that this should be configured in the xbean though. +for the new integration test +Theres no need for the G build to grab and build all of +I was envisaging bringing the rules config file in via a version artifact as a plugin dependency +Moved to wishlist +Can we close this +As I said my intention is that receiveResponse never times out in normal state no matter how slow the other side is +common was archived in war and in ear +But have tried withcacheMana +So the bugfix has been verified as well +Cleanup marking bug as fixed so it disappears from the weekly +Also related to updated version of the archetype also includes examples of filename substitution for two +If you need this to work with you can turn on using the XO flag to but you do so at your own peril performance will be significantly worse +Updated createAttribute methods in +yeah! This is awesome +For these I would try to generalize with JTS +Fixed +You are the man +Anil committed it but didnt close. +What else is keeping government employees and the agencies they work for from signing and joining in the fun? to everyone for making this happen! I can hardly disagree with youThe fact that this particular body acquiesced to our CLAA shouldnt be treated as evidence that it is worded properlyBut since the NSA accepts this agreement I urge you to your thoughts in their entirety and take this to for further deliberation +If their JMX is messed up then the deployment location wont be added properly and all sorts of other stuff wont happen +Queue Apug has nagitive Number Of Pending you please provide more context +partitions are required for vnodes but not vice versa +Hmm there do seem to be a few odd interactions here + may be a impression of a unexplained bug which relates to the communication between and +Thanks Eric +Should be fixed after regenerating from the new Odf Spec running codegenI checked the old Odf Spec and discovered several errors regarding default values +We are storing the timestamp of when the appMaster registered but not using it anywhere yet +Also do the same for the project. +Therefore I think we might always get a failed merge if we try to apply our merging strategy +There is certainly a problem with Groovy classes and +That also explains why I the buffer at the beginning because once the bytes were read into my debugging buffer they were not able to be out of the buffer by thrift as there is no method for that type of streamThis is more than needed to be said because jacques fixed it but we were looking at the code together and I couldnt remember for the life of me why it was thereI will create a modified patch with only the related changes +I experience exactly the same problem using Axis and IBM jdk on Windows XP +patch fixes the failure issue +I guess that there may be more potential deadlock issues +Please find attached a log output of a test case to reproduce the issue +I committed this +Rejected authentication would look like a failed connection however the Endpoint would never know about failed incoming authentication because such connection attempts would never make it in to the Endpoint at all +FadyCan you provide some sample code showing what you are doing? Perhaps you can upload a small webapp showing the problem +thanks all for the code reviews. +Before committing I will try to replicate the problem you are seeing by changing the integration test do index a document +Which version of JBAS are you seeing this in? Can you try a dev build of and see if the problem occurs there +bq +after resolving the licensing issues we can reapply this to whoever commits thisThe license file would go into tomahawksandboxcoresrcmainresourceslicenses and would need to be menitioned in the tomahawksandboxcoresrcmainresourceslicenses file +By the way this issue seems to be +Many thanks sending the vc project +Mikhail I think we can close this JIRA now +Leo patch applied at revision r thanks a lot please verify +For row lock based database locker I use the same configuration file for all brokersThis makes deployment job simple really +AnilSo you mean its mostly intended for component but there are any call for that and in this case why your patch target workeffort compoent? Do you really want that patch commited in its current form +What do you thinkParsing the rules once per test run is not a problem since there are messages to process +Commit a fix for branch at fix to branch at Rev trunk at Rev +I tend to believe Maven found a way to workaround the problem by using a jar with a manifest containing a entry +The release deadline is approaching so we need to differentiate between critical issues and minor issues among the open ones +bq +But running repair is easier and cheaper than stores N copies of same data +indeed doesnt work when using an annotated controller interface in combination with AOP proxying +We dont need to mention Jetty +HiI have tried configuring the mosaic through the last geoserver trunk on both Windows and Linux and it worked fineI was able to add the store and view it onCan you try using the last nightly build to see if the error still existIn affirmative case can you please provide me the exact steps sequence to perform which raises the issuePlease let me knowDaniele +I would suggest to commit the current patch and open a jira forWe need a major refactoring of the use of in the code +I also think this is an improvement not a bug +Yes this appears to be an error +Regression test works for me but I excluded it for x architectures +An overlapping of solrwriterDoes it make sence to spend more time looking into this if will be removed anyway in +We plan to support nested foreach in future release to solve the problem +Committed and trunk +Heres a work in progress that attempts to derive a reasonable size of the LAB from the block sizeIts not configurable yet it may crash and burn your machineAlso still experimenting with who should own the LAB +Adding bfederle to the bug in case he can comment +Hopefully this patch will apply cleanly +Any thought to doing one last Ruby release and targeting the next release to be Ruby minimum +If history is any guidance theres high chance of wanting something similar eventually even if it wasnt immediately obviousBut I think we can split this in two parts hook to call should probably differ anyway and it might be that one rather wants hooks + +If you want more assurance still since Kafka is easy to consume from its easy enough to build a consumer that backs up the data elsewhere +implemented suggestions from Jacobs good +Sorry for the issue +When waiting for the final application status to be failed you can use a smaller sleep inside a loop +Heres the right patch +Trunk is almost identical to branch at this point +Once thats done we can add the code to Tika to use the library +We really need the svn patch command so that it would have locally added that file on me applying the patch +HiActually you should file a bug against Weld because this is exactly what the spec requires at the moment With we I was talking about hte Expert GroupBut yes we will fix this problem nontheless +Worked for meipv has also been know to radically inflate the number of file descriptors we have going on at any one time +Good on you Dave +This is for Accumulo +dupe +Carl is it already committed +btm as the extension +bq +Changed the checks to be rather than looking at the whole job +So I am including Cordova along with jQuery MobileHowever in my real project I have omitted file completely and only left jQuery Mobile and it seemed to show the compass value on its ownSo it would seem jQuery Mobile code is the winner here +Fixed +It also removes the limitation of needing to use a special pattern for thrown POJO exceptions +Can anyone point me to documentation on how I go about applying the maven patches? I cant seem to find anything in the maven site +One collection worked as should while the other two never recovered +We had using this patch in our production environment for about month any plan to merge into the next release +Optimization for the simple LOADSTREAM and STREAMLOAD cases where the serializationdeserialization isnt necessary including a new split by qualifier to LOAD to aid the optimization only for cases where maps are consuming complete files +this patch fixes the issue +Thanks for making it you still doing a force param +Let me know after tonights nightly build +You should probably just modify the FTL to say Not available and disable that choice +Hi Daniel +Ill do it tomorrow with more complete description of the system schema java versions etc +I need to figure out for which scenarios this process can be skipped +Wrong +I am curious if this patch makes performance a bit faster for the affected use case +As far as I can tell now this should definitely make into the release +The change being reverted here seems to be have been accidently introduced inIt affects only debug option +Attached a trunk patch that skips STAGEGETFIELDS if the fl param is just id or idscore +Please review and resolve if you are satisfied +Java is now deprecated from most modern oss and throws a running mvn clean install on a new system will throwCaused by Compilation failureFailure executing javac but could not parse the errorjavac source release requires target release As java is usually not installed and has a of updating this to resolves the problem since the source code is already based makes sense to update the target as well +Verified on JBT. +What about other RHELi boxes +Thanks Jonathan! +This is d say this is not specifically a duplicate of the +However now we have replication steps +It looks like setting skiptrue before running could be a way to avoid this issue when component is new. +Sorry I will never have time to get to this and I am sure kris has already looked at it +Added which version is affected and marked this as Path is a replacement for +There is a similar call that simply reads the local value of the ledger handle once the ledger is open +I followed the I Add JSF Capability I end up with exactly the same corrupt tried a couple of times to make sure I wasnt making a mistake but the end result is the same every time +We modified the derby code to use the regular which just uses inputoutput streams +No I dont see any delays here +Unload this zip into arbitrary directory review and edit and then run with ant f I prefer this to a shell script tooJohn +Its not a but just because of the document + the patch looks good +Committed revision . +filename will not be changed for this release as the community release is already the comments in the file has been updated and this is documented in the Server Configuration Guide. +BJIs it ready for commiting +Nice work. +There are things that map to a core channelDurable subscription are managed in the channel mapping dont think it would make sense to have the durable subscription channel mapping split out into a different it may be an idea to split out other diurable subsciption attributes into a different the mapping between destination name and id can be done in a different table +Just committed the missing file and added it to the build scripts +Ok thanks +is it really ok to remove those configure calls +Im a Hama newbieI will add the setup and cleanup modifications to and attach a patch +Ah okay I probably should have asked what you meant by does not workWhether it is intentional or a side effect of something else is another question +Use generatorConfig to set configuration directories +I adapted the patch to the apache versions of the code line +Closing all resolved tickets from or older +For it can still be used but it is not recommended +I missed it because it wasnt in alphabetical order +I might have to make my app work in Android as well at a later stage so keep up the good work + +patch committed +that the shape fileDBF have the wrong binary codes for those we are working on fixing the shape fileDBF before testing this further +Clicking on the link could produce a language selection popup window similar to the current lookup windows +The test just initializes a parent context then tries to build a child context. +except maybe if you ignored a thrown exception or something +Yes I tested against +If they havent been it looks like it would be a pretty simple migration to the other method + +I think the real feature would be to be able to completely disable the journal +Remove as an affected version. +It still found weird though that both systems would be corrupted at the same time +so the place to look is +similarly with hdfs +The new test with abort will come in when is merged. +We already have the plumbing of supporting multiple identical entries to support isolation +I cant wait to see what it does for our client code coverage numbersIll commit this patch +Will evaluate upgrade for ER build. +patch has fallen out of date can you please regenerate it +Happy to do and trunk patches if people are ok with thisSeems like is a bit already but didnt see anywhere else that was better for this test +Which eclipse toolbar do you mean? The Run As +Alright good +Version Released +The idea being that reusable expressions are created in advance and then they are picked from a list by regular usersCalendar event screens that support temporal expressions will have UI elements to createedit a temporal expression specific to an event +Currently only one compaction per store is allowed +Makes sense +Otherwise wed have to do fine grained on all and granting a user access to a table would require granting them access to the files involved in that table onlyThats one option +This doesnt fit with osgi jndi which aggressively caches services from service factories +What examples are you looking for +The state is still stopping after I click Disable Query +Is there a simple guide to achieving this some place? +change component from scripts to close for Solr +Looks good to me. +fixed in r +I added the methods to and as well which allows you to set this on component level as well and also see the value in JMX +Zheng did you mean the above comment for some other jira? +This does not use in patch. +I merged the patch to trunk and with no problems but has a few merge conflicts +Thanks +This also allowed me to remove extendingHowever this has API issues +Cant wait that long sorry +name cloud print shows +No capability to build on other JDK implementations. +Our current aries blueprint version is should upgrade blueprint version +Barry and I discussed a different approach for consideration and inclusion in teiid +Ive fixed this as well +Scheduling to fix in widget list is displayed an instance for the test user is created +But this makes me feel very bad if maven is not capable to take the same decision at every build +I tested the Phoenix patch under and it works. +lgtm is in +Until this is addressed in the spec we cannot say what a proper behavior is for this +If we want to implement partitioning the processing logic should be migrated into core modulePersonally I think we can hold the partitioning for later on +Being worked on by Rupert Smith +Fixed on Camel trunk revision +On the viewId does not change so unaccessed conversations are never discarded due to validation failure. +Screenshot showing resources available to JEE app in websphere +looks good +Im running the with set to true +Can we not have platform specific settings in libtoolautoconf? I remember doing something along those lines sometime back but not sure +This is the documentation task associated with which Kurt is the info to the admin guide also +Can you please start a parallel run locally and post results +Between two deletes relinquish lock to allow namenode operations for other clients +Updated patch for trunk +patch against sea +committed. +Instead of pasting the code here I attached a patch +Daniel please verify +Map Bean seems useful so putting in however might get punted to a later version if we decide we dont want big new features in there. +Ill attach diff of my working copy in a moment maybe someone else can pick this up and see what am I doing wrong +Patched and jars with the patch built in +Pushing off to release +Verified +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Then what do you think about adding an option that disables autostart of SSL handshake? Would it resolve your issue? +Please try latest nightly +Awaiting documentation +Fixed affected +Closing issues in +set to fixed in +However at the Reducer side we do not know the size of the map outputs beforehand and the reducer needs to request specific map ids It cannot just specify a single size as the TT will not know which maps have already been fetched by a given reducer and which have not +If you think this does seem right then that should clear this for theSteev Coco +This information is enough to reproduce the issue +I just committed this +Weve started already using maven trunk as the example to work with to try and reproduce my problems +If you try the attached WAR you may be able to discover what is taking secondsAnyway if is disabled by default in that works for meThanksRichard +Patch applied with some changes +Sounds like the ulong typedef isnt present on Mac +Im attaching a test program and that demonstrate the problem +So it has nothing to do with Atlassian connectors +All source files should have svn property set to native so that the EOL is adapted for all platformsIve fixed moduleslunisrctestapicommonorgapacheharmonylunitestsjavautil moduleslunisrctestapicommonorgapacheharmonylunitestsjavautil moduleslunisrctestapicommonorgapacheharmonylunitestsjavautil at revision r +If we are going to remove it as a breaking change this should be +Scott is there still more to do here? +A comment is added to ease dependencies tracking +Ive just committed this +The tests itself run now but some are failing on my laptop but I think it is not related to the file +Is this still an issue under? Much of the code has been rewritten for +It might be better to fix the profiles issue first and then fix this one since well have to do it for every profile anyway +bq +Can not am glad it works for you +We could do better by letting the client know that the region is now fully available somewhere else and that there is no need to wait before retrying +just prefix all other props with +I dont understand the question +So no matter what value you set for indexinterval in system tables always use the default of indexinterval +Thanks Peace! committed +Thanks Claus just to know is there a way for me to be able to find that out? Or is it just a question of it going into the next build usuallyCheers. +Patch applied with rv thanks ivanI still cannot confirm that the test is failing but wil check your mail on the list on this +Hope that makes sense +If theres something more going on that Im missing please the bug. +What happened with these ones Mike +Fabrice could you make the Delete all filters link a buttonCould you rename Save filters to Save exclusion filters and Delete all filters to Delete all exclusion filters +This affects any app using the Springcache plugin as well as the ehcache dependency is packaged as pom +Alright closing +Resolving as fixed please close if satisfied +Thanks. +I will be checking my emailintermittently +support if now +bq +I just commited the majority of this patch +Can you send a diff file? it will be more easy to apply. +However I am open to contributions of a better more integrated support for SOCKSI committed the proposed fix to SVN trunkOleg +studioeclipseplugins +Go ahead +I might want to recover on RM restart but not on AM failures or viceversa In case of AM failure user might want to check for user errors and hence not recover +will be in ER +I believe we are safe +My suggestion would be to make the focus of this issue the part of expansion and contraction and to define hooks that specific services can use to handle adding and removing nodes from a running cluster which as Tom said can be tricky in a series of cases +Nathan the synchronization is not needed here +rescue Object or rescue Exception should pick them up +Not sure if that can be helped but if its an attachment on the wiki we could at least capture noteschanges there +I just want to comment and say Im not sure if GA will please reporter on this issue +It looks like Charles has some idea now +Okay so would this change actually make anything better or is timezone handling with Postgre not working even with that fix +having combined jars plus dependencies in the distribution tar ball +Seems ok +Hmmm +You claimed that this patch is to provide soft mount feature to dfs +If we want to make further test improvements I think those should go in a separate JIRA +Submitted this patch to trunk +patch looks good +Also check the doc +Yeah you need a special tool either a Java IDE or a build tool like Ant or Maven to add these headers +Changed in jacorb releases in the comment and the commented out properties as follows connection idle timeout +Did you eventually replace it with a newer versionJuergen +New jacorb has been uploaded to repository +Failover works as long as supressExceptions is set to false +close of resolved issues in. +Provided more info on for not commiting on trunk that would be troublesome +bulk close of obsolete debugger issues +They look good to me +BTW do you think we should branch now and restore when we separate the descriptors forI guess branching would cause additional work of merging changes from Branch to Trunk +We need a maven section in the book +Currently in branch soon to be also in trunkSee recent comments in +Closing old resolved issues +but thats what you want rightThe thing is that we want checksums for everything other than reading hfile blocks +Additionally if our own JS submit function is used we handle exceptions during submit by removing the transitioning pane and displaying an alert +I assume JRE upgrade fixed the problem and there is nothing that needs fixing in itself +Emanuel will have a look at your patches and get back to you re getting a pull request +testviewfstrash failure is unrelated to this change +So allowing the server to change magic project is fine because the server stores the magic project in the server settings +This is partially core functionality is there the serialization needs further for the non nested jar handlers +Got a on RB +I tested a nightly build and can confirm the patch is working +Still in more investigation +Sorry for thatTed yes I should pay more attention to the precommit report. +Based on the fix I proposed +Looks like to me +I agree the pm should be closed at the end of the test +Note that this function is not a part of the SQL standard +Making this patch available +Thank you Bruno! +I dont think scheduling of can solve this if a user or app is pouring enough ops in to affect QOS of others then its going to cause backpressure which is of necessity notI think there would need to be some kind of at the coordinator node instead since thats the only place you can backpressure instead of entire peer nodes +Unchecking Patch Available after Mamtas comments +The exception is thrown in and this patch doesnt touch +Please checkthat you dont have any old versions of FOP in the classpath +For the attached the generated stubs omit the properties for the attributes in the attribute group +So am closing this and opening more focused JIRA where appropriate +Does this generated POM file need to support? Im not sure about the branch but scope is supported at least in and +No worries on the resizing of the logo well attack all of that stuff when we get a chance to reskin the ecommerce demo and backend etcI do like it at the larger size just for readability but having a more light weight header never really hurt anyone +The referrals are now handled in the Core API +If so a warning is currently suggest to introduce a property create which defaults to false +Commit on will be on hold for a bit +It is possible to make a https connection without authentication +A minimum implementation should also probably provide a simple timeout that will respond with nothing if the timeout expiresI think the simple implementation by itself a huge win and you might consider is that good enoughFor me it is I would prefer to get the simple thing in the short term and wait for the harder thing in the +Ill commit after hearing from rough day +No Apache headers yetShould we fork this into a submodule? I think this can be majorly improved over time and would be a good start in making Hama more scalable +Entity Local and Action Java has not been an issue since before release. +Committed to trunk thanks Prasad +Closing all resolved tickets from or older +Another simpler and probably more practical approach would be to allow specifying jvm options for a specific input split +And when you account for that it seems that locking would not be significantly simpler to implement than versioning +Original patch was fine + +CSS Preview view Properties View has been added +The whole point here coming is how to query from client the result of an old operation happened in serverThat is why I advocate above whether we should make nonces visible to the client and enable accepting nonces +Please if this also happens in compatibility mode +is also committed +Bulk close for +hi this is in CHANGES but is it resolved +m not certaint he bug I fixed was the one you are mentioning but it seems very likely that it is +When the page is created or editted it will return a blank page instead of the portal you are debugging the code and the other thread is able to catch up it will sometimes render the page properly +The exported target itself does not contain +Ok. +But I believe its solid +Roman can you comment on what testing you did of this patch? pending Jenkins and an answer to the above question +Indeed I had CLASSPATH set +Thanks for cleaning this up! +All of those changes are in the patch +Unit tests pass and caching seems to workIll make some small javadoc and cosmetic fixes upload the prettified patch and commit on FridayThis will give more days to others to review your changes and raise any issues they may see +Remaining issue re central featureplugin will be addressed in. +We need to update to version of proton in order to fix this +I restarted the cluster +Wdyt +Ubuntu makes me dizzy after upgrading to +Definitely complete had been waiting on my access to resolve and then it fell by the wayside once I got that access +Attached is which incorporatesthe last few suggestions from reviews as well as someadditional commentsderbyall and were cleaning so Im proceedingto commit this patch +I cant see how it prevents attacks from trusted nodesIf the node is in fact trusted then youre screwed +The unit tests of that release would be retrieved from subversion compiled and run against the current core code +Shouldbe OK now +The issue wasnt reproduced on EUT tests r +Needs review +verified thanks +Need some logging somewhere so that excluded tables can be discovered and reconciled +This is going to be a blocker right? Maybe we should reorganize removal of under an umbrella and make it a blocker +I added the +Closing. +I ran workload e which has some scanners but youre probably correct about the length +SVN patch of my implementation +v is what I committedHopefully this will decrease the failure rate of these tests +But to big global locks that prevent that kind of stuff +PB makes it easy to extend the headers +Sounds good I noticed there isnt a branch? Is this something in the works or is it a bit further off? Perhaps there is somewhere where I could read about it +Eli filed to address the test failure and has some good analysis for what the root cause may be +Not that its likely but the whole point is to avoid problems or catch them earlyPlease hold off for a few days at least to allow others to commentThis might actually have been better to start as a mailing list thread but Im guessing you werent expecting real objections +For example in that case I can use components from warlib in EJB module and I get if libs are not in lib folder +I am attaching the patch here please review and let me know your feedbackWe will update these classes when required +All tests are passing now +Also please correct me if I am wrong but it doesnt look like the index file associated with the is opened with Direct IO and the new being written +Provisional fix is really not good enough +Im now continue to develop the rest +Indexes attachments from various types like ppt word txt and anything that Tika supportsTODO recurse into folders performance tuning support filter criteria for folders supprt more than one mail box support pop USAGEFor each mail it creates a document with the following attributes Created fields single valued subject from sentdate sentdatedisplay XMailer multi valued allto flags content Attachement flag values answered deleted draft flagged recent seenCOMPILEDependencies API jarActivation jarTika and its dependent jarsHow should we go about adding these dependencies +bulk defer of issues to +ask rpenate what the correct favicon at eclipse was copying the resource over but maven wasnt +its the name of the boolean that will be set to hasNext during the loop +Same problem with version issue has been fixed by in CARGO version. +Our algorithm for checking if a property has been touched had a bug that is covered by the description +Certification for DB is deferred +HiDid you have any idea of fixed versionThx +Kelvin patch applied at revision r thanks a lot please verify +The field type that you provided doesnt use the the though Have you pasted the correct one? Also have you seen the other configuration attributes like generate and split? This may be a configuration problem and not a bug probably you would get more help on the users list +The downside to is you cannot interrogate the exception for the path that had the problem +This problem seems to be caused by implementing rather than +Johannes this code is being almost completely rewritten in the types branch +Have you added your component as an extension to the build? If that doesnt work try adding it as a dependency in the plugin configuration +The last patch from Ravi looks good +Sample implementation of batchUpdate method +It removes completely replacing it with two factory methods in +Although this change started off simple and elegant it quickly rippled to many more changes through out the code that really didnt add any value +It made meinvestigate the problem more closely +Hithis patch has the fix you suggested +I think should get fixed somehow though it can be quiie annoying to not be able to clear the attribute values without having to enter dummy whitespace +the additions we are proposing for integration with trunk +We may not be able to access to +I am able to reach c +Attached patch removes the folder and its files +Passing a batch writer to the batch deleter makes assumptions about its implementation +So Im glad to see you even updated the documentation good jobFor linking issues this jira instance is the first one I use what I know is that its very configurable so its certainly possible +Im not sure it is a drop in replacement need to try it +Thanks for your help Adrien. +running tests +StackCan you comment on the latest patch +I will look into it +Change was merged onto the branch after approval on the dev list by Justin. +Upload a patch create a separate removeAppdispatcher in for handing app removal requests +Patch applied thank you. +Committed +Please upload the updated patch with the seed data added +I forgot to say that Im using Muse. +Im guessing that should happen if the clicked component is focusable. +Yeah good idea +I use with Tomcat andIn Code Editor highlight an red error without run application andseveral Unfinished Tag +This causes the my own I solved the problem by providing my own class loader factory that caches process class loaders for each process definition id in this way I am sure that only one instance of a class object exists for classes deployed in the PAR +The original fix for incorporated into EAP through removed the caching behaviour in +I will look into this over the weekend +Please let me know ASAP if you see it again +Draft only tests and support to be added +I forgot to include +This is a duplicate +RS completes the merge so that A and B are offline and C is online merging node C is in merged state now +The additional methods do not affect jdk build +Is the admin webapp tested for embedded broker in tomcatThanksBhupi +Ill be adding unit tests to this but in the meantime could some committer please bless it +I guess having on top of the +Since the request in this issue is to mark unreleased versions as released which would make the bug database contain incorrect information Im closing the issue as Wont Fix. +Is it enoughHi Tony that should be enough +Agaon which version are you using + +This is a very good idea but Im not sure this must be done by the release plugin +Verified on +all BrianThanks a lot how do we get existing artifacts to the from the repo? Or are they already moved? If so where? existing artifacts to the releases repo +for a JSON LoaderStorer that are part of +really we should be able to backport this to too we just have to make decisions about whether we should break or add lots of deprecated ctors +Nates closing all issues that have been in resolved state for months or more without any feedback or update +Marking it as patch available to check test results +Hi guysjust checking if anyone of is working on this? Im just scared that the more we wait the more the source will differ from the patch +The default status code has changed from to now so it should work you please try a recent snapshot and see if it works for you? are being added again in +I think its readyNote that still cannot consume a graph so eg you cannot apply it after WDF or after Kuromoji +I attribute this oversight to running clean instead of +The fundamental solution to the problem is to implement and deprecate this attribute. +Config files in Hadoop dont seem to include a header so you can probably skip the one for +Brandon For the record we had HH turned off so I dont know +closed +This patch removes the documentation of the property +This is not a bug +Bulk move would be extremely helpful to me +I have the same problem but only with JDK +Why make assumption about how one wants to handle titles in Solr. +It now functions at normal speed. +sorry +should we have this in branch +really added in r +quickstart maven app version to change into the unit test is not that critical. +can I ask why this ticket is in a state of ? What do you need from me +only optimised queries handle search by document contains results of tests for getNodes and getProperties by pattern +Hi GiriWith this change only bundle core hadoop classes +That was what I meant. +Link the required changes together +The reporter has not been able to repeat these problems when the client code was changed to not refresh the session from the thread on which the observer is called +That is ineffective because the object under construction may belong to a subclass with its own Class objectFixing that bug resolved the original QA test failures but was not relevant to the failures +Check for null as sstablecompression is removed users can use sstablecompression set to to disable compression +thanks +IO needs a lot of performance work. +Right so maybe we should create a new issue on edit modes and just do the solution for this issue. +Not that it was very difficult to produce +coretargettestmainruntest +dont cause this issue though +at +Closing resolved issues for already released versions +Tagged and staged +Closing all resolved tickets from or older +It is impossible to run any test of CTF test suite as long as JEE category isavailable in +I see this is noted as fixed on any way it can be applied to? It needs to be available for Java as well as Java +If we do that then the need to have majorminor numbers can be considered an optimization isnt it +there is a lot of code which is there just becaus eof this alias works now even after the alias command +Committed at subversion revision +With this rule itll be easy to create a new Sonar Plugin to count the number of occurrences and follow this number over time +We can deal with that in a separate JIRA ifwhen the time comes +lgtmThe tests failed because I had missed committing some files during +Please reopen if Im wrong! +This should be resolved for and cases for RC +some reason the nd patch failed to be applied +with latest Eclipse on Unbuntu +Lets put this on the list for this is documented behavior its not really appropriate to call it a bug +Add layer +Stefan Bodewing +I can try to configure it not too +The bug reporter? Reviewers? thanksIm verifying this issue and the document looks fine to me except a broken link to classlib pageThank youPavel +Sorry +Now merged to the release branch. +SvetlanaPlease confirm that this problem no longer exists +I have a much more heavy weight patch for this that works in all cases I need to prepare for commit +Yes create template from volume still is not implemented yet +Duplication is bad and all that. +Thanks for clarifying! Should we also try to figure out which of the two was corruptI dont think its worth it at least not for the first passWhat should we do if the replication level is one for the corrupt block? Keep it I think +What use cases do you have in mind +Ill try to make time very soon +Integrated to as well +At that time the output will be hadooppsusrlocalhadoopoldhadoopoldbin ps ef grep collecor wc lhadoop pts grep collecor But if collector is running then output will be hadooppsusrlocalhadoopoldhadoopoldbin ps ef grep collector wc lpostgres Sep? postgres stats collector process hadoop pts grep collector from this we will come to know the status of the collector +Added setUp code and inner works with Hellbound. +I prefer for trying to catch the subtleties +Not sure what the right thing to do here is yet +The fronted webserver should be a proxy which forwards the request to the application server +This loses some information but on the other hand allows type inheritance rules to work correctly in cases like when a type hint like suggests a more specific media type. +Requirement for protletbrige distribution should be created +Please mark as patch available +So we are releasing without a fix for this one +Please review +Tomorrow I will clean a few small things then I will write an email to the dev list to explain the work that was done the next steps and how to test it +Do you mean the patch is causing the problem or does it not work without the patch +Please create a patch against the federated build that updates the jdktools tree to include the new tool +Hadoop users want to know how their jobs will perform in a new Hadoop release +Attached HTML output from the tool +The reasoning is that starting the ensemble then shutting down the servers then starting them again is an avoidable time sink and the tests take long enough to run already +You will need the fix for +Are there tests for this I should modify +Ill commit this patch but still leave the issue open for a bit to give others time to review the changes +We need to put some notes and comments on the CSV files and documentations. +The patch no longer applies to trunk +running requires that the user is an administrator on the local machine +I verified all the HA tests passed on branch before committing. +If I put my local repositories first everything is fine +think we can have the Form submit normally bypassing validation and recognize that on the server +Thanks Andrew +You can use a Timestamp as if it is a Date object but not the other way around because there are other attributes in Timestamp object that Date object does not have +Humm It seems not the size it is an other am trying to find free to reopen it if you can provide additional info about how to reproduce it. +Gary could you provide a final patch with your and Mickys latest suggesting includedI will then be able to commit the patch +I didnt mean to rush you just trying to find out who was reviewingI definitely agree that we need good documentation for this feature and I will file a new JIRA issue accordinglyIm concerned that the current ALTER TABLE doc is already a bit unwieldy and so Im interested to get some suggestions about how we can rework that doc to make it more comprehensible and usableI think we can deal with all those topics as part of the documentation JIRA +Moving to and +If there is such an effect we have to look at the other + I tested with Eclipse +Tests ran cleanly for me +Example of pom I am using +Hi this is the diff of the patched installation guide +Comments are more than welcome +I agree that it is a security breach but we dont have time to fic it for this versionPostponed to t you fix this Em +successfully build in but still need to fix tag issue so that can built it in has the same error can be seen on RHEL with installed +This is the wsdl from the axis version which is a good bit different than the one generated by +Mike the patch looks good to me +skewed datacombiner job job totalpatch min sec min sec min sectrunk min sec min sec min seccombiner and using skewed joinpatch min sec min sec min sectrunk min sec min sec min secno combinerpatch min sec min sec min sectrunk min sec min sec min secno combiner and using skewed joinpatch min sec min sec min sectrunk sec min sec min sec +I fave fixed all Seam Core tests but test +So disappointed that its arched for i but when set to armv the error pop up again. +Dont laugh I dont know python at allThis adds one fail fast check for the testNotice check +Closing and back to dev after specifying the main goal of the issue. +I have temporarily switched to the until I figure out how Im supposed to run Zookeeper +I just committed this + set fix version +It is a bug indeedI think I see where the problem is working on it +attach a straightforward patch addressing this +Could you please tweak the filesproject in srctestresourcesnorunnableTests so running a mvn test will reproduce the problem you are fixing +Thanks to Aidan Hollinshead +not in this would love to see some generic property attribute type extensions to page metadata +I think Jodatime can do this +Moved to since its not close to being done and its not that important for +Maybe we even make it an alternative for a real deployment as well as you like it Ill design Java based DSL configuration API which be matched Google Guice +Inclined to close this as cannot reproduce +All entries are retained for me +This should resolve the issues please reopen this task if it does not. +It appears that just setting the cidrs to allows zookeeper service to open up as the service itself handles the firewall rulesSo I changed the title to match this +as long as the unit test is for tests +Shall I supply a patchIf so what should the flag be called? call it require +This is minor since the workaround is to use which is the recommended approach for setting system properties that affect JVM behavior +Sounds good +If its value is true then each XML is open with Tools XML editor + could see cases where it would be nice to have the PK values along the top with the Row line +it appears +Yes also would be nice if reporter could verify fix in master +The corresponding left outer join form should get pushed down as a relationship query +Start +patch applied thank you much +Same thing as a patch file +productiondeploy +Got the last remains of in stuff which occurs when in processing is not switched on +It turns out the test is failing due to a race condition +We are also not dependent on any sleep calls +It should not be the case but there could be some flaw in the implementation +NN should be added for and +how long does a scheduler have to be up before the long compaction times happen at the next start up? Days? Weeks +On the plus side renaming the current implementation to might give the developer a better hint of the methods of its superclass tooWould you expect the cost of traversing the postings to be higher than reading the payloadPeter +But using the failure detector seems to me +Felix in the end I went in the direction of your original patch +Is this a duplicate of +Thanks Eli patch command did not show any conflicts and committed based on above but has will resolve and update it +If the combiner does the same then we have the combiner write text while the reducer expects typedbytes +Pull it back in if you think different +Maven core continues to provide the versions for the lifecycle mappings +The problem appears to be that a permission which was granted to the jar files was not granted to the classtree +Ben In Branch build still specifies If we can change this to I can check in the FIELD replication stuff for +Hey MateiSorry for not responding on this for a whileWe have put this in production for a long while and it fixes the problemI think it is OK to commit this +Updated patch with bug fixes and a simple test for reordering that passes +changed to correct Seam to jfreechart with as fixed issue to release notes as per. +maybe group events by parent then sort by timestamp? It would make the log hard to tail via ajax since newer events could appear anywhere +Sounds good except for use directly the object instead of a copy +ve added the missing class +that sounds completely unreleated to how keywords must be quoted and a +I didnt mention that explicitly my bad +Its valid to see two different packages of the same name +Why critical as a priority +also makes partial files visitible no? Is this breaking something? If not Id leave things +I totally missed that the expectedVal and actualVal were cast as Strings! My comments would apply to a more general method but not to this case +Commits applied everywhere +In BSP case it looks like data is needed +Also the marginal benefit of making this a singleton doesnt fix what ails this class +Every project is different and has different set of requirements +FYI The code is pretty polished at this point +File Contains old and included a basic junit test for this feature +The system test only tests the watcher code +Patch applied +still a problem for you +I dont know why that is but I have reopened the issue at lower severity because of it +A Derby database is a directory containing the following contents +Is there any documents for these? Thank you +One last observation could you please amend the reference guide as well? A patch is just fine +Have you tried again with these fixes +all others went wellIts loops per version +Here is one of my html files +Verified by Richard. + +I think maybe we could break this down into a smaller number of changes each with a single themeFor instance changing all the memory allocation into macrosstatic inlines would be one fixing up the vlas might be another the struct initialiseers another the char const char changes might be another etc +Then the issue is resolved +There is one more warning in writeChunk that is not introduced by this patch and is actually safe +Thanks +About the version number then not sure if that is better or worse than or without which normally is used as qualifier separator +Looks like review board is down +i type f ctrl space and no components event or metadata is present in my completion listHe doesnt know component and obviously his the project created by ToolsCould you show your +Alena Can you possible to handle the NPE +i just noticed that you can have headers with deep hierarchies this patch only adds support for one level of nesting +I think you need to update your trunk and regenerate tha patch +original resolution of this seemed correct +Do we keep the old wiki or move to wiki or create docuementation on a repo +However there are several queries that I think can be tuned andor lumped together +The changes are committed +The most recent development has been done by Gabriel so I Would ask him +The older release of the plugin are not available +Thanks for looking into it +Thanks for reviewing Ted +In any event yes I bet you can spot the small bug an suggested a way to handle the null here +This patch should apply +Okay Ive added this line to the end ofif not OSWindowsNT cd PWDPlease test if that +This fixes the problem for me +Patches has been applied +Thanks a lot for the examples that helps a lot to identify and solve the issues fast. +Will post updated version soon +Heres a patch with unit tests which fixes it +Thanks for fixing +Heres a possible implementation with +That explains why Subjects are compared for equalityidentity +Version Released +Will look at it later. +Also it does not guarantee that validating happens only when editing is done +Could someone provide an example domain to test on? Jeff could you provide a patch or diff explaining the change you made as it is not clear to for now as no test app as no reproducable test app supplied +Working on it now +Is there a benefit to making these changesWill it break any important user features +Snjezana Ive moved the Find Seam +Can you try to emulate the actual problem thats been reported and see if you generate an exception +Thank you +As far as the application is concerned the Statement and s are closedI see it as really an internal optimization +Resolving this as a duplicated of and. +Closed after release. +You are right +Verified on EAP Affects and Fix in Version +Patch applied and pushed in bdbc +But is static +My BRMS was started and with my Mac this was greyed out I guess you have tested on a mac and are not seeing this + +So anything we do will be bolted on and break should the ESB schema change release to release which it does frequently +pig has been released +batch transition resolvedfixed bugs to closedfixed +Widgets are now installed dynamically as part of the build process +So needs more investigation. +Great +Thanks Ning +HelloI am looking into thisRegards! ThanksThere is a partially built one in the already +Would that live you enough time? if you want them in the beta release on the th July but perhaps for middle of next week +please check the service reference for other examples that start with isperhaps you did not notice the name is is Invoice In Foreign Currencythe output is yesnoRegards you did not notice the header of this issueI checked the files for methods with their name starting with is +Because these are still BETA quality I will update them all after public review by Japanese you. +The parent JIRA is to fix areas where trunk actually fell behind +The bug referenced in that tilt comment appears to have been fixed some time ago +I think we can change lock node to Ephemeral Node +Retargeting to fix in since has been shortened and made more focused + see this is fixed talking to Steve he mentioned that XMLTABLE can accept full expressions and XPATH is subset of the Xquery +They are different retval is not in loop its single value +Theres been too many changes related to parameters and jdbcdatabasemetadata + +Fixed thanks for the review Grant! +Pushed the fix into both master and +I primarily plan to modify the noticelicensing information by replacing Yahoo with apacheHow much history we could transfer and how? Any pointer will help +Moving all issues to the CR +is that even possible +It will be necessary to update existing code +Let me figure out where that would be +long divided by long returns a in Groovy this part is ok +That was possible because of the changes to the render processI need to look into to evaluate the is that fragile +Use cases for QE Open runtime pref page add new runtime path then a assert new runtime is added b press cancel c pref page assert runtime is still added in server view but runtime PATH is not on the pref page create one runtime path press ok runtime pref page see one runtime path +I committed the patch to trunkThank you Ivan! Thanks Daryn for the review +Can you please put some more descriptive text about what you are planning on nuking or removing here +Leave it opened for now and Ill return with feedback when is released +advance if the filter looks sparse +Only line now differs +No backwards compatibility issues +we can attach this directly to the wiki no need to check into the source tree +Ok +many of the exceptions have no info other than the stack trace +are not final versions so its not needed +If everyone else is OK with the packaging and changes I would say go ahead and commit +give me some time to test in cvs head +Verified on. +That is good both for humans and IDE if you need a partial class its only letters to add. +Before it can be merged well need to hook it into the maven build also +To contrast I have only Transport threads like this oneactivemqlogs jstack grep nid grep Transport wc l +replacing with +the code already used to return a sequencefilerecordreader +Much thanks to Tim Bruce and all for having shown so much patience in helping to resolve this issue issues involving reading from streams are always very difficult +merged to +rev +this one can be set to resolved I think. +Sorry I missed your message +It seems to be working in tomcat +I had a question when you place the translation on a revision does that means that the opentaps demo or the ofbiz demo is update on the weekly updatesOne more thing Ive never created a patch for a translation files how do I do that? or its just the new version of the files +I can get around this by reading the whole document and modifying the part I want and saving again but would rather to a find for speed and atomicity +need to remove some unneeded logging informationone main problem here is to determine when to download file +It didnt help though +Introduced +Why is this something that youd like to do through MavenHow would you find? Will it be in the repository? Wont it be easier to just load the jar into a classloader and run the commands in the same vm +I think I couldnt distinguish the terms on each record and reset on each withinI have created a pull request for my implementation on github if someone has time please review +Committed to trunk +With this patch I was able to index lines of multiple files +jrebinjava jar pdklib you go Doug customer just sent this in the sigar version infoSigar version +So far We have only implemented methods in +reopened to implement further deferring of application loading and merging into th e list of created to +It solves the issue but it looks like there is a more Plexus specific way to do it based on the archiver field xdoc so feel free to correct +At revision thanks for the patch on Aug s build. +Ill be presenting whats behind this service at the upcoming Lucene Revolution conference so it would be great to see the new search up on +Thanks for the quick response I was about to submit a patch myself until I noticed your incoming change +Is there any one working in this ticket +Im not sure if my comment was clear +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +So CDI should be able to do thisThanks +Any reason flume artifacts follow a naming convention of instead of a regular +refer to comment in patch per v this patch is not required. +But on DRLVM this function for some reason returns which means according to MSDNI didnt find out the reason for this failure so far +At a stage where exists in a stripped skeletal form +I didnt know this method was existing and it fulfil my need +are you saying this isnt a bug or that its been fixed? wrong selection +When NetSSH attempts to load it nothing happensLine in libbuildrcore netsshauthenticationpageant if RUBYPLATFORM java +The Quote button would setup a reply and automatically insert the content of the entry in a quoted block of that said there is a workaround +Fixed the POM to use the endorsed standards mechanism instead of boot class path Modified the POM to package the jar before running testsThe tests themselves have only been modified very little +I am working on this is now committed +Lets create a new bug report uncomment this one particular testagain +The test case can be replicated by creating the grails app dup copying the into the newly created app domain classes controllers hibernate am currently getting this very exception while trying to add some hibernate mapping through files to a grails domain I this am using grails so I think it is strange if this old bug should be my problem though it sounds like it is. +If this is really important lets fix it and include it in the beta RC +Bulk move issues to and +Then you could set steps to three and you could set the ETA to times the processing time of the first step +That is true if Im using the input calendar then that fails too. +Exchange messages as SOAPUI displays provided dump and your last comment it seems you have a valid suggestion when connection is closed after response headers are received +easy fixable +Note that it is the Kb file that was uploaded on the th August that contains the additional fix to attach source directories to JAR dependencies that relate to modules in the reactor +so it may not be the appropriate solution +I think both of them is fine right now +Big from me +Grant write Is the slightly in the noise? Seems to be +I am using Axis and I consistently get this error +The buffer implementations may be removed in favor of the Queue Deque interfaces implementation available in +Ill upload a new patch soon +I dont really see a reason to stick the version into the block filename though +released issues. +required for class unloading +Actually downgrading now to enhancement with a priority of minor and a fix version of later +could be used +Probably the method selection process needs a refinement here in using the distance to Object as distance element for this +the I attached the same file twice please choose either of them they are CarstenFixed your patch is in trunk at rIt was pretty recent and only introduced in trunk at r +Patch applied in Thanks for the patch Michael Osipov. +If user doesnt change pattern in the drop down list selected date in the calendar will be displayed according to Mmm dd yyyy hhss +Thanks. +Targeted to both and +This may not be safe in case theres a crash or somethingComments +Ill look at the rest next week +That sounds fine Laura +If the service is rejecting it then its a bug in that service + +The reduce bars show a stack display of copy sort reduce phases and for each phase the average progress is computed +Theres already support in for the escaping methods we need +But could you please attach the changed +Please ignore the earlier attachment +I think the latest Sonar version is able to combine those reports so we could simply look at the Sonar output +This would allow to have multiple services to provide roles just as multiple services for users are currently possibleYour proposal to change the list title would make it clearer what happens +Attaching +Take good care +thanks for the patch and please close this when you have verified the updated server code. +New make dist target generates source tarball +Issue was +Should cure the exception +Testcase for the error condition and corrective fiddling for a while with this bug it looks that even if im the reporter i miss the authority to reopen it or simply i couldnt manage to understand how to do it +Christian I agree with Willem that the ability to flow SOAP over any transport merits a new component +If you believe the issue is not resolved please reopen or open a new issue +The AOP tests now run with a security manager enabled +Im applying this fix but we should really start to look at writing some specs as well for becomejava +Please review +We have implemented our own Predicates as the requirement is too specific +Anyway I linked the cause of the actual bug to this ticket +I think that proposal was rejected because the fix would introduce a compatibility change within a major version +Hi LarsIm working on thatI will open another JIRA and submit the patch +addresses this issue. +Straightforward patch +any idea why The patch does not appear to apply with p to p +Would it be appropriate to determine the version at configure timeIf so we could create a file modpython which would containversionMPVERSIONand configure would grep srcinclude for MPVSTRING and generate fromIf this seems reasonable I can put something together. +Congratulations on a great web site +I am not convinced this functionality should be a part of but if you feel strongly about it go ahead and commit the patch +I would be happier if i knew the was passing on windows and linux with itBut smoketester is currently not even passing at all +Thanks to Konstantin Hairong Dhruba Nigel Doug Sameer and others for helping a lot for this to happen +property interested in it + +Fixed with. +Reopen for backport and assign to myself temporarily +it touches on every invocation of make and doesnt work with Windows. +Im not convinced that having the loader skip empty files is right yet +The patch I just attached to contains a fix for this bug +Closing this jira it works with trunk thanks! +AS has the same issue you can expandcollapse tree only by clicking +An implementation should optimize for sequential reads and perhaps also parallel reads but thats a secondary priority +But i think you have written the class at first and maybe its function just to provide as default similarityMaybe i have not describe the feature clearly +possible operations on suspended deployments are Delete and Resume +Please find the stack traces logs and conf you created a test case that can reproduce this +Just committed this trunk +This would get someone familiar with the intricacies of Supersonic and how we can plug other things into it +Fixed in r just needs site deployment to update public website +If we do this and maybe another place or two would need to be updated to look at this new variable. +Manually tested +I have a candidate fix for this in the and will create a pull request for it +The emission from java version does not have the args that we actually startup with applied. +Are there somehow two jars on the classpath somehow +We dont want new file we want to the existing log file and just put more infomration thereI am still looking at the second half of the patch +Im afraid that classes are less flexible than XML config files in that respect. +For trunk at revision are for the trunk only +Do we have a after? I still like the old collapse component better because it had the afterI think there is another ticket but not sure if it is being worked on +Ive got a possible fix on the jclouds side but thatll require moving to jclouds which means needs to be put to bed first +If removed and they comply later Ill them no problemNote that I have updated the howtomirror docs to reflect we dont allow cusrom redirects of +Martin can you please have a close look at the patch provided by you +Well aforementioned brittleness is a great benefit as well as a downside of code injection +Fixed +mgrep yuiUsersplindner +Perhaps even an archetype isnt necessary just instructions on how to include Seam in your project using Maven dependencies is fine +Last I heard Yahoo! was still using Leopard +the nio utility class to provide support for JNI methods +I forgot to add the addPackage functionality to this task +No Solr change to add the textko field type +If mmap does not provide a performance improvement on that platform users of that platform can disable itOK + +Do you want to try the latest +applied org +Patch for adding ignorableWhitespace if the last character is no whitespace +patch allows to make modifications to system keyspaces data but not to its schema +I have changed mb limit to mbAlso implemented this feature for apache module and for simple tcp server +Just from a JMS spec point of view is it allowed to add support for extensions such as function callsDespite it could be a nice feature +Added a Listener to run on dispatch only +Proc when getting set up as a thread attempts to access the static scope of the block so it can ensure it wont share the lastlinebackref scope where the threads closure was created +On my system it appeared to pass all tests and it passes our own set including ones that check that orphan tasktrackers and namenodes can be shut down +Where is the above method calledDynamically thats how works todayBut its just a scenario to reproduce the problem there is no fix +Seems like it is data binding +Punting issues from RC to final +Oops heh I meant to type projections not projects in the summary +Need more log and what was happening over on server +Not accepted by osgi EEG. +That seems fine to me +You cant create an entity with key test +Will it make getting a jdbc connection without even specifying a URI? In this we can remove such checks for or something like this +To the icefaces guys +Youre right adding these setters to would be much betterI solved the working directory issue by setting the maven project the same dirty way +Also I am willing to make those additions to your patch +A nice to have feature to be incorporated into if time there any urgent requirement for this issue +And I still do not see why exiting API is not sufficient for this use case +Patch committed +Ok look at this applied with minor changes +This is not a bug +bug brought forward to visibility +Author tags would be imported from the source and will not be changedCustom tags could change any registered user. +Again the app server is managing this entire process so we have no control over how the object is constructed +I ran into a problem updating thrift to +The other changes need to be added asap +Thanks Christian for great job But when will community release the fixed camel version for TLV parameters so that i can used new camel version in my current project requirement. +however based on my testing the hex escaping by itself isnt URL friendly so it had to be URL encoded +bq +And so there is no need to gsh shell scripts anymore +This issue is fixed in svn commit r +Patch for review that refactors the code to use more functionality +Thanks +I would say marks all queried pc instances as dirty by calling +ignore the above comment +As soon as there is a way for me to run with this in Eclipse I can get my guys to test it. +Same problem however for the asfplugin +Seems like a couple comments about this here is appropriate to me +Resolved as duplicate of +Thanks Torsten didnt realize you were also working on it. +I just have a feeling we have provided too many options here and when we do that we cannot afford to have any holes +If anyone it interested in merging to let me know. +I feel that it should show some exception before shutting down +We have put in many fixes in the last month +There are loads of tools and techniques and other Spring stuff out there for cacheing +The first time I ran the examples the progress bar disappeared too early and the application was still downloading while presenting me an ugly gray box +I am not sure whether it is worth it to try to converge the logging approaches into in the Tika Bundle itself +Thanks Marco +Change committed to subversion +Via the discussion on the lists +Removed both. +Verified on JBDS and Fedora . +The current implementation of demux annotation has an issue to support multiple columnFamilies +will be fixed later +We spent some time starting the server and seeing no issues +Of course this means it does not get deployed but thats no change from the current situation +too badIm not sure at least is ready but fine by me +I think this is a duplicate of bug +Still a problem in some cases +Thanks for reviews lads +Cool +marking patch available for +To clarify no trim is only for fixed width parsing and can be specified per column. + +Made the change to OPP in ccbecfebcbefefaa +Hi EricIn revisions and I have refatored things a bit +Any suggestions +Hum +I just saw it and was going to submit a fixed patch +It wont update the password thenWould change mechanism to judge in VR rather than mgmt server side +cest la vie +I have implemented the failover in keeping with the steps mentioned in the aforementioned commentsI have made one small change in that the failover implementation will now throw an exceptionin both the cases of the failover succeeding and failing +Velocity is powerful enough to be able to handle markup so surely it ought to be possible to fix this somehow +Suggested patch +Marking as fix for +Unfortunatley nobody has had time to really into this +This is fixed in Hadoop. +Just need the english properties file +I used a to create the initial schema on the db then two distinct to fill tables for the two distinct phasesThe rest of the code in only a translation using assertfail statements in place of the old checksPlease reviewUgo +My concern is this will create a situation where if you Query the using you must do it one way but if you are using the File Managers included querytool then you must do it a different way +I will now incorporate those two blog posts in the first comment +bulk fixing the version info for and all affected issues have in attempt for patch +Should be easy to fix though +Jessica this issue is still present in network detail page when it is accessed from Network tab list network network detail pageReopening to fix this issue +Does someone want to work on this agrees has ideas +The bug still occurs +The Doctype is one of those +I am ashamed of myself I should have done it myself without coming up with excuses +What is causing the null to be inserted in the first place? Is it a database platform issue? Can they be ignored +Thanks Roshan! +Do you mean the stat option of ls command +Notice the +That is what I would spend my time doing if I had unlimited time for this again +thank you! +Dickson and I am looking forward for any comments +While I was testing with Karam I worked around it by putting in a null check +Committed to trunk +Instead could Master restarts the RS instead +Hello Per ChristianThe proposed patch does not solve your problem since it deals with the same symptom but a different situation +test block +Patch for tomahawk wo role addition used for autogenerating +Changes look reasonable to me. +As for switching the config order I wouldnt recommend it without a lot of testing +Its got the most duplication and the most JIRA items openThanks for both of your optimization suggestions +Hi RekaCorrect we need to update files under +Patch file to remove synchronisation in +You are very much on the right The minimum data points requires that a baseline must be calculated from a minimum of number of data points +The code is deployed in the Codehaus Maven Snapshot Repo. +Slava any way we can add the additional properties to this thing +Thanks for the patch Anoop +One way to cut down the list requires postgres cut it requires more than one server or specialized server configurationdoctoring cut it requires something other than EAP cut it does not import into JBDS cut it does not deploy from JBDS to EAP cut itand we are only looking for about quickstarts that pass these tests +blah lets not do that +For assistance press CTRLSPACE or type hint then hit database reverse engineer schema myschema package +Here is the patch for this problem +Havent located an obvious solution somebody needs to put time in on this but it may not be me +replaced all uses by is applied see commits of this issueThank you Ganesh Jung for providing this patch I will resolve this issue now! +My mistake for not marking this closed +would these partial fixes have forced the android keys to be lowercased even when they are typed in camelCase? The check is still case sensitive and the default is camelCase so no matter how it is typed in the it always fails to match +ran script against trunk and it gets lock timeout at end. +Gabriellewhat is myfaces doing +I do have dispatcher ERRORI am using Tomcat perhaps that is the difference +Let us use the plugin +What do you think? GuysThanks Martin Dimitri for the patches +This isnt required for the present issue but would be nice +Rado please try to reproduce this once the above changes are merged into master +Sorry attachment is not related to Batik but to one of my bugs in another Apache Project +if the would simple be comparable and say it is equal to what was open last time there would be no need for this listener IMO to begin with. +is already loaded ealier on so why is the static method fixup invoked at Thu Oct given that the class has already been loadedWell the verbose output did show was loaded +This adds a release audit warning as I mentioned it would +This issue fixes a bug that was only present in trunk and was never released +out file for the new test +Looks good +you try it out with inputCalendarIf you confirm this problem with inputCalendar I will take overregardsI tried inputCalendar the same problem occursbest regards you check with a recent buildBecause I had the problem too but it seems that it has been corrected recently +Cheers! +Tilman Hausherr hello finally it worksI just downloaded the revision and bulded it with Antand added the new +This comes from the implementation of pp which uses a hash to monitor recursion +We will verify it and we will close JIRA as resolved for verification +This now works +The thing that we need to take care of is the mapreduce side files +After further discussion were just removing this from JBDS not from JBDS +This works fine on my environment. +Batching removed deleted ledgers would reduce number of in topic queuer while is asynchronous op which also block other requests in topic queuer +Im inclined to close this even though we still +Marking the Release note needed flag because of the backward incompatibility introduced by changing SCOPECATLOG to SCOPECATALOG +Patch with absolute paths removed +I attach log dir +and the like +Im confused +Seems like we can safely change this +Ive added some info to tell me what was expected and what Im getting when a methods expected but the arguments are t say whats already been invoked yet and I think there are other places which still throw with minimal mesage +Ok thanks for the clarification +You are using both in setting up the hash table and in reading the fragmented table before the join +Just checked out sources from CVS and am still receiving the same invalidoutput using Sun JDK under Eclipse +In the meantime I have checked in the new concurrency test at subversion revision +Thanks again Dag +defer all X issues to +Can this be reopened? Yes its currently incomplete but leave it open until the request is complete please +I think all of these things are pretty much taken care of at this point. +Lee do you +Assign back to Michael when done +Ant didnt get support for tests until so if youre running an earlier version its runner wont pick them up +Also the etc classes in the JDK perform appropriate pattern caching when building new instances making it even more +Brian +HiHow do we use the fixed revision? I tried using the nightly build ofAxis that does not even recognise rampart module +Along the same line I dont think get belongs to especially since its more there as a facility method and I dont think those should find their way in interfacesFor those reasons if the patch doesnt seem to break anything I would prefer commit it as it saving me some time for now +Ive just committed this +It is limited to the JDK level here +Ill take down the confluence entry until this is approved +Attaching patchIncludes java implementation and documentationMoving shell modification to a ticket +The soap action is definitely a bug in ESB +svn rev Closing code change only. +Attached the complete results in case anybody is interested +This method if returned false would also prevent the XHR call being made +I have edited the JIRA subjectline inorder to reduce the scope of this task +i dont think anyone in our community is building or needs to build with these days and im certainly not interested in encouraging it at this point. +So it doesnt harm anyone but it is very annoying to see all these lines printed in on console. +Committed as r to Committed r to trunkThanks guys. +test think zkHost will likely fall out of so that we can put those in zk +Owen do you have specific concerns +After some discussion via perhaps it would be best to go with Repositories as we did with GuvnorBRMS? A Repository seems like a better fit than server in this case +Ilya the problem disappears when I checkout DRLVM codes and rebuild again perhaps there is something wrong in my formal working copy + +You want me to convert them as subtasks here? or only reporters for that issues only can dobq +the section is about extensions +Typo +Ive started to review the patches and will post some comments and commit part of this code over the weekend +Unfortunately for the moment I cannot afford the time it would take to make the changes required to test out the patchThat said thanks for your efforts! From the looks of it the bug would be fixed with this +In meeting so cant review too much now but at first it looks like a good list Ilya +Ill suppose this was related to the bad classloader association during startup which has been fixed some time ago. +Need fix for CP in CP by r. +A patch is provided in +However only a single line of code needs to be changed and the description contains both the old and the new version of that line +Moving these to for now +bq +asking for feedback from other devs +Yes +It holds a reference to a underlying list which is shared with all sublists +DROP COLUMN ALTER TABLE +Im not intending it to stay in +Patch for ctc recheck this was this a test issue? Can it be closed? Thx. +Yeah we had the assumption that the source data did not have duplicate docs across shards turns out that it does +There are a number of complex editing operations within designer that are NOT undoable +Title is a bit general and suggests cross rack and cross node group dependencies +The update in the for the file was not committed to the svnrepository +Still cant get the case to work as we get a ZK error in the asyncOpen callback +Log showing us missing writes +Thanks so much Tatu +This can be done in follow up jira +Nit in the table in the exponents e are given with varying caseAs for further changes I havent gone through all the manuals to look for occurrences of the limits but I imagine you didIf so to the patch +Hey Permaineplease correct me if Im wrong since QA needs the javadocs for the distribution then once you know which ones you are including would you be able to attach in zip format the that need to be included on docs? If the ones online need to be different API docs can you please explain that a bit to me +Logged against a console error seen with and +attaching cleaned up version of the patch +It is fixed in on most recent update site +Added a test app for issue +Thank you heshan for the contribution +Verified in BRMS as resolved in the BRMS release notes you used a class from a model in the expect condition of a test scenario and later removed the model the editor would throw a Null Pointer Exception if you tried to open the test scenario again +Maybe this issues is to hard to fix +sigh It seems like well have to move the tests in to +Mike Overall it is a step in the right direction +Thank you very much for the review! Yes I agree that we can handle combiner per inputoutput in a separate jira +Will reopen ticket if it arises again +if does not have the resources to fix it soon I could try to have a look at wouldnt say its major the solution is to eagerly initialize your association when querying or declaratively +Thanks fixed. +Im closing this. +Marking bugs with invalid fixed for release as fixed in. +The failure is a known issue +Yes Im sorry it was incorrectly configured testcase +Just uploaded the trunk version of the patch +It is actually sec +Thanks Chris +closing +Close database connection at end of the process +Things appear to have drifted a bit in the Corona code base since last I checked +There was some discussion about cartridge metadata with libra +Committed w one final change to to include the size of its superclass +Spec updated revision +Is this still an issue using +solving cacheloader inefficiencies +Couldnt the javadocsimply state that collections and indexed properties are not yet supported andwe could add such support in a future release? +defer all issues to +I committed this to and trunk. +We can write the tokens to a file and set the env +Sounds good Daniel +Or does your setup write them? AdamI am using hbm files and not Annotations to define my mapping maybe thats why the Version property is getting audited +I think we must be working on similar types of projectsI did something similar to the above but in two different ways I extended to take optional pattern replacement parameters +So it seems to depend on the default encoding +Improved version +No longer supported under +Unfortunately I realized that I had forgoten to handle one important case inside the antrun configuration +It is not acceptable to say were going to create this anyway and deal with the consequences later +Proposed patch for fixing this +a few implementation notes +default to CWDcall set +Unfortunately this is not split into Hive infrastructure refactoring and support they are all mixed in together +Only difference was I got different totals then mentioned above +Changes almost done +That allows me to build it successfully as noted aboveHowever Im not sure how to test that the build is valid at that point. +For the moment I put pubbranch paramer in ivypublish task in ant script +this was fixed in r thanks! +Through some means it becomes aware that new data is available so it invalidates the data is Teiids cache +Infinispan will append the cache name automatically. +I already fixed this once some time ago after a forum thread but did not get to test it Ill do it again +This is an annoying problem +this was released in but is now being reverted in due to overwhelming community issue went fixed in then reverted inThis ticket was changed to wont fix against +Resolving. +Could you post more details what you mean +This has been superseded by. +BTW who has reviewed the patchI reviewed the patch +I guess youre right +Hi NoelleIve processed out version of the logo into different sizes and and attached them to an email +Close after release +exists path does not match return false + +Freeman Youre right the Aries doesnt see that bean instance whereas Spring would +Also happens ongoto hostecommerceclick Drop Ship ProductsAdd x dropship with addcart buttonClick Checkout in minicartUse l admin pofbiz in login screenClick next on checkoutoptions screenselect no shipping click nextselect cod click continue to final orderclick on main from this screenclick Drop Ship Productsclick addcart buttonIf you logout cart will work again +And for those of you who read offline Ive attached the patch and resulting +Agreed the trouble is I havent tested the removal of the workaround with the current version of Oozie +I think we need to do this sometime +Rebased on current trunk to pass compile +This is first stab at refactoring +Sorry that i didnt answer your question earlier +Patch no additional work performed since the patch I reviewed and applied. +No Im not working on it +In my rush to run through closing outstanding issues I missed the point you was making +Thanks Karthik +What do you think about two tabulations treemap and list above the dashboard treemap +Martin please reopen if you encounter problems with the EAP. +But I could have sworn Ive used this before in WW long time ago so it should work +I also noticed a lot of and instances but they were regularly so it doesnt seem to be are configurations I tried Tomcat Mojarra Tomcat Mojarra Tomcat Mojarra Tomcat Tomcat Oracle JDK Tomcat Oracle JDK Could you please write me which JDK are you using and how many concurrent clients does your application have +Theres something very wrong with how Im uploading +Nice find Markus +I would want to test this fix by using a database with the MSSQL database detached and unavailable just to verify that there are absolutely places where there is a hard dependency on MSSQL +Koen You are right your files were valid against property files arent filling the database anymore but looks like the files are still incorrectly created a Jira for this issue see you want to look into it that would be great one thing to check is that XML files would be correctly read +reported to NN the new replica +If that later processing fails then a warning message should be output as with the other stores. +BrianYou can use the Apache Commons Compress library to add support for file permissions +Attached patch prevents the message getting built always when endpoint format is set to restNow the message get built only if there is a content aware mediator is present in the mediation flowIf there is a message format conversion from SOAP to REST happens within the mediation flow we usually set the messageType property which is a propertySo it is safe not to built message always when endpoint format is set to rest and allow rest passthrough scenario +When you select Verbose Output can it keep that setting for a subsequent callWhat do you think about trying full width layout rather then the fixed px +I think most of the differences between and versions are to do with classes like etc which IMHO will be rarely used +Backported to issue is fixed and released as part of release +Looking forward to hearing from youYour camel rider AndrejYou patch looks good I just applied it into trunk and fix some CS related issuePlease update the wiki page if you have time +To check file content we could instead simply implement a job that streams through all the files in the fs +We can create another jira wherein we can implement an interface that allows random fast lookups which would be easy to do +Ill post a patch to demonstrate +Changes to should not affect the test cases +Ill try to work on this today +Attempt to add an optional boolean parameter to get additional params with tests +Fixed typos in that break javascript +The upstream bug has also been closed as Cannot Reproduce Bug +Sounds like RAT has some of the code already but rather than comparing against the default encoding Im suggesting checking against basic ASCIIThis is because ASCII is a proper of many other encodingsIts a lot harder to mangle ASCII characters inadvertently its quite easy to mangle characters outside the ASCII rangeBTW checking against the platform encoding seems wrong surely the checks should be against the actual encoding used by the projectThis is generally or +Agree Hibernate plugin is a correct place for the shortcut +There is a bug in the release which causes it to never find the local copy of the DTD +That can have a performance impactIf you use instead of the directly we have detected some of the common and SAML usecases and have certain optimizations in place to avoid the SAAJ stuff if we dont require it +It looks like a great start but I think there is some more wiring that needs to be done to make this work +the last thing we need is another highlighter that falls back to this this already existsif we want to break index back compat default offsets to ON and require reindexing to upgrade to then Im ok with the fallbackotherwise I am as its just going to be a bad slow default +This seems to indicate a broader problem with scm wagon etc +Updated the fixed bundle +Dennis +Is gone in. +I also made the Set in the final there doesnt seem much point in lazy loading this member +Agreed on both points +to trunk. +Thanks Vikasyour patch is in rev +See also TCK test +Its easily reproducible but im exhausted at the moment +It is disabled by default because of backward have to specify the following JVM property on the java commandline to allow this feature resolved issues for already released versions +is it possible for a character to show up as the first or second character of a sequence in a String that represents a character beyond the standard set of x to xFFFF +Id like to use Derby as a replacement for Oracle during development to allow the database per developer principleI cant change my hibernatejpa annotations and have to set the generator to sequence +It depends on the proposed +perform a browse from Hermes JMS and notice the increase of +It seems that some nested jar logic may be to blame +Change to critical as it kills our client during +Session Bean Package Name Entity Bean Package Name Test Package Name dont work either. +Attached is the patch +This looks like what Ive fixed in r +All the major options are supported already +I added the example in the issue not specifically in response to your example but for future searches +thanks I did not mean the fix in trunk should be backported just that the same bug pattern had already been fixed once and maybe the changelog between trunk and branch could give a lead in how to fix the new one +I hadnt really thought about duplicates it doesnt really matter that much as long as the behaviour of the system is consistent with the set of policies in the policy fileI think merging duplicates should be OK as long as you never replaceupdate a DENY with ALLOW +Any progress +I will upgrade as soon as the artifacts are present on the central repo +The transition would not take place if the method returned false + This bug has been marked as a duplicate of +A small patch to fix the issue +cancelling for on Hudson +But I still think is valid so I will do my best to revive it and convince other HDFS developers of that +I update everything to current trunk and made the Field Values API somewhat easier to use +Now we have to wait again +yea but please add a comment about forcing a less aggressive splitting +it seems like license expired! am I correcthow do I fix this issue? do I need to purchase the licenseThanks +Rereading Patricks comments Im not sure this is necessaryIf we just read and send stuff to avro were going to get dataloss +Gladly +If you want to get it in +gray out the parameters in the suggestion that are optional Im not proposing we do that just thinking out loud +patch proposal for parent updated the pom to version of the apache parent pom as proposed by Pierre Huttin in revision I also fixed some minor pom issues and added the missing committers to the team list. +Do we concatenate their messages or take the highest one in the stack +has been committed to trunk as followsSending Sending srcmainjavaorgapachegoracassandrastoreCommitted revision +Bulk close after release of +Please review and commit the +If I change topic to a queue it works as expected +i believe you are looking for the stale resolved issues +But somehow I am getting the feeling that this is unrelated to this patch +Everything went ok except for for that there were too many conflicts +This is a flaw in maven and should be fixed in rc which is due out sometime in the next few weeks +SVN revision release come with the fixed version of activeio +However it also includes snapshots since paths may be provided by users +From that perspective doing it consistently for all such parameter annotations seems better than doing it Juergen +Well I would suggest to call it something like dummy +Please verify if possible +I am punting on howif we get the new HDFS token to to be used for log aggregation +Reading what you linked to and the example hacking instructions Im not understanding why we need a separate repository for this +The fix will be bundled with committed. +Eli do you still see that failing +Its not necessarily that and it is a bit of an idea +Good to hearoLet me know if there is something else what I could do +css editor association in RHDS +I couldnt come up with a way to salvage the green box +I dont have an environment to try this out in so I suspect its just setting up an environment if possible using virtual machines which I doubt +So just help command in shell will be printing out all commands functions and arguments? Am i rightVasanth JarcecCurrent code can do help command showing its function name alone but not description of each command +Guess Im not sure I understand why not give it Wlrpath +So just check in the class that the options are not used togetherThanks! have rewritten this +Thanks for reporting! Could you attach the heap dump of the OOM moment +So can this be reproduced on something that is a part of the project +fixing javadoc to avoid warningthe warnings on the are on areas of the file not affected by this patch +so this is the algorithm we are already using except starting result at instead of +Clint points out that this patch changed filter interface to abstract class +The is not parsed unless its a string +fixed by revision fixed issues and +I havent looked at this particular patch in ages but let me try and check it out again because I think this patch fixes a real problem and should get in not archived +patch appliedThanks! +Left comment on RB +I tried many locations inside the webapps folder but relative path didnt work +Patch applied along with some updates from joint review with Phil. +The simplest solution would be to modify the compiler to always add an ASL header but it would be nicer if the license was configurable +The file used to scan the lib directory was being constructed incorrectly. +So hibernate jars are only used for building and tests then presumablyI simply stumbled about the entries inSo only the hibernate update is relevant then + +I was thinking someone will work on that later and add some specific Groovy buttons +And i tested it in both the browsersThanks +Perhaps there is something we can do with the JSON annotations jar that would allow a value to be base encoded using the annotations +Ill look into this issue + upgrade affects only the JVM used to run Sonar not the Java code you can analyse +We can allow for multiple methods but I am not sure this should be a bus specific feature +resubmitting for bug is not a blocker so it must wait until this is a fix Im marking this open to get it out of our long patch build queue on Hudson +and selecting particular feature within features table +Forgot to resolve this earlier in the week +Anyway +It takes less time to make a github project then to interact on JIRA and that way the author can keep iterating on the code unencumbered +You now use pure Spring to start and then you need to define I will polish a bit more on this tomorrow +Thanks for the review guys. +At the very least the compiler will force then to think about an exception handling strategy of a sort +Committed Bens new patch up in the hadoop issue +I am going to close this issue +Logj is using the to create osgi manifests +I changed the function to call the checkDest again with srcName as the new destination +We will have to work around it in code +Thanks your code was quite useful +Heres something that solves at least some of the problem +Therefore I have introduced a separate project to home Bnd specific extensions +Mr Hadoop QA do I have to do anything special to your and test of version of the patch? Thanks +New patch is really simple +In fact the same thing happens when you create a that includes tests from other classes Maven generates just one XML file for the entire in Ant I see that Ant does the same thing too +urlcompressor has always been an experimental feature which was never fully flushed outi made it so that pages that have a compressor are no longer considered stateless which of course wont help you much not sure if we can support this usecase +Sorry false alarm +Prashanth if you have some ideas here let me know There is still a failed test in I was not sure what with errors and offsets in handle in building the for what the values should be in that block +second iteration of jmx s use the branch I created for this and other JMX work +Sorry forgot to add that an increased heap size is a solution to the problem +Yes Max +Accepting bug +please +works for me +So the code now checks for this and sets the transition probability to when it happens +Part of a bulk update all resolutions changed to done please review history to original resolution type +I own you thisIll test more grouplens data +We should probably resolve nodes and racks to make sure they are all valid locations +Added cleanup routines to phaseCommit and phaseAbort +Thanks for the update KaiThis revision is much more concise than the previous ppt deck and has certainly incorporated ideas from the discussions that were started at summitI would encourage you to bring your comments to the DISCUSS thread that was started on for identifying the moving pieces +resolved on revision resolved issues. +Uwe was already modified in Lucene see As for I will take a second shot at an expert API next week +Shouldnt be too tough +Bumping and changing I nabbed this one in ecf on master +Ill wait a bit to see if Yonik pipes in first +Fixed on cellar trunk revision +more flexible to just expose an errorListener property on allowing you to inject your own implementationYes definitely +Linking to because generated columns would be a workaround for some of the use cases listed in the referenced email thread +IMO this seems a bit out of scope for dont understand why you think so +Resolving this issue +A page refresh would get the panelGroup rendered fully +Hi RuchiIm not sure I understand you fully +I think more cohesion is better by casting +We would need to ensure that shared edit dirs is always first in the list of dirs that is being written +Refer for proposed patch +So all issues not directly assigned to an AS release are being closed +Verified fix in trunk +TomaszCan you please explain the exact sequence of actions you are taking? I would like to know which commands are issued and in which sequence +Ive commit the fix at r Please verify +I think the simpliest way is to do almost the same thing as An option like KARAFETC would be perfect. +This is the resource file +This patch would show the failed tests on top of hadoop goes into there should be no need to set vmem config parameter +is this not a bug? +Thanks Bryan for catching that +Applied patch at r thanks! +All yours Gary +Attaching already tested this and its not a valid use case +bulk defer of featurestaskswishes from to +Thanks for the patch David +Hope logj will stabilize eventually +There were actually two commits to fix the above bug +Final patch attached +When the thread exists it rechecks to see if fsRunning is false +At Ds Dmitriy the feedback loop between clouderas rb and JIRA has been broke w a while so fellas have been manually flagging postings to RB by noting them here in JIRA +Thanks jaikiran +Fix confirmed +Once is enough +Applied patch thanks +Ah I see now that it creates ext filesystem on EBS volume we use the default filesystem type for selected OS then +need not init again +If the default location is localhostfoobar thats what should go in the properties file +Now that YARN is in trunk the patch really does not apply any more +findbugs +Thanks Sriram! +But it did not appear to help performance at all and actually hurt it +Will close when I fix that and get it checked in +Fixed +we could probably implement a second AOP aspect to fix up the +Is it intermittent? I will keep an eye on it to see whether it appears again +Added the missing to trunk +Do you want to get this approved through PM first +Triage for +not sure +On version there is no problem is on first position +BTW Im using jdkClosing this bug. +Many thanks for such a thorough analysis +The default format will be used like in the dashboard header for example Oct +Also needs to use mapping of column names to track common keys among +Thanks Sandy +I think we will have to live with the fact that things dont work under sigwin till this issue is resolved by javacc +while marking missingReplica of BK got NEE +the reactor based cluster goal copies the expanded directories so update tracking is there already closing +Sent CLA request to Michael +will if i can repeat when i eventually get round to checking it out. +Patch adds validation to Java Editor for EL in Java StringsPlease review +etc +IT added +Patch committed to trunk +We dont need to timestamp the jar we can just download stable version from the web site +No problemI just quickly glanced over the pull request I still need to do a more code review +Yes patch committed +See +Id like to do a bit more testing with the final release +Ill let you guys decide if you shoudl but I think this should be fixed in Geoserver +JSP eats of a bulk update all resolutions changed to done please review history to original resolution type +reopen as we need r. +heh well if you want to be lazy with the admin code you can just plug in +If anyone gets a chance to look at the stack trace error above here is the info I think would be interesting what is the stack traces of the server threads when the timeoutdeadlock happens +Categorizing bugs related primarily to stylesheet elements etc +Reviewed all these issues and marking them as closed. +Propose to close no further info from submitter. +did you think if I close this issue as duplicate ofI will handle and remove the deprecated style on this issueThanks actually I already asked such a question in another issueThanks Marco +Committed revision . +Can you provide a sample request that reproduces the issue +Patch applied under r +Thanks Robert +OK indeed I can get the failure to happen using your Test running against a partial Wikipedia index I have +Please find the same and if everything looks fine please commit +Attaching and changing the example to what you recommendM srcrefPlease let me know if more changes are needed +The key is only used to correlate the various points during copysort prior to the summing process and is not actually used within itI think Edwards patch represents a small performance improvement in that shorter keys would presumably be faster than longer ones +committed +The encoding problem would still be there +Its just an arbitrary safety limit to protect the system against buggy sources sinks +The sad thing is that I even saw the recent post on the ML about that like five times +Committed the patch to trunk with revision and backported to with + +Makes sense +Its not a big deal but having the expiry validation outside the secret manager seems a bit odd since the expiry is an implementation detail of the secret managerAfter more throughly reviewing the code it doesnt appear the tokens expire after the container completes? More importantly the token doesnt appear to become invalid for launching a container after the token has already been used to launch a container? +The tests that failed appear unrelated to this changeThese tests have been sporadically failing in prior buildsThese tests have been consistently failing for a long timeRequesting approval despite the test failures +css files are checked in now +All link tags are being parsed and included +New patch will commit momentarily +bumped to current approach which tries DNS resolution is not hurting anoyne +The patch move jthread library to vmcore site +Resolving +In that the nkeycontact column is not created by the automatic schema generationShould I raise this as a new issue +some more cleanups to reduce the code size of this test +Hi Lars I just have merged this to +This fixes some capitalization and spelling errors in the English messages +Very trickyWell i can of course change the code again to rely on the config option only +Only the explanatory comments Ill get rid of the extra ones +Im happy with the layout in the screenshot +I would be nice if all the decoration solutions would have some nice logging feedback stating that the desired element doesnt exist +Im rejecting this issue for the following reasons Appender behavior is no longer the realm of in any This is the JBAS project logging enhancements go to you would like future versions of AS to include automatic file compression please open a new issue in the JBLOGGING project under the +Hey guys Im sorry two of my mentors are on vacation and since were a git based project I need someone to setup an SVN repo for this +Please run and the unit tests + needs to be addresssed so that the matrix testsuite can stop servers when passing the serialization flag +There are no logged against it that might require them to change may be additions in future releases +Please post logging configuration questions to or +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Then will commit there +Ill look at it +removing the priority as suggested by the last commentFor the moment you can create the axis service your own and use it +Tests have been added +Ian and I had a misunderstanding + +Bulk closing stale resolved issues +Non logged in users can make suggestions and logged in users can accept themThe user jsc has full permissions on all projectsYes please have the pootle developer contact me directly to see if anything is amiss +When folks ask me Can I trust Lucene? this is the kind of thread I point them to +When those changes are later merged thechildOrder property will show up as conflicting and we havent specified how a MK is supposed to handle such conflicts +This was done for a customer and its being contributed back nowBut we can contribute this to the right place +Attached patch +How about adding libtestjdo +Willem applied the fix thanks Willem! +At last count we have about modules each with independent versions +Added a patch that fixes he issue +get this in real quick +Ill go back and see if I can figure out the difference +I just committed this to branch +Ive created a patch for this issue +HiI would like to solve this problem but as I am new in Hadoop development community I do not how and where to start +Please check that the bug is fixed now +Release will happen in the build agent where the last build of the project was happens if that build agent is busy building some other project when you click ReleaseWhat happens if that build agent has been disabled or has been deleted from the system +Although I hope to be able to reduce log noise and in general improve the way extensions are added it is good to know that this does not necessarily affect loading times as much as previously thought +Attaching a patch for +doing a cleanup prior to a upgradesstables still results in sstables that were not repair and upgrade all seem to skip the same sstables which seems to go back to the original guess that perform is skipping random sstableswhen the node is restarted the sstables in question that have yet to be upgraded are logged as Opening and no exceptions are thrown +Integrated to TRUNK +I agree that flexibility of layout is useful +Ill put the test here as soon as I know how long it takes to complete +There should be just one line that changes to add the new topic +See we missed a spot. + +Thanks Shaz +If they fail we have no way of telling what the real error is since exceptions in the finally block hide exceptions in the try block +Thanks. +ThomasPLEASE upload the latest WSDL that you use for generating the codethanks generated by Axis with soapencfix +I am currently unable to upgrade to a newer version than because my code will not compile in anything newer +The term earliest version of windows may be +Duplicate of. +Will commit it shortly +where do the temporary files get deleted with this patch +Brilliant! Thanks very much. +ill fill in the details for upgrade later when we commit the upgrade patch +committed +Assigned by accident +The inclusion of randomwriter and randomtextwriter in tests is just a hack to get around dependency issues in mvn. +Isnt this done already via +Never mind just looks like a formatting issue in the patch eg one of the brackets is indented to the wrong level +If nothing is going to be streamed waiting on the latch will return immediately and finish the bootstrap +postponing to +That said a note about that in the README is probably warranted +Yes we are cool with it + and VS Pro installedLet me know what configs you need for buildbot +Adjust logging path to new class location +Fixes to ensure that only happens if the client did a keytab login earlier +I patched the released Jena site yesterday just ran the patch tool on each javadoc tree and republished the site. +patch for scribe data copier +Clients can control the locality of their data through careful choices in their schemas +Not necessary to resolve this in time for +Not sure if this is the right fix +Adding patch with functional test using real server +Are you asking that the plugin be included in core? If not please close your issueAnd its not a patch so Im removing that flag +I find it to be a nice way to get involved with development but still feel comfortable with all the tools Im used to using from the Java world. +Revision +fixed in r +Logged In YES userid is not going to change +Merged into the branch and into the master branch. +How do we proceed do I write up another software grant? Its K lines of code not counting the example application which is another K lines +I just tested it and it works perfect +the groovy forum url seems not to be working anymore + updated the example removing these two classes +Where exactly is the b branch +NET Compact Framework FYI +In youll dont need to add it in or +There will not be a moving to for doc work should be done in EAP +I switched my Firefox theme on my OSX powerbook back to the default theme and it didnt make any difference +My current team is using Java and Hibernate with XML mappings +Im making some more changes and Ill make another version of the patch +on no screen scraping yuck! Given that Hadoop counters are pretty limited this would be nice to have + +Why dont you open a JIRA for that +reopening as i am now able to reproduce the issue has been fixed as a of fixing. +Deepal were you able to verify this bug? The fix appears to be simple. +Im not able to reproduce this on either Ubuntu or Mac OS X +Will if I can duplicate +Committed revision . +Im leaning towards adjusting as this new feature fits naturally in it +It seems the gbean is still in the running state after undeployment +Thanks a lot Devaraj K. +The only trick you have to know with monkey patches is that the patch has to be in your main app same package than the original and you have to compile with the option merge into memoryThanks for reporting +Docs in branch are in +Is there any remaining work on this Jira ticket or could it be marked Resolved +I believe it is necessary when registering the test bean under jdk +This is still a major bug and we still see it in our production server sporadicallyI hope someone else can pick it up and run with it +Simone this has been merged on trunk no? Can we close it +Its simple and complete afaict +DONEAlso exec and run commands are misclasified as file commands in grant +Zero itself is not a problem +Vikas +I have a list of urls and it didnt work +Ill add documentation about applying transactions but long term I would like to find a way to fail correctly in this scenario +Peter has suggested a fix in however the traceback above shows a problem with so it looks like there are a few places that need the same fix so the fix is not completeJohn Gentilin remembers that in a previous bug triage that we would use the the code forA work around is to move the namespace declaration up to the parent element of the sort element +If you commit then I can include the assertion in the test +Can you attach the failsafe configuration from mvn help +Resolved with commit bcdcceafbcbffbAll the tests checking exceptions thrown with die havent been ported yet +EJB specific integration might as well become part of the EJB module which would become relevant once EJB becomes a standalone EJB cannot change implementation details without modifying the relevant WS integration points in the webservice might be necessary to modify SPI details for Beta +This will be superceded by another feature request to provide a pluggable eviction thread +All tasks completed. +modifies Erics patch to support export as well +Patch applied to both HEAD branch +Making it a method means changing find to +someVariable needs to be defined before it can be used and it has to have the value null if it should cause the NPE +Thanks! this patch fixed problems. +Patch for the Tiff you but please select Grant licence to ASF for inclusion on that patch otherwise we cant apply it +I think this can ble closed Derby has a way to achieve this functionality already +Verified by Oleg. +See the patch committed to trunk on +Updated patch including change log entries +Fixed +solaris +RASTER Type +BTW +found more description and made some changes to descripitionsthis is not ready to be put in the trunk or yetit is here for others to comment on so I can be complete before committing +Ill commit the patch within the next coupleof days committed +Bulk closing for +I think the simplest solution would just be to open all segments as rw since that is effectively what they are nowI can take this if you like since I was just mucking around on that stuff and it is possible we have the same bug in the +Hudson appears to be down +Test added in r. +Uwe it looks like this has been caused by adding support for validating and +Hi BryanHere I found that this exception occurs with the presence of the single quotations with table nameij DESCRIBE COLUMNNAME TYPENAMEDECNUMCOLUMCOLUMNDEFCHAROCTEISNULLI INTEGER AUTOINCRENULL NO D DECIMAL NULL NULL YES TEST VARCHAR NULLNULL NULL YES ij DESCRIBE COLUMNNAME TYPENAMEDECNUMCOLUMCOLUMNDEFCHAROCTEISNULLI INTEGER AUTOINCRENULL NO D DECIMAL NULL NULL YES TEST VARCHAR NULLNULL NULL YES ij DESCRIBE vCOLUMNNAME TYPENAMEDECNUMCOLUMCOLUMNDEFCHAROCTEISNULLI INTEGER NULL NULL NO D DECIMAL NULL NULL YES TEST VARCHAR NULLNULL NULL YES ij should find the table as quoted string case is preserveddescribe JAVA ERROR ij should fail as case is wrongdescribe JAVA ERROR ij should work note that schema name must be upper casedescribe JAVA ERROR ij set SCHEMA USER rows insertedupdateddeletedij should work even after changing default schema so long as schema is rightdescribe JAVA ERROR ij should fail since table is in the other schemadescribe JAVA ERROR ij Can use as a wildcard for table namedescribe JAVA ERROR ij describe APP +The original author of this probably was modeling it after s addData which has since been marked as NEVER because its meant to do single INSERT at a time and autocommit +However the problem with writes on the column family it was simply doing range scans of still persists +I committed the patch to and +I only used value which worked just fine with the version of theBest regards Henrik Uffe patch nunit tests +I will improve my code with your suggestion and submit a patch with unit case +Also does pmackinn have an Apache JIRA account? Have to assign this ticket to him +As verify pom files within the repository this way we cant obviously compile anything here +Got it makes sense! Thx +There is a problem with the increased value ofdefine MAXINLINEGROWTHFACTORPROF define MAXINLINEGROWTHFACTORPROF regression tests in debug mode fail by timeoutIs it necessarily change? Can we put here or leave the old value +Marking this as invalid and removing the release note needed flag since the language that would have required this notice in the JDBC spec license has been removed. +I reduced the priority of the bug to standard because this is no new bug and exists already since Groovy close the issue as Wont Fix since this is a in general ambiguous part of the grammar without way of fixing +Anyway its nice to see that you have integrated Axis +Removes jquery files from static +Could you please provide a concrete example +Committed patch as is after Udai run multiple tests without how about switching to nano seconds? +I dont have this problem in +Think this is a good idea and incorporating it into my implementation +How do we use the locker with poller today or do we have to wait for release? Since this was a defect in XSD and not code can you publish the new as reported earlier The spring initialization code prevents the JOB configuration XML file having a poller and a locker to be is going to take a wile to be released +Committers can you please review +Why is it that temporary files are being deleted when the application is destroyed +This issue has not been resolved yetThanks Allen for letting us know about the varopt specific case on Solaris as well your other pointsIll address this forI think this issue should be moved to the DIRSERVER category and assigned to versionWDYT +It would be introducing unnecessary complexity +Changing the Seam Facet Version + patch +It does is nicer to include the name of the property for which the value is invalid irregardless of the bind marker problem +Could you rebaserebase from kafka Thanks for the review! And theyre addressed in v patch +will commit shortly +Good idea be sure to run ant checkstyle to maintain proper formattingYou can see if is an option +There are already other feature requests for returning the number of group +Instead of using HDFS for testing the pluggability which has resulted in being broken several times I suggest writing a test in common along the lines of +This set of screen shots shows how to reproduce the call me at if you need additional details +OK good to hear that it works for your purposesFYI Spring will be released any day now mainly as a bug fix release for so you might even upgrade to right +The additional interface is really an internal value but having it share the same name as the other two is more confusing and can lead to programming errorsThis internal interface was renamed to +Here is my missing informationlegal name Claudio Corsiemail ccorsi at progress dot comuserid ccorsiproject to whatever you were doing again it hadnt been done when you tried. +I just found out that you have to be a full JIRA administrator to make fetching the custom Release Note field work the way it did before the upgrade +Thats fine the problem is that the context isnt actually cleaned up afterwards +No usable patch. +For me the unit tests initially had a different porpose to ensure we reproduce the original exactly in java +pong! I created the project request a while back I think +THANKS +Otherwise the default Aegis databinding is usedIm wondering if this should be split into two propertiesand The Databinding and the frontend are two separate things +I noticed that a shingle test usesSo I undeprecated it and also by using the same trick as inAll tests pass +Closing. +That is very oldIf you can provide a patch for I will try to find time to test and commit it +I will provide a patch in the next days! Erwanwith this patch the content list will be updated after the upload +Ive figured out that a works if called from setupRenderafterRender whether XHR or notb does not work when called from XHR event handlersa works for XHR event handlersb does not work if called from setupRenderafterRender in an XHR requestTherefore in the case of the setupRenderafterRender in an XHR request is tricky +Note had to make changes to jdbc as was apparently fixed but not closed +ZK is designed for high read modest write workloads +attach a correct patch +Assign to Svante for review +What are the modifications needed to this patch after which we can this and get it into the repo +Did you use the Bulk deletion page to delete those projects? Just in case have you stopped your Sonar server instance while the bulk deletion operation was running +Thanks for pointing that out Alejandro +So Would like to know if anyone there has a workaround for this issue +I will look at it while Ray looks at in roo git branch commit eedaebc +The sample of course does not showthe foreign key being built or more than one thread but I assume it wasprovided to enable the recreation of the problemSince the code wasnt posted I have a few questions +Andrea thoughts +Ha wait I will remove the dependency on like I did in earlier patches +Thats the reason why Ive explicitly specified in that implementation should be independent on any channel internal structuresWhat do you think about having that functionality provided by completely independent piece of code? I understand from is that it proposes primary and secondary channels as the first class higher level concept in FLUME where a new channel can be created by composing any arbitrary channels in a hierarchyHowever here the difference is that we are creating a very specific channel type whose contract and transaction semantics are clearly well defined +Since already include the library it is better that we get rid of mvel library in the is using the while is using the need to be tested against these both. +If you look at the error it appears that there is something really wrong with the XML libraries on your classpath +bq +We could add a new imagedeleted state to be processed by vcld +is now copied instead of. +The convention for us has been that it is ok to drop support from to but just wanted to make it explicit that it will indeed be the case +Nice one Uwe +Something like FINE so it would not show up unless you are debugging anyways +Its not possible anymore to override multiproject Ill check this +The correct correction +Can you move it there from its current location while you do this or would you like me to do that first +Support for positional ngrams is another thing to consider +Buildbot has been triggered staging site looks ok +I can no longer reproduce this +for the patch +We could decide to support some compressed xmi format +And many times questions do go unanswered they are missed people dont have the time at the moment so its best to supply all of this stuff unless you are prepared for a wait if their is no current interest in going over the patch +Thanks a lotBest regards by Richard. +It does not apply to subtree user class because that class does not even allow using refinements +adds a small clarification to the comment for zoocreate +featurestargetclassesfeaturesindata +This requires documentation +I get CPU usage from eclipse after a few minutes of working +We need to take care of this scenario during ASF to Upgrade +changed DATATRANSFERVERSION from to changed DATATRANSFERVERSION from to +documenting how to use our modified version of jsdoc +Sorry I see it nowBut Id suggest moving the licence out from the Quick Links to make it more obviousI cant see the point of hiding the links under Quick links except to confuse me! +Thanks for reporting these issues Matt! Ill update the patch soon +The Converter classes ought to be in +Recently JD raised an issueChunhuiYou have any patch for +Could you clarify what you had in mind for this? Thanks +Obtain guide from s not possible to include these changes in the GA moving to +Fixed with commit with latest build from Default template download gets initiated once SSVM is up. +The question is Should we provide such a ZIP filter at allTomcat provides GZIP compression as part of its HTTP connector rather than as Servlet Filter +Includes unit tests +Where is it located? just reread your last post and I see that things are working when you switch to jdk +This seems entirely wrong +passes for me both when snappy is not installed and when snappy is installed and native compilation is enabled +For its a difficult question +I encountered the same error +thankscommitted to as commit dbbebdcadcedabace +Thanks Stefan +Give ppl one pig version to get off the pipe and then we kill the old oneShip it +If the issue is hive itself is trying to create this dir multiple times I would like to understand that +Closing resolved issue no response from reporter so assuming ok. +Is it a need to be satisfied by code or a documentation? We can already bypass checks in the mailet chain for whitelisted senders using matchers such as and +Make sure we doc +CURRENTDATE will return a different value in any statements that execute after midnight thus leading to queries returning a different number of rows. +Thanks Brett +Nice workRobert youre blaming Uwe for your own future bad behavior when he threatens to use your tactics against you? Sweet +to trunk + the patch looks food +Unless I get a patch pretty quickly I have to defer this +Will rebase and upload again +Rename my file toI think a good solution would be to just remove the from the file name when creating the generated class name +I manually verified that using the old property name the AM was looking for still works +second issue is that after I restart my second node I would expect to see k messages on each node but again what I usually see if K node A and or on node B +Will commit it if Hudson signs OK +Thx. +patch added +Woops thats silly Ill commit shortly +Alejandro I havent had a change to look at whats going on there due to dayjob consuming all of my cycles I am happy to purge the git repo and restore write access to svn until we get the issue figured out +I have added the project examples from the weld is a problem with the example that is related to annotation processing described in you import you will get compile workaround is to the project select PropertiesJava CompilerAnnotation Processing uncheck the Enable annotation processing checkbox click Apply check the Enable annotation processing checkbox and click Apply again +Heres the patch +of the root elementanyone knows how to solvet this problem? Thanks adam +Please commit this patch +Issues resolved in +btw +In the meantime is there a realistic issue with block numbers being reused such that an old block coming online would pervert things? Should a datanode maybe be periodically requesting the for its blocks and checking to see if they still match? This generally falls into the space of a good idea anyway since barring reading a block theres no way to tell if its disk has gone bad +Did you ever update the Wiki with this new functionality? Thatd be awesomeYes I added and the timeDecay option to the Wiki page. +If you generate the report at the end with its fine however +Fixing Affects and Fix in Version +Excellent +I assume you would like to see this inThe handling of the various spatial predicates is in +Im not an expert in this area but maybe others could give some good advices however it would be nice to see some fixes and security improvements in the forum stuff +Thanks dmytro i will commit it +fix doesnt reflect keys consistently is workaround for a known issue of drlvm thread blocked on io will be interrupted by unrelated signal +Hi you create an application named and ran griffon on it? or did you attempt running griffon on a pluginIf the latter then Im afraid that wont work +sorry for lots of interruption. +I run derbyall and tests +Addressed comments +I mean actually create a new version +Work completed in archivatrunk revision everywhereRemoving +Im for option or for opening +Sorry about that fix is in revision fix is in revision . +Sorry I just found the download today and was excited +specifically decrementing the counter +There were quite a few unit tests failing +If someone wait for this fix please test it at first with will go on testing for this on more scene in some hours +The works by alone +This is test first so hopefully the tests are actually correct +And the patched jar. +Some people may have some dependencies on OGNL but since the change to reflection code I would be surprisedregards Malcolm I should have uploaded a zip not rarAnyway if anybody wants to have a look at the benchmark both attachments contains the same content +Great work Clebert +I didnt see this problem on any supported platform +Im having the same problem with both the vote script and my own have no need for what kind of pluralization what so ever but apparently its part of the MVC addon +So either allow it for both or do not +Fixed as part of blockers and critical issues are resolved but not verified +Yes this is correct fix for that ve fixed that in the community branch. +now records who is creating the file +give the web service builder a default parentId it can add to a configuration under construction +eli I think it was a case of tucu error run the tests from the IDE and it looks I didnt force a recompile and it was pickup the previous compiled class +If I remember correctly the bug fixed by was causing the affected transfers to fail silently +Logging should be controlled by the underlying log package and not subclasses +If you really want to support these other mappings I think the much better place are +With JIRA feedback is provided to the user as the query runs see attached screenshot named ButWhen the query completes the user must still explicitly open the Tasks view to see the query results is there a way to automatically open this view when the query completes +Let me try to answer your questions +The primary indicator is the presence on an exclusive lock on a well known fileSo how about initially creating the lock file if it is not there and then use only file locking to prevent multiple instances running on the same repository data +is incorrect +The other option is that we allow one or the other but not both +PatrickI had done that earlier + +Patch appliedalthough have to manually insert then test them because they got stale +hi thanks to everyone for all the help with this work ColinIll look at the code within hours +Patch added +As soon as the data source becomes available it becomes valid +Finally I found out that the UI is frozen because of waiting on push dialog and not because of trying to publish project without git repo associated +Good idea +Creating a new for each message would be pretty ludicrousThe genesis of this readLength business is hidden in the murky archives of the Thrift incubator svn repro +Issue rejected because Move to Git will reduce the number of jobs so well less suffer from big locked by upstream ping Further improvement on build flow should start on investigating +For more details I attached thread dump fileThanks +It would be useful to test if Groovy master which uses ASM instead of ASM still has the problem +batch transition resolvedfixed bugs to closedfixed +It does look like a typo doesnt it? Ill make that change +This patch enhances the tag doc by giving users more information about the relationships between tagsThis patch makes it so that on facets in the tag doc the allowed child component tags are listed as links which user can click on to be taken to that tag doc +Fix of Thread ID bugPatch radices quantity of Thread to xFFF to match lock word format and add opportunity to reuse dead threads in new threadsPatch fix described problem but doesnt allow test to works OK +Thanks for starting work on thisThanks +Currently it adds code to FLE and QCM to support observers later +Well Ive figured out one way around it but have an issue with it + +No problem like I said we have dealt with ANT quite a lot in the past so this comes easily +This patch fixed to release find! to master and. +Logged In YES useridshould be fixed in Username djencks +However the ui will be broken too if there are other exceptions right? What can do for that +The problem was that an error was thrown when the parameter was omitted from the config block in the POM even though it isnt necessary +grin +Let me review this a bit more and get back to you +Random thought what do you think about making your schema code output a proto idl? For the heterogeneous array option Id use a type of repeated bytes with the assumption that each bytes value will be the schema followed by the dataYes +Thanks for looking at it +The problem is the new XML validator now have red underline as it do not understand the new you double check in if this is now fixed seems OK to me now +Especially on examples found in specifications +Seems reasonable +Candidate for Heiko could you apply this change to the branch or if if codebase if you could point me to the relevant feature in specific dont need to be console change for you +Hi Kathey does this test fail reliably on AIX? The test may be vulnerable to timing problemsit is suspicious that the test cranks up threadsThis error may flag a bug in the automatic collection of statistics however +can we please push this a little further or can someone else please pick this upThanks reference attached the svn diff for the changes related to and +helper fixed in +Please provide a demo application that fails due to the escaped entitiesThanks +If you dont restart eclipse the old tables and constraints are still there and if delete of tables fail you will also see create failing +all issues that Ive reported and are resolved. +All tests passed in integration so committing +Hi Mariothx for the info + +I would like to continue the discussion and resolve this critical issue in Use composition instead of inheritance between DOM and DOC levelIf you agree I will mark this issue as resolvedThanks and regards status of this issue or fixed or no responsed maybe can be closed +The issue wasnt with the number editors but the fact that the scaffolded views do not format numbers based on the locale +Hi RobCan you review this change please? I havent actually gone so far as removing scavenge yet Id like to rework a bit to make it much more frequent before doing that this is just a little something that needs done before we could remove it +Can we close it or is this still an open issue +Thanks for the heads up +Ive just been too lazy to file a bug +And most importantly it works for our windows users while head does not +Yes it can be installed from into but as per the screenshot I only see one of the required wizards listed in the New +With the original circular linked list may be broken after converting an to an +Is a easy way to do this in +Now it only locks correctly before creating the at all and when requestingputting entries +Verified on JBT. +Issues fixed are +Just run mvn site and look at the checkstyle report or run the checks from the command line +pushed to change is failing the release build so Im reverting the commit +Oops was already fixed was testing a slightly older version. +Should not take hours to fix that +Reinhold if you can make tests feel free to tell us. +Heres few comments +This would allow scripts to consume the same information without resorting to string processingAlso I always found it quite annoying to be writing user facing text deep in the various internals like view indexing or compaction + +bq +really should always be combined with setting the startRow in the Scan +patch looks good to me waiting for to give us its verdict +Lets get back to this after we put more thoughts +only correction on title Olivier I have committed a modified version of your work in rev +Or maybe using +A token should be able to be signed and encrypted and admin can renew or cancel the token +Net fix parameter +OK +This means that all schedulers would have to be modified to support this but I can make the code generic so it should be fairly simple to do +When you say which value what are you referring to? and the same for the search algorithm + +groupId has to be jboss other versions are are also other versions in group the jboss group be used now? +Claus thanks for the very valuable lesson. +I am srowen at Gmail +just in SVN +some deliberation the previous patch wasnt enough to resolve this issue for us +Please see attached committed a typo fix at revision +I have now attached a quicsktart test for super easily adding this feature. +It then uses that SAML token to get the conversation token from the service +I also added a link to the topic on for usabilityAttaching and with these changes +Thanks +Oh one other thing +should not be throwing they should be throwing a or something of the sortWhen is moved we should rename it to Nothing Gadget should be in the common directory +Version of the patch after discussing w Randall +Applied and doneThanks for the patches! +Dart bumped to for now +Completed At revision +Would likely need to see the full wsdl and test caseRequest for a complete testcase that demonstrates the issue unanswered for over a month. +As I cant help you and you seem to misinterpret what I say it seems we cant have a discussion about the bug reported +I think now is a good time to ask againRegards +JMSSERVERHEADERS is now replaced with JMSSERVERREQUESTHEADERS +I have back ported the rev into Seam and FP the FP from the fixed versions as this is implied from it being fixed in the equivalent CP and thus it does not require a release note for the FP release + for the patch +Can someone else please reopen this +Ive added another optional interface in revision This is a POC for +Patch committed to both trunk and branch. +See some of the other for examples to copyAnd in the file there is some javadoc grouping of packages +Entity files that generate the described error +Modified this issue summary +This is working correctly nowVerified in +Patch for and released to customer as a fresh to merge the patch into the branch. +This is done by using EAP server adapter + with Fuse Camel Editor Feature +Ive committed a patch to trunk which should fix the problemCould you give it a try and post your feedback hereThanks! fixed in SVN trunk +If Override eclipse at least will complain if nothing overridden and will underline the fact that there is an to considerThanks for doing this Lars +Thanks +I hope you can look at this when you are free +xcodeproject +Bulk move of M to M +Since thats cleaner than invoking directly I amended the patch +Please give it another goAnother reason why I feel so frustrated is that I had a definite feeling that JAMES community was turning the corner +Components dont have an explicit actionExpression definition +I still couldnt get pass the compilation error +We need a base class for all these things to use +I posted a program to which would be interesting to run with this patch with tracing on to understand the server round trips in various scenarios +Based on a cursory glance all replicas should be holding the same data so I think streaming should be happy but please confirm +Verified new version is called down +Release Notes Docs Status field changed again as its not really a new feature rather its an enhancement +One possible application ofrestricting sets of entries by a specific valueA table with several columns for comments title text author +I can fix the javadoc warning in if it is ok. +I think this should go to xalanC component +Closed again based on comments that release notes are not required. +the outtertag prints html header and footer info and the inner tagdoes a DB query +I was under the impression that SSL was required at all times thanks for putting me straight on that one +Mine is Python from usrbinpython +Otherwise were stuck to using clientProperties for this kind of metadata which is available on trunk anyway +The source code I used was updated tonight at about pm EST +It makes sense if its a general design decision. +Im going to need some more explanation possibly in improved commentsWhat I understand so far removetoken force is the force method +Also check other flow controller examples and update is implemented such that no synchronization is needed because the framework implicitly will be locking calls to flow controller methods and flow object methods on the associated flow controller instance then this Jira should have as its action to simplify the examples by removing the synchronization it will no longer be needed +Closing. +I too have been looking into this without much success +Strangely I dont see it on my development machine but only on the production one +Assume this ticket will also cover the impls for these +This patch fixing obvious issues +apply my first then then +Thanks +Anyone disagree? +I see that the test was succeeding before this commit and failing afterwards +We could probably ditch the JMX stuff as we dont currently use it +Our problem is not state being in ZK it is that the state is in multiple places in ZK itself for different parts of the same regions state plus some state in master to reconcile these plus some state that is not in ZK but only in master plus also metaMaster is the Actor +I have attached output as an attachment +This patch addresses some transaction related code that could potentially lead to lock problems and transaction leaks +All tests pass +Holy Bacon Thx Sandy +Schedule more fixes for the version +Should it be closed +Btw I agree that the method name is not optimal at least atm +Could you also attach your +I agree would be more powerful +That forum thread might indeed be related +Were going to address this in because there is enough of a workaround for examples +So if its the new one work as we want +Kendall what sort of chargeback are you thinking about? Just reportingOne of the reasons we implemented hierarchical queues is to ease management for cases like youve talked about here wont it suffice to haveI have sympathy for the notion of chargebacks but Im wary of adding new concepts to YARN without really into whether its really necessary + for master in git +Merged into the master branch. +Attached new patches that does not cause potential backwards incompatible signature changes +Dependencies look alright too +defer all issues to +Its always there +If you finish it before we release beta please correct the versionThanks +My first jruby patch be gentle +That way the server will not be overloaded in the case of fetchAll +I have sent an email to Dimitrios who Im told should be able to assist with moving things around within the to a limitation in JIRA this issue cannot be fixed. +Attaching patch which fixes this problem now the same configuration data will flow forward through the map task +If you still have tis issue please reopen with all necessary resources to locally reproduce this issue +I could not reproduce it if it is an issue it will be some kind of race condition and probably more to do with servlet rather than REST +I think I am going to close this Wont fix because The test doesnt exist anymore so it is not hanging up test runs +this version worksvn up r devtrunk +Fixed in r +because initially we have intended to support is a limitation coming from IDM integration and how it is used +It is clear that it is not an issue it must be a project configuration issueRegards Thank you for following up on the Glassfish side of this +The logo looks good +bq +i mentioned in irc and he responded with i thought i fixed that +Attached Sample PDF with missing version string in Patch file for applied the patch in revision as proposedThanks for the contribution +Open other bugs for new or unrelated issues. +This still doesnt work for Types with subtypes like and m attaching another patch with some more testsAlso updated the to use the existing logic instead of replicating it where the fix is applied +Verified with JBDS in order to update description +The latest patch looks clear and correct to me +I added configuration mentioned above and solved the problem. +I was initially thinking of doing something similar to the Protocol buffer implementation of +Doh! There goes minutes +should go in first to avoid any breakage until the commit of the hdfs portionsI will do more testing and add test cases +Test case to reproduce issue +The URL problem is fixed as far as I can see please open another ticket with a specific test case if you still have problems. +it will work with cfreturn as well problem was that expression parser not end properly because that the following tag parser runs into a message depends on the state of the tag parser because that you get different message. +i ran against latest base patch and got no errors on a full test suite run against ibm jvm on XP +And that is for to be see +it is often the case where one wants all values that arent explicitly mapped to go into a default fieldWhats the real to be able to search all metadata? One could use a dynamic copyField into a single indexed field +r on trunk +I accidentally set that value when creating the bug +RE encryption in memory we might want to punt on this for now and wait for a future hardware solution that just encrypts everything in RAM +Fixed in the latest svn +Would you be able to check it out for me HowardThanks for any help you can give me JaredI think you need to add to Tyronnes listJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEclientJBOSSHOMEserverYOURSERVERPROFILEdeploy +The multiple goal is very and a little tedious when you try to include any set of dependencies +There is no way to tell the difference between a jar and a zip file without looking at the file extension. +This should be easy +I see it already there orgzkosszkforgeel from september made a mistake that the previous version shall be +I have written and attached a Testcase for the problem +Right now all the existing users have push rights to the new repos so we can close pull requests etc but Ill try to come up with something more elegant. +How are you sending the field namesIf you are really making a GET request with arbitrarily long stuff I suspect that will cause other problemsI dont know much about the updateextract handlerWhat Request object are you using to send stuff to it? Can you just change the method to POST +VFS listener receives a plain text message +Ive been out of the community for a while doing other things but am about to start a large project that I hope will lead to some interesting contributions along the way +I agree Charles +Should update documentation to explain this +There seems to be some issue applying the stuff +It is relative to the code so it duplicates +Groups will not be synchronized when property Dn is empty +Im probably between that and just disabling the code entirely +last lines of patch but looks good +not blocking release. +I think this approach might bear more fruit than the others as it worked before in and I probably should have incorporated into the locality patch already +cr is used before it is tested for NULL REVERSEINULL +JBDS Update in this case can be performed only by root then +Maybe we can label it as a improvement rather than a bug +This is for reference only +Hi Benoit in fact weve introduced Reviews in Sonar and its already possible to add a comment on a review containing a link to a JIRA issue +KonstantinIll fix most of the nits you have pointed out apart from this one max Having this variable simplifies testing +joining with the dependency direction +closing with proper status +Closing all tickets +So there is a warning comes to synapse end when vfs lib tries to close the streamSince we handover the inputstream closing job to another process at synapse end I think we can avoid this warning coming from library +thanks for taking this issue +though on linux also if somehow it happens that HIVEHADOOPCLASSPATH env variable is present in hive process we might run into issues +Patch tests passI had to fix up Edge to work w spoon feeding but otherwise no analyzers seem to be failing at least on one runI had to do some sneaky things with to work around its state machine +Test failure is unrelated +Before tracking the bug it is hard to resolve this task +I know the team is busy but if anybody has an idea of where to start looking I would be more then happy to dig and provide a patch +I have committed the changes in support of that do not depend on it in ron trunk and in r on +Two questions you say after creating a WAR project I assume this would be relevant for all Java project types? Got an eclipse bugid for it +Im raising the Xerces example with the developers to highlight that perhaps the compiler choice shouldnt be part of the build style +I am able to find the cause but try to find how to fix it +Hi StazekNow that Lucene is final can we finalize the jar for this one? Also this final JAR will handle the license and stuff rightThanks GrantIve built Carrot binaries and tested them with Solr trunk +Is it possible that build takes some previous version from branch +In popup mode input is disabled and button is enabled +Lowering priority +Then I will widen its coverage +Since that time to work around the issue we have upgraded our machines with more RAM moved to Java and given our JVM more RAM to work with +In most cases the colorvalues will be just wrong because of the missing transformation +And by waiting to reach the lower barrier with a constant heavy load youll always run into +Uploading a patch for branch +Committed. +I thought about moving into application +Why is it attempting to read it now +Note that this depends upon the patch posted on +Should have code up soon +Scratch that last comment I hadnt deleted the celtix package that had been downloaded and unpacked to scabindingssunjarsBuild still fails with or without the s the full build trace this one actually finishes early due to a maven network problem but the failure to install the celtix jars has already happened further up +ANDapachectl executes usrsbinhttpd instead of usrsbinhttpdWhich one needs documentingThanks +Not sure how its related +if it is. +keeping in mind we added the restriction that a components parent cannot be switched +Patch applied merciPlease note a slight tweak on the patch Ive expanded the tests and renamed theComparators statics to CLASSNAMECOMPARATOR and PACKAGENAMECOMPARATOR tobetter reflect the fact that there Comparators work on the names only +Rick was going to investigate +Sending srcmainjavaorgapachecamelutilTransmitting file dataCommitted revision +Apologies for the mistake seems I havent fully woken up today. +Can you add some comments to the new test that explain what exactly its doing? Its really hard to follow what with the test inheritance going on +I still see but it is empty +FreemanYoure right I can see how that would be a problem +Thanks Daryn +Torsten can you have a look at this? I didnt see the problem quicklyIf you apply this patch and expierence trouble with the testcases please try to copy the and the file to javamainresources +im on it +Could not stop the domain far seems stable +Is there any reason why this shouldnt be the only behaviour? I dont see what you gain from not making it the only way +everything needed for this should be available at the branch only needs to be ported to trunk and changes have been ported to and +Resolved with revision of Commit message Update patch for too +I left it as it is because I thought there might have been a reason for it +Stan Im assigning this to you for now on the assumption is has something to do with the CLI GUI +Java a bug in + build does not support Cordova Windows is not Windows Phone pesky confusing naming conventions. + +Lucenes default behavior is nowSee for a more generic approach +Anyway was just a thought +Adjust some timing parameters +I found a bug and fixed problem was thatif a is made from Node object which is a instance other than Document a tree of is constructed with only root node +Maybe someone could try it out with pcntl im running on windows +Thanks for notifying us +Even though this is derived from an EE requirement this should all be done in a separate security subsystem as this a universal concept in Java +I did not commit it since there was no unit test for it but +This issue can be closed because the issue replace this oneThanks a lot I guess my process recommendations were not sufficiently clear +Looks good +Check Disable some items +Not surprising it failed to compile it depends upon +It fails because is a style test and it cant find any of the old style test cases in it +There were many IPC Handler threads to service the calls but only one Responder thread to send out the rather large response buffers +In a netbooted configuration isloaded via simple workaround is to add to theclasspath on the machine where the netbooted instance is described in so I am closing thisas a Duplicate +Closing old resolved issues +If they work fine is it ok if I mark them fixedAlso since the GSEs with all the dependency checking is only there in trunk are these issues only trunk specific? Or some of them may be present in too? I havent had a look at yet +Its cleaner and less error prone +Has this patch been scheduled for commit? I wonder what release is going to have this because otherwise well have to use an existing release and apply this patch exclusively for our deployment +Ill review the fix and check the thread naming too +So this one can savely be resolved as duplicate +PR for master merged in master branch +Why not set RS port to and use whatever ephemeral port it picks +However the warning could be over look +No response from reporter closing bug as cannot reproduce +Description is added and committed +Resolving this as there were no issues that were uncovered in the basic QE testing and unfixed that needs to be issues raised in the community to be addressed in EAP the person raising it has to be a paid customer and has to raise it through support +bq +This is already implemented but the developers are not including it in the source +Ill add your patch and if my testing shows its fine I can put it in +opening so it can be resolved webconsole can be already be used to monitor processes that run inare controlled by standalone clients +Fixed race added barrier to keep existing behavior for close +Alan updated to your latest patch +Patch available +bq +This is taken care of by the Logging POJO Service. +The doAs parameter name was chosen to mimic the Java API is kind of confusing do you mean the proxy user or the doAs user +If the supplied JRE installation failed no warning was given +Has worked for me now with FOP yay! +This isnt going to be possible to fix anytime soon. +Please test it for me. +Logj think that root logger is using a default appender and is using PIGCONSOLE +More often than not they would like to selectively archivedelete process instances after the fact possibly from the console +Thats a bug in documentation +Few minor nits posted on RB. +patched for are recent but the geotools version is unmaintained +At first glance it seems to work If someone needs i can attach the POM for wrapper +Done through an addendum +Why do we want to suck a splits worth of data? If its because you dont have running on your Cassandra nodes well go fix thatIf its because Hadoop has created too many tasks and all the local replicas have their task queue full wont assigning it to a TT just cause more contention than waiting for a local slot to free up? +We do not use the resolved status. +Ah never mind Im an idiot +This patch registers the schema included in the plugin in the xml catalogThe dependency had to be added in order to do thisOnly is provided I couldnt find the schema for settings fileThe patch was created against latest trunk in svnIve had a hardtime trying to test it so it needs someone with good knoledge on internals and cache to certify this is actually workingIf you find any issue with this please let me know +This is fixed now +Was resolved by up priority +This patch is intended for a replacement of the first +I have committed it to branch. +The issue lacks any description or you please provide more information so that this can be verifiedI verified that its possible to correctly chose Deafult or custom installation etcReopened the jira as it is not clear what the actual fix is and hence cannot be verifiedclosed for there is no paneloption to test this bug for in changing the fix version to +This issue is probably similar to +I just committed this. +Ill be finishing this JIRA for Lin +If you want to close it I can just reopen if needed +Use Case Thirteen Password Guessing Example Drop any connetion that fails to authenticate three Case Fourteen Negotiation Example Drop any connection which attempts to send plain text commands during Case Fifteen Monitoring Violations Example When any connection fails to authenticate times post a mail to the spool +We could probably safely drop it down towards where our lease lengths were seconds I guess +Committed to trunk a while back. +Then maybe they need to be refactored? Why do they have jbpm in their name +If you havent started Ill finish and actually commit it in the next day or two +Does it look ok? Should I commit + source and this patch going to be included in release? thanks +Sadhu Can you point to the test case? It seems there is some confusion about this feature. +I cannot reproduce this warning +Either use postObject directly or apply the attached patchIn both case the property cmisobject must be defined with the right value +Jason If you decide to keep the shading yes it can be closed +oh ok I searched the directory and could not find the jar anywhere there +The precommit failure is due to the fact that is required +I agree with Musachy commit it +Anyway will take a closer look at itLarsFor i will check if i can make the to pass +Please close this case +remove unused imports remove use of byte stream size char buffer that buffer can be many times too large for text remove redundant buffer size argument in readerstring +That way mvn will not pick them up and I will still have them available for archive purposes when looking for how did i do that before +Thanks Navis! +Since uncacheable objects will still be cached via findBy operation this patch will evict the cache entries of the uncacacheable objects from the cache whenever the entity objects are modifiied in memory and flushed into the database +Can you please attach the schema or a sample schema which also demonstrates the problem +Jim will make changes as indicated +When the typical group size gets a lot bigger than the number of bits in a long another implementation might be faster +I have a patch to cleanup the standalone iocore building that based on +For now we have worked around the issues and added some properties to the persister to control how it encodes +applied thankyoutook some tweaks to get this to apply please check youve svn upd completely before generating the patch. +But Mark Miller fixed it +this is just a failure to push the conflictstrue parameter through to the include docs code +Close this issueDevin +That works for me +Keeping the message but downgrading to DEBUG level so that they are not seen by the user +Nice and simple idea hopefully time just occurred to me that this hook exists already in the form of +As a side note the conf in hadoop is which controls the number of maximum permitted failures for each task +If the modules are missing do a modprobe kvm modprobe modprobe and see if the host gets added successfully without reboot afterward +When are attached on shared storage we can use node groups to group data nodes by either VM fault domain or storage fault domain +Yep I have seen it now +Prompting for data is not going to cut it for Hudson jobs and the like +Committed. +by the way when I add the project I checked For multi modules project load only root as recursive build otherwise continuum checks out all child projects as a flat structure +The trick is to add enough rowsso the sort is not in memory +Heres the patch to to Fixed having been released it is not likely to contain this fix as for it will delegate the indexing to SOLR and wont contain any Lucene related code +test cases used a few edge cases that cannot be valid paths on Windows +Currently we only get this welcome page when running standalone +The problem is that there is no fixed ruleset to check this on so that is probably why browsers and curl dont check this +Judy is this still an issue +If I get some troubles Ill report it hereThank very much you for your help and also thanks to GaryRegards alexejGreat does it mean with mysql database this issue is no longer validRegards have a look at attached file +placement where all replicas of a block are placed in the same rack +Updating fix Version to reproduce install me import the Maven plugins as Maven projects check their +Note that the nullability annotations in are marked as deprecated so maybe we really should use the JSR variant +There is already way too much hardcoded stuff which is specific for single in there +only just came out of alphabeta with the correct way is to update cassandra using right +Thats exactly what I would do +uploaded core dump occurs after applying the patch +I believe Hiroshi reopened this because it was not committed with test code +We should get his attention first I think +The ? above should be ampxdampxdda +Maybe I missed something in my a RegardsDonny +Do you have a sample application or code that you can provide to illustrate the exceptionThis will make it easy to reproduce and provide a tried to reproduce the issue to no avail using this JSF and CODI I can guarantee that this is not needed +The failure is not related to the just committed this +Hi David +I have committed this +Be sure to include your plans for migrating your websiteoff of confluence as that is a blocker for site migration +The latest archiva build doesnt handle relocation from the DAV ve reopend and will investigate this one when solved. +Another option is to throttle the rate at which the appends happen on the follower that reduces the sudden influx of messages at the follower and fetch requests at the leader +Not necessarily horrible just a gotcha +Patch to branch please ensure that it is applied there +Fixing this would be very much appreciated +Thanks +Overall it doesnt fit the current model entirely as thats centered around parsing attributes elements rather than handling attribute values +You can change hosts in this script to include a cluster of nodes or a single machine as you like to +I think Alena may have misinterpreted it +RM changes to handle NM heartbeat during overload +Code looks clean and simple as expected and has a test so Im guessing it worksI do think its interesting to push into the graph algorithms area as a new talent for the project so I do think its worth stacking in some clean simple initial implementations like this. +this is core task for migration of Seam to examples except WIKI are using maven and integration testing with Embedded as container. +Assigning to Stefan since he told me he would have checked this +Im making use of interceptors and the so were really seeing all the variations +Please verify that it works as expected now. +Brad What is you problem exactlyI made a new release of with in directory srcgroovyWhen application context shutdown look at for the reason +Thanks Juergen +Patch for +just verified with the fix in Origin one can currently log into Online again using JBDSThus closing as RESOLVED +using version of the site plugin fixes the problem +There is an issue with concurrency in this see all resolved tickets from or older +Except I was going to replace the existing get functionsI think H as a prefix should be reserved for stuff part of rather then thrift stuff +NET compiler +Need a clean way to load common init params for the dist search components such as timeout thread pool size and search handler pattern +Thank you for testing Marcel and Mike +See comment in patch +What tool did you actually use to create the patch? In case you didnt use svn diff can I ask you to do so and a new patch +Oh makes sense because the uses it as well +Hi Aileen this feature is now complete and in will make it into the release but it needs doco +a call to listNodes took me about minutesi noticed that the reason was in fact the time it takes to convert tois there any way that i can configure the result fields i want in the response? for instance all i am interested in when listing the nodes is the ip addresses and the names and i do not want the location the image the hardware +Hope that helps give some clarity on the bug we are seeing +The final fix will be dependent on the new UI and reimplementing the original flow which allowed the user to choose an ESME username that is then associated with the +You can configure the executed goals for prepare and perform yourself +The issue should be reported to IT department. +We have not throughly tested thisCamel should be used for SMX +Just for quick reference the prototype code Chris mentioned is in and +How would you approach the implementation of this +You should start a new project on Sourceforge or Javaforge and release the code under LGPL +Could you give me the versions of the bundles youre usingm just a little busy at this time +Can you regenerate the patch I would like to get it in before releasing and are mutually exclusive actually +? +there we go updated patch with the logic previously Screens to show what it will typically look like +I will now test the generated descriptors +patch for trunk +Also I think we want to log this at info level so that we get it by defaultI will change exceeded part +yeah we are seeing this in the Drools BRMS as well FWIW +Thanks for the patch Sergio +Ive switched to XR locally to check this new XR version are still crashing as before +Also the constraint make address sensitive which isnt ideal for managementaccessauthorization but is appropriate for many +So my previous proposal is incorrect +No Functional change +Fritz is working on the problem +Unassigned issues we get this on the Wiki somewhere? Ive been looking around havent been able to find it +Sorry not starting versions recent versions +In Starting server log entry ill show it with screenshot coz i cant extract text string from this not actual issue +Click on Provide Patch under Available Workflow Actions +Here it is might be worth to mention in the wiki that you really like one single concern per patch though. +GKOP fix commited in r. + +Lets mark this as resolved then +Do you have a number or link for the JRE bug? Do you know what versions it applies toThanks +I will create a patch and upload in some time +note im thinking about why there is a hardcoded sqlserver dialect in the generation class +That said there is still the issue of orphaned preferences for example when a user or portlet gets deleted +Dont tell people that they can put filters on associations its not whats happening + to fix fix version +Verified by Mark. +This should be fixed +Ran tests on line line and +As I said the change is to do with how inlining is done +I tried to change the log levels in the too but there was no change +one more comment +Lincolnisnt this issue obsolete through the forge commands as deploy as exploded etcRegards actually +bin line exporthomehudsonhudsonjobspatch No such file or directoryPATCH APPLICATION FAILEDRecording test resultsNo test report files were found +Wont Fix. +Curious if the idea could be pushed even further and allow the merge to analyze the regions to find pairs to be merged then specifically offline those regions and merge them rather than disabling the table +Note that issue looks like to be solved in Oracle driver Nicolas is this fixed in Oracle driver? +Also if the first line is not a newline a newline is not automatically added to the textarea rendering. +doneI suspect I wont like the leading separator I will look into it and commit a fix in approx h or less +So we need to keep the Proxy up to date with Accumulo to make sure all the features are fully available +Adds several new span query position checking queries +Unfortunately we would need to move the code down into Common +Nothing here either +Attaching a new patch which includes updated documentation +the fop script provided by FOP is not the one used by +Dupe of +Version April April JoergIve made modifications along the lines we discussed +Could you provide a reproducible test case? +Does this rule out bucketed or will continue to work? If it is the earlier shouldnt fixing that be a blocker of thisThat will continue to work +This is what the format tools give you +The second because of the firstIf this issue cannot be fixed Id suggest the grunt shell throw a more helpful error when it encounters preprocessor directives and warn when it sees likely attempts at interpolation +the steps to reproduce? server runtime configuration is in multiple places in the ui +This should return results but provides the error instead +Verified at +and indeed i wont get round to it +Oh okay do you think it is a solvable problemId like to add compatibility to a gem and this is a blocker at the moment +Committed to branch +fixed the obvious issues to adding findbugs to reporting tools as suggested in closed as resolved and released +Ill post it here as a progress indicator tomorrow anyways +In the case of Exception we dont populate cache +Yes this will be a very nice benefitI am now going through the tests changing them to try to assess which way is better +I have to kill the process +Thanks for this +I actually found another way to resolve the issue I had that correlated with this request +Just retested and on two different runs it took seconds and seconds so the exact timing varies +Would it be possible to integrate the multiphase for the slave with the existing codeI must also admit that I do not quite understand why there is two files +Bumping this up to critical because any row operations seem ot break the table +Thanks Todd +Fast +batch update of issues with fix instead of the correct atVersion id +The way we should handle this is by supporting XML namespaces +Dont forget to also rename the Metric in the Java source code +Randallgood points +I need to install a SUSE environment to reproduce the problem +This lets us keep consistency in almost all cases +Done IF +nitpick indentation seems wrong in the beginning of the test case +It might need more changes after is committed +MikhailIm not sure exactly what youre asking here +job file is used with Hadoop not MavenTo use Mahout in Maven just reference it as a dependency in your project pom fileTo create a +Resubmitting +For now Ill use the libs in my previous version but thats my temp solution +This is now fixed in Mojarra +As an enhancement can we add this to the Hibernate Settings in the Project Properties? It is not very intuitive for the users to find the Hibernate Perspective and click around to figure out how to update or test the connection +So I can say the preliminary version of the site would be ready to go online +Any more of the DEBUG level messages should go to TRACE +Joris Yes coding of measurement tool is simple +We dont need a new patch just a separate hadoop jar that works with the official hadoop release +Attached you find a patch for review adding two new tck test classes I havent added the two new test classes to the tck conf files +Defer all unresolved issues from to +Since that doesnt happen then here is the preferred attachment for you to use is the place to start. +didnt timeout in latest TRUNK buildIt didnt timeout when I ran test suite on Linux + I strongly disagree with this +Slight clarificationThe schema above has worked for me in the past but I am not sure if it has always worked that was too strong of a statement +The object being tested is env not ENV because we want to test what the shell thinks its ENV is +Youre right didnt look into the patch deeply enough +with administration capabilities and parallel builds in Im not sure if this is necessary any this issue amounts to making configurable for the webapp without a system property which I think is a good thing so Ill adjust the subject +PaulIve done rebase and will take a look at the examples thanks +The requestst patch +Please report an issue agains Maven Embedder Ive added some additional logging to the plugin and it should look better now +But it will cause build failurebecause there is no maven artifact for Aries tools when we build +bq +b +to trunk +I would love to have cp and classpath work but unfortunately I think this would actually be more confusing since it requires the that the regular java parameters does not +After applying the changes as described in the attachments our time went down to under a minute again +Finished r see the attached screenshot +Ive committed a fix into SVN trunkCould you give it a try to see if your problem has been solved +This is resolved as part of using new Layout class and +Is that not deemed enough? Perhaps the distribution directory needs some more +deferred to deferred to or depending on the VOTE which is currently runningYou still didnt proof that this is still used! All your arguments in this direction so far got trashed in the ground. +I hope Ill check one in two days +we just need to fix the javac and audit warnings +Youre actually right its only available at marshalling and thats not what you originally requested +On cursory look they seems useful but lets do that in a seprate jira +Information on these new exporter types is required +Also on a slightly related note the getting started page says you run plexus with the command bin +I was testing the fallback functionality which works fine but I just figured out that there is a related problem with the form component models not being updated correctly +OK so as a temporary solution I included patching the config file to the corresponding Hudson job its now in the queue +generated logLevel INFO style OBF war Ctempwebtarget localWorkers ve updated GWT issue with details on how to reproduce without +I think if is relative then we should probably use somethingrelative to the PWD of the JVMIf is relative then should it be relative to the PWDor to the index directory? For consistency with perhaps thisshould also be relative to PWDBut when both are null then the lock directory should be the index directory +Not reproducible in JBT move of M to have tried one more time and it isnt reproducible +Just because its public in Lucene that doesnt mean it should be public in SolrFor each little thing its arguably never that big a deal but a good policy keeps a bunch of little things from creeping +How should this workIf the SMSC which closed the connection the SMSC should decrease the number of active sessionsIf a firewall in between your service and the SMSC is dropping the connection chose a lower limit for the enquire to prevent thisI will close this issue because I think there is nothing we can do +It can return the substring up to the last occurrence of the underscore rather than the first +Committed in M and trunk +The API specification for getNamespace readsThe value returned by this method should be prefixed or appended to elements such as variables or function names to ensure they are unique in the context of the portal page +is not instance of Activity +We will come up with a patch sooner +Reassigning bug to Xalan +See the patch for +Colin Before I commit have you tested in other Linux distros as well +Sending auditAdding auditsrcmainjavaorgapacheservicemixnmrauditcommandsAdding auditsrcmainjavaorgapacheservicemixnmrauditcommandsAdding auditsrcmainjavaorgapacheservicemixnmrauditcommandsAdding auditsrcmainjavaorgapacheservicemixnmrauditcommandsAdding auditsrcmainjavaorgapacheservicemixnmrauditcommandsAdding auditsrcmainjavaorgapacheservicemixnmrauditcommandsSending auditsrcmainjavaorgapacheservicemixnmrauditjdbcAdding auditsrcmainresourcesAdding auditsrcmainresourcesspringAdding auditsrcmainresourcesspringAdding auditsrcmainresourcesorgAdding auditsrcmainresourcesorgapacheAdding auditsrcmainresourcesorgapacheservicemixAdding auditsrcmainresourcesorgapacheservicemixnmrAdding auditsrcmainresourcesorgapacheservicemixnmrauditAdding auditsrcmainresourcesorgapacheservicemixnmrauditcommandsAdding auditsrcmainresourcesorgapacheservicemixnmrauditcommandsAdding auditsrcmainresourcesorgapacheservicemixnmrauditcommandsAdding auditsrcmainresourcesorgapacheservicemixnmrauditcommandsAdding auditsrcmainresourcesorgapacheservicemixnmrauditcommandsTransmitting file dataCommitted revision +META +Somehow there is a race condition where the region is moved but the does not know about this immediately when the region is attempted to be moved again the does not know about it +Please take a lookThanksve submitted code that fixes this thanks for the report +DONE + +It should be on both places i think +I am getting it as well and do not see the ifx in the upcoming release +Makes sense especially since theres not a heckuva lot going on via anymore +For this iteration the zip files contain only the topics that changed as a result of the latest review +If that is really what you wanted you could have initialized the byte array to xFF xFE +I think them relate to the apr +Reopened to set the Fix Versions field to JDO beta. +this issue into the general SSL is broken JIRA +Running tests from +This by design? Or just for the current version? +Build fails on machines that did not already have the parent BOM in the local Maven repository +We should probably make a configuration option that completely disables the http browsing and hftp servlets in hdfs +what about using optimizer hints to fix the query plan in the desired shape and then check ifpushdown is happening? The test is about pushdown rather than hash vs scan choices +FixedPlease verify +This issue after extension property development + +Thanks +Perhaps similar technique can be used here +Will post a proposal soon +So this was resolved and waits for me to check right +Im going to attach a simple WAR that reproduces the issue with Majarro as well as + +If thats the case then my apologies its too late to fix it now +I have already completed the coding as part of the patch required by the customer +added for a few missing cases and also unified for and +Nice ve often wanted a Grails console for the currently running app just to try out some HQL or whatever without restarting +When iterating through the query results an instance was returned even if it should have been a after the first +bq +This applies to both EAP and EWP +tools jar one is not flexible the ext seem to be more now +That way we keep the learning curve and maintains low +Moving to CR since there is less than weeks until Beta +buildivylibHadoopcommongriddevbin +Open a new issue +I made my own modifications so that it seems to work pretty well I am at home right and my source is at my officeI will attach the source tomorrow +can I ask why the assignee change? I shouldnt work on this anymore +The tests should therefore live in testsuite +This patch provides only a temporary fix if gets committed later +Whats in the trunk now works but does not have CLI to encrypt +falseWill investigate other solutions ASAP +This patch appears to take out the printed messages saying whether or not the vectorization path is being used +Could we doing overrride in +The normal cascade could take care of that +Yes JSR adds few missing things but above logic remains. +So lets Lex make those icons +Sorry to be unclear +As requested closing. +Juergen +svn commit m minor typographical updates +firebug shows that some of the get requests just timeout when using ssl +the point is that the reason you use a column is to get behavior you have an append operation basicallyI am saying the whole concept of older versions doesnt really make sense when you are using it that way +patch it +The general approach seems like it should work but there are a few things missing +Thanks for pointing out the error in the javadoc +We dont have to state that now thoughIt doesnt necessarily have to be ve either +This is a known issue +Is there a way I can turn on some debugging or other diagnostics +I updated the as an example +Resolved for a while now +I looked over the results and it is due to testRecovery failing due to unreasonable length +A good side benefit of this approach is it will allow multiple versions of a services component to be installed and configured differently +If so then it happens because we always repeat the last message sent +bq +Fixed in branch. +I can remove the prepare build successful mail notification since it seems like youll get notification anyway during the build which also means the prepare build was successful +If it is resolved which version of axis it is resolvedThanksRahul +The one I submitted works only for D and even that only for lines through the origin and thats something I just used as a test here +Is this going to happen for AS GA? Im now having some issues with and wonder if some changes has been made +Geir how does this impact previous releases of Geronimo +This will raise dependency on and after upgrade a blame of all files will be done because detection of changed files is not the same +I will test GB files in +This is because all other lines happen to be commented out +Closing +Having this in one step would make things simpler and clearer +The only caution I have is that if the person is running off of the net the download might fail +The query works fine and the NPE is gone as well +Implemented using an array header containing popped state seems like lowest common denominator +Will this feature be available in? Thanks +I changed the priority to blocker +This is a duplicate of which has been fixed in trunk. +Once we know this works fine on a standalone case we should add a junit test to put it into our regression suite for this change +I have the audit ui up and running in my dev environment as soon as I get a chance to patch our producers I should be able to submit a couple of tweaks to this patch for and +moving this to as theres a workaround for this +but it doesnt workDo I have to change any componentI dont know if I have to change some componentswbrGeorg to work in as other table related endless loops +Map TRACE to lower than the logj TRACE so both of them work public static final int TRACEINT +Gary +Still researching this. +Then the WMS could also show something if no default was defined + +Are you ok to commit these files but no put them in seed or demo +Thanks Ryota +I hope I have some results at the beginning of the next weekSorryChristian +Hello YINGPatch is great but could you attach SVN diff as a patch therefore we are able to see what you change +L files are special could overlap with multiple stripes +Triaged for + +Correct me if am wrongStackDo we still need to prioritize the reference files so that compaction of ref files surely happen? So i think after this patch still the reference file of the great grand father may not get compacted and we may never allow the CJ to remove the entry from META? Correct me if am wrong Stack +Okay ive managed to work around it using the above patch but the default jackson way of resolution is to use the class name right? which then would be deserialized into af property that must exist? My specific use case is +Thanks for the quick reply btw It works if I set the threads culture to but not with failing function seems to be NielsI am closing this issueBut feel free to reopen it if you think it is something that should be handled byDIGY +take care of this problem in a more generic way +They will continue to work with this patch +Related tests pass +Thanks Jeffrey +The CMS API isnt going to tell them the specific quirks of a particular implementation +The testcase ispackage public class resolved issues. +Jostling around the release goals and numbers +It depends where you are and whether you are connecting to HTTP sites or HTTPS sites +like mvn package +The basis for an initial commit +synching or starting probably because of something is asked to rerender or something +Can you maybe describe in more details which problem you are experiencing +I just committed this +bq +Marking PA for Henry so runs +As such I wrote a runner that met my needs +Clean patch changing position delta encoding from vbyte to vshort +Introduces different KEYS and TESTKEYS in +The examples in the examples directory and the script for running those examples are included in the project distribution + is probably missing +No problem +In the same line set permissions to tasklog attempt directory to restrictive permissions +I learned ant an hour before the patch +Has anyone tested this on appservers +Along the same lines as doing things like disabling block cache or increasing scanner caching +I have filed this issue to have native support by Spring to overwrite mappings +With default value being false queries will still fail +Not following why this should be a bug on eclipse side we are the one maintaining this is not +I dont think its very likely that someone will step up to fix it on older branches so Im closing the issue as Wont Fix. +However there have been many important fixes in since +Even if Commons Configuration is not fully thread safe I think we can at least fix the most frequent failure points in a multi threaded environment +Added in the preinstall scripts to fix the issue for packages that are already installed +I will check my environment setup +I tried building the Geronimo M source thru Eclipse using Suns SDK and IBMs SDK +Reopening until we can tell what is going on or determine when this issue was resolved and close it +Thanks Giri and Matt +bq +This integration makes our life easier +We wont add Groovy to the default distribution so Ill just see what we can do to get rid of the ugly message +Fixed for only. +thanks Matteo and Flavio. +The only way I have found to reliably figure out what is missing is to attach a debugger with an exception breakpoint set. +I filed for the findbugs warnings turns out they are legit but I want to keep this as a refactor and commit the bug fix right afterwardsFor me I was getting an error that the unchecked suppression wasnt necessary where is there an unchecked cast warning +But because of some additional changes are required +We should consider adding a configuration option for this though +Created to track this +As Dave C and I just discovered for the umpteenth time the error message one gets when the inference gets confused like this on a constructor call can be quite baffling +Added first cut of the new async DSL replacing the old thread DSLtrunk cut of the async DSL +The patch checks whether the page was visited via the form or not +But setting it to FIXED was not your intention I guessCan someone review this patch immediately +done +Manually tested +Thanks Paulex +You can set environment property +A new patch + This bug happen on GTN in this case Click on from date field mini calendar appears Click on scroll bar and click on from date again mini calendar doesnt appears +Ping +Fixed JMX behaviour now server is visible in a JMX console when server is functionality to client side to create RO client one have to pass additional boolean param to s ctor ctors with old signatures create usual clientsNow if server is partitioned it doesnt allow usual clients to connectDecided not to create separate ticket for change since its anyway related to this one +Implemented default deserialization constructor generation in Java backend in r +Session attributes are cleared but they are immediately filled in again because of implicit model +Any caveats Andrew +Thanks +Runping can you give some information about the job +You spotted a couple of things that go caught in the patch that I didnt intend to check in +CommittedSending srcrefTransmitting file dataCommitted revision +It is simple enough +push to +I dont have this version any more because we decided to use the version or +So nice you say that +Not declaring this property should fallback to previous behavior +Thats a great newThanks again TimothyIll be there this weekend too +I am asked to undo this patch +Looks like they are all waiting on the cores lock +patch looks good +Under heavy load we got deadlocking in this cache +Patch committed +Now it only looks for about times each file before moving on so at best I would say it is partially fixed +Sorry for calling it rejectingInvocation its misleading here +Lets wait for Jenkins to bless the patch before I commit this +It seems like value is not passed to the bean at all +Hi KimThanks for mentioning it +Please open a new issue with fix version Beta for the bad wordings you find. +it would be great if the back variable wouldnt have to be set in the +I attached a patch that fixes the th millisecs was used instead of mircosecs. +Not sure how appfuse can help the tooling in being more flexible about project archetypes are worseequal to what we have today since it is all code generation instead of adjusting existing projects artifact style random maven or any other codegen will not result for a guaranteed working does not mean maven archetypes is bad just mean that as a foundation for a wizard doing a specific thing I do not belive it is a good enough foundation +Attached is a patch that backs out the part +Who should I poke for the subversion module creation +and advantage of having properties inside the project is that jbosstools eventually will be able to reflect upon them too +Should be verified in gatein master and reopened if necessary +Are you implying this should go into before trunk? The release is also important and having featuresimprovements in that arent in trunk is unacceptable IMO +When you persist in an order that would violate foreign key constraints attempts to insert null and then update the foreign key value in a separate statement +tested implementation couldnt find a way to make it any faster +The test cases handle the situation of where reverse DNS fails on loopback patch not sure why hudson is playing and resubmitting Im not sure why hudson is unhappy with the patch but there dont appear to be any author tags and looking at the console output there appear to be missing directories in the committed this +Also filling out the methods of Map causes the code dirty +Resolving as invalid per creators last Ive used already So close forget. +Transitioning to waiting for adjusted the script to automatically push latest changes to +Patch file +Committed patch as revision I hope there will be a patch that unwraps the into for these methodsThank you for the review and commit OysteinI will submit a patch for this issue +These files are changed according to HTTPlegal NOTICE files java files javajet files xsd file xml have partially applied the patch at revision +filed to fix fallback +I thought this worked already +The important thing is the issue against which the commits were made which in this case was +Thanks Sandy! +Hi NicolasThis would work on one shipGroup only while the current implementation works on all shipGroups +Marking as resolved. +That is you wouldnt be faking up the API so much as adding some fixed data into security to make it look more like a doc +Youre right Amber in that case only loads the +Are you using WAR or? +Phils original review comments still stand +Reviewing this patch I came across some more issues in related to deletes + tests tests for itemSeparator committed +The choice of datatype implementation is done a ddl create time based on size that user gives us and we deliver varchar standard behaviour +I prefer if the source simply says in comments THIS IS INTENTIONALLY WRONG per Max or points to a JIRA +I will try to look at the EMF code but since I dont have this pulled down it may take a while to get to it +Henry asked this question before in and a patch was committed before I had a chance to respond +Patch looks good +Would try to isolate the issue and see if a reproducible setup can be created +just a quick question tho are you able to obtain the desired behaviour when launching maven from command line? i bet you are not +Clearly they can retain a reference toit but it will create lots of questions over whether they stay validHaving the context object passed into the map and reduce methods makes itvery easy and clear +Also note that when you click on a child queue it doesnt properly show you the applications running in that queue +Please see we just need to marry up these conflicting requests +This is a Java backend issue +I have seen and absolutely believe that they should be accomplished together and that they compliment one another +So I dont think its resolved +Thanks Yip this looks good +Split into subtasks +Issue is resolved. +Attaching which cleans up the last set of Exporting type through public API warnings I see in the implsql subtreeThe patch reduces the visibility of the fields and methods to package or private if possible +Patch for just committed this +But locking could certainly be involved +Patch to include user as a property of committed to revision Thanks Nandana +JackThank you for the tip regarding the heap size +Therefore a returned Iterator will work as long as the transaction is still contrast to the above scenario will use one per operation when called outside a transaction +Also if we decide to change our defaults I think that we should include and possibly other things listed in the optimization guide in such a changeAny other opinions? +I am somewhat reluctant to have to do multiple lookups and am leaning towards making implementations require keys to have an explicit port number +Thank you Venkat for your contribution +you could have in a node but split across cores +If this affects the community version as well a jira must be opened and the problem fixed in Branch and HEAD too +Please know if there is anything I can do to progress this issue. +I just added a regression test case to reproduce this problem +Attaching submitted along with run instructions +Thanks +This would have broken the tests so they were renamed as +I believe this is now working correctly even on Windows. +Java ME support has been removed on trunk and I dont expect anyone is going to change the drivers handling on the older branches so Im closing this issue as wont fix. +Still linked to the closed project +In this senario broker B have dynamic IP address while broker A have fixed IP addressSo the duplex network connector are used to connenct from broker B to broker AI have tested with a different topology +This is purely within the context of a single plugin +I also see that the signature ofpom is incorrect +Thanks +Modified the code a little to only use the synchronization if in an XA tx possibly this could be if not in localTx such that this will allow the managed case to workwe need a test case that shows the managed usage from the tck if this change breaks the tck. +is now bug fix only so no new implementation will be added to that branch +If it works all right with recipientlist then I will close this issue. +Doc is now visible on the Derby site +Third try better remove the temp dests before decrementing the count on the connection pool in case that causes the underlying to close +I believe the readers can be cleaned up further but Im short on time and dont want to delay it for another week or two and then rebase it against updated trunk once again +Now marked it as EdisonIs this fixed? If so can you change status thankssudha +Edit toc +is a flaky test +Thanks Lukas +Such an overlay implies that the latter WARs classes function at runtime in the same classpath +into trunk and +Thanks for the patch but Im going to close as WONTFIX +Thanks +Dont forget to consider tests too depends on a lot of bundles from all other modules +I know enough to make it work but not enough about the details behind everything to want to take ownership of the new copypaste module that could have been implemented by extending an existing class and a single method +I added the to the list of default providers +From Langer and Kreft Standard C and Locales pg The relationship between global streams and their counterparts is undefined which means that you do not know what is going to happen if you use both in the same programSo it is more than just an weakness in one STL implementation and the source is more definitive without diving into the STL specpg says that you can mix C stdio functions on stdout and cout and the order will be preservedThat suggests to me that the better approach would be to bypass STL in this case and go directly to the C stdio methods which should preserve the ability to use either stdcout or stdwcout in the calling programs +theres still remaining issues with this according to Ricardo so we should actually look at the attached patch and try to apply it. +defer all X issues to +How does it make sense to use a GROUP BY if you are sorting by ID? Can you provide a bit more detail about what you want to do and what the resulting query should be +Emmanuel yes by server side hook I meant hash interceptorAgain this represents problem for us and we would like if this issue gets resolved. +Check for missing for use of primitive types in what is the state of the sub tasks have you started working on themIf you have not started with them then you can assign it to me I am planning to take this JIRA +Pete since you worked on maybe you can look into this too +Integrated to as wellThanks for the patch DanielThanks for the review Stack and Lars. +Merged to for at request +Worth making the output buffer size configurable vs to k while youre at itWith this patch it respects whatever is passed in from + sync the codes in with trunk +Finding the same file each time is not a problem now it will download it as a new revision +The use case is the code that intends to register a listener but gets the from another layer outside of its own control +ant Dtest issues resolved in released versions +Needs work to finish though +Ideally we don t wish to generate gsp views if we only make a small change to the filtering of a controllers list method perhaps to use a session object to filter according to a user permissions +Well the exception actually is from the in the rtcore If you turn on the validation tools do give you the error msg but by default we turn off the validation during wsdl definition build Theres couple rules opened by default during the wsdl service buildI guess to fix this issue the best option is turn on the validation by default and fix all the wrong wsdls in the cxf tests but its a big task +Also I think there might be a case where a mapping or permissions problem can occur where a suggested time will not work either +but it does appear to be loading you able to at least do jruby e require ffi? If you can do that its at least loading the native bits +project archives should not run at am not working with a project needing project archives to mylyn makes it possible to distinguish if a jboss project is active or not and thus project archives it would be an easy way to prevent jboss tools from doing unwanted activities +Revision close for release +This bug may be closed +Once we have that we can close this issue off +I think that is also perfectly reasonable +Maybe it is a directive valid only for the EAP +Then Ill see if I can break out the enum test into a single JSP that you can use independently +A quicksafe have a spurious whitespace change at the end of the patch and the patch does not have a unit test caseAdded unit test for the cases calling write method with null key andor valuea new patch with unit test code for the unit test Runping! I think it would be a much stronger test if you checked the results of what was put out on the streamOf courseAnd we need to have unit tests for many other aspects of +I do not know whether youve all seen Martins comments above and I am sure wehave already seen similar threads at this very place +Patch for CDI injections and refactor +It fixes the race of saveNamespace with setting safemode +Is there a date in mind for this release? If possible I would like to work on the next release of Common to test the maven release process +There are two pages for the Tuningguide and pages for the reference guideComments on any or all of the pages are most welcome +I think that would be good in tooWell have to release an +Attached patch fixes this problem +I plan to commit the code patch tomorrow and resolve the issue but I can reopen it if necessary +I just had the same problem yesterday with n in a double quoted string +A forked version would be sad indeed and would be more work for us to resolve what you guys did with what we need instead of just agreeing at the beginning +Please apply the attached patch for the new packaging structure +If this really isnt going to get fixed then I cant deploy even in a testing environment +I or Sorin will look into fixing the anyone wants to try Bills original timer macro just drop the +Its not clear to me if we need to continue persisting if you never have to define cores there +Closing duplicate. +move to next as known issue with workaround in the EAP release notes +Patch that passes regression +If Im right about this then we know this will rise up to bite us again. +GertAh of course +So this behavior is currently expected +Please verify and close +I think the resolution is simple We keep the servlet API around for standards compliance reasonse but we do not use it for any moreWDYT +tnx for input +Sorry meant to say +This sounds great +Is there a particular reason why you decided to split the unit tests into another issue? Adding unit tests to as one patch is better +Will try with Win XP and Win K tonight +Changed the code to copy between the camel Exchange and camel Message and +The same is used for other features like queue administration and I know of code throughout that hardcodes this same assumption for the lack of good API to deconstruct an ACL or modify an already constructed ACLAre you proposing we change the character separator to a? Any reason for this? Now +Right my bad for not closing it out. +Strangely enough restarting Eclipse seemed to make this duplicated toolbar disappear +Committed in jbossws revision and +It seems to be a race condition with writeBuffers and the allocation of pages +Im developing my own application over year in tapestry with extensive use of javascript and it drives me really mad to parts of my javascripts each time when tapestry updates from to +Sounds reasonable to me +We will be on the lookout for these instances while working on the guides +Have anyone seen this issue on release? Could we add some logging so that we will have some clue when it happens again +Done Ive created an which works similar to the +Patch application was voted down by has been some recent talk about it using but the parser doesnt make textSymbolizers and the renderer doesnt know anything about them +Patch applied to trunk Jevica! +is there any fix available now? +Please dont stop selling if you think Im wrong though +Kannan have you applied on your hadoop test cluster? Theres a patch on that JIRA thats compatible withI also recommend and which are in s release candidate but perhaps not in your version +For adding a second filter administrators could define a filter chain via filter mapping +Requests seem to be fine +With the patch applied I can see memory decreasing up to a certain limit where GC kicks in an the lower bound on free memory keeps stable +also the generated test suites are generated under qlsrctestorg +BI tools such as Birt Jasper or Talend may be the right thing to use +Jessica can you confirm this is OK +Please try this patchBTW some test cases for Vector are also added in this patchThanks a lot +a Ill remove the data dir changesb Because the byte buffer has to be allocated up front and has to be of fixed size +very same about my comment at the endAlso Im suspicious of the logic for handling parent markup containers in +The bundles vote has passed +I will leave this ticket open for a while so you can add details on the problem or close it if the problem is gone +Yes I think I understand the issue +Patch to solve the leaksIn httpsender it keeps a copy of the client in connectionmap in order to release it when svcclient is destroyedIn samples it releases the namespaces usedAlso it solves an mem leak when setting the same value to a propertySolves a mem leak when the server returns a error +Thanks for dealing with this guys +it for the meanwhile +bin logs from PM EST this afternoon + the patch looks good to meI dont think this should be considered an incompatible change since old clients will in fact never request a when data encryption isnt enabled +See the attached patch +Thanks Hitesh +Well for now the sites are separate +As a result we are closing this bugIf you can reproduce this bug against a currently maintained version of AS please feel free to reopen this bug against that versionThank you for reporting this bug and we are sorry it could not be fixed. +Also this causes multiple round trips to the serverAny other ways to address this +Support IP failover +Let me know and Ill make the changesCheersCheyenne +Some exception should have been thrown before that since this is NOT a registry issue but rather a issue +IT added. +Maven +Works for me +Its a bug of CDI tools model +You should really set in your +All right BobbyHere you as patch as before except with but the value of needs to be removed or commented out +in ea commit in master +It should only happen when the minor version changes +Solaris sparc OK +Ok I dont think this bug should be fixed that way otherwise Java serialization will be used for non Object messagesAlso Im wondering why JBAS performance tests are using? use serialization so you dont use them if you want the best performance +It helps +Using the library framework would be the New Maven Project then Ill need to look again but last time it seem to be hard to adjust with respect to adding additonal UI for specific seam questions +Can I provided more information? should I debug it? Can you point me to sourcecode where to debug +Committed revision +My main concern here was that with the movement of into a deploycluster dir that earsejb jars with names lexicographically before cluster would deploy before the needed clustering services +If theres a problem logged the user can decide to a fix it b avoid loading a module they wont use c ignore it. +if it works for you all +Changes in new patchNew column is not automatically added when a list row is selected +The existing error message is extremely vague +Its class response for property sheet view +Did we not already handle the case with the non deterministic function support that we had added to predicate pushdown +test added +That format looks fine +Thanks for reopening Jared +Simple patch +with the fix test passes on Linuxia with Jitrino built in debug mode rthanks +Verified with the build from latest ACS branch +I know nothing about Maven so help me out here +This feature is very important to us at Wanted Technologies +It seems like you want to implement a different strategy which should be done in a separate entity command +We should also clarify what happens if a scope type has members + +Thats an excellent suggestionIll log a separate issue for adding MTOM support +This is already committed to +I committed this +Thanks Heshan +Errors must be explicitely caught and logged in the dashboard ruby page +Perhaps walking through delete as an example +I think thats would had in mind Vladimir right +Thanks for the test code! I fixed the drivers. +So we would effectively be fixing that problem by updating theIm also concerned that the comments that were in the get stripped out +Rali will move the required classes from testsuite to test and will work with Rajesh to get a release of that project out ASAP +Quotas with append have zero test coverage +I definitely was not using the t option +Ran a import job importing rows in to regions on serversWith outAverage time taken by Reduce tasks mins secWithAverage time taken by Reduce tasks mins sec improvementSaves quite a bit time on imports with just a small set of server and spreads out the average load on the regions so no one server is getting hit hard while othersare set idleWould still like to see some stats form someone that has a larger server group then me +thats because jbpm and jbpm are not included in JBDS +I cant see why checking less often uses more memoryObviously something strange is happening +I committed it to +Move the fix to +So all issues not directly assigned to an AS release are being closed +Need to make some updates theyre coming in a sec +I had to back it out because it broke the main build +I am going to create another issue we can pick up later with a patch that contains the sorted source impls we have today +I tested this on on SUSE Linux Enterprise Server SP using cmake version patch +MattIsnt this a duplicate of and? I dont care what JIRA we take this issue forward with but lets link to any related issues and resolve any duplicates +Is there a way to avoid such issueWed need to preserve status and version for backwards compatibility but otherwise you could prepend a sigil to these administrative path roots something less likely to collide with a name a user might choose for a table +You can comment out the doxygen stuff in and see if that allows you to build +We are having this issue for long time and looking for a solution immediately +Be careful with your line endings +Ok so adding the valve is reasonable but I still dislike the API design +This issue shouldnt be too complicated +Lets see what hadoopqa says +Spent some time trying to manufacture the condition to no avail +Resolving this. + should fix this issue it did for us +Ok i just changed ! andfoo to work more naturally +should be quick to review +So this seems more like a blocker to me +Patch reviewed and Committed revision Thanks Shivantha +I reverted this change +The fact that it works for a particular implementation does not implies that it will for allAs the javadoc says the source should be a SAX or DOM one ill ensure that this is the case +Though I checked with Label Manager and no errors nor duplicates have been introduced which is noteworthy for a such large effort +Switching their order should take care of it +Not that I know of +We need to either do a ant or remove ivy cache before launching the build +Perhaps I can extract that from the context some other way +Not worth trying to add a testcase for this as it is about the log format of the pig launcher logs +Thanks Anton for improvement the patch was applied to TOOLS module at r +We should probably be looking at this for not take on the ticket +jars with the right license +Therefore there should be no name conflict as described in jakirans comment from +I have resolved that doing some changes to the above templates +Please take a look when you have a moment +You will need to rebuild the whole thing +Yes the problem still exists inI recompiled the code against the new assembly as well +Your class currently has no annotations + +The new patch looks fine +See for details of how to build +Taking the bug +Until JAMES was chaning the of messages in many many places and the was changed to almost every message relayed +Find the fix attached here +This issue wont be resolved until this Spring IDE will be used in of update to springide +Add featured attribute to widget object and persist this new attribute to the databaseAdd a check box on the widget detail page to allow admin users to designate widgets as featuredDisplay featured widgets first on the widget store page +Havent seen the tests stall so closing the issue +Good point Johann +Oh well while I was at it I committed it and also fixed anyways since I had done half of the work already before realizing the validator used was not actually the geotools one +I send you reformatted file in unix +It would be good if someone else more familiar with this code could review +Also need to create a fix on the backend because it will take some time before every user had a new app version +Join also should not be deduplicated +No problem at all AdrianThank you +or am I missing something here +Well for one users may have another version +A little bit strange if you consider that Ive imported it immediately +bulk close of all resolved issues prior to release. +this is better fixed wrong reference to plugin from send a much better version soon. +This patch applies to branchIt will apply to trunk also +Sorry for missing that one Myrna! ThanksDag +Some description of what the patch does would be helpful for review +I will run tests and commit +Couldnt this cause problems if a statement is after changing the isolation level +PR is s applied. +Tests pass I think this is good to commit +Great catchAnd an bug not just a test harness problem in the end too +could you please elaborate on url encoding state logic in portlets means? does that mean theres something to specify in liferay? what +is used by mercury as this component has limited dependencies on trunk +committed to trunk revision on. +readtimeout is set to secs and conntimeout is set to secs +I still think it belongs in the Repeat chapter just not necessarily as a sub section to +Karma can only be given to those with cwiki accts +Issue closed with release. +Classpath loading issue +reopencomment if you think this is incorrect +It got all the bells and whistles to be fast and leverages NIO and memory mapped IO and whatnotBut that can be done when we overhaul stream caching in once we get the time for that +Everything works fine except the problem mentioned above +no longer depends on to decode encoded streams +I could not reproduce this problem +Closing old issues +Will open additional tickets for additional work to be done. +sentence says +In the core framework exceptions can be handled in a more elegant way with writing code +To back up a bit and provide reason for the questionAndre has some integration tests for Deltacloud which he runs locally against an actual Deltacloud server +Hey Matteo there was a test script that executed the commands attached to the review +All tests should likely backport this to patch messed up the posInc fixed it in this your fix makes sense to me +will alter the meaning of the handler count and will change the way it is set +I am working on my proposal and am almost finished with the paper +I have already d +This is not resolved and will likely be reverted +But I have trouble reproducing this bug which is why I asked whether you could reproduce it +Fixed with testcase in r +If not the submitted needs more advice from us +patch againstIve attached the patch for the branchCheersEric +This is really really bad +Also if its neither then we should throw an exception so that invalid cannot be created +Saminda can you please take a look at how to make it work on? I am attaching the patch and i tried to make most of the changes but still not working +Unassigning so that a committer can pickup the patch for review +Hi folksI was shocked and quite happy to discover a thread that discusses the obscure issue Im having +It compiles but it is untested +The documentation was simply incorrect and made use of a bug that was fixed in version. +very nice collection reminds me of myna amr +It tries to respond with rather than +Delaying it to Beta and reduced priority since its not something affecting daily usagedevelopment of Tools +verified on CR +More recent versions of Xerces will work as will the versions built into any recent JDK. +Closing all issues + +Patch for attached +Yep its a bugThe jsvc was compiled with hard coded paths +Fix integrated in infinispaninfinispan +I wont port it further back than right now +I could understand that if the user specifically overrode the defaults to declare that as the desired publishing structure but as a default to only include a +BRANCH is still as it was before moving to Juno +I checked out and compiled the branch then replaced the trunk version that was in +Hopefully this is the final patch +Alan Im moving it to +Sub classes have to call the superclass constructor as per the KVM spec +Updated the patch to match trunk fix up some javadoc and tests +I fixed a issue with the same root cause as mentioned by Rajeshs comment +Maybe there is indeed something that should be done about it +This has various applications such as for diagnosing failures and visualizing behavior and using log statements that are already being currently generated gives us a nice way for doing this +mojo is the location of the actual plugin +It would also be nice to have some soak tests to test clusters but thats another issue +The current also has the asymmetric deserialization problemThat is is serialized to but it is deserialized from +The filtering of unpacked artifacts is available in the rest in updates +As for I could not find entry for Everyone group in directory +However I reduced the accuracy to because the tests failed the Longley data when using +Closing due to lack of response. +Fine +Thanks Crispin Ill take a look at these changes +Touches the following filesM buildsiteA buildsitereleasesD buildsitereleasesA buildsitereleasesM buildsitereleasesM srcdocumentationconfA srcdocumentationcontentxdocsreleasesD srcdocumentationcontentxdocsreleasesA srcdocumentationcontentxdocsreleasesM srcdocumentationcontentxdocsreleasesM srcdocumentationcontentxdocsCommitted at subversion revision +See. +now to start moving things in the repositorythis will have to wait until the release then we can remove the jars from jdbcjdbc for example. +resetting P open bugs to P pending further priority for bugs with a t this already fixed by by +The attached patch fixes this issue +ImprovedThanks for the quickstart! +One final comment would not be better named I thought adding a generic exception to indicate a directive was given to the service and it failed to execute it +Attached patch for this issue +Of course this should not prevent us from providing a better solution +Will commit in hours +Simple patch to add a deprecation message +bq +Yes storing complete distributions is very beneficial one can never know if complete project and not only library jars can become someones ltdependencygt +By the way I think the MB upload limit is per file so it might be possible to split the zip file in smaller pieces and upload each piece separately +Committed first cut of these changes in r +That way everyone gets a JIRA assigned to them which they can addressCan you do so since youre owning this baseline improvement and the codegenauditor stuff + +We will go with a schema based on the should try to use this schema if schema validation is available in the version of the xml schema having spent minutes ripping out obvious MC stuff from the old DTD should still be part of the dist for backwards compatibility +Please reopen this issue when soapr leaks fds with newer versions of +passes test +Thanks Ivan +Would you be so kind and attach the failing test? This would make my life a bit easier nailing down the problem +Originally we have handling the reference to the pojo and also reference counting +Will try to catch up with +David could you take a look +Even I feel the same +What is? Are you using custom preferences for eclipse? May be you specify in this file such property +So I would prefer to make it be anComments Currently int would be enough for counting namespace size +Looks like the trail is cold and its too likely that the index was simply corrupted in a manner that involves compass +update patch for trunk. +Ive also published the change but it may take up to an hour to be reflected in the website +Defer all unresolved issues from to + +for WDF improvements +Thanks Jerome. + +In this case we do nothing +Hey GuysLet me take a crack at this I should have some time today or tomorrowThanksCheers of r were pretty close +I didnt but from your description here it sounds like what I was hoping for +Thanks for your +Patch to you Dimitry! Your patch looks good and I have applied it to the trunk +Maybe returning an enumerator with partial match complete match no match +This particular issue may be closed +I assume youre using Springs defined in your? This should basically work then +Add a slew of tests add exceptions for too manyfew allowing parsing a list +No the sender will block until there is more space as I use a for this +Fix for is set to because its not Eclipse to M +this issue also occurs on rhel same problem exists on rpm distribution RHEL +Im downgrading from Blocker +rename offset to num +If theres enough demand for a or patch release with this andor other fixes then we can certainly do that +Sorry i must reopen the i have attached a valid schema with the described can you please test it +Also the way that row works in this patch seems like it might be broke +Attaching management server log for a deploy VM the async job response time calculated from the DB for K averaged for every the attached log I see that are getting created as part of deployVM +to commit this change to the build system! That was really needed +Looks like a good enhancement for +It also removes assertions +the UI for mappings is there along with hints on how to setup single file rundebug +Audit warning fix missing the Apache header in the new test file +Is there an agreement on whether connectionURL hook should look for default DB and create it in case it doesnt exist +Changes exception type to thanks +Without an interface making this method public would be odd since the View contract is the only public interface +And this time I am checking the Grant license to ASF option +The new version is attached to this issue now +match the standard HICC committed this +Fixed +could take a rather than a +I would look for correlation of when page alloc or page dealloc happens and when log records happen for the bad pageThe stack printed has the feel of maybe looking for a page that is past the end of file? Putting some debugging about the page table and the end of file at that stack might help +Applied a modified patch which just embedds the configuration handler from the latest config admin release in revision +This will distinguish between blocks that have not been reported yet and those that will never be reported +Verified that this issue has been resolved correctly +Note r is not updated I will rebase patch and update it feedback from r also updated read the code looking for stuff like multithreading issue +So I would move this issue to test will probably pass the next build so I dont think its necessary to disable it for +I agree that this is how it should work would be great to be able to enable disable use of the CCCAdjusting the type of this into an improvement. +Looks like this may have broken a test on see. +create scriptnode dependency should be hidden from view with automatic configuration of node script +I will take care of itWillem +to the next issues to next release +Hi Ian I just installed the rpms on a fresh machine to test again and guess what the admin console works fine + +comparator that doesnt actually match the dataI note that were passing null back in as well what changed +BrunoHmm that is strange I was sure it was all working at the time and I doubt anything has changed in the framework that would have caused it to stop working +I think WALA is a X kind of issue if you disagree please go ahead and change target to a. +Brian I think we should be able to address this as part of which I hope we will have for alpha +Added to resolve so we could probably address with the new dialect +Thanks Konstantin Second try with your comments +By default is set to true in the configuration +it works in if ajaxSinglefalse hmmm +subject to nothing being reported by jenkins +My last comment means that we have two config files and +We should probably add it to the C requirements there too +Patch checked in +When attempts to deploy my EAR file the main thread ends up calling start +No a regular is fine Java Strings are immutable and a put operation should just be a pointer swap +What if we wanted to build a similar scenario with Manifold or at least offer users to be able to search content updated within a minute? that would be lots of jobs as wellI am not saying remocing the current feature just adding another oneHope it helpsFran +The export wizard is intended only to work with projects of that type +Bug fixed by forcing timestamp to think it was already fixed in +document the usagewhat do you think? Maybe we should put logic of this patch into some common place extend it for all symbol resolving check +Here is the patch +TSCCM was subject to major refactoring since alpha Lets give it some time to settle before extending functionality +fixed in provides also a method to add new languages to the platform +for that Jarcec +Seems to work fine except that selecting and deselecting the Show runtime sites checkbox on preferences page does not update the table +Committed revision +This looks good + +Flintan I assume you have correct this +AshishThanks for your commentsWe are happy to answer questions that you asked Purpose of default map in forms It tells the form what default map is to be used +Applied patch to CVS +ews tomcat has been rebased on in CR build +marking as intended for +Yes youre right the patches are not independent and you need to apply first then this you want me to push everything in my github repository let me know +To ensure the object manager is initialised correctly we need to rely on DS in inject the Dbo service into the manager + +To avoid to loose users I renamed the full distribution to the normal name revision +Ill improve the by adding some gettersetter for the main attributes of a +I noticed that the blamelist is not reported is that normal +It is quite consistent +agreed modified patch to code to add support for standard JMS exceptions +Attaching patch which adds the configuration parameter to have an accept queue size of to along with a reference to this issue and a boilerplate warning +Lets ship it +FOO which version was committed +We both agreed that any metadata service built around the files should have an interface that their metadata service can easily connect to so that if a user wishes to use both they can do so without needing to register metadata in bothAs for documentation were working on getting ready for external release +Done. +Just like with Maven +Join a test project for this is not a bug and I moved the report of what youre experiencing in a log message at warning level +Passed tests and committed to trunk +And a needs to be added to the package +Also I think the line endings may be wrong in those diff files +Ive confirmed its inclusion in later versions of the Android API as well +Not a big issue thoughCheersAre T +I should also mention in the above numbers Im not doing real memory management +POINT +After the fix it will be successful +I will be DOING algumas but searches in the history +Philipp any findings on your side +Looks good +Honestly Jason you made my night I enjoyed our dialogue honestly +Ive also noticed a few problems with the stuff +Thanks +bq +groovy files assuming it doesnt break the existing integration tests or mess up some ide muck or something like that +push to +These are the PDF files usedThe file worksThe Andreas Is there anyway I can handle this by myself +which is expected +That doesnt help for CDI beans as you still getUnauthorized deserialization attempt assume a workaround would be to add this one and others to a custom but it is still a how should I know if or other classes that may be required are safe to be added to the whitelist +Committed to trunk with revision + +The patch looks fine thank youBest by Spark. +Im not sure even I am in favor of such an option though Just a thought +If you can find it we welcome a patch +Thats much better +Can you attach a sample applicationIm not sure I totally understand the +I firmly believe there is a way to do interfaces and abstracts in a proper way in Java +Verified at +New patch for trunk it was missing the poi upgrade +Closing as this should be verified with state gatein master +I guess I was waiting for additional approval and I guess I just got that Ill do a another mvn clean install Dall just to confirm that the changes still pass all unit test commit it. +Just want to sanity check the cleanup with everyone +Thanks for the patch Committed revision . +IMHO maven needs a way to specify the the jvm on the command line when invoking the tool +As pointed out by Guillaume the put method erroneously store the value using instead of Patch Guillaume for your patch Ive appled it in revision Please cross check and close this bug. +WDYT +Ive integrated the Envers support directly into the implementation of Hibernate but declared the dependency as optional +for the changes +If you create a vpc with vm then stop the vm for a bit the NICnetwork gets cleaned up +HiI hope we can get the code drop of the Zeta project running soonish +bat file +Very interesting +do the substitution inside then in the package target copy conf from to just committed this thanks Jerome +Ill make a list of svn trees and start down them one by one though Ill likely get multiple done in a single sitting +Not blocker +I backed out again +failed again +Reference can be used irrespective of access modifier of the fieldI will raise an issue against the SCA Java CI spec to ensure that this is made explicit in future versions of the specification +I have previous experience of the existing methods for deactivating and reactivating components and their references when implementing some domainnode features before Christmas +Bulk close for the release +Hey JunpingThat was my initial thought but test checks that should have distance and has distance rightAlso the previous test stably sorts to order as well +finally in the code where you change the classloader on the context +Removing the feature from the cellar one fixes the CPU consumption issue and a couple of other issues. +And then your patch would probably be handy. +Patch has been applied thanks! +I also figured out that Signature Encrypt works fine with both approachesNo one an ideaMarking this issue wont fix as per reasons outlined in the commentColm. +Sigh +FolksWould anyone really miss line number information in the debug output? Can we do away with altogether? The stream truncation can easily be moved toll happily take over from here all issues fixed previously after a brief review of each. +But a caution this will not trick Ibator into doing joins for you +Bulk close of old resolved issues +Vinod If you havent had setup to mirror then will still reflect svn rather than +By the way this isnt specific to view +was updated to include the Reference book + +It uses a class that extends to provide custom initializations for Configuration and +Code ReviewedOne issue that came up in the code review was the persistence of Network Topology in the case of namenode restart +Done +Patch to test. +Burr and me are unable to reproduce the issue against JBDS Mb +Could others who have encountered this issue please try this fix to see what results you are getting? Thanks +Hi this bug is already fixed in the trunk +looks good +Admin users must have the admin role in the tableThere is a utility called that you can use to add the admin roll but you will have to configure it by editing and possibly the script itselfIll be documenting this stuff for the upcoming Roller release. +Added check for the +Patch ran successfully on my test rig but this begs the question should this really take hours +On iPhone abroad right now so this will be brief +Do we need to create an adapter layer for older sequencers or should we just expect users to convert their custom sequencers +should be good for +Fixed on trunk revision +Is this not a dupe of +But because of a bug in this block size of is ignored by the namenode +Hi PeterI have tried the patch attached to HHH but the problem is still thereI dont know why but with joined subclass some queries have the formselectwhereorder by idthe problem is in the unqualified id reference in the order clause +for the fix otherwise +Thanks +Raising this to Critical since once a Topic or Queue has been created the clustered property can no longer be modified. +Please reopen ifthis is still a problem. +Is there a proposed workaround for this? This combination worked with previous releases of Roo! Thanks +Applied at revision +They both use String keys so thats why it could be relatively easy to retrieve something that belongs to a REST server from a Memcached client. +Since we are not working on can we close this one? I cant reproduce this issue at +Will commit if tests pass +Hi OliNo major objections +optimization comes from stuff like but in Java its the JVM JIT that does optimization +Ive tested different combinations on a couple of people +Willem have you tested CXF on SMX? Are there more work on this upgrade +Ill leave this issue to you! +What happened? It ran tests for minutes then gave up +Also merged the change into all resolved tickets from or older +But a I just downed loaded b the problem didnt show up on what passes for search and c I still cant buildMostly I care about c +It seems that the rDNS problem was because the host has a number of IP addresses and I assumed that the IP address was looked up against the name in the HELO command +This would really help investigating the problemThanks in subversion repository as of rThe problem was due to some missing permutations in the root matrix +Attachment has been added with description bug in already took a look inside I will try to fix itI also attached a small example showing the bugFurther help is always welcome +The classes I moved to core didnt seem to belong in the util module anyway since they are in the packages and not the package update files to refer to the new core jar fileThis patch modifies the following files listed in wsm netuisrcutil netuisrcpageflow netuisrc netuisrc netuiant netuiant netuisrcorgapachebeehivenetuitagstree netuisrcutilorgapachebeehivenetuicoreurls netuisrcscopingorgapachebeehivenetuipageflowscopinginternal wsmdrt netuisrcorgapachebeehivenetuitagsAdditionally you have to perform the following steps in svn move from wsmexternal to external move the package netuisrcutilorgapachebeehivenetuicore to netuisrccoreorgapachebeehivenetuicore after you have patched add netuisrccore add netuisrcutilorgapachebeehivenetuiutil +Looking at the most recent patch it appear to me that and could be independent of +Marking it volatile wouldnt cause any harm but I think we assume pretty much everywhere that the leases arent mucked with without holding the coarse grain lockBTW on an unrelated note I found that times out due to the changes imported by will upload a fix for that here soon +So what looks like a promising strategyOff the top I am thinking something as simple asstart with ORD with no fallback on the largestif the next segments are fairly large use ORDVALif the segments get somewhat smaller move to ORDDEMOddly Ive seen VAL perform well in certain situations so maybe it has its place but I dont know where yeteditOh yeah queue size should also play a roll in the switching +Please verify +If Im understanding correctly that means youd still need to do seeks to read the entire row except in the edge case of each update modifying every column +Spot entity you use nr while it was suggested to use number +Yes this is an often asked for request but I still havent found a good way of represent efficientlycorrect +patch looks good the only thing we need to fix in it is that the default TZ should be UTC instead GMT +I just committed +Okay +Committed as r +Im looking into itUpdate this isnt related to trie +Marking patch available so that runs +Not covered in chapter +The time taken is in proportion tothe size of the segment +This approach becomes non trivial and unlikely to survive long term upgrades +I hope following explaination will suffice all your queriesIntroduction Trainings in HR module includes Training Calendar where we can create new schedules from trainings available in an organization +Daniel moving this out to please provide a patch +pushed +I wasnt aware what the usual flow was +I feel like an archeologist by patch for hiveqa +yes if we ever decide to rename the Solver interfaces and classes too I ll open a separate issue. +I wonder if anyone would be willing to take a look at the change and validate I didnt muck anything up? +files for importing and exporting data from tablesExtra scripts used to create the jars and execute the jobs are attached for reference +I would also rename to its private so does not matter but wozuld be nice to haveOtherwise I agree with committing to trunk +No that does not need to be committed +Bulk closing for nd warning is still here +Input file Pig Script and Parameter somewhat related bug is JIRA the difficulty of debugging issues related to passed parametersIf Pig produced an output file of the code with parameter substitutions made we could have more rapidly isolated the problem +is not supported with release +By branching we can officially do a feature freeze so we can finalize our testing on it without forcing you and other contributors to sit on your patches until we deem otherwise +Sorry for lateness in replying Jrg +Uncovers a wrong assumption following lazy loading integration +Patch including implementation sample and documentation +Only AddRemove Blok Comment make sense to fix because Toggle comment action does nothing for web pages in stuctured editor for xhtmlhtmljsp and jspx previous versions of JBT Toggle comment action for +Committed to trunkThanks Jarek! +The sample code you have posted is too complex to work through to narrow down to a specific problem +Its the productized SOA version of it +Dominique can you verify this please +Thanks for getting started on +Moving out of +Closing this issue on Amys behalf +Commit attachedThere are one problem yaws working fine but crashing too many times i creating new issue +Very usefull feature thanks Tricia +for the ve just committed this! +This is pretty much done now isnt it +It breaksBut on the patch otherwise +I thought it would be possible to determine the keystore type by trying to load the keystore file with various keystore implementations available in the JVM +was doing some optimisation of the imports hence the inconsistencies +To follow additional unit patch with additional unit testing +Final patch. +Wed love to have it but my intuition is flashing complexity when I consider it +If a real network connection is required then this problem still exists +svn revision now contains the NIO based code which I think should mark the files as binary for you +JochenI downloaded todays snapshot and this test still fails for me +Looks good! +Closing. +Ill add to that patch the pom for replication +If theres still something we can improve reopen with a clean reproduction. +The timing issue aou mention under is the motivation for this is no longer an issue as the latest patch uses instead of +Also made the string static final +Updated patch Changes the configuration field to userPayload and as a Includes proto changes and serializationThis isnt being used yet +This has not always been clear in the documentation +Marius means bean archive scoped service +this is +A quick fix +But there doesnt seem to be a way to alter the actual URI on the Link instance passed to the Listner +I agree the wait is arbitrary hopefully at tleast if it still breaks the new errors will make it more obvious what is going onMaybe Rick could comment if there is a better approach for the test or if the above interpretation of intent of the test is right +I meant is fixed in +Ive been looking into your tests trying to figure out how it was possible for them to succeed indeed a subtle difference + is a proposed patch for review +At the end the view pool gives an small improvement in cpu when the page is rendered but it reduce the amount of created objects +Agreed I was implying that if were to be rejected for some reason wed need to have fix +replication may not be the best term but is used by the DBA community for these setups for or other RDBMS +shows an example +Attaching patch and regression test +But one test case test of is failingCan you please help me to clear this testcase? Instead of this testcase everything is working finePlease find the patch and let me know your feedbackOverall the patch looks good but I cant apply it to either branch or master +JacopoScott if you have time I will be really glad if you could prepare the patch for the license change of the entity component. +That exmaple is the origin code the is customized you can try run this according the you are right there is need to be improved I will try to fix this +Thanks Harish! +Yeah well that was a pretty easy one +marking this as wont fix since we agreed that the one to one correspondence between a zookeeper client and session keeps things simple for the users. +You need to use instead of if you want to get the full URL in a transparent deploymentOtherwise youll get the URL sent by the client which wont have the hostname +Thanks David for reviewing +Showing that last seen date somewhere in the UI probably makes sense +If not ill have to go back and check. +Mars HsuThis ticket was created for the browser thrift +Thanks for the patch Ivan. +Feel free to provide more translations. +It is unlikely that an error message referencing the xml file will be provided certainly not in AS +Nowadays on NM resync with RM all containers including AM is killed before NM can reregister with RM +and not EL stuff +One more reason to take the aproach is that the ivyconf file can be stored on a webserver +Can you review and make sure things look goodOne thing in particular to look at is the guava dependency I had to add to make things work +Committed for in rev +So had to build the bit binary ourselves +The shell currently supports the followingecho flush w t bloom accumulo shell u user p passThe example could be changed to this +This one fails +IgorExactly the same problem is reproduced with hdataTable +the patch looks good +Scheduled as new issue +I think this is working ok in and trunk. +Candidate patch v +If you want me to take a look on this again during I can look at it +Hi SahooCould be similar if you have a concrete testcase please attach it to the issue so Peter has something to work fromFYI the update of BND was postponed to the release of the bundleplugin so is still using the old versionCheers case that shows that uses upto minor component of the version string +Since they dont behave as such they really shouldnt be considered instances of Module +Nothing yet but sometimes googlegroups can be slow +We will have to modify the antlr version though once we move to hive +thinking about the feature to not start accepting connections until +Fixed in resolved issues. +If no objection Ill backport it +it hasnt been released yet no. +So I will say that the optimization could be thought as requiredThe fast mode should be used in the generated text by the which is an integer +This patch fixes the have applied the patch locally successfully here I have to do some integration tests firstbefor checking it in +Nitay committing. +Closing because this was a jvm issue +for restoted component have nodeId property but in RI case its nullBut I dont undestand full working logik for component +Else a cached file or other static file with a +Please apply the patch from the xdocs patch is for Axis folder which is not in the code base anymore +Marcel let me know if you want me to submit this +There is not really much to say that is specific to AS +I know you previously closed this issue with the following comment I dont see any compelling reason why SQL Maps should arbitrarily insert whitespace that wasnot intended +Oh it was Alexandr Shulman who noticed my error. +I will add the checks curious how this is happening though +Ill try to prepare some graphs of node metrics withwithout the patches +You might want to run mvn with X option to see where each version of got used +bq +Your patch is in Alby +note I can see in the spec that reference is required +Ok about update version but its using JDBC driver with Java and from If you are using the or JVM then you should use the JDBC version +This entails a change from get so that it gets s instead of the actual thing and then puts in a new get to fill in the actual object +Thoughts? for this strategy even if it manes having more jar files to manage +Thats how we generally do it +Can you confirm it works for your? Thanks +In that vein I should say that Im running on sun hotspot +What if we just made the compaction manager queue a priority queue instead of FIFOYeah thats what Im trying to do its just that with the current behavior we cant determine the priority when inserting the tasks into the queue because the sstables arent picked until tasks are removed from the queue +Couple of main issues The ivy jar isnt part of right now and so the ivy classes cannot be found to satisfy the Grab +Please include an example that matches your description +A few comments you note that HS should only support true here +thanks +The osgi converter would convert the url into native jar or file urls +Ive merged this to. +Patch has been applied thanks +I think this is an issue for the highlighter mostly which can score in the document so as to return the best snippet +All MR apis will not throw anymore they will wrap and throw out the test case patch based on the latest patch looks good to me except for a in which isnt bad +Let me know if I should open another ticket +Yes that sounds reasonable and is one way of solving problem IMHO +For instance Your file is malformed +Internally we are working on this one at the moment +bq +I took a pass at the changes +Still worth investigating though +We just need to write tests +Bulk close for Solr +Working on the remainder +Use +Debian Ubuntu the RHEL variants and openSUSE SLES +fixed in last fixversion +Ive been able to do this regularly by doing a bad upgarde procedure +Patch incorporating review comments +Craig Im using signin with facebook +Patch which includes the and Equals +My worry is that this would translate to a bunch more syscalls no +Heres a very simple patch that fixes this issueAdded with a default value arg and used this in +Does Teiid provide easy access to these relevant properties through the API +Its too late to get this in +Updated I just committed this patch. +Would it be a problem to downgrade this to? If not I can provide another patch +Please find patch attached for issue +Kevin please take a look at this one +is meant for cases where an intervening save makes the underlying content incompatible with the changes that the current session is trying to save +Yes you are right +This release is replaced by. +Makes perfectly sense +If any node exists the version will start from +It was released with the fix in +Modified files svn statusM srcref +Any more information about this problem? If not we should just close the issue and reopen if we discover a problem again +Its Mauros current working task. +Nice catch Henry! I was looking at the class and couldnt understand why that class wasnt using the Configuration object no more +Yes Aster is the one I had looked at too +Ill see other tickets as well +The request latency metrics are part of the request object +Backwards compatiple patch to allow salted versions of current password algorithms additional algorithm BcryptUses Jasypt for salted versions and Spring for BcryptJasypt also has the possibility to create passwords but this was not implementedAdmin passwords can also be hashed with algorithm of choice +I cannot manage to get plaintext formatting to work +Thanks +comments on you refresh +Fixed in master at ff +Because the reasons for using our own aprpool are no longer there in the current codebase +For example the files analyzed in the last days are probably not the same at all than the files updated btween and days ago. +Since WODEN issue is resolved it is possible build this branch successfully +I was confused because the FASTFAILMAP container was still being assigned +Have you tried uninstalling the plugins first before specify the new path to them with close this issue +clean up a few there any remaining issues I need to address +formatting and redundant import +how can manage it +attaching java file instead ofI put a file up for you to test +Is that OK? Is there an automatic fallback to normal connection when the proxy isnt working +This patch Adds page title and section title to in +FYI the code from Erica referenced above is using predicates to against an array of a class that she has created from the data +For example from JMX console +I cannot reprodice this in my environment +where the types attribute of a protocol is always deemed to come before the messages attributeWorks for me +Reschedule +The solution is not suitable for us +Would you like to expand a bit on that +If is intended to be made a public interface at some point I agree with you +This issue should be fixed after fixedShweta could you verify it +Spelling is my forte +Patch looks goodDo you mind submitting for QA run again +Sorry I just glanced at it and did not realise this had already been have a pull request to drop the dead code. +Can confirm this works +Which JDK are you using? It works for me as it is with Sun JDK and the message is slightly more friendly because it gives you the bean name as well +Attaching a patch to align this version as well +This test case is very similar to test +Wait I see +So all issues not directly assigned to an AS release are being closed +backported the fix to with revision . +Here is the patch to fix the resolver so that it uses the local and doesnt try to connect to the internetPlease test and commit to trunk +A fix to some issues with the script +Updated documentation and created schema +I dont get the nextString methods though whats after MORE? and whats with the two parameter version +Actually as it turns out this patch applies cleanly to branch and the tests passI think this is safe to commit +It is good feature but it is not even in workorder for JBT and JBDS mean move it to rightunfortunately I agree with you its a very big chunk of functionallity to introduce in in the the seam wizards and code completions would be better spent at the current time instead of rushing this feature + +Changed priority because workaround exists +The type attribute was in and was deprecated in +Ive seen this come up on the mailing list pretty often a clear message pointing them to the config and the current setting would be helpful +What do you want to do about thisWe can do commit access one of two wayshttps or svnsshAt present we cant mix and match due to operating system choice +Nicholas I had sent you Rob and Daryn an email letting you knowSure I got your emails +One to many clicks +for the patchThanks Ivan! I tested the patch on Mac and Windows +Fantastic! Phrase queries have often been a bottleneck +Id be happy too. +I just looked a little bit at the trails code links here but Im not certain where the ASO gets stored. +If youd like to dig in and see if you can find anything new wed appreciate the help +I dont remember if LDIFDE extract a version number out of AD so in this case its really important to accept a ldif file without any version number it will allow a M compatibilityThis entry could be changed from Bug to Improvment +Patch applied +Hum the encoding is performed inside so Im not sure how to pass the decimal positions there is more GML uses double not decimal for coordinate encoding so Im not too sure the decimal position setting is still relevant +However setting the resolution to done for this issue is incorrect I believe +So putting the classes there would be no good idea +Mostly as a sanity check when reading and writing through the different APIs +Cause I dont have the permission to reopen an issue the same report from another user on IRC +Thanks Jiannan +After that itll be up to our project managers to decide whether to pursue or defer this task +Looks like a bug in how we handle colormapped imagery in the +same bug on a linux box +A test case +Please provide an update on your issue as a comment +They will switch from using to the new rolling logs during the soft upgrade and then back to original May be someone else in the community might have more thoughts on this +Will check it out of course +Some of those issues will be addressed for +So no doubt it will start bugging me soon and Ill implement it in my down time sometime before the new yearI do think our current use of codahale is very bad at reporting latency spikes which is more of a problem for our project than for others given how we advertise real time characteristics. +Close this issue as Not A Problem +Done or fix later or not an issue now +According to Clover the condition line isnever evaluated to false that means the line is never run and openElementsdoesnt return null +So Ill close this issue since its duplicating commentreopen if I misunderstood +I will close this jiraThank youMirek +No hackeryChatting today the character came up again +Tested locally passes +Fixed on trunk revision +AckI dont know what our next step is + has a pull request but remains to be seen if it will be to +Release Note Text must be added to this issue by Thursday th September otherwise it will be excluded from the EAP Release Notes +Closing this issue as duplicate since is committed +This fix was not in the version that will be released with EAP +If one wants to add complexity to the projects the burden of proof lies with youIt works both ways isnt it? There is original proposals and counter one w symlinks +Shoot +Since could lead to generics removal +I have not had the chance to look at it recently but when I looked at it a little while ago it was failing due to lack of heapspace issue +Please test. +Should pass now. +Thanks for the review Stack. +Attached is a patch I might ask for review at Apache CXF and then commit +Max because conversation never amounts to resolution wo a JIRA and an assignee +FYI I just changed the environment from jetty to tomcat bc tomcat is potentially more often used than jetty +The client just need much less code than the serverwebservice example +bulk defer of open issues to +Components should be Spring Security instead of Samples +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsermodel +Closed due to the release of Tiles. +Verified by Leo. +Ive modified following two things +Patched versions will be rolled out in next release +Verified in JBDS. +And by design they can be ignored +Right I can see what youre doing now +partial little of the work for this was started by +If you want it to be skinned then put it in the xdocs directory +Sorry Ill get it straightened out either this evening or tomorrow morning and upload a single patch +Its still the as you can see location to the is generated during building the framework and not during using archetype +fo object doesnt have either nor propertyThese properties apply to all formatting objects and they are notinherited therefore there is no much sense to define them on a table rows +Will be implementing the same once get an approval for socialdata +Jared any news on this one + to trunk. +HOSTID is a full app state after now so we can revisit this +Would be great if we could get it into +JIRA stuff is all setup for +Also the provided schema is from an OASIS standard so I cant touch it to make ADB happy since it will affect the interoperabilityAnyway the resolution is wont fix so I guess I will have to live with my current implementation until Axiss JAXB integration is mature enough +This is a dupe anyway the issue is fixed and it was only logging not failing. +HiThis issue is marked as fix +bq +Just copypaste exceptions and whatnot +The When called section could be updated but the Requirements say destroy should release resources to get ready for another initialize which implies that it should undo the effect of any previous initialize even if partial +Update the description +Are you running this with an old version of on the classpath or something? Or is the test not ? The runner is supposed to catch and treat it as a skipped test not a failed test +Thanks Slava +Patchfile which filters the restlet headers +Ok but I thought it was a good idea as far as its just a matter of changing a couple of lines of code and allow users to set the propertyfactory from resourcesettings +Are there actually any LDAP services out there that have a default schema that does this? Surely the more users you add to the directory the less scalable this becomes what would the advantages be +As long as theres a reasonable hook Im not particularly attached to a specific implementation +The biggest issue is that apps often introduce new html rather than rehashing the available +Does anyone know what files invoke this by any chance? Ill continue pokingsearching around +Sample Roo script to reproduce the project with errors and the output of the mvn gwtrun command +I would look for a large in the heap dump +Im going to add you to our contributors list in JIRA which allows you more control over tickets +Ive been testing it by hand on a real project meaning Ive been running mvn compile in there and checking if it does relink if nothing changed or not and if it does it if smth changed + +I looked at for a call to setParallelism and createTasks but I dont seem to find where parallelism is being set as claimed herePartial work for Its not there yet just added the comment +zookeeper or others +It makes the more difficult to understand as well complicating the implementation. +scheduling for its too close to the due date of RC while I wait for have layers which takes seconds for each request +This I made be using sonar infrastructure outside of theWhen I hopefully have a breaktrough Id like to present you my work to motivate you and the other sonar folks to help me get it doneWould you agree +bq +Described CSS Preview in CSS Editing Perspective chapter in Doc . +Field copying with ctor is restrictied to basic types so that isnt a general could try some heuristic to mach property types to constructor arguments +There are several writable rpc tests as well along similar lines +Thanks +Hence reopening the defect and attaching the logs +We can add a hook in wssj to set the classloader to be used for loading the classes +That should be removed as done in the are right I will apply your patch to remove UPPER from Stefan +next patch follows asap +Patch for added. +a test case +Additional tests and changes to and +I miss to say that this problem occour in server side problem is fixed with +Andre Thats it + Fix mainteners Fix some for upstream project Remove some mentions to Cloudera +b Pig logs a warning message and continue assuming user wants literal and not the substitutionAdvantage for b is there will not be a need of escaping +For others implementing this work around Ill list some things to think The POM published with your artifacts isnt being truthful with its dependencies +I agree with you BJI also expected Quick Lookup and Lookup screens to filter products by store +Set it to loose and you got thesame result +Hi BartYoure right Ill leave this open as a feature requestJon +Youve been hanging out with Eli too muchAlso Id say we should just log it at DEBUG level for the Not checking case +The client just checks its local prefetch buffer and if a message is there its returned if not it returns null +Well I would consider this to be important +. +This patch modifies the column pruner to create a special case for the lateral view join operator +So I change resetValue method to set to true +Since the mapping is set to an absolute path I guess it should indicate whether the path is correct in the server admin +The only licenses that were in the old build but are no longer downloaded are the now uses the apache license so it should be covered by I think these dependencies are not used anymore so we shouldnt need their +Thank you for the report +This was my mistake +Upgraded priority to critical +Great thank you! +According to DTD for message is required by record so this is RIs bug +Adding a patch to see how stable trunk is right now +Give it a shot +Hi ClausI have updated the patch with the changes you suggested +Makes sense +Make sure you have RF in the project class path +Committed revision . +Perhaps it would be best to just audit at the Thrift interface level +xn so if the application requests such sequences in order it will receive entries in order +Tested on JBDS me JBT maven plugins from the project imports without errors and run on tomcat out of the +Was solved by adding the recent button. +How are you generating your patches? I have never had trouble with anyone elses but I consistently have to use for yoursHere though even didnt workfatal bad inconsistent new filename on line +added javadoc and committed +Ideally this would be an option in the upstream project +Wha? How do I vote against something +BTW does setFollow even work? I cant find any documentation on it and I cant see how it switches from STDERR or STDOUT to the other without royally screwing things up +Id like to address that issue in a separate JIRA after this patch gets committed if possibleAlso I have removed the interop tests for now +Its impossible to see the differences from the patch itself +I want a consistent layout between tarballs andI dont want to include the platform name simply because then all of the scripts need to consistently figure out the platform name +I saw these errors on trunk. +Thanks Sergey +I didnt put it there +You have a point there +I would like to reproduce the pb +This should work fine with fresh compilation of your expect that a recompile of your JSP will help +A total of tests and they all passed +Im wondering what should be done when the referenced attribute doesnt exist in the entry? I think that to keep it simple for now we should just ignore the reference +Steps to reproduce +Docs updated +Patch file for +John I obviously failed to explain this well to you as this is what I was getting at when we were chatting Sorry about that +Ideally it would be like a hash of some kind but I just took the request number and appended a dash and the current timestamp +It should be quite clean now +Bryan would you have the possibility to provide a demonstrating the regression failure you mentioned or provide maybe a snippet of your routing logic so that we can reproduce this regression failure on the trunk +The build runs just fine for me ouf of the boxThe module folders are in in the SVN and do currently not reflect the artifact names +bq +Thus this is an edge case +Could you please add a sample project for testing +I just need to test it out on some examples +should we just provide something like that for the thats generated for clients +If you need more immediate response please use thefollowing contactsFor Sales and Contract related issues please contact Mike OlsonFor Support and Training please contact support ortrainingIf it is urgent please send sms to my mobile get hadoop hadooponline training blog blogtwitter cloudera +For updates and deletes the sequence ids introduce a small overhead in the since I have to add a long value to each item +Assigning to Jesper +Verified in +For running the project as a plugin it is also required to execute mvn clean install as the jars folder and Manifest files are required which are generated by maven +Just to add to Martijns comments Im a novice and have never worked with Wicket before +Attached +In the future we can introduce a if use cases for this arise +I applied the intest of the patch test and all seems good. +I just committed this. +Can you elaborate +the patch to include the documentation for the extension point and the new interface classes +Patch adding glob support to the parameterThis patch also contains the fix for +Patch applied at r please verify +Its probably not worth putting too much work into our parser at the moment +I also did a full test suite that also shows how to use the +This has been fixed I dont know when but views are visible among the other tables in Geoserver. +Proposed SOA TP Release Notedocs has been removed from +In fact its already there as part of the +Also if you happen to have debugging on for some reason you will see much more numeric output as well as well formed debugging on the bottom of the template +jaxws builds now with both m and mthanksdims +This is not a complete solution as it only exposes what I explicitly needed +There are two transformers one for and the other one operating on a generic so working for JPA and any other Instrumentation agent based transformerWhile this solution is overkill for this specific bug it was the last part of the maven plugin still waiting to be properly rewritten and also rewriting this part was needed to try to optimize the performance of the LTW system to have more productive developersNow there are hooks to introduce enhanced +If there is a problem please open a new jira and link this one with that. +Isnt that required anymore +Works well Linda Ive just updated the issue message in the specification +Without this the original intension is useless sinceeclipseclean cleans up customclean patch is included with applied thanks. +I guess the real question is why doesnt the test work if rewritten like thisBug inBug in properties loading functionality itself +I need to upgrade jboss AS finals infinispan toSo I have to upgrade ispn jar to jar because of the old jar contains wrong xsdWhere do I downlaod +Thanks for the patch Alexey. +Is there a way to ignore this object when flushing the sessionThanks closing stale resolved issues +Patch applied +done but need the mahout repo folder gw so I can move it to archived and set a read only folder in its place on folder has been archived. +There is a huge need for this +This makes sense Timo +This also gave me wrong resultSo some setting to create jar archive is suspicious +And blushing due to forget uploading my manual test output lead to waste Jons extra time +Thanks Tony +Therefore Ive dropped the rest component +lgtm committed +Now it works +Yes this is a limitation of the features in karaf and we havent worked on a yet +In my opinion thats not the way to goRe +Its the way its coupled with sasl challengeresponse itselfbq +All classes are in the same in classes and an empty as with annotated +Heres a patch that generates the POM file from Ivy +I guess we should try to the fix at least to +Applied revised version of fix to trunk under revision r. +Its really interesting the effect of peeling back those abstractions +This has already been fixed in Axis trunk and hence will be available in Axis +Anyway I manually editted the patch removing the erroroneous +I replayed the scenario in and could not reproduce the issue found in. +The main issue would be that the script is dynamically generated to include initilization statements for each service and reference +just a bump to review and see if there are other jiras that fix this +I do know that atthe time I did the analysis there were something like instances of the DTMcreated during a single transformation +The code doesnt seem to be too frightening +Will try on the latest when I get a breather and report back that said I tried this on a recent nightly build and still got the error. +Ive taken a run at this to document the failure procedure while keeping the internal implementation details out of view +Blake could be possible to the the task logs for the mr action launcher job from hadoop? +Bumped to +Also make sure to read Sylvains comment. +Punting until other bugs are resolved. +No test ran and the build failed but its got overall +The dictionaries are local to only the row group so that row groups can be processed independently of each other +Reviewed by Darryl +I like the idea of NOW as seqidflushid +Thanks TomYour patch is in trunk at revision You should not have included the changes since you already included them in the button fix +When I apply the patch in my environment the feature seems to be working fineThe two patches attached to this issue provide two alternate syntax possibilities +If an administrator wants to use a different value for the initial permission heshe can set it in +Based on some changes in RS it gets updated +I removed the o +Im actively working on this +If you have a set of zookeeper endpoints that changes on a somewhat frequent basis it is entirely possible that a very old instance of a client is sitting out there with a list of servers that only include one valid endpoint +Please close +The search never has been very good though. +I will add factory methods and commit them +This is a problem from JSF responseWriter +Right now the links are broken +th patch Fixed Javadocs errors +Thanks Claus +And THANK YOU for raising these issues formally so that we can make sure they get addressed +nd patch committed to goraamazon branch patch Committed revision in goraamazon new version of the patch including tests upgrades +Will try to do it +Thanks for your answer +patch applied just doing the mirror image change to the getter body method +Sorry the above issue is in sonar please update Prem could you please try to reproduce the issue with SonarThanks I have tried the same project with sonar it worked but we have some developed for our internal purposes for sonar and we want to use those to the let me know the exact problem so as to justify the error or issue so that we can proceed further +can be solved by +But I dont consider that to be a good solution because I think the appended parameters and the included flag should not be separated in two different classesI will manage to fix the class to resolve that problem in a more beautiful way when I have timeRegardsZHENG agree with the direction were heading but need to have a clean patch before committing +Committed +The problem has been fixed using the option of weblogic +Please verify +It was decided that we wont fix this in the compiler but instead change the language to disallow shadowing see +Not for right now EAP and SOA are the only products that use an installer like this currently. +Patch with necessary changes with the new dir layout +AFAIK if peoples jobs generate so much data that we cant even contain it on the disk then our only hope is to find a way to better parallelize the problem +Use a patch in the. +Maybe a in +Cloning may be controlled explicitly byreplacing entries in a cloned params object or theparams classes can be derived to clone specific entries +Im going to commit this as is as theres one and no +Karel I have refactored tobut doesnt agree with renaming of to original object which handles is called so it looks consistent form me +ThanksThomas +we discussed using for this +See attached file +Looks like a modified version of addDocument will work that operates on and Documents manufactured from the +Changes to secret key management cannot impact them +Thanks Joe +However foreach doesnt belong the the core tasks of Ant and requires additional lib +The other way is to have setter methods for pk attributes but throw an exception if the method is called and the attribute value is not null +Good patch +Fixed + +Just committed fixes for and removing for jee builds +I created a user library in Eclipse for all of the logj jars source code and javadocs +assigning away since its not on my near future todo that it initially feels dangerous +Running the test locally also seems to consume less memory with the patch +If we are changing the node structure on ZK via this patch will be affected no +I attached to this JIRA entry the files to fix and +Estimated effort on the assumption the problem still exists. +Ill get Bill Burke to close this one +Committed thanks for the review Stack. +ah sorry I understand now its being killed after the first tests because of timeouts +Committed rev cddfeafbbceffdc +Sorry Uwe college is taking my time againHere are the latest changes for trunk and xTests are passing for both +Committed trunk revision branch revision I will keep this open for possible backport to move I hit close by accident +I think wed love to fix this but we have no way to reproduce it right now +So if you want to have a look at it please do soThe issue itself looks a bit odd Obviously the class loader behavior is different +its not working at all +I like the idea of controlling the caching at the translator level +No new tests needed since the changes are simpleI will wait for a day before committing this +Rest are all IO or BSP framework bugsIm close this issue and will upload the benchmark results to wiki +I might help and would be interested to follow the Apache Commons also though I have not much time at the moment +Setting component +Patch looks good Fay go for it +is probably more interesting than schema +Where can I download the whole adaptive scheduler +Skimmed patch +portlettests +Isnt this generally necessary? Or to put it differently when would validation occur otherwise +Committed revision Thanks Jeff! +Just to make sure Im using Spring +Its working for me as well +I tried my test case again after clearing out some meta information Pulse keeps on update sites and everything is working fine +Sorry that is Eclipse on Windows doing its thing +I have committed a fix for this +added javadoc for +richfacesrichfacesarchivesasseamstrutsbirtportletruntime +didnt mean to surpriseHere is more background +That will come later but this issue is now yes you should definately open up a new bug when you find one! I was just letting you know that I was working on rewriting some of the location +Thanks gives me info I gives me info I we close this then? And I guess +From experience I suspect getting the code for IP based automated failover working in a approach to where the community is ready to see it as a stable codebaserelease will probably take some more investmentWhen we are there it would be good to have both the attached design documents and the test plans and test cases updated to reflect the use of for IP based automated HA failover +once finished Ill submit the patch +This is kind of a serious issue +havent changed a thing Ill go have a look at the setup +This seems to be resolved via the expression language introduced in M +where are these files syncing from? +In this case we normally add it locally into a lib directory and remove it later again if the official version is available + you please tell us your reasons +committed to trunk +I wouldnt think endequeue is a bottleneck so Im curious what points you in that direction +Note that default values must be specified as a json string since Java does not permit arbitrary types to be specified as annotation valuesIf this looks suitable no one objects then I will commit this +An informal explanation of the +We expect that maven treats the dependency as an overlay like it does with war type +part of release. +Docbook is not available from the archive jumps from to +The invalid values for the first byte in the sequence are used almost as if we had half a byte extra in the header reserved for future use +What kind of test case you want me to createI still can provide example document but the only different thing is execution timeBest regards SergeyI understand it is not a bug and the change doesnt fix an error +Not sure +Fixed +Three clicks no typingAssuming the steps are complete I cant reproduce. +Its working well for me +Is this still related to the barrier sync? Increasing the timeout wont fix the problem with it +Thanks Chris I appreciate it +Justin Im wondering if these schemas should be then managed by +test flapped locally for me but got it to pass +But Dan mentioned that people using JPA will already have that in the classpath +Ed I thought meant stableWe can support backwards compatibility for the old way But the project is too young IMO to start squeezing new features in through loopholes in the current api +Specify more info then reopen if needed. +But in this case testRecursive put in a boolean true rather than a string to the recursive attribute +The color brewer module pom says youre the maintainer Ian +Rebased +cordovaI dont really understand what this bug is asking for + on JBT +Should the AOPAS integration not live with the AS +Just let me know what needs doing and I can add it to our product backlog +Thanks for all your investigation and work on thisFor the sitestage I will discuss on the listI confirm this as fixed then and await +Copyright would need to be transferred to ASF before doing anything with the ontology in the VCL. +However one might argue that +Added new category for Query and Search packages in and marking as resolved. +So mvn clean findbugscheck doesnt do anything but mvn install findbugscheck would do +thats the current design of this feature as shown by the examples above + +But anyway we should make it the default for all r u saying that I can assume it is ok to assume V is default and apply the patch +About of the time we get the error Connection already openMaybe something isnt getting cleaned up properly +When you canceled the patch it seemed to be because you didnt like the degree of complexity I placed in the ant scriptYes +Yes sorry +Tar is supposed to be self contained because it is the basis of our packaging and release +However for the time being Im not going to use it per my design which relies heavily on for storing the distributions with a large key space equal to S where S is the number of hidden statesThe sequential command line utils are convenient for validating the results against the variant so they are definitely useful in my case +in FAQ +Is the problem that do is a javascript keyword and it cant be used as a variable name? If so the easiest solution is probably to just skip do when generating names +Committed revision +doesnt look like will make it into and is using so Im moving this up to fix is already present in and it will be in SOA + Username letiemble +If this is the case we may want not to allow to kill an application when we have confirmed it is at FINISHINGChecked added it back +OK I see the issue the UI is effectively not very clear +who do review it? seems like an harmless addtion +Josh are you willing to do a code review of itMike +Can you plese post the patches to the hereregards +Verified by Alexei. +Or we have a interface and a with the readwrite +Fixed with in JBT +If you wish to have this issue looked into please provide us some sample code otherwise we will keep testing with our test suite and we will keep marking this bug as not reproducible. +Thanks for the review. +We can see when you have something which is worth including +I just committed this +This works locally on dual processor Fedora core and single processor Windows XP +Issue closed after the release but was included in it. +Er I read that wrong +probably fixed already but forgot to update this jira. +The eclipse log +Assigning to Adrian as he should get credit for the fix. +This feature will we discussed in the ML before proper changes should be applied +bulk close of obsolete debugger issues +push out to +batch transition resolvedfixed bugs to closedfixed +Committed to issue as it has been released as part of release +minor thing to move to on user poll and since Id rather do this post install in the admin interface this is really a post M thing. +bq +This is real essential! +Dan can you look at thisMax or Alexey can you tell Dan when you need this investigated by +This patch now only fixes all the headers adds checkstyle to keep it that way +on patch +The changes to the spec are not correctly processed by Forrest for me +I have added new version of the patch which now adds support for Acctg Trans Entry Type Id and it works with currently latest SVN versionI still think this is a useful patchbtw could somebody explain why update service allows to modify posted transactions? If it wouldnt then there wouldnt be any imbalance risk mentioned aboveIf update should be able to modify posted transactions then isnt it enough that the transaction editing screen doesnt show posted transactions thus practical risk that somebody would change posted transaction is minimal +yes I think the globber can be fixed to avoid listStatus on path components +But I forgot to check for the hashcode methodDo you think we could include the source code of the Arrays class from jdk or just a part of it in a new class to have the hashcode method? I dont know the detail of the jdk licence +htaccess file committed revision Also did the svn up on +failure is unrelated see +Lets add a check that values inserted into an indexed column are less than the short length +I just ran all the tests and this fix does the trick +The idea is that they should never need to trycatch a build session but always check the results list +Please let me know if you have any problem while applying the in patch with revision +For example you cant store a null value for a key +Ah yes +If you find further issues please create a new jira and link it to this one. +The are placed in the lib directory only if you build the ICU using the CYGWIN option +examples possibly depending on if they are raw output that needs to be matched etc +I cannot reproduce the issueFresh installation of Eclipse on Windows no extra plugins +Directory Studio version has been released +Fixed on you have a build +not at the moment its just the minimum that most clients can aim longer term is an API for more meta data based remote access webdav has failed as a standard for more then basic filesystem type access. +should work now with the current trunk +Committed revision +William do you have a patch for the bug with finding the mbean server by domain name +This is my first post +stacktrace of the problem +Just dont have it in there and set reducer to zero in the job setup +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +A patch increasing the segment overhead constant +upgrade to wperms off +Then you can debug your implementation with the flag set to true and the TCK can run faster for the normal case +Jukka will put this on the Hudson master +Let me add the running of the fixup to master joining an already up cluster +We need to investigate the possible avenues to determine the best way to add features for REST +I then use a count down in the session itself and use the Ajax action to decrement the timer and to inform the client how many minutes remain +Release builds works ok +Now I am getting different error as toURI I have very less time to resolve this +Some fine tuning You defined package protected abstract methods but made them public in the enum constant +Will do some more testing before committing the fix. +Since you already know it there is no point to repeat the effort of figuring it out myselfCould you provide a patch +Derby should be able to boot in mode against databases in jars or databases on mediaI thought there was an existing bug but I cant find it yet +Committed to both and trunk. +Now we have to revert the operation we inject A back into the server BANGThe server does not allow addition of aliases pointing on invalid entries Check For Target Existance We do not allow the creation of inconsistant aliases +Submitted the patch +I wonder if there was any progress on creating a reproduction for the cause of the corruption on this issue +Could you give that another goThanks has been has been updated and is currently being builtthe features project has been updated to use the latest release versions the only snapshot dependency remaining is theNMR has beenFeatures has been +Hi JesseI could not reproduce this issue with the sample applicationCan you help me out i am not able to proceed with this issue +This issue has been fixed and released as part of release +The total sleep in sec +this happens only in the cfcatch block of a process that runs in its own thread and throws exception +patch applied sorry for the long wait. +On Mac the failure wasnt reproducibleOn the Linux box I have access to the test is blocked by +Im guessing needs similar changes made to its too +This somehow is not in and +Note to self most maven related NN stuff are actually scattered among other components therefore should be added to the appropriate NN descriptors Maven new Quickfixes for or redhat missing dependencies new profile template Central new archetypes new simplified UI for the wizard support for the enterprise flag Commons improvements on the Materialize Library related NN committed in trunk for Maven Central Project Examples and Commons +This will address the case when a client is accessing a file and the first location it gets is always a random replica +I can confirm those errors w JDK and Eclipse solution was to create an project specific compiler target which is stored by Eclipse under +Are you still having troubles with this? If so please post your and the file referred to in your examples shipped with Mapbuilder should all be working +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Zoni report +we can document that as a possible workaround until it is fixedI also have to research the areas where we accept the file piped or redirected to the parsers and tokenizers on the CLI +thanks +I presume that the miniCluster part is what makes my attempt different +Also tried on Ubuntu using same version of SMX but did not get this error +example application supplied with is not working for validations both and +Dont know how to use fix versions +I also took the liberty of cleaning up this whole class a little bit while I was in here +Reproduced myself on Windows s how the original reporter reproduced it on Windows Pro SP bit +commit the new changes At revision in you Jarekcommit to trunk At revision If input nothing and click GO you will get the full list of all packages. +Carl Steinbach notes that sudo port install coreutils will get both of these utilities for Mac users +I put a site up at +Now that work is well underway with AS all previous community releases are +This implies that wed have to manage the enlistment though since in local mode we cant do recovery so wed have to be sure we are enlisted as volatile +Comparator does not implement Serializable +shell etcIll attach a separate version for the only real difference is the domain where the show up + +This has been fixed in +Hopefully this will get done some time soon +Patch that use the solution proposed by Dimitris +Thanks. +The log and snapshot files for an occurrence of this with a single node using ZK +In fact they are standard properties from Sun and Felix current URL Handlers implementation already supports themThe properties for stream and content handler packages are respectively These properties are a separated list of packages +Here is an implementation of circular dependency checker based onJust a of validation logic using your new planAfter going through all the current Pig commands yesterday I realized that DAG implementation is not necessary as it seems impossible to create a cycle using Pig languageThis validation logic may only be useful in development time or in the future where new constructs are added to the language +Looked into it and found it is really a flaky one +I dont think it is necessary to fix this for since it apparently is not a common kind of situation +There was a suggestion to include it in the product folder which Im happy to do next time if that is where it will end upPS Just noticed that this carries the old headers so will fix those for the next time around as wellPPS Im also adding another patch to the product module that hopefully can be reviewed and committed as is +Surely I can but I have nothing in my descriptor regarding filtering +Well it has to be one or the other +is done +Perhaps the message could even include a link to the gory details +Ignore trigger action possibly changing the lob value +Resolved for a while +Better cleanup is good anyway +STS versionVersion Id should be fixed in STS and you please give this a +henk can this be closed now please sir? +Otherwise it can be provided as an option +Fixed with revision I didnt want to take the time to figure out the section anchor for my link +Dont endlessly add and then remove and such +You can set one cluster to +OK this patch does a better job with the matchVersion as per Muir +Without this which is in cant load the yoko impl +Also what would happen if another store function like is used? like our comments crossed +Rather it is a classloader problem +Thanks Jitendra +bulk defer of open issues to +Sorry everybody Ive been thinking about this the wrong wayOne of things about Pivot that keeps biting me is when widgets have similar names to the ones in SwingAWT but act differentlyPivots is most similar to SwingsAnd Rogers is most similar to Swings and is basically a simplified copy of PivotsMaybe I should just document this in the code so I dont forget itRoger I withdraw my objection to your patch +thanks so much +Committed a fix to this +Available in both Central and Extras site +The dependency plugin does not read the pom file directly so theres nothing to be done here +It will be under srchdfsoahdfs Anything under srcjava that does not go into srchdfs and srcmapred goes into srccore There is no new jar +As far as I could tell my patch does still consider the resources configuration reason that this is needed is for those who have resources that need to be included in the build but are not part of in bundles +When I mentioned a workaround maybe its possible to disable the JSP custom tag libraries support in Grails in exchange for this support? As I dont use at all I would happily just do agree with Graeme though that the real thing that needs to be done is a GSP editor for Eclipse +If you still have a problem with this please or ask on the mailing list +How do you double clickDo you wait first for the page to be loadedElse you would click on the exact same url and that url is something like wicketinterfacexxxxxYou say you fast click +The semantic of sizeMax is specifically the maximum allowed size of a complete request +I think an argument could be made to throw an error when the prefix is file but defaulting to only allow hdfs is odd even if it is configurable +gitignore file to not track the eclipse project settings Pushing to would have no effect to the running application though +I did both +I committed this +closing all issues in resolved state that have not been reopened nor commented over in the last month +testcase and m resolving this for MR since Ill need to open a separate YARN jira forThanks Sergey +Can you try the suggested fix +We ended up doing it differently. +Any other opinion? +Element excludeTransitive commented out using goal jnlp +yes youre rightI think a reasonable behavior is in the case of a for a warning to get logged and the connection hung up on with the intent that the client should come back laterIve got a patch that does that and Ill add a test case for this situation to make sure that its handled as gracefully as possible +I mentioned it mostly as a reminder that there is a solution to your problem already it just needs to get tested and committed +Maybe it will be incompatible to your software +works a charm now +Cool +Thanks KnutMaking volatile does solve the synchronization problem between the threadsI do not see OOME with the above change along with the fix suggested in previous commentsI have attached a patch now +Hi Michaelsure +Rob what exactly was this issue about? Is it still relevant given all of the refinements that we did in the meantime? issue is no longer necessary +Thank you +I believe that the diffs in the jdbc tests are caused by additional overloads recently added to JDBC +and actually since false is passed here for ignoreCase the constant does nothing +Fixed dependencies +No +Also adds some unit tests for the code +My suggestion does make overloading a bit more likely as requests take a little longer to serve but the implementation mechanism is just the same so the difference is just one of degree rather than of kind +The containers have been corrected and optimized as of rev +Xing the patch was not attached +Thanks! +Overloading the one enum seems a little dangerous +What do you think +should be modified to allow the knowledgeable user to choose between XA recover and no XA recovery and make clear that the XA recovery choice implies changing the transaction manager +I had already opened a under as +The file looks like it tests this but only for test failure looks unrelated +Thanks Paul +Rosta is it ok to have in endorsed even when running with JDK? We dont want to ship separate distributions for JDK and if they require different packaging we can only make one the default and the other has to be documented as an installation step +constant name constant name are right Im you veryChema. +This jira has been closed as part of a review and archiving of the aj project +Committed to trunk! +I mean we reach the split size etc just because of the flush and will not probably effect the results but i wonder if forcing the flush is the best case +Im not strictly against changing the current behavior but I would like to better understand how bad the situation is +In my initial investigation I copied made no changes other than the class name and command ID ran it and it successfully opened an rsh to a +See for progress on providing in JBDS s CentralUsers who want a newer version of should uninstall the feature from JBDS then install it directly from Springsource +Can you attach the log for the failing case +bulk close of all resolved issues. +is read by a lot of developers +Thought of implementing a way to output query plans to XML and convert into HTML like thing will be doable going through Bryans he says there are tools to convert XML to HTML such as or Stylus XSLT Editor toolsetcThank you +How live is wiser +The previous patch corrected the problem in +patch that add the config which will be used by +Owen recommended to put the jar files into HADOOPPREFIXsharehadoopprojectlib to reduce custom script to construct class path for contrib projects +The renaming patch looks good to me and seems like a fine place for it +Would need to synchronize all puts to HBC since the Configuration backing Properties is not accessibleIm going to cancel patch and move this out of +Can anyone commit the fix that Jose provided? Its an important issue with an apparently easy fix +Committed to trunk and +Its using the bundled but calling methods from +Roger Unfortunately we dont have enough time to make good patch right now +I have no idea why this had been closed as its not fixed +But without manually flushing it I would hope that entries in this cache would automatically expire after a given timeout? If so what is this timeout and how could it be configured easy or not +If the general feeling is that the Zookeeper codebase isnt ready for case Im happy to host my Activator elsewhere until it is ready for it +For now Ive put in a hack that redirects calls to the old reader to the new one so I could make sure all the unit tests pass +Yes I am just using the same benchmark and reproduced many a oddities with K maps and was extolling you on the way for the benchmarkPlaying with and profilers on this benchmark now +Sorry my mistake the patch was indeed committed by mistake onI did not see it because I normally just work on trunk where the patch was not mergedIm going to revert it +What is your opinion Jason add to clean or create a target +Released as a part of JDF Boms was released as a part of. +Can you please retest the changes? txs +Any browser that has a cookie with as path mustsend that cookie before it sends the cookie for jakarta whenvisiting a based url +So I think we need to handle specific classes as we become aware of themAre you OK changing this report to be Need serial for +I am not sure this patch is enough since theoretically I could deploy the same app more than once +It is testing the not any mapred functionality +Needs xsite so we can build the manual which we had to left out of the first build. +Yes I think it should be along with a few other small tweaks +No worries +Radoslaw line number from the original Ruby code? Because since its a it should have a full Ruby stack trace already +Thanks Chris and Ganeshan! Will commit shortly +This patch makes all tasks and works serializableCurrently there are no additional testsOnce the whole is serializable we will serialize the whole query plan after compilation and deserialize the whole plan before execution +Setting this issue to resolved. +unless you think the get to awkward to use and that we should get rid of it and turn all calls to get into get +Im not sure this is a bug but rather how it works by design +What exactly do you want to patch +Andrej patch contains the change in parent pom +Thanks for finding pointing this out +Sending srcjavaorgapachelucenesearchspansSending srctestorgapachelucenesearchspansTransmitting file dataCommitted revision Thanks Paul! +Awesome +The behavior is changed now to throw a if the given is not found +I dont think the patch is getting submitted to jenkins +Dhruba it was decided that we commit patch in HADOOP which is to have first replica on local node second replica on node on different rack and third on same rack as of nd replica but different node +I dont understand the proposal here more explanation is required +Ill be looking into your proposal to move access in our application to within a static block in the same class that the converter registrations occur +I believe sstableloader uses an int so ill follow that and will fix it later today +the target is eap as the javadoc bundled with EAP still contains internal API. +updated the patch +Silly of me to have declared on class not an interface +Login as adminadmin create the users we wanted + +Attaching new rev of the design document which is with the implementation +Attaching my jmx failure +Reporters of these issues please verify the fixes and help close these issues +Heres a new patch this one has a sentence in the escaped syntax section referencing the two functions and linking them to where they are in the section +Can you pl explain in greater detail the alternative that you propose +Also replaced HDFS with local FS for the cluster so as to quicken the test a bit +File which is used to test the restful web serviceJust attached at the top +Closing wo any other testingvalidation +No patch file. +Added on trunk revision . +Marked for keeping it on the radar +Im not closing the bug because this doesnt address the issue of reporting progress and then reporting failures which is also part of the bug +Im not sure what would cause that error +Will do new issue to address your comments +Is synced to fuse trunk +Removing Fix Version since dupesinvalidwontfix bugs clutter the release notes. +Closing resolved issues. +Hi could you attach a git patch please + provided which doesnt use the pre for the testcase provided including for RF M and Guava r +Ill revert this +Okay i will do itI have no idea Luk all this AS and vfs all r killing me since we started development in AS +In the first instance can you please review my proposal and code submission? If it looks good then Ill proceed with the unit test and update the wiki once committedKind regardsChristopher +Thanks +Found another problem in when processjvmargs is in jvm args should be OK since is given a string with all the args for the java cmd which must be parsed before being passed to duccling +I can reproduce this issue manually with the following steps +Well no one has objected so Im going to start scoping out this patch +is the next bugfix release +If this can be confirmed I will be resolving this JIRA +Thanks Arun! +Rene can you provide a patch along with unit tests? Have you or can you run contribbenchmarks and include your before the changes and after the changes results here so we can see what difference this change makes? Thanks +Deleting and Archiving Events and Alerts +Heres the patch +That release number should be uniquely associated with a particular branch which could identify and use +Specifically need to look for for the query +We dont have a very reliable repro path +And Abhinav if you test the case again with the latest code you will more detailed vmware error message such as stacktrace etc to help us to find the root cause +I usually run dosunix on the file before I look at it with vi so Im not distracted by the end of line +I took a look at +My question was just asking if I was missing a step in the process as I assumed everything was configured +Most likely WS doesnt wait until the container is ready +The button to save configuration disappeared too +The system falls into infinite loop trying to recover a lease making the namespace image inconsistent so that the cannot notBoth of them are critical for the release +The patch +The junitreport task collects all the reports in the junit directory into a single test report +in case of an upgrade to jackrabbit we still have a jcruuid property for each ntresource node in content +Sending srcmainjavaorgapachestrutsdispatcherSending srctestjavaorgapachestrutsdispatcherTransmitting file dataCommitted revision . +No this only works for annotations at the method level and the type level +might be a similar cause. +It can actually enable a lot of components to get information wo raising new issues in the is available from everywhere +I seem to have implemented everything else holding up commit +strange +JBN EM shows availability of both of those nodes as being green +Patch committed +we noticed the issue when solr became slow or unresponsive on the machine which solr has not been restarted for a while +I have not tested that wraps around +First version of patch that doesnt quite work +Source could you attach a complete webapp eventually stripping the libraries if it seems too bigThank you + an issue on but onIve just upgraded to Sorry that its a dupe I couldnt figure out how to have JIRA do an AND search on Comment and the number of matching issues was huge I gave up after pages and just decided to enter this issue as being better safe than sorry vs +It is up to you +Attaching the patch with is due to just committed this +Committed to and trunk +Oh I see is to work around the breakage the Thrift guys introduced +looks good yes +RickCommitted revision +Its been like this for far too long +Its fine to the CD change in a separate jira +Because of that I created to upgrade to version of that module +agreed +Switching all issues that have been in resolved state for more than one month without further comments to closed status +I think this will be very useful for many usersWe are freezing on Monday so moving this to release +its not +It is based on have trouble accessing SF these updates to a patch +I do not want to protect a put overwriting another put +I agree it probably meanswe should limit the set of properties to a fixed one +It will be closed if you dont provide any code or tests +The retries are putting such a load on the master its not able to do much elseThis patch should fix the issue +This will be hard to do now that we are sharing codebases between and +Also please review this and confirm its fit to commit +Thanks Stuart Please give us any update you have on this issue +first implementation try svn diff file +I can revert +Oh it is good news +I checked in a partial fix under r +But subprocess then signals the super process root token which is already ended and the main path of execution continues +I looked in these files and could not find any significant errorI was using before this one with no errorsregardsDanny +wilhelmbierbaum True +Heres the initial implementation +I just got confused because of the subject of this JIRA +And the war file is built from it +Many of the HTML element collections uses appear more than once on a page +So its really temporary +Alan would you please take a look at this one +Thanks for sending this in Scott +Anty I spent a bit of time thinking of this +Ok the Ant build now takes a property which defaults to true but can be overridden by the caller by using +The original NPE fixed by the uploaded patch is visible all way back to marking +Take youre time to finish the worg +Instead having a system property set the default timeout lets specify a with default by doing this the ratio can be also applied to tests that have a custom timeout +Thanks +It would be great if this patch could also be included in to also +I dont think there are any missing dependencies +could you push a version to github while were waiting on te official patch +Rgds merged to. +page You can use namespaces with Xpath annoation +Tested and confirmed working with Sun JDK +Could you please move them to the top of the files for consistency? I am otherwiseDoes the same fix work in? If so you could consider attaching a patch for although that certainly wont prevent getting this in trunk +I was thinking about how this could be done and had the following thoughts A tablet should advertise different ports for reading and writing so that reader can go directly to the process running scans +fixed rThanks +Id be interested to see a performance comparison with file tailing +After giving this some more thought it seems like the right way to do this is option +applied. +PatchWill commit when tests passed +The test case fail for me on Windows XP using Suns JDK andI did some simple debugging and printed out the underlying int for both the Calendar and Date objects +Got rid of many sleeps in the ftp component cutting down approx secs +The Sales Order checkout process is functional and can be tested +Found the js as well +I am convinced that it is the right way to go +Please give them a try +looks fine thank you George +These issues have been resolved but never closed +your patch looks much better then mine +Actually its an old problem I saw when commiting but it seems that noboby has a better solution for now +Committed patch to documentation trunk at revision Merged to doc branch at revision . +This is fixed by which allows for the persistence adapter to implement or for a custom scheduler to be configured via the set method on broker service +I like Carls approach +Can you please verify whether you get the same behavior when running against Spring? on problem still exists as described above +I replaced the actual file names of xxx because actual filehost names are internal to our organization +But I think it would be nice to have the rounding method in a helper class as you suggest +Someone has overriden it but forgotton to delete the target dir in doing so + +Unlinking as Windows support is not in. +it calls getResources but it doesnt check to see if getResources returns an empty list +NOTE Repeat detection is based on the cascading pattern where one of the choice paths is consistent in all of the choice instances in the hierarchy +I think Ive sorted this out +Ahh finallyDont forget that we need to add apache headers to new files thoughAdding apache headers +The classpath containers dont include all jars from a server +Lot of things have changed since this bug has been filed +Error in the way the installer is getting string values for localization + I dont think we should include in the release candidate changes that are not checked inIf we decide to refresh the candidate then we can add this change as well as the demo html changes if we get a patch for those +Im willing to live with thatIf youre not we could just just check for BCL and only enable this if theyre in batch mode +This patch does this +Something like a small test project to demonstrate the issue is going to be necessary +This appears to be fixed by no +Since uses a you will always get true so you cant count on this navigator property to give you proper network statusNow I know that when Android devices transition from G to wifi and vice versa they are offline for a short period of time +This patch is for patch looks similar to +Isnt wcs branch going to become geoserver? If so the fix for version number is wrong +Poor performance is a major detractor for any framework +compatability for +Whoops I was a minute too late +So I have a seperate junit test for DFS and +Fix propably should be applied to JBMETA and all custom hacks should be removed from AS modules +Attach an example that reproduces the attached the file that causes the out of memory error +Remove unused +bulk close of all resolved issues. +Moved to move from to to to +compilation is handled by maven tool in case of the maven projects +I will update the will take a look +MD generation is CPU intensive +Intead of having a selection box in the header just display the currently selected language and provide a link to change it +For bootstrap class loader it not only looks up a class but tries to load it if lookup failedI suggest separating these two functionalities for bootstrap class loader rewriting implementation of kernel classes accordingly +fixed in close for the release +Ryan could you please figure out NPE is thrown on which operation in the edit log? Otherwise I need a copy of your fs image and edit log to figure out the problem +If not please reopen this ticket. +Passes local tests +So Im going to stick for now and itll be relatively easy to change later on if it proves more confusing than I think it isSo anyway felt we were good on the overall patch so committed +Changing fix version because this issue isnt am moving the ending question to cleanup the associated FIXME +Please dont delete old files +Available on container integration to be ported to AS code base during webservice component update +I can find no difficulty with the code and it seems to work for me +From the capture it looks like your timeLimit on the search is set to seconds so that may be limiting the returned results +Steps to reproduce Edit title of page portalclassichomepage in page management as described in previous comment Go to httplocalhostportalprivateclassicportalnavigation Click Edit navigation on classic portla Click Add node +With your new patch you can just attach it and leave the old one for referenceKathey +Fix was applied as expected +The same piece of code is used in Secondary Name etc whoever needs to connect to +have a look at your own +Restored original patch +These work for the Documentation Guide which has a daily build on hudson +Thank you Chelsey for contribution +Feel free to edit if needed +The AVG and all VARIANCE related aggregates need to use double internaly even on LONG data types +I think we can just cancel all the builds of that project when cancel button is clickedFrom the queues page you can select which build you want to cancel +Verified the includes and excludes functions worked correctly when used in a Query. +People can choose to prefill if they want that default is You have to decide if you want to run as part of creating the datasource from the console +Ah right that makes sense. +Test is removed at +committed svn version +thing is not related to a main problem so I should have filed a ticket +Seems like we need to handle these types of failures anywayRight but I was seeing them as two types of failures possibly +testcase is the space simple workaround remove the space. +This allows the directory to be properly removed on iseries +The way it works is files are nbproject which is shareable in cvssvn and nbprojectprivate which is not contains just the token that designates if its a jboss or tomcat or has the actual instance of that type for this particular user in the property top of that it also stores the property with jars in the app server and puts them on classpath for compilation etc +Ive followed your suggestion to commit this patch and will file an additional patch for just the one file that needs another changeCommitted patch to documentation trunk at revision +Can use an editing pass +Moving to next CP quite a number of issues to address not possible for this CP +Merged for in revision About the API change Im accepting this in the patch release as the main API in this case is the PROPPATCH method whose implementation is corrected by this change +Used implementation is Apache BVAL +After fixing to use remove as well this looks fine to me now since our are quite carefully designed to avoid specific Spring types if possible already +This could be quite useful for replication too as a way to the metadata check between copies of +One can plan to avoid this by reducing the business logic inside Action objects +Yes I think it would be much better to generate the XML form of the queryplan statistics by querying the XPLAIN tables than by trying to parse thetext output of the logged query plan +committed to branch +Thanks for the review Hyunsik +I would like to see the pushed down to the rather than put on the +All the expressions in the filter predicate may not be pushed +Defer all open issues to +This code will also be modified with the integration of ZK +Ari What o you mean by performance testing +Failure to throw the correct exception will cause the system to enter an inconsistent stateI therefore suggest we stick to simple exception handling mechanism where the processor catches all exceptions and backs off from retries for a predictable amount of time +Thanks +Thanks again Henri. +As a guard against spam your account will have to be approved by one of the committers but that should be fairly straightforwardThanks couldnt find a way to edit the Wiki page on whitespace management +The RM does not have blacklisted it has unhealty +What Java and AS version are you using +Thanks Paulex patch applied and the issue is resolved as expected +Updated design document +The other nodes will auto extract the war when you start jettysolr +Thoughts +mitosis is an outdated issue +Jitendra Youre definitely not using +Assigning it to Min to have a lookThanks from UI in creating Zone no matter whether user is selecting or UI will always send the following commandhttplocalhostclientapicommandcreateZonenetworktypeBasicnametestZonelocalstorageenabledtruednsinternaldnsresponsejsonsessionkeyeuDthat is missing securitygroupenabled flag +Martin please can you comment on the work done so farAlso is nearly ready or shall we close this issue and another time +Thanks Jacques for pointing out that some Dutch help files are still incomplete +q filedfs rmr seems to fail if the directory doesnt actually exist +I restarted the eclipse and now I can generate the quickstart jsf project +Verified on nightly build with. +Yeah sure +push out to +If the main use case is security then I think were better off adding this functionality in a way that makes the additional work visible to the optimizer +Here is a patch incorporating Owens suggestion using instead of http for finding +Patch for trunk +Per your suggestion of thoughts from others I have waited a few days +if you have no output why configure the option? I realise some test cases will be different to others but its more confusing to be missing a file than to have it and it be empty +How can I help you +here is a patch that uses a single array if possible +xml would issue is detecting files and not trying to deploy them until that is complete +My inclination is that we are better off leaving this as it is +Thanks JodyThis issue is caused by the WFS changes in and is not present on stable +In that case there is no need to clear reporter at all +I found the balance ago in branch is invalid for our scenario +in +Ive been looking at this one for a good while this morning and I dont see anything obviously bad +Please object if you see any concernThen I will do similar refactorings for and Dirichlet +Great idea +we still dont have a way to associate a call to sstableloaderbulkload with a specific streaming session rightNick you are right +Additionally I think commands that only read information should not ask even if you want to display osgiinfo for a system rethinking of that I think its the expected behavior as the is set to all bundles in this feature has thisAnd as the threshold is set to the bundle is considered as a system bundle which is quite the caseHowever Im agree with Christian that commands shouldnt ask for system bundle confirmationI close this issue and raise a new one +Making thisconfigurable might be appropriate +Still need to consider if I should just make bundleframework implement these interfaces directly rather than having separate objects +Can be postponed until +Actually the bouncycastle classes which reside in ext directory should be give all permissions according to the policy file +This is just to provide a simple option to enable to use with simple use cases not target all possible use cases +one thing still needs to be investigated whether users should call the or not +After all I was able to resolve all problems and also cleaned up codechanges after the here you get the final patch for review. +Most notably correlations +Ill monitor this for a bit on x before bringing it back to x +Not to bother you too much with internals but if use case isdifferent then a separate JIRA please +Use the classloader associated with the conf +Logged In YES useridAre you really usingOr is itEither way you need to provide more information wrt to the way you are specifying your transactions and your use cases +I therefore suggest you to open a new issue for resource prefices if you think this is something to think about Werner +can you please submit a patch +nocommit we should not init cache w full capacity? init it at and only start evicting once entries is over our max Same here I took the same thing the old cache was doing +Manually tested! +on the patch +Indeed no reason not to expose this attribute +There has to be some such code so that the web beans injections can be integrated into the EE containers component creation process +that result in a job being launched were modified to check if the REPL is running and if so create a jar of code compiled from REPL input and ship that jar with the job so that its on the classpath of run tasks +Mark Any update on this +Rejected +I think this solution might be adequate +Lets make this as the Jira with failover being a subtask of it along with other things it needs to do +Idea is basically to allow MR to be by maintaining cached FS objects that are distinct from the FS objects used by user code +Im debating whether to make usetrue the default for AS +That only leaves DB +Thanks. +Having two different versions of same jar causes unwanted griefHives execution mechanism ensures that Hadoops classpath gets preference over Hives +Does that sound reasonable +Thanks for this perfect patch! +Reverting availability till then +can you submit this to the proper jira? This isnt hdfs +Attach a fix +Committed to trunk. +Its going to take some time to merge that change and adapt it to work correctly on WindowsInstead Ive reset target version on this jira to just +This table would be generated on the fly +Would be nice if we could make it a bit easier. +Have added Makefile to fix missing makefiles +We have a problem if it never goes away +I think this would be very nice feature to have in jBPM +New patch without any warnings +Let me verify if it works in real databases if it does I will disable sql for this test or do some workaround for derby +May add rbCan you describe the capabilities of predicate pushdhown with ORC and how you modified it to work with vectorization +Exactly why does running the war the first time not work? WAR and EAR projects started everytime at my workstation +Good point +Does anyone know if this has been or will be fixed in Seam so it can use Hibernate +Am I missing something +is the only remaining issue affecting conformance we arent going to fix that for so Ill this issue and punt to to fix along with m closing this as we arent planning to implement the only remaining issue +The Model updata phase was skipped for the form +It will be transparent for your team but it will be in the maven way +a couple of problems so far when hibernate generates the SQL schema the foreign key constraint refers to the incorrect column in the primary table when inserting objects the primary key of the join table is set to the primary key of the primary table instead of the columnOnce Ive fixed these issues Ill submit another patch +Just added this property to the just committed this +srcjavaorgapachesolrcore +So the released version does contain the problem +Heres a patch +As far as I can see you have to be using the old http binding to get into this trouble +Thus HADOOPLOGDIR and HADOOPLOGFILE are not set when executing Nutch commands +Between and there are various differences in ways that it collects the tasktrackers where tasks ran +Back to the topic I reinstalled groovy with the windows installer and let native launcher to be installed +Login as mavensync select Account Maintenance in the left menu and have a look at the bottom of the page +Im not really sure why we have a dependency on Xerces at allThe dependency probably couldshould be removed +testcase attached +It was set to a high number before because we used to not sort by update time DESC +I implemented it that way now +Hi Jacques no problem yes i saw it was very late yesterday night as you made the comment No Problemthanks for the commitCheers +We are specifying the wsdl location +Zhenxiao please see my comments on reviewboard +Thanks to Deepak Dixit on dev ML. +Heres another patch that checks if a repository has been scanned by getting there ids and placing them in a Miss Deng +JuergenI know is needed for running Wicket on applications as AFAIK especification does not support filters +HiCan someone apply my patch please +Only two bugs are actually fixed between RC and +Raghu thank you for pointing out that is sent in RPC in some hidden way +Changes look fine to me +See +This bug doesnt reproduce on rev +I am planning to create a wiki page with ideas how to ideally define features endpoints +If you dont support maybe the best place to put this is then your Wiki +Usually they are the one that actually create the osgi metadata so not actaully using them would just break the whole thing imho +So this can be closed. + assumes getParent never returns null in a lot of places +Patch accept and applied +Finer grained locks are only make it less probable the deadlock will arise but this will also make it harder to debug deadlocks when they arise because they will be more difficult to reproduce +So that the concat code takes care of removing or adding replicas of the blocks of the new fileBoris could you please revert massive refactoring of imports in and submit it as a separate patch +Just wondering whether you care about the marshalling side of things only +lgtm nice find +jpa ftp asf +stuff thanks! I committed the patch in revision +On party profile page you will find a tab for Billing Account +I only have a concern about how the serialization filter would work with UDP transport +Protocol has replay from a known state which does allow compensation for this +Kabir or somebody with javassist knowledge would be better placed I guess +Unsetting Fix Version on unassigned issues +Please reopen it if you have any additional information +Thanks KevinPatch applied to BEANS module at repo revision rPlease check it was applied as you expected +Is it acceptable to just leave the class annotations in place and add more visible annotations to certain methods? Or is it that the annotation is the one that applies? Or would folks prefer that we create a new class or two which just selectively expose these interfaces +After merging it the jar fails to start +Fixed in rMerged Branch branches the branch r. +This patch adds support for the configuration described in the issue description +Worked fine +Incorporated code review to trunk +Current code have been deployed with the check your application with a new implementation +Sorry it should be with or without the patch. +Some hit ratio have recent ones tooOne possible way is we can provide incompatible new metrics using native Metrics library types while preserving current metrics by measuring twice for some metrics but we will not providing recent metrics for new system +The patch works fine but the spec fails +The latest patches resolve the issue by avoiding the buggy ctors +I would also note that Python and allow optional plus for exponents as well +Doesnt look like people where disturbed by that limitation +Not for commit to Apache. +Im going to switch the over to use a hash join on long keys +So I can compare whether the result is correct or not +Here I have attached the Java Doc for and Interfaces If you have any suggestions for improvements Let me know +So if there should be really problems to implement the flag on a per service base I would draw back my feature request +mrepositorycomsunjerseyUserstucu +Is there any particular reason the escaping has to be related to strict mode? This is an improvement as it shows the user intended to change escaping behavior +In particular this change makes writing directly to the output stream significantly more complicated +Thank you +fixed a bug in +Attached an updated patch for trunk with test timeout +I think I fixed this in master branch by making sure the tonr authentication filter is applied before the fb oauth filter +Per the JIRA meeting on April this should be a warning not an errorAssigning to Brian M +That would also solve similar problems that exist for custom delimiters that span block boundariesCurrently the line reader is buffered and it would be a shame to have to give that up +Thanks Ill make sure not to add the +Mark are you saying my initial assumption is correct? Should it be detected and scoped to the webapp inside the EAR + +It is common with aj that you abandon requests +Please change back if you disagree but Ill need help in understanding the connection +Therefore I think the default behavior should be to include those file in the +a we will not send events to disabled beans anymoreb since in we also find observer methods in superclasses the workflow is different anywayc we do not search for the most specialized bean anymore +Fixed with revision +Suspicious in the master is the server Map +Fixed Javadoc +Fixed the nit and committed +Therefore order items page should also be updated to provide a quantity field for every ship groups for each order item +Nit In make the instance variables finalDone +Read checkcheckMatteo I feel these two requires READ and WRITE together +One of them is now written in lower case to demonstrate JPQL case insensitivity +Fix isnt working on Liferay Tapestry +Sorry for the delays on your requestsbq + contains reworked versions of the content type setting participants +This should be added to the new features ticket and shouldnt be too hard to implement +of course +I took a quick look ad it seemed the routes were coming out in the reverse order so I added a little to grab it by name +hbase Loop This shows num nodes per mapper not M +Note that creating an instance of a filter does not fire the event. +ive checked and looks like they use +This fixes actual bug +just committed this +Max can this patch make it into beta? commit to trunk and +It seems youre assuming that an HA setup would have three nodes active standby and NN with the NN failing over to do checkpointing against the standby after a failure of the active +Service has been attached +Attachment has been added with description patch has been added with description patch +Now replace the examples WEBINFlib with the commons version +This patch has been released since w no problems. +Added for follow up. +I dont want to stand in the way of innovation but I also dont want Pigs internals exposed to the point that the next time we make a change to our Operator class someone complains because we broke his tool +I have also made a change to class +BTW the maven docs should state that the artifact goal only exists in the newer artifact plugins only compatible with maven +I have picked up the code from richfaces live demo to demonstarate model am using right to left rendering of this page by setting the dir attribute of the page to rtl +This bug occurs on Tomcat as well +This was with Xcode +When you request a Base encoded version of images from a or mega pixel camera it will cause and out of memory error +Yep like I said in no problem turning the Patterns into static fields +BULK EDIT These issues are open to be picked up +Does the work for you? you do not experience the SEGV +Have a look at the openSessionget methods A Session is explicitly turned to there if being enlisted in a in the case of the Session will be in outside a transaction and in within a transaction +So something must have been changed from the official version to the nightly builds if you can call the webservice without problems? Or we have different configurations? Since encoding was the problem in the simple service perhaps it has something to do with the locale on the client machine? I am situated in Denmark and we have our extra characters in the alphabetOk +This one may be really hard to fix +Thanks +quotes removed +logging configuration is echoing log messages twice to console + +CSV export should be the easiest thing +Thanks Alexey! +And then people who want details need to get that too so they can see if a specific thread is consuming all CPU and why +proxies for collection members dont seem to be registered in the +HiI have some questions please help to answer +Thank you Lu Jiang for this patchBasically its good to demo the part attribute and new annotations of servlet +Sometimes errors can get reported against the wrong files if you are still having this issue is there some other type in the same project that is dealing with Calendar and Comparable +Left couple of comments on RB +I am on having an +Set the threshold in DN say k disks +Attached reordered patch for Igor the patch seems to resolve the problem +for the review +Shouldnt there be a resolution pass dealing with these issues before handing it down to the generators? +Can we plan to upgrade our custom Jelly to fix this in the RC +The test case exposed another bug log truncation was not being done properly with the buffered inputstream +In structure it seems to depend only on transport +Maybe when it does the is empty and therefore the version is null. +fixed in +Ill help +Suggested changes have been about changes to book structure need to be passed onto the Drools team +I see +SVN is up to date +Thanks Sanjay +Upload the file to this JIRAgit works better for all of us because its easy to attribute the commit to you in the git history since it contains your name and email address +The problem above is resolved +Juca please verify the fixes on the CR docs bundle and close the issue +in release +WillemCan you point me to some good examples of configuring the binding and the serviceFactoryThanksve added a task to document how to configure the data binding +Thanks Robert! +This jira has reverted the protocol level to for all the commands +DanielSo this is interesting +Thanks Bryan for the updated patch +submitting patch +uploaded new patch to ignore case sensitive for the filter appStates and appTypes +Adding this as a high value fix candidate to the list so that it might be picked up for +HSQLDB occasionally throws an exception when shutting down +After a fresh test it doesnt work on trunk neither +Yes feel free to activate the site with the new banner +The works fine with removing the locks +The only reason I made a copy was to have the ovr testcase completly independent of +Resolution verified to work without the workaround +I dont have the duplicate JIRA for this +Fixed in for patch for the patch +Fixed with commit bfdcdeeeaeaedeed +Any other suggestions that you may have would be greatly appreciated +And where teh tests should be checked intoBernd +Here is how I found it I was working on a program which loads geoserver data directly from the data folder +How about adding a property on for the object and setting it in the appropriate places instead? Then you can access all the schema information you want +Whats that status of the patch for the checksum based approach? I am using and am willing to back port such a patch to if necessary +I am waiting for feedback whether to allow such redirects or not +I also attached a patch that adds back the sorted source so we can spin off a new issue and make them efficient without writing it from the scratchSimon can you invert this patch and open a new issue for removing +Doh! Please use ignore and +Im surprised at a bug like this only showing up now +Ok Im going to err on the side of giving you too much information rather than little + +Thanks +Hi +Done +Sooner the betterRegardsThis is modified patch for the issueThe attached patch could not be made after taking update because server was down so If someone applies the patch may come across will upload the patch as soon as possible when server will be goodRegardSantosh Malviya +bin to +Changed session timeout from minutes to minutues +Cant remember why I have I think its because of the installer I used +Ive created for the follow up +That way the minmaxvalue will all be interpreted with the same pattern and timezone offset +Too small for some uses by far +this is necessary to preserve the order of KVs comparision before and after comparisionDid you mean to say before and after compactionIf this JIRA is for branch please edit the subject accordingly +Thats intentional when there is a complex boolean expression i do this on purpose to make it more readable and intent and precedence clearI dont see a benefit of using! here it only makes code more difficult to read +I think they should be added to incubator group since that is what tuscany is still inajborley antelder bdaniel dims dkulpedslattery edwardsmjekoneilfrankbgeirmisilvaljboynes jmarino jsdelfino kelvingoodsonkentamkwilliams meerajk rajith rineholt rfengrobbinspgsvkrish +Usage link also not working and throwing the same get this problem in an RHEL based system with firefox +Same node rows not work +Attached patch changes archive names of EJB dependencies to +Can you verify if the problem also exists in todays release and reopen if necessary +Attached patch applied and tested. +I am pretty sure it was part of the doc change pull request +How about exporting shiro extensionsplugins as services from the bundle that offers them +Pull requests welcome Shouldnt take long to implement in and the bulk of the work is experimentation and naming design +I have removed this section of the Getting Started Guide +Thats interesting +srcjavaorgapachecommonsconfigurationimport +Opened to take care of Nicolas comments made against. +I just committed this with a test. +Thanks for reporting this some of the validations are certainly too aggressive +Same problem on all Weld on EAP ER +I would have expected them to be tied via LDAPIs there something I need to do to get this to moveSomething must have changed since this morning +related to security arent adding a parallel mechanism that might hypothetically improve things in still unnamed cases +There results changes of records in systemm tablesSYSCOLUMNS SYSCONGLOMERATE SYSTABLESgenerated to javatestingorgapachederbyTestingfunctionTestsmaster +The word reqest is misspelled +Not in branch +I will play with this little change to see how it does with other and I like so far +Can this get some attention from strutsxwork committers? Getting tired of my continuous integration machine croaking over every or builds because of strutsSean you sir are a true giant among men + +The security packages are only needed if the security provider is installed +The error can essentially be ignored +Progress is slow though want to lenda handAnyway please dont use Bugzilla to ask questions use the mailing lists +I am working on some other at the same time +Tentatively scheduling it for +However I seem to recall there was disagreement on this a while ago +We should see if the new parser makes this easier and if so fix it +thanks for catching upon release of. +Looking at this right now +We should benchmark various options and try to find a good combination of performance without memory leaks +This is happening in FF for me too +If not fall back to the default +This has to be repeated for every line which is slow and somewhat ugly +Thanks Shaz! I had no idea we had to break out. +fixed in trunk. +bulk close of all resolved issues. +Thanks for your contribution Alexey +Thanks Jacques for your involvement! +Please review the patch +With Teds patch it seems the shell works but the master web interface gets the exn still +Simply its illegal to add checks for each of attributes like disabled readOnly display its hard to maintain too +Fixed on and +Added ASF license text +Not an issue applied streaming patch as mentioned in IRC. +patch with fix and failing test ve applied your patch into SVN trunk and the branch so that it will get included in the final releaseThanksMaarten +Should be working now +looks to trunk and branches +Tested on firefox and chrome +Instead of inventing a new syntax let us use the local params syntax +Looks good +on the new patch +Bulk close for Solr +send me the fileand Ill make the required adoptions +Good to to get actual user testingfeedback on new features +This one hooks Oozie and HTTPFS to the +Confirmed the broker no longer crashes. +Package has a Bigram class that implements binary comparison +I bet thats related to my last change to reduce the memory allowed during tests in order to avoid a different kind of problem +When the transaction in the test is supposed to be committed the timeout is set to seconds +Is there a way getting attributes when parsing a XML Schema? We only defined column annotations in the context of elements of an Schema +Could not spend much time today +daniel you still have commit rights +bq +The timeout is also unrelated see +ShashiAny thoughts on whether we can just pull out the appropriate pieces +Well it would be nice to have something to test this against Vs us inventing something based on our interpretation of the issue +OK inspired by Uwes persistence on I realized one cleanway to fix the break here is by using reflection whencreating the Analyzer as to whether the class overrides thetokenStream method +The patch is appliedThanks for the investigation +Wouldnt it be conceivable that they might change the encoding to at some point? Or that someone else has created trec compatible data in another encoding? Or trec has data in different encodings? If something reads a source of files and the files could technically be in any encoding I would expect things to be configurable so that I can specify what encoding the files are in +Maybe we should only deliver GEP via an file and stop building altogether +Marking all current RESOLUTION LATER issues to be against the Nightly Build +Verified in JBT +What the does is to accept a trace dictionary and a model in an update method +I did not change the behavior of nor should my changes break existing code +Not sure i follow you since with this you datasource is jta +There is now commited in the trunk inspired by the attached implementations it should work with both and but currently tested only with flat reuse the add integration tests with +This is a very important issue to my company as we automated schema generation and diff it against a previously validated schema to ensure the output is the same +I lost track of this +is a concrete class and it is final +Not sure what you mean +Lets plan on that +If you find it necessary please feel free to open a new Jira issue +and +hi todd would appreciate it if u can write some sort of a unit test for this one that will help getting this into trunk +If I understand the scenario described the standby NN never purged any files from the shared storage only the active did that +I think thats one reason we are exploring the idea of having a implementation +Fixed +It is the messageStore causing the problems since the starts checking right away +Galder the fix version was changed to +Closed after testing. +why does it want to grab the wrong one +Modify the example by making the consumer use the JMS transport and set synchronousfalse to cause it send messages to the SMX NMR asynchronously +Those I didnt Ill respond to in next comment +sounds good I will work on this approach +Ive encountered the same issue on Tomcat when can we expect a bugfix release? +Fixed for +ok Ill start with this +So two s for this patch and no objections so far +A new version of the patch that fixes the single bloom filter per query issue and makes it so the user can specify desired false positive rate and estimate number of elements instead of number of bits and hash functions +Could you please the remaining parts of this patch to trunk +defer all issues to +Added a version +Wouldnt that match CF or KS names or even other parts of the path +Jobtrackers memory mainly user for objects + +The java module is an issue with java application client container not being started that being worked on +Hmmm on master for me it appears to get through all k fibers without problems +Alright the windows native builds support is in +Any thoughts +namespaceUri values are explicitly trimmed now discarding all +I would suggest hourly is a reasonable compromise? +Hi This is a issue to us +But let le ask you one question if we add such a method to where in the generated source code should this arbitrary code fragment be addedinserted? What does this method look like in your fork? PS Is this actually the fork many folks have mentioned to me over the past three years in terms of Maven forked Castor +Maybe the updates happen because of deletes after the index has been created? See for example and which are invoked from +When building a Debian package with Maven the Debian policies prevent from downloading anything from the Internet +Committed the addendum to TRUNK just now. +The same problem occurs on Spring as well +svn ciSending shellwrappersrcmainresourcesorgapachekarafshellwrapperwindowsTransmitting file dataCommitted revision +Added a unit test in trunk source under module that would test consuming of expired messages +Done for all old versions. +It wont be a direct match with the package but will provide everything needed for doing things the Hibernate way within a application +Ok +Having worked a little more with this I would like to amend my original suggestionI think the first table should be titled something like Features and the second table Parameters following the XML of the fileSo perhaps the heading for the pane including these two tables should be renamed to avoid confusion of having a section and table both titled Features +Ensure that multiple select values are passed correctly +push to +Closing issue Oozie is released +Line contains the call to which doesnt seem to exist +Doesnt seem like this one should be targeted for since there arent any patches here yet even and that its still being discussed +However its not worse as it is now either +It just something Ive never seen before +verified + +Committed. +Supporting request and response seems very easy patch attachedLooking at the spec and looking at the code theres no similarity between the two list of variables so even without the thoughts of backwards compat Im dubious on removing pageContext +oh this is interesting +Thanks +Updated javadoc and the groups to the there any open source project that uses Java security frame work for the security similar to we want to use? The current HDFS use case is to get user info for every RPC call +This is resolved by adding timestamp to the log file +Committed +If manually placed there property code completion works fine +Moved to t make applied in r thanks for the applied thanks for the fix! +This might be a separate issue to this one though +now Solr packages too are OSGI with in rHowever this still does not have support in the official buildFor that it seems like the right approach would be with bndtools +Will open new issue if see it again. +Committed revision . +The MVC Java config on the other hand can have the flag on without breaking compatibility +If I restart from server Admin screen times I get a Java Heap message +Yes sorry I did not see that ns mapping +Moving this into the next release also atm I feel we should not change this at all +In general please review and try to reduce the number of dynamic allocations where possible +Verified on my side +Additionally a namespace will remove our original reason for using there isnt any action that needs to be taken I would like to mark this issue as resolved +Pavel can you take a look at this jiradocumentation +CVS now +Certainly looks like a split brain decision +And debug it thru give us a solutionD +Heres a screenshot of the error +Ive stumbled into a similar problem while working on an improved patch for +Attaching the current version +Even running project with RC but with Tomcat I wasnt able to reproduce the problem +Id say warn level as there is some unknown problem during authentication that should have been handled by the authentication logic. +Thanks Daniel for the review. +Well hibernatetools does need to connect to the database when you either are reverse engineering from it or executing queriesIm very much open for making it easier to use so any ideas in form of patches would be welcomeOne thing I definitly need to do is to improve the jpaconfiguration option to not have to require to be located in a specific locationBut merging persistence units just feels wrong +unfortunately i cant check the development trunk +Small things +I generally recommend using Woodstox though I am not sure if that would help in this it be possible for you to attach a sample project that triggers the problems? That would make it a lot easier for me to +What is the scope here? I am not sure what custom filters are supposed to be and how they plug in +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Assuming the stack trace above is using DBCP DBCP is propagating an exception thrown by driver +The value add from the use of the component is that it encapsulates the additional administrative details +After deploy with admin api I connect to the version from the client with the version specified as +good catch and nice solution +The patch solved the problem for blocking case but not for the non blocking caseI managed to fix it for both using the given patch as the basisNow it should work correctly +Previous one uses PUT methods for the engines service +Has anyone else +It would be great if this fix could make it into that release +Please try with latest code. + +Done Done Done I have a bunch to initial setup that is not present in that test and they dont require itNeha Done It seemed safer and intuitive to create a listener to a topic after it is created Done Done the reason to expose it is for manual replica assignment +Refresh and merged on run you refresh again? The test is failing while applying the and updated the test result +I will attach a zip with a source pdf and the splitted pdfsI tested this with revision Adobe will remove unused fonts in one pageIf I analyse the content stream I can remove unused objects with pdfbox +Attaching a PR +Are we really going to do this? This touches an awful lot of tests +Will take a look later +EL expression for selectedTab is +Thanks Sammy for the fix I did not realize you had spotted this too +Patches welcome + +So I will suggest you to use trunk for this work +now has a unit test that exercises the interface +Ill also discuss the servlet registration on the dev list +The current one could be called the one without external dependencies should create a slimmed down that only contains jboss messaging classes this is essential to avoid dependency conflicts on the client could use Classpath annotation on Manifest to those external that will require stop scopping +Oops this was not fixed in Rescheduling for now +can coexist with in the same classpath because they are not in the same package +Are we doing anything in compressordecompressor? Do we also need something there +with of course +Thanks Jason +I just committed this +So what we should do is provide a base class that does the common things +Or else you will get the memory set your server location +Checked on Ubuntu x no problems found +IMO weve already deprecated them by removing them from the public API so the note could rather say may be removed in the future +patch looks good +Potential +When I disable my Clearcase Connection I am able to import the project +In answer to the second question yes this was done because of a cyclic dependency. +it is a bug in the jdocbook maven plugin id think and I will try to release a new version of just released closing stale resolved issues +Used in to override the default of The overridden value works with addr to check for Channel instancesThe overridden value also appears in the JMX stack parameters for UDPSo expected values are there +Hi Rayfeedback like yours is really greatly appreciated! I didnt mean to look defensive on my work sorry for giving you this impressionI look forward at getting your feedback on my last commitJacopo +Verified in BRMS as resolved in BRMS that used external functions would cause a NULL Pointer exception to be thrown in the guided editor +for us to recreate the problem? If you have a patch that would be great as wellthanks simple WSDL to illustrate the problemNotes This WSDL does get processed properly by When the faults for the operation twoFaults are commented out of the WSDL the service works just fine +Converting to issue with commit of +So all issues not directly assigned to an AS release are being closed +jj fileOne escape testcase is still failing Ill look at that tomorrowMike if you have time please review itthis code does not contain your patchI did it on the latest trunk +The checkin fixes that problem +We already have some reflection going on that tests for presence of methods +There are literal strings referenced in many places that need to be changed to match the changes in the patch +Added it to the added the and I updated the wiki to reflect the two new you Joost. +Is the available on JIRA? What would happen if I tried to suck all issues onto my computer via SOAP? Only the last might have attachmentsOtherwise I guess I will go back to the community and see if we can just import the old issues and attachments into a new project +Reopening +So calling notifyDelete will remove the line from the content store db if it exists +must have been a pebkac on my end. +Tentative patch against Rev +And it has no associated testsIm interested in understanding if logic improvements to either canopy or kmeans can be made but from this patch it is too difficult to understand what is being proposed +The latter is mainly for communication with Keith +Verified +havent looked at patch in details but touches only upon the code generation part so should be safe +disable major compactions I cant remember the last time I told someone to run them more often but just last week I told someone to either disable them or have them running weekly +Are you working on merging this? or shall i proceed with the mergingThanks personally wont have time to look into merging trunk into the branch until later in the weekweekend +Ive incorporated the info as we agreed Ivo +Missing options addedCould use some java samples using the fluent API etc +SEs expect the scripts to be available Apple for instance +lets fix that for after release. +Can you also change the configuration of the dev and users lists to add a footer with some minimal unsubscribe instructions? Such asTo unsubscribe For additional commands Thanks +Its been deprecated in trunk in favour of but this has been deprecated too in favour of methods on Job +Nice work thanks! And sorry for the delay in responding +bulk defer to defer to +At some point I commented that out +This issue was resolved with fix of. +I just want to implement something like that as a client API +I started the first two containers skipped the steps +Ill commit this soon unless someone objects + default multicast is enabled so other servers started but without loadbean deployment also join this cluster group fail over happens on these server making lookup fail +Create a new promotion +Is this still an issue for +Not sure if it is worth adding or if we are happy without it +on the patchWe should probably add an and image tarball to these tests too given we have the infrastructure but we can do that separately for sure +Right now ETCALL ETNET and ETTASK are the same unless taskthreads is set So ETDEFAULTTHREADTYPE could just be ETTASK if you want and it would accomplish thesame thing. +My openmeeting is configured with for me as expected in +However I am not convinced that there is no other hidden issue causing the print err +If we changed some API in component A and some code that uses this API in component B then Hudson fails to build component B +should be fine it last version build from trunk +Closed after release. +Ill investigate more once TC is back up +Bulk close for Solr +Pretty amazed this hasnt been resolved yet +Also provide approximate details of how big the output is number of mappers number of reducers etc +Linking to of the failingaborting tests have been marked ignore with the JIRA that needs to be fixed to uncomment them +Fix committed in r +What you say makes but for someone new to SI it may not be obvious +After applying the patch the ESQL sample now works for me OOTB and I have an Italian localePreviously it failed so the patch seems to work okPatch applied please. +Im going with the approach for now +The name with which you edit the wiki pages is visible in every page edit +Any particular reason this isnt ON by defaultIt makes no sense to me that this needs to be turned ON when it seems like normal behavoir for commandline utilities in any Operating system +bulk close of all resolved issues. +Passes store test for original error +Patch supplied +To much generification might irritate users to much at this pointSince the feature is not yet documented well we can try it out as it is now and redesign it forcheersRainer +At least we could reproduce the problem of in tomcat by setting the log level to debugWe had feared the fix of did not solve this finaly +No QA needed so skipping QA QA required closing this improvement +All tests pass expect the one has been mysteriously failing for me since the last couple of daysOleg +i think that you have security settings or policy settings in IE that is a bit to restrictiveall examples i test for IE work fine +I checked in a fix but this is IMO a workaround for what is a bug in the core +Tests passed cleanly for me on +I am guessing the ns come from the freemarker templates? At least the templates also use n +This could be costly and think the behavior we want is if the schema exists use it if the schema does not exist behave as it behaves currentlyIf we get rid of the schema thing and instead always only look for a +If that will work then the description of this issue should be changed +Closing issue. +This patch broke data given by Ganglia so I just could filter data only for Ambari diagrams +Three simple ideas for fixing this +Sure one compiles on a trunk version +Tim are there a combibation of older versions of AS Messaging remoting where this bug doesnt appear? thanks for the very quick fix +Looks like user defined compaction has not been working since data directory structure change inPatch fixes this by pointing proper directory based on given file names +clustersrcetc +the processor could first check if input is already cached and if not then notify to start fetching +No objections from me. +I think both issues have been fixed and I have deployed a SNAPSHOT build last issue might be related to the problem fixed in Rev +Attached make patch to remove from exclude listIm not sure if this should be applied as test test case that test contains fails for now +So thats fine no revert wars hereThe question is revert or disable by default I think +Container +the only quick fix for this is NOT to remove the temp behaviors but to let them be disabledbecause i dont see that we fix this differently in the branchmaybe but in matej is busy with thisWhat we maybe also could do quickly is generating a hashkey for all the behaviors and use that oneinstead of an index in a listThis key can be as simple as or but it is stored in the behavior +lack of time +Yongqiang Can you have a look at my patch and give me some suggestions on how to fix it? Thanks +I have upmerged this patch to deal with it +Wildcard problem is fixed by accepting multiple args +If the patch is ok can some one commit it +Excuse my ignorance but why does this matter +committed. +Im running derbyall now and will post the results when it finishesderbyall on Red Hat Linux with ibm ran with no new failuers against sane jars built with applied + +back port the in EAP noted inOn injection of a jbpm to a Seam component the process instance returned was registered for +Thanks Suhas +Both are fine +Latest x patch is uploaded a bug and attaching new patch +what would be most bang for the buck +I just the compilation to see if my unix shortcut problem had been fixed but now it doesnt even generate anything +Thats right +Attached was very noisy so I replaced it with one created with svn diff x contribmiscellaneousIt is relative to trunkA test is added to test that verifies exactly what are the values of norms after modification +trunk Committed revision +Hi PhillipJust to be clear you had the FORWARD dispatcher defined and the jspforward tag still wasnt filtered by the +What to do with versions +If the fix is ok for you please close the issue +The standard public API for such things is not +Thanks again John +But are those existing applications relying on questionable behavior? Id argue yes + node is null has appeared on page with closed be checked in selenium testsall methods of should be checked in selenium tests separate test to two tests +Tests fail without the fix and pass withWill push this in ifwhen Jenkins blesses +The only way that this will be be fixed is if you provide a pull request +Id rather see how this all comes together in a bigger patch rather than these little patches makes it easier to get an idea of how it all will work together +In such case I need to remember to the strings in the set and then to the strings before checking +the artefacts filter was removed. +So Ill close the JIRA +It wont involve a web browser so should not use selenium +Having said that there is always a choice of using a custom helper for to string or to byte array conversions where additional validation logic could be implemented +There is an example that takes a regex over rows +Yes will look at the doco this is still worth looking into as it is probably hiding something else +thanks! transition resolvedfixed bugs to closedfixed +Sorry about that +These should be updates early in M to allow for correct component creation +If the value is null then the appropriate sql is column is null +this line is in the static initializer for the class so it will typically be called once per run of the right + you elaborate on what you need default handler for pointing +get permission restrictions are too old resolved issues +Hi AngelineIn your setup added the below egress rules rules for network adminsrx +We would need a way of triggering teiid to update the model based on commits to the repository +or the token +Yup +I would first check with perhaps marketingbrandingI will point them to this issue +I think its odd that we have to do a full dist to get all the necessary into the Seam any event Ive updated the docs. +This will be called at parse time not run time +Tested with on cloudservers +For built in shipped gems like rake and rspec I added generatebatstubs to the end of maybeinstallgems +If the defaults in the file are not fine for you you should first change them and then build the war and deploy the war +This is not going to happen unless and until we get a clear use case for it and some in the spec +Marking as Seen in production because of the duplicate issue. +However we need to track the fix in there too +Yes I added the code +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +stupid error in warning seems to be from other build +Maybe the correct fix is to update the parsers to generate the metadata before they call endDocument +I have checked the content +richcolorPicker VPE screenshot was atteched +Observed the bug on a build made on July th +Your patch looks goodCould you press the Submit Patch button? You may press it once you think that the patch is patch +bulk close of issues as part of making release notes. +This gives them a control which they are already using to have a boundExplain plan can be enhanced to show input size for root tasksWithout going into the details there needs to be query stage specific variables for number of reducers which overrides everything +Your patch here should be applicable to that work without hassle because it is pretty much the same codeIn other words its gonna be committed soon too +I rearranged the code from the existing readers into a new set of readers that the schemas to allow for much faster reading +ve added setsetetc methods to as well as an with Oracle JDBC API types +I was about to add the chek of the vm state in the code +I came to the same conclusion after debugging and looking at the code +After the change I didnt manage to get any of them again when rerunning the same sets of operations so Id say yes close them for now +Oops has attached wrong files were missed from the previous patch +Also I dont know how to run Accumulo now +In any case we probably do not need two dummy Input formats in the project +Okay in camel was on a wrong foot with the configurationApache Mina has distint configuration for acceptor and connector in Camel this was the same configuration +OK heres a follow on patch from me +Weve also run into this problem +But try it out and see how it goesCharles the build server has just finished the build +Proposed thrift changes +Ahh thanks for bringing closure Nick! Although Im a little confused how removing files from the index while readers are using it could lead to those exceptionsNote that its perfectly fine to pass createtrue to IW over an existing index even while readers are using it IW will gracefully remove the old files itself even if open are still using them + +Committed to branch +Committed revision +hi peterright +Thanks Oliver this is really +These are web resources that are different to normal classpath resources +And its supported in iBATIS which is ready for Beta any time +Theres an issue called Java Codegen for X Structs +Ive run the sample application with Apache Flex and I cant reproduce your issue on OSXSafari and the popup works and theres no extra characters displayed +First Iteration BYON still not compiling and not tested + prevents by extending the condition for stopping tree traversal in to include an exceptional case not considered previously + has been marked as a duplicate of this bug +Thanks EliOne thing though is that there seems to be some formatting issues +Second iteration of consistency testing this one keeps polling even after the file is consistent to see if it stays attempt at a fix in this to reupload patch I dont think I hit the right sequence of buttons the first time around +I am not against this fix +Heres one more iteration with s binary methods restored along with tests +Making your app production ready should include reading some more docs as you normally would like to customise interceptor stacks switch devMode of and suchPerhaps we should rather think of giving prominent hints on devMode state in initialization log output and maybe on the sample of blank app +Remaining work will be committed in +Oh right I forgot to add that it would be very helpful if the util packagecould provide some adderror functions so that one could registercustom error codes and messages +bulk defer of open issues to +All the tests pass so I assume this migration succeeded +As usual Wicket provides a solution +Applied last remaining applicable fix from the provided patch +Can this be merged in? +Ive just run your test and wonder how to be able to repeat thisagain and again +for the to trunk. +No answer from the user. +Thanks +yeah this is annoying +This issue is resolved although we still have no good way to include driver so that plugins can see them +There is no such thing as pure Hibernate anymore +Perhaps you would like to put it on the wiki +All of these classes now have a test called test which covers this requirement so I think it is safe to close this one. +Propose to close it is assuming a JCR based implementation which we have dropped now. +It should be placed into the bin directory of castor +Verified with JBDS + +Typo +code will make the warning go away but just does text formatting +Thanks Nicholas! +Connectors report DD errors as you would expect so I guess we need to get the web deployer to do that too +We may need to reassign specific jiras to a Red Hat employee based on priority of theMike C +Thanks Strong +Bryan The patch sorts fields after the last one +The stateful session bean extended persistence context feature is not complete and is being tracked in issue +So that we dont force the sm on people. +Patch which eliminates the race condition on the lock test +Dont forget to delete ewshomeworkCatalinalocalhost directory after making changes to +Please review it first +You can make sure that I am the administrator and main developer of this project +Currently the tag only escapes the most common characters +It needs more work to be a generic test of concurrent queries in LARQ +Anyway I understand your position +We need to focus on windowsxp first +but there is problem is that when open the Add and Modify Projects dialog there will be two modules for every ESB project since the ESB project contains two WTP facet and its treated as two it seems is not ok for this issue + +Thanks Aleksey +A new Bundles set has been released +Assign to Shane he knows what is going on IIRC +patch doesnt solve the rpm build process is replacing usrlocalchukwa in two place +please review this patch +but how about Banten +One question +Please report here whether the workaround succeeds for the source for PGAS is not available so you will have to use the binary distribution +Your basically saying its ok for all threads to wait sequentially in several areas of the code during an in memory readonly index search +Fourth possible solution is to configure rewrite rules on httpd to handle wrong URL to improved integration between JIRA and there is now a Source tab available which performs the same functionality as the Subversion PluginAs such we should investigate whether it makes sense to remove the Subversion Plugin altogether +I have verified that this issue is fixed in +bq +Someone could still chmod something in the trash and get into the same situation +bulk close of all resolved issues as part of closing items. +bq +Acceptable imho +I will verify thatThanks again folksVishal +This also applies to +I have added a simple test case for backward compatibility! The tests really cover almost everythingNot that I think that it is required but looking at this maybe it would also be possible to add a forward compatibility check +The changes in this patch are that I have added a new test lang which does basic testing like aattempt to update statistics of a table that does not exist battempt to update statistics of a index on a table ccreate table create index load data in table make sure there is no statistics created for index because there was no data in the table when it was created create another index it should get statistics created because there is data in the table now perform update statistics on first index and it should also get statistics created for it dissue internally generated ALTER TABLE +Im not sure how to implement the compaction optimizations the main JIRA description mentionsPlease provide feedback on what needs to be addedchangedThanks +Note that I still have random failures with I am not sure if removing the clone will change so much +How is this issue related to Buildr? Could you elaborate plz +The way it is setup now the alttype follows the constructor calls ans is set accordingly +tag youre it + to trunk branch and. +Done +Every node MUST have a unique id including those just connected by message node must have a unique irrespective of whether you are using clustering +Wont the individual slfj implementations then expose their own +I also verified that this patch still covers testing of +Are you sure that is correctI just checked and from the download page requires only while from Maven Central requires as reportedNow both from the download page and MC requireThis is fine by me but I wanted to holler anyway since it may have been a mistake that source and target have been set to instead of. +My welcome screen looks like that in Win x JDK x +We can work on speeding up MTTR elsewhereChange Default +read lib is I came across this bug while upgrading via maven +Im not sure if there is some setting on the JDBC driver that needs to be adjusted or not +eg +Otis please let me know if you want to be pulled into a future release +The platforms that used to work need to keep on working in addition to the new platforms added +Implementation of the proposed parameter +Just to clarify I think its ok for the archetype to extend from jboss parent its the generated project that should not use +thanx +Is there a design reason we do not have log methods that take a log level in the Logger interface? Seems pretty useful +Reopen for backport consideration +smileI just committed this. +Since version you can also prefix a ruleset filename with file to designate that it should be read from a filesystem path rather than the classpath +Improving support for Complex types and elements +Not fixed +It does look like the same issue though. +All! +Oops somehow I pressed a key combination which created the issue in the middle of typingAnyway I think I made my point that +Please open another issue for the +I see what you mean and am trying to build my projects this way +Its uses include undeclared errors a base class for declared errors a wrapper for error data in the generic APIWe might keep for the second use add a new named a new extending thats used by the Generic API +Made a new JIRA issue for that. +I dont think we should do this until after +They are represented by two different constructs in bpmn + +tnx for input +Must be verified using a postgre database +All if anyone would be willing to grab todays release and confirm this fix wed sincerely appreciate it! +The following sentences Subquery materialization is detected prior to optimization which allows the optimizer to see a materialized subquery as an unknown constant value +And if people complain too much we will revisit +I think the HTML report should include this information +I update the patch file which should solve this problem +Looks like we should also create new listssecurity and announcelist with zwoop the moderatorowner +The file is what Ive come up with to resolve this issue +Priority changed to MINOR because a user should be able to work out why dragging stops working when you move outside the Shorter +what version of nutch are you running +Moved code from a test to general setup as it was needed by more than one test and broke if tests were run in oposite order +hi I have the same problem how to avoid it? Where to configur with CXXFLAGSmarchi +Dan thanks for catching that! I should have thought more thoroughly before applying thissort of optimization suggestion +I cant reproduce this error locally even after getting a clean workspace from svn +Text looks good Sijie +By default people would expect that they can see jobs in the completed list for some time after the job is finished +At this point no new instances of are created for this session +How do we proceed from here? Is there a committer thats willing to look at the codeI have commit rights but Id like to find some time to add the benchmarking code first and also trial it in a live environment +Closing as fixed please reopen if there are still issues. +ok will do +hmmm +Committed to trunk. +Unassigning due to inactivity +Slack parameter code path is removed now. +After waiting for more than a week I think that there is low chance of getting a fix change from +Verified on latest master build +This has been done and its called closing all issues that have been in resolved state for months or more without any feedback or update +Removed Feature and Property components from Woden +Im still having difficulties reproducing the test failure after modifying the test to fail if it doesnt get a timeout on the first attempt +still missing +Moves Cell pb back intoIncludes aAdds a primitive test harness to compare the old school KV serialization to manual Cell serialization to pb Message serialization of Cell +in. +Sorry for the spam +Cant we keep the same as we had before? +Do you have the full list of failures +I suspect the problem is that it gets deleted before deployingIm a little confused +project +gitignore consistent +Patch that fixes the issue +mattdiezActually in just about any time series application usingIE or Firefox this problem comes up +I review the code and ensure that all versions have this problem +Im not an AWT expert but probably you somehow have to configure the used to render the text correctly +Note that when we run the +If future tickets can you leave the status unchanged and just set the patch available tag if a ticket is resolved then committers will not look at it as this state signifies there is no action needed +Any plans to add this in +I basically know what you mean +bulk close of wont fix unreproduced and duplicate issues +That is the problem occurs in all components regardless of whether the developer is using spreadradio for their layout or just the default rendering for +The testcases depend on via git diff r aaeccaafafdafaaafbfe qa is messing up probably due to dependency on ? Will mark as available once that is committed +committed +I forgot to svn add a couple files in my previous diff +Maybe that Luc can give his opinion as to whether it has to be so +I made changes to it as suggested + to trunk +It ran fine +should we defer this to or? WE need to release on mon to +Attached is which is a straightforwardconversion of into a patch toI think its valuable having these tests even if we decide thatthe current behavior is current and that we dont want to change itIn that case the tests will serve to verify the desired behaviorAlso if someone should experiment with changing the behaviorin the future having these tests in the system will draw peoplesattention to this JIRA entry and thus to the discussion and commentsin the issueSo I propose to commit the patch and resolve the issue as invalid +Would anyone want to try this +Please give look at attached most other place where a list or table of things have operations you put the buttons to the right of the table +bringing back to b since it can cause build failure when user have annotated +bulk move from M to on +It was defined as a convenience for Mappers before we provided them access to the directly +If so how does it relate to +I think the default master file for needs to be updated +The only interesting usage of this return value is accumulating the number of bytes read which definitely should be stored in a long +Commited to trunk +Rejecting +bq +Patch which fixes the issueI hereby disclaim all rights to this patch release this patch into the PUBLIC DOMAIN +output from failed that this affects only WFK so we should patch it just to make sure we are consistent. +test on master if possible +closed as resolved and released +These are the same files again +You should set up FQDN on the hosts so that hostname f shows the FQDN +is because they will assume the user knows best and try to use the already set HIVEHOME location +do some crud operations and watch url in jessionid remains the same which is what i would expect + +Switched off behaviors wrapping for in M and added diagnostic warning for this +Weird it doesnt for me? Ill get some more details +Closing released issues. +This bug should still be verified seperatly as fixed once the patch for has been commited due to runtime environment differences +Slightly modified patch was applied with Christines suggestion to move the Contact us to the bottom of the left side menuThe patch was applied as was the new part + can be extended with without pain +Closed without integration test. +Remove the set stuff and just set it in the loadI will likely commit tomorrow +Setting aside Hibernates suspicious claims to using POJOs I think that actually my preferred solution is to remove the call to toString but it sounds like that will meet resistance +Nicks feedback +I also forgot about the bogus values set by +Thanks Steve +We are moving to dotted group and since Ivy completely leverages the Maven repo they generally dont +Not an issue for me any more but please comment if you can still reproduce. +Checked in +I committed this to trunk +You will still be affected by this if you are not using the through +Im not sure we need streaming or whether calls are sufficient +if there is no one I would create one for ittwo comments about the patch +To be confirmed +With the same priority it will be up to the installation and eclipses extension manager to decide which is truly a higher priority and which gets called first +Any suggestions as to how to overcome this before a proper solution could be implemented +I dont think that is going to be portable across or Locales +We already have a little since its quite unlikely that snapshots of across multiple replicas contain exactly the same updates +This is a four year old Bugzilla issue against Xerces +Is there a place where proposals are collected? Maybe a wiki pageNot yet +NeoIs the problem specific to super CF or does it show up in regular CF too? Also have you tried flushing smaller amount of data to disk +Thanks Mark +If you get any ideas for improvement please suggest here +ok thanks ill assume it was caused by some transient issue and the build scripts are ok and Ill go respin the RC artifacts +screen of which will be fixed eventually see there. +I dont think the problems caused to distributed caches by itself is a good justification for preventing caching of an entity +Attached a sample test that fails with the previously described error +This is still publishing hcatalog artifacts in separate namespace +I can vouch for this person +For example If an project use the pdfbox and do some cryptographics and stay at bc with this acutal pdfbox no problem will occure +The patch looks good for me +Issue may already be resolved may be invalid or may never be fixed +Triaged for repro attached and setting normal urgency. +except that it resolves against the target platform instead of the source site +Committed the a patch to trunk revision Attaching a new patch which removes all usages of Vector in and replaces them with +put a temporary hold on the Cache logo design +However on a monitor you dont have so many changing areas that you can see a difference at all even at fps +Committed the fix +Paul Yang this looks good to me +This has been raised to the OSOA specification collaboration +Moved the scripts from different repo to cloudstack +Fix for failure +I managed to forget to add the operations folder when comitting the DSL refactoring +We cannot maintain issues fixed in multiple places +There is some issue with sorting now when a group sort normal sort has null values +No I have no more deadlocks +According to Joe is fine +We are going to get warnings in the idltowsdl tools when a user chooses to split the generated wsdl into logical and physical wsdls if we want to validate them +has been by mistake +Im a little bit confused about tyour keepContent flag +Thanks Ahmed! +Verified that the instructions are still the same as on Nicks screenshot +is mostly a I can changes in for this event can be undone +In my application a work around is difficult as the queries are dynamic and I will have to do a lot of rework to control the size on the in clauses +To make matters worse if I shutdown Tomcat and restart again The same messages will process again becuase they were not deleted from the DB +All are now reusable. +Thanks +So Im marking this one as as closed and will stay there to remind us that we need to provide some configuration for the otherSRS list +DRLVM file contains the mainpage after applying +Is there any other I dont know +I can provide a patch next week +Core tests are failing +Amit Could you elaborate on the remaining issues with distributed search +There is a better chances of developers taking a look at it if it is already in the trunk +If you have a better patch feel free to attach it hereOne of the interesting specs that have come out is the annotation based validation support +INFO Running for rows took ms +Can you add your first comment to the class javadoc so that users do not have to read this jira or whole code +Im inclined to table this issue until we see whether that leads us in a useful direction +The final version will have one +When application thread calls castMessage underlying plumbing in sends a which in turns obtains a lock held until response arrives +I believe referring to a specific version number is intentional I think the packaging plugin is intendended to be relatively independent of the bleeding edge code so it is trying to refer to a specific known version of the kernel and not depending on the latest and greatest SNAPSHOT +Thanks TedNew change addressing all your points +awesome +Im going to mark as resolved though please reopen if this does not fix the issue. +Can you please attach a standalone example of this problemFirst glance at the code looks right +For installer that contains EAP No I already have it installed should be just No If No option selected in EAP installer step Server List should be still available to configure additional servers +Sending fileinstallsrcmainjavaorgapachefelixfileinstallinternalTransmitting file dataCommitted revision Please open a new issue to discuss the other problem +Yes I think this is actually a typo in the spec +is back up closing issue +sorry +Thanks Heshan +I will be working on that nextIn the meantime if anyone has time to reviewcommit the I welcome the feedback +Of course +updated is done +Ill leave open for the time beingColm. +What is the package name that you give to PA +Mallikarjun +Quoting from his bug report Neither winzip or gnu unzip can open thearchive anymore +Works +The problem comes when we run the really same thing under sonar analysis +Retains all method signature the same +thanks +It would be great if we could merge the functionality of and as a part of this migration since having two is confusing for users +Resolved +I think solved this +Hi Arun If you have alternate ideas please feel free to post +Patch attached +Should be good once we have a clean build from buildbot +Setting isnt saved and after editor reopening its default blocking was solved in M but the fixes where at first not included in JBDS +It only means that if there is a bug in JBM at the place where byteman throws an exception and this bug causes an exception like byteman does the message will be stuckThanksHoward +Finally done at r +Committed to and trunk +Ok using entities I was able to duplicate this pretty quickly +If no kerberos information is provided the script will automatically setup as cluster +Fixed with conversion function for cast a type parameter type to primitive +I believe this is fixed no in SVN. +This site is for reporting bugs in the Apache Roller softwareSorry my mistake +I will change the release accordingly as well. +Could you put a patched jar on a web server like Abel didI hope Abel will have the time to reintegrate his patch with the lastest SVN a lot of people are waiting on this bug +It might be a good idea to run a DFSIO benchmark to validate that it does not impact read performance +But seems most parsers are lacking a good regex feature +I am in agreement with this as a content provider I would not want integration to go so far as to cloud or obfuscate the systems in favor of +All created files showed there it looks as Package Explorer issue +The in the does not close the dialog immediately +Make sure you really have Weld in your JBOSSHOMEmodulesorgjbossweldcoremain +For me the question is how many people are affected by a issue and is there a workaround to handle itmicha +and future qfile tests will be addressed in future jiras. + +As already reported all CTF tests pass successfuly +index directory andor clear out the db? Im not sure but the old index might have caused the indexer to break +The patch did not handle null values correctly on priority values +the file that we are reading b +I will deploy a jre and see what the differences are +Other than that Id like to on the patch +dupe of +It seems this patch is for +Wondering did you make any change in your checkout that might explain what happened on geoArne reported also cpu usage which I think I managed to reproduce by using a very big layer and cutting the connection mid air +This just adds up and once we will reach codebase just for workarounds +I just asked a colleague who is running a Linux distro and he had the same failure +Have you narrowed the problem by trying it on one slave and then a different one +As always your explanations are much appreciated +In fact we are already doing so on the project Im currently working on +Then well need to document its use on the wiki and make it available to people +Uwes patch to reduce bytecode for the readLong methods +i see we would have to negate the bits maybe we should pull out s anon scorer into a separate +It doesnt work on my windows either +I fixed it added some more tests and committed. + +The client program can then just be launched using the Tuscany launcher as describedIve also added a sample that shows the same pattern applied to our calculator sample +Hi ClausThanks for your comments and input +I have tried to disable the and it workedI have tried then to set the to version and and the to and it worked for me tooThanks for the hint +We might use a more advanced algorithm in the future that displays the output in a manner that reflects the time granularity but for now seconds is the most commit feacaaeeded. +Ill attach the new versions of the source files too +the the actual problem is in the call site handling stuff because the signature with is right +I will add a link to that one +Id really like instead to make our job easier by at least combining the framework main jars together into one target first along the lines I described above +Is there any disadvantage of running it as a separate process? If it runs as a separate process then making it a plugin is not necessary yes +Assuming that the problem reported went away +Thanks +Updating the patch Eelco noticed that if you were doing a removal the code should never reach the switch statement below +Makes sense +The onlyproblem coming along with the patch is luke +Looking at it +Appears to have been a network problem +is your tool simulating separate sessions for each threadEach thread have its own session +testng runs junit tests not junit +Look at the linked issue for more details +The static is return code +never mind the issue was caused by a version conflictjackrabbit code was used to access a JCR repository +With dropping soon and this only existing in a branch Im guessing wont be seeing Groovy config? Is this a sign that it wont be in at all? I only ask because Ive been asked about this a few times in the past couple of weeks and am not sure how to respondand honestly Id be keen on doing config if it were part of Spring +ToddIt is very difficult to bring out race condition between data nodes and activestandby namenode with the current test infrastructure +When the source filesystem is a replicate does nothingI just checked in a fix that will hopefully avoid replication collisions +The difference in size is M vs M +Took a whirl at this and after running mvn assemblyassembly in the directory I got my jar +Hi FernandoI created jira problem of the archive is that its fully supported feature of and for this reason should be signed +All the tests passed +Were talking about containers reloading an app so all those references should be disposed by the container even if they were being held in error +So how do you distingush between the case where a master comes on line and the RIT is OPENING but NO exception has happened processing +Will review +Ive also committed to . +This causes the Weld CDI bean proxies to blow up because it needs to use the correct TCCL to do its thing +Will be enabled by +So far Ive come up with the following A connection uses k and is limited by the number of sockets A session uses A consumer uses A producer uses this sort of info is vital for boxing or hardware provisioning +Maybe temporary resources shouldnt be handled through at all +Thanks TimThe fix looks good +Can you do us the favor of reading the DTD and finding anymore errors? Thankfully the document is short +Im assigning this task to you since youre already working on it +closing it +I just finished moving logging Ill figure out the rest and move those forward as well. +Corrected compareTo and simplified hashCode implementations +patch attaches the license header to the following java classes +Not sure how the fix version changed errant key click +I just committed this to trunk already in +Moving this issue to the release since the release will be a simple release +Please debug the root cause and update the bug +It might be worth keeping track of the specific involved in the failed compaction and skipping those + Randy! +Attached patch resolves this issue by not attempting to the generated context file +Is that important for anyoneThe change in was to ensure Jackrabbit does not shutdown Derby because Derby is a separate bundle and cares for shutting down itself and also because in such a setup it is not Jackrabbit deciding on whether Derby is still used or notI would assume that for the pooling Jackrabbit just does not care about shutting Derby down or not +This is a very trappy thing and we should catch any violation IMO very quicklyI agree its trappy and its great to add this checkIm simply proposing moving it to less of a and I dontthink this will affect how quickly we catch violations but shouldreduce the cost of this added assertionIn fact I think find isnt great in this regard +Looking at all the references between the dc and store there is nothing shared in the map and doesnt make much sense as to why this occurs +The Ant task is now done and has an Ant build that uses it in combination with Ivy +Website doneOnly item I know isnt done is putting a note in the downloads +Here is a patch to try out to see if the server locks up as reported in the issue +There was an interesting conversation among MINA users that it would be great if MINA provide session groupingWe could rename to and provide more advance features like that +Now when clicking the recompile button once all have recompiled the DSLD box will automatically refresh itself. +make it more distinguishable thanComments are welcome +We ahve now have entry and exit trace +I cant see any obvious flaw in that argument although there might occasionally intentionally be spaces in fields we should at least be consistent if we are going to nuke them +Dupe of. +Viktor please check if it is a problem of Tools XML editor +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I see region offline exceptions and during this new upload +Could you send me the tomcat you use so that I can diff and see what are the differences in my environmentMoreover please close this issue if you are sure it is no longer relevant +It also throws a exception on a timeout +Hi +Examples should show any differences also +JgrayIf youll note there is almost an order of magnitude of context switches when run on the same hostOn a different host we are hovering at about cssec +Any progress applying the fix +No worries +But on the other hand perhaps simple cutn paste approach would work have for stax implemented in with suitable license copy it from there +Thanks Viknes +Ill work on that over the weekend let me see if I can have that done by Monday +Here a new patch with test case for the positionHistory +Further adjustments to reflect advice from Uwe +I think I may have found a bug +Correct +I will go ahead and synchronize the getter and setter for the variable since it is not final anymore +modulesall produces anotherIn addition to that the root POM also produces anI dont know what others think about this but I believe that this is definitely too much +Im a little nervous about applying this patch especially so close to release does it effect the ordering of items returned in the maps? +Do you think you could provide a patch for thatMaybe use a wsa prefix if none is provided +Its possible youre leaping to conclusions by assuming its NFSCan you boil this down into a simple class test case that initializes Velocity successfully in environment and not in another? Then attach that class with your and we can take a look at it +the codehaus ssh server is down with no scheduled to be up I can not deploy status on thisAlso can we get a release of this out? I need it for and Maven is not very happy about releasing anything with snaps +Simple fix +Patch is attached +I need to also either create multiple instances of Listener or make the instance in the Session wake up all Continuations +bulk move of all unresolved issues from to +Here are the binaries of the samples from Sun +Is this not fixed in trunk +rebased +Hmm +But insert the method that handles is the wrong place to do this +It seems to work at least for simple cases +Assigning back to Van +If does not solve this issue will have to be implemented but its not as straightforwardSo what can be done for the moment? Clearly the minute lock and replication timeout in the test were exaggerated but it helps see the issue better +I am updating the patch here and hoping that this will be committed soon +I believe I have properly filtered the extra events that could occur in rare cases +JOBID and STEPID columns renamed to JOBINSTANCEID and closed as resolved and released +Guothanks for your hard workI am reviewing your patch and hopefully I will commit your work asapPlease see the following minor points there are two patches for frameworkwidgettemplates in your file which one should I use? I have noticed that you have added the request and response objects as input parameters for the class if I am not wrong they are only needed to render hiperlinks for html this is ok for now but we should find a better way of handling this because I would like to make the as much webapp independent as possibleJacopo +Many thanks and for the fast turnaround. +Patches applied. +Marking bugs with invalid fixed for release as fixed in. +If the directory is empty it may generate a job with no input splits and thus no tasksAside from that looks good to me +I would actually suggest that this plugin is removed since is superior and having two plugins trying doing the same thing will just split the community +Changed Target Version to upon release of +pull requests merged +Committed to trunk. +Please see Ross comments on Jun + +checked in +Apologies to David Graff +just committed this patch. +This bugis also reproducable with Tomcat +What does this mean if anything for Pig and +Thanks Ashutosh! +I am about to add a general search box that allows you to filter the displayed names +Ill leave it to your discretion whether some kind of test is worth pursuing +Committed patch as is. +The problem here is that the Errors interface is centered around this notion of an error code that gets expanded into message codes through a +bq +For the header the metadata name is + +for committing +Hi DeyanThis works OOTB so I guess you have made some configurations before could you please detail Jacques I havent done any special configurations +I dont see what we can do with it in the source tree though. +With my fix you need to specify a quality of so the image is not compressed +For example start the server with long or v or vv arguments +Committed to trunk with revision Closing issue. +as I need a new release of cdk for this +I misunderstood one of the previous posts +Thanks for your help +bulk defer of unresolved bugs to +Patch I applied which removes my catch of added by the first patch attached here and instead Ignores this failing test +We may be able to do something with the way the states are saved but I dont think that will be a magic silver bullet +I am back to Trinidad after more than a yearGoogling for solutions and still finding the same old bugs I feel right at homeEvery decent web GUI framework works around the IE select issue using iframesHow come nobody is interested in fixing important bugs +Main source excluding Davids +IIRC the deployer has hard coded + +Please review contents of group but the UNIX group has been populated. +Thanks Steve +Again this might have unexpected side effectsHow can I test if Clerezza still works as expected? I myself use only a small subset of clerezza +Removed the private tag in the asdoc comments +Look for example at Glassfish they use MavenIt would not be that difficult to mavenize Hadoop +Ive written a good chunk of code for this but there is one big issue with this Im not sure how to solve best this will break isolation from a user perspectiveMore precisely the goal is that from the user point of view all of this is transparent it does a query and will then fetch query one by one and the driver will use that feature underneath to having getting into OOM situation and to avoid timeout for query yielding lots of result +No unit tests added as there are no addition or code changes +I guess it works as always +I have fixed this issue in the revision Now all albums with out owners will be visible to everyone +Merged +MEAD has been modified to only sign the contents of zips instead of the actual zip +Migration issues are mentioned but are they API differences? Is there anyway we can point to a document that points out the differences instead of saying Hey customer you have to find out yourself +So I think we should use jboss as not so raise that on the jbosstools mailing list and Nick or Denis can let us know how to get this up and running +The attached patch accomplishes the initial merge that we need to enable further testing on WindowsThis patch consolidates the work of multiple contributors on Raja Aluri Sumadhur Reddy Bolli Ahmed El Baz Kanna Karanam David Lao Chuan Liu Ivan Mitic Chris Nauroth and Bikas Saha +Alternatively the entity class could help us I am having a little trouble replicating this +The mirror is now available at git Jukka! +Testcase necessary to reproduce this issue +The specification recommends that any request scope values in the tag should be reset in the method doFinally +I didnt use the bean declaration mechanism primarily because I hadnt looked at how that worked when I wrote that part of the patch +Improvement in existing code and service logicThanks to Rishi Solanki and Sumit Pandit for help +Basically done +Understood BinglinI dont think there are implications by checking if a GPL library is loaded or not but not being a lawyer +Reopening +This will still make the code go througha code path different then if no timeout was set +From what I have seen previously entering the realm of code sharing between the two drivers can result in quite a lot more work +I was always wondering about this inconsistency +Updated patch post changes +this module can easily be moved as it has no naming clashes heavily with the module that we distribute with +There was one less Assigned from earlier entry in the log than there were failed map attempts that received containers +Committed revision . +Well +LGTM committed +If the original poster has concluded that the leak does occur in production ignore me +hibernatetoolstests +The patch looks good to me +I buy the the users app then thinks its a valid name and other stuff breaks argument on the patch then +Arguably this is a flaw in the filesystem but addressing that is a different discussionAll this issue does is make files visible in the namespace as they are created and while they are being written +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Very nice reads +My feedback on this in addition to fixing the bug mentioned there is quite a bit of refactoring done here +If you submit an updated patch I will look at it and commit it once everything looks ok +upgraded the bug back to critical as ctrl and tab keys are important on console operations +Hi Ondrej whats the progress on this? Did you get the SVN link you were after? Thanks +building against hbase +Thanks +Thanks Roman +If this occurs they need to fix their codeIts useful to know which thread triggered the connect +Closing all resolved tickets from or older +tipping long grid inverse grey circularc +Distcp Guide to be updated via Dhruba we already have for updating distcp docAlso thanks for reviewing the patch. +Attaching updated patch which was got committed +If we add that the split log worker daemon thread will die and thisregion server will not help on distributed log splitting any moreHere it is normal to terminate a log writer thread since the task is taken bya different region server +Just my cents it just needs to show that some number of variables are defined so when users uses nextback does not have to click around on the wizard to know if they actually did fill out the right stuff and not having to click on several buttons to see so +So its still needed but well revisit the patch to see if its still valid or not and get back to you +WidoHugo Wanted to check if you would be able to fix this defect +released closing. +how do you think thousands of other people use wicket with databases without messing with repeater ids? or do all wicket apps out there suffer from concurrency issues where pressing a delete button deletes the wrong thing +A is not bound to a request and thus using it for the factory sounds wrong +worked around by r in go in the next modcluster version +I did something similar and it seems to work +Does this break anything +If nobody complains Ill commit the patch tonight +patch +Attached updated that Blah +We changed the configuration since then and we havent had any of these +Done closing. +Reopen the report if you feel I am wrong +for removing them +See second screenshot +Andrew I suppose you can fix this youself +I took a look at parsing the response etc +There are lines changed in comments that have no discernible change +Trying Hudson if the old chap is alive again yet +Looks good to me +Everything would need to be handled by overriding the segment readerDrop the current back compat code for its mostly unnecessary I think +Changed the patch to include license and more decent error handling +I made also extendThanks +Initial just committed this +Java Datalog wrapper package that deals with importing logical operators into the datalog library +This now works perfectly in almost all cases +An easier feature might be to automatically set up the proper codec when reading the file extension as input to a job +At this point its just missing pagination and having the query criteria actually subset the prod list +This isnt windows specific right? See +Just bc you dont find something useful doesnt mean others wont +Thanks +IMHO deprecating that stuff is fairly clueless and adds clutter to our code base +They are not exactly the same issue but Ill attached a new patch for both issues shortly +and by shutdown of course I mean kill +At this time were not willing to get into a situation that requires platform specific testing of scripts +rev fixes seems that before this commit we couldnt deploy anything from the agent +Check the ASF bugizilla for that +If Im using Solr and want to feed it dates in a particular time zone or perhaps a of day and clients expect this then why should Solr force me to specify a timezone? I find it irritating +I dont think you can relax the conditions around distinct since that involves hashing the input across reducersYeah thats right +The abstract contract in wsdl schema is not effected by the choice of endpoint a consequence if the contract permits an empty message the endpoint should be able to consume it regardless of the coice between bare know we talked about this before but I forgot what you said +I agreeThe position of these buttons let think that it only applies to the Applications ans containers panel so it will close this panel if we click on the close button not the whole pageThese buttons should be at a more global place +moves code out of the constructor for +Brett this seems to stem from the work on +Will start adding more and more tests to show the same for all other artefacts described above. +I am unsure about the status of this issue +Hi Lionel thanks for closing the loop. +Kev waddya think +We may be able to solve this with optimistic locking +If you think this is a hack or something make sure to inform users about it by throwing something more reasonable than +Furthermore its not just the greater than operator thats affected +If you are in +Thanks +Yes an hour is way too long +Sorry meant to close this yesterday +Im not saying it should be able to handle everything I just assumed a more usable notification instead of a stack overflow is part of the Apache Batik SVG library +Did you mean that is not released to +Also please search the archive before opening any new JIRA issues +Please reopen with a test project. +Now that work is underway with AS all previous community releases are +Marking for +Attached PhilIve added the package that you mentioned and also some more packages that are referenced from the bundle to the for raising this +It will find everything on that paragraph except Folic acid +The goal is to support students who are likely to become useful and productive members of the community +From now WMF stays at applicaton but application is no longer an alias but a completely separate mime type with a separate magic +Ignore that last commit it was intended for. +I dont think I have permissions to assign things to myself are you able to grant that to me +what is the operation you were running for this to occur +Did the nd patch ever get committed +As for maybe we could discuss this further on or offline for a future patch +I attached a patch to a little while ago that adds a preference option to disable the download of source pachages +Second path from previous comment should be ProjectsWin sorry +since we have and it is would benefit from it but I doubt that we can do the backport so late in the cycle +OK I confirm this one is fixed many thanks to Yoko and Charlie for the help on IRC +Running with java +Glad you caught this Julie +Dup +For my information what do you mean exaclty by AJP gets clogged up +Fix is in CR tag +Confirming that works for me too +Thanks for the review +Hi ImeshI applied the patch with commit eaffdcecfacebaaThanks! +uploading an updated patch with the above points considered +If a base project has a knowledge about an upstream module this is plain wrongI missed that during the first review I think the name of the patch file confused meSo it is big +It should pick an unused rule number or ask user for input +I considered your proposed change but turned it down as I thought it too general +As this is fixed by assigning same person. +Brian assigning to you since we talked about it previously +I do not think having candidate in the name helps +Could you provide a test app? I dont know what changes your booking application has got. +markerFile too +targeting for now +There would need to be a bit more work on the exceptions that are allowed to be thrown I believe but the code works just fine. +Should be done post allWe do have all the necessary infra to fix this +If so we can reject this task if not we can send it on to QA +OdiI asked Brant to file this bug report +There is the same problem with richselect component +Closed upon release of. +Authorization already covers this +Test cases and log output from the test cases when running the test sequence as described +It seems to me that we will have to start with the threats and usecases that need to be accounted for rather than starting with a lot of implementation details +I fixed this and applied your test patch in ed +So yeah I would say add the checks and its good to go +Committed revision Thanks Akshay and Noble +So I was able to remove the need for the +Another patch against the now current continuum trunk and some more up to date plexus comitted by jason earlier today +Where in SVN is this example btw +I am using the in a lot of projects and never found this bug until todayBut sometimes it happens that one is commenting out all code +Forgot to add one files mods for this change reuploading the new patch shortly +v +Attachment has been added with description There is an error while writing the TLEs while writing the attribute qualifier in the afp outputPatch applied to FOP trunk see rThanks for the fix +You can now make this work by using excluded subsystems in that will allow you to use CXF +Then it gives an error and have generated json using rumen +Please be more specific about what you are looking for +Updated patch +Im not using +Im still waiting however for verification that systems work properly with the new NTLM code +Thanks +added ritchiem as a moderator +Do you or anyone else here know though how to specify the binary keysnames on the shell command line +Try it again +Ill take a look at +I moved the stuff for textcurve replacement to a subclass of so I attach it for those who are interested but I suspect itwont be very useful of the FOP library itself +Im in the process of updating the file so a patch probably wouldnt work anyway +Danielthanks for your patch and comment your patch is in rev +Please take out the latest and test. +Making it patch available +Mang can you write it? +FWIW the changes to Kryo also fixed my JDK problems +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Resolved issues related to the private gateway creation +We still had to pin the vms to a single blade howeverWe really are trying to keep these on VMs so that we can begin to think about running these with elastic nodes and plan on working on solution to get our nodes off the same blade so we do not take a outage if that blade fails and they flight to another ESX hostOur next steps are to build out a private VLAN between blades for cluster communication +Thanks looks good to me. +ivy directory? I saw it until I did that +Express is not supported +Not a bug +assigned to to correct component + +I am out of office until Aug th +The test class itself creates a memory leak since it never close the session and the session keeps on growing due to the itemPool +Instead of using file for the file scheme please declare it as a constant +Fixed at revision missed to check for the srcWorkspace param in Workspacecopy +I built the web site and confirmed the faq was there +The fix checks the return code and throws an in batch mode +Other changes look good +did not add tests either +The jobtrackers uid would need to be a member of that group +We now enforce that there is a security configuration if there is a security handler and vice versa +really nice +But my problem is with the API an is absolutely the worst place to have a getter for a moving target because its all about searchYeah I can agree with thatMy comments were about the ability to grab a schema off of which is desirableneededSpeaking of which I should go check out how is handled in this patch +Thanks Vanessa some comments to the patch What the geronimo branch this patch is created from? please dont forget we have jetty +I renamed it to be consistent with the name of parameter passed to ctor +Sure you can do this either of those approaches you mention +I have confirmed that the Maven dependencies for now resolve correctly from Maven Central +I have also been gone through the evolution from JRE +I just checked in a fix for the +This is one that a noob could get stuck into +I finally managed to update all dependencies and feature requirements and tested that is installed out of local repo to eclipse +Seems that the the test result shows no failure +Fix checked in as revision +For other servers it is common to have some mechanism to view the requests currently being handled could also be a very useful debug tool +bq +I would like to know if this has been solved. +Also get this warning with your patch AM WARNING The unmanaged resource xxxxxxdesc was feature is deprecated and support may be removed in future versions. +If anything is missing from the documentation it should be added it to the wiki page mentioned above +Alexey please verify +This is independent of Woden or WSDLJ +Wonder if some line breaks event is not captured so the XML gets reformatted to a single line +Yes the case we are looking at was indeed really pagination +It can be assumed that a class of a helper method is initialized +Ok here is a version with an added rat exclude and added submodule decl in top level pom +you please attach a test or link to the github that reproduces it +Well done Emmanuel +If you know of any reasons why it is bad to change the URI on POSTS please let me know +Until them we stick to Hibernate +If its not going to be fixed I will need to work around it +Seems to be an issue with resources from classes +Thanks Raghu! +I have created a sample project that reproduces the bug and will attach if JIRA permits me to do so +Pls provide your comments +I have seen this as well when running on WLS +HiPlease disregard the last patch +updated patch to work with latest version in vs bugremember to run tests first next time doh +Better to remove it now than to ship a release with a bad API. +Several revisions all toldtrunk and x +Could someone please look at fixing this? +Nick and AndrewIve got a gem which requires thrift +Or we could just add that single extra method +Curiousbq +Validator already supports numeric ranges and minmax validations +This has been fixed in svn and will be included in onwards +We have to wait till SOAP is supported to fix this +war and ear just email me +Ill try to make sure the documentation on my end is clear about what version of the plugin works with each FlumeCassandra versionSince Im not as familiar with the Flume community what versions of Flume do you suggest supporting? Do a lot of people follow trunkAlso dont forget that you can fork the repo or copy the code somewhere into the Flume repo at any time +I just have no idea what part of my app might have caused itApp was running on an Android VM +An example is if you have many AND queries which causes a lot large skips +And Im having difficulties reproducing this +Therefore time to close this +Potential landing spot for this utility would be in foundationcore +Not all of those versions are binary compatible +added unittests and an case in +Bringing into as it to myself +I guess my main question is what are the different classes of exception we need? This issue presents one what are the others +css files Ive created a new more generic issue for this which is to automatically add scripts and css from lib script and style directories see +But Im unsure if this will potential break other parts in the MessagingIn my opinion this should be transparent to the JEE developer and should be handle either by the AS or MessagingThe patch seems fine but I need to run the test suite to make sure it doesnt break any tests before I bring the patch in +golden on stacktrace for selector the problem is accept method which falls back to legacy Selector in this case +just pointing quartzProperties to an empty properties file works as a +By implementing the marshaller will use the classloader used by the bean factory which I think is the desired behavior +I do not sure but his code should be in SVN +lets upgrade for appears that the API breaks the code pretty good will need some work to upgrade to the newer API +Is there an issue for API in lucene? The thread looks like people were generally in support +My first attempt to this issue +Thanks a lot! +Sugessted patch +Mylyn currently does not honour the security level setting for projects and is thus unusable for us. +Heres the same code but with no tabs +Last tests of this passed no complaints from other users since +Committed revision +A patch would be very much appreciated +Ive just fixed this as worked with +In any case using a Date strikes me as odd +I ran this failing sequence this morning with over K iterations and a reasonable amount of concurrency across several test beds +I committed it nevertheless presuming its something small +reopening to fix fix version +Argh I meant NOT +I count on you for quick feedback +Modifying the definitions to apply to releases of the product requiring us to have a tedious discussion about whether these rules are recursive and malleable within a majorpatch release doesnt seem like a useful capability +Ive tested it with sstable created on +Closing all resolved issues for. +I agree +CLI supports comment notationplease try that first line connect localhost second line show keyspaces +I would say explicitly throwing an exception when trying to access a cql column family from thrift would be an acceptable solution +I also checked in the JMX and the sample you have provided +Ive brought the bundle into line with proxy uber bundle in importing osgi framework at rather than +thanksRangan +Attached jUnits for inclusion in the bean validation integration module +How can I test it? with older versions it worked perfect +Its kind of rare so I think we can generalize that when we have a second use caseWe should also move file format check function into the specific file format as well + pages for hdataTable were for JSF components will not be covered so far +on previous patch I have forgot to remove refreshPortlet on and so replace refreshPortlet by showPortlet in previous patch I have forgot to remove a at rev +Thanks for taking care of it +Well Ive built and master switched default config to what is present in description and no exception so? +I will have override it and do what it used to do and leave s as the general +We started off saying for simplicity reasons mainly that the annotation on class level is the one indicator whether a class is to be looked at as a componentOnly later was support for abstract base components added but the requirement to have a tag was never removedI think it is probably not a big backwards compatibility issue if we start supporting classes without tags to be used as components +This is a bug in Smooks and before +On PDA this DOM support is not available +trying upload file again +All the incubator references in sanbox are cleaned except those in and the files +PDF file manifesting the has been added with description PDF file I built that manifests the problem +Attached patch fixes issue with pytest +where do we actually compare it w another buffer +Is this a specific issues? If so we should bump it out of M +Im seeing this problem on Windows with MINA +This is in the depot but will nothas not been pushed to the server as that usually happens during the next releaseClosing +Fix Verified and automated in V closing +pushing this to has not been any changes to this issue since February and it wont be included in Marking it as fix version unknown +I cant reproduce this on Android on the HTC One +Add another param to the updatecommit cmd that lets you force the opening of a new with the latest settings +Nick could you help me with creating the appropriate jenkins job please +I havent looked into it too deeply but the script blocks on MRI and as well +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Descoping items not being worked on for M into Unknown Fix Version for not issue out as part of JIRA cleanup +if the AJ editor is not the default editor then the programmer is prompted to make AJ editor default +Thank you for your consideration +Brucethe targetOperation attribute is already implemented in smx branch and in trunkSo this patch is only needed for branch +orm files to use xsd srcormdatastoreidentity srcormapplicationidentitySending srcormapplicationidentityorgapachejdotckpccompanySending srcormapplicationidentityorgapachejdotckpccompanySending srcormapplicationidentityorgapachejdotckpccompanySending srcormapplicationidentityorgapachejdotckpccompanySending srcormapplicationidentityorgapachejdotckpccompanySending srcormdatastoreidentityorgapachejdotckpccompanySending srcormdatastoreidentityorgapachejdotckpccompanySending srcormdatastoreidentityorgapachejdotckpccompanySending srcormdatastoreidentityorgapachejdotckpccompanySending srcormdatastoreidentityorgapachejdotckpccompanyTransmitting file dataCommitted revision svn commit m Change some +Tammosvn patches are not quite move friendly +Thanks Thilina +Rebasing this one was a bit harder than usual + +Or does the name of the spheroid have to match as well? I see in this case they are indeed different WGS vs WGS +If only nodetool ring then its a minor bug +I should set the conf value in the test but not change +and strips this from the redirect URL leading to a blank redirect URLA core dev may see a better way of fixing this but this patch works +duplicate of issues in hbx +LucIm all done with chapts +Single patch for and +Hithx for informing on how to am requesting that the default location should be changed from current work dir to +It helps when looking for new committers if the fixed jiras are assigned to the person who submitted the patch +This is reproducible in the latest trunk also + +Then there were two Tablet objects in the server at the same time and each was allowed to update theMETADATA table +Its on github on both master and closing stale resolved issues +Ill look at it +I get this when I try to access a nested entity which has generator assigned from another entity +I will pull a patch into github +Bulk close for release +Please submit a patch only including the fix here +Applying Hotfixes to a vSphere doc patch has been reviewed by David Nalley and submitted back in January +I pretty much cant get my cluster up without this patch +screentshot. +Fixed the EOL style in trunk +Thanks for the reply Vera please dont take my comment as a negative view +is now Sharable +This makes index consistent +Thanks Eranda +Do not add datetime format methods that do not include Locale and parameters +was any decision made on this patch? +serverDiscoveryURL is now the same as serverDiscovery +we have here a clear conflict between the services thing being looked up through a classloader and the containing all compilation related configuration elements which should contain this information +Take a look at the core +Thanks +Thanks Ive applied your patches to both codebases +Good to go +Thanks for the tips on formatting +But at least now you simply can maintain your own plugin and your own and Solrs build allows you to incorporate it. +Thanks +Is it supposed to be fixed on or the Wont Fix mean that you do not consider it an issueRegardst fix means it is not an issue and it wont be fixed +Attaching + +If an example is avaialble then please reopen it +because you cant really use in a Flash or Silverlight application without being able to read a file +BMThanks for the clarification on union v intersection +Ive modified the patch to return Nil instead of David! I pushed your fix in raef and updated tags in raef. +Also it contains a unit test but it requires a presence of a proxy at localhost so it probably shouldnt be added permanentlyI was able to confirm that it works for http as well as https +Paul do you see any problem in changing to +I mean optional +This also lets us share all of the pieces for testing +Obviously I meant instead of +As noted in this is caused by a bug HP aCC doesnt generate negative on Theres nothing we can do about it +New use cases are now covered +I have just tested having a client in the elasticsearchsink talking to a cluster and it fails tooThus I think that users should choose which elasticsearch jar they want to use and place it in the lib directory +I could trigger the bug as part of a larger test +Thanks Pablo I was looking to document this feature temporarilyUnfortunately as I am on Windows I need it to be there +I switched to using dependency since it works within the reactormultimodule build +So it should be sufficient if we find a solution for the concrete thrown there +This also means the report isnt generated if the user just wants to run the tests for example in a continuous integration system +Retargeting to +Thats not truncated thats an overflowIm not sure what the options are and this might be an thing anywayOther than converting to a equivalent first and comparing against first then proceeding with the existing validation Im not sure what to suggest +oh ok +This patch fixes the issue for me +As far as the Actions button and the tiny one with a small arrow is concerned if you have not noticed yet the tiny small arrow is actually part of the Action button +What exception are you actually seeing? You might want to try a masterdev build of as well since there have been improvements +bulk close of resolved issues whose fix target is either unspecified or from a prior release. +but thats out of source features +Same comments as in other bug report +Daniel you are correct that the cost of the exception catching is only incurredif the exception is actually thrown +Sorry first one wasnt good +However I think it will take me few weeks to reach there +Fixed. +Patch is for trinidad trunk +But maybe Im misunderstanding your workaround solution. +This needs to be checked to see if the issue is still present +On the surface it looks like the file transferred OK but if you do cksum on the files youll see that some bytes are missingI +on the proposal so far for these reasons If the goal is traceability from a token to a job how do you plan to handle clients that lie or jobs that fail to propagate the conf value to a subjob? The core objection is the identifier must be opaque to the clientTheres no guarantee the client can decode identifiers +I wish some warning in the NEWS file for had been added with to explain that streaming would be broken during to upgradesTotally agree +see also +Could we just use instead of as the moduleId +They extend from now. +Moved from to just after merge of old branch into TRUNK +But I am interested in working with you to build off this +I debug I find the range contains only one key +The level of support here needs further discussion + after release. +bulk close of resolved issues whose fix target is either unspecified or from a prior release. +Same behavior for and this a couchdb bug? it seems you are simply hitting your systems limit +Patch for the issue +Yes it is based on Seam community +The on this issue just silently argue that my concerns dont matterMike gave his opinion on it tooStating that our concerns are meaningless is not the way to create consensus towards a good solution here +Thank you all for the focus and help with this +Ive created to fix it +Its pretty simple +times the issue appears +Attached is the patch for stub of +Could you please post the exceptions youre getting? What dependency versions of Spring Data JPA and Spring Data Commons are you using? +further work will be done there +Before we generate sequential messages padded with a series of character xs which leads to very good compression ratio which is not typical in real case +I guess we should support though +bulk defer of featurestaskswishes from to +Investigating +Patch that fixes problems +The file should be +Will post the patch after tests run +Patch for for Avro +It seems like it will be a little more annoying to make changes on the fly by a user but Im sure the tradeoff should be worth itI do think I get a test fail or two still but overall its looking good +script for integrity tracking slightly code for easier maintenance separated data mining and html formatting by subroutines added support for tracking results timestamps added history file for tracking last good results +insert is not supported unless all required columns are projectedIts not clear to me yet where you are putting the generalspecific warningserrors +In order to get the exact behavior we need to go through the snapshot chain find out the leaf snapshot then stored in in snapshotstoreref table which will take a lot of time during the upgrade to do that +Ive had the same problem with throttling so I am running with in ATS +Go for it Ram +fixed on revision equivalent patch for trunk +I just ran the tests with the following results +Fixed in version +Thanks for taking a look Alejandro +Along with tests and an IT then sure. +I think this is somehow connected to the next bug +Maybe we dont need to check for an empty value +Thanks for doing so +Tucu I think we are close but I dont want MR AM or AM configs in +bq +rebased to trunk +The attached file contains at implementation and corresponding and implementations to ease using certificates with Spring WS +My comments on this patch There is a common logging mechanism to enabledisable logs in compiler +If you have further questions feel free to post them on users. +Does it work for your Ted? Should we show readwrite in the Master main page too per region or is that too much +I will apply Marizio idea and you will be able to download it from +Thanks +add uninitialize in to release resources used by ledger manager factory +Fixing this via an addendum patch at. +Just for the record This also applies to binjruby in! Were pushing out RC soon +Will submit a new patch with test case that should happen in about a week to two latest +Release Note schedules now use FIREONCENOW instead of DONOTHING as their misfire behaviour +Should of been resolved ages ago! +action pageSo not sure if this is a incompatibility issue between struts and weblogic g or if there is some permissions configuration issue with weblogic and struts +Added missing properties wrote about the consumer +think Im confused +A very preliminary patch without tests just to get some you say local files do you mean local to the machines in the cluster? If so it is a sort of blobstore right? If you meant local files to the whirr call blobstore has a persistent filesystem provider that can be used to avoid unnecessary code duplicationthoughts +Many cases fixed in r +is this a map c +This has been slightly changed +you +findbugs +Here is a patch to fix issue of correct refunded amount to Google Checkout by Quick Refund and Create Return process +sthree Object methodsthree Objects +is slightly different because it involves the Bundle List plugin. +we should mark it as fixed and close this jira +I used Unix style paths as well as negating drive names for portability purposes +sorry I mixed this up with general generate the wizard it should be trivial to do since you already know the file names upfront +can you test it again? If this persists maybe we should work together at it +JBDS repo contains +snap lolYeah webservicebpel requires EJB +I would look at the connection rate and the loading on the system and then try to find a balance that meets your particular set of fault tolerance requirements +Hints and reference docs updated as well +Havent run all the tests yet though but the added test passes now with the fix +The underlying assumption is that the application which passed the zk instance is responsible for passing a working zk handle +this isnt an issue of getting the IO object thats taken care of by toio logic +For this type of stuff no one has complained doesnt mean much thats why these changes are so insidious they are easy not to notice docs just disappear and users likely dont know they ever existed +As you mention we use SES extensively and likely all those timed tasks have also quit firing which could lead to an untold amount of confusion if we gossip since there would be no immediate red flag to indicate a problem +n +Stephen do you have a reason to say List specifically or would Iterable satisfy you? If so it sounds like were zeroing in on agreement +Ok point taken Ill open a separate issue for this and move it to for review +Once a is created this can be closed +I dont know why I sent a svn patch Here is the git one! +XMLWSDL Importers assist in generating SQL +I think I finally have everything working properly now but apache svn is having some problems so I cant commit the change +In the stream the JBT plugins are not compatible with so I dont see why youre adding its update on the other hand is absolutely required to be able to install me +However I created a related bug to track the issue where rollback was allowed on the newer release. +Other problemsThe is dense and a sequentialsparse one would also help for subsampling giant data models +There is no right place for this code +applied the patch +Unfortunately I think the problem is only reproducible if you have running on your system +I applied the last Dzmitrys patch with only cosmetic changes +This probably slowed ingest too but the slowest node will slow all other nodes during ingest unlike verification so its not as easily detectable on the monitor page +In SLES Environment with the unexpected behavior there are only two CXF for the outgoing transport nothing else +Closing as fixed up in hadoop as per Jamess recommendation +Image quality has improved considerably +I try it on DEV and it redirects to login page as expected + inline removing the useless Stage wrappers centralize stage creation in and standardize stage naming +Patch applied to and trunk +This has a JIRA component set was there anything to do in JIRA for this +JaroslavDoes this error occur every time +Uploading patch so I can submit this to hudson +The current Solr can be deprecated and or replaced with one that accepts a Solr core +I have implemented this in my version but extracting that diff from my other changes is a fair amount of effort +Youre definitely right that interrupting one thread shouldnt take down the RPC connectionAdding yet another thread to IPC seems a bit complicated though +This does not fix the issue it hides it +If I remember right the merge code has huge input buffers for the files which is totally unnecessary for ramfs +What would be best is to keep it in a consumable format +Actually it works for me now with the latest from HEAD +t see the previous comments before writing my commentsim deleting mine as it is irrelevant. +Resolving as cannot reproduce please reopen with test case against Designer +patch for and sample java service and +can be fixed by upgrading to +I committed this +Thanks guys the patch was applied at the revision r +Those are fortunately just warnings that Tomcat spits out your application should nevertheless start up correctly +Ill be able to wrap this up next weekend +This is an updated version of the patch to reflect the updated build system +I think this would be clearer across thriftJDO to distinguish the grantor from the grantee in all cases but if you want to leave it as is just ignore that comment +Although I can understand the problem you are describing isnt the proper solution to define the alternate join table as you suggested? If you want to maintain two separate relationships between two entitiy types you need the ability to separate those concerns +My experience with the ant project is that they are generally receptive to small fixes like this but the time is long +Looks fine in ER +My ITBLL test was green too. + +Updated the review board +bulk close of resolved issues. +JamesDo I understand it correctly that you are trying to test the async caching client by wrapping it into a sync! This is how I found out that and were out of sync wrt they implementations +Recovery does not complete until all attempts of all completedTasks are removed +I dont agree with the patch for this issue which just outputs an alt attribute with empty text if it is not definedThis fulfils the letter of the xhtml specification but not its intention +Please test it. +Entitymanager is constructor of the factories is not part of any public API so I guess it is ok to change them +For this reason we shouldnt include the patch in a bugfix release since it is potentially incompatible +the patch looks good +is this for the latest nightly? +Patch works for me but with the addition of urllib requirement perhaps we should push this to +Thanks Arun +patch looks goodBTW why these previously public classes do not have javadoc +Maven to make sure what Java version are you on +Please review and test and then close +This is not true +Skimmed through the patch seems ok +An upgrade will only occur if the new plugin version has been tried and tested successfully in a real world environment +Ill attach a patch shortly +Ive uploaded the updated patchThank you +I believe one can do a similar test even in rtfrontendjaxrs but Ill need to review how it can be doneCheers Sergey +And I attach it here for you reference and comment +A few comments Could you see if you can have the class extend so that you could refactor implementation to avoid code duplication +Ive applied the patch at r plus an additional fix +I have seen emails back and forth about moving to Corenerstone to incubation +It is likely issue or has been already fixed as part of something else +You might have to override the addChild method of and might want to check for the derivations of as for example the way you import from a and from a is different +Then when it needs to be resolved again if that file exists its loaded rather than +The qe project already supports setting the file encoding of the eclipse project according to the setting within the maven compiler plugin configuration +When testing the patch I reduced timeout in and validated that thread dump is presented in the test output fileAFAIK we cant really do anything with the surefire timeout +Resolving it now +However since then the locking in the JT has changed significantly hence my advise to revisit +Thanks Karel +JIRA where this code was this is the function and it is the same issue +I suppose youre right that is a smarter way to do itCan we keep the edit log stuff in this jira please? I feel like our subtasks are already pretty and we have a lot more to go +This should be checked by default but Ive seen some cases where this was disabled on Linux +Let me verify and change if required in the release note +Is it the right dependency +Have you confirmed that router has been fully boot up before trying the ssh? Its likely mgmt hasnt got chance to program the VR yet +Verified in JBT +I assumed it also implemented generic retries for transient failures +integration of any of the tools out of box would help +But if you are able to get some simple test cases well be able to fix each case in turn. +The best is to put them in the release profile +Small change already tested in production +Thanks Pete +An alternative to could be a config option safePathsetcsolrconfextotherpath option to let RL see selected safe paths outside instanceDir +produce patch using svn diff from top level added blank I really dont know how to figure out what is wrong with my patch based on the above +See in +Patch to wait until process exits before returning success on shutdown +Pushing out to as this is not urgent +serveralldeploylib +From the description it is not clear what kind of scripts would be effected by this issue +Wouldnt you need to ppr the afbody to do that? In which case youre essentially ppring the entire content +Stan Yes we need to wait for the servicemix patch to be applied +Seems the parser doesnt take into account the expression and always uses the class to assign the constants value +The file should not be included as is +strange process. +Fixed in revision +Were just starting to get interesting use cases +app in our distro the supported one +I think Ivans last proposal sounds good get in the JMX part of the patch now including changes for Lukes last round of feedback and then queue up a separate jira for consistency around the sinks +Heres a patch that fixes this +I dont want to mention first step because in my case I have all the forms visible at onceIf we go with above name it can replace +A bit more info Our view is a translator extending from which rewrites the query to mimic views +When generating the project the project explorer initially shows it but is then removed propably because the wizard overwrites +In our case the default was about MB for a certain field while the patched was about +Is this still relevant? I could give it a shot +I have to change my and rerun +Forgot to include some files in the previous version of the patch +I dont see any problem here +Did you subset the font output fromttfsvg? It is possible that all of the diacritics are supposedto be done by ligatures but when subsetting ttfsvg doesnt outputthese and you are left with sub optimal generic diacritic handling +Patch applied. +The problem has been resolved with the patch for the Oleg +JonathanThanksLuke LuUnfortunately we need to shim a little bit more since the actual has also changed +To say it more strict I generally dislike the usage of a default encoding in any kind +I did a couple of runs today in the cluster labs and the results so far indicate that the home grown marshalling framework is faster than Marshalling +The distribution contains several for each sample +Could you checkout the code for Branch and let us know if this issue has been fixed +Ive gone ahead and committed this now that I have it working locally +Its been a month w no progress +Ill start working on a non coordinating version perhaps as a superclass of +Fixed +Now you know how to manage AS servers +Thanks Hairong +Automated running in nightly +NA post mavenization. +Im using tcServer from STS +This is an svn diff which includes some trailing junk about setting the properties for the executable appropriately +The value returned indicates the number of codes not the number of Unicode codepoints +Poor choice of some point prior to attempting to deploy to the jboss server a copy must have been deployed to the repo therefore a deploy must be called +But it might not be suitable for +! to roberts comment +Closing resolvedwont fix issues. +We could certainly put some work in to get a profiling setup finally added to but the chances of matching exactly are rather you sure rails is going to start depending on it? It seems like a very strange dependency to add +My mistake I should be more careful when creating patches +Then I guess I should not worry about builds failing for patches +Committed to. +I added an isolated testcase which fails without this patch +So Id consider change that allows importing of external factory methods a hack even if a very useful one it would not just attach annotations but also static methodsI need to think bit more about this because I see usefulness there and I agree in that there should not be actual risk involved +The lib folder currently only auto adds karaf +Now file can be uploaded successfully +Right +why is runid removed as argumentIt was not used +seems out of date couldnt reproduce on iteration components dev sample and metamer with believe should works in that way +assuming that the tests pass when I run them manually +Lets see how this patch does +On HDFS we do not support +Aaron +So I think it is OK +I will see if it fixes the problem I had +your thoughts? misspoke +Looking at the screenshot I have never seen continuum starting another build when there is still a build in progress though +about the scratch file in the patch +afterwords for the Japanase one +what I do not know is what you consider to be a wicket component in the context of seam for this annotations to wok +adding parameter listAlltrue can list volumes of all fact the volume should NOT be destroyed is reason enough to list it in a regular list call immediately after the vm its attached to is destroyed +Ah I didnt realize that the tests jar was a superset I thought it was only tests and supporting classes not hbase itselfWe still pull in we just no longer get it from lib and instead fetch via maven +I updated theproject to use jdk and then I did not get the error jdk vs +Im unable to produce any error from the can you confirm that this is fixed in +I just tested this in and the reported behavior still exists +Will be in +Accomplished in a manner similar to what happens in the path of the code +Thanks +Patch updated for latest trunkI ran a small crawl with it with no problems +When using the mapping support anything annotated with Id will be stored in the id field +Thanks for looking at my revised patch Knut +thats good to know +Thank you for your many contributions Andy! +Awesome thanks Ted! on commit +Let me know what you think if necessary I can back out all the changes before creating the source drop this think there may be a larger issue at hand here with the tree having local changes for branding that we do not have +Perhaps request may have some effect on how the exception format is selectedI will do some more tests later today with another computer +Thats kinda obsolete now +Ive attached a +I will dl and recompile and check it out later +Apache Stanbol Client source code +Dhruba Its good trying for first +A property is set in the message before dispatch to the DLQ but this can invalidate the message for delivery to other topic subscribers +I reverted both patches +As I recall there were a few additional items you were investigating +We need more work for this +Support of positional reference is not on roadmap as of now +Doc looks good to me +Ill go with suggesting the Dispatch authors to switch from using classes +In portlet applications create portlet definitions having identical portlet name +It can be modified to redirect to different based on the capabilities of web browsers +If all admin locks are essentially shared by all admin sessions this will break big time +In provide the comments in java docs format +committed thanks! +the patch for now based on QA bots found issues +I had the createtrue on the tests which are not required by postgres so I removed them +If someone really wants to discuss using another method we should open another issue. +I see this as a that adds complexity as Thiago has noted on the mailing list +The test failure is ve ran this patch with committed and everything looks Ok +Instead of to be able to associate a coding rule to a specific project I think it should be better to be able to set a coding rules set to a specific projet +At least other Hibernate users wont have the same headache because its not documented. +Updated the patch with addional comments for new classes patch was generated with incorrect version +I committed this. +I have done only small change +Yea its a naive comparison based only on the trees +Doing this nowI grant licence to ASF for +Thanks +bulk close of resolved issues for release. +The fix looks good +Seems reasonable on paperAs for what to do with compaction either locking them during the repair or skipping anyone who happened to get compacted concurrently should be workable but Id lean towards the latter +simpleimpl does not make use of reflection and does it similar to jdbc preparedstatements +I associate more permanent association of a process to a role unlike state changes +Although this bug is filed under MAPREDUCE and surfaces there and in hdfs the patch is actually for looks reasonable to me +Why you dont want to offer the stub to classlib +Fixed an obvious bug which caused unit test failures +findbugs issues covered by +Simply we install modeshape and own dependencies inside a OSGI repository and it worksI could add a documentation about it +Details for this issue are documented in the linked issue +Brock I know you are working on the Protobufs patch so would it be ok if I take this over +WTP shuoldnt just take the latest and greatest and set as default it should be controllable +Closing old issue which was integrated with revision on +I think that a more general policy is needed here +In addition fields are generated with the proper simple type instead of the primitive base type +So the master would be trying to communicate with the remote agent over and getting access denied +Cancel idea is interestingMy dont like the idea of buttons above forms for actions performed on the form button to submit form is expected towards the bottom of the form +INFO level is good for me +It is a folder named. +As we may be recategorising the projects in the future anyway Ill see if it makes sense thenI hadnt addressed point yet which is why Id left the issue open +that and could both extend +If you finish it before we release beta please correct the versionThanks +Need a confirmation that this is truly the case which will mean that it is most likely a bug +Lets try this +We could also link to it from the documentation +After discussion on the dev list I will be adding a new interface that a can optionally implement with a method is +Fell off my radar +If you want to put in the change as is I am not going to block it but I suspect that the security people here will force me to change it before we can deploy into production so I would prefer to just have a separate field that is not a URL. +Well it compiles with no problem under Eclipse +Ah ok +I agree with Tim that throwing a when the connection pool is exhausted doesnt make sense +This will make Chukwa less platform dependent +The repository exists only to give access to jars which cant be downloaded from other wed like the repo and the docs to be out of CVS on a real filesystem +Patch looks goodCommitted to trunk +I couldnt reproduce this with the following configurationFirefox Tomcat JDK Windows XP SPAfter logging out I get correctly redirected to the login page when trying to view the created pageAre you sure this is no browser cache issue +Fixed the other archetype i r. +Here datanode generates a storage ID of the form +In which profile do you deactivate the rule? I thought it was obvious but Ill make sure +The attached patch tells the to use the instead of the +Added code to clean up unused snapshot images in new sub removesnapshots Added working code for the powerstatus sub Added working code for the powerreset sub Added working code for the poweroff sub Added working code for the Poweron sub Finished the controlVM sub to support off remove pause and resume Cleaned up the load sub a bit Added code to properly shutdown and remove existing on the current vmclient before creating a new one +Thanks Rodrigo! +It also has the nice property that datasize disksize after compactionYoure right that the work Bob and I did only included user data in the size computation +Closing per the release. +UI problem assign to thuy be verified in gatein master and reopened if still valid +I just committed this thanks Kevin. +I it and it seems working now +pending Jenkins +Verified in JBDS Build id +If you want to change some logic then do these changes in separate commits +This patch also contains a refactoring of the AES code to prepare for unit tests of column families with other clock types + +Im having the same problem +Should we copy that code +Please provide a quickstart demonstrating the problem +are far easier to review +This PDF shows the proposed approach for default security configurations +Ive prepared small update that fixes this issueRegardsIlya +I thought of that initially +The interest in getting this right is to ensure a critical component of HDFS is rightThe reason to do this is to make sure not only the design is correct but is simple and completely documented +Please post your additional comments to the new issue. +This wouldnt clash with grouping or queryCurious to hear your thoughts on this ticket and where youd like to see it go +Applied +Its fixable but I cant look into it tonight +Alright closing +Would you please check if that is the case and close this this one as dup of is so +Hi MarcIts great to be able to get in touch with someone from IBM since this is quite important to at least a few IBM customers +just makes the implementation more obscure or am I missing something? +This is not jPDL but pageflow +And when I attempt to apply it to my project I get an invalid patch format error +Closing due to lack of activity +I committed this to trunk and +Closing the issue for the moment +In the meantime Ill continue with local tests. +If this issue does not reoccur during the M test period then it should be safe to close the JIRA +I think thispach is good to go +The +Attaching which is an additional change that had to be committed +This has been fixed in custom AOP patch +I just wonder whether the new should go to the same package as all others +We believe its better to reject these queries than to return bad resultsSOLUTIONThe fix involves rewriting how Derby interprets queries with GROUP BY or HAVING clausesWORKAROUNDYou can work around this limitation by rewriting your query so that the SELECT list does not alias an expression to the same name as another column or alias +as bb and bb arent exception handling blocks it appears safe to merge thembbsome PEI with an exception edge to +ceteraThe needs more feedback from Trustin and Enrique for us to agree on the SEDA to provider communication mechanism we will use +Per Emmanuel anything this might break was wrongnoncompliant anyhow +Looks good +Sorry I was pretty busy these days but hope to get back in the game during the next week +v is was I plan to commit if the tests are okI kept the beforeCall as it was finally I had actually a bug in the v as the case without retried was not managed +I will capture the logs again +Fixed on master in commit ade +working in +Prachi can you check on this issue +One should provide a formatted page for Win users and a plain one for ixI will look into it asap but not before next monday +We just know all the column families of a given hbase table +Hi guysIndeed this patch is using very similar serialization to hbase as I just copied it from there to get the patch moving and figure out the harder parts + +Service challenges Gateway with Negotiate for authentication +Tom I made a first pass at the release notes text +Are you sure that is the correct link +I just took any VDB with source model and flipped to the Source Bindings tab and checked then clicked add to see the behavior +The test case should call like set +It would essentially be a root package not a generic of things +Fixed already in latest version. +You could attach a patch to the build file too +concurrent still fails but this time it does not fail with nonexisting commit but rather fails with a conflicting commit +But this doesnt apply to your case I thinkYour program does GUI thing so I wonder you are using sandbox security or something? Such restriction might have thread related limitation. +Further inspection on the consequences of concurrent access must be looked at though +It doesnt look very consistent if all methods that get injected by GORM and not defined in a class itself ARE transactional but classs own methods ARE NOT +Maria good idea can you create a new ticket? I also think it should be configurable +Errors like that usually mean you need a clean build +is inspired by Unix error code EISDIR which means that expected input was a file but a directory is being passed +on latest patchA few empty lines can be removed at time of commitRan test times which all passed based on latest patch +Probably we should go and finish +Should be fixed in next upgrade of Fuse Camel +Pick your poison. +Ralf I am sort of able to anticipate your vote +See details in linked jbossaop has been added +The latest revision number I have used is It seems to work fine from my limited testing although I havent written any unit tests to check the functionalityAs each shard already returns a pivot facet its just a case of merging the pivots together as per the other facets +The option will still need to be defined at global scope in main Apache configuration outside of all configuration containersOn this basis I will now make some of the changes to the C code bits of modpython which need to be made so that new importer works properly in all cases +r updates the plugin descriptor generator to handle overloaded operations as described in my previous comment +I will investigate this bug for +This patch is based off of that for +Here it goes again I have an additional use case where decoupling the minimum and the multiplier would lead to a significantly better resource utilization +Also the and header are set correctly so gmail should have enough info to thread it unless they only use the subjects for threadingCan other people confirm they see this too or is this just me +Recent work on arg processing has made it clear we need to overhaul block args before final so this will probably get visited anyway +Sorry I cant reproduce this +Thanks +Maybe remove from Overview and only show it inside package javadocs +Committed revision +I changed to in and use prettify in in the patch +We will also break anyones poms that are depending on the existing jarsAs part of moving code into I can create a pom for that pull in the new jars +Committed revision Committed revision +Hrm Im not seeing that +I think John is working on the backend code for S and he needs to confirm on the listSs API call behavior +Updated patch +Ive got a patch for that but will be adding a hack to set the format to KMZ when using the flat mode against a raster layer before posting +I think for now I solved this +I agree with Chad +Can I change parent version to +the problem is not so much if it picks now Callable instead of Runnable +bq +Can you prune libext as and when those services get broken out +Regressions passed on Solaris Windows with JDK and +On yes I think that backporting wont be enough +Is there a reason that decimal can not be supported or is the support for decimal incompleteIf SARG can support decimal we might be better off not adding protection instead we should ensure that our unit tests cover all types +Ivy should pick up the right version the build machine doesnt have the problem and I cant reproduce it +Thanks. +But I thought you wanted to be able to set this on a basis +Have confirmed that the calledElement field of the activity is not being set this would provide the link between the calling and called choreographiesNot sure what you mean by the first sentence? A separate Eclipse project will be created per serviceparticipant as they are intended to be separate projects that could evolve to include implementation details but there are not multiple folders +Thanks for adding the unit testsThis actually isnt an issue from what Ive read since it only uses the hooks in java and not an actual implementation we made ourselves +Maybe this issue is a side jUnit test which demonstrates the issue +Similar patch for trunk. +Modified files include KaushikThanks for contributing the patch +Requesting close of hlog ERROR Error while syncing requesting close of hlog FATAL Could not sync +At this late stage I think Id rather not spend too much time thinking about it and just remove sysinfo from and come up with a better solution later +Hi JonThat would be very fine for me +Please apply to b +bq +When I run the tests with the all works fine only with the CVD the exception occurs +These variables arent even used otherwise +bq +OTOH checked exceptions all over the place might not be helpful either +Doing svn switch relocate to change to fixed the errors I was seeing +It is mostly a refactoring of the code so that a certain functionality is separated out more cleanly +I think the only manual change you are required to do is to extendGive it a try and let me know if it worked +Now exists with tests in EJB for invocation over the Store +Im reluctant to spend more time coming up with a better fix since we want to discourage use of this kind of joinThe fix is to have the base tables tell the VTI where their rows are +In Fedora added thi line in etcfstabtmpfstmptmpfsdefaults +I fixed it +Im particularly interested in the start stop scriptsIf its at all feasible Id like to be able to able to have options on to +Good candidate for a +Can you try changing it to Web Port startup poller? I assume that will fix it for +Im not sure how to fix this problem I could of course just change the javadoc +rodrigo would you be uploading an updated patch +Descriptions are adequate as they stand and will be rewritten for. +Closing issue as it has been released as part of release +Note I had Stack review this patch last night since he did the rolling restart description originally +work after i put the with in tabbedPanel tag VERY STRANGE! +If you believe this issues should be fixed please reopen and add reason for reopening +Back in the portlet business for a quick visit again I stumbled upon this bug +Those from +Thanks for the help hunting this down Casey and thanks for the fix Wayne! Ill see if I can merge the patch over to as well. +Question should the be before or after the call +Perform ajax submit +Id suggest that the forked lifecycle is not cloning the list of already resolved projects and their locations +Manish why not put your content in a different field without stemmingYou can use +With the kind help of Guillaume I now figured out to introduce a typ converter which now allows to correctly announce the configured auth methods on the wire +Updated patch using svn di command line +Placed at wiki and forum for public discussion +However if we had a more generic it might fit +Nak a banner like this would break many usages +In trunk r fixed queues page to show all current and queued builds categorized builds that are distributed and built locally in the queues I enable distributed builds then I want all the builds your comments its seems that if I turn on parallel builds now things start building on the masterTo me being able to use parallel and distributed builds together means that the agent can do parallel builds +Can we get this to the and branches as well +Resolved with patch under +Just tested using the network test page +As it is not so well commented I also will work a bit on the comments +Using the patch for the trunk code one can run on every and +Committed to trunk and +Wont break +Please open new bug in Jira +Any hope someone solve it +Thanks to Adrians expanations +Scheduling for later +The change cleaned up the warnings +Good point I didnt think of that +I was originally trying this on an old Beta build of embedded console +But ant can find javac successfully from RI +Youre right Bryan Im developing in Windows VistaIll update it in my environment too thanks for pointing it out +have fun +In default settings it uses builtin from Hadoop +Ive added the missing tags to the tld and also updated it to use an XML schema instead of the DTD in the mater branch for you please fix this also in the branch? +The was moved to the common module in Qpid trunk at rev as the first stepThe common module also contains code for the client and therefore contains configuration required for the client to be accessible from within the common module +Heres a patch against current +I think you have a local classpath output from export JAVAOPTSverbose +Ive never heard of this before but I have seen LOG messages severely impacting performance because of strings generated unnecessarily and that one reason for adding the if blocks +I excluded the false transitive deps +SPS recompilation may use internal nested transaction and user transaction to update system tables +Ivan I think there is an issue with handling of the config option to override the configuration directory +The patch looks ok we had a similar problem with the main makefile too +ok sounds some actual relevant text is better than data +Turns out this issue is really about badly naming the body element as when we already have something called +I just committed this. +Hey DaveId favor using option here +Its contributed by WTP +Another version more cleaned up +HiIts hard to do an ETA but please follow the discussions inThe community has voted on and produced and were trying to make maven artifacts. +Works with single multi or no keyspace arguments +Changed so it doesnt give out regions if region was on dead server not unless logs have been split +This issue is confirmed resolved in of +Here is a patch that applies cleanly to trunk +the later speculative read removed the first reads readCompletion callback +Lars Im pretty close now testing some reflection stuff have one of the MR tests failing currently +That needs a little more work on getting this to work properlyI have tried to shutdown the default task runner factory fix but when I hot deploy the container again +This might be related to have reread again the WTP bug comments it actually says that WTP doesnt support s workspace resolver +SCR has been released +It works in transition resolvedfixed bugs to closedfixed +I dont know much about security or SSL but the func spec looked good to me +the main idea is that the compiler removes completely the code for cfdump in your case cfdump is still loaded and called for a couple of cfdump this is ok but if we have lets say calls of cflog this makes a difference +I have a modul A for what I like to write some specs +Shouldnt have rejected this issue as it was reopened incorrectly. +Konstantin can you please file a new issue to switch Avro back to? Thanks! +I dont believe they are requiring the opposite either thought +This is committed. +Sorry for not testing all the scenarios with this change +I cannot use this workaround +I signed it +Fix for. +That would not restrict its range and might use more space in some cases +In this example the textfield is bound to the cardNumber property +And if in this case the JAAS configuration cannot be loaded this would again be a configuration error +Im fineIt is easy for the developer to fix the compilation problem with Optimize imports IDE command. +prep closing stale resolved issues +Maven will think these are different and include them both +It seems that multiple optimizations are possible +Heres a quick patch to fix things +Ok Jay I already setup the build tied to branch it was actually modified properly by Nick and we are currently waiting for successful build the branch is after merge still broken. +With this change the parsing logic becomes fully independent of the transport traditional IO NIO email files on disk streams of logging data or whatever +Maybe rename to send or something +I changed my mind I wont convert the attribute type back to string instead I will create a renderer string property that will do it +I think the path to finishing the code by adding more tests and implementations for bytes shorts ints floats and doubles is easy enough but it wont affect backwards compatibility so not a blocker for. +And together with fixed columns and resizing feature its more important so separate attribute used instead of style +So it would be great if you can provide me with profiling informationI am not able to check this too deeply until the next weekend because I have some exams to take +The first highlights OK for me but not the second +What is perhaps the real issue is that the Plus sign is dropped as it reads in the error message +The reason why I did not do that when I introduced Mutation was the the difference in which Increment handles its familyMap +on +Reopening issue because the Reference Guides documentation of needs to be addressed. +I am done with the testing for this patch +Yes it would be great to store the multiValues as array instead of concatenation +This is to pull across the trunk fix into the CP in revision . +See also possibly related +It looks like bndlib v has made it to maven central +Marking as fixed with. +I committed this to the feature branch +Could you please provide your memory settings and try to capture a set of thread dumps to help locating the problem +Cancelling patch to incorporate offline review comments from DevarajComments include +Latest Patch file attached uses with instrumented still fails to get the association to stick outside of this editor either +are being madewe have sorted data files containing same sets of keys but corresponding hdfs chunks of each file may not have the same set of keys +Thus we can detect possible problem more efficiently instead of having hard to reproduce crashes +Ive created two pull requests one for and another one for a possible branchThe latter one is based on the latest commit from master which is not related to BV and thus could separately released +I think its best to offer an option on the seda endpoint to dictate which behavior you want current behavior with an exception waiting and blockingAnd maybe this one to offer eg a sec timeout waiting blocking but with current behavior +Ill post again when it is resolved +added to the nanoc this issue as fixed +I run my tests with the changes posted in AMQ and it definitively fixes the problem! Thanks a lot also for fixing it in too since this is the version we are targeting to useBest RegardsMark +With I can get the count correctly +Actually any variable I try use result in the nPE above +Docs must be updated as well +Please add a new API in there +If you reduce replication factor your durability also goes down +I vote that we pull this into as well +failed but the patch in no way touches code in this test case +Fixed the link the site and it looks good +Ive successfully logged SQL statements against postgresql after enabling SQL proxy weeks ago +Okay thanks for the comment Brian +No further feedback given assuming no longer a relevant problem. +I have committed a fix +Note that the patch on is required or the test errors out before getting to the part that makes it crash +Ill attach the test log +Adding it in my todo yeah this makes sense to me +CI setup +The cb will never be completed +Im sure the change is good but still want to do that. +Once install is completed copy from INSTALLDIR to users home directory +Thats why we have this Source object which a points to a source file and b points to a class file to processWe can change the process how these source objects are constructred in any way however we need a pointer to the source fileI guess it would be interesting in the scala case to get an error reported on the scala source as well +So perhaps thered need to be more testing +The patch is right it will use user input +A first try leads to trouble +Perhaps we should simply include this in as this provides no value and isnt really a bug until is implemented +If it is agreed that a container such as Spring should be used then the wiring of should take advantage of that +Also limit DISI to looks good Uwe +This code also doesnt exhibit the same one interpretation is this issue occurs when using s eachchar returning from the block of eachchar and in combination +push to +What shall we tell them +Column added at r +Looks good on a quick review +closing as duplicate of closed. +Good catching XutingVivek I assume this is an invalid issue can you double check +there is such option but referenced libraries are hidden by default in Java Package Explorer view it can be done from Navigator view where there is no such more feedbackcomments assumed ok. +thanks to Kirti Dhakad for patchRavindra uploaded patch is too old need to upload the patch for the latest trunkI do not remember what was the purpose of this issue but I will look into this then either resolve it with updated patch or discard this issueApologies for late. +Carry on +If that is indeed the cause of these particular test failures then perhaps a check can be put in so the fixtures dont run if the locale is set to something else than expected +Ill commit this in the next day or so unless someone objects +I managed to fix this and it is working now. +Select projectdir Ok +I have been banging my head against brick walls this last week because of a lack of this features so I would say noAlasdair +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I also like the that allows modifications of the jsp pages +So we only need to call update in MR +Fixed by Maxim +Thanks Lily and Rick for running the testsI think there are multiple issues with this test that causes problems when the test cases are reordered some tests leave rows in tables that are shared between the test cases some tests dont drop the triggers they create static counters are not reset between the test revision . +Then Camel offers JMX API to do that +Changing the title to reflect that its not just a web UI problem +Update Bulgarian localization for the information in for providing the translations +Thanks Sergey and Nick for the review. +Out of the points that were outlined in the original post it was concluded that only points and are not necessary +Has been already fixed and now works ok for of a bulk update all resolutions changed to done please review history to original resolution type +I confirmed that the temp streams are appropriately limited to and I can see that we already have logic that should be freeing them but the inputstreamfactory created by the objectdecoder doesnt implement the free method +As it stands right now this patch now completes successfully against Pointbase and mySQL +Both are in though +When reverting to the old version again the triggers should work correctlyWorkaround After reverting to the old version drop and recreate the affected triggers. +Committed revision thanks robert. +This patch fixes some problems in the client main programHowever I faced some genereted code problems on windowsNeed to retest for looks to me that with the fix for this issue has worsened +Thus B would also be marked as provided somehow? Or am I interpreting the scope transitivity wrong? + +apxs is broken on this machine and modcluster seems compiled with the wrong the build of EAP should fix the problem +No special subclassing or interface for widget to be including in dashboard +Could you squash your patches and upload a new one for commitThanksJon +Given those priorities it may take a couple minutes more to call a set of services from a instead of using a service group to implement a service but it is a lot less flexible and ultimately leads to less than ideal work aroundsChris your solution to this of wrapping the service group in another service is the way to go if youre trying to avoid redoing the service group +committed +Ok I now think it was to consider a GAE platform +Please note if you proceed with a bulk move will not be Let me know once you are ready as you will need to modify the roles for the project to allow access etc +Patch applied at +Thanks think this might be a problem with your build +Anyways I added a couple unit tests to make sure it does not come back +Indeed it can be closed +Fixed as part of in +sounds fine + +Config parameters in +Committed to main trunk with revision . +This example shows how to test a jar who contains struts action and need some file in the webapp classpathI will attach other files will apply the changes +Thats why suggested removing the jar because there is no conflict then +minor bug Wrong return codes +try authenticating against your acctYou aretrying to unsubscribe your email address using gmail so it requires your apache ldap credentials +i want to attach some made the same mistake as you did at then i created the attached results with the hsqldb in a separate process +Sorry for raising an objection on the patch +Please add release notes for this and related +I will verify what I have with svn head and pack it here as an updated patchRegards Doron Your patch uses JDK + I propose we bump up the default heapsize to M too many people are getting burnt by MWhy the rush? This is for trunk not an urgent fix to a blocker where a fix might be appropriate +Feel free to attach a patch +The client stub is generated only for the last porttype in the wsdl +Closed with the release of. +Gregor do you work on a patch +resetting P open bugs to P pending further review +We could however include a patch in case you have one +Are You using mvn site to generate the report? Can You attach the reporting part of Your pom here +I just dont think that ords without fallback is going to get very good +revised concepts w EPS accepted by stakeholders +I need it +The final resolution for this issue will involve rewriting the test output parser to search backwards in the file for the summary tag patterns +Is this issue going to be assigned to an AS version and component +Bumping to for any future perf m going to mark this fixed since we have put the in place and it solved the vast bulk of performance problems +What do I have to do +Fix posted to mailing list waiting review +Can be changed to not throw an if there is an existing object there already? It could instead return a boolean indicating whether or not it was able to successfully store the value +we should stick with hadoop not CDH and make to work with it +Thanks Bryant +Triaged for assigned normal urgency. +The issues with national characters and should now be fixed +Verified by running the same command directly +I also cant manage to reproduce this issue. +Thanks Hiram +Hi Ben FlavioThanks +Just committed the fix +Otherwise you can directly reassigned it to me so that I proceed with +Deferred to to jBPM releases can only contain bug fixes +This bug also applies to regular TCP change limits the impact of +How about renaming it +Amit I was able to repro this issue +Attaching the pdf having state transition diagrams +The current fix does not respect plugin configuration declared inside executions +It seems to fit all our needs use the quickest and more standard way to declare plugins in a transparent fashion +Updated the test to use your settings and large file still works fine +Millions of thanks if you can help +This is actually why the strategy of moving nodes around works too you move those that are super frequent but theyll most likely be reordered at the top suffix frequencies of the automaton anyway. +Given that RC date is Aug and the suggested solution from Wido works on KVM only I would move it to the next release +I think youll need these if you want the init scripts to run on boot +User should use this option if his aggregated values are very large objects +Since this was omitted by mistake and marked Critical it is approved for +seems to reproduce it so lets just keep it open for now. +I am closing this issue but if you see it again please reopen and post an updated stack trace +Hi TedDo you have an estimate for how much network traffic or heap footprint that this would saveJust curious not an objection +Hi Rakesh Thanks for reviewing it Im fine with your change +Thanks for taking this +Tommaso WDYT +Attachment has been added with description +I added myself as a contributor +If you want to do it yourself please hop on the developer mailing list and discuss what changes youd like to make to the code +clear the changed types and elements +Also at the same time keeping the abstraction of Executor creation away from the user so that it could be tweaked depending on the startup memory constraints +Fixed in new version on OperaVersion Build PlatformWin SystemWindows XP JavaSun Java Runtime Environment version Pluginnot available +Will reattach the same patch again for one final run through Jenkins +For your open question I dont think we should support upgrade from a namespace that was not cleanly shut down +Any objections to the latest patch +I agree with Max +Hi JacquesI am still full time writing Ofbiz code although module this was finished some time ago for meSkip +With update It seems better to compare source dir with both stagedir target dir for filesizes and checksums in first try alsoThis is basically when distcp fails after copying some of the files to stagedir and then when user wants to copy the remaining files using update it is simple for the user to give the same stagedir if distcp could avoid recopying the files existing in stagedir already +the timeout check fails because existing tests have the function start on the same line as the TestI modified the test to try to open numerous connections and then verify the max is not reached +The only error I get in this application is about the contentType not being set it seems the upload does not have content type info +Removing init just makes it stay at trueMy point of asking is that saving password should be false by default hence I wonder why it is set to truethe patch is fine +Consider upgrading to Oleg sorry for the delayed response +We arent really doing memory management +I looked through your code briefly and it looks reasonable but it was a brief look +Pull request has been merged. +Hehe race conditionsbq +xsltc is failing the following numbering conf test in nontrivial waysnumbering numbering numbering numbering numbering numberingnumbering numbering retest show numbering and numbering passnumbering and show only a minor difference at the top xsltc has and Xalan doesntnumbering and still show nontrivial diffs with the expected output +Jacek thanks for the report +Similar to the suggested patch thanks. +I tried it +If you think that would be a useful addition you can go ahead and create a new Jira closing does not prevent further work +Attaching and adding notes to topics that discuss the security mechanism in the documentationNo change is planned for the documentation since this release probably has few if any users at this point and since I dont know how to generate this documentation +Please confirm this fixes any other leaks +I am not exactly sure what we are testing with this jira +I tested this in the latest nightly and it does appear broken +Hi AllI have attached the Maven project of the prototype created using Apache BatikThanks +Yes all the integration tests run +Attaching patch a which removes the redundant check Dag spotted +No need to do this task more complex than it needs to be +Right now there is a lack of detailsregarding how to actually build the site tree +That would slow down invocation against singleton instancestypes +If there is a problem please open a new jira and link this one with that. +Its a fairly primitive fix for now scoring methods both on whether the arguments are assignable AND an exactness count but it should solve the majority of cases +xsd and +According to the performance numbers run by Tim the s creation of was expensive +use instead of +Please have a lookThis change will also require a small change to MR as well +I still have pending changes since there were many failures for a few of thepatches +Obviously for our usage the thing I had in mind was being able to pass in an avro schema or url to get one +I have added a new configuration property to disable implicit boot delegation +Jeff I just committed a fix to the svn can you test please +Put it in the objectstore +As an aside I think we need to start making more frequent releases +StackIn our internal version of which has changes as in along with Timeout monitor related changes we did what ever Maryann did +Would be nice to have the old naming too as default and flag to the edireader that you want to use the id instead of xmltag when naming the elements +rars are missing +At this point we can safely assume that the broker will not be sending any more messages +Hi Alejandro Arun Thanks to both of you for committing this +I have attached the failures received when running with host besides localhost +Might as well add support too come to think of it +This is the java program that I used to debug this problem +Works for meThank you Alexey and Gregory. +Last thing to say setting the to is so far the best workaround Ive discovered +For now it is required to place in CHUKWACONFDIR but this could be fixed by fixing the shell script. +It was written before the TX manager was provided in ESB +IVe committed a first simple for the patch Ive applied it to get it into the code idea this seems to finally fix all casesIve attached a new patch which includes this functionality. +The and versions do not bump rpc to maintain compatibility but notes in comments that has been doneYou are going to bump the rpc version in +This is fixed Felix +Marking this issue as Needs more info as from a user perspective does this change actually need a release note? Does the user need to know that these error messages will no longer be logged every five seconds +No need for +Changed the summary as I suggested earlier +Sergei though the reply is a bit late I hope youd still be interested to see this your work integrated into the cpdebase +Moving to TBD +Makes sense and lgtm +I fixed the whitespace issue +this file should just stay there in both M and M for now +This new patch replaces the older it follows the layout and other patterns used elsewhere in that file +Sorry I meant pruneA design document based conflict resolution function wouldnt discard users writes its a mechanism to allow users to more easily do what we already suggest they do to resolve conflicts write a new document andor delete conflicting leafs a read resolver is similiar however the conflicts would still exist which isnt a particularly a bad thing however it makes the semantics of alldocs views and changes more so Dale does mean resolve in the delete a leaf off disk sense +This issue is pretty well handled in the current release +This needs more discussion +I give up on this. +HbaseOther dependant users need need end up in reflections for every version compatibilityI think you meant need not +The documents are updated as requested +close opened file channel during garbage collection so deleted file space can be until we sort out the api issue raised +So PERTXSYNC? DELAYEDSYNC? DEFERREDSYNC +Thanks Raghu. +Working +I had to revert you change back +I will look at it +Main class in this case is probably itself +Should be removed from Fix Version +generated logLevel INFO style OBF port war Djavatargetitsrcmainwebapp startupUrl hello +this issue just the reminder of the test case and problem with current auto layout algorithm +Comment included on +Also please feel free to comment after the fact +Unit test for reproducting the issue +r +Before the region server died I see in the master that a log split was going on +To throw a bit more light on to this subject the only thing that seemed to really work for me was removing the from the grailslib directory +In both cases the servlet container has to create the and related objects +Allows to change the style of required content and label of all input patch uses Adams modification for intercepted states rather than the previous patchs code path. +PRPs now persist look good but that Array must not be zero length error still comes up when dealing with remove and not setting a new value +Agreed +My vote would be to make left and right ends optionally open or closed +The source file that is used in my +Attachment has been added with description fo to generate attached used to generate the attached transition to closed remaining resolved bugs +For now it runs chmod x on the file if its present and accessible +Only problemremaining appears to be when doing a make install but that is easily workedaround and might be another issue I thinkThis issue is that make install only copies intousrlocallib +I committed the above change +htaccess TODO Need to recycle the httpd instance to activate +So I think we could address this by adding a sessionid system function and by adding a config property to enable the env function +Its the combination of all that causes the bugC just seems like the simplest one to change +Jamie since JB isnt online today feel free to resolve this issue and get the Release Train rolling +slight refactoring by changing the tableName property when it refers to a tableId +Here are entries from myjlibCjrelibjlibCjrelibNotice that the colon is NOT escaped +Thanks for your comments Stack! +If no one has any objection I can add the whitespace myself and commitWhat version of Solaris is this +I have been working on reskinning the pages in the Hadoop branch using Bootstrap +Patch for Y! to trunk +DUPLICATE +Link to modcluster config generator has been included in both EAP and EWP versions of this publication +Then we could add the feature that returns a different WAL for META alone +This issue only occurs if you explode the war file to the servers deploy directory +Testcase showing the code for stub and ADB classes +here is a short description of root of the problemthe start timer is parsed in parse method in Bpmnparse +Reopening issue since the commits in question have potential IP ownership questions +See getLocale +I doubt it would cause any problems but I wanted to make sure I had a completely clean zip of the project +thanks a patch would get it applied do +However it would still be nice to have it as an interceptor as Im porting to the simple front end and this would aleive the amount of subclassing needed to achieve a JAXWS +Yes in the browse dialog is an empty line under root node +Is not controllable by SOA plugins on their ownQuestion is though if this change enables things that is not supportedallowed +No further feedback assuming fixed. +Heres a snapshot of me clustering wikipedia docs +Please add a edump switch to your command line and post the resulting stack trace here +But Ill wait for more feedback +code looks good +make log format more verbose logic error caused erroneous patch master and region server do not join with tweaks +bulk close for the release +I must have been tired last night and accidentally deployed the instead of the +My name is Patrick Lightbody and I approve this message +This way a single replication session still shows up as only one entry in the history log but if only one log gets written during some checkpoint pass the comparerephistory will still find an older checkpoint if it existsIts not even a problem if the recordedseq doesnt match +Thanks RonI reverted to which I know works on the iSeries time constraints on proj ll be patient and wait for Mr Burke to correct in his next again for your prompt problem Ronnie +Please reviewOleg +Will upload patch when tests complete +was fixed to include LDLIBRARYPATH pointing to pathtosourcedeploylib +Committed patch that provides a locking implementation based on tracking the of locked nodes +Looks like it works in JBDS RobBut i believe this is caused by AS M or M changing its startupshutdown mechanismthus JBDS which just targeted AS M or M wont work +Thanks again for your work! +The others deal with other topics which are outside the scope of this JIRA +Well when i set VERBOSE on a plugin it is because i need a debuging aidONLY for this plugin +Default collection names are working for me in for query and update requests +Patch committed to branch +Are you using an old build of the compiler? This is probably a stupid question but is there any chance you are looking at a service class rather than a struct +Thanks Amareshwari +Cool +Both tests failed on Hudson in many recent builds +Closing this as a duplicate. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Add the libs to the classpath and run the Tester class to see the issue +ESB standalone distribution if the user use this type of ESB runtime ESB tools will pick up jars from directories liblib and lib +Attaching which attempts to incorporate your recent observationsI found that if I set JDKID to JSE for the small devices then database boot failed because Derby tried to bring up a storage system which relied on the nio packages +Hi NirmitI just checked and found that it is working and I got the original wsdlpls check with nightly build and check whether it is working for you +As Simone says the two config interfaces could be merged into one tooI was thinking other folks would just compare the end result class hierarchies and see how they smelled and felt +Another thing about the latest patch this line needs to be removed in RPbq +Chris See +But I do not suppose that is what expected here. +When I use with it gave me error +Ill keep absorbing for a few more days +Thus Arun pointed in earlier comment does not copy any files to dfs +andjbossjboss applied to the old EAP maven repo so marking as out of date +Thanks for your advice Juergen +I certainly would have shouted if any warning bells were rung by the brief look that I did take at the timeHTHs move on +So all issues not directly assigned to an AS release are being closed +BrunoIm confused +Ok created a test case that reproduces the issue +To create Javadocs for and they should be marked Public Stable +Committed this to trunk and branch. +Actually i am also getting the same am using eclipse anyone help me to solve this +For information it does not occur on Linux too +I just wanted to know if this is a regression or not +Alena You have a valid point +End users do not dictate code level organization but they do have an influence when compatibility is involved +Complete patch that uses libtool to make the modulesplugins +This is why I reply on brighter collaborators to namerename things +Uploaded patch with fixes based on Sebastians feedback +See Exclude groupe +I have applied the patch +The client can set either but the NN will translate relative time to absolute time for the and for serialization to the edit log and fsimage +I believe these are two different tasks +This can be accomplished by changing the log listener for the log file so it picks up INFO level log eventsWe dont want to do this by default because it would drastically increase the number of log files created by Pig since now we only create the file when there is a real problem executing it +Thanks Plamen +See Wish +has been marked as a duplicate of this bug +Thanks Ravi +Please provide your feedback +When I ran it in a networkServer environment it failed with a problem that I logged as a new Jira +Document generation should use docbook as the document format with freemarker templating capabilities to access relevant artifacts to complete the specific documents +Whats odd is that I do not get an errors with regards to a missing file +This may cause checkstyle errors +The first fix introduced Blands rule to prevent cycling +However this all especially youridea of using a query optimizer is really a bug of work +trunk supported verified in JBDS H Fedora +Closing issue as version has been released on. +I just committed this +Not sure if we should continue executing commends after a failure +Thanks for pointing that out Virag Ive missed that +Not enough information to reproduce or test the issue +Typically you work on translations as bundles ie +Ben can you update the patch to apply cleanly +This concept is essential to support a generic handling of and its contentSo I guess there some more things we can do to improve the performance +Either way there have been huge speed improvements in project archives in teh upcoming stream +Thanks Amareshwari! +Adjusting the priority to fit the definitions Minor loss of function or other problem where easy workaround is present. +Screenshot showing the improvements in Firefox at r +fix bug uncovered by jenkins add a little more debug in +This can be tested using the version of the m planning to release the update on Monday +curl output log fetching binary resource from output log fetching binary resource from this was disabled in because there were issues with applying GZIP on top of byte arrays Ive updated the runtime to optionally put again specifically for byte arrays but will actually have to be configured explicitly and have its newly introduced report boolean property set to true the RS runtime itself can not guarantee some CXF interceptors like GZIP one wont affect the length so if the user knows that it is safe then this new property has to be setthanks +Closing this feature +Anyone able to comment on whether that commit is indeed the cause of this problem +Thanks Scott for updating the patch with the three configuration variables +upmerge to resolve conflict from removal of other again third time is a charm +There are a couple of things we could do that we should talk about I had a couple ideas that worked when I was originally doing this but did not encapsulate into the input split organizer so well +All other accesses on tasks are synchronized on this +This issue is specific to the and should be able to be validated without using DB. +seems reasonable to disallow it then +Whatever it says now is of no consequence +Previous commit missed this fileSending modulessrcjavaorgapachegeronimosecuritydeploymentTransmitting file dataCommitted revision +Does this fix locate in in svn? +Since we have run automatedtesting with various kinds of inputs I would have expected this to be caught +Still no way to fix this in the frontend until we have a spec +Thanks MamtaIm not really certain which behavior is better the first time or the second timeIt is definitely confusing to have two different behaviors thoughWhat was the final outcome after the experiment? Was the TIMINGS tablepresent or had it been successfully dropped? +bulk close of wont fix unreproduced and duplicate issues +Uma +I would recommend using entities for this ampamp for CRLF just amp for LFIf this is processed by an XML parser thatll already work for free anyway +Please move this to GRAILSPLUGINS and assign it to out the flush leads me to believe that youre just missing an exception that occurs later onor not making any changes to your DB at this bug occur outside of? I never testedintended plugin to work there +Openidjava released in June +This is awesome Marshall +requires OS dependent setting in to address the CR +But it still have some benefits that we discussed above +with a slight modification The method only updates the service registration after also validating service references +This is because the dependencies were not yet released +But it will be in scheduled to be released in about a week +Jon Yes we should go with some form of persisting the splitkeys +fix committed in r +Booleans are greateasy of course +I am missing your comment +Pushed to anybody know the reason for Hadoops object pooling? MR applications are typically I so garbage collector overhead shouldnt matter much +Since you are familiar with the code change that went in I will be fine with your judgement call for backporting. +bat file you might need to check the contents of commit r to see if there were any accompanying changes therecheersj +will close when code is pulled up to branch +I dare to comment without looking at the dependant code but it would be great to have it in the ioc module + +Both layouts will work with JBT +Please delete from SVN the +The patch includes a testcase too +Patch committed to branch and trunk +Commit changes to trunk at revision +Committed to trunk +Patch removed trailing spaces per RB comments +patch introducing the possibility to use resource properties with resource adapters +some test for your patch but i am not able to apply this patchi am not using any svn version please tell me how to apply this patch +YES! I am so stupid that misdeemed that cm equals to mmI will fix it tommorrowSvante the vector is the old synchronized collection and has performance issue then which should be used instead? List the non container +Reviewed all these issues and marking them as closed. +Ill use that next timeGood news about main methods +Since we have the Warp Alpha I can do that and prepare the release +Committed revision . +Pavel When debugging issues I have to use the builds as I need to play with fixes +Tomcat is holding onto the core library and I would theorize its due to this issue. +Recreate the patch after updating my local codeThanks am just going to commit this small patch +Applied patch +Hi allDid anyone have a chance to examine the latest patch? Is everthing fineThanks + for a no arg constructor equivalent to the default format +In fact all etc resources are not copied into the NMR distribution +Thanks for the feedback! +If you want to update your patch that would be dandy +pushed +Patch which converts the last of the Analyzers over to using +edson this is a duplicate of similar jiras assigning to you to deal with +If there are any comments in the future I will fix it in another jira +After extending deltaspike API well be able to implement custom properties resolver +would that cause any trouble in the case of OSGI? Sorry I didnt think about that possibility until I had already committed +Then ujax is probably betteri dont think sothe jcrprimaryType should be used since its already defined by JCR +bulk defer of unresolved bugs to +Serge does this match what you are seeing +no there are reasons for forbidding such a is this a bad idea? Is seems like it goes against the normal inheritance model of Java +Can you post the tests that you are failing that you see +Two tests are broken by this patch I linked jiras for the failing test cases +Thus your assumption fails +I will announce plans on the mailinglist and create issues likewise when the dust on my desk settles Best regards Wolfgang Andreas Again Im wondering if you have any solid plans for unicode text API in the near futureThanks for any info or ETA on +Marking this for beta but disposition as you see folks I mistakenly changed the status of this to Resolved +This prevents blocks from getting orphaned in the blocks directory forever File recovery happens after reassigning lease to an NNRecovery client This also includes safeguards and tests to ensure that straggling commit calls cannot incorrectly overwrite the last block of a file with an old generation stamp or a different block ID +Page is a bit better now. +I can confirm this issue broke my application after updating from to +Last time I tried staging reported Internal ErrorAs far as I can tell the code in the SVN staging area has not been updated since then. +Sorry for not being in patch format but Im not at that stage yet +As discussed on the ML here is a patch that will allow the code to run and compile on +added a partitioning skip flag debug Im currently evaluating a job that makes the partitioningThe problem with the partitioning itself is that the parts are named like the groomsBut I can overload the partition method once again and deal with and just rename itI open a task for the error logs +version requires v works without having on the branch +Subtask fixed. +The TIW OOM is still occurring however +Its quite close to the Jamess one but I put Ignore for threads to keep build green after apply and I also assert that are properly destroyedIm going to create sibling issue for DIH and put most of my prev patch there +Thanks Alvaro! +tdocument is required for theCould you please try a tinputCalendar with and without ADDRESOURCECLASSYou should get the same exception in both cases +The network code should probably try to reconnect when this happens +Ok so we need to abstract out the allocation part of the block cache then implement the same download and install system as we used for LZO +First there are no known issues with it and any changes may destabilize it requiring more maintenance releases +Committed to trunk and +Work was completed. +Testing will be done by hand for now +As a first step Ive added a list with all committers to the website with version +you are right it goes away on trunk +Again note that in the case of heap corruption the native stack traces dont actually indicate where the real problem is +Test case to illustrate the as part of +Youre probably using the mirror where there canbe delays caused by the svnsync operation +Nick does this mean it makes senseis ok to commit fixes for into trunk for the various modules nowAbout xulrunner im still awaiting info from Maxim about if and how about since it apparently removed the selection API we depended on +wrong project There seems to be no Apache Logging TLP on jira otherwise I would have moved it alreadyMaybe move it to +I havent checked if the th bit will be set only by but otherwise the change seems good to me nice catch +Adding the for checking the result of invocation +I havent noticed any bad effects. +Done +There wont be a release +Committed revision +For the attached test if you put in from or M test fails +lemme know brett +This was fixed in transactions project and fixed version is already used in AS codebase +I guess changing the retries to should also fix the problem as well? We shouldnt get a NODEEXISTS return for the RESCAN because we create it as EPHEMERALSEQUENTIAL +However none of the form data can be saved +In day we should have A B but we got A BI see thats becouse the revision number for day is and quering entity B for a revision equal or lower than we got which leads to the value BSummarizing how can I insert a past event with envers? Does valid time really solve my problemsThanks +The NPE is fixed but in the UI the highlighting is still horrible +I believe the management API exposes everything that HQ exposes +created a patch and have tested it +added skip config parameters in just committed this +example of error of in +I will submit a revised patch +attaching patch that opens an in create mode to effectively overwrite the main index +The current trunk solves the issue +While you are touching this would it be reasonable to also switch from using the to a blank? It runs a lot quicker if there are a large number of topics and you run the tool from outside the ZK clusters DC +Can you please rebase? Thanks +I have implemented Functionality according to comment givenuploading the patch for the same ThanksAwdesh Singh Parihar +It occures to me that we shouldnt even need to sort anythingStay tuned +website move complete and synced is now showing +Yeah ideally the stress test would be testing at a range of load levels instead of just picking one + us know if we can close this issue. +But I am attaching the directory that was available +rRemoved the deprecated methods from the previous design of prior to this JIRA +I saw some patch files are missing from source code +I dont recall exactly how JSW works but not waiting until the framework cleanly shuts down is certainly one cause of the problem +A unit test log file and jstack dump of the VM running the results of run +Hi any chance to fix this as soon as possible in a snapshotThis bug drives me mad +Am talking to Scott on forum about how to solve this +Obviously this is a completely untested change +The API locates logging implementations by searching for in the classpath +Thanks Sandy +Ive adjusted this issue summary accordinglyAlso the Portals Bridges common dependency will be upgraded to release version which now is available from ibiblio too and there is no need anymore to configure additional remote repositories like bluesunrise +The final patch has been committed for this issue +It should have said this issue is Blocked By instead of Blocks +I think I am asking to enable the sliders on the bottom so that stuff that wraps off the edge can be viewed +bulk defer of features from to +I dont have postgresql installed so could you file a new issueAnd does the pg gem work ok once you have it installed against our +Updated patch to fix the unit test failures +Verified in +Daryoush since you wrote this sample any chance youll be able to take care of these changesThanks +Committed to trunk +Nicholas! Yes We should make it private and put synchronized on access method +Maybe this can be merged before the final release? +And I dont want him to have to modify his all the time +The script simply compare the number of javac warnings with and without the patch +Thanks a lot Ivan. +According to last commentor this problem has been fixed. +Id prefer that he had a look because he already expressed some suggestions and concerns. +Thanks Raghu and Thejas for reviews +doesnt handle the An easy mitigation would be to modify the wsdl to point to https instead of http +Updated to collect some more data in the logs +Very sure we wont be able to haveem all sorted out +Marking the parent issue as resolved +Uploaded renamed files +for the patch I committed it to branchThanks a lot for taking care of the including many of mine! +I dont know why we have the dependency in the first place +Rebased to the current trunk and updated content +Resolved issue but did not assign a fix version as it is only related to the site +Why not just make a dummy request after the restart? +So Jason what should we do +So a stacktrace is needed +Reworked patch to avoid ignoring charset issue +The old version must have been cached somewhere or the mirrors had sync issue or something. +I am not able to reproduce this behavior with the current development versionI think the problem has already been solved on June th as part ofCould you check with the development version from our subversion repository +But its a simple patch so once it is generated from a newer working copy it will most likely apply cleanly +The HV type and version columns are not marked as unique keys in the hypervisorcapabilities table and so the insert ignore statement is not effective +Thanks Alexis for all of your hard work on this +Im not working on this issue. +Please Kamesh for the patchSource code changes looks good to me +bq +There is no need for another reinit state +The index table can be created with a name which is the actual table name suffixed with a predefined pattern idxUsing fixed table name suffix is not flexibleFor diagram the flow is different from that of diagram there is second createTable call issued by Master +Uploaded a patch that implements a workaround that enables clients to fix types to schemas upfront with +Im gonna take a look on that +This looks fantastic! Dumb question what does the do or rather what feature is it providing +Logged In YES useridReopen if its still occurring +How to view build version +Closing. +If there is a new jar somewhere I can download +Hey Rich thanks for the time and help reviewing this as well as committing it +I am leaving this bug open and rescheduling it for since this seems to be a temporary fix +However I think ease of coding trumps that +might be considered late for somebody +However it was noticed on testsWe found this issue through but Im not sure how other Resource Managers would usually behave +we change the version node version number Do not touch the data written under version New data will be written with version New RM can read data and write data When downgraded old RM can read and data but write data With a major upgrade say to we create a new and NOT touch any of the New RM reads data but writes data in a different hierarchy In case of disastrous rollback all new apps in hierarchy are unreadablegone +Great! Thanks for applying the patches +Adding this param means that we need to ensure that older config behavior does not change +CLASSPATH Xmxm ryPlantrue Main Will run with rows +I guess that disabeling this fixes your problem +gems a different implementation of say or which uses the and requires would still have compat for all that is possible in but allow users that use Pico with do so +I just add the fact that it is also true with version too you can remove it if you point in fact the xml api is the which is a dependency of stax +Brian hows the investigation going +However it only ocurs if a JPA entity is replaced with another JPA entity of the same typeHeres an improved test case that integrates in the existing test suite +works good on Hudson +Fixed by fix for +The task might have succeeded if there were fewer concurrent tasks +I am not in a position to be able to reproduce the deadlock issue I was seeing when I reported this issue +this is not an easy fix +Yes I only tested against H and Oracle again all I have easy access to here +No rush its the wrong version and should be fixed +Just reproduced this I can take a look at it. +Add direct endpointsAdd JBINMR Add the newer REST stuff Remove Camel Saxon component needs to be added to the list of phase one components +Uploading a revised patch +That works +And xfire caller never access directly to theses URL +Relation column was added to selectdisplay related property Owner possible values are yesno or undef +Im submitting another patch perhaps this bug needs to be reopened +Lets use followup issues with tighter scopes for further improvements. +to trunk +Jetty log for a clean I still dont know why this problem Postgresql version do you use? How do you create the schema +kinda require we get the dTP connection support in now just add a warning when running +the patch is generated based on +reopened just to set the resolution to fixed +is a major release which was never intended to be backwards compatible with Acegi +Test case you for reporting and providing test us know please if you face any other related problem. +Can you confirm with the attached patch +We have also experienced this problem at my company +If you dont mind Daryn Ill reopen it while we are still discussing +Attached web service used to trigger problem +ok +Committed without the xml file +the new patch looks good +Permaine Cheung also confirmed that was the case for. +Now its clear we should use logging API not a specific implementationAgain we have following two candidates +I have the same problem in other projects where im on the commiter side so i fully understand your point of view +If we can move to jUnit with minimal impact to users then we would be able to use jUnits built in Ignore annotation instead of adding our own +This is a bug +I am leaving it as resolved for EAP + +This can simply be an under the covers changes which could be viewed as a bug fix to allow missing columns if they occur at the end of values +Id rather just add the dependence on Snappy +The return value of and the check versions is taken into consideration while retuning the return code +It now also allows to run ant test from a source distribution ZIPTGZ which does not contain the backwards folder +test patch attached +Please review and apply +I cannot reproduce this +We need a flag to control whether the response should include the xml declaration or not +This will be implemented with a svn mv rather than a delete and add correct +Shai can you please commit the patch to the branch? Thanks +Please give me a little time to check the patch +And Viola has fixed this in trunk +Well I think the patch still stands but I havent had time to do the auth stuff +This patch fixes the reported issue +Yes thanks +In an earlier comment you mentioned Since you are using a prepared statement derby chooses a query plan without knowing what the values for the? operators are +Did the fix possibly not make it into the release +to check with compilerReport understand better you are using gwtsoyc or add the mojo in the reporting section +Whoops forgot my additional local tweak +This will resolve the unintuitive behavior that we see today where a message is accepted at the source but gets dropped at one or more of the intermediate pipelines +Until release it worked perfectly +Add a FAQ reattached from bugzilla +for application on in Tools +Probably it is relevant to your but once we start down this path we will have people asking for other sort ordersThat being said Id hate for your work to go waste +patch a to the branch with revision Attached first revision of a release note as the change may break existing applications using the back end +Have you come to a conclusion on how to solve this issue? We frequently use String constants in annotations and its very awkward that Groovy forces us to inline them. +Thanks for the quick reviews Suresh and others who commented. +Integrated to trunkThanks for the patch ChristopheThanks for the review Lars +It no longer has the in the list which should presumably satsify this requirement +Thank YouOK we will try StefanoWe did some more tests with your proposed changes again and it still did not workAlso searching in the documentation did not help usThen we switched on more debug traces but were not able to find the reason for the Validation exceptionWe did upload the logfile to ftpoutgoing and did mark three lines with these labels PFA Starting Deployment of PFA Starting message driven bean PFA ERRORWe did expect to find some records beetween the Starting message driven bean and the ERROR mark that would help usWe found a lot of Finding class xxx and corresponding Class xxx not found pairs but don not know if this could be the problem in the last few record before the ERROR mark +John do you think you have time to fix thisIts the last issue left in the releaseIf you dont have time right now Ill just push it to the next release +can you not simply redeploy the version to the main repoIn the mean time the best solution for dependencies is to specifically include the dep in your own POM +Ive got a visibility macro in trunk that seems to be working +So I decide just to display the existing roles and then the user has to retype them +wicket does not validate semantics of markup other than that it is valid xml. +Ill add it to the Wiki +Moving out to for now we can discuss if we want to support forward proxy and HTTPS this way +Here is the patch to fulfill what I know of this issue +Ill have to check it out +Im opening a bug against the arcsde datastore for Gabriel to Gabriel did we do something for this one? I think so but cant really remember + +Is that what you are using? +Look at screenshot plz +They are named operational because they are meaningful inside the server only so if you dont explicitly request them there is no reason to get them backFix the client. +Soon few more bugs are waiting. +on v +I can commit that now if you want +I think it was just misunderstanding from Pauls side to reopen this issue to wait for release +There is one code tweak the rest is updates to tests now that POI exposes more +See for details of the modifications +Added line number input and resource path to message +Please review +I assume its the resolution of this issue that prevents the tests from suceeding. +Giri java dependency issue has been filed asRamya make sure the HBASEHOME is defined correctly. +properties file +Hi If I am reading this issue properly the integration tests do NOT work properly when using Hibernate with combined with the sample script above +Could you create a pull request Andr +Hi TimThis is not fair you should not make such proposition to an addict like me Of course I want +Could someone direct me to the test that I need to change as well? for the patch +A cluster with million files +LGTM +I had to research issues that I should not have had too. +same for cannot reproduce this issue using the and the tomahawk candidate release for +flow flowScope for JSF binding prefix to variables. +This version info will hopefully go away with my next patch to allow running the test as part of harness +Ive updated it based on your last comment and the refactoring you have been doing on etc +Mysql and other drivers dont seem to exhibit this problem since they use getString not getClob +description of patchI have added set methods for blob in embedded driver +I just committed +My application has the same problem after upgrading from to +Need to regenerate + you get a stacktrace from firebugsamplecontainer seems to be fine for me +million bytessecAfter ms million entriessec +Im seeing this on EAP and. +Its better to just use all of the existing to store the timestamp and store any and all additional information in additional bytes +Thanks! it mean that we have consensus for implementing option +Attaching new Common patch for hudson +The Oracle bug report is +Once the remote tests are integrated then will test the success condition. +Hi sirthis is unfortunately general issue in Sqoop +I doubt another hadoop module will import bookkeeper so its fine to not shade it +to replicate the issue Extract all contents into a tmp directory run see VM debug if you want to run with JPDA listening on port just got a new VM crash +Instead of providing many different launchers we should improve doc and allow people to easily create their own launcher +Patch updated to the latest trunkA few notes tests pass tests pass tests pass only partially failures occur during merge operations +I hope everyone is fine with the idea of coupling message storage? Please let me know in +See also +The problem is knowing whether the logic is true in the method +Will add the null check to and +Took diff from wrong directory +Hmmm this requires some thought since you can chain cache stores + logs +Support problems users on the mailing list asking what it does or reporting that it doesnt do X or YGiven the above its a hard decision to make +HiSorry to disturb but I would like to have some news about the progress of issues +Changing the relationship +I think it best to wait till that happens to move on this +This should be closed because the reply is now truncated +Manually tested! +methods +I can optimize the parsing about x more dont parse from text! My question is whether the use cases this approach is for are simply better suited to binary input since it is even faster results in less IO and has no gotchas +Same can be done on client but will incur overhead +During indexing if url field is changed host and site fields are updated accordingly +My version of tools is at while my JBDS plugins are using mostly GA plugins +reopened for fixVersion +There are commits for this JIRA from months ago so im going to guess its done resolving. +Attached wrong patch first time +test case eclipse maven project that uses previous one uses just to trunk also where script is refactored a lot +I think an subclass is a cleaner way to do this as it doesnt introduce an extra frame in the exception stack traceI implemented a new class for this and committed the adapted patch in revision +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +It is syntax that was added in +I personally dont think its worth mentioning the number by which it will be off +Reporters of these issues please verify the fixes and help close these vpc network tests now can create LB without specifying the scheme explicitly. +Verified in EAP script name in EAP is not seam but seamseam +I believe he has something in the works +to me that sounds like a very risky and big change and there are no details herewhy would we do this at hive layer given we have HAR alreadyi really dont understand why we wouldnt start with and then add HAR as an optimization to reduce number of files for small partitions +How come we dont package the commons file upload in the standard lib directory? It seems to me like we should since we have a default dependency on them +Thanks Devaraj K! +I was even contemplating to show that stat for last min too +Lets take these tests one at a time +I had seen this issue as well but have not seen it in days since the mentioned fixes +Why is it using etcpasswd as a source file to begin with +Committed +It was quite interestingMaybe we should need something equal to the graph database would be an option +I didnt think to try the we have someone revisiting the rabbit work right now so maybe it makes sense to have them try this +Thank you Ivan. +This isnt for testing Cassandra but for testing apps that depend on Cassandra +However many customers use jBPM in this way and may have written their applications to take this behaviour into account +Thanks Alex! Ive committed this on branchesnewtrunk +This will should save you another loop through checkDest +However for Java plugins uses +Please note that theres an outstanding patch for INDEXOF and LASTINDEXOF in ll review this patch +defer all issues to +change println to Harsh I added +Just click on the Queue type in the nav after youve created the first resource and you should be fine +please run the and browse to httplocalhostbugbooklist +The workaround is disabling security +Did you create the Github mirror with or with Githubs svn import mechanism? If you used what options you used to set up the mirror +I think ultimately the information should dump only for errors or higher if the derby engine is booted but while developing you may want to leave the boot check out and the use this corrupt database which will fail to connect and produce a errorLily also mentioned to me two other types of errors that might need to be reported java exceptions that dont get caught and wrapped in like some and also it would be nice ultimate to include low cost asserts in insane builds +This is because in AS server all the JMS destinations are deployed at a leter time than those in a standalone server which means when the stored JNDI are loaded the destinations are not deployed yet so the server wont apply those loaded. +I have added a patch which solves the issue and can be used as an example to see how it works and what it looks like +Im not sure what you mean what is a membership lock on the node +Thank you for your comments and commit +It is the problem with last +Cool +If you are commenting on or the flag will not be added because jboss and did not support the b flag +fix applied at revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Patch after incorporating review comments +It is not an trivial task and needs some time +Thus the usage feature really shouldnt be listed in JBDS as a required feature only the usage plugin is important to getAnd isnt this issue about it being wrongweird that Core has the feature listed as installable when it really shouldnt be shownadding the usage feature to JBDS would be a working but not proper way to fix his IMO +Can you please rb this? Also verify that is reverted since it is no longer needed +patch for preliminary AWESOME! So glad to see a patch posted here +Conversely if you define xxxyyyand yyy doesnt exist in the module xxx it doesnt complain whenit really should +It is possible to save subscriptions +Im able to reproduce the same test failure both with and without my patch applied +uploaded for the patch +this patch removes duplicate calls to in the just committed this +Ive debugged this and its because parser is trying to create an instance of this classThis class requires several classes in the +The RPC client needs to acquire the lock to UTFOBUF in order to receive the response from the server +The changes are +This patch adds the header +Whats the stacktrace of the exception? What does your transaction demarcation look like? Any log messages worth mentioning? seem to do manual transaction handling there calling methods on the +But I dont see the interference with delta scoring as the global is set before the solver starts and never changes during solving +Committed to branch +Fix part of +This is not an issue for or JBDS Ma +ok i was able to reproduce it the client connects to the follower issues a sync the error message shows up in the log of the leader +Had commented out umask setting which is reverted in this patch +merged to branch in . +Im adding a that refers to an unknown idI will attach a XSLFO file that reproduces what is probably a variation of the same problem +So this issue will be resolved when PR will be applied +Patched files and a slightly different version of the patch +No but we implement Central so we can make it do what we wantneed +When a URL hasfoobar and filters are added by the support the URL is preserved correctly +Changing to other supervised learning algorithms shouldnt be difficult and I have made the regularization pluggable but I would as soon get this working as is before adding too much generality +maybe the component of this bug is JDK +Hi HolgerThanks for the reportIs that possible that you also provide the file you are trying to openThanks I try to open comes from CProgram FilesApache Directory Serverconf of a ADS installation on Windows has been releasedClosing this issue. +It is a patch and I want to spend some time to understand it +I think we should fix this when we move hcatalog into +Hi RamanI think this is due to calling unsubscribe method twice if you get me a the full exception I can look in to this +For me switching to a traversing sounds like a bug in the cost calculation or the query engine +prj file which is created empty +Going back through our existing websites and updating them to make every link a component is not an option +for the reviews Uma and Todd +Perhaps flush the os file buffers between runsThis may be a race condition where the change in zookeeper needs to propagate to all the tservers and this is done asynchronously +Closing as Incomplete because there is no feedback from the reporter for months now. +Maybe use workflow or Python ways. +How can WTP find the src +Thanks RohitPart of your patch that is applicable only for Secured in FTL is in trunk rev +Am I missing something or can we close this bug +So I will mark this one back as Resolved since its in domain now +added a test patch +I do agree that it is a good idea to keep this close to Java API +This runtime issue is being tracked as +I changed the fix target of this to with the intention to make an annotation on an assignment not be a declaration +Issue can be reproduced on can be reproduced in RF +Not usre if its the jdk problem or netbeans +So there is no direct link between Seam facet and via Project properties page +While the existing implementation supports authorization on some operations it doesnt secure all of the operations provide a way to bootstrap the system etc +Looks good Nick +The is what I used to create the standalone version +Your goal of having the defaults be separately configurable without changing code is being done today +When I drag the top the scrollbar shows up but when I release it the top of the dialog moves upward out of view I cant get it back into have rewritten the application with manually configured pages that access the bean +To make sure that all dependencies are explicitly linked together +Hi Im not sure if this is a clone of s a pretty trivial patch +Well revert if someone feels strongly this is a violation of the freeze. +Reducing but it has been this way sinceI can see it is in there a long time ve tested it as well and there no error occurs +LDAP provider is pretty much in place now so we can probably use more specific issues now +If you want to apply the patch with the beautiful icons then unzip in your root and it will place the imgs in +Felipe conveniently just posted the debug with Eclipse note too so I included that option is to place the new MAVENOPTS text on its own references page +Isnt this a duplicate of which has been resolved +The initial version is in trunk +work with used in bundle to load classesin osgi evnthats why we need wrap it like its done nowAlready explain on irc but still put here to track thisthough it finds now the needed protocol providers the datatype handlers are still not workingI tested a lot yesterday but wasnt able to find a solution hereLars +that documents are still being passed to the indexer even though says they are not to be indexed + +Dirk comments? These are your turf obviously. +This patch makes sure that a dfs client tries to read from a different replica of the checksum file when a checksum error is detected +This code needs additional patches from and to function +I think I like this better +well is that a new alternate introduced +Actually my analysis is incompleteif we haveoption then we should end up with a collection containing a single empty value which would be consistent with what I said above thus having an empty collection as requested by this JIRA is indeed correct when no a given query key exists +Here is the patch created using diffIf not OK please give me the command you want me to execute +Hi Marko That test wont work against the source because the version of in doesnt start an nio connector + +Thanks +I would like to see this added as is sometimes necessary to compile all classes but package only a subset +This happened always innot just only in suppose its a bug in +Second version of the patch that correctly handles subselects with no quantifier +The specific request was to include the vdb nameversion too +I would also like to see this in but dont know if a schema change is OK in that branch +Hi DanielThis is issue was fixed in and will be included in the Solr +Revision implements the above logic see for detailsLeaving this issue open as we might want to discuss the magic value thing for the ujaxredirect parameter +Thanks for catching that +and forName currently seem to generate a single exception string only irrespective of the few ifelse conditions +Not a bug. +As Leigh appears to have supplied the required patch nearly a year ago is there any reason why this has been integrated into a release? +Reporters of these issues please verify the fixes and help close these issues +But from the users perspective JBDS appeared to have the project correctly deployed +Forgot to tick off the ASF license grant +I believe for the test to work properly youll need to install the capture file and the media plugins. +I already found and quick fixed the problem but I cant commit it yet as it will require a bit more of refactoring +applied to and is not active and infact is not either it must have been created at some point during the refactorings we probably need to clean that is still developed in trunk so this is where it should be fixed if needed too +I believe the answer here may be to use instead of the +This bug is fixedPKCS is defined and working +There is a version inconsistency the bundle show the version as while the bundle names have correct one or the other +Sorry for the late responseNo this job does not have a combiner +That should be pulled into trunk once this has a patch for trunk +And for the time being I wont be investigating this any further +Closing the issue +Ill file this one as a separate issue. +Currently one can specify in POJO and it will get mapped to XSD date +It will be interesting to see how things look years from now +So the choreography is no longer annotated with any informationWe still need to work out how best to link a choreography to the activity information but this may occur based on the service types. +I ran the test locally and it passes +Supply a in your dd or plan and a in the plan +defer all issues to. +Quite frankly I dont think even think Lucene committers can even have a look at it to consider it for inclusion with a GPL licenseIf you have written all the code or otherwise own all copyrights would you mind switching to Apache License? That way I at least think it would be possible to have a close look to see if this is a good fit for Lucene +batch transition to closed remaining resolved bugs +builds cleared plugin installed +Committed fixes to trunk and to and branches +Well you can find that information in the CONFIG message and in the my thinking was that it would be nice to tell which server a given request was directed from the client but youre right its easy enough to look up server address from the jvm route via a INFODUMP +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I simply did a git blame on the class in question and found that Juergen added that method on Feb th +The connection test on a dual Opteron RHEL bit Java bit serverJetty NIO connector persistents Jetty connector persistents Situation reversed from! The persistent connections are horribly slow +May be that helps to find possible css class overwritesGood luck +New patch incorporating review comments +confirmed the fix +server on the seems to optimize method calls a far better than client +Hear hear! Debug logging is definitely very unhelpful as it works now +Thanks for fixing +There were problems The test was setting up the security realm with authenticationjaas +Anyway not sure what we can do on this +Hmmm +Marking patch available as a patch is on RB +Thanks Jeff +rob could you look at this again and if continues to be a stupid WTP error we cannot really fix let us report it to eclipse and put a ref here +Verified fixed at rev. +flume that does it the new way +Do one at a time +Did something go wrong with the history doesnt appear to be there +I tested it with Jetty and Tomcat and it working +Thanks for the thorough report +Works for me thanks a bunch I appreciate the quick turnaround +Can you provide a reproducible test case for this bug +For your convenience I attached a screenshot +fix at the same time +Yes you can use these properties when configuring DBCP +Two more pieces of this +The only reason I can see how EPSGnull would end up in a crs field woudl be for the lookup call to return null +UweYour comments were a home run thanksThere are places in the code where constructors needlessly try to assign that +And a JIRA to wire it has been seating idle for almost years +Ive pulled your change testing it now +Thanks +We can and should revisit this if there is a good reason to see that this isnt the cause +I think I have a patch for this floating around somewhere +Patch attached +gson seems like a nice project as well +As well as explanation of how things work to help understand configuration +Worse yet everyone has a different opinion on what the default returned value should be so you end up with a complicated and convoluted API +BCEL is not perported to be thread safe its up to you to enforce that +Still digging into fails with compound file formatStill only seems to happen withOddly it seems that the searcher opened after the replication is somehow using a different Directory than the one the replication goes to and the writer is reopened on +This is just an initial patch +Ahmed would you please upload a scriptpatch based on the lastest patch +In Java the US languages go from enUS to enUS esUSThis is a problem once we need to compile on but it doesnt cause a problem I think for the usage of a Lang jar under the JVMSo punting this to +Heck I dont WANT to use it +What are my next steps? Create a new issue with a different module? I am wondering how I can get this fixed and confirm if its a bug or not +Why? Only for testing? If so add otherwise the java comment needs to be fixed +Thanks much +Closed with the release of. +Some more advice on your test you dont need to provide your schema as Hibernate can generate that +Well Thank you for your reply +If you can recommend paths to refactor so as to reduce code duplication Im all ears +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I did not post it in patch form bc there are a lot of moves and reformatting so it would be pretty messy +by Richard. +The fix for RAD helps the property to persist +We could be backwards compatible with Hazelcast if we want to? Or just cut and require Hazelcast There is a slight API change in Hazelcast in the +This is a patch from the tag because the trunk didnt install cleanly on checkout it may or may not be easily portable to the trunk +It was not fixing the root cause +So if you do see this problem again please raise a new issue +What do you think +Various cleanup on application components commited in rev +This is the same as with the used by default to build a jar packaging project. +The JCR layer is properly handling the delete but the above exception is in the federation engine +I fixed and as suggested and committed this. +Let me know if theres anything else I can help out with +I dont think we need support with we can stick with and to varying degreesEdit Fix accidental use of markup +Point takenThrowing an exception is an API change so this will have to wait until the current implementations in trunk factories cannot be null theres a check in ctors so this issue can be closed because we prevents null factoriesOR I didnt understand how the pool should work so null factories are allowedWhats of the two +I increased the gcgrace for the CF in question to a large number like year +Im hitting this problem with Wicket but not using the Strategy referenced above +And if serialization is the way should be static to make it work +requires the file to be present +x null is probably not what the user meant but as long as the filter passed no records through we did the right thingThe documentation needs to be updated to state this behaviour +Sure will remove tabs before committing +Copied the gist to. +In an attempt to make this more focused which settings in particular are direly needed and unreachable otherwiseIm specifically looking for datafiledirectories so that I might locate all the snapshots +Yes pleaseWe are in the process of change so we might as well jump to svnpubsub now +Yes it does not disable other beans completely +It was the first second strategy is to produce some custom header element inside like or with some attribute fromtruefalse +Assuming fixed +no response details or test case +Thanks Mark. +If youre using this patch comment out the given block using regexes if you want it to compile and work +A high level interface would allow more efficient support for partially structured solutions +I am not an expert on the reader nor am I an expert on how streaming loads and uses these so I am giving this patch a but I would appreciate someone else to take a look especially in relation toIf no one else takes a look in the next day or so Ill assume that this is OK and check it in +What version of your DB server? Because on our side all test passed well +Could you attach a complete stack trace of your first exception +StefanI had not bothered looking too closely at the errors I was getting +Id say just commit it +If I persist state afterwards I dont get an error +A new issue will be created for that variation +Still a bunch to do +if it isnt present itd ignore and not add the jar fileI dont think it will be ignored +if nobody objects i will mark the interface deprecated in trunk and adjust the subject accordingly and resolve this issue fixed +I was working on implementing Solution as described above but Im on hold now until we can finish our discussion of potential issues around this solution and what the best alternative might be if any +I have fixed and it will be available in the next dev build of groovy Eclipse on. +Could you please attach a screenshot +fixed +Thats not an option +Jouchen if you could also create some minimal project which works on previous version and doesnt work on new version this also could help us +No need for IT +Other Apache projects have this automated QA job setup that applies the patch and does some checks like did coverage come down? find bugs increased? compiles tests pass etcI dont know how to set it up but I think we need to do some clean up in the existing test framework so that we get a reliable mechanism to say if a patch broke the buildWe should create a separate jira for sanitizing test cases +If user specify lzo and lzo cannot be loaded we should output an error instead of changing it to +Committed to trunk and +As said previously this issue is no longer relevant +Seems J has some other problem it fails with the same diagnostics as beforeDo we need a native regression test? Volunteers welcome +I was not suggesting otherwise +So I need to update the Phase patch to account for thisI already have the required fixes locally I want to run derbylang to make sure nothing breaks and then I will post another version of the patch +about where this conf is used +Hence resolving. +You get a relocation warning not an error everything still but doesnt that still mean that the pom should be updated? Otherwise whats the point of the warning +Agree with Leleu we badly need a server module instead of client module +modify to use prefer to create new service may called which uses Apache common file upload library then we deprecate the old one to make sure there will no regression in next coming versionMoreover in the we are still having an issue of the uploaded resources in the case the user just uploads files then do nothing +I just tried every interpolation method including Nearest Neighbor all with the same result +I could not find any example which is using +updatedThis file contains the extraced attributes which are referenced by is expected to lie within the same directory asPlease read the comments in the +Patch applied in revision I also created a test case for this functionality in sandboxcluster. +META +Thanks +Not going to have time to focus on it. +Hi WeinanWe maymay not continue to use emma in the future so I would be reluctant to add that especially as we are using a custom fork but only for community testing +Ashutosh Apologies for the delay in getting this committed +If we go this way I think the only piece of code which would need to be changed in method +That makes Whirr flexible although there is still a big need for recipe testing +see also patch for +Closing resolved issues for already released versions +The RM sets the secret key in the but the NM only sees a null come out for it +Please confirm the current status of this issue and if its incorrect change it as required +Is there any followup on this +Wasnt able to run +I dont see anything that looks wrong in your subclass +reopening to change title to its real have confirmed that this is still happening in AS GA release +Closing resolved issues. +This is the way the spec is going right not +Some of the comments are valid +Fixed that +I found the root cause it seems +Experiencing the same here just running a test on Query is enough to kill my machineGalder had to terminate hudson today as wellCriticalWorking fine here if I change it to a CHM +In keeping with the more generic script direction I refactored the taglib extracting the code that actually creates static and dynamic scripts into its own class implementing +I will submit a patch in a few days +Thanks for the test case. +I have committed this +Any chance of trunk and patches +Fixed in branch and merged into trunk +reopening to give visibility to the vs considerationIm with Yonik leaning towards switching to for counts anybody else have an opinion +HelloIve read the entire thread and installed but i still have the issue reproducing itselfIs there anything to do to Cordova besides installing it in order to fix the issueIs there some code or files i should search for to be sure Ive installed correctly? Thanks a lot for your excellent work Shazron +Committed to trunk branch +I would mark this as a DUPLICATE +Updates Jackson core and asl to latest in so looks like only has the patch applied + for the patch +Do you think its better to keep this in the Apache repo or somewhere else and just link to it in our wiki? +Thanks Andrew. +Decided to leave modules as is with the idea that google support will be expanded within the existing set rather than adding new +faebdaccfecbfdc seems to have clobbered that Im putting it back +I agree that this would be very useful as this makes it easier for the programmer to see documentation for an element without having to refer to the schema +Ill try setting it down to and see if the problem disappears +Can you please rebase? Thanks +r completes the fix that r started I just had to remove the import of that clientapi class. +Thus the patch is still valid and I still whould like to apply itIn short our interrupt scheme does not allow us notifying particular thread and thus we have to notify all of them +So to not break a test case which excluded the main query I routed only requests that use the to the new method that scores while creating the docsetThe next step is to throw an exception if someone tries to exclude the main query while using the +AlexanderUnfortunately it does not make sense to append them to richpage component +for patch +The script does not delete old generated Java files +the same one since hudson had some problem and I want it to run the same patch again +Is there any update on the status of my patch? I dont want to start work on a new patch to deal with unique key constraints until this patch gets committed +But I can certainly make the fault appear and disappear simply by enabling and disabling rampart in +Committed to trunk and +Maybe retry the async dispatch fix again to see if this was the failure you were seeing +is the correct version for branch +Sorry I just noticed test at the bottom +How about simply merging it +Did not create all Japanese messages yet +This patch fixes the signature of get and adds testcases of +I used revision this can help +It prevents from initialize properly in Collectors vm +Yes we could get this to too +You should be aware of this fact as in order to use you must explicitly enable ll fix it once we have a chance to work on the +Yes +So in this case the list version should be good? It triggered more full than simple version in my benchmark perhaps because the threads were delayed by more complex codeallocating a bigger block from completing activity that would have otherwise released some memory +Yep works as expected now and will be included with the next update thanks +too bad Jira does not allow editing by person who submitted issue +tgz is created from scratch when I wanted to pack a scenario for you +Source result set +Its working fine now +One other note +RTC work item defer minortrivial issues to +Also this would require keeping the old plugin repositories which would originally have been consolidated into one file +Perhaps just telling people dont is enoughNot sure about removing install things via Makefile but I would agree the scope is probably outside the level of +They would probably be added to the view toolbar rather than the main toolbar +thriftfs contrib with changed encoding for reading and version of double post and I dont know how to remove itThe comment should be that it was enough to change into latin for me to get the binary data to work see attached file above +Generally we should add a +Is there a test case that could have caught this issue that we can add +If so I suppose the alt syntax could be extended to allow this flag +Agree on carls opinion +This issue is superceded by all resolved issues with a status. +Pull requests merged. +Its great +Note that all of the Remoting issues linked to this issue are still open because they apply to both the remotingGA branch from which will be released and also to the remotingx branch from which will be released +Have not seen these since ie not reproduceable but lets keep an eye open +Retargeting to because of schedule constraints +Added doc details per the bug desccomments +Should we backport to? If its easy it might be a good idea +Marking this done +Followup patch +Yes thats pretty much exactly the approach I want to take +I hadnt thought about multiple parsers +Without such locality your first fractions could cause all the blocks to be read +Spinning a cluster to give it a spin via Hive +Upon further testing I also found the source repos are now pointing at the asf repo instead of incubating but there is a small bug with the dir pathing for the components +Id suggest starting the so version at and going from there +could you please explain where you shut off the parsing from Tomcat? It would appear that you are adding the servlets as well as Tomcat so you would have double the servlet based objects created +sounds good to the DOJO listIE has core language bugs that we cant support +is not known for its scan performance and this is one features to point to where we allow to not even look at another column family unless a filter is matched for potentially significant speedups +Also see and which have been marked as duplicates of this entry. +Hi Venkatesh no worries please disregard the Jenkins failure for now +So close it +Jon Agreed +Thanks Ning +Now I wouldnt mind if you assigned this to yourself +Since this is removing bits of JOBCONFXML should that be removed from as well +ReminderUpdate in kit and web release notes to reflect last minute issues +ThanksC +What are your ideas Edward +Committed to TRUNK +Ciao Ivanthis issue should depend on please test one of the latest nightly buids and let me know if that works. +They also have a kind which represents what kind of token they are and service which represents which instance of the service the token is for +Again I dont see how to do that with JIRA hereHeres the feature Id like +Sorry about that +New LDAP based unix group and committee group createdYour LDAP password which is for logging in to AND for committing via SVN should be your old passwordPlease email root if any of you guys have any password troublesMailing list moves coming next +Thanks Ted +DavidAgreed +Im going to move this bug out to I think unless someone wants to work on this for +I would need to look into it to be sure though perhaps it can be solved using some kind of lazy loadingrenderingWe should also upgrade to jQuery see +It seems the trunk does not have this problem but I still think my patch is the nicer fix because it makes the code more similar to the and the +Unfortunately not yet +On top of fixing the perspective we may consider spinning the stream of bits making an helix +Verified on revison +I tested this with eclipse running on a Java VM and the little AOP markers are shown! Sorry that it took a while +therefore i reverted my commit that moved all the login modules to the spi space which is exportedso lets reduce this issue to the typo +David your patch has been applied in the WAR plugin Thanks +Ill keep you informed + +The object does appear as deleted in my collection but it shouldnt appear at all +i have the same NPE in do method of class and also in get method of these classes when you get the files list from a you doesnt test if files list is null or the problem is with Synergy SCM you can complete a task without checkout you can have a in without file hope it will be corrected as soon as first contains only the correction for the for the classes as described s better to fix this in the Synergy SCM provider +I see fair enough +isnt an agreed approach for XML parsing in in the developer team due to the additional dependencies JAXB will be the better choice but this is pretty much work +I dont think we should do this +Aaron out of curiosity what happens when you run that same benchmark against an NFS drive? I am worried that the caching is going to be useful for anyone serving up content from a remote FS even through the JavaExcellent question +Agree +Made some improvements on FLE and the test case please make sure this is ok +if you can fix it easily go for it +resources script +Olegthis means that the parameter contentCharset in has no impact how the request is encoded? It is only used for encoding the response if the response entity hasnt this information? Only to clarify this for meThanks +This if why it is how it is now +Patch for the and projects that includes Inclusive Gateway for the patch Tom +So basicly wath it would need to be done its a change to display a more comprehensive message on the screen instead of all this rightBecause if thats so I can take a look into it +Sweet +so possible to have something better +already in branch. +So in the Mapper class it only get one tuple each time how can it get multiple tuples +The best would be to change the mechanism and dont rely on setTimestamp +just a code style tweak a test to make sure we get BS with random access filters and randomization of the parameter value in +It will still exit from the CLI on first exception if session was marked as batch +Hitesh can you please open a new jira for that? I went ahead and committed this +I think it would be good to get the names to reflect the reality that CDCFoundationJSR is a of the JDBC specnot that JDBC is an extended version of JSR +should not be editing anything under the generated package by hand +That should dothe trickThis does not work! The part headers are still included with nothing after them! What would help is a method that allows us to remove the header from the part entirely! If you look at the example I sent earlier +This patch is just to quickly plug the action but I think there is still some little work to do in order to improve this action + a javadoc warning in the earlier patch +I was using JDK Some tests dont pass when is +IMO this issue generated because body background color set as a black in the at line there is a class set background color as grey for alternate rowThanks and RegardsBrajesh Patel +Markus another race around updating to the new index and looking for the size of the index +it looks good to me +looks good +Yes there are some prolific users +Done for trunk +jarwill cause restart on both +But then this is actually a bug in the +I took a look at the patch and am wondering whether it matters if is created using the path passed to the method instead of letting the constructor determine everything from the request +It wont take severity into account but simpley process errorswarnings in the order of the bindings +remove sstableLock +break the tight couplingAnd what is the advantage of this +Scalate just delegates to the Scala compiler to make the template so it doesnt necessarily grok the function declarations and so forth wed need the Scala compiler plugin to help with that could just annotate the objectsclasses that contain tags or we could maybe analyse all the functions used in all pages and generate an XSD from that? +bq +Collecting related packages into one javadoc build is immensely useful as well as propery hyperlinking say the object from all the other packagesA secondary benefit of improving accessibility to the javadoc is that it encourages us to write more documentation +Theres still a long way to go before we can retire the old framework for good though. +I dont seem able to assign to myself perhaps because the issue is closed +That sounds good to me +parent +were not developing Deltacloud any longer +Closing as per statement +prior to generating release notes +As to changing its name that is hardly an option unless you want to break a lot of code that relies on that particular name +See second screenshot in doesnt suffer from this problem but thats because the PDF style doesnt use the jboss styled colors for the warning note and important elements +Thanks Jesse +SubbuDo you want to put the patch on review board? It is of decent size +Yep aware of that still nice to have a ticket here users will have something to look at in the release notes when we fix it. +out filesThis patch rotates and keeps last +Thank there still any glue code that needs to be written to achieve this? or is ready for testing? Are there any instructions somewhere on how to build any with this patch so I can test it +Attachment has been added with description input has been added with description generated fo you have reported similar issue. +Remove duplicate declaration in C +and both print class names as their exception message which then shows up as a class name where youd expect to see a message like after Caused by +Does success mean the abort was a success? What situations can occur that abort calls the error callbackAlso the example contains bytes sent stuff in the success callback so on success abort returns how many bytes were send? If so what is the point of that since Cordova usually the size of the file to storage +It makes it easy to isolate tests for individual things in the whiteboard +Ive committed this. +This patch will fix the issue described in +thanks Uwe +Since it is a user API it would be nice to improve itYou mean hard to read as javadoc or in the code +HiIm getting this exact same error +And yes we need to clean up those other subprojects as well but I wasfocussing on core at the time +Wrong consider offering a patch for instead of writing your own system +Is this related to + object added +But not all +Thank you very much Myrna and Knut for resolving this +Increased the default startlevel for new bundles so that core bundles are started before new bundles +fixed in trunk as of rThe patch has been applied with slight changesthanks for the patch +Maybe I will manage to rewrite the testcase without dropping the collection +i agree with slava combobox are the most flexible +I just wanted to emphasize that it is not a weired configuration I have and because it spent some hours on figuring it out that my time was not in Brian +I ported the SWF samples to use the new form tags and it seems to work just fine +Mike This class should be in the otherwise it will appear as if its part of Java Lucene too so once Nick gives us an OK for I will move it and commit +Martin have already done it everything is fine + +I tried to mimic the way transformation specific parameters are handled w pdf renderer +I understand its Im saying is that there should be a constraint and a reasonable exception when that happens +including configuration and more +Fix verifiedThanks a lot! +Update to handle requests for initialization breakdown relative to job process delivered. +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Rescheduled to to has been releaseClosing all related resolved issues. +Resolving as its duplicate since fix there will fix this issue as well +All processes are on the same serverI think I forgot to say that the session stays up in the client process when BSV and client want to stop simultaneously +Besides fixed typos the guide is the same as in Beta +I cant +Attached is the file containing the cosmetic issues I identified +Note you must include the units in the string eg px +lets put them in liblicenses then imo its too cluttered all in lib +moved the Java notes to a new issue as it is not really resolvable right now whereas the main topic of this issue is +Right now we could reconstruct the actuals based on the time of the time entries but we dont have any historical data for the estimate +Tim have your already dealt with this kind of issue before? How do you cope with it usually +Attached a patch containg a fix and a have updated the patch as I noticed some errors in one of the regex tests that I have tidied up +This would be best held as a discussion on the Commons Developer list are you subscribed to that +Thats strange +Also if we dont use a prefix a closure could not be accessed anymore from a variable as it would always be evaluated. +If you still experience the problem please create a small test project and reopen +A workaround for this would be to just make ENV be a concrete class rather than a singletonized Hash object +Another one off the list. +Its not a JCA subsystem issues because those thead pools operations are defined in threads subsystem +avro using provides a generic avro Writable implementation in see the test in srctestjava to see how this can be used mvn clean install will run the whole shebang +Thank again fo all the help. +See original Hibernate issue for more information +Yes we can do it in place on trunk itself +All these changes seem to be stabilized over the pasr month or so +All test failed because there are a problem in userroles tests +But I will close this issue to create more scoped ones +It seems to me that validwhen was never built to work with mapped properties and therefore this should be an enhancement request rather than bug +This patch is sticks closely to the original test design without making manual entries +There is a high risk of overflowing the command bufferAh finally a material argument! Thanks Allen that makes sense +Just committed to as well. +all the time +Ive committed this to trunk and merged to +Ive just committed this +patch for the provideruses the previous method to check wether the host is in the non or not +fixed with rev refactors the query test so that the tests are run with multiple configurations +I dont thinkthe boolean would be necessary +bq +build number should be i I checked it works with build Is there a way to have an automated test to prevent a regression on this JIRAAlex. +As long as the esb examples can be made to work out of box via a runtime and users can manually fix if they got their own ESB variant then im fine for GA +Ill set that and test anyway however just copying the file yields a race condition and the deployment fails +Im directly closing this issue as we are down to failures on windows platform with the testsuite +This particular problem is fixed but still has some breakpoint related issues. +It would be great if the content would be overwritten if I update a bundle +Ryan can you reproduce +This test passed when I ran with +This affects all released versions +Denis is this caused by the build not using the right branch or was something commited to branch that did not make it into the main trunk +I have tried to review the patch but I can not make any sense of the original code +This patch isnt really going to solve anything for you except remove the Could not retrieve endpoint ranges prefix which is still technically correct +Committed solution fix patch to trunk. +Includes version from the into so fusedfs version prints it out +bq + +Heres a profiler trace of this issue +Also each time the editor is saved the file will be so that parser errors will also be exposed to the userHow do these validation messages look? Did I miss anyValidation MessagesCNDWARNING No namespace declarations or node type definitions existERROR Duplicate namespace mapping prefixERROR Duplicate namespace mapping URIERROR Duplicate node type definition namesNamespace MappingERROR Empty or invalid prefixERROR Empty or invalid URINode Type DefinitionWARNING No property definitions or child node definitions existERROR Empty or invalid node type nameERROR Duplicate super type nameERROR Invalid super type nameERROR Cannot have explicit super types when super types is marked as a variantERROR Cannot have a primary item name when the primary item node type attribute is marked as a variantERROR Duplicate property definition namesERROR Duplicate child node definition namesProperty DefinitionERROR Empty or invalid property definition nameERROR Invalid property typeERROR Cannot have multiple default values when the property definition is ERROR Default value is not valid for the property definition type ERROR Duplicate default valueERROR Cannot have explicit default values when default values is marked as a variantERROR Invalid value constraint ERROR Duplicate value constraintERROR Cannot have explicit value constraints when value constraints is marked as a variant ERROR Invalid query operator ERROR Duplicate query operatorERROR Cannot have explicit query operators when query operators is marked as a variantChild Node DefinitionERROR Empty or invalid child node definition nameERROR Invalid required type name ERROR Duplicate required type nameERROR Cannot have explicit required types when required types is marked as a variantERROR Invalid default type nameERROR Cannot have explicit default type when default type is marked as a variant +Ill investigate that and open a new JiraAlso I will open a Jira for a new tool to create POS Tag dictionaries that optionally checks if the tagset is valid maybe looking at the training corpus to extract the tagset using a cutoff +Metadata seems such a delicate area right now and as you say the tests are behind the feature devopment +dims +I have the same problem which seems to be fixed works as expected with the snapshot Im using version but still getting the same problem +Moving to to moved wrong issue by mistake +to scrubbing driver references from the section +Dimitris can this issue be closed +HiThanks +Resolved. + +for Marco Sandrinis syntaxAlejandro Guizar sorry for being offtopic but what do you mean by If you want to avoid duplication simply use XML entities? Would you have an example +I think it might be better to leave that to the user as its likely to remain a special case anyway +Heres a patch for this +Since all patches attached to the Jira is committed and this thread is becoming big and confusing I would suggest to close this ticket and create a separate ticket to track unit test fixes for hadoop agree +groupId should be redhill +It looks like a socket timeout problem +Hi ScottCan this one be closed +fixed as specified by Benjamin +Do you already have the fix for? That was a similar error where we were creating an invalid access node position which lead to an in +Fixed with close of all resolved issues as part of closing items. +Test case for MINA case for MINA like the regression is between M M because the test work ok with MProlly something in the last core modifications +Is there an instantiated Price object serialized in the testing jar? If so could an old version of have been used on this platform by some accident +Im not done yet but Id bet that repainting after you scroll will fix this issue alleviating you from having to repaint the whole component. +bq +I cannot commit this to the branch as SVN is not working +Committed revision added a new unassignable base role for the repository managers to inherit that has just the permission for managing configuration +Was any changes made to the code to fix the Deployment Planner behavior that you mentioned in your comment +The global lock leads to a potential deadlockThanks Dhruba I overlooked the global lock which we did not have before +Here are all the logs that contain that region ID +Emmanuel nutch seems to pass the junit test case without applying your patch to +Uploaded a new patch with windows line terminatorsPerhaps one of the committers could set svnnative on the canon files in the j directory +If I saw that message I would not know that the purge was not executed because the build agent is busy +Although you simplified it it still contains a bug in line to +Attaching the thread dump +fixed title +Committed to branch +All the order dates are shown in the calendar and the entity through usedCapacity field keeps the information for all these reservations +ssh dir on and let us know then we can enable the account. +I submitted to trunk and will backport to and before closing this issue. +What link are you refering to when you say that the link is not here? If youre refering to the support case link it is already in the JIRA issue +is committed so this patch is unblocked +Bulk close resolved issues +Please provied a diffpatch with only the changes these files have been formatted so there are a number of irrelevant changes in each patch changes relate to targetArch changes relate to includeDir uses those changes plus fix for the in version +Might be good for you to work on +Fine Rajesh +Later we added partial class names which should have forced a of the decision +Applied in revision with minor changes +You just cant have a using such a field without surrounding it with braces +GWTs provides basic support for rule flow elements however connectors would admittedly need to be an extension to GWT +This has been fixed by the change to the low level javaassist API +Closed upon release of +Initial attempt at ubuntu support +Tested +batch update of issues with fix instead of the correct was not updated +bq +Unfortunately I have no concrete case. +Minor change for Schema backward compatibility change is to add the default value optional double dclocalreadrepairchance +Committed for in rev +This issue is for a very lightweight HTTP Service implementation one that could run on very small devices so this likely wont be done with Jetty +Feel free to close the issue +trunk +in +env requires an agent on the other side to set the env apart from the info coming from the api itself +bq +There are new files +Please that Git commit ID eefcdbbfcfcbccfe fixes the underlying issue +The is written in a way so that when the user hasnt set the AXISHOME env variable the script will try to get it set properly. +I understand the problem but I am not sure about the solution +Sorry the version Im using is understand only problem that I see is that this design forces you to have spring contexts and requires that action classes use dynamic lookup instead of IOC to obtain references to the beans in the Spring root web application context + +Ive been mulling over the best way to approach this for several weeks now +But I have no idea exactly what semantics Gavin wanted for it +Thanks for the patch. +Patch for the test is on the way +I am reworking the tests to deal with constant values +Unfortunately after migration Clearspace to SBS first name and last name is not propagated to SBS profileBut at the moment were working on movement registrationloginediting profiles to SBS so well be able use special characters very soon +Well I forgot to say that there is a possibility to set report URL after its initialization right in the code +Fixed in svn trunk r. +Attaching +I can provide a patch if Im sure that it will be accepted +After reading the documentation I was under the impression that it worked in all environments and was using it should be made more clear under the usage section that although the plugin is enabled by default in all environments it should only be enabled for environments that deploys a apologies for incorrectly filing this as a bug +Workaround for this issue is in revision be included in next merge +In context of this issue also fixed CREATED responses building in PUT COPY and MKCOL methods +I will commit this shortly +regioninfo from somewhere else +Particularly for the XX TB case mentioned above where balancing in a new node from scratch would involve a huge data transfer with time and network consequences +Here is the slightly different patch I closing stale resolved issues +These tests subclass from but make no attempt to use the provided and so wreak havoc +I forgot to say patch is against +If you can provide a full trace for that error it would help +Updated patch based on recent commits fixed up the javadocs and a fewother small things +if were very conservative +Anyways i will try to find it out and get back to you +Closing as a dup of . +That would be fine. +Unfortunately there appears to be a bug in the thrift system +might take some time +Ive also added an implementation in Tika +This is falseRather strongblanket statement +Return number of bytes written not the length of the string written. +No probs Im not actually using it myself just spotted it when noticing +Will report back with results when that finishes +who else uses false I wonder? Would changing these profiles so they dont get automatically turned on when false affect anyone but QEBounce to Max for comment +should always be detached +Is a third party LICENSE file part of the products documentation +Thanks +An RSSATOM example is in the works +If we move all the code out of and into the plugin we have a timing issue as the plugin may not be loaded when the main activity wants to show the splash screen even if the plugin has the onload attribute set to true +Attached patch for reviewNote that the corresponding docbook updates are outstanding +I believe the fixes to and reverting the workarounds for it in may have fixed this +Thanks for the patch Maysam will look into this and commit the same +Patch from Brian committed. +I think it seems that we need to fix this as the point is not successfully counteredBut I am postponing the fix to. +It could also be modified so has not to break the method signature +Yesterday Ive checked another project and bundles from the were resolved for it +This how we are reopening the writer on the latest commit +With the installer plugin problem dissapeared. +Perhaps theres a more elegant way to deal with empty arrays in Puppet but I am not aware of any + +I ended up managing to reproducing it on my I copy code from within IDEA and pasting into the console I also see those pointed at the related IDEA issue tracker ticket which seems to be related to some Suns specific Ill close the issue. +Admittedly I dont yet have a good benchmarking setup for these spanqueries yetBut from doing a quick test on a k doc corpus the on a common term like the tookabout half the time +This bug has been marked as a duplicate of +Not sure of your component strategy but this is related to the component of the toolsThe parser barfs if you replace the HQL WITH with any invalid keyword say XXXX +bq +I removed the prefix exporthometmpjagtmpautoderbyNregressioninstalllib from the Java classpath and broke the lines manually +Thanks for the patch Knut +Transaction cleanup is affected by the Work class not porperly cleaning up the trasnaction when an is this issue as I cannot reproduce it +You say it sometimes results in logFile paths which cases are a problem? Can you describe how to reproduce this please +Thanks +The error I am experiencing with is described in issue + +Luke I agree very much +Hi MarkThank you for your work +this was in a relocation right +The resolution is submitted to the board and received an acknowledgment that it is now on Sept th board agenda to discuss +But now Im realizing that is probably caused by this issueI wonder if could just catch the exception and block instead +Even Id be interested to have this feature which is to have the list of in a separated file out of the http section +Yes so it appears +hiverc before running any query +Make sense +The patch is correct the location step cant be empty +Committed. +I was not aware of the necessity to have regular analysis +This was a problem with my environment +Great work guys! +Replace it also in the configuration of the +Thank you. +Visual Studio error cannot apply binary patch to dotnetsrccontribJsonrBinCompactNew without full index lineerror dotnetsrccontribJsonrBinCompact patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinCompactNew without full index lineerror dotnetsrccontribJsonrBinCompact patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinNewt without full index lineerror dotnetsrccontribJsonrBin patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinNewt without full index lineerror dotnetsrccontribJsonrBin patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinNe without full index lineerror dotnetsrccontribJsonrBin patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinNe without full index lineerror dotnetsrccontribJsonrBin patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinSilverlight without full index lineerror dotnetsrccontribJsonrBinSilverlightll patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrBinSilverlight without full index lineerror dotnetsrccontribJsonrBinSilverlightdb patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrhm without full index lineerror dotnetsrccontribJsonr patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrSrcDocdonategif without full index lineerror dotnetsrccontribJsonrSrcDoc patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrSrcLib without full index lineerror dotnetsrccontribJsonrSrcLib patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrSrcLib without full index lineerror dotnetsrccontribJsonrSrcLib patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrSrcLib without full index lineerror dotnetsrccontribJsonrSrcLib patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrSrcLib without full index lineerror dotnetsrccontribJsonrSrcLib patch does not applyerror cannot apply binary patch to dotnetsrccontribJsonrSrcNewtonsoft +Also TLF and OSMF are NOT optional +Instead of based on file name choosing block placement policy during file creationappend would be more generic +The node will never respond to HTTP requests all ZK connections time out very high resource consumption +The lines of XML config that this patch saved were significant so if theres a better way can you please enlighten us +The idea is tomake the code modular easy to understand and simplify furthermodifications +Is there any status update on this? waiting to hear back from Weston on this case +Hi LukaszUnless the reporter or someone smarter than I can correct me I think it may be closed +BQ see associated Karaf issue. +Any progress please? +We used to load the table desc then acquire the lock but we should swap their order to guard against the race between alterrename table and snapshot operation +Once your code is in the right format your more likely to get a committer to spend some time with it +Untar it and it compilesOne note we need to do a fresh checkout before running ant otherwise there may be garbage inside +and me +Patch which appends arbritary elements to the surefire first patch was delete the first attachment and use this using testresources and resources the resources will end up in the jar this is not what we if using a targetPath for the testresources these will not be available on the classpath while could also be external resources which has to be available while executing proposed patch would fix all these issues +No reason to change this +This patch adds that capability and also added a test case for it +Why did we get rid of in the explain plan output Table Information has been replaced by Table Information without a new line +getTracker was removed because the static variable holding the job tracker was removed +As a minimum we should provide a comment indicating that this feature is experimental and may change in the future +use where applicable for more readable code This is a good suggestion except for one thing Its possible to include quotes within a delimited string +Which version of Spring have you been using when encountering that stacktrace? The line numbers dont seem to match with a few versions of that Ive checked +you can reference to the current user locale by +neat stuff +Sorry +Do we want to consider supplying open timestamp from the master tooWould that close the holes in this mechanism the ones that we could have if the server times diverge? Building a mechanism based on comparing server times will work most of the time but therell be folks who will have drifting clocks and then well have new interesting issues +And the CI also failed for Job for this under +It can return from there +It is used to capture the JAXB generatedrequest wrapper bean and the element name and namespace for marshalling unmarshalling the beanThe default value of localName element is the operationName as defined in annotation andthe default value for the targetNamespace element is the target namespace of the SEISo I think we should use the as the target not the one from implementation class +Im thinking that I should make it into an xdocs article and will check it into srcdocs +add code believe this was fixed as a of +Added the null check. +fixed in trunk +How are you running it? What do the logs say? Have you tried running it after a make install +The question is ejb starts looking for a security manager by doing javajaassometthing +Apply the current patch before applying +I see the problem with the patch +Would that fall under this issue or should it be a new oneWith this feature you would have two setups one of which would replicate from the other +The module is gone now but I just checked out the server client and plugin assemblies on a Windows XP Pro SP machine and none of them have +svn ciSending assembliesDeleting assembliessrcmaindemosDeleting demoswebNOTICESending demoswebAdding demoswebsrcmainAdding demoswebsrcmainresourcesAdding demoswebsrcmainresourcesdemosAdding demoswebsrcmainresourcesdemoswebAdding demoswebsrcmainresourcesdemoswebDeleting demoswebsrcmainjavaAdding demoswebsrcmainresourcesAdding demoswebsrcmainresourcesresourcesAdding demoswebsrcmainresourcesresourcesdemosAdding demoswebsrcmainresourcesresourcesdemoswebAdding demoswebsrcmainresourcesresourcesdemoswebNOTICEAdding demoswebsrcmainresourcesresourcesdemoswebAdding demoswebsrcmainresourcesresourcesdemoswebsrcAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainjavaAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainjavaorgAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainjavaorgapacheAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainjavaorgapachekarafAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainjavaorgapachekarafwebAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainjavaorgapachekarafwebAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainwebappAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainwebappAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainwebappAdding demoswebsrcmainresourcesresourcesdemoswebsrcmainwebappDeleting demoswebsrcmainwebappTransmitting file dataCommitted revision . +Resolved for a while +Like I said syntax is fine but properly not existing +place for the original cause of the issue +Hi GeorgeThank you for your checkMy issue is tracked as nowWishes Anton +The screenshot was captured in Mac with running VM and pulled to latest master branch + +It is really annoying I would fix it in if that is think this is only Windows bug +didnt we fix this +seems to be a bug +Move the fix to +We dont know when this will happen but it is innevitable at some pointIt would be helpful if you could outline your use case and why switching to would be a problem +svn ci Sending Transmitting file dataCommitted revision +I tried it as a plugin in Nexus OSS but it does not seem to be recognized as the Nexus plugin +IT done +Deferring all post soap service xml and console issues. +That version has not yet been released but you can try out if you really need this functionality right away. +And in my machine only AS can works correctly for this case +Its populating now +Excluded +If there are two xsds or wsdls with the same namespace like your the returns the first file if you want to have the second one +back port the the fix from +Tested +We ran into the same issue with full copies happening with each incremental +Im confident that the test failures are unrelated +you starting your nicluded page started from html tag? use uicomposition t reproduce on attach war file +Dan how does your review goes on? Need more samples test cases documentattion? Hopefully im now more on the xfire way of handling with xml doc by using stax and co +Your fix will come from there not here. +If the interface still needs to be changed it will happen under the issue requiring the change. +Ive already made my commit +I will take care tomorrow on changint the suffix +Hi I have tested my test application with the fix u have given +Patch prepared for passed with the patch +Im attaching the test output showing where it is hanging +Exposes idcf lookups +should the switch cases for MOVECURRENT and MOVELASTCHECKPOINT be switchedGot it this is what you meant by switching +rebased +adjusted subject since its not the repository access servlet that needs to do a reconnect but rather the rmi layer +I should mention this change is available with commit bccfecfeebeee. +Every kafka broker maintains data structures leader cache the list of alive brokers the broker id hostport mapping for all brokers +Its a subtle but quite important difference that is seen for example in the fact that the Importer interface has no dependencies on any XML handling it deals solely with and objects instantiated by the XML import handlers +That is worth a separate bug reportIll have a look at the geotools code and report it if is an equivalent to not set +There seems to be a lot of unrelated formatting changes in important classes like etcCan we factor these out and do any formatting changes separately +Committed revision . +Not something we catch +The reason of failure was inconsistent loop structure +Please try again with this script but you will need to change the userName password and databaseName values to match your you build the Roo source from Git and try again? There have been substantial changes since the can you please attach the DDL? I can do extensive testing with this in I saw the th Nov date and thought nothing had changed but moreover I got this mixed up with +And for all that it doesnt really added any needed new featureI dont mind veering away from strict pass by name as i have already said +I added rendering for TTF fonts in revision +James can you action this and close if its not a bug we can fix in Roo +A workaround is to reverse patch the change in r and to build Transports with that +Committing shortly +bq +Taking the bug a fix is on the way +In that case request will be data +Youre fixing the consequence not the causeThe underlying issue is that collections of elements should not be processed in the second passSo Ill not apply itYou can use it in the mean while I imagine +I just committed this to both TRUNK and +Am working on it +I would thankful to you if you can help with those +Sorry not my fault you did not add any test +Its simply using URL instead of URI when dealing with the connection url +So if the physical files was located at Cusrprojectscurrentjbossnetworkhypericruntimesserverdefaultdeploy it would be great if what got stored in the config response was deploy +What version of Spring Integration you are using +Chapter +Hi Ann no sure to understand why you dont buy Simons proposal +Have tried a variety of configurations and also created some basic tests for this feature to verify defaults and other basic configurations + +in JBDS +why I cannot use prototype param for localObject? well it was not working when we started with ROP +Screenshot of empty news ane empty feedburner there is no entries in the Blogs or News site Central returns the correct message No entries found. +I guess its safe to assume there would be at least one more release before +Waiting for the new catalog and new UI to kick in to work on s open a related jira to add a flag to turn off this functionality when we have enough work donein the catalognew ui to fix this? Or else I can rollback the patch +Right +Waiting for to be available as fixed version to move jira to resolved. +I agree this can be resolved for now. +Closing thanks for explanation. +great thanks! to help! +Also let say that we applied suggested fix and also solve the issue with importing all columns +Thanks Suresh! +Allen is reaffirming my point that we dont know the complete set of configuration formats +Please check it has been applied as expected +Thats not what OTP intends by includedapplications in the +Applied the patch to r +Attachment for method getFormatted implemented docs updated +As Part of the patch provided to this JIRA I have commented out the complete junit test for the below mentioned modules as due to any issue that needs to be resolved forsamples itest and few test cases in module +Basically instead of using a Im using a to store the map +Dave what kind of invalid configuration only gets logged at DEBUG level? Typical configuration errors should show up in the error console immediately +Closed after release. +Lets see how the Jenkins build goes this time +Divided is another similar option +Thanks a lotThe patch passed my test case toowhy this issue is not fixed in version +Set Urgency to normal +Still not fixed for Version I see. +I dont think the should use the naming store cache as the are themselves cached so an extra level of caching would just add more complexity with no benefit IMHO +Thanks Willem. +Again this server starts up and the service runs fine so long as jms is removed +I suspect that as long as the DLL and the Xalan executable build correctly the workaround is OKI will leave this bug active in case we can patch the project file for the next release +Clicking on help does go to login page +Theres no reason it needs to be there and I dont think we should rework the test to account for it +Im confused what does the make compliant actually mean and was this work completedIm looking at from and the MANIFEST is the same as in the release and certainly isnt has anyone seen jgroups working in an environment such as Equinox +Im pretty sure this is fixed will give you a solid answer after testing today +But I agree that is confusing so I replaced it + +Must use gnutar +There are if else statements in all the methods there +After some observation I realized that there is no way of modifyingthe Logger permanently since the Logger is loaded from which is in +Its best to let Jira handle link to its issues +That is not the summary Im referring to +Forgot to add Rakesh as the main author of this patch to +Testcase for reproducing the problem +Ill update you once I got it run over certain HosseinThanks a lot for the info! think one important thing to reproduce the problem is to cause clientserver failover due to timeouts +Gem names adjusted to remove the torquebox of now have even easier but docs still need updating again. +Committed revision If there are any problems caused by this let me know! +Resolving as this is fixed +Attached patch implements this functionalityNote to retrieve the value of structure corresponding decoding functionality has been implemented +Cancel woks as Emily find Jodys hints useful but will add a you press Cancel nothing will be displayed in the Table view and if you want to see something or have the message again you have to restart uDig +Unapproved licenses coresrcmainjavaorgapacheoozieDeleted the above file as there was a RAT warning against it +Also the driver is always returning the result of queries so result seems to be a reasonable name for the table +Leaving this open to fix. +relative offset for content area on frame on inactive tab has been fixed too. +Thanks for the fix but I misspoke earlier +Why not a separate Maven module? Why a separate repo? People arent going to want tools shipped with +JSR +This should have a single well know place it logs +Patch against svn trunk version +I remember sometime back doing a test cleanup cleanup effort but cant exactly remember my motivation +according to this change this patch doesnt apply any longer +Ultimately I think just passing a CQL query will solve all of this but if someone wants to do this now what I recommend is having external utilities do the conversion then use pigs parameter substitution to execute and pass in the slice converted parameters. +Thanks a lot Ian! +If you think of a service contract each interface can have several policies which specify the behaviour of the service +Hi EricIt is hard for me to understand how the patch solves your problem +Closing. +This may not be required for the completed file however from consistency point of view your change makes senseComments Please addmodify existing unit test for checking token is set for last block +We can remove it if desired +Resolved as part of +Move to no known way to fix this +I worked with nutch for one month for a projected project never executed +IMHO its not a PLX issue but a MWAR issue because its possible to use a custom manifest in the the jar plugin +If you only need control over the assembly name while its in the target directory then you can use the finalName configuration in the assembly plugin configuration block. +Please confirm the current status of this issue and if its incorrect change it as required + +This looks something we need to investigate and fix soon +Wow + +Christian can you please try with the latest? Should be fixed as we now ignore annotations interfaces and enums as being subject to injection +Super +This document is required for GA +Radim Could you kind educate me why we prefer to use bigger prime number in the hashCode function +It works for meClosing issueThanks. +Here is an example while you wait for me to learn how to write JSF tests click startclick redirectclick dummyrefresh browser or retry url +The patch doesnt appear to include any new or modified testsAll my patch proposes to change is s class javadoc +Part of tck now +Yes makes sense for me too +Since there is no session id we can use a custom url parameter with temporary id that can be used in the constructed key +plain text of this still an issue +what was the problem exactlyBTW nice to see you back in JIRA Andy +Bug has been marked as a duplicate of this bug +tests included +Its latest revision is at November Can you check the git mirror of giraph +in particular locking down to should get the Camel build working with all recent Maven versions +Ok you can have a small VM but if the performance starts to affect other then we will have to address thatThis will still be weeks away +Finished backporting to the subset of which restores the original lock timeout on completion of and and see if that helps with the intermittent failure on codeline +Move it back in if you think differently +I also need to clean up the Tomcat example in hibernate and retest the WAS examples +Looks good again +An FP is essentially a minor release small changes regressions are acceptable +I am doing the latter +Well as allen said on My technical objection is that there is a high likelihood of getting this wrong +Created a link to the Mailing Lists page in appropriate place +A separated operations to get owner info of a topic would be better +Any chance this could be fixed or a patch supplied for +The XSD used for validation is an old legacy xsd that imported a namespace without any ref +I spoke with Mohammad Islam and Virag Kothari +Updated in revision . +yes it does +What I meant was more of a SW organization point of view +Thanks! +Just putting it in here for out of +I think it would be better to check in the generated files but not the generator +Verified in ER +Otherwise it is missing classes that may be expected by other packages +So that is needed by the client in the version in EAP and is needed for JDG +It depends on the order that the class loader returns them +finding AndyThat could be the cause. +Please realize that I have close to a dozen tickets now in some stage of review +This action of creating a containerafter the prepare is caught by store and the error is thrownI am not sure the right fix but here are some options somehow figure out how to move the work to prepare rather than commit so that there is no work to do at commit time +command will be consistent with the add. +Hey Michael any chance of looking at this weird issue yet +Is there a test we can add for this? Thanks for getting the patch in Yoko +Committed. +Theres a lot of new stuff in there +Quickly looking at the method in question shows one way of reducing the performance hit instead of iterating over the collection every time keep a cached dataLength member and return that +Should be closed or marked as an enhancement request +Instead just let it become an identifier and then get treated like a method name +Its included now and is default jetty version. +You can either give the new patch the same name and Jira will nicely gray out the old ones or you can give it a different name +As agreed on the ml fixed +Attachment has been added with description the property is is defined on the table in FOP and probably earlier transition resolvedfixed bugs to closedfixed +sonar subir but it still will the target dir the will be used by default to store the xml filesI guess if you agree to do something with it it wil be a low priority anyway as there are workarounds to achieve it. +I tested this locally by verifying that tmp is no longer allowed but a valid path like hometodd works fine +Could be +Without a patch this isnt realistically going to get added due to the superiorJSTL tags supporting this feature. +databinderfcst +getting the commit log context and switching the memtable is not done atomically with respect to writesIsnt this easily fixed by moving the context request to after the loopbq +The listener is correctly notified of a dependent MD change but this happens before the is notified so it can provide a accessor + +please apply the patch so we get more green +I think it would be good to add the CAS Editor there +Looks like the same bug as +I would be happy to do it. +The mapping file should be written manually + Oleg +My bad version check is another Jira. +Thank you ChuanTodd please do post your comments +We can not ship this way was there any information on the WAS forums on this or in googleThis is not a good way to go +An update on this one my colleague found that if you pass the non serializable object like DB connection to injected bean then we get this exception +RFC sticks to the definition of a number which must start with something different than a when there is more than one digitHowever I dont see the value of rejecting such +Verified +Ill fix those ASAP +Thanks for the patch Dave. +How about a file to set the various props? If file exists but is empty it could be a simple marker that this is a core and assume defaults for all settings +This was included in. +Marking closed. +Andrews fix for has same effect as the one I suggested for listTables where we only look in the info column +Fixed in the project will be in the next merge +Committed to trunk and +The patch retains Eclipse classpath integrity +Hi ThanhCan you please make a test for this against? The tests in should be a good model +dependencies in poms that are not in the current not relevant with the database in place +Please raise new more specific issues if you are affected by anything similar to this. +looks like this is so close +Im closing this one +Thanks for checking DanielI will get your patch in today or tomorrow if all tests pass +So it may return confusing result there was no messages or messages are still not visible? Would it be useful to add start and finish timestamps as explicit parameters +is out +It is Brew +Over the next few weeks hopefully more commiters will offer up code to complete the plugin and we can publish it out to the sandbox for review. +Attached is a patch the trunk +Patch with test for the Annotation Ontology Mapping Strategy code and bug fix +I also had some trouble with dynamic revisionsBut I guess sometimes the version would exist and my resolver could make use of it +Patch applied with thank to Glen. +I am attaching the patch if anyone wants to look at it +Forgot to mention that the doc should briefly describe the outputs of LsXMLIndented +Converted this to a issue per Todds comment. +its even worst +Are you saying that this would go awayTalking to Stu it sounds like there are sources within the cassandra core that rely on the jar dependency +And it makes placement of exceptions pretty easy to understand +Thanks DagCommitted patch to documentation trunk at revision +We only need to update the Japanese translation for the following pages of the website index +It still fails with master with a clean rebuild +to branch in efe +Id even appreciate if you could clean those above SaschaSince you already have patch available for this can you please apply it Trunk +Initial patch for the UpdatedThanks Konst for your commentsI am working on patch will put it shortlyThanks Patch for MRThanks am still working on adding more tests will update the latest soonThanks Mayank +And lets discuss in the same issue why we have richbeanValidator tag in taglib +Still should expose the collection of not just the first one +So a cluster with a periodic write load which was too high for LCSs increased IO needs would revert to STCS of L only until the load dropped +fixed in revision after release of Jackrabbit Server bundle +See for further details. +The previous computation had a problem in some corner cases +TimIronic this should come up now +Perhaps you want it to have a more serisuo title? This is a major waste of time caused by known bugs +Attaching +Not sure if I should be choosing Done of Wont fix the documentation has been made explicit conditions on transitions are ONLY accepted for decisions +nextDocs are used +Does this mean that the issue will not be solved at all +I tried with both rubyzip and +Hi the fix was for not tested and got same result +Cant be reproduced with +I am not sure what the jbi servicemix build is +pending for the reviews guys +Per your previous comment Im going to mark this issue as invalid +reopen to set fix version to again with correct fix a possible fix in and would like to get some confirmation that this solution is good before I commit this fixAny feedback appreciated + makes the plugin use a maven Resource to include the binary goo rather than copying it by hand +JAR stuff doesnt just work +expandSingle condition is violated +Addition of a few existing test fixes +Removed macros and code formatted in modulescore applied Many thanks Diluka +Thank you very much Timothy +studioeclipsefeaturesmaven +why is that code even running in my case +Fixed on trunk waiting for a little testing before backporting on on as well +I believe it should definetely be a version with the eclipse artifact resolution +Its no fun debugging intermittent hung flushing or closing on a distributed cluster +The previous results was run on this good assuming tests all pass +move reporter from Steve to t need a development opened issue for this +Attaching a patch for the branch +Im not sure I understood your and are implementing this I missing something +Thanks John I put this into trunk and +x +should be added +Fixed thanks for the patch +Hey Zied thanks for the heads up + test fs rather than jFS +If we settle on as the encoding of arguments then I believe we must accompany it with a indicator of how it was encoded +But I couldnt reproduce your leak +Not looking to be a pain in the neck but I would like to close this out if it is indeed fixed +It must be something more subtle then +slaves name is hardcoded as a file name confslaves in hadoop +If circumstances permit I should have more time for mimej sometime next weekCheers Heres another attempt at fixing the problem +It seems that there are some problems in Jenkins or the build +Improved documentation around subclass hooks in. +resolved issues for already released versions +I suppose that is when the runtime name is invalid +diff between my modified file and the original file in the forrest distribution +Byte arrays are serialized in XMI as hex strings not Base +I just checked in the currcent code base and it does not have that bugcan I resolve the issue +Yes thats still an issue Im planning on looking into that tomorrow +To find the included stylesheet module a number of things are done but since the primary stylesheet has no system ID set it isnt relative to that URI so it starts to look in other places +Attached is a screenshot of the orphan Runjar processes on you tried this on or trunk +At a preliminary review it seam to me that that it will not be possible to support both with one parser but we have to point out where the problems of the support for both dialects I think we should add support for some Castor specific extensions to EJB QL +Sadly it was much easier for us to use a common subclass that provided synchronization between the constructor and all the static accessors than it was get set up to build our own version of the package then distribute and install it +Once that is done we can implement it in the the AS project for feature requestsThis sounds like a of. +Logged In YES useridThis URL is one that likes if I use this URL in my browser I get that addressing schema +We wont add globally named queries at the moment +Cannot reproduce with current trunk state +Patch committed to trunk and and to so updating Fixin unchecking Patch Available and resolving the issue +The more tests we have the more stable our releases are going to be +Will open another issue for multiple security handlers support. +for we might be force flushing due to the maximum hlog fairlyfrequently anyways +If someone knows the opposite please let me know and I will revert the change +Some kind of user fields apparentlyI have removed the offending snippet from +Open source Open mapping Good proposal IMO +hence the TODO there +Please clarify +Yes I referring toI want to know when Jonathan H will check in +If you need additional info about the submitted changes please feel free to contact methanks application coplet has changed in the meantime Now it uses the usual configuration mechanism and doesnt need an external configuration file +Solved thanks for reporting! +If you have decided to not use the configuration its ok but it can be tedious to modify a high number of files +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Patch looks fine to me +Will apply it right now +Thanks +Fix committed at revision for isnt fixed yet + Is it needed in too +However reload makes error occur againRestarted again deleted the store and layer hit reload catalogue and error againI receive this error for ever workspace that enabled is checked in workspace settings +JustinThanks for your input +You can reupload the same patch and that might trigger hudson +Added config properties for usernamepassword +Low went anywhere with this +NARROWCOLUMNS optimizes SEEKNEXTCOL NARROWROWS optimizes both SEEKNEXTCOL and SEEKNEXTROW +Link the generic dependency controller to the reworking +This is started for M but wont be complete until M rev of this is done minus the aggregate bodies which is in another issue. +Trunk only in my opinion +Now I need to check with NSS setup that that will need to wait till tomorrow. +Moving back to for now + +has been created to deal with the sorting of running jobs +We tend to use a special Java exception to realize this +it would be better to limit that in the config instead of failing at an assert later onSplit that out to +Theorically the spec does not handle this case but it is hard to find a use case when this could be a problem because most of the time the component state is not relevant after an ajax redirect +But perhaps we should allow that through another API in case someone relies on it elsewhere +Applied this patch with a small changeTo prevent a possible race condition when is nulled whilst in the middle of this method I read the member once into a local variable and use that throughout the rest of the codebaseTo be honest I have the feeling that there are other possible concurrency issues in this code but for now I wanted to allow +Nice follow ups on the comments! +We can get this in later releases of +Then as long as you follow the convention in placing modules in a subdirectory named for the module it should just workOf course it would be even better if the child code was changed to use the relative path between module and parent rather than artifactId +Hi RonGiven the hang stack trace is exactly same as I saw and commented in and I believe this hang already get resolved as part of fix forWhats the servicemix version are you usingWould you please try with the latest released Fuse ESBIf you still see the problem with Fuse ESB would you please append your testcaseThanks please find the SA that hangs on deployment and the source project used to create the SA +This requires explicit linking of schedulers and triggers +Id propose to remove all those filter actions in the entry editor preferences show object class attribute show must attributes show may attributesWe have a more powerful feature to filter the attributes in the entry editor the quick filter +I will check it again with some different code +Does this problem still occur? If so could you provide a small standalonetest? The document caching code does synchronise on accesses to the cache soits not immediately obvious to me where the problem lies +Unfortunately we wont be able to accept the code if you claim a piece of the copyright +Perfect +It will be supported by when I get round to it +Fixed in r +Any chance this could be looked at +Gregory patch verified at rThank you for applying it +Were experiencing the same behaviour we did in +Brian it has been resolved +The paranoid classloader solved the problem in revision . +Its a pretty small task to add the feature +I had a fair amount of functionality working nicely +Possibly my disk free space is too fragmented and its really taking java minute to find enough disk space. +Transitioning to closed. +Simple class to showcase you please let us know which operating system are you usingJust for you to know in jBPM we have removed Mina completely +find was moved from frameworkcommonservicedef to componentaccountingservicedef +I tried to include the sources in the main jar but your builds the classpath by only adding the dependencies with classifier sources so this does not work +re patch aIn discover when the user has specified an explicit log classthen we shouldnt do any discovery yes? We should just load that logging classor fail +Jon Yep now you got what is in my mindDuring the st run can we store region ids selected? Second run we mark as rerun and check only region id against this list +dynammic validations on step and toThanks Jaimin. +Simple testm which is a very simple test involving no failure is failing +I am waiting for feedback on this +Fixed another test error +There were some emails that I exchanged with someone using Grails with Eclipse and they were focused around not being able to create compile a class without a package declaration +This is a backend issue +Committed to trunk. +Sure I can help with the testing Werner +I also changed the scope of my conversational scoped beans to session scope +Seeing how fast API is changing Im to wait for their release of M before doing anything serious +Well that works fine but is it possible to change the sort order to descending via this collection sort +Im not using ssh repositories but maven repositories over https +Bulk close resolved issues +Should be closed since this is fixed in jBPM +seems to do this by using ZK to enforce mutual exclusion which means at its heart it too relies on a consensus protocol to cope with these tricky failure cases +Committed revision +That worked for me +This is caused by response headers with instead of the expected stringAttaching patch to do comparison of headers +Im saying lets make it unified by putting the content we need into a structure we can easily download from and have existing mechanism to cache with to save time +This is a research tutorial of what I was able to simulate +I have attached an alternative patch which iterates multiple times over nodes until it all blocks have been grouped or no node if left with more than minsize of blocks while trying to keep number of iterations low +Ill look at the cost functions taking too long in another issue if it can be reproduced. +Thanks +Clean target can not remove projectplugins because there might be scripts that listen to Clean itself +The trunk and patches are essentially identical and actually the latest Jenkins run commented on the patch instead of the trunk patchUploading the same trunk patch again to kick Jenkins now that has been committed + Committed to JBESBGACP with revision + +Ok +I did not mean increasing combiner buffer size I meant overall memory that is given to the process +This is a rather critical issue people are waiting to be fixed! So instead of suggesting to execute the phases locally by everybody how about providing something like an official version? Or maybe since we were told to not use SNAPSHOT versions of plugins? What are the plansJava is out now for quite a while and its annoying how the CheckstyleMaven toolchain continues to ignore this fact so please lets have progress here +As it is to keep my builds going I have to change my proxy url every time the remote repository goes down just to point to something else that I know is up to avoid the timeouts +Please find attached my solution to link to screens a locally available Confluence end user doc HTML exportThis is a first trial to share the ideaIf we are ok with this we should have a cleaner Confluence space dedicated to online documentationThe file could be periodically and attached to a special Confluence page to be downloaded and installed in theAnother option is to have the help application committed to SVN with the HTML files and periodically update them with a fresh export from ConfluenceWDYT +Really frustrating +This is basically a duplicate of and and the same concern applies here implementing the cache like that will result in a class loader leak +The library comes with a number of unit and integration tests recommend you see if you can add test scenarios there that reproduce these sorts of issues +Yes its ok nowthanx +Send by Fabian Christ on MLyes we discussed such chains a while ago and they are definitely onthe roadmap +Linking to the issue that introduced +Related to processing of headers we should consider how the current certificate login handler operates relative to the SOAP credential extractor +I see that has been fixed by Tim +On the other hand we need some way to handle dates or other parameters that cant be converted into literals +Closed due to the release of Tiles. +Leif can you review and commit +However Im not sure if using a Map with get checks is the ideal solution +This case will be addressed in a future patch +You cant execute such code via a server without barriers +Ive also added validating of types +Fixed in Rev +Updated documentation and tests. +I just committed this to trunk and +chaqnged from type enhancement to bug +Ok +Would it be possible to put and under their respective jars too +ill add that +Swamy need to test on Vmware set to critical as this si failing even for the fix eth didnt get additional ips when happens +changed replace with patternmatcher with your regex in the escapeQuotes method +The issue +Committed to branch and trunk +Max Ive made version corrs and Ive played a little bit with colors +The purpose is to use the same indication function that indicates that the submit has started as in the mentioned ajax implementations +Please review +BenIs there a spring standard for these type of tools within pom files? I remember seeing a presentation by Juergen about using pmd but it wasnt within maven +The patch is for trunk branch only +Committed to and trunk +changes committed please can you hit the Accept button if youre happy +images andor scripts and other content is missing and trickier specially the img src of the action gifs +Fixed +Using a single role to represent multiple privileges can lead to arbitrary combinations of role strings eventually turning into soupLike I said Im ambivalent +I am not calling the +Apologies if this does not belong m using JSF on Weblogic and Tomcat +rev enables the hover plugin to show plugin description per Jareks suggestionrev splits the page into +Im closing the issue +This can theoretically be changed but as of now theres no UI for that +The reporter should try with Hibernate as I remember Column issues including composite keys to be fixed with that version + issues solved in released versions +Could you tryand submit a clean patch for that +Please check that it was applied as expected +patch committed +this patch fixes the debug messages +That and the fact that findbugsfindbugs doesnt use the same reporting configuration that it uses when it runs inside the site plugin +If you objects are prototypes for dont they need to be scopeprototype +That data source does not exist in Derby versions before +Is this purely an issue of the size of your resulting jar? You can do this more effectively with a one liner with proguard in your build +a note that this patch doesnt address the testcases +This really really smells of backing buffers getting chewed up and spit out too rapidly +But what can I do to quiet this down? +I think practically its not a large problem since the write lock is used for only as you say +I follow now +Working for me +Ilya seemed to agree that we should leave it open until it was fixed for linux and not just windows. +Reopened because of users comment +Could not reproduce the bug. +This patch adds a and pulls the code out of +So thats whats going onTheres an easy way to fix this and a hard way +As we have to share the artifacts with EAP and EAP this requires to be unsigned although I am sure there is also a rule for that unsigned artifacts in maven repository +setting Fix targets to and so this doesnt fall off our believe this is a bug in Jetty +Is this for +I expected it not to work on windows too +The per thread context class loader is accessed using the Thread method classes used by an application will typically be loaded by a hierarchy of class loaders +Makes sense I was actually considering that but my newb nerves felt safer to go with the flow +FWIW we got the same problem once more yesterday +I second Marcos comments in that having a native wrapper a plugin for html audio that would allow us to background would be a good start towards making cordova the default way to handle audio on mob +UI Mockup with default is actually useful to users who have submitted jobs +This is still waiting on jarjar working correctly its not really our bug but we really really ought to have a mavenable jarjar release as our actual release dependency +We should document this behavior and provide the workaround for users who try to do this so they can go forward +Updated patch generated with so that it actually applies +patch to keep invalid characters out of index namesif you can afford to lose the data the easiest fix for this CF is to drop and recreate it +Since theyre primarily meantfor inheritance I think they should not be interpolated before theyre deployed +stocktraderdotnetcommon +Ive only converted the component tag so far +Suggest rolling back for further investigation +fix a bug in converting milliseconds to strings +Thanks +closed +shall I close this bug then? Really it is more me nagging you then a bug +One advantage of the old requirement methods was that we could easily distinguish requirements that were a consequence of a customer order versus those that were a result of +It must be related to libraries in RuntimeI create simple Java project with one Java source file that references no libraries +Attaching and with these changesM srcdevguideM srcref +All the core and contrib tests passed with the patch +i wonder if we should just reorder the code so that we validate the session before doing the zxid check +Thanks! +I pulled this into too +Hi HelgeI hadnt noticed it before but its come to my attention that two of the files you attached to this bug report and might contain Sun copyrighted code + +Can you please write what was the resolution to this problem +The patch do fix the problem that I can deploy a cluster and run job successfully without specifying value now +for ex my configuration files which reside in confconfigs +Completed +Perhaps this needs documenting somewhere +It is planned for version +Logged In YES useridI will try to post a FAQ on the forum if it isnt doneyet. +So I would have to run the build with Maven I guess + +We will back port it and test again +Ive just committed it. +Workaround set WarningIgnore for Preferences ToolsWebSeamValidatorComponent at conf call if it is possible for s possible +When we update to newer version we can look at implementing sprites for the rest of the common elements. +Seems to me the problem is much more likely due to you trying to use an entity name as an alias in the query +Im having the same problem using Spring Modules and Spring RC +Jenkins slaves are using default settings for ulimit which isnt a viable option once youre dealing with applications at scale +I cant say something like that will definitely be required but I cant rule it out either +docs added with svn revision +Okay stuff commited to trunk +This also jives with it not showing up on and showing up on was committed recently and between the release and release candidate +Merged to doc branch at revision +Slava thanks for diving deeper and finding that stuff +I attached the Log file while analyzing the project by using the command mvn dotnetfxcop +The patch has rotted +patch looks good +In a very small cluster clients may run out of datanodes and fail before retrying times +This seems like the wrong direction +Patch attached +Im hoping to make it next week unless something unexpected appears +Should there be parens around the size of the macro +well yes +fixes the nant test issues +Heres a draft of a full patch +We recently decided to use the date of the source file change as the idea in short format +Note Ive left out the libhdfs work and plan to address that in a separate ticket since I think the changes we talked about might need to be reflected in the existing read path making this patch more cumbersome +Carl adding code that is not much used is always no harm except a lot of maintenance and document pain +Uma the reporter clearly said they expected flushsync to work for local file systems +Ill get to it this week Harry can anyone get us the proxying stuff in place so we can start using our wikisthanksHarry +This does not mean that it is not set +The idea of a lightweight timeout is looking better all the time since we could just toss a thread pool at it and have supahcheap timeout capabilities +General problem for any to m because its more serious and eGit is widely was able to reproduce the issue and more importantly to solve itI have no clue as to what the role of the eGit plugin is in this issue though +I dont have set up right now +Should not we expand advanced section automatically in this case +Verified and closed it. +Could we add a class level javadoc to to indicate that it exists because uses sendFile? aside from that +thanks for the patch and please close this issue once you verify the fix. + Test could you include this test and see if this is also fixed by the hibernate stuff that we did this week +Including those libraries ought to be done already shouldnt it? I mean if theyre already in the runtime installation and you have a target runtime specified for the project +I think it is a good idea to go with the established naming conventionthat if a class provides a set of static utility methods it should be calledUtilsIve committed the class and the test +So I think if we are going to provide test stuff to users we need to be extra cautious about hindering our own development +Committed this to trunk and +Ah oh this is visual stuff worth a blog entry imho +Files include patch for the issue and two binary files ODG PNG formats of FI framework pdf version of Forrest generated lines length has been reduced for better PDF by NigelSome typos are corrected by CorinneSeems to be good to go +Its not beautiful but it works +If you find further issues please create a new jira and link it to this one. +Attaching a new patch incorporating Amareshwaris comment to do with logging of and +This fixes these bugs please patch in when possible +Thanks +The installer now installs all the jars but only the configs that you select +Havent seen this lately Im closing it. +Aleksey are you sure the test is from jdtcorecompiler suite? Im not able to reproduce the failures +Please guys help us +The real issue turned out that the equals was broken as it was not comparing the port component correctly +Im using the JSF that comes with AS the subversion of Gravel has the same issue +Correction to my above comment was not timing out for me from a bit of debugging I am relatively sure started failing after +Added the mappings I still need to add unittests before I can mark this as fixed +HiI just tried to reproduce this and didnt manageI see the normal behavior until the number of messages in the queue reaches the page size which was explained by GaryI noticed that in your code you use client ack and dont ack messages anywhere +Sounds reasonable to me +What do you think +Actually the real bug here is that the framework shouldnt allow you to specify any +yes but that will also imply that we should weight it in memory size instead of number of entries so need to use jamm which is calculation overhead better just remove unused because we know precisely when to do that +We can authenticate the DN there before the client sends any private data +Not sure why it was changed but in an earlier version of CXF there was no problem having a Provider handle a POST with an empty body +this is a problem with using Maven +I just committed this +This is very important feature +I would vote for gradle from my experience +That would make things a lot simpler +These are deleted at the end of the job atleast AFAIK the data is deleted but sometimes the directories themselves are not cleaned up which looks like a bug in HODThat said I see what you are proposing has advantages Currently HOD requires the temp directory to be world writable so different users can write to it +Ron +because you would still need to keep the references alive otherwise there would be no online documentation for older versionsI can keep still up for some time its not a question of server resources but more about legal stuff. +I think you deduce this later in test but it is never explicitly stated +Deployer +The unit test does not pass any principals keytabs in the configuration therefore spnego will always be disabledIm not an expert of the YARN code but it seems to me that you wont be able to get the right token when spnego is disabled +which uses XSLT elements such as xslIf you need a replacement to support the Saxon parser just put this file in the GRAILSHOMElib folder +The difference with my case is that my Item class has id field as String +And doing that gives me other fun exceptions +I havent had the crash again but I would like to reduce the memory and see if I can carry on as before +Made it an ImprovementCan you be a bit more explicit? An example could help here +JonYou are not implementing a proxy! Only a cache +Zach is this something that we need to do before shipping a of Beehive? If so can you take a look? If not lets move to TBD for the fix release +As a matter of fact Im updating this guide as we speak +Verified on +Regarding your comment we may not need to expose fileId in then you mean we should repurpose this JIRA to undo the fileId field both on the server and client side of +Which means that this API shouldnt be public +Im hacking on this now +jar forgot to escape the wildcard ChristianThank you so much for your reply +diff that only include the tests on top of +If we can identify any code from anyone who hasnt agreed then we should replace it but theryre several people we havent been able to reach and whom Im ot sure why theyre in the author lists in the first place +thanks rick now I understand why the change +files and directories +I made sure it took effect for the user I originally logged in as ubuntu for my ec instance and jeremy for my local vm +Reporters of these issues please verify the fixes and help close these few more automation runs result to verify this i will close this after verifying next two I think the KVMFILELOCK is left over on secondary storagebecause its unable to lock the primary storage and doesnt clean upafter itself +At this point we dont produce a consumable AST of incoming queries +Get and the test will succeed +Latest revision which addresses Namits comments is on review board +Deferring to this for Im just not sure of this one yet +This issue needs more discussion and the feature it proposes is not strictly necessary so I recommend that we move it to. +I commented out a number of methods I couldnt quickly figure out how to port +Relates to +That has a workaround at user level remove the dupsAre we OK with this patch for + +The way Im planning to do it if the first cores config is missing the value itll start off with the new default which might ignore any values in the other configs +Support added to SVN revision +Cancelling the patch for the javac issues +Our new Jca layer will be based on this case I dont think we need expose the subjectFactory as JMX Mbean +Removed in revision . +I should add my opinion is the other issue with user minimum limit can be fixed in a separate jira +Not sure about what we are talking about here +removed unrelated changed suggestions for this patchIve done some tests on our cluster and found the job that always got dead lock works well now. +I think this may be an issue with your unzip program +The purpose of this is to make view entities more visible thus encouraging the use of of to leverage a normalized data model instead of denormalizing the data model to fit a specific needLicense is granted for inclusion however this patch is not necessarily meant to be included as is +Cool +Id argue that the specification means the exception should not be thrown if the declare has an empty field but that if the declare specifies an alternate exchange that should match any setting + for the patch +Hi Sriramwould you mind sharing details of your use case? What exact commands are you executing? What are you trying to achieve here? the syntax +the patch all resolvedfixed issues of already released versions of Roller. +Sending daytraderejbSending daytraderstreamerSending daytraderwebSending daytraderwsappclientTransmitting file dataCommitted revision +About access control arent there cases where parents are not accessible but children are? So isnt it a problem anyway to traverse down the tree +I think you can save a lot of time by reusing already tested code since it is pretty similar to this task +I would say second wait is good enough for failing the test because thread sleep time is set to milliseconds +You didnt answer if removing the implements from and writing a new empty class class extends implements works for you or notI cant understand the need for to be proxy to another class considering it is already an hidden implementation returned by a factoryservice locator +Special characters in text in HTMLXML output isescaped based on the value of the setting +Thanks Mark +Handler for case when the substring is longer than the specified string has been added +Change to so that the enum does not try to populate the reverse map upon construction +solves also for similar problems with regular text form also see also Bug for a more proactive management of excessive sizes +Itd help me test your patch +Want to take a stab at a patch +the proposed solution still uses javascript to initiate the roundtrip so what really is the advantage here? +Looks more civilized I hope +harry was an attempt to refer to the userid or auth id and Harry as the person +NET Service +libgriddevbin +the duplicate error +Before we choose the underlying technology Im going to make sure all synchronization passes through an abstraction layer +As I am a noob does anyone have thoughts on what the proper approach might beThanks another class is probably the correct way +We already talked about waiting for rehash to finish before allowing cache start process to finish in order to avoid put or similar calls before rehash has think this would fix the invalidation issue as well +The other database examples is based on jpa crapola which is complex and overkill for most peopleWhere is the link for this +I tried also increasing the property but the collection is never fetched +Yes works just fine here +Then to find plugin jars we are just scanning the classpath for certain named jars +Removes the try catch block +The fix doesnt work on AS environment +Code that participates in the build should in general be implemented with a separate Builder rather than a +Well there is only finishStage in the patch which is of no use alone +Click on the Search button then one of the dataScroller links +will have to determine the proper course of action in the case of a reactored release but otherwise should be doable +Resolving +OK +a contrib testcase is failing with a whitespaces up patch with trunk +However Camel FTP has had a major refactor in Camel compared toNew Revision +We havent tried upgrading to the latest version yet because its a production system +Fix version says it was fixed in +This was already fixed in additional unittests +Hi Alan I think we should stop exposing to users otherwise this will keep happening +Change checked in at r +This is reason why Im asking if Jira should be created +Knocking down to major +New files that need to be added to SVN under the srctestresourcesunit directory +Thanks for the fast response Daniel Ill apply this to my fix for and see if it helps +This gives us more freedom in future as we are limit to completely bits are already used this only adds more not By the way for performance reasons all constants should be declared as int not byte as the byte read from index is already an int +Now that I am planning to merge my code into mumak +I hope to get and send pull request +You can presently set a separate read or write timeout. +Thanks Sangjin + +Thanks Andy. +import +Thanks for the patch Sharmarke +The proxy lazily report the target +Note that all content has been updated by hand +Committed r +We might choose not to support that +They give no hint on how they do this or whether it applicable to greater ns though +One thing brought up in the past on the dev list is that Struts doesnt complain when there are duplicates in the config file +Oh good point +Second persistence becomes dependent on the compression implementation +Maybe could be utilized for this purpose +All unit tests pass +Hi RandyWhere is this script you mentioned +So what does the resource method look like to reproduce this problem? And what exactly does the Accept header look likethanks +Please make it Patch Available after making the change or run tests and locally +current status is wont fix +we may need your ideas on this targetthanks for you review kuotai will attach the patch later +I have also included the two new txt files that should be added to the regression test along with the file +Sounds good! Thanks for posting the RB +I actually think that you are going to use it somewhere +No clients cannot get conflicting locks zookeeper will guarantee that will read again to +Applied in r thanks for the patch Dan. +Is there something Im overlooking? I checked via mvn dependencytree and mvn dependencytree +Already solved +Earlier versions of hadoop have a dir +Im not sure where the correct place is for pear packaging desc for testing srctestresources +Plain scp failed because of host key rejected while scpexe just silently did nothing +Did you add the jar to you lib directory +patch moves into +Any update +Then used this for by adding a new concrete class to test the connector +The first patch +This looks good to me +Yeah looks like we concluded the same thing +I altered the method to return when the input is +batch transition to closed remaining resolved bugs +The assembly plugin can run on +This looks like it is already done +I was wondering what other issues where going to show up now we know It may be that we have to stipulate that in order to use the DLL you must ensure that you build it linked against the same that your application is +Not but pretty sure still +Himanshu its more of a feature add than anything doesnt feel worthy unless we want to do a feature branch +If we think this is critical we should file another JIRA for it +Its simpler for us to maintain a Java API than a URL API particularly because we already have mechanisms for this +Nicer test failure message +The last visited viewId is always renderedThe runs into a different scenario +Cant reproduce +Which provider are we talking about +Thanks for the fixup Sebastian +IIRC on the development branchwork has been done on image caching +It will be in the temp build. +However I couldnt support and yet because they use complex Java such as I dont have a good idea to make them inI wouldnt be able to make them by the Click releaseI wish to make them in the future release. +Thanks for the patch +JSF searches its classpath for files +Keith just looking at some old issues + +Its not that big of a hack to move from to +Thanks +the issue is closed +We can always reopen or create a new issue if the problem reappears. +Attached zip includes the patch as well as html output for the new functions +That is if something were to go wrong a complete rollback of the plugin install can be performed +One is as before with java and the other is for cases where you have an XML document you want to transform and send to the server +bulk close of resolved issues in. +It might be a sickness +httplocalhost problem is described on the page itself +Committed to the svn trunk as revision Thanks for the patch +What I think in the divide method is returning if either of operand is null +in a file loaded by orThe bean should not be loaded by the again else youll end up with two instances one in the root context and one in your context +pending results +I think youre right should pass the instance that it loads instead of creating a +This avoids Pig needing to implement these commands +Committed to and trunkThanks Kihwal and Luke +I will mark it resolved +so any users created after the table could be given automatic access to it +I did add the bug id inthe description though +Cool thanks Ryan +patch has some conflicts that need to be resolved against current trunk +Bug has been marked as a duplicate of this bug +The number of idle connections that that pool limits via maxIdle is the number of database connections that are waiting for clients to check them out of the pool +OK its all clear thanks again +Merged to pinged Paul Ferraro on IRC +Ive pointed David Parsons to this Jira so hes read up on the conversation +The safest thing to do is to stop the cluster and start it again that triggers a process that tries to pick the most up to date replica and trades up to updates or so among each other if some are on some replicas and not othersIf you are sure the leader is simply behind you can just bounce it and let a replica take over as leader +Daryn could you please take a look soon to allay any concerns you might have +Thanks Arpit +The blueprint stuff seems fine +Configure Add JSF Capabilities +create create create can be inlined and you use the factory methods in the records +Update fix version +The difference between the two is the doesnt contain dependencies for the projects that are in the eclipse there a better solution +Attached surefire report +Ivan what the problem is all tests pass for me +Heres the ones i could applied in revision thanks Robert! +Attaching the patch +Please your application against the latest SVN snapshot +After updating the index I was able to browse the uploaded artifact but not the other artifacts +Is there a reason to have two projects instead of one single project? Separation of the Scala and the Java parts +Need an examplesrcmainresources is the configuration file for the jboss microcontainer +The Groovy version uses remove on a clone of set and uses Groovy compare logics +the new +Verified in +The support is not finished yet and this will be handled as part of the implementation of these issues +Closing for now. +Many if not most of the checksum errors Ive heard of traced back to memory errors +Has this issue been resolved? If so can you resolve the JIRA +bq +I upmerged and changed get to get because I did not like exposing the queue in that wayEric I looked more closely at your comments and I think you will want to use the API get instead of creating a yourself +Maybe there is a problem with the processing because Im sure that showed all as member of bothHowever I did not check the actual LDAP groupIve already raised the issue of skhudiky being in the group with the have sent an email out and included Marvin on it and will follow up on that skhudiky has been removed from ipmc group. +Hi It works fine when there is one elementbut when two element occurs then it throws up same scenario was working well with there any workaround without changing the xsd +Get hudson to run the actual patch +Steps for testing +Kishan pl +So indeed this was an issue but has recently been fixed as per +It works for me +That solved the issue. +Thanks Gavin and thanks for the headsup about the upgrade. +Ask to hudson +Bouncing to Denis +Will wait for you to come back to itFeel free to take it tooI think is also very important for speeding up merging especially because its so unexpected that just by adding different fields to your docs or the same fields in different orders can so severely impact merge performance +Besides options Xverifynone and noverify added +This should be fixed now +Unfortunately annotations on the target controller class arent seen on the proxy at all and we cant easily introspect the target class it be good enough for your purposes to move the annotation to the controller interfaces? JuergenThat would be good enough +Ill put that back as well and port to. +I agree it is not really necessary but I consider it nice towards our users +Can I do the usual to branch +We probably should have a set of documents with different size and complexity to make the test more scientific +This is already handleable via the chain logic in hivemind +We would have to do the check for dirty in +What is the status of the initial importThxs +has been marked as a duplicate of this bug +This appears to be a case where it is has not been implemented therefore it is a new feature +Need to become a feature as the failure URL will be a mandatory property that therefore requires additional configuration +I noticed that I made an error in the patch +Works well Linda Ive just slightly updated the specification to make the issue message contain the class name improve the pieces of tested! +Im considering to add two attributes retry and retry to the Connection Factory to fix this +Missed during first review but thats not true +closed +Component has been deprecated and will be removed inFurther development of this will occur in tacos. +bq +Nice cleanup +Looks good to me +Here is a patch +Ok the NPE was caused by a starting jetty via run instead of and then not having alternative pickup paths set I fixed the npe issue first by replacing it with a warning along the lines another bug was in the variable resolver a missing null check which nped orchestra in that area fixed as well the demo project bruno is working on now is running perfectly with full success. +Oh yeah I do remember that +Agree +Correction my previous statement is incorrect please does appear that prior to the urlMapping handler was not matching on content and with it started matching still looking into why +Committed to trunk +Any objections speak now +Im having the same issue as well and. +An ant package generates the API in all of the above languages +This no longer seems to be an issue with the console the console is integrated into both the project and the platform and has been tested quite a bit +But again this is for deployment correctI dont want to change our release process for +bq +ok thanks for your feedback +Cool its fixed now and I was able to install GWT Designer in JBDS +And its fixed now +committed after adding a test case that threw an exception from a method call. +It seems like we have the ability to break compatibility very rarely so we should use that to clean up what we can +I still think just dropping the interface at this stage is going to make upgrading to Solr a hassle so I think the best option is to move the interface into under the same package name +Yes it is +Good point Seth! Just changed +Its been a year +Given that this issue has been outstanding for over a year should it be consider supportedDoes jboss no longer support netbootIm trying to set this up now and running into issues and having found this defect entry I wanted to make sure Im not trying to do something that is no longer supported beforeSam W. +Oh should we add a profile for? Its same as? +But wouldnt a better solution be to change to default to true? It seems like a cleaner way rather than having to add a parameter every time you run the testsuite +You may give the head revision a try +And I cant find a way to edit am sorry to bother serialization and deserialization into protected methods in +Thoughts +We may not want to check this in as its hard to qualify it as a bug fix +Please check that tracing is disabled in release now +pushed to all with JBDSVersion Build id +Merged to the branch in revision . +Currently if you want precendence you can just your mappings surely +Needs to be applied first +The title of this ticket is +Ive added a simple test for +Added in improves this code and fixes a minor problem introduced by this issue linked to tab handling +Was just surprised to find it wasnt there it never struck me to look before but upon implementation in our codebase it surprised me +cool feature gerhard +Pat I didnt understand from your last comment if you have tried the patch I uploaded yesterday +It is a fair one and is something that Ive done manually before so I think it makes sense to ask +All tools issues are completed for this release +Thanks. +Can we these patches to version +But stick with this conservative approach until this is resolvedThats my two cents worth +Dont expect the needed generic mechanisms to become available any time soon +Patch to fix the version to be which works with applied with svn revision +In the case that the float field is less than for example then the JSON encoder spits out +push to +patch for again +is influencedimproved by the and not the other way round. +is the doc patchI see what you mean + does not support reusing tunneled HTTPS connections +You are free to change the MVC pages to prevent update from being allowed if required. +Using the app server guide from CXF didnt help +There is some regression from the previous patch +Unnecessary listHeight in VPE if the sum of heights of items less than the value of listHeight +I have just committed this. +bq +Tested on +Geoffrey do you have a sample project to reproduce this stuff which I could add as an IT +Well I have to say this is probably a caseAccording to the JSP specification the default character encoding of JSP source files is assumed to be +Thanksreplication strategy Yes currently thats the case but DSS can handle all the required levels it will just forward the requests to if other than or +The existing widget models try to do this form of prediction and it has caused problems in the pastIm currently working on porting the macro renderer over to the visitor interface to see what kind of problems we might encounter +Moving the integration tests to the module cause unexpected test termination in json module +I think the point of the move is to make them available for insane builds so they can be used by the engine on a crash so if it cannot be in shared I guess instead of moving the code there will have to be separate new code for producing thread dumps in javaengineorgapachederbyiapiservicesinfo +Slightly improved patchThis handles getting the clientId of a component after it has been rendered +Not sure if there are any side effects to this +Fixing Affects and Fix in Version +It is backwards compatible +I just enabled saving test output +fix got in cvs today +So you have +This is godd! How about adding another test method that runs modulo queries for all integral data types? Please have a look at the other test classes in the operators package +So far Ive focused on important usability improvements or issues that would prevent users to model processes that could be executed bu jBPM +If it turns out that I get access in time Ill revisit the issue for +Any chance you can also try on the command line or groovyConsole? That would help us narrow down our search +Especially this is inconsistent with the generated which uses +Also included a test case which fails without patch but passes with the patch +Yeah +Okay the tricky part is to aggregateI got the completion predicate running now but you can configure it to use details from the aggregation which now happens after the completion checkTherefore we need to aggregate as well which takes much more work to archive +Fixed in when it is released. +Same from containing only the victim connection +Second patch for the new style to CMSContent SetupData SetupLayoutData Resouces forms +I attached a patch with the change of my previous comment plus the change that allows fielded queries +attached is a patch +Hi Lewis thanks for your brief class description and trunk patch review +But its not only that we need to execute the right statements the TCK also checks for proper return values from the provider and many other things! For example the not only has to add the given files to the repository but it also has to return an which must contain the list of changed files! And all scm providers have to return the same list +If you have any difficulty to run the application let me know can be disable by removing aspectJ classes is solely used for logging purposeLibraries needed are et dependency JSF Mojorra Please take the time to setup the application and I will assist you to pinpoint the bugs for Then +nickthis is right fix for tycho based product build or it would not work +Its really not acceptable to break existing things with refactoring +Gavin yes it is tucu remove works +It looks like there isnt much interest in this issue +Skipping offsets and offsets hardwire the file pointers of frq prox files yet I need to change these to block offset etcDoes the offset imply that there is also a need for random access into each blockFor such blocks PFOR patching might better be avoidedEven with patching random access is possible but it is not available yet at +screenshots of issues +Regression test show how the procedure can be executed only by dba unless dba grants execute permission to other users +Working now in Liferay Portal Liferay Faces Bridge but still have in Pluto +I tried to verify the osgi bundle deploys ok but tempDeploy dir is not recreated anyway +I dont think it is possible to have it there +The result for every guide is in separate folder +Opinions +I just renamed to in javabin +A simpler solution for now would of course be to trigger the gc in a way that doesnt call schedule from the trigger point but somewhere safer where we can lock down tighter our calling context +Support instant schema updates with out masters intervention and bulk assignunassign +Unfortunately I have been swamped with other higher priority items so I am unassigning it in case someone else wants to take it on +Patch makes the following changes to javadoc generation If points at a jdk installation then we include classes in publishedapi and derbydocs and we use the jdk javadoc tool +This issue can be closed +for commit to this is broken there too +Adding the unit test to ensure no regression in the future +Interesting +Working on patch for now +Closing it now +Under Language Tools Visual C check X Compilers and Tools +Here is latest +The attachment is an eclipse Tomahawk no longer has a dependency on core +And per recent discussion on the dev list I believe the the consensus is the conservative approach is appropriate for and earlier FYI +Other solution is to write custom protocol handler to support multiple proxiesc but does not support it since it does not use internallyI think there is no other solution with to support multiple proxies at this timeI have customized and classes to tunnel through configured number of proxies and working finePlease consider it for next major releaseregardsMurty +Uploading a version of this patch +If not we may not get to this issue especially since we cant test it that well +Hey Brian +Argh +Patch attached +Not fixed in latest Jenkins restart +There is no way with Java to have a universal timestamp across nodes that is guaranteed to be ordered properly at a precision less than millisecondsHowever given that the wobble in network latency is generally millisecond anyway that is FAR less of a concern +attached the fix for trinidad trunk +Yes thats definitely something that can be done through an service +updated affectsfix versions +More strategies would be added later as this matures +You are going to need to supply more information regarding the environment +Can you send the of the module that fails please? And can you also send the result of mvn dependencylist executed from that module +I will post a message on Cobertura JIRA +Same issue +This change allows you to factor out that common header into the Caller API +pushed +Unfortunately the code does not handle thisBut anyway my proposal is to disallow the across volume rename for now +Your third point is a good one +its not wired in so we can leave it out for the moment +Giri thanks for incorporating the feedback +Im not saying they dont exist only that you and I couldnt find them +Such an approach could allow some form of denial of service attack because a failed attempt to do something still would call refresh thus repeated calls to the disallowed operation would still have a potential effect on the VM +Closing as dupe of +After running ant in the project root I runbuildjava jar The issue appears as described +Tested +so how to change it? how to make it multiplex instead of replicating? I tried adding multiplexingbut doesnt seem to understand this I am not trying to multiplex based on header values I am trying to load balance requests in a round robin fashion so is that supported +Pull request was committed in + patch for Hibernate Tools is the right thing to do for fixing the issues with tx manager not being available for query executions +Well probably want to handle the to upgrade in a separate improvement issue and come up with a separate solution to this problem +This issue is now resolved. +I deleted the earlier patches as they dont reflect the change to +We have seen the huge kills JT +Agree with Jay here +Description formatting +I dont know what I was thinking +That is the reason we are using dynamic router and also creating dynamic resource for same user in xmpp connection url to avoid Already logged in error. +The site generated for in aggregate modeEmpty Surefire report page is displayed +You dont need to produce a patch +Can somebody review the patch? +Added a search in the object so as to make the use of Ant +Ill have some time for that by the end of next week +Closing tickets +Your patch wouldve thrown a NPE if the wasnt a +Applied in r along with a unit test based on the sample document from the bug. +In summaryIn initialization would try to get s address which throws the expected exceptionWhile in Initialization would try which protocol to chooseIf MR it would do the same asIf MR it does not try to get s address in initialization +Patch looks good +Revised the description +I actually fixed it in the branch as the freeze has been lifted for a little while + IMO border is too thick +I would be very happy to see it in the sources asap +current plugin javawsdl and wsdljava wizards have fixed layout it run x resolution it has some will provide the implementation to get it +No user feedback +and +Some more cleanups and this now works. +Tests explained above +Either way sounds good +Decided to back port to backport is also rejected or is there a new issue +I agree with Vinod +Essentially it really forces you to do things on the client side that shouldnt need to be done +Committed with revision +Resolving this issue again +perhaps you can find a good java but this seems like a big heavy piece of codeWhy do we need to do this +Its an additional array of bytesI will take a quick look this evening at the patch and see if I can bring it back to life +All schedulers MUST handle it +This is the more correct approach +We can try to optimize that in trunk +I saw this same bug +I have described situation there is no any HTTP post +Is that javascript +Perhaps I was using an older version of the source code. +mostly after a week +Username genman +Let me retry a submission +done +or ROOT is being opened +Update the s eclipse root poms version +Moving to +Even if not the string will be cached in the static string space of the class file and this should just amount to setting a register address each timeAny other blockers +Please see for more details on this the change on RHEL and the changereposjbossasbranchesreposjbossasbranchesreposjbossastrunk +I have mixed feelings about both +Updated thrift +Move to backlog until when it can be done is clearer stop just rolling from release to release +The patch fixes the issue +The controller model has also moved on a lot since Spring Security first developed its filter chain so as long as a suitable configuration mechanism can be worked out using controllers may be a good alternative. +Hmmm +yes it seems it works +Only the transports to go now +I removed the update parameter from a temp file +IE the current behavior is whatI would have expected based on the specWhereas if I wanted to see the gadget in a fullscreen overlay I wouldexpect that to come from an openGadget where our canvas overlay is adefined viewTarget +If the client gets an error it must close the file delete the file then againThe idea is that the client should avoid getting an error + +Using the quick edit on the ticket itself did not resolve the state issue the ticket was in +juddi has been updated to v and this no longer contains the changes for updating juddi. +The server now states Publishing + +Well I can assign it to myself +Unit tests has been applied to branch +So disabling the tests for now +Is already required on Windows? If so then updating to across the board probably wont do much harmThe Newtonsoft dependency is already there not added by this patch +Is it really too much to ask for users to upload something that actually works for bug reportsYour bundle is missing referenced classes and the package structure is not maintained +Ill look into it. +my suggestion is to remove the assert leave the code in +If I have any concrete comments or suggestions Ill barge in but so far its all looked quite good to me +I had to make a small change in to make it so the output table wasnt created with buckets +Was this patch made against the trunk? Anyway I manually merged the changes reformatted the new test removed from and committedKathey +went in resubmit +It is a bit of a hack but I dont see a better way to do this +Yet able to ping all public +No need for IT +Forward progress towards where we want to go is better than no progress. +Still discussing +Thinking of it that proposal has the drawback of being hard to optimize for when the client gets the size and starts iterating though the children +Please add some unit tests to prove that it works as tests are added +Im Ok with external test lists but for two reasons they require an extra maintenance step is needed at this moment HDFS seems to be the only Hadoops project which uses test lists +as i said on the mailing list wicket cant really do something about it +I tried it out on and but only with the example provided I will hopefully have time to do more testing in the next couple of just had a quick look at and none of these changes should break anything so backwards compatibility should not be a problem +Hasnt been tested extensively +changes look fine to just committed this. +Also tried to clarify a bit in the row cache description and standardize on word instead of both that and word +Dont see any actionable items for SDR here +Gabriel reworked all WMS tests this should not happen anymore +Connections were mostly to the data node +rev ports the server trunk changes to the branch +Heres a fix for this issue +thanks +I dont really have a preference +Added a but omitted the crawler bean policy + is called before commitlog replay +Ill get Pat or Jason to comment on this issue +It might be easiest to implement this as two buffers one andIt could well be so +Moving to +Ill try to get some time in the following week or two to finish up or at least publish a draft for you to take a look at + This bug has been marked as a duplicate of +Add which is a clone of +Since such methods are just gettersetters for the property I think we can always use a getter return type for labels +Closing as this works for me with M. +I have a factory to produce the correct type of using the type of as discriminator +Closing as duplicate of cant work on the JBDS issue so reopening this one +This will not be considered for issue was triaged for SOA decision was to leave it in Future +Now that work is underway with AS all previous community releases are +If people are happy then close this one out +Maybe the bundle lock was acquired before the global lock was acquired but its true it should not really disturb the refresh processingI think I ran into this problem once during load testing but I may have been misleaded because the issue I was really looking for wasnt actually in Felix +Dialect patch on version patch on +Supplemental patch to synchronize the remaining events +If appStates is null or empty on it should return apps from all states +New patches attached +Hi ToddI fixed the comments can you please check +The client should call to obtain a new generation stamp any way I believe this what Raghu meantYes +OkI agree that if OID macros are not really defined in the standard they should not be usedMorever it would be great to have a standard schema representation so an application can provide a schema file that could be inserted in any standard LDAP directories +The proposed patch should fix this issue as well as of on Windows XP Pro the call to IOclosewrite now causes the exception +The patch is for drools git repo +hehethnx +It seems to occur only for the first term of the document if its positionIncrement is equal to +Patch committed in revision thanks! +Same thing happens if you insert richcalendar as filter element after table rerendering it not opens +It works for insert +promoted to highest feature request priority +Fixed for Endpoints and Beans in Palette and Context Menu. +Please review the issues and create change requests if you feel the issues are valid otherwise resolve them as the linked JIRA from Jan to make sure we have clarity the specific feature are talking about is correct the linked JIRA is not relevant to this issue +Combined with Simons check and Libbys check Im thinking that this may be an issue with the manufacturer not a general Android issue +It also includes a test for this feature +Not needed +The test reads testsamples and writes a to disk +I have a srcexamples could I make a new root for that +we like markup packaged next to components not in +thanks for the feedback keep up the great news on this? Is there a good place to for information on using Spring security and GWT with Roo? +This requires some sort of parsing +We may have to define if its a bug or not +look good just committed this thanks Terence. +Bryan do you agree that this will not result in a performance improvement? If so what is the benefit +I have uploaded a patch for both and with snapshot version numbers and also fixed a typo +No I wasnt upset I was just trying to avoid confusion +huzzah! +JDBC is available only to developers who download Java SE and use it to add JDBC support to Derby +If everything looks good overral then I will tie up the loose ends and add more testsFor review the code is broken into removal of old code new code test +Perhaps this is only true on ASM that I have been usingI will with Beta of JBDS and ASM +is related to +Remove save actions for groovy files except for organize imports and format +I will look into it further +So just to be clear my use case of access control on versioned nodes is invalid? Or the manner in which Im trying to accomplish it invalid +So there is no need in minor version for builds +This is apparently fixed now +didnt get +Thanks a lot for your helpHappy to close the ticket +Thanks for fixing these issues Kristian. +The problem still exists for when the label decorations are disabled +push out to +OK on a whim I removed one of the two coverages that was exhibiting the problem and recreated it with a different name +So this analyzer works correct +When an MDI page is closed there is no need to do anything in set +OK now I am only seeing one resource for JMS however it is the wrong one +Thanks Knut Anders and Dan for the clarification +joram can you see this in the mailing list +Moved assignment of temporaries per Suresh +Thanks all for the review +The that represent each are created dynamically as the start or stop +It has been done! This is now pushed back to masterThanks to everyone who helped review bug stomp +v cleans up data files from obsolete indexes +Todd can you let me know if you are very busy and cannot review the patch? I will work with others +The primary motivation for this one is that Weka is pretty broadly used and it would be good to be able to ingest Weka input docs into Mahout +Same pattern later +Denis is this still a problem on the latest +I think Ive got this mostly addressed for a first pass +is not truly a JEE compliant server they knowingly include these hairbrained behaviours in spite of what the Servlet spec says they should doObviously there isnt much we can do with IBM +This is in place in branchesjmxmcgsoc and will be merged to Aidan can you review this change please thanks. +Actually the analysis I presented in my previous comment iscompletely wrong +Same issues in issue has nothing to do with AS. +would appreciate if it can be fixed it inIn fact the problem Im facing is to load the payload in the into a SDO data object using the and it fails due the namespace issue +That means that the count should be something like count value and there should be an upstream combiner somewhereThat will do much more good than using listsThat comment from Owen is essentially the same as what I and others have said +It is simply not normal that those tags do not perform well under RF components +Let us break the Manipulator interfaces at the type boundariesThe types like durationsdatetime etc bother me too +I thought of a way to fix it in the tutorial +Fantastic nice use of ObjectshashCode will double check there was something about that test and the ordering that was important but it might be as simple as making sure good is first +No I have not been able to reproduce it +Adding HDFS would eliminate themI would suggest that any replication strategy should focus on delivery +Could be a problematic to fix for M because currently build with target platform fails on wit error about resolving first plugin in dependencies declared in manifest +Hi I have no permission for closing the issue +not blocking no longer can reproduce this closing to not clutter the anyone find a way to replicate please reopen. +How does this look? It adds an empty file called where is the last directory created and so has the result +OK I will update the javadoc for to make this clear +It makes the side exception with a more meaningful message but no change at the client side +Why is that? Could that be a potential solution to this problem +Ping? Anyone willing to tackle this request? Is there somethingblocking progress here +I see memory footprint reduction in my case and saved me running ove GB limit in windows +See prior issue for more detail +I agree with Konstantins analysis +I will try to do it. +I think we can the edit title of this JIRA as wellThanksUma +If you define all of this in the plugin config then will end up duplicating every dependency in your project twice +Certainly the problem is not versions of Commons CLI since is used throughout is clearly a real problem but it is to do with Gant Common CLI jars in Grails and Groovy lib directories and the embedding of Commons CLI using in am therefore closing this should perhaps raise an issue in the Gant JIRA to cover the problem + +is an incremental patch can be applied against svn I dont think my changes are very clean +and rev +Do you have a test that reproduces this problem? Would be great if you attached it +I changed Filter architecture to be more Basically yes Filter will be applied to all messages +Especially given Knuts note that in worst casewe are likely to be trying to allocate an array that is +Full thread dump looks OK to meAdding a to the filter chain might show who is writing the message repetitively +In the interim Ive deleted the git repo and restored write access to your svn tree +I uploaded a simple testcasestep by step test procedure +there will be something better in mahout in the future +Looks good +initial javadoc cut please review FIXME javadoc locations +Works pretty well but As soon as you start using the pagination mechanism available in the widget the link available on the total number of issues becomes incorrect +I had mixed up some code relating to some other tests I was doingPlease try this one and let me know +Changing this into a only issue given my comment above on tests passing with native +Hi Tomaz I checked with our system admin and he sees the JBOSS is killing the system with the load it is generating +Attached is a patch for this issue +No the tutorial zip files +Interesting note with respect to +to investigate the the hang problem a bit +new patch for missing getter and setter and testcase +I agree that the wording of the spec does not reflect this so it most likely will be changedSince you are past the problem I assume that this issue can be closed? +I dont think we need this substates are rarely used anyway +If using some other backend would be an overkill then lets stick to +Let me try this patch on a cluster where reverse dns is broke to make sure we dont see the issue +Seems a strange choice to meWhy not have extend +The issue is the clocks being out of sync between nodes +Ill look at this +Updated the content and navigation divs to float top instead of float left +Thanks Jyotsna Rathore for your patchYour patch is in rev Ashish Vijaywargiya +I wont press the issue if the tide of opinion is against me just honestly didnt think I would meet with this much resistance +Jason is this activated when check snapshots option is enabled? +The idea is change into an abstract class and create private implementations customized to be as small as possible +Whoops this should have been filed under the Maven War Plugin project not +OK Ill move it to FRN instead +forrest passes fine now thanks +RCMED CRU in revision +However Ive been thinking that Tika has a lot of dependencies by default and we should probably go through them and reduce the number of dependencies +it was certainly broken for a period yesterday + should fix this it didnt +Please verify whether it is resolved as you expectedGood luckLeo. +Assigned to Jeremy for completion for upcoming +if null is returned the app just dies with no notice but if the exception is thrown than without doing anything special the developer will know it by virtue of a log entry or the exception handler mechanism kicking in +same result as previous patch but applies to current trunk +Upload a patch that creates a module under +Thank you! +Thanks for the benchmark Jie +Please apply the patch under +Hm so this is probably a documentation bug combined with the fact that I have been producing plugins for both GS and GS out of the same that with GEOSERVERWMSURL set or is not accessible through httplocalhostgeoserver +New patch addresses merge conflit +Changes in the code patch were applied. +Sorry no news yet +and are used to compare the Container objects and the getters of all the three are referred in multiple places tens of times +A Map would support get and remove better but I think we want to move people to using Iterators and the remove method is there for a case we dont know of yetIll create a patch with these ideas shortlyCheers +You are talking about a different strategy that should be impl in a different command +Patch for trunk ready for commit +And whether the table have many history versions or not also totally depends on the usage of the table +Any objections? this should definitely be in. +I dont know when we supported Vpn for nat ip +Thanks for looking at this Knut +You are welcome to review the code and comment +This issue is primarily concerned with adhering to Androids whitelist for those modules that might be used in Android development +Lets not do this inAgreed +have separate implementations for server and client because of how auth protocols work on the wire +pull request merged into master and BranchAS you! +If the algorithm simply tries to grab the lock and then boots the db RO if that fails then that is what user should be told +Temporary fix until comes in +committed. +I just followed the pattern that was there but I can change it +Lets continue the discussion there instead of on this CLOSED issue +An added benefit of doing what Im suggesting is that can refactor much better +Metrics being evaluated in +to trunk. +Command ran hadoop dfs ls webhdfshostls Exception obtaining parameters +git checkout vectorization git pull git checkout b teddy +Hi BJ just applied you f patch +Thank you apologize for not providing all details at once and I will remove the info about from bug ElenaI already committed a quick fix for it last you verify my fix of it by running some test on the lastest Fuse snapshot kitRegardsWillem +Marking as blocker since this has been seen more than once +Could you try to remove the directory from your Maven local and then relaunch a Sonar analysis? an additional note you should not modify which were released this is not a solution at all +Hi Guillaumewell I havent seen any source tarball like yours since Im on the Internet and the current layout is causing any distribution head aches +Hi Bart cache manager set to info check checkYou are right this makes it easier to backport to right now way before the stats are stable +take back my comments +Im going to try and look at this some more in the next few days and clean it up a bit and add in some of your suggested unit tests +Ive added a junit test and fixed the bug +Updated description as it was a bit have no problems with renaming it at the time I started the patch and were the polyfields I saw in the repo at this point so much has probably changed naming things appropriately makes sense +More specific messages are in place +Amit can you attach a sample application here so the testing is easier? have attached a zip file with sample app +I hear this is no longer an issue so closing as Invalid. +as for the question of outdated items this is a discussion of the policy and should be raised on the dev list +Github isnt a mind reader and neither are we for a gihub mirror to be created one has to ask for itLets treat this jira ticket as a request to add it rather than a bug +This is a patch to resolve yoko ORB related port setting problems CORBA Naming Service ORB Adapter ORB Adapterby adding a binding host part to the ORB initializing property +Is there a SQL Standard reference for TRIM and LOCATE using collation +lgtm +adds closing of the which I assume is the cause for the reported errorI will resolve the issue if the daily test run confirms that the test runs without failuresCommitted to trunk with revision . +So It is not worth the effort of making it write negativevaluesSince this is a special case which is true only for I was wondering if this can be handled by writereadExternal methods of itself +Looks like its the same. +while this is an improvement on the old totally buggy behavior I think itd be great if as much of the path was aliased as possibleToby if this is something you can submit a patch for in relatively short order please do so +Can you remove from the patch I dont think it is being it looks good Please add more comments and I will accept and commit it +hi thomasi think you meanare you using icefacesregards thomasi seecould you please provide a bit more information about the bound property +Work on is done +So theres very limited parallelism you can introduce in the L L stage which is the biggest bottleneck +If such lists exists the aggregate does not ask the FC for the next step and instead will remove the next delegate from the cached single step list +Could you make sure it works for you +Should note Ive only hooked it up for lucenelib and solrlib and not any of the modules or contrib +Thanks for that +Archetypes release has been submitted to VOTE +And jars work with Ant +Ive nothing in the lib +The context menu has been created by adding extension the the +What does the path to the shapefile look like +The context is now correctly used to determine the appropriate classloader and the Exception no longer occursis an accurate summation of this ticket? I require an acknowledgement by the th of February and confirmation or clarification by the rd of March +I know depending on a SNAPSHOT version is really odd nasty but that would be the best option we could currently have +To answer my own question a calze is a chalice +That begs the question was ported incorrectly? Otherwise why doesnt trunk have this issue? If there is just a fundamental difference in vs the minimalist approach is probably best for +for the latest my +Thanks all for the reviews and suggestions and assistance running testsand preparing the patch +Guvnor is usually my default project +And I dont really have a clue where to look yes that is most probably the thingYou are using JOINED inheritance right? If so the child table only needs columns to reference the primary key of the parent table which constis of the revision number and the original id +Wont be done for Moving out +Thank you very much Gary I will do so. +But i never ran into the problem you have mentioned in the issue +Is there an address I can it to +Please test. +If this port were to be occupied shuffle could fail + has been released now +Bulk closing issues related to +A diff comparison over +About the propertyOpt could you add a description about its purpose and format? Also its not used in the init method of +If there is a problem please open a new jira and link this one with that. +Things get fixed faster when people provide property was only used internally so no need to document it +Heres patch for original approach to +I am investigating the test failures +Do we want to depreciate the method for then +Unless someone shows that its on a public repo I think well have to stick with the dojo foundation would set up a repo that is automatically synched with maven central or arrange with sonatype to use their rd party OSS repo that is already +Juergen +Committing this +jnlpThank you for responseJRE is installed but i try to run +Added module that mirrors + +Please use something between kb and mb as the defaultYou should derive from for +I couldnt create this issue with a quick start applicationLooks like something has to be tweaked in my project +Thanks Kristian Knut and Rick. +The latter is used if you want to override a specific Codecs behavior +The last incremental update overwrote some important objects and cause the acrobat reader stop correct parsing the document +Jar FilesWith regards to Jar files The installed jar files are not logged but system catalog updates are logged when jar files are addeddeleted +You might look there and for naming conventions that are compatible with the approaches being taken in thoseDonebq +This situation takes place for all error FF for other error windows How it works? panel contains Upload Images for IE IE IE on revision +The patch is split into formatting and changes parts for reviewer convenience +thanks purpose of refactoring is trying to keep consistent on exception handling for different cases +Thanks pat! +Hellowe have the same Problem with an realy simple CriteriaCan you tell us your detailed pleaseWe use for our Tests Linux an Java for compiling and Linux with HSQL for RuntimeThanks closing stale resolved this issue happens occasionally in our production environment. +Gavin any chance you will be able to get to this soon? We are holding off the release on this task +Hi Namit is dependent on this JIRA are you able to work on this? Also also needs to be made dependent on the new HADOOPFS parameter +on removing the themes +No published release note for just a bug. +Sorry I cant reproduce this nor offer any deep insight into JMX +There is just one thing I realy dont like Why is there an containing the parent pom? A simpler way would be to put the parent pom as ancestor of the submodules for me its a simpler and better approach +Yes Ive tested this yesterday with the latest snapshot build from Tools and build seems to include the latest fixes +The patches submitted in this ticket are diffs on top of the changes in +Annotations should always trigger functionality that is configurable with the facade because not everyone wants to have a hard reference to in the projects model nor do you always control the classes where you would like to add the for the upcoming release Id simply expose the test for a transient field by moving the condition from the nested code part in method visit into method field +I think the magic byte sounds like an elegant solution to the problem +Attaching a second rev which removes the TINYINT line +From the logs I couldnt see anything else suspiciousI suppose that the pattern you posted keep repeating itself indefinitely +Missing bit binary lib would force users to build the hdfs project which might be simple for us but it could be a deal breaker for end usersAs it has been pointed out the arch specific libs had been provided in past versions +Oops priority should be minor +how do you propose we show that in UI? We could put it as the root of the tree I suppose but that will move everything down a level in the tree when there will only ever be one root project +Seems to me something about the resource loading here is not right +Filter creation will return back types +Hi Adrien thanks for the fixes +Thank you all for your responses. +I was unable to reproduce it running this test from Eclipse +Hi PascalThanks for reporting the bug +code lgtm +as I told there is no stack trace at all +Are you sure this is not a App server issue? Can you provide a test case showing that Forge should be doing that +one is simulating the real world deployment where projects and modules from the internet namespace are mapped to residing within a root node of a filesystem +And if you do patch E it removes them +Roller now provides feedback in case you get the math question wrong or forget to answer it +yeah seems to be. +Removed the Java code which had crept in Two tests still fail need to solve Breaks backward compatibility so need to find a way around that +Actually thinking about this again can you the creation of such a faulty mapping file through the +Hi Martin +Fixed +Fixed and examples updated and working AFAICT +Fixed will be in Designer release +Looking +patches supplied in issue is fixed by the patch applied for jira +All does is for the client to pick a more optimal node +I think I can add a test to jdbcapi for an url with spaces in the database name +Took the liberty of removing an unused import from +Releases and archives have been moved from jakarta to the relevant locations for turbine. +Ill commit after checking a little bit +This is still pending cluster testing +Hi EdellHow can I test on my PC the patch +by Richard. +currently assumes aggregationsoperations on statistics produced by tests and does not consider aggregation across a group +consolerootsystemvm cat etcCloudstack Release Mon Feb PST On the hypervisor cat varloglibvirtqemu starting upLCALLC PATHsbinusrsbinbinusrbin HOMEroot USERroot LOGNAMEroot QEMUAUDIODRVnone usrlibexec name S M m smp socketscoresthreads uuid nodefconfig nodefaults chardev socketidcharmonitorpathvarliblibvirtqemuservernowait mon chardevcharmonitoridmonitormodecontrol rtc baseutc deviceidusbbusaddr deviceidbusaddrx drive filemntifnoneidformatqcowcachenone devicescsioffbusaddrxdriveidbootindex drive fileusrsharevmsifnonemediacdromidreadonlyonformatrawcachenone devicebusunitdriveidbootindex netdev tapfdidhostnetvhostonvhostfd devicenetdevhostnetidnetmaceafebusaddrx netdev tapfdidhostnetvhostonvhostfd devicenetdevhostnetidnetmaccddabusaddrx netdev tapfdidhostnetvhostonvhostfd devicenetdevhostnetidnetmacdbebusaddrx netdev tapfdidhostnetvhostonvhostfd devicenetdevhostnetidnetmaccabusaddrx chardev ptyidcharserial devicechardevcharserialidserial chardev socketidcharchannelpathvarliblibvirtqemuservernowait device virtserialportbusnrchardevcharchannelidchannelname deviceidinput vncpassword vga cirrus deviceidballoonbusaddrxDomain id is tainted char device redirected to devpts means you are using the old systemvm templaterootsystemvm cat etc Cloudstack Release Mon Feb PST The right output should be Cloudstack Release +Nothing bad happens with ports in +I have run ant on +I wont have time to look at this before so anyone feel free to give it a shot +Fixed all the broken unit tests except for +I do not object +test and passed +If possible please provide some indication of what is actually being instantiated when this error occurs +Here you go after you have given this some thought send a briefing on how you intend to tackle it to the dev list so we can all share some thoughts on the package how type inference for enumerators works today and how we can better integrate the solution with generics in the future +Attached patch of related work +The method in lucene +I have refreshed the patch due to your advice +Vijay if you agree that this is an implementation limitation in Java backend please resolve this as not a bug +hostportd +Solution foundAfter some experimentation I discovered that the solution was to simply move the from sharedlibto commonendorsedIm not sure this is the optimal place for this jar but I think it does demonstrate the the core issue is some kind of problem with how the classloaders are being used +we need to address it for M +Like Seams approach to POJOEJB optional natural Seam in a pure TomcatSpring environment is extremely easy and efficient and this platform is sufficient for many solutions +I thought I had migrated them over +let user manually delete might be more safe I think +Since there is a lot of latency in IO we have an improvement of a few times in the time to generate the block reportThe test for this is just running the directory scanner test twice with parallel execution and without it +by Dave Grovewell have to redesign before release +Looks good to me +Thanks to everyone for comments +Cancelling patch to a new one with whitespacestabs cleaned up +This is something that must be looked intoYoure more than welcome to tackle all this CheersAlex +If there are no objections I will commit it +Works in my environment +issue as fixed as per comments above from mfojtik. +Please add implicit parameters as another issue so we can discuss it better +I migrated the functionality from the branch +Revision basic implementation +PavelI dont have time to look over the patch just now but could you make sure that these changes include updated documentation in doccqlThanks +Theyre both needed for interoperability +attributes can then go away from the mapping files +If you want to request your desired behaviour as a new feature go ahead +Resolving since the original failing test is gone even though the root cause may have been different +since after setting closing flag all the following put requestes will fail immediately there is no neccessray to verify the result at all +I think a single checksum per map task output would be fine +Both NMR and Features use cglib bundle which the latest one available. +Version M has been releasedClosing the issue. + contains the changes for this +This issue has been fixed and released as part of release +Thus another service should be manage except regular Source and so on +We will open separate tickets to address +However given Id like to get some fixes in before tomorrow Id like to update this after the feature freeze date +A recent change in Pig causes setLocation to be called twice and if setLocation isnt idempotent then you get twice the output + +It is another regression of the fix for as +If you can reproduce the issue let me know. +By the way in case you hadnt noticed Web Flow is now on github so you can submit pull requests for changes like this one. +I will mark this bug as invalid and close it. +BTW there is one issue which is really buggy but more in general compiles this without any complaints and makes an installer of it +Corrupted tables are not consistent meaning some times some tables and some times another tables +HowardThe wiki will be ready as soon as you confirm the options I suggested in my previous comment on this Jira issue +Proposed patch for integration to make it work still need to add three export packages in the bundle +Committed Thanks Shai! +I am still thinking about the difference between function querys and the new and I would really like to combine bothI know for sorting the array approach is faster but maybe the new could provide both ways to access +I think the issue with this as it relates to realtime search is in order to sort well need to freeze indexing +That would seem to point at something other than the JS +Im inclined to agree that it should probably remain the same +In what case would schema for source path need to be registered under the hood? I would assume a user will always need to reference source path before using it in which case storing schema would occur as is done currently +Its in trunk now if you want to check it out. +Where these changed really not committed or were they committed without a comment with? Either way we need to understand what happened +With Maven it does seem to run + push to +from here as well +The convenient way will be putting instance of cache on the so there would be separate caches for message body readers writers and context resolversSo the cache will be dumped only if the provider of the same type is addedThe cache must not be static there can be multiple instances of Wink in the system with the different configurationsIf the cache is put on instance of there is no need to synchronize since the is already synchronized +Used s escapeString and unescapeString directly +New patch with some cleanups and that lists it as a new feature on version +For Data Scroller Ive committed a fix in +and of course next natural idea which comes to mind is provide json inheritance like maven which has a parent etccome on if you serializing classes then json docs also have inherit +So just +Thanks Richard +Arvind have to put Nemon to Sqoop contributor list before we will be able to assign ticket to him +The proposed solution causes test failures suggestions +Moved back to to currently DNS server is not change on these DNS issues didnt work editing again to set the fix version to to The DNS server is not maintained atm +but joint compilation using groovyc works or not +Thanks for the reminder ClausI have something ready that I will attach to this JiraRegards you can assign this task to me if you want to +I dont think that the JMX interfaces should expose any rate type counters +Hi Michel Im pretty sure there is an issue with your configuration +pubunpubusingrest that will demonstrate publishing of files a step that is needed prior to querying the metadata +Occassionally we have run into bugs in the Sun RI +In fact we both found nearly identical fixes +Im confident the test failures and release audit warning are unrelated to this patchThanks a lot Harsh +bq +Scheduling for you so much for this patch Bernhard +Thanks atm +Hope we will get response this timeRegardsAbhinav VaidiLabs L T Infotech LtdMumbai +Fixed in and M +one could also think of it as null being a valid attribute value to pass and the encoding of that shall be to leave it out entirely +provide a more tutorial that the quick start since the rest of the documentation is now slightly terser +The exception change and the selector change are independent +Should be fixed +In this case the default document GET servlet would probably be an which does not handle the request for an empty request thus allowing the default GET servlet to kick in +note that this patch should be reverted after releaseec cloudservers patch warns before converting ec and cloudservers tested with cassandra on both ec and cloudservers +Would prefer to see us closing one jira and creating another +Thanks Thiruvel +Issue simulated seems like not implemented feature +Heh I didnt notice the last comment that it is still reproducible in a special mode +Specifically add compatible version for add compatible version for add compatible version for Once modified the snapshot for EAP performs well under EAP integration Darren +Reading the code it defaults to true which is a behavior change +Another thing ive found out is that if i revert to version of the site plugin the links are generated fine on both windows and linux machines +We need documentation and unit tests in you have initialization logic etc +Namespace support has now also been added +Resolved +For the time being I added some documentation to the GPG Plugin that points out a workaround +Some parts of use the new code +If you stick with the current default of XML they are perfectly compatibleIMHO its OK to require special work to enable compatibility +Site works fine back and patch applied. +Toby has attached a patch for +Is there any further work being done on this issueIm looking for something similar not necessarily aggregating a score but I would like to able to use sort and together such that the is applied with in the group first and the first document of each group is then used as the representative document to perform the overall sorting of groups +I could be missing an important use case here +Also javadoc should be trimmed since its too much blabbering However is not about disabling versioning for particular requests +Since this is a limitation of the ASA appliance there is no need to change the code to handle this +commit this patch at revision +This will be fixed in +Comment above? You mean in? No it isnt needed +I added an entity called that can be used to give each user their own DN +I think itd be better to Assert here +Im trying to poke through a few other things right now +Another test case for the same problem +CR is what I used +The check highlighted is a protection for the data structure being used so that it does not overflow +Thank you Chuan. +Just a suggestion +Thanks for the clarification +I think you need to either revert the changes you made and use svn move on each file or if you want just finish this off you can do svn adds of all the files you moved and then create the patch from that svn diff +Let me put back this bit of hackery w comments so we get retry on open and nexting +Thanks Subramaniam! +Ill work on the unit tests and then post the resulting patch hereThanks! you add more tests +New patch attached with apache header and style test case +One could provide an option that allowed you to override the default mechanism and force Castor to decapitalize in any case + Committed revision trunk Committed revision +This looks like a good thing to get into +Looks like i have not searched hard enough before entering this issueI understand the concerns in +I will open a new issue for these +Ive added a comment into the ticket and had addressed it to Brad Maxwell but he hasnt yet responded +Ive added a junit test and committed a patch into SVN trunkCould you give it a try +From the patch an observer observes the leader +Yes Jacques I tested it and I am using right nowDid you antIn the patch there is a new seed data that needs to be installed +Attached is a totally self contained version of the test program +Fixed there were a few properties missing so the framework started up with the wrong configuration +Or I should say this syntax works now +is indented incorrectly the goal links dont work how to use Usage how to use section says Plugin Name instead of deploy plugin comment outremove the empty faq for the examples Id suggest grouping them by project deployment and file deployment there should be an issue opened and linked from MNG to point at these docs instead of the guides already on the site which are now duplicatesthanks +Its solid lines in Linux +Any objection to commit it +Hi ToddOne minor comment instead of returning null for this function can we return empty string array? This was the cleanup we have done in the past with couple of JDK +to another way to implement itThe anonymous viewer test now in the Shindig level it not even hit the data storage level +I should have it committed soon +I know Leo but somewhen we had to do it + +This is new to me what else causes pausesThe Safepoints in the output were Deoptimize no vm operation and +Fixed it by updating the status from QUEUED to MATCHED once we find it matched and when dequeue the correlators only those are not in the MATCHED status will be tried. +Modified the implementation as per review feedback Introduced an abstract base class and reverted the resolvers to extend that instead of directly implementing the new functionality +andreas thank you for the clarificationPlease see attached patch which removed and substantiates on documentation within Lewis Thanks for you contribution +Not sure what effect this will have on the CALL statement in triggers apart from the combination is unlikely to have been tested +this codebase is still importing JDK stuff so its not buildable with JDK +Clients would need to set command line options into this conf +I had already configured websphere to use port this didnt help +Lets continue the discussion thereIll do closer review early next week. +I guess ideally Modeler should automatically pick up new adapters based on the contents of SERVERADAPTERDETECTORSLIST in +Indeed Dinesh +Here are some simple test cases for this enhancement +On the one we could deploy it with no problems and on the other we couldnt without the above fix +I dont think is a necessary return type when we already have LinkUsing is the lazy unsafe way but creating an that implements Link is the best way to goCheersNick +Pushed this to upstream master +Close all resolved issues for older releases. +Thanks Bhushan +Do you have a wiki account? Its different than this jira account +So we need to stay in what is provided now by confluence and docBook export. +The one thing i am against is somethign that needs manual maintance everytime something changes +Btrees can under go some major changes when there are no locks +Fix in place +StartShutdown operations work without NPE +Please pardon my limited knowledge about but I have a few questions about the approach you are proposing +I know that you Aaron are working on it so I will surely just be sitting and waiting for the patch to arriveThanks alot +The last stack trace is +Does solve the problem +There are a lot of changes in the patch but its fairly straightforward +Im not talking about Groovys whole collection type system wrt arrays +Implements path resolution as discussed in and Added an interface to to indicate if the file system supports symlinks and a test in to test the behavior of createSymlink when the file system does not support links +Its a feature branch not a branchId like to see removing shared doc stores backported in particularbecause its a good simplification of the code ie it can preventsneaky bugsAnd I think we should backport DWPT? But first things first we gottaland it on trunk And let it bake some +shall I submit that patch +This is a new one on me I kind of expect resources stored in the classpath to be platform independent +We need to revert to achieve this +jar +The war file is based on the latest SNAPSHOT including the formidable upload stuff but the issue existed before it may have come in at the big cleanup months ago +You could try something with some custom htmljavascript and richcomponentControl +All other browsers will work well now including ChromeThis makes using forms in modaldialog more clear and clean +Is this fineYes that should be good to work with +If the answer is yes I wonder why do we need create method it would be simpler to bind and lookup directly from the Registry +It is indeed not introduced by this patch +You can use any login module you choose in the fallback domain so it does not need to be a database module the most common choice in the scenario you describe would be to use one of the LDAP login modules in the second domain to authenticate against LDAP using the username and password +Please give some feedback +Maybe this matters +And seeing the whole stack in a single view is IMHO valuable since Im the person that has to monitor everything during stack builds respins and releases +Let me know what you think +From the comments this sounds like this has resolved itself +on this +committing this I want to double check a couple things to make sure there are no cases where we end up making more packets than before +I identified is a font problem +Hi liitIll have a look at this +Add descriptions for all of the targets in Move some common properties to +Thanks Mikhail +is this against trunk? It fails against for me +Theres gotta be some way to fix it +Hi Rajiv no software can ensure that the administrator not change the config +Here is the failure message displayed on the Testsuite page Assertion failed got server info Pluto expected Pluto +My first thought on making Eager work for Enums was to introduce a new ENUM type code in +I agree too +I still need to turn it into a test case though we can run regularly. +Currently you can workaround this with fixing integration url in the Im gonna use your st proposal as getting the info out of PS seems too ugly although it would be the most valid +Test and patch to support thanks. +Well what is this new bug about? I cannot understand itBTW its not easy to fix the Filter Praser reentrance pb +We can do that +I would think the xss would be more in the POST caseJavascript can be injected via the url query parameters too +What do you think about introducing a configuration parameter likeThere is a line in the patch marking the right spotboolean post true X make this configurableIt seems to run fine in my real world application but I just need to start the full compile with tests enabled over the night +You could argue that the syntax of the request is incorrect +The seems to be a hint that somethings wrong in the overall class loading could also try to run the relevant parts of your Spring application context against with standalone bootstrapping in a test for example +Sorry little mistake I can work with Mongo and this exception seems not to influence work at all but it is still confusing and it happens when multiple transactions are sequentially called + +Using the Remote Systems explorer Im able to list the files pretty fast +So if the partitions consumed in a consumer doesnt change the memory needed for metrics shouldnt grow +For now I just copy the code over top of the or code on Linux or Solaris +Click the blue buttonAbout graphs will go by and then you will get an exception +all the following should give out this servlet binsuffix bin bin binmyservlet +I fixed this in r with a different approach to the one you suggested in your original patchNow we determine the number of available bytes for a read using a faster ioctl call. +Struts seems rather limited without nested tags +Yes apparently the STS step is necessary +Integration test has been added. +that one is weird +the maven tree changes have caused considerable distraction for us and we havent been committing recipes just because of the maven changes +This would be nice to have in a release so client writers would not have to special case offsetrequest offsetresponse not maintaining the correlationid +moving to BETA can you answer Brians questionsLooks like the preloader fix went in +Yes i agree we should getit inDavid +Patch addressing Owens comments +cant run on a JVM right? Then this wont work since I need to have access to the COM objects because that installation of Microsoft Office is +Added javadoc and toXmlfromXml implementation +Looks very goodBy the way we welcome testcases as well +yes for version and is already on had the same issue and had to add the charset in the accept attribute on the receiving end +Plz improve the fix version +Im definitely not going to commit the same patch which is attached +I was unable to figure out how to the option to make wsgen generate an for each also see that the generated classes are in a different package than the one their namespace specifi +I just happened to notice this one. +Is there a way to unify the two? They full fill the same purpose group puts and deletes togetherNot at a computer right now but I can take a look laterOtherwise I have no problem renaming +reopened for fixVersion +I think its fine now +input paths with spaces +The services now refer to libext. +this to with the refactoring coming up +We can add arbitrary expression later when we see such a need +The first put on the was probably fine +Try with or as it should be fixed +this was critical for usability as the thread seemed to have frozen but didnt +Se also more word on the size +The example room number annotator description says that it findsroom numbers in the IBM Watson research buildings +Bulk closing stale resolved issues +Fixed in. +Might it be ported +The reporter mentions that there are versions of the class loaded and the application server is wrongly choosing the one during casting +Built on both windows MSVC and linux w gcc +I however run into another problem on unresolved view +Adding to SerializersDeserializers component +This was resolved when was resolved +Could the classes be left in the codeline and just excluded by the buildDo they even show up in the jar files? I would have thought not for embedded and I thought the client moved to a dependency build +The intent as far as insertsupdates are concerned is that setting the column to null is equivalent to not setting it at allMore precisely its equivalent to deleting it +We need to triage this at a minimum +In and can you move the new pathelements from the to the test classpath in ql +Closing the issue. +just did a commit for this +Theres at least one issue and that is that the Voldemort client libraries are not in any Maven repositories +If you make individual patches to that issue you might specify when you are done so that we dont close the issue to quickly but one big patch would be fine too +Context is now serialized and deserialized correctly +And while Im making a patch Ill check whether somewhere to modify or add to the unit test next timeThanks for your suggestions stack +The build does not work with it +I have also created the add button invoking discovery wizard with +The need flag on the termpos was getting reset on the last term in the span by +Yes which is why the spec states The Supports transaction attribute must be used with caution +From a code point of view this class should be removed but from a users perspective I think its more clear to have a class with a more accurate nameThe other possibility is to rename the already mentioned as and make inheriit from it +For example JPP ER had the following contents at path modulessystemlayersbaseorgjbossasproductjppdir Portal jppToday that folder does not exist and has instead been replaced with modulessystemlayersbaseorgjbossasproducteapdir contents of this new folder are dont conflict they show the actual layersproducts +one could select +Then that name could be referenced in discussionbq +bq +The Gora API has a flush call +Initial patch for review that adds an assembly descriptor to codegensrcassembly and configures the assembly plugin to be used +here is a new patch adds Field API for new int types adds tests for getArray hasArray adds tests for new Int types unifies some of the existing tests adds javadocsI think we ready here +Ive asked an editor friend of mineSame question for String syntax in ij commands Identifier syntax in ij commands +will save the update Info from heartBeat event +Bug has been marked as a duplicate of this bug +I have verified that this was a regression in my own code +Maybe later Ill check if those members can be removed or does this warning indicate some errors +I applied patch by Andrus +because java do not generates no arg constructor for you if constructor with argument existsAh my incremental test build succeeded for irrelevant reasons when I did mvn clean test it failed as expected +This patch is the trunk +on windows notepad is a good one to open pdfsthe first object obj is the document catalog and the root object of the document +Hi NicolasI my patch and made sure to select the license grant +Reopening to let QE double QE the mavenized seam projects were removed from the examples as they contain too many errors so the examples to test are Simple Example Richfaces Simple Application Simplest Hello World Portlet JSF Hello World in JBT +Closing issue as version has been released on. +Assigning to myself however because we may want to create some similar code in Castor in a util class to do these sorts of checks +Should be easy fix +AXISSTRCMP +Ive also done a regression test by running all the tests in Annotations and +Sure +that have been deprecated but still exist should be in the header file +There may still be an issue with the code which loads main relying on finalization thus leaving the file descriptors until gc catches them but currently running on with using GGC and not seeing the files stack up +Cant duplicate in not fixed in meant the aggregated feed box. +Attached is one with the same preamble used in the other files +We have a prototype solution in place for This solution solves the immediate problems for configuring pluginsReducing priority and moving to when we will complete this functionaliy +I think this patch should be committed +Yes exactly that but I applied the patch anyway and the tests are actually running if it doesnt produce more failed tests it will be in the good way to be applied on the develop branch and I hope be in the very next release +If youd like I can recreate and attach the relevant files +No trigger present but generated key will be returned by insert currently does not support this and throws a +bulk close of resolved issues. +Suggested way of doing this is through what I see this method is activated when user opens properties dialog for jar entries in Maven classpath container and like I said manually specified source and javadoc location is overwritted by this forced by the time editor is opened sources will be already downloaded if user enabled that in the settings +Patch applied at r thanks for your contribution please verify +Merged err and out stream for spawned process +Adding an exported dependency to the top level deployment to make it available to the whole deployment is actually already documented in the class loading guide. +Also appreciate if we could open a jira and fix the docs +I have committed this to trunk +Patch for source and unit tests +The proposed fix would require changing two test cases +If you remove it I believe there is only one compilation error to fix +They claim to be fixed in but Ill be sure to check for that +my +Added patch Fix for the issueThis assertion seems to be redundant +Please reopen +The simple case would be two threads in the async dispatcher +Thanks for the patch +TMU usually passes IIRC +No commit Information or Fix information present to verify the fix +You need to come up with a unit test that demonstrates the problem +Use setConf rather than parseConf +I have not verified using cqlsh but I have verified by calling directly with the following query DELETE FROM users USING CONSISTENCY QUORUM WHERE KEY and it works +Please check it +When adding the dependency the problem my two cents. +Can you also provide a test case please? +Okay Ill look into it more +Committed to trunk and for coming late to thisIm not sure this is the right fix we could just handle the NPE with a null check in + Bug has been marked as a duplicate of this bug +defer all issues to +imo ERROR is correct for the server case as theres no way that could be right and we want to bring to operator attention they need to fix it +Could we commit now Uwe? Its easy to tweak if we have to +It does not compile in however it compiles in and by the way indicates that incompatible with +to use an event driven architecture? For exampleThe interface would support a that could be used to register an event handler for processing the message +if you change format wont it influence other parts too? And do we want this behavior there +Committed to trunk with revision . +Moving this issue over to that is correct +JIRA seems to have mangled the tests so Ive attached them in a test I found the problem +Cool +changed the backgrounds on the admin and jdbc so that its obvious they are not editable. +Hi JustinPlease take a look at my patch +Thanks for the insight! I am sure that SQL is a beast internally +Otherwise the block is reported as missing rather than having a bad replicas +The test source is also updated +Existing committer list added to group shindig DNS entry for +Committed to upon release of Hadoop. +So exactly one consumer thread will be responsible for those messages +I explicitly needed to make aware of the implementationLets try to avoid that if at all possible +Fixed with build fresh got the same error today +Passed all tests in trunk +Getting feedback from people would be good from those who favour Antness as to whether things are going int he right direction and from those who prefer not to be driven by Antness whether the current state is objectionable it is worth noting that Gant unlike Ant supports a silent mode where no output is it seems outputs errors to and all other information to currently Gant always outputs to +Since the original patch can be cleanly merged to +Thanks John! +I was unaware of this scoping rule +wont fix it then +The changes are straightforward though so it should work +Do you have a pointer to some web page that says otherwise? Ill make sure to republish the site to get the since markings up there +great thanks very much +provided everything still compiles feel free +An issue has been created for to take Richards comment into account +attach a patch which releases topic when encountering bookkeeper exception when persisting message +We now restrict the ability to add the quartzscheduling binding to Services only not References +Attaching patch incorporating Hemanths comments +The problem here was that the night test did not have access to the default temporary directory +May be better included in Conversation? What do you thinkThanks a Cache Control +My to commit +Added more test cases +I think this behaviour is purely usecase based +and comes with a superb ratio of core code to test code +How are you going to handle those +Tests passed +Besides this rather massive testing done by everyone using cloudstack on Ubuntu we have also tested using genisoimage to generate the cloudstack system ISO in our own production systems +Incorrect names of jars it should be just and +Would you please try this patchBest patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +since Seam Tools model cant load some core components +Checked both have been closing all resolved issue that have not gotten any more feedback or comment in the last month +However the spec does not require this and Felix does not do thisI think we can close this bug right? +The expiration time would be the length of time that a host cannot have more than one connection. +The connection control message is always sent on new connection start so ends up having all the broker info by default. +That default may itself be a bug +Thank you Ted +If I comment the filter out things go back to normal +batch transition to closed remaining resolved bugs +But that your VE resolves to null is your problem +This should go naturally because the processing is based on the replica state not the list it belongs to +Heath I used +Hi BenjaminIf you could verify whether setting the Java formatters space and tab usage preferences works for you then I can close this issue as resolvedThanks the dialog I mentioned is only for plain text filesThe right dialog is Java Code Style Formatter +Atleast to understand the flushed file has a sequence id and all hlog entries which smaller than the max sequence id will be skipped when replaying edit logsIt means all data in memstore smaller than the sequence id has been flushed to storefile when flushing but broken by the case that snapshot called again without clearing previous because of last failure +for the patch +Let us do the rename as another patchcommit +Its not correctly IMHO +ill look into it. +Trivial mods fixed javadocs and a warning added a few comments +There are two roles for the web console and we should always assume JMX access is the admin access to the broker. +Im attaching a patch for the Windows cmd scripts +I see let me check +Reopening to investigate wise this is and the title was changed to reflect this +Hi Bryan +It loaded the question is how come something wrong with the metadata in your local repository can happenthis is the kind of problem that may be frustrating when everything works fine except for on plugin and you have nothing that lead you to suspect the repository contents +This seems to be more than a nuisance now +Hi AndyId like to see this happening as well +Who else is accessing the bag since it is no longer managed by spillable manager? Memory fraction should be a java property so that users can control it they choose so Why do we have limit of only tuples in memory since we already have memory limit? Also if we do need it shouldnt it be configurable +An updated version of the patch +It will of course mean patches that change the query language will have more changes in them but on the other hand its more consistent with including the generated Thrift code as is already done +Sorry I just sort of glanced the emails and this caught my eye +So whatever is wrong seems related to payload size or the nature of the payload in some way +Attaching patch +This was caused when the parser was either throwing an exception during parsing the parser should never throw any exceptions directly not providing any line number information about the errorWe believe most of these errors should have been resolved +What version of hadoop are you running? Have you recompiled accumulo for your environment +Thats what I get for copypasting too liberallyI also updated the file to mention cmake instead of the autotools and put the cmake URL into the C librarys INSTALL file +that the problem does not occur with Sonar +If you need it ask me +Yep seems it fails if mouse or keyboard is touched during the test run +it does influences how many documents per file how many files etc etc Optimize will follow those requirements possibly merging files etc Lucene in Action smaller mergeFactor values are actually better for searching +And this is only happening in +Also can you make sure that if is used the ordinal is setHmm and also a test case for this? Thanks +Fix is highly appreciated +So it was a problem in my code because I still use my own converter and have not upgraded yet +If we leverage the jbossws integration test suite all tests under webservices can be are nonsense anyway +welcomePerspective The preferred perspective in which to show the welcome page +cannot be redefined in a module. +I have followed most of what is in the specs to map the generated to +bq +Jiri can you supply more specific info on the issue please? Whats missing how are you reproducing it etc? +This is only necessary if we have to create a new user in the users table +Will it also work as easily with Maps +If I recall there was an issue related to persistence of primitives through Hibernate +All effort goes into jBPM +Thanks Harsh for resolving this +Thanks Ravi +If a full publish were the default users with large projects changing only an html file would be very very annoyed + on to add Fix Versions +svn revision Thank you for reporting this issue +here is the url for the image placeholderhttpdomainnamegeoserverwmsREQUESTVERSIONFORMATimagepngWIDTHHEIGHTSTRICTfalsestyledemBut this style works well in the layer previewIs there something still missing +The process is run the Ant script run then run ant mits +Both nodes look nearly identicalThe host doing the insertupdates shows a peak of around open connections and around in TIMEWAIT +Should be fixed on trunk in dbecbcffcdcdcbdbbf +The change is pretty simpleIf a cluster is using service RPC address namenode address in JSPS is set to RPC address +Release Note contexts can now be propagated between ESB services and jBPM Orchestrations +Thanks Steve took me a bit but Im about ready to commit to x +bulk close of all resolved issues. +Might be in a few hours or tomorrow morning at the latestThanks for checking back +bulk defer of unresolved bugs to +Please reschedule if necessary +I tried it on a PE and created tables in the shell I like it +If the above fixes arent enough and deadlocks will still be reported around indexing another option is available the index method which is invoked from separate threads can be changed so that the indexing operations are serialized per workspace +I it and test +caches where possible +Currently Pig does the sorting but Hadoops secondary sort can be used instead +Where are these and artifacts coming from +I think you need to add an implementation for now too +Thanks Kan +It looks like if no one takes an action this valuable contribution will be on hold forever +There are no layers added yet +Im sure this issue is whats hitting me when trying to run rspec tests from an rspec gem in a jar file as the stack trace is exactly the same for is pretty serious then +Thanks for the patch +i added the add line again but now in the bind method itself and not in get +Since Dan committed test cases and since fix has been present in trunk and codeline I am closing this issue. +For your concernsIf there are double separators contained in the boot class path provided by a VM it is definitely a VM defect +Thanks +The source code is also there +what were you trying to do +Same goes for the DOM of course as Cameron pointed out. +However something that might be a somewhat related bug is if you have Named and Positional parameters in the same query the parameters can also be bound in the incorrect order as well +bq +Stopping work on this until we have an EAP from the platform guys +has been released closing issues. +Yep +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +I observed the duplicate class error only when running it in defaultroot package +Originally I only issued the start once and never issued a stop. +Recommending a subclass seems completely reasonable +I am looking into the test failures +what version of are you working with +Closing tickets +Red bar with EAP related jars +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +Fixed some minor issuesRenamed the metric to raw +Is the old behavior able to be reenabled at allNo +Thank you Xiaobo +Thought I read this was from jbpm not from jbds +The name is confusing +batch transition to closed remaining resolved bugs +David last we spoke you seemed to have decided not to do this +rolled back due to +MartinOk +I can reproduce this issue and looking into it +Makes sense +Committed revision Committed revision +Yes I Im seeing the same error duplicate class definition when running web service +This is a temporary situation until I have migrated to WicketMany thanks again +Resolved without tip for menu +Thanks Krzysztof and Ashish for these patchesIts in rev +Yes that is my fault +Downgrading this issue to Major because it is not blocking the release as indicated by its reassignment to +AFAIK the write pipeline has always had this issue since the ancient days right? Or did the rewrite end up changing the order of the ack send and the CRC verification +patch looks good +on the latest patch +MDP container patch to support Nikolay I didnt read through the code with a fine enough comb Ill give it a shot in a proof of concept project and see how it goes +One other thing to note the JSON protocol also currently assumes that map keys are numbers or stringsI think we would want to exclude floating point types from being map keys because equals for floating points is pretty wonky +Corrected the ruleassignoutput handling +But that seems like a local folder + Update the Axiom tutorial to reflect made the examples compile fix the namespace samplesfix the Build and Serialize you please create a diff and attach here? Its much easy to review a diff than the complete file +Bulk close for Solr +Committed to trunk + +I just realised that there is no warning or error if the resource root does not exist which I have fixed upstream +reopened these issues so that I can move them to Future and remove wishes +However the fix adds a test case to which verifies the behavior of +I have only found one test classBut it seems that there are a high number of classes generated by +Ah +since it sounds like we should still fix the silent hang on unexpected exception +I would have no problem doing this but I am having a hard time understanding James dev environment +Added information about behavior of and to the documentation. +Ill drag this into since the patch is ready anyway +If I remember correct we only make API breaks between major versions so this would have to go in which I think could be a long way away +I am looking into Rave st as now it does not even allow to add this gadget and then look into this deeper +Yeah it caused +I let you close as you like +Ill keep an eye on this but letting you know about its current status. +In Camel its somewhere elseYou should really use as the type if you want to read an XML file +Vladimir can you create a list of the poms currently in the Maven repository which are still referring to jboss parent version ? If possible put the list into a format that can be read by a script maybe one line for each GAV that contains a reference to the older parent pom +Thats if this features going in +Please test to see whether that cures this problem as well +If ever spins once it spins foreverBut clearly we need a unit test for this! And we need to evaluate how we managed to release with such a bug clearly were missing something in our prerelease QA procedure +Works in Windows XP with Internet bug occurs only with Nautilus and Dolphin clients in Linux +Tried today again on my development computer and it worked without any difficulty +It probably is in some lost cluster of my memoryI think that if the server correctly starts up without exception during the initialization then you can ignore my null pointer. +Both patch look good to me +Any comments? Is that what you needTo be more precise I was looking for a utility which would tell if a block is corrupt or not given the OR given a datanode and a file display if that replica is corrupt or not +Ill push today +Rob this is bad if true could you let us know what causes itAnton does this only fail on import? what if you use existing workspaceIs there other update regressions? You did complete the regerssion test previously against a recent GA or +Now we mostly do not throw just make a note so this is code is no longer unusual as I see it +But now we fail in body check setting the right smtp. +Looking at old pending issue I found this one again +v Minor style and commentdoc fixes +sorry +Committed at subversion revision . +wholly undocumented? perhaps in the groovydoc itself but I can assure you it has seen plenty of printed and electronic ink elsewhere pretty much every Groovy book out there that touches the subject makes sure it mentions the mkp reference +Hi MikhailI didnt apply the +Moving this issue to to verify when we have an update from Messaging +Let me know if this isnt the case and i will the bugthanksdims +Now the patched tests work on Linux also +The variables arent substituted if they have no values +Could you pleas post a minimal test caseproject which would demonstrate thisI mean I wonder what Im doing wrong that Im not getting the exceptionI will discuss releaseing a version of the POM with the HEM team +Updated patch incorporating Toms feedback +I believe many people thought that was dead before the releaseSo could you commit it to +Try running ant clean buildin the root of Seam you are seeing instrumented classes from having run a coverage run +Here are test groovlets scripts you can use to test my code +Marking as resolved since the last changes complete the feature. +Hi just to let you know that there is a new release of that covers several bugfixes and enhancements in case youll continue to use the library +I am not asking to be a member of the JIRA Administrators +A +We can shred Protocol Buffer and Thrift data structures to it regardless of where it lives in the code tree +Updated spell checker class diagram of the adaptive spell checker with all java docs as lots of for deleteDocuments in andSomewhat hacky solution to pick up the deletions in but it is a solution +It applies cleanly for me but I was also able to upload it to Review Board successfully +if we are lucky the old transaction logs may help tell what was going on with that page +valueOf cannot be over written for enums so implemented a new method for the same +Committed the new integration test after the refactoring was in +So Im going to keep this open but mark it for resolution in on these issues until we have somebody committed to fixing them +Doug I agree Ill make this be the key +See for further details. +Since the input is invalid it should throw a checked exceptionWe want to switch to but we have a policy not to use software in beta +Posted the request to the customer +Testcase to show a blocked call when deleting an alias could confirm this seems to be a problem in ther server +Integers are not converted into marshalled values Vladimir +When I filed this Jira I meant to support pluggable compactions like level compactionI think we dont need two levels of plugs for compaction which may cause confusion +to D +The value is an bit integer thus cannot be printfed with d +This patch fixes it and also adds a comment to explain whats going on +Added element check and error logging to all decorations +m not sure what this would take but I think it would be good if we specified that its a string with a length between and in the WSDL +Unless I hear objections I will make the requested change only in both and branches +Basically Im requesting JDO support against all datastores supported by basically similar to the level of JPA support such that if I select JDO as my preferred persistence API there are no references to any JPA artifacts in the project at is worth noting that JDO introduced a standardized enhancer API in so that would be the preferred mechanism by which to enhance classes +Setting status to closed as no problems have reported +Didnt actually have the handle changes +The bug does not appear in but it does appear in +This also affects the newly released now +Hi MusIve sent a pull request upstream for moving to using JAXB impl which has the fix for this issue according to the description +Please verify. +Since we are upgrading anyway we should try for the latest version unless theres a reason not to +Ill take a look at the changes to and see if they allow me to resolve my issues with the tagsThanksEdgar +patch looks good some nits Convets should be Converts ioe should be static Cange should be Change +Hi Paul attached has the fix +The clock changes would get into trunk quicker if we didnt avoiding the extra overhead of a big patch during reviews merge with trunk code updates and publication of a new patchIf the concern is that we wont attend to the clocks once this patch is in I can promise that well look at it straight away +But from the other point of view there is a question of compatibility presents here +Please review +Since thats irrelevant to the Application Server Im not going to hold up the release. +Suggested patch +Need to check those headers were added in and then we can close this one. +Now fails with a different error +jarExtracting systemorgapacheariesblueprint systemorgapacheariesjmx +Jason any chance of this getting included in? Currently Im maintaining my own branch of the code for my company to use +I dont know because Continuum do nothing +All I got is in the first comment of this jira which is not sufficient to help with tracking the problem +This is fixed in SVN +Thanks very much Andrew +Closing +If you find further issues please create a new jira and link it to this one. +Committed +While I can see that huge god classes are a pattern in the code base I would rather stick to one class one responsibility and use packages to structure related code +I opened as the blanket issue for addressing on next POI upgrade so I think we can now resolve this one. +Looks like this is dead for now +A global changes would make changing away from update notification easier +I have added a further testcase for the Compressor analogue the same problem manifests there so I have also made the corresponding change to +Paulex when I changed patches and I lost some code which fixes Test its my inattentionCould you apply new patches +In the meanwhile we could look at the above mentioned drawbacks and decide whether the they are worth fixing or not either by restricting solution above or some other solution +add an explicit statement to set default replication to in the junit test +Please wait until the mailing lists have been +irbrc where you can define this per table so that these classes are loaded implicitly +Ive put a null pointer guard on this for now leaving open while we consider other options for WC work on WC Widgets conformance testing has made this issue redundant +Attaching patch +This is because at no time does Acegi Security itself go and instantiate +Tooling could also be employed to help turn existing methods into methods with a wrapper object +The is similar to what Im looking for but an is generally more usefulAlso does not seem to support more than a single term querytypeAUTHOR typePUBLISHERstill lets typePUBLISHER through +EXECUTE Add a jar with bundles to the class path of a CDI project that has Seam International module +with Enterprise Web MichalYes selecting EWP results in the behaviour you report +Long was saved to the disk using the serialization method in so we need to use the deserialization method in to read it outI incorporated the second comment in the new patch +To prevent the cyclic error portal and cannot be both used in one reactor taskThis was the case in the subClean taskTo fix this I put the in the excludes list and added a clean task for on its own +series there is a specific property to add your own jars before the MR framework jars +bq +I think we found the worst problems +Some implementations use the DDL for configuration +For Pis latest comment I will try hard coding the temp file path to see if that fixes the issue +Went through and reviewed gateway code did not find any instances where should be used and is not. +Modifies the boost function parsing to utilize use the class +Im leaving the Tuscany code as it is for now +The addtion to the test is a method which creates and uses statements in such a way as to provoke of statements and result sets on network server +Maven version of the project attached +I know the spec says methods arent supposed to throw but there appear to be bugs in Suns JDK implementation in which are thrown in some cases +Committed to +Any thoughts on this whether this is a reasonable thing to add to Lucene +done +Attached file contains the schema and exact query which fails on HiveIt is worthwhile to note that the same query executes successfully on Hive +Can you reduce this to a standalone test case we can run? Punting to . +These two enforce a limit in YARN itself default is Looking back again at your stack trace I agree that it is very likely you are hitting the K limitCan I close this as a duplicate of? You can verify the fix on when it is out +Yeah as this will no block with release if there be no room to introduce changes to the apps +Unassigning for now as I am working on and first. +Why not just put them into the root directory in SVN? We have the Eclipse +Unassigned Rohit +I dont think we should have the backwards compatibility allowing people to make options without descriptions we want to ensure that all future options have descriptions + +Hi RashmiThank you for the contribution +I just committed this to branch +Now waiting for more information on the POM that also has to be uploaded +In this latest patch I have now removed the code to set the default for the wsaTo header +Not reproducable +It may not all be relevant anymore +Youll see through each log where I sanitized private data +We did all tests within the weblogic container and used the weblogic jms implementation for the asynchronous dont know if the problem occurs in the asynchronous proposals in this issue were independently developed in the following issues +Im closing this because there are good reasons against doing things this wayAnyone needing a Foxpro compatible file can use the file +Attaching new patch that incorporates Eniss commentsEnis Ive made the change to add DDDBRR extending DBRR and added a on top of that after all +If all region servers are dead except just one alive this could happen +Ive just gave it a try and I still cant get to details page of the painting +Committed revision Song thanks for the patch +Manually tested +Adding release note text +The problem was that maven was taking value of some from cache +That would best be A file of our ownIts the test itself at fault I wouldnt say its Windows at fault there +Thanks for reporting +The failing test has been commented +for Expert configuration with defaults set to failures +Anna Im currently working on a fix for this +I did +Accepting the bug +The classes use all kinds of variables that are protected. +This is a broad symptom of compaction failure I suspect +Bugger +The patch aggregator +Both show a different A h duration gets magically converted to a h duration The Left to do is calculated correctly because h are substracted from h +To try it out requires also downloading FOP and replacing the fop jars in lib with the jars from FOP and then including all of them in your classpath before starting a pdf build +Its for both injections and events +updated fix version to reflect merge to +Ill check if its possible to be rebased on trunk. +So the changes are limited to changing the parameter and field typesAdded filesjavaclientorgapachederbyjdbcjavaclientorgapachederbyjdbcThese data source extend from and respectivly they implement methodjavaclientorgapachederbyclientThis is implementation of +committed +Vinodbq +Then shutdown and undeployment is performed DS is undeployed and then the job is it be simulated by modifing source code +regards solving the same problem in you check in the changes? +I can no longer seem to reproduce this with the latest build of trunk +I thought I saw the status as unsigned when Iassigned this issue to me +It also contains a regression test which verifies that the nullability is as expected in a left outer join and in a right outer joinIll post an updated patch which contains the necessary updates to the test canons once Ive gone through all the test failures and verified that they are expected +packaging coremath and collections to one Jar package for is very imperative in real environmentIn a real projectif we used and want to package the whole project to one single jarwe have to package coremath and collections to one Jar package manually because mahout update quicklyso it is very annoying so thisif Mahout project can provide thiswe can depend it in the Maven POM file and package the whole project and its dependent jar easily +Ive updated the static pages to reflect the new console lookfeel +We get around that and several other classloader issues when we create the run scripts +Now the missing pieces are enabling it by default documenting itThe first bit I would like to do only after some feedback which means we first have to blog about the curved labels easter egg in +See bug for a real solution to the problem + +The NPE resulted due to fetching the for an INSERT statement which was of course null +Would you like to commit it yourselfCheers +Theres no reason we cant also include an as well +SJC is now a compliant bundle. +The file patch assumes this is at the top level +NOTMy Suggestion is to updateSTEPYESMy Suggestion is to updatepipeline +This issue is caused by not finding the wsproviderinvokecommand class in the plugins +The above comments are the latest port information for EAP +So I need to make it a plugin first so I can initialize the hyperlink detectors when the plugin own lifecycle little more convoluted than expected +Hi Manik This was on the forum post +OK the reason seems to be a wrong and btw +didnt mean to do that +Jared I fixed your Release Noted Text where you wrote turn off Seam EL in +This is not because of any of the code but because I am in the process of having a Hudson job setup to run the testsuite and get a credible integration status +It could be on a different rack but if you are following best practices and alternating racks then it is likely either on the same rack or there is only one copy on that rack and the best case possible is streaming from another rack anyway +So the platformcompiler definitions should be unified with the common gcc settings +Was that intentionally introduced by this patch +Bulk closing stale resolved issues +This was not against you but I am hacking native Windows since Win was started But you dont understand how its IO works +Fixed in r. +Thanks +They were introduced to support locatorsNormally I would wait for your reply on this one but because the release candidate is getting very close and you say you have no objections to the current patch I push forward and commit itThank you for your comment +You mean the plugin? Works pretty well and it has its benefits +It seems safer in any case to use the size of the records list for iterating rather than the size property +Implementation of HTTP digest good thanks! Some comments The class is probably best placed in +There are now examples in the wiki for php python and java +That explains why the value of window state is always the default the other hand the field current is of type which does not implement there would need to be a switch to which is be verified in gatein master and reopened if necessary +activemq startAre we missing doing something for the recoveryI am another log file named +Ive just assigned it to you +We have a user forum if you have questions about usageThanks +no need to wait +Alan if you explain Ill try to work on it. +patch looks balancer tests are passed that verify the correctness of the patchIve just committed this. +All three subprojects have links to some amount of common modules that are our framework +Thanks for the update RomanMapReduce binary depends on common for other things too and so is not really so we will not need this anymore +committed +Improoved version of just came to my mind that we could obtain both the private method and the private property as well via reflection +The getDocument method will now check to see if the parser is null and will throw an if it is +Verified works fine now. +This patch fixes the and contains two testcases to reproduce the problem +This is because of the old jar files +Heres a patch that includes the patch from and updates the descriptionI also added a test but the test passes for me with or without the change to in either Ruby or +I still intend to do some more testing but I think the patch is ready for review so Im setting the patch available flag +based on this change the Getting Started with and Xcode Template needs to be updated as updated. +Supersedes +For other vendors its fine to skipIts fine we cannot solve all the worlds problems +The init method of wars servlet has s sleep +Daniel As it is very little of our custom code uses the shopping cart order helper java methods +applied the patch given by chathuri +Cumulative service pack fixes the following issues Fixed issue Predefined variant types not supporting dynamic event types Fixed issue incorrectly posts remove stream data for shared state under view sharing that this problem exhibits only if event type alias and filter are the same between two or more statements and only for the stdunique view +Comparing the diffs for and this issue will show what is actually changed for this issue +The client side of this has been moved to as thats going to the release +Maybe I am wrong but it seems to me that you are shutting down the entire system +Changes for and under this bug ensured that a Query too complex exception is thrown if the resulting generated class will not be loadable by the JVM due to exceeding limits +Did you also try with a fresh checkout of the branch +Yes mine seems to be a duplicate +Working properly in HEAD +Applied the new reviewed fix on the branches and trunk. +Now Images search result page is opened without Image Dimensioninput and info icon +Maybe Im doing things not correctly +Using maven shade plugin or some such +I forgot to update the Mocks in testThanks againChristian +Hi AdrianoYou just forgot to delete the query parser files generated by javaccand regenerate the code from the using javaccThis created new classes called and but your code was still using the named classesNot a big deal +Considering resolved unless you have a test case showing the problem still exists on the latest build from jbossastrunk. +Actually I was thinking of categoryChooser +This is a lock type the user has explicitly requested +Fixed in revision +I just selected axiscar to export the server in admin console +this adds all the current prod store emails as default to Note this does not cure all the problemswill still get errors till the rest of the patches are adds all the current prod store emails as default to Note this does not cure all the problemswill still get errors till the rest of the patches are implementedps sorry of all the extra notices +Actually the real problem here is that we dont keep track of the operations on a handleclient and therefore cannot cancel them +extends the pluggable to toggle between Required changes to and +We are planning a centralized Synchronized solution after interceptors are done +We hope to add this feature soon +Moved docs page to extension in user guideCommitted to master fa +Closing +Clearing fix version +Thanks again all +That did the trick! I was so close! Thank you so much for figuring this out for me you just made my month! +Targeting Hivemind or later for utilizing JDK specifics seems reasonableWhen is Hivemind scheduled for release? This year? No earlier than would be a nice to have +Good catch +Albert +User confirm that now it works for him. +Can you list the tests that failed on your machineFor security profile fails consistently and is tracked byOther than that test we should be careful +It happens when using AS and EAP included in JBDS +Pls review it if no problem i will apply it thanks in advance +If we get a status back then we probably need to authenticate otherwise we don thoughts on this +server crashes AFTER weve verified it up in assign but before we get to do a scan of +goes ahead it might be a natural place to make everything in deal in FieldabesWhat stops us from resolving +If you guys are fine with itchange dyn from hashcode of the parvitionValues map to something unique such as the path stringsplit into two and there is really not much code sharing and this adds a lot of complexity because of having two almost mutually exclusive code paths in the same classesinstead of serializing hiveConf why not include it as part of not sure why we are doing this +just for that feature +Committed revision . +Stack the shell outputs the binary keysnames just fine see the example up top +or will I need to build Roo locallyThanks again for your time +Removes an if statement that has always been skipped and uses entrySet rather than keySet a get +Verified in Metamer. +Id love any pointers to guidelines you have +MahadevOK +It will be checked in tomorrow evening if all goes wellThanks for supplying the patch +Thanks +are any of the other build servers seeing an issue? The opengeo one seems ok and I cant reproduce this failure locally +Shalin Shekhar Mangar how can I use the proposed attribute feature to be used for masterslave configuration? Do you have a code snippet? +See +Maybe we rename this JIRA to Support XA Transactions in SJMSRaul would be the perfect place to do that change +It nearly took days out of me +Alan this isnt proof that the client has a problem For exactly the same reason the server might breakIn other words you should repeat your example tracing the servers outputApart from that the important information is missing what transport factory you are using Obviously thread safety is actually a problem of the transport factory +If it works for you and you can risk the wrath of webmasters thats fine you are free to use this patch but Nutch as a project cannot encourage such practiceConsequently Im going to mark this issue as Wont Fix +If it is feel free to use it locally since you clearly know how to generate it +I tested Alis suggestion but I got the same errorIt seems to be related to in my weblogic classpath +Will commit if tests pass +All Done. +Currently the implementation only provides for the training logic. +lets close this JIRA +Brian dont we already have this? I thought this were exactly what we are using in the webservice wizards today +It was full eap testsuite like running for one off regression checking +Commit changes to trunk at revision and snapshot at revision thanks Han hong fang for the patch +Patch catching to the new class with this functionality +Changing the default implementation to return false would be an incompatible change potentially breaking existing subclasses +This is profiling snapshot under for with list +Looks like the text paste of the patch had issues +This umbrella issue is all about maintaining kind of tests inside +Hi ClausI found the duplicate and closed the other we need two different mechanisms one for OSGI and non OSGI + +No yelling here maybe a stern talking to though +Then the feature could be used in other scenarios +Targeting bnd which fixes an incompatibility with the property change +This stores the of the TT corresponding to the +As expected removing the primary copyright notice removes just the generated copyright from remove just the generated copyright does not seem to be an issue in the current codebase +Reviewed all these issues and marking them as closed. +Reverted the used version to be again and removed the entry about this inside the release notes +Also you have to give target directory for heap dumps +Resolved as a part of +My patch fixed the bug +I am going to check in to trunk once tests pass +Sorry for all the extra comments +Fixed in Committed revision Thanks Suran +Bulk closing all of these resolved issues. +For me it looks like this may be a problem with AS then +It passes precommit java tests but I have not yet checked for regressions +implementedConfigurable only via configuration files currentlyTODO tru the ui +Are these the same Oracle g and databases that are being used with the Jackrabbit runs +I have published and +Juergen +Thanks Mircea +The adapter naming should reflect the specific mapping layer being used +Firing an event if the quality profile to be used has changed would be a good start +bq +Please recheck with current svn and if problem persist re open with more details +Can I ask what coding is being done for this? From my perspective this bug is fixed in +I didnt see it previously +I have also unified the Persistence Unit classloader with the JPA bundles +Done. +The real fix is for the block cache to be instantiated in HRS and not be staticThis slightly complicates things but is possible +I can still see both files in SOA IR release which has ESB included +Its too much for me and is bound to lead to confusion +Please take a look +Will edit the ticket to specify the correct version +I have verified the proposed modification fixes properly and. +bulk close of resolved issues as part of JIRA housekeeping for release. +attached is a patch its a little ugly since doesnt implement Replaceable +Damn! You are right +Thank you very much +Ill definitely add this to but probably not the branch +A test case to validate the behavior +Also would be nice to show how a JDK Name implementation behaves in comparison so we can see whats happening in the test case +So it sounds like downversion clients can talk to upversion servers but what about the reverse? If the client sends a message that includes the force parameter will a downversion server simply ignore the additional parameter +I am using SOAP HTTP +Not sure how well the two would interact +Yes we can rename a file +USE THE USER FORUM +Im going to close this +Are you suggesting we should just enable to enter some data manually +Let me try out the test here with the patch there +And youre comfortable the other test issue is something else? This passed the test suite for me that that was a while back and Im not extremely knowledgeable on all the areas of the code Im touching here +Code applied +Tests passed cleanly for me on +Committed to branch. +what test did you run to get those numbers +This issue is out of date I will upgrade to wrt +The quick fix seems not to save the files correctly +Lowering priority cant be criticalRequires Spring as spring has to do some internal changes to allow custom spring handlers to get access to the property place holder stuff +It looks okay to meCould anyone please take a look at the diff and also verify the fix +Ales WDYT? Do we just delete it +Enterprise Maven repository do not seem to be available +Here Ill reiterate so you can avoid any sort of disillusion in the future the Hibernate team has not ever and will not ever upload artifacts into any Maven repoThis has of course all been discussed many times on this very JIRA and on the development mailing list. +File it as a bug with ruby core and link to that issue or discussion here +This away any attribute can be injected into an entry to effect evaluation regardless of the partition implementation +Needs manual adjustments of a few paths depending on your configuration +why QA complained The patch appears to include new or modified tests? comfusing me it should be just one test expected +Closing the issue as not a problem. +As I tried to explain above the reads from the WAL files at offsets that are stored in ZK +Fixed in trunkPlease verify +is using the Thread Context Class loader in order to resolve the class +For my local change I did start off by doing it as a protected factory method but then changed it to taking a factory in the constructor primarily to keep symmetry with the constructor in +I tried even using Fetch and but that that was worthless +Shows odd behavior in web is called now thanks JDCurrent issue with this patch is that following first split META table eventually loses all references to the new split regions +Jrgen has fixed the core bug in Spring and +I committed the fix +Though its not a big issue it maybe worthwhile to put the zookeeper data in separate directories for kafka and to avoid obstacles +The new version of dispatch a JBI message for better demonstrationThe basic question is does a binding component must have default marshaler or not? Service unit using Axis binding component must define client andor server classes +In our dev i just create some templates to contain only the form so there will be and both called from and template +Done +Fixed in revision +Thanks Kamil! +It can be the case that finish throws exception before unregister event is sent out +Add more debug printouts on failure +This setting will be ignorederrored out for vmware deployments +Yet I dont see a NPE here thoughI fixed some whitespace issues in the patch and refactored your impl to use a switch statement instead of if else I think is less verbose and has less duplication but as you said thats a style issue mainlyI will look into the FC issue and move forward here ASAP +The forum has some replies from Adam Winer from Oracle but Im not sure if hes still associated with this +I prefer this to the separate class construct +Hmm okay and purged some messages from their tables +Patch applied on and trunk +The above build did not even get to and +Unchecking Existing Application Impact +Thanks guys. +do we want to have an additional param which tell us what data structures to use? and by default it uses int? whats the best way +There is no way to run against Hadoop +bulk defer of unresolved bugs to +Vladimir could you please provide a patch script for generating all missed empty files +Fixed on trunk +No response assuming ok. +Changing to Doc hopefully someone picks these up. +thxhint when using windows dont forget to use binary mode with fopen rb wbmfg for the hint +It is working +Fixed as suggested by FelixAt revision for the fixUnfortunately this only partly solves the problem because the is still implemented +I dont have an access to a Chinese localized XP where I encountered this bug but Ive reviewed the patch and I think that it should work +Kevin would like to have more flexibility when building the JNLP file and thus requires more per dependency +May have been reported on another Jira ticket +Also the new code you added and are missing in this patch +To reproduce this the bug set this constant to false +The immutable file and the copy would share immutable blocks +Yes Brian I think it is all is working +I commented on the pull request perhaps we could try to add the logic that goes to the Hash itself? Or is this the way MRI handles the logic +Maybe you give it a try +Closing as duplicate +I am guessing that they wont be lost as they are now stored in the workspace +I spent few hours after realizing that the real implementation of indexes was not the one described in the proposal page +We just saw a user report where the statistics were messed up due to a bug in Derby and this caused the istat daemon to go crazy +Can you elaborate why using is very inflexibleSee Annops comment Oct about limitation for patch vbq +Also adding the themeName to the javascript location for the theme specific resources allows us to mirror the setup for themes we have the following directory structure that we use jscookmenu other image files +I set the test to randomly kill tservers or data nodes every minutes with a rest before each restartI let the test run over the weekend with the HDFS Trash turned on for analysis +Covered with selenium test. +The grant revoke tests passed with these changesCan someone please reviewcommit this patch? wrote I have moved the drop trigger code into makeInvalid call +Punting minor improvement +Committed +Initial looks good to me +That second issue is probably browser specific +It only happens on domain mode on standalone mode works without any problem it could be a issue related to the differents jvms involved on the domain startupI decided to work with standalone servers for my cluster while do more tests with the domain mode +I am looking at logs now to see where the problems areTracking current failures here the short term if all the multicast addresses are set to the value of udpGroup this will be provide all with a reasonable default +Ive just committed this +assign to Nadav as he is the person who wrote it in the first place and therefor the best person to commit the change +That way its delinked from the the sleep period of the loop +The runtime would pretty much undeploy everything that was in that folder then redeploy those that were left after the seems like a lot more delay on the runtime side +Attaching a modified document incorporating changes from Dick +For me on and work fine but the arrow keys do not +the compiler looks for implicit conversions for conditional blocks +Rakesh +The version of servicemix include springframework packages so even if I include them in the exports of the system fragment I still get the error +I am fine with your proposal +The patch is attached +The problem is that the method is being used by other methods which do expect it do encode the URL +on patch +We can configure the threshold +Sorry for being so slow to understand +has its own package space and is therefor shielded from such problems +i am using +backend server needs to be ready also sure the latest version of the Push SDK is being. +patch after yesterdays merge from trunk to vectorization branch +We need to include and +I cant have a HOME +I will investigate the issue next week when I have a few days for ossThanks for reporting the bug +Ok got itHere is a patch which add the emptydirectory test and also fix itIts still throwing the which will most probably be fixed when I will apply you patch from +someone please add Harsh to the contributors list for the HADOOP jira so we can assign this to him? +I think that Sharads patch fixes my problem +Perhaps we could now remove the generation comments? Thanks again for your timely processing of my patchesCheers Adrian. +If you drop this in a install do you still see the problem you described? conversation a couple weeks ago w Bela he saw the not working properly whether or not shun was set to false +Ive created for issues and from above +Thanks Richard +I can confirm that still hangs onAlso I dont know but this might be related to. +Thats fine but the longer term issue is the latency in getting new releases up there +Update the patch to see if we can get rid of the complain +Some of these issues are awaiting retesting after upgrade of ODE version and merging of the JPA work from the ODE trunk to the branch +I dont understand the Further calls to +Found a couple of tests under the ID reopen to remember to investigate if now fixed +Final patch for preview +ant test and passed on my machine +As Marius says any managed bean that is not itself an interceptor or decorator +Add the plugin in the jspc profile +Here is the patch for the plugin that generates for +Filed a new one +Arno I do not have the ability to close this issue +Not even with the same seed +But now they are configured with threads with two instances per node so the memory requirement increases byThe end effect of both lead to on everything running on that node +Mark did you include the tests anyway and just set them to only run on Oracle for now +Your best bet is to implement the annotation yourself and to submit a patch +The nice little patch looks fine except for the point that adding or removing the wild card character either as part of users list of the groups doesnt change the ACL at allYou can also separate the newly added cases into a new test +Probably would be changed for Maven so that it can be tied to the POM model version +Yes that was a bad patch +misunderstood the exception handling in +Awesome patch looks great to me and does indeed our most famous todo gadgetThanks for the fix patchs been committed +This can change with though +If anyone wants to fix this can check for errors of this type automatically +Is that really going to be faster that doing the filter query that was discussedAnd +INFOconsoleDRFA classpath PIGHOMEbin +Curious why did you close this? Seems like it should still be a bug shouldnt it? Even if theres a workaroundIm seeing this exact same behavior I was having a hell of a time figuring out why my filtered ones would be different +Lets continue the discussion on. +very nice patch +Retrying hudson the tests failed due to dfs problems +Indeed thanks for pointing this out +We also found it necessary to provide an implementation of as the alternative version within the application server codebase relied on a specific ordering of the synchronisations which is not supported by the JTA or JTS specification +Thanks Mayank +For now I am hoping this issue will be used to add message types for stop and failoverin +I have deleted the +Push to timeframe and look at the state of Avros CC bindings then +Compound triggers appear to be an extension +Again if someone has an incompatible browser theyre still no worse off with the patch than they were before +Long saga of this code change +Its on on its way to being an interface +This was a failover of the master case? It was not a fresh start? Only on failvover do we do the contact of all regionservers +BULK EDIT These issues are open to be picked up +I think its too late in the process to change that we need to care about backward compatibility +I can also do this +Reassigning to Doug +Will this also be applied to and the branch +I assume theres no chance you will be able to test? Also do you know which line its crashing on? And what the values on parameters and local variables are +Logs included. +css need cleanup to remove our local changes +Suresh is there a spec for me to look at for this issueId like to understand the documentation hits +Hmm +I added support of KARAFETC and JMX RBAC in +We have other validation checks that we can only handle at runtime too +If you havent signed the CLA could you please do so +Any idea for breaking this into smaller chunks without sacrificing accuracy +It works for me on Linux but Im the one who submitted the patch so thats expected and I figured you needed some external confirmation +This is caused by the same issue as +Is this really a major issue +Verified in and +This might be termed as an incompatible change +Does not Operal would need a plugin to show SWF movie? Does your installation have this plugin +PS I should have logged on IRC but I usually dont do this before leaving for work +I guess people started using CONFIG more broadly than I intended +els enjocs sn alts +We dont want old task attempts authenticating with the new app attempt at least not at this point +But in either case something to communicate back to the developer of the C code would be interesting +will commit soonregarding the actual issue its do do with failing over while the transaction manager is recovering +can you whip something togetherIts in trunk and will go out in in its current state but Id sure love to clean it up a bit especially the matching logic +is configured as a however it make not distinction between an activate for an alias PID and the real PIDIt should look at the properties to determine why its being activated +I assume you meant +If localname is always empty it means you didnt turn on namespace processing and you should look to the you are right! I forgot to check that! Thank you very much for your help +better +This has been done in scope of M. +I have been working on some code that uses Spring Integration to process chunks and I think it should go out as a sample in +Suspect user error as per comment above +REST of the encapsulation is done in +is fixed now +The latest patch applies to both trunk and +So on that notion +heres test with fix toI tried to rearrange this in a way that its not confusing +My guess is that you dont need a huge machinery and will be able to implement the local disk and HDFS versions with the same code +Thanks Konstantin. +Closing it due to no answer feel free to reopen if needed +Test into the master branch of the repository +Source files to be considered for contribution to Kato +Thanks +Thanks for patch NitayI tried it and seems to work. +Not sure about AS. +Actually I added this test case and used it for proving my fix +leave this open for the forgot to do extends Serializable besides implements Serializable +In other words it takes two clicks to expand or collapse +Developers should have some that could be run to check that their change is compliantCurrently I run something ant clean test javadocs dist to validate things before checkin +Can you try out the latest when you get a chance Markus +Patch applied thanks ThomasI tweaked it slightly to work with thus the identified intersection is removed from the hash set to avoid duplicatesIt feels to me that cpu performance would increase if we built the hashset on the larger of the datasets as we would end up walking the smaller list but it also feels that that would hurt memory more oftensvn ci m Applying Thomas Rogans patch to improving the performance to in the manner described by Jilles van GurpSending srcjavaorgapachecommonscollectionsTransmitting file dataCommitted revision . +But can it bethat this is nowhere evaluated? I tried to set but the db is stillcreated in pwd +We now generate separate style sheets for secure pages +If that seems to work correctly perhaps the solution is to remove the if entirely +The final is only set for the last attempt +Hi SenthyRight now Forge does not comprehend settings in projects +I received no error but server log contains warning and queue size on A is on B there is message in delivering expectation was that after restart with only one message in queue B and no message in queue A I would be able to move it from B to A +there is quite a bit of juice to squeeze from the Java implementationI agree with +Hey guysStill looking for what action Im supposed to take here +Henri look at line to see the method which is the equivalent version +We can move this to next CP +and handle this as null +I will report back +I also faced this issue +Ill apply the patch and commit the changes +But how do I then send additional properties from the SONAR plugin to JIRA +It lets you have a smaller number of regions at the same time as having the most efficient compactionsI agree +Clarification based on Allens comment in my case the files were not in the conf directories they were the ones from the jar files +Illustrative patch +So all you guys work for big companies and make big money using our little free software + +Hi JulienYou can see in patch I take this information from Generate Mark Counter +Additionally Ashutosh ran several ee tests and also verified that this change fixed user problem +and Compound expose a token api in a similar way +Upayavira has the karma to press the magic button on the website to the whole structure +Dup with sorry for duplicated submit +If NN is a new server the client send the new get or get which send back that carries only a local name in byte formatPotentially could also written in an backward compatible way +For now +Hello Clauswe edit this in paralel so that i havent seen your second answer beforeBy the way ther is a small bug my name is Dario +This could certainly be enhanced to support rendering a page from a partial data set containing pages of data just havent gotten there yetThe tests committed with the change provide examples of how this works +Hey guys any chance to see these changes in a future version of thriftIm willing to port the patch to current HEADbestd commit a patch that applied cleanly and implemented comments on individual elements style +Ive committed this +Here is a patch with your comments incorporated +It would be nice to have this info available on site +They could probably be better with more details on which track has been messages +batch transition to closed for remaining resolved bugs +Btw there are also spaces between the list of required bundles +Committed as r +Thanks John! I have now committed your patch to with revision . +ItsI fixed the issue Im retesting before uploading the patch +I have no unit tests yet to verify that it all works but Ill get some together early next week +According to the Basic profile you need to declare the part you going to use at the soap binding if more thanone part is given in the messageAxis Codegen follows most of the recommendations in order to make the code generation simplified + +Please attach your workspace +Looks like a checkin comment typo caused some commit build results to be posted to the wrong JIRA. +Hican you change class so that accept optionally some kind of factory for the facet collector? so that we can decide outside if we want a different implementation of the to be usedNicola +This option means that if some file has changes on ST server and that same file has been changed in the local working directory it will get merged automatically and without though this situation would be not normal to happen since no one should be supposed to go into Continuum working folder and modify any files it is conceptually wrong to my opinion +However it would be extremely tedious and verbose code +Fixes javadoc and the timestamp I forgot to check +In any case for your first question the size limit can be checked every time we roll the log +I check the log and find out that our last sync revision is r +Committed fix for +This is going to miss upgrades to the thrift library should make this easier +Initially the formatter option only applied to the session +Ive moved out of you paste in the discussionexplain why you dont want to move up protobuf? I know its a traumatic migration because google broke the but as we did it for Hadoop and in near lockstep itd be good to have bookkeeper in Steve My understanding is that folks felt that it would be best to shade so that we dont have to be in sync with HDFS with respect to the version of protobuf +Ah ok +goes in xercescsax +However Im totally unwilling to defend this against any opposition I just really dont like two container elements with the same name holding different things +Changes the default to be generate default message using same static var as used for default adjust default to be or at least number of reducers +Attaching the patch with new name so that it is obvious it is for +ok +worked unit test were wrong +And then a failure on ubuntu +This is nice +Let me know if I can do anything to regardsI will check tomorrow +Is this something that is almost fixed or nowhere near +Lets wait for Hudson +Duplicate of which is a simpler but sufficient solution. +It all works correctly now +Please use clean rebuild in DRLVM to resolve the issue. +Any chance you could provide a simple test? This would help tracking down the problem +I actually moved the contribution from being an Import wizard to a popup menu since I want it to be available only on the project root because I believe it doesnt make sense for the checkout to be made on arbitrary folder +Customized entries now take precedence as expected +added exempt from verification flag to the transport in trunk +Its private in both cases +I did a quick sniff test by mucking around with start levels +Committed as +efccbccebedebcf SPDY plugin sends malformed HTTP version string +I will take a look +We did a load of work to change splitting so both came up on same server improving MTTR +If I override schemaLocation in my code I do not want nor expect Xerces to look at any schema location hints in any documents +Im not sure I understand +i guess with current pom you mean +If I then hit the second server I have no problems loading the page but if I go back to the first node any operation results in that exception +that means a file is to be compiled which needs another file which then is compiled but depends on the first file +Do we need to add requestID in response also? I do not see the need +I also do not see the reason why we need to reference to the article rather than to have more correct introduction to the technology that we can see there references to useful resources is a common practice for books +Look like the patch is using the framework to implement tests +moving off of this milestone m pretty sure this is redundant now that the provider filters have been heavily refactored +If you would like to help out the give us a shout +Move the offline image viewer and the offline edit log viewer into the package +Im not either +Might require some modification to the view drop down etcAlso this throws an error just like all other queryparseerrors so I think its technically a error instead of a as suggestedIncludes test case in the in r +i wont be able to do it this weekend +The PR help to Fix login issue as described above now we check user existence by instead of query for user profile Handle case for disabled amended the PR to avoid creating a new and merged it. +You can test it yourself by using the class +Hi Luciothe problem with this modification is that it breaks backward compatibility +its probably faster to insert the code by hand than to apply the patch anyway +host has address So i assume youre james server is running on the machine with +Please confirm the current status of this issue and if its incorrect change it as required +It would be nice if every SQL construct worked but not having full sql syntax is not a burden to most people and supporting it all would be a burden to us. +Thanks. +Going to going to see if I cant flatten out the impl of it a bit +I propose the shell be updated to invoke get is a privileged operation while get is not +If there is something else youd like done with this just let me know Otherwise please resolve again +I will work out a testcase that I hope will convince you the need to remove that throws +I was waiting for the pom to be fixed +The normal casting doesnt work because we need to compile those casts in a special way and we currently dont +Hican we close the issue? Seems to work since +html make sure to check the of the +Todd do multiple junit tests reuse JVM? If so you are better off adding this to and +For problem and is OK. +There was a bug where only the first bytes of the attachment was read probably introduced after changes +I deeply regret having so MANY protected methods and variables should have been private instead +Thanks Thejas +Im still in the process of perf +Attaching a simple patch which inserts a call to wait between the failOver and the connectPingThis should make sure the slave has enough time to fail over before the call to connectPing is made if that is indeed the issue here +It would need someone to work up a patch +I also see that SSL is part of the same patch +If you could just add a small comment of why you are doing that it would be great +Ok a bit of bad news the included in camel core is an order of magnitude slower than either commons collections or the in the JDK +Have you tested this somehow? eg running a with the new client pointed at the old cluster and killing a node or two while its going? And vice versa +Attached patch brings the reverted patch back to life in addition it saves the resetedip in system table so when a connection is reopened to the host we will use the resetedip instead of endpoint address +Please review the patch +Mental I know this isnt something we want to just fix on a basis but perhaps its time we discussed a more comprehensive solution to creating interruptible threads that dont blow up things like IO +Id like to implement this before the release but perhaps only after most other things we intend for that release are committed so we break fewer patches +I think it would be a nice feature +Is this part of the JEE specCan you point to an URL +I will submit my solution soon +is not current for release and if it is for me personally I can change on my system. +Thanks for reporting this issueUsing the added method in to detect overflow in the long case is goodThere should be a way to avoid the while loop at the end of the long implWe also need test cases to close this +The Screen Shot Video attached shows fix for The White Stripe effect +to the console +Ive checked it in the last snapshot and it still doesnt work +I forgot to mention hereIntention of this Jira is to create the zk nodes ledgers and ledgersavailable which will not be done automatically +Closing as wont fix. +Fixed w commit Windows build works fine. +pertest and perTest still work but Ive changed it to always which seems consistent with once and also changed none to never. + Username ektak +use svn props svnexecutable to enable execution of scripts +To implement this we will need to change to check for a PU name during its configuration +At the time I think I decided this was a fundamental problem of using readerwriterlocks in this setting +Ill go ahead and check the tests in with an Ignore in +Lowering the severity now that we are skipping the test +Thanks! +you wanna do another iteration to fill TODO +Since apache releases source code this shouldnt have any derived binary artifacts in it +In I dont think we need the target +Thanks! Ill try and get a look at it as soon as i can after the holidays +Maybe it is a hardware issue? I will try to spin up a VM and see if I canreplicate that way. +Tomcat only reports that you have a memory leak related error in your application +That is right this has nothing to do with map joinWhenever a predicate is pushed down it is also retained thereby having identical filtersIs this creating a performance problem? It can definitely be optimizedI totally agree with your proposed solution +Should we do something about that as well +Thanks Arun. +Reporters of these issues please verify the fixes and help close these repro +This is one of my pet peeves and the reason why scanning with block encoding is so much slower and more GC intensive than without +George there are more patches in There may be a test that has an issue after those are applied if so Ill take care of it +Hi Aleksey +fixed in CR branch and on +I think that is because that glibc aio is a user thread implement like our own AIO and we need some performance improvement in this patch in our testing the cpu usage may get less than origin threading AIO and the disk hit response time may get into that is much better than beforeWeijin should post some more data here added the AIO unit tests to the build +Its ok if not every message that appeared to be delivered was deliveredThis is a classic CAP tradeoff does Kafka provide C or A? I propose it continue to focus on A +I tried that once for XML but got stuck at a very early stage as there is no documentation on this available and the XML test framework is not intuitive to be used +This patch causes HDFS to throw a in instead of a generic +Then we have at least these two options Configure s so that external can be retrieved without setting the system property +that totally slipped under my radar didnt get the JIRA email? Hmm +Patch is included in the build and will be in the RC build. +Bug has been marked as a duplicate of this bug +because not every storage support integer version +Sure I will review this +To fix a couple of issues with pc. +Bug has been marked as a duplicate of this bug +Tested +If you look at the current vdbless example the import settings are on the connector binding +ACK this patch fixes the issue I saw +Please double checkMany thanks again +Attached patch adds small versions of input files which are used by default and a mvn exec configuration that uses the original large files +Any other comments? +We currently support only. +What about the fact the currently Deletes and Puts are idempotent? With this change a failed Put or Delete cannot just be redone because the effect might be different +All this leaking happen when running fsck with move +trunk +Thanks for the update +keeps track of resources on each VM host and could easily overload a host that has on it that doesnt know about +Since attribute values are not executable content I cant see why we would continue allowing those values through unfiltered + for the patch +Could you send me the output of svn diff specialpurposepossrcorgofbizposdeviceimplAlso have you run a make clean lately? I checked the patch and it looks like all the code is included +Why bother creating a test branch? Just commit your debugging Andrew to trunkCould do that +All modules were processed +After looking more into it I dont think much can be done about this at this point +the attached patch is for previous version of HADOOP +Personally Ive been using in Eclipse with but I use a trunk build with Ant to resolve updated snapshot artifacts +This would be handy for projects like ours where a large number of alternative compressors are already available although not necessarily possible to contribute +closing since is now released +For the other Clirr reports does this breakage warrant a label or can we file the changes under the fact that these classes are internal to the codec? Should these classes be Javadocd as package private? The changes seem inherent in the implementation of the performance improvements +for border highlighting. +This is the correct order +to trunk +Hopefully a release can be made in the next couple of weeksThere is a recent email thread in the dev archives talking about release stuffOf course all help is welcome +Ill get a wiki page sorted for this in due course +Note that is the patch + +I have fixed this for MYSQL case +Go ahead and commit +Brain was not in gear! +The introspector stuff is in and looks fine but I have some problems to put my test classes into the patch files +Added patch Raise area turned on for event +I have no objections. +Am running against a build root container if it matters at all +I think its ok to change these SQL statements to not use prepared statements and ijs using clauseThe return value of the fn abs is what is being tested here not the fact that if you pass into a prepared statement and compare it to you get false +The exact same patch also applies to the jenkins again +Is there a known workaround for know +Ah makes sense +You should just be able to mvn archetype and just have it work +The maximum value I got was about without disabling Nagles algorithm +Heres a version of the patch that packs booleans +Theres a factory method in Action getInstance that anyone can override toinstantiate a different +Could this be something specific to SolarisI have not run them yet myself +simple patch with simple unit test that wont compile if they are finalupdated javadocs as well +Source files for test project demonstrating this Maven project file sourcesNOTE For this project to work on it needs minor modifications in the dependencies and in replacing +Is already done by Reporte in +allIt looks like I misunderstood the documentation +This is on purpose +Its possible that exception was not hit by logging or there is another exception in the given time +This fixes the build problems with clang +Actually I now realize this has a direct connection to though Im not sure if that fix addresses the JPQL issue discussed here +I am in favor of +orgapachehadoophbasesecurityaccess +It doesnt know if the region is opened yet +Talked with Pete +That at any rate is my take +Theres only one line change in source code +Will monitor the updates for the next couple of days to see if it works +youre rightI think we should add back the configuration which allows it to persist blocks even if HA is not enabled +This wont break abstractions +Attaching a new patch inclusive of previous patch plus implementation of Local Data Storage and unit committed the patch +closing with release +I will try it soon +True that is because the test attempts to user a test repo and downloads the entire bunch of dependencies +Thanks for that suggestion Knut +implements most of the DROP COLUMN behaviors but did not handle the interaction with column privileges +Did you code that all yourself or was it based from some other project etc +Yes noticed this issue in +It was discussed on the list a while ago IIRC +Updated the tools document with the API +Granting apache for the patch. +Thanks for the comments Jun +On unix we could store them in the user dir where normal ssh stores them +Lets change it +There might be another solution see patch +Reverted from at revision +They are already gone in master +Thank you Jonathan +I will take care of it soon +The patch is attached +I actually couldnt care less about how the functionality is implemented it is the functionality itself that is important to me +Updated based on RAT output added licence fileAlso updated readme to point to docs +Its mostly for my own selfishness but Id prefer a jar of the javadocs anyday +fixed patch +Im wondering if this can fit in with solved by? The only thing that I dont see in there is the ability to use the parameters passed to the method +If I select my project right click and select maven Donwload Sources I obtain messages like CET Cant download for artifact CET Error CET Cant download for artifact CET Error CET Cant download for artifact CET Error CET Cant download for artifact CET Error shows my local repository after executing mvn dependencyresolve Dtypejar DclassifiersourcesThen If I select my project right click and select maven Donwload Sources I obtain messages like CET Error CET Error CET Error CET Error sources are now resolved correctly by eclipse. +good point reopened +Until we support methods of doing this its not going to work to have a verified datanodeAs far as a config as mentioned above it would essentially be which is not a good idea for the same reasons as above its a huge configuration mistake waiting to happen and moreover will be unnecessary once a fully pluggable system is in place +thank you for reporting revision Hans please also apply this change in +Patch committed to trunk. +But only the problem is that stale block may not be cleaned until we restart the cluster +All happens in your code its a little bit out of our control especially if you can create the databse structure +I presume building the POM should be analogous right? +This is a patch with tests of tomahawk coreAfter component generation this issue became invalid since all problems detected was solved when the generation was appliedThe idea of this test was to detect the problems. +can you point out what am i missing please +Thanks Glen for pointing this out. +fix logic for detecting changes in the kerberos ticket cache file mtime +committed the addition of to the file with revision Ill look into the maven stuff tomorrow +Per Radhika the document is done. +Gabriels patch committed as r +Is some of the generation methods behave differently in JDK and JDK or this behavior specific to your test cases +The simplest thing to do appears to be inside check preemption when assigning the container and then if it is there the TAKILL eventI am not sure that this is the correct thing to do as a TAKILL has already been issued +q file and data files in the patch +BTW wont run the offending testcase because the changes are in common and testcase is in hdfsThe reviewer should apply the patch locally and run the testcase from trunk +Please verify and close. +reopened just to set the resolution to fixed +You should use Ant ll have to change the API since it sounds like theyve expanded theirs +Nope the default is false +I have now deployed the version of the xsd +Moving all issues to the CR +The Attached SQL produces this think I have enough to reproduce this +This removes the need for which appears to have already been changed from its original implementation +Tucu mentioned offline that the main motivation for making this provided is making sure we dont end up with jdiff in the hadoop classpath need to check that +The test passes consistentlyMind giving us some more detail? +Quince can you try this again with the current latest in SVN? I think both of these problems should be fixed in the latest in SVN as of this last weekend +Fixes some typos in the document Signing Releases +committed. +The rollback process has been corrected so that this issue no longer if you can correctadd to this that would be great. +Committed as r +Please could you help me to find the changesThank to the docs maintenance Im assuming Miroslav is having difficulty locating the changes because this change was made to the EAP docs not the community docs +However we need to ensure this will work with etc. +Seems that this special switch for websphere development is not working anymore + +I guess +This is no longer a problem for me. +This will eventually force the content to be read in to the memory as the whole content needs to be base encoded inorder to return it as a string +Simply a rename to make it clear that this patch is for +Ive tried the patch and it works ok coolThere are only two things to define IMHO the configuration is to be set in it should be in the patch also with a comment I dont understand the need for two different parameters could you please explain this in more detailThank you patch works correctly and helps a lot when you have to use different encodingsThe author of the patch Miyabe Tatsuhiko kindly explained some features in more detail with also a reference to a discussion on the listTherefore IMHO it can be added to CVS +All tests complete succesfullyy after applying this patch +Attachment has been added with description The +I made some adjustments to the patch last time but they were basically cosmeticLooks like I didnt escape much work this time though Ill review and commit shortlyThanks a lot +Ran out of development time. +That patch is no longer valid due to the switch over to Plastic +Performance differences with patch +Made changes to give users the option of outputting to the filesystem or to cassandra +No this still doesnt workIf not please help enrich my understanding of relative URI syntaxWell thats just the problem +to branch. +This specific fix will appear in used by the new proxies. +Applying the change that I described let those test pass as well +Latest HTML file is included for review +Rather it could send requests in one thread and receive response data in anotherFor writes the client typically wants an ack response that data was successfully writtenSends wo response might be useful for status updates where an error would be ignored but in most other cases I think we do care about the response + +This is a great start thanks should definitely look at this for a future release a simplified edit and deployment lifecycle would be great and this is a nice start +We use the RBL to reduce spam +This looks very good to me +removed the remaining code +ToplevelFoo should resolve to ModFoo since Mod has been included into the +Locking transactions might be an option though +It also appears the create does not set the fieldsTried on this thing named foo top projectFolder foopersistence setup provider HIBERNATE container JBOSSAS named Foo string fieldName namescaffold New rows are created when I create new Foos but the name column is always empty in the database +svn ci m Applying the patch from +The bigger issue than bundle not getting killed was that the bundle continued to run for days without any error indicating to user that one of the coordinator jobs never got insertedstarted +It was released with the fix in +Depending on the node to which the user connected they would see different results +efaecceebdbcacef hash in ebbfcdfadfbd hash in master. +as it is and create a new script named +Looking at it is obvious why +Im assuming you do not register any custom provider? In which case it is probably a default provider for writing Strings is used +Verified by the community member +This seems to me like part of one of the most crucial tasks for the next release deprecating and then removing or fixing anything not using the newer Hadoop +Cool +Attached a patch fixing the a noddy utility class used to load the relvant client classes in multiple classloaders and provoke multiple registration attempts +Shouldnt NOACCESS simply be the absence of permission? Why do we need a separate status for it +Reviewed by Keith Philip and Robbie +Any word on a suggested fix for this? Getting ready to deploy +If you do not want to have it unique you could go ahead and remove the database constraint yourself we do not think of any problems that should cause in the internal engine. +I just see that theres another problem the exception also occurs when you have a field but it isnt indexed +Specifically that implementation walked up the tree of nodes all the way to the root document making it most noticeable in deeply nested documents +This one also uses includes as well +Ive searched the jboss docs read the release notes and I see no reference to nor files +I will be in some meetings let me thing about itFrom code perspective looks ok +dev +And workspace can have links to external projects not located as a child of the workspace thought this is custom plugin but maybe we have to create expert version Definitely I will test your proposal properly if it works as the old one +Added test in r +Is it supposed to be private or public +Ive looked a bit more at the streams from on down and made a few changes to writeInt and writeLong +this time after doing a mvn clean +lgtmTrying against +But many are missing +and flexmojos has a LOT of configurations so wouldnt be a nice idea to duplicate itThe IT I posted is just a oversimplified version of the problem I found +That makes more sense then wrapping subprocess in a popen wrapper +Attaching a new patch fixing a minor bug +Sorry for delay on this +So I went ahead and added the test case and changed the implementation of auths to use a so it will always iterate over the keys consistently regardless of insertion order +Should I add ASL v in that +add test case for the use case of filtering while deserializing +Attaching the i have a null check on the jobqueue manager the same error is getting reported and the scheduler fails to start +What youre suggesting only works since the projected value is the same for all union branches +The workaround already seems to be implemented in the adaptor I got but it does not seem to work +revision created a java EE Web project using central and tried the cd functionality +Thanks Steve +I just want to evaluate the code generation and right now the persistence provider doesnt matter to me +Oh Dag its fortunate to see your message +That sounds good to me +Trying to trigger +It adds openejb and into the tomcat server configuration +Will try unit tests on top of it +We need sample code to pursue this further +minor changes added based on review feedback +Uploading patch +This class is not to be included in JR code +Applied the patch into trunk and branches +We are just talking +OK there is a utility to do this +IIRC you had to specify it in a similar fassion with the old annotations as you dont want a parameter to be specifiable from because it would break some best practice such as reproducibility for example. +Wang yep Id be interested in taking a look +Then he modified the destinations clustered attribute to be false +Please verify +If anyone finds any other dependencies that also need to be upgraded from can they please post them here also +Closing +now closing and removed from the Fix Versions field +Right now kit does work as expected +added this to the documentation +The patch fixed findbugs errors +Remove the formerly duplicated code +Committed. +I have upgraded to XDT but I still cant find where to change the x or g compiler options +So it looks like its an issue on the CRM side +Also could you try to define new Sonar installation for testing with defaults no password no host and so on? And run your project with this configuration +So its a fix for your that theres also a problem with this it will split commands incorrectly that have spaces in the path to the executable +Biggest change in Arthurs patch is a split mechanism that groups together regions that are on the one regionservers thats kinda nice and others will want to do that but the map per region strikes me as more general and therefore should be the defaultTo test filled table then ran the rowcounter it uses +This is all good now +If you agree please feel free to do soBTW styles are just the Java bean properties of the skin so you can look at the skins Javadoc to see what styles are supportedThanksGreg +databinderfcst +Im guessing its because patch changes meaning of is +Google plugins are available +Thanks DagCommitted patch a to trunk with revision Closing issue. +In the case of OPTIONS or TRACE methods we erroneously also count these two methods as extensions +Modified srcmodulesextensionxsdsrcmainjavaorggeotoolsgmlbindings Modified srcmodulesextensionxsdsrcmainjavaorggeotoolsgmlbindings Completed At revision + wanna get the release out +use a template based on the taskname with +Ive updated the patch with a fix for which was missing the taglib import for +I have seen this situation with MS AD +Although not that important from a user point of view this problem is Open Group LDAP certification relevant +BertrandI committed run method is different between mapred and mapreduce +if you run builds in the same directory as different users youll run into permission issues deleting files from previous run unless the user running the second time is a superuser +Are we comfortable with closing this issue or is there still research going on with regards to generating the archetype +However anything not Opentaps specific is given freely +Revision +Shouldnt this be closed? Looks like the patch made it to at least +Fixed +part of release. +I will mark as resolved for now and not close it until youve had a chance to confirm the fix on your end +My linkid is viniciusbarrosSorry for taking so long to submit it but just got a free time this weekend college stuffs are keeping me busyUwe I added to my proposal the idea of enabling the user to enter a query that searches for a single numeric value example ageAbout I think this can be decided later in the end its just a class with some format options the user may choose +Moved this to the JBAS project as this would be an AS service running on the +Worth reviewing this as the getTarget method had a TODO indicating its not finished and I didint compeletly understand how these were happening +Thanks Justin +srcorgjbosscacheaop +Sorry about that +IT added +Tested on single node job run. +Looks like a design issue in pdf. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +It was commented for JSF that such a hook should exist and the EG dropped the ball on it and it was not included into the spec +Marking this as wont fix +Wizards validation calls this method to check if the server can work with modulesWe should not call this check every time project name is modified because the result does not depend on name + has been released and surefire pom updated to use it in rev +The fix looks fine. +Configuration issue can be solved by increasing value of property +However would at least be acceptable. +Patch looks good to meAndr Please click on Submit Patch in the left side menu so that Hudson can test your patch against the current trunk +If I recall the remaining issues were easy to fix in the jruby ssl code +If you created this against the latest version which I assume you did can you just attach the whole file? Ill commit it thenThilo +The only difference I can see is that your checkbox is disabled +It would get hard to exchange the JPA implementation of a server +If NOTHING is selected that first time you on the canvas I can consistently reproduce it about of the time +Tested the feature on latest build closing the issue. +Replaced check of namesnull with +bq +It discovers validators in aprivate find method +Matt I dont have access over that to Barry +all issues to the CR +Closing it has been in use for ages anyways and its working all issues that have been in resolved state for more than one month without further comments to closed status +May it be a direct buffer or the factory should create a for use +might be too much +Committed +These should be deprecated +should apply to both and the +Im not sure if my comment is what is preventing it from being committed +go for it +Can you please regenerate the patch using the option and make sure you check the license grant while attaching +How is this related to JBDS? JBDS does not have an abridged category it has a Developer Studio feature that installs this +So your arguments come in as GB and we leave them as so theyll print out and be encoded correctly +Thanks for the patchI added a note to s javadoc explaining how to deal with custom URL schemes +So not sure now it makes the code more complicated than is warranted for this feature +Please reopen this ticket if you disagree. +Fix Fired a second change event with the id of the autocompletecomponent +It would be a separated jira +did we find a way how to test thisi think were not strong enough to beat captchacan i close the issue +Four policy files have been created in the srcetc directory these are the permissions needed to run all of the Remoting facilities +Fixing both in trunk and in +Could you rebase? It no longer applies because of trigger changes to +Adding to identify missing CR issues +Good call that is more clear +What do you think +feel free to close the bug +vms created after overcommit change +And yes we need to move away from snapshot dependencies before we can do looks like an immature change we arent ready for that yet afaics + adds very simple stub server to the test +A better fix is still needed in geotools +Hi Kathey do you happen to have the for the above NPE case +We already have one in trunk +Closing resolved issues for already released versions +Bulk closing stale resolved issues +No idea how we would recover we cant start the so yeah we abandon them +Thanks thats much betterv patch applied at r +tests ran ok except an exception in which looks unrelated +Is the logging redundant Ram? Is the state logged earlier? Are these states of SPLIT and SPLITTING legit here? Its ok to just break out +Jay this is a long going conversation and the reason we need Groovy is things like +This will become obsolete ifwhen DIH is included in the Solr Common jarDIH is a contrib project which means it will always be a separate artifactbq +Unfortunately I still get the same error +Yes I meant adding licenses +thanks +rst files +I added a few comments over on rb Mingjie +The richfaces jars have to be added manually due to the size limit for the upload +Upping priority as it is a regression on the stable ve found out the converter was already in place +The problem is if we have IW writing field A with codec Standard then open a new IW with field A using we get problems with the comparator if those segments are merged though +Moved TTLTIMESTAMPCONSISTENCY to the Attributes class +One thingI did not add methods taking start and end as parameter for the specialized cases +Tested with WebsphereOracle and TomcatUCP tons +so the equals fails +If you do not want to do it just file another Jira issue and assign it to me and Ill do it +I didnt do this from scratch +Pushing the last patch through just committed this. +That is correct +Good +The trunk patch above assumes a couple of fixes made since +This patch is for branch +patch looks good to me +Thanks for reviewing this +Because of I have disabled the country code support for now but that should not make a difference for you anyway +I used +Doc team will use that as a guide. +I think that we need to clarifly what the property does +I have included escaping functionality for URI path segments with commit eebedadefff +Dag for the review +Will look into contributing to the project a patch for this feature aligned with Rossens suggestion. +Is this ticket still required considering the changes being made to the produce and fetch requests? If weve agreed on implementing the changes this ticket is uncessary right +A check can be done to see whether the packageName element value is empty and throw a in case of an empty value for this element +Committed +So I will mark this ticket as Wont fix +Thanks for the patch Mikael +Fixed. +Specifically the slfj and logj dependencies +So the explicit check is so you get the error immediately instead of far into an indexing process +ericCan you post a patch for the proposed fix +I assumed people might copy some or all of those strings from the Hadoop docs and use them and I wanted it to work whether they used jenkins jenkinshash or On the definition by number of elements and desired accuracy I agree that would be nice +Marking as resolved +Please review +We can keep this here as a note +This one could go away instead the Analyzer would store a resuseable reader in the TS cache +is a patch for removing trailing spaces in andIm committing this without review it only deletes trailing spaces +This has been fixed for oracle and stores are not affected by this closing all issues that have been in resolved state for months or more without any feedback or update +Can you update the with these changes as wellThanks +Added javadoc for the getConfiguration method in the interface +Sending srcmainjavaorgapachecommonslangTransmitting file dataCommitted revision +With the new patch I ran the tests and in all frameworks and JME +Without the patch it grew to something so scary I aborted the test for fear of running out of space on my laptop +new architechture file old one was current default template is attached as am assuming that by going this route we can close as wont fix? +Ill try out and see if the issue continues to surface +Simply updating the reference in the seems to take care of the issue see attached patch +This is broken for dropTable not for add. +In Maven it passes +Its the Maybe we should pass a superclass of here +so let me just commit +Resolving as wont fix Will add endRow using new mechanism by Add to +Is there any chance of the patch being applied to the branch +This is the way a project would have to be entered to make it work with the Code Generation is the incorrect project is where I had to update the Seam Gen settings to make Seam Gen Work +I think now the issue is fixed I invoke about requests and it work fine for me +Sorry tested with only setting the submit host and submit patch with only setting comments on the patch +Thanks a lot! +starting applying the patch from Suresh Attanayake in rev +Ill leave it up to Geir to apply the patch and the build file +Had to reverse engineer the patch since it was not applying cleanly +Weve got to have stable unit tests on the branch +Closing issue as it has been released as part of release +Please use this one instead +It was added in +Creating the folder during the build would not solve Christophs problem overall +by compilation error in. +Thanks! +Thanks. +the classpath thing is fixed on windows though there are still some issues with fixed now phew! +The screens are updated. +Patch for which used more fine grained locking to avoid deadlock +Attachment has been added with description the has been added with description the corresponding has been added with description My calling I produce the pdf with my own in the attachement Without the record xmlnscounterext it did work anyhowWhen I call my Class or the in my directory I get this endless loopI thought the whole stuff of the download is +I have checked in all changes to the ECS SVN as r +defer all issues to +The fix already on the CVS +This makes the following changes Make the reading VTI handle continuation lines Adjust the message column in that VTI so that the is not included Fix a typo in the script which runs that VTI Also make the script select all of the columns from that VTI not just the messageID columnCommitted at subversion revision . +Close all issues. +They dont +Ive rebuild the patch and successfully ran checkstyle against my code +I think it would be clearer to keep for improvements to the core binding and restrict this Jira to the stated scope +No concerns about usage at build time +Full Log of Bug with SQL Trace +quick note The reason why failed in jenkins is that the consolidated patch didnt change the binary edit log file used by that test +Integration test added +bq +That seems to be tradeoff that project has taken. +Huh curious +great thanks +IMHO the default behaviour of an application is to treat user input as first class data overruling any default values used in set attributes +Under manual testing this works for me +DOMAIN stratos controller hostname echo ip elb +Derby affects version field to based on reporters feedback +So suggestions and improvements for the test are welcome +for Steves patch great work and you beat me to it +Ideally we could even reuse existing checksum files so that all blocks would not need to be read +The headers now match the actual package locations and class patch vb replaces v +I looked at the transition diagrams and the changes made look fine +I used +Bulk close of old resolved issues. +Heres a first stab +a new patch for Carls comments +Though removing the recursive node type makes it work again! Ill pin down the problem and bring back more clues +How amazing! shows great improvement on both memory and qps +and then could be used to pass back into an Indexer +Many thanks +Anyway this issue is not the place for this discussion +Thanks Rick! Committed patch to documentation trunk at revision +introduced the new failover transport behavior that causes this issue +Committed revision on branch. + +This was fixed by the fix for. +I will resolve issues over the rest with Yang +Since has been tagged Im not going to mess with it for now so its less confusing +the antlr has setup to depend on which using transititive depends on +More fixes +Improved patch which uses fix from for NPE in determination of content length +If you can at least sort out what that exception looks like it would help +I fixed all other issues and all tests including the highlighter pass now +also passes with this patch +So I will close this issue as WONT fix +Added will break if you put any form components into modal window and have the modal window inside a formIf you use form in modal window you always have to put modal window in another form +Sorry for the false alarm +patch looks good +No tests are included because expanded test suites in prior jiras provide code coverage +If the session succeeds then a new reply queue is created and the request sent +That was just back from the time at codehaus +regardless of deep or shallow GC if JVM is low on memory it will do full GC so if you get OOE that means soemthing went terribly wrongmoreover there is not that much data we are talking elements which roughly translates to MB total XML message size if AXIStomcat fails with that it is not good indicator and points that there are some serious memory cachingholding problems IMHO +before Object. +Please mark this as resolved +bq +Thanks Ben and others +I should be able to tackle this Tuesday or Wednesday +The important part is that people be able to do facettrue +Patch applied on and trunk +Hi Apostolos +One small thing at the moment you are loading the config twice each request which is not necessary +Ill have time for this after May th +Can we close this +For internal or private issues while is running we can do private communication +will open bug after i have some good trace way itll make it into any release sure when well consume this from in JBDS +On sandbox branch removed usage of and from in revision and the definition ids are now only used to diff node types replaced with replaced with Direct instantiation of Q should be replaced with Q but those are currently located in a CND package which makes it a bit awkward to use +thanks Ronald! +Thanks Scott and Harsh! +So how about you create a new which takes the list of fields not to load? It can extend by overriding needsField +Ill try it on my Mac OS X Snow Leopard to see if this is OS related +So either you migrate to integer or string +Im generally assuming that the exception responses first line is some kind of readable error message which shouldnt change in the future +To elaborate on this bug a bitA UDDI bindingTemplate does contain a but UDDI does not +So Im just resolving the issue by now +David Blevins too confirmed this for me. +So I guess Im suggesting that bundling it with some test casehopefully of existing test cases that are directly relevant to this feature seemslike a good idea to me. +If we start Oozie then user will not be aware that Oozie actually did not start +The issue was that the spell check component reused the query component to check collations and the query component wasnt originally written with that in mind and it modified the list of filters even though that list of filters had already been usedThis simple patch just makes a copy of any existing filter list before adding new filters from the fq list +Is this Add method the factory method you referenced in your document or is there another mechanism to create components? I think that there should be a single approach to create components +The attached patch replaces all calls to constructors for classes in the class hierarchy or with static factory methods or with where thats possible and more convenient in the client and drda code treesAll the regression tests passed with the patch +Its necessary to check all screencasts and newly made changes in guides +ideally itd be nice to reconcile with the bits but this isnt that high level a feature +VeithenNo unfortunately I pasted this line from my original policy into this one +This issue is still valid for M +still not seen this on any other project +should stay +Spring Security was released before Spring and the two projects are versioned the fact that it is built against Spring should be irrelevant +Interestingly load is going to the same code path as but not save +it is now a subtask of which is scheduled for +I completed the missing files but this bundle really needs some reworkI am not sure why we need to wrap it like it is done now +Looking forward to it +Edell +decdbfecdfbefdaaececdabfdcafd +ben can is also marked for then +The issues are now closed. +Marking as to be fixed in CP as this is a regression relative to title to reflect impact of the fix into JBESBGACP branch +Verified against SVN using Carlins example +Would it be helpful for YARN to supply a public API that reads the files for you +Thanks Tom +Marking this PA since went in +What do you mean Dan may I add this to PROD tooGo for itbq +Copying paramsin the constructor would leave me without a chance to avoid additionalobject creation even though the one I pass in is exactly the one I wantto be used by the method +Ok that was my fault +Have it been created +Yeah I should make the test fail on these but its kinda hard to put your fingers in there +Addressed Mohammads comments +Also this is not with the referenced example code +in the attached file included are the following files +Its work could mostly be replaced with a method in SS that calls into the static methods currently in Bootstrapper +Verified that the queue is not protected and the read lock around that code is not needed once the code is switched to aNice catch thanks! +I have committed this +Fixed leading zeros problem +I believe contains the fix +The runtime exception should in some way reiterate the deploy time warning +Oops +My plan is to filter out any and dont provide any input values for them +Downgraded to WSSJ since has issues +Grid can you please either +Thus the groupIds groupIds have to be changed along with version the BOM has been removed from WFK maven repository +Bulk closing stale resolved issues +Ive made a patch for this bugIm not sure whether it is correct if someone could look at it +Yes distinguishing between the test levels will be very helpful to anyone looking to adopt DNA or replace their existing repository +What about existing methods like should and should? Change those tooUgh +Dominik can you add your deployment to this issue in case it is related in some way +r +Sample schema documents illustrating the crashI havent attempted to reduce them to a minimum set of data to reproduce the problem +Verified on Tools +As for looking for the number youre using the Whitespace tokenizer and the numbers in the sample text arent delimited by whitespace so you need to refine your analysis chain to accommodate + +Hi LewisSince the proposal is not accepted Im using my summer time to work on my undergrad thesis +I dont see how that screenshot is different from what we have todayThe previous screenshot was taken from STS + you! +The adapter template will be fixed by +is a patch to address this issue +I dont see any obvious issue from your log +Ok makes sensethanks for the quick reply +When this occurs manual intervention is needed +In one case the corrupt block ended up in the middle of a file +Patch for would suggest starting with a patch for TRUNK which can be verified by Hadoop QAOnce code review passes you can backport to branch +we havent found any way to reproduce this +Updated patch with tabs removed +committed the fix +finds with RSA from +I want javadoc to handle my project and all its dependencies as if it were one large multimodule project and generate a combined javadoc based on all the sources +The number of failurestask is configurable but not dynamic. +I just committed this + +The performance gains are impressive under an ideal read heavy load fair lock k ops unfair lock k ops unfair lock unbuffered appender k opssecA single thread consuming log messages from a queue populated by the rpc handlers is sufficient to improve performance +If that method wereprotected instead of private I could make very small adjustments withsmall amounts of codeIf the two factories were one class then I could just override thefunction that creates the server engine itself +New keywords revHere is the patch forIll attach the others remaining patch ASAPThanksAshish Ashishyour patch is in rev +Thanks Raghu! +Committed this to trunk and +patch that makes sure we use canonical amis in but not the testing or daily ones +Oh I see I think im nervous about test tooMaybe we can comment it out and just mention its more of a benchmarkThe problem could be that the test is +I think Ill use it especially the set of property editors for Scala collections +I think Ill take a stab at it. +Please do not use JIRA for users questions there is a special forum at Use or to fix the problem +Issue is fixed +Yes thats a good idea +Anyway when you say Import the projects what sort of import do you mean? Import as a maven project? Import as a filesystem? At what level do you do the import. +The patch looks like its doing the right thing so unless you have a clue as to whats up I can have a go +Will commit to both trunk and branch as it is +Hi I am using with Spring MVC +Im not sure whether its solution or workaroundI think there should be method to clear or set null value in Manager. +Deploying another file is reasonable when the command is going to be a multiline script +I tried the WAR attached to and it works fine with Mojarra which is the version we are going with for AS Final +Jaroslav there seem to be other issues with web services in EAP as well +Not relevant anymore for me. +Testing now with attached patch +This would up the hdfs load on startup +really shouldnt span versions if it can be helped and there havent been any commits to this issue explicitly in a long time +Forget to mention JRE +Thanks +edit just created for the import +Patch which takes care of second and third bullet points of this jira +It seems unlikely that this is the case but we ought to run the grep all the applied +Resolving as invalid +Ive thought about this a bit since almost every time a ship a package they ask about this but strangely I havent really received a request about this +Version +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen + looks goodwill merge if the tests +In that case I submit mine as an alternate implimentation +I hate to see this patch languishing here +The approach is going to be JVM and even GC implementation specific right? For instance right now I have CMS old gen CMS perm gen code cache par eden space and par survivor space memory pools with the JDK CMS GC +You can them from the log +Isnt some documentation the real solution here? Maybe a special section for restricted envs like applets in the wikiJeff you can also sign your applet which will give your applet an unrestricted access to the system and thus eliminate the issue +Slava please look at it +I think that it is fine to have them in different base directories if it becomes a problem in practice because tasks are trying to share data with each other then it can be fixed then +Some property of this particular svg file dramatically cripples the performance of batik +can you elaborate on ifhow will work with your embedderThe test case attached to doesnt check that +This option is really a must since it is impossible to compile a reasonable amount of sources using a forked compiler +and a client to try to connect to +Test cases included +The patch excludes some multiquery cases where more information is needed to correlate and determine the files to consolidate +Ill leave it open to track the was just released. +Please check that it was applied as expected +The problem is apparent on my machine with JDK so basically this bug shows up regardless of JDK versionIve looked at the source for and theres a custom implementation of a map being used which is a bit scarry +uploading v which should include them +mrepository and it works ok. +Thanks Chris! Resolving as fixed. +Then my comment is just that the hash code can be defined such that it doesnt need elements in order to be computed so shouldnt +looking into it now +I just did a quick review +Here is a sample of various driver versions aTIMESTAMP aTIMESTAMP The switch from in to in seems to be the problem here +I had to remove some path references for security reasons sorry +findbugs was see jira +Duplicate of of +Done. +Waiting on two other issues in Plugman to be resolved first before that gets pushedIn the meantime if you are keen on getting this functionality ASAP you can pull the latest cli code from master and run from source +To use the patch parser should be recompiled using +CheersAshwin +The bug was previously fixed in CVS under bug report +for the patch +Updated test cases +Perhaps we should take and associated classes and create two seperate modules for hadoop and hadoop +If this is not stricly needed for JB we can move it to need for a separate mbean service. +I dont see any issues with extension tag in examples from +Yep I think were good to go! Thanks for your help with this +however i wouldnt take the documentation too far +Eric +Closing stale resolved issues +Its not clear from the test output what the problem is so I am going to revert +Please confirm if this is correct or notIn my opinion ZIP installation needs to be and not touch system configuration files +Not sure what you mean these methods are available in the class and are implemented +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Updated patch to reflect recent commits and fixed to use +To be reopened if any issues +Sigh +Is their someway to get a git repo for just drivers? like but? This is causing major pain +Updating the fix versions. +Snjezana any specific reason this is using mysql and not hsqldb +Hello Robert many thanks for your help and sorry for posting this bogus! +Committed with nits above fixed thanks Pavel +What version is that? Is it released to central +Attached is a patch which removed unused variablesimports + for the to trunk. +Need to think about where and how to set that property for a single test +The former does not run the large tests if the small or medium failed +Now this isnt of course but would be an indicator +The requires should be at the top of the script not in the thread and as mentioned on the mailing list you also need to require jrubypathhelper +trying hudson again +Whitespace isnt allowed in any of the aboveYoull need to make sure the regexes selected are appropriate for each field and take note on what happens to existing fields that dont match +re Youre right on +verified +Hi HowardThanks for the quick response +The OOM could have occured any number of placed in BTM and caused similar havoc +Forgot to clearify that the above error shows when in Admin and trying to view the Home tab +Thus if the namenode could not find the token during the transition it can wrap the exception in the and ask the client to retry +so xxx part is not the same in hdfsxxx +That seems a valid bug as well +My run was less than successful +basic has proven to be insufficient for getting our internal test passduring the last oakathon we agreed that we have to add a proper implementation for setup and provide a hook for other cases as outlined by michael above +What did you to for this to happen? Uwe the problem happened to Mark +the solution could be like in this thing is that the format of the and entries in the manifest are not the same as in the +Attaching the release note for this patches submitted and committed and release note issue so that I can scrub the release note +Investigate for if they should be in SOA Van will set the Fixed Version +As discussed offline there was a problem in the way the incremental backup comparison was done against the of the backup store +here is the latest patch up to trunk changes +Therefore I do not think the issue should be closed +Attaching which contains an integration test reproducing the problem +I can confirm this exists on on my Galaxy Nexus running +manual refresh a new table doesnt show up even after more than hourOh sorry I misunderstood what you were asking for +cant be backported to and earlier because it had required system level changes which cant be backported to Derby products +Thanks to Nicolas who pointed me to this resolved issue as it seems I did not receive any email. +in common seems to be a more generic location +looks good +I took a look at and I believe this is the issue I experienced +Both and are in this attachment +Is keeping author information in the Apache Way +True only a half of problem is fixed in previous patchNew weak references detected after processing finalizible objects cannot be safely ignored +Think you could make a failing test case andor patch to show the behaviour +Postponed to +Not yet after we make sure that its working properly and efficiently well be sending a patch to nutch +Thanks for implementing this and adding the xml support +This is a very complicated problem not a trivial feature so manual tuning right now is the best strategy until someone wants to dedicate enough time resources to conquer that problem +Rich as the issue reporter please confirm that this is fixed to your satisfaction in the release then we can close this issue +Agree what stack said +Why would you use group by for a guaranteed relationship +User was always null and start time was always +This patch makes minor changes to Dhrubas patch +I used Pauls workaround but now it should be fixed +It passes tests on is a updated version of the test to work with trunk +Attaching +Not sure if it would be best to enable or disable by default +Patch for optimizaitons committed +Further frequent cleaning and setting up may slow down the build process drasticallyWDYT +get token on bootstrap that gives us half of the keys from the most node +Reopening as theres no documentation for this feature AFAIK +Interestingly the task has a defaultConf and a final conf +Addressed comments from Sid and Ravi +Fails the tests though +Here is the part of the patch from +once thats addressedI do also find it a little odd that could conceivably run dfsadminfsck and arguably the balancer as well +Hello! How can I use the tool to run multiple datanodes in one nodeThx Kate +Changing these fields will not change the ports the server itself listens on +Yes Leon I understand your problem + +Im asking on the main list for some direction +Patch attached for looks good to me +Theyre all different from the ones I got locally +Hihere are Ankits patch with a few fixed the checkout process works now +OK I take into account your remark about testsI have new version of classes and Patch but I have problem to create patch from old version to current version of file I can create patch only for new fileI think problem with testing class Instrument bind with class Patch because I use it in test and class Patch now is only stub +The same operation in JSF project creation wizard already was processed that way so I hope that now the lockup problem is solved +Then concurrently a thread could be started for each regionserver containing the keys and the regionserver could find all the rows in one method call +Wont it apply to all partitions where merge is being performed +Ive been pulling my hair out over this +However this test class is not intended to test storing of float or double values instead it should test the! operator in JDOQL queries +I have started working at this issue +Updated to also match ctemp like paths which are also accepted on there are objections I plan to commit this revision +Adding patch please review +Can you please regenerate it +Version has been releaseClosing all related resolved issues. +Let me know if you can get to it now +that is because you never enter a the event on the process definitino instead +Now works side effect of Hajo refactor that makes the decoration call into all issues that have been in resolved state for more than one month without further comments to closed status +As far as I know the ultimate use of sqoop would be on a linux box with Hadoop on Linux +I will take a look as soon as I get back from vacation +Commited to +Resolved with +When I click Apply then Ok on this dialog the retrieve pattern is reverted and modifications arent kept +The complete body of the missing email is attached +The v patch still has the drop trigger code in +Are you interested in implementing it +I see cygdrive +If youre reallyconcerned wouldnt it be possible to wrap the Method lookup in a privilegedblock as well? +LarsDo you want to assign this issue to yourself +dear hiYou mean that I use other versions have a try +Im using Debian unstable but it seems this is the same package as stable +This patch has broken the hudson build Ill be submitting a patch to fix this shortly +Patch doesnt apply on trunk however +Please let me know how your benchmarking goes +Is this a problem on staging too? If so is there a way we can get the next release rolled out to staging sooner +bulk defer of unresolved bugs to +Thank you! I could easily confirm the reproduction of the errors +For being or negative value other than uses the earlier logic of determining a pageSize capped by based on size +Theres still a bit of cleanup to do +mrepositoryUserstucu +Ill do the test and if positive will commit +Not sure if this is an error on EMFs part but to fix I had to add a content iterator after the load process that ensures all contents has a UUID +There it reads the port file from the properties and sends the shutdown command +In order to fix this problem it would be necessary to completely rewrite our own new class wizard implementation +To prevent deleted data from appearing againDoes this mean that right now we have a data correctness problem in and trunk +it doesnt look like its related to jaakkos. +bulk defer of featurestaskswishes from to +The particular case not solved today that Im running into is aSolr server that simply takes too long and slows down the entiredistributed query +Added unit tests. +May I suggest or something like that +Removed the target +Note with Connection Pools this is can be a real problem +Thanks for making the changes Jeff +DoneThanks +I should have broken it down to an example first +It just means changing more libraries to be compatible +A customer of our partner Red Pill is also having this issue +can only be on this list +Do you think you will be able to look at why my target roo project is not right? The redirects to the signin controller but that gives the Requested Resource Not Found +New error checking for existing partitions and changes to the actual DDL for field schemas and such? could you update the jira description with everything the patch is doing or separate them into different +Closing it as duplicate. +It is closed as wont fix +Looking at this now the problem really seems like its in somewhere +Thank you Chris. +Id support the catch clause too +That makes sense I did not pay close attention to the stacktrace before +scope handlers improve use of type handlers vs artifact factory any artifact stuff in should be called instead of vs treatment of as artifacts vs best to look at in relation to the dependency mediation features +This patch seems to do two things +r +HDFS does not have a version that exact matches + +CMF will be used to access keying material for column level encryption. +AnoopCan you explain why you feel client does not know that the operation is successful in May be you had some scenario in mindIs it like before the expiry time the client retries without knowing that the previous operation was successful? The usage of nonce should solve the problem right +Modified both and to return false upon JS submit +Fix the failing unit test it had a match against the error message text which needed to be updated +Thanks Doug +The fix I made was too much +Looked briefly at the patch I think this is a move in the right direction +I just committed this to trunk and +WillemWell spotted I will add the strategy method so end users can customize this behaviorI was wondering if the current behavior to return the exception asMESSAGESTACKTRACEMaybe should just be a plainSTACKTRACE +I think it should check whether or not there is any existing graphs related to the server before deleting it +added drafts +But the maintainer is already working on integrating the code so any instructions will be obsolete in a few days +I am already using the work around provided in +Tentatively set for Gadgets section of the Reference Guide has been removedThese issues have been corrected however in case it is reinstated in a future versionThanks Marek Thomas +I think this makes a lot of sense and it works well +This looks to be working guys +Thanks for this Andrei +Moved dependency of to test compilation dependency on by removing. +And do agree that this could be confusing for people trying to upgradeI will try to get in touch with Richard Kettelerij who has denoted this component +The proposal above to turn on and off features on the write path according to flags is a good one a generic and extensible mechanism that can support many different policies +The main problem with the patch is that the performanceresource consumption is unacceptable + is on EAP Affects and Fix in Version +Few classes seems already committed +Kian Win OngYou can mail me your modifications with a patch to my email address +a bit the previous version because is not serializablePlease review and comment if there is something wrong. +The patch cant be used without some manual fixups +Added require for the attachfunction the minimal example with MRI and got no error +Fixed into revtrunk is the advanced search in category after the correctionThanksMarco +Awesome +As discussed the other day the information added to the code generator properties is excellent in terms of content and amount but still I think that most of the information should go to a new page +There will need to be some memory management cleanup +Can you reproduce this in? What is the actual procedure invocation and procedure metadata that you are calling? Did the metadta come from the designer or from a dynamic scenario +Key performance gain comes with async set ups where directing requests to main owner allows for replication to be async and only go to a different owner upon failover or exceptional conditional making the chances of not finding information slimmer +Patch for Maven doesnt allow more than one dep with a different type but the same id +Is SOOOO small +Thank you Xuefu! Yes all the unit tests pass now + +Nicolas the above should be put on the dev list +iPojo is not reflection based its also using ASM +See issue in the deploy plugin +Id be interested if anyone knew of a way to do that +TedI feel this JIRA is needed for as it is related to the recovery time +But this should be discussed on the developer list first +Many of the changes were nudging the authentication scheme towards a pluggable design youve even taken advantage of some of those changesThe new hadoop SASL related interfaces shouldnt be necessary +I previously argued against such parameters at +This change is not required +Yeah I just went through the callers of that pass in Client protocols and they are all testing append or retry in the client +This is a new feature so belongs to rightJulian please attach your code as a patch! At least when trunk is It seems very interesting +Patch committed in rev +Please build +Why not? This require us to do releases of jboss tools even when we dont have too just because seam did a new release! It should just be a warningIf we had a runtime AND runtime support then I would be ok with you making it an error to try and use a seam with your runtime forcing the user to use +Im attaching the new one +day +thanks Jonathan +bq +So its still alogical line + ran just fine I have enabled it +additional info all recovers fine if the database once got successfully accessed time and then the db dies and will be restarted +I would prefer keeping it in over creating new factories for command +Excluding an undefined tag has been defined to be the same as excluding nothing hence the facet counts will be unchanged +Attaching my version of just in case its needed +its silly to be in the memory contrib instead of wordnet where it belongs +Hi GarvinThanks for your it works I will ignore the thing is on my list of things to do +Successfully tested on a secure clusterTrunk have moved files around caused dependency issues that I need to resolve +This issue can be closed as Wont Fix + covers only a small part of this larger issue and is superseded by this issue +to throw exception as earlier +Propose to close in a few days +Hrm not off the top of my head but Im guessing no +Go ahead and work on it in the meantime +x +Closes issue after bundle release. +Handy for batches of many thousands of documents +However what would be your preferred approach to implementing this functionality? I was thinking that the annotation would be extended with typeRequestResponseBoth or similiar +Patch applied and committed +However a new datasource was added which is almost identical to the now depreciated +And request was merged. +The change was requested by Karthik + +bq +typo No they are still see at least empty files that should have never been committed in this form +This requirement comes from end users who wants a one click print buttonIts useful when some paper workflow still existexample in manufacturing when order approval should print a production run in workshop when administrative sent a lot of paper mail for patch because of replacement by +Looks like a relatively simple changes to use cp along with minimal works ok with simple sample EBCDIC files I have access to +patch applied thanks! +this is the latest patch which throws Runtimeexcpetions with interrupted exceptions and gets rid of the while true loops +Hi KarlChristianI have developed a custom output connector which connect to a repository AND adds messages to Rabbit +wow that is pretty intense for what appears to be a simple mistake in the builder +However the patch does not apply cleanly to the latest codeline and needs to be updated +Release note draftThis update is of interest only to jBPM installations employing multiple business calendar property files AND repeating timers +Will commit in a day +java files in the patch file due to insufficient privileges +Reopen to avoid forget to remove attribute when we move to version of Dynamicproxy. +Added isDouble +The main instance has been migrated +It line functions not worth the dependency +It seems that you have a different JNDI name between the one defined in the WSN components and in the main fileIm trying by updating the JNDI entries +it is impossible to have a verifier with auth flavor AUTHSYS +Im still getting a duplicated jaxb will reopen and investigate more +Having said that i do not have any idea why using a key generator for the refered entity influences creation of the refering entity +I believe what you need is something like this +should have equalshashCode etc +deleting while running works fine and this seems to be the intended pushed to master with commit hash abafefcfafdefb +Well one is for sure Im not gonna eat such error just to make up for other framework bad behavior +Could someone else add them? I havent changed this as the change that I used for integers does not have the desired effect for doubles +I set the path to manageTo test I set this element on available on lookupContentNicolas +Usually its the same one every time its just a code convenience to not hard wire the producer with a specific Destination +I have been using a little it didnt encounter any provided for the skip paramter inCedric +They wont have admin rights of course. +Ill get this fixed ASAP +Its a pretty complex task +Resolved for EWP. +Jon Should I push the generated code in githubve updated the patch to fix a few things due to intervening patches +It works for html tags +Pending char limitjavac warning seems not related to this precommit build failed due to ubuntu node +Attached the patch with unit Arun +Thanks for making rat pickier + +btw thanks Steven for testing +Will provide patch +to verify we keep backward compatibility +Actually copy it to project lib +I recently raised an issue again a testcase failure just because the Puts got the same timestamp and the test was failing randomlyBased on the discussion over here i can provide a patch for that +I dont understand this is this fulfilling what I asked above +Yes KVM +Chip thanksI just realized that I forgot to publish it at last Friday and the patch has been staying Ive closed it in the review board +I understand the practical objective here but I dont think the exception class is the right place to address it +Im not sure how works but since the bug was about the files being deleted too soon Im inclined to think its effective +Are you still interested in a release of this project +Good to see this! We can make this as a part of contrib +CMS has no internal tracing you are free to instrument the code using whichever tracing tool your particular organization uses +HmmI can not reproduce it + +OK that should be easy +The work was already committed directly to trunk +Raghu is right +Thanks +txs Joe! Somehow my svn client didnt show as dirty +For the complex type the object of JSON could be mapped into Tuple of Pig and the array of JSON could be mapped into of PigAnd I also plan to write a storer to store data in JSON formatAny thoughtThanksChao +Thanks Uwe +I think this is a hudson display bug +Moving out on it will provide a patch soon +but it would be trivial to treat it as one and get all the features that come with tables as a result +Thanks for the patch +Add parsing of command line suggestions by Nicholas +Then I defined a bean with my subclass and set up this bean as in the +the minimized test case will come tomorrow since here is near middle am still using ant no available now +Ok I fixed this by allowing you to pass false to becomejava! to turn off child classloader +Uploaded a new patch addresses Bikas commentsbq +has a dependency on and ui +Furthermore the removal of both of these does not clean up properly because the deletion that takes place does not notice the intrinsiclink rows corresponding to the documents for the same reasonIn other words what we may have is the result of either a transaction isolation bug in or improper usage of transaction isolation in the driver andor the hopcount code +First release of three unit tests +Ill update Roo in Action to reflect this in chapter +I no longer work with Sling project that caused that error though. +In addition I have also added a master file for the phaseTester test in javatestingorgapachederbyTestingfunctionTestsmasterOutput of the manual run of this test can be diffed against the master that I am checking in +Your patch didnt change that one though +Patrick I agree with that idea +Multiple can work when sometimes unified dont in such case we need multiple to be published while we are working on fixing repo URL vs artifact I think that forcing one or the other to change for better alignment reduce readability so I would keep it as it is now +Tools and launcher sources were moved from classlib to jdktools component and jpda module was added there +Thanks for the explanation and changes Narayanan +Wait +looking at it in and added to jboss on +But there is a problem in the new code +Have you had a look at the maven archetypes we have? I would think follow something similar +junit was fixed +Ive just created a confluence user bamboo +Go ahead and do svn mvs for these tests and commit +Fixed for AS EAP Affects and Fix in Version +It should however be easier now to add this support if requiredPatch ready for review +It would really help if you could show me the client code you want to write and the output you would prefer to include the complete usual stack trace starting with the stack trace for the and working back from there +Ill run the regression tests and commit +Any HA action initiated when a KVM host is not offline could result in problemsIf we know that a KVM host has gone offline and is not on network we can allow it to be manually disconnected +The algorithm is simply though +A specialized proxy would just forward the mutations in parallel and report back failures +The JAR appears to be corrupt in addition you might want a groupId of tanukisoft and an artifactId of wrapper +Could you give it a try? have confirmed that this is fixed +If the bug doesnt occur within seconds the experiment is repeated +Sure dont i need a key to login +Fixes a small cosmetic bug in the previous patch +Does this issue need to be documented as a Known Issue in EAP? Thanks +Can you update the patch to include it? Thanks +Tests are passing in hudson. +Ported from trunk to branch at subversion revision +Also I have rarely seen this enabled in prod +Ok as long as the credentials can never appear in the log im fine. +If we really need to protect ourselves down at that level we should do a or key crc when it comes out of the map and keep it with the record all the way through the framework until it hits the reduce +and committed! +categorized in seam to be fixed in +Thanks for verifying the fix. +Did you run this with delimited identifier mode turned on +New patch fixing previous nocommits downgrading to +Changing fix for to an unreleased version +I will proceed as I mentioned above soon +Thanks Brian +I meant Everything is working well before integrating Trinidad into the project +Currently regions are an implementation detail +Closing all ResolvedFixed issues lacking a Fixed Version +table to the new format +In the capabilities transformer WMS +Another small tweak should delete output directory if empty +The authenticationauthorization successesfailures are captured +Could you provide any details about your environment? What version of are you using? Did you implement your own user manager or made any changes to the codebase +LarsI think patch wont make some problem but I dont think its the reason of the failed testsIf all failed test were caused by maybe theres no problem in current logicCorrect me if wrong +Per discussion with Ashutosh opening new jira for a feature weve wanted a couple of times now to be able to inject overrides for and once thats in well need to change our tests to test all modes. +Discussed this with Mark +Pepijn this patch breaks specs forIm going to try to integrate the change you describe in the description of the bug though +I know its probably asking too much that a JCP spec actually be detailed enough to get into details like this but Ill have to go back and follow whatever guidance they provide in this regard +Added on revision +If youre running mappers on significantly fewer machines you may just be wasting time as mappers are competing for the same diskAlso I find it important to increase to or more and increase quite a bit to perhaps half the workers heap +Verified with +Issue resolved when we put the cluster on the same host and added another nic where cache communications was error free +So are you still on Spring? Im considering a backport of this stuff to Spring +Committed revision . +If this should be reconsidered please add the label +Aftermessage has finished there will be bytes in the buffer or another whole message in the bufferso the second invocation of size will behave correctly +Great! Seems that its exactly what we expected from this issue +Updated to allow as peer +Ive put this in production monday after i had it baked for two weeks orso in a test environment +Currently implemented for +That is there may be a simpler fix +I have added support for this +Im moving to as a reminder to check this again in the environment used for that release +This demo application has a named something other than app +Regardless on lock patch just for the gains +But how many people have encountered this problem? Waiting for a aspectj release build for would be fine by me +Marking as a blocker and moving into and +new patch with based onconfirmed to be working with IE and FF +Patch b committed as revision . +Is a fix for this bug available for community edition GA or must I migrate to or above to get the fix? + +Some additional changes made in rev . +Would you mind checking that you are able to connect to your database from all nodes in your cluster? Its pretty easy task with the but Im not sure how to do it for Microsoft SQL Server and Im not sure that Microsoft is shipping any linux compatible client +Triaged for +Ive modified it in the ASF repository +jar be fixed in current trunk. +This should not be much code to write +Directly accessing is not a best practice and I dont think we should change a bunch of tags to support it +Hence the method is removedbq +Both indicated through exit codes +Source as this is being ported to HEAD and some cleanup done on +Im pretty sure this is the same as the other issues and has been fixed +The core hooks are simply final because of the template method pattern to enforce the specialized taken on reuse of the code there however thats going to be hard since the primary point of the current test classes are protected fields and methods +before doing the rest +I didnt realize the are cached in theSo for v for from meWe should have separate jiras for and +Also modified the messages so that they are more descriptive and relevant to the failure. +Done with manifestEntries in the +If it finds a descrepancy between the number of bytes that should be in the file and the actual size it throws an which will effectively stop the job +Attaching a screenshot image +Actually I do not know where the site docs reside +thanks +Lets just put them on for now +For example when a is passed to a service method. +Yunkai Lets land this asap and Ill respin a release +Looks fine but could you add some comments so I understand the rationale for what you are doing +Model Project Set that contains a VDB which can be used to demonstrate visible default options +A simpler way to handle this would be to let write binary already exists for the same. +I could simply fix the templates but if youre converting many static HTML templates to Tapestry this error makes life difficult because the exact line number of the error isnt always reported +Changed version from to +I have a much better idea in mind will post example shortly +Any update on where we stand with getting this issue addressed +trivial patch adding public access to currently package private Steve Ill commit shortly! +Since we cannot update this issue maybe we should close it +Thanks +But Quick filter creation alone works ll retest this problem for latest during planned upgrade and raise upstream issues if problem will persist +I need a to check this in +Ive verified mvn clean install +Hello folks +Paul I think we corresponded offline about this bug +Attached a patch that includes a test illustrating the behavior +Ill start looking at your earlier patch just in case to see if youve missed anything +Then if some users try to install our release on other machines will it be just not work +redundant as of +what is resolved? where do I find the updated hibernate jar that supports Left Outer Join on Unmapped Tables +because there is file lease in the namenodeonly one client can own the lease +as all subtasks were finished close it + +For instance if you know the first tokens of your docs are XMore to follow +even when creating fields as unions of their type and nulls it doesnt work +I will create a forum thread with details and post a link back to this JIRA +checked in +Ive attached the patch +Meanwhile let me test your patch +field number class +This is done +The version shouldnt be hard coded in scriptspatch so that upgrading to new version requires just change of version in main pom +And after checking it also fails with the latest X code +Yes Im not expecting this to improve performance its more for completeness I have an other need +Thanks for the report and checking back +The attached file is a testcase of the problem illuminated in the description +Uploaded a patch to address this on single node cluster all the metrics can show up on metrics dump the test case +raised the priority as the related console issue has a dependency on this and its also scheduled for this release +Much better alternative then using the fusesource repo and it will benefit others as well +Trust me people will forget to do a separate build for rat +Seems fine to me +Not that it isnt possible +Hibernate doesnt bother validating String lengths and I am very on having it do so +Ill update the docs with some of the detail m not aware of any problem with VARCHAR in general and Unicode +As far as I know there is no such functionality. +Patch that establishes the Queries module and moves the classes +Files must be ASCII encoded you can use the nativeascii tool embedded with the Sun JDK some properties editor can also do that automatically +Do I need to again? just committed this +Please apply this patch if you are seeing random unit test failures before investigating further +I corrected it in my xml though will be waiting for Mark to give the full version of the xml file +config element is now optional again +svn looks like we no longer handle the case sensitivity but the rest of what I said was validProposed patch +try catches all now +Lincoln what are all those Setup XYZ? Actual wizards exposed in eclipse or how +Try this one +Codec is free to record whatever attributes it wants on this FI +more demos to go as of this morning +That is why this patch got a for Release Audit +It has been tested but please review +in particular our bug tracking system +Once well have the separation of the configuration and the EL well add this support +From Yep +If you dont set those system properties and still use long queue names then there is no warning or error alerting those queues will disappear after broker restart +Attaching +We could have the list generated from the file in the case where there are no attributes presents but this seems like a bandaid +I went through this task deploying apps on EAP and found following As for simpleapp For both mobile and desktop I created As for kitchensink Created +Users will not see this as here run this script against your source tree +on Int and Long is Native and the classes are so the C backend never sees the use of in the call to the static method in an AST so the referencedType logic doesnt kick inFix is to manually do the proper inclusion in +Let me know if you continue to have trouble reproducing the issue and Ill cook up a simple test case outside of Infinispan to demonstrate +Once that implementation is removed it worked fineThanks. + than remove since it referenced fix it +for the committed the patch to trunk andThank you Brandon! +on patch +We still need to explain how to build and we could link to the page about building plugins +With +It is public to provide an easy way of testing a Reducer alone with more complex inputs +rev builds bundles instead of plain jars and makes a few osgi package imports optional +Improved with r +bq +it generates error to scratchxenvgumtreereleasesserversquokkaconfiguration and it is attached +Verified in JBDS +I reported a defect against some of the EAP docs +bq +Also the patches are submitted by Ravi but the JIRA is assigned to Hairong +I tried a couple more times with fresh installs of Yoxos distros along with the manual installs of the XUL jars +But I manually tested this patch with the following tests Ran a sleep job with a huge sleep time +xargs L iFILE gpg clearsign +Both can be done with importer options when using dynamic vdbs +I reopen it for reasons I missed to put in LICENSE and NOTICE Its an issue for Birt see at revision +this is a cayenne tutorial so it has to be about Cayenne not Tapestry +Fixed ordering +Thanks UmaAlso thanks Daryn for reviewing it +Hi Ted resolving this JIRA as well with the following explanation +This is easily done by creating a Filter and setting it on the in the layers should be an extension point +Unused imports should be removed when updating the file for some other reason like fixing a bug or adding a new feature +Attaching patch and gzip of corresponding htmlpdf output for the Developers Guide +It might even be easier to review in that case +Seriously thanks for all your hard work. +The +Where is the error message coming from +Yeah I can change the default to use a dot I think that is prolly more standard +committed fix again +Same as above from second contributor for legal +Both have been implemented in trunk alreadyautodetection in svn revision string constants in svn revision the constant could better be public are public as of svn revision +If this is a bug it is a bug in Grails not the plugin +I am going ahead to close this one and track only the issue from. +Would anyone want to try this +Assume closed as resolved and released +Now I see and showing up with a google search for siteBut not +Cant speak for CDH but should never be public closing as Invalid. +nope not yet +WorkaroundThanks for posting this +step cluster group mvnsomegroupartifactidstep cluster groupgrep i artifactidartifact artifact +Thank you for the review +Thanks for the clarification Paolo +Still not converted to but now as an own module with fragmentmapperI have NOT modified + I would like this API exposes this. +I agree +I found the bug while doing a code review and wasnt aware that this bug had already been filed +Bumping to blocker for release +Do you have any other suggestions to overcome that problem +Its not clear to me that these different algorithms operate on logically the same input though I imagine the command line options and such could be more standardized +This is caused by the order of the static initialisers within +when spring wants to create a bean through the Bean annotated method this error is thrown +Judy assigning to you for assesment +Community feedback appreciated +Ive tested it in firefox and curl and I get the expected resultsNot being super expert at either the codebase or Erlang I think this is sort of a naive approach and would appreciate some feedbackIt merely checks if the client is sending texthtml as part of the Accept header and like it doesnt respect qvalues so if the client can do both html and json it will get htmlThanks +Attaching +new patch addressed latest comments +Uses antlr +Test case should go inside srctest +huh +Ed I changed the Stack to a Deque with the corresponding change of method names +It does not fix it +So I dont think its related and this patch is good to goBut maybe an interested committer can verify +It takes minutes for arround mio files on nodes +Attached is a simple patch that implements this functionality +There are common classes for ocsp and crl such as +GI for Geotools Implementation? I understand that it looks like a soldier +This writes files in the conf directory that it does not remove +Hi EdisonThis issue still happening with latest blocker critcal was created before July please review and resolve we are approaching code freeze in days +Thanks Chuan! Ill commit this shortly +Weve broken off the Nightly Downloads page under the Development tab +I set up a remote session with this users to see that they werent doing anything unusual to cause the issue +Note that also contains a timer service implementation +Created and attached test case. +It should be there +Is there a chance this can be converted to a subversion patch? We dont have git since were a windows shop and the version of git for windows is too unstable for +Assume closed as resolved and released +That is hierarchy levels are skipped +batch transition resolvedfixed bugs to closedfixed +This latest patch doesnt compile +fixed in +Fixed the test file as commented by Namit +Dan commented that this is the correct behavior +Thanks Raghu for providing an implementation of a Throttler that can be shared by mutiple threads +A smarter way to test that is to ping the URL from each repository specified and check if no is request merged +IMHO we can do the same with less changes +For The createSession method now waits until the connection state OPEN before it issues a session attach +Oops +Ok this is verified now. +Seeding and the separation of inference annotations are not done yet +In my code I had to create a hack for users with few items +Setting the Fix version so this gets picked up in +Sorry for all of the comings and goings on this +Server is unblocked from rsyncing and has been synced as of +Fixed this +When I ran a unit test in it didnt print the version number +Hitesh Ive quickly gone through your patch for +if you like the idea could you fire a jira issue as an enhancement? +There are quite a lot of PD +thanks steven +Ok I think I kinda figured out why deviceready isnt firing In the file it deals with firing the onlineoffline events for Cordova so removing this will cause these events not to fire +This is not to say that throwing an exception in case of null is not a good thingThanks +The unsats are gone so closingThe test does still fail a ton of assertions and then hangs +I like to test with CR again. +This patch fixes the build hang on one CPU +will fix that and make stopGracefully work by SVN revision +note the problem here is that Oracle is extremely bad at jdbc metadata concerning their SYSTEM db +Downside is just that Im currently using an in house developed library on top of JCR which unfortuantely I cannot pass aroundWorking on an isolated test driver that talks just plain JCR though +This issue is not always reproducible I will post the infostream output once I get it reproduced again +Ill link it and close it +Now working when CMDCTRL Shift U is invoked from the editor +This patch includes fixes for the project which I missed in the previous patchCommitters can you please review +for this ability for users to choose among which selectors are available is s an extra requirement that I am bringing forward not sure what the complexity on this isThe list will be populated based on user permissions of some not all the users can deploy production rules but all can deploy QA rules for example. +Thanks +Username djencks +Last I tested fixing the payload issue was not quite enough +Patch looks good +Trivial patch to fix this documentation +I think that murmurhash is here and a simple prime number congruential approach versus the indexes is likely just as good +There wont be a release +Same from head this morning been running it in our build today and its working great no more sourcetest pollution and a nice side effect it also seems to fix +Even better patch that positions the tip better for cases near the upper right corner and still works well in the lower right of a narrow display +Another topology that is popular is having the message broker remote from the actual jboss cluster with its own backup which means if a cluster node crashes the messages are still available +Rebase updated javadoc and findbug warnings +Good catch Sailaja +This was due to a configuration problem on the users site which we found and corrected +Interesting +New patch +fixed in trunk and also fixed this for xwork because Confluence uses xwork +The first two patches are incorrect +Yes please could you look at this error +Two of the blog sample tests are still failing but those failures are related to changes made under. +Had to change the extension of the file and reattach so that JIRA treats it as text +bq +Parser for response files in Findbugs is broken +Attaching the refactoring patch +Testing the new patch now will post soon +This is great but are you sure this is the thread that was segfaulting? There are probably a couple dozen threads so posting all of their backtraces might not be a good use of your time but can you hunt around a bit and see if anything pops up +That issue seems at a glance to be about which I dont use at all +Assigning to Hui as she is working on EAP releases primarily +If you can take this up with and make progress toward the original behavior we can try to reinstate it +This issue is caused by fixingSince the issue described in doesnt happen in Eclipse Juno I have removed the fix from +Marking closed. +Here the patch for Robert which fails the tie break test +Committed. +Deepa is this still an issue for the JME runs? If so can you prepare a patch for this? I will be glad to review and commit it if so +Attached patch to fix this +here is the patch of both encoding and s great if itll be in +So havent used CXF at all + is a very problem specific solution and may not be what other users expectIn fact it depends on the email client used which filename will be set for the inner email message +Please patches after correcting followings +azarutin per brandon please retest on +Thats been preconfiguration and rest apiIt also shouldnt stop working I see this use case work all the time +more refactoring of couchkeytree new etaps test for +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +The cygwin bunzip also has no problems with the +Do you think you could give that a try? Also the Eclipse plugin should now build and run +In the mean time your welcome to try to write one yourself +Unfortunately thats a pretty big architectural change and will be difficult to apply on +But as there are multiple ways to call serialization I was hoping you could point out call you are making that has this behavior so I can fix code to pass proper message. +Patch available +but shouldnt the output of X say overriden by as it does for other use cases? +The example above with a and a plain Controller did not make me think of a Spring app +update the patch for recent changes in +when to support ipv in +Bug has been marked as a duplicate of this bug +Hive should do this already +Cool thanksThere is one problem with the patch I am still thinking aboutIn the bulk assignment I keep the async ZK node offline for the performance reason +When i deploy my WAR file in tomcat and run in browser that time its give me error like thisWhen i run this in eclipse its run very wellCan you help me how to solve this +FritzI just commited a change to the Journal and pushed out a new SNAPSHOT out +I think Teds patch is good +The name of the transport is a tad ambiguous though +When you say that the class is still used by legacy code is there some reason this legacy code cant be regeneratedFor our use cases we have not updated thrift for a while +this patch removes the kill command updates the docs and also updates the script to use pids and unix kill cli to manage the processNote I removed kill entirely due to its potential security issues +only see references in pom now. +class +If its not a core dependency and its only used by contribs we ship it in the dir of that contrib +jbossxb CR available and so it is Beta +Ok seems to work ok now with the latest changes to the create script +The groups all get updated but inside i still have etc +please that run trees with expand nodes are now working properly in Opera +Freddy what do you mean by full content? According to patch it should print only absolute path to the file +I verified by hand that the istat tests now fail if you run them the required statistics tuples no longer turn up +at that line the prop property is not proposed +Verified in JBDS +Possible fix for the issue +This fix does a couple of thingsa It adds operation selector support for all types of operation selectorb It fixes an odd issue where if you had two bindings on a service that were the same type it didnt refresh the UI properly on the bindings property page +Thanks Daryn for the reviewI have committed this. +Alan this is great thanks so much +I am not aware if this is supported in Hadoop +You also have to consider how often this happens +Sending srcmainjavaorgapachecommonslangexceptionTransmitting file dataCommitted revision . +Not getting done in the next few hours +Please flush the browser cache by pressing CtrlF while page is opened +Attaching the correct patch +Hi Devaraj and MahadevI will merge this to +What do you think about that +ErickI am not sure how helpful my script would be it is very targeted to the way I install solr and OTOH my unix skills are not too sharp and I might be doing stupidless than optimal thingsOther than that I would have no issue sharing it + +Perhaps you can put this info on the wiki for future reference. +From the referenced email threadbq + +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Yes indeed +Which is what we should consider supporting as well like we do for Hibernate +committed to trunk. +Attaching first pass at a solution +This patch doesnt need to include any tests +Regression tests added. +MarcThis sounds great! Many thanks for such a thorough analysis of the problem +Default jnlp generated by nb ant +Your broken component architecture is not our faultTo change the flushing behavior set as per the documentation +Thanks +No the initial integration is pretty rudimentary automatically places tags on pages and allows for the editing of the tags properties by the property inspector rather than by hand +How could you differentiate a previous failure from an existing node working protected by the lock +Note We need to make sure we are using the country field if its accurate + +a patch including load in the caching in a is a new patch makeing the loadSource load protected +If can think of a good spot either then if someone could explain how this fits into the overall JBM component Im sure I could find its home +is just the class Im used to and my fingers automatically typebq +Thanks GregoryThe patch is applied as expected +Updated patch in synchrony with the latest sources +Apparently the scope field has been added for to preserve the scope in consecutive eval calls +Ive done a full test it so it works for the online help index and also the docbook html file also works for viewing +is a TP in SOA and cannot be a blocker. +OK +I believe this is a known bug in CXF +You did forget to attach the screenshot +replace the iconssvgicons icon with the attached oneThe icon is part of the svgicons library +Please review it and confirm whether or not its an accurate for providing valuable and detailed information about the underlying causes of this issue. +I applied this but I think we should somehow make this schema available no? Not sure how we should go about publishing it right now though +See. +This bug was fixed in Xalan Java. +Patch to add to committed to SVN trunk and branch +If no objections Ill commit the proposed solution on the patch soon +Definitely seems like it +Ill look into it +Is it worth writing a visitor that removes unnecessary casts +Craig great +and M arent working anymore because theres only one runtime +Since this indentation strategy is a bit flaky I will only apply it on the first two pasted lines and preserve indentation relatively from there on +Is the implementation very different? Eg has which this patch introduces but is not in trunk +bq +Also changed everything to use newHistogram as that ensures consistent naming for with the rest of the metrics +Thomas could you comment here +Not something that affects code much but it definitely impacts the unit tests +I think it covers everything described in the email +Thanks Martin +Thanks for applying the patch +it stalls the maven build like every seconds +Things should continue to work if we simply remove that code since the token already has the address and there is no need to it +Is it OK to apply the NIO changes to get around the connection dropping until jetty is out +Juergen +Thanks for your bug report but you are completely confusing me +I have changed the splitLocation to splitRatio to indicate that it is float we are looking for +I mark it at resolve as the unit tests does not fail for nor for +Attempt to add clarity to when the words workspace and resource are being used in the publishing wizard +Fixed problems with iterators +This appears to be a duplicate of +We get the option of dropping messages and we can work properly with inbound and outbound messagesCheers interface commited +works fine with current Sling and custom importers +However I went ahead and made a very minor editorial change to it. +Is there any reason not to commit them now +Moving this out until we align the usernamepasswords in Fabric ESB MQ +Let me try if its still the case on +I know this chunk of code went throughsome birthing pains! +Please if this issue is still valid. +Committed to as part of +Patch looks good to me +So when the Flex release comes +Fixed on closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +spec file to make it work +There are committed to branch +so resolve it for now. + to change the default classpathScope to runtime +Ive added a super simple test for this +bq +RI output on it is ACCEPTED connection got request sent proxy authentication request ACCEPTED new connection got authorized request sent OK responceConnection got responce code Connection sent authorized proxy request truePASSEDbut Harmony acts as follows ACCEPTED connection got request sent proxy authentication requestConnection got responce code Connection sent authorized proxy request false patch fixes this problem and implements regression test +uninitialized constant SocketSOREUSEPORTIn terms of the differences in constant definitions there are a very large amount of differences and some of them are likely platform specific +submitting patch +Looked more at the code not something that can easily do with the current IO API so killing this issue +Committed to trunk +There are patches for this issue +If the ratio is bad assume that theuser made a mistake and switch to non aggregation plan +Same with branches +it has come up in the past and is usually traced back to the XML parser you are using +This issue is handled in and ll resurrect this jira and use this to get track the hadoop configuration user guides +When reinitialize they are binding again to port zero and hence they are getting a new trackername +Move the fix to +I do not know why Finally it can run +Applied in commit abf. +MIS thanks for the patch! Committed. +It fails without the patch too +I deliberately did not use the subprocess module so this could would run on python +Attached patch modifies openejb but fixes an issue that is present only in geronimo +a session leak in the test runner +Thanks a lot Ivan for fixing them +in r +is propertyDetails nullregards Gerhardactually my application now generally works using bean validation together with extval +Mike OdiI have made some heavy modifications to the and code in order to enable connections support +Ive fixed this in revision As we allowed null explicitly in the past i kept thisBut I added an additional check if the resource is the same as the current and noother attribute has been specified to alter the include it throws an dont think that this is the right approach +in please if the issue is still is still a problem on the latest code streamWhen I get the code from svn and build the build is successful but does not have and the above problemexistsI am unable to reopen this issue +Ive assigned the final fix version to M from final. +Similar to hadoop behavior +Works for meThank you Alexey. +IMO the latter requires that you parse the query in order to find out what and where the bind variable names are +The new files need stability and audience annotations +has example math in the patch +If Hibernate throws a plain in our face all we can do is rethrow it as +Use FOP +Attached java the diff file +revisionId +Thanks CarlI applied this to trunk +Niall If this is fixed I would like to assign this to and resolve it +After going through the patch briefly I noticed there were a few iterations over the s Running list +When that happens it is impossible to have the reducer read typedbytes and write text when the combiner is on since the reducer expects typedbytes in input and the combiner will provide text +I guess I will reattach the most recent patch so that it has it as the most recent attachment +Look at the list of +I dont remember in which version it was fixed +It generates one wsdl per service +It can be opened again if there is still a problem that can be demonstrated +No use in doing it twice +this patch applies on branch without conflicts reducing diffs between and can we consider applying it on with something like a h timeout +Due to a heavy travel schedule ill have limited email access over the month of July +Id like to give this a try over the upcoming holidaysWould it be possible to post a single patchA single patch can be made by locally svn adding all new filesand then doing an svn diff on all files involved from the top directoryRegardsPaul Elschot +i think the issue of reinitializing a servers data and reintegrating into a zookeeper ensemble is very reasonable +It is compiled from Apache Commons Daemo rev without any changes +This update to and fixes the issue +Lets see if I can get it running with a patched jline in Apache Karaf. +Will ask on the list +Ill move checkstyle to and update the path to be a relative one +I am planning another release for the meta model generator by november +Committed to trunk and Thanks Shreepadma +really need a reduced case on this one or were going to have to punt it +PS +I will commit the changes after I run the full axis test suite +Committed at subversion revision +is being logged now by the default Monitor implementationThe default monitor can be overriden as people might want to monitor progress in a better way than reading logs what is missing here is to define a good interface. +If the branch fails to compile I wont be able to do thatIm having holidays here this week so will be back on Monday and will reviewmerge and if required discuss the changes to the switchboard github repo that you pointed to +Sorry about that +Have you tested it on a real cluster for any length of time +Merging with the trunk again +If it is shutting down or refreshing you could simply evict the context specific caches +Will it be solved +Attaching a second version of the repro +Tested with multicore configuration on Tomcat and single core configuration on Jetty +Ive tried to comment this option and do some UI testing and it seems that disabling it does not have any side effects +It is experimentally done +bin startstop rest works fineunit test is green locallyIntegrated into trunk. +Unfortunately Im not familiar with cmake so I am unable to provide a fixRegardsAlfonso. +Ranie not sure I agree sometimes searching and not finding anything is an exceptional situation and sometimes not +Patch diff for and new file +HiThis is the command Cjava versionCgroovy versioneach time they are not recognised as an internal or external command +Todd Ive committed this +Changed fix version to no longer barfs against pending trunk +add operator definition in r +And consider creating a branch of the ant code at codehaus +For me that would be the perfect solution and I think the most common use case as well +Fair enough Ill do that. +Manually tested +Dan TranAssuming the directory is added to the source path for the compile goal then yes the filter will apply to the generated source files as need to update this patch once more per a discussion with Brett regarding the layout of the configuration items +Resolving the issue since did not heard form Benjamin I tried installing the plugin as mentioned and then removing it with the temp files and installing the plugin it working for me +The use case is we can delete invoices but we cannot delete the related products +another example that causes the same attach a simple runnable testcase so we can reproduce this +Velocity should be fixed now try again in a few I understand in it does works with but you want to use other version am I right? if thats the case its not a problem of poms you have to cnfigure maven properly +Built on different local machines and no javadoc warnings +Ive check this version and I just see those issue Portal Actions arrow cursor arrow cursor when you are hovering that people iconCant reproduce other issue +There were some things that could be refactored but I tried to transfer as is a lot of the logic from the classes to the File classes and the +Extract this example to the Seam examples folder and read +Would you post the full stacktrace of the exception? I just wonder why nobody else complains for this error +Tim I added you to the wiki ACL + added to dns config files in svn just need a dnsadmin to update the zone file and reload the zone +Hi Martijn any idea when range and query facets will be added +If someone tries to paste it into the file hell have to make it back on a single stand alone test to reproduce the reprojection problems +Closing all issues +Committed to branch as well as master +Ruth you are right patch applied at revision r thanks a lot for this please verify +Might be a right time to ping them +How do you determine if a file is real or not? How do I undo a symlink? What does this look like from an audit log perspective? Is this really just a short cut to providing the equivalent of mountIf we ignore what I said above Im trying to think of a use case where I wouldnt want to symlink an entire dir +Hi Dave turn on verboseBoot and verbose class loading to see whats going on +Committed +Its all Maven committers anyway now working on this. +Because it worked with back compat the people that added did not have to update all the tests at the time something thats commonly doneI think that is totally okay though because we are here to do it now that it actually must be done for Its actually probably simpler to deal with now than it was back then +Now the sucker password need to be specified in encrypted form +Thanks for submitting the patch +Change repository location in so there are no spaces in the path. +Please update your eclipse formatter by going to Preferences Java Code Style Formatter ImportAnd then select the file and apply +So Nicholas if know how to do this please reopen the issue and correct it +Im wondering if there is actually more deep seated problems here +Please reopen this bug with sample project and steps to reproduce the problem if this does not work for you +Fixed on and closing all issues that have been in resolved state for months or more without any feedback or update +Patch +What is your exact environment? OS version JDK version etcIm running the app in the debuggerI have three windows in a tight loop executing requests on the appwhile true do curl O g httplocalhostvalidform doneI did find one potential deadlock situation inside but I havent been able to get an exception thrown +Yes I stumbled across this several times as wellI agree that we should make the default for the next disagree because it does not match the DS spec which defaults to +This still works with checkstyle beta +Thanks! +I ran a bunch of tests locally but well see if Jenkins can find any other issues +It seems like it works fine for me +Fixed in branch in +Forgot to unassign +Inadvertently attached a yesterday here is a CP CR standalone log. +Fixed I will wait until later today to check this in when we resolve the foundation profile changes +Shouldnt have closed in the first place +This is not regression from +The result will be a single oversized region at the high end of the key spaceSo as far as I can tell the existing behavior does the wrong thing and furthermore does it wrongly +JonThanks for committing it +The individual key processing counts need to be updated as well +Attached hotspot error dump +Please create a separate jira for HDFS +Removed link to as that JIRA is closed +yes +I updated checkDirs to make use of the flag +Hello MishuraThe fix looks good +It seems like we should generalize this a bit more to cover et al as well since I imagine the same problem exists there +Thanks for the patch Andrew. +You are probably just as likely to find it cranking for seconds on vms timdes a day as on one machine cranking for hours or whatever +Its just that ifa Query wants to use an scorer it should also implementan one +With Spring IDE M we shipped support for the requested feature +BTW this only effect the UI +the counts are just incremental operations on I dont really see the added value in using TSTheres two main benefits we get a full time series of the collected values and we dont need the extra interface or custom code for accessing this piece of information +Now only IWC is Cloenable +The changes in the patch look fine to me +So in this case you didnt use as the data type? Is the above the exact code that you usedCheers improvements made to simplify this use case. +Hi GeorgeThe fix looks good +hard code strings consists of in the source codesshould be also be addressedMice +In the description in the it would be good if we have a link to account verification related documentation in the +Furthermore you cant implement context sensitive serializations without the changes toI think they are orthogonal +Hi david Did you mean a separate commons project? thanks checked in the CN code to revision +Shouldnt return in this case +are right I meant just a typo I know it has been fixedI followed your advice and opened a new issue +I have added the output from mvn v as requested +This issue has been fixed and released as part of release +Are you getting any errors in the logs +Forgot to add it again though +Maybe this should go into rather than? OTOH standardizing on a known and proven interface provided by a commonssubproject is probably not the worst thing to do +Carl shall we close this since is already in or is there anything to be gained from it +The issue is observed in found that on accessing at end socket goes in CLOSEWAIT state and would remain in that state for long period of time +So this bug is reproducible under Linux too +Revision . +Attaching jprofiler snapshot with unfinished RUNNABLE http threads stucked in +perhaps the or Struts taglib folks would as well if you offered it to themill give this another day or so so that other folks have a chance to chime in if they have an opinion +but need to allow exceptions here +I can only add newer version to classpath what version of works with your fix +Reind reports fixed +But if I spend more then seconds a week thinking about logging my head explodes +Screenshot with a new toolbar position was attached +Holger have you figured out how +I think a better approach would be to do something like Torquebox does +Didnt work either +An alternative would be change the default appender to be the which is defined within the logj configuration file +Thanks for the thorough review Ted +I do not think that the test failure has anything to do with this patch but will resubmit the patch to make sure that this is true +It still has no CDI nature +Additional note from the user on IRC I adjusted the live server config to run the solr jetty and now all the stats are showing up even on openjdk so the issue seems to be with older jetty versions +Yes they are all in now and were ready to put this one behind us +As new plugins or features are added you can log them selectively +bq +unless someone wants to open a new ticket and submit a patch? +TIA +A regression with that will need backported as well +Thanks Suraj +Just wait that somebody have a chance to look at it I guess +To be clear I suggest changing the default to some pretty high number like or possibly be a function of stack sizeheap sizeAdditionally if my analysis is correct the async stuff should be completely deprecated as no one should ever be using it for any real cluster unless possibly if they use Cassandra as local storageI also suggest changing comments to reflect the real meaning of this and what it affectsI will provide a patch for the comments and changing the HSHA defaults to a fixed number if anyone s the idea +See for further details. +Should be ready for large scale testing +Should we change from bug to enhancement +I just tested it and it works fine using a snapshot build +Looks like those hardcoded port numbers are not really needed +They have planned a RC and then the GA but no dates have been set a side note they are planning a release of no date yet either. +Basically you need to turn off the autoreformat in your IDE so you can create a patch that includes only the changes associated with the issueThanks again for this contribution +I dont have a solution but Ive noticed that decodeexactly is less lossy then decode but still google code is lossless +He has tested it in production but has not been able to provide a regression test +Parameter changed +Thanks Ravi +I tested with Tomcat via our IDE but also directly with Tomcat deloying by hand to ensure the IDE classpath will not interfereIt is confirmed it doesnt work +Also if there are delays in this could still slip back in but it is not blocking the releaseQuestions Would a protected settergetter work for you? That would expose less of the internals than a protected member variable +I do not think I want to change this behavior in +I propose to make it part of specialpurpose application set +Created a new issue +Trivial fix replaces literal input with input name and same with output +settings folder into the resources to add and commit with git +Nevertheless the API allows skip of default values we should fix it +Already in trunk. +Brings in the Wikipedia capabilities from Lucene adds an Analyzer for Wikipedia cleans up the input parameters etc +note added to +I didnt look into file before but I agree that its working after removing broken +Closing. +The plugin will now remove the Java Builder when the option to generate classfiles has been selected +Thanks Wendy! +MS log is MBPlease access the log from here cpgvolcommonlogsThanks is because all your hosts are in Disconnected state so when we need to send a to a hypervisor host to copy a snapshot from secondary storage to primary we could not find a Up host to send command to and thus failed in create volume +Would you mind doing just the genericizing part in this JIRA and we can discuss the other use case separatelyAlso please add a test which uses this new facility to initialize BKJM edits if you dont mind +I can also reproduce this issue by running the using Ant +Not done yet +Adding a more colorful logo +Its easy enough for me to run +We backported this to sorry! You can close it +Btw In ij we have the command show settableroles etc +It committed as of revision . +I did not want to complain it was just already the idea behind! Thanks for reminding that we may need some documentationSorry for problems but maybe the code pasted above helps +Im sure using Brand Fuel is adding a ton of expense +The test failure is not introduced by this patch +Thinking more about it packages in services probably shouldnt be in the Store boundary +No idea +Based on Hairongs measurements we expect even larger benefits with which have longer path names and more files per directoryIf it passes full testing I recommend for commit +I looked into the code to see if it can be done + solution would be to not kill tomcatbut always do a nice shutdown +Dont know if this helps to diagnose but I thought Id mention it +I understand this complicate matters because the parser probably needs to be updated accordingly but I believe wed better bite the bullet nowIt would also be nice to add indextype support and to update the CQL doc +KonstantinApologies +I dont have that section +This can be avoided if the table scan operator is not replicated and has multiple children instead but this optimization has not been done in this patch +Thanks for the mammoth patch Enis +bq +Debug build is just OK +Thanks for the review AlejandroIve committed this to trunk and +Wed like to collect as much information as possible if it fails so that we can triage any failure quickThanks for the review +All Pig test pass +Verified in Cr +The release branch is no more actively maintained. +It works just fine here as well +Kiling resolve this issueThanks Sergio + +I have just committed it. +Nevertheless its really better to fix them as youve done +We have a customer using entity sync and experiencing the problems in this issueWe are interested working with other parties on this subject +Added a single test case for the last item and synced everything to beaver will deploy my updates to the actual codebase to beaver later tonight with other be able to test results usingcd repoclean +Thanks Crisitan for finding the issue and pointing it out. +When I update a class from the outside it simply worksAnd I also see that there is no restart of the console +Hopefully this time I got it right. +For some clusters the current sometimes replicas is optimal for others it may not be +Verified in ER +And my last experiment before I had to give the machine back to users was to reboot with stopped before I rebooted +and framework +The issue can be closed after verifying whether the serializer jar is actually required +Thanks BobYour patch is in trunk at r Not much to say but please take care of tabs +You can reach me by email or mobile +Whether to use coprocessor it is not a big concern to me +If you are working on a plugin feel free to start a thread on the dev mailing list as we can give you some pointers on where to look +me has not yet finished initializing its internal state +Fixed small issue +Click on Quick Ship Entire Order and followed by Create Return from the page +Spring IDE added +Moved this to as it is probable well include this as experimental inMaybe some and furtherwork will be opened assigned to trunk and merged to branch if agreed + causes s pathInfo field to start with +I agree with you +Anything else related to this can be done as enhancement later. +When I move to the thread solution the application shutdowns fine and the exceptions arrives as well +Configuring JAXB Annotation Introductions in Should be removed +The recursive option for info is not requiredUploading the patch with the other above comments implemented +This is just a suggestion and I wont insistThanks + +I need to investigate more how to solve this +Is this still a problem +Could you please try the recent JAR and reply to meThanksTrustin +and have no license +We should fix asap +I have been a bit lazy myself in not following it up +fix findbug warning +I created an initial implementation based on the +I am closing the issue. +Doing some investigations on available vector data I have an average failure rate of +I am still getting the following errorDescriptionResourcePathLocationTypeProject hivelocalws is missing required source folder jdbcsrcjavahivelocalwsBuild pathBuild Path ProblemProject hivelocalws is missing required source folder jdbcsrctesthivelocalwsBuild pathBuild Path ProblemThe project cannot be built until build path errors are resolvedhivelocalwsUnknownJava Problem + the issue missed resolution +Bulk close after is released +Also while it is a hardcoded value in this action the action brings up the creation dialog where the user is free to replace this value with another if they choose +Source of test patching the test case it no longer triggers the bug +I will assign this to me and will keep an eye on it for a few weeks +figured that I still have a few bugs with deserialisation +If youve got the chance to please give it a tryThe actual release is planned for Monday afternoon Ive verified that last nights snapshot reverts these behaviors to be like such that my application works despite the order issue +and two test looks great +Logged In YES useridSorry forgot to add I have downloaded bundle from projectsjetty and replaced tomcats sar with jetty +appliedThanks! +I completely agree and you already have the contract documented in the interface +Sorry for the noise. +Version has been releasedClosing this issue. +fix patch format issue updated javadoc +get is not in hbase +Similar issue but not sure if it is the same. +Added logic that will not send notifications if the event user is also the notification request user +Here is a new patch which incorporates Jochens changes on the Groovy side. + +Martin how do I run the single test in against mysqlIn this would be mvn Ddatabasemysql Dtest build of that branch fails for me with BUILD FAILEDhometdieslersvnjbpmjbpmbranchesbuild The following error occurred while executing this linehometdieslersvnjbpmjbpmbranchesbuild The following error occurred while executing this linehometdieslersvnjbpmjbpmbranchesbuild The following error occurred while executing this linehometdieslersvnjbpmjbpmbranchesjpdluserguide Cannot find hometdieslerjbpmrepository imported from hometdieslersvnjbpmjbpmbranchesjpdluserguide you please fix the build or how the dependency on is handled +worth exploring +Resolving this one as a wontfix instead well discuss convert on. +Yes but this one is mainly for mkdir +Applied to and trunk. +Will pass by hadoopqa again +I added a link to the +If you do please update the and this issueThank youGary +This problem is not their in latest versionof xercesWorkaround Use the below mentioned command in vi to remove controlcharactersHere M is a control charactersM +You should move the from import up to the top of the fileCant due to some fun circular importingbq +Would you please verifyI noticed that in your xml document you are using xsino to refer to +Patch attached + is now has been upgraded to version will this JIRA be fixed for this EAP CP cycle +Does Dhruba plan to continue with or does it make sense to incorporate themTodd if you have the time and energy please feel free to reassign to yourself +It works now after rolling back to then to the current version +Temporarily reopen to update release note field only. +The patch deals with ? but we also need to handle at least ! and + +i met this problem toowhenever i got this exception i can do nothing but restart the activemq server +Underscore should not match aa because aa is not a single character. + +Hmm +This removes the need to replace them with references ? in the sql when writing into the index. +Would it be possible for someone with an understanding of mind telling me what its purpose is? It would seem that its more than just a transport and some context would be usefulSome comments on the current version of the patch No author tags are allowed in apache code Would be nice to have a class comment explaining the Event inner class chunkState inner class should be consider writing method defs across multiple lines instead of all packed up on one line Should be? Please move tests to a separate test file that lives under test +I may have some time soon though. +i could not reproduce this against the latest have verified both the quickstart and my applicaiton do not exhibit this behavior with the latest build. +Committed revision for client patch +Should I post this in MEV instead? If so this issue can be asmasmjar Try downloading the file manually from the project website +Moved to but will probably be closed as fixed in as the submitter claims it doesnt work in +Updated file in +Further to this is that no persistence methods will be generated as persistence operations will delegate to Spring Data +I only can reproduce this error when the BRMS is down and the target Guvnor repository is really not available +Done +the initial in the lines is just a numbering symbol +Thanks Vinod +i am working on it +The new binary drop resolves this issue +Im tempted to close this one since we have an existing open issue directly related to this however thats on Android and not on Android +Ive been experiencing this issue as well +CALCM makes more good from the commits will review more in depth later +Thanks for the new constructor tooIm not sure if it makes sense or not but there is the class that could be used instead of a two point array +Can you please give it another try. +Patch to fix this problem +Just a quick note regarding +Reopening to backport to to in r +Committed at subversion revision I ran the upgrade tests against trunk upgrading from the following versionsTouches the following filesM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsA javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsM javatestingorgapachederbyTestingfunctionTeststestsupgradeTests +How do we move on? Tibor Lars Adrian +In my environment this is broken on Windows but works on Mac + an hello world returning a value notnull and validated manually should fail this test doesnt valid the returned value +Latest patch for trunk based on changes from Rick Curtis +Fixed it to match what is being done in distribution because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +cool thanks for fixing that +This is by design and older Camel releases had this as a bug +Where these Archetypes should be released? or Maven Central +At the worst case on RS crash we might end up already compacted files which wont affect semantics +a perspective extension feel free to reopen +Thanks Tim! Much appreciated +Of course when youre pooling heavyweight objects that dont implement Runnable the approach doesnt really apply and the callback is overkill +Attached a potential patch +We should discuss it for M +I ended up doing both returning as the estimatedrowcount after detecting retfraction and then at the bottom I made sure that the rounding did not end up causing an estimated row count +Only one note forms action will not be worked in this case because JSF engine can not process multipart requests +We should not open up internal without having a compelling and concrete caseOleg + +Attached is a fix for the namespace table is not being editable issue +Besides Im not sure how I feel about a ton of different each with small features that separate them +Thus in r changed to not include the servlet context path in the request substitution variable. +Not sure if its still valid is old +I tried saving the transform and it and now I get a correct graphic position for the file +Closing resolved Release +Deferred to for issues resolved in released versions +Also moved test node type into class to clean up the package and make more consistent with the approach in +Breaks into several tests +Needs to be fixed in using openjdk and Sun JDK +We need to crash the node when retry fails so that the failover can happen +Nice +Committed revision +I kind of feel that it should go in the recovery service +I did some testing up on cluster creating tables and then bulk starting and killing a server during startup +suggests that we add to our regression tests +Thinking about it I wonder if we might just make a file the default response +I feel pretty strongly that this is a good solution towards those bigger jiras +The startup scripts need to be updated to work with these changes +Heres a second patch that works on the branch +so maybe lets put this onhold till release +Ill upload an updated patch with the fix by tonight +Do you have a hmessages in your JSF view so that you can see any validationconverter messageIt looks like JSF catched some error and so skips +Thanks Brian for help with verification +these api additions are significant enough to be what id consider suitable for master only +However there is no urgency for it so were moving it out to another release +The root cause is that a Date value is added to the variable resolver instead of a correctly formatted date string +If you finish it before we release beta please correct the version +This produced WSDL with the wrappers in the default namespace and the wrapper contents in the specified namespace just as Tuscany is doing +See initialize method is not called this should setup predefined classloaders +Sample to reproduce the against a snapshot thanks for fixing it. +Simple patch to use v does some cosmetics changes based on Ians patch +Its all good nowThanks for the help and input everybody +Added varargs and extension properties +I added the API docs from the old site to the new site copying over the content in revision +for the patch +The nameserver entry wasnt updated on so the mail server handling JIRA mail was not resolving MX records +Is this also the case for commitWithin on adds deletes deleteI suppose not this just covers the commit case +Great! Lets first target get it vetted and QAed then we can back port it to + +Graph of current LuceneSolr cyclic showing dependencies after applying the work improving this Steven what a mess +Thanks Sriram +Im going to run this through integration tests and then commit it thank you +They should only differ in the import statement +Could make it so we just put up the dfs for things like the zk and startstop of hbase cluster tests and then group more of these weirdo tests into the one suite +Please test and let me to in Git ID effaebafaaaefb +My first haphazard guess is that its not +after submitting the fix we dont have to rerun whole suite but only the relevant ones +Being done by volunteercontributor Giovanni Giardina +Using that information I can properly update the whenever the view anyone can register for these events +Please check that it was applied as expected +Same modification in the method save +Everything is ok now +Nice work Mike pretty polished +I ran the maven build successfully and eclipse doesnt complain either +I think this exception may be unrelated to this warning example if you try a greeter or greeterrest demo then you will see it works even with this you retest with trunk +In general it is better to provide a diff I can for example not see what you changed in guess this can be closed. +So here it is The change made by the patch is orthogonal to the question of whether or not droptrue can be combined with other attributes and it does make the sentence more correct +fixed in trunk is a patch for Beta branch +Fix for the bug +So that patch made the situation worse +I have a few minor enhancements in +I think making the reference count atomic is reasonable +bytes per a character on utf charset inso hive works with on CHARSETlatin inregardsyw kim +If this process ran smoother it would save them a bunch of headache +Closed due to the release of Tiles +The classloader used by OIS is the first one discovered by traversing the stack +I will check into this +Committed improved fix in branch rev The trunk fix completely ignores the main point of the lock which is to prevent copies of activemq in the same vm but in different classloaders from attempting to use the same file +However property is not yet documented +Can you give us a patchregards probelm is the of floadBundleThe bundle is stored in and not availiable in case of validation error +Patch applied on the branch see +is fine well go with that +its impossible to add an icon to itYou can easily add an icon to input controls by using the CSS background property. +Such a context automatically has the root application context as parent so can refer to all beans defined +Its not directly related +Ill update the patches +Patch accommodating Shravans suggestions +Close all resolved issues for Engine release. +I fixed this problem for japanese version Ill give my was good please use it +Committed revision +Done +Also on Windows note PC and false this always happens +patch v fixes a couple findbugs warnings adds tests rebases on trunk and fixes the jira number in the patch v to trunk +Thanks Andrea +This patch also passed full tests on trunk +RegardsLeslin +Yes this feature will be implemented but I dont know when for the time being. +Thanks for the patch +If you are interested in looking at what ive written so far let me know and ill cleanup a bit add some comments and email you or somethingregJoe +I rerolled +hopefully now fixed +But the new client will get the now such method exception and will know that he is talking to an older client +see history +I tested it with the and it is the same effect +id also suggest renaming to say to further highlight the environment its meant to be used inid also like to see this change made tothanks jukka +I will send this to them asking for an updated plugin and also decouple the version of the plugin from the actual dependencies being usedThanks for getting back to me so quickly. +Ive added some unit tests which seem to show it working Ill add your mimetypes fix test shortlyOn a different note its generally best to generate patches with svn diff if possible a single patch file from that is normally easier to review than a zip of single ve applied your fix in r thanks for thatIve also added in your test but to a different file its a general mimetype detection one not a special container one so belongs in instead +I hope this will be corrected soon +Should once this issue is resolved +which is the right exception make sync interface and async interface consistent to use same exception +What would be the affect of deploying this patch? Would it be best to manually delete those directories first +The patch was made using svn revision +Changed query builder to interpret AND OR and NOT as general query termFixed inSVN revision +One more tiny suggestionWouldnt it be easier to extend from and just switch the flag on before the test and back to its original value afterThis way the test would have its own dedicated workspace just like the indexing config test we already haveWould that make the test simpler +Scott says it is in the in the deployers directory and that he doesnt see this issue with the testsuite but I have the exact same xml file that he has and so it should work on both +like this is fixed so closing. +fixed +Forgot to ask if there are any important test cases I have missed +This should in fact help us to preserve API compatibility longer since the mapreduce kernel will be independent of the APII have always thought of hadoop MR as assembly language and gradually no one will use it directly +But Pat has not published any comment on this issue until now +Hey Ryanthanks for pointing this outThis is same code as inSo should we address this in JSF as well +However can you please help verify if G samples release works well withThanks +Patch allows to pass against hadoop is only a problem when running compiled against on rightI think we have more cases where it is required to build against if it is meant to be run against +has a typo in javadoc The mrOwner and admins always are is missing +So we need to see a high level design of how this is all going to work +That makes database backups more difficultBut maybe have a separate feature for wanting to run a job on all tenantsThere are several ideas floating here each one would need to be done separately +Passing back to you Philip as Ive resolved +WDYT +Fixed in the latest revision Ill mark it as resolved after adding a couple of test great! Just one thing can you please add the new command line options to the usage information of +Id be glad if you could test the for the report +Yeah thats what I was wondering +We were wrongfully skipping deleted rows with no columns during compaction +Whats the right move on the continuous build stuff +requires some harness Im not sure we want to invest right now +Can you test itDidnt you call a clean before to call the eclipseeclipse a second time +I am planning to put together a that would cover not only endpoint mutation but also ability to mutate segments of route +I feel all these are having readonly behaviours +Note if the specification does not support this concept we will need to give your function another name +Added some clean up between tests as tests got run in opposite order +Got this one +Closing resolved issues for already released versions +bq +You can have the most exacting whitelist but it doesnt mean its secureThis is one of the reasons why I think we should have a execution environment completely isolated from as a third party module and have a simple and secure embedded implementation built in +I dont plan to backport further. +You correctly found the two bugs but only solved one We dont use because it doesnt allow for heterogeneous deployments +New patch added + +I would assume then that the that is injected into the producer is going to be the one corresponding to the injection point which would be the right oneIf my explanation is incorrect can you please suggest where we should be looking to solve this problem + +Release was now published. +Max it was fixed in trunk but not for its fixed for GA +Im going to have to close this one. +Latest patch Ill commit this soon +Patch is good +Victor please take a look at this issue +But as far as Im concerned this issue is solved +It provides MRSW. +I reopen this issue because I forgot to merge the change to the TILESX branch + +We dont set that header at all +Im attaching a patch that disables bulk fetch if the table scan returns or +Revision enables to be run with just from the set of derby jars +We could possible also add some check in Fuse IDE so when people try to use the JMX explorer with the local processes then we can check whether that settings is correct and the can be found +Marking as added some faqs +Hi Alberto +Im going to cancel the patch for now pending a current use caseDo you count the merge into the reduce? +applied to and HEADOleg +may need a doc issue +Ill upgrade the plugin to the latest version +As before River developers have permission to use it in test environments but it must not be distributed so there is no ASF license grant +The Enterprise BRMS platform package contains two files and +Ill take a look before applying the patch +Also if you use a REVOKE or DROP ROLE statement to revoke privileges on objects that a view depends on the view will be dropped +Now the project will build with either the maven classpath container or the plugin classpath container +David are you able to work through some of the suggestions you make and analyze where we can save memory by optimizing our use of maps and setsOther things to consider include are results cached unnecessarily if a result is not used often it may be better to not cache and compute it each time dont create different objects for declared and normal will do this in phases that make the most sense +You can use the same name jira will take care of the versions +It is great to see this direction +This looks fine +trunk r +API +KR Koen +Implemented tested by ECS installed into tool is now improved to provide better error messages if problem is in files. +Ill have a look +Maybe changing the example to use only one queue would avoid the confusion. +The output of svn stat q is as follows +Applied a fix to branches please test and if it fixes the issue Ill apply it to trunk as well +Actually we need to remove the default value from if the value isnt defined let the lower layer generate random bytesSaving the random bytes lowers the security of the system +upgraded to latest janino in trunk + + Added as part of REST API +ThanksVenkatesh this will go in +I committed a fix +Bulk close after release of +The reason is we can do aggregation at source instead of in Teiid to reduce the volume of data being transferred across the wire and performance reason too +Ill continue to keep an eye on it +Committed to trunk. +Jody Ive committed a patch and updated test case at r as I need the geoserver build going +IMO this is such an obscure usecase that we shouldnt introduce any extra complexity to implement itFor example we should not guarantee exactly when a change to set will be seen + +One way to resolve this is to add an applicable branch in qpidcppincludeqpidsyst I just replace all occurrences with unsigned int +Dave right +Improved patch for review incl +Will commit shortly +Strictly speaking is an elusive goal we cant really claim more than a best shot using heuristics +That was running the Infinispan configuration that came in the ZIP of that performance user forum post we had a look to last Thursday +Resolving as incomplete +However weve not been committing v changes to only and trunk +Edit this wont flush correctly unless the patch in is applied as well +Packages have been rebuilt and updated +Picking which tasks to kill is clearly in the domain of the scheduler +Can you try again with undertow upstream and see if you still have the problem +If this fixes your problem you can close the issue. +bulk move from M to to m since no updates done by because of no update +The old patch file should be deleted +Cool +seems unnecessary +Patch looks good thanks +This can be set in the but if it is not the value specified in install will be used +test test case that uses random intermediate paths and user so it can be run multiple time also switches to common concurrent test utilitiesCleanup remains a problem in the absence of a way to remove Users +BJWill you provide also a new patch for this one? +Committed revision +Chris the component freeze date is +just committed this thanks Cheng. +If a couple weeks go by and this problem doesnt crop up I think its probably safe to say it was a thread safety issue all along. +JustinCould you please upload your java test code that you use to send the message? I am assuming you ran wj on the wsdls in the prev commentthanksdims +Ill try it on other machines +To resubmit your information and view this Web page click the Refresh button +Its ok +Do you think that this setup would reproduce this problem Create a small java class that uses commons logging Put logj in the class path Put a valid configuration for logj in the class path that tries to use an smtp appender which requires and to be present in the class path Do not put and in the class path Log a message using commons loggingAfter this do I also need to call logj directly +True +Thanks for your thorough response Kai I agree on having support for tokens with pluggable token validation +What version of Maven +Changin component to to xom since it doesnt seem to be a core bug +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year + +Even if we used asm to compile we would still run into this problem at runtime +It would make Struts worse not better if we supported markupIf Greg had submitted a patch it would be wrong to invalidate the bug +Yes Ill take a look at the issue this week +We have a namenode which will not work well if we upgrade to a release with this feature since it will need extra for the operation +this patch contains the fix and regression tests +If you could remove all the Spring related code contributed into the service configuration and verify that it all runs that would be helpful +In this patch removed unnecessary methods from embedded attribute +And we are avoiding array copy in those case +Correcting wrong exception method Exception specifies the name of the method not supported +Added the setter method. +Your application needs to be tolerant of reexecution of tasks +Downloaded MS SQL Server Enterprise and Developer Editions into Tims download area +I will close it Monday +In my examples I typically use require bundle for the cxf dependency +I am using current svn HEAD +Moving out to for now which means unless someone moves it back to this will not go into. +Were still seeing these issues occasionally using +If we simply allocate a new buffer per request whose initial size is k +It hasnt been reviewed. +Does AS has a separate plug in for JON? If not then it may be not advisable to Teiid either +I have gone through the patch on the review board but I still need to contrast the patch in against this one +Could be closely related to because its about a proxy key as well +I dont think this is a bug on the console but rather on the coreFix confirmed thanksColm +Provided patch with updates to the header of resource files +Supun +It is a large grain component and it just feels right to me +logic is moved to f +The parser will ask your your entity resolver for the appropriate input source +Are you proposing that these should replace? Your startstop scripts lack many features of like using a pid file to keep from starting multiple daemons of the same type on a host optionally rsyncing code from a master specifying jvm options etc +Yeah its Please review itI was curious how this worked so adding documentation and unit testing is a great technique to learn it +Attachment has been added with description Support for Arabic PDF rendering using patch uses ICUJ to do and BIDI transformation of rendered text +and querycore process request return solrdoclistsbut you get the exception which do not cast object type +Probably by our subsequent Cocoon upgrade. +Fixed +It used Java build path before +Thanks +Closing the issue as a duplicate of. +Checked into trunk +Fixed with includes SP and Riftsaw requires +was pretty much broken because it contained errors in the plugin configactually I think there is a much better solution possibleWe should update our to describe how to add an profile in +Please dont ask me whyZlib consumes input buffer at this returns needInput even if I call Inflaterrun at the same may need to implement our original InflaterDeflater in the future but I keep it as it is now. +Yeah I meant the DEFAULTENHANCEDSUFFIXORDER array not the whole cant actually find the subsystem in the actual source +Hello GeorgeThe fix looks good +sounds API changes at this point are being tracked under discussion glad to see this coming along well +Also with your proposition m could be local to getNext instead of private at class although I believe its at the class level as a performance optimization +Hmm ok moving out until we have something with a quantified benefit. +or I have seen the test fail with the signature you describe but it has the observers error in +Nice work robert while I only looked briefly at this patch but I wonder why is an interface while it could be an abstract class +Do you still see any problems with the ideapatch +For example when we find the template if it is a COLUMN marker we can take the row and column for the scanned row if it a version we take the row column and ts from the scanned row etc +Again I highly doubt that this is a tooling related issue +should the transient fields of the acl instance when it is reloaded from the may be seeing the same behaviour that was reported in +Now the system will be serializing the exact same documents twice that is extremely redundant +We can document how minversions maxversions and TTL settings interact with each other simply enough +. +That works Ill implement that advanced searches could possibly be added later +Which is wrong a security hole +This has been a problem with chunked encoding enabled for GET requests +bulk close for the release +So warning is not necessary. +That is because the SYSUSERS table does not exist in +was just looking for the same thing +is from and it doesYes flushing at MB is better than MB on my machine +I have incorporated all changes you suggested except for the based on Johns response +I agree with you that logging the entire stack trace at WARN is too much this is why we changed to log it at DEBUG +The following scenarioes are tried with a new created keystoreScenario step +If you cant log in with the ssh key then the ssvm wont get any commands either +Thanks +given all the bugs we have been able to find in the protocol implementation recently I prefer an approach that is perhaps less efficient but easier to read and interpret +Good catch! This is caused by signature differences between Hibernate and Hibernate which we have to bridge in our scenario +Thanks! Who would take care of updating also the documentation part +This patch gave me some problems when I was testing it so here are some steps to follow You should apply this to thats aready been built because the projects will look for binaries like etc +hama shell uses the javacc to parse the scripts +IMHO this is a bug in maven +Just a few little nits I think doing public static class is a little more common than static public class throughout the project +For example this testutility would use a thread pool on the client side and would invoke as quickly as possible against a test server +JonathonI think you are right Ive applied the fix you have suggestedRaj can you please test nowThanks. +As a result of talking with Alejandro and Tom we agree that well give up the integration into Hadoop core and cancelbecause to avoid confusion with usersdevelopers the code should be in one place either a separate project or integrated into hadoop coreThe as an independent project have the flexibility of doing releases and bug fixes independent of Hadoop releasesso I think Ill continue to develop onlyThanks +Yes the thought is to have a different key that the client generates during job submission +And just to clarify by multiple inputs we are talking about giving a mapper or reducer class multiple sets of key value pairs not using the class +on removing Chukwa from +Note that one of the goals is to propose an authorization model such that finer grained privileges can be added in as necessary later +This patch addresses Nicholass latest comments +For me it is fine to commit above patch +includes details on how to go about tracking the underlying issue which is very much a configuration need to move up to a recent version of Hadoop the fixes for your problem are already in the codebase +Keith Ive created bug for the problem described in comment +I was also never very happy with the number watchers that we were juggling +Joshok your mirror will be added to the list soonRegardsHenk Penning +BTW This works great in should know that not only the RTF output is very very alpha the PDF output of FOP is too +Thus columns of column families that do not contain any of the selected columns will not be passed to the filtersVasus patch would make this behavior predictableI think we should have this in and +Closed after release. +but I also ended up with the owning session closed like errors afterwards +Necessary tests seem already to be implementedWe also dont have a transposeView which is very handy some days +marking it for +the patch looks good to me +I changed to my trunkcore directory and ran mvn clean package +I did not see the fixup you did Jingguo +Ken Thanks for posting that fix for handling refcursors +Your comments are welcome +Thanks ChrisIm happy to participate others like Bobby Mayank would be interested too I bet +Can you make a patch to do the above please so I can test and commit if alls well +Committed revision . +Releasing with a standard version is obviously better then a dev build +Integrated into trunk and +Yes sure +The application can analysis the data make choices of prefixes set the prefix map appropriately and call the writer +It will be apparent when I run against our bulk data since it will end up requiring more memory than I have +Hope this is fixable +I cannot see any usage of config in the seam catch example +Please let me know if you want to access a database with this table remotely for testing I can set up a small sample set for you +Resolved in commit. +committed to beta +Neither atomikos or not present in it +Its impossible without some hackery for my callers to set their content type appropriately and still get the response format requested in the Accept header +However we will need to find a permanent solution for this +I am using which points to trunksmina +Reopen attempts to capture the current position from any incoming readable stream and it to force it to flush before it returns from reopen +Theres the beanvalidator module +Allow jobs to request encryption of the map output using a second job specific random key + will commit if tests pass +You didnt do it for one of your patch files +guvnor doesnt detect equivalent rules though brms did +why we use this Groschupf Thanks to your contribution in katta but do you think the lucene can supply some functions like facet and in solr? We all know the patch use Hadoop rpc connect the slave nodes like katta your want build a cloud protoype with Web APP server features +BTW I tested the patch with data migrated from a database +This patches will solve the issue +Just merged the initial draft into the master branch and deployed a snapshot build +You are right +SETXPLAINSCHEMA now takes a schema name and creates that schema ifit doesnt yet exist +What spinner though we dont ship with a spinner except during the splashscreen +Patch file is attached +Update trunk test resource Fail rather than error as per ALR replaces jpadeployers and patches +This is a new patch to accommodate trunk changes in locations of defined literals +I believe there are other cases that it returns null so attached patch just checks null after FilelistFiles is performed +I think it would be worth at least mentioning that value provides the initial text for the field. +Jean noted the problem with the ICLA I have sent email to Halley Pacheco de Oliveira and Jean T +Need this if were to scale and be able to keep taking on writes fast or even just so we look good in single server install +One of the main driver was simplifying the writing of custom namspace handlers in some cases but with the addition of the which allows passing an already created object it might become less useful +Verified +The entire space available in the data directories are free disk space +the servletPath pathInfo is used but its not really something that can be regarded as a bug. +The logging initialisation that was contributing to the deadlock has now been delayed +Applied the patch in f. +webhdfsv stuff off and invoking HDFS directly +Sorry I submitted it a bit first I dont even understand how I made such a mistake the assertions at the end of the test were wrong +Editorial and committed patch on editorial improvements +Yes we are using instead of the defaultWe copied an existing project that used it +The basically delegates all functionality to the original Feature except for getsetparent +Can you elaborate further or even provide a little code sample illustrating your point? Stefan just to clarify what Mario means is that process doesnt take in account the entity fields added in example the Base entity field modifiedDate isnt rendered in any view page +is not dead +I committed to and trunk. +This task is a forward looking task for the next version of the C standard +BTW in case it might be helpful attached are two patches I have that allows Hadoop to be compilable with true +As long as yaml is an acceptable format for storing in the meta data table Im fine with it. +Please verify before closing. +Here we go +Bulk close of old resolved issues. +Confirmed an issue with Apache Flex a couple of place in the SDK where are dispatched in the items array +This bug is invalid +Problem was not disposing of temporary +Thanks Jeff. + +In branch changed to return instead of in the above scenario. +Added a small workaround to handle the behaviour in. +house number ranges to the leftright side of the road or bollards at the start or end of the road segment +Click on the button assign and then in the dialog assign to me +I was also able to remove seen since we can instead use current! null +Thanks for looking Sergey +Attache is a patch solving this issue. +Neophytos has verified that my change actually works +This patch would greatly help me + is a spring patch to default loading support a liked interface to write test case contains a testCase to use using it will facilitate you to debug the testcase in Eclipse ide But still has its disadvantage to especially its resolve strategy only resolve the first started +Will you be interested in posting a patch for the same +bulk move of all unresolved issues from to +This is not what you need here you simply need a reference that internally pools Connections and Sessions +Ill have a look +I checked in a fix please verify +Applied from in SVN revision +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen + +I am now able to reproduce this issue +Updated STP BPMN feature list for latest Ganymede of required for BPMN feature that are not included with a first iteration +L L Fixed to store edit in start +The changes are minimally invasive and negligible in overhead +Possibly Javassist doesnt handle a static method accessing a private instance variable correctly when it converts field access into a method invocation +This change changes implicit casts in and to use the current schema collation when doing an implicit castI could not come up with a test case for as substr will not accept a date argument +The build doesnt start automatically after a commit created for that +I tested locally using a local version of JSTL and it seems to work now +Thank you very much for your time LarsJarcec +Great! Can we put something like that into too in addition to adding in Syncable support to the output streams as described previously? I doubt there are many people wanting to run production on top of so I would not consider it critical but IMO for testing getting its behavior closer to HDFS would be very nice +I shouldnt have closed this issue +I think we have to go with buildr maven builds to axis +sigh +Steve this may have been fixed by your reworking of the Replay compilation infrastructure +The dump is logged both on the and is displayed on the in the Errors tab +bulk close of all resolved issues. +Nicholas thanks for the comments! Update the new patch to address the comments +A plugin can only contribute one version of a XSD +Kila and significant errors that emerged when using this platform +I am working on updated javadocs to the Sink interface Thanks Lukas I really appreciate your efforts on this +Can you either give us a script that reproduces this behavior or try again with current trunk? Charles has reworked the IO subsystem lately and it might well already be fixed +Im still not sure how that prevents me from stealing the new session in just the same way with an XSS attack +You are right +skh pushing to to to to +Have you verified it? I did and it turned out to increase current JBT installation from to megs and I have Logic example compiled and running +I thought about using something like an increment or fallback that a cluster can use when there is an existing ZK ensemble using the port +As soon as is reviewed Im going to commit this modification run for and commit it as well to minimize time of broken. +Confirming that aria attributes are working fine for my usage +test case applied along with a patch many thanks +Moving out of + +I just tried the reproduce this and the wizard does not crash for me +hmm actually in the latest example its stop so because i think your other fix will solve linked pull request should adequately work around. +heres a fix to the problem +push out to +After we have the last checkpoint time shown on the webUI +Just wanted to indicate that were looking at the issue +to convey the structured label map as another request attribute with an internal after release +closed +Yep this has been fixed. +test case now passesIt actually passes in as well +That looks like exactly the right fix +Addressed reviewboard suggestions +I was complaining that the general community doesnt seem interested in maintaining the code +never mind I just applied it and also verified it seem to work great together with runtime detection popping up too +My ongoing jar parsing work will follow this design +As far as I remember the EL specs also says that only public methods are accessible +That solved the issue and Ill close it +Added comments on methods which are present in +This jira has been closed as part of a review and archiving of the aj project +Thats because the call to couchfileexpand isnt happening until the call to couchstreamwrite which makes the patch slightly less trivial than I thought but not that much harderDont force k inflation test to demonstrate lack of sparseness +Heres a patch that modifies the wicket hotel example to instrument the wicket components at build time if the instrument property is set which it is by default +It is approaching +UliIt looks like everything is being redirected to the Wiki Restructure page rather than everything except the old homepage but that will sufficeThanksRobbie +If you arent sure its a bug please post your sample FO to the mailing list +a user can have rights to FICO to store invoices and agreements +for the patch + issues associated with existing releases +dojo will be used I only used the release matthias is right +eddie im just testing plumbing on Jira +But these strings are hardcoded in the external +Someone holler if not +Fixed on trunk +NP for me +Was included in respective releases. +So Im kind of interested in what environment Enis is seeing ms +And if someone specifies both a name and a namespace I think the name should win as the namespace is like a package declaration determining only the defaultHeres a new patch with the tests you suggested added +Yes that would be useful +Could you try again the index command after removing both +Marking closed. +seems to me that the test case is just wrong that the profile work altered the test case behavior +This is what happens when the About Pluto portlet is maximized on Firefox +It would be good if there was an easy way to generate the POM file at some point before or during deployment +I filed for this +Ive just added a skip property to which defaults to false but when set to true skips validation if no header is present +I still need to add in some unit tests but I am not super familiar with the code so I thought I would throw this up first to get feedback on it +svn ci Sending Sending Transmitting file dataCommitted revision Thanks Donald! +This area no longer mentioned This area no longer exists +Oh yeah you mentioned cliffclick +It shoud be applied at the same time to avoid a build break +I committed the patch to both trunk and +Patch add in all matrices including dense sparse block and distributed ones +Note to self I do think were of the way to here so I should give that to you nextEdit Maybe looks like only takes a single Range as of yet +Yep it is same as +Thanks for the continued advice Dan +done. +So it looks like system property is not set for some reason even D option is in command lineWhat is your Target field value for Uninstall JBDS menu item +patch looks good new test passes + for putting this in an category away from communityenterprise categorization +Updated pom to reflect new packaging of jericho +Committed to trunk and x +Fair enough +We screwed this up we should fix it as soon as possible +Im not ecstatic about the current patch but it passes tests on my machine and resolves the issue called out in this JIRA +Yes there is probably no reason for not fixing this +Great work +Since there is only one handoff per target at a time we dont need to worry about having to skip tombstones in a meaningful volumebq +WSDL from CXF DanielJust now I realize that we do have but I dont think it is used for the call which I sent you but not sure +Patch to add a getsstables command to nodetool +Your issue is not really a common use case +re job still takes a loooooong time to display processes tab of details page +Any chance you could let me know what version of tar is installed on your system? Running tar version should tell you +This feature has been added +I checked the ejb plugin and found that it doesnt include any of its dep in its jar +Secondly an is thrown to indicate the logj Logger class could not be found +Merge from master waiting on +Resolved as of revision want to open a new jira for? So that this gets tracked on there +Hi Martijn this fixed the problem thanks +Something like this +Moved Named classes and class to this packageReformatted a number of classes according to the formatting rules I uploaded in +Fixed in too. +patch committed +Sample project that demonstrates the problem +Using the provided sample code and the latest source I have not been able to reproduce a crash +But basic natural ordering support would be a good first step +Patch to apply escaping rules to both and elements plus support optional newline escaping in both +Please review +Found a bug in the v patch that caused the multiple producer throughput degradation +Thanks Senaka for the Issue +Closing this +Uses osgi dependencies instead of +numRows +I just added volatile to all fields. +For method chaining to work you would have to pass back the Builder for each set call +Lets revert this change until we track these issues down +This is an interesting limitation of how AS ties all of these operations to that of the m not sure how we can work around this +Made all three writable +Maybe even rename the IT to +Id have trust you going ninja style on that one +Closed +sorry Doug +Thanks Robert! +This was fixed in the most recent JBTJBDS in. +Ill fix it in the layer +Cool thanks + +I could see where it would be nice for the component to configure the topic permissions during the creation of the topic +At the moment I am thinking that we would have an extra operation redundancy table alongside the current replication logs which would have two columns one being the CSN of a new operation the other being the CSN of an older operation that is made redundant by the new one +I forgot this was a significant patch so I did not add it back to the banch that is in +Sometimes this doesnt work smooth but happens quite often + +Excluded other platforms at r +I recommend closing this bug as WONTFIX + +Codec can choose to look at and pull the right DVF +Please let me know if this doesnt fix the issue and I will reopen it +Verified on. +Daniel please look at failed tests +It also updates the jquery version to +We can use the default value in those cases +ScreencastCan you still reproduce the problem? am closing this issue because the bug is no longer reproducibleTested in JBDS +Chris? Stefan? Owen +Commit alongside a test case +true +And in my initial patch as well +This would allow aggregating to be developed and used with distributed search +I dont see any point why we should return null +However the branch does not compile after a recent trunk merge +Emmanuel Assigning this to you since youre driving most of the research behind this at the moment +Otherwise it cannot be compiled +Were seeing this behaviour on as well although you have to sendreceive about messages before it starts to show up +Used instead which seems to do no harm for now +Anyway moving to is on the roadmap along with correcting these types of assumption issues listed above +Thanks Andy for quick answer +Also information like username priority etc can be passed using job profile +Canceling patch pending addressing Juliens comments +The image loader is not thread safe either +Thanks for the patch Jon + the revision r + +Looks good +I am not planning on putting any build related automation work as they may be very time consuming right now we will build it manually at release time and include in any kits +everything is working fine on Windows XPIt should be sufficient to set the host type to build host CC and CXX is auto detected if cross compiler is installed correctlystrip is usually done on make install in that case we probably have to do it from script +This is not a problem in trunk anymore +I am using the follwoing +The problem with this is all checks that compare objects for equality will fail in enumsAh you are right +I have committed it to Rev +Are there any places that are served by jettys file servlet? I think the user might be able to do something like drop a symlink into their log directory which points to the TTs keytab or another users tokens +Fixed as suggested +If unchecked when the user on a node it opens the ESB editor with that particular component selected +Previous Solr versions would silently ignore lib directories that didnt exist and the ignored entry at the time had an expletive in it +appliedOleg +fixing this issue would probably fix as well as +In essence this is no different than the RM restart today +Raise priority +Committed revision +batch transition resolvedfixed bugs to closedfixed +HelloThis ticket was left on hold due to the process of graduation from Apache Flex incubator to be a projectNow that we have achieved the status we would pick up and be able to undertake the process of migrating from SVN to GIT that have been pending for several monthsI look forward to know what steps we have to or we can helpThank you very much +verified on +I tried import Marek Goldmans blog from and it went smoothly except one tweak SBS didnt understand two tags which held some metadata information +It should link to the admin assigning to you since you are working it +Updated example showing that the provided workaround doesnt work for mixins +Attachment has been added with description Allows Group D file should be put in testimages +Shall I do it on followup jirabq +Thanks again +Let me see if I can reproduce thisThere are potentially three sources of exported packages ordinary bundle deployments packages provided by the system bundle packages that come from modules that are registered with the resolver +Thanks +Is fixed by this change? If so can you link it to this jira and mark that as closed +Committed as svn resolving +The pom is already providing the Felix plugin and the build already provides the required stuff please do a fresh checkout and try packaging the application the read the manifestThanks for contributing! in Apache Commons Digester release +Oups! Obviously I didnt tested the port well enough +This fix will be in the this into the next pending merge to merged +Steps are documented properly +Ive just committed this +Thanks Prasad +Just the unit tests Ill find some time this week to work on it then Ill resolve this issue +I just had a quick look at the patch and it seems fine to me though someone should definitely try it out for good +This would also be resolved by but this new code is probably clearer anyhow +the problem might be that the region is being marked as closed in meta and thus never eligible for reassignment +Deferred until. +Test passes locally +Also no real reason for having run t remember why I kept them in the toolbar might have been because of some problems with having the shortcuts working need to figure out +Last bits fixed by Emmanuel and were golden +I just committed this +Employ cache of Services data refreshed each SM publication interval +Committed revision in branch +Thx +This seems to be fixed by impl +I tried searching on for error set error error validationvalidation dynamic method all sorts of stuff but the only page that turned up in the search was Validation when I searched for validation +I started a test to check that for whole night +I added your name as one of the authorsof theThis will show up in the nightly build and in the releaseThanks! +Ill have to play around with this a bit but I do want to make sure that the current workaround is sufficient for you. +Seam and have with attribute +It also looks like it would be necessary to split the validation and persistence components into a separate jar so that one or the other can be included by application projects +Ive started working on this again +As well a new for was introduced to support custom and class lookup configuration. +OK +Moving back to for now +Committed to master. +Hi MarkAuthor is responsible for creation of patch whereas committer responsible for acceptance of patch so from my point of view they both responsible for quality of commit +I am a host using to agent +Uploaded to Marmotta Website SVN repository and thus formally contributed to the Apache Software Foundation. +The desire is to integrate the runConfigure decisions into the autoconf and automake toolset as was done with +Closed. +for committing this one +Easy to reproduce on my side +Even with two cascaded buffers the code seems quite complex +The Jenkins failure above doesnt seem to be related to this patch +Guillaume or NicolasCan you try Weis suggestion from above and let us know +The tool shows the first line of a multiline trace statement in the trace window and subsequent lines are shown in the detail window +I have tried to reproduce this behaviour with no luck +During setup Merlin looks for the keystore to loadThe kestore is specified using the property namein the signature or encryption property filePls check if the keystroe file is available and accessibleThanks +The patch has been applied. +Thanks. +Lokking at the scm url in your pom it would seem that Abhijit is missing the section userhostport from the scm url +Also JVMTI tests are not compiled with debug information any more and fail because of it +Excellent suggestion +Still doesnt work in +From what it looks like you are basically adding as another charset correctAt this point we support thre charsets public static final String KOIR public static final String ISO public static final String USASCII I am wondering if there is some sort of master list of charsets we should add +Marcelo thanks for digging in sounds good to meEdit Fix typo +The special char detokenizer rules are now commited +Please double check +Typical scenario user should not be doing this +Will modify it to simply write a known set of letters and cat the same thing out +Phabricator seems down for me + do have duplicated messages becouse they are the will be remove together with whole web subsystem in coming weeks +Do you know what SELECT statement triggers this corruptionThanks Oliver +bryan anything outstanding for preparing this that we need to track +And my other AOP definitions are domain specific as far as I can tell there is no reason why my Spring AOP definitions should be proxying an +bq +Attempting to queue the patch again +Mostly fixed except for hibernate is an relevant issue +The should simply do the right thing in this context and add racks for the nodes provided by the user if the user has already not done so +Theres currently versioned RPC over HTTP for Java in Avro +closing as incomplete which is about the most appropriate resolution +Murphys law I just released few minutes agoThank you for the fix yes not sure how I missed this one +When I ran it seemed that later iterations were running slower so I thought maybe the lobs were not getting garbage collected properly since they were not being freed so perhaps thrashing was causing the slow down +Oh and I think idea of a shutdown hook a good one +Or the Master can sanity check on its own +If you beleive that should be please and contact the program team +If the user explicitly set to jenkins we can detect that and blow upWe cannot delete the directory after a job finishes in local mode because more jobs could still be running +Done should be usable in a couple of hours. +Since this is really a corner case it does not really matter whether it gets into or not +This one should mimic the old though as the performance gains will be very beneficial to a lot of current users +Testcases are supplied. +Sorry the wrong JIRA was linked this is an old one +Definitely RSS for recent changes +code minor merge conflicts +I think we may close it now + +Damon We discussed this several weeks ago and the plan was to get the server upgraded then convert rules and the other do NOT want to do any major changes to the current repositories given the problems we ran into w this extraction +just open and hide the panel at demosite sample +Its the removal of that I suggest as a thought experiment +Note that we would have thought threshold being set to would be rare +As demonstrated by the profile this issue has been covered +I think you are supposed to keep the +Would you mind putting the unlock call behind a finally block? Technically its not really required +Moving milestone to sorry for the incovenience +This may cause spurious failures maybe the recommendations get better after a change! but best to fail when it shouldnt than the other way aroundOK will commit what I have now to get this in place since I think its good enough to go in SVN at this point +do you want to provide a patch +The new API FOC in trunk at least already seem to be qualifying the path +What is intended to be edited is the users local HOSTS file +I will try to run this test against Derby andor Pointbase +Agreed do you have any interest in doing this yourself? Wed love to have that sort of flexibility here +Use trunk for main development +I would imagine that a nonsuccess exit code would arrest the build process +If Im the only intellij user left I guess I should go back to just maintaining my own simple config again since it seems like Im getting blown out of the water every other week or soIm also still using +Closing resolved issues for already released versions +most containers arent quite there yet +patch does not compile with latest patch with latest trunk +Resolved. +We have optional but we do need user to specify a partitioning spec +is much cleaner and relies a bit more on JDK +Thats OK +We need to figure out how to integrate part of rev before +Can you build it from trunk and test it or can I send you a snapshot to see if you reproduce this error +This patch passes all unit tests and successfully runs the example +I missed cre +This has been resolvedJSVC has replaced the daemonize binary and the init script on RHEL based platforms has been fixed since its lacking a good LSB package +We can talk about maven publishing once the patch is in at least +I wrote the support for thriftclient +There is probably some other issue regarding version compatibility but I am not able to replicate the scenario +Method signature is changed to avoid dependency from classes in user bean +Merged into the master branch. +Change needs to be applied to +What is strange is that I dont see the dashboards in the node management I would have expected to see a node in there set as visible with no page attached +To julien niocheI missed your comment about speculative execution +Alternative standalone core bridge example +If I dont hear from them by Monday Ill and I can cc you on any further correspondence +I dont like that array solution very much +This patch only adds transient markers to the main class fieldsWhat about the nested classesSome of those have fields should they be serialised or not +This patch fixes the problem With OPWRITEBLOCK we also send a boolean +So I guess we could keep for and then get rid of it forHowever needs to be resolved +If this is a learner and the zxid of the client is higher than ours then we issue a SYNC to the leader +The only question is where to put the that the plugin installer depends on +This issue is fixed and released as part of release +Note that the URL mapping in question is +However somehow tomcatrun and package must be made to work to get JSON output for REST systems regards Eberhard I have made a change in commit bfdfddfabaddfdbf which allows me to access the mvc app via the root and the GWT app via +I will go ahead and commit the patch to backport the changes to the backport to with revision for has been backported to +I agree with Simon +I have also attached a sample configuration file. +No response assuming ok. +I am also interested in this issue and would be willing to contribute +The patch should be applied to. +synced to FUSE MR trunk syncs is to FUSE MR trunk . +mixing of the insertupdatedelete statements is now supported +Can you please resubmit the patch so it compiles with java as well +Scary +println At the time I reported this issue I used at the moment I use and the problem is still the sameAs IDE I use Eclipse Juno SR Groovy Eclipse Eclipse project setup is direct as jar in the classpath no usage of Eclipse Groovy LibrariesThanks for analysing +Except its called aThe is just there to handle a back compat issue its something that we would want to get around and use the reader valuesource for instead but that patch still had a long way to goOverall from what I can see the approach was about the samebq +Ive moved parsing of DEFINE and STORE to from and ensured it passes all unit tests +Right? The tests also seem to be strictly following their mapreduce counterparts which is good except for one thing There seems to be some difference in between and the new one in +can you provide a server thread dump when it freezes? +This was committed +can you verify this happens with a clean browser cache? +More report show this should be considered for final fix +The scm url is not readily available unless the brachBase parameter has been set +Add strict verification property and use it in tests and similar situations where config is constrained + +We could have better error reporting scheme for this operation +Kindly clarify the above +Yes thats will start implementing the geotools stuff and need the new hints in the Hints class +Please verify the fix in the latest +here is a new patch removing the file based store load methods and adding a fileName method to solrs factories to maintain compatibility +I found a way to make the +Added running total functionality +I will wait for a stable build before trying it in my production EclipseWell done! Thanks +An example would be a that wrote records to an external data store and thus might want to disable speculative task execution on the MR job that it was a part of but we wouldnt necessarily want to disable speculative task execution in other MR jobs that ran as part of the overall pipeline and did not involve that +Committed revision Thanks Otis +Hi Scott please dont create a JIRA ticket before having first discussed your issue on the Sonar mailing list as Im pretty sure that this issue relates to your environment +Can you please provide me some hints so that I dont spend lot of time going the wrong way? Your help much appreciated +Looks good to me +Fixed a typo in the patch otherwise it was fine. +I will run the tests and let you know. +Attachment has been added with description This one has been added with description This one doesnt in FOP and probably earlier transition resolvedfixed bugs to closedfixed +The good news is that the change worked on the REST WAR project +Havent found any awareness of this problem or a workaround +The problem is due to having old contacts that were not properly cleaned up +Updated patch looked good to me has been committed thanks for the hard work +However I believe that was using GEOTOOLSDEVELOPER logging level +Committed +StefanBy coincidence this issue was just brought up by another user a few days ago +And the feature is already flexibleparametrizable enough to allow more or missing if needed +To be portable however applications should always provide these values for and REF parameters +Marking as done as per Rameshs comment +You have a tip for meKind Regards TomI followed your instructions and built the snapshot and then used it in the Maven dependency +Also fixed as a part of this +If you still have any concerns or if you think that the information given is not sufficient feel free to this issue at any time. +Is there anyone deploying these snapshots to the snapshot repository +Ive added support for throwing an upon allocation error +This looks good to me +Have you tried explicitly telling Derby that the database should only be accessed in mode? You should be able to do that by setting the following Derby property readThanks +Fix enumeration safety state inconsistence in file +the version of being shipped with the core patch +Any progress on this one +By the way Derby seems to simply ignored user name and password +I just committed this +this time the lines are diff slightly off from my src do you have any custom changes in those couple classes +Its fixed! +all tickets +quick top of my head commentsCreating yet another pluginmodule system sucks Using classifier of maven artifacts would be a misuse IMO could we index it someway elseConsider using something like P from to handle the installationautomatic updates +bq +Related with locale I can translate all the labels of it fast +Ive rebased this patch and fixed a typo error +I dont know how to fix it myselfThis comment is part of an effort to close old issues that I have reported +The only problem with doing it quickly is we have an internal repo set up with some custom stuff in it so duplicating an exact project dependencies wise will be a little tricky +I should probably point out that while remains unrectified the in loader will not prevent a memory leak if the application uses in a container that creates and destroys that extend during the lifecycle of the application +I would attach any files if requiredThanks +So closing this issue now. +This is an upstream issue +The class comments have been updated to reflect the changed image names. +Any news +was never really intended to function automatically across VM boundaries +Wouldnt that work even today? Do we expect anything more +Updated About image +Updated the patch to successfully apply against the latest codebase +I was using the Boxed classes rather than primitives +Since the name of the file needs to be parsed in the current design a delimiter underscore was chosen +And again if you can come up with a couple simple test cases I would really appreciate it +So concurrent users acting on the same controller may potentially share data. +Simon watch out for we have to be careful when labeling to not use the label with a trailing comma stuck onIe this issue now has two such labels gosc and mentor +Im running in this issue as well +Are you really expecting to have more than rows in any given tableWith that said Im a big fan of consistency and if HQL is returning a different type than Criteria then I could see the case for that +Justin +Resolving as a duplicate of which has a lot more discussion about how best to do this. +Sounds good Matthew +I will have a look at it today Jimmy +There is a bit of a red herring here I realized this when I was looking over our public and private conversations you have a whole pile of related switches setIt turns out to be cache off because what happens is that with no caching its possible for the VM namespace to get dumped during the parse of one thread while another thread is the AST +FWIW Im an Apache committer with a contributor agreement on file +We need to develop some persistence based examples at some point in order to exercise this area +fine thanks for the with +Seems to be no longer an issue. +All the functions already has aliases +m not following what youre saying here + +parameter aliases did not work so I added the new Configure and configure parameters and deprecated the old ones. +Ive done some of the work the file is now generated from the file of plugins in the core and whiteboard plugin directories +Attaching which does just what you said +No assignee and no recent movement +Hi KiranThank you for your helpI tired the below solutions removing space from and some codechanges based on the isHR checkIt does not work please let me know if you can of anything elseYou dont have to try it will check itRegardsAnusha +In the process of testing these defects it also became apparent that a few other changes needed to be made to support the new property types +Rails use of signalsTERM is used in Transactions to raise a as a way to shutdown Rails after it is detached from the console +Patch was applied as expected + to simons assessment and proposed solutionunless of course weve both missed something +Ill get it in as soon as its released +I think theres still a race +I think this should be applied as long as it is only used as a fallback right? This seems like a good solution to avoiding hangs at the least +already tracks this issue. +We need to do more investigation on this issueThe patch enforces that each datanode belongs to a rack +Yes + +Sounds like the test case you have here relates to some of the scenarios described by some of testplans cases +. +I was reading the issue in my JIRA client yesterday and somehow missed all the info you already gave +This is the write part of RDFJSON support +Try again +Patch of which waits until the local Brokers ID is known before transmitting the to the remote transport +Whats the mechanism for preventing bitrot of these tests? +That is a reason why I am interested in building micro instance +A script that benchmarks system and smartsplitbenchmark on +We should have some kind of registration system +Majority quorums trivially intersect I dont think the failure in the test is due to this patch +bulk assign all unscheduled issues to next release +Added txt document showing example session +maybe you can check it out HBASEThis patch seems to add a lot of logging but we can turn that down laterRyan does this patch resolveThanks for the sweet patch Ryan +Simpler things come to mind after sleepingFixed. +svn +Changes committed to trunk revision +Sounds like a relatively good way of having notifications when artifacts are being added to be the repo +I dont think it would make sense to choose utcrandom unless you had already planned to keep your clocks pretty accurately synced +Hi AlexIf you search for chukwa against the Jira website youll see a list ofpatches that we want to commit but were depending on external Apachecommitters to get them committedJerome. +And there should be a trace +I noticed this problem when giving the geoscript class and changed the way it is executed +we have the capability to fix that and release that fix on a schedule which suits the Hadoop team +running precommit now should commit shortly +At first glance that incorporates several nodes into a single view shouldnt be problematic unless the cache is filled up with the nodes in view after the view id is set +Yes I think RAM is not that costly these days +GIT version is our current one +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +Not quite the right place for the discussion here but probably one that we should have +Yeah there is unfortunately nothing in the Commons FTP API for keep aliveKinda hate to have to do a pool in Camel to periodically keep the connections alive +I am not seeing the error after applying just committed this +Cool +Same stat file +Forgot to add we are using the with a temp in the +Please keep this issue open +Sorry for not mentioning this early +the values in the list of related entities are unsorted unless my patch is applied +thanks! +Thanks +I dont think we can do much about it since it is related to Ant classloading me the only thing to do is to add a note in the documentation this is the patch Im attaching +Thanks for the patch +We will be testing this and the tomorrow against a few thousand documents to see how it impacts our unique term counts +Please change title of jira to refer to this specific optimization so we dont have local mode is slow tickets as we keep chipping away at this +As a please do get the sources from the tag of the SVN repo. +I believe that this was already fixed +We are working to remove it +When adressing this can be added as well +Do you mind suggesting how I can test with Jetty ? All builds of the http service seem to be using Jetty +Looks a dot is missing after insteadis run instead JON also mistakenly +However rather than changing the code I think we should just not right the random seed to the directory to begin with and instead right them somewhere else +We do have some checking in start +I havent been able to reproduce this issue +This has been resoled in the Fulcrum Parser Service +Think need a day or two to fix that +our existing isnt really setup for testing config failures in cloud modeneed to look into that a bit revision +export vmhome +When I click the pencil and nothing happens Firefox error console saysError b is nullSource File httplocalhostLine +At revision I added log warnings for accessing sessions that have been idle for more than a minute and that are not configured to never refresh +P +they dont use any of the params they are given so no reason to waste the cycles wrapping the defaultsjust in case though i added a comment about the defaults in case those methods ever do start using defaultsCommitted revision +Fixed in close after is released +Found when testing using RF that there is a chance that the local node could be included in the workMap and be the first node ordered by proximity +It would be simpler to rebase this on that. +to best of my knowledge I have covered all the tests that old test covers +s descriptor file + +The previous patch did a null check after the error would have happened typoAnyway here is the correct Steve for finding the problem I just modified it +Multiple cases of images that are too unclear to be read +Thanks it to master branch +My badThe patch is trivial and just adds a experimental config variables and was around for a while before I pushed it to make +This was refactored already as part of +trunk branches on JBT +Seem uses to get the channel decriptor in and I suppose it doesnt work with directories +On an after thought I think that these context maps could be useful when we want to support sessions +Closing this issue as the initially planned work has been completedTesting can always be improved and this work should continue outside of this Jira +Thanks +Running and still seeing this issue +Could you give us the message headers for the message which hit this +Feature which is better implemented in flatpack +WDYTThanksImesh +smile +Looks like a WMS test case +OK just submitted a fix for this I can you check if it looks can you see if this fixes the Swing scenarios +do you guys know what is a proper the number of handlers +Attached patch removed the variable from +Will check mojarra has this for validator behaviour is not there for +bq +My only concern with setting the capacity to would be it could be possible to create a queue where a user would never be able to run anything even if it is just to drain the queue of apps currently running +I have written in the meantime a number of its but with other dependencies it might be a good idea to take a deeper look into the mrm plugin which can be preloaded with those dependencies for the integration test +This happens pretty much everytime somebody updates the libraries +I agree last note though what I put in my last patch is much better than what we have in trunk currenlty +Last time I looked Im pretty sure an email address of all spaces is considered valid +fixed +Now I can reproduce your original issue where resumed jirb does not echo characters +This has been committed. +agreed +yeah tried with ant from cmd line and that test works with ant +Also increasing the default buffer size can have a significant impact on memory usage since some applications open lots of streams and each open stream frequently has several buffers +No links at all +I think we should fix that so that even if recovery isnt supported we atleast parse and get the previous AM attempt info +So now if the test case doesnt find the text it prints a warning and moves on +Ive tried all the options given in the related JIRA issues. +Then this would be a serious bug as two jobs would be overwriting each others parameters continuouslyReally there are two separable concerns here Representing one canopy Implementing the canopy algorithmThese are bundled together into one class with the static stuff representing the algorithm and the stuff representing the canopy +Thus sooner or later there will be only one store remainAnd there are various other situation +Thanks Milind! +This is exactly what the javap tool does +Attached a patch for where an assignment made to a local variable that is never used has been changed to just call the method without assigning the result +This looks like quite a positive change at the macro and micro level +At least that is my understanding +Just before the exception theres alsoWARN Problems unmarshalling remote command from byte bufferDont know if related +we need this addressed for release is frozen + looks good +Ill try your suggestion and get back to youI took your suggestion and it worked perfectly +Please feel free to review this +Note that did have that same +The CR quickstarts work fine on my XP s strange +docsupdated +The property must be accessed in a safe mode like thisvar input ? See patch above +But with some thought we could eliminate nearly all the integration tests we run under the python harness +there is nothing like thati cannot find the on the agent only on the management serveri installed from debsis the deb installed for the iso +So whilethis patch addresses the serialization of configuration items it seems there isstill work to do with the validator +Karl you are the only qualified person to look into it +defer all issues to +Any router that can return multiple channels should actually have the configuration option +a more complete you please review the latest patch and accept it if it fixes the problem? Thanks +You can run the tests in your own workspace via running something similar toant homedhrubaforrest homedhrubafindbugs RESULTS usrlocalbinsvn bingrep usrbinpatch usrlocal you please run at least the target which is a test rather than just oneIn general you should run in order to have higher confidence +BULK EDIT These issues are open to be picked up +wait a couple of minutes +This seems like an awful lot of code for something thats or lines of code if is available from JDK +Verified in ER +This is a duplicate of aka +You mean I should commit them to the branch +target youI will inform you of my +I can try it +Just to make sure have you fixed namespace and namespace also +is done in this patch +Make a to backport it +see that works +I think the fix you suggested is a pretty good one and probably the best course for now +I think some users might make use of this type system because it is part of the UIMA SDK +Please feel free to ask if there is more to be clarified +done as requested granted the required perms as wellThanks Musachy +Hi Jos +Thank you for your contribution! +On the branch there are teststesttest in TRUNK and backported to SP as well. +resolving because erik said its ok +It should be noted that it makes sense that test took longer with the fix +New patch which Sreekanth and I worked on together +this issue resolves many nunit tests +Looks like a plugin is the solution here +Is this completed at this stage? I think it has been but just double checking +GertThanks for applying this patch and I think the patch should also be applied to servicemix branchBecause this feature is important to Servicemix it will make Camel Context to be recoverable from deploy failure or JMX management stop operation +We can clearly document this +Aggregated patch that includes the code changes as well as the test for the xmlctf suite as requestet. +Blocker for release +The group I work for deals with cloudbase not accumulo so it would still be very beneficial to get it working +Do we need to do anything special when deploying to Nexus +Most of the XML files Ive seen that run on a server are documented internally with comments before sample statements that a user can uncomment and use as a template +I suppressed one warning concerning template redefinition which is rather VC error than Harmony problem +If the proxydoes not support HTTP it should have returned an appropriate status +Only issues that are actually fixed should have a fix release. +yes exactly +Nick thanks for keeping a watchful eye out +Id like to move a bunch of these methods into a new class +I think we can resolve this now? is open to address on next POI upgrade +I was going to open a new issue for composite subtype matching but found this issue filed a while back +hpp +We might need an additional JIRA module for this +Ill kick the build again I think Giri fixed something this morning that was breaking these links +But it can be done in The Security Auditing feature is not related to the Managed mode auditing +But year Oracle forbid half of the English dictionary youd better switch to French +If you arent I can give it a try myself or ask Piergiorgio if he wants to have a look at itHave you given any thought as to how best to test this connector? Is there any hope of having the tests bring up a standalone instance which exists only for the life of the test before it is shut down +See also for a possible duplicate issue +Since the will probably be replaced by working on a finer is not the first priority +This issue is related to because both of them relate to the phase at which we create files assuming there will be a web application emitted +I think we should wait until hbase actually publishes those hadoop binaries before considering this patch for commit +We should also have a flag for disabling type generation so that a developer can use hisher own type generation +Change the comparison logic a little bit +I will have a look ASAP +With the closesocket instead of the nightly quick crashes trying to close the socket after its impl has been freedTest case Start broker summary count +Would be cool if you could report back if you still see issues with a current snapshot build +I agree that we need to add test cases to test themI will add a test case +I am pushing Alt Shiftand Caps Lock to as these issues are similar to the issues inAlso resolving the ticket + +It does not exist in is a blocker for our project to upgrade from +bq +Do we have any guidelines for thisIn general when I use libraries I dont care about how they do their things as long as they do it fast and correctThat said I really think we should keep the implementationalgorithm description Im just not sure what is the right placeformat +Caused by table or view does not exist since you created the schema objects for user scott manually you might have done something wrong +However it will not cancel a project build that is still doing scm update it will just skip it. +Thanks. +Removing the entry still lets me compile the whole project in the IDE without a compilation error +Adding link to older duplicate Jira issue. +issue has been fixed and released as part of release +I only did use as affect version since it isnt released +Was not interpolating systemvars +bq +David the links and info seems to be accurate the only thing that should be fixed is the unsubsribe email address for the user mailing list it should probably be without the s in users. +I dont think my patch is causing this +java and +moving to M + for guess a desirable solution has been found. +So thanks for confirming +According to Zhengs suggestion removing the version number in and get it during ant and passing it to Test +Its not complete testcase but it create required JAXB objectUnit test is in the bundle It creates JAXB object by loading classes from two different bundles and The classloader of the second bundle able to load all classesbut the classloader of root class located in the first bundle and unable to see classes located in the second bundleThat is the problemIs you will try to pass object through SCA it will cause error mentioned +bq +More details are added for the various algorithms +Folks could then run the target from within a release if they want to build a tarball +These will be in the next release +Lauras suggestion of using an Attention note works perfectly +jar changes +Im looking into it right now +This patch makes the implementation of listFiles in is the same as the one in except that it does not handle symbolic links +Regardless of how we achieve it we have many classes that arent very extensible due to their use of private instance variables and we should consider how to resolve that problem +HiAbout we have to discuss it on the dev list +Where should I start +Closed upon release of. +I have committed this. +With this patch the split will succeed at customized point even if there is no store files in the regionIn short the split decision should be made by upper level which knows more about application logic instead of low level +The bug would be that the reset gets applied not only to documents that actually failed but also to documents that had been already handed to active worker threads +note the change looks good to me just trying to get docs on concrete usecase for why we need ve added livereload as possible affected projects since it also uses jetty +A potential work around would be to throw in an extra ignored parameter with a constantly changing value to prevent the inappropriate cache hit however this does not work since returns a queryparseerror for the extra parameterI have not observed other queries accepting extra parameters either with or without an ignoreextra parameter +Robert very nice +So it s transparentDIGY +Updated with new ruby module name +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Thishas the added advantage of providing the user their incorrect input +It will be complete by Sat a simple addition of the appropriate forEach applied with slight modifications +Ill be honest I havent done a thorough review of the code in the patch and only have had time to review the functionality +looks good anirban +The comments here might prove useful if we revisit any of these questions for a future version. + +What about if the bundle transitively imports any package that is included in the superset of exported packages in the dp + +Thanks for the logs! Are you able to access any other services? Can you point out exactly which line causes the crash +May be this idea is not correct +It add example in the case of the is obtained from an http serverBut is different +Nice +As such it seems that it should be shutdown to prevent new Messages as part of the same step that shuts down the pollers +Is this still happening with the latest code +Common part Make class pluggable +Checked into and. +At the moment I have removed the getId method in the interface and introduce a little bit more reflection to get this value +Note When the cluster is initially created all ZK nodes should be cleared +We can always move things around if we need to +seems more about setting than getting and is only for I do anything to have this patch applied +this should work better and NOT contain the confusing zh Deleting old patches isnt necessary Jira grays out the older ones and highlights the newest one +Hi OndraI think we should raise an Arquillian issue rather than the workaround is that possible? It should be quite straight forward to create a simple Arquillian test that just starts the server to replicate this +Mikio has done a nice job of this I think we have enough now that the issue can be closed +After applying this patch it is possible to generate the also +Do you mind linking in the JIRA that addresses this issue on trunk +ELOSflush calls ELFOSflush which does a force on the underlying file channel +ve just committed this to trunk andThanks a lot for the reviews Eli. +Additionally the for supports locking seems to be redundant +If the behavior of reconfiguration is defined by the implementation then can we have more comments with this patch +Patch to branchPatch applied +Verified. +This is a log of the git repo just before yesterdays push +May be in the future if the OS and JVM support some real async disk IO performance could improve +Task Killing patch incorporating Ravis comments +Since it works upstream we should make sure the product build doesnt have a regression +Its an essential feature I think +Implementation and unit test for passive aggressive +is having problem it didnt run test suiteJinchaoPlease also add the timeout parameter as Jonathan suggested +Thanks a lot for the explanation +due in Dec for AS Beta BETA due Jan for AS BETA GA due by Feb for AS CR. +Got sidetracked and confused +Also it can evolve at its own pace +Original issue fixed. +AmarThxs for the clarification +Maybe if the artifact is downloaded by Maven Maven skips updating due to some incompatibility +Attached updated test program +please open new issues for new thingsimprovement instead of on this issue please +This is a patch to that causes it to import all documents defined in the config file +sameer solve the problem of detecting replicas from deleted filesIs that actually a problem? Well only generate such garbage blocks when nodes fail and only a few then so it doesnt happen at that high of a rate and the current blockreport mechanism handles that +batch transition resolvedfixed bugs to closedfixed +Patch handling other invalid javascript which handles other invalid characters as patch with fix to catch more bad variables +Hi CamSorry for the have uploaded the patch please review it and let me know the modification if any? Kavita Patch looks good to me thanks! I like the test particularly + +Committed a fix for this + to more clearly describe the precise feature we are looking for in the new build issue description more precise +Error doesnt seem to appear anymore +Been trying to get to it all day but looks like itll be tomorrow now +Moving this no way well have this ready in time for this will be released +to this one as just committed this thanks Ari. +Attaching to addresses Mamta and Dans suggestions on the testcase +Hmm well that patch passes tests but up on cluster it does weird stuff +I was out on vacation recently and just missed these updates +All tests are passing for meIve made some mostly minor changes as well as Removed and the Router workaround for it +all tests was not cced when Justin made the comment so added him as a watcher and everyone about the patch +Thanks Ryan! Committed +Now the JSFSeam portlet facet offers portletbridge from EPP when the user selects EPP portlet JSF portlet and Seam portlet can be created using default portlet has to be created using the New Seam Web Project portlet and JSF portlet are being created using the New Dynamic Web Project wizard +Lets do this fix along with new file format +Many are coming from trove hibernate +Make take a as input +Closed upon release of. +Thank you +Yeah Ive got a version that handles the user case +The message has been updated now and in ER will read BIRT service has not been activated +You are absolutely right +Any update Phil +So if you can reproduce this issue after running update project configuration please provide test project that we can use to reproduce it +I will be on vacation for the next week so I wanted to attach the current state of the patch incase someone else sees my mistake causing the extra memory to be used +This broke the ability to run HDFS out of checked out source trees +Reviewed and applied the patch +committed to enabling support on existing project file is added to +Each proxy has a different token so theyre cached separately at the IPC layer +bq +Is there any chance we dont have to change the directory name from that used by karaf? Perhaps if the existing geronimo hot deployer ignored anything it found that was a bundle both deployers could work off the same directory +The final solution was a little bit different in that I added new coercions for each of the specific primitive array types rather than bolting the funtionality onto the Object List coercion +This was applied to job in view too +So all issues not directly assigned to an AS release are being closed +Second test case where first fix wont work +Who knows anything about it +this still the case + +You can unzip this into tuscanysca folder +OK +Thanks for your quick replyI can not say if it is a great amount of work +Setting version to for now fix version will change once is firmed up +rev and rev Do not allow creation of by activemq and should resolve pathuri relative to GERONIMOBASE +Yes I will provide you with such but need to handle some higher priority stuff first +Do you agree ThejasAgreed +On the other hand since this is now DDL and no longer DML I think its not a big deal +It only simulates its overhead +Hi HowardI realize that I am replying several months later +I can commit it today +Yash has tried but was unable to reproduce the problem + +Here are a few comments Please add documentation +Heres the patch that fixes regression for Hadoop +I also see a couple questions per week pertaining to HA Federation or other features +I would like to get more information on this project +Only relevant to trunk +The spec only says for each Bean another must exist +Do we need to fix Windows create script +the stack trace from the client that file for the slave file for the master capture of the log from the master +Patch checked in +smile + stale resolved issues +Im working on a solution +Resolving the issue for now +The only thing I havent worked out in the test is how to get a remote instance started +Cheers +Is part of the public API +Patch has been submitted to review this record since is now released. +Unit Test +It can be enabled by setting the system property to true +rev +Uploading a patch for adding a parameter map field for database +Python translation sounds easier anyway +patch to move the logj initialization into the real entry point rebase but +A minor priority seems completely inappropriate here at least if you want to use Maven at a enterprise level +The change has been made when we were working on +Closing wont fix +war archive containing the endpoint used to reproduce the by Undertow upgrade to Alpha +So far I have checked only zookeeper hadoop hbase and phoenix packages +Thanks a lot for the quick review! +Committed in rev +Quick patch not run tests on it yet +This results in the creation of a new workspace instead of using the default workspace +I have committed the backport patch to +Yonas patch improves the creation of images with tranparency within a pdf +Looking good +Patch applied +Applied second patch from Alex and Andrew. +after all InputsOutputs have been initialized +No new tests required since test testcase covers it +Beat please open new ticket with any new quickfix enhacements you may haveWe want to close this ticket so we have a record for the change in the camel all resolved tickets from or older +This problem was already fixed +assigning to rob since theres a question for made +some additional info for the issue there are Controller service +Agreed +links to fix in cdk are fixed in cdk with links to fix in migration is attached +Verify startup shutdown of the cache and a simple put case for got rid of cache loader for security +The issue was tested on tag +If ones cache is not updated it fails as above +I hadnt realized that the final spec was released and that different from CR +The presumed issue is not happening on on though as this patch is absent. + Bulk move to that selection of connection profile is not accessible in seam facet wizard page when Java Persistance facet installed. +Could you please upload a patch to +If a datanode really becomes the dfs client is able to detect the problemThe issue here is that the test simulates a block receiver +Wouldnt that be a bit ugly in +Needs testing +Its actually closing the connection +How are you authenticating to the application? Are you modifying the authorities anywhere? Can you post your security configuration? Also if you have any custom code around authentication can you post it? In particular interest would be how you are creating the Authentication object +Assigning you since you probably have best notion of the provenance of this script +At this moment Im not sure those happened for the same reason +redirect viewid for an exception by dragging not implemented yet. +Hints for Preferences Table Style when hovering over table heading code delivered +Things to look into would be undoing reflection though this seems to be relatively inexpensive going by recent profilings and it makes extending RPC simple and wed look at adding to the clientside and sending big data in chunks +I do not think we should abandon older versions of Hadoop for the newest version which is not even complete yet +how did you get around it without needing to modify +patch updates alias map and docs +I agree that redirect after post is a standard behavior for SWF but SWF itself is not a standardThat said if all that needs to happen is for us to return a null I have no issues with it +Is there an additional step to do? an other question is there a way to get the component into the release +I wonder why not do that? The other alternative would be to use application which would be handled by the +Thats why I kept it in here +So Id like to see this developed in a branch +Ill do plz advise +insertCSS looks to be relatively straigtforward to implement executeScript will probably be more difficult as it requires an isolated JS execution environment +for shouldcould add a warning but not really easy to detect IMO. +Also to see the stacktrace run the development mode command when the Roo shell AlanI was thinking about it yesterday and yes it is not a bug +I dont have the means to test the patch but I assume it works as intended +In there there is some standardization around this +Closed. +Fixed in eaecabedaeeabe + relates to Kerberos support for the +If this is critical for you you can always ask for our paid support +Could you perhaps attach a small test which produces the problem for you that I can try? The sooner the better as were starting the release cycle soon and if theres an issue wed like to find it before goes finalThanksGlen +Still need requiresPayloads boolean +consistency is better removing all properties with empty strings is ok then if they are not mandatory and if they dont have a defaultValue assigned +When this JNDI is configured in WAS all I enter is cwpServices and WAS adds the beginning prefix s odd is that clearly Spring is finding the reference from JNDI but it seems odd that it is thinking the stub file is incorrect +Some small improvements +patches welcome +done in r +Good point on the other environments +What does hadoop do for baseexample task classes? Does this patch work w h and h? The keys and values have to be serializable +Editorial pass over all release notes prior to publication of. +If I were in a worse mood I would argue in favor of not setting the defaults in the failure branch of because of cognitive ease of understanding but its not something Im gonna fight for at this pointOh you mean the macro as a dependency +Verified in CR +Applied patch in revision +Any chance you could this against Eric +Applied to and to trunk +t this solved in +basically the credit pot starts at MAXINT and for every n messages sent the pot is decreased by the window size +Currently handles this in +Wasnt there something that version of the enforcer plugin didnt cause this warning +Implemented persistence mechanism which allows projection federated node mappings to be stored and retrieved from the system the into the federation branch. +In one qt I have a long list of qf fieldsand in the other qt I have a different list of qf fieldsI separate the two do that the amount of incoming bandwidthis reduced +This implementation is based on a generic JMX which is available in the module +Another test just failed in same way +Eli can you update the patch and? Thanks +Perhaps the simplest possible example of a strict mode web application that fails is the place to begin +bq +this isnt going to change anything but applied patch seems to resolve this issue +I suspect the latter but am open to opinions while I research how the frame size is negotiated +verified to set the Fix Versions field to JDO beta. +Pull request sent +It was based on revision +OK then Ill take a closer look +Not sure Ill have any time to look into this soon +Ideally the wsdlj object graph should not be part of our meta this is fixed synchronising access to the wsdl is ok +That would be very nice +Closing all resolvedfixed issues of already released versions of Roller. +bq +Out of curiosity why do the tests appear to except trailing whitespace in the outputIm not sure of the original motivation for that +but in two stages unfortunately +bulk close of resolved issues. +bulk defer of open issues to +You must be using an old version of since this message is physically removed from the code in the most recent versions +Could you please add io session and context logging as described in? Feel free to tweak and add more detailed logs including SSL engine status or wrap unwrap operation resultsOleg +Im not convinced there is anything Struts is doing wrong here as it seems to be an issue with IE and how you are configuring the form tag +This is how it looksI havent figured out a reliable way to reproduce this every time +I hope this helps +Lets use your ultra cool hudson build to test this +Thanks again for the patch MikeIve applied and tested it works fine so committed now as well. +This change was overriden by. +Would appreciate it a lot if any of you can help in this regard +ant Error with the that I did remove the line from file as per Kevins suggestion +This patch does not solve that problem at all +fixed in +sure i will also take a look +We can revisit debate about other Java escape sequences when we work on version. +Updated patch +I think we should get rid of the casting to +With fix for and VM side patch to call Agent before killing threads all tests dont crash any more on Linux x! Committed and verified at . +Ill create a new issue to solve the layout question +svn diff file and the John H +Just scan META and return false as soon as it finds a region that is still onlined +In particular Hibernates creation of CGLIB proxies takes a very significant amount of time at are some ways to speed up Hibernates startup check out the Hibernate docs and forums +So is supposed to check that the database name is correct if the DB client is used as well? Have you tested that it works with the DB client +Another approach would be to come up with an event driven way of noticing the schema arrive and then set the status to normal +Designing optimal withoutusability testing is difficult to impossible +See. +No cycles foundOnly other potential possibility is that each transaction requires so many locks that it is almost impossible to get through it without running into deadlock so it just makes no progress +Thanks for the for fixing +Id really like to wait to apply this until we see an Spring Security M release +Thanks for looking at my patch Manish +Verified passes now +I think you fixed this +Barring objections Ill commit this tomorrow +true does not work +linking to that issue +I should have removed Endian so people wouldnt be tempted to use it but forgot +You would have to move the to a shared location like the conf directory +Please log the error +It would be a performance enhancement and youd need to maintain backward compatibility for older data +Personally I will prefer to the second one and the third oneAny comment? Thanks +All my sources belong u +After running for hours no failures so id consider this last fix to be good +reopen to change fix target this shouldnt be showing in the release notes. +Hymmm the semantics of this could be argued either wayI can see arguments for bothFor sure we can not change the behavior within a release without a fair bit of consideration +Hows this? Its just a small patch that shows the boo version its running with and some of the global commands +Actually I was not much familiar with this system and I was in doubt about itThanks for your help Lars. +Great suggestion Chris and thanks for reviewingI have updated the patch +I ll update if I see it again +Changes have already been applied in svn. +A small Eclipse testproject demonstrating the resource injection that is not workingthe logging output is also attachedI didnt attach any libs because of the +This error happens while editing codeMaybe its related to my constant refactoring +Committed in rev +Fixed by commit b +If thats too much trouble it may be easier to just manually add the element to the descriptor each time you have to the generatorId recommend compiling against the latest final release of RHQ +I am happy to add this feature in as well +Attaching a proper patch sorry not used to jira +The one you want to review has the blue colored link and is K +We need to research what we can do for specific JPA providers there but generally speaking you wont be able to rely on custom isolation levels being available when you use JPA does not support custom isolation levels because the JTA specifies it +Im willing to work on itCan someone give me some pointers on where to start? Ill start once I get the go ahead +The patch looks fine to me +I will reopen the issue now +Patch with proposed fix for the issue +This patch file also includes the enhancement to the help files in the locale EN +And with it may even soon become a no problem so its likely not worth the complication +After applying this patch I will provide the patch that enables the user to define his password custom policy +Workaround confirmed +Here updated patch +Users should only be using the public interfaces like etc +This did make it into upstream +Directory structure has been created but the translation isnt done +Committed to branch +MMM is default in some locales so should be supported +when you set underlines text to the first word +This does not look like it is specific to I have other dependencies and am experiencing the same issue +In order to cherrypick this in the EAP ER build we would need a resolution by Febth +In the future we may consider the overall performance of all tasks it has run so far and make some decisions based on that for this special caseI meant tasks across all jobs +Patches with test case for this problem +How about thisIf the file exists in the same directory as the it is used +Hi Amichaisomehow when I apply the patch it shows only one line changed +Alexander if the release doesnt do the trick for ya +It looks good Ive one query its currently using and from the package is there a real need to do that or any way it could be done using classes? We try to avoid using the sun +Then we map the MM query model to the normal hbase client api. +The most important one is that the version will not affect those sets as the pass false to ingoreCase and we have full control over the stopwords +The table entries seem to be repeated times! I will attached the generated file + to trunk +Its on the list and pretty high priority I just have to get the time when were not right up against a release +This issue is closed now +Phabricator seemed gone +any chance we get a test before the release? days from now. +Ming what does that fix it +Frankly I dont really know if that is the right thing to do but I assume we can go with is it that you dont know whether is right to do? I am curious +Its OK +Added patch with new component +With set to false it is notified immediately after the failure +Is this beeing resolved? I have exactly this problem as Kenney stated on Nov +Tested this by changing and returning the hostname to localhost times each +wondering what changed +The second patch was applied at r +Is there a way I can set a threshold and make the maven build fail if the coverage rate is under that threshold? Thank you in advance. +Thanks for explanation +This adds property setting support for Sun on Solaris +Moving outI checked the issue and was able to repro the issue once but not alwaysThere are two issues delete using retry count new zk client instance created during master abort may not be seen by other threads due to no volatile declaration Attached patch including refactoring code to handle ZK session expired consistently in all zk async callback functions as we currently do in async callbacks retry deletion in where other maintenance work are done +Only if you want to I can make these changes upon commmit +As I see it Seam still uses the older richfaces and aj jars afaik! The tooling should work with I miss +I tryed to implement it but calculate the actual size of received messages is uselessly complex and slow while here perfomance are really neededIll reduce the minutes timeout and increase server memoryUPDATE ok i see now you last comment +If affectedKeys belonged to the local transaction then assuming the local transaction was found fine the code would work +I believe that was my mistakeIt seems ok to me +I would also vastly prefer that we make such an incompatible change once rather than piecemeal providing a complete fix for +Ill reply as well +I will go through your logs +Sorry for that wrong bug +Is that necessary? If so perhaps add a method that allows you to pass an attempts value +I would pass the result set as parameter from executeQuery to processData method or merge both methods into only reason to have a global counter is that we need to pass the inital index between the bind methods +An contribution that is not needed by DRLVM at the moment +I am desperately seeking the solutionThx for all the hard work +I moved metrics about getputdeleteicv into a helper class that mutates data in +Do you think this is still neededAaron +What are we trying to solve here? What feature do we need? Or what bug are we fixing +also updated +memcache may be an option +Deployed snapshots +Also test cases for the utility function using a I picked the package to match the other classes in contribmiscellaneous +Also I guess Windows Mobile emulator can be installed without MSDN account +Hi Kathey thanks for the feedbackI see what you meanI also notice that IJs code is calling the version of setObject notthe version and I wonder if that is why your repro didnt provoke the errorI will look more closely at the client code using your repro as a basis +Note Im not necessarily opposed to this change but please justify its usefulness +Changed and to +I would like to fix it could someone help me out in assigning the bug to me +One of the reasons I am slow in hooking up the old fault injection code back is that I have a better idea how to make more flexible and totally fit it into runtime +That is the unchanged cloneNode method +thanks works great on transition to closed remaining resolved bugs +Piggybacking on the parser seems a little more elegant and should be nearly as fast +Hopefully this is what you wanted +Assume closed as resolved and released +Also note that Solandra does use directly instead of going through Thrift +Can you try your test case with and without to verify +I think this bundle might not be needed unless you think otherwise thanks +This method assumes that we are using the Max Visited logic +This is the case for all the methods +Fixed in rUse runjava to launch java programs that call x programs +good +There were no issues with the last Validator release either +Rick is adding more information in and +Then you need to use configuration to point to own implementation of renderer +And youre right regarding its really a weird ll deprecate it +OK agreed that is a bug in. + patch looks good to me +Generated Download Page Notes Some binary distros use the bin prefix for binary distros others dont Commons and Collections currently have two release versions listed Daemon only has files Latka lists a single zip file Email also lists javadoc and sources jars Launcher has examples distros Proxy has a download page but isnt yet released +This is not ready for review yet +Why is that? Doesnt the UIMA implement? If it does it should work if not whyThe truth is Ive not had a look at the UIMA wrappers i dont know whats happening thereOk ill give it a go tomorrow morning +Moved to +Will not implement. +Make sure you have both outputfalse on the function and the component +Ill upmerge instead. +We have changes filters now +BingoNoel I confirm receipt of your manual messageVincenzo +Will be delivered with API +Thanks IliaSeems i was wrong and ERRORINVALIDUSERBUFFER is really caused by simultaneous read and not too many requests as i gueseed before +Cleanup of code and error handling improved in +I dont think its needed and I think our solution is only temporary +Where exactly do you want this changeAnd what is it good for +Rollback Chris change +Please upload them +Rolling back assumes one client has complete control over the index and the commits +Here is just a few things that doc team should remember to align with Red Hat Developer Studio Perspective will either be renamed or removed We will have a Seam To fix this issue Daniel has added new step in Installer to select appropriate JRE to start eclipse see last nightly build There is new item in menu on Hibernate Configurations view that allows show daigram for selected entity +Its probably something in the +Pat was this committed to branch? I dont see it +Grid can you look into this one + +Ok good start +Looks good +test with legacy pooled invoker which uses the JRMP invoker proxy ha so dont foresee any issues +Added backwards compatibility for mojo +After looking the code this is definitlely archetype specific and specific to any mojo using resolution capabilties that are not done in the same way that Maven does it +To reproduce the issue please use this EAR instead +This may also help reduce the number of hanging Github pull requests on since Github will autoclose a pull request is the HEAD SHA matches a pull request SHA +Ideally IMO the filter parsing would be done by javacc or antlr so wed have a real grammar +Really moreshould be done to parse the argument as a stringified URLAdded some test cases for it +Patches look good to me +So we defined that at compile time a vanilla name that does not refer to a local variable or field is first looked up as a class and if that fails the name is seen as dynamic property +Hyrum Can this one be closed now +I think only get was missing +Submit a testcase that reproduces the problem otherwise Im going to close this u run refresh it serveral times you will find that the memory continue to rise but without drop +Yes thats a viable way to fix actually created an interceptor to do this for me so I didnt have to modify the spring core classes +The fix for fixes at least one case of this +Ok np +Sorry a typo +Remove author patch with author tag removed +Thanks +Hmm I dont know anything about Glassfish +All the properties are shown in the tablePage Shows the list of all string with the generated keys +Thanks for the work +Best regards to respond to your comment about the continuing number of failures in +svn directory to the generated +I havent verified this yet but if what Runping is saying is true does it warrant a release +HansI dont have a lot of time to help you investigate but I can give you some clues that might help +has been opened for the discussion of similar optimization of wire format for homogeneous arrays of objects +It seems that trunk revision of doesnt match +And it also occurs when using the which can be downloaded from Maven Central +this is happeing because a beta version of the was used +BenIf the mojo project heads in this direction that could be a lot of requests for rewrite rules +In fact we should reuse the file encoding defined on the Eclipse am taking encoding from Eclipse project level and passing it to parameter +I think I spoke too soon +What if the option is contentType and its default textplain +Awesome thanks! Ill schedule some time in the coming week to patch our dev installation and sic some QA guys on it +Do not issues marked as resolved against a released version if an issue occurs again it is a new issue and a new Jira needs to be created feel free to link to the original problem for context but do not break the release notes history in Jira by moving already resolved issues to different releases. +Thanks Rob +Ill review that because I find it odd +committed to branch revision more changes to beta branch at +Committed to r. +In all the +Great I just committed this +Committed patches to trunk with revision and with revision . +JacquesI think the patch is a good demonstration of how it could be done but would need quite a bit of work before it should be committed +Thanks Vasco +my ctsplease make sure that all these scripts startup in a restricted protectiondomain assigned by the classloader +Flat presentation results in selecting the last package folder +Steve while its great to set such a baseline we need to make sure that we arent creating new testing infrastructure incompatible with what we have so far +Can you give more specifics like the exception +Does that work for you +Integration tests were updated +bq +Bulk change of reporter as per +This is supplanted by. +Attached a rough patch for will attach a patch for trunk if this looks goodThanks think +Hi There are still problems with encoding parametersa parameters from hidden fields sending with POST methodb parameters sending during response typeI prepared patches solving these problems for me +is out the door adjusting on all remaining unresolved issues to by new externals should fix this. +I attached also fix the store func setup issue +The odd parameter passing between the compiler and the execution engine was cleaned up by Knut in his patch to +This patch introduces a few new classes for the handling of hits +added blin but could not find jliu maybe subsribed under another id +no time befor enext release +still requires the patch to be applied +Customer reports Firefox plugin was cause of the issue +I cannot replicate this issue om M build using steps to reproduce +The default for these jar files is that they are jar files I dont want my production system running with all java code in it +But weve added a system flag to bypass the FLUSH check so you can still use a protocol stack +Trivial patch fixes the issue +this may need a wfly issue this likely for tomorrow? I see its your last major for M +Yes youre right. +Would you only ever want to set a store? Or would you want to reset other objects? If the former then I guess something like POST resettoppstatesshapefile works +for the to trunk. +Side note I was surprised that nothing in UIMA detected this situation at +Attachment has been added with description Testcase showing my Ill have a look ASAP +Please can you assign it to someone from the AS team +Added a client attribute to richselect +get need not be synchronizedbq +QA can tell me if Im wrong or not +Changed error message in need to change for Branch and Branch is another error message which can arise in this multicast binding we try to bind to a multicast address using an stack we get the following exception Cannot assign requested addressIt appears that binding to an multicast address using an stack is outlwaed +Column and relations extend that type +Doing this with the would be greatCan we rename the issue title toSupport defaulting Index name in the ? +We have also not changed anything from original Seam working code +Initial patch +ok please use the attached patch called in patch. +That should allow us to grow to new in the future. +not important for changes to fix this in CR were made previously +Trunk Committed revision +I believe we can close it +Added startup and shutdown script Added torque metrics data loader Improve handling of Exec Plugin Added Test cases for File Tailing Adaptors Added Test cases for Start streaming at specific patch depends on the patch for +Fred Bricon These are the correct when you build and deploy the quickstarts using the command line +A workaround is to manually rename the jar sha and md to the latest version after update but it is of course not a very elegant solution +finally +However the DFA probably is the ideal model to go after +I ran it against about partitions input and didnt see obvious client time increase +Nice work. +As far as I saw from DN and DN databases the schema upgrade might indeed be required +is this fixed Sergey? Isnt used to indicate infinite +for also will be fixed which has an inappropriate break statement location +Move it back in if you think differently. +Verified that the realm is usable in a web app. +But let us examine also what you are trying to accomplish +ideally that should not depend on the java version IMOfor my side I will release a new version of the and since everyone tends to move towards newer versions of java jruby things should be fine +Current mult operation algorithm is too bad +I dont see reason for this +Yep sounds best for nowI think Jareks advice is good but currently I just choose solution to make this patch for +Hi James no problem +jdo files according to the latest and updates +Fixed +Ideally Id like to only expose interfaces in the WSDL but interface inheritance is not regarding the Im configuring my services in Spring with the and did not have any problems with I assume doing it this way meant that the XML namespace was correct by Im reopening this so we can take another peek at it. +The simple insert case to Blob and Clob data type can be hitting it like the does +I moved everything back into the +I tried specifying this however and it caused unit tests to run considerably slower not a good thing. +Patch committed to branch +I think that all the week numbers are wrongAnd i will fix also of courseRegards the screen youre looking at is using a Freemarker template then the problem is due to how Freemarker handles date transforms +Added documentation in and just committed this +Could you clarify if the problem is that the ipojo package in question actually uses asm classes in a method signature or by extending or implementing one of the classes so that the uses clause is correct or is this another example of bnd generating incorrect uses clauses for packages containing implementation classes because the uses clause generator scans the whole class not just the class parents and method signatures +Trying against hadoopqa to see whats broke +just need some testing on! +application and these need to be escaped too +How can i use this revision +In Jetty this error is caught because Jetty needs to load the servlet for further processing notbecause it is trying to validate the +The problem with adding it however is that gets called every gossip round and runs through the hint removal processBut it only runs hint removal once per node right? Or is onChange not an accurate method namebq +I ve configured a different version of jira cli that works with the current jira version +Ill add Ignore once we know what the problem is +This problem also exists in the current HTML content handlerThe problem with an embedded element in a string does not exist if markup is not included but i couldnt manage to figure out how to replice Boilerpipes proper handling of these +Marek has just checked in support for svn workflows as well +Daryn patch looks good +And we better do it now before are frozen for branch +But is possible the Cordova App and the Camara Intent consume the MB of Ram that is freeWith the plugin of Zxing for read qrcode that uses the camera there isnt any problems like that but may be Ive been considering that the Camera Intent is more simple than the Reader IntentThanks a lot I will try it in other mobile phones +This replaces with the corresponding Felix sources +Of course so simple I cant remember last time opening itYes you are right +Closing issue out as part of JIRA cleanup +We need to support temporary queues across networks +What did this cause to happen +Fix spelling in patch +Completed At revision +Ill provide theadmin a list of subscribers to add from current back intothe new testershttpd listPlease act moderating these very quiet but oftenspammed lists is a massive waste of cycles +Once commit we will see the right schema +This is the is the fixes from me +Not the most elegant solution but it should work. +One possibility is make call to for the name and return truefalse based on null value +If you are sure that this problem will not reappear then I can remove the block +Heres a patch to pass that flag to testsId appreciate if someone can test this on WindowsOSX to make sure it doesnt break their platforms +Unfortunately this cannot be added to derbyall before we can use JSE since there is currently no way to turn write access back on again +JanneIt is version +to a separate issue for moving the length default checks to DDL timeAm I correct that the current patch is limited in scope to allowing defaults greater than for varchar columns? If so I think it might be good to make a separate issue or subtask with the specific description of the change + +Since the default logger will write the stderr that will handle the standard case quite well +the patch just needs to be applied to the kernel patch on the shouldnt of been includedI did do an svn update before creating the patch but I needed to change the Repository completer +I would make the access method final. +I dont have resources to hand to test this at mo +wicket can simply detect this kind of url attack and abort request processing + sounds great +Deepal Here are the files that caused this bug to occur +Committed to trunk +The test failed because of too many open files +who commits the patch then? i would if i could can i have access to that part of the tree +From Hadoop Dev Mtg Owen will try to get more context on this from the Y! branch merge +the separator and format +The number of corrupted files is usually low +Ideally there should be no multiple RH have fixed the main issue and the issue of handling multiple replicationhandler instances can be handled in a separatebq +and recognize bothThen so long as someone is upgrading from the most recent to theyd be fine +It works well on hardware its unsynchronized and its fast +My fault for using a class without a packageSorry +ve committed this +This is exactly the purpose of those other changes but its not really the responsibility of the release plugin +i couldnt find anything +Ive read the PCI guidelines to research just this and communicated with others familiar with PCI about this and my understanding Persisting the CVV code is allowed by the PCI requirements but only within the scope of the initial card authorization +And so I consider this as not a bug +The screen is exactly same but without debug message. +Please attach a sample application as it seems to me you are probably misusing the API as it was intended to be used +Ill go grab it thanks bud +Mass close all issues in resolved state that have not received further comments in the last month +If by someone you mean you and by a little effort you mean a pull request +This is how it was before where we had to move it out of the +Added a simple check in the listener method to see if editor is being disposed or not +Luke we do have the patch for classic MR trunk in can you pls port it and commit? ThanksThis will ensure nothing breaks +Regarding XSD specs assuming you finish the patches to the newer version we need to sift through documentation and update any usage of the specs to the new version +That appears to be whatyou imply in your original statement +As you said the problem is in and the way it handles all the possible JPA implementations that exists in the classpath +This time without absolute path names +Attaching test patch +Just changed the log and moved the error processing there to should be added in catch which adds it Should it be in another JIRA? This is already in I wanted to resolve +maybe we should bring this to the Dev you for your reply and posting this to the List +Then we can debate the value and issues in the particular patch and how to move the project forward +The fix for this is localized into and the change is to map the? to BLOBCLOB if the other operand to CONCAT operation is BLOBCLOB +Youre right sorry +Slight temporal anomaly +I second Davids comment +Everything else works. +Not sure why the testbot didnt catch this but it looks like this commit broke the system test compilation +the changes look good +I think he meant latter +Remove parseArguments +I have not written a reproducible for this as yet +released! +jUDDI is incorrectly overriding the JNDI defaults rather than allowing to do the work +I want to address all of the questions you pose in your comment but wont have time until early February +Typo above should have been +This problem also applies to setURI setValue and set +Added a sourceEncoding parameter +Verified in +ElliottThanks for the info +Fixed in the javaee module. +So just having null check should be fine +The root cause of all these problems is schema validation introduced in Spring +Assuming you mean this +Patch attached +patch for trunkNutchgora branch has a couple of additional dependencies so I will get a patch sorted for it when I have time +patch +Re Index srcorgapachepigimplplanI will make the changes you suggested in comments in a new patchThanks for reviewing the code and for the comments +Thanks Nuwan for the contribution. +Ended up not needing to change the existing although I did reuse the request and response interceptors +It wont be a performance bottleneck as it is called when stopsThe biggest change is using state machine in the newest patch +thanks ben! +Yes we need to have it as well as import static +Between html comments and exlanatory sample text I hope that these templates can beI think that toolsreleasetemplates is a fine place to park the new templates +Downgrading severity from Blocker in light of the previous comment and JIRA name change +Thank you Matthew +is a screenshot that incorporates the derby logo from on a site built by forrest +Patch to comment out the assert for now +It seems like it was not properly formatted by the author and there was new lines and white spaces which was not recognized by tortoise +Ugh +should post soontry running hadoop in debug logj mode +First theres fetchValue attribute acting as converterSecond you can do the necessary conversion directly in suggestion action isnt it possible +Marking LATER pending a course of action +Fixed via +It is fixed for vIt was reopened to remeber to merge into. +Ive attached a patch that would not use failover when the forwarding bridge is sending a shutdown info +Ok I still dont know what stopped me from upgrading from to did manage however to upgrade by doing it step by step from recommend people to do the same since it worked for me +Everything seems fine after the fix +The next step is to update the core API to use and complete fleshing out the Context interface +Anyone have comments to this? Ill try to add some tests and prepare for commit +Sorry Avdhesh I forgot to add these two files to the patch here is the new patch containing the missing applied it would be good to cleanup some of the consolelog outputs to at least have some meaningful information being output if any +This issue has been fixed and released as part of release +This issue is caused by the http chunk mode +now has persistent read write and rollbackCount +patch that provides component queuing support on vote for adding it to because this feature does not compromise anything it is just an alternate way of adding components +Not closing until Doc updates are for were mega merged from trunk to branch on Aug +Perhaps it can be made more reproducible if purge is run in a thread parallel to the consumer +Since this is an external codebase at the least well need a mapping from each external author ID to ASF or well need to simply mark them all in some wayGarrett Rooney ought to be able to provide details +Thanks Owen +Moving it to the next CP and assigning it back to will take a look on Monday +Thanks for the feedback +Iam having the same pb I have svn copy a folder in my trunk and since mvn changelogchangelog failed +Reproduces the error when Network Server has NOT been started +AXISPRINTFFORMATINT +The port chosen by the client is random so if the client is behind a firewall this may fail +And during that time only partial of last edit was written +closed with release of +I will wait for that before working on the rest of the issue +Currently the Build Queue appears in the menu when the user has the been fixed by in r. +Id like to discuss this before we call it resolved +Do you want to do that or should ILet me give a try +Theres a file in the tarball explaining how to use it +We dont allow true anonymous access to the data but we do allow any valid ASF LDAP user to view the data +Ideally the cache stores would be fast enough that we dont need and the consistencydurability concerns that come with it. +Applied the patch jira can be closed +Is there any update on this? We are moving more and more projects to Gradle and I would love to be able to populate the Project Links on the dashboard through the build +Actually the title of the issue is misleading the actual exception is coming from get before the scroll even gets called +Added code snippets +The patch makes most of what is private in DR protected +Remember the javadoc that Knut Anders quoted from is from some draft of the JDBC spec +I think that its more common to see things like than isnt it +Il lalso do some testing and verification at my end +I am still not able to reproduce this even if I change the as you suggest +I would be very glad to review and apply your patch if you extract the messages +Tests for the binary value object +Produced patch by copying new source from MR trunk +I have reproduce this issue on my machine and there is still hanging on checking cmdIts strange that works OK but quickstarts NOT + +we populate the information to the shared http client only the first time +Was checking every variable declarations now checking only class fields +I could not get the following lines of code will be great help if u add more comments to the code +This issue was fixed for Solder and Seam Faces. +Fixed chars is quite a bit +Ugly ugly names +The problem still exists it just takes a more complex example to show it +Please supply example project if issue still occurs with next release. +Note all the colors in Recent area of the color chooser are greenish on RI whereas they are gray on Harmony +The two basic modes out of the box will just work wo edits +you wont believe this +We can wait for but it looks like a test in might fail as its encoding base data and expecting it to be decodedOne more question which perhaps isnt related but it looks like is also encoding data as Base +Try to reproduce issue before delete image. +Yeah the test is commented out +Theres also something seriously wrong with this on Java +Closing as you can now plug in custom logic to make the decision. + +However when I created a new project in my workspace with the defined server as the target runtime the plugin will get activated on GEP startup +I based the design somewhat on the work in Hue but I found your SASL lib to be less than ideal where a library could easily work in its place without any strict external dependenciesIve primarily tested the GSSAPI mechanism but have done some light testing of the other mechanisms +Ctrlspace dont bring the class too confirms that there is no bean in CDI model +Ill review and amend my Release Notes entry if required +HiIve corrected this bug a long time ago but never submit a patch +running tests +Well keep HDFS naming convention which is along the lines of the naming convention used in Common +I committed it now. +Updated the patch against trunk +I am not sure if we should do that for +Behavior is not a bug but design as planned +My confusion was related to thatbq +Starts up ok logs to only now and output from main goes to log too. +Hmm interesting the thread dump you included shows waiting nothing that should be consuming major cpu cyclesI see you are running the full config +Same error when saving a Network Proxy +I have opened a new ticket about this +Its fixed and the unit test is part of the standard build +Unfortunately we cant apply your workaround to Seam as it alters a spec +I just did a checkout of and this test runs successfully on Windows +I think it is good to have the unit test +This patch changes all methods that call check to be synchronized +I have noticed that other changes are tracked in JIRA and +Interesting +I dont know when is expected though +Committed revision . +Seems like we dont have test for Subscribe in + +Because Java arrays mixin Enumerable they also had first for a short have not seen a demand for features yet and so we backed out the handful of changes we made +So this is purely a extension to the STOMP spec and I think we should be able to make it just work as expected even if the type info is not part of each header +The formulae for determining the pending tasks have to be changed to the following in order to have the correct number of pending tasks to be determinedThe formulae would benumber of desired task number of running tasks number of failed TIP for task finished task speculative running through the logs Ive realized that the negative count on running tasks results because of that report failed tasks after the job is killed due to maximum TIP failures being hit +This ensures that the analyzed data would be loaded into database +Try to break it and if you can send back how +Moving out of +YEAAAAAAHHHHHHHH WAY TO GO +The same as above seems logical to me the is taken from final project why not overwrite another file that exists in both the dependency WAR and final project with final filet +Reopen for clarification +The fact that error message is pretty much meaningless is a different story +upgrading nowThanks! +Im fine with those nits if someone wants to update the patch or open a new JIRA but Id like to get this committed now and free Hudson +services shutdown nicely +Thanks Simon! +Committed to +I confirm your Issue with Firefox +ant test passed on my box +Im going to mark this as fixed +Andrei I dont have any clue what part of codes I need to look +The easiest way to check this is to modify the management port in and then try to connect to the server in Explorer + to from +Hi Helderthanks for your help and for providing the two test cases +Based on the feedbacks the hint is changed to using collected +workaround dont work with richfaces and Portlet Bridge +I updated the error message but perhaps the can just replaces or remove the whitespace before deploying? I leave this issue open until we take a decision +rUbuntu smoke kernel +What do you reckonRegards +Patch applied +Thanks Mikhail +If adding this config knob is not trivial and they have a wildly unusual setup it isnt worth the committer effort +This issue is not approved for SOA CPO +As a of the discussion on OK I think that leaves only the patent wording +You dont often need to edit the template especially now it does not have a fixed version in itMostly you only need to edit and then runmvn changes Check that the new file looks OK and check it inThats itTry it and see +This JIRA is to provide an API layer that will abstract away the caching and query of the jobs from how that data is used whether it be a web service or a web GUI or an RPC call +Irrelevant in what sense? Irrelevant in that use of that rewriter always obligates the server to perform a? Yes which could depending on the data and the rewrite being performed call into question the wisdom of using that particular pass +Reviewed the patch and it looks good +and tested +How do we deal with our currently corrupted edits file? This patch needs to fix that as well +Selection doesnt work +I just committed this +Needs verifying +HTML output file is included for review +Not sure what partial install actually mean +The attached patch fixes the problem trivially by moving the JNI native methodimplementation to with appropriatefunction renameThe output on Linuxia after the bug fix +Bruce are you working on thisIf not Ill start to work on it and prepare some ideas to discuss about soon +one seemed to like the normal encode methods output when the other did think this is a larger issue than simply file +Matt Can you share the solution to the problem on +This is done by the line execute goalfind bugsThis will execute findbugsfindbugs before findbugscheck +synced with the branch +Verified release notes +EricWhen you kill servers do you stop the whole hardware just the accumulo processes or both hdfs accumulo +Closing all resolved issues due to a successful release +committed to in JBT +We used to do the throwing an exception thing but the particularly bad thing in this case is that doesnt declare that it throws any exceptions so you have to throw a runtime exception and then catch that runtime exception in the outer code +The map needs to count spilled records as it writes to disk thats straightforward +to clone the all directory +I dont think its possible to eliminate all the duplicates programmatically +Im not sure why the conflict was not there before +sure thats great +Seems to be fixed +Average size of sstables read per request is on my tests with and without counters ABSC still show better results comparing to TMBSC +The method that you proposes is not very efficient why do I need to list all commits just to retrieve the latests commitDataIf you only have commit it does not matter +The reason is that the client hangs when the chat server is not running +Where is the inside the etc folder fromits not the one provided by Pax Web +This issue doesnt depends on? What is the relation to ELs in Java sourcesIm removing the link +The testcase runs with a reverse unicode comparator which triggers terms to appear in reverse order during indexing +Our goal was better support for for example financial applications +Any update on this issueIs there any way that I can resolve my SNAPSHOT dependencies running the maven in batch +Somewhere on the advanced options screen. +thats strange +I cant see why +We can revisit this if we get all the WAL optimizations from trunk back to for a release. + +Looks good so Ive merged your patch into trunk andThanks. +Patch includes some better error handling as well as some general cleanup +I added debugging to but then I couldnt get it to happen in the time I had to play with it +Initial docfix patch that gets rid of the staging talk +master tries to change state to REALLOCATE +Please disregard my previous conclusion in SLES selector wakes up when read on sink channel occurs +I would like to be able to swap out an implementation via the Maven build process whether controlled by a maven profile or simply editing the pom +But it seems counter intuitive to a user who just made a BPEL project that nothing is showing addition to this the project is a modulecore project but completely lacks a +Both are working fine for me in and master +Because uses the return type to decide what getNext method to call it would fail with the use of +should fix this issue +I take that back all individual connectors are clean from javac warnings +Why do you want to disallow anyone adding a new serialization type in the future its one byte you are saving +I guess it has been fixed by Adam at an issue +Committed to trunkThank you Chris! +Checked in a fix for this really get should be a core part of but for now Ive implemented it in an class. +The if exists syntax is not valid for series databases +Is it same as +This patch introduces JDK dependencies which should be removed both from trunk and against revision removing dependency on taking a cause +Please review this code again when you have timeThanks +Committed the most recent part patch to trunk and branchxThanks again Uwe and Hlavac! +I use Ctrl because for java editor Ctrl +No more pending patches and No backports requiredChanging the status to ResolvedThanks Kristian for pointing this out. +When another client is holding the lease I do not think we should allow create with overwritetrue to delete the file +Thanks Jing! +bq +It looks like the underlying mechanism is already set up this way private double force +We cannot support the zOS versions of these products because we dont build them and we dont have the environments available to do so +We still havent decided how to handle cases like this this is a bug in MRI up til +Commit the changes to trunk at revision thanks Shawn! +I have added that new test case in file which checks if the keeps the right semantics +Inside the patch itself it looks like there are a few places where the formatting is off + +I will test it as soon as I can figure how to get a nightly build which is something I am completely unfamiliar with since I have never used Git +I see a potential for a hang here if the delegate doesnt become available +Let me with them thanks +I just see it as a fracture in functionalityI would feel better about leaving the querytool alone if it were deprecated that way there would be ONE tool to query the +I have taken a look at the source for both and and noted that the cleanup indead has changed + + pending jenkins +Interesting +If it is not persisted during active replication it is possible that replicas end up in ways where the block placement policy is not satisfiedstack and nkeywal when a node dies there is a correlated failure and replica count goes down to two +Based on the feedback on this doesnt appear to be a priority in the JBDS timeframe which is what would use for tooling +If the error is in the then the statement is actually printedI would rather not see more code in the alter table code trying to figure out who called it +The biggest worry is that immutability is broken not that a copy is made +the task will always be run regardlessi quite like the idea that i can specify a task that runs against a bean only if the bean is required by something else +i commented on the seperate ticktes +Stale property values are never merged so this looks like some code is updating the property concurrently with the code committing your transaction +Zip files containing patches should be typebinary +is able to write exceptions Im not sure if yours is doing that +I sent it Jozef to look try out +I have rewritten all the related code as a part of +After code comments +Out of date +As we discussed this issue over IRC and phone I assigning back to you +In terms of the what we arelooking for threshold violations that we define over things like HDD smartdiagnostics and other log data to make the determination on if a node isheathy or suspiciousFor servermaint requires operator invention to execute it +Yes + +Committed to trunk and +It is not a problem if Ivan wants to have a look +Thank you for your precious feedback +Can you clarify whether you are focused on Python reading unicode strings or Python writing unicode strings +It looks like we will go with a policy of having no provisional API in our releases so we should keep the packages in the felix namespace and guard them with a mandatory attribute to make them opt in. +Committed to branch thanks for the sense. +The result was the same +Hildaset is a method on the interface which we cant reduce the visibility on +The third parameter of each of the three signatures of these methods is defined as long instead of int like it is defined in which inherits fromAlso the same problems occur in aswell +Reopen if there are any remaining issues to address. +I am reviewing the v patch currently but will likely concentrate on the store level changes +Integrated to trunkThanks for the patch SergeyThanks for the reviews Stack and Nicolas + +Verified with JBT. +Is there really any reason to name it +When an XML import is performed the API that is being used for creating the node hierarchy is the same API that is used when a client modifies nodes +It passes when run against a common with appliedCommitted to trunk and +The new jira number is SYSDEPENDS may be keeping redundant dependency info +Thus this issue will also fix no +I think Todd was concerned changing to make an RPC is unexpected +Youre right Simon + +Committed to trunk +I dont understand your comment did you mean that you cant see in the source code the access to the nested beans? I checked in the zip and its OKCan you explain a little so I can give you a better answerThanks in advance + will fix the Windows build issue +The will probably just apply cleanly +Fixed in trunk +added Alex +Any security vulnerability can be exploited to corrupt data +Then you can enable that option and it will transfer the exchange as a serialized object over http +Maybe Ill include an example of thatThanks also for the confirmation on the bit mask +I missed and inactive icons were reversed. +Variant of patch applied in r. +I had seen these in the suite in our product +you are right +In current implementation we inherited from master branch it will check if the volume is attached to a running VM +That seemed out of scope +Is there something specific youre looking for in the new code +We will add the docs when final is released +Resolved with revision of srcmainpatcheseap message Use version +The fix will be released in +you could try latest test case in each instance creating new dbs for each test run agains jdk jdk and jdk to make sure +Thanks Jon this is awesome +Patched both trunk and branch +Pavol framework version has been incremented tocould you please update also QECI builds +Not being an obj +Change the name of the parameterAdd the new parameter in Also can you add a review board entry +In particular the general approach of using interrupts to stop the thread is rather fragile +This is why some tests relying on the real Aries Resolver failed sometimes +Ran test case against works as expected +So better to generate your patches from shindigphp or shindigphpsrc that way no one will get lost about where to apply it +Im attaching the pub key so someone can put it there +I think we should remove the Mglicherweise ein Session Timeout! part if we cant handle this case with a full reloadBut without pointing to one of many possible reasonsThe internal reason is It was not possible to restore the on the server +Moving to maybe there was some confusion about parsing vs generating but this has not been implemented in the ndim branch +Initial patch +patch attached +Still the default size should be readable its a minor issue but making the text scrollable would make it look better +some quick multicore testing makes me think its improved the error reporting in some situations there as well but im sure i havent tried all of the edge cases it may have broken something +This probably belongs in +Attached the wrong patch +Committed this. +closing the issue as duplicate of the newly created one in issuezilla +Daniel Ive committed a fix for +Has it been already fixed +If we are moving stuff out of core and into contrib dont we need to have a version that matches for a first releaseNo fixing core sorta assumes that is a major releasebut we can do the patch on the issue now since because contrib has no backwards compat policy. +Also one can choose to disable this by not configuring the proxy user settings for the REST server +do you have a WAR +The UI seems to be passing invalid value undefined to the scope parameter +I have updatedadded Licence information +The fix looks fine. +Glad to see the problem has been solvedIve also merged the change in the. +We still need a pluggable implementation of scanner than can use a was fixed. +Thanks Ryan for your comments and Jacques for commit the patch. +good improvement but bigger issues to fry +This could be either done by compiling all sources with compiler or compiling java classes with regular compiler and then weaving the binaries +Providing a distribution build capability to +Usually all the gbeans codes could be loaded from the artifacts in the defaultEnvironment I found an extra field is addes in why not define an artifact list in the as class sources and directly use them as the source artifacts? Thanks +Ill remove them +It heartbeats every seconds +Now that work is underway with AS all previous community releases are +Logs and other stuff required to run this test +Charles Thanks for the feedback! +Use a failover sink processor instead of a default sink processor +However there was a typo in one of the initCombos methods that prevented a couple of tests from running +Ahhh thanks +Thanks Rohini +Can you check it +Done +ok if we follow here again or betterthe eclipse java plugin then it should be an error +Likely this was caused because of a disconnect between the project as imported and as authorized in Nexus +Kevin Parkerson Im not sure which version of Shindig youre using +Help menu under Windowsthumbnail +Im use project files with BCB compiler all of the time without any troubles +Verified +And this approach is not equivalent to the old behavior +No the way stop works is correct as is you really should turn on VERBOSEyes if you want to see those messagesSo our problem is that Cassandra is taking a long time to shut itself down +Closing since is released. +Committed to main trunk with revision needs to match the deployed to main trunk with revision . +Thanks Mubarak +Patch committed. +behavior of is restored +However there is no issue with internal materialization and external materialization is not really considered a process so this isnt necessarily an issue. +This silently broke our builds because we also had a case where we needed to execute SQL in a POM module +I dont see any stuff either +Not sure but it could be that we have fixed this with +task is done. +If they change the meaning though please fix them back againThe output HTML is in +I assume that the goal is still to make the RPC completely pulggable +Borislav could you verify if the fix solved your problem? You can download the latest nightly build which has this fix +i want to run a few tests with this patch in place up on cluster to be sure its refactoring of such as the changing registry of store files under load hasnt broke anything + support using Spring Security login ve attached a new patch which allows roller to support login and registration using simple registration extensionNote if you are gonna use the patch youll need to change the schema file reference in the and download a new library attached to the issue +Can you describe your application a bit more so that I can create an Eclipse project that simulates it +And we insure that because the scan positions at the leftmost matching key now and movesright +Sorry binaries +If you develop on a Windows box I recommend which has a handy utility for applying patches +Thanks Arvind! +Add check HTTP status ofplugin never check http statusI submit a patch to check http statusI hope it can be include in releaseRegards +I have just sent the same pull request to the richfaces repository +Feel free to reopen if there is remaining work +I think it was reported recently by Stan and Stale had a fix for it and a pull request + patchNew factory and implementation and code to make use of them and support rd party implementations +Let me know if you want it expanded or +Header because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Shall we close and create another ticket later? +Modified code to take appropriate action when a listener fails to connect to a broker +Patch committed for real this time. +As a task may be assignedtaken many times it can be created only once +forgot fix version +The WSDL generated by from the deployedNote no wsdlfault elements +a patch proposal +Calling brew install says that its already installed +I will fix this issue and if you find any issues with your testcase please reopen This was the problem with the +Please comment if you are seeing problems with othersI havent yet tested the infinite loop symptom on the main page described in the report but will investigate that further as wellNo diagnosis yetPlease see bug for a list of the affected themes +verified by Vladimir. +some methodsinterfaces need pruning +A maven pom has been included with the example and is configured to use the jetty plugin for easy testing +Please add description as similar as for other project +added new properties as suggestedindicatorindicatorStyletitletitleStyleIt is a very simple task to do with myfaces builder plugin +The profanity part doesnt bother me but theres probably someone out there who will bust a blood vessel if they saw it +Committed revision Committed revision + +If no one is ever going to wait the notifys become dead code +Bulk move issues to and +The problem is coming because in case of deleteDomain with cleanup true CS is not allowing deletion of domain if there are any account under that domain needs clean up thoughthese accounts are removed and admin cant see them in the listaccountsSo CS should not restrict the deleteDomain in case of accounts needs cleanup +Again found serious log messages without did some more investigation and found out that of these messages are inherited from other messages with same number of arguments and valid value patter so this issue was resolved by previous fix made by James Closing +We could just allocate a new flyweight Column object when we want to read it +The workaround of placing Jars to lib folder was not a choiceMy issues were solved with the combination of this patched and Annotation Scanning configuration in XMLPlease note that the attached does not contain orgjboss that causeERROR Failed to load orgjbossweb ignoring +Applied patch submitted by David Graff + + is already closed +This report is supposed to be posted at Panel Component ve tried searching to edit or delete the report but i cant find the link to delete or even updatecan anyone help me with this? i am sorry this happens because this is the first time i post at jira codehaus component to Panels as requested by reporter +Fix provided by Michael Watzek +Im still exploring why +Thanks SianPatch applied to PACK module at repo revision rPlease check it was applied as expected +One one hand that makes it seem silly to go out of our way to support the API but on the other hand if the API provides all the needed functionality we may as well provide support via what might someday be a common interfaceAFAIK the only other common interface that caching products freqently support is which for our purposes might actually be sufficient for a simplistic integration +Right now we need a webservice optional test case and run this with tomcat and axis +Probably tens of thousands +If you run into more problems along this line Id recommend trying to get a test working and reopen or create a new issue +We need to decide what to do with db module +Currently one test is failing because of this check I have to figure out why this happens maybe a test bug the failing test seems to do some fake requests to default select handler and passes some remote streams +Here is a patch carrying arraycopy optimization in the separate PMF pass +Canceling patch so that findbugs issues can be the patch to fix findbug Most of the code is reserved from earlier old API code only making the changes in places where new API is required +I am using the Apache UIMA Cas Editor plus the plugin and the Corpus Server to annotate data but we are trying to move away from that to web based tooling +Earl please attach this patch as a separate file rather than putting it in the comment box +t is related to Canopy for sure +I know that you can set which inputs you want bound within a view state in webflow but is there a way you can exclude inputs from attempting to be bound? +To the best of my understanding these two issues will are dealing with different issues +For the record my SVN is that out of sync with the main repository +So it can be problem of the example as well +Thanks! +Thanks Tony! +Fixed for +See for my request of this very thing +This has been implemented +Create a network offering as given in Section Creating an Internal LB Rule +META +I just tested the latest build and it seems to me that the problem has gone +And it also allows for bulk assigns +bulk defer of open issues to +Its just the compression test that fails FWIW so it seems that something is broken in compression or the way we read them +Ah or is this a client module you can use in the Python interpreter? Seems moderately useful. +Test work of workspace ws via +Addressed Sureshs comment + +Can you at least try to find out which YAML document it fails to load +The continuum stuff from trunk was merged to NBCOMPATIBLE branch +Added support for the two definitions in the same generic class in the Java +The patch changes the ALTER TABLEhandling so that the parts of the identify column that you arent alteringare preserved and not incorrectly reset +Committed revision . +Nicolas I think I follow +Also once those comments are addressed +Im moving this issue out of and +Now it is back to what it wasThe problem that I see without this patch is that sometimes the new peer and the leader end up in an infinite loopI have seen this when there is a new peer that has a higher epoch than the leaderShould that be reported in a different JIRA? +We should just return the header as an as we always need to retrieve all headers anyway +I provide a way of embedding jars over at which could be modified to work inside the without the need to modify poms or use the project layoutHowever Im about to leave for a week trip to the UK visiting relatives so wont have much time to spend on this but can provide advicesupport and would be willing to help with coding once Im back in Malaysia +bulk close of all resolved issues in preparation for release. +I want to apolagize +Nit Document the return cacheReportbq +Ive refactored the constants used in and into a separate class +htaccess file is that? Do you need special karma to edit that file? Thanks. +Fixed with checkin +Attachment has been added with description testcase for this bug +this is a duplicate to in Developer StudioVersion Build id Build date +Ported from trunk to branch at subversion revision +I ve tested the solder develop branch with guvnor this issue is resolvedJason Thank you +The breakage was introduced by the changes for so we should confirm that defaulting to doesnt break your fix for that issue +The underlying Adobe AIR implementation of has limitations requiring a between scrolling and softkeyboard feature control +Im sorry and thank you Carsten +It would need a sponsor and work to make it useful along the lines that Todd and you suggest +to branch thanks for reviewing Eli +Fixed. +this is causing the to fail +patch for failonerror ll fix this if you provide a patch for the documentation and the for the xdocs +The attached WSDL generates working code +We should return an error response and the applicationjson content type +Can it have its own config to to do different policies for different subtreesSure can +Can we give it an optional keyspace name argument to override the directory name +Perl PHP installedLet us know if all is well and can close this issue or if you need anything else doing +To allow a server to bind to multiple ip addresses might be better implemented as a flag second address string or otherwise +make test for me still fails because of the wrap option +Out of scope for the GPG Plugin itself added a FAQ mentioning the requirement to update the Site Plugin instead. +bq +Im for this change but do not see how this change will stop me from using java methods since that is dependent on what library I use not what compiler settings I use +it works great thanks a lot! +resolving as fixed +bq +Glad to hear it works +Thanks for the review Jun! Thats a bug fixed it +From this information we will write the final release notes for all issues +We have a patch from Peter Kriens for the plugin that appears to fix a least some of this issue along with adding nice features like header generation among others +We could add for users who want to access orgapacheavrotool from code inside another projectCan you describe your use case in more detail Stu +Sorry I missed thisInterpreting ls output is guaranteed to be not safe if the user name or group name is longer than characters +Attached the source code for someone to commit on will start creating the components page on the Wiki +I have made proper changes in testcase +Moving this to the bucket until were ready to put it into a particular release +which could be either socialrest or socialdata +ok so its an old bug not a regression +But why have you factored out the toolbox support into a class? Couldnt it simply stay inset looked kind of messy on first impression +so if we want also to reuse that here we can share +This was resolved back in February with and in JBT JBDS and JBT JBDS +Closing all resolved tickets from or older +zip file that represents the modified contents of to include a feature as well as branding information to the to support about boxes in Eclipse RCP applications +The reason for mixing is that apache cassandra is one name the parts separated by dashes stand alone if you see what I mean +So this issue may be NFS related +svn diff was taken from root of the gsoc sadbox areaThanksShameera +This is a trivial patch +bq +The only restriction is that the bean cannot be an inner bean +Good catchThis patch fixes that printing the for tasks with disableCounting set +The number validations dont use locales +Fixing Affects and Fix in Version +bulk defer of unresolved bugs to +your understanding correct RaguThe only undecided is the location of the of the hdfs does it go in srccore or srchdfs +the problem seems to be that on windows there was some kin dof newline character on the error line and that made the regexp pattern to fail. +Note that these are intentionally not targeted at because they dont belong in the release notes +Patch appliedGoing to leave this open until we can update the SDK Quickstart instructions +Patch to define the config and apply it to the MR AM biggest comment is that this needs to be added to so that there is documentation for it +Verified in Developer StudioVersion Build id Build date +Ok +Then we will be able to reset timer on +Few things to addFrom Kishore Replication in the recipe was done using rsync +Fixed by reverting +Please and close the bug +Balancing after placement causes more overall IO then placing it right the first time +I will check what can I do +Then we should create new issue to merge them +Attached is a patch that implements the proposed fixI tested that both loadstore with secure hbase and hbase +I uploaded a patch +Docs added +The patch is fine thanks Tim. +Maybe we should backport the Bits interface +Sorry guys I am getting lost in all your happiness +How are we going to handle static in? I have uploaded this patch to replace it with local variables in the class +Yeah I will try to use tab first and then will think what better widget that GWT had for this requirement +Weve been running with SVN HEAD of Polyglot for a while now with no issues +Ramesh can you evaluate this for inclusion into +Is supporting Trash useful for users running with deleteTo me yesIve seen many of our users deleting their files accidentally +changed affected versions from and to +A proper fix may be to have storages report if they are or otherwise +Are you sure thats the problem +By I dont see any reason why the current state of this should be considered as not meeting criteria for a release +Also fixed the same problem for group by result sets +Then I figured that it was made for So yes this needs an update too I fear +one question why we need to move in both get and flushLedger? I think we only need to move when flushLedger +Just committed this +Ill check out the continuum logs should have thought of that +Move the fix to +No its not the one in the JDK +bq +Olivier can you check this +Im not sure how to active the profile depending on the jdk +Nice catch let me check +Committed the changes +Ah HAThe reason the tests werent failing is because i was an idiot and hadnt configured the query properly in the test configsnow that ive made the tests fail i can finally start trying to figure out why the tests fail +Thomas can you do a quick review of this issue so it can be closed +sounds good +Ok +I did run all of the mr tests +So you must be in the Felix directory to start normally +Heres something that jumped out at me could you break up the no credentials row into two rows like the other rows for credential settings? That would help me reason about the completeness of that table +You can get all attributes from theThe clicked row is the current row of the or you can get the selected row from sheet StateIt should be not diffence between a row with a tclink or row with a columnEventIt should be possible to reload parts outside the sheet without reloading the sheet if not its a bug +So far I did not come up with a satisfactory roadmap +When we are about to start moving files around after a compaction we write our intention in there and when were done we remove it +Works fine on the latest its not possible to reproduce in Showcase +And it doesnt help that these field names are looongI suspect youre going to have to change your design to not create such a huge number of fieldsHow does your app use these fields +The simplest would be on bit GNULinux distributions which for most of them would deploy native libraries in usrlib and use usrlib for bit native libraries +Defer all open issues to +Looks like Martin fixed this already +Since theres no ldap the auth is imported at the time the project is setup we dont have it tracking the svn auth file in real time +The about dialog is too crowded +Agreed these speedups are awesome + +I was unable to submit too long messages using either decoding path +Fixed for . +The logging will only happen on node startup and there are other delays in there so its cost is low +Marking patch available and release note needed since users will have to drop and recreate their impacted view if it was created with a version of Derby with the bug +On the other hand I would really like to see better support for Strict but I think it will require a config setting so the controls can check to see which html they should target +I applied to +Actually adding more keyvalue pairs to the MAPPERS attribute would be even better so the original names could not be changed but new database names introduced +Deferred until. +has the same issue +Thanks Sijie. +Could it be that this is an issue caused by different package RE libs with JDKand so even something about endorsed libs? This would make it really ugly tohandle this correctly +Attaching patch +The logic could be the following one If the source directory exists do nothing If the source directory doesnt exist replace all by after this conversion if the source directory exists use this path +Not sure about the latest mysql version but it definitely worked in one of the older versions +I was able to resolve it by ticking Use default system libraries in the Eclipse preferences for the installed JRE that the IDE uses. +The patch looks good to me +Otherwise the test is skipped +But I really dont see anyone ever going back to clean them up no matter how much I would like to have it happen +Would like to see this included in if possible given that its solved a problem for one of our users that has been there since someone review this +Thanks Dave first of all! Let me check the patch out as soon as I sort out my large plate of items +in rThanks! +Seeing the same error +Ill be away for at least a week possibly longer so would you please commit this one? Thanks +Paul I am this jira to you as you are assignee also on other tx bridge demo issues as I can see +Whoops! I forgot to merge the changes +I posted the results for ant clean test in the attached fileThere are several failures +The patch looks good +Update as Nicholas suggestion in v patch +I dunno that the whole server should die +I added the patch as proposed in revision Thanks for the Im extracting text only is it possible to adapt your patch against the method prepareOr is invoking prepare for text extraction required +The which is used for this test and others behaves differently from the regular dispatcher +By that way the region wont be opening on the region server any more +Test patch in a follow up +For all those errors its not rendering all that badly +Ive just filed to fix +FO files can be used to produce other report types +released a fix to master to use a busy indicator instead of a progress dialog +Fred can provide the exact version of Maven embedded in the tools +Minor naming convention suggestion Probably better to rename to or +Integrated to branchThanks for the patch JeffThanks for the review Lars. +the provided patch makes the plugin support gzip and deflate compression +According to I can assign this issue to you +Here is an initial prototype of how the API looks likeTo compile and run add to the classpathsee the example of usage in Generated Pig scripts are dumped on stdout while executing themIt is not complete and runs only in local mode but is meant only as an illustrationThe real implementation would create Pig operator objects instead of Pig latin statements + I concur +The following command strips of this folderpatch p i I tried it on a clean checkout of the project locally and it seems to workHope it works now sorry I did not try it myself patch file worked properlySince the automated tests do not exercise the commands it would be good to set up a database instance from scratch using the changed code +Need to do some testing and clean up the abstract classes a bit but wanted to get a working version in first. +Please raise the severity of to get to the root of your problem +Second thought never mind wresubmittingI can handle these current patches +Might as well. +Verified on JBT GA H Linux JDK +cannot reproduce +Please use instead +as Stu mentioned equals and hashcode are unnecessarywe can also get rid of getLong by using bb absolute methods above IMO we should treat positionlimit as immutable we could even check if assertions are enabled and wrap BB in methods that throw errors on methods although that might be more paranoid than +Thanks to you two +I have attached a new patch as the old one need updating before using with Nutch +Yikes thanks for the heads up! I like the ASAP option +fixed +Cool +ugh mvn can me sooooo annoying at times +The new implementation detects every startup as an update when there are no write options to use startup mode from if available +Forgot to mention the audit issue from above is a known bug +Thanks +Here is a patch that fixes the I attached the wrong patch before here is the correct patch for this issue to Jim Wright for this patchregardsMartin +Test failure is not related to the patch +I should have mentioned that the getEnvironment stuff is not required +Closed upon release of Hadoop. +Make a patch Harsh? Maybe add documentation to your patch that lists s way of doing it and your min method +Thanks a lot! +Among other state information isolation level is saved as part of the connection objects state information +Im not seeing this error on Ubuntu +I ended up putting back the way it is in CVS and changed to start a transaction AFTER restore view instead of BEFORE restore view +It doesnt look like the test failures are related to this change +I also already tried it out with and it works just fineSimple easy fast +Fix was included in EAP. +In general Struts apps including those that are part of the Struts distribution work just fine with Tomcat +Verified instance creation was not needed here has now been removed. +SiIll clean this up as part of my Webtools refactoring effort +This one is up to date +I agree with Thomas comment about the deadlock not causing the exception +Im not sure why but Id rather not change the hashcode if we can easily help it +can be added to further work. +Ive updated the patch +bitmap will have two bitmaps internally one for Male one for Female +I am not sure if I had an incomplete thought or I cut half my messageIn any case I like your idea of bringing jars into HDFS +Patch that allows usage of references to system properties in the idea thanks AlexanderIf no one has objections I think it should go in +The reliance on Axis which jUDDI holds has been removed +Building with ant will not create these headers +Then compile using maven mvn Dfull clean installAlso final should be out next week +Now after integrating Annotations and back into Hibenrate Core we even have a unified profile handling for all closing stale resolved issues +Released targeting Swingx +In other words I dont have a ready solution and this only serves as a reminder that ideally I would like to see things made easier for users +OK thanks! I kind of knew that but it must have slipped my mind when writing the code +I cannot seem to attach file to this jira as wellException trying to establish attachment directory +Try building the kernel project after you nuked +Running the build a second time went to completion with the same set of test failures +to make sure we give the right infoSanjiva +Proposed patch +this one gives a different stack trace +If need we have to complete another implementation for eclipse +Committed to project trunk. +to fix broken and trunk patch to fix broken a more fix for the +Thanks for the comments have addressed them in new commit +Manually am seeing this problem when using pluginflex +Attachment has been added with description has been added with description history browsing +if Hmaster dont shutdown itself it will skipping splitlog and wait for assign Meta table or root table + +As the stack trace shows it is Xerces which is correctly rejecting the invalid URI +confirmed on plain ears updated title and descriptionMarking as Blocker for ill check why I thought I reproduced itRob could we have it relay to the first War in the ear if a war is included +Oops youre right +If this is truly an issue someone needs to provide a patch +Clayton It looks like its simply having trouble finding the native library for the sparc jar +Simon wants to go forward with other patches and there are heavy cahnges in it so I need to do heavy committingThe discussion should have been here before because most individual setters are now removed this is the really only relict +Does this new play well with NRT indexingIt doesnt right now but that should be simple to fix since its just a Lucene index under the hood +Thanks for addressing my comments Jim +trunkCommitted revision + +whats tests + +I added it to my CLASSPATH in my linux home directory +They should use after flex has landed +Modulies allow to you to create different build variants as you need with only one distribution +I started out with a different hierarchy where being abstract made sense but didnt fix it when I should haveSee what you think +Just to clarify The issue here is not the character direction +I can see that +This was caused by +scan until the end of the region and then use that last key to find the next region +Even with patch wrong component definition is used if component has Name annotation and if components properties redefined two times in with higher precedence +The patch with the new components +If we add warning message for such potential problem Im afraid we may need add too many warning message everywhere elseFreeman +The bug is in a specific compiler pass it is not properly handling casts of null literals to function types +Does Derby throw twice long as the value set to set? How does user know when get throw depending on the value they set to set +from what I can see could be mistaken +Juergen +Applied the given patch +Should we fix this for? If so how? My inclination is simply to delete +Closing old issues +Thanks +Thanks Chris +Ok got it +Sure +Dave +The fix for may have provided enough information to find these dangling triggers at DROP COLUMN time +Some minor comments It is better to compare Class object instead of String +Bad jira never mind. +Benchmark to detect interface call optimizations +Lets +Headers cannot be set reliably after the response is looked at our code and we are actually eagerly committing the response before in before delegating to Tiles to render fragments +new exception make rollWriter public deleteRegion throws exception rather than returning status Replace deprecatate with throws exception rather than returning status remove main program and baggage that it carried replace close with mergeAdjacent add new method merge merging is now done by copy and compaction added debugging method listFiles make compactStores public added new private method row make create public added new public method open make add public deleteRegion throws exceptions rather than return status new private method deleteRegion uses removed method merge checkVersion throws exception rather than returning status optionally prints message to console if upgrade required +That revision is associated with which makes sense because uses which I changed in that release to make sure the port number isnt getting stripped out of the import URL. +in we should add a warning when encountering this in raw markup while in we should try to escape think there is no need to log warnings or escape the value +I will continue to create a better issue description tonight and will hopefully present a piece of code which reproduces to problem +Is this complete? I saw flurries of commits earlier +Holler if you see any that I missed +Thanks Ben works great now no bounces anymore! +Thanks +ThanksSuresh +If your view would use or something adaptable to to represent poms then existing launch action would be automatically contributed to its popup menu +For other blocking operators obv we can relax it +I did an svn update on the trunk and have been building some HTML single books and am still seeing XSL FO Document +it is already in the dictionary in both directions +this wont be done in bumping to issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +Thank you +Some of these images are also missed in the Administration And Configuration Guide of the Application ServerPlease see +although in the case Im not sure how to get rid of them +If this is found to be needed for Van will set the Fixed Version to is needed for +Since posting this issue I have switched employers and hence email addresses +Patch applied in r +which column is failing? Something in? Then I think the workaround is to mark columns as not problem with the missing join table is because associations are mismapped +the patch based on trunkHi EliTodd Can you just take look on thisThanks the same patch again to trigger Hudson +In general Im a fan of failing hard over warnings +The context names are also changed +Updated the feature branch name +Bug doesnt reproduce anymore at envirement winXP PHT Harmony revision r +Were on a machine pointers at bytes each +no response closing +I would implement an option that will make the dfs client behave like a soft nfs mount +Thankstrunk +I committed the patch there which undoes the anyways since Mikhail said the change was good for some tests he wasnt sure about trunk +actually +Patch submitted +I google for it and found Apache has one Commons CLI +Bulk closing stale resolved issues +Hopefullyover time we can tease these details apartIn the meantime I dont think that the proposed patch makes matters much worse +I added new patch inspired at fix +Ive committed a few files to the branch for this improvment +We cannot apply it to as it is adding a new feature not fixing a bugIf someone were to fork the main Hibernate repo and redo this work there as a rebased pull request it would help out immensely +LG to me +In future it should be possible to combine the Hive support into a single implementation +It also worked without error for my attempts with Java +Thanks +Fixed in rev this go into M +So this should no longer fail. +You may have a look at it to see how it should be doneThanks again +No one is working on this parser moduleAnt build does not even try to build this +No thanks Ill take your word for it +Lets wait +implementation for generator changes +Upstream deployable caches the references and seems not to update them +Fixed in r. +Patch for have applied the patch +I assumed the number issue was because of a glitch in the build not in the plugin it self +i agree that there should be means to deal with protected properties +At this point making those jars part of the update site will be great for most of the usersI was thinking about talking with ivy maintainers to pull the resolvers which require rd party jars into a separate jar so you can use classpath element +Please update if you want this reopened +I had intended to add as an affects version but unfortunately that replaced what was there. +When this exception occurs there are no SASL mechanisms available which match the caller parameters +Making Packedcalc the new Packed seems like a safe bet +also its necessary to add links to outside resources mentioned in the guide and in the guide +It worked fine +No more work is being planned for this release. +What I am planning to do is implement a unit test which uses Bleis implementation to decode a dataset that was encoded with a plate model that is the same as the LDA plate model +If you can provide a reliable mirror please let us know when this is fixed +Attachment has been added with description XSL file for markers test has been added with description XML file for markers testPlease ignore my preceding attachments +Im backporting this fix to and will do so for any existing overrides we decide to document + +As of revision r the XTS demo now comes in two flavours +Looks good to me. +Some components like the tokenizer or sentence detector do not need to extract the probabilities and only look for the best outcome +however I have no idea how to configure maven to use this +The proposed change will also change the for this kind of error so it will need a release note +Submitted code without review. +A simple fix in +Closing as cannot reproduce until someone can +Hi RobWere you able to reproduce this +Fixes redhat init script for multiple small issues most notably support multiple start calls and return valid status codes +Is this still an issue for you +Youre probably right about the missing compatibility test caseI committed patch to trunk with revision Ill leave it up to you to resolve close the issue +Marking this one as done. +The simple solution is to make the method publicThe alternative of not having write to this magic directory seems more daunting considering this bit of Hadoop changes every release +Also as far as i can see the code seems to be identical in both and for the method to where the patch for your help +We might need to clarify the documentation +Im a bit stuck with reworking so v patch coming only tomorrow FYI +This bug is old and the feature is working fine as far as I could see so I close itMichel if you do find issues again please reopen. +stdio can go over network if required but the latency might be a problemc Addressed in my proposal to some extentd Disagree +Thanks a lot for sharing your patch! Just for completeness sake Im attaching a similar patch that I created for CDH +Upon further investigation Im not sure if this is a bug or intended behavior +Attaching patch that should fix this +trunk +Definitely something to reconsider for +Ok +commit if hadoopqa is seems like failure has nothing to do with this patch +I think Archiva do it already maybe well can reuse the code +or better yet we put the prefix into the name attr +patch committed to and trunk +Still would be useful to have some of the other intermediate stuff +Nik +You can use multiple keys in a solution but only one per project. +Thanks Pat! +Moved to is good to go I guess since has commented on it since it went in. +Closing all issues +Already backported +Use multipart payloads though it might also do better even with the regular requests now as Ive removed a call leading to creating a temp one seems to link to the latest one I can see from a pom that it includes a CXF dep but Im not sure the attachment related code is there too +It seems that its not so flexilbe to configure +Unless someone else wants to work on it which is fine by me +Youd pass a method named check or check or something and it would run what you describe above +Its better to free it rather than put it into the if the events cancel flag was set +The conversion is only being done one way for now +Thanks Ivan + +David Van Couvering wroteIm uncomfortable with the build trick it feels like a hack +This is actually scheduled for +If the real instance fails to create the YAML representation is removed +An even safer way to fix this would be to have throw an exception if the output directory already exists +In the patch Im removing the old bookies registration path if exists and continue bookie startup +Attaching a patch that makes both test cases create PMDD first and drop it when completed +To reconstitute the object you would have to mine the records +The examples in the previous code are to highlight the issue with some explanation but before the fix +The problem of handling pluginplatform removal comes up a lot for various reasons like this +Branch looks goodBut we still seem to have s methods? Shouldnt that class be +However we should remove dependencies within the engineRegards devsShall we have this done for Regards I can see where existing transport like xmpp and amqp supported in the code the place where these are hard coded can be avoided +Then we might have separate datanode and namenode plugin interfaces and configuration properties +bq +Its a very nice win with little that we need to do +propertyExists to properly throw a checked instead of an unchecked when a relative path is provided +This will be fixed in the releases +Solves problem on our clusters +Looks like a good improvement to me +I am sorry for thatThanks again for reporting and proposing how to fix it + since needs to be committed to to be backported to maintenance decided to do a big bulk commit to maintenance patch for MAINTENANCE +Same as above patch but actually has a fix for the issue plus tests that reproduce it +Please find the patch attached with the moduleskeysThanks regards +Closing +agreed +Commit rev by foconer seems to have addressed this. +The patch in modifies to only modify CASSANDRACONF if it is not already set +I made some changes but I dont know if it is tested and the issue still exists in RC +But as a a is necessary for any Hive query running on production clusterPaul do you see other potential problems by changing the default value of in +Applies atop + +Also localized exceptions were used where possible. +Closing resolved issues. +to this patch and your plan +List all jars we needed for +it needs to be designed not to have scary race conditionsOf course +This patch depends on the object parsing approach in +error after importing ldif file containing containing spaces instead of after importing ldif containing spaces +Thanks Dean I didnt know thatI guess this bug can be closed +On some OSs using the RWS flag results in only a single IO per write to a preallocated log file +Burr not sure what you expect if users rearrange their views and perspectives things like this will happen +Sounds like a good case for having a whitelist in the config and removingthe blacklist stuff + +Does anyone remember how to do this +Implemented Zookeeper clients for both and +Ill discuss the author tag issue later +Looks good +Closing this issue for lack of activity since the last one year +Its only here because its the version I started withThank you very much +The problem is adequately described +Ill do this +The development cutoff for was the tag and only a very limited set of blockerish items are being for inclusion in the engineering builds that QE is testing for. +I find how to resolve this issue +Ive seen people package their database in their application then on the first run of the app they move the database file to the proper platform dependent locationI think the best way to handle this type of situation is to create a plugin for each platform that does the work and shares the same +Downgraded the severity as I think this is primarily an issue with the test rather than exposing underlying unreliability in the messenger implementation itself +Defer to C hosted +Assigning to Mazz to provide more info +The generated NOTICE will be used to update our current build system and we should notify the dependants that have missing information in their POM of the changes they need to make +This tag should not and will not contain html specific behavior +Thanks. +Fixed the indentation +Those OOMEs in the IPC handlers are bad for us +Actually this is a blocker for cloud environment where nodes are ephemeral +So they dont have to be renamedI am still going through your codeWould user be able to create table in default namespace after migrationPut in another way what privilege is required to do the above +Willem can you take a look at this? It involves and JPA +I do think its a valuable feature but we dont need this in. +Hi Russel are you still planning to wrap this up before Pig release +has been marked as a duplicate of this bug +Update usage test +Hei Marat John IDEA is now and soon version do you want me to support now +vesta is ready to be on all jenkins build slaves unfortunately not but Im also not a ruby expert +Sending srcjavaorgapachecommonscliSending srctestorgapachecommonscliSending srctestorgapachecommonscliTransmitting file dataCommitted revision . +Thanks Corinne +Hi KimThat sounds like the right spec to me +Glad to see that was useful +All the core and contrib tests pass on my machine +Thanks Paul +Instead of unloading the service I keep the dllhandler in dlldesc +To do that just execute it with the X parameter +Testing done closing this Test plan for this QA work? +Ill take a look at this patch this afternoon +Not for or I presumeAnd how would you want to do this in code extends with no added features +Attachment has been added with description Patch to add missing charsets submitted one some time ago when another Apache project asked the same thing +This is something that changed in at least the sun jdk +But if its still doesnt work for same cases then please reopen the issue with steps to reproduce +curious that this wasnt caught at the same time. +As I understand Stefan tracks case is fix personal TODO but probably a new separate issue should be opened for for further tracking +The outer most exception will be the exception declared on the interface and the nested exception will be the root most cause +Yep I figured something like that. +Are we missing any of its transitive dependencies because of that +Sending srcmainjavaorgapacheservicemixbeanTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixbeanTransmitting file dataCommitted revision . +btw +It should be and not +On that tangent do you recall why its supposed to be like that or is there a jira with a discussion +George if you can open separate issues with steps to reproduce for the other problems we can look at them. +Patch looks good +If you do that will run them as tests and not honor any of the annotations +Uploaded which replaces +should I open a new issue for it? +Look at it from a practical point of view +Yes you can close this issue +marking as fixed +Do we have any timeline by when this feature will be incorporated +Merged. +Marking closed. +I will add a test suite as soon as this is fixed +Would just branch of the branch +Can you try the latest released CXF please +IIUC maven does not compile + +Patch with unit test +Instead of giving the actual offset alldocs returns the skip value as the offset +Also note that protobuf was recently updated from m on hive +Thank you for the patch Jordan! think we need to revert this commit +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Jonathans patch should also work in the branch I believe +I also can add that sortBy attribute is ignored by columns that are put inside scrollable component +Unit but I dont get what you are trying to do with this test +start from of +If we try to do it the other way not making the writable as comparable and relying on T being comparable I do not think that we would be able to tap the power hadoop as the data could not be written as key in hadoop and thus could not be sorted +In a private communication JDBC spec lead Lance Andersen says that NULL is not a valid value for UNSIGNEDATTRIBUTE and AUTOINCREMENT +OK I opened +enum could also be moved into and can be renamed as +Work from here will be siphoned off into other projects +eclipse and the files +Committed to trunk at revision Thank janel Zhang for this test case! +It essentially just checks that every path from a fork eventually goes to an associated join +Unit test coming soon +Ive reopened reopening this one for sure +Noise +updated patch please looks good +Its been a while so can we get a short update +I will try to use in provided scope and for tests +The messages must be persistent and there must be a forkjoin in the are applying the patch and will hand it over to QE to try the test aginst to reproduce the with ant deploy ant deployProcess ant was able to reproduce at last +Can you try increasing the connect timeout in your Solr connection and see if that makes any difference? The default is try making it I think what is happening is that your Solr is extremely overloaded and simply cannot process the requests in a timely manner +Call setupJob in just before the submitJob +Can you provide a unit test as well that covers the nickname propertyAnd we should remember to update the wiki documentation +Ran tests in OKProcess finished with exit code to and trunk. +Attached test patch +Michael you should have permissions to fix this yourself if you desire +Fernando the created tag does not contain +A patch for this bug please check it out +Do you think you can make a quickstart for that? If so please create a new ticket +Is there a way to make the error message contextual? Ive been using this on and its very confusing to just see Invalid text expecting SLASH found when Ive got a few paragraphs with quite a few and slashes +Please have a look and let me know what you think +This is how combo box works within the browser +added to release nots as resolved Configuration files have been added for the use of Microsoft SQL Server as the ESB database +As those measures are not inserted the values calculated by the and will be correct +Ive changed to the name of the issue since its not actually configurable via a system property but via the xml or the command argument. +If it isnt then nothing changes +Done +In a script the method does now use the binding which menas s is available +Oops I meant not. +Weve not seen Sams original issue in a while and itll look different in current codebase lets open new issue then +Sorry this simple fix took so long to get to! +I cannot reproduce it so I assume it fixed. +If the git repository will end up being Gb could we look at reducing the size of the problem by eliminating some branches or discarding history older than a certain age +That makes sense and is exactly what I was wondering it would be worth adding a implementation as well for those where it does meet the requirements. +fix committed. +Do you say that this is already supported by the use of roles? If so howI do understand that you already have to specify what you want to access in web services explicitly so the primary focus for this feature would be the other transports +javalicenses +It may have something to do with the WTP support in meclipse but I only can make your thoughts on my description with an existing GWT maven multi module projectI am absolutely aware of the fact that it is hard to setup projects for each error report and if my time will allow it I will provide a test project but my time is limited and I cannot post the original project files thanks for commenting. +Ill put a test case together from the current Branch that I dont duplicate our efforts what did you reproduce and fix? From the code it looks like your checking if the OCA was initialize to NULL +Unit tests and have passed +Adds setting sequence number into whether they go via WAL or not etc +Regarding the proposal I will look into it +bq +Some comments on phabricator. +Last patch was more then lines +We need a working JCA in +This has been fixed in zookeeper +Cell level security is very interesting to some people high ingest rates to even more but custom distributed joins is interesting to a much larger user base +For instance if the source is just Apache licensed the binary distribution can easily include other artifacts pulled from Maven Central each with their own particular licensenotice needs +Thanks for the reviews +Sure just get him someone to send me the licence and keys and Ill add it to Xircles +I guess you need to resync again +Can happen with any new kscf if some node becomes visible between the CREATE KEYSPACE and CREATE TABLE calls +Yes we should add more test cases for parallel execution +This is an issue with the styles not the plugin +Description of patch Remove processing to expand data from of blobclob to memory before sending to client +So you have a bit system +JBDS set url to httplocalhostwebdav now instead of the url used everywhere elseAbout this issue I dont see any projectproduct difference needed or am I wrongHere it just sounds like the user should be warned files are being overwritten and be able to say CancelNo before it happensBut it sounds like BRMSDrools should implement a runtime detector for the BRMS distribution to be configuredused +Other tests are not there at +Including an option to encode bignums as strings would be pretty easy too but Im not sure how much I like it +mrepositoryUserstucu +Please read the for more information +Integrated into trunk +Done in transition resolvedfixed bugs to closedfixed +I just committed this +Closing thess issues out as they relate to the old and +Closing all resolved tickets from or older +patch from my github in r +Improvement done on projectmgr also for the update note +Yes we did not rebuild the AS ust replaced the and in the ZIPSo the sources weve built from still used to is avoid a full respin it was accorded in the meeting that wed patch the ZIP and installer with a branch tag sources everything still says which was the version before the patching +Thanks for the update Yonik! I will see if I can get this and using the same classes +Committed svn +Not really satisfying +Verified fixed in CR +Thanks Eranda for all the work on this patchThe latest test looks fine and it runs successfully in my environmentCommitted to the subversion trunk as revision +A small ij repro attached +applied +I with Designer based VDB looks like some needed metadata is not correctly being loaded for foreign keys thus metadata load failing +The build was also broken somehow so I assigned to myself +instance vs +Password properties are now using password field automatically +A primitive test has the writeString being a good bit faster than the UTF equivs so would be great if RPC moved to off UTF and on to Text +with this kind of tone I think you better off with a commercial brings panels validators BTW +B I dont know +orgapachehadoophbaseipc +If it is not just the more reason for your patches +or something like that +item can come later too +however it doesnt throw me this exception now i have mny other problemsWe can close this rd since this was given by a incompatibility between strutstiles versions +Igor please follow instructions as in the Release Notes. +Ive also included patch files off revision +I am adding incompatible change flag because +Vishal in case you need this for can you please upload a patch by friday +Its am here II wonder what JAXB version xjcs using thoughXJC uses a SAX parser instead of a parser +Fixed +It sounds like a bug in +Up to you +What happened? +Thank you for taking time to look into this. +wiki updated to reflect the answers is excellent news! Thanks a bunch +users vinodkv and anew added as administrators for the twill project chtyim added as a contributor as he is only listed as being apart of the incubator +cant apply patch to trunk +Justin didnt we fix this one along with the wfs reprojection patches +The attached patch removes usage from HDFS +Thanks both Mark and Oleg for your help +The latest release is Pax Web +jenkins dont know how to apply asrc +Changing to normal urgency +Elevated to Blocker +during initialization +I checked in an automated test for this closing. +mrepositoryxppxppmin +Thank you Karl and Richard +Hopefully we add more test methods to this base class +this approach reduces the chance of accidentally removing something we shouldnt +got to mirrors and select a random few to go to +hi brettcan you post the log from the dialog box? or any infos that will isolate the problem +Merged into master branch. +Thanks for applying the patch +on the idea +If I fix it and delete the project then I get only a BUILD SUCCESSFUL message +Such as any Windows text file +Ill make a new patch that covers the rest of the contracts later +I just thought I would suggest it +Well it is a fix but it cant be done until after +of small files is wasting system resources and may be slower than just reading a few bytes with. +And we cant really say the improvement level only by this change as there were series of changes done even in Strings +BTW as a work around if you subscribe first before unsubscribing it should work +Hi AllenFor the cluster monitoring issue we could not expect an integrated monitoring substitute the external monitoring system +We need this for setting theThis should be done together with +Ohhhh I see +Josh MatthiasI was able to make it run by overridingdowngrading slfj version to Thanks and keep up the good out the issus for the release. +Committed with changes for +Attachment has been added with description option used transition resolvedinvalid to closedinvalid +I have now updated these bundles at the same location +Ive committed this and merged to +Ill try to fix them today +if there are further changes desired we can discuss and create a new jira that represents the desire. +Provide the correct settings for Derby please +That limitation sounds too important to be left to the Reference guide alone +btw I filed as a to add more documentation +Please review +Ill upload the log file from the test +From memory there is a recommendation even outside Java not to split the line +The funky java version string is because Im using a java build that my company created internally +Resolving this bug. +Verified in ER +Thanks Olga +Setting the referral method to follow really should be the way to go about this and the library should now support that fully. +Thats reasonable +Dag +This is known limitation of the present GWT release and will be rectified in the GWT codebase +and yeah knowing when this change actually happend and what we recommend users to do for this would be much appreciated +Otherwise the treemap is really very huge whatever is your screen resolution +OK I would take this issueAnd also I would fix format pattern problem too +Is this based off of the commit message or have you seen it behave incorrectly? I tested this on PHP +You can probably test if exists in your classpath +Thanks Ivan! +Iow what else could it be +This is done via an ant task +The same input may produce a corrupted file or it may not +ah! my mistake should have done that while attaching patch +Delivered fix to in Brian +Test complete. +Hello HansOk +Straightforward patch attached +bq +Moving this out + throws a more meaningful exception +for me too +tgz of sounds good +This was already done for plugin repositories +is faster that DBCP so we should be good here +Added a Filter class +Thanks for fixing my mess up +Updated dependent features and committed with revision bbdffacabdadddcbff +Unassigning myself after commiting the fix for CREATE ROLE as partialsolution for this issue +And there should be a entry added to the features as wellI wonder if the google +Will post the patch as soon as regression completes successfully +look good just committed this thanks Terence. +M released +Basically it wouldnt be a problem groovyc has to solve any more +solutions update to trunk HEAD grab the penultimate version of the schemaClosing as not a do not quite understand because the ofbiz sources are updated to the trunk HEAD is at svn ofbiz is at svn Could you check +Thanks Rakesh +We should just ignore it for IBM jdk this issue based on Jasons comment. +Nothing is said about adding contextPath to the URI +It is very common to use to indicate no selection +In dynamic partition we shall disable customized external partition location + in works without issue if using the recommended persistence adapter. +committed to trunk and +Committed clearing of map on each next invocation. +Not sure if I mentioned this before but your config is exotic you have both the SEQUENCER and FLUSH! Im not sure if this combination of total order and virtual synchrony makes sense and this is completely untested too +Now that work is well underway with AS all previous community releases are +Fully implemented in r and is efficient after r +Can we add a that the Query invokes when it makes the WeightScorerInstead of can we just check for null +resolved at Through the use of the only the hash of the password is recorded in a file on the broker +Silvo your patch helped to identify the issue with stale artifacts in the model manager +Ill first fix in only and then raise a bug for the same change in Groovy +This is my first issue posting +WIP +the ramdir size is the allow each to grow on request +bq +as I understand it we resolved this to be a JRE if its triggered by a change on the Maven end then we should look into whether its a project with a simple case that shows that the IBM JDK fails while the SUN JDK example needs more work to show the same NPEs found in other more complex projects +That way we could print lot more information about the directory it selfHow about putting as a field ofI somehow feel that it may not be right to have extending especially adding new fields to +If we lookup only by IP then as soon as the second datanode from the same IP registers well accidentally think its replacing the first datanode which isnt the intent +Merged in the competing PR instead since it had tests +this issue without any further backportThe portion of this issue suitable for backport was backported all the way to with +Final addition that adds the properties to +I have added some javadoc to to demonstrate how to format fieldsThe rest of this goes beyond lang at the moment into the realms of convert and the text package +I think we could also solve this particular problem by putting the content to be unpacked in a directory in the plugin +fixed in SVN please check +Thank you for the amazingly fast response! My only concern is that it is ridiculously easy to get a login +Only for Ola explaining that the introductory paragraph can be interpreted as pick one wed never have settled on jruby +I have a fix for the issue the is missing the ability to configure HTTP Clients max attribute which defaults to +The latest patch looks fineTime to wake up Jenkins +curl G d mimepropjkmanager grep activation This will initialize subworkers +As you note below when applets share a common VM instance and classloader all static values are also shared +Can you commit? Take it back all yours +Attached a simple patch that seems to do the job +The problem described here is that junit tests are not running correctly anymoreAnd porting from junit to junit though being desirable is no solution as it is not just changing the inherited class but parts of the tests have to be due to different behavior of the test runnersAnd if I remember it correctly grails release notes explicitly mentioned backwards compatibility in the area of testing +Commit acebacaaedba introduced unintended side effects as it copies the platform before the www contents while it used to be the other way aroundI use a shim in my www directory to allow testing stuff in the browser +this issue needs to be fixed +AndersIs the only thing that is failing in your tests? Are you actually using it in production? So basically the question is whether this issue is still a release blocker +Committed as r +Ah I looked at the wrong jar +the fix is committed to svn repository and will be included in closed. +JIRA will simply gray out earlier versions but wont delete them +I have concerns with this change I thought the proposal from was to add +edit of my last comment was meant to add test hL to address testing for symlinksewhere is L +We can probably get rid of the following information from the test output servershutdownunixtimestamp servershutdownunixtimestampms electedleaderunixtimestamp electedleaderunixtimestampms fullelectedleaderunixtimestamp fullservershutdownunixtimestamp +Quick update I havent had time to progress these issues because Ive been working at a customer site but I hope to devote some cycles when Im back in Malaysia next week +Suggest Thilo or Adam as the it is not clear on how to handle this issue and is just a Helper class I modified the target to +Already fixed by latest version +Change looks great thanks for contributing! Mind updating to cover the new behavior +Closing all issues created in and before which are marked resolved +Were also experiencing this issue and it prevents us from upgrading from to a newer versionUnfortunately were also affected by from time to time and therefore really want to upgrade to the latest T +What do you think +It appears that I only have the asm jar files that came with the plugin +Thanks for the confirmation +and do not support pipelining +btw this is easier to test using a file repo than a test repo on sure if the actual that led to this helps but I had project depending on Hibernate Hibernate depending on JTA JTA hosted on my repoSo JTA on my repo couldnt possibly be referenced by the ibiblio Hibernate POM and so m needs help finding it +without that the test is just testing the select operator and this is tested elsewhere in the same file anywayso i took out the old code and left a comment that this test needs to be enhanced to cover more comprehensively in the future +Hi Alexey I cannot reproduce this problem in the luni nor the nio modules where your patch suggests you are seeing the problem +I agree with Ning +fails because the enhancement is not +Ive added the changes entry. +we can keep the backword compatible with some small patch +So it hangs with some specific clientprotocols but not with allIts well possible for now trunk does not work with mac clientsAs we changed quite a lot since beta could you try with a trunk snapshot and report the client details which hangThx I think you are talking about hanging on imappop which is not the topic of this jiraPlease check in the open jira if one corresponds to the issue you describe and add your comment on itIf you dont find any please open a new oneThx Eric +Now if more racks are added to the system will HDFS automatically rebalance the replicas to conform to your logic +Good point will follow up on +The former is bash specific and performs pattern matching which is not needed here +the point is that this constellation doesnt work but it we should test it with jboss as +Error message only shows up onceIt happened when a large number of files were deleted on the cluster +Given this situation continued support for xmlrpc should in my mind be continued +Patch is in parent bug +These class loading changes will not be accepted for as they change the semantics of calls from throwing CNFE to returning null +test for CDI project is added +Ooops sorry +Please review are not sure why this is happening but people seen it occurring thus we will have to bite the bullet and play on the safe side an addcommit +I would like to hear reasoning behind this new feature request + for the patch +The patch has been committed in r. +These will override the automatic versioning +what if i rename the folder? shouldnt such checks be done on every start of the server instead? And if the folder is missing informask the user about it +Thanks Adam +Ive also added specs in rubyspec to check the different behavior between ruby and commmit think that this issue has already been fixed in but I committed the regression test in db +Any further action required on my part +svn ci m Applying my patch from moving to using nanoTime under the hood +I vote for a dedicated thread on each name node using of disk IO to validate blocks continuously +An named will be automatically generatedMoreover the patch solves bug in the internal logger which does not remove its service listener +Merged into master. +Can you please release the new version +with JBT. +Added checks for stored identifier case for catalog schema and table name patterns when retrieving tables and columns from db metadata as well as more informative messages for join tables in Git ID fdbabacbdfacceba +The point is to have a starting point for different areas that others can review and improve as approaches perfection +Just had a quick look over what gets packaged in the releasesSrc version the various libraries used in widgets connector and features such as etc source code for WAR version packages src plusStandalone Same as WAR plusIf we remove the ASF project jars we just getWAR version packages src plusStandalone Same as WAR plus +Patch attached at +With this change we can view like a workload generator only +Thanks Jasha for fixing my breakage of the trunk! to the fix although we should discuss it with Shindig if they shouldnt upgrade to using a compliant json format +Thanks Konstantin +I will have a look on why the tests are failing +Perfect +The in branch RESEARCHCLIROXSPRING seems to be ok +Doesnt sound natural for me +Ill wait with the patch forTo work with right now on already existing projects +Updated patch with documentation changes. +I think this patch is an improvement although it takes a different approach to the issue than what was reported +Verified on the latest build Works fine. +The correct behaviour should be look in its configuration if not found look for its parent and combine its parent and seld defined properties to make the correct set of the property definitions +original in rev +Looks good to me +Old JIRA patch preserves backward compatibility in constructor by wrapping a float target or default in a +So it may be that some ownership changes are needed +The trunk patch looks greatThe patch Will give some really weird numbers I think +The TCK already has a number of binary property test cases so do we need another one specific to Jackrabbit? Whats the added value of having and running that test caseIMHO we shouldnt just test things because we can +That seems fine by me but Im curious why you chose to do it this way +also assigned to in jira. +Hi RishiI did that because of Deepak comment at +Cant reproduce this on trunk or in +the usage of generics breaks the too many idea frames open +This is one of the situations that was fixed as part of the mods for + +here is a patch and the extended Simon I took this one and will commit soon +It was released with the fix in +checks are applied and tested +Strong for the functionality Ill be more than happy to review the integration from Sqoop side +This was required for the logarithm udf function +As I commented in I would prefer throwing exceptions when a filedirectory is deleted during listing +Please reopen this issue and +We have Struts final working fine with WAS under NT And anyway shouldnt this be considered a bug and not a struts bug +I mistakenly closed this issue while there still was more work to be done +There are few unused imports in patch +No changes required in the Admin Console code +So here you set after read to the code it will become +Attaching patch that relaxes the sanity checks and updates the documentationNote that while working on this patch I discovered we never enforced a stated condition for the merge join which is that you cannot call on the data between a load and a merge join +Again the suggested change might seem proper but really it adds nothing to the user and only hurts usability +fixed the lines longer than the patch results even bigger because there were lines near the changed code longer then My Jenkins is green +HiThis is because of which is a new feature in +I want to get this working in branch and trunk +No thats not necessary +Thanks Wei! +bq +All effort goes into jBPM +This would be great to get in place for the webinar +Thanks! Nicholas +I assume the bug is in Verifiyerget +I have tested all of the seam example applications and the testify test suite and have not ran into any unexplained EL errors +pbd the compaction description added to the WAL +Why arent they in the poms directories +If rule is specified valid values are REALTIME and PSEUDO default REALTIME +I will have a look at that you tested the templates provided by the content assit in +This is a Grails application +You lose the ability to use language feaures and +I agree that ideally we would use a template but we could add that later +XTS now builds again + can now be publicly constructed as of just released few days ago +We have however used this patch in our local version for quite some time and it seems to work just fine +More patches are attached +Fixed for revision . +Agreed that it will not break anything if this continues to use CRC but it might just be better to use the same as in Output stream unless there is a compelling reason otherwise +closing since fixes are in released code +The prebuilt WAR contains this may be causing the issue as itll get picked up before WASs servlet API +test no longer exists and there appears to be no reference to Jason in the code +use asf masterThe environment field suggests Hypervisor with SP and with Support Pack +This can actually be reproduced consistently by undeploying all wires for a reference +Fix for and +Is Dynamic Web Project not the correct project to use to generate a Web service top down? Which project type should we use instead +Nice +Hi Jake thanks for the patch +You are rigth your code works correctly +Please close if this is ok for you +The graphs represent the cumulative time for adding new users to a group +I have used too but it cant solve my issue +Meant to say that +The waiting interval is about of the assignment time out +It sounds we are not agreeing on terminology +The headers and footers are now fixed thanksBTW the server is responding with the headerServer nginxalthough the page footer saysApache Server at Port this a problem? The reason of this is because nginx used to be a cache server which serves downloads +Marking this as a blocker after discussions with Mukund and Sameer +Also updated properties view for service name and operation name +I agree with the approach Paul +YMMV +This patch will fix the problem +Initial thoughts on this is a region slave can get notice from the region owner via zk that a log has rolled and process the new edits from there +Hence closing the ticket. +It would be nice if this were possible with the Java library as well +Closing issues that I resolved in already released versions +The initial description seems to indicate that you will provide some extra syntax to allow the creation of unique indexes which allow nulls is that still the case? I didnt see any parser changes but again understand that this is a preview patchAlso what is the intent of the change on existing indexes +Im attaching the files separetely since jar and tar files can not be part of the patch +So we will be able to use some of the features from this release such as changes in NTLM authentication etc +While I am at it I am also going to this for Append +I would be really interested what you think about doing these two things +Ill try to run the with the version and report what happens +That sounds like a bug in existing trunk +Im dealing with this as well using Grails +Ok guys were going to revert this behavior and with an meaningful error message will be raised with when a provided source directory doesnt exist +to use if we do nothing else on this issue +By my Timetrack apps lib folder do you mean lib? Is your init servlet loading beforeNote that Struts uses Commons Logging which will usually just find Logj in your classpath and use it +OK fixed on. +Thank you again +I added a param that defaults to trueIt fixes the issue +I forget it was used in two places when I fixed +Good on you lads +Then the value attribute would behave like any other Component valuebut the handlerId attribute would be used to name the Handler +perhaps to override the bundle configured in your validator configAnyway id rather not add something i only so ill probably just commit your patch unless you or someone can vouch for the purpose of the rest +I will fix and later +On my all tests pass +please suggest me what font i have to used and what methodPlease give me some sample it will be very helpful for me to close this issue. +Marking it as Wont fix because we wont make efforts regarding that request +Jeremy sounds right +I couldnt reproduce it either this time. +Please rephrase +if curious about your comment about reads being file based +Check please again +The classes start with capital letter +Great thanks +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Not sure if this is an issue with just the webview or what +Committed in r +Correct its the same with me +will commit if tests +Hi PavolThe cd command should not open the resource in an editor +Used valueOf instead of creating new Integers +I will add this to the pom soon +Grails by default generates the web if you dont want it generated you run and specify your own +If not please reschedule to +Sugggest you raise a JBLOGGING issue +I miss some code +Should it then the remote server? Users often complain about this +Committed to trunk with revision . +ChunhuiPlease check the issue +A project that we have no dependencies on +I fixed it and uploaded a new patch +I did a test run with a simpler new logic vs the old logic by running in a foor loop with a size of New codeOld codeAs you can see there is a flaw in the old code pickup the last element more often +While away I will have limited access to +What OS and JVM are you using? Does this also occur on other OS or when using otherIf the error occurs are all your connections lost or just the one you created? +Hey Todd in short I agree we should be looking at moving performance improvements over to the main FS implementation +Thats why I vote +Ill try to find a xalanj rpmIm not sure jpackage started to work on it yet +Cloned to track for EAP +patch conversation with Tijs Rademakers about this +Great work guys +Found the root cause Map parameterization was being lost when trying to locate serializer for contents of Collection +I had reopened it already some time back + removed the lock as its not needed added comments for when it might be needed refactored app application to make it consistent across the file added a couple of logs to make it consistent across the processing a few Eclipse code style +i was using whitespace so best +LamonYou should be able to create new issue after the matt reload the new migration i did today +the last patch went stale with commiting of +Cool +Username nobody +This should report an exception now +I didnt know that it was a known issue so I figured Id open a JIRA for it +The carrot contrib is not really an integration in the same way its a fully formed solr module we distrib all the files you need to run it in the contrib +I had a look at that patch and it appears to be a subset of this one since the patch fails two tests in this patch test and test this patch passes all tests provided in the patchHence Id like to mark as a duplicate of this is that okay with you +Hi StanThe tests will need some type of public repo to resolve things that are not builtowned by Fuse +Bulk move of M to move from to to to +Reviewed Adams patch and Im on applying itdch The fact that you can kill couchjs to free up the VM is quite intriguing +This was removed because it used some that were was originally hoping to use much of the same code I wrote for extract constant but it seems that the analysis that JDT uses for its extract method refactoring is incompatible with what we need to do for I am going to revisit the original extract method use what I can and rewrite the rest +Patch to switch test to use Italian instead of thanks! +Trunk patch is also applicable toSeparate version for + +Repl queues will then be processed later when the NN does automatically leave startup safemode once sufficient report +Since we should not have a static getter I am saying we can use and get the value of Id agree with exposing it publicly if documented +I would like to debug this but cannot do so because youve pasted the sources files into the bug rather than attaching them so there are linefeed characters in the middle of tag names etcPlease attach a copy of the source files to the bug report +We have improved our validators and now they work faster +Sorry attached the wrong patch +A primary key class must be defined to representa composite primary key +Retargeting to fix in since has been shortened and made more focused +The same not is true for URL based deployment of the VDB where the name of the VDB file is unknownAlso in order to do this JOPR needs to load the file inside the VDB to determine the version +Perhaps the out channel for the runtime starting the console is getting routed to the real? It is peculiar to be sure. +That is the case +Mamta I have the text about PUBLIC and privileges re views triggers and contstrants +The patch doesnt appear to include any new or modified tests +Bug to tomcat was created +I think lowering the number of search and index iterations would be fine for regular builds and nightly builds can keep the current number of maximum iterations +is a datanode servlet but it is misplaced in the namenode package +The QJM is now in +Here is the patch for irc link in developers pages +Thanks Sameera for the contribution. +Please find all in on +Therefore we have to take special care on them at refactoring +Have you verified this is still a problem on +This is critical for us and applies to as well +Thank you for pointing that out +That worked like a charm! Thank you Jun +This version contains a little test to show that actually decreases in size when theres memory usage +You should see stack trace on RajAnd sorry for the confusion +Added to the repository +I will commit it shortly +The developers parameter of the mojo does not get populated with the values from the project pom +I think the patch may need to be rebased a bit but the contents look good to meSo pending a clean Jenkins run of a rebased patch +Almost thereCan you please try making that top layer more uniform these horizontal line segments having them in the center to be at similar brightnesscolorscheme like the ones in both sides? This would remove the feeling that something is missing at the center +However I think using logical names in the configuration keys than physical address is a better idea +Verified by Sean. +Add comprehensive unit tests along the line of what is code base so as not to depend on. +Can you please start a thread on dev list regarding addition of this utility as contrib? This way we can all agree where this should go long term and where we the best place to host this code shorter term +committed to trunk and branch. +Someone said they would work on a patch for this +Yeah this would be nice +Agree +Same here +Same asFixed in nightly build thx Tim +Looks like Adrians this should certainly be taken into consideration +We also need to handle NOT FOUND errors and the like +David I am going to try the solution of Jrg with entities + contains a workaround for this issue +Committed to trunk +Updates applied +Well I wasnt sure what to name the output columns for the explode UDTF +Thanks RobertPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +Im not sure about findbugs nor about the checkstyle section so I havent checked that +However before we decide to enforce or break that we need to ask ourselves whether the new behavior is the right thing to do and how bad the old behavior is +Is this ticket still open? is now closed and commited +Displayed in result if found If url field is filled it is used as href on the title link else fallback to fileresourcename or to plain id Detects file type from contenttype field with fallback to filename suffixWill commit r to trunk and r to suggestions for what we should tell people to index to test? I think the most fun is indexing my own docs folders I was thinking instead of bundling some synthetic docs in exampledocs we could use a dump of the web sitewiki javadocs or some other real docs? +Thanks! +thanks gavin and yes u can remove the obsolete amber project +Im markingthis as WONTFIX for the reasons mentioned above +Im still trying to understand the need to initialize all the beans in the factory? Whats the purpose +Touching the files doesnt seem to have worked +I only committed the fix to trunk +through failure is due tobq +Though maybe I shouldnt have focused on the exception messageTo people with alot of Java Tapestry experience its probably pretty obviousbut I spent a number of hours banging my head +Thanks Anurag the v patch looks good +However some people use both Gnome and KDE +the directory path is dprojects default Did not override in Your test application runs fine on my machineWhen I add targetin the error is gone +As Niall noted in theres a corner case where the added code can throw a +this fix uncovered the bug in the test case +For MRU Im not sure whether it will lead to better performance or not +Are there any updates on this issue? Weve similar problem with and would like to know if its been fixed in later versionst this fixed now with the new in version? Havent tested it yet but will any time soon +requires and +Thats exactly right using the Named annotation binding from a properties is the preferred way to go about this as is done everywhere else in the code base where a global configuration value is used +Once I get the patch integrated Ill build a small sample that shows what Im talking about and we can see how to solve that at the Axiom level +These arent unit tests +I removed the patch since something modifed was commited in how And also because I wanted to check at each step before commitingbecame And also because I wanted to check at it step before commitingin my brain +The old directory is deleted prior to the new directory being moved in +I can confirm that the bug is fixed with the latest source in the repository +Committed to trunk +Patch attached +I dont think it is a big deal and it certainly doesnt make sense to hold because of this issue +Could it be that they are broken +Please could you confirm if it has been applied as expected +We may not want to put it in but we may want to put it into the product and if so well need to decide whether it belongs in or +It has +all guides are checked +In that case I think we should suck it up and take the backwards incompatibility to fix thisYeah I think I agree really +The thread on mentions two where it might prevent a race condition upgrading and the database +verified on Version +defer all issues to +No user feedback +Im assigning to Daniel but assign back to me and I can look into it again probably after is closed +Implemented for in master branch at github. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Complete source generating stale resolved issues +Marking closed. +Hey Bit Are you interested in providing a patchIf you do so this week Ill integrate that for the next release +Details mailed offlist to rbowen + +Ok I did some more tests using confirm that the simple properties like title queryable are correctly updated but if you try to change the defaultStyle for instance the update is simply curl u admingeoserver httplocalhostgeoserverrestlayers tmp edit the defaultStyle using any other style name present on the data dir curl u admingeoserver XPUT H applicationjson d tmp httplocalhostgeoserverrestlayers +Ill let you guys commit to at your own pace hope thats OK +now it works +We are using the new compendium build now in feature core +Actually this stunt might have the potential to eliminate the need for +But make +Bulk changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Example project with the encoding issue in be i but is i +I reverted back the proposed change to +The size of all primitives except is defined in the specification and constant for all time and platforms just like in Java +lgtm +im getting stung by this as well in a huge multimodule build this patch would make it easier at least to grep for something more specific than repository +The file Ive been trying to send is the example file provided with the distribution called after its been transformed by +Thanks Miikhail +Hi Daniel thanks in advance for your supportI send you a simple eclipse CXF webservice project using spring and a simple eclipse CXF client project +I restarted the nodeA server but the nodeA automatically join the clustereven if i install new +Unfortunately this goes deep down in Geotools and its notRan into the same issue here too would really like this to be option would be that if it doesnt get fixed it would show some logging information throwing some kind of not implemented exception +The contribution archive +But when I tried to apply the patch found conflicts +Will commented that the patch is ok and he wants to put it in +If you are going to get irritated at least wait until you have a reason +reopen for set correct fix change for resolved issues that pointed to instaed of concrete +Seriously nothing goes to nexus release before we got these things cleaned upAbout TP tell me why other tycho projects uses it over then? and is not something you can describe in a mvn project a TP you can +The project was originally created using grails +Composite queries can then pass their children to the processor during their rewriteFor backwards compat and simplicity Ive created a which directly calls rewrite +Committed the bulgarian tranlations in Rev +This is usually caused by an older JVM loading a class that was built for a later JVM +Reopening for fixing the fix version +STEP src Apply the patch inside QPIDCPPROOTsrc +The wiki works by using a special tag +This new extension will work with Mozilla and Firefox versions +So I believe the formula should be good +Hi! Is there any movement on this? Any updates on perhaps what has been decided +Please see. +Then when new kv that has an extant key but whose value is different the key object in Map stays same but it looks like the new kv goes in as value +On failover we can trigger all the blocks of all files to be copied local +I realized that some of the classes which were earlier had to be made public and therefore are lacking in class comments +should be good The current tooling we have only allows it to be sourced from the latter right now +Im ok if the user is asked if he wants these additional bits we dont support but it should be a choice not a we can do for that +From a logging perspective the easiest solution is to just add a retaining handler that keeps the last N messages above the given threshold the console would then access this handler just like any other to get these messages +my last comment a patch and our global my last comment a patch and our global it remains unknown whether env vars are supported now by this env vars were supported previously and worked with meclipse for many revs this is a forward to your best +Ive just created using your config and can confirm that your scenario works in and fails in. +No problem +DBRE gets all its information from the object obtained from the JDBC driver when it makes a live connection from the shell +Eventually we could turn the field access into a utility method that uses the geometry descriptor instead +I just added a call to purge in an undeploy method which should fix the leak. +This issue is fixed automatically once first issue is in +Thanks a lotBest regards by Richard. +This is a basic template to help you get started +That makes sense I was hoping to deal with this during the transition but as long as we are dedicated to removing the properties after the transition this is a better plan +huge. +This appears to be being held onto by the thread context class loader +svn ci m Exposing the unescapeXml and escapeXml methods that take Writers +Sorry it still not workFinallyIts work. +My changes to the fsadaptor for should only affect perfomance when an immediate change folder is registered +No sense living in the past +Details of the failures +See also for an earlier issue and related discussion about how we could allow Apache committers close pull requests on +Reviews and tests are always welcomeChatreeAs Japoco outlined it introduces a risk and for now I prefer to put this aside because it adds not much benefits +Thank You for pointing this out +Defer all static analysis issues until someone can be found to work on it +Should probably call it quits for today and look at this again tomorrow +The purgatory size sets the buffer size +patch looks good +Sorry for letting this sit so long +The ID here being the uniqueKey +fix typo in code file name +rar deployment in +Integrated to trunkThanks for the patch SamarThanks for the review Ram. +Hi MirekCan you please upload the thread dump for me? ThanksHoward +bq +Im fine w the quick dirty approach for now +Ill check this next days and will comment here with the results +Reviewed all these issues and marking them as closed. +I just committed this +The javadoc warning was from a remnant of +What is the prob happening here? +what you think it should be +GreatThanks Carl Amareshwari for seeing this through. +Also attaching patch here too +Thanks Thomas for researching and providing the information on the issue +I will attach a of my program when I manage to strip out as much as I can to simplify the failing scenario the code is a real mess at the minute +I am going to remove the patch +The reason I chose to start here was that I felt it required the creation of some of the base entities required for and +It has a much better integration with Hadoop +Simple patch that just calls put methodI havent tested this patch but I think it should work +Mike what your thoughts about the following suggestion +Patch seems to work great to fix this issue +Yeah I thought about doing the same +Patch applied as described it would be handy if Martin could test andor tell us a server that has trouble that we should test against? +Patch attached +No I mean that after I build the smoke tests dont pass + +Public getters and private setters usually indicate that +It is being attached here to grant the appropriate permissions for inclusion +The above hold true for all or objects that interact with values provided by the developer +I dont know what to say +Simple fix replaced with +The addition of the suffix version of the method is just to keep everything thats in one spotGora could just as easily append the suffix using the version of the method. +This is not obvious and the script should fail fast with a useful error message rather than hang +Patch committed to both trunk and branch +Note the download page mentions how to check hashes but there are no links to the hashes and there are no hash files on the distribution site +Maybe related to +All fixes and tests committed to trunk +Ups +how about slfj? so that user can pick their own provider? like logback +I think we can remove the mechanism for number of acceptable findbugs release audit and javadoc warnings completely +current input is like Where is the sleep time in minutes between kills and tup and is the sleep time between tup and kills +There must be a bunch of classloaders in the mix that can result in multiple instances of that static member +I will take a look at the code and let you know if I can find the place to change as you have suggested +Closing verified in ER build. +reopencomment if you got some info to shed some light on it. +Is there really a need to store that much data in your session attributes +Patch for the to try using instead of to get rid of all the extra dependency stuff +Good idea +Alan could you resubmit your patch without the path prefixes so that I can actually apply it? +Not that that is exactly an automated test but +Already NOT REOPEN ISSUES FOR ALREADY RELEASED open a new issue for this. +PS Is the issue really error reporting as the label indicates or is it an issue with not being authenticated anymore +Isnt that the case? + +I thought that Christians suggestion of a method to configure path to svn binary used by eminently reasonable but noticed it had not yet been acted upon. +most likely as part of the rewrite of reduce for +Basically I wonder if there wont be a number of scenario where because you get a bit behind on L once then the IO you waste doing STSC in L will help you get even more and more behind on your levelingThats exacty the case which is why we only apply STCS to L when its fairly badly behind +That was an issue in the test +Thanks for the patch Clint! I didnt knew it was that easy to find the root of the pattern +Er make that +I dont think the curren provides that type of functionality but thats really what takes so long is the rule construction for a digester +I think this does what we need + on both and patches +this seems to be fixed +How does it work in hadoop then? From what I see the externally visible behavior of JT and RM is identical in both cases +Im going to change the component of this defect to Blueprint +Ill dig and see if I can isolate it some more and either update appropriately or report the issue to the correct group +I am now working on the CSS and will patch and go as I progress this MOTD issue will get sorted as a side effect of the revamp +moved to since we ran out of time in priority to minor as discussed above +Roman I didnt notice tha the code box has a scrollbar on the in my screen the others were truncated +It makes sense but Im wonderingwhether this means that the cleanup is in the wrong place +There are where this plugin is used where srcmainjs files are processed to targetclasses thats the default for this plugin +assumed fixed a while back. +I like the relocation of pigtestconf to buildclasses +I dont think it has ever worked +Same project backported to think this should be fixed in +It is likely that the failure of one of the two tests was caused byWill try test run again after that is fixed +Brian merged the PR. +I see this too I think I put a pause in there once before to make it go away but now I have this faster computer and I still see it sometimes +I deployed new snapshots +Attaching new patch removing from a catch block of shutdownHook +The groovy shell simply failed to start +Please verify. +Comments on the patch Normally does not appear in a patch +All I did was create a Roo web project and deploy to my ISP who had me on Tomcat +Assigning to Gytis so that he can track whether this is fixed by +Good catch Rich +Say in jboss portal GAwhen i create new file after clicking the create submit button im in the Manage Content screen where it is possible to click on the language link +attaching patch against more complete patch with exception handling in eip endpoint +I plan this bug to be fixed in ve added a junit test reproducing the problem and the test now passes successfully. +Unfortuantely I must use this library due to corporate policy and cannot change it to avoid using env vars +Roman the patch is fine +Riftsaw and pull requests have been applied but will need to wait for next release before available for switchyard +Here is another situation +Im all for restricting the release authority to one person and I opt for Vaclav +It looks like the problem reported in the subject of this jira is now fixed and the remaining ignored test was incorrectly reporting a failure so ive corrected that +The +having them check in alot +The point is that chaos monkey actions shouldnt be trying to avoid exceptions +on other platforms than linux +Some are commonly used by people others less fequently +Standard prompt was displayed and I was able to log in with no issues. +Additionally Ive updated the documentation on the websiteThanks to Antony for his contribution +Im gonna be honest that from my location I am having trouble checking out from SVN so all my looking is being done remotely and my comment here is further based on the branch +This should be done now +sorry +Only in wrapers for protected and public methods we set caller as current class and in wrapper for private method validate that caller is current class already have private method which work with private property do not generate default gettersetter for this private property +Also to unit test a part of improving the test suite we should disable state transfer for all tests that dont need it +Giraph Issue Patch to add Identity Patch updated and corrected checkstyle I have corrected the checkstyle violations and mvn clean verify does not show any errorsThank you +release the lock +Does v fix the bad decompress that you found above testing with PE +With the work Ive been doing on Decorators I am close to a solution for this +Native SOAP stack is not supporting JAXWS any more +It is simpler to simply create a new class called I dont see how this is simpler +Simple one line fix +Here is a better patch that puts back the test suite as part of the tests and also fixes the broken test +Perhaps an answer would be to have a stylized mammoth next to one of the wire pngs? In the end it would be cool if the mammoth logo was ubiquitously associated with Accumulo in the same way that the penguin is synonymous with Linux +Using common sense I would condense multiple sequential wildcards into a single wildcard then process +There were some changes in the internal environment handling +Still looking and testing but it looks good so far +Any examples that actually look decentWe can improve the site design lookfeelnavigation any other time for now getting rid of the Forrest mess would be a big improvement many more people would be able to do a full build package for testing +if this issue is done please resolve it and create a seperate issue for remaining issues +I have misgivings about the original Derby design which makes an artifact of +your comment really makes the implementation quite easy +This was fixed by the to refactoring a while ago. +ehh +Since jirb was working fine before that change the change itself is not acceptable +slip it again +I think Id rather just leave this one in JMX rather than add it to nodetool +Complete patch including the missing files. +Sushanth has graciously agreed to pursue HIVEm going to mark this issue as a duplicate of since it attempts to correct the same problem although with a different approach. +It is the only classfrom this branch of the Boilerplater hierarchy that the Build script dealswith directlyAll compiled code will be linked into a single shared object which willbelong to the module invoking use Lucy will load the entirelibrary +war file must be loadable immediately without any customizing which should come in a succeeding step +We dont use it BP anymoreWhat do you mean by that? I looked at earlier today and saw BP there +In your case you may create and instead of +sql script to create db structures and populate with test attachments as unified patch +The stars didnt work properly +The reason I find it extremely unlikely is if the call to STOP scanning goes through odds are very strong that the call to scanning will also go through +Im not sure why its generating multiple Foreign Key Constraints +Thanks Patrick +telling me +Igor is the only reason you need environment variables for +This appears to only happen with the HTTP Protocol and not the SASL protocol +Reviewed all these issues and marking them as closed. +Have not seen the patch yet +Hey Koji does the work done in meet your requirements? If so we can close this issue +We can ether change it over to be sessionStorage instead of localStorage because it goes away after the session ends +What I take Swapnil to be saying is that all properties should be in the form and the scala variable should always be the same thing but in camel case aWould be great to do this in with the other compatibility changes +This sounds reasonable to me +Thank you +fix for after on after +At least for Jackrabbit I suggest to resolve this as wont fix +It does +with Sonatypes repository hosting or bundle upload service you can still deploy your artifacts to maven central +I think we will need to rename some packages and whatnot +Thanks Kate! +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It is a melting pot of and that I provide as is +Marking that this is a limitation of the release +If you have time please take a look at rev +I am for merging +bq +Good point +You can also replicate to a server if you supply the proper credentials in the POST bodyIf you do encounter the problem can you include the syntax you used to trigger the replication in a comment? Thanks feedback in a while +Changing it on will probably break existing apps it wont be fixed in the branch +Hi Sebb see the attached file txs +pr approved and merged for in JBDS B +proposedIt seems this will help though I cannot guarantee it with accuracy until I my current approach based on +Sorry +Cleaned up tests locally +OperationsTypeMeanSDththththththPSBBh +Move the codec package up out of io packageYeah could +Thanks +Attaching the common patch +Patch is attached +Include patches from on this one? not on the others described in? Strange +this is now committed +The and failures are still unrelated and are tracked by and respectively +Thanks +Address all the latest changes seem unnecessary +In general it will take two read check cycles before the client gets booted however in your case since the client simply connects and then does nothing the inactivity monitor has settled and it ends up taking a third cycle normally if you client connected and did some other interaction with the broker this wouldnt happen +This looks looks like it is related to or a duplicate of. +Anyway Ive delivered fixes for this +We should document this in the installguide +The patch for review +Thanks to Ross for the contribution and to Gavin for preparing the initial plugin. +PR applied on master and branches +Updated file based on Scotts suggestion. +Attaching patch that has been generated for trunk +Ive never seen that test spuriously fail +Well take a look +Okay I give up +it is used to allow regeneration +Maybe consider adding pretty printing later. +Great feature +The older clients were not setting the property which is always present in the newer version of the driver +I guess b was just a weird OS X build of Java causing errors +So you would have the same problem withhmm not sure if I understand this correctly +I have a patch for this Ill be adding within the next day +Patch provided Hiro thinks it looks good +sanne no that does not but that duplicates names in the not so uncommon multiple constraints needs +Ill check it again with CP and then close the issue +Extending the Trash API might be ok in the short term but does not sound too appealing from a perspectiveOne suggestion to solve this problem is to have a API that allows applications to register shutdown hooks +Make the test wait till all replicas check in with the namenode +i tried to reproduce this scenario that he mentioned +As said this fails for me with Maven and. +already provides a method for generating the contents of +I cant tell if this change is making it worse or the code is already just very confusing +Committed to TRUNK +I dont believe it should filter out tables +A stripe is like a +like serviceNamespace service endpointNamespace endpointIf we move these method into URI can be a good URI citizen +We should simply backport +bq +Any update +Richard patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I added the fix inchangeset +Other imports I found are affected +The README for this one will also need instructions for AS and EAP +pushed in Developer StudioVersion Build id Build date +This patch replaces the bullet item in about runtime compression with a warning about keeping enough space on the disk +Im fine going the route of getting container status from the RM when required +this was fixed in +committed as of closing all issues that have been in resolved state for months or more without any feedback or update +Would save me a lot of work Ive otherwise been doing on andIll take over this one since it looks to have gone stale +Token Storage class should be +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Ill make a pass over this soon +Have you tried +Will work on getting those other patches in now +Ill provide a patch later if no one else is faster +I the uploaded File there was a new A bytes after the original A +The issue does not appear in these cases +LGTM +The drawback is that if the index does not fit in cache this may impact seek performance +Commented out failing compiles as they dont seem to affect results. +Would you be willing to dig into this a bit deeperDoes it work with the IBM JVMColm +Im thinking the shutdown hook could be not used when we terminate normally so the successful flow of events does not have to depend on a shutdown hook like it used to be +Thank you for the comments againThe situation you have pointed out is when the database is +My bad +As youve pointed out its been done beforeIs it really installs? I thought it was two GanymedeWTP and GEP +Short was fixed in commit A new EXIF interface was added in commit and fixed up in the subsequent commit +I can confirm that since we pushed to production we have not seen this exception. +I will be organizing my thoughts and questions over the next couple days and will hopefully have some thoughts for you in a day or two +Thanks to all for the inputthis is now fixed in revthis is a good point thanksRev +Patch with the above code +proposedcommitted to as dedcfbffccbcaecdcbfddbea +Committed revision +Ill commit this in a few days if there is no objections +Thanks for your fix +hpp. +A future improvement in this area would be to include crypto ibrowse mochiweb etc +This work should fix +Verified on JBT. +I have renamed the option from loaderCache to contentCache +Im not sure if this is a problem unique to our environment or a general issue +ant only cleans datas from the derby database not any external database +close this issue. +Tests passed in my environment too +Right but that means releasing the parent which has references to some snapshot plugins which are part of the project + +I try to do good Job This time I think I got in too much hurry +Im investigating some other avenues because even if we are willing to unsign a jar on a case by case basis it seems that isnt a good solution in the long term +Works well Evgeny but the type of the xpathQuery property should be TEXT +I just committed this +I skipped javadocs by purpose because I thought similarity with the well known stuff would suffice by now +I like its very clearNo strong opinions on the thing I trust you to do it right +Committed in branch and trunk. +Im fine with the exception from your test +If you have direct access into the node to start the external system you are trusted +defer to +The Unicode Standard requires decoders to +yeah the linux distro +Perhaps use PermissionVERSIONMNGMT after all +Closing. +It seems that in Beta there was some as dependency in ws feature but was not sufficient reason of the original issue +Please +are you sure you have got running LDAP server on localhost? what kind of LDAP server are you using? can you connect to that server with any other LDAP client? is thrown when the client is unable to communicate with the directory or naming service +Has been fixed in. +The jclouds documentation is a bit out of date it should be not are the relevant and coordinated properties I use to launch a private image in for provider stateavailablemachinearchitecturex +Thanks +update versions pushd build devnull sed i sVERSIONtagg ant f popd devnullI think this section can be updated to fill in the current date and update the substitutions used in the docs target +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the codebase +updated patch to actually check the header +The server part is the demo found in the distribution file +It also Adds a constructor that takes a userlog subdirectory URL +The check to see if any of the node types to be unregistered are currently assigned to a node will be fairly trivial once queries are around the release goals and numbers +It is not useful for our purpose though which is the reason for +Changes made as discusses this morning + +To configure it through Felix look for configuration named Apache Jackrabbit Oak ServiceIt would register a with property think this can be resolved as fixed right +So maybe I am wrong and they are not really options after allAnyway +This has been completed +seamlib The package does not have Seam nor distributes JAR files in wonder if you meant the ZIP produced by +Werner I used for Windows to create this patch from schema of root directory of castor project +I am sorry that I gave my ok now +Thank you for the patch EugeneMichi +May i ask for more info for this issue? Id like to work on message +That makes sense +Committed thanks for the great work Eli +now added to CP Branch as well +Committed +Originally I posted that Tomahawk was causing the compatibility problems +Supporting are attached in file to run the BR Management Client Removed the run method as Arnaud suggested + for the patch +the maven integration will probably not work correctly for you and maven integration will ignore it +Shall we do it in another subtask +thought about toggleControl and not a componentControl for some reason +Fixed on the trunk thanks Eric for the has been fixed with the patch on the +Agree with you Gianmarco +Okay I have the issue locally and have debugged +Patch committedTrunk GeotoolsCompleted At revision Completed At revision BranchCompleted At revision Completed At revision closing all issues that have been in resolved state for months or more without any feedback or update +Trivial test included +We should send something like Invalid media type in the soap fault +Committers please review patchThanks +the podlings PPMC and community. +This looks reasonable for +can you point me to a public wfs that returns when counting is too difficult +libs +They seem related to the patch so Im removing the patch available flag for now +Anyway one comment I have is that I removed the jdom deps because I thought somewhere down the line a transitive dependency was mucking things up +Compile the list from this but leave the output directories as they +Implementation is generic that the counter need not count number of files +Hey MartinThanks for the fix and sorry that I didnt get back to you with more look forward to the new next release of m still experiencing this problem in version +For example this is possible +note that anything under +Any luck reproducing this or should we close it +I have attached the source file with the changes in it +And the general solution would be simple too Drop the entire circular reference resolution mechanism always throw an exception in case of a circular reference instead +I let a few of them process through the Listener and then I shutdown Tomcat +Thanks for correction Ramya Nicholas please consider this patch for Hadoop +Ill update the matrix considering +Is this still an issue with +The reason my xalan build was failing after my xercesbuilt successfully was the build being partly stale since part of the buildused the old xerces +In my opinion it is allowed to include patterns without constraint before other expert doc for version statesText in the first cell below CONDITION develops into a pattern for the rule condition with the snippet in the next line becoming a constraint +Thats my +That is what is happening here +Elliotte has indicated that that issue is now fixed to his satisfaction and that issue is closedGordon please verify that this issue is resolved so we can close this issue. +I didnt do the part tests are becoming out of sync with the codebase +The issue was that the sample was out of the build and thus not generating the dependency xml +This only affects the current trunk this class does not exist in branch +How are your templates with PV tools created? Cant you set the memory properties when you create that template +Fixed with +Just replace with the content of the script it calls +Im still cleaning up some of the item providers and samples but should be able to commit tomorrow. + +Lets all go w Lists over arrays from here on out. +There are been at least two bugs with unit tests failing due to not using +Here is the sample application as a maven enabled source tree +OK sounds like this patch should be applied for +So I cant stick a or extra line feed or anything like that +elements handled. +it is certainly not used for the creation of an invoice from a project where the projectrole assignment should be usedJacques concerning your questionThe default setting will preset the organizationParty in the drop down but it can still be adjusted by the user +Lately every job fails because of +Your user story above needs to be written against the eclipse plugin +When the server gets into the described state we dont see activity on the +Thanks for the report I will integrate your tests and get back to you later this evening +Rick I am sorry just noticed I confused your names in a previous comment. +Initial and final patch basically moving responsibility for instantiation of to and refactoring the code to make this naming scheme available everywhere required by passing it around +Sure we should create a test after we understand the cause of the NPE +kernel tests started to ve updated the patch so it passes all kernel tests on Windows now +BTW this may well be an appengine issue but Ive created the issue here since it may also be caused by Roos and the libs it ships toThanks for reporting this +I could have guessed +Attaching the patch to fix the issue +Please deals with test +fix checked in +I have a portable extension that adds a Synthetic bean id annotation to its so it can identify them in the event +the above which I noted on review board +Closing old bugs +cheers +Hi WernerWe were using to create the java bean object here is from our scriptSET GENERATORREM Generate Schemas java GENERATOR verbose i XSDHOME package noticed that you were using class to create the data bean object +Thanks for pointing this out! Well fix this for both M and ve lowered the priority to Major though since it is not a blocker for the overall framework +That is how they are exposed to any production user will have their own tailored to their needs +no feedback on this +ok if central does already what is the problem? isnt this jira then solved +Lowering the priority as this is an undesirable behavior but not reall harmful afaik. +Plugin might better be called since while not it is specific to services that are started and stopped +It is easy to workaround but you need to add these properties to two additional places to make Maven use them add them to vmargs section in or to Eclipse command line to make Maven launch configuration use them use Window Preferences Java Installed Edit Default VM argumentsTo make both Maven cli and Eclipse use the preferences they would need to come from the Mavens but that would require changes in the Maven itself. +I am using camel +The largest supported chunk can be GB so that it will work with all browsers +treepages of a bulk update all resolutions changed to done please review history to original resolution type +Hi bc can you talk a bit more about the use case in your mind +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +This should be fixed nevertheless +That helpedallow works fine for me +Closing issue as it has been released as part of release +Tried a test project and it doesnt flail anymore. +We dont have multiple GS instances on same tomcat +I wanted to update in row I have double value in Increment api provides support for long and int to update column value So I am unable to update column with double values. +This was bad anyway because this used implementation knowledge of the resource resolver which IMHO is not allowed anyway +Some questionsAre you saying that VFS should add a to certain requests? Which ones? Are you saying that this is a code bug or a documentation bug? Should VFS offer an option to follow redirectsCan you post your whole configuration with instructions such that it can be reproduced by anyone +applied patch from venkat +We dont even have to differentiate between String and Block anymore because a String will be coerced to a Block +OGNL was implemented to sustain the inverse property of the +Thank you Gunturi Does this include the Hortonworks releaseCharlesSent using +Fixed by r +This looks like something naming all issues to the CR +Please fix it in jcr fixed for +There are also thousands of people that spend a day or more trying to make that leapAlso it still doesnt answer the question as to why the child entity doesnt get updatedOh and by the way +And then somehow we walk that tree enumeratingscoring individual valid fragdocs that are created from that treeSomething like that +So I dont think it should be a singleton at allbq +Can this be resolved Rick? + on patch +The patching was smooth and overall it works good +The sanity pass rate is now whereas previously it was so we have regressed slightly +Im ok changing scope if others agree +Claus looks like upgrade to xstream solved the problem +The holidays got in the way +I committed the patch +not going to happen in the series. +If I have NO file it works fine even on a Mac +all tests pass +Does seem to be something odd going on here I updated your test to actually timeout and fail if things dont go as expect as well as run a persistent variant for comparison +Im all for having it back +The idea of separated from tree has its excuse +is coming along pretty and should be close to getting published soon + has been closed +How about to share related codes by making extend +Im not sure how CL interacts w protocol level +I know we wont have too many explain formats +Only needs to be signed but still needs to be part of the version of the attached +we dont have an easy scale down solution +Ive finished this implementation and am cleaning it up for contributionIn the meantime Im attaching the results of some performance tests +For this issue I created a new branch called +ClausI think we should keep this issue open until we upgrade the CXF version to +I need to wait for results from Jenkins though. +Will attach initial version tomorrow +Patch attached +Rob Frank thanks for running all these benchmarksThe fastest implementation varies from platform to platform +Theres been a lot of work in the session handling recently with new session attrbitue and so on this is old and could be outdatedI would like to close this +The error in console is Node cannot be inserted at the specified point in the hierarchyIn Chrome the error is HIERARCHYREQUESTERR DOM Exception . +Makes sense +Thanks again +Well post it to the news section when people can actually get it +Only code change if possible and not convoluted is always better IMO +NOTE If you are using Sonar Light then file will be created and existing overwritten. +We want to keep the release backward compatible to reduce impact on the patch since theres concern about backward compatibility +I have no idea is this is correct you know from product can you verify and close this issue +Not a simple DRL will compile +Especially for YARN where we are already changing a lot +You can club them into small groups and file separate jiras for them as you fix them +I still get an access denied error when deploying a snapshot from my account or from vmbuild +Sorry it took so long. +I am not surprised that fixed the problem as it seems the issue is in just the last few pages of the file that are marked free but not actually on disk +Passing user data via one file and ssh key data via a disk layout in a separate vm payload would require updates to to know to look for the optional nd fileIf the were not to be used simultaneously with the key data injectionvia a config disk layout it would be possible to inject the key data via the current vm payload user datapath +Yes that is a nice idea and more consistent +Anyway a more comprehensive investigation of dojo custom builds would probably be better. +bq +Can you please do your work before asking me to do mine? +here is a patch show my code on that +here is the updated +Is the intention that the system tests are run outside of SBT or would it be OK to execute them as part of the test phase on the master project? If the latter is OK I think this is done fairly easily +Thanks for your efforts! JuergenConfirmed with build asMichael +Duplicate of +The sooner my test runs the sooner I can play around with the possibilities write the howto and attach the any menutabscatalogue lang files I create aside from what is currently in the distro +I believe I have completed all steps required +Is there any reason I shouldnt commit it for AS as well +Seems to be a Hibernate mapping then +KF fixes the IO issues this issues since Felix doesnt support fragments and thus this functionality doesnt apply to it. +Assign to during tidy up prior to +The src for and provide the base that the product docs are forked from but they are not shipped as built by the maven build +Let me know if you see some room for improvement +Applied thank you Oliver +We should not back port this to not need to back port to for like this was reopened for backport by mistake and left open as a result. +The change is in and I raised to get this pushed into groovy core +If you dont this duplicates that issue we can leave this one open +What are the chances that this patch will go in Ofbiz +Updated with the following changes Fixed to return the correct value it was returning false in all cases +That seems cleaner as it is not something that should be used directly +I disagree with that +After a successful authentication a signed cookie is issued and used for subsequent requests +Sample source referenced in issue +Can you please provide some background info and possibly some sample data test this +Cant find the Runtime zip so I suspect thats been removed +That is a known issue that I am planning to look at this week +All JREs have been updated to +IT added +And the timeout used between the application and the client must be longer yet +There is no intent to overide it +I assumed that the code that marks a RBW replica as corrupt when received for a COMPLETED block is probably there for a good reason +The new patch still depends on jdk but Ill see if I can fix that reminds me that I was meant to follow up on this one +checked the capabilities and describefeaturetype outputs and they dont seem to reference any sort of link +Also has been copied from hadoop and modified to make it work for avro data also +fixes the bug and renames mutable methods to get +See any difference? Whats the math like? How large is the Map if M regions in it? We remove when region is bad but ones we dont access could stick around for ever +Performed modifications as specified in the previous comment +Its a large +Verified in +Damn! I forgot to give this ticket a much lower prio than major +This can be solved in the distributed mode by actually having the mode in the task +Got it +It looks like OBLSYNC can be used only when user knows exactly how messages will be exchanged between N peers +there seems to be no regression there +Im sorry I just dont follow +Initial patch for review +bq +Keep this one that way the history of the discussion is all together +Can we have confirmation that fixes are in place for +The introduction of the setURI method wouldrequire modification of the interface +will that work with both with Portlet Bridge and +Great work Sijie +This is truly a bug in eclipse and caused by the Show Referenced Libraries Node featue in package explorer +bulk defer of open issues to +I wonder why you want to align the version number string with that of specific versioning scheme +Why does extendsIt seems to me that does the that has correct +in +The new allocation should use INITIALRESPBUFSIZE instead of hard coding right? May require modifying the test +Thank you Leonardo for theAs recommended I tried setting the SUPPORTJSPANDFACESEL to false in addition to the existing EXPRESSIONFACTORY being set to and that didnt resolveThis time the application started without errors but when navigating to a JSF page the following exception was received Target Unreachable identifier xyzBean resolved to it normal for the files to not go through the JSF lifecycle? I would think that they shouldIf files are supposed to go through the JSF lifecycle any thoughts on why they arent in this scenario +Thanks for the pull request +Still there is a reason why they are in examples directory +If not we might consider checking for this during DN startupRegardless that can be done as a separate JIRA +Ping +What about this issue in the coming new releaseIt should be nice to confirm it has been fixedThanks +obsolete +Platform supported included with the JDK provides remote connectivity with a very small amount of configuration +I might have already fixed this +For instance consider a N node joining a cluster of nodes +Dont think tests are required for this one +It should be put somewhere else in front of the chain +should be fixed now please and its now working nicely. +Sorry for confusion +Or perhaps an explanation of whats different about this code from a typical Rails setup +inode numbers +We dont have to support that already proved source attachments couldnt use variables so be it +Will get back if i find any issues +Now is the difference between phonegap and cordova more and more confusing +My definition of a balanced cluster was in term of used space +Thanks again +They are lazily initialized whereas all other class nodes are generated directly from Class objects +i think this is it +The demo runs successfully and the console does show messages that convey that the logging and authentication policies are working +Destination promary storage is a primary storagemysql select from imagestore id name imageprovidername protocol url datacenterid scope role uuid parent created removed totalsize nfsexporthomesangeethasecondary NFS nfs nfs exporthomesangeethasecondary ZONE Image NULL NULL nfsexporthomesangeethasecondary NFS nfs nfs exporthomesangeethasecondary ZONE Image NULL NULL NULL +I could run all test without exceptions +Im waiting to see when varrun no longer exists before running start +Thanks Ravi +I reproduced it with JBDS on Ubuntu x +Can we close this issue +I have written different tests but I do not sure I have tested all possible scenarios +Closed Issue +The patch is to be used due to rename both categories for JSF EL Proposals and JSF Tag Proposals +This is actually the bug that we identified late +Xuting thanks for the patch +Youre right +I think it is a very good idea to unify log configuration +Please use this zip found a small problem +On trunk I will preserve the behaviour defined above +Can you send me link to the specific doc you are looking at +Ive added and committed tweaked version for as wellLets create a new issue for the new disabling test case with the offlinedisable regions and discuss there +Josh G Can you update this to work with trunk? Id like to be able to include it in +it needed to be specified as a property in the +Looking at the code very little is actually shared between these two use cases so the need for a base implementation and concerns about the appropriate location are not so much of an issue +After i had allocated Mb for my tests all tests passed successfully +Thx +Fixed in the test was fixed +An exporter with a lower bundle ID is preferred over a bundle with a higher ID + +AaronI changed the code according to your guidelines and merged it to the trunkAs far as the unit tests for supporting new Oracle data types I modified the file to include data types TIMESTAMP WITH TIME ZONE and TIMESTAMP WITH LOCAL TIME ZONE +I think we should keep that +Yea sounds like a bug in the method the balancer uses to determine the namenodes +Logged In YES useridOK the only I see is that setting the values inthe preparedStatement for the finder is done with typeVARCHAR instead of CHAR which ist the jdbcType for the my original post for problem with an oracle seem to found documents on the oracle site and discussion threadsfor agree with the posters there that oracle wont be willingto change the behaviour +It would be convenient to see what state the app was in when we expected it to reach some other state +Hi Mitesh please try to write a proposal like above my outline +This property is enforced at adaptor creation time if an adaptor name lacks that prefix it is addedThe main motive for this change was to make the logs easier to read and parse +This should not include any discussion of the private classifications only our compatibility promises for public stuff and that anything undocumented has no compatibility promisesDoes this make sense + +link added refres page +Heres a patch that solves the issue I was seeing on a clusterMost of the issue was upping the move cost so that all cost functions below it couldnt overcome the cost of a move +A quick patch to solve the problem + svn ciAdding assembliessrcmaindemosdeployerfeatureAdding assembliessrcmaindemosdeployerfeatureAdding demosdeployerfeatureAdding demosdeployerfeatureNOTICEAdding demosdeployerfeatureSending demosdeployerTransmitting file dataCommitted revision +The User Guide provides information on how to include the CXF based configuration with the BPEL deployment and there are some quickstarts available to demonstrate username and digital signature authentication. +Do you have any specific suggestions where we should announce this in addition to the three places named above? Any suggestions are welcome. +Or you could upload the pcap file +mvn clean install Dmodule the resolved issueLahiru +on the latest patch +Space created Greg has all perms +jBPM Tools Reference Guide is checked +When we find a third case necessitating a change to the we will admit we need to refactor +This one is for Klaus I we upgraded the dlls to and compiled x and x version of dll. +Basic t this be done with some kind of script context global or something rather than adding a logger to the method signatureAnd if it is added to the signature will this work with methods without it too +Updated the environment +This is a new consistency check introduced in Spring s saveupdatedelete methods on a Session in is potentially dangerous It means that you are either doing this in a transaction which will never flush the Hibernate Session +Actually you should use to clear the field in the +Woops youre right sorry I was looking at an old version +Here is the other issue that relates to and +fix made it to R Early Access +Otherwise Select type will not show up +Thanks Sid +This is now split up into separate more manageable JIRA issues so that we can discuss each piece in to change the component to Module API +Yes Olga I agree with will be resolved as part of turing complete work +Can this also colocate regions of different tables maybe based on a key prefixIn our case wed tenants share tables and each row key would be a prefixed by a tenant id +Fix is now in CVS +The problem mayan manifestan in Weld +I would like to hold off on closing this +I attached this two files for more researching +thank you Tim +The cumulative patch gives boost on Dacapojython boost on Dacapopmd boost on Dacapochart Dacapofop and DacapoluindexUpdated patch with changes at and +rename next release to +Its getting deleting successfuly umounting in kvm host and is not showing the poolWill verify adding another KVM host +Committed revision I also removed the custom file copy method with method +Could the metadata API be modified with an flag +think on the same topicjob type for different tenants with different configurations +I meant remove warning directory does not exists +The values come from weights and can be changed to +Bruce any comments for the patch +The patch should also apply successfully to +This is a good idea +v attached wo extra Builder fields +snapshot I think +Its written for +Ive made versions one where the feature is enabled by default one disabled by default +SLFJ will be replaced in just a few minutes with SLFJ +Rather than make something custom for heap doesnt it make more sense to just process the command line parameters and strip duplicatesFor example right now Im fighting with the NN logging because I want to set differently per service +I think it was hbase testing that was encountering this issue +Need to run through this with Hadoop capture session output for an example and incorporate it + fails intermittently +Should we go ahead with this patch +Will this address your concern? If not can you provide a concrete example +So that cant generate the issue +Perhaps a full heart beat from server to client then client to server would be a better mechanism by default +Any other component that isnt included in screenshots is invisible +Is it done then? Isabel you had suggested moving this to so I suppose youre saying its not done but wonder what the delta is thenGrant I tend to agree with quick review and commits since patches very quickly go stale +I dont have to select first so not sure what the problem isnote the selection is there in case we detect it wrong and allow user to set it. +Could you try out showcase and see if it works there? Theres an example under tags ui tags lots of option transfer select Could you also post some snippet of how the jsp looks like +added code can you attach an unit test that demonstrates thisIs helpful to implement this smarter in the core +pending minor update to take Todds earlier suggesetion of having the message reference the config option +Proposed patch also removing the class which is not used any longer because we dont scan the class loaders for the service files +Scheduled for +I think it was +While I agree this behavior is strange I think it is expected behavior when you consider Velocitys whitespace rules +The RooJavabean ITD is only created if you have fields in the +A separate SQL is issued for each entity referred by the +Dependent upon the commit of will request review whenif that work is committed + you +Take for example the spring librariesbundles etc +I assume the exception thrown because the node selected in the tree was removed from the model +A small note you can install the demo with Java or the previous comment isnt completely trueJetspeed itself runs fine under Java but the wicket examples do need Java m going to provide a new demo installer for the latest at branchThis will be based on the release. +This issue should be fixed in subversion repository as of r +The runtime now removes the intent and the test passes +I have provided this so at least the Groovy team can easily decide rather than wading through Grails examplesI went so far as to do so as respected members of my team have hit this problem and would have preferred not to +An implementation can easily have a single method for creating a cookie and call it from both methods +Fixed at rev . +Hii have exactly the same issue +When I work with the most recent nightly build dec I still encounter the problem +I think when we fix this issue users who created their views with or higher will have to drop and recreate the view to see the fix +Ok thank youBut now I dont see the dependency inside the of the example application +It looks like the sample size for Timer objects isIts only kB per Timer though which is pretty much a rounding error when compared with Solrs memory usage +Jacques let me look back into this and see if it works for me +I just committed this +I understand your concern Onno I just believe the more common case is to have it the other way around wanting the page to render as quickly as possible on the first visitYou say the overhead is minimal since the script is cached but that also applies to your own use caseYour problem occurs when the page is fully rendered and the client waits for those scripts at the bottom to load before you get to manipulate the DOM +Committed revisions and . +debugging would be off by default though +OK on A +Can you please elaborate what you mean by When you upgrade to drawd +I made for doc +Removed default in send field revision that some additional changes were made in r to improve upgradeability on both databases +you dont need to exclude them to chenge the scope just add it with the right the code patch but I added a simpler unit test. +made a few changes regarding logging and formatting +Applying the patch to my couchdb install resolves this issue. +We use a lot of HQL and cache most of the queries +Happens when are renamed across file systems +Creating instance each time a parser is needed may be too expensive +Should Geronimo immediately upgrade +I summarize the main infos it seems that is configured with file descriptors we have some leaks the patch itself could be committed imho +Thanks Stephen Ill make sure the commitWithin is accounted for at the request level +It might be bc we timeout on the wait +Should you modify as well +Step Sending trunkapplicationssrcjavaorgapachegeronimoconsolejmsmanagerhandlersSending trunkmodulessrcjavaorgapachegeronimoclientbuilderSending trunkmodulessrcjavaorgapachegeronimoconnectordeploymentSending trunkmodulessrctestorgapachegeronimoconnectordeploymentSending trunkmodulesdeploymentsrcjavaorgapachegeronimodeploymentSending trunkmodulessrcjavaorgapachegeronimojeeSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojeedeploymentSending trunkmodulessrctestorgapachegeronimojeedeploymentSending trunkmodulessrctestorgapachegeronimojeedeploymentSending trunkmodulessrcjavaorgapachegeronimojettydeploymentSending trunkmodulessrctestorgapachegeronimojettydeploymentSending trunkmodulessrctestorgapachegeronimojettydeploymentSending trunkmoduleskernelsrcjavaorgapachegeronimokernelconfigSending trunkmoduleskernelsrcjavaorgapachegeronimokernelconfigSending trunkmoduleskernelsrcjavaorgapachegeronimokernelconfigSending trunkmodulessrcjavaorgapachegeronimodeploymentserviceSending trunkmodulessrctestorgapachegeronimodeploymentserviceSending trunkmodulessrcjavaorgapachegeronimoservicemixdeploymentSending trunkmodulessrcjavaorgapachegeronimospringdeploymentSending trunkmodulessrcjavaorgapachegeronimospringdeploymentSending trunkmodulessystemsrcjavaorgapachegeronimosystemconfigurationSending trunkmodulessrcjavaorgapachegeronimotomcatdeploymentSending trunkmodulessrctestorgapachegeronimotomcatdeploymentTransmitting file dataCommitted revision Checking in modulessrcjavaorgopenejbdeploymentnew revision previous revision Checking in modulessrctestorgopenejbdeploymentnew revision previous revision Adding kernelsrcjavaorgapachegeronimokernelconfigAdding kernelsrctestorgapachegeronimokernelconfigTransmitting file dataCommitted revision +But reviewing and committing the two changes at the same time makes sense to me +You have in your config file not +Thanks Owen! +I think we can close this too +i can provide patch even if im not committer +Stick it in the EPR +log message at warn level in r +Re Implement service createNow service Check For OutgoingIncoming Payment And then Create Payment Group MemberAt Ui payment is non editableThanks to Rishi +Please verify in ER are tracked on will check final version for EWP after is closed +orgapachehadoophbaseexceptions +Thats the way the API was initially set up +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +This stuff is getting resolved more and more for example in JDK new settings +Also the SCHEMAAID column which we are changing is not part of any index in SYSSCHEMAS +Closing. +Fixed patch need to check the contents of the +Added a runtime dep on logj to the checkstyle plugin POM to avoid the problem and made the more resilient to null error parameters +Duplicates which is fixed in trunk and patch is ready for +Thanks Dhruba! +There are hundreds of us here but Ill have a few try the nightly build + duplicates this issue but all the comments are in +Postponed +The issue is already on the Store interface level where get return a +Patches updated to match tab policy used in the original Alexey Cpp and both set TCPNODELAY +Playing with this when a region gets a lot of writes we could lower the compaction time but it would happen more often +Jamie and ChristainHere is the work that I completed that uses a different mechanism to associate a master instance given or more karaf instances +I put this code in common since thats the easiest integration point for additional native codeOkay but Id then vote that we not commit this until either theres another dependency on it thats committed or were about to commitLonger term perhaps well add native code to the HDFS project and could become a patch +Our nightly QA also found a related issue +Tom any feedback on this +Resolved +Fixed all of the problems in this section except for the issues with the graphics +It can only learn that U is owned by X via gossip and T will be down until it learns about Y +I ran into this today for a property of type long that maps via a to BINARY in the database +I think this is because some client classes use so that that class is included both in and in +The problem was the error message +Not sure I am the right person to take this one +Linked to JIRA for transactional support. +You set out with a goal and I guess this was to make performance gains +Note this is only true in the wars application +Is the test failure related +Closing as wont fix since the root cause is not in +Having said that I think I have found a potential solution +This seems to have been fixed within Maven. +Committed revision +Thanks for the reviewAlso Im gonna open new JIRA for the incorrect NNs behavior and move this one into the append branch +As mentioned above swiping on a does not move the scroller up which to me is not a big issue +Assign back to me if you have any questions +From Sonar point of view I think it is better to raise a warningerror when convention is not followed +RobbieLet me check if moving the property into the pom helpsThanks this issue the recommended path is for maven users to specify the required property in the maven build. +Not the end of the world and workaround +As this is not a blocker for release Ill move the issue to there anything else needed before the patch can be applied and the issue closed +My frustrationis that constraint ought to be made clear on the page describingcontributions +ok new version of the patch +so I dont have a better name for it right now +Thanks +ivy to the working user and it is still working +fix findbug warnings +unscheduled this. +Patch attached +Improved the component Example with an so that the last created widget instance can be displayed +Documentation in schema updated. +Attaching the patch with above mentioned Vinay Im not sure I agree with the premise of this JIRA +Thanks for your work on this Bikas +Go for it. +Committed at subversion revision +Just made a full build on windows on the problem is no more there +When asked about adding the found runtimes press Cancel +Patch looks good and changes will surely improve datanode startup time +Ok got it +The following files are missing headers package package package package package package package package package package package package package package package package package package package package package package package package package package package +Theoretically only value Helix cares is the id which is mandatoryFor step lets try to continue the discussion in +Ashutosh I agree that the information can already be viewed via a describe formatted command however for tables with a large number of properties or even a large number of columns or large column comments it becomes difficult for users to quickly find the property theyre looking for +I mean Fields instead of there a chance you come up with some example code snippets on what you would like to see? I can hardly follow your explanation +Patch with unit tests that expresses the NPE on deserialization and during the roundtrip for serializationAlso shows that the object inspector is behaving correctly +This was a valid issue but I saw it and fixed it quite a while ago +No needs to say endless scrolling should always be an user pref option +Ported the XML sequencer from into the branch. +When I change doc newDoc everything works as expected. +Migration process completed +Closing as invalid because this is pretty vague +INFO nntpserver NNTP Service is running on INFO nntpserver NNTP Service handler hello name is INFO nntpserver NNTP Service handler connection timeout is INFO nntpserver NNTP Service connection backlog is INFO nntpserver NNTP Service will allow a maximum of connections +I deployed the webapp in trunktestsuite onThis error didnt occurI agree that the problem is because the colon in the windows path +There could be a number of reasons +this works fine for me with the trunk +Daniel any updates regarding this? Do you think there is a better way we could fix this issue? +Hit httplocalhostportal +Thanks for the patchFixed in eaebcfbdcccddea IT added in afddfcaeedfbaSorry about the delay +This patch is the same except that needs to call retrieve rather than get +is fixed in code but a unit test still needs to be written +This issues history with different reported problems is simply too vague. +version of previous patch +Perhaps if null is passed in we should just do nothing? If they try and retrieve a value with that key theyll still get null +Okay will remove it from the branch as well. +Thanks Michi +ccwWithout a heap dump a stack trace or even process logs for the hung process I am not really sure what more we can do but tell you that it looks like it was an intermittent issue because it recovered after rerunning the task +but Im not going to argue for a better patch in here when this one is already fat and its gotten a from hadoopqa +Committed to too +Instead of calling it CONSTANTSSET how about we make it POSSIBLEVALUES or something like that? I dont want to confuse CONSTANTSSET with the Ruby constants method +I just realized that I hadnt checked in the changes yet. +I just committed this +These regression tests run as before on both Java SE and on CP and as far as I can se I havent undone any of your code from your patches in or even though the code has changed location +We should revert this change +Thats weird +Changed the behavior to first check for related headers +Although it is not marked IIRC this issue affects both and +closing this as rejected and creating a new and closed as issue was accidentally created for EAP but problem is with reopened to change Release Notes status field. +Is it possible your changes altered the calculation of +Kannan I guess you said already that that would be tricky for you +I have started looking at this for current idea is to make use of the Groovy MOP and do the following compile the contents of the display view into a script pass in all the local variables into the scripts binding for all variables and instances referenced change their metaclass to one that overrides getsetProperty and invokeMethod +Looking at this again I can see some very elegant solutions +Maybe yonik could try this out with Solr? +I verified this on Windows and Linux using. +Done +Still I just wanted to warn everyone about possible mysterious resultsSergeyDid you use the Givens or Householder method +Sample style sheet to trigger the XML file needed by the style checked xslvariable or xslparam made no difference +Committed a better solution using for the locking issues +Theme and localization support is now available by as of revision +Sorry about the delay I didnt have time until now +I vote to close this issue as resolvedWe fixed some stuff for ZK that should prevent this from happening +To be applied in cocoon trunk +Closing +I have no real idea how this can implemented sanely. +This required a forced upgrade +Of course its the difference between allocating one object and many +to leverage Correct +andThanks +Reopening so that I can attach this issue to +The patch removes from +lives in the tigersrc directory +er take two uploaded wrong thing +Specs to exercise subclassing classes with public and constructors +I had an easy time reproducing the bug +Whats interesting is that none of the local log files are logging the ERRORUNKNOWN at all that is only showing up on the collation server +pushed +Will investigate +I cant reproduce this +Thanks Sid +hierarchical network topology +Try a mvn clean install Pfastinstall in your CXF workspace first +We should change the title to mean that it is a bug beyond +It looks like regression is ok and patch is attached +Implementing first would give us more options on resolving this issue +This does not seem to work for me +I believe these are all resolved now +Add the test package to a MINA project and run the class +It occrur error message like that whenever I execute a join on OCI +Bulk closing stale resolved issues +I saw EAP earlier in the Alpha builds and now I verified it again with JBDS B +similar issue solved in the same way +Since James changed his mind Ill also attach a standalone patch as well +Currently we check the info for deletes however with thispatch I think we need to check the segmentReader which couldhave deletes that dont show up in the info. +I was trying to fix this bug and created a patch file +Granttwo comments How often are you planning two publish a snapshot? In every nightly build? Or in every nightly build unless no commit happened? Seems too often I took a look a the other projects that publish snapshots in that repository and they dont seem to publish more often than once per month +pushed +as discussed infra will be used +But this is just an implementation detail +The problem as mentioned in is that t injected cache resolver is not being used as alternative implementation for the cache resolver +Committed to SVN rev +Jar file with a simple test case demonstrating the bug +Im adding an include method to and an that will take this switch into account when the meta data is retrieved +There are some places where performance degredation could occur which need to be addressed and some other issues in comments in the code +Since theres no code in the containing class I see no point to having it +Add a second constructor to which accepts both the name of the offending field and the message +What about if you make usage optionalplugable +Its tfw for tif and jpegw for jpg +I hit this as soon as I compiled my app using JDK even running on a JVM +as previous discussion we preferred similar filesystem semantic when a concurrent deletion happened the writer could still write data to it but those entries will be garbage collected eventually +This is the last sign of live and it can be used by humans to decide to stop or notKeep failed jobs depends in the use case +decided to take it externally for Ralph to optional dependency seems OK to me +This patch adds GRANTREVOKE USAGE support for aggregates +This is the current lib built from the release tarballThanks Stack I think Im good +Rather I suspect changes in how methods are mapped or how Java Integration wires up classes as possible areas to investigate +Thats a better way to fix such things rather than just increasing the sleep time as we have before +A Junit test has been added on the get method +I started to implement this and it is currently partially implemented along with the fix for +r on trunk +Ive attached some basic examples An example script that runs through all source files and outputs as properties each types various members annotations docs etc +there is maybe something that is doing that the others dont +A dependent libraryYou have fixed the first issue but now you need to deal with the second +Refactored unit tests that fails when trace is set to ignore + styleClass of calendar is please update Metamer to check styleClass +attach my patch +here on my german the acrobat reader doesnt show anything +Hey Gary +Thanks. +patch is included in +We recognize this could cause a priority inversion problem +If you just want to generate a JAR without running any tests try executingmvn compile jarjar AngelaA question like the above posted onto the mailing list would possibly get you the views from a larger segment of the community +Once I switched to Eclipse RC these problems went away +And the block multiplication run correctly +Committed patch with revision +No user response seems like issues is fixedThanksDeepal +Do I have to create templates for existing Synapse components? If yes for what sort of components would that be? Is it the case where there are different sets of mediators which I need to make adhere to a certain pattern where it will be easier to call functions by setting the parameters in a easy mannerPlease elaborate the scope of the idea a little bit more + +Without seeing the code mods youve made its hard to say what the issue could be +I would like to see some input from the folks who did the security implementation first +setup transaction context before then roll it back on finish +What about formatShould we place it in dfsadmin group too +Attaching version which unfortunately cannot have the unit test because the coprocessor hook for injecting failure is not there +Gonna revoke this as it will get detected at startup and noone liked to implement a more user friendly solution +Need runner API instead of at the same time as! +commit initial version at revision +Sure it is not critical +We will be putting out first thing Monday unless anything else comes up +OkSo why not mod to check zk rather than do the filesystem check? Filesystem check is more comprehensive +ok will do and commit +Are you referring to the PDF version? Im still not able to see them there either although theyre fine in the html +martin if you have time search the codebase for i remember writing code that fixed relative paths for error urls there +also can not Sharing removal with default grouping using standard configdelete command +I dont know what should be done. +pushed +Thus the Hudsons comment could be disregarded +So what if we said anything ending in eats any newlines after it unless a special symbol is used +Knut saidI agree that it may be good for compatibility to allow shutdown to continue even without this permission +I also fixed a couple of errors missing param and typo in param name +BTW There is also a ticket about a component +Ok Ill do it with a single flag +Doh! Just realized the changelog website hasnt been published since Jul so the documentation is probably just not visible yet as you mentioned adding documentation +a lot for the quick review Eli +Please review it +Yet to come verify new code works when bloom filters are enabled +Hmm I think Busch might be on VacationId hate to release with a known bugI already owe Uwe beer so not sure what else I can offer +Ill close the issue as Wont fix as a comment has already been added to the sourcebase. +Havent pinpointed which yetSnjezana you said you wanted to take a look be my guest since this one is really annyoing +karaf branchPatch care of Heath Kesler svn ci Sending manualsrcmainwebappTransmitting file dataCommitted revision +But I was terribly wrongDoes this make more sense +Marking as resolved please reopen if necessary or file new bugs. +This is not a duplicate of +Its there now thanks +A new JIRA will be created for the new design and implementation. +If you like you can add classpath to and as they are always my cents +what we can do today is to first do a JOIN and BITMAP AND operation on the index tables and then find all the matching blocks which is ok but there requires a join operation +I sent the project a pull request adding an explicit license file +In the secured environment a client should make a privileged RPC call to access a instance from an NN +See in the stack trace it says Remote unexpected exception so it was indeed cast as +not sure I understand what youre asking here +SC gen is now done with Velocity +After the patchmethod for Reader calls method forTest of this patch Execute derbyall and found no error other than and +Ive added code to capture and log runtime errors in run methods +The patch is committed +Pushing a fix for this to and +is fixed +It took a bit of thought but Ive come to understand the error of my ways. +Would you please check the attached image? I agree that a new request has to be sent to the server asking for a new secret but I was thinking that only the latest image should appear on the screen +Have you had any luck profiling this to see where the references are being held onto +Patch committed +By design to resolve maintaining context during page forwarding +I will add the comments to the tests +reopenning issue to add patch which will cleanup of usage in tmp +Removing keyword substition and releaseprepare worked as not be exactly the same issue but something changed that locks the causing writesupdates of it on commit to fail +Im closing it anyway as I dont have time to investigate please open another ticket for any remaining issues. +Do you reprocuce the problem with my sampleSorry for my enlish +Yes you did sorry for missing this first time. +The javadoc tool appears to have generated warning messagesThese javadoc warnings are not related to this patch +Project setup. +BTW Im reading Valiants BSP and MSs dryad papers and just wondered why these jobs should running in on a cluster +defer all issues to +Ill keep an eye on just in +Alternatives have even been suggested that would make it redundant +Reopening because the random issues are still a zip file with the surefire reports showing random test failures within the module collected from successive runs for hours +We can alternatively just disable line checking if wc L is not available +patchThe setter had a documentation and overflow problem +I added a new view class that allows for the format to be determined at runtime +Using a Method as a Bus extension pollutes the use of the Bus +Now we can upgrade to and replace the datagrid cachemanager subsystem with ASs infinispan subsystem + +Build successful! So now we have all necessary files for patch + +Did a check through the rest of the classes the property order of also seems incorrect changing it from instanceParms description overviewDoc description overviewDoc instanceParms +Can you investigate why the Byteman rule is firing when it shouldnt beThanks +That would handle this checks +I just asked user how to recover from bad records when using +I attached a quick Java test that illustrates the different ways RC can be used +Issue resolved +Thanks for raising this and thanks in advance for the patch There is no problem using an enumeration to represent the type of test +Ummmm +The latter wouldnt help anyway because the default could be null toocheers Roland +At a very low polling delay it should be possible to differIf a parent directory has some files or folders stored inside then you have n delete events and also n create events where only the topmost folder differs +You can leave your plugin as for as long as you like and as long as you keep updating the branchtag names to match the version of Forge that that particular branchtag supports youll be summarize +I decided to join steps two and threeHere is first patch which introduces class and incapsulates all access to class constant pool into interfaces +House cleaning old unresolved reports +Closing all resolved tickets from or older +Reproduced only on page because inside page layout div with overflowauto style + +As I thought might be the case tests went through fine on my test VM +Ill open a separate issue for removal and link in the relevant subtasks +push to +The official Solr binary distribution includes the API jars outside of the +We get a complexity that looks something like dakkde which mightbe ok when all those value are small but if da k de you get comparisons that you have to doIm proposing more like a merge approach where you merge compare da and de andproduce a survivor list this list is then compared to k +I agree with removing from getComparatorI agree also with moving weight to a long +Sorry not sure Im following +Closing due to lack of reply. +Looks like it got fixed at some point. +Patch for is provided in will try to write a for for spec for +Arun can this issue be targeted for or? It will be really bad for us if we would have to wait till to get a official stable release that includes this fix +I like that approach Martin +JieHave you run unit tests for your patchI think certain code may actually rely on the existing behavior to work +I would expect the same behavior as Cleven is expecting +Ive checked now that Jexl has its own package namespace so it is fine to have both versions on classpath + +writercache packageGood idea Ill move thembq +Thanks Anoop! +Is anyone available to review this? Id like for this to make the freezethanks +The default value would still apply even if the property isnt set +I got the latest version and check the performance it works betterThank you very much +Of course I can but Im not about to until someone who really understands the status of this patch weighs in +So I resolve this issue with setting Assignee to Tom for attribution. +This appears to work in uDig rgould can you confirm or deny? +The fact it does not have JAXB annotations is not important becauseit represents a query parameter not the message payload and hence the processing rules are differentCan you try please? You should see the imports section updated and the compilation succeeding + probably has the same behaviour and it would be worth trimming the URL when it is injected anyway and not just trim to check for comments or empty +This will ensure that when the NN restarts the lease does not revert back to the original client +However due to the behavior is not reliable +removes the deprecated classes there were two public fields declared of a deprecated type they were not used so they are pulled in this patchSpatial could really use some love its not in the best shape for its debut there is no good overview no to explaining anything and extremely sparse javadoc +Here is a command line for testing the verifierant HHHHHHHHH +Hopefully take addresses all of your comments including refactoring splitTable to use +checked in thanks +bq +I am not sure which bug this one duplicates but I am sure Ive seen other JIRA issues which did the same thing +Xavier can you look at this one +I hope youll consider pushing the patch through and fixing the FLE issue separately +Will reclaim if it is still open by the time I get to it. +And including a total msg counter might not be a bad thing either +guess we should deprecate the finalResult methods in and suggest using insteadAlso why not put your logic to sometimes choose a impl into by default +Thats odd +Now it shows that it waits for the approval of the project lead +One version uses to invoke the web services and employs the WS API classes and implementation +Used the snapshot +First of all understanding of BSP computing model is important +Will verify and upload a patch if it still exists +Back to occasional deadlocks +Hi PeterHow about the status of this issue? have reproduced a bug with first usecase delete an user with the user management portletBut there is interest usecase change a password for user delete an user success +I attach to make it more clear +Jackrabbit will use native as implemented in available in Lucene +Attached an updated patch in which the java coding style is corrected and javadoc is added +Committed branch and trunk +Dan +Actually on the patch both reads times will impact on the only read counter +No objections on my endJarcec +Im working on a fix for this +A patch for trunk was attached +not backported yet +One possible solution is to add a method to distinguish the two +ok I will attach a patch soonThanks Roger +Making distribution inherit from the pom means that a distribution would get built every time you did a build which is undesirable +The intent was to create a container that ran on a users desktop which the Jini community could provide demo software to run inside of +Attached patch should resolve the issue +There is nothing g compiler can do about it +Form for the object makes possible to create edit and remove properties included into the I think dialog Add Property may remain as is for adding some items not included into objects +The only interesting bug I found in the mailing list was which was fixed in Hivemind already used by +I am a little bit worry that more generic a approach might affect performance more than a approach +is this got fixed in hbase +I had this problem as well +That is the same cert that I exported +This issue is still open due to the couple minor subtasks that are still open +It causes the Cannot remove the same component twice during the phase when trying to copy the facets in this wayTested with Mojarra. +I removed it and uploaded correct version of the patch +Furthermore changing the name breaks backwards compatibility +These nodes should be connected with linesarrows to show processing of incoming message +Also Hassan are you planning to complete the port of the unit tests +sends NEWLEADER concurrently with the leader adding NEWLEADER to outstandingRequests so even fixing the barriers as you did does not prevent the race I mentioned above I thinkI must say that I didnt find my proposed patch very elegant but Im not entirely sure that yours covers all cases so let me know what you think +nope widgets are not at all associated with bookmarks +or which have models that are not supported by the version +NicolasThere is an existing function find found in that seems to do nearly what you have submitted here? It doesnt extract the first result and isolate it but that sounds like a very bespoke requirement and the service would better suit reuse if it just returned a complete listSome general commentsThe code in the patch does not follow the standard formatting guidelines recommended for Youve completely ignored error handling and are throwing general database exceptions back through the service call which is a problemYoure using the view so why not just return that rather than loop through and look up all the real Product entities? Chances are most callers would find the information they require on the view entity and if required they can get the real Product entity +Applied +bulk close of issues as part of making release notes. +There are different jiras one for ensuring the bucketing properties and one for ensuring the sorted propertiesCurrently even though the tables are sorted and bucketed during the table creation they are not enforcedIt is up to the user to make sure the data is bucketedsorted appropriately while loadingSince it is not enforced the optimizer cannot take advantage of that because it doesnt know whether the data is actually sortedThere was a jira previously which took advantage of the fact that the data is sorted for processing for group byThis is controlled by configurable parametersGoing forward we want to use them for joining specifically for sort merge joinsEdward currently we are not doing skipping based on sorting propertiesCurrently we create an additional job for bucketingsortingEven if there is a cluster by and the data is already bucketedsorted by the correct key we dont use that +Ok thanksRe formatting These must have creaped in while adding the synchronized stuff +It appears to generally be a simple ordering issueAgreed +removing fix version +A chukwa record contains both key and value hashes +So integrating the fix would be much cheaper than moving the due date every monthPlease reconsider this issues priority. +I think I should change ADVANCED to config and a couple other minor things so please dont commit +This is actually pretty funnyOur logic for turning a Method into a Proc uses something called a implementation +So during startup if the checkpoint file exists we know it is a checkpoint which was completely written +prior to generating release notes +It is related to some work for optimizations +Applied patch and changed the test package structure as suggested. +It can of course be that there is a bug on the side of Groovyc or osgi in the way that something unexpected is in the bytecode +any comments +I just tried the trunk and I still get the same problem when running srcexamples gdb +The really major issue is the message header +Sent CLA +Currently our intended approach is to add a method to to check the status of a connection +This doesnt happen in case of EAP +I am not even sure of a simple way to restrict it to just the RM +Thanks. +We packaged the jars and war related juddi into directory in release +Just want to say great bug report! I will they were all this detailed! MDB have not been working in Geronimo either +user can use Explorer or command line for now +Thanks Doug +Thank you +I added it so Im not going to remove it. +That patch also fixes +invokeHttp contenttype should also be an OUT parameter +Here is the patch +I guess maybe the former makes more sense since you have already started the UI stuff +More detailed log about this revision Committed revision trunk +Followup improvement in r +otherwise I would tend to think if ZK is tolerably secure for Hadoop then it should be tolerable for Hive +Thanks for the comments KameshThat would be a good idea +One minor comment on Phabricator otherwise looks good +Should be pretty close to MRI now. +Marking release note field as not required as this issue will be captured in the documenting of the main task. +Either A cannot be exploded because it is not defined as direct dependency or A cannot be included as transitive dependency because it has already been exploded as a direct dependency +Pull back if you feel otherwise +I have created a new issue for this problem with a better solution than I suggested before +The OS deals well with keeping hot mmapd data paged in so we want to let it do its job there +Iwein can you run these tests against SI tomorrow? I completed the upgrade to Spring so this should now be was indeed flawed behavior is correct. +I have not looked at javac warnings +Patch looks good to me apply it Id say but aint we hosed if we OOME in first place +ok i assume thats a issue +Issues seems to have been resolved +jar +Ivy now supports gzip by wrapping the in aCould you give it a try to see if it works for youThanks Ive add support for deflate as wellI did refactor the code a bit for this could you give it another try to see I gzip still works for youthanks revision works for meThanks +To clarify In the scheme the list has tasks sorted by the priority of their jobs +The fix for will probably fix this one as for fixes this bug +I looked at applying this patch but it failed with the following warning not sure if it is my environment or that the patch needs to be regeneratedmibmpwdCpmopensourcemibmpatch p i ctmppatching file javaengineorgapachederbyimplsqlcompile +Changing resolution state to Deferred +New patch includes documentation updates and changes required to work with updated class +Except I think we can now simplify to thisYay this make the nrt check simpler too +index and +Code freeze was on +i just say one of the largest contributor is block lru cache or memstore depends on users configuration +I can probably provide better guidance +I cant reproduce what you see +add bean decleration in master ebfafcbaddfacfadabe +How did you make collapsing work for distributed search? Which parameters did you use while doing a search? What I can tell is that the latest patches do not support field collapsing for distributed search +most up to date patch +There are pros and cons to all of this and forcing Thread per connection IMHO is not the right way to go +I thought of trying to resolve the other font inconsistencies but decided to keep it simple +patch checked in +You get what you are looking for +AFAIK members are not allowed to subscribe to all lists there are some exceptionsAlso members dont always subscribe using their ASF the script would still have to convert to the availid and then look up in the members list +So I wonder how to set the properties in junit Testcase just like inThanks for your advice +Going to check in the code today. +Patch applied. +all inside of an empty workspace problem is it theres something about my project thats preventing meclipse from resolving the workspace dependency but which doesnt stop qe hmm +need to fix or remove the duplicate Unregister actions removed +So this looks like a bug +access control is not something we were trying to do in Solr +It seems better to me too I did not look in code +Pushing back versioning bugs to has been fixed some time ago as well there is actually a test for this in the wfsv module +bulk close of resolved issues. +If we copy a file from ext with K blocks to ext with K blocks we are not trying to preserve any block sizes +trunk +That said Id rather be alerted to look into the fact that the job failed because no records were found +Im wondering if a more appropriate solution would be to making batch size fuzzy +I am thinking that we could add the generated to the same fileDitto +Sorry the last suggestion should have been to remove the from the mbean +Bulk closing all of these resolved issues. +The only issue may be the parent pom he is currently using but they may also be open to deploying that there +Sorry for the misunderstanding my errorI do not see all upcoming events in the dashboard +Please vote for the issue +Ive made a change that should address this if you could check out the updated and try it Id appreciate it +We can then simulate failure by unplugging the USB drive +very nice +Work in pagesize +Postponed till next will break backword compatibilityyou should file a issue for html tidy validator +On red hat this will include jna and attempt to mlock but fail due to the memlock ulimit being set to low for the cassandra user +Anyway we need more discuss to find the best reasonable way +Need to find out where to package the +No some housekeeping is needed +Current problems arein project list of projects having all permissions delete project link is not shown not having build permission the release link is still shownIn projectView No links are shown at all probably because the projectGroup object is not in the its not project but projectsummarycomponent +We are able to connect to AD browse using which is a Microsoft Utility we felt that the issue was not on AD sidePlease can you confirm if the plugin has been tested for with AD and tested OKRegardsSuneel +No the Select file upload dialog works well on every client configuration we tested +Already fixed in which is released few weeks is still issue about hangling character in names but Ill leave that one till migration to repo manger indexer. +release lines explicitly says thisWe need to fix it +bq +Its probably not a good idea to change the scoring for what to a user could appear to be arbitrary queries +Uninstall bundle locationA configuration to be assigned to locationBpid serviceAlso includes integration tests and fixes of existing test cases which do not expect this functionality all issues now this version has been released +In addition to direct requirements from the mailing list it may help to use analytics to measure browser types +Patch that allows specifying and +ZhiyongThat looks like an issue with the jboss implementation +Bruno see Richs the eval creates a subshell and ! reports the pid of the subshell instead of java have reverted the changes until figure out how to get that pid +You are as free as anyone to contribute a feature that matters to you +No comments. +properties files will always be treated as if they are generated during compilation +Hi JacquesIm thinking that the patch in is a very important task to be done but I think its not valid for Order and Ecommerce applications because its not enough to move label duplicated into and toI think that every single label that is used only in Ecommerce has to be put into and instead if its used on both application it has to be renamed to and change all the references to this oneFor example the label has to be renamed to because it was used on both applicationsThanksfill free to do as you like +Attached sample t reproduce locally which PLF version did you use +But no it would remove the namespace +Patch applied to LUNI module at repo revision rPaulex Please check that the patch was applied correctlyVladimir Please check that this resolves the problem +Sequence is different to identity column +Thanks! +you may trace the loader object cfe for details +This is working now for my use case using from the Apache People have changed over to it. +Timestamp related tests have been failing maybe related to +I dont understand why you would use for this at all but i am probably out of my league there +I got the same output +Right now it wait for s for the SPLITTING znode to be removed +Lets use this JIRA for the new test case and Ill open a new one for the Configuration change +in +need to apply to to branches in rev along with +Bulk close for release +I reverted v patch and applied this simple v it just ups the timeout +Kind of similar realy similar issue +Ted Will add that +I agree its unfortunate that bad symlinks but also that this is significantly better than the alternative of silently missing them since apps might follow the semantics of isDirAnother crazy thought Id like to throw out what if we just returned false for isDir if we cannot resolve the symlink rather than throw an exception? This has the benefit of treating the unresolved symlink as a file which means apps can still move them etc +Might have to wait a while AFTER GA to verify this has reduced the of errors reported in Marketplace. +Hey Konstantin NicholasThe report command gives you a lot more information than what getStats does +I think we should change the operation to remove instead of removeAll +Checked the screens of the VPE in Doc of them are. +The request is for a bit buildbotThe comments relating to vm or not and tethys are just rehashing previous conversations with infra on best approaches +Fixed in edfdbecccadfcbdbda with testcase updates +The upgrade wouldnt be as smooth as when selecting clean ranges randomly and would require metadata conversion +Seems to be a problem of SWT layout in Linux for this wizard +I think what we are discussing here is discussions +after they are committed this jira could be resolved as implemented. +Can you write a unit test that demonstrates the bug? Please attach it to this issue +Just committed this +Separate JIRA should make things easier though +Tests run complete successfully for both and withDaniel is this running or +Thanks Kim +The good think is that code becomes easier to read and understandWell of course you can choose a different approach +The new plugins I installed were Groovy Compiler Feature Feature JDT Patch Sources Feature Sources Feature JDT Core patch for plugin for the +Hope new bundle is valid. +may need to be enhanced to use in the retry case but will probably need to become a bit more sophisticated +But for deploying is needed when I look into the source of the corresponding ant task +The file is one I accidentally called in the issue text +Youre right I missed that were always updating the conf the job is launched with the latest known good dirs in TTlaunch +Based on Shawns suggestion create a new patch pls review it thanks in advance +Make sure to apply the latest Nature patch from +Committed to trunk +the only question left to be resolved is for the embedded what do we do with the esb extra drectories ie +I also put in code that detects readapidenied errors and attempts a login if that is seenI need to still do something about the possibility of infinite login loops will tackle that next +Resolving. +Im planning to come up with a solution as soon as I have cleaned up the profile design I currently plan to come up with everything for the in December +Added to the list of affected versions and estimated effort. +I dont get it when I run on the Nexus S on Galaxy S running +I think this would be a hacky regardless of whether its epoch nanos or not. +I have attached the testcase and the description of the steps to reproduce also very clear you can try to add the wsdlLocation and remove the wsdlLocation from the annotation in the SEI +Put the new class in core if it depends on something that is in core +I think it looks ok +As a workaround I was able to override find in a derived class and include the code to continue if headerItem is null +Needed to be rebased to trunk following the new final modifier +Changing file to seems to work now scandic letters are printed fine Odd that C files worked they are probably by default +Wrapping the query in a CDATA section is not specified by CMIS +If you are testing it with smaller example it can be the cause of the problem you described since flush will be the first moment when messages get sent and processed +What was the original shebang line in the rails before you changed it +Without comma separated paths it is not possible to use it with oozie if one is consuming data from more than one directory +This means if we deliver nexttoreceive would be set to and therefore and would be discardedI think the only thing we can do here is to maintain a hashset of capacity N that is used to check for dupes +Maybe something related to MavenIvy not pulling the right dependencies +these changes made in rev +Thanks in advance +I am running tests nowTouches the following filesM javatoolsorgapachederbytoolsM javatoolsorgapachederbyimpltoolsijM javatoolsorgapachederbyimpltoolsplanexporterM javaengineorgapachederbyvtiM javaengineorgapachederbyvti +generator has the ability to have multiple databinding frameworks but only at the WSDL level +Also in some sense it is an Improvement only +Thanks +Thanks SureshAttached patch fixes both +Fixed code formatting in patch to Maven standards +Martin you were correct fixing the first problem solved the second one as well +Its the same of having a dependency between the modules +Thanks Stepan +The code works but following things can be improved +Now that Ive gone and rewritten the patch +It looks like some of the JAXB objects are not being unmarshalled to their expected concrete types +Patch against +I havent used junit for the test cases because right now the doesnt provides any method to return and the test case needs to test method +Its tests still pass +for patch +Please retest upon receipt of the ER build to verify +This is necessary to use the new service to deploy them + +What you see on the view page is what values are actually set and stored in the db +for the patchCan you please create a test jira as I suggested earlier and link it to this jira +This issue is also addressed and patched in +Same patch for guard against negative start or len +Well make sure this is documented in the release. +If I have time to Ill do a fresh checkout and test it +Please direct questions to the user mailing list +Downgrading back to fixed it +Search likely searching for now + +Updated and final patch. +Will rather close as to be fixed I think we should better reload security when its missing than hidding these real believe this patch is finally the right way to do it +Tong Fin reports All the plugins work for me in the following environments Windows Eclipse Linux Eclipse +break caused by. +My BadPlease refresh your project example section +For example if a bundle got installed etc +It would also not need to be called but could use any appropriate name +I tried the clean trunk and it sometimes fails as well +OK is there any other tracker where you prefer this to be logged appropriately? +If I can get it to work then I will create a spec and attach it +I am checking it again and putting it here in hour +Ah nice good catch +Hiour client is waiting for fix he is waiting for project production task is required to fix this good news! Congratulations Alex! This one was really heavy. +Reopening to set the fix version +Look at the new one project as an are two files which I changed +Modified benchmark that shows major slow down when loading AOT compiled files +I like it I assume it hasnt been merged by now +Its hard to understand what is it about +ThanksIll handle the patch tomorrow +However it is low to prevent unnecessary replication in face of DN restarts and for that reason it does well not to consume too much network +Probably need a patch made against GPE +This bug has been fixed in Xalan Java. +Otherwise this sort of scenario may continue to be a treacherous time sink for others that stumble into the situation. +This version of the patch moves and into the main pig package and greatly reduces the size of the interfaces +Snappy name for a class isnt itBest regards could reproduce the described issue on IBM VME only +If you use the classes as your when launching it will do the right thing +See +They looks good nice jobThe road to seems to be shortening +processor threads and the acceptor threads are dead +Any other problems Id be interesting in hearing about +Why is not documented on web site +Im happy to close as wont fix +zip is actually Maven projectI think it will be good if we add it as a in wicketstuff for while to incubateThis way other users can try it on different containers than Liferay +Justin do we have another jira that already covers this? I believe this one is about having DFT run in a workspace specific url but I may be wrong +I remember bringing this up with David and I cant remember why we could not do so +Two very small comments Recommend appending caching strategy to the error message this stream does not support setting the readahead like you did in the case of +The approach I would take is to stick with only channel and connection exceptions and have them simply include the error code and description as fields +yes that sounds right +all issues to the CR +Fixed onNeed to fix these on also commited to Committed revision +minor comments on phabricator +Aha +The GS preview makes it use if results are found in the topmost layer only the first layer will return data otherwise it will move to the second and so on +Is thissomething you can change for meThanksSeanPatch appliedColm. +Cleaned up version of patch I sent to test for splittability +Attached the Im using +Hello TimThe fix looks fine thank youBest regards by Andrew. +That would be a reference to Ning I missed +will see +I have integrated it +Then we probably should choose the best one to commit +I also tested the examples that you worked with and confirmed everything works fine so lat step is for me to try with more sophisticated example +I will look into it. +Fixed with commit feaaceaeafeceafbb +Which module does the patch need applying against +screenshot showing some of the portlet css classes this should be updated to be more black and grey rather than attached file improved more css classes +So by moving it by it was really skipping bytes +Exclude Nodes Patch +This sounds complicated +NOTE I run it on Ubuntu I understand correctly that the issues are not the issues you originally wanted to fix +Saw it just from rake test and rake testcimiI get a similar error with the CIMI ui client when clicking on machines in the UI +Please pull the original and remake your change without reformatting in any way +Closing this issue I uploaded yet another attempt at a release not for. +for the patch +By shipping the JNI libraries with the Java code it will always work regardless of where it was installed +The problem is we shouldnt create eventdrivenroute if it is an aggregator +Reassign to stack +That said I want Seam to have an official archetype so people dont have to go hunting the web for something that may or may not be up to date +I attached a file which shows an incorrect behavior with classifiers +Thanks Freeman sounds good +May be something like this +I wanted to mention it as an option to remove obstacles to committing something early +Or may be just tell me which kind of requests youre sending so I made a similar test plan over the release configuration +Fortunately I have another instance to test right now +Want to make a patch that applies to Kay Kay? Things are a good bit different between the two versions +I have and am experiencing this problem with both the release and the dev build update sites +what do people think of meta? It seems like it could make things easier on clients removing the need forIm on the idea +Yep you got me apologies I neglected to look in one of my jar files +The second patch fixes the problem in the same way as in the constructorWhatever is provided as only use it as mechanism +the UDIL is not necessary becuase the Javamailcontainer has a routine to check for duplicatesonce it is fixed I should not get duplicates a dayStill want to add the markmail seen flag +Ill check it in +Backported patch a to the branch with revision +But for consistency I have used LPWSTR instead of WCHAR instead +Closing all resolvedfixed issues of already released versions of Roller. +One thing I forgot +There is that I should be able to remove as well +The formatting was a bit off for the code example here is the actual script +can this issue be closed +Abhinav Looks like this is a different issue compared to original issue +Fixes the whacky formatting in the last one +Terence I just committed this +Backported this fix to the branch as svn +For +I committed this. +Any progress on this +Hmm +Reopened +Thanks for humoring me. +This then allowed me to move the registerPoly static method to the I kept the point +Closing this for now since weve had no test case or stack trace since the last comment in August +Verified on JBT +I did not see overlappings here cause hibernate configuration name used in the file name to store editor state +typo of the classes are subclassable on top of the public interfaces +run ThanksBest regardsve added the files to the repository +Fixes are committed now +And also mathematicans have to life with that because for matrices for example that is a common thing +Ill fix it asap disable the test meanwhile if required. +downgrade priority for rd party issues +Logger is supposed to decide which underlying logging layer to use based on the that having logjslsj in resteasy module is a bug you would be wrong as Logjlogger depends on Logj +The phonetic package only contains the plain and those are only +not too keen on synchronized methods added but patch applies and worksapproved +Reviewed by Ken. +Thanks for reviews +bulk defer of items from to. +In Infinispan this region name is ignored and the query key is simply used as key in the cache itselfWhat Im trying to understand is what is it that youre trying to gain out of having a different cache instance for each query region? Is it setting different eviction settings for each query region? The main advantage of using region names in the old Cache module is that you could then go to the jboss cache config file and set evictions for the specific FQN for the query region specified +The problem cannot be reproduced with any Apache stdcxx release. +Ill be back tomorrow the earliest. +But Im not sure I agree with the premise +Im having the exact same problem only the first level of xsd files is having the schemaLocation rewritten +This line of code was also one of the first things I was thinking aboutI also think this is a bug in explain + for the to trunk +Reusing the term to mean essentially the same thing I think we lessen Seams learning curve +Sorry for misunderstood your comments about the logsYes I copied the logs together but make up them with the wrong order +What is the output of the following commandsbrctl delbr usrbin looks goodThe vm should use notIt could be a misconfigurationWhat is the in? +Rick Raymond thanks a lot for setting this up +Also added a test to verify this behavior doesnt change +Carl added the option name to the title of this jira +Reassigned this to Mazi + +Trivial patch attached +Unfortunately at this stage I am clueless where to start +commited again +bulk defer to +Thanks for the patchWe will study it soonOne comment though you can use the in a DN its allowed +cppsrcqpidframing +nfsfc +Imesh This service is no longer used but the source is not yet removed from the code base +This does not seem to be a Spring Framework issue at all We dont even ship Dojo support its something that you as a user choose to combine with I agree +Thus their scripts should be with them +My mistake +Rebased patch for branch +nobody could come up with a more appropriate name then either +And verified both the setup tasks are launched on in one heartbeatSince gridmix has multiple simultaneous jobs ran gridmix and verified from the logs that multiple job setup tasks and job cleanup tasks are given to tracker in single heartbeatRan test and Verified from logs multiple task cleanup tasks are assigned in single heartbeat +I think yarn jar is basically indistinguishable from hadoop jar for purposes of launching MR jobs +I didnt have intention to I was looking for libraries for that and I saw that XT Ajax had got something like that +Fix committed at apache will sync into apache when we turn syncing for back on shortly. +What seems to be wrong is the local variable table which ignores that the value is a Reference instead a boolean nowSo this bug is about a variable table entry done it be possible to backport the fix to? This issue causes confusion for developers trying to debug their Grails code in a debugger +Thanks for attending to this so quickly +but if you put lowercasefilter after it it passes The problem is this stemmer incorrectly creates some uppercase stems from lowercase words +Ive never messed around with the eviction times but theyre easy enough to specify via or I can split them into their own properties too if youd prefer that +Hi Ron Ill try to look at it this afternoon and append my thoughts later today +Thanks for the input +This project depends on an ejb jar that isnt on repo +Could you please assign it to me +The problem is that I have to update them every release so the more examples we add the more work this task requires +Simply delegating the logging to another thread isnt likely to improve the throughput of logging +Thanks Devaraj for kindly reminder +That way calling forceFlush wait will be guaranteed that the action waitFF puts on will run after the flush completes +But Ive not checked details +This unfortunate synch is ok because the secret manager is already heavily synched so I made it no worse than it already is +I managed to get a test app for which made it possible for me to both reproduce this issue and demonstrate that my fix resolves it +Thanks much AustinIm marking this as patch available on behalf of the author for consideration in +As an idea could it be possible to use a code generator to override the finalize method of the pojo to include a Spring callback so that when the pojo is garbage collected the user defined destruction method is called +the bug was open for years +Hey travis I have made changes to have one method +hyperichqbuildsqlbuildsqletcinstallersrcnethyperichqinstall has a bunch of Hyperic HQ in it +Seems to work for me fine after applying this patch +Also I dodnt add the readonly tags +removed the onload event +m repository if Dresolversinternal is passed AND local repository is present +Please do not say anything about acceptance or rejection before Google does +I am testing on bit +If the context initialization is moved from after to before a conversation is not correctly stored especially in case of nested conversation or long running conversation +A bettercomment in this test would help +Duplicate ticket nothing to be done. +Committed as dbbddfffebffbde to master. +This will help debug the real cause of the problem +eap runtime excluded from and included in +Checked. +and tests now run on AS trunk +Fixed F too +This is using code from CVSRegarding other options Im open to alternatives +Adding capabilities +Using as the abstraction for method security avoids having to write separate versions for and Spring AOP +Maintaining our own history on Android has made the most recent refactor much harder +To help out Toms checksum issue we could support sync markers +Fixed points are +After applying the patch national characters are still displayed incorrectly + for the patch +Also since there is a workaround the fix isnt critical +Yes timeout condition is moved above just like for process we can set the default value to frequency or minutes if user has not specified the timeout or timeout exp is broken +Yes you right +Fixed a regression. +Applied in for a while +rvt still around are cruft Im going to remove right away +Todd thanks a bunch +so need to sort this out soon +the other risk with all the options is that someone has to test them allOne thing Ive been thinking of is making it easier to install plugin rest endpoints into manager and worker nodes the way is hard coded for and the standard built in servlet pages +A new test cannot be created since some classes need to be made public in order to plug in an external implementation +Find this properties file orgjbosstoolssmooksgraphicaleditors in this file you will seeTask Tasks do these two strings actually appear? I still just see Tasks and Task instead of Tasks Map and Task Properties in the the CR stream Im still seeing Tasks and Task for those labels +If putting it in we dont need to implement the checking in each of its callers +okIll tryBut I think it may be only as a temporary solution +The lack of Team API is blocking our multi VM work on BC and applications +It should be add instead of add +Something like The encoderdecoder has generated a Null Pointer Exception +The wasnt pushed at the when doing a polymorphic call through the block protocol like blockfoo in your example this made subsequent block protocol call be called relative to the wrong context of the super block A instead of relative the context of BPlease check if it solves the problem for you +We can revisit it in the future once we figure out how to make it work nicely but for now we simply dont have the timeresources to make it work in the short term +I just committed this. +Tested on Spring Integration and Spring and I couldnt replicate it has been fixed along the way +I just committed this +Attaching new zip to reflect changes called please use this one for testing for the heads up Steven ill go grab latest snapshot now. +OK thanks for the explanation +A disadvantage is that it requires modifying the scheduler data structures to separately account for requestsIn another suggested by Arun in and implemented by me in the March patch if I want a container only at node I set the flag on the rack that node is on +Maven does not understand that a dependency on a different version of the same artifact is in fact not a cycle +marking this as resolved since Nick already have a script that nags about branchingtagging. +take of the total time of my benchmark +It does sound plausible to me that the nonce is connection specific and theregets accepted by Apache HTTP server as long as the connection is being kept aliveI also think that the fix to the redirect problem is relatively simple +Im able to install the version of groovy plugin with both and plugins in it +In patch v I think we should online the region only if the server is still online +Normally Id push that effort back to you to address but in the interest of getting this in a build Ill fix it myself +Now that work is well underway with AS all previous community releases are +Lets fix the issue at handI have made the following changes allows variable in stripHTML attribute allows variable in formatStyle and locale attributesIll commit this shortly +I dont really know Perl anymore +The current rtfdtm may change between the push and the popI dont believe so +A reason to change the package name would be if you wanted to use the old and new version side by side but that would not be a common usage pattern for ACM I think +Any objections to committing this patch +Patch removing prefixes in file names generated due to incorrect diff command +same as the one above with different order of the concurrent paths +Here is the updated patch +Does it work in the booking example +Theres roughly bytes of overhead for an empty mutationWe could encode the lengths and timestamps as integers +So what is exactly the problem does the selector hides some resources? Could you please describe the usecase to be supported + +That does I doubt you want to use if someone can attach an example which shows this it XA transactions resolved this issue for me +Thanks! +Committed the a patch with revision +should be part of has been set for release on December still current? If not please bump to Unscheduled +tested on windows xp sp and osx under tomcatx jdk start and stop output are present in +Perhaps Ersin you can provide an assessment of how much an effort it would be to not have position factor into the terms of the ACI +bq +If you want the results set columns post resolve then use the get method +cordova +I like Pauls idea of having the code catch this and throw a more meaning exception messageMaybe we need to expose some way for the Portlet to ask whether this option should be enabled without requiring the user to try and create a connector first +Can you try it with the full path usrbinpatch +Is this intentional? I dont see stopSlave any other place in the admin guide +What problems are you having +in the case of a search +Apart from that it works well +Could this be a consequence of not declaring a transaction manager in? We use JPA but not +patch applied in dacfedcbbbdbddecThanks Conan! +Ive attached the missing file +Thanks Daisy +I can make changes if required before we start syncing our +I had some experiences with hupa projects gsocI have web designing and java development experiences as well According to the description its may interesting work for me and I think I can make a good valuable participation for it +Now paths are generated wrt issue +Ill spend to work on this this week Ive done some testing however I am starting to think the problem is larger than just here as I get the same performance with or without synchronization +Attached patch adds a check to the JAXR test case in scout workspace +This also happened in a Stateful Bean where the producer and consumer of an event are in the same stateful bean +is for feature development +I felt that we would be more likely to cause the default value to change than to make a change that influenced serialization on such a trivial exception typeLow risk no matter which way we go Id rather focus our energies elsewhere +FTP Message filter also has attributes directory so I added the directory pickup for them as well +Ive created a draft of a release note in the Release Notes Text box could someone please look at it and let me know how to make it accurate? Thanks +Neither help +DimsYep I can take a look at the JMX layer +My close this issue +Since the upgrade to webflow webflow no longer throws these exceptions but returns a for invalid that should be that the flow is automatically restarted for invalid states +Its surely not an critical issue since the users could workaround this by setting the correct hostnameip +Im going to regenerate it and submit it to again for another round of tests against current trunk +Has this been fixed? The comments indicate that it should have been fixed for +attached with tests for Repositories and NetworkTests run Failures Errors Skipped applied in close the issue and will open a new one with the specific tests to be added soon +Here is an intercpetor that works around the Websphere query string issue +Thanks for clarification +I verified this on both Windows and Linux using +Jrn can you test this +If you are using the you must duplicate the bytebuffer otherwise the in does not finish correctlyYes its basic hygiene not to modify the keys out from under the mapper or more generally Dont mutate method parameters since the caller may still be using them +Resolving for now to make sure it makes it into the release notesPlease reopen if it turns out the fix is inadequate +Personally I am OK with anything in depending on +patch is committed to trunk r and openjpa r +Other tests are passing +Restarted jboss server and found there is no +Are you sure your configuration is correct? Seems like the login module cannot be found +but Im not sure how far we can protect people from shooting themselves in the footOne argument would be we could go as far as to not give them a dangerous loaded weapon and instead do it properly with APIs once we get there +I think I will commit this patch and then file another one for the fix you suggest +Thanks for fixing itTeresa +Applied in r thanks again! +SVN diff for the is attached here +For ZWPS feature it should change the way how we get hypervisortype for a volume since template creation is using volumes hypervisortype to populate hypervisortype column in template table +Trivial patch attached that just changes the expected exception text +For Hudson we might have to reduce the number of transaction in this testActually this was not common till build or so +Thanks for the patch Sean +Thanks +Hue over here and support for it is definitely scheduled to appear in Bigtop releaseOnce that happens our hope is that Hue will get to be used as a platform for writing any kind of UI tools to interact with the Hadoop cluster +added link to keytool docs in +Ill give it a try tomorrow morning +Im not going to do this right now but its something well want to look into in the future for cases like the Ruby parser +Patch fixed the failed test case +We are not saying whats in contrib is Apache Accumulo code so it probably does not need to be in Aapche svn +Thanks for the clarification +Closing all resolved tickets from or older +Can you provide a sample your Spring configuration files? I suspect there is a problem elsewhere in your security configuration +Please report back if you still find any issue +The status would include a counter that would get incremented when the application called progress +and +This version of the patch represents a way better fix of the result of execution on my machine overall +But anyway Ill just keep plugging on a couple of hours a night until its all finished +So codeph is a good one to use +The patch failed contrib unit testsThe test failure is due to +Thanks very much for the feedback Ive taken your suggestions and amended the patch +I added it because I noticed people were comparing Strings containing xml +It is just not easy to see what is going on in the code for the patch! Im running a test crawl with this patch so far itsrunning well +What do you mean after a URL redirect? Are you on a new page after the redirect? If so you shouldnt expect your callback to work since your state has completely changed when you loaded a new document +Thanks for your contribution. +I think you need to change the port in the custom configuration context or you need to reuse thatThanksDeepal +I will commit this once Vinay reverts back with the system test results +Repro Im not understanding your program very well but it seems like you are creating a statement in but I dont see where you close that statement +Committed to trunk. +Thanks +Unassigning myself since Im no longer working on that book. +Test case attached +Sounds good then +On the other hand we could do some minor chekcing on the server to see if the path is null starts with and others +If you felt like writing one and submitting it that would be a nice addition to the test suite +Also little things are getting cleaned up in the realtime branch +Are these based on interfaces +and leave them deprecated of that someone still uses them somewhere + I solved this problem by delete it and create new one +This modification will solve the problem +Ryan added in IRCthis may be the root of the problems I was describing above some threads may be dying due to OOMIm skeptical that OOM could cause CME though +As an administrator currently has to publish the gadget as a manual step there is an explicit action being taken by a human before any gadget is available for general consumptionWe should make it configurable whether a rave instance allows this feature to be enabled but I given the constraints above what are your concerns +May not be a need gravitation of data in buddy replication will solve this +For this case I cannot recreate whatever John is placed the in default +The implementation class remains hidden +attached to can be used to asses the CPU penalty of this patch for larger buffers +the only reason I can imagine for this OOME to happen is that javac in needs more memory since it works without changing the Groovy version +Thanks Ivan +Corrected the package name +how to decide what is the relevant part of a stack trace? How to detect recurring log messages? The problem is more complex than you think. +Checked in Ankits +is basically just a hack around hardlinksYep +Also I dont want to create a namespace element since it is not a mainstream requirement +Hiregardless on whether to start with branches or not I would like to investigate a possible solution on a fork keeping restricted environments like Android and GWT in mind +Actually Ive been looking into that and similar requirement recently so well do something with it for +ill submit patch once ive tested it some +Got an example code snip +The solutions is downgrade from to uses hornetq and uses hornetq which can be the root cause since AS uses Quickstarts probably should use the lastest version +I like tightening up our inspection of an event such that now we also check source server not just even type +After more looking I think I have a better understanding of what is happening +Hi Erick a change log entry was added with my first commit. +Is there anything more I can do to help resolve this issue +need to also handle support a pull request for +Tomek Korzeniewski found a bug when global context is defined +sorryIm stil looking to understant the need of redirect versus render! +for PR +I dont want it to grow again +When your registrar asks for the context it forces a premature initialization and that causes the use of the XSD namespace URI with unpleasant consequences +Duplicate of + +Thank you I thought that because the other issue was closed that my comment would slip between the tracks so I created a new issue and linked it to the other +to trunk +Updated patch with better delta update capabilities for +I am building on my system now looking into a couple of options +Thanks for pointing this outThat additional check is indeed necessary +Yes please +I added paths of Java sources from these projects to the set of scanned paths and model by default created links to folders outside project +has been marked as a duplicate of this bug +Sorry I dont know more about how starts up to help troubleshoot this more + +If this is an open issue due to a problem in the JVM it should never have been resolved +I think the only thing to do for the issue is to also port the changes made in to +Chris +is the most recent dump that we took +patch includes Treecache Cache and ehcache providerThere are two components to be configured Pagefragmentcache and the is provided in the patch as other should be implemented as an abstract base class with an instance method The decision about which cache provider to use should be defaulted based on what classes are available and precedence cfg the variable name is wrong it should be configurationFile the configurationFile should be defaulted based on provider the page fragment cache should use the cache provider installed with the option to override if needed the class is then unnecessary Versions of dependencies are declared in not Why are methods like start stop Why add not put? Exceptions should not be wrapped I didnt review the looks great thanks +Please change to if you intend to submit a fix for. +Patch is getting big enough that it will be really annoying to keep up to date so Id like to commit and start baking in trunk as soon as possible +My mistake the error is actually the inability to load +Same problem here +Hi RickThere is no bug on this +It calls directly on the rmi stub so the marshallerunmarshaller are not used at all +and should adequately address this issue +I see your point +If I have something I stay in touch +I would argue that the maven jira projects at codehaus are managed by codehaus on behalf of the ASF Maven TLP +Please read has been added with description font files used in example part has been added with description font files used in example part am sending the font files too +Applied in r +If we can avoid it Id prefer to not add it provided we can fulfill your use case that is +Please if this comes back. +Works like a charm with thanks. +I am terribly sorry that I have not yet submitted the ICLA I have signed it already but had to wait for a final clearance from the University that the source code was mine to donate +Kirk as we have fixed a couple of issues related to the use of both the location attribute and derive could I please ask you to try your test fixture against the latest sources? I hope I will manage to ship +Add an editable flag to the viewport so that a client can lock it down or allow it to be to this to again will do under its own issue +We could also consider what happens if the two hints hit different magic matches +Completed At revision +Management server log file attached +fixed in rev . +Please verify +If anyone feels strongly I can change it later +Cant reproduce this with the latest embedder +It is easy to fix +The API for managing task instances which will be part of the task management component and external to the BPM engine +attached a patch to be able to run a build tru mavenNOTE some tests are failing I will provide a better patch laterPing me when its committed and I will activate the jenkins build for sonar analysis +Applications will need to set a new permission in their policy file +Graeme Hmm I cannot reproduce it either both with Grails and +So Im afraid we might not be able to do much about this other than copying the relevant code +This is an initial demand for my try on implementing +Patch works for me if there is an existing testcase id be happy to look at that as well +Changed +is a sample implementation of which is used by +exposing EJB as service support +the client and broker compete to register their respective SASL plugins however the broker should always win but if the broker fails to register then non of the Qpid SASL plugins will be available and authentication cannot take place +Ive reviewed and approve +Bumping release target from to further input and anticipating the release shortly Im bumping this issue again to for it further to as still waiting on further input +I cant find a way to remove that gray barLets mark this as resolved and try to fix the grat bar issue later +See and there is anything else still outstanding please open a new bug report. +I kept it going for rounds and found no issuesThanks a lotVishal +If you finish it before we release beta please correct the versionThanks +Ran svn up and updated the diff +Verified +behavioural change is okay with me as it is in agreement with the specsOrtwin Glck +The search should be ordered from the implementation class through its superclasses to find the most specific implementing method +This doesnt seem to be fixed in CR +Thanks Tom for reviewing +This does exponential backoff on the logging to limit log file size in pathological cases +HiIve tested it at the time when the bug was fixed by Sebb on following OS Ubuntu Windows Macand the solution worked fine +I think that so long as someone else doesdid the work im fine with this +I was hoping that we could eliminate another discrete manual step in release generation +The patch submitted by Stefan has been applied and the class and reference into file have been changed to and defaultI have tested the different cases I found in all the comments and it seems to be ok +The patch for nested queries seem to have broken some grouping scenarios +Put another way why shouldnt pull in Groovy as a dependency? What sort of problem does it create +It would be cool if functionality could be merged +. +Thanks +Since Im new to Thrift Im quite unsure if the sequence of bytes read from file handle fd is already in host byte order +It would be nice to fix the bug +I think renaming DEFAULTBUILDQUEUE then will also not work since it is referenced by name at various places +I have actually been able to connect using jconsole remotely but in this case I had to use more parameters +committed to trunk in revision thanks varun +And if make it to load user libraries I got which is a result of different class loaders for the same class +I committed the patches to trunk and +Committed to the branch too +This makes things more difficult on our end however it should result in better QOS for clients +fixed at with the release of. + +js drop +Commited fix to trunk +After the wait it will check for thread events so that if a kill or raise interrupted the target thread it will propagate think this is a reasonably good start but Id like to have a look and see if theres any reason not to expand this to the general case +Thanks for reporting this +The javadoc warning is from common and is unrelated to this patch +It should be possible for each consumer to choose its preffered provider based on memorynetwork +Committed to both trunk and +for to trunk +domj to dom parser was modified in order to load processes using DOM objects +It then recovers the inprogressznode which will fence the ledger which it is using +When that is complete your manufacturing help in German will be presented in the new format to anyone invoking help from any manufacturing page when their locale is set to de +In our environment buffering query results is not a problem but perhaps that is not the case for Paolo +How embarrasing +Less charset decodingencoding overhead +and +How I applied the patch +No problem there! So the problem might be the tomcat el implementation +Since is still typed you cant assign a List to it +Minor comments at phabricator +Think itll work but not sure +mvn clean really helps +applied the patch and added some enhancements to itthanks to ukasz +Thats what Im trying to say +pushing off remaining issues in order to release version in JIRA and create release notes +The spaces are removed when I save and the weird strike though comes in +I get the same error as in the description of this jira issue so it does appear that the error occurs for lambert conf conic SP Geotiffs +Using throws a argument type mismatch exception +Covered by documentation and sample. +I just backed out from branch because of +This issue looks like a subfeature of agree +Working with I found that it would be useful to support CDATA sections in the XML that is being processed +I do not see these exceptions at all in my environment using the Woodstox parser so this must be a parser specific issue making it difficult for me to test +This is what I think we should do if no transaction present enabled +Included that fix in for patch v +no it is just similar to it +databindings +Causes an issue withRevert Reduce number of requests when querying for subtasksCommitted revision . + +Ill try to get a patch up tonight if its blocking you I can revert it +run main method of in attached application to reproduce to release since is being released +Hi DerekI get your point but I chose those class names because in my mind The handles exceptions generated during Ajax request processing by +That explains the version number differences +Passes all tests locally +Simon thanks for your fast responseYes it is a multiple module project +Have not been able to look into indentation issues for +bq +I would appreciate the communitys continued feedback +Please provide it because Id like to see what you are timing +Uses for simplicity +Response cookies from the target host are automatically +Both propose transaction support +Refer to the repository problem reports for more information +Hi Guys feels a bit like ground hog day here It would be great if someone that has been working on this issue could review and close if this has been sorted out +Is there better terminology for the fact that they used the ! notationQuestion for Noble If a shard field is specified and there is a ! on a document key which takes precedence +Bulk closing stale resolved issues +I think this patch will be the best solution that the Java programming language can support +on the latest patch +Java code to reproduce the a fix for the problem +Look again at the surrounding loop +I just committed this to trunk and +But Im not sure its really a problem +We dont support starting a queue from outside of +the policy to fix LGTMTwo comment +I can not reproduce the failure on my Fedora ia box with IBM VMEWill try with DRL VM +I have created for adding new HDFS client +Patch attached +As Vinod tells me the code is already tainted with the uber AM so Ill go ahead and make the quick fix for cleaning the staging directoriesWell have to rethink the code when we do DAG of jobs in the AM +The only real difference appeared to be the TableKeyspace naming of things + +Or test for POSIX compliance at configure time +If this is a temporary problem please let us know when it is reliably fixed +Changes to use the new property +Previously it said EAP in some log messages and Enterprise Application Platform in other messages +Committed as svn closing. +Need to write a unit test investigate +Dhruba sorry for this late thought +Applied. +You sure about this? This is how we have uploaded documentation manually for years +Thanks to Jaimin Jetly. +I tried to join the installer jar but thats not possible since it is about Mb and the upper allowed limit for attachment is Mb +New Java Project +But still it does not work +You have to start from scratch +I am experiencing the same issue as described above +etcWe found some things very strangeIf you apply the version you can use this version but you have to use the library of version then it works +lgtm will commit shortly. +Checking for appropriate if blocks around costly debug logging should be a responsibility of the patch reviewer +It worked fine in dev environment but when we take the tar file and install it we see this issue +Synonym + seems to be used for properties elements and their parents containint the stylenameTherefore the given patch seems to be the best solution ad hocCould you please review it when you come back from vacationThanks in advance and push this issueBTW I face the same issue which you faced under JDKW but my test environment is JDK +Thanks Eric! +Going to commit to trunk and branches soon unless objection +Hence Ill mark this as Duplicate +I committed the patch to trunk and +User specified directory is not created by review +Marcus your contribution is extremely welcome +Fixed in r +I think adding a configuration for viewfs for the home directory pattern would be the best route +There should be no subtleties I mainly moved code from the main method into a new method and indeed a bit of +Or do you know how an inherited constructor of an already existing object can be called? mimics Java serialization as close as possible but it has its limits due to this internal stuffYou might get away using the but thats a question for the users list +I think forcing a client close when the server closes in this situation is what we need to do +Sorry havent seen that +Agree but why not providing the dual packages in or +Heres a patch +Ill take a look at it +thanks for cleaning up the patch stack +here are correct and issue to write release note +patch is done again after the previous patch failed on the again +I cannot reproduce it now +i do in fact agree that a global lucene option is a good patch was committed to the upgrade branch +Thanks +Fixed in r +Either increase number of allowed files or file upload component to unable new files to be uploaded. +Previously the URI would have been updated +The include order of the +I also added the tests to and +Please publish the war and the patch at your convenienceThanksDave +Id rather not count on file length as well +A patch to fix NPE in start stop and shutdown methods of class + +Nick can you take a look and see if I missed anything? It looks ok but Im not sure I hit absolutely everything +gradlew build or other targets to double check. +But I didnt find the method using +So if you want this fix this defintely isnt a bug but at most an enhancement or request for this kind of functionality +srcmaingroovygroovyxjavafxfactory onIs it just that has changed and this needs to be encoded in the source +Thanks. +I dont like testing +Can you provide an example which illustrates your issue +When we start job tracker and task tracker using sometimes job tracker is still in initialization and task tracker is stopping because the job tracker is not available by that time +Thanks +Updated patch removes the vestiges of the htrace related code noticed it looking at the attached example output just now +I hoping the xerces issue is resolved as well + pending ve just committed this to trunk and +Ive added a patch to allow support for +This value seems to translate into that awful non ascii character that prints in error messages ala +I solved all the issues about integration tests I think that now this first version of the connector could be considered completed +I just realized I can edit +looks good will commit if the tests +inlined the PJCRC constructor and committed +Thanks for the instruction and patch JrgenCommitted as revision +that is with blank field and filesystem it also resolves the relative path as if the project name is specified +Stream pumping threads from system call were generating debug events turned them off for those threads. +The full configuration settings for the plugin will be needed to fix this one +Just realized I never reported this +Afert I deleted them everything now works as expected again +Code looks good +I couldnt wait to try GA into production +The WSDL is generated from choreography using Savara wsdl generation so we need to understand the implications +I have an idea I switched between Maven and Maven +downgrading from blockerMain concern here is to have the max versions set to be helios so users does not accidentally install on indigo +This is an expansion of will be split up into multiple patches +Ill include this check and request a pull request +This changes doesnt affect the behavior of the function or not reduced any functionality of the script +Also in the future as the page says it would be good to try and avoid making unrelated changes such as the expansion of imports +srccorejavaorgapachedroidshelperfactories is not synchronized since their access may have to be synchronized depending on the usecase but ATM I cannot think of oneThank you very much for the patch please go ahead and commit it +It sounds like these are internal classes so renaming them should not affect the API that Jexl users are likely to needBut Ill leave for others to comment +The EAP dependencies should now all be using redhat version qualifiers +Thanks Sergey for looking +Merged with trunk again and fixed all tests +I think it is fixable before GA +Not only at shutdown but if there are concurrent region close operation in the HRS there is a chance for this +In I can only see used rather then just Envelope +Thanks for pointing this out +Updated karaf scripts to match karaf functionalityPatch based upon patch provided by Jon Anstey. +I dont think this belongs in +comes into my mind +push to +I agree with you its serious problem +That is fine +Some are still open iirc +Attached a spam email in moderation queue +I use solr but I looked for this changes in solr too +Manually tested the patch by running failing streaming job and successful streaming job +some touch ups to do for Final but most resource descriptions are now present. +Technically this isnt a bug +Ill test the code and then if it works ill commit the patch +It seems safer and is more intuitive to go with a straight overloading think the log message should note the ambiguity though +is this something we can add a unit test for +Ive created to address the var expansion +IMHO it is a different feature so I opened this issue +This looks like a duplicate of. +A lot has been committed beyond the last patch +I would look at synchronizinglocking around the get and put +back to extension points were commented out +Since the work path and staging path was same for both the cluster this issue came up +Delink? you me remove it as? If so Id like it to stay as subtask as they are related +please review and update +Morris pointed out that it is better to implement every methods defined on interface for innerclass Attribute in order to avoid future confusion +Committed to branch +The error was introduced by not handled well inPlease review the patch again thanks +Client will send OPSTATUSCHECKSUMERROR to datanode to inform that there was potential checksum error +fixed we need a release +I generated the patch from trunk directory +No need to port +Nice! +When I call get on line am I not using same address that failed above when we called verify on the line above at line ? Thanks +Would it be useful to propagate the parent process arguments in addition to what you have hereMakes sense +Committed to trunk only + passed in build Can you publish your patch along with the test failure? excuse me about the formatting +Its the first time that I have tried + patch looks good +A note has been added to the reference documentation and the javadoc of the appropriate Spring classes +Ill try a few things and see what looks least ugly +What are the semantics for triggers when the commit log is being skipped +Detection works for the two relevant servers now. +The AIOOB exceptions I was hitting were truly bizarre +All unit tests passed save which is +Exception handling has been covered as part of Chapter mostly as part of the listener interfaces +Closing this issue for now. +add a patch for chinese resource pls help review and submit it thanks +When opening the above link to fisheye and also on a fresh checkout from CVS there is no message that refers to +Updated the patch to match the new source tree +Maybe now is a good time to move most of these features into the hive binary +Its ok +This is a nice patch and saves some annoying manual work +I actually updated the pull request about an hour ago +I believe these two may be the same is fixed in with the webflow upgrade +here is a fix for the nocommit robert put into +Is there a full stack trace? Is this occuring in a Handler or maybe in security processing or something +My assumptionof what is happening is that the page actually gets corrupted before it gets recognized +The patch can wait until I have time or the community has enough interest to create a test +Cannot reproduce +Ive opened to have this fixed +Thanks for the unit testThe problem is a bit chicken and egg +Got two questions though why did you choose blocks to check at a time and do you think it might be worth to make it configurable? And the other one is how much memory pressure would the checkedBlocks hash set add +Update Association column as discussed and fixed NPE with dev cockpitProvisioning and batch is working fine for meBut still an issue on the provisioning page even if theres an error the project is provisioned +just instantiate the List where its declaredColm +I strongly agree + +Thats the whole point of the token endpoint whatever the grant type +I have executed this in loop +these bpel project examples should be done as bpel project examples provided by the bpel plugins not in the shared project examples +without luck +We have downloaded the current CVS code and it has not made a difference for me +Thanks will do +This is my task designer based on Oryx that can export jPDL +I guess if you have hundreds hmm thousands even generators you might get a GC hit you would not have with Reset. +Instead of resolving just the discovered fragment bundles the current code resolves the application content and the fragment bundles +Im checking whats inside +Fixed by patch on of +I do not like the ideaof just ignoring exceptions altogether or putting a errorwarning message forwhat is essentially a normal behaviour +will make sure and fix +I agree with Geir +By running continually in a long running process you get the advantage of JIT optimizations as well as one time startup costs of both Maven and +Im using windows Jruby Jdk Rails +committed in revision workIm not sure if we should place JCR extensions in after all they are still subject to change and it might well be that the interfaces will change before JCR is final +Heres an image for the installer +With this patch the emulators will return progress when the plugin is disabled indicating that no emulation is required +Before creating a graph remove the hash +mrepositoryjunitjunitUserstucu +It is simple to use the plugin +So I wonder why the unversioned XSD in the web features this attribute at all +Also the chinese and japanese characters sets do not use unicode encoded values my commit would most likely have corrupted the characters +I didnt start the server with b option +Max Seam project enabled by project preferences have no deployment type preference +I will circle back to this one and finish it off after finishing up with the indexable binary stringsIll implement with a platform setting and an ifelse at this time + for option +addresses Dougs concernsIve also filed to remove methods deprecated by this patch +Joel does that mean that the base functionality of having configurable collector is gone now? I dont see for example a in the related issues +New patch changes the CLI argument to mysqlmediumtext and has a minor update to the help text +Image showing the original set of mappings +Pull requests merged +patch to for did two fixes on joachims great work fix the path where the messages get stored connect the store +by clean i meanyou dont have to patch the wicket jaryou dont have to have a package in your application to override a classif you take the hybrid strategy source and paste it into a new class in your application while tweaking it at this point in lifecycle i would consider even that clean. +Is the log on the client system? Is the log from a Web Application? Any other information +Im on the road +I will commit the current patch shortly +I have started work on this +Updated patch that completes the +But Im not so familiar with the JNDI SPI I dont know if this would be really helpful +I dont think we should support changing dependencies through if aliasing through isnt enough then we should concentrate on xml configuration in car files so its relatively easy to change a compiled module +As requested +Unless there are objections I am going to commit in a bit +This would allow using a query anywhere one currently can use a positive queryOne could simply and naturally do fqid to filter out a single document +For XMI serialization compatible means the services type system is a proper subset of the clients type system +Patch including unit tests which appears to correct the bug +Changed fixed in version to FUTURE as Sybase is not yet supported by +I agree with Knut +The zip the a java class which implements certain functions required by the and the a java class representing the objects to be exposed from the tree +Above is a wrong patch update a correct one of patch +Minor change some of the example code incorrectly referred to the resultset reference as rs instead of uprs +Thanks Todd Fast for this workaroundThat also solved running Nutch within Eclipse +I changed it to treat block size with an argument based on a previous test result. +The patch cant be used in the PDE Runtime +Changing to Master fixversion +Closing resolved issues. +Here is the patch for the sameThanks and RegardsAmit Amit for the contribution committed in trunk at r in release branch at rAshish AshishRegardsAmit Sharma +Did you see for example any IDEs content assist or another desctop application assists +Can you take a stab at this Mikhail +Here is yet another version of this patch updated to trunk as of +Explicitly the methods on the Principal interface on the +Java is required to run EAP Show it statically on the page of the JVM on the page of server setup add it only if user chose VM +Committed fixes for this and local I can verify ESB but need full rebuild to see links to Teiid working well +Added synchronized on parent when obtaining the deleted docs in +Similarly for HCD +This way we dont have to duplicate any code dealing with the modeler download into the distro build which saves us some time and users can still patch their own SignavioOryx wars with what they get from the this issue to me if you want me to do it that way. +Multiple things affect the downgrade of this JIRA +Marking as closed and renaming next release to + +Ive just committed this +I dont know enough about Felix versions to tell if is intended to be R compliant but the exception is telling me the BundleResource objects it produces do not have the capability +From your point of view would that be ok or is there a reason why I shouldnt do that +I checked in a workaround to this +Still reproducible at +It doesnt fix the rendering issue with transparency +The difference with the issues that Ralph mentions is that in this case the Qpid test code would have dependencies to GPL code and that test code is part of the Apache release of Qpid which IMO is not okPhilip did you try contacting the author to see if a change of license would be possible? Maybe GPL was just chosen by default with no specific reason +Accidentally I found this issue that introduced a problem of mutual calls between classes over packages when initializing the classes which I had carefully unbound in. +Users dont have to know how to run their app server to run with this agent because the client classes the agent would be trying to hide are not in the +Perhaps look into treating gaps created by OOB message reception differently? We do know whether a message is OOB or not +comments are welcome +During this sleep if we change the blocking mode it should not have any affect +since this wasnt fixed +This seems to fix the group by aggregate issue for us +Also attaching the generated release notes for review Committed at subversion revision +I took a look at this tonight +This is available in AS the questions is why it is not in the EAP The environments the TEIID Jopr plugin needs to run in are the admin console in EAP and +I totally understand your concerns regrading the usage of these attributes for the lookup +or maybe the ALTER TABLE activemqacks drop constraint activemqackspkey syntax will work across the board +cpp +Rewrote so that it will be much more efficient creating unique names +Weve got to discuss on the mailing list whether this patch will make it as it is or whether additional changes will be required +Fixed within +I created for the nightly docs +The fix and test code +Luckily I did not run anymore in this exception under load So seems to be better now +You have a patch Igor +I am close to considering this issue fixed but I dont have the power to make the appropriate change to this issue. + +Please verify that it has been applied as expected +I would do more than happy to assist with this plugin since no one is stepping up to do it +Had to make this change for our build system to display logos correctly on cover pages +I tried it +Looking at github seems like this commit was not pulled back in master +this patch there isnt much difference in performance with the two buffer sizes so it must be related to the NIO implementation in some way +at least one has an unrelated temp prob +Is there a JIRA for this issue? If so direct me to it? If not please direct as to how to proceed +It is a good idea to document and +Smooks editor is no longer being developed +New features should not be blockers. +other than thought this doesnt seem to be related to the Eclipse issue +On the other side different port for management interface can be defined for standalone and domainSo question is Can be shutdown script extended to reflect custom port for management interface? Is it enough to add information about port limitation for shutdown in documentation +actually the was a bit old sorry for that heres the fixed oneeverything else seems I noticed that in contrib we have the same logFile name in the target as the one that appears in ql +Version v fixed the warning above wo changing the guava versionHadoop branch does have the same problems as hbase but only on Journal and in the tests +Sigh this is a problem I dont have a developer environment on Windows +if there is no replyDestination explictly set then get chance to create temp queue srcmainjavaorgapacheservicemixjmsendpointsTransmitting file dataCommitted revision +What we propose here is standard +bq +Reopening in order change resolution +Right if thats the case then its not something specific to its really a feature to add this capability to any Cache annotation +I did a bit more testing of shell +Here is a working pom for staxs missing info license url description +Ok that makes sense for it working on Mac +Hi This feature will be great for form based login pages which currently have to be built without using a Form Click controlyes good use casenow that the release is out there is not problem to add this feature rightregards +If theres further issues file them anew. +Something more subtle is going on +Thanks for the clarification +It worked fineI have committed this +Fixed to always use even in a managed transaction scenario +Without the right values in their things like producer audit and such break +Closing. +Thanks for the patchI improved the patch to check that the FILENAME was already an Expression and use it as is +they look good +Now Camel jdbc set the auto commit flag to false and commitrollback the database operation at the end +patch looks good +This is fairly easy to address +There are still a bunch of commons projects on can I move all of them over to Nexus and get the rest of this setup properly +Thanks for the fast comeback Dave + +Please find more information about the crash in attachment +Thanks! +Any opinions on that? We could either backport the offerService refactor or simply leak this again as proposed in my option above. +Andy have you considered using Result Grouping Field Collapsing for your +Using revisions for versioning could prove useful but the fear that compaction is triggered accidentally might prevent its widespread use +This one applies cleanly on svn trunk as r. +Seems like an error would be better +Marking closed. +Any feedback appreciated. +Global sync is expensiveDo you think flag files are good? This looks a lot like race condition problems to meZK would be an option yes +Done in rev +Attached diff for component adding extra Javadoc +Hi BarendI know realize that I have committed your patch but that you didnt checked the licensing checkboxIs it possible that you check it so I dont have to revert the patch? Thanks +Looked at the subsets of the patch +Perhaps we could add a search page for schema search +If no one objects Ill commit this one in the next couple of days +Thank you for the commit and the guidance through the patch OysteinNarayanan +Not sure how I can help out here to Jrgen as hes most familiar with the LTW infrastructure in Spring +I see two options either keep id as a required field but use it as the thread prefix only in the nested case or else create the name from the owning beans name pool +This issue blocks us from testing http session replication and ejb SFSB replication this was not seen in the alpha build which just had Beta build was upgraded to and we see this any comments +There is a default repository or repository but the can connect to and choose their repos from a list +added additional refactorings and unit tests in rev this to Bernd as he performed the major changes +Maybe thats just how its showing up in the diff tool or something +Verified by Vladimir. +Update +Otherwise would be rounded to and to +The testsuite still works as before but please retest with your custom portlet. + +Vinay thank you for the patch +some even more +I have modified the CSS controlling the display of this console to make it look like the attached screenshot +Related issue should be solved at the same time +Please explain the steps that how you createdeploy the +Grid was this delivered to just the branch or the branch and trunk? Trying to determine if it made it into the M build +are you a PMC member for Shindig? currently Hudson logins are only +Lets create a new HDFS issue with the same name and link it to this one +BTW thereporter is now a cocoon committer! +maven test fix committedtrunk rbranchx r +I will go ahead and put it in trunk as the issue is fixed with +JBDS java the eap java what are they +I am on vacation from to If you have any question on deployment and JEE bugs please contact Saurabh Arora or my manager Maruthi NuthikattuFor emergency contact me at. +Done +Added patch Adds TI support to Linux x helpers for monitor enter exit +regarding masking issues i would suggest that we dont stick to one particular setup for thedifferent types of benchmarks +Retargeted the issue to since is being released very shortly +I have already added some tests in what is already a start +Attaching second version of this patch first version was lacking the policy file change +Merged onto the master and branches. +Hmm +reference fieldName parenttype cardinality MANYTO mvc setupweb mvc all package +should be tracked in separate jirachanged SLASUMMARY userusername for consistentcydelete bean type from SLAREGISTRATION +any progress on this bug? There seems to have been no discussion as of late +Applied to the trunk. +If any of these need fixing urgently feel free to keep bugging me until theyre fixed +I will attach the error logs +Do we have data about the performance impact of this? Perhaps itd be easier and more performant to deal with the rare and redo the controller! null code path +Thanks +Verified that this issue has been resolved correctly +Delivered in July +Add initialization of database writer to find database meta data +corrected capitalization in the deployment now the extended test runs through on hudson +Is that known? Please advise whether I should log details of differences on this issue or create another +The change always reported the exit value of the script as meaning that it was always successful +It would be easy enough for us to extend the and add that functionality but it also looks like it would be simple to pull it up into a new abstract class +If I dont set it all images will be written as a first type listed on produces +The jar has got a to define some managed beans but it needs facelets to run the contained xhtmls +Forget my above comment +I think when we get the expected outout we set the exit code as otherwise set the exit code as + for which component +For the example above I was not using a community build +Applied the patch into trunk. +Issue was verified closing. +Try also this command line from components site it work as in succeed or does it work as in generating the project reports? I thought the latter would fail with this patch +Thats rather unfortunate +oops bug entered before I finished completing the see the attached Ant file +Bulk update to fix version +patch and +If files appear in the same directory in different overlays then that may be the case +Cargill +Dup of! I thought I had reported it but couldnt find it +Massive comments are not useful being shown in their glory in that list anyways +Reviewed all these issues and marking them as closed. +So instead of using stomp or whatever we use tcpwireFormatstomp this way the the transport that is being used is unambiguous as is the wireFormat +In above you mentioned performance gain +Thanks Dianne +Ive thought about that before long time ago as that is something we dont do now either +I am assigning this to in case we have some spare cycles +The unit tests are all passing on Windows save for one +Im raising it as critical because it has more meaning to me my project than to youI dont see how this npe check will slow down the framework as well I dont see it going bigger because of more npe cheksI dont see how you save my time with fast small sdk code if I had to monkey patch every thing myself doing it as rsl and loading before any stuff this way you save my time my efforts? rightOur project is huge financial business application I cannot provide you with the exact use case how to reproduce it Ive got this trace stack from the end users +Status update Sept Progress On Due Date pre FuseFirst testcase is now running in nightly trunk to project view +Verified on Jenkins. +This occurs with socketTest caseRun in parallel test client test client test client test on trunk and +why wouldnt update semantics be handled on a basis +looks good +Suggested fix attached +Nice you keep momentum on this improvements JimI would suggest to focus on making it easier to plugin your custom tracerAnd then create a new ticket for the filtering of nodes +provided with the new Desktop +It will happen when there is an existing file or NOT +Why cant the s file be run manually? Does it need to be invoked by maven +keeping open until I can confirm is ve updated the html on docs should be available tomorrow. +Also it replaces tab characters with spaces +The DToolsnothing you are seeing is all that is left of the error message which should look more likeException thrown DTools you for responding on this easier debugging it would also be very useful to have the stacktrace printed to the console as well +Perhaps threading subsystem +Ive tried to use anything from to the present +tentative look for Ryan for looking into the meantime do you have any pointers on how work around this nowMy guess is I have to prevent Enunciate from compiling attribute doCompilefalseBut Im lost from there +If there is sufficient interest we can revisit this in the future. +Im working on another issue right now so hopefully I can fix both of them at the same time if I can recreate your problem +Can we use the ZK config file +Thanks Chris +When content streams are present in the request requests are always used +There are some Timers went are marked as daemon +Reverted previous commit due to issue with multiple triggered off one Roo introduced for and Test in Git ID bcbccbfceefbef +For example we check the file descriptors on all the hostsThere are other things to check the permissions on conf the existence of the wal directory the existence of the log directory checksum the configuration across the clusterthis should probably be optional +Right I set that up when the code was in this issue reversed now +So the problem seems to be related to the stack that the base driver uses +So if we want async behaviour then we cannot find if the znode exists or notEven if it exists we should take some action on that +There was a problem with and with Guththila all the samples are working +Then comment out everything in allblocks profile in core and in core repeat these two steps a uncomment one dependency in b run mvn clean package jettyrunThis way one could easily find offending part and we could investigate further +adrian does this solve your logging usecase +tested within Equinox environment +Two things currently has no dependencies on MR +pushing this out to as more information and analysis is needed +If you want to specify a zip dependency youll need to explicitly specify the could certainly make the argument that jar zip but one could also make the reverse question about the artifacts If we manage to get the artifacts you mentioned into Nexus would your build work? There is an option in Nexus to disable the artifact validation +And we can readily drop that restriction for file objects in Project Explorer +Therefore issue can not be reproducedThanksAsela +You are probably right but until proved I wanted to move looking at this up in priority as it is a possible regression until we fix itI will look at this one next first see how often I can reproduce and see if changing the timeout helps +should be changed to only give the warning once +For configuring capacity values no need to put +It is most certainly not the most optimal option to include in your model every time but there are the rare occasions where it really makes sense and simplifies development +You can easily just create a property which uses the embedded buildnumber property +Patch looks good think this is the same as +Now with this fix in place the node will not block but throw an exception instead + are designed specifically to prevent using them in the model layer +This would address the issue of changes as well as the issue of memory use for massive numbers of cores +This updates the tool to accomodate changes which have been made to the format of JIRA reports +This problem should now be fixed +However our solution is not yet in compatible +net and mono on windows and same issue of for ALadded more configuration take +It sounds like you have more experience with NIX anonymous pipes +Just updated the summary to be more specific to the issue +The patch should get more testing before it is committed +Ill also keep in mind Dougs thoughts on documenting the forthanks! raw appendnext interface for is intended to get the raw bytes from the file +Committed to trunk. +Thanks for the suggestion but we use the same parent POM structure as the server so the samples will inherit the same build prereqs as the server build they are targeted towards +Patch applied in trunk +Thanks +I spoke to Ben about this just now +They fix the issue but after their fix after typing value becomes +Dont know if I have the perms to do it +File naming convention for schema file is all files we regularly use +Can you please provide a simplified test case that demonstrates the problem? +Suggest be changed to on summary line of issue +Just dropping the jar from the CI server into your install should suffice if you dont have time to rebuild +Each reboot takes the server out for a few minutes +Then Camel will take it from there to remove the exchange from inflight registry and whatnotSomething along the lines of that +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian this issue. +Found that I forgot part of the merge for +Merged +Which tests are you referring? I removed a log message which was causing the find bugs warning +another take at solving the problem +Moving out of +Users will then be required to download it separatelyI will let you know when I hear back from legalThanks P open bugs to P pending further am in touch with HP on this subject +can you please take a look on Balancer issue also +See for further details. +Patch applied. +Note that I chose not to use a for this because I believe the readwrite pattern for this collection is mostly reads with very rare writes. +Committed to trunk +I meant to say +Please get more information on this issue +My mistake. +Works for me +Strangely this war file works as expected in t you find out anything I could try to start a service through a custom sources dont have annotations +The fix for breaks the JSF CSS Hiperlinks functionality +Update stacktrace to be patch to the issue +By adding Named to the Farm bean it will simply cause CDI to expose the bean through EL NOT JSF hence rendering the test incorrect. +Assigning this to Stan since it is related to JSF +Mass closing all issues that have been in resolved state for months or more without any feedback or update +pdf file while the text extracted from the +Also i would really appreciate if someone helps me in fixing the reported issueRegardsBansi +examplestarget +I have updated the patch +Use of the enumeration pattern limit changes like this to a simple refactoring +Would you please checkout latest trunk or grab a new snapshot and test again? +Thanks guys well update the version in Hibernate it be possible to also have a release that targets Java ? The release appears to target Java +On my system applying the patch leads to startup degradation from ms to ms so around ms which is significant +Btw I noticed that in site are defined some references but they seem to be obsolete +We could release without this change in the scripts but with a change in the docs +Assigning to Stefan as he did some enhancements to services to support a broader range of JSR constraints +forgot stress test +Verification of depends on fix spec has removed this API +I have verified the fix addresses the problem in build +The simplest change and what is applicable to is to change the to use a to prevent current errors +hod contrib was removed +bulk move from M to is eclipse behavior for pages that were filtered as part of disabled capabilities +Im going to jump on the pile and ask that a test plan be included for this feature +I think youve pointed out that whatever we use we need to be able to remove from both ends of the list depending on whether we want the or used connection +Am I wrongRight +Alexey please verify +Will be fixed on when dojo is refactored on its own plugin. +Do you have more information to attach to the caseLike graph calls or something similar +That is correct +Just have this JIRA as a reminder for future +Oh yes Im stupid Im sorry for the noise +Ill commit all the way back to +Customer is happy with the workaround and there doesnt seem to be much if anything we can do from the IDE side of things +Do you still want createMetrics method added? it would be better if it can be added +The error message appears to be coming from the JVM and Ive never seen this error before +Closing resolved issues. +Excellent am on vacation until early April but I will review when I am back for sureIf we can do the same for Tomcat then we are in great shape + +Working on other issues involving reuse of streaming it has occurred to me that I can reuse the stream cloning code found in I plan to split the original patch into two parts but first Id like some feedback on a few questions +Moving to target version it to future releases +its not just a add random number gametheres a bug in the first rows added let me fix it +Sorry this bit you +These unsafe codes are commented by +I havent checked yet the CND registration due to lack of time + +See and my longish comment at the bottom of this would be totally compatible with that +That is now changed to AND so it will be selected or informed iy should +Fixing it was already on my list of things to do as I also saw this error in my own application yesterday +I attached a patch with the code and a few tests +added a result section +those schemas are really only used to publish to clients and not by geoserver internals that is sort of by design +OK your patch and suggestion covers all bases so I agree thats better +Use any input to reproduce the problem +Done +Need to run tests again +This proves that the isse is within Hibernate codeThanksKrishna +If you think that part is overkill and just makes it harder to understand which might very well be the case I can easily revert back to the model I had beforeNo objections to trying a bit harder +It is not an ideal way to do this but the only way that JSONP supportsIt is not needed but is also not that hard to support is it +And as there are no competing validation frameworks I know of this would be pointless +Attaching a new patch incorporating your suggestions on testing +Actually a first version has already been done but it needs improvements +I fount this project and I am interested in doing this +They have been excluded to be removed in future +In the next chapter we will discuss how to use the installed AS runtimes in your web projects +Refactor using Eclipse +Verified with beta builds closing. +thanks Ivan +is a patch to correct this test case +I think it may take long to create a new fromThis can be a cause for Anands case +attached is a patch that seems to help for me it doesnt create such long unicode strings in the testIs there some reason why the test would want very long strings +For example the next thing to try with this patch is see what it feels like to replace ivars in with a object +Im expecting that we will just read the pom version to determine how we parse a different model that way a single version of Maven would be able to handle all the previous pom versions +trunk +Backported the code changes and some additional test to with revision changes to with revision changes to with revision for backporting this to +Ill come up with a basic cache API definition and share shortly +Changed fix version of to to have a complete changelog for +looks good to me +you posted while I did +I like the ideas of making static and using +Seems like a good start though +Later on add method an improved version of y called method z a protocol +I just tested it for Oracle +Note that the previous lib and arent needed with this process +I have applied the PR to the master and sure if it should be applied to the have also created +Thanks! +We add new deletes as we encounter them so that candidates in older store files dont shine through if theyve been deleted earlier +OK +I set that ignoring setting values in the past is optional and off by default +The problem comes from the fact that remote repositories are removed from the object if you simply go to the repositories admin page youll see that the remote repositories are not displayed any found where the bug comes from and will fix that +This is what I have in mind as a solution +It appears like the election process is OK +It only happens with Blueprint DSL though I was unable to reproduce in Spring DSL. +has been marked as a duplicate of this bug +Workfixes has been done in the Query framework +Ira Is there any compiled test for this bug +I think I agree and Im more inclined to let the behaviour of endsWithstartsWith dictate how this is handled +Cool +The initial sentence should be to be +Ill get it checked out on Resin by tomorrow. +If it finds one then it decorates it otherwise it skips that property +See diff to the trunk withCommitted revision Commited to the branch withCommitted revision +I think this is fine then but Im going to poke some people about additional feedback real quick +it asked to send unavailabe message to user +The old static variable caching was cached for all transactions in the same JVM which has implications in a server environmentAs an instance field it is safeI have reviewed and approve the patch +added stylelayout and CSS Demo Portlet +This issue should be assigned again by +Thanks Brock! +Thanks Kan +No available time has been had to investigate any attempts to run domain mode +The problem is caused by line in where it determines whether to prepend the base path by checking to see if the name starts with which is defined to be +With the build works again +any updates on this JIRA? the code concerned here causes a more serious issue in so Id like to see a unified solution for boththanks attached w that approach +btw +looks good to me +jcrspi is a bundle now in revision +Dave QE think thats ok as a fix +Committed +I think youve fixed this w your recent HCM change +Latest patch is misnamed +Ill take a look through the patches and give them a go +If you want to put your stamp on it again go for it +I was a little enthusiastic purging httpclient turns out its needed by stargate so I put it back +am not clear on what is going on hereDinesh can you run the tests? Or are you still having problems with the fileCan you run an individual test using parameter + + Committed thanks +After that the original stream is added +I ran the Java unit tests and they all passed. +Theres a simple fix though +Thanks Kim +If resolved before is released please assign to +I plan on making that change and also plan on changing the stack stored in the to be immutable +The demo consists of processes A Derby server which hosts the demo data +Closing the issueI do not plan to port this to. +Sending Transmitting file dataCommitted revision . +This is one of the reason I suggest to quickly push the release +Need to commit to Branchx and trunk. +I can reproduce this in geronimo and snapshot but not in latest buildIll keep looking into it +I would say go ahead and commit the latest patch to trunk and we can continue to work on validation tests +OK +What kind of conditions lead to this problem and could we address those +It seems to be OK except when you use DESC keywordUnit tests attached +Thanks +M Chen can you try +There are still a fair number of in there but theyre mostly optional or performance enhancement ideasA very simple benchmark says that it takes about times as long to serialize a struct in the compact protocol as in the binary protocol +Metrics do make sense there +And cannot be started me it seems that the JDT patch was not uninstalled +I think if can not get blocks ids sortting the output according to is pretty much the same +Thanks +Just need to decide which one to some consideration Im going to suggest that we get rid of the server name uniquness setup +It can be argued that should have more information +diffs applied thanks for the updates. +This patch copy all the required libs configuration files to the AXISCHOME +Created upon release of +I only made it possible to change the timer period so I could do some testing on the command line +Bulk close of old resolved issues. +Ive checked the svn logs again but I seem to not have committed the second patch +This is the error element from the file +Once the problem manifests all those TIMEWAIT connections clear +I agree that providing an iterator based interface is a good thing to do in this caseI quickly glanced over the patch and it looks good to me +Ill try to verify this todayTom +key issued resolved in The for attribute cant be derived from the key as the key provides the labels id rather than the target elements id +Ok this one is based onIt also adds a few minor fixes adjustments to change so it applies cleanly after up old if necessary. +Patch committed +hadoop fsck getblockinfo blktmpnobodyroot reason why this is part of fsck +comitted by pranav to master bddffedcadcbd +Desktop launcher created on Linux has wrong version either in Label and in Type +If the problem is permanent it will eventually be resolved by human intervention and the backoff mechanism will ensure that it does not tax the system too much +See also. +Jon can you quickly change this so that we can get this in? Thanks +Checked into SVN head is the code that no longer generates the GWT callbacks that unwrap the parameter type in favor of just exposing a parameterized parameter to the Async service +Let me know if this still doesnt work +I dont check for it for the hints written from during write timeout +It was indeed a duplicate of or at least it is now resolved +I cant influence ifwhen it goes into Android Im afraidSend more issues +was the one missing the implementation of these new methods +an old feature request for regionarray +However when a session is established we dont have access to the session id through +and +However I have to admit that I did not test it very deeply +Matthias do you will want this space +Jochen that is exactly one of the scenarios I want to cover once we move beyond Negotiation as I mentioned in the other issue the current architecture of the authentication valves means that the SPNEGO valve can only extend one existing mechanism for all other mechanisms it means a duplication within the SPNEGO valve +It looks more like a networktransport issue at first glance +But I do think this my point that this should be an interface with a defaultreference implementation of but which others are free to implement themselves +Updated based on comments and also added a little more information +And ant test target? +Is that a work +If the state of the job is comleted or exiting we deallocate else we just print an error as before +Feel free to add more nodes for testing +Ive committed the patch +Thanks Liang for the patch +Work on Nicholass review comments +Closing this as Wont Fix +Yes its a documentation issue +you LeonardoI created the issue and referendeced this one. +Could you provide a build with such clean up of the hash mapThanks a lot for your efforts +These checks have been moved into class that encapsulates behaviors depended on particular libraries +so you have to run maven in the shell? as a work aroundstartupCmd on full package has its own running in same command line window after using +Resolving issue +Added missing to accounting for accountinginvoiceitems CreateEdit Person screen you added firstNamelastName fields labeled as required +The patch includes all changes of the previous one except changes to +Another workaround would be to have an internal Maven repo that has those dependencies stored in it +Ok Ill add just removes updates jdocs and updates to latest trunk changes +Patch for guess this is fixed +Because you have no substantial evidence that removing the method call will impact performance in a noticeable way it shouldnt be done as you are increasing the complexity of the code and reducing maintainability for a benefit that you dont know is thereIn other words this is premature optimization +Mass transition all resolved issue that did not see any further comment in the last month to closed in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +With regard to the last comment I have seen bugs posted about getLength failing for with large files for file locking and the explanation was that being exceeded +Some one wants to take this up +Thanks +This is a different trie than the one used for visibility parsing +I know you said top right but I look and its just not there +Thanks Eugene +Reminder to self try creating an interface with lots of ints and doubles and implementing it + also address the thread safe issue of theCan you verify it with latest released Camel +Its stayed up for a few days so Ill close this issue. +test fails because this JIRA depends patch in +passes mvn verify etc +You may need to clean out the data dir after doing this +All the core and contrib tests passed on my machine +Is that right? What happens when multiple exceptions in the stack claim to be right +And if you are still able to please attach you with the exception here because it will have more info on the error that should help +Map Collection and Enum support is fully implemented on Axis trunk now targeting release +Yeah I think this will be the right thing to do +More folks will see your question if you put it there +As far as I know is you provide a reference explaining what callsite caching isThank youMichael +Yes evenif we have YARN problems that I cant figure out after some looking I knowwho to ask among the YARN folks now so we will track down all the problemsfor sureThe other issue is that the main Hadoop stack guy here at Etsy is quittingto work at Apple next week so instead of a month to learn about thesystem I have a week to take over his job! So I will be super busy for thenext few weeks just getting myself up to speed in case their workflowsstart crashing or other plumbing starts to breakGood news they are very curious about Giraph Tajo Tez and Stinger +Ive now checked in changes to the derby script too +Attached files should reproduce +yes +It does seem that the old behaviour is still better and requires less user interaction though +The query has Employee hardcoded yet should be a name +that does seem to be the best option for windows servers at this point +I have committed the initial features which we can invoke a service with embedded mode by invoking from +btw +Not yet fixing this turned out to be trickier than expected but Im on it +Thanks for the clarificationSet to closed +So I think the problem is that the WB deployer is putting itself on the classpath in such a way that the sees you checkdebug why this is so +Almost working something like projectInfo plugin was not found or some such +was it really downloaded by a fair number of users and was it used in an obvious way on the Cloudera website for a notable period of timeThe fundamental question from the trademarks point of view is the likely association in the minds of normal users and consumers of our software not those of us already familiar with HadoopThat being said a logo also needs to fit with the community so if its causing this much controversy then from the community health point of view +As is there any change in load on the cluster compared to when it was running +IMAP server log using have removed the custom authenticator but still unable to read the mail +Actually it would be great to just fix all of the findbugs errors instead of having either this count or the exclude files +this has been fixed in head +Runnable jar file and main java files +Postponing till after +And restarted tomcat and sonar attempted to create the database and perform the necessary is the start of the exception trace An error has occurred all later migrations canceledTable rulescategories does not existfrom optlocalsharejavatomcatwebappssonargemsgemslibactiverecordin attributesfromcolumndefinitionfrom optlocalsharejavatomcatwebappssonargemsgemslibactiverecordlockingin attributesfromcolumndefinitionwithlockfrom optlocalsharejavatomcatwebappssonargemsgemslibactiverecordin initializefrom optlocalsharejavatomcatwebappssonargemsgemslibactiverecordin createfrom optlocalsharejavatomcatwebappssonarconfig +cannot be used with source +Picking something reasonable like should be fine +Feel free to commit to as well +From IRC bbrowning I think as a first pass try getting things working with an unmodified kickstart file bbrowning so choose sane defaults take the name of the appliance from the name of the kickstart file bbrowning CPU version bbrowning maybe? that way someone can give a plain kickstart file and get a working image +The unit test has been added +Sample symm binding policy symm binding policy class which test the assertion implantation and the relevant policy files +What was happening was that read was being called both in the partialDeserialize method and the read method in some edge cases +committedisnt the file location anywhere on the classpath? if not then IMO it should be +So first impression we have to warn user on java selection page if selected java is not a would the message sayYou must select Java version to work with EAP +At some point in the past some of the asserts were commented out to aid in debugging and I never them +Setting to future pending a patch +Thanks Owen! +This is not different than but we were using less threads back then +This was caused due to memory space issue in the filer where the data dir was present. +rvmrubieslibrubynetin start +Please reopen if the problem is seen again. +bq +Hi SalihThats all the info I need thanks +Lets change it from warning level to debug level to prevent warning in online cluster +Which might be out of scope for but I dont expect that adding the tracking will be very invasive +I fixed this in +Also while you are at it it would really help if you implmented a way to do exclude lists specific to xsltc and a given flavor +Any chance we could get this in +That makes life easier in any case. +Resolving wont fix since the diffs are expected +thanks for reporting back! +Max this was an issue with the previous stable tag and should be resolved now we updated to the latest tag +CR Correct jBPM version in place +Lets fix that +There has been no feedback from the reporter and I believe I have answered the question. +Patch in trunk +Anders why? Im usually not too much bureaucratic but IMHO looks more a new feature +But if the batchlog write is successful it should return TOE not UEUE means I stopped because I knew the replicas were down before I wrote anything +Do you mind trying with ? I think I just need to get a version out +However we do not encourage anyone to use field names starting with capital letters as it is not a generally accepted Java programming practice and may cause issues with Roos use of reflection to determine accessors and mutators of javabean properties. +AS versions +I reviewed just the comments not sure if this is ok for backport +Closing this due to lack of interest from community. +Im going to do that +Second your decision Sam +This only happens for IE and its compatibility mode +Patch committed. +Fixing this bug by working around rather than providing a complete fix for now +Well upgrade Sqoop when we do it for all Hadoop in the future +When changing the locale to French the s portlet title disappears immediately and the displays as the title instead of using the default title +I just hit this again while compilingIs there one of the strategies in the comments that you prefere I use in the patch? +GregI think the lock paths here are unlocked in finally block +afaics adds a notion of esb which does not exist yet and in my suggestion mihgt be called instead +However Im not sure whether that query was the first one run or not so I dont know the status of the searcher etc +Though at Berkeley we were using a shared mysql not installed specifically for Chukwa +Agreed +now contains your suggested fixes on +OK so it seems that the comments on this were all about and which was never upgraded +As part of the generation the jaxws tooling attempts to create a classpath based on the location of specific classes but it does not understand the vfszip protocol used within AS +Exactly when will the method names be consulted? On each call or only when a optimizedFoo is being called instead of the regular foo +adds new functions to that call endFunction with the old signature +Closing. +Previous patch contained a dummy error +OK I have done an svn up and redeployed +It gets complex with inheritance +Thequestion are who maintains Excalibur event and do we want a dependency ofCocoon core on another third party software +In this case cp dependency appears twice jdbc cachestore module depends on it in compile scope and in test scope +We are going to upgrade our H version to the latest one to see if the bugfix solves the problem +We have addressed all your concerns in the wiki and have very actively responded to all the commentsWe will revert the patch and make it patch available for nowWe need it soon so please try to review asap +So which guarantee are we losing +Weve had this at least since. +thanks Zheng +It depends on the purposenature of the test in question +Closing resolved issues. +fixed in revision Updated Minotaur +not making beta and so probably shouldnt make unless max really declares it needs to move to +orgmortbayutil matchesBinary file +bq +Verified on +Option Separate the abstraction to Multitenant and Singletenant classes +If we state in our documentation that we arent unless you specify volatile then we dont have a need for the to be volatile right? Or are you saying that you believe because GC might be in a separate thread we need volatile even for user code? +Please dont consider performance issue before you can make it correctThanks. +Rebooting clears everything for the first test run to lock up and become zombied +when you perform a WFS request the resulting xml has schemaLocation set to an url with port or may be you just need to configure proxy so that knows its being routed through port +the patch into trunk and +Thanks for committing this SureshI filed to fix the test. +I was tempted to try and run our Confluence under Jetty but I keep wondering why we are running Jetty at all +Resolving this. +Or whateverWould be great to get done for but I dont think its necessary +IMO should have test for this not breaking in trunkWill try to find a way to test this case +Ive just committed this +you might track the percentage of times that each had been allocated from and allocate from the node whose percentage is farthest from its desired probability +Thats fine thats exactly what I am suggesting +Fixed in revision fix break some of the functionality +unless ? arches iiii else arches x end +Attaching the patch which adds the group to the cache during initializationrefresh +In any case draft patch has reliably failed x in a row and a la tcpdump confirms the lost messages +This would allow Tika to avoid casting the instances it uses down to and would potentially enable other archive formats to expose similar information as we now do with zip. +Any chance of getting this into so I dont have to patch it +Would you mind uploading your patch as a file and and the same time uploading it to Review board? attach the final patch that has passed the reviewThank you guys! +Theres no need to mark as well +When replication factor is it is impossible to recover data so we shouldnt wait for confirmation +Im creating summary of proposal in and +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content Services +Please update and the patch +Tried this out and it works beautifully +QA Closing. +Fixed at Committed revision Bug with comparing float types tracked in. +anyone see an issue with backporting this fix to the branches +patch looks goodI have committed this +Its the same problem +From the stack trace it looks like Trejkaz was getting an exception in the call to +in order to get a grasp of the issue easily and also to resolve similar issues lately +Feel free to help with examples for GET and for the examples I have added them to the wiki all resolved tickets from or older +settings directories in each project folder +So if I use same database for both datasources I can see exceptions in server log +Thanks. + +I reproduce the exact same issue same stack trace using Infinispan Hibernate Atomikos Hibernate Search a XA DS +Thanks Milind! +I tried that project out and I didnt get the result I I run with the torqueboxstore I get the behavior you describe +On the actual patch by Junjie the table is being dropped at the end of each fixture but Im not sure about thisOn the original test it seems to me that the table is created and populated once and then it is queried on several tests having only the function procedure and the table dropped at the end of the whole setCant this influence the outcome and the purpose of this test? And maybe even make it slower unnecessarilyFinally I will make it so that we also look at the actual data since that is actually done on the original test +Thanks +rd times a charm +The patch is against top of +Will update the docs to reflect +I have attached the corrcted wsdl here with pls have a look at that. +Now site is. +probably need a custom order based multicast UI like the table based CBR some sample files that work fine now in +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Since has nothing specific to an rpc engine it makes sense to move that into a separate class which any rpc engine implementation can use +Commit changes to trunk at rev any other concern please drop a comment +Test added +StackMy patch is for branch while Teds for trunk +Id appreciate a little guidance where its up to the features I shall check for +Zhao Im happy to later versions when I get a chance to +If you read my comment its what I explained +Because there is no failed request but only empty response +but we must do upgrade on the active NN +Thanks +Fixes committed +Committed +Once moved everything worked +Nicolas You referring to this method add? Should I open issue to fix +Moved to Tools since there isnt any JBDS specific issue here it is a matter of the runtime bundling something different than what jbpm plugin marked it as a fix that needs to be available in since this sounds like a serious bugregression +We have another test case which fails even the timeout is set using +using https is unjustified for LOGCXX and HYPLUGIN +Assign me the former and someone else will have to look at the HQ issue because I havent tried that before +dup as was +I believe this issue can be closed now correct? +How about the branch +could somebody more powerful throw away all the other source files +I have had good success decompiling the generated code by using the free Jode decompiler +Committed revision +I am using a static method in a supporter class for the DAO which fetches the queires using message source +Please integrate it as soon as possible itd be much appreciated +Ill probably be able to hack that out by tomorrow. +Please your application against the latest SVN snapshot and confirm the fixOleg +Closed. +Here is the patch both for the URI implemenation and + +There seems to be something wrong with the fabric plugins +patch works +Added code to handle text prefixes based on code supplied +Verified in ER. +Having a java client that reads from an old server and writes to a new server does not seem like a crazy thing to supportThere needs to be a good reason to drop this basic functionality +shows a timestamp of +I see a ptest trunk build ran successfully after this so I kicked it off again +Why is this marked Resolved +Id like to avoid writing two copies of the base dataCassandra only needs to store the row UID as a Lucene document +I also did something similar for the same problem in but I found many ways to break it +Here also if the issue is with namespace like then I could definitely rename it to something like +And maybe change the default period to second instead of or even? On small clusters probably not a prob +No way to read the type in Java . +Hi Bin +It appears that your site has not been updated since June +Can someone more familiar with the date conversion please have a look at the patch +Could you describe a usecase +Unable to replicate in JBDS +So here is a better version +This is the same logic I have in +It will work fine +Thank you Jing for the backport +seems like the was the only one without a version +This is not probably notgoing to change in the near future unless some sort of disk caching is introduced +implements an index without and has some code in patch to make secondary indexes pluggable +we never setCaching to the Scan object in Maybe a bug now something that hasnt been examined I think since +Lets get it in +Verified. +When you refer to removing the dependency on the Servlet API Im assuming thisis for the purpose of supporting portlets? In that case creating an interfacethat has the common methods would do the trick +fix is also possible +Andrew Given the high load on all our VM hosts we should probably wait until the new host arrives in March +This causes a few test cases in to be brokenHence the attached patch needs to be modified to distinguish between streams that were obtained from a result set and streams that were obtained from a Blob object +applied with minor changes + this issue for Jackrabbit trunk +I need to look at what happened to that +I need to explain to people how to set their confluence username in Xircles +Need new patch for trunk +Thanks for the patch Ioan +Nothing to verify Savara and pisoa are building fine using the JBT parent pom version +This said I dothink that digest scheme user names are not meant to have content asit would constitute a gross violation of the RFC +Expected behavior Error message provided on the page header indicating the name is duplicated with another realm and user can not go ahead to next page until heshe uses another unique realm name +Same as the previous patch with the apache license added +If you need help simply shout at me +Any objections additions +Is it rendering correctly + +Bug has been marked as a duplicate of this bug +As per your resolved status its done to +Hi OliverCould you provide additinal details for the not working custom installation the default base URI the URI of the pdf the URI which you enter in your browserif the latter are identical I dont see why it shouldnt be working +Thanks Nicolas +We need a minimal complete example to debug +Still does not work as expectedAnd actually the bulk deletion of projects does not seem to work as expected also +Committed the fix +See the longer term solution here +Its be also great to have a separate placeholder to list all issues found by tools in Hadoop +Closing in on this one +without this patch the difference is even largerSo this is not about index key issueI think the overhead is that the fake key need to be construct for a seek operation +bulk close of all resolved issues. +The reason the test fails is that the is only seton the environment by the processing pipeline during actual processing whereas at the time is called the pipeline has only been setup andprepared +Fixed in trunk +That is if a float value of +LGTM +an you elaborate here +reassigning to arik +There is a nice GIT extension for Eclipse that makes it really easy to manage branchesCheers Lance +fixedDate Thu May New Revision for in revision . +Would be easier to comment here It does not interact withCan you fix that? I mean if you want to do it in a thats fine +What do you thinkLastly Ive added a lock during the process failover imho we can meet race conditions without this +A typo in the above in paragraph bar should be equivalent to componentbar +I didnt add an argument please add it if you prefer the explicit shutdown +Though we should think of ways in which the UDAF does not need to create a sub class for the evaluator +Approved for Apache Helix. +Id like to see a diff of the pom in the api project versus your generated pom hereThe pom for the tck is a good start +A patch against trunk which should fix all the problems mentioned in this issue +no we cant for security reasons +Committed to trunk +Committed to CP branchCommitted revision +I agree i think its nice to see a patch to lucene that includes any changes to the backwards testsMike did this with and i was shocked untilI found out he was doing it manually with cat +to trunk +I also checked all existing tag libs for the pattern that causes the problem iterating down through ones Scripts checking their type with instanceof +Thanks Brandon! +can be fixed by moving the elements below the elements +It just copies the current JPA connection profile instead of picking up the current JPA projects connection on initialization +to be clear jboss supports the b flag but and do not +Fixed +Most of us spend their spare time for you to keep Wicket up to date +Also it will be good to file a separate JIRA for Swapnils suggestion for the patches committed v to +thanks +Refresh the frame to see the changes. +Attached patch +This is an issue with exception handling in jdk +The client was closing FORWARDONLY result sets after the last row +The only thing added to by is the member fontInfo and the corresponding fontSetup method +Thatd be great +Juergen +I have made an update to change how the priorities are compared on queue entries in order to avoid retrieving the message metadata and thus avoid trying to access it after it may have been removed from the storeI have attached the patch I expect it should apply to the source as well if you want to do that +Thanks Feng +Also we now have an R initiative well underway in the Incubator which is a better home for this patch +krb under home directoryThis is configurable and each install might have different setup +I strongly wish this to happen +oops +The stack trace should get logged to a file +Otherwise I would have no idea how to encode coordinates with z andor m values +Yes using an annotation for Java and above could be an option +But theres no test for it so Im blameless if I broke added that behavior will have to fix it themselves and add a test +ohI see +I fixed the indentation and logging in the attached patch +Its no problem Adrian +Look at +Thanks! +At any rate I think the ODE community should have this code +I chose to make the amount of time waited for a task and then for a task to be the current heartbeat interval each +I think this one is fixed in the master now +Both the original code logic and the proposed patch has a we always use the same URL for bundle installation +Moved to next milestone release. +In some cases probablly not all threads are stopped before check that caused error in test +Thanks created +I agree there is room for improvement in the javawsdl mappingmatching +Ok I think I figured this out +All artifacts are on stage + +Grant license + +lyum install rpm qagrep javals l usrlibjvmjavabin +Updated patch patch d like to confirm my understanding of this patch and propose a possible useful extension that may be better off in its own ticketConfirm this is purely a notification for the caller +The following is the code that I ranant test Dtestcase trunk is broken +An upstream issue +reserved in the target +Pmuir To add to what Sande has asked The throws an when creating an Asynch entry + patch throw if constructor is called with old syntax +OK Ill backport the fix to which will be released way before +thanks +Given that its been broken since and is undocumented I dont think we need to worry about breaking plugins with changes hereIdeas Inline logic into Media plugin Make a separate plugin called +typo in issue title +gz files +Fixed in commit ceafdfbeffdfedbafFor future reference it broke due to the move to from apache http +I think the current behaviour is acceptable for our needs +This patch is for trunk +graphValidator put all validation messages into global scope because it can be invalid values for a bean in the this example grephValidator used different instance +What you think? Meantime Im testing +stores the secret and the in its +Hope that helps +Heres a sample timings that have been experienced using groups second groups seconds groups seconds groups seconds groups removed a lot of codes that I think are not needed + published to Maven Central +Updated patch +My initial thought is to encapsulate the priority indices inside which is where the priority queue and indices live anywayWe could also guarantee the appropriate index is decremented properly on each call to removeI do not think we can know in most cases whether a particular block lies to the left or right of the index since the random of blocks is implemented as a hash whereas the index is an index into a list +The change looks good and I ran tests with the IBM JDK and did not see any problems. +Would add a log because else someone will think the call is not happening +I noticed this also yesterday +Is it really worth the hassleHaving the around in an might not necessarily introduce a memory leak and the object itself is not that big two array lists and one reference. +reassigned for was just working with a client who had this cause an issue in so Im reopening to have it fixed in the branch as well +Maybe a text configuration of how each element was written out +attached the same to trunk thanks again Sho +bq +For making this functionality I made two new methods is and is in Thanks again is the new patch for the same functionality but here I have used groovy script instead of bsh script +Vincent Just make sure that you dont make this plugin too Sun specific +Do you want to run any tests Pavel +Preliminary fix pushed in +Fix in original contributed DITA documentation +Thanks. +Trying for the upcoming can you make a trunk patch +This issue will be verified on the ER build +Detaching this issue from +OK so you want one line that reflects the balance rather than a deviation from the average for eachHow do you measure the balance? Do you have any particular measurement in mind +Looks good to me +This situation happened to someone on the mailing list +no it was a miss diagnostic +I can find no record of a dtubello user in Confluence +I use this pretty frequently and from recent list xfic it seems others do as well. +No problem Ill mark this issue as invalid during constructor resolution the factory may try to resolve several constructors +Thanks to everyone for sticking with this spec process and making this spec better +shazron hahaha i thought you are talking to me because late night coding is common in here Thanks a lot for this fixing this issue you are the savior of many of us i am pretty sureAwaiting for your zip file plugin +JAXP describes the value argument of as follows value The value object +I just saw that the old html attributes are not supported in strict xhtml +I have checked the modified files according to the lists on svn commits +will look into it +Ill check the model generated is working correctly asap. +So that Ive to commit a new version of the including the needed patch +The stack trace is the same in other cases too +When select Customer input type wizard dialog will show all input type +If I return yes as on a drag and drop would seem reasonable then any other place that API is called Id be turning that object into something API is also called when on any resource as the creation of the context menu calls the API to see if the object is already deployable and can be run on server +Id like to see this fixed as well since one reason Ive enabled map output compression is to reduce disk space usage by the mapreduce framework +Assuming that going to disk is slow the extra check of cache should be inconsequential but this made a big difference in throughput of a workload +It cause timeout failure in iteration +Is there a way to add the files in the new repo while preserving the history +I mean i downloaded the Apacheds zip file extract it and start the +Im not too familiar with the use cases around this but the bigger issue is this isnt ready from community perspective +We serialize the RDN not the DN +I got only connection +HiYeah I dont see any reason to change that until we want to extend the Log class at which point it will be quite easy to change +Patch applied on trunk and closing all issues that have been in resolved state for more than a month without further comments +Scheduling for ER +As you are a PSC member all it should take is your +The userconfig file probably contained a setting telling FOP to producea configuration dump and then exit +Do we apply the patch or solve as wont fix +I have added Derby test as a performance test case and I have run this only using Junit directly +I planned to do something with those return codes but i wasnt sure how to proceed with it +Committed as r +I dont see the interrupted status is restored in other placesIf we do restore the interrupted status the thread will die +We have the exact same issue in our project +I will put together a solution to hopefully make it and allow stream saving to be disabled +I think this patch is ready to commit +added message sender we are using to reproduce the issue with a client app +Is it OK to use it? Or should I wait for a new one +bulk close of all resolved issues as part of closing items. +Implemented new topology view which layouts the server groups hosts and server instances +Committed this to trunk and +I was expecting encoding from this utiliny +Hum the spec also says the typename element is optional and this is not handled in our it seems to me on trunk parsing still generates multiple filters instead of just one +Will take a look and put the patch back up +was found while testing the changes to this bug +This seems to be a symptom of the same problem that the locks filesIll note that I dont see this behavior on Linux or Mac so I believe its by changes to +Then I had to change a bit the test to change the for the body to be reset +Please review it We will need to update the document and mention that SEI must be an interface +The parsing would be ambiguous yes +But otherwise I agree with your comments +it was correct because session was that I was not able to recover my lost probably your error was caused by session expiration +Will update the patch +Its difficult for downstream projects to build against a branch or an uncommitted patchUmm? We could just write up the patches and test them either or via a and then commit +The Welcome page contributions are still to be done and will be written up in a new Jira +Im not sure if we already have an issue for this so lets keep this one open +Is it appropriate for me to modify the behavior of the render method within or do I have to extend and make my changes in a subclass +Possibly related to barrier issues addressed in have to wait and see +It is much cleaner and the tests pass for JDK and JDK +Attaching a stat file which shows all the various jars that are added changed removed +It works. +Ok I applied workaround and tests patches at +Yes we do need to patch to support deep message iteration in order to do index verification +Committed as cdebbcadacaeaae +last changes fixed the problem! the system module was in conflictthank you again +Best regards +this patch file has unix line endings +What I tried to explain was that you could just allocate a new to decode into and then pass that to the Appendable instance +I will try again +Ok no problem +The attached patch fixes these two patterns so with this fix you should no longer see these errors if the node no longer exists it should never appear in the results unit and integration tests pass with these changes +I think these problems are fixed now +For and this is OK since there is a correspondence between these kinds of statements on the client and on the server +is there any update on this issue? We were able to get this resolved by not using encryption +These are configuration related warnings that are resolved by properly configuring the broker for your environment later versions of AMQ will adjust down to the available space as needed on start +Change committed +After some consideration Ive fixed this through a different kind of check along the lines that you suggested +I we take a look on the code then we see that s code use reload in redeploy method thats incorrect since +There has been talk of using an to abstract the details of each issue tracker +Note the fixes for this issue will be present in the upcoming release of Grails and the Grails Maven Plugin Archetype +I am thinking that giving that we dont care if this is solved a few milliseconds later one way to handle the backward compatibility issues would be this FollowerObserver sends an increased protocol number in FOLLOWERINFOOBSERVERINFO +Deploy the war and point to httplocalhost patch was incomplete due to a missing folder +Thanks for the patch! +Closing this issue +and then report it to it make a difference to call set first and setEvent afterwards? Or does that show the exact same behaviorJuergen +I was thinking of a cache whose scope was the length of the request +I applied patch for on trunk revision +Read subject as keeps refreshing shared Connection even if container is not running +spec +I tried grepping around for but didnt find that somehow +No activity closing. +The problem has never been reported with other browsers. +Hey Jiri having downloaded and installed the latest JBT nightly builds I am still unable to see the Exporters tab is this something that has to be turned on or only appears under certain circumstancesThanks JBDS has now gone out and the book was QEed for that is this JIRA now Out of date +No damage was caused just some of my queries stopped working +Done +Can you validate on your side +This may not be a forum but the issue continues to exist + is the QA test results which show which versions are currently not following the standard +Guess I should check trunk as well tomorrow I should get time to do that as this is becoming a pretty big problem for us +Reproduce the error in beta +Doesnt fit the framework well so no great interest in implementing this. +Can you try to figure that outIt might also help if you show how your serialization changed I did not see a commit referenced +Changed the resolution to Wont fix it is more appropriate +Attach patch to address Richards comments and hudson feedback +for the patch +This is fixed by putting the close and the into the finally block +OlegI definitely agree that some major refactoring is needed +Using we still see the last row of a table overlapping the footer +This would help ensure that the DRDA changes made as part of are correct +Yeah same +Do you still have this problem with a more recent version of the site plugin +Ultimately Id like to contain the js files for all the javascript libraries it supports +Will apply to all three projects instead of creating jiras as all files are identical +Will it be possible for you to this patch with changes to lines removed? thanks +I cant really remember! I think it was Archiva and I followed the instructions on the Archiva site for seem to remember that the property name was also appearing in the Archiva page for the artifact too rather than the value of the property +The latest patch addresses the concern raised by Alan +Verified at r +Youll need to go down the component tree and compare ids one by one to find to the correct component +I did only wait for your you use the provider then it does use the Groovy the runtime which uses the normal Groovy stub gen. +All we can do about it is add a simple check to prevent it happening. +The list box contains only the repositories the current user have cannot I wonder how you can have in the list repositories you dont have access. +Does this fix the problem seen in +bullet is doable and then the simple language in general will be improved with this featurebullet is required for so I doubt its easy to abstract a remote FTP filebullet as bullet the file language needs access to so you can use all the file methods such as getName get getPath etc +Sounds great +Its a pretty tough problem to work through though +fixed +Im still confused as to how fits into all of this +Simple archetype using addtional properties propA and as quickstart archetype except this will used the values of propA and propB in the element of the created pom +Recursive component properties inheritance yesno +fixed the view calculation the happy case now seems to workit still needs some cleaning up for the pathReviews more than welcome now +Havent seen the issue in several gridmix runs with the patchAlso went through other objects used by appears to be the only one which would run into such a race +Closed after release. +Committed branch and is a result of changes in Result from delayed deserialization +For example a Purchase order has a header record and multiple line records +Specific issues will be tracked separately. +Attaching patch a which makes JQL the default querying mode for obtaining the list of issues from JIRAAlso removes some unnecessary code dealing with the reportDisqualifications propertyI tested the ant target manuallyPatch ready for review +Sorry this held up the QP refactor +Some specific questions +Changed blocks to is depended upon since the roles part of the issueis fixed and thus the issue no longer blocks even if not fully resolved +Thanks James! +Attaching minor logging cleanup on previous patch +sounds good Erin +Addressed all the issues ppointed out +The connections are automatically closed by the connection pool +Many thanks BorisBruno +concerns would suggest the default value for this attribute would be something like all +It simply tracks emitted artifacts that are in the list of reactor artifacts and avoids emitting duplicate entries for them +the doc for or also say no word about not contains or not contain +When I sync up new codes and rebase some tests break +The is code from the SVNAll my changes are in a seperate applicationI dont make changes to the application base or framework except for those to cofigure ofbizstarthttpslocalhostpartymgrcontrolget a list of with showingat the bottom is the next buttonclick on next get the errorI have tracked it down to listlist is not in the code anywhere +AaronPlease apply this patch too +URL and it worked fine is this URL problem Windows specific +Good to know. +arent necessary and I dont intend on including them +The description of the patch is in the comment for the file Ive ThorstenI think this is much more convenient am looking forward to its releasecheersg +Hi I didnt want to hurt youCan you help me to improve I didnt get the point a Is it ugly to mix uppers and lower cases? b Is it wrong to implement a bidirectional in this way? c are you talking about the missing inverse? d Was the issue not important enought? e have I missed an issue dealing the samethx in advance Michael +How ironic this issue is created nearly on St +Is the new release of hibernate filter? But I cant find class under srcgroovy folder +Unscheduling due to inactivity +The create check logic was moved into so what I made the check in for Im not entirely sure +In order to backport it and the related bug fixes we have the following JIRA sequence Should DFS outputstreams close wait forever? Terminate when all writing files are closed +Should I fix the nits and create another patchBtw I ran mvn javadocjavadoc and could not find any javadoc warnings caused directly by the patchPlease let me know if I am missing somethingCheers to and trunk +Thanks though. +If you can still reproduce it please post more details on forum +Patch applied. +serviceDnssupportedservicesVpntraffictypeGUESTserviceproviderlistBserviceproviderlistBserviceproviderlistBnameVPCNetworkguestiptypeIsolatedserviceproviderlistBserviceproviderlistBVpnserviceproviderlistBLbservicecapabilitylistBperaccountuseVpconserviceproviderlistBDhcpcommandcreatesignaturepD with blockers and critical issues are resolved but not verified +Is that what you wanted me to do +Thats a good start +Edward great to hear + +See +Is there a way to get both caches to persist before upgrading with the fix +Im removing besides revamping there is the docsexamples which is referenced by examples +I am not using a proxy +Good to know +Changed to HSQL and MVCC model and all is well. +Ill take care of this issue when the new JSF impl is released +Can we do this for instead +Regarding JSON content type this may be related to which makes the configurable but defaults to something that will display in a browser by default +We update the slides and the scripts we used +Here is the patch which implements the improvement +Ill have this fixed +will take min in the best case and min in the worst case assuming default config + +Future issuespatches will address individual componentsfeatures that requires changes to work on the Browser +As is +The fix for this bug was late for. +Cassandra will be upgrading to thrift at which point we should work with Hive without conflicts. +It looks like a JDK solution is to delete a file and then recreate it +The test passes now and others too +Thinking about this a little more Im not sure if adding the license headers is practical +With an upgraded package this patch shouldnt be required +now it is in branch and will be there for in JBDS +r ldimaggi ldimaggi Mar +One extra modification is that the property of max should be set as for tomcat connector +LGTMIf passes I think we can commit to the meeting in Linkedin we decided to do Tez work on a branch and that Cheolsoo will initiate conversation thread on mailing list for it and take up the task of creating the branch +Changed Target Version to upon release of +Reopen to upgrade to ll to commit to Branchx and trunk the upgrade to SP. +If you disagree please comment. +Does it means that we should release final now +Where is that bpel cheatsheet reference coming fromthe bpel examples should only be added to project examples if you have installed the bpel plugins right +seems issue was fixed in SVN quite a while ago +Spreadsheet as HTML not XHTML +Is this change also required for trunk +If I add the upgrade tests back andrun the whole suite I get the encryption errors +I found out that this is a more general problem not only related to JMX connection created another JIRA here move to +Rock! But +Now that made it to HDFS let me commit this patch +You can close this ticket if you want its gonna take us too much time +Thanks Laura +It appeared from inspection of the source that is was only or patches off without any jboss specific customisations this made upgrading easier! Integration to the stream would be great +Add Japanese translation added and testedr +If I need to dump any big lumps of data I cab go via CSV or bump the heap right up for the job +These Generic ifaceclasses will be of the current Remote +inputs are now properly set +Reopen the issue if you add that info +Alright I guess Ill be adding some migration code to the master to handle this changeover then as well +I have attached my preliminary patch for this issue +Closing all jars now start with logj apitarget coretarget coretarget target target target target target webtarget +nicely done +The has all of the jar code refactored and tested +Running tests in the background +I dont really think this is a good fix +If necessary a private inner class or anonymous class can extend the class providing a protected method so that it isnt visible to most users but can still be made available as needed in other classesAutomatic warnings are great but not exposing these things in the classes weve identified as public API in the first place is even better +I think the statement compilation should remain strongly typed however and not allow to compare objects of any type +Perhaps the recent upgrade in jars fixed this issue +Alex could you quickly tell me what was the reason for this change? Im not against the change but would like to understand why it was doneI reopened the issue because arguably the class now has the wrong name as it is no longer bound to a session +I dont understand why you change the repository of the artifact +Hmm the dup vfs instances have nothing to do with classloadingIts just that were not able to link previous same jar with the new oneSo I doubt that is actually what solves your problem +xsltHere is the current list sorted alphabetically +Thanks +Any chance of including fix for that N +Could you run test target for locally +There have been problems with using +Nice finding +Hi to install and I am using FC cant see other packages besides these two +Fazer get the work Numa conusulta using with clauses +That is why you moved the to onInit as wellAlso the price one pays for stateful pages seems cheap initially +Please verify +Ups havent seen them +Verified on Tools +Fixed pom Brent for applying this because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The caused by the synchronized block should be gonepatch applied bug fixed +Bumping to so as to remove version +is meant to make the build work on Windows and not to provide feature parity with Windows as it is clearly stated in the name of the ticket Attempt to build Flume on Windows x fails +plugins has many unattractive names +No need for IT +Attached a patch adding broker configuration store and replacing existing xml file configuration with json review changes were made for via +Thanks for the fix! I look forward to trying it outCheersEliot +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Or it could involve registering a more specialized type that provides access to the result +In the interest of finalizing I deleted the rest of the feature +Still discussing +There is additional work needed to split methods more intelligently but at least for JIT mode we did institute a bytecode size cap +Because things will continue to change on both sides and I dont see myself updating patches for you +patched will be released in next into PROD +In my case the logo is a square so it is much better in the menu section than in the top location +It works OK now +The earlier pull request was closed without merging. +If release note is required please provide causeconsequencefixresult +I attached another screenshot of showing that in fact million documents were processed in the end even though it said throughout the entire time. +This patch reenables this test now it is working +User reported that the problem was in their application. +Update to handle to application transform and add a testcase to +It will remain in existence +If you want me to change it to fail the build just give the word +All done and dusted +Are you generating the docs from XML tags on the methodsAlso if combining the aspnet properties into pattern converter is still desired I could look into developing the pattern +Test it manually and it works +It was released with the fix in +Could you please the results that you get here? so this issue can be considered as resolved now I +revision thanx hiram! +See tail of for discussion. +Hi PrashantThank you very much for the patch! I tested it in local and mr mode and it works fine +sesametools? +It is possible that a map task copied a few blocks on the file before it encountered an error +Can you submit your issue to the mailing list prior to opening a bug +Thank you Bharath +describes the very same issue and was created by the same user. +resolved issues. +Current log shows a lot maybe even too much +The missing elements seems to be a separate issue which I have filed separately asWatch that if interested +Ok can this be addressed today and need to reach to conclusion thanks VM template is not downloaded to primary storage pool due to permissions issue mentioned in above commentsIs this resolved now? Is system VM template downloaded successfully after permissions are updated manually or through cloud spec file +We need more QA resources to handle other containersversions of them. +Attached patch will solve the issue +Reupload the patch to trigger jenkins +Thanks Tim +Failure was a timeout in +Thanks Liang Xie and JMS +Adds infrastructure to to be able to start HA +The issues are now closed. +I guess this has been updated in while doing the move violations issues. +If this is something that you find really quite pressing feel free to raise a new issue and Ill work on this as part of theRob +thanks for the patch +The problem is related to the jaxb connector developed by external partyHowever be adviced that you do not need an extra connector for +Yes the main query still has relevance applied +It was present in the source distro +I think this issue is resolved +Attachment has been added with description Ignore the empty class loader +attached a new patch that throws when parent dir exists as a file +I would like to keep the synchronization as it was in the original patchbq +closes and sets this instance to null when the JT is shutdown +patch ve committed this and merged to +Resolved as described in my comment above. +Closing the ticket. +I arrange my Groovy installation to only use the indy artefacts I have not tried with the artefacts +Can anyone recommend a more obvious test of the solution to this than kicking off a process to extract text from the document and capturing? It would be nice to have something that we can generalize to other documents that trigger this issue because of a different set of missing beans +This sounds a lot like which will be in. +It appears to default to false in so you may need to explicitly set schedulerSupporttrue on your broker when you update from +Im closing this out as there isnt anything more we can do here. +Marking a blocker for just committed this. +the archive contains an HTML file and graphics as well as a copy of the CSS file for viewing the document separately from the website +should mean in Groovy +Let me know if there are any problems with the patch. +Adding in the dependency I just filed +Thanks DeepakYour patch is in r r +Committed fix for this jira entry with revision +Hi Jacopo I am working on this issue with Karl +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues not use a Factory that could be changed from the file +Added the channel message convert methods +patch jueljuel dependency into juel and juel juel and juel version to Third Party to juel and juel console with still reports this problemerror parsing exception Provider for cannot be found Provider for cannot be the default service factory finder code to match juel and hid factory finder that BPM console did not report an exception anymore. +About the nd question you mean within as defaults +batch transition resolvedfixed bugs to closedfixed +I think the patch would be simpler and we would avoid breaking the client API if we kept Stat for clients and added the size and of children and then used for the current Stat structure +Prepared for next +Will look into this when I get back from vacation +Updated title to more accurately represent the to will be bug fixes docs only +is about this +The syntax has been changed in the revision which results in our current code generation library in an exceptionAnother reason to switch to the new source code generator created by my collegue Schaal +Model tried no more than jdt already tried +I just tried to simulate the error and I used the derby JAR file for the driver +This ticket and its solution is assigned to Spring M +I committed this to. +algo needs those items to maintain the error boundsRight +In Iinserted code to serialize the form to file then ran the app with the oldStruts jar and the new one containing the patches +Apparently parts of thiscode originates from the Cryptix project +Well get this straightened away in the maintenance release in the next week or so +We might as well remove it or have it computed in the doc automatically +What happens when we call it again later out in finishInitialization? Could we end up processing same server twice at allThanks +It might also make sense to optimize the output with csstidy or something else +sodlls for every environment we come across and the unsafe can be default +thanks! +It can be enabled by default +Something likePN OR EPEPEPEPEPEPEPEP end of block! right sheet border +Moving issue to a next issues to next release +Committed on trunk will wait for release for on as well +FYI we have the tests where the filters pass the properties around though they are not persisted into HTTP request +it overrides any setting +It was not yet completely finalized and more important it is dedicated for the release which isnt yet out +Removed and all its subclassesSome testcases fail still +It seems youre being inconsistent here why is it that shouldnt take an arbitrary delimiter but should? What I mean here is that either the concept of a line is a sequence of characters with a newline or its a sequence of characters with an arbitrary delimiter +I think the trace file parsing test might be pretty fragile so am not such a big fan +As is set for so wont be updated to until then +This client is the barebones implementation and hence has no smarts +Good multipoint as well +Run bin +HiI think this patch is almost complete +That problem is now also solved +it looks like this entire flow of the shell could use more unit tests but I cant jump into that right away +Thanks JonathanThe patch was applied to trunk r please verify +I think a unit test should be feasible +setting to MINOR since the backend is reporting the problem correctly Invalid cartridge +Getting started documents are now finished except for needs to be a part about Mocking too +So as a attribute it makes sense to query the masterNNQuerying just one node and making the assumption all are the same seems a little fragile you could easily end up in a state where some of the in the cluster have version A and others have version B and clients will be able to talk to some but not others depending on which one they happen to RPC to first +The question is if there are cases that require retry when a client receivesOne correction the jira that I referred to in previous comments should be +I think we have another Jira about that with the fix version defined. +In my point of view it is an indicator for a folder renaming +Hi AlejadroI attached a patch for that addresses all your commentsCheers just started looking at patch +Since there is a workaround reducing priority +Whats the status on this? Should we close it +I assume this is something that INFRA handles +Hi Clebert reopening because QA is blocked by test this manually and I use a new see attachment +Shall I make a pull request of these +Jeff! +Time has the same problem with the hours off +Bill any progress +contains definition of parameter that refers to global but the global is not defined I think it is missing for this action +Updated the files and committed themIve attached the diff file and updated html filesCommitted revision +All thats needed is a new release of the PC module. +I would love to see this in as it is something we need to use fairly soon but I have not had time to start the patchI would take a look at my comments above about Current Faults +One point we came to was that making these changes would allow single MR jobs to talk to separate Accumulo clusters instead of a single clusterI did settle on a change that Im not completely happy about that is reliant on the fact that splits are generated by one host in serial +reopen to set fix version to again with correct fix version +A new JIRA will be created for the new design and implementation. +When ant is run build fails withBUILD FAILEDfoo Error while expanding foolib is looking for jars with as version but trunklib has jars with name containing Just renaming the jars in lib to have is solving the build issue +I tried to do some digging through the source and while some changes perhaps inaround may resolve some of the issues the HQL parsed queries also require to reflect these changes +Seems to fix the problem +In addition rest proxy increases the latency for response time by one hop +Thanks for looking into this +Im closing this jira because a lot of this was related to getting the feature in +Cool will take a look this evening +Thanks for the commit Stack +Finally I decided to log that as a trace. +Assign to Vicky to look at if he has cycles Weston is not working on this anymore +Unfortunately my understanding of portlets is almost and Id love to hear your comments on the issue with regards to the portlet environment. +Personally I prefer the design of having take a factory +Nothing has changed for the majority of the docs +modified src +All supported types have also methods with a fallback to a defaultValue all except the wrapped classes of primitive types +or make the auth cache aware of default ports? Which would entail having to hard couple it with if memory serves auth cache is expected to find the narrowest matching credentials anyway rightI believe you are confusing it with +Im not that attached to Lright +Committed as r +We need to check the spec as well for exception handling when theres exception on server question is whether such exceptions have to be propagated to client +Will post results once theyve finished + +What would be good is to have an Ant script to copy the +I ran all the tests locally +The two changes were Increase max permgen size since theres simply more code compiling now +Thanks Jesse +The patch passed all unit tests and integration tests +Check your keystore settings +Im still using the local FSBTW we dont have HDFS as a dependency in maven should we add thisThe version in trunk should work +Created +Committed to trunk and. +As I commented above the server still cant recognize it has started correctly when the management port is not exposed +attached +If you want to take a crack at finishing them off that would be awesome! +Partially applied +OK added spec it passes alrightwbuildrsvn ci specjava m fix for Make sure is the first entry of the jarSending specjavaTransmitting file dataCommitted revision . +So the issue is incorrect +More improvements to the new classes +Ive committed this with Stacks suggestion to trunk and +The problem is that jboss is to eager to destroy non matched connections +sure enough it is exactly as charles describes +Javadocs are already generated separately for solrj and DIH but not as jars +It is weird that the config wont validate but the configuration still works as you would expect when using it at runtime. +as needed +Passed +What I am seeing is as follows Create project and server Open server editor page +Mostly are overwrite of test files +In what other situations could we disambiguate semantically +Ill simply leave it as String +Regarding Mikes comment on the th of FebruaryThe failures are only indirectly JVM specific as you will only see the failures when the connections are obtained through aSince the JME platform doesnt support a is used +Verified this is still a problem +Sampler is fast though +The Multiplexer now sends out views so from the point of view of the service the behavior is the same as before +But process using rest is trying to do something with key which is obviously the problem +I know it doesnt help you now but in Forge this is basically resolved already because all addon installation actually does work via maven so if the project happens to be built from source first it still ultimately gets installed via maven dependency question is do you want to wait for that? And Im guessing the answer is no +Ok sure +Here is the patch for the test case +Issue may already be resolved may be invalid or may never be fixed +Ill add this to the main patch this week +Sushanth thanks for going through the patch +Enjoy +Maven artifacts deployed at Maven central repo have license info included +I can redo the dm patch as necessary to sync up with this one +This message does not appear any longer with the latest versions. +the first lock addl with an operand of is interesting +Oleg I made second patch from Mar could you please take a look +if the test is ok +The current patch leaves the old access path in place and has the on the side +Hi Paulex Changes committed in revision Many thanks for this enhancement +Awesome! Thanks Josh +Thank revision in nutchgora +Marking as resolved with the latest commit +Im reporting these issues back to the OASIS specification group and will be holding off on committing these changes as I cant get a clean build even with several workarounds on our side +Another example upload is sometimes an arrow sometimes a button with text +In the code is mistakenly done on cmdDir instead of str +Discovered that it is Hypersonic only +The example you describe is sort of at the heart of the Hunspell algorithm and outputting those three different tokens is one of its major advantages +Ann could you post some more info about your code and your +Contains Serialization change and test case +I just committed this +Also please try this with latest snapshot the DOM replacement method has changed since beta IIRC +It was a pain to explain why this step is needed and to enforce +I prefer to define them as this and true by default correspondingly and hide from the taglib +Bulk closing stale resolved issues +The class wide pool has been restored and synchronization cleaned up +I think the performance hit for the new over is acceptable +Could you please give an example to elaborate the use of base +Rebased patch to trunkOwenTodd final comments +Ive attached small patch to show the package structure Ive created for the social API +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +should be up on the Red Hat portal within the nest hours +We have already changed the icon for server actions +no tests included because this simply fixes a minor memory leak +The problem was the different in versions between the dependency management and the transitive dependency have added a warning and will use the version from the transitive dependency for packaging. +From the NN UIJMX the administrator cant tell which Active storage directory has fsimage inside and which doesnt +Any update on this +Ill check this out by wiping out the workspace and launching a build then well see if the build gets stuck because the sanitizer install script is waiting for input or not +It doesnt seem to be fixed for and any news +That is also true and as you say require some customization +Dont know what a pure java solution would be in this case +bulk close of resolved issues. +Moving to until after Daniel and I finish the performance work +the jstl tld would be a good testcase +Needs to go with +In fact I think its a feature of people using sstablejson are Doing It Wrong its meant to be a debugging tool end of story +committed to some puppet dependency issues + +Draft patch that implements Query timeout supportSo far only tested on Derby +Patch for and files for src didnt use the attached assembly files +Ok this is not en entirely correct +The patch Did not give the class +No problem I will do that in product branch +Theserules can be found in chapter in XSLIf you want me to point out to you exactly why the border isnt shown in yourspecial case please reopen the issue and attach a full FO file +For me the clincher argument is the need to review the release notes +No It need review and merge +Thank you very much NickHow long I need to wait for version? Can I get a copy of version to test out my code +Verified on Jenkins all tests are passing +Fixed the typo +I tested with two other IE instances and is working as expected +Issue will be fixed in that exists within installed EAP for after installing from installer jar +Use a command line tool to browse the completed jobs information +The error condition is actually defined by the protocol as meaning that a bad value was read off of the wire and as such the documentation doesnt acurrately reflect what has occured where you use it in +Im releasing the version of the Emma plugin. +Under WindowPreferences +This patch uses the caching parameter +After that method get prints the progress of running queriesSo it doesnt matter what I skipped API +I havent seen this test fail for a while now so this can be closed +Attached is a simple Eclipse Ant project to deploy an EJB based endpoint with Username Token being used for the authentication +Somehow related to its a pretty big oneCurrently people cannot have query with both having and group byFrom our side this is clearly not an easy fix +on Elis patch +Are you sure? This issue didnt exist when using WTP and I upgraded and the javascript errors are there WTP the WTP error already reported +bq +There are possible cache performance implications with for the detective work maybe we should open a new JIRA to look at the performance issues of TIB constant folding +The same with the previous one but create it from the root folder +This issue has aspects of logging in HCM so can try to tell whats happening around rediscovery of ROOT killed investigation of this issue +The one failing test is due to not handling migrations yet +I dont know that we actually need a M heap +The other issue is expected as well +Can someone please close this issue? I dont have the priveledge of doing so +At this point I think it would be risky to modify the behavior of the default template +This draft incorporates Dags comments +But by hidding the parameter table the users cant set the parameters on the ftl resource anymore +Ill take a look by EOD today +mccraig could you check this behaviour with the latest trunk? I just fixed the other related issues and perhaps this could be also fixed +Attached patch makes the Offheap allocation pluggable and hasTo Test Plz set export LDLIBRARYPATHxxxlibJVM Property xxxlib an updateJason Evan says LDPRELOADing jemalloc should be okay as long as the JVM doesnt statically link a different malloc implementation +v and w should be different objects +Ill look to make a release of the archetype +I am working on this +Applied patch with thanks to Karl +OK fixed the section in the reference docs and added clarifying comments to s +orgjbosslogging +Also hoping that resets will be unnecessary but this is a measure +Integrated to TRUNKThanks for the patch Ashutosh +The trouble with this is that it breaks binary compatiblity so if we want to do this we should then upgrade version number to +was failing with null pointer exception in because was set to null after namenode was shutdown but the pointer was being accessed +Ok thanks for the hint I missed that one +Nope closing +There are two quite separate uses of Chain here and I believe we should treat them differentlyThe first use is the request processor chain itself +Is corrected in rev +It seems your patch has lot of text modifications and really hard to review it would be great if you can provide a new patch based on latest codes in the trunk +The timeout currently is seconds +Closing issues. +It means you wouldnt use counters as a system of record for an advertising system counting dollars and cents for example +I the test using debugtrace for and piped to to isolate the messaging operations more clearly +versions scattered about is a real mess +I changed synchronized to volatile for modifed and initialized +Thanks +I suggest creating a linked data package +a revised patch that does things removes bloom filter flags from enables binary key pretty printing +released today please check if the problem still persists +Johan Hammar +We close issues when the code has been released +Please test the latest code and ensure that this fix is working correctly +Master now on will use that to communicate with RS but RS side there is no RPC server bound with this hostnameip +I left it this way because of batch operations you would want to know the SQL associated with each step of a batch +found while investigating delta failures with +NET projects and the developer can to choose an option to build clean install and +I just had a look at the and Servicemix uses old version of whereas the fix is applied to version and will be available in Ulhas +Committed svn revision +The larger file contains clover classes +After coming to terms I have committed the version of the patch +The doesnt include as a dependency shouldnt that be done as part of this fix +If it failed please attach datadump before and after upgrade besides ms log +What are you missing in the CHANGELOG +No mentioned issues found in JBDS +return newly generated itemThat will always return the same item unless some other process updates the cache map +Ill see what I can do about the performance m not sure we should use the current patch but rather start from it an idea to base a performance on since they too should beindependent of one another +A query facet is just a query that executed inside the main query and for this query the hit count is computed as if it is a facet +Perd! Ve de la. +The two major topics of this issue are resolved +Fixed as part of +Your implementation supersedes that one +No longer a problem. +Cool yeah this was the real reason I was doing that last fix I just now datastores a bit better so started there +as +Should we remove the boolean return value and always throw an? This is the same case with copy +This error will kill the thread that pulls new sockets off the queue and will prevent any new connections from working +The Security provider also load the default policy +Geronimo is currently evaluating the usage of this plugin for assembling its binary distribution +Having a timer associate with each target is to avoid under replication when the replication command gets lost or the target data node shuts down before sending the replication notification back to namenode +As soon as I changed the parameter into a primitive it it helps someone +I think it might be related to IMAP connection pooling +If not please reopen. +Committed to trunk and x. +It was a clean install of the Platform Runtime Binary followed by an installation of the JDTMylyn from the Ganymede update site and then the rd party plugins from their update sites +So this solution violate xmls encoding specification +Check if the problem is known on jmxmonitor project and if yes then port the patch to our forked sources +Modifies binary target to depend on libhdfs target +Description of patch Remove processing to expand data from of blobclob to memory before sending to client +Also committed to. +A refresh of the project however may be required +Looks good +Works well Dinesh but we generate some on classes which doesnt implements the Collection interface +This needs more discussion +Verified +Sounds like a potentially useful thing to yes help is always appreciated and welcome +Approved for +Thanks again for the feedbackI should have something to show soon +abebbdbecefdbcefcbranch master +Can you please clarify where you saw this? Many thanks +Hi RenCould you please specify version of SCM Activity PluginI suppose that its +SaschaThe approach I described is used in a number of Ajax requests already just take a look at some of the current Ajax requests +For consistency I would suggest we change all the scripts to use bash to guard against this regression in the future +Closing all tickets +If you have the content for these checkboxes just let me know Ill take care of themThanks the template registration wizard you could use thisPublic Make this template accessible to all users +First valid war is whats opened +needs an overhaul for the configuration +The minor change mentioned above is still recommended. +I ran tests with both FD and FDALL and this always worked +So Im skeptical that there isnt more work to do +Fil Ive submitted pull requests for and to resolve thisIm not familiar with the release process for plugman and cli cli appears to depend on a particular released version of plugman and it would break if the CLI changes were applied without the plugman changes +When something finalizes we can add it to the wiki along with a few other tools +Keeping this open so we dont forget +The does not work properly +yes we have the ESB module I think we should have an icon for ESB projects which I have ever talk about +See. +but then it cant be the interfaces fixed +Committed rev Thanks Bruno! +Ill sync to trunk once you investigate and fix thisRegards like JAXB already encodes the certificates when marshaling the JAXB objects +Why more complexBecause I do not understand how it works +Last snapshot version works for AS EAP Ill update to the new release version when its I see no sign of a version either in Nexus or Maven Central +I think it should be in a separate thread +Maybe because flexmojos compiled with and working only with it? I think that is not rightBy the way why Flex SDK must be mavenized? So far as I know Java SDK for example is not needed in thisOk maybe mavenizing Flex SDK is a good simple way to use it +There was new measurement added +I believe that would just be changing some import statements to remove the classes and use the version +For code packaged as a tapestry library in a separate jar you can share the ln files +Fix the findbugs +Done +Cant check right now will do +think configuration for exampe +I think the problem becomes then the +My position would be bug in ACFIt shouldnt remove delimiters and the behavior should not depend on which element of the list you are replacing +settled on for a while now just closing issue as there are no immediate plans to change to something else. +Now test if contact is in parametersYes you are right for creation i meant after creating a contactMech the association with a purpose was mandatory without this patch and that surprised meThank you for taking care of it +Okay good point +At first limitRender works fine +Can you give an example code I never had this happen not even today +I also meant to add that its been running on one of our clusters since yesterday and its fine +I had committed it only in the branch now its also in the +closing since fixes are in released code +The commit associated with this defect was incomplete +Resolving as Wont Fix since I dont see much benefit in this change and since the patch no longer applies cleanly +RobertCan you add a test case as well +Attaching a patch file with the fix for the +If by you mean then yes +patch with testsIll include some details in the upgrading section of when committing +A minimal test case or at least a process definition would be very +retargeting from SVN head to so they dont get overlooked in release planning +most of the code here is the new objects to hold that common code and that are annotated for jerseyjaxb to automatically handle things +So I just want to know if I should wait a few days or if I should find a workaround. +Id be up for spending ten minutes around a white boardPrefetching is useless +As with the latest trunk version of Guvnor you are able to build packge using Use selector option without actually writing a selector file +defer all issues to +But I can learn it in the community bonding period if I get selected to this project +I found resembling problem for message of MyrnaIm linking this issue to +To update the whole tree point ajaxKeys attribute to set containing single created using default constructor +I just checked and Row extends which in turn extends Writable +Feel free to reassign this to a specific release but only if you have the time to do resolved per Martins comment. +Problem was that the SNAPSHOT builds were still picking up parts of M +Thanks +Please dont use in package names +This looks like it is a core server issue rather than an ESB issue which version of the server are you using +If the later apps wanting to be nice could wrap the IO streams with throttling aware ones +Closing as is already released months. +Test case added +What are you referring to +Other packages made from those sources are just a convenience +Otherwise it is hard to fix bugs like +There is not a clear separation is what to put in the session as field and what to put in the session as statemap valueImho you can add it where you prefer the whole thing needs a and well decide later how to handle it +Example of custom properties problem in was creating spans when I should have been creating divs. +This has been deferred from the current ESB trunk until we integrate juddi v as the scout release is already complete +Yeah I think costly but perfect minimization is the right default. + +Username nobody +If outstanding work please capture in a new JIRA for M. +examples and not +Build is an example +oops thanks for catching this! +Hi ErlendSince the parameters are completely local to the postgresql driver Id prefer to move them into that class +rick added impl needs some unit tests +I will add a symlink from binruby to binjruby and well see what happensAt the very least Ill have to be more careful in my environment about what ruby command I run +Im not yet sure how the casell work which isprobably the most logical to implement +Resolved and committed. +Right but thats just impl not design +wait on patch there seams to be a logic error somewhere let me run some more test I am seeing major compactions more often then I should be and not when they should be +Carlo thanks for the info +I should be done with Webtools sometime this weekend or Monday at the latest +exception seems ideal to me but logging would at least be something +We could consider rewriting the rest if it makes it easier for users to merge our rest API inside their own webapp +So it might be a while beforeI finish my implementation +Top panel side +Thanks Carl +Looks like there is a typo in the latest commit +This patch is not intended to be committed +Using separator should be enough but the built in regexp logic in dojo blows up because it doesnt take this into account +For we could remote it no? I can do that if you dont have the time but youre ok with the idea +Ive just committed this to trunk andThanks a lot for the contribution Andrew +Im closing this issue since it isnt a bug but a application configuration issue +Shall I open a new jiraYes please +what sort of error do you get? The wrong number of rows? The wrong values inthe columns of those rows? The rows are in the wrong order? What resultsdid you expect to get and how do they differ from the actual results +thanks Ill close it +The same goes for this variable that if it is specified in that value will be used otherwise the value from is let me know you would like it to be another way +because then you actually have a way for the user to workaround it +Encountered with the same issue +Simply choose acceptExisting over shared now for sharing by cache manager name and for controlled shutdown +discussion with Dan in +Added new integration test and ran on windows and linux and found no message has to be sent to a topic for the test to fail +approved for by Chuck +If more methods are protected itll help +Committed in revision s not really harmful as theres no state in but I agree that the if statement there is not needed +Attaching first version of the patch which deprecates +The zombie threads are from andSo this is looking good +Cheers Paul +Updated patch +I totally agree that it needs to be investigated +Patch that uses reflection to see if syncfs is available and if it is uses this syncing +unfortunately quite a lot of effort elsewhere in the core +The new proposal makes sense +I will have a look at your patch and let you know what I think! +newer version with for looking at this Si +I used the custom resolver because it seemed to fit into the pattern of other JUEL resolvers for retrieving properties from an object just happens to be a very powerful way of doing that for nodesAnd what do you mean by simpler that it would have been simpler if I hadnt already written the code +It can tell us its stopped when its actually started +Im not sure how much parallelism this patch gives +As discussed above. +I will just set this as a duplicate of so people would find it +Have not investigated what breaks +However it is intended that access can be performed concurrentlySo I am going to change the subject of this ticket +Changing fix version to +Throw an +to investigate interaction with +So its there but I still think the should be listed and maybe link to this advanced part at the end of the GORM section +I recommitted with the fix from +Applied at revision +For the period of time it takes for a broken network to be recognised mesages can be acked in error +The correct brew builds for and are now in the corresponding tags for EAPEWPIll update the brew builds when we get the errata back +Please review +Hi We implement our Converter by scanning JAR services +Reopening and assigning to me as a fun post arraycopy correctly is one of the highest impact single decisions we make in the opt compiler +HiThanks for pointing out this issueSent an alarm mail around to community +Oh I had already done this +Im looking to move the quorum to other machines +Appreciate the help +Thanks very much for catching this error +The problem here is that the way we have defined selection indexes it is impossible to distinguish between the index that indicates before the first newline and the index that indicates before the second newline +Sid kindly agreed that hell take this up +Hi JorgeCan we get some more information on this +Thanks! +the latest but not committing ran into a couple of issues I went over with this applied now +The issue is not a duplicate it just happens to have the same fix as another issue +This is intended to support external management such as AmbariIm not clicking Submit Patch yet because this patch is dependent on the patch +This is cool Patrice +Then commits came in in September for Openshift is now + +Next Version including form to modify the selected file and button to save the changed file back on diskto I just downloaded our other patchAnyway itll be a busy morning here but I should be able to take a look this afternoonIm geeked by this! I think itll let people get started using Solr easier +With that said this can definitely be done if there is a need for it +Yeah perhaps I should have posted the patch onI think this work requires to first be backported to +Ok I updated and retriedjoernkarkanduima test bin Usage specify XXXX where XXXX is a string of file paths to directories or jar files separated using the proper path separator character +or ROOT table +Hey John how are you importing the project into Eclipse? Are you running mvn eclipseeclipse prior to import? Can you also please provide the script for the entities which are failing to be enhanced +Venkat +No you havent Thats the whole point of processing a processor must ignore the xsl instructions it doesnt know about + pending Jenkins +Have you benchmarked this at all yet +See Section of the installation guide for details +Clearly Im proposing and not Hairongs daughters picture +Thanks ATM forgot to change after +This patch is also needed to ensure that searches on a multisearcher will be the same regardless of how documents are distributed among the subsearchers +next question if they are checked into the main code line andbundled in one of our jars what would be the argument for not them? Name space pollution +I might be wrong though +DavidThanks for putting in Fixed Asset Maintenance component +read records from Hbase which is a how to know the schemaless schema +I more or less agree +Well when we found those classes they really were just a proof of concept and the jdbc datastores took them to something that actually worked +Portal will not work with this default configuration so I have added the manual steps required to make portal work with the Installer EJB module selectedYou must move some libraries and services that Portal depends on in to the defaultdeploy directory now a +I thought that was no longer the case? I can no longer find the mail I thought I saw fly by on this though + +thanks for the patch! +I will take this patch and commit it to the branch +This IS the code duplication +Committed to and Thanks for the Christmas review Ted. +thx a cramp its in the proper location now +this mean that this fix will be included in final +usesorg +Shouldnt you change your code to use the class that replaced the deprecated one +Publishing to Dist is instant syncing to Archives is dailyIts there now +All done I think. +Thanks Javi for this patch +Committed to CR +Later when listsnapshots command reveals the rollback snapshots people are surprisedI think dropping the rollback snapshot is desirableI will provide a patch as outlined by Matteo above +bq +So the reason this happens for Lists is that decorates a List association with a that creates a new instance of the domain type +I like all of this patch except for the separate +lets comment that and move it to debug then +JarkkoI ran the test and now Im curious +But where should it live +Further testing using the attached test case proved that the change only fixed a java Boolean deserialization issue +Im not convinced they have to show exactly the same thing it has different purposes +I looked into your modification in detail +We are actually spawning multiple processes and using a flag to pass values to them +To be precise the idea was to optionally let a Table or declare whether or not it would allow a client to set timestamps +Thanks Dag +I can add you to it if you give me your confluence username +It would be great to add this functionality as an external +And sending the counters is capped at min +Pls provide your comments +Shoudl we reject this issue +Committed +However we are forced to iterate all messages in the produce request +I think Hiram is a bit unconvinced about this patch but the removed code is clearly wrong and I cant see anythign wrong with the patch +Heres a patch with the Im using in my project +Also there are many so having to reference each individually would be a pain +I think I had trouble recreating it on a small project +I am creating nodes as well as saving and loading them all correctly +Still encountering this issue often. +Appears they are back Ill try the nightly builds and let you know as well +Note that the above list of statistics is a bunch of Cache internals reflecting an overall JBC instance and not the region related to a particular entity +gwt setupweb gwt all proxyPackage +Test Plan in the works but not ready yet +I verified that claims raised in this issue no longer apply +This issue is resolved in M. +Is there something preventing us from committing this fix as it is Jukka +Heres an initial patch to make sure folks are OK with the approach +as still +Its next to impossible to change that +This is the patch for trunk including the hack for +strongly disagree with this +Maybe can provide some inspirationThe command is the only difference between a client and a broker connectionTo see if a broker connection is complete get can give you some high level viewNot sure if this answers your questions but I hope it helps +Starting new issue related to to request permission to merge to branch +Note I get the same exception when I try to update an existing XA or Datasource via the Profile Service + implements which member method on currently handles updating the package name when the Rule Flow file is attached to the asset +bq +Just looked at the Servlet code its pretty bare bones now that we have the +I committed this to and to trunk +some +Verified at r +Can you have a look at trunk and propose a small change to make public what you want to get at +I think the notice may not yet be fixedI think theyd like attribution +What version of Tomcat did you use +This class should get additional code to filter the resulting classes on the Test annotation +the problem with the previous settings was that they were too flexible and allowed for different encodings in cache and in memory +I ll update the patch on to reflect this text +Assume closed as resolved and released +Do you know what build be the earliest with a fix +The attached patch adds and to the list of jars in toolsjava and moves the paragraph below section. +This change should not be merged back to branchI willl submit another patch to add new upgrade test for upgrade from to +Close ticket +These tests use configuration file and so should detect whether AS is able to start up with this conf file or not +Thanks! I applied the style and header patch in revision . +And this seems like an easy fix btw +Im sure there is an advantage but Im failing to see it at the moment +You are probably already seeing them because the DGM variant of these methods are being proposed instead of the PGM methods +Added that plus new test cases +The following is replication queue initialization times +Duh! I confirm that it works pretty stupid of me +A better way might be to add a mechanism to add fields to a document +Then on first start runtime detection check this location and if server is copied it creates runtime and serverIt is done in class from plugin +Nothing breaks indeed +But this is the case currently the pages header is sliced into too much the panels header is handled as a block thus possibly resulting in duplicated js contribution +No worries Daryn +It doesnt need to testing +Also opened as an umbrella task for gathering and discussing many possible such improvements +I agree with Kojis idea of adding a config hook for auto for those that want it though but i dont think it should be on by default eitherThanks +Thanks Elli +Please change javadoc for provides to not just say return true implementations that override this to return false might forget to provide their own javadoc inherit this and this claim behavior opposite from what they actually do +I committed the patch to trunk +Patch will be up in a minute +Summary its implied here that Guvnor Tools are Eclipse Eclipse IDE with its workspace is used +Upcoming TB will use AS +ThanksPatch appliedColm. +Oh I forgot to set the fix versions and to thank Shivendra for this contribution thanks again +Had to revert to an earlier version to avoid this +Any primitive type works for me as long as agrees with the XSD +Apparently I was not aware that the second job was already doing the sortingWill test and commit +but yes it feels like much wasted space +Dev should provide the converter for that case on his own. +Hi Marcos can you add the steps to reproduce such as how you sign your entity etc? Thx MarcosIt looks like the signaturepresent is not fully implemented +thanks for the patch +Hi Sascha Why and how did you chosse elRTE? Could you give us some elements? spoke about elRTE a few days and i find it was the best looking open source editor i found with the right license +You can have a look at the accociated test case resolved issues for already released versions +I am going to run the tests with and report back as well +Itd be nice to see if xavier can still replicate it +Any chance you could create a matching test case for itLook for orghibernatetestsql which contains test cases for thisMuch appreciated +opening just to changed the fixed version +I think that we all agree that the server nameid negotiation is needed for general sasl support +Slating for fix in you! +Thanks Andrew for the pointer +There is no file descriptor or socket leak we are aware of in +Resolving this now +Seems like removing allows tests to run +Can no longer recreate believed fixedPlease this bug if the problem recurs +Regarding comment from Martin are you in a position to help usout with getting as many tests included as possible in srctests forHSQL +application was written to help developers debug their YARN applicationsframeworks by running AM on the client machinebq +is a patch applied to +Corrected mixing up the digits of the issue number +Should we do that as a part of this issue or in a separate issue +When invoking it on the whole framework a number of issues arise around individual bundles that dont like to be refreshed arise for which is filed. +If this fails the cache will be assumed to be a path to a file +Can you make that the same here or shall we change the default in m +Applied to branch in commit review and apply the patch to branch. +and after the cleaning the problem reappared so javascript seems to be part of the problem +Could we move this discussion to the mailing list +IMO it would be nice if we could have at least some configurable logger that allows user to redirect stderrstdout to file or some other placeIn other words I dont think the JIRA by itself is invalid +If youre satisfied then you can resolve this issue. +I guess we might want to define a new method say append in the API +This is an interesting case +I tried Ravinders patch but it wasnt working for me for starters the dropping and closing left or created locks + +Thanks Nicholas +I agree it seems reasonable practice to deprecate the class in and give people enough time to prepare for removal +Namit please see my comments on phabricator +trunk Thanks a lot Stan for the patch +Unscheduling this aggregation issue will be scheduled and marked solved when all the subtasks are done +The MR client would have to be modified to set this boolean variable +The comment in the code explains why MB was chosen rather than something larger +For some reason hitting OK on the dialog where you specify ESB interface details clears the service name + +But Im not an Eclipse user so my opinion is of limited value +With this fix if fails to connect to one NN it still can connect other if any without throwing NPE +So all we need is a tck test caseIll take care of the specification update +A cache was always part of the intention of this package. +the link was dead the last two times i tried +Username jsimone +Git is way too flexible and confusing but I cant live without it nowYou can do a full GIT checkout from the links on +Thanks Simon I feel better now that we get our tracking back +Marking as resolved per Patricks comments on April th +But if you are using Hibernate and Maven for your build try going down to Maven and check if you get the same error. +Resolved see bugzilla +And the cert I used was from another connector that worksIf that was true you should not have seen the error in the log you postedTheres a disconnect somewhere either the log is not the right one or the code you are running is not from +Talking to Bela yesterday adding the flag is the quickest solution here and Ive narrowed down the use case when to use the flag so it will be accompanied by plenty of javadoc and Ill blog about it +The point of the second patch was to prevent us from using the file system +RBW B and RBW Cs GS gets bumped up by pipeline recovery since the client detects that A in pipeline is downEven if the client fails before pipeline recovery starts the lease recovery does not takes RWR into consideration when there is RBW or finalized replica available +Thank you guys for your reviews! there any objections to me changingGood point about the compiler error you could change the name in from withCounter to addCounter and then you wont have the compiler errorOne other suggestion is that the expected counter size check is unnecessary at the beginning of validate success will remain true and neither for loop will run if both are +Thanks for the fix Jon. +With an upgraded package this patch shouldnt be requiredOk that seems broken +I also see that we are hitting the because we are not presizing the +So looks like since the client cant determine that the new exception class is a sub class of all of the reties are being issuedIll post a new patch +For the moment though in production we are not using the feature as advised in the documentation. +I managed to close the other hand in another thread +Im not setting this as patch available because itll break things until and are committed +Hi Jake the issues you raise are worth investigating and yes both syntaxes should work equivalently +Do you want me to repackage things? In effect only the src and test folders need to be processed +Mikhail did the work he deserves the credit +Eventually certainly thats the right thing to do +is very good +this has been seems to be working fine +Thanks for the clarifications +Good catch Amareshwari +Keith did this fix made to the SFW? +Just setting to M for now +Changes include Included source file which was missing in the previous patch Fixed a few issues which were uncovered in some tests regenerated the patch with the latest svn we get this patch committed +I think the issue is SUBMITTED And ACCEPTED APP transitions does not support for ATTEMPTKILLED which is actuall being sent when KILL event is dispatched +Please verify the problem is fully resolved as you expected +The only docset that is right is the bundled distro +Attached is a screenshot from our heap dump analysis showing the large number of entries in the the in that screenshot indicate the number of instances? If so what container are you running +I created a separate module in the examples app and installed this +Committed to trunk +yay it seems to have worked +Very soon trunk will be branched for Pig so if this feature is useful to feature in Pig we should get this committed soon +Attached file contains fix in the execute method that releases the connection after execution + and look good +Globals can now be referenced direcly in the LHS +FYI Im trying to revert the change from this jira at. +Sean thanks for offering to help with questions about the patch process +Was this ever completed since all the are done? I know were already on rc +Ive committed the fix and merged to. +The findbugs are within trunk +It might be and depends on what structured model provides for entity declarations and references +If you find that only people with shell access to the server should be able to make backups then why not disable the backup function inside of OM completely? That would have saved both my and your timeGreetings no organization name I might have forgotten to fill it in while installing OM +This is a working callback argument makes this approach untenable +That is except that it might trick users into writing SQL +Attaching server patch + +Test application a file with jbosss logs script for testing the application +I suggested we allow both modes because we have been using the model at Facebook +This patch uses configuration to leave out the etc +I think I will change this again to put the on the install page as my plan is to have this info together for each releaseI dont think you have to define a geronimo update site in order for download additional server adapters to show Geronimo +I now understand youre suggesting to disable the integrated JSF facelets feature for older versions +I would not support value compression +ok marking fixed +I have merged the subtasks of this jira into +BTW I fixed this bug I found an interesting thing in speculative op each time the speculative read op would be scheduled even the first response is already received +The User Preference feature has been moved to the common component +Tried messing around with the way uses a regex but failed to get that to work at the single feature would include mockito which is only necessary for testing feature included in Designers test feature while the rest included in Designers main Designer can be modified to include plugins rather than its features +maybe rename to? or just? Elsewhere we have the strong convention that Writable implements Writableafter fixing those from todds last have merged this to. +Screenshot is in the issue and I got the error when I clicked on ToolsRuby gems +me +I contributed the first patch to make possible see +b duplicated words which represent represent services page chapter +So by convention it should be hadoop +I think the qdox behaviour is wrong but for compatibility reasons Im unsure if we should change this +I had a quick follow up question on the less radical patch particularly when we have increment operation +Why would client applications need s memory information +Hi KnutThanks for reviewing the patchI have attached the patch this will allow those queries you have mentioned and broke the code into another method for the ease of readPS Ill attach a different patch including the changes to theThanks +Ill be working on this on Thursday and will let you know how I get onFeel free to close the issue or assign it to me in the meantime if you like +here is a patch from playing around this morningIm afraid of specialization here but this one should help the relatively low bpv I think by using readShort +The attached patch can fix this problemThe logic seems a little confusing so I added some comments in itAny suggestions are welcomeThank you +Regardless persistent connections is a nice to have anyway +I can no longer reproduce it either +Should be easier to do after this work goes in. +Back to the start gentlemen +Patch is in rev +Old wizard has a feature missing in the new one list of namespaces available in the project to select some of them to be inserted into the created page +Despite that difference the memory consumption is equalIll also restart a node with NIOFS but i still expect memory to increase as with mmap +strtoul and strtol are not giving expected values specially when deserializing values near to +Fixed in still has a SNAPSHOT dependency so I will not close the issue yet +How can the event type be invalid? If in the future we add more event type into we should also add the method to handle it +Thanks Knut AndersCommitted to trunk with revision Closing. +Attached for POJO service +Thanks +thanks for the great test case issue resolved on trunk in rstate on the client broker and failover transport needed to be reclaimed on prepare readonly +Fixed by +I try with RELEASE and I check again with RELEASE +Attached an IT for this to simply for anyone that wants to take a stab at this +Hi Tim Sorry for forgetting to close this issue +Thanks for taking care of the autoconf stuff +Alan can this be resolved as fixed now? +Im running into this one when using cached gems on my app +Im not all that familiar with but the thought that came to my mind was why wouldnt we do all this through JMX +patch fixes some of the iconv specs. +Note that these new failures are not platform specific +Im not calling for a configurable number of mapperssplits +sam Yes +Discussions around the deployment scenarios will help drive how any layering will be done between Knox at the perimeter and HSSO and TAS inside the clusterI do have some questions about the authorization approach as well +See for further details. +It seems that the number of operations in the SOAP binding are not the same as the number of operations in the port type +Is there a particular test thats failing? Perhaps the issue is our tests maybe were not using something correctly. +The problem would then be in createScript +The JDBC version information is not essential and removing it definitely keeps things simpler +Works fineBut lets commit the fix to branch too +Interesting if that is indeed true I am for what do we do here? will fix this forBut weve broken the promise that all versions of and are compatible with each other +Bummer +So hudsons checkout will also need to be at least Thanks +was removed. +Currently some are broken and ignored +ill write it on my listCheersYes I plan to have a look +Otherwise I am definitely on adding that extra complexity +Thanks. +Most of the time its the assert on line that fails but sometimes its the assert on line With the attached patch which makes the test wait for the post commit thread to complete before it queries the lock table Im not able to reproduce the failure anymore +The patch seems to make things compatible again but I want to make sure it doesnt break anything else +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +our adds on context params were not picked up anymore +it realted to review +Continuing on the issue at hand I looked at the code and it seems the publish flag in has no effect if you are running in Direct Mode as has this value set to True and it doesnt look it up from any of the config xmls neither it is exposed in + +Moving these to please move any bugs back that will be worked on before the November release of +Whoops sorry didnt see your comment above before I posted the above + +yes this seems to be a good hack +it looks like the collection editor is unable to gather itself +fop c has been added with description the has been added with description Attachment includes minimal test for embedding SVG and console only receive the error for SVG files +They can only be accessed by getting child nodes from a particular node +setting LATER since this is a bug in the server we have to keep an eye on +This time with line +says its the editor for +I made it sync explicitly so that in future it cant become a problemThe problem here had nothing specific to +This is not appropriate in this particular context if no encoding is specified the transport should let the XML parser determine it +Attaching patch to implement this so I dont loose it +Should I need to apply both patched or just one +While we are at it we should change the zoom out from a tool to a button +Thanks +I think I now understand the functionality being proposed though the spec could use a little gramatical cleanup +Ensures that it is set at each stage so that they comparator can be derived correctly +After discussing this with Tom C we wont need it. +Fixes the issue with browsers dying on SCBADREQUEST HTTP response code +According to the history of BCIR athrows have always been optimized to gotos thereby creating a normal edge to an exception handling block +has been released closing issues. +Attachment has been added with description the PDF output of the problem is also present in subversion HEAD rev +I took a slightly different route and have posted my patch for review +I was discussing this with Aslak and developers and it seems that their remote container is behaving as managed automaticallyI am going to play with that a little +For a normal server instance this will never be present +Yes were aware +Committed revision +for this makes perfect sense to do this for all +For example we implemented a deploy and manage toolkit for our hadoop clusters most of the operations are done in shell command line we do not want to specify several config files for each cluster we use the toolkit to generate the config at runtime and pass it by +Thanks binlijin +Since this is not critical for I suggest that we make the patch for trunk +Would not the best way to do this would be to do a map that formats and sorts the data per column family then a reduce that writes a mapfiles directly to the regions columnsThen that would skip the api and speed up the loading of the data and it would not matter so much if we has region or not sense all we would be doing is adding a mapfile to hdfsCourse the map would have to know if there is region or and split the data correctly but even if each map only produces a few lines of data per column family the compactor will come along sooner or later and clean up and split where neededSo if we add map files to one column I would assume that it would slow reads down a little bit havening to sort threw all the map files while scanning but that would be a temporary speed problem +I understand how it works now but Im pretty sure Ill forget that before the next time I need to know itThe idea of generalizing to allow including in it sounds interesting +HiHas anyone been able to look into the patch and apply it? Can you please let me know your ideas about these modifications +Macgit +I believe that if you delete the exclusion you will still see the problem +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Are we changing this to something different if so what +This is a better representation of the filesystem browser which will be used where applicable within the UI +I dont think it is being used +It seems soAnd my patch should be reworked +Reloading the page makes the Favorite panel display recent change fixed this +Along the top now appears a title for each panel as well as a number indicating which step of the process the user is in +I currently have no idea how the existing graph data is gathered and displayed +The persistence on requires even more edge case analysis testing than this feature to meet its requirements +Thx again DanielEric +reserved +Changes committed in revision +Igor the patch was applied at the revision +The error appears after I have created the embeddable class and use the roo command field a check for embedded objects in commit acfddecdccfbbaa +I dont think this is likely to be a software error in at this time +See for further details. +Trying to determine if its something we need to get into +and also Java centric +Project has been retired so wont fix +As a last desperate solution i deleted my complete local maven repository and the build still failed +Infact I thought we already had this to turn onoff indexing but it looks like it wasnt done +Unfortunately no +We have a global exception hook set up to shut down the process for those +Thanks for the report Anton Ill investigate and see if I can replicate the problem +Please provide a sample application that illustrates what you are talking about +Yeah Scala is weird I know. +Patch that fixes descriptions +So system property resolution does work as you you intend to resolve against is called system environment variables from the JVMs perspective +The bahaviour was corrected while fixing +Mathieus patch synchronizes methods in but that class should be kept unsynchronized so I moved the synchronization up in the call chain in Since so many people have hit this I would appreciate you guys trying out this patch and letting me know if it works +lgtm +VERIFIED +Patch test +I still cant reproduce the issue +I will fix that +we need to avoid using external resource to avoid such blocker problems +You have used Helios in build +AlekseyI wasnt able to reproduce the problem using test on current Windows build +I verified that is already applied to CVSand the change in the documentation looks OK to me +The changes I made probably wont make much difference for you just some minor +Oops my bad sorry +an of which appears to contain as many objects as there are on diskNot sure how badly validation needs all of these at once +You are right the difference is really small and your code is less complicated then mine +Zhenxiao Please see my comments on phabricator +As it is we will not merge it into this project +Any reason not to default the jsp pages toWe do want to default to UTF but we still sort of support Tomcat with this release which is only up to servlet spec +Ill change the name of this issue to reflect that + +Only by doing a google search for grails date were we able to discover that solution. +Committed to +Thanks Stephen. +The war itself does contain a manifest that still references the Spring libs from inside this way you wont get the neededreferences which are provided by karaf +The binary for the plugin +The test that we currently have for detecting when the server came back is less than robust +I think OP is saying that replication by id will not replicate a delete but instead throw an error +Attaching patch incorporating comments +Bulk closing stale resolved issues +We think that this value is correct in pom because pom is the maven reference and its used by lot of ll reopen this issue if we change it in future or if we allow users to choose what is the value the most important. +Will it be supported in the near future +Looks good at first glance if it didnt break a bunch of tests +Please reviewcommit this patch +note with last patch both tests fail and +I had issues building jars from the source but I downloaded the binaries and manually placed them in my drools runtime folder +I just committed this +Verified on JBT +add diagram for check an additional precondition for all streams must start on the same txid fix typos patch doesnt change so I dont think the build failures there are relevantStill I will resubmit this so that Jenkins can it +Attached +Why dont you just create patches and attach them to this issue +sorry for the delay on this I somehow missed your commentsI assume by setup routine you mean Before +Other than that it looks good! Thanks! that is good +This issue was avoided by the removal of the reboot capability inWhenIf we look to this capability the behavior and symptoms will have changed so Im closing this JIRA. +Ive checked with dependent uncompiled jse ant based projects and it seems to work there +Pushing it in +Fixed +I attach a patch for item +The javadoc warnings from hadoopqa are valid though +already fixed this issue +Dmitry before making any changes to the patch I suggest you check whether the classes provide everything that does +Does anyone from Apache assist with the mailing list? I do not have the time to waste a week waiting for someone to assist +You are right logging should be improvedI really believe that one of project libraries contains corrupted class file which cant be read by ASMIn the meantime as workaround you can simply disable bytecode analysis +It makes it impossible to change it dynamically which is not necessary for currently envisioned operations other than small optimizationsIt makes loadingadding files much simplermore reliable though +I believe you are misunderstanding the library design +Each components monitorRunnable is scheduled to run immediately each of which schedules itself to run after seconds +Cascading is going to allow for comparators to be passed in +HDFS makes every effort to chunk a file into blocks of size specified by blockSize +Please add license to patches with suggested fixes +Also see duplicate cr +Will patch for javadoc findbugs warnings seems to be normal right just committed this +Update the patch to address comments from Aaron and Nicholas +allI attach another patch to make the dojo style looks more prettyRex would you kindly help me review the patch and commit it? is slow wont fix this. +New available based on the other store samples +Hopefully by tomorrow +Currently we are working on a solution +Ive applied the patches but unfortunately some tests fail +Once youinternet telnet gwexternalip works then you just need to use gwexternalip as your browser socks proxy on internet +Do all new tests pass right now? How long will it take to run all tests +Id like to keep it in case its needed later +Lets commit that partRe +Well keep you updated +That is broken +see subversion commit tab for locally before commit fix +Bulk close for release +Yes it seems the editor is acting glitchy +Moving it +Not sure what this is trying to say +This work looks really interested +This thread is a system thread and has nothing to do with LuceneExactly so we shouldnt be starting it in our code +I just committed this +Oh I ran this on cluster to prove that new way of doing memstore size calculation is not going haywire +Hi RafaelIve posted a PR for discussion purposes +Yep most likely fixed in cceabcdafaceccc +So on later +Everything looks to be done for the Harmony TLP move. +Added an overloaded method to take a set +Have you tested your code against a SNAPSHOT +May be should be done in +There is no Ignore in the to remove +Looking forward to more applications to commons math and other commons components +Thanks Ning +HDFS proxy was removed hftp is documented. +Ok I see +bulk close of resolved issues as part of JIRA housekeeping for release. +Sorry about that should have given a heads up on that change +EJB +committed spec jars +Fixed in trunk r +As a first step im going to write some tests to cover the current code +I tried the attached example and took about minutes +but of course the folks at Cassandra are free to choose whichever way works for them +Size attribute was added into encoded sfileUpload template +Dan already implemented the workaround in the pull req for +This issue has been filed in our bug database under the original Bug ID +we support paged search now. +Ah one thing in Java passes input to the parent ctor because all synchronization should be done on the source reader +I agree with Jorg the should be have the ability to make partial release without touching the workaround for this issue so far as been to use profiles in order to choose what module are released +Added a similar fix for uninstalling resources in rev +I agree with Giovanni there is still a bug in Wicket +I think we should go back to write version and you should also use the found in because it fixes a serious bug that with large arrays hasGiven that I think we can finish this up and call it doneThanks +Please resubmit as feature request if you still need the functionality. +applied thanks +Let us track the error with init method separately with +We can go back to our patch review process again +Changed the title and description to use the name +Thanks this has now been fixed in trunk. +In the edition mode thats enough Stas +The piece of code you have adjusted in is only dealing about Embedded objects it seems +should be all fixed now however if you switch from hibernate to openjpa the first build after the switch fails but then things work fine subsequently +not the branch +seems descanalysisengine was deleted since r is it due to this or something else +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Otherwise there is always the slight risk of changing an operand that is not intended as a there other issues here +just committed this + added emptytrue on Connector elements in should use this value to add correct session cookie +is minimally maintained to serve a few other Apache projects that use it +Added complete DUCC source code base to the svn +As some of the used fonts doesnt provide a mapping to readable characters some parts of the extracted text are unreadableThanks for the report! +Please verify the bug has been fixed. +The previously removed path appears againJBT of the preference dialog seem toget the right state back just punting to next but marking as blocker since its confusing +I found another bug when I was working on +Thus if we remove one of these an example needs it wont workThe server runtime adapter list you show are things that works with range of servers and versions thus they do not match up since very different conceptsAnd yes as Snjezana says what runtimes is it you are missingAll the current releases are there +This patch will switch the webapp into the project modelGiven our proximity please review +There is no anymore +I am currently in a process of debugging it in Jenkins which takes time as Jenkins is under significant load last daysI have talked with Vladimir that he will try running the Java ACC locally meantime and I will then try to modularize the script to run in Jenkins in our automated setup +But clearly were in a world of behaviors here +Commenting the log mover for is an option for the interim but would resort to it only if there serious time pressures +push out to +The difference could be that the ANTLR isnt back tracking properly and that path might have changed with the inclusion of metastoreCheck command +You still have not explained why you want this +we need a changes entry but from my side this looks good +Im closing. +Many thanks Jochen! I see that another has verified the fix inAny idea when well see released +This one massive issue has not seen any progress. +One comment it would be nice to have the tmpReader logic be triggered when such a flag is passed in the constructor of Reader +We need a NPE check at + +Attached patch prints additional information to FSCK report +And I think youre on the right track with the side but Id break that into more sections like the editor +by the way it means that currently the consumed memory is proportional to the data inserted in the JCR +Im make sure to do that for future patches +v +on the mail core ofthe entity doesnt have a name and neither on the solr mail tutorial +Hi Raghu I am perfectly ok with the fix you are proposing +Were fine in now. +When a null or is registered with the SAX requires an NPE to bethrown +is a better version +Patch includes quite a few unnecessary white space or format changes +Because long character types are not the best thing to use in the most important index in your database +You might want to find as the release and as the release +we could also key in the code for n too but its a lot more and it would be better to figure out the smallest representation in java and make a code generator +hmmm the whole way cobertura instrumentation had been implemented is a little messy and with this fix I have concerns about GPL leaking as cobertura is always on the classpath even for tests +User can have different actions attached to these events especially for assignment events since they can be fired many times +Will submit patch after the dependency is checked in +Let QA run +This issue is fixed and released as part of release +This issue is identical to +Unassigning so that a committer can pickup +Well leave it as implemented and Ill port the changes made here over to the EAP streamThanks for your help with verification on these Messaging issues. +Is there some functionality missing from this new way of starting geronimo +Hes your friend Q can we have this by or should we push it into +Created project now builds OOTB +I think the real issue here is that the view table name is generated with the full file name +Patch works for me also. +I suspect its a problem with the typename keyword but theres no way I can prove it +Does look like we improved some of the cloning costs though +I think the fail tests may be flakytoo +I missed svn add +SGTM +Hii have the same problem +This way after unmarshalling the Binder and DOM document can be discarded while the necessary expression objects would still have the namespacecontext property available for expression evaluationAnother enhancement would be to create custom coded JAXB classes for the lax types that contain content that needs to be unmarshalled properly +I will also provide the enhanced parameter features to and the Xalan command line executable +I cannot this with my test case. +It sounds dumb but building Hibernate from scratch is not all that straightforward and even if you know how to go about it takes quite a while. +Committed to trunk. +Resolved as a dup of +If you finish it before we release beta please correct the versionThanks +This looks good Daryn +This patch is intended for commit into branch after append code is committed +All done added support for the Enter key and Tab already worked +The tuple source sharing logic caused the nodeid to not be propagated which is required for the reuse logic +I dont see the original being changed but I may be missing you confirm it is broken if you redownload the patch you uploaded? maybe there is a local difference +Thanks Tim added messages are as expected + + Attached +Thanks +The feedback from Trademarks was to include it on the site +I agree that I dont understand the difference in sequenceperformance on trunk it was puzzling at the time +Also keep in mind that you can use a set of parameter file for the same queryThis way in order to add another cluster you only have to drop one additional parameter file +One good approach to verify this would be to run the test with reduced memory heap +Will close this jira once the corresponding JON issue is fixed +Provided there are no critical issues we try to have a release every to monthsOleg +lgtm we expect backport of patches to +Which hudson do you meanThe new one from Eclipse +Closing +However the provider is close to stateless as all its methods require a instance to get the contextual information +Will file a followup JIRA +The same bug is in beta +They are either still on or +Its been months since that project so I cant remember too clearly +If you are not sure how to integrate your solution into you should just attach the and well work together to integrate itIMO this would be just a temporarily workaround +Good point keepRunning is a bit too abruptHow does this work +This seems to be a design choice made by Sun for scaling selector for handling groups of thousands of connections +This is just the diff file +Thanks for spotting that +Agreed that if nobody is interested we should not spend cycles +Easy workaround is to use an explicit join btw. +Yep addressed in later minor version of +team please resolveclose this issue as you see fit +Based on that we could simply keep a set of external files we could read and execute +Shalin did the changes in the latest patch resolve the issues you referred to +I havnt had time to checkUntil we can confirm that we are symetric I would rather keep this open as a reminder +But they are not watching this one +Username rafciom +This includes adding unit tests to verify correct interpretation of the namespace and also proper behaviour with respect to new configuration options availableThis closes this issue +Committed to trunk. +It might cover now but next year this wont be the case any moreWhats worse the JDK it finds might not be the one that you wish to build uponTo simplify the things I dropped the entire guessing mechanismJAVAHOME or is more then enough +This issue has become important to us upgrading to blocker for some reason this patch causes reducers to hang when fetching data from the mappers +Updated patch for all services and tested it on AWS +While I agree with you that the specification states the two fields are mutually exclusive I would say the specification still allows for the existence of the field with empty content +Thanks Dag +Is it possible to use that for this purpose +this will update or replace the configure email for purchase orders +feature is now provided byOleg +Its a technical limitation the proxy cannot be crecreated anymore due to missing information +Very nice could we emphasize the infinity between the p and a? It can get a little lost +Derbyprovides a couple ways to do case insenstive searching that uses indexeso setting database wide collation to be case insensitiveo sql generated columns +currently has tests and this really helps preventing us breaking thingsThis has been a serious investment and we profit from that every day +Shouldnt MAPPINGFILE be added to as well +Is there anything else I can try to make it work +Verified on Windows XP with +Justin spot on it worked! Most of the changes are in geotools so Im going to open a jira are the changes to be applied after the geot patch is applied +Two small comments on this patch lets not modify equals and hashcode methods for now since it adds messy code please use spaces code indenting instead of tabsAfter those fixes it looks good to apply +did the recent confluence upgrade fix this for you +This has been checked in nowWe still have to fix or replace Jansson for a build to actually work on Windows +Where can I find the useable jar +We currently have about M projects on the builds causes no memory oftenly when you click on Show projects the JVM takes some tenths of Mo that are never given back until it comes to the Xmx +moving open issues in X DT to X DT as discussed with Bob +I have applied the patch to my working branch +I am trying out generating the site +It doesnt throw this error for meCan you provide instructionssamplescode on how to reproduce itwhen it happened platform OS etc +Marking bugs with invalid fixed for release as fixed in. +Fixed with SVN +can you comment on this behavior +will fix while notification +Good stuff +This bug also causes a cacheTest to fail in UI example +code looks goodTwo minor comments +Patch applied passes tests preliminary testing by reporter passedPlease reopen if further testing again now properly recorded against Fix Version as on a patch to layouts that may be the root cause +How about thisSo Id like to separate output folders per subproject but currentlyI have no idea for realize this +Patches applied fine +Ill look into minimizing the diffs after your next patch. +Also logging operations are removed from +Wait dont we have JMX for that +bulk close all resolved issues in preparation for and releases. +I just committed this +The behavior should probably be configurable though in the case where we legitimately do not want to flush children unless our buffer is absolutely full +Cool +I am not worring about additional attributes because launch configuration file rewrited every timeJust be aware that today users can go and rerun andor copyedit the generated console configuration +We should not release Camel with any SNAPSHOT deps +However the same strange result is generated when using twiddle command +Is that correct +XavierOK but I cant see any UI to edit Organization Accounting Preferences did you use a loaded file? BTW which version of Opentaps are you using +That would be great +How can one reproduce this bug? The query parser produces a normal term query for numeric queriesSomething like this does not cause an error on trunkhttplocalhostsolrselectqpopularityhltruepopularity +I take a look at it +However Im having a hard time imagining any meaning other than all ports on that host +Have post another solution +The good thing is that SCOPE is a separate protocol and so no code changes to NAKACK or UNICAST are needed +Bulk change of Fix Version for all issues fixed or closed for +Id rather do this to be honest as Castor allows for object references for map values and there needs to be some validation that such a reference is present upon unmarshalling +bq +Committed patch with some tests in r +See AllI am using +Thanks Hairong! +Verified in Beta for proper fix in b so setting fix version to that. +The class mentioned was moved to another package +Attachment has been added with description Overlapping J and o +I implemented octetlength +However this plugin still throws the internally. +Better try with +Is it possible to add to the + +Covers backwards tooI will work on the trunk version now +Your previous messages has been captured by my spam filter +Is there any harm in doing what Lars is suggesting? If no why not commit? I can imagine removing a dir locally for whatever reason and regardless of whether Im working with a svn checkout or downloaded dist. +Changing the fix version to since this is not critical for functionality and is more of an optimization for larger numbers of Session instances to the same workspace +So I take it the incubator release is no longer holding this back +I was talking with Astrid today she was really looking for a way to shave off columns from the feature type configuration +Could you please have a look at the forum thread? I commented on this issue and waiting for your feedback +If etcservicename does not work then it is an initscript bug something else or a bug in the LSB functions +pisoa is currently using so its possible that the caching properties only take effect from +True +I cant add or tothrows. +Anyway Ill run some tests and let you know +We need to analyze what the benefits and costs are before we decide to do this +All of the unit and integration tests pass and all of the examples and their test cases all pass +will commit momentarily. +Do you agree? If so then I dont think we need the patch in at all your new version of load should take care of itbq +Sounds interesting +Breakup of sort and combinespill timings is attached in +Attaching a patch implementing compression emulation support in Gridmix +Attaching patch +Sam fixed it for me +Ah! I wasnt even thinking about it from that perspective +Backported in r r r +fixed in reopening the h headings no longer disappear but they dont appear in the table of contents +Not enough information here +Thanks for the review Stack +Committed as r thanks Sijie +Is there an alternative toIm a big fan of minimal dependencies +Can you try the code from Git +Reopening per my comment from earlier today +Replaced by concept. +At first glance it looks like Derby thinks a page has been persisted to disk but somehow it didnt actually get written cf +Another fix for releaseaudit target +I pass this patch to hudson +New version of the patch addedChanges IDs are now handling now separated from to a dedicated class new class extending the old Enclosure class +For links in the content that dont use the site indirection then links are not made relative +Those attributes do not have project properties substituted +UT was also added. +the problem should be fixed in the current master or releases and later please report back if it still valid +we should create separate bug for wf name +Fair point +We place an upper bound on this count +verified on local build from on JBDS +New version + +This patch simply sorts those executions think the original poster wanted the old functionality back where plugins could be declared multiple times and run in order however Im not prepared to undo that on ve applied the patch to resolve the issue +Accidentally uploaded the wrong patch first this is the correct one +Never mind as it is included in your first patchthough +I am not sure with the implications of excluding lib +Im going to next look at and then + +I reattached it granting the license let me know if I can help with anything else +Thus if your application registeres a newcustom one it would likely override the one hooked into the configadmin stuff +The second type the definition is processed and added to the Map of typedefs it simply overwrites the earlier entry +Its not a simple solution +fixed the issue with revision +Last try +This will not break rightAlso I dont see set at all in +Having said that I think we should add Maven profiles to allow easy execution of those tests on any other database +notation should NEVER fall back to the property access to avoid confusion +I also agree that the current behavior is weird and not as expected so a rewrite to the way Windows works for example would probably be wise +Thanks for the patch. +The error is thrown also using only from JavaIf it may be useful to anyone Character style is referenced but wasnt defined in the rtf stylesheet and the java fails because of this +It can cause Jenkins to hang because of default password entry +updated patch to trunk +It works reliably for annotations and string method names but the parameter passing is a bit trickier +Kurt can you take a look at this one please +Fix NPE in bean service wizardFix title in service test wizardAdd custom Maven repository for repo +If you dont we have virtual user tablesMy view is that the old Alias and Forward support should be removed and replaced by a management interface for the virtual user table system +Other issues not relevant anymore +If you are not using our shipped Jetty your probably not doing it right if you ask mebq +Just a few more little nits Recommend sverifyverifyg Why not make instance variables in be private +OK I agree its not a BUG lets reclassify it to a wish +I wish I could provide you with more info but this is handled by Android Webkit and not us +Thank you +I will reject this Jira +It is done with the +Due to time constrains I have to post pone this to +try it and lets have hadoopqa see if things break? Id imagine it would be ok because it would have nped before +Hi chris snowI tried run your code +Closing all ResolvedFixed issues lacking a Fixed Version +Missing api javadoc was for stuff as explained gapi not excluding private members I looked into it a bit and groovydoc tool doesnt seem to be dealing correctly with the modifiers at all when it creates the AST for the sources +Thanks for the recommendations looks great +Oops +Other then the Deque and I am +Fixed in +Thanks Ill come up a patch +Its still a unique request +If we say lifecycle is a session then I expect to get that object in the next sessionIf the cached object is going to OOM the JVM then clearing the object from the cache and then it from scratch is also going to OOM the JVM right? All things else being the sameId be spooked if I said lifecycle was a vertex and the object disappeared after the task finished because GC kicked in and decide to remove it +That will of course not work for the we can of course display a message in that case but i guess it should be possible to ensure that we return valid json +Well the default precision step is in Lucene +Focus is now on PBR able to resolve the problem of IE and Chrome not working with set to ALL and WRAPSCRIPTS true +that illustrate the NPE the file cd into and run maven multiproject +correction warwarresourcesearresourcesm running from latest svn and I dont see this in there +Thanks Varun. +Logged In YES userid os compiled for Java you want to use with Java you should downloadthe source and rebuild using your sdk +Disable flush and compaction logging by historian +Is it fixable? Or is it better not to apply the fix at all +We can always this issue if necessary in a future release. +So changed that +Only issues that are actually fixed should have a fix release. +I am using and I have run into this issue +Extended version of the unit test which shows the problem +We will also at some point provide encapsulated native query support +So its faster than stop and start since it just restarts the configuration instead of loading and starting +for patching +Both cases work +If it really is in Java Classpath then I wonder if we cannot just access this code directly +This one will be fun +This is not going to be done prior to +Many of our users leverage the slot millis counter as a metric for cluster utilization for jobs +metadata +This is the solution +See what it thinks +It seems a vote is necessary to resolve these differences +I havent been able to test the performance improvement so I cant judge what it might be +Thanks Isuru for the contribution. +However there is no such customization machinery available in the VM yet so as a bare minimum we should keep switch +oops sorry about thatFixed it in revision tests are in place and part of the suiteResolving this issue as fixed. +AdditionsNow upgraded to requires and Working on an object data structure to capture more information on dependency relationships +This will break BC so requires a major version bump +I just wanted to see when buffer grew +Normally it is probably fine to leave contexts around +Ill merge back +rightwe need to batch update all fixes on to but i didnt find this feature on JIRA so still in embedded in both webprofile and introdoces itself as not +Hi Andy i would prefer the patch added with Grant license to ASF for inclusion in ASF works +Someone to work on thisve tested and can confirm that both new Mojos work well with relatively big someone else please try tooIs it possible to get at least mojo included in? With it and with existing clean and instrument mojos one can have more complete control over what and when gets instrumented run tests and then get a coverage report only without rerunning of tests just for report +Looks like it is safeto apply the patch when you get a chanceThank you +Hmm Im still getting the heap space issue I think its always been somewhat intermittent sometimes it doesnt happen usually it happens when you run all the tests sometimes not though +Thanks +Dont paste changes directly as commentAlso verify that all tests pass +Therefore I think this can be resolved as duplicate of +to be clear out intent with the meclipsejboss tools maven integration is that importing the as a maven project should be enough +looks like you caught the bottom of my comments others still stand on the latest patchMy latest patch was to fix the test failures and not to address your comments +The IT test passes now +Good idea +Thanks for the suggestion +Closing. +I have the same problem +Other issues found will be tracked by separate. +its easy to build the bnd code with maven you just get it in an appropriate structure +It should be possible to make the case for +ThanksIll let Luc revise the contents but I already noticed that the patch file contains tab characters +Closed again now properly recorded against Fix Version as well +What do you mean by validation exactly? I guess this is not distributed at all but rather happening locallt? +I moved this Bug to Grails no longer supported in this plugin. +Can one of persons that have jira admin rights please give me permissions or set this flag +I forgot a couple notes about the test application changes Ignore the CNFE exceptions I didnt want to take the time to chase out all the optional Seam dependencies to resolve them +This sort of targets the Emptier thread run at the noMakes sense +Hi Thejas Ill take your suggestions +And since on systems where the is installed it uses the and the SASL library instead of the the EXTERNAL mechanism doesnt workI prepared and attached another patch which fixes this problem as well +Now this should be able to stand alone and be committed if deemed acceptable +Patch was applied to Apache HEAD CVS +The last issue may not be a real problem as I cant think of another at this moment that wouldnt also be subject to heapsize it works around this issue and this issue is going to pretty important for when its used as an enterprise repositorys backing store +I suspect that your multicast is not working +Does the existing parameter not meet this need +If you have SDK questions contact Thanks! +When not required it is simply a waste to save state redundantly when there is no difference across the views saved for a page in different user sessions +AndreiUsually tests are performed with which allows you to run your tests inside an environmentI could try to put some tests there +I agree that a check for Null is not a good idea +I mean to tell it to bind to NO port and not allow any remote invocation requests to come in at all nor to attempt to reserve any port +This patch is to fix it +Same as the previous patch with a small fix for the unit test failure +Alternate suggestion +The information comes from the request header if the contains we haven an ajax requestYour right but you working with permissions and without the auth option +verified +Yikes just realised there is a README too in the yarn directorybq +Thanks for the patch and the backports JimmyCommited to +Ola I had to revert the rev since it breaks jirb on posix systems very bad making the terminal unuseable once jirb exits since no echo on the terminal +Whoops here what do you think good to commit +Columt Type should be added in list between Version and seam and jbpm runtimes be created for EAP and accordingly +thantakes the path of the request and stores the cookie with itGET onWFS dwrequestidiGETonWFSenUSUSD dwrequestidiTherefore both cookies are different +No new issues without going through Kevin +Fixed in fixed in r +sorted and committed revision Lewis +forget to mention that use java JDK for this plugin is if reported against the next release using a compatible JDK +Hi PierreAs i described in the email +Review Board not necessary +The information you mentioned can get derived at runtime after parsing stage just right before execution +Some are checking the return value and we are running them on Windows +Hi MirkoThanks for your patch +Stefan we should remove the usage of in +They also use direct byte buffer +Someone will need to set me up with the CVS details +Here is an update patch +I dont think we are going to introduce new interfaces or abstract classes in +Committed to trunk with revision +If Stack has completed its review then no prob pls go ahead and commit +As for replicating I think that should probably be off by default as it could cause complications but it would be needed for a cluster configuration so we would need to support it at some point +But probably it doesnt make sense to worry about that at this stage +It overwrites! and it is the expected behavior its not a should be executed just after the to avoid this problem +What if I want to change the formatting output? I have to go hunt it down in three places in the three different routines that are doing the printing to change it rather than make changes in one place in one method. +Fixed the naming part at least lets open new defects at github issue tracker as necessary. +Mark right my mistake +Patch adds helper methods to utility to allow removal of regions during tests +New stack trace for revision try this updated version +Tomorrow I will try plain old java serialization +I have an early implementation of Hadoop RPC using Boost ASIO that I could probably throw on github if people find it a useful starting point +Jan thanks for you reply +Thanks Gary looks good +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content Services +Awesome! Gene splicing gone wrong never looked so good + +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +How does work did you add a package private method to update the requestsI think I have a preference for as it keeps the Context design immutable +Please switch to Link pull request state once you will want to push it. +It might be OK to break compatibility +Right Ive never had one reported +Yes now you have setup you can go to the page I gave you the link to and use it to created you entity models from your foxpro db you have set up in the bye default they will fit your since they are taking them from the DB directly +Hum forgot to include them +Seems we can drop get +Ive attach the file Im using for performance benchmarking +Add a new unit test +Verified on revision +Moving to since this is only an issue for those with large numbers of procedures in their vdb +Canceling patch to correct and resubmit +Will you like to work on this? You can reference my patch at for how to do repackaging easily +Should I submit it as a pull request +No i am actually using i just used the reference to The bug as sample what The jacoco Team thinks causes such issues +Christian do you use the same profile for every project? This plugin mesures the activation level of the profile of the project +Didnt run the test enough times to ensure all codecs are fixed yet +Having the management client or user know that much about the model doesnt seem right +The meta data is created statically in +you might want to refer to the setup application in the commonext component you could even the help screen for that +Further inspection has also revealed bad links within the site source repository +bq +Fixed intrunk r r r r +Adding more features without adding more people to the project willing to maintain the code hardly helps +Thanks Tim +Basically it could still say files changed if no resources have changed in teh WORKSPACE +Can you include a unit test for that verfies the exception will be thrown? are a few comments that should be stripped out including a printlnThanks for looking at this Jim I have addressed most of the issues you mentioned and right now Im just some issues around the unit tests as from some reason the way Im initializing the scdl load is failing as its not finding the loader for Id appreciate if someone could look at the path and give me some ideas on how to properly process the scdl in a unit test +closed issues for this done when integrated with console +But it looks like there is a different error path +Created for the English doc problems +It would be great if someone could this patch on the platform which shows the problem +Do you have a fix for this Clint +thanks Kamil! +triaged for without a repro not much can be done +This is a list of issues or hacks in the current implementation +new patch looks good +That was just to avoid the JIRA workflow resolving this issue +for the to trunk. +Whenever a container fails on a host the host will be blacklisted and will add to blacklistAdditions if blacklist is not ignored +However I fixed the class by overriding testPoint instead +You should either remove it in favour of the install or have it call the install to handle the deployment +None of the Vectors used in that class have a life span that exceeds the method in which they are created and they are not visible to other threads so it should be safe to replace them with unsynchronizedAll regression tests passed +unless we want to enforce the checkstyle plugin +Ill send you a patch to review right now +Query log with explanation of what is have reached my mailbox I will be out of the office until Januaryth without access to emailsKind regards +Once we do that the actual AM can pick up the credentials just like it does in an managed AMI like this idea because it keeps the AM logiccode identical in managed and unmanaged modes wrt tokens +why is the status Patch Available when there is no patch +Will upload a updated patch that incorporates Amareshwaris commentsThanks Amareshwari for the commentsUploaded a updated patch that incorporates all the comments +Committed revision on trunk +Fine there is a page but obviously I could not find it there are no links to it from the Home page which is a pretty big omissionThat needs to be resolved I would say at least a More news link in the news panel and really a BLOG item in the criticial though is it +It also includes the fix for. +I will create a new dump and post back with the location of it when that is done +Then I think the problem areas would reveal themselves +Attached fix +In particular looks like the same thing as +See for resolution +This jira is unassigned but contains a patch that needs to be reviewed for inclusion in +I checked the site today and it was updated +Well cary that notion forward for the tree and address this larger problem in trunkThoughts +Sigh +Patch verified. +The code I committed doesnt have the quirk where youll get a proxied instance when you issue a query where you only are selecting a Calendar field +Ill file a new issue to follow up on the suggested approach +off of and +What criteria would be used by the dynamic enhancer to discover and enhance this class? If its not being done by the javaagent when would that discovery be performed +ok +This patch also adds a number of unit tests +Julien i finally found the difference between our environments i am using and you guys tried +If the problem persists please reopen +But it may require more refinement therefore I appreciate any commentary +Oh and I think you should also change the CHANGES entry you added on +Please verify +There is no need to exclude anything for cpa if I dont exclude those currently excluded theres at least one test failing +Can I move forward with committing this patch +Continue to have the legacy code for parsing robots file +Thanks DarynAlso thanks Tanping for the review +to trunk and branch. +The sitemap engine doesnt care about xmlnsxxx attributes that may have leftby some previous processing +I also wrote a test to verify that signed int arithmetic behaves the same as long arithmetic truncated to int +Yes it is a Signavio issue +Attachment has been added with description Now the PDF! +The snapshot of has been updated in the repository +This sounds good but lets be careful about making things clear for the average user for instance seems a bit vague and nondescript +Heres an updated patch which should address your comments +I will post this to the dev list but we need a place to do inline commenting of patches +is probably being a little eager in picking up resources +We need to create a gateway that behaves like the others part of the lifecycle etc and exposes the UDP functionality as part of know that you have looked at netty so one solution could be to add support for a +Heres a fairly good stab +There are a lot of custom one do you mean +I just copied the code from Yoniks method from Solr my original code donation back in the past used float +A test timed out probably transientUploading identical patch +I dont want the test to fail if the person running it doesnt have a that supports it bc that wouldnt fail in the live case eitherThoughts? Id like to close this one out +this patch add license to java class which hasnt apache license +Would this then cause a duplication of all dependencies? Would you still produce a distribution where we can just point a runtime to it +And how would you suggest we implement such a feature? Is there a JIRA plugin +and at the same time we provide the access for the old web UITo be more explicit the experimental UI needs to show a working hyperlink to the classic UI without requiring js being enabled +bulk move of all unresolved issues from to +all scopes EXCEPT system will be forced to runtime for dependencies. +Heres an updated patch which should address all of your comments +Then when writing if the index gets too big we can downsample at that point discarding entry or somesuch to keep the index within a certain bound +I built a project using the aj compiler plugin and got some compilation errors i ran the build again right away and it said something like no sources to compile even though half the source didnt compile the first time and I would expect to see the exact same compilation errors again and again +updated hard upgrade section +If the base model is never actually used because the factory always returns the injected version then to be honest I dont see what benefits are gained by not touching it directly +Ill make the same argument i made a long time ago but i guess never wrapped up this limit should be MAXINT for explicit stuff! Yes the more BQs you explicity put the worse your perf will be +Thanks a lot for this quick action +Feel free to modify +Now camel ships with own managed resource classes in core module +If something is missinglet us know asap while we still have zfs snapshotsto retrieve from. +Change to Release Notes made as a result of feedback from Marek in JIRA +Hive calls get and get just to report the number of mappers and reducers +most likely over the weekend. +Support for providing name at API level is already provided the remaining part is to invoke it from UI +Presence of SNAPSHOT where exactly? Could be a dupe of + +Ah I see +though Arjen is probably the one who should make that decision +Lets commit it and see what happensBTW how to announce the extension? I think I havent done that yet +Committed the patch +We could probably just do a JNDI lookup for this +Thanks Prasad for quick turnaround on this one +lars I hate to go backwards a bit with how we do testing but this seems like a reasonable case for doing some OO in our testing namely creating an class that we use when writing tests that has things like this +hmm this is the code ported from hadoop commons right? Well spotted +cordovalibs version +trunk Committed revision branch Committed revision . +Ive just committed this +I intend to fax the ICLA later today or tomorrow +Given that is no longer maintained Im closing this issue as wont fix +Test passing hanging on winxp and some that I only half committed this other half in of the patch was properly applied +Maybe something relative to HDFS then +Ive just committed this +SergeyAfter change to org the just goneIm still running the test more times to ensure the message lost problem get fixed since the message lost problem not happen to me every time before need more test nowWill get you postedThanks again for your help SergeyUnfortunately Im still seeing the message lost problem even no exception thrown on server sideThe fix absolutely make things better since I rarely reproduce the message lost problem now but the message lost still can happenRegards FreemanWhat is the value of max that you set for the test? Suppose a loss of messages starts withmaxI believe that practically it should bemax knowing that is that dangerous limit where the loss might start occurring +In my cluster ROOT and +This is all I need +That is why I find the problem a little bit +Resolving +defer to +Cheers +If there are no objections I will commit it +I think I messed that up I selected fix instead of affected. +Of course any deployment will ultimately end up in a container duhTo me any use of the term testing is completely redundant within a TESTING frameworkMaybe something like extra or frozen or arq +The patch works for me +Yep closing. +I can not reproduce the bug on my Windows XP +Attaching a test case for this bug. +It is up to the tasks called with the targets to not do things if it is not necessary +Attachment has been added with description patch for the has been added with description Small test has been added with description PDF result after applying the has been added with description Updated patch added support for and ipdauto +Could you try this +logs for the D computation +Will upload one soon after I move the test under lucenecorebq +Thanks Paul! +Could we try to implement this for? +Hi Trung Should we resolve this issue too since it arose as part of a workaround for? Thanks +appears to fail in all +Hello DavidI agree with what you say and yes we can use your suggested way but this tag is also present in the services so a point came in my mind why we cant use the same tag in screens and forms +But did you also fix the that the displacement of a graphic is actually rendered? Otherwise i would reopen the issue +Committed to trunk. +Ive committed the patch. +Committing +I have made a start should be able to send a patch in over the next day or two +And yeah we should open another JIRA for wrapping curl calls in retry logic. +Sigh messed that one up +There is two ways Completely replace current implementation and take care about upgrade Leave current as is and add new which can get configuration from you explain why you want to make the Sonar plugin a +We need more info +First of all JBDS has no update site yet so there is no means to update JBDS this way +I think the was renamed to +Added screenshots for Ubuntu JavaDisplays nicely fonts look more clear now especiall bold fonts +there is still a single class but underneath it would be multiple My approach is different from the aboveThe new interface should be general enough that can be implemented without the requirement that have multiple writers +Not sure what the cleanest way for us to get this retry logic across would be +Let me assign this jira to you +Note this seems to effect both the plugin and whatever version was part of install as I get the same behavior +Right +We need to just remove our +is missing the is method need this implementedEJB still works flawless JMX testing is held up by this currently +No I am working on it but it generates a different error on master +patch forupdate removeChilds behaviour avoid to remove the installed application with the same context +If thats the only log you get so far then the node is waiting merkle tree response from Check if you have any error on that node +verified at r +The workaround was introduced by and is related just to JBOSSHOMEserveralllib +Looks goodThanksMayank +Go ahead and reopen the other issue for this then since it will be committed to. +When system and response encoding differ result destroy non latin charactersRegardsMarek Sikl +Closing all tickets +Funny since cookies must start with Version +Any suggestions please + +Olga +Two comments +Part of a bulk update all resolutions changed to done please review history to original resolution type +Else Ill call it out and I expect the others to do the same for me +It doesnt assumes that score means score anymore +I know performance in groovy is an issue with loops and math +Something like would be more useful on grids where people use pig plus other tools +Verify ifhow both issues are related +Suggestions +Changing it around may bite in unexpected waysRegards the internal comment it talks about calling beginOp then you must call +Regarding the docs were working to update them +Michi just it for a patch +The second one is an enhancement and the first one is a bug +I believe you compared the performances of and +Youre both right +Again adding JAVAOPTS as DEBUG to will make it visible in out of the box just like system properties +Completed initial pass through +Q Do we move the nsis installer scripts to repository as wellA They stay with the they only bundle a JRE if the build machine has been configured with one +Looked over the code +We never did that versions stick to the Java version they were developed with +checks for any running tasks on the +Increasing the max in the policyEntry seems to help the issue but eventually the consumer on a separate correlation will lock upAny ideas this is has prevented up from upgrading since +See transactional client wiki page +The delta from patch to patch was Committed revision Note that the change for for get was committed as it was part of patch Can you now mark issues as resolved if you believe you have completed them +also we should understand selection in visual part but that should be different issue selection in visual partselection in source part it select the same source nodes but with different presentation +Lets scrap this ticket you will never agree that exception handling during repair is broken and fix that exception instead of exception handling rightI am not saying that fixing exception which caused this hang is bad thing +Thank you Alexey. +No we dont have the example as Werner have applied the patch by hand and commited it +A seperate Jira will be filed about changes to changes to support generic sources for data and checksum so that most of the code can be shared +Thanks Sergey! +Verified on latest branch commit ecdcbffaccbcfdaNow able to create LB rule with inline mode deployment +The NPE was fixed in r +sshidrsa cd PROJECTDIRECTORY maven sitesshdeployshould contain newlines +Latest patch doesnt produce any errors for me with either lucene or solrCool thanks for testingIll continue working on finishing the conversion +I tested this and on my Ubuntu the Menu item is there but grayed out +I just applied the latest patch to trunk and I dont quite understand how the numFound in the response list is computed +Also allows for quicker bug fixes and granular upgrades when issues are identified +I rearranged the wording a bit so that the prepositional phrases would be closest to the words they modify +I think this was a upstream issue that has been resolved in the meantime +Turns out that the registration of protocols does not work properly unless the is loaded by the root classloader +Any idea when the version will be released +documentation issue pom snippet fixed in configuring the dependency as a dependency works but its not what we the dependency as a plugin dependency doesnt work either if the checkstyle plugin is called as a report neither if its called using checkstylecheck directly see a Maven fix +If it doesnt exit normally then the SIGTERMSIGKILL approach and would be relevant +You can get them for any svn revision but they look like release tarballsDo you mean that they end up in the wrong place on x? Thats strange +Basically only works if getPortlets has been previously called which is not always the case the first time check is invoked +I looked through the codebaee and could not find the source of the generation +It also includes a unit test +I dont usually like greclipse producing different errors than groovy does for the same code and groovyc doesnt report that as a problem +the same JVM we need to connect to some kerberos secured and some unsecured clustersGood idea! This provides another strong case to validate the change to support multiple clusters for client +Thanks SamI did not test but this looks good to me +This stuff compiles with the method commented out +this to dev as this is on the usage of this +OK closing as Wont Fix based on Peters comments and the fact that these entries can be explicitly excluded. +I send you a new update for multiple forums for forums adding modifies about guest +When I start progress I can give more now what I can say these ws preference page reused jsf module codes when I create the ws module +It is the most common so perhaps you could start out with just that one and wait for people to notify us of others +An early patch having +To review this patch youll first need to apply the path for since trunk builds are broken in at the moment +Sorry if my posting two separate patches lead to any confusion +Needed to add dependency on rpc into BobI assume that request is a pretty container specific implementation +oops I meant +Luke is right that the protoserverid is used by to form which currently has no bearing on the authentication +I of course know that the purpose of is not to automatically generate a fully features ERP but I do think it should be able to offer CRUD for all the tables in a data model regardless of its complexity +Thanks Mark I will look into the issues youve listedI want to keep this ticket open for if this is ok with s RM? +The resource leak was actually due to the fact that the sender doesnt close the JMS connections when the transport is stopped +So you should correct them accordingto your workspace +I thought it was fine the way it was +Text extraction rendering and printing work fine. +Add a test that invalid flags value is disallowedAlso tweak errno on invalid flags return EINVAL rather than ENOTSUP +Committed revision A big thanks to Noble for designing these cool features and to Grant for his reviews feedback and support! Thanks to everybody who helped us by using the patch giving suggestions and pointing out bugsSolr rocks +you just end up with the +This bug only manifests when wrapped connections are not +What is scriptomWhy are you asking us for samplesAlso your username was stupidly named and confusing so I deleted it and created a new one more appropriate +That is what I have done +Commit +to +Slightly updated version +Will check this in +verified at subversion revision +Moreover the logic of different distribution and packaging conventions would then be spread across multiple projects replicating logic that might better be consolidated in +I just committed this +In addition we are currently adding many new cache engine implementations that might be used instead +on Release Notes page under Steps for doing a Build with Message Localization Support +Moving to this method is not used in +The first search is slow but by configuring query listeners the first user wouldnt notice +Addressing Ashutoshs comments and making the default connection pool for Hive +I think the exception is logged in addition to displaying it so I will provide the log +Similarly the diff for is against the generated version from this file is not checked in anyway +However why does from BB as b where ? work? But form BB as b where ? not work? I assumed that the bug was because the super class mapping BA hid the BB mapping when it was aliased because from BB as b where field ? works +Does this mean Spring ALLREADY supports field injection without making this an explicit feature +Which components do we want to addIll leave it up to someone who is more familiar with the codebase then me but perhaps some of these would make itcqltestcommitlogmemtablegossiper +the exact test WAR file plan is included in the know +Finally fixed with +wouldnt combining the mapcontexts prevent the optimisation of not having to redraw the map every time a highlight or selction changed +Ive attached test cases for canonical xml implementation for axiom +Deleting javatestingorgapachederbyTestingfunctionTeststestsjdbcCommitted revision +Ive checked but the support for composite column types also doesnt seem to support writing +Im attaching a first cut at the patch +I have finally made all of those and want someone to really run all possible tests with this fix before it can be committed +see we were to do this then the tests would behave differently based on whether it were in a or run from the individual project +The launching plugin works worksThe empty packages can be removed +I dont think that would work anyway +fixed rThanks! +Patch for uploadedThe deprecation of older create API is removed +Im closing the issue + did +Do check whether you are using jarsTool has been updated to work for latest +Then people can override the createRegistry method and return their impl of choice +missing apostrophe mount points Fixed +This may have had a purpose previously but it would seem that the reason is no more +If you want to try the latest build and provide feedback it would be appreciated +Ill try to do better next time +There is surely also something missing to support multiple Camel contexts +sehr! +There is a cross cutting concern across import and debug +This should be available in the branch +Thanks +Wont this be a potential security issue to store ssh private keys in the database? I dont know of any who does that +However I dont think that will have much effect since the body of the method consists of binary rather than standard AST +on Failover you may have to retry +No problem Ill submit a new patch tonight +After a few research I found that the parent pom requires srcmainresourcetherefore the child has to have it too +Now its in central where it continues to be accepted by almost all released Maven versions +As by the meaning of resign which indicates a valid leader actively resign its role as the leader deleting its election path is the correct way of resigning +thanks +same test coded in a slightly different way to show how this can commonly happenMichael I urge you to reconsider this +Although the fix was not very nice it shouldnt cause any problems further + +I added a set that should be what you needse +Yes said I should fix these issues for GA so I dont commit it into RC branch. +Did not any tests bc its a trivial typo fix +Im going to fill out the wiki page +As I saw In your solution you populate the new elements to a new different schema +Fixed in r +it is ok to rename toand it is jscalendars normalI am very sorry +Code and documentation updates complete +dirent pointer passed to scandir function is freed twice +Simon based upon my comments above I am marking this as resolved +The code works by doing a sort of WFS query against the geometries not against the means basically that if you have big symbols you may be tricked into thinking youve clicked on the point because you clicked on the symbol whilst the geoserver code will look for the actual geometry and not at its I agree making the distance configurable will at least make it easier to handle the situation from the user point of view +Backing bean for test case +I update and stop trying to make the and related class pluggable +Cannot reproduce with test case added to trunk +Erick dup of +Skinny rows is more special casing than just partitioning indexes and doesnt solve the how do we keep the index consistent w the actual data in the case of failures problem +Fix doesnt work for interfaces +No need to do anything here +bq +The patch has been applied with the lowercase classpath variation +is this a valid path in windows? + is a patch to fix this issue +Closing the bug after verifying the fix on latest stable branch build +I just committed! Thanks Sreekanth Ramakrishnan! +people in Harmony will not have this benchmark available +Seems a very good solution for enabling debugging +needs java or more we can provide rest including Jetty +I think it will solve the problem +The real cause is that the container has completed before we can display its page +Problem fixed in M +Rory If not in the Commons Net where should this be addressed in your opinion +Will upload the remaining patches on review remaining patches have been submitted for review on review patches listed above have been reviewed and submitted to master andLikthaCan you confirm that the feature is supported now +There could be a single that handles the lifecycle of all If converting them to services is really necessary that is the same approach that I was thinking about +Because I like better isnt a very compelling reason to change a core constructCommitted v thanks. +This of course is modulo +I dont know if that patch is of any use now as i see that the is now being moved into the profile service +remove statements +On the other hand its more complexYou could also add it in so that grouped Puts and Deletes wouldnt have to carry the same information N times +LasanthaPatch is incomplete +Its undefined in the callback to the deviceready event +Sure sorry about that +FYI you can modify any start parameters vm args system vars you like by modifying the launch configuration. +Did you try with +For the moment we prefer to focus on making user use the interface to perform configuration +I dont think this is something that should be considered +Sorry you are right I use JDK of course +Converted +The upgrade tests intention is to verify that sysdepedns row that may have been dropped of in earlier releases will be restored during the various phases of upgrade +to commit +Works for merev +What is this waiting on +Sorry linked an issue in the wrong window and cant figure out how to delete the link +NoBack then I got stuck at the time was changed so much and didnt work in at all +I see that the last attache +since it isnt the default and there are other alternatives there is currently no plan to provide a workaround +Instead of altering the signatures of existing public methods please add new methods with the new signatures then change the old methods to call the new methods with a default value for the new parameter +I dont think it would take much +It basically remember what listeners are registered before master initialization so that we can add them back during master recovery when we cleaned all old listeners +does allow setting the property but it should only be used in the absence of any other setting of user credentialsCurrently the call to get looks first at the existing properties and then at the internal userId attribute +Ive rebased the patch and fixed some comments +Readme file included +Committed +Pat Dont we need to do this for files in recipesrest and others +Deferred for +In my opinion it is ok to show the link only in the expanded header because changing visual theme is not something that the user will do very oftenJacopo +please review patch for looks good +Adjusted priority according to the priority definitionsSo perhaps this is a symptom of an invalid Composite Component and the exception is getting triggered earlier then normal? Thats interesting +Another committer asked me if I had started creating the patch yet and if not they would commit the patch +See forum thread and subtask for complete discussion on how this should really be implemented +leaks have disappeared in this version +afaik the publishing method is taking the settings from the server where it should use the util method which honours overrides +Committing this +Move as an inner class of +Next time I will do it as the first thing before reporting And big thanks for your effort on this +batch transition to closed remaining resolved bugs +I need to fix too +The new version of reflects modifications made by hand to make the sequence of any entities have code similar to that generated for a list of a simple type and removes the verification of the entity names against the artificially generated name extraElement +The only thing it does do is compile your additional classes and create a new warIt seems clear to me that were going to have to do our own thing instead of using to change the ant build to expect the war file in the connectorsalfresco folder +Looking at the code it seems to be a conflict in the order and the way we include and exclude +Nodetool will only query metrics about local C +This way the modproxy and modproxybalancer documentation would be relevant to understanding what modbalancer is doing under the covers +Reponeing to set release note feature request does not need to be documented as the entire tooling module is new for this issue as release note flag has now been set +Oops Its my mistake +No problem Matt! +ill try and get this in for Jukka that really helps +Please note the comments from tsurdilo make necessary corrective actions and test against +Hi MariusI have checked in just nowThanks for the patches +Ok closing. +That prevents deleting SYSTEM node from UI but makes the delete action much heavier +But does this cause any problem to? If this is not a blocking issue to I would prefer not to commit this to append +Lets do this in then for the reasons I stated on the pull in master +backported to together with the changes for with revision . +Please make the new patch with these changes Patch Available +The core issue I had was a low message rate and a trigger rate that was less than the default receive timeout +Verified the following scenariosAttaching the Local data to shared root Shared data to local rootWorking closing the issue. +I am not sure what the consensus warnings are but as much as possible leaving out would be better so that it is transparent in the future to address should the dependent library change +As far as I know it is a backward compatible change isnt itI am facing the same problem Riccardo Tasso had +Seems like the arent in there! I am pretty sure I added those to replace the paragraphs +committed an implementation on trunk and the road to +Sorry there was no problemsThe problem came from my side as i am currently developping a sonar extension plugin +Also we make sure we get all events notified + +Reopening for commit to into for uploaded +The documents were created can we remove this issue +Just looking to narrow down the possible source of the issue +Bulk close of old resolved issues. +I have automated the following parts Test WS internal tests sample tests against and Test samples test against Tomcat Assumes that you already have AS buld and Tomcat working on Automating and building of AS and tomcat +batch transition resolvedfixed bugs to closedfixed +Rossen yes your argumentation makes sense +It seems the script just calls java when attempting to shutdown the changed it to call JAVAHOMEbinjava and everything am closing this issue +The response received from the target service will only have doingtrue +We can fairly easily generate a byte array that falls halfway between the of two but we cant make it back into a String without understanding the JVM we are running in +errr +this is also applicable to generama and +The attached zip file contains both an AIR installer as well as an FXP export of the Flashbuilder project +The issue actually has more to do with multiValued fields with a real precisionStepIm sorry I dont understand what you mean by that +The factory can keep that aroundGood idea +Fixed in +One more thought on size then again I could do what does with its overloaded constructor use that size initially and then if needed +Will file a separate JIRA for thatWill checkin +You can place a tool like middle of the transport to capture messages sent by Axis +The display value is not correct +I took your sample code and ran with it +bulk defer of open issues to +js fix for IE and Firefox patch looks fine to me +A great workaround for this is to use even if the rest of Tapestry is +The sar deployment logic needs to check if a classpath refers to already deployed jars to prevent the duplicate access +Good eye +but a couple things come to mind we can verify the genstamp before we do anything to the files on disk + +Thanks for the contribution +Yes sure david there is a way to test if you have a look at the changes in revision where i used a work around by copying the form +Thanks! it took me so long +Did you press Make deployable option from context menu? If this option wasnt enabled datasource will be invisible for add and remove dialog +If you are looking for the latter you might want to check out something like SOLR +The only modification is in the checkSchema method +Justin could you review this? Put it on trunk first hopefully we can get a beta release out of trunk and then backport for or +the patch was committed in +Bang NPE is shown in error log +Isnt that a bug in the OBR resolver +The reason why I thought the fluent api would be valuable is because it isnt completely clear from the interface that the last parameter is the throwable +I guess my view is that any client that sees a session expiration should assume that they need to regatherreset whatever state +And if compaction cannot be done at the moment just go and rescan tobmstones trading efficiency for consistency +then it sounds like you made an explicit install instead of an Help did you update these +We have to change the convertion from Generalized Time syntax to a objectLets postpone this to +doesnt even get to the point of rendering or anything else just does nothing at allSo my only option will be to either create the share link in the inappbrowser or something similar to externalfileutil in the phone gap plugin directory + good start +That would substantially increase the chance of missed problems +The workaround is as you have specifiedTheres a related issue however or anyone else could you point me to where this is known? A customer hit this today and Id like to share that reference with them +You can remove the error from the project by removing the metadata from the project +Categorizing bugs related primarily to stylesheet elements etc + Username esriram +One should be available later this week +Just document the purpose and more importantly what it doesnt do +Looks good. +But its relatively short so maybe its ok +Rob problem is the archive name edition is synchronized with server for the first time +One thing I need to make sure is if we populate all columns in the list of needed columns for select from +Assuming problem has been addressed +Not sure exactly how to solve this though +to the update for the current script +submitting for RAT warning is for the srctest which has been added here +Thats true but it is not fixed in branches +Rendered correctly with transition resolvedfixed bugs to closedfixed +Thanks Mikhail +Marking these as targets for backport to so we remember that they are important +Now the dialog will display the content of the properties file for the default locale or if such a file does not exist any other suitable for the locale properties file will be displayed +VerifiedThanks Alexey +The actual svn move was requested in This issue canbe used as a reminder to update the svn links on the sites +Because of new methods introduced in there is a new class for +No Kerberos tokens do not contain group membership information but tend to get pretty large base encodedI guess kb header size would be enoughMaking that configurable is your choice I personally wouldnt as I know no things that cause header sizes larger than Kerberos but then if it was configurable already this ticket wouldnt exist +re my previous comment in it now sets the fields fine on create but fields are still not updated on saveorgjbossseamforgejpaSave and delete properly +Another component is needed for a reasoneable workflow cut a cas and copy the new view to another cas as initial view +Id be very eager to know the plans to get this implemented +I just committed this +Please read the file for more information +First one is to fit into and alter to support this as a date field type +If the user doesnt use it he can already change the corresponding properties in +just realized +It seems to fix Chris NPE +Please review the attached patch +Thanks for reporting and for the patch! +I think best way to understand this bug get a repro that we can run through the full upgrade suite +We will see what happensThat being said the test changes look fine to me +Id like +Add new system property +I tried updating the file to include the ios changes that havent been released yet but it didnt effect anything in our codebaseIm going to keep trying to get a working test project +The only viable solution I found is to match the multiproject settings with dependencies of the project +I only modified their values +The other complaints are about optional parameters +Proposed patch for this issue +Cool! +Im attaching three patches +I did confirm that the fix worked for +Fix bug +Categorizing bugs related primarily to stylesheet elements etc + +If you just notified me somehow when a new snapshot with this patch is available that would be fine with me +That might be good time for you to startup +Possibly that is the issue I reverted to tomcat and I can now add imagespyramids to geoserver fine +However its strange that it works when pacakged to a war +Turns out it needs the startup and shutdown jars in the bin directory used to be a problem in early versions of but they were resolved a long time ago +Ive applied the latest patch and tested again +In that case abc will be resolved in the context of client side mount table +I am all for a contrasting background but dont think we should necessarily restrict ourselves to the grey that was there before +Committed +The change from Carl will sidestep the issue +Lets see if we can get this added +Makes the index mapping think these are fine changes and Sebastians comment is correct +We have the same problem when trying to convert dates to +Thanks Jonathan +Glad it worked though. +It feels wrong moving all the work into +HelloCould you give me some more infos about how to set journal and jdbc for persistenceBest regards +Simple patch to make sure that htable is closed on execption +Hi FelixYes I think you can close itWe can always open another jira later for the other james project to go to maven TksEric +This patch will do the package to qname AllI think RPCMR should NOT accept user defined WSDLRPCMR generates WSDL types section from Java by reflection and doing databinding according to that schemaIf user defined WSDL and Java doesnt match wellno one can tell how that should workThis will cause many troublesWhich case users want to put their own WSDL +Itd be simpler to write a line piece of python code to load the XML file and extract the relevant settingsReturning to the patch which Im not over enthusiastic about as you can guess conversion to should include the inner exception for anyone trying to diagnose what went wrong +Also the patch as Jurgen said if the Leader server crashes will bring down the whole cluster +Applying this patch put a big smile on my face it does indeed fix the main offset calculation error +Wahoo +I tried several ways to rename a package with multiple classes or moving classes to other packages and all the time the refactoring support changed the class names in the spring config files correctly +This is the first time that I use iBatis and the getEntry test doesnt pass +Were we to allow the behavior requested a cluster could end up partially unsecured with no warning that it is in such a state which is not acceptableWhat youre asking for is essentially to make securing the datanodes ports pluggable which we fully expect and plan to do +Attaching +This can be useful when you call resolve from Ant and only want to reuse the same resolve result within +User guide was updated to remove obvious references +Moving out +batch transition to closed remaining resolved bugs +Hi DushshanthaNo thanksI always feel very pleasure for contributing to OSS community for Ive also benefited from the community very muchRegardsMichael +Please verify Ive changed the naming for everything under trunk that looks +Also and more importantly since the optimizer is someone complicated now I think it would be good to put a large comment in the package header for +Its a good suggestion BryanIn fact I have tried to do some unit testing before submitting the patch +Ditto for all the others +Another attempt +Thanks Vinay +trying to make hudson pick the thanks for the review + +Sorry I get those errors with in Directory Studio version has been released. +Thanks Thomas! +Like we did in we could divide the implementation into multiple jiras +I just wanted to note that in my case where I use a stream script for TRANSFORM which only happens at the very end a workaround was to simply force it to have stages so that by the last stage the input list has been merged +That is a separate issue just rename both files uniquely and deploy and connect for your testing +the URL can get huge and exceed the servlet engines internal two slashes in the URL were a red herring +All tests are passing for me on the three branchesI believe accessingreturning the lastzxid as I do in is valid but it took me some time to convince myself +I will investigate further +the exception was accidental reproduced but I cant provide exact steps +Paul Guillaume and myself discussed quite a while about the issue and we decided it might be best to roll it back +Nick let me know if we should use another release version +But the same approach works care is just needed. +As of both and have a validators parameter that allows the component to be validatedAdditionally has a translator parameter that defines how the text should be translated into an object. +Is there any way to convince subprocess to use the python pty module for its fork +Im cool on adding genericsThat comment local addresses was kind of a remainder for me +Assigning to Marius as he wrote the original text +bq +Suggestions +When to run add partition command is depending on when you want to make the partition available for use +Thanks ClausIm heading out now +It seems that when computing how much data we can mlock we must round up mmaped regions to the operating system page size +and often times need some sort of customization to make them work in user environments? This involve changes to the core or are you finding yourself copypasting to add some needed facility that youd like to commit back and have available at a rate that is higher than that at which we releaseI suppose this facility would be for toolsmiths to share their improvementsDo we need to move current tools to the toolsmiths repo? Or can they stay where they are and toolsmiths use repo for new stuff or amendments to core tools +Applied to the current code +Thanks Sangeetha + +Keep up the good work +Given the original statement The Apache Connectors Framework has an implementation of NTLM that it would like to include in its release there is no pressing need to look for an alternativeIf the license requirements to implement NTLM are problematic it doesnt help us all that much that somebody else has the same problem as us but just isnt as diligentHopefully we will find a way to make this work +JBDS can now connect to the Dev env with appropriate settingsVerified on Developer StudioVersion Build id Build date +Patch for review +issue been fixed +This will be the primary focus of though +but the cce needs some stack traces which seem to relate to bundle replacement and classloading will be suppressed via remove printstacktrace from errors on close replace with debug logging +On the other hand I do not think that there is left one use case that requires creating links from +Extract the thread dump information from +Its cryptic and ugly for users and really doesnt tell them what they should do about the problem +nullable false should be the default in Grails validation all the time +In order to keep the work separate I backed out the changes from item and made sure I worked on pristine versions of the classes from the Apache Imaging development trunk +Applied in r +Please provide update on your items as a comment +I am seeing this crash about in every times I run the pyactivemq unit tests which only does normal things with AMQCPPThe code I posted might be slightly unrealistic but it was the quickest way I could duplicate the crash outside of pyactivemq +Verified with JBT. +Closing. +If you try to start theanimation while another animation is still running it works +I think just fixing it is probably fine here +My fault +Yes I found all the riftsaw samples are changed to +Hi ChristophI appreciate your interest here but can you please look at the patch files? It looks like a formatter was run on them and not much else +Uwe Im not home right now will look tonight! Thanks for writing the patch +With mojarra it works with myfaces or trunk NPE happens +bulk move looks like the same issue as +a separate issue add a specific that extends to throw for Configuration problems +Thanks David lgtm +Committed +JIRAs username rename implemented +Everyone is running with Xen now +code to fix that can be found in r Tests and megre still merge now with a test in +Yup Monday is the pencils down date +Maybe ask on the list +Test Case is addedPlease verify +Ill check in the patch in a few days if nobody complains checked in +I dont know +You can try to install WIAB locally and see what happens +In addition to the new unit test I have successfully tested this patch on a cluster +Okey thanks for not sure I would call it hibernate misbehaving but more a changing classloaders for entities will by its nature cause problems if you dont rebuild the sessionfactorybehavior but im just guesssing here +Iirc it required the user to generate an addition class +Closing resolved issues. +In index writing its typically a batch system indexing docs and theres no fairness to mediate? Breaking it out into a utility class seems like a good idea anyway +I havent seen any problems as a result of these changes so Im marking this issue as resolved together with the related issues and. +Im not sure exactly who does use the etcdefault systemI think the idea isnt too bad but once again principle of least suprise +Thanks for finding that! +works not perfect +Jimmy Xiang i just find this one +to close as Wont Fix so that it will be documented in the Release Notes as a known issue +Thanks Tom for taking a look +Let me remove the first two test address pairs if theres no objection +After that dependent project could use it +bulk defer of unresolved bugs to +no feedback pushing to next +Title was manually set after titlekey resolution and this is wrong according to jsp tag try the new snapshot and let me know if it works for you so that I can definitively close this the latest snapshot and the problem does indeed appear to have been fixed +bq +Salutlets take in consideration that shielding from data races is different to keeping the data structure thread safe anyway if you are confident that the used locking level is robust enough to keep the data structure I trust youAnyway if youll need to traverse the structure take in consideration themerci a toiSimo +LarsPassing null is same as not passing any splitkey in which one region is created with empty start and end row +When I get to improving tracing then that would be proper place for such information. +Yeah we have vs on JenkinsIm more the buildbot area so will have to defer you to the Jenkins admins now for this issue but it will get doneGav +Dont have the cycles rightnow to understand how the hive build system deals with and local jars +We only need to cleanup and make this SPI thing look a bit more professional +This patch is to be apploied against the standalone branch it adds the required properties some notes in the readme and some documentation in the properties file +just committed this to trunk +But Im okay saying this is the users problem +not include time spent waiting for mappers or reducers +close off release +Will check test cases later today +I sent an email to jenkins dev asking for access +Confirmed that this problem is consistent on both LCS and +There is a DELIMITER in already +bq +qpidjavabrokersrcmainjavaorgapacheqpidserverpluginsqpidjavabrokersrcmainjavaorgapacheqpidserverpluginsqpidjavabrokersrcmainjavaorgapacheqpidserverpluginsqpidjavabrokersrctestjavaorgapacheqpidserverpluginsjavaextrasqpidjavaextrasqpidjavaextrassrcmainjavaorgapacheqpidextrasexchangesdiagnosticqpidjavaextrassrcmainjavaorgapacheqpidextrasexchangesexampleqpidjavaextrassrctestjavaorgapacheqpidserverpluginsqpidjavaextrassrctestjavaorgapacheqpidserverplugins +Id rather add it back later if it turns out to be neededinstead of keeping all the old cruft in case some of it turns out tobe useful +If they typed abc I wouldnt know if that was relative to the filesystem or the eclipse workspace +I will provide my comments on that jira +Is there a doc reference that explains the restrictions? Also would it be a reasonable enhancement request for the future constants to match the collation of the column with which they are being compared or is that just too hard +Weve moved to Felix HTTP service. +Thanks PierreYour patch is in trunk at r +x repositories on a dailyweeklymonthly basis is too hard +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Changed description to JBC as that includes the fix +Patch applied +No youre rightMy apologies I was looking at the bad url. +is not looking for this class since is not using SI +Again thanks for taking the time to look at the stuff and incorporate some of the ideas +The test passes after I rebuild the whole trunk +You can use the redirect method to show a view +Verified on +It looks as if you are attempting to restart the server during the publish which may be the cause +NET it was decided some time ago to cease focusing on Mono support for SPRNET +Is this a regression or was this broken in tooThis is not broken in +The value of the output is actually translated when the pom is read and the parameters are injected +I completely ignored that this is about the zip +Fixed in CVS +the implementation should also try to arrange to checksum that data as it is delivered to the application from the stream rather than as it is read into the streams internal buffer +Youre seeing a first indication of lack of join support but if you proceeded youd see that the actual join machinery on shapefiles is completely missing +Defer all open issues to +Ron can you help look at if this is a known issue or some regression failure in? Thanks +I will do soSbastienAdditional note Phil you are referring to r +SSH will start and find the missing daughter +Tell me if I can do more on this issue +trunk patch attached +The will have to be obtained via a HEAD or file stat when no is returned and the response isnt chunked. +So we need this patchBest looks like not exception compatible but calling stack compatible +Wont be working on these and theyre old +Have a era cluster with TTL where repair turns gb of actual data into +The new patch addressed your comments and added more test steps +As always patches are welcome +Future jBPM releases can only contain bug fixes +Done +How does this avoid leaking index entries that will never be cleaned up by compaction +Hi Arvindthank you very much for your feedback +which is a shame because if it was an interface it would work just fine what you can try is adding a default constructor +Do you mean +For those not running on Java I believe that a separated JAR file is available +You can use GraemeTo use you need to be on the requestresponse cycle hence is impossible to do so when you are initializing the is not a bug is an improvement as I see +Hello SumitI have implemented following mentioned points Prepared Type data for fax billing and shipping +I have checked and this issue is already fixed in Karaf so I am going to close this issue as resolved. +Great SAP are happy for me to follow Route and just attach my patch which I will do now +Networkremoting code has been assimilated into +I used to work on windows I work on Mac now +Apache Shale has been retiredShould we also remove the Link to Apache Shale +Sorry for the confusion +A patch for the bug +Are you looking for a PP presentation on iBatis or are you looking for a tutorial of how to use iBatis? but this is not an acceptable request for JIRA. +Hiif you run java jar for you PDF directly from the command line do you get any output? If yes its how is called from the PHP program which is a PHP related question +Username genman +It seems that you applied the first patch not the second one +Please read the comments above to understand why this is not a problemTo sum up is a generic mouse event so its not executed correctly as you say +dims +Management server is +After asking infra my understanding is that current thinking is that the use of jira groups is not recommended instead roles should be sufficient +I need to beef thisup either here or in the JPA component doc page +Oopps i just saw it greatI will post my feedback soon +Thanks Tim patch applied as expected +HiAny word on thisWe can use slfj in as soon as these dependencies are on a repo +Now this file contains only the method fix +Moving to commons project as enhancement because this isnt Struts code +Moving to +Denis yes it could be an useful feature +Let me retry your v against hadoopqa +So all issues not directly assigned to an AS release are being closed +I committed this +The downside there is that you wouldnt be able to use the namespace tag for configuring the destination +Ian with fixed should this issue be marked as fixed as well +fixed a bug +Will this be fixed +This method should be marked asdeprected +These are just the changes I need please add this diff! +This situation will happen if the vdb is initially inactive which is likely the case with dynamic vdbs +The tests needs to be passed before a patch can be considered for commiting +Were doing an initial synchronization so no cron job yet +Sorry for the delay on this Ill followup with a patch that removes the necessary bits from attached patch combined with an svn rm drivers should be enough to call this a Done Deal +I guess the patch was submitted while I was away and did not notice it in the pile of mail that accumulates in a few days of on and trunk + +The corrected manifest where import is optional +could you describe what behavior you expect from the current algorithm in the following casedeferable constraint in deferred modexact inserts vdoes not commitxact inserts duplicate vreading above it seems like xact is going to block doing the ahead check which does not seemto be what a deferred constraint wants to happen +Close after release +No it was just me being a clown + +Re my item I just read about your suggestion for the to wrap passwords sounds like a good idea +Committed! svn ci containersrcmainjavaorgapacheopenejbjeeAdding containersrcmainjavaorgapacheopenejbjeeSending containersrcmainjavaorgapacheopenejbjeeTransmitting file dataCommitted revision +I am not for or against it totally +The same credentials were working before +The source code of is taken as the fix +Dup with . +Part of a bulk update all resolutions changed to done please review history to original resolution type +punting to next +Its a question of dividing the problem into manageable chunks +Thanks Matthias I believe this issue is not only related with TODO app but it affects and +The log file for using maven plugin snapshot with a container +Changes that fix the error when addingediting value constraints and default values +Ok all that needs to be done is remove images user and developer folders from the documents directory +Fixed with svn revision +Im attaching a new patch which contains the full test case. +Ive created a general CR for request payload handling added a fixing the payload issues doesnt resolve all the Subversion problems please move this bug out to or +I think this is now fixed properly +Maciej feel free to close this issueBut I still think that there should one unique Id which should remain unchanged regardless the process is embedded reusable or +I usually mark it as Deprecated in XX to be removed in XX so people can easily tell when they can safely delete the class. +The best thing to do is to redo the test decrease the timeout and run the test with a sleep in it +But if others really prefer something more obvious its ok for me +Heres a patch which addresses the issue +move to value and use ALL as default. +this is quite critical but we wont be able to fix this before deadline +So instead of having one instance handling values and the other handling values the second instance now handles It works but it is definitely not a clean solution +Moving out some bugs to please rearrange if you plan on fixing any of these in the next few weeks + +Patch staticHowever passing Stats to the methods you mentioned would only be possible if Stats already defined every possible statistic either as public members or getter methods +Looks like a duplicate of which was fixed in so the feature should already be there +Saving the main controller after you have made the change to the second controller would have fixed your problem +I started on this but am holding off +Bob for you +Jesper feel free to move this to an issue and reopen if you feel it could apply to the latest version of the software. +is done +We didnt define a hard limit but most lines are to chars long +Not sure if it will be today but it will be within the next couple + for the patch + on patch +Stackbq +ThanksRevision +Already tried to generate trace logs for hosts with the DB dump that i had +Yes I know if I add one of the facts to the esb project renaming will work but it also generate some useless files and folders the ESB project has its own project structure +We were on where the class hierarchy ispublic final class implements Configurable It seems that is added inI also agree the caching should be done in the implementation because different impl may have different caching policies etcI will close this jira +If they have a then it will need to be renamed when deployed in to the ODEtomcat environment +The first comment defines the initial proposal Which says that when both new and old values are set in configuration we give preferenceto new values +Sorry Waldemar Im not sure how to read your description +looks goodCan you change the comment to remove mention of date string since thats not always the case +Assigning to Len for second review +Just doing a bit more before I check it in +Hence just regenerated the patch on Linux and uploading itAs theres no difference at all in the patch I will commit it +Adding support for in and +Thanks for catching that AlanUpdated in trunk as Rev and as Rev. +is also slowishI am thinking this can be speeded up using the secondary sort trick so that values need not be cached +covers security and the IBM jdk +Since neither enable not update sources meant to be used for non JDT project. +How would the authorization check at create db time happenIf the authorizationDB has the roles declared with privileges granted to it would we be using the existing apparatus to do the actual authorization +IMO depending on the app server either a classloader based strategy as is used at present will work fine or the app server will need to use something else for every ear +Thought Id link it in. +Applied patches in r +This is same as the earlier one plus an additional change to handle the failures from application +We need to do a sweep and add this header EVERYWHERE +In general we need to do some work to make it so cluster boots as fast as possible and is immediately useable not online but soaked running major compactions and splits etc +This seems like a bigger feature we should think through more +pls review it +It does seem like this should get ported there +But I am not an HDFS committer just and common so it might be best to get an HDFS committer to put in both +Do you think that on a busy forward proxy its safe to go with shareserversessions? Do you need a bt with shareserversessions? Thanks in advance +I thank shutdown should close all regions unless compacting and close the last one after the compaction is done +Part seems ok to me for the sake of the form +Per Burr and Max at call these wont make JBDS in but were uncommit so no change request required. +Can you add more detailed comments here +Going to commit +PS the link is for the x bit version of Eclipse +Ben where is that parse call coming from would be good to see the complete stacktrace there +Just committed the patch +The reason for enclosing the values into quotes is to put a delimiter or a line separator in the value +Note that this is a of the same workaround in patch also works my windows envThanks for the patch +testsMore probably is that some VPE tests freeze +Any news regarding this issue? Is there a patchnew version available? to add additional lookup of Bob +Grzegorz thanks for checking +This may have been caused by the container not being fully initialized prior to installing the feature. +Marking as closed. +has a long history for failing +removed unnecessary import +any real chance of solving it for +This patch alters the max amount of time that a deadline can be exceeded from ms to ms or of the recv timeout in +Thanks Xuan! +I dont think this is something that I am responsible for as the readme in question is contained in the code think that has been consumed in one of the Guide documents though +Good catch +How does that interact with tomcat +I think we should do this for? MarkGrant is this ready to go in +The same applies for changing the signature of cleanupJob +Updated patch. +Fixed in jrubyc returns a status if there were any errors during compilation +I committed trunk patch and then found that patch doesnt work +But its ok to set font info as long as you use relative sizes +It passed for me +Attached logs are from running the test with the unpatched version of ServerThe failure logs has a about miliseconds in +Ive been unable to reproduce +Manually need for IT. +Michel could you confirm that youve activated the property in the configuration file +With RC I am bumping into +Please check that it indeed works as expected +It is the cost of connecting the components to the pages and doing all the other work +The method name and parameters would be a clear indication as to what it does and separate it from the methods used on standard column familiesWith the interest shown in this feature I hope we can reach a compromise that works for all involved +I know this wont help much then but shows that there will be something wrong in your installation maybe the xml parser is somehow puzzledIf this do not help could you please pack a little war and attach it to this jiraThanks +Bear in mind that BLAS and LAPACK tend to assume storage so for Java implementations it can be more efficient to change the algorithms so that D matrices are stored as their transposes and their indices are swapped within the algorithm +Are you planning to provide some information +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I would say that we can also classify this jira as a bug as well in order to put this into branch +Also the patch seems to include +The commit notifications should go to scm +Examples are when the code is referring to Cloudscape releases prior to it being open sourced as Derby +Perhaps during FOSSG +OK gadget to show one job is available in marvelution plugin from comment only +Once I get the thumbs up from you Ill see about fixing this another wayThanks bypass of assertChangeable revision to trunk +But fine with the other approach too +In our build machines they have system property for this purpose +Fix is confirmed +Fixed in rev by using ant code instead of plexus. +I will fix this asap which means you will have the fix in forBy definition a direct call to the has to handle a response call with the standard parameters with no internal ones. +This is unlikely to be a common use case and the implementor could reasonably just use a to store the contextGive that it isnt clear which methods in the API a context may be required for it seems better to opt for the setter +Sorry for the late reply +Two brokers connecting each other using none duplex connection on both sideThe message loss situation does not changeThis means activemq cant provide reliable message delivery when doing store and forward style message transfer +on code review +A couple of comments The list of end points will be specified via configuration either looked up from a named file or passed as properties to the constructorfactory +Our performance test environment show this bug +Example tcphostkeepAlivetrue +When the bug is fixed the test should be updated +Quite a bit of work went into issues related to what happens when you press newline +and I dont think we should discuss those in the public should we? btw whatever the default is an import should not leave the transient space in a weird state like the missing jcrprincipalName properties after a user importthats the consequence of the compromise I agreed on although I already wished several time I hadnt +Marshall take a look at to see how to start a real quorom set of servers +Which is more generalIm just wondering why the programmatic API never used trash in the first place? pete +It did that for me too the first time I patched but then the next time I compiled it was fine +Uma good suggestion +Creating shared network with all the services set to VR deploy a VM using the above network create PF firewall staticNat LB rules and they are functioning fine. +bq +so rather than being able to see the result of all requests initiated before the sync the follower will see all requests completed before the sync +Version fixes the issue on +Ill commit the three patches soon +Closing as is already released months. +The solution we are going to with is to consistently use the maven version for the zip directory and maven artifacts. +Most apps will not see the async API since they will use +This issue supersedes the attached patch for the +See it was fixed in +For once it will affect Hudsons configurations and will introduce an dependencies thereI believe a better way is to rely on the content of SNAPSHOT file which includes all bin scripts in it packed into file +Patch applied in revision +Forgot to mention that it was tested on Ubuntu LTS and Windows without any issues at all. +This sounds reasonable +Can you test it too and see if you see the same behavoir +Thanks Bikas +I agree that bytes per task is large but I was doing a back of the envelope estimate for worst case approximations +New tests written to identify issue fix checked into CVS tests pass. +EAP has seam stuff as well + +No changes are made except for changing the filename +It means I can issue Join Queries with tables from other schema +This issue has been fixed and released as part of release + +We already do remove members who replied with MERGEREJECTED from the test for this is +ChristianCould you please parameterize max value using class and resubmit the patchOleg +I manually tested the patch on +Dropping priority as we are creating a for tested a quick fix for this specific use case so that SOA QA could continue their testing the diff is attached +Thanks Hairong! +The patch works for meThanks +Removes legacy tar creationRenames remaining profile to be tarThe TAR itself does not have the bin postfix anymore +Closing +Ups rightOk Ive added the to the api and added a first simple implementation to the jcr resource bundle +Id like to resolve this as a duplicate of that issue +Test reports with success +This bug has been marked as a duplicate of +I think that is fine Lily +We should allow to specify bit packages +Ah yes you have fixed my mistake +Daryn Sharp Im a bit puzzled by this +To avoid that a second thread reads an old value of cached while the first one is between the two lines inside the if block this would require locking which probably defeats the performance gain of the cache +I understand that there is an easy workaround +Im not sure when well be able to get to this comletely +I didnt get this Warning +since This also corrects a matter concerning escaping of special XML characters when generating the +easy now its trivial to fix and will be fixed for +Attachign a new patch with exactly the same changes. +Oh sorry +No word on the others +Marking as fixed +issue now release note status has been set +Oh well Ive just remove the package and specified manually the path to when compiling and I dont get the same error message nowIm now having a return code so Ill investigate a bit more +Once is done this should be resolvable +it might still be nice to include this information in the OPTIONALLIBRARIES file +AFAIK this was a of the Jacoco conflict and was fixed along wiht +Triaged for +Thanks Chinna Rao Lalam +Hi Andreas Thanks alot for responding +Maybe we should just call it +htaccess +WorksI am not taking this up in this JIRA +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +That really can jus screw things up +But is still open and needs discussion +Checked and the entirety of the patch seems to be included with that both dvdstore and hotel booking shortcuts are able to deploy the demos and initiate demo startup page in on windows +fixed by the above comment this issue needs to be +Committed to and trunk thanks for taking a look at the patch Ted. +Derby sees that a permission has already been granted to that authorization id so that authorization id is not available as the name of a new role +Tried with kvm xenserver +You could file an enhancement request titled Allow Weblog Ping system to use proxy settings but I think its a very low priority feature +Fixed in. +Im going to run tests locally since seems a little funky recently and then commitI manually verified that the unsafe comparator is getting used by introducing a bug and making sure failed +Could you add a unit test just to avoid I mess up everything later +However a more appropriate fix would be to update the saveaccessfile method to report success to the calling methods rather than it being assumed as present and the calling methods updated to take appropriate action in the event it should fail +We do not want to workaround a bug in some other component by removing a desired behavior in the plugin +Still these needs to be completed for windows environents +Done +The tool was very likely Nmap or Nessus but Ive not been able to reproduce the problem myself. +Note that this is a feature and should fix this issue in the right way +Yes this is done now +With a workaround available and a move towards each module providing its own config this issue will no longer be relevant. +bulk move of all unresolved issues from to +My recommendation is to move the issue to classlib component +Build console configuration +It looks like this issue might be related to Read only mode after applying the patch from I could not get this to happen again +Thanks for the test case JimIf by the same scenario you mean the workload that left your commitlog throwing exceptions then yes please +If Im not around just dont wait for me +Youre right that this occurs whether or not the view is open and this probably needs to be fixed +anyway the behavior you report isnot normal and there should be some trace in the log +Conversely producing all that output might impact the test slightly +The language may need some changes thoughI hope the changes to the new Admin Guide topic are satisfactory +You have one table with some eventsdatainformation with a VERSION limit because you want to keep if small and efficient +etc +Why this is needed? Do you know if the AS handles the unzipped in deployment? this also complicates the deployment through Admin API and CLI +I intend to submit this patch today +Two of the tests failed but re running those succeeded again +The username is titmus +If a valve cannot be installed via JMX because the context is not registered we fall back on installing via the Tomcat API +Will be great to apply Customize after Add version thanks for the patches +Implemented fix at Apache today and synced to fuse trunk and branch. +Restarting Archiva temporarily resolves this +Committed in revision +I will try and implement the bean deployer for inclusion in the JB branches +Where does the master copy of that schema liveHowever if you want to kick back the old way of doing so i have no objections +Could we resolve this issue for now +I shall run tests on this patch and revert back +If anyone wants to take care of this one feel free to take it off my list +Yep that is what I meantExceptions are currently caught and stored in a field named fault of type ExceptionChange it so that fault is a Throwable Throwables are caught and stored in fault and the getter returns a Throwable +The intermittent error encountered can be attributed to something else +Until then I dont think much can be done for this specific feature in the CDI EG +Nits feel free to address wo another patch +You just need to set BEEHIVEHOME in your shell +Thanks Amar! +I couldnt attach the original process or even the original message but Im sure that it was failing with the xmlns prefix +Should we still punt this from? Doubtful anything will be done in time +Move this minor issue to +Added the license headers +It would just be nice to have Tapestry work with application code compiled under JDK +Does it always work to set cglib as the byte code provider in? If so its probably a better way to set the property than setting cglib to JAVAOPTS in bin +Lets say that Keith Roy and myself all share a springsource Twitter account +Thanks Ravi and Jakob! +Luke I have not seen any failure related to this +richgraphValidator I dont seem to get notification on I save password also when password is too short +Volkerdo you have a patch for this? If so please upload it so we can merge the change into the by is right the issue still exists Why the JIRA is Closed +Can it be that stops feeding new into because of this bug +on patchI ran through security test suite and tests passed +Chances are there that someone who is very knew to Sqoop might not even come across it +Im not upset at all! I think its just me not being very clear and you were perfectly right asking for more explanationsSo conclusion again close the issue as Ive created to track the improvement we discussed here with hopefully enough details and if its not the case dont hesitate to continue your comments they are always very gladly welcome and appreciatedThanks again for your really great work Roshan +Can you check again please there has been a release in the last month +test files to go in test for unit testing +Thanks John +The patch fixes some. +This here is definitely not a forum for discussion so please stop it +What is the behaviour on EMR pig +committed to trunk in a few days it will be committed to in a combo with other maven related patches. +I dont think this is serious enough to spend time on +Well bugger +Hi Van I think you said that at least one of the GSG examples would include this +I really think that older objects that dont have interfaces or cant be should be limited to using the primitive model +reopen for clarificaiton Well if copying is the only way of fixing the semantics then it is the right thing to this were copied from EPL code then dont forget to put it in an +I applied the patch and it is working +I would rather see the expiration callback handle the hint write then see the call storageproxy +Aiming for the for trunk attached including tests +Will do that this afternoonIm very happy to see some activity hereBTW +Running with correct versions of jars in top level libjdori resolves most but not all test errors +Expect a new build in the stream shortly +The patch with fixes for default publisher was applied to branch at r +Hi LukeCan you please provide some links to the forums where can put this and get resolved? +Sorry Luke my mistake This is not a bug +The reason I made a builder was to separate how data is storedsent from how the client logically thinks of and creates a request +Tests for jsp templates were added +Ill open a separate issue +unsigned signed version of attached please test this patch against EAP progress on this +I couldnt recreate it with the file you sent me unfortunately +The problem showed up in a stylesheet that uses xsl on the enclosing element +That should give Hayato an easier way to test at least +No particular fix tied to this just generally a lot of cleanup effort by a lot of people +I passed somke tests on +Thanks Jarcec! +about that assignment +So any verdict on this for? We should make a decision and close the JIRA either way +Yep that was what I was implying +Fortunately the server code already checks for such a paramter and we also have the displayed text somewhere so no need to enhance the translations etc +Fix checked in to the trunk and branches +In some way we have different understanding of fixed? I thought it means that the problem is corrected +heres the version of patch +Heres the start of a test for the +I can fix it though as part of the commit +Sishen Are you going to update this patch to accomodate s suggestion +Theres got to be a better way of removing the ancestor paths +CarlosI checked out equinox and the pom doesnt have changelog plugin configured in it +Expected for +Thanks for the working on this Tsuyoshi! Some comments so far Please add annotation to the new class +It should reread the created shapefile to verify that the geometry has been correctly copied the long name fileds are correctly copiedIll try to find some time to write a more correct test a lot +correcting type and in svn rev +Good catch RobertIll fold this into the back compat test +Attached from server +Do not merge form the attached diff +Absolutely I missed that fieldname usageUpdated patch +Leosigh its back +The Geometric distribution test needs to be fixed +Closed. +This bug has been fixed in Xalan Java +By the same token pycassa and the driver need to be mashed up as well +Thanks Albert Chu +crc to the file name +Committed to branch +eg when creating a dont specify a limit in the default the has a limit of characters +The two queries are not identical! In a distributed environment the first query asks each shard for the n document Ids with the top scorces +It is not a patch but it can be easily integrated +It does make it even smaller and simpler +Thanks Ning +Not sure whether we have conclusion for this I would like to add the operations of removing bundle in the unloadConfiguration method first for currently while undeploying the applcations the bundle is still there +I just realized that this JIRA has been back and forth for and a half months incredible patience on your accountNew patch looks quite good to me and all comments seemed to be addressed +you Jacopono I did not see the screen you have pointed outI wiil give a look +Patch reading the remainder of the segment +I added an exclusion for this one and enabled link checking for solr too. +I can reproduce it in about of cases +th so if you want it in the release please target no later than Jan +I hope thats clearCheersClinton +Tagging for if this is still valid + class class to back out patch is not safe to use and in any case has not been applied correctly to the current am backing out all these changes +Reopening it after users it as a Bug in monitored handlers are being wrapped in which implements +All tests pass! Ill commit this tomorrow unless I hear otherwise. +If the Axis Engine receives a REST request and if it cannot find a relevant builder that matches the it uses that builder +It seems now that the correctly reports the portType cannot be resolved +This has been fixed upstream If thats there do we still need this +My only hesitation with that is that it disturbs the ordering where right now Comparable is consulted ahead of falling back on equals +oops the patch based on couchdb +But it is not a bug +Patch implements Shalins suggested changes +We should be able to fix this forAs near as I can tell this setting is simply ignored when in a Groovy editor +I made some progress since this morning but havent finished I am running into some issues where the coverage tests pass in Eclipse but fail in Maven +No we should just open the HQL Editor with the query dont execute anything +Presumably the lowest value would be the value of the count from the first query to the server whatever that happens to be +The TODO above can be droppedMy understanding of that todo is that it shouldnt be setting as many PRI as theres normal handlers so it still standsbq +Ive committed this! +Opened to fix it +Release Note A currently unused part of the JMS address was being initialized with incorrect information +Done now you have to add an external jar file to the project classpath then its tld entries will appear under node Tag Libraries in Web Projects view and can be dragged to the Palette +This issue is when the end user is sending using the reply manuallyIf using Camel requestreply out of the box Camel will not enlist temporary queues as endpoints +Same behavior in STS Version Build Id behavior in STS Version that its eclipse problem. +What is this restart from the failure point feature about? Do you have more infos on that? How does it work +Hi everyoneHrvoje would it be possible to attach the resulting drools osgi jar that you were able to create? +Have you tried Beta? Does it still work or work betterThanks for reporting this even if it turned out to go away on its own! it does work ever since I couldnt figure what was wrong. +But maybe thats not important for this issue +leaving this in progress until we have tests written and have also confirmed the feature on a device +There are several problems at least with the C code +Closing resolved issues. +In case of timeout or job interrupt the is cancelled with optional attempt to interrupt the executing thread +and more over in same project it self if they want SNAPSHOT version of C and release version of A then this is valid use case +Here is the patch for the branch x +Hi AndreaBase on the my testing the problem started from our internal build dated Nov +We should get lots of bug reports if its not working since it makes it difficult to upload a SLD document when its broken +The patch containing the changes +geoerver wont even start up though Tomcat which version of geoserver did you threw it in +This just does that for META incase of no region plan available but with a sleep +Moved to JBTM as it needs fixing in the Narayana +Let us see if this patch fix it +But recently I noticed that I can simply use HelpInstall Software instead without first uninstalling the plugin +Lars you assigned it back to yourself again +can you change the summary and description accordingly +committed. +Adds field name to exception message for missingincorrect typeIf a field is a destination for multiple copyFields and it is not defined as multiValuedtrue then it shows a Warning during schema parsing +Thanks! +thanks Arun +Do we have plans to push to versionNot that I know of hopefully we can merge the couple changes that introduced a big delta to so merges like this will be easier +Richard can you take a short look at the changes and maybe do some testing with your use cases +Ive just tried to reproduce this again but cannot + +Matt whats the date for providing a patch? Id really like to see this fixed in +Moving this to the you need to decide if you need this in the release or not +Okay I just switched my browser to Italian and browsed some Italian sites and you are both correctProceed +The problem is the LAST newInstance which runs when the class is set and which is outside the trybenson +Thats helpful +Screenshot attached +Thanks +Changed Target Version to upon release of +so just have to copy the corresponding files to the appropriate Can you please submit a svn patch? Based on the svn client you are using comand line or eclipse you will be able to google and find instructions how to take a patch +Here is the patch to revert the ampersands +migration guide added to closed as resolved and released +This does not seem to be a regression and I propose that we fix it for +I dont mind working backwards from a configuration if there is a case when that makes more sense +I think this is sensible to implement because client code can never rely on getting back more than one copy of a given key +Backported the code changes to but not the test changes as they need byteman and does not use that. +One use case which I come across is after installing and testing user wants to ship the nodes and update etchosts by reconfigure the IP addresses +we can derive from external classes with generic bug is about the special case as stated in the new title +Did not mean to link the PR against this +This problem is really serious +not in the series +Can you please create patches against the project root it makes it a little easier to apply themThanks +better patch version prev version doesnt work in all cases. +Hi TimI see that doc +A testcase that shows that the name attribute works fine with reports but does not work with modules +I will add the rest of HDFS rpc protocols +Todd please run tests and manually so this can be committed +I updated all scripts calling and I updated all docs I could find that reference the command line except for the PDF +By leaving this approach do you mean keeping using it or abandoning it? I guess the latter +Thanks to whoever Ronald and Alex for pointing this out + for the patch +Isnt that part of +I am guessing you are actually hitting an issue where the channel being full is causing the replay of even the last seconds to take a long time +JUG also provides a native library for getting the mac address +This is actually one of the reasons we want to remove the dependency on zkclient +The problem is the web console plugin which we now can move to a separate bundle if we want to +Hows this as a requested solutionCan we rip out the whole cache make it a plugin define a default API and come up with some way of doing that as a part of the cache API accessed directly and proprietary to the cache implementation in use rather than adding lots of thisHeres the reasonIm working on an replacement for the memcached server theres already a java memcached client but its terribly primitive +OK im waiting for this release +Closing resolved issues for already released versions +This is why we intend to keep it as a simple DN +Alin could you check to see if this meets your needs? If so could you close this bug +We must have been inheriting from some file that has changed in updated gccIll add the include and we should be good. +Especially in these stacks commontests jsfvpejst mavenseamhibernate +One issue should suffice for the get problem +The top ancestor is returned back to a client only if all of its children are fully processed +Any future changes to the URI structure should be opened as separate issues. +Magnus would you be interested in contributing a patch +Ahhh right I didnt create the patch to add the book to books section +Btw if we choose to have a separate application I agree with the name the one hand push is a really important feature that may need to be on its the other hand I dont want yet another getting starteddemosample project +Are you saying that if we just dont put in any allowed values then all values will be allowed? Id rather have some level of control and not have to put in some additional code to do that +Youre right that in this case it doesnt matter +Opening a different JIRA to investigate the option of not storing disk info in the DB and have the details held in memory +Ive reviewed the attached patches looks pretty good +But it is not that big of a deal really as long as it is documented +This is probably a mysql server issue. +Is there interest in getting this back into trunk +Moreover the selection follows the same rulePlease close if youre satisfied +I think that work on this issue is done. +So there may be some problems with the patch mainly things like so use this with careI will wait for your feedback before checking this in the subversion repository +However would that be a big deal for you guys to implement it? Im using Solr in a web project as search solution and desperately need this feature along with the great grouping functionality +I will keep this bug report open as a reminder for us to try to improve the error message but it is not a bug per se +actually a +the same thing will happen with an id field +First I suspect that your issue is very similar to the original issue so your offering to help is much appreciatedIm not sure by what you said if the jar containing your custom appender was placed into the same location where the logj jars were placed +Mine are submitted under and it might be worthwhile to add those to your watch list since any changes to fix those issues may fix yours as well +This patch sets the process limits like RLIMITNOFILE in Traffic Manager when run as root before the TM changes to the configured usergroupPlease review and let me know if there are any issuesTested on Ubuntu and Fedora +Hello RohitMay I know what are the vmware complaint strings that can be used to replace the vbox specific strings in the rd Step +That means that the default file need not have these entries by default +Verified +Closing the bug per comments +Took out the sleep added in the first patch +Done +Leaving this open to track documenting how to build the UI plugin from scratch so that it includes all the correct files + +My current hypothesis is that the decorator tries to set up a new server before the previous server instance has been shut down properly +its right inside Karaf itself I was looking on the wrong place +I have attached a second patch which also resolves the issue while theoretically maintaining ABI stability. +At this point the requestresponse headers are already consumed +existed before this +Then we perhaps wouldnt even need to intern the namespace strings +All patches submitted and committed. +And I could not follow the comment exactly +The Accumulo tickets provide more information on the needs being addressed by this effort and its progress +Ive committed the shared params issue under and added Per to the Changes entryOn which branch are you committing Mark +Will file a Jira and investigate the timeout problem separately +For numeric columns in the catalogs length seems to refer to the max number of characters needed to display the column value +Fixed in r +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +A better implementation would use instead +running Christians Transport Planner app in deployment mode the all seem to be working ok +This works for me thank you for the quick fix. +I set the maximum heap size to MB and the build ultimately used MB +Correct version of theCould you please add the test cases to and provide a patch for this classThank you in advance +Really its just saying the divisor is off +Thanks for your experimentation and testcase +I see my mistake +It wont wait for ever if wait is set to true +Look for in the patch on +Attaching +We can do this in +I imagine we would have todo this a few times to get other references toapplication out which can be confused withapplication scope +Updated word doc for unit test +Thanks +I moved some grant revoke tests so may confuse patch programNext I will work on implementing authorization for Trigger View and Constraints followed by some upgrade migration and metadata changes +Changed status to critical as this is an essential capability +I found a better way to handle this specific condition without removing the intended behavior of purposely removing the end slashed when computing the local has a better test tests just to make sure i dont break any thing +in +Snapshot the Volume +Thanks +two commands are added disableReplication and enableReplication +IMHO this has been fixed by extracting the and creating a dedicated +That doesnt seem to be the case since the tasktrackers in fact are doing the heartbeats and hence they wont be lost +Bulk closing stale resolved issues +To make sre Delivery the same packaging for all project we have develop a complexe mojo wich generate site with custom reports generate javadoc copy jar and war into spcific dirs +Also check under the source tab for the changes that Steve did as we mentioned before you can back patch your version if u want but you are on your own for doing that task +Slava could you take a look please? It seems to be a problem of CDI model +Breadcrumb dividers are broken need to be rebuilt with spans as described in the Bootstrap docs +What do you think +In we removed the header trimming in order to not discard spaces that a client had intentionally placed there +Fixed in trunk with provided patch and one minor typo fix. +Duplicate of +If your context is being loaded twice its not really a scheduling bug youre rather initializing the context twice in some place a different issueJuergen +see patch attempts to make Locale and Version optional in the cache key +The RS didnt shut down though it just kept spewing those exceptions +It would be nice to get a fix for this problem. +I can write my own class +Now that we have custom norm encoders is this one obselete? you can just use to encdec your norms +So if I spot a place where it is being used directly but string concatenation is incorrectly being used can I reportpatch it as a bug +Im now running version and this problem still persists +Ill port to other branches after I see about a test caseBTW this specific problem of the injection causing the wrong is not likely to be a problem on jetty since there we create all the servlets up front before any requests are encountered +Reproducible in at of a bulk update all resolutions changed to done please review history to original resolution type +Im not sure if anyone is interested but I wanted to post the results of the internal test +Added signaling so that mappers can completed +Does this issue cover upgrading dependency to httpclient or just the documentation +Any luck with this? It is also failing with current master in generated apps +This patch would seriously break this implementation +There is an error in the url configuration on the server that causes maven to spit out some errorwarning messages but it works even with the old url. +is set to INFO +Thanks Boris +Also there should be some intro text about what the heck snappy is before launching into these detailsbq +I guess what Im arguing here is just to not forget the case where you use because your column key do is intrinsically composed of multiple partsI think something got lost along the way here because Ive been very clear from the beginning that we do need to support the dense case and that I prefer the component syntax in part because it does accommodate that +I agree the intersect looks wrong indeed +You can emulate the desired behavior by removing the abstract keyword on User +Slip to is available in JBDS +I think you are right that is what it is doing +failed on QA machineShould investigate +For example if I have an ETL cluster +Config and and stack +Edit the sentence Note the insertable false updatable false on relationships +Committed to trunk. +If you want to remove it go ahead and file a JIRA but it may have further ramifications +I took modeshape from the SOA build which is suppose to be the same bits that BRMS is pulling and I dont see the same issue + ticket could also cover other PMD rules +Attaching and with changes to three topicsM srcrefM srcrefM srcdevguideI noticed that in the Developers Guide topic the paragraph introducing the note also needed some fixes +Man do I hate that bug +This is a replacement xmap for resources +If no id property is specified manually the id is injected at compile time by the plugin +Closing +Hence marking this issue as resolved +correction +Thanks for fixing the as expected in the latest master +It is NOT a patch but a temporary solution for someone who wants to debug Doxia or to create new Doxia issue could break the backward compatibility with Maven or we need to distribute on the repository the new doxia artifacts +forget tomorrow +Attached patch +yes that one is a good example also ihtsdo +It would be great if someone who has access to the solaris machineenv where the diff occurred can test out by modifying the as Myrna suggested +Part of a bulk update all resolutions changed to done please review history to original resolution type +BenI am going to reproduce the situation with the contact example +I think it is a transposition error linking in have merged the changes into CP as part of +Im in the process of wrapping up a release which will hopefully go out soon +This morning I talked to Mickael Istria about how we could change where the Jenkins job puts the +Looks like it automatically does this right nowQuestion Have you checked for other locations where to make sure that adding another enum wont cause issues? I ask because MISSING should probably behave like FAILURE if not otherwise specifiedThoughts +further updates rev +bq +There are conflicts on +Thomas yes that dump is what were looking for thought it maybe be quite big so if you could make it available at some publicly available address instead of attaching a MB file it would be example you could use dropbox they provide GB free at the heap dump by the objects I see on the heap this seems the same as which should have been fixed before releasing you renaming the feature type by any chance? Can you try out and see if you can still reproduce the problem +The committed fix solves the issue +Thats fine +Different story forcheers Roland +I am fine with the change +png and +Removed excessed files +Thanks Deepak Committed at trunk r rAshish Ashish for committing the patchThanks Rishi for the discussion +Committed +JDK SPNEGO puts the Authorization header in the request so we can check that in the if statement that was added in +Assume I know little or nothing about your test runs +Verified at rev +I will investigate +Maybe Ill make the mapper classed private and remove all exception handling +reopened these issues so that I can move them to Future and remove wishes +Instead a querying api findObjects is added and checked in +Thanks Ramzi! +Well there was an issue there all right thanks to Arne for spotting it on and trunk +svn diff for the msvc diff for the msvc mscvch msvcencoder msvcgcgenuncomp msvcport msvcverifier msvcvmi revision rFrancisplease check that there were no errors with the patch +Maybe the driver close is not getting called for this failure +A HOD JT this should send logs to a user specified directory Central logging of job summary data that works in either of the above cases +Do I correctly understand that your proposing a change to the way a connector specifies its capabilities? If so I think removing the interface or changing the mechanism to be like is fine +Should I get this as a project despot? Might this affect commit access to svn +Still need a little more work to automate collection of the coverage results +I am going to commit this big patch and then create and commit another patch with just that fix +a pattern +Is that correct? Bear with me while I Then the only issue we have is one that cannot be solved +The rule annotation should not use an enumeration for the new property status but a string in order to not be forced to migrate to new version of Sonar when a new status is added +Thanks for stepping up help out the Apache Commons people +Closing this +satisfies some of the goals that the linked JIRA was meant to achieve +In particular the token ranges are not set correctly +Thats what my testing time shows +Run forrest projecthelp to list I was aware of that when building my site using forrest +The editor that I was using was inserting an endash instead of two single dashes +this will be a great addition to the cli +Closing. +This is set by the server as part of processing the publication of the document and cannot be changed by the user +Ldif exported from default Apache DS M configuration by Apache Directory Studio rightI thought you were opening it via File OpenIndeed I get the error when I try import it using the editor button. +That worked as a workaround +I think that the user will need to explicitly group sub queries together to get the desired behavior +you were working on this for M right +This is probably what should be fixed +see if Jerkins can reproduce the failure + +Cool +Proposed Patch for the issue +Well spotted WillemI have updated the javadoc on and added a note on the camel release page about API changes + is a patch that I would suggest for use in implementing the change +No sense having jiras all entitled Fix races in slab cache +This could be abstracted out probably should be abstracted into the Refactory +By it should explicit the column names do you mean that Hibernate Tools should add the referenced attributeIf not could you explain how to correctly setup Hibernate Annotations for a multi column join without using referenced? +Should have been using the zzh param insteadVerified all tests still pass c and java also added a new client test to verify that dup state change events are no longer sent to client +Ready for review +Now that work is well underway with AS all previous community releases are +Thanks Luke for the update and the actual fix and sorry for missing the duplicate. +I think it is still not working +I think close a tab is a very common operation +Thanks Ashutosh +So only one version of the bundle can be active at the same timeThis will be ignored by Felix +So let me describe the steps again +This obviously doesnt affect use of the var attribute +Smooks editor is no longer being developed +This was my case I had the credentials remembered from previous Betac build and even the deletion of my storage did not removed that Thats why I thought your fix does not work +Ill assign subtasks for module owners to confirm +Hi BrettI think you resolved wrong JIRA I suppose the comment you put here was intended for +This doesnt seem to be a regression or critical +Contributed eager batch cache has the same interceptor but the code is more +Thanks Bill +It was easy! Took me hour. +Any chance of a patch with a regression test +Just because it passes the existing tests does not mean the current code is right there was very minimal testing effort originally +Functionality added +normal pig input output format +Im just going to close this since I submitted the patch yesterday. + +Does it make sense to remove the config and schema links on the left after youve selected the core? They seem unnecessary with this new capabilityNo big deal either way. + +Fixed in trunk +though Im starting to lean towards disabling jline for windows +Its very strange that you run out of memory as ive run my tests more than once without that +Same code wrapper script +I would like to make more verbose documentation to the goal +The patch has been appliedPlease verify +Oh wait +The contents of this block was already fetched by the preceeding read call +Applied patches in r +Are these make files specific to BCC? If not should I name the folder generically as ProjectsWinBCCor ProjectsWinBCCinstead of +No followup since November +rev Add descriptive text to messenger docs index checked in new C API doxygen comments in +Verified in ER in both JON and the embedded admin console. +Thanks for the feedback Tomas +VivekAre you saying there was an issue while copying the data? You found some bugAm not clear on this JIRA +adding a patch includes a test case which reproduce the hedwig issue as thread stacks provided in this jira includes a test case to test ledger recovery with limited permits which also introduce same deadlock issue as hedwig add a possible fix to avoid call bookkeeper operation in bookkeeper callbackthis patch is based on s patch because I need to limit the number of worker threads in the test case to so I could let the ledger operations are executed in same thread to reproduce deadlock issue +Seems this bug was introduced by +The problem is that the grammar ambiguous like most grammars +i know debian packages can use scripts to do things like appending a line to +This is a potential fix +code freeze to trunk could you please add some details about the fix so we can properly verify that? as fixed +bulk close of all resolved issues. +Fixed leading zeros problem +is it really a separate task from? Or is this intended to represent the underlying AS issue +Can you check your java version? Please post the output from java version command +so we can pick the license that Apache supposedly permits in the repositories +In addition the default schema is generated for the Derby database +Verified in build both tomcat and jetty +Fixed for future releases by we fix this for the same way as you suggested it for copying the new files of and into commonslib +If there are missing deps create a bundle and use MAVENUPLOAD +Ive committed this. +All the changes to the WS Tester for beta were under the covers and dont really impact the user +I think Tomcat does the right thing here +Attached is which includes some simpletests that verify that substrBind now works correctlyI believe this patch is ready for review +This could use some more eyes please test if you can +Patch looks good overall two things Lets do protected +Changed component from dfs to just committed this +Popen class provides it but Popen are not available on WindowsSo Windows needs to use subprocess instead of popen +updated RPC Server to add authenticated to the ugi associated with the connection +bq +i think this is obsolete +Please try this patchThanks a lot +Troy just checking which of these are you choosing to do +In Axis we do not have method call invoke in service clientI think this is an Axis bug +The interesting bits are to determine override logic but I argue that is not the job of Bean Validation either +Attached another patch that reuses the existing problems sink instead of adding a new one and setting a precedent that users will have to check problems in many different places +Fixed in and trunk +No comments so issue closed. +asciidoc file in JBT more timesERROR Only first change to file is displayed within Chrome web browser +Also removed some commented out code +Wow! Did I really do that? +It seems cleaner to modify instead of adding more parsingmanipulation of query strings in +Fix is now in for this +It really needs to be made simpler in the future +I havent been around the or trunk for a while + can not be incorporated into the CTS until this is issues solved in released versions +Applied patch to SVN trunk in rev to branch in rev +This looks good except I think you forgot to add +Same type of fix just different numerical classThe current behavior is per specification. +Attaching new patch which applies cleanly to current TRUNK +Verified on. +Can you attach a sample project that reproduces this issue please +and the requirement has universal utility Agreed +once we integrate with BAPAT +Whats the new theory? Is this just the class to use all by itself? If so I can submit a patch to the docThe successor of goes by name in trunk +Do not include its outdated +Thanks for the detail PeterCurrent version of sstableloader seems it does not pick up sstables compression info before start streaming and that causes receiver node handle streaming data incorrectlyAttached patch let sstableloader to pick up compression information if available +I know i have done that in the past somewhere +npe check in +Please take over Oleg +tried to be more clear in the message text +After specifying +Id rather extend to add an option for generating a full URL +There were always some hurdles which are difficult to overcome and they made me wait for the next release +Attaching test case +Reopen Todd please if I got it wrong + this messes with the component hierarchy definition we need to ensure no problems occur +Done +Additional options were provided to enabledisable support and support in CXF to support using Jersey andor RI for support if desired +Probably but lets do that in a separate patch +Stuart do you know who could help +Now it does +to DOS format +I have committed this +In most scenes it works well +Once we get those jars the DDL importer plugin needs to be updated to reference those new jars +Im going to close this as invalid because I think our scripts are in fact POSIX compliant +Ill investigate this property as well + test case takes care of this +I like this clean up even more +Yes option USERONLYSECURITY makes sense to me too +are there any known issues? Besides the minor performance degradation there are none +This activated the DTD validator on the child pipeline if the validation feature was set to true +So all issues not directly assigned to an AS release are being closed +The documentation has entire chapters dedicated to each subject so each developer can make an informed +I think Werner was right on what the code suppose to do and why the codedoesnt work +Hi Serhiy Im closing this ticket because we both cant manage to reproduce it +Grabbing the big ugly component +Would be nice to post a patch here that would enable Jikes as well if thats supported by you have to do is change in the default to jikes +integrating it into jBPM was too hard +Applied the patch and run the again Joachim +Sorry for inconvenience +Dear KristianThank you for your advices and encouragement! I will check the patch again and improve it! Please wait for it +Restart after RM restart is already covered by setting max am retries to by the app client on app submission +in order to know if you can reproduce with the exact same version +is a waiting patch for in +But both of those look fine to me +Either a default sort field annotation or means to specify fields on cmd line would be useful +Malith appreciate your contribution +This should really be fixed properly if at all possible +So I think forkjoin should always be in pair +Must be done by docs and assertions to importersexporters clarifying the need for real files not dirs on input +I made a patch +Yes I just run the luni tests there is no more tests failure or error after applying the patchAbove patch only contains fix I will create another one also contains tests soon +The attached file does not contain all of the log you requested only the interesting part ie the and some stacktraces +Support for Username Token without to run in the parse phase +And if it cannot we need to reactdocument it. +Changed subject target version since the original problem is fixed and perhaps the goal of having a default security realm could be debated +Shelly have you talked to Jason about this already? Have you switched the Branch to use these already +I have committed this +I like the patch but I am a bit concerned about the scope of +This test case is not used to generate any java source files instead the xsd files are parsed and then the castor model of the schema is investigated +Its weird +A fix would be niceI understand that some people thing that the arguments should not be optional but unfortunately I have a published service contract based on what was available at the time so I would like to start workingThanks issue is already fixed in current trunk +Please consider these voteswatchersBug has Votes Watchers Bug has Votes Watchers also needs to go into the compatibility wiki if it is not going to be allowed going forward +In Cloudplatform varlogcloudagent is generated on hostInvestigating why there is no varlogcloudstackagent or anything similar in ASF Im asking is that you look on the agent host when the state is instarting and do a virsh list +More changes will be required for the data transfer protocol between the client and datanode to implement the full functionality +Wiped out the dir per suggestions which is cleaner +Thanks Stack +once we are on this here we should think about breaking out storing fields from the indexing chain +Finally it also repairs a couple of unit tests to hopefully make them more reliable +The first test completes successfully the second fails +server and agent to Charles +Updated local repository jars with Pluto versions in both and +Reopen and moving into in case we ever release that +Pushing again +It isnt currently possible to specify by CF so maybe that should still be addressed +Theres some management issues regarding Yecht itself but for now things are working fine and acceptable to the reporter. +Committed Rev +Marking these as fixed against rather than against. +Committed to both trunk and on deltacloud +Thats rightIt serves as a programmatic hint +and as a side effect would fix the one above as I wouldnt have to hack the lineup to only docs I can find for hacking is the source code itself although the docs has some hints but mainly on customizing for existing languages not writing new ones +Attached is the patch +Same behaviour with +John are you actually working on this +Can you provide a bit more context to this issue? I would need to see how your app looks like so we can figure out what is going wrong +Ive added a new version of the demo app +The approach addresses this fairly well by keeping more frequently accessed data available no matter what the GC state isAs you noted the equality check isnt an issue since youre never comparing the values in the cache +changed from plugin to import in +Thanks a lotI put jod folder to opt and specified path optlib in configOne more question +the old webconsole is deprecated +I think Ive started using version to indicate +Would be better if the formatters were refactored a bit to have the hard parts of getting default values and descriptions pulled into helper objects but I copied what was there for now +This issue seemed to be a test timing problem and does not occur any more +Bug has been marked as a duplicate of this bug +Chaning for for branch revision Committed branch revision +Igor the column tag allows you to define the maxLength attribute on a basis to the look and feel of tables +patch seems to have an issue when using the Psrc profile the common JAR files as MIA from theRoman is investigating this assigning JIRA to him +May be nice to check for things that make accumulo suffer and have a page on the monitor that reports on offending serversSome things we could check for swappiness OS open files xcievers namenode threads datanode threads zookeeper max client connections setting low walog space memory sounds pretty cool +Example job to demonstrate problem +This will be performed as part of but they dont affect normal users wishing to use Maven to build Roo. +The +There is an underlying error that the Transaction cannot be looked up at the specified JNDI location +But I actually am happy to do that myself for the app I have in mind +No patch was checked in for it +Very basic LRU cell cache +Ill close this issue now as no changes were needed to Roo itself. +This gets past the Lanczos issue by checking the size +This issue is preventing me from creating a usable archetype that doesnt require to modify the of the submodule +So all issues not directly assigned to an AS release are being closed +Wed also need an api in to set the final output directory for the side files +it will have following main model classes a history variable represents one variable for a process instance +And beacuse the uses model that caches the feedback messages until it is detached content of the feedback panel will remain stale +Not loaded at all +Verified on EAP in src zip SVN repository and decompiled from + not designed to be invoked directly and recommended command line is mvn sonarsonar or mvnsonar +a Without these or similar patches cant run in secure mode on various OS distributions without crippling them +Fixes +Is something like that what you have in mindYes +In future when people have identities that have not been identified its important to uploade the analysed byte code in order to be able to repeat the bug +Request numbers added in revision have added the request numbers in front of the requests links in revision +kindly available its fix in next apache flex release +but it relies on the fact that we dont keep anything to track the snapshots +One possibility is to use the ip octets like the +Thanks for pointing out +Please dont add constructors which are only used by tests +Verified +Wrong bug +patch attached get list of live nodes and issue takeSnapshot to all of them +Your change would eliminate the copyI think the real issue here is that the data hasnt been copied into the new instance before validation occurs +Hi EnisI tried your patch on and both status and kill works +on latest patch +Granted it was a pretty ugly if there is a previous copy there the jbossdeploy will succeed with an old copy +I just committed this to +Sumit can you please take a look? The install fails withbinsh c cd usrliboozie chown ooziehadoop vartmpoozie mkdir p usrliboozielibext cp usrshare usrliboozielibext cp usrliboozielibextNote the last copy in the command there is only one argument to cp + + attached patch +I would like any feedback on this proposal that the community has +So it is ok to have two deployer looking at the same extension but in different foldersThank youPlease take a look at the code +Sorry if I wasnt clear in my latest comment but adding the new alias is OK for me +t believe we did not have it already! thanks +Doh very sorryConfused about our new packaging stuff and I had both versionsSeems to be working fineThanks +Added a test case to demonstrate the implements Stomp which has no heartbeat mechanism to detect that a client connection has dropped +I feel much better about this now +Reopen if this is still an issue. +The patch looks good +Office Create Blank Sheet Data Get External Data From Text Delimited Comma Finish +I would however fix the error message to say use subquery for union instead of union valuePlease file a separate enhancement for the full union fix +Also removes iterations when a file is addedExtra copy still exists for files that are created at the beginning while reading fsimage +fix pushed with commit bdfdeddfdd +Applied patch to current codebase and didnt see any failure at test suite +This would be a very rare race condition where someone is trying to kill their job right after starting it +some screenshots to show how it looks live +Changing a font does not make it look selectedPlease just use common UI colorsselection mechanisms +Nice + +closing. +Looks good +I just ran into this bug any progress with it? Its two years old now and marked as major. +Im compiling with jdk running on Tomcat which is running under the same JDK and I get the error +This should work +As the console work is ongoing Im moving the parent tasks to the next point release +Old JIRA cleanup +Changing it unlimited gives me about a improvement in performance on my benchmark testing +for code review +yes files thats a lot of manual processingPlease send diffs +Emilio could you attach the patch for? Id like to take a look to consider applying it to as Im working on changes to the exports +It seems you forgot to set the parameter in maven settings +these classes will make their first appearance on geotools closing all issues that have been in resolved state for months or more without any feedback or update +Im guessing its the recent pkging changes Eric would you mind triaging +I actually was testing in the main app rather than the test code I sent you +I assume they go in srcmainresourcesimages +Ive assigned myself to this issue and will work on it for the releaseIt will also go in parallel to +Raising priority this would effectively reduce load on name nodeHaving the manifest file would also give us clearer view of the snapshot +test case works fine when is used in builder but NOT when +Please share the patch ID I can review them on Apache Review Board +Once you know the type and method there is already the descriptor API to determine whether a method is constrained +I just committed this +Do you have evidence this provides better properties than that? Otherwise I am not clear on the reason for adding this to the projectAs you say maybe reopen this later. +Its now just a Serializable List of Properties +Done API is implementeda testcase is in place. +indexinterval would not changeEDIT Well the meaning of indexinterval might change because even in a sparse index you might want to store something even sparser in memory +I tried against snapshot from March and the same issue occurs +The default is true +This is never going to be supported +This patch fixes that +maven project testcase stacktrace Bartosz heres a standalone testcaseYoull need to use and profileYoull find in the attachement and +Thanks Nicholas! +that is my question +there must be a conflict with the EL lib included in +We have quite a few projects set up like this +Looks ok to me. +Its simple to use alsoLets see what others will say about this +Maybe Im misunderstanding but I thought the plan for this JIRA was to add a more structured version number with majorminorpatch components +Would any committer help to commit the patch? Thanks +Even more strange if I run just that test from maven all is fine mvn test Dtest oHmm +sar be inside the ear project itself where it should sit +Check the code oras the XSL project for how to correct this +Does that make sense +I think the patch should be good for and will backport to later +Let us know when you get linestrings and polygons +If it were a true UTF issue wouldnt it drop bytes on the single column writes as well batch mutate +Old one is obsoleteThese scripts were tested on windows and linux +If your contributions are significant and of a high enough quality we can grant commit access +the first patch I uploaded was for methanks! +Charles reports hes found a workaround +attaching patch for a test case to illustrate problem with not found property on decisions condition tag +And the build used in JBDS is EAP not just AS +I think we can postpone all that for the future. +Fixed by forced builds are working again on vmbuild. +To solve this problem I need to be able to do some preprocessing before running the command +No progress since last move and there is weeks until the Beta release is do + thanks +Can you provide a runnable test case? The path should be normalized before so such problem should not may need Spring latest snapshot as I tested with the hsqldb comes with Spring petclinic exampleTest environmentWindows XP SPTomcat +Patch attached +Its all good There have been a lot of changes in this ticket to work through +attached is the thrift definition the two files generated on OSX Linux +Id like to see a private list created and then let the community decide how they want to use itIf you could make Matthew and I moderators of both of these lists that would be great +removed please tell me if I need to put it back added the following to the directory listingconf contains required port bindings reopen this if I need to change anything +Jetty support is in the tarball attached to issue to set fix versions. +bulk assign all unscheduled issues to next release +Looks good thanks Doug +I have tested integration with Mozilla Thunderbird and the address book search is able to find entries from a current build +Ive asked Burr to comment on this +This could be done if we change the clientdatanode protocol +call site generation means a certain overhead if you do it all the time then it becomes a performance problem +Im unable to verify it at the moment maybe one of the other watchers can +I dont know the right way to solve this +Hey Todd are you going to commit the patch? Or I could commit it +That will help me figure out exactly what goes wrong +Committed to trunk and branchOf note the repair uses a hardcoded interval of rather than our configurable interval w a default of +It seems like the code should throw a lock timeout error rather than XY +Marking this as FIXED since the issue is no longer present in FOP Trunk has been fixed as of IIRC +Of course we could duplicate this piece of code from into our own code base but in the spirit of making all our code changes compatible with Apache copen source I would like to request the community to analyze this requested code change and approve of it +fat fingers Ill do axiom from +Thanks. +I can no longer reproduce it and that makes sense looking at the commit +Should proxy in this tickets summary be changed to NTLM proxy +Anyways here is the patch +While adding this it is probably worth looking at and incorporating its performance improvements as well +Rebased patch to current trunk +for the get type when generating dynamic serde do you not need to translate types like int to inice catch +Probably you will need more blocks of parallel to the existing blocksThanksRick +Im attaching a zip file of a sample project that fails the IT under Maven but succeeds under +Removing as affects version +Myklebust all that will be included in will be the commits referenced in the STATUS file which is only fedabdadabacaeceIgor should there be multiple in the STATUS file should I be backporting based on your comment above and or start a new vote +But I was assuming that changes would require a constructing a new object +exception messages have been altered +I think the error message can be more descriptive +Patch addressing review comments +bq +RECORD MAP ARRAY with Super CF implementation can store single column ARRAY in addition to all primitive types and FIXED +The transform EIP is for transforming the message body as documentedTo prevent side effects Camel will do a safe copy of the Exchange which is passed in the method +Thank you for the link and the more detailed information +war does not include The aim is to provide a logj implementation that can be used if the end user implements logging with logjAgain I am happy to put this elsewhere but given that logj is pretty common it would be nice to make it work The first question is if we want to support logj out of the boxIf so I think the right approach is to add a new lib folder to webapp and put all the +Submit my patch + close +Attached file. +The zip file contains the complete sample +I will try to run jdbcapiSuite separately and see if itreproduces +Ill slim down my client and see if the bug still the user guide detail the minimal jar requirements for a simple client +Clean up the use of the global transaction log +The client would need a way to go from a block id to a +Fixed in revision in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Maybe we could add it to a faq somewhere +done +Once we have an ER build with the artifact in +Thanks Mamta for the review +Do you have an estimate on how much space this will save in NNs memory footprint +Typically branches are restricted to bug fixes and minor improvements +Thanks for the patch John and BikasIvan for reviews. +rebase on trunk +Not a blocker +Best regards George +form +i disagreelogs should be readable +A patch for is coming shortly +The patch passed system tests framework compile +If an existing class is to be used the error string has to be made a constant with the test referencing it directly rather than having a copy of the after on the latest good straightforward move +Can you find a way to test this without making the constant name lie? After all CODECMULTIPLEOF must be a multiple of three for a good reason! There +And if the beanio guys in the future is osgi compliant then we may not want to create a bundle at SMX repo +Probably unnecessary as can take place during authoring. +Trying for ve added a test for this to which will be in +when the dep is downloaded into the project it could simply be extracted and used by the project being is actually a bug in the deployer however that causes deployment of snapshots to get messed up +Might we instead put this in rather than in util? Right now things in the util package depend only on the JDK not on other parts of Avro +Breaking change for spring +as soon as I find out where Griffin was going with this feature I will implement it +on disabling RAID until changes are resolved +This issues is NOT a bug +This issue probably has to wait until either Vaadin supports this out of the box or someone spends the effort to implement something client side +I need to do the actual checkin because it involves a series of svn mvs that this patch does not fully capture +Yes the footer is currently taking the menu into account +Test failure is unrelated. +I think this is related to +Okay just let me know which to go +When updating the same file there was a check that avoided to update the file when the whole message was not changed but there is no solution for only headers changedIm currently investigating these solutions remove the load only the headers optimization and always loadparse the full message in memory change the file repository to write to a temporary file and rename it when the stream is closedI already have found problems with both solutions and thats why I dont have committed a patch yet +I am including an HTML output file for review with this patch +Muchos gracias! +I agree with Brandon on this what do you think Jonathan +Got what I needed. +The values to put in markers have to come from the input XML +Id be happy with it for that reason alone +committed thanks guys +marking this issue fixed since the changes for this were committed during SVN version +Yea i didnt look in before post if and versions are unused maybe remove them from list for JSF? If this possible of course +In the future it would be nice if we didnt support system level properties and instead only supported them at a namespace level +After talking to Christian some more we likely should add some sort of property check onto the lookups to determine if they should apply or not. +Committed +Thanks +Committed changes to trunk and. +Do we have any control over the ordering though? My hunch is that user supplied ones should probably be used in preference to Tika ones and the parser based detectors in Tika should be used in preference to the Mime Type onesOne situation where the mimetype detection is better is with truncated files +This may be okay? With the changes I have for we only update the L cache on a read if it is able to acquire the lock immediately +The set methods are provided by introducing a new class +patch for exclusive locking +Jesse can I close this now? you said youll check on this during your next release process. +Is there documentation or a plan sketched as to how to bring immutable recordsWhats the timeline to remove setters +I do most of my thinking after the work +The problem was that Java was not found which caused some targets to fail +An svn diff to resolve issue for the patch +How is that being handled? In my opinion the basic problem is that we are using the same service field for two different purposes +It is a if youre not sure if something is a bug please ask on the user list before adding a bug report in transition all resolved issue that did not see any further comment in the last month to closed status +People wanting to help should take a look at all of the linked items on this issue and see where they can contribute +in the screenshot i just attached it is more obvious what the problem in its ready to be closed. +No it does not work in that case either +Due to the huge refactoring Ill need some time to review it +Disregard I was looking at an older version +In if I am not mistaken the script looks for a +Just add in to the mix +I will attach a zip of everything that Ive changed +Fix a hang in the service listener that prevents the listener from stopping +Jason I assigned this to you as the main task here is to decide if we want to do it +Extracted out all messages and added Japanese messages +Nope I plan to do first and then rebase this patch on top with this new implementation being enabled by an experimental config initially +Committed at subversion revision +I dont think the are distinct enough to require a separate banner +Id think this is a blocker for +So this fix is going to make it to MThank you for this timely bug report +This is more like a Cant Fix +fixed in subversion as of resolved issue for release +Patch for the suggested addition +I assume this is xmodelcontent navigator integration that causes thisCould we somehow tell content navigator there are no children for obvious empty ones like + +the build process is very different now the javadocs have been updated to remove most inconsistencies and errors. +fix committed to another patch so http response headers are working now too +It continues malfunctioning on +I dont think we can limit the number of threads thoughbecause wed need to then implement interleaving doc ids +Upgraded to apache parent in rev +I think we can support the basic features so far +Do you complain about a being in there or really about that the cannot be cast to String? The last is since in Java domain nothing we can change +Id prefer not to give out root privs on brutus just for dealing with occasional bugzilla issuesThere are lots of services on brutus sharing resources and I think people with root on brutusneed to be familiar with all of the services running thereIf you need to be in the group or need anything else besides root Id be happy to arrange that +In this is correct you might want to make numDatanodes numIterations +Ill take a look +Merged with latest trunk +These are usually created by instantiating a route via the dsl +Probably this would be interessting to be added before moving to Tomahawk. +Patch v I plan to integrate +committed in r +derby have lot of issues on OS X +just add a more complete regex +With this I dont believe youll need to make all the changes in your patchShould I file a separate issue for this or would you just like to include this change in your patch +Moving this out to ER +bq +Patch is applied at +Closing all issues created in and before which are marked resolved +I will reopen this jira so we can do that for +Yes the flatpack component needs an overhaul for Camel +With the patch now passes with and nodes if Unfortunately setting now breaks it +patch added +And I think youre right on the pk metadata its definitely more efficient if separate queries dont have to be run +revision +I committed the patch as is +Ok I will create a follow up jira for this +Shall we postpond the recording of the policy to another future Jira +Committed to x and x thanks Andy! +Ill add the ones for the source and XML files since those are trivial +I agree they are similar but I am not sure it is possible to merge them +Final patch. +Response from KevinThis is very interesting because it suggests that something is calling into the pool from an invalid classloader +Jozef as discussed this is the first step for you +The process is because you need to traverse all postings for all terms so it might be impractical for larger indexes +Dan has made changes to make this thread safe. +Im not sure that some of existing coverage tools can provide such information +The extension is to add some randomness in picking the farthest point +The problem has been fixed within +These tests were done on out of the box server settings that comes with perf example and I did not make any change except the test settings in to send messages durable with batch size of +Just a FYI I havent been able to reproduce this locally +I also added a sorter to make sure that the list of operations in the WSDL selection dialog is sorted alphabetically +on v +Sorry the patch is against a copy of the source that ive mirrored on our own git repo so the file paths may look funky +Fixed in r to modified the Axis JMS transport to use Tuscany listenerthreads when set doesnt work +it feels like working in the dark a bit +The issue might be updated a little to reflect the work youve done on datetime support +If so we will not need +Guessing this has been handled for a long time. +and both pass in batch runs of times +Hi DanielsYour fix for Reporter is right +Attached the concurrent the second tar gz file with framework test sample code +This issue is not happening anymore in any tested branch +Ok that works! Thank you for the magic +Hi Jakob are you working on a patch for this for? If so many thanks! Im going to mark this for +Invalid code reverted and simple test case added at repo revision rPlease verify this fixes the issue +hbase and datajoin +to during tidy up prior to +We have now acked an entry written to the old bookie +This patch applies to the trunk +did a clean buildinstall and run mvn dependencytree in to verify is gone +Ill commitIt would be nice if you can provide also unit tests +Patch committed to trunk. +It seems that this subtle bug was introduced by where a method call was dropped +We should disable synonyms in SESSION schema +This will become more of an issue as people start using Forge to reverse engineer from existing domains because existing domains will have Bean Validations like on themAt the moment people are generally playing with Forge for new domains but I believe Lincoln is intending to move more and more into the reverse engineering space +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Sorry nope dont have one. +This isnt a great equivalent as it isnt colloquial like the Catalan expression but I guess it will do for now until someone comes up with something better. +Removing the solved this problem. +Uneeded attachments removedYour work to be distributed into the LGPL +Coincidentally we were just about to submit our own port +sallow hbase to do additional iopsallow hbase to avoid additional iops +Hi I am new to HDFS dev and I would like to take this issue as my first +Laura I presume you will be doing the enterprise docs for the Infinispan EDG tech preview +Please review +For us if we lose all cache replicas its not a big deal +Do you think is it possible to use just in Task runner? If so lets do like that at the momentIf we can continue to maintain low complexity of syncServer and no big differences about performance lets get rid of zookeeper +We can either remove them or I can try to combine all modes of roll and recovery in one new test to check for any interferenceAlso should we have a check for illegal values in getTopic methods in Utils +Please review this patch +Please try and validate it +Is this on windows unix or will this error occur regardless +Even although I didnt used themPlease have a look at this before closing this issue +No longer an issue +Does this represent a change to the design or is there a yet to be built Pig classThe patch breaks the Pig class interface into several class to register or define in global scope to compile a Pig Latin script into a object +is that a possibility? And if so whats the best to way avoid this failure +This will help in understanding exactly what is going on +Yeah I was just thinking about the use of ZK could be very similar +I would really like to set the store api cleared up in general with this +Hi Marvin Thanks for the quick reply and sorry for not reading properly +Thanks for doing that test +address nits +here is the right patch +That used to be Javas default but I thought may have had a higher default +i was getting confused i think the errors to do with annotations were only in the geoapi code not the yeah +ping +I suspect they did with v as well because the clearset thing isnt a bug its just inefficient +A separate pull request will be sent for EAP +This issue needs a forum thread I suppose? Cant figure out whats the real issue +So we will have to compile internal lib with the new version of aspectj +A suggest approach to resolving this what do you think of this patch +The missing methods have now been implemented and this error no longer occurs +I just meant that if the pool was told that factories could never return null then it could legitimately reject null as a returned object +I believe that Ive already reported this issue in +this issue has been hanging around long enough +Identifying that this error still exists in the latest released version of the eclipse +Already fixed +Peter can you take a look +Lets make sure we get this in for. +Also will be fixed wrong mime type recordmifapplicationvndmif mifapplication. +ORIGINALPATH The path to the original fileORIGINALFULL The full name path name something with BEAN to invoke a bean method that can compute and return the filename +Would certainly try it out +Thank you RemyI needed some clarification before I close the. +Verified and closed the bug. +I was using the wrong version +r caused complete test run failure reverting it in fix again in is working well now thanks +If you need the +Patch that includes all changed mentioned in the summaryTo test the error page at work the most easy thing to do is take a jsp and introduce a compiler error +This is a quick solution that might be taken out later in favor of better solutions like the one you are talking about +Patch doesnt apply clearly because of Ms within patchdosunix utility applied to patch cures this problem +bq +Ill send a mail to the list and ask them +Now we need a little bit of documentation on how to use it +i applied the proposed updates and introduced the message alias. +It seems p has a different NetHTTP implementation for than whereas in p theyre unified +Also please note that I am executing these scripts from java code and I use +Fixed in JBDS and Test was added to to add back catalogPropertyschemaProperty to template files to fix it +Relating to +Giving patch a shot +One change is required to the nd example make the privilege available at after the object is created And show that when the PUBLIC level privilege is revoked Derby currently does not use available privilege +If you notice in the assignment it is doing a on to that thought +Submitting cleanup patch + can be instantiated with a instance so if your object pool implements that interface it should work +This was not implemented yet +Removed paragraph about replacing +In the PG case you are converting from string to timestamp +Binary feature is generated out of +Im a little lost at what this patch introduces that is imposing? Ant itself has maven support as part of its trunk code base so its clearly not too imposing for themIs your issue that this patch introduces things that get in your way somehow with using ant to do builds? or are you against committing this due to your general concerns with Maven +mvn package Pbintar creates the new layout +this simple patch bumps theThanks for the patch +toDid you get your answer or is this still an issue? +Attaching and with the information Kathey recommended +I didnt get a chance to look at this today and I dont have access to Oracle from home +It seems also that when we have this failure the class files are not hitting the filesystem seemingly though the cmpile of all hbase java files completes fine +We might want to think about a way to provide expected output files for these kinds of examples +Mario is it possible that you upgrade the ibiblio maven repositoryMany ll cut a new RC in the next few days and put in on ibiblio too +bulk close of resolved issues for release. +Closing. +Is there a way to somehow unify this work with the existing callout mediatorThe issue description says that its difficult to create chaining sequences with the existing send mediator +Closed due to the release of Tiles +I ran and derbynetclientmats with jdk +Would this include for example dfs replication setting? If so yes this improvement would be very helpful +This patch makes local transforms work in the presence of a custom class loader +The patch failed these core unit testsThanks +To make it truly in revision I enabled the ignoreBogons feature that prevents tagsoup from adding encountered unknown HTML tags to the schema +To make it doubly useful to do this I fixed HTML tag usage and grammer in these files +concurrent open mapfiles set to and xceiverCount should larger than it so set to +well +To quantify this I can run and report numbers for something like stress on an EC cluster +Now I have seen this failure on a machine as well +I was trying a litte more complex transformations than what was shown in +closing as per marcs recommendation +skinsxyzcss in the tags attributes which is a problem when you bundle files +I tried synching back to just before the checkin below but had trouble building +Thanks for reporting it Alexei +It is also broken in the sense that we ocassionally see it fail with since there is no synchronization construct available for the to use before the moving the temporary files to their permanent abode +I suspect that circular dependencies are common in configs +doesnt need any modification I changed and precisely to test the behaviour of the handlerHave you made a manual diff of testincludeeventcompare and bintestincludeevent to try to understand what was going wrong +No new tests needed since changes are only javadoc and additional logging at the DEBUG level per JIRA description +The improvements extracted from dedicated the the phase improvements +We will then need to document how to best provide secrets to camel like properties files with permissions not use them as arguments in command lines etcObviously the credit still goes to Lorrin for reporting this +You know what +Glad to hear that all was functioning as expected. +However all this spinning in the reactor execution loop is inefficient +Documentation for demux largely adapted from what Bill Graham put on the wikiCan someone review for accuracy and style +Jonathan Ill get the ring state checked at work and report backBrandon Looking at the patch source itll likely take care of this issue +This appears to actually be working correctly +I think thats the best solution +Seems correct will check the up the good work +as additional demo to steps to reproduce +Transaction support for Tomcat will follow independently in +How many contributors have there been to Luke outside of the ASF? I know Andrzej and Mark +The adds the gbean for sharedlib into the assembly +But then Im not sure I completely understand yet what youre proposing +I do agree it should not have been changed and will revert it back to. +Just fixed it +Tom anything to be done about this or should we bump it +for this patch just focuses on findbugs error +Ive got some suggested improvements and the UI could use a little work but there is no reason not to move this work over to the trunk so we can polish it up for a release +Fixed in rev +This is a minor task however +Nope I wasnt aware that a gadget author can set the id explicitly on embedFlash anyway +Hi AaronI think patch has to be applied on branch +API now tidied up as far as possible whilst still being compatible +A few remaining undocumented directives are listed in the last section +okIll produce a patch +The fix is to stub out the RuntimeTheres also another fix in this patch related to the fact that changed the expected exception stack in the case of an error from transition +hopefully we have these issues covered +Thanks for the patch Sergey. +This is likely caused by see +And the user will be conscious about the existence of +In the meantime if you have only one node having problem and unless your replication factor is you safest and probably quicker way to go back to a correct state may well be to replace the node with the potentially bad disk +committed to closed with release. +One thing to look out for is that the sockets might not get broken depending on your network hardware switches routers etc +As such it is a very natural extension of groovyc that both fixes immediate problems and simplifies future efforts such as all without changing the class loading behavior of existing compiler clients +resolving as fixed +How is this? It seems to work locally +Springs Web MVC module is essentially a request dispatcher framework it never was a framework to begin also that Springs Portlet MVC module and the Struts support module are not included in the main either +You should be able to get a patch from the change set when applied without waiting for the full release +Ive committed a fix for this in r with a modified test that should be a bit more reliable +commited the fixup thanks +it does +In fact the server should not log the exceptions that it is sending the client actually! may be except +Adaptor will not finish tailing the file if collector is down +Navigation icons are in place to be pushed live in the next release. +Deleted the data instances and the lock file +A patch trying to find a form that keeps the original intent but respects the idea of a +haproxy was not installed on the template +So trying hudson +It is NOT reproduced when writing to a pdf that only contain an imagePoint leads me to think that it might connected to the embedded fonts in the document +So I guess here we can reuse the nameservice loading code from JDK and use allpart of the nameservices to get those possible hostnames for a given IP +bq +I have a similar problem with an application I wrote the service stalls on reboot but works fine if I manually startrestart the service +This is the last patch that in part solve the issue on product tab but it is not equal to the others style +The problem was that the new methods in used the associated with instead of the associated with the file argument +Close for release +Also affects version is wrong and the fix version should be removed imo +Thanks for the patch +However the root cause is op once again +Added class from JDK to and adapted referencesFixed in Rev +It will hang +Last Saturday I had verified it and its working +The app is executing queries and this happens at any time also when all connetions are alive and wellIn any case DBCP is used in my application as is evident from the stack trace +make sure it is not mentioned anywhere else in WFK still appears in the +This is fixed in and it seems the same fix cant be applied to without generating some new test failures +This affects the reflective signature for the method +Fixed in r +Is that any failures only failures of a certain class of exception? +bq +By eliminating short data field we get bytes of saving due to byte code alignment on a bit JVM +Good observationYour current and patches still remove the log messages +You can safely remove confidential information as long as the overall configuration structure stays the same +Thanks Remus for patch +Added my environment raised Priority as the plugin stops working properly in my case +That provides the name by which the deployment is known to all the subsystem deployers defaulting to the name value if not set +But we should open the corresponding CDI wizard +ok cool to hear its working now +saves us from all this! We should compareTo in +patch is committed to branch +I dont have exact benchmark comparison on hand since I never really wanted to compared it but its quite apparent even with map tasks +AdrianJust to be sure to understandUEL calculation is done with methods when field type are? +Resolving the change in the title has been made heading for years further work can go in a new JIRA. +So all issues not directly assigned to an AS release are being closed +also overrode the logic which will also be added to Teiid Here again all the teiidembbeddedquery classes will be removed for Teiid Designer as completed +I applied and commited locally that patch on the latest trunk +BASEDIRoptasreceiver should be BASEDIRoptxxx sorry for you create a minimal example which shows the problem? May be on so I can take a deeper look into it +Hi didnt have a try because were using APT now +Attaching + under Windows locks a number of filesdirectories preventing the unexplode target from deleting them successfully +Committed revision +Then how is the performance of KB? Is it similar to KB? If not we probably should increase the value +Thanks Amareshwari +Closing. +Fixed with latest commits +Hello can this issue be acted upon? Or do I need to provide further info +A fork that is maintained by could use git modules for exampleIn we have a case where API is used and exported publicly and one artefact uses the same API in version embeddedprivately +Im unable to reproduce this failure when I checkout master and run the tests locally +Useful debugging feature +defer all X issues to +Ive replaced with calling the equals method in SVN trunkThanks for reportingMaarten +Not a bug +Patch applied with fix for checking JMX state before trying to access the management context +schedule for main issue we have to consider is +test application to demonstrate the to check this bug +Nitay I hope you will look throw my changes to maybe something can go to ZK testing core +I totally agree the static approach is a real problem +Here it goes the zip file +This project references the project included in the next attachment +I would still vote for applying the patch since it adds configuration flexibility namely a way to affect all pauses via delayfactor setting. +Yes this is complete and can be closed +The code changes look good but there are a bunch of conflictsCan you refresh +Keys could alternatively be stored as nodes with each key having a map with the key and value stored in it +Kristian thank you very much for the review and the testingI committed the patch to subversion as revision +Problem has been resolved +Revision +Only issues that are actually fixed should have a fix release. +Attaching a simple patch +Ill handle that separately. +take a look at this issue I found that it is difficult to add a test case over it +Whoever fixes this please add a to ensure the problem doesnt return unnoticed +Reopen for. +Once a release is rolled it assigned a point number that never changes +to trunk +I redesigned the resource test class so you can use it in a subclass +is needed or at least one slfj concrete jar since MINA uses slj +There is of course the related problem that having the NN resolve any links is wrong +It dont work properly means few of charactors is correct in HTML source and other is wrong charactors display in HTML source same as xEF xBF xBD but no in +What matters more than anything else is consistency +Attaching patch with the above comments from DevarajUnit tests passed on my local machine +a first rev of a functional spec for this feature +This is where some tracing logging code would be useful +not a high priority fix as pluginRepositories are and plugin repositories are fully partitioned in. +Tested OK +Adding a RELEASE NOTE in the description field +Without the patch the new test will fail +Miroslav could you do me a favor and try it one time again with +Participation on the code is encouraged via the standard patch contribution +MaxWe cant assume that people will be using forge in eclipse but we can surely provide that integration with eclipse for those users who do +The compiler now always mangles type parameters. +I personally dont feel a need for a separate project here as of now +I introduced that when I was having more LOG output for debugging +I was referring to the jmx bulk load call +This was valid XML and conforming to the schema +Is there any standart format for that stuff? what are the minimal information to provide in the responseThanks for the idea +We can get an overall high performance for all integers if we use different algorithm for different integers +ANTHOMElib contains the file basedir for the Ant project must contain a mysql jar file +Moreover we have to make sure that more than one datanodes can run in the same machine when this feature is disabled + in and +Added testcase as test which reproduces the bug reported in this issueGrant would you like to review this patch? It contains minor API change so youll need to revisit the custom implementation you had +Thanks Rick for all the advice +Can you reviewthanks looks like a runtime problem TCP connection towards database server is broken please check to see if the problem is problem shown in log indicates similar issue with +Added test in to test for thisAnswering the question from a few comments ago I dont think this patch needs to wait on becoming read only +Hi Drew What do you mean by one configuration at a time? Do you want to say DOTNETCLIENT connects to PHPBS as well as DOTNETCLIENT connects to DOTNETBS? Obviously it is not possible +plus is passing locally +I am not familiar with module properties that seems like workable could remove the default code based configuration and add the same explicitly through xml configuration to provide a default template with properties if you want +Can you create a new issue for this with a test project +ant test was run manually +Guava uses contains for each element using the objects equals method with that +for instance etc +Any thoughts on the use case MichaelLarsWorth doing? Or push people to actually log the results in the client? If its worthwhile what granularity +Well have others review as well when you are finished +note ant Dlibhdfs passed +If we want to make this filter support turkish lowercasing in the future for all of unicode not just NFC composed form we cannot do it with +fixed the flow of the whole sentence not just that phrase. +The zip with kb is the new attachment +I was having a similar issue switching from maven to did the trick for me +ok +Dunno thats why I wanted you to review +This is as designed. +This doesnt preclude hooking with a security manager later or adding alternative hooks that do something aside from throwThanks patch attached moves and to common +pending same patch for our friend jenkins to take notice +just committed this to trunk and +Please review this +Updated patch +will not support mavens pojo approach anymore as this conflict with junit tests +subclassing via an extends keyword +Thanks RohiniAnd thanks Cheolsoo for you consideration in It is just fine that Rohini took this one as it was her who got me looking at these tests in the first place +Sorry if this isnt the right place to ask this question +Thanks John With Surreshs and my own I have merged this into and to add the new file? + here is outdated +Moved off of M Release +Im therefore closing this issue fully aware of the seemingly broken build but confident that this will be resolved ASAP as per my justifications as above. +Resolving as implemented +I suppose if you just cant describe it dont write anything. +Ill to verify +Actually could probably write a that worked for anything but the would always have to be +Given the previous comment are you OK to close this issue +Ill take it out of the Params classes and only deal with those classes that are part of this patchNew patch forthcomingCheers HossI went ahead and got some time for an updated patch +I think triggering an Exception in the process would be the correct behaviour in this kind of situation +That should cover Justins needs as well as retaining backwards compatibility for existing configurations. +Hi KimI had not noticed the commit comments indicating that you werent the original contributor of theIts good that you gave credit in the commit message where it was due and wonderful of Raymond to help outStill its a little unusual to not hear from the contributor in personPerhaps we can get a comment from Raymond indicating the contribution is officially granted to Apache? Kim and MyrnaNote that almost all of our message localizations are prepared by people who dont appear anywhere as Derby contributors +The issue can be closed +The zk failures seem to be related to jdkIll take a look at those +If i recall correctly this was fixed some time ago +The test works on all platforms both on JIT and interpreter +Fixed syntax error in the sample code +If you can do that you know how to do a netstat a p tcp and work out for themselves what they are doing wrong +Matt do you mind working on an unit test as well +for the patch +Other code coverage such as the emma maven plugin can easily be added to the build as needed. +This is needed for +Thanks for the quick fix and the information! +Yeah Im sure I meant to remove the underscores and just didnt quite catch it in the final patch thanks for getting rid of those. +It also affectsThis is a game dude +I dont follow you +It looks to me like the UIMA contrib was committed before was released using a version of the jar and then never upgraded after the releaseI think it makes sense to switch the version of the jar in Solrs source tree to the released version and then stop publishing a jar +Youre welcome Jaikiran and thanks for looking at this issue +I have implemented the algorithm described in Pregel paper without using Hama to get a feel of the algorithm +It seems a little strange to me that a can take the whole things down without any way for the user to handle it +the update to revision is still done +Please post a stack trace for and NPE in second issue about please fill in separate Roland +Moving to CR since we have less than weeks till Beta +fixed in rev investigation reveals that we are supposed to throw if the ejb client view is used +bulk move of all unresolved issues from to +good morning Michaelgood points +On we can actually alias rasters by changing the name of the is dead but if you still want release a with this patch go closing all resolved issue that have not gotten any more feedback or comment in the last month +Also agreed my bad +Dispatch count shown in summary line on Job Details page is sometimes too big +Do you have anything at the log files that can hint what is going on +Maybe this could go in the wiki instead +That was quick Thanks +You need to add a +What good does it make to have a cookie policy in place and to not parse cookie headers? One would send a cookie to the target server only to end up ignoring it when it comes back with the response +The component works correct but when a user uploads one or more files then there are multiple input tags with the same id +Use adaptor is good +Error message changed in r +I got pulled away from this a while back so either the docs need updating or the prefix fully implemented +Removed tag Will rename to s this? Shouldnt it be closed +At the same time it does not say anything about TCCL and classloaders in general which you are referring to +Once is in will revisit again +how about via user agent? What kind of user agent string is coming up through your proxy +Apparently fixed updated gold was checked in. +Verified by Mark. +Yup I leaved this intentionnaly in archetypecreate as archetypegenerate encompass the archetypecreate it obviously works with archetypegenerate and new kind of archetypes +Also we may be unhappy with md checks and caching +Ensure that you have all of your change as commit which has the correct commit message something like Custom users home directorys creation fails in +Added code to set in the second patch I also setjavaagentGERONIMOBASEbin GERONIMOBASElibextJREHOMElibextto match current geronimo start scripts +I think we need to get a consensus on how we are going to handle this issue +Marking closed. +Attached updated patch file +If the latter then no change is needed +I do not think I follow your comment here +HiAmila can you please clarify what you mean by a databinding? An example would be good I am not sure what you mean by this? Is this information held in the WSDLThanksStuart +Repository configuration +Since uses it must be compiled against +I added this feature +Any thoughts Joris? Otherwise Im inclined to close this as Wont Fix this as Wont Fix for the time being with the customizeContext hook as recommended solution. +fixed in and +GertThe patch looks good please go ahead to apply it into the trunk +reproduction script is +The column names utf assumption was a bit too hardwired +For code maintainability purposes we should deprecate log versions that occured release back seems like a good option +Ill take a look at this in a bit but no promises +Tested and ran fineCommitted +This is because if we pick the first replica to be the largest broker id the rest of the replica ids are smaller than the preferred replica in this case +Seems like robert wants to roll a soonish +Thanks +At least that has to be supported by Mimej +It also includes a combined user add and edit screen +Patch +r +I think were only waiting on confirmation that JNA releases have been spun for us +The issue was resolved +In my case I have the same error message even using the W flag +We are including it to show how cool Rich Faces is +Made the changes in SVN Rev + +Thanks for review. +I cant assign to myself for some reason btwFor the recordTheres a workaround of adding AS module to the apps deps intoDependencies AS classloader will then have msc classes at hand. +Thus the input buffer is never consumed as the encoding of the input to the output buffer in fillBuffer never succeeds leading to the infinite loopWe should check the buffer size in the constructor and fail if it is too small for the selected charset +Updated patch includes a patch adds license header to +Brian Your updated version of the patch seems to have broken the caching mecanism +Thanks for the fix Stepan. +NovYes of course you are right well keep it open until everything isThe nowtoday issue is solved with the new plugin but still marked as false +Sorry is resolved +Patch upgrades existing docbook from docbook version to version docbook structure and build environment for linux and windows are available in trunkdoc +Ive not sure this is valid anymore please reopen this issue if it is still a problem. +Is this still an issue given the remote deploymnt thats been added? Moving to are you talking about? +But this should go into noYes +Thank you! +should have been a solr should have been in solr +Currently these fields are sorted by whcih is inefficient +I expect youve hit one of those situations +I can easily set up maven to pass in the and have the tests only run when the binary build is also enabled +Attaching a patch +I had a quick Google and there doesnt appear to be anything providing this function in the WWW ether +Please let me know if I have missed somewhere +When the version is finished we can commit that and close the issue +Fixed on and did a minor additional modification + +Changing components +Hmm do we want to put a lib named beta in a major release? Can we get something more official released there to includeOtherwise +I see thanks EvgueniAnd why is it a special instruction? Why dont we just mark operands as +Seems fine everywhere else. +We might want to create a separate ticket for that if theres demand. +Now that work is well underway with AS all previous community releases are +It will take all three weeks till code freeze to fix it and that means Alex Kazakov will have less time to take care about other issues +The point is that Strutstags should not be allowed to generate HTML that does not conform to the HTML spec +We chose to make use of the Commons Logging API to give people a choice of whichlogging implemenation to use +This test is still failing even after applying the new MTOM flow. +The class name is available in the primitive descriptor of the AE +Ill try to get this done in time for the release +from a join b + +Thanks Bruce +I agree that we can record the timestamp when an open occurred in namenode memory mark the inode as dirty and then allow a async flush daemon to persist these dirty inodes to disk once every period +Removing fix version since this issue hasnt gotten much attention and doesnt appear to be a priority for anyone +or commit as two separate patchesOne issue one commit is cleaner +One small issue is that all the continuation markers ended up at the same lineEXTERNAL NAME +This was before I added the exclusions but after I upgraded to Ant +Thanks very much Dyre. +Committed revision . +Getters added in r. +Thanks Raghu. +the text is plain HTML + +lgtm to trunk and branch thanks Ryan +Will in context with +I found icons x I attached itsWhat type of icon images do we need? ico gif png +Ill close the vote shortly. +Thanks +But it used to apply per volume and it will be if reintroduced as wellThe remaining action of discussion here is to decide if we should go with Percentage opt one value across all disks +The attached patch does the following checks if the ranking algorithm is with set to REMOVED finds all indices with in the input arrays removes the corresponding rows in all input data passes the modified data to the ranking algorithmThis should ideally be handled by the but updating the interface would break compatibility afaikThe class is the only one in CM that uses a to rank correlated data so it is a kind of compromise imho +As far as I know they dont have any such override setting where the developer can simply set the domain they want for the sessionclient cookies +Andrew can you review again +Arun thanks for working on this +I think theres two other options +A separate plugin could be firedThe only thing I really dislike is the probable command linemvn clean warexploded cocoondeploy cocoonxupdate jettyrun o truecouldnt we make it any shorter +Thank you. +The remaining work for this issue is getting tracked in +Can this issue be safely closedSure +is not fun +Hi Michael option and option patches are uploadedYou do not need both patches either one of them should work +My only concern is the wait time between each scan is seconds +and languages removed from added explicitly defined that is rdf List in +Lets pull it into if we do another +A patch for this +Removed the plugin from the trunk + projet projet fix of fix this issue +The wrinkle here is concurrent schema changes how can we make sure each node uses the same column ids for each name? I see two possible approaches embed something like Zookeeper to standardize the id map punt let each node use a map and translate back and forth to full column name across node boundaries +The changing is not a problem however since the iOS exec function caches the the first time +changes in to check access permissions also when parsed via rev +I took a cache diagnostic trace +In a set youll probably get an NPE since null isnt hashable +Please scrub with debug logging enabledYou can set disk io mode the same way you always did its just not part of the example config anymore. +This is intentional and by design +I dont see what the issue is +Can you please update your svn tree and regenerate the patch +two clarifying points this is not for +portion +are we going to fix this for? It would be nice to have this in an unit testDoes anybody has one that we can include +The patch looks dont have commit rigths +Its also not clear if the data being added are merely for the one layer or for all layers in the current map +I have tried with and without versions for the without any luck +After fixing this everything works okSorry +minor comments on per comments +I left the optional mainClass detection in place for now but am still open to removing it for the sake of simplicity +In the end it is not a good idea to use a lossy to compress the data +Reproduced locally +It fails on SQL Server +lisa +Yep sendfalse is the same as not having a dlq configured for a destination because the message has been poisoned ack by the consumer so it has gone from the queue. +In fact I am doubting whether we are doing the right thing in the current code base for redefining +an additional test for that I forgot to move from Solr +Could you modify the patch so that I can test on +Committed revision +now uses the underlying peek instead of trying to read +Bulk close resolved issues +Patch w test fix +The latest code doesnt have this problem anymore. +It is not missing it is a nonsense length zero +Closing resolved Release +batch transition resolvedfixed bugs to closedfixed +is much more concise than Cppunit +To use it you are supposed to set the attribute to the font name in the collection +Patch applied +Fixed +Heh +Couple of niggles though that I might have to change +In the previous patch the wasnt aware of this and everything was done in the but there was a valid concern that the balancer implementation shouldnt do things like META table update +The patch attached contains code that groom server periodically reports back to bsp master +Thanks +I will update the matrix accordingly. +The parameter name provider is currently scoped globally to however is should be treated the same as and Co local to the Validator +webworkpackage name extends namespace abstract external includes delimiter is webwork package name class name package class webwork name package delimiter is webworkresult for name value package type webwork name package webworkaction name include namespace converter param success error input webwork name value webwork name value webwork name value webwork name value type flowing is method level +They will have to access these via the JON API +on Seans proposal to set minmax +Thanks for taking care of this Adrian +Now with this one Im unable to reproduce +Would it be nice if we also had methods in Context to do these things +Should be code drop +Updated patch against the latest trunkTests were run manually +In what component should we this? should be kept to a bare minimum of its +This fixes the bug described in the previous comment +Logged for Kevins is now working Len installed the patch +Remove repository since is now in Maven Central repo. +par is wrapped in a +Ensure the downloading tasks are completed before verifying the results +I would vote for changing the fix version to and close it as trunk is where most of the code clean up is taking place +I dont understand why it is failing +once the tests pass and you commit it I can do the rest then +Thanks for the report +Issue has been resolved for a long time waiting for potential confirmation from the reporter +Do you have an example of where this has happened? Ie is it worth the effort +Attached a test case file Apache Java license header to the test case file +Note that the header issue I see in the Ignore test is affected by a timing factor +This is also tricky how would you use this to flush cloned ram indexes? Cant there be deletes coming in against those during the merge +on addendum +Im not too familiar with Trevni and didnt want to take a risk so I added a change in to preserve the original behaviour rather than update the filenames in the test +The JIRA issue is not the place for this discussion please take it to the forums +META +Ok got exactly the same problem NPE +If you resolve this in wouldnt that be OK +This is done in Branchx will do trunk tomorrow. +Conversion for screen +As I said who goes first doesnt make much difference we just need to have it clear +While it never happens on embedding +This patch uses and which is available as part of patch inThis patch introduces the jdbc implementation of and extends from and extends from +Heres another option nice and simple +oops +Rob im not sure what it is you want here? creating a empty plugin seems a bit weird? Didnt you have something specific you wanted in it +I guess the point is that if the featureType doesnt have a geometry it shouldnt do a hard failure it should just gray it out on the list +The RM was setting aside those resources on behalf of that application therefore the application should be accountableBugs in the RM or scheduler may cause extraneous containers to be allocated but they are still allocated on behalf of a particular application and should be charged accordingly +If you can verify the console still works eve wen using these plugins please lower the Priority +verified in JBDS again +Apart from that the trunk patch looks good to me +from me +At this point there are small in the Store +For patch based on branch you can still upload to reviewboardHere was from Jonathan HsiehAssuming you are using git and have a single committed patch which is yourcurrent branch Ive found using gits git HEAD and usingthe generated file with the project to be the easiest + Actually many people around me questions about this feature with processing +So far it seems that sharing a connection is better than a connection per thread in YCSB +This would require that performWork also received a taxonomy writer +I just committed this +Verified +We need to reinforce the requirement to have all type declarations and shared assets loaded before a ruleflow that uses them is loaded. +in the Annotation Styles view then all annotations of the type are highlighted +Dart +Youre going to have to do better than that if you want support +It was also suggested that a single commit should work like this svn commit m commit message +Ive tried a few testcases and all of them seemed to work +Logged In YES useridGood catch +This is the last remaining issue for the release +From the stats +I hear from windows admins that this is possible though I have no idea of how database was already running on a different machine +It seems entirely reasonable that I should want to create a CLOBBLOB field from the Roo shell +bat commands +provided patches fixed some issues but did not make the autotools files set stdc +I vote we resolve this issue as a dupe of what happens here is that the region is processing stuff even after a GC when it shouldnt and what we need is more like IO fencing +Lets put it differently The bean already has the as its input and is now throwing an exceptionAm I misusing the DSL +We must have a common place for this job because the same code its usedduplicate in some place resources plugin assembly about in shared +It seems to be fine Thanks for the explanation +localCatalog in the we will make a release this week for testing. +The script I posted are using ij +Please confirm they have been applied as expected +From reading the logj code there are two places that logj looks for the the loader that loaded logj itself and the context class loaderWe might have to look into the context class loader as the source for the +Sorry Doug +A patch that also includes a prototype of the new Async APISee I forgot to track different change list so the patch is all my local changes +I wouldnt delay the release because of this issue if that is the question +patches issue is marked Resolved but I dont see the fixes in the trunk +It might be worth adding to the manual a note that after adding or replacing drives on a when its temporarily offline anyway that blocks and their associated metadata file both can be moved to any defined data directory for local rebalancing +No response assuming ok. +branchesJBESBCP is a standalone reproduce a quickstartant deployant runtestThe message should be routed to the red way +Please commit this when you can Feng and change the Fix Version to +Just run the script which will call an url until the error is thrown +Fixed in r and r which were included in release +I assume that it is like that to keep consistency +I tried with a very simple project it works for me +Rather than asking for domain when creating a password the tool should ask for mask name when creating a password maskThis will allow us to use a consistent masking metaphor in both the documentation and the user experience right up to the point where the bean annotation happens and there we just say put the password mask name as the value forAt the moment were inconsistent about the amount of implementation were exposing the user to +The handler delegates to a map of keyed by the correlationId property of received Messages +Attached containing implementation source +Only we have to ensure is that Client should stream all packets to DN till the fsync length +Because of that generated a totally wrong sql queries for those cases +setting minDf and minSupport to for all wikipedia chunks and countries generated using the commands above causes a drop from features to features and labelfeature pairs to labelfeature pairs +We need a complete story for how we supergroup mappings mappings and acls across RM +If this BC is created due to its use of the GPL it will have to live at the Codehaus +The reason sometimes fails is if you see that stack trace then you can disregard +I just cant see it as being generally useful or interesting +HTML file is included for review +Problem is still exist in version +The unit tests run fine as expected +if you have groups in process and have set the +It would make it easier to read it if you could split the require +If there are other mutable schemas that arent handled please build upon the thread assumes behavior I believe in the worst case hash code could still be calculated multiple times due to multithreading visibility issues but since its stored as a primitive there should be no chance of having an invalid state + +Lets move this to so that we can perform more testing +However a GB file should result in minutes on most installations so should be OK +Would you like to assign this issue to me +Perhaps someone else can shed some light on this onet be addressed in more discussion guys just looking at our critical issues and hadnt noticed this one previously did anyone have a look at this issue and can we reproduce? +From where exactly do you want me to remove the dot? This happens at the servers startup and I guess the server determines this path internally +was released which fixed this bug +I fixed the issue with on AS +The right thing to do is to hunt down the root cause of new interrupt state to preserve previous behavior +Err autologged in as my old account iksrazal +fixed in r +I didnt notice earlier. +Confirmed by Craig. +patch lgtm +However note that curl plays nicely when you provide compressed +Hooray it is resolved! May I ask what the resolution was since there appeared to be a lack of certainty in some comments above about which direction this was going to go +There is but Im not sure I understand the problem completely +I believe I can wait about weeks for this you for the prompt responses +CMS sorts merges by segments size and not +Thanks a lot for the review Steve +Today HBCK is using a lot of filesdirs names checks to find lost regions and things like that +Reverse lookup is done in several a new attribute on connection factories providing the ability to disable the lookups +Has the build environment been updated between the old and current build or you implemented some fixAnyways this can now really get closed +I also noticed somebody mentioned as integer rather than a boolean is for such replication purposeThat was me a long time ago +Sveta thanks for the fixes +You mean in addition to this patch +No worries +refactor can you rebase it +Is that still correct or should it say there too +Ive just committed this +Proposed SOA TP Release NoteAdded an expected output portion to jmstransacted Quick Start +Any further input? If not I will close the issue +Its the same problem under Linux confirm this issue under Windows Sun JDK and grails +Takes up screenestate +This issue is a particluar case of node limitation in inlining +Verified at r +Thanks Sharad Sreekanth and everybody else +Built in brew updates have been verified in the ER build +Works like a dream. +Unit Test ResultsTests run Failures Errors Skipped is looking good +SamarthErik How does this patch lookOverall its looking good +A patch for the existing and my new unit test +Unit test passed and got improvement of cpu usage per nodeWhat workload? Terasort or something else +I notice that your patch has a mixture of DOS and UNIX +From what you describe the management seems to have changed +i dont know what value will be adequate +The new versions are somehow calling into too early +looks like something unintended to always throws integration test showing the issueThanks mirko for spotting for reporting Ive fixed this in trunk now the exception is only thrown if an error occurs or the noderesource does not right forgot to commit that Ive added a slightly modified version to our integration tests. +committed to trunk +Rob can you commit this please +You probably could create an abstract class that all of them inherit to share the common codeOf course its still a bit weird bc in your case the type value is going to be set to ACRONYM when your example is clearly not one +Ok +It should definitely be set to true by default though +We can make a separate issue to do this after your shell work goes into hbaseWould appreciate feedback on problems found and questions raised above when you get a chance +This means the IO scheduler has more time to reorder them etc since no one is waiting on the result +Please backport to SMX branch if possible +Ran with patch applied in +Thanks Joe +This is a code maintenance improvement so what matters is that we fix it in the codebase +when I stopped mine it was several hundred threads all waiting on Kernelsystem output +allows STARTTIME for tasks which are started before JT restart if the history file has the key RESTARTCOUNT +There are also some failing tests with mysql that I have to look this one is ready to commit. +ascending or descending order at order item +I dont want to block this change I just want to make sure we dont follow this no public methods approach for Jackrabbit +Note to self Switch to git +This will also help when Oozie supports hive and other actions through tests +Work complete awaiting PR for +Hi PaulexThe fix looks fine +Hi YokoThanks for putting the fix in I assume this is the fix for the +this has been fixed for some time +This patch allows the bug to be reproduced every time +keystore files on so they should have made it in +Would you mind checking if this is still present in the latest? +Not critical though +In these situations we would issue an error message Cogroup by is only allowed if the input has a schema and error out +Sorry +Yes but this is DEBUG +For this reason anything I do would not work for the standard Collection classes but it could be done with modification of other Collection classes in CommonsCollections +for pulling this in. +Running site will put the aggregated report into targetsitecobertura running sitestage will put it into targetstaginglocalhosthomeusermysitecobertura so that cross links from within the staged site work correctly +The value is that I will often if not constantly want to know form looking at a step instance when did it last run +My benchmark codeOr alternatively checkout git client patch to support buffersizePS how do I write a test for this +I am no particular fan of this feature but thatd have to be a ML discussion of sorts +To handle the issue you described we can send a leader and isr request just to the leader with the reduced isr +Hi PatrickCould you please submit this patch as an attachment and additionally grant ASF the right to use itIf you have a test case please submit that as wellThanks +If you have no objections I will commit that one tomorrow +We do this mapping manually at present by attempting to match MRIs names for each platform +The attached files contain a patch to fix the is contains an xsdany in the objectList sequence to let custom parsers add their own the method is changed to use the one the registered parsers +i dont think its critical but maybe? u tried Axis +Currently the job puts it heredownloadshtdocsmodeshapebuildsstagingallrepoYou mentioned you didnt like that path +Thanks +I think your fix is OK Maarten so I mark this issue resolved I suggest users to open new issues if other packaging are not recognized. +Thanks Lars +Since we will be using a pojo everywhere internally +The DB shows the version field is set to by the unit test as you would expect +And add a note about this change in the release notes +to fix +I cant update the bug entry to show that it has both a test case and patch for evaluation +has been released and wont generate and should no longer generate user threads which prevent the test from ending. +The site like any other build output affectedcreated by plugins should be reproducible which requires to lock down the report plugin versions +fix for this issue was made in SVN revision +was removed from upstream AS so this is no longer an issue. +A quick glance at JIRA shows most of our tickets are around the tags which frankly see little work outside Musachys heroic efforts with the Dojo tagsIm not saying Im totally against putting this in but Id like to see a truly compelling reason to justify the cost +Do we have some use cases? If so we can prefer st patch itselfThanksUma +cleaned more the hibernate is the counter patch +Can you check please and report back if that is a problem +IANAL but based on the statements in my previous comment recorded above in this Jira issue I believe we can distribute these under the Apache license. +Verified with the update listed in +verified +The test needs to be changed +Nice work Sachith +Also it seems Principal has implemented all interface methods so it does not need to be an abstract classIt is good to keep Principal method since it is an abstract concept +should supports servlet if there is no servlet API it will not create the +This relies on you having an Awestruct based project site rather than Magnolia +Im found the method which Im using here in response to a forum persist catch instancesThis makes it possible to see hibernate validation errors that are caught on persist and update as messages +I changed the name from to +Two small comments I think the summary is supposed to be a +done at both the ui and technical level now. +btw +Check the document for the allowed properties +We can then deprecate the original. +Tested very lightly with a short fetchlist please review test +Supported is limited to Enhancer created proxies using CGLIBs no super class or a superclass with an accessible default constructor and a single Callback. +If servers hosting regions should see X +Fixed on branch r +It would be better if we could add Success messages as well + fixes build against hadoop +Love it! From working with the schemas you use for the C code it would be nice to have some information about just what corner case the schema is meant to test +Youd then be able maxEntries lifespan +This patch doesnt apply cleanly anymore +Current search implementation covers all queries except Descendant or self in SQL dialect but RI do not support such queries too and there is no TCK covering such usecase +Per Dans comment this is not a bug it has to do with type incompatibility not data incompatibility. +Could you please reattach it with grant licence checked +it seems that the pom file was unparsable after your change thus the loading of the pom failed +Applied the patch +Mahadev Sanjay reported this first here and tries to get this accomplised via directions from the to the +But now I think that it was an unnecessary class because of +Some of the method names of the old and new may accidentally match but their semantics will be differentI had missed this +Paulex you are right +Carsten? +i know what it is and in my opinion thats only a crutch for people who comes from cdms and trying to match their content against +Since they are independent changes they should be tracked in their respective projects +If you get a chance could you run your test again and confirm there are no more problems? If so this can be marked as fixed for that the problem is resolved using the latest build. +and changed buffer size about MB in the recordsat this time max bandwidth is G +I created for this +I believe the solution to this and other initialization problems is to wrap theentire init method implementation in one trycatch catching Exception andrethrowing it as a telling the container to notallow any requests and make it clear it didnt start up correctlyIll wait a few days but if no one objects Ill make the change +Applied the patch +I agree with Kris and see it as an issue outside the scope of Struts +HTH +Appears fixed +Programmers guide says As a convenience you should simply extend +This patch seems to fix the problem pretty nicely +Cool! I did not realize that already does this I will take a look +No need for +I spent around hours with debugging to find whats the problem because I thought there is a problem on my JSF login page which have sent null as password value +We need a runnable test case +Verified on EAP is failing agiain with can you please have a look if this is the same issue or new one +Why does the RM not notice abnormal failure rates on such an NM and put it out of rotation for schedulingCurrently the RM doesnt track container failures on nodes for purposes of blacklisting them +Fix IgorPatch applied to AWT module at repo revision rI realize that the effect can be seen using the demo applet but do you have any suggestions for an automated regression test for this problem? I wasnt clear to me how we can detect this case easily in a test case +Needs to be fixed +What do you think about the modification in the attached patch +With the patch the keyspace disappearing bug didnt happen +It is true that now all printers are initialized on that first rendering but for me that is acceptable +According to the spec there is no parameter for +committed thanks! +Attaching two patches one for trunk and one for which should get the two files in sync +That should be sufficientShelly what do you think +This patch is even better +Therefore I see that there is a requirement for a README in here too +Even if you dont want to modify the interface just exposing that field in the subclass would be helpful +I have attached the patch for this classBest regards +Deferred until. +The application then either needs to know what all services the containers used on that particular nm or we broadcast the stop message to the services and they stop as necessary +on to trunkThanks for the patch EnisThanks for the review Stack +Im glad its ok. +Will look further tomorrow +This should wrap up the change to make the classfile sequencer output compatible with the java file sequencer tried to build the patch from git so apologies in advance if it doesnt work for anyone else +the problem I reported previously is likely only a local issue Ill set this jira back to resolved until Ive confirmed just checked on RHEL zip didnt get that error and jsvc started properly +The splitshard API returns at this point +Fix version +Theme CRUD UI artifacts +ok I will review againthanks +Ill have a look +We need to discuss as part of the updated to having the services broken out into their own to make this task easier + +It would be even more useful if it was ported to +The resources needed in the Map phase typically depends on the amount of data it processes +Thanks Siying +Its tricky +Thanks AnahitaYour patch is in trunk at revision This was indeed totally uselessly bloated in r r r +I think this should be supportedThe fact that it doesnt just work is an artifact of our codegen strategy for main methods in the C backendThere is no technical reason I am aware of why it cant be made to work with some hours of fooling around +Tom would appreciate it a lot if you can review it one more time +If this needs to be reconsidered please raise the issue with the SOA PM team. +Hi Freeman your comments make sense thanks for the feedbackI will backport the test to the karaf also +This does not look like a bug report to me rather it is a request for help +Hithanks for your support your mirror will beadded to the list later todayregardsHenk Penning +We should open another JIRA if there is any +Its due to use zone wide primary storage the hyervisor type is not detected +It seems that this should be fixed in as well +It deleted more deprecations than it tests but I promise to catch up +moving off to to next version +To support all the dynamic features that it does Groovy does type conversions so commonly that I dont know if it will be a very useful thing to get all those right now I wanted to explain the behaviour that you were observing +For example our has properties used only for tests +Did this work with a previous version of +Eric We should probably add a test case for this +After I installed Cloudstack there was no indication that the installation failed +no you cant a class in its constructor youd have to use a factory you could disassociate the mbean from the PFEPS is being instantiated in near line which is what is throwing. +Are you sure is deprecated by +I think it is better to run the tests with the default configs as far as possible +Have a look at as well as the suggested patch in which seems to fix the problem for usRegards Toadie +Can you please grab the nightly and give it a try +Attaching a simple job that fails its maps andor reduces +Attachment has been added with description Patch for the bug with correct code formatting +The current demo setup only includes it from the download site +The bug is reproduced when using a truetypefont as well +all skins were updated. +Ramesh how would this tie into what you have added in the? And in general for or are you thinking that the timer based full refresh should apply to internal mat views as well +Version Ok I changed the whole thing +Fixed at in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +Delete existing snapshots of the given table +What would be different about this one +Resolving +I will port to as soon as tests complete +In this case iteration over complete set is acceptable +While its possible to automatically escape characters its probably unwise because it balloons the response size +Ok lets switch and discuss on dev should consider switching to JDK m not sure we should require Java for users currently +thanks +Basically Commons CLI seems stalled but not dead there is still some willingness on the part of Henri Brian and others to do something +Hi Dmitriy Can we have this patch checked in for older versions also like and +ThanksChatra +That seems reasonableWhy did Brians only need SS and not CS +Ive just looked for uses of pk myself and indeed it seems it should be removed +the tail files console is not supposed to reopen again and again its supposed to be a persistent connection that shows the logs matching the initial it not work to close the console and then start it again +I will try these out this evening after work and report back with my findings +This issue is now closed +Patch applied +But verified that file with zero size can be uploaded without problems in IE with RF. +Meant to say Regarding the check for cachingDisabled its outside incrementRefdecrementRef since I wanted to avoid method invocations when caching is enabled +I think inserting semicolon is the better idea rather than expecting users to do the same +This issue should be the same with it was caused by the semicolon in the parameters. +To the new assignee this is not a Transactions issue it is a small issue in the integration code where you are meant to configure a that can connect to known servers to see if they have any transactions that were started at this is a distributed JTA issue not JTAJTS +ve committed this +This may cause a conflict with s patch when being applied because they both edit in the same spot +I just dont had time yet to do it a contribution would be very welcomeI also worked on the there I started to integrate the name finder in a way it can use the user input to enhance the recall of the recognition and then the user can confirm annotations + checked in +followed by camel case unless theres some other reason not to +Hi FreemanYou need to get the test case running and see the behavior because the connections do not get closed the way you describe +Also it would be helpful if the new API would return all permissions for a given context +But if we can do it what do you think about this idea +VinodAccepting a mapping of arbitrary commands is indeed the most powerful approach +Attaching complete patch with tests and Spring NS handler support +Looks ok to me and I got a clean build +But removing the sort completely from the framework would make many applications impossible to write +Merged into master. +Id still like to add a unit test that triggers this problem before committing the patch +Reopening +Line breaks should be preserved if is set to preserve +can you create a simple quickstart that demonstrates this usecase +Id like to extend the test cases to cover more cases including any combination of addremove node by one session and move node tofrom by another sessionIll post an updated patch when it is ready +I just committed this +I mostly get itAlso looked through the comments on +Ill go ahead and close this for now since youve already created a replacement issue but what I said was to handle differently these cases if it is a replacement of a previous patch put it in the same issue if it is in addition to a previous patch and separate from it then create a new issue +Addressed Amars offline comments +now +Sigh + +The latest patch from reviewboard +bulk defer to to have +Ive see that test fail a number of times +Not sure why I didnt try that +What if the client is talking to a leaderfollower connection drops and it connects to an? Im actually wondering if this case is even possible +Functionality is already present using loggers and handlers. +I think this can be removed +I dont think more seconds in QA run are noticeableCan you put the in the right place and perform testing on real cluster? executor to now in your persistance on this patch +bulk move of all unresolved issues from to +Bump I will fix the formatting later +I think it is better if prints something when the wait is over +If it is a issue I will create a separate Jira for thatThanks did not attach the log file yesterday +Restarting the cluster fixed the problem +Tedattached an addendum +Fixed in jive + latest snapshot certainly works for meThanks. +cool +Seems to work for you so I close it +Please verify it was applied as you expected +This interface is currently only used in the zip mode however it would be possible to use it in html mode as well like showing an image in the console etc +Ah +So should this jira be opened against G +Those on may wish to do more +I have replaced and it works immediately. +Name it openssl +Agreed +I will look into it +Sorry for wasting anyones timeEarnie! +A test case that signals error if the logger gets an ERROR event from the tests invocation +Closing resolved issues for already released versions +when I say no DOM Parser +Move the issues to as not critical +Patch adds include parameter to the plugin chance this patch will be included soonI also need these empty directories and the current behaviour is different from eclipse behaviour +Attached patch resolves the problem with the client side error messagesNot all SSHMSGCHANNELOPEN request from the client have wantReply true +Added myself as CC +attached you will find an updated version of the patch because there is another problem oncethe first pathc is applied +So no need for scripts +This issue occurs due to billing contact mech for order does not created +Bulk close for +Well Im not speaking about outputting xml prolog to users +I havent look to deeply at JDK concurrency support but something like the might be useful hereUnless someone beats me to it Ill have a look at this over the weekend + +You are right Nicholas +Software ext mmorin +Than obvious question is Why this issue hasnt been assigned to someone +Closing the issue now that it has been resolved. +to the branch in revision . +If there are no objections I will commit this beautiful addition to our analysis tests soon +I mean users dont care about your automaton at all much less transitions +thank you +This is a problem if the field is used for sorting of documents and if there is a secondary sort criteria. +Thanks Sujee +I think this is broken again +I thought I left them in the DTD for backward compatibility but looking through the logs I did indeed remove them +I improved the error message to thisCannot cast object null with class null to class float +Resolving. +It now compiles and runs under rhel bit +You will not see this behavior w the same jars of the release +Struts ships with Retrotranslator version +Very crude limit connections +Libor can you add the annotated java file you used to test this as well as any information from the file or console +I just committed this +After I imported the projects the srcmainjava folder was not added as a src directory in eclipse +Fixing the links +In case we dont want the official to depend on httpclient another solution is to check as is move this mojo to a plugin to the mojo I think this issue needs to be reassigned to the webstart plugin in the MOJO project +upgraded to xalan for HEAD too. +thanks for cooperation guys +Defer to defer open issues to defer issues to +This issue can be reproduced for another components +I doubt this is isolated to Windows but have not tested that theory +Maybe a biased round robin +Thanks for the patch IvanWhat do we need to do before we can check in this patch? to answer your question all we need is a careful review +I i restart mcf I see some delete command sent to solr +Integrated to trunkThanks for the patch RajeshThanks for the review Ram +Same patch on RB +This is a duplicate of. +Fixed improved the test improved the error checking involving directory permisions and added docs +having a teiid connection profile just enables the eclipse based tools to connect it doesnt enable it for the command line tools nor does it make any Teiid specific dialects available to neither command line or IDE +bq +Width OK BUT NOT HEIGHTDoes it versions +This is the test project I used +So maybe thats the easiest way to go +That means it would have to reuse the +Hi all we have just tried to apply to +Test applications to show this is an EJB issue. +Also as Nexus uses Apache for security and Im on the team I am more than happy to write any components necessary for integration within the ASF if the ones arent sufficient +One more thing what the name of the merged porlet should be? Freezone or docs? Im voting for freezone with an indication in the docs that it can be used for docs +common +How do I collect the equivalent traces Kathy +This is a problem that Emmanuel came across +The difference is that we dont want to be bothered with many authentication providers at each service +I think its reasonable to permit other filesystem implementations besides local whose uris do not include an authority +Marking bugs with invalid fixed for release as fixed in. +I dont think have a similar property for the ID of the BUNDLE that launched them though +Hi TuomasCan you please give us An explanation of what youre trying to achieve Your file The command youre running The output of the command when run with the mvn e option What youre expecting this command to doIve fixed the NPE but I dont understand why the is used when youre attempting to use a remote configurationThank TuomasOK I get it now This hack should then solve your issue sorry about the hassle causedThank you +Heres a new patch +I will try looking into it and think about your suggestion using +apbackup svnwc Jan +I also changed the fields declared in to private final +Would Knuts proposed change work with existing user applications? It seems like a better solution if it works with existing applications +thanks +A number of intermediate buffers are replaced with one thread local buffer +Patch that fixes the issue +The last comment on indicates the Jenkins slaves for Hadoop are controlled internally by Yahoo? So it would seem cmake has not been installed yet +However there is no option o to uname on Mac OS XDebianUsing straight uname with no options works for those systems +Looks like will have a abstraction layer in Tapestry soon +I mean you can assert the real value of partition is this case +Thanks for making the changes and pulling it in. +Fix memset issue in first versionAnd do some optimization on linux platform +Bulk move issues to and +Initial updates to in kit Release notes svn ciSending Transmitting file dataCommitted revision +Complement patch for rethrowing is availableThanks +Looks good verified in JBDS h +Hi BJI wonder if the new Setup component is not the solution now +The above sample deployment plan can be opened after applying the patch in +Marking resolvedfixed +are some other problems with this code +bq +Thanks Raul Gutierrez Segales! +We have a task item which involves a redesign of the percentage handling +Project +I found the patch already applied +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +what about trunk? seems to be a fix for the issue you describe +Closing issue +It works as expected +Added which proves that the WB CL isolation deployers kick in this should also be tested in testsuite +Attaching a sample Java program that reproduces the problem +Chaining actions is generally regarded as a bad practice so were highly unlikely to provide support for it as part of Struts +This schema is referenced in the attached WSDL +I dont know how practical this is or even if we would be able to update the symlinks by that point in the upgrade sequence but what about having versioned in separate directories and symlinks to them in the classpath? I do think that dropping the versions from names would be annoying when trying to debug and ascertain which versions of each artifact is being used so if we do need to drop them to enable smoother upgrades then I definitely think we should provide another easy way to find that out +Please confirm they have been applied as expected +Please see attached patch +Heres a patch for this problem +But I didnt test APIIf user enters some queries at calls method submitQuery ofIt is a async method and calls method get +My guess if the CSS is still the same as it was is that the area for each product summary has a fixed height in the CSS file +Yes there are nexts problem while trying to implement non attached doesnt works so I cant just use component control together with f to use ajsupport and oncomplete I cant show tooltip at all because of problem in ffcheck function +As Scott indicated he no longer needs this feature and I see no pressing need either Ill resolve it as Wont FixScott if you agree then please close this issue. +Cardinality Exception Error fixed by the inclusion of an extra MR +I take it back the new test in the patch does NOT fail without the code changes probably because its an absolute rather than relative path +Ive now finalized the upgrade and were seeing better performance of blocks got processed in msecsAlthough having the datanodes lose contact with the namenode because its checking disk usage seems like quite a serious bug to menot sure why sda is more busy although that is where the logs are looks like DU goes through previous also +The application would need to set a custom property in the RESPONSE header if it needs it in a reply +ok the problem is the additional encoding of the url for the script src attribute likesrchttpparamvalueampparamvalue +Hello PaulexThe fix looks good +Primary data node was forcely powered down so logRoll was requested +My synonyms file +The new modified patch for +Also itll be hard to figure out which tuple the schema is supposed to apply to well need to prefix the tuple with something anyway +They require all classes to be available even if they are not used +I dont have a list of the exact changes +updated component type +Patch I think its ready to commitSeparately we should think about how suggest module should set these +Mike at what version did you do this patch against? I tried it against latest in as well as latest official one but both failedPS Big thank you for fixing it +I realize this isnt being done right now in shale but its possible as an interfaceJoshua Bloch has a good write up on this in Effective Java +. +I put as resolvedincomplete to facilitate searches with My resolved pending own in r r r +You should create and attach patch files for your test case and proposed fix for review +Thanks +Future callout development can reference this change or start from scratch +Reopen if not please +this is fixed in describes adding such a configuration setting +Will migrate it to very soon +All next changes are not is the right the ui could at least put a red cross on it and provide a way to see the error message which is already in error log +Nice catch! Fixs been committed to both the trunk and the tree thanks +Can someone verify? I am still encountering the same issue +Ill have a closer look at this asap +So I guess I am not going to get the latest class file +its +Closing +The changes arent too much work to implement +See. +Only two? Thats cant be right I posted about a half a dozen +Just tried the snapshot and it works perfectly now +Advice to users is to turn off +Not during compilation phase but during fails due to builds modules servletint and loggerwhich is dependent on need to remove the dep on both in the POM and in the code +I think isnt quite right in fact neither nor is quite right I checked both +Hey Dhruba a couple of quick questions Why is excludedNodes a? A should perform at least as well and use less memory +Third error in the build files is a non existent target in the userguide buildI thought I had similar issues but atm I do not seem to have them using +Thanks for the great work Dian +In my case I am using which wraps an injected Castor marshaller but does not expose these two methods +Adrian you mentioned the Common component wouldnt be better to move the artifacts to the Webtools component instead +branchSending tutorialssrcorgapachepivottutorialsstocktrackerTransmitting file dataCommitted revision . +Goal is to obtain an authorization cookie using a single Camel endpoint +Is this still an issue? Can you provide a sample project andor some more explanation of the issue +This is expected behavior and the generated names look like etc +The parameter is indeed a number of seconds +I think this one has been fixed +The workaround is to do the build when it fails set OBJECTMODE to then manually finish the build of the native code and the packaging of the resulting files into lib +Fixed in revision +Any chance of a patch Bob? Thanks +Ill retry with your latest patch +This is my first implementation for the Bracha Toueg algorithm +Window styles can belong in a window style category +Can you please upload a new patch creating a JIRA in projectThanksChinthaka +andrew Yes the TTAK should default to ryan If we explain that when a user set timestamps then the TTKAV will apply to it +I think having some consistent serialization would be nice +Updated patch as per comments +I have created a patch to try and show the problem +Much appreciated. +Currently Hive does not execute anything as part of EXPLAIN +I will get to this in a few weeks +This is from RCI built it using java though +Its all generated source anyway +few comments +This one requires more information +I dont think this is a Palo Alto specific case because the id field is already defined in the rule before the Palo Alto has any control over the rule +Committed +I am running tests right Namit can you resolve that diff manually? I just tried that testcase but I did not get the diffAlso can we increase the interval time in another jira? Right now I often got wrong patch file so i do not want to risk getting a wrong diff again +I fixed by removing that address option that dont make sense. +Looks like needs to have latest version of servlet for the example to work +ill take care next timethanks for your mentorship +AFAIU the format is proprietary but open source libraries exist that allow unarchiving the license explicitly prohibits of the compression algorithmThe ASF lists the license of the UNRAR lib as one that is acceptable for a binary depencency. +all cases have the same error +This should fix this issue. +This is because the CP version does not have which is where was updated in Branch. +JasonThe is an action that exposes only Webservice endpoints that are internally hosted in the same container via +done + Plan looks reasonable to me +It is unclear what changed caused this problem +This solution should do the trick! it can handle all kind of special type array think you are rightWe should change it in myfaces shared code. +Servlet spec dictates that get should return whats after and before query string parameters +Does this require JDK +I am also attaching the source file +Bulk defer all frontend issues to +We rely on the Ant libraries for our behaviour +In this case Im just using it as a visual confirmation that the users are running the correct application and its connecting to the correct database etc +The MVC exception handler should be setting that header I think +Hi Peterwould you say this is a recent regression or have you not tried this setup previouslyI made some changes to ast transforms for grails just before M went out you see +We should not avoid the decision but make it +Enabling profile release builds the userguide for inclusion in the distromvn Prelease package +srcjavaorgapachecommonslang +Ive committed a fix to the tree that should resolve this issue +Includes unit test +The suite passed on Hudson +This patch will force a when the Cache is configured without configuring the +Thanks for the review and fix Arpit! Update the whole patch with addendum +Well I completely agree now but since currently storing the image file on disc and storing in memory state are kind of two different abstractions that are communicating via String paths this seems like a clean way to get some performance improvement +the Cancel link only is visible if the box width is enlarged +I had a look at the changes this morning the look goodregards Malcolm Edgar +The WAR files are extracted in special folders of the server installation when the application is deployed on the server +Have you tried to reproduce it? Put some bean configuration in a sar that also contains a war +You should be able to find some of those changes in the SVN history +Thanks for raising the issue! Ill take care of this +Good catch thanks! Added oi in commit bd. +Removed the ant target from the ant target until can get this resolved +In the outgoing direction when the xmlid attribute is passed back through libxml the resulting message includes an explicit reference to the xmlnsxml namespace +Heres a patch for the root +But its a real I dont think that in any Java API I can think of or HDFS you would have a successful write and successful close but fail to write +My is now experiencing the same problem while not using any plugins except one that I wrote that is nothing more than a grails project with a custom script +Hi DavidIm slating this for further review in when well be taking care of a number of similar issues +But once you close Eclipse and reopen it again it stays at Republish state for ever and ever +Andrei we should wait until is done before committing this +This is working but +I would just caution one thing +I think that the issue is almost resolved +In this way the processing will be much faster +If getcompletedsnapshots is not too long I will attach a patch with that name +Hence clients to a job submission API +Agree +Here is the diff file for the change +INFO Running for rows took ms +Is there a way to set the build number at deploy time to work around this issue +Still happening with +Slightly improved patch +The patch looks fine but I dont yet understand the need for it +Thanks Melvinshould we move to an instance method of Message? This would make timeouts easier in the future as wellLets make the default values in Config s as well to avoid surprising people who upgrade but keep their old config file aroundOtherwise LGTM +Screen shot of the current JBDS editor very wonky +Provided patch +Could it be that annotation support in groovy would only be a compile time annotations that could be used to create a rather than a runtime annotation? This would make it possible to compile the groovy plugin with Java be able to run in an the annotations would only really be needed when compiling the groovy mojo plugin +Can you attach an example marcel +Jochen I tried this change locally and didnt face any concern if I go ahead with this change +Youre using an old version of ASM in your classpath +added comments to added test for zooset also the async apis are tested via athe sync apis +Updating Fix version to reflect actual one after Todds backport. +look good to meMike +Pls review again +Can you show the race condition that illustrates this +None of us besides Ole have been in contact with the folks +this is correct semantics +Tested and fixed with both and JDK. +I ended up doing some fixes and refactoring and adding couple of unit tests as well as a new class which will search the collocated terms if provided with a term +Looking at API docs it looks like this should work but the IBM JDK doesnt like it +strange i could have sworn i resolved this issue +When is scheduled for release? +Raise a http instead + +The right fix is to change the initialization order if it is doable +All examples use the new streamlined XML syntax. +The location of the cache dir in the patch above is in ACCUMULOHOME if its not specified +prasanna How to do that? +Can you please provide your test classes and configuration +And I have tested the byte array from on Harmony and it is the same as that of RI +Just take a look at some other classes on core for the style infos +removed it in . +bq +Yes you are right Scott +If youre OK with this Id like to remove the transaction module from the NMR build again as initially intended and then mark this issue resolved for the upcoming NMR release. +I just ran the build tests and integration tests and the current code base works fine with java. +enjoy +defer all issues to. +its a custom extension of webdav that isnt definedby either of theIf you want to move the specific stuff to the webdav project they should at least berenamed from to or and put into a different Constants class in order to make sure that those constants have nothing to do with standard webdavi really dont want JCR to appear in the dav library + +Hi Rajinithank you for your good supportThe warning column is mostly for us +Yip I am still reviewing the patch +Thanks! +Make one of the data directory unwritable +The user name and group name of default derby user can NOT be changed but the default password can with following steps +After removing the file the Select file button started to work +If any of these need fixing urgently feel free to keep bugging me until theyre fixed +oops relevant output does not show up very neatly +Added the in question the the mvn X output +I just fired the tests using time java true +The Stomp specification states that Any client frame other than CONNECT may specify a receipt header with an arbitrary value +Reopened to correct Release Notes field. +is presnt when page is rendered from browser but still fails +Compilation works +four related tests passed by Hadoop QAWill integrate Brians patch soon +Test passing with this fix +Hi Marek I believe this was meant to be assigned to you +Marek When these operations via the main or context menu will occurWhen you choose variable then the context menu should have additionremoval operation? Or somewhere +Hi +resync with trunk +The url mode works just like the same mode in Mojarra does and it is better to keep this mode in that way to ensure compatibility in testingInstead we could create a new mode that do what you proposeNote the client mode implemented in do the redirect +Also updated the descriptors in the examples +Please take a look +Could some one please check this points and patch +Does it work for you using without the firewall exception or with it +Here come the logs pertaining to the this works and this also works code snippets in the attachmentsHarald +batch transition to closed remaining resolved bugs +Can we close this out +was a issue to +In mapreduces case the comparators are created through so in some way you can pass more information through the Configuration object + contains a patch for modules +Cleaned +I will add both and version bundles for elasticsearchThe Maven URL iselasticsearchand its present on OSS Sonatype repository +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Improved realization in trunk +In any case though it would be nice not to have to create a custom just to get this one sanity check +updated patch i tried to clean up these tests some +It seems to be waiting more than a little while +If you think they are critical for CR please them +Just my comment I mean the threads attribute not maxthreads +This issue is resolved when everthing works out of the you elaborate +This sound like an issue for the Site Plugin rather than MPIR +logj Xerces need to be included on the CLASSPATH before launching Ant? I have commented them in a local copy of the and everything still works +Changed recursion to loop +Will run additional testing on v patch tonight +This can be closed +This issue was fixed with a respin of the beta bits. +we will need to use a flag to mark if conversion has happened or not since null can be a valid result +Do you have example flusherThis is a good questionFrom the presentation at User Group I think there is already another implementationShowing that implementation would help us understand the use case +Hellocan you attach the testcase in some format +This is the same patch as but this one has been changed for This one will fail the automated Hadoop QA because it is +I have attached a new patch that fixes the problem without getting rid of the Proxy +fixed a bug for path checking +i dont have the extra spaces now in my test casein regard to the umlautethere are other special unicode characters also in the textnot only to indicate umlautebut also to indicate list itemsexample jeKOUS schoner die Spatzen singen destoKON spater ist jeKOUS spater der Abend umAPPR soADV schoner die Gaste +Closing all resolved tickets from or older +If I were to use class to identify the classname of the triggering method that would be ok because there cannot be a local or parameter variable with name class its a keyword +The Mep client had a built in way to handle faults +validator done +Will keep it open a while longer because it bugs me +Create a new ticket with a quickstart please. +Patch looks good Committing +No exceptions in tests failed against manual verification +Committed at subversion revision +Mylar is no where to be found for download +Hi Rodionthanks for your interest +Verified on Tools +Still I think it is better to have this feature work when you are not using such a mirror than not to work at allWell it makes the build different based on local configuration maven version and existing metadata in local repository +According to some info on the users list the plugin can be absolutely specified like this mvnunpackIdeally the site FAQ should be changed to include The fact there is a namespace clash between and +New version of patch that updates hashcodes with a common method hash +I will try to build a small test case to reproduce the case. +That does not solve the problem of how to do it from the explorer +Hey Gary can you provide a series of repro steps for this one? Im not quite sure how to set up the test case to repro + +I cant think of any other platform specific text files +The patch fixes the scenario where the provider incorrectly handled a DONE exchangePlease backport this patch to SMX +Thanks Stepan +More difficult than s give it another try! It should be easier using the SSDK +I was stung by this today +Please try in a couple of a lot! +Update the Fix Versions field +Heres a new version of the patch that fixes the compiler test issues +Initial version +Yes I think we should set both values this way we know what is compliance property to source versionFixed in revision +However we should definitely do more on that front +So with that in place I am switching over to the Pig side of things as the actual Tez runtime cannot handle more than MRR +Part of it is the general structurePartially waiting on getting some well known url redirects set up so that we can use them from within the documentation +The amount of the affect depends on how many block updates done per measured benchmark and if the benchmark must wait on those updates +usesuses lusesor usesorg + of code performance is the same +It means Tools and hibernate tools wold be which matches your wish to have component versions under or equal overall version +Ok get it thanks +test failure was due to timeout +One concern I have with SASL or other forms of mechanism negotiation is that the authenticity of the mechanism can be confirmed in order to avoid downgrade attacks +and added testcases in Rev +Michael Thanks for pointing that out to me Ive asked them for clarification +example setup that shows the +Well my son wants to go play in the park now +no a jcr node name does never reflect a jcr property valueplease explain what you try to do and what should be added +I am going to commit this one soon +Also the while loop should be fixed the first sleep will lead to return +If T is a big table we dont want to create small versions of T for the selected table +Thank you for taking the time to do thisNew patch coming patch to merely add to +Ill take a look into assigning you should at least be able to assign the issue to yourselfUsually you dont close an issue in the workflow this is done by a project admin or the owner +In the latter case we need to alter the request protocol so that incoming requests can be buffered until they are complete and then dispatched to a worker thread +Thanks +Great +Hi I have just seen that the version I attached does not contain the latest bunch of bugfixes +Hopefully this will work as a workaround for others until we can get this issue fixed +Thats overly cautious +Sorry I forgot to include a newly added file should work fine now +Maven changes the artifact version to the managed version and deploys it to the need a clear test case for this issue or its going to be nearly impossible to be sure we have it located and fixed. +The change made to cat makes the output of cat different when the src does not exist or is a directory +Merged into the master branch. +This would be an awesome simplification and it really should be a private decision to the Codec +Unit test for this included +Might be related to to which is fixed in Maven confirm that Maven solves this issue. +So I wonder if there are differences in behavior here betweenrunning queries using IJ and running queries directly to JDBCAttached is which tries to use modeboth in and in a new test scriptWhen I use mode in it seems to work correctly I dont get the not openBut when I use mode in I get the not open message just as describe in the issuedescriptionThis patch is not ready for commit just shows some of thetesting work Ive been doing to better understand the behavior +examplary Java Code unlikely +I am totally naive in authentication systems but I am assuming only access details are stored in metastore and authentication is done by one of the systems discussed +Rejected +Ah got it +taken care of a while ago. +It is suppressed for in +We probably should add to the cmd line to support setting minDF maxDF +Fixed +Ill have to see what the generated javadoc HTML looks like though +have to be loaded before the compaction strategy +Anyway this is the result I get when I run java jar x +Is this assessment correct +Marking as invalid DO NOT THIS BUG REPORTCastor supports Writers its up to the user to configure their accordingly if they are using +Bulk closing stale resolved issues +How did you configure to usedefault windows installation on a xeon single core that is misunderstanding +Thanks Ezra I will test this and try to get it in Final. +The could easily solve this problem cause you know the field type and the provide also type information. +is there any response about this +Verified on SVN Revision . +Failing test should be fixed in trunk now through smarter guessing of the common element +To facilitate the authorization of these pages we will need the job that are introduced by and used subsequently in +Improved region selection using Priority is close to what I described inI used to alternately choose regions from head and tail of regionsBoth and pass +Same problem here +debugger info we can observe the locale is not correctly loaded +we will be focussing on this quit soon +The patch changes the javadoc comments so that they dont imply that the cache is populated with actual objects on creationCommitted revision . +Log file with verbose logging has been running stable for a long time now +Anyways its fine +Trust me if we knew how to catch exceptions that JSF swallows we wouldIdeas are welcome +The attached patch fixes recognition of Oracle g databases in the class +I can wait +Hmm +It might be the case you uninstall a bundle and a new one at the same location and want it to have the previous configuration +I had a similar problem but it turned out that I needed to clean the build first and then it worked. +Thanks for the quick response Woonsan +Ive finally merged and integrated the at revision . +Also Dominic if you cannot reproduce the issue anymore I suggest we close this and open a new issue if you stumble over it decided to close the bug for now +Please add it to docs +This problem should be solved nowCould you please test with a latest cvs checkout +Closed upon release of +Here is the patch for changes mentionedRegardsSurya Surya SumitDone at rAshish +Consequently I have removed the previous interfacesI have also moved some constants we use to determine the path of znodes in toIt adds a unit test and all tests pass for me +Thank you Freeman for the commentI really dont understand wha do you means by across applications but for sandbox i just switch the database of the to database of the user by changing the +Thank you for the test runs also +Remove all while job complete wait loops in the test cases +Feel free to assign any task to me +If you just want the empty map fix apply the patch on If you want the whole enchilada apply the patch attached up here +visually inspecting the html output to verify jQuery column def are correct as well as aaSorting + +classify +ill be more blunt since you are wasting my time here +hithan you for replythis error happens when spring context loading I dont know how to write a simple test case for itso I create a new project it is a maven project +to trunk +The user may not care how it works with rfiles and walogs they just want to turn on encryption of Accumulos persisted data +Patch quick comment from a cursory try of this there doesnt seem to be a way to go back up the panels once youve gone in to a detail panel could you add some sort of navigation facility even just a back button would probably do +Assigning to the JS expert +also added couple of more reusable methods +Is it still too early? Eclipse CDT pleaseAll the bestsMarci +If I fix will you upgrade? If not Im just going to close as will not fix because there is a workaround +Thanks Matt for the patchThis change is very much required +Creating a new data file +Doesnt pass all unit tests we might need to rewrite some +CoolJust a note to say that even with a privilegerole system in place we need to think about newlines +Redundant as we no longer use regex but ant style paths for URI matching +I have seen a static in the class holding such a references for example +We tried upgrading the memory ingaea and hyperion today but forsome reason the new memory causedthe machines to fail to POSTWe may try again tomorrow +Marking closed. +Thats the cause of oscillation at startupIm thinking of the possibility that during startup we just use assignevenly to all region servers +Found a way to handle and and split up the testing for easier testing +not blocking B for now becausea not time to actually fix itb not consistently reproducable for others than reporterc build from Forge just calls mvn and that can be done in so many other ways from eclipse or command line so not a showstopper +Just got back from vacations Ill try the fix you mentionned and let you know +I wonder could it be a timeout or something? I havent been able to reproduce the failure +Once is resolved anyone using columns will require this patch +Ive had this issue too but the classpath entry doesnt stay invalid on my side a USERHOME dir is created in the projects rootIts really annoying that STS cant use a dir outside the project absolute or relative paths wont work I was thinking of using a link so it at least pointed to the right directory in but figured it would be deleted every time I ran clean +Reopening the issue as the problem exisit in a fix for the solutionThe patch refractors the code with the similar occurrences and removed unused imports from the files +If we decide to keep any kind of reference to the comparator instance Id go with the hash code approachIt seems really lightweight and it provides enough uniqueness for the cache key to avoid potential collision +See PR . +Address issues with the mechanism to turn off prune columns the unit changes per my last review comments looks good patch for hudson review +Even though the feature is disabled by default I think that it deserves top billing because users are waiting for it +Would need some discussion here on the designimplementation decisions before finalizing the design for trunk +covers most cases. +this is really annoying +Thanks Yannis for all the digging in +And until its impact could be determined wasnt sure of the implications on the current release +I agree +But the fact that the callers keep the old address in their immutable object can cause problems +We will see +sure. +Nice one Markus +The POM in the artifact will at least initially be the untouched unresolved version +If this is the case release would be leaving our users behind +has several child classes defined for all primitive types such as for Boolean for Integer and so on +Ill let you know when its doneThanksHoward +Lookup Identifies the Lookups available to the Interpolator +It sounds like we still have problems with handles and GDI objects +Moving this to as Id like to get out as quickly as possible. +Thanks Stack for review +Restricting the instances that are to be terminated to the version of the Hadoop AMI specified in the config file +I applied this patch to my branch and ran both the queries above and they both run successfully now +Defer to next release +No that just added additional logging. +After modifying the showcase demo to use myfaces and use and Digits I am also unable to reproduce this error +This patch includes a backport of LRU ordering of deletion and throttling +However not that is out I plan on upgrading the patch and working with the guys to get it into trunk or at least a branch +Patch applied +I havent included the argument you make so far and think its valid but Id keep it in for that reason +This essentially means that is not generated until the project is launched +Attaching the updated conceptual file There are still some collation questions on that need to be answeredAlso one lingering question from me about JVM in the comments on this issue +Im sorry I dont understand the question +The applied patch does not increase the total number of javac compiler warnings +I will do that part also in the next version of patch +This is not true +Specifying it in doesnt work and hibernate claims that the bytecode provide is cglibShould I open a new bug ticket +This must be related to the queue switching at the end of the sync +Whether its a feature request or bug is slightly debatable cos no one has still answered my question Hmmm whats the point of Servers Host Name then? +If you do not have JAVAHOME set in needs to be detected correctly +Bug is not recreating with one node Per can you try your test case against to verify that this was indeed introduce in somehow +Its practically the same argument as to why not extract OBS into commons +As discussed with Mark moving this to FUTURE. +This sounds promising +Check whether you can remove initialContact flag and use only the restarted flag in the heartbeat method +fixed with revision . +Im a little bit busy at the moment but hope to find time in the next days +pending jenkins +I am running into the same NPE +What is your problem exactly? +Sorry about that +Since theres no indication that this will be done im moving Fix Version to Unknown +I apologize for the confusion +DoneRunning Tests run Failures Errors Skipped Time elapsed sec +However I cant see any changes in SVN trunkHave you committed the patches or was there some issue with them? Or perhaps they were committed on some branchI dont want to start working on while the state of the previous patches is in doubt +Hi Myrna I have submitted which tests jdbcnoonefruitflycreatetrue +Lets make it optional first and then see how realistic it is to make it automatic +This is the scenario that I am facing currently. +I wasnt aware that you could even set the cardinality of a mapping class +The bitmap will become sparse and we will need some sort of bitmap compression scheme +Because as you know Lin is working on the new server assembly portlets I think we could improve this inThanks Donald manual refresh is not user friendly as how do we convey to the user when they need to refreshA couple second delay while the list is rebuilt isnt a major problem +Thanks for testing it +class that allows searching over multiple remote indexes via a +I dontt know why you are talking about running sitesite after clean install did you mean sonarsonar? mvn clean install sonarsonar should work if you remove the checkstyle configuration +Thanks for working on this David + +Tony go ahead Im preparing a script to do it right +There are also a number of methods and classes that are deprecated and will be removed in Lucene +Now it is very simple Youre running installer as root user +Once tested I will merge + +That is good but it is bad for us to send new users to after the create a weblog +added guillaume and sent info offline on moderation +You need to update to TRUNK +This allows for the case where a JSP expression is evaluated before the tag gets control +i was not aware of this method +patch was applied on thanks this issue since I am finished with the work targeted for release +Committed in revision +Sorry forgot to tell just need to be commited +I have a small test case that shows the problem +Each TT is configured to allow only four concurrent tasks mappers andor reducers +Fixed in the revision Now support both and the platform ServerMove to the latest iPOJO version +yes I think in my isolted test the logging was not connected +Youll need to add your latest release to that path first then I can +Torben check out libphpsrcprotocol Thomas looks like the right placeI can see some bugs within these lines i will provide a patch and add a new ticket for itMy should not be included +I remember looking and they are not really used but they are still there +Its fixed in +Marking closed. +However keeping them in node id sorted order did cross my mind +Do you have some real use cases for this? I think is not realy a good use case as it is about a general feature where overriding seems to be overkill to dont have a use case besides which is an excellent use case for this +Moving this out of to as there is a work around for and in the future we can look if theres anything alse that can be problem is in the test case which calls the Node twice one for client and the other for service +With regard to a global service Davor Hrg already explained how to achieve thatHaving a class per page is one Tapestry architectural decision that was done after analyzing Tapestry which had it +Ive committed a fix in SVN rev +Step If the collation derivation for DTD for this node and otherDTS do not match and none of them have the derivation of NONE then it means that we are dealing with collation derivation of IMPLICIT and EXPLICIT and hence the dominant DTD should get collation derivation of NONE +I thought we already reached a decision on this but while continue reading emails I noticed the discussion started again +looks like an old issue that has been fixed time ago. +log WARNING JSF Unable to find or serve resource from library +This is my mistake thanks for fixing +Ive committed the change to the CVS +So it seems as though the webinterface is going through the api properly to find the status on the table +Added actual test for this at the low levels too. +Do you think you can fix that? Maybe you could just remove the changes in temporarily +rework and Mathieus patch and applied +confirmed +Adding XML file +It produces undesirable results and potential conflicts so Ill revert that change +bulk close of resolved issues. +Couple things on the test Hould be jnuit format Please use constants instead of direct config names +Thanks for your time and can work around the problem for now by rebuilding the jscience jar with RADIAN also present in the NONSI first this doesnt seem to cause any other problems +fixed all major issues with Hadoop. +Thanks very much for updating the wiki! If theres any remaining issue here please file a separate bug. +Patch v illustrates my proposalI also created a singleton for the null that signifies there is no server to assign region +Fixed +This jira has been closed as part of a review and archiving of the aj project +renamed property classes per issue description +Hi Kaz +Perhaps we should do the same with the pages +I will take a look tomorrow +This is not in +The map phase took roughly patch does not include waiting on the reducer fetch if the memory fs is full? or does it +Bad +Siva has confirmed this is an issue in Weld integration +Applied the patch into master branches. +Updated previous patch with minor optimizations +If the number of times the task tracker got blacklisted is greater than or equal to then the job tracker declares the task tracker as deadCan we count this only from successful jobs +Updated patch to address both comments +As per the mailing list discussion proposing undoing this change then subsequently providing the facility via an alternate route in +Scott can you please take a look? Is the use case generic enough to log this message as an error or it will have other unwanted +I just committed this +Attaching which address the javadoc issue +Committed to trunk with revision +Im working on a trunk port for this +have other people experienced tourbles with andor this patch that i should be wary of +The final version of the patch +I can add a comment +This passed locally testing with Hadoop and Hadoop on a private Jenkins with and on but appears to have introduced regressions on EC Jenkins build +It will also be good to expose it via a Hadoop Metric +It does not modify the executed code itself it just adds categories +The unit test is working here +I added the additional configuration setting to make sure xceivers are always available anyway but it can be removed +I am on this too +Now updated the file against the latest sorry ignore comment above +So dropping thisThe patches can be reviewed while I try any possible code reuserefactoring and making comprehensive +It checks for Null in the lookup for the job in and returns nothing to the TT when such a state is determined +th nightly build and Im getting the same issue as the Oct +Was an issue discovered in the CDH distribution that was already fixed in the Apache codestream +Or maybe you can raise upstream? you have the screenshot +Requires + +Should be now handled by EJB should not be doing its own scanningrepresentation of a metamodel by +Regardless of outcome Ido think we should still try to factor out reusable code from andsubmit it to appropriate Commons +I have been making this mistake too often in the last few days +Thanks again. +Fixed +This is superceded by and the plugin proposal +Ive added the attribute which can be used as an alternative to. +Hi Gavinas i said i can login +This seemed OK for us in but is broken and returning result again in Grails in Git reopen this issue is not fixed for webflow tests. +Hence lowering the priority +but may contain same. +This does not have any priority at present hence Im closing this issue for the time being +adds a unit test and adds support for some versions of Hadoop with peculiar distributed cache +Thanks Alexey removing the file helped +Yup +As one example jar referenced in the error in the description is required by not by hama + added +I do believe its fixed +First of all this defect has nothing to do with aborts let me be clear about this! removeIterator is not protected and it should be +The whole escaping logic is dubious there are a lot of corner cases +with clean repo and could not reproduce. +This is more burden on the AM NM and confusing to tracking services like logsAHS +we should bump up s version not sOops fixed it +So this JIRA can stay closed phew. +Hi Eli Sure +Another thingAnother random thing that occurs to me is that we should check that the request comes from a cordova view before handling a gapexec request or else chance leaking Cordova API access to. +A couple of points about testing we should avoid making tests depend on example so that its easier to change in the future when we want we should avoid creating entire jetty instances unless necessary should work to test basic test to use and not depend on example as Yonik suggested +Done +Now this variable bytes was removed from +will also release it to with +Patch file for trunk +Covered by integration tests. +I would like to have different base files for different environments +This now works. +Added and tasks to that allow debugging and profiling within a generic run target to nbproject to point to proper tasks +I think Tom nailed the problem +Update in light of modifications +Fixed this by adding license file to and confjmx +applied on and trunk +Not sure how to check for the mapping document usage inside the transformation +However itd be good to have a review from someone with better MR knowledge +Hi NeilThe parser doesnt know the semantics of the mydepend URI scheme +zip so no asynchronous action is created to compress the file +See the for details on how to build +I already sent a mail to Ben trying to get those snapshots back and move them into to be caused by were working on it +Old unit tests were also updated to the new signatureThis patch should only be applied when upgrading to Hadoop issue to reflect changed Hadoop trunk version +Hey hey Tim mentions that the section Accessing Messaging from a remote client seems to have been removed and it seems that it is still removed in the latest version of the JBM docs from the community is this done on purpose? Cheers please dont set to resolution done status resolved until the task is done +All Kepler IS component content accounted for FSW Beta. +Like stated on ml before I would be really interested in factored out last week +After removal I obtained Errors Failures out of +According to Egors comment I replaced the three lines with belowsXemoptXXyumainyugetXXworkspaceharmonytestlogjitlogseqnbclass +Committed the patch to +Just glad to help with such an amazing project +Thank you +I guess Cassandra would only need a fixed count of exactly making it aAs you said we would need a counting filter with bits per bucket if both bits are set the bucket has collidedIf any of the buckets for a key have collided you cant perform the delete but that isnt the end of the world hereEDIT Actually if you couldnt perform the delete you might end up superseding a given sstable multiple times so this is something wed want to avoid +Unless I get more information on this I will have to close this as cant reproduce +Ok just found and fixed the bug +Attached is the patch to fix this +please make sure you add more java docs and forrest docs for bookkeeper +As for your questions I dont know how to answer at the moment +It seems the ui for dataimporthandler doesnt pass through the clean optimize or commit commands when clicking execute +Great these leaks has gone please commit if you canRegards +Fixed in SVN trunk. +I have tried with all these versions it is giving same problem +Closing issues before next release. +small ui patch with a approval +So I am marking this to go to release notes +Ive tried both safari and but I cannot reproduce the issue +Patch applied with thanks! +So make sure you pullYou can check how you are doing by running mvncheck in the root of the project +Generally speaking the Axis plugin is enabled now and we have other to track the left issues. +Gunther which version of Hadoop does this appear in +Then of course you could also add more but this would be the basics +Very nicely done +I agree that this is difference +When we undertook cleaning up editlog one of the things that was good about the design was NN could roll editlogs independently of the backupsecondarycheckpointer +Im seeing the same exception under more or less same circumstances +I did an experiment in a dfs cluster with node +New version of patch with comments old code has been removed +Yes ant test does workThanks +Thanks Timpatch was applied as expected +For some reason this hasnt been run through hudson could you resubmit the patch Brian +bdudney could you check my work +Im very wary of making the client API any more complex +Shouldnt the I have X tombstones in this sstable code kick inAs in the code in STS? I suspect were ending up under the min compaction threshold all the time +This member is initialized in the Constructor and set from the Parser +Yes the same behaviour is in +fixed by revision +Maybe a note in the guide on this possibility would be useful +Updated patch attached +I committed this change to trunk. +committed the missed file in r. +Correct the web app class loader used for the enc was not readily available +See the use of +This patch is well unit tested by existing as well as new tests +Otherwise we will continue with the for loop +Wont fix in the JBR support +I have tested the private scope change on jetty and tomcat with and war target it seems to be a valid solution +Updates appear in Latest Alpha Manuals doc so this issue can be closed. +Thanks Mamta for looking at the issue and point out our intenstion is to throw the error +Thanks +A JEE compliant appserver should deploy compliant JBOSS really require repackaging of a compliant EAR into a JBOSS specific formatI always try to have that are accepted by the SUN EAR verifier from the JEE reference implementation +Yes and we probably wont do a dedicated build for now +Added question in FAQ chapter How to use subTable in combination with dataTable? and link to the topic on the forum where this problem is discussed +Here is a mockup of how I envision it +I think Christian is right Id be handy to have javadocs and sources boundled with jar files available in Maven central d like to take a look a that +In revision updated Jetty not to serve files out of or directories as per rfc spec +Assuming that we get resolved you should be able to use the browser history which will fix iFrames but force you to deal with the old and? bug in Android and. +Where did you find it? My aim is to close this issue with a valid argument +While its possible that there was an error in that code completion tool there is still something wrong on the Struts side because as I mentioned setting emptyOptiontrue had no effect. +number of handles is still fairly high but not actually a problem IMO +Attaching you mind writing some standard junit tests on this one +something along these lines is needed for patch should take care of the problem +There is a also a Reconnect action that will refresh the selected servers children +Will let you know either way in a day or two +I made a couple of additional minor edits to the patch particularly in and +Integrated to trunkThanks for the patch Francis +The idea is to be able to tell the UI user what the current cluster configuration looks like +An example of a workaround to force is available on the JIRA +That Trie code is not actually used +has been added to source repository. +Great tested +Adding support for the Camel SQL +No solution patch proposed +Modified applications and applications not to bundle the jars in war +javac apparently has problems when source and target differ +Actually the active goes into safe mode on my machine because it thinks there is not enough space +To clarify I used raw sockets inappropriately as pointed out by Chad Harrington on the maling list +It would be better to hold on to a few TB of replication logs then replay that after hours of downtime than to mess with the stuff since youd have to be careful to hopefully avoid duplicating keyvalues +As the data of request is null an exception happen in line +This will be available in tomorrows snapshot feel free to give it an early try! is solved in nightly build thanks! +Would you change how score is handled so its like other pseudo fields? I doubt it +Hi RobbieI tested your commit and it seems to work fine +The previous patch didnt help +Also I think we can remove the sync on do +Maybe someone can adjust the Affects Versions field so this issue does not get forgotten. +I dont fully understand why this is necessary +In practice a lot of domain object graphs are not appropriate for use as backing beans +This patch adds the to +I think it happens when I have an active session open via a facebook login and then stoprestart tomcat +Damn I have only just noticed that I quoted the wrong example +We dont want to a structure that doesnt reflect reality very well +Results in ambiguous answer about existence of a cell so elevating to Major +I agree the priority of this issue should be higher +The patch is not attached to this ticket anymore +How to change timeout for for IOS? +bq +Reams of rules exceptions to rules strange corner cases unexpected consequences +A tomcat problem +Verified +I think by overriding the method in the subclass with a nop implementation he is effectively disabling the test +oh right thats true +The changes to Configuration in Common project are as youve said needed only in for the sake ofSo I think we should do the first for too perhaps as part of itself and do only the configurtion changes here +I believe this was resolved in when I created a Populator interface and a piece of generation code that runs at build time +Patch for BXML Chris the patch seem Ok to meHave you tried to run the ant doc task after applying itIf all works and the generated javadoc is ok go with the commit so we can get even this for theBye +for the new patches +Im not sure how much memory a float field takes in a class as compared to a double +But you are correct for other operations we could be strict and tie them to versions we know about +So the server would apply Increment requests with the same nounce only onceObviously +I have committed this +looks good +Looks good to me +Yes we would definitely store the raw HTML field +applied the main patch I think the thread safety test is problematic due to timing and threading so we can just retain the used the method level syncs from the patch since it would be equivalent to blocks that sync on listeners +defer issues to +Thanks to Sharad for helping me debug the javadoc issue +We then need to adjust proxying requirements to only affect business methods for doing? Or not a problem in reality +pwdjbosslocalCRedsfind +Thank you Adrianthis new patch should be better +After being able to consistently deadlock my cluster Ive applied this patch in production and dont have any further issues +Have anyone found a workaround for this problem +Thanks Josh +Apart from this information no call to any other method in the is even valid within the context of a +I think Fuad has a point +Forgot to set proper issue possible solution is to add to groovyc args an option disabling ASTT by its +changed the test to what? +Patch to resolve test hang +It seems to nor have that +That can lead to problems since request objects can be stateful and essentially reused +This bug has been marked as a duplicate of +Ok I am battling the stupidity of drop box right now +I suggest we assume that name will be a required attribute +Closed. +Ok Im good for a v then +The issue is not caused by Spring alone it is caused by the confliction between ORM layer and Spring that both use CGLIB to generate proxy classes Spring is confused and goes into panic for every DAO objects it created hundreds of proxy objects behind itThe workaround can reduce memory consumption from M to just a couple of M however the new proxy mode has an some impact to DAO I still have to spend some time to get all issues resolved +Not totally finished +There seems to be no way to override or remove this configuration inherited from a parent pom +right +It would be possible to remove the possibility of collisions but I dont know how to do that quickly with minimal codeJust compare the name +I closed this issue because it seems to me that it is already deploying an uninherited uninjected uninterpolated POM to the remote was fixed in the release plugin. +Reversing order of file attachments for Hadoop QA +Applied proposed patch +Second I agree that of course URL is not uRL but conversely you have to agree that uRL is not URL +This is the +Translation of the German textA runtime error you want to start the debug modeLine Error style is null or no of Microsofts debugger after the Javascript error window pops up +Please commit to +Included in release. +Please ask on the user list or provide more details if you think this should be reopened +Thrift tests pass cleanly +Thank you Lily for testing the patch +Hiin order evaluate patch from more easily I moved code to class and I created a test case for this class with escaped quotes +I am using and was trying to use the existing extension rather than a copy paste clone +Thats what this implementation doesCould be used in anything that requires a request to multiple solr instances not just distributed search +Please confirm this works or not +Because I thought that both cases werecaused by the same problem I created one bug +Now that hcatalog and hive jars are merged all hcatalog jars will automatically get loaded +My guess is that this is caused by but I opened a separate issue since Im not sureThis is the same type of JVM performance issues reported by Mike in and myself inSetup Same test index I used to test faceting M doc index with fields fs a single valued string field with unique values fs a single valued field with unique values fs a single valued field with unique values fs a single valued field with unique values fs a single valued field with unique values I tested against Solr are of the formhttplocalhostsolrselectqrowssortfsascbranchx fs sort only mssort against random field mstrunk fs sort only mssort against random field msThis is not due to garbage collection or cache effects +Could you provide your +I asked the lawyer of to fax the grant again +I could replicate the same error the user got on the users at the forums for more detailsreserve need to generate uniqueID. +Proposed patch adds support to server and client also adds test case +what I dont get is why we cant have SOA builds just depend on aggregate site and integrate often that wayBuilding of latest milestone is fine for but that isnt where we are at now +Maybe was improved upon by others after I submitted it to make it also handle deletions with a row locked +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +If not reverting is preferred Im attaching an additional patch which fixes the broken tests + +the source is here authordisplayARQReferenceGuideupdate as you see fit +pending trailing space on commit +Subsequently user anita revokes SELECT privileges from PUBLIC on table t +I am curious to know what the problem was +this is fixed now isnt it +I prepared variants with external and internal link names +Spinoff from +Class and test case +cool thanks! make sure to mention claudio in +I retract my hasty judgment +FAILEDFRAMEWORK attempts should make status as KILLED +Ive meanwhile managed to run the test from IDEA and it passes +Im unassigning myself from this issue as there are so many votes and I consider it a hack to add a change whos soul purpose is to change the behavior of a query parser and I dont think such a thing should be committed +Everything works fine and new application is now working but Delete original directory and you will get an exception +Your fix is minimally invasive and seems like the most straightforward way to solve this so well go with it +This is related to the issues that AS is having with on JSF is critical that the embedded jopr tests be executed with JSF and the latest code base +Committed to trunk with revision Committed to with revision . +For anybody browsing there is a discussion about how the should be constructed to handle this +closed with release +I do hope this issue will be solved by I have attempted to reproduce this on multiple platforms with no success even with your detailed steps +Managed to reproduce the problem and test the patch before I checked it in +Do you see more live instances of than there are live threads +This is a valid issue +Since weve agreed to put in bugfix mode Im moving this one on if we should make this one a admin level config or a request param +in reply to your earlier comment no workaround was needed. +patch to docdoc adapting to proposed moved file +This is why Web Flow manages the hidden field +patch just from a eclipse rename of and +id prefer a separate module for the remoting layer instead of putting everything into the current module +The commit comments were as followsAdding a junit test for the standalone test case provided by Dag for +Closing all tickets +jstplugins + the patch looks good +Im looking forward your comments on this issue +Please find a Unit Test attached which is generating the test dataThe test crates a temporary ZIP file with files Dir DirImportant +Logged In YES useridWhen turning echo on I can confirm the double backslashbut the batch file works fine for me from anywhere +Unfortunately with the compatibility behavior between the existing types I dont think we can actually hit a situation where well run into this issue with the current types +It reports when safe mode enters the extended wait period and says how many seconds remained till the will leave it +True +Raghus block crc upgrade code throttles the deletion of +And I think the fix here is not reasonable +Fixed the issue by converting the URL string into a URI and then get its pathJesse did you submit the code of the to Shindig as patch? If so can you check if it works with a space in the classpath +I guess if you dont use and HTML mime type and put any HTML tags it will be plain text? +has a little weird way of getting the access modifier set through private protected public package s weird because if someone sets private true and public false it will be meaningless unless if you change other properties when one Javadoc task on the other hand provides something cleaner through just one property called access that can be set to protectedpublicprivatepackage so no chance of conflicting access modifier from being sure how freely we can change the weird but public API of +On my machine Im able to initialize the cache at a rate of k nodes a second +What operation were you running when it produced that backtrace +Version Released +Any update on this +Agreed Tim +I thought it was using the global app tokens file but now realize its using the tokens +testing now +This issue duplicates +I was using +issue one patch wasnt applied team is working on respin +Multiple blog posts and tutorials mention having to manually import in user bundles +Once this is applied there should be test failure +please attach an ssh pubkey +v only adds some instructions in the javadoc for the example no code or pom changes +Then +bulk close of all resolved issues. +switched the handling to dateadddatediff to the timestamp functions +to trunk +Yes there are incompatibilities between AS and and Hibernate Not sure how we can fix this without completely externalizing the query execution +Attached new patch to fix the intermittant failure +My DEFAULTSUFFIX is also +Had to change the import from to but than it worked! Thanks for the great helpCheers this helped +Our current thought is to generate a pom or some other artifact as part of the actual release process so we will be working on that route leading up to the release +Hmmm I dont see where this CGLIB thing comes from +This looks good try it out and see if it helps +when any of the goals in the rundebug actions contains reference to warinplace then the project infrastructure will redirect the deployment to the correct location +Fixed in commit afedccddfffeebfaadafdedd +that user should make sure dataDir is good +done +Using I get a similar issue to this one when using and. +Thanks! +Moving to release to make it available as soon as possible for ESB and the SOA platform +change fix version from to on resolved issues so they will show up in release notes +A workaround for this is to give each subdirectory a directory +So we have a plan that looks likelimit order by projectWhat you are looking for is an enhancement to push the application of the order by and the limit ahead of the projection or alternative to raise the subquery evaluation above the ordered limitThat should be easily doable in +will have the same +Does this only happen with one particular test run or with any test run +The VM needs to be stopped and started to do so +Hey BobSince this is in your area theres not much I understand about it currently +Jochen Thanks as well +Patch in revision +During Preview its pulled from your localreferenced Connection Profile if it still exists +Thanks Paul +Alexey please verify +As a consequence build still contains the old code +Finding out the simpler solution is better is always pleasant but many avoid doing the tests to check it +on the current patch +Resolved. +It was only not working because of something I could avoid in the Gentoo scripts +Fixed the index new more scalable version will be uploaded pretty soonPlease do not download this patch as of nowThank EliYes thats exactly what I am doing to pass in the argumentsThanks for the suggestion +moving to M +We can simply svn cp everything from trunk into new repositories for each new TLP and let them figure out what to keep and what to drop from their respective projectBefore we do that we should probably decide if we want or abstract class extension +Will raise a different issue if seen again +Nice thanks a lot! +Heres a patch for LucyBuild the ModuleBuild subclass which is used tobuild Lucy for Perl +I did some profiling with the +See +Ted +I just wanted to make sure whether they are really needed +Personally Ive given up on interfaces for stuff with more than one method with at most one parameter +But also occurres on STS Eclipse JDK on OSX and other platforms +But we shied away from it as it may introduce another component to support this feature +After removing the jar it works +contains the standalone logging level configuration +Future version of Drools will address this issue properly +Thus any time that categories is added as a search term it is going to require a more complicated search algorithm +Sorry to have to ask this but where in Cocoon CVS is this thing +We dont require final local variables for inner classes and dont make full generics checks +Itcan only add unecessary cost for those callers that will never pass an indexI also dont like the API inconsistency that this would introduce to currently all methods require index be +If this is still problematic and the workaround not sufficient please reopen. +Permissions schemes is yet another one +test you collect the jclouds logs? Maybe Adrian can help +FWIW this is not a SOA plugin issue + +I agree with for +Can I get a sanity check for this? It changes some of the WFS bindings +WFK went GA on Friday and the maven repository is in the process now of being updated +Committed the clean up patch +Here is the patch that fixes the problem +Steps to repro +No response assuming ok. +What I fear is that some sensitive informations such as passwords could end up in the log and that would be a big security issue +Btw looking at the patch I was wondering if it would be interesting to have a correlationId so that we could trace the exchange to its parent exchange and fromthe parent to the and so forth +Fixed +Personally I would like to see some numbers before committing resources otherwise I would say that the experiment was a failure. +patch attached +Committed in revision +Address last of Teds comments +Ok here is an updated patch +Yes I considered it to be a special case of b But yes youre right +What is interesting here is the fact I made a similar test myself with the differnece that I created the classes and mapping by myself and the tests run without an error +Create a patch file based on Daniels suggestion two interceptors are created for close the created readerI am not quite sure about the PHASE now I use the POSTINVOKESlightly modified patch applied +we really ought to write up some directions about how to modify Hadoop to log via chukwa +Sending a newline after the Forge plugin installation seems to solve the problem +Mac OSX is better linked with framework then ljvm or lserver directly +Im uploading the patch I have committed +from the table since they are also internal and not set by user +This demonstrates how easy it is to mobilize SQL data stored on a server and load that data into Derby databases running on handsets +The issue was happening due to incompatibility of Blackberry MDS tool and emulator +Added div that clears floating to fix the broken IE CSS box modelFixed in SVN revision . +Resolved with. +Filtering out fields for and properties for. +Its safe because we cannot run an infinite number of them because of other resource constraints like memory +The last two failing tests are due to +So im not sure i can even change the POM after reading that +Thank you very much indeed + +Please +dunno why hadoopQA tests did not trigger +ThanksSameera +Thanks +Ive incorporated it into the newest patch +Work completed +Castors Maven POM +Please feel free to reopen with a patch if you think youve got something. +Attached the new patch +Hrm +Its better to use the constant in and getDonebq +The situation persists irrespective of +Not reproduced anymore in at +Just ran into today and was a bit surprised myself +RTE was probably ok +I implemented the approach with svn r +Bumping to CR so Gary has a chance to no response will close as cannot reproduce +Duplicate ofPlease see the comments on that bug +It does support exploded jars but I seems like it cant find the manifest at the beginning +Increasing the amount of memory doesnt seem to have much effect since its the property that decides whether or not hash joins can be chosen +Another question concerns watching for authorizedkeys changes +Thanks fix applied. +Assigning this to since it has been identified as a regression +Thanks +Looks good +So updating to was almost painless +I assume the does not exist in the any of the +I just applied the patch to the main CVS branchThanks to William Lee for this patchThe code loops over elements in a Vector incrementing the index jSome elements should not be processed so they were removed from the VectorThe removal of element j shifted all higher elements down and made theVector shorter +plugin now uses encoding to generate the adequate +Why not put a right in front of your that only has a single transition to your +exceptions you need to remove the author tag and follow the hive style rules mainly spaces for an indent not tabs + has a workaround for this issue +changing scope to reflect impact of the issue +I dont believe the globbing code I cleaned up for v has these issues +Thank you Binglin. +this seems to be the most likely the result of those changes +Thiw patch fixes the problemRegardsRashko Rashkoyour patch is in rev +This problem was addressed in Update of CBuilder by Embarcadero in their CBuilder Update +Rohit can you check on this +for when we want to do threaded splitting of logsNeeds testing +This is our DBSQL Anywhere Network Server Version for the feedback Ive downloaded Sybase Anywhere so I can do some m planning on adding a new that uses the syntax that works for SQL Anywhere should be able to do that for +The difference is that other client commands that arent listed here wont pick up the CLIENTGCOPTS but frankly I dont think gc logging is very important for client commandsAnother alternative if we dont care about client gc logging is to put include the SERVERGCOPTS directly in the HBASERESTOPTS etc in +Please disregard +I agree with you that this exception is misleadingThe problem I found when doing this was that wsdl specification does not define the HTTP binding properly and there no details even in the basic profilethe only sample given in the wsdl spec has a rpc type sample +Other than that please go ahead and commit +Here is an html file that needs to be converted into overview doc +Committed to master. +I will close but I think first we need to make a new issue for the error reporting improvement patch Mike attached +Thatd just be a oozie change +Please read the issue I linked. +The place this is controlled is in the xml file for the events mbeans and not on the properties in eamconfigprops +Hi AlexanderPhysical operators are in srcmainjavatajoengineplannerphysical +Yup hence theres already a pull request for it into the master branch +Fixed in both the branch and the trunk +reformatted runtime using felix eclipse template +In a regular CF the parameter should be just CF +Will have a look at what is wrong shortly +We are seeing no leaked file descriptors +You can prevent it by adding next line to the of ASJAVAOPTSJAVAOPTS XXM +I think we should change this in for the reasons Uwe pointed out to disallow changing the delegate after construction +Sorry for the false alarm. +svn ciAdding downloadTransmitting file dataCommitted revision +That is goes from performance thoughts and JCR doesnt need that multilingual support in JCR on use dialect on singlebyte default encoding of a database +Thanks very much +Additionally in case the existing lists lack moderators feel free to add me there too +There is EAP Runtime container that is unbound and the Portlet Libraries container containing no entry +Still ongoing +The NPE I mentioned above seems a different issue +I assume that you were just going to count parse invokations right? This really tells you the number of statements not number of lines in the file since you can have empty lines which are skipped and also statements +Maybe the nondeterministic detection by the CME that is in place right now is enough? the record the attached patch implements a simple check that prevents an observation listener from accessing the session it was registered with +I cannot reproduce this with current +Should look at at the same time +I also tested that this change did not break the fix for +I think its better for Hadoop to do it in an authentication independent wayOkay +Your original plan is to have separated patches for various components +Eddiewith your r is generated into controlstestbuildbeansrc but not compiledWhich might be the cause of when compiling controlstestsrcdriversorgapachebeehivecontrolstestdriverthreading which is checked in by r but remain inactive +Reverted this functionality in favor of style +Especially if rm deletes more files then they want +It is not complaining about not finding the key store file or any other keyStore related error +Verified that bug exists with but seems to be gone with +In DB trim is broken for a different reason +I plan on rewriting a TCP transport once were on JDK anyway to run on NIO so I dont want to make too many changes to TCP as it stands for will check how it works the way I would recommend to set usesendqueuestrue in the default in sourcesexamples +Thanks owen for review +Committed to trunk +looks good! thanx for making the changes + +I have promised to try to help with this but summer got the best of me +Shall we really consider this or mark it as wontfix +Verified with EAP. +Hi +Otherwise the problem still exists exactly as shown +I think Ive discovered another issueIn line readsscp SSHOPTS PRIVATEKEYPATH rootMASTERECHOSTroot +Although is a bit flaky in trunk this patch should still apply +Committed at subversion revision Touches the following filesM javaengineorgapachederbyiapiservicesloaderM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslang +If you want I can make changes to code so that parameters of local service invocations are checked also against nested attribute definitions +Thanks Thiru! +Thanks! +Updated product code changes to resolve mapped superclass object id creation issue and supporting test cases +I just dont like running on libraries hence my question +This issue has been fixed in of the site plugin +I tried Safari and Firefox BTW +Yes this solved the problemClosed the bug marking it INVALIDBest regardsThorsten +I have hard time understanding what usable issues you are solving +I committed these a whlie ago +s not a patent issue +And this rule was actually merged with another one so you wont even find the RSPEC ticket implemented Its sufficient to just focus on RSPEC tickets. +Given the type of data mostly stored on HDFS how useful is it to provide an internal ability to print line numbersIt could be done perhaps a small bit by external programs using the output of s cat today +I feel implementation must be threading safe +Patch for and branch +Sorry about this +However I still can get things to work +You could use here actually since fsyncing it seems reasonable +Feel free to repriorize. +looks like a cutpaste error but i went ahead and added a test to reduce the risk of future regressionCommitted revision +The improvements look good but Im actually not sure why we update all votes for the same id to have the same zxid +Nope +Having a version indication on all pages might lead the users to think that it only document that specific version +Thanks for the backport Brandon! +The name of the directory in candidates and the names of the tags is maintained by to follow the RHEL standards so we dont have control over those +Marking for because although it is useful there is no patch yet +This is also true for sink if a throwable is thrown +Rescheduling for sooner as it needs more testing +Per Kihwals first comment we havent seen the error in but this is the same code so we suspect the problem must be thereWe should perhaps post a patch for also and get it in in a subsequent release +Running now +Locally I changed implement so that its possible to override +So IMHO it is logical if DRLVM java kernel classes are builded with debug information included +Both are GET requests +Closing issues for beta . +had unnecessary Runnable Sam +metadata folder in my workspace it may be from prior to reinstalling the project Im not sure I was worried that Id hose JBDS if I deleted it but could it have old settings that are messing me up +The master needs to blast region assignments out there in a threadedperformant manner rather than waiting for checkins to do their magicFinally I have decided to up the size of a region on my own +wondering if someone could review this +Remember +Trustin I dont have a clear idea of how you intend to do this so Im going to move this to the next release and assign it over to you +I had a look at those tests and it seems to me that the issue can not be tested in since the bug is not in a Hibernate support for stored procedures but in codeThe test would be more suitable to be present in +Correct me if Im wrong but the injected introduce the proxy around the domain instance already +Note that this is with the default compaction strategy +The attached patch removes it +This is definitely not a solution forI agree the complexity would be raised +on patch with test +is frozen and only security fixes go inPlease update the quickstart and the patch to Wicket and we will consider itThank you +patch for branch for tests included as this is a setup script +Ill fix this as soon as I am back at my workstation in a few hours +We defined that an unresolvable type variable is one which cant be inferred from the Event injection point. +This issue can be closed since it is implemented and fixed. + +Exactly this is a real feature request for a Seam framework to allow jbpm state maintain ajax requests on a same page +The startup output from Geronimo reports that the RMI port has successfully bound to port Since the connection was refused there is either a firewall blocking access to port on the specified IP Address or Geronimos RMI service is not listening on port +This bug has been marked as a duplicate of +But I will see what I can do about it +I have a basic registration page that I need to check in +Thank you Norman! Youre doing a great job! I can receive files now + +I am thinking that Gant should behave in the same way that Ant does in the same situation +Instead we should have a numkilledmaps +Integrated to and trunkThanks for the patch clockflyThanks for the reviews Sergey and Chunhui +Attached patch should fix the issue +it is a pain to have to specify the URI if you really want to dump the contents of a file in the cwd +Im fixing them right now +when looking up a WAL for example the object is used to check existence +Hi BatoszI checked with your pom an it works! so i made the changes to my original project and i have a trouble with the Testable class and after fix that the is there again +How does one add them? Or are they always there? +I think Alfresco does that and so does Penrose +Widget is all yours +Rev +I stand corrected they have changed their delivery optionsSignatures were standard then they stopped signatures to residentialnow They charge Delivery Confirmation per package Delivery Confirmation Signature Required per package Delivery Confirmation Adult Signature Required per packageI suggest Tie to the will cover both and use the carrier +We can assume the average person will be using a local repo at least from a getting started perspective If they need a remote repo then we should instruct them to see their local Maven Admin person and provide the URL to the how to install the EAP WFK repos +Of course I only askedThanks for fixing the bugJoerg +The tests warn Ive explained in my previous comment +Verified on revision . +Thanks Akash Pranay Others for your contributionChanges are in trunk at revision Ashish AshishThanks and Regards Akash Jain +Like in the is loaded on every request +Cool I agree with your points above +looks like the EOF bug was fixed in +which version of the site plugin are you using? is put the above pom into an empty diretory added in the build section and got no Exception +If you dont like the API weve designed feel free to open a discussion on the mailing list or even a JIRA ticket as Wish or Improvement +One more condition is that the independent values must exceed the threshold of issuing a single query against a dependent side +None of the recent hudson builds have had avro failures +Patch updated it was adding a warning on +Hi EricAre you sure that is the cause for that hang? I have the very same configuration here working lot with +Check if we can proceed with committing the changes without test if we are not able to find out the actual problemAs I already mentioned patch has been tested manually +Can someone please commit the patch I provided here for? The patch should fix the problem if its spurious +Brett have you made any progress on your investigations into this issue you reported +or trunk or both? Just is fine Please add to the fetched stored headers +Use the native instead of a wrapper on suppression an object creation for each call to the iterator +Just to clarify +Almost every class we maintain contains commons logging imports and usage +Here is a design doc for this featureWe will be creating a new branch for this work +However this is undesirable when not wanting to introduce dependencies on classes in Controller method parameters +Thanks John +Changed the issue subject as this applies to all plugins that provide supporting de la revisin +The new scripts will default to using JAVAHOME for the JDK location but you will be able to override this for each Server instance without affecting the default +Committed to trunk +I dont really understand completely the possible codepaths using this code +I expect any installation of spring would do that +Hi ElliHmmm +It seems that is a standard way in JerseyBTW do you see any performance problem in trunk +Approved for +We may have to modify super classes +wicketsrctestjavaorgapachewicketproperties +For small chunks it is far faster +It seemed okay in the end so closing. +Will update my working copy of trunk and try again +Ill see if I can dig it upKihwal are you on commit now as well +The interesting thing is that after importing the Server project the launcher plugin sees the existing launch since it will not let you create a new server of the same name +In other words it seems to work in Windows debug version both in debug and release versionit doesnt work in Windows release version +Thats why I have provided a patchwhich java +I renamed the parameter from allowedViolations to max to match better with the parameter names suggested in if we later decide to implement that. +How about WITH FILTERING ALLOWEDAgreed thats a better name +Isnt the latest version of workflow and coordinator xsd +This issue related to the modcluster documentation before it underwent a major overhaul as part of the HTTP Connectors Load Balancing Guide +Even if it is called twice it shouldnt be an issue because does its own check to see that it hasnt been released more than for the noise +My pleasure +If someone can figure out whats broken Ill fix it +Russell please open a followup JIRA for that one now link it to this one as related and then edit the title for this one +Attachment has been added with description This is patch for FOP to solve problem with borders inside table cells with background color specified +I think the actual calculation of refSize has improved since the change and that is what is making the test fail on the bit machines +Documented as tech preview component in the EAP Beta release notes +We should not show the permission tab when edit portlet in user should remove permission tab for container has updated +Only remaining question is Should I create a separate inner class for inside? Earlier code was reusing it from the Generator class +proper patch that actually shows the missing filenamedirectory +Does close this issue +Ive not been able to reproduce this made s of secure shell connections via bash script with and Hadoop +will do +A showDefaults query parameter causes them to be included in the responseBy default all requests are indented JSONThere are tests for everything ant test passes under Solr ant precommit passes and Ive added entries +Could you rebase your patch +With my modified Trinidad version I can run your test case flawlesslyNot sure my solution is the good way to go though +I used the style and svg attached to this ticket with and the geoserver preview it worked without problems I could get feature info right up to the border of the square +Patch for +The patch looks great +I dont think balancer handles this right now +I now see that it is much smaller than the other oneShould I have committed both +The source from these tests is part of the Findbugs project tests +With fewer threads loops Ive seen this occasionally succeed +Close as a duplicated issue of +The reporter does not follow up his request +If you have more suggestions for logging let us know we wont start the next run until tomorrow anyway +It might be outside AS and MC but the wiki and jira calls it Reloaded +Closing this one +Patch looks good +Still in this case the documentation should change to be explicit as to when a class is required versus an interface. +This is followed by information on the skills and software required to perform the activities presented in this document +Previous patch forgot to add a delegator method in + is also required to avoid javaxjws . +heres a delta of cleanup to clarify the above +I also started to investigate this +Maybe for we could consider this or even better provide patch to Luna to provide this shortcut by default? +However this wont affect . +Can you review for me pleaseThanks verified that the comment is there +Can you post your document and schema so that others can reproduce your problem +I guess I have no permissions to close the issue +Im still looking over the patch but the only comment I have so far is that it seems to include diffs for a couple of files it shouldnt +Attached is a patch +I see thanks for comment Adam +I didnt add the requested as it could be or some other prefix +Committed to and trunk thanks for the reviews Aaron. +whats the difference between corelib lib coretarget target +Can we please make sure it gets resolved next time we do a release +Committed at subversion revision +Again the rational is that Since that code was added weve introduced a better way to deal with node being overwhelm by hints in sTHE +Thanks +Thanks +There is no timeline implied here +All it takes is to look away for a couple of minutes in my case the error was a reference that matched the name of an instance field in the static method +Of course does dependencies matter we should not add these lightheartedAfter looking at the code there is no need for to actually be here since it is just used for classnamesChanged title of the bug to reflect what its about +Any hint for this issue? I dont have an Oracle JVM and there is no chance Ill install one any suggestion +Categorizing bugs related primarily to stylesheet elements etc +Committed to trunk +There was a bug in the implementation provided three years agoDue to a wrong increment the old code could cause an index Exception when renaming nodesThe new patch fixes this. +Given the changes made in response to should I submit new patches based on the state of CVS now? The old patches did not change the contract of this family of methods and neither would new ones +We should consider the amount of computing involved in the mapreduce tasksThe assumption expressed in the description may not be satisfied in various scenariosI think we can provide abstraction over key space partitioning by introducing an interfaceThe idea would be one implementation +Collecting meta data will help a lot +I will add a comment to +Thanks for the patch +was thinking about this issue +will I reopen so this issue catches a patch to change log level +This problem and the attached patch are similar the previously fixed and its patches +Incorporated changes for Exception +Thanks for reviewing +ve committed srcmail and examplesmail to CVS +Makes sense to me +what is the contents when the warning occurs? Can you show us the warning that you get also +Issues like this have the tendency to stay open for years without anything happening +Bill I dont undersand why you need to modify the +I was excited to use evaluate grails for a project at I was stopped dead in my tracks by +Depending on our delivery schedule well try to use the SP and delete our temporary for resolving this issue +If they were meant to be included in this code drop then this bug issue should remain open +localhost in is not sufficient though right? What if user wants to run in mode +This is fixed in SVN now isnt it? Can we close the issue now +An updatedsimplified generic version that uses the The format is different but seems to match the JS needs betterI will commit this soon so the UI stuff can continue in a different in revision Lets keep the existing servlet in place until the admin UICan we close this JIRA? Or are there additional issues pending +I think Luc did a great job providing the right information to users +With map side hash based aggregation added in queries see improvements upto +I have since reviewedthe plugin installation web page and it provides guidance on cleanup of theold pluginsThanks for you assistanceCheersJason +I just committed thistahnks pat. +It does not +The code generated by does not seem to do this +Who wants to do the legwork to resolve these issues completely +Attaching patch +Fixed in cfbffccdbfdefdcc +Maybe we can massage the persistence unit root URL that we pass to Hibernate there +This new attachment contains copyrightlicense information for the test class +This way the css stylesheet can replace the existing one with confidence that no ofbiz functionality be put at risk +We can relax it later once we have the multiwiki stuff up and running and we know exactly what the limitations for the space names are +Reviewed all these issues and marking them as closed. +Sorry yes after talking to Stu I found that out +Done as Niklas suggested +Ill try and polish the patch a smidgen and commit it as a start soon +Added the test case +sorry it took so long but it is in +So is this only a modification to branch? Id actually like to see a better method hash on trunk as well since is being voted out and the original global method cache is just a hash too +And for the safety of the double check its better to add volatile modifier to the +If I run the script with the code via groovy the include has no effect on the nested javac task +Sorry Olga my mistake the xml files have now been committed +This is a departure from the common practice of other JDBC expressions not inheriting from equivalent kernel expressions +Thanks +Please resubmit the patches with a grant to ASF +Last week we moved from postgres to postgres and this revealed a problem with parameter binding on some hql queries +v +Log in as that user +As for the mapping feature collection issue I leave it to you to use whatever is appropriate for complex features since that class is specific to that +Not resolved yet. +GreatWelcome to Tajo David +Moved from contrib . +Were using MINA via DS RC +I will provide patches just need to grow few more neurons in this protection domain first +How do i get the latest version of those files? I am using maven as build tool in my project. +New tests pass against both the latest Harmony and the RI +The directory needs a bump can you edit the KEYS file perhaps with another key or whitespace to trigger a commit +Here is the patch for this issue +Reopening to correctly set the Resolution it since its not a bug but is intentional. +I have incorporated your suggestions in a second rev of the patch and +This may be revisited at a later time +Or was my verification flawed? +Applied at r +Upgraded to the newer version of the parser in. +Thats cool thanks for your update +So if the user configures the HC to use a profile with this subsystem on an AS server it will fail on the Other handlers are changed to do nothing in +It would be good to have a test for this +The patch is to apply to the svn version tagged EJBRCPatch +why do we need it other than to infer types +bq +compile and run Pig without any manual intervention +if you use the fix i added making LDAPCASOPENID inactive would be using the following configie +so i dont talk about subpages +Simon can i close this? I think you have fixed it with in +Can you identify where we are making the initial socket connection so we are only doing this once +might change that back to before committing +I didnt write new test because this patch removes one overridden method and I have tried the test in the contribution of Harmony JIRA the result is same with before so I considered this patch passed regression test +and if Google is still handling maps search crawling this way. +I forgot to add the license comment to sorry +Hopefully it will fix Glenns scenario +I think this would be totally confusing especially when we consider script maintenance not just script writingYes I do worry that pig queries will become like many perl programs that I have seenI dont find it very confusing to use anything that doesnt have a name conflict as the meaning of +When reviewing further please look for related popup panel issues like this +Explanation added to LM models query boost added +Nobody is working on this right nowUnassigning it and moving it to the Unknown fix think its a useful issue +Your wish is my command Stephen +we should run the comparisons as a public static void main i think instead of extending +Core tests got a for failure of Ive checked the console output to realize that is hit again + +Thank you for reporting this I will try to get it fixed for +I like the simpler approach +The cost to benefit ratio seems really bad please dont delay the release of for this +Still they both extendsbq +query plan +Hi all in case someone had a chance to try the latest patch please let me know your feedback +Same as for will we have the script +The current one is lines long! Perhaps delete entities and fields till you have a minimal script that reproduces the problem. +Patch applied. +Cygwin on Windows also has binbash as well as binsh +Id like to chirp in here and mention that performance of jruby is of the utmost important to users like me and many others +Pls submit the patch to keep status true after +The client connects to the server on the main port and notifies the server which port it has openened on the client side +Attachment has been added with description Patch to change Output web page to show that PCL renderer doesnt support applied thanks +thanks Jakob for the review. +Bug has been marked as a duplicate of this bug +Simple patch extending runtime type checking to Circle and using the Shape bounding box +Idea about rollback came from this feature in Camel +Makes sense to me +Any interested in porting it to the new runtime +If it is SOAP encoding we cant +I also fixed a in package installation +Interesting +for precommit tests +Seems some problem with hudson +Happy to go about it another way but I assumed since this is an init script and Macs do not use init scripts that wouldnt be a requirement +Ok Ive gotten rid of all the duplicates leaving the one page +Verified by Elena. +Will come through in beta. +In that case since hasnt gone out yet either including error information here or starting another JIRA is fineId recommend including the error information here until we can determine if your problem is the one that this patch is supposed to fix +But even without those suggestions this gets a +I just committed this to trunk and +If the RM allocates a container for an application whether the application ultimately uses it or not it is a resource that is unavailable for other applications to use +GirishSailaja have brought up the tests on rb. +Please do again with svn diff +I will submit a patch and test for this +This is an issue with the plexus archiver bringing in the wrong plexus container again to Brian and John +The first was to use the WC XML Schema to describe our XML messages +Ive changed a faulty unit test for this case to verify the fix +This patch adds functionality to add licence files to those missing oneIt will either create a new version of the file with +FYI weve fixed the problems with the repository +Thanks for the two suggestions +Thanks Yip +Backported fix a to with revision Regression tests ran cleanly on the branch +Rebased version of new test cases to test tuple and bag scenarios moved to a new test fileFixed the exception handlingAdded detailed comments +This removes QE and productization team as bottleneck since you dont have to wait for them to setup and fix your hudson issues +Did I request the wrong partition to follow? totaltimeblocked is always zero as well as constant queue size of zero +The enginens is now deferred started until has been fully started similar to what we do in +This way it listens on local IP address and the machines domain name +Disable icms and add a bit of doc that newer jvms are a bit better around avoiding fragmentation +I have not had time to really think through it +batch transition resolvedfixed bugs to closedfixed +And stabilize rest registry and follow up with release which will remove this limitation +The next step will be for someone to take this user documentation and use it to the feature +Of course the ParserWriter implementation should come with a variety of unit tests validating the functionalities +I had seen similar issues before and this should be a generic fix for any case like this in a foreach +Hi Todd sorry the lag in my replies +Repository project contains an abstract test class and a host of test files that can used by detector implementation test classes +according to the JCA spec the following are the standard activation propertiesdestinationdestinationTypemessageSelectoracknowledgeModeclientIdsubscriptionDurabilitysubscriptionNameAll others including mappedName are vendor specific so if you want your tool to be portable you should use destination instead of mappedName for both +patch looks good +Choreo showing the currently detached Message and Choreography Task nodes +Sounds similar to +All patches are based upon trunk can be applied independently from each other. +So thanks for the catch +Attaching a release note for this issue +And since theyre rare its pretty hard to repro and investigate further +Im hitting the same problem for both and s a problem with LDAP will be resolved shortly +Oops! It looks like this patch never made it into trunkCommitted revision This was reported in Bugzilla at bpel +usrsharewebappsclientAlso I did not see this UI issue after I have installed the latest master buildPlease verify once againThanksPradeep is observed only when upgrading form to Master and when cache is not clearedClearing cache by rm rf varcachecloudstack varcacheyum varlibcloudstack usrlibcloudstack usrsharecloudstack varcachecloudstackTo avoid this issue +Apparently my test was incorrect +Thanks much +Manually tested! +StrangeIm going to have to try the same case with a straight trunk build just to be sure I didnt mess anything up +verified at have tried but it doesnt work +They storeretrieve the timeouts in UI and pass it over to core +It will become available in when the new stack passes the CTS and is backported to is a working implementation available in +Ive never used Jonas but I think that if you run javawsdl and then run wsdljava then the generated WSDD will contain reverseReturn +includes +May be either this one or should be closed +the javadoc warnings seem unrelated to this patch +Ill get a note out to dev +This problem should be fixed in new service +I think Justin included this plugin in the build so I am reassigning it to him for his consideration +Patch is now in CVS +Patch abides by Fix conditions in JIRA description +So postponing again. +Resolving +If colors still exceed the range we could look into alternative designs with fewer colors +IBM made the same statement and Daisy thought about exchanging the temporary file handling to swap into memory only and loading only the parts on demand to the serverSwapping from memory to is seen as part of the using container instead of the ODFDOM libraryA further idea is the setting of a environment variable for a temporary destination but I would stick on the memory only solution first and get someAssign it to Daisy +Attaching a patch for this issue +This is the nd attempt +The list consists of mapping elements +Can someone it +This has actually be done already as part of. +I tried both work arounds and still have the same outcome +What am I missing +The is also updated of the resources names +This is not a surprise that we have bugs +The same scenario for the signature for the pom doesnt verify while the signature for the jar does +mention that projects could be exported into imported fromHibernate Tools using Eclipse provided connectionRenaming Java project with Hibernate +Your workaround seems to suggest the customer neednt specify the replyDestination +Yes as its essentially an artificial field we control we can define the rules here not so much with the id as you discovered +Canceling patch so that comments can be addressed +Thanks for the prompt patch Ted +There are two reasons the when a spill happens is already there +I suspect that there will be more Vector implementations in the future +Its a more complex test with custom usage +We will use Maven and +Daniel makes sense +I agree with LeonardoAlex please write a mail describing your problem to users thanks! +Do you have any idea of when a release version of the plugin will be availableThis would allow the use of the release plugin on grails projects +CLA received ac requested +Fixes for other multigeometries under construction +We are making an application which uses this offline event of CodovaThis is Copyrighted work if my boss finds out +That forum is full of spam +I assume we are still convinced that this is easier than shipping them our challenge was being able to point our QA efforts as well as our customers at the right release version of the content +Documentation done +For now to speed up the implementation I am thinking of using a simple Java to keep track of duplicate keyed values +But Caching is still faster than if dont misinterpret your numbers +Anyway Im thinking I shoudl put up a JIRA for this and try it as a separate issueWhen the code died using my attempt at spill to disk it died right at the beginning of the run as soon as the graph data was loaded and the messaging started same as without it +But there will be anyhow be a risk that valid commands are intercepted and filtered +And because Im using the push system in a couple of places the result is that about objects are created every minuteAlso in the test environment the number of objects never declines +Updated patchset with working tests +If you find further issues please create a new jira and link it to this one. +Couldnt test with Groovy and because a joint compilation bug made it impossible to compile Gradle +Now with automatic index stat it is more needed +Added a check in the constructor of +On the getting started page in the compile section Have the mvn command for building against cdh I keep having to look through the mail archives for it +I think one of the Linux guys should try to reproduce this +Feel free to me +The patch instead keeps around and just sets the namenode to null +However I think that silently truncating oversized strings would be more useful +More +Ive been doing so for a while now +If you commmit the patch then I will test if I can really remove the code and if ok will create a patch to do that against the latest rev +Ill try to get the Tomcat guys to do it +Then they took the list of worker nodes and mapped to where they were in the machine room +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Patch file WADL documents and the imported XSD files +Javadoc is easiest to keep in sync with changes to the code since it is in the same files +Reproduced with target on d like to investigate this problem and try to fix this if possible unless someone else is willing to fix this this weekThanks! +thanks +With the problem is gone. +The article is changed and posted on Wiki +I have tried to incorporate most of the comments +Delivered to +Fixed while working on +This bug affests branch too +access will need to be created but should NOT be used throughout the rest of pluto +Second Im curious about your test environment +Fixed in after much pain and suffering +Interesting I will check the code in +HelloI cant reproduce this bug after I the sources from will close this will reopen this if I can find a way to reproduce Daobing +I changed it simple because its an easier to use data structure however Date can be used solely with some api calls +Thanks Ivan for the patch and Uma for reviewing +Ray I know nothing about Pulse but p updater seem to be able to install meclipse components without any Igor can you please take a look at this maybe this need to be bounced to p component at Eclipse since Tycho is delegating to p app to generate those descriptors +I +Updated patch that should now apply +clean +Could do Encoding Encoder Though the are not really encoders at heart theyre just setting up the environment for the actual encoders +Thanks James +Apache SJMS SJMS Component BabakDtest is used to specify the name of a test +Metawidget is the only use case for this right now so Im happy for it to live inside for the momentHowever this doesnt get you off the hook for +Hi RiniTo answer you last question first I dont have commit access +Issue resolved. +The patch looks good +Hi TimTed I would want to use Ganymede and Jse for all the illustrations +Do you have any reference of starteam about locale is set at server side? I think we can create test input file per your locale +Ill close this one as invalid because there there is no way to reproduce it and checking the latest code everything seems to be fine. +Yea I originally wrote the code that did the copy but in hindsight I think it was a mistake +Indexes for and Oracle ODE Hibernate RafalI was thinking of doing some optimizations on the queries queries around the largedata table are problematicLet me know if you found any other bottlenecks and thanks for taking care of this one +But of course this is just a start +Im undoing it +How to customize for workaroundI have a important complete project and my customer started to use Opera +Arun please revert commit from trunk +ftl to +Please create a new JIRA subcategory Bluesky for the Bluesky incubator podling +bulk close of all resolved issues. +needs to be fixed before t that be calculated be asking the view what is visible instead of having a seperate state +I dont particularly like the name Basic nor do I see how the abstraction adds enough power to balance the lines of code it changes and addsAs for the stacks in and these may affect performance and I would prefer not to add such overhead to our simplest fastest implementation +This patch will fix the problem +in this case we still need to release topic ownership by deleting the znode so we should not checkExistence otherwise it would callback directly without deleting the znode +I am resolving this as wont fixThe codebase in has changed sinceSo after the release feel free to take a look at it again. +Ive checked the sample page and it does function the way Id like to see Counter work +Please supply the patch for trunk onlyAlso there should no nullpointerexception error message checked trunk analyze command returns FAILED null so its already fixed in trunkI remove patch for trunk +I dont like automagic behaviors +Essentially the default implementation of this method does nothing +QE please continue with trying to reproduce this one taking into account the comments made in capture of the problem +committed thanks! +You have a couple of possible solutions +Set this to resolvedThanks to Nico for his contribution +Could you tell me what we should do? sent to contegix s been applied +I will add the documentation to and I will commit to trunk branch +Im assuming you copied the whole folder from pabgacmsil thats why it worked on your endCan you confirm if you can still successfully start even if you only have the in the directoryThanks +It however removes the shutdown hook if all the processes exited normally +Im working on a project which embeds version information to decide whether or not things get cached +Also updated some of the examplesHowever I didnt notice any other error +Hi Chad you are talking about the chapter right? That was an example to run it like a normal java program I am glad that you made it work on the cluster +These stacks arent exactly the same so Ill leave it open but feel free to resolve as dup if its the same root cause +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the codebase +Retargeting to pending review of patch +It seems there is no problem whether you defined the element or a ref +I dont think that provided would work well in that casebut maybe it would be a good idea to update the Spring dependency version to latest stable series version? Theres a lot of improvements goodbackwards compatibility and in of cases that should bemore correct than the current springFor a better solution you should probably improve somethingin Maven +A SVN account would help here +Postponing to after issue topic from +Hi just some minor comments about packaging I think its customary for an initial release to have patch level +Accepting +Just to clarify in this case should we respect the spec and not allow concurrent access or should we just always allow threads to proceed without waiting +Whew things got complicated in distributed search +Using the latest nightly build of Axis from Oct +This sounds goodCould we include the in the cache key? Then itd be functionally equivalent we could enable it by default +Im hoping we can do a if not this week then definitely nextNot sure what the deal is with that and other file +issues +So not strangely it is behaved like normal start event and the next activity is started just after starting instanceAfter this first execution a event handler is created by +ignore my previous attachment. +How many of our customers use Windows and will be pissed if this doesnt work +Such conversions are highly specific to particular index implementations so to me it seems like the best place to do something like this would be in the index implementations rather than in the main query engine +Most of these remove unused code but there is also one switch to using a mask rather than a string in a permissions constructor argumentIve tried to minimize whitespace changes in this patch +i will provide them soon +svn ci m Adding Locale overloads for to and to as provided by Benjamin Bentmann in srcSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +reported many times before and the answer is simple dont run schemaupdate when it cannot commit +to check if the case is covered in +Fixed in dc. +Yes months is much and also not so muchIndeed the best for now is to monitor maven guys initiative and hopefully get a solution for free +Renamed to since that makes more more readable code +We had a need for this feature as well to distribute decorator CPU load among multiple flowsthreads within one collector nodeThe attached patch is based off of trunkaaccfeaeffaacfde in git +The patch seems to deal with my issue +Will investigate as time allows. +if you have different schemas that should be shared with multiple cores +for quick response! Probably a stupid question but how should I become a commiter? Whom do I need to ask for permission +Took a lot of googling to get to this page. +And I am wondering if we should do this also when we do the rollback by throwing the wrapped exceptionI can add that however it has no effect on JMS tx being rolled back as there is a bug in the way tx management is implemented in Springs JMS We are adding a TRANSACTED property on the Exchange that is never removedThats not quite true I believe +Maybe its an os specific problem Ill look for that + Username pinguine +Yeah Uma shuffling logic will help if have more ledgers +Im fine with making these classes Luke already uses a gateway class to access friends +The build has an initial fix for this +Duplicate of. +Normally when replication and shared cache stores are in use the cache instance where the data is put is responsible for storing it in the cache store +The UI should change to a UI that starts at a fast summary of the index schema and fields then drills down to deep scans of different fields +In the end I am just a user trying to improve thingsFor the fileUpload tag I read the documentation and the recommendation but I decided to give it a try +Patch checked into trunk and branch +I havent looked into the command line part yet. +Thanks a lot Jacques to review this patchIn this moment I cannot check it but I will let you know you something more in the next days and I will provide a new working patchMarco +Applied changes from to branch in SVN rev to allow testing of patches applied to this issue in a working since is still in the middle of refactoring +cant resolve this issue unless we actually usebundle sdk +Mithun can you also add a test for completeness +Ive just been able to reproduce this locally +The patch must be supplied to the off release +We have another way to resolve without removing reuse +If the new bookie isnt in the metadata the entries may be GCd +Also if the mallopt setting doesnt do anything useful we should definitely remove it. +Same as before but all string literals are replaced with proper named constants from container classes +Changed type from bug to improvement +How did you get round this in your build? +I have committed this to and above. +This also means the regressions dont run for +Done. +It would appear syswrite is continuing to raise ErrnoEAGAIN even after the socket is closedAs an aside is there some particular reason this is on a write instead of using blocking IO or blocking for using select +Verified in jbt +Jacquesthe sfacontrol has been fixed in rev +Attaching patch from previous ticket +Can you test it on yours side as well +Does it fix the issue +I have updated the title and description to reflect our decisions +Manfred the jokey chest +bulk close of all resolved issues. +Whats worse it doesnt actually work for meI had JBDS B with EAP +We have no response on no probs +Rolled a new patch this one should apply ll need some help with this one my antlr experience is very limited +OASIS needs to fix it +I am willing to learn but you have to tell me +rofl +I do not have too much experience from JNI but Ill try to map something up unless someone beats me to itHow can we distribute the C lib? Should we it for multiple platforms +Added final and volatile as appropriate in domain entities +Fixed open range and mixed include and exclude bugs on the javadocs changes in rev +I was just trying to understand the proposal especially the part where you say rather than checking against cluster averageCurrent code is trying to distribute the load among datanodes +Thanks Hans for report Dukian for patchProperly fixed at r +Lets create a ticket to do this right for but this patch is definitely a minimally inasive improvement on the existing status quo so for. +The log message was extended to cover all possible scenarios. + +Marking bugs with invalid fixed for release as fixed in. +There also needs to be an interface to determine how many of the maps and reduces failed +Thanks KristianCommitted revision . +Its also possible that this issue has already been addressed in subsequent versions +commit it. +Lets not put in another configuration like in MRV +Bulk close of resolved issues of +Merged into the master branch. +This is as designed by default the message consumers producers lifecycle is bound the the Sessions lifecycle as some JMS implementation reserve limited resources for these objects +If the number of threads goes over the pool size then we could log a warning message that asks the user to switch to a pool +This is about server GROUPS and is about SERVERS +I finally decided to show the pages dropdown only if needed done in trunk at r and at r +This affects as well as +this feature has been implemented in Fuse ESB a long time ago as part of +Patch lgtm +In fact ifyoure using Resin why are you also using Commons +This is the behavior of +something subtle +Ivan any more details for me +Still think this is useful but we need a fix for the example +Thats too bad I checked the sandbox example and there it works as expectedMaybe you are able to kick in a debugger and see if the system entersand later if will be you use the update tag on your pageIf so do it point to a property which holds a listmapFor some reason JSF cant determine the type of the property +Please verify +Its too soon to tell whether a branch would be useful +Close this JIRA based on janels posting. +Ill fix this for +Will I even commented on this one +I guess is pretty stupid +I fixed a bug that I had introduced which skipped log recovery on startup for all leaders instead of only sub shard leaders +It should also be noted that any object with either a finalizer or a phantom reference to it cannot be cleaned up in the eden by the hotspot generational collector and is forced to proceed to the next generation where gc is significantly more expensive +Updated patch to add units to min and add units to fair +For SerializerDeserializer we should always open but never close +Attaching proposed fix for review +Hi NarayanThank you for the patch +I have no time to do this. +Apologies that it has taken so long +I added a parameter +But the setup is not with me the basic zone apply method will be called to apply the LB rules polices are getting missed while packing them into before sending to LB device to executeWill fix it and send it for review +It has been Released. +Right away regions were assigned out +Yes Im proposing solely a rename not an API change +shrug +Im sorry for not having noticed this earlier but I think I need to reopen this issue +But if based on those conditions it is decided that the class itself is not cacheable then other contexts where instances of the class is used properly will not be cached either +Have run this on every region in one of my clusters a few times in the past two days without a problem +Thanks +Thank you Andrew! +Hard to break whatI just meant in general I agree w the build failed principal but some of these development toolstests are really useful so Im inclined to want to add them even if it makes things marginally harder for usAnd yes I agree we should do the whole jar +bq +what does your last sentence mean +Some code cleanup planned after the release. +This occurs with no CLASSPATH set and only uses the libraries built into the JRE +defer to +Thanks Tyler +Im not going to change the assignee though until I know I will get it completed. +To repro ant from controls had been updated and passing. +The point is that the marshalling layer should be able work with elementunqualified which is the cannot constraint our implementation to use qualified because then it wont interop with any rolled back your changes to the test case +I also assume that the system is running Derby network server configuration when the problem appears +adds testcase +Hmmm +Waiting for JBT to verify with and close this Jira +Hi Jacques Ashish did test it on his machine and it was working fine but not with a commit to release +I realize thats a little silly for materialized views but we share this code with views +Closed. +Therefore forcing all of the metadata into a format just to avoid quoting in one format is extremely premature +Now I am trying to install the features +Current code does not the patch does +Add It went +Feel free to assign it to me unless someone is already working on it +Thanks Jarcec! +Thanks. +Would it possibly make sense to also allow inserting into Named Windows trough FAF queries +HiThank you for your quick answer +The build I named earlier is the one that should have it it is currently in tag but I do not control the tags so I guess it is possible that it did not get tagged into the ER release +Stack If you need help with this patch please let me know because I can make the time to work on it +It also doesnt change my view that credentials must be protected by SSL +Latest version of attached +The intail reason I had for this enhancment was to allow the user to see what was causing the problem +I just committed this +Could you attach a screenshot +This isnt common at allI assume you mean uncommon and Im not sure which youre referring to +Ill file a separate JIRA to address this as well +I see increases in code coverage in both and with this patch + for the to trunk. +Change the MXMLC compiler +Overall this looks great +However this is more a feature request than a bugIf possible the source of the PDF documents should be corrected to produce specification conform PDF docs +The build script on his get project doesnt do this repackaging automatically +I have committed the fix both to the trunk and +Verified by Richard. +Quite close i think +Seems to work pretty good +Even if the timeout is sec youd probably want to know if your IO is taking more than s of ms really so sec should be conservative +How do we handle settings that override each other as in two files specifying different values for the same setting? Last one loaded wins +So it seems for we could fix this problemWhats the fixworkaround though? Solr currently assumes a disk resident index and the patch here makes up a directory +Did you ask the freemarker guys? this looks like a freemarker problem to me +Needed a little fixup filename is diferent for the qos function and wouldnt apply in +Your opinion +Ive checked this into trunk +Think its safe to merge +See Ive introduced some fixes into the pessimistic lock interceptor to test the validity of a transaction before acquiring locks on it etc +SVN trunk you could simple run svn diff after you have made your modifications and attach the patch produced to this Jira issue +Extracted Tika based detector and updated the way mime type detectors are loaded and initialized +In the Rev +Can you post the full stack trace please or at least the parts that immediately proceed the error which include classes +Reopen if you feel otherwise +Is this a duplicate of +Greg is modifying the EMS library to help with this problem +This patch fixes all of the mentioned shortcomings and removes from example as well as from Velocity templates +Clearly it is important to remain backwards compatible +Folder plugins is added to images screenshots in the guide are scale for screenshots pdf version is set +tlog is not hooked up to rollback +PrashanthI am ready to commit your patch +Fixed the local writes case with commit cffaeecbaeceaae to trunk +It is from using the command +It will fail if the fix is not applied +Ive checked the NOTICE file generated for the jetty bundle and all the components listed there are only the jars that is included in the distribution +It may be better to raise another JIRA issue for the WSDL bitsAnyway the patch is quite difficult to analyze because the whole file has been reformated so it becomes very difficult to see where the real changes areWould it be possible to provide a cleaner patch and attach it to a newly created issue +Hi RobIs there any chance you can give me heads up on when you may have time to take a look at this? Thanks +Patch attached +Want to make a patch Bing + +This issue needs to be wont fix +TC should be a usage problem see my solution in below we could explicitly set the to APPLICATIONXMLTYPE to fix the problem +This could be a flaw in how were handling the rescue of Java Exceptions +v works with git apply +The problem is with inaccurate targetting of the source element from its visual representation +but it means it normally works +Ashutosh can you commit it to as well +At least Ii works on my computer +In WF admin users are not all going to have superuser privileges +I agree that we can revisit this bug if that solution proves to be inadequate +Will you please backported this to seam +Why are the calls not going out in a rational manner as they do in hadoop and showing up in the job tracker HTML pages +Andrew Ive added it to the schedule +The code base does not have an owner +Once a hardupgrade has happenned then the new backing index for unique constraint over null columns will be enabled +It converts the datetime to the seam timezone which is the current users local timezone +Can you remove your projectplugins dir and see if that does it? For some reason the patch doesnt seem to delete that deleting the projectplugins directory +Is there still a bug open here? What exactly is it +Fixed in SVN thanks for the patchBTW I changed the property name to runtimeSchema versus runtime to keep consistency +This change broke a number of unit tests in namely about that +In many of the tests we need to sequentially shutdown datanodes in a cluster +Fixed in latest SVN +Where are you pasting from? You must be pasting from a Java editor not a text editor +If this is insufficient it can be reopened after isreleased +I therefore added a method to Util that takes a List of filenames and returns a list of corresponding hadoop Pathes for those which are indeed files and for those which are directories it returns Pathes to all the files inside of them +I will upload a patch soonI guess the advantage of using the same file is well less files to manage and perhaps a bit ofI cant think of more uses for besides debugging +Hoping to come round to this in due course +eepSo we might need to draft up a nice effort towards writing a bunch once we get this incorporatedI guess we can do this include the code first but dont change any implementation to use Colt +Ill hold committing this patch for a bit then +Thanks Juergen for updating the example and for the detail explanationmakes sense for me too +This is happening to me too in as well as +Checked in to Web CP branch revision fix version to EAPEWP as scheduled by update which incorporates this fix. +This would break HDFS if HDFS and HCFS exist in the same stack +So its not sure the same JAXB implementation can be loaded all the time +Thanks for the great work Amit +Assigned to Henry Zongaro during the Feb bug triage +Okre even when we are closing the host rsreJD +Buffer size varies and at least a couple of times we should have a read in between of the two rsIve fixed all other nits that you commented onbq +in the layout name and ignore anything that looks like that +This change is needed but is this scenario possibleThis case usually happen with balance and disableTable race +new patches include all changes from code for the patches BJ +Verified with JBDS. +And another note need to add support for the skip hadoop tests system property as well +All tests pass +CoolChris do you want to provide a patch to add MALLOCARENAMAX automatically to containers? We can do this in the NM +JBPAPPGACP switched to SP CP +Test suite passed under for the review NicolasPatch v addresses your comments +Conflicting patch incorporating reviews +The renaming of to stopped the cron job that builds the binaries for the nightly tests +The infinite loop could be due to Synapse ignoring exceptions and trying to keep itself alive +Please do the same for quality profiles +Thanks Elsif for creating this JIRA +close region out on regionserver remove from +Verified in JBT Linux +Very happy thank you +The patch attached to this issue appears to be incomplete since the class is not included +I think I am hitting this bug +Yes Ruth has plenty of ideasD +Hello Anton and TimI cannot reproduce this issue on revision r +Also all unittests pass since a new snapshot of jUDDI has been deployed which includesJarek do we need to open an jira in jUDDI on the Concept name issue +Oh I forgot to say very good catch +Hmm +Ive updated the udfbitmapand and udfbitmapor tests so that they would have detected the bug in the old code +Fix is to use iterator for the modification +Attachment has been added with description patch for making as a that it is not enough not to use it as a Bean but it would be great tobe able to drag and drop the component on the +Now if you create a request to add the cookie and read the cookie from the response then on a second request that cookie should still be there if its makes sense for this caseRegards testing a form where Cookies are required the flow is REQUEST RESPONSE REQUEST RESPONSE REQUEST and so on +The problem is the startstop code was blindly copied from ubuntus version of service script on centoss +Is there a maven model for war files too? I understand maven was primarily conceived as a way of handling jar dependencies which may make it less suitable for dealing with ACF than wed hoped +I havent had a chance to check out the whole patch in context +I have already created some enumeration into but it is still not been usedThere are still a lot of things that has been not implemented or not working so contribution from other people are welcomeRemoving the hardcoded value into at line is something still to be doneIm thinking to add a combo box after logged in to choose the pos but I didnt get the time to do itThanks Jacqueswhen I developed this application I taken some screen from ecommerce component but I decide to not share label between webpos and the ecommerce and pos components to avoid what its happened between ordermgr and ecommerce applicationAnd anyway it seems to me that pos application does not seen the labels of the others componentsActually webpos is an independent component and it is living aloneThanksMarco +tinderbox completed without errors +The unit test passes with this change and I have successfully applied the workaround to my project code as well +in large apps validation is reusable a service can be accessed from a web ui or a remote call +Agreed +proposed proposed change fetches K nodes at once +Suggestedr +The patch is created against revision If you would like to change something let me know I will see what I can do about itIt would be great if someone can test this on other platforms and versions of Python +Ignore that part of the bug thats actually working as intendedThe bug is that if you have an image tag or a script tag that references an external site thats not whitelisted it should return an empty document +Well it was a bit different problem so my troubles are not related to this issue +It just comes from nowhere +Changed class to an enumRemoved problem with setting parenthesis and adding tablenames when more than one join is contained in a qualifiers +So same commentsPatch submitted +separate issue not for I think we should make completely final +Jon you are right in that your archive does contain everything to execute the test +This patch adds a method to for silently terminating replication monitor even when namesystem is up and running +OK +It also addresses the major slowdown issues with creatingreleasing filesystem objects per operation + the screenshot show it as grey but its actually pulsating blue +Octavian thanks for reporting this +Checking it right away +I have sent a reminder to asf t because the DOAP file has not been processed yet +Thanks Vasil +The problem is the way generics are implemented + +Clint did you commit this in to the trunk +The whole idea of having seperate jars I thought was to run them without being dependent on each other +It would be good if test cases for these paths were also generated +I tried to reproduce this by running the test times but it did not fail +ok + there is many public generated classes like etc Is there any way to do thoses classes package protected tooAs far as I can see they are all hardcoded as public +The fix proposed in the forum messages DO NOT work for me +This is important if the factorys constructor needed any dependencies it wanted to pass to instances of whatever it wants to create +but think we should omit the deregister for versions that have the fix where the deregister should happen automatically +Having said that I have no idea whether Maven uses a sane XML parser +Ran the unit tests with the patch and no issues found +So punting it for the future would be quite a distant future indeedGiven this is an internal API there shouldnt be any resistance to changing it in the future +zip file still reproduces the to review the fix and the IT I have the same problem in both sample projects provided with I had the problem in +i dont think their is anything like passing known exceptions to valgrind + fails to load when using and works with Maybe it is not related to +Try a more recent version +Anyone else tried their luck? +It should be zekzekus or zekzekus! My wiki username is +Changed from implementing Writable and Comparable to implement +Alternate patch that sticks the theme aware template loader into the loader stack in the instead of the +As we are approaching a production release with Portlet Bridge Beta announced already I wonder if this issue can be closed +I found it really good and very similar to other versions +It aggregates all of the previous fixes +It needs a plugin then extends some extension point +Can closest then be added to +After reading this more carefully the problem is on the sidejarWe need to update the dependency to be or higher to close this issue +Hey Gav so what you are saying is that this should work then? Seems like Mesos RB activity has picked up so maybe its working again +that we make it clear that whats provided is just an exampleYou can make this clear in other ways the folder name is a blunt for this +Ive added a strategy +So we cant know what the original indentation of the first line wasI could still implement this idea but it will only work when the user selects code from the start of a line +Released on Mar to fix some problems with rsync +Patch to update the Exception for the patch Drew +As John Kaputin has committed the latest patch described above I believe this JIRA can now be closed +pass +Each node would still require a small file to tell it wherewhich resource to locateYes and that makes me think that something is not right here in order obtain configuration I need to specify the location of the configuration server and still distribute this configuration file to nodes +Even if this new impl didnt exactly match the results of the old Id have no problem just deprecating the old and saying we are starting over with a real fuzzyquery the old one goes away on the next major anyone dying to stick with it can just pull the query into their own codeI think if we keep the prefix it should be for a good reason in itself rather than just back compat with the old crappy versioneditNot that I cant see people wanting just the tail end to be fuzzy in that case fine do your dirty work I just dont know how much of a real req that is compared to a prefix query +I am going to commit this one but I am not sure what needs to be updated besides and +The problem is that caches its log message name +Ill let Jeff comment on alternative options +Ive tried reproducing this against the exampleThe example has row itemsrows works finerows works fineReopen this issue when theres a reproducable test case against the latest snapshot. +Thanks for the review and trials guys. +It solves the problem both forold clientnew server and vice versa as far as I can seeJust a minor code nit Magic number in declaration of diagnosticLevelshould used +Id rather see that than not be able to deploy a datasourceThoughts +It is just a bigger task +trunk +Here are a couple of more things I have noticed +yeah ive fixed it one way and just havent gone back to do it the way you suggested +Thanks Oliver! Youre a starDo you know when the next release is planned for? problem Jeff +I looked at the kitchensink error first +Dup of! I mean dup of. +I think that the problem is solved +Sorry I wasnt watching this before so I missed your work til you mentioned it on the JIRA +Without details we cant do anythign about this +Enabling the patch again since unit tests now pass +However if you add both a connection factory and a destination into the same plan and deploy it both the connection factory and destination will remain under global context even after uninstall +Coulda sworn I deployed itYou should be able to deploy it yourself just checkout jnr run mvn deploy and it should push everything to codehaus maven repoIf it doesnt work Ill retry deploying when I get home +Shall we first check if tar command is available and then execute it? If not available then invoke untar task only +This is because discovery tests are disabled afaik +This is fixed in thus closing too. +Following tasks were completed Build and package Apache Stratos components +Could you let me know how you encountered this issue? I am unable to reproduce this since the forbidden message is rendered from the java code as an exception which comes only when you trigger the migrate API call for which you need to have an extra host in the pool to which you can migrate and this would happen only when you click on OK button there +Logged In YES useridUsing your test WAR I am not able to replicate the is it that I stumbled across this in my code? Could youprovide the source to the WAR and I will compare it to mycode to see what the differences are +Also tried adding a to the label provider but no clean way to make that show up consistently +The update processor has primarily been for documents not whole indexes +Am I missing the point +The patch includes a test which fails under current build and passes with the patch to +Refactor the Party permission to follow up new pattern +If all goes well the release is hours awayPlease join us on the mailing list and cast your vote if you have the possibility to test the release +thanks +Also I would suggest a regular name for the test case file something like or something shorter +Resolved at by applying Yangs because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Jack! +If done on a container level groups protocols and extensions wont profit of this is that a correct assumption +rNext step is to look into whether can be conditionalized to use different fonts for different parts of the tree +Sid can you please take this up? Thanks +This appears to be a duplicate of bug but I am unable to confirm that without a test case +NNBs hardware is getting retired so it is removed and a new node NNC brought up +Committed to subversion as revision Bernd if you get a chance please let us know if this resolvesthe problem in your environmentThanks much to Knut Anders and Satheesh for the reviews and suggestions +Thanks Sam +Lowered Priority to Minor. +Probably a counter in the readwriter that determines when it should fail needs to be reset +A long test against verified cleanCan you give more details about the test +This time with +You just need to add the updated release to the dependencies of the anyone have a patch that will still works with +Same here +It is great that Jakob has reviewed it. + +Also added a test case for multigraph vertex classes +Manually tested! +Ill hopefully have a fix today or tomorrow +Updated enhanceddrlvmtrunkvmvmcoresrcjvmti to build +the same two actions are in editor context no patch yet +If you use JSF beans Im sure they are stored inside session properly +As it is now all the blobs are spooled to the local file systems but deleted only when I stop tomcat +Attached is a patchPatch definitely not acceptedWhy does our code need to use reflection to get working sets info +The discussion from is related in this regardFrancis started with storing group information on hdfs +Marking as close of issues +It is really confusing when mixing rename with the real logical changesSecondly do we really need to do the rename? Whats wrong to call it storageID? An ID could be a UUID +Pick a type for a NULL literal in CTAS +It is shown correctly in the mapViewer example with the new widget. +Yes in my patch there will be a class that is initialized from an just likethe one you are currently using +Jeff CoWhat is the status of this as far as getting it into the repositorySomething to do right after is fully releasedIts still in sync with codebase and works right Jeff +The parent does not have access to the properties defined in the childs POM +ex org orgfoo orgfoobar filegenerated matches all files whose one parent directory is generated +That is an issue of Seam really not +We still need to monitor that system table and we dont currently have any indication that a node was repaired +Harsh any update on this +I have uploaded the patch for this issue +OK +Something really odd is going on here +yes in english it means missing left you dont have to list everyone that has issues with the DML +Now included only in xulrunner feature which is not installed by default to let installing richfaces feature even on unsupported t do more for this on M + untested and unused nuke it +comments in Namits comments +SVN does not show moved files in diff so the first patch is broken because it only deletes the bundle finder classes +This would actually be quite nice to have +OK If the dependency would be updated to the code of the project will break without update to the new classes +Lars Ive committed your versionThanks guys! +The prefix is a config key +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +As it has been discussed in the current design is to add permissions on directories only which would actually not allow this kind of restriction directlyWe could achieve this by inserting to allow any access restriction scheme or else well need to redesign to provide permissions +Bulk Close for all issues before incubation +Lazy consensus seems to me to be on the PRC is bothered to comment +Perhaps +The broader issue here is how customizations are done +My intuition is to make this two separate operations truncate and +not worth the rippling work it would cause in all the testing infrastructure +Hi Sadeep You have attached the patch to fixing the bug in generated +By now the Karaf project is well established in its own TLP and our website has been completed refactored to remove all links of Kernel. +Does it need to be done nowI think that the system property pollution is making it difficult to breakthe utility method out on its own so that it can be easily reusedIn general setting up the security artifacts should be as close to thecontributors as possibleThere really shouldnt be consumer code adding credentials that should bepart of the contributor workConsumers should really only access them and use themThe fact that we are going to remove that short term workaround makes itless critical to fix at the moment but I want to express what the usagepattern should be for +The goal is to test the as part of +Move the fix to +Taking over the issue going to provide a patch for this problem +Also realize that these libraries DOES NOT require openssl to be available +Yes this only improves log message +However if you submit one that is it will be fine +To fix in later release +Arsen this is trivial with +bq +For some reason I had it in my head that maven would sweep each phase across all projects before proceeding to the next phase buts its clear that the build system cannot work like this when dependencies are catered for +Something like +Adding productservicessoapsrcmainjavaorgjbosssoaesbactionssoapproxyAdding productservicessoapsrcmainjavaorgjbosssoaesbactionssoapproxy productservicessoapsrcmainjavaorgjbosssoaesbactionssoapproxy productservicessoapsrcmainjavaorgjbosssoaesbactionssoapproxy productservicessoapsrcmainjavaorgjbosssoaesbactionssoapproxy file data +Thanks for pointing that out +It was like a scavenger hunt +I believe that nobody is too adversely affected by the extra It will certainly not make the code less complex but we have reached a complexitiy level where this increment in complexity is almost neglectible +FromThanks for the tip to run Process Explorer! That was right I found that there were two one of them was from Intel SDK +Install JBDS without EAP +Thanks for all your must confess that I have not found the time to find out why the unit testing of the authenticator method does not work when I build my application with maven as we are currently in the development process and the deadline is coming feel free to close this issue +Moving massive stat calls out of the critical section is good +Elegantly supportingCollection List and Set properties will take some careful work and issomewhat beyond the scope of a simple bug fix +the plan is to get something out within the next couple months +Anyone have any thoughts +Thansk Pradeep for contributing +Thanks Jaydeep! +Ive wanted this since PigIs anyone talented in this manner or do we need to pass a hat around and commission something +The patch looks good +Still to do move the unit tests over add back the APT classes once their depencency is added to the unit tests were moved to the core test dirThe apt related classes were not commonly used so they were removedSince xwork depends on Java they shouldnt be required eitherClosing now. +Seam context variables are not picked up by Seam Tools model in this test project +Verified with Seam from EAP on +Your issues would fit in there +I second Victors comment +Polished code +maybe the nbantext artifact should be moved as well +The plugin unit tests launches an instance of eclipse and then runinside the eclipse instance +thanks for reporting +Probably open an issue on github for tracking +This is generally a good design choiceOrtwin Glck +Sync has been performedI performed the scalatedeploy. +To be evaluated if it it is doable and easy to understand for final users +My initial guess was that thrift would chunk the Mutation down to fit within the frame +The in released and do the cloning so it was kind of our recommended implementation +Looks great for me +Denny heres another one to look into +Will be solved by +The one test fails as the BCEL rewriter does not support synchronization +Andreasthank you for the reply +It has to update the connection state object and then trigger responseReady even which in its turn can safely submit the response on the IO dispatch threadAt the moment I would like to minimize the extent of changes to +Attribute java +bq +Please ask this question on the Struts user mailing list not in the bugtrackerTheres a page devoted to the performance tuning in the Guides section of the Confluence wiki +Anyone object to adding get to? It appears to be a good idea +Laura is now in charge of this issue lbailey present the system wont allow any bugs to be assigned to her which is why i am making this special to Laura as she is the new lead for EAP docs +Will commit if it passes since hard to verify in other than test runs +In my opinion it seems like it would be good to pursue sureshs suggestion since then we wont need to add a extra class +This contains test cases in a separate Eclipse project +Thanks for the spec FrancoisIn addition to the comments made by Knut and Rick I think it would be interesting to know the latency of log shipping something like Log instant X was sent to the slave ms after it was written to disk on the master +variant of previous patch +I havent been able to find a solution to this +I believe that need to escape is documented as part of parameter substitution but we could make this more clear when talking about I was wrongI was thinking about the use in Java programs where you use the dot +Thanks for the patch always welcome and especially with test cases! Ive applied this. +Also has a fix for the number of reducers for order by when parallel is not used in the script +UI support is merged for this task +In this patch I have update Hadoop to declare the required roles unfortunately they fail I think we also need to update the scripts +I believe this requires an upgrade to? VFS has a dependency on Jackrabbit for Webdav and Jackrabbit still uses httpclient +Cron disabled +Does this work with Jbossweb +If in the future someone would like to tackle this feel free to open is and attach a patch. +The problem is to serverdefaultconfprops does not reflect the modified username +Sure it does its just annoying for newbies like me thinking were doing something wrong when were not +Committed at subversion revision +if you didAddAdd need to load them at the same order +I believe this issue is resolved with the new file upload handler which is available in nightly builds as of and will be available in +Will make sure the documentation is updated in the next few is still no documentation for sdb component I will add updated docs good Bilgin! Let me know if you need any help with itAnd sorry for the bigger refactoring +Thanks for taking care of this Giri +James set it up +Hi Laurafix in the docs for EAP is in the docs for EAP is been fixed asxmlnsjndiurnjboss but correct version should bexmlnsjndiurnjboss +closing this bug as the code is available now. +the attached file is a fix to the attached file is a patch to the patches in r +It sounds like this is just a configuration issue and does not have anything to do with slow machines +Isnt there a situation with an ant task to figure out the compiler? How does the ant task get compiled +In the past I have problems with all XML +Patch to provide a place to build a proper package from within the source tree +the fun thing is that in JBDS the urls in schemaLocation is highlighted independently where as when snjezana runs the latest svn the whole attribute gets highlighted +FYI In latest snapshot showcase Im not seeing this behavior +Categorizing bugs related primarily to stylesheet elements etc +Id commit a patch that modified the visibility of these methods +Manually tested! +I focussed more onR An INSTANCE MUST return a Internal Server Error HTTP status code if the response envelope is a R says that the status code should be taken as a hint only +It would be good to get some development on the branch +Also the process of creating secondary indexes needs to use the same merge policy so that it arrives at segments with exactly the same count and same sequence of docIds +The issue with named Apis like this is that you end up with ugly code likeIf option then Call apiElse Call apiDont ignoreCaseSo do we provide both flavors? Posted from Bugbox for enum +Updated images zip file the last one was missing a file +Ive rewritten +If we want to limit the amount of memory used in the call queue we can keep a counter tracking the total size of the unmarshaled requests +Closing upon as this is due to underlying problem in the IBM provided tasks. +This is because internally we dont use any other form than the OID form +Commit at subversion revision +When I tried it on a standalone Java app this problem does not exists +Reopening to move to as was issues after move to in +has gone through many upgrades large and small +Just give me some hours +Yes indeed I wasnt trying with no arguments +After check the serviceName on content is also use on order component on service +Seems like I didnt choose the right component correctly and I cant see how to edit the component in JIRA? Can somebody with more rights assign this to the please +Without the patch for there is no warning issued though even in this case +You are free to give your opinion on the matter of course the marked as a duplicate was just a way to say that the discussion probably belongs to to avoid dispersion. +Trunk user guide updated revision +to both branches as well. +They were all f versions +Should we make it a user payload and then convert payload to conf inside the? +AFAICT this has been committed to SVN after the release was done so it should be in +But I have seen cases with other load on the test host when both startmaster and failover can take timeIt would be good to see if this change do make the test more stable +And of course it only happens sometimes and never in the debugger +allows you to say I specifically dont want a container on this node +We dont really support using the commands by hand. +Tell me if this works. +I also removed as per your reco +on to trunk and branch +There is probably the same bug with +Proposing to contentSeveral fixes have been done on the to improve compatibility with IEPlease verify if still issues with IE otherwise suggest to closeNote we dont target compatibility with other js libraries such as YUI +I thought only these +That stopped being viable due to different projects using different repositories and things +Eclipse IDE metadata changed in relation to this. +If we can cache it it may be ok +I must agree with AbhiThe message bean in his sample is NOT annotated with +I do not have any resources available for gdata at the moment thats also the reason why there is no development at the moment +bulk close of obsolete debugger issues +I dont see how its possible for DFOS to know when it is safe to delete the temporary fileSeems to me that has to be the responsibility of the caller +Not the container +Sample code + +Minor mistake in the two links to the Archiva useradmin guides when the patch was adjusted from the docs to the site module +properties files in CR is +It would be possible to extend the reader to recognize an ORC file and to have it delegate to the ORC File readerit will be great to have this support +Tom I think this should be working in if the two operations are done in a CLI batch +Pavel Gregory thank you very much! The patch is fine closing the issue. +oh yeah sorry forgot I left that typo in my commit +is it possibleYes you can dynamically change an update processor I see +This is a bit like arguing we should capture invalidly specified provider specific configuration options +Removing the author tags without the consent of the authors is in my humble opinion a questionable action at best +Moving to the next milestone +Minor typo +You may want to ad a dir at line +lite version of test project for error reproduction +Thanks for the quick review Jon +is there a function which is using viewProfile based on username +The DTD and style sheet should be changed to handle more than one ke generator of the same type +I get still the error +If a bootstrapped node streams such stale data from an existing replica the bootstrap operation does not change the staleness! You had stale data and you still do +Since I dont know how to do it as a test case you will have to manually switch the default codec in conf from none to html to see the effect in the edit view +Fixed in trunk +If you have something just open a JIRA file a patch and assign it to me and Ill check it outThanks for the contribution! +Added files to run AMQ with Java Service Wrapper on can now be run with Java Service Wrapper on win linux and +Yes this is a dup of . +Distributed log splitting not very familiar with +Given that is used in it is a definite option +Just remove the HADOOP environment variables and run the app +I may not be able to tackle this for release marking it for future. +We then ask for a compaction and wait for it +Not sure if that would be your case thoughCannot reproduce and request for testcase unanswered for over months. + Bug has been marked as a duplicate of this bug +Values are chunked into k compressed blocks and only those overlapping with the specified range would need to be decompressed and processed by a task +So the need is to know the conditions under which a user created by an admin can be deleted +Or rethink the configuration time of your Couldnt you configure it right on initialization possibly via a? This will work for any properties source which defines values for your placeholders +Moved to upon release of +Thanks Yongqiang +Hi Paulex The fix looks fine thanksBest regards by Andrew. +It is actually seems to be an independent action +They are executed by exec Ant taskwhich does not propagate the failure status code to the adaptor +Will try to reproduce first and provide fix ASAP +This looks great but Im having trouble patching it into trunk +Fixed. +Bruce can I please get your opinions? Ill be able to dedicate abouthalf a day tomorrow or Wednesday morning and thats about the last slot beforebuilding RC +Its always possible something does break and we will provide a patch as soon as is possible +Is it possible to send the xslt that you are using to transform the message so that I can have a look again +I did not change the since there is no existing having the space +I believe the attached patch should work +After the Metamer fix in Ill test changes on Metamer and close this bug +I meant All I want to do is flatten the tuple project out two of its fields and leave every other field alone +the proposed patch supresses the exception and just logs a warning +Does that make sense? If you are confident that it is not being used anywhere then we could consider changing it into a private variable +Why the removal of the note at the top? All of our extension pages have a similar note +Damien can you review this patch when you have a chance +Is that the exact code? And where is the located? Same jar as the MDB? Can you attach a reproducible application +Which one would that be? +Im going to release the latest into test and time will tell if it occurs again in our test environmentFor the moment i suggest you close this and if it occurs again with the latest i shall let you knowThanks for the support Sanne to you for all the insight its good to have more careful testig +we also port this to +is a good option as it supports multiple transport system but server side will be bit complex in that case so its better to go with native websocket implementation first then migrate to the for multiple transport support +Improper test config causes previous reported failures so close this one +I think this should be default behavior +Hi Claus other committer colleaguesI have added a patch for this particular issue based on the Nabble description and the description for the issue aboveCan you please take a look and let me know what you think +What is the existing format and which are the changes discussedI think it would be no great hardship to change scripts for running nightly tests but Im missing some info in this bug +after release. +someField in many placesit make Groovy very different from such a feature is not do you think abut this +Too large to upload as well it will be on github til I figure out how to make it smaller +Ah yes that is a good point but then this will still require that the user includes a Processor that will extract data from the message and apply it as a Property in the Exchange +patch applied in rev +I can remove the samplespetclinic changes it you wish +use +Thanks for the review +implemented in rthank you Chatree for the patch and Scott and Adriam for the comments +I am having to try to work on the now but I am sure I will miss something else up +Ill investigate Monday. +Fixed at the merging of all functionality. +Ive started a download of once I get it installed in a instance I will try and get things building and check the unit tests on that OS +You have an addendum then Andrew Purtell? This is totally opaqueWorking on it +Hope they are okay this time +Just a tracker Jira for EAP discussion see upstream Jira +Both logos could be extented for additional use casesMy preference would be to go for the left hand side version because it a has a more modern look b has the Apache name attached to it c works better without the text d is easier to extendLooking forward to your thoughtsBR logo is nice but my first TM in mind was dropbox not pdfboxthe right one is more like the existing one an i dont know any other TM with a similar icon +I havent tried the trunk version yet but I dont see any relevant changes in svn so figured I should ask here first +ISDerby derby jvm exe start ClasspathCworklibCworklib JvmCjrebinserver result of sc command I came to think that some of program should be included in derby to make it possible to run as serviceConsidering license issue procrun may be appropriate for that because it also is under apache software license +Bug is fixed at + +Against svn +never mind +I will look into this +Wont Fix as the user will see critical alerts anyway. +Luke did you implement this cleanroom in Hypertable? Do you hold the entire copyright thereof? If so would you consider relicensing the code as Apache license for use in HadoopIt would be great to have something like this that is Apache licensed so we can freely distribute it with the rest of Hadoop rather than making people separately compile and install a dependency +Attached bumps the tez version and fixes accordingly +If I can help with further testing or anything just let me know +Patch v change the method name from quickAssign to assignAll failed tests are passedAlso I have done a performance test +test used to path but it was wrong the assertions were checked against the initial data instead of the processed data +This is closely related to as any validation errors will throw a and thus pass through the error framework +I wish we had more data on the subject of how common network errors were +Argh +This approach has been discussed but not documented +Just in case my wording was too confusing +Applied the suggested fix in trunk +I suspect this was a mistake +Sorry to come in late the patch has gone stale +The context prevents an exception while shuting down but it will through when used after the shutdown +I wonder if it would be feasible for one of the nightly runs to put first. +Hey KevinDo you have an example of such a plugin that we could incorporateId be glad to add this I think its very useful +Should have made me wonder when I had to rename the +ve made the flag default to true for and kept it as false in. +Instead attention got diverted to a related but not equal issue the process class loader was not involved in the revert understanding of the problem is +Support for final modifier on fields +The patch seems to have fixed the problem +Committed revision +Thanks! +A lot of MarketingManufacturingWorkeffort labels cleaning +Attached is a patch to that resolves this issue by making the iv optional +The problem was fixed in need to check if reproducible with that patch +Removed CTP from help screen +However its better than nothing +mrepositoryeuiksproject mvn clean install mvn dependencyresolve Uwithout any errors +This should cover for now +Okay I see the node parameter is redundant and error prone +Does it mean that released is not fixed +Changed to future release work from. + +Committed to trunk Target Versions to match Fix Versions. +Ive updated the marketplace info could you verify it works as you expect and then assign back to me +After some digging it seems you are totally right BC doesnt have any support for SSL and your proposed patch is the best again sorry for this +This duplicates which was resolved as Cannot Reproduce Bug by Mark Newton +Anyways lets not guess and allow Michael explain what problem he having +I agree it should be a separate goal however it also needs some way to determine which files to deploy +just make sure still +Hopefully well get a new working snapshot soon +A test on your part would prove very helpful +Filed to followup with dependency resolution issue. +Thanks Sandy. +As far as I remember I had created also a test case for thisI will have a closer look on the changes on MondayFranz +As JDO descriptor generation capabilities wont be available till Castor release the JDO extension integration tests stay commented out +Change to sign requests if either the SP has stated it will sign them or if the IDP has requested that they be signed +Attached better in case Ive added a patch +I guess you could use to do this in HB +Thanks Liyin +The RMI related classes this issue is about will just be moved to contrib but the problem which are you talking about will stay the same +As a result the list go massive and the listeners never gcd +Patch reviewed refactored a little and appliedMany thanks +Would you rather the host be configured by the client itself and remove the dependency on +Then write the site +Please change the title to Convert SASL to use and add lengths for processing +bq +Closing all resolved issues for. +Checked in a fix svn revision dims + Why not move on up to the junit suite as well while we are at it +Changed summary to make the issue more obvious +Thanks Giannis Added and will appear after next update +Thanks PaulexNew types added to LUNI module at repo revision Please check that the patch was applied as you expected + +already! +A few comments on the current patchsrcjavamainorgapachezookeeperL Theres a lot of instanceof used with exceptions +ok Ill add this to AxisBut I think the example you gave for boolean type is about bad programming +I still find the difference puzzling I cant justify your version being x faster +any thoughts +RegardsLars +Note sure about what to do in but may be I can commit this part first? Waiting for a before doing so +will have descriptions of each of these metrics. +two rows of type +DAGverify should catch this +Isnt workspace resolution a meclipse feature and not known in the normal Maven world? Thats how Ive understood so why should I report a bug against the Maven assembly plugin as it does nothing wrongOtherwise Ill gladly do it +So i cleaned it and reattached the patchI have concerns with this fix The performance isnt synchronized hashmap slow? We still dont know what the other colliding threads are doing if not also creating helper contexts? Looking at the code the constructor is supposed to be initializing the map so I wonder whats going on that would result in another path into itFrank +When this is applied PigCheckin and PigCheckin pass +Please close this issue if you are satisfied +Heres a new version +I do not think we should hack to support +Thanks! +perfecti had checked your last patch on and hadnt seen any explicit handling of this situation yet so just wanted to call it Tim +Patch with the updated test case +Thanks for the review +the purpose of NIO is to let a pool of threads service a larger number of IO operations rather one threadoperationIf NFS is the problem it may be the block size used for copies is somehow inefficient with NFS +Im going to fix that +Second aspect to this is the exit status is checked on completion of map or reduce tasksThis isnt clear to me more details please +Upgraded to Hibernate and my test code now Hibernate Annotations does not compile with RC so I had toi take some classes from CVS head and run with these +Fixed in rev +rev trunkee jndi requires a new instance be created on each lookup +We can do major streaming surgery on FST or possibly wait for to get past this +It seems the solution for this will be too big to fit into this release +We will take a look +give it a try +That I did not change and I am not sure what problems that might cause but for this case it worked to our advantageTests are running onKathey + +needs to be closed as a dup of in case the fix for fixes the former +Bulk close for Solr +first? last? splitSo Ill put it this way +The problems related to lookups with all these themes is that a very small changes may have annoying side effects +can match or bis stateful can hold business code +Is there any chance that you can send over a project that has this error? I tried reproducing naively from your comment but cant +Unused getProperty methods are removed from the gml and gml bindings +That would require a fix to the archive file system we should file a separate jira to fix that in the future +! this is exciting as i feel actually motivated to improve these docs +Declaring variables return values or parameters of type is not allowed +Fixed overwrite handling MojoProgram parameter +Thanks +Rescheduled for +Fixed +Thanks for the suggestion! Ive added a corresponding constructor that allows for customization through specifying fastest JIRA fix ever! Thanks looking forward to the release +This completes the work for this JIRA issue. +OK Ill do the fix +Looks fine for me +My apologies for this dragging out this long +For tests which which arent setting it to true assumptions and intent of original test will be changed and thus semantic they are testing will be different +For a router it is not +The session will be closed after the request anyway so its not available +Can build after this patch run tests build jars +So I guess at least for this property configuration outside of doesnt work +I just noticed that on infra someone just asked about importanting a Confluence space and Jeff Turner said to make the dump available on and he would take care of it +Reduced verbosity to FINER. +The cast is fixed in Branch Branch and HEAD but we need a unit test for this +Because running and waiting information can be tested side by sideThe disclaimer which I was testing has been removed as I have changed the disclaimer + to avoid any misusageWhat kind of misuse were you thinking about? Anyone who instantiates the Oak class in any case has full control over the repository so its their problem if they want to misuse it +thanks +Are you looking for oracle style handling where only one such global table is defined but will have session scoped data? Is sufficient to handle the external case +It even breaks any negative trend that might have been there before +for to trunk. +final version after CarstenGood catch its a pity we have no unit tests on theseYour patch is in trunk at revision +so yes Id suggest putting in for + +do you still experience such problems with? if yes are they reproduciblei was under the impression that we had fixed all issues which could potentially lead to repository corruption +which you might think is ok BUT we still have the Failover Mutex at this point so we BLOCK failover occuringPotential Ugly Solution +In class renamed the variables to create and overwrite as well as their corresponding getter methods +But Id just wait and see what happens whensomeone tries to implement Start TLS they will surely figure it out. +Hi RayYou can raise a new JIRA for the issue you mentionedAnd also fix applied at r please verify if it was applied as you expected +DEBUG or even TRACE might be betterI will all my tests with your next version +Fixed checked in trunk branch +Thanks Jody +Thanks +i would strongly recommend not to do this but keep the original node type definitions as they are but just deprecate themwhy? those are internal nodetype definitions and are not used by users +Not enough time +Tom no problem please see and +I think it comes down to dealing with customer problemscomplaints in the short term versus supporting two implementations of the same spec +If you would allow finish wizard without selecting classespackages and add good starting message likeAdd classes and packages or press next to create empty mapping fileThis could be more user friendly than the way I selected +The patch for Guvnor to work with Oryx +I have a few requests On the form page +bq +So I dont understand Brians comment +What does it slow down or degrade? I can only come up with the possibility that it degrades accuracy in iterative algorithms like +I wont have any time in near future for this +and have additional methods in JDK +Perhaps we could create an abstract class that leafs through a users preferences and returns a sorted list of those above the threshold? Then we can use that function in our splitters +As project despot I approve Richard being given write access to the Mapbuilder repository +Sometimes it seems like they are incorrectly removed and nothing is added in its place +Lowering severity due to lack of a detailled issue description +Perhaps but we dont have the mandate to commit on the community version only on the CP branch +Working on a complete rewrite of the plugin using modern Grails approaches and a much more thorough knowledge of Grails inner workings than when I first wrote it +What I dont understand is why do catches when is thrown by loadClass +Shindig uses only JUL +Wish there was an edit option in JIRA +Yes it looks like a bug but distinct from this current Jira +Patch to fix this issue +verified at +Would be great if you could do that +Attachment has been added with description source fo and desired for your bug report +Patch applied. +Emmanuel in your fixes are you also going to pick up the fix to this as well? Should we assign this one to you as well +Because I set +so isnt this just expected behavior +I intend to remove the meta migration code from and add meta migration from but if we remove those parts then wont be supported +This is fixed on and branch +rolled we can just remove start and end from the hashcode calc and perhaps instead base it on index name and type +This makes me want to release soon +Updated patch now tested against hsql mysql oracle and to comment out util as this one caused with postgres +I went by my wishful thinking +Ivan I delete srcwinutilstests directory and its content +At worst it will result in an unstable sessionThe typical usage pattern here is to open a new Session using the same JDBC connection and use that other sessionAll that being said if you want to continue with this pattern for this particular issue try which halts aggressive releasing of connections during a flush such that the same connection is used throughout a flush cycle +However just as the file states you will be responsible not only for new line handling but escaping quotes etc +But yeah put up a patch so fellas can take a look first +We will release after Camel has been released so I set the fix version to the next release +bulk defer to defer to +has been released +Is any the final version +I have reproduced the problem only with button Link with Editor case is fixed +OK new patch Improved to test and miles Implemented flooring logic as described in I think its ready to commit +He is aware of this and said they will help us put that stuff together +We will have two versions in CP +Thanks Eli! +but the most recent patch seems to have now lost log again Can you give it one more try? JIRA wont let me grab the old superceded log patch anymore +Well actually it does work fine now without any further Check that now that the underlying core problem is fixed this use case work +It actually seems like is different +Thanks for the clarification +Full cache is cleared to remove also potential dependencies +The patch is bad since it doesnt take care of multiple part messages +Verified on Tools +I will work on getting you a tests case which reproduces this issueI have tried using the SunOracle javamail instead of geronimo and I ran into a different issue +pl amthe containers will be moved out of the Arquillian build at some point +I guess Ill have to delay this until +The root cause seems to be that watchers map keeps growing +Transitioning to waiting for Infra +applied +If the total limit is violated will kill the task with highest amount of memory to relief the memory pressureSo if we kill a task with highest amount of memory these are more chances that the task will be executed on the same machine +Can you try to turn this into a test case or spec that runs properly on Ruby? Once we have that it should be trivial to fix the rest +Ive attached a patch that fixes this issue will someone please commit this for me +Try this one +Thanks for the reviews and commit Vinod +and yes this seems like a good idea in intend to commit in a day or two +We are seeing the same issue with a large object set +This was totally my mistake +I will commit the fix +Dude read research and think before you write +In a proper debugging situation you would run all the cases and the errors would be corrected in development +I dont think we ever set sessionTimeout less than +Where will this discussion happen +Thanks Anurag! Committed revision +The code looks good +Committed check removals and additions to trunk with revision Thanks a lot! +less than useful +if we model the type to be part of the Row Key then the problem is for or so use case i will be doing a Whats the objection here? still has all the problems of with the added problem that it doesnt parallelize across machines +changed to tests pass cleanly +Correct the way to do it is to register tx synchronization to call dispose after tx is completed it could be done like a dedicated command executed on the session +Looking forward to a resolution of this bugBest believe this should be fixed in the release coming out today +OK +Basic WS Security support is good +The multipart RPC literal issue is fixed +Once the is accepted and I sync up Ill be submitting another patch to bring the changes over into the Tomahawk tld files +Hi TilmanIve got Mapbuilder trunk running over HTTPS with and PHP +Anonymous use should see No issue +Hi allI believe we can simply use a sed an fix this issue +testcase for input +Well as described in we could make validation part of the catalog api but there is still the question of how to we in exception messages +v updates javadocs of the new method and handles more cases in the methods matteo mentionedThis passesmvn clean test DtestTestSnapshotsnapshot true true +The javacode is parent and use now as parent +get this into fix for +the raison dtre of both indexinterval and columnindexsizeinkb is not because we have the notion of rows in the formatIf Im understanding what Ellis is suggesting it is that the entire sstable index could become sparse that would mean that columnindexsizeinkb could be renamed to indexsizeinkb +ThanksColm +If youre interested in getting involved with this effort please let us know +So were talking about the same thing +Annoyingly Im unable to find any information in the error logs about which foreign key constraint is being violated +Stephen if you want to open a fixing this feel free to assign this issue to you +pushed to b as part of provide steps to verify or close this Jira +Of course you can add them manually but for me it sounds like a hack which may lead to problems +Reuploading patch with p this time so it should apply automatically +Merged on branch as of revision +New rev +I guess I might have used the wrong dll to test the other day +I am interested in getting some feedback from those that understand the code a little more about this bug and getting a possible fix generated for it +Very nice +I wanted to just test a fix to if that was right but can contain any span queries so I guess all of them might need clone impls and we may have to clone the whole chainA little tired to think about it at the moment Looks like the issue is with the cloning in though. +Including here just in case +looks good +So this attribute worked as its definition on geronimo latest build +Verified the issue +job file certainly but doesnt it need to be present for the junit tests or any of the utils that work with Sequence filesFWIW in light of the commit for I need to rework this patch +closed shipped in +I know DELIVERSM works fine though +For large numbers of files that will be less efficient than the hashmap but lets face it we arent going to have large numbers of files and anyway the amount we save by not loading the schema files multiple times will more than cover the cost of a few string comparesThat way you can a get away from the hashmap construct that requires special handling on different platforms and b simplify the code change by getting rid of the stuff +Deferred to. +Thanks +Err +is very complex to implement and I suspect a bug was introduced while we split it from +Which creates some overhead for admins and developers +Thanks +I chose BSF over because the later is part of JDK +Searching for all issues in XALANJ takes me s +OlegI can understand not wanting to take on additional code for maintenanceI am going to work with the team to add the scheme to their library as a for http clientOnce we have a stable implementation if you could include a FAQ entry referencing the so that if people wish to use oauth with http client they can find out how to do itIf you ever decide to add oauth support to the core I would be happy at that point in donating the codeThanksPaul +Might be a rat deeper in the JDBC test infrastructure wood pile + fixed the commons stuff will take more time although I dont understand whats the problem with and are usually bundled with the application server and can cause classloader problems if included in the lib folder +It appears that the printed stack trace is because we do capture the and print it out +Dejan it appears that you fixed this is there more to do or can we resolve this one +reinitializeConsumer This method is called every time a new topic is discovered +Why is this not a bug? User enters something that is not parseable to an integer into a form field Iexpect that to be caught by the validator framework instead hereis what happensIn the process of transferring the form data into the actionform mybean property is defined as an integer so the try to convertthe form field to an integer fails and defaults it to zeroZero is valid so it passes my validator which is setupto verify that its an integer +for the review +I think I know what the problem is now +its svn rev was checked in a while ago this should have been resolved +yep you can ignore all attachments +on patch v +The java doc says that the iterator will be fine as long as its only used single threaded +Thank you very much for your suggestions +Thanks for the feedback Stan +bq +Thanks Werner for quick ve attached the whole xml and xsd file and the complete test file +Looks just committed this +There are also several load methods in which should be moved or adapted in this process +Looked at the patch +Please fix this +Sorry for letting this sit for so longThe latest patch looks pretty good to me +Another option is to inject the Cache reference and push it manually +On Thu at John Hawkins wrote Ive always questioned why we have processException in the first place ! Ive never understood why you would want it +Patches look good +Commit message format has been updated on all of the repos +One problem here is that embedded ZK currently expects to start on solrport As far as I remember thats simply a default and you should already be able to run on whatever port you want with zkRun +Already fixed in +This issue still seems to affects my files with Grails +This patch is provisonal and it is not intended to be included on myfaces core +This is not correct +Such proxy does not extend but rather implement +Thanks Tom! +this permits the generated classes to be in the same package as the interface or different packages however its specified in the protocol +Maybe well get rid of excludes for simplicity later +I will post here what they tell us +Note that I incremented all attached files suffixes to rev for consistency +doesnt web container like Tomcat or Jetty so this improvement is not an option. +Thank you for your contribution Miomir! +duplicates it and has not be resolved +Do what you think is best after considering the input put up a patch and lets see +Hello Helmuti think keypress is a better choice instead of keyup +Added new method +Ah now I see the issue in +Username lkral +other things to work on that will surely help +Indeed it looks like this is a symptom of the bug +Just an idea though +Most of the remaining missing source jars appear to be thirdparty Ant builds +So we need? +Attaching list of being closed +I find out that resteasy build in Brew is missing usrsharemavenpoms in rpm +How about we add two new options filelimit n and sizelimit n to distcp? Sounds fine to me +I noticed though that if you have a plugin call the package plugin command generates a file +Thanks for finding and fixingThe patch applied at +Is it fixed? not fixed? Is there any further action needed +Works on Helios and Indigo s working fine for me on m +Does not hurt to do it now +a lot for all your helpchris ll have a thorough look at the class and Springs implementation in the timeframe +It seems to work well and now matches the SVG drawing +I removed the error handling from hadoopuserinfogetgroups. +i have fixed the copying of drools libs into EAR and added a check for existence +ant tests passed on my box +Backport to hadoop would be nice +Can A MTQ also iterate over docvalues? I think we should commit what we currently have first and then address joining by docvalues in a new issue +I admit I was worried about someone catching this +right as it is an object we can simply check if it has been set or not +The timers are created as it should be and they fire as expected +Its not a bugAdded unit test to verify correct behavior in CamelDanatello thanks for the report +Some files seem to still have copyright statements +Hi GuysHave to reopen this one +Sorry about this +Thanks for the detailed information on how to recreate +Closing it. +is done now so startstop operations on the app server should be working therefore we should be able to test patching end to never could have worked because the never got implemented +I will make a separate issue for handling deserialization issuesThanks on trunk +Python XSP had same problem. +Attach a figure of the benchmark +MarcoAre you monitoring the dev mailing list? There are issues with your commits +iTests issue now fixedUpdate of the wiki in progress +I dont usually archive those +Thanks for the reviewHere is a patch that is synched up with trunk +contains a version of the supplychain sample which uses a combination of and POJO implementation types for the components +The more smaller pieces we have the more we can make progress on a single component at a time so to get the tests to verify the approach more frequently rather than having to code for long without feedback +The work is done by the bookie and the sync thread which are the entities interested in marking +HiIve created a for this issue +If there is any enhancement to be done one has to incorporate it into both code bases +Yeah we should give this some thought +Hey Davidpatch applies cleanly but needs to be brought up to the date for the new Vector iterators +Dynamic RTL was on for the projectI also checked all my IDE project options against the Xerces BCB project options +If there is anything in question lets leave it in for now and can work with smaller patches to fixThanks Earwin +Checkin completed +Apply from modulesweb directoryBundles in the lib +Committed revision +UI change are not working +attach quickstart +Committed to hive svn trunk +Can you open a seperate Jira for thisThanks to capture theIt looks like this issue is happening again in release +In addition lets see whether I can find all the corresponding places where things are meant to change after we have completed our refactoring +Looking forward to seeing your patch +Are there similar changes on the web UIWe should also refactor the code so that the same string is not constructed in two different places +Its on my todo list for about weeks now to help this guy getting it to central +Its not fixed yet it will be fixed in version +Work on this can invalid assign from t Fix is going to get removed +Im using Eclipse +Does this database type require a new schema file +Obtain guide from +changed some fixed some rearrange some codesSince the file sizes to readwrite are very small the should not be CPU bounded on tasktracker nodes whether you have e or mappers per tracker +Do you think you could add this backoff mechanism to the SDK load balancing Rpc client also +Bulk close of that have been completed or are now invalid. +I also ran into a couple occasions where the test suite got hung +Need to be able to point at an alternative before can deprecate +I believe the number of duplicates created is proportional to the numbers of servlets added. +My ID and my Skype ID is leducbao +Thanks Hairong. +if you have Server provided libraries it wont be dont understand this issue +to trunk as dbdcfdfedc +This may not be relevant but I got misled by this bug so want to mention that you can support authentication and forms authentication for the same url +This patch makes raid compile but the tests fail like crazy +That is actually the M +Not possible to generalize +hrm committing this to both version may cause trouble on a forward mergeDo you mind to give a once the patch is ready? we can do a forward merge of TRUNK before you commit this particular fix +It should therefore include all such convention properties and methods eg beforeUpdate and so on +Thanks very much! +Easy fix should merge segments with the smallest segment +also fixes an issue which caused serverport pattern to give error in glob expansion +All patches submitted and committed. +However that may change later +I like your approach for dealing with the straight forward and efficientI am assigning this back to you please apply the patchregards Malcolm has been applied +Attached is a maven project with a single class containing a main method +If not then we can see if I should sharpen my axeThe sniplets wee there to explain the setup but you should mimick it in an existing unit testYou can PROBABLY get around this by adding an annotation pointing at the Users class to the service interfaces +Genesis of this Jira from MarshallDUCC has a project which Im guessing has examples for theuser to learn from and build upon +version no longer supported +It is +When we resolved it for one user everyone was fixed +There is a minor norms bug +Ben I dunno what happened +changed to store as that is where the routine being described is located +Have a nice day +Could you please provide a patch on top of the current state of trunk +Upload a patch that changes to contain a field and MR AM and are changed to retry till it becomes true +Initial edit in Git commit bfeaebebaab +This is related to +I have verified it works with now. +Hi PavelThis testcase fails on RI too for me +Patch looks fine +like martin and ivo I have the same issue fixed by specifying pluginshomein webtestconf running your tests via using will avoid any classpath issues. +Thats fine +Or it could be a delayed write thing on windows +verified the fix and it works fineThanks for your quick turn aroundDo you want me to go ahead and close the issueAlso when do you think this fix will be releasedThanksPraveen +Is the plan to keep out all the downstream vendor stuff and replace w bigtopYes +Can you regen the patch against trunk +Not updating would be good for to bring in line with its copy of +you were right one entry for each top level groupId +Well start a release pretty soon +Patch to disable this case for the time beingLooks like the test for tokens wasnt forward ported to trunk so no need to update anything there +Nice addition with removing the classloading all together! +That means the serializers can be more efficient because they need not attempt to verify the incoming dataLook at it from the perspective of an event producer like an XML parser +What is Felix and why should its source be in the Derby repository +Hi Neil the conclusion guys have reached is that we wont fix this jira but a release note instead +It can for sure be moved somewhere else once the system is upgraded +Hiyes it does Thanks for that +Successfully tested on PG +The way that the Thrift server code handles this is to log the exception and close the client connection +Specify as hdfslocalhost in conf +I think there are changes in the C client that should be made as well to make it more resilent to clock adjustments as well +defer all issues to + is missing from RC +Ill commit there and leave others to review the branch +And this patch ensures pulsing support for and For postings format +Unfortunately this happens most in the IDE it takes some time to discover the problem since you trusted in the IDE +I am very sorry I have rechecked this test again and found a bug there +I agreeThe reason for the difference is probably because at the seperation between two normal numbers change due to exponent change +Just Remove unnecessary semicolons +We upgraded to Version and encrypted the database with dataEncryptiontrue +This will keep the behavior similar +Its pretty terrible +Is the fact that not all of the dependencies are exposed at still an issue +Thanks for creating the new option Andrew! Ive reassigned this issue to that release +It looks to me like youve implemented solution from my bug report above +Seems to be working fine +Checked repro attached +usually all nodes in a cluster are configured the same waythis is also a blocker for us as far as taking our application into production +the fix is quite straightforward +POM changes get consideration for +In can so happen that a cluster administrator might actually want to watch a variety of services Hadoop being only one of them +Is this fix relevant to EAP CP as well? Cheers think so +What if end users just want unix format mac format +If we dont tolerate the mismatch it is possible that all region servers in the cluster abortI will try to think of a way to tackle the above scenarioHowever the majority of the problem raised by this JIRA has been solvedI am fine with patch v +Patch that resolves this problem +But what is not clear to me is why is at the moment? In my mind it should redirect to homeofbizbigfiles folder in machine +any thoughts +This is the patch that fixes the problemDFSCURRENTVERSION has been changed to since internal file layouts have changedI created a new package for exceptions +Cos could you please try verifying the whole file instead of the newly hflushed data +Please attach an example the same issue than +Logged In YES useridFixed in recent releases +This looks like its generated from which uses references to idsof +There is so little interest in this patchfunctionality now that I doubt it will get distributed search support in time for I would like to commit Bojans patch for though +launches does not exist +if you are seeing a different scenario to the one outlined please should be considered a bug then because it explicitly depends on SNAPSHOT versions when compiling those are not correctly updated when using timestamped snapshots then it should account for that and use the latest timestamped version in compiletest shouldnt it +Test enrichment possible typo in very first sentence When you use a unit testing framework like or your test case lives in a world all its own +Good call Todd thats definitely the canonical case for multiplexing plus it should show some benefit because I recycle file handles across requests by pooling them rather than opening a new file per request +Bob Buildbot can run any number of tasks in sequence currently I think it doessvn coant ant ant then some other stuff for the nightlyIf the user is running a separate ant task that downloads and executes a library then the user will be aware of it +Resolving as fixed +Ive applied your patch into SVN trunkThanks a lot for the contributionMaarten +Its kinda hard to do this from a contextInstead we count the number of regions that belong to a table on the current regionserver +v +patch is now fixed Andy committed my patch +that properprxim dia may corr +the nd issue the mapping of for informix was not changed because it would affect the cmp subsystem. +This is undesirable for a couple of reasons If you arent testing the binding as part of your test its not necessary to actually initialize it +This may not be obvious +Im afraid it is different case wrtIn AFAIK CONTAINERFINISHED is likely to arrive as early as an is at SCHEDULED while its container is at ALLOCATED where the container may send CONTAINERFINISHED if it is killed or expired +Thanks for all your time and help with this +Multiple occurrence of the same option results a semantic error when script is parsed +I had originally attempted to add this +The method for commit is split in function one is pvt with an extra boolean variable to avoid concurrent modification while committing in a iterator loop +Say you flushed the logs and then ran a compaction and waited for the cluster to chill out +Updated summary and description to indicate this is to do with a standalone based on standalone +Can you please verify the testcase is not related it doesnt seem so just to be sure +Moving this out to for now in prep for the release +Here is the uploaded patch +Would like to get this in to +Daviddo you have a patch that i can apply forthanks +net +There you are getting a connection for your own use +It looks like all the files now have the correct copyright +Then construct a String from it and use it in the test +Terry how were you actually seeing this happen +Applied patch with some mods thank youThere has to be a better class name than lets discuss +redundant now that we have bound inserts via Commands. +Removed Erlang from rules and control so will not be generated + +it would be nice if we didnt have to do this Right it will probably take a while as that test is quite long +Integration test checked in revision . +Not sure if this works with the lifecycle +Was trying to make the smallest change to proceed ahead +Patch looks good +One way we could do this is to add an API method that would allow the connector to set the hostowner which we then store somewhere +I also noticed this issue while working on other jiras +I am working on extending the Chained Superstep API from current state +Converters for other types will have to be registered manually +I dont get this error +can we close it? because its already fixed in trunk +Adding a pluggable layer which permits users select different BSF implementation +Reopen this JIRA for backport to +It is an issue with my patch for actually +Please revise +Is it enough of a limitation that someone actually started working up a patch for this? Umm apparently not +Patch to fix the error +switching off patch available I commetted the latest patch with revision +Basically this is intentional and it also prevents JDT to compile +HansWhy would you want to change from rate to rate? Are we sure that a rate would ever be for a currency? It seems tome that here a rate concept is more general and open +I have also added Spring unit tests and verified that everything is in order including checkstyle checks to ensure the code is style compliantPlease let me know if you have would like any other updates +I just needed to clean and rebuild +Committed the patch +If is applied to an and beans implementing or cannot be serviced properly then an will be thrown +Test manually and it is successful +But I cant manage to push my changes back into github! How can I do it +This is now fixed in re welcome +populate had an error +Committed by Ashish. +Our mirror monitoring site changes on a basis and we dont have the resources to keep track of mirrors that require special actionsFeel free to continue to offer the mirror locally but we wont be listing it in our database. +Let me take a look at the coverage numbers before this patch went in +Applying LB on multiple using List view is same is successful when selecting multiple using individual check boxes. +Instead of the underlying database connection you should have the JCR Session included in a potential distributed transactionResolving this as Wont Fix as the described behaviour is a feature of the current design. +Hi Fangmin Could you upload a design document for this feature please just so that anyone interested can check your proposal +So that you can use directly xmlbeans generated codeI tried to generate the getting an error in codegeneration time +Erm Im not seeing whats wrong in the log +And if i select custom setup of JSF implementation libraries this component libraries appears in my project +if agent is specified and points to a path then the jvm will not we should download the file with a slightly different name +is not a working copyIs there any way that I can build my project without having to check it into our remote SVN repo +Use this one insteadextraneous material purged +does not access any of the +Yes I have shipped him the patch and waiting for his confirmation +Can you confirm the same? Do you suggest us to do the dependency changes to or move toRegardsJana +This is because there is no other way to stop normal logging dumping out the full trace +Passing against +That task in itself is large enough +Ill add the code to unsign the jars in the for both EAP and EWP +Postponing to to unblock work on release +How we deal with these global statics when multiple cleanupdestroy events can be called is a good question +well its fine if you are doing matching on something really short you could index with keywordtokenizer and use this for some use cases +We generalized to to be able to enlarge the index when we load a segment +After unchecking Group items by category the original plugins are visible + +at the moment you must dispatch +Unit Test created in reports this bug has been previously fixed by related issues. +This change will not result in API breakage +Felix could you have a look at this +It is strongly recommended to use only one form on a page and use the ssubForm feature for handling partial and validationregards is a duplicate of which was fixed long time ago. + Username kosulin +We should add note in the Flume user guide +Ive committed it +Would you please check this against as the annotation now has a versionField attribute you should set to if you do not want a version field +AdrianCan you provide patch for? this even an issue in? Looks like all of this is new in trunk +I am not sure packages would build as isthe conf files are already globed byt the statement at line and rpm may complain these files are declared twice +Thanks a lot for the patch +Fixed by incrementing the version number +current trunk does not build due to the lack of not relevant to this hudson testing +Sorry I accidentally left the last commit off the diff +Steve Im talking about hibernateexttools which has much better and broader ant support than what is in corePaul That is the other alternative yes +The would make a good unit test if it were updated to use with no filterPath +I just tried this on another project and I can not reproduce what you describe +Pushing this in +Christian do you still have the issue? Does your project compile if you run mvn install +Would you please try this patchBest patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected. +Stack Actually this is an old bug in branch only + for the patch +I feel that is a good time and that is the implementation Im using however some sinks do not throw it +files +For allmost databases DECIMAL should also map to Hibernated register the datatype at the since this is true of all Oracle setups +No tests since this is trivial and was tested manually +bq +Also I think there are some problems with the current solution for hcommandLink +Thanks for catching that! Heres a new version of the patch that includes the test +From what I understand it makes a formtransform tag superfluous +The patch is not complete +Will commit in days if I dont get no go + +And I wouldnt use a GET to send my usernamepassword over the wire +Example +I tested by removing the explicit dependency from srccontribhdfsproxy and the test still passedGiven this the for the contrib projects can probably be pared down a bit but I agree with Aaron we should probably leave it up to those authors if its not causing problems for the overall build +lets use the proper cocoon somebody explain why property cant be set inThats what parameter is for. +Will commit soon +Maybe there should be a param to indicate which scope to usemvn Dscopetest groovy there is some more work to be done here for this to work properly. +The problem is that we cant just go linking to Java What exactly would be the problems in linking to Java or maybe just Java ? I mean as long as no methodstypes have really been removed from the Sun docs the links will work and will point users to the latest documentation +or +for me +And he was the one against the required functionality by Wicket +Could someone edit the attachment if needed and close this issue +Uploaded new patch to fix the findbugs warnings +extra comma +Some beans are not proxied while everything is fine when using CXFWould need a test case for this +Running each test in a separate JVM solves the problem so its not an issue but in general Id rather avoid the use of static state in a class like thisThe latest patch still has a change to which I think is unnecessary +For blockScanner I will attach a patch that differs the initialization +bq +That work was done on the trunk under +I think updating the docs as you suggest is OK for now Bryan +Hi DIGYThe first patch diverges from the Java code +There is no such thing as a repository wide default node type in JCR +It has already been implemented. +But to me following a consistent pattern is far more important than the number of charactersAlso I wasnt sure whether a and b had become flipped in your original example +Slip out of this ever done +closed for final release. +Reopening +Committed revision . +Yeah this can be pushed back +Looks reasonable to do for the next be fixed can you please confirm? +Put back the support from the old into a shared test base to cut the copypaste previously present +Thanks +Nevertheless the versions in the released CVS zip file appear to have the DOS line delimiters. +EugeneLooks like all related regressions have disappeared so you can close the issue. +Yep it should happen automatically without explicitly calling focus +Merged into master +And thats why I was talking about wrapping Reader in an Attribute not copying its contentsHow to do so is much less obvious +The patch is applied perfectly +Moved of GA as agreed on call +I dont know exactly when but it has now been addressed +Not quite sure but this seems related to if its not an like should be solved by +More detail is now provided. +Is there a way to reproduce the error +Done +I guess Idea can catch parameters that are only used recursivelyNew patch +Then this patch removes all deprecations in indexer nowIll wait for comments about the patch from Uwe first +Closing issue it has been resolved for a long time. +Got sidetrackedSimple patch to only throw out exceptions in DEBUG mode or when really necessary +verified on demo works ok +dimmed icon is there. +Hi DanielI attached a new version of the that solves the previously mentioned issuesThanks! +for In my opinion Boo should pick the delegate that most closely matches the callable signature you pass +If the JIRA import fails Ill revert back to BugzillaCreate and edit will be disabled in Bugzilla after a successful migration +Verified problem solved +uses resource request for push so it should solve the problem +Best regards George +Thanks Jimmy! + Username sflexus +Thanks Tim! The patch is applied as expected +Thanks Jitendra! +Mike here is final also +Committed to trunk +P localTests to still work previous addendum +After few restarts server successfully joins but others in cluster have this errors in log files +Move range offset table to the alt header +This is just some really old test code from when I first wrote +patch committed to be applied to issue as it has been released as part of release +thanks MartijnYes I ported to as contribgrouping so this will be released when we release +So I close this JIRA. +Thanks Iulius! +is an improved version of my previous patch interface replaced be used to interact with both Body and Entity objects Made Disposable interface optional +But I think we should push this back to +To prevent Tycho from running them but running the rest of his unit tests weve added this extra exclude rule to his test pluginThe question is therefore should the test class name +Ive just committed this to and +Currently one must compare your new implementation with the existing somewhat modified implementation without the benefit of diff or any other documentation to see what has actually changedIt appears you wish to augment the canopy code to produce an additional output folder and that kmeans would be able to utilize this folder to optimize its measurements +It might as well be added to io as +Looks like none of the uses in MR are affected so the patch should be good to go +but why search text input is in lower part imo all will expect it in upperthe same thing is for Relationship search +AndyIndeed the weaver jars should not have to be present +I had a quick look at and it didnt look very is it something that could live in instead? No youve got the binary dependencies worked out it may not matter much but I thought Id mention it anyway +does not address things like system that dont grow large enough to flush to disk on their own +Thanks for the patch Li. +Some questions I have is Is there a problem if we kept the triggers even active after file is closed? Isnt rollinterval and idleTimeout concept completely different? In other words once file is rolled over idleTimeout has no meaning left? I am guessing this issue triggers only when rollover occurs before idleTimeoutTrying to understand the conceptsThanks for the detailed notes +Also updated the rtfilter zip readme +It would be case by case +Yahor do you reproduce it with two simple or with large projects +experienced here too with my own be related to can confirm that the example project fails with MPIR can also confirm that the example project works with MPIR +Version Released +Can you give the exact error you get and also the startup log? resolving as Invalid. +tdieslersatellite cygdrivedprojectstestsuitesrcresourceswebservicejbws wscompile cp +Also Spring comes with several enhancements that better support the use of generics in bean property methods +I have this issue with version +Once a version of is released with Java support we can apply this a fix for the surefire forking mode issue as well +Im unsure I understand the reason that you need to retain the filenames even if they come with that name you can choose a more appropriate artifact ID when you are populating the the problem that you need to support a legacy Maven repository +Now click on submit patch on the top to trigger the compilation and tests +This feature is implemented only for java projects +Bulk close for +I downloaded the sourcecode and started to debug the the project should be commandline is parsed well and variables are properly it parse the try to find projectartifactId and it is not able to find checked if the file is properly loaded what is the seems that the xml parser used on mine Mac isnt the same as the one you using +I dont think it is a good idea to keep the pid file under tmp +Fixed the plugin name in patch +I am going ahead creating the patch for my suggestion +the default is the empty list which sends allBut AES does NOT send the empty list so an target will reply with the wrong data +Manually testedJust to be sure of my understanding with this ticket were just supporting any kind of exception propagation mechanism? Propagation was already supported but some use cases not +What I can do though is to make it throw a subclass of that can be retried using a custom retry handlerOleg +We have had versioning in postgis for a long time already. +This JIRA applicable only for +I took a quick look at the patch +The encoding detection mechanism still falls back to the ICUJ code if juniversalchardet wasnt able to determine the character encoding so the risk of regressions should be pretty low. +This is odd +Without a clear statement of how things should work I think what we should do is catch and log anything inside of the stopconfigurestart +Added junit test case +Attachment has been added with description proposed for the patch +Ill try to submit a patch tomorrow +Thanks +there is IO that happens when this is locked +true +There are following questionsWhere can i get information about which directories should be cleaned upHave I choosed correct directoriesShould be included only into main? +NP +Closing all tickets +But Jonathan feel free to reopen the bug +I also would like to introduce an authSupplier for proxy auth and handling of responses so we can support multi phase authentications with a proxy +After talking with Nicholas we decided to split the task of reviewing the patch +This explanation might provide some more context to how it affects MR jobsOut of my curiosity why disable permission in your cluster? I beg most existing libraries such as the MR code assume permission is enabled and do not handle the case that permission is disabled +Hi ChathuriYou can apply the +fix submitted in +I just committed this +Also they should handle with a warningerror since wont build without boost itll just set qpidboostroot to no +Marking these as fixed against rather than against. +Heres a patch which uses Charmonizer to probe for variantsof toutflower +Subentries are now supported by +cloned to track this separately +Merge pull request into master branch. +The report indicates a unix problem and OS X is unless I am mistaken a port of unix +Thanks for hint MR! if GRAILSHOME SET GRAILSHOMEGRAILSHOME rem if GRAILSHOME SET GRAILSHOMEGRAILSHOME +pushed +Hey Darren can you give one of our snapshots a try +By using the callback hook to toggle settings the conditional processing in can be asserted +Closing the bug. +In anycase please try and see whether it that section of the spec saysXML Schema Structures introduces a mechanism for signaling that an element should be accepted as valid when it has no content despite a content type which does not require or even necessarily allow empty content +Attach a preliminary fix +When destPos srcPos we are using reverse copying method while we could use direct method when arrays are different +After network restart router it has the same ipaddress +This bug has been marked as a duplicate of +The cache implementation is not yet complete +Reopen and resolve in correct version +closed +Initial commit done +I added a check for this NPE situation since the JDT API indicates that this object could be null +This can happen on trunk too albeit for a different reason +Not sure theres a specific issue to address this but we need a test of some sort to reproduce so if you can create one then its a good idea to test against the latest snapshot as there are a lot of fixes for other issues in there and one of them might just resolve whatever you are seeing +parallel loading index the way we use the buffer to read index files +writing such entries has been split out asWith svn revision I have added tests that prove the tar package now understands both the starGNUBSD variant as well as the POSIXPAX one +Hi Bill could you please post a comment letting us know whether this is fixable so that we could look for a workaround +At least it needs to be in line with plugin syntax with a system +But somehow thrown in the doAs in submitJob is not throwncaught anymore +Thanks +I will submit a patch after testing +Should be able to give it a try on MondayIf thats just a script issue you can just put it intoWe have to get rid of that now +The recent change to use as the maxprefetch default has caused the consume rate to drop sufficiently for this test to failSetting maxprefetch back to resolves this issue +Fixed in r of branchr of trunkThanks! +I am not checking the Existing Application Impact box as I dont know what the rules are for regressions +bq +These are queries sent at a very high rate to stress test the Lucene server which runs on an cpu system +Patch to reduce default memtable size from of heap to +Ive altered the test setup to use REPLSYNC to avoid this kind of spurious failure +Removed Roo source from in Git ID Roo wrapping source from in Git ID Roo backend services source from in Git ID faeddbabfadd +So I suggest the following solution We make the constructor private and provide a static factory method to acquire an instance +Done will reopen if some problem comes up with the upgrade. +Id like a code review as I am not sure if this is a hack or a proper solution +AFAIK Im not overiding the configuration +Test case with this JIRA number in it +Slightly updated version of Alans patch +Ive committed your patchThank you! +This is a great idea but we also have to take into account that some interceptors to produce their net effect have to alter the entry on its way out the door +Also modify the called method in the class to sleep for an extended period of time to represent a longer running task +The point is that I have no a deep knowledge of how fids should be consistently treated across versions storage strategies try to investigate it further and come back with a reasonable implemented using +From my point of view this is all about adding new custom rules +more details I need do you have a dependency of type war in your project +I know this can be enforced with something like +Any update on the failing testsPS +Do you really think we should fix this now? I mean it is a pretty big and fundamental changeI would say this should be fixed +Modified Result private buffer reallocation to pad to a size equal to the smallest multiple of a configurable constant +This wasnt a bug just a misunderstanding of the use of scmURL in Clearcase +Didnt mean to imply that its not a web service +Just import the package and you will be fine +moving out to CP but continuing to check that process works so fixes if needed can get into jon are now more specific issues. +new +Log file for patch file forI tried both and and am still seeing the problem +The order got inverted when we switched naming from weight to precedence I think +Im not being able to create a new review request and it seems to be due to moving the repository recently +Done +This is probably a duh for most everyone else but it took mea few minutes to see it +Updated patch includes added license to added to simplify creating an index outside the Java API +If you know the solution for this there isnt really need to clone project and you are more then welcome to contribute a patch. +This would be very useful +LGTM committed +This adds a bug fix note for +Hi PaulFor testing these use cases your input is required without that getting stuck +image of the actual installation Mark Should I be removing the second screenshot in section with the Registered Servers option? Have you got a new example of this screenCheers is a sample screenshot for section +Attaching proposed patch +except that I do not particularly like the node in the name but I succumb to the majority +If the broker dies spring does not notify the uima AS service of the failure +has been renamed to +This issue was originally raised by Jirka he created a process containing a timer which was cancelled before firing +Disregard pre patch which got mixed up with the patch currently attached to +I create java project create hib +Hi Erwanjap thats now problem i can simply extend the in js mapping and load the validate language filesIll do that in the next daysCheers +Would like to give some more information the application is thread based +Should lowerBound start off as in maybe if we keep sized segments +I think that is OK +This is a minimal maven project which creates +However after that isnt officially supported by Oracle +open eclipse and import the existing projectIn case checkout is done using subclipse in eclipse one can refresh it after running ant +Documentation change only. +patch for orgspringframeworkcoreconvert for as part of a general overhaul of and +Copying a template is not any solution at all +Closing since is released. +If you cannot reproduce the bug I will close the ticket +Cant it be used with ? after addressing above +Would be great if this could be change in one place +Sounds good to me +Thanks a lot +Check the version and reopen withthe correct version +please look at some of the other examples in the codebase and follow what they do +If you use to view it then you should see two spaces +This looks pretty good in a first pass +Comments +scheduling to go along with RDF addition it appears that properties cannot be used as versions which increases the need for specifying the version in the parent pom for a large project +m +Closed again now properly recorded against Fix Version as well +Updated version of the patch slighly changed to avoid getting the artifact handler twice and code reformatted +This fits nicely as that test has test casesCommitted patch a to trunk with revision +getting closer! I added the xcopy location to my path variable and now Im seeing more friendly errors +Unfortunately we have no good way of automatically installing the library +Works wonderfully now +For example I have created a schedule component which is shown the hour time +So while you can view them in my output above they arent in the patchsvn add your new files before you create the patch for example svn add srcadminguide svn propset svn native srcadminguideIll remove this patch so nobody will attempt to apply it +though theres no jira issue to keep track of it reopening issue +Thank you Mayank. +I think the primary use case that was envisioned for was tocopy from Strings to specific types +header management fixed +looks good +thanks Julien +is reserved for modpython use +We have similar open issues and need to have one more issues for this problem with Seam model loading. +If he doesnt want them Im fine with dropping them +test that conforms smoke test standards +I dont see any errors during startupdeployment in the studio console view +The change to shutdownNow is not good +This seems to be causing warnings with the functional least one of the functional tests blow this away +I think there might be issues here +When connected to the network javadocs built from a source release will attempt to download the Oracle Java file +Great that did work for me +Catch the error and continue +Would that work +Documentation NoteIf WEBCHATCONFDIR env var is not set will try to look for config its files in default locationsIt will look for and in +closing a connection closing all the session a starting idea for to up the great work +Lets see what we can do here +Fixed +The algorithm that we use was that if youve already gotten to a page you should be able to return to it +I think those cover most of the scenarios this Jira was after. +updated patch that cuts over the remaining two qps the flexible queryparser and precedence this issue to make it easier to understand +Verify that their names descriptions etc +I hadnt checked vs +Reopened to set the Fix Versions field to JDO beta. +What exactly is in it and what should I test? Is this about the direct mode corruption +defer all X issues to +bq +I have just started to study Spring Batch to see how to use it in my I briefly describe a bit on my situation and see if there is better way to achieve or it is in fact a suitable use case for this featureI am planning to use Spring Batch for a Settlement systems day end + +Fixed in rThanks for the report. +abouton no longer even exists solr can now start up and server some cores even if other cores fail and the admin UI can now display errors about cores that failed to init either on startup or because of an explicit CREATERELOAD request +You are right +But they too should be allowed multiple times +I just committed this +Ok +Hello TonyI have removed the attachment as you required +Is this still important? I vote that we dont attempt to inject behaviour based on an empty class name +Is this for building EAP before a DRER release or doing a rebuild after a DRER release +Demo looks good but sholdnt the quick fix actually tell what header it is going to insert +Datanodes may receive the same buffer twice so buffers will need revision numbers or somesuch +Will file a ticket +Sorry I dont see the patch in the zip file +In order to share smth +This is the first patch for implement this new functionalityThanks second patch solve most of the issues during packing now the products are grouped and the packing will break only on the complete button if a product has more than one reservationsThanks patch is a trivial improve on packing to see the description of the products after searching by order number on the first screen of packingFurthermore now on the list in case of marketing package are showed also the descriptioncode of the components associated with the marketing packageThanks in advance is now implemented +Closing this ticket because there are comments in Tomcat Closing this ticket because there are comments in Tomcat bug thatindicate this is an XML parser issue and has been fixed in early versions ofTomcat +I made tweaks to the JSP so that it would compile +Most of the time people dont use qualified names within the body of the classes +Fix bug in Yarn AM that was not starting JVM metricsI tested locally by starting an AM and container +the code that is run when is found does an improper setting of the language because it copies the bit set instead of setting each language +We can offer community level support afterwards if necessary +Applied patch in r. +thanks +Thanks for the changes +Leave it open for now Id say +Given that supports so many data stores out of the box it seems like the Isis team should appropriately make it the first choice to support +HiThanks for the information +I just realized youre using popen not popenpopen in MRI does not do the sh adding logic +As I said I wrote it blindThe change compiled in Eclipse and I inserted the resulting +Leaving JIRA open until I can rerun the entire test suite +The default should be if you dont label it as then it is not +to trunk +Bug Solved with currend HEAD version. +Integrated to and TRUNKThanks for the patch Jinchao. +Attached patch made the minor changes to +is an unreleased branchTrue but the branch date is also the feature no? Documentation tests and bugfixes are added after the branch date but not features +I am in favor of Johans proposal for a compromise +Sorry for not having looked into the problem deeply +Here is the revised patch +Sid is this a wont fix +Ive read the spec and it fulfills all the requirements I think +Updated patch format and spelling +This is full recovery debug log from the repro +Attaching patch that simply refuse such queries +DominikNow that the dust has settled I plan to update the documents describing the management interface and the Infinispan management interface +Created a common application context Map +A patch that resolves this issue and solves my problem +Fixed +Moving out of + +Now that work is well underway with AS all previous community releases are +The problem occurs when lvtIndex if that helps when setting breakpoints +It appears that my shop has made modifications to our local install of this plugin +Maybe you can try to remove wrongly added item and then add it again with the proper configuration +It falls into the same category as the taskController +Fixed with commit of. +Small hack that has worked well for the release work +It is a duplicate of. +It will always contain key value pairs and no other things +The code that calls these has been written for the RI +Context that shows the fix +Perhaps this can be broken up into subtasks but hopefully wont be too big ones the other three are done +Thanks! +I got the same warning when I deploy the xml with Camel and Karaf +Confirmed +Closing since fixes now appear in Latest Alpha Manuals. +This is a duplicate of +I am just not sure if we want to mark the datanode as stale if the region server crashes alone +Implemented and tested +I just committed this +The resolver support is also now more generic +Thanks Claus +fixed a while ago with all the solr cloud stuff together +was committed this one is now a dup. +Committed revision . +looking for more inputs +Id like to see it fixed for the M release +resubmitting for the tasktracker should do a join on the cleanup thread in order to give it a chance to complete cleanup of the paths the tasktracker asked it to +are you still planning to work on this +Please let me know if any questions or problems merging the patch from master to +It was a TCCL issue +I applied your patch and can you please test itthankx you send us a patch? I am unable to reproduce this bug since i work on linux +Thanks Nicolas +Eclipse automatically does that for all my java code +Closing this ticket after more documentation was added +Nice find Todd +Ive committed to trunk +fbedaaebacbacab Move down to libmgmtbbdcaccaeaedfa use rather than similar custom codeccdddfaddabdcdad reload at runtimeEthan I made some additional changes after applying your patch +It works as designed +list +For now I would simply suggest to use a snapshot build +Couple of minor observations +Help in resolution is appreciated +I suspect that the earlier resolution was a coincidence as the above poster ll create an example EAR that demonstrates this bug +per kevin. +Surely the patch here has bitrotted +But about the tracing property not set after the first run I posted a seperate mail on this regard and didnt get any reply +Close issue in preparation for release. +The shared api is done here. +Bulk closing for +Tez is already a client side deploy via option . +I should add that the issue can persist once the map editor is may be memory related when trying to take a screenshot windows reported insufficient memory to do so +I tried removing the tags from the host and then restarting the MS but it didnt work and i still saw those errors +Modify freely as you wish +Show me an existing database that lets you attach random crap to a table definition and I will reconsider my position +I am planning to commit these +bringing things up from scratch was trivial after ant idea +Hi Johann Further to the above Ive removed the basedir and built a local snapshot and all the above scenarios properly generate my ddls +in refguide Jimmy? In slab is on by default +Benoit have you succeeded in running the benchmarks? We are considering using asyncbase and would love to see your numbers from the comparison +If folks dont think its useful I can go the filter route +It is possible this bug will be fixed with a switching on a design mode of +AFAIK JDK is not on production systems yet I would postpone the problem and provide a temporary solution for todays technologythat makes me smile anyway please dont get me wrong! because in the dev ML people are still worried about maintaining JDK +And all of that has to be done at JSP compile time when runtime valuesof set attributes are unknown +Thanks Selckin I will make sure I not you in the for all your testing help! +still doesnt have any tests +The problem was I was calling registerError on the parent of the tag and then reportError in the +Hopefully the final patch for this JIRA +done +This patch adds a class and its test to the project so that maven version can be checked when activating a profileI added the dependency to the pom because I used the class to handle version numbersAfter applying the patches the mavenVersion tag can be used with version ranges and exact versions to activate a profile by maven versionRegardsGabor the two patches +Not sure if this setting is available in solr but as Otis marked as improvement I would guess not and the idea is to add it +Thanks DarynHave we been considering these changes to be incompatible changes since they change the error output? +Converted to enhancement request +thanks +would you please try my patch +Yes that was the design decision made in and for accessing HDFS through +Link to the ticket for Antony +Problem occurs when an active profile is specified for a profile that does not exist in the or project hierarchy. +PatchNo objections here +Versions of could be different +Denis can you try again? Or how else can I help you +The right way depends on whether the type is a qualified type and requires parameters +This issue is marked as and is in the release notes. +This happens due to lazy loading hosts when we click select all not all hosts are loaded to UI so we will select only hosts that are already loaded at that moment +Will somebody try this +thanks for the clarification indeed it makes sense to be able to stop a map reduce task before storming all the entries in the cache +Were moving away from the old test harness to junit +I think in the future we should get rid of all of these et type of methods and solely use the jobconfAny thoughts +This is now fixed by fixing the underlying problem which caused the creation of records for a single login in the first place +The jar file was uploaded to jcifsjcifs while it should have been uploaded to orgsambajcifsjcifs are right but your pom groupId is jcifs and should be +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +However deflate does not use checksums and when we use the we should do checksums +After doing some investigating it looks like the applies to the parent bean +gitignore +I think I have fixed the first and the second cases in the SVN trunk +Closing resolved and verified issue. +Fixing the component assigned to this JIRA +For I have tried just the first two because defaultOperator is deprecated +We do not have a way to test such a major change this late in the game forWell Daryn will be doing such tests on Y clusters +Corrected Fix Versions and Target Versions to correspond to actual commits +Including the queries both with and without ORDER BY sounds like a valuable improvementto the test suite even if it doesnt reproduce this problem trivially +Please post your maven config here +I changed the code to use from to avoid introducing the dependency of +It includes the neccessary new classes as well as testsM srcmainjavaorgapachecommonscodecdigestA srcmainjavaorgapachecommonscodecdigestM srcmainjavaorgapachecommonscodecdigestA srcmainjavaorgapachecommonscodecdigestA srcmainjavaorgapachecommonscodecdigestA srcmainjavaorgapachecommonscodecdigestA srcmainjavaorgapachecommonscodecdigestM srctestjavaorgapachecommonscodecdigestA srctestjavaorgapachecommonscodecdigestA srctestjavaorgapachecommonscodecdigestA srctestjavaorgapachecommonscodecdigestA srctestjavaorgapachecommonscodecdigest contains some notes to the reviewer and should be deleted thenThe files already include the Apache license text but the original utexas copyright notice is still left +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +thanks man not its just a bug in the sure that will solve the problem you observed Michael +Resolving as fixed per previous comment and no additional feedback. +We can just bring these frankenstein codecs upto speed there +Ive deployed a new snapshot can you try it and let me know if it works? If not could you post your +Seems was fixed. +Yes this has been resolved +Hi RomainCould you please describe use case which you want to cover by this patchProblem is that configuration of triggers becomes too complex whereas we want to keep it as simple as possible +Latest dev builds include the changes please let me know if it makes any difference for you +BULK EDIT These issues are open to be picked up +No problems on. +Ive reproduced what appears to be this bug and this patch fixes the issue +Furthermore in whole guide refers to all minimal default in different terms profile server profile config configurationThis is more tricky to synchronize than the issue I added above +Thanks Tim integrated as expected +Updated patch that resolves the problems mentioned above + +Done. +I would have to open a new issue as it will not let me attach on this one +Rescheduling for CR. +Any idea why this now fails +the spec does not impose any restrictions so extending to support functions is fine from that perspective +Thanks everyone. +I have the same issue with a project that uses both xdoclet and ant +Last few Jenkins runs confirm my expectation that this problem was a bug in the test case itself that is now fixed. +I applied the above patch +Another way to look at it is that the name defined in JSON schema is what a user sees while the names stored in schema objects and maintained in internal dictionary are always renamed or in compiler terms mangled symbols +Ill have to take a look at this tomorrow +I am running derbyall now +I opened the view Server Server Log in Eclipse but there is nothing listed there +oh i already had implemented those methods as i used them for a project just had to add them to the github and thats it +The codebase now support property replacement for the following replacement occurs within the deployer and therefore only works within and an AS installation of ESB +Ill wait the version for definitive a lot. +You said Try buying a select black then then size +DO NOT USE JIRA FORVery kind of you but Im NOT asking for support +Thanks guys for looking it over +I am not sure if it is functioning or not with the current codebase +Before proceeding could you please restart your Eclipse with the clean command line option +classic package +but for now this is strictly impossible to use jaxen with too. +Do the same changes made for to apply toCurrently extends +for the patch +So be it then +Hi Grzegorz KossakowskiPlease let me know how to do this add mine +I think you should be able to set it while you are useing IW +Marking this as fixed +Just tried out the current snapshot today +I will update my authar tag also +To compile unpack in cd geronimomaven multiprojectinstallThis will build the new classes build a derby database build an ear including jetspeed and the sample portlet apps build a geronimo configuration for jetspeed and build a geronimo server including jetspeed +They fall into Jacks and categories generate less code and split methods up into multiple methods +Ive made this test to make myself more clearLooks like in the development version this problem doesnt occur anymore +Before the patch this report could not be generated due to the +It might handleclass name conflicts but it doesnt handle property conflictsIve added some methods to the dynamic client that now allow specifying binding files that can be used to overcome JAXB issues just like can be done with wsdljava +Mike can you sort out FIXME part +Please open new issues for improvements. +Move back the mailbox specific stuff to Would MPT be a good home for +What tree was this patch generated from +Thanks for the reviewPlan to integrate after Hadoop QA reports backbq +So Im just assigning it to you +Dumping my current state queries and faceting are working but I havent tackled Stats yetThe current faceting fc code works for the single valued DV case but i broke it out into a different implementation to handle multiple values and walk to avoid the binary search in +Nonetheless the Spring Social Facebook API binding will return whatever the Graph API gives inconsistent or not. +Patch that resolves the issue +More on the constraint is the override limited to indexes backing unique and primary key constaints or can indexes backing foreign key constraints also be used? I guess this is really a question against the functional spec +Igor if you did any checkstesting for the patch please specify such info +Ill submit a patch for both of them asap +Run this from the root qpid directory +Closed after release. +Fixed +The documentation is committed +Thats how I successfully get the plugin to work properly +wasnt referenced so I did a svn copy to put it in the www directory +Ive put in a fix for this +This should not happen because the line information is lost that way +could you +One reason I never did this is because it greatly constrains implementationIf delete queries ever became buffered like delete terms currently are then it would not be possible to return the number deleted +Updated version attached +Thank you Peter +Plus process authors can accidentally overwrite a configured template if the mail activity has the same name as some ve made the following changes +OK these are the EAP versions so the upgrade was not complete for some reason +Thanks for fixing it Rick. +sure jar that interferes the most is +Please change to if you intend to submit a fix for +we can make users aware of the reporters with some documentation +Given that this happened before can you please also add a test to prevent us from screwing it up again in the future +I wonder if it is possible to at least address the binscripts part and also to provide an environment that the components can use if available +editor so lets keep it flexible enough for both +I created a jar file of current castor and added it for +Calendar is left as it is no further link is provided to the calendar tab +Not a problem thank you for your effort +dfs already has a tool called that allows to run multiple datanodes in a node +Added some comments on Review Board +Ill try to test it out myself and dig into the problem + as a fixed issue in the release notes. +kfs is an optional component anyway +I just need to implement thisClaudio +This addresses and and fixes the SAMPLE issueI didnt tackle theres a lot going on there and Im not sure I understand it all +closing it then. +Attaching a patch +would be a good thing +done +Marking this issue as partially completed +HiIs this because of the same reason toI am working on it and will create a patch for it later +This patch moves all the classes and tests adds needed imports all over the place and makes some previously methods public mostly for the sake of tests + +Associated this issue with the newly added Deployment Admin component for easier future reference. +Lets delay this until after to date patch applied to bigbang branch in rev +Having generate keys can get expensivePlease elaborate? Im probably naive about a detail but it seems like a simple operation to generate the key register it with the RM which can cache it in its secret managerThat was supposed to read having generate tokens can get expensive +patch with file format docs and comment fixesI think this is ready to go +I submit it to your consideration +The patch looks good to me +Maybe rather than a developer section we need a new book a developers manual? Maybe Ill start out w a developer chapter for now +The issue here is that the src directory included contains a different set of files than the src zip file +Also the allusion is arcane cryptic +thats only change in? I am using todayIs it safe to use snapshot in my prod envThere are huge changes in compared to +Thanks Nicholas +It may work in some cases but it cannot be considered secureThe micro benchmark is impressive but it would be good to also find a macro benchmark that benefits significantly from this +I do +I think its way too much work to deal with way to small a percentage of the requests +Since I didnt get a response on this for so many days I am assuming this is not an issue and closing it +Applied +I get the argument that Velocity ideally shouldnt be required to embed Solr though +Also note that I changed the bevels in the so their shadow matches the other elements +Adding a shutdown of the default task runner will remove this issue but will raise another issue when using the shared classloader of the application server with multiple containersThe default task runner should be associated with the broker server and not a self contained static class +While your at it go ahead and fix the spelling of successfull successful +The large format icons dont exist +Additional runtimes was optional feature +If fixes this issue yeah backportI had to make up a little script that manually disabled regions using just so I could add a column to a table +modified pom in s the exception you got without applying the property +The API could basically already reflect that and are the same and simply specify a type like vs +I think Ive got it around now perhaps a little higherIll post the patch fairly soon still struggling merging with your latest and trunkI think Ive got it all except an issue with one of the contribs must have gotten a little mis merge +Ill look more closely and then comment further +The fix will prevent future growth but wont reclaim old space +for the patch +Its easy for the user to read them using +Should look better now +Do you think it makes sense to mark this bug as resolved and another to track this issue +I am wondering if this error is related to surefire report +Arrggh +Oracleg should also work in to trunk at r +Thanks Pavitra for the patch +Derbyall ran cleanly for me on cygwinxp under modulo failures in wisconsin syscat and xa +I found the job arguments are not set correctly in Job RSL +Be sure to use the class as described in +What is the problem and is it resolved in the example bundle were shipping with? Unless it involves grammar or spelling its likely not an issue ECS can solve + +I tried to add all editing features such as Copy Paste etc but I couldnt get it workingThis improvement will be part of +We could trigger these optimizations using another Column Prefix Delete Bloom filter but that might be an overkill +It may not properly handle the no password case +Fixed. +Closing issue as it has been released as part of release +So I added it with the README CHANGELOG etcShankarI think that ultimately this should go on the wiki but Im still waiting on access +bulk close of all resolved issues. +Committed revision +HiIs it possible to get a unit test reproducing it +Thx! +What OS and AIR version are you using +please try BCEL is correct +dont see how this would help +Even if the IO bandwidth is saturated by these threads it wont slow down IO at individual disk level which is the major factor of startup time in this phase +Testing Instructions hit secured static page login to form let the session timeout hit secured static page login to form login form should be returned when valid credentials are supplied hitting refresh should take you to the static page without logging in +This is not a critical issue +This is actually a bug +Moving all unscheduled issues to +Optionally copy opt directory if it exist +Enabling this so that this is true even when the reaching the syncInterval does not cause a flush unless the flush method is explicitly called +To improve the error messages it catches to print an error about aggregation rather than pass along the exception to the end user +But for some reason it doesnt bubble out +I meant what is the use case for passing in startend positions per file? I support the idea of a bulk call but dont see the need to pass startend positions per file +need ensure this testcase run fix with test +Found a difference in the way css file get loadedIncorrectAll the accessesservertokenccfcdebafcffecjsframerightnocacheCorrectFirst accessserverlangencollationconnectionutfgeneralcitokenaaacfcfbddfffbjsframerightnocacheSecond accessservertokenbbefeddeccadecejsframerightnocacheCan suspect due to an encording that get dropped at ELB +Sorry +I will open a separate JIRA for that +update on this one +Since patches to fix failing tests are generally committed quickly without much time for community input in the future it will be best to save improvements to those tests for subsequent where the community can get a chance to see the changes +The log entry is at the WARN level its not an errorIt could be argued that this should not be uncommented in the config but it is not an error and doesnt prevent startup its there to illustrate what happens with a misconfiguration +This should be considered for the next major release. +We could of course to the latter now but since I have no time to do it in the short time and dont want to twist your arm Ill let you decideIn that case ill go with this quicker fix +Adding the java version of that looks really easy +I will investigate it further and raise JIRA for this if its a bug in HarmonyI verified the commits +ouch Furthermore real blocks whichare still some time away of course will make all this uselessRecent Xerces versions have xinclude functionality when parsing the file maybeyou get Xerces running in the correct mode inside CocoonAnd the least option is to use entities and file include +Would you mind committing if you dont object Hairong? +The check that fails is cproducervctype HTHTTPSERVER in my scenario cproducervctype HTTRANSFORM +As I have helped him some times I dont want to put directly or review here and push this in +Scheduling for as per discussion with Marius +Here is the patch file +So what shall we do with this +The nightly builds of the series or the release do not suffer from the problem closing all resolved issue that have not gotten any more feedback or comment in the last month +This patch adds singleton access to the Includes several unit tests +The use case it to first read the headers and then the content of the fileI agree on the null return value an empty array is better to avoid a null check + updated the tests +Hadoop and Gson s an update were onto the actual Solr node portion of the code and some tests around that +Whats different between these AS versions is the default for XML validation +still is probably +We might add something to the Python getting started guide about this +To get around this I had to switch to using the command line options +It passes the test suite and doesnt cause the strange errors that I ended up with from the other patch +Its gotta be on your end +This is taken care of by throwing an exception from setupConnection and thereby never returning a connection +In commit dfcffdcacbbadefe was renamed to do better reflect its purpose and who should use it +use the conical id also +Im wondering what we want to have out of the gate for this release +Reading the patch very cursorily made it seem as if we werent checking the exit status of the script +It looks like isnt handling STATUSNULL correctly +Feel free to rename the variable +Are you seeing this on? If so how reproducable is it +Hey Mika LaukkaCan you place the compiled jar file here? Javier but I dont have a patched jar file handy + +Patch updated ready to commit +Patch attached +but when i try to access it fails +What test should pass to confirm ENV and ENVJAVA work +It may be null +In the case ofan empty directory it seems to me that an empty array is unquestionably thecorrect result +I didnt understood why the implementation of aggregation would not be obvious changing the implementation setup +The remoting server requires us to mark an invocation redundantly once by wrapping an within an and then placing a in the same s metadata +The new implementation makes It safe and cheap to recreate instances now as the delegate is cached internally +In fact while Im at it Ill amend the English example slightly as well as very handsome doesnt really sit easily with the colloquial minted +Any progress with this Andrew? +Thanks Eric +Basic conformance checking implemented does not support quick fixes though. +Any ETA on when the bug might be fixed +Oh dear oh dear my fault completely +Im interested in getting your opinion on this change which essentially adds a Properties cache to the Configuration class +If I follow the specification correctly the factory implementation classs name should be able to be found even if the context classloader is changed +The motivation described in that comment you linked to seems to be Continue to support configuration of HA addresses +I am interested in creating a package for Ubuntu I will work on doing this and keep people posted if no one has allready started this +Martijn could you look into this +It was released with the fix in +Trivial patch to generate a working python artifact +I had to patch so I wouldnt loose other things that have already been +I agree with your prognosis and will look to doing those synchronisations for +Thanks Sanne + move issues to and +Im kind of annoyed that I had to download patch and build my own version of Hibernate for such an easy fix +Now each module gets added and initialized then the next module is considered +Added patch. +After thus explanation Im with leaving this in s make this a simple decision snapshots as is for or option is to only support format in releases and leave format to trunkSince is coming soon this seems to be a better choice +Java DSL exceptionSpring DSL onExceptionI think we should align these +Also changed the admin functions to return inputstreams. +I think its fine +since this is not fixed yet the onload param has not been added to the yet Added link to that this bug depends on +I think sticky would be ok as get nervous using the Back button with datbase apps. +watch this space +I would think if someone is using encryption they would appropriately configure swap +Not required due to of teiid server from jboss Final is released +Testing this now +This means that we may unnecessarily do an unclean shutdown +Currently testing this patch +link are still not relative necessary to fix +I think something broke this in the new version +If there is a then this should be for it +This is not a replacement for the existing as it does not support the requirements of the UI but it will allow users to access the job history in a way that conforms to the API framework +as long as you followed the remainder of the instructions +Let me know how you want to handle itCould you upload the newest here or give me a github branch I could from? Ill commit the command line parsing stuff from this patch on this go but there will probably be more work and discussion on over thereThanksJon +What precisely does that mean? Is there any flexibility to integrate the Intel MPI integration which is pretty muchChristoph unscheduled issues to the next release +i think that theres a good chance that someone sooner or later will end up needing the extra functionality provided orginally but using fromStack wlll simplify maintenance and make the API more intuitive +I just committed thisThanks Alan! +For arguments sake I tried running the TCK tests using the new IMR as a backing store +I will have a look at it later +This might be expensive so I was wondering if it is a better approach to use versions instead +Yes if your mail server attempts to validate the senders address then that will failWe dont want folks sending to the address the BZ sends it from so this is a deliberate configuration decisionThe is to white list the senders address +Issue closed with release. +Well need to merge all these changes to AS trunk on release +Im running regression tests on the attached patch which cleans up the warnings in the jdbcapi test package +Unable to delete rules at a time in guvnor +Oops I just realize that its the specific recursive case +So we can consider this as a new feature +Ive been meaning to find time to go through and clean up the btree code considerably but havent gotten around to it +Enhancement can wait until. +i think that is what we are doing w +The user either has to be the superuser creator or in a role that is assigned to the group which contains this resource +all issues to the CR +use direct assignments instead of swimlanes for that behaviour +The person who reported the JIRA has an earlier version of the patch that works for himBut I cant say for sure if there is a problem +Ok thank you +Could you please test a recent snapshot to see if it works for you +Unless someone else wants to t fix in favor of a brand new gui +I will check this tomorrow in the office +patch integrating feedback from Zhijie and Sandy as per my previous comment +Hibernate and am having a similar issue with association in Embeddable +The API change is the addition of methods +Has anyone found a solution to the Internet Explorer issues? We are planning an implementation and are researching this same for any information that can be provided +Ive done a nightly run of this test on interpreter on CPU linux server +This is the functionality that limits the number of files that can be compacted at the same time at startup +Please advise +It was created per the instructions on how to add a new service as part of the stack definition but I had no luck getting it to work +This thing is years old and I dont recall this being mentioned on the list in the last half a decade +All patches committed +and are gone +Please correct me as necessaryThe cases arise for customers who do BOTH of the followingA Run with Authorization turned onB Run with a Java Security ManagerFor these customers the cases are If the customer has written their own policy file then the customer will need to add some more permissions to it +Alans patch with some modifications and with the new test cases +Committed thanksIve also created to explore the idea of controlling the number of allowed filtered records and to allow expressions. +Gary all attributes support lookups with a single +But I just tried it with the default server profile and same problem occurs there +This is nowhere near suitable but I am attaching it this early on so I dont loose it or revert it when mucking around +This is new code and were talking about the data files +I definitely would like to see a fix for this +Youve done a great job so far with the patch +bq +Without beingfully immersed in VFS and thus having a slight ignorance of context IMO itseems if you do go the way of the you could get correctbehavior if your filesystem cache is somehow tied to its parent Project +From a design perspective its better to use method calls rather than directly accessing the protected members of a superclass so I support Dougs approach +OK this patch solves some of the problems +Is that what you would expect? If it works for you please close issue +So Id suggest checking the size as a but wouldnt bother checking the content +Need to verify if you see other sessions that were jdbc connections that have vdbNamenull +Ive attached the single band image +Just a strange +Running it through Hudson + thanks Ivan +Attaching an example patch for branch +Your suggestion was to configure the configuration using the system properties as default +the KILL to the containers should come from the RM or the AM exclusively but not from both +but you wont backport that to branchI think replication retry settings should be same layer as or Snapshot settings +OK I think I have some more questions but Ill post them to the list since this bug clearly isnt a bug and needs to be closed +I think its just about ready for checkinI propose to put this file parallel to trunk and branches +Heres a patch that addresses these issues by a allowing the to have multiple Instrumentations simultaneously and b adding a statusUpdate method to that is called on state changes and gives the caller a reference to the complete task state +Ive just removed that method as part of extracting the property name expression processing into a separate pluggable type see Leaving this open for the moment to see if there are any objections to if there are no objections Ill close this as WONT FIX +I want to protect that work and see it adopted by more customers +Update the task with lower priority +Its true that working with a data source using a simple jdbc connection would be all the existing phpmysql apps out there would be ripe for about piggybacking on something like JAG +Any problem that I missed? Yes SPI implementations need to read cnd files and create + Could you port this fix to the branch? Gladly! We just need to identify which fix it wasIll try to find some time to do this but if you can help identify the fix that would be greatFor example the first step would be if you could fetch thecurrent tip of the branch and build it yourself andverify that the problem still occurs since its possible thatthe fix has already been merged to the branch afterthe release was made +bq +I think we can live with the current MOVE and DELETE optionRajika +Yes Thats correct and myfix is not correct per since it impacts balancer once security is enabled +If there is another issue please reopen. +And the lightbulb goes on +not blocking CR lets just make sure we get this bug reported or find existing bug at eclipse with this and link to it +This is definitely a good step in cleaning up! I have updated the jira to mark it for release +The previous one got stale +Simply increasing ROWSCOLUMNS integer value in a Matrix class and commit it at onceThen itll be solved +Also must be used with care as that class loader violates certain class loading constraints from time to time +Pushed fails on cloudbees due to WORKSPACE +Append for the local fs is a new feature and should go in +Thanks AndrewPatch applied to vmtt at repo revision rPlease check it was applied as you expected +StevenThanks for the patch +test for the feature request +I fixed the style nit on commit. +Ok i have tested this plugin in IE and it worked with adobe svg viewer but the percentage graphs are not so elegant +we havent focused on domain mode yet since its fairly rigid and not really good for iterative yes we should at least be able to setup a a deployment now that how domain mode works is better knownmore used +Note that the example in the description is a bit contrived as it involved a Final release I just did mvn deploy and promoted the release in nexus and all was well +Latest version doesnt have refresh +My local mods attached for future on +Everything in between is warming time +Thanks Konstantin! +awesome +I think its better to make sure in rare cases where the permission is denied the clear error info percolates up to right places +All of aries appears to build OK with it. +I think Usage is better and more descriptive +sStandardSimpleThat is is not appropriate for many languages +Can you fix that + id date +Know of no workaround +Otherwise the patch looks good to meNits Some small typos in the comments in that you might want to fix Creat the required tables creat create Its trigger action Its Its trigger action with have with will +Thanks Sean. +removing SearchSearch part +Failed to reproduce bug using Suns JDK +Permission Nicholas use octal but Hoop use rwxrwxrwx + +I just switched to my mac and tried it again exact same behavior +This rar includes dojo +If the broker is configured to not use jmx the optimization is not performed and every message is processed +Committed revision +Oh I see +Spent a little time on getting this to work today but made minimal progress +I uploaded the result from the Buldocs schema parser +Yep i mean the dispose buttonI thought looking at the icon that stopped flow. +Was it fixed in as described +Make scrollable is one of decisions +its not going to be exactly pretty but if users did install this on Juno their install are botched anyway so not good answers. +beans the last few days Ive completed this work +So is officially broken +You are actually moving protocol version check to the call for saving the initial round trip in proxy creation but not removing it +Now I am trying to separate them into a set of small changes each one doing an small thing +From the admin console attempt to upload the application +It isnt clear to me why this is useful +Upon second thought they need to be Serializable in order to be used within a clustered Hibernate Cache +Could you update? Thanks +AFAIK what we have now is closing stale resolved issues +Same problem for for in trunk and +I still need to improve the Javadocs of those methods per se and make the Use +Kevin can you ask your colleagues if they would consider defining the gadget XML spec in the scope of the shindig reference implementation? This would insure that the RI is in sync with the XML spec and give a place to publish the DTD and xml namespaceI feel we will lose a lot of momentum if we have to resort to JSR or IETF style dickering to get additions to the gadget spec. +I can no longer replicate this issue on CXF Sergey may close this issue when he wishes +I saw this in the wild too on a customer test server that had been booted so sooner or later it will become an issue +Possibly the wrong patch was inadvertently applied +Good idea but lacks see this as quite serious security problem +After all the message properties can still be displayed +Where I started the daemons using one user and then confirmed that a different user and the user started the daemons can both successfully run distributed shell jobs +Done +fix the problem of setting the ours attribute in javaBean writer +Logged In YES useridYour example does not reflect the described problem at all +Ill set the jira to resolved for now +Ive addressed your feedback abovebq +I did not use the protocolVersion before +I just have noticed that this issue might be of similar interest as issue Best regardsSvante +I just merged this to +dubto que hi hagi cap expressi catalana equivalent +I also rolled this into since its affecting Redcar +Thanks for catching this +Yes please +The intention is to also provide accessmanipulation of models artifacts and repositories +Fix committed will be in ER +Hmmm Im very concerned about adding more UA specific code into the core +Can this be committed to since it has not been changed to use +Committed to trunk and. +In most cases you should first subscribe to the myfaces user email list and only when someone confirms that there is a bug should an issue be created here +Tested! +Attach my patch +The test case for the telephoneNumber comparison is still included in my protocol test suite therefore we will not forget to implement the missing normalizer +There is no much demand for it. +The API is cohesive it does not involve interfaces as did inI think you forget the kryo dependence in the pom since you said it will be in and there are no kyro API usage in the patchAlso I think we can remove and since we are providing a pluggable place to user set how serialization will to be done +Thanks Matthew +Anyone know of a better way or is this already doneI think the open question is more what behavior you want +Moreover as you wrote in the ML probably there are no users on upgrading the web part except community and it is known is no more activeAs you wrote in the ML its maybe time to speak about totally dropping the web module in? what someone else thinks about it? Keeping it just because it exists shouldnt be enough to justify its existence if no users community is interested anymore +He says this is configurable so we need to expose this in the IDE configuration with the default as single sentences +If you are willing to use trunk the JDBC connector code there has some notable resilience improvements as well mostly concerning removing the need to cast CLOB columns as strings in your query if you are using them for id or version results +Do you really want a BK application to directly interact with ZK? Isnt it better if the znodes are all hidden behind a BK api so that it is easy to upgrade to a possibly new layout in future without impacting BK clientsis there any other reason why a BK api directly interacts with Znodesanother option would be to wrap the Znode access via a seperate BK api +Closing the bug +Fixed in trunk some time ago +this was the only change i had in mind with the passwords +the right way is to wrap segments of the buffer the reader is fillingadmittedly this is a much bigger change and should probably be split into another ticket +I didnt see an attachment on does that need some assistance +From ircchat i know what the problem is! earlier openshift command line tools modified HOME +I elected to change the actual semantics rather than just doc it better I think this makes more sense +i used old source +split by delimiters and do nothing more +This all works great +Thanks to Szehon for the contribution and to Navis for the review +So closing it here. +Website up unix group created +Currently jbpm is in development but updating integration for jbpm is too much effort for Seam team as we are here just for maintenance now after releasingjar should work with hibernate just fine +Ok I got it to merge cleanly and committed it to trunk +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I didnt see it because I wasnt looking for resolved issues +Curses +Ill let you know soon +This is stuff many asked already + +We could always poll to see though +Please review it +I think the issue still exists +Hi I believe you are building the trunk instead of release +Wow +Is there a restore script used to restore mysqls initial state +Ill need some time to investigate this +Allocate a object before read in its contents from the fsimage +Had a better look the problem is trying to get a connection passing a null state passing instead the auto commit transaction works but still the test fails in H because it is meant to work only against prepared statement dialects +a small patch +Thanks +sneaky. +This zip file contains files You must put in datagml and You must put into libwidget Contains an example of how to works the widget +Accept texthtml Both texthtml and applicationjson have a of they are exactly equally preferred + +We can switch once snapshots are published +Thanks Ashish for your suggestions +Yours seems more about the deployment mode +Problem with the users Gant fileCan I get you to attach the Gant file you are seeing the problems with and a transript of a terminal session showing the problem +But this is the standard Menu widget from gwtext +Jamie I added a new class for locking and added tests +The proxying mechanism for old operations implemented here does not work it looks like it hasnt been tested at allReopening to add some tests and fix related and issues. +The common case by the time we start looping through rows is that the row were looking for existsAhh that makes perfect sense +this bug still exists in Although I dont think this has always been there +What if its subWeight is a and I use a Collector which accepts docs? Will I have a problem if I ask for an Scorer +No one else is seeing that cpu usage +I have moved it some how my patch doesnt reflect that change it seems +We will probably configure maven to run a script that generates the index file +Pablo or ChintanaThis patch has been invalidated by committing Can one of you update these patches and remove the old oneWhat Ben commented is correct +Attached file doesnt include JSF and tag in +I would propose that they should be after CDI interceptors +The patch has been applied +Seam project only may need jsfnature in order to be shown in Web Project View as JSF project all other Tools does not need this it removed automatically user can add it correctly if needed by Add JSF Capabilities +Thanks for the contribution Garry! +Foxpro support would be an incredible help for many developers stuck with ugly legacy applicationsIf such an integration cant be tested because of the Apache license restrictions maybe it would be a good idea to host those tests somewhere else +See attached +Is it possible to upload it the project documentation section or should I just upload it to some site and just ask you to list the translation url +Might be good to do some benchmarking in order to get a sense of how much speed we gain by such a modification +Not a problem this is a nice addition +Instead the preamble points the reader at the buildnotice target which shows how the NOTICE file is builtTouches the following filesA toolsreleasenoticesA toolsreleasenoticesA toolsreleasenoticesA toolsreleasenoticesA toolsreleasenoticesA toolsreleasenoticesA toolsreleasenoticesA toolsreleasenoticesNew notice files one for each grantM M toolsantpropertiesMachinery to generate the NOTICE file when the release is built +And make the second complete scan including failures normally in +this patch only adds a on the end if it is needed +The same problem as in linked issue. +The addendum is in. + and are not in can we port these back? Seems like pretty self contained changes +My sincere apologies for posting this issue and I will personally validate any information from my Python team before posting it here again. +Thanks RaulI am trying your big patch +Im heads down with my maven drools plugin so Id actually be quite happy if you take on this subjectBestAnsgar +Ill create a script to reproduce +Yes same behavior in thanks for checking and for the test caseIs this a problem for a particular dialect +ivycache both clearing +thanks Hitesh for digging into that +Tagging features we could implement in the near term +just have runned this test localy on eclipse +Can anybody else run the benchmark on hisher server +Incorrect sql are still being createdIs this a bug then or do I have rhe wrong end of the stick +This problem is now resolved +this is the same as which is now resolved +I added some release notes +Thanks to branch as r +Thanks for the review and comments! Ive integrated the patch into and trunk. +v patch adds missing changes to config sorry for the flood but I messed up my local commits and am backtracking. +I built Railo with your commit and it works for me +It got delayed due to the upcoming holiday also got feedback to try something out concerning this later +I have a tool that will work with Roller and blosxom to chew the filesystem and import it via Rollers internal to the roller system +This works for local mode +The problem is that its hard for the SPI implementation to decide whether it needs to internally refetch the children if the client first asks for the count and then starts iterating +Patch become stale i will rebase the patch and i will upload +OK this issue should be fully resolved by the r commitPlease verify +Committed revision close for +Then I did chmod ax HOME and the tests pass +Randy would you please review my commit on this issue +I started trying to review this +If so we will spit out a warning Id like to do something a bit more elaborate that automatically disables any JSF implementation that is bundled with the WAR +The remaining time for the transaction is available via the OTS propagation context which will be a costly operation to perform on every client server invocation so the preferred implementation would be to pass the remaining time via a message property +day of work in QADEV environment +adds one line +As far as I can tell the dependency on is for test scope only +Fixed in SVN +The is able to all binary packages I believe +Thanks Zheng +Please close this one. +applyed the patch +Thanks NathanApplied in LUNI module at repo revision Please check is was applied as you expected +Thanks for the patch +Added docs suggesting users with recent versions of Oracle to update the driver +I remember being surprised but didnt pursue it in more detail +patch includes all fixes listed in issue was changed to with some mods and tweaks. +The solution mentioned attempts to solve two issues uri will probably contain hostname so no need for a dns lookup to get the hostname to dfs handle +This cache is on which takes care of removing obsolete readersThe Scoring part is contained in andThe will simply load the from based on flagThe gathers rows across segments and returns a globally +I actually knew that and my brain has just been retarded I guess +Thats what I was thinking too +Maybe I am missing something? Here is a list of all my +You can easily reproduce it by trying webserviceconsumer just check the +Being able to add new functions to the parser is probably useful to people +You need to provide copies of your stylesheets along with where they live in the current filesystem +In production you dont have much control over s but since struts makes use of explicit mappings its good to know about s when in development +Merged the patch to +However is problematic +Again this doesnt seem to affect jdk so Im removing the common as fix why did you put fix version? That was already released +all in one swoop! Message gets split and routed to different destinations with each destination receiving different subsets of the as well as receiving them in different formats +Patch adds apache header adds annotations for description adds annotation for Deterministic we go on this? I have three sets of I want to follow up with and since they all share the +I added stubs in several days ago for these methods untested so suggest that you apply your patch +to trunk +Sorry I originally posted this report and patch in the wrong area. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +A more complete patch wintegration tests and some more bug fixes +Sounds like ti might be related to +Please see the attached screen shotIt would be nice if you can check it again in your environment +This includes Asynchronous issues resolved in released versions +Thats fine stack +These tests were done outside of unit tests through umount l etc +Thanks Pallavi to report the issue +However the use of is absolutely wrong so still more work is needed +Weve marked similar versions of this as wont fix +I tried the other versions older and newer and did not see this issue +Applied Eds patch and a minor update for consistent code style +At this time we cannot upgrade to +On the other hand Im getting these failures without the changes as well so it is not clear that Ive caused themIve implemented this additional featue if the set method is called with null or an empty dictionary or the activate or modify method with a Map or Dictionary return type returns null or an empty mapdictionary then the service properties override is turned off and the service properties are reset to the combination of the defaults from and the config admin settingsIm not thrilled with how the return value for activate and modify methods is returned to the caller and what Ive done might have concurrency issues +I also commented on your pull requestNext time +I am looking the flaky test will fix it soonbq +Juca this should be resolved please confirm +srctesteehcatalogtoolsgeneratejavaorgapachehadoophive should be in srctesteehcatalogtoolsgeneratejavaorgapachehadoophcatlog +Why? Do the raw files that I downloaded need further modification? If yes could any hero here please teach me how to do thatIf any hero here already have the patched files please share them with me +I am trying to check if I can manually execute this procedure and check the vmware specific System VM Templates validityThank youChandan +So why dont you just add tableId as I did for app servers where id cannot be dynamic +Rohits and Mins comments about changes seem to contradict +Note that +The IP can take over the token at that point but the state is never updated +Reopen to merge fix to +This would definitely be valuable to me +RS A wakes up and removes logs +Can you either advise how to get around it or release it would beMikael you release a version? The bug is problematic when you use too would like this to be released +This can often decrease the number of heap insertions by several orders of magnitude +So all issues not directly assigned to an AS release are being closed +All tests passCommitted revision + checksums are useful as they permit efficient checksummed without scanning the entire block +I committed a workaround for this issue +Im not too worried about that +Just a mere copy paste +That little bit of code is all thats needed to build an extension +Ok thank you Tatu for you as this is not a Jackson issue. +I fixed some typos and reworded one or two sentences +Since this is a new feature it is difficult to get this into release isnt it +I started work on some generic GUI components that might help Ive got a which gets built by a Listener and started work on some forms for crtieria verifications + +Im sure this is a dup of +Correctly encoded please take a look +Perhaps it isnt worth putting in Groovy +Committing shortly +Never mind I see from looking at the patch that the DN when the client and stored checksums differ so we dont have mixed checksums +Ive backported and all my tests pass +Merged the two factories into a single component factored out the resource provider this be marked as resolved? +For instance container tokens are always used to prevent from launching containers with different resource values than requested from the RM +This could reduce the amount of time required to execute this test case +Attached a patch to handle this like the +Thanks +I am glad that you pick this up +Fixed r in EAP CR. +Oops make that PRENVIRONMENT rather than +Alexey please verify +Its a plain of a bulk update all resolutions changed to done please review history to original resolution type +Thanks Earl. +This should have been resolved in forgot to close it off +files from projectis to be implemented inOther items are implemented for previous versions. +The latter is something of a convention but nothing is fixed reallyUpayavira +Thanks! +To me though both approaches feel a little +slip all minor and move from M to to M +patch applied +Test case Simple showing the unexpected behaviour +ACS feature freeze date is November th +TODO patches have been removed due to conflicts +can you give an update and a pointer to the solution if there is one think i forgot to mention that this should go somewhere into the jpdldistsrcmainresourcesruntime release notes with an update statement that should work with the DDL listed already in that file. +Reviewing it right now +This looks fine except I dont see the required addition to the LICENSE file +fails query when the given partition does not exist +the AWESOME help of Wendy Smoak this issue can finally be put to rest +Including here so we dont lose it +Please check and commit +Hi Stan I would prefer not to add this repository since its just a single project snapshot repo +Perhaps introduce a configurable exception handling strategy implementation executed here +This is part of bug which is in a kind of messy state rightnow +Candidate patch for bringing Nutchs Metadata framework over into Folks Okay attached is a patch that ports Nutchs Metadata framework over into Tika +Please verify +checked in ensure that either expr attribute or text nested element is given revision +I think this was already fixed +tested and the change allows the scp from the top level to work during the +Username bcotton +Each RTF now has its own factory which is much more expensiveIm going to work on doing something more sophisticated in the hopes theres a better compromise between memory usage and speed. +tested tests still fail due to device variable missing as expected +Its not possible to include these changes in the GA moving to +Its also possible that the JVM was killed +Its probably true this patch is undoing some resent changes but if it is anything related to layout and file structure it was probably coded by me in the first place and it didnt work so this should be an improvement +Ive verified mvn clean install +fixed and added test +And heres the patch w tests +This is one way to do it however +Eliminate some extra synchronization for and +Using the original job id is not a good idea unless you block collisions in the job names +Unit tests pass on cruisecontrol +After seeing a demo of Roo at we are also interested in using it for our newest project +mainly the code you use for insertion preferably in plain thrift or CQL as this would eliminate the possibility of a client library bug +See also the attached patch for +So thanks for changing the Issue Type +That will get attended to eventually Im thinkingI think the hang is probably a issue +Yup we have updated to fckeditorBut i am not sure i dont think its going to make it through the incubation process due to licensing issue mentioned by Ted +Works for me now closing. +OK I see +Meh those arent important +Users might not want to see an exception trace if there is an invalid argumentFair enough for invalid arguments though I think at least a debug level log with the full stack trace would be helpful for unit testsbq +Perhaps you could pull from github and try to build one to confirm it +Thanks Enis +Secure hadoop should be the default for future version of hadoop hence I would vote for instead of +You can apply the same kind of +But no others users seem to be impacted +Its also readily available on most platforms +Sorry I forgot +Ping Doug +might be tricky to check because youd need to see if a is used in a String expression +Hi Dhruba how the datanode is done? Just kill the machines +for the new patchwill commit it now +I think a Note in the section explaining the attribute would be useful hereLet me know when you guys are done and Ill make the change in the docs +Added setters where needed +Just the way keys can be externalized Map keys can also be this is backward compatibleMaps are not used very commonly in SOLR +thanks andy +Well see what they say +By the attached patch not optimalworking i meant the one attached in + +I dont think its asking too much of people who need a server to require it themselves +commit in GEP At revision and server at revision Like mark the removed applications when server is stopped and pass them to server when server is started next time the server wont start them and then in GEP could delete them when synchronizing with serveralso using the true system property to enable the feature +I get the impression that this issue is fixed with my patch for this since both Denis and I patched some aspects to make this issue disappear and right now I dont see any other paths to skip cleanup +Someone else seems to have done the work here. +Unless I hear otherwise I will resolve this issue in favor of the server time zone +My preference would be just to build x libraries at the same version IIRC as the x ones rather than clutter the ant code with conditions to set for different library names for each architectureIm also hoping that Oliver will produce an x VME to help us make faster progress on this platform and having to ifdef the hardcoded references to these libraries just makes that take longerAlternatively we could just rename the files and hope for the best +This is done by creating an iapi interface that implements and accessing through that interface instead of through the Util classAll the regression tests passed with the patch +UPDATE PROPERTIES SET PROPVALUE my license WHERE PROPKEY +Are you sure you want to completely delete itWe could also make it read onlyDo you want a redirect to the new cwiki +Im not sure if after a DWPT is flushing we need to decrement what would effectively be a projected RAM usage post current DWPT flush completion +Thanks applied +Client libraries such as this Go library are no longer being maintained in the main projectPlease close +methods getsetEnvironment have been removed from Context +I know there are ways to specifydetermine which versions of things to use and make this kind of thing work but it has been my experience that it always becomes overcomplicated after a while it is just bad jujuLooks like the Hadoop classes in examplestargetdependency are tacked onto the classpath inside binmahout which is why standalone mode currently just works on a box with no hadoop install +bq +Note that all references to modeshape in this doc need to be prefaced with the fact that it is only supported in and later versions of references removed +Donebrook and anew will need to create reviewboard accts before I can add them to the oozie auth group +For this file I have the a lot I havent +Can you provide a minimal WSDL that reproduces the problem +Shall we do some judgements before we create the I think if no default subject exists return the is more reasonable +It then frees us up to do all kinds of changes to the underlying logic without the definition of their documents having to change +New JIRA is +There are multiple but the problem is still not fixed +is best for now and lets us brute force the issue for now +I ran through CT and various tests and everything appeared to workI will post snapshots of everything including a framework distro so if you could check in your own scenario that would be great +This patch obsoletes my last one and fixes an extra get of this that was causing verification errors +By tearing down after each test All the ee tests pass nowThe new patch includes the following changes Makes the ee tests more deterministic +I will test and commit +IIRC Subversion changed the output format and the SCM library hasnt been updated to you be interested in contributing your fix as a patch using the matcher as you suggested +TBH Im not sure I just ran across the change in the EE spec so I wanted to raise the issue now rather than possibly have it become one more TCK problem to fix +Patch for here we go again +QA Verified The server is already started. +I look up the code if we change to a or null then we need to change more code +Updates should propagate in time +More doc coming +Attached patch to be committed to SVN +Have compiled XR under linux x and created eclipse bundle for it +I seeI will commit it +Attached patch should fix this +When Datanode is started it checks if volumes tolerated is honored +Claus sorry for the misunderstanding my fault I thought it was just camel case I didnt catch that should start with Camel +xthanks again for the great patch Antonio +Im not opposed to moving it if others agree thats the correct thing to do in this case +Otherwise just use patch +poms to upgrade to and this disruption seems unnecessary. +Lets make sure the quickstart pull request is ok and merged in repo before we try to integrate these into the archetype +Juergen please take a look and let me know if theres anything else youd like to see done there +Personally Id like to see the methods in Me too Ill make that change before submitting the patch for inclusionHi that sounds coolAfter sitting there in open state for almost years its nice to see a patch is availableThanks a lot TomRunping +Please refer to the release note on that issue +What are you getting? Show us the need for this + This bug has been marked as a duplicate of +Now the functionality is complete i think Sorry that it took so long but know it schould be completeCheers and have a good day +I am uploading the files seperately from the patch +I have done some research and it seems to me the rendering the calendar is not made visible when clicking on the popup button +to see test results and warnings before it can be committed +and have been released +Mass closing all issues that have been in resolved state for months or more without any feedback or update +No test case here to validate +OK lets add it to toplevel too! Always better to error on the stricter side with licenses +I also cleaned up the syncedCount variable that was redundant +Commit was accompanied by a case to verify functionality +Claudio F +bulk defer to to +bulk close of resolved issues in. +We cant exactly do that but in cases where it is clear we can +Still not working properly +Committed to the trunk and branch. +Patch applied! Thanks Kevin! BTW next time use svn diff to generate a patch file +r was backed out and a different simpler change put in r +Yes this needs to get fixed ASAP +I am assigning this to you as you master the buildmagic files better than me + +committed +That spec update may supply the framework you need +renamed the service names from AIM to aim to be more similar to the service names standards add the possibility to specify the certificate alias name when send it to AIMThanks into rev +can be solved with the Pattern annotation Im not gonna add it to the common validator pool +At that point both the network wide deny rule which was not created at network creation gets added as well as the user created allow rule +finder without osgi is the standard imo +Looks good committed to trunk and branch +I think we should improve this by minimizing the locking +Ill double check what the server does with cas foo rnbarvarn to be more precise and get back to you +and any pointers to what I might have overlooked +OK I see the email thread about Unicode data in general does certainly cover this +If has a writer no lock is acquired on updates +One of the biggest issues with log aggregation on a large busy cluster is the number of files it generates and the write load it places on the namenode +trunkCommitted revision +This has been completed for current schema versions only +If nothing has changed it wont rebuild the JAR +No answer from they side yet +Additional copyright language should include Copyright The Apache Software Foundation Licensed under the Apache License VersionApache Cordova and the Apache Cordova bot logos are trademarks of The Apache Software FoundationOr something similar +Vinod Kumar Vavilapalli checking just in case unmanaged still work after this rightYes the tests passes. +Also patch in test needs a small correction in the empty comment +William is again busy working on patch for Camel +Also fixed a minor bug with the collection change events where it was ignoring ADD and not performing any of the REVEND functionality which can lead to issues when removing something from a relationship and then adding it back +Oops good catch +It seems like Ralf wants to disable LDAP and to enable LDAPS only for security reasons +A few comments get and get seem a bit redundant in the interface +Does this error also occur for you on a default Indigo Java Eclipse IDEDid you specify a name for the process and a folder in the Activiti project +The simply contains the topic used by the example. +If no fault handler is present to catch them they are propagated up the scope hierarchy until they hit the process instance and cause it to end above procedure was performed without emitting any diagnostic message at all +ok got it +Now the steps are misspellingsIm sorry for inconveniences is it OK now +Pavel attached is a work in progress that implements the insert and multiget tests +Patch committed to SVN trunk with some changes +This is not correct since such Blob objects should live until end of transaction unless explicitly closed +Ah scratch that just beat you to it +Could please describe your environment in details? In addition we do not have the functionality to populate list with any suitable +I didnt realise this was a comment to the above the final is not so much an issue in this context but the missing static should it be necessary to explicitely mark the fields as public? I am a bit hazy on my groovy but I was under the impression that by default members were public if not qualified in any other you dont need to mark them as public wo access modifiers the default is public +I think XML directly is best +Some changes the groupId is ommons and not and Ive removed the announce section +Just revert changes and modify a little bit +Test code does not have to be pretty +Closing issue now that has been released. +Make another copy of the default configuration +I will review in the next day or so +Also please note my question asked above +Ive added this to SVN but it runs without an errorDuh sorry forgot Id been working on a fix so it was testing that not the original code +It looks like what is on the trunk is trying to do the same thing as the patch I submitted +It should get pulled into the next release of AS +Oops I meant did you have a good idea how to specify ordering +ChunhuiI have started with my review +Blech +Closing. +Jena is correct in its behavior here with regards to the latest version of the Turtle Damians wording is more are more than welcome to submit a patch for Jena for the RDFXML parserThe parser you are using is the old Turtle which is going to be removed +yeah test needs to depend on package i believe +Thanks +The string check on the is not ideal but I cannot think of anything better somebody with more knowledge about our RPC should chime inIs the needed? Why cant not hold the sequence number and get a getsetSeq method +Got rid of the logs Renamed to and to +I can remove it before committing + +Already fixed in trunkbranch +Bulk close for the release +whose class or superclasses lack a public +So now we just throw an exception if it is found thatunlogged operations were running +Closing the resolved issue. +MRI does not explicitly load the thread library but it does present all thread library features from boot +I dont think its critical that this issue is fixed in but it should be fixed before the release +What is plan from s POV WRT to migration to codec in this area +Without additional details as to where exactly in the code those spikes seem to occur Ill have no other choice but to close this issue as cant reproduce +This does not seem very efficient +please help me anybody +Namely finding a job execution by its job instance +The stop semantics do not match the corresponding service startstop needs to be a seperate jsr service that maps from the client side operations onto upload undeploy events as described in the associated jboss forum posting. +if you you see the full in a tooltip o dont remember if there was a reason for doing it like this instead of showing the full date +Thanks + +Do you have any special Maven Checkstyle configuration in your file +properties files +Oops ignore that semicolon at the end of the id line +Hi MartinBasically the site crashes on selection of the drop down choice on I pad +moved up to next group escalations on ticket creation by adding second to the ticketgroup created dateOrdered in this orderCustomer groupSEG TestCreate a case or several cases with different usergroups +This is spring property placeholders +added exclude tests minimr tests are excluded from regular clientpositive tests did some subtle changes in how and are specified to allow testing against external hadoop clusters +We can always add creation date later if we need it but right now its just unusable baggage +Can we make a package protected class in? This way we dont have to make it part of the public API +Removal of this code also reopens the discussion on expressions in POM artifact coordinates but reverting this feature is intentional for Maven +Patrick +Is there a way to force Ivy to generate and upload Thanks +I think Ive found a solution and need to reopen the issue to attach the patch +Ok +Instead of adding an extra Ignore processor class I fixed this in revision simply by making ignore all operators that are listed in the properties file without a corresponding class name +I think we need to reset it to no? otherwise. +JacopoNow that there is better locale support in the framework there is a possibility that this problem has been solved +I fixed the immediate problem in DS in your last stack trace rev have not tried again to upload a snapshot +I agree however that granting allPermission to junit is not the way to go and that privileged blocks should solve this +Committed partial revert as r +Since we use reflection for everything using the string impl was also quite convenient +If only core Attributes are used everything is working fine as the passing junits showI added a corresponding comment to the javadocs of that class +Ok I am closing the ticket then. +okey so for now its the workaround just assumed that the settings would be not saved and then in a later fix we could store these in some preferences per file in the workspace settings +I dont think removing Date is the right solution +I dont think wed have the problem we did from before with the thread already being tied to a transactionThis follows the EE model but I have my questions as to whether this is the best way + +This patch accomplishes the merge +that would make sense to have as a child of the project instead or +bulk defer minortrivial issues to +Kuru yes the JVM you run against should understand the server argument aka on windows that means installing a JDKSimplest thing for that is to use a JDK or as you have done cheat by copying the client in to look like a server. +Nisset pointed out weve used the same cache region for objects and query cache and seems a very plausible source of the problempossible mitigation may be done by turning the query cache off by configuration +which we will want to do at some point +Reattaching the same introduced not related to this patch +Verified on JBT. +Something you can look into +Sure Im good w simplifying the license descriptor +Ill get a fix up for this todayWhy is it that Hudson didnt include in the +XXfalse and XXtrue are needed for CU to turn on patchBug which was found by Rana is fixed here +It is also not part of JAMES +How does the user verify whether the above commands returned inaccurate resultsOperations on the table will behave as if the table is not actually in the state returned by isenabled or isdisabled +The log shows that the seam application starts first but it is while deploying the contents of that it hits the issue +The are no longer appearing +This setup methos will be called for each mapper for each input split +The problem is how to use it compatibly with the Native lot of the methods in my current XTS build are currently parameterised with instances or employ local values of this type +Added missing licenses to all xsds wslds jsps filesPlease review and apply the patchThanks +Fixed in Branch as well +Could we use the registered Derby driver to anchor state? Are there other Derby objects which we could use +I love those +for QA Test scenarios covered Ran normal sleep randomwriter jobs and verified if the slots like running map tasks Running reduce tasks occupied map slots occupied Reduce Slots Reserved map slots reserved Reduce Slots +Will file another jira to fix filedirectorysnapshot deletion +release bulk close +dont you think? +What does it mean in frame of Seam? I think nothing this issue is in Kevins comment unrelated to Seam or am I wrong +Patch to add the actions +edit memstore +One drawback of is that it may increase the load of NN +Tim bad luckYour changes break building the javac tool +bq +So this is not fixed at all by +For anything like this please allow at least a couple of days for review before committing +I can confirm that this can go wrong at codehaus but Ive installed a complete system on a laptop and here the release stops if the tag already exists +Sure Im attaching two screenshots of git and svn email log messages upon commits +Could someone review my patch pleaseThen we can create another ticket for the block report problem +when one of them is a shared library then it should be sufficient to include its parent directory into is a patch adding tasks cwd and parents of to the LDLIBRARYPATH +ipa to +Looks like my unit test was not good and did not properly enable JAXB annotation introspectorNeed to try to fix properly. +The second test was trying to create collection with maxsize but collection with name logscapped already existed so the old collection with maxsize was used instead +I propose to commit the patch if you guys are ok +i dont treally want to check it in with hard coded references to this temporary repository you asked for +I take a look if you have a specific example it helps me to better understand what you mean +Closing this issueThanks for doing the work Yun! +I got mixed up +People please do not reformat code at the same time that you modify itI wanted to review this patch for correctness as I have done a lot of work on the calendar component +Got bitten commenting so I get notifications about updatesUsing the TMPDIR workaround from is working for me for now. +Thanks. +What would be sufficient in this case would be that the compiler setsunsets some kind of property that says yes grab is being used and if the launcher sees this property set it can do the right thing when dont believe you are restricted to having to run them as scripts +Committed with regression testsNow common DGM methods that take a closure and iterate over the target object will have the implicit variable appropriately typed. +I close this as Wont fix +If you are seeing indexwriter already closed type bs its probably this +applicationssrcjavaorgapachegeronimoconsolejmxmanagerM +Jan this patch has a good encoder and includes a good number of test cases for normal use +It definitely worked in +We found this issue too + +Do you have any idea +Issue is not coming in the latest trunk code +It doesnt try to validate the though perhaps it should +I have also tried going back selecting all going forward going back and reselecting tomcat and the result is the same +batch transition resolvedfixed bugs to closedfixed +Using this ticket for protocol related authorization checks +fixed committed in r +Sure Ive come to the same conclusion in the time that this issue has been open +Bringing back in as critical +Simple scenarios as in your example should be supported now I have added a new unit test class to verify thisI still have to update the Javadocs and the documentation accordingly +So the input file with name of the file to read is local to task but not the file that it then readsNot saying there is anything wrong with your results I just think it needs more explanation +No this section in Admin Guide and close issue +there is no upgraded to the current trunk and I still see exception every time I run itClaus there is a bug in the unit test in my nd example two routes are concatenated by dot not by semicolon +Ignore this jira +I will be looking to confirm this is a me problem as opposed to the plugin needing an update on what it emits +It adds debug printouts to the version using a logger would be equivalent the goal would be to extract a trace history of the memory cache +Fixed in trunk backported for +Was the tier created along with any ACLAns +fixed in r +Use and configure it to pickup artifact at your internal repo first then repoOn the second thought It is a real pain for this setup +Committed revision +I do have some minor quibbles about implementation but i want to try and get this into +hadoop jar s patch along with the fix which makes work again +I just would like to give it to youRegardsPhilippe +This bug is the same as +I think that this is the normalWhen you delete cookie your session is still valid on the server side until session times outinvalidated +Manually tested. +This change is wanted to be abe to have like in maven tests classes and ressources in the classpath first while you are executing your tests +Ive added this check thanks. +Im going to apply the patches +This requires new code that parses the HAR index file and maps a partfileoffset datafileoffset +Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadcomponent Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadcomponent Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadcomponent Sending content CTrinidadTrunksrcmainorgapachemyfacestrinidadcomponent Sending content CTrinidadTrunksrcmainorgapachemyfacestrinidadcomponent Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadcomponent Completed At revision +I have so many questions Im not sure where to start +Yarn is required to bring up embedded starts RM and NM +id prefer start and stop as the method names though +If someone could please review and commit this to I really would appreciate it +Done +I edited some of the flags of this issue +like the fix is to make it use different files for compressed and uncompressed tests +Dimitris can you help assign this JIRA to a suitable owner +Can you please handle testing and committing this? Thanks +I just submitted +Apache Felix Declarative Services has been released +I just committed this. +According to the Online Doc on will be called whenever an item is skipped while reading +Patch attached +Patch available +Chinna Compilation still fails for me +Verified on Eclipse with Tools and JBDS M +For a given type and size it always produces the same set +Recently I updated this patch to trunk and got rid of the threadlocal usage and Query rewriting that was the reason we had to pull this from trunk long ago then I attempted to override stats on with global stats this is when I realized that had no affect on scoring anymore this will now be addressed +Our customers all updated their jobs +Screenshot shows the error that occurs when running the test application with Tiles jars +It only happens when you freshly create a user +See +So embedding API is context class loader safe nowI read the stackoverflow question and the but I dont think I can fix classloader setting issue for +here is a patch against +There seems to be some code duplication in the tests and the coding standards arent quite right for the if blocks but it works and all the existing expression tests are passing for me +Since it scales the buffer linearly it will do order n array copies +Oh +If you find further issues please create a new jira and link it to this one. +Fixed in rev +Also there is no secure artifact in the maven repo so depending on when is cut out we might want to push as well +If the presence of author tags is a barrier to graduation then we cant have them +Sorry I dont fully understand this issue +These are the last of files for this patch +I missed the comments on. +There is no single interface we can look for +I change it to future release +I have run all the unit tests +Committed to trunk with revision Awaiting daily test runs to confirm that everything went well. +OK Fast Forward size set to records. + +Thank you Pete. +There may be a lot of regions + patch looks good +These attributes are value bindings that should return a string message to be displayed +John what are the test errors that you are getting? We should fix that as soon as possible +Dont be afraid to ask for assistance if you need any +Thanks Mike +hiThat would be a nice to have +Committed in revision +Should we close this issue +Applied at r with minor modification of the testWould you please verify it +I no longer see the NPE exception and the service works asexpectedThanksAnthony Bargnesi +I tried downloading another patchFrom the addressftpsoftwarewebsphereappservsupportfixesPKBut i Couldnt get this problem solved +Should we go ahead and commit the Enis patch and publish snapshots w the hadoop versions published with a version that has a hadoop suffix? Patch looks good +The region servers should watch their own znodes to make sure that they dont continue to live even if the master sees them as dead +yes sure +Misuse of. +Six months has been entirely too long for them to be in an untestable state +Thanks Jothi! +In this peer setup one peer grabbed leadership to start and then lost it going into an infinite loop while the other peers went about their businessBut by far the most common log messages that are repeated are the ones referring to opening and closing transaction logs and very little information is given about why the election process keeps restarting +applied. +YuanBug fix doesnt need release note +I modified my copy of the mailet in this past week to allow for SMTP Auth in response to some of my ISPs suprise changes +Please reopen after we release if you still have this issue +I do recommend that this patch get committed +Please check that the bug is fixed +What is the exact syntax of the command that I can run to see the failure on my machine +So this patch throws a VCEVENTDOCLOSE event at a but it doesnt record that fact that the event is outstanding? What prevents the from being deallocated before the event is processed +overall +Upload the generated html file by the latest patch. +These metadata tables can be used to find the actual column can be added to the Dialect class and overridden within individual Dialect classes such as RegardsAshish +Therefore this rpc engine needs to call get +Reopening to backport to branch +I suggest to create a custom implementation reusing the api provided by into its shared package +Sure! by better compression I mean to see if its possible to reuse the duplicate jars if we really need them without taking space +TranslatorI did a massive of files in the physical layer package +The new child node is also displayed +Lowering the urgency this looks like likely a test issue of some sort +Attaching a patch that allows job to continue even if acls change across restart +Can we commit it nowThanks + +Ive just committed this +I can reproduce it on Ubuntu +Stefan a few comments from a cursory look +This feature needs lots of effort +I was just thinking how I could make better use of in this situation +Moved the pending over to +Thanks Sean TonyI applied a modified version of your patch to LUNI module at repo revision rI enhanced the comments and chose a message that does not imply we need to implement these methodsPlease check that this resolves the issue +I was just meaning why IT would always return null + +Id rather commit it and worry about later +Were keeping this issue open as this is something we plan to add in the future. +CR is using. +AFAIK this is on hold until the dependencies are available +Andrew I dont see the branch and it looks as though the update commit is incompleteReopeningCan you push up the branch? I can fix the rest +Had a bad experience with Mule where you had to use a special context listener in to init mule +reopened to tidy up the transactional collection wrappers used in map daos theres no point in using them its just the redundant transactional collectionsince were just modyfing a single object during step execution there is no way to achieve transactional behavior in the map repository +If we have more than streaming sessions Ill live w the overflow +I think this is one of the problem we faced of having +Doesnt the cache check for the modification date of file? Seems like anatural thing to do +Rebased the patch to latest trunk +for me as well +Keep in mind that OID macro is not an valid form for an OID +tnx for iplemented +Thanks for taking care! +Thanks for making these changes Aniket +Committed to the branch +I wanted to confirm that this is still an issue in the release +This is hard for me to test since I have not been able to reproduce the error myself +Editor too +Eli Im fine with your fast move +Did the problem appears when is used +just adding for tracking +So changing the configuration name and semantics will have impact on those releases +ExcellentThanks muchJohn +good +Jose I committed the third patch thanks! It doesnt actually run the tests for me though +Wrong script to generate hadoop and hadoop poms that have modulesand profiles appropriately setAlso cleaned up our dependency settings after review of dependencytreeand dependencyanalyze w hadoop and hadoop setupsPurged our use of slfj +It turned out that my viewer was the problem not the. +Created a patch for branchI did both a few unit testing and manual testing for both and +patch for +You had brought this up originally and I think it would be ok if you just created a simple FAQ on what needs to be considered. +This patch addresses the last suggestions given by Camille and Henry at the review board +Also fixed a couple of problems with out of date config in tests +Is this fix also applicable to Commons Configuration? Strikes me that it wouldbe but I havent verified +works for me under tested Report Problems wizard on Windows was created with correct log fileMax +Best regards. +Any reason we dont want to add it as part of the fix for this JIRA +The bridge would generally only be used without a local consumer +Does the fact that the container is reserved change that somehow? +No it still fails +cannot stop the test +Thanks for info glad to know that proposed workaround works fine! However indeed this case should be handled by plugin automatically +frameworkwebtoolsconfig +Ant dist only gets called if you are calling the create script from the src +Wontfix based on Rodneys comments. +Corrected a bug due to a typo +Resolved on my Infinispan fork pull request issued for both master and +Invalidates excess corrupt replicaPlease review +A test which updates a map key passes +Another reason to follow the normal way for patches +Thanks Nicholas +Here is my patch +Creating patch with git diff +patch with Simons suggestions +Raise an issue in just created for this +Where it is +Marking these as targets for backport to so we remember that they are important +It includes all libs and can be run directly on a web server tested on tomcat Deploy on Web server then go to httpservernameport and click on the calendar icon you will see for example with firebug that it tries to load a resource from this url httpservernameporttcdvaluelocdeenc which return a +This is not secure but it is possible +Hi Mike is released under the License a liberal Open Source license similar to the BSD or MIT it is BSD like is compatible with apache license +I misunderstood how some of the debian packaging works +And nothing prevents us aferwards to implement the one for enabling s wait and see if other opinions are I will try to patch it this way probably in March or April depending on our projectsSylvain +Seems to not compile in TrunkYep there has been drift +We broke in Avro formerly we used Utf and made maps with Utf keys +What is the impact of not having the Java compiler on the system classpath when starting the as? Is this impacting Tomcat jsp compiling? Im just curious what the issue was +according to your suggestion i cleared the load and patched on my cluster to testuntil nowit works will try to test more senarios and then provide the new patch to you for reviewing againBTWone hserverinfo object occupy about B memory though cleared the loadif we dont use my ugly refreshing solution in worst caseone region need one hserverinfo objectif a big hbase cluster have regionsthe hserverinfo objects will occupy about B you think this can be acceptable +This change also fix an existing issue of old entry logs being removed wo forcing new entry logs flushed +The console log above shows this happens for INSTALL not UNINSTALL +Rather than upping OKRELEASEAUDITWARNINGS I think its better to add this file to the list of exclusions in in the configuration section +but you will notice both the Lucene qp and Solr dont allow leading wildcard by defaultif we add this automaton stuff I think we should reconsider this ruleinstead of dont allow leading or? by default just dont allow just leading by defaulti wont really argue for it though because a wildcard query of abc probably just as bad as abc +From there I want it to not cache s generated +May be u just dont see that there are unnecessary figures +A Test revealing the bug for Test revealing the bug for sampling is under effect the original is replaces with a wrapper which takes under account different sampling related parameters namedThis wrapping class modifies only a small set of parameters in the request and should otherwise delegate everything to the original one but it does notSome of the information that is lost from the original request number of results to lable and DepthA patch would be available shortly +Ill fix the ones Ive introduced myself in this patch +QA instructions after the UI is in place UI verify that Vlan can be set for Isolated network with the source nat service only from the Network tabCreate network and only seen in ROOT admin UI +Set affected versions to and +Merged only into master thus fixed version is still you should just set the fix version to the first community release where the issue is will be fixed +But I think that the benefits outweigh the drawbacks +So here is the new patch which is latest and updated +The correct solution in that case is to embed the JAR into the Abdera bundle that contains these extensions. +In the end the only modules which require Maven repo are docreference and docxslt +We should address the nocommits I put in here as wellI checked and they are not related to +Im more than happy to setup the scanner myself but then the AS tries deploying the items twice +so the designer deployment feature is not needed in we should offer that functionality out of the box to our developers in the project in the product people do not expose this fnctionality on the internet without them knowing about it +The issue wiil be in +Your example will provide a Long result +It also seems to violate the exception hierarchy you have created since is a which should indicate that the peer closed the connection but would not in this case +You said so all DOM tests fail +If you change that to an idea we have to change these things inthe demo setup as well that we create the demo config in the database forthis user id instead of the nameThe REST webapp then has the user id instead of the name? Then this shouldbe the only change I think of at the momentCheers big impact +As it is already shown in the chart +It is a problem with the cleanup code not a concurrency issue + +Some other code reformating changes in this commit also. +I have fixed the ulimit issue via ssh to vesta and hemera Pre Fix ssh pctony ulimit n Post Fix ssh pctony ulimit n Pre Fix ssh pctony ulimit n Post Fix ssh pctony ulimit n +Java on Vista +Not sure why i removed the line but i changed the in js file loading a little bit +Committed to trunk +is this resolved then +just at a glance it appears the test tries to create a file with a double quote in it +ear files is there +Do you have a m snapshot repo available as well +Hi ClausI kicked the servicemix bundle deploy build after I got the CI build error emailNow the bundle snapshot is published and one of camel Apache CI build is passed +Unless there are objections or improvements ill commit this one in the next few days +If we add locking then there is a time duration between when the local key would be locked that the owning node could have received an update causing a L invalidation +The security interceptor does not do the logging +Committed revision +It will go into the next patch +Im marking the issue as resolved +This one is correct +So we would no longer be able to do a normal push from Codehaus to Github wed have to do a force push and then that would impact forks of the Github repos +The only thing is that I havent gotten it to work quite yet +Dan Im assigning this to you +Other proposal for writing the list of region names in wal header and reading them to determine which tasks should be complete before to make sure the assignment seems more cleaner to me +Resolved also for +Changing the issue to be a launchpad issue to update to the latest Felix Declarative Services release once the vote has passed + are conceptual issues that prevent me from opening this API +So this issue is only when unit testing and using that feature. +problem on Daniels comments the issue should be reopened +I see a fix that is localized to a single source file and every change appears to be directly related to addressing that bugAs Oracles implementation was originally based on Apaches it seems quite likely that they would have a similar bug and that a similar fix would be in order +The patch looks good but I havent applied it to be tested yet +I somehow missed that one +This also improves dependencies in our generated Maven POM file +I think we can probably add XML escaping to the output by default and this is probably a good idea +I just committed this +I attach an POC patch to is the real use case behind this issue? Can you link the original issue with itMoreover here you mix perf and memory for me I dont believe that the perf part can be improved because I unfortunately dont think that the permissions can be skippedLet me give you an exampleWe have results to the query but the current user can access to node node and node Offset and limit Current code returns node OK Code without permission check until the offset returns node OKOffset and limit Current code returns node OK Code without permission check until the offset returns node OKOffset and limit Current code returns node OK Code without permission check until the offset returns node KO +It may be due to a bug in the FS layer +I merged the patch to. +is there now but it doesnt really fitscale anymore +This seems to be caused by the linked issue +bq +I could not actually reproduce the same issues but Ive fixed a few things in the renderer lately that might account for the m still seeing the problem but only far away from the area where the projection is any useful for example that stereographic in the screenshot should have a linear deformation of near to the areas where the rendering issues are showing up +Thanks for pointing thisDid you see a way to change the final status of contentLoader inYou caching class can be use inside droids? If yes can you provide a patch from the projects root? Its more expressive and understandable that an attached class +But whithout tag ajax rerender wont work +Weird we havent run into this issue before +See attached patchIm still not happy about the jmx stats though +Looks fine for in rev +The fix is currently targeted for the release +Doesnt the wiki have instructins about how to add these files now +moix gat is in the DIEC nd edition +Ive been very bad at giving links yesterday but indeed the crash caused by GTK and is already reported in. +PM team decided this is a nice to have for SOA +Not for commit here +Applied in r +Could you include the code snippet that shows how you are getting this +bq +it seems that before this back port example as a fixed issue in the EAP release notes. +Try to use different from the parent pom +sorry I missed the checkbox +Closing the resolved issue. +Thanks Hairong! +Use formatpgna to use it or choose the png alpha choice in the more serious patch +Its actually build that fails +The change you suggest cannot be implemented because there is no restriction ona that it contain only String values +Thanks +Normal users dont hover mouse over message and they dont use tab either + +Nice find and great issue with a lot of detail +this example wont work in previous versions of JBT nor JBDS + +Ive committed your fix to SVN. +So we just log the null reference and blockinput +that looks great Rob +is too eager I think +this is a test case timeout in this test case scheduler will not be used +bq +However I recommend to deploy HICC with read only user independent from the regular data processing node +Marking this a blocker until we get to the bottom of just committed this +Olaf can you look into this? Thx +This is not the best solution +Patches applied all works as should closing issue +Current state of development +Merged to the branch in revision . +Patch for both x and trunk +It as functionality to allow multiple type of job resources including jar files classpath jars and classes files and most importantly directories +This is where the AM sends the startContainer request to the NM +release notes updated with feedback new known Darrin I was combining browsers and section +I committed Kristians version of and will leave to Lily to submit a follow up patch for the issues pointed out in his last comment +Ima fix this one +Would be just great to have it working without own implementation of +Jasvir thoughts on this one +I noticed that the new files are not committed +Will commit shortly +I guess our Perspective is inheriting from the Java EE Perspective thus leading to a fairly crowded menu and we have a few more things to add so making some room would be nice +Please use a different name for now +Now it seems that is fixed but maybe there is a problem with multiple static inner classes +I close this for now. +I am sure I was a little aggressive with the refactoring for SVN +Looks good to me as well +starting on a patch for this patch as discussedAdded class to handle reflection code for servlet init params +libhomeedwardworkspacebin +I dont want high entry barriers to this module +After updated with the patch in is passed in Windows and Linux but is still failed in Windows +I am unable to validate this behavior right now +Thanks Dag + +this removes the need for any changes +Agreed findbugs warning is unrelated +Try to clear clear browser cash +I dont think that is a good idea +Sure Ill get to work on that +If you find anything that seams to be wrong or can be improved further let me know and discuss further improvements +I can confirm that convert hangs everytime I use a java JVM regardless of which version of I use +rakesh are you working on this checking now? if not id like to take a run at it +Thanks Ivan I commit the code according to your suggestionCommitted code in revision at trunk revision at s also fixed in trunk +For I ended up using most of the javascript features before I was done +Attached is a modified patch which mainly add a clean compliant wsdl generationThe documentrpc style is discovered given the abstract wsdlWe will need a more abstrat description of the binding to handle WSDL and rest bindings and also enhance existing components to use the wsdl jbi wrapper +Subsequently we can add support for calling the C implementation from Java +How did you configure your war plugin to use the version of the jar dependencies? I feel like that is a little cleaner than keeping the timestamp versions around in the EARs lib folder. +and visibility is one of the primary reasons were moving to Activiti +I think Ive spotted the problem the deleteRaw method throws an if the inode doesnt exist unlike the DFS or Local implementation +This request is instead similar to what does adding a text in the message bodyMaybe it should be better to add a flag in the mailet to put the footer at the top +Cascading is good software but it is GPLd and does not currently accept outside submissions therefore its incompatible with Hadoop from a licensing perspectiveIt would be great for Hadoop to have a more generic workflow scheduler that is Apache licensedAs such we should avoid a discussion cascading vs HWS +It looks like changes from structured model is not processed get param on +Ive been doing some searching and dont know where it is coming from +To see if the bug still exists I will out test again the latest build tomorrow and report back +Thanks for your contribution. +Closing released issues. + already has the base classes required for most of this stuff +PLUGINS JSON END +Ive tested this locally and this fixes the problem for us +Modified patch +Attaching patch for the proposed continuum ve been modifying the white site so the patch is outdated now +Making sure all Ps are at least marked Critical to make them standout compared to P and Ps +For example when you have class A and class B extends A and then you try to load class B through the +Anyone care to set a target milestone on this one +for the patch +A job consumer returns a result success or fail so we know that the processing failed or ve added a prototype of an api with rev A new interface is an enhanced version of the it allows to return a which provides more return information than just the state like an additional message etcA new is passed to the process method of it allows to log messages set the progress etcJust a first quick if you implement this feature +Ensuring tests pass should be good enough +Verified present in CP +But I am not sure if itself could get them passed or not +Yes in terms of functionality one can do this with a filter but it is wasteful to perform the validation calculation on the entire index where the candidates to be in the hits set is small +This seems like a crucial bit of syntax so Im marking for +Looks like he forgot to update the Jira. +V patch is based on the latest fixes the build issue for pig test +Attached from pluginmanager doesnt locate the plugin as the pluginmanaager looks only at classes from classpath which extend the implementation from in next attachment is ignoredto reproducemvn of not able to locate goal java inYour example uses not as you stated in the environments sectionSo which version of that plugin do you want to use? +Closing all resolved tickets from or older +you should be able to solve it with or without additional optimizations like this methinks +There are many such jiras created in Hadoop where jira is misused for asking questions +It is not very handy to use for someone just playing with +CR becomes GA so for now no issues to be put there +Ill put that up here +Daniel could you reattach the file with ASF license granted +This bug has been marked as a duplicate of +I tested with Flat Grey and from the change in your patch I wonder if its not another commit which has changed things +It simply returns an empty list +This is an expected behavior +Fixed as discussed +The code in my diffs check the bean as found this out the hard way +What about the changes in r? Do they fix the issue? +OK I will +Oh then its just a versioning issue were still building against +the simple workaround is not to click the transition linksif people in the community submit a patch for it they can reopen this case and we gladly incorporated it. +Thanks Shalin for pointing to which is very advanced solution to term counting approach. +The change looks good to me +In one scenario I have a web service callback to add ws security header to the outgoing message and in another I need to add a cookie http header +I can see that it would also be nice to have a etap test that checks that the suffix remains unchanged for multiple utcidsuffix +Users that had used the old ant script +The patch only partially solves the bug entry +Im uploading version of the patch +actually this is fixed in the assignment to was a typo. +Chukwa is based components Agent Collector Data Processor Node HiccI think that each one should have is own rpm and we should be able to install all of them into the same machine without any conflictbinconf could link tocontain files for all the installed components binconfThen we could have one directory per componentAgentCollectorHicc +That was as a result of we trying to do a cross region transaction kinda thingHere I mean the row locks dead lock +Perhaps inserting before each +Committed to trunk. +Would be good to fix for +The broker also logs an error of the form mech ANONYMOUS is too weak +Closed after release. +Finished a first pass at Greenhouse updates +fixed rev +The names of the job directories start with job +Ive inserted all informations about markup container and markup stream inside class +any objections to committing this oneI havent had a chance to try this patch out but if it works for you thatsgood enough for meIt all can wait +bob can we make the parser add an error like bad rule name or somesuch +Thanks Colin +I think its good that it can be checked +As part ofThe messages were removedJ J J The message XJ was updated to remove TYPESCROLLSENSITIVE or +How about this Stack +This should already be fixed in trunk could you give it a try to see if it also fixes your problem +Please attach a patch including the test case to proof the fixed problemAfterwards you might assign to myself for reviewThanks integrated to the master and closing this issuechangeset cecdauser DeveloperDevin Reviewersvanteschubertdate Sun May summary bug Test case for split fix given prior with revision changeset dfaecbuser DeveloperlucianadercoliReviewerweihuaWangdate Mon Jun summary bugbug fix for transform cell address string to column and row index +I reverted the commit +Does it make sense to factor common dependencies and symlink with Puppet in say the distro framework RomanSince we support multiple versions of Hadoop its unlikely we can avoid multiple versions of jars in the class path from tarball distributions +I hope this capability wont end up mixed with direct lookups like in Enterprise JNDI. +This will need a couple of passes through the build farm to shake it out +If we reset the size of that every time we start wouldnt that preventpeople from customizing that size then +Juergen file upload component does not require to be placed into multipart form +add code +There exists even script syntax error! Dont know how the code can be in the first found this issue in latest runs +Added the functionality +This is a test issue +Thinking about this again one could move this into a separate module and changecreate new build files that produce a Web ready to be deployed +You can still run the AOP tests locally by doing mvn install from common and running with resolversinternal in hdfs. +I think that would be somewhat useful for simplifying configuration although they dont mix well with +Let me forward it to them on listThis is actually caused by JDK Dynamic proxy scheme +Patch for branch attached +So I set the integration tests to the root in Rev +I have compared them and found to be superior in terms of match to current +For a temporary workaround set JRUBYOPTS to turn off JIT +Does having hints set such as axis order or standard unit influence WKT decoding +Its still missing a few things that I will add to on Monday but its a place to some more tests for param assertions +On a report we have code that will reject HRS with lease still held BUT itll tickle the shutdown processing +If your patch were instead organized to make as few changes to the code as possible and if these changes were well documented it would be easier to evaluate +This will match the network client behavior again with the logic being that it is simply a case where the columns are not specified so the key should be returnedThese changes would be backported to and then we would file a separate issue for the remaining clientembedded differences which can be addressed in the trunkSorry I keep switching back and forth on this but I think I have at least settled my mind on a sensible solution for +I hope I get to this soon +Mind went walkabout there +I will instead use a second variable to keep track of whether an is in use + would extend and obsolete this issuepatch as instead of reporting problems makeInstance should raise an exception listing all failing directories and the reasons and at least one nested cause +Correction that should be KB. +Ah thanks for the explanation +Please run the patch from inside the apacheds module +looks good overall +The branded EAP docs pdfs have been checked into the soa repository +Yes Mute microphone is the statusClick here to unmute hisher microphone globally +Just had a look at the patchI like things coming from devnull devnull Thanks H +Not required but a feature +I made the highlighter module depend on the query string parser and analyzer modules for testing +Hence my opinion doesnt matter +Ive checked that the codebase of Branch can be built and XML binding tests in the testsuite pass with the WS team should confirm that there are not conflicts and urgent issues to be fixed before we can close this includes fixes for the following issue NPE for enum type attribute that restricts an xsstring with a default jbas to jbossxb and the jbossws too +Must check if a delivery channel can wait for sendSync is used the copied exchanges are send to subscribers using sendSync alsoIn addition a dummy component set the original exchange to DONE and send it back when there is no destination for the message. +bulk defer of open issues to +Especially loosing the ns prefix scares me +create new class selects the src folder automatically for a project even if the selected element is somewhere it does listen to the selected element to find the project and if there are more than one possible output folder then the selection is also used for choosing it +sorry for the misspelling +If you commit after the close the test will work as expected and the first consumer is shutdown +It looks pretty much like a race condition +I changed the superclasses implementation to be placed a less public place plus not being called too often +You can easily get around stale documentation +Fixed +What use case are you thinking of? What deploy target is intended +Knowing your background on this project and its sponsoring project we will not dismiss your concerns lightly +Didnt know every test in a suite was running parallely +When do you plan a release and what will be includedDo you need resources and for which issues? We might be able to contribute resources for a limited time +Fixed in master +I also checked the j option. +I can make a sample code for easier there a workaround for this? Its a serious bug in that it blocks the bidirectional mapping on a joined subclass +Example helloworldaction not fixed still relies on queueB to be deployed which is not present on +I am more than happy to look into it if you havent already +I found that also had wrong cast +Im not sure how these components are used andor if these need their own build configurationxdocsDavid please clarify +Move to +OK +I think its as fast but even if its a bit slower Id rather have a slower clean test class than a dirtier fast one +Here is updated for the patch Vinay and thanks for the review Haohui! +In the mean time are there any known workarounds for this or will we just have to handle this on a case by case basis +Oh if you comment out then everything works fine that indicates the system is not detecting the EJB module correctly +Im going to adjust the velocity templates for the users to apply metadata to arbitrary text and existing ODF components +First off the installer for and use the same code base which means that uninstalling and installing it again should have the same effects as usingWhen installing the installer checks whether +Still valid +So not sure why Oozie is still having problems +duhh +If really needed I think it could be merged manually but wont do that now. +Ill update the issue again once I get the test import completed +afaict we are doing the best we can with at this point its pretty clear if this issue occurs an error is printed to the consolelog and the watchdog eventually backs off +using one side of split for both splitA and splitB and + +Hmm I got similar results from a crude test generating user vectors +Fixed in branch and trunk +Any chance of a unit test asserting your change does as you suggest? Thanks +There has not been anything added to indicate that is ejb is selected that jdk is required to run jboss +If a user logs in to bill for that and the organization is associated with the user then it doesnt matter who the service provider organization is the users organization will be the from Party on the invoice +If you remove it make sure that you update all dependent source code like pig seemed more descriptive to me +proto files between versions to see what new fields are added etcMaybe we need separate tags instead of just one incompatible change checkbox? Ive always understood the purpose of this checkbox to alert users and operators who are upgrading between versions of things they need to be aware of that can break their applications +It should be page size table space +This behaviour is not correct global modules have no relationship to AS extensions +Xiangtan has a patch +I dont see any good choice a framework like camel can make for you in such situations +I tried to get someone that did have an interest in it to test it out at the time but since no one responded I figured it was an improvement over no support at all if it didnt work in all cases JIRAs would be filed things would move forward blah ve just put up a patch for that should fix this issue +Rick the trigger test failure you noticed is related to my checkin yesterday for +Clover isnt in this category because it works on source code level. +So I think its safe to apply the patch +duplicated withby a to have a look at this aftercurrent trunk builds with JDK +Right so can we quit the discussion and just fix the bug nowWere looking at it now +needs a better ui to select categories maybe via trees and a way to select rules +Are you going to reviewapply the patch +The patch was applied at the revision +Forums thread added for discussions +the patch would be nice if it makes it into added taking boolean to supress refresh. +Have you tried this with +But after a couple days of testing it looks to be standing up +Fix confirmed after building against current trunk and apply that attd patched with git apply +and it is generally best if it is not NFS mountedThanks Hairong +was split up with most of the original code going into +Marking as resolved +Untested just quick code dump +I dont have a preference with rmr r but I believe some library would help us a lot here and in the future to be more flexible +I think the issue I filed would be a subtask under this one. +So tests wont find the bugOn the other hand we dont want to NOT wrap in it because some of that crazy stuff we want to always dolike checkindex on closeSo I will open an issue to clean this up +Looks good +Second core does in fact currently have the hbmddl tasks as well +Note that patch is relative to trunk +We have jarprovided on maven dependencies and compilation with maven ends with success but there are errors in eclipse project generated from eclipse plugin in version +But since I couldnt get the tests running I was unable to verify the fixYou thingsSee my test port simply adjusted your test to run against ArquillianIm trying your patch fix now. +cglib +surefire re concerned about the cyclic topology with a network of brokersIf a durable topic subscription is requested more than once on the same broker will ignore itthanks +Case closed. +Committed +I committed the patch can someone resolve this ticket? or give me rights to do it +All tests passed +I am inclined to agree if the amount of work to handle dashes during streaming is trivial +Well let others know what changes are made for every new patch +Is there a way to manually disable this check so the error doesnt appear +How can I make sure it is used when put on a private repository? Btw I dont think overwriting jvm default properties from POM is a big issue since that would be kind of hack and should be done on cmdline if for the patch and hope it will be included in next versions +But that means that I cannot open XMI files anymore that were created by another pipeline using a different type system +Please see the new patch file +I vote for proposal One more thing I think we should calculate distance once for each replica and use the value in sort +I dont understand how this is a problem given that the only supports GETPushing to. +I tend to think that it is the best option indeed +Like the comment states in I too am unsure if the preemption is translated into a kill command to the NM by the RM directly or if the scheduler is relying on the AM to see the finished container status from the RM and issue the kill to the AMIf its the latter then the container will be killed after the AM has already determined the container status correctly +My apologies I was confused by your statement that you were using the used Xalan and XSLTC there +As Sanjay pointed out the permission checker throws an exception with only the path fragment that access permissions allow you to see +Would it be worth splitting hints out into objects? Would be fairly simple in hintFor +That was the default option and I didnt think about it +See trunkbuildaggregate +version control setting is overwritten to no version controlthanks fixed there thank +Could you update it +Could you prepare an example Maven basedIve tested with your code snippet and everything works you dont have to specify at all by default Struts will convert and create List and User without any additional code +Sean please remove the section from guide and then set fix version for that to behavior was not implemented and postponed to future. +Id appreciate it if you could do that and close this bug +Could any security expert comment on this +Thanks for checking Mahadev +Would you mind attaching the patch to this jiraWill commit the change early tomorrow if no more comment objection +Got rid of asserts +s +Artifacts files will be created but this is not the same thing +Probably a commit made it workNow that it passes the next timeout is Is this last timeout related +We corrected the nameserver entry in so it should be fixed +Concurrency tests fail and Im not sure why could anyone take a look +Furthermore because the argument to has to be a completely static identifier for ACF its not possible to create a wrapper functionError thrown is thisComplex constructs are not supported with function +I answered quick because I wanted this bug fixed already and then forgot about it +Eli yes it seems like that +is this still an issue? or just a problem with the configuration of the repositories? in between and there have been changes in the repo configuration the managed and remote repositories have been separated. +Im going to close it as Not reproducible unless I hear otherwise +The is being taken over by as per previous comment. +No there isnt one yet +I finaly reopen this JIRA and add a patch to fix NPE when shopping cart have no product PierreYour patch is in trunk at you Pierrecould you secure your modifications with unit testsTIA +yep thats actually by design after the profile is created the view filters the list by the profile name so if you want you can click on the i icon and edit it +hey stefanI need to tell you this is great stuff here +Perhaps it could be logged as WARN? +So maybe the second test case I added is not that important even though it wouldnt hurt to keep it +Uwe do you plan on getting this into +See +First step is just to update the parent version + is the only method used outside of +If either of these locks can be heavily contended then it might make sense to run something which exercises this contention to make sure the change doesnt cause our latency tail to grow dramatically +AFAIK Hadoop has no concept of uid as everything in the HDFS etc is stored as a string +and if this task is longer than expected maybe we should process it differentlyrobert exactly +Andera navigation toolbar works with datatable and in datatable the number of rows is specified by dataprovider so it is known before render +I only tested this with HSQLDB so far +Test failure is not relatedWill wait one day in case Jon has more comments +Not a bug per Aung Maw +Thanks a lot for digging in that issue Michael much more than appreciatedThat is why anyway nobody cut and open a votes for a new fluido Id propose to not taking an action for that issue and wait for next bootstrap release before releasing next bootstrap version WDYT +but not in JBDS using the update sites above +Change made in Git ID fcbfefcfdfdceeea +A client side handler prints out the headers + +Draft version +but if you are just getting the straight string representation back it might not hurt +Changing it back to the way it was should not cause big problems for the users of this method +Attaching patch to fix links in the navigation menu +This has nothing to do with the command the command line is a one goal lifecycle so the executions have nowhere to bind and so they are separate +Thanks Killua! +where Nutch hardcodes the same field names +This patch was applied to only because of the integration need to support querying thru Teiid +Andrew might be Ive been building with the profile but running on top of a hadoop installation +It will not help us IMHO +I havent noticed this before. +I also ran and derbyall without seeing any failures +Thanks +applied to branch rev +Resolving as Fixed +with which build is this fix available +Tests are done well +Was is something like this you had in mind Rick +layout page? Please check that all resources were loaded properly on Net Firebug I still havent got the point about two lists +HiI have implemented this feature and attached a patch of the affected files above +Committed to trunk a while back +Verified in JBDS H Fedora +Are there any changes in using the library +Sorry I didnt notice the radio buttons when I first uploaded +By the way Is the absence of the file name and the line number in the message another bug +the import wizard does not show this shapes by default +Similar expression doesnt appear in the current sourceThanks again for your great courtesyMichelle Baert +I have other installed and none appear to have issues with spaces in the directory paths +It is now obsolete but I think using a value of to mean ignore the page limit and use all available memory could make sense +A funny related anecdote that Ive heard +I can not see what commits Alex did for this fix in branch +My unit tests are coming out fine and returning expected results +log file from the +Ive applied the patch +i love to do like that but one framework is depends on other is making me to push it up together +At this point it saves the version of the writeLock znode +It looks like we can commit quite soon so reverting this patch from will only make more work for usMy bad for not noticing this test failure on +Ive now removed the disabled parameter altogether but the problem is still there +Erlang can certainly make use of multiple cores when the DB is experiencing a bunch of concurrent requests and I worry that all the context switches involved with couchjscore might actually hurt performance +Cancelling patch +These changes look fine to me +Wasnt sure if we needed to discuss the refactoring that you suggested +Attaching patch incorporating comments +Ive updated the patch and it solves the corner the fix was so simple that I dont know how to write a spec about that so I wrote a test with those corner cases +I guess the experiment to try is replacing the Gant version in Grails with something more up to date +Is it possible to add that back inThanks +Hi AndrewNo I must change the source code for my case +back to you for the ZK stuff +This is the war file that gives the error +Modification Rollback previous two patchesTest Execute derbyall and found no error other thanSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaSending javaengineorgapachederbyiapireferenceTransmitting file dataCommitted revision +Adding reference to register portlet in and takes care of the problem +No source files are ingested because they are not there +works well thanks +Thanks Matthew that is very helpful +Unfortunately the only other web services plugin for Grails Axis also has a fatal bug that makes its invalid so right now there isnt a single usable web services plugin for Grails +If debugtrue the resulting classes are hot deployed to dev +it takes time to do the manual uploads +I agree with robert that this is maybe tailored +Ill ask around +Id like to see a compelling reason for the extra complication before pulling in thoseAlso it looks like the changes to Plugman are relying on the CLI folder structure which isnt allowed +In this case the async op when completed may try to add something to the executor and get a runtime exception +The namenode has not yet been altered to automatically empty the trash however users can manually purge the trash of expired items from the command line +Committed to trunk. +OK only one reducer can do this but I dont see another way +We are putting our main efford into JAXWSJAXRPC tools issues may handled by community contribution +I think we need to generalize the notion of an input ot output class to an input or output descriptorGotta run now literallydk +And vertical orientation + But that results in because in the existingHandler is properly started and the new one is notWould it help if we explicitly started the newly added handler +to trunk as dadeefedebca +Im getting the same exception +I dont think I have looked at it before but I can see that both and trunks are still failing +Heres the updated patch +for adapting! I have seen your fix in CVS also understood about the is a good sample app +is in sync with trunk +Will be part of In case explicit is specified and it does not exists procrun will exit with error. +Just a remark about +Hi JarekThank you very much for your suggestion +bulk cleanup of Jira versioning +Thank patch is kinda of a question +committed + for the patch +Attached Augusts patch from mailing list +Its too easy to add a field and have it show up in the serverclient thus my suggestion to separate the and types ala +Is this a blocker for given has been committed to +Committed to trunk. +Im not involved or familiar with the RSS wars but I will say that this is a quick example of getting RSS out of Solr in the easiest possible way +Forgot to add the Apache license header to the new fileIll be running and derbyall with this patch tonight +What version of Seam are you using +Ive checked the issue in Windows and ran times now sample runs might fixed when fixing memory leaks in the issue is still there please reopen the jiraRegs seem to say that this is fixed now +Possible solutions here When adding Models to VDB the is generating a JNDI name for the user +Thanks +move sybase issues to next as wont fix if someone run into this issue feel free reopen it is good to see a workaround is documented the system I am working with requires me to build a custom query per request +Thanks +This definitely works in trunk and some portion of that fix was ported to the codebase +Olaf should patch his local copy and wait for +Patches that fix the patch +Hi ClausSorry about the earlier patch +a way to specify inline images first then the text with the references still getting resolved properly? have no idea i havent used inlines before +It was just quicker for me to write a benchmark and Akka was already bundled +click on Test Webflow +Forgot to mention that the wiki was from the SVN trunk +Heres the file that handles the spaces +Ive removed the limit +With Ruby behavior where each eval gets its own scope this would not be an issue +ftl screens show dates in the current sortable format It can cause conflicts when form and as the java methodminilang will surely only recognize datestimes in sortable format +Why dont we stick with itJing thanks for helping out here +This jiras title mentions but you changed is it what you meant to do +Please review both +And also move optimization over to a new issue I want to go test more and maybe add a cmd line flag to enable the optimization +This is because of the different transactionalsemantics provided by the two possible modes of execution +This issue still happens with SVN dont know if it is an issue on SVN or on Maven side but the issue can be resolved if we tag from repository instead of tagging from the local is an open issue on this subject just after posting my comment I noticed this was already fixed in I guess the next step then to make this official is updating manager? Anyone know the plans for this +put default view ids and belong to the remove setter methods for parameters that shouldnt be changed by application Request Phase save state etc +Thanks +The warning is still there +Commits for the PR updated +bq +Since nobody commented or reopened the issue Im now closing it. +Testcase added with rev issue got fixed with various changes done as part of where changes in different session within same thread get reflected +Committed +We might want to have this feature configurable. +Nice start +Thanks +Is it ok if I commit Henris patch +Jakob has suggested to use s expected notation instead of the trick with assertion +Thanks +Now you have current B and null +patch that just adds a Test case for Configuration in order to test the changing of default values for minOccurs Nillable and +This duplicates +Hello we run into the same problem with DB +After sleeping on this I had second thoughts about the original way I did the patch +The first test still uses mergeSort +Not for commit here. +Latest patch didnt apply cleanly to me so Im uploading what I have committed. +Im testing on trunk and it worksThe ticket is duplicate of. +Committed to trunk + +random uuids generation is the least surprising of the three behaviors so it heads off interminable why do my ids look kinda similar? discussionsIs this going into +Lets transfer this to the mailing list first +Can you think of any ways we could get tricky at the HDFS level to prevent from deleting files we need? Otherwise we just add some logic to this special mode to prevent it from doing anything like that +Makes a few changes to code and how filters are run +They do not have to end in conf so that should be up to us +Thank you for cleaning up this messThe patch looks good to me +First cut patch +includes an example of what it would look like if properly handled its input as a graph +GuysUsing unique on each box resolved the issue +I mean here putfield to final fields here +differ from current log? All it has in it is events no? Regards your question about process Id say no need of an issue per metric +Confirmed still an issue in Apache Flex in previous version or in Apche Flex develop branch +To keep track changing the values didnt seem to influence the test at all and it failed +I just went and grabbed the fantastic mr +Since this requires a KFS deployment this testing is done elsewhere +Thanks. +HiMade NKF to decode all diffirent parts in a string +Its currently stating that it can handle static resources but you end up with a with the components parent class and provided path sticked together +Heres the patch to take o for the scripts +Just moved the DTD folder out of the way +Will registered editors be applied in this case as well +No need to keep this bug open +Could you rephrase +Can you please attach a test case that includes a sample XML document instance and the Java classes required to unmarshal things successfully + Enterprise Application Platform Runtime properties selection of Connection Properties Connection Properties Driver Properties get message box Connection Properties Connection Properties why same items +Thanks everyone +Some comments about I propose to add a second test query that switches the order of param and param but keeps the order of the actula parameter values +Do we just want to add the functionality to iterate? After all SQL is and Java doesnt convert that so theres precedent for supporting both + + to trunk and to branch +RakeshCould you please convert the comments to javadoc and at unit tests before we commit the code +Part of the problem is that we have so many versions +Thanks +please set fix versions +verified +Can you provide a test URL that we can verify with? verification +Assertions are now +I fixed this to be +A temporary fix could be done by using the configuration directly instead of asking the region server +Sounds OK to me but Im only looking at it from a console perspective +Second PR to fix remaining applied on master and branches +Pretty nice. +Patch was committed to Subversion trunk +Merged into master branch. +I plan to commit this later today or tomorrow morning +Finally got enough information to figure out whats going on with this problem +is going to fix those since it added those codes in its patch +Moving out to move back to if this will be work on soon +Objections +I am looking at this now and will have some comments over to you today +Instead they should use to pass in a mock to the loader which will replay parsing the SCDL +patch applied commit no +Tested manually as per description and it passes for EAP +Patch integrated to trunk. + pom same except version Ive uploaded those and +As of the sort cases are fixed +This is a very good change for the stability of the system +If other requirements will show up then they will be tackled in new issues. +is returning null? That doesnt seem right +Sorry bout that +I have applied the patch also fixed the exception type generation +Should the contents of connectors in JBT Central and JBDS Central be the same? In the past weve had +Does the revert for meta tables only +Mostly done +Is it intended to support jars from different Lucene versions? Would a unit test for this project include old versions of jars retained as binaries +Another for bundling the database with LCF +Attaching a patch for this issueI moved the related tests to their own filesA javatestingorgapachederbyTestingfunctionTeststeststoolsA javatestingorgapachederbyTestingfunctionTestsmasterAnd I changed to only run these tests if JDBC is available +heres a prototype patchMike tested with his wikipedia testrunqueryqps oldqps new diffunitedunitedrunqueryqps oldqps new diffunitedunited +added log as we delete them after some time apparently +It looks fairly straight forward +The initial then allows the client to switch on any number of options incuding the tightEncoding. +Here is a patch containing my implementation of the +Would anyone want to try this + committed to trunk and branch. +The deployer has been split into two piecesdeployers various metadata related deployers that produce Temporary workaround to support controlling the lifecycle of the webserver runtime portion of via a JMX service in the deploy directory +Patch attached +To be honest Id like to get this committed +Jesse Are you referring to the fact that we have to add these magic lines to each test? Yeah there must be a better way to do that +Thank you for applying this patch +oops +Yes everything works in old unsupported you share the dataset and the requests to help us reproduce +The patch didnt apply cleanly because of a recent commit I did +communicationtiming problem +hibernate tools jar was in JBT Linux +I have never seen a CVS module likeI hope you referred to c +tests JSFWebbeanJPA Scope Annotation Inject annotation events interceptors decorators +Here are the results for the slow parts of the project +At this point there hasnt been an outcry for the petstore dashboard in the distribution so Im going to resolve this unless there are objectionsWill also remove the reference from the petstoreWeb documentation +ScottIt will be useful if you can also paste the Pig script which produced this exception +Hi Alejandro I need to rerun the failed by rerunning the main workflow associated with that subworkflow +I tried this but get is declared final in you cannot override in +I have attached two modified classes from the wicket examples +Sure Suresh +files +Thanks! +Id prefer that we catch at the top level not print anything and exit the thread +Added in dependency for IBM JDK builds. + Looks great for me +If you can just validate that the works properly after the refactorings then we can get this in +Hive QA wont run try again +The solution proposed by Thejas does not work directly because the is greedy but I worked around itI think it is ready for review +Note In order to use global temporary tables in db I had to first create a user temporary table spacecreate user temporary tablespace MYTMPSPACE managed by automatic storage +Try disabling jline and see if kill works +See test in patch for more detailsI just hacked some code in the class +Darrin this is really good +thats the dreaded signal segmentation fault crash +Resolving the issue please if you still run into the to reopen but Im still hitting thisI think that the problem is that Im hitting the I wont be the merge leader bit in MERGE +To run test applied at . +NickPlease to me if you think I need to do something more +With this Task notification job has become never ending jobHence reopening the bug +this issue be closed? Whirr was released with a dependency onIve created asking for an upgrade to jclouds ideally in WhirrOr is the intent to keep this sole issue open until there is a whirr stable release that depends on a final jclouds release +Resolving as a duplicate of. +should that be a new bug since the original was actually fixed +If the data is going to be in my comments regarding the approach for this patch are not applicable +Do you work on the HEAD or on the branch +Could you try it with both settingsfor this property to see if the behavior changes +The issue was already resolved in master so it will be picked up in the next release +Development is continuing on Windows platforms +Try the query more than once to compare performance +Yeah I do see this in eclipse +Hi Matthias +But I think the check for both properties absent should also be there right +If you have your aspect class implement you can look up the DAO by name this is not ideal but it should work until I get a fix into +Ok were all done here +We were using and only k concurrent connections created by +I am only putting it up here as a first cut of the abstractions I have in mind so please provide feedback based on that especially in regards to the class +How about this +Since this is a decently big checkin I will create a branch with the name in which the work will be done +This is now enabled by default via the flag on the +Either way I think a new JMX statistic should probably be in a second ticket +Im not sure the shading will work yet either +Java class to run Weblogic goes in integrationantsrcconfsshare goes in integration goes in integrationantsrcjavaorgapachecactusintegrationant realise now that the is not actually needed reference to it could be removed from the java class I providedLooks like Cargo is the way forward but we needed a quick fix so writing this was the only answer and someone else may find it useful +in revision did not use the latest patch which was simplier and had extra checks? The am pm problem again +Resolving as fixed +Hi again +Done +Hairong will take the lead investigating thisDevaraj thanks for the clue +we dont have to worry about it +Seems like they dont notice comments on WTP issues lets create a new one +Thanks Jing for the review and ve merged this to +And I verified your changesThank you very much for your explanation and fix Eran +I just catch this in my application unit test code and this assert is valid and good for robustness +is obviously not suitable as it is the internal stuff of RM +I have added parameters to the BIRT Hibernate ODA data set and committed my changes to the trunk +returns null when there are no tablesI assumed this was expected so handling the null is needed +I believe updating the dependency and leaving the generated code wont work +blocking? AFAIK we can still release using svn +Our URL plugin simply allows you to use the page title in place of the partThere are no plans to allow multiple to a single page as this could cause considerable complexity +Dhruba the above patch included a small unrelated change in +Address Brandons comments +Possibly we need to recurse into child exceptions when doing this Im not sureWe dont currently have a unit test for this so Ill leave this issue open for now while we work out if we have the best route and while we sort out tests +Can you please try to ping host from your Linux machine +Im kind of surprised at this result +Thanks a ton Ben! You are a lifesaver I was really getting worried about this one but couldnt make it happen in my own tap example app +How to workaround this problem using default configuration Stan +r. +remove the list from and add more counters +Its a bug of can reproduce it with New Dynamic Web Project report it to Eclipse bugzilla +Joaquin could you please explain what you mean by Saturate the effect of frequency with k? Thanks +I will have look at it tonight +A few use an embedded ftp server like +Thank you so much Army for the detailed write upAfter reading it a few dozen times I think I have a much better understanding of what is going on with the modification of the child result set +This is going to bite other projects Im sure so this needs to be fixed +AFAICT plugin works fine. +looks like it has been been fixed in the trunk as a part of feacdaaefeHence the patch is only good for Hive we target it for instead of simply closing it +Regardless since all of the examples above are shown using an E escaped string literal the clients are not relying on the behavior as the E literal makes the escape handling explicit +Otherwise please close it +Hi Guys Are you planning to release this issue inIts a very important issue to me +Changed to pass number of items per old resolved issues +Issue was the culmination of a very intelligent conversation that took place on the dev mailing list +Windows server is better than windows client for extra fun +Both methods use the same implementation to obtain a context with the users credentials +looks to be the same problem with creating directories as +INCOMPLETE french translation +Corrected patch +This issue will prevent you from seeing the actual exception being thrown. +Thanks for the followup on this +Im not familiar with a fBSD jail but whatever works Ill learn +Fixed on trunk of by adding back the former check and now also searching for both tables and views in the call. +Like page actions these are conditional page navigation +The other issue is how do we do regular testing +verified +Yes only to trunk +was the missing JVM option FrancescoGlassfish seems to be working now as good a tomcat for the stuff I have testedThank you and Massi for your effortsrgds close for +Looking at the code it seems to be also confusedThe javadoc forbq +committed to can you please suggest how to verify this change in the UI? Or is this just a change in the logic that cannot be easily verified? Thanks +Thanks Jimmy. +core test failure is a known issue and is not related to this patch +Thanks for the merging across I missed this method and missed catching it because of my local env +Then the child will register its own styles +All it knows is that there was no matching handler method ll revisit that in the context of Spring s general REST support which will also refine the matching should work now +Which perspective are you talking about +Please take a look atll have a look at this but it will not be able to make it into +a ditto patch as it looks like the QA bot didnt pick it up yet +I am running it through hudson just for sanity check +for sure it isnt that nice as wellfor simple like primitives dates +Franscisco has performed the upgrade on both Branch and HEAD. +You guys are familiar with the log stuff is it doable to get in a fix to drop the buffers for +Ah yes +Tested and committed +Teds proposed patch extracted from parent +Almost two years later I ran into this bug again in another project +At least I think all other failure should be pinpointed at injection point accuracy. +Ive also attached a copy of all the surefire reports for the module where all the errors occur +What I should do in addition is to investigate whether theres an already existing builder configuration property that I can +why do you want to move translate from varwwwtranslate to xWhat is the purpsose of xCan you summarise the disucssion you had on IRC as I dont remember being part of it +Ive got the point +Committed changes and updated the site dir at hopefully it will sync to the real site soon. +This patch fixes the issue +The conflict handler was initially exposed through the Oak API and only recently moved to the SPI +Dropping priority since this bug has not +So there was no report created in targetwebtestThere should be a stacktrace logged somewhere I would have thoughtIm not sure why installing would fix Webtest very strangeIf you find any more information please this issuethanks reopen this issue becose it does not workI uninstall the Grails Functional Testing plugin and the webtest does not runIt was no report created in targetwebtestOnly empty dir and dir resources with dir images and files +Hi Robbie I agreed with your changes and have applied the patch +It maybe useful as a demonstrative example as well as a job that does nothing but sleep used for manual testing of some code framework +Tests pass locally +Its also th in votes for major issuesI noticed your workaround Vladimir in the forum +We have xml and dtd checks in automatic tests so their is no need for that in runtime +So this issue isnt specific to this parameterAs Andrew suggested on the mailing list I think making all the parameters by default and the platform checks is the best route +claims that they are upto X faster then DBCP and CP +are not deployed to the serverdeployquickstartESB dir they are deployed to the database. +Added a test case and I couldnt reproduce +Thanks for the information but Im not sure what youre saying +I will give a try tomorrow morning +At least with the entrylog approach the coupling can be hidden behind the interface and the Journal doesnt need to be touched +bulk defer to defer to +I committed the one I have locally instead of the one attached because it didnt contains my changes about plexus removed too the use of analyzers in Jar class because it isnt the best place to use them +I would rather stick to less libraries and use Commons Lang +Youll get nice integration with IDE if you use maven plugin for instance +Patch applied in trunk. +The wisdom of its usage is up to the user +I will create a new one +Thank you for the comment +When we go to do that we get +The key expressions are visited as a field if it correctly maps to a field. +to only appearing in the context of that perspective +Verified on ER +bulk close of all resolved issues. +Committed to trunk +After talking to Raghu and Nicholas I realize that a datanode does not need to read all bytes of a rbw replica to get its valid length +Patch checked into branch as well. +Reopening so can submit improved patch +These jars are added to the project by a special GWT SDK classpath container theyre not present in the project itselfThe local deployer traverses the classpath containers too and includes these jars in the deployment artifact +Tests still run fine +more files which need to have relativePath removed from them +Going going +v attached that do that after the call to get +What exactly is the exception that youre getting there? Also wouldnt overriding register do the job as well? We are catching there already so I wonder what the actual problem could be +less than coverage? In particular I use it to check that orchestra is still and compatible +Sadly there is nothing we can do as the Samsung audio recorder never calls on. +Mmmm +Finally it includes minor edits of SKMD in terms of the output format +Patch committed to trunkr thanks Viola for the patch +Multiple classes may be defined within a single +I suggest we close this issue with no action if we want to keep it it should be opened in WSIF not here as the code will go there not hereIve been thinking about a WSIF kinda thing which takes Axiom and WSDL into account and provides the same function thru the Axis API +Looks goodCommitted to trunk. +Having a datanode refuse to start in a secure cluster if not running under jsvc is the easiest path to not accidentally getting one running without it the current behavior is fine +Includes api changes for upgrading to +We dont seem to be able to reproduce this issue +Know nothing about it nor I use the assigning to Mike hopefully he knows a bit more about its my Latex setup on Linux during the PDF generation I get a bazillion +Hive now only support hadoop +Justin I might have done more analysis but I have a lot less experience of real use in this area so I am willing to do whatever consensus dictates +Okay big mistake on my part +The bug I was testing when I found the bug is actually not P open bugs to P pending further review +Then you wont see any warning +We will have to keep this a sequence and be careful while adding a replica id to the assigned replica will iterate the elements in the same sequence they were inserted +Remember that old options will be honoured for backward compatibility with threat of them later being deprecated +It should not be a problem now that weve stopped supporting JDK +The bug is marked as fixed so you are likely having a similar but different problem +did +Kara Thank you for confirming it with which means this issue isnt stale and we should look into itSide note CR release was delayed from earlier this week it should be out soon +This is a issue +patch applied. +Whilst you are in the code any chance of considering both have patches attached +Based on my unit testing reversing the order of your dependencies in the pom should work around this until +Implementation complete modified method + for the patch +Moving out to for now which means unless someone moves it back to this will not go into. +Sorry if this is a dupeGdevdrubydownloadsjrubylibrubysiterubyrubygems warning already initialized constant Cachein convert Could not initialize class fromin check fromin is fromin handle fromin interpret fromin interpret fromin INTERPRETMETHOD fromin call fromin call fromin cache fromin call from binscreensaverin file from binscreensaverin load fromin runScript fromin runNormally fromin run fromin do fromin internalRun fromin run fromin run fromin main +The concept and the patch look pretty good +Ive made a few changes to the Web UI to show more better info Pools now have their scheduling mode and fair shares displayed +Committed to trunk and +Isnt it +This is the visual result of the exception +Bruce can you attach a patch in terms of what youve got so far +Hi James +Hi MarkI cannot do anything against this but I can still say that I dont agree with you +ok just committed a fix and added a test like this to for Intersects disjoint contains and making the above change fixes the build in trunk fully +To do that we need to collect the largest timestamp among columns in a CF and then delete the CF with the largest timestamp +went a head to upgrade per and its associated plexus IT test are now passed on both windows and linuxRevision Author dantranDate AM Wednesday December Messageupdate to fix space in path issue which only found on linux +Committed the patch +Please dont raise multiple bugs for the same problem! This bug has been marked as a duplicate of +What a helpful comment +If the tests can avoid that its better because then they will all work on too +I tripped on this bug last week myself and its great to have a fix so quickly + to trunk. +Is this something that should be resolved in +Patches that fix some of the problems in the junit test +Why would need to get restored fromI reopen because it might not have a valid solution. +In that case Id close this JIRA as wont fix +I think the ideas to clean up the API are okay +Also please make sure you replace tabs with spaces +I hope that you wont be against +Id love for someone familiar with to give this a review +It looks great +If so this issue will become relevant again +That work can continue under +The dispatcher on makes implementing reflectors quite a bit different so I dont think there is much gain in spending a bunch of time synching them up on. +I am having this same problem with version I am trying to run init and I get the above error and stack trace +Nor for +This trivial patch fixes the problemCommitters can you please review +We just need to confirm and close +Closing tickets +I was able to fix this problem by creating a bit version of the file + Changes to incorporate comments from Richard +I suppose that SEAM should give more descriptive error messages. +Maybe Ill try find to give it a while locallyOne major change Ive wanted to propose for a while is a move to a data message paradigm +current system can skip unknown structsPoint taken +Just an update that I have this fixed in Plexus the wrong realm was being grabbed and so the right place for the components wasnt being used +this makes sense because if a url is not a child of that url space wicket will not be handling ithowever error urls are an exception +What output would have expected? those further examples to the +All databases pass this test. +plugins worked in support in plugins has been removed +The namenode should ensure that datanodes match the namenodes build version if the value of a configuration parameter is set to true +Per your request from the mailing list +Thanks Raghu and Michi. +So these imports can probably be couldnt find any unused private methods or variables though +checking branch soon +Mailing lists createdmoved +untested +Ill exclude them for now and try to find a better approach. +Will not be commited to to M +If this is intentional then it doesnt make sense to close the issue +Unfortunately this was required because we have a circular dependency between Seam Security and Seam Faces +Should read. +This will only execute non scheduled Asynchronous calls so Ive added the ability to fallback to another Dispatcher for scheduled calls. +See commit +We need to summarize the results and document the state of this we need to reach a preliminary conclusion for RC here +So I am resolving this issue but please reopen if you are still having a problem compiling. +patch was applied. +Yes youre quite right actually +We have to add the missingdot when cookies are parsedcheers Roland +Im assuming here that nanoTime is not affected by ntpd adjusting the system clockI think thats implied by its design +Some things I am wondering about +One still has to do with the fact that events are still not delivered synchronously since theres a thread pool in the dispatcher +Meaningful words and highlighting have been checked +test case added and now it also fixed the original failing code that I had. +Bulk closing stale resolved issues +Sorry about that new to this +Not required anymore see Richards comment on linked +I had this using the JDK Windows and also under linux with the JDKOn Windows the memory usage went up to MB before stopping +The text and shouldnt appear anywhere even by default +OK +The fixes seem to have fixed Derby performance too! I still get occasional output for the stuffer query but on the whole it looks goodSo Im going to resolve this ticket. +this is certainly handy but its a bit slow to do every timeWorth thinking about some more either an uptodate or a separate target for this might be worthwhile +See +NCan you add Apache license to the new classes? I used Jenkins time to do the split +Confirmed bug +Tests dont fail when run individually but if all the tests are run then of those fail +Id hope our plan is to move towards a standard wire protocol either adopting an existing standard protocol or developing our own interoperable standard +Somehow the RB integration is borked +I took a look at fixing this but its a relatively major undertaking +With it still does NOT work without the is sort of nasty to would be very nice if this was really fixed in further releases. +Looks like we need to alter the patch a little for since the federation stuff isnt there +I have Reverse Proxy Filter enabled at first +The mem leak turns out to be caused by the MRU cache manager in JDBM +I cant think of a solution that really is ideal there +From the global administration it does not save the JDBC Driver and when you Edit one it saves it a second time so you have multiple on the build plan when you select sonar profiles it would be nice if it switched the text below to reflect better idea is to have options one to select a predefined sonar profile and the second to override it and then you can edit the Larry thanks for the feedback +Created in atsintegration +If were serious about running arbitrary Java programs we have to allow suggestion would be that we fail the setAccessible for all +And that is what is broken +Wont fix +This work is now done +I believe that we need a goal to remove jar from localrepository and also for repositoryyes for SNAPSHOT I do find that some release plugins that are havingdependencies on SNAPSHOT version and very often that SNAPSHOT maybe broken or no longer exist +Working thanks +Yea sorry for mixing the up +Confirmed same behavior on HEAD as of to the preferred replica tool tests that you are doing +For example To create an encrypted db encDB using an external keyjdbcderbyencDBcreatetruedataEncryptiontrueencryptionAlgorithmDESCBCencryptionKeyNote If you lose the encryptionKey you will not be able to boot the database +This issue stands for some time now +I tried my best to follow the convention about what exception to throw in UDF +I am pretty sure that i am not behind proxy +FixWe are just extending +Trying Hudson again +So unless that default is changed calling the option unsecure is a little misleading when the username and password go out over the wire as clear text +Major as there is no workaround +Trying mvn install execjavaworks but the README says mvn exec like somethings amiss in the dependency tree or the execjava config +Do you expect this to show more perf gains along with your patch +Norman will contact the instrument scientist for the list of devices +It works fine for JspressoLet me know if this will be part of next snapshotmilestonerelease for the patch +Hmm that is a problem i think the message corresponding to the seqId from the acks may not be in the store because it could already have been removedI think the brokerSeq should be set to the seq in that case as a fallbackOn trunk there is a null check but brokerSeq defaults to which means that ids will be reused in a failover message producer this could lead to duplicatesI think trunk needs to default the brokerSeq to the seq value returned from the storefyi the root problem is that the brokerSeq incremented when a messages is received cannot be used as the orderIdseqId in the store because they values get out of sync when transactions are used commits may be out of order to initial sends +I had some more strategies that worked for me but werent complete or perfect +Ok so currently it is something likeOperation HistoryExecuted At Operation NameStatus ActionsOctober PMTest ConnectionSuccessful Show Details BelowAnd then upon further scrolling down it saysResult Yes Was a connection obtainedThis is ofcourse on a successful runLet me know what you would like the initial Successful to say instead +I guessed there is something wrong in jira +Im wondering if its possible that you can move the following code inside unsign itself immediately afterI thought about this too +This combiner is same in concept to the combiner we use for Hadoop jobs +Just realized the subject of the third patch is silly +Because mail datasources and jms resources have a configuration aspect of a service that a URL binding does not +The section ends with a brief description of what is presented by each of the Working With Derby activities +Other methods such as flush are also affected +If this is happening on successful runs something is broken and should be fixed +should be amended in the same way as +Thanks Ted +See +I would vote for having a higher priority for this issue or disregard priority formality and just get it fixed +Ive deleted How to use jsFunction with JSON +So the problem is in release plugin +Which is very good thing that can help Sqoop be more successful so I definitely support the idea to natively support Windows platformI do have concerns about destabilizing our stable branch as the patch seems to be referencing files that are not inside the patch and it seems that were planning to add much more related changes in near future +Included in Hibernate Known Issues in Release action required Strong +I hope I wont forget to do it in the future +its not just about the change but none of the async methods throw exceptions +There are two ways to do what you want make your template use quiet references like b which prevents b from being rendered if there is no object in the context for b Make a little context wrapper that will return for anything not found in the context +I had the one before and now its workingagain +I add for development build +Ear to reproduce the problem +The issue will be closed in weeks otherwise +Needs to be fixed for close of all resolved issues. +Yea Ive seen one fix that works with HTTPS but nothing so far for regular ol Andrew where would I find +This is what I committed +I tested the modified Makefile on my bit laptop which successfully built only the bit native libraryAlso I added you to the Accumulo contributors group. +This is not a Tomcat issue anyway +This case and some small other checks were added +Yes I need to update the unarchiver version which brings in a new container that needs a new harness so its on my roadmap to some how deal with the unit tests in a more common way +I did consider checking the types before trying the array copy in which case the issue would not have arisenHowever it should be cheaper to wait until a problem occurs and also slightly safer in case the check is too strictIll update the code shortly +It is not perfect for device but should be improved +I may try although I hadnt use git before give me the git address plz +Especially when you pause the jobsBut I also think having a simple option in makes it much easier for end users which may not dig deep into the many quartz options and instructions it providesI would argue that the skip should skip all past jobs +typecss or custom parameters if I add any +Thanks Ted I was not aware of this tool +BTW the version of this artifact appears to have the necessary VPP deps as the cgen ant task does not fail +The new comments do indeed make things clearer as does the renaming of the two methodsI vote for committing the patch with this additional javadocespecially since as you said its solid and fixes the identified issueThanks for addressing my comments +Verified by Alexey. +Thanx a lot! Could you tell me how can I download source or binary of OM +I have been running my projects with many dependencies on my windows vm to try to match the initial reporters setup but have not yest run into the same issueIf anyone watching cannot give a working example of the issue I will close the issue +Resolving this issue for GA now +Lets finish this up and put it in needs oAuth to be implemented to work have a twitter widget but not one that supports posting new tweets for which we really need to wait until Twitter supports API is a moving target +So we cannot tells whether the failed fault injection tests can fail the build +Please assign this to me +In addition to bug fix there are few updates in comparing to snowball stemmers in Lucene and Hungarian stemmer was added +this imports the Supplier inventory info +Signature PADES using have retested it with your last comments and it signed signed a second time by Itext all is signed a second time by issue on you attach the class you are currently using with the made changesoptional it would be also good to see the double signed document that the pdfbox created +Yes it works ok on my notebook with original delay +Your interactions with other committers and conduct over time speaks for itself Im not going to code battle you Im just going to express my opinion +Would it make sense to remove the extra method completely from and simply do all in one method? good idea! Ill rework the patch +This is probably an importer issue as its likely mapping the OBJECT BINARY column to the Teiid string type +This has been completed. +hey there was an umlaut +I dont really think that asking people to look in a log for an error is that huge a deal vs shaving seconds off of mttrI can haz both? Should those two things even be related? Or could we at least show this line like we did beforebq +I thought that a way to have a similar result on a finer grain was a good idea though +I continue to see this problem even on revision of the branch +also similar user threads at forum was posted before forums down +What exactly is the issue here +Preparing for release +small sharded systems can afford the second call while large systems should avoid the second in + to trunk + was a different issue +The internal representation is a BSP tree and one of this part of the outer boundary creates an hyperplane that splits the inner triangle + + +Pushed to master +I committed a patch to implement this approach for JRE package version numberingWe actually cant use +Closed +For example we need to the ZK state for replication +Yeah thought it could go into +I attached screenshots of my tests +had this too +Without changing the NodeRawElement classes I couldnt see another way to fix it +This issue has been fixed and released as part of release +This tool is useful but as Nick pointed out we should really be using es for test synchronization +Looks fine to me if it works +implemented verify uri is already declared on the page but with another prefix there is no quick fixesI think there may be quick fixes suggest adding one more namespace for the uri with new prefix +Changed title to better reflect the outstanding in r. +The result of this isa new session is created for every request until they timeout +Closing out the issus for the release. +committed this +Verified on stage. +Patch is long stale +The memory sensitive Lru and the tester class still working on some details about thesize of the memory used for different classes and will update as soon as Im done +If there are no objections Ill commit these in a couple of days +Also I think we are talking the wrong approach on this +And how do you create staging previous? By moving staging currentBTW you mentioned has problems +Sorry about that +I just committed this. +It seems service should be invocable and you shouldnt get at all +bulk defer minor issues to +Here is another relatively small patch for the Admin Guide source files plus the map file +If charles proposal for Kernelruby is accepted then the default should be true still? Of course until then I think it should default to false so we finally pulled the trigger and made this change. +It seems it was an on WAS which has been corrected on WAS! +JohnI cant reproduce it in Windows +Ok so we ignore data loss +Doing an explicit levenshtein calculation here sort of defeats the entire purpose of having levenshtein automata at all +scheduling for update +I added simple patch +Do you have an ETA? Ill be happy to monkey apply it on my cluster and be the guinea pig for it +Doesnt this fall under the land? Is this open to guards that catch and throw up appropriate messages? Im fine with that just asking if that is the goal of this JIRA +Thanks very much Andrea +WDYT +bq +too bad that we didnt get this in sooner looks like we missed a translation effort +Seems like were putting the wrong name in the frame for this +It might be addressing or fixing this issue +ant clean test Dtestcase pass for me on trunk +jar files using git and maven +This issue also affect the commonslog for this issue have been applied so this issue can be closed +It usually recovers +Committed revision +Previously it wasnt needed because we never actually did initialize class with security turned on due to the circular initialization bug in +Fixed in trunk rev +ok Ill leave open for the cancel build button bit +looks good to me +The attached patch sets and in the test when running on platforms +bulk defer of featurestaskswishes from to +Ive verified mvn clean install +Anyone who wants this fix on should comment +test m not quite sure that statement should be independent from each other is correct +Improvement to the exception committed to masteraed and da +Logged In YES useridAs it seems to work I close the issue +Unless they are daemon threads this will prevent the timer from being GCed +Verified on JBDS. +This patch adds as a contrib project into Solr +When u do this please move the single line from to the new thanks deepal +We already require protoc in order to compile Hadoop so there isnt any need to avoid it here +Its good that the position increment gap is always applied now but there is some other increment logic in this class that foils my attempts to coordinate the positions across multiple fields particularly when the first value is for a field +If an error occurs during bulk import all changes are rolled back +The only is to replace the Hibernate entirely and remove completely +Commit please. +Run nodetool scrub on both nodes +One for and one for +The CVV gets its own response which should be stored and potentially acted upon I think +Dropping from the schedule as this is not a current priority +There is PR in wfly in pool +So will provide a crude workaround for now and this feature would be in at the earliest +Thanks StefanYour patch is in revision +Moving to work on is closed +Generally not they represent code cleanups rather than bug fixes and only gets the latter + +Applied Jeff Browns patch to fix the issue +Probably my attach here a new version with unit tests this time +Ok I have committed the perf tests in revision +We should also make sure that in newer version tests pass in minimr mode too. +Add David for review he implemented the current version of the legend ability to filter legend layers by decoration m not able to apply this patch cleanly on trunk or the branch +With this one I couldnt get it to work +I want to confirm that this is happening while the application is running in production mode not development mode +Thanks Robert! +If it adds some speedup to server itd be a nice addition +There is in fact an error in the log but just because I added a dependency on the datasource +This issue has been fixed and released as part of release +Priority is lowered since it has almost all seam distribution needs seam tests +Please Pierre if you want to contribute then try to understand the community fundamentals first +Right patch this time +Were there any error messages logged +Ive updated my snapshot and recreated the patch +Thanks Mike +If you see the It has the targetNamespace +Having said that i dont have an idea what could cause TC to fail without testing it. +At a glance I think I like the idea myself +Attaching an updated patch which includes a test case that verifies that the engine can be shut down even without the permissionThe command svn cp javatestingorgapachederbyTestingfunctionTestsutil javatestingorgapachederbyTestingfunctionTeststestsengine must be run before the patch is appliedWhen to older branches the test must be disabled on because it uses + +replay buffer on the mqis there a way to say hey Im brand new +Now the container and the compose stack are got from the Request objects inAll the methods that get compose stack and container from objects other than Request have been deleted +Well considering its been tagged a major issue and slated to be fixed in version it would most certainly seem to be a real code defect and not a misunderstanding. +Currently the count is now incremented upon lost nodes but is not decremented upon rejoin +Great +Thank you Anil. +Any additional information about this one +It also added support for integer type columns to +Can confirm that the fix above works +Components used by the admin consoleajcommandButtonajlogajpollajrepeatajsupportrichcolumnrichcolumnGrouprichdatascrollerrichdataTablerichmodalPanelrichpanelrichsimplespacerrichtreerich and Alex completed this investigation and have created jiras related for that support +Are we assuming that the admin console can only manage instance of? If not then under in the tree the next level should contain a node that represents the instance and its configuration tab should contain its configuration info +you for the review +Chris this is a good idea to keep the modification date as just the long which we get from the java file API +Could you please explain the problem and copy the jsp tag source which you are using for this +Added support for two currently undocumented instance states Deprovisionpending Restartpending. +defer all X issues to + could you please have a look at this? There seems nothing to be wrong with the Spring ORM manifest so I suppose it is rather an issue with our Hibernate bundle in EBR +I dont know how that got there +This patch looks good to be committed to me +Closing resolved and verified issue. +Rebased onto the patch applying it depends on which fixes error handling on the bk output stream +Yep we have the same problem in Persistence and Security. +Back ported in +The main problem is that we have to add the operation to the queue of so that we guarantee that it knows about the operation once we receive responses from bookies +Can gvnix automate the creation of services by moving all finders into the services layer which would delegate to the finder methods created when doing DBRE? That capable of generating the layers out of the box would have great value imo +Can you give examples where dia cop +Would that mean no balancing? Or just that balancing would be every server has same number of regions? If latter lets do that for now +The previous implementations would work maybe of the time +I have raised this SR for the BUG in JBOSS for the errors I am receiving in the logs? Atleast help on that issueAppreciate your positive response. + +Proposed SolutionUse a instead of a in the class +This patch is with m planning on moving this to quite soon +Ate beat me to the commit though Try checking out fresh from the cvs +Passes smoke tests will run full suite overnight +So we dont introduce any additional enable jaxp just modify the property as suchCastor will automatically check for a JAXP parser in this still need to get rid of the Xerces Serializer dependency but thats a different issue. +Patch is attached +Can you attached the modeshape configuration if its been changed from the default +Same here +failed file since even though the ops dont roll back they still failed +compiles with trunk Far from complete but works for my immediate d like to use it +unsets the parameter at the end ofqlsrctestqueriesclientpositiveqlsrctestqueriesclientpositiveAs part of this JIRA the unsetting should be removed and the ant test command in the Description should still work. +Would you cross check and close the bug +What was the conclusion on the user list? If this is all too hard why not just create a separate issue to add unless and confine this issue to the idea of +See also the comment on by Gilles Scokart the classifier is suposed to be an extra attribute that should be placed in a different xml namespace +Should be a subclass of same as? Do you plan to unwrap ALL in? Currently we unwrap only a few exception while others are still thrown as +thats just asking me to do even more work + +But an extension point could possible solve do you mean by when the server runtime changes location? Could you explain this use case Max +There are pairs of code contribution +patch looks great +So we cannot simply say this is the right way since it is consistent with what we did years ago in a completely different contextWhatever decision we make about symlinks we should not overlook the fact that the situation is different now +Checked in a fix for the bug + +Investigate for if they should be in SOA Van will set the Fixed Version +I agree its better to factor that logic out into the conflict better yet find a way to rethink all error handling in srcgen while the conflict resolution changes are being made +If the other application is not deployed then deployment will fail +findbugs +I would argue that it should go into the stream +Yes Pauls suggestion is a good approach but if possible I would like to solve the problem without changing the API +We will not fix this for. +in jbds +My only concern is that having the static configuration methods in the might make them harder to find for the users most input formats I have worked with have them in the input format class itself +Warning and Info icons were nullified +Done. + future cannot really get a decent sample to test the feature on +Emphasis is on basic +Why publishing of XMI files fails and publishing TXT and CSV files works I am not getting GZIP exception in case of TXT and CSV files +This jira is covered by. +I think so +Added an appendix for sample configurations I can currently only test mysql the generic JDBC connector +Review comments welcome +Bulk change to CR Error Log with following env same error reproduceble for win same error with win SDK +We know that the issue is with the buffer size +Please Review the patch ThanksKathey +Now as its up to the admin to manage its database I would rather not pollute the server with such checks in. +Why submit again? You are unlikely to get a better result out of hudson a single fail in a test that has been failing all day? +I will submit a patch very soon +demo should be corrected after tooolbarGroup also. +Thanks for testing guys +Thanks for the detailed review +I think that we can assume if there is no heartbeat then any request to the host will fail +Another alternative that isnt quite as clean is to generate the files directly into the expanded war location prior to the package phase so that they will be included in the war itself +This will be fixed by adding and as dependencies of the console configuraitions Im waiting to find out the best way to do this +The has been removed from the exclude list so I assume this is resolved. +SumedhaPlease check if this works now +Tests passed cleanly for me on except for Heisenbugs in testPing and +Other issues will work on improving it in a acceptable manner +suggested was applied to the code +The stack trace in indicates that a thread was interrupted while waiting for a lock +Updated patch +Now these tests should work again +Attached is a file listing what i believe are the current known dependencies for have attached the finished documentation. +Admin rights granted to Olivier +Logged In YES useridThanks for trying at least! I really appreciate it +closing +Add increment +Is there a way to get the code changes associated with this resolution? Thanks +Still present in master +Or the command line groovyc compiler maybeI assume you reuse the AST after compilation but the AST is not designed for that +You are understandably looking at it from the scopes intent and I am looking at it from how it works today + tested! +Having a default precisionscale here could allow most users to continue using decimal as they are +Thanks for reporting a critical bug +Correction to a typo +Sorry for the lag Im only just getting back to work after the holidays +Thanks for the patches Vadim and Dennis +Ill need to investigate this +Now I know that Ill have to convert to Faceted project first and then add ESB Facet and everything would be OKDo you think ESB Facet should be added by default when ESB config file is added to project +bug brought forward to doing nightly builds +New patch using instead of +Im thinking of implementing some of the new code the same way so it can be applied to tables in as well +Is not against but against CVS HEAD +m install trueif this still didnt work please confirm the version of the you usedit should be did the trick +Will make a big difference once we have the new site up and running +Ok thanks Dan +fixed checked into trunk and backported to +you mean that you first have to run mvn install right? Because I dont t think Enunciate will start any process without a maven command or am I wrong? And about the disabling of a goal +I have prepared a patch that fixes an install issue related to the recent changesThe fix is to have configure set the INSTALL variables in each Makefile so the On systems that do not have a compatible install application or script installed the install target should default to the provided xmlxalanc script +All users are expected to move to or higher +Verified the patch locally +Bulk close for release +Could also be interesting to have some kind of MR job that scanned over all storefiles and generated aggregated statistics about all your tables and regionsA quick row count is certainly useful +I also meant to mention that this patch can apply to trunk and then merge to +If you assure me this is the correct patch and you can justify the changes and that all of the patch applies to this jira Id be glad to accept it +I think you should apply the patch and then ant portlib to get native code compiled or the patch does nothing for you + +No +Released with version in syntax +Attaching one more patch to fix following issues in ebay and ebayStore component Fixed issue of set complete order status in ebay when we complete order in ofbiz +Im now streaming jars from the URL to HDFS instead of copy so that would work from any URL +Retest show position and now pass. +Yes this is a bug and will be fixed for +Thanks for looking into that! Ive some questions thoughWe dont need ESB bits to be copied to the test profile for tests +Use the java code from to implement archive packing and unpacking in JavaI prefer +Fixed in the and Fuse doc source +I could be wrong because I have not done much research myself but because there are quite a few compile time nodes involved for user defined functions I thought we would need changes in handful of compile time nodes to make sure that the collation info from gets propogated all the way to bind and generate time +batch transition to closed remaining resolved bugs +Ahh sorry about that +gerolf can we not make the first step automatic by defning wicketstuff as a plugin repo in our parent pom +Ive made corrections to your patch +we need to check is null or not before calling its methods +There still several holes to get stream udf work under Hadoop I would suggest commit the patch first mark ee tests as not valid in hadoop then fix them after branch +please test +there are some people who are thinking that a release could occur in the fairly near future is the work on this issue likely to be finished and tested and stabilized very quickly like within the next few weeks? It sounds more like a if not +This is described in +Thanks Army and Dag for the comments and suggestionsAttached is a patch proposal for the docs and the resulting HTML page for the LOCATE function in the ref guidefor easier review +Running tests now +Hello TonyIve fixed the issue at revision r with minor change of your test case +DougIm getting there so far the and the are working +and are already fixed +Kristian given the problems I had it would be nice if you could check if the commit is ok for you. +With this format that is exactly what will happen +New stuff looks good +Means that if there is no open files with a particular then that clinet will not be there with renewer +Sorry updating the patch to also include some code that properly sets the wsdlLocation on returned too. +Weve run some evaluation of this speicifcation statement and decided not to fix this as Suns JVM skip this checkSo I would propose to close this jira as WONTFIX and disable corresponding VTSVM test permanently +it appears as though the SNAPSHOT version contains stuff we need +Ill submit a modified form of this patch again after I take a nap +Im not sure I understand the problem yet +Policy exceptions are synchronized across the broker so if you exceed the limit on a queue the sender will receive an exception and the cluster will continue as normalAny time you add a node to the cluster it will synchronize with the other members when it joins +It would be really nice to get confirmation that this fixes the reporters problems +I just committed this patch +Can we close this now +bulk close of all resolved issues as part of closing items. +Done through. +JBPM console works well with JDK +NoteThis usecase is an issue only in the portal env +refers to a Tomcat SSL configuration issue +to correct release. +From todays IRC discussion until this is sorted out the related work in the Axis binding is blocked +It also means that it is one less thing that we need to you run is starts the productions server not the default server +Hmm given that there is a working I hope you dont mind if I change the priority to low +mrepositoryorgjacocodestfileCDEVELOPINGWORKSPACEINDIGO +Verified on JBDS. +Resolving. +Attachment has been added with description The has been added with description New testcases and transition resolvedfixed bugs to closedfixed +If one has a mixed cluster with some nodes and some nodes then one might reasonably set these differently on different nodes +If see any value in any of the content in my FAQ they can use it feely. +Is this a dup of +Negative +Thank you for making this change +in rev +I am going to unlink the subtask for conversion of legacy metadata and resolve thisThe feature was added let legacy metadata remain legacy closed. +Im new to the JIRA experience so Ill try to have a proper testcase drawn up later +I also included my corrector +Verified by Spark. +needs some of the listed missing methods to return to instantiating s as opposed to an extended +Ill reconsider test more and reconsider when the access block mechanism is in place +Revision New API applied to +Max that happens when I try from Brno office +added RTT is no more an issue on the latest build +Thanks closed. +Key Embedded Mapping +Please provide a patch for httpd and let us know which OSarch build is neededWell work on this as time permits +new version up on review boardYes its a big patch +I checked in a fix +If it passes hadoopqa go ahead commit +Daniel I tried with both and the current code from Git and the Table annotation is removed from the entity ITD after it is added manually to the +The EAR is preconfigured for +Converted the Set C class to Boo +env +Did it occur with PLF on Tomcat +I am closing the issue. +I tried to fix the issue by subclassing at first for the project i needed the functionality for but ended up copyng the entire class as extending it didnt work as well as expected + at +Could someone from the Hibernate team comment on this? Im running a hacked Hibernate version to get around this +We dont know how to run the mini cluster with the namenode ha rightThe above requires effortThe in depends on parameter to tell whether the cluster has HAHowever doesnt exist in The first step would be to enrich shim layer possibly through srctestjavaorgapachehadoophbase so that HA config can be specifiedI suggest we add test case for HA Namenode in separate issue +Source code +When i open of project A then i click on any project in Project Explorer focus still stays on overwiev now when i open profile selection view using shortkey or right click it opens profile selection for project A because i dont have a focus on Project Explorer but of project A +Pushing issues to the next CR +Its a issue. +Marking +I created a fresh IDE today but this project was created under an older version so it might have some bad baggage it could have persisted Im surprised that I would be the only one getting this when I have gotten it every time I have used it +Can this be prioritized? Im not the only one blocked by this +Just as an FYI I was able to execute the seam test suite with this setting under JDK +Or is it there just as a + +We started using bash because we could not find a way to make it work with simple sh +Looks like safe fix +Marking the log as corrupt makes sense since that enforces the variant that all valid logs have the STARTLOGSEGMENT op +Typo But if they are notshould be If they are not +seen that reported in user forum too I believe this is an AS issue more than a tool you attach the ear that causes it +The Show In Forge seems to work normally in the Java and the perspectives +Two little things +is submitted for hudson tests +Its working on my local box but no on about this its working +Manually tested! +Closing since there have been no other comments since this was marked resolved. +duplicate by closing stale resolved issues +Cleaned up the class and improved error handling and shutdown code in. +So we can target Sybase support for the next update. +Patch became stale updated the patch +Yes it is missing the annotation +Can you please do a rebase +And where do we track style issues? +Also there may still be issues with retrieving the + +This may work though it is not a supported use case +So you can try updating to the latest snapshot version and using the compilerBut again theres not much more we can do until we can reproduce the problem ourselves +nice +It also has the following performance improvements use a hashtable rather than list for outstanding requests only flush the txnlog buffer we we absolutely have to only flush to buffer for the tcp stream on the leader when there arent pending graph shows the throughput of zab as the number of servers increase +Stack trace generated when endpoint is accessed +For inspiration I just listened to a winston churchill speech +I have committed the patch to trunk and +check out the blue text near the end I saw the HTML that renders the labels it is exactly what You said +Project containing for ACKing a HL message +I see that you have tagged this issue to version +You have to do a delete flush compaction and put +Glad the package isnt doing this anymore +is still out of dateDesperate to get hold of these changes as quick as I can Ive been waiting for reverse engineering of relationships for agesThanks we didnt get the ui in place im working on that +For the other suggestionI will do a modify laterThanks for Stacks review +Patch looks good to ve applied Felix patch in revision +Since I needed to change the skin to include the trademark in the footer I removed the extern and created our own +Doug Thanks for the kickMoving this issue to Resume Progress stateEdward Please regenerate the patch +All these issues seem to that the special char problem in have just created a new unit test to exercise this behaviour in string based XSD types and it would appear there is no problem for these +But it involves further patching our branch of the Jasper compiler +Include in the patch +OK thats good if youve found an alternative solution thanks for your response +Im beginning to think this is actually a problem in my app an NPE somewhere +Are you sure you are running in a jruby environmentLooking at your stack traceCprogjrubylibrubysiterubyrubygemsit looks like you are running the MRI version of gem not the jruby version +patch attached +I have applied it in revision r. +add the attached and your +Can you please attach a minimal source file and stylesheet which reproduces the problem youre seeing +Commit some dependency changes to trunk At revision For they need to update the configurations to accept servlet spec +Am I missing somethingId support a clean abstraction but this change modifies substantial parts of the framework in ways difficult to understand for a relatively uncommon +Is this a duplicate of + issues after release. +Also yoy should be able to get the full error from Eclipse log or from the Error Log view +Here is our deployment POM for your further AliI think we can handle that now that we know the functionality is actually workingDo you have an estimate for the release which will include this changeBest regards RayCARGO usually gets a release every weeks so we are not too far away from thereBy the way the only change is a DEBUG log the functionality is working fineCheers +This is where reading editlogs from active will be handled +A comment states in case xsi is not null type should also be not null +I need more information to determine the root cause as the exception message only indicates a runtime vCenter call failure +HiYes the SAPI should be in the class you try with JavaThink the problem is how JVM handles we would need a special method that would be called outside JNI +We should probably check that the HADOOPHOME directory and the config file that it finds both have the same group ID as the itself and are not +I think I can remove that I will upload it again +should readis not a completely replaced byIn other words the set statement doesnt workThe does +problem with user agent is that we noticed some requests in this scenario would come through with the wrong value! another ie issue im sure + When using the custom deploy directory the Prefix deployment scanner doesnt seem to be used +Just to alleviate future confusion the original example implies that implicit joins across a are broken and this is not the case +not able to reproduce with latest kvm agent and master +But that workaround may not be to implement Optimistic locking +Please see for this +bq +Committed to trunk and branch +from me Konstantin would you mind taking one last look before we commit this +orig files should have been committed too should they? Copyright in the new files should be notes version should be version Rev Date Noted for the future +Marshalling work is moved to +Sales discount entered at nothing shown +JacquesThe first link performs millisecond arithmetic a definiteThe second link looks more like the code I came up with but it still falls short of giving the exact elapsed time +When I use workspace metadata deployment with the scanner is added but not removed after server stopSteps +It happen as soon as youre using a rich component in your only happen on macosx above workaround also works for me +the issue is that you dont have a primary key defined for MAPFUNDPAYER +I just committed this +Patch applied. +Please see of this can be added to betaas there is no code change just readmes hereRegards applied +break the tight coupling And what is the advantage of this? So I can do svn move session in without having to expose Jackrabbit internals to JCR clients +The attached stylesheet is an invalid XML document and there is also an invalid xpath expression +Our workaround was to redefine the other authentication policy to use our login module +I love these features and have wanted them for a while +Attached patch a which makes the original repro pass +Probably the same occurs even in maintenance branch +What are the odds this is actually a Grails issue? I thought that only tweaking the file pointed the finger at Spring Security +Patch is applied at +release first +Done please check and confirm +Can you try that again? Thanks +We can easily consider a case where there may be additional tests in mapreduce which work with different filesystem implementations +Thanks for the patch Eli +Ive attached revisions of and that all have various performance tweaksThese are based on the TRUNK versions of the code as of jan +with EAP. +Bulk move issues to and ve encountered several issues recently because of this patch I committed and I am considering reverting it +Do you agree +Synchronized is sometimes good because it makes sure that all the variable are retrieved from main memory rather than a local cache +This is a general static initialization issue which will be solved once static initializers broadcast their values +The problem reported by Rohit is fixed in rev +Ive committed it to the trunk and branchbq +yes ephemeral is a bad name +Simple patch and test throw an if we get an empty list +The to get the Deployment might not even run in the same JVM as the Instance +bq +is no longer an issue +So we really need to have package built exactly the same as Maven would build reason is that sometimes its not as easy to build an archive as copying source files to a jar file +The CP ones have started to show up of late +Currently it goes through testing +fieldName homePage sizeMax add field string class +Since we need to support everything we ship in the repo we try to include only the minimum +Thanks Jitendra! +Yoav is working on the proto issues +Patch applied in revision thanks for your contributionPlease and close this issue if ok +Im working on this +In addition to the splits by Owen above I would also suggest the following package restructuringpackage package +I committed thisI added a simple file that handles test and clean targets only no interop targets or binary release packaging yetI also added license headers to a few files that were missing them +We should also fix +Clere kindly informed me that the Whirr JIRA notifications are currently being sent to the Amber mailing list +This got fixed in the proxy handler +Cant replicate as well +Lets think about it carefully and work on it separatelySure here or separate JIRA doesnt matter to me +One small comment +This site now references +Lets run it through one more time +then we can show info in and info in the schedulers info and get away with the issueI personally dont think you want to introduce such differences +We could of course reduce the sleep time from ms to anything else perhaps thats even better now that I think about it +Attaching the first rev of a release note for this issue +wow! This is just great! Nice work Eugene +I used the original code and XSD from this ticket to test it in Eclipse with Java +The plugin encapsulates about of the Twitter API but most of the core areas are ready +After adding debugging info this was very apparent +Note that the relevant Neoj module version hasnt been released yet and the release made is not officially announced yet +Would you also return JSON or in this case? I am not sure if this is a valid path to go +Resolving as Fixed! As of now we have no known issues regarding JSR +Thanks +The instructions are Create the Eclipse Project files Import the Projects Clean up some Build Path errors +I am at a loss still +Why not just stick this url in a properties file so that anyone running this can modify it for whatever theyre running in +Changed to use instead of. +As part of updating this patch could you please root it in recipes rather than contrib +Specifically the +If activation succeeds they are moved to the correct location +patch looks good +Field cacheswill require custom dynamic growing of the field cache arrays perhaps wellneed to add a +Files modified the following changes in Changed the change function to click function Added one if condition to each click function +Attaching another patch addressing my minor comment +for eg the scripts have Drop tables commands at the start and then the create table commands +Hello Mr +tests run and pass for me code looks good +Although aftergetting my feet wet I did also consider moving along the direction of modifying the byte buffer itself to includethe memstoreTS +gtr invalid backslash escape at end of stringecho i tr sed s +Some times people may implement this kind of factory themselves depending on the design they useSo I think the better idea is to discuss this in userdev list and add this feature if people find it more usefulthanksAmila +Need to resolve these issues for maintenance release +patch added +These changes are a simplification of the interface +I was willing and I am still willing to integrate it into Hibernate Tools +This is also a problem in oracle +provide steps to verify this issue +Oliviertake a look at the attached patch and tell me WDYT pleaseI thought the aside the first match policy we could add a exactly match approach +One possible solution is to let the consumer read the partition data from ZK directly +I also have the build first run some shell commands hostname ulimit a and jps +kggg +Hi RayeesIn the BVT tests are the create template from snapshot tests running fineThanksNitin +wouldnt that be sufficientEven with the above you wont be able to do sorted nodes since the chidlren are still sets +Since we already released a version of this in Id rather keep this closed and open a new issue for any further changes +It would be great to verify that all the tags can be built after Migration +A classloader is better optimized for searching classesresources and would certainly increase performanceIt is also possible to use the Eclipse Java model that doesnt lock jars and uses Java indexer for searchingThere is yet another issue in +Dont suppose you can give us a test case to reproduce +also article on on Oreillys ON Java is required +META +The command stty sane will cause terminal abnormal on AIX so remove it and use devnull to quiet +Done. +This is for the same should be able to apply to JeremyYour patch is in trunk at r at r +For example Xerces and Xalan implements on JAXP while Tomcat depens on it to pick the parser +Stans point about separation of aggregation and parent is relevant. +Yeah some sort of lazy loading should do the trick though I cant think off hand how complex it will getIdeally youd be able to specify the encoder class in each column familys schema params without going through an enum +SateeshIf UI is tracked seprately can you move this to resolved state +The bugs Ive seen are infinite recursion in RA reconnect processing deadlock in Connection close processing +Changing this to be an enhancement request instead of a bug report + +Please remove me as a moderator for all Kafka listsNeha and Jun Kafka PMC members and current moderators for Kafka dev and users will be the new moderators for Kafka private +There is a configuration file where valid domains are placed one per line +It also has the same error on on also fixed this in masterad but well leave the hack in place for for now. +if mapper has no data to handle +Yes I think youre right +There should be a comment in the log file stating the sharedLib absolute path directory beginning with loading shared library +This is not essential but convenient +I read through the patch and it looks good to me +Thanks for reviewing the patch Paul +cloned from the problem is in the construction of the tests pleas let me know what I did wrong +That might helpYou might also want to look at how does things which stores the numbers in binary while still indexing it +If we did that and Ed logged in then the problem would still exist just in reverse +Fix wait to give up after sleeps +Small change to KV +I think it was this way because behaviorally it was the same +I have a patch available running derbyall suite now +Just commited a fix so that the add bookmark and add task actions are available from the gutter. +To check this just make sure that has not empty JPA Details view even it is not added to persistence unit +But the save page looks not working porperly +I know you havent committed this yet its just a friendly reminder +This would be a useful feature and Ive wanted to have it in the past +Please see for more details +It is very picky about serialuids for classes +You can still use Seam for Ear bad that Seam can not be used although it is a GA version +It would be nice to have the connection state diagram as part of the documentation +I suspect that there is something related to the usage of temp queues that is causing this problem +Committed. +Thanks Harsh! Ive updated the patch +The intermediate transactions between NEWLEADER and UPTODATE are never written to a persistent store on the follower unless it manages to stay alive long enough to do another snapshot +Resolved for a while +I think we should +A snapshot of the fixed which is compatible with for anyone who needs it is attached +dooh here is a patch that includes I just ran svn up and ant test a bunch of solrj things fail i cant look into just now but Ill post anyway +Ah thanks for the feedback! Are you able to share your project setup +sometimes the docs that are there are very opaque +Use the tool interface that lets you define properties and config files on the command lineEither the command line or the environment variable should work for you +Thanks patch merged on and trunk. +On Tomcat gives a which is caught by same thing should happen when using servlet but the kicks in from what I can see in the logs Hibernate starts up ll report this to the tomcat team and keep you updated +I did extensive mods to it as part of +Ive started to work on this issue so if anyone is also involved in implementing this component please contact me so as to avoid effort +deals with a similar problem where richmessage is same root cause? +I shall incorporate it in the next release +I understand that +The tests that the processor on the mina consumer side can recognize session creation and can asynchronously send messages to a producer client +Some weird issue with JIRA fyi +Right +It is a red flag for me that existing applications could be impacted by the code change +what OS are you running? I will see what I can do I just use Netbeans execution engine so it could be a more general issue +Edgar its interesting Tools also has ESB project creation wizard if you are interested in it you can get it from tools nightly build +The move is virtually done all that remains is to the query tests as soon as is enabled the lucene query tests with rev Anything I might have missed during the transition +Excellent work Benjamin! I enhanced the patch a bit Made take throwable Found more places where it could be used Made extend so we can skip wrapping in a few casesAlan or Olga can you review so that we can get this committed +looks like its missing the which wizard did you use to create this Seam project +Titles adjusted with abdfdcffefbfbfdcee +It is a common scenario described Apr in early versionWe have done many works to fix casesSo it wont be a problem now +As committed in Rev +Oops +Based on feedback Ive seen resource producer fields can only be dependent +Okay Ill put up a patch +This one seemed the best to me but others may feel differentlyThis idea came from amc +See for Maven support in Spring and for other Spring modules is a start but he update referred to above only covers the there any chance of getting the rest of put into the maven repo +It will break the BNN functionality. +It looks like this is exactly what you are looking! Thats what I needed +the latest patch looks good to me +If the original sort was by relevance this doesnt give you the groups sorted by relevance right +The bug occurred in version +Closing. +That is the only real differenceDisclaimer Im not familiar with the code base but Im guessing that one fix may solve both issues +Ill certainly test it by the end of the weekend +If you offer a getShort somebody could accidentally downcast its bit values into a bit values without realizing +Youre welcome Doug +This would probably be a bit more consistent +its a shame I didnt document that change back in +This is a feature that i believe few know about or use +on the fixes on trunk and +I still need to pick up on this change for deletion however the logic there now requires documents that are in activerescanneeded to be put back into active and not actually deleted +fix typos remove the setpreflightheaders functionTested on safari chrome firefox +Implemented in revision to the branch in revision . +bq +The ngram dictionary is created from the sample data +to master. +The proposed change will inline the translations which becomes part of the page markup +Now not all buttonsfeatures are going to work and that primarily is problem with test applicaiton itself +Sorry again +That should work but Im not sure about that so please give it a try and let me know if that works or not +But I think we should have real use cases before we hack on it +This file will be at the same level as +So great! Manually tests added +Reopening for release note +Not to mention that wed be doubling the number of OS level processes usedMy vote would still be for using heart if possible. +Canceling the patch until we figure out if the test failure is due to this patch +Id be happy to see any patches you can offer. +It requires mocking the constructor of +needed when resolving different versions +Since the default encoding is platform and location dependent this test will fail on other machines or produce different resultsTo fix this always specify the encoding when opening the data and it should be retrieved via the class path instead +Yes same solution I was thinking +To be double checked + + +Attaching the updated patch +Removing backport target until we have a patch andor more info to reproduce this +Do you want to help +This is because the disabled status of the field may change dynamically due to +That is instead of the individual language generators deciding the values for valueless enum constants now they will be set during the grammar parsing phase +Profiles cant be done until we more clearly have defined other parts of hte behaviour +is there a way to integrate this into the mechanism? That will be nice +verified by Vladimir. +to trunk branchesThanks for the patch Hari! +Regarding your performance concern for now we have to address issue +Thats the reason why I would like to get this into +See the full list in +On the current trunk single and double both fail the others pass +CheersSid +From the history it seems that Jetty releases are made approximately every days +gitignore rules +Rather than blindly casting I think its worth checking instanceof and bailing out with an error if one of the journals isnt a +But I think we should allow it +One nit is your patch leaks connections due to not closing the or you create in the +Any comment +Feedback timed out +AmitHere is my comment from earlier in the thread dealing with a similar had to put the ntlmFilter on all of my to get IE to work otherwise the post parameters seemed to get lost +The problem was not with but with +Updated patch which ensures that the root directory is created +Many thanks Dmitry for contributing itOleg +Covered by an IT. +This was already fixed and will be available in the next release due tomorrow August . +It should also be in the file but it is more important that the file should have the checksum embedded such that walking the whole file gives a fixed value +I thought we were now caching client so is this still necessary +Patch and screenshot with updated name to am also fine with instead of or +Attachment has been added with description Full thread dump showing deadlock in Batik +If I can have both info I indeed can try to provide a do it was an honest oversight on my part! Do we need any extra support to inject particular cache or is it already supported by our CDI extension? +Applied patch in trunk in revision to the standalone codebase in revision . +What is the motivation for change +Stay tuned +The schema tool does not currently support being run more than once to changeto a different database +I ran a build and reloaded the edits log for sanity +ve just committed this. +The patch fixes the issue +gradlewtest still says the tests failed I suspect because they never ran +Solr support for the improvements implemented here revision +This has the unfortunate of requiring some deprecated methods to be made public again instead of protected in the +Wiki reference manual has been updated to reflect how to perform this type of configuration + +If this is indeed an issue with Apache please feel free to file a jira with details specific to that +This will create a repo for you under the portals profile +bulk defer minortrivial issues to +on patchWhen committing please add profuse comments that this is a hack and a better way needs to be figured +hasnt been an issue in a very long while +Please at rev +bulk defer of unresolved bugs to +If the Log syncer is on suppose thro the we have done a WAL edit and in the main flow we do another but both needs to be synced together as atomic then we need to the logsyncer to be disabledOr else the logsyncer may sync up the first Edit that happened thro and then when the main sync happens there could be a failure leading to inconsistency +Always requeuing the request if we couldnt get the lock seemed to work only if I put a sleep in before closing the connection to give the post commit tasks a chance to catch up +java print xargs grep l All the java files which contains the and need to replace with +Look like thats yet another regression hopefully easy enough to +Marking resolved per user comment. +From the documentation of the Form component I see a focustruefalse parameter +I did not add new tests since the changes are simpleI have committed this. +So you just need to extend the class call new on your version and call the system procedure +I havent found the hdfs jira for thisDoes svn blamegit bisecting not turn up the issue that fixed this +Any thing that adds unnecessary complexity will be vetoed by me +Website up. +Shouldnt we add new category? Something like Support +I did not care much about the report at the time but seeing the code on trunk I realized there is no way to force into config and ft like in only way a ft is cleaned in trunk seems to be saving the which I guess would be triggered only if the user went to the specific feature type page and saved it without chaning anything +Incorporated Dougs comment +Please correct me if the reported fix is incorrect +Added tld attributes +No +Thanks Jonathan +How long do I have to get this into? Is that still possible +The interesting part is how frequently they occur +updates per bens messed up the last patch this has the proper updates +Oh right +It involves some significant rework of column prune optimizer +I cannot edit the issue to change affect fix ve reproduced this issue on my side in de Design service +Thank you for resolving this +helloworldejbwswss The UDDI httplocalhostjuddi +If colloq folratfolrada +The Subject does not contain the correct set of principals +This patch uses and Version to preserve backwards compatibility +All of the explicitly reported syntax errors in the grammar became +Patch for based on a release +Hi Carlo who is going to produce a patch for the CP +However its not clear why we have only one test case that is failing since we are doing the same kind of operation in almost all test casesSince this happen only on Solaris it may be caused by a delay before we can actually reuse the same port but in order to validate this I first need to be able to reproduce the problemAlso we dont have any code in place to randomize the port but that could be done +This one may be a bit better? +I dont think so. +It is also fine that we have the option of manually setting the cell type +Tests verify that the config has been correctly read and loadstore operations happen as expected +But just for starters we could add the jar at least onto the classpath? +So only Strings and primitives are alowedIt would be even better if class really had only those put and get methods +We use to abbreviate Apache Directory Server all over the website in the navigation +Mark how do you feel about this syntax would you adopt the same in OWB making it consistent +Heres a patch that includes a JPF to test the struts validator support of the bundle attribute +I went ahead and committed some code towards this issue +Yes please move them to or unassign them based on your own willingness to work on it for + +You are using the wrong branchtag +Reopen to close as WONTFIX. +here is a patch cutting over solrcore +Wonderful! +example pom was missing the client dependency +Verified +Some typos in the Admin guide on the page titled Running the client with SSLTLSa The phrase If the server does client authetication should read If the server does client authenticationb The heading Running the client when both parties does peer authentication should read Running the client when both parties do peer authentication A little grammar sanding in the Developers guide section titled Derby and Security The phrase in Derby Server and Administration Guide would read more smoothly as in the Derby Server and Administration Guide Same comment in the Reference guide section titled sslsslMode attributeThanks +see +is there any http server in front +Ive added an extra constructor as per the original request. +Kind Regards Wolf +thanksd characterise it more as suggesting that people walk on top of a complete bridge rather than clamber along underneath where there are no handholds +Anyone using an earlier version should use the documentation that is bundled with that version +Adam are you seeing it hang or just not download at all? And were talking about right +The DSRAE Statement is closed problem is also mentioned in +Problems were introduced because of merging with trunk problems +Gordon ROOT refers to the projects directory +I will commit it after running unit tests +With resolved itd be great if people could check out the branch +Verified by Tony +I move summation work to map +The configurableness of the progress is nice to match the zookeeper configurableness of the timeouts +the test was left out of patch reattaching plus a changed the wording on the rationale for not catching the to it may be better to expose itb removed the around the final commit in the nested tr case you are rightThe regressions passed so Ill commit this version if there are no further comments +Patch committed to trunk thanks Chi Runhua for the patch +You have some mandatory requirements to keep EVERYTHING +I am thinking this might be due to a mismatch between junit versions or something +attached correct schema this note to self avrojs rpcsend utility uses an HTTP transport so its not going to work with my Python raw sockets server +I dont believe that this is true +This would be nice +The only transaction was rollbacked +May be we were not waiting +patch applied in r +Its path will be stored in rat report in target +As I tried to explain before a command line like mvn +I am looking at creating the statistics automatically if they are found to be absent when needed +is no more used +I think we should be able to use better approach +Scrub cant help there +Data from committed transactions may be totally missing or partialchanges may exist +Unpack attached anywhere +Committed +Why not take it one step further and a slice of the with this block size a la? Reserve say of the for slab management and leave the rest for the blocksInstead of explicitly setting the buffer size why not sample existing and calculate a guesstimate +This will have to be done through ZK acls though +Hi I havent got time to go through the patch in depth so some of my comments might be irrelevantOverall I like the idea a lot +But once I restart the other two zk processes then everyone syncs up and were OK again +I can see that there are as much as Configurations in fact if you look at the source code of the test class I attached youll see there is only one Configuration instance involved +Which changes in particular are you referring to +Hey Ivan +I dont think we should take responsibility for deleting peoples database directories +The changes that I have made will keep estimated update everytime that Speculator calls +Thomas Im going to try to fix the pdk problem later today +Mice this issue is reopened can you check on this before which is our planned RC on the latest build git HEADfbcfbaeadcfdfeaadbadfe +After applying the patch the ant test task fails on +soant realcleanantis supposed to fail? any idea how you get it to fail +I think we can close this +Are there anything wrong with the unit tests for the scan delete tracker? Or is there some other issue that makes you think it is broken? We havent made a full clientserverclient test yet so I cant confirm that is works all the way but please let me know what is brokenStackYeah I think that applying the deletes in memcache will keep it cleaner and we use less resources for the gets and the scans after that +What if you set JREHOME correctlyset JREHOMEcPrograjava had tried this earlier and got many more errors than just using the javahome JanTry these +Kafka request log is a daily rolling file appender that contains every request received and served by a kafka broker +Higetting the same issue after using the Axis version. +Do we have such a page +Any solution for that +Would you please commit the patch? Thank youBest regards thanksThe patch is committed. +Reopeningso we can document when that problem is repaired +That would only require inclusion of the Geronimo servlet API into the build instead of the EJB and potentially other libraries. +what is dom +Extending a line does work as I tested on an example myself +Unfortunately the solution is not that simple +Is the issue reproducible in nightly builds or even better in +Thanks for this Bryan +This is now fixed in SVN +I actually think it would better to document the setup instructions in all of the tutorials since the Page Flow one might not be read first even if its at the beginning of the listResolving as wont fixLets address holes in the other tutorials individually +Please review and let me know if anything needs to be added +Here is a new patch with the other approach +In this case the first OOM occurred here during writing the checksum +buttons so user can manager CPs on the fly from this wizard similar to JDBC Importer +Its actually the class that has the extra reconnect logic that I was wondering about +Closing this one +Then it became obvious that it was really broken +Note that the testcase would only fail on JDK patch to include fix for other method highlighted by to branch in revision +Cant you use the resolutionoptional setting for the dependency? Ideally Id expect the bundle plugin to automatically add that to any dependencies that have been declared optional in the POM +Works with too so probably IBM JRE specific. +For + +hairy and long in the tooth is new for meo +This patch will be resubmitted if and when it can be reworked for the trunk +I did something like that +Maybe we can have a flush callback on we can hook into the flush call +Committed to and trunk +A fix +If not use default +The claim should probably have been that the is compatible with trunkPerhaps it would be better to ship James rather than cutting a Mailet yes please even if the patch is just two linesIf you have a couple of ideas submit both as the since tags. +I know it is not very but it is necessary because it references projects located on different places on filesystem every user has different location for projects that are related +Thanks +Wouldnt we need to support this in other handlers too +Seam Persistence to the latest snapshot of Seam Persistence fixes the problem +whats your server config like? is there any special configurations other than the defaults +previous patch had issues patch with guava calls replaced with version that removes the new schema file and just modifies schema instead +But yeah the original issue still stands +Updated files +If I tried to truncate the length of the concatenation so you dont get a column width error thats a stronger solution although it can also fail as you sometimes need the full length of the name and blog to fully get uniqueness +Hmm except in s case it just has one static method +Probably fixed in latest repository versions +I am using and go the same error +Pending resolution +Ok +Pull request merged +Within a phase interceptors can order themselves +The test cases and need to be updated +I was merely posting it here for folks to take a look at it +Patch to implement the change +bq +Merged into master branch. +My understanding of is as followsA query is exectued and the best plan for it is picked up based on the current data in the databaseThe data changes quite a bit which makes the plan chosen for the query not an optimal plan but during the next execution of the query we continue to use the plan chosen in step +Try Hudson +As per Daniels comment marking as not a bug for +to run tests in slow env just give a larger factor value from command lineant details at precisely I have verified rev having the necessary changes to tune these sleeps according to needs. +Please correct me I do not understand the issue with iframes +Ive committed this to branch and trunk +This breakage to the error log entries is also what induced me to drop attempts to access the data more directlySomething is calculating a bogus length to the URL to be logged +Thank you Andrey! +Thanks! Also could you create another issue to track the merge to trunk? +Ok Brian mistake is on my side +Drop on an Tomcat with Java browse to and see the behaviour of an and an +is not configured either way in our application +Thanks Brock! + +for the shim layer refactoring you can use this jira If you want to change the source level compatibility you should open a JIRA for this +In general the producers and consumers in CMS and JMS are intended to be accessed from only one thread +Thanks Mark much appreciated +I agree with you but I cant see how this would be better than not using the head tag on your page after all all the head tag does is setup dojo or Im missing something obvious + +I originally created a patch containing the binary keystore files but when I tried applying the patch to test it it balked at applying the binary data +Well need to upload these artifacts somewhere along with other supported architectures in the futureI did attempt to make both the build and runtime code work if youre not on a supported platform but I havent extensively tested itAt this point the patch just adds support for interacting with zookeeper via the native code +The question in my mind is why is this in here at all +Im not sure that I see how these assertions have anything to do with attachdetach behaviors +For the SASL work which is on the I had a patch that almost fixed this +For example an HS with concurrent users will have connections to zookeeper even though zk connections are used for fairly short periods +Yeah youre right +If it is fixed then close the issue +This will make it into. +Im trying to use the Velocity escape tool in Maven so that I can escape strings that look like property references +Now that work is underway with AS all previous community releases are +Note that was using ant I switched to use ant and it honored the ignoreThanks Thomas +duplicates m seeing this issue with recently grabbed from the update site +Is the plan for to include support for something similar to the logback formatting scheme? I like the idea of specifying both padding alignment and replacement chars all in one conversion string instead of in several maps +Hopefully done Daniel please verify +The attached patch moves locking from the class loader object to a locking approach based on class name and thread +It improves the handling of property values by the class and fixes the unit test + definitely belongs into every server startup script along with sensible values for heap DGC intervals etc +look around days old is nothing please be patient a bit longer many thanks for your commentWill give my best for it +Available tracing levels and values and Table +The only functional change is that a on its own at the end of a Java file will be removed not kept +Even though those objects are I have found that it can have a big impact on performance +I hadnt tried +Will this work as far as accepting the patch is concerned +Updated patch is coming in a few minutes +Now that our focus is on and in the future support our plan is to not fix pluto issues +the test failure is due to soem c test failure and has nothing to do with this patch +I created a new issue +Being the review positive I guess you can just go and commit Kenneth? Or youre still waiting for futher dicussionsomeone else to commit it +We should also handle +AgreedAfter sleeping on it before removing the +Scheduling for Final since we should fix anything like this we can for Final thereafter it cant be changed +been deprecated for a while +Thanks for the patch Jeffrey +I applied a slightly modified version of this +Sorry for the confusion +This way we have access to all the region names we need as well as who has them +let me see what i can do +Hi CharlesYou can keep use the SOAP over JMS by setting the option of synchronoustrue it will force the cxf consumer use the sync invocation to wrok around the issue that we meet +Attached patch for backport to patch to apply in JBDS alpha fixes for iu in discovery plugins +Please add your detailed comments if you wish to make the case for reopening it +The other alert actions to not ask for an to Charles for further investigation +It will use a derivate of the code in this issue for single word suggestions +Thanks Tom Chris Ive opened to track the larger fix +Im using the packaging type of pom in the because all it does it run ant to generate the libraries via cmake and there are no jar files +These effects dissapeared and application acts as intended +Ive added logging information to the which now write the percentage of +Thanks Ted +I update the wiki there more work to be done? Is the upgrade complete +I dont think there is a way to avoid this interface +My only concern was that if we back out then people wont be able to run on the tip of since the protocol wont be the same +ThanksFolks outside the US are experiencing slowness in acccessing the git repos +Looks to be a bug +Rohini Palaniswamy Koji Noguchi Im curious why didnt you hit this one when running with hadoop ? We havent started using viewfs on production yet +I dont want to start a parallel universe outside Apache just to get around some legal requirementsHow much code is it? How obvious is it? Is it part of a tutorial? How much work would it be to redoThis open source license stuff can be frustrating +Ah youre talking about shared library +This bug was fixed with +I will do my best +Im bothered by the fact that you have to explicitly state a delimiter if you ever want to use the other parameters but I agree in retrospect that its not worth removing the feature entirely +You mean that its done partially +Fixed as part of as part of in in +When look at this please also update information from as it is related to manification. +Committed revision +Core test failures are unrelated +I just this in trunk and it does not work at all +Filed to address it +Ive moved the class and run the build +We already have a write on Mutation +Communicated with Simon Kelley and he considers the configuration file in the upstream package to be public domain +Hmm now that I rebuild everything from scratch it works for me too +Sorry wasnt clear this issue is just going to cover HDFS and Hue +Let me finish track it down +Fixes error with links to Archiva useradmin guides +committed the patch with revision to trunk backported to with revision . +Closing this record since is now released. +Resolving +Thanks for fixing this. +fixed on trunk please review Service re right Joerg after reading the doc Im convinced that the commit related to this issue has to be rolled backthanks for reviewing +So the Message type implementations in the serializable package all seem to implement SerializableExternalizable whereas their equivalents in the xml package do not +This is solved in build tried it with after our discussion this afternoon and looking at the stack trace the only thing i can think of is the maven repo +Regressed with latest builds +My bad +Well Im not sure its a great idea to have some transformers turned on by default +Any kind of content transformation is out of scope +This can be reproduced even in latest AS snapshots with the attached dummy application +but a better approach might be to simply construct a servlet under Tomcat that does something similar and see how it behaves +The related bug is fixed but not yet released by +on patch +during import from bugzilla ticket had resolved field set to unresolved with status set to closed +bq +Ted how are you testing it? I ran mvn clean compile and mvn clean compile and both worked fine +There was a problem with add event and add user in which is now resolved +Gerhard this looks like an exciting enhancement and youre right a key one to support redundancy +Thanks for the patch slightly modified the patch and then merged you check this +If there were any exceptions theyd already be logged +fixed in trunk and +Also it would be cool if you could skip certain keys with sstablejson so that there is an avenue for removing the problem row in the pathological case where you cant possibly allocate enough memory to get it removed by compaction +bq +This patch applies cleanly with E p +Please dont create any JIRA ticket before having first discussed your issue on the Sonar User mailing list +There are several fixes that have gone in which I think this maybe related to +Please attach that gif file to this is the gif mentioned in the sld +There was no big speed improvement observable +If you use and point to a random foo class that wont be detected until the cast occurs in the message receiver which is at runtime +if the readers record schema has a field that contains a default value and writers schema does not have a field with the same name then the reader should use the default value from its field +Tested and closed. +Looks good to me +Not tested on IPF. +The attached testcase will reproduce the issue of which the network connection were not being established after one of the brokers wwas testcase does not reproduce this issue deterministically and it is possible that it will require several iterations of the testcase to produce a hang of the consumer testcase will perform the following steps start broker and broker start a producer application in the background and wait until it has completed start a consumer application in the background and wait until it has completed stopstart on of the brokers alternating between the two for each iteration goto step By default the script will iterate through this times unless the consumer process is waiting this point you can perform a script will then generate a stack trace for the running consumer the two brokers and testcase just needs to be unarchived within a activemq distribution directory and run from that script is bin script was only developed for a nix system +I just committed this to trunk and branchTrunk and but not +Kristian I have removed the wrong import in my new patch +replace Vector by List +And Ive set up the Johns email to be for sending messages once any of these builds fail by his fault +Could you give a me a little project describing the problem The license of a has nothing to see with a file noMoreover the in the goals has no link with your license definied inThanks +A pattern Ive had luck with lately is to use erlangsendinterval to replace loops like that +Here is the patch for the issueIts natural place is in but as it is excluded now ive created and put the test there +Using the stylesheet above There are leaked objects before the patch +If we made the quick fix discussed above then wed have a harder time subsequently adding real support for Json annotations without breaking applicationsI agree that adding full support for Json annotations to Protocol Schema and Field would be an improvement +This looks simple and sweet to me +Thanks +I just committed this +Lastly IMO Notepad is a poor excuse for a text editor +Also we need a much better indication that JDK actually does something beneficial with this you can provide ALL of this information THEN reopen the issue +patch for the diff at to and to trunk +Import Maven projects wizard same projects should still work SnjezanaI wanted to make such a screencast ages ago No I should really go back fixing mes EAR support +Verified on JBT. +The patch submitted did in fact work but only for restricted message content ascii charactersSomewhat embarrassed by SVN revision +These need to be fixed so that the patch can be applied without breaking the test run for everyone +Max how can I find out if a property is primitive +yes that is a problem +Downloaded the jar and tested +That is what the creator tool property is for +Ill apply this after confirming the changes on the dev list +Any ideas on how to deal with the last bit of data? In my test the last doesnt get invalidated because there is no lag for the last fadvise which is done immediately prior to close +The patch was integrated. +with sources +Nope it doesnt +new variables headElements and jsElements +Logj just moves the encoding into the Layouts so that Appenders can handle any kind of data +Graph example works fine on huge graph now as I expected +Anders It is possible that we cover what needs to be covered with the patch +Rejected +Do you want me to spin a new patch +We display counters percent complete and elapsed time per step which is also aggregated at the phase level +We are close to releasing GEP so the sooner we get these issues resolved the betterThanks for your help +That is actually a good point about the src created by the dist goal +I missed that this is a windows subtask +That was before the last major shindig update so maybe that is no longer the caseYour explanation of enums is correct +In the meanwhile got committed +The changes which disable the logging contention will also be required as will changes to the process definition as it currently stands contains a contention on the variable assignment when handling the service responses + is a program I used to see if this problem would reproduce with upgrade with a +it would be good either as part of this project or a follow on to enhance the consistency checker to make sure keys in these indexes that support deferable constraints are all unique +text which is being hidden by the top seems to be a dead project. +The build is not working at least in my Linux Box as it looks for +Thanks for the patch and for the anaylsis. +The problem is neither with the NHTTP transport nor with the transports from the Axis Transports project +rating as minor though it does need to get fixed before putting up any artifacts to the central ivymaven for the key change is the target lists the dependencies for all configurations so is more useful +And besides we might wanna thing if we can do something for this as well in the Spring DSL +Yes well get this into +Fix committed into JBPAPPGACP revision +But making an aggregated sources JAR available is definitely a good idea so Ive added that capability to the parent POM +Ive restarted one node in the cluster and it worked again for minutes then went back to giving the same error +Global permissions A and C dominate namespace perms and also grant admin and create perms on the namespace itself +Ravi if you figured anything out please chime in there. +Shouldnt this issue be marked as fixed in M now that Barend gived ASF license to use his patch +I guess in most cases the job files will just get deleted anyway at the end of the jobWhen you say submitted job files you are just talking about the job conf and jar file that correspond to a job correct +Could it be because of lack of progress reporting while the task is waiting for the datanode? +This is partially caused by how was implemented and partially just by the nature of how ugly the code unit tests will be fixed by +assignee so mails go to list. +Since testclassifier outputs summary while running in a sequential mode while doesnt do so in mapreduce mode it cud be confusing to user +This is really the same as resolved Release +Verified by Tony. +I feel strongly that randomness has no business in unit tests +yes close. +It looks really appropriate +wheres stage? can I have a link to it +No new tests added so far +It would be great if someone could test that this has been fixed on Linux +I tried several ways several times and never got deployed on Its like this because this file is in resources +Missing a pb deserialization in +Yes THAT bug is already fixed in svn +Marking derbybackportreject as the client datasource structure is different on earlier branches and this would be a manual merge +After you verify it I will publish it +and Crimson in doesnt though does +Thanks Anita +Id still like for it to be optional since it appears the permissions are correct before the chmod is run +Committed this to trunk +is not getting initialized +Configuration Database class file patches +This is a simpler test case however it returns same error as another test case +I remember addressing it in so may it didnt get properly ported to +Note the patch contains the instrumentation as wellBy removing the synchronizations I added you should be able to see the bug in the repro +Please dont forget to describe this new parameter in the has been added +Attached MS this issue on both and is it possible for you to attach the cloud DB dump to the bug as I dont have environment? I need the DB dump taken after you deploy the vm in shared network but before you start destroying VMVR +is threadsafe so its overkill when you dont need threadsafetyThe compiled code will run under which satisfies our requirements +Thank youYour change in +How do we deal with map failures? Do we get multiple completion events when there are +A question In most of the code the have been done away with and replaced with Object + applied thanks a lot for the contribution! +At top level you can run mvn install PeverythingBut I tried this from the distribution folder +cleanup +Be sure that you get a consensus before working on this application unless you need it for your own purposes +I have the same problem +And it does not work correctly because you have to establish a match between manipulation methods and fields before blindly deleting them by looking at their method name +But we need to think twice before we take that decision +I should have something to submit soon +Well spotted! Spring does declare that dependency could you please this one for the timeframe? I thought I fixed it in but apparently I didnt +reference as these two are related + deployments from template copy from template on secondary storage direct to primary storagefcadbbdbaabdaea +Attaching a patch that makes the suggested change to the project +to and trunkThanks Kai for reporting and Prashant for fixing +This patch brings windows up to same ICU level as unix series and allows people who need to compile ICU from source to do soThe original issue for this ticket remains but at least people can compile from source if needed +This is indeed resolved on AS I just with a clean server install and fresh builds +Why is it so? Is not imap faster than pop when receiving and searching mails +Trunk for patch with the same changes +yup working like a charm! +HiPlease let me know if anyone is bothered to look at this issue +Now I got itpro Karaf Just two necessary feature entries for http and war this fits the specialized need for karafcon Karaf why make up a feature xml when there is one providedpro the project knows about the needed dependenciescon the feature xml is slightly more fragmented and the dependencies can not be seen as clearly as in the karaf onejust my +I am not sure if it is the best idea not to use as base foryour fileObjectThere is so much logic in this class like the event handlingCouldnt you change your fileObject to override the protected do Methods +I noted that both were affected in the problem description but the testcase I supplied only demonstrated the problem for +Sanjeev in investigating this blocker defect we realized a bug introduced back in in handling dynamically downloading system and templates when a different type of hypervisor host is added to cloudstack +Ryan can you see if you can reproduce +Please assign back to me when done +AlexandreThe issue was closed as already fixed in the current version +Im upgrading this to blocker since it makes it almost impossible to run mapreduce jobs +Should it be removed +Do you have test reports available to look at when this testing was done +bulk close of issues as part of making release notes. +the problems was create new issue for new problem when in +That can be in a new issue too +fixed now also inside the X in no release mistake +This feature turns out to be hard to test +But I have no way verifying the fix now +Ben I suspect all of these are obsolete but please take a look and keep openretarget any of them that are still useful to youthanksdave +Can you check it works for you and also update the wiki +Reviewed by Philip +Thank thanks for letting us know +This can be the reason of our misunderstanding +See corresponding class in branch nutchgora someone can identify the outstanding bug here I propose to mark as wont fix and close this issue out + +shrug +Patch for both issues +Please check out the latest from trunk and confirm it is now working as expected +I was trying to explain the the API sends the Vlan list when we give some values rather than keeping it blank and that is the correct behavior we are expecting even in the case of the blank valuesRegardsKiran +Reassigning to jgray FB team is taking this and running with it +No other plugins +If you fix this bug both should show cversion of +How to deploy and test is covered +Attached a patch for supporting suspend and resume process instance in in r +Isnt this similar to a jira opened by you already? The issue being that the scheduler puts a reservation on a node whose total capacity is smaller than the reservation resource size +I have seen in some comment that you have the idea of distributed Auditor +Danny +Youre amazing +This have some flaws so we will have to change it so that Leaders makes this decision instead +bulk defer to +I would love to get us back on the Apache Flex track so will definitely check this when I can +Please note that the next EAP are due out at the end of April +Can you provide a patch +MarkCan you submit a patch with the proposed changes? am not exactly sure what you mean or how to do it +Moved to please move bugs back to which you will work on in the next weeks +Patch prevents users from clearing namespace quota on +Until then Im going to carry on trying things +The fact that this feature doesnt work is missing in online suggest to use to track this. +We need figure out why the tests are broken with your patchFreeman +This fixes the problem in some cases +We should leave this as is +Maybe we can use Objenesis by default then it only applies to CGLIB proxies anyway and just have a setting to switch it off +This bug is reproduced depending the operating system MINA runs on +Hi Patthanks a million for attaching this test webapp Ill investigate the problem shortlyCheers was also resolved +Unassigning due to inactivity +and +But as Burr said my web service module is a traditional +I still have similar prob +Did you manage to get commit access +Another alternative approaches could have been to store tokens in hdfs or mysql server +Perhaps it is merely because I am bad +The user controls the the intensity of the load by adjusting parameters for the number of worker threads and the delay between operations +patch should not introduce any failures +Hi i have make a in your format +verified in JBDS +Now does not implement anymore exceptions pass through +I am still waiting for because of another small bug but it is pending for long time now +Applied patch in Rev +I know exactly where that is. +The failure at apears only every now and then and the failures at are open issues at the moment +Will that generate an ATTEMPTFAILED? Can the node crash and cause an ATTEMPTFAILED +When running multiple on the same machine why not set info port to ? Then it will be dynamically assigned afaik +sure but when I took out the Resin started workingResin is fast you dont need no stinkin page cache +Aha yeah I noticed those and figured they were not needed since there were copies of them in images +Thanks for testing this +Id resolve this if I had the right perms but the new version of surefire is compatible up to version of +Parking in the next CP cycle as this one has already been released and will be taken from the assigning to you for the next CP +I had the same problem on +David this is great! How did you happen to find that out +Robert please close if that looks good +Then we might see which transaction is leaving a lock and causing the timeout +These are usually test which depend on other eclipse pluginsIn these case the two test classes test if the project model behaves correct if resources in the workspace are changed +I am leaving GreenplumPivotal on May th and am no longer involved in Hamster +Further tweaks welcome +Still do some housekeeping +Easy enough lets put this in with all the authnz to we wont release a s do that later +One issue I see is that there is no way to limit the maximum number of threads specified at query time +How about I post my version of the patch nowI have not run it +Exactly the same phenomenon +So we will have to recompile the core without the call to the right now? except there is another way to do it +Thank you it is OK nowThanksElena +I bet were SEVERAL times faster than here +Thanks Hernan! +Hi Yonik thanks for the info I dont see much of a direct link with and Trie based fields so Im not sure what the difference is between Trie based fields and is possible that its a problem with the implementation of toObject in the s +But when you restart Eclipse with new workspace and add the again and set the deployment to tmpdeploy no new deployment scanner is added so it doesnt workSteps +Thanks a lotPatch applied +Ok +Resolving as Incomplete for a few reasons now raises real Java exceptions from Java libraries +This actually rejected by Oracle at least with the out of the box understand this is difficultimpossible to fix and is anyway a pretty silly abuse of the API. +What is the purpose of unsetting the redirect? If you dont want a redirect why construct an +Attaching an incomplete patch for branch +In IMAP its a bit different +It would only help when creating multiple violations via the +There is a menu bar for imports that will be in the my page for all three importsthere will me a import menu in Products and partiesthey all point to the same Screens for import that lets you select the excel file on your desktop click on the import linkthe rest is done by services +Check this issue with please. +I forgot to say that the test failures are not related to thisSee for and. +Jan two issues for converters for complex row keys were fixed +Done. +It would be great in job control could use this as an optimizationYou need to have a method where the application declares that all of the input has been added +issue here is that the aggregate option will force javadoc generation to only occur at the top lvl execution root of the project and it does this by putting adding all of the source roots of the reactor projects into the list of source paths to process +Thanks Thiruvel for taking this up +We can explain this a bit more clearly in the documentation but changing the implementation to support multiple threads is not an urgent issue in my view +Other implementations presumably would not +What about using an introspection cache again? Kind of nasty depending on how soon we could get rid of itPersonally I think Id prefer a break to a solution that doesnt solve every case +Pull it back in if you think different +thanks for testing my previous patch and suggestions +Attached the patch as discussed +Peter has kindly updated the BND tool to handle the include directive in properties passed in from the plugin as well as handle manifest files differently to property files +Maybe we can bump the version for I think the zookeeper team is now recommending. +Added documentation +If this issue not going to fix I would suggest to resolve the issue with the resolution of wont fixIf someone may be interested later he could open it +The Dojo default was chosen early on because we decided to use Spring JS which supports Dojo only +in revision +Uma +Most definitely +Ryan your attachment doesnt look like a patch +vs which you instantiate tweak and use +fraction insteadGood suggestion +Hi TonyIve tried to set LCMESSAGES variable to enUS and still could intermittently reproduce the descrinbed problem on +As pointed out it doesnt work as it relies on the build and layout of webwork and we have a replacement in the jetty plugin as well as good IDE support for running app servers in a development environment. +There was a discussion on this on the mailing list +Actually the list of modules is updated properly but the modules are not engagedSo the fix consists in engaging the module mentioned in programmatically. +Applied the patch +Will integrate in the afternoon if there is no objection +Bulk close for +the output is already in the maven covention default repositoryhowever there is a problem with the since its unable to use the target directory as the holder of dependencies +I originally had my own swizzling code and ripped it out because made the code look nicer +committed thanks. +do you define the property in your USERHOME or elsewhere? and the property with the list of remote repositories +Galder this is working as expected +If someone is interested +Because we just found this issue when we enabled the cache Ill let it run for a day or two to see if any other issues are uncoveredThanks +If you find further issues please create a new jira and link it to this one. +If the intention is to pauserestart a listener then we need to look into extending the current mechanism +Ive tried replacing logj through slfj with jdk logging but portlet container cannot start because of the hard dependency +Try updating it to the one listed in confluence +Yep +Ahhhhhh +Therefore even though it looks like youre running the version of wordcount youre really running the version because of the default classpath setupThats why HADOOPUSERCLASSPATHFIRSTtrue and true are necessary to make sure youre really running the code +issue fixed in iPOJO +I want to make sure this is in line with the Struts tags before we release. +Attach your patch to this report +I dont know even if this would be a store change or a language change + Vladimir for improvement the patch was applied at rPlease check that the patch was applied as you expected +resolved in just picked up the nightly build and it works! Thank you so much +Docs and migration guide updated. +Fixed +In fact we do separate modules mostly at package boundaries +and not days as indicated in the component is entirely finished and works fine on tomcat or glassfish and all browser support. +Fixed format javadocs too +Mona +Youre right that we cant support Jonathans use case but it isnt a valid use case for Hadoop +Patch to avoid loading empty invoices within applied +For some reason I thought httpclient had a separate bugzilla project but as its just a component of Commons the import isnt as straight forwardI will discuss with the other JIRA administrators what they think is the best approach to importing the bugzilla data and get back to you +Were talking about the Admin Console of the code to allow the to be longer than the older default of characters will now truncate at chars. +Start Solr +Integration test updated +A few example queries both source and the resulting parsed query +Hi Francesco MaximilianoI reopened this issue and today Im going to add the patch file for is the new version of the Crontab Cointainer properties translated for portuguese +checkAnd do the right thing +I completely agree with not packing required libs in our repo and using maven to get them +Is it sufficient to prevent them to run on small devices +If you ever get the chance to translate those too let me know +This feature is not supported in Java EE +But this isnt related to the patch committed right? The committed patch doesnt touch any of the affected files +Closed as per state in Bugzilla. +I made fix for this inI was not able to find any information about this in PDF referenceSo I dont know whether annotation subtype has to be handled in different way than other annotation subtypes or notBut my fix works for attached sample PDF file +Too many of the packages under +Problem still exists on sonar. +We should apply this fix to all XTS tests as I think they could all be susceptible +Please check with them +Resolved in r +I dont think this is going to have a good costbenefit ratio +thanks for pointing it in svn revision thanksIm not sure about this change +Dan can you or Van look into this +Considering that all the changes here are going into you can resolve this as a duplicate of +Hi ve given my concern about logging to Ceki at slfj mailing want him to develop new logger API for I close this taskI expect that hell resolve my concern in his new think that there is no nessecity I develop same API +Let me know if you need further infocode +The versiontimestamp column in question however is a new column of DB type timestampI am using the IBM toolkit +closed for release +This patch replaces stacks for linked lists in the implementation of callback routing chainsThe patch passes all test cases in spi core and as well as the supplychain simplecallback and samples and the ws async samples +Come to think of it I think all of our examples could use a review when it comes to running in both local and Hadoop aware mode +If the issue does not repro with the client libraries then it could be a bug in your code +I looked into it and found out why deleteTable fails one of the region is not closedI will put up another patch soon +I am new to the hive community and was not aware of the process and did not see it in page also +Im just waiting for access to vmbuild to insert one for continuum +Tar archiver uses the config from so it was made compliant the minute days ago I also updated to use that setting so it should also be compliant +the actual issue it addresses which originally came about through discussion with some of the guys over at +Better shift this issue to Denis what was the end result last time +Added a unit test that uses a mock implementation to drive the +please stop marking everything as a blocker look at the help icon to understand what these priority names mean +This should be added roadmap +Is this JIRA not a duplicate of? They seem to be aiming to do the same thing +Confirmed it happens only on project reopening and never happens on IDEA reopening +While investigating this issue two more problems were found and +Manually without integration test +To use GRANT you must have the GRANT OPTION privilege and you must have the privileges that you are grantingad +However map is Deprecated as it will be dropped in Drone you want Graphene to drop it now +If dataloss or probs around recoverLease andor append can examine in another issue +I briefly tried a few methods of trimming the problem files size but none reproduced the issue in the resulting fileCommitted a check for null in r. +Ok thanks for the feedback guys Ill push it to the repo +Technically this ticket is complete +I agree independently verified CS seems to work fineLooks like might be broken +I think that would be a great test of our compatibility story +Behaviour as the patch will be applied I can also send you the patch for for PatrickI could not apply the patch +Ouch +I like +Thanks for the patch +Id rather have something that works in and remove the remaining part in in +bq +Updated the feature to use JAXWS JAXRS and Jetty with release versionNeed updated the JAXB when servicemix released the wrap bundle +Committed to and trunk thx +Happens intermittently perhaps because of some permissions issues +Looking forward for the patch +The view geographycolumns in postgres was not readable for theThanks for your suggestions +The only thing I could find to complain about was the use of as opposed to using the logger +Lets target to +Have a look at +I followed the same steps to reproduced as described above +Lets use a separate issue for promoting this from sandbox once the code is stable +Therefor just asking are you sure this will be the solution +Here is what I committed includes fix of Ram comments +Updated patch to reflect +It will still accept w so we have backwards compatibility +Attachment has been added with description logs full url not just relative part if there is a failed warns about base urls not ending with has been added with description logs full url not just relative part if there is a failed sure I like the warning +Most users dont go through Hive Server so we need to ensure that we dont break their use case of using the cli +Imagine a servlet code that needs to call a web service +Im seeding the DS Name with the VDB name for starters +The GSE looksup are not page requests in the first place they are lookups for groovy sources and classes +Fixed in branch under r and under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Committed. +Thats my guess + +AFAICT this dependency only appears in the testsuite +skh pushing to +Carl wed like to start helping get these patches in shape for commit to the trunk +Thank you for taking a look Daryn +need to evaluate this before a Id say to use the There appears to be a problem with generating the join conditions here for join dialects +Stepan Do we have an ACQ from Vladimir? I dont see one +Not using max was part of debugging +Dont forget all the classes with in their name that are missing as well I assume the intention is to use +Patch file to fix dependency + +I see +please delete this I pressed submit before completing the as requested. +Applied to branch in SVN rev +Looks good to me + +Currently the most common reason for a send to stall is that we are being producer flow controlled +Hi AllThis ticket doesnt alter the existing behavior but just involves coderefactoringAs far as the test cases are concerned a should be maintained result of test case execution which is maintainedI know every patch needs to have test cases included but in this scenariowill it hold goodPlease let me know in case of any issues which are blocking this patch frombeing committedThanksMIS +what version of are you using? and how do you start your sample and see the bugIt would get us much faster up to speed how to see the issue has been split into two issues and issues +I have modified the method and the javadoc to accept pos parameter in char length +Brian I dont exactly following the below +Amar I suggested we cleanup the files here because The root cause of the issue is that job files are not deleted upon check failure +So all issues not directly assigned to an AS release are being closed +mrepositorynetsourceforgecoberturacoberturahomesonar +Then paste that back into this issue? Thanks +It would be nice to reference the in any case as that was the original source +I think this would be convenient since during development tables are dropped often only to be recreated +Taking a heap dump yes but hopefully were not going to reimplement jmap next. +The important thing in the decode will be to have a fixed memory overhead +Great find +Tested with the release version of IDE +Thanks in I have just tried to create classes from and the generated classes look fine to me +RobertWhat should we do about this issueId like to release Site Plugin soon +Thanks release note instructions need to updated so that users also install yum install I pretty sure this can be combined into a single call yum install +I have added the remaining missing license information in an extra configuration file src +Should be done now added it to both sets of LICENSE and NOTICE files can you please give it a quick look Marshall +trunk +I integrated the fix for problem but Im keeping the issue open for the second problem +I believe the preferred solution will be for the fix from the XSLTCDTM branch to be applied to the MAIN branch +It requires some coordination between two threads so my first guess would be that there is a timing issue in the test +Thanks for the patch Jacob +Comment crossing again +I will backport it to to branch in revision . +I commited some support for inheritance it needs polishing and I didnt write any tests yet but I think the schema generated should be ok so you can check if it works forI checked on the trunk version persisting a entity I think its representative it inherits from and uses a secondary table and all went well without problems +so when you say that you only add or remove one node at a time +Daniel Let me take a lookScott Its worth noting that projects can include Avro support as they wish just as Avro can incorporate that work as it wishes +Not sure when the regression occurred but it should be fixed in the upcoming release +Yahor could you review the PR? If its ok I will merge it to master +Can you test this with the trunk code or at least? Can you attach your data file? Thank you +JacquesI was thinking we could do that in the Example component only +Ill try and update the per your instructions and test real quick to see if I can get this committed +Andy any chance you could provide a test +BTW Today the SVN server seems has problems I cant commit the codes to svnCant close file mntsvnreposjbossasdbtransactions No space left on device wow is our svn server full with codes +looks unrelated to remoting +let me know if there is anything that is unclear +It has CellsWhy should Get and Delete haveA CBM instead of inheriting? It makes things messierLet me play with your last suggestion +Some more results the change I proposed in my previous comment certainly fixes one problem but not all +Attached improved patch +I think its even more unlikely in that case but still should be done. +correct for applying so quickly. +Could you up load the latest patch that applies to trunk +patch looks good +Instead you should destroy the old one and create a new one for each parse +Both scheduled polling and direct routes are available for eachWhat do you need from us to proceedTheres certainly other areas of the API that I plan to include but I can release those as patches later on +Unicode objects that the application puts into Thrift structures will be encode with on serialization +Marking as resolved. +Is this still a problem +PatrickThanks for the hint +the patch makes sense is at fault for not clearing out those variables before reusing the +I found an interesting item in the bouncycastle dev mailing list +Which version of Maven are you using +If we dont get any feedback I think we should resolve the issue at the latest for the release +For now you cannot compute the confusion matrix with the mapreduce version +not blocking B +fixes the problem +Done +removing from queue since there has been no activity ve reported to take care of this if its still relevant +Have a look at how we did protobufs same thing +Unfortunately marking it as deprecated doesnt allow you to communicate ones intent +Sorry +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Much appreciated +But honestly I did so at least times +Can you please elaborate it more? I would like to understand your thinking +Resolving as per the previous comment by Asela +I dont think the v patch still works as the security profile changes were made after that +I managed to spend a bit of time on this last night +the idea from the description can also be achieved by making a tab notdisplaying text but instead a little graphics that displays the text in the way I want it to appearfor example as suggested in the issue descriptionThe suggestion is not uncommon so we all have seen it +If we need additional tests then please open a new issue +I found the root cause is the makeForm function in cannot work correctly with dependsOn field +I am researching as time permits +I this and tried all kinds of ways to get this query to work but it wontEven after ajusting my path and making sure I have the necessry libraries for rc I still experience this problemIm including a zip file with almost all my libraries and a description of the problem +Its just too few to really get a handle on it +Heres an updated patch which has some basic benchmarking +RobertDo you agree with the +It doesnt semantically make sense and I dont think that anybody was directly interacting it outside of the test casesWhatever changes to Ambrose and Lipstick should be communicated clearly also +I think this has the same cause as +Thanks +But contributors are very much welcome to do soAnd users are welcome to indicate whether they successfully use the current code even with the identified shortcomingsThere have been recurrent discussions on this on the dev ML. +Rather if you make a mistake add a new comment correcting it +Just committed to trunk +Unassigning from myself so its clear its open to be worked on by whoever has cycles +Committed patch a to trunk with revision +Do you know how it needs to be appliedor can you provide a single patch that covers the entire changes to this file +If instance with a persistent identity does not mean the obvious thing then document what it means +Will wait for the Jenkins before committing it +Talking about the T REST orientation +In other words the User Preferences feature could operate just as well without the entity +Verified correct operation in +Sure thanks for pointing this out the patch is now a bit to old +Looks like there is an issue with the patches we have applied on top of stock +I appreciate your +GuillaumeGo ahead and commit if you like +persistence is brokenmulticoresharedLib gets written as multicorelibDir so loading the multicore configuration after saving will fail +So this happens in and but not trunk? Or also trunk +Attaching and with changes to filesM srcdevguideM srcdevguideM srcdevguideM srcgetstartM srcgetstartM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefI hope I have provided some overviews and links in the Admin Guide that guide users in the right directionsOne of the Developers Guide topics I mentioned earlier turned out to be more about authorization than backups so I added Admin Guide to only topics in that bookThe changes to the What next with Derby? topic in Getting Started really have nothing to do with the backup issues except that the topic now points to the Derby documentation +will this still work? So if I had cores sharing schema and cores sharing a different schema will that be possible with this implementation +can you attach a patch from svn diff please +Verified by Sean at thanks so much Alexey. + +I committed this fix in r +I use Acrobat Reader +Hi BenYeah it would be nice to see if the patch still works +Its fixed now +Anyways Ill see if I can work around this +between is defined in the spec as equivalent to AND criteria so that is how we treat it +its designed to be accessed as and the attached verifies it works +About knowing which features will changestay that is impossible to knowBase stayed stable in this lastest releaseIn a next update Base might be the only thing that needs an update and the rest on top does notshould not need rebuilding just because it has an updatebugfixThus Im not following why you would not want to simply keep each released site +Am closing this bug since the pig changes are in and hadoop changes are in trunk this should work once we use the appropriate hadoop release + on adding an API similar to obtain that does not launch a task +Please wait h then reach out to builds as Tony suggests if it is still not updated +I just realized that JBOSSHOME is missing in the environment of the Jboss user +I will commit the test patch to get the test running again but it seems problematic to me that the client jar is required to run the embeded tests or is required for client tests +validator should define target message component in code during request processing +is a standalone program to invoke the service it doesnt actually write out the local file but the exception in question is still generated +This ticket is about the same in multiple files +Improved fixDebug build returns correct exit codeReduced native heap usage for stack iterationsFixed found memory leaks +This patch looks backwards its removing a instead of addingCan you fix it and resubmit +Waiting for resolving and. +As theres been no response can we close this as fixed in +My patch for does cause a minor change in the behavior that Sergey noted in his comment on Case When adding the first child to a node it is expanded and the child node is displayed +Sorry for the confusion. +But the first patch is not right +One query +Were hesitant to deploy this without it being in mainline though +The attached diff provides the following Increased granularity in probing and reporting of large file support instead of defining HASLARGEFILESUPPORT Charmonizer will now define HASBITOFFSETTYPE HASBITLSEEK etc +I will keep it open until this is merged into Realtime Branch +I have already written some examples to illustrate how to call Java from XBecause it is not clear how complete document is needed I will close this issue +Would you consider this the same as adding RSS support +Patch to fix the transitions +That cant change and I will never support the widespread production of multiple three is up the second point and I will apply the patch +Jie Huang thanks for keeping me honest. +The issue with dashes in the name will be fixed in the Teiid Designer community release +Included in latest Deployers + Username mkling +Verified at +Looking at it now +just back from vacation There is no open source project logo so far for Fuse IDE or should I say Fuse Tooling for something to be done please dont use product branding on JBTJBTIS content +Added xmlserialize to clob for derby retrieval + Things that we expect folks to edit in their installations should not be in Hadoops repo +Just today we had new rubbish pages added +Please apply the patch +I downgraded my trunk checkout to rev +the patches for trunk and sure what you meanIm not sure it need to be asynchronous +Please cite tests are happy I am happy So I say go ahead and commit +Anyone care to try the version with the num of groups ? You will likely see this problem from unless a patch is available +test failed without the patch and passes with the patch +At least it looks like it uses from the comments +Attaching +I do not use any toString especially I just want to give the user the ability to enter date without useing tcalendar or tdate +draft attached +Im marking this for our M release though because there isnt time to do it before then +Reopening +Issue tagging notwithstanding this patch looks pretty innocuous +Otherwise the job wont fail until map tasks are launchedI will incorporate the rest of the comments +Hi RobSorry for the delay! I missed your comment and we just ran into this issue again when we upgraded to AMQ +Leandro I havent +The target uses the built in ant task for genkey which can be adapted as needed +for +I will fix that soon +Hi JacquesFor the user point of view BIRT Web Viewer provides many options to play over the report Showing table of content Changing reports parameters over AJAX dialog Exporting data by selecting required columns Exporting report with many options Printing report AJAX paginationFor the developer point if view they can test the report that they are working on by just specify reports path on the address bar dont need to create request map view map screen for testing +After adding the the bug seems to have been patch applied. +No apparent migration issues +WDYT +Sorry for the inconvenience. +Thanks for the patch and description Dag +Im still wondering if theres something wrong at libpthread level that prevents other distributions operating systems from crashing +what do you mean by java land +Hey +FYI The Portals properties button now is replaced by Sites config ve followed all the steps in the description but i cant reproduce this issue +Oh and now that youve submitted a patch for trunk I guess you can simply commit to both branches and close this issue +Not sure that this is good though to hide those inside RowAs the code is written now you can mix and match all of the types into one batch but I need to do more testing on this and we might need to extend the compare code to include something more than rowYeah I guess that it what that means Not really sure what would be the best alternative here and if we are looking to make thread safeI have done some test for recovery that are not in the patch but those required modifying the actual code but will try to figure something out so we can put it in a unit test +I dont think one can do this directly in because text output formats would not work +Should we close this issue +your latest description is different case and its a bug of WTP. +This has been fixed in revsandI apologize for any problems this may have caused +Hi just executed the code I posted again to be sure but I have no duplicated objects in my database only one instance of the object I have created is I wonder why +If they should be then I can create another patch adding a test that does multiple file uploads runs the GC and then checks the disk usage of the dataStore folder to see if it decreased in size +Is Serialize Printing working like expected +By moving them to different annotations you adhere more to the principle the system is open to extension but the individual annotations are closed to modification +Matthias my apologies conflicts with +okay so Im down from errors to only +But that is also consistent with Pinakis direction to have the application control whether these generated SQLs should be cached or not +Nobody will try to allocate until B is exhausted in which case they will all create the new Region and CAS again +Will commit tomorrow afternoon unless I hear otherwise +just committed this. +So instead of having a methodadded hook to clear methods when theyre replaced it now checks whether the hierarchy its being called against has been modified or has a different method than the previous call +Pushed to in SVN trunkOleg +Asking on IRC users in other parts of the world are doing okay +Error is possibly from packet res objectName CNAGENCYIHEATON dominoSuspect that packet res may also error once is fixedobjectName OIHEATON domino +I tend to use multiple classes per file when the extra classes are trivial and supporting which is not the case for Socket for instance +Merged to to docs branch at revision +I am not sure how easy such a test will be other than a stress test like yours though the real fix is probably to fix and avoid all together +This did immediately decrease the time needed to process all our seems to work for the use case which we have for Spring +In most cases user will default code snippets which will be available by clicking by clicking CTRLSPACE not by calling New File Wizard +Please release +Since you agree this doesnt needto be addressed Ill close it and mark it as WONTFIX +The remaining issues will probably require more than +removeDeleted should just drop tombstones +You have been added as an author for the modified filesMike +The patch is made on top of patch +Essentially the selected item specifies the current action for the entire component thats why itself is theI used a similar technique for a navigation component in my book. +Patch to make the message pattern being checked a bit more lenient on the memory usage format +Latest patch by incorporating some more offline comments from Amar +remove these methods in trunkI think this was used long ago before the javabin format took over serializing +The services field is volatile so readers will in any case see the correct value +In any case we must stop using this deprecated method +Closing here as this will now be handled by in and tracked there. +Seems ok to me +I decided to leave the Edit View as it is +closing this issue as this is guest os specific issue. +That may be fixed in trunk but if not I can patch that up its pretty straightforward +This one doesnt include test cases +I have replaced usage of library with Jansson library +The AM may get killed when the state is in DISABLING but the regions are closedSo can we check both for DISABLING and DISABLED stateI will provide my patch ASAP +Keep the already defined in the config in order to remain compatible with JCR only the mimetypes not managed by our current implementations will rely on +Ah I thought I read it was happening during abort +Possibly due to +Freeman do you mind looking into this +this will fix the issue but I wonder if it is doable with a smaller scope +btw I fixed line in but replace the with dd +And from build to build I dont have a perfect reproductibilityAlso if I clean some nodes in the tests tearDown I have some other tests going redI feel ill at ease with this behaviour +From my point of view it makes not to much sense returning a in case of request with applicationjson as it makes things nontransparent +Because these event logs are used for restart I dont think making it pluggable is a good plan +I just tried it with the and it m going to close this as its unlikely we will fix this for +AFAIK I dont have access to edit your submission so Im bouncing this to you to fix +We can just say in the docs to use post that it works far better and will certainly not work for heavy in +Thats quite a long time +Committed to trunk and. +Reopening to fix the RAM balancing problems +This patch reuses the Coord Rerun filtering logic +If it would be helpful I can put the fix on the branch as well +esbrefguidehtmlsingleimagesesbproject +We should turn off invariant checking by default and only turn it on in nightly tests or local builds +Perhaps we could harvest the necessary and validate them ourselves +should work nowplease test againregardsits still doesnt work form me +Attached is a new zip file with the dependency on the missing jar file fixed +in svn +I dont think fixes the case of a daemon being left with leaked leases on renamed files + to trunk +And actually is related to +Its just a matter of updating the cache when restarting the server +AS uses AOP +no was marked as fixed for? not CR +There is no project at all and Roo will create the project from an empty directory +In fact theres nothingwrong with the patch itself +Once you fix the you should have a fully functioning think this captures it even if the socketProxy is only covering one half of the duplex connection +You can execute the code from the file +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Patch available +The original issue in this JIRA has been resolved +for the Java only workaround +My pull is wrong I will look again +This patch should make this instability go away running regressions +To get the test case build you need to run mvn clean install eclipseeclipse against the root folder of the interopdoc +Hello Yousif do you want to donate your patch to the ASF +Unless I am grossly mistaken +Ping! Any thoughts +Hitried it just now on and its fixed please close it +Will check this out tonight +Fixed and IT tested +Hi Andreas +I just committed this +Ok Galder so I will get back to you when I test the CR +Small differences in source code alignment are just a fact of distributed working and only become difficult when there are overlapping or related changes +Getting better support from the would be greatIn terms of atomicity Im leaning toward a best effort +runnable test with mvn test +Thanks for the quick patch Eli +Create table Cat with unique column name and without primary key +Does the inter node communication use the RPC server? If it doesnt is there a need to potentially cap that communication channel as well +Well the fix Ive made fixes different issue +OK so this is all my fault I suckI fixed in CVS +An alternative would be to just print a warning if the permission is not granted +stalled reopening to make sure we test this is still the case +rebased patch no longer applies +Meant to added it to a different ticket +This patch file patches the corresponding test to reveal the problem and then patches the class to fix the problem. +Why I want thisWith this a user will can organize projects like he want and hell can remove the project id for the checkout directory and the user would can recreate their own project directory structure that can be useful for some project with a flat agree I think its too big of a change for its either a model change or a change in the way an existing field is used +We should have a look at for batch operation +Should be fixed is still needing resolution +No response assuming ok. +Resolving this issue to note that it is fixed in the trunk and at the head of the branch. +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be translated message bundles can be added as people contribute them. +But we need clarification from runtime folks +Thanks Adrienbq +we might want to open another JIRA to track this issue but still dont know why cannot be injected +This is a zipfile of an Eclipse project that implements the beans plus Java source to run tests +Thanks Mark for taking care of this issue +Works thanks! +If no one objects Ill merge this to +Proposed SOA TP Release Note no longer uses the default session +should help stress test +Im struggling to use HTTP authentication +I think for now we should just leave it as it is without implementing the lock interface and see what our users have to say +is async replication is sync seems the NAKACK error is diffecult to reproduce +Assigning Brandon as reviewer because he was initially involved +Checked the and bundle other software +patched for ordering and location of the Clerck servlet Jaco please also subscribe to the list +Oh Im happy to do it +Since the Xerces repository evolved since I provided the patch the first patch does not merge anymore without conflictsSo I herewith provide an updated functionally equivalent patch solving the merge conflicts of the older patchIt would be nice if this issue could be solved soon. +scratch my comment this is fixed in HEM this fixed in HEAD? We use HEM there +Any chance I can have more detailed repro steps +just add the providerurl and it will work +exiting from eclipse just wais forever since the stopping job is also not stoppablecancable +I updated the patch +Added identity keygen test to and started test test case passed post patch to +apurtells fix of seems to have fixed this oneRyan is fixing thrift server testMerge is still broke +Fixed. +Have you tried posting on the spring security forum +unofficially targeted to kepler was pushed to sr but is only in sr rc as far as I can tell +defer all issues to +thx by Toby +Patch looks good thanks ShaiIll commit in a day or two +r committed the latest patch +I removed it too just now. +From above comments apparently not a problem on +Patch that adds Ignore annotation support for test methods and classesAlso adds output when a test is skipped due to using the existing annotationUpdates annotation so it can also be used on individual methods instead of only at the class levelUpdates so the user can supply an Annotation class to search for instead of the default and provides usage details if an arg with ? is given +Is that possible is part of hadoop but is not? Surely large is not good +Maybe do this for too +Improving the documentation is always a good thingNo objection at all on the contrary +I was not able to reproduce it +Of course I have solved the problem for myself +Will do sorry about this + for the patch +What email address were you trying to subscribe withCheck your junk mail too sometimes they get caught up in it +ATM these are nothing more than a random collection of microbenchmarks without any unifying idea or logical structure +is our standard default password +Yeah I checked now and the branch is there thanks Pete. +Uploading is failing at +This happens because the patch changes the command protocol and the different versions of the would have different versions of the protocol +ekh +Thanks for the review Stack +May I have a binary copy of with the patch applied please +The same patch for +Adding release note section of to this release note. +Ill check on that first +However in case we fall back to backport we shouldnt be accessing it through reflection +Our team here used above given instructions for testing on Postgres +This issue is blocked until the relative path refactoring is ported to not fix it in trunk +Would be nice to try it out +Have recreated on manually and retried passed again +Hi KiranI am checking how to connect through +Closing as this works for me +The other choice would be to change the API of the content object to cleanly document that a other object that has a different behavior than the properties object is used +No Kathey thats the one that I tested last night and it failed +any performance impact +The default is to allow any individual header to be at most characters +The reason cleanup and compaction are different things is that rows that do not belong to the current node can be generated by writes as well as left behind by token changes +Deleting an entry now flushes the main page cache +Sorry for the noise Ill move the issue over there +Fixed on trunk revision +utf files but thats not obvious +I really object to using system properties for configuration when we have config admin +If you find further issues please create a new jira and link it to this one. +Since there is not much of use case to look up a divPanel I guess we can lower the priority on this one + version of the patch generated from the top hive directory instead of deep inside s ee test infrastructure +QE please have also reproduce it on container. +in to see only SUCCEEDED FAILED or NONE filtered messages +Instead it should just try to compile and link a test application to detect if libcurl is available +ThanksSending CHANGELOGSending libbuildrcoreTransmitting file dataCommitted revision . +I agree that a compromise is to continue the deployment and WARN the developer +Based on Ray feedback I think we dont take a big risk updating +Simple HTML app no server dependencies +So this is using the default typing and +bulk close of all resolved issues prior to release. +Let me see if I can reproduce it +seems to happen when there are no nodes in the cluster +The page includes the differences listed by Mamta +This issue is fix in FUSE ESB release +If you finish it before we release beta please correct the version +But users of Solr plugging this filter after a will have problems +Im not sure how useful that would be +Try to set it +If you find further issues please create a new jira and link it to this one. +Since my torquebox server is behind Apache if I disable Apache then I can deploy without a problemI also confirmed that removing standalonetmp allows for a successful deploy +Teiid has committed the changes to the branch that will make the defaults and the recommended value the same +Postponed to +Filed +For example I configured Apache Maven in the meclipse preferences +Killed off a zombie process should hopefully take care of it resolved this problem! +More tests for to cater for phrase clauses that produce no Junit test to test for phrases with clauses that produce no do you think about this for Mark Hbq +Two databases should be created namely umahost and umaauthzmanager then use the files to generate the database +once a couple of small changes are made to this patch we should be good to go +Adding as a dependency to getting this done +this looks good to me +Thanks this has been merged into both master and +Hello WernerI found the defined by the system administrators install a usefullness mess and they append a JRE at the beginning of my PATHWith my regular JRE it is working to bother problem +However I believe this is an exceedingly rare case +Doesnt look like terminate fails to me +I have moved it as per your +Patch was applied to Apache CVS HEAD +This bug has been fixed in Xalan Java +Once you have provided those instructions and I can verify that it works the bug can be marked resolved +Could do this in +IMHO simply moving them out of the mapreduce package is a good start +If the underlying database supports the UNION keyword Hibernate appears to accept it just fine +I looked at the patch and it does not seem to include +Loaded VM Summary screen +Will retry in a week or so +FROM DESTquery SELECT DEST +Attached is a test for time and date values and a fix +HelloWe have exactly the same pb with Struts into Websphere ZOSHave you a prevision for fixe date? When the release will be publishedIs it possible to patch ourselves pending this releaseThanks a lot +However the Summary metrics table for datasources is displayed used r and Branchx r +Thanks to Jerry and Jean for their inputs and review as well. +It may be a problem with the wrapper itself and not Apache +The users would have to keep track of if any keytab was generated for a given principal to know when to use the norandkey option +Resolving this as fixed since we now have a basic performance test suite in place +Where is this file located +After two code upgrades this would crash every couchfile in the VM simultaneouslyFourth as Ive mentioned numerous times before the proper way to synchronously start a process that might fail to initialize is to use proclibstartlink and proclibinitackFifth has anyone considered using a write buffer outside of the couchfile API that would allow clients more precise control +Can you take a look please +As you no doubt noticed there are plenty more examples of this issue in the tree so feel free to chip away at this issue as you have time energy and desire +Also added a bit of javadoc to patch looks perfect to me +M +Patch demonstrating the implementation of a generic insertion operator for const wchart +Maybe this is some internal name used by JBM? No this isnt an internal you using? Looks like some queue created by that +I missed a couple of places where we construct new operations instead of looking up the ones in +Even if due to some rare zookeeper session expiration issue it is not the leader the test will not fail +please change the java main to class +I also removed two useless log guards in +So all issues not directly assigned to an AS release are being closed +This patch fixes the issue for me +Please see +Fair enoughSo given that does it make sense to open this Jira under Dev cockpit instead as an enhancement? I still feels the functionality addition is justified. +Verified on EAP ER +Abstract beans are very useful for defining common attributes that other beans use as a parentIm still trying to understand the need to initialize all the beans in the factory? Whats the purpose +Updated patch addresses feedback from Dan and Deepa the test should close the statement objects at the end in order to avoid wasting client driver resources and to match standard JDBC style +Ok +The existing hardcodings are equally has a different module layout structure so dont assume that will work if you are hardcoding this +I think the correct approach is to check the derivation method of the complexType before creating the +Added a new test test which contains the test cases from the reproRunning regressions +Closed issues related to Roller release. +Should be fixed in recent code +using startkey and endkey works like classic RQRight but since you dont get Token strings back when you query you cant requery using a new TokenEDIT Hmm I guess for the second query you can use a startkey and a finishtoken +Whil reviewing this patch noticed parameter description is incorrect for all the backup procedures +Uri the patch is missing a class +to +This patch resolves the problems with forrest docs for c client acl api not matching the current file +Other Execuotr implementations do not have synchronized keyword may be we can remove this +In any case its almost certain that whatever it is youre doing theres a better way to be doing it +its a attribute of +Thanks +for the patch +You can copy this property file from the original Jar to the patched one solving the problem. +Robert please close if satisfied +this is less trivial than I originally thought since in order to use a you will have to have access to the other dependencies from the project that created the +Untested patch +Heres a start +What Im noting is that on line the project has no primary runtime set yet +Also these patches arent originally by me Ill add in the names of each of the contributors in the individual patches theyre from contributors in Microsoft who developed against hive and asked for my help in reviewing and to trunk +Couldnt reproduce +Barry has fixed this in other text fields before +I am also running into this shouldnt this be fixed somehow? Either address why it suddenly increased or just change some build settings perhapsReopening this anyways. +Tests are currently passing after having manually added the to AS currently waiting on Marcus to add system prop support in security domain declaration of userrole prop files +The UI just says fragmentation +Sounds like a good solution to me +bq +Reducing priority on externalizing cluster project until AS is out unless there is a need related to AS +The patches are a way to see if integration with Ocean is possible +thanks for the test! the test fails for me though +apk and logcat m going to send a simplified apk sorry give me about min +Same patch updated against latest trunk +Please use the Roo backup command and attach the zip here +r should fix this but that depends on whether a deadlock error will cause the entire transaction to be forcibly aborted +This just seems like something that roo is tailor made for removing the need for such bulk manual editing +Please feel free to reopen if you can get some confirmation proving otherwise +Heres the concepts for the shirt +I investigated this issue in detail +Fine +libs +I could not reproduce the problem anymore +So the check on win is clean +Please explain me more in detailThanks Brett! Looking forward to hear more from you +Fixed in CVS +I have done that so your test now passes and converted all the tabs to spaces +I meant you can try hive trunk and see if the error also exist +If there was a rollback and a flush then there was a system or internal error so it seems like all we can do is barf and ask for user intervention + +Reopening +The collections are lazy after resolution from the query cache +Hi Gareth ClausI have made the changes you suggested and have checked in the code into the Camel trunk +Thanks +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsermodel +I believe you would be able to figured it out quickly since its your code while I would need some time to know your code. +How do you know that a key will exceed a threshold until after youve passed itI am thinking this can be pushed down to the deserialization code in rfile +Ilya were you able to verify this fixThanks problem thanks a lot +I ran into this issue just today and it practically forced whole app to return to previous library version +Axis has been released some weeks ago so its time to fix this issue +Perhaps repo should leave the bussiness of maintinaing that jar to s what i did but they changed it after sync what can I dothe only thing that can be done right now is upload the good jar as version state in the pom description that its a fix of you remove it from repo +doc done. +Sorry my bad +Looks like the empty file caused RAT check to failr tyu staff Feb srcmainjavaorgapachehbasecodecprefixtreescanner closed. +Has been fixed. +The sample imports +Verified by Sian. +Sreekanth sorry for my distracting comments +Ive checked in a fix for that. +That of course fails as a result of the core changes made by this +Thanks Marco +Bulk Close for all issues before incubation +It looks like this wasnt committed to the CP branch so I went ahead and committed the fix along with the test case +Ill upload the patch tomorrow +So all issues not directly assigned to an AS release are being closed +The dispatcher needs an aggregation of +Thanks for fixing this martin +Ooops yes this DLL is missing in the archive +Its more important that extensions are working and that users dont get weird errors when they try to update the extension +Attached patch to fix errors in build due to change in fields in +Hi schemas were properly checked for case but not for Applications so I added that +Yes sorry I forgot to mention that +It would be good to read through archives on where we were + thanks Jens +Yes agreed what I meant was the logic would be the same as a merge even down to configurablepluggable behaviour +I have tested it in current trunk all works fine +etc +Heres what Ill commit +I dont plan to work on this any time soon +I have applied it to the trunk at rev +branch merged into SPCP revision . +The Ant task only uses the module property to load the manifest which is already available in the NBMs Infore probably better off writing the code you need directly rather than trying to call into some old crufty Ant tasks the logic is quite simple anyway + +Other than that just merging +Hey Owen is the latest patch acceptable to you +attaching the diff of the modifications to the as the patch is the new one pls ignore the previous SandakithI tried to apply this patch but the original file has changed a lot and I can not simply apply this +Soemthing to think on is doing some more docs +Ive created the relationship link. +Q +number fieldName volume type class +Attached is a very simple patch that adds the required null checks +Ive added some test code that ensures that this does indeed work +Add support for background operation +Thanks Alejandro! +Counting down from failures to failures +This is not what optional means +I do think we should do an realm for +Nick license header is now fixed. +Hey Maurice Ill take this now and update the User Guide etc +Ill look at EWP and see whats the problemFrom the log I could firstly give out some clueshomebrqpslaviceewplibendorsed is this one a default included jar by EWP? If so then the should be excluded from resteasy test +I also wrote Ruyue Ma for the of his patch above but no answer as yet +patch setting log in to be used by all reports +Preliminary patch +I dont see any problem when compiling the generated code with exactly the same wj command line params +Ive upped the httpd timeout from mins to +This has arisen because each doc is in a different git repo and it is a huge hassle to fork different repos make one tiny change in each repo push all the changes back wait for each dev in charge of their repo to pull the change in and then run the jenkins job again +bq +Will post the results as soon as its done +the patch looks good +is it ok to propose patches or do you like to have the little improvements some other way +I do think you should ensure that the returned module isnt null though. +Believe this is a ESB issue +Which version of Spring Batch are you using +I will apply your change into our local patch also for now and wait for it to go to master and be eventually released officially +Im actually having this fail with fs directory as well same stacktraceJust that functionality is not affected in this case as it kills the thread in the executor after having applied the work on the index +Even after deleting the dead build agent the Queues page still shows an the system is still trying to get information from the agent even after it has been related I notice that the config file still contains this deleted syncs up the file with reality and makes the Queues page work should not be necessary to in this situation +I actually like that better than shutting down the NN it fails the particular operation without shutting down the NN itself which allows the logger to recover if it can +Marked this a blocker as it can break functionality of reclaiming capacity +However after that i trapped into the same old issue of having on my entity for which I am creating by Error creating bean with name i threw exception on object creation nested exception is No supertype be you can attach the jar +If an issue has not be assigned a type Select type will be the first value in Type list +Im afraid I dont see the need for maintaining two independent ways to manage java routine securitybut I think you are proposing two different security mechanismsI think you are proposing that if I have a jar file then I can control USAGE on it with GRANTREVOKE but also USAGE can be given to others without my knowledge by the dbo granting the right to set the propertyIm saying that if I have a jar file then the I control USAGE on it purely with GRANTREVOKESeems to be the former is more confusing +Attaching +Added Http level compression based on the patch Claudio provided +When executing the following command from the source directory the views sqale scm activity +Perhaps picking any type in the AUTO sorting code would work since it wont yield any hits +By leaving logj unconfigured we force the application to become interested in logging +Thanks that explains itI was just curious about the apparent discrepancy +Thanks Jonathan +Hi Can you attach a diff instead of the whole fille +If NW is unreachable from issuing refreshNodes and Standby +Closing as per reporters comments. +BULK EDIT These issues are open to be picked up +Thoughts + +add and These were well tested with release therefore skipping with this release as no fixes in this area. +NTLM proxy support in is indeed broken +Let me know if the attached patch fixes it +Some examples or even a link to a basic LDAP reference might help +I have the exact same issue as Olaf +Thanks Nicolas +Required or Length +Attached is a reproduction for this issueRun the script with and then the program with The test case can probably be narrowed down some +to the JAVAOPTS because them once more resulting in bogus command line passed to the is very problematic BTW and it will break posix as can try by setting JAVAOPTSsome value and executing t make it for CR. +Andrew Stefan this issue is unassigned could you introduce it in Roo roadmap? I think it would be great for Roo GWT community +On a more deeper look id suggest not to have hardcoded command names and environment variables +this feature was pushed alive on Oct +But not sure of the right solution right now +As Andrew says you need to confirm what versions you are using if you are using a groovy that is targetting groovy then you need to be on a dev build not the release +Anything else +batch transition resolvedfixed bugs to closedfixed +noarch we have no control over the yum package what we can do is to update the instructions in the Release Notes +Of course the tool requires that this table also have no other columns +Sample FO to replicate has been added with description Error is a sample fo +Trivial patch to fix the issues hence no tests added +Fixed in in arch gogo +Anton see Seam perspectiveSelect ProjectNew Seam Generate Entities wizard +Can at least name the protocol on the patch changes the add dialog to show the protocol +This is a Blocker just like all the other EDG eccp failures that prevent the library from building +Also no title or additional buttons are allowed +The issue is not the news on this issueDo you understand what I meanIf you want I can patch the plugin if you are ok to include it in your next release. +Okay you are right after spending some time on the code Ive patch it to be compliant with new API and it worksThanks +is the file on the classpath +It assumes that the war packaging will be responsible for creating the final lib directory +verified by Paulex +Tested the patch and gave it a go and it seemed valid +In the patch a flag is is introduced to make sure that no one can write the edit log buffer once it is determined that an automatic sync should be done and before double buffer is swapped +This means that the interaction between and would be unchanged from the current implementation +To doug +Already done +Changed wait to RUNNING in Junit the seems to only produce applications with the state of ACCEPTED +which build you are using i tested it on latest build seems no problem when adding both recheck it thanks + for patch +The problem we have with associating the threads automatically is that we have no way to unregister them +I dont believe it makes sense to busy wait given that it is so small duration +and all seem related +Sounds like this affects all of Windows +What JDK version do you use +Congrats first week first bug Thanks +This will be available in the next snapshot of +Removal of a dependency from a project does not remove the library from the Maven Artifact Repository. +Web should now have a dependency on all EE components from the same module +Wasnt handling the mode properly +Bulk closing stale resolved issues +Great catch Hairong! Thank you +Reuploading same patch because it looks like it didnt run all of the tests +Committed patch in revision +Alas it still only helps somewhat and might not be worth the hassle +Also fixed the limits as suggested +The right fix for is to use the kill task to kill off the redundant speculative reduces +INFO Running for rows took msNow number wiseHere is the average for before your patch and after your patch +Committed r. +Hello MaciejArg +So meaning if I were to use I need to sacrifice the use of property +Cheers +? Actually incrementing the AS version like this in EAP releases was Andys original recommendation which you were opposed to +what version of maven are you using +This patch removes the methods Knut mentioned in the previous comment +An authentication provider instances can now be set when a session is created +Luke would very likely work if you used it with the Solr version of Lucene rather than modifying Solrs schema +applied patch Fabien can you test +Changes committed to branches and test +Thanks a lot Paulex +If this is correct shall we first upgrade Datanucleus version to and then switch to Datanucleus functionality instead of our own retrials + patch looks goodFor a longer term we should fix hdfsproxy so that it wont depend on external urls +The first one is the actual pluggable config that the uses via +Yes I can do it but should i do what more to specify where is the error in what page what bean +If the host responds with I think we should +is using the Hibernate version directly from as opposed to the one packaged with +This failed test is unrelated over your change Marshall thanks for the extremely quick response! Some questions Why in createcompletionentry do you LOGERROR to LOGSTREAM instead of LOGCALLBACK? Sometimes logging to the callback and sometimes not seems like confusing behavior for an end user +i think the is right applications shouldnt throw this kind oferrors +The way the lookup works is an optimisation if the code worked correctly this behaviour should be the same +maven quick start with a upload button and a working multiple instances of your are used inside a form which of these on should be called +The servlet could take a similar approach +Okay mavens parallel execution on testcases tricked me once againJust a short note must the stay on info while deleting nodes? This is verbose I put that on DEBUGThank you Suraj! +apachectl stop +Attachment has been added with description resulting pdf file with extra unwanted is normal and expected behaviour +But uncommenting either the OOB or timer executors leads to issues. +Do we also want to trim the redundancy in? A wider variety of classes use and would have to be modified if we change that too not that this is a problem but do we want to leave in some of the client convenienceAnswer Yes trim it all +Im going to commit this momentarily +Committed trunk revision +Changed to a snapshot in rev +All the plugins related has been moved to plugins directory +Cosmin Can we close this? what we have checked into hbase and what we expect to run on now has +Are you working on trunk version +The javadoc should be updated to reflect this +compilation log is at r thanks! +There are some minor inefficiencies and some code +First Draft patch for comments +Thanks +Should be fixed now +Not intended will remove on commit +in a directory which contains eclipse projectI have patch which fixes this issue +Yeah thatll work +But this still isnt addressing the error which concerns what Hadoop sends to workersPlease see the discussion above +ArjenStill working fine +The classes and interfaces have already been deprecated as of +srcjavaorgapachehadoophbasePlease let me know how to address thisthanksSubbu +I dont know why this was marked closed by anonymous dont think this is really doable without reworking the AJAX strategy for Wookie which at the moment is based on DWR reflecting Java +Just because another thread has logged an error and published all the buffered events doesnt mean the current event should be ignored +fixed with with +Can you post the updated archetypes as a short term workaround +The documentation looks good +expected version is ready for backup +Merged into the master and branches. +I am also experiencing this issue and am working on a patch that will add the ability to specify the fetch size to the configuration options of the relevant classes with the default being +I think the issue is with the printedpublished documentationFor example when I print a pdf copy of the spec the only thing that indicates a version is the title Avro SpecificationHowever the spec is much different from one I printed several months ago which also had the title Avro Specification +Was unable to see this using latest trunk perhaps it was a bug of M? If it still happens please do +You misunderstood I cannot mess with the class itself +I dont believe it is near the top so I am lowering to a minor +My code has changed so I cant be sure the problem is fixed but its very likely it is so Im fine with closing this bug +By default it should be empty and this means ALL the types +Lets try to fix it for to +This is not true +I thought this was the same behavior of the prior get aDir implementation but I could be wrong about thisIn any case I am OK with leaving it as wont fix at least this jira now documents the fact that get aDir does not work in the current versions of Hadoop. +Xerces does not +I see that both gwtcompile and gwtsoyc have skip parameters so this is sufficient +Thanks Thiruvel! +The release notes made me think +Pete I suspected that But may we have any problems having jars in lib folder and in list them in in the same time? It seems to work in my +I think it might be better to just have SQLXML except perhaps in a concept topic like cdevstandardsxml XML data type and operators Where most of the overview and general information should be +marked as a duplicate of this bug +I tried this a few notes P deb runs on OSX but P rpm does not +Fixed by moving the markup in the +Attaching first attempt +Committed to TRUNK and dev branch. +ATSROOT vs TSROOT TSROOT is the default +After created applying to reverted source code and confirmed applied source codes can be build +The best solution I see here is adding the method you need to be inlined into Inliners bonus tableModification of XXXXis overkill here. +How about an RSS feed for a topic? That would achieve roughly the same effect +forgot to choose asf with patch with Vishal This jira is marked for but the patch only applies to trunk +is this ready for review +Thanks Sharad +Moving this issue post Needs a patch file some description of the param in and more importantly some experimentation to see how it impacts the performance of the noticed this open issue +Slfj is definitely the best API around there +please refer to patch submitted for bug This bug and are actually caused by the same problem +user reports it is not the attached two files to srccomjbossdvdseam of seam example dvdstore and running the example can replicate the Windows XPJDK myfaces or +Good point on the config name +The patch is for the Groovy branchShould it be merged in master and the branch as well +Committed with it removed +I believe this has been resolved with the resolution of +Updating the patch +I will ask the SAAJ mailing list about this and probably write some workaround for that as a workaround you can use Axiom rather than SAAJ +Fix in trunk now allows me to run with an IP address set to the ENV var for the host +I only found that I have to do one additional thing +meta refresh tag in dist build queues pager of branchr of trunk +Sorry for not noticing your patch before +Hey Jitendra the patch looks pretty good to me Heres a slightly updated patch which is rebased on the HA branch and also makes the following two minor changes Make the mini cluster start zero datanodes instead of +Which in turns is the result of will try to take a look at this one +I give up on trying to get jira to take this the way I mean it +However we dont do method interception now only field addition we will need as well +Thats a screenshot of the labels info form +Anders I think you should post your diff here so the maintainers can review it and hopefully incorporate it on the official build. +Please provide your valuable comments on this +Automating it with a script in bin would be ideal +Reuploading patch for precommit test to pickup +Current changes are only for Axis need to do similar things on CXF integration +There may be other errors in there which I cant see through the blizzard of expected SEVERE warningsLet me know if this helps or if it just moves the problem around +I will have a look at it this week end +Ah ok thanks for the information didnt notice +That is why we still havent committed the patches to trunk +Please test with the latest CP zip and close this jira accordingly +Adding a filesytem API is fine like I first mentioned +Biggest question is around +What platform were you having this issue with +Verified in build r. +Fixed now with +Also I searched for Domain Object that turned up nothing nor did validation reference +If you find further issues please create a new jira and link it to this one. +So does that mean you agree that this is a configuration problem and not a bug in? If you agree may I ask you to close this bug report as invalid +Committed revision +It occurred to me too +Will look into the one now +You would rather have them always having to use completion twice to get proper completion instead of just enabling it upfront for Criteria Editor +Can you port the unit test to the patch too please? +Patch fixing the bug +A new patch has been attached +Thanks Torsten +It seems that this is an issue with how WAS writes the responses headers +Build code on my laptop +Deletion of app znode should end up deleting the entire subtree +The underlying issue seems to be in handling the dependency between and wanting to tolerate failures of some of these and not othersGiven that adding the config wont solve the issue completely I agree that it is not a good idea to fix it for RM restart alone +I can see the kinit and issue going either way +if not this might be a nice s DejanThanks very much +We can that discussion but it needs to be discussed more widelyI also assume that the change to fix is accidentalIm glad to see the move to use methods on Token I think that is appropriate +I already d it +Mikio Can you construct a example for Igor to work withThanks +Remove local files out of the diff +Martin is working on a patch +Since there has already been a request for internationalization support Im adding that issue as a dependency of this one +Just to confirm +it +For business blogging where accountability and audit trail are important I dont think we can allow the deletion of users +I thought we would be having one in par with the other + +If we use the procedure you describe then well also need to modify version for in every time a new build and new qualifier is produced for +patch with just the stopped check +I was not able to reproduce this in +on your client try that might help minimize the duplicates +Formerly the code only worked if richfaces setup was called firstRichard +I think the qlmapred directory is created when running unit tests +If there are disks on a data node will it be legal to create volumes of type HDD on that Datanode with disks each? I am guessing that the volumetype list for a datanode will come from config +Thats not realistic +bulk close of resolved issues. +Thanks everyone for the guys +Follow the steps below to see the behavior +Looks good to me +Better patch +we put ESTEDT in the text and move from. +mrepositoryorgapachemavensurefire cDOCUMEcsanchezLOCALSTempsurefiretmp cDOCUMEcsanchezLOCALSTemp to attach the olivierI understand there are other workarounds from the users perspective that is what i am doing currently but why shouldnt it be fixed in the future release of surefire +Fixed and committed +If a way does not exist then I dont understand why this issue was closed +Perhaps also ruling out the owner type would be useful +We already have the karaf file which lists the bundles that will still get started after a clean +The current design is much more flexiblebq +If no one mind I will commit it in the evening +Version should be +We need to remove the working memory reference from the logger +Any comments on this? Im assuming this isnt on the table for +Simple eclipse move plus change it to abstract class and add factory method +Reviewed to work in portlets fixed case when two input fileupload are used on a single form and an file size exception occur fixed problem with ignored getLabel property on core +I cant find a way to reproduce it +Yeah just got accosted by Koco for being a sticklerIm leaning fairly far in favor of just abandoning attempts to handle reloading +If you find further issues please create a new jira and link it to this one. +on patch +The s deadline is carefully chosen with the idea that a randomwriter job with maps will need at least heartbeats map map and +Bringing into +Problem is resolvedThanksGeaaru +Thanks for the review Kihwal +We also some way should provide compiled for bit windowsBoth bundles for Indigo seem to contain the for bit windows +I observe the same as Chanwit on Win XP SP for JDK and +perhaps there are a few places that miss it but for the most part all statements should be logged +I think we should have one for this patch but it would be better if the test dont rely too much on internals since it will be hard to maintain the code moving forwardThe reason I create the first version on is because we still have pending on trunk which may require major rework if that patch get committed +We should probably convert that into some documentation improvement +Patch incorporates suggestions from comments +Please excuse my ignorance and thank you very much for your help. +Thanks +This looks fine except for the spurious whitespace change +Also note that I added an INFO level log message announcing which directory will be used for the hot deployment classloader +Please note the test will only fail on platforms without configuration changesCan we leverage the lucene constant to simply skip the test on windows +I think this information was added a while back by kenney when he worked with the build queues hopefully just need to percolate it in rev +So that failure is being said Im investigating the balance of these problems +Everything seemed to be working and therefore Ive committed your patchThank you for your contributionJarcec +By enabling WARN level logging for the package you should see a warning when the beforeCompletion failsShure but logging the original exception is not enough +Problem occurs on as will have to correct my previous comment +Fixed for the patch +Would you like a new patch +So if things dont suit their needs they can make the name explicit +this one will still use the add parameter unlike the prev patch where it was triggered by managementArtifacts. +I dont think much of umbrella tickets +Thanks ErwanSo you confirm these comments are not of any interest anymore +I would like to support this too as we have the same problem as what Mark has mentioned here +This class has been implemented but no unit tests have been written yet +Thanks for pointing this out! I just resolved it by adding in place of ref +Rebased for trunk. +To do this in detail will require GUI changes to add or Set to NULL widgets so we wouldnt have to interpret the String values entered by the user +It already passes good +Please verify. +For we decided that users are not allowed to call get twice on the same column for the same row +well yeah +Possibly based on time of day the test is run +Looping this patch over night I ran into thats a separate issue +With revision there should be no old files in trunk now +And that is what my current patch does +Im not able to reproduce in any way by wangc reported that he can hence assigning this issue to him +For what personal purpose is this for and the likelihood of such usage to occur? CommerciallyWhat are the advantagesdisadvantages of the coupling? Or is it a hack? No offense I am just trying to get advisejustify +I dont really care much about shorter and I dont buy the prototypinglearning factor +After determining that Sphinx or the is the culprit for causing the errors during the build I would recommend two things a either revert which introduced Sphinx or switch to another document generation technology and b switch to a normal maven site build where the site is only built when you run mvn site and deployed with mvn sitedeploy or mvn site +The problem lies in the GMS transport apisdlls in conjunction with the JDK and issue I am facing with also +in the base component references classes in the service component +Hebrew is the only RTL language that I know a bit of it so it was my main test languageWhen we get to it Adobe will donate the new Spark tooltip which provides even better support for all kinds of tooltips including RTL +Are you running the Quickstarts from against a server by any chance +I will provide an additional patch with a junit test for trunk only +How about if I add a config option to control whether the new behavior in this patch is enabledThis sounds ok +Fixed +Thanks for reporting and plz let me know if you face any other problem. +This is also a problem in jruby on master in ad and in db +confRegards this occur in linux as wellAccording to the given description the isnt following url should show thehttplocalhostaxisservicesVersionxsd +Disabling functionality of course has the risk of regressing someone that is using itIf running in the same JVM you can also shutdown the engine with or +Also its not quite as simple as ripping out M and replacing with M we need to continue to support M for compatibility with old data files +Most people have their workspace in so will be breaking the properties as you mention +In this case the internal check done is certainly not goodSame for the second issueEach time I wonder why we dont go for a RC you bring new bugs that make us feel we have made the good choice when we decided to release milestones instead ofAleks I must say you are a doing very valuable work +Hi Raphaelle thanks for the fix didnt really know how to fix it and had no priority Ill apply it when I am back from Easter holidays +Whats your problem? It works for me mvn X testForking command linejava classpath CDocuments and Settingscsanchez +I think this would need to acquire the write lock before starting all and releasing it once they have all been started +Please review +doesnt show up anymore. +I tried to insert the time zone support into the service engine but I ended up chasing down a lot of bunny trails and I finally gave upThis is where I could really use some help from the gurus +Clean Comparediff w +The first time you invoke the expression it is compiled using as a reference the type of the object you use to invoke it +In . +I too would like to see this happen +Unfortunatly I dont speak anything other than English so could someone who is fluent in Japanese update this file +Including not only the source but as well the documentation and Unfortunately I was not able to generate a test TAR archive containing sparse files even invoking GNU Tar with sparse and all the TAR archives I have with such sparse files contain information Im not allowed discloseIf someone is able to generate a TAR archive containing sparse files and provide such pleas let me know so I can include the necessary unit tests +But these mails are not sent by bugzilla but by +The code fragment looks fine to me where did you place the class? In a jar file or in the war? Is there any output in the log file +On the if the class is ever made public it may be useful to have its intended public API and even if its never made public declaring some methods public has some documentation value +The fix says! I see this issue with version of the plugin +For example I put the sonar exclusions below in the additional propertiesfor the plugin + +Thanks Scott! +The message is ok. +There is a use case +Patch with m default preserved +BTW why do we have TNV profile collector named IMHO makes sense to rename it to TNV profile collector is a kind of Value Profile Collector so generally the name is correct +Later might make sense making abstract making a and as appropriate. +However after some discussion the needs of the Directory studio product seemed not to require this modification however the control can still be implemented later on since it does server a purpose +Committed in. +Actually the +Are dependency projects taken into account when is kb built for the project +Austin when you say our leader election unit tests is that something you could contribute back to the Apache ZK community? If so please create a new jira and attach a patch because Id love to include them +bq +Easier this way than providing feedback +Ive manually tested this in a enabled cluster as well and caches were shared as expected in both a service and message processors. +My goal is to have my own that will deal with the output at the last minute +Im inclined to agree with Aaron +It must be negotiated as to speak of custom headers +Unit tests pass +Verified in Developer StudioVersion Build id Build date +I am checking the failed tasks +Hi Ferdy +This time in html format easier to read I coverageio and imageio modules in unsupported directory on trunk. +Then it is a CLI issue +Robert why would you want to +Using a library for arg checking might be a good idea Ive filed to investigate +Everything else would violate the specification. +Yes GWT needs to be in otherwise the GWT examples does not work +Im not sure I entirely understand your use case but couldnt this be achieved by fully wrapping a transport without any lib code changes +KVM agent instance is invoked through java reflection need to initialize logj in proper place +Not able to reproduce this issue on the latest asfmaster code +I have some time this weekend so I could apply the patches unless someone objects strongly +This file was extracted with no encoding specified +When using directly youd have to pass a custom into the getValue method with that configured to know about a systemProperties an example of how this can be accomplished have a look at admittedly its not quite trivial + +Please reassign yourself if you would like to work on this issue. +I think the case that it was trying to cover is when there is a child view open but it was always returning true even when there wasnt a child view +I found the missing feature +for and for we do a test for this before it goes into and trunk +In both there are sometimes patches that unfortunately languish because while theyre important to the contributor they fail to sufficiently engage a committer +That would suit us perfectly +Thanks for opening this I will have a look into this +There is no RPC stuck but a lot of Connection reset by peer +But Im always willing to give a hand if somebody is going to take this up +quartz plugin needs to be updated to depend on if it is to continue to use mocking to solve this problem +Thats what the v is about +What happens if the client disconnects form one server and moves to another? Or you want to be able to disable that feature as well +If further is required on this JIRA then someone take ownership of s the maven incantation to get the current DN SVN +Applied to trunk +But how I process the last every update +It seems the exception is caused by instances of where instances are expected in m not sure how instances should be handled +Can we reproduce this in real time? It would be great if we can reproduce so that we are clear of the actual problem +I have the same issue with not using a target namespace in the schema element +I suggest to close this issue +Test case committed in revision +I noticed a test case has been added for this checkin +for the patch +hbasck should take a and clean up any leftovers +I havent been able to reproduce the problem yet +Let me know if this is an issue that can be worked on or if it is something I need to take a further look into myself +There may be some other problems but I am feeling more confident about this so probably nothing big. +Please verify and close this Jira or provide steps how to verify the BIRT preset has been removed the issue cant be reproduced. +I like the version Justins presenting quite a lot its in line with the cartoon style of both Hadoop and Hive and has a lot of character +The question for you is whether the interfaces as they are work for your use case +The forminput tag already allows dynamic attributes which means you can add HTML attributes like required placeholder and others +I think maybe we should put up a FAQ somewhere for using Designer because this question has been asked before and it would be useful to have a concise description up somewhere to refer to +mention writing the port to a do we need to update the logging section +I had and files in the bundle +So the message reported in the seems to be erroneous +defer all issues to +Please review the attached patch and judge it again +At this point however I thought that it would only complicate things because users would have to install Nutch classes on Solr in order to use Signature implementations that we use +However I will do this Make an interface Use a default Allow you to plug in your own typeThis should then allow you to add hooks to the replication queue logic +On production systems where you have only the exception stacktrace you have now way of knowing where it occured when the methods are on the same lineregards Malcolm DemetriosHave been looking at the implementation which this class is modelled after so I think your request is validregards Malcolm Edgar +In particular we get burned by in a generic parameter. +i think as well that there is not an urgent need for a generic adapter factory returning value maps for resources that do not provide one for themselvesbut coming back to the initial cause of this ticket i think it still would be a good idea to return a valid value map specific for the resource provider +Code changes look good +Updated the release notes fields +So picking a single pool and setting a threshold seems fragile and so does taking the average or max of all pools +Also the autoLogin method being final prevents from overriding the way the token is passed to the to +some of the joins can always be parameterized by adding where col ? but I think that will only feed the result through a? And changing the parameter will then only affect the? Im not sure but thats how it looksAlso the ps used to create seems like a candidate for adding parameters +The attached code no longer produces an error with the object model +Fixed by +This should only affect the I dont think tomcat supports embedding +Also jruby runtimes count in the jmx keeps going upTomcat crashes after few redeploys. +Resolved leaving scripting language request open as a separate ticket. +Also check the demo installation +In the absence of an alternative making them public is ok but do you have any ideas about an improved API +Hi MarkIt works here under Windows! in rtxs for the patch Florent! +For getting rid of trailing white space most IDEs have this function built inWhat are you using to write this +Users should work around this by ensuring that they ignore all contents of the output directory that start with +Verified +Do you want to submit a patch removing the test +What I am saying is there is nothing we can do about this it is not under our scheduling is handled by the operating collection can also introduce large pauses of ms or even several seconds depending on various is not specific to the JMS you want guaranteed response times you need to look at a real time operating system RHEL with real time kernel and a real time version of Java +That requires everyone to write their own switch statement though +batch transition resolvedfixed bugs to closedfixed +Attached is a patch proposalThe idea of this patch is that during the tree rewritingwe replace expressions with references in two passes we first handle all expression replacements then once all the expressions have been replaced we replace all column referencesThis patch fixes the repro script and makes almost allof the regression tests passHowever two of the test cases in fail these are the two test cases that I listed in my previoustwo commentsIts not clear to me whether these two test cases are valid or notthey are certainly very unusual expressions and I think an argumentcan be made that they should be rejected by the compiler as invalidPlease consider this patch as for discussion intended torecord a possible alternate implementation and to stimulatesome feedback regarding what the correct rules ought to befor correlating elements in the GROUP BY list against elementsin the SELECT list +However you need to flight with Eli for it since he probably will oppose it according to his previous comment +Update ofbiz commited the labels in trunk rev +Sorry about the have uploaded a new hope that you can put it soon on iBiblio as some dev are waiting for for your helpBenoit +Im pretty sure we can just then assign copyright to TOPP though since its outside work we never have the option to relicense it +Please also make sure that the does not reset and lose state if it gets an when talking to theHowever rather than have s store additional information about the final task status of each completed task in ram I think we should reconsider the option of using the as a transaction log for each job +I have sent details of it to Stefan +Works locally +Is that patch ready for review? you mentioned it still had some +This patch adds the reference counting to the so that only one should ever be registeredIt also adds the methods and behaviors that Ant sugested be added tois has been addedand unregisterMapping now returns the Servlet which was unregistered +to CFRR changeswasnt immediately clear to me what CFIF changes are doing can you elaborate +Minor edit of the did the test with maven and I havent anymore the would it be possible to use this release of Maven in mEclipse? +Removed reflection which fixed the classpath issues +However for built in hive it will almost always be better to have vectorized implementation for performance +provides fix for this patch in Rev +Hi sorry that I keep you you please try the attached gem built from Duncanmaks development branch? Ive not yet tested this on other than SunOracles u though +Verified at r +Havent found repository with and there are more than dependencies +Changed the fix version and think the name init is miss leading but I couldnt imagine a better name can someone please help me on figuring out the proper name for this sequence +This patch hopefully contains all the code it should and none of the debugging code it shouldnt +Attached patch file in commit cecbfbbfcddcddccbbThanks +When you run the bat file itll only write it for you +Also exclude accessors from complexity counter +I think its pretty decent but it would be good if someone else my work +When a method is overridden which is identified as innersource the template or not take care of the related imports +is there any patch or workaround for this issue in camel +Patched to check the and skip SDE managed columns patched trunk as of revision do you want to apply this patch to stream +The attachment is a +DawidWhy is there a problem with having some functional application stuff in Collections? not push all duplicates back to Collections +Warnings are useless as they are universally ignored +bulk defer minortrivial issues to +This seemed to make sense but I guess may not be the only way +WotjekSorry I dont understand your comment +it makes sense for me +fixed +and while youre at it adding the option to make the combiner smart enough to protect against any combined url getting too largeif the url is bigger than K of K then it should start a new combiner url +Patch is attached +Why does it only affect the Minimr tests? Specifically which parameter is too long? Can you please provide a reference to a class and method? Thanks +Verified with a tarball that a NN is not startedstopped when is not set and startedstopped when it is +COMPONENT SHOULD SHOULD SHOULD APPEAR +pointerDefinicionsca Smbol que apareix en una pantalla dordinador generalment en forma de fletxa que indica els moviments del ratol o de qualsevol altre dispositiu de dileg. +We could only really factor out parts of those conditional checks into separate private methods +This is similar to except that the key is to provide access to JMX resources instead of registering them +The fourth example of the schedule component demonstrates this by assigning a different color to each entry +bq +Part of a bulk update all resolutions changed to done please review history to original resolution type +Sajith Have we removed this already? If so may be we could close the issue +Im not seeing this problem with the latest code thanks. +ant clean then ant build fixed it up for me + for jukkas proposal with explicit workspace name param and clear separation between credential attributes and the new attributes map +Dear AndreasI hope this will be OK now +Thanks for spotting thisJuergen +Setting even a bogus agent name is an insignificant effort compared to the further complication of the code and configuration optionsI dont see how it complicates code if checking for data needed for http is done in a place that only affects http +Iainabout WS it is strange +The latter are minor but you can see there is drift +Thanks Vinod and chakaravarthy! +The will be uploaded by Fernando Nasser presumably to +Thanks Tom +commit a patch before having your commentsquestions sorted outIts easier to work on incrementally improving this when the code is already in svn +included source and class files in to make sure this is genericised and fits with the public +We appreciate all you patient work with us and the Camel. +Any more comments on this? Id like to resolve this by extending the solution of to cover also standard that can be freely redistributed in unmodified form +Please give latest nightly build a try and let us know if it works for you +I have tested the nightly build and it is working now +I got around the problem by taking a common component and adding an Ajax action that keeps the session alive +Verified. +Still need to finish Japanese messages properties file +Verified by Leo. +Im not sure why weve not noticed this before +to be served unless theyre in +Id appreciate if you could lead me to the right direction although I fully understand that itll be much faster if you fix it yourself +So in my opinion an stack should contain SCR +Solutions Showing date times in the server time zone is already implemented +An XML File is not a filesystem by itself like an Tar file but itsstructure is the same as for a normal filesystem +If you have any assistance you would like with this please let me know +Also if I remove the jars then the compilation will break +bq +We dont have support for beanwrite and similar tags so will have to look into it after +The default Exec for is the JVM you have configured in eclipse to is though a valid point in why the combo is referring to and not something more recentcorrect +Thanks +Just making sure that everyone is aware that the andanything else from cannot be committed to the Castor CVS due to licenseincompatibility +This is fixed in +In patch v removing lock race in and skipping region who is flushing when flush +Now that work is well underway with AS all previous community releases are +This causes in the Sun VM all declared methods to be accessed which in turn causes the Resource interface to be requiredBy refactoring the class in Rev +Ive been thinking about this +Patch applied +This is similar to what I did with the spring configuration +When there is not a descriptor entry for the EJB the ws endpoint registration code throws a NPE exception +Hi BenjaminTrunk is already on +It is trivial to swap out for couchjsUsers can try V already +If so make it provided if the database is google app to make the jstl dependency provided if the database is google app engine +I can confirm that it seems to work now with Firefox The Opera issue still persists but hopefully this will not affect too many users +However the issue that bothers us is the general case +It could be connected with bug so I am going to start on that first +The merge failed again in CI +fix some extraction error +Simplest app to show the problem +Um is failing to select a provider +I wasnt recommending violating the spec and there is much to be said for knowing the spec and thereby being aware that one cannot create method in such a manner +Youre right +So the current code that copies the response message properties into the response context should not create a recursive reference +I just committed the Java implementation of this. +pavels fix +Assigning to Luc so he can confirm this is a duplicate +When will the schema be updated on the web so the expression attribute will be resolved in STS +Fixed and Ive tested it. +It will just make that merge harder +Im glad you liked it and approved it. +This new version would say indicate wal was now able to do extended compression api AND includes flush and compaction markers +Actually no one else has seemed to need this patch so probably best just to close the issue +I can take a look +Passes unit tests +Applied by Gavin +Modified with all lines containing go removed +Before I read s proposals thoroughly I had these thoughtsI would support removing the optimize button from the GUI or at least removing it from the Overview page +I have therefore resolved this as Wont Fix does that seems the approrpiate way forward +Theres a comment about num being a multiple of num +Of course one can use reflection but thats just atrocious +defer issues to +That may explain why the last time I tried to verify thisit didnt seem to create a class file +Local unit tests failed this is my first patch and I may have missed something any comments feedback are for the feedback +Committed revision . +And use a final release not some arbitrary release candidate. +For your issue thought you are missing it should come before the processor factory +The API for was changed back to its original form in order to maintain Backward patch +studioeclipseplugins +I used to think total nodes including active nodes and standby nodes +Committed the changes to subversion as revision +This requried some code changes and Ill look into it +This incorporates and removes contrib from build and test targets +From a spot check it looks like all these changes were included in the patchfile for so Im going to mark this as FIXED as well +Ive resolved the latest batch of solaris issues. +There is no other output generated as the service itself isnt used at all +Thanks Yan +If the calls are prevented from executing in parallel then I can not reproduce the problem any more +The was introduced in bash +If we can assume that this case would be rare it wouldnt be a big problem as the plan would be and placed back into the strong cache usurping the existing entry +was jira with test cases only no code because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Still i think we shouldnt close this jira yet +Removes links in and editors to the custom property editor +was my fault for not testing it soon enough +This causes Derby to raise an exception if you try to implicitly cast to or from BOOLEAN in a way which violates the SQL Standard +Could you please check if works for you +Needs emailed socreccb about the jboss user +Fix is committed in r comment thanks Thomas +Since were making a commitment to output I simplified the test regex to check only for letters +verified in JBDS +I can put up another patch if you wantIf mvn package runs on a clean checkout I think you should check it in +HiThe same problem occurs on with Firefox and Safari tooBest Regards looks like this behaviour pops up in too +please let me know if this can be integerated in axis or what i need to do to get it into the codebase +And what about branch? +Verified by Vladimir. +A system that combines the configuration two is intriguing but would represent a pretty big shift in API +As a next step I will introduce Derby into the game +Can you explain why this module and how it relates to? It sounds a bit strange to have ways to provide push support in Wicket +I will take a look when I get a chance +I wish HDFS would handle this stuff for usLGTM +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Thanks for new patch AndrewIve been playing around with the migration +We had no yet so Ive added this to +So this issue seems to be limited to the hdfs side +The problem is even deeperThe problem with the many instances of is easily solved by making the counter variable staticBut there is still a problem when is repeated in Wicket repeater +But in your patch its reversed to find the alias name for real field +fixed in the trunk +pushing through to verify that it doesnt break anything else +A developers dream come true +Tanju can you help me out by providing simple test case? I want to fully understand the problem before we apply a patch +ing issue as part of JIRA cleanup +But lets answer to your question first +Thanks Matt applied + + +Attaching new patch for trunk with review comment incorporated +Its the same problem as but in different tests +toolsconsole +Patch added +Were limited by the Hibernate API though +Patch for aggressive GC in svn rev +The setup is what I expected +attach a patch addressing long length lines and trailings +I see but that is only if we are proactively populating the cache +Hi here is the requested test case patch +Should work now +With Gurkans recent changes my second app still fails +passed on my local machine +Hi NathanAny comments on patch +We have seen stores that needed to compact files and this is why we have a limitSo I question the feasibility of this jiraIn the particular case of waiting on flushes waiting on too many store files what I said is that its by setting a very low number of that we easily hit the limit +Only affected abstract base types with a single String id field +I just cant imagine a case where a developer will make changes attempt to package and NOT follow it up with just doing the sortingBut I guess it is the better way to go +If the patch is that much slower on your system it is likely that it will be slow on others +Patch has been applied to current ChristineCVS should be for us because we migrated to Subversion +Stack Jons comments makes sense +Throws if proxy is not closeable +Can someone close this issue as duplicate ofThanks by +I took a look at the patch +Further if you want to compact the META table the instance should not even be +I created two new issues for the default values and dates and +I have applied the patch but we should really see about getting you version control access +For the file name of the tck distribution we should use the standard naming convention for JCP distributions +I have committed this. +Saw an issue with fragmentation today in the web UI I refreshed it around the same time a region was deleted and was thrown +What do you mean with similar error message? Please post a stacktrace +The original issue is actually related to the server side email address checking +Committed revision . +yes because a little title on the images would be helpful and the layout is ok too. +This is specified in section +I was advised to use the package prefix by the JAXB group when hosting the plugin at +Im pleasured we can understand each other at lastSure I try to test when readyThank you +I took a shot at a patch for this myself applicable to the trunk +Because if they set the value to enabled or disabled they need a way to get back to the login option +Right +See and patch +Jerry original closed this as fixed for +I am assigning this to myself since David has not responded for a couple days +This is an unrelated problem and should not affect the behavior of the plugins +Closing +An alternative is that the test could be skipped if running in a separate processmachineWould need to see how much extra value having a particular test running with a remote server would be if there were many other tests running remote +So you mean that in UI jsps or in Jetty wrapper code we poll ZK who is master and if not current machine send request to other master? Could be a? Anyone working on this? Should I have a looksee +Ideally I would like to see also default to true but I am fine doing it later +Fixed filename error +Hi AllI am running +Ive downloaded the release today and tested the download of the javadoc via has changed in the behaviour +We wont have a proper fix for to Adrian if he can sort this out +The new files for unit closed and comitted to CVS +stephenI have run the patch on my notebook +no exceptions with that +Ive whipped up a small patch that alleviates the problem for my specific use case +But I will certainly refrain from editing the ticket itselfAre there any other suggestions for making the proposal more viable +One Small Step For Me One Giant Leap For The Grails Community I will upload them soon +Also refactors how key cache is instantiated and removes from test suite +I should have picked up the problem with the Release Notes TextI will update the Release Notes entry for this issue and again resolve this issue when thats done +Whats the situation here? Is someone working on it +If I wanted to work with an or that required a file generically I cant +the rest jar files from +So Stack Maybe you can have a safety look +component required readonly private dependency +Add your notes and wish list here Galder! I forgot the details by now +point being characters has been a very common limit for a long time either strictly enforced or silently chopping off whatever was left overI am fine with moving the upper limit to characters and setting the default to that +It seems the might not be the right one for the job. +Ill share my ideas you can use them if you want +Linbinis this additional patch or replacement for old one +Normally this would be but this jira is handy if we need to backport it into +If either one of you can give me a French translation I can patch the English text for this key isNo changes in this release +just committed the latest patch to trunk +patch just committed this. +I tried to apply the Quick Fix to several of the Examples but once I change the targeted runtime other problems appear missing classes +resetting P open bugs to P pending further review +needs to be serviced before Doxygen generation of API Reference Manual can be performed. +small correction needed procedure description is not clearcurrent decriptionThe system procedure disables log archive mode for the databaseshould be changed toThis system procedure disables the log archive mode and deletes any existingonline archived log files if the input parameter DELETEARCHIVEDLOGFILES is +with all the required imports? I know its a minor point I just like to keep Karaf mostly free from import alls +Subsequently the contents of these tarballs are not referenced at all we expect users to supply their own installation of Hadoop +Execute derbyall on jdbrc environment and found no error except for intermittent error of derbynet in framework +the threads that would have called I dont think were very rigorous about calling so I think this will actually deadlock it will wait for daemon threads to die and the daemon threads will park at the lock acquisition +The code will no longer associate the analysis record with cached results +Perfect now thanks. +version of the patch removed factories and switches +bq +In the import causes a compilation failure +Apologies attaching a second patch with minor changes make visibility of methods in new private static inner classes consistent use rowcolumn parameter names rather than rowNocolumnNo as POI does +Patch attached fixes the bugThanks RegardsAtul Atul for the contribution +Fixed on closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +I have not been able to reproduce this behavior with Oracle or postgres using same or similar settings to above +Will be available in next snapshot build. +However what definitely looks wrong is that a response causes unnecessary termination of otherwise perfectly connection +Here is a patch that adds the as an argument to the constructor +ChrisReally Im trying to lure you into the discussion +this tarball creates a wicket test case in srctestjavaorgapachewicketwicketuntar it in wicket +OK got a second to take a look at this +Lowering priority since Eclipse support is not in the current scope +I created for the caching issue +A subtask per module is probably going overboardAny other volunteers interested in helping with the move to m +batch transition resolvedfixed bugs to closedfixed +Transitioning to closed. +After each job is executed I decide whether or not to delete update the job entry +Would it make sense to allow folks to pass in a localized string for the Done button? That way we wouldnt need to get that string translated for all languages +Pushing tried this on a cluster +Assiging back to you in the case you agree that more meaningful error should be reported in such case +I see +Latest hudson run confirms the failure is because of +So the easiest conclusion was to define multiple groups and to do the conditional thing in the view +if its not a break why do we have the method there throwing +Youll need to run mvn package to generate a jar file that you can throw into your classpath +patch that moves the clone also only does it if we are actually going to distribute the committed thanks Jim +finalized and committed to trunk +Updated the title and description of the bug +Simpler fix +Thanks for your tip I tried in this way but not enough hard I guess +I wonder why this build fragment did not work for contrib? The only problem is that this also copies the package +Ill add that I just upgraded from where we have been running a suite of tests fine for a month or so +Here are the corrected +Thats what oddIf there is no assert how on earth am I supposed to reproduce a problem that manifests itself by throwing? Please provide a reasonable test case or I will have no other choice but close this issue as invalid +bulk close of all resolved issues. +Heres what I was thinking for C api +Issue fixed +Heartbeats have been removed from the bus +As per the above tasks which revolve mainly around reports it may seem an overkill +Thanks Kim +However for me in Windows the patch supplied appears to fix this issue for both sliders in all directions so I am happy with the patch from a Windows perspective. +Patch to include another testset to the CORBA testsuite along with some minor modifications to the other testset in the same suite +Patch added +Ill create a new issue for it +From the message Caused by Connection refused connectEither your ftp server is not running or your configuration is wrong +Confirmed here as well +ClausI will add new parameter in the annotation to define the character to be used for the carriage returnLine feed and extend the class +But the bug in fact is not Xerces asit only appears in xslimport so it is rahter XalanI see similar bug which is about dtd validation in Xalan +I dont know I didnt use the EJB plugin since this issue +derbynetmats had no failures +Guess what +I was reluctant to contribute it earlier since the tests require an AWS account and I wasnt sure how this would work during regular builds +We should fix this before to unblock downstream projects +Craig picked this one up in. +I got it runningPlease review the changes in so we can submit this as quickly as possibleBTW I changed the use of a thread per task to a polling solution mainly because it should take too much memory for larger number of tasksAnd I excluded the testcase because the yarn testcluster seems to be not available in the releasesnapshot of oo +Invalid +For view controllers its also common to use. +While I was at it I also added content lenght and last modified date +Certainly done +It happened only once on AS bootI agree with Davit its the aforementioned bug +Ive just noticed that youve done the same before +I wonder how much effort it would be to add a new test to the CTF master test suite so that we catch any harmful changes to this code area in the future? Just in gerneral but as you do not seem to be breaking any existing as you know I am absoltely behind you that we should add tests to our test suites for any new feature we implement +Unless they arent marked as exclusive and if you install both it picks up the Sun JDK +So need to implement v instead +I posted output from my screen and included my development usernamepassword for my database +it gives us ints but it complicates things a bit +rows appear in wrong order +OR whether it is a Tika doc or HTML in which case it prints the title dynamic teaser document size document typeMIME etcFinally we could add some to the exampledocs folderDo you want to attempt a first patch +no problem thanks! +Hello Erwan Changes you made in your patch are not appropriate +I note here that the fix has been applied in the branch and will be included in the release. +Sample application to reproduce the issue is fixed in trunk +approved +Adapted the HTTP Base authenticator in Rev +But as I said when I close my clearcase connection everything is fine +I dont see a forum topic with that name +moving to since its better to have a bigger community to collect different point of view added method without mapper +All regression tests ran cleanlyCommitted revision . +As you know MR and based implementations are removed +Have you added to your? Changes made in CR seem to require all roles that you want propagated to your portlet be present in a +is about adding attributes to classpathentries in +Also please can you file another JIRA to convert the existing mirroring system test to use this new framework +Verified by I meant verified by Alexei. +Oh yeah Ill update it +Changing module the only thing I do not get is how Im supposed to know whether an issue I have when using SEAM from Tools under JBDS is a SEAM issue or a tooling issue +Patch to make the visitors lenient towards the number type to be for sure what patches for you are referring to +Then I tried adding a print statement after the last what is happening is that sleep is waking up spuriously permitting the main thread to exit indirectly killing the other threads before they have a chance to finish counting down +Yongqiang have you started this JIRA +Jason this would be on the host where the agent runs +If your testing differs there is something going on +Thanks in advance +youre addressing a pain point that people just arent feelingSo Id be more inclined to say lets do it the intuitive way +This is needed for Pig unit test to run overI ran MR test target +So I suggest to mark this test as resolved +As the new system template for xenserver does not have XS tools we cannot support system vm scaling on Xenserver +Manfred could you please send the you created for Seam +Did this firstgrep Index +Should we close this? seems to be a more general approach. +Explanation above +Imagine that I had already done for them +Applied thanks +This is only preliminary draft work +Visible text is probably not the best criterion +Ive released metadata which fixes these tests and updated the component matrix to use it +I searched the testsuite and there was one usage of the Java EE one and no JEE one +Issue fixed included in the iPOJO release. +Attached patch PleasePlease can someone take a look at patch? +Jay anything we can fix without doing the hack? Does all the templates have to use resource to lookup instead or + +Excellent if youre able to make a we might even be able to avoid this mess in the future +Someone can help to review and commit it? Thanks +Test case for the bug fixed +Resolved as part of +Good point Sid +We can always do minor tweaking afterwardsOleg +Any update +This project is a web project and is the only project of a group +Issue has been resolved for a long time waiting for potential confirmation from the reporter +Having the style rules be part of the query is fine but not all Datastores know how to deal with Function expressions yet +please reopen when final is out +Hi Hiram unfortunately I reproduced the issue with this build using the army swiss example and sending message at the time +Youre right I had forgotten to update the JIRA issue sorry. +So it actually appears we are inconsistent about acls +as per Johns s open up discussion for a better extension mechanism in. +Yes its probably useful to have output json format. +Currently task cant be completed because there is no possibility to build richFaces and get actual attributes table +Committed revision Jeff Levitts patch that documents errors messages +Can you assign this to someone to validate what should be m a little bit hesitant to just update the sha hash of a file since it points to a broader issue +Sorry the damage once done will remain in the indexThis new assert will only trip if the index is recreated ie when a segment is first written with the wrong countCan you run on your index and report back? Im curious how many segments show the wrong delete count and how much the counts are off +It does not work doing for the basic case c list +Additionally the change in to set the s blocks as null may be necessary in this patch as well +Thanks folks this sounds promising +Is it in megabytes or in bytes +Also from the Camel API you can use that id +Removing the Jenkins component as this seems to be a Sonar looks the build need some external tools +Build fails due to dependency on which is not in trunk yetThe patch affects large number of tests setting timeout for all of them shouldnt be done as part of this patch +Implemented for Jackson +Not sure about the malformed issue +Looking at the patch its not clear to me whether the reloaded core would return proper Commits right after its been reloaded or only after the subsequent commit +Ill try to take a look at this later today +Attaching a new patch with following changesMade modification to to use the new API for the web UI and command line interfaceAdded test case which makes use of the new methods introduced in the the class because of a being thrown for the default scheduler pointed out by HemanthFixed a findbugs warnings in regarding incorrect synchronization +Patch to address test issue on slow machines +Will that be okay to do it separately soon +Of course I stole the idea from theexisting filter in MINAIf everything else looks okay and if you think the packageclassnames are good as they are I will go ahead and try to wrap this upthis week +Agreed +I actually read which indeed needs a small fix for +The compiler errors were due to the fact that a REST procedure had no return value +John for youLen please report issues directly in JBIDE things moves faster than when issues reported against in Drools HEAD +Please add your newly contributed tests to newtests folder +After backing out this patch doesnt fail any more on my +Are there any objections if I separate them again? I would start with and +Im prepared to do this work btw +Setting fix version +As you can see now all the elements are in russian +I can see this mentioned in the release notes but cant find the jar anywhere is it going to be available somewhere +Im not sure if we should changeenhance the resource resolver api what about adding a method to? +The problem is that if the connection gets hung then it wont time out since the flag is set +PDF rendered with current trunk plus some changes by me and by others that havent been gets weirder I get the same result as Luis at home the previous result was at work +Please have a look at +Docs for tried to follow the instructions in the zip but couldnt get it working +This patch includes fixes +the resolution is in the hands of the dependency plugin so if this were to be supported it would need to locate the reactor projects as well +we do not need to prevent files to get to code generation when that could result in Java errors +gone hunting +Works for me +Heres the new chaining since its clear that this is a different feature Id rather open up a new JIRA issue rather than roll it into this one +This fix to the URI implementation is in Xerces. +I consider it important for old urls to still work after the move +Thanks Simon my original test is working now as well. +d suggest we put an according test in the codebase with all combinations of field and accessor visibility +Hey Joydeep Ill try to implement it however we will make our temporarily on Thrift Server mode not CLIThat a great feature you mentioned +Is there a way to fix this problem by removing the invalid versionsThanks. +Thoughts +It also solves in another way i left that code in there though since the same amount of work needs to be done +Still cannot reproduce so hopefully its all good +x patch +The only reasons I used zip utils is because in lucene test they used it and I tried to stay as close to them as possible +As a result we will suppress all cast warnings in +This should be made conditional by a Java property or env var +Note that it requires changes in the javac code as the classpath changes when calling javac +I initially designed it to be just a normal class with recursive functions +Thanks +the disadvantages of having everything in a single index are you cant use the same field name for different things they must be the same type efficiency and sparse fields sorting on a field takes some memory for every document in the index regardless of how many documents have that fieldPerhaps you could give a small example of how something could work in indexes but not +Not sure what you had this problem with my project when it was on version +It isnt strictly necessary to open an issue there however +The simple addition of Bsymbolic to linker flags of hyzlib solves the problemI would suggest to apply this patch until the problem is solved in a proper way of removing duplication hyzlib system +Ming maYes what you say is correct +The fix did make it into the source code however I can reproduce the issue in +The only change relative to is an extra test for SET ROLE? syntax +Moving issue to a next issues to next release +Sets me up to start putting up monster YARN patches +There were similar issues fixed in release please the tests with and report back +BTW comments on issues that are already resolved are most likely beeing resolved issues for already released versions +Change the name to just Journal +Currently I am uploading the new +The scripts can be loaded from the bottom of the page without changing the delete logic just by looping over the widgetRegions twice +We need to make other parts of the build ignore such files +Tom Ive remade the patch to work against trunk +Adds to the chosen comparison +No new tests since successful compilation is all this attempts to fix +I understand that point +It might be better to avoid assuming that the storage url is actually hdfs +Close out issues. +Also its probably a good idea to return or null if value parameter is null +I added a menu item in the navigation that will hold the link to editing the users own account +Might be related to DFT changes for +svn statM javaengineorgapachederbyimplstoreaccesssortI ran derbyall on windowsjdk ok with this patchCan someone please review +I still have problems loading pages in my project which uses the angular jqm adapter but thats probably not a cordova problem ill investigateDifferent question more with the JQM WP theme is there a reason page transitions and pushstate is disabled? Can I enable them? Transitions do not seem to work. +fix is committed under by mistake +Unless of course you would really like to use this product in production and not just as a development curiosity +Update that removes the need for server files +No youre right! I did count it to be when preparing the script and before doing the commit but apparently I screwed something upLet me prepare the diff list and move them in properly +Open sources was great isnt ithave a good day this fixed in RC? this issue need to have an Affects Versions assigned to it or some other action taken in order to see any progress? It is surprising to us that incorrect SSL handling has not received immediate attention +pending +i think im changing my mind on whether or not we as a web framework should be attempting to fix thisa wicket application is not always composed of just the wicket filter a great number of them use other filters and servlets thus we cant truly say that your wicket application is safe from this kind of attack because other filters and servlets in the application are not safei think we should drop this and let container folks fix it which is the only place where something like this can be fixedalternatively someone can release a servlet filter that can sit infront of everything else and look for this kind of thing +Two is fine for a default for now although I could see setting this higher to make apps more robust to node failures +Tested under Windows using latest trunk and it works Jay can you please verify +bulk close of resolved issues for release. +I see RC being out +integrates Nings offline comments +However once enabled any resources should always be removed once its been validated and +Attached patch makes the required changes in to use if the load location ends with extension +Aside from the job manager the others work fine with the current approach because you control their instantiation and can supply the delegator. +See SOA issue for committed in revision . +Same as the v patch just in contrib instead of core +Closing this ticket as JDO has been released +So the new PR fixes the hibernate tools issue +Do we need this in trunk too Andy +Im confused by what Scott means by EAP +The reason this dynamic loading was done may have been because of a temporary implementation of the transport earlier on in development +For that we have no chance to replay the logic to start a new attempt App itself should be able to start a new attempt +Just a heads up that the svn tree is RO while I migrate +Definitely still an issue +So that its better to closed this issue and if find some memory leaks can reopen the issue. +The issue is that when upgradestack is run it does not update the current stack version value associated with the cluster +I picked in analogy to the existing and because the are loaded per region +we get a summary of what those issues are rather than pouring through all of the comments +As correctly noted by Mike Mller this issue is actually superceded by the changes for The new logout method is of course present in the new Authenticator interface The interface is not needed because the new interface already includes the added method +Thanks Hiram +We will evaluate traffic over the coming month or two and close the when appropriateThe list has been closed already as part of the Jakarta dev lists merge couple of months ago +First you should compile projects from directory subprojectsIt would be better if you can ask in dev mailing list before creation of such issues +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Its the more heated non technical issues that would likely boil up regardless of CTR or RTC that cause issue +Regardless without a test file to compare against Im not sure how to test to see if these have been fixed +The should probably detect by type rather than by media type +The plugin is complaining about missing is only a warning +Attribute is present tested on NB +Could you please provide a test case that makes use of to reproduce the problem and attach it to this issue as a complete file? there isnt even an assert +Marking closed +Done +Also there is no longer a cursor in the shell view +Committed +The forked lifecycle must be dealt with before we can do this accurately +without stack traces we wont know the cause +To work around this issue use Seam for integration on Enterprise Web ServerPlease ping me if this requires corrections +for option I dont see a reason why a null termination is needed here +Ah +The JIRA should clearly target and above +Rerunning tests again right now because when I updated my tree I got a conflict inPlease review +Tests are failing on my run +Max wanted it assigned to Alexey +Roland could you resubmit a readable patch +Would you like to me to change the code in some way +But doesnt that mean that all rows will go to the same reducer which might be bad +for the patch +Docs should be updated before release +Heres my attempt at a fixI know that there are two variables called queue in the new class but it seemed prettier than casting +Thanks issue is closed now +There was a corner case in the previous patch where in if the history is accessed after the history file is created and before the job status is logged then the cached result will never be refreshed +Thanks +It returned the length of table path before but now its returning +Thanks Amareshwari! +Postponed due to the absence of an updated patch +That is the only difference I made to the config and now its been almost weeks strong +Obviously we do now support Greek characters and as such the test should passThe second regression was with something I overlooked +Close all resolved issues for older releases. +Cant reproduce closing. +Shai are you absolutely dead set against the name forceMergeNo I am not dead set against it +IIRC the only ones that seemed to have any kind of support were +I added a note to the wiki page +The documentation should be changed to agree with the preference is for since that would make my tests a little easier + makes this issue obsolete +Manually tested! +Unassigning the issue since I am not working on it +for applied in trunk and Beta branch +You have to be careful in using though +DaveNesting can sometime unintendly exist when collapse is left open in a previous bracketAnyway I will fix this nesting issue in a next commitI tried testing withwihtout spaces after the bullet and it didnt make a difference +Hi JulienI am sorry for the delay +Committed revision thanks Javier +Appreciate if this can be committed quickly +Patch integrated to trunk and branch +I have left the announcement untouched as I am unsure what needs to be update +Committed a fix for it +Customers should be discouraged form opening that kind of security hole +I think we cant do it with junit or at least I dont know how +are those lines in the debug output consecutive or snipped? Very weirdTheres some special rules in Maven about but Maven should be giving you the version you asked for +oops fixed test case to use constant throughout +A new seeder only for a line annotationDate and time are yet more complex to define and rather a use case for rules that are used as a libraryId rather pursue different approaches adding those annotation to the casxmi before rules are applied +If the pool had threads that would be more than enough to prevent slow blocks from completely blocking the +Works still in progress Handle special cases in +HiI can not access the FTP mirrorCan you change your frequency to be some figure above hrs thanks +Changed the component to Build to reflect thisCheersMike C +Is this an oversight or intention +Moving to in preparation for release +We need a snapshot of the wsdl in SVN so we can run wsconsume at test build time +It is ASL so wed have no issue including the source and we dont want two different artifacts in the first place +Thanks Cos for the review +dont know +If this is fixed then what version of the xerces parser do I need +Will it hold up the RPC handler during the call? what happens when data can not be written or read without blocking +I suppose we could have a separate spec for RPCI filed for thatIs there a problem using HTTP? Rubys Mongrel should be easy to use for servers +orgapachehadoophbaseutiltest +Here is what I have so far +Thanks +Check for null pointer exception when reading for columns and there is no result in the selected range +Patch looks fine to meIve added a static constructor on to create from a as well as whichll let that bit be tidied up thanks NickIll add a TODO where we use to cutover to once we upgrade POI +The indexterm entries are exactly the same for all releases for page at least + +Test source code and data files +This is already fixed for trunk and here I am attaching patch for branch +Simple patch for the problem +Patch for trunk +Now Eclipse debugger shows more then threads in a thread group so the problem looks fixed +I have been thinking about this issue and wonderIf we put the name of the database in the name of the thread and then check for the existence of that thread before booting the database and fail if it exists would that be a reasonable solution +Fixed in r +Editorial and committedThanksChatra +That is goodJust we are closing down onBut if your patch is ready before we start cutting the RC then we can get it included +The interesting thing to note is the empty in it failed reliably +Sounds good +JonathanCan you share your patch +Thanks Mikhail. +BVT runs are failing because of this issue +whoops mangled the what version is this forkeeping it assigned to Yuro +and if you think about it you realize that not only are you constructing the model in your view component constructor your inevitably doing logic in there as the application gets more complex which would definitely be an in my bookSure I can emulate it or design a new class that helps to avoid that sort of thing but why should I have toAnyway something to think about +Does that make senseIdeally we should probably configure Ivy so that it looks in the lib directory for dependencies so that offline builds work when a release is first unpacked +Right that was just me not getting you idea at first +Jonathan feel free to upgrade as you have time +The log in info for them isuser is in the group but is not +Good point +Its about compiler deprecation warnings +As all core development tasks are complete and resolved I am marking this jira as resolved. +Quite a few javac warnings are getting in without being detected +Since there is some user expectation that +Yeah as an alternative Im considering just implementing the XML serialization directly without having to deal with the complexities of the JAXP libraries +to the suggested approach +Per my understanding some legacy systems already have dedicated session cache cluster for session resumption when they upgrade to the latest openssl session ticket is supported by default sessticketenabled option provides a flexibility for them to disable session ticket and continue using their session cache service +The code is a direct port of the C version +Can you try making your security config conform to the schema and see if that helps and if not attach the modified config? The schema should be in schemasIm working on improving the validation. +Appears to me this is more than a minor issue for m users but if you have the work around then you can proceed +The attachement contains the full source code of our custom sdo implementation including our data mapping extensionThe included maven pom file might need some tweaking to get it to build +Minor comment +as long as covers all cases now describes I guess +Review posted in reviewboard +Sorry meant to say Im not sure how feasible it is to remove the dependency +jar already +If its not lets go ahead and close out. +Thanks Sean +Heres a patch that implements this complete with a testI intend to wait until at least has proven this to be a useful approach before committing it +Roo already supports case which is excellent +This patch addresses the backwards incompatibility +What makes a difference is the position and that we cannot easily express it with what we have +I just committed this thanks Jiaqi. + utility has been modified to take the clusters and points directory as input instead of sequence file and points file +Rolling back is impossible so it makes sense that engine boot fails because this error is unrecoverable for activiti +we have done perf testsWe see no indexing throughput improvement query throughput improved by +Thanks to Pawel Koloszko for this fixregards revision to the change applied has been added to +Looking forward to seeing the results +This seems like it might be expected behavior and if you want to disallow nulls you might need to add notNull +There should not be an NPE here +The variable name is always underscore underscore uid underscore underscore +This will pass when will be changed the to use insteadof the tmp solutionIf this tested will be committed I can increase the test coverage +Fine I was worried by using spring components from different versions +Up to date ESB kits are available now. +Thoughts? +Building a parent site jar absolute links from the parent should be made relative +Downgrading severity as workaround exists +Ill look into it +BUMP! This plugin is useless at this point +Throw exception instead of silent return on unrecognised type +This patch also relies on using Mockito for unit tests so theres a modified that adds this as a dependencyIm hoping its OK to add Mockito to the test scope +I was supposed to make a new patch but forgot about this +plugins dependenvy to +Thanks +This patch does the improvement +For consistency and to avoid checks for nullness IMHO it is easier to prevent fields to be null altogetherIn the test case both null and empty byte arrays are used to initialize identifier and password. +Please let me know if there is any way I can return the favor +Heres the patch +I think we need the resizing to deal with overflowing headings +Also added a new billing account subscreen in which orders with of type EXTBILLACT with status PAYMENTNOTRECEIVED is shown +Hi PaddyThanks for reporting + failed can you take a look +Please note that in order to avoid the NPE on conversion you can use the twoargument version of the API which allow one to supply a default version for nullinput entries +Fixed in Jopr SVN branch r +Fixing release note typo. + +This will fix the parser to handle right parenthesis correctly but reintroduces the issue reported in +Id be happy to look into this closer if anyone can provide som input on how to reproduce the issue reported +Let me retry this one more time on pseudo cluster +ok we should be under the max column name now +nand depending on which part of the framework looked at the particular value they would take the value from conf +updated patch with refing in its so that javadocs workthis is ready for trunk. +I can think of one more way of solving this +Waiting for metamer release and then tests will be retested +fixed in w the release. +Fixed on revision . +metadata +Please try my patch +I havent had the chance to test it yet. +Once the new patch os submitted I will review and commit it +I know how to fix it but I need to clarify some questions before creating PR +This bug has been marked as a duplicate of +So an alternative would be to bind our filters to some sort of scope so that its not a singleton but right now Guice is attempting to inject the into our filters before has actually registered the hence the warning +On first glance this may be as it has a patch attached Ill focus my efforts on that ticket +The code particularly the ecommerce cartorder UI is very ugly and full of my debug comments and working notes +That looks like but may be occurring after the original protocol error during the subsequent cleanup +I saw some more problems when using alternative names or of attributes and object classes +In run with nodes I started to get loads of view changes and NAKACK errors +Since you are trying to dynamically change the Table annotation at runtime doing the byte code enhancement at build time would make no sense +Sorry I forgot to do svn add to add the new file before generating the last patch +Patch is applied at +s seems to be reasonable though +Is this a new behavior + +our only simple +I guess this part is primitive and was written by referencing other engines for other languages +Defer all open issues to +You should set expandedtrue values for all expanded but then why does it work on first subgroupDoesnt seem a very consistent behavior +If I dont hear back on this one by EOD Thu Jun this will have to be deferred to GA as theres no runway left for to Brian to monitor as Ill be on PTO Friday +A patch for solr with improved clarity +JCA wraps it in an XA interface and tags it as a last transaction manager handles this using the Last Resource Gambit +This will require making changes to which currently doesnt bring up the so initialization will fail because itll try to make the HTTP REST call when the service isnt upThis might be tricky because the test enables security and we dont have a good way of running unit tests against a secure yet without extra work to get an external KDC up or use the apacheDS from +No response assuming ok. +I will work side by side with them for the following months +I cloned since the issue with the transform code is still what was required for this again? +Ill ask Stuart to take a look at that PR +And continues to work after the a fix has been applied for Final such that the original exception will be logged and you wont get this NPE +This jira is about implementing the correct semantics according to the SQL standard with respect to grantable privileges +work tooone question though +Committed to with revision . +Good catch +Let me look into creating a test I agree this makes sense +Do you have a test case that tries to parse resultsets with columns and rows both using columns names and column indices? Can you post is some where so that we can verify itKaj +This is the version I use to compile and the version passing the TCK for Hibernate and ASyou should not have the dependency pulled if you use the pom I pushed you look at the history of and they have not always been aligned +For those needing shindig I avoided this issue by using guice instead of. +Just the subordinates or descendant count is not enough IMHO +A whitelist is a configuration as I wrote if we need a config a blacklist is imho more appropriate and far easier to handle +It constructs a version using this for example buildNumber + Fix attached +You should open a bug +Set version to for all issues since more true than no fix versionAdjust the fix version to match actual aerogear plan +From the log it does not seem to be realted to the VM state for this latest error +This means that the adminstration of the cluster becomes slightly more complex +Needs these to be installed previously +I just committed this +The will have locally attached volumes while the has only IIRC +Not completed yet +build SDKROOT VALIDARCHS ixcodebuild error The project does not contain a target named Whereas when I use a location CONFIGURATIONBUILDDIR without spaces things work finealdo aldo ioscordovadebug Build settings from command line ARCHS i CONFIGURATIONBUILDDIR UsersaldoDocumentsxcodeiosbuild SDKROOT VALIDARCHS i CLEAN NATIVE TARGET OF PROJECT WITH CONFIGURATION Debug Check dependencies +It looks like its worth the merge when we have decided on where to go with ftp +You access queryCaches outside of a synchronized block +has been marked as a duplicate of this bug +Note that Im talking here about operation specific phases and not about global phases +So I have a the position stored in pos and update it accordingly +Ram are you looking at the right patch? The part you quote was removed with thisOr are you saying it should be possible to do thatI think that would not correct as I want an atomic operation and I realized in that I need to write a single +what is this about? Improvements for the series corrections +Wink now uses SLFJ instead of JCL for logging facadeNOTICE Wink is shipped with bridge +for +My aplogies for wasting time +Attaching a new version of +The class now contains a set of utility methods that map between and typed Iterables +Output of info command +Once you got local copies of required artifacts building inside Eclipse will probably work faster +In any case your dirty fix should be fine for now Ill try to fix the real problem asap +I think we should add support for delimitermaybe we should add support for multiple values in this issue alsodont forget to add unit tests andor system tests before closing this issue +Alas its in the hands of who is potentially away from all the secret squirrel stuff that is required for getting SSL certs on the other hand am at my desk but have none of the legal instrumentation to do anything +Duplicate of +OK +The difference being that similarity computations is parallelized in +One reason for using this technique is that it uses standard interfaces +You could try to nest your products List as part of your form object and register the custom editor on the binder that is used for the form object +Closing this issue since I have not seen it since +Ill run it past the other engineers to see what they think +I know my workaround is working but still this is a regression introduced in. +push out to +This is because the filesystem never gets initilized with the right configuration +Thanks Virag. +release audit License header is not added to html and jsp files +This most likely wont get fixed in but Ill leave it open in case someone is motivated enough to try to fix it later. +Paths belong to file systems +Not fixed in +Thanks for looking out for our thrift APIFor first error make the hadoop jar come before the hbase jar or the dir with hbase webapps in it +However the main issue is that the causes large waves of major compactions versus a small batch of or +Please do so +Default is true +Yes it works with Karaf as well +a patch attached +Jackson output is more sane and more intuitive but different than Jettison unfortunately +So ok lets go down this path +Ran some testsbash script doing create table with regions disable drop +Bulk assign all unscheduled issues with priority major or less to target +Tom thanks for taking a look +When you first select the row the extra pixel is not there +We do need this patch +Revision . +HelloWould you please try this patch? Thanks a lotBest regards a lot Stepan +This property is called +Has that scenario been tested with this change +Its reverted now let me know if the issue reappears +Commitet a modifed version in r +The api is in place but the new methods on the shard server are not implemented +Lookinginto it I found out that the tika jars included with the JR distributiondid not include all the needed dependencies +So reject it. +This appears to be broken again in SVN HEAD +Can you take a look +Same issue with SP +A patch that displays performance counters in nanoseconds +Hi RyanThe implementation does only support json protocol +In became a subclass of and both include the query component. +You mean get it into the trunk and backport to? +We arent going to add some Scala code in the project unless we have people supporting it +We can do more complete solutions yeah but this seems like a simple thing to do that would narrow the gap significantly +Then if it works Ill apply itActually I dont think creating a lot of issues for every findbugs bug and applying them separately makes sense +Triaged for +You will have to set the www to a file protocol path inside your sandbox. +probably does not work if the mapped superclass or embeddable object is reused multiple time +I will take a look at this issue as soon as I get some free time +Thanks for the comment Alessio +OK I now understand why you made that change +We will fail some specs and for people who want the extra precision it would suck to remove think we can probably figure out some clever way of invalidating it +I believe the problem is that Hibernate interprets in a query as the start of a placeholder for a paramterPlease try escaping as in select from a where asdf asdf +paginatedList above is of course my own implementation of use listin there any solutions for this problem +Updated path in patch +We need both patches +Obtain guide from is a formatting issue in the chapter Pooling Strategy on the last row of that big table on page values from the second and third cell are unreadable +we try to improve the repo with some minimal packaging and license information +Cool I think that is the biggest single win for this method yes +I will follow the same policy in. +Thanks for reminding me my lazyness Todd trunk is now clean +WDYT +applied thanks! +Annoying we moved IW as required into MPI think Im to blame for thisbq +This hybrid approach is what many native mobile apps are using like s iPhone and Android apps +Whats? Is findbugs also configured in reports section +Anyway it works now +You might want to rename the helper test bc it causes the Test andor Test includes in ANT to think it is a test caseReally starting to come together +Note This removes the unit test! I tried to get the test to work in but eventually turned to doing manual testing on my cluster with my jobsAll my testing shows that the patch works as advertisedIm moving this into a busy production system and will report back if there are any issues +Merged to the branch. +But a umask needs to be negated before it is applied to the user permission +Also I changed a few other classes where null was being passed instead of empty string for a namespace uri such as in descriptor serializationIt would be great if you can verify that the trunk version now works for you. +The problem occurred for him as well +Committed revision +Thanks +Also included is a small change in the way the parameter was used by +How do I use this +Merged to the branch. +If the index was rather than unique current code would just work +Im puzzled as to why this bit me but the fix seems clear enough +I didnt realise you had commented Enis thanks for thatThe nightly target seems to be working now for me although it fails due to broken tests throughout moduleI would like to commit subject to further testing and begin work on some of the tests in the near futureThanks hasnt quite worked as expected +were not developing Deltacloud any longer +For me your proposition above is ok could you provide a patch please +no longer related to Maven. +Then when I attempted to upload the file via flash it gives me an HTTP error invoked by on +Fixed the broken values as discussed in this thread +This is a struts problem +Ondra did you generated some discussion for this +stack +The model MED can now be updated to the version stored in the registry +Roger It looks like I didnt set Unix EOL on the previous patch +I would not hold up for this issue +This is the full description of the release including the overview the list of fixed bugs and the list of significant +Finally getting a chance to look at this. +Also many of those projects dont have logos +Passes derbynetclientmats and jdbc suites. +I tried building with IBM SDK SR on XP and the build was successful +However the other tasktrackers are unaware of this +This may result in a plan +Patch attached +In strict reference mode theres no tolerance for completely unexpected references though it can certainly handle null values quietly +That appears to be the reason I cannot apply them +ThanksBest regards by Andrew. +the last part is +Splitting into inheritance and aggregation poms should improve to +Bouncing to to better reflect the work that has been done under this issue +already possible using current code +committedi think we want to keep flushlargestmemtablesat since it measures directly from the GC its a good complementagreed re deprecating settings but lets see how this shakes out w more testing first. +Thanks Deepak Rishi for the contribution Your changes are committed in trunk at rSpecial thanks to Scott for providing useful commentsAshish Scott for valuable commentsAnd Thanks Ashish for commiting the patch. +I voted for this issueRegards +Work completed we now have a fully working scout maven build +I think this would be a useful addition +No I dont think I can contribute this +Thx you probably noticed the patch says rev +ImprovedYes it works that way thanks a lot +thanks for the bug report +add another test setting the actually configs and running a job fixed missing Test and upmerged to latest +I added a test case for the flag truefalseAbout the logging uniqueKey yeah I could get the uniqueKey but it cannot be taken from cmd without schema +Thanks Dan! +This may have contributed to your issue +Basically the kit remains as is we do not need to add any jars +I am enhacing them using the maven plugin as the site suggests +regression tests ran without failures committed the patch with revision Will wait until theres a clean run with the jME before closing this. +Well there is a test right over thereI specifically asked activemq to wait for my before sending a new message and send single message at a time until I acknowledge the last message +Most of the problems described in comments have been fixed in one or another of the of +That is they are not related to any interfaces +resolved in master branch +Rebased and merged into the master branch. +My client hides most of those headers so was only seeing the To field. +Rich can you take a look? verify +Thanks +Thank you. +It should be a runtime expression but there should be no null pointer exception +finds the number of cores or NUMA nodes or physical packages and expresses their relationship +Some one from mapreduce can comment on thisbq +Given the recent work on in Axis I think we can have this for M we will need to use the nightly snapshots for testing +lets not bikeshed here +Maybe someone else will pick it up +That shouldnt happen if the system decides to retry the transactionIve requested more information from the person reporting the problem +Seems like we forgot to change f to only for getIncluding this patch +I also tried using kill instead of shutdown with no s not easy breakpoints arent very useful because its fairly random as to where it hangs so Ive just used ioformat I think the example shows that its not in escript though you can just put a copy of that on your source path and add some ioformats to those functions I commented on +Updated patch job scheduler now uses in addition to +I just committed this +Also updated and to determine the of a stop +If an exception during syncing is handled in anotherthread this can then decRef the newly merged segment to deletingit too early +It seems to be working for me on OSX +Thank you for your contribution +the stars are finally aligned and the test is working now +much more consistently than looking for particular jars in particular also updates the message that comes back from the jar test to be a warning instead of an error which allows the wizard to continue despite not finding the jars if the user wishes to do so + +Made changes throughout the source to support W Wall +Theres no way for a wrapping class to specify do not let row through if column not found or vice versa depending on default behavior of SCVFI will merge our patches and try to get some more tests finished +trunk +Regarding updating a binding via adminapi if the same binding is used across multiple vdbs will it be updated in all places? If not will the adminapi be changed to specify the vdb when updating the binding +It will be included in resolved issues for already released versions +I think the doc still needs to be moved as it is an admin moved in rev . +I will mark as resolved waiting for feedback +Current version is this is unlikely to change +I should mention caveats to my previous commentFirstly I tested on fedora and secondly I used the quickstart from github but I should be using the quickstart from the EAP download +Merged to and committed revision +thanksnandana +This issue is out of date +Maksim congratulations with your first contribution! it is really awesome +This was already included in and was created for the one added in +Reopening because implementation was removed in +Yep it works well +Committed +Ive attached what I think to be the correct code to structure the LCF security support as two plugins into this framework +bq +New patch +I modified the master file for jdk but have some trouble with jdk environment on my machine +If an action mapping doesnt have a success mapping then it uses the global one +patch and unit test that address the data for unit this patch in revision . +The problem is the format +Here is a patch that compares the value of each bucket to to ensure that buckets stay in range +I committed this +Sound good +Failure of seems unrelated + I was wondering if anyone besides me has come across this error and whether it is the result of my naivet or an actual problem? How are other people handling thisIm pretty sure its an actual problem as Ive never tried a post request that large +sorry defintely not a major issue but it seems I cant change it anymore +Good analyze thanks again! +no matter how much backwards stuff we try to add this is easiest and safest on everyone +Needs a bit of polish but that is the general idea +And big on setting up some classes that make it easier to achieve consistent paging and sorting +Can it be done with just one querySay if I wanted the first so I was expecting all SC and SC data can I query Cassandra for the first distinct values of the first component of a columnThanks again. +iocoredns +Anyone have any better ideas +decision delayed until the merge of the pipelines +Fixed in r +Damn +See comment on +Ill try to get your patch in before that + Bug has been marked as a duplicate of this bug +Hey guys sorry for all the trouble +Marking closed. +So how multiple partitions can be sent to one reduce task will depend on the custom partitioner logic +Minor comment Can you please add some description to the test as to what you are testing +Robbie please checkin your patch to to out old issue. +Hi jayghost +for this patchCraig +Its not released yet though +this is perhaps two or three issues here +Since we have had the noSelection attribute on gselect so this can be implemented quite easily now I in mentioned about use use noSelection and then when a property is nullable it will pass validationdata binding +However as I said this is not a regression +HiInvestigating the build script we dont were the site is created before being useOur reading of the script is that site might be necessary if you do a on SVN +Want to roll an RC and this is kind of important polish +committed load info rename and compaction nodeprobe +We should just do the Jim suggestions in a new issue? If you aint doing it Bryan give it to me and Ill do it +Remember that the logical plan doesnt give any sense of the physical nature of implementation +Please check that it doesnt cause any other regressions +Its been running all day and no warnings +verified documentation note copied belowNoteWhen you create an image from a stopped instance in vSphere cloud this particular instance is marked as template and it is also removed from InstancesUnlike other providers vSphere does not support assigning a name and a description to the image when you create an image from a stopped instance +Hence this JIRA can be closed as invalid +I just committed this. +I think we agreed what the underlying request is to make it possible to compile the generated classes without the destination dir for the xsb on the classpath +A Unit test would be good here so that someone can take a look into it +Agreed we need to perform the operation atomically and the skeleton you present is pretty much what I had in mind +As for unnecessary ZK reads in this should be fixed as part of when moving the leader election logic to the controller. +Lorenzo and Anthonyhave you tested the fix I suggested on your buildsIf not then please do we need verification that the fix works before we can release it +Justin I got slogged by other commitments last week but Im dedicating time and resources to fixing this in the coming week as I really need this functionality for my project +ResolvedYour patch is applied to trunk +It would be useful to see the retry in trace log We should capture all Throwable +So I am assuming that the JSON api will stop to work if you switch to +nvm +Thanks for patch v +pls post your +Can do in another issue +You can change it as you wish. +defer to +patch attached +Sounds like we will have a Lucene release in the next week or so with the fixes in place +HelloI have create a patch for the H platformIs there a problem with the patch? I think its better to commit the patch now to make merging simplerRegards is an updated H patch against the latest svn HEAD +It should fade away when we obsoleteStill having one problem with the tests +Id also question why Oracle is throwing an instead of an then Oracle are always a law unto treat any as a broken connection +Pull request on EAP ER +working +LGTM +This issue is still very much present gentlemen +Done. +Patch for branchI also fixed this for passivationA short review would be great +This patch makes the obvious fix and tests the results +has been updated to include the enum COMPLETETABLE. +implemented this feature by moving to the new connector added in. +Parting shot The JEE thing just seems a bit odd to me +I agree that we need a framework for optimizations in the backend +Shall we also delete all of the apache mirrors too +Its issue of WTP +going back to worked +Thanks and Im sorry about the name messup +Moving out of +No the first comment above with the hyperlink is just intended to define which Quick Install Guide we are talking about +Very nice thank you Suraj +Workaround is completely ok for me! +plugin was renamed from jbossCache to +The issue reproducible with on also +Alejandro Thanks for your comments +working on an interface for it +Hi JoshIm sorry I didnt check all the path that could effect the http method urlAfter went through the code and I found a way to verify my change and find a minor issue of http componentI will commit a better patch shortly +Sorry I hadnt updated the test +Thanks for the backport Jerry. +The patch looks good to me +rayeesn to add moreThis problem with KVM runs is seen intermittently so in this case it is possibleI was able to reproduce this problem even through UI on vmware +classes updated to take advantage ofthe new preference architecuteLet me know what you think patch adds receive buffer size parameter for and classesFolks any feedback on the patch so far? patch committedI am planning to submit one more hopefully the final patch shortly The patch removes those parameters that are better served bythe specific parameters changed to takeadvantage of the new preference architecture Do you think that the following parameters should be deprecatedgetsetgetsetA few comments I think the connection should be left on the +Changed Target Version to upon release of +could we know in which version of mahout will user based recommendation be available and what exactly they are doing +This got fixed at some point +Resolved in revision +Doing this will allow modeshape to reuse the templates to create its file that have its defined JNDI name +Sorry for duplication +trunk Committed revision branchx Committed revision +Except a soonI think it should be in the main Pig branchAs an aside what is the plan +Fixed. +The modified code still has a leak in at least one place +Struts in in order developed by Deepak file extension +Not a single glitch +ugh I just noticed we have vinting in Bytes class +The changes described above were applied in revision +Kiran did you already start porting it to +reopening to look into whether an can be setup while is called on the domain services +Other than that it makes more sense to encapsulate the use of in the as the current version does +Similarly is not persistent for now +Just to clarify my above comment on REST it seems that if we would have had someone in the past to drive Thrift we would not be in this dual interface predicament +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +sounds good to me +Ive patched release of solr +mrepositoryorgjruby should just do it? Shell is important interface +we can insert the range that failed into a system table on a replica that succeeded via thriftYoure right we pretty much need to do that anyway since we have no guarantees that whatever process the Hadoop job is running in will be around later +I implemented the functionality for doing this on xupgrade to rails branch +For and the OS is LinuxUnix for which there shouldnt be quotes in those parameters +it behaves exactly the was more of a feature request +Im going to mark this as being also a bug and target it for when well look at moving to +Attaching patch against trunk +Arun I dont see the ceil to rint changes in this patch +The user is also expected to customize the plugin settings by registry URL if they need to +Attaching the initial patch +Accept wildcard args for copy constructors and Unmodifiable decorators +I still would not try and fix the malformed URL +Updated patch to cover both for the updateCommitted the patch to trunk with revision +will enable the to be specified on the Roo command line for the entity command +I havent removed the preference +Obviously the is not used in there +I dont know what changes this suggestion implies though +All licenses on the distributed software need to appear in the LICENSE file +merging to +The configuration for SSL support only works when using the java based http processor and not with the APR based transport. +There are some build issues on AIX +Requires no command line arguments +Thanks for the extra info Sean +META +Please attach as separate zips your project and the project using the Roo backup command in each case +But I dont think that I really like thisIf you want to integrate this you can have an Maven Eclipse plugin setup to have the process run on demand +idea how I can get this fix +Thanks Stack for reviewing it. +Why put in place a solution youll just be changing again in a couple months? Cant we just wait for and go to the final solution instead of a temporary oneHow does pointing the builds at the +Should be fixed. +Could you please test it with the upcoming release? If you still have problems please reopen this issue +Patch available +revision See also +It takes out a read lock while its caching the row so snapshot aint cleared while reading row +I think you can close this issue since I cannot reproduce the problem problem I had must have been with the test application I originally created with grails +If performance becomes an issue we can backfill the endRow parameter later +Upgraded bundle plugin to with svn revision artifacts has been created and added to the repo. +Reopening for handling in +Since PUT returning an inserted object back is also perfectly valid scenario in REST Spring should allow is of course calling execute method manually but I dont see reason not to include those few convenient methods into interface itself +Only little initial effort is required to move towards stable world of Unix derivative OSSorry for the OFF topic comment but I guess this would be fineAshish +low priority ve tried the site on Opera but I didnt see anything wrong +Avro should be able to have a union of two different records with two different namespaces +This is working in rev +In your use case I would suggest using a dict from uuid to structure +bulk close of resolved issues for release. +Thats right this doesnt work +Committed to and trunk. +Before committing can you remove the extra commented line in the reduce script and add a comment explaining what it is +for the modifications +I was very confused because the output files were showing SQL Exception while the code definitely sh owed it should be saying Then I discovered that the +Closed. +fix version +Can you provide a code sample that demonstrates the issue by opening a dialog via a context menu +bq +Also if we go this route we have to think of the case of process error andor process input stream buffers being full and thus blocking the process from finnishing what it had to tell us +About to apply it but faced some password troubles +fix findbugs described earlier the test failure is just the fact that jenkins failed to apply the binary diff to the editsStored file +no such class transition resolvedwontfix to closedwontfix +I will update our internal Oracle bug with the new with does not work for me +Yeah I fully agree +I think I found the remaining load path when running inside a jar includes an entry like this under fileUsersheadiusprojectsjrubyliblibrubyThe problem however is that this entry does not exist in the complete jar +Thanks! +The issue is caused by +It works for both servlet and portlet transparentlyIt should be applied to a Tapestry final source tree +Submitting patch for review +new patch version +I updated to and now show seems to work but autoHide does not +Maybe it can be an error of JSF view +If we have to respin for some reason we can also add it +Few lines of code which bring a big benefit I think its worth +Regarding my first commentI just tried again setting the Spring version to and everything works as it ought to +There is also an issue with the blocks queue that make it worse +we have to migrate our code towards the castor branch anyway so it would be no problem for us using an do you plan the next regular release of castor? if its out until early august i wont get in a hurry +If I understand correctly storing tokens in something third party that is not kerberized defeats the purpose of kerberos +Looks like weve reproed the case +Hello WernerDid you manage to compile and run my codeAny idea of what can be the problemThanks a lot and best regardsGonzalo Abollado +I think we can adjust priority slightly for consistency with our other issues +Agree that we should move the source jars off the classpath +Freeman thats essentially what I did take a look and let me know if you see any issues +Created a profile in assembly tests which can set the related Rahul! Committed sha faebcefcbfbeedaa Im now triggering a new build +Still need to add back local optimizations mostly doing rename for a move rather than copy from one Directory impl to anotherIn pretty good shape though +My home directory gets auto mounted from the NFS server and the trunk branch is checked out in my home directory +After hrs of build time I had to kill this test again via VNC? Sl qatoolsoptjrebinjava true qahudsonwsworkspacetestsjboss qahudsonwsworkspacetestsjbossseam qahudsonwsworkspacerequirements qahudsonwsworkspacerequirementsjbossas qahudsonwsworkspacerequirementsjbossas qahudsonwsworkspacerequirementsjbossas Xmsm Xmxm Xbootclasspatha +Looks like I messed up something with this one +cancelling it +Any ETA on a release of of pdf plugin? +Fixed in revision now run successfully with the default heap size of mb. +They are identical except that the broken one contains lib! a problem with AS but with JSF +Patch incorporates comments and bq +I am able to reproduce this issue under when running groovy from the against uncompiled groovy code +thats life +Basically this repro is locking with a timeout and I cant work out how to get around it to actually get the repro to work +The identifier does not get underlined +Much much faster +Patch applied at +Sorry for being delay please find the attached war file please add your jar fileand test the following URLhttplocalhostfileu test it on view scope it is not JSI attached the sources but without the jar file please checkNote Appserver Jboss Thanks just now Ive noticed that you have upgraded JSF version of your container +This class can read the XMI format introduced in UIMA as well as the XCAS format from previous versions +It turns out that for JDK because there are so many different types of exceptions in we have replaced the specific exception class name with SQL Exception +I write a test case and run it against with the trunk code +The code was already committed +Is there a description of what happens when a client wants to append data to a file +You are in the group +Then people move on in life and there is no one to answer a question on the feature +I am still facing this issue in the released version of Apache Flex +I didnt noticed this dont know why was closed without to fix the me we must change the schema and to avoid errors from for me the choice is id or groupId need the model and the xsd can become a version instead of a +Sorry I didnt see that before creating this issue +I dont think there is anything we can fix +the pom cant be changed youd need to request a new upload with a new for the quick response. +Regarding your changes to remove the unused dependencies +where did we document how to use this +I think this is holding onto information somehow Ive just been repointing my drivers to the new jar when I upgrade +This fix should be in JBDS and closed. +Lower priority to minor see previous patch referred to in my previous comment. +Ill look at it when it will be would like to contribute to this feature but dont know how to get started +if this is done at the coordinator level on a failed write the leader has to be maintained and the mutation has to go through hinted +Interestingly it seems to be reported for a task that finishes without problem +I removed it here +For tooltip API are correct in s necessary to check it for other components find components that have API in guide its a separate section run component check whether it works with do or without do if with do check that in guide its written with do if without do check that in guide its written without do if both work in guide it should be written without do for for for panelmenu +As a general rule if you need an ordered set or map as a result of an operation in Java it is much faster to use a or and then order the results afterwards than to use a or and iterate +I tried solving the issue by doing this also +BULK EDIT These issues are open to be picked up +I know I have put up a patch I think +Isnt this solved by? +added in the joinDesc to indicate the tableDesc of the inputs to the join operator +Yes we should automatically add this attribute. +The testpatch result is incomplete + +Hmmm +The patch utilizes the instance now +Suggestion minimalistic fix for this problem is the followingIn the class add a private boolean member started +Fixed in some old version of +Any concerns on this +I was mistaken and mentioned but the shared library I am forced to use is actually +Let me ask you this how big of an effort is it for hibernate to enable autogeneration for non primary key columns? If you say it is going to be a huge one then I am ok with not pursuing it +my mistake as my repository got messed up +Ok that makes sense should only set this variable if the WMS is not accessible through the default path which is httplocalhostgeoserverwmsBut I have also made a mistake in the documentation the getcapabilities document is no longer the correct URL it should just be the WMS path. +There seemed to be a number of issues expecting results to be logged when the burst size is configured to be and a lack of sleeps between tests +personal reasons +Awesome thanks RobertNikola can you switch to the fieldtypeconflicted branch and resolve the conflicts fix the newly added code? If we do this quickly then hopefully trunk wont change much in the meantime +All collection proceeds from an equator record serialization moves forward in the buffer while metadata moves backward both as circular buffers as in the existing design +I disagree that this should not be fixedWhy would this be a special case? If touchingchanging is supposed to redeploy the app it is a bug and not a special case +Build definitions +HiI think this bug is fixed in Ive experiment the exemples and all work well with the groovysh +emulator is correctBut need to fix build carlos not sure if this is resolved if it is can you close the how are you testing? This should be in the latest bits not any released version I believe +We could use the same approach taken in a patch ripped off directly from I kept the old target commented just in case +Cleaning up old if you verify this was fixed +fixup +If you see a more elegant Java compatible solution please advise uswait at the moment is used for just one highly specific case namelywaiting for response +binhbase shell you now get raw jirb +Attaching a repro for this issueThe idea for reproduction is to create a table with large number of records and delete most of them so that atleast the last is empty and then attempt to do a inplacecompress +The project is an ordinary jar packaging project me should add a java nature unless someone does something to prevent it +For the moment I wrote a sieve matches pattern to java regex pattern translator and replaced the whole GlobORO stuff in jSieveCode is not tested but I think this is the right solution raise your voice to complain Ill add some tests next week if noone preceed me +Maven has a strong preference for convention over configuration which means that requiring the source and target entries in the pom is not really reasonable IMO +key to the bug is requested +batch transition to closed remaining resolved bugs +Windows errors are fixed in revs r and r. +If you change your test to extend I think the goes away. +I dont think it can be marked as because exec allows you to execute any command that may damage parallel builds +Steven I argue that your quote from the standard agrees with this issue more than it disagreesagain its up for interpretation though +New rules DLSDEADLOCALINCREMENTINRETURN DMBOXEDPRIMITIVEFORPARSING NPMETHODRETURNRELAXINGANNOTATION issues with Maven Central resolved +Michael B sounds good +Thanks Robert +Any solution to this needs to be part of the larger solution to allow definition of metadata. +The patch and the regression test are attachedPlease mark this issue as Patch available +Hi Jacopo Here I am attaching the page of the work if there is some improvements then let me knowAnil! I missed that +Most of the merges with in and releases is fairly straightforward +It works. +Yes version is correct +You only need to add this parameter if you are running from the root directory +Thank you! I knew i forgot something important +You still need to manually remove a Added missing +Cos let me try this on +I will remove some of the unnecessary log statementsThe patch consists of two parts +One more bit of advice try to find a way to extract the popup text from the online help file even if that means the online help system needs to be redesigned a little + +You should be able to start with the current group operator as reference +I think this is also valid for most of the decorators +Seems to run fine also for meDo you have the associated log file? +Correct patch +Wait no not SIGINT but I seem to recall two that the kernel doesnt allow a process to override +I think we can offer value on the list by pointing to examples or anecdotes without getting into the world of legal advice +is only in Java +Integration tests still pass thus this change is stable with respect to existing HTTP POST we resolve this issue +bumped to +Its buid againt revision +Holding open for a bit for Lars to give the version a look if hes so inclined +Its good to hear that mina ssh client is up to the task +This patch is nearly the same as v but does not included the generated thirft code and changes the reference to keyspaces in the commit message to Obviously the same change was meant in my comment that came with the v patch +Added switch to turn off processing of undeclared results. +Ok that section of docs has been updated. +Forgot to add the flush forgot the getConnections +fails so im guessing whatever broke this was committed in between beta and of +ChangeThis change may be brought back at a later date if a send command is implement to send input to the STDIN of a back ground process +Added more tests for slow operation counting +And do not use archetypecreate but archetypegenerate of +Reviewed by Ken Giusti. +I believe the signature should be changed +re not marking enhancements as LATER +Fixed in rev +added an excludes field as it you can add some file patterns to exclude. +Now that is resolved this regression only affects and which use lob locators +We want hdfsproxy to act as datastore servers +All Solr tests pass +client Xcvs buildoutputclient Xcvs ublicbuildoutputclient Xcvsbuildoutputclientjmx Xcvsbuildoutp utclient Xcvs adbuildoutputclient Xcvs licbuildoutputclient Xcvsbuildoutputcli ent think Ive fixed it +Thanks very much +I maybe doing something wrong here +plz make a clean definition on for all +I am going to change it to enhancementJust imagine situation when client has filter which is applied to the different types of messages and some messages do not have property which is used in filter expression +This is not junit or jetty something like that where a common place to rev the version would be nice +Rake Rant +Wed Feb CST orgpicketlinkidentityfederationweblisteners +added a fix in fbdbafcfceedfaafcb to avoid infinite compaction on a single L sstable that isnt large enough to promote. +Hence the tests are being skipped in these two modules +Fixed on trunk in rv +Yes I understand and I felt that the first release was cannot move away from castor because its a transitive dependency from an other project and I am not sure that we could have the source code +In the mapper the heap is going to grow huge and GC will not do anything +Moving this to Backlog per email from Grant. +I believe this is done please if not. +I now see smaller GC pauses and I am achieving higher hbase perf than before. +Cool +Actually this may not throw the as expected in the supplied unit test +p typo may want to customize for group of brokers +Reopening to get. +Solved in EJB for pessimistic locking in Q and for optimistic in Q +It just invites questions of legality +is this error message expected +Did you manage to get hold of an existing despot? +It should be tested there dont know what Im doing but the two attached fixes worked for me +I tend to agree with Thiru on this issue +can we have the option to set diff keytabs for hbasemaster and region servers +Sounds good Brian appreciate it +Once aliases are used column references must use the correlation name. +Please review +We can reopen if still a prob +proposed patch +But when you ask to returnPrev a working open core is guaranteed to be returned if a previous one existed +Issue can be closed I guessAttached youll find a hotspot list for the same request with a patched +why does your timeout callback not close the connectionA timeout is just an indication of some problem which may or may not be temporaryIt should be left to applications whether they want to close a connection or retry the communicationjust my Roland +Let me try to bring clean up with null checks and close the streams in side try itself and retry on finally with +Data is not read sequentially that depends on the larger table being read by the mapperIt includes random reads +Whilst the client does have some issues with handling of asynchronous exceptions I dont believe that this issue will improve that situation as the waiters will not change during the connection creationclean shutdown processes +Reopened to set Release Notes Docs Status field +Please run all the unit tests so we dont introduce any other failures should be OK but best to be safe on the stable branch +Please take a look +The javascript is identical to the javascript for the Jira Repository Connector and seems to work fineThis is tested with Firefox +After patch integration all Classlib tests related to this bean passed on Win and Linux x +Attached new patch with the Apache License information +Since each slave starts differently we cannot set the since they would get different indexes based on when they start +otherwise i cant apply the patchthanks patch with license granted to patch with license granted to apply the attach patch which handles issues with the WAR Minimal Standard distros +The portlets will then be able to process events and actions and serve resources but avoid rendering which is obviously a waste of CPU timeThis assumes the portlet developer is not changing any state in the render method which is the best practice anyway +This problem was originally framed in the context of the writer and I guess I just decided to give up on matching a CRS to an authority code ergo it got closed +Well we are building EAP now so we need it tagged and uploaded to asap +Create two users both in the users group +please see discussion in +mostly con is readySTAB is almost ready but has a bug when using concurrent sweep after dozens of GC iteration +Rado please verify and close +Patch attached in +in other words as it appears now it would show up above all of the servers listing I think this is valuable screen real estate so this info can go after the listing of all the groups nodes etcGoing back to the support of http get actions on the server from a browser I am not sure what you are saying with a servlet or https? For example if I am listening on +Builds changed as detailed above +Following patch corrects the issue +The move is done in the CVSI think we need to update the docs about this too +Is that even defined in your +It is possible the numDecimals value in the file is being ignored or set to zero +jar but correct is deploylibjuddi +Hi BryanHere is the java file I reproduce the bug +Attached is an updated from SHALEHOMEframeworktrunksrcsitexdoc +Resolved +Ivy now throws an error if the parent pom cannot be found +I just tried this out and it works well in the src distro +RepeatFeature requests without adding at least a sound usecase is almost certainly not going to be implemented anytime soon +What could be happening here is that your consumer is using an ack mode of auto and the broker has sent the consumer all the pending messages because its prefetch limit has not been set +We should probably make it configurable though whether to rely on the driver implementation or actually fire the validation query +Forge documentation is not a priority and will probably be deferred +Thanks Thilo! It looks good +Resolving +Whe do you think Pavel +This action would be used if there are downloaded runtimes in the file systemThe Download and Install button will download the WTPSeam Runtime from the internet and installsetup it using the Runtime Detection API +jsonp GET or POST hangs with CPU Shameerayes youre trunk build works like a charmThanks +Thank you Dag for looking at this issue and the IBM JVM +Several tests are there speaking for bar being an invalid name for a field in Groovy +At the moment I have to use a debugger to see what Id ideally like to see in a run log file +modified etc to include in the copy to ant task +Moving to Portal Driver category +A couple of bugs have been fixed since the last time I looked at this +NOT applied to just to and trunk +This is source compatible but not binary compatible +Also if you dont like the names and want the mixed case names you can configure the introspector as guess I cant have both of them correct? And thanks for the detailedexplanations +Ive never seen anything like that +May be you can suggest some other secrets how you reproduce the issue +In an ideal world everyone would use the same library versions but since thats not always possible users may find themselves in a situation where they need to change a library version +Just want to make sure this doesnt get hung up waiting for a commit +Adds a processing instruction +is still open but do you want to verify this one and resolve it +initialhost list is realy quite long host but portrange is +Not a good idea +Yes oops I just caught this bug myself! Sorry Im still kind of testing the patch myself I may have posted it prematurely. +On the main trunk Committed revision On the branch Committed revision +This patch is oddly formed +Otherwise it will be in a later buildrelease +I have a patch for this in the review process +Closing please reopen if it becomes an issue again +Attachment has been added with description source +The overhead is probably due to increased state in the object as well as an unpredictable branch in the inner loopIm not too worried though +Thats exactly the case +That is why i initialized to trueWant to change it +the comment the issue and checking it out now +Added proper comments in the patch +Thank you AndrewPatch applied to NIO module at repo revision rPlease check that the patch was applied as you expected +Rooting out all of this type of thing is a bign +Yeah +Im going to send an email to the user mailing list to explain this choice. +submitted for hudson tests +I generally try to only use that when youre exposing some piece of internal state that shouldnt normally be used from the main code +I wanted to write a reproducibleconsistent regression test fot this bug never got around to it +For some reason one of my scripts was callingmaven Dgoal build multiprojectgoalwhen it should have beenmaven Dgoalbuild multiprojectgoalSorry bout that! +Awesome +Closing tickets +Patch looks good Ill commit shortly +Thanks very much for the dependencies though helped me fix a production branch build that went bad over the weekend in the wee hours. +PAX URL is able to handle version rangeI see that the version that you report is +First iteration available with commit aeaffdbfaaccd +Committed to trunk +Resolving as fixed by +If you want to have the jar download a more recent nightly +Both are importantNot sure why the code bothers reading the fork and compilerVersion settings +Thanks! +Batch files based on the current batch files are next +without leader election we wont be able to run +The only real reason I can think of to support defining a column as unique is if there happen to be databases that do not support constraint definitions but that do support marking a column unique +Currently every new file gets their permissions from default +To clarify Remy Yoav I are owners for the various Tomcat lists +It is still happening on groovy +This should be committed after Avro is released +Wouldnt we potentially do a lot of splitting when there are many regionserversEach regionserver would split with the same growing reluctance +Attaching a patch that updates the Javadoc in interpretive that more accurately describeshow to overrideset xsloutput properties via the JAXP API and it includes the interpretive extensionsto the standard xsloutput propertiesThis patch will resolve which is now marked as a duplicate of this issue +Export application wizard is now available +I committed the patch +I could make a patch that processes a basic CSV file of m points but I still wouldnt consider this issue closed as the discussion regarding standardizing the inputoutput formats for Mahout still seems to be ongoing and isnt something that will be resolved for +New patch fixes writesave writtensaved etc +WilliamWhat do you mean with +I translate the Chinese simplified version of the site and resource bundles and attach them here could anyone help to commit them to the svn repository +Hmmm +Waiting for wstools issues to be resolved issues for already released versions +Gail assign this to you since you asked the test case +Then I tried putting set back and simply add and my app uploaded one picture but crashed on the second one +Those sound like useful improvements +and are for performance testing only +Thanks! +to commit +FYI we do already support a fallback to FORM based authentication and do have a configuration option to strip the realm name +I do have access to a windows x I am missing is the proper knowledge to actually be able to compile the dllI opened a new bug for it +I committed to subversion as revision +Cant reproduce on questions fix version is and but the commit log only says it is applied in Isnt there a better fix for this than catching? Cant the stackoverflow be avoided upfront +I have successfully tested it in my local environment using glassfish v and v UR on JDK +The fix is OK +This allow the updateDependencies afterwards to remove the reference rather than saying the reference is still existing +I think Ive resolved this one +The message memory usage instance is set to the topicWhen the message is sent to the queue the memory usage instance is not set to the queue memory usage but it still uses the topic memory usageThe message set method is used to set the message memory usage but this is only performed if and only if the current message memory usage instance is null +Attached the changes to ensure small tests run in a separate jvm similar to the large or medium tests +We just caught up with Tiles for Spring after Spring having been tied to Tiles +I just committed this to trunk and branchPlease pick up for other branches if applicable. +It should be fixed in and shortly on the stage +patch included and testedchanging to fix in +I hacked together an initial patch that uses a per url with thread and a queue of +Sorry for the false alarm +Just ran remoting testsuite to see if it would pick up this deadlock and did not see +committed thanks! +You should change these to match whichever ftp server is accessible in your local The ftp server The user name for The password for The base directory for the testsKurt has recently asked for the build to be changed to use an embedded ftp server but overriding in this fashion will still be email address is hudson using to sent reports to the mailing list? We will need to add it to the group + branch no more under active development +Where are the Android guys to affirm my nd hand knowledge? Question is what is should do in this platform because someone who uses will probably expect that his check will succeed for Java also on Android +Attached patch passes most of the existing testsThe is still in a broken state on the branch but will be cleaned up later +Attaching a patch for get that appears to work +Resubmitting with minor change +what is the bug id +changed the error packages old date issues. +Folks is it WIP? can you also provide fix for +I have XML and properties on my projects and it still have no idea how important is to escape colon but if this is truly required should be done only at +All the regression tests ran cleanly +Sorry for not thinking that through sooner! +It would be best perhaps if you tidy up the comments here so that it is clear what you think is achievable in the scope of +When I accessed the repo group I was still able to browse everything and when I tried accessing the repo which the guest user has read access I was prompted for authentication +A fix is in SVN +a project admin of A shouldnt even see the roles for project B +No its an error with STATICCALLSIt doesnt do proper substitution for property methods +In other words its impossible to cast a bag of tuples of bytearrays into a bag of tuples of longs without this patch +I think Andreas patch is OKThe old headers will be obsolete after the transformation the content length most probably will be different the encoding may be changed as well the content type tooId also like to add this as a test case to +Committed to trunk in rGonna wait till I backport to before closing +I will look into supporting multiple reports for the next release +Thanks +I have no clue what can be causing this so far but it does help narrowing it down +metadata +Can you please try the latest marvin? As I showed almost any associate call shouldve failed regardless of the testcase since the change is in the library +hi Deepalsorry for any inconvenience caused +The new allows to seamlessly integrate this pattern +Thanks Andy for the implementation +And its different thanFor numeric queries its no problem at all for of course it is slow +It can be worked around by specifying the value manually +Syncing patch with trunk +If any setting are added in the future they will be. +This made the commit succeed without the NPEAndrus thanks for looking into it +Im happy to update the file but Im not sure I know how to go about getting myself setup for that or even if Im welcome to do so +As I said though I wouldnt object if someone made this parameterizable +Please if you are still having a problem. +Project Examples wizard is documented though Quick fixes options and how to call Project Examples wizard from the Help menu are not described yet +for Trygves suggestion +Have you got any idea about this problem and solution? +Thank you very much again +cwhen the cli prompt comes upreloadThen look at the console log in the st window +I kept the Listener and the Responder as separate threads +This needs to go onto branches and trunk +Well make it a megabyte then +Thanks again for the report +Amar Can you please please share the numbers for data locality which you are seeing? I am running Tera sort for my benchmarkingThomas tasks are off rack and rest are rack local +Patch applied with thanks +Does the fix I proposed work for you or notSorry about the protected method you will have to extend to expose it publicly I guess. +Andrea does this look reasonable? Id like to commit it on trunk and stable +yes +Many thanks guys for helping resolve this +Patch to fix appendingsetting shared data +They are very helpful +Disallow nested snapshots in the first phase +When I looked in the log of the producing broker I see the same stack trace +this was done for beta. +I have applied it to and as well +The assumes that the is some sort of deployment descriptor and therefore doesnt accept the Jar +Hey Markus it was more a question rather than anything else really +How odd I dont think I got the emails for this +The ha bundled is made to run on a cluster you cannot run on a cluster so i think its better to let the user define its own database anyway since it will break im not convinced we should ship a at all +Patch applied and then some. +It should really be DEFAULT which is the reserved keyword +Pls refrain from sending further messagesI will just personally contact you with my new email addressFor any concerns issues inputs relating to your transactions withExist kindly forward email to hrThank youHR Dept +Something in me makes me avoid using the as I see it as a hack but I have no ground +This issue is not approved for SOA CP +Attaching a prototype where message will stay on a recovery area for a timeout +Tom I have some questions +Should be resolvable through a registry key please do not specify the probing path manually +will reuse the partitioner etc code in the server to determine which nodes a given key is mapped into +file is now updated w newer versions of the +As I commented in the other jira we should see why they were not doing soWe should do some reconciliation between what we find in ZK and what we think is online based on but not sure exactly what course we would take in a state like this +Patch committed to trunk and branchThank you Cheolsoo for your contributionJarcec +Thanks +Is this waiting on some action from me? Just let me know if there is something you would like me to doit might be easiest for now to just remove the tests which use the binary files if there is an issue around those +I can run it in my eclipse workspace just run a junit plugin test it can work wellAnd I run these tests by the command mvn clean install Phelios +Already fixed locally +Ported from trunk to branch at subversion revision +It was added to reduce the cost of Map +Sorry this patch is not applied to the trunk +Adding patch with code changes +Now that work is well underway with AS all previous community releases are +A tuple returned by a loadfunction has nothing to do with the comparator logic +I added a note to the javadoc of the behavior of the implementation of loadText. +Good oneI appreciate your bringing me along +committed to master +Would it solve your issue +Ie onException is now always calledIn addition in the event that a close is done async from an onException there is still an opportunity to have a normal interleaved with a Forced exceptionI think this is the same as with a pass through on exception a close can get called twice but this is handled ok by closeMostly though I am wary of the change in behaviour introduced by the exception handlerAs this is a filter that is added by choice it is not such a big deal but we may as well iron out the detail +No need to increment pisoa teiid savara since they do not depend on JBT components +Please cross check and close the bug. +Closing +Retry template now rethrows unclassified Throwables as private inner and unwraps this type of exception before passing throwable to collaborators +Please LMK if you still have problems +job file its a +It actually is dependent and is not a function of the service itself +Attached +Closing as not a bug then +Can we perform these tasks at the startup of the service rather than at initializationYes sure +I will do no +Because T is still not in the cache of the ANN we can trigger the same issue as described in this jira +Maybe Ted Dunning will grace us w an opinion if we give him a chance + +As I discussed with Daniel please do such things incremental instead of big bang +This patch makes hdfs tab completion work makes the bash completion work for hadoop being anywhere on PATH and adds job completion to tab completion for currently running jobs +subselect join at the physical level +Main changes no freemarker new service create is now minilang less new labelsMy limited understanding of services suggests it cant be used for create which creates two entities +For some reason they were quite class name based in Hibernate and they still are for many customization purposes in Hibernate +In fact this feature was disabled since because of a misunderstanding i added a new portal template called Basic Portal it is different from Classic Portal that one has navigation another has none +So i think i can try to write a class for Geronimo +Thanks Suresh for the review. +If you wish to create a separate document describing all the quickstarts then we can certainly look at doing that for a future release +I added the test case and it runs green without problems. +nope will look at this first thing +I had to fix it to ignore SUCCESS files found in Hadoop distros +Testing then will +This is probably reasonably as hashCode is a system function and not a business function +So if this is caused by chunked then it really is an issue of the servlet containerCan this problem be reproduced with Tomcat +trunk +Can you fix that? ThanksRegardsHenk t really tell from here whats missing is path corrent in theis there a parent for the vhostdoes it have the icon definitionsIncluding in the vhost may workfyi your cronjob appears to have stopped last rsync connect RegardsHenk crond wasnt runningTook care of that and resyncd Ill look into the icon issue soon as I have some time for it +You could use here to notify sync thread to force a full flush or just force flush entry logger after that it is safe to force flush an ledger to steal its ledger cache page +Performance part I will ask on users list +bq +fixed in rev are willing to compensate anyon ewho could help us resolve this issuePlease contact me directly at to inquire howThis is a great opportunity to earn CAH for your open source contribution +Thanks Hairong! +sorry Ive been busy with stuff thanks for understandingIve spent sometime thinking about this and I feel we can do something far simpler to address Syncsorts goal of plugging in your proprietary sort while mitigating risk to MR itselfHow about this I feel we could accomplish both goals by something very simple +Its just NOT OK to change the license and drop the original credits! +Weve been discussing this on the mailing list and weve decided it clutters the API more than it improves it thus not making it worth it +We have enough open ones already. +new patch I renamed to and made the createSlicer method public otherwise Directory impls outside of can not delegate to it +patch still needs to handle the case where the folder might not exist but is not needed +We using Maven for the builds +I will change filename format because use and this character is not allowed in filename for Windows +Search for updates of the currently installed features Next click IDE to expand the tree select Eclipse Next etc +sorry resolved by mistake +It was sent to hadoop but it seems the machine is the change requested by think reported corrupt files might still convey the wrong idea but I cant think of any better term +The problem is the current doesnt acknowledge that yet +If all browsers were as strict as the specs thenno website on this planet could be parsed +Rewrote the source repository page and all are correct now +It more likely clashes inside the server +Which approach unordered lists or stacked divs should I commitAnother issue I ran into while building the tree was that there is no +Option Use the attached patch to set only on AIOption Bump all mailets to Comments + +Im referring to the roundtrip your patch introduces by responding with negotiate if its a authThe roundtrip is mandatory for auth in your patch +Please advise +Will be fixing it in new patch +It had a different extract directory name thats why the templates arent found. +To be honest I dont remember my actual setup anymore but only that I saw not much difference between normal stream IO and NIO but the data was by far not gigabytesI used default direct buffers and buffered io +mrepositorycomgooglegwtCDocumeUSERNAME +Max bytes M +Thereis probably a semantic explanation why that behavior is completely valid butstill +Would it be better to add attributes in the appropriate spots in the file instead? +Therefore a simple solution that simply checks how long the query processing is taking is not enough +Following changes I have done regarding this jira +Thanks for the explanation RickIs there any way to avoid that ugly exceptionA much better message like Cannot update obr deployed bundles with the update command use obr deploy X instead +Committed to trunkI dont think its worth it to backport to the branch since the only danger here was if the set of recognized uppercase variants of HTML character entities ever grew one of them might contain an i since branch is though that set will never grow. +Pei ping +Also the patch adds a Java object to new system tables +Otherwise wed have to throw a not unique guess what you wonder is whether the case should be restricted to the method name match as well +Heres a context diff +If there are other releases that are missing i can import them for you directly +SPARQL will be only available in Stanbol when the Clerezza version supporting it gets released. +Attached patch exercises the servlet at scheduler while running the jobs +parameters now added for an portlet mode only. +There must be some logics i dont understand in my assumptions of +patch correcting the PNG predictor committed the proposed patch in revision +the resolved issueLahiru +I thought bump out past GA to next version +Are you able to share a project that has this problem in +We have a draft version of an index rebuilder +Right I dont think we need them under collections anymore? They will likely just get out of sync with the clusterstate I think we only need that info in one spotCan you do one thing? Back out the rename from the patch? Then I can do the proper SVN moves for that class rename as a separate issue +Change to Major +Better patch this one follows the Maven code and patch conventions +need to support sibling semantics +Using a approach to code formatting and style guidelines to me means that it will never happen +Closing this issue then. +Nice workaround Closing this issue then. +Its not just +You need to fix the findbugs warning +Im a maverick. +close off release +Yes perfectThanks GavDan +Now it is fixed +I was not able to back port in +Committed to head +Is it OK with you bamboo +So if a new API makes it into the distribution feel free to provide a plugin maybe as part of your library distributionA small tip even if we would want to include the lib into our distro we couldnt +In many companys setups it is very hard to upgrade both the client server at the same time +Other than that I was holding off on this in the hopes that switching to the Eclipse menu extension point would help matters +It would be great if Hudson could detect such problems +Done +List of features and required by STP BPMN editor in Ganymede should do the preparation for BPMN to be in the build but since we dont have anything use it yet im not sure if it makes sense to enable it yet +Many thanks to you for filing the issue. +If the file is not there everything will function exactly as before +Coolwith the version which was build an hour ago it works as expectedthanksaV +I can reproduce the error of this issue +trunk diff on patch into trunk and branches with thanks to Aki. +Per Martins request closing as duplicate of +Can you please test now? Ive at least removed the doPrivileged for you +Next release +I am not very familiar with the internals of smooks so I cannot comment intelligently on the solution you have proposed but from the sound of your solution why do you need a set attribute? Couldnt the new visitor you described use the data attribute to determine this? Also notice that this worked in version so I assume this is a recent change that broke this functionality +IIUC this didnt make it intoI wonder if it would be possible to package this template as a plugin that when you installed it got copied into the new server location? I havent thought this through but wonder if there is some way to use the plugin installer facilities to make this work +Verified in JBDS +Will this be fixed? +Marking bugs with invalid fixed for release as fixed in. +Were you ever able to discover the issue or work around it + on the patch +Perhaps this is related to issue was fixed with Native release. +Assigning to Dyre although this is an example of team work with Dyre Dan and Knut Anders contributing. +Does the hour vote start when the release candidate is created? Or when someone explicitly calls for a vote? Id propose an explicit call so that we can test the rc and then someone can call a vote +FWIW I think everything that remains can best be handled as new tickets and incremental improvements to the connector +Show In Cheatsheet View +The test should be ready to add into the suite once the issue is fixed +Patch attached + Namit for creating a review request +Again what is the settings for this scheduled task in the Railo admin +First as sketchy but with a white bg +The portlet was tested to work on IE and Firefox +Patch suitable for and suitable for +I honestly hadnt thought about thatI was actually in the middle of an editing session when it came time to go home so I saved my draft work in the Release Notes field + +What do you think +In such case I didnt see the mentioned error msg +small project which exhibits this behavior via antlrGenerates an executable jar which should print hello world +I spent a few hours trying to figure out how to use Jackson and together in such a way that I could put inline JSON in certain places in genavro +Thanks for the patch Gwen! +Improved explanations for the cluster wide properties +This would have actually been useful for ETL jobs +Also using child tags works +The issue here is was that any that arises from is always blamed on map causing the map to be +Again I am for removing it entirely after running it by the list to see if anyone desires it +Here is my understanding out of the comments Ive seen here including my own observations The code we have for and enable rolling upgrades in the way you guys have been suggesting to people +I have tested it it is a issue +Guess the exception could be changed into a log that said the setValue type does not match the defined type +Nows the time to break back compat if need beIs the right value to use? Is it less useful than +patch to fix existing ee failures due to test related problem +Is anyone working on this? If no I would like to +For the asserts it would be nice to have an associated message for better readability + is used by MR only it should be moved to MR project +Yeah we could go the download route and that is fine +Verified by Egor +I think the check on whether a conf files been modified to reload the core can borrow from the replication handler and check the diff based on the checksum of the files +Any idea how to do this conceptionally +Mark this as an improvement rather than a bug since it doesnt make a documented feature fail but its rather a missing feature +Ive just committed this + +which are normally populated for every request +lgtm for the may be invalid due to timestamps but if we are assuming this method is only used for meta maybe its ok +Thanks for bringing closure +Patch against the current trunk tested to the best of my abilities +The ODE is released I found this problem on this versionWhich version the problem should submit on +Committed revision Thanks Andrzej for reviewing it +if it is unchecked it has a value of false +And I didnt want to mess with the JDOcaching docs where debug seems to be another feature entirely +Attaching new patch for that backports the trunk testcase which caught the problem +committed this to master +Will commit if tests pass +Editorial pass over all release notes prior to publication of +A very high value equivalent to blocking the value is starts a process and returns immediately +If everyone agrees Ill split that out into a thrift patch +Then you can close it +Patch for java and c implementations of locking in zookeeper +There are all together + closing stale resolved issues +looks good +committed fix +Based on comments from this issue is more a function of their implementation and spec ambiguity than a issue +I will replace the fields table with a simple text control +Just some basic clarification on my part +Good Catch Olav +on cookie based approach +Committed revision +thanksRegards the last version of xmpbox the attached document is considered as valid +Heres a compiled version for testing +Ive attached patches that should fix the problem +The failed test is also seen in and should be due to +Will investigate more and later file a bug against EAP +Thanks for fixing the s test +I personally disagree with that assessment but there is a workaround available +If you think that this is not a general problem feel free to close this issue +For me this issue and maybe follow up issues of its kind are evidence I can point to whenever I hear but your code is full of bugs +Update settingPUT dbmetaOptions could be passed as a meta document when creating the db too rather than passing an empty body +Fixed by SVN revision +bq +Theyre particularly relevant if you add a column to a table definition after initial creation +Unfortunately same thing +It just looped forever +bq +I just committed this +The issue with that approach is you dont get a nice report at the end +Lets see if it fixes build issue +Please fix +For jira issues please user Jira project +Ive made this little example which shows the problem +This is not what i had expected +patch fixing the failsThis directory was seriously screwed if you use compound file format +Verified on. +It is good idea Jacob +bulk close for +I committed Michaels patch in revision so I guess we can resolve this as fixed now +After everything is setup browse to httpservernameEnter user jmchughpassword noiseHover the mouse over system maintenance and choose maintain product categoriesChoose one of the categories from the list it should fill in the remaining fieldsHover the mouse over system maintenance and choose maintain product categories againChoose one of the categories from the list it should again fill in the remaining fieldsFor me right now it will not even work on the first try there is already an error +Ive found a better solution which seems to work for all the browsers without errorsIll attach it ASAP +So all issues not directly assigned to an AS release are being closed +Adding a setProgress api for pipes applications +Whether or not they also contribute to ASF under different licenses are irrelevant +is tested much more rigorously with +This contains and which are causing thread dumps +for a no arg constructor equivalent to the default formatIf you mean the DEFAULT format here then that achieves nothing as the user would have to override any unwanted settings and the use would have to know which ones to override +We faced the same problem in our cluster +Above patch holds the remainder of the modifications as suggested by Davethough in unified format +fixed now +Capturing tracking code linked with campaign At the time of lead creation Select lead source as campaign Provide the campaign tracking code It will relate lead with campaign +If we change snitch we also need to change the schema for the existing clusterNot the schema per se but the datacenter name +Heres a example and the logs that show the problem. +This means you can run the cassandra startup scripts without copying jars around or running an ant build +Verified with JBDS L +We mainly want to run demo services do you think a memory of GB of more is possible? Thanks is quite fair np +Upgraded to with EAP release notes Component Upgrades are not documented +I do appreciate your attention to and engagement on this issue +Carry on +That is definitely not related to this schema issue at all although they may touch the same files +The author cant write too well and so the title is a bit off it ought to read Doco that explains the protocols required to allow or any to inject values into or retrieve values from another! Ive checked out the Maven code +Just committed revised patch +Also the doc passed to explain from needs to be doc debase and top level reader to for weight +Actually just recently Ive added a default checkbox in the UI so you can control which one is the default I think its just not shown in an easy way in the workspace lists +These can be a good start toward creating tests +I still dont like the double wrapping +Thats the behavior that Id like +Anatoliy could you try out version and see whether that fixes your problem? Thanks + this would make the API simplerI think we should be able to do all work in rewrite? And then throw UOE from createWeight +Thank you for a very good question Anil +Patch from Bob +If patch looks good please apply this patch in location and add the newly added file in to svn tooRegards it looks like your patch addsmovedelete some files +I recently enabled it specifically for one test in but we need to do the same for some of the ql tests +There have been quite a few changes in the area of content negotiation with the introduction of the produces and consumes conditions in Spring as well as a new for processing requests with annotated controllers +Just a heads up if this is the sole reason for thisendeavor +Sorry. +Add dependency to UIMA Eric is there some simple test we are missing in the UIMA module that would fail +This has been fixed +It also only sees just the HEAD request +Kindly get back to me ASAPThanks in Advance. +Ouch! I must have really disengaged my brain when I made those might have to roll back that whole change +marking patch available so that qa can run against it +Confirm the issue is fixed +I took a quick look +Pushed to master +Couldnt see the patch +Not heard anything about this being broken in knowledge agent. +Opps I used +The latest patch also removed testing of but I believe the current test isnt really the right way neither is it sufficient +distDebug laxis lpthread +Warning I dont writejava +to trunk and. +Here is an updated patch to fix the test failure +Arnoud We are still using zookeeper and never tested with zookeeper trunk branch +to the target directory so that it is deleted by mvn cleanThank youRajini +Some links to the jdiff output should be added somewhere in the docs +Will Not Fix +Resolved on Apr +get function can be found on codebase This issue seems only exists on and before. +My preference would be to have a property file configuration similar to the proxy that would allow a user to fire up the mini cluster quickly with deterministic parameters +Roland using a flag to shortcut a double call is probably impossible as the method depends a lot on the state which is likely to change between calls +Maybe we can use the appropriate http headers for this purpose to figure out if it has changed +Can you try itI ported the testcase back to and got the above exception + +Hi ColmThe nameFormat property is token specific and should be in instead of the token agnostic Claim classOliNote that and retain the old behaviour for backwards compatibility +With Stefans patch the problems were provide my WSDL and mapping here because they are a bit shorter than Stefans +Another option is to simply always lowercase the hostname in the master just prior to using it to determine RS identity +Agreed with Harsh J it looks like the upgrade script was not run +Edwinyou are in the right space +Ive attached a quick fix that implements exactly what I described +Tweaked description and summary since were adding an mxbean method not a metric +Assert statement to detect dashes in constructor needed to be moved as other artifacts could not be created like in the +Just a short update to say Im alive but still crunching numbers on my cluster +Similar error is seen with oraclei +Issue reopened to change Release Notes field. +Can be solved by binding file or placing in classpath +Ill startEnsure hadoop zookeeper and java are pathed correctlyJump through the various levels of symlinks which may be involved with getting to the fileHandle being executed from a symlink in another directoryhandle being executed from a script in another directory +You are right Doug +Were using OGNL version reasonable to me +Why dont you please very carefully check my posted solution and actually try the new code which i say fixes this issue +Its ok now please test +This is a nice idea +glad to hear you have it working +That patch file should be assigned to this jira issue +I was reporting that its referring to instead of or +I have emailed you my guvnor me know if you need anything I got it +No reason to have bugs open for the same thing +After some offline discussion I think it makes sense to split this issue into two and do just the changes in patch +Committed to and trunk thanks Ashutosh! +As stated above with original patch unit tests are not complete and I plan on finishing in a follow up jira +car into the bin +Do want the urx behavior added too? If so would it be ok to be done on a separate jira +Let me know if it needs revising +If you do an mvn site then youll get the style conformance reports +Yes +I dont know if well need it but it seems like a typicaly thing youd want to do +My bad thanks Jimmy +Im still checking on an interim iFix +Burr Abridged tools is the first element shown and the rest is for those doing individual installs +d jdbcderbydblooktestusertestpasswordtestit ran three commands java cp +Patch that does what I described +Applied patch to resolve issue +However its compiled against +Thanks Michi! +bq +Of course async support is not relevant for components that support onlyFrom my perspective sync vs +Most likely the server will be written ok and the client will be the blame +Anyway as I said I will have a look +Deferring it to +Thanks for clarification! +Thanks Rohini for the review! +Im also wondering though if you can accomplish what youre trying to do using the API? +Moving out to +rename as and as +I havent looked at the code lately but I recall a commit that turned off some javamail logs which has effectively hidden this bug +Would you please modify it to sothat we dont mention wrong path of the jar +Since many of the people interested in working on this are in the Oozie project it might make sense to host it there +This is runtime issue is still just verified the scenario on +The performance improvement is IMO sufficient enough with just theAt revision diff which introduces a cache +I tested your code with the attached groovy code and it does not show the issue +This seems like a great way to inject tokens into a process that it shouldnt have +Running patch through hudson +the two commits you post are the sameI assume one of them should be efadeaefdeebebaec which is required because of the new clobber function +Closing as requested. +Closing issue. +We cant go back and patch the code you have on your machine +I confirm that I managed to commit the artifacts this time +Long time ago was fixed on when was introduced +Hi Simon now I am happy +I took the liberty of renaming some of the functions to be more obvious what they do +This works great for Hive itself but presents integration issues if users need a different dependency for some reason +maybe thrift async should have release notes to that effecteric +Its a new file +we have a patch that seems to work so we should dfinitely try to get this into the next release +Hey Jeff do you plan on working on this ticket any time soon? Id be happy to help debug +but this looks different +Right +test cases around alternativeproducerdisposes. +Why not just set the operation itself? I assumed there was a complicated reason and left it alone +In that case it seems that the pending record cannot be removed until timeout +It would have been goodbetter if some pointers to the issue would have been given even if they will not resolve the issue +Its hard to justify performance optimizations over functionality +have this and have not reproduced any more errors +Sorry for the slow reply Monday was a holiday here and I was away on Tuesday tooI realise that doesnt use I only mentioned that for comparison with the workflow that we useIm still not sure how the user creates the template in the first place +Try to do as much as u can +This bug has been introduced by the fix to to issue fixed in released version. +Reassign to a release that you feel comfortable resolving this issue in or leave as is +here tooWould using digit precision all the time be better +I confirmed that works. +Here is the patch for the same +But ideally we could configure a fetching job so that +Per comment from SA Uday Aarkoti If I mess with the project even a bit like doing a rebuild before closing those steps do not seem to fix the problem +Loading all of the modules individually when they change +Ill include relative paths to where these files need to be added +Ive traced through and cant see anything it still happen when you dont have the default mapping there +Setting this to resolved. +Thanks +Correction it is only +Allright thanks a lot! Xavier +Regression with multi JVM should be resolved before or reverted +Dont really know much about the code but am happy to runbuild test cases and try against different versions of the code +I tried to verify this use case in an eclipse that has no pde just in case but I was unable to even get a wtp download due to missing some apache stuff which it seems the wtp update site doesnt provide +Thanks +Would you accept some logic that checks the feature tree for files at startup and opens everything found in there +Cant recall much what I did without looking into it but I thought I would share it anyways if it is of any help. +I think the toString of should be maintained and cannot be changed it is a constraint +This is duplicate of +I will check it with JBDS build +This is the first example of the m not sure I can fully endorse this patch +Please split these into separate items by moving the service group item to a new request under the service groups category +on this patch but patch may have problem +Sigh this is technically masking an underlying bug of a double fwdrev resolve which can return a different hostname and foul the use of CNAMES + For components which do not have tests in Metamer yet they need to be focussed more for manual testing +See +I guess you can do the same with the serializer for your previous patch as well so that the work is really completeThank you very much for the contribution +See also +I run the test in the latest version and found the bug exist. +Defaults are now Xmx m and XXmA default server seems to be using around megabytes of permgen space and megabytes of heap. +Could you please add a test? Thanks +All done. +I explained what each saved view does in one of the emails to franocis jim and luc +Thanks +UPGRADE PASSEDWoo +I am running tests nowThe error arises because at execution time the ALTER TABLE statement cooks up an UPDATE statement to run the generation expression on every row of the table which has just been given a new generated column +Uploading patch again a wrong version of was accidentally included +Thanks Rahul they look good + +Thanks +I just built the project without this dependency it worked +Sorry +The cache of encodings could be with commonly used known values but seemed simpler just to test and cache +Copied code changes Tom Jenkinson made to the jbossasquickstart code base +Suat has applied the patch. +Great thanks +For the checkin I just referenced this ticket. +I hope the upload function works +I made some minor wording changes and committed the Admin Guide patch at subversion revision + regions are not removed from assignmentsI think we do this in +I assume it will also fail on +If we dont have that fallback then a misconfigured or otherwise missing JAAS configuration is easily detected as the repository can throw an exception and simply refuse to start up +These scripts should be useful for the release too so I suggest they get committed after is done +Attached the patch for trunk using the config key +Just resolving. +Commited a fix for Ivar. +Fixed in the trunk +I tested this out on a small cluster and found one small issue +Also this means that neither throught options nor JAVAOPTS I cant set any other JAVA properties what is wrong +Thanks Joel for the patch +I was unsure about the new dependency on Guava but apparently people have already agreed on adding it to so I guess its not an issue +Ill keep trying +If you plan to at least note in the case the issue you encountered +It includes a testcase that shows the bug +Replacing previous diff with one that doesnt include local changes to +Verified artifacts and output with jdk build. +Thanks Zhijie! +Thanks for the patch Sishen +Thanx for the information +Address inane extra white space not added by my patchFix broke test Ted found +Examples +Will doIt might take me a couple of days but Ill submit a draft +For instance line of +the home page? +Hi i updated my trunk and i installed feature correctly +I was in particular interested in handling very large XML instances with BASE or any XML that are backed by file system or database but that was deemed impossible and out of context for XML beansif things changes could you point to documents mailing list discussion or source code to show itthanks does not do soap encoding +Ok I removed concurrency and still the issue was reproducible so please ignore my previous after our IRC chat yesterday we managed to reproduce the production case of and +Changed addPage to importPage but no change to the stack traces + +After a discussion with Nigel Sameer and Mukund it was decided that the Hudson testing process would be fixed to provide a version string with a major and minor release number +Im picturing this issue resulting in a small change to the login code that would call an LDAP authentication method in addition to the current authentication +Yes WISE is probably a good choice now that I know its around and a project +Done in r +Thanks Lohit +Due to a group limit on beaver svn files have to be owned by apache in order for httpssvn commits to workWhich access method do you want to work for commitshttps or svnsshThe repository will still be readable by either method +And yes changes are not needed +I didnt use it where a specific contrast between SE and SE is made because ifwhen the minimum supported JDK becomes SE those topics will need to be rewrittenI found occurrences of JDK as well as JDK and also some occurrences of JDK and earlier +I got disproportionately teed off by the implication that I hadnt read the documentation +I would just help get it setup to run from the DIH module and offer any help around getting DIH configured and running with it +I have reduced the priority because the workaround is sufficient +Ive got the same problem in you are correct +this is passing through our load tests here for a while now +Since LOCKDIR is depreciated with and not used anymore in I plan to commit the soonll see if I can change the code +Thank you Ashutosh! I have committed this to trunk. +Please post your additional comments to the new issue. +seems caused by related issues +Looks like I meant to close this for. +Daniel I stand corrected +now the random test passes always +You can build from trunk or wait for Teiid Alpha +I see that this was merged into the branch as well and is in the final version +If a service has not been installed then all widgets related to this service will not show up at all +Gossip Router and traces for the problem explained above + Causing resolve this as rejected or specify what need to be done to resolve +Looked at this with Suresh +I suppose its safe to close this +Yes +He created the patch removing from but without and without other options I presume so this will not cover the entire +And I still dont feel comfortable fixing in the way youre proposing for simplicity +this fix affects the error messages wrt jira issue +Thanks for the patch Matteo +I believe we should commit this change even if you are not going to work on the python scrips +In trunk this bug was introduced in +Patch you split the patch into multiple patch files? These run across many different classes and fix many different things +That adds needless overhead and clustters the POM with pieces of Ant +Also the parsing part relies on the enum order implicitly +This could be wrong when HDFS is integrated into clients file system namespace +Do we care at all if it doesnt shut down cleanly +bq +And bugzilla +Patch relative to trunk commit formatting in the pull request branch didnt have the right Java style formatting so Ive fixed that in +Juan looks good do you have commit accessOh one odd thing Ive just noticed we have km and miles meters +show role propagation status after create update close for +Ive reverted this for the moment +I believe this is fixed +clashes and Ill treat those errors as fatal srcgen errors if no mapping is used +They are simply binary files +trunk +The asinacos patch fixing the errors identified by Jeff and adding the asin and acos functions has been committed in the subversion repository as of rThe tests for depending on the dfp library have also been added as dfp in now a package in Thanks for this contribution and the various fixes +It seems theres no documentation about configuring queue managers +I wont be offended at all if somebody else wants to take their hand at finishing thisMy thoughts on it were this +I thought the real issue here was how to write clean code to detect if a job has a task to run +And we need to tweak them +Nice! +I have out in r to fix the build +We have already internally tested this operator and it works just fine +Peters reasoning is similar to what weve talked about + Sorry this is an oversight in the release build that did not show up in QA +This should be fixed in the HEAD SVN +commit fafcececcfcfbc on +Currently planned framework and tests for trunk and branch are committed and used in nightly tests ClosingAdditional improvements and additions will be traced by new JIRA issues +The classes have been updated so that they implement the required methods according to the Java spec +The patch with implementation +The fix works as expected +In the common case it shouldnt be an issue since all the classes will presumably be in the same classloader as the class +An other issue has been opened to improve documentation to close as the documentation has been fixed +Im unable to recreate this issue which build of JBDS are you using? Thx +do look at the other documentation I have mixed png and +I think this resolution is not perfect +good catch erwin +I created to track that +verified in just to set version fields +Identified as a Groovy bug +Is there any kind of best practices for porting code between incompatible licenses +I will check on my end tonight unless I find something new I would agree +It is workingI appreciate all your effort and speed in resolving this problemThanksRindress +Committed +Theres still the bug I mentioned on a previous comment concerning a missing in the invoked URL but this should be easy to fix +committed r +Also please note that Knox isnt released yet so this doc and its contents are a work in progress +bad testing all is ok +If you look carefully those lines fix a spelling error +The above list is not comprehensive +this one still needs work +This change went into +In this way the user is aware of what is going on and can make the decision +I think the events are better raised in the Container state machine +Please feel free to change as makes sense for the project I just want to ensure that every PMC has a DOAP checked in so we can start using as a real resource +Thanks Julien. +The default behavior is unchanged +Reworded the issue +Should be fixed now +Fixed when using PUT to update a widget it uses the widget id passed in the PUT resource path rather than create a new widget using a new generated id +Hi DavidI dont really use CC ATM +Matts last comment seems to be related to perhaps an incomplete checkout +AST is OK after Postoptimization but broken after Lowerer so this is a issue +Thank you Jessica for the update +User Error. +Thanks Erik for your comments and suggestions +A final alternative is just setting up a testbed SOAP web server that receives requests and responds accordingly +Good by +Please verify whether the problem is resolved as you expected +documentation of this feature is being stored insrcsiteapt + +In recycle we always delete it if its the wrong size +Do we need to enhance the tests to exercise this scenario +I wasnt expecting it to fail for you +yes thats correct +I think you can commit this now so that it is easier for others to take it for a spin +patch was applied for closing. +Xml files names changed pictures sorted into corresponding folders and named as chaptername +it certainly seems to be built in to most parts of Tapestry already +Yes trivial modifications can go in and comments fall into this category +Closing all resolved tickets from or older +Attached a patch +This is about marshallingunmarshalling object instances +Unless there are specific problems that where not addressed within the original JIRA I see no reason for keeping them in this state when most of them where fixed and resolved almost a year ago. +The change to asm has been made and is in has further API changes whcih are being worked this bug as the issue is now under control. +believe the is source compatible with sJames +I remembered hadoop used ln command to do upgrading too +New patch soon +Ran the same test on my machine and it passed +Bulk close resolved issues +There was a YAML parser issue that was recently fixed +So a sample is still appreciated +And although the NPE does not look very nice it is not any better or worse behavior +Currently it uses default size of I will leave it as it is now +Patch added +I could reproduce the exception +info was added into about Request Errors and Session Expiration Handling was added to the and FAQ +All subsequent issues are probably related to this first issue +Thanks Bill for the patch also Kirk and Eric for testing. +I had a chance to talk to Blake about this issue +If we take a strict definition of applying coverage at the unit then we should apply as much as we can in the SPI module +The Central editor will print the No entries found message if a connection is made and a feed xml file is empty +I consider it best practice to check the message on the exception in a unit test +Hi AllHere is patch attached for UI enhancements +Log file for detailed error message +old date issues. +You can try +As pointed out the patch is considered to be inappropriate. +Logging file opens should be good enough right? How much is transaction logging a bottleneck currently? How much worse would this make it? If files average ten or more blocks and were reading files not much more often than were writing them then the impact might be smallAnother option to consider is making this a separate log thats buffered since its data is not as critical to filesystem function +we cant compute all the way up to infinity this is why i think something like is reasonable if it requires more edits than that go with the old brute force algorithm +Thomas Thanks for helping with this improvment +Intended behavior you are doing a lookup by pk +answering myself when recovery is enabled its the tx completion notification cleans up the recovery information from cluster +No locking filesThanks so much for your prompt feedback and helping us track down the issue +Committed revision I plan to backport the fix to the branch so Ill keep the issue open for now +Analysis of the heap indicates the memory decrease is due to bloom filters and index sampling so I think were safe without a fudge factor especially since the fudge factor predated the +I have opened for this +thanks! +Is the test outside of the scope of hibernate annotations? Isnt it possible to state such a mapping explicitly? Was the mapping itself in errorAs an end user I would expect to be able to explicitly state the fields that should be part of a mapping +Committed revision +verify method refers block but its not synchronized +I believe checked a fix for this bug into CVS yesterday can you please verify that it works +Thomas StefanThanks for your answerers! Obviously I need to give it a lot more thoughtBart +It looks like its timing out waiting for the to shutdown +Big companies are so stupid sometimes I would know Ive worked for several of them +So I suppose that in blocking IO it is not safe to do socket write inside a do not think that socket buffer sizes can help +Adding patch for Hadoop Thanks KonstThanks for the review I updated your comments ThanksMayank +Everything seems to be upstream. +The test now runs cleanly both standalone and as part of the upgrade suite. +But it does not if you build in a stand alone eclipseWhich it did before the patch and that is why Im saying it imposes more depeneciesI think it should notHere is a practical advice that solved the problem for meInclude Ccygwinbin and CProgram FilesSubversionbin into the Pathcygwinbin contains and Subversionbin has which are nowrequired to build hadoop +Being able to index the forms so that they didnt use the same name has come upbefore +Thank you Tao. +When the message is decompressed the compressed flag should be set to false before sending the message to the belated consumer +While an considerable amount of work has been done on this already there is not enough time left to properly test the current state before the releaseReschedule to when it will be properly resolved +thanks +Verified on +I dont think an option that allows opting in to incorrect results is a good idea +I think this would weed out the pom a little +Here are a few comments +Fixed by +Accounts have been created. +Its odd that this works in tomcat bug not in jboss especially since tomcats being used by both +does NOT support that anymoreFell free to work on this component to require Hibernate Spring +using accuracy using +trunk disable Nagles by default +Just adding more comments to your reshape function can be icing on the cakeCan you now submit your code to the repositoryKyo +for the test failure fix +Quickstart showing the problem +Mayve I am blind but I cannot find a URL for an XML document that happens to be a valid XML document that I could use for unmarshalling and that validates successfully against the given set of schemas +For now I would wait before we do big changes to this. +The default we could argue about my preference would be for it to be on +a war without a is definitly not a suggest you add an empty to your smaller war and then configure the overlaying to keep the relevant +Verified on the Latest Build +Ruchith will fix this +No more exceptions +Problem fixed +Please track the status of the original report +However this doesnt work either +The version code has this resolved +field is necessary only when youre using a +is kept as long as the Thread is alive. +After patch over Ran Sort twice and did not observe this issue + This bug has been marked as a duplicate of +The attached patch ensures that when an operand is to string datatype the result type will take the collation of the current schema +People could then restrict it to just one way if they find it appropriate +ThanksCatalin +New patch blocks any new connections being made after close is called along with the change in ordering of the executor shutdown from the previous patches +Is there anymore to the stack trace +Hi GustavoWe prefer to leave issues resolved on a released version to stay closed +Patch looks good +Attaching new patch which corrects the issue +batch transition to closed remaining resolved bugs +I commit the chagnes based on the patch from viola to trunk at revision +What kind of issues would this problem causeSince the arrays were not resized before trying to assign new values if you were trying to deserialize a BL with multiple hosts or topologyPaths you would get an array index out of bounds exception which is a runtime exception and wouldnt have been caught +See +I checked the log files and there was nothing of that sort in there +The only change I would say is to remove this line +When the client reads a file and fails because namenode is in safemode throwing an exception explicitly makes the cause obvious and also gives the client option to handle it differently with real missing blocksThanksBrandon +blocker for +I have committed this +Thanks for getting it fixedTammy +The attached patch put the cleaning of two invalidate sets in one place when a datanode is removed +This is of course wrongThis fix corrects the problem by checking for null +Thanks for the bug report however is not currently supported +Since I am using Spring I thought that it was the Spring integration stuff that was failing +This is intentional +Heres a test program which illustrates the there a temporary workaround +Changed Target Version to upon release of +On Ubuntu and AIX however several tests take more than minutes +Action Collapse will remove that that is it removes ability to browse the content of tld file +See joinCluster +Please confine the discussion to the users list and raise a JIRA if there is a issue with the Solr code that would require a code change +This issue should be fixed in Sonar +back Thanks +This failed if a generation clause had a forward reference to another generated column declared later in the tables signature +Oups I forgot to select licensed for inclusion in ASF works +to +Part of a bulk update all resolutions changed to done please review history to original resolution type +Attaching patch for trunkThis is the same fix that went into in +Also if you look at a lot of the current documentation it is under yarn even things that should be moved out like HDFS Federation and also things that arguably should either be for the entire project so split up into sub pages like +Actually only files are not conflicted +Slava resolved this for the M build +OK I agree! It looks like in its possible to achieve these same capabilities +is not included inbut inso dependency to should be required forEven if is rewritten with Hectorsit still requires at runtimeAs long as using in test I think it is the only way to include in dependency +Seen through direct DELETE statements as wellDELETE FROM +bq +No the older JBDS had the similar issue but Denis rolled back the changes that caused the issue +Just some minor polishings +looks fine thank you Tim +It works both with Softerra UI and the command line examples in the issue descriptionI modified the original solution idea Emmanuel suggested a bit +I committed the proposed change with revision Cheers also think the syntax without space is the good one +I really need to put this off until resteasy +Have reproduced this with a newer jvm? If not please close +The soa zip contains not +Attaching some simple integration tests for properties using the itest plugin +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +Since a patch for is provided here I let this issue opened with a clear title to indicate it targets which is not an Apache version. +I checked in the fix +When I added the required libraries it started working +Patch attached +But I am little confused as to how I should proceed +Initial patch for lazy binary columnar serde +Im able to handle a relative path redirection when I disable strict modeCan you reproduce this with the latest nightly build and strict mode disabled +does have it +Using as persistent storetrunk +Thanks John for fixing this issue +Maven Site plugin is used in WFK project to get a correct Maven dependency tree for mvn q X site command +Im not seeing an actual leak +I did this for a day and I had no issues but this can occur coincidentally too +i am fine with making it reflect reality +Ive added not all viewChange events +This is somehow caused by the threading of the junit tasks in conjunction with HSQLDB +Cassandra System Logs for node Arya +There might be multiple map join operations in one local work +If present MUST be set to +Please if you still have any problems or see any exceptions in your Error log +This is with the xlC on rlsecompIBM XL CC for AIX bug report to the xlC team +Hi Robbie can you please review this change +Perhaps asking on the user list andor the forums could help +Validatable was part of Wickets public API and not marked as deprecated therefore I think it should stay available +Otherwise get me the steps to reproduce and I will try to look at your specific scenario +I have changed the xdocs for the release which were not XHTML compliant to be XHTML compliant and also fixed the issue of the broken internal links in the maven generated sitePlease review and apply the patch attached +Besides if you change the colors the home sites AND the doc sites will have to be upated +Yes it does +I have jai and installed +However this problem particularly impacts me so Ill keep an eye out for solutions +Patch switching Solr s Jetty dependencies scope from compile to runtime +Let me know if there are any problems +Attachment has been added with description the fo for has been added with description the generated our own test case exhibits the problem we just dont test for it +Perhaps we can add the ability to throttle Directory level IO to this issue +This patch reinstates the old enums and tests that they work as aliases for the new equivalents +bq +and also exclude target directory completely +The patch is also makes other class literals loaded to static fields just not to break the style of entire classWe can open another issue about class literal performance +Patches welcome for anyone who wants to work on it +The test passes for me tooI will add the test anyhow and resolve this issue + causes breakage +And support should make it possible to adjust what gets displayed. +Error message seen in the log file is Caused by Child not found lib for +Marking as resolved +Applied at +Marking these as fixed against rather than against. +In this case to target +Thanks James seen your comment too late +Since its simpler to just cause frequent rolls we may as well do it this way and also solve the problem for BK at the same time IMO +Im going to call this one fixed if it causes problems later because were doing the authentication outside of Wicket then we can deal with it then. +I can make file work but not ganglia even w applied +to the committed to +The initial issue in the imageio module has been already fixed +It is just a documentation issue as in EAP the contains the excludedContexts list +I should probably add a test in the http component also since thats where the actual code change takes place +cleaning up Open Issues list +I did some refactors but I have to prepare a demo app to check your case +Logged In YES useridIm working on this in the context of some other scopedclass loading issues +equals in +This issue has been fixed and released as part of release +This addresses for Map tasks and for HDFS in a uniform way +He doesnt go over to the other side +Another option is maybe to make encoding a property under +Likewise maybe the title produce entity content could appear as subtitle or subsection? From a users standpoint the main task is POST +Updated the dev guide topic per this issue +I seem to remember it addresses some limitation in using as a source folder as it used to be +In response to this I guess that its easy to simulate MPUTMGETfunctionality at the API level or via the Ant task +Yes I can write the docs for the wiki +and revision respectively +scustomercustom + for the patch +Unchecking Regession checkbox as this is not a confirmed product this against on February on trunk +At the time we solved the problem by doing some XSLT optimization and switching to using the XSLTC compiler +Committed thanks +This is just like mutable able to be mutedchanged +defer all issues to +Verified on. +Cannot replicate that anymore +verified ok Tools +Whatever the cause the unexpected disappearance of any part of the filesystem is something that datanodes may encounter and should be reported more meaningfully than with an the shouldRun false assignment up to earlier on in the server shutdown process appears to make this message go away altogether +Hardcoded property names removed and default values filled in and some more tidying of code +Defer to defer open issues to defer to not essential to address for +Patch for current work on seen with Guillaume the interesting code is inside the classIt retrieves the WSDL description from the target endpoint decorate it to make it a real SOAP WSDL then creating a Binding from itIts the behavior of the old HTTP endpointsFor the new HTTP endpoints the SOAP Marshaler could do the sameIm going to patch the SOAP Marshaler and test the behavior +I added the dependencies mentioned but I made some changesI for opensaml and santuario I made all three of these dependencies as optionaltrue and scoperuntime +It simplifies the start logic a good bit and the test failures I was seeing are there regardless seems theres some random failures related to the races in temp dests and consumers on the bridges +restart felix with no update done for the new jar +Can definitely create a story for is commited. +So I think this is not the culprit +If not do you mind I taking over this JIRA? There is a task that was blocked by this JIRA +Could you please add some unit tests +Could you please file a jiraspoiled by always building g +For the recordI have reviewed the submitted patch and it looks goodThanksAbhi +Some conveniences for experimentingalias closecurl httplocalhostoderestcounterclosecounterNamefooalias getcurl httplocalhostoderestcountergetcounterNamefooalias getcurl httplocalhostoderestcountergetcounterNamefooalias initializecurl httplocalhostoderestcounterinitializecounterNamefooalias undeploycurl httplocalhostoderestundeploypackageNamecounter +This branch is targeted at performance improvements and as such the compare view includes changes that will eventually be committed against +Therefore all operations have to executed in one session +Thanks for adding the test Enis. +Would be really great if you implement that +Attached is a patch which adds a and profile to the project +Implemented in branch +Thanks Prasad! +Other hostings might return null for +Since it is a requirement to have jdk on user path the default javah provider implementationdoes not invoke javah executable full path any more +Seems like we are on the same page Excellent +Commited +I successfully ran all unit tests and a tests +It was written by me with Facebook JS SDK as reference +It seems to me like there is very little benefit to offering this feature +So I dont really know if that particular line ever becomes a problem when block cache is enabled +Ive restored backup on test cluster so hostnames must change +This is a part of the storage implementation +A solution for this one was committed on +Committed the patch to both and trunk +It should show up on pig in an hour or so. +Patch for applied +Elliotte can you please elaborate on what the issues are +In fact upon further discussion with Joydeep and Dhruba we may drop deficits altogether once we add preemption and just use a similar concept for guaranteed shares to make sure pools get their min share in order of how long theyve been waiting for it +patch to classpath to somewhat avoid long file name problems +The biggest piece left is signing the artifacts which appears to be a requirement for deploying to the staging repository +If we dont have any remaining attribute the server does nothingA very special corner case IMO. +I would personally close this issue with Wont Fix but I think thats up to you anyway Let me know what you decide +There is a problem in document +I confirm this issue has been fixed for +If we can test the specific aspect that caused then we could consider writing a test for only that specific aspect +Max please reopen it if you still have any questions +It worked fine +Having said that I also dont like the info level to much but I understand the requirement and also agree that there should be a possibility to configure that in one or the other way +Comitted the doc patch to the docs trunk as subversion revision Once the updated doc appears on the web site Ill close this issue +Good find +That was the other thing that was bothering me! There was no stack trace in the junit report just Ill take another look and see if I can get some more information +fixed +I think Eli has a point and im on Alessandros solution. +If you have a patch for this issue please attach it and it. +Here is the patch +Aleksei if I apply the last patch in case of scenario failure I get notification that is MbIs it possible to shrink the notification size +And one should be able to implement enumerating snapshots and mounting them from libhdfs +site is svnpubsub and will need to remain as such so skipping for mirror req +Should have been rejected. +Lewis could you unset fix version ? Thanks +Looking at it clearly creates a new Configuration object if null is passed in to the constructor and does exactly this +Zebra needs column names and cannot work with positionsThat is then the limitation of Zebra which it should overcome someone point in time +Yeah this is bad +Verified changes +Thanks Vikram +Works fine for me happy to close +Sorry reopened a wrong issue as i found the error messages similar to the template creations issue for that is which is currently not reproduced in the latest build +Thanks for the patch Takuya +I suspect that errors are reported during processing resource change events in our code +If any other issue is found that is outside the test cases provided with JIRA please provide a separate issue for that +But will be glad to change with more appropriate name +might be trickier than it seemsIm closing this issue but Ill call for a review of all serializable classes before we go betaPreferably with test cases that can detect incompatible changes that try to creep incheers Roland +I set the fix versions and if Ioannis has no time I will tackle it +Resolved seeCould you please? Links in that issue as well +Inside the file you could mention JIRA number in the commentRaghu +Not sure why this bug is a blocker if it is not a regression +Dali is not supported +Layer B streaming for clob and traditional way for blobI think rd one is best one and work forward it +I think this is implemented but and more testing is needed +Any chance someone would port this to Android +Check for the condition in the finally block when releasing the rowLock +Updated to now output type of the attribute instead of value and target master and. +Can it also tell us which resources the conf has actually been loaded from if that informations recoverable +The function Graphicsnew tries to create a new Graphics object +So Ive put a test in that skips the uri calculation if and are the same object because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +To make it pluggable is to encourage people to write new remote and I dont even want for any longer than is necessary to make a transition +Maybe wait for this version that should be the last of the series +Committed revision +Rob Please verify that things work fine for you too +pls ignore patch i just uploaded here attached it to the wrongbug +Initially I named it but it wasnt clear that it should be used in the context of so was suggested +Deferring to +Would it be possible for someone to upgrade the trunk to since it actually is backward compatible +While I appreciate the effort this is a setup for disaster as Maven has no way to differentiate from +changing status to patch available for trunk +Some outstanding dependency errors exist with the showcase creating a new issue to address this +Whats the error? does seem to get the code fragment above into the generated code +Upgraded the used by plugin +Awesome that the patch was accepted +Unlike getListing and list this patch simply returns all the current snapshottable directories considering the number of snapshottable directories may be not very large and currently all the snapshottable directories are stored in an unsorted list in the thus it is hard to continue the listing from the middle +jaekyung can you try and see if this is still an issue +until we have a better set of to justify this feature. +Wow I hate conversations in Jira +Committed +So Iam afraid there is a bug in the tooling +Any idea when is this going to be fixedThanks +It looks like the performance is indeed in the metadata path since a read lock is held on the Namespace while the reverse lookup is performedI think a config param that turns off this check still makes sense +So please forgive any mistakes on my part +Please take a look if you get a chance +Project also include corresponding files and +Committed to trunk +From my point of view the issue that the present ticket is about is resolved with the patch from Koji +Seems about right to me +About the flipping observed on trunk it may be related to which was fixed on trunk only +try SIZE works. +May be related to +It speeds up after solving +It is exactly thatIT is not possible to define a ressource as encoding and another as encodingSo when filter ressource is activated the file created will loose the original encondingTwo solutionsAuto detect file encodingPossibility to set encoding for a list of is exactly what I want thants a lot +Should I ignore themYes if they are indeed false positives +This fails for me with Maven and Maven on Windows XPJava as well +EJB FR The Optional InterfaceThe bean class is not required to implement the interface defines two methods set and set method is called by the beans container to associate a bean instance with its context maintained by the container +By allowing client side caching of css and js resources the number of such requests has drammatically been reduced so I assume this issue is obsolete now +Would it work if the given permission does not have x for example +Well look at this during the timeline +Patch to rename to +Why are you printing empty string as error in? Command description name node to namenode In the script you are setting errorFlag before for loop +If you enable trace logging youll still see some messages but at least they wont include the stack trace +Also please add Release Notes to describe why this is incompatible. +This behavior still occurs with the attached +Looks that Linux does not observe the backlog parameter +Then later includes explicitly +Nice idea though to have JANINO in an Extension +Robert +The changes to deployjms was related to related parallel final config file didnt change in this s fine to be removed from thanks for checking +Reopening for branch backport +The must hold a reference to the first and last Attribute but only linkage from first to lastAs Attributes can only added to one and not to multiple ones I see no problem with it +Not done yet +This is easily identified with some basic troubleshooting but thought Id try and save someone the minutes +Fixed in r +I looked extensively at the code already and did not see a way to force it to use multipart post in fact I created a Solr ticket for this reason +could probably use some love and flexibility +This has already been fixed in nightly build +allow containers to override HADOOPHOME etc +Bump +HiI mailed root on Monday +Committed to just pulled this into too +It causes an unnecessary burden on the server +The name should be whatever it was when the URI was submitted +Seems to work for us +The javadoc plugin is skipped because the aggregate project has a packaging of pom + the rest is testing +Hmm are you referring to the should stuff? We are passing session through there with just the local variable in the should implementation named session +The configuration patch will solve some basic problems here please vote it for better hadoop local mode support in the future +On a simple setup META was on node the only user region was on node and showed that it was on node. +Commit applied verified and committed +Ive seen a few exceptions to that in practice such as for trivial documentation fixes or for from other branches +I think a simple comment with the attachment is fineBy not clicking the ASF licence flag you stop others from working on the patch thus removing the chance for community development +resolving as fixed +This involved deleting the class which is not the right way to do things +I think moving much of the logic from binhadoop into Java is a good idea +Please bugs back if you intend to work on them in the very near future. +Uploading patch that fixes this issue +Should httpClient send the header? +please help me out +A likely scenario today is a site with wildcard dns +anyways if you think its too much a pita to build then just close this as wont t fix as synchronous ajax requests are pure evil +Sets a default threshold just after like it needs more research to me +In any case future workNo it implied too many changes and the guy who did it didnt push very hard either +See the discussion in where it was decided to remove the distribution pluggability in +And you can recognize that only if you have actually tests that use the missing class indirectly +Checked in test casesvn commit m Add getMetadata test srcconf srcjavaorgapachejdotckapipersistencemanagerfactorymetadataSending srcconfAdding srcjavaorgapachejdotckapipersistencemanagerfactorymetadataAdding srcjavaorgapachejdotckapipersistencemanagerfactorymetadata +You probably didnt understand this bug can be fixed by swapping source with arg on that line +This fix was done in geotools +If you can think of a better way to do it please reopen. +May be we can support this in the AxisC +yep both +Patch committed. +I created for the other problems I now see with groovy files in eclipse +Please verify if this is still there if it is please reopenwith Fix Version verify in ER +A new patch is uploaded +Completed from the perspective of it generates legal code but I would like to revisit the import management +closed +Plus the used by the test itself would need to be Im not really clear what your goal is here +I did a test using Tapestry beta and OGNL to verify the bugIn Tapestry current version the expression for a component objectEnabled is OK but objectEnabled not +We should also add the usual as a suffice to the username prefix +Download composite site compositeContent and +But yes it is pretty solid at the momentAndrew anything to add +to Yura it is necessary to test a functionality of the css dialog completely +Otherwise good luck forThanks thats what you would say but I had to try I understand everyone needs to pay the bills +Could you elaborate on the buggy scenario +I would think that most folks are going to check to see what specification version is running not what specific implementation version +Resolved in branch +How about this one? +I actually made those up earlier as well so here they areUnfortunately the numbering on JIRA seems to be added alphabetically after each attachment is added so the above numbers are gonna be wrong nowAidans will be Steves will be and his will be +Both were rejected +bq +to be more specific problem is with the last true in the looks fixed in +We need to evaluate if this is still and issue +An overflow in does result in a valid number result just not the expected one +But its enough to get started +We need to rework this patch to fix the problem before applying itBasically the doesnt generate the idltype attribute for the corbaparam element +Note that I am not in love with the method name at all! If we can find a better name then that would be great +old news +Cant you help with themAntonyCould you please refer to your underlying logging implementation manual whether it logj or logback +On closer look realized the previous patch might mitigate but would not solve the problem +Assuming comes back clean looks good +Documentation updated +Attachment has been added with description PCL Image showing text being P open bugs to P pending further review +Please ignore that patch for now +Creator Javier Mozo mail to add that these actions are usually taken by the creator of the work. +I cant see anything like the pull request code in the current master branch +This patch ensures that all job properties are saved in the json file under jobProperties tag +It shouldnt be difficult to pull up those methods into theyre on Object already and just need to be rebound in mode as being on +controller automatic name +Patch applied +You also need to lookup the components youre using thats why youre getting the NPE when running the tests +Could you please take a look if I made some mistake or forgot somethingThanks in advance MarcoEverything looks pretty good but I went through and made a few changes in rev +Remove interpolation during resource copy +clarify impact of using linkersinclude further similar types of says the patch for fixes this issue +Is there another Jira for this? Ill probably have a patch ready by the weekend +With this issue Ill dig this hanging problem +It does +This is the same as the fix by Andrea adding a xalan test dependency +looks fine to me +war and registers +I cant reproduce +And a box as a visual has a benefit over a toolbox as its more generic + should not create more than the number of connections set as max on the connection pool did you get that error message +WINNTTEMPTMPDocuments and Settings should either seemlessly support both forward and back slashes or use one consistently. +Unassigning myself since if its a JDK bug I doubt Ill have any time to think about a workaround +I backported the changes in and DFSNAMENODEEDITSTOLERATIONLENGTHDEFAULT +Thanks Sandy! +In which case you will need to open a WTP bug as they should be prevent the restart action from being enabled if a publish job is running +This means that is needed at compile time but not at runtimeunless you wish to use the preferences framework +Someone added it for some reason +do removeconst Coberturare right I can use moduleeval removeconst +If you wanna do a PR to that end by all means +Thanks to Jeanne Waldman for the patch. +The patch contains +Behavior might also be different depending on if you deploy in exploded form which is an expectation for Roller +static? This is fixed in and +Bulk move issues to and +As far as Im concerned its a Im not going to treat it any differently than I would any other persistence exception and where there is no behavioral difference IMO there is no need for a distinction +Main change is upping retries from to and a change of theRETRYBACKUP array so we ramp up quickly to retrying every ten secondsM srcmainjavaorgapachehadoophbaseclient Print out elapsed time over all retries +When I have a fix to prove it Ill make them duplicates +There are a bunch of tests that extend to do that sort of thing +Editing an obscure properties file does not count as a simple toolI would urge the maintainers of geotools to provide an interface with the EPSG database so that users can download the latest version and straightforwardly update their projection interest I have attached an updated version of the file to this I am not volunteering to maintian this file +Adding some more information regarding impact of this on size +but +These blockers and critical issues are resolved but not verified +I just committed this to trunk and +From the description above I missed this +For HIR users its convinient to have and operate one new instruction than a new operand kind +See the attached screenshot the help content is included but the top level summary is missing +Indeed this was inconsistent and with commit dacfdddecabffbfddd list views do not contain columns for relationships any more +Could you be a bit more specific please +Changes look good and have been merged to +but perhaps it just needs one or both of the previous patches Looking forward to getting them all in apply issues aside they look great +I just committed this +The content page uses the font +It might help to reframe the problem since the ticket proposes a solution instead of framing a problemProblem As an what I would like to see is a uniform interface for interacting with and managing Cassandra and a CLI was a natural place to put that +This patch addresses most of Raghus comments +Resolutions other than done should not have a fix version set. +Hi FrancoisDid your show any problems? Do you think the fix will make it into +Georg Are you the code from or do you have your own configurator? It is a littlebit complicated to set up the appropriate facets and validation settings for a dynamic web project +Committed to trunk. +Trivial patch to fix issue +Not reproducible under +Committed to +I meant to write The patch mostly looks good but lost patch along the way +RandallIf theres no technical reason why we need to bump the seq for changes that dont modify a document we could just stop doing that +We wont have anything to parse feeds if we remove now +Like you say the first thing I noticed was that he is using byte arrays and I think would be better suited for this +Raising priority as this task is about making the correct design choices +It does make more sense to add the umask test in +Queue might accumulate many regions to flush just by way of normal operation +I would prefer someone with a very special need that already uses complex JNLP file spec come to me with a real life use case +Use optimist for arg parsing +IIUC the only argument against it is that it is possible to get the same effect by putting the classes into a separate project +testcase with default server configuration issue with workaround true applies issue doesnt occur correct jar location returnedProvider code base filetmp believe the issue only occurs with SunOracle JDK though and not right? I think that does not verify the signature on jces +Thanks for the response +tableinfo file with the correct table name column family and default properties + +Taking the status bar timestamp info out of the game +removed the generated html from addressed +It seems the patch applies when I run it on the command line using the patch tool in Eclipse it seems to behave strange +See that spec for more information. +Closing issues. +Committed revision . +For stateful restarts of RMSorry I didnt understand what that means. +The refresh method is based on the reconfigurable mechanism by folksDo you think it should continue use the startup option when add new disk? or we should only simply load or format itMoreover is it a common case for you to load blocks from a new disk or may I format all new inserted disks +I tried to find something to read in lot of places +So implement the interface +Hmm +Updated patch that applies cleanly +Closing issues before next release. +See if patch v makes the test better +If the default view of the gadget you are rendering is of type URL than the metadata response you get back will have an IFR URL for the URL view +Thanks Andrei every memory improvement helps +I added some links also note that this issue duplicates think this is the most serious bug and we should fix it in groovy first program prints now and the second I think that is what should be expected +I see this problem in Grails and Grails too. +Its been a while what is the summary for this issueI see it is flagged as wontfix +Few comments srcmainpuppetmodulesmanifests what will happen if someone configures the hive conf dir to etchiveServer changes LGTM +The issue seems to be related to the use of cygwin. +If not well push it into +Todd here is the patch I have so far +Ill implement and add a few tests +I have Super scoring as I do not know how to do it correctly +There are no new features since that release +Hey Todd I think you again forgot to include the file in the patch +adds long to the NLTAGS tests for your assumption was spot on definitely a bug the way was ignore longthank you so much for the testCommitted revision +We almost finished with vpe design options dialog see think its better to leave on denis maliarevich and move to next version +Hmm maybe the problem is the interface? Instead of perhaps we should have etc? +im also wondering about the heardFrom +Thanks. +In bafae. +Its all as it should be +This feature seems to work only when the parent of project configured references resourcesArtifact artifact as provided sample if I remove the modules section of root then run mvn clean javadocjavadoc under test folder files are not correctly this the intended behavior? +It still surprises people like me that filter their incoming mail traffic into different folders +Jukka That page is now a Do you have the info +remove great Colin +Ive committed this +Fix for in rev +Works fine when tried from API by replacing the sourcezoneid with proper zone id in the API +We really should get this fixed +Fixed in revisions bfeaaaaccefefecd and baeadfacacbdacddaThey should be rolled in to into branch a couple hours ago. +Just the mapred tests to do +Eventually it will be used to create choreographies instead of using +David AaronThanks for your attention on this +There are no actual imports only references to classes in. +My vision for this is a script which will pull down the dependencies you need for a given container and stick them in a folder somewhere +Not +The patch fixes the problem of currentHost being null and also handles the case of update being called twice without migration completing in between the calls +Hi BurkardI cant this issue can you explain further how you produced it and the output you saw that showed it running in production fixed when moveing my grailsapp from to +I have built with the detour verified it fixes the problem in our application +Thank you Eli +adding resolution fixed version +Lets see whether we get some sources as well +Updating it with more fixes as in not been picked up in days +Yes good point no need to keep it on disk +Given that you made dramatic improvements in speed by moving less data to the reducer this should have very significant effect +Had a discussion with Rajesh this afternoon and we decided that the best way to continue progress here is to create an experimental branch apply the fixes identified thus far and run the experimental branch under that end I have created a branch JBPAPPCPIPV and will be setting up a Hudson job with the same name to run it under the group of Hudson further work will be applied to that branch +I suppose this is irrelevant now since were using bundlor instead of bnd +Fixed under r in trunk and r in the branch +how are you opening multiple CRAR for the same sstable in a single request? Secondary index reads +Thanks Jitendra! +Deferring soap service work beyond xml and console issues. +This turns out to be a bigger issue with body tag handling +Patch for trunk +Kailas you will have to provide more information for me to help +This PR is just a workaround that partially fixes the issue +Added support to extract information from a compute node description +Move pom configuration from project to root I think that makes sense if the project POM is simple and the amount of boilerplate you get for having a separate POM is significant which was certainly the case for. + to trunk. +Attaching and the patch for changes in the under javatesting directory +I dont run into problems with the new embedder +Message of issues is abbreviated to characters if its too long +Maybe a problem in Jira lost itCheers no sorry PaulI guess because Sascha switched the status to resolved and I reopened my bad for this wrong advice +After this fix each block update will have to wait for an IO to hit disk at commit time +Does logj buy us anything else? Im definitely a fan of logbacks of config. +If we understand how it can be reproduced we will reopen issue with video attachment. +I am pushing this in +committed to +Modules depending on the new interceptor module wont build +fixed thanks +I also believe youll find that the other places in the ejb containerthat are interested in logging the ejb dump the ejb context to the log rather than the ejb instance. +From a first scan nothing of the kind is available yetI briefly scanned for header and cookie serialization to JCR but didnt see anything. +Have you seen this outside of EC? Only in multinode setup +Is the test failure related +The patch which enables to load resources from the classpath +Only slight adjustment would be required to adapt the subclasses +I adjusted the unit tests to trigger it and have made a fixThe patch includes only the new fix +I will personnaly wait the jQuery branch merge +Sooner or later indexing hits io limit therefore there is no profit to utilize CPUs by huge amount of indexing threads +Quickstart showing decryption exceptions +Thanks for submitting it +The MIT license for is included in the license file since we distribute this in the binary tarball +I need to look into what the new tar assembly in module is but I think we could probably just calculate the version somewhere common and use it in hdfscommonmapred + Username juanmartinez +Second this breaks in AJAX context +So theres no reason not to do it right and add a use this snapshot parameter to the +Within the generated package there would be subpackages +six sample projects for uniformity across examples +I see you updated one of the source license headers if you wanna point me to the official one Ill supply a patch for the rest +Im quite confident that the test failures are unrelated +Pete I noticed that you had removed the dependency in the pom but I added back temporarily to resolve the build failure +Removed throws clause +Thanks good +So a separate issue is filed on that problem +Is installing bash on difficult? If it is and your group will be regular Hadoop users then perhaps we should commit this patch +the problem is almost certainly that theres no java command on JAVAHOME or else it just cant find it +There are previous issues in Jira where issues like saving the request body were discussed +I have the following suggestion Source artifact is named by the projectversion as usual Binary artifacts are named by projectversion with a +IMO until WFGEN is functional complete we should not include it in distro tarball +Smooks are using those now +SVG has been added with description SVG original would like to know if some news regarding this issue occured +seconds to run mvn clean install +Marking the issue resolved as the fix went into the release +testcase using user interceptor was added +Is that writersClosed flag being used? Agree we should not try and close again if first close fails +I was doing tests on my local machine with an ssd and loading is definitely cpu boundOur index has terms +Can you give us some more details? +Also in the code force rewrite the producers property to and the property to syncThank you for the comments its really careful of you and careless of me +I suggest that you review the improvements and we continue the discussion at that issue +The new patch conflicts with +It fixes just one symptom reported +Patch looks good +Right now the Solr output connector tells the world it accepts all mime types but we can readily put that under user control +Maybe this could leverageThe no codebase issue is also worked around with this approach as we would include it but parameterise via the JNLP Servlet and use in theIve tested this manually and successfully replaced codebase title and argument entries with wildcards and achieved secure status +I wouldnt expect it to but we definitely should do some testing to make sure +Its more likely that the cluster output file was ready only on the third attempt +Also one needs to know Kafka code to figure out how to construct the thread name from all this information which is even worseI think this is making things more difficult for us which is a down side to disallowing +This has opened up multiple paths in Metastore code which needs to be reasoned about +Assuming this is a duplicate of +Further investigation showed its a DCE pass that removes creation of the instance Finalizer object as it is not used in the program +Im planning to commit this when I finish testing please reviewtest and let me know of any issues +Should be considered if everything that goes into an editor wouldnt fit better in a view and vice versa +Special logic was inserted to allow those simple typesNOTEAn issue has been raised with the Schema Working Group in hopes the document will be modified so that the types are no longer presented as global components +All tasks have been resolved. +Separately we can decided if ALL interfaces need to be labelled +But In a suggestion for us is making an auto sync repository so we did it in this there any other thing I should doThanks a lot. +Screenshot snippet of Checkstyle report showing severity as Error when it is defined as Warning in the file +thanks +Max this is trivial and safe +Neat how it Hudson noticed it was just a documentation change and didnt require tests +Is it the case that hive is expecting for an empty file a split size of and then the record reader will emit no records +This patch needs some work +I have been out of business for the last four weeks so my answer is very late +I know when we initially moved to the new dispatch system the ows version was used +Updated patch with more features and some debugging +Apply this patch to your folder +Verified in JBDS B +Looks good to me +I agree interwiki links are a special case +Take component security for example we can annotate only the component methods that we want to secure with Restrict +mind to upload that +Thats perfectly correct! default maxAssign and assignMultipletrue so fairScheduler will infinite loop +Shahzada would you mind if we close this issue given the functionality you were looking for already exists? I dont really think the regular expression suggestion would really add much value here and the includeField option is just the opposite to what is offered already +However Not able to build using these details + does your example from above not work with generics +I propose the following In the case where there is a conversion we should throw a warning There can be a flag which sets whether or not we should instead error out +Thanks for the contribution +This strikes me more as a kit instability issue than a hawtio issue can we once we have a more recent build +This example app showcases the problem +yes these codes are for ARM. +It may simply be normal behavior for Windows that links are considered open until their creator is closed +Otherwise the patch looks good +So you actually have to make sure youre not running a postgres instance on your local machine +please note that if you do a in a maven build somewhere the IDE is capable of catching that one even though its not routed through the loggers +Stumbled upon that issue with exports SLD by default so if your style in qgis contains graphicFill for polygonSymbolizer you need to manually convert it to SLD +The response code is not use case is sending a http always and having the actual status code in response read the ticket +Does this only happen in a node ring +This can give us a clear picture on pb rpc overhead right +I presume this was put in for some testing +I agree with Anton +I confirm that the fix resolves the problem of no code being generated. +Okay guys I played with version and it seems to works fine +Thanks Mortiz +No worries Richard Ill be the first to agree that parts of the spec could be clarified +Backend code is completed UI part is left +In your case can you set the property true and it should help +Didnt find any problems with adding JSFSeam capabilities to the project with Servlets JSF either +Patch LGTM +Some more flexibility can be gained with for the different features but I suggest to leave this another issue +Verified in EAP to add release notes +Hmm +Work done long time ago. +I see AS is crashing on something most likely an import statement that cannot be resolved due to the fact that Spring is not bundled in the application +Added code for fix in list of in file can be uploaded and in one go lot of can be attached to product promo +Hi Lucwhat do you think of moving this contribution to the package? I think this would be a better fit thanThomas +A map function that emits no keys coupled with a lot of deleted documents means that the partialupdate frequency will be quite large +Any ideas where to start debugging +You can add a wrapper to your application that modifies the callers or you can use an incrementer as you suggest +It makes sense +bq +Lets revisit this issue for the we are revisiting in lets update the fix version to revisit this some time later +Ready for review. +This has now been has to be updated definitelyOn I am not sure +the decision to switch to Nexus was not made by the Tomcat community on the Tomcat dev listMy assumption is that the Tomcat community will retrospectively approve this change +Changing the port doesnt change the port of the server youre specifiying the port that the server is configured to which is by default +It now works correctly for me +We can check if the hadoop command is available and fall back to the old method with a warning if it is not +Dominik is there an upstream JIRA for this issue +If that isnt done upgrade from +Looking at the +I would assume for it is a bit its indeed a bit dangerous but it should go in yes +II think the listPaths API on a directory used to return the size of the entire directory subtree +Other transformer components such as appear to copy properties by default and provide the option to disable this behavior if desired +Actually I fixed all the sites of thread creation when I rolled in the security manager several months ago +Both current JIRA and seem to be about the same thing +Ooops +Thanks! +Committed to trunk +was added as a public class in December of in + +Attachment has been added with description Test case that shows the exception +The method toString now outputs only a few will improve it following suggestions from +Im not actively working on this +If you want to pick this up feel free +If all the remaining replicas are corrupted Namenode will keep on trying +Doing now +However the missing class is present in the Jetty release can confirm that switching to in jetty fixes the problemPlease consider releasing a update including the mentioned jetty upgrade +I guess that what my suggestion means in practice is that the memtable in the system that was using the largest fraction of its local threshold would be flushed would be applied when a keyspace threshold is exceeded rather than when a system threshold is exceededWhen a server threshold is exceeded you would first look for the keyspace that is using the largest fraction of its threshold then flush the memtable in that keyspace that is using the largest fraction of its local threshold +The reason I did this was to actually prevent concurrent processing by Mail VFS etc +currently returns the absolute offset which is ideal +Good +Could you please review and apply if ok +Yes but in each of those real data corruptions data can be salvaged by switching to a valid instance of the blockAssuming the corruption happened after replication +DWR and nightly Dojo build were used in this patch +Reopen to update the javadoc is updated with Igors comment. +But for this particular use case it is and seemingly wrong handling +New patch applied in revision which uses format version number in +command +Im going to mark this as a duplicate of because the problem is almost certainly the same +Patch to allow PO order entry without requiring a no longer required for PO entry as of rev +My patch continues to do that but Im wondering if it needs an additional test for BinaryThis patch is absolutely what I think we need for trunk +Still this is not enough for two reasons +of string manipulation +this one may make immediate sense to you if not we can do some more investigation +Attach a patch +defer all issues to +I will correct them +Committed to trunk! +That might be what causes the other failures +No it happens in all browsers. +Bulk close of resolved issues for. +i like version b colorful is cool +Added to the SOA release notes as resolved is now supported by jBPM JCA inflow +yeah thats what Im asking +I committed this to the branch as well +Nope +In several files I added an import where the import serves only to blockwarnings by javadoc and is not needed by the code itself +I committed the patch to +m Userswsmoak +Requesting has SSH keysTrying Pageant key Key authentication refusedNo supported authentication methods left to tryNo supported authentications offered +Forgot to tick the patch available box when I uploaded the latest patch +I did testing with the data and the encoding is working now without the +Reproduced in what is the exception for this time +Thanks for taking a look nick and stack. +The tests ran cleanly +This bug has been marked as a duplicate of +FWIW an ANSI SQL compliant TRIM function would be fabulous +Hence the test I added was failing +You dont have to use mvn command line in order to get source folders un Eclipse +I think the latest patch represent a net improvement to Derby so I propose to commit it by the end of the week if I see no objections +KennyThere will be no new features added to the line +Yes Lars I am adding the logic to count the rows deleted as this way +After I have replaced foinline by fowrapper that bug dont occur +When resolving make a fix version as well. +Maybe you should try to attachthe full error message +Thanks Edward I will look into it +This helped us to find few more issues as wellThanksDeepal +With this arrangement when I deploy the exploded war to application works just flawlessly but when I deploy the same to AS CR I get the exception stack trace for which I have pasted in this bug and which seems to be giving the impression that its a bug in my code +Are you using Windows? Havnt gotten it to fail on Linux yetPerhaps we should add a check of the status returned from index in there +I have merged my code for the UI support for AWS Health Check to asfmaster +I am closing this Jira. +I tried something else that maybe changes thisI changed policy at server side so as to not have to encrypt wsaTo and wsa from the client and it worked for the first case +Im working on a fix for the original race condition +The explanation why the tests fail on ibm jdk is that ibm jdk does not support jps will do a pull request for the fix which works for me +I do however like Michals idea so ive left this around as a reminder +this is not a regressiontrue enough +We felt the current flow is probably not the right way to do it +One thing I added the max size parameter to the ctor rather late so there is probably some room to use more of the fast +Id throw IOE instead of to be consisten with the rest of the method You dont need to use since Java optimizes string concats and this code path is rarely run The patch has uses tabs instead of spaces +Please comment here if you have updates so that I reopen this one instead of creating a new issue +requires each capacity be between and + +Bumping the Priority and Urgency +Please accept this as authorization to treat this as an open contribution to Wicket. +C is stopped +Alloc prefetch support for GC of in r +One is a driver for the other five +is now pushed in the codehaus maven repositories +Is this plugin available in any public repository or is this part of maven build helper plugin as Pauls comment saysThanks +Brief comments may have overlap withJuns should be removed +currently that isnt the contract of thead yes but remember that this only happens if the bean gets used on different views which use GET! This is not often the case +ea has to go in MAVENOPTS until surefire can fork the tests and we dont want that enabled on Maven itself +ALSO still includes plugin +Hi Trevor is jsvc definitely accessibleexecutable in the package you built? Did you definitely have some low ports configured for the DN +But it has enough transactions in it and it shows that each time the complete CONNECTPOST sequence is performed +This issue seems to be become critical on Sonar +As a consequence the factory is just logging those exceptions to master already +Corrected typo in issue summary +Ive just committed this +done +I can whip up a quick diff to do this +bulk close of all resolved issues. +Were therefore going to need to try and rejigger the queries in question to use IN clauses which Derby knows how to use and hopefully not break any of the other databases +Patch for existing be in a new token folder +are now used in a direct more usual way +Also the noop timer uses this callSo far Ive most of this code working so Ill produce a patch in order to be reviewed by youLet me know your opinionRegardsManolo +The script shouldnt have to reference and by name for example +We will happily accept a patch patch avoids parsing of authority when it would be inappropriate when URI is a simple path +So those java files use small primes and seed value as to generate the hash value +Agg site job runs when composite site changes +How about we make them configurable and default to metricsjmx? If they happen to use them as table names they can change them to whatever they want +htaccess file is on the machine at wwwdtd i planned to move it to forrestdtd and enhance it +For examplejava httpsxxxxyyyxxxx best regardsF +There might be situations where startingstart are called but stoppingstopped are not. +libs + Bug has been marked as a duplicate of this bug +IMO this is not good for Maven +As version may be some time away is there any workaround +Yeah is definitely more than scheduling +What is the synchronization for them when blocks are deleted +Now this looks like we dosent want to fix this +Closing resolved Release +Its a very easy upgrade +Or do you mean change the path and removing dots from this directoryWell I can try that on Monday +Well that obviously didnt work +However I downloaded the I attached last night and with that jar the test failed +Ignore the description issue was cloned from a previous issue I pointed to Alex as a dtubello to the tynamo developers group in Xircles will cause this to happen automaticallyWhy not just do that +Lots of things in EE may care if weld is present or not so I think it makes sense for it to live there +We need to implement the ability to add a storage pool before any host is added to the cluster +Removed an additional strategy change not directly related to handling of transients the safeEquals check +Moving these all to a Doc release version. +The syntax specified here doesnt work +Yes the write throughput to the base table when there are or more index are there to be measured +It was changed because of Tom Baeyens concerns for more concise XML +Can you please make sure this is the case +Jasperthanks for the hint regarding the validator configuration DTDI just added it for Please check it outRegarding the improvement Id like to hear what other think about thischeers it would be nice but its hardly an issue for me +But that applies to the core AS as well +committed converted test and removed redundant tests xa xaMorph and xa +Are you going to make the change to have wmsscaledenom call the OGC scale method +The files will be updated for +Thanks Prasad! +It simply takes which facets your server supports and makes the default the highest version that the server supports +Its a modified version of which is the sample configuration file of conntrackd which is license +And there will be an extension to the that writes Avro +This is probably another good question to ask on the EMF newsgroup +Or is there something special about layered pdf files? Whats the difference? Would this be a new output format or just an output format option on the existing pdf output +Do we need this in too +Follow the source code and diff files +Anyway Ill add demo later +This does occur in as m having the same problem with a WAR project +This is the patch +we havent heard back from you since we asked for the stack trace a few weeks ago +I installed the bit version of the library should be fixed now. +No response so well assume its okay to close. +My fix was good I believe the packaging process for the snapshot was broken +Would it then make sense to package and publish this separately? Publishing it in Hadoops contrib may hide it from those who could use failure monitoring outside Hadoop but do not know to look for this gem in Hadoops contrib +Any kind of limitation is therefore very specific to your object model and the consequences reaching a limit is even more specific +the latest patch looks good to me +Or do you think there is more work to get it to? I cando the committest running just let me know +The patch can go into both and +yes I realized that maybe handling this by vm is better +The fix adds a configuration property called that defaults to false +It would be very fine if the model itself were a writable +Updated to support include at bottom of page +uploading sample again +Oh sigh +Thank you for the complete test +Can you see if you can come up with a way to refactor the pom information so that we are minimizing repeated information +However the GC maps are naive and expect l to be live in the call to resolved new array +Thank you for committing this. +do you have a test project that fails if you use standard but works with your version? The ones Ive tested worked without making the substitution and I want to check the shade plugin is working as expected +Thanks Mike +Thanks. +If I can get to it tonight at least the one error message +s the status of this? are query results being properly sorted based on distance + This bug has been marked as a duplicate of +The attached test and fix work well. +The depends on it because of the stax dependency and this has worked in the removed from the maven repo on purpose? +The change from DERBYINSTALL to DERBYHOME was made in srcadminguide as part of another fix so this bug can be closed +Committed revision Fixing groupId and repository location for the daisyplugin +Should we +Closing stale resolved issues +RWUSEFACET changed to the resolved issue. +Steve thanks for testing this patch and putting a lot thoughts on this issue +I feel that this issue is too important not to address prior to Seam +bcca forward declaration does not match implementation +Resolved not in ER +committed +Looks good so farI am opting for to make graceful shutdown the default in Camel and have it failsafe with a timeout that can tackle if there should be a weird situation where there are inflight exchanges which seems to never completeIt will of course be pluggable and configurableIt works with pending messages on seda queues and it waits for inflight exchanges to be completedAnd it stops those routes which can safely be stopped so Camel wont intake new exchanges while other routes are still routing those remaining exchangesThere will be an interface to use by other components to indicate that they are like the seda and have pending in memory messages that Camel should wait for as well +Thanks +Ill state a few examples Class Level on Attribute definitionsQueueWhat is garbagesupportWhats the difference between enqueue count and total message countalways default value +API which admittedly only works on systems +You the man Karan! svn ciSending containersrctestjavaorgapacheopenejbassemblerclassicTransmitting file dataCommitted revision . +Have to confirm with Peter and Ivo if its true for the rpms too +This is a hole in the typesystem implementation so its critical that this gets fixed by +The patch adds the attribute deregistere +Patch applied in in is all complete and the patches work for me +The document that I referenced above regarding the SAML standard is not an official OASIS standard but rather an independent attempt to create a profile that maximizes interoperability +But I could be wrong about have tried restarting spring roo and it still doesnt take out the annotation +An email was sent to dev list +Fixed in trunk +Closed after release. +You should use the shortcut mvn sonarsonar +maybe a lot of files to fix but the fix has a minor impact and easy to fix +Can you give me some hints how the test look like? Is it simply JPA stuff or some with business logic involved? Or maybe you can share test code itself? That would be ideal +Some of these tests are dependent on each other so its not easy to split them +I cant reproduce it +The deployer should establish a dependency between the application and the component modules the runtime should have visibility into that class loader and it should be used when loading classes related to transformation +It was in the patchThanks to Kentaro Kinebuchi for the patch +We still have many customers running applications on MVS mainframes +The missing properties file for the French and Dutch locales are fixed now +It is a major feature that has a significant footprint and was not all that stable when it first went in +removed Beta since this is not critical enough to respin B of +In any case since the conclusion of was that both patterns will be retained so serviceName will not be deprecated the latter is what we will go with. +No reply so I assume its fixed +just someone should made this decision +See +This makes it a bit hard when regenerating the patches +should those extra actions be a separate JIRA and this one closed +I did find a couple of files in the patch that look anomalous some files in target folder some +But IANAL +PAGECOUNT +Now if you have other extensions that are not but something else then we need the specific setup +After that initial set of messages was presented to the translators we added more messages to both the development trunk and the branch +In fact I added this to as it was not using that setting. +Just verifying the number of dirs expected dirs should be sufficient +Oops sorry seems like I missed that one +couldja give me a tad more informationAnd I can do +Unfortunately the result was not that useful just clutter +Ill just leave the issue in RESOLVED state for now +Thanks for reporting it. +Would like to know reasons for this proposal +Then should be closed if we go a reasonable amount of time without seeing it in nightlys or reported by anyone +Rebase on latest trunk +Patch regenerated. +Covered with selenium test +Vinods suggestion ensures it transparent to users who use either CLI or programmatically +Yes changing the description would Fix the problem +Remove incorrect description of in example server config +then the system procedure creates the six XPLAIN tables in that schemaI still need to do the work so that the captured XPLAIN data is stored into those tablesPlus theres an enormous amount of testing and clean up that will need to be doneBut Im pleased with the progress so far and felt it was worthy of posting +This is on going +At a minimum this should be called out in the reference guide replication page +mvn clean package is equal mvn clean package +Additionally I dont have the cycles to tackle this right now +Thanks for hint i opened a new issue +DRLVM now works with the launcher although a few problems seemingly major problems have cropped upThis patch wasnt applied directly but bits were used such as the fix to the stack trace et al +Ive removed the class and removed unnecessary methods inside the other classes +Result PXE server added in DB +James your observation is right but I would ask you to download and apply the patch +Fixed +Dont we need to exclude the from the generated jars incase they are on a clusters classpath +Sorry about the delay but Im just finishing up the unit tests around +The first three patches I think should be deferred after the release GEP when we have more time to create both a manual and automated testsuite for Abbot testcases +I actually consider this as enhancement rather than as bug + there is no chance it could be called concurrently correct +This patch does the following +Were using Metrics with Graphite which is quite nice +I will attach a patch soon +In response to Ricks first comment had a larger problem that a Spring managed transaction was bound to the thread and never closed so requests to our webapp kept using the same transaction +I kept the isInstance asserts just because why notDoes that work for you +Committed to master. +fixed will verify in the next fixed in build +I hope its not a big problem +I can But let me first figure out what really causes the problem +I switched to using which supports the SYNCBLOCK file creation flag +i think this is for JSF xmodel +Meanwhile it would be great if you could usetest the plugin as it is now in trunk +I dont mean that its a bugI want to request the new featureIve post in forum to ask a solution for a login timebut I didnt get a well and complete solutionso I want to do add a encodingProvider for hibernatebut I didnt find the best area to add that +Reopend until the work is +It applies to BranchSPCP revision +Sending trunkassembliesSending trunketcSending trunkmodulesassemblySending trunkmodulesassemblySending trunkmodulesassemblysrcizpackSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanDeleting trunkmodulesSending trunksandboxpetstoreSending trunksandboxpetstorerev changes to openejb are no longer relevant. +I ll test that try to understand changes and take it further asap. +We really need more information if we are expected to verify or reproduce the output it appears that the JAVAHOME value is not set correctly so the script cant find a java executable to run the broker +If in doubt svn update mvn Pdev clean install. +Bernd any update on this? Have you find any solution to the problemI would like to close this because I think it should not be a problem in the james codebase but I would like to have a confirmation from you that you found the problem before +Yeah we need to support because thats what shipped with our +Verified on +I only hope users will understand this and they realize that is not backwards compatible to previous versions! This code in our codebase was added by knowledgeable Lucene developers +not too worried about the duplication its more the changing of URLS as it makes it difficult to deep link to ie it use to not be duplicated and now it is and I guess long term you want to correct it. +The batch gets larger and larger as a result with the same docs sent multiple times build your patch against the trunk not a released version some things might have moved since move the README file to the srcplugin dir populate the language field from the value generated by the with a default to en if its not there in the readme and in your comments above maybe explain where to find the doc for how to create a domain and declare the fields etc +Logged In YES useridIts not always question of wanting or not wanting +Removed the headless check +Attaching patch +If you are to do that youd have to take this into i understand that +Which method? can u please paste a snippet here? just generate code for your wsdl and it generate stub with correct method signature so I have attached that for your reference +Created to address this +Try the version +This makes is much easier to troubleshoot any issue with metadata cache refreshing +I think Hitesh might have tried it already +The Eclipse test was using a specified data dir while the Tomcat test was using the default one +If I install special methods that cant be used for dispatch purposes method caching breaks because we have false methods in the lookup chainThere were other variations but ultimately having the methods exist on the proxies as real Ruby methods always ran into chickenegg scenarios that broke dispatchingThe best we might be able to do is provide standard utilities for querying Java methods separate from Ruby methods +Sounds good to me +Thanks Mahadev! +Good call simplify the signaturesUpdating the JIRA description now patch to follow +The average search time for grouping by indexed values is ms +It is believed to be a testsuite issue and not an actual bug +There are groups of users +This missing Affects Version +See +think this can be resolved for now +for your input but I dont think this solves the code still runs with the but the byteman rule that sets the flag does not work Just wanted to check to see if theres been any progress on this +Looks good +externalRedirect url params must be explictly provided now. +test to make sure that Hadoop was covered +Hudson is now running the CTC tests nightly +Submitting patch for a qa run +Latest war to view the stats +Yes you can manually maintain your and interfaces and we expect most RF users will do problem is that the is hard coded to use reflection to look for a static findFoo method on your Foo domain class and is also hard coded to fire all methods not tagged as Instance by finding a like named static method on the service want to replace that hard coded hackiness with something pluggable but that pluggability may not make our GA +Something more elaborate than a combo box allowing filtering of records to select across a foreign key relationship +Please focus on the hard stuff +was deprecated or not +Closing issue. +Resolving +The application is now released from a separate project. +trying to run the tests again +ANTHOME is +Reopening to reassign to Luca +Any other thoughts if I dont have mavenThanks +Version Released +Lets move the last pieces to as osgi is not supported by leveldb yet +That way everyone will have better confidence in whatever approach gets takenAwait your feedback and any more comments +if its an array it will create a new if this is a repeat post wasnt sure how to attach the patch +Although the itself will not run into threading issues that is not to say that the code which uses it wontTruth be told aside from the deadlock issue this isnt even guaranteed to work because there is no guarentee the session instance is the same between threadsIn later Trinidad releases I would say this should be replaced with a lock +Also you need to enforce HADOOPCONFDIR settings otherwise test cant find Hadoop configurations +I am just not positive what those optimizations should be yet + for the to trunk. +Mike Any thoughts on this one? Is my suggested change not advisable? Maybe it could be put in properties or somehow externalizedAs it stands I have added the modified to to my classpath so will make my errai upgrades more difficultThanks +If you want to restore some data you need to copy which should follow the current copy semantics +This was resolved as but the id attribute entry in the tld was overwritten by Patrick with revision +Good suggestion Steve +just generated by giving server somebody given an update on this issueWe are facing this issue now +Do you have the full stack trace +fixed in on latest ACS build +Credentials have been stored for user mary but not for MARY the DBO +I said I think its probably reasonable to have a separate message for the specific purpose of printing JAVAOPTS +Dmitry did you find out what were the cause +Its basically automatic unless there are merge conflicts +Could you fill in those blanks? ThanksThilo +Patch attached in a few momentsIt succeeds in phases first failurecp +That includes the numeric types along with CHAR VARCHAR and the datetime typesAlso you are correct that Roles should be in the list of objects in the Tools Guide sectionThanksRick +Im not sure if it work and maybe we need to kick the build process again +Even in your own book your mentioning that cascading not always is the best way +I feel letting access is Okay + +Second attachment +Yep Tom didnt check that one after the backport +Fixed by removing the undoability of the action +Applied your patch and made the following changes +Applied patch with thanks to Christiantrunk Also I removed the dep on as it was not needed +Thanks for integrating the patch. +The fuse team can decide whether its worthwhile putting that minor pom configuration change on the apache branch +Moving to wish list as no action is required for g release +Anil can you take a look with this +I tend to agree that before initialization is appropriate since init methods may expect proper instance variable state already +And Im not sure the Board is going to simply create a new TLP because you guys dont know how to operate correctly +I cant reproduce it +Should we ignore and +Will not fix due to deprications involved the lazy approch of sorting new content into a useful sub package will have to do. +Can you increase them by up to an order of magnitude to check that? +Konstantin might have more comments on this one +Please apply this patch on the features module. +The attached test case creates both a one and two member containment cycle +is the root package name of the code +This forsurely replicates the problem when using the NMS frameworkIf you could do this that would be great because maybe we are missing something in the unit testI think if you run the reporters producerconsumer and everything works correctly then its safe to assume this is an issue with the NMS frameworkRegardsPS Sorry about the previous comments +We want to avoid that until we are closer to release +Hmm that will be the JTS version eh? Ok Ill patch that too +The motivation from Hairong was in our production clusters we often see that the number of get calls is multiple times of the open calls +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first. +By all means better right than right now +So for instance T is not more specific than U but if you swap T and U then it parameters of U can be converted to parameters of T which by definition means its more specific +Bryan I havent read your comment completely but I noticed I wish I had a clearer grasp on the concept of a partial join order +Help in resolution is appreciated +Thank you! +I have a reproducible test case now based on the consuming broker hovering near full memory utilisation +Fix invalid creating the jira should of been enough +Then remove the reflection call declare the dependency and create the class now it is loading classes in a conformant way and there are no clear dependencies declared +Patch applied thanks! I need to do some house cleaning but I will soon attach a more complete patch to that will allow you to my results +Thats not a diffSorry by diff I meant the difference in search performance on a vs where the files are all cached by the OS +Examples are no longer built with Ant the build file in question was removed during the migration to Maven +merely a GUI problem +Well see if further support is required. +The fix was verified by Mikhail +The issue here is that the text currenttimestamp is not being recognized by the translator as a method node +The wont handle all of the cases because can write multiple files +I think this patch is for +Good catch +One thing we should probably address is the brittle cmd cloning +That really starts to blur the line between a router and channel +Fixed in trunk will merge to to branch as applicable +Its called webhcat these days rather than Templeton +the branch code is now merged in and might start making this easier or at least give us a framework to add it into and we can adjust legacy code +I see no reason to automatically create a jar if the +Im closing this issue. +Ive been pointing people with related questions on to this though +At the same time I agree with Dave Medinets about the documentation in its current form it doesnt give much info +Hi ive seen some issues being committed to x branch but not for trunk yet this and other issues are marked as resolved +ABCD +bulk close all resolved issues in preparation for and releases. +Lets keep this to make compaction pluggable and add extra strategies separately +We are ironing out some wrinkles so that we are sure is something that is going to be a solid starting point. +Ill have them posted in an hour +Turned the internal datastructures into Queues and added synchronization. +Attaching a patch file +This problem and the attached patches are similar the previously fixed and its patches +Moving up to we should look into writing a clone method +During the release process we need to be sure to use when building the release then update branchtrunk to thereafter +But they are in the initial text file Nigel attached +Seem that this has not been fixed +I dont see any tests with empty files do you? We always write something into it before closingChecking pendingFile is redundant since if fileBlocks is null then pendingFile is always null as well +Thanks Tom. +Thanks for the review Suresh! Ive committed this to trunk and. +It can be disabled by setting metadatanone +People who want to embed Tika in other products should preferably use Maven to get the dependencies +So I am closing this ticket as a duplicate +txt files to +This will likely be closed as not reproducable unless you can provide a test case or at least sample code on which the problem has been observed +Server log for the all server first log for the all server second used to deploy jbossws cxf into the jboss properties used to install into the tried with HSQL and and got same error with both. +I have omitted to change the level of the logging +m moving this to since we are already going to be reorganizing deployment because of the planned work +A test for has been enabled in Tests run Failures Errors Skipped Time elapsed sec +java and one +delaying until milestone until I know what to do with this module +Im thinking about this and I want to get this done and get it committed +add log warning message to to notify user that they have added no inputsadd checks in old and new api Map and to check for null mapper or reducer throw if null with message such as No Mapper class was providedthis exception carries over to the class in the old and new api and the Pipeline class +It can be tested with the testcase attached to said that the patch need to be reapplied after commit of +Moved out of for the late response +Except this does not work with the release plugin +to removing the limit for thread local pool +I committed this +added to list of known to list of known from the list of resolved diskspace requirements as per +Attaching a rebased version of my other patch +So the idea is to have an enumeration which defines all this in one place +I will follow up with Madhu and see where this really like the idea of this patch +I am glad we have understood each other patch looks good +Verified withOracle JDK Oracle JDK Oracle JDK +Enabled the remember me feature with default in +Ive tried Sun JRE and +You could go around this if youre doing embedded +This way the module is actually placed in the lib target folder +I believe it is this way because the peer will still respond to a notification from a peer that it is looking so the message is not strictly necessary +Marcin please apply this to tag +Ok looks like we need to do a quick soon here to fix this bug +I mean forwarding from an action directly to a jsp I can see the multiple select whilst if the action forwards to a tile that contains the same jsp in the body the combobox are not displayed +I have converted Java Service into Simple Method +So no it shouldnt be cropping up againJames how did you obtain your +Added a note for the Throwable instead of Exception per Kojis recommendation +Remove all dependencies +Attaching a patch with Amareshwaris comment incorporated +Also if it helps I have following plugins installed charts gdal gml oracle and restconfig +Not sure what you mean when you say the rename is not atomic +Luckily theres still though +Downside system specificFor both the upside is we dont need to build a Derby specific system level authentication mechanism +Is that a symbol or a propertyI think the enhancements I have planned for the expression language will cover this +Committed svn revision with the updates to from Ian and Shaz +the result is the exception reported in the description of this ll open a new issue to get this fixed in the next Maven release +I think this pretty critical I couldnt successfully split a log for a long period of time as the log splitting was moved about among machines failing on each + +I interrupted the test several times during execution +it will continue to delete the rest of the files even if it encountered a locked file or any file that it cant delete and will throw an at the end replaced with this it be better to use s deleteQuietly method then just test for existence at the end +It would be hard to return a better message in the because the region server doesnt know whether it was a scanner whose lease expired or if it is a genuine unknown scannerSo I think the scan telling the server the lease period does seem like the best bet +jar +bulk defer of open issues to +Close after release +Reopen if you want to work on this +Still keeping around for backwards compatibility to be removed inbulk closing issues that have been resolved for more than a year. +Hi VladimirEnvironment X Java trunk in rev +In EAP ER Section still contains the following textIf your default JDK in Eclipse is not a Java SE or Java SE JDK you will need to select a Java SE compliant JDK +for me I would even make the build fail if checkstyle find errors +I intend to commit this and backport tomorrow if there are no further commentsobjections +onclick defined on table and onitemclick defined on items only. +What do you think? we should keep the code in sync across trunk and until finally there is some incompatible change on trunk then as close as possible +This fixes the problem because at the time of SELECT when we do privilege lookup in SYSCOLPERMS we have info on the newly added columnI have added few tests for this Jira entry in lang The derbyall suite ran fine on Windows XP with Suns jdk Any feedback will be greatly appreciated +Can this be reproduced using any DB +but an admin can always change it for finercoarser granularity +How do you feel about this patch? Checks hadoopjars instead of +As long as the current configuration works as it stands today Oozie should be fine with any other enhancements +Will do so for +I modified the Hibernate tutorial application a bit but its structure should be familiar to anyone whos walked through the Hibernate tutorial +We probably should backport to +the patch looks good to me +I will take a look into it +tnx for attached +DamienThe purpose of my comment was to set your expectations at a reasonable level +Reopening to get this fix pushed to master and branch +Instead duplicate objects are silently the current version can identify this kind of loop shouldnt some kind of reference tag be used instead +This patch gets rid of the interface and which keeping almost everything else like it is +about plugin deps rabbit mq put all deps in the plugin repo +Thanks! +Rick you diagnosed the issue + +I was just copy pasting some of the dependencies created by Dims in a previous commitI normally send an email to the dev list regarding my patchesLooks like this patch has been applied to the trunk +Not sure if the Solaris or update was helping but we dont encounter this issue anymore +Is that right +attaching a jard that contains a complete listingof the log records of the database +Ill add some more tests based on the pvm enterprise timer which works under the same idea +Attaching patch for latest patch +Antonio if my understanding is correct there is no Sonar issue here +Changing to an improvement +For now just record a warning message that inform user they are deploying a wab not a java ee warWill commit this change to M but not trunk for now +Alexey well would be a bad thing if the purpose was to create a String that shows all characters correctly +Would it be possible to revisit this issue? Are they the same? +The last two are from cloudstack +I have also removed the message body from the signature now signature is computed on timestamp element only otherwise encrypted key and signature elements causes an error when checking signature on the client sideNevertheless this issue has to be fixed because it is frustrating to not be able to serve Axis clients box +can you reproduce against the default set of columnfamilies +Attached you find a new patch for review +Thanks to Deepa for pointing the issue with A in the svn statThanksSunitha + +Updated with changes from javassist which is the version imported by Seam +So you may want to create and maintain your own global object that has the right characteristics for session management withinYou will obviously also need to handle the case of session expiration properly when a content fetch fails because a login session has expired and relogin is necessary +Thanks Uma +configuration or properties file locationIf so is there any chance you can attach a failing? I attempted to reproduce it but its working fine for me +Do we have that is the question +Minor clarification was modified in and trunk has some additional mayhem to clear out first. +We already have a bunch of stuff marked Deprecated that was meant to mean not really tested or supported or maybe used but one day someone will make a decision and work on it and it wasnt really used or acted onI am also not sure the project needs more stuff +patch or +Scott Ive only tested this on my Mac and it happens fairly reliably +Revision added the new system for download mirrors +maybe for just open a separate issue? In this case I thinkwe should just give it another Scorer impl +Theres a way to get the current value of a sequence in OracleSELECT FROM DUALIs that what you need +for each parent resume max results from employment detail children and max results from education background childrenNot sure that behaviour is possible in the approach +closing self reported issues resolved on versions already released +We are generating the right html +These tests are commented and not runnning +the correct one should be jdbcsqlserver +The caching registry interceptor is now enabled by default using the same value as the service invoker cache life unless overriddenUpdated in revision . +Resolving issue per Ritikas comments on the dev mailing list +A compare with instanceof may be slightly faster but causes much more problems at classloading of interfaces +Am I rightEclipse does not generally provide understandable P error messages +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +svn ci m Applying unnecessary cast patches from Sending srcjavaorgapachecommonsbeanutilsSending srcjavaorgapachecommonsbeanutilsSending srcjavaorgapachecommonsbeanutilsSending srcjavaorgapachecommonsbeanutilsSending srcjavaorgapachecommonsbeanutilslocaleSending srcjavaorgapachecommonsbeanutilslocaleSending srcjavaorgapachecommonsbeanutilslocaleconvertersSending srcjavaorgapachecommonsbeanutilslocaleconvertersSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilsSending srctestorgapachecommonsbeanutilslocaleTransmitting file dataCommitted revision . +Please upload your complete wsdlthanks RichardI just used a wsdl which has a custom fault defined in it +That doesnt fully solve the problem though +Ive reviewed the patch and it looks good to me +MS Tools preference pages now have a root preference page they can attach to +Or push my social skills to get someone else to do it +Hmm +Needs to remove label. +Creating these Jira groups wont give anyone access to any wiki +The above comment really belongs to +I think it should be removed from trunk +Oh I see this actually doesnt implement testtraining splitting which is the other thing I thought should be separatedWell why dont I have a crack at fixing up the formatting and such and adding that and then committing the whole lot +That document is not valid anymore +Much more information are displayed now when Sonar console is in debug mode +Are you going to move it to srcmainnativeTheres a proposal to move fusedfs out of contrib at +I am closing this however should we decide to go in a different route based on the discussion in feel free to reopen so we can handle this discussion. +Rejected see previous comment +Sorry was easy to reproduce with the example i was playing with +Hi Andersthanks for your reply +I made a patch for the draft which output slow rpc request to log +Updated comments + +Hey Uma can we either close this JIRA or it as I previously suggested? Thanks a lot. +This has been fixed by Sanjaya +It is fine either way I just was not sure if you intended to use it but then forgot to change codeAlso I assume that this patch is to be applied to the trunk +Thanks for the test file +This registration system should be done as soon as possible. +Paul thanks for the clarification +Without a jar id you should get a syntax error when declaring a procedurefunctionDerby already provides the extension to allow method names without a jar identifier +The code was unnecessarily using a char so changed to use intcommit m replace char by int as a work variable to allow compare with Committed revision +Posted from Bugbox for Van you dont need to set the defaults just move the import to the top of will take care of setting the properties +I will look into incorporatingcommenting on it so we can proceed +Uploaded the patch to fix this +second DSA added. +I have a simple change to jetty that AFAICT fixes the problem Im discussing it with the jetty community +Wrt I understand why we check cluster for null but per my above comment cluster could never be null here rightAh touche +Ill set this jira as resolved and close later +I believe these were intended for future releases +Added back some more methods that were removed from +if you project folder name with jar just like projarit will failure to start the struts +OK cool Im fine with JAN actually it makes more sense for a Japanese reader. + with for +The text extraction works as expected +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +will have to wait until. +bq +Ill try to have a closer look at your proposal some time this week +Mostly fixed needs more investigation +That is an AJAX issue +A reasonable pravctice would be to recommed using a base config that loaded and configured all the plugins +thanks +Thanks +is in my classpath +when is it planned to have a solution for this +When you use Ctrl in source you can select to make the query that should be shown in HQL or Criteria query opened that way would make sense to allow saving back to where it came from just an idea +but other nodes in the cluster shouldnt be sending the data until the schema is settled and we are ready by doing the wait for schema after update token that might be violated right +Lukasz and I have been working a bit on the build and getting some other mechanical things done but we should be getting close to attempting a release real soon +Well stick with + +Thanks for the resolution +So we dont need to additionally mirror the JAR on beehiveThanks +In my humble opinion and accepting that different domains might have different ways of dealing with exceptions I believe it is important that someone from Grails document a working design for how to handle concurrent access to Acegi controlled domain objects +Yes I will try to write a simple test that can be part of the urltemplates web test +This was caused most probably due to the errors in Sandesha moduleThe latest Sandesha module does not seem to have such problems +With that fix we can survive loss of a single datanode and recover the last WAL within secondsI have not heard from the HDFS community on it yet but I think once we lose a datanode that should not be chosen as the Primary Datanode for lease recovery nor for block reconciliation +Thanks +Jim if that patch is fine and no one else objects Ill commit it later today +Now if user does not see a visual preview it is clear to him what is the problem +Some minor changes to the patch added Glassfish to the set and an extra testLutz it would be nice if you could test this on an OCJ server +One question +No worries I got a set of entities to reproduce this +Committed to trunk +Turns out it is required for development +Tested with one more testj while true do ik kj jexpr j echo double dj dk di ij done tee the patch to +Thanks Matei +Workaround using Chrome +In the later case we want to preserve the logical separation +However I see such a behaviour also critical as Olivers example with the Windows path shows. +I added a couple of simple tests just in case +Basically I added a default block to that throws an +Hi MarkI dont understand why the number of BZ users is an issue that prevents FOP being transferred from BZ to JIRA +In our project we were using exactly the behaviour which was removed by this fix we were using a converter to create a body from object +KatheyI will upload a write up of that was done and how to expand it +Ill inform him +I will wait a bit more and close this issue +dont you really want to check whether or not you heard from guy that you think is the leader +This is a problem of the jsf specWe cant solved this in tobago this must be solved in the specMaybe it is solved in a future version of the spec but I dont expect a solution so farThe other problem is the bookmarking +This probably requires the symbol table computation to be updated accordingly so I let you handle this Evgeny +I have emailed seperately my testsresults with in case where Geoserverwould be doing the gzipping +Its fine to have this issue deal with both issues +Verified on revision . +Robin is this something I should commit or do you have an updated version +Venkatesh is working on the new option dryrun that displays the files to be copied by distcp also +This jar contains test codeone java file with test casesone xsd file with XML Schema definitions for the test cases andthree xml files for testing XML Schema primitive datatypes derived datatypes and assorted complex datatypes +Hmm +Here is a new patch that renames the base class to +Dupes +I have merged the latest PR into so we can continue with this +here are the three test projectsi found out that the problem can reliably be reproduced by deployingpublishing the two webapps at the same timeif the applications were deployed separately everything works fine +Im certain its good enough to get into the repo now if you believe its readyIn about weeks I am beginning writing the chapter so this is very timely +nor +BTW when im investigate this bug im found a functions and +Thanks for the comments! I have no problem is comments the latest patch for the defect contains them +Second patch applied at rPlease check that this fully resolves the issue +Im looking to retrieve the documents added to the index in the order they were added or sorted by descending order of doc ids +Only certain people should have access to this screen presumably those that are authorized to manage this sort of thingI like it the way it is +OK thanks! +One of them probably isnt strictly compliant with the SPNEGO spec +Instead of fixing the Wiki I copied what I considered to be the most relevant bits to the Quick Start page +Like Olle already pointed out this issue is a pain for all Maven users including the joint compilation can be delegated to the Eclipse JDT compiler then this issue should be resolved +Will not get to for the release +Corrected in SVN revision OK with the following PATH elementCProgram FilesbinCProgram FilesbinZhomebalexbin +Easy to fix +Committed +Thanks Mladen for the Solaris and Windows buildsAs this is a regression we have rebuilt with the patch for the next handoff +Start of my tasks getting assigned to you +Then cycle jetty and attach EclipseBy no means am I proposing that a more complex delivery mechanism is appropriate for production use +Hence resolving this +Yes same ve generated a patch using git +I committed the patch to +That one issue could simply list the general intent and all of the touch points in which is usually how we do implementation planning +Another easy one to implement via hudpull new regexp pattern on all Trunk and jobs +can we set this option also from spring file +actually I think this is related to the fact that I dont see old searchers being cleaned up +We can do this in +Unassigning due to inactivity. +Initial version +That means call before +We could have something like instead of inside +Verified on Jenkins +Rico Id appreciate some testing in your environment +I want to look into a light mvc framework something that will get all the code into servletsjava classes +This was the logo preferred by everyone +IMHO should have the same behavior +also added a check to see if multiple sources exist under a model. +Retargeting to and +Committed to trunk +However maybe until I get that up it would be usefulOn I dont think removing them interface completely is really the way to go since both the replication metrics and the region server metrics are mostly dynamic metrics ie they arent like the master metrics +So no issue at all +This should be tackled after the other two the deprecations be to? If we did wed be able to remove the actual code that much sooner +There is not much more that I can do at this point. +I just installed the build and didnt observe a change in behavior +Please verify that it was applied as expected +Recent fixes have fixed this in develop branch. +What do the urls look like? How is it securedCan you write a page of docs for it or explain more about it here? I didnt see any +Uploaded candidate patch and ejb testsuite reports + +Committed r to trunkCommitted r to +move defines after the include of the openssl headers +Patch was applied as expected +Duplicate +Did you create Userssvnpigpigbenchmarkcache +Hello Pascalthe comment from Werner was ment for me and I created a test as he described it +get API in Job to be used later for the UI +schema tool does not check anything and just add switch to db instructions to the document +right thanks for the clarification. +If you feel you want this changed for any given list you may want to ask these to ask for the change to be made +Im using version and its ok +schedule +Agree on above +fixed in revision alexei please check that it was applied as fine thanks Mikhail +Is there any? Invalidating client sessions is the goal of this patch +I think we need to make this a blocker +Did you trim down the stacks? I would expect to see more +Thanks for the specification URL +Its only available from CVS +Thanks Ashish +Committed addendum thanks for watching Elliot +This issue duplicates +Im tempted to push this to as it doesnt appear to be terribly importantessential and it would add a fair amount of complexity close to the release +Fixed and verified. +It does warnings but I cant see them +Since we went to the new manifest its been assumed that the model name represents a schema and the file extensions are stripped off the file names and the type value is used to distinguish the model all we need to know is can the accept as a valid model name value +Committed to trunk and. +That sounds good Im happy that this difficult problem will find a solution +IMO the answer is yes too +ChristianCan you show me where in the RFC it is stated that redirects to the samelocation are not allowed? In my humble opinion multiple redirects to the samelocation! an infinite loopOleg +As it was discussed on Im going to provide an update patch for messages marked with tags by mistake in earlier commits +Resource ids need to be URL encoded and decoded. +The ability to unregister services is irrelevant to this issue this is a manifestation of but with juddi v +I also tried xcf etc +Using jboss seam same issue with asynchronous email rendering with quartz dispatcher +Thank you very much. +I sent the error here because while I was running it ask me to send this error to JIRA +Those two tests are only the test suites holding the other tests +fixed the issue in current svn code will do the changesThanksDeepal +This is is my todo list still but hasnt risen to the top +Thanks for the patch +Doing first bullet point only sounds good +Like to hear how to apply the patch +The usage pattern must allow outside a JEE container so may be better off not relaying on Workmanager stuff +Are you defining a procedure or a view? You cannot define a view with a procedure +Uploading a patch for this is same as the patch that got committed at but with different directory structureAmareshwari can you please do a sanity check and see if it works for you +That seems to reinforce the for separate jboss packaging and whats the next step to getting the attached plugin into the sandbox? Do we have a committer amongst us +As soon as one regionserver aborts its filesystem is cached andor closed by user the master file systems DFS client is closed and all hell breaks loose +Added example wiki documentation DONE +The message is now no longer a BUG! message and it contains valuable information that should help to solve this problem. +Were never going to fix this because of the drastic impact that block retry has on several jit optimizations and because Matz agreed its a baddangerous feature +In the future it might be a good idea to use failure detection techniques like phi accrual for the speculative read timeout instead of having a fixed value +is now and returns the stylesnullText + +In order to be able to handle a larger number of simultaneous block losses we should have the option of dispatching map reduce jobs to fix the affected files in a distributed fashion +Suits me we need to push on getting RFAJ restructured then. +Kudos. +The IBM JRE loads for DSIG +project and +Any feedback on this patch? Safe to apply it +is it ready for review? I skimmed a bit it makes sense +Patch committed revision Thanks Saurabh VyasBTW I see that you do not have an ICLA on file at Apache if you are intending to work more on Derby I would encourage you to submit an ICLA ahead of time rather than waiting for a significant contribution +This has been running for weeks now without issue. +Like to see AXIOM not depending on anything other than +the regression happened on Saturday night and is expected to be fixed RSN so there isnt a JIRA to search for +There is no mention about new proposals for attributes of seam mail and in the file +You are right +Heres the modified now thanks for the for getting that included so quickly +Since the test module is being moved outside the framework and python doesnt like outside the parent module sharing the parent module the test segment will now be known at springpythontest +Look forward to getting this working it seems a neat extension to anakia +Jira freeze was on this is not possible for unforseen reason this must be escalated to the release manageer in a timely we dont stick to these simple rules our release planing becomes meaningless +Gary Ideally all of HC but cookie classes would be a startAll Given this turned out to be such a controversial subject I am kind of inclined to close this issue as WONTFIX unless someone would be willing to take upon a task of making toString methods consistent across the entire library +Now considering this issue fixed. +r for release branch +Patch to provide the fix +However might be the wrong target to introduce this changeOK +With M docs and K hits I get almost similar results +Patch is committed. +my patch +it triggered the creation fo this jira and we are waiting since then that is is corrected +I tested this manually using the program also attached to this JIRA +The tests should not be run under Maven +Thats all I wanted to make the test failures go away +Thanks! + thanks +Is the overhead of wrapping a string in a list causing performance problems +Id like to be able to specify an arbitrary location for my and files +Or do we have some staging? And only need to add it to the available software sites in eclipse +See +I sent you some files that can be recreated by following the steps I detailed in my original comment +Sorry I forgot to say that I was trying the injboss example not the JEE dukes bank example so Im not sure this fixes every thing for that work is well underway with AS all previous community releases are +ThanksNeil Hauge right +Thanks +Libvirt issue needs to downgrade libvirt to +I remember the discussion going somewhere along the lines of if users wanted to customize everything that was added for its commit they could use the git menu +on commit to +fix memory when but not Delete issue +Please delegate to someone who can accurately document not using ARC in iOS. +Vincent why are you not comfortable to remove the aggregator? The current tagging functionality basically doesnt work with any flat project layout +But is it a practical point? Lucene shines on big indexes and queries resulting with full index are not very useful I guessI think its good idea to hide the caching reader class and utilize mechanism to make the fix more effective + +I had to go into an old backup of my Eclipse setup and manually pull all the proper files and folders into a set that we can use +Linking to +It might be more esthetically pleasing to define some new entry points axiomsoapfaultvaluedetachbasenode etc +Ping on this bug? What do we do with it? +Attachment has been added with description cheap cloning of the foreign transition resolvedfixed bugs to closedfixed +Hi Andrea sorry yes I think Im all set +Through the use of inner class structures and a new public method the generated packages and classes may be inspected +So for we can only collect request level stats. +But please fix it as a separate commit +no were not planning to migrate yet +Since this has the potential of breaking other users due to browser bugs should there be a new setting for it in +Because its JSP based chances of dynamic ids in components are also very low so please list the ones that you need as you encounter far as your previous comment is concerned you need to explain bit more. +I think in or is fine but I think its important line number is before file name with a meaningful separator +After looking at this and the discussion on nabble I think we should go the route of making the read methods const and making the stream mutable +The problem is that the load group mask was switch to abitmask of a long +Can you send me an example of this? Itd be good to get a unit test that captures this +I added the repository to the project so it can be run standalone +Those are currently done with and friends +Attaching a patch of test with Albert and Ill take this todo off his plate +committed to on +This patch allow delete default project group and project groups which have associated build templates +Igor forced me to work out the exercise for the reader so hopefully everything is completely explicit now and you are in agreement +Your wish is my command +Thats the same library as I mentioned in the original issue description +I think that a better solution might be to create an that would be used for all calls +It should really be looking for should be a +Those OOM issues go away with increased LCS file sizeBased upon the data from this experiment an LCS file size of MB would be an optimal default value +Any suggestion how to translate into italian +is a complete patch +Close after resolve. +ear on latest AS trunk and still see the same error +Do you need me to find out Permaine or were you asking Fernando +In my tests after a while rebalancing is inconsistent +who push this in central repo? currently its not in +The RS proceeded to open the region and since it had come in under the wire the RS stayed up holding up test completionA RS that has been asked to stop by user via or in a unit test testing balancing should reject requests to open regions +Have opened another ticket to address the caching idea +I got the same error above +I cannot reproduce this either in Rev +We have four servers +I checked this with JBDS with enabled debug in in test project +committed to SVN trunk +It uses the build steps now so I think the plugin is no longer able to determine the correct pom and I beleive it uses f by default +Attachment oe has been added with description Solves problems with gaps in +Yeah this has been on my todo list for a while now +See is fixed with revision of +Whirr cannot list clusters +Committed to and trunk +It should be a fairly transparent change +Sure it suggests there is a problem +For hot NIC plugin in Vmware must be installed otherwise an exception will be thrownCan you recheck if the issues still exists if are installed +We plan to send the code to you for review and check in to the projectwhen completeKarl Pieleck +Look at and the proposal to use partial persistent data structure +Richard which one is my second issue the problem or the problem? I think you mean to say that one should not worry about initializing fields +This is fixed in trunk +Yep it should be in the IAPI only +Then we dont need to manage two repositoriesThe project folder and its are already set up to be included into trunkcomponents and be build together with +Ok Ill take care of issues to next release +Strings or? takes less mem but requires two array lookups +Good point we should load classes lazily there +Feel free to reopen with a reproducible app if this issue still exists in the latest version. +Thanks IvanAlso thanks Chris for reviewing this +This limitation is mostly due to the having the limitation of not allowing multiple keys of the same value +is very big +I dont know if there are any other issues with memory stream transformations but if there are those would be something differentXerces does not use any of the standard template library so this shouldnt be a problem + +Then since theres no modify method the component has to be deleted +I always agonize about whether to include cs tests or not when the changes are seemingly embedded onlyI dont have a general rule but in this particular case it looks like we run the entire test inside a stored procedure in order to make sure that only embedded is tested + +Those sorts of comments in a patchseem like reminders to me of things Id want to resolve before committing the patch so Ithought Id check to see if they were things you wanted to fix prior to commit +I have reported this problem to Jetbrains +Similar for javadocs which this does not touch at all +However Im not sure what we can do to fix this one + +what do you mean +Btw Jon how extensively did you test this? I wonder if something like RF Strategy could be something that could be handled in a distributed test +Typo in Acl associatd +Hey Zied I cant reproduce your issue. +Fixed in SVN +what exactly does fail? building project dependencies something elseeverything +Do you have a patch for trunk that applies? Thanks +Before looking into this can you please provide a more elaborate scenario ofhow you intend to be using this? What is the main driver for this? And why isit not an option for you to take the JVM running Castor down +Would it be worth checking the return value when deleting the temporary file? If the file couldnt be removed then throw the exceptionApart from that this looks good to me +Has anyone found the missing files? Has anyone instead rolled back to the previous andRevision Author junraoDate Tuesday October PMMessageAdd compression to C client patched by Eric Hauser +Yes it now works to add the parameter to the configuration +I think Accumulos behavior in the presence of multiple identical cells is undefined and Im not sure it should be defined but rather clients should probably avoid this situation +patch +Forgot u on the diff +I would like to mark it as duplicate of bug both related to unimplemented methods in the subclass of +This patch passes all unit tests and is ready for code review +I just committed this. +ThanksOm +I think Dans proposal above seems like the way to go +So what exactly seems to be the root exception +As for timer +Ben please use this updated patch file for the basis of your fixThis patch includes everything from the last patch except patched against latest svn head cleans up a number of LOG messages in particular adds a number of DEBUG level logs that help in tracking down issues when things go wrong +Updated patch which uses the value of the badness threshold to control which behavior is used +Please do not file a bug reports untill we have a full QA build and test infrastructure up and for spotting the bug but there was some changes from the beta that will give more infofor such errors. +No longer an issue probably thanks to changes in +Yes but what about the specs? We dont support a feature that is part of the JSR +Assign to Daisy review +Anyway if you set the property attribute then this attribute will be used for the sort +In my installer I want to be able to enter a path or to just leave it empty +Right I missed that +I wont ask you why +Ill work with Simba on this as getcount look like a much better API +r into +This was their responseThese references Developers Guide and Working with user authentication have been paraphrased in translation +Fixed in in this since has been released +Ive also created a test for this issue and applied at the same revision +Thanks for the release Mark +committed to on +I corrected the comment in revision and I also removed twoTracing and logging is completely separate yesThe changes look good to me to commitThere is one trick that may be useful in rare cases since the properties are static and that is to generate a stack trace and look for the daemon thread to determine if the daemon is active or not +Therefore a timer is used to trigger a configuration in regular intervals +delete all childs +Thanks Chinna. +But the album ownership concept is not supported yet +Cleaned up +Thanks this is working +Thanks +Caching the null value is what should happenRegarding the patch of Felix Meschberger I see the problem however Im not sure it is a correct solution to default to the system classloader +Its used by the stand alone build and duplicates much of what is in +As the status changes can get complicated +The first issue is solved sorry my fault. +Checking this in +Patch to add bzip splitting support +Target version no longer exists do changing to unknown fix version +JIRA Triage March agreed to make but subject to performance testing +Closing stale resolved issues +This got opened another time as +Can checksum be returned with roll +Note that while the lowest Datanode should be orchestrating block recovery the Datanode order for data should not change +Also thanks for doing I was going to do that as my next should have been quite a tweak to BLM to reduce the IPD +breaks sort validation +Any update on this +Making it creates a link between and the object that uses it +For example in todays DWPT code we will stillblock document additions while flushing because were tying a thread to a givenDWPT +Looks good except for the duplicated code in fixSo it is ok passing an encoded name only to the master? It will just use it? Does the assignunassign in the master then figure whether it has been handed a regionname or an encoded name +Thanks Tony. +Mass closing all issues that have been in resolved state for months or more without any feedback or update +What was the impact of this if anyYour would not start in a environment unless the values of and were the same in the NN conf +Hi RichardWhat version are you using +Failures are mostly either timeouts or class not found or errors +No idea why this has been done worst idea ever +Proposed patch trims out +In this case should we add the original replica into finalized version +This patch doesnt seem to work +Curious what it might bebq +Im not a strong Java programmer so Im not sure if I have properly handled the full range of exceptions or not +The list of cores is nicely sorted now! It seems to take a bit longer than I would expect to display the core list on a page refresh +We still need to make some ajax examples work well +I have the same problem when trying to build xerces on this config uname a XXXX sunu sparc SUNW gcc vReading specs from toolsgccbin +I was able to get it going by adding the following to the profilemvn seems these need to be added to profile +I think set is what you want +Id argue that we may want to support both interfaces +This issue was taken care of as part of +Still limited function compared to JMS binding spec +I have changed execPbsz to be optionalWe need to allow end users to configure as they want +thanks committed to trunk as bcfecfbbdddc +User indicates that the email still does not get through requests that we change contact address to simboss +just committed this +The result is not but should be shouldnt it? added some logs to +Finding a place to expose it is the hardest part for me and always sends me into a spiral of designing the view server +As to the JDBC Connection it seems to because the same JNDI is used by both the service the webapp both with s specified +and also plz make the guide more exclude the long descriptions of how something was implemented but add more info on how to use the functionality +I have a patch ready that should resolve the remaining test issues +Thanks Chris! +Yes it works if I put the specVersion on the Component +Committed +closing this for now can reopen later if required. +How about returning both current KS and CF in the metadata response on all requestsIf all requests here includes USE instructions then that would solve the problem too +Vikas +lgtmMind formatting a patch for trunk as well +Why do you ask +Hi +Ah yeah thats an issue thats been fixed since has been released +It was changed a while ago to use an svn checkout to avoid dirty extra crap being in the release +This issue should be marked as Fixed Javier and Paul +I find the names compile and runtime dependencies misleading +I never liked and never used the approach with the anonymous package +Applied the supplied patch and added the binaries. +I want to warn that is going to drop the ZK dependency from the core POM +on commit if this hadoopqa run does not uncover something ve committed this +It failed the same way +Added IT in r +Give me some time and I could provide you with a test case if you think this would be of any help +could you attach the file that is causing the problem +Applied to x and branches +I have no feel for how much this is being used but it does allow for a solution where the configuration can be relatively easy scripted from a shell or similar +Closing this again because i still cant see a bug but application three weeks no futher comment and i still cant see why this should b a bug +DRLVM asserts on the second attempt to raise an exception +Closed upon release of Hadoop. +Attachment has been added with description generated has been added with description data has been added with description behaviour sum of widths exceeds available IPD +I did a second experiment in which I killed a client after it created a file +Will open a new issue for the column aliasing issues remaining in the repro script attached to this issue. +It should be except +This is the graph I obtained using JUNG library +I do not see this behavior myself although I am running STS +good suggestion thanksregards tom +This will work out just fine for us +Since it doesnt seem that maven issue is going to be fixed any time soon isnt there a way to say plug another impl +Sorry to revive this old bug but its still present in version and I couldnt find any workarounds or updates +Also i dont know why this template was removing other attributes such as title and target which some people want to use +Just dont expect a lot of commits from me +Im sorry for extremely short issue description +It would indeed be a very good fit forThe last remaining issue though is how it does streaming of entity bodies +However I was using the IDEA Axis plugin which doesnt allow you to specify this when you go down the Custom route which many ppl will want to do to override the default package location for the emitted classes +Fixed upstream by yours truly +Will be fixed in next has been now fixed in these builds this is verified fixed before EWP release please remove the release notes flag if it affects only we dont need to note it +If this is an interop issue then we should investigate more on that and fix it in RampartIt would be really helpful if you can attach a sample message which is accepted by service without any issue +the actual issue yes I think it is pretty safe to say that it is a duplicate and should be marked as such +Its also broken going against any app all the way up to EAP Alpha +What is the API that going to use? directly +In this case input summary for the table directory always contains file and length which might confuse mapjoin resolver to take the table small enough to be hashedIll make a review board entry. +Thank you Faranak! +Attached is a screenshot +The new implementation uses DOM and the routines in the class Eddie committedThe patch updates the core package for ant and deletes the config file for the schema + additional comments about contains java source including xdoclet tags code for processingand Atn build it contains archive that is a result of processing the same source code by xdoclet is an example for EJB plugin how it should really work. +Can you attach example manifests that recreate the issue +Luckily I didnt get around to applying those changes yet +is no guarantee that we support older versions of quartz +Ran into a problem that I didnt notice with the HDFS version of +Its IP if needed is I also have the gossipinfo from after the assassinate if needed +I believe the intent is to put these changes back in sometime after we cut a branchThanks back to Pinaki +Inside partial consumption is enabled with a set command so it should work regardless of what the setting for partial consumption was beforehand +It would be SO much easier if JBDS came as an RPM +This clarifies the followingo Varargs are also allowed if the parameter style is DERBYJDBCRESULTSETAt this point I think that the feature is ready to be documented and +etc +This screenshot shows a standard Eclipse with only subclipse and meclipse added to it +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +No additional unit tests since this is fixing an existing test +Will be implemented in Seam as clean up +Added help comment and committed Thanks! +Heres are the log entries produced during the updateENTRY MESSAGE Error parsing metadata repositorySUBENTRY MESSAGE Error at line column Missing required attribute in repository urlSUBENTRY MESSAGE Error at line column Illegal value for attribute url of element repository nullSUBENTRY MESSAGE Error at line column Missing required attribute in repository urlSUBENTRY MESSAGE Error at line column Illegal value for attribute url of element repository nullSUBENTRY MESSAGE Error at line column Missing required attribute in repository urlSUBENTRY MESSAGE Error at line column Illegal value for attribute url of element repository nullSUBENTRY MESSAGE Error at line column Missing required attribute in repository urlSUBENTRY MESSAGE Error at line column Illegal value for attribute url of element repository nullENTRY MESSAGE Error parsing metadata repositorySUBENTRY MESSAGE Error at line column Missing required attribute in repository urlSUBENTRY MESSAGE Error at line column Illegal value for attribute url of element repository nullSUBENTRY MESSAGE Error at line column Missing required attribute in repository urlSUBENTRY MESSAGE Error at line column Illegal value for attribute url of element repository nullENTRY MESSAGE An error occurred while collecting items to be installedSUBENTRY MESSAGE No repository found containing have the same problem as Stefan Scheidt +Test was failing becuase insert thread was committing the unlogged operation that was suppose to block the backup +For example I send a byte message I only receive bytes +Thanks +Its either that or it needs to be duplicated here until Solr is on of Lucene and can incorporate your changes there +This patch does not include Ive it +I am planning to use exclude filter file to exclude the invalid bugs +I couldnt find a decent location for so Im putting it in libext +update Fix Version from to +Hence the confusionUpdated version to +Adding the release notes and documentation Eva can you please verify the progress on this and resolve please? Thanks +I cant define all validation rules with the hibernate validator annotations +Ill try to close this up ASAP for youMike +Heres a patch which addresses Sureshs comment +is there a test case where the source file is already at the replication level that is expected post replication bu the modtime of the parity file does not match the modtime of the src file? In that case the src file should be isnt it +Also bumps the JDK and Ant versions to latest +Thanks for the clearing this +classpath +It works fine but we need a test for this wizard +As you suggested +I found none of them were better then the next +there should be a toggle button to change order as it is done on java in trunk +Fixed by bafeebecedafab +I just checked out the source but I cant find the class +verified on reopen to be able to target correct fix version. +Accepted +CXF JAXRS bundle will also need to be stripped of few dependencies it does not actually need +There is no Flow Control on the message stores +The key separation of the service vs dl jar file is simply the fact that you dont want implementation in the classpath only in the codebase +Is there a thread forum for this? We might collect further information on where the required classes live and so on +Component should be HTML for the patch +SVN implements this feature and adds a partial flag to the configurePager tag currently this only supports rendering a data set of page size each time the data grid renders +As such the debian package should have the same files and metadata as the rpm package Presently a generated deb package will not have the typical debian control scripts that set up user accounts init the db etc +review and tested patch applied thanks! +same resolution as +Should this happen for or can it wait btw +as a debian packager and ops guy I strongly recommend allowing additions to the classpath in etcdefaultcassandra +I will give trunk a go then +What was the bug? I couldnt tell from the Lucene issue description +Hi Arun Yes there is connectivity between the virtual machines in my cluster +Many thanks for this enhancement to the tests suite +I just want to point out this problem does not affect trunk because trunk has moved to use Maven +Ill start the list I can index varchars but not clobs o increasing the limit on varchars is not going to help this +Not quite sure why didnt run on this latest patch +Currently thats +Make blocker for +Heres a patch that repairs the changes of so that is no longer broken +KarlAs asked by Ashu post some information for us to replicate and work on the issue viz +I dont think this will be very useful but it is good for think this will be primarily for that want to force their own named argument method proposals in certain situations +Patch looks good +I just committed this +I like the idea of +Fixed for V of the plugin. +as I wrote occurs at the session for pointing out the example it was badly written and did not prove anything the consumers were created and consuming in a fashion +Have you made any progress yet? Is there anything we could do to support you? We would be really glad to have this feature +I am unable to reproduce the dependency problem in my environment +I will provide a patch done +Attachment has been added with description The attachments shows a cell with background and span one with only background and one with only span +I dont see a big problem adding this to +Pretty please +Committed on trunk as svn +I did as you suggested and included SOAP actions for every operation inmy WSDL +It shows the message for me too using the blank application and I cant reproduce the problem on my test I dont know why +Tested +DROP COLUMN locates such triggers by asking the for a list of all triggers on this table then iterates through thelist asking each trigger which columns it references and then either cascadesthe DROP to the affected trigger or fails the DROP depending on whether theuser said CASCADE or RESTRICTUnfortunately there dont appear to be analogous methods for columnpermissions in the data dictionary to allow retrieval of all the column permissionson this table for example +Without giving it a great deal of thought Id think that if rbac is enabled an unsecured interface is the same as a secured interface where the user doesnt map to any roles +I noticed that in the repository cache on the working server machine there was in fact a file of size that was being created +The patch looks good to me +Fixes the comments issue in +This one supersedes which appears to miss one class modifcations. +Please do not reopen the jira +yes is in and good to me +Instead it may only cause some periods in which some process believes incorrectly to be the leader but again the service is robust to such scenarios +Hi Kayne this ticket is a subset of so Im closing it +Fell free to update at any time +Cant you use the existing translatehbaseexceptions method for this purpose +Sorry for the multiple comments +The times can be configured in the fair scheduler config file and thus modified at runtime and the guaranteed share timeouts are per pool +Also reviewed on reviewboard by Ken Giusti +Wed never recoverWill make a patch in morning +Updated the patch +I put it under mandatory +war +The patch was committed dfbbcbbedaeaefeeedfaeThanks Sebastien +I think its the that adds these intrinsics I dont know if theyve been backported to updates to but this should be relatively easy to verify empirically +But first I want to post some tables showing what annotation support will be initially provided +This looks great to me +One solution might be to have a selectablesingle +I just imported this WSDL +committed to should also be fixed in Beta its a blocker +This should be after is committed as it will make the configuration side much easier +In order to speed up the patch delivery Im going to submit the new one including these files and run it through process one more time before committing +A viable workaround for people who want to use attributes is to create custom implmentations of the HTML taglib not dont use the HTML taglib +is still very important in Velocity so I would be keen to use alternative compatible implementationsandregards Malcolm Edgar +This was committed and was requested to close +Its not a trivial task and new features for JSP support is not our priority +He did the work +Please review the good to me +plugins Developer Studio debug homeeskimo +The solution is to attempt reflective access to add the additional compilation phase operation if the project looks like a grails project +Hi Kathey This is a customer issue +If you pass mr to it will use Hadoop to classify the data +Will do that point on the qualifier probably being a little over zealous in my qualifiers in an attempt to prevent the init method I thought from the Weld documentation that Inject was the right thing to do as well as from other module code I have seen +Closed the ticket as I did not get a response +Paul doesnt plan on looking at this later +JamesI tried to reproduce the problem here but cant seem to be able toThe build file that is attached to the bug had your system dependencies +We should instead deprecate all existing public constructors and refer folks to the static factory methods +Ive showed Jim Moran how to save views including filters +The file location information has been updated for the mentioned upcoming. +Such strategy would give an easy hook to any property if users ever need oneSounds good +Hi Evgenyyes I can confirm this +Swapping out one filter for the other seems like its the correct thing to do since the characteristics of the stream change once the connect operator is finished +Returned not used methods +why am I doing that? You are a maven committer! +hence the patch is based struts +Seems like the only way expiration can be done properly is to do the timestamp shift as per the and then do the expiration test on the broker when it is fetching the next message to deliverDoes anybody agree with this +The tests the standalone implementation of the +I was doing the translation because Im interesting in that functionality Its my native language so also excuse my bad english and is really nice to know Rodrigo has that great advanceRodrigo if you want to continue and finish the accounting translation is ok or Ill do if you prefer by the moment I cant to help in any other activity because Im begining to get involved with ofbiz Ill wait your commentsGreetings +the failed hudson report is because this patch is against the HA branch +ThanksChunrong +Complete core pages translation for Simplified +thank you. +It was only seen after I upgraded to +I should note I also ran and against this common patch and they both passed +This is already planed for you can change the in Eclipse wo any problem. +Punting to for now +In particular if you have R reduces it means there are R files open each with a buffer associated with it +AFAIR the +If this is could we make the scheme something like hdfshttp or httphdfs to indicate the encapsulation? I always found hftp to be very confusing to users who thought it had something to do with FTP +Thats right its to forward to another controller +Current code changes BI so needs upgrade code or else no old database with any index will be able to be readI do have code for soft and hard upgrade in my todo list +Please find the modified patch and I hope I tried to address all your comments in that patch +disable +Sample class attached +But lets not prematurely optimizeSanjay I see your argument to be equivalent to arguing against service level authorization and that method level authorization is sufficientNo but we will eventually probably need authorization too and it would be nice if whatever support we add now also helps then +Further testing seems to prove that its only an issue with gdal +thanks again +Yes both comments make a lot of sense +Thus the problem doesexist in releases previous to inplace compress +Only done if the +Thank youBest regards by Richard. +hasnt been applied right +with Spring configuration + +I have started work on the above examples +Im not sure +Done as part of +Default behaviour is unchanged so the is correct +When jobs finish all running tasks should be killed via Here the job did not finish it looks like job is marked for completion only when all the maps and reducers are finished +I dont understand the problem where else would you put sstables than under the keyspace directory +How can I prevent the loading excluded appenders? you have removed plugins from your jar but they are still listed in then you are going to have problems +Update to latest on +log the throwable if commit rp commit to trunk this will allow us to see why commit rp is just committed +There should be only text in proposal list +Luke should be able to show any index even if it was not created with solr +Tricky to test we need an LDAP Server with TLS enabled to perform integration testing +Everythingh needed to reproduce the bug +HolgerThanks for the good news +I would say that if one project takes advantage of it that more than makes up for the perpetual maintenance cost of the module but we can debate that if you like +Ive just committed this to trunk andSuresh I see all of and +Personally I think the default requiredness is confusing this is the problem you end up with +a small nit in the earlier ydist patch rectified in this +Its definitely something I wanted to look into but it was quicker to copy and make changes to the existing just to get things going +xhtml pages only in the +rev for EJB that work is well underway with AS all previous community releases are +cmake filesYou can remove the +Thanks Varun! +It seems like this issue is fixedKathey can you resolve if it is +Cannot reproduce with. +What I meant by was making the data available in the Java API so that these stats could be accessed programatically like the things currently available in +parameters for this passwordnew parameters for this of these two new parametersnew parameters for this issue my my only needs one parameterproxy so we might look at how it implements this and use this one parameter instead by Brett + +TedYes +Lars agree that is probably should just be in the regular jar but lets do it in another issue +Ill open a separate JIRA for thatThis patch applies on top of rebased on top of on top of the current that has been committed heres a patch for this against trunk +So please be patientRegardsLars + Id suggest to name it The comment of postContent talks about query parameter while it expects The response could contain an explanatory entity please review comments what happens when the specified resource already exists? an response seems unappriate in this changed the comment trying to more accuratedly describe what the method does +What about WL +What happens if you want some entries to have multiple values and some single? Id rather move attribute to entry element or better yet simply use values instead of value to specify that entry has multiple values +That would be one way to solve these kind of issues. +That depends on how runtime is implemented +Im seeing same for in an environment +Framework is being voted on +Can you double check what version of Groovy you are usingprintln When I run your example using the GroovyGrails Tool Suite based on Groovy Eclipse with the Groovy feature enabled it runs fine for me +So it is going and adding the duplicate private gatewaymysql SELECT FROM id ipaddress networkid reservationid macaddress vpcid taken sourcenat NULL +Take a look at the default implementation of WSTs XML editor. +I agree that by default we shouldnt store all the hostnames in each etchosts on the cluster +I testd with the given WSDL and it does not trow an exception now and the code is generated. +do you see this function getting called a lot in a write workload? Your comments on the mailing list thread indicate that its performance sensitive but I dont see how that would be the case. +for the committed this earlier but forgot to resolve the issue + +Ported fix to +Note that different downloadable runtimes are availableNow do this instead Window Preferences Server Runtime Environment Click Add Click Community AS Click next click Download and Install Runtime Note that only runtimes show upThe dialog itself has the notion of filtering now +we could catch the oom in this case and instead return an error to theclient +This should be solved at the kernel level see the linked Marek I fix it into the kernel could you please test it to validate my fix +Agree its easy to set up ganglia but imho its important to have the master know if a regionserver is sick for a variety of reasons +Committed your changes to recompiled works greatThank you a lot! +I just tested this with the latest Castor trunk and it seemed to work +Im fine with pushing this one +FYI +Maybe it would make sense to use the provider for those since we dont want them to contain some BUILDSTREAM and other JBT specific labels +I should probably mention that I am not convinced that moving the gen code into the init methodmakes things much clearer +Wow what a quick response thanks +Now that the definition of the Hibernate session factory is factored out into its own Spring configuration file this issue is already half solved +No worries Jan +Thanks selckin I think I will let Steve or someone more familiar review your patchAbout the there is a problem somewhere +As reported in above bug we are not able to invoke +You should now have perms to share this works we just received the first message +I do not have permissions yet to close the issues I am waiting on those +screenshot of the wizard selection window and the sample project +I have the working fairly well and I think it is a good solution for shutting down the AM if we run into some unexpected errors +Thats why the additional files are not found +The only problem remaining seems to be As showing up in the Server tab +The screenshots in the GSG in Welcome to Developer Studio section should be updated because of the old title +Thinking again lets mark it as PublicUnstable +I will explain what I have in mind there shortly +although we did test and continue to use the free license we have been granted whenever we get the time to make sure we release resources etc +I spent a little bit time debugging the issue and found the problem might be caused in +I have build system changes +as previous comment said this is already in and I dont want to add new features into since it is a maintains branch which should only accept bug fixes thanks for looking this +I think disabled gbeans in a configuration are a much better solution to the kind of problems I think this is intended to address +I agree with you and I would follow up with this issue +Otherwise I have to start all that much later in either case +Applied at r fixed a couple of warnings in. +Thanks +It also doesnt yet have any inputformats so it cannot be used from mapreduce +AFAICT this was a misunderstanding of what you need to do for artifact aliases. +Sam Ruby wrote Per Sebbs comment above we have been presented with a license that gives us access to unspecified IP under conditions that we ourselves dont meet +Currently the partition column value is checked for the st row and assumes all the rows have the same partitioning column value +for removing it if the license can not be figured ODMG support in vbranch. +So the check should be +Canceling +Heres what the x version would look like if anyones interested +If someone can come up with a suitable workaroundpatch then please submit anew issue. +Moved over +i sort of wish i hadnt pushed and prodded yall down this road +I personally do not consider this as a blocker for Axis to move forwardThese days our devs are busy in stabilizing the code + uses inode id fixes a bug from previous trunk merge +Not having custom prinicipals available in our is a major problem for us +Affected components are Central Hibernate JPT plugins Common Openshif Maven GWT IntegrationWe did try to make as many plugins as possible to be compatible and there is no way to support both and in the same branch. +Hence I upload this patch to correct the build break +I think Id first like to know what problems are being solved by making the registry easier to get access to +However at least it makes you think about the roles of your business objects and their entry applying AOP to method calls within a business object youd need a solution like or which modify the byte code of your classes +Closing no response from reporter assuming OK. +I want to backup Dhruba that passing each and every op through the will substantially increase load on the and that holding lock longer will reduce its productivity +Thanks Randall +I found a bug cause by the Patch when creating the order +The now contains a parameter that defines IP addresses of trusted proxies +Obviously wed like to have a single set of docs that generate this but Im not familiar with the toolchain for building mainpages +But before we start fixing up exception messages we need to define a standard pattern for all exception messages across Web Beans so thata users know what they are looking atb tooling can use itDo you want to have a go at defining a standard pattern enough of the code is in place that we could do this +Ill wait for Hudson to run and commit it +Fix was to next release. +Im currently looking into this and will post when I know more +My understanding was that automake has supported since +Id like to commit it in the next few days assuming the licensing is no problem +this creates and maintains the Catalog in the productstore linked to the website +I agree with Erwan this is purposely done so I close +function queries arent weighted +Each servce or in case of out scaling each instance of the service must have an unique id +Scroll to the bottom and uncheck Enable semantic highlighting + +Using code borrowed from +Suppose you are using the build from trunk +Pushing the button took me back to the choose plugins to install page which also seems odd +This cleans up the sockets hanging around in the data committed this to branch +shouldnt we still handle YARNHOME as deprecated +No exceptionThe code itself does not swallow this exception +If not then we can close this +While without this patch that part of logic would be executed inside multiple mappers +We dont need to maintain strict on trunk +Fix for all platforms +The package name change is from to +Patch checked in +The code of s Portlet enhancement was committed so this issue should be closed as duplicate. +Which is only possible in or previousPerhaps this does not actually require any fixes for +Ok thanks for explaining. +Even after this patch the AM will be localized as the original user since the RM picks up the username from ugiMaybe we should remove the user option completely? and use in the AM which is anyway set by the RM based on the logged in user +Those are the kind of blocks that initially triggers a into this state +Compare with the has event handlers for all buttons on the form why Action cannot have the same features? Ive been promoting approach for about two years +These two patches mainly put different table in the same page into different +Was missing an exit statement +Its a new branch +spre regionper region +Firefox Chrome behaviour +v looks great committed with a thrift minor version bump +separate from + is also tickets +See this as it seems aims at much more than just SSL for Cloud and thus may take considerable time to completeSo if anyone feels like cooking up a small clean patch for this specific issue then please go ahead +I applied the patch to current CVS the issue is resolved. +Patch looks good +Since the shutting broker is sent the it will stop the fetcher and fall out of ISR +features +I think things should be a good bit better now after went in +Heres an updated patch which adds the comment you asked for +From the Hbase point of view all that really really matters is maprfs but the rest is still there. +why should the first token have an increment of ? I think the first token should always be +All s leave it open until weve fully tested it +First the classpath of Tomcat was used with the oldSo to use JAXP with Tomcat we must update the of TomcatWith Tomcat there is no problem +This looks done then. +But found that client data sources are also tested in jdbcapi which is run only in embedded framework +What is your OSI see in your log that your svn send message in french it must write output in english because scm provider support only the english +Moving it out +The full fix will require additional output for the simple history report +please refer to my last comment thx +Resolving + the patch looks good to meI think it would be best to move the checks out of and into prepost test case methods which can probably be done via a similarly to the patch thats posted to +Hey JoeSince these are categorical values I suspect it should not be a problem +Upgrading to Maven solved the issue +This now works perfectly +Howard The JIRA Issue described the internal implementation details I dont want to care your internal implementation +Andrew thanks for pointing out +i mean the fisrt time the page is loaded the attributes and the parameters are there +I believe the site gets generated nightly so it should be fixed within hours. +The package private collection converters do call back to the implementation which we can revisit if we really need such converter spi flexibility +please when you cut the new branch. +Old table of contents +This error happends on a regular basis in my production environment when the server has been runing for about days +I cant see anything wrong with the delivery either +AFAIK is a replacement for +Change session timeout to seconds +If the jar required for the job gets deleted in the middle while job is running then job should be failed and AM for the specific job should die isnt it a expected behaviour? If it is a expected then we can gracefully kill the AM with an appropriate diagnosticsPlease correct me if am wrong +Committed +Thanks Daniel +My +and doneWill finish up the rest tomorrow morning +Committed to +I get a bad version number in the class file for the model in that repository +ok the previous approach does not work +Used for comparing two cluster centroids +on somehow getting seq source to log exceptions was hit by this myself +Fixed in branch and now the fix is the same as all my fixes are this selection has been added to the list of overrides when setting up the launch configuration rather than just in the launch +Committed javadoc fixes. +Found an issue with contrib project vaidya which accidentally re includes the top level this was causing the compilation of the contrib project itself breaking +Note that simply waiting for containers to complete doesnt address auxiliary services like s that want to serve up data after containers have completed +This patch enables the plugin to retrieve the list pf plugins available from the plugin descriptors specified in the went wrong and I seem to have uploaded the patch twice +You can control the ssh options so you should be ok configuring it +Will be committing it shortly +The default encoding isMaybe the author uses an editor to watch the result which isnt capable of handling encoded textSet to resolved. +Relating issue JIRA +Resolve this issue as the portlets have been enabled +so comparing embedded to clientserver is likely not very useful in the setObject and setString case +Im just trying to understand what is the use case for having the fileId avail in the returned by the REST APIFirst of all I like to emphasize fileId is not in REST API +When set to false the fix will add only the root warning and the warning chain should be read as shown in this JIRA. +Thanks +before the tests we even create the directory on file system etcLet me know what you think +I suggested that the submitted of this discuss the idea on instead. +Fix provided by commit can be pulled onto product branch +this is the very reason i thought that JMXPORT could also be configurable where do i think it should stop? i think easy wins like port numbers should be configurable via environmentthx +on patch v +I think we should target Apache with a fix for this as well we do have a build on Apache hudson and this is an easy fix that will help that out no +Thanks Sijie. +Hi DoughAttached is a patch to add to TLP of Lucene siteThanks +It is an error to use this annotation on an interface +This helps identify when messages flowed throughThanks +I am on Windows RC +applied after a few bug fixesMike +Ive updated it with some javadoc and more UT +I ran several single node cluster tests manually with this patch +Thanks a lotRegardsNareshWont fix as per the commentColm. +covered as part of +It is not the functionality of the to order the children as per the business logic of the application using itFurther I think we should not burden the Zookeeper to do things which hold little relevance to the core functionality of the +Thanks for patch Dave. +This is on release. +Remote deployment of WAR files need their original filenames so the context root will be set correctly for default cases +But the optional dependencies are necessary to start the so they appear in the OBR +It also addresses other documentation fixes for the testing webpages that are required in the upcoming Xalan Java release + for embedded + +The failed task is because ofIt is not related to this patch +I think that this is addressed by the patch for +Thanks Eli I will take a look on thisThanks attached with fixes for branch security +Exactly how much capacity does it claim to be showing me +That wouldnt be a very good approach because any actual implementation would have to use variable sized collections and then convert to an array at the end which is exactly the thing that the user will do if that is what he wants +HelloIm unable to find the correct method that is called when doubleclicking on the attributeI can send you a VM just tell me where +Previously you had to set an Analyzer even if you added NOTANALYZEDNONORMS or otherwise youd hit NPEIf the tests pass it means its ok because DTW is not extendable in a way that it allows you to control how documents are added +The spec in fofootnote Constraints hasIt is an error if the fofootnote occurs as a descendant of a flow that is notassigned to a or of an fo that generates absolutelypositioned areas +Rewrote the title and description asking for support for the richcalendar +I removed the inline styles from that paragraph and put them up into the they want to modify the css in the header or apply it to the other pages they can find it by clicking Page Info Javascript +went in so marking this as duplicate. +Will do an automatic merge where possible with subequent imports +We could pass in a custom Configuration and specify a new portno including a random one +Added several unit tests that check the existence and validity of In constants and messages for all annotations referencing them +Tomasz Is it your intention to have this all as one long page or are each of the sections in the PSD separate +The version of logj specified in ivy is however that version will not work for Maven build since it pulls in many unnecessary dependencies +The input should already be cleaned in this regard +v actually applies +Next time dont give me twice the work needed +What exactly do you mean with a member of the communityOr do you mean an ASF committer +Hi guysI had added these samples to the place like Snjezana you can take a lookIf need I to add others please add your NickI dont understand what you said +so please try again from HEAD +Here is a sample project showing an implementation of improvedthe skin class is added to develop branch +suggested patch +check now it should be corrected +Ill try to be responsive although Im a little mired in investigating some of the defects which have come up during testing +I want to use the executor service that is already there +I bugfixed it on the Source code +bq +I cant test it right now but it looks good +Ive committed this to trunk and +fixed in r +Can you please run Maven Update Project Configuration and then try to run your tests again +I for one would be against this that the inherited attributes are in the wrong place Id agree and would probably like them better on top but otherwise id rather have the tld declare the attributes in the same order as my code which also influences generated documentation and as such allows me to put the more importantrelevant attributes first +Committed to trunk and. +Thanks +The patch also applies to +Only than produces correct relative urls +Ive just committed this +Closing as verified. +Please speak up if you know anything about this + +Y! distribution just committed this +Does you know some analags where such feature exists? Its not user friendly behaviour by my opinion and doesnt needed on web projects where to include in some files used one file for which we include js css or others and then include this file to other files +This trick is also used for the +HmmSorry that youre having problems +What is the status on this? Are you still seeing groovy files formatted when you dont want them to +oops +For the opened files NFS gateway also saves their latest file size +Need to talk about tested +this happens to work for normal rendering but by no means is by designif you want to continue to rely on this you can simply replace the enclosure with a webmarkupcontainer + +If you agree please also fix to not use versionAlso how do you plan to set the sytem property? Need some changes to files too +Fixed +dba +Note that the process will deploy fine if you add the queryLanguage attribute to the vpropquery element in the WSDL and specify xpath +Does it work now or why did you close this issue +Hello +Please delete this directory manually since the files have been moved to +Friendly ping +Nick can you please attach separate files if possible +Committed to and trunk Thanks! +Yea I hate how android has these little bugs that are such a pain +XML should be added in to M since we still need to implement the approach +Patch LGTM has been committed thanks! +Hi Sergey Im using IE as youve already figured outVersion number Ill see if I can attach the html file I use +ThanksYou are right about +It takes away too much control from implementations and replaces it with a fixed set of options for score caclulation which sort of defeats the purpose of this APIMy objection to the original patch was specifically what the should do in this case and not what any should be able to do +pending Jenkins +Thanks. +Removed example not a good idea +Bob Can I take this issue +Please verify. +findbugs and test failure are unrelated +If a user already exists in the users table then they already have a group +ing issue as part of JIRA cleanup +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +Thanks +From what I understand the aim of this jira is not to reduce latency of an RPC but rather to reduce the network bandwidth usage +Can you please give me a link to the installer you used +Changing status to patch available so that this will show up in the review queue +No worries happy to be able to contribute. +I think its getting pretty darn close to ready to commit +Attachment has been added with description Test FO to reproduce has been added with description table scaling transition resolvedfixed bugs to closedfixed +Smooks Transformations +See also the reproduction project attached to duplicate issue +As Gary points out though the application consuming STDERR and STDOUT needs to be able to handle the data effectivelyI missed your last comment when I first posted this comment +console config it is correctly inform me about The Java project zzz does not exist +Do you have any ideas in mind on how Id check to see if there was something there before that we replaced and put it back +Ooops +bq +Sorry when changing from webwork to struts there was a misunderstanding +Please see the comments to that TR on how to obtain andtest the candidate fix to the blank page problem +As Vishal pointed out there has been some fixes to leader election and to the logic that connects a leader and it followers since +Therefore the exception will just output the exception name and no details about the overloaded method +Good idea +I dont know about case three David you can simlpy copyField it around in Nutch mapping or Solr +fixed for all the distribution projects as well +Has the deployment truly failed or just not succeeded? isLoading can still be true with a missing datasource as that is considered a recoverable error and loading will resume once the datasource is available +Ok now were getting somewhere +Robert I just wanted to have a fuller record of whats going on with various Hadoop deployments models on this JIRA for folks to to get a full disclosure +The only thing left to do here is port over the changes to Felix and submit them as a patch to +often the small number of rows is actually growing but we happen to compile at app startup +Thanks a lot for the contributionWill you be able to attach the documentation as wellThanks will add a documentation patch for the samples and +Operator error +This picture shows the imports and the order they occur in +I opened a new entry so that we can discuss it there +xsd files are not found are not located in platformplugindtd directory as described in catalog descriptionand IDE is not able to locate them and use them for new XML file generation. +During the duration of this incident the regionserver in question was being pounded while the rest of the cluster was idle +no +Looks like the query index has an invalid match possibly for an entry that once existed but has been deletedmoved in the meantime + has been updated +In the IDE and especially on restart one cannot start resolving dependencies and possibly download ve added a patch to embedder in issue +Can you take a look +I got quite a long way with this but need someone to finish the will work on Petes code +Release Notes marked Not Required as this issue relates to Red Hat internal testing +Im not sure that disabling the IBM JDKs validation is the right way to go if nothing else it helps to show any cases where we have invalid tags +It isnt clear to me why they are treated differentlyI dont see how they are treated differently +Full agreement from me on your comment Roy but Ive felt many were disagreeing +Can an option flag be set to fail the build on a cycle? +In Rev +can we avoid this extra dependency? is just a client and is not supposed to depend on lucene or solr core +echo +I cant see why this couldnt happen when writing other type of thoughts +This pretty much seems to have been dealt with elsewise +Also is jruby still using this instance of Jira? +Posted a patch that overloads the method that is used by the build from the +The patch currently just gets an application delegation token from RM +I think this issue should be about solving that problem only +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +oh i did not see the title +Thanks Mark +Looks goodApplied the to trunk and +Patch applied with slight apt formatting modifications Thanks very much! +Pers patch does fix this issue +Did you run it on JTS transactions? It failed whenever I ran it with JTS +Ive attached the same patches here +We used to detect difference between a binary hbase install and a checkout by looking for presence of +Minor fix error in LCE tests +committed thanks +Feel free to submit a patch +appear to be pretty complicated about their baseline support but particularly for the requirement can be simplified down toI want my label and my Combo to have their baselines line upNote that this requires more than just a change to +Id hate to see customskip this configurations in POM files for all IDEs and their specific bugs. +Sync on write is now the default mechanism without the buffer cache +bq +Patch attached +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Thats a good idea and should make the schedulers more efficient as well +And leads it to the behaviour described in code annotationsI do not know which solution is more appropriate here +thanks to Mario Ivankovits for this patch +Can someone point me to a clear unambiguous statement regarding the policy on versioning and bugs? At what point does a version become unsupported? So far as I can tell becomes unsupported the moment is released +In any case I believe most or all functionality has been removed +What is the correct practice for having service method return the correct status of As save? Is a flush nessecary after saving A? Is there a way to in general force a flush after every save? What are the implications of this way +All these failing tests pass for me now +I have an excuse for committing this during freeze VMTT is not a part of the release +Applied at r thanks! +What worries me is that a lot of effort is being made to make this plugin work on when it is as yet not clear whether rpmbuild works properly on at allI looked last night to see whether rpm was available for and Ive found a lot of people complaining that while they can get rpm installed they cannot get rpmbuild to work for them +cant you just add a sort by clause to the query instead of storing it into a temporary table and then doing cluster by +Also note that this patch adds one more buffering to input and output streams +Thanks Mark I just committed that. +It will be a bash script but a windows user should be able to convert it easily to +Some threads are blocked for more than seconds which would lead to users refreshing the page and making the situation even worseAn additional problem seems to be that one has to aquire the lock at the object pool to return a page to the tool so if one thread renders a page the whole application gets stuck immediatelyWe also retested with Tapestry it does not solve the problem +Sorry I found the solution to this +For example there is no way to differentiate this issue from one where the dropdown actually became unusable because it didnt its current state locating the entry you want simply means remembering where its ordered in the list of projects +commit ccfddfeeaabffbce +Its Eduardo not Steve pIve done the change you asked +are they all SQL +Ill enhance path support to be able to use folders or +I was wondering about what happens in the following situationo Alice invokes procedure which executes with definers rightso Procedure sets role to HR does some work then returnso What is the role of Alices session after the returnThanks version of this patch +Backported changes into codeline with revision . +Id consider that to be a superior enhancement +is this what you intended +This one contains all the files being indexed into not just the first level +I committed the patch to trunk +Id been considering this issue some myself after coming across an email from Jan way back on the mailing listOn the whole just being consistent made sense to me but I was weighing that vs changing things on everyone already out there using the current stuff a couple blog posts talks etc +Two possible solutions areuse character instead of write a custom and use it in Wicket +expandedtrue on group doesnt expand parent group +Jongwook sorry for not getting to this sooner the queue for RC is still long +The Rails philosophy says we should put all our database structural changes in migrations +How do I run that pages? Where is +Added section for Java System Properties +fixes ID gaps produced by standard implementation +This is pleminary patch +Is this completed +Attaching the same patch +IMO it doesnt require a release note +Adding the as I did is just another way to deal with this problem +So either stop your moaning and let us get on with it or pitch up and offer some help +The thing is the referencing module in gt assumes if two datums have different names they are different so that will never match + +Should actually addreplace its own during launch like the other platforms +annotations need to be updated in the EAP and release notes. +For you could use a with as JNDI indeed it helped +As part of the coprocessor API I think its critical to get this in +The reason you get different results on different servers is likely because of how Tapestry attempts to determine if the client supports your CDN must be sending headers that confuse Tapestry about this. +stack startup is a mess with our assigning out regions an rebalancing at same time +I dont think so since is still holding on to the s Request object which is the produceRequest +Username rafciom +Ill investigate more +yes i listed them above and marked them in code +I can go ahead and give the MAVENOPTS a try but I am doubting that this is the fix I need +Can we explicitly test for? Thatll be great if possibleSure +Similar to the older issue it fixes itself if I make the package name a single word. +baz would be the same as foobaz DavidFrom my view Geronimo behaviors correctly it does handle those relative to the module not the ear root directoryI will post an example that I used to test this patch and the updated fileIf anyone has time please help to reivew it +to the patch +This patch works fine in our production cluster +So it must be something in the workspace +vs +all issues to the CR +Sorry actually I think the entire logic of my patch needs to be +Ill try to get the testing going asap +Will fix that in next version of the patch +Maybe something to experiment for +class with extended trace level debugging and my solution to fix the timeout error +Attached is the diff necessary to apply this changes to the ivy project +Yes a check would be fine +I wasnt able to find any option to run unit tests in a mode +Please provide an example +Committed to trunk rev Thanks BJ +Dyslexia on my part +mrepository +This patch adds the Edit user info link in the side navigation menu +Further adjustments with commit eeffddafeafafff. +Two uses of unset that I see Readers schema has a field with no default value and the writers schema does not have that field +Thanks Rick +Alexey please verify +Is there documentation on symlink input format for the case where both the symlink file and data bucket is on s +Alexey please verify +I think the events arent getting hooked up correctly but I am not sure +pushed to in Developer StudioVersion Build id Build date +I should have tried original reporters testcase again when I started commentingResolving. +Having said that it used to be ok on too! +Its not possible to upgrade only when needed +If the changes were compatible then why did we bump the version number in the first placeIf we do not bump the version number old client can continue to talk to a new server but we also want a new client to talk to an old server +Im resolving this issue as Done since theres no Nothing to do resolution. +Theres server side file system based implementation at the client is not recoverable from restarts in our initial recoverable RM client will be provided in next releases +Erik I think that makes a lot of sense +Hack was committed to and trunk +Attaching which bumps the version on the trunk to +At this point the ability to limit methods is in the Struts and code but it isnt exposed to the user in any real way +Wouldnt it be good to have a command line option to exclude certain modules from coverage analysis instead of skipping it entirely? Or Is it already possible to achieve this via command lineCurrently we are doing it via pom file and it would be good if some option exists to prevent frequent changes to the pom fileThanksGayathri +same storyCant write any scripts without Grab +At that pointthe connection is availableMike what exactly is forceClose? The only reference I see to it are a coupleof test cases that send Connection close headers which is not appropriate inthis caseThanks patch for the should have taken care of this one too +Any chance it could make into the Xerces any time soon? +this is important should get handled so users can extend their wait important as long as openshift still can time out over long distances if too hard to fix can be postponed for CR but should be in GA +Makes sense +See +Links is notpart of +Committed v patch +Fixing them to get the tests working as before +The attached patch fixes these bugs which are all part of the replication functionality and is not in use in databasesBugfix patch va passes all tests and is ready for review +I have a feeling that the problems were probably fixed in other ways but I still prefer to take a cautious approach and not try and fix this unless there are some strong opinions +and without specifying the dependency in teiidmain module the is should never modify our module need to validate that is the name of the module and that it completely defines all of the necessary dependencies +I am using Tomcat on Struts +Unresolved as of +how did you run scrubto me it looks like you ran binsstablescrub not nodetool scrub +With git you need to use so that Hudson can apply the patch +Xuan took a quick look at the patch +Thanks Jiri +Jesse thanks for pointing these out +Possibly i will look into attached sample application runs fine for me against current SVN head is there something im missing? What are the steps to reproduce +The patch has been appliedPlease verify +Ill look into that and figure out a design however my only concern becomes the Session beans should we replace them by regular java classes or use a transactional based framework such as Spring? I understand the philosophy of to keep things simple and abiding by core java what are your suggestions gavin +Hi KenI know you cant share the actual classes but is it possible for you to createa test case which will reproduce the problemThanks Ken The inner class problem is probably related to the fact that Castor onlysupports static inner classes +This should go to I to TRUNK +sounds goodFeel free to apply it yourself +I think I could make use of this feature as well +Verified that the shared flow is not null in the derived class using Carlins repro +bq +cocoon database of tens of thousands of emails Bayesian classification ends up with emails +Patch has been issue open until it has been committed. +and its autocompile triggers creation of query classes +So people using Java already do not have to bother about these dependenciesAlternatively we could add the in the module +For your Parent and Child class types that is +But the meaning is the same +Issue may already be resolved may be invalid or may never be fixed +What good is it when the manifest method is implemented +Waiting for this patch to be committed +Patch updated to latest svn +The lists are used only in upgrade path one for functions and other for procedures +This patch introduces the changes to the +to trunk +Yeah that is the schema I intended but it took me a while to figure out why it was failing +Please cross check and close the bug. +Thanks! +My problem was with using the old cvs repository +I am not even sure that the differences are +Its really not worth debating anymore but if the client decided to guess and blast a INITIATE it could simply set a ignoreNegotiate boolean + +Moving to to correctly identify default namespace and add to the SQL transformation. +I got this exact error message when I had a mistake in the actual driver definition +So that work regarding and can begin +del the view all routes pagefixed the bug willem and jonathan mentioned here +Patch applied thanks +I thought for Accumulo we were sticking with as the default version and moving to for Accumulo +This log comes from the original file which I later reduced to the test case +Grid can you look into this +There is indeed something wrong with the content assist +Pushed to trunk branches +Sure just give me a few days to find some spare time +The text is arbitrary and comes from the CNN website discussing General Motors +Release completed +Is this a dup of maybe +Could you reapply pleaseThx this code has been changed a bit by other commits attach a new patch for easy applying +Sorry for the alarm +Excluding versioning tests +Marking closed. +Hairong this patch no longer cleanly applies to trunk +New patch does contains trivial updates +thats a very simple thing +That is bad I will correct it so even if it fails you have a chance to correct the things causing the failure +For you above being a user of has a very narrow version range due to the subclassing but in many use cases a much wider version range are expected to work and still yet for some projects even the exact dot dot release are needed +Volunteers +see +Second change reviewed by Keith approved for +Done +if version exists but doesnt match throw exception crash RM +Contains and test for that +Fixed at r please verify it works as expected +This will lead to ignoring of status updates from a task once it sends the done +As of the xalanjcompat code has been removed from the repository and is no longer included with releases +Did you ever see this recur? +I have the same problem Im using RF to so that we have one bug per issue + Takes care of converting soap and soap to the qualified forms and +Only the review requester can update the patch for the reviewSo the person composing the patch should be creating review request +I will apply the patch as it solves the custom activity issue +Note that that does not mean people can not use solrj to talk across different versions they may have to use xml though +But until some clear andconcise steps are provided advising us how to reproduce the bug we cannotcontinue to dig +Moving this to for investigation +Can you please check? Should the empty textmarker package be removed +I dont have Websphere but maybe someone will see something +If so lets change the for a together with replacement. +I ran into this issue over in and did same fix there +Agreed +My employer has pulled me off of this task so I wont be able to do much work on it anymore +However before that Im pretty sure I can get the same result as Mirek did +modulessrcmainjavaorgapacheairavataxbayaappwrapper +Ah right now I see it in the image +Switching off Kaspersky IS everything was fineNevertheless looks funny to me that Metro workedThanksHannesHmm +Uses Michaels NMTOKEN regex +Even if the query is reordered as Chris suggested the literal is still not quotedSo it sounds like my suggestion addressed at least the parsing issue just not the quoting +To that respect batching is in the right place here +Let me check whats up and I will come back with more information tomorrow as Paul is currently not available +was reopened to address them +Also realized that its loading the configuration fresh off disk not displaying whats actually running which makes it a little less useful +For example consider the case of doing range partitioning to achieve distributed order by +Ive committed my patch at +Regressions passed +Closing old bugs. +Thank you Manu for clarifying agree your comments and close this jira. +The proposed change is meant more for debugging purposes +lead left someone else got miffed and took project to. +Marking outside of release cycle didnt make it +add null pointer protection +aftersBut you get an error message in the couch log rightYes I think its intentional +Great if you can come with a patch unit think we can take by default without having to force the user to add a new configuration property +For the record Im not in love with the statefulness involved in this algorithm but that predates this issue +In this case wed run the splitter in SSH not via executor +The reduce side is reading the bytes it knows about hence the result corruption +Sebastien I really sorry but I dont understand your use case +Ive committed the latest patch which addresses your new comments. +It uses the nodeid info from A to workB container request is processed +The only way to fix them would be to recover chains of blocks and put them into lostfound and here we could do better than fsck because we know the file name they belong to +Are you assuming that this has been done prior to the refreshDoes Redhat deal with hot pluggable disks +Youre right cant really generalize this +This one includes them +I dont have any EAR projects in my workspace +nativelib is varlibnativelib +the patch +I still disagree with applying server logic to the process of constructing messages on client side but I think a made enough arguments and Im glad that they were at least read and taken into consideration +Used different names for different modules +Result is Invalid argument in JDBC call parameter index out of range . +I will investigate this issue +Future revision may be necessary +It would be harmless but the spew of errors would worry people +Much have been done and commited +service resources +But it should be a nice reminder when someone actually has to do a release +Thanks Alexey! +Andrea I think an extra directory got added when you committed this patch +Patch added +Just to confirm have tested teir with latest nightly snapshot and same happens +Resolved for a while + +Patch applied but I agree it still is confusing. +Although it would only need a few px enlargement there +Committed to trunk as r +server side is also done +this seems to work havent checked for any differences in null pointer behavior may want to check that as well +The good thing though is that the discriminator value could be inferred from the presence of some properties and avoid the pollution of the data model +The file is only usable when its located +the word application stands for Derby Network Server and the sentence could beThe Derby Network Server can run as a server with Derby as an embedded part of the Derby Network Server +please dont set the fix in SVN. +I like this idea +Agreed that any issues will be dealt with through separate issues +Can this issue be placed under the changelog component please +This meant modifying the resource file parsers with naive comment skippingIve also added javadoc for all visible types and membersRe documenting the resources Im not sure what I can do +Output of yum list before the installation yum list tomcatLoaded plugins rhnplugin securityAvailable tried it again on clean system with exactarch set to and it still took tomcat from RHEL base seems that for the only successful method isyum install would be a workaround for now +Yes I do agree we need to fix our hard coded Upper to Lower conversion issues once Annogen fix their issues +Please give it a whirlNot only is supported but I now also provide the actual expected and diff messages. +Can someone explain +Patch to quote relevant values +In addition the aforementioned changes Ive updated the and to verify the two spots where is used +the patch does not apply +compat and simply update to Unicode seems to be more important +Diff versus tag +Bart Let us know how it goes in production and if you continue to see issues +for this patch +To get the same length on trunk wed need to put back and dial the retries up to I can just revert the change from trunk and +Is it possible to provide us with a sample pdf which demonstrates the issue +Nick how goes it +Thanks Benson +is preliminary I havent run through tests yet +for a pull request on this one ll send along the pull request +Thanks +Heres my first cut at this currently compiles the core fine doesnt cater for the contrib modules yetAs said previously Ive just simply adding a new core directory to contain the pom to fake Maven into thinking it a with the core referencing the java src via +Attached patch collects the metrics in jobtracker +I prefer followup issues +We can delay the rewrite until source query formation and pushdown support would need to be added +Fix has been made to refresh the instance data if the Properties Viewer is openedon a RUNNING instance that has no key data filled in +I would argue that with Manishs patch committed and should be closed and a new issue opened to cover the remaining problems with column aliasing shown in the derbyrepro scriptSounds good to me +I am not sure how we missed that one +was used to mean the highest value representable by the type since the types are unsigned +Ill need to call this out in the release notes +Verified on a cluster that the patch fixes the page load slowness +Fixed issue in type for now carries the name and equality includes name +Did you mean this code +It compiles and passed my testsIll check the current patch into now and assuming you get a patch up for the test Ill throw that into trunk too +Looks good to me +Delaying this until M +Thank you for all your work and the explanation +BTW Galder whats the CNFE raised? Like the Entity classes are not found it that what happens? +HiThis seems more a rhinohttpunit issue than a Cactus one +However the code in question was added for deployment fixes so Im not sure if its totally cool to remove that logic yet +However Ill check that this does not affect the ROWID handling +What do you think is the maximum value I would want to useand again Thank you for your quick reply +Update console can filter messages add test example move query test logic to common checkpoint agent and console test code successfully performs simple test of agent discovery query for packages query for schema ids query for schema +But the schitzo ISs should never escape out of the top IS that has an ES +Have you made any changes to the queries before posting them here? They dont seem to be generated by the sql compiler +All completed +Spoke with Scott Marlow who had some thought on the issue and will comment +I dont see much added value to this but we could still consider it +I getCUsersgroovy versionCUsersProgram FilesGroovybin is not recognizedas an internal or external commandoperable program or batch dont think both the quotes and the apostrophe are right +Will commit if tests pass +But if you just comment out that code other code will break because then default primitive values could trigger a validation exception for I agree +Hi Myrna and is created to convert jdbc and jdbc to +Solution copied to trunk as well +Bulk close all issues. +I just committed this +Im not sure what i can do about the tomcat since it seems to signed in both cases +AngelinePlease provide database dumpIm unable to reproduce this bug in my environment without database dumpthank youJessica +Phew +Free for dev pay for production +Ill try to find time and troubleshoot the issue in my favorite themeThanks +Thanks for the patch Michael! +I much prefer the second version for both its clarity and extra flexibility +Work reviewed and committed by Martin +Yes I think continuing without may not be a good behaviourBecause that particular regionServer may have more capacity to take up the new regions +I am going to use HIBERNATE for now +Do you see where it went awry earlier in the logs? Where master had regions as k and then of a sudden starting thinking one region out on the cluster only? Good on you D +Simple change to the shell to make sure that it knows where the bloom filter class is +Puppet changes +do you think Debug Information is really sufficiently descriptive if you dont find Debug enoughPerhaps Debug title with debugrelease values are not optimalBut Im not convinced by Debug information title with yesno objections fixing this to Title Debug Information Values Yes or NoIt is at least better than the currect situation +The test was updated to show additional details on RI behavior group can be destroyed with empty subgroup group can be destroyed with a subgroup thread group cannot be destroyed with a subgroup with terminated thread + +Will get back to this for sure +Patch not yet ready for review +Im not looking for every possible exampleBut since we give an example of a custom update chain it seems we might give an example or pointer on how to actually put it in use? Its kind of a frustrating example otherwiseIt should be in the Wiki as well +Something like this +If you still encounter the issue please reopen the bug with adequate steps to recreate +If your concern is that we might miss other kinds of exceptions hidden in LEE then I think we could grep the exception message for File does not exist and otherwise let the exception come out like it currently does +This should be implemented in the +Committed +Richard patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Likewise if youre looking to Get Involved then its with the Community collection of projects in generalIts obvious this is the case once you click on the links and this is part of a larger effort to help people understand how to engage with the whole Community not just one project +Wouldnt be hard to change this Interface name to? +Attached patch removes the guard against nodes not being in the nodes map in +Do you agree? The new data structure will be similar to others like etc +We should fit core in this dictionary as well +yes the numbers are greater than +The refactor looks good +I have tried buliding patched axis and confirmed E option is valid and solve the problemThank youWhich version of axis will include this fix? +Office program with a preparestatement which is useful to reproduce the issue +ALOS ALOS Prism Image was Imported to a native raster format of a COTS tool +I came across this as well +Rebased the patch +I couldnt get it passing tests however and was unable to reproduce leaking need to fix this definitely but I cant find where it is and it would be a bad idea to rip apart the threading logic immediately before +Verified with JBT +My guess is that the intention is that the default value is not serialized +just to add Im using osgi +Hi AlexCould you review the attached patchcheers issue now also incorporates see the linked JIRA for details +Closing as per comments above +tested the the issue if there are any issues we have to reopen this issueSvante +This shows a sample test case attempting to attached to springpythonspringpythonspringpythonlocalhost +However this is not easy to fix if only we can fix itPostponed to to but it will soon become a feature +Maybe someone can find why we get this serialization errorre right DIGYIt seems like it passes sporadicalyAfter adding the attribute it passes so I figured it solved it +Confirmed to work now. + set fix committed a more complete implementation of this in revision +So thats why it works with SEDA as its using a instead +doesnt a hashmap have to box into an Integer for lookup? +Any ideas you might have are very welcome +I suggest we postpone this issue until OlegSorry I did not look in this thread for a long time and now my reply is probably too late but anyway I hope it may be useful +I didnt know that the was available in JNDI as well on Server +Such as and give user a chance to change the jms destination at the runtimeAs uses the temporary queues for reply by default the jms producer can always get the response message +Yeap! Great +With Hibernate or higher the Session will automatically reconnect when an operating requires JDBC access during a subsequent request +In case of error the created engine instance should be freed +We are looking for a Read Only user who can only query the confiuguration but wont be able to update any configuration values +maybe +The test that was added in r was not a real tvm test but just an additional test case in the harness package +Now tests work regularly for me +DIH cannot do anything specific for one type of driver +marking as fixed. +Thats an activation that allows people to skip specific module testsAll of the protobufs were moved as they can potentially be pushed across the wire +Committed + +It worked thank you for you +Defer to all open issues to +From communication come out that Toni Rikkola changed the name by mistake and he will revert this change in brms branch so this should be probably fixed in new build +This is unfortunately not trivial to solve since Spring relies on the JDKs Introspector for bean property detection and its exactly that Introspector that determines the accessor methods and apparently isnt quite capable of handling generic overloading there +This patch looks good to on trunk +Im withdrawing this. +Another workaround is to use human readable columnproperty names instead of a mangled mess like SGCOLLTYPE +ESB project wizard seems not to work has new method I have committed the fix by adding default method +please try with last maven to close the loop I confirm this is fixed for my problem project using Maven +Ive just updated of our Maven plugins to use annotations instead of Javadoc ones well see in a couple of days if everything is OkBut from what I saw the generation works Ok now +Fixed in Rev +PMD extensions work but not checkstyle ones +So I want to avoid Polymorphic queries hibernate feature but i couldnt because this bug +Would you be happy to contribute them to project waiving any IP rights to them under the terms and conditions of the license +This change needs to be backed out for now and we need to figure out to solve the side effects first +I went with the full method name split to make it absolutely clear that character type was still the split fallback after camel case. +were backported to as well adding the latter fix version. +Thanks Yan for the fix and Jing for the test! +Thanks for your clarifications concerning the proper usage of XML namespace in FOP +please cross checkthanksmy last consolidated patch was incomplete +If it doesnt work then please provide a test case +Now synced +The worst case should be that you request a meta block and it isnt found +and I will commit it shortly +Sergey thank you for finding this possible to include a regression test in the patch +Otherwise other rules based on may not work +The patch has been successfully applied on Hadoop and branches +Thanks Janel for the patch. +Your approach seems to be better also for save actions +Obsolete Helloworld JSF removed from the list of quickstarts appearing in remove Hello World JSF quickstart also from JBDS +We will probably want to talk about the extra base dependencies this implies also +I committed the latest patch +If you remove it the server willnot be able to parse the request body correctly +I just reassigned the index number of acls maps to the deserialized maximum long from the map +Opinions anyone +Please refer to bug number for details +Sounds like your server is in a different timezone than you think or your locale settings for the server are off +Im also a little skeptical on this being a Curator issueIf you just create does that work with a chroot +Gant cannot execute without knowing where a Groovy installation is so should never execute without GROOVYHOME either being set explicitly or being deduced +Guo Can you post a review request on reviewboard or phabricator? Directions explaining how to do this are on the wiki +The JNI part still needs some work mostly moving a few things around +controller automatic name +It is good idea to add attributes like sorted and optional comparator +Seebq +Yes this makes sense to fix in both places +removed bogus as currently connections open for more than one minute are do the hardcoded timeout value +Ill commit this shortly for +bumped to critical to avoid losing bug +I also killed the menu registered a new one +Without volatile there is a minimal chance of the object referenced by the not being fully initialized as well as the itself not being fully initialized of course +I should have used Searchable instead of Searchable on JDK or later machine +Expresso double to documented in EAP Beta release notes +Are there lots of columns and few lines or lots of lines and few columns? this problem might be related to the number of lines as it is probably trying to do it all in a single transaction and the parsed strings end up sharing the raw character array data in some places but duplicating it in others +IMHO they are a step backward and like this one not needed for the purpose of +The copied was modified to support indexes inI think what should be removed is the main method from +We downloadthe Hadoop tarballs in order to obtain the jars which are referenced when theshims are built in order to satisfy the compiler +These changes are in branches and branchesI sent a note to regarding appropriate attribution needed for our jar files +During my debugging I had completely cleared my +Say Id love to commit this but I seem to be having technical problems on my end +Will works fine at livedemo and will be resolved in general till release +Fourth set of fixes for these issues wasnt rendering components +The only difference is that it has no dependency on additional image filesNote that still needs to be removed as it serves no purpose +It requests a url representing a directory instead thinking its looking for the file. +calling that method from a filter returned the uri plus jsessionid +please use svn diff in the future patching manually is a pain. +Im sure your fix will work to but I should be able to figure out how to make it work right then Ill upload a patch +The generated sort link shows something like httpmypagesortpage +skips non FETCHSUCCESS s +Its the setting of the WSDLOPERATION that seems to be causing the issue +I am now seeing an exception on restart +Good point +Fixed in r + +Can I get perms to create and edit my own page on the hbase wiki? Otherwise I can just attach what I have so far in ugly text format +Integration test checked in and it appears the XAR is produced and is intact +No its not the same values persist just fine outside of the container and the values are legalNotice the connection already null in cleanup no action message thats the issue into legacy component +BULK EDIT These issues are open to be picked up +push out to +Please try a build and see if this addresses your issues +As an attribute of the element +Is it also intended behavior to bind it to the object? That is a breaking change for anyone upgrading from Spring or lesser +Can you apply the update to the branch also +The class restarts the cluster after every method +Hmm +Committed revision +the fonts +Likitha do you mind checking on this +Ive tested with and the delete still fails +hmm i dont see it replacing the scanner in do so this would still be valid right +See Steves comment +If need be I can help with but I dont see anything in the JPA spec that looks like a problem + to master as edebefeaaaebefadcfa and to as bfffbafbfeffaccefb. +m on linuxsearch from root on widnows +It shouldnt be added to schema +However we need to know that this is where it should reside before we can make this change +Its blank because those issues are not assigned +Added an if statement to avoid the ctor of is the correct file +runs great in my environment so I committedit to the trunk as revision Thank you for the contribution +Thanks a ton Markos! was struggling with this for a long timeosgiinstall works for me too +The documentation can have details +Added bindings jar to esbserver zip +Index page and News Archive maybe its better to use H everywhereH has CSS property associated with it which makes all text be in upper case +Ive accepted with minor changes +Inserting a dummy access to the class could be another solution but to make Avis launcher general tool hard coding is not acceptable +While testing it failed on which I think is not affected by this patch +All references to this setting should be removed and we should point out that the XML log format config is the only format supported +that didnt make any sense lets add it as default to the archetype pom though +I dont think oozie or the java builderProcess recognize the for loop syntax in shell +Updated translation for Simplified DavidCould you attach the patch to? Janne had the idea of collecting all the localization patches thereRegardsThe patch has been attached to +This patch fixes the access qualifiers in because JDK has more strict checking hereIn it fixes the conflict with the Map interface that is extended to fix errors about generic type erasure +Attaching a new one +We should try to fix this so folks dont have to use that true to comic mxmlc and +Bulk close of resolved wont fix cant reproduce etc +Thank you for supplying working code + +Changing all primary keys explicitly from NUMERIC to INTEGER seems to have workedHopefully a on ID typeNUMERIC istrue got them all +After some digging I found out that the cant really deal with multiple +Maybe somthing like onlyrootsdirectdependencies or onlydirectdep +I would prefer if it did use the offset and size but as temporary solution you could just validate the offset and size and throw a if it does not match +Werner When you find time please check and commit theEmir +Thanks ToddI put this into trunk and +Made Andreis suggested changes minus integration testing which is being worked on currently +Could you try the patch as wellthanks for running the patchThe client does not seem to differentiate between a closed connection and a timeout +Would it be possible to read the metadata first from there before starting the parsing +I agreed to separated it into three jiras and we fix the first one +Its passed in the recent run +This does it +I think it is fine to leave them out for the sake JME if we document there is no XML support. +Attachment has been added with description Input fo file and associated external has been added with description Pdf showing the has been added with description Pdf showing how it should be after patch is attached file contains an example that shows how it fails thisexample is an edited version of the example in the source tree +this patch will use the api remove function that will make PURGE request handled correctlytested on lastest cleanup please let me know what you think +Ive fixed this issue in test +Lukas you can see it on httplocalhostmetamerfacescomponentsrichTabrich doesnt have one and that attribute isnt present on httplocalhostmetamerfacescomponentsrich to remove issue tracking +Thanks for more simplifications and cleanups Knut! Out of curiosity you have removed a lot of code that uses synchronization now is it measurable +Is the difference significant? The two tests are nearly identical and could probably share more code than they currently do +Ari incrementing a seqId on the file name will work but then you need control over the file refactorings elsewhere +Fixed with suggestion by Bob S +Please commit it +Thanks for submitting the patch Juhani! +The above analysis forTrunk will verify and then see if there is any problem +I have to remove the pluginsfeatures manually but something is left behind that is brking my Eclipse has been running ok until I installed via the update site +I will review the code tonight +Ill update the patch shortly +Simple fix is to recurse as is done in after patch is applied +Hi Lukas of course +So an input output description will be of great the styles are way off +I almost feel like the sync should be enabled by default +That was a configuration problem +Looks that some conclusion should be made or real case test scenario found. + on QA it is done closing. +Explicitly calling persist of the message route being created by addRoute fixed this problem +Best regards by Andrew. +Can you create a review on +This comes from the fact the pdf was created within Word whereas the others were generated using +Ok +Changed Target Version to upon release of +Yes +Looks like the problem was when using Joda +Syslog w Cuanto havent duplicated the problem since I posted the bug +For our this is not a problem as nobody can accidently readwrite the buffer because the reference to it is nulled in our code +Michal can you verify this in EAP CR +class sed s done +When doing queries with a timestamps you maywill want the timestamps to be consistent between the different tables across the machines +I left a TODO to examine in its place +Fails with latest deployment resolved issues for already released versions +Thanks Ashish! +can also be observed on derby so the bug might have crept in earlier +Patch to implement configuration of the +Bugs Features should be reported worked on as new issues. +Thus the reduce task knows aboutk M map ids +I plan to commit them later this afternoon +As a consequence close method is not +Linux flash archive for eServer xSeries x version BIOS update +The issue must have been fixed between and however I didnt see it documented in the Change Log +It would be a big win if our client apps did not have to know about those dont know the maven system well enough to know if this breaks anything else +Applied +Please Verify +bq +The patch looks reasonable +Changes to not delete underlying data on drop partition +Applied all patches except the patch because thisclass is still being worked +I tested both case when the process is under the root or other sub directoryThis transformer produce and empty string and the fix adds instead +Im not sure if we should make the head tag take attributes for this type of stuff +the SVN revision is +Will add a workaround for this specific deployment +No custom code required +Hmmm not sure what this is about but Im going to close it and if you still have issues getting ATS to work at all please email the users mailing list +The same AM cannot have stop and start logs rightThat logs may indicate that MR fails in the starting phase because the staging dir is cleaned by the first completed AM and immediately stops after that +However its only available on Apache and theres a better solution in this case +Thanks Sean the patch was applied to BTI branch at rPlease check that the patch was applied as you expected +please adjust the priority to indicate that the Jira should be considered for CR +Why arent file names sufficient +Hi which object happens to be null? In my current copy a line points to but I dont understand how a class can be null can you explain pleaseAlso returning null there would result in NPE immediately in the next call +As discussed above. +The attributes siterackmachine were moved from subsysteminfinispanXtransportTRANSPORT to subsystemjgroupsstackXtransportTRANSPORT a long while back and the attributes were not removed from the resource description of subsysteminfinispanXtransportTRANSPORT at that timeIll fix this and set up a pull request +I havent put in anything thats difficult to shift around +I consider the issue as fixed. +Customized Grid Pager displaying number of pages items per page number of items and allowing client to select the number of items he wants to display per page +So we need to understand what is different about your configuration +This is a strange requestAs I said earlier these are trivial changes to add a constant to be reused +I have added an explanatory comment and committed the change as r +Thanks Arun! +This is a brand new interface the first pass at it didnt work as well as expected we should fix it now while it is still new. +Same patch without changes +The patch looks fine to me +You are assuming that Column Pruning is happening before map join conversion which is true today but may be a bad assumption in the long termIdeally the optimizations should be completely independent of each otherTherefore both kinds of trees before and after map join conversion should be supported +I wish there were a way to get the entire string over the protocol though +run our test application with facelets +sounds odd +defer all X issues to +That is unless we define a way to transcribe that +Actually I think it was my fault I thought it was unassigned when I assigned it to me +A Solr client library could run the distributed searchmerge and return that to its calling appDo you mean to say that the client library must handle the distributed search? That may not be a good idea +care if i set versionreally hoping to remove these after the lucene release +Can you please attach an example showing the issue +the main site docs also reference the contribs inWhoa OK Ill update that as well! Thanks for the pointers +Closing all resolved issues due to a successful release +Thanks Kim +It means an xml parser was parsing a particular file and came across an element with an attribute type which isnt defined in the tapestry dtd being used to parse that document + file is the file used for deployment +See revision for a simple fix +Hi MartinI agree with this in general as Im using Eclipse also and see all the warningsbut Im in the middle of making a number of changes to the XML side of the showand have been cleaning up all the compiler warnings on the files Im you can hold off on touching the package that would be great andIll do that package myself as I continue to plug along +In this case I think its better to just them and relegate the common setupteardown to shared Before and After methodsI agree though that generally if each test needs a different conf or setup then we should just do it in each test separately +It also happened for me once when I did not have war deployed +Sure that makes sense +The issue here is that when a layer has a default style that is local to the workspace that is not indicated within the style reference +Thanks +I do not feel strongly +Thanks +I would suggest to not advertise them when they are not supported +But this did not help alsoWhat I have noticed is that these errors com in bursts in a row every couple of minutes or so +push to work Ross! +Implementation committed in revision One further simplification would be to remove the constructor that allows wrapping of exceptions but there might be cases where wrapping will still be useful. +I have only one other item to add +Anyone have feedback here? There are some definite benefits to using Guice though theyre not critical as of yet +to trunk! +This one should work +Bean overriding is actually working just fine its the bungled ordering thats the problem. +jar creating a dependency on the commons codec +for the last patch +Yet another change +This problem isnt appearing when the database driver is set to HSQL +I just committed this +The tests testPre are testXmp are fixed +I am not sure if this calls for a revert +Maybe we can open a separate issue for thisYeah Ill open one +Great Nick I will take a look at it +This is a first draft and much remains to do +Its really nice to add new formatting features here +If we adopt a proposal that if a TT fails consecutive tasks from N different jobs then the tasktracker is shutdown +Both the old and new test suites passin my configuration and the new test program looks thorough and completeIm intending to commit this patch +Id like to put a button on the top of mapping viewer and its name is Auto mapping when user click it smooks tools will find out the target node which has a similar name with source node and connect them +How do we use both per service and globally +DIGY! Commit away and for removing andAlso I see this patch as a temporally solution +Attaching test case and patch +But that may be more trouble than it is worth +Heres what Ive got setup +Thanks +What is the issue? That you dont like Millies instead of Millis +Vikas +So I would not leave the comments mangled +Just adding suids doesnt achieve that so far as I can see +Patch for attched +We arent ignoring this just a few things in the queue ahead of you +Definition for provider fixed to used parameter as is in SVN +The tests testPre are testXmp are fixed +if we dont have the ts fix the sleep sounds ok to Matteo test failure seems unrelated +I didnt want change Camel core in last minute +The project folder persists even after being deleted from the Projects view +Nothing to do here since issues will not contain blank comments any longer. +Fixed with revision of release. +There are container system events like which can be used for that +Hi Lev +Will try to do this ASAP +BTW in my previous post n is the of effective bits needed to represent the integer without the sign bit +Forgot to mention that the failing test is in CDI TCK branch but problem is related to CDI spec +Hi Gianfrancothis bug concerned Beta and was fixed back in that release +Agreed but decompiling the distributable shows that the patch hasnt been applied +As far as I can tell the overhead of maintaining a map of string values is sufficiently low compared to the timespace savings +bulk defer minor issues to +We havent seen this impact our scan times +See the installation guide for more details +Which is indicative of a client side abort that has not yet filtered through as a socket closetimeout on the brokerCan you again restate what you want to achieve or what you think the issue or bug is +not it imply making available out of central repository as a versioned artifacts another possibility but I still think the Javadoc plugin needs to be told the class name to load +usrlib usrlib you have a patch file capturing what you did to get the build working +This is not something we can fix in jboss tools its upstream issue +Dont know if this is of any help this is the full trace of the problem +And as the patch is still there and working I think it should be possible with only little effort to apply it officiallyRegards it be possible to add support for making the tag uppercase +examples +Resubmitting patch as is +Ive tried and had no thanks for quick fix and workaround works fine now +As far as I remember originalValues +All instances of localhost in srcresources replaced by NODE +Sorry +Works for me +No idea I probably copied it from an example on the internet if your PR passes just merge it thanks +feel free to it if you have a patchIn the future it may be a good idea to pass ideas through the user groups to get some initial feedback +Manually tested +Example garbage collector and object header definitions +I take back it does seem to do this +studioeclipseplugins +Upgraded to CamelRevision . +Ill call for a RC release later today now that this is taken care of +Tomcat is out +HiIs there an idea about when this fix gets into a stable releaseI see there are betas of velocity tools but not for the velocity engineRonald +Thanks Edward! +My first instinct would be to pass a List of domain names into the constructor +Patch Committed revision an issue in the policy reloading stored procedure +fix to Point committed in r +I got it confused with my local copy which has it +Latest patch +Currently for example checks for duplicated field + Looks good to me +The patch will be updated with that nevertheless +Im no expert on how slink works but Im not really sure how this could have ever worked since it would seem that JSF itself would have called before renderResponse with as the before Seam would have time to render would seem that in order to fix this Pages would have to take into account not the of the but rather the of the page it will be rendering which would seem possible? and Im not really sure why that would have to change if it was working that way before +Ian is working on this right now and will mark this resolved once EJB support is nailed Jopr r. +Add support for column readwrite fragments in HB Core +I am changing it back to byte array +Marking as fixed +We can list the type attribute in add a setter to and use the provided value instead of automatically writing the type out as text +Attached from xenserver +Were using an updated version internally which is working well Ill post an updated patch after the weekend +Demo of snssqs the wiki pages for and +Replaces all patches it works have adapted from your patch +I was about to file this same bug but found this issue in a search +Postponed to beyond RC +Seems like the most minimally invasive way to do this +Can someone please apply the attached simple patches to etcThanks +I wonder ifwickets own feedback panels couldnt be implemented this way instead ofrelying on being explicitly notified +This matches the previously uploaded patch +Will rebase this for alone for now and +Static public fields are excluded from list of valid proposals +I did a few tests and it seems to get us a few extra rows per second +Rebasing on trunk +that reflects the actual configuraiton + +There should be a few coverage samples but not mb ones. +System description +Hi ErikThanksTwo options we might consider I didnt see a for loop and in some cases there could be multiple collations? When debugging seeing the raw suggestions prior to collation filtering can be nice +Hello Maciekcan you tell me how to test it? Or do you have a test case? you test it and close the bug afterwardsThanksJoerg +Thanks Sijie. +It looks like this is reading entire rows into memory +Includes test cases for the new patch without dependency on Java dont remember exactly but I guess we wanted to keep the interface as simple as possible +This patch fixes the problem for me +Actually if it is possible in extension we should close this issue and create one for the itself +xyz +There is a separating FC and FRAG protocols but of course this should be However correcting this made no noticable difference to the application behaviour The bug still exists +Its actually the same in +i gave it a spin and it looks good so i have committed this patch to roller trunk revision +Thanks for doing this Daniel! I think it is not a bad idea to disable secondary key optimization for nested foreach if it is too complex to get it right as you suggested +Since this is not expected to be fixed in time for JDO release the test will be put onto the exclude list +No glossary index and citations are all quite differentGlossary a list of terms with definitions occurences of glossary terms i the text can be linked to the coresponding entry in the glossaryIndex a list of terms with to where they are used in the textCitation a list of sources for more information with links in the text to specific citations for further reading +It is better to have one jira issue per problemAlso it would be great if you could provide more details on your aspx problem +yes applies to trunk fine +I actually had to delete my local copy and then do an update to getthe cvs status right +Marking fixed since the original problem is no longer present +I have recently been working on deployments using both includes and imports without any issues. +In that case shall we leave this jira to and use to fix the misconfigured service and exception logging +Hi Saurabh is just a test case +Please tell us your problem +We will remove checked exceptions soon so this issue will be closed soon +If you need more help or have questions please use the user forum +I just remembered that setting the alignment context is only possible by because the constructors of are either private or protected so the accessibility is limited to the package +This is more like a fully materialized snapshot rather than a true snapshot +it would be great to get this into looks ok compiles and applied +I hate to be anal but to me yes these sorts of things are important +Have you considered including how long the grace period is in the response +Ron +Attached note we may be dealing with badly implemented HTTP server end points in terms of handing Expect request header +Im packaging three +Block recovery tests have been committed to both and the trunk +I have installed erlang on a development server applied the attached patch and can confirm that it prevents the termtobinary snippet above from crashing erlang +Great thanks for the tickets +Suggested fix reset BAOS if larger than N initial buffer size +Leveraging Ant path is smart but I think well quickly run into users not understanding why something work in a place and not in the other +Then they should be readable by user and group but not world +We do in fact have timers running while the app is in the background fwiw +will get a numerical long value and put it there so it does not do that by default +This is by design in unit tests use the mockDomain method otherwise create an integration test +Patricias notes on refactoring +Patch applied in commit + I think that we should be able to turn that assumption off +please point out which of them are zips so I can upload them Carlos I am not seeing the zip files up on the maven or maven sites +For example you can write your own bypass filter and select that in rather than spring having it skip the for you dont want to be secured but direct everything else through it +It is based on the names +Committed code changes to trunk and +Cleaned out broker permissions on the Windows slave +A dependency on SLFJ has the advantage of clearly documenting the logging that a library usesThat said itd be very nice to have it documented somewhere that can log using JUL + push to +Revision . +Hey NicholasPer your second point That would have been the natural thing to do + Final is released +Ill prepare a new patch the reason for these deficiencies is that I worked against trunk just before the generics patches were applied +Thanks +But you need to be careful there as someone could seek to a specific Key which could actually be the first entry in the row +Marking the script file as in error and then leaving the compiled remains is confusing to say the least +Ive had a crack at implementing a fix based on suggestions in +Youre the man Lars +Completely removed +Thanks for the tests cases +Why is the singleton INSTANCE removed from? I thought this would be done as a part of +bq +Currently it is not +This is a significant undertaking +NOTE The datasets are generated by Thomass MR program +bq +Yes I meant causeThe existing multi code fills in a list of results for each op +No changes to the logic how FFIR works removed now redundant +Issue closed with release. +nitpicky but just trying to get up to speed on the codebase and just happened to notice it +Sohan The test failures were a result of some trailing whitespace getting excluded from the updates for and but Im still not exactly sure how this happened. +Hello Edgar have you defined several Sonar schema in your oracle db like sonartest sonarprod +Resolver errors come from the Felix Resolver +However I think most of the code formatting changes will consist of white space manipulation +I suspect Windows will have the same issue as that seems to be the pattern with this issue in the past +I should have added specs in spec unwrapped exception causes a now +I followed from the beginning +Ok tested every parameter given at command line overriddes the value on the profileSo everything works perfectly for me. +Here is a thought if a compression algorithm is able to somehow compress usernames and groups should it also be able to compress common path prefixes +I have tested this on another PC and another PDF and i am still having the same problemI have attached another PDF Another book Wrong extracted f +Pushed upstream. +I guess its plugged to or something +One way to fix this is to only convert from and maybeFlush to +submitted patch for same +Its also interesting is that I dont understand why the flow is resuming anything since this is the first time I hit the flow after Ive started my server with a clean browser +I understand your point +Anton please check it here is some description of Ctrl feature. +Hi TomDo you mean that the change introduced for is causing this issue or +However the still would +Look at for instance which will use the type of the concrete class regardless if the field exists or for the patch but just a note that if we decide to branch out and use other converters unintended side affects may occur +I think you can leave the natural language ref to asideIn terms of module that it affects Id imagine this would be a module of Hibernate or OGM +bq +This is a configuration option in logj +Marking Fix In at subversion revision +Seems like with were almost all the way there then + +Delaying till since we do not have an alternative in patch tests need to be updated +Somehow that got changed to only one when Jacques worked with it +Yes when processing a bbw block report we put bbw replicas into tagets arrays but not in blocksMapBasically with a separate bbw block report add could be made much cleaner in append and a bbw replica can be handled more semantic correctly +What Benjamin Wirtz and myself are referring to is removing this annotation from the class and rebuilding the jar +I think updating our plugin to a more recent version of BND is of paramount importanceDoes anyone have any objection to just directly including the BND source in our project +Verify filtered data behaviorrows with null values doesnt display after filtering +I did a lot of poking around +Per your suggestion I filed for javadoc warnings +If possible then It will be good to consider this in +fixed the way to count checksum in producer JunI just uploaded a patch which reduces sleep time +reporting the new status string in nodeprobe ring would help +Bulk close for +The cleanest way to resolve this is to trim the values as they are supplied such that all subsequent uses will be of trimmed values +Rob please review and if its ok I will commit it to trunk +btw maybe the features should be sorted A +Fixes the problem for me +Bernd made a request for this issue in BPMN the swimlane is stateless while in jBPM the swimlane was use cases Ive seen didnt need to memorize the swimlane which means that every new tasks is put in the pool for a certain group +Should there be a way for the client of this class to specify the prefix and suffix charIs having for example h as the first token really the right thing to do? Would he make more sense? I know that makes it characters long but its chars from the input string +Its pretty ugly but works fine +My attached patch changes this. +To resolve how about To start the installer the installer file or open a command line and enter +I have also run the lang Suite against both sets of jars +Right now for we wanted to fix this error as it was bocking. +Id rather see the subscriptions dropped and no further queuing but maybe that wasnt doable +Sorry for the bullshit +It was because there was a space created in the latest patch in the testcase +Can you please rebase on current trunk +Is anyone working on this +This must be fixed in a CR +Milind have you looked at the latest patch +There may be something wrong with my for your help anyway +it does not occur any more +patch store attributes with duplicate names in list for freeing later on +thanks for the explanations +It appears that the design calls for no error message so our curent behavior is fine +This is due me not handling the case of application clients where the local receivers arent present +I examined the three failures notedThe two hdfsproxy issues dont seem to have anything to do with this patch nor the associated patch forThe issue also doesnt and the noted failure Too many open files does not occur when I run it locally even on a very small boxRequest approval of this patch despite these unit test failures since they appear unrelated +The final fields values would have to bechanged during specific life cycle events such as retrieval from thedatastore flush commit rollback etc +Thanks Gopal for digging in +jdb session of a running namenode which receives a kill datahadoop dies automagically after a whileIt seems as if the problem is related to some socketipcrpc things +I committed this change as well +core module svn module svn module svn module svn on various modifications done outside of the jonas container impl done on jonas The impl features a remote deployer based on JMX remoting or MEJB an installed local deployer a runtime local existing local standalone looong ago +Could you name patches like? Because log suffix is confusing +To run the code above to reproduce this issue put from the attachment to the current directory +patch applied txs gerhard +A duplicate of this was filed against nhibernate as +is still proprietary but I have contacts in IBM that might be able to do this +Actually if we think that not clearing the selection on mouse down when the item is already selected makes sense I could see moving this back up to +Thanks SaschaI will double check and maybe also contact Jonatan directly +No problem since this is a new plugin that nobody uses yet +Both compiled and ran the tests fine +This is the failing test where the operation isnt is a modified version of the broken test that uses an eip component to add the operation name during the flow between the camel and cxf components +yes you should force users to define the value in their and cannot provide a default value in pom for those that didnt do their configThis is exactly what we do but it does no longer work with anything newer than. +Making this issue depend on as the current implementation uses wildcard highlighting +Alternatively you could use to add the annotation to the class as it ignores Target +and were changed in Remoting so that if they are configured with version they will turn on connection checking by was also an incompatibility in the rmi transport +Is the PDF still generate in the Lucas buildWhen I opened this issue I looked at the PDF but thought it is the generate one +Moved from to just after merge of old branch into TRUNK +Could you try again tomorrow or building camel yourselfFWIW Im actively working on the blueprint stuff to bring it on par with the spring support so let me know if you have any more problems +Could you say a Continuum window where I can see something similar +bq +The test case has an incorrect option deliveryDelay which should be redeliveryDelay +Yes it takes time and requires generally making a nuisance of yourself but it can be done at least I got a handful of Oracle app server bugs fixed that way +If he wants to do a mv after committing great +Fixed with revision + Username dsundstrom +ok Ive reproduced the problem using the jdk that Dan the file from didnt help anything +I will check this locally in my pom file +You would need to add extra logic to the parser to detect the merged cells and output the appropriate rowcol span attributes if detectedIf you do get this working please upload a patch! +Experimental patch Mark! Im looking into the issue +I think you may have the versions in JIRA wrong? It currently says Affect Version and Fix in Version +Im no specialist in that particular area but both your last two comments makeme suspect that the real problem lies somewhere in Ant or rather its javac taskAFAIK GCJ can compile to +I was trying to make the changes and found some other problemsThe is only available in ql while is in common right now +Corrected few places where files were not closed properly +Found an error in reduced javascript +Moving to one knows what this means +see my reply to in for details +Railo is running on a Centos box +A couple of things that might be interesting to take into account while thinking about this issue As noted in the Part stuff is actually not part of the public API +Tested on +Well this issue may cause problems with other JS event handlers too +Start contributing to Geronimo and thats it youre done subscribe to devuser mailing list read the code respond to user questions contribute patches solve JIRA issues simply be active and before Christmas Eve your wish becomes true +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +That being said I know there are probably other places in the app where we display a stack trace for some sort of known error condition +Fine Im currently working on removing all the me errors +I pushed part of this but there are still two outstanding pull requests that need to be processed +If you are seeing a +ps my hadoop version is +update configurations to use database schema configured in step +Thanks a lotIs there a possibility that it makes it into the next official release +Semaphore is not the right approach +Looks good thanks! Ill get it included still in. +The GSG needs to be updated with the newly supported the XSD and WSDL sequencers to the standard configuration +updates comments derp derp +Are you seeing failures? The reason why these error messages appear for some jobs and not for others is primarily got to do with the size of map outputs and the heap size allocated for the reducer +It was a bad idea +Now the unit test gives an +This patch blocks issue and this commit will help me in submitting a final patch to that issue +I dont think its a huge API change at all especially the solution Dave has put forth +Unsure how I can add some tests for subinterfaces even with mockito + +Looking at the cfg if a chaining cache store is used the sharedunshared flag applies to all cache stores in the chain +Its a Xalan problem we wont fix it here +I vote against a line len limit because I have never like how eclipse automatically wraps things +Since closing a file removes it from the map there is no need to clear the map at the end +Gold files have been committed +Attached patch to fix the bug +JCR folder and the +Thanks again DennisPasting the original Release Note draft here for referenceCause processes requests for HAJNDI while its being shut downConsequence Client can get errors from components that are not in the correct stateFix now throws if a request comes in during shutdown and the shutdown waits for existing requests to completeResult A known exception is thrown for this use case which causes an automatic retry in the client HAJNDI. +Ill be working on that this morning updating tests as I fix stuffYou can keep tabs on the lazy branch if you want to be on these changespatches as I go +Attachment has been added with description sample fo and has been added with description Input FOour local BW AFP Printer handles this and converts the colour SVG to BW on the fly +Verified +Patch looks great Dawid! I like the blue expanded node dot coloring +I have provided the tests separately in Common itself along with nowThanks +that patch looks good to me +Have fun +Deferred to close of resolved issues. +Going to resolve this as wont fix for now Since gems in jar files as somewhat dynamic and rubygems only builds the gem path when its loaded Im inclined to say that you need to ensure all gem jars are already loaded or on the classpath before you require rubygems +Sounds great to me +Thanks Glenn your patch looked fine to me so Ive committed it +I think cliProperty would be more appropriate as system property use is being eradicated in the core as it makes embedding not work +This is mostly a matter of documentation +A much better idea is to centralize this information and encapsulate it in a seperate configuration file +Yes when the call is removed +reopen to add fix version +You should always target highest +Ill commit this +Committed revision I keep this open for further discussing +If its not accepted the conversion will be a waste of time +Although this fix is quite simple and straightforward I still could not be convinced using this solution +I will be reviewing this and the UGI changes today. +Paulex thanks for applying my patchIssue is fixed as expected +I am using it in the example and it works fine there +Seems did not capture and store any of the Table request variables such as page number sort column name etcThis patch also tries to sort enable the control so that it will work correctly when sorting on different columns +Attaching and a more concise reproduction of the issues in question +Some introductory material would be nice +Sorry missed this +It would be much more efficient to use in memory streams +This does happen on Android in most cases +q file got left behind and needs to be updated +is included in this patch by accident +I just committed this +This work fine but we release that it is not the best way if we plan to move to another you are interested we can attach the any patch is welcome. +Maybe it could be standardized on some bigger value +Need to make sure it is invoked after optimization in both old and new logical plans +It initializes a too long readParams array Therefore the last params of the array are always null +Its unclear to me whether its an incompatibility with Java and Gradle Scala Junit or some combination +Ill move the lists after movedI think the only thing remaining is that the mail archives arent appearing in modmboxThis isnt completely resolved yet +htaccess has been excluded from the rsync with the option exclude +Would you please reopen this issue so the misleading warning message can be addressed for buildsNo I checked with some other Maven developers and we agreed that the warning as produced in Maven is fine + +Resolving +I did this without need to using dynamic casts by limiting the nodes to those that are owned by the current document +This isnt going to be fixed any time soon and I dont want broken Android behaviour to hold up our release +All passed on both OS +The new patch will also eliminate the need to modify the s read static methods +No need for separate QA so skipping QA Task +Fixed in the following files in luni at repo revision javalang javalang javautil +Sure in magic fairy land Id love that too but the question here is what can we improve for +Thanks lars +Default interval is now day in second unit with validation +Lets put wont fix +I changed the issue description because the TLD should use version of JSP +However that is a separate issue as compared to the problem related in this JIRA +In branch release +See in connector directory +The needs a redesign but my understanding of the concepts of Cocoon is not deep enough to do such a work I thinkSorry +Right that would make sense to me +I think it depends on what you are doing +Moving to the correct component +At the moment the user must try each one just to see that it forgot to add one driver +This should make integration of easier since I can rebase it on top of the current trunk to get a working AE instead of having to do it in two steps. + classify the we know the status of this please +Still a unresolved issue in + in +Let me fix it +Note that version also allowed HTML markup which was not in a CDATA sectionFor me HTML markup only works in a CDATA section even with version s an attempt to objectively summarize the comments on this issue so far +Ive committed this and merged to Thank you Jason! +we miss a validator to prevent that kind of invalid names +I also started developing in production mode to avoid a proliferation of images css and js files in their caches +pending Jenkins +Other small It wont require anymore changes +Is this what you were trying to address +You should probably copy them on top of the existing files to get the encoding right +Test case to reproduce the you for the detailed bug report and the test caseThe reason why this happens is the same as reported in. +From my experience those refactorings tend to loose some bits +No please leave open +I love how this patch also speeds up operations like table disables region reassignments etc +Chris This has been a MUCH larger task to tackle than I had originally anticipatedWhich is why I suggested we continue to the existing Java classesFirst step is to clean up the codes in those classes +I dont think the is complex enough to dedicate a subchapter of its own for it +Will commit when tests pass +Record IO generated test files which were compiled everytime +I think this would really put Roo on a very strong footing +There is no update on these items and are in open state if they are not ready for please move them out to Future +Checked Crash and Seen in production +love the added test +Does anyone know what this warning is supposed to tell usOne small concern I have with the patch is that Im not sure it preserves the intention of the runtime statistics tests +Fixed in trunk could you please give it a try +This JIRA issue has been open for a while and its one of the last task remaining to be done before I can open the main web page to the public at ASF for Regards George Aroush +Ive seen this work on chrome in linux and osx as a data point +Feel free to do the change or wait for me to come back from PTO in about weeks +Thanks! +Can I specify the local ant path during configure? Please let me know if there is a way to do itThanksSantosh +Sending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesenginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesenginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision enginesSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision bindingsSending Transmitting file dataCommitted revision Thanks a lot for this patch Jon! +Quoting frombq +Moving this one to its not as critical now that isnt going to be part of +SOLUTION Modified the wsdl service generation tool to produce an application descriptor object +Fixed in rev and updated snapshot deployed. +may be just a result of the problems in but Im not sure. +This one replaces the previous +I move the gapstopper to its own jira link it and move this issue to beta +studioeclipsefeatures +Closing all tickets +Indeed emma shows that Properties handling in is not tested +Yeah makes sense to me +Yes I have found service with name createContact in sfaI did few changes in this service to make it more generic and made changes at other places wherever requiredThanks Ashish Sumit and Rishi +Over all the region servers would self fix this problem by default over timeSo only time this would be an issue is on a restart or after a failed region server +Created course there will be annotation processing for JSF beans but this is done before +looks good to me. +Fixed with r +What about creating the at teststartup out of the +We cant remove some of the things deprecated in or +I believe this patch fixes both issues and I dont have a strong opinion as to follow the RI here or not +We no longer publish a Java jar as such we do not need to fix this issue. +It is very annoying not being able to paste into an irb console session. +Patch from Billy improving javadocs +Thanks too for committing it +I suggest to add this as a known issue +Can you clue us in on your thinking? I mean rollback is a traditional function of database transactions which is what a sequence of uncommitted updates followed by a commit seems to be to meAnd if it is removed from Solr what workaround will you suggest for situations where a user needs or wants to roll back latest uncommitted rangesIn any case it would be nice to hear more of a discussion about what exactly is broken and why it cant be readily fixedAlso if it is marked as a major bug then maybe the feature is more useful than you suggest +This is going to hamper Solrs existing strategy for supporting solr +Sebb we do not set user mailing lists to subscription moderated +This patch switches the normal explain statement to use my walk tree implementation without introducing new tokensSomebody please mark this as patch available +He thought that I could not +And it would affect the basic functionality of password enabled VM so QA must test this +Please your scenario and let me know if that still didnt fix your issue +Marking closed. +Can we resolve and include this in +Proxied Repository page no longer exists. +Names need to be consistent +could Writer has namesystem as a field +tcrawley chaps my hide. +Pushing off to release +Any opinionsAlternatively one could submit some patch to Rake itself but again not clear what that could be +No tabs +Havent looked fully at the patch but one thing I am curious about is why remove the Fieldable interface +If the column is missing in a current fresh install does that mean we dont need this update or that the column needs to be added and the alter table was missed +As Jing pointed out offline the for loop covers the if check +If i find correct reason +Ok since youre ok with it ill go with the simpler implementation +Committed to trunk +Fixed several problems and extended tests to cover imports of services and inherited services. +I can not thank you enough! You are saving my sanityProblem is closedBest regards Martin +The unified meta data model lives in module server resolved issues for already released versions +Moving as unlikely to be finished by because sitemesh will not be out before then +In users have to explicitly send mkdir p +Ill port the changes to thenThe fix also works for jxformatDate as well as jxformatNumber +Nitin can you have a look at this? I believe you worked on this featureThanks a UI bug and so please assign it to Jessica can you take a look at this one now? Im not too sure of the best way to determine if the template is linux vs windows to conditionally display the message +Thanks all for the review and discussion. +But are not being handled as Beans anyway +Anita Ideally wed only provide one since two sets of checksums really noises up the download locations +I checked the source code on trunk and cannot find any changes regarding this issue so it is likely not resolved yet +Made the following changes to ensure proper behavior when a process is killed use an exception holder for the initexception in the adminapi servicedata fix the jgroups remote method call exception handling and message remove all children on a process add so that a new process after a kill does not keep the stale entries of the old make only local updates to the registry to ensure that only authoritative changes are sent across the cluster process polling needs to catch messaging exceptionsWe still however need to check for duplicate hostcontrollerprocess launches to fully address this defect +Still pick PIGCLASSPATH first so user can override hadoop conf dir +IanCan you post the endpoint URL you use eg the complete endpoint uri with all parameters etc +I may have overlooked a couple of things or looked at an older version +That way wed never have them out of sync again +If you need that functionality modify + +I am seeing the exact same error any work around or anything that might be done to alleviate the problem +Can you provide an additional patch that updates the test case +looks awesome +Is it that is using the same flag for multiple disjoint purposes? If so then a new flag should be introduced to cover the second caseIf cannot then isnt it broken +An is created with a timestamp carried by an initial PING message and it will discard the contents of an update PING message if its timestamp is older than the previously processed the initiating PING message does not have a time stamp then the lease will assume that the client is using an older version of Remoting and it will accept all updating PING tests has been added to the versioning test suite +A patch for the issue +The endorsed mechanism was designed for overriding classes which have the same name as those already in the JDK + committed to trunk. +no problems there +Thanks Russell! +Bumping to LATER +The updateTuple code seems to batch though +Verified in Fedora +We combine them in afterwards +If you have any further followup feel free to comment here and well address it. +I will add Apache License and reformat in next patch +The HSQLDB implementation required HSQLDB improvements to make it possible +Ill close this issue +class file is no better than a file that does not have the +Here is a full from a run with diffs in updatelocks due to locks in background by the stat daemon +We can use as a separator +Logged In YES useridCan you attach a sample war that shows this on redeploy +But after noticing it I spent the afternoon creating a new partition and moving my source code there +Ive created to track the issue +I have a fix and a testcase for thisCan you take a look +I will take a look on the other Jira +This issue is about uploading to the central s not just an issue for +It looks like it even contains a spec which is awesome +Please check that the bug is fixed for you +Nobody is trying to take over the mark and I think thats obviousI do not favor having it because it is not essential cassandra code +In the that I ported for we had a filter property that I used for this +For merging from a branch the three s cannot be cast from any of the committers who worked on the branchDoes it make sense? I will post proposed change +YARN is for graceful decommission which could tolerate longer waiting time +I hope I can receive notifications from now onThis patch is to show you the idea +Nothing really actionable here right now +But this patch has too much code duplication +This appears to be different than since I dont see any exceptions in the test output +So I had to comment out action are few things that need to be done move those into Run As +Then you wouldnt have to extend +In my case it never finds the gemI find it a bit risky to call the gems index for real on different platforms different situations this stuff might break +Here is the updated patch for this issue +Maybe I could check in my code and you could take a critical look at it Id be happy to take a look as Id like to learn more about the EJC code but I dont think Ill be able to give you any advice as you know this much better then me +Ronald had already fixed this issue on Revision and he didnt tell anybody about still need a testcase to verify the scenario +fyi for future if you name the patch the same name JIRA will take care of it making the most recent patch active and the older patches of the same name greyd out +All tests pass +Fixed one more failed unit test +This is my first use of it +I created a it myselfAttached a need to include test cases will a patch +Makes sense +Net clients which were removed from the repository months ago and not the current +The major problem is that you are determining the type of the parameter based on the value +its up to you +At least to coref stuff should be applied after we have proper training and testing over there +Yes +Strange that it works on the desktopI have created to provide a responsive template for MetamerClosing this one. +Excellent +Ive also made the subclass a class so that people can specifically test for you please take a look at the current SVN and see if everything is in orderCheers old issues +ear not +LGTM +for Dougs two comments +is right out the door +Let me know if it is not reproducible +The same applies to example +Weve got over session beans and tracking down the exact dependencies that doesnt like takes a long problem is that doesnt work when you have EJB references +It has not been properly yet so cant promise for + +Thanks Aaron! +I will submit a patch shortly +Do you mean always print out the full trace when print is called? I suppose that would be fine since most people will expect to see a normal Java trace for a Java print +etc +for as well +good suggestion at least to make tests that is not realted to validation run +Latest patch to support encrypted header +It retries at most a configurable times using a fixed interval when receivingIt also modifies the package a litle bit +Nick this feature is important for the look of extendeddatatable component cant reflect the capacities it you do something pleaseI have to say sorry but this issue isnt planned to be resolved in release + +The new patch preserves the change done in the first one and makes two additional changes It modifies the queue comparator so that queues which do not have a Guaranteed capacity come to the end of the list of queues No jobs are picked from a queue that does not have a guaranteed capacityIt adds a test case that tests that GC is correctly computed and jobs are not assigned if are zero for a queueVivek when the cluster is starting up and are reporting in based on the percentage GC for a queue could be zero +Thanks Sandy +But we are notCan we close this issue + +Might be useful to add it to reduce the hassle for new contributors +I will be doing my own testing and put here if I found something that its not working properly +That seems reasonable +does not have a formal automated tests suite available for it yet +A test and some comments and typos are modified +Milos doesnt want to bind to a specific version of the EAR plugin for example because he was a generic JEE module in Netbeans +Thanks +Packet dump from the issue. +Saving changes to a connection pool where the value of at least one attribute is null triggers a NPE on the background timer thread that rewrites +OK tests passed +add a mandatory source field that is the prefix in the log +We have not tried I tried psql that failed but others may work +before deploying +will be not supported on +True will update +Please find and screenshots with attachment + +But it still might be worth trying here too +Performance is an issue +So Im not sure there is a down side this yet. +You can change it any way you want +Username cuplan + +I would appreciate all suggestions how to improve this situation +Note that are not tested for file protocol +no there are problems involving compoundfilewriter as well +now release note status has been set +Logged In YES useridAdrian Thanks +The extent of the diff may scary someone just looking at the size of the diffs +Next time I get to a machine that I need to reboot Ill put it in DEBUG +The code to produce them is also more concise +Agreed with Jakob we should put it in the documentation +The master node would be the NFS server and the worker nodes would be NFS clients +Yes it is a problem with the secured deployments +Werner has been debugging his AJAX code +OK i just try to create PO on demo Trunck and demo is empty on Trunck is NOT empty on So i will create a new JIRA on this regression and close this one if you agree with mePierre +Is it not workingYou could also try passing distribfalse to manually prevent distributed search from kicking in. +added as a client property that can be set via the url datasource or with a set statement for CR +Verifiable by setting to SBAP and turning auth on before running hcat e show databases +Using a separate thread to handle viewAccepted seems to me not a simple solution +This changes collection implementation classes to use schema properties too +The handlers esp +My client still gets exception +Its a good practice to attach stuff directly in Jira issues for online pastes do not last long. +Army I think the metadata queries get compiled in SYS schema and that is why the metadata queries dont fail for parameters since parameters will take the collation of SYS schema +I see it in also +committed w private change +Thanks a lot for the patch +patch which fixed the Ive rework Maxims patch a little to get rid of update junits related with new to be clear this patch just seem to open up internal apis thus its not a guarantee for future versions to maintain itFor that it needs proper separation of API and implementation +Rich and I discussed various approaches over IM and modifying the pageflow context design at this point seems unsafe +Thx Luke +Yeah maybe we should get the AMQ people to see if they think its a good idea to have this baked into the vm cursor + +What I do see is that there are some error markers on the generated +applied +This patch would use the above pure Java pipe implementation on Windows but continue to use the native Pipe implementations on other operating systems where there are no issues +Committed to +When do we want to restart the module for the user? Itd be inefficient to restart the module after every incremental publish because then every time the user changes a jsp or html file the module must essentially +The interesting was that the declaration was inside a private method +Relates to +Dont be sorry it is great to have someone checking on these issues +please let me know if you need any help with this task +Incompatible behavior of the stack size option may be another issue but of less priority than AWT one +numeric task id d +Thanks Cheng. +Yes +I apologize for delaying reply +Henri Can you please attach the relevant parts of this patch to to fix the JMX issues +better patch now also shared registry among sinks that use the same morphlineFile and some more metrics +I moved our CSS to top position just below due to performance because its the best to have CSS as first elements and then Javascripts +IF we are making a MR specific change then we might as well ignore all deprecated warnings using Xlintdeprecation +Came to this conclusion after trying to debug touchdown vs touchup in Automation. +Is there a concrete example for this issue +The reason for this move is that the schedule for releasing the final versions of the Java EE specification are still unclear +Thanks Miguel but query criterias by are more dynamically then HQL queryMy applications have criterias by only and run correctly with two changed classes in first CLASSPATH +So I am removing these useless tests rather than spend any time tracking down the instability +In the case of cleanup compactions we could utilize the compaction range and patch from to determine a more accurate size for sstables which should allow us to complete cleanup the majority of the timeEDIT isnt quite ready but you get the idea. +However I still have data binding problems which I have not been able to trace +Andrei cool thanks +Thanks +The problem however is that since A modified its own state A is also free to wipe that state out either by restarting or simple remove another token because theres only space for one +sample document showing formula use not being parsed. +if you want to look for a better one just reopen this is not working again due to dispatcher restructuring +dont you want to encourage people to move to? My project works on SWF and if this isnt fixed Ill probably keep it there +Nick yes I saw it thanks +Lets use it as first step +done +Hi I need this optionA job failed in our production environment because the file which was used as distributed cache was modified at fixed intervals + +And what about following Eclipse versioningSo we remove juno from and and we use and so wed see orgjbosstoolsmultiple orgjbosstoolsunified orgjbosstoolsmultiple orgjbosstoolsunified +I think this discussion rather belongs to +I will include the project a RossenI think we can close this issue +I will close this jira +Committed +Oh and its brilliant idea +It would be unable to find the proper jmx classes and would error out +Fixed by SVN revision worries Ill try and set aside some time next week to check it out +Why would purge implementation based on a periodically executing job be superior to using? At least to me the latter sounds cleaner and simpler +will hold working on this issue until Abdel Adridi provides some guidance on how to run We have another task on the roadmap for soak tests anyway +that way all extractors can take advantage of it +According to Jacopo it needs a new patch +Ive already taken care of this +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +See explanation above +Kevin if you take as default then you risk to get into deep troubles +Has anyone a first cut on top of from the proposed interfaces above +The modification looks pretty clean +And if they are proper attributes are they stable and supportable for many yearsI realize this last question argues in an opposite direction from my previous question +This means that a database may contain passwords that are hashed using different algorithms +reopened just to set the resolution to fixed +Temporarily reopening to update release note info. +rsync is a great solution but it is manual +mrepositorygwtupload everything was ok. +Thx +Intial fix for the removal of the uac and pab directories +Patch applied to trunk +However this would give a fix relaease of +Basically there are now different avvailable all extendingHowever the UI should only work on correct +Depended on by +Thanks +How can we avoid removing imports for this call when inside a groovy compilation unit +Attach patch for the described is not shared across therefore this issue is blocked by +In addition Ill submit an issue to SF +Attaching +The WC includes a report that tells us if the actual results are equal to the expected results +Thanks Hairong +Okay well hopefully just having minutes or more between balances should suffice +attached for discussion +Should be possible to then do compare that is not susceptible to changed ordering just because startkey contains delimiterMight have to have a version for root and another for meta given that the root has rows made of the meta tables rows +Thanks for the review +Andrew Sorry I hasnt tried to reproduce this issue on tomcat the following in the Archiva docs in ro instructions in setting and in Windowso more troubleshooting tips for startup errors +Sorry guys I have broken the build for jdk runsAm handling it now +Gabriel I can confirm that it works on Linux thank you very much +In this case the error message is telling you that the only configurations you have are applicant and patent +I believe view compaction is also an issue to be handled + +For relational operators that require multiple inputs the list will correspond to each of its inputs +In case should we just try another region server instead of waiting till it is up? I think we should do our best to use the same region server at first because it could be a region server requested by the user or there is some locality issue +cool! I was scanning through it but just escaped this sentenceThanks +It looks like something went very bad lot of namespaces are missed ie +Was the archives from projects merged to the ? Where can i find the archives? +TODO merge w +Would you mind asking the question on the user lists working as expected +Forgot that was done still not sure if that should be done I guess many users like seeing familiar linefeed +If you could respond with some specific instructions to force this issue to happen thatd be the server is started a listener is added to the process so that if for any reason the process is blown away the server immediately changes to stopped and the poller should be canceled +Works on Mac OS X mysteriously +Fixed in both HEAD and maintenance branch CVS +Master already has plain text auth along with sha +Renamed ticket to make actual change more obvious. +Works fine for trunk +thanx for quick fix +were backported to as well adding the latter fix version. +I am changing that test and will check that in here +Sure please update the summary and description of this JIRA +Sorry to come in late the patch has gone stale +sysctl A changed to use HWMEMSIZE this should properly report on your machine. +Since its also using JS it could just be handled by the JS client but then youre only getting in app notifications while its open +Awesome! Good jobThank you for working on this with me +This is why we are not using maven for Penrose +Thanks Arun DiveshArun your patch is in trunk rev +Im going to resolve this as Cannot Reproduce if servolux causes issues in the future we will revisit it. +test are you not able to reproduce it now +Tyler At this point theres no plan to release a +bug is really annoying +It really needs to be refactored to be unit testable ASAP +do we speak of a webservice call or wddx call +Committed the patch and marking as resolved. +Alex Youre right I hadnt looked into setTypeRakesh If you check the first comment of this jira in the sample config it sounds like Alex is proposing this new pattern hostnameportporttype +at least we could get some experience with it and a better understanding for what it can be used and for what not +if the packaging to +test +Luis Did you ever get to investigate this on Sun JDK with debugging symbolsDylan Looking forward to any more information you can provide +Patch is in CVS +Ive asked a colleague with a test setup to check if he sees the same thingThe reason I asked Ahmet to check if indexing is actually taking place is because I suspect that these errors are in fact warnings because the connector does not even use the methods in question +As a cleanup step we should look at cleaning up methods like get from the superclass and have the new subclass deal with collation +Checked in the fix rev for branch +I have applied this patch and have been using it since +This portlet should be modified to support JEE changes etc +Yes Id like this in +But couldnt one make the counter argument that by forcing snapshots to the same volume as the live data you could end up in one of two scenarios Your snapshotted data has eaten up so much of the existing volume that cassandra fails as it tries to add more data because there is no more disk left +We check that project is synchronized at least for one level before loading or rejecting JSF nature +Id like to see authorization based replication +I succesfully reproduced on Ubuntu bit and +Dag the following JIRA is probably relevant Kristian +can be looked if you opened a new enhancement request and stated your precise requirements +I think you misunderstood me +Perhaps the jar used in a is not compatible with Java +Can the severity of this be bumped up if possible +bq +I havnt tested this with a JSP page however I have integrated it into my freemarker forms library and it works a treat +resolved in r +But it wasnt complete after hours +It looks like this was never fully done +I do not see why this is related to the foreach issue +I will do some experimenting when I have access to my OS X boxI am thinking of using find and xargs +I enjoy colorful ticket names like spilling like its as much as anyone but I think we should rope it in a bit +It might help if the program dumped more timing information along with print outsFor the timeout case ignoring that it should be a deadlock it seems like we should see a timeout after seconds for Thread and then a timeout for Thread after seconds but it looks like they both timeout at the same time +And if you want to simulate failures with replication youll want an additional slave beyond . +Should we change this +Do you mean what the current head is of a branch? Id say its simply the highest revId of the commits for a given branch +But I dontknow how much effort it is to add and evaluate that flag therecheers OlegThis looks pretty good + +But if it is at least configurable Im happy. +Thanks Swati! +I just pushed the final version of. +But why would that not be a bug? The garbage collection seems to stop working for SFSB in session scope which holds entities if the session timeout is set higher than minutes +I havent had a chance to really play with this lately can you give an example of the query you are running +This change was in the new importer code but has now been incorporated into the current importer code as well +Hard to tell what you approach was from just the patch +I just assumed that the reason that this hadnt been resolved was due to the Maven Embedder +Edison said he will take a look at the root cause +unit tests fixed +The patch adds a known configuration problem to the README file +QuickstartWatch +It depends on patches and which must all have been applied before applying this patch +Example of the x Knight covering Chess problem +sure replacing Fieldable is good just noticed quick win when deprecations from our code base +bulk defer of open issues to +Rescheduling to ALPHA since ALPHA is supposed to be released could you let me know the progress of this issue? I didnt see recently +Im not too keen on this +The thing I dont understand is the following We generate Java objects from this XSD use these objects with a via its setter and then when marshalling castor complains because the milliseconds are +Do you have the security domain configuration for +It is a indexing system +And you can give it a try +Here is what I propose How about moving the depends into antcalls inside the? by doing so the autoreconf wont get triggered if is not setFinally if we think target should be controlled with flag then we should mark this jira incompatible At least for me I do ant and expect it to build +Only issues that are actually fixed should have a fix release. +NP! Thanks for the nice patch +Applied Patch and test casethanks out patch and test cases since this breaks the WSDL for doclit Object arrays and causes to throw warnings about Object types when it shouldnt +If reading if cant find the filter file continue with warning +I found the line number from other parser are not matching the qdox numbers +The same problem occurs in when using this interceptor with an EJB stateless session bean +Yonik I isolated a testcase which fails very quickly +interesting +I understand Eugene but I really dont see much of a difference in the Project Explorer oder the Java EE perspective +It is still not totally concurrent but the procedure meant to be used to allow concurrent access to the table and compressing the table at same time isThe COMPRESSTABLE function guarantees best compression but will always do so by getting exclusive access to the table and internally creating brand new base tables and indexes and loadingthose from the old table while holding the exclusive lock +Ok well +For now focus will be permutations that return timestampdata always +Lets not break any API in the +I think we can close this one +You really only need to drop relocation poms in the old location which we could help do theres no moving of artifacts +Using these errors are filling up my logs making it harder to diagnose issues when they do happen +Check out libjavaSince theres no overhead I suggest including the Scalac compiler by default and not having to require it explicitly +All of the old functionality should still work it is just that more memory will be released to be collected when a call finishes +Having a particular user skin change reflect globally across to all users creates a major support issue +There is already a setting for this not exactly this but nearby +It is different from mine +Are you available to test a snapshot and verify the results when I do commit a change here Stephen +This happens as well with intermediate nodes but those cases are not extreme since references to the new nodes can always be retrieved if there is a reference to the DOM root node +Removing the complain from +Thanks Mamta for looking into this Im still concerned that the object has the ability to change its hashcode and its worrying when you saySince the earlier entries in hashtable is step above were based on the hashcode of tableUUID and grantee we will not find the entry in the hash table based on the UUID +Otherwise well close it with a REJECTED +Attaching patch a which moves assert to a better home +Ive installed the patch and it works for me +Mapper +patch committed svn revision seems that a fair amount of work has been done in this area +Re checking logs readers are to check for events of Warning andor Error +What do you think +Initially I also doubted the same +sufix could have latters such as +Also code was changed recently to check null and the commit message says it has done it to avoid possible +Updating the title to describe the change being committed since it does not completely remove the dependency on deployed MR jars +Ive committed it +For some reason the patch on did not apply cleanly for meShould we just commit this instead? It should be the same change +would help you since the downstream message is constructed form the object that was returned by the transformer +Perhaps you can clean it up a bit and reduce the amount of things going on here tests takes minutes or more on my machine +fixed in w the release. +Please check it out and confirm that thisaddresses your issue +So I think this is fixed with the latest launchpad release +unlinked would be better +okSounds like a bug dubious development flaw purported as a limitation or at least a doc bug +It happens quite often actually + the following text has not been fixedbq + Its a bug +This issue did not break the build in and the exception was the same one as +New embedder should include fix for this issue but I havent got chance to test it with project yet +Patch verifiedThanks Regis! +other than ads +To cope with the existing code shouldnt be public interface extends Serializable +this will be done by arch figure +Thanks for pointing this out Ive addressed this in the same style as for bean classes with the target type names resolved lazily if no bean class loader +I havent yet opened a ticket for this behavior +It is not envisioned to change in the near future so Im closing this out as wont fix. +Thanks a lot Eli +I tried the steps but still cannot reproduce with plain JSF +could someone please remove it +If you have the space in path issue on Windows install Java in a different place +Resolved as duplicate. +. +the createFilter works fineSorry for the bug opened but Ive not found information on changes on this behaviour in the rc and on the hibernate forum emmanuel told me to post this here! +I do not understand +But we have seen issues in the past +I checked that this bug was fixed in version. +I wont remove any s from the class +I was thinking more about what you can do now +Close the parent task track the subtasks. +The engine will use this to enhance performance. +Removing the calls to get still results in a reverse DNS lookup so theres probably more than just that one call to deal with +Locking error investigated and filed as a separate issue +Assigning myself to this issue since I submitted most of the url patchesApart from adding junit tests for the patches the only work left on this issue is to enable the failover command on the slave side +Thanks a lot for reporting it and giving a sample codeThe fix is in the CVS +Ive gotta be honest I dont think the proxy feature should even exist +Not a problem of Commons Configuration. +Ill commit this feature in the coming days +file is not schema to reproduce was a timing issue when using the testAutomatically option +In I believe the annotation is Validated not Valid +source code for and attached +Weve been working on this recently +This works ok now +Close it now. +Please tell me how I can resolve this issue! Thank youRegards Robsonthe problem is solved with the latest and plugin packs. +Administ Administrators Jun +i try upgrade Gatein to in ExoAfter i hope ill find a workaround if i cant upgrade +Since loadEdits already hanlded stream closeneed not handle this. +Im not familiar with Hibernate but I would hope our position is that Aries should work with any compliant implementation of the Java Persistence APIAlso theres at least one case where Aries JPA does not work well withIm marking this as a duplicate of. +Darran could you please paste here your build order again +More investigation in the code is needed to fix this issue completely particularly the classes in package that deal with exceptions +Thanks Eran +Another puzzle piece includes a completed Fetcher alongwith some minor changes to the api and a newly added to pass around +push out to prep for +asfeatures +I need the header class to be placed in the parent td just as it is in the main content of the table +I have got exception with listener class but I cant reproduce it constantly I cant stepify exact steps +I think this has been fixed in latest CVS +enough I will use thanks for spending time discussing and looking into this issue forRichard +All of these will be fixed when the new creation after becoming a TLP +Hi BensonI found a file that seems to map the property name to attProp +Would it be useful to take a shot at rewriting to use thisYou could do that or an alternate implementation +Im not sure if the pom of is already updated +Resolving it as invalid based on the fact that the functionality already exists +Username sflexus +Was this fix included in? Doesnt look like it am I right +This resulted in lots of imports that had two candidates one from the system bundle and one from another bundleThis ended up being a worst case scenario since the system bundle export was chosen first since it was resolved already and resolved packages are given preference +I agree that RI is not right at this point but I wonder if we break some existing applications by leaving this difference in HarmonyProbably we should draw some more opinions to this issue +Latest patch looks pretty good to me Andrew +test failure fix +Dmitriy up for trying to make a patch to address this? Sounds like mostly changing some declarations rather than implementations +This version of the patch updates the code to work with the new request objects and correctly respect the minbytes and maxfetchwait settingsPlease review the new configs and make sure we are happy with the naming +Well want to close this one too when the new authentication system lands on and operational in all issues that have been in resolved state for more than one month without further comments to closed status +This problem cannot be solved by just overriding the equals function in the options +contrib failures are just streaming +It was released with the fix in +merged +are you able to share any of your data or a subset of it that reproduces the problem +This issue may have been fixed and code may already be present in the old trunk +We could use a rw lock to get around that but I dont think it will be an issue in practiceUpdated patch attach +Not too many. +Mikhail you are right this method should be removed It has been added at the early prototyping stage and it is not actially used +Part of a bulk update all resolutions changed to done please review history to original resolution type +For example use the most compact encoding during major compaction but use the fastest encoding if memstore flushes are backloggedWe probably cant get it perfect in this first iteration +going to be fixed in isnt working out so were going to implement this instead +You need to generate it with git diff otherwise we cant apply it +Heres some graphs after implementing a solution based purely on and pooling marshaller with thread locals + +Why is it that your destroy methods actually fail in such senseless form? What kind of stacktraces do you see there that happen all the time but can be safely ignoredve revised bean destruction logging to log at WARN level with the exception toString only +can you provide a diff file patch +Committed it to trunk. +bq +The problem is unrelated the specification +Ill if I encounter it again +Pull requests merged on close all issues in resolved state that have not received further comments in the last month + has been added to Hadoop to enabled backward compatibility for downstream projects that used Hadoop in their testcases +so +In Camel you can enable transferring exceptions over JMS with the transferException option +And I agree with Karls suggestion we need to cause the active start level to take the first step into the lowhigh range before we start pulling off bundles and testing them +just edit the fop script to havefopexecargstrue +Estimating w d remaining on this split between the open subtasks +I dont have any use cases where the storer is used in my cluster though so I cant easily test that +retrying hudson +filePlace in this directory trunksrcdocssrcdocumentationresources just committed committed +I believe that the javadoc for those classes is good enoughDo these changes look adequate +You never can externally +Works for me +Fixed in Branch BranchSPCP Branch and trunk +Part of to trunk port +Could you please send a sample to demonstrate the problem +see for details +can you pl see if you started seeing these problems since was applied? Also if possible can you apply this patch and rerun the benchmark on the node cluster to see if it fixes your problem? Thanks +Owen one can have compatibility with and use more performant jre right +created a patch of the proposed a +defines all different types of replicas temporary finalized replica being written replica waiting to be recovered replica under recovery +We could just pull out of the registry in runtime and determine the valid source file applied and integration tests added all works great thanks for the work Brian and Amol +Id hope that we could be a bit more accommodating of existing deployment plans +previous version patches +Theres even an incrementupdateseq API that just bumps it directlyHaving a lastupdate in the db info blob that lists the last update for managing changes feeds might be useful in general and seems like itd be an easy enough addition +Thanks for all the help on this one +When a tasktracker with a completed map task failed the map task will be and all reduce tasks that havent read the data from that tasktracker should be +Hey Ill change the copyright Well +Do we really want that check disabled the vast majority of the time +Thanks Thomas +I just did a build on the latest snapshot of trunk +Fixed and pushed +This code was originally created when integrating Katta +Lets hope it makes this week +Currently this is not allowed although may change this +The metadata is all consistent and the problem seems to be localized to that one compression block +sorry reopened wrong issue +Hi AkiI just did a test with your patch it worksI will commit it shortly thanks for the patchWillem +Is this a JSE or JEE test? How are you enhancing your Entities? Can you provide a simple test to recreate the issue +If it occurs on the next build of the update site then please reopen. +Removed the logging monitor gunk +It should be fixed for yours too Jess please verify and if it still doesnt work for seems to have fixed the problem +I think this one was solved in right +Bulk closing stale resolved issues +The service name is sampleThanks the patch thanks Thilina for the contribution. +same behavior should be present in more IMHO +For branch +That way we can at least get a commit of the PTOF without further delay +applied thanks! +But there arent actually significant functional changes to existing codeSince the patch passes unit tests and doesnt involve significant functional changes I intend to commit this barring objection +This is impossible to understand. +The next release has been added and the roadmap is updated now +fieldName validUntil type notNull Customeradd field reference jpa class +how do you use it +I did have one DOWN event on the load balancer during that import but it was not caused by a GC pause +Thanks. +Ill add a save action that strips tailing white spaces when Im back home from workHave a nice dayBenedikt +path to branch and trunk +Applied patch with thanks to Jonathan +The problem you are encountering is that the general ledger is not configured did you load in financialsdata for example? If so and problem still continues please create another issue on for the financials module +Moved init test cases to a preGoalFixed in svn revision thanks a lot for the simple and effective fix! +Moving out of +I still see SQL Exceptions on the setup +This also avoids side effects if some unrelated exception might be declared on the service locator interface which the original simply ignored based on the signatures seemed like a way to create the strongly typed service locator +I havent confirmed that is causing the problem +Joris well as far as I remember the goal was not to make things fast here but rather to make them possible +Thanks. +Yes Im not a committer I just review things when I can Ill keep an eye out Im stuck in a hotel trying to get a flight out this morning +The OP did not answer to our questions and his request is incomplete +But this is probably the way we should be doing it as it appears to be the Pig approach +Resolved previously +Please try the latest RC and report back if you still see this behavior +No extra code is needed +Moving issues that arent going to make it into to provides this +The RF dt again does so the dt and edt should be consistent IMHO +Please find time to test and let us know if its still there +Unless anyone has any objections Ill change them to File objects +Looks like it WAS committed to trunk which is actually +Bump +applied your patched to trunk. +Patches welcome +As part of an Ajax request your controller should return the content your client expects for example errors test for a tooltip or some other partial page content. +No problem +Marnie thanks for those links +This patch formats the output of local task +I can confirm this bug +we may not want to release a whole new version of the default container for just this fix but its fixed in of. +We need a design doc explaining how all of these bits work together +But why is a filter in the first place instead of a Servlet? This is somewhat but if perhaps it should be a servlet then this issue is majorly disrupted by such a change +And its not following the spec so it means someone will have to do it +Much of this has already been done with more coming +I think that I did receive a reply message in the past +So I am closing the issue. +If not how to optimize and fix this problem +Check out the saxon component where the servicemix repo is added as a maven repo +Issue resolved +bq +Just improved names of xml element and associated class and its methods +hmmm +It is most likely Upstream issue in WTP Web Project wizard so Im scheduling it for next JBDS version to check it and report an issue in WTP is this issue still relevant? Did last comment help +The patch passes all tests of CTF and should resolve the problems seen at current some additional work needs to be done to fully solve the duality of source and configURI in +The other versions are patches that dont break backward compatibility so Pig does not need to be updated since on the backend it does not use built into itIt is possible that in some cases Pig might need to be recompiled with patch changes but that has not happened to us yet +Committed at subversion revision Touches the following fileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyvtiBretts original patchM javaengineorgapachederbyvtiTouched up the javadoc to say that! is a supported operatorM javatestingorgapachederbyTestingfunctionTeststestslangTest case for the! operator + is out the door adjusting on all remaining unresolved issues to by concrete feature request can be solved with changes now +I will look into alternatives +Im not able to reproduce this +What process is opening up ? If it is the node manager do you have log aggregation enabled? Could you try running the test with log aggregation disabled and let us know if the problem is still reproducible +Its a bit obtuse for something that is basically a + +It might be very well be releated to the security provider +If you have anymore information please let us know. +Great thanks for your work +Lets see how it goes. +Unless you left out the part about changing the db schema so that the parent knows the childs primary key I cant seem to get this to workI have this as the parent nameangusChild class cascadeall falseAnd this on the child nameid foreign type columnid null uniquetrue nameproperty valueangusParentSeems to be a real bug to me. +Attached is another repro case showing what works with the default configuration and what does not +rebase +So for now we can help check to publish mvn repository to a location of our choice and see if we would be able to integrate with thrift + +Theres a stray printf comment and I think OK is more common than Ok +me and the related Maven plugins listed in should be installed by these should be uninstallable if user wants to +Although it seems like a better answer for get would be instead of +Ugh! Thats really badWell fixed in CVSWe really need tests for this shit +However I can see nothing in the standard that backs up this assumption +just did is also now so add users as you like set up notification schemes etc. +Do you have the SLD and a sample of your data with the feature that makes it NPE +in other words to my knowledge sharing a token should not possible in +And our build is failing because it has detected it is not present and the PDF needs to be rebuilt +add it as a jar in lib folderPlease have a try again with these additional jars +Final patch attached. +Ok but can you compare andThere mus be a diff between the two as on our case successfully reports all the files +Ive changed the code so it should cover both cases now +This is resolved in version which is now available on Maven Central +Okay fair enough +For Review an early version not finalized yet +We should consider also doing this for +It still has the same issue in both and with JVM on LinuxIts a crazy issue somewhere +Yes you are right +TT should get an instance of from a configuration variable Make TT use a string for a pid instead of an integer +One specific use case is finding the distribution of a directorys data blocks +It has been left at false for backwards compatibility +The problem with testing is that its really hard to reproduce this issue on staging I guess because we have small number of slaves there +Bulk close for +Thanks Todd for the reviewI have committed this. +Please help review the patch and commit it if you think everything is ok +Thanks for the patch applied to trunk +Changed the issue Summary since the problem was elsewhere +No problemJuergen +I think that was fixed in +In the long term we need to create an accessDenied exception along with several others and have the framework handle them correctly similar to how Jersey handles it +I think this copy operation needs a deep copy following all the references etc +thrift has no released version +As jeff said this has been fixed +An easy fix is to not do merging minor compactions on the metadata tableAlternatively we could move the delete markers for the METADATA table into the root tablet +Good catch on that bug and enhancements +Committed. +Thanks thanks so much Namit +Fixed in the trunk +I wont have a chance to do it anytime soon +Thanks! +This looks like a regression caused by +This is a good idea +I will have a look at it sometime this week or early next weekSorry for picking it up so late but this bug went under my radar since it was not marked as an ajax or jsf bug +Reviewed made slight update. +import is before compilation and it fails on putting the project structure together +This is puzzling +one diff for the mapreduce bit of the source tree one diff for the section submitted as bugs against yarn and mapreduce projects with the submit patch button set to trigger the relevant build and testWhen I apply the patch which I will do I will do the unified patch as that will mean only one patch backport from trunk its just that we want jenkins to do the workActually you could just get away with creating for the two things and submitting the same patch to all +Please open one more bug when you have another patchthanksdims +Committed the change to the trunk as revision Ill investigate merging this change back to the branch +Bug. +I committed a fix it was a test bug when the number of indexed terms was mod term +Also looks like Hudson is AWOL again so you may wish to run tests manually +javaimport is there and we have been recommending it but we need to add a warning to import and includeclass that says theyre deprecated or not recommended +Could you give me a minimal pom with a configuration to display the problemthanks +Okay I thought you were talking about the configuration stuff not the code itself + +Alan did you run your benchmark with the new option we suggested? Also you might want to update hive to trunk to take advantage of new hive performance improvements introduced by. +Fixed the reference guide link but the other links are actually not broken +Users may set whether they are selected or not at their leisure +Committed to trunk. +Hoss that sounds like a reasonable feature requestimprovement +Yep thats basically what its doing +Changes were applied to the soa build project and should be in the ER release +The only availableworkaround that leaves us with is to instruct the customers to useFirefox for downloading content +One of the transformed columns had both infinite and elements so the test of the norm was never met and a column index was never setThe fix consist in detecting norms and throwing an stating the decomposition could not be performed +Note that repeating this for number of supersteps would get really costly +Im sorry you are right in saying a didnt provide enough details +thoughts +Did you manage to confirm that the IPTC Spec license allows us to include parts of the spec in our javadocs +It only changes on the bookkeeper side right? One option I see is to keep your changes to hedwig and not have recovery acquiring permitsOne clarification if you dont mind Sijie +It would be uncommon that a person would both want to print the selected value and store it in a page attribute at the same time +Perhaps an underscore or something +Gabe can this be closed +well +Yes I would say that when you decide to use condition on transitions for fork activity you should use multiplicity attribute on join to make sure your execution will continue +Win Xp bit and Win bit works fine +I have one major concern shouldnt we file this against VME and wait for the next snapshotAnd several concerns about the patch Probably misprint ends ends If we add workaround into class library then we have to check all boot class path provided by a VM so it would not contain double separators inside it not only at the end +Thanks for the updated patch Ill try to get round to applying it over the Easter break +Must have overlooked it +Most of components are up to date including new file + AS M will come with Weld built for info Pete! And it will also make the archetype generated jee project deploy well too +The difference that probably causes the issue is a different date representation +I have attached an image with the mentioned option checked +The performance on Chrome after setting the maximum of consumers windows on was improved however when using Firefox the performance is again very bad when some of the consumer is replaced with new consumer then the sent message is shown on consumer after sec +As a consequence only matching rows are returned and the other rows for the same feature are logic should instead be to return all the rows having a gmlid for which the name matches +Do I fix it correctly? Why the hdfs sink file have a repeated character in some line +Im having the same problem +But I found there is a plugin that has include the bpel project example +If the file was empty then it was probably the result of a deletion patch getting applied +Resolved in now since fixed. +First pass at a class to implement fixed point types +the above +The unit test failed on the experimental branch without thatPlease verify and close. +This issue has been resolved by changing default value of the property to srcmain in additional property is added in file and checked in to svn +Hi guysThis was a tricky one I have a local patch for it +Eclipse version isSteps +see problem easiest kind of ticket to close +Please attach a testcase as we have tests that show it working + depends on a bugfix for this plugin +After discussion we decided to get rid of Backlog all together since there are simpler and safer solutions +Is there some design pattern or process I am not familiar with which recommends this? Because it seems to have really trivial pathological and confusing cases +So I dont care +do not forget to close this bug once you are happy +I have this working at work but Im at home now + +On an another box in full test run these tests indeed passed +Is this still valid for the released versionOtherwise attaching a sample project would be very helpful +There is no support yet for the fileset extension to provide an action to perform on +Im switching the fix version on this to none +Ive began to work on this issue and it almost work +Choose a command that exists on either OS or something something like I suggested one test for Windows one for POSIX systems +Its used by s constructorOh +Net add does +The patch +It is hard to give meaningful and it would require a major rewrite of the DRDA code since this is detected way donw in the call stack +following up on henrys comment it appears that we are indeed calling some of these from synchronized blocks but they are really only synchronized for the call itself +Thats Right +unnecessary fields are removed from List in +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks V +of replication factor +Updates committed and pushed to the web site. +Gary Yep Ill be working on a trunk patch as well +But yeah for now go ahead and commit it +Implementing the above changes seems to break Some EXTERNAL tables are created in that test but they throw an error if I try to specify a LOCATIONAny thoughts +commited to svn +Integration tests were updated +I took a quick look at the latest patch and it is badly out of date now + for the to trunk +Duplicate of to +Thanks for responding +pem files generated in Windows as well +BTW at this moment if an added value exists it must be accounted +I this I have fixed the issue +encapsulates both and within it +Ill commit this patch with bogus nocommit removed passing precommit checks +Cool stuff +set JAVAHOME for run of kernel cunit and smoke +Committed thanks. +fixed in for add proper resolution version fixed +Is that intentionalAlso should it be or +Given that the problem is a large regression on jess it seems extremely likely that the problem is that arraycopy inlining has been broken +fixed in latest greclipse dev builds +Thank you for your reply +The previous executeStatement has been restructured and modular based on the functionality implemented in that +defer all issues to +This one actually passes +dont see how that is possible +for updating text to point out the requirements in no openshift should only be used if there is an existing git repo there AFAIK +Works for me +Thanks for the patch Doug. +For now get the Roo source from Git and try out +Fixed in to Vijay until the specification becomes available +Ill put a slightly better error message inbut processing will terminate neverthelessCorrect the source +Fixed in trunk +Patch applied at +Added catalog file example + +Let me know if you think that this requires a test case and I will be happy to include it +Can you trycollectionLet us know if that works +bump +Ryan will do so Was just on the Go but wanted to point out that the issue is a simple javascript fixUwe will try to add some global for Issues like that +This should be resolved with the delivery of the new tabs +Thanks Sergey! +use some temporal before we call storeDocument and then use results from storeDocument to exchange with real one +what happens if one remove fails but the others dont? How can the client get information about each specific remove that happened? What if the client has permission to remove one but not the others +We will try and begin these discussions before the summit as our time will be limited there +and ? Isnt +Can this be added such that MHOME M are the default if neither are set it falls back to M +Committed to. +Yakov would you mind attaching your patch as an attachment with appropriate ASF license agreement? As an attachment its also easier to test itAs for your second problem that sounds like a new bug needs to be filed? Would you mind doing so +looks good Lars +html in your so the request is not routed through the Tapestry +i think your client library should also let it flow to the keepset api level +Yes I did +Postponed to +Just to ask why do these features have to be coverage tool specific? What is the difficulty in supporting other than +Done also for the JBPM Tools BPMN Convert guides and ESB Editor Reference Guide +Hibernate should have a listener for bulk updates not the standard entity listener but a listener that receivers the hql query executed so we could create a that parses the bulk update HQL +It looks like was not present in earlier releases so no need to go back further. +FolksPardon my ignorance but I am not sure at all that chunked encoding can be usedwhen submitting multipart formatted requests +At the same time we need to remove the dependence on +bq +Patch committed in r +Heres how you run these scriptso First run using an earlier release like +True but Struts doesnt keep references to the tag objects the container does +Re Id put the extra distribution build stuff inside the profile so it wont get invoked in a normal build but will automatically be run when cutting a release +Never mind I found the code style scheme +Committed on rev +ogcsrcnetopengiswfsmodified +Attach patch v to fix a busy wait issue in +scripting is not easy especially when one of those platforms is Windows +r fixes remaining build errors +Oddly the test no longer fails on my system wo this patch +Okay I clearly underestimated the difficulties of this due to +Of course this is just tests oddly failing doesnt mean the core build itself fails +The is also supposed to be readablewritable in other languages with the spec being clear enough on how to read and write Tfiles +LukeIm a user of this project not a developer +The diff file really mess up +To your first question No memSize is not what I am looking for +Rebased with avro changes +Yes a faulty customer logger will affect the functionality of the NN +Aliaksy just to be clear everything works right? you can launch the server etc +So in the past sorting certainly added a cost to multiple segments as you moved from segment to segment did that go away in some issue? That must be completely different code these days if segments or more performs like one +Thanks +This issue has been reproducable on windows windows server r and ubuntuWhen this exception occurs it has been when using both and +Please confirm the current status of this issue and if its incorrect change it as required +Perhaps someone else will find it useful +Committed revision Thanks all! +Greg if you think this is important you should take the initiative and commit your changes +The idea behind using ZKs logj logs was to modify ZKs code as less as possible +Is that what were trying to address +Tagged for I believe the addition of this feature may have broken integration +reopened since I had to rollback the CONFIGUREDFORUSEDINTEST public boolean flag +Kevin found that breaking up the URL into its components and storing them as separate columns results in significant space savings +Did not use patch +If you have any questions even if they seem trivial you should post them and well try to answer as best we can +this looks good +It seems to me that you created that output file in teh Project Archives view and then told it to deploy repeatedly on any change +Also a restart leads to SEVERE Error listenerStart +Mojo class +I have introduced a parameters in a new file to allow showing the virtual keyboard or not +These blockers and critical issues are resolved but not verified +Fixed in and trunk. +These methods will be helpdul thanks for adding them. +Assign to me to add code enabling the Pig counters +Oh you mean the regular docsexamples in the dist? The references to and jbossweb confused me +The mandatory thing needs to configured to invoke ajaxjsf filter before myfacess +Thanks for catching this Torkjel! +XN patch resolves this issue +Could some one review this +Looks like Auto Config from the compendium spec is the name of the game +I am getting same compilation issue while trying to build commons module with mavenChanges to default and default type from int to Integer in method get of class fixed the issue for me +Please advise to disable the test +Will address these forms of issue over separate smaller and more divided in futureClosing out parent JIRA and hence closing out this. +Added you as contributor Bob. +Pushed +Im not sure it was more than eight months ago +Ok than we should discuss to delete the applicationIn other case the jCarousel is a nice image content slide plugin we can commit it i make an Widget Example from it and we can delete the application when everyone agree with itIts up to you to decide agree for jCarousel for the component we have to discuss about it on dev ML +You basically can never set listenaddress to an unspecified IP address because gossip needs a way to communicate with the node +Basically whenever anything needs to wait on a given object it calls into to do wait +See for further details. +I want to know exactly what the test is biggest concerns in order of priority are what effect does this patch have on Clean install start time +Thanks +I was running Sep nightly build +Ive made it painfully clear why I offered the patch to the community +I dont think the current terminology is to users as these are really internal enums within HDFS +Notably I wont need to sync on the secret manager since makes the ADTSM +Makes sense +Im assigning this to Anil who will provide the release tag and then assign back to me +Assigning this issue to Don Brown since the original issue was assigned to himSorry Don +Still more to it than that there are a number of files that have not been copied across +backported to r has a lot to do before leaving on vacation so Ill finish cleaning up my own mess +Consider as an alternative for hosting artifacts not suitable for maven central +Ill commit this patch on Friday if there are no comments by then +The used in java source files need also be taken care of +as before dont use the stuff its just an example +I think nospc means that theres no space left on the page +Patch to remove all private classes from LuceneSolr trunk sources including tests +I updated the patch based on this suggestionFor the sleep time when I tested it I increase it because I thought I have to wait some time for the update countersBut the counters will be updated at the very beginning of the task +Are all of the failures related to JDBC or could there be some problem with running junit on this platform? revision . +So if a Map join is involved in a plan and the output of this join will be consumed by another subsequent operator we should leave the Map join in the Map phase of the subsequent operator instead of cutting Map join to a separate MR job +Mark as fixed since this is included in. +Hi Colindo you think i can attach my code to the ticket? i think that by checking the code you can understand better than my explanation what i have implementedthanksAndrea +The Admin UI will need to use this to display the actual list then make a STATUS request when the user is drilling downI suppose another solution would be an extra parameter for STATUS namesOnlytrue or some such but I dont really like that although Id be willing to be persuaded otherwise +Ive heard from others that a ant will fix the issue if you see the same stuff +The incompatibilty is within the connection to Zookeeper +Please commit +Expected behaviour customized manifest entries should overwrite generated bundle entriesSo need to reverse the manifest merge also should check for customized manifest sections +i would apply the patch as soon as possible +Looks fine could we just add a check somewhere that the path must start with and throw an exception explaining to the user the problem + May be we should postpone creating the dn block map until the upgrade is done +lets create another task where we define what should be the nd tier testing for wsrp. +We do have which gets a deserialized message from we can get around the problem by just removing the assert which was added to make sure we are not assuming it is a echo message +I will try the PIII at the weekend and see how that goes +Here is an updated patch +Affected version should be +You also need which updates Hive to use Thrift +Thank you for the great patch Marshall +The seam example doesnt require jgroups so it can be run in the default config +Addressed Tims two comments as well. +So +All done +Jup check out sharewwwscripttest +Yes this looks like a good catch +The problem has been checked again and still exists +Ilya please work on the first part of this task +Ive double check also +Pavel why do we need to touch code here +Any suggestion on howwhere to implement this +Do you know where the code is for that or the exact text of the message I can search on? I suspect it was an INFO message and is now in the log but not on the screen +APP is the database owner +more commentDuring the st call the URL is found in bundle During the nd call this bundle is listet in bundles but the URL is not found! RajiniWe have seen that all Tuscany bundles declare This should be used rare because it will cause slow classes loading and may cause errors if few bundles will share class with the same name for example different versions of one classBeside we dont know if this is used for Serialization issuesbecause we dont understand how Tuscany can find classes to deserialize objects that were transferred with Tuscany Thanks in advanceSebastian +The test case raises NPE without the fix +all footnotes are displayed none is dropped +Awesome work Maksym +Were you able to find anything in the release notes on this? I couldnt +sorry does not disable so you have to set the threashold to a value greater than ever will be met like +Im not sure if this is neccessary as OPIC is not really considered a proper method for scoring incremental crawls which is what most do +It doesnt bring any advantages as well as it doesnt make any harmHow ever some effects could be achieved in runtime while using different set of CSS class names with the same styles +I should have sent you link directly for your information but I added link to message with other details +IMHO this was a plain oversight +normally it works fine +Verified on ER. +Integrated to branch as well. +You have fixed this havent you Todd? I see RIT in master UI now +Even if classloader scanning was remotely reliable this is not what you want +The use of the +I am able to reproduce the problem when disabling the cache. +Committed revision +There is still some cleanup to do some more stuff to add for actually running forrestNext task is to gradually add dependencies until the is successful +It is easy to expose object +Heres a patch to revert the changes made by +Hence I think it is OK to address it in a separate JIRA +Added in to for release in late +It can handle more annotations and should be safe +I had used a correct repository and the issue is still +Waiting for a patch. +Hipatch with converts the first bunch of filesPlease keep the issue open +Let me write a unit test to see what gives +Suggest testing with a newer version of the broker SNAPSHOT has many fixes +Hey Min I believe registering the metastore jar is necessary in both the frontendbackend +Thanks +Daniel please verify +resolving wont fix. +AshutoshI dug deeper into the issue +We are trying to move off anyway + +Committed to trunk. +we might want to review the advanced only portlets to see if there are more good candiates for basic view +But it does work for JBT wonder why I dont see this? My search menu is just fine. +Only appy in the patch file +We pick up only one service from a wsdl so please split into wsdls one for Service and one for Service for now till we fix thisthanks aar is fine +Thanks a lot for the quick turn around and keep up the good work +CDI validator should mark such we will have a dialog with multiple options anyway +As long as you realise this can only ever be a best effort attempt + +Closing +Charles thanks it looks good to meGuillaume agree its more elegant to put it in the + Knocking off the refactoring code +The derived datasource property was not expanded in +Wouldnt help too much though since RM management would interfere with the client managing the tokens +However some subclasses of such as does blocking writes to a queue +You seemed to be complaining that classes were too specific for this case so I said I was okay with if you thought that more appropirate here although perhaps thats too general for your taste in this case and youd rather separate +Hi Myrna I think your patch from the th is safe to commit +Done +This may be an issue with Windows file locking +on and +So the dirty problem doesnt occur in my test case +Ill see if I can add something to also +But theres only one cookie attribute path and offhand that looked like it should match +This seems a good idea to me +It should be placed in trunktoolssrcmainjavaorgapacheyokotoolsprocessorsidlThe new test I added for the new types requires two new files in the resource directory +Until then were making do with the Thrift calls +How could a unique field which often implies a physical index be nullable +The main goal being to forcibly drop support for them before branchingThis sounds like a good idea +Can you create a patch? I dont think the unbreakable space has been added on purpose it can be removed +The webapp used for testing +Grails shouldnt log such standard exceptions at error level there +thanks +This is running on EAP so way to reproduce is to download JBDS which includes EAP and follow Jims instructions +Hias explained on I had to create a patch that removes all checked exceptions from the API +not amount is added but +Np Robert! Was a little confusing indeedxieguiming Are you willing to contribute a patch against trunk for this featureimprovement? Wed be happy to review it +So for each of the ant commands above use option to pull in +Thanks! If that does not work I will try the mailing list +Will have a look +bulk close of all resolved issues. +Thanks +Thanks for the patch though +At a glance the method it occurs ingetlooks to have a complexity of about + push to +Yes probably +maybe caused by some dependencies which should not be there when deploying to GAE or some other configuration required +I then ran the application from and it ran fine +Ill go with the method as this is more generic for example you can use it for and other purposesIf a resource is not adaptable it is ignored if one wants to distinguish between resources that can be adapted to the given class and then fallback to something for others one should programm the whole loop +We probably need to delete beta and create betaThoughts +do not critical for final I think +Patch which makes the test to be evicted if the Dialect is the Sybase one based on similar eviction code from other tests +you may change the source code if you need to compatible it with java as you have mentioned +Should it be untabfied? If it should be untabfied what number is best for tab widthTab width is just a configuration in my development environmentI didnt think this configuration in untabfy was problem because I have seen tab in source code with this configuration and have been familiaredI think best configuration would be different according to person +This class will be removed when the binding metadata API is fixed +Graeme I respectfully disagree +No it does not matter and that is fine +They will help with internal book keeping and tracingI am not quite in favor of exposing TERMINATING states +Thanks Mike! +If the cache is running in disk mode SWAP then the items in memory will go to disk otherwise they will not +Luckily it appears that you guys already fixed the problem +An initial patch for a deployment interface +Lets drop the DB and try this again +Im rebuilding the causing bundle as described in OPSJs Jira and see if it still happens +Methods in the script are not carried over classes are but only because your script and the script before share a common class loader +This should be available in tonights snapshot +I found your Decoder implementation is invalid +URL filtering though would always be appropriate +Glad to help track that one down +I think all were doing is skipping some callbacks that would otherwise have whcih maybe clients dont want to see anyway +JVMTI correctly sends an event to the agent but for some reason Eclipse doesnt see it +Patch conforming to trunk and +G +Ive submitted changes that hopefully resolve the issues +If I have fixed this I will update the patch +I just got this message starting my because thekeystore was not in place +Thanks +Added some docs on it today. +However I am still waiting on confirmation as to whether or not the default is BLOCK or PAGE +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +I would say one new parameter called showThis has a default value of pages to remain compatible with the current behaviorIt can have a value of count showing you the number of pagesIt can have a value of lastModified showing you the count followed by the datetime of the most recent changed pageIs that what you would like +is defined +But I dont see it from the JIRA +So the only workaround I can see is to review and rewrite all Roo tags to avoid JSP EL features +The project activity includes coding writing doc testing infrastructure related activities and generating developer interest +Ok today I was able to resolve my problem it had to do with a corrupted plugin jar in my repository on solaris everything runs fine now +So to my knowledge it all works now +Done +Will most probably do the job for trunk too +We have some problems it can be reproduces very easily for example by restarting database server and i agree with poster that Websphere doesnt have this problem +At that point sufficient privilege to open restricted ports cannot be regained +Resolving +I was wondering how often does java get run? Dont think it is part of junit suite +Unfortunately it looks like this bug has somehow come back in FUSE ESB +This is what I was alluding to in my first comment with the TODO item Interactive cert generation process more like binaccumulo init rather than running with a big pile of switches +Thanks +added a default for when RS in r +It is the most widespread logging system in use +See also + is a visitor that walks the entire plan of +Ill tryW +Digging more into it I see was missing from thriftif target in serde I added that so I now have and getting generated +Fixed +If anything we might need to add something to andor include a copy of Snowballs BSD license too as something like +This test does basically the same as your test but it fits in our overall test test runs fine +Once guarantees delivery order I can ignore the seq field in heartbeat and pings messages +My misunderstanding of DS +Where did this commit go +I used Charles to test itOne other aspect is whether we want to enable connection reuse for proxied connections +The JDK is very brief about reporting zip errors +This is able to load from +bq +Mohan or anyone in Taobao is this patch still relevant? If so can you rebase it for current master and post again +Logged In YES useridfixed in and HEAD. +I can confirm that this works for me in and fails with. +hi guysin result of Ive decided to migrate to the now I get this problem too +IMHO this speaks for some code trying to match by the number of arguments there any plans to fix this yet +Hi Doug Beattie I cant believe that delete button doesnt remove project data +very minor though +This work looks great as a prototype of the join order optimization for the distributed joinThis patch contains some changes of global planning of other operators such as sort as well as the join orderingAre these changes just a refactoring of the global plannerI think that it would be better that the refactoring is proceeded in a new issue but it looks hard to be separatedOther parts of the patch looks good to meHere is my Please remove unused imports from some tests before you commit +Thanks Bikas Vinod Hitesh Alejandro for the detailed discussionThe issue however exists with dependent and need to be handled may be in Ooize for now? In the long term would it make any sense for YARN to support +unless you just mean to make it slightly better by freezing a snapshot likely its going to be a snapshot than just making a copy but still people must be able to live with the fact its not a perfect match +If the user wants to override the set TTL it should be done by calling the appropriate producer send method +Patch to achieve this +Sounds good committed the latest patch +Stuart has sent in an update which I have committed which fixes this problem. +Updated patch for this specific issue +But it still exists on IE +I am using version +We probably need to change the append response to have the server tell the client which checksum to use +Removes reference to LONGBYTES +thanks. +Changes is in class +in your application? If yes remove those jars and try it out. +Partial fix for JNI function +move issues to and +This form used in inliner and escanalyzer to start new compilation +repo is just for QE it is not buildable on its own thus not useful for us for CI builds +See Javadocs on for information on how to use this. +Fix by it only affect AS when using the configuration +I kind of think that having an offline scrub is not a bad idea anywayWith that you should be able to stop the node run sstablescrub ksname cfname and then restart the node and you should be good to go +Is it +Your database is not thats the Web History DBThis shouldnt affect anything with your app since your app uses a different database. +r +Hi StephaneAny chance of getting this in? Looks like the above change will not affect the normal flow +To see properties info you need to type help properties +That should keep me going +i fixed it in different way could you verify that it fixes the problem +instead of pt mm I changed the units to match pt pt The rounding errors are then avoided and the changing IPD algorithm is not started +Which JAR are you referring too +Anyone have any thoughts on this one? Its not clear to me how to extract the information required from XML SecurityIMHO we need to dig into xmlsec here a bit +This also surfaced in other contexts +Hi Ann do you agree that the best solution is to not ask anything to the Sonar administrator by providing those abbreviations? problem Ann +This is a patch that fixes the problem and adds a test to verify the correct behavior +So please post a question in the forum. +This introduces a release page and wires it into the download page +is for is for great we got a patch for this next time please wait for review before applying only solves one part of this though it does not solve the ESB plugin side which still would have to just to support a new updated esb plugin that more than need to change afaics +Ok let me check that +With the new fix this should be solved +Modified patch which registers a service listener to update the system state timestamp and registers theregistered services with their registering bundles instead of with the system bundle +But feel free to reopen or create a new issue if still experiencing problems with indentation on tab presses +can you check it out for me +I patched solrIndexsearcher with just your changes +I have create one hadoop ticket to fix jenkin and make sure it runs all the tests before commit +It really helps if you put in a bit more details such as an example of the current error message and suggestion how it should be alteredAnd even better a patch +The fixed one is already at that URL +I have resolved as fixed +I am getting these same errors however here is some additional informationWhen I run mvn javadoc from within either a standalone project or project that has test sources it works I run mvn javadoc from a parent project it fails when it gets to the +I hope this can solve your issue as well until the jruby guys come up with a fix +But why is partName parameter ignored when name is declared and parameterStyle is BARE? It still seems to me as an unexpected behaviour +Sorry I needed to remove the document because it contained some personal information +run Hadoop on all of the Cassandra nodesthis is the approach weve been going with. +First thought it was by extension but my + +This is being fixed in maven the meantime about all i can do is make a new enforcer goal that doesnt require dependencies but then certain rules would fail +Someone could be reusing the same driver for several tests +Hi Catalin we use targets to generate eclipse projects +Resolving as fixed. +its a poc nothing more +issue will be moot when goes in. +Seems to work fine +If not should probably go in a release or LATER +Update Center is back +Problem was only in the test file +Mass closing all resolved issues that have been closed for more than one month without further comments +Please give us more is your POM which produce this error +I have no earthly idea if this patch will help you or if it will even work +The reason the last character is still misplaced is related to the fact that the character is drawn independently from the rest of the line already within the pdf +Thus Ill keep this issue and postpone it +All critical bugs resolved +Oktest latest snapshot version +My last comment was ambiguous +i am also getting same issue +Desided to attach the interface and the implementation that still needs to be seperated into a global and a local one before throwing it away +Never been able to reproduce I would close closing all resolved issue that have not gotten any more feedback or comment in the last month +lets get it in +Hey Andrew it depends on what kind of database you are using in your project +I can work on it now together with BIXO team andThe problem is but Ill try to minimize changes +Nice! +Dynamic shadow proxies are now completelly integrated into core and compiler. +XufengHave you tested the patch in real cluster +Documented in the Release Notes as above +Thanks for following up on that Daniel +the jboss tools JSF projects wizards are not facet based as it is now +and if this is not global depenency it would be better off in spezific pom +Thanks that was just about the kind of details I was looking forI will try to refresh my Java security soon I think it is needed +Fixed in rev +Update for new test output semantics +For all file systems that I know of that have a static number of inodes the ratio of inodes to disk space is configurable and even if it requires reformatting the drive knowing that it is a problem will allow operations to possibly adjust that ratio appropriately +For lets have just a plain text area on mobileFor we can look at an editor with a UX tailored to mobile devices +Is there a work around for this? Im experiencing this in Glassfish v as well +I am not sure what code changes would be involved in the DIH I guess you would need a CDATA section in the XML file we are currently using MYSQL move fixing the version info for and all affected issues have in is no indication that anyone is actively working on this issue and it has no current patch so removing from the fixVersion +There is even a bug in the test its only using like bits per valueSo this is really even worse +default should work +Could you please give the incorrect sentence and perhaps some context around it so that I can grep for it +We could imagine to support two kind of exclusions Ignore duplications in source files whose paths match a provided pattern ex also would be relevant for complexity in general I would like to be able to exclude some files from all analysis + +Thanks Aaron! +Please do not require a cluster to run the example programs +The question should be rephrased as Should the eventual boundary split be amru or mruThe current code will produce a key split of amruI was assuming you wanted a test code just for you guys to test if it works as intended Do you want this test code also in the source considering it is a test extremely local to a particular function +Otherwise it looks good +In the next time I will try to do that +Ich bin sehr froh dass dieser Bug nicht mein Fehler war! und vielen Dank +Reassigning this to Thomas to see what we can do for Oracle there Thomas are you aware of a limitation in Oracles JDBC driver in that respect? thats what I thought thanks for the clarification ThomasIll close this issue for the time being since we cannot do much about it in I used of Oracle driver and Spring for spring framework it still retruns not the right affected rows of sql executed result. +Crucial issue for GA release +ajax version of page event test is missing in epp +Close this pending issue. +We need to post to DIH for the various transformsjoins specified in to take effect +bq +If he doesnt find something similar he will report a duplicateBut I have the impression that you dont care about your users +please provide more details +As I understand thingsA little work on the creation order should fix things +Any reason this patch hasnt been landed? Im still encountering this issue in trying to replicate over Back To My Macs VPN which is only +EBS is faster than LTQ +This is deprecated so it should just be ignored in the stats to clarify its deprecated for all types of datasources and connection factories or just for XA datasources and connection factories +For example this bug affects TPTP profiler running with DRLVM +Classpaths can be quite confusing +Thanks Mahadev! +This patch additionally handles acks with seqno and some corner cases of the last DN in pipeline +What is HADOOPHEAPSIZE set to? How much memory do you have on the box +Comment from chipc via IRCAnd is this actually a blocker severity? Does it still occur after the expunge interval has cleaned out the datastore in question +Common part of large which has been actually made for +Patch applied with thanks +hi everyone If there are no objections I plan to commit this as soon as I write a testcase +If the Cocoa ticket is a duplicate of this C ticket it should be generalized to include all of the libraries +Reminder it was implemented after these main events me rember of an easy to follow implementation by Doug Lea based on Semaphore Rod suggesting not to rely on any port and me giving up with think the API is well done and ready for the release +Shouldnt be too hard tbhThat said I will also try making that part of the comparator faster +Does it affect UIusers? You think the fix is safe +However I dont think JCA will have that set +I dont think there is currently a case where a token will have a renewer when setKind is called but it couldWhen isManaged renew or cancel are called on a token it first calls getRenewer +Closing no response from the reporter assuming OK +msec SLOWER Djava Test create MB file +Any update +removed the overview +Similar to this +In the case we end up writing to the ledger device before we write to the journal as you report here it is a problem if we have a pointer in the index of a ledger to an entry but the entry doesnt exist +applied by chirino in the revision at Sep . +the description tag or the content tag +Thanks +Sorry about the break of the build earlier argh +No longer seems to occur using SDK on Mac OS XThe traced output only showed the text that was pasted the second time +Thanks for review Ted thanks for patches Jan! +Would you please verify. +Will hold off until mesos API supports querying for list of known tasks in the cluster and use that to reconcile our internal state +Thank you for your review Jing! I agree with the points you mentioned +I cant rely on because it is ignored for those where bind does not delegate to bind +I can verify that this patch causes the test case to succeed +Dave can we really make a change like that in a point release? Wont this change break everyones +Please verify +I am attaching the logs for hdfsRMNM +It doesnt cause message redelivery +OK I updated w the cool and tests all pass with this patch and +And a fresh checkout gets a completely empty geoscript externalFetching external item into build with spacessrccommunitypythongeoscriptChecked out external at revision +Patch to combine all incremental reports in a single call to NN +upload patch for resolve this issue +Id love to figure out some way to take advantage of the work you guys are doing to achieve some of these alternate goals +Sorry I forgot DB Oracle XE is a dupeJoined and secondary table cannot work together is the duplicated bug? Can you explain the Joined and secondary table cannot work together atm comment I dont understand +It is not clear to me how to download the project from the link you sent me and I am not setup to build it even if it was clear +Please note that this problem is also present in Chrome and IE +Please supply ASAP +in palette works correctly +HDFSs is currently implemented with +Is that Ok for feature to be under MPL and meantime it contains fragment I mean which is under EPL +I did re check after Grants commit of with and this is issue is now resolvedTo be counter tested by someone else and set to resolved status Grant if you read me +And in mode with master and worker on the same node we have two addresses corresponding to the same taskid +Our DFS cluster nodes all have the same amount of disk space so youd expect mean of individual datanodes to be the same as DFS full but its not quite the same +This looks like a duplicate of +A client verified the patch I uploaded works to fix the issue +Alternatively it could be method instead of constructor but because root process pid do not change then constructor is better +was do to having in stack name +Still not sure what the issue was originally +As I have something that now works fine I dont think I will do it +fix it and release one week later +A copy of the email can be seen on +It is no more fired for the scrollbars +I was a little confused +I should really have spoken to you guys firstAt the same time some documentation on how to use Apache Chemistry in a Kerberos based environment would be handy +There was some issue with the file and somehow we were able to convert the file in DOS format then we were able to make the build +I am borderline about whether this one is appropriate for the branch since it changes the behaviour of +Ill take another look at it. +Im assuming its because there are two copies of loaded by different extensions or there somewhere I can put the jars so they dont get loaded by extension plugin or reportingPlugin classloaders that conflict +We dont provide this for other case we should not provide it here +py or we could move these into sharetest +See previous comment for the suggested fix +Templates in VPE are in general very advanced and user friendly framework and are nearer to be API than many other features in Studio +Ive updated the patch once again forAm I wasting my time submitting this patch to the Xerces development team? Should I just keep it as a patch +As of remaing work is mostly better annotation support which is under the jira entries +My personal pref is to call them recipes perhaps this is due to my background in networking +Lets keep it +jar file with new new files for the showcase I forgot one modified file It also fixes a link problem with the model driven example +It is a component dedicated to efficient computation not pure theoretical constructs +The context will let us expand on other items thatmight indicate how a URI should be displayed without having to modify the rewriting API +Youll need to copy them into your eclipse installation like so should be copied toECLIPSEHOMEplugins should be copied toECLIPSEHOMEpluginsLet me know if this resolves the issue for you +I assume you dont do this for performance reasons perhapsWell I let the issue churn on the back burner for a little bit but I dont see a slam dunk solution +Otherwise You will have to wait for the next release that maybe will include a rewrite of the +So my comment above is not really valid ill database value and the configuration value are availible at the related source code position +Thanks I dont see the NPE now. +like the change is correct +This is a trivial utility method +If you believe that there are lingering issues then the way to determine is with a test cause that validates your assumptions +Im ready to commit soon +Closing old date issues. +Im currently considering demanding either a type or an initializer for a field +And could you present action class? especially field you want to map to +The def keyword is going to stay +Yes Konstantin I meant access time for files for the getblocklocations RPC onlyRegarding Joydeeps requirement about recording the of last access I agree with Raghu that it is more likely a case for +Patches applied at +Ok thanksthat makes sense. +Hi Nathan +Unit tests came back broken +i have a working patch for this it makes an amazing difference! I tested it on a version of hbase with that META deadlock and it cleared it right up Ill post a patch patch changes the api for rpc +But it no longer applies +How about a builder +Yes please file a JIRA +Temporarily reopening to update release note info. +Simple enough patch +makes many claims and assertions that are not true +for putting this in +There are two possible solutions Let the onus be on the administrators to be extra careful while configuring +Retargeting fixes for SCR release +is used to communicate whereas is used to persistent info into fsImage +This is handled by the parser itself. +Ignore! +Ill use GRECLIPSE from now on +I think something is off myself +So I restarted the first instance and now it works there too +Also look into SCX which the ESB is moving to +This is resolved by +To remove the node entirely you can delete the +Neither xfire nor groups contain the aforementioned jars +committed this patch to +Thanks Mostafa! on v pending jenkins +Since Im the only one who has ever really noticed this I dont think that blocking is necessary. +bq +Done +Each folder may contain a normal script or an empty script or an commented out script +Request review for the Vivek! I run a cluster in my laptop with your patch and all three node lists looks good +Hi DanielMatthieu Firstly thanks for finding the issues with the +First the eclipse wsdl validator does not work correctly and will say that many valid are invalid +Accept the return from return header by selecting status Requested Return Accepted +Im attaching a Java JAR +bq +At least the fix I was going to apply for this issue was already applied even though it doesnt seem specifically related to +By the way Mark in case you are interested the here still has problems with kleene star as I have mentioned many timesSo wildcard ofabacadaba is fast wildcard of abacadaba is still slowin the same manner regex of +This can be used in place of the helloworld quickstarts current +on v committed. +VERIFIED +We dont need to test how castor behaves at failures +Probably my expectations for automatic recovery after a delete failure in the implementation were too excessive +I agree it would be nice to remove cached results as soon as they arent needed but it looks like quite a change Id suggest to move it to unless theres going to be another CR too many changes for a release imho. +Resolved so now closing it +Test are also not packaged Failed to deploy filehomealrubingerbusinessjbosswcjbossasprojectsejbtrunkcoretarget need to be added to both and mvn runner configuration if valid issue +I finally applied the patch on the trunk +I dont think there was actually a problem but please shout if you disagree. +Ive committed a fix to SVN +Tested +Bowen +Thus it would be great to have it fixed in the new release and I think it can probably be implemented rather quickly +Ok +It was made based on V not V +my fault +On the plus side though the logic minimally impacts the client and processing logic +fix on +Anything in or inside still needs to be done +This was happening in CVS HEAD +Still broken inWhat was the intended previous fix +page is submitted with validation errors the first hinputText is blank with an error this behavior is OK and the second hinputText is blank too but it shouldntRemoving the code mentioned in the previous post would really helpThanks +i agree we should have the extra test +A lot of development has been going on so I cant be on it +That would be really weird +adds support for some configuration from bundle context properties root URL of the console HTTP Basic realm HTTP Basic user name HTTP Basic passwordSuch framework property configuration will also be assumed as default configuration in the generated Metatype descriptorNote that I use a prefix of +this part of wicket has been rebuilt in +Sorry for first patch +oops +Confirmed the flush flag is completely ignored in +fix MVN build +All Hail the Big Central Singleton +Robert we cant have an external schemaLocation +Closing +If someone wants to take care of this problem be my attached file fixes the problem BUT causes a test to fail in the shell launcher +Another patch for Create and List methods and property +Beige design also by Nicolas +if you set something as invisible it should be invisiblethe regular pattern of dealing with this is to add the invisible components parent to the ajax the user has css that says displayblockimportant on that div who knowsthe only foolproof way to do this is to repaint the this is for usecases when a component is hidden initially and is later version still dont think this is a way to go +Images are updated. +Fixed into trunk rev +Thanks Todd for the comments +Committed at revision +I havent had a chance to test something like this so Im not sure what the actual problem is +Thanks! +Compound indexes can be useful but they explode in size far too quickly as you add index terms to be the only solution offered +In the case of Facility its not the same since I dont expect Facilities to ever be in heaps I put the changes in trunk at r at rThanks for the refreshing discussion +Not sure if it makes sense from the editor probably not + looks great +I have changed Brian patch and commit toHi BrianI made some changes for your patch +It should be fine interrupting ongoing splitcompactions +Well you ask the processor to do something not quite possible and which causesconflicting user expectations +bq +The code and issue remain here in JIRA +Patch for a SOAP configuration +Looks like Spring JMS Template have default configuration for eg +We have a financial EWS customer that by policy needs to install on dGSS tells me that this isnt possible at this time and I should vote here +im with yonik lets stick with primitive types in the response +Thanks. +Should we change thatSometimes + all old Resolved issues +All it does is preventing the product form showing up at the Enter Bug page +Patch for incorporate Owens comments +Inserting require nethttps into BUILDRHOMElibbuildrcore at line solves it +We could already provide a generic version for the Frequency class before the release we would just have to name it differentlyThis would also allow to keep the old one in case some users prefer it like thatAny ideas on a nameMy thoughts Frequency +Currently the renderer handle valid area checking +Branching is not a big deal really we just create a branch from the tag apply the fix do a release remove the branch +This is a patch for the Yahoo! distribution not for committing. +It is difficult in practice for users to decide upfront whether or not this flag should be turned on in production +Makes sense to me +Thanks +I have been working on cleaning this up and hoping to have a couple examples done soon that mirror along with what is in the code base tutorials have started the javascript tutorialabout of it done +Id love to see the discussion on the matter on why on earth this should be permitted +kpiwko Looks good +What will become of the freemarker templates? I am assuming they will be going away in favor of a more modern approach to content that I need right now to power the API is JSON but Id imagine there should be and xml variant as well and potentially html +Dennis for the pathIts probably the only solution to have that fixed. +Thanks for testing the patch Dag +More tests added and bugfixes including a switch from the md choice to the md choice as this balances more nicely with the code to work out the actual number of days between the two +I have attached a build output +add paramMap to request attributes in case those values lost again when error LeonYour patch is in trunk at rI guess you meant a variant product rather. +looking at this took me a long time to remove all the comments so please if there are any further changes to be made make them on top of the new revision of the patch that I am attachingThere were lots of changes to the code that were cosmetic only and in areas where there were no functional changes too which made the job of reviewing the patch quite difficultI note that there is a TODO in but I couldnt tell what was to be done +I probably saw only that it is a header fileStill I wonder how adding a new function which isnt apparently used anywhere in the patch solves the problemSorry for the late replay +Hey Jan I have already made available build with exception of can start with a build verification +Let me know if something is wrong +Patch applied +Filed. +Closing. +If you finish it before we release beta please correct the version +Sigh +How about? The toString method misses to include the value of the field l +this patch makes TermcompareTo support Terms with null text +Id prefer a more general fix if possible +Im having the same problem using the search API also +Another possibility would be to prefix the reserved view model keys with grails or something like that +Hit an issues where I expected a single result to be returned but an update count was returned as well. +It seems that its going to take some time before this issue is fixed in ZK +Yes it makes sense to log the log file name +So anyone is welcome to have this fixed +It doesnt honour that someone could also provide an own ExternalizablewriteExternalreadExternal or Serializable readObjectwriteObject implementation +I committed this with a bunch of whitespace fixes +bq +I have a couple of performance issues +I cannot answer your question +Albeit getting an IO error rather than a permission error is a little weird +It is a similar problem indeed +Bulk closing stale resolved issues +The problem is the Description assertion attempts to open a stream to the target namespace to test if it is dereferencable +Uninstall JBDS via shortcut +Hi IngridFirst of all please do not create issue in Jira before discussion in Users Mailing ListSecond from your logs I see two separate problems during execution of Cobertura during execution of Sonar PDF PluginMost probably after fix of second problem analysis will passAnd in any case Im closing this ticket and suggest you to switch to Users Mailing List to continue investigations of your problems +Disabling the orders demo allows the jaxb quickstart to deploy normally with the other quickstarts. +Open a bug if thats not the case +Resubmiting the same trunk patch for jenkins to pick up + they also have injection support which can be used to specify data providers to create test data based on the method name calling on the injection +Im attaching the updates to the +Added a note in the documentation. +Added test cases and a new patch +Sure no problem +LGTM +Changes have appeared in Latest Alpha Manuals so closing. +Thank You Everett +Hi MikeYes I like to have a try +Spring support is now being addressed by the team at Oracle +Thanks Daniel yeah fell free to provide a patch with the START END snippets +Concerns raised here by any review +Attaching performance results matrix +Is this intended as a replacement for too? yesOk I guess my misunderstanding was due to Owens commentDoug Cutting Sep AM I wonder whether tfile should start out in contrib until it is more? Without support for java comparators or random access it is not yet a replacement for +Fix checked in +Still awaiting feedback with the info I requested +IMO making it work with snapCount set to would be the best way to resolve this +Contrib tests failures are unrelated to changes made in this patch +the tags raise some interesting questions in my mind like why isnt everything in the KV implemented as a tag? For example id love to see TTLs someday +Promising +When I change this to compilation fails +Commit seems to have fixed this issue +For backward compatibility the old styled commands could be present as aliases +Updated the binding to match the way the ORB handles things +The license and each localization item should occur only once per each file +for the patch +The patch however needs to add a regression test case to one of the language tests to verify that the bug is fixed +Attachment has been added with description Test case for margins on regions +Thanks for tracking this down ChristianIve committed Chucks version of your patch in +Thanks +BJThere is already the possibilty to use a file so it seems redundant to me +Bryan thanks +Jason can you please verify this with the embedder +Thanks Harry +Empty result imo +Attachment has been added with description Changes the check to add the end instruction length to the local variable thanks for the patch Committed revision . +This seems consistent with Bills cactus results +defer all X issues to +ill attach jstack files +The idea is good and I would certainly consider applying such a patch +Can you link to where you saw this element mentioned in the Enterprise User please ask Andy Taylor if you need more info about to the it is described in the Procedure +If I add sonarsonar to the Maven goals Sonar is executed information collected and published in the dashboard +Please find the log attached +Submitting to Hudson anyways to see if tests are passing or not +Today I spent a while refactoring what they have and its okay for us to use their solution incorporating it in does exactly what you describe and integrates with Acegi to provide security but also more control such as security based on and events +Ive attached a copy of as well +Since everything in that directory is supposed to be images anyway simply exlude anything in that directory from the checks +No in the sense that I am basically just describing two independent trains of thought but you are absolutely right that there is no reason to have an applicationid and a clientid +Committed revision +As Java allows it and Groovy is derived from it a working Java grammar I was thinking it should be possible to support this +Good news is I have clearance from work to release our patched code. +Can this issue be closed +Thats why need it on too +patch committed +The defaultcatalog only solve this issue if your default is the same used in the table. +Does this problem still occur with the classloader policy change I mentioned above? Does it only occur on SOAP? I am just requestingwsdl +canceling the patch until comments addressed +I noticed that my first patch contained a bug anyway heres another one with a slightly different solutionThe unit test now fails though because of a NPE bug in it cant deal with a destination file that has no parent dir and it throws NPE +I tested the patch manually +Could you try with beta +If we ever want to let users specify it we could fold in Richs solution in the futureI plan on committing shortly. +I dont have a Windows box around but I do have Wine installed and it has the notepad application +Jeff what you said is New BPEL Process File wizard? Based on the latest codes it is ok +This is the patches for the hdfs JAR to bring the NN and DN under the lifecycle patch supplied with the latest version of +We should make commands so at least TAB completion works fully for one or the other ie addon install id search and and addon install bundle bundle +BTW this jira should have been a and not jiraAgreed +I have added a new snapshot which should fix this issue to the thirdparty repository +I doubt youll see a slowdown from this +implemented added option in configuration prarmeters and as an action +I can reproduce this +committed to trunk and +In the future please open separate JIRA issues instead of putting everything into a single I missed the Ivy configuration file in the plugin directorySee for the new Ivy issue. +You are welcome to add missing hooksRealistically if this feature is not contributed by an external contributor it is very unlikely I will find time to implement it in my free time +Thanks Dick! Thanks also to Guanying Wang who worked on this +ok sure that worksill get it me know if the double brace initialization in is not allowed then ill remove itpatch against my mbeanjmx skills fail me here but seems when naming the method set it is not exposed via JMX naming it crc does any clue why? naming convention rules i idea +Actually I suggest we rename this issue to be no failover should be default and create a new one for keys failover Erik talks about! +bq +I have applied the provided patch and added an integration test ensuring correct bindunbind order in Rev +Leaving it to +I changed the JIRA description as per my comment from st July and attached patch +One more comment +Yet the upper levels of the om will be discarded when the SOAP response message is discarded upon the next request through the service client +Like I said Im a newbie to geo search but it seems like the should delegate the parsing of the q param to some other parser and then it would only do distance calculations on the docset returned from the +So if this has nothing to do with Git why does it default to Usersburrgit? the default value might be Usersburrprojects with a note that says we recommend that this NOT be your workspace directory it will be imported to look like it is in your workspace but will in fact live at the directory specified here +Yes that is the process you should use going forward +closed wrong closed wrong issue +Patch on original patch on version of Olegv attachedThe main difference is it cuts down on the special cases in by creating a hard link to the new segment as soon as it is created +The tests for were based on from Branch +Is there a workaround for this bug since Im experiencing the same problem +Okay I just tried creating a new Fuse project using the wizard and selecting and it generates an XML similar to above which cannot be work with the UI editor +Linking to which actually already requests supporting nested Config classes as the default +May be you should use the option uENCODED since we need to encode the Array type +If its still fail could you please attach the logs +The files in the inPlace configuration dir was not copied to the deployed repo artifact +After also renaming both +Travis you need to add true +i just wanted to make sure people saw this issue +I think the approach outlined by Paul in is more appropriate + +Ill take a look when jigger fix for this profusion of logging +Thats what I get for skimping out on some kind of integration test +Youre right evgeny! Im going to decrease the priority to minor + +Both of these attributes are configured in the deployment descriptor +Vikas +To help usability in this area we are going to attempt to include soon +Attached the output of David Wheelers sloccount on stdcxx sources +I just committed this +I think its just important to get a name in now because I could sure use itI can make these changes to the component and serializer but am not sure how the project is run +Im not sure whats the best way to report this so just posting my small fixbtw I really like the new cleaner API +Regarding your question I will leave my thoughts on so we can track all discussions there. +Patch applied. +A similar approach is to catch inThis way existence check can be omitted +Im not familiar with the new project wizard for projects and youve already done this once for ESB project +Tested +bulk defer of issues to +Should be worked with the upgrade in if it more than trivial effort in +It has the same signature though so we could easily refactor the to implement at some point +The new patch looks OK to me +No problemJuergen +Note +Are you sure the upload got it right +write no longer writes the image due to a null referenceI fixed that in revision . +When our rigs fail to complete it is hard for us to have confidence in the bits +I know Gav is busy so I tried to lend a handCommitted with svn revision Thanks! +We removed fairly quickly and unless there is a compelling reason Id rather keep old around just a bit longer +Facebook comments allow additions of moderation before submit +This issue was a playground for me to check if I have enough rights to help with Git administration +Lets do that for +Thanks JessiePatch applied to CRYPTO module at repo revision rPlease check it was applied as you expected +Updated. +This is not ready to be committedIn I move issues like these to the next point release if I think they are generally useful for but not quite ready yet otherwise I unschedule them +Thanks! +I haveadded a to the repository that shows how you can solve this issue usingthe class. +The problem is due simply by the name of the + Btw +Moving back to Post Release +Thanks +xls include the key details and link back to this jira. +Committed revision I leave the issue open until the bugs are fixed +Nicolas could you check if there is entry for Everyone group in Novel eDirectory and if it is somehow marked as default group +I think committing patch version would be better isModal property should be protected rather than private +only hickup is that maybe jbossws dtds should be done via the jboss WS plugin and not jbossas +Great will review when you post it +That was why I left this openI think it is confusing because the patch on this issue has already been committed to the target branches +I was confused +Alas the build went further but still fails +I think weve already fixed this so unless you see any problems we can go ahead and close it +Thus can catch issues like BPEL attributes being set on BPEL where they are not supportedThoughts +as commented in having all those extra methodson root looks a bit awkward to me +Ive removed from the Fix Versions field of this JIRA +Hi AdrienI apologize if I have duplicated any of the features of jclouds +I shall run the tests and shall revert back +WDYT +reopened just to set the resolution to fixed +Uploading gzipped pig core jar since its M +Increasing the priority as gridmix compilation is catch the compilation failures early gridmix should be added to the binarypackage target itselfUser should be able to build the jar just by doing in ant in the srcbenchmarksgridmix directory +I see this issue multiple times in the master log on the apurtell cluster +for the patch +Yes it does +I think this is addressed with +Sorry about that +So best option would seem to be that you expose it as a real maven repository and refer to it in your so everything works have already exposed these dependencies in public CVS so I cant see why making them completely public would be an issue +try to use same znode to hold all disablingenabledisabled states +Although and do not work well I think the failover for is still considerableIf on one server down we need to restart or migrate to another server +Im too +ThanxFred +Maxim there is absolutely no regression when you analyse the root project of a Maven project +Seam FP doesnt work too +Thanks Vinay and Chris +Thanks for the quick fix and prompt to mention Ive added clear tooJuergen +Strange that the automaton author did not add this? Have you reported upstream +Aleksei the patch was applied to BTI branch at r +This patch stop getting type information via Interface which can be used only after jdk inside classInstead type information is gotten from directlyNow I will examine patch with derbyall +targetting for as seems like something that needs that this should not always check for modifications just check for modifications at points where it is not expecting them +Untested applied at +attach a patch to change ledger by providing size +nix expects a timeval struct in the setsockopt call for setting timeouts and Windows expects a DWORD + is a sample script with a class definition +I will post all this work when all the unit tests run for security against Ldap probably early next week +then goes to FAILED state either on AM CONTAINERFINISHED event sent by NM or EXPIRED event will then be notified and during of restarts a new attempt depending on the number of +But new development is moved on to tapestry so this is not important. +Suggest testing with a current release of the broker +Moving these to for now +Thanks a lot Hari! +Feel free to if there are new aspects to be discussed. +Im admittedly no UE expert but not clearing the selection on mouse down when the item is already selected is my vote thats what Windows Explorer and Nautilus do +Initial patch for review +I will commit this shortly as this needs to be fixed in + for patch +to trunk +Patch attached +Removed unnecessary code and made it more precise +I did add logging to that project but its odd if that would make just those two tests fail. +Looks good +Done +And markers are added if any problems are found +Further itself is not only an affront to efficiency but also type safety when used casually +Initially the project reports were included because we had little contentsHowever I also think that the project reports can be removed if there are no objectionNow we are getting new contents +No its waiting on a committer to review and commit this patch +Every queue keeps list of paged in messages which defaults to max entires +this has been resolved for a while now +It might help you while the error has not been fixed +We should also allow mapping of multiple security domain roles onto multiple application roles so a many to many relationship between the backend security domain to the application exists +Good find! I can certainly help test a patch or collect more info if needed +Patch for tests added as this is a change to the shell scripts +Will see if I get the time after +I did not understand what you mean by And further to that are other databases +only comment I have is about the change of the serial for +All assigned reproduced in or and verified in +All of them passed +I wouldnt use new though +Thanks Laura +demonstrate this +remove file tab at interview room or just separate the exit button at topbar so if we create a room we can just hide the topbar but sill have the exit button b +committed solution on both branches but since there is no confirmation of the patch provided solves the problem this will be let ll close this one as fixed if it appears again Ill reopen could be good to know which UI issues you have when upgrading so if it is a bug on myfaces we can fix it +Added apache header +xsd whereas the former maps to a trouble element in the +looking this aint as bad as I thought but behavior is a little indeterminate i dont see a check to see if region is closing or closed +Please review and if looks good please the stat file which will reflect the removal of the data directory from both the tests +Any chance you can help point out if something is missing in the current systemThank you for the feedback +but is what Jeff +ThanksI implemented a similar patch in HB. +Thanks Uwe I can only trust the XML policeman here +Reopening for at revision +Back to you to verifyclose +The test that verifies the fix went in along with the other changes for so Im closing this issue now. +Im trying to figure out whats wrong with those first +Patch looks goodOne question Filters run in a row scope their state cannot span rows +My opinion is that it is easier for an to select a directory than writing a class +Options if it does not worka disable webserver login orb provide +This check in the code can be added with WARN logged in the log file or provide a force option to continue operation in absence of vmware vmtools +Reopening so that I can update the fix version field. +This happening when the cache fills up was also something that I think washappening when I saw the error +Makes sense. +feel free to commit this Daniel +Im wary of catching and logging an but ignoring the problem is no good either +Can you double check and commit it +Username pinguine +As you indicate you could even decide not to store the sources but only the classes for this exercise is relatively simple and results in more clearly scoped modules +If this is true it explains why someone pointed me to this bug for the two issues Ive been experiencing with replication +bq +Same patch applies for branch +There seem to be no activity +this one fixes wouldnt happen to know which profiles to enable to run the tests under tests +batch transition resolvedfixed bugs to closedfixed +Integrated into +So I think the bridge is going to have to change to explicitly save and restore the view across the actionrender at which point we wont need to preserverestore the JSF attrs anymore. +Assigning to Heiko Braun in case he wants to do anything w the console around this forI dont think the Host Controller behavior will change but we can debate that after is out if necessary +Reopen to fix it. +I updatedpublished the site with the change +For example you can override method +Is that decision set in stone or would you consider a patch allowing that supportI also realized my earlier description was slightly off +I dont think it would work with Postgres unless one was very careful with the create statement and Postgres is very standards compliant +Thanks for the quick! +Also moves some files which do not belong to into appropriate locations engine or to branch +You may need to clean hadoop out of your local maven repo first if you have newer versions locally it will pick those up instead of that is present in maven central repo. +RohiniNot sure what happened +this happens to anything with tests that use the plexus container +the last comment on this issue seems to indicate that this is no longer happening +Thats a fair point and Ill concede the CSV point especially given the current scopeSo the question becomes how generic to consider making it +Normally on production environment server are configured with a default gw that doesnt permit to go on internet but on owner nets +Tom can you keep this as a separate issue for had this down as a requirement and I would like to hear his reasoning before development starts +Michaelis there anyway I can change the GWT defaults so that the button will work for rows less than ? If so where would this code change need to be made and then how would I submit this change to GUVNOR? Wendell +for +The plan is to stabilize it by the end of Sept +Thanks for review Bikasupdate with changes based on code review +But in the context of CMT you dont really need to control transactions yourself do you +Angela Note that I had to change a test case in which fails after applying the patch +seems better +linking issues as we need to resolve the patch before this can go to looks great Simon much simpler! And faster +Committed patch a to trunk with revision +Hmmm +We can followup with a general purge of if that remains a pressing issueSorry for the churn Aleksey +I will get to thatDont fear were taking your use cases onboard Probably with your issue fixed we would likely be restricting inclusionexclusion to classes in the defining module +committed to and trunkthanks +Closed wrong bug +We should put classes and in vectorudf +accept am also using the IMS Content Packaging schema and experiencing the sameproblem outlined above +Ill investigate +Steps to verifyIf you dont a in SOA Softwares registry itll create one for you SOA Software run through the SOA Software Integration guide and see if it with CP CR +Im not sure if we can do that ourselves or if we have to file an INFRA is going out in. +Ive made the first pass +Would something like this work on OS Xln index snapshot +It works for me now. +I think that might bring a potential issueI think after we provide our own method to insert an image we need to replace the innovation ofClose this issueRegardsDaisy +of the +Can you run svn diff from the top level Oozie dir and NOT from inside any sub dir like webapp +Chinese translation for password expiration alert and password expired message +Merged to trunk also. +A request for further information and a test case for this issue was not responded to +The test will fail +Thanks Hairong! +does not have to beIm also OK with synchronizing the directlyWDOT + +If you would be interested in doing that please create a new jira and we can discuss it on the dev forums +Reassign to Stan this is his area +Attaching Before and After debug log for this use case for reference +Based on David Jencks comments Im going to remove the derby version instead of applying. +Violated again in r +Can someone please include a stacktrace +As requested for Alan +Ive run this against the Sun RI and it passes all tests +I felt like adding both parts in one jira will be too hugeDmytro when are planning to commit the above change +Ok Ill take a look +This is a question not an issue +No response assuming ok. +hash tableAgreed that the details of this are open for consideration rather than locked down on choosing uthash +In case youre wondering what the use case is my application figures out some information at the beginning of the request and stores it in its custom implementation so that it can use it later in the request +Looks like i have a patch seems to work for me +A couple of minor things that I would like to see fixed before commit the indentation in should be two spaces not three and there shouldnt be any author tagsLooking forward to seeing how this will be used +Youre right I did some tests with hibernate core and we can make it play nice if we explicit the type somehowBut the reflection layer has to discover the type properly I dont know how complex it is +It additionally avoids raw types makes some fields final and makes the stopping field volatile +follows redirect and sends the NTLM type messageServer has no knowledge of the nonce that was used for the type message and so rejects it and sends back an NTLM header +I should add this version provides Order and no ordered data types +Verified in +Firstly I apologize for the strong language +Basically initialize add a new protected method for creating intercept on and call back on +Can you update the app to demonstrate the problem using a single +fixed the issue with revision +No need for a separate Jira for such a simple change +Does this still make any sense after the changes in the security subsystem +The problem is that the Manifests bundles versions were out of sync with that of the projects POM +Product constraintProductAKNumwhere Num HGThe first query will get the row from the base table whereas the second query will fetch it from the index +It includes a test for the protocol that fairly comprehensively tests the protocol +This has already been fixed for +Patch applied in released +Thus when the number of requests increases time increases in a near exponential manner +that was super fast +This bug was confirmed by developers of jTidy that are waiting to port next Tidy version +on patch +if you have access to Zos +I figure with the default MB block size thats a TB fileFeedback appreciated especially if these defaults should be altered +Additionally upgrade was not correctly evaluating if was running +Thank you Laura. +didnot work +Does that seem right to you guys +The restructuring looks good to me +Until then I see no problems with the package name that warrant changing it +Is that safe +Help in resolution is appreciated. +The is already included in the build process through the ant build script descrbined in the guide. +Applied +Can we please just drop this and get on with real work once you get a bit more experience with Seam and insight into state management in JSF and Seam you will understand why what you are trying to ask for is +The same patch with one small correction +Am i missing any thingRevision Author dantranDate PM Tuesday December Message add IT not able to reproduce the issue +The spec has this to say about changing the message listener on a consumerThe effect of calling while messages are being consumed by an existing listener or the consumer is being used to consume messages synchronously is undefined +After applied the patch in +Committed per Pauls request +Ive added support to try and catch this recursion +Please file separate one in HDFS project for test if you want +Version Released +and the patch was a missing catch for a socket exception +Workaround Put the whole tree for which you do not want the xmljson into a zip +For and UUID part I added a new class providing a set of helper methods converting bytesstring to stringbytes for client IDThe patch depends on thus will trigger Jenkins later +Im looking in the object but none of the parameters come right and say something like CSV +This one fixes those problems and had additional test cases +Take a look at other modules under trunkperllib for examples + +Closing. +As I mentioned in the summary the accumulo thing seems to just be a build profile problemso this should be good to go +Closing all issues +now directly subclasses tooSeems to be way more consistent though +The also fires up a +Okay the is being fired to early for async routing engine +Cool we were planning to upgrade to anyway +Fuhwei Lwo is looking at this for me +Just missed in migration to new Type base classes +Ignore my previous comment about the crash I was stupidly trying to run it using an RVM in the same directory as where I was building! +Expect once the stdlib works without needing the sourcepath other libraries should work too +Documentation for included and anticipated to Lukas and Sanne +Marking as closed builds quite differently now anyway. +Attaching a patch involves attaching a file named something likeTo attach a file look under More ActionsYour change is fine but distinguishing variables by difference in case is a spanking offenseMuch better to pick a more descriptive term like np for +This will be important ifwhen Avro becomes a TLP but could perhaps be a separate issue if folks prefer +Agree with +Updated patches same format as before +Fixed in s no more development going on for EJB on community AS versions +Once on AS there is no longer an issue. +Please move the bkjournal package to contrib +And this as well +u MYUDPGROUP +However I concern with the platform dependency of leveldb +I could see it going either way to be honest +Added chart that shows how much time are the Guvnor operations taking +Seems to be a little bit confusing to me that there are two directory which are pronounced the same way but looks again and so long Bernd +Lowering the priority to minor +Can you confirm if this is still an issue for Drools trunk +An oversite I just added itHows that for service! +I think we can move to Xerces c without any major changes to the system +Lets commit this to made a new version of generator with usage of matrix data structures classes +It doesnt come up with the camera the second time it just crashes +This bug has been marked as a duplicate of +There are too much different counter naming strategies in the code to include them in a quick refactoring effort here +Just wanted to let you know that what you described is exactly how I imagined a solution +HQ Tomcat is now JBN Tomcat +The missing headers are fixed in my copy +the API only says is this object convertable into an? and I return yes or no +Both should solve the race condition +Address the review comments. +Apparently it didntWell thats another way to resolve an issue. +Pushing a out now for you to try out +Nice catch does this relate to +If this update going to happen anytime soon? This might be a good time for configuring the databases to handle Chinese character as being discussed in looks like the current Confluence version is so we probably want to skip +See workaround description paste from Darrans to I am not sure if the workaround should be considered a resolutionthe optimisation is supposed to allow JAXB strongly typed accessto the fields instead of using reflection setting the system propertydoes eliminate the error you are seeing but it forces all access to bereflection based +Pardon me you already made the cheaper clone fix over in +No node should serve readwrites if there is no schemaRight but thats a subset of what Im claiming no node should serve reads if it doesnt have the data its claiming it has +I havent looked into it at all +applied. +What this means? You will correct this bug in +Same ought to be done for an extra newline that got in +Making scheduler inform the Node about containers that it doesnt know aboutFortunately I could change my unit tests a bit to reproduce the corner case also +What needs to be done for this to make it into the trunk? I mean any additional development or testing? Maybe I can be of use +Closed upon successful release of. +Also it is weird since this patch has only test code changes +Continuum sends email +Probably a better idea to use a proper connection pooling library like cp or Apaches +That is the reason I am attaching a zipThe patch file should be applied for merging the modifications to existing files +Hi Dag thanks for taking on this big task +Ok thanks +I have a similar problem +Closing this issue for now. +Patch that solves the the same happens when you declare static loggers with slfj +Great thanks! I committed one more minor fix in revision PS +Yes noticed about the patch and was working on correcting it in parallel +They can be swapped out for each other will no ill effects AFAIK +Please attach a simple war demonstrating this problem +Thomas +Closed. +It was a regression that I introduced into the code from a refactoring I did a few days ago and not related to the problem you are describingHmmm +Where is the actual patch +Since the realm gbean was found by virtue of being in a its running so showing that its running doesnt provide any informationIm leaving this open until I find out if we need to change the resource bundles +I can always learn new things! I think I can envision the scenario upload the config link it to the collection that doesnt exist yet then skip the collections API and manually create each core with +RonI attached patchesCould you check if this is what you wanted +Better inn with all At revision +This patch passes all tests for me +Ahhh! Magic windows backslash +Such a case can happen for order by query where we pull columns if needed for order by columnsNext I plan to make the code changes less of a hack and then run the derbyall and junit suite to see if we catch any failures +Regis patch applied at r thanks a lot please verify +Heres a WIP in case ill accidentally send it all to the litter bin +Also affects Branch +I still have to investigate but I presume the metaclass for list is a little special and is doing unexpected things +Dont bother +Martin said that this patch must be applied before is applied +Some mock classes for the adapter manager have to be located in the package to access methods +I dupd to this JIRA since it looked like the patchs approach was a bit cleaner +thanks for the heads up. +Patch looks to commit +Try to remove from your app andsee what happenedWhat is your packaging model? WAR or EAR +I named it +Deferred until. +No response in over six months problem with workaround closing. +But fail directory didnt go away and it removes some other files from generated and classes directory +When will this issue is resolved in a formal Hibernate releaseMany questions hope someone will answer +Fix for this has been in and since last October so Im marking it as closed +I think this will fix your problem would you like to give it a try and let me know what happens +Thanks for submitting this issue and fix TimoMerged into the master and branches. +A +The stopped with local storage on the host gets destroyed and the host gets successfully removed. +Looking at the description isnt cancel field is misleading to Back OrdersWhat about something like cancel or cancel +I think were talking about different things here I just tried petclinic with Spring and it seems to be working just fine +bulk move from M to M +Tested on mysql by jpechane +Closed Issue +ThanksArvind. +Hi KiranCould you elaborate on whenwhere to set categoryid and pcategory +To be honest this bug isnt so easy to reproduce as decribed because it only happens on certain circunstances +for the patch +I am trying to test this today +This issue has been fixed and released as part of release +I confirm that it is fixed in per Michals comments that the fix is verified as good. +Give me a week or so +Daan markmail does not provide us with an account they just get subscribed to all our lists by default +This is probably the best option create a new component type METADATASTORE which will hold namespaced keyvalue pairs on a basisWeve been sticking metadata into the statistics component fwiw +With no replies in over a month to Bretts inquiry and as yet no volunteers to do the work Im closing this as Wont Fix +I am truly sorry its up now +Applied in rev +In order to fix I should know what the file will look like after adding EAP and you please attach an example +Committed trivial patch to trunk and branch +Ill attach a sample app and instructions on how to recreate +Thanks for closing itGreat work all +Updated patch for handling NPE +This has been fixed on trunk +Lukas Youre right +The there only stores the stopKey because the startKey is implicit in the name +Will provide a patch soon +Maybe the right error is catched somewhere and then replaced with what we see this still a problem? There have been two releases of Gant since this was reported. +Ive made some tests in Friday +Hmm +Ive completed the component without the ajax features so for now it does now work with client side state saving +damn made the change for port but not for host although you can use +The class B may be available on the normal classpath so the class is not loaded through the groovy clasloader but through the +Hi hear you and indeed a full chapter on logging is of the questions that repeatedly comes up on the forums is how to configure logging though so a small appendix is definitely in order in part so that Costin can get another few forum postings by referring people to said appendix get my of logj etc +The uses now +SVN Revision +Anyway I fixed some namespace handling on the jUDDI trunk a few months back +The patch command could not apply the patchPlease generate the patch from the project root so that the patch command could apply it +Back to Steves v patch fix minor UT issue and address two additional exception swallow issue in Configuration +I was also working on this issue under taking a slightly different approach and not getting as far as KnutI was moving the methods over to the jutil package as this is very much part of that package and extends beyond functional testsThe other differences in approach were driving the data source bean properties from the and using a to define the data source properties not a Properites object +Only within Eclipse IDE +The issue is still present with following setupMaven version Java version have committed a fix for this in new has been deployed to the +st this is not bug so I close as it could not even be an improvementSorry but your patch does not make any sense to me +oh it was just one missing place fixed hereTest passes now +Reassigning to Arnaud +Can these be addressed in additional jiras at minor priority +Stack trace from agent crash reported on +Should we ignore public fields with annotationsI think that way we will be sure to not generate any false field with some annotation becomes a kind of m thinking about the way passes parameters to s done using the annotation works whether your field is public or not +Improved algorithm of setting escape state for connection graph nodes +Closing. +Patch that adds a have used the same template as in the hadoop source codePlease let me know if there are some correctionsadditions to be made +modify to access timeout as a valid responseIf I remember correctly if a bookie is crashed the operation would return could not connect exception not timeout exception +And post a separate MR only patch at +Attaching the patch which implements function calls on alternative stack for WindowsThe patch fixes the problem for both and Windows +from me +updated patch with no to include files +Will it fail mysteriously or will it be easy to know that the failure is because of the invalid syntax? If not I think it would be useful to make it easy to know this is the cause in some way +in the end i would like to point to the webdavssl support feature +Not quite done reopening +Changed to not extend Logger and just delegate to Logger +defer all issues to +Could you provide more information about how to reproduce this issue? In that way there are more chances this issue could be fixed soon +The RPC is accepted by the new RS but that RS failed to change the assignment znode and just gave up assuming timeout +potential performance issue +push out to +I didnt generate eclipse in my testingYes please enter a new jira for this +Indeed pretty clear +It would be good to distinguish between failures based on +Its fairly complex +Test results for revision patch has been applied Thanks for your contribution +Thanks Sid +hdfslocalhost +Not as pretty as render as but works and I was running out of time to really investigate how it should have been done +Ive just tried and problem is not attached +Punting to +for the patch +The code is the vestige of revision Thanks Amit +While researching your Jtidy properties patch i found that there is a property yes so used that instead. +I think the problem is a is a stopword and strips stopwords? Try using instead +When I used the same config for testing via the console I didnt get enough info to figure out the error +Theres way too many and its a real pain to reopen all of the tickets +The patch is useful as is but is the only way to actually optimize the groupbyorderby case to do the ratio thing as a conditional task? And if so would that be this or a follow up jira +Thanks Talat! +Trivial patch which changes rws to r for the file open mode +We are using the Joda DateTime Converters in our domain objects which could be contained within a Web Form + +The patch includes functionality and tests for adding dispatch policies for weblogic ejbve also did some refactoring for better code reuse +does not add specification and implementation entries by default any longer and allows you to override their values anyhow. +It would be good to prevent the NPE in this case though should be an easy fix +Yes and afaics the view itself does not need to implement it +Anything to do with is removed and some of what it does is incorporated in and +The current workaround is to build the template +I have verified it and it works +Thanks +we should default always to true +It seems to work +I cant find a R compendium in maven right now so Im using reflection + is a step forward to support +Thanks Jason +Modal panel is the component that doesnt depend on another so you can try to +I can use this to make Jaxen throw an exception not just return an incorrect value +for me the build works without the added repositories +If issues a warning then we can give them until to migrate away +This allows the to be generated when the correct auth method is known +From those reviews I was initially leaning towards deprecating until it became clear wed need to bump the rpc version numbers in both casesThe patch is broken out so it is easy to pick one path or the other +Can you please submit a new patch removing the deprecation and just changing the api? Please also remember to remove the checks for null in the toString method as the object is not expected to be null after this changeBTW I spoke to Amar about this and we agree on this point now +its arguable if it should but to be compatible with MQTT wildcards on topics it does need to. +JSF artifacts are now in maven repository. +The client just runs its client logic in a method annotated with an SCA Init +Can we replace it by Description is here like a javadoc comment parameter aliasoptional name that replace the field name requiredtrue validator expression String outputDirectoryafor field +Created and separated some tests into testing v +Can someone from dev review this patch and comment on its likeliness ofgetting applied +Hi Allderbyall Tests ran successfully +verified +added the issue to the right component and patch has to be applied back to as well because the issue is present there too +Update to match doneFurther replication tests as new. +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first +Any objections +Closing as worksforme +The essence of the issue that this fix caused is that row selections are tried to be remembered via +Let me know what you think + Username mikezzz +Added test cases for defaults processing into changes +I looking in to it +Anthony youll need to upgrade to. +Just skimmed over it +Does this happen using +It definitely should not come out as null in the error +Resolving as it is working fine in Vmware +This is my test scenarion +I attached npe because it is also not available by link nowNow I think I see the cause of the problem +Todd Ok +File which exposes the after release. +Its quite possible that my previous patchset wasnt quite right +Being in the log is evidence that there were no reasons for the update to fail if the data store is actually working correctlyAgain the assumption of correct operation at this scale may be too strong +This version is in svn somewhere under sandbox or something +The main piece of data in the features directory is XML +GavinI do currently set the into my Interceptor with Spring +Its always easier to remember only Infinispan than From this perspective it could be better to have something like wikiInfinispanid whereid will not be mandatory +Temporarily reopening for release note status. +All tests pass in this patch +So that means we need to load the snapshot on D with to send that Row to CRow is stored on D and A though +Create volume attempt fails because there is no image file available to clone fromAlso found following events in datastore historyRemoved datastore bddafbcdbbe from AdministratorDiscovered datastore bddafbcdbbe on info +Thanks Arun for uploading patchI will review your patch shortly +Only a ESB project creation wizard is finished any others wizard with ESB facet will not open this file so it will not conflict with other behavior +But I dont know what you mean exactly +I will update it if need +Again I had to get from usrshareaclocal not usrlocalshareaclocal Im not sure if thats important +I will reassign this issue for verification when the docs are available on +The report template file is located inside reports +bq +applicationidentitypkgI probably got this wrong but you get the idea + +Committed to master. +It seems would could communicate this better to save time in the future however. +Im an idiot who needs more coffee before doing anything Monday mornings +Above resp +Your bug report looks nicely detailed! The key thing to keep in mind whensubmitting a fix is to include a unit test! This makes it easier for themaintainers to have something to verify your fix against and ensures the bugdoesnt come back +My platform default file encoding isCan you set your default file encoding to with tried the current source from svn but with no effectBtw I use +A max of GB is still excessive IMHO +The problem is pretty clear from the log Too many open files your host is hosed +Rossen please create a Backport subtask for this if you see fit +I just committed this +I asked about this on the list +I committed a version without factory method for interface +did you have any large change in artifacts rebuild the index in that periodId like to encourage you to try or the release the hourly peaks will certainly be reduced +NET Compact Framework doesnt as yet support SSL so the Stomp client wouldnt be able to operate over SSL on +IF you run Camel on Windows but connect to a unix FTP server then I suggest to use separatorUnix as option to tell Camel that its a unix ftp server. +Thanks +Peter I get the ANT warning while starting the bam examplerunWarning Reference has not been set at runtime but was found duringbuild file parsing attempting to resolve +Why would it be OK for instance and revert to bad URL and now it look OKceo +The goals of are beyond job scheduling and include local resource management and optimization +The changes applied at have been preserved. +Reopening as a doc issue +was released last month +Detect the protocol http mailto etc then you can refine the regexp accordingly +the fix looks fine +Converting to upper case fixed the problem +added remarks +has been fixed by Darrin +Moving these out to which probably means they wont go into +Didnt know the ticket existed +Ifthis is not what you mean please let me know and Ill see if I can reproducethe steps +Assigning to Jody who is the official maintainer of this module +Apart from refactoring this will help in some speedup in compilation as well +included in estimate time for outlining time to develop user story writing time time for preparing note that this section is unlikely to have anything on writing plugins using Ruby since that support is still very much a work in progress. +Communication service usage was removed from code in version so it is safe to remove the whole project +JFYI junit based tests suffer from the same problem it seems that in my test case just the different timing hides the problem when changing the test from junit to junit +Thanks Derek! +Because of we can run standalone BSP system with local filesystemOf course the cluster can be implemented later +The mirror is now available at git was fast! Thank you Jukka. +It prints out Shutting down on signal multiple times as I repeatedly hit however it doesnt actually shut down +for the patch +Lets either do this in or break it in +Also seems like it would be useful to include the readme in each platforms instance of the hello world project +I think it will improve self managing aspects of the cluster +The latest design document introduces which will manage all the journals for the namenode +Im not actually sure what this task is +Are you considering rekeying +Ill try to come up with a smart definition of the heuristic and see what you think +Im not yet finished with all documentation stuff +Cannot reproduce this +Thanks Amareshwari +Updated the description since it does not work even with Jetty +If not it uses the previous rules with the caveat that there might a error if the schema is ambiguous + LGTM +Did you gave him karma +One more elephant please resolve it as WONT FIX or FIXED +I think its convenient to include java files in the source tree +After installing the error persisted but the error message show me the way + This bug has been marked as a duplicate of +Archiva responds with a HTTP on the request for that researching cause now +A simple t hard was it +This is just my own observation based on my data +This idea seems to have a significant relevance to the Whirr project and I really want to give my contribute through Gsoc +Update of the previous patch to also allow any valid character in the field name not only in the query string +After a restart the created connections are goneThis was detected in LDAP studio under Windows XP professional +Fixed with unit test in cbcdabbbbafcabbeeI was unable to make an IT of your supplied testcase but I think the fix covers your problem +I added this to these two classes and when a block cache is asked for during the creation of a Reader it skips it when blockcache is set to false +is quite an old JVM +This bug is hitting me too +I was able to create a new jira normally today +The type should be changed from improvement to we got this nailed and fixed +The last failure will be resolved with JBCACHE and testCompatibility showing this mismatch in latest test runFAIL serial error for current and testCompatibility mismatch FAIL serial error for current failures for CR. +I agree that it would be good to see this in is it really that hardto migrate this to RF? it doesnt come across as the sort of thing thatshould take more than a few days to do but then I know little about theCDK at this time +DTD style descriptors are not supported for EJB deployments. +In the minimum should be changed to use localhost for RM instead of +Id rather have a slower clean test class than a dirtier fast oneIt depends on the time difference I thinkIn this case it doesnt look like a problem +Closing this as NMI +Would be great to have fix or a workaround that does not require patching buildr +Note I had to do some changes to the patch since some of its code was moved into this patch +Its a good news but its a bit strange +I dont think it is showing there are no specific changes in thereWhat I think it is doing is showing but doing it twice so there seems to be a lot more content than there shouldIll put a patch in shortly that from my end at least seems to cure it but I only have the plugin itself to test with I have no real world use case +What exact exception do you get there the nested exception of the? I assume its aThis means that you havent even hit the database +It is called mt and exists both in cygwin and in MSVC binaries +Executing a test GET or HEAD against the target URL to ensure authorizationis another +I was thinking saving row cache into different files for keys and payload so in event of shutdown we could just persist keys and on startup we could check timestamps of the files and decide if we need to load it from CF or it is possible to read payload from save +DominikDo you still see this in tests? I think not +The current behavior of the apt parser is the historic origin of the mess we have right now +AFAIK the current implementation is not the one discussed as the desired implementation in the mailing this over as it is holding up a release and we have had no response on the mailing list +There are even classes that have zero test coverage although they seem very important +Committed to trunk and Thanks Vandana +You are probably getting duplicate CSS and sent down and causing JS issues +I think I had noticed that also +The patch also makes the appropriate additions for openejb to get it up to speed with jetty +Use and. +if saveNamespace fails in one storage directory but not all of them it still throws an exception back to the caller +Are you sure that you submitted this issue to the right project? invalid as this appears to be a Javassist issue not a Spring Framework issue. +The FAQ use the word query my situation when linking to the docs but the docs does not even mention the word +Made further changes in +Ill try to create a reasonable unit test but first I have to do some further investigation in order to emulate environment and communication correctly +change will carry forward on next synch of plugin +Im fine with the k limit since Ive always thought of collections as places to denormalize small amounts of data not a replacement for creating a new table +The pc file is installed but not packaged +It then gets reported by with a failure automaticallyokbq +Please reopen this issue +First pass on a patch + Username vbkrishnan +IsuruPlease dont add a new method to please change that into a property am talking about setgetthanks should have pluggable senders in addition to pluggable builders +So please use your EclipseSTS JPA entities from tables wizard and point it to your db and generate the entities +So for many users of a table and many new users of a table they are using the same table and want to refer to it as such it is the canonical tableBut sometimes the table was originally named with crummy names and itd be better and cleaner to document and train new people on the appropriate namesViews eat up the namespace and provide a level of misdirection that is not always desirable but here are the two biggest limitations of views SELECT is not fast +war file appropriately +this was observed while debugging a Sling test thus the package nameif you want me to provide a patch against the JR source tree please let me know what test to put this in or where I should create a new so far adjusting priority +Hi UmaOne more scenario where same block with different genstamps can be present at DNScenario Both blocks in current File is written to DN DN DN and file closed with generation stamp +Usecase is to be able to run the tests without having to rebuild them checking if other components are affecting time we might find some tests are not built for that but then we should filter them out or adjust the tests +Then dependencies need to be adjusted so that things still workExactly! This patch just does that +I guess I should extend that a bit +this is a good goal but the strategy mentioned here would have almost no downtime and very little code change +As I see it it is supposed to be this way +Ive filed upstream +considering tools page we could accesstoolstoolsetcScheduling for +The documentation doesnt state that this functions throws an exception leaving me a bit confusedThanks +Verified in JBDS h Fedora +Patch doesnt apply on trunk +I just committed this +That is the problem +Seems like I was wrong +Minimal test case thats definitely a bug +The Import JSF Project wizard still doesnt support it in beta + is also necessary but is already an umbrella parts of this should go into +The patch appears to cause mvn site goal to fail +im getting stung by exact the same thing +How are those two bounds differentThanksRick +Maybe it should be legacy approach that is no longer supported +verified. +Thanks Naoki +Thanks Bill both for the patch and your other feedback on our documentation + Stack I committed to branch and trunk. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Links in EULA and in JMX Security screen now open in separate window. +somehow dont think thats it +Thanks Yonik Ill fix it +bq +This is useful for reverse engineering feature hashed modelsThe idea is that the hashed encoders have the option of having a trace dictionary +verified in Tools also with using of to write release note now release note text has been written and fields set +What is the sync? +Implemented an abstract +Changed classes in +I can create them if you wantNo thanks Ive taken care of it +I have traditionally used Ubuntu package names in that file since thats what I install +ant binary fails in factThe following targets need unless datajoin jar mumak package streaming jar test vaidya init packageCare do this in this issue? Or another one +was ok with trying unlimited keepalive as the default as the reasons for closing the socket are non obvious +Verified by Ilya. +Hence removed componentCorrected bug title accordingly +This is the way it is handled in the server so it would be better if the tooling followed better solution may be to have check whereby the user can be prompted if there are only gateway listeners +The standard industry trick in this area is to only index the first N characters of a long string and then do magic in the query plan to use this info but know it cant get the actual value of the column from the index +That would be quite heavy to do relying on at least one rd party libraryWhat would be the benefit +for commit +Reverted the patch until we address Gregorys question +looks good to Eugene +To upgrade this correctly we do need to upgrade Xerces both in the main lib jars and within the plugin +It is introduced with and it is not a regression +oops only votes are from the pmc +I fixed this issue +Hi Andy seems like this functionality is done and AS is able to handle HQ in collocated topology +Thanks for the great software. +The Greenhouse server does not yet support posting geolocation data for a tweet +I agree a better test for this is to try executing of the system routines as a user other than database owner +from me as well. +I agree that providing an option to specify a custom loader in cases where the loader of the service class does not see the runtime classes is useful +The bug is marked as Critical as it affects resource limits +Moved from to just after merge of old branch into TRUNK +Andy do you have an outline for this chapter? Can I help in any way? +Was it backported but then removed? If so the fix versions field must be fixed +Also your comment in is slightly incorrect the Python implementation can be built and installed without setuptools but cannot be packaged for distribution as an egg without itRight +Thus the server does not receive heartbeat +Hi HeathIts OK for me basically I dont think we need the whole original but if you can indicate which port cant bind that would be great something likeIt is unable to bind to port xxxThanks for the patch +Srinivasan I concur with Stuart that a event isnt advisable in this context but I love the idea of an event of some sort being issued one that would inform each terminated conversation of the exact circumstances of its termination +Indentation of log +Looking at the patch for final reviewcommit +Jody can you test if my commit at r works for youOf the two proposed alternatives above Im not happy with any so I went with a plain hack on WFS being I dont want to pollute new stuff with deprecated filter stuff nor Im planning to port the current WFS datastore to the new filter interfaces as itd be too much work and the actual thing to do in the module is to port the whole WFS DS to the new design used for news Gabriel it worked I can render and edit again! +My receiving terminal was under osx which fully supportUnfortunately in our project we are forced to execute tika under windows and im trying to extract text only arabic char +Proposed patch for review +I think theres a deadlock scenario if you basically synchronize on the service registration +Back ported to closing after release +I will you a copy of new one and remove old ones +verified at rThanks Tim +should be fixed too now +can any body review the patch +I have some AOP advice applied to my controllers but I need to keep cglib in order to generate these proxies +I unfolded the function myself when I was writing it and realized the fact that function unfolding made a lot of difference after the code now +check columns timestamp values I bet they are from the futureCan we fix that on startup for other upgraders +The main consumer outside Spring of XBDP is Spring IDE and many of the changes I made are to accomodate features that Spring IDE +as well +it +Special thanks to Brian Minchau for providing the patch for and +The issue exists not just for elapsed time but also for elapsed time shuffle elapsed time merge and elapsed time reduce +Tests included wiki page is created +No one is blaming you or +For rows that dont contain CFA the code has nothing to load thus dropping these type of rows quickly +Im actually going to start a discussion about removing append entirely on +Integrated addendum to Thanks for spotting the problem Lars +comments on NavisCould you kindly refresh this patch as it does not apply cleanly anymoreThanksVikram +The remaining content of the path component will not be modifiedPlease your application against the latest SVN snapshotOleg +Attach a patch +No tests as just a pom file change +I committed the patch to +larsi am less enthousiastic about thomas proposal +Synced with trunk +The sameexact for FOP and Batik on my Win machine work without any problems +use looks like we accept jmxport as an option but it doesnt get passed to s d rather create the object with the correct final values than rely on mutating them later but it does work functionally +Thus both these components should work ok with a externally configured replication policy isnt it? Replication policies may have other requirements other than number of racks for example +The user has some other ways to read the entire message hover mouse over the message in one second a tooltip with the entire text appears transfer focus to the message area by clicking mouse in it or by repeatedly pushing Tab after that the user can either navigate cursor with arrows making last lines visible or copy content to clipboard by CtrlA Ctrl dialogs in Eclipse inherit which implements these features of title area and if some better features may be suggested they may be as well submitted upstream for improving +Preliminary version +Ill wait for to be s needed to improve to use tmp before merging. +filed separately as as wont fix now was most probably a problem in surefire plugin. + +Fixed in aether will be fixed in maven when a new release is cut that includes +Hi Hansthis should already be fixed by or are you saying that that commit didnt fix the problem? for the quick response +Just committed patch to fix items you list +It was applied to but we might want to apply to +There are both periods hyphens numbers and strings +I have verified that this repository metadata file is created on startup of the meclipse plugin so this isnt an issue with the maven core +It is not critical for CR though as i the error only occurs in test code and ii the second tests add little by way of coverage +I figured out my prior post by incorporating NilsI am also having the same issue with the C client communicating with the C server after the third request throws an exception on the client application +The compiled code has to retrieve the value as an Object via reflection and then unbox it to a boolean +Unfortunately the buttons cannot be disabled in JOPR +Thanks Paulex +Already applied. +if so can you closeresolve this issue +Replaced deleted patch for historical pusposes +Are you saying the field is not being indexed when the keyword indexing is done on your instance? This really should be working +Essentially the problem was that we did not update cancellation tests to use getMembers but instead used old Transport lets integrate those tests as well! Where are they +I could not find separate limits for regular files and tcp connections in LinuxThe rpc layer needs to limit the number of tcp connections to be a little less than the max fds +Thanks for this Roman +Mass close all issues in resolved state that have not received further comments in the last month +This is really disappointing +So in your case you have no feature types but you have coverages right +Thanks Ning +The stepExecutions reference to Step and s reference to Job should be removed first +Christian If its unrelated please attach it to a new issueDennis as a hack this is fine I am just somewhat reluctant as introducing this means that we officially support this feature which was never the intention +Hope this process is a smooth one +Actually I get your exception now +I imagine thats related +Any nested maps however although printed were not similarly treated +I will double check all works as expected when I enable validation on the bc and let you know how it s DaveI just mark it as resolved currently feel free to reopen it if +This allows for checking whether enhancement is required +OK when you get to M data points itll take seconds +We could use this as a fallback mechanism in case the locale language is not supported +I used this pattern string and it was able to find it when running the test within Eclipse on my local machine however when I tried to deploy to Weblogic this did not I am using Springs EJB Support classes and I am defining in the the env entries for ejb +ok the UI is definitly not listening to the version property and does not handle relative +I can see people using this too and i dont really have any problem with allowing null values to be keptset in the context as long as it still passes the test suite +I verified this locally +Patch checked in +I think this is already fixed since I am not getting any errors in building surefire +Ill get it right this time +This error was occurring because of the mismatched responsibilities on the and Configuration +Fixes the issue +Id be interested to see if all the Highlighter Junit tests could be adapted to work with Highlighter and get some comparative benchmarks +and even more important a failing testcase +This contention will result in the persistence layer throwing causing the encompassing transaction to be rolled back and retriedIn order to avoid this point of contention it is important that all instance variables are initialised before the process definition introduces any concurrency +Verified on EAP ER +Could you try removing generatedtrue from the table element if I remember correct when generated keys are being used we will ignore them when sending the data to rdb +IT updated +The next step will be to simplify the api exposed so that using it will be easierElliot do you have something in mind for this? While working on I also noticed that needs some simplification as well +Im inclined to mark it complete as weve done all that we can for the release +Great +Marking closed. +Closing resolvedwont fix issues. +Fix attempt to read from locked file that fails on Windows +there is no instanceof check in the patch at +this is priceless Carlos thanks a lot! maybe it is needed just a lower alpha level to put them in the backgroundcan you attach please the sourcesmany thanks in advance for the prompt help I dont have the words to say thank you +We managed to reproduce the bug in at approximately when we moved messages from the queue to +Barring any further considerations that may come up the fix should be available the next release +This would cause Weld to see dupes I think +Closing this one off. +This also passes with some changes to avoid using and nested connections +Sorry Miguel I didnt get a chance to check it out last night +Fore sure various scenarios are possible +Ill add a test case to show thatt this decomposition to be handled by the XML parser? Or may be by Model builder? I think SAAJ Text object should not have to parse its content again looking for such compelx possiblities +Looks like this is handled by which adds a configuration variable that lets using standard sockets in +You are of course free to use itGreetsMarinus +For new improvements we should create new specific issues. +FTL files still need to be checked for redundant DOCTYPE declarations so Ill leave this issue open. +Looks like you have s alreadyI had put up some comments in RB +If TX servicesz are disabled there should be no deps on jta +Patch broke some Temp Destination functionality as the class has its own init method +Whenever we fix any issue we will move it over to asyncweb for considerationThis issue has to do with handling message bodies +Tests for and idl files and generated classes added +What does strike me as odd though is that your test has no schema at all hence the uuid +This happened just once and with an older version of the JVM so I think we wont fix it unless it comes up again +Ill apply it +Patch checked in +I have removed cglib from spring and jeebooking examples +Updated patch for version +Folks I spent some more time on this trying to straighten out the legal stuff +Its sort of similar to a table you cant insert new records but you can update whats thereCurrently I dont need to rebuild or copy these tables but at some future point itd be nice to specify how they really work in the modeler. +An example of integration with crunch +We can remove it and look again at this if we get user requests to support JSP on Tomcat I guess. +partial Datanode log of INFO of blocks got processed in for is attached +I have some spare time to try this component to route FIX messages from FIX gateway to various FIX sessions through Camel +Current test pass. +Thanks Rob a few commentsin mapreduce is looks like we arent packaging the sources forYou might also package the sources jars +relative path issues as well +I believe the change is safe it didnt break any of the unit testsIf you know of some properties that do need to be copied perhaps we can some up with a cleaner way of handing them without blindly copying everything perhaps a new plug point would make senseEqually its not critical to clean those methods up right now so if you like you could apply the rest of the patch +MattCan you please attach a sample message which is returned from that proxy service so that I can try thisAt the same time if you can attach the synapse debug log that will be helpful for the diagnosisThanks is the web service response from the end point service which has a reference to xsi within itIll follow up with the log filesMatt +Just a page style name in enum is too simple it should incude more responsibilities +More over the API should discourage having a separate topic id the experiment id should be unique across a deployment and for all tracking purposes including event topic we should start using the experiment id +All subtasks are done so Im closing this issue. +Committed patch +Martin if you still see issues using next weeks release please +Please verify +Ben I have a feeling that this is still an issue +We depend on a functioning get in thrift +This should be fixed now could you check with the latest from SVN? +has been created to ensure that we also update the HTTP reference chapter. +This seems a reasonable RFE other customers would likely benefit as well and is with our focus on operational eas of use +Anyway I am writing my TSS article now so itd be nice to have the code in my article in sync with the current API +New patch which forces fields to be required have a default value in so that it can be changed on a basis +There are unused imports in +Do you know how to fix this easily or you want me to look into it +With this it works for in the ide and in the command line +However its possible that the same command was already forwarded to us so there are two nodes we must avoid forwarding the command to +OK great Ill test that +Works fine +Are you sure about ErrnoENOENT? Looks like a weird errno to report esp +Because the clean plugin clears out all the classes etc which are then rebuilt by default when running mvn siteThis can waste a lot of time +The fact that you cant build unified is not necessary a big issue +Can you give more details on what dependencies are missing and etc +updated version with fix for the future use with arquillian +NachoYes this would be a good improvement +I would rather saty away from providing ever possible combination of options in the Expression factory methods +I changed this so that it uses a path convention instead of an explicit path +The referenced page has been removed from the community site and consequently the issue should be closed +Thanks Nicholas I just filed the JIRAGoal for that JIRA is to move the atime and mtime fields down to i as this will save the memory consumption with dirs and also support setTimes access only for directories +You should see rvztr tkrfrgp +Can someone test if in is posibleto compile and run DELI using java? I am not sure but I remember it dont run. +The resource bundle of a plugin is accessible only if you open it +Something to discuss +I am keeping thatThe issue should probably not be in JBPAPP maybe TAGMay I decouple the BZ somehow to stop unrelated comments being copied to it? +Instead of checking for we should just check for null because that is what we want to guard against in this jira +Friends dont let friends run WindowsI wonder if the line ending thing is the issue here +It has been open for more than a year and is very serious for those of us who need multiple mail sessions +This is a bit of a special case since most applications dont allow a user to authenticate within a session which is already authenticated with a different user +At least the behavior should be different since I could not find Could not get listing in the ls codes +I disagree that interning of Regions even for array literals is a reasonable request +At least make such hashes optional +Did you configure a token store +jarShould I file a new issue for that +Not sure where this leaves the ve just also removed the patch from my local grails seems that this patch breaks the wiring of spring resources into grails also discovered that the post processors are executed in unpatched grails if the beans defined in are only wired into grails controllers +Generated a wrong patch +The attached patch for review implements this proposal +Ive created a class and changed to use that instead of +Thx guys Im very happy on buildbot! +Its dangerous but someones gotta do itWe still need to update the underlying xml from which the web pages are created +Fixed in r on trunk and r on branch +Youre welcome. +Committed to +I suggest thatr equests with unsupported Accept headers return Not AcceptableSuch responses provide a list of all supported formatsbest +I have added the test cases to exclude list in +Not completely happy with it at the moment +Fantastic +Along with that patch we should also probably be more explicit regarding our file sections +on mailing listNo +Committed to branch and TRUNK. +With doc changes I meant User Guide changes. +Thanks for the quick review Arun. +I left deprecated copies of the old public methods behind +All subtasks except one are resolved +We should fix this on at least as well rightWould be great if you had the bandwidth to fix that +Its been a while since I wrote this patch Im little off the topic now +the package is version specific so it cant be loaded in +Looking at the catalina source I see URLs are being produced pointing into this jar and theyre being registered with digester +It will then look in that directory for a file named after the function that it is trying to resolve +Why the need to get the scorer etc? I wonder if it would be more efficient to just have a simple Collector that did the work and we skipped scoring etcThis allows subsampling by the document relevance +chmod in shouldnt be necessary the committer will set the exec flag before committingNOTE to comitter be sure to set the exec flag +So yeah no bug in +Thanks Gavin. +FYI I am not calling the same from two threads I am calling from only one thread its just the calls are made very rapidly like calls per second so one call starts before the other finishes +x +Resolved +Moving to Beta since the Beta console is in +I am not sure if there has been a change +There are some public methods on the JPA config that Im not sure if we need to expose to users +Optional means that the component is not required forstandard use of the project or for the project to achieve a desirable level of quality +I filed to do this work on. + into the as well +Leaving the namenode up and running without the thread is not an option because blocks do not get replicated +Its in + +it also doesnt solve the problem of we only want a select group of nodes inYes it does +Ill update the Wiki pages accordingly. +It passes many less +the patch actually also fixes a bug in maintenance +Moving forward it should never be needed +thanks Roman +Moving to +Was afraid Id remove too much +commited rthanks Rafal +gavin you can decide on the priority for this issue +trunk +I assumed it was deliberate and that a Jenkins Admin had to do this for me +i downloaded the release tomcat package as is +I generally also do not use the webapp directly so thats not uncommon +It is a completely different build setup on EAP and it ishard to know if any of this do not reopen this JIRA it has a mixture of issues thataffect different components different people involved and it is hard to track the problems +Sample works ok on Websphere Environment Application Server Linux RHEL the issue now that it has been resolved. +That breaks some of the utilities + +Approved for. +I was actually thinking that there should be a single try block followed by two catch blocks no? You may also consider fixing the other two issues Findbugs is complaining about +Please investigate and determine what is going on +All builds succeeded after removal so marking as closed. +I remember something about Extended Jar now but I couldnt find anything on the net that looked like that +I agree with both Kevins +Even though this is older perhaps it could be closed as duplicate of? +There may be better metrics for that +Any chance you tried this yet? has the fixes for reverse and redirect map so really curious if this is still a problem +Sorry for the delay Radim +And there is also a session in comments from Google drive +is a final method and cant be mocked by MockitoThe Mock classes cant be moved to a different package +Added tracing for the numb of rows for each indexed value and what index was used during the query to highest +Exclude list feature typos and workarounding minor EUT scripts issues on Linux I found +Closing all resolved tickets from or older +accordion +same trick for +I guess ETA will be for the four XML schema types mentioned above +a few versions after reportedNot sure how or whythis can be closed +Okay committed to x and x things should be in better shape at this point +Patch toThis new feature is for the javadoc moved the issue +committed to and trunk +Hi PhilippeOf course Id never come to the disgusting idea to provide code in another way than a patch because this would violateHowever one could take a look into the attached maybe accidently some classes slipped intoRegardsKarl +Both to already use JSF with an appropriate description or create a new issue +Ok so this ticket closed with resolution wont fix and Ive created another one for . +Please replace them with what you have I can my changes if necessary but it would be nice to have the revision history back +can be configured to run on mode as well all you have to do is addruby version to your or configuration file +Replace old with new to old forrest site +Strip them from the source document +Patch +It should be fixed in todays kit +The reason why we built them as separate daemons is mainly because we want this to run on multiple clusters as Dhruba mentionedbq +The main query then runs and the post filter is appliedIm exploring another scenario that will perform x faster then the current pjoin +I have tested it locally and with this fix we can use the axis with Synapses nhttp transport +fieldName fpfGewicht type field number class +the executor approach is looking really good but thats worth keeping in mind down the road +We know we should have set it to be the system property and will do so in the future if it doesnt harm existing users +Changed the java code to remove the Exception suffix from faultseg +Committed to trunk +I have created to track the permission issue +Resolving as not a problem because ARQ follows the SPARQL federate query specIf the space should change so will ARQ. +filtered is a list not a set so correctness wont be changed but it will be slower to use it +This should be closed me thinks. +Correcting incomplete the required the required feature +So far it looks good +That is possible but you have to specify zkhost as a sys prop on the cmd line thenAs Ive said the file is optional it will get there because the user decides to create itI think all of this is pretty simple Im happy to implement it +There is a new script SE called which makes effort ofIt has packaged groovy and Rhino JS. +Yes a table with two distinct surogate keys is not justification in my mind I can just hear my DBA friends cringing +Tests ran cleanly for me + removed in r on trunksvnignore updates in r on trunk +If you still want it open a different issue by preference in the JBIDE JIRA. +all code examples used with callouts are not syntax highlighted by purpose because that doesnt work with them +Thank You +Thanks Brett! The patch looks good to me +OK guys I agree with you I see also any benefits to have those almost empty languages + +The data passed intofrom the setgetextraElementat routines were declared to be axisarraylistt +So the curl operation ld becomecurl XDELETE httplocalhostindexgenerictypehttplocalhostchemistryfilenameinstead ofcurl XDELETE httplocalhostindexgenerictypefilenameLet me know agree with Karl we should merge this branch into trunk and then I think that it should work correctlyAnyway we can fix new issues quickly without any specific issue related to the CMIS server +Above patch contains the complete set of tests as discussed above and behavesas expected +Looks good +If your validation code executes any queries you may want to temporarily disable to avoid your flush entity listener being called recursively +Problem was solved when I rolled back to M without modification of my code +modified to bundle in all resolved issues. +I think we shoul add cleanup code in MCstop to get rid of this thread in the case that the module is unbooted without calling stopMaster or failover first +Its really up to you +Changed components to MiscellaneousLinking to and as duplicates +Yes that is the exact issue +Flip spent several days in the user IRC channel trying to deal with the load spikesDoes changing this patch solve his problem or are we assuming that +If we could avoid that it would have a substantial impact on overall performance +The save menu option is for editors I kind of thought changes to projects were always saved? So Save is not it is not shown as an editor we will need to use Export Project instead of Save kind of requirements come from the eclipse user interface guidelines +Since I wrote a question to Tom +Thanks for the comments Knut AndersIm wondering if you could create a separate Jira for these issuesMy plan was to create a utility class that works nicely which obviously requires some more work +It improves the situation enough to use but we may need to this bug later +I too was able to get it working with a naked test application +Shouldnt this be fixed for +Marking as RESOLVE LATER pending course of action +We could change the modtime but thats a little perverse +It would be good to evaluate what work remains on this issue. +Attaching patch for SurendraPatch looks goodThanks a lot for niticing this +But I have extracted the linux file and ran it no problem on at least different machines from the one that it was packaged on. +Maybe you just ran out of items and didnt create a new reader for the next step execution +Fixed the javac warnings in earlier two testcases that failed are +Absolutely +Thanks Shreepadma +Yup exactly +Patrick? +Ondra we originally put most of the instructions in the root file but people complained that there was too much navigation back and forth so at some point we decided to copy most of the basic processed into each individual quickstart and only link to the root README for unusual or complicated instructions +Id like to ask folks what is wrong with the notation currently defined by the SCA Assembly spec which follows the WSDL specThe advantages I see area Were not inventing anything new here it simply follows what is laid down by the WSDL specb No use of namespaces hence no need to author the namespace references at the top of the SCDL filesAre there good reasons to choose something elseYours Mike +If you hve any other questions let me know +Patch committed rev eefeadeecafffcde +Fixed +If anyone can explain what it does we can always reimplement it from scratch if we have to +It looks like was resolved and is also resolved. +Candidate for please apply the same change to the EWS builds +Here is a to trunk patch Ill try and get to something that should be fairly committable before to long as I said to a large degree that just means a thorough review myself as well as working with any additional feedback +Patch for missing entry +I was not able to start a single node cluster since my network did not have reverse DNS lookup +Maybe Tom can take a look at this case for the next release +Any news on this +What do you think +Seems CXF and use different way to process the oneway I think we should unify bothFreeman +Bulk closing stale resolved issues +The attached patch makes a map of column names to column number +committedplease try last SNAPSHOTnote you must add various slfj files in extraDependencies section to have those included in root class loader. +Yourre right this code snippet seems to be problematicIll investigate +Within that ticket it would really be nice if you could explain what you are trying to do and why you are trying to do it +to backport to as partially fixed until backported +Committed trunk revision Committed revision Uwe! +It seems to me you would only need this is you wanted to implement some access control within the definers rights method? +Attaching a patch which includes a new test case and the code fix for the release +does the failure still occur in your environment? +Now that Hudsons is working again submit this patch for Hudsons +However since the underlying Cassandra does not yet work correctly with using them marks the Deletion as invalidThis can be fixed when goes in +I dont know the exact the problem was the CSS copying the updated one from ORM is another possible solution +looks coolIn your example URL it looks like mindf is repeated +Koushik why do you think its a setup issue? the same scripts setup works fine on environment +I just committed the. +Rather than guess about it and then try and write some sort of awkward tests for it Id just as soon say no especially considring I dont foresee it even being used +Added a warning message in if classes can be loaded. +We could also add a flag to jirb +Committed to branch and trunk. +Not sure if this is related to the Hibernate Tools Eclipse integration +Thanks guys the disabled button is really nice +Updated test and test to close connections +mrepositoryeuiksprojectand then start a Maven build using the top level Stanbol POM +I executed build on my windows box once without problems +In commandArguments is unnecessary +Proposed changes to resolve compilation errors on Linux systems +Please help to review the testcases thanks +This issue is reported against a version which is no longer maintainedwhich means that it will not receive a fix +But we still show Unable to create envelope from given source in the response text box +Closing the issue. +Be aware that this will only work for Maven + +Ill ask the codehaus guys to hook me up again +Same view as before just in firefox all menu items are visible +By the way i decompiler the class in the jar see theas I stated in a former comment Struts comes with which does not contain the latest fix +The GUI has been fixed to make this kind of problem non reproducable on trunk only +Attaching for Space Quota feature. +I have add two attributes for UI TagextraAttributesSet extra attributes render as html element attributesextraParametersSet extra parameters if need to custom templatethis will not override parameter m not so sure this should go into +Per I think it uses the first image image and edit log regardless of the date +I would even go a step further +Hence it is a very rear case +Assigning to have appeared on the website +One of you literate guys would make me very happy if you could pick this patch up and add the finishing touches +Added a root element +I gave up with Hibernate and switched to +MAVENOPTSXmxm is I think such an option +if you have connected successfully with some cert and then in the same process you try to connect with a different cert you could get a cached connected transportI think it would be worthwhile addressing or at the very least opening a subticket to track the issue so its not forgotten +Raw patch +So Ive made comparable changes in the wiki to add this note +Upgrading to Blocker +Actually I believe that item above is already possible via the UDF context +Of course the problem there is defining how many records there are total in any given input +You must set of a catalog +Your patch is in r +I read through it got some idea but not fully +Hello FlorianTo be honest I need more details on the problems you have encountered +Hi ArnoudI perfectly understanding your motivation for reusing reports but have a look to the following ticket to understand why we have started to depreciate all mechanisms to import existing reports I too dont see how removing the ability to reuse unit test and coverage reports will cause problems with consistency +Done for build +You can close the issue + +It contains the following +Once I have the requested sample XML document Ill commit +Of course we can mix all this with the current server configuration but I think that setting the parameters of a server and declaring quorums are different concerns and should be in separate files +Im not able to view the testReport or even console output +Seems to be OK +This sort of thing could probably go on forever though +In JSF you can provide your custom and override that part +for the change Alejandro +Should be covered with the module assembly stuff Rob have been working on +Thanks Sijie. +But I think that we should follow architecture approach divide data and data processing for realize low coupling between components +the same route which doent work on works fine on this snapshot version. +Not sure why you wanted me to take a peek this isnt part of the original Solr function query stuff so I wont know any more than anyone elseAnyway the current code looks like its working as designed? Perhaps it wasnt the best interface but not worth breaking compatibility over and not necessary to improve for IMO +Attachment has been added with description Another has been added with description Render of Attached transition to closed for remaining resolved bugs +Moved to JBIDE since there is no JBDS specific in this looks like a rather annoying bug in the new copyclone setup of configurationsOr is it just the ui being wrong calculating the default +Thanks for the patch Saeid. +It is possible to delete the variable instances iff logging is not enabled +Closing this issue +I like this idea and its not hard to do +Moving out some bugs to. +They are by accident! Sorry +Moving to if you need it sooner let us know. +Id like to leave this ticket open until is solved. +Thank you Ivan +If the previous heartbeat says it has space for blocks but the namenode can schedule as many as blocks to this datanode the overassigned blocks will fail with the +Im changing my opinion. +I assume that the old code should work fine if the implementation returns from getGroup a shared set that is still used by the +Splendid! Thanks guys +Committed +Works fine with reuters corpus +Hold on Im reproducing it now +Proposed patch for this issue +Im on the same boat being a backend programmer Had a good day todayD + +I look forward to implementing those and seeing how this API evolves to suit. +So doesnt need to become This example shows why leading zeros should be preservedThere are different kinds of text search +looks good for security branch +Leaving open until it can be into the branch. +Your latest desciption of relativeabsolutefullyqualified link targets make sense +Added a test and applied the fixI decided to change only the linearprocessor because when other code clone a mail doesnt make sense to keep the stateIts left to the mailet code to set a given state before calling sendMail +Whats the present state +Closing. +Revision Temporarily truncate revisions that are older than minutes if the document is larger than think this may break the repository +Stein can I assume that you will be checkstyling your patch before committing +It definitely needs the treatment +Patch looks good +Anyway I think that this issue is not a showstopper for a release since its a testing problem which occurs on specific environmentsAgreed not a showstopper +Well commit them to if will be fixed before release +I think we will need a class say which will read the hiveconf parameters remove them from the command line and then invoke the user specified class with the remaining command line arguments +Manikumar would you mind fixing this? Notice that the logj reference also seems incorrect etc +The patch has been applied +Locally I have removed all author stuff +For example if SBN was checkpointing holding the write lock and sufficient number of rpc connections are made during this HA admin command will get blocked until checkpointing is done +All other exceptions will lead to a warning and continuing as if the resolver had returned null +We have a saying where i live +I added s for all of the caches as this is where the group search appears to be getting from +by executing antrun plugin and printing out a message TODO Enable the testsAd +To +Sweet that works for me +Just removed myself from the watch since this was a necessary change of moving from AS to AS and was forced upon us +Should we remove that change from this patch +this is a patch to continue. +All other mutations have a map from CF Set of KV +Committed to trunk +Title should read Add a pre select update option to JDBC +Im hoping this will make it into that release +sticky should be changed in docs to am not sure about what to change in sticky desc +jspx emission etc +Looking at the current code it creates the json on the fly with the drawback of serializing the json which is due to the json writer used in all places +Hi James I was able to reproduce the plugin download problem in Geronimo +We will not have time to get this for fixed for +The new method is added with r. +Someone from QA needs to fix thatMy suggestion is to delete the binding and datasource both before and after the test that would make it stableAssigning to Jan Lanik since IIRC he was the one coding the datasource tests right Honzo +Thank you +Im going to investigate the crash +HiThough not take a close look at this issue just a quick hint since CXF is already an bundle so you neednt package it as a bundle yourself +I think this can be resolved The exports are not in the default system exports any longer and we include the fragements in the bundle list. +Moving to Jacobs old resolved issues +Patch should include modifications for is the same except for mods +GeronimoSending etcTransmitting file dataCommitted revision OpenejbChecking in etcnew revision previous revision +So pilot error. +The Agent is now checked to ensure that the latest version of the rule base has been set. +I just committed this. +You are right +Sending srcmainjavaorgapachecommonslangSending srctestjavaorgapachecommonslangTransmitting file dataCommitted revision . +Merged to the branch in revision . +Thanks for checking and fixing +I have to bent over backward to make my calendar work across multiple timezones +The thread local in is not static so I think it can only leak if there is a reference keeping a instance live +Documenting as a Resolved Issue and setting back to Closed Done. +Matt I dont understand why SF is using anyType with a String value +So it will go through all existing endpoints +This issue has been fixed in Eclipse. +Yep with trunk the exception is gone +Attaching the updated SQL +Given installation and removal is such an infrequent occurrence I dont think its a major issue if its slightly expensive to perform on Windows boxes +Adding q file and file for negative test +Ive heard more reports of thisCan someone please provide a test case that demonstrates this +I dont know if I like the idea of multiple applications being rolled up in to one either but just wanted to say that I have seen other people do similar things with rolling up +These concurrent DB access errors are successfully handled by the PAM retries but cannot be filtered from the logs since they are reported by the Digester used to parse files in +I have attached a patch for this case +Attaching proposed fix in diff file +Adding the as listener to the Jetty Context to allow Jetty to send the events to the to forward them to the registered listener issues after release +instead of into the source tree +As mentioned above the exception ends up bubbling all the way up to the handler thread which forcibly exits the process +Thank you Willem for the patch +This is the real patch +In general though I think the modules affords us a lot of opportunity through more exposure at the loss of a little bit of flexibility in Solr which as you said can still be handled by doing that little piece in just Solr +Going to attempt to fix it using Long only when neccessary +My preference would be not to have hsqldb on an open socket in the first place + +Ioannis +I will discuss with Jieshan tomorrow +I committed the fix in since it is quite stable hereLet us know if you have any improvements +I should not have included the first instanceThe second two however are in Ofbiz code +Derby has been built to allow for just these kinds of extensions it would be nice if we could somehow improve the release mechanism to allow for this but whether myrna wants to do that is up to her +This ticket is for or does this apply to and tooWould you please tell me whats the current issue +Installed +My preference would be to not have anything related to the site in the trunk and keep the sources of the site in asfnutchsite as well but Id be happy to do as you suggested as well +While doing that a mysterious line has been added +Since this only affects HV the Bean Validation reference implementation from the EE umbrella we only plan to address this in the Spring Framework line where it has been fixed for the upcoming M alreadyNote that Spring Framework keeps supporting Bean Validation and hence the Hibernate Validator generation but requires HV or higher nowFor Spring Framework Id recommend staying on Hibernate Validator or using native Bean Validation message lookup mechanisms only +Im assuming this is a watchmaker example that you modified +I beleive that should do the trick committed to CVS HEADOleg +Another huh? You own it then +Andy +Right now we should change comparators of the schema for that to work and add a check to the and CQL validations to prevent users from setting comparators to the new along side with check when schema is converted to +Patch applied. +See on the directory layout change proposal +The reason is the EJB support in that namespace which requires parts of the if youre using the jee namespace either including the jar or the full spring jar in your classpath +You have these options +already resolved +Any date as to when is going to be released +The hardware has been removed and replaced with a Linux box +bulk defer minor issues to +The goal is to define that in the JIRA +Marking all issues that need to be addressed in the EWS GA build as blocker Option seems feasible for we plan to do option in a CP we need to evaluate now if there would be no installationupgrading problems especially though rpms +Marked issue as incomplete because of lack of feedback from reporter. +As such I am closing this one. +This might be the reason why RM page did not have the information +i dont need what i asked for in last commenti have added some other solution to my problem but it will be good to have this feature in later versions. +A few little comments No need for throws Exception doesnt help document anything +Our reasoning can be found in the comments above logged onThanks issue in order to attach a new patch +This snippet of the WSDL exhibits the problem +Im not real familiar with that code so I wasnt sure the best way to do that +In the meantime I will test your fix today. +I added Freemaker recently +I am fine there is only a typo copied from my patch in the +Im not sure if I can move this bug to another project so Ill leave that to you +Todd I just confused with the last point +Radek I commented in that forum thread from where you were directed to this JIRA +Isnt that what is for? If you dont want any reducers to run until all of the maps have finished then you want to set it to not the that is the default +This patch adds a lint directive to the compilation target which compiles +The log rolling makes for a deadlock +Thanks +Thanks +Im curious how you were getting this to fail +This is a lines change which Im attaching here +Patch looks great as do the commit comments +Juergen +Issue has been filed by Yip Ng +It is a known weakness in jsf standard converters and validator +However the two fixed issue should fix the problem extracting the zip file on Windows and eliminate several of the duplicatesYeah that should be enough IMO. +assume fixed + +The mirror is now available at git my gosh! Thanks! That was an incredibly fast response +The code in is structured somewhat differently than so its hard to tell if it fixes this issue but Im hoping will solve it +I could have sworn that was not an issue with the build a few weeks ago +Merged into the master branch. +This doesnt appear to affect +No longer relevant. +Reopened following above comments by Thomas ModeneisI notice that the priority is currently major thats perhaps a trifle strong and I would make no objection to it being lowered +The changes look good +Put updated jar into Jboss and ran my original symptoms are exactly the same as before WL still in place after for the moment when TX tries to obtain write lock for that node +Now its possible to set a max number of errors during rendering past that the request will I think a layer connection failure will just count as one. +Now fixed +So we will fix as soon we move to eclipse +Did you also import the related server runtime your project depends onIf that also happened then this for sure is a bug could you provide a screenshoterror log of what you are seeing +updated the logj handling to use trace +Mauricio it looks like your editoride is using tabs instead of spaces +No release note for this massively incompatible change +If an interface method is not going to be used then we are better of not exposing it +particularly since its using the most generic assembly project possible I think +Rubygems changed their GemPackage rake task +Dumb tried to use my design doc as a patch +I thought being a different user was necessary doing stuff like lease recovery and for security testing +fine to push to after JBDS this is just a quirk we have to live with +I just committed this +Attached the correct in r +However it isnt atomic +Perhaps we should just remove itSounds good to me +This could be integrated with to have a class that exposes the various methods in my own code Ive patched a on top that has the named methods and one to return the number of live workers for a attach an approximate patch for this jira alone if it not much of work to separate it from +I have not added the BEA case but for someone working with that now they should be able to see what Ive done and add support +Do you have any releases of this component after this one was fixed +This last test it is complex and must be run manually by executing inside the it m going to commit what I have for this and if regressions pop up I guess well fix them +I have been able to reproduce the issue +Committed +I dont know an easy way to include binary files in the patch so Ive attached those separately +Therefore I introduced an attribute delete which can be set toyes to remove the file after sending the messageAttachment replaces Thanks. +I just thought using a bundled binary jar would make it easier +What do you think about the patch should we go that way +Thanks Cos! +Saw some work on trunk for this too what needs to be done to finish +Committed to trunk +jBPM logging is now being disabled in dbinstall tool +Seriously quit assigning me to this +I guess we can go with this implementation and see how people also updated the docs in several places to mention this feature. +Ok sorry I didnt understand that youve already done it +As per the issue details I was testing on EAP so only the JCA deployment could be used +Duplicate of +DhrubaCould you review the updated patchThis does not include a unit test since it is correlated more with sun spots +Also findbugs before committing +can you elaborate on ifhow will work with your embedder? thats a crucial functionality for my netbeans ide integration +this patch takes the string after the sign as the attribute a bug in the bigfix sorry about applied thanks +WellWe dont use Thrift directly we are using Hector to access cassandraIs there any issue using Hector with appThanksOlivier +The same exception appears with ajqueue +Tyronne reassigning to you so that you can provide us with the updated code examples for +Ill take a look at the notice plugin it looks like a nicer solution +Note no multiple threads required it is a question of stopping child nodes and removing nodes from the active list in the same loop single thread +Cleaned up the toString on table descriptor and column descriptor to not emit values and to not emit bloomfilter state +Some documentation about the classes in the new concurrent package +What tool is reporting leaks? When I run a debug build on Windows with memory leak checking turned on I do not see any leaks +This issue will be tracked as asPlease reopen if you spot the failure in +bq +Looks good please commit. + +So why would I ever want to use a Dumb compound filter? Any suggestions for a better name? No need for a patch I can just make the change +fix file necessary only when writing +As I have not heard anything more on this I am going to assume that the changes fixed the issue +Sorry for the duplicate +Well the advanced installation worked +No that should work fine because as I said slfj doesnt rely on accessibility of members +fix bug on groupingand do not support update on you can add order of score +Theres now a space that causes this to fail instead of a newline character +I wanted to attach the +perfect +Konstantin can you verify +Added a pull request its a workaround Im not going to accept the pullrequest because these Maven plugins must stay backwards compatible +Only comment is that we might want to at least log warnings on the XXX stop processing exit spots at least for debugging purposes I have not for the moment but if you agree I will add them +Also why arent usrjavadefault and usrlibjvm checked firstOverall maybe globs arent a great idea since it may lead to a lot more paths being added which I think is Elis concern +Not really patch available Im testing scripts which need patches attached and marked as patch available before it can pull them +Okay Ill ask on the forum +The patch did not apply cleanly so I have created my own +Updated the files from commit above traced their history to find they were contributed in +Verified on. +were all waiting for this +Attachment has been added with description Patch created from proposed with main build process was improved Square brackets around ant tasks removed Echo messages XML encoding CVS changed to SVN for excludes +Now nonshared shelves arent present but albums from these shelves are present in search result +In general polishing our snapshot support to allow for this kind of usecase is likely to open up other interesting possibilities +Committed prototype for this feature taking into account the lessons learned from deltaspike +BTW all of this was a result of me investigating why when we hit stop in the Windows Services window that the server doesnt shutdown properly +But does not need a conf to communicate with dfs +Backport +rebase on trunk add unrelated unit test change to work around the fact that and are not in yet +can this be related to a bad jetty version that runs GAE +Didnt see this option earlier +Any ideas about whats going on? It looks like the WSDLs schema is incomplete + +JMS says If a client sets the client identifier explicitly it must do so immediately after it creates the connection and before any other action on the connection is taken +There is no service to set in those cases and we are better off getting a runtime error in that case +And other thing is dose not have direct relationship with addressing module +I send the client a custom patch for and then backport after the release +I had revised the tests and made simple modifications. +I have added the following ProjectsCompanies so far Bluesunrise Groundwork Jahia at Convergys Corp +Notion of selectedcertified isshould be done in central IMO +Hi Donald Thanks a lot for applying all these patches +Component set to tck +That would be same version as the other featureThe editor feature only became available inThanks +MS Up login as admin +Updated srcdocssrcdocumentationcontentxdocs but I can not verify it as doc files are copied from doc directory instead of being generated from docbook files +To clarify why Im still waiting for a build the as component job was failing in Hudson for the last few days but Rob says it should be ok now +The way we are defining defaults for capacities actually does not help us accomplish this purpose + Central is implemented as an editor part rather than as a view +Which part exactly is dupplicated? I cant find +How about this +Sure it resolves the issue but it causes regressions in other areas of the product +Thanks for the patch +On the side there are improvements like this as well for example I thinkthe lucene jars used by solr are rebuilt in the process +Editorial pass over all release notes prior to publication of. +So I think we would need to modify jasper and Geronimo code to pass classpath somehow or make jasper compiler work with a regular classloader somehow +KapilCan you try the CVS version of Castor I checked in some patches back in AprilforHave you tried this with a more recent version of Castor? +In the future if the code affect both engine and client developer has to fix in two places +That is a good idea but it needs to make sure that whatever it is doing is applicable to all web apps that would be including the jar +Can this issue be closed now? +The assertion is still kept but modified to handle valuehaving said that I will be glad to consider any other approach than creating these child classes +I think enough has been done on this that we can move this out of and address as we go forward +I will verify the use cases mentioned in this bug as part of verifying . +revision for trunk revision for branch +I will submit a patch for this +Zipped JBDS Welcome Screen extension example project is attached +MRM MRM MRM MRM are fixedThanks +Closing. +The description Exception thrown to indicate that a feature is not supported by a particular database +Thanks! Hemanth for necessary information +Thinking on it it would be nice to ship with this metrics servlet enabled its better than the devnull we currently ship with +libhomeedwardworkspacebin +The change is in the main normally it should be in contribs or notWoops youre right +I can share you the patch tomorrow +please note that the load average on the system is slowly rising Even though ATS isnt doing anything on this machine after hours the load average was +The comment for tokenizeDatabase method in can be changed from return database name and attributes to return database name +However using the example fo provided I cannot successfully retrieve the image to a pdf document +If its no longer an issue feel free to resolve it +Well I need to read up on the Mail API +Hence I went with multiple configuration files +Do you think it will be included in the next release +Integrated to and TRUNKThanks for the patch AshleyThanks for the review Stack +orm and +Thanks Isabel! +Stuff like compression and caching would be a better example for b +Regarding the broken dependency on bouncycastle +There is already the file now so closing this the work done for Maven integration a Grails Ant task should be fairly straightforward to implement and work in a more flexible manner +Its description says Show the last KB of the file and the shown content in the above two examples are both of exact K size. +I have already fixed this in +Cant figure out what is different in my environments I dont believe its an issue between XP and Ubuntu more like how the environment or its variables have been set up +Ah you are right Thrift closes the scanner +Currently there is no plan for a and this may be misleading to FUSEMatt +I found you added the dependency of the jms transport to the soap binding pom +Without this its just a vague description of the problem and it leaves us without a definitive confirmation that weve replicated the problem in a test case + +Fortunately Spring Security has an abstract implementation called that handles much of the grunt work and only requires that I implement methods +Does this affect the code committed withYou set the affect versions to so Im not sure +Yes it is Leon + has been upgraded to use Axis +Bug has been marked as a duplicate of this bug +This would make the includesis distinction much easier to make +Patch to set and upper limit on the number of cached extraction works fine since rev and the use didnt come back +However even the later simple conditions need the prepend removed and added in the tag or an error will be thrown as an invalid sql statementThanks for your timed be happier of one of the official maintainers were able to comment on this issue but as best as I can tell the prepend value of and will only be output if another is element is satisfied before the one where your prependand is specified +fixed in shrinkwrap + for the patch +Patch was committed a long time ago. +such as +Users who expect the packages to be at version are expecting something which does not follow the best practices for semantic versioning. +I tested it first in a new install of Eclipse and saw the warnings then put in the ignore and saw the warnings go away +Marking as duplicate since was already there +mirror has been again! Please remove our site as a mirror! We were contacted by GOOGLE stating we were involved in unnatural link scheme and were told it had to do with a mirror +I fixed thatThanks that was an oversight +Daryoush are we trying to fix this before v +no but now it is +if any of those names are all workspaces are observedin the latter case the event will have a property called workspace containing the workspace revision if a user is not allowed to login into a different workspace than the default and tries to access a resource in such a workspacethis is treated as a resource not found +Hong thanks for the guidance thus far. +Dawid could you the files with the apache feather? I dont think that we can use attachments unless they have the feather +Hi JulienAgreed to wait a while before committing next time +Fine just remember to this for the Java broker when is fixed +Hello VasilyIve tried some test data but cannot reproduce this issue +Targeting version +I was not aware SVN is no longer synchronized +Looks like is obsolete since has a Executor for streams +However using it I have got a version of the Eclipse plugin with working +Thanks Namit +I will check the mailinglist for any possible reactions +I dont want all the ms dependencies dragged into m by default +Jason thanks for the detailed explanation +patch looks good +If this doesnt work please reopen the issue. +Added to Release Notes CP for EAP merged the two notes from both tickets into this one admonition because I felt they were related +Verified +Added message to exception call in readMapping +Seems low risk +The key pair must contain either a pair or a pair +I proposed on that we revert that patch since it only applies to iOS and we only support iOS and greater now +fails before +OlegI tested the CHM parser from Tika on a few sample CHM files and found that many valid CHM entries are skipped +The change has been committed and will be effective in the ER release as well as brms release +I think this is the version Ill commit if there arent any objections +Addition the tester report the ROOT is not excluded in fact +Does offer similar functionality +Damn I had to backtrack as it wasnt possible to have JMX merge custom mbean operationsattributes with the default in an automatic way +The attached patch makes sysinfo print JRE JDBC Java SE JDBC when running under JVM version +Similar work will be required to integrate Knox with +Only urls which match domains in the domain set are considered valid +As long as there is no runtime dependency and the pieces are clearly marked so that those that would be troubled by this could quickly and easily remove it I have no problem +The patch +Zheng does your change work with +Im sorry for the confusion and I hope that this is not too much trouble to bundle project urls in each of the requests now point to the correct bundle with the proper you and Im sorry for any Safris +Would you like me to reapply the style and submit the diff? Ill also keep that in mind about the dependencies +In extereme cases you should provide your own Dispatcher and build a URL to the Dispatcher not to a page or component. +converter +With plain JSF there is no issue +teted in both windows andany other pointers? we can reopen the bug if we figure out a way to recreate the problemthanksdims +Please provide input and assign to docs team +is now completely on the new escaper system in this and I mostly just need to javadoc and strip out debugging to have a like for like replacement +I will try to make the switch and I will let you know as soon as I do it +I have fixed this issue locally +This is an experimental project to understand the usage of Apache Batik +Test script to create a collection uploadlink a config and create two cores +Thanks for the patch Andrew +Okay I double check +The same exception happens the I forget to sign my jars +Projects that use Hadoop already pick up a lot of jars and increasing the set when all of the versions are the same is a problem +I think this issue can be resolvedAnd of course the integration tests still succeed Please close if it works ok for you +This tends to be ugly platforms may interpret them differently +But until then we can do the approach +Tests with the latest code have not shown any indications of this issue +Hi DanielIs there a way to work around this issue elegantly? Basically information on split needs to be available to loadfunc +No tests simple UI fix +The patch applies fine for me +I see resolutions +Jarek tried this on Windows too the same result blankIE on Win XP k not server specific that build never works even in vanilla AS looks like GWT code was not checked in before that tag was created. +Proposed at revision +seems to have CVS problems Im afraid I wont be able to commit this before tomorrow +My next trick will be to try maven +behaves correctly +attached patch +I havent delved into the code yet but Im guessing that there is further work to do with inner classes +I would suggest checking the server side logs +possibly useful datapoint +Will work on porting +Fixed in rev +Ive built this patch times now and had only ones a Failure in test instead of multiple failures on every build +Thanks JianMarked incompatible just in case someone was using these in the alpha releases. +Might have to reopen to look at qualifiers used in source build +Added test for Trinidad and capability of handling the Listener tag in a tld +Hi Ted I actually already have unit test showing how should run it checks the states and right now just run through couple simple different json objsIm going to update it so it checks all output +Hi JacopoHere is the patch to remove the deprecated operation from the MinilangAnd if something is missing please let me knowThanks for your support and timeRegardsAshishHere comes the patch to remove the deprecated method from the MinilangPlease take a look at thisRegardsAshish Ashishthis is awesomeYour last patch is in rev +The whole code will be replaced in Marat I cant simulate this memory leak can you send an quickstart +So I dont really see a need for this action to stay on the screen all the when import feature will be implemented maven will be enabled we have New Maven project wizard in the trunk +Thank you +Patch available Ive tested it +Moving this to installers category +This is already in the help desk queue +Duplicate of have tried patching the js with and the problem persists +It will provide us what we need outsideof a JEE container +down to the eb version works as well +For the ZK reconnect condition the if logic would bypass the part to change the server back to the not suspended state +Thank you WernerCould you please update dependencies for in order to run tests +Hi SagaraThe files are now attachedThanks + +Not sure why I set it to ER as the Affects Version as that doesnt make any sense seeing that the fix was in the ER release +Marshall take a look at to see how to start a real quorom set of servers +Can we have confirmation that fixes are in place for +need to raise this question on Yes please start the thread +M Released +Hello +Closing. +Used xsltproc Bxsltproc version Using libxml CVS libxslt CVS and libexslt CVS xsltproc was compiled against libxml libxslt and libexslt libxslt was compiled against libxml libexslt was compiled against libxml XSL Stylesheets has been added with description reproduces this error works +hmm not doing the close on the connection used to create the XA Resource will leave a connection leak +Peter could you commit the attached patch please? Thanks +I tried building it on a local partition on the same machine and it seems to have the same result +Derbyall still running will confirm if it passes successfully in subsequent comment +The methods that are supposed to return appId for the return null for this so that it crashes deterministically if it gets used in the future +After testing I found that in one more on deletecascade wasnt changed to on deletenoneSo I upload new patch for thisYoure correct that the MSSQL database support isnt correct yetBut your patch isnt the right way to fix that removing the cascade actiions from those constraints isnt enough the children rows still need to be removed somehow rightIve reopened for that as it better targets the whole problem +Atlassian support installed in JBDS +Ilya can u pls rebase your patch after? Tx! +If theres more work to be done for this issue please or open a sub task for the remaining work +Will post soon +I agree with Andrew +I would assume that this is fairly important and having duplicate ids could be bad +Created and finished building are not the same thing +attached that moves filtering on resource to itself away from execute +I believe that the original problem is caused by the minus sign in the class name I checked with a friend who works on the System i JVM and he says that the hashcode native method will sometimes return a negative number which is behavior permitted by the Java specification +Reopened the issue due the fix did not work +Robert this is an interesting bug since it seems to be a very old working on fixing this I was wondering where to place the error marker? I cannot put the marker next to where the property +I think Gregorys post makes senseI will wait for further validation result against RC and watch out the discussion thread +I looked at providing the same approach for the Jdbc but its not quite as straightforward because the restartSql thats used in the single column case uses an object array +I tested the Chinese Input from Windows which appeared as box +The first round of tests migration +Other ideas +However it looks like patch includes a bunch of changes already applied such as the changes in +We can make it configurable and document that I guess but its kinda odd to have to do two toggles to get an override done +If you or Mike have any suggestions on more test scenarios please let me know and I will add them +Just letting folks know +Duplicate issue +Just to try and push things I altered my schema to use LZO instead of GZ compression ran major compaction and kept the writers going +An alternative would have been to use a temp file with a unique name in the basedir +Are you intending this issue to cover both renaming a table and renaming column families? They are two very different and independently complex issues +you are right I will add validation to create method to ensure that Snappy lib is in place +Here is a patch +Incorporated dougs suggestions in this new patch +I cant recreate it locally either and I cant log into the hudson machine to test there either unfortunately +Looks like we should expose scanner lease expiration through new coprocessor API +I personally also use but my idea was that I can extends some class and and run his methods with existing IDE Junit support +please see dependent +So the issue might be an edge case related to deploying while the server is up +Just checked it yes the major of sql server version is +patch +Does Xen support this or would it be KVM specific for now +The test still fails if I force LTC to never use an ES in newSearcher +I couldnt find the original but this is my best guess at applying it to recent source code +Trivial patch to catch this but only mention it at debug +Fix looks good +Im doing maven clean jar by the way in case youre referring to the Ant build with +In isolation it definitely isnt thread safe and I dont really think its a good idea to make assumptions about the usage patterns +For example a hotel could have more than one deluxe room +Patch looks goodTest failure is because of +If the performance were good enough one might even use it all the time +This should be working in CR +I dont use m hence I just want to know what the directory structure needs to be +Can you if you confirm steps to reproduce? Thanks Kristian +It came up during testing related to as a result of moving classes from one sub deployment to another +Thanks for your comments Sanjay +Its still the case that we need to uncomment the dispatcher lines for Tomcat after installing + +Maurizio my patch was actually just for benchmarking the various approaches its far from what Id consider good for inclusion +It seems just calling URL like this httplocalhostexamplesjsptoto fails due to fragmentIt does not with Java or rewriting of absolute request was inconsistent +Committed to trunk and close resolved issues +The default size of cache is which turned out to be too much in the current case +Up to you +The mentioned lines has xA xD window style line endings +Well have to wait for upcoming JMX revisions to address this this might mean waiting until Java +Two votes is good enough for me +Yes we have considered Curator +signededsmodeshapeexamplespubunpubusingrest jcoleman soa Dec +A patch fixes the immediate issue encountered in the aforementioned library but I really think that data type conversion has to be implemented more generally not just for the Double and Float cool thanks +Applied in r thanks for the patch. +is the standard for plugin meta dataWhat am I missing +But without even so much as a response from a Hibernate team member I dont even know where to start +I thought that security module was huge but sum of all is appeared to be twise bigger +I think the issue is fixed now +Maybe this is related to? What happens if a heartbeat is not received in the time specified? Is there an accepted way to recover from such issues +I think we just have to change the link in +ricif you search for Eric the returned records should be records and +I know you are not supporting but any idea would be regardsNicolasIt seems to be more complicated that just replacing the castor jar file +For more than a few thousand columns we could fall back to the current implementation +It is caused by the scanning behaviour in the +Also attached patch for trunk but i havent succesfully compiled trunk yet can someone else confirm the tests pass and the exceptions are gone +Looks like FF cant reproduce of a bulk update all resolutions changed to done please review history to original resolution type + +Fixed by Avantika committed by drewbai for by inspection +Native mobile app developers will still be able to leverage the pages generated by using in their app +Just a short questionWhy do we stick with lock files instead of rpc calls and handling locks on the master +pending Jenkings +theres no indication why this is happening +Please add more cases to the patch +Looks like for every module its trying to update again which is wrong +Kan you mentioned that fixed the exception so please close this issue if it is fixed or add additional details on the new failure scenario +Reallocating to rc +Hi ystein thanks for reviewing this patchI agree that the first connection should give a more meaningful exception +I think we may be ok on thisDownloaded and scanned the source for GNU references and by my count they only occur in source files all of which seem to be part of the code to handle JSPSince JSP isnt a requirement for I think we can safely unbundle the JSP code and avoid any issuesThe other license references I saw were mostly BSD but Ill do a proper scan of the other source file licenses at some stage +Ill try and determine if we really do have a circular reference +would say no we will not change this for +Soright +Hmmm yes thats because the theyre essentially regarded as different classes if youre reloading that is +Clients not using the library will ruin the scheme +Still investigating +Fixed in closing stale resolved issues +The original issue was caused by multiple processes logging to the same file +Marty took the action to ensure FAQ and any other emails were constructed and sent through a system such as Eloqua +Karaf updated to Aries Blueprint let see if the problem occurs with this version +This solves the integration failure issues on my platformI suspect the problems to be related to the older default setting for session pooling which seems to break with Jackrabbit +Retry Js patch +we could generate a custom Solr XML Response Reader using JAXB we could generate sample Solr XML Responses with ease using a tool like Doxygen or +Please find the problem project attached +I appropriate we might want to report this as a separate issue +That requires a bit more work and also may complicated the configuration a bit as you would need to configure expression to evaluate the correlation keyAnd then the throttler logic need to have a map of correlation key +The difference in accuracy is very close to machine precision +Ok +For instance there were previous suggestions on writing back to the last file where the setting was specified or by specifying the file to write to on the command line +Ill try to add something similar to BG. +Resolving for the time being +Test to verify the bug +Can we rewrite the patch this wayOne concern I have What if the file is actually gone for some reason? In that case it seems wed never stop retrying +The workaround I know is to add a property to the embedded object that cannot be set to null e +Ok the changes to appear to solve the lock up issues for the JDBC tests which make a lot of HTTP queries +the query use thewhen i use the connection through JNDI the query use the +Still broken in any code changed to address this +Do you agree +This issue is closed now +When youre working on this make sure you are NOT changing the existing application flow just create an alternative flow +Closing issueThe test stressing concurrent bootaccessshutdowndrop should be added under if appropriate. +In this way different URI of the same are different +Im going to close this as invalid for Xalan then +This issue was originally scheduled for but its still incomplete +Struts forms are intended to be String based to capture the user input +See resubmitted patch in +Looks Good To Me +Ive attached a patch with some further improvements to the demo instructions +The Queue Has Been Deleted +Ive seen your patch in the branch +patch file that they can just applyIt prevent us from using Zookepper or newer releases inIn fact if the Zookeeper team is NOT listening to the community it would be better to remove all the bits from their fileThen at least the release is NOT attempting to be bundles +Liu +Then the NN tries to close the file +We can add a note in the instructions for other app servers +If you want to try it out you need to build from trunk or use the latest snapshot +See if you can repro and fix in IE if you can do it otherwise drop it to major and we will leave it for a GWT update in future +Sure we can do that +Reasonable future change +So users who want to emulate the forced even when user didnt say so explicitly can create QPThis should be an option not an emulation +with +The quick view makes API calls to load the response and since the context is not defined it throws that error +Yes I now can indeed confirm that it works +Since that time most of the tests from regression have been moved to master tests addition I expect that you need to do the work yourself as there had never been response to my comments wether from Ullrich nor from Chris +Set it to a magic number which is large enough but not too large is a viable solution. +Carl Edward Ive updated the patch on reviewboard +Simple Chinese properties +Triggered a new build on Jenkins that completed without errorClosing this issue for now +Attached original +id rather see a global query parameter or cookie that controls the version selection +After running the test suite against sapdb engine Ive found more related attach patch later and reopen the bug +It would help to understand what LCEnonsecure doesnt solve yet +Patches applied +I think its now functionally correct but I still need to compare perf vs existing syn filter and there are still a few minor nocommits to work outIdeally if we get perf close enough since RAM is much much less w this new syn filter I think we should replace the old one with this new one +attached updated which uses the correct offsets +Do we need here features like defaut default agree with you +Wasnt sure but made the spec pass by just testing instead of eql +Further debugging confirms that which initializes the during SOAP calls is never invoked +files changed M srctoolsM srcptBRgetstartM srcgetstartI modified some minor details in the sysinfo example as I was looking at it including the revision number +Attached patch for review +This also allows to specify different schemas for different namedOutputs +Around september I might be able to allow some time to work this issue out +Or check your cloudsearch domain and see if the documents made it there +I am not sure why maven fails at all here it should not even attempt of locating this artifact +I would rather not do the testing with completely static indexes it would be better to have data being added and deleted normally +is this more specific to the features added in that allow the MAC to be run as a single node development environmentThats what I was thinking + native config files from the defaultdeployers folder and the log file generating the information associated with using the riftsaw simpleinvoke example +If you try to create a new key you get asked to give it a name +All the dependencies are moved up to the as well +need native EJB test execution +The WFS spec definitely has an option to let the user specify the FID instead of having it set by the server +Bringing this back to trunk again and adding one small change to make config files more backwards incompatible +Thanks Rick for posting the results +Added an assertion to the related unit test and fixed the query +now to the which is the with Scotts modification and the patches from needs to be updated with the new jbossas package also we need to verify if the regular can be excluded from the errata or if it is an indirect dependency of some other package +I thought about the final approach mentioned Roger but Im not sure Im following +Ill take a look at it +As they both are separate +However we notice that when we use xsdinclude the data types defined in that include file could not be resolved +assume fixed + t reproduce this issue +Committed. +Perhaps we need to somehow formalize this converter logic so that this screen can also take advantage of else we could just allow editing variables of certain types +Yes this is current code this is part that actually makes my change to fial +Looks good to me +I have locally tested this change +bulk defer to +Where does your patch fit Mark +This patch marks all analyzers in contrib as final and removes the backwards compat tests checking if subclasses implement reusable +I also noticed that when running the demo target more than once the classes were not recompiled +We cannot reproduce this bug after months of trying +let us know if you are still seeing it +Could you provide a example or even better a test case? Thanks +Thanks +derbyall and SuitesAll did NOT show any new failures for the patch v +Makes sense to me +Looks like the code uses generics which limit the use to JDK platforms +just running the upgrade tests passed also +So this also applies to queue type destinations +or come up with more unit tests +I tried the example and yes we are not cleaning the viewMap when there is a redirect +Note that marking completed reduces of running jobs as failed is really bad because if the file already exists I believe the tip will fail thereby killing the job +As it stands the implementations discourage using the Map +Not only do we need a way of getting the list of options in XRX we also need a way of passing such a option to XRX +Disabling thrift and gossip is seen as a way to isolate a node from clients and the other nodes +Verified by Daniel. +Thanks Kathey! I guess this is a small device limitation + thanks Ivanps please update your editor settings to use spaces rather than tabs for indentation +Pull request with the workaround only please leave the bug open +So what Im trying to say is that init methods are not always bad and sometimes ctors limit you +Overall looks good +A couple of comments The string constant NUMINPUTFILES need not be defined in +Sorry seems I cannot close it +I have removed the unused service definition removed omProperty as we dont yet have a good design or working code for handling this case removed the duplicate code copied from and changed the package name to dbecho in order to distinguish this code from that in the sample +Consequently all access errors are showing up as not found exceptionsMore logic will need to be added to determine the actual cause behind the not found exceptions +a shutdown hook on the JMS stuff to ensure cleanup of session pooling +This bug has been fixed in Xalan Java +See the attached graph +Marking as invalid. +This should have Spring Integration as component not Core +If Im not mistaken pushing updates is just a matter of updating www +You could take a look at the and Inactivity monitor classes in and contribute a scaled down version of these to the project we love contributions +Thanks for the quick response +Ok I confirm break with has been introduced in as its only milestone builds Im tempted to not support +Would be great to see this developed voting for it +New patch which passes tests + +This is why I like to see a stack trace so I understand what the problem is +qualifier anyway to be able to get users to try out nightly builds +Verified at +Ill kick off a build for this again +Stop connection +then there is a way to remove the entire completely if it has become obsoleteso our artifact id can just be kafka or core or something +A few nits The changes to output were garbled +Please keep us posted on future developments on your sideMeanwhile I shall continue my work on this and play around to see if I can get more ideasapproaches +Adding v that does thatBut actually v adds a clone variant that allows to specify the factory which allow to change the backing map in the clone +Sorry I do not understand if this is fixed and in which versions +Attached migrated version. + +forgot to select the fix for release +May be we can add jsf projects which will be SIMULAR to our existing project to get smth done before helious +This is still present +There is some scant documentation in the plugin +add missing spring bean context file for this is a dupe +Maybe supporting the accept header directly in Sling would be a good idea not really sure +This case opens fine in Archive Manager is you name it +Not a Hibernate issue +This is just one example but Im not sure how much we can expect from automagically figuring out language support +Since I got no answer on if this is fixed as indicated Im moving it to TBD. +The current code that updates the conglomerate numbers indexes assumesthere are duplicate entries in the with same conglomerid entries when multiple indexes are referring to a conglomerate +Hi KarlYes that looks like a cycle and Derby should have been able to detect it and throw an +My current opinion is that we should be making sure that the leader really has the highest zxid among a quorum before exercising leadership +files since I havent been able to run the tests yet but I am posting it so folks can start the review +this issue wont occur in cluster wide storage +Not sure which version of EAP is picking that up +In XML filter copy cookies from wrapped response to AJAX redirect response +I propose we just remove it considering it was deprecated in september and probably no one is using it +However if you find an instance of this difference affecting an application please inform us +bq +After refactoring many things in the triple store the error cannot be reproduced anymore +This bug is out of our control on the tools underlying issue have been fixed in Seam and JSF AS +A patch is attached +Patch to remove sstable stats from the system table and make them their own sstable component +To less sleep last night +I did not get what Best effort for the directory means and why it is followed by in the first occasion but with the +I am not sure why this issue got assigned to me +So definitly should stick to bytes for the buffer size +Seems to be due to another constellation that I did not have in the previous +There is too high a likelihood that pid files could be altered or deletedIf this was wikipedia Id add a citation needed flag +I believe the proposed patch is incorrect +As I said its easy for me to work around +I will go ahead and cleanup the code and check it in +If a client is reading a complete file and that file gets deleted or renamed then the reader will fail +The new patch only uses the hadoop modules when the profile is active +BTW +The patch looks good to me and very simple too +JornThe released +I just want to make sure we are informative in the logs and that the tasks are allowed to be if applicable etc +there are two things to mention +There are a lot of possible ways to have something like karafrun +if you try to access it directly fails to recognize it as a metadata file because the check in line compares it to the string and fails +Reassigning to Sneha +I feel that opening a features file and parsing this is a bit awkward +Please redo this patch +Patch that parses the rules on configuration and throws if the rules are not valid +I think we should still disallow ordering w no key restriction instead of allowing but ignoringbuild looks expensive can we compare wo that copyNit prefer isEmpty to size +The testcase is updated accordingtly to test this new method +our DLLBut we didnt resolve problemIf you want we can send compiled exe file and youll try it on your side +Can u please post your code snippet on howwhere you set the option to true? better yet a full sample for us to recreate the problemthanks have a difficult time believing that this has anything to do with since the compression should be taking place at the transport layer +HiPlease take a look of this new patchARGS variable was still in unix modeBest +Thanks Szehon! +Not that Im claiming it was a good idea it does indeed help to have some way of creating instances for testing +BenjaminGood point +hackaround that prevents the problem by capturing bad and preventing them from being accessed my hackaround +Testing now takes a few hours to ramp up through the map phase +This way bug fixes and improvements can be easily backported by simply copying the latest files into an older installation +Ill have a look at the issue hopefully the fix will find its way in either in the old Oracle datastore or in the new one we have been developing during the last month +If you can take a look that will be great +Patch applied at r please verify +v +Khoi Nguyen Can you please fix this issue ASAP +The patch simply fails to apply to trunkThat explained why the unit test failedThe patch failed to apply because r did some format change on which caused conflictsI will the patch nextA new patch with conflict with trunk resolved +the compiler hook then groks the the comment tags and transforms the AST with new line numbers +In the patch I do three things +This seems like a dupe of +If not I do not think this is somethingthat will be changed +trunk +No needed anymore in context of +so we can push it to until we iron out the detailsBut what is already done lets me share solrconfig objects and reduce core load time drastically for our massively multicore search system +I have aversion of at home that does this which I will attachtonight +Moving this jira to release since this is not critical for +Closing out now merged to. +it looks like a dup +Fixed within ok verified in JBDS h Fedora +test case illustrating the does that ring a bell +Let me take another look at this and see if I can identify which server import is causing the problem +I have committed this to trunk +Ive also fixed the real issue in but with the workaround we should not need a new release of paxlogging for now +thanks Aaron +Reopened to keep it on the radar for +fixed locally but havent tested can you verify this please? Thx +Given that I would argue that for the case of running the command ant test this patch is better than the fix proposed in although that has value for other cases +Fixed with testcase in r +Pushed to trunk and branchesThanks for the code patch Tim and thanks for the review and doc patch EdwardBy the way you have both been added as Flume contributors so you can assign yourselves now +fixed formatting using dots cuz spaces are parsed out it this on the agenda and roadmap +FYI the same tests pass on the original project in worst case we can also put it as utility into no longer calling it inij +Just to rule out the obvious you dont see this behavior wo this patch +Looks just committed this thanks Noel. +Patch committed. +The way it is the individual project is always consistent its only the fork modes which arenm not sure thats the its certainly true that you cant have it be intuitive both ways +Weve got a build weve got a process for building out all the components +Update the patch to make using configurable +Only sometimes is replicatable +This patch comprises comments suggested by Steve +We should see if this fixes it +on the right side the icon is used by me as server image will be displayed on server view but it seems not be conventional server image could you design a new icon for it thanks +Let me know if you spot the error +I then sent another message to the less active consumer and it received the message +Mailing lists should be set up and working now please confirm +Closed after release. +Attaching a patch that adds a gettersetter for the checkpointInterval +Once is done to put something read from an external source the other to update the value of something in the cache. +Thanks Dave +It has been backporting to and as of now +I have a beginner level experience with Hadoop +Strike that +Ok so I did a little research and I could find no evidence that suffers from the same spurious wakeups as Objectwait +Please submit a test case that does not rely on external servers and does some properassertions that show the problemOrtwin Glckps +Committed revision . +Version control is only used to check out the source +jruby is released and in central maven repo +This modifies and +I pushed in ffb and Im going to mark this resolved +Need to find out more +This JIRA patch depends on the changes submitted in +This patch puts some basic support in but the test app must include an offered sequence +Yes +But using the tool it shows the same fault as the echoblockingdual + +since there are not much dependencies in this class it should work to overwrite your old directly in the JAR +its not OK to be missing some needed query parameter but it should be OK to have extra they simple wont bind wont be passed to the query +I am merging and testing right now and will commit it once passed +It was just a problem of wrong version of ASM. +This is taken directly from the default config file just put in a document format +Or test the new engine inside the passing the classname to the servlet configuration +Im reopening the issue +Ted what should i do for QA run was aborted? or could you kindly just kick another QA run instead +The given patch realizes the described version of the patch applied in +I think the first option would be better but I dont know about the activity there +Correct patch for the trunk and +How do I mark this bug verified +of partitions configurable no +wed better fix the problem introduced by fairscheduler in fairscheduler +Ive checked the changes in +i cant see this on my installation +Making connector attribute optional does not seem reasonable as we need that as dependency for the subsystemUsing lists instead seems to only workaround the issue of not supporting runtime mode for attributes properly +It looks like we are making the class part of the core Whirr API aware of the BYON provider +it out since its not a blocker +Thanks sorry for the noise guess that the whole folder missing made me doubt +Integration tests added +Reviewed and ported to +An uninstall can be run after other UDDI data has been save so just deleting the entities saved on the install is not sufficient. +Done +Otis Im working on the changes I described in do you think those changes are better suited for a new? I was adding new request parameters to use the fields query analyzer as described in this issue +I ran w no arguments +Still needs to accomodate DTDXSD conversion for proper validation of the TLD XML files +Ive applied that patch in commit resolving fixed. +Sorry to be such a pain Thanks for al lthe good work +thanks Donal wed need to write a fairly specific code to do with parsing the base path and figuring out candidate chars which can be given that the usual path encoder wont encode it because is a valid path char and even does not touch it +Marica says it was fixed in and Ive seen it work correctly as well +Seems generally safe to pull into all branchesEmphasis mine +This is not observed with the latest Build. +This was on a dedicated machine without HadoopMore interesting in my opinion is the reduced memory consumption +The user cant proceed through the wizard to get to the operation that commits the atomic conversation since validation will get run and fail just after the model is numeric data if JSF cant do a type conversion it already doesnt update the model + +fixed in CVS +Dropping the target as theres still more to do +Patch attached +Resolved in rev +If so someone close this issue and I will deploy the plugin to the snapshot repo +Replied +on patchNice work Ramkrishna +fixes where there is not data for a fieldit also fixes the eol +Doesnt work for all roles +I tried with Mahout +thanks flavio. +Gump is no longer building Excalibur now +Commit +Please could I have a reason why the bug has been rejected +Since this the JBPM alread fixed this issue It does not need to add in documentation +Uses the shorter less informative but more compatible method overload. +Edited the readme in a few places please review my editsCarl +Agorava core has now a new engine to produce +In my idea this should be possibe with subflows too another approach for just this element would break the concept and would irritate the user +This is not required for the initial shipment of +The drivers should be able to handle that fine +I just wanted to clarify my observation +Code changes look fine +I wonder if using is that much of an advantage when the keys are Type objects +Updated the screen pics in the dynamic help for the preference pages so that they show the latest preference page tree entries +I agree with Van +Yes there is no difference we could have a generic goal for xjc and then more specific ones to build the projects +The namespace implementation is still outstanding but thats another issue. +One possible explanation could be that the error is transient and that the attempt to handle the processing takes so long that the error becomes likely for example one of the responses in my testing exceeded K in sizeI will try to investigate this further during the week and see if anything else can be discovered although the impact of this has been reduced dramatically as a consequence of the performance changes in as the business detail should no longer be requested during normal EPR operations +Hoping to hear from you guys soonkthxMahouter +Patch applied with svn revision +There are a few wrinkles to take care of in the next couple of days and then Ill start the release process +Since Wicket makes an heavy use of slfj Im trying to understand how to use a custom Logger to keep track of generated logs and test them against expected string formatHas anyone experience of using custom slfj Logger +isdisabled may return true incorrectly and a subsequent call to enable may throw a +Hi Fuhwei your attachment only has the fix and a test schema but is missing the +Hello ManuI downloaded the code created an Eclipse project with it and run it with the attached +No worries so James just glad its working +Definitely for this patch and printing statistics about unique value counts for SORTED and SORTEDSET +Its easy enough for me to add this to my own metadata but ideally it would be part of the commons metadata no? +Sharad what Solr revision have you applied the latest copy of this this patch against? I know that the r commit caused all kinds of havoc in the source tree but Id really like to try it out and I dont mind using an older revision to get it workingAlso do you have any newer versions of the patchThanks a lot +So IMHO it would be nice to switch to that one +Integration tests already verify that the different work on all the supported databases +wont compile against the releases in that repository it requires changes made in +This is a raw javavm crash so not sure what could cause plain Eclipse Juno work for you with the same steps same jvm same osHave you tried update your jvm +Its a pity if it is break +Server is the exception stack trace attached as being thrown by Test I am getting this particular type of error in most of the cases +Like the Authorization Codes in it can be persisted and utilized by many types of applications upon expiry it needs to be refreshed or the user needs to be +Can this ticket be to be able to reviewcommit with the patches that make it configurable +YesNew patch with println at the end +In that case the representations for layers and layer groups changes in that the style reference instead of just containing name contains a reference to the workspace as well +Feel free to roll up the sleeves and provide some help to get it resolved +I am doubting it was already fixed by changing Semaphore to a guava +A lot of this is implemented by now in rev and previous work +David thank you for contributing this plugin back to the community +Getting the handler involves loading classes which triggers another classload via the and there we go +ignore the first Kindly do the honorsthanksdims +Oozie has distcp action for doing copy with Robert and ViragCopy could very expensive in cases +so remove shutdown just close its underlying socket works well +Its still broken +applied in revision Paulex please check it was applied as expectedthanks for the tests +It should be fix +The attached file is a patch +with a failing later version of the attached project includes a failing parent application context explicitly defines a with the name taskScheduler +D +As I mentioned the new integration tests verify that sequencing of CND files is now creating the correct output +That result type is tested in every build and also every nightly by the official OGC compliance tests +Ive updated the patch to also passed down the client address into the bind requests +Some want solution where they do not want ip failover nor special load balancerbq +The latter is by no means a common use case +This patch looks good to me +upgraded priority since if it is affecting other things it will delay our release +Sorry is it two patches for the same JIRA? If so please combine them together having multiple patches for the same fix is a pretty bad idea for some apparent reasons. +BTW I assume that this is incorrect for other cache types as well +I actually have something working in my local repository Ill compare to yours and comment when I get a chance +Seems like fixed Opera +Patch +Ill let you know when I did the ok +If it is compatible RM can load it and proceed as normalIn case of incompatible throw exception and indicate user to upgrade the state with a separate tool to do things like rearranging the directory structure or renaming the fields back etc +Whoops +I guess I should have looked a little harder +this patch fixes the described issue by getting the id out of the hidden idcl field if no element is providedfurther it again adds the to buttons in the callback to reenable updated and select need to call the oam funciton in the to enable them as partialTriggers as shown in the example +I am away in Asia now and I dont have access to any dev environment +I also never have used fixed reply queues with more than one consumer on them but I guess camel will handle this caseI will try to work out a patch in any case +Perhaps +So when we close a file the SBNN loses its block locations info for that file and is no longer hotI have an ugly hack which copies over the old from the existing but Im not convinced its the right way +erwinben Would it be possible for you to convert the documentation into forrest? Documentaiton as part of the release and accessible via website docs will really help adoption of hedwig +Here is a new patch that fixes the comments +Can you give us more detail about unload the application? WillemI had planned on leaving it up to the application to call this method when it is unloaded from the app container +Thanks for the reportThe links should all be fixe now please reopen with details if there are any still broken +theyre start to pile up and will soon reach the point of no return and all collapse if I dont commit them soon Im really happy to help and to do the dirty work +The issues you raise are real when you are moving search indexes and such +I found that a number of internal exceptions including decoding and processing of sasl and context protobufs caused on the client +Applied the patch into master branch with thanks to Grald +Running the trunk tests against I did not see the assertion but after minutes saw server side exceptions +Looks like theres now an OSGI bundle for that release +Fixing it in the next EAP and EWP should be fine +Thanks Lukas +Changed to major as I think this is an important optimization +Closing all tickets +This is caused by there is old data in zookeeper +So in this case there is only space for Maybe ij could have some minimum display size for a column so that at least some of the column name was displayed +Wonder if you were up for refactoring it so that they use the same code path +committed +v fixes a couple bugs and updates the javadoc forPatricio can you test that hint creation replay works as expected +Plz note it is a copied and modified to test only types +Updated Substitution Group Sample attachment +I merged this into trunk and. +Added the link in all of the places +Weird +Im also finding that if is used in an application the seam sfileUpload component doesnt work +has been attached +While creating I didnt notice this issue is already present +We dont hope to see that +Since the information is simply unavailable to Java programs in my personal opinions the problem might be better addressed in the JVM or even lower abstraction levelsThe JVM has some problems but this isnt one of them +Rebasing the reassigned to myself by accident when submitting latest patch +Boss Rishi! +bumping to Beta since that is where these should be verifiedresolved +Marking for +I think Stacks changes are still desirabletest isnt the last test in +I also thing some methods in the would require privilegded blocks because the methods in them are called indirectly through the java procedures using the derby engine +So no longer will anyone be able monitor individual servlets +Closing +this issue is resolved in trunk and installer works verified JBDS build H +If not I would like to close this ticket as duplicate of +The failing tests are unrelated and respectively +The plan was to create a artifact that includes the metadata and RFC code so we have a competitive offering to Eclipselink +The search method was mocked to give a particular return value regardless of the string passed in so the hdfs user was totally arbitrary in this case +tested +Probable +IMO we could remove this from and just leave it in as a standard ticket then you can close the release +This portlet was also contributed by Chris Mills and will be available on portletswap for now +My custom Field Type spits out the query it gets and from the console I can not get it to receive a backslash +latest patch looks good + +Not issue per se +In other cases the command line values usually override what is specified in configuration file +I think we should start with some monitoring of whats going on especially on HDD servers do some plugin point where people can play with block placement Let people with time on their hands do the good pluginsFor the complex decisions you probably do want some history filename created last read owner +I fixed the testcases that had an Ignore on them and run tests +Thanks Alejandro I solved it by writing my own components based on the proposed onesClosingRegardsMatas +Now the patch is okStill to my thinking mallocing small local blocks is not a good practice +More details on +Attaching +I dont think we should introduce a second API when it looks like your problem can be easily addressed with the current one +Then rerun the script without specifying credentials for anyone on the vm boot command +Need to discuss the auditing features when community is interested and ready to work on +Heres the Component Plugin . +All this keeps the memory footprint exceptionally optimizations enable AS to run with stock JVM settings and on small devices +Check it out +Attached incorporating Cryanss comments aboveYou have my lets wait for Jenkins then Ill commit +Best regards +RN entry corrected as requested by Marek +Please look on the related JIRA issue +Reason why we are calling it Writer because it is intermediate Writer for Mapper Threads and It eventually uses Writer passed to Job to do actual readwrite +Vijay please confirm the proposed syntax for annotating void methods and the LHS of an assignment +they may revisit their redesign +I was playing with an xml serialization library that used annotations like the JSON library but it was not suppose to be included in the patch +Logged In YES useridYou could already use it with as the generatedscript imports a file named will port the modification to Packaging +Ill try out and get back to you +I agree this needs to be done and merged into trunk +Here is the updated patch this include the cancel received item flowITEMREJECTED status change will be handled by flow. +You are correct that I missed it +Part of +thank you Shaun. +Thanks so much for fixing this! It has made my day +The parsing is deferred until the moment the class from that jar file is requested +Thanks for checking back +Hi you sure that youre using Maven the newest one? All previous version where for trash when you where going into details like the execution order in one phase +How will this relate to the stuff in +batch transition to closed remaining resolved bugs +IS this going to be fixed for? If not please unschedule it +Basically the micro build sacrifices capabilities and performance for minimal size +If the new test differs from the one in the original patch you need to upload again and tick the grant license to apache +M released +Thanks + makes sense but how would work with selectors +Im going to spin up a couple clusters to test but I wanted jenkins to take a crack at it +This issue should be split up into a bug and an improvement issue +Because of that Im marking this as DONE when itd normally be marked as NOT REPRODUCIBLE. +Added patch +If an exception is thrown go to step +Proxied all repos on buildnumber mvn clean install with mvn went down from to mins +after this patch it would also resolve the issue described in +I dont commit the codes this is the patch for this issue and I will ask Tom to apply it +Looks long back there was a discussion about this behaviors considering performance and other factors +for the webapp module is and finalName is set to product +Luke would you please take a look at this +Is it possible to use inAny existing test you can point me to as an example +Please review this +svn ciSending Sending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexcommunitySending srcmainwebappSending srcmainwebappsiteTransmitting file dataCommitted revision . +Trying to figure out a workaround +moved to installer t fix because of the installer redesign. +I started seeing this last week but took various steps first like switching from JRE u back to u +More precisely the Hadoop test are being published but the class has not been moved to the Maven part of the build so is missing from the test +When I run mvn test what is the magic that causes the HDFS tests to be ignored on Windows and run on Linux? I can see the test fail on Windows with m test Dtest I just do not see the code that says dont run me on WindowsThank you it I found the hdfs Maven profile +No! It is a learning process for me +Thanks a lot Stefan looking good +rev fixes the problem in a new spec jar +Plan action have been fixedTransition action should also work usingactionstransitionconfirm for +Sigh +See attribseterr +Closing. +So the problem is that you have a cycle like thisMarshal Serialize MarshalEven though our objects implement serialization they may not be fully serializable +Check your server configuration for this you qualify a long time? Are you talking about minutesIn Studio the connection is seen as closed by the remote peer and this could occur for many different reasons but in any case Studio is not closing the connection itself AFAICT +Also reorganized the documentation to be clearer in general and include basic coverage for database selection and deployment model selection +The proposal is to add a new event which is between BBD and ABD and can do such a job +Bulk Close for all issues before incubation +I confirm it works now +Closing issues resolved in released versions +This patch refactors filtering and parsing of outlinks to a static method in +I also started writing the blog post for the plugins release +bq +not empty is not validated in following Metamer samplehttplocalhostmetamerfacescomponentsrichSelect +added Expiration memory cache cache is using to store elements on the basis of their expiration naming the new abstract methods in the base is inconsistent +Ive prepared a simple patch that solves this issue +An exactmatch of the entire candidate line is not required it is sufficientthat the boundary appear in its entirety following the CRLFHow do you read itIs a line having XXXsomething a correct ending delimiter for the XXX boundaryDoes this apply to intermediate delimitersI already created test messages and expected results but we should fix expected results based on our understanding of the rfc first and then try to fix the code +Made changes from previous patch to fix test cases in core which couldnt find or +throwing this into the bucket in hopes for fixing it in next release +comments on mostly goodMost of the comments are minor the only major ones are around lack of testing +Thus far we havent been able to come up with a suitable solution that does not break our release process and create cyclic dependencies +First in rev +so why isnt imacat or rbircher doing the changes required? The main reason for their access was to perform maintenance and upgrades this sounds like a perfect match for their access +This issue is pretty much a for us with our Rails upgradeWe appear to be caught in a nasty dependency vortex +I have tried with the word Hello to super Impose on a document and have run over this code multiple times in my server and one of the response is no english text but symbols displayed on the pdf and imageThe same gets repeated until i restart my server againRegards Harippriya didnt come back to give us some feedback +There is a new findbugs warning about the constructor of class Messenger in but there were no changes in that constructorAnd I think the core tests complains about the tests that have been moved from package to +Well its more complicated than that +Then on startup it can discard any that are corrupted +I had no time to deeply study the epilogue status. +Although deprecated could you please fix it so it returns true only when get throws a +Verified on revision +I will work on your comments and submit a new patch for the sameAttached patch is only applicable for Java level +Patch implementing three new methods removeAction set and didnt find any tests to modify and not sure how to go about making a new test +So even if I see no pb there I have backported to Jacques for committing my patchBharat Soni +I have to confess that we patched this class quite some time ago by using and directly +Reassigning to Ben for +Im not sure I applied all the relevant patches and understand exactly what is covered but this is at least a start toward very nice validation +I just committed this. +This is like catch problem +ve committed this although I just noticed that Hudson hadnt run +not with stringDoes anyone know how to the problemI use version of Hibernate and Oracle g driverThank you in and as well. +This branch is based on the branch +Added a section in the configuration chapter on using alternate JPA providers +Things definitely seems to be improved overall but weird things still happensSo +Didnt mean to close additional unit tests created as checksum fs doesnt support append +Committed to trunk. +Edited title to indicate an important detail since problem is related to type information inclusion +So I added the dependencies to +Beta branch is branched from code isnt it? So shouldnt we branch beta jobs from trunk jobs too +Looking through the spec its not clear on the naming expectations when using a Map +In reply to comment Hi Taciano youre right this feature is not you think this feature is reasonable +One last minor rev of this patch +We should add that too +Attaching second rev of patch +Closing again I reopened it by mistake. +bc master master +Thanks Vinod +But it gives a hint about the problem +My intuition is that this type of functionality could open up a new set of use cases for +Check exclusions before closing +The spec even says that it can have no value but the implementation is throwing an exception if we it to null +As initialises itself for each new context it detects useful information will be output about which logging library is bound to and why +Do you know where I could find the service names for the affected ports? I need the full Service name as per the other entries in this tableCheersif you follow the serviceName attribute in conf for each port you should get to the service that uses this particular binding +I have committed this to and above. +AlexCan you build the latest PBR from github and try your portal againThe fix Ive committed should resolve your problem but would be good if you could verify as I dont have a way to do soThanksKen +You just mention it only once in the entire proposal +Jannes a member of the ASF board +Theres also the fact that the memstores will be filled during write load so with too many of them therell be too much unnecessary flushing +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +I will attach a unit test in the hope somebody will fix this ASAP for JSR name needs rework +I dont have access to close it so can somebody else close this issue +COUNTY TREASURER C +but were okay if we change the avro client api in ways rightid say adding OUTPUTSCHEMAKEY is worth putting in the ifwhen its actually a problem humbug +BTW I took the shortcut with int values to avoid string comparison or keyvalue logic or something to cache those m generally happy with the results Im getting +Mike should we open another issue in LUCENE? Lucene Demo has the same packaging problem like this solr one +Why not use +Patch applies to a clean checkout fine and clean test dist works +Is the problem reproducible with a JSE environmentclose is meant to terminate the connection in an orderly manner by flushing output buffers and gracefully closing out input and output streams +Before implementing the pluggable navigation we need to extend the interface to be able register resources such as scrips and style sheets +Same applies to the Richfaces Component +CedricI think this alwaysRun kind of doubles up for run in all groups regardless of whether it has been excludedand run even after failureIs this correct? If so we could somehow split this up into two distinct casesI think the reason Arq is using it is to make sure that operations are done if something fails +At one point Pete Rafael and I discussed that each product repository should probably be configured to have its own and root file +The email for this completely skipped my eyes +If the dbf corruption happens again Ill let you know and open a separate bug the initial problem is still true even with the uncorrupted +I thought it might be a bug since I assumed that the should apply to all symbolizers +Guys it might not be clear from my wall of text however I did retest on a bit machine with the same resultsI also was thinking it might be an issue with that +The java version works as expected The given pdf uses a JBIG encoded image which isnt supported out of the box +Added cannot see a line with a trailing space so I dont know why the QA is failing +Ive built and run this on Ubuntu Precise and Fedora no problems +Here is the next iteration fixing all javadocs and removes the acquireLatest and release utilities from +make sense +I deleted just from the local repo and Im using the latest maven ve attached my log so you can check it for have verified with an SVN build as of a few minutes ago the issue seems to have been fixed +Well need to change that but can do that in another patch +Fixed a few spelling errors +No other changes between +Ugh yeah +is a support patch port of the fix to +Once COMMIT processing has sarted it just ignores the timeout after all the TX is going to go away soon anyway so introducing a heuristic outcome needlessly muddies the watersIf the legacy code cannot be made to do this then the best solution is to upgrade to a version of the which includes the Arjuna TS code +Committed in rev +TODO m sorry but documenting this feature is not very high in my current priorities +Sorry and discard my comments +Heres an current code is based necessarily somewhat loosely on RFC HTTP Remote Variant Selection Algorithm +Hi ysteinThe patch looks correct to me and the test plan also sounds good +resolved. +But I think it should +should be in +I guess that the bitwixt dependeny is a pending issue has that been resolved +This patch is intended to solve the issue with ruby arrays hashes only +Also updated the count in. +Moving to upstream and LATER +Resolved +If you find further issues please create a new jira and link it to this one. +Functionally same as last patch with addressing the comments from Eli +I see the problem it seems is that the same raw boot ops are used to do the reverse check boot up the legacy changes the ops which happens sometimes in the current controller we might end up with problems in +and again for separating directives out into query parameters leaving the path strictly for cell addressing +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +Also added the test case you wrote to the official test suite +patch applied! Thanks Issue +Patch for branch and trunk +Maybe we just tell user to get used to such output +This feature could be of general interest I think but didnt make it into any release unfortunatelyI couldnt find a version the patch can be applied to without much difficulty +Added sample FO file to test performance SVN patch as generated from Eclipse! Hope to hear some more results from the tests soonRegarding the note can you please provide some pointers to replacing the svg rectangles with fo based ones? +I verified it works when using via Install new software using the Alphaa updatesite +patch applied in revision Denis please check that it was applied as thanks! +The join is added to the base mapping select but I am guessing it would be best to put it in the eager select only +Also replaces FBOSgetBytes with DOBgetData in a couple places that look +I dont think this has already been handled exception rethrowing is still not configurable it just adheres to the sensible default of rethrowing only write exceptions +The langugeName is provided as a parameter so you should check for that insteadAlso we have a class in which may need to be adjusted as well +Committed to trunk and +Patch looks good to me +I tried to read more Axis codes and find that it might be required to consider more about throwing exceptionThe problem I think that Axis has different expected value for method +also has checks for being in and returning JDBC objectsThe embedded JDBC driver already has a factory for JDBC objects +okay +Caveats still apply to this version of the patch +Can the commit of this ticket include putting the state diagram in the master comments in one of these +You dont mention any defaults for min and max are there any +So we need to be very careful about throwing this only if Solr runs out of disk space. + +Previously the plugin installer displayed Processing null +Doug what are the relevant policies around using code that is in the Incubator +You guys are currently arguing about this one so the issue is still open +Patch applied on branch and JBDS CR +Its intermittent and it only occurs on AIX and then its in +A fix is in CVS +Sun Java RHEL HibernateIt is not a problem of Hibernate +I have posted a message on the mailing list +Delaying this fix till we address along with. +I know it doesnt now thats why I wrote should +Still failing occasionally +btw thanks for taking the first crack at this mind I found it at +StefanThanks for investigatingUsing a different version of ldapsych it does appear that this is using a single value for namingcontextsdnobjectclass topsubschemasubentry cnschemanamingcontexts AAsupportedextension supportedextension supportedsaslmechanisms EXTERNALsupportedldapversion supportedldapversion vendorname IBM Lotus Softwarevendorversion Release +Forked +commands are working fine + +Fixed +I am facing exactly the problem that I have to exclude all parameters but one when is a pretty useful feature +checked in. +Not convinced this is significant but something to be aware of +Lets open a more precise one later +I was wondering if it will be used somewhere else in the future +It is small but I think a good change +Closing the book was updated in the release. +Never mind +Heres an example project +Closed upon release of Hadoop. +Waiting for review to Mar +Note that the older is now obsolete with the newer patch +Closing resolvedwont fix issues. +Please disregard. +The change also confuses the purpose of the column family and Im not sure that we could document our way out of that. +This issue is probably fixed +I think the JDK one will be really hard to reproduce I have only noticed it once or twice and I might have the JDK in that instanceAs far as the ESB Runtime that is more common where is the log file for Eclipse that you would like to see +No need for them with regard to configuration +Please post a patch I will review it and commit it +bulk defer of open issues to +Per suggestions from Tigris removed prefusejavahlcommithack suggestion was to simply remove from the Central connector so that theres no default svn interface set to but ONLY +Todd +Still Programmers Guide says nothing about it and I built UI based on its advice +Requires rewriting a big chunk of the query engine so we can reuse symbols across plans +What should we state for an affects version when testing a specific unreleased version and it is not a problem with the previous version? +I agree that static list is not the best way but I dont think that including extra column is worth the overhead it creates +Moving to Id like to see this be a little bit more generic in terms of where the original vectors come from + for this patch +Probably true for Algorithm too +So I just removed the connect statement from the +Interesting to note that we do not use the in parallel processing modeI tried the suggestion of Ellis to use the current Request object as lock but this does not help requests hang the same way as before +Anyone else willing +Yes I did confirm offline with Ramya about this change before I even started this +The patch with problem solution +These make it difficult to track changes +For the record with v applied I got table not found exception within mins which suggests the error is in memcache +But that will make our installation slow +not sure what I should do +No apologies +I suggest to leave the flush there +Patch relaxes copyField source validation to allow any valid glob including those that dont match any explicit or dynamic fields adds test for this caseRetains the validation check that copyField sources must match either an explicit or a dynamic field and adds a test for this caseAdds tests for copyField source and dest glob validityCommitting shortly +r removes Perl license from and adds ASL license header +Believe there is no more work to do under this issueSealing and other security features can be added separately. +is intended to permit reading specific and generic objects too so this would be an incompatible change +This is a limitation of JavaUnlike C Java does not have a syntax to represent the of type hierarchy +But I presume the limit at which paging to disk happens may be configurable +A couple files missing headers package package +Hum this is a tricky one also broken in previous revs of HibernateWe need lazytrue fetchjoin to behave differently depending upon whether object is coming from or even cache or from a HQL query +This is definitely not an issue in beta due to examples merge. +need to ask miller if this is a for catching this RyanCommitted revision +In such situation eclipse closes file and then renames it but in our case page is still opened +Standalone example attached +Yes +Fixed in revision +It would be even more awesome if it didnt assume the entities were extending +JDK is automatically selected in the New Java Project wizard I didnt change simply cant create the projectsourcepathclasspath structure in to run a Seam I the only one who thinks that this is a waste of time and that we need a and a custom project structure for development? Masochists can still use this stuff and even deploy their into production +Changes discussed with Jim +I ran the test application provided by the filer on various versions of the Jboss +They really dont need to be at debug. +same patch without the apt doc update +Sorry should have said the line changes should be in webapp Ruth fixed in trunk r and r +The Other sources node is just a simplified view over srcmain folder +on war is gone +Will roll that into this patch +At the result these issues developments have to be delayed +Sounds good although I preferred the displayable name to the visible nameDisplayable sounds like what it means I wantdont want to be displayedVisible sounds like dynamic state am I currently on the screen +Attached docs on distributed builds +You should be able to drop these in place of the current site output and join up all the other sites that have been generated. +Patch as suggestion to fix +I can see similar optional arguments for various other tasks +Never mind if its too much work +Confirmed fix with test case +Hi Jeff believe this is now fixed using the new internal representation +Fixed under revision +The commit should be after libraries are bumped to. +AgreedSomething like skin or shall we switchProbably make now and eventually switch later +That implies something is stored and is going to be unstored. +I can see value in this change and I agree that it shouldnt affect existing codebases too much +Revised the example in the Database Creation section accordingly +Not perfect but it does the job +committed along with moving to +I still need to do lot more code cleanup +Looks this to +Please ask for help in the user mailing list. +fixed thanks for the patch +Thanks for fixing this! Please close the issue if the committed changes work for you +This should provide the desired functionality +Children of these node are appended to parent of encrypted node +Rodrigo can you reproduce this +Probably a week or so no set date right now +I agree with what was pointed out regarding the intended use of ijOne thing that puzzled me was the get and getTable entries in +hi armando +Ralph Goers at Cocoon has implemented the new to address the issues with +Please attach an example stylesheet and input document to help us reproduce the problem +This is a patch with tests that prevents duplicate feature name registration +This is my first patch submitI am not familiar with JIRA +Finally switch text to generated on the fly +Here the entire request is empty +using upload pom right? that never worked and cant work as the file is uploaded to the server where the modules poms are not isnt possible actually +Sounds like current speed is acceptable +These two patches add test cases for this fix +Put back if it can be ready in the next couple of weeks or so +This issue was caused by the introduced in which cased a whole host of issues +Seems a bit more than a low priority bug +I have some ideas on how this could be developed further to +You already helped a lot reporting the problem! But if you are interested in commiting some improvements just let me know and I will be more than happy to guide you +Simpler code less dependencies and basically does the same thing +For example it may run in out of the box without that special JAR +Ive pulled in the changes from the examples still fails +This would mean conversation with the same CID will overwrite each other +Please feel free to provide a patchRegardsPierre +tested manually on the command line +We have a few questions about our available options +And it was easier to code this way +Ive enabled an environment variable with the name screenshot on how Ive used the environment variable in my project configuration +Since Record extends Writable will not work here +Congratulations Andreas! This one was not easyI have updated several of the test cases with the output from this version +just committed this +I was trying with gutter +Youll be looking at and friendsMake sure your test coverage is good and Im certain this patch will be picked up quickly +m +Attaching a patch which enables jdbcapi to run in client framework in branch +Two things Perhaps we should change some of the colors? The bright red is kinda hard of the eyes and the dark grey on black for the end node makes it difficult to read +Thank you for your response +to remove a lock even when ppolicy is disabled +ditto! +sorry I missed the build break +Bulk close for +new controller automatic name +changes made and new snapshot has been deployed thanks +Bu classifier is marked as deprecated +Perhaps this alternate implementation is merely to bake in via a contribmodule one of these libraries underneath the hood +Gary thanks for clarification +HiFollowing the current discussion on a relatively close topic I provide you my own dirty yet functional solutionSorry I cant give you a patch that would be applicable to Lucenes because my has some extra functionalitiesHope this help but Im very satisfied with it in the context of an arabic analysis +Do you see a LOG message that occurs outside of lease recovery +ACK pushed as fbbececbdabc +for the patch +many of the exceptions have no info other than the stack trace +Sending Csvnorgapachecommonsiosrcchanges Sending Csvnorgapachecommonsiosrcmainjavaorgapachecommonsioinput Sending Csvnorgapachecommonsiosrctestjavaorgapachecommonsioinput Transmitting file data +Assigned a fix version of +I would only remove the blocks in the test methods and let the test method declare the exception +Did I miss something or cant constructors be compared using +Link to community AS feature +With previous key locking used for isolation level implementation the codecurrently may or may not lock the left key depending on isolation level +Anyone else have a comment before I commit +After you will end with them reassign it back to me to look for another components +Actually i can not the last available built console doesnt allow to add user without a domain +Adds in several implementations +Test harness for running multiple transforms by multiple threads in a controlled manner +Fix for the problem +This issue actually seems to be the same issue as +Since this is our public interface we dont want to change this unless we absolutely have to +He has credit in the commits also +No issues found on the fix +It may potentially need to be handled at the step level but Im not quite sure yet +I just realized that I didnt assign the issue to a component so if one of the admins could move it that would be great +Please try RC and submit a patch against RCthanks problems with NTLM use with should be fixed by now +sorting and filtering added new information and screens +Hmm Im not immediately sure how and would integrate together +For these functions we only have specialisations and no fall back general instantiationAll that is required is to explicitly export the specialisations. +Anyone want to tie this up today so that itll be committed before I cut the branch +Jeff can you test the latest builds to verify this is indeed fixed +Patch includes a new class which supports multiple table names database names partition filters and mapper classes +Here is the exploded WAR I took out hibernate jar files spring jars and aspectj weaver files to fit it under megs +Thus it may be better to keep two enumsBTW the enum names do sound confusing +bulk close of all resolved issues. +So the changes to and are extraneous +Perhaps a list of that redirected to each page could be kept in the metadata? Can anyone think of a better way to implement this +would be if master expected no changes except for line fixes +You might want to look at the enhanced image query syntax for scheduled for jclouds Looks really nice +Right now I see it too +Doing so will correctly ensure the dependency but it will not recompile Bean the first time +These issues have been resolved but never closed +I solved the issue by not allowing any other shapes other than polygons or multi polygons. +Lacks a test no patch is committed without there is a need +gets one new object reference +likewise namenode etcthe idea is that the test is done after the tcp connection setup as the first message passed between client and server +Hi AlanIf you just want to get this behaviour as a fair scheduler user the properties can be set in your +Thanks Amareshwari! +Obviously the methods still need to check that the passed queue! this and that the queue is on the same vhost +Heres my latest patch which incorporates a unit test further refinements as well as related changes we discussed on a separate thread +Hmm the memory portion may not be feasible for KVM +Thanks for the confirmation +I added john michael and proctor in on this to avoid keep spreading the discussion of this over too many channels +Reopen so we remember to change to COMPACT +Oops just noticed that the timeout messages are coming from the referencing module +The only case when this does not happen is if the file is flagged as binary which is not the case hereThere is nothing to change in the buildId suggest you perform a new clean checkout just in case +I wanted to do a variant of this for the following reasonsIf accidentally files related to block replicas are deleted on the datanode without a periodic block reports the namenode is not aware of this loss +I meant to add an argument to +Is there a region server log stillavailable from when this happened +Closing issue fixed in released version. +Ari is correct +Will soon post to +Could do with some tidying up and parameterisation but generally happy with it +Added license and javadoc toI ran locally and got all s +I suggest to use the same formatlogic that we already have on browse on +The plan is to eventually move to a more criteria API at which time we can add support for that +Ill also create a branch so this can go in a release +This is holding up our deploymentsRegards committed will be in EAP ER +This bug occurred after hours of continuous ingest +Since seemed to be working on this I put together a basic patch that addresses the core issue of potentially spawning too many threadsIf the core thread pool size needs to be configured too maybe that could be turned into a separate issue +A rename is as atomic an operation as possible with a file system in that it works with the file systems indexcatalog rather than the file itself +Are there logs from the AM attempts when this occurred? Last I checked the AM will recover map tasks if they completed before the AM restarted +It allows to inject other resources into features +Thanks Amos can you upload them to nexus? +Patch added +Can you please provide more information about these methods +The computation takes min per iteration and this is only for nodes and regions +Added documentation so I guess that the issue can be closed +With this change when the war is built by Maven the jar is put in the same place as when the war is built by the Ant build under lib + +Other message sets can be copied to a if desired do we need a new api in the interface? I recommend we just add the to the existing buffer field +The difference have been discussed to death in that jiraAlso see extensive analysis of Namenode protocols and adding annotation for operations in that was done over a period of weeks! Not to mention all the other operations that are made idempotent to make HA finally work correctly without boundary condition application failures! IdempotentAlso the methods without these annotations will not be retried +The above comment is our hypothesis +I also tried starting up after adding it and connecting in ExplorerClosing. +However if I understand correctly when a container is explicitly killed by the RM the special value of is reported to the AM instead of any exit code reported by the NMIf the RM is indeed telling the NM to kill the container then we would have a race with tasks failing due to the notifying the AM before the AM sees the container status from the RM +Some new tests including an observers version of +The interfaces discussed have been inplemented and will be part of the first version. +It will however be very rare for this to break in the future since its unlikely new Ruby revisions will substantially alter the set of methods defined on and +lib + latest patch from RB +Not sure about this +If anyone experiences this issue on websphere refer them to httpsupportuidswgPK +But we would not the mathematical computation of SIN +Thanks +I dont see what value this brings to the table but the overrides properties map would have to be added as additional parameter to all constructors which adds a lot of could easily write a factory method in your application code which creates a and applies the properties. +done! username is sebb passwords sent separately +Thanks Shelly +This is the same as working in HEAD and probably in +Tested the patch on cluster +Closing the book was updated in the release. +This class is used internally in the JTS buffer code to produce raw offset curves for buffers +We can either patch it now or wait for javacc release and patch everything all together +Configuration with BDBJE cache loader and eviction +Updated file for new patch releasesvn ciSending Transmitting file dataCommitted revision +Im calling this fixed. +This is not a bug and certainly not critical +submitting patch +static struct S is a different meaning of static though its static as in the OO part of the class not part of the instance therefore no this pointer meaning as opposed to static as in the procedural meaning global rather than automatic storage durationI can see it being useful even in an instance method to define a local struct that does not have an implicit this pointer and for that only the static keyword makes sense +Closing the issue. +See r +They are pretty well aware of Springs use of the there so they should hopefully be able to advise on how to avoid this problematic scenario +Thanks for the patch please could you verify that it has been applied as expected +Its to overwhelming with all these methods +All this could be attributed to the usual flakiness +Fixed in rev +I think we can implement the issue described by doing everything in the in the lines of how the links for FAILEDKILLED tasks are created on the job web ui +I wouldnt be surprised if cleaning +It looks good to memvn clean install finished successfullyI fixed the function name in the description as defined in the patch +Any update or ETA on this one? In the past its only taken a couple days for these tickets to be closed out so not sure whats going on? you please process this asap or point out any problems that prevent you from doing soThe latest Dozer jars I found in any Maven repository where which dates from +I can take a look at this +I have been hesitant to go beyond one level of association because once I go two levels deep someone will ask for three and so on +Do you have a reproducible test case +Thanks for your suggestion +Hi RolandYoull be glad to know this has already been done! See and you can take advantage of this through the latest nightlies +it still shows it correctly Did job history hdfsjob dfs +When these two requests reach the leader replica the current value of local column is +I am experiencing a similar issue with an app +If you think they are critical for CR please them +Agreed +Even version seems to retain to main artifacts manifest entries +Patch against SVN trunk checked in +Thanks again for contributing! Ill push the driver right after you send updated patch +for trying to add capabilities to detect whether a job has tasks to run without significant redesign +This is an indication that we need more client side tests for major functions +This is similar to is no longer supported in V of the plugin +I could add it in the patch +Thanks Anne! +From my reserahc into the problem it appears that the object is stored in the cache incorrectly + This bug has been marked as a duplicate of +Has this been tested when running inside a war file +I disabled the sort column functions on the page as it is not quite working and I dont feel sort is needed now that we allow users to filter contents +Well done ElijahPatch applied see r apologize I took a little more I stumbled once again to refused chunks +So lets create another JIRA for the +Ive stripped down the constructors to make it easier to pinpoint this issue +Committed revision . +Closing as this seems to have been applied correctly in the current code. +One with the producers and another with the consumer and run them on different machines +you are correctUnfortunately we cant call Regions equals method from a type constraint so I weakened the static constraint on copyTocopyFrom in r and added a dynamic check for region equality +looks good +You can reopen if you think otherwise +Closing old date issues +I already mailed Dims regarding this due to problems with Jira but Im sure hes a busy man so Im raising it here now +I closed as a duplicate but it has some good info in the comments that could be incorporated here. +Too many store files compaction is requested it completes but does not actually compact anything +Have the same problem +If works but fails the build this is not reproducible +So reopening this ticket +So becomes fieldTest +Different layout should be used because trace panel assumes is used and is added to with constrain +Ive committed this and merged to. +Could you please point me to the exact AS binary +I think Raymonds proposal is superior to mine so that should probably be the focus for the merge +Clearly its more simple and cute +This is using the from the trunk +I ran the tests a fwe times and did not see any failures +First version of the nutch crawl script +I can reorganize this and resubmit the patch accordingly +The file isnt read if the local binding is configured +Lets delete that dir do the same on and mail lucene pmc +will look for more holes and try some more times +The expected output isinvoked interceptorinvoked interceptor +It is supposed to be created during DSLD refreshing if it doesnt already exist +Fixed in revision +upgrade to released version +Thankscommitted +I dont think theres a good way to fix this in CQL. +Additional patch is applied at +A block report is considered to be the second block report if the number of blocks for a DN is During the test a DN sends block received before sending block report +Patch checked in. +The default was Mb I think I increase it to Mb +First created. +This patch let can replicate blocks belongs to under construction files but not the last blockAnd if the decommissioning only have some blocks which are the last blocks of under construction files and have more than live replicates left behind then could set it to DECOMMISSIONED +Could you provide a stack trace? Whats the Weld version +committed thanks. +I committed this to trunk and to unblock builds +Hum seems that I pushed the wrong file +So no manual intervention required in case of dependency updateSince we dont embed dependencies in Eclipse plugins itself we dont need to worry about syncing Manifests +Looks good +ok reproducible +Issues and inessential change from +Attached patch makes this test case actually run and fixes it up to work with the new edits log layout +Fixed r +But in your machine I think no +I agreed and we need more test for end of async stream +This issue is fixed in version of the plugin which uses Cobertura as Paul mentions above +under contrib but I dont think it belongs at top level in the distribution +If youd like to take ownership of this issue please be my guest and thanks +srcmainjavaorgapachehadoopyarnapirecords +By the way the same URL is referenced at line of +Thanks to Nicholas for suggesting the workaround +Also when running in Firefox on Windows selecting an item in the layered lookup does get moved to the field on the calling form but the layered lookup does not close +Yes I think should be in the same way Replace the hardcoded strings with an enum Use a builder instead of direct constructionI was planning to do this myself but I was hoping to get some kind of comment from somebody core to the project before writing too much code +We need to review and investigate our primary version would likely be jsf unless jsf is released prior to our releases +Fixed along with +works fine +In the meantime the not freed noise does not indicate a there an hserrpid file this time +I am closing this with Cannot Reproduce Bug status as it seems to be working and a for this has been added to the AS testsuiteFeel free to if you hit the issue again. +The patch in fixes sellitem at least +Other languages allow it and i think it reasonable for a user to expect this to work +If you dont provide a solution for your pb in clearcase provider well cant fix it. +patch is available marking it a candidate for for working on this issue Yip +If you have downloaded then take a look at webhelpdocs in a browserIf not here is another screen shotTom +One other note this change speeds up data loading by in our environment. +orghsqldb matchesWere these ruled out as last time? +should not be backported to as SYSUSERS is a new catalog added in +I forgot to remove them prior to posting and I dont have permissions to remove the attachment +Could you fix it and upload a new version Ivan please +I suspect theres a optimization to be performed to reduce the output of but in its current state this patch works and should address the problem +There are also many sets used where would be more efficient since they can be truncated to fit the data or initialized to the proper size to begin with +If it is also missing component please set both so that the issue can be properly triaged +You just need to use the right port number to talk with them which I believe is Im not sure this is causing your difficulties or not but it certainly isnt good +Original problem with incorrect namespace is now fixed +fixed this a while ago but forgot to close the JIRA. +c Is the commented line still required or shall we remove ita yesb dbadba is a default SAP DB connection string for the DBA userc that line just indicates that the SAPDB driver should be usable working in the SAP DB Oracle SQL mode +The is not even in Apaches htdocs directory +The way major compaction currently works in Accumulo older data will eventually end up in the largest file +this patch is out of date needs to be wrote shock and it was contributed as part of Pig to Apache so it shouldnt be considered a separate piece of code in that respect +I will upload patch +because getOK has a dynamic return type we are losing the type information for the x variableI think a reasonable rule here is that if a variable that already has an inferred type has a new inferred type that is either Object or def then we ignore the new type and keep the old one +no further see the ONLY in you open a new bug report with a runnable test case attached that will help me track down the issue +small updates started working on the solution +It was considering compressed bytes size for calculating the totalBytes because of this it is showing the progress as more than +So all issues not directly assigned to an AS release are being closed +UriThese all sound like good suggestions +Injection is supported within this object realm under the simplified API docs +bin +What is version of Hadoop? Is it +or you can set the prefetch on the broker via a destinationPolicy to set the prefetch for all or a matching set of destinations. +the sampling job fails when is not configured in pig property file +I agree with Uma this works as designed +Since fixed in trunk. +I also ran all the tests with the patch +Please complete the bug report as it stands now its going to be closed as incompletenot reproducibleEither add your complete source code or debug your code to find out what are the threads doing when the unexpected behavior occurs +Im this issue and setting back to its original resolved state as it was erroneously not verfied under the WSM framework +Now I see my thanks to you Mladen were truncated from the original comment? So thank you for your help! + and applyNowload paths use directly instead of converting to Rows first +Do you have a good test case that can reproduce? Some of the tests in are supposed to test +What about a new class +We could also accept both scenarios contained in the gliderwhere the glider can have as many scenarios as the user requires +The included in the metro trunk should be a more up to date than this one +Otherwise we would need to make it backwards compatible somehow +I would like to see much more work into creating proper widgets that can be used in multiple places +i gave you permissions you are the best take a first cut at the document and post the link hereWe can always review it +I would have to write the unit test against because my rule is failing to compile with +All tests passing. +things need to be tested +Android already implements this! has been implemented +and it would make the code more clearly with netty styles +This works correctly in both JBDS and JBDS +Youre right Freddy so closed. +Commit to trunk patch works for me +Committed to trunk andCreated to continue discussion on using error codes and enums. +Just verified this bug also occurs under on Win XP has been verified onServer with JDBC Driver Server with JDBC Driver Server with JDBC Driver comment is intended to show why Ive reported it as a bug if someone comes back and says Nope its part of the design and heres why Im all earsNow maybe we can focus on the problem and not your misunderstanding of my language +if you can create a quickstart that reproduces it please reopen and attach. +Thoughts? Thanks +Revision fixes the issue by only running the fixtures that require Subject if the class is available +Move tests in that dont need the full stack to and remove some tests that dont seem to improve code coverageAs for application specific logic one option is to do that outside of Producer + implemented status page changed default redirection behaviour adjusted some tests +Tests passed cleanly for me on +Hi WillemI uninstalled bundle to avoid exception with uninstalling feature +and Currently for all types of response we have moved to the model that there is no global error code at the response level +We still need Ajax push for Recent Activity +Its also why Im slow to give out random +I didnt know we were discouraging this +Resolved with revision of Commit message In do not create a separate +Addressed Mayanks comment +This however is a problem is the Bean Validation specification which ships with a simple default resolver +Mahadev thanks for your comments +Is this a default behavior you need? Or you want to have some value being the default and configurate via webdriver extension +The patch file applies to trunk only +This is the sample image after the fix I does not suffer from these issues has been retired for quite a bit of closing all issues that have been in resolved state for months or more without any feedback or update +I have another issue found and will create another ticket so that this can be closed now I you could add a ticket about the chickenegg problem you can better describe what the problem is and how it may be resolved +This patch is well tested along with patch for +Deleted tcktestjdoapplicationidentityorgapachejdotckpcfieldtypes tcktestjdodatastoreidentityorgapachejdotckpcfieldtypesrevision to set the Fix Versions field to JDO beta. +If a first request has its response delayed should processing of a second request continue immediately and that second response be queued until the first response is sent? It could have a performance benefit +I thought usually the submitter of a bug could do that +I did do a functional test and it works fine +File Contains a one line change to add the NOWAIT directive to the FOR UPDATE statementThis will allow slave instances to gracefully skip blocking on a locked table +Do we really want to store one blob per file? That could create too many files on a node +Richard Friedman askedAlexey I noticed our EJBs are being generated with xdoclet and jboss Does it matter? Should this be upgraded for jbossAlexy respondedIts not problem for me +This is not the case because what is returned is the Trinidad which does not have this functionalityNormally we could simply wrap the Trinidad object in our own and change the +Hello Alfcan you attach the example code please +So the EJB project is not in the WAR classpath +I was not planning to increase default buffer size in this patch +Mass closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Let me try other browsers +I suppose we should look up if there are any policies on how often to publish snapshots +Submitting patch for Hudson to one test failed in the previous Hudson build and it looks unrelatedI run the full test suite in local with the patch applied to trunk and all tests passed +We also need to upgrade to Tomcat which includes new security fixes +More updates to be carried out on this attachedModified to properly fetch dependencies for std and min distros and also to reflect the proper std binary ditribution +The bug is also present on Sachins fix doesnt work for me +RomanWhat version of Cassandra are your server nodes running? RC +Somehow Cyclebreaker is involved again in the issue +I should add that a JTA tx should really not be used for this use case its incredibly the absence of any explicit transaction management type indicates the default CMT with default REQUIRED tx semanticsAs for the message redelivery semantics the EJB spec doesnt list down the rules and leaves it to the JMS providers to handle it in their own specific way +I moved the to apps and changed its content so that the notes node is createdWhen requesting I got errors because of body properties of child nodesIm not sure if my changes make sense or if my installation approach was wrong but if youre interested I can attach a patch with my changes. +Patch checked into branch. +The intention of is absolutely to provide the necessary hooks to be able to support either one +Committed to branch. +cant wait to see you patch its gonna be huge I guess I will be here to help you splitting it apart and get you good work into contrib related patchCouple of still noted in the patch related to distances +Since Hive seems to work with both version and of Im having trouble seeing how this can cause problems. +Hi Alexei! First concerning and it is right fills BITMAPINFO structure in two stepsSecond I will prepare patch for Linux ASAPThanksIgor +You can make it pretty small and still be able to read it +I agree that cluster ID is a nicer construct than namespace ID +Will apply this tomorrow if no s +pending messages in a TX is kept in memory +Looking a bit deeper this looks reminiscent to this ticket where an older version of hibernate is being included +How is it looking now +requires python needs to be run from the base directory +I updated this patch to make it work with the last changes in issue due to lack of activity and without satisfying resultsIm still happy to discuss how we could address this but lets file a new issue for when that happens since these patches are very ChrisAm actually trying to work out a stable patch here that Ill be using for my project +Will make the static long member in Writer as volatile +Hi Richard This looks like one for you +as were doing in +myself via Brocks thanks for the review Brock +I end up having to force quit uDig and start I repeat all the steps above I cannot get the of schema names to work at all even after I click the Lookup button +Ashish have you had a chance to test this? All existing tests pass for meUnless there are objections Ill commit it soon +The only failing test has failed for the last builds +Applied patch from Allan Schrum +Well now it fails because of +If we create a specific proxy for each user type the need could be worked around +Attached is a patch proposalThis patch modifies the processing of GROUP BY columnswhich are generated into the selects result column listSpecifically the patch causes columns to be generatedinto the select RCL only if the select does not specify DISTINCTIf the select specifies DISTINCT we dont want to includeany additional columns into the RCL because we want tobe sure that we only perform DISTINCT processing onthe columns that were specified by the userThe patch also includes some additional test cases basedon the repro case in the issue descriptionI ran a complete set of regression tests with the modified codeand it passed all the existing tests as well as the new testsadded by this patchPlease have a look and let me know what you think +The connector should also be able to call the getUpdated operation as well +Trying hudson +You can try setting the updatejdbcdebug flag and running the script to see if some sort of error is occurring that isnt being reported or you could comment out the cleanup of the jdbcclasses at the end to inspect the that was jarred up +So basically we want to enable the user to have the ability to add new args or other args but any changes related to the runtime will be overriddenAny suggestions here +is the lookup path for native libraries loaded by the java program being run in most cases the place where the JDK lives is not relevant as the native libs will be somewhere near the +It is simply impossible to lock at a finer granularity than that +Jarkko feel free to if i missed something or messed up something +Can you look at this and add this to the plugin tree +Additional cleanup of a document sourceScaled jpg files to fit page in a docbook document +on the other hand are eligible for collection the first time the GC sees there are no hard references anymoreIm not sure of the best way to solve this +I use the +Reto are those modules already release? If so we should upgrade before the next Stanbol they are in use in the ng branch +Agree I vote for without LF +Post we will look into better debug support which will include looking at scrapbook pages for groovy. +merged and committed the fixed from Carlos Thanks +HiThis issue can be considered as a fix for transport after applying the patch i provided for thanks the patch on the and resolved that hence this has to be resolved as well as per the last comment from Charith. +Yep Im working on the snapshot downloading now +Have not used it before but I will work it out ahead I would like to share formative ideas about the packaging and Roo commands +Changed to an assert +Dont waste time trying to understand if you can help it +Hmmm yeah +Sorry this has been a hectic week for me too +That seems like a relatively reasonable fix to me +already adds to the WAR file so a fresh WAR deployment will have this dependency installed +Updated patch resolves issues with wildcards with too many matches +Well actually I think it has more to do with the fact that the method coerceArgument is protected +Running it through Hudson +Apparently that is not a correct assumption Thanks. +Status rules can have the same salience value it is also logical to have itPage object paths now esb and mappingPage key field is visiblePage signalPage now bpm and esb millis is everything I was able to Fixed up the and as this task had already been moved out of the against +Is this contained within a single class? If so maybe you could attach the entire thing +This feature would be really interesting but please keep in mind that dosgi hadnt been touched for about a year and there are a lot of upgradescleanups required before it is usable again with the latest versions +Changing the fix version from to so that this issue will appear in the filters used by the release notes. +I will investigate the differences and try to post a patch +Attachment has been added with description fix attempt updated the patch to take into account Helder Magalhes has been added with description animation of with onload ecmascript to modify the has been added with description scripting of the has been added with description minor formatting fix +Is using the correct media type maybe undertow is stricter about what is will accept +I didnt modify the logs +Second existing partitioned union or just general union optimizations will not be used since the final union form is not added until after planning is finished +Each client is different on the wiki and there was no uniform tutorial to follow for all clients +Create test for update of issues with fix instead of the correct +tools core docs site s do it! Heres the breakdown at the moment from what I can tellTools Docs Site Core everything else outstanding +Unless something unregisters the shutdown hook it runs on a clean exit too +Attaching also separate patch version for since trunk version does not apply cleanly +This is the qnd patch to the core that puts only the elements into the map that are actually present on the feature registry +While I think from a representation it probably does make sense to include the workspace as part of the entity itself rather than having to parse the uri but in the interest of keeping things consistent I think leaving it out makes sense +Exactly how the inheritance works is somewhat constrained by abdera +I will be ready together with my english dictionary at the hand +Mind uploading a patch so can run? Ill review the combined patch +the caja test which were missing from the last patch +Awaiting clarification from working group on whether the spec should state an assertion requiring that the http location must conform to the grammar +test case + +No luck theres apparently not a Windows Server instance available in AWS +Is this still the case +OK lets use differentnew to continue UI development +Also adds instance management functions with lifecycle as an optionAll unit and integration tests pass +Uh doesnt look like I have commit access to +For the record I think that the compact override syntax is stable now +From what I have seen so far yes I believe this change should just not have been made to the schema the addition of the schema which was subsequently renamed to should have been sufficient +But for this Im unaware and we never fully specd it out +Does anyone have a clever idea of how to take a snapshot of a running very active and very large? Without things timing out like crazy? Were ok with the idea of just stopping everything once in a while and restarting the namenode +v serializes to a raw byte array to avoid the copy +Rebased the patches slightly +The break up of the word is incorrect +Carl What does that mean from the resolution of this bug? Will this be addressed or not by Apache Hive +Which files are requiered for a complete test caseBTWWhy do you need a test case its obvious that with collections is broken since +Thanks +No these finalize methods have been by me +Theres now a patch available in that fixes the Forrest JDK problembq +Thanks Tim +Thatll allow those updates to be addressed separately +I have generated and tested the Service +I didnt delete the original from the derbyall +It seems its a random error present only in hudson +I have committed this +The patch adds one more manual command in the manual step of the wizard sudo su l hdfs c hdfs zkfc formatZK This also removes stop namenode command from which was introduced as a requirement in +Edited summary for readability +I personally think that we do not want to but I also want us to fully think about the problem so we dont paint ourselves into a cornerYour initial ideas seem very reasonable to me +This fixes the hang on this might merit backporting to at some stage +Frankly this ticket is wrong the threads EIP has fluent builders for setting thread name and any of the other options +Len could you let us know what is for? EAP doesnt bundle nor integrate ESB afaikMarshall could you please double verify that we are using the current release of EAP in the build +method in in branch now to reproduce this +Fixed +Hi DevinI have reviewed and pushed this patch +Ill fix that sometime today +Now that work is underway with AS all previous community releases are +I changed the tense to be consistent with the rest of the paragraph +Hi TeamMay I have any update on this issue? As it may be an underlying risk for our project deployment we hope to have a clearer sight on this unexpected you help on it? Thank you! And if you required more log trace about it we will try to provide all our collected regardsRaymond you using TRACE level logging +I see +just committed this. +Leonid I need more details on this one +additional fix to keep the stack trace at debug level committed to +LGTM +Username crazyfoam +for auto as default this is IMHO the best fit for most about a configuration option for auto so it can be reconfigured to +Isnt there a state to mark done but uncommited +Yes it should +I changed of the action providers +Nobody is working on this or has plans for doing soScaling down to a Low Priority Wish and Unknown fix target +also the original file that gets copied is supposed to be anyway but that doesnt work either +We shouldnt be doing validation if there arent any BPEL resources in the workspace +So I have not yet checked it in +Duplicated here there is an object that can be returned by an event handler to send an error response to the client +To do thisput the version in libcommonpluginput the iOS version in libiospluginYou should also update the documentation in by making edits to the repo +Moved this back to Known Issues as per s comment +I will share my findings if something interesting comes out of these experiments. +Can you please upload the difference between test execution times before and after the patch? Just checking to see if we are bloating the execution time by an order of magnitude +However untill then I cannot workaround it from stated in having proper jaas auth also within osgi is on our todo list and we are aware of the problem +There is not much I can do about this until someone provides me with an example that reproduces the problem as I have not been able to reproduce it locally. +before GPF +Please check that Ive applied it correctlyIm going to the Jakarta Commons Lang tests that caused me to raise this bug +Publish new bits adjacent to old ones +I believe this feature is working as designed in Roller so Im taking this off the listIf you want to change this behavior please propose a solution to list. +Seems thats it +Thanks + +Closed after release. +Anyway it works on AS +From snip from cvs logrevision date author mschachter state Exp lines +bulk move of all unresolved issues from to +If that doesnt work we set the default content type as application +I thought were falling into Douglas Adams kind of narrative +All issues mentioned here have been resolved on the branch for release +Delta to HA branch +Ive confirmed that the cause of this bus is in fact the +Its keys are very basic +Also included is the crux of the the job submission and monitoring part +I updated the svn and then created this patch after all unit tests ran successfully +Fixed on and closing all issues that have been in resolved state for more than a month without further comments +Updated to include the exclude patches submitted and yet its still not applied AND the module is now broken +These notes describe how to recover from the upgrade incompatibilities introduced by the secure server work +I meant JDK reported this is not our bug. +Please discard it and try this patch +It appears that simply setting TRACE in the simplelog properties MHOME was not sufficient +So I have changed the code for merging such definitiions +Hi SteveI started adding JMX interfaces prototype on trunk do you have spare time to review them? Feel free to rearrange things Im not JMX expertMany thanks in advance +New patch which uses wrapper as suggested on the dev list by Blake +Therefore this test again fails because bean is not passivation capable scope +However if you want others end users to use your libraryyou need to make it as simple to use as possible that includes minimising thenumber of dependencies +I have committed this +There is an unrelated change in the file +Ill post a patch if I make any progress +Andrew was that you when refactoring the Server classes +But its failing in our setup what may be cause +Hi includes the fixes needed in srcjniRegardsSenaka +Manually integration tests +Since this would be an incompatible change it is unlikely this can be pushed to a release +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issuesCan you give more detail as to what kind of data you want to save in the properties file and where that data comes from +will commit Tom! I committed this to trunk and. +See for further details. +Its only an issue if the KV is from an hbase catalog table +Updates each time the server is started if the file has changed +Made executable and committed in revision . +fixed in revision in +Please discuss task allocation in the forum threads +removed and +Kevan thanks for the link +after analyzing I may found a bug resulting in the unexpected behavior mentioned in this issue +If there was a problem as you mentioned then I would expect that the document would not be indexedI am toying with the idea of using TIKA however I cant find an example from which I could work from +Fixed with revision +Please give it a scan and post any updates or correctionsIf nothing is found I will update the site in hoursThank you all +I updated the patch so that everything would continue to cleanly applyThis is actually against the sqoop committed +good catch into trunk and +Do we have any plan to implement this featureThanks + as a bandaid for a DR release +Ive committed thisStack I committed this because this bug sometimes broke our build so it was very annoying +push off for when we have time for a significant has been done +Waiting on QA confirmation that it still t reproduce on EPP or trunk. +Verified. +All regression tests passed +Yes they are checked during initialization to make sure they are directories +test case reports refresh instead of find but thats just another with mvn clean compile exec +Im leaving the ticket open however +Knut and Rick thanks for your feedbacks +v contains everything except stop use a notification on sleeper to lower shutdown time by sI will create a specific Jira for this one as it implies some changes ineugene with a specific jira it will be simpler +Never mind +Now that being said I would want to propose a new issue to use the check methods of the Guava Preconditions API to their full potential later on + +I am resolving it as Wont Fix since it is not a problem with SI but rather a minor STS issue which based on Christians comment has an elegant it will be resolved wit the next release of STS +I tried reproducing this in an application of my own but I havent been able to reproduce the deadlock +flushwriter blocking is a feature otherwise you OOMThat is assuming that the memtables are large the user had a large number of and so each memtable would be relatively small. +It could be removed by adding a check for JAVACMD in all the scripts that call or perhaps by letting the script call the end label of the parent scripts directly but I fear that the latter option might be unnecessarily complicated to implement +Hi ChrisSorry for the delay I didnt notice your commentIm ok with your changesYou can commit this +If we the mechanism from then we will probably see more holes in the lists of generated identity values +Add a kind message in to delivery the vertices size messageOther changes are in +Committed revision +what was the install information? I havent used that mojo in quite awhile +All the findbugs warnings are due to sources generated by +The change was contributed by Daryoush and applied by Michael. +I wonder if this is related to instead? Ivan is it reliably reproducible every time you sleep your machine or only occasionally +Patch of modifies not only entity but also class molder of proposed entity +Ive tested with the test application and then it seems to work fine +This is only happening on the RHEL machine and I did manage to get ant installed and the issue was still happening +Sorry about that +Fixed an import warning +Hi RaviThanks for the commentsPatch is updated with the commentsAbout the test part Executed the test cases in node manager project locally +I still think it a good idea to leave it out here +Patch applied +This matches Words own default way of building the table of contents +Thanks +Patch committed to trunk +prj file +New patch +I extend to take an array of +Well I forgot there is the matter of the name of the property as we discussed with regard to my question above +fixed IT + +right now we are deploying to a project local directory +Agreed option is bestOne minor issue is that the would need a new negociation code +binjruby samplesffi binjruby samplesffi qsort After qsort binjruby e puts it still fails after updating from git here +My stands across a rebase +KILLWAIT should handle cleaning up any lingering attempts and by the time the task transitions from KILLWAIT to KILLED there should be no active task attempts and therefore no chance of receiving TATTEMPT events +Should I go ahead and move this to be resolved +testadmin failed because of too many open files +Raising to Critical as this this shows almost every time the subsystem is started +the last two comments are posted to this tasks mistakingly the comments correspond to Dev jira tasks review for Tools javahqleditingandvalidation +Those are not used anymore with your changesSorry missed it yesterday +I had one small issue highlighting did not seem to work +Fixed the jbossweb branches +See the HTTP Connector Guide for information on disabling stickysessions then +plugins +Replacing the algorithm wont change anything to the problem and adding a salt is not difficult at all +Sorry I tried to have a look in the documentation but I didnt see it +i appear to have the problem too when Ive got those projects defined +Although the inner class has a reference to the outer class they are nevertheless completely separate objects +Committed to trunk +FYI my more intensive tests on monday have been done with success works better than Maven Ant Tasks for lets work on to get a release +JM thanks for the patch btw +Please see updated demo at +I thought about doing all those changes in this patch but changes were more involved then I thought +So User can not add modules like Addressing using his if that is disabled in +Central includes maven stuff and Extras contains EXTRA maven stuff +Thank you Prasad! +I opened. +But it is held back because of bugs or missing features in modello +Done +picked the wrong jira issue +I dont care about the Netscape bug but we should look at adding validations when we look at adding radio validations + describes same behavior with WAR component +Reported upstream issue fixed Metamer and it seems to be OK now +Based on discussion with Gary there is a valid use case with NIO where must be done by evaluating custom header in place of the sequenceNumber header +Please provide steps how to verify or close this Jira +I guess we will need to address the introduction of many more build time configuration parameters as I can easily imagine the scenario where we want to stay with existing arrangement want most flexibility in managing memory allocation and thus almost all spaces but the nursery become overlapping on bit architectures where virtual address ranges are cheap make all spaces non patch still needs some cleaning up and straightening out but I think its fairly close to being ready to Ive timed out for today +Thanks +Ill post the results laterPlease note patch is also pending reviewcommit +This will ensure that production systems do not get impacted by an infinite loop at the same time inspecting the log will tell us that the software is not running optimally +PDF processed with pdfbox processed with pdfbox +i waited a couple of days james but didnt see a patch so i put this in myself as it is also useful to us for our selenium definately do not think we need to go this far just find a character that works perhaps escape with its good cheers Igor! +Also remove the scope on your since it should be a singleon +Thanks Narayanan +Removed dependency on +Changing to an m not suggesting an improvement Im suggesting a fix for the bug that Roo does not parse the POM properly +Next we perform the copy down +I dont like the feature that changing settings in the configuration after the FS was created is reflected in open streams +Evgeny I couldnt reproduce this problem +Thanks for your contribution Jarcec +For me it was because there was a change in it from another developers commit that I didnt resolve +From this information we will write the final release notes for all issues +Looks like we have incomplete merge for awsapi +So I dont think its worth blocking this patch on it given it makes good progress getting the tests back running at allThanks for the contribution Henry! Nice speedup + is now released +Verfied on Affects and Fix in Version +Committed for trunk in rev . +ve committed this to trunk and. +In addition as suggested by Steven on Friday we could move towards using fully qualified class names rather than +Setting the property to null didnt reset the timeout +I do not know what its complaining about +On my machine RIs execution time is ms without Xmx setting and DRLVMs execution time is ms +Trying on Hudson +Let me know if you can do that +cant reproduce this any more +They were zapped as part of the migration to the new version of new host for Bugzilla +Bumping issues to except those that involve with Struts or +I suppose we should leave it open for now then +We could instead specify the dev dependencies such as javacc as optional in the pom transisivefalse in ivyTwo things jump out as definitely important to separate +Randall I sent you the maven module per and I attached also here +we should never use the attribute names in or directly but always do a lookup to the object +Thanks for patch Jon. +I had a sport injury and have been offline waiting for it to heal +I dont believe so I think the new append design in trunk prevents this issue +what does not mean that we could not extend it by some timeSo first it will just deal with SP and positive weights +Thanks Tom +In our case it uses +Thanks for the test case I was able to reproduce +Generate the ecore model for the JPDL definition +I was writing to quickly +Hi DanCould you please let me know whether this is a valid bug and if it is can you let me know whether it is something being worked upon for next releasesYour help is really appreciatedRegardsPrakash +The JAXB framework requires them to be defined in the order of declaration in the file +Errors were caused by backslash in workspace path. +this happens on all cases where the user has no other two seem to be redhat for prod +If I am right I can work on it for a patch +helps but still could result in a race +Thanks for the patch Andreas really building the distribution you should be able to do that with a ant either from the svn sources or from the source distribution +Another example is that touching can throttle webhdfs accidentally since the code is shared by Hftp Hsftp etcI assume that youre running your jobs in a trusted environment +It passed + +Havent looked yet but iirc this file is a port like the rest of the new spatial package which I ported last year +can you try with a nightly build +The files will not have identical content! The POM files in the old location must have the relocation entry while the ones in the new location may not have em +The model data could not be loaded during the +Cant reproduce after r thanks Gurkan +Issues Mailing List Project Name Hama Project Key HAMA Jira Admin Edward J +As Robert points out if in the future we get more sophisticated about checking the classpathlibs it will fit well there +weird +It looks like theres a ton of effort put in to avoiding making sort work w nodes +So on one of the places it has to be removed to avoid the conflictlet me know if you have any questions and I would be glad to help +StackIts better to set it to instead of fixed port +Its alpha stage and now too late here to come up with something more reasonable so if you feel please commit it +Thanks very much Kathey +Nit it would be cleaner and easier to read if we add a new allocate method and have the existing allocate method simply call it with rather than change all those places to pass Speaking of leaking containers is there something we can do to auditassert that applications that have completed dont have running containers? If we lose track of a container finished event the consumed resources are going to keep increasing indefinitely +Note even though the POM in the attached project has a jetty plugin I couldnt figure out how to get the embedded Jetty to run with endorsed dirs +resolve as out of date because of month no activity +Does not matter if Seam logger is used or not +Cant reproduce this anymore +yes please +Heres my +Please do not open issues as critical bugs when they clearly arent +Maybe should file a ticket for to allow Exceptions to be checked for the a possible annotated expected Exception typeRegards agree that you should drop automatic flushing and so allow support for expected have used this approach a long time and found the new semantics suddenly broke a whole lot of my old tests +New patch A fix to the fix and some more random testing that attempts to catch this it doesnt seem to yet though +In its place a new file is created highlighting the differences between the embedded and network clients as outlined in this issue +fixed the grammar right nowDerek +Thanks +Juergen can you please attach a sample + +Ok so after a couple of days we have the same result in the test wiki alsoSomeone needs to check the the autoexport settings I think +No sorry not at all +applied to trunk +Fixed. +It probably is related with some bug in classes which manage +If the next call on the iterator doesnt get any message within that time an exception is thrown during the next call +Thanks for clarifying BrianWell leave it here for now +Shall we check the directory name ends with +It was a complex structure that had a map a list of maps and a map of maps +requires small modification to of Final +As for yourcould you pleae fill a new issue for this +On controller failover this is the value read from zookeeperOther than that fixed couple other issues +I should also add that is another use case +I have applied the fix for this issue in the was also going to fix this in Trunk but noticed it was no loger there is that expected +Removing messages from the ARW needs to be governed by a CAS as multiple OOB threads could be doing this test is SCOPETest +I prepare new one +I confirmed that updates work with the latest nightly +patch committed to trunk. +We can prevent these exceptions from being thrown by writing correct code so we should +Sorry for my english +bq +Using our dll makes things work correctly +ExcellentI think the JPA annotation may use updatable +We try to avoid long discussion in jira +Im guessing that the performance difference would be reduced if couchdbjsondecode were modified so it only called xmerlucsfromutf when it encountered a byte value xF in the binary +Ive committed this to trunk and +That would be great +Thanks Matt for nailing this one. +PID now in logs you DanSending srcscriptssnapinstallerTransmitting file dataCommitted revision +YesThe problem here is it seems that is not designed to work taking url with parameters for GET operations +Two in as and jst were fixed by copying particular functions over to the JBT source code +Maybe for later +Messages with an expiration cause problems when the producers clock is not synchronized with the server +please if you have a better idea +I believe that in Ruby storage will remain active as long as the class is defined +Since Ive never been able to reproduce the issue Im not sure if that will really help +i figured out the timing issue for +It handles comments and continuations correctly +Pinging Ted Jones +Ive created app from scratch and it fails same as before +It is pretty big even if you divide it by the replication factor of for writesAlthough I agree it would be better to have a method of calculating the limit based on some natural criteria like hardware configuration or heap size +Yeah thats fine +but if you get this resourcesmarkup how on earth do we know what to get? Because we lost everything that is related how to get it We lost the classname that we want to search for +We have an article ready to publish soon there we differentiate where we want to use this and when Arquillian stay tunedNils Is this fixed for +Second part of application +It will also be impossible for users to deploy existing SOA P projects to is that really what you want? +There was a small mistake in the first patch +Bulk move issues to and +I noticed that you might need the switch on the generator name Visual Studio Win +Have a draft just need to commit to the repo +on patch v +but we still need it +I think that is a bigger issue than the length limitationAgreed +Raised in mapreduce project to fix the compilation issue because of this api change +Ping +My guess a text node is being added twice probably one in front of the other +That sounds correct to me +Please add the apache license to if you grant the license to the +closing the issue +Ioannis this looks great +Disregard the comment about PEAR files its not relevant to this issue +can still communicate with new version because the enum values are still the same +The version does not have this overflow in its gcd implementation and it can be easily preventedBut as written in Id drop the Fraction class in favor of Fraction +some may just need connection settings changing while others could require users to be modified or even test operations to allow the behaviour of a realm to be verified +In my local tests it seems that Value is always reevaluated but expressions in XML property values are cached +hey guys had this been merged by now +I should probably mention that the patch was generated against version +I have resolved this job as the necessary code change has been made and the associated documentation has been corrected. +Ah ok that makes it clear +Also the methods such as get are trying to call the static method find but are using the Policy class and not the class +Marking as a parser issue +I dont have the source here +Thanks Eric +Move it back if I called it wrong. +Hi Sanchit these are neat +Scheduling for test case has been replaced by maven project thanks for the test case +We have lot of constant strings in our codebase +Hudson has been the most disagreeable of late +It works on Linux probably because the widget autoselects and so the event fires +Okay +Diff to Karthiks +include an ugly fix to make Hadoop working +Patch with the tests as instructed by David for the patches Manu T George. +The JTS implementation uses so made them match +You can remove the backward compatibility tests as part of the patch +resolved by thanks +would still be an issue in my case so thanks for pointing that out. +I would suggest we have these checks at the API layerSecondly I can add some basic checks at the UI layer too along with the tooltips which highlight what sort of input values are valid ones +did you have something specific in mind +It works great +distdevforrest distreleaseforrest created +move to by End of March +Nite the finally clause should not be on its own line +I committed the patch to +Thanks much +The Parent dup class problem is solved by. +Planning to commit today +Patch files against the head revisions of trunk branches as of October The code makes use of Java language features which is in conflict with current project style +Need to resolve these issues for maintenance release +If my hunch is correct you could be trying to return more characters that fit varchar and then you could get in the CLOB area which is then probably not supported by the type you specifiedRegards further details after months +Personally I dont think we should have +If I understand the request correctly then this is not a comment if you think otherwise +In any case though I agree we can make the protection more Tapestry aware +The key is that operators will be pluggable in the full execution engine and records will native buffer based for easy passing between java and cpp +Thanks Deva patch looks goodThomas do you have any comments +As you stated since this property is unlikely to be ever used I see no reason to keep it +Performs simple operations on an int creating bookmarks and allowing the user to go back to those bookmarks +Closing. +I would really like some comments +this looks good +I overlooked the option to diff in the attached patch with the loopback transport has been committed to the CVS repository +cloned for this still a problem with the latest releaseArent we dependent on Maven SCM for this support? If so we should close this and open specific issues there. +That is awesome thanksEAN is the barcode norm does it really translate to JAN? Can you check if the algorithm match the Japanese barcode numbers +If going back to plexus is the easiest way to get the existing database migration working again its fine with has been talk of switching to a different way of migrating the database +Please provide simple test case to reproduce the issueIt seems to be packaging issue to me +Committed as r on trunk and r on +Not a blocker for will go into backlog for +Bumping to since this is a pretty obscure failure and the amount of work to fix it sort of sucks +adding attachments +Submitting to the patch since the previous one wasnt generated from patch since it was not picked up by hudson +I added some more checks around the code that was throwing the NPE +Thank you for reporting this and for the patch +Tony +Erroring out in this case means notifying the so that we can release the operation +Included as part of +You should be able to followthe example of the Metaphone unit tests +I just committed a new class loader that should eliminate the file locks created by the class loader +Maybe later and even then chances are it might be easy enough to retrofit +The goal of the patch was to enable the decoupling of hibernate from transaction management without having to declare the hibernate interceptor around all the beans that used it in the config file +However if I try to change it in the UI it wont be saved when using WTP will check this today but think the easiest way is to skip creating EAP when starting JBDS with JDK way would be to add a dialog that would enable the user to add JDK +Can we remove all the isassignablefrom stuffwe should instead use proper fieldtype methods +Created assembly module that creates a binary assembly similar to the one created previously inside the moduleThe source assembly collects together the sources of and assembly modules +Divisions are now calculated in floating point sizes and rounded to integer on creation of subregions +I have tested update of JRE and JDK in bits of version +But there is linked from any other place +r +The patch details the issue that can NOT be worked around. +I think thats what I was missing if you come from smallest you have to map on the ord switch +We could add an order by clause on the creation time but that might not suit everyone because it would require an extra index +You maybe didnt notice that the piece of code you copied is inside the Attributes iteration +It was released with the fix in +Isnt there an M release now? Can you try it with that +Things work if I comment out the compression codecs in but than Im having to maintain a separate copy for oozie +I have tried locally with Apache Tomcat and and it works fineMaybe it is problem with CI environment +The problem with this is if the fields is only indexed without vector offsets and not stored +So far weve punted this to some magic that will let us load both clients into the same VM +Ive written a plugin to generate batch files to execute these admin tools +I have not done any testing on max row size but cols rows of bytes total is much faster with the filter +Perhaps it is seeing the driver does not and is simply reverting to JDBC level support +Added a sync test case to show that both paths suffer the same problem and also to verify the async case was suffering timing issues +integrated on master +think it is fixed on it wasnt released yet? because I have with version in my you can mark this issue as duplicated +Stan can you have a look at this? I think it is much more likely to be related to the JSF upgrade rather than Undertow +where do you get that mvn package is more use in javaee land? isnt mvn install sufficient +Could you try again on your side +Persisted and transacted messages size bytes Kill C Cluster failed over to A Restarted C exception appears in C log during. +I will spawn another ticket for similar issues I have already spotted. +Test and patch attached +Bryan thank you for those suggestions about the MAX and MIN pages and Tomohito thanks for renaming this issue that is the right thing to doShould I change the language in MIN about blank spaces to match the language in MAX +Marking this as completed. +Felix Meschberger Yes the Jackrabbit Server bundle embeds Jackrabbit Core +makes name was not happening by accident this follows the naming pattern of query column functions query query query query +And sorry for coming late +On that thread pool executer at the and we have OME +clean v attached against trunk +rd time lucky line D is an issue with the way smooks is loading classes has the patches which fix this issue and will apply them to the smooks should have some new smooks jars for ESB +How about creating a patch and appending it to this bug? This way maybe more people could give the new code a test driveConcerning the points and +Michelle +Patch was made one level too deep +Before svn load please remove the current incubatorgiraphtrunkThank you in advance for your help +Our goal is to get the information to the operator so they can diagnose and fix the problem and seal the problem off from the clusterThis is distinct from the case that the node is simply full +I only tested the filter part of the patch new normalizations are not yet tested and I noticed at least one bug in the patch for normalizations +Its not solution the problem come from here portal java script first +Doesnt this use case violate the spec? A getService call cannot return an instance before its activate method has returned +I have also started derbyall run with Sun JDK and so far no failures +intRange I believe is only part of itI think they want to use the notation MAXINT MININTfor range values +By default the limit should be hits +i placed the hashtable which is used for in proxyhttp and but i dont think this is a good way +When this comes around let us know if we can help prepare for the upgrade +Rebased to trunk +We should try to examine this before if not fix it +This is reasonable +If you believe the issue is not resolved please reopen or open a new issue +It only seems to happen with zip files permissions are preserved with and +It would be nice to have a configurable thread property for MapReduce tasks too +can I take a look? whats the URL +fixed findBugs errors +Like i said I disagree +Ok I might need to enlist your help with figuring this out +Encapsulate proxy details into a class as was done in just submitted a patch to the mailing list +I was not able to reproduce this + +Patch committed to SVN trunkOleg +This is a general issue of static initialization +Ive noticed this started to happen after Ive installed soapUI but it couldve been a coincidence since Ive tried a brand new installation of Eclipse with GRE plugin no soapUI and pointing to a new workspace +since is created during JIRA encounters problem +It crashes after new session callsRandom number provider should beI proved this with a simple java programAxis uses random number generation to create session idThe random number generator crashes after callsConsider it resolved. +Frederic thanks for the latest info +Rather than clean up every row inserted the focus of this issue should be eliminating all schema manipulation in the tests +Then we can use suggest you open a new can a malicious user lock pagespagemaps so create those kind of errorsThese errors are more or less programmingweb application errors that you need to fix +I think it should be more general approach not only the related +Im assigning to you Ive added all the ones you listed except the intellij files +Ive replaced reRender attribute by rendered attribute in section of the Reference +However Id claim that if the file fails there shouldnt be any way repeated reads should cause problems +Duplicate of +I think this is already covered in the tests Andy added in +Marcos I had to still add the in commonresourcesdependsjars to the build classpath this jar doesnt currently get built into the hdk so I didnt add it as a ve just done another test run and I can see the reason why they appeared to pass successfully the test target prints BUILD SUCCESSFUL after running the jdktools tests whatever the outcome! Not sure what is causing that to happen when I get some time to look at it Ill see if I can fix it +Great work. +So anytime I make a change in Eclipse project the classes are recompiled and Karaf updates the bundle +Next add verification of migration +After minutes of inserting the will create a for the first +when the archetype is created no dependent artifacts are resolveddo you mean the checkbox should trigger a parameter which in turn is used by our archetypewe could also just have this as default in the created yeah +For you then buddy +Have tried with the latest snapshot today and are unable to reproduce with the quick start +Suggested patch introducing timeouts and seeddetails reporting +Seems like somebody has already improved the page but not regenerated the web page +Blueprint logs during container startup showing creation of the duplicate service instance +This is a candidate but not committed for EAP YannickIt works for me tooWhere did you see this trick? On forumsI hope someone put this instructions in the Official soonMany Thanks +Not reproduced in new versions of +Yes I tend to agree +In MR at least let the task fail earlier and have it retried +See +The correct default value is +in fact the modcluster directives arent in the files +Thanks Steve +All datanodes are bad +Its really up to Stan to decide since JSF is still in this point the updates for this release are more or less done so unless a new version can go in today or tomorrow the latest this will be deferred for a future release +Use provider +The directions on the wiki are outdated +defer all issues to +It runs the plugins in the reporting section but is not a report plugin itself +Currently the size is to or +Now that work is underway with AS all previous community releases are +Was committed to not. +This also cleans up the code a bit since having undoable actions handle enablement proves to be very error prone +The patch removes the dash and adds a test to verify that it still works +Marking this as wont fix per the discussions +probably worth measuring but not a lotof different cases if the design does not expect a research +I corrected the error message to report the application that could not get imported as maven project since we have no maven project name yet +bq +patch to add to the search patch for submitted +for a test to catch this +This is the initial version of the code refactoring to remove HTML snippets +Hi! I have to disagree the operation should be consistentAnd more importantly this is a regression as in the past it was consistent +I wouldnt want to expose the timestamp field to users as we are using it internally for dealing with concurrent writes and reads on a table +all current RESOLUTION LATER issues to be against the Nightly Build +Putting something in the config doesnt sound like a good idea +This JIRA item marks the record of the optionaleasily removed and clearly marked mysql dependency being fine. +Perhaps the TT can request one more task than it has slots in the hope that it can be assigned when the next task finishes? This is all about collation after map tasks +the new release is awesome +Its animmutable collection of files of a specific date +If we do we can add some logic there to allow overriding it though +The patch to fix errors in the document Built In Functions in trunk +Attached in case people want to check it out +class file from Groovy theres no import for slfj +Theyve been a source of confusion and frustration and I imagine our users wont miss them either +Alright got my changes pushed to Teds multi branch +These are the patches for trunk to add this +Closing all tickets +CP should be fixed in revision should be fixed in revision +TomazWe dont currently have a QA set up to test messaging against it possible for you to checkout the current Branch build and verify that the changes were applied correctly to work against DBThanks for long replay +you can declare a variable in an imported stylesheet andreference it in the stylesheet +I think the plugin will be a very good until plugin infrastructure is in place +So there is no installation process? Is it similar to tomcat install? Is there a way to have it up at all times without having on hanging through terminal +My bad to have missed the annotationUploading a patch with the annotation +Thanks for the review +We were working with JSF pages and when we tried to open an specific xhtml it occurred +Thats been on my mind since you mentioned it earlier Robert +New version of patch that includes tests + This bug has been marked as a duplicate of +It was almost a year ago so perhaps you forgot +There should be some way to default the column names +Here is the patch for the sameRegardsAmit in r +Well it is implemented for empty property sets You are right regarding supporting arbitrary properties but my impression after reading the is that this is not the way this mechanism is intended to be used +The new patch takes care of your first comments +I will keep your tip in mind about comments +I was going to try with RC +Heres my currently biggest problem the SQL statement we use for loading aclass only loads fields +Probably due to conflicts with libraries installed on Ubuntu. +Done like dinner. +Damian first thanks for the improvements your raised +I also fixed an occurence of START WITH in stead of START WITH +Also thanks for that tip I saw that expression several times in the existing implementation so thought its ok to use it +Sorry for not looking +So I will close this JIRAIf you want to use this featurejust try geronimo +patch was committed. +Hi VladimirI have decided to drop the class for now since it does not fit in with the rest of the layouts which return strings +Im not sure what is the cleanest way to improve those tests +Sebastien +MartinThe patch looks good I dont see any reason to not integrate it. +Thanks for the patch Roman +This does not comply with a unified and standardizedor provides two logic for users to me there is no difference in IME appearance +I have committed this +Actually the new name isnt propagated to windows title either +might work in tests but not in the real world +I have the snapshot repos defined in my + change isnt needed +I went to commit this to the branch but it looks like is already passing on a fresh checkout without this patch +I dont see this issue any more +You have valid points above and we do intend to revise the code accordingly +Thanks for the contribution Arpit +After applying this patch I still see an extra on linux after deployjdkjrebindefaultMark +That would mean changing getConf to get +If you are happy to leave the default limit at then this bug can be closed +Please provide more details about what you see there since it is no longer similar to the design +Quick correction +Rob couple of questions If this is for version why is it called version in the descriptor? is the ui updated so if a user edits version files in version will he be asked if he want it converted +updated the parser to replace with and with so that the input string is a valid embedded java string. +File is checked in by Andrew in trunk rev +We should fix the ICE +By the way I forgot to say that I had the same problem with Eclipse +I guess notThis is unrelated to mmap etc +Told to ignore it +Can you elaborate on what you mean by some collections were being fetched as proxies with missing session context leading to ? Is there something I can do to reproduce this +batch transition resolvedfixed bugs to closedfixed +v logs the exception to server log via logj as well as sending it to the misread +Looks like hudson is not able to run findbugs may be because of dependency of the tar files inBut I ran findbugs on my machine and there are no new findbug warnings introduced +Todd is it okay for me to commit the patch? Im going to commit it to both trunk and +Having it on the face of the web page means that i cannot download a copyThe extra stuff in the diff that you mention is probably due to your XML editor making whitespace changes when it saves the file +Yes I do have that code +I just committed this + +Thanks Adrianyour patch is in rev +unit test showing the issue +Just that both make senseIts been almost months now with a lot of words exchangedI personally still dont understand what is the resistance based on and dont see why we should not move this forwardLets have some action here +Havent lost any blocks yet but had some close callsNot sure where to mention it but it exposes a limitation of the when down replication happens clear blocks from the fullest node first working against the try to distribute replications across racks +Fixing the load balancing strategy for the option on the Producer +Right I meant evaluating several groups in the same traversal +An important part of the recipe to reproduce is artificial DFS and master CPU loading +Im not attaching a patch +Spaces are now sorted alphabetically +Problem doesnt occur when using Java +Im not sure the deals with tracking while this one synchronizes changes in services delivering +Above I explain why I think its rare it will be queries onlyDoes load hdfshbase work out of interest? Its not important if it doesnt +This issue must have been resolved with +If you really need flexible urls you could consider using a proxy like apache or nginx which gives you all the power of rewriting urls even better than by utilizing s restricted possibilties +On entering of one of them a seam managed action handler is called +Super all tests pass for me too +Interesting the tests didnt catch this? How did you run into it +This one is about the proto files under srcprotobuf which should be easier +Reopening it to get it fixed in +Just to make thinhs clear I wrote linker still cygwins link command is not linker but it is Call the link function to create a link named FILE to an existing FILEsorry for confusionAlso I wrote The known workaround is to remove cygwin linker from PATH +Maybe is just an impression of mine +This has been placed on the backburner for me at the moment sorry +bq +Bulk closing stale resolved issues +You mentioned that you were working on a revision that added more comments and cleaned out some debugging code do you have an updated patch with that or would you like me to handle it +Thanks for the thoughts Yonik +Has your fix been made available somewhere? Id love to test itWhen trying to build the project with the enable which does not do anything useful it must be said I get for +Our network is fairly low bandwidthhighlatency so we were wondering if there is a way or one could be addedto adjust the inactivity timeout value +But rather than how about the sharing a singleton map +Patch includes a test for this issue +Also on Windows Im not sure which tmp to setShall we create tmp directory in tasks working directory and make +But it seemed like it could lead to confusion about where matching artifacts are +Not sure for the colorsD but great initiative yes +Didnt know about rm +I can implement my own hook and configure the yaml to use it +Ill commit the patch +Thanks Chaoyu +Changing the type to reflect this. +I will file a jira on that. +Clean attribute packagespackagetodirPlease this to me when you answer +It creates more problems than solves in its current state +This issue is the type of issue registered handlers will potentially run into +Fixed typo in Jira comment too +Changes hae appeared in Latest Alpha Manuals so closing. +bulk defer minor issues to +This patch fixes the issue for me with swiftMQ +Only people asking for access get karma to the later and I just considered this entry such a request Go ahead Bill +Setting an exclude on the dependencys could work but that dependency is managed by Me and Id like to avoid hacking more than strictly needed +keep Wed Sep UTC webappsjobhistory +The only issue I can see is that a server may have a different notion of quorums but I think this can happen even if we enforce that we set the only during initialization +It doesnt make sense to do something more heavyweight since we already have that +Vidhya +cQ +all issues to the CR +So we only remove deleted columns if there is a CF level tombstone to ensure the delete makes it into an +Yeah adding encoding would be great +bulk defer of open issues to +patchFeel free to change the class name method signature etc +I switched to Accumulo and everything runs fineI would still like to know why cloudbase hangs on setting the instance name though +Im running regressions with Knuts patch and a patched +This is an issue for me least the build shouldnt fail but instead cache the values +Rolling over to reduce workload for pre is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +I believe we can add that without breaking compatibility which allows us to move forward with this Jira +I think I fixed this in trunk by making the corba ref builder look around to see if there are css bean or tss bean refs or if the orb is present before adding the corba ref +maybe the unit tests should set the Xss parameter in order to be as close as possible of a running cassandra instanceNot a bad idea +If you are using then the transaction is if you havent change the value of to false +agenda and seminar PP presentation attachment +NOTE The file is indeed tried it on Fedora and Vista system using tools and Jboss tools it works well +I didnt applied Christians Patch but used his idea +The logic in question in already does catch and raise it as EINPROGRESS. +You convinced me but dont ask for in the next three months +Also worth noting is that the build now depends on libtool +Verified that usage in Beta is reporting correctly +Help would be appreciated on this to see how stack trace elision is performed see and last +I think it is better to use the Richfaces pagination component. +Security OverviewModified books staged and sources committed to SVN repos +Even so its not important to ask this question +that way Im stuck until its committed and Ill have to start messing with patches afterwards Ill submit all at once then when its completelyIve +Ive tried to introduce sleeps in between even with s pauses between them I still see the same in the logs +Marking as open while Sharads comment is addressed +Ive modified it to first produce all messages with a correlationid and then try to receive the same number of messages and checking the received idsAlso added copying the correlationid in the handler +This would also allow to handle is only question to me is if this is something so basic it should be handled by the normal parser similar to the regular meta tag +Update the patch +Good to know that compilers nowadays warn on this as well +Marc unless you can provide a way to reproduce I will be resolving this issue +Uploaded patch v added a prefix as additional salt +Is the Administration chapter of the user manual the appropriate place? Maybe right after Monitoring +Thanks +works as desire this time +I led you down the wrong path with using uptodate sorry about that this patch also includes changes to generate the property file so it doesnt get regenerated on a recompilePlease review the patch +Would you like us to make an official release with this fix +I applied the patch and made the change you suggested but I get tons of spec failures in +Shouldnt be more than a few lines of code to do that +Hamlet since you are the expert here what do you think of this +PR Reopened +Seems like a minimal impact sort of change as this api appears to be a little protected +Fixes a hang during shutdown caused by invalid AE descriptor +See +Couldnt we build a list when readingwriting of the the first time so that after new write is finished we just mmap through the page cache and we already know the new locations on the first pass +though we may hit file descriptor limits in some tests +Fix for this has been committed +Working directly with Amit on all these issues +SNAPSHOT +Im going to apply first then and take the test for a drive +If its broken will back out this patch in and cut a new release candidate +reverted +Ah yes it looks like you were passing a size to the constructor of the transport when what it actually takes is an initial buffer value +It seems very doubtful to me that the type checks could be added for everything +Trivial but helpful +This looks like debug mode to me +However the exact same thing happens whan i install subclipse so i suspect that it has something to do with a common dependency that is installed either by Subclipse or by LDAP Browser +Thanks Aniket I just realize the patch try to reuse splitIndex +Thanks +Three classes are added in +Another new approach +however is separate from this issueLets nail down these then see whats left and address those in subsequentjiras +Thanks. +Help in resolution is appreciated +It seems logical to have the same tool generate Java classes and their mapping to XML the mapping file going to be optional? If so lets first try wo the client will use the same tool to generate Java classes serializersdeserializers mapping file as the one used on the server sideIt seems to be a weakness to me since how can you be sure the client has this same tool? Assuming there is no standard for these tools output +DoneThanks Mike. +Added asserting code +I managed to solve the problem by including the various projects sources jars using the classifier sources and scope test +Hi sorry about not specifying which documents +Commit +Updated patch to reflect changes to trunk +Sorry closed this prematurely +This needs a release of apache xbean the new dependent feature is currently only available in a but we cant release with a snapshot dependency +The next patch looks good +As for the private conversations not all of us are privy to them so how does the PMC shut them down? Besides people who live in glass houses shouldnt throw stones +Marking as duplicate. +Closing +The code for this is in the trunk under toolseclipse but it isnt in the build and has never been released and its not clear what state its inFor I think its too much work and too risky to try to get the toolseclipse version working and tested and switch over to using it +A patch attached +postpone needs changes in maven embedder binaries +Patch committed +verify through some hdfs call the size is correct +the fix is simply to verify that the figures for this cache service reflect the correct stats +BTW I think we should use this ticket as a container for for each change +There was some confusion on this issue due to being a bit vague so were waiting on fixing helloworlddbregistration until tests are working in CI and Kevin gives the okay +Thanks +Additional mapping files can be created to map to other tablescolums defined by the adminstratorTks Eric + has been marked as a duplicate of this bug +A patch file that applies my fix to the code and creates a junit test to test it +Might take me a few days though +Can you upgrade your jdk +Guys I certainly not excited with the selected approach it is not very flexibleIn fact the problem of crash detection is global enough to try finding common solution which would fit all existing harmony suites classlib unit tests drlvm kernel jvmti reg etcPlease take a look on alternative solution at it is being actively discussed on dev list +The proposed changes dont really match with the current code base +by passing valid arguments performing necessary initializations and so on unchecked if the problem is outside of the callers control such as network connectivity file system errors and so onBased on this policy throwing a for bad parametersis appropriate +Hi DeepakIm not sure I understand the benefit what good is it to render the doctype here when so much more is required of an html document such as html head and body tags +Its hard to just up the handlers above +WernerI hope that you have not confused my about the other problem that I was facing which has been resolved with +Other hints are appreciated +No more no less +In a super CF the parameter could also be CFsuperColumn +Hi Igor could you send me the patch please +Committed revision +Weve talked with Tony about the implications of migration and think we have a handle on dealing with those issues +Sounds very interesting please feel free to go ahead and put together some code in a patch +The rest were assigned out a few minutes later +bq +Only downside I see is that round was not synced before but now it is +Please save the connection timeout in a final member varilable in Client instead of getting it from the configuration every timeOther comments final modifier for static methods seems unnecessary +This change corrects the code to prevent an NPE +Ensure there is just version jar files in yourclasspathRussell +This type is present in the tests array ILLEGALJDBCTYPES but not in the code used for checking check and on the client similarly in check +Makes sense +should have been this one that includes init method +So Ill do a round on this patchbq +The unneeded JDBM reference has been taken out of NOTICE as well as removing the incorrect uses text +I have provided my configuration and test results +Is there a new release pending to try this out or is it a question of being a bit more patient +We dont define a fix version until a fix is committed +Is it straight forward to fix +Fixed with revision +The value of the expiration time is calculated as follows time of request expiration minus current want such a solution because the request sender determines the time the sending system is waiting on the response message +Add a annotation to aspects module +Im still curious whether custom interceptors can be attached to a passivation capable bean and what will happen in this scenario if they are +Verified token renewal requests on a secure cluster +I didnt know about the issue seen in there a reference for putting the ipv address in square brackets apart from my comment aboveJust wondered if some research had gone in this or you just trusted my brief commentMy searching was very quick early morning and before coffee +Closing issues that are still marked as resolved but were part of the release. +has changes to the file and is worked around in the jsfunit on cargo changesrequirements also +is a separate issue for trunk +some more information about your hibernate console configuraiton would also be very beneficial so i can reproduce it +Thanks to Jakob Korherr for provide this +The build is designed to allow There is lib and lib and all binaries live in bin and sbinWe can have that for but too demanding easiest thing is This is because user can choose either or bit mode on bit machineand only if using bit kernel on Solaris and always on Solaris This is why we cannot choose the because we dont know weatheruser wishes to run in or bit mode +settings file is created butThe remote does not seem to be used as the key to allow for multiple ones as discussed previouslyThe deployfolder is stored in the project shouldnt this be on the server +Operations teams dont want to specify the memory on each machine +Moving to Open to Community as I cant commit theyll be resources to do this for +Would love to get resolved for not sure who to do this Im not familiar of the changes if not I would have done it +Patch to verify weve met the slowstart criteria before unleashing all reducers once all maps have been assignedBesides the automated tests that were addedmodified I also ran manual tests on a local cluster with sleep and wordcount jobs +Hi Mamta If we have to check the error string I think DRDA is a better string to search for as it wont change with localization +HiSorry for very late reply +Im working on a patch +I removed the forcing of forkMode when I closed this and I tested the use of forkMode at least somewhat +refers to a JDK class bootstrapped by AS +Yeah sure go ahead +Fixed post. +Committed. +To solve this I added a flag call enablesecurity to axisconf +Definitely fix then general task here should be to examine Math Fraction for changes that need to be applied to Lang Fraction. +This is valid only for as it is fixed on branches and abovePlease close this as WONT FIX if there is no intention to provide a patch for branch. +OK this patch switches over all uses of the old names to the new onesI plan to commit in a day or two +Can you code it up +This was not because of class loading +Fixedwbuildrsvn ci CHANGELOG libbuildrjava m Setting a system property in the buildfile causes a Sending CHANGELOGSending libbuildrjavaTransmitting file dataCommitted revision . +Thanks for the great work +They should defer to the ongoing scans +I downgrade Felix to to be JDK compliant. +However even we just read the partial history data of application for example the application information we still need to load the complete file +That seems unfortunate +Diff of suggested changes +Closing as WONFIX. +Can you describe the worst case scenario you imagine will happen if is exposedSure someone who thinks they know what they are doing closes the IW when it shouldnt be closed causing exceptions etc and emails to etc +I wont admit my lack of git voodoo powers publicly but due to the new global cache design in I have to port rather than rebase anyway +Thanks SianPatch applied at repo revision rPlease check that it was applied as you expected +Thanks Edward +Introduced new methods in the instead +Shall I prepare a patch for this +I hope to be able to remove them all by the time we release +Closing all old Resolved issues +attached the latest patch available on review board +Konstantin IlyaPlease provide some workarounds if there are any +Just curious what are the consideration to not include them as the other frameworks? Size? I made the change locally and the difference is kbI always had the impression that I can rely on the framework implementation to export all the packages defined in the core specsAnyhow is not an hard issue +See for reasons we will defer providing such integration. +However youve reformatted the with tabs +This issue was resolved by. +Can you not leave the component of the RI as it is and only change the behaviour of the component in tomahawk? This way we still can use the requiredtrue attribute on tselect tags +Im updating the priority and component to better reflect the current status. +This is the error message as it appears to an Outlook user +bulk defer of issues to +In other words it is the applications responsibility to ensure that at most one thread is using a object at any given moment +So all issues not directly assigned to an AS release are being closed +The memory model after this patch is very complicated it all depends on what operations are performed in the second MR job +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +This patch updates all filesOur continuous integration environment cannot handle this patch yet due to changes in the protocol from the recently committed +I believe Charlie did remove most of this last week +Not against it just need to know the use caseRegardless of that can we leave that as a future improvement +As far as Ive observed its enough to have the CA certificate present in servers keystore +This patch is for hadoop +See +Also strips the single quotes as they make cut and paste more just committed this +Committed revision with unit test +Unfortunately I am currently under agreement with my employer that I can only commit my own original code no code +Ill have a look into that +Bulk update to fix version +bq +Perhaps it would make more sense to put it in the log service subproject as an option command similar to the obr command in the bundle repositoryThe downside I suppose is that it cannot be used independently with out log service implementations + move chown operation from post install into rpm package phase +I might be able to work in this in the future but probably not in the short term +Just allowing Hudson to validate the patch with current trunk +But why a custom mapper? I was thinking along the lines of letting user set a standard named property in and the presence of it would trigger the status to be set +Its not entirely the fact that creating a per request is too heavy +I just uploaded the whole file because svn diff claimed Cannot display file marked as a binary type +Committed revision for this patch +must be fixed for final +So congrats to all on that +lets see if we can get this resolved for the next release +On procedure that returns a resutset I used similar logic to return collections of Complex Objects +just verified the code +servers start is very descriptive net start not so muchIm curious to hear if there are any other suggestions +Thanks Vandana +Thanks Will! +Part of the value of using scaffolded views is you get a bunch of default behavior in the view without having to write it and one of the defaults in the scaffolded views is the layout +The primary intent of this issue is to test that Avro reflect can correctly characterize all of the Namenodes protocols +But we could implement the minmax in the a way where we put hadoop sorting to some use +Patch integrated to trunk and. +Otherwise the default adapter is used +Alternative patch that logs an info message when no password is set and attempts to continue +right this is duplicate. +A couple of comments +Please review the proposed change and either fix or reassign back to me +Sorry I use JDK and not +Ive checked in an integration test for forkModealways in revision . +Cool ill see if the librato folks want it if not ill publish something to maven central once Kafka is there +Are the images css necessary +Ill submit a patch in the next couple of days for the component +theres some inconsistency between Usage and How to Use between the plugins that have been submitted recently +Redirect handling has undergone significant changes since this issue was opened and we still need to take a hard look at redirects and possibly how scores are represented +Thanks for following up Uma +Thanks Doug. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +patch to add default file location of license file as per m leaving issue open for solution to the new feature +Thanks Brock +Fixed on and closing all issues that have been in resolved state for more than a month without further comments +Fixed in revision +I do remember an issue with fillIcon some time ago + demonstrates a problem +Test reproduces error and patch fixes it +Attaching a version of the patch with a testRuss please start writing testsAll piggybank tests passCommitting to and trunk +Toby any update on this +I havent been able to get to Marks machine and am unlikely to do so to +It seems that this option is only useful for people who are using Maven resource paths since it provides no benefit for people using thus it is ultimately encouraging people to use Maven resource pathsAnother way to look at this is that you want to change how the default value of is calculated +Your analysis is correct attached is slightly revised patch to emphasize that we already read exactly one Fabien +modalPanel works fine without dom attribute using +k +IMHO it ought to be logged at INFO but WARN would be understandable too +Tests addedCommitted revision Committed revision . +Some minor updates bundle is in the wrong startlevel Shell bundles should optionally be added to the Rev +You can have a look at meclipse and qe which handles it. +bq +Test attached +I think what we need from the ESB project is toolsconsolemanagementtoolsconsoletoolsconsole these are laid down we need to copy the JSP files from the project into +Since the input stream has aleady been consumed by that earlier process it is no longer available for parsing by CommonsAm i required to write my own handler to get im very new to this strutsCan you tell me is it will solve my problem +Ondra anything marked testsuite is getting automatically assigned to you +Ive just committed this to trunk andThanks a lot for the contribution Colin. +This is quite likely the same problem. +However this brings up the deeper question of how do we prevent segments with identical starting offset from being created during log roll? Maybe we should add a check in to guard this +Its definitely not in Seam code its in the JAAS +HiNo only cobertura affected. +Added connection option on the to let user control the connection live time +Bug can be fixed in the CVS when it is first fixed in the Excalibur were is gone +Thanks for the explanation +To summarize the patchAdds a Xmxm and XXm as a default to bin bin and etc +Also dont forget the servlet component as it extends the HTTP all resolved tickets from or older +I think that our hudson instance my might some lag it didnt peeked up changes right away +issue fixed +if +Requires moving Commons Chain to a mimimum of JDK s additional overhead to using is more efficient especially if youre creating the Map on every request +This is not a CR release +agreed +Please reopen if you have any problems using the provided paths to the did not report about this because we have another issue with cycle dependencies building the javadocs there +appears twice in heap size test and removes the advanced tunables from had a look at the patch it is looking good + + thanks. +Resolved +anybody working on this or planning to in the near future +In general we try to leave stdio streams alone since s shutdown may or may not be the last thing to happen with those streamsIm not sure why THREADSAFE makes a difference +I uploaded a log file and new service configurations to illustrate the problem +cost of the additional checks in are neglectible as they only occur when a competitive hit is really inserted into the queueThis should be benchmarked for and too as they also use +Tran Le Xuan this issue is closed as duplicate so please add the patch to if that appliesCheers +I believe the buffer may appear from memory mapped implementation so a proper fix of may fix as well. +To set the priority of order manually go to order view page and choose the priority from drop down menu and click Reserve Inventory buttonNote By default the priority is set to normal while creating sales order +If you choose to use a custom implementation which delegates to multiple handlers then you should ensure that they behave would regard it as an error to invoke code which calls redirect twice on the same response and I would prefer to have an exception raised +FYI enable to use the for the tip +How far would you go with this tools consolidation? All tools across the whole of Scenolunr? Or just the ones under modulesanalysisI just meant under the analyzers module +Was that intentional? I thought you were trying to pass it to Liang +Closing issues in +oops correcting slight error in previous like this was fixed in +Can you please provide more information +lgtm as well +I thought there should be mp but it is not +Doesnt change the default behavior except for the unlikely case where people were setting JAVALIBRARYPATH and it was getting stomped on and when its no longer stomped on theyll break +If no longer relevant please close +See. +Fixed +thats a mvn task inside ant though +Can you write the pom and counsel on where we should put them when we release? Thanks +Thanks +By default custom password types are rejected +Actually I just tested and reviewed the patch and here are some issues I have noticed Tests fail with new changes New extension methods are missing docstringsHappy to merge it as soon as those issues are addresses +In http procotol we verify receiving remote response only by or chunk +Verified in CR. +Sometimes its necessary to shake things up with new ideas +In that case we committed the jar +I also tried to remove the ear from and add it again with no effect +Thats reasonable for broker configs +Fair enough but please open a separate issue unless its guaranteed this will be done and integrated within say week +This is tracked as a different bugClosing this as package upgrade works. +Can you please tell me how to resolve this issue +Closed after release. +Verified on revision +The java class and its form one whole compilation unit +This sounds correct to me as GWT requires JDKCan you please check if you still experience this with the latest version +Peter could you commit it please +Does the maven jar plugin depends on the same plexus library as the maven +What nkeywal said above +Fixed in release +I had granted the license but I forgot to add the file header. +Linking to where the code was introduced +What does the int returned by is represent? Please see javadocs +or we could put it in a data structure apart from serverloadDo we have a route through which such data structure can be channeled to master nowadaysbq +So the pattern used in this test class was not working +I hadnt had time to review your case yet but that would indeed cause some problems +So I think maybe we can close it now. +Perhaps this would be one of those casesIn any event why doesnt the interrupt method on the thread work for you? As long as youre making the connection to the database on an interruptable channel the blocking connection should be forced closed and return with an appropriate exception. + +bq +release bulk close +Of course well blog about the changes so that everybody know they are there and go look for nightly better wait for the next release and let us others test it thoroughly +So yes I think having a copy of the Sybase entry should be enough +Given the exact same versions of the software I cannot reproduce this issueI think you are picking up some different version of junit probably the one that is packaged with the spring framework instead of the stock junitjunit version +log file maybe or the console +I agree using standard way to store file origin is better +While the apis are public in technical terms they consider it friend meaning they can change whatever they want +Dave GC maps sound like a distraction +Revisit Jira assignments +So it may be better to add the specific info to the Wiki +My issue centered on the use of and such for anonymous inner classes where binsh was interpreting this as an argument to the present script +Tested and working +But if the pool has a different nameuuid and is using the same path it doesnt reuse it and fails to create the pool we need +May be for a begining a simple patch would be enough +to commit and remove this part of the trap +And if we find something to adapt at your patch dont worry either +Hi JoeIll apply the fix to master then close it +Cool thanks for the explanation niels +Ill let you know when its committed +Patches are highly appreciated +I dont know what it was but maybe the last update +So for example the security configuration in the vote sample application is copied there properlyjar tf target grep classesclassesclassesmavenvote your mvn package command execute correctlyWe do not package the application contexts in srcmainwebapp because not all applications produced by Roo are necessarily +Fixed in forms block +and finally came to youId greatly appreciate if you can spare some time and look DEEP into the problembugThanks. +Thanks Raghu I have attached another patch which fixes +test now pass +Further work will be to remove and +fixed in SVN +So in this case we are actually going in the opposite direction of what we have been trying to do which is to rely less on the messages and more on the exception hierarchy itself to convey context info +Changing type to bug fix version to fix seems to work just as it did inAs to your suggestion I created to express the need for handling the encoding of XML files +Manually tested! +So I dont want this one to linger too much longer is it cool can we commit it? Its probably up to Bryan since hell need phpize installed when creating distros +You dont need to change the PATH +Bulk move of M to m converting all that is not using resources as it explained in description then well see what needs to be done for particular is does that worth to commit something for M or do it after branch for m is created +retargetting for after in both and M. +I believe the same is true in AS and as wellThe only difference in AS is we write out the +t removed it because the report generates always on our code +Then it reads version from manifest and it is +Is that correct? If so I like that simplification +Merge with the branch +Minor correction in release notes went when +Yeah looks like that one caused them a little consternation +I can run on devices and simulators of course +It also includes a unit test to show the problem +Oh ok +I committed this +Leaving open till see it passing +the patch for Jenkins +Thanks +Pluggable interfaceclass should be public with the annotation +any ideas? rightLets fix that. +Have had the same problem with version +Resolved by commit cfcbcfcbadfeacbeaca on masterand commit aeafefaffbacbddf on this record since is now released. +Thanks for the patch Karthick. +It creates less api churn also is more amenable to Hadoop Pipes and such others +I already had to heck the schema files in order to get them to generate properly due to the mass circular references and the like +Still should probably be fixed there as well just in case +However I think the current interfaces to the scheduler which send it events on NM and AM heartbeats are the correct ones +We confirmed that logs were consistent correctly +property renamed to network server control bean implemented by this JIRA adds another way to shutdown a server +I think wed better rewrite this with other framework than dojo for +Hello PeterI have a few detailed questions regarding your componentIs there any way to get in touch with you directlyKind regardsAlexander +Committed as part of fixed +I am closing this now since the icefaces is already removed +Its fine to do some incompatible changes before application and API is frozen for branchBut actually I dont care how int is called in CQL as long it is included in Cassy release +Stop everything I get a collaps! Again wrong patch +Not sure how that file is in there +Do you know any way to reproduce the issue +or else I will take over in the evening +Hence our use of one builder and many disposable clients +defer all issues to. +Can you reproduce this with final? Does it happen for every project you try to release or is it only one +the hsqldb issue is caused by the G table not having a primary key +This will ensure that the value get correctly reset at the end of the test +We implemented it in the AM tracking resource requests made for a specific host with relaxLocalityfalse and then if they are not filled by the scheduler after n heartbeats dropping host constraint and switching to relaxLocalitytrue +Missed +Please can someone look into this new year party is almost over here +The given patch has been applied to the CVS. +But I have problems reproducing the issue +This patch modifies the so that it does not create the extra closing tag in its markup +Same issue as described in parent JIRA and as further described in +That temporary dump file is an so we should be good with as long as the same configs apply there +I noticed that dates get this format is string but cant remember if that is something I added or not +Added to SVN +I assume you mean on the public website +Dave what RDBMS are you using to execute aboce code +The config changes lowered memory from G to G and walog size from G to M +As for the network of brokers you should see on the remote broker connection like that with revision +bulk defer to +Patch doing the suggested incorporating offline comments from Sharad +No it is fine like this +If you look at my screen grab doesnt the new sizes look a little small for a tablet? +Looking in job tracker here are the properties in the configuration for the job which failed starting with crunch +Assume closed as resolved and released +If it works i will close this issue out +Hi SubhoPlease redirect your question to the user lists + was the original blame command jira which got implemented in a fix +resolved in +There is a lot of noise on this subject and I believe Dan is correct in that adding support will increase the adoption rate of Seam. +Even if downstream project uses it seems that it does not activatedeactivate the profile in hbase pom which is a dependency +Fix include r +Changes are in SVN +Hello MiceI saw the bug on a VM with one ROOT and two DATA disks +Updated to the latest trunk +I think this has been resolved with the of makefiles not to use convenience libraries +I cant but agree +For Ive committed the patch fixing this behaviour +This is an AS issue +Validated by Vladimir. +Patch solves issue +I cannot reproduce getAnnotation as +I guess was not yet merged into the trunk at that timeNow it has been mergedCould you try the same patch again +Well as I decided to comment this out it still awaits resolution +Will commit it shortly +The same BUG can be reproduced if AMQ is configured without journal and with +Working on web expressions and standard way of configuration a handlerparser +Awesome MikeI am applying the patch right now before we get too far out of sync +Fixed. +change passes locally +The unit test suite should pick this up if it does happen +Looks good +The attached includes a test for the bug described in the bug report testing each of the three methods +Ill add to the fix version. +Moving this to it may well get done before that though +Its probably being used by java script +This strategy provides fairness to all datanodes +reopen to be able to target correct fix version. +Hi guys I added a simple patch to bug which solves that bug for me +If the default does not exist in the readers schema an error will be thrownThe only way that a default value in the persisted file itself will matter is if you are deciding what the schema to read as is based on the contents of a file + +They look very cool and I think they are a great addition to the community however before I check them in Id like to get past some of the errors Im having +Attaching the patch +Ive not seen others with this problem nor any advice to try it in the documentation so Im not sure why you would want to do it +Then if that is set all files indexed would be prefixed with httplocalhost instead of fileAlso in my opinion this configuration information definitely belongs as part of the process of setting up a file connection not setting up a file job +seems to address the need and changing the default to suit particular platformJVM considerations doesnt seem correctThat said its very tempting to just raise the limit +The same testcase should work +I think there is some issue with patch v +After my experience with the Slide server I can only recommend non blocking locks +ok +Maybe Im misunderstanding but wont the servers response to acceptcounter the clients proposal introduce the same delay? It also introduces a complication whereby the client is still guessing if it has the required credentials for the auth methods even though it may need information from the server to determine if it has the credentials +But really not think this is critical +Joachim assigning this to you as you are running on Unix +Attached patch adds the featureNotesI piggybacked the time zone setting onto the service so that we wouldnt have to modify all of the filesThe list contents can be controlled with the file just like the localesThe list contents may seem redundant +So all issues not directly assigned to an AS release are being closed +dont ask for the password for a second time if it is not acceptable but we do still need to output the requirements. +Pull it back in if you think different +Im assuming this should be part of the official documentation not just a Wiki page right +Now to move the assembly Anil made in EJB Core to security module Ill take this on + +sounds like we should split out the add acceptDocs to get as a separate issue just like we did for +size of M and split size of G +After some thinking +Applied the patch and fixed tests to pass the needed loggers to the componentsNot changed the default level for logging of imapserver it is experimentalalpha code I prefer to keep DEBUG by default. +metadata + + +So now this issue should be resolved because jar publishing into maven repository must be done after the release +it works now very thanksSean Owen +The MED editor now receives registry events and updates its error messages when a registry event is received +I have attached a patch for a svalidateEntity tag +Note that the script attached here has had some changes made since being there may still be some further fixes to be made +This bug showed up in are you sure? The surefire plugin works for most people even with that published and the plugin itself does not depend on any providers or the providers pom itself +Due to the license on this code it cannot be included for redistribution by the Castor Project +has that changed +This is the patch created off +Good work Bertrand +Its in rev . +It will then need to iterate over the held flow definitions and destroy them +Could the progress information just go into the standard Eclipse Progress tab? Then there would be no need for the additional window +Reporting use discovered this was a problem with his NAT setup. +Shouldnt the snapshot client say Im done with themI understand about the problems involved in this it seems like theres a hole in the state sequences +On Scott Stark commented onThis is something that would have to be investigated as a jboss proprietary as there is no interest in supporting this at the tomcat level since its not defined as allowed by the servlet spec +Thanks for the contribution Jeff! Resolving as fixed. +Hmm +Dear PhilipWe are also facing very high CPU utilization after we migrated from struts to strtus +right +I have tried ant tests and a slew of othersI would be glad to pick up work on this otherwise though +The magic byte pattern we added in for detecting iWork documents seems to be too eager as it matches also this document +Updated patch for trunk +typo on the commit line actually I wanted to refer to that case perhaps amend the log comment accordingly +Ill commit this momentarily +Actually I am using mac + srcmainjavaorgcodehausmojojspc file data +Thanks for the patch Jonathan +Better but still not right +VERIFIED +It can be changed in the root +It seems related to all right +Closing based on workarounds in comments +Remove properties that are redundant +Please note that this patch has been created from the project and it has been created from JCR new patch sounds good +This patch fixes that +It looks good to memvn clean install finished successfullyHowever I think that issue name should be changed to Implement moddiv function +Also Restore Defaults and Apply buttons could be added to the Preferences Page as a standard Eclipse behavior for Preferences Pages +Make makeQualified a method on Path and make the call it +I already worked on thisand i has prepared patch for thiswith that patch it will work in required manneri am attaching that patch here +I dont have a strong opinion but I havent had time to learn about each in detail yet +What I tried to say with my last comment was that I wouldnt be worried if we documented the feature in its current state in the tuning guide +The is on the applets classpath but it subsequntly downloadsthe application jars on the local disk and uses a classloader to launch thatapplication and it somehow causes the classloader failure to load the messageproperties +For the version of Confluence we have it involves hacking an export of the space to find and replace various parts where the name is +Will start implementation + +Michal? +Please try to see if it is possible to actually update the namenode with corrupt blocks which fsck can detect +How would thisinterfer with your ideas +Forget my rambling about the constructor which has nothing to do with the issue at all +I filed to discuss optimizations for symlinks +Can you check what AMI its actually using when you dont specify +While RELAY does forward the multicast to NYC where it will be invoked on C and D which in turn send responses to the caller in LON the latter wont wait for the responses from C and D as it only blocks on A and reason is that null in LON means the current membership which is A and B +fixedso possible to mark this fixed. +Logged In YES useridYes the foreign key column type was SERIAL +patch added +Reporters of these issues please verify the fixes and help close these issue is not observed now on a upgraded setup now +also updated related javadocs +Ping +I have a controller thread that asks for checkpoints every ten minutes and all my logs showed were about of these checkpoint requests falling on deaf ears +I have the RHEL yum remove command +We could even do a release with the first two IMOIf we keep adding features wo releasing its pointless +There are lots of advantages but Im missing the combined unitit test coverage I used to get from EmmaRichard +Thanks Alexey +Converting to issue with commit of +except the jclouds version in this patch needs to change once we get title and description to match the patch +sorry for the multiple posts +I have committed this +I am running regression tests nowTouches the following filesM javaengineorgapachederbydiagM javatestingorgapachederbyTestingfunctionTeststestslang +Clusters are of poor quality but some appear among top coordinatescontent type jira browse hi all apache software pgp signature version Overview of clusters obtained after iteration of with +Fixed by reverting an accidental change introduced by. +I have updated the test not to use the anonymous inner class. +Thanks for following up Sergey Shelukhin +I am going to propose to costin that we work this into a future Spring DM release +Due to problems with the public Jenkins instance they are not yet visible from the outer network +Thanks for review Bikas! added edge hookup in is now cleaner and instances are shared by source and dest vertices +The primary problem with is it doesnt participate in the infinite loop and metadata scanning detection algorithms correctly so much so we explicitly detect it and give up early +Thanks +Ive updated the project home page and sbs space overview page with the updated videoThanks again Burr! +I would prefer that a separate method be broken out to mirror the case rather than set the properties inline +I guess this issue can be closed then + +Why does Java Script conflict? is bundled with the same ones as Prototype Scriptaculous In the the file says Thu Jan +We might fix this by just using a different method name in and deprecating but keeping the method in +The hashmap was replaced by a threadlocal +Then well be able to provide the code against the +Carl its just like there is rm and there is rm r +Im still not sure this is fixed Im seeing a random spin up +Here is my config and zipped traces of the issue with schedulerSupporttrue and with schedulerSupport get the same issue when try activemq on linux with +Geirthe only reason is that Jitrino was built without files from vmjitrinosrctest folderDid you installed my patch before testing +In that case I suggest that the documentation around exception handling should be improved +Tests with external SSO provider applied +Added check to make sure doctype is present if it isnt dont use it when writing the xml file +Frankly Im not even sure what its supposed to do since it is not documented +How do you expect people to adopt Groovy if fundamental mathematical expressions dont work? +I upgraded this to critical bc it confuses customers and wastes resources +bq +Committed under revision r +The latest patch looks good to me +It would be better to have your own start script +includes a few incremental patches the first applies v of the second comments out the use of and other add instrumentation around the RSs internal scanners next and row delimiting functions +only close when thanks +This file could be located where the pidfile is +Thanks Joe for this patch +So we left the staging and production trees intact from what arieswas using on the incubator site +From my side we can close this issue. +Closing all tickets +to deal with this one +preferrably a unit test +This issue has been resolved on trunk through previous changes +Michael +Retargetting this issue at the Sling Mime Type service +The tld is now shipped in the jar and the deployment tasks no longer adds the library to the. +It seems similar to is that right +Only if you remove the block between the inner andouter will the vertical positioning work again +TODO Add wiki documentation and write what file language headers is wiki for file language to list what tokens is supported by fileftp consumerproducer +Well try to fix it soonFor the other problem I guess the maxwait parameter would have fixed your issueAFAICT you get that exception on windows if you try to get connections too fast and the OS has yet to free up that resource +I cant catch it in the latest build +Whats the status of this oneCan it be reproduced on the sandbox +Alex add schema binding resolver +Very strange +NET version so thats most likely the reason it was used instead +The net result is that strings dont get escaped unless the content also contains an ampersand char andor a less than char andor a bad char +could you create patch for and attach? Of course I can +QuestionDo you need collapsetrue when you can detect whether has been specified or not +unless Im looking at the wrong patch if theres no authority its immediately dropping it and relying on it being later instead of just adding the authority immediately +Heres my work it seems to do the of this was modelled on similar code in the package +This patch makes use some new methods on to do things like create a completion service and do a distributed request with a software load balancer +youre implementing a standard there +I think this is a bad idea +We use maven version it solved in maven +I hope there no timeout on the consumption right? Processing each batch can also require additional queries which now requires new connections etc +Youre right +The attachments are meant to show how the project may want to consider a modularization of the River codebase using either Maven or Gradle +Still broken in +fixed in rev thanks for catching this! the fix in build and confirm it works now +Updated based on Richard Halls naming suggestion. +Looks like the relevant commit was on Apr one day after the rebase on Apr +Thanks someone helped me with this on IRC + I dont understand your comment +Are you sure it still occurs +Are the versions were using the latest ones + +Thanks a lot for this feature and to respect our code styleFYI I renamed goals to be consistent +On the other hand I agree with Christian to not preserving leading zeros +I agree that its not necessary for column names especially in a first release since the comparator type is always available +This really needs to get fixed for +will have startup issues with static initializers +Does the new test pass? It calls getState but the new command is get +The jbossws lib is not included by resteasy itself its configured for +See if hadoopqa will run it now +Thanks for the patch Michael +Exactly +Thanks Thomas resolving this as a duplicate of +Commands supported INSERTUPDATEDELETETRUNCATE +That would be controlled by XSLT so its not j +While in the values are which lead to the security failure when creating using the wizard +Please verify +pete +updated with trunk +AllI have attached is a simple test case which demonstrates that Envers fails to handle any entity which has the Annotation on the class +If a user is manually entering credentials it never challenges them again on subsequent requests +I am still seeing the same problems without the patch I provided applied +woops thanks. +Thanks Radim +I also tried doing both in one run and I filed for that issue +Integration tests were updated. +Transitioning to waiting for for user have access but not sudo +sorry this patch got missed +Closing resolved issues. +Is this going to be fixed soon +Is SLAVEDB a good name here? It seems to me that weare talking about the database at the master here +Actually I cannot reproduce this +Ill fix this +To many changes in code to include in a patch release code must work in +move the util method which sets read column id list in conf from to fetch the read column id list from table scan operator in store the read coliumn id list in table scan when does column remove in the patch when test and commit +The performance cost for that feature is simply too high +I was able to apply by splitting up the two hunks in the patch since the patch utility was complaining about the second hunk and not knowing what file it was for +That is for the current policy of shutting down the NN on such errors +It isnt supposed to point to the dev list +Having a closer look it will be tricky to do without making the code really messy +When generating code for a call the Java backend checks if the method is a property getter and uses the Native annotation on the field +I can open a subtask for the documentation for this issue +html template are available in SVN at beehivesite and on the website +Good catch Samarth +Found that the modifyColumn or modifyTable works fine even if HCD is changed in modifyTable +Let me know if you need any help + should be should be +Hi Joerguhm not fully agree on the normalThe use case is release a parent with a SNAPSHOT in depMgmtThen release a child with dependency which is finally now SNAPSHOT will failed due to the version inheritanceYes we could make to configurable stuff but IMHO it should failed by default in such caseSure it will backward non compatibleOthers WDYT here +When it reopens it it works fine so why cant it do this under the covers +Yesterday when I looked its not there m closing this issue now. +DOES apply to as well! Only not as frequentapparent +The updated patch looks good in this regardbq +It appears there is no support for query params and using the cat command in this example is not essential for this demo so I have removed the lines from the README and will create a bug at apache to get this fixed. +Thanks +Please verify +That aside that property seems to be just meant for programmatic use anyway since its inherently not to specify the target object to unmarshal into there +Since the does not provide an API to pass the cert as a stream the earlier option is the available option for now +Formatted code +Deadlock still occurs in r +In several files there is a url element that refers to the sandbox +I vaguely remember a situation several years back when we updated the query plan in the ref file due to an ordered null semantics diff and we couldnt figure out what had caused it or why +Edited these issues to set the fix version. +Frankly its turning nasty assumptions about parsing XML are scattered all over the placeDoes anyone have some grand scheme in mind for handling this? Ive got an approach but its tedious mostly making a new class that provides a thunking layer +Fixed on revision . +If you pressed this button and DD anything to the page then as soon as your mouse reaches the editor area the selection in Package Explorer will be set to the file currently opened with the if you have this button enabled then Eclipse tries to drop the same file you are currently editing +At this point editing has not been enabled via the editor since getting the editor to make workspace changes will take a bit of work +That could be implemented +do you need to mention new facility in the Scan help for shell? If you can get it in add on commit +I have made appropriate changes to regarding the EBR to handle this. +Attaching project which illustrates the off release +Whether storing in zookeeper is under reviewI am fine with storing operation directive on hdfs. +For the future can you write a +the lookup failsOn the current with this test fails by getting a from the lookup +NET programmers who wants to extract content and metatdata from documents without using JavaThe only think they have to know is an API to process XML in there specific programming language +Attached is a quickstart project showing the draft patch to solve this problem +Will include in list once we approach GA or when all installation guide issues are finished. +JonathanYours is what I first tried but instead I tried to do it in SSTR and I think that is what we can do best forOne thing to point out is that it NPE when fpChance is null and try to convert it to double at +sounds good +The JTOC implementation of and the differ because of s isolate support. +Patch contains fixes for adding in header support to camel messages +ha! No worries +Looks good to me too +In that case should swallow the exception and let code continue so maybe we need to add some code in that part +Committed to trunk +Note that I think this is the preferable way to solve this since it allows you to change snitches for though your consistency guarantees will likely be violated and require repair afterwards. +Just had a quick look and it looks fine to me +Move to backlog until when it can be done is clearer stop just rolling from release to release +bq +Was misreading +Documentation for review +I tested that so Im sure you must have done some mistake in your source code +Thank you for taking the time to read the specificaiton +proceed is changed to throws Exception instead of issues +Everything works fineRobertAccording to your stacktrace the embedded couldnt be loaded which leads to the described NPE +ehh +i could make it happen simply by starting the node +has been deprecated. +closed with release of +We have other domain classes with up to a dozen such fields +trunk +In general there is no difference from an engine perspective between connections that are made through a datasource and those made through the Driver. +Thanks a lot MarkoI understand that you would like to see two things wildcard support global configuration for each project of a solution in addition to specific project configurationsIs that correct +bq +Almost works +lgtm +STATUS file copied from trunk to tags after moving CCLA agreement notices from Pending to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Dawid Please let us name our tests like in our old test framework has a good test runner improvement for that we dont like the verbosity on imports and this annotation +Please review and give us a go for commit or some feedback if you would like any further changes +Lets see what others will tell on this +on patch +We have yet to perform performance testing but the overnight and performance so far hasnt generated an exception on client or server +logs and server for mystification +Would love to hear about other ideas having the cluster become inaccessible if the user account for a certain user happens to be on a node that is down seems like very problematic to me +Would like to get some tests for this also +default mode is client +Thanks +Hi Hugo looks like you have first launched Sonar at parent level and then at each module level but I guess this is not the case? The first unexpected behavior is to see all your Maven modules in the All projects page +BrianI see +I have attempted to alter the audit log domain object mapping to eliminate synthetic keys which allows for a simpler insert statement but creates problems reviewing the log in standard scaffolding +Whenever you do cvs update then you should explicitly do build clean then build especially when there have been major updates +Hi NicholasDo you mind taking a look at this patch +resolved by +What is the identity strategy used by? While reproducing the error in JSE environment I am noticing that the crucial factor is the datastore identity used by a class +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Im not sure if this is an acceptable solution for you + +However if I click preference and configure portlets then I can drag and drop portletsVery weird +Thanks for the patch David +The same thing was already proposed and implemented in the +n kr kn input inputinputfile mapper bincat reducer output output For any special charecters like and p and letters instead of numeric it still sorts it +Any chance this is low hanging fruit? This would be a very welcome addition +Added now +Custom zkclient jar +The KEYS have to be cleaned +How does this relate to? I was looking at that +Attaching Added unit tests for Added a check for Nondeterministic UDF +I have not received any feedback on environment so this wont get done for this release +Hey I also want to be listed in as I did the new Lucene API with its sophisticated v and logicOtherwise looks fine +I deployed a snapshot and the site +Seems to me that the entire build should not fail just because a single file is missingIve worked on the code in the past so Ill take another look shortly +So +The images in the book use the Drools logos and do not show the recent change to the navigation entry from packages to knowledge base +There is in SOA but not in from in beta +I am on this +Permitting arbitrary flexibility would also make quality control harder when we already have too many licensing documentation bugs in Apache releases +I committed this but dont plan to include it in as it could break applications that use illegal names in schemas. +Please verify using latest from trunk +I tried this test on Batik and the bug is still present in that version +Move the issues to as not critical +in release Thanks MarcoOK for me I just fixed labels +Cool applied cleanly +added methods to reserve and unreserve to need to ensure that any component which is using a commit point should reserve it till it is done +This issue is closed +Let me test again with the latest code +Works in my environment +Then people could change the colors to anything they wanted +Puyoman Did you build the xerceslib with icu? if yes how? Can you reproduce the same error wo xalan if yes can u forward your testxml instance file? thanksRgds came to be able to do normally when recompiling not by Forte but by gccThanks for all. +Thus so is Exception and so is +Assume closed as resolved and released +is coming out on Friday +Since the config isnt shared between IPO well likely need to set it everywhere +Maybe you should use it in this style old new it oldhope it is helpful. +Added patch without stuff in it +Assign to is working for me. +Precise easy to understand +Agreed +Making this critical so we make sure to triage it +For the first case the question would be should thereturned dynamic result sets simply be closed or should each row be fetched and then closed +Thanks for the reviews everyone. +during time constraints im fine with this moving to next denis if you got something already use it otherwise postpone +Fixed when the issues were fixed. +If issues are discovered please feel free to. +Yes for sure +If you move the definition of the mytype so that it is BEFORE the definition of mytype then you do not get a null pointer exception +I am not using WTP because Jetty is enough for my seperate project configuration is not really satisfying right now +Simpliefied implementation Moved everything to impl which is not public Drop Change and Drop and use to gather information directly Add argument to send method indicating whether to set the actual status code or notI removed the Change and classes and the interface because these bring no added value IMHO it is better to gather all the information in one location before sending it and the Change and classes are simply not needed as the information is one way +It was simple cut and paste +This patch adds interface to the metastore package +I like it all except Im not sure about the cleanup methods +The problem with the scenario is that using a inner class causes the instantiation to fail +I forget jsp and classes directory into the patch +Max if you can attach a sample web app showing how schemas are referred to then it would help otherwise Id be keen on closing this issueIve run a quick test today with Tomcat Ive added a temp log warning showing schemas can be loadedsportFactoryserviceBeanbookserviceschemaHoldertheSchemasjsonProvider root of factory hierarchy createSchemaWARNING Schema resource classpathschemas is available on classpath initDestinationINFO Setting the servers publish address to be bookstore initationContextCXF libs are in libSee the attached sample web app as well +Sure you can just include a template into but +That might be moved to some other testcase +It is risky to appoint resource and static files separately it causes bug with resources loading +Will try fix that and then maybe I can reproduce the customer issue +It will probably be a while before I can spend sometime on this. +method calls method so to be fully safe one has to set Transaction logger to error level +I am not a user of Pax Runner but if you can detail the steps to reproduce I will take a look at it on trunk +The was misleading +After applying patch do a ant clean else you will get lining errors +Attaching the seems reasonable to me +Fixed in current Issue +Helping with the shells that the scripts work with is fine +In most casesthrowing an exception on bad data is safer and this can be done for all theconverters just by ensuring the converter has no default value setThe referenced bug makes good arguments for having struts form beans have onlystring properties ie struts is not a justification for this feature as far as Ican seeTrimming whitespace does sound reasonable at first glance +But it doesnt fit our needs We need to arrange rights for the moment space adminmoderator can move from Development space User space of his project but not to another project space +Its conceivable that we may take several directions including push shuffle etc +Attachment patch has been added with description patch against to fix the length to and patch to seems to have addressed this issue in at least one example +Seems that was not the right should be can you give me a little test case? I just deployed a new snapshot which improves forking output so that may help +seems like similar +we should not support the bamconsole +Fixed in rev . +It does not matter how many clusters you use +A difference would be observed if you started writing to an empty Accumulo instance w native maps +patch for hudson +Done either for org and for com styles +messed up the last patch +is this something we can handlefix +So which components would be necessary to develop to handle those plugins? do you consider to develop any of those plugins as examples +Dupe of closing. +There is no property cmistitle +Wanted to avoid thrown an exception but does the same thing in the end +There is a catch though the patch does nothing about the view groups for a corrupted DB +Answer is that streaming is discontinued and exception is sent using as same as other placeOK +txt after the patch +Is this an improvement or a bug +Wizard is replaced by invoking Properties dialog with Project Facets page selected +Im not sure I want to be the judge as to whether thats a good thing +I also changed the sentence after each list in the data type topicsPlease let me know what changes are needed +Especially not +Ive verified that wildly RSE now actually sends the proper launch args to the remote machine but I couldnt test the full flow because my remote box does not have java +I also dont think wewant to base enforcement of a database level action on whetherauthentication is db local or system wideSeems there are more than Ricks three toggle dimensions in this spacethat matter system vs db level authentication and readOnly vsfullAccess connection authorization +This is fixed +This will be done in a branch off of head +Fix delivered +Whats going on about this JIRA? We have clusters with Hadoop or series so we need HA for JT +if C is relative which path is it relative to? If it is workingDir what if working dir is at another drive? For example Dfoo +Did you finally find the way to resolve this issue? I would appreciate if you let me knowmany thanks! +I dont know if well have time but it should resolve all remaining flow control issues +Thanks for the quick turnaround! +In the there is a test that shows the problemIf you leave the program run till end the output isOUT main startINFOmain startOUT runINFOrunOUT run HELLO INFOrun HELLO OUT main OUT run HELLO INFOmain INFOrun HELLO OUT main OUT run HELLO INFOmain INFOrun HELLO OUT main OUT run HELLO INFOmain INFOrun HELLO OUT run HELLO OUT main INFOrun HELLO INFOmain OUT main OUT run HELLO INFOmain INFOrun HELLO OUT main OUT run HELLO INFOmain INFOrun HELLO OUT main OUT run HELLO INFOmain INFOrun HELLO OUT run HELLO OUT main INFOrun HELLO INFOmain OUT run HELLO OUT main INFOrun HELLO INFOmain OUT run doneOUT main INFOrun doneOUT run exitINFOmain INFOrun exitOUT main terminatedINFOmain terminatedOUT main exitINFOmain exit DEBUG Shutting down SYSTEMOUTIf you hit the output isOUT main startINFOmain startOUT runINFOrunOUT run HELLO INFOrun HELLO OUT run HELLO INFOrun HELLO OUT main INFOmain OUT run HELLO OUT main INFOrun HELLO INFOmain OUT run HELLO INFOrun HELLO OUT main INFOmain C DEBUG Shutting down SYSTEMOUTOUT run HELLO OUT main OUT run HELLO OUT main OUT run HELLO OUT main OUT run HELLO OUT main OUT run HELLO OUT main OUT run HELLO OUT main OUT run doneOUT run exitOUT main OUT main killedOUT main exitNote that all the last logs are LOST +Got a link +Docs inline with changes +As Mark said the dependency plugin has already this feature. +Fixed +Yes I can confirm that the default in the UI is +It does NOT look like this was addressed in the patch +It is fixed in. +Thank you very much for assigning to this issue KimI guess you are right that we only need to document replication in the Admin Guide +Why do we need that task? I removed the notfound node before GA release because it was not working correctly and there were many bugs +Thank you very much I wasnt aware of we surely dont want to rely on feature and I will change our implementation accordingly +the OAK API should not expose information thatare not accessible to user in JCR +Yes tries to delete the users home directory! Which raise the question of what if this testcase fails and it successfully deletes the users home directory +As we port to we can remove these as much as we can. +But Im not necessarily keen on this approach true srcmainjava is usually where the folder is but thats not written in stone +for the patch +Update the review board with some commentsCurrent patch is building the jars properly +Same environment +I swapped in and both tests in failed +Thanks for taking the trouble to produce a nice clean patch Ron +The purpose of allowing different numbers of map and reduce slots on is to efficiently use resource on a node +The start index is for the replica of the first partition we create +Slashs in the end or now slash in the end whatever I try empty directories are left out +In that case just fall back to the original code. +Not integrated yet. +You guys have any more informations? Im happy to commit the patch if that solves a problem +Jakob could you please commit the additional tests +I committedSending srcrefTransmitting file dataCommitted revision . +It adds deferred rendering to the HS jobs page and HS Job Tasks page +The first pass of this merge is complete and I have things running locally +Looks like we flip back for now in either I have looked at but havent seen any patch available +Heres a test case that generates the exceptions +This patch adds a templateEncoding parameter to the goal + +It may be a problem with just one of the worries about the hack for the requests I wasnt expecting much more +If the action can still be executed with the action class name then this is not really secure +We can remove them +He knows better how evolved +for the patch +The fix is +Added groupBy property to queries so Seam can do the right thing here +Is this Ubuntu only +The constructor needs to keep the two arg version so it can know the date rangeto iterate over +I committed this to trunk +I applied some changes in trunk that you are welcome to test +Are there other admin to consider? +Thanks DarynAlso thanks Matt for the review. +I remove the SAXDOM in +No way to fix. +To reproduceDeploy the quickstart to the qajmeter I used threads second warmupms other comment I did not see this error when testing in the same manner against MR +Why do we have thatIt feels strange to me that a could have no +Still need to check next RCP build +No failures +Here is the above patch as Subversion diff +Isnt there a way to makethe metrics cumulative? That sounds like the only fair way +You should create a new issue or at least this issue +It seems that it is there by intention of minor optimization +Scheduled build Result all three project groups were distributed to the three build queues +Resolving as a duplicate of. +Jody this is exactly an example where lack of performance planning makes code correct but not usable beyond toy data store is orders of magnitude more ll have a look at JTS but given our continued performance issues with JTS I wont hold my hopes highthought I maybe try to fix the performance issues there too it would be sensible to share instead something we can ship now? If first option does not work third one is where I would put my money on +If we add one and the email contains a word document with an embedded excel spreadsheet then suddenly because we added a Parser to the context recursion will see your point it makes sense no hidden surprises +I just committed this +CHM makes sense as these objects are stateless +User documentation for Rank operator +Makes sense +Can you give any specifics? This isnt true in general +No need for fixed length headerThe Maven project I have requires protoc at maven build time +for the patchCommitted to trunk +HiIt makes sense that to check if the visibilityTimeout is set when the extend option is trueI will commit the patch shortly +overallIt still has minor issues tests findBugs warnings and a couple of bugs in the PB wrappers +Ive been swamped at work so I havent had a chance to look at this in depthI have not tested it yet but in principal the patch looks fine +Your approach does not let me change the default timeout value of minute +Higher confidence more likely to commit +I incline to the latter choice +Im currently working on a streaming solution for Chukwa but open to share the work Just let me know what you have in mindMy current plan is to use Thrift for principal reasons +Kevin can you take a quick glance at this patch and let me know if its good to commit? Alternately we can just commit +Looks good +this is a pretty serious change while it makes all previous incompatible it also allows for minor changes to the interface without altering the version number +After that i start the job again documents sent to solr but without any metadata +Ill go reopen +Not to argue featurebugfix but I think there was an actual bug here which iswhat got me looking in the first placeThe server I was accessing was trying to negotiate the algorithm whichwas silently ignored and the client responded with an MD algorithm which causeda Bad request response from the serverId suggest that if this does not go into that at least log a warningwhen the algorithm is not MD +SSH can move on to next server and queue up more log splitting but wont get blocked +this evening or tomorrow morning and make those minor changes before checking inDid you run +I dont like it +Changing fix version to unknown +The changes looks good to me +So closing +ok Ill not mark issue resolved till the code commit +Cleand up Javadocs and added them where not added yet +Using makes me nervous +The attached patch file updates the and to allow the dispatch servlet to be configured as described in the description +Record number is like the line number. +Same as for initial version of source code forward patch +Nathanael theres no hard date on release were working hard on making it happen in the next couple of monthsMark can you post your client code and configuration to the user forumCheers +My mistake +thanks +Thanks good work +One Blocker gone +This would have the disadvantage of having incomplete information for the first few splits +Thanks for comment RayI guess we may wait Scotts patch on before going further on this +This change should be fairly simple so why not do it before submitting? srcjavamainorgapachezookeeperLAre all these trycatches necessary? A single try catch around the whole lot would do +If the thread pool is a workaround sounds like a bug is in our thenfwiw I highly recommend using Java concurrency bits where possible +Raising priority in the hope someone will at least all old versions should now be pointing to +Should I add some sort of grouping based on the role prefix? Yes I would just sort the list +First pass for comments this issue only handles the index side of things +Are you sure you have tested this against current CVS? If you have I need a copy of your very simple test case +Your system gives me the following definition of a critical bug A critical bug is one that crashes the application causes loss of data or severe memory leak +Doing this halfway doesnt gain much and the current state is obviously confusing as it was causing your problems with your shard splitting patch itselfFurther even in the shard splitting issue Shalin talks about splitting that patch out into smaller issuesI dont know that that is necessary but this looks like a perfect such sub issue here lets improve the API call naming and start making the right calls? That would seem to mean that clusterstate updating code uses getSlices shard id finding code uses getSlices and a fair amount of other things should use getActive slices +Suggested fix +Since the build changes so much in M this needs to wait until we work through M release so this can be completed early in is resolved because Paul posted the correct way to use the stage in the build space. +The bridge is missing a dependency on +Thanks Dart +I always do that for conditional I think it makes it easier to find the block bounds +is the right place to create zip archive +Observers really shouldnt send and if they do we have a bug +This behaviour is expected given the current code base and has been present since the original failover work was done +But I mean change the keyword to boolean as well +Hi MarcoI suppose its normal practice internal releases +Please disable warnings for the time being in a bit compiler +explicitly state in the contract that the method will return false if the property doesnt existI dont like too much since it breaks separation of concerns removing a property and checking for its existence +This makes most sense +I also ran unit and manual grunt tests +Hmm +Would anyone like to rebase this patch +So I think we need to instead copy into Avro +Requires some investigation on the surfire way to listen testResult and generate the TEST +If we require the check then this has consequences how Batches are implemented +CXXFLAGS for C and CFLAGS for C +Moving this issue to the correct component +Were using a modified from the fuse php client +Thanks Todd. +here it ve applied your patch thanks for your contribution! +It would make it easier to apply if you provided a proper patch for the required changes +This patch adds and rolesYou can have only instance +Marking this as fixed after Johns changes to the freelist pointers Im unable to reproduce this bug +As an example I tried it with Wssjs handler +Added support by adding if enabled will use the default pretty printer +When I opened this bug I had a few key pieces of functionality in mind +ive reopened this issue until weve got test cases and documentation done +GuillaumePlease provide guidance here +Theres no way to see generated ATM The Realm is automagically generated and yes its created destroyed along with the application. +OK after some further investigation we found the following Activiti does not update variables but deletes them and then does an insert again +Having the user appear as the TGT user on a secure cluster and as the unix user to a insecure cluster may be a bit jarring +Overall though the idea is to prioritize data integrity consistency over availability +but I would feel comfortable ifwe could test these and if they really turn out to be problems at least state it somewherethanks +If we make it easy to spin up an ec cluster or similar that lets people easily and cheaply test their changes which seems practical +New patch with test case from version of the patch ported to trunk +So this behavior is probably correct although clearly not desirable in the case of a sortable data table +If I send a single request before starting the performance test then the issue doesnt occur + +Verified in +did a simple hack that allow write AND N read to goes through at any given time and didnt see significant performance improvement in the scenario that I tested +The validation messages uses wicket resources instead of the +Closing this issueverified that mock driver shows correct Instance States image. +When primary host is created the host information is created using property replacement in the which accounts for all properties +Hmm +bq +You xml work as it was necessary +Ive commited the validator to Im waing for proper error message text. +Verified during test cycle looks fixed closing. +Using Logback would also allow people to bring knowledge in and out of Grails projects +Sorry about the false alarm +John from that bugzilla entry it says The data is not wrong its just confusing +not sure what a hat issue isthe person is doing email as part of his main projectnot sure he wants to take on more than he has envisioned +Does anybody care to work on this? If not Ill close it +I took a closer look at this and I got the same results +There is also the problem that for the moment the artifacts are not available from any Maven repository. +Committed your patch in revision +Oh BTW I do sense that we should rename to +Proposed patch add property super to and +batch transition to closed remaining resolved bugs +Thanks for spotting this Sethre too slow Juergen +Latest patch from rev dbccddafdfeed +Juergen +Shai I think we should close this +Thanks Nicholas for the information +For hadoopBundle profile besides creating bundle this would make a jars folder and generate in projectFor working with eclipse we can invko mvn eclipseclean eclipseeclipse and then with the hadoopBundle profile +The is not publicSolution make publicSolution move the back to its original packageI dont know what OWB usually do in this case but seems solution +updated patch some refactoring and started adding remove collection code though currently we do not remove all collection info from zk even when you unload every shard something we should probably start doing +waiting until all replicas are placed +Im out of time on other commitments. +You really have to download and step through the code or just guess if you dont have diagnositc outputSecondly an extra check in config code has trivial performance impact because it is typically only run at startupsetup +Fixed reviewed by Vinod +Having JSF available globally led to some weird issues for applications that werent even using JSF +from tika you must download the source code and then compile it +Is this bug even still there +bq +Verified in due to +If so than this patch doesnt do that much as of now +What is the problem with coming up with a for this + +I created this issue but upon further analysis It seems attribute attachmentFile will be always setSo this does not seem to be the causeTherefore I am closing this issueaki +There are two issues about this and +Any chance of a backport to the branch +At least thats the reasoning I remember. +Run eitherjava jar java jar +ThanksBest regards! JIRA closed. +Can we just apply original suggestion and resolve this now +Then add it into the form and populate it on adding a registered application as well as listing the registered applications +Added a perform method in to return the release results +Closed as per state in Bugzilla. +Patch applied +Thanks for the clarification of the Class loader stuff +if all tests pass without this jar why do we need it +Thats strange +Thanks Alex +For the record +But this didnt help the text is still not displayed and no error occurs +Dependent on posting a session rating should return a if the session has not completed yet +I was hoping it could be more isolated and maybe there was some code already somewhere that had to special case user access to system catalogs and could special case the argument to the store openScan call rather than pushing and popping transaction isolation level +Can you provide a test case or at least describe your setup in more detail +Obtained usingsvn diff +Good call I will update +Jawahar if these answers are insufficient please provide a sample of the exact code that is failing which covers any relavant declarations and the Exception log entry that results from the failure. +Right now the unregister method checks if a given key is registered with the server and if so unregisters it +Run random writer sort and then sort validation +Sorry I had attached wrong a wrong patchNow I have attached a new patch which is successfully compiledAnd I have added no test code because this change is only about exception message. +Here comes my test left out the libraries so in order to set up Unpack castor add commons logging and logj jars to the directory copy these files into the castor directory and run ant from command have included files that dont work and one that does in order to switch between them uncomment comment in the respective lines in regardsmf +Hadrian Sorry about that +to Iterator is a component that allows iteration over its child components +Thanks applied this with minor changes. +Please point me to the instructions on deleting the index +Is this the same issue as and comments of +Prasad I am assuming you are looking at is missing from patch +Thanks for looking +Please confirm the current status of this issue and if its incorrect change it as required +Thanks +Id like to set the number of concurrently running map tasks per node for a single job +Thank you for your help EdwardI am working on it +I wonder if anyone can shed some light on this +We need to rethink of the algorithm before we englarge it +Windows bat works fine. +Cool thanks as expected to save memory. +Defer to C hosted +Hi Rick thanks for the extra information regarding why Principals cannot use multiple string constructors in the policy file +The attached patch will determine the console framework servlet path as follows get the path info for the portlet from the portlet request get the portlet name by splitting the path info on create a renderURL from the portlet response truncate the renderURL at the first occurrence of the portlet the patch fixes the problem +all artifacts from this component are being built creates artifacts and puts them in the etc directory under test +The fix does allow the form to be submitted in inline frame +In attached patch I added one more offset line and two more testcases +please implement this in this way I will be able to use tomcat plugin for integration I see the source of this issue the project generated by the artifact includes the path at projectbuildpluginManagementpluginspluginconfigurationpath since I had it at projectbuildpluginspluginconfigurationpathBut reorganizing my pom as it should be I got the following errorThe plugin does not exist or no valid version could be foundNotice the wrong groupId of the plugin which Im not sure from where it comes from a bug +woot! close as dupe +In in the compile method we find that it is attempting to compile a Source of for the className org +This submission popped my Android cherry so be gentleGlad I could help out even a tiny bit. +Youre welcome to it though +REST api gateway may be difficult to secure the system +the ones that came from the original Continuum installation +Excuse me +I am guessing the component but the converter package is well named +One statement in prevent all classes which is load from classpath to be included +We should definitely be sharing both indexes and Readers across all gets and scanners + to trunk +I think the issue here is that Hibernate simply cannot lazy load the entities in this and must hydrate them before adding them to the setPerhaps it should be a configurable option if some users are using the ID and want the lazy loading but when weve learned from authors on the Hibernate team not use the PK in equalshashcode and have been doing this practice for many years this definitely seems like an implementation bug with +Im resolving this as Ive had no response from Keith and Im pretty sure that this works correctly anyway. +Its getting very late down there im being lazyThanks a lot for your help. +xcss files or have the common one so I cant generate summary table for every component +That is what scares me the most and makes me wish for programmatic transactions +You could have searched for the same issue before posting another one +For just one thread is used to process assignment events +Securing images is a valid use of Spring Security +The meta hlog file is created in the same root hlog directory but is suffixed with +Sorry for the delay +Applied in r +added as part of commit +I propose making it convenient for direct reading and supply a special parser with a tool that will convert this log into tool specific format +I think the idea to leave the current behaviour as default and add a flag for enabling of caching the image would be ok +This is leading to leaky abstraction +Thanks +Determine whether the profiles are included in the default set of server profiles shipped or if they are only available with the EC images + has an attached patch for this has been resolved with the patch for +hmm is it fixed on? maybe an error in the version of the application triggers the again in and still broken +Thanks in trunk srcwfssrcmainjavaorggeoserverwfsxml Completed At revision Committed in srcwfssrcmainjavaorggeoserverwfsxml Completed At revision +I have checked language and content of half of the chapter +bulk defer of open issues to +LarsOk Lars +and merged to the branch in revision . +Im going to commit this momentarily +A +never mind my fault extracted to ctest and it worked fine. +Attention MarekOndrejPlease review the updated version of this document and confirm whether or not its now correct +However the json page fails with orgcodehausjettisonmappedConfiguration +This is not technically a bug +What happens if we restart the sync thread? What if we just continue to the next iteration of the while loop? Also it would be good to have a test case +Closing as fixed as per last comment +what i miss in the doc is the livespan of the mapping does it exit only for the current request or the application context livespan or is it persistent +That is actually a good idea except we should remove it altogether +b localhost as well as opening it with in JBOSS we can bind the address to my local IP address while running +We can have and used in conjunction +Would someone more familiar with the PDF plugin please comment +Therefore I close this bug remove all advertising of the footer feature and suggest to use either embedded JAVA code to achieve the same result or XSLT to transpose columns +Thanks for volunteering +Yes I have a surrounding form +How does this differ from the way that false works already? Is this a documentation issue or are there functional changes requiredCurrently when false setOwner and setPermission does not check permission either +Let me try a bisect +I did svn delete to and +Dont worry no rush! Anyway I wasnt able to assign the issue to you since you dont appear in the list do you know why +Move site building thereThis patch will build a binary tgz +Ok +Im tempted to resolve this so we dont have to maintain two CSV web servicesIf I dont hear any objections in the next hours Ill resolve this issue as such +Should be fixed +Just a note Only framework build went fine +seems to be garbled can you resolve that +That is not correct +Attached patch introduces a new rule to push foreach with flatten down the tree +I have been running a concurrent test on Nov +patch got messed up due to diff svn patch +We now catch exceptions and log them and continue +Committed to and trunk +Thanks a lot Daniel Im processing it right now! +Unfortunately I wont be able to look at the test failure until this evening I have to get out from behind this firewall first +I would like to avoid adding multiple dependencies like thin and rack if possible and keep the clients as minimal as possible +Strong can you comment +for applying patch to since it does not haveThanks Tom and thanks Karthik for the review! I committed this to. +Registration with null would be unnecessary +My impression is that its a lot of work with limited benefits but I can take a closer look +I also find that writing it down helps me think. +try deleting what you wanted to delete from the save it then close the project and reopen it +I have verified the results for several of the properties it deems UNUSED +domain classes +Neither this or should have slowed things down because they only affect multisegment and multiindex searches I thought +I do wonder what itd be good for after the fix has been properly implemented guess we should return the value consistently I consider it an oversight that we didnt do that in +Sure Attaching v rebased onto eaffefbfabbafee +Added released +Committed to trunk +Ive looked at the gc through the profile it was around minor gcs with CMS s with G w or wo the patch +Thanks Nicholas +Hi and thank you! Seems I missed this major aspect +Its very error prone to maintain a list of exclusions in the WARs pom +Chris One more Minor nit Your latest patch does not have the Apache licence grant check +field reference jpa class +Wonderful! Ive executed selenium tests and all tests have passed. +All issues has to be filed against from now +col value we could try to determine the appropriate form from the length +patch for HIVE QA to run precommit tests +I might close this let me check one thing I think I was referring to a branch and the trunk appears to have this issue resolved +Committed +As we have source code of JSCH maybe we can consider to build the jar ourself +v is the winner +The corresponding event fired has to be fired in any case. +Can I close this issue +I was doing other work on this class and noticed the missing serial so this was fixed for the AS release +this fix works for me +Nadia could you please retry with revision could you please launch a daily test on mysqlinnodb with CQ with rev ? +A difference between and Sun Java is definitely unexpected theyre the same codebase +Please let me know of any improvements needed +Javadoc for the Java skeleton +Entry is on top of your list there +The issue with the official release is likely coming +The patch is integrated at +Ooops I didnt though of that Nope it will not work +Im not comfortable removing checks though unless it is clear that it is not verifying anything +Closed after release. +By in the passed zk instance doesnt wait until connected to construct a bookkeeper instance +Hrmmpf +doc reader could peek into the file and return either textplain or application +Im running a soak test overnight of with the mixed workload Ill report back on its success tomorrow +Clearly more tests are needed in this area but no reason not to mark this bug fixed now change has been made. +Duplicate issue. +In my case wraps the Exception with aI will attach a patch with also catch the +Do you try to use different versions to fit your requirementMy point is if you change a protocol you should change the protocol for all existent language +Do you get the same test errors in a Roo shell? I can only change Roo code not STS and specific issues related to STS need to be raised with the STS team +Len and I received an email from our point person at HP saying that this patch has been incorporated into SOA Registry Foundation +Please check that it was applied as expected +doneyou can close this bugthx +We need +Id still love to have a working test case which I could add the test suite +Ok now I understand + +It would be great if you could try it from Roo StefanGreat yes is working like butter Thanks for fast fix +committed this. +It looks like this patch hardcodes the number of partitions to keep +Perhaps we can dicuss that on dev to get a clear picture what to put where +Entering code and text with special characters Special characters such as and and HTML characters such as ltltliteral ltliteral and amp can be escaped with should probably be displayed like this instead Special characters such as and and HTML characters such as lt and amp can be escaped with Chapter +I failed to find that add method the first time I saw Configuration class yesterday +Until then woodstox will be the default stax implementation for java with released. +This way will contain bundle information headers +Well test this case too +I dont think there is any problem in doing that specially since both these projects are under ASF +now uses a suitable mapping to get the correct width value for a fontThanks for the fixed an mapping issue in revision which led to a regression in one our test cases +You can assign this back to me if you like +James Cobb +If the problem describe here still exists please reopen the issue +However if you use a hyphen in any of the parameters to create script it will generate this error +Configure script now checks for Bundler availability to allow ci servers to build successfully +Its really my fault +The problem is that any IPC call returns only +Wouldnt it be more approproate to add the classifier as a element in the pom just like groupId artifactId and version? It should apply to all types of artifacts and not just jars +Only thing left to do here is to document it +Closing it as a dup. +I see +for the patch +Hi GavinThe problems are fixed +Id appreciate it if the others were deleted +It should be to of screenshot +Hang i see the snapshot you attached is not the default +Patch build isnt runningNew patch needs to be attached +again if I grab the CVS version it works the version I built myselfapplying the patch to beta still seemed broken but if I go with the CVS Iagree that this is workingThanks again for all your time and help with this. +Thanks a lot +If the caller is trying to commit by closing thats really the problem +If its default null in the migration script then it should be default null in the create script +Any plans for committing the common jar to mapreducehdfs rightaway? Or wait till some other patch does this +I tried on latest build no this error pls check it +We would like to see the behaviour of presious versions of reinstated which is that it only warns the user +I think you need to have Aileen get me another hard drive then for my MacD I can get a copy of through my MSDN subscription but not sure I have enough room thing you can do is add some logging to the test that are failing to dump each frame the test receives so we can get an idea of what is coming back from the Broker right now its a bit of a mystery +Can you please provide more information? There are some test cases which are removing exploded deployments so it might be something different +While the spec says that files should not have this issue most readers dont have a problem rendering the file because they use the information from the xref table to figure out which objects to use +It can also allow commits but we have the commit buttons so that is not a good option any +At the same time if the follower reacts fast the learner handler will update tick it sooner after the sync up phase +Without a real unit test all I can say is that the code in looks different now and null values are cached +First patch renames ringid in peers to hostid because thats what we call it everywhere else including nodetool so we should be consistentSecond patch updates the hostid in TMD when it updates the tokens +lgtm +seems thatthe bug has disapeared +I just need to do a little more and then Ill post the patch. +It worked fineThanks is related to with duplicates definitely +I committed this +i hate binary order in general to be honestBut binary order in this case is code point order +Shall we postpone the work on this issue to after + for trunk +Fixed in link to API docs +I take it you were hoping to see additional statements in the log +This patch provides a simple mechanism for adding the license files until the more flexible approach described in can be implemented +we believe composition would be better than inheritance in this instance +Ticket is good to go +The main purpose of this chapter is to tell you about +I agree that the source tarball looks wrong every single file is undersrcWhereas there should be NL etc under +bq +Fixed a small problem with the initial code committed against this task in commit aaccdedbffdb +Hiare you saying the soap fault from a service is not interpreted by the client? The soap fault in this case should be interpreted fully by the clientCould it be that your case is about a oneway serviceRegards aki +Ah! I see +I will try it a lot its in +So Ive removed this patch until I can address that +The exact timing of that move I do not know +which in turn is run in building process engine and before even starting process instance +Committed +Duplicates duplicate. +Thanks Eugene for the contribution +screenshot with beta and snapshot versions on the update +So if everything is ok it shows them both if not just the version +I really hope this project will make Maven publication part of the xerces release process +Ill watch the other issue +With my limited knowledge the patch looks very good to me! Ive applied it and Hamlet for your effort +changed to be an improvement rather than had looked into implementing this and its not exactly trivial to do due to the way that a vm connection can cause an implicit broker to be created +and I ran the same test +It doesnt need to be too obvious +But another user can grant extra rights or revoke them +Not in all cases I think it somehow depends on how you initially specify the dataDir Im seeing this exposed in a test after working on +Lets discuss that offline +Works without problems under +This is very probably fixed by recent changes +Actually the second patch seems to have some whitespace changes that dont seem warranted +Ben can you remember what the final decision was on this +Please open an issue for that in the MPIR JIRA projectThe warning from the Surefire Report Plugin might be because there is no test source to generate the JXR forThe distribution management is still a requirement +EAR child recognition is commited +Thanks for the patch Bryan +You specify the logging level of the by specifying it on the configuration element which is documentedYes we need a section just on the system properties and also that they can be loaded from a file +Im guessing it will take me a week or two to find the time to to make these changesIll also document the feature once it makes it into the source tree +Added for part +I am fine moving current schemes from configuration file to based mechanismWe should also retain the configuration based mechanism +We just decided to revert instead +Closing issues for beta . +Ive run into issues with this in etc +Merged changes from trunk to +Which sounds very much like type mapping something that eventually Rob co should provide support for with the communityapplication schema +I had a small merge with the changes so Ill do one more clean build and test run before committing +Looked into this and I dont think synchronization changes are required +That is the result returned from native the empty string +I tested after your checkins and found that it would nor work correctly when buildr was called with generate +the fix could maybe be better +I also have a need for multiple +Very cool +Though we dont have anyone to work on the tooling at the moment +Hmm +Fixed on trunk based on Bills suggestion for. +According to Charles Crouch the version we need for those issues wont be out in time +Many thanksBest regards by Andrew. +Do you have a stack trace for those messages +Work done for issue will be finalized when preparing. +I am going to assume my first results where invalid +On windows When run j cprogramme filesJavabin there is an parse errorLexical Error at line Encountered P after But on linux +I think it might be useful to have it in the secondary master as well +On cleaning RPC layer +Can no longer reproduce this +We have to start Eclipse when some runtime paths already existCommitted to the branch too +Feedbackis Sam thank you for the improvement +Its not only used by UID +after reviewing this issue it looks like group is the right thing to use +this code should be ready to replace the old base adapter +I wish this patch had updated +I rephrased the sentence slightly hopefully making it easier to understand +I vaguely remember this being a conscious decision at some point but maybe I made that up +I used sudo and now it works normally. +an alternate is cleaner and cheaper but may have side effects if subsequent plugins expected a directory +So what I believe is that by having a set of generic state machines we can delegate the work to the state machines which makes the generated code much simpler +Thanks for finding this out +It needs to be installer and then the bundle for needs to be refreshed +Ill open a discussion for this on the dev and maybe even the user list and well see where it goes +Prepping +Locked toNeither of these are handled by the pluginManagement section of Maven so the latest version available was used when Cargo was built +resolved by PR for with JBDS +I applied the patch and verified that the Webdav functional test still works properly +A high level design doc would help reviewers understand this feature +Ive just committed this trivial patch. +It would also help if the directory name were more obvious like but like you mentioned youll see if the location can be configured +fixed in trunk +BTW I agree with Hong I dont see why this in the test at all +Also I might add that this patch passes all the tests +not enough support for this + +Peter could you please care for merging the patch inIts minimal but I need it +Thanks Awdesh Singh Parihar for patch and testing stepsTesting Scenario for AP Go to Url httpslocalhostapcontrolfindInvoices Search invoice by clicking on find button +Instead I have copied your classes into a new Eclipse Project and set all dependencies manually +Works well Fabrice but looks like there is a side effect +Thanks Narayanan for the fix +Seems like here is the same problem as in so the similar fix as for should be made hereIll provide the patch soon +If it is a bug in jaxen its probably in the domj navigator but Im not sure +so you now depend on +If this is still a problem with the new code reopen this issue +I addressed all the comments except for using variable substitution for logging in +Ive been out of the office for some time but I can confirm that this is now fixed thanks +For example Amazon S correctly implements HTTP while Google Cloud Storage via the S API does not. +has some code that is hard to reproduce in JBC +Please review and let me know what you would say please go ahead and apply the patch +What do we do in this case? +I too tried logging outin and that didnt help +fixed by as is already released months. +I didnt want to change the public API so I copied it privately instead +This is now fixed in CVS. +as you want to extend you dont need to change from static inner class to inner class +I dont know if its officially a design pattern or what the name of it is but its easy to implement +Roy is there a list of licenses that are compatible with the apache license? Just to make sure we dont spend too much time in the future for extensions that we cannot include +If server has an alias it should be resolved to actual host name so that the KDC doesnt complain +This issue still occurs after updating to the snapshot +Sonar should load extensions for all plugin in order to determine which extensions should be executed +Youre correct +Only the operators that were submitted with the earlier patch may need another glance +Therefore +Sorry no update just yet +Closing old issues +only the second issue is in fact resolved +I suppose it is probably related with the fact that we only run using we need to understand if this is really a memory leak on our side or a transient jdk behaviour on the particular to pinpoint with the help of you testsuite memory checks the exact we can schedule the fix in the linked JBAS issue +test is attached to in svn +It looks like a flash player issue as its possible to reproduce it directly in the player needs to be dug for the feedbackSo how should we proceed is there something that we can be done or does it seems like a bug in the player and needs to be fixed there only +This operation is effectively doing what register does in case of overriding an existing definition just that it doesnt register a new it with careJuergen +Thanks Jing! +I believe the fix also address the server does not send any piggybacking update and improve the performance concern since we are getting the right isolation level +I hope to see no problems in using Caliper for I am having a hard time hunting down a maven dep for caliper +How is this looking for CR +Should be available soon +Do you have exit codes? But TRUNK works correctly with your fix? works well +This time granting license to patch applies perfectly and all tests pass +downloaded patch and tested examples now work for me +Patch applied to both trunk and branch +This problem reproduces cleanly against after iterations but with the current trunk code I was able to run two million iterations so it would seem that the bug is fixed +I am assuming we want to hit code like +Not ready for commit no testsdebugging done +All right I will submit a patch +There are a number of Tika issues filed that relate to this +ArunNo disagreement expect for the fact that the current implementation of passing on the standard deps for applications is suboptimal since it can not be fully parameterized +Attachment has been added with description Patch to fix the made a change to so that there is only one idea of a documentsURI irrespective of whether it was set by set or set +Great! Id love to see how CR performs perf patches were being submitted right up until the moment CR was cut +The private inner class is not used in any publicprotected method and the equals method is never used for this classOtoh the patch looks ok so either we take it or just close this issue as Wont Fix. +I will get back to it soon +this is due to not being found probably since it has generic type information +One you post your review comments it can be addressed in another Jira +This patch removes the pass statement from interface methods as its no longer bad +As for the Findbugs this is stuck with the file tracking the expected findbugs warnings not being updated +added to the release notes as resolved RSS feeds were returning a Code HTTP error instead of the expected RSS feed. +Conditions are always similar UI define a job repo is Web output is Solr +I removed the debug tag Patrick is referring to since +must have been fixed a while ago +Could anyone help with this +I am talking about listMessages channelID clientID and name +Actually there seems to be a certain digit pattern in every of those numbers with a leading as first non zero digit my personal guess is the workaround fixfor this issue is to break this pattern one way or the other and split it into two separate values which then should be readded on native double level +the codes according to Raghus commentsThanks Raghu and Dhruba for reviewing +A thread stopped in a finally clause run during sorts has been for a long time +While this may by fine in other peoples cases it was causing the path to not include mvn or java since I dont have these defined for the root user +Committed revision +The logos are of fine resolution that print well on tabloid x and bannersIf you need graphics for sizes I can provide those alsoThe is my professional description of the trademark information for review. +Have a draft +if it finds that the version is actually not changed it should just continue talking to other nodes in the shard and not invalidate the cache +Is also subject to the stdoutstderr handling bug +once these are addressed +Resolved in close of all resolved issues in preparation for release. +Also I defused a kind of race condition by providing an actual +Updated JTA Programmers Guide to make it clear how the shipped recovery instances work but also that they are intended primarily for. +As the logic is also related to the Tracer etc +bulk defer minortrivial issues to +So we strongly discourage you to use final classes or methods as facts specially if those methods are property accessors +Bulk closing stale resolved issues +I agree that this could be somewhat confusing +Hi Elijahpatch applied see r for more detailsI just adjusted minor signatures to get advantage from generics and avoid string concatenation but the rest is OK Thanks for contributing! +What do you think +I suppose let me keep the official version of the patch for now since I need to make some fixes +dhruba Ive been running all of my tests with these patches for the last several weeks and its been very stable +bulk close of resolved issues in. +Only have one question in get do we need null check if Initial always generate legitimate bag? Is it just for bullet proof +Hi JacquesI can answer this one the line was most likely uncommented because the message string declared above the line was showing as unused by Eclipse so to fix the warning the log line was uncommented +Anyone +Added tests for job history parsing to verify each task attempts shuffle port value is the same and being written and read from the job history just committed this +However it is pretty weird that I cannot mark this issue as patch availableYeah sorry working with asfinfra to fix the new YARN jira workflow +Awaiting feedback +Having a Transceiver and Server would be useful and Harrys patch appears to provide a complete implementation +the reason i think CG should implement icolumn is so that we can port the rest of the code to using it and drop CFSC entirelyAhh +But since it is no more a requirement for Hama I tend to write a Database +Todd thanks for taking a look +just committed this +would be good if you could reduce this code to an actual simple testcase that just shows the error +Please let me know if you would like me to change something +But will you have different sets of jaxb XML binding java classes for each XSD? if so how will you handle that from the WFGEN easily +This issue implements part of +With Rob being pulled in many different directions Im going to push this to LATER +It is a member of which extends which is Serializable so the problem persistsAdmittedly we are running expanded but changing that would seem to be a side issue +Applied thanks! +JCA RequestResponse enhancements +Patch for performance using valueOf can be times faster because the values can be cached by the compiler and JVM +I have just checked and the +Attaching and with the latest comments incorporated +If we have to we can print the summary at the end +This is a blocker issue for me and it seems to be unrelated to the amount of available disk space +Ultimately the next release of Felix will then use the correctly compiled artifacts and trunk will use the correct snapshots for now +The way to check the process alive on windows is using with the wmic command so there might be a problem +Aaron Kimball actually made this change on my computer while we were together Im just committing it +Patch v incorporates Andys feedback about using the help from and I saw pass against hadoop am also testing the v patch now with and also applied +Instead of doing string substitution the patch should probably only substitute the namespace field specifically +Verified in +Committed to trunk rRequest no risk to however its not fixing a defect either so severity is low and it shouldnt hold anything up +Closed. +But thats an implementation detail. +We do not need to worry about the type of the response +The hard part was coming up with a proper estimate for the number of segments that should be owned by each node +Rather than putting in a hack it should define the same JMX operations otherwise random things could stop working +Hello Robert as it seems this patch hasnt been committed yet I wonder if there is anything I should do or may help withIf so please let me know that +Patch committed. +verified on +grails and +Thank you + +I might actually have two different things here +What is? Sounds more like a method name +The fix was committed at the revision +Ill revert then and we can rethink this +The cordova version appears at the top of the file +I suppose I could copy the info into a holding structure but that seems like a waste +Scheduled for should be easy to add as an option for windows installs that I think people would like +Bruce does postgreSQL have support for sequences +Should we open again thenregards they are seperate and is still present +Yes ctime will prevent from previous versions connecting to NN started with upgradeCan we add a condition to allow only if there are no in progressThat is one should first finalize or rollback the then do rolling upgradesThe scenario that scares me is if somebody does a snapshot then several rolling upgrades and then decides to rollback + +Bobby Im still using the response which I think is reasonable +I have fixed the issue by adding touchpoint instructions to the product feature +Putting another repository to Karaf with only one artifact is mistake +The junit package is for classes associated with running testsutilitly assert methods and decoratorsA better approach may be to make the methods part of the same class as the actual test that uses them or even a separate but related class +If we feel we need to add this to that can be done after is tagged +currently I made some code review and I made some small made model structure more clear and I format code in one think this is not a big modification Ill commit ok be ok? cause it could look like a big change but really it is not so big +complete but for test +Merged to doc branch at revision . +Release Note for When constraints were changed tables were not dropped correctly causing errors to be thrown +hot deployment and instant reloading of mapping files is getting more and more important +This patch needs to be rebased against trunk +Passed all local tests +Those are referenced by the links through their PDF object number +This has already been fixed +Note that still works as before in javascript currentNode is just a convenienceIn fact is also convenience and of the kind I dont like because it is JS only +r +might be a bit more efficient +Ive got into the habit and my IDE nags me if I dont +Related to +Thanks SaschaYour changes are in trunk at revision +Please apply +Ive added synchronized modifier to methods of because queuedCounter is assessed concurrently from several thread +Changed first time Ant task +Please check that the bug is fixed for you +TomI am happy to help out how I can +No more pesky for spotting +I think I have a simple solution for thisCopy the InRename the goal from emma to reportRemove the executeDoes this look reasonable +Similar to but the properties can change dynamically +This issue is still open +Certainly the ideal answer is yes but the practicality of achieving that is the the ejb layer and har deployer should be tied to the same hibernate version when bundled in jbossas + for the to trunk. +In fact our central use case for Hadoop is analysis of massive structured logfilesif you guys are in the Bay Area maybe you can swing by Palo Alto for a demo? Also were headed to Santa Clara to talk with the Pig team in two weeks if that works better +yep I think there is also something open for this +If you have more questions please post those toI am so sorry for making a mistake while reporting the bug there are slight change there please look into that again. + see issue for reporting this +because of the ordering guarantees if you issue an asynchronous sync the next call whether synchronous or asynchronous will see the updated state +Sounds good Andrew +It just would have been nice to either commit them at the same time or give a heads up on the mailing list that you were going to break the build for a little while. +Perhaps afile encoding issue? +Needs review +Is there something I can give you to help you help me avoid waiting minutes for a already available jar? Is there a log file that might be of help? +After that they will be able to log in +This problem still seems to exist +move issues to and +No Sonar should be okayThis is exactly the same error as before +A under the branch that perform tests million messages exchange with size of K and KAs of yet only Mina client and server are implemented Netty to come thanksIll add the module in the project +All patches submitted and committed +Is this really a critical bug? +Ended up having to move to the +It would also be nice if user can override it +The code that is in SVN now handles trace level logging for logj and. +Is that necessary? Should we flush before compacting when user triggers a minor or major compaction +closing issues associated with existing releases +jar files from SVNWe could put a file there telling peoplewhat JAR files they should download and copy thereOr at least where to look for that information on ourweb page +OTCP will give the required abstraction for the Internal IPPublic IP and Encrypt this also provides abstraction for streaming to happen via the private for aws users +Feel free to if you encounter this issue with a jar we built +jdo and +The is attached. +Uploaded a patch to +So in summary my question is how can we do modify verifyTrust to validate certificate used to sign? To fix this issue we did updated verifyTrust as follows Take the exponentmodules of certificate used to sign and take the exponentmodule of client key sitting in service side and compare +Closing all issues +I just committed this. +Closed after release. +Thanks Ken I was planning to look at this today but got delayed by some Swig Python problems +Removes rest from binhbaseAlso removes srcwebapps patch available +The overcommit ratios are dynamically plugged into the capacity calculations +There is an html parser in plugin +You are talking about right +A conforming parser will look for the last EOF and simply skip the content after it +Some corrections for proposed by Tim Januarios implementation looks solid +Proposed s fine with me +All the fixes such as this will break your tests but they should get the credentials from the config file +There are too many other higher priority issues +In that case will be recognized as a suitable update for +FWIW Im using +Will have to look at it more tomorrow +Are we sure this is fixed? I cannot get the directory in srcmainwebapp to be excluded. +Vassil says Using views is not a very clean approach because it mixes up the presentation xml and the controller code +This is not really needed +in M +Weve been discussing this for over a year and it looks like were finally fixing it +Sorry about thatThen after only I will start the patch +It will still be the cause of problems when you have defined additional types in your own filesIn one of my previous jobs we had defined some extra types and it was this setting that caused a lot of problems with the usage of especially when new types were added +Same thing under HSQLDB +Actually looking more into i see that topicOwner ops are in now also +it shouldnt be calling unlock unless it has reason to think there is a stale lock that needs to be closed ie this is a bug in that you have only discovered because this bug was fixedI would suggest a new issue for tracking and a patch in which doesnt call unlock unless isLocked returns true +I also think that it is safer to postpone the persistent lease feature for the next releaseAs it could be seen Dhruba tried several approaches to fix the issue but all of them turned out to be problematic in one aspect or another +Quite nasty +must take a closer look +This is a language design issue +Resolved now. +Yeah I committed the patch a few weeks ago +Hi all feel free to test it! Im curious to hear about your numbers +We could name change to insertArray? +Itd be great if you dupe too when you resolve this one + +Moved to doable in attempt patch is added +Cool closed. +While I do not see any immediate way to make this work on JMX there may be ways to achieve the same for the time being since there does not seem to be a viable way to make this work for JMX +Its slightly more complicated than that because we just started to support encryption so theres some metadata at the beginning of the fileAccumulo uses a simpler file format to avoid the check for an empty file that occurs in +I think for this we need to look at the underlying error generated by the Aries application code andor Felix OBR itself +I dont quite remember anymore but I suspect it wasI do note that does not work for enums in different class loadersShould it +Please see for my version of this fix +Could it be related to class unloading +Im about sure that the syntax Chris put in there wont work but I havent tested it so it may be that this feature slipped by me unawaresThe first thing to do with this is see a more detailed error message +Was that intentional +This issue was migrated to +They are smaller and run faster +Try another event as a workaround +The modify button should be there but not sure we want to pollute the world with more presets +One more thing to add I ran the same build with Maven it works as desired +For my your solution works +Can this be marked as resolved? +PranayYes I think a new field on would do the trick +my colleague suggests actually getting rid of the gradients as well so Ill play with that too and post more +I have added documentation to the API explaining to users how to work around this +Thanks Babak +HaliThanks for the instructions I will try it today and response ASAP. +I think for sake of completeness that test should fail if no exception is thrown +See for more details +Thus we try to connect to the datastore +Sorry make that I am only trying to make it better +It may not all be relevant anymore +So my JBDS should be quite virgin yetBtw +Thanks Stefano +Bulk close for Solr +With taskstip it is lame but not that harmful +Actually the test fails with any number of threads but number of threads it suit is at it is not good idea to call to objects methods without checking it is null +I am attachinganother patch that addresses your commentsreviewYour commentsreview is appreciatedThanks KatheyThanks for your review +It turns out that join nodes are reintroducedin modify which are not involved during theflattening process which happens during preprocess time +Ill add a test case based on this and commitThanks Arjohn! +fine apache is screwed up close it +Made it a little more aggressive instead of +My build settings and error message +We might also use a NETBEANSINFO file as I already proposed above +Most look pretty good now is still hanging though +Patch for hadoop . +I forgot to copy the return value when it is not a! Fix on its way +If one wants to regenerate the index with the stored fields decompressed he could simply use the contrib module recently added +Sorry for causing the troubleI should have run findbugs before checking this in +WDYT +Added a unit test Thread handling more robust If the consumer thread dies the producing thread is interrupted as well up old if necessary. +Hadoop is not using directly as you say it is a transitive dependency via Avro +I created a new one would somebody pls try it? Thanks +Sounds similar to two other issues and +nodelete might be a confusing name +what are the implications of this? Is this a wise direction to go in +The list of implemented api and list of introduced code improvements with desired benefits would be nice +I therefore decided to revert back to previous version of till I had time to look into the problem in details +Since is rolling back I had to change the unit test cases for +Linking the issues for future reference + +This functionality has been completed and the issue can be closed for now. +yyang did the work for Upgrade JSF impl version to +Heres another Looks like I have one left to hunt down +If you actually do need to delete just a given qualified column name on every row in your table you should write a mapreduce job or some other highly parallel way to do that +If it is not required to change the password after using Email password link then it shouldnt be also after generating a random passwordStill it is not a big deal since it is Bilgin for review +Can you provide us with that pdf +and ONLY on Windows +resolving cause I resolved. +rename variable err log and fallback bump minor a quick update ill try to understand data block encoding readwrite internal path such that we can make this optimization benefits on more scenarios +This should be revisited once is fixed one way or another. +Libor Whenever you reproduce this can you take a few threaddumps so we can identify what component is responsible for that bugI seriously doubt its a maven component issue but you never know +Would you be so kind and check it? Thank you +A suggested patch needs detailed review from more experienced Sling patch Moved up in the package hierarchy and have the servlet extend from theElse your patch looks okWDYT +The tests may be a maintenance problem in the future and therefore you may not want to commit them +I have changed the unit test to always set that field to the same value before comparing +What kind of feedback is this issue waiting on Ollie? Feedback between the reporter and the assignee? Seriously lets sort this out for otherwise well never get to it +see agree with Otto placing tempfiles in tha same directory is not the best way +Thanks for taking care of it lads. +Victor please take a look +should I ask to get this committed to or shall we change it to just trunk +Committed r trunk r ant example the second time now comes in at sec for meThis helps big time doing testingdebugging using the server +Its just that the whole machine was slowing to a crawl due to swapping +The diff of +Also needs dep to run Weceem at least +I checked and this fix only seems to be in nightly build could the same fix beupdated and release as or something because the nightly build is failingits junit application test +it does indeed make sense to use latest +Good feature to have not urgent +I just built the guide and checked as well as greped for cloud platform and I cannot find anythingI think its not a issue +Please reopen the ticketwhen requesting production support +I think I ran into +My Bad +Does anyone know if this ever worked properly +Please attach Surefire logs +defect to the changes +Nice catch +Thanks for the patch +I confirm that it works also for SATJ +Stylesheet file +Given that my first patch was k and the latest is it seems the first was missing a bunch of things due to me not yet adding them to svn sorry about that the latest patch should be complete +Yet the Base postfix is pretty common throughout the JDK similarly to the Abstract prefixI added little more which brings some clarification when to subclass instead of AnalyzerI guess this is ready to go in though +On further testing it seems that this is only a problem if there is an auto cookie in place I havent looked at the code but if it is looking at the auto another criteria might be better +all fixed. +Attached patch adds the line Dan requested +fixing some errors in sections +my original bundle must have been bad the jar is in the repo +Whoops should be +This might be a good time to get the struts docs moved over +This will be very helpful for system administrators +Scott good point +this JIRA should be follow up by a JIRA that fixes dependencies based on the dependenciesanalyze yours thx +Patch added +The executionCount was also removed from +I ran the indexer over the CR repository so the index should be available as soon as it propagates out to +Button Save is already change +eclipsekeyring showlocation +Id assume that it isnt +Ran tests of mapreduce and hdfs with the core jar built with this patchAll tests passed except for in mapreduce and in hdfs +Deferring to to JBESBGACP with revision . +closing issues for released versions +Thank you Ivan. +essentially we need to call get and get in the same transaction this way the isolation level will guarantee that we see consistent data scross multiple queries +to Todds patch +Release includes some minor fixes which locate syntax errors more precisely and provide for partial recovery from errors +It would also involve UI work for new API +Closed as suggested by Vladimir +What I am trying to understand is the side effects of excluding a component from the Tools beta works fine when I use the update manager +In the end it might be cleaner to unify this with rather than having two mechanisms to define part of the split policy +I dont think anyones opposed +Updated has run his profiler tests and with this change he has found that there are no classloaders issues in the GC process +Would you like commit access? The stream stuff is my favorite part of swizzle and the reason I started the project but most people use the jiraconfluence stuff +Its attached to along with my subclass. +Our patch only covers JDT core and it is complex to maintain +i just committed this +This patch includesbackend for image taggingtagging is also implemented in the UI where the user can add search and remove tags from picturessearch results are now displayed in a table layoutif no field is defined in the query string the query will be executed against all these fields tag name and albumwhen the user clicks on a search result he is redirected directly to the album and the selected picture result is set as current album imageStill remaindsthe search results should display the picture miniature instead of just text it can be easily implemented later when an API for retrieving picture miniatures becomes availablethe tag remove button should be an image button something like an X +LukaszIn order to be considered the final comments in the report should I request an Enhancement for themYes you are right and I will try to move Struts archetypes inside Struts project to be a submodule and included in release process of whole projectIt isnt so obvious but there were only few archetypes released so far so it will be better to put a list of available versions of archetypesRegards first thing that Maven users not familiar with Struts do is try to create a project from an archetype using the command from the online documentationPlease make them happy +Id be great if you can attach them as it makes it easier for us to reproduce your exact issue +I dont really understand what theyre supposed to be used for and I find the implementation a bit strangeI dont think its possible to convert them in a way anyways and writing a replacement seems not really worthwhile or does someone use these +Its not very elegant and we should have created the at that time but it wasnt the case +Additionally since we have been trying to avoid copying unless necessary this means that manual patching has not been too onerous +Dont know where I took Micha from +Im not aware of any progress for this issueShould we close it as wont fix +Im continuing to investigate +We should have a mechanism to retain only the relevant constraints in a type for error reporting +it might take different statements also +New patch addressing the comment +Moving above does the trick +Yes otherwise I need to add Xss option to increase the stack size +At least not for and Rails +Please open a new ticket and describe what the code does +I think there is a valid need +So is the test invalid +Committed to trunk r +Try as I might I havent yet got a gc collection to dispose of the exception object and therefore trigger potential segfaults when the zookeeper module raises +Quantity should be a type and Money should be a type +They were generated using IKVM version +This patch hasnt been tested thoroughly the only test I made so far was to change a Party entity and verified that the Materialized View changedA few things are missingpending Insert entity data to new Materialized View upon creation Re build Materialized View upon entities modification + change is to is +Marking patch available so we hopefully get some comments on this patch +committed this to and! Thank you for your contribution +Task is in progress you could see from related issueCorrected behaviour of the subject attributes is if the value of the attached attribute is true component is attached to the component specified in the attachTo attribute or to the parent component if attachTo is not attached false components doesnt attached to any element and can be activated only via JS API +Heres a patch +Not a very big issue +Thats what one CPU can do +Christian do you know how many blacklisted were present when you noticed thisThe reason I askA is blacklisted once it has more than failed tasks however once more than of machines in your cluster are blacklisted the blacklist is ignored and the job is allowed to be executed anywhere to ensure it fails +Should be ok without a test this is a configuration change +You can change the method argument to Exception or split into two methods +However I then realized wed have to have multiple sets of code for the same functionality wed have to have some way of determining at runtime which implementation to chooseBoth problems seemed quite troubling the second problem seemed very importantbecause if the selection of the appropriate algorithm is based on information aboutthe size and distribution of the data being processed by the query then the decision oughtto be made by the optimizer which appeared like it would dramatically increase thecomplexity of this projectSo it would be great if the single implementation was good enough for the queries weexpect to runIll try to put a benchmark together and we can see what the results say and then wellhave a better idea of how big a problem we have here +bq +The name change certainly makes some sense but even as is a core that is returned closed is not very useful +I havent tested there however were seeing this in production with +Its not illegal +You made a good point +Committed at subversion revision +By gaining more opinions I thought to follow your suggestions so here are the new class names update to the latest CVSCheersSergio B. +yes that is expectedThe list of deserialized objects on which dirty checking needs to happen is maintained in great +Accidentally missed a change in the last patch +Because of the context at the time of assignment hasnt changed when the second assignment happens the two assignments go to the same host +Its only when deployed in my own container as a WAR that it gets into trouble +Ill look at adding that +A simple change but much appreciated +A huge thank you to Supun for this patch to an issue that Id reported long ago +Not sure what you mean by not right fix? Pl +An implementation of that handles partial keys +To summarize where we are if approve the changes in my pull request and update the package name of clientutils I think we are ready to push this into the master branchThese discussions should be filed as individual issue reports as feature requests so we can track the progress +However if appletviewer is resized along nothing happens with the graph +Together with this enhancement the display of the bundles and components has been has been implemented some time ago +Any errors with the Spring archetype on the front tab of Central +Theres lots of code thats used on both the client and the server +This is the kind of CC issue that make me love Java +I can test nowLee can you please attach to the ticket the test app that you used so that we are on the same page +To implement sliding windows we can be a little trickycounter superstep of counter counterval time of superstep of counter counter superstep of counter counterval time of superstep of counter In this way we have to use n counters to display the last n supersteps but we need not remove any countersAgreed on the MR work +Small addion to my last commentThe imports I removed from the schema are imports that are used to reference namespaces of schemas inside the same wsdl so no location is provided here +path is neededOops should be path is not needed +Thanks is irrelevant in this context of figuring out why something that takes miliseconds on both my low and highend machine is taking minute on if you open a second file? does that take agestooWhat about Eclipse WTP installed what happens then +I do not sure is this comment to this issue or may be some other from list +See +We can do whatever you want +How about adding simple input check in Path that throws an exception with a descriptive message +Its somewhere on my todo list but hasnt floated up anywhere near the top yet +In a large cluster we could have hundreds or thousands of resources to populate in the JMX property provider for a single request +Are there any implications from the standards work being done hereAs I recall there was some thinking around the coupling between views and sofas looking toward generalizations of this +Brian this was character encoding ironically it displayed properly in transifexs web console but not in the text file +Please test current CVS and let us know +Instead we should honour the input bounds and the process should transform each feature to the output CRS prior to rasterizing +patch for bad for providing the patch +Indexes and Constraints +Comment from David Jencks on rev and You are assuming that any connector thathas a called Driver uses that property for a javaclass name that is expected to be on the classpath +fieldName houseNumber notNull sizeMin sizeMax add field string class +Actually sorry this is consistent +I looked and looked but couldnt find it +It is sometimes a timing problem so you have to click more than once +This is actually way more primitive than I remembered +is the impact on now making the RS to transition the node to OPENING? Thanks SergeyNow the main RPC thread for asking the RS to open the region is doing an additional work of transitioning then node to OPENING and then calling theThis should have an impact right? But the above change is needed if the RS goes down before changing the state to PENDINGOPEN as said in +Lets move this to updated as per Anoops comment +cant reproduce now +Im guessing there is some inaccuracy in the tests +What do you think? +Just committed this +The patch looks good +Ill try to update the mapping files to the latest from pluto where I think these problems have been fixed +There is a Procedure and a step +then I will close this issue if you see this problem again feel free to reopen it. +excellent work you guys +Another patch which enforces creating corrupted old storage and image files in all cases +Anyone interested in more thorough logs +The problem was the usage of tmp dir which caused uri syntax exception +It looks like Bing has fixed this now +That might be better in general if you rely on always accessing the raw value because in case of a binding error will always return the rejected String value even if the underlying property is a Long or a guess the consequence is Use only if you can accept a String value too be it rendered by a custom or a rejected user input String +Zip file containing test portlets and a test case this hits a page called mark which needs to be set up beforehand with the portlets present on it +Was meant for +rundebug project on ide On command line type mvn true gwtcompile datanucleusenhance package reload web server and now it work +implemented comments +for to trunk. +I tested my patch against both on a fully distributed node setup not with unit tests +Spent some time again investigating the issue +Im awaiting more feedback before I commit it +Were the old election messages resent when the Fresh server was contactable? I would have thought that election messages should be based on the current state and never send old state +hmm interesting point +The original patch could not be ported directly because it sets the version id to rather than +Yes that property can only be set when the deployment is created via the template +Why was this issue cloned +I also updates the as generator for this same chance we could apply this patch to the branch also? Currently were using a forked version of it which incorporates this patch but Id rather not have to if at all possible +there isnt a solution which fits all requirements ill introduce a new spi +It works with fully qualified name for host in etchostsThank you very much +Adam are you still having problem with this one? Have you tried just setting the mail body as a simple String object? You can still have all the mail headers set as plain headers on the messageSee the unit tests in Claus Thanks for feedback on that one +The zookeeper bugs above are filed for the same issue on the zookeeper server +Hi Jacquesafter that has been committed we can close this issue because it seems to me that now everything will be cleanedDo you think is it not trueThanks JacquesI have recreate the patch from svn client instead of Eclipse hoping it is working nowThanks JacquesIm uploading a new patch with the remaining labelsCan you try to commit itThanks a lot for your patient a lot for all your work MarcoYour last patch is in trunk revision +kinda hard to review because of all the side effects of passing the HTD but Stack walked me through +In this file on and RHEL. +Attached a pdf file that loads and contains proper content and is accessible for example throught Adobe Reader but not with +it test added rThanks +Thanks +TomI was thinking that this would be good for Amos to investigate +Didnt realize they werent in hdfs anymore kudos to whoever moved them +Deploy the attached web application and view it at httplocalhostroletestThen click the links on that page and note that the link to httplocalhostroletestsecureadminonly does not cause a that changing the to match the causes security to be applied. +Patch available +Should such tracking be added to instead? implements +This does not definitively disqualify the Birt component but shows another reason to put it aside and improve it +The jira is not reproducibleI used Windows x MSVC debugrelease build +Getting this issue right counts on we mark this for Lucene as well +Set this issue to you DaisyFeel free to reset the status to new and the owner back to issues if your team do not plan to address this issue at allI made this issue dependent of issue our style refactoring as the style handling should be refactored in the lower layers firstThis issue was set to criticial by the submitterStill it is only a missing feature of convenience of using the Table DOC APIA development priority is set in relation of our other prioritiesIf the issue is criticial to you Ravisankar +great +patch that forces failure for multivalued fields along with an example of how i think we should call this behavior change out in this uncovered a lot of tests that were missusing fields declared as multivalued and are still failing in this patch largely because im so confused about how they work its not even clear to me what file they are using to try and fix it +can you provide an example of this pleaseIve just created tests for this and in both cases the threw will attach the test I created +Thanks +Attaching +So I report it as a new issue and link the original one to itGood luckLeo +I updated the status of this JIRAIts a new added feature we normally dont pack port this kind of patch +please lower the priority since I think i get to prevent this condition from reaching agree this is incorrect but disagree its a is the staging plugin trying to read it as a directory? Shouldnt it be interpreting that from the directory listing +Other than the EE application client and IIOP remote invocation is largely unspecified +Its something to consider separately from the main conversation +Thanks +Fix in? We need help here +No passivation +If this fails then we will probably need to implement the treat large maxOccurs as unbounded hack for the time being +HiWe are not using mechanically generated classes as a mapped object but these are legacy property names for a system being migrated to hibI believe the fix I posted with my original comment will allow the to be used anywhere in a field name +Seems to be working fine for me +Node hotness derived from edge profile is enough to avoid inlining on hot pathsor remove this option completely and simplify inlined method analysis use direct pass over all nodes instead of dominator treeAny votes +I just wanted to add that Im seeing the same behaviour under Maven +using xerces artifacts is the old way and using is the new and preferred way +resolving also resolves the remaining concerns for this issue +It is just how we are going to implement properties +Oh you already committed this +Thanks Suresh! +Yes I think this can make it +class inside a jar file +I suspect this will improve socket resuse should limit slow start but good to checkMeant fs get here since were caching sockets on reads and not writes +Many fixes went in to the ID constraint support between and so its quite possible that this problem was fixed in that +Lars I believe the port to and are likely identical and nearly trivial and I was intending on doing it +Hi Willafter rereading your issue and think about it I am no longer sure that this is really a problem of in general +contains update to ask to save when closing a tab some refactoringApply the patch to trunkmodules +bulk close of resolved issues. +patch was broken by a hadoop update about to fix +issue is described well cant replicate here by default all users logged in should be able to delete all their own comments from any issue created by them or notLeave it with us +This also needs to be fixed for but I need to get the fix out first +Thank you Rick for making this tool +We still need releases of both but patching is done +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch. + will commit after tests! Thanks Namit! +I personally use to generate and manage my Erlang cookies +This is not a regression rather a bug that has been there for a whileThis bug is due to the address resolution code not issuing a after it creates the queueThe above address works as expected for consumers bcos to the registerConsumer method in invokes the bindQueue method for both addressing and binding URL based destinations therefore compensating for the above mistake +Cannot reproduce it under Linux +Jason assigning to you for triaging +Ive made the about graphic for the IDE about screen +so I will make the change you request +Looking forward to to this new compiler work +Assigning to Darrin to fix the this received feedback from Tom F +Its scheduled for fix version. +There doesnt appear to be a consensus just yet +On which version do you have the issue +Added a comment for the hash function +Yes I have the source set up +QE please ThomasIm back from vacation I will be back to you as soon as I can +Experienced this issue on as well +Just my two centsAnil Ill try and take a peek at this this weekend +Stupid question but where can I find this wiki? I would like to link to it from our website and put a test plan for our upcoming release in it +I suspect that will be too intrusive +bat version as well +src +What would that break? Isnt a Being able to do scmupdate whatever the pom contains sounds closer to the maven philosophy to me +generation should be deterministic +Would you like me to create a formal code review or is this enoughAlso I may be adding some more online schema change monkeys but Ill file a separate jira for that +There are actually two kml files in the zip and +No it can be fixed as part of this ticket +No other client classes are affected +dims +I picked the ones that I seemed to like best out of the ones that were provided +Also be sure to set the new property to the JNDI namespace where the instance is bound +Attached a patch +Hi LarsIf possible can you please have a look at and comment on that JIRA whether it is valid or not +I can correctly execute tests using view ids with complex paths with all phases being executed +Is this intended just to show what would be needed is it a typo or is the intention to backport client adapter support to? The document does not talk about port numbers or protocol detection +Cloudbees account created and users added for the team. +Hi Andrew Craig +Not a torquebox issue +good point +Reopening issue until comments and test are both addressed just to have it noticed +Note by default deep serialization for json documents is disabled as it can have unwanted side effects +Patch that adds the errorPosition attribute to the and implements top and bottom positions for xhtml and cssxthml templates +Im strongly in favor of a whitelist +We need to document it but the description right now is simple +Deferred to defer of unresolved bugs to +Committed revision in trunkThanks for the patch Kazuomi great work. +Leo patch applied at r thanks a lot please verify +cTAKES distribution had a single testdata top level directory sibling to bin +ChrisIn your last patch you forgot my advice aboutp +No near plans for it +I have also cleaned up some code in and which had some unused imports and variables +These classes are so small and simple I dont mind a duplicate set +Refactored +This is a basic feature that demands further consideration +pom file which would have pulled in the dependency automatically +Reverting this change in both trunk and. +Hi all Ive also been seeing some issues with long compile times +theres a bug in wagon so this needs to be correspondig IT indicates that this is in principle also fixed in as expected from the SVN history +It is convincing +By the way I am planning a serious rewrite ofdocumentation using wiki as a scaffolding +Patch works for me thanks Doron +Narayanan feel free to close this now +pushing off to as this is a fairly minor defect +Dupe. +Folks I am working on clean up of internal exception handling structure +If you find further issues please create a new jira and link it to this one. +Please can you attach a complete FO file for analysis +Cannot make it work so far +All the regression tests ran cleanly on SolarisThere were two failures on Windows +Nick can you please attach a file which holds table creation statements for theobjects in use +I guess we could base it on some fraction of the avg row size per CF +patch applied at r thanks for your contribution +Please pull back if you disagree +pictures for uDig same behaviour can be reproduced for uDigIn addition another problem exists labels are on top of the legendSee pictures attached. +Added link to Dev Forum post regarding this issue +I can upload a image with eDirectory installed somewhere if you want to test it? +There are some cases where there is a tentative resolution and you dont want to close the case because you want the reporter or someone to review and test the fix to make sure it was what was intended +That will make scripts a bit simpler and make maintenance easier +Shouldnt this case be closed +Flavio this issue would be fixed in +Im happy for you to update the JAXR project Thomas but it will require an EAP update +Is there demand for also a release with this fix +It seems to be a sun thing +In that case user is not left with blank workflow panal +Has fixed the NPE on our startup good news as it means we can move back to trunk versionLet me know if I should close the issue or if youd rather leave open until youve had a chance to investigate correct behaviour +foo bar +Bulk closing stale resolved issues +How can the optimum buffer size be deduced? Is it a factor of hard disk sector size? Memory page size? Lucene write behavior pattern? Level X Cache size +Or close this issue as diff +Also fix the same issue of endpoint +I would rather avoid patching existing and published jar +Would be great if you could create a unit test to demonstrate what you are seeing wont have time myself to try to reproduce this for a bit +What were the issues +Removing fix version and patch label as there is no patch yet and the issue is not yet fixedCould you please attach a java reproduction for the issue +This isnt needed for the battery plugin +Or you can login into ssvm take a look at agent log +I guess this is one is addressed by your patch on Rick +thanks Edell +I think this is just the normal log messages for a client closing down its connection +Will commit if tests pass +Adding a sources JAR to the binary distribution would add another in size not a good idea IMO +thread dumps of slow is the minimal revert +I removed all occurrences of or lowercase since this restriction is not valid any more when providing names for domains or applications +Please can see just one set of packages in installation instructions and this set is full +This is a slightly modified file for a junit file in the JAMES server sources +It returns the value after jsessionid +Committed revision +Cdevjbossrulessrcmainjavaorgdroolslang Cdevjbossrulessrctestjavaorgdroolsxml Cdevjbossrulessrctestjavaorgdroolsxml Cdevjbossrulessrctestresourcesorgdroolsxml Cdevjbossrulessrctestresourcesorgdroolsxml Sending devjbossrulessrcmainjavaorgdroolslang Sending devjbossrulessrctestjavaorgdroolsxml Sending devjbossrulessrctestjavaorgdroolsxml Sending devjbossrulessrctestresourcesorgdroolsxml Sending devjbossrulessrctestresourcesorgdroolsxml Transmitting file data +I checked and Ifound too +NOTE this source code is based on an early draft version of JSR andnot intended for product implementations +We are closing down on and want to keep its API stableAlso as API design naming classes whatnot is hard its good to have months for settling down on this new fluent builder for the next minor release eg instead of doing a last minute commit and then break the API in next release +I have reopened the issue +Modified the testcase to handle this scenario +Closing all resolved tickets from or older +Do you confirm that youve recently changed the configuration to use SSL +Ill try to use that patch too if its ok +In the meantime I have also recognized some failing tests on derby but there I have to investigate further as I embedded derby instead of using a derby server +Thanks for the patch Mike +However there is an error I get in the module +unfortunately gets a Hunk failure at the problem is with +I think Im confusing myself with namespaces and unions but Im not sure how +I agree the messages are too long and printing them every heartbeat is wasteful +Isnt the best time now +Triaged for. +Can I work around it by running the job as the user mapred? From the bug report that should workRaj +Do you have the same problem on different OS JDK +Your case would have to be strong because while anyone can offer a patch a relatively small number of people are authorized to commit changes and theyd have to be convinced that the change is a good ideaPersonally I would not offer such a patch +Now marking as resolved since the fixes have been applied to all branches. +Check the GRAILSHOME variable +Thanks Milind! +Please revert this Could you please reverse this + + +If we started to include more specific utilities like the one youre suggesting we would end up having to offer any permutation of technologies around +This adds signature matching tests for DECIMAL DATE TIME and TIMESTAMP types +Im closing this since the patch has already been committed +Mark as resolved based on last comment +Demonstration of bug in junit with maven +I think it is good to scale with the amount of memory given to +Thanks lads +Now that work is well underway with AS all previous community releases are +No problem I will also add changes to mapred +Hey Scott Latest patch looks good to me +I will play around some more with it but I dont think there is anything I can must have changed the persistence provider to Hibernate as the script uses +For the remaining issues and functionality new issues should be logged +Im personally very excited about this improvement to Lucene +I think this type of confusion can be prevented by generating a log entry stating that a Typed annotation is implicitly applied or something like that +Verified in dist built from SVN . +HiIn other words you want a junit testcase using template in srcbugs +Andrus do you want to keep it +I have slightly fixed assert messages and made the test independent from the Java implementation +to talk to Mic Neale about how best to handle it +If only this part of the key had been named version instead +sounds like a good default +Closing per Bryants approval to do so. +Not upstream. +acl rule can be adding to the user created acl list +the most important documents are where i try to explain my proceeding and the definition of the bug is fixed in the current svn attached the test case in AxisJava SVN Revision +Ill just add itThank you! Its important to note that we dont need to add the optimizations for this jira which should focus on fixing the protocol +whats the approach you are planning to take +Thanks for investigating Dave +Pardon me I mean it would be neat if Avro worked with ugly strings +Let me look over your patch more closely +While Im happy to confine the ClouderaYahoo catfight to things like logos and keeping it out of engineering decisions could we just evaluate the logo on its merits? The yellow guy is owned by apache Cloudera also uses it in a bunch of marketing materials so what? The main hadoop logo is all over the front of Clouderas webpage right now I just checked and its not cramping my style +heres a patch i also fixed not to reverse the query string but instead the actual automatonThis way it has no problems with what is escaped and what isntadditionally its tests needed some revamping +This is patch containing a new sample showing callbacks across Web services +I just committed this. +Observation it appears that we never need the actual JCR lock token of a session scoped lock +I had to change the methods slightly to allow access to the generated equals methods but it works well. +Setting to resolution state before reopening to change affects. +Currently any legal identity enjoys authorization to perform the following operations i Create databases +I dont know how to kick the build bot to make it do the right thing for different versions +ESB Editor Features +I dont follow the reasoning +Ill check and merge Daniels is the GA release date +Please upload a new patch merged with latest trunk and rerun hudson tests manually +Ill run some tests to see what effect it has to remove one of them +tentatively evaluate for approach for using for caching a Portal Page based on the state of just the CMS Window will technically not workThis assumes only a single CMS is present on the Page but in reality there could be multiple Windows for other Portlets including other CMS Windows on the same value for such a Page is a nutshell its not possible to cache dynamic content and produce consistent for binary resources served up by the CMS is of course feasible and fixed in an earlier task referenced on this issue +Will submit it soon +This patch changes the signature removing the offset and size +Heres a patch that adds three new events for locale change theme change timezone! applied +version of that applies to branch +Hmm Im not sure if this is the correct approach since it will only give you problems that occur during parsing +Updated patch is cleaner and almost ready to commit interfaces are more clean now and most javadocs is in place +or initializing database for sequence generator testing +this is my public user name do you want +Alexei this issue is very similar to +Are there plans to integrate the SAML extension into the base Spring Security +previous file was out of date attach new one +Probably will release another version so displays more information on the active threads +rtf files +Oo +Hibernate suffers from this problem +Maven users are unaffected in terms of acquiring these extra jars but they still need to be aware of the extra needed dependencies +However its pretty well integrated with the DFS API and only replaces one method on the filesystem class +I just did it at the project level for jaxrsbasichttps specifically +Having read the comments I think that this substance of this issue is in the way that shade interacts with multimodule projects +Sian has converted the zip into patch format under. +Yeah I think thats ultimate what we want to do +Verified in the +Ooops wrong key to start with new key attached by will +This makes trunk difficult to release +Seemed good enough to ll attach the files right away +Pretty much what I was thinking. +What if someone would like to take Junit instead of the version +Test passes while it should fail +I seem to be getting slightly different exceptions randomly theyre all of the constant not defined type last one I got was constant Api not defined for Object +Just tried out the latest patch works fine for me too +Applied the changes +IMHO this change is not helpful +Hi All I was also studying about how Derby works and how also query plans can be displayed in an informative way +MaxDenis are there any other components that require Datatools +Is this issue still current for Branch +Can you retest with or more ideally with and see if you are still having an issue? You can also use the emails on our profile pages if you want to reach us directly +for the patch. +Please review +I couldnt replicate this but edwin is going to try it again and the close it if is not valid any this issue now +Thanks +Apparently its a class loader issue with the overlapping singletons +Paul are you planning to provide a patch for this issue +Vincent is this the case addressed in the xar? If so I can close this now since Ive checked and thats working now +On Linux Iget the no route to host exception within seconds already +I dont really see any way round this the namespace relies on the GUI authentication mechanism elements being able to modify the list of providers being maintained by the +branch +It is adressedin the redesign currently under way +batch transition resolvedfixed bugs to closedfixed +I have attached my JUNIT Test case which reproduces it out of timesStreps to Reproduce Run it until it inserts atleaset records or until you see a Deleted files message on your console +This is a common complaint for Windows users in general so I think the best thing we can do is recommend disabling system bell globally in +Do you have a test that can reliably reproduce the problem assuming its the same problem +Using thread local variables is likely to break the resource for reliability we now have transaction support for JMS +How is this doing? Anybody running this yet +Ive learned something from your explanation +The change was saved but is not being displayed in the detail displayAlso Note The term requestName is used by both for the form target as well as a dependent drop field name +The result is that the guide category links on the menu do no get expanded + +Dont apply the patch yet +I also looked at Mikes patch and the null pointer but I cant get that one to work here +This applies in particular to memory considerations and stuff like that +So wheres the patch +Hi ClementCould you format your patches using diff so that I can apply them via patchColm +If this is happening with EAP then this might be fixed already +Hi GuillaumeMy use case is only for properties +If there are additional questions about this issue please continue the conversation on the Deltacloud mailing list dev. +np closing ticket +In general the LOB machinery has now changed so much that its getting hard to backport the fixes toUnless I hear anything I will close the issue in a few days +Some comments The class could have destroy methods that take a boolean to decide whether to destroy the process group or a single process +Add an acceptor thread responsible for accepting connections +However there is an ordering dependency where the default portal needs to be imported first and Im not familiar enough with the code to assume this +Fixed in dedcbfaaabe IT based on by Larry Shatzer in aadccfcaadc +If with applied patch for the tests will work ok I think this issue should be closed as invalid +Few were checked in which fixed it +Committed revision . +Any reasons why you hidded that interface +But even that may be a rare case if you use as your client which will route requests at the client level +Patch in testing +anyone see any objections to including this as examplesolrconfxslt assuming we +duplicate with problem if it works with you then its fineThanks +john feel free to reopen if does not work for you +patch looks goodI can reproduce the failure by setting umask to +Thank you for your improvement + +Hmmm +Thanks for the patch IvanBuilt with the fix will be in CR +jstack attached +I have not found any changes between and that would have caused this regression +Thanks MarkusTrunk Committed revision Still needs to be ported to +Thanks Zhenxiao +If it werent thread local we wouldnt need this struct at all and it would cease to exist +Not an issue and not required +Were dealing with a lot of artefacts and a lot of dependencies +However I still want to confirm that this fixes made in the patch mentioned for this issue did make into version of Grails +As I understand the proposal its not obvious to me that this change would necessarily affect users since the could theoretically remain the same both before and after this change only the underlying configs would changebq +thats awesome mike +the failing zk test +passes mvn verify etc +So that only means it has to be under connectorsjdbcconnector +Ill pay more attention next time +mvn site works correctly nowThanks +The sampler interface is changed you can add samples iterate the current list and check whether the sample would be dropped +Sorry my initial patch was getting entries in the reverse order +If the other class is precompiled we will see the class instead of the Closure and then it can work +I get a bunch of tests failing when I build the latest trunk even without applying my patch +Hi Aruba staffI guess renaming is not a big problem but the modus operandi is that you are free to do the changes and execute the commit and open an issue ticket if something go wrong or not get acceptedCheers you! +Hopefully this will not ding too many users. +What do you think +in version too +thanks Devaraj Owen for review +Sample maven Just one question why tomcat field is used to override the usual global one if Im using an embedded Tomcat serverShouldnt thus exist the embedded field as wellRegards +I committed the patch to trunk andThank you Colin for the review +This will require corresponding changes to the runtime and tck projects as well +Thanks and regardsSascha +If I specify encryptionKey and an invalid encryption without specifying an encryptionAlgorithm theres no error jdench java jar DERBYHOMEjarsinsane ijij version ij connect jdbcderbyencDBcreatetruedataEncryptiontrueencryptionKeyencryptionij Derby seems to ignore the length if the key is specifiedThe following also succeed with no error specifying the default algorithm and either the default key length or an incorrect key lengthij connect jdbcderbyencDBcreatetruedataEncryptiontrueencryptionKeyencryptionAlgorithmDESCBCij connect jdbcderbyencDBcreatetruedataEncryptiontrueencryptionKeyencryptionAlgorithmDESCBCencryptionij connect jdbcderbyencDBcreatetruedataEncryptiontrueencryptionKeyencryptionAlgorithmDESCBCencryptionOn the other hand if I specify an encryptionKey of the default length with a encryptionAlgorithm I get an errorij connect jdbcderbyencDBcreatetruedataEncryptiontrueencryptionKeyencryptionAlgorithmAESCBCERROR XJ Failed to create database encDB see the next exception for detailsERROR XBM Startup failed due to an exception +Can we include this fix for the comming release +If it makes you feel any better my idea here is that youre interested enough to eventually perhaps be a committer and following that process is the usual first step +updates to apply to trunk and adds a stab at reconciling Man uses Chris Hostetter in Changes? Weak +minor corrections to the docs for trunk +Attached zip file containing source files and unit tests +After talking with Kev moving for now +You can see the Status in the Old UI screen shot this gets updated as the job runs +Running derbyall now +Can you take a look +patch against trunk dont think changing to not static solves the problemWhat if you want to use the same connector in multiple bundles? eg you want to use port in bundle A and bundle B +With the patch all the tests are passed +Thanks Erwin +but keep in mind that the mailing lists are archived and thus can be searched by everybody in other words any other user can benefit from your question and possible replies. +No response assuming ok. +So either the button should be disabled until there is a text selection or in the absence of any selection the button should copy the whole buffer +A couple quick comments on a very good patch Please use the in component for internationalization stuff +Given that the column statistics schema consists of simple data types and is portable across DB vendors and versions Id prefer to keep things the way they are +Marking as closed for now gcc compiles master fine +This issue is associated to an inappropriate hibernate configuration see +This is now in the branch +Adding the MR components to the classpath was moved to +Interesting +Dont forget to update the changelog +I have removed the static field so its thread safe nowPlease take a look at it againRegards! Will give it a try. +woops +same patch just code clean up +Thanks for reporting. +please submit patches instead of colourized html pages thank youThat said I never imagined users would prefer to list large list of xml files via the mapping file optionAny reason why you dont have them in your +Tested on +Ales wrote sections on VFS tuning which will be helpful +Your next step might be to try to create a java program based on what your application is doing and see if you can reproduce the problem +This is a nice requirement +i do not want to know about tomcat i install jboss studio and i want all work smooth first example +You see a indicating your credentials are bad but the doesnt indicate the usage of how to set the user and password +Hard coded image as discussed and documented this change +Andre Could you please explain how there might be duplicates? A server can only hold one copy of each module on it +Lots of deletes and inserts of hole blocks +As of revision the wraps all incoming streams to regardless of whether they claim mark support or not +Set Estimated Hours and this time really assigned to Travis +Attached and the latter does not compile because the constraint solver complains that the real clause is recursively defined +So I think this is ok for now +Joshua Fixed in trunk please see if this addresses your problem +Applied to and to TRUNK +I could use some help here from you Mircea +moved patch level generation up to capture tests and source changes in same patch +Ask Edison to review it +Thanks Suhas Do you think it would be possible for you to post the steps to reproduce the issue +Im in process right is not a duplicate at least Im not aware of any other issue for JSR support +You are not supposed to rebuild it or change with original jar from project +Did not see the commented code +For reference lift the bar for tinyintsmallint in Pig +Ill commit attached patch woith some fixes for typos etc +This would be more reliably correct +Let me know if anything doesnt pass in or since I added Spring Test support in both of those modulesAlso let me know when you push your changes up so I can review the method implementation you addedEDIT I see you already pushed the changes up +I do not see why the append branch needs to be compatible with +Passed all tests +However indeed all code that wasbinary linked to the code will not work with these changesapplied since the signature of the method will changeBut assuming that the definition of the freeze is strict I have no objectionsto marking this one as WONTFIX +Code looks good to mes been almost a year since the last comment +well fixed in maintenance branch and going to be in +Try now +Closing old issues +I was misreading my exceptions +Will review and apply the patch +I would vote for handheldIve applied your last patch and Ive added a comment to the code to remember that this is something that should be improved a bit by the way for now I think that its fineThanks +It seems the same issue of the reopened bug Should I mark this bug as duplicate +Please try it outI think that some of these changes are not specific for macos and they might eventually be required in other platforms so it might make sense to move the changes to another JIRA if this one doesnt make it +Verified on JBDS +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I had checked Eclipses log file no stacktrace here +Ill try to check other parts of the code related to make this point easier +Committed at revision +Mea culpa +Thanks for attaching your app +However one of them is public readable + because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Reduced number of connections to a Broker managing reply queues +Ive fixed this in CVS but Id really like to get the test for this +it should not be written to the wire if it has not been set +Very nice thanks we will apply this +This is the last patch I put up on reviewboard +If they are different then we analyse the file DB entry further to see if an update is could also greatly reduce the synchronization issue by moving the file upon successful import to a under the extension modules hot directory +And shouldnt generally be usedIf you use file connectivity then you should either delete or move files after they are processed and not keep then around +So the additional issues should be tracked here +No not complete +now is fixed and the problem here is not jrebel but just happens when editing something in server you retry +views or set on save if not already set +I cant reproduce it +Verified to work +marking this as duplicate of which has some further details. +fixed in revision more thought is needed but that is always the case +bq +Ill improve the constructor to copy that behaviour +Version Final is released +At a customer site this week +Thanks +A one line patch should fix this issue and help the users could try using category feature for simple LR training as an example +has been marked as a duplicate of this bug +Spring is autocreating the instance under the by upgrade to Spring SHA afbebadaacbcbeaefd +As part of this work I also discovered that the Resteasy binding does actually extend which needs to be tweaked as part of the EMF model + fixed the compilation problem +New SNAPSHOT deployed again +ugh +I have tested the patch in cluster by configuring different ips for master and regionserver and also with localhost in some tests +Maybe informix generates as an id or something like that +We can expect that youd have to add the latest mysql if you add any version of phpmyadmin but were not completely sure how things will get implemented on +fixed as larry decribed +Additional connections will be dropped until authentication succeeds or the expires for a connectionThe default is So we can just add the key of host to the VRs +The problem with baseurl is that a fully blown system must be able to generate RSS feeds and those need absolute +Add distributionManagement section that only has website stuff in it +Hi ViniciusGood work the patch looks good +I did include an example of using traceDirectory and traceFile together +The typical sequence is as follows +Because of the casing issue the wrong summary line was plucked out by the release note generator. +Closing this issue per Michaels request +This takes care of the leak for but the same change on trunk still shows a leak related to comibmlangmanagementThere must be some additional change related to jmx that is necessary +The above flag fixes the tests. +Coverage using marvin verified the following tests and they went fine with latest builds +Attached a sample repository +Patch applied thanks +The complete bug report +We just need an existing database with the old schema to work on. +I need to make some editorial changes to conform to the Commons coding standards +You may want to leave the listSnap call out and add a TODO and wait for Nicholas to complete that part of the work +The memory usage is only for the time it takes to processthe query and is then released +Closing as invalid +Patch is a little larger due to the fact of changing tabs into spaces +Ill get to work on this +It still refers to the old pdfbox project from sourceforge or do yue really use an outdated version like? +Would be great to get some feedback on this +Since theres a new config option added by this patch should it be added to the? Probably my bad for not including it originally +Willem we have added a separate CXF blueprint example in the examples +Resolved some time ago. +Patches and committed +Committed in rev +We would need phase of the feature to support multiple storages +Fixed in master request for a to and this record since is now released. +Look at the number of posts to the mailing list that are just people who havent fully read or understood something +oops closed the wrong due to inactivity. +Here is the diff file against the current trunk +Any update on getting this into a release +Wont go in w backport soon +How was it fixed? Anything the build system need to call to set the right version infoname +This bug has been marked as a duplicate of +I would use this feature extensively as James implemented it +Is that right? If so this can probably be done just as well in user code collect all the values for a given key then sort them in multiple ways using different comparatorsThe datajoin package in contrib provides user level implementation of this featureIt is an solution for now +Then the test could use it +Rebased patch for trunk. +cac +Regressions passed +The fix I committed only works if onlytrue +trunk Added in Camel a pre check that will deny polling if both starting and poll endpoint are file based as its not supported in Camel +Nits only +Updated patch version according to s comment +Nice job Robert thanks! BTW your results show an effect that was reported in the papers on this subject namely that some metrics may actually improve like MRR and P above +I think we should probably go ahead and apply the patch when others get s few moments to comment +So what is the current understanding of this issue? It is a genuine bug +Patches have been applied and committed to HEADNice of you to include the patch and associated tests +The attached patch incorporates the first two comments + for the to trunk +This would not happen with a String array but it would greatly reduce the functionalityThe input array could be at least then. +The patch +We currently use of spring framework +I just found after digging more onto the hivemind configuration that the documented bean translator is actually not a translatorbut a configured objectranslatorSo the configuration is rather missleadingSo translatorbean should be replaced by translatorobject and all configured values should be prefixed with beanREM Documentation might be updated +This patch fixes the problem by pushing open and close actions into a separate thread +Mattias can you please attach a valid and sample XML document that unmarshals successfully against the schemas attached +I also used some JNDI test cases that fetch schema information via API calls +screens to update imagesruntimesservers to update +but then the method selection would result in a method selection error +Im not on this mostly because servicemix should properly handle multiple elements by default since the SOAP spec seems to indicate that this is a valid fault +Duplicate of Patch welcome. +Powell has been filed in Apache Software Foundation recordsSo this administrative step is over we can include the code as soon as we want to +The link points to a blank web page +Documented as a Migration Issue in the release notesThe serialization of binary resources used in Enterprise BRMS Platform and later is not backwards compatible with the release +update the title to make this issue clearly +Do you have a script or method to reproduce this +Now during the query youll have to apply the same analyzer which one? If you process the user input with the same analyzer your users will never be able to find ID while if you dont process them they will need to type exactly Title of this book matching the same case same whitespace and all terms in exactly the same orderSo in both cases youll have elements in the index which cant be found and cant possibly produce hits you would be better off with not indexing them at allIndexing multiple properties in the same field is indeed useful for the reasons as you say but Im afraid that to be useful they should all use the same Analyzer and Index optionsI think we should throw an exception or log a warning unless someone can make a good example of a valid combination of a mismatched set of optionsAlso +Links are added to source +Hi Ivancan you send us some of your files to reproduce the issue? SimoneDaniele I have some deadlines this week so I cant devote my time to it but this weekend Ill prepare you some files and some more info! news on the sample dataset to test this problem +Im attaching a version that fixes the misnamed parameter force deleted that method and did import static Now the last assertion fails but Im wondering if the Test is expecting the wrong row key +The issue reproducer +GreatEverything is under our control +This will need to be fixed before is resolved by patch provided in. +remove all the blocks associated with that node from the +Since I added this I cant make it fail locally +I also agree its counter intuitive like this and having it prompt for the first time surely must be what most users would want and expect +setting of running true should be removed from run method +Blocks +Its not possible declare two bean in security xml +The idea of using gnus auto toolset is to not rely on paths since they are systemdistribution specificSo well any information you provide does help in solving the problem for the next user +ongoing issue +It would be really nice to get rid of the me problems for the lib plugins and thus we should maybe search for a workaround +Fixed in r +Ill commit this tomorrow. +Neither Rhino or Diceros are a TLP or incubation project in Apache +I see thanks +Its confusing +Well yesJCR standardizes on importexport format and there is an API to do it +Even then theyd only rebuild it to accommodate incompatible API changes +The servlet transport works exactly the same in and outside +I was sidetracked on a different issue yesterday and when I synched and rebuilt today the test ran ok standalone. +Dimitris Ch +It is because ofThe same test passed on my machine +This does not appear to be fixed as the exception could occur when configuring the not at load are now created in a lazy manner so the class should not be loaded if it is not required by a component +The cause of this problem is that unlike MRI jruby does not search the PATH before attempting to execute the arguments to exec or system +Then at least one knows they are consistent +AlexeyI just tried with the sample HTML project from Central +regarding your first question +With given patch Help argument at end of whole command using command history then user will erase only last help argument alone without moving cursorVasanth kumar +Will it be better if I create the patch for all benchmarks and upload them or wait for this one to pass through review and commit and then proceed with the other ones? +Thanks Vikram! +Ill try to adapt +Please start this after if you need help with this release let me know +Sorry for the delay +Is this issue fixed in release? Im using and in my unit tests autowired is null +I do not mind moving the test there I assumed doing the change and providing tests was the way to go? I will add the super column support +Its committed and you can try the next snapshot or the release that should be out any day +distDebug L +Why would we put any obstacle in the way of the group that is running the largest hbase deploy? Dont they have enough headache already wo having to jump a gratuitous incompatibility hurdle Anyone even need this feature in? Suggest removing it for so our man Dave can just go there +yup will do +One more quick question do you disable annotation processing explicitly? I am asking because one code path that seems relevant is only taken if this is the case +Does fail with snapshots +but in fact is the same problem +I dont think itll make though. +I like purely becuase of m going to spend some more time working on this next week so I should finish this up +This fixes the script to use the current libraries the to use the JTS class shipped in the h datastore and adds the see the patch has been applied + for patch +And this problem occured too with a non snapshot version was included instead of we will try your workaround! issues for filed against but I cannot reproduce using that version +I hate UI programming +You appear to be running your mirror on win +Patch file created from trunk directory +This should apply to trunk as well +Both events are currently raised with being deprecated and slated to be removed in Seam +What sense does a lib artifact in group have? close to none +This bug has been marked as a duplicate of +Fixed on SVN trunk +Ted I have the patch locally and have added to it as well as fixed a few more typos +The test suite to check +Ill then try to investigate +A patch to add and their respective tests +added a method to the interface to perform a cache access thus letting the cache implementor to define synchronization policy +Thanks for your help and Gabriel for your suggestion +Huh youre right +Agreed +bq +Patch to add requirevaliduser to httpd config section +Thanks Adrian Here is the updated patch with locate and timeZone +patch does not include the pipes binaries +For now though it would be good to focus on getting a better error for Derby users using +for the patch +Fixed thanks for the patch Peter! +Done +It still applies. +Fixed zebra table rendering in Rev +This patches follows similar scheme +The system could be smart about caching the jar on nodes +Looks complete +This would be broken by generating a hashmap snapshotInstead I propose adding a component that extends LMC explicitly for this situation and so can use a map to cache the +Closing the issue per previous comment +Please ignore there are problems in the current patch will update later today +You need to figure out where this address comes from +so if the client resends the pending requests to the server when it reconnects the server should be able to either replay the responses or execute the request +Reattaching patch to run QA again +This does not really help us though as we presently cannot use LGPL licensed code in ASF projects +Graphics shows average time of method invocation +Jimmy patch applied at r please verify thanks. +Hence closing this. +Posting the patch for his fix on JIRA +This sounds like the best option to me +HiI just retried this LOB problem with data files bigger than kB using theOracle JDBC driver +See and for protocol specific issues +I will add set to overide looking for the special target named finalize and use essentially the same design for the only downside here I think is that finalize shouldnt be used as the name of a target that isnt a finalizer +Now if users clicks on Expire button then all the active records are shown instead of recently expired recordThanks Ratnesh for reporting this issue +Please create a ESB JIRA and link it to this +are you a surein my test there was no downside with custom tag invocation only when in customtagpaths was defined it gets extremely slow +For surefire Im currently stuck +Embedded WAR type does not exist for AS plugin + +Try to remove it +The images must be resized that it fits the page +Incorporated Raghus comments +The fix attached to this issue has two issues The imports on spring security are wrong its missing a web It does not fix the problem with filternone +I opened for the score diff issue its a bug +Its default logging is JUL but it can be configured to use SLFJ by setting set CLASSPATHCLASSPATHAPPHOMElibAPPHOMEconfand in the lib folderThis obviously works because we saw the DEBUG outputI will try to use the logger directly +Thanks Rick for the proposed simplification I believe there is no differenceI found I could simplify the last part of the query in the same wayUploading revision with that fix +this bug was falsely assigned to jboss tools +I have deployed the test application in Portal and was not able to reproduce the went on the page httplocalhostportalduplicatePortletclicked on submitthen went on click submit again and it does not I miss something? +If we had a single disk partition the numbers would be even closer +serverproductiontmpteiidbuffer +bug fixed on cr +Have committed the merged changes into codeline using revision +Bard once confirmed please put this in the spec +A better way to fix the bug you found is to use the method to construct a cart object if one does not already exist +As best I can tell all patches for this issue have been committed +That doesnt look like a AS bug +I suspect this is a localisation thing here I get for ve just removed the upper limit for now +Thanks for the patch! +Tried several times on fresh workspace after jbds restart yesterday and I always saw that issue +project top setup provider HIBERNATE database HYPERSONICIN class testAutomatically field string fieldName nombre notNull sizeMin entity class testAutomatically field string fieldName pais notNull sizeMin entity class testAutomatically field reference fieldName idioma type notNull field reference fieldName pais type notNull controller all package test controller test controller test controller like the problem is that the enum is being treated like an entity by using field ref instead of field the script to the attached code should address the issue enum type class +ok +on the new patch thanks Swapnil +I was using Karaf which is similar to Karaf +forrestdtd +Thank youI read your blog often I respect for youOK Ill investigate that parameter assigned expect Thank you +Instead we have a MODE that is either MOVING or NORMAL corresponding to bootstrap normal operation +Lets commit all changes +waiting for you then +receives only security fixesIf you need this fix then youll have to upgrade to latest version +to TRUNK and backported to +Hithanks for the note I cant find the mirror that does this ideasGroetenHenk it possible someone else removed it already +This is something else that the riftsaw deployment is overwriting we need to prevent riftsaw from changing any files in other to Julian so that this can be prevented as part of the platform changes to deployments should now be requested via SOA jiras +These are the work for this changes to the attached should address this problem +Works OK. +It may give rise to too much confusion +Dirty state is interesting i think to support undoing and other interesting stuff +Seems I forgot to handle an exception lets fix that +the change was already in trunk and since jesse verified the fix closing as fixed. +Added timestamp validation to the form widget +This is the output from buildqltestlogsclientpositive after running the patch after resolving the conflicts +Done +CLASSPATH Main Will run with rows +Hi GlenI installed maven and I am having problems with junit dependencies +If we decide to do this it will be a later feature enhancement +Would be good to fix for and +I have verified the build locally +To show this more clearly I have re added this patch with test as the first test to force the build to run that test first on jdk +In cases where the client wants the majority of cells in the block youd basically be sending the same amount of data over the wire but it could save the regionserver all the work of decompressing decoding iterating reencoding etc +Lets get this in as its an annoying bug +So its nothing specific to SQL authorization its just a quirk in the display of that property I guess +Hmmmm okay +Can you give an example of a classpath where quoting breaks it? It definitely works for me right now +Testcases added thanks for your transition resolvedfixed bugs to closedfixed +Marking as fixedwbuildrsvn ci libbuildrcore speccore m fix for Continuous Compilation Support for Sending libbuildrcoreSending speccoreTransmitting file dataCommitted revision . +But I think having a separate special config file for the config items is a better approach than having a new config item +This looks REALLY nice +Setting closed with resolution as later as this will have to be approached differently. +If necessary hard code the value +Just tested on and it works. +May also want to try specifying in to see if its related to mmap +Set fix version and closed again. +filehomexiaoxiatempmaventesttargetclasses filehomexi aoxiatempmaventesttarget filehomexiaoxia +Upps right +And theyre not +committed to and master +Added patternExpression property +on committing +Integrated into trunk and +Maven is absolutely right how it behaves +Ill let you know if I have any problems related to +In principle but Id like to have a short conversation about this first +Uploaded patch +Freddy small project attached which demonstrates resolution of this ticketYou willl not see any violations when rule checkstyle activated with parameter allowtrue and will see one violation when allowfalse +Sending srcmainjavaorgapachecamelcomponentbeanAdding srctestjavaorgapachecamelcomponentfileTransmitting file dataCommitted revision +The problem of restart was already fixed in the previous jira +Ill attach test my test case +Fix was to remove the source feature as in include from inside the core feature +Pom version and references to artifacts updated +Compiles clean now on on my machine +Jos Since you seem interested in tackling that one do you want me to assign it to you +Attaching a patch with the logs removed +Fix for the dependency +Ok attached is the patch I referred to in my first additional comment +After doing everything it can to make sure that a filesystem directory should exist does exist and contains nothing proceeds to remove that index directory and log a warning saying that it did so +This now works +Thanks for looking at it +But if identifiers are not that common then it would be beneficial to show a different example one without caps and quotation marks +Although without a more specific example of what headers were being set its difficult to judge + these packages from the following comps package will no longer be included in needs the jbpm library IT SHOULDNT BE REMOVED from seamlib +Same failure I guess I will have to wait until is solved +This may be mute issue after the latest changes to the Teiid to allow by default access to schema information as these calls are based on sys tables +you tried to + +Do you mean separating out all three from the inner decoderencoderschema layer? Or separating out those individuallySeparating Specific from the rest was easy +Undoing r for +The proper fix for this issue is. +Can you try upgrading to the latest version? If your problem persists please post your full pool configuration +its detected that the locationmap is not valid the cache needs to be flushed or the reload of the locationmap will have no effect +Bulk move to CR +See previous comments +Updated to JSF Mojarra Beta +This part refers the design of +Added to cc list +This demo was a start for a Firefox extension +cant find any source reference to lucene so we may even be able to remove it +That way if a node is lightly loaded it will checkpoint every hour and would still be much smaller and if it busy we wont add extra checkpointing loadAnother thing to implement image file token to avoid steps +Sure Generation fails because of simpleElement defined in wsdl generator expect only complexTypesMaybe try to replace your simpletype with complextype +Sample EclipseThats rightUnfortunately all the example i made test with used redirect in the bad +hi I am Tharindu first year undergraduate student from university of Colombo in Sri lanka This is my very first Gsoc and I really interest this project I think that I can contribute it with my new ideas more than java or any other technical skills +Perhaps by grabing the original if thats stored somewhere +I extracted the matching logic into and added MC based adds to dependeciesbut it shouldnt be cyclic as classloading should only depend on +This seems to work now. +Fixed the find not to cause NPE +Fixed by SVN revision npe for above with SVN revision +I have relaxed the resource processing so that a warning is logged rather than aborting the deployment if an Resource cannot be resolved +See bin +That is not an option for us +I say commit away +Some curious observationsThe timeout exceptions occur after perf node joins +v LGTM +I am attaching the files here. +The data is like following cleanmaxlevel inlinemaxlevel inlineserial From the data we can find the benefit distribution is inline sync elimiation scalar replacement +Hi Jayyour suggested change seems to fully solve this a lot for the good work so farPlease consider to add the changes from the attached together with your latest change to help others to build a working WLS example regards you post this on forums with more information on steps you followed when you get the error and any other information +This is already handled as part of i tested version conversion for all maven dependencies and added unit I made the modification +Not by Groovyc and not by javac either +Tried many times on one and one newly downloaded Eclipse on more than PCand it really OPS is simple Chinese Win Or it possible to post an error log? This should be available in the error log view of Eclipse +Thanks Brock! +ciemo you can find start and finish times of every task in job history +Oleg I know we already discussed more than once +Patch to fix problem +Patch applies ok and the integration tests passesHowever I have reverted the changes to the Sling API to eliminate the need to depend on a later version of the APIAlso there was a license header missing added inOther than that LGTMI will go and find the doc and update that as well +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content verify +Patch which adds a log line +So I guess the que satino be comes white defines sabe consumer +You may close itIgor +This could possibly be implemented using JNA +Attached patch fixes this issue +This allows to be a shared library which improves memory usage +I committed the patch +If I left INFO level logging in thats my bad +I am going to start a discussionon this issue on the list +IMO doAs is a easier not to get confusedI am fine with using doAs instead of realUserBTW hadoop auth use +Same patch that was attached as the rd revision on the reject +Committed +done +so its already fixed +Nickolay Martinov has been removed with references to that user replaced with Nickolay Martinov +Click on the button Change Tree and a tree made with the tree tag is displayed +Fixed as it was in encoding and was treated as binary +Committed the patch to trunk +Originally the max and min configs WERE scheduler specific +trunk +Hi Jay I believe that I can is M targeting end Oct or mid Nov or some other time + +This example also show cases the view definitions using DDL in the Dynamic +I will definitely look at your fix and apply it soon +Perhaps a less contentious solution would be to make addValueaddField ALWAYS defensively copy collections but keep the current setValuesetField behaviour +This could help to keep the position payloads smallso they would compress nicelybq +It does match my expectation that matching shouldnt throw but at a directory component level rather than theThe new behavior seems a bit cumbersome since callers now have to check for both an empty array and catch FNFE to handle the case of nothing found +It looks like svn commit r from Andy not only added functionality but also fixed some errors hereRegardsYou mean you tried with trunk and it worked? In such case we should extract and backport relevant fixes from this commit +Theres no patch attached +Awesome +I altered a unit test to reveal the problem +FYI we are keeping all versions so we can do inserts and we have a mapred job that runs once in a while to merge versions +My mistake. +Attaching which I committed at subversion revision +I opened for the outstanding simulator work. +As per alts comment please assign back to me when the info is ready +Viktor could you confirm that +I read about using Java bin transport in in another ticket if possible I would like to test this +ok good +And as you are the unicode policeman I dont unterstand how you can live with default charsets +Again I think a test case would be helpful +Feel free to reopen if you see the problem again. +Revisions and +Attaching a new patch for this issueCurrently Session objects hold reference for prepared statements cursorscursor statements and asynchronous statementsAll now accept the notation to design an identifier outside of the current connection +Thanks Sijie. +yeah! patch available very soon +It is paired with the related problem that you can use EXPORT and BACKUP to trash and data filesand even files depending on the privileges granted to the account which boots the server +This was fixed as a result of finishing +BriceUnfortunately nobody is currently working on it +then wouldnt you agree then its a issue as opposed to a CXF JAXRS issueGive it a try please +I guess this problem has been addressed in trunk when the installers have been refactored +Here is the modified patch for the same +This method returns a sorted list of the keyvalue pairs +bulk close for the release +I sure can rename it to and remove setParent method +caused by missing framework bundle tidied up immediate problem by logging error message vs throwing issues raised by me as closed +Or to put it another way thats not specific to a split and in general the need is already met earlier in the pipeline so to speak +Please add the description how can I reproduce that +It sounds suspiciously like it could be an array bug so Ill have a look this weekend +Since the test case I have depends on the work Ive done in Ill attach that change as a separate patch +Path applied with thanks to Jonathan and the revision is +First stab at the problem +Its of general utility +Nice one +Thanks a lot Steve +Is trunk unaffected by this behavior? Last I checked we still blow up if we do the chain inside a task +using it now thanks +Chunhui Let me try in +Jakob Michael I think this is resolved now isnt +There is a workaround in the latest alpha releaseIf you specify a different resolveId property for each resolve task different XML reports are created for each parallel tasks another workaround in latest ivy version use a separate resolution cache for concurrent resolve +reopening since issue is not fixed null is still shown in case of network errorsThat should not happen on the first and most viewed page in the tools +Until all the above issues are resolved we can use our legacy PPR code as a temporary fix +Most the duplicate logging was related to +Im using newest JDK and Tomcat +The nextback control could be like my suggestion o remain in the old positionMany thanks think that makes a lot senseBut I would rather make the paint tools a single column instead of two columns. +Thanks Robert! +Thanks very much +I agree that in a perfect world we would be able to move things around but a project that is migrating to maven may not have the ability to do so +from me go for it Ross +after applying the patch and running javap on the class file for which should have inner classes the attribute only lists +Only if the service responds we take additional actions as described belowUima AS client would maintain a number of outstanding requests for each remote delegate +jar tvf on an exported file works as well +Wasnt able to reproduce on latest either closing. +Please take a look at the installation guide +try there! Thanks for fixing this! +I am working on trying to get real test numbers from our cluster on this +IMO it makes the most sense for most resources to be associated with only one that way when results are analyzed you wont have artificially high results +Ok looks like I need to edit svn properties manually thats not applied by hive +Sorry I didnt see Hosss earlier commentsThanksJohn +Please verify that they were appliead as expected + from the Kuromoji code anybody taking care of this? If not what do I need to do to make progress here +Dummy data and unit tests need to be added +Can I edit? Anyway rvm was used here +There isnt much to abstract out and keeping it a task lets users customize their jars better via jar task options +is not a browser +Wont be working on these and theyre old +Trivial patch +The same program no longer seems to crash and the proposed remedy is troublesome +Which implementation wise for us is equivalent to applying the same featureId filter to all the typeName listed in the as mud +I already ran there was only one error on the Media plugin +lgtm on have just committed this to trunk +This issue pertains to adding caching in the PHP application +bq +Im not sure this is the best approach but it was the only way I could get it to work +I can continue later +Confirmed script operates on Windows with minor change to +The integration tests simply dont run and give the error above +i still think its too complex +Not quite the solution suggested by the description thoughI havent found a way to determine within the plugin execution if it was triggered from command line or as part of the build lifecycle so the committed change might be the second best thing. +Fixed in trunk. +yes and no +Rebased patch and fixed assertions +I dropped shortbyte support for now because I somehow forgot to add in and on the related Lucene patch +Please submit another pull request if I missed something +Ive added the address to the allowed posters list +Fixed +Issue solved based on the fact that the tests passed last week when manually adding the security domain now the security domain is available but the tests are failing for other reasons unrelated to this jira. +Leo Thanks for your patient explanation +The source file headers should be changed according to whatever the discussion was on The CoreThe EULA should be in the base dir of the distribution. +wraps the s in class +I can confirm this behavior on snapshot I had to move the logging entries to the top +Especially EXPRESSIONS field should be efficiently searchable from some range filter +It is now easy to do first match or last match in the code +Is there any reason why the Bash array are used rather than just doing string manipulation +Therefore the target was created which runs a single test for each of the Remoting transports +looks good Ben +The discussion in stackflow on when to catch Error is if you are planning to handle the error +try hudson with the correct patch +See my comments in the discussion. +It is ASSUMED that in the normal suite runthat there will be enough memory for the tests to pass and that when the same tests are run as part of thesmall memory test suite that the bugs will be caught +BTW Can somebody disable testing DIH? It fails of all time in any test out of allCan you be more specific Im trying to decipher of all time in any test out of allI know one particular test keeps failing that I see on jenkins am I missing something or is it more tests than thatLets open an issue with the particulars and perhaps we can draw some attention to it +Excited to get this in +It seems like some test that ran prior to it didnt close resources properly? Does it fail on your machine + +To be clear when the server is started up is when it adds all the possible locations for the deployment scanner to scan +bulk close of resolved issues. +Looks like this one is the same and has a patch already. +Though not in itself a priority change for it sensibly goes along with +is wrong issue to myself +The user will be able to migrate their existing projects to EDS +Updated example application +I was installing mahout yesterday and just did an svn update a couple of hours ago and ran the command for the first time today +I am working on this +But I cant make further work on this at the moment since weve abandoned the idea to work on RAR development on our project and have since that time a lot of pressure on our daily work +Applied +Compressed files can be corrupted during download +further work should be attached as an additional rename Traversal to Visit +Some of the issues addressed as opposed to the ctors +Here is patch for +Which version uses EAP? As it is not fixed AS Id like know which version this bug fix for +I do wonder if we should be as liberal in the classlib +I have done a full build locally and start the servers and it seems to work for me +Ive just found which provides what I needed +Adding logging statements per Devarajs comment +Thanks +is an error of this extension? and another thing that makes me think the current extension not determining the size of file when in any other request can solve itcan u test it +But i think it is good to keep this as it is because we can stick to the java reflection standardsThanksShameera +The original value is almost never useful since the unescape processing is rather expensive and the cookie API is rather nice. +I dont think this failure was seen more than once +Thanks will try to dig into this +Do you have a time frame for when is expected to be released? Im seeing this issue in a production environment and would like to get the fix in place as quickly as possible +See. +I think the new tests could be added to +It must only do that if it sees a conflictI believe this is already fixed in the +Note Currently its Servers view +Please could you verify that it has been applied as expected +Until then +Stefan This request is still several weeks away form being sorted +HiI have a auction application being developed and have multiple users login at the same time for bidding +Does it not reproduce with embedded +This breaks the scenario and in the as it uses to test those scenarios +The other option is to only store it inside the server in plain text as is done now +Deemed to dangerous for GA we will just keep the current version from richfaces builder still relevant going forward +Regarding Jsch I read about it after you point that out I totally agree with you that this seems to be better solution than running a shell command from the java code using the helper class +Wont fix we already have in the scratchpad! +Not good for for the reminder JacquesCommitted to trunk in r + +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Specifically I am looking for references to generated classes closures etc that have not been garbage collected +Daniel please do not do part without discussing with me +Attaching same patch to kick Jenkins +It turns out that the client get a when the starts to become available +The diff is updated and a current patch is attached here +Problem above was due to Maven +Im not trying to sound superior and if I am coming across that way Iapologise +We are not using uuid based routing values anymore. +Fixed in trunk and big bang +The correct page is shown in Jetty using +Thanks! Did my best! +The behaviour shared by Harmony and the RI does not comply with the specificationWould it be better to close this bug in corresponding way and to file a bug against Sun +Ive updated the summary field to make it clearer that using a name for gcc and g isnt actually doing anything +This configuration sets as dont think this is one for us +I like it! Simple straightforward to use consistent with established convention +Patch Im committing +The only functional change here in RC was to remove the notion of the injecting the attribute onto the Hibernate mbean instead the mbean itself ascertains this information +Im only changing the profile that inherits from the base profile +Committed to trunk with revision +It would be cool to support a services sales order that feeds into a which already supports time tracking that can feed into an invoice +Patch gets us closer it opens the internal web browser inside of Eclipse and points it to the first page of the minute tutorial +So please move it when you can +I believe the mapper and reducer both default to Identity functions +the build process and testing integration with Pluto works well with Maven +So should only be forced for cases where Source defines that Woodstox uses for constructing an input stream or Reader +did you try Axis which is the latest? tried ont on tomcat exactly the same excetion thrown +In without decoding means without decoding to theconfigured +Certainly makes sense at least for +For example a sender that wants to add a message gives up its thread to allow a receiver to clear some requires from the invoker +Believe this fixes the problem +The first patch did not handle collation sequences properly +Sorry if this doesnt belong hereIndeed youd better bring this kind of issue to the dev MLThe more so that there have been recent discussions about changing the matrix API and decisions ought to be made quite soon now +The has never been added to the schema or to persistence classes +Having said that what is this different from going to a particular project with a and do a mvn dependency +Sure Army I will port it to and +were you able to correct this +Is there an error message the shdfs via s gives you with the double escaped version +We are going to start coding a fix for this at our company within the next few days +Verified in the RC release notes. +If a symlink could be created we would not have to copy them prior to restoring +Nicolas I will wait for validation and your to commit this patch +r to put in context modifiers to appropriate getString method calls. +I think that this is now fixed +Fixed in trunk +This is significant functionality I think we should postpone until at least a +Thanks for the patch James +Kevin would you mind submitting a patch? It would be very helpfulwelcome +Patch seems to have been supplied but Resolution is set to Wont Fix +One last question though Where would you want to see the documentation to be changed +I think its quite straightforward though +Test case XML and XSD files for this patch +Thanks for addressing my comments Narayanan +will add additional unit tests if there is no strong opposition to the config param +Tools should not be dealing with that right +ThanksBest regards by Andrew. +Hi Pierre I believe spec mandates that and similar responses caused by mismatches have no response body status only possiblysome headers such as Allow etc +This was fixed in +Can you clarify what tests fail? A genserver crash in the log file is not that alarmingDo you have a script that shows an incorrect response from couchdb through its HTTP interface +Yeah it applies against trunk not +Verified that this test fails without this patch on Windows and succeeds with patch applied +A quickstart showing the problem +Closed again now properly recorded against Fix Version as well +bulk defer of issues to +Bump this into criteria since this is blocking KVM BVT test run where KVM agent cannot connect to MS due to this +not a the patch as Adrian suggested +There are at least two threads asking for it to be public like +Thanks Alexeythe problem disappeared. +jules is done +Yes it worksThanks you Richard! +That mirrors last commit is from months ago +However there are sometimes multiple rows with the same ID so we combine them into one contact to be returned +with the new methods added in JDBC by to +The question is we call this new mojo copyResources orMy for are welcomeThanksOlivier +Only then do I want to see something that effects every single RPC call incorporated into the code baseI understand that this is not how you would like to proceed and I understand that this is not historically how you would have proceeded with a project like this in Hadoop +Am I understanding that the main thing this does is allow you to specify one input and get back analysis for each field you specify? Well that and the GET right +There is no release in maven on related subject is missing tag on github +So this sounds like a bug you should report to TPTP or P since it is p trying to do a chmod on optappidepluginsagentcontrollerbin that file actually exist? Maybe the casing is wrong? in any case a bug worthy for I would reckon +Sian is there any specific way of generating +The test fails +Do you think that it would make sense for QE to provide you such a report +bat +If its killed for any reason other interrupt then the it doesnt clear the doneLatch which leaves the pumper thread waiting forever +Logged In YES useridI coudnt reproduce it +Third and final patch has been submitted +Let me rebase +I have a proposed solution +ships with Rosetta so it can run ppc binaries +We really need to get the properties andfieldsfields in codehaus refers to public members of the classIf you see any serious downside to doing this let me know other wise I can provide a patchEven if we decide not to support this the code should print a warning and not spit out incorrect WSDLDimsThe issue is that the generated WSDL is incorrect +Current is working fine with junit and easymock +So now the test runs in seconds maximally waiting for a minute +Sounds good thanks for the followup. +I will also move the target to free to move it back to M and resolve it if you think its trivial +Patch applied +I think it happened again after regionserver went down with oome and the region was reassigned twice +Assigning component + +One nice thing that might be required for people already owning a database journal is there a way to easily detect whether the LOCALREVISIONS table is missing and to tell the user to upgrade their schemaCheers DominiqueGood point! When this patch is applied to a Jackrabbit installation that already uses the clustering feature it will break if the LOCALREVISIONS table is not added manually +The script launches the service under the apacheds userI believe giving the correct rights to the apacheds user should fix this issueHope this helps +Im probably not going to be able to review this until January +Thanks Tony +It has been fixed +the wizard informs correctly it doesnt offer a resolution though youd have to do this manually +So we set src to to prevent IE from complaining about security on https pages +Did you add that for the generated files warnings +Yes I will do that in a separate jira +Btw what does Java do in this regard +lite doesnt include timerservice functionality +Can you actually build wink with this patch with no issues +Marking as Wont fix since this is superseded by +How did you test the performance of the map? My very naive approach of running the unittest a few times in eclipse and looking at the reported execution time of the heavyLoad test shows you actually made the performance better not worse +We no longer block on the socket buffers if they get full + +committed to trunk and and according to the patch is for not. +Have you guys at Yahoo seen this on any clusters running secure YDH + +What do u mean by test ensuring that roshan extension is not used +Removed unused method with revision Thanks Dag +And I would have voted for it as well +Ill post a patch that shows what Im thinking in a little bit +In adb case contains an array of +Committed revision Thanks allOn the others told me to open an issue in INFRA to get an account I could then also take the clover update +Derby uses a network protocol which is based on DRDA +But is staticugh right +The new category name starts with +maven repo based jars +info threads bt bt fullOnce the stress test has run the server out of handles it always crashes at the same point +Wont fix since plugin version not depends on meclipse. +You can install the and then add the featureUrl in the +Was there an actual commit made to that branch for this? I am all svn loged out +It should probably refer to buildjars or buildjarsclean instead +What is the installer jar name? I need to know when it was built +No unit tests since this is an optimizationDid some rudimentary performance testing on a sleep job with maps and reduces and forcing the reducers Fetcher to always fail +The original resolution may not be wont fix but as the bulk change required it that is what it is. +If multiple vips are returned a typo in a hostname will cause the rpc client to pound away on the vips foreverNote The changes for supporting the use of the original hostname used in the connection have been split out from to +swp and +Im guessing this is no longer a problem since you havent complained about it since summer of last year. +Please find the fix in the latest snapshot release andor the SVN trunk +Sure I am happy to compromise in this case but consistently where in this cookie hierarchy? All of +The tests run successfully and have been passing in the bvt suite. +Simple fix for the +This seems to have broken compilation of +Bugs and required changes will be addressed under separate + +always had so little traffic that it doesnt make sense to maintain it +Fixed the failed test and tested all the metrics testsmvn test P localTests DtestTestMetrics +So whenever a region is opened it return a null value in the master side +Sounds great +We have webspheres workmanager feature which needs to be migrate using commonj +moves the native code for zip and zlib into the archive module and luni and vmi into the luni module +Thanks to Bob Morley for notifying it +Attaching a screenshot of resulting dependency tree +What I am trying to do is have most of the code for in Java and just be able to call the users C code as a JNI method +Payload is the jobId actionId statusTopic is the type should be part of the payload as well +I would like to keep the default implementation more +Thanks for reporting! +Darryl? Whats the benefit of overriding PROTONBLAH instead of just overriding LIBDIR directly +I just committed this +branch updated revision . +svn delta of changes +Damn +Ive committed this. +I will put the patch on the X branch so it should make it into +as far as I remember the cql expression expansion is done inside the svg factory should receive the url already expanded just like all other factories +a permanent solution will require a code rewrite for the core search functionality which uses the old POI library and is not forward compatible with the new lib so that will take more youre not using cfindex you can try the workaround described above in the Feb comment +Attached and seems it was failing because some files were missing when the jacc server config was created from all +Hmmm they are labelled instead of +They are not conditions out of the control of the code in question +I like the idea in general can we come up with some criteria when a job needs to be kept after its finished? I dont want to keep all jobs all the as you suggest we can configure a time for how long the jobs are kept but I think we need an additional way to identify the jobs to keepSimplest solution would be to add this as a configuration property of the queue but this doesnt feel quiet right +Thanks Teddy +This is when Im using +need toremove h commands after Alex will fix issue with ajax updatesremove icons if not going to be there till final +Fixed in . +This needs to be assigned to an engineer to outline this content +Fixed in +I dont think its too late until we do an RC +Looking at your stack trace I think that we might be starting from the wrong place +Oh I see now that the and patches were designed to be applied togetherI missed that in the first examination and was only running with the patchI will also apply the patch and do a bit more testing +I doubt this problem is new in this release +This definitely works much better however now it seems that treads go off to nana land and stop taking work +Does the prefetch have to be done inlineNot really +All of the tests should pass +I will add the ability to specify regex filename filters per partition +If it comes up again Ill debug it and really drill down and figure it out +Appears to be a regression affecting +Hmm there are a couple ways to find out what kind of authentication is being used but it really does sound like it is just SSL still not establishing a reasonable session +The need to add resources tags to direct Installer to xml spec files which replace original strings to translated is a wiki page explaining the latest localizationinternationaization infrastructure for Installer +If we keep in and put future servlet classes as static classes in it will make unnecessary messy +Its not perfect but its Good Enough +Retrying hudson +Nothing has changed in the FD in a long time. +Not sure what version this stacktrace is supposed to be against +these keys are in the middle fileK iteminfoservervlen VK iteminfoserverPutvlen VK iteminfoserverstartcodevlen VK iteminfoserverstartcodePutvlen V xxxxxCxExAjIn this case we have a on infoserver and infoserverstartcode at the timestamp +Please dont reopen issues that have already been closed for a releaseRegressions should be filed as new bug reports and additional improvements as new improvement requests. +The problem is that the implementation of statics in the single VM does not fully support the language specification +See for the resolution of this issue. +. +Sample Tomcat need a build script for this. +Maybe need to disable ok button to prevent adding AS as EAP +Fixed +attaching attribute node to DOM Element using the wrong method +Moving to unknown fix version plugins have their own release t we specify a plugin version? +I like the idea of describing why a query listed in INVALIDQUERIES is not valid for delete by query +In the next heartbeat the TT came back reporting the attempt as a success thereby preventing launch of any new map attempts for this task +Shall I create a new one +Let me clarify that +Base a Gif and PNG version of the eclipse icon with the HTML logo +Maybe we can make the logs clearer +I should say on commit to branch and TRUNK. +I ran selected mapreduce jobs against hadoop on Mac and they passedmt Dtest patch v with log of number of entries added to conf +Scott you are doing the merging deployer +This makes sense as uidpwd would probably be better to be persisted during a session not permanently in the workspace +Two comments +javadoc is saying Bookie Shell to readcheck bookie files +Resolving this issue since I got no feedback and Im not observing tests failing in my environment +I propose to commit the current patch for now to get alignment between the drivers +That is the correct operation +The plugin works as specified in the documentation for +I reading at the moment the documentation step by step +does start a network server in the same JVM as the client +But yes it has to be recoverable +When the other tests are enabled it fails +Committed to trunk +Manually tested +Defer all open issues to +Do we have any idea how often people implement? Im wondering if the calls to initial in and should be wrapped with a catch for in case people are using old listeners +youre right I guess I hit the wrong momentNow worked perfectlyHave a nice Sunday +post +Closing all resolved tickets from or older +But this is probably better be looked at in the light of +After decompiling the webservice we found that the webservice does not serve any binary valuesWhat we have done to get around this is Created a new operation in the webservice called +The periodic however is done from memory not from disk and the initial at startup is gated by the initialization not the +This allows the system to correctly find the summary and cached contents +pigbootup configuration should import piggybank and define all its functions to short callable names +This is a really great change +These blockers and critical issues are resolved but not verified +Sorry Pete The EMF is just from the example so I guess that example should get updatedfixed +Yes I was using but after you mentioned it I upgraded to and still the same issue +A simple fix is to change to L +I agree that the usage should be consistent within a class +I see what you are saying +So we have to before the also +Previously it used to not print anythinginfo will print that the cluster directory is invalid and print the cluster id if the cluster directory was deleteddeallocate will print that the cluster directory is invalid and deallocate the cluster based on the cluster id if the cluster directory was deleted thus freeing up nodesThe patch also adds unit tests for testing the above conditions +And the example Christian posted is correct eg use for OR in reg exp +Can we close this one +Please let us know when the proxy is set up so we can make sure everything is working properly +if you could provide a patch file for this then its more likely to get into hbmjava thanks +And a diff from the root +Users are free to reopen this issueMaybe keep this issue open so it can be voted upon? +Thanks for making the changes +Thanks AnkitYour patch is in trunk r r r r +gitignore I use as a start +trunk r +Please let me know if there is anything else you want me to look at or verify +To verify that similar items are identified correctly +Added a spec to test this behavior in trunk in +That is why the schemaLocation is now necessary +Thanks for reporting this AlexCommitted revision Committed revision +Closing as wont fix. +Not terribly important but Im wondering if is a better module name than in order to stay consistent with other MK related modules like etc +So all issues not directly assigned to an AS release are being closed +But since such a feature is missing right now and heading text is very important in relevance ranking i feel we should add this in anyway +I now agree the idea has merit. +How shall JVMTI functions lock the state of a thread +Rick please resolve andor close if this takes care of this issue +Committed revision Im marking the issue as resolved since the problems the initial intended to solve have been fixed. +Second Davids remarks +Oh understood +I compared the Screens with the revision from months ago and it looks like this is indeed the complete patch +In the same way that offline compress does the same work as droppingrecreating the users indexes and constraints this would take care of badsuboptimal triggers +Thanks! +I tried to produce an example of the type of thing im worried about with the body of a comment getting slurped up +This patch simply removes an unused method +AppliedThanks for the patches! +duplicate to +login details mailed to yahoo user jin with sudo access. +Here is Properties page for Validation +created to move out the from the access controller +Thanks for release +Applied the fix to include std namespace usage in the SOAP Fault header +Hi FrancescoGreat thanks Ill review apply it +Basically my last comment just meant +So the issue is real in +It does seem possible that the expected test case output is incorrect and Xalan is now correct but Id prefer not to close this issue unless we can confirm that. +Contribution accepted via vote on dev +Sigh +Works! Thank a lot Emmanuel +I also added more xpath comparisons to show how the debugQuery will be displayed now +Here is an updated version that uses the noggit library to write JSONI dont have a good zookeeper setup +Havent been able to get a status from Avinash on his multiget changes so I will carry on with this and deal with the conflicts if and when they arise +for responses it would be not +Lets just think this through rather than put in a quick fix +As is the current loop is certainly not a bottleneck were already iterating through the keys several times + errors seem related to errors were fixed using and verified using and on. +default version for JBDS is so is no such thing as default version +bq +It also fixes a bug occured during large scale testing +The issue might be related to because the Google Bot sends some unknown unsupported Client Agent +Attach for option +Bulk close for Solr +I plan to commit in a day or two +Thank you +You can already see that with ant v +This is a Phonegap Build problem not a Cordova problem +This is of course a big API breakingIdeally we cannot break model from rootspi in Camel but defer that to Camel +Theyre working on that +From what can one conclude who owns the issue? AssigneeI do have an opinion on compression type versioning +r added a workaround to Benchmarks +This is not complete but is being posted now in order to get feedbackIn addition to provides web services for this also provides the ability to submit jobs for Pig Hive and +Possibly an issue with the +Perfect +Updated patch for running some manual tests to verify +Ok forget my last comment its easy to add the sequence numberSo lets talk about the api currently all querying is done through the how do you get your event objects +The same solution seems to be valid forI dont see any side effects of always skipping fields which are not stored +Also there is no diagram for this all the from to some point content was added to the README so Im closing this JIRA +They will need to study Infinispan and then do the productization of the docsWhat resources are QE dedicating to this? +Feel free to reopen if necessary +I didnt look at your patch closely enough +The former code seemed to accept that the namespace uri or prefix was declared somewhere without requiring that the very same structure be used +incorporated zhengs to both and trunk +Hi Volker could you attach a test plan to reproduce the problem? so far I just made one that makes concurrent requests times and nothing +Tried again the patch +There is no input for this operation +now it is possible Save As workable so FIXED +on patch +It would involve great changes to avoid the static way of creating instances and it may affect performance to createdispose them on each request I think that in this case it is better to keep one set of resources for the session. +As i said before im not a soapwsdl expert and for that im already trying to make simple things and understand how they work or notSo your help is obviously welcome +HiI am not aware of any problems with cobertura but I think we are mainly using the versions of Maven +Our Lucene index often becomes unusable when the file is not renamed to segments +My build now works fine +Then on to Interesting approach for the exact algorithm for Wilcoxon +Closed after release. +The mint time I commented above +Eg if is in your site root directory and contains a link +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +We wont revert this but we will try to fix the issue you are having. +That is done via the option to set default an instance as the default +I never have any issues with other downloads +remove misc +This filters i often used with a large number of terms in a single field and usually cached for best performance +doesnt look like it will make PhaniI cant recreate this on Windows or my Mac +Please try again using our latest distribution which has a set of updated BCB project files and a fixed +Thanks. +in this case +Ian please copypaste an example of your own stats gathered with your changes so its easy for people to see what the output is thus making it easy to understand and evaluate +The vote has passed +Then distcp will use the corresponding file system according to the specified scheme +Already created issue. +Avg time +Ill review this +A server failure +IT added. +If it succeeds to connect or tried all servers and failed to connect we return to the normal mode of just going over the new shuffled list +Linked to issue with background information +JianhongI just committed a patch to replace the withPlease checkout and build it in your box +I can access all jconsole tabs and on the tab I can see both JVM and together +I will retest as asked however the MFC stuff should not be an issue as the worker class was platform independentI will take your code and rerun using the versions yo uhave listed +also update the wiki tickets +Basic functionality for inboundoutbound http email message queueing authentication and authorization now available +I think we dont need chkconfig at all but Id have to verify itchkconfig is for RHEL based on Debian based we have can also be fixed in the branch by just removing chkconfig as a dependencyBoth patches wont apply on though since the debiancontrol file has changed a lot due to packaging changes +The workaround was to go into the geronimoactivemqmodulesra directory and run maven +XML Column Mapping +This doesnt have the smarts +All the tests passed with latest patch Ning for the commentsPatch incorporates the review comments +Start Tomcat +This is a defense against really messing up on the serverStill this would contradict the synchronized purpose of the client snapshot requestbq +With the OCM components now included in the main Jackrabbit build I have removed the separate build from Hudson. +But we know JAMM works reasonably well so thats fine too +Todd the goal of this jira is to write a library that could be used by failover controller +closing stale resolved issues +I agree with the optionality of the pid and factoryPid attributes and with the respective changesYet I do not agree with the OCDname attribute being optional According to the schema this attribute is required +This is not something the example did but something one of my coworkers was thinking about once. +Sun is not and will not likely be open to treat this as a regression issue since in their minds the feature in question probably never should have worked in the first place +Hopefully there arent many differences +Aniruddha could you also quickly comment on the following questions? where do you store the stats? are the stats persisted? Or is it volatile? Do you have a bookkeeper stat JIRA opened already? Could you also comment on how the current packaging would allow implementing a vendor specific stats repository underneath theThanks a lot for the good workYi +Thats fine +Cool thanks +it will return all prepared transactions irrespective if they were inboubt from recovery or from current in progress TM sees the xid and does not have a reference to it so it calls rollback +scrollable when scrolling rows mode does not work properly +Truncate HTML nicely is no longer used +Fair enough + +I cannot completely control Ivy so I just try to trick it +no idea +Is there no way that such a crucial fix will land in or even +Interestingly there is a subset of Thread and Mutex which are loaded by default with the rest getting loaded by +Couple of updated to fix unit test failures +Attached the patch since the community will want it for the release +The bulk of the patch was written by Rapleaf intern Jonathan Yip +Im trying to understand this and its a confusing description at presentThe optimizer does not know about semantically the sameAre you saying that use oftest in the OPTIONAL where the query is materially affected does not cause the bug +Committed at subversion revision +Thank you Lukasz +bq +Any new overrides should have separate issues for both the feature and the documentation +Assigning to Peter Newcomb +Management since the beginning of the installation process +Forgot to mention +Im happy to look at this +Attachment has been added with description AFP JPEG has been added with description afp in rev Thank you Jeremias and Mehdi! +Facets names are was not described. +On the other hand Jetty is tested +The good news is that if I sort the reduces by an ID that the maps dont know about the maps page just ignores it but it resets the sorting for the reducers not tooBut this produces even stranger behavior in the counters page +Would it be appropriate to remove the Documentation flag if it doesnt relate to any documentationOr does the Documentation flag mean something other than documentationCheers issue is closed as out of date as per the middleware process rules for aging of bug reports +I am closing this issue. +there could really be worse thingsbut if spaces get incorrectly added to words thats going to mess up tokenization. +Manually tested +Any more thoughts on what the best solution might look like +can you think of any reason why the namenode is going to the safemode as soon as we start running it +If the object at runtime is no instance of a is thrown +Attaching patch from after rebasing with FS changes +As per comments below +Can you post some more information about this dead lock? At minimum the stacks for both threads involved would be a good start +An empty string is not a proper date format and you are feeding the converter with undefined data +The html file still does not satisfy the WC validator but I think that it is at least +but +Its actually a bug in so hopefully there are no side effects to fixing it +I know that this is now a community framework so ask about timming is not the philosophy of this but now I have an idea of the next step and I can try to gain more time to postpone the dpi support on this projectThanks again +This means that Claim Check can very well be used as just a holder of the message address to be used for future countless retrievals of the actual data +Added method to explicitly refresh the instances and should automatically do this when setting the class loader factory +Added check for service being defined with port type for now +Attached volume in step and it got successfully attachedAlso for the uploadad volume scenario once the copy from sec +Im not sure of a way to resolve this +Im closing it as its probably no longer needed years later +That one will probably be the next for which Saxon fails +After looking at the patch briefly what is the reason for now doing this similar to all the other JMXHttp mechanism? That should require very little change +Verified by Sean at rThanks Tony. +The next thing I would want is a stable release with all the new features and bugfixes in it +Most of the tests now work apart from some which are mentioned at the end of +I installed the Seam Tools and now its working thank youBut Im surprised that the installer didnt indicate me on the dependency with the Seam Tools +However you can still create and and then rename to ending with two +and the output looks LOTS betterObviously you fixed what was messing up the line width +Will create a separate JIRA for this +However the list for Affects version does not yet have a value + patch looks to trunk +updates to docs as Bryan turns out even option is not requiring explicit only +Will review patch and get in before final +Patch for this uses prepared statements so no escaping is closing all issues that have been in resolved state for months or more without any feedback or update +Bulk move of M to its related to +Email missmatch was solved by IRC exchange. +Works with +FYI we deployed the fix with XX config change to our clusters only to find out this option would hang each jvm for seconds when shutting downA single ls call javaplatform dfsclient used to take less than second now took seconds after the change +It may be hard to reproduce this issue +Could you point me to previous report +The patch looks good to me +This may be the only way around this issue though +Patch with test run the test mvn clean installmvn exec for reporting and the patch. +The following has been done Moved the framework into test +ShwetaI just checked in an UI fix but unable to verify it since I dont have your database dump and API call infoPlease verify itIf its still not fixed please reopen the bug with database dump and API call info providedthank youJessica +Although the pattern in the name is a bit unfortunate it avoids confusion if we dont introduce a new attribute to do the exact same thing at this point +Ive donne some more testing and i dont think it has anything to do with the hivemind version +If you are already using the latestrelease I suspect the NPE is generated elsewhere. +Drools does not have a native solution but JPA has been used can also communicate with a database using an hibernate query in a from clause. +Thanks for supplying the patch +should readAt some point we should rename all method names to be lowercase +Thanks for the patch Daniel +Im going to upload sources for both snapshot and release bits +The solaris man pages specified full paths to ls variants and it did not include binls +Probably shouldnt be volatile at all then +In YCSB I dont see any performance difference over the trunk it just uses the available Java API to avoid launching our own threads +The daemons heap would stay small so forks would stay cheap and it wed have a scalable portable solution but for higher performance andor security the native code could be used. +Thanks Nufail! +Its broken again in newest snapshot +Verified +Ive incorporated Dougs comments +Removed references thanks for the report +Note that on my Windows I saw a number of deadlock warnings while I didnt see them on my Linux machine +I wouldnt worry about it till someone else complains +This issue is fixed in branch +There are also a couple extra things in them but you can just ignore thatJust declare the queue run the consumer and then run the producer +VikashYou have not clicked to accept ASF licence +From previous discussion on this ticket I think there is evidence that this class has some useful credentials +Allright I will take a move on this one +This was resolved in. +The build was changed so that ant only repackages without doing any rebuilding allowing a docs respin that doesnt affect full split will happen in zips now produced for each contains the systems and the other the docs and release notes +it does not fit in os cacheI really should rerun tests since its been quite some timeWhat do you think about CONSTANTAUTO doing your max trick +Closing resolved issue no response from reporter so assuming ok. +If are needed then either set it on set or perhaps move the methods to methods of +Your question about complexity is similar to one of why do we bother introducing all the HA complexity if all primary and standby can fail at once +Paul not really a maven issue but can you look into it +now provides the message The service implementation selected does not follow Java naming conventions +Attached patch to change the wrapping exception type +Very interesting this shows if its a QUIT that is received at it is lost between MINA receives it from the network and before its handed off to the handler +HDFS exposes an API that allows an application to retrieve the CRC of each block of an HDFS file +Though you dont need to give us all the the last one would be enough Your patch is in trunk at r at r you Jacques for committing my patchThanks and RegardsRohit jain +Following tests failed on latest patch +All licenses are fixed +We are creating mailets that relies on Avalon components to provide services +Patch for Trunk with overloaded which takes instance rather than table names +Itll let enable you to compare at two instances of tmodels Enjoy +My bad I did run the two controller of the same domain directory +can you explain the issues? On the dev list perhaps +well thats a good thing to know i suppose +One more question here +Press Enter to start the shell now +Thanks for your help. +this was already still get this error all time +Resolving + to both the trunk and branch. +Basically the right way to do this is to start your services when the config is present +Nick if these installs are failures for users installing on Indigo then Juno installs are issues with jaxrs and maven should have been fixed with but it seems not very weird +Spring is now used as version range. +Please verify that this is still a problem with Geronimo which now supports usage of Java runtime +Wow that works nicely well done Jason! Ive clicked around a bit and the url bar is updating +Resolving as duplicate of. +If I run with trunk plugins I see it properly defaulting to +Only you can test whether your two entries are fine or not +This issue is fixed and released as part of release +push out to m not seeing any broken links to PCS input and without a specific link that doesnt work I am not sure this is an issue +For the sake of calling skip listeners before successful completion of chunk we need to cache the skipped item itself anyway +Verified BVT test has been passing successfully +Internally we encounter the same problem as well +Will attach a patch patch that installs base version of pom artifacts +looks good the test failure is unrelated no new test since its the script +This issue has been resolved for over a year with no further movement +Additionally there is a patch to to get rid of the restriction on selectqtwhatever from working with prefixed handlers +This will make intermediate review easier +Patches applied at +I believe this has been fixed for Lucenec and OJB now +Committed revision +A concern having it off by default is that could break existing applications +This patch created and applicable to the root of the j source improves on the formatting and corrects the missing div in applicationssrcwebappsecuritypermissionsI dont know if it was related but epiphany and mozilla were crashing often on this page and the Constraints page +src +Eclipse project files dont support this so its not straight forward +One corrupted one not corrupted +The Hibernate runtime +pulling all the relevant fixes from trunk to the branch Should have a build by cob today. +Attaching a modified trial patch that hard code the fatal error codesEven though this might fix the problem with Postgresql its far from the final solutionI realize that tranql does not take advantange of the implemenations that are available in most existing DBMS JDBC drivers +Hi can you try this patch? It should remove the WASUSERSCRIPT variable before executing the shell +Please attach a quickstart if you are able to reproduce itThanks! +We are using a golang wrapper around libhdfs +patch applied thanks. +From IRC discussions I think Peter Palaga wants us to use the root Maven install command to verify the JPP quickstarts compile +The profile doesnt mandate using SKI as a so its perfectly acceptable to support the XKEYIDENTIFIER behaviour +OK forget about patch +Derby provides ways to log the execution plans of every query executed +Did you tried to access org packages without Packages +Yes Mark is aware of this +This is not an issue for AS +Hmm +Once Maven respository was built from scratch all of the correct jars were closing this issue +Bulk move is a patch that adds support to +Im just starting to work on it will try to finish it over the weekend +This is expected to be the global the user of the interface is not supposed to be able to change second thing that I did was make the a global variable so that its easier to access +next patch checks the cache and if not found uses current w change to extract the cache code from getCF into and pass the raw row there from RIF so you dont have the chance of a row being in the cache when we check the raw call but pushed out by the time we call getCF. +Hi KameshThe concern raised by you is already addressed as part ofHence closing the issue +More like a choice of how wants to present those +We need to talk to Globus online group and figure this out +AaronI can think of below major places where we need to take care of family +Thank you for your contribution. +Also pre region metrics and per table metrics need to be removed when they are no longer on a rs +For I just dumped the columncell labels out separately +Indeed there is no drawback in allowing for that +Investigating failure found by the precommit patch depends on a change brought in by +Canceling patch to address sids comments +No response from request for feedback +I cannot run anything from eclipse dev environment right now +Seems to be the same issue for me +Hi Christian I guess you wanted to say Add tooltips to the treemap but this is already the case +Im resolving this issue +Closing is dup of +Sscoping items for early release of Java Eclipse Mgt Console +If there was a stack trace with this can it be added please +Unfortunately our resource constraints dont all us do this right now +If you or anyone wants to submit a patch Ill commit it +One way is to add a main method in. +Fixed incomplete handling of values in +There are no missing files in the +We had trouble reproducing this uses an Ordered interface to perform ordering between processors +Are you accessing the internet directly or do you have you have a proxyAnyone else have some ideas? MikeYour log still contains the same problem as mine Request to add Host header ignored header already addedIm using and I cant bypass it and I dont have an environment setup on the firewall box to test itIs there a problem doing what I suggested in my initial report commenting out the get +Added an eclipse sample project to reproduce the removed the following files from the lib subdirectory to reduce attachment files must be copied into the project from the castor distribution +The NM sends a containerStarted event to the RM which ends up causing the NPEFrom a quick look at the code if the AM release event had gone out after the NM containerStarted things wouldve been handled +We need to commit as many of these files as possible for Release Candidate +Good +For the moment only the old configuration method is available since the new one based on seems to have strong dependencies on Metro +WDYT +applied in revision you Mikhail +for not changing the CLI if possible +As of binhcat is still part of the release so we should still test it +v +standard naming conventions are Grant! +I think that Flavio is simply pointing out that if can follow but cannot follow then a rolling upgrade works +With this patch those are fixed +Also when I profiled the code the vast majority of the time is at the level so it would appear the slow down is either JPA or H +adds a test and rewrites some exception messages +One more reason for the upgrade is already fixed for UNICAST but the fix wont be backported to UNICAST +The text file called attached to contains the notes I made while I was reading through the chapter +Im not passing it but it works properly in ACF +I think we should remove this generic? Its too dangerous that well accidentally use it in a hotspot some time going forward +I use Fedora core and And for internals of windows I think it is OT for Cocoon +I see +Fix is in the copyExecutable method +Hi ToddAny further updates on this? If you want I would like to write test cases +As said then Im willing to send my code and then you can see what you want to do with itTom +Do you still have those problemsissues when using the individual smaller tooling projects +header added in rfeel free to reopen if anyone wants to add the timestamp in the explorer +or do I miss something here +The patch did not work +Piggybacking on jobtrackers memory for this may not be a good idea +Commenting out the ZMQSocket finalizer eliminated the segfault adding the context finalizer back did not hurtLooking deeper I appear to have found a reproducible bug in the mq framework +Thanks for posting this +Ive backed that part of the change out +Sorry I will fix that +Patch to Hibernate the first patch +Im wondering how unusual is a case with a custom SOAP header definition and why it has not been addressed already +The logfiles do indicate this may be the case +Do you want to attach a patchs an old one Let me read it again +If anyone finds I cracked open something else in the process please reopen the JIRA or create a new one. +If this is to work for splittable codecs per and maintain progress as in well need to come up with a better solution +Would you please try the patch? Thanks very much +Having done all of that I still think it would be good to be able to dynamically label all the build artifacts with the same version including the pom +How can I post this issue to mailing lists +We were requested by BEAs legal dept to make it difficult to get to while they clarify some licensing issues +Simple patch to log locations per host rack +Thanks VladimirMarking resolved as part of. +Closing as cannot reproduce as per Gordons last comment. +or add for the latter caseOtherwise looks ok +Now it should be fixed isnt it? +There isnt load tester that Im aware of +It does support text and image but those data types are mapped to LONGVARCHAR and LONGVARBINARY +Fixed along with the issue with latest cloudplatform fineHence closing the bug. +In trunk there is an rpc out of +Both pathces apply fine +My apologies for the funky formatting +Thanks Lars +Im removing the patch available flag since the patch does not seem to work with IBMs JVM + for the review Nicolas +I see now Paul thanks very much for the explanation +Thanks for the report Vince fixed in rev + for the patch +Ill patch that shortly +I am posting it now in hopes of getting some feedback on it while I run the tests +One admittedly low tech idea would be to just make different coveragestores that are named according to their version +Patch committed thank you! +Thats wrong +Matt the committed patch cuts the block report processing time +I like it +One more comment on the API +Fix and test case applied in trunk +There is a new method added to the but it only use in the test case created by you +Thanks for testing this against Spring + fix all import warnings in hdfs and +Need to produce a final version of these +sec sec sec sec Could you take a look +sorry for the for Teds suggestions +Im trying to reproduce could you please paste the info from the log file hometsztmpThanksm able to reproduce the failureAfter the the fix for the package target depends on docs target with inturn depends on and When we call the target through ant though we pass the the script doesnt seem to pass the env variables while calling the tar and findbugs target which is causing the ant failureThis patch cannot be tested as it modifies the itself +But checking the profiler revealed that for a mistake the fix was not applied correctly and was ineffective Ive sent a new pull request to JBMAR + +Then he calls +Edit fields in rows after the first row +Im not sure how useful these will be to replicating the issue but more logging is better than no m marking this as resolved in b since user hasnt seen it in a long time +Attaching the patch +The developer ticket can be closed when ready +At the very least part of our testing this round should revolve around java as well just to ensure proper functionality for users of it +Leaving the test pages in the initial content is ok +This issue seems to be too Oracle specific to add hacks to fix it +Im not sure how to make it easy to set up zookeeper +If the exotic config gets say a mention or two on the mailing list +If we want to prevent the centralized thread from removing things in a users trash that that user could not themselves delete then we can explicitly check that before removing things +this issue because the options should be forced when installing for the first time +I also notice that just minutes after the commit Abe had a comment that resulted in another change +Sorry missed that +See for updates on the remainingrelated issue. +note inverted logic +child modules publishing consolidated +Tests are fine Im cutting off the release +Ugh +Marking as resolved since this looks like its going to fix the issue +If you could provide a combiner that implements that so much the better +That resolved it +Here I updated the test as follows +Yes I know about this issue in or +New JIRA for cores see +Passes check style checks +This seems to be specific to +As a workaround we can suggest to remove the repository directory or clean up the DB if one is used +its not needed should get moved to an own issue. +Verified +Thejas can you treat this as high priority since it can cause failures in pretty basic Pig Reset fieldschema of all expressions from constructor instead of doing it in each visit function +I had initially assumed it was because I moved to and the other successfully tagged projects had used +I left it at in quickstarts as moving to introduced a compilation error. +You will have issues as soon as you move to EAP Someone should open a JIRA to fix those properties on all the tests you guys have with colocated +If you have questions on how to go about writing a patch or Jelly questions please send them to the mailing list and Ill be happy to assistSomething I am a bit leery of is adding too much complexity +In this case we initialize all DSLD scripts +It still uses the same config parameters but doesnt modify the stack +This has currently been worked around by always using query string in the of remoteLink and remoteFunction and not passing any parameters to however means that these tags are unlikely to work with methodpost if query parameters are used. +If so would you please close this issue +As soon as you need to do assertions against anything else on the exception besides the type and message using shouldFail breaks down and you have to resort to less groovy assertions +Can you please resolve this? Thanks +All set you still need to migrate your content out of the contentjspwiki subdirand into content +Sorry for the slow responses lately Ive busier than normal at work +Useful changes to get tests to run faster +Fixed since revision When the archetype is present locally then the plugin doesnt it still does when the archetype is not present. +I havent been able to find any timeout for proposals waiting for acknowledgement is that really the caseI really think I must be wrong there +This version has one menu completed to show how it worksclick on Setup TabthenDo setups for the Applicationsthenthenyou will see the website setup pagecontentcontrolyou notice that is ask for product Store in the pageso this menu would be grayed out till the Product Store setup was doneany other setup code wold not be in the Setup folder +on the patch +start fetching segment +good points nicholasfor the space issue you are right ill try and fix that issue in a seperate jira! good catch +According to your st comment there is no need to export package in but I found that class is added in the Import Package list so when I start the bundle in servicemix I am getting below exception The bundle could not be resolved +If no hangs are reported in a week Ill consider this part done +defer to +Patch applied +Dan can you supply a sample project that shows that this issue still exists +The patch should just be applied to a newer version +The question is +Attaching a proposed patch that triggers a reindex on the initial repo content just before the validatorscommit hooks are in placeIts not the best solution maybe this can help start a conversation about how we can fix this issueWith this in place we can even remove the reindex true properties on the index definitionsAgain the purpose of this is to have the initial content indexed so that the query engine can leverage defined indexes from the very beginning +Any news on this? Its been almost a year since the bug was reported and no one touched it yet +This is for trunk looking into backporting this to other branches now +Also though I probably still need to override PERSIST with PERSISTEJB +Feel free to contact me if theres need for help +This is a general requirement writing a node does exactly that write one node +Its awesome to go away for a month svn update and have the tests take a fifth the time to run that they used to +Increasing could affect all statements but I dont think its likely that youll notice any difference +It looks good +In a sense these issues are related +Creating those files is just very fast +Can you write a simple unit test? Most of the components needed are already mocked up +Commit at subversion revision +Marcel Would you mind having a quick look at this diff to make sure it looks OK to you +Committed patch to trunkTed it looked to me like Jesse answered your question up in RB +The adaptation of API is no longer including in pending it was not ready for the release +I will verify with that +Could this be a race condition +Update patch prevent NPE if source has null think this feature should at least be made configurable during development I just want to change the profile xml do a new login and use the new need for this feature to be optional with the updated patch profiles are reloaded when changed because the validity object returned by the source is patch is a very good idea +Hi BrettThis patch looks good to meCould you please publish a snapshot jar including this fix so that I can run further testThanks Brett really appreciate it +I only found in the class to do some dependencies s very similar to and is mavenProject related not +Are there errors in your system log +fixed in r +Bulk closing stale resolved issues +Included in trunk for the next release +Postponed to anyone seeing this problem with now +It is always good to have some sort of reference materialI checked other sites and indeed it works +We might or might not decide to carry this over to trunk +Added sample based code for secondary attributes to see if they exceed the max ndv +Also I disabled jndi EMF lookup by default in the properties filesIncludes quite a few minor javadoc fixes in the modified java filesAs noted before the servlet would be nice but is not essential for geronimo integration +The engineer was unable to reproduce this issue but indicates there have been changes since the release that look to have fixed the issue +on the audit recovery query limitIt needs to be max max maxDestinationsbut with jmstemplate producers can change with every send so the max can be large and also the max value needs to be large +For the KS or CF drop this seems necessary to try to wait until all running compactions finish otherwise it would end up in errors like one in the description also other operations create update are not affected by this +Formatted output should be the default behavior of DESCRIBE TABLE EXTENDED but due to backwards compatibility issues DESCRIBE FORMATTED was introduced instead +is the SOAP request after applying the proposed patch +Ive added additional details for all instance types including environment for managed instances +Ignore all of the commented havent had time to look at the source code but wanted to give some hints where to look +is a strace of the C java process and children while running the query to about k open files +This seems plexus related +QA please in be related to +Sometimes JSF patch releases break the integration code +cheers nd patch +Hi Im still having some strange issues +Fix one javadoc issue in patch +Im not able to remove older file I uploaded so please ignore it and take a look at +The sorting order is By media type By generic type By priorityRegarding system providers they are added with priority while the default priority is so the last key will ensure that user providers are invoked before the system providers if keys one and two are equalIve commented one of the unit tests in since it assumes that should override +This patch provides the first version of an ant Task wrappingIn addition to providing a for commands it also implements test and touch in +We have seen this before and it was related to the test automation setup +I dont have bandwidth at the moment to address this myself +Resolution Wont Fix cannot have a Fix keep this open and see if it is needed +All of them pass without any issues +the bug with screenshot was found in GSG guide fixed with width parameterscreenshot Anatily Fedosik will look how to fix it with scale affect versions were bug was found again and should be is done. +The compiler complained for not knowing about uintt sockaddrin After adding this COMMONDEFDHAVENETINETINH DHAVEINTTYPESHto Makefile the problem is gone +Take a look at repro to see how to get an XA connection +Thanks! +Ineed to add a sensible cvs log message +There is also related in mode which causes to be available in by default which is actually wrong +Lucenes build system just isnt quite clean enough at the moment to be so critical of javadoc issuesas for the Multiple sources of package comments im using Standard Doclet version FWIW +Defer all static analysis issues until someone can be found to work on it +we dont need to delegate to metastore +I plan to review this in a timely manner +It would be good to have an SQL parser for unit testing the +John please the trunk patch to trigger CI buildAlso for consistency could you please set the timeout in to rather than to match the other test cases unless theres a reason it needs to be differentThanks +I think the return true is preferable too because can be indirectly used in some Agree Im not sure that isEnabled method just returns true maybe its created for a bit more complex functionalityBut anyway Im relying on your commiter authority and proposing patches one with throws in the method signature and one without +the mods to the French file are in svn with rev +It should extend extends +This issue is also present in the command line javawsdl tools and others +set to blocker +If we are consolidating all queries to support positions and payloads then it shouldnt be needed right +Its there to be tested so if there is no maitnainer we just have to say so? Afaik the maintainer issue was there only for modules part of +Just to be clear is it required to make these changes? I was under the impression these configs were a safety net in case there was a truly incompatible change between YARN versions +Probably best to concentrate on top level for now +We probably should have some way of distinguishing between s and serialization failures +Not found with last run and closing found in latest runs +Thomas you are right +It does not say whether an exception should be thrown or wether the provider should silently ignore these type of validation +I did a run of the AS last night using the new classpaths and the testsuite executed to completion against EAP BETA with errors all known and unrelated to the signed jars issue +Update +It is also backward compatible so we arent taking any risk here +trunk +Im on this for reasons already stated +Lets see how it doesWait on thread joins +Patch that solves this issue +This fixes my problem with the sysinfo tests +Dunno if that was intentional or notIm even more terrible with names so Ill defer to Nicks suggestions + patch looks good +lib +Theres an added process during checkoutupdate of a project with distributed builds that tries to get the latest update date from the change log because sometimes the date in the changefile is this is the cause +proposed patch at revision to the branch in revision . +fixed the issue with revision now schema location space is replaced by needs to be reopened +new patch including support for keystore password as requested in +Ill take care of these as soon as I get home +Does anyone have access to other where they could test this query and see which columns they return +The audit is complaining about and missing the apache license header +If I set the threads count to it works ok +Looks like this is complete +Thanks again for your help with narrowing down this issue +Hmmm has a dependency on +it is already fixed in please as working in in +All the changes are made need to verify API changes link when release. + +If I understand your previous message you have two plugins that are each writing to the +this patch adds couple of ant targets to execute the cppunit tests +Ive added should be available in the next nightly buildI have a few flavours on my web site but as theres nothing much to most of them therew doesnt seem much point in adding the restNiall +should work the same way as this code was done against +I suppose however that we should specify how the API version will be represented +I agree on the application of this patch but this would make it easier for the common user to run a server +So yes this is betterIt just occurred to me that there is a neat way to handle deletes thatare flushed during a concurrent merge +But that doesnt lead to very serious problems anymore as it least the exception isnt thrown up in anymore. +That meets my requirements and seems generally useful +Ok I see +Unfortunately the bug was not discovered by any of our tests since cleanup happens before it should only with one specific order of events +A solution would be to cache sort object +I verified that the issue is fixed +This is a problem for append write +Simple patch which does what is described above +Looks like this column INTEGERIDX should included in the primary key of SKEWEDSTRINGLISTVALUES + is not provided by Geronimo annotation spec bundleIt seems that the issue is located on bundle others bundles dont look to have a dependencies to annotation +Doesnt apply to either or trunk for this was supposed to apply on trunk Ill rebase and resubmit +Thats what is already is +If a symlink to forrest is created and the symlink is a relative path the determination of FORRESTHOME fails +Attached patch to prevent splitmerged region from being unassigned +Building the javadoc works nicely for me on the Sun JDK on Windows for both Spring and the current codebase thats about to become Spring this only fails on the Blackdown JDK? for the time being. +Thanks for the patch I was actually thinking about update the Expresso tests to Mocha but I guess we can keep nodeunit since it is pretty standard too +At this point I have made all the changes and backporting I know about and am not going to do more work on this until more info is available +patches pushed as cbcbfbeffcbefaeabcebeeeebaebfabbdafabeebf +Good job. +There should not be a need to introduce any xmbean descriptor for this change +So we will keep some thing like time bound flushes for Meta +Resolving invalid per reporter request. +I didnt test it so I can not confirm it works in that mode too +Updated patch adds a config option to enabledisable this feature +Flavia is this complete for CR +Added a patch to handle this scenario +Hence all updates in the editor were performed on an outdated model and lost when the model now correctly updates the editor as well. +please deal with this +Assigned for comments feedback +In Erick asked me to raise this issue and assign it to him +xml files extended in +Thanks for reading the spec on this Knut! Adding a comment that this is an extension sounds prudent +I use cached encoderdecoder to avoid getting them from corresponding hashmaps each time +In case of a call this list is getting smaller and smaller and for the other cases it gets bigger and bigger +From a code review perspective +The only problem here is that the archetypes needs to be updated +You have to decide on the priority of this of course I cant argue that its enormous. +Can you please explain this a bit moreRegards dont think that we can handle this case +At least we should be consistent between the various importing toolsIt seems overkill to burden these tools with thatThat said I am not opposed just questioning the usefulness +I ran and didnt get test failure with Jons patch +the never changes but redirects to a specific maintained by that global instance. +Please commit +But Im not sure how adding the classes to the web application will help make them available to the security realm +Seam is not supported fully in anyways is it +Roman the link you give is using systemd +Please review and let me know what you thinkOlrh +This is because a keeps only a single column in memory and this only happens within the write method +It would perhaps make sense to reduce it to loglevel WARN? From Garys description it appears to be harmless when using failover connections +Bumpimng priority since we see users running into this describe statement kicks of the logical plan type checker optimizer process +Feedback welcome +Ive tried the examples and shipped by EPP with this new +I confirm that our test passes now! Your solution make sense. +Also the james jars need to be listed in the tika bundle pom so they get properly included Ok did not know about that +For the function zero could we pass as input parameter type because the function doesnt support other types anyway? Thanks +I thought would fix this but I just tried it and it doesnt +Looking at the patch I think it looks good to me but I would like Vinods opinion on it before checking it in because the last time I jumped the gun and that is why we need this patch +If you look at this class youll see its pretty much crafted to try to get around the sun bug constantly attempting to delete the file before doing the renameTo +After talking with Konstantin we decided to create a throttler on the fly on each file transfer +Ersin we might need your help on this one because it has something to do with resetting these reusable parsers +I see you attached patches twice whats wrong with plain old files? Every user should be able to attach patchesThanks anyway Ill review it as soon as I get the chance. +can be removed in favour of the Cargo alternatives +We should be able to apply this patch and verify it fices the problem +Similar we can remove tokens too +Would you have any problem making the pig code part of trial projectI think were pretty sold on the benefits of having the base Pig loadstorefunc stuff +Resolving this as Not A Bug +Im going to start work on this in weeks +. +The local and dfs flag is set depending upon whether we are doingINSERT OVERWRITE DIRECTORYor INSERT OVERWRITE LOCAL DIRECTORYWe will have to fix that semantics and language constructs before we can address this more holistically +First listed first resource root +Let me update the UI so that at least there its explicit that its requests per interval. +Shanyu thank you for the patch +Its fixed you will now have more accurate messages +I still dont know well the internal of maven to be ready to the deadline proposal so i lower could use this for a current project if it existed +Thanks +Updated and with revision to address this problem +this should be resolved now. +Issue number seems to be resolved +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +i guess this is covered if there is no need for new tests to be written we should close this +Fix verified by Mark +Vincent I believe you have the required skills to carry out this dangerous assignmentWell have to get you an account sorted on the jira let me know if you accept this mission +I am going to do two things +Furthermore Help About Plugin Details does not list the Hadoop plugin in its listThat having been said under Configuration details under Configured plugins it does listfileChomeaaronDesktopplugins +Updated the patch due to +Rebased with trunk and no four lines of license expired for Intellij so I cant cleanup a nice version with indentation fixed +Attaching a new version of the patch since the test with threads once in a while experienced locking timeoutsHence I am going back to threads +MarekI made a new patch now using domj to assemble XML because I needed to make an escape in cases where it url to contain special characters +Can be easily reproduced by creating a manual measure with an integer value +While its still a hack from a code point of view I plan to keep the general approach of the implementation +Debugging this issue in the issue happens inside when we delegate to AxisI looked into as suggested in some of posts in the web but nothing really has changed on that file for a whileNeed to investigate it further but ideas as welcome +Please hold off reviewing the patch in detail i havent kind finished the one small part +I just pulled this into +Sorry if my titledescription were not very clear my understanding of the MR behavior around this isnt super strong + in revision Vladimir please check if it fully ersolves the thanks +closed with the release of +Committed revision . +Something like httphostnamejmxqrymbeanName to implement querying per mbean property +The issues are now closed. +Nice work. +So perhaps just hold off on committing it for a few days +Thanks Steve. +That is if we have a patch up for this branch and we can use timeoutimplicit +at that point the collection doesnt exist so it cant be linkedYou can link before the collection exists this feature was added to support some more complicated scenarios +Thanks for review + +At that time you will have to install the jar file on the former slave before clients that depend on this jar file can continueWith the second approach you stop and restart replication in order to get the jar files installed at the slave before a failover is actually needed +see +I tested this out by altering the ruby code on my box and it worked +The default currently seems to be grouping size grouping separator at least for numbers expressed via value +I wont get to it immediately but know what to do +Personally I avoid all the external content stuff when I use jQuery mobile inline everything so not a lot of experience hereNot sure if this is the problem but pretty sure there is some way to have jQuery Mobile work via page reloading in the case of traversing to a new page that is actually external +Just for the sake of completeness This bug does not only affect the Criteria API but also HQL +I will do the trivial change before commit +See +Yep still require mvn installAs for Java Mail do we know which rd party lib has the dependency on Java Mail? We may need to put in an exclusion +So the hibernate search version should be +Hi I reopen as it works with cachepath +v is just a test for I blame windows + +Fixing is dependant upon grammar which changes in +The file not a correct PDF +The code is almost correct but cincludeaxisserver needs to change on line as followsFrom typedef unsigned char xsdbyteTo typedef signed char xsdbyte +indeed the tester has found other case quite closed to this Go to Group management then add a user to any group go back to user management and we got an +It works as expected throws exception if there are multiple layout versions +A developer I work with is working on a patch for the ear plugin to implement this feature +That looks harmless +It fixes the problems weve had +At least needs investigation before we can make release +Then the from each could be mergedbased on that value +Can others watching this thread verify Joes findings? If so we may just need to update our Minimum Technical Requirements for which version of Firefox we support and close this issue +Continue as you described with and and create a +So if we just changed the throws statement with throws statement then the new implementation would be compatible with the new standards? Thanks in advance Gilles +It is simple and removes the need for MockitoJunit from the +The use of the operator for appending strings is inefficient +Removed Michael Purdues entry because it includes the IBM Cloudscape logo added Andrew s to the list +Transitioning to closed. +bulk defer to +For example I had a problem with poll timing out and dropping tasks +Im going to commit this to the snapshots branch if therere no to the snapshots closed. +It would be great to in the labels from the context menuHopefully the guy releases the final version of the tree soonThanks for youre helpHave a good dayI had just a quick look at sortable indeed it would be a better choice +is this fix released somewhere +I have been unable to find any workaround other than to not use the nothingLabel at all which adversely affects the usability of the suggestion box +I tried that same thing without settings the xmlnls in list and it did not work +Where do you see the problem +Also it needs to be reformatted as a +is a patch removing the call to patch that removes IWuse entirely only access via MP +Done +Patch seems good +In the future please provide more technical detail as to the problem you are seeing and your suggested fix if a version for the plugin is simply how Maven works +Do we need these jars as dependencies at all? IMO since the runtime is packaged with Tomcat the application should not even have to depend on these Jars at all +Sven that was my problem exactly +Though I hope you wont mind that I renamed the methodsAbsolutely not Its much clearer this way +Should never happen +Ill go through itI just created layer which has start stop isRunning apis just to plugin as a service +Committed revision Jeff Levitts patch that documents TIMESTAMPADD and TIMESTAMPDIFF functionsModified files svn statusM srcrefM srcref +Heres full patch +Attached graphic summary +I have fixed the instead of and bridge to logging from JULBut the test is still failing due missing correct initialization of mock test environment for JSF +If it finally is persisted in to DB you cannot be sure when +Maybe I am asking for a status URL on the instance that I can check to see if I can communicate with the named httpd instance? Somehow there needs to be a documented way to test out and see how well things are working or more importantly the steps to go through if things are not working +to have it wokring with current tooling it would need to be done via the hibernate tools we should get something like the naming tooling they have in and put that into the tooling when we will do the new a release of was fixed in Seam +Patch still applies cleanly to trunk +Sorry Tim would you please submit that pull request again based on current code in Github? I had merged two requests from someone else that caused the integration tests to fail and unfortunately reverted all requests which included see you have in the pull request and after syncing up this morning the change behaves as I had hoped +To upgrade I replaced the following jarswith And I left the other jar files unchanged they are +Arguably this is wrong on CXFs side Why put an Resource annotation when this is not an object that must be injected? Why put Resource there in a framework class in the first placeI do see value in turning off the default Spring annotation support even in an integration testing environment +All files of the same table are part of a single pool +Im setting the targets to and M +First working version of the patch +In r there are still broken links in the Web UI when reconfiguring the alias +Source code for grails application +You said its a limitation but SO link says it should be possible +It would suck if it turns out we allow Thrift messages just under the limit and silently drop the resulting Message +Theres already a ticket tracking this defect and we are looking into this +delay added in is completed in have committed the part that goes for +I receive always the wrong numFound results +sha md files are present the is trunkated to bytes +this error also affects the webservicewsaddressing is a duplicate issue I am going to close it. +ThxIm running fince since yesterday with Time for a new attempt for +Find an approach for extracting data from block types +IIRC this issue should already be partly solved at least for the git status part +I think we should make the jobtracker as light as possible +I modified the code to use a switch statement and ignore views and vti tablesDidnt want to fix this bug by ignoring synoymns and temp tables because maybe its not the correct approach maybe the table under the synoymns should be compressed and the temp table compressedThough maybe a good short term easy soution is to ignore synonyms and temp tables +attach my patch +Having the same issue? Is there any work around? I am using the latest STS +I guess it would nice to get an when you try to insert a keyspace with a strategy incompatible with your snitch +I have applied the patch with commit Regards implementation +random write because it seems iterative applications require this feature for which hdfs is not designed +Thanks Mike Ill remove the call from and updated the javadocs accordingly +I am not adding generated files to patchPlease review and let me know your feedbackRegards got the sense out of it but didnt understand completely +Attached a patch for this improvement +Every time a watch event arrives the client delegates event processing to an associated watcher object if one existsThe patch also includes a suite of unit tests for both single and multithreaded libraries as well as for the Sync and Async +you will need to add throws Exception on each of these methods to successfully compile +I flush ROOT region then I kill the ROOT server +Checked in the fixed isse closed. +And it probably could be closed now +Thanks Jarcec! +Particularly the implications of changing result formats is much greater than the implications of providing the stuff in a different format. +This advantage would be undermined if you had to downloadunzip or copy a MB distribution of the framework for each task which involves quite a bit disk IO +I need to merge across a fix that was not copied to the new location for +Some new tests and some fixes let me know what you whats the advantage of using a synchronize block and checking whether eviction has been completed yet vs using a RWL +Have you tested it with the latest maven +No longer failing to compile after AS and Openshift respins in Jenkins +bq +Thanks Ramya! +But Im not sure that is what the code is doing +I don think so +The spaces should be four from the previous line when a single line code continues to the next line +hopefully I caught everythingcontent a formal point of view its goodThanks for the ideaI used my PFE to replace the tabs and remove the extra path info from the file +Makes things quite clearer now +that this doesnt add a new feature +Yes I am using trunks code + notes that we are currently passing all sequencing operations through the JCR layer but do not provide any type safety for setting is the issue for completing the observation layer +Read version it looks great now Kim thank you! Feel free to update it to further clarify cf my answer to your question +i dont have the original accident log right nowKeeping log would be niceLooks like there is some issue with build environment +Sangeetha I confirmed that these patches are in branch issues are not related to Spring support at all +Chatting whenever we read in a KV from a store file we could insert into the memstorets field the storefiles sequence number all edits in a storefile would have the same sequence number +Axtually Infotech has contributed some patches in the HR area these last times and there were not much pb I guess my confidence came from this lesson learned +switched teradata to use positional ordering and corrected +Without there is a stack trace and no image in the top right with there is no stack trace and the image shows upThis PDF combined with this patch shows up a clipping bug in Ill post another ticket for that +one more thing for us to be able to use this contribution you should upload it with the Grant license to ASF for inclusion in ASF works button checked +Fix in place +Integrated to trunkThanks for the patch Shrijeet +OmarAny thoughts on including this for +Know this was a bear but itll really simplify our maintenance choirs in the future. +The way to get the priority raised is to provide specific use cases that you would like to see was talking about a how do want to use the feature if it would be presenttabIndex has also some issues on more complex components +But in certain cases it can happen. +I need to think a bit more about this one +The following file contains the source used to reproduce this issue +Tattletale +will update the patch shortly merging in the changes to +were backported to as well adding the latter fix version +r including the images sent separately +not meI counter propose that if we rename this to we put this into the root directory and remove docsIm TOTALLY opposed that a user needs to go the web site to do anything with binary distribution +poss + to have this feature inAshish +Is there more work to do on this issue or is fixed? It would be great to get the fix into +I suppose I can wait for the document to be released to the public but it sounds as if youre creating a separate extended attributes service to host file and directory metadata separately from the NN +Before that the counter would be redundant +tomcat dumps fine but tomcat has some tagbranch problems that Ill need to look atJust to let you know its in progress +Im closing it since it is working right with geoserver WMS +Dont know what else it may silence perf logging code removed in r +okthanks +You have to write a program to make this migrationFrom my point of view this patch introduces very neccessary funtionality into Jackrabbit +I fixed that and adjusted the Keynote test and the fileI will work on the Numbers support in the coming days +i tested it on final and it worked without error for test run +LarsDo you think this feature should go toPlease assign Fix Version accordingly +Thanks +information is is done. +Please verify in EAP documentation on stage +Enterprise Basic +all issues to the CR +Ok +Attachment has been added with description example has been added with description example xsl +I understand +META +Committed the patch to trunk +There probably needs to be some code in the to deal with this value but that can be created as a new issue. +Incorporated Virags review comments +If we are seeing lots of very small files maybe we should consider just working with the uncompressed data for small inputs? That may use less total RAM and it should be easy to determine that on the fly +Yes we shouldnt be connecting aggressively to remote or even local resources on an eclipse startup unless user actually explicitly said he want to connect at startup +This version describes the new behavior which was just checked in +Please delete the following files from your plugins folder those files have been deleted please start Eclipse or STS with the clean flag only me know if that fixes the issue for Mike +Looks like is not released yet +Patches to the Eclipse project files to allow Portal to be imported into Eclipse without errors +Attach a proposed patch for SSO an issue on trunk applied and fixed in SSO trunk +Dan had indicated that this could cause the wrong statement to be returned and unexpected and upredictable behavior so may be serious +can only be interpreted as it reads if there has been no component manager it is created the new version of the phrase is shorter and does not repeat component manager two times about compound nouns indeed the hyphen is widely used and helps avoid ambiguity a quote from a and respected book on technical writing Microsoft Manual of Style for Technical Publications Hyphenate two or more words that precede and modify a noun as a unit if confusion might otherwise result +There was no other information further down the stack trace which is why I only pasted that segment +One minor comment +Is that because of the copies? Got benchmarks? AgreedI will attach my current idea of what the stream interface should look like +in MR +I dont know the commond to run a test like can anybody tell me how +Are there any docs to which you can refer me for submitting patchespull requests +Closed after release. +Now it only returns true to say the key press has been handled if there is a child view +Both? +Can you check what version you are picking up + server s workingThanks +I was able to connect with jdbcderbyabcdbcreatetrue but now I consistently get JAVA ERROR when I try to connectAlso when I attempt to connect with correct properties I get this error after a different stack trace after which the jvm just shuts downI know Ive been able to create databases with ij before with JSR not sure what is going on +Another sentence in install doc to be fixed Templeton is deep in the middle of development and does not yet have a smooth install procedure +Missed removing an unused constructor new patch attached +Please run the test suites also without internet connection +Ill cut a in the next few days +But if we say that the current behavior is wrong this is a bug and we should fix it and dont have to keep the current behavior +Patch applied and compilation successful +I have updated the dependencies to reflect the latest jar files and names +Hi David I forwarded the attachment to your vmware email for your reference +including some sample graphs of appropriate size for people to test on would be fantastic to me +I ran iterations of the test with this patch I think we are good to go +adjusting adjusting priority +It seems you have a stronger background here and can be of more assistance to Andrey than I can +From the stacktrace I see that you run FOP embedded +I moved them before I committed +Im happy with the default format on publish to be different between the and however I think we should find a better way than a JVM system property for specifying this +oh I tested with a mojo which implemented disposable but not with a component injected in a mojothis need more investigations and changes in sisuWith the provided patch you can at least implements Disposable in the mojo and call dispose of the injected component +Patch for trunk contains fixes for what Stack wrote +Id like to ultimately add an integ test that mixes in a into config and validates that it actually works from outside the AS +The slurper is used inThe slurper is built for the dev environment in too +patch +This would be something fairly easy to write in a higher level language like Perl or Python but that would create an additional dependency on an interpreter +import adding should not be needed in this case + for the patch +It will work with the C backend +Yes that seems like the best way to proceedAlso we can work on each test one at a time we dont have to find and fixall of them at once +This will be supported beginning with +bulk defer of unresolved bugs to +Yeah I also like the flag suggestion better +Attached a test project +Attempting to add a single document also results in the same NPE at line +It certainly is possible to get new you just like the standard set +Already in safe mode +Thank you for the great workThank you for all the help with reviewing and testing it has been great +Please use includes proper use of the javadoc plugin +Waiting for ant tests to complete before marking patch available +Saying that its typed is sufficient +No response assuming ok. +Fixing testThanks committed this to trunk and +but we can look into see if we can hook in some actions +Also keep force return quietly if it tries to kill a application +It is harmless for now since you have the source you can edit it as a workaround +Hi Dianne youre right per the spec a managed entity is one that has a persistent identity and is associated with the persistence context +Anything else we should still change +In that case you should later your path so that your groovy launcher and jvm are both or bit versions +We still need to figure out how to attach things like jira issues and authors from a git push +Wired up appropriately now routes to. +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Discussed with FrancoisThe proper solution is to implement your own Criterion and use it +You can grab it from subversion or the test source is included in the download bundles as well +IIRC there were a couple of fixes wrt to the interpolation in so it could be due to the old data in the db when the repos were scanned again +Open too long and I think I applied the magic to fix this +Provide an interface for Python classes first +Please move to if youre not planning to work on this in the next few daysI updated the patch for trunk and made some minor changes +I think this is for only there is an heuristic when a region move we send back the new location to the client to save a call to meta +Do we want an annoucements mailing list +I might be able to alleviate it by changing the Map implementation from a to a custom one that uses the as the Entry +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +The patch was committed at revisions and +No but there is statspartialscanautogather is there eg which looks like ll run for also +SerializergetSize in itself goes a long way! smile +Wouldnt it be possible to autowire the available message converters from the context and if non of them respond to the given media type Spring could throw the resulting in a response instead of a ? +Fixed +is not supported +This is a bug I introduced into the code and reverted just after the release +And the stream can specify the mode of transport and we chose appropriate protocol +Heres a patch that always attempts reconnecting to ZK when a ZK connection is needed +patch applied in +bq +Fixed in this has been fixed in the +When we look at the stats for Flume the number of events sent directly correlates with number in the source file however in the version in HDFS a whole lot is missing +Updated the pull request forgot that they can tap outside of the dialog to dismiss which explains the +Otherwise all maven java projects in your workspace will execute that routine on update project configuration +Committed in . +But I think it might be more consistent to just operate on the abstract syntax layer in also two Graphs are considered equals if they are isomorphic mutual entailment isnt enough for to return true +It will continue to evolve however +If its an easy fix Ill fix it for RC +interestingHowever it does make operations inherently idempotent + to trunk +The remote stack trace is now used as the exception message string in the local exception +And it also improved the code readability +Sorry I meant not +You now have jira priveledges so am going to assign this one to you +The Oracle i drivers are known to work alternatively explicitly pass in a SQL this is also an issue for MS Access. +I have tested this defect +Will do that later maybe in a separate jira +It enables to initialize all the transport receivers defined in but HTTPIve successfully run the WAR in Tomcat and saw the listener is up +we will get itthanksdims +I agree with the observation that we can close this. +for the patch will commit once please port this to trunk now that the kinks are worked out +please reopen if you prove me wrong +initial implementation and tests are checked in +I forgot about the binding making that impossible to do so +It has been created with exactly the setup you have discribed +I recommend upgrading to the latest beta +Committing this in a bit +Keep it simple but convey the idea +bq +A second version of the patch with an error check as suggested by Sushanth +If anyone wants to volunteer to write unit tests andor refactor it would be good +Bulk move issues to and +Im committing it +Neither Dain nor Alan was able to reproduce this +looks good +This patch is for trunk +these two sections of code look suspiciously similar refactoring opportunity +Id be very interested to see the patch mentioned above +Ive written some notes about this on +I checked both the webapp jars and the commonlib just in casePaul I was not able to figure out how to download but I tried the jars and found the same problem +Outside scope of Apache are a couple of reasons this is out of scope +For larger riskier or more controversial patches Ill often say something like Ill commit this later todaytomorrow unless there are further comments +Closing resolved issues. +Is their an update or should I do it +The volume path is inconsistent inside the DB table and the config file +These can be handled by the site plugin and the release plugin which I will have a look at next +Future work on truly compiled stylesheets is being done in the xsltc module. +The features service somehow needs to make sure this doesnt happen or a much simpler way would be to make jpa and jndi dependent on the blueprint feature and therefore remove the util bundle from those +Just uploaded a patch that should fix this bug +Im not sure you need to track read across restarts is there a compelling reason to do thisFor it would be useful for determining appropriate initial sizes for the index summary +To confirm Pradeep can you please attach the fresh logs from your deployment +Not sure it would work with latest version + +thnx sadeep for the contribution +jsr test app and see what happens with the TCCLWorks properly as expected +This is why im asking for it +You can create from content httpslocalhostcontentcontrol and then can be associated to Stores from httpslocalhostcatalogcontrolview BJThat menu button is on the Product Store not the Product Store Websites thats why it appears regardless of what Product Store subscreen you are inIf you wanted a button for creating a Website it would go inside the screenlet not outside it + +Tested + +Patch attached +I assume this bug is related to the commit r which looks rather odd in that it removes these lines FIXME VM doesnot support the capbility right now +yep +Committed to trunk +As we continue integrating the code path better with the support we will hopefully resolve that +Can you show me a valid sample deployment that shows this issue +In the latter approach of using the bulk loader would just lose the ability to know when indexes are done building which doesnt seem too bad for the s you gain skipping RINGDELAY +Proposed Patch to properly propagate classloaders from the creator to the createe. +Ill have a look monday +Pending Jenkins +Yep tis true +Changes ported to branches in rev +bq +Closing the bug based on my previous comment +context +use and to get search results in solr browsethe test steps would be +unzip the zip file +Will do so once the current patch is reviewed +Goodie +Therefore we have to adapt our test cases +The problem obviously lies deeper Actually all plist configuration files that do not have an outer dict element cause this exception +If a timeout has occurred the code should and throw an exception +Hi YartThanks for this new patch +Tough it is not trivial wasnt implemented in the first place +Since relies on to notify it of changes on remote cluster nodes its unable to broadcast events are three possible solutions may specify whats changed when a node is modified +forgotten the issue number its as requested +It will take me time to actually set up an external machine for testshib. + +Java has its own internal version of Kerberos +Adding CLI component +Minor cleanups +For we now propagate the actual error and for we made rescue of Object or Exception also rescue Java errors. +Ran the test in a loop and noticed no failures +Same again but with previous comments +Thanks Jean for commiting the clarifications to that I wanted to target with DERBY +I had to upgrade the to version and use the ea repo instead of release +One would expect that if you transform the same data using the same XSLT that the result would always be the same +svn add +Hi DaisyAny update on thisI am unable to do further enhancements in other issues due to upcoming merge conflicts +This project should be reanalyzed to remove these obsolete duplicated blocs +Also it is not just the repository having spaces but any project shouldnt have spaces in their directory names +The assertions always expected while the server was started with localhost +defer all X issues to +I will commit this one soon +That doesnt sound too bad +Submitted changes to ejbql test from patch as they are separate from the code change to ij +this is a regression + +This is for hadoop QA plus for you to fiddle with +It is reproducable with JAVAHOMECjava +Changed Target Version to upon release of +Since the error message is changed with this patch I had to regenerate the golden files +Hi AndersI had reported the same on Sonar community but the issue was closed saying its problem with maven as the project is not getting compiled at first placeIf you suspect this as user issue then I can open this case in user mailing list +Thank you Carol! +Thanks for catching this to fix the ditamap file +Try creating new setup and then check for this exception +Why is the functionality of creating a ticket token in a filter? Isnt it a one time action following successful authentication +Looks ok unless VivekDeepak objects. +I have put this in my project management so you can delete the sub projects if you BJPlease which issues should be closed? Which issues should be reviewedAlso I forgot to mention work done before on this subject by Nicolas Malin like would solve the problemonce it is appliedso can close this once it is +Since it has the capability to do fc could we also enable that as another test +If there are no comments on this patch I will commit it over the +When I separated those out into their own module the UDF no longer required a dependency to be added to teiidmain. +Jim I wasnt clear on whether you were connecting to the domain controller or to another host controller +Looks like this has timed out. +batch transition resolvedfixed bugs to closedfixed +Ive tested Craigs work around with Mojarra and can confirm that it works. +Only the log records when the web service is invoked a GET request like httplocalhosttestcaseRESTcustomermanagementcustomer I found the reason in has the level INFO +He refused because he doesnt want to depend onMy suggestion is to simply have the ARE class implement a method called getCause which does exactly the same thing that the current get method does +as r +Usually once a few issues are fixed Ill post to the dev mailing list directly asking for someone to commit multiple the meantime can you take a look at the site output for your patch changes? I believe there are checkstyle errors in the patch +Made repMember node and properties protected revision user importer in revision credits to of the performance difference between the two content models for group memberships +Testing Done someone please commit the patch +Test is checked on revision +Ill pass the information along +The doc improvements will definitely help thats what tripped me up in the first place simply not knowing that there was an implicitly registered. +Will commit this to trunk if no to trunkThanks very much for the review Ted LarsAndrew and out of wait lars to confirm when backport to to see it was in trunk now! will revist our internal reverse scan bug fixed list to verify weather existing in trunk version or not possible need to port a bit +Ill do it shortly +After few hours all regions got a major compaction +Although is not very flexible at this time it is an improvement over the current approach and solves this issue +As usual does not like my patches the patch applies cleanly on trunk +Good catch +Not a problem hopefully there wasnt duplication of effort in either +Could you try with latest snapshots +Exporters should support nested tables +For this reason this patch is not allowed in our official released +Thanks again +It seems it is possible to have Spring Security with JSF so it would be great that security setup is also provided after web jsf +Since java only lets me set one printStream to capture stdout per jvm I have to check and then decide where to write the captured outputI am hoping to incorporate some of this code into my custom jetty server that submits hadoop jobs +Other Konstantin is working on we have discussed having a true Replicas clone might help to avoid implementations for all +Need followup on here too. +Just saw a case in where the scanner was not passing this info along +see last comment. +Please let me knowThanks +I will commit this change soon +patch applies cleanly tests passed lets see how it goes through the CI +If were going for the type mantra then is probably what most new users of Solr mean when they want to specify a float field +on the fix although most of the patch appears to be whitespace changes. +These are probably all obsolete +Committed revision . +On the other hands you have more control over the access rights you need in the current you have an example for for example then it would really help to see that because many people here a code guys like me I get it faster when I have some code in my hands +Merged to branch. +So no unit test is written +How about breaking out another directory to separate functional tests from unit tests? eg srcunittest parallel to srctest +Okay William you are the best to add the ANT for your great loan broker is in the svn repository. +I agree that this would be a neat feature +I had to copy from because its now private +Dont we make a copy for the row cache entry? If we dont we probably should +Thanks Omkar! +It seems to hang after messages +The test case does not validate this environment variable in the right way + +Presumably since we now have rubygems installed we can now install +This means that i no longer need Configurable on the and this regression is no longer an issue for me +Please attach a runnable test case +The will check for set and use it +Just remove this file +Updated patch to used unique UUID to identify a client based on my previous comment +Committed revision . +Ok +Coming back to this after the scrambleIt sounds like you have a reasonable solution here is there any reason not to implement it for +Point taken but this is an interface not a functional change +To remove the and the classI modify the class and the classQ This patch depends on +is this our oldest outstanding issue? I didnt bother looking that far back. +I guess our future selves will just have to read the discussion here to figure out what was going on +In the beginning it will serve little purpose +The patch allows for detection of document signature dictionaries +set output folder for each source folder when invoking there is an issue with doing things this way +will try to check on this +release bulk close +fix included with the patch in with close for the release +Ive added the dependency to my project today but just after it the project stops running when I remove the dependency everything gets back to normal +Please use attached zip to reproduce the problem +Gavincould you check this issue again +is the correct class for this case +I have supplied Joe with a contact at Oracle and there is an outstanding question surrounding the process of the transfer +but it does not include the class so I cannot comment on javadocs +When I added this in my custom configuration everything worked fine +Hitesh could you please rebase this on top of which should go in soon +sigh +reproduced bug +i was actually taking this requirement from the other enterprise searchengines that ive worked with that do this by default +Far better would be to make it? Surely a user who is running some projects on would want this capability +The fix is not really perfect we are replicating the code out of the link renderer here +Confirmed that this works on two systems +Thanks JyotsnaYour patch for Clone Lead Functionality is in trunk rev +I need test more when I have time +Updated the diff with some of your suggestions +Jdbc implementation of token repository is now also complete. +Its a for now until we add the more advanced version +Provided patch the correct working has been confirmed with tests against a version of Seven that had all JAR files cached locally and for which the JAR file server was brought down +probably need another JIRA so as not to forget to do this +I did not noticie any change +Please ignore last comment +Fixed tab to +Closing along with. +It also has senses embriaguesa embriac and in the plural sordonaia +Since it does not have associated that panel is marked to remove but it should notThe fix is just check that case on +Create Ear or Web or ESB project on project properties and select Deployment Assembly Click the Add +Thanks Vinay! +A test case would be useful to ensure we dont regress in the futureAlso pls mark the patch as Patch Available when you think its ready so I can track them +I have checked the logs the additional encode has been added as part of a refactoring in revision . +See if you can +At this point all the major pieces are there +If not let me know +closing in bulk all issues marked as Resolved +Hi Matt youre right and perhaps one day well reopen this ticket +Now it works fineThank you Dmitry and Alexey. +Opened to track this work. +Cool thanks for the explanation Uwe. + Using the resource name I fear the full path with get too long +Verified in JBDS respin a. +right Im still interested in looking into it but will need to do it in my hobby timeSome help would be welcome +I am going to take this read thread where Brian said he is doing this +Here is an updated patch. +As PMC chair you should be able to update the SVN files yourselfHowever I have now done it and will update the working copy on minotaurpeople shortly +After we do not shutdown the DN any more +Can you please confirmHari +Even public API if it is too great to ignore it may change in minior versions but there will be a note in the release notes +As far as I can see there is only left using this base class +Can you do that please +As no one has volunteered to do anything about this Im closing it out +Without it we give up schema evolution a major feature of Avro +Please verify the fix in the latest. +Hi AndyThanks very much for opening this +With this most of the EJB deployment time issues have been identified and fixed through various remainingnew performance issues with EJB deployments will be taken up in separate +Nothing has changed for ESB since so do we really need to change anything +In my months experience with Maven the always came before classes +Awesome! Thank you for verifying this. +See Todd +Cool stuff! Will merge it in the next do you change from SHORT to INT? We cant go beyond K do we? go into the next version of JANINO +Checked in to revision clustersrcmainorgjbosshasingleton clustersrcmainorgjbosshasingleton clustersrcmainorgjbosshasingleton clustersrcmainorgjbosshasingleton clustersrcmainorgjbosshasingleton clustersrcmainorgjbosshasingleton clustersrcmainorgjbosshasingleton testsuitesrcmainorgjbosstestclustertest testsuitesrcresourceshaA testsuitesrcresourceshaelectionpolicyA testsuitesrcresourceshaelectionpolicy testsuitesrcresourceshaelectionpolicy testsuiteimportssections testsuiteimports testsuite understand this should not be applied to Branch too correct +was applied to latest code +Theres another advanced test case I want to implement that will probably run ONLY against +HiI have the same kind of problem with the +What are we supposed to do with this +Do you have the Jira that refactored the +Thanks for the info Fritz +It does not contain Weld SE features +Moving out of +Proposed doc changes for above +are some updated docs +Weird submission but true +Datanode table can be sorted on any field etc +I cant seem to shell inssh Alan Ive reset the passwords and mailed them to adcapache. +Awesome glad to know this +Attached patch to resolve the issue new unit tests included in the committed thanks pradeep! +I committed this +When people first start using C it is common to be a one DC cluster then later it may grow up to a multiple DC cluster +Logged In YES useridCan you please try with the tool provided in CVS contribs even a configuration file that Ive tested and this bug if that tool solves your problemcheers Username rmaher +Namit Updated patch is available on phabricator +Closing this test addition. +But at least we avoid failures +Hence the +I simplified and created a method that allows me to run multiple rounds +Only a handful require a second pass +MichalI guess you have some prototype alreadyCan you let us know the percentage of performance improvement? I dont know exact numbers. +what is this trying to accomplish that cant be done +Lets be nice and say that some drivers do a better job of this than others +Unfortunately I cant be much help with that because Ive been unable to reproduce the problem on my end +At a minimum we should reverse the order in which these edits run +But anyway let me check the code of is it was modified to have ONLY one object in createBus method +Thanks! +Do you know which component releases the SAX events with the null uri? It shouldnot do that +Released in. +How do I upload PNG files +version with some trailing spaces sync with now that some tests have been fixed now that is in failing test is unrelated an as Jenkins seems behind on its tests Im going and commit unused import +Marking all current RESOLUTION LATER issues to be against the Nightly Build +The repro app +I think a architecture diagram on what you are doing will help us all understand this better +I think we should just replace the word Tomcat with Jetty there +Yes +If it is waiting then it should connect when the vdb is active +If I set JAVAHOME to no problem +The opposite is not true however I feel that taking this away forces Struts developers to make reports usable in standalone iReport where some simply dont need that functionality. +Would like to tackle that in a followup and suggest the scope of this issue is installing Phoenix as a library +frameworkwebtoolswebappwebtools find sample xls and zip files +Attaching a patch for +So I just linked the issue hereWe can add backport to the title later on for similar issues when this kind things happen again. +Simple patch + +It seems like did not correctly throw an exception to cause the component to be dropped +bq +Any reason not to commit this now +Changed the target mileston and marked closed +This is followed in some other tests +Gavin whats the pain with going back toHenriYou cannot see the tab because I disabled the tab in the plugin config +Jim subsequently fixed a problem where memcache constructions could fail leaving outstanding read locks preventing other threads from ever again being able to get a write lock on memcache +The toolbar is fine with IE on Source +Added test to cover this good message +I dont know if this is important enough to bother about +I wonder how can I view the statistics in +this doesnt work for me in beta so reopened +If they forget theyll get an exception like the one above at which point they can register the function and try again +The patch actually does a pretty good job of staying out of the Server and Client files +This fix should be implemented in the WSDL to Relational Importer not the SOAP to Relational importer should put the in quotes by default but have a checkbox allowing the user to turn off the wrapping of if they know that the be the right thing for the service +Are these ignored because theyve gone stale or because using makes tests run too slow? If the latter is the case is there another good way to write a test for this +Unable to compile the odd pixel bender files +were looking only at the attributes and not system properties +also patchset needs forgot to mention that my earlier comments about the row cache problems are no longer valid +Can you clarify further? How did the signature for that another GET method look like? What that extra method had to do with PUT method not capturing username +is related to this JIRA +Rob seem your personal QA missed something? +Makes sense was assuming that was the context just wanted to double check because I didnt see other jiras that would remove the other remaining refs +Another test you could try is to upgrade to connectorj betaI know that much of the code I wrote and not committed about reading streams from db is very much bugged in previous releases but I saw no problems with the code for writing streams that I decided to commit +The long is now injected and I dont have to prefix it with a zero +I cant see the problem +It might be necessary for those reserved words +Considered a blocking issue for but not blocking for + +Updated patch fixing findbugs warnings +This one only compares native accelerated PNG compression but with three different images using PNG and formats +Thanks MichaelI will apply to shell +Provide a high version first policy which means that the same export packages with high versions will be considered in the dependency tree will provide a parent first laterCurrently I only enable it in the deployment time for there is still issues as that plugin explicitly defines the framework export packages there are different with runtime +I dont have a network drive to test with but Im wondering if the script is actually passing but just giving some scary output +John so for you just this file is generating the exception? I have the project in my workspace now and I am not getting it +Go on with Native side +Fixed in Before upgradeVersion Build id Build date After upgradeVersion Build id Build date +Closing issues. +I will improve constructors while working with the Monitors +Depending on how we tried to trace the processingan would occur +Cleaning up old inactive bugs +That behavior comforms to APIThe latest Tuscany SDO implementation is already throwing for test case +I just committed this with a few javadoc updates Jim suggested. +This patch is a progress of this issue +If there are areas that are not using sourcetarget I think we should fix that +What OS did you try it on +Its not clear whether it reverts the independent fix in that is mentioned in the description +Shravan why is it always a good idea to do this? This is not specific +Also do not forget that each build needs to unpack requirement before using and it should NOT be in central location +I think the alternative option Eric is referring to is to default to true which means people with secure clusters who just update their bits dont switch from kssl to spnego automatically ie theyd have to explicitly enable spnego via setting to falseFortunately people can override this and ship with true +I think it would be good to limit the downloading to the case where is needed and doesnt already exist +Cells can have arbitrary tags +Added as base class to make it easier to implement custom to branch as well +Id let it simmer a little longer while Greg Jan and other Jetty folks get the deamons out +Im wondering whats needed to port it over to trunk? Are GS datastore pages not ported over yet? +The test fails for me when splitting the first region the client sees regions for some time +any chance of a new alpha in the near future? Id like to include this in some jclouds tooling +However I dont see this as a blocker for autotest anymore so Im resolving this. +I dont think full index pluggability is a goal here +I think theyve made some changes +ah yes i see what you mean but i think i would prefer a outstandingRequests that might be stale to a possible deadlocki agree that this doesnt need a test +bulk close of obsolete debugger issues +I believe this one to also be resolved by the simple TTL fix. +Its a little disturbing +AdrianI think it is because of Script changes previously I found this type of issue in forms also +Now if shutdown is called and the server manager has not been started nothing happens + +Im not sure which one fixed it but one of them didCan i just make a quick feature request as well an auto increment feature for the docid would be really useful or be able to get the highest docid via an xml query so I know what result id to start adding from +This is generally useful probably doesnt need a configuration value to enableconfigure +Same patch against for just committed this +The observed behavior doesnt make much sense to meWhy would the logger on restart keep file unchanged overwrites and leaves and in place for the time being and overwrite it later eventuallyIt looks like the logger doesnt find the correct end of the log for appendingLoosing log information is certainly a severe problem for me +has just been fixed so we may have a reasonable hdfs version as made it as well +From there I will add in the +CXF operates on which is supposed to return an actual absolute request URI and in your case it is decoded too earlyAs I confirmed in one my above comments I could see it working just fine I suspect it is a Tomcat bug +So looks like the problem is when the it deploys under the AS +Now merged to the release branch. +added a test +Im creating a separate issue for the new failure +Investigate for if they should be in SOA Van will set the Fixed Version + +changed so that thrift directories are excluded from checkstyle +Basically what happens is if the active region false within a component branch that is renderedfalse i get an empty http response +no problem! I will squash it before merge the pull +Ill have a look at the other issues as time permits +I have tested your test case and passes with the latest code from remoting +So I am not submiiting another patch for this +Sorry folks +I can now see that it is not good destroy the session beans on a container shutdownHowever there are maybe lots of active requests on a productive environment and I think those should definitely be destroyed correctlyThe beans are stored in the current view and thus in the view state +Ok I can see where they are declared in the source code Ill add them in +the patch with Grant License to ASF selected +This issues covers both text and windows t do step of until we know B has been T have been completed and are part of the code drop +can you confirm what your local metadata looks like and the contents of the remote server +How about ra for refresh all +Are you proposing getting rid of all of the sample configs and having a script that ask some questions and generates an initial set of configs +i dont think its a problem with this particular dependency +NoNot unless you give us the minimal information needed to get a clue about what your problem can beAt least the very minimal is the MINA version you are using plus a test case exposing clearly the way you use mina +Closed fixed +Anyways changed it so that the UI puts the URI right away +Could you please try to reproduce the problem with as well? will give it a try +see test output added in the comment above +Will be in as well as in +Interesting a mirror where you cant download anythingPlease check your configuration +However I might be able to come up with another solution based on this idea +Marking resolved no further work planned on this issue. +Thanks +Manually tested +Shelly is seeing if she can work around this issue +The others are related to functionality provided by a process +Great sample thanks! Ive commit it nowTheres an issue with how to have it run with both the testcase and when deployed as a webapp as the WS endpoint changes +Thanks for reviewing guys +Again feel free to disregard +Can you provide some more details on how you setup your environment and provide a sample application that demonstrates the problem +Patch to add looks good +This patch is based on Hadoop release +Also all the +I believe Corridor normally handle these +Thanks for all reviews +I am using default Aegis bindings in other words I have not defined acustom Aegies bindings file +I think he meant to comment out +I noticed the same problem rarely occured with running these tests against RI on Windows +Attached is a patch for this issue +Do you mean that you are going back to the solution of merging monitoring with actual workIf so Id prefer that you focus on the actual work part and leave the monitoring on the side for the time being +Resolved +Stack thanks for the review comments +According to earlier comments we may want to close wont fix. +Mind posting an updated patch +I thought was a different issue by looking that the description of that Jira and its comments +Have a look at some of the other unit tests in trunk to get an idea of how they are structured +Ill apply it when I get a little time which may or may not be today +Concatenating all our names the attribution sounds as followsApache Directory Apache Directory Server Apache Directory Studio Apache LDAP API Apache Triplesec Triplesec Apache the Apache feather logo and the Apache Directory project logos are trademarks of The Apache Software FoundationDid I forget something +Thank you +Thanks for the report and fix! +Ill submit one +Reinvented wheels tend to come out square in my experiencebq +The reason it matters for this is because logj depends on those classesSo on a fresh build for zookeeper or any thing that depends on zookeeper via maven would not work +Thanks for your comments guysUnfortunately the scenario for this bug is not easy to recreate so dont count on me to verify it when is available. +A suggestion i have is in current case we can connect to registry while loading Xbaya and load the objects to avoid user confusion +Tests passed +Hey Linconln! Im so sorry for no update it before +I bet this has been fixed already thanks for verifying Jeremy. +Oh and Thanks for the reviews Sandy! +Verified in +files for a large number of jarsDo you have trickFor now Im doing it by handFor example mvn install Dfile icu Dversion Dpackagingjar The complete command line with checksums generation should be mvn install Dfile icu Dversion Dpackagingjar true Felix +from meBriefly looked at the patch applied it and built Derby without errors after doing ant clobber +Namit Yes it has dictionary encoding for strings +No this particular run didnt have the fix yet +Missed your note +Attaching bpmn process problem lays in Specifically The Knowledge agent is not initialized before calling for the list of packages from GuvnorI have a solution which I have verified locally am relaying details to Tihomir +Problem describe is filed as and more details described there +It would be nice if this patch gets into +Tycho is not yet released +I use version +Geir apparently you have no time to handle this issue hope you dont mind that Im going to manage it myself +I think more than that is hard to specify at this point +Unpacking jars and older zips has native support but I am unaware of any packages that do that for the variety of archive types claimed by Aperture + +run the same patch again to see whether the C test failure persists +This would allow an embeeded broker to be shut down without it leaving behind a shutdown hook that would later run when the JVM was shut down which I believe is the only reasonable explanation behind the request to directly access the threadChanging title from Make it possible to access broker shutdown hook from outside to make the unregister its shutdown hook when closed by means other than the shutdown hook to reflect the actual change being made +integrate Chris Mills done on the side now the only piece missing is AUTH itself +RightIf we save shellcommand into a file then the command launch for AM will becomeJAVAHOME binjava Xmxm numcontainers Then AM can read shellcommand back from the file and launch containers with shellcommand pwdls +FWIW I like the patch you posted better than the proposed solution +Also the indexes look like a very good contributionIll investigate both suggested improvements further and get back to you +I opened up the code for the webdav project as I wanted to see if I could help get this task kicked offHowever Im not sure which under webdav one should start looking related to this task its very confusing there is a server a webapp and a webdav project. +An error is thrown if the partition is missing from the current cluster but is present in the primary cluster +Fixes minor issue enables shell support in scan and count +Patrick the patch looks great +Thanks Rahul! +on the patch +Great! Tests pass and it looks like the accidental changes are gone +I patched the serializer and tests pass in rr +It is not possible to override classes from JSE +Thanks for the spec Rick +Canceling patch until Konstantins comments are addressed. +Please replace this file in the repositoryorgapachegeronimoframework +It seems this may just be an oversight that needs correction +Each stage carries set of information including requestId +If the command exits the shell will not exit +but contrib tests would not pickup from srctest +Did you try using the examples provided to test the db integration +On the other hand removing Directoryclose makes sense on a first think we should catch exceptions that come from the IWDirectory +I can make the required changes if people agree on it +I havent tested this patch though +There are a couple references to in maven that are causing the build to break on r now that corelib is gone +but important one incase we provide pluggability for the loadbalancer +update the code patch based on Bryans suggestion +Teiid Designer uses DTPs SQL Results view which we dont contribute to or manage. +To use apply the patch to trunk and build +Rajika +Having said that I do not know yet what goes wrong here +I think this patch is bigger and more complex than the patches in so it is probably better to commit this patch first then update resolve conflicts and commit the patch in +patch available +Was able to deploy instance +Resolving. +Attached the patch for YS branch +for what its worth wicket can come from the namespace prefix that you will make configurable in +Also the fix may be involved +Hi JustinSorry about the delay +So I desperately need the fix back ported to JBOSSGA +Dont start split or even compaction if weve been asked stop +If you too reach the same conclusion please do state so and then resolve +Then Futon using a simple switch logic can display View Group Compaction or something else +hibernate module updated. +Deprecated in Release NotesDeprecated in next release next release Release Notes +jar files in core and the following jars do not contain a license within the jar file itselfe about the Eclipse License you are completely right I have just added it on either branch and trunk thanks for finding +The patch isnt a very good solution to the problem it results in duplicated classes +It looks like you are trying to get rid of the Journal Spool in BN +out of curiousity can someone explain to me how the behavior of differs from the Portuguese snowball analyzer +OddCan you try that with a log level of debug +Checking for loading of availibility if visual editor plugin was added +I was looking at improving the error messages when a column is accessedmore than once and I realized that doing so will be much easier once thepatch for is in which records whether lob columns have been accessedI would like to propose that we go ahead and commit the patch to trunk and then attack the improved error messages after goesin as a separate issueThoughts +Thanks Andy +could you regenerate this patch I am getting failures in applying the test portions of the patch has been a hot file +Grgory do you already have a patch for this +mjdksvn commitSendingSending javaengineorgapachederbyimplsqlconnSending javaengineorgapachederbyjdbcSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiTransmitting file dataCommitted revision +SCABPEL +This is a pain point for our users both new and experienced +Prakash neither can I +bq +Committed this to trunk and +Closing resolved issues for already released versions +Modified the test case to include an attempt after login. +Superceded by YARN though a lot of the work here needs to be pulled in there which can be done on a case by case basis +Patch uploaded +yeah Id like that too then we dont have to enable that by configure +The warChild module dependes on tools and has container configuration parameter container is of type File so it will not find your file since it is inside a dependency +Available in head +This way iterators do not need to be disabled across the table for an edge case in the data possibly impacting operations +Can you zip down the project removing stuff which is not needed so that I can reproduce this +I was never very happy that resolve mutates the auth request +While I understand the reasoning for it doesnt that weaken security? An impersonator can publish an alternate principal for which it has a keytab forPlease note that server advertised principals wont be honored by default +This was just something that I thought about while looking at existing issues +Needs to be merged to as well +is left for the followup ticket +Maybe startfinish arent allowed to be empty strings anymore +This patch should do it +Is is possible for the provided patch to be applied for version +Bug has been marked as a duplicate of this bug +Radim can you please update your patch as it does not apply cleanly to most recent trunk +I also changed the unit test to use itThen in gets I had to parse the data from ZK so Im making that method public in +Verified both patches using ant clean package test Dtestcasenothing +thanks pat for the suggestion +Reviewed content committed to SVN revision . +Weve never tried to build on an ia system beforeIts likely that this will more or less just work but I think were going to have to get patches from you to make it happen +Committed to trunk and +Do you think you could help with +bulk close of all resolved issues. +Updated Patch includes the focus on last possible row after deletionchangewill commit that intrunk rbranchx close resolved issues +I also renamed finish to flush since thats what it really is +We may spot some additional methods there and there but I dont think there is many remaining. +So all should be good now +Is this also going to remove the dependency on logj +Thinking caps on +bulk edit of issues that for some reason were automatically assigned to me +It will be fixed in the next AS release +Hi Royyour patches are quite at a risk of being ignored because the WFS module does not have a maintainerIve added Gabriel and Jody as watchers to this Jira as they are the last ones to have done any work in the module +Closed based on Ilias report. +The thanks +Was for and not for trunk +I could not reproduce the Unexpected ordered exception myself so Id like a test case for that +The links are marked by link so any class or classloader starting with link may not work properly but I dont think we will have any classloader class or interface name starting with link +Thanks Tom Ill post a to allow the version to be passed throughFor now I found a problem with etcprofile is not executed in the context where bincassandra is called so CASSANDRAHOMECONF dont end up getting set +Thanks the patch looks good +One more note annotations in X are definitely not erasable and can change the semantics of the program +Oups sorry you edited the file +Please dont assign time estimates +Its a slightly larger undertaking is that desirable +Remove old related +Close after release +updated amq config file and common code used to authenticate by mistake +Revised patch +As long as were not testing authentication or authorization the repository used by those tests shouldnt touch that code at all. +Time zones are almost as ugly to work with as character sets +The updated diff removes an erroneous change in the bkjournal +Leave the dutch site as an example in an other language +mopenFilevmargsXmsmXmxm +Moved until after the should probably really be in the managed project anyway +fails under TEXT IMAGE and UNITEXT datatypes may not be used in a WHERE clause except with the LIKE test will have to be excluded from the Sybase suite until a workaround for the above limitation is found +New improved version of patch +Now that its for a new release we can update the status response to include option error detailswill update the patch +What about replacing the add button with a success message and a link to add another +Applied to the trunk +I need to investigate this further and either determine that it isnt a problem or find a better way +Jiras priority guide thingy for minor says easy workaround is present +Reading the code produces no clues as to why this behavior would occur +The fix would then be to cancel getting lock from a crash node when a new view is acceptedThe implications of this are that the sessions trying to get sync repl lock timeout or reply after the timeout when a node crashed even though FD detected it earlierI might be totally wrong here +Done +would this be easier to grok if it were using Objects? obviously I got turned around thinking about what its doing +At current the profile is only envisioned for use by the Teiid JDBC importer +in the meantime ill tell you that I dont use the relativePath element explicitly in any of my poms so it should default to +issue supercedes Meaning that the prerequisites added by the older issue will get removed and a prerequisite for maven will get added to the parent +to clarifying javadocs +This can be pretty dangerous unless it is explicitly stated that social networks need to change thisSuggestion Issue a WARNING while starting shindig that the security token decoder needs to be modified +are you using freemarker as template? maybe it cause by dynamic attributescould you paste your ssubmit +Theres something weird with the way IBM JDK treats buffers when underIf I initialize any buffer after the intialization from buffers will stop to be supported on the IBM JDK +I need to think about it more carefully to understand how treating compare of nulls differently in the same tree might affect things like logical undo during restart recovery after an uncommitted insert where the row may have moved drastically since the insert do to lots of other insertsI will write some tests related to this scenario and get back +After using m for a few weeks now I am beginning to better understand how things work I believe that this should be implemented as a seperate goal of the assembly plugin +If the command is directly an argument of a AssigmentDeclareLoopCommand its not a subcommand +We could leverage this by not checking for null but catch a null pointer exception though I believe the extra overhead of a trycatch block would be more expensive than just checkingThe performance hit is likely only in the case where it isnt equal null in the equal null case you end up saving a multiply plus a memory reference that is likely going to cause many CPU cache faults over the coarse of a search long enough to matter which together probably add up to more than the cost of the checkAn alternate approach that would eliminate this overhead is to subclass the scorers that use norms and create an appropriate scorer that doesnt require the check +Thanks for the applicationType is null or empty applications for all types will return +Verified on JBT. +Thanks Naela +looks like this actually done but apache jenkins truncates the data so that it is not available from the builds logs +The javadoc output must be unrelated it says I fixed warnings but I didnt modify any javadoc in this patch at all +Thanks Olivier! +Patch addresses the proposed changes +The existing format encodes the number of items for arraysmaps before items themselves are encoded +Looking at this in more detail these two tests have been disabled since the initial commit in +Thanks Greg can you generate your patch from git with to make it svn compatible +I modified myenvironment and the registry so I can run commands written in groovy from my bin withoutthe +could replicate up to clustersize blocks in an iteration +Moved to as has already been to There is a new test framework which no longer depends on these and simplifies writing tests and allows us to define a i +My code generator was initialising the local slots type array by pushing the type of the method owner and then the types of each method parameter +If it does please reopen the issue +Includes context is it will include at +but I left Graemes author tag on it because all I did was his classsrcpersistenceorgcodehausgroovygrailsormhibernatemetaclass can change the author tag to me if you like but I am not sure it is warranted +HADOOPROOTLOGGER can be used to switch between the different appender +I know the patch process means creating svn diff patches and opening JIRA issues +The case of pooling environments is already tested in several other testsI also tried running the two new tests in a XA environment and they passedPatch ready for review +In the attached patch Ive assigned the exception to a variable and only printed it out in one place at the end +Im afraid my files where messed and by Keith +That would explain why the is incorrect +Changes in the patch look good +The patch adds dense returns rollup and void to the reserved keywords +Other than that I suspect the actual logic is the same so its probably just a cut and paste job to create another patch +I think that right way is using of extension points +will disable invokedynamic by default so you shouldnt see this error. +This test passesmvn clean test DtestTestSnapshotsnapshot true to address teds commentsFor the last one I didnt change to warn but added more description in the text these are expected since so isnt really a warning level log message +Feel free to reopen this ticket if this is not the case +Which version of Commons Configuration do you useThe problem should be fixed in version +Could not reproduce myself and reporter claims updated Java no longer has the problem. +Style changes have to be performed in a separate issue If not Its very hard to distinguish the intended change and just style modification +Im obviously missing something but If it only needs it why does it need to export it +This issue is still not resolved +This issue is in CS product and not in the testscript +From what I know at Google they do nothing on a global scale +Some day I will try itBut still dont see anything bad in improving of Cayenns pool +Ok thanks try it again now please +Shubham and SaschaSascha your patch is in jQuery branch at rBTW Its all great to see good progress in this branch but should we not take care also on the remaining bugs in the trunk related to the similar area +In this patch I moved the get methods from functionTestsutil to a new file functionTestsutil and the method get into functionTestsutilI adjusted affected tests that I could find and verified the affected tests where applicable with +Are the current prefetch tests sufficient? I guess theres a bug in that were not reusing the existing Configuration in any case + svn ciSending Transmitting file dataCommitted revision +But for now Im calling this one done. +This may help until we identify the waits on multiple schedulers +For example having an entity named User is not a problem in or Hsql but is an error in Derby + +Let me recollect and apply the patch and will provide an update on thisThanks +The suite doesnt use infras functionality to download dacapo benchmark because i didnt think its a good idea to keep benchmark in the buildlibs directory where all external dependencies go +Hum those two names actually come out of but the reality seems to be worse both bands are named UNDEFINED which obviously prevents meaningful usage of the raster from the raster symbolizer channel selection or processesIm going to cook up a quick patch since I need this solved to move into a bug since bands cannot have the same Andreathe patch looks goodPlease proceed with the fix +Its missing the +Ideally wed fix that as well +The dependency should rather be excluded explicitly within sample applications build files +I will update the patch addressing the comments +I know intellij gets confused by classloadersdifferent sometimes +comments +has a member xcontext which is never usedSorry looks like a bit of debug code I left inbq +Please verify +The current patch fixes the two failures on and two of the three failures on trunk +Suggestions welcomeTests are running now +Alexey any idea on what is causing this +I consider this to be a low priority item for the release +With this change it is expected that NN is allocated more memory say +But I leave it to you to decide +It should return false if the string is empty or only contains whitespace characters +Attach modified for s a user error so close it. +FYII am going to test the v of this patch today and report back +In fact this seems to be causing a compile error in which was initially undetected by Eclipse for some reason +A simple query to alldbs gives somebody a list of the number of subscribers you have and potentially from that list of DB names may be able to identify businesses or individuals who are actually using that service if youre using which is a common couch scenario +Do we let Cell out via theirI think Jon did filters already +Fixed +Its a handy check for people who are building from source +The patch looks good. +This is the simplest solution and prepares them for a move to a different class post +Creating encrypted database with external keyTo create an encrypted database using an external key you can use the encryptionKey attribute as part of the connection url +fixed with svn revision fixed input and output +for the patchThank you Arpit +Added guard clauses at constructor to prevent violating these not null not empty constraints. +Lars Hofhansl yes it perfectly solve this issueYou can close this ticket now +Create a gemset rvm gemset create +Thanks David +Removed all Java specific attributes from the +Out isnt supported +It appears that I may have made a mistake during my installation as I am not able to recreate this problem today +Working with Continuum trunk I noticed that if you +Weve flattened transitive deps pretty much from day one + push to this is not a bug within Gora so to speak but related to an incorrectinsufficient mapping configuration for a specific use case +Unfortunately we dont have control on how resource folders are rendered in the Package Explorer view however it might be possible to handle that in the Project Explorer view so if you are interested please submit a separate enhancement request for regards to the project description since reporter havent provided requested information and because it is implemented like that by design I think it makes sense to close this issue +with the remaining failure in test +Ive renamed to and renamed to in the branchAny further comments +Thanks Aaron! +Modified to run under interactive modeFor hudson invoke examplesbin niUsers will invoke examplesbin and then they can choose between kmeans and has been knocking around a while and looks like a substantial and useful bit of work and its been reviewed +I am starting the zk namespace manager in the regionserver so that I dont have to start one for each region regionserver +See attached test project +So Im just resolving this as rejected. +The Travis build with the feature enabled was successful so if nobody objects I will remove the old code soon +Applied to TRUNK and branch. +Fixed on and trunk +Is it possible that you upgraded the core jar but not the api jarThe removeContext method was added with beta +I dont think so no +Released in +Maybe include a new parameter on the annotation +One thing I did to come close is to extend use the default implementation of hasError and override handleError with a noop +Now the sources and javadocs are included in the repository +One more reason to make them illegal to create +Should I hold off committing this fix and just check in the test to verify that the problem is solved once youve finished upI think thats the better option +Ah nice I had to it for one or two spots will be nice to just keep the oldThat sounds great Uwe I had thought thats where this was going but Michael seemed to have some doubtsIll try it out with the Highlighter changes I have made +Ive also attached the stack trace from the test failure +Changes look good to meSuggestions for improvement remove the line of blanks fix indentation and add for the new method in Out of curiosity do you know if the property is in the j VM +Im planning to do a release asap so let me know when you think youll have a working patch +Converter Test Unit Tests +What we can do for first version +Heres a better one +Failing unit test patch and fix patch attached in +Its just a name for a zip file +EAP requires JDK so it should be a perfect fit +Resolving as fixed +not But as far as I can see its the only one left standing between us and release + on patch +I will work on it when I have time +Just to clear up some confusion trunk is going to be so what youre playing with is probably the table name in the outer map is only used for grouping purposes I think ensemble is a better name than dummytableThe code snippets in second comment are equivalent +I have attached a new version of the stop patch +Assigning to Todd for comment or resolution +The approach can very well be implemented within or +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Reported fixed in JON Sept build fixed in JON Sept build change to in JON Sept build change to JON +type ant all to deploy jar into default configuration +cordovaThis is basically the first step in treating platforms folder as a build artifact +Confirm and wrote only unittests for the +new excludeLibs regime seems to work nicely now. +In any case it is a real issue +v attached w that to avoid the startColumn +Looks like this got incorporated into committed this patch by mistake as part of commit. +If you can confirm that the issue is fixed I will merge the fix toColm +Whenever you feel it is ready +Ok Im resolving for easier to write filter format will be handled bysupport for post requests in me know if you spot any bug +Can we add an isAlive to ZKW +So all issues not directly assigned to an AS release are being closed +Attaching a patch +You explained it well enough I was just trying to justify my concern thats all +v +Yes I use portlet +When an error occur you should grab more information like FFDC of IBM this can be done using Aspect retrieving all parameters of the stack when unwind then like a popup should show up to let user to submit the more useful error report back to you +Added the flags to so they were together +This seems to be a duplicate of +Thanks for getting it this far Doug Ill dig into itIf the failure is always when appending compressed to uncompressed then either the extra data comes from when the block was compressed or is a flaw in handling buffers when we are decompressing +We could update to no problem for the next drop +i added a openshift task to the grunt build so grunt openshift will create a build for OS replacing the with +One retry +However they will not be included in the distribution classpath as they are not needed. +Ive rebased the patch so that shouldnt happen this time +Axis is now at version +Done +This new deserializer inserts a subclass of HTD that also has a different deserializer one that can read v +In order for the DB to know that an incoming query only differs from any other query based on the defined return aliases the DB would need to perform at least some degree of lexical analysis on that incoming query +a minimalistic Mavenized project that should demonstrate the problem +HiI have found that the problem is due to Eclipse? compilerIf you download Xalan sources recompile them with Eclipseand then run the test I attached you get a heapdump and the jvmhangs at least on the platform described in the enviroment describedin the first post of this bug +There is no advantage to dropping the connection +It now works as expectedAs with the previous bug should I close this or it is something you should doThanks again patch against CVS HEAD is virtually identical to that for branch +I believe the fix for this cast error is just to check the key format before attempting to save itI still have no succeeded in defaulting to using the instead of the +I verified manually using the IBM Heap analyzer and the test program supplied by the userI am running regresssion tests now +Hmm thats odd +As per the linked issue +Sorry I have to pass on the hot potato back to you This JIRA has nothing to do with +You need to post your the datasource Test does not exist in the configuration +Thoughts +Now that work is well underway with AS all previous community releases are +Thanks +Ran in a loop of I had to add a sleep of sec before completing the test method I find it necessary as the Server manager still holds the reference of dead regionserver whereas the cluster is down +I just committed this +If the specified id is newer than what is loaded load it else ignore it +These issues are on the same topic however I think the approach here is cleaner +does it impack scmupdate? does continuum return list of changed files? +is also in the same situation +Anyhow where in the code would this behavior be applied? I might not be too smart or the code might be more complicated than I thought but I have not seen anything that would do that +While missing simply renders yarn unusable the historyserver configuration provides just a minor annoyance jobs take longer trying to connect to port +Alternatively we can follow the convention in Change the method name from lookup to put Add methods get and remove for the class instead +is included in both and +I think it is impossible to support this directly with GSE +Is that correct +I got a similar problem trying to validate against a document that used work with release does not work with the CVS version +Number of fragments for example should be per field value instance +Right I prefer to let annotations drive things instead of config files where possible +So you may want to add a statement +New patch with I have used facade as well and it does give you a deterministic test +Verified in JBDS Linux +I agree with you in some sense +Should we close it as wont fix +Thanks Souvik +POSTPONED SPEC Testabletrue Section following methods of the must provide the functionality defined by the Servlet Specification get getMethod is getSession is is and is +zip so now both dictionaries are in the patch +The code looks like using the new parser so I guess Gabriel and Justin are the people that may know how to fix that code is not using the new xml parsers s on the totally unsupported wfs and old xml stuff land so +The following codes repeat twice +Patch is commited +Verified duplicated by +I dont doubt that there is a problem but I dont believe your explanation +fixed in revision Im going to add regression test be able to accept block comments in the policy bug is caused by root cause bug was fixed block comment was returned please check that it MikhailThe fix looks good +Still would really like a test case but its fixed at any rate. +For the Core MANIFEST bug please file an issue against Core! Thanks +I fixed the catalog and payment properties rev +fix committed in r +The class java is from java the is sourced from the STAX project so this should work with as long as you have the dependendies correct. +Nils did this but his perfectly reasonable request has been ignored in the fix for that bug +EJB throwing a is mandatory +bq +Prasad feel free to open it back if you think the patch needs to be backported to also +Ill file a separate issue for that. +Well Ill take it over +Surely this is proof of engagement no +This patch ensures that we dont send duplicate state change notifications to the watcher +Too much day job in the way +this issue is really solved +Is this Issue a Blocker or Critical priority for EAP GA? If yes +Verified fix. + added interface +If the newly proposed extraction feature is implemented then all the tooltips will be revisited anyhow when we make them simply tags within the doc XML source files. +Yes it would be tricky to create the situation where safemode object is not null and blockmanager not up completely +Even though we are not strictly violating the spec we will address this issue with the next not always that though Sanjeeb in the general case +Ok im waiting for new release +Ill likely commit at some point later today or tomorrow unless Hoss wants to discuss the weight thing more in which case Im happy to hold off +By reach do you mean the progress bar? Currently the progress bar does not accurately reflect the progress through the build as all the groovy code is treated as one thing so it will jump from the point it hits the groovy code to +It really looks nice +Is this patch ok for the issue + +Could be a new tab in this page +Fix for NPE in + persists in can you verify if this is a regression from EAP +My production testing shows that there may be another similar race condition +A separate API could be added that just returns the full path but it should be easy to build off whats there right now +We can then from the Runnable parameter cast that to and then get access to the Exchange +The calculation of trends assumes that the measures are done in regular intervalsCould you please give me more explanation on that? I have designed the CI platform of my company in order to save CPU +I just tried reverting in my local branch and doing so does indeed fix the test failures +Luckily I ran the tests myself as well so far one failure is left zombie test +Not sure what I can do here +We ended up getting a patch for the compiler from Sun +Hi Ashutosh are you seeing the latest revision? Sushanth and Ashutosh Thanks for taking time to discuss the changes at length +Not fixed +okay +Fixed in SVN headThanks for pointing the problem +Thanks +Is there anything special that I need to do in order to get these files loaded into the Ibiblio repositories? Thanks +Heres the patch I tested +All tests passed +started working on making L optionalDo you think that could be giving users too many options +Not ready for review yet +Only those fields that are dependent on RMs configuration are checked here as they have to be validated whether they are part of new submission or just being recovered +in should work without manifest rewriting as long as the WAB imports the standard jsp packages +I was hesitating to do this since I have not studied this component code closely and was worried about and multiple rounds of testing since it is a heavily used componentBTW I found this at a customer site and I have given them an identical patch for camel versionI will make this change in the coming weeksCheersAshwin +All you need to write is hdfs reader writer driver for spring batch +Is it possible for you to make this configurable behaviour? Perhaps an extra param in the file leave the default behavior as it is now and let people with complex multiproject structures switch back to the old system if they need it +Does that means the problems is in the parsing of the inChanges are committed in the revision r +Therefore if I build hadoop from trunk and run Sqoop it will not work +Duplicates which has been resolved. +No patch undoing fix versions +You use Byteman much as you would for any Java appThe only thing that differs is that the AS module loading may mean that classes are not in scope when rules are injected +Test failures are not related to the patchWill integrate tomorrow if there is no objection +JasonThank you for the detailed analysis of the problem and the fix +Accepting +then we can move this to +bq +Should probably split it up but assigning to me since welcome page needs a revamp anyway +Change in enhancements release +without the quotes I see its harder to readIm not so sure its necessary since the native stuff has gotten by without it for some time +rev +One by one might be better +I modified the resolver to rethrow the deeper exception rather than consume it +Not strictly true +Thanks Gavin +Well this is kind of ugly because it increases the memory footprint of the build phase that was the whole point of using Iterator in the Dictionary so that you dont have to cache all dictionary data in memory dictionaries could be large and they are not guaranteed to be sorted and with unique keysBut if there are no better options for now then yes we could do this just in +Changed to minor +This is not in the context of making a search but when creating a directory server configuration my directory server is based on the empty DN + +We are running Hibernate on a cluster of servers and it was kind of painful to track the biggest offenders in the database profiling tool +I logged an IBM internal bug for this The defect number is +If theres no sourceModification in the project we must use the package name and the directory +Thanks +Thanks for the report +bq +bulk defer of issues to + +However the rest of the jbedsp integration related tests need to be evaluated for moving to Teiid +JeffI tested the patch and it looks good +Ive run the testcase that verifies the fix +Be careful though +Its only thoughI am not quite sure about how this is related to schema information fetching though +d +Going with were not fixing this instead get to put the jars and a better pom in the repository. +This one is a pretty localized change +even if youre the should send a bug issue to Jiras Jira +The precommit tried to apply the patch from inside so it didnt apply +We should use the same format for all of them +Another problem with introducing is that has to be thrown for each subsequent write request to notify all write failures +Beta build of EAP will need to be tested with Beta of JBDS no need to bundle just certify +RobYou probably forgot to commit your changes from the have some compile errors in the plugin +should be out possibly tonight or otherwise tomorrow morning +writer or calc documents does no harmAbout the order I have it somewhere in the back of my head that the order of files in the ZIP file is somehow part of the standard + +That is problematic to say the least +One would have to ask the original authorBTW one change I did make was to fix the fact that the +Ill get with people about this next week andsee if we want to continue to pursue this. +I noticed in the file there were several references to the original Guava code! I realize what FBC does is expose the indexed comparisons that Guava does not but I suspect Guava is already a part of Hadoop and not yet for Avro +Even better if there is an option to change policy of a file +However it would create of course the wrong impression since the dependency is of course not optionalAlso if you apply the relocation functionality to for example to the must apply some byte code modifications +Works already in where I have done several fixes in the bean component already +Used this patch and the branch patch for and now passes. +Some functionality of parameter management has been reworked. +Thanks for the patch Markus +related and relevant tests mvn test Dtestroot cause cursor out of sync with store messages to the store added during a dispatch resulted in finding messages that were not tracked by cusror counts +Heres a patch for hbase to make it work w jetty +if requests charset is set all the following processing will use this to decode the inputstream +Please try turning it off just to exclude that possibility +Attached the wrong file +no more features making it in to block GA +The creation of test configs by limiting what is copied by the base config has been addressed in AS trunk already as similar issues were encountered. +We have experienced several deadlocks and general breakages due to code being called multiple times +setting to X +It can also be viewed by pushing the Messages buttonIm fairly certain that there are other jiras that address the fact that we need better reporting of these messages +Patch updated for trunk +Ive made som e other changes in the wrapper meanwhile can you check that it still worksI think we can make this into a jca adapter and have both local and xa described in the same Im working on how to do thisAt the moment I dont seem to have permissions to assign or close this issue +might have not initialized due to some exception +v +Issue resolution delayed until following uptake in some time. +po file with the appropriate language code +Please test against any other pages that were broken by +Checked for errors rich components + +Hi Thomas In clients security policy can you check whether you have set encryptionUser parameter? Cause for this exception seems to be not setting the encryptionUser parameterthanks WSDL with different policies that Im trying out at the same time +In our user case when we do not want to skip those killer records +Do i need to add anything else? +While I can agree with the argument that language detection in Carrot is relatively simplistic and we should use other mechanisms if available this patch removed also stemmers and stopword lists for other languages without the ability to provide them so that Lingo uses them when extracting text features +Writing directly to result directory if the file system does not support move +let me also check other actions then +bq +FWIW it might be helpful to know how you arrived at the conclusion that the release plugin expects all modules to have the same parent +I want to check a few last things before committing +in my environment shows the same resultc a c b c ad c a c c +Or it may require a special assertEquals for objects either way we need to be able to test the return value of without having to convert it to a String +Again the only change to existing code is. +An inputformat for use with ChukwaHas been used at Berkeley but extensive tests are lacking +It is clearly ok for mapred to depend on core +Yeah it might be technically impossible but if you can do it for byte you can probably do it for Float too and you certainly should if you can +According to their headers they should be generated by rubygen I download rubygen folder but cant find a clue how to use it +When creating the example in JBDS and then using FileExport +I forgot to mention that I think youll find fixingupdating the rack placement tests will be exceedingly difficult wo doing something very hacky +It should not be included in the distribution and the docs should reflect that +whose ball is this +This was already working just needed a test with. +Committed at revision +Yes this is for core in Seam +Jens to answer part of your questions the log config files should never be in srcmainresources +May be i have problems because of it +Moving from resolved to closed +Sorry its not +Attaching a patch that uses the to distinguish between and +I only created accounts for Ubuntu as you requested +Thanks for reporting this Ive just the protected versions of is and I dont see the cause of the parser issue +Rick thanks for working on this jira +Patch looks good to me +r on trunk +However AM need all the datanode information in its file in order to init +I was under the mistaken impression that setUp and tearDown were executedonce per class not beforeafter each testThis fix looks good to me as the test environment is now cleanly set up beforeeach test test order no longer matters + +Change to throw and +We keep the current white list +This will improve bug reports tremendously +Quickstart change in revision Tested by deploying helloworld and executing ant sendesb +I think importSorted should check that the keys are actually in order to prevent writing any bad sstables +Log that indicates the checkstyle failures in the current code is what happens when I run maven at the top +Applied the patch into master and branches with thanks to EugeneI also updated the blueprint test wiki page for this new feature +Proposed patch to prevent the crash +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Only one line fix +Does not seem to be critical to anyone +This patch deletes the clojure client code +Fixed in revision +Feel free to add a comment if you disagreeThanks Andr you mean for instance to be able to see which were the duplicated blocks of code in a previous run? +Applied the workaround for now and it worked +Patch supplied +Added as a repository attribute otherwise it always appears as undefined +Verified on JBT. +hum well maybe I messed something with the new plugin repo config +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Unfortunately the secret manager only looks at renewer +If someone with more knowledge wants to help or fix the bug let me knowNote that the same problem occur if you try to change the increment of the identity column +Trying it out +Committed +Misspelling fixed. +I think this is not urgent enough to hold for it +Hellothis patch is great but i had to remove BSD license in favour to Apache License i removed some doc tags like author +Username lafr +Please feel free to attach or open a new issue +Its a more natural code ordering than what exists currently which is user permissions then group +I also want to add that the HS API is capable of supporting multiple resultset serialization formats +Classlibs patch merged with the main trunkPlease use provides more mature implementation of shutdown sequencePlease use instead of patch doesnt apply cleanly to the VM +I fixed the problems in that were leading to loss of login info when accessing an unsecured resource after login +You basically use the repository or you dont it cant be mixed and matched +targetclasses as the eclipse project output directory you just type mvn If you want to use +The content assist was temporary disabled in visual part of in +I bet youve been doing that for so many years it just makes you feel comfortable +I will perform some final tests and attach a patch +That didnt quite apply right to +Lets organize an IRC meetup to whoeever is interested and beat this subject to death +Also it would be nice to stop the JT from taking new submissions so the administrator can allow the running jobs to complete before upgrading the cluster +I agree with your nd point I think allowing users to directly stop containers will lead to problems +Im guessing it should be fairly easy for someone with knowledge of the internal communication flows +In hdfs I recommend you add the username associated with the connection to the log message youre printing +Yes called with the wrong key size throws +Applied patch at r. +Maybe there is a difference between running the context in the unit test and running it natively that influences the behavior? I am able to reproduce the problem with Camel but not with +Do we need to commit to +So probably there is no bug in Felix but you can check that on Felix mailing listThanks for the patchRegardsAdam +no test case provided +Hey Arun +I feel that the master is sure to see the new children +for your help +Perhaps for file system it may not be severebq +Is that the scope or only part of the scope you are considering +Ive just answered your mail on the mailing list thanks for the report +Bulk closing stale resolved issues +Also it does delete the data file which I am thinking is a bad idea no matter what state its in +Maybe instead of focusing on the name it would be good to first focus on the actual spec +I think srcmainjavaorgapachehadoophbaseprotobuf is in wrong package +It is the second case +The idea is that type promotion will fix the diverging types during merges but we guarantee persistence per segment +Im wondering if the spring hibernate template may be an issue +By providing a system that is fully functional after running antFrom CHUKWAHOME only thing required download untar run ant default config should just work bin +Please revise. +thank you very much Rupert thats awesome +I see the unlocked fields as metadata which isnt really part of the state of the entity +revision +the test looks good to me +ok without option now +Can you provide details? I dont understand enough from your bug to reproduce itAn example reproducing the bug is best + +Looking at the resulting +Specific implementations will be posted as hdfsmr patches +Ill go ahead and commit this +This patch has fixed for the prob for me +This patch builds on the previous changespatches and ensures the http location context is absolute and includes the interpreted fragment identifier +In Cassandra +That exception is really just a sanity check +Ah interesting +or the HEAD file edited in all the repos +How can i assign this issue to myself? I am unable to do so +fixed in rBTW itd be great if you could post a Wiki page about running Sling on Equinox +Applied +the source repository of the websiteThe README in the project root is having some TBDs in there wondering if its decided to be ASL although I was expecting some kind of creative commons since its about content and strictly code +Went to do that the other day and realized that an example would be hard to show without donating my default action +Fix to follow shortly. +I have tried this again and it works for me +And now have to get back to setter injection or create an interface which dont really need +You can always make it a plugin + retrieved codes from and for +If you see this problem with your WSDL +idea why only judge the booleanintdoublelong before +I just committed! Thanks Namit! +SVN revision id is +as a patch against axis after the last version of that I posted? If that is difficult Ill work with the new version of the complete patch that you postedMy issue is that Deepal has been changing the old implementation every day for about the past week +Tests should run OK nowNiall +Ill commit this once Jenkins is okayOne other thing I noticed is that as they are in the patch dont work well with hierarchical queues +If this is not the case please provide a full testcase that shows that it is a different problemRegardsBrian Minchau This bug has been marked as a duplicate of +Applied patch closing. + +Verified by Tony. +Its a scaleabiltity issueIs it hard to to remove the strong reference +I personally will be experimenting with this new featureThank you Hoss for applying the patches and following through on making this fit the current Solr architectureOpen source development at its finest! +I did a silly mistake thanks for your tests Parsing is now working correctly. +bq +bq +Most likely the answer here is going to be excluding the classes in that package from the test since they arent meant to be serialized via java serialization +Please let me know +If outstanding requests can be reached as suggested by Jay above then it does seem like an issue +At this point with the merge we discussed at todays meeting it would be best to get this into the repo +Committed to trunk. +this should be fixed with query works +The changes I proposed to log message parameterisation on the dev list will supercede this anyway. +committed thanks! +This would be consistent with JDK classes such as or should be ed not ing and doesnt need Wrapper either +This is the kind of thing which greatly helps normal use of Boo by making it as as is definitely acceptable in case of d +Verified by Andrew. +moved to RFPL +Is built upon Helios or Indigo +Im not a committer thereNo new links have come live to manipulate fields or workflow +in other words +derbyall runs did not show new failuresMarking patch available flag +its reasonable to move the call to meter too +This issue is much broader and need to be resolved outside of meclipse first +will commit if tests pass +Committed to trunk +So admin can see this logged event and act accordingly +Thanks Christian it is now fixed in rev +The C wrapper would have similar class structure and API as +So I guess a complete solution is better than partial and yes this is important issue that must be fixed before webconsole is released +Any problems involving filetest are captured by issues. +cannot submit patch i guess due to the status of the JIRA so I am going with the old attach file route +Added missing arguments to was bug fix and was creating problem on relase branch too so I have committed it in rb at r +But that assumption may not be valid as the resource could be any resource +Sorry some combination of buttons lead to reassigning +love simplifying these things. +No integration test +If you require automatic detection of Hibernate mapping metadata be it in annotations or in XML mapping files deploy Hibernate as a standard Java Persistence provider with Hibernate +If that is the case then if we try to check a string is in between and we might not probably get a return result would try to work on this issue when I get time and come up with possible luckPradeep s bug. +The options have a comment but are not marked as deprecated as far as i remember. +The developer doesnt have to enable the interceptor in the The application method does not need to be annotated if we decide the API would be better without it +Identified that the delegators object is missing in context while displaying form hence a proposed solution is to set delegator object in context in s constructor +v attached to track allocations larger than + +oops +My congratulations! I will remove them from the exclude list +Trivial patch fixes in nutchgora branch +Patch to provide fix to compile problem in on z ColinPatch was applied at r please verify +Thanks +There are mapping since the english keyboard is the first mapping is to convert from the English keyboard to the countryOS keyboard its a simple mapping telling only the difference between the reference and the country VKQ VKA +This bug report is old as I mentioned the code has changed since have downloaded the new sources and the way filtered consumers and iterators are handled has been completely changed so this bug may have been solved because of those changes +I know this is not the best way of handling it but for the time being I am able to proceed with this patch +Pending PM decision +Fixed on and branches +I tested +So I added the message to let them know that their specified class does not exists and we are defaulting +Hi AlexanderI just wonder the reason for moving this issue to +Also this string is passed to so java serialization also maintains identity. +Thanks for all people involved in resolving of this issue especially for Emmanuel +This is expected behavior +I think we should introduce a new matrix type to handle matrices +And yes as well as for the fadvice hints +This is clearly not desirable in the case of Sling therefore I removed this automatic setting of the status code. +svn rev +move it to CXF s the status of this? This is preventing me from the development I had with +PR merged in master +Thanks for providing the patch +Well build almost passed only two tests in example module failed. +Applied the patch in a rewritten form +Hi MaciejThank you for reviewing +Staring at your backtrace I see something +committed to trunk +defines three basic +agreed +bq +Did anyone investigate the use of +This should fix it +This is a test page i made and when you click the header it show thewhich in this case is very ugly cause I add a css transition on the header +Committed. +Thanks ChrisYour last patch is in trunk rev +if you can deduce it automatically would you not be better of by using a HibernateHow many systems will have a naming scheme that works with just adding prefixesI guess one way would be to allow you to write a naming strategy and apply it to the configuration before we write out the welcome +Let me know if you want to work on it and I will reassign. +Also wicket shouldnt have any input up to this point as the security filters should all come before the wicket filter + server side invocation process benchmarks jprofiler snapshot +Not receiving any confirmation Im marking this ticket as resolved. +The columns that result No +Thanks to Kathey +People do not use binary alone nor the source pack aloneFor example we use source checkout and build on various platforms +Hi Cos hope the HDFS compilation problem could be fixed soon +Did you mean? +Another valid choice would be but the branch is a little bit newer so +Its clearly a bug that occurs either under load or other circumstances + change please status of issue +Thanks +need to port to fix to trunk r +However Apache servicemix already do this job for you you can take a look at the shipped with sevicemix and you can tell the difference with that in karaf +Nice +Continuum Ive updated the common settings so most components should be +We can then open an issue when we have a clear idea of the things to be done +Ive split the groups but havent split the code into the classes +This was accomplished but committing the related tasks +Its nice when +Please if necessary. +Right I missed that +Ill see if I can reproduce it might take a few days at least. +Lets ship this in +committed to in Developer StudioVersion Build id Build date +V patch attached if all is good Ill go ahead and commit to the branch +There are no more and the logic to support new windowtab is different. +Thanks Ill get to processing the pull request very soon. +This patch catch the exception +What I applied +In a first time we are going to implement the went ahead and logged to track the XML suppression. +Comments from Hanio Show how to access service once it is regisitered via SR and servleto Document requirements default ctoro How to set own Invokero Clearly document when clients workdont work pointer to other optionso How to overrideextend updated as its going to get. +Would like to hear your opinion about the direction +Thanks patch applied in r. +If the VDB or the where the VDB is contained in is deleted this editor now closes +So get the latest from the trunk until we release Synapse Asankha +the hard part is figuring what jar within the nbm is the module jar I think this is just an artifact of using an Ant task in a way it was not designed for +Resolving as Cannot Reproduce. +In my experience the files published to the public site are usually outdated +Ill try to tackle this one +By the way the reason I need this is to allow search screens which result in bookmarkable results pages +OK + +I think it may have been in order to support more complex boolean expressions +Reopening for another bug in +Why does the model think you like Shrek? Is it because you rated Shrek stars or clicked it times? Either both or some of both could make sense +Good pointVerified in Developer StudioVersion Build id Build date +Almost load time +Mind if I spelunk the code for other similar issues +adds a unit test that would have found the committed +Updated patch for current trunk and added parallel configure script execution on all machines +Even with containers that know to pass in the unique EMF name application managed entity managers will not likely specify the unique EMF name +Actually the same issues +either that or say that directionsave entails unconditional updates +The configuration dialog seems to miss some options +It sounds great but it shouldnt be a literal copy cause its a sure way for code duplication and creating a very different version of the little framework in the futureIn the past I was able to from HDFS and make it independent for the sake of Owl project +Cant reproduce it with Sun JVMs which JVM are you using +Thanks Fabian! It worked for Maven +This helps with thread cleanup and other tasks +Deschedule interceptordecorator solved decorators remain +Duplicate of +I will put an update in the snapshot repository to test +What J feature was I using? Ill have to try with to see what doenst work +With this working we will not need to implement byte block +The small desktop container with GUI allows services to be built into JAR files and then you can just drag them and drop them on the service container and presto you can launch instances and interact with any they haveIt does require you to have a somewhere and I did have a jard version ready to drop on the container and launch +Patch file for adding max +Zip all my scaffolding src and give it to you +Working on this one as prerequisite for +Thanks for the pointer Lance +The first patch here moves to make the approach to go support fit current go practices which then had some implications on pomacks github work +this is teh screenshoot of the problem with the scrollabledatatable +Should work now please verify +Couldnt seem to get things working with that ve generated a DSA key and attached d be most grateful if would set up my public key as attached +Patch looks good to me +This feature had a fatal flaw so it has been removed for this release. +My apology I thought you were referring to the package name for +committed to the snapshot branch thanks guys for the review +well it reflect my lack of experience in term of unit test +Theonly reason that the are there is for the convenience of users +Also relative file paths will now be resolved against base URL if available by input factory +Heres the zip containing the WS external service and entry point projects for the +Lars ramI has done for +I would imagine people would expect it to behave in a same way as the +In the script query is well formed syntactically so technically its not a syntax exception and I will argue that throwing parse exception here is an incorrect behavior +bulk close for +Suresh I was looking for that and it appears to be fromI am guessing you can ignore my patch and just merge in +Second patch is fine +Looks like this was a likely cause of the unspecific test failures although I do not understand why +I have to change jxr multiproject and dasboard to provide references to the variable it was broken in all of them +However in my tomcat application this fails +Ive tried it out +I dont want to break existing should do the trick +So I am OK if does the same +thanks tom and ben +sorry invalid bugdocu is as rejected due to reporters comment +So all issues not directly assigned to an AS release are being closed +Ill rely on the Design Time to warn the user +I would suggest a social item in your project to ask that those that Resolve issues also Close them or make it a responsibility of folks to Validate resolved issues be closing them +It will not apply cleanly to trunk instead depending on and patches already uploaded +Ill add a test case m in the process of trying to replicate it in trunk but I encountered a hurdle +An alternative would be to extend forceclean in jmx to allow you to specify both the Keyspace and a CF so that you could manually do specific +Thanks for the patch it works! Your patch is simpler and denotates your full knowledge of the code. +it seems this is fixed in at least the attached test shows that +Issue closed with release. +Please contact the repository vendor. +That could be tightened up at the cost of more complexity but I dont see a good reason +Can you create a sample app that has this problem and attach it to this issue? Id really need to see what is doing to know what is going on +Ok thats fair +The failure is expected here +Len seems to be the same problem we were having the other day +have been updatedconditionsbyline is not involved in the coverage calculation +Here is the testcase demonstrating the problems described +Bumped to critical for CR only to get expression support for the +Thanks for reporting the bug Howard +please adjust the priority to indicate that the Jira should be considered for CR +minor problem +Yes I have also found this to be a massive pain in the assI think Max is trying to optimize things for huge schemas but this has sacrificed usability + +Though in this case so far as I was looking at code the cluster is not showing signs of recovering +Can the FST be made concurrent for use in as the terms dict implementationJust an idea maybe you could use something like for thisuntil mike makes fst or somethingits only in java but you can poach from apache harmony +please use new applying your patch Id like to understand the problem +Patch available +I think we should keep the name and just remove the root class from itas it could be usefulthen fix as it looks like to have more problems tooIll send you some thoughts by mail as Im not sure of what should happen there. +Build fails when applying to latest master +As you can see in the definition of the enum there were strings missing for some newer added +Resolving issue as all have been resolved + please fix the indentation on commit +will not be fixed in b +properties file etcAnd for that you need to use the API from such as the resolve method that is possible the easiestYou would then need to fund a hook into velocity so you can call the resolve with the value and return the result to velocity. +Weve also considered having independent releases for core services but we concluded this is complicated from an operational point of view and too early the Whirr core still needs many improvements and we prefer to be forced to change the services at the same time +Ill just commit it as a separate rev +You passed a NULL POINTER +Will commit patch in a second an close this issue. +bertrand the sudo system on the vm requires you to use the One Time Password system in this case you run ortpasswd on the VM and locally create OTP key to type in as the resultJump on asfinfra if you need any help in this area + caused thisSid can you please take a look? Thanks +Ive updated the RF files to and downloaded your fix + was my first suspect as well +However in the context of its important to not the difference between the pushdown modes optional never alwaysWe have to be careful not to conflate a never pushdown UDF with simply UDF +Then after this is done we could easily write up the bindings +Thanks Rick for the update +Thanks for the patch and for testing things but I have to reopen the issue as the changes have to be committed before we set the issue to resolved +I can fix it easily desired +Good +This is the same as the method that is called during database creation +Closing tickets +Thanks for looking at it AdrianI did some tests and the issue seems related to the presence of the divs in the left columns in fact the big blue box is rendered to be approximately of the same height of the left column +Jetty is going to be doing a lot of package changes as it moves to eclipse so I will wait till that is further along before moving the sandbox branch to trunk +Ive upped the precommit build timeout to minutes to see if it still fails +just reports it as a potential deadlock spot if the threads interleaved differently +Just add the as an to the of your servlet. +It isThanks Emmanuel for this quick fix. +Not sure if you meant it that way but wanted to point that out +Thank you very much for the prompt response +Please find the fix have added enabling osrpath to the patch +Need to add a few more then cluster test +Hence the documentation comment is all we can really do about +no longer extends Message and Field constructors rethrow as enabling the user to implement a simple exception recovery with a single catch clausePlease reviewAre there any objections to committing this patchOleg +Oracle jvm and IBM jvm will both pay attention to the severity levels and print thread dump information to +Is there anyway in which we can send the or as application in the axis ADB generated stub? I need to send either or as application in my SOAP request from ADB generated stub +We clone the results now and lock the WM during the operation. +Nice work +You tell me +The tests at the client site have passed so the fix is working +It can be worked around but the fix would be better. +The original solution didnt work in my env +I played a bit more around this issue and it look like the problem disappear when I run my application within a tomcat container +This supports clients such as Ambari that may want to display the data in a different UI +Reschedule for issues to the next CR +Wildfly will no longer contain the CMP subsytem as it is not required by EE +I committed this +Yup my mistake please close it as a duplicate of +Also is calling up into java land from JNI while the VM is in a GC cycle safe? It must be +I think this issue should be converted to a more general one of implementing some mechanism for binding queues to exchanges which should probably done in the configuration where the queues are declaredNonetheless a convention for simple subscription to a topic exchange might make sense +Ok heres another go +Committed thanks! +This was the class causing the problem and can be used prettty much anywhere jsr is used so this seems reasonable +Ill work on this change and hopefully to make it available as soon as with latest build fromThe exception reported in this bug is not encountered when deploying vms in parallel +that actually makes a lot more sense +previously only for Criteria Hibernate have resulttransformer for hql and sql too. +Has problem both caused by figure out a way to run this spring sample on geronimo +Fixed in Trunk +Another bug is that I had to pass at allI think its bad that docvalues gobbles up so much ram when mergingCant we merge this stuff from disk +I committed the patch +Pushed into commit ceed and pushed to branch +run java jar quickstartshowcase +This patch ported the feature to Yahoo branch +Thank you Claus! +Closing issues before next release. +Will commit for shortly +reviewmakes sense since exchange name durability properties are not propagated to subscription queues created on this topic those properties should not be set on the topic otherwise its confusing +Thanks +why this issue is not detected by our test suiteI dont know +There was an inconsistent read of the METADATA table due to a split occurring in the tablet during the merge +One common scenario that you will have to encode is when we generate a temp file to use as an input +Thanks Jitendra for the reviewI have committed this +Patch adds the working directory and the security realm name to the Web General Page for the deploment plan editor and saves the data +Committed this to trunk and +Mike I have backed out the fix from and +Replay positions from older metadata will be treated as NONEThis will force a full replay the first restart on afterwards any newly flushed sstables will have the marker and future restarts will not need to replay data unnecessarilyWhat do you think +it might not be so necessary anymore but please give usecase where a should not throw one of the three exceptions it already can throwIm not much for changing it without a compelling usecase +I see the bug on both trunk and +I already need to carry the field name corresponding to each column so its not that big of a deal but it just doesnt feel right to me +If you do that feel free to link that issue as related to this one +I just restore the origin log level in the memory now +I have changed to using ulimit m rather than ulimit v since it seems better to control the process resident memory rather than its virtual memory limitWith this change will not work because the test assumes virtual memory is limited by ulimit +Ok decision made to set lock acquisition timeout to for the eviction thread. +core tests +for the second patch +The Landing Page was something like Welcome! Start creating a realm. +Patch trunk +Thanks John +Thanks again Phil +Ha OK meaned nothing to me I understand nowHans is very active these last times thats why I was thinking about his time empathyo +Diff of branch using developer svn repo +Im unfamiliar with jira +Hi Igorattached is a simple test case Id appreciate it if you could review to ensure that it conforms with the style of +After succeeding for to executions it seems to get into this state +if controls controls both center then set controls default position controlsright controlscenter + +Already fixed on trunk +I did use same password file for both +As part of the redesign of the copying API for made methods static changed name to asyncCopy have a order of arguments +Fixed the problems mentioned in Chriss comment +Sorry to come in late the patch has gone stale +And when you try to serialize entire OOME seems like a valid error +does network restart stop all on the network? if not this is a critical issue because existing may not get DNS service they are using previous DNS servervirtual router restart does not stop Vms in the network +If the team dont plan to move forwardI guess it is difficult to work on the mavenized projectThanks for your infoBest RegardsXiaofeng +push to +Ive tried to apply it but it doesnt pass the build +Ill try to send a patch to +Thanks Mithun for the patch +Verified +Closing the bug unless I find how the environment became corrupted. +No response +marking regression test failure and as it would be good to get these regression test issues cleaned up for the QA run +This should be nearly automatic +updated patch to apply +Youve basically done three things Pass memorymgr into mutex creation +Awesome! Was just waiting for your confirmation before closing +reopen for the connection info how are this reusable from ant +does new patch added more testcases and fixed some bugsnamitI agree that will make the code more clear +Thanks Nicholas +if it doesnt apply anymore then yes +What should happen for other event generators? +yes the grammar has handled parsing super columns all along +Agreed it would be better if both plugins used the same mechanism +We dont need the layout attribute as the widget itself layout +The memory manager code doesnt propagate failures across its processing of multiple tasks +I tested it out locally and it appears to work just fine +That said if we have to rush Id say the proposed patch might be a reasonable workaround +I have committed this +Ill try the patch for on Monday +The program that generates these source code do not understand limit +The patch to shared HARDCODE the version number +So if I wanted tests to use spaces instead of tabs I just create a formatter and override the options that I want +The case of a nonexistent file seems a bit more arguable but Istill think that a case can be made for an empty array +So maybe all is well with fixing now +The question is whether that happens definitivelyThe fact that they go to D while there is no real physical IO here smells like a kernel issue to me. +I got it working missing lessInstall via npmnpm install g lessthen run web server against +variable and condition will be added to to control including or not including EAP into installer and what kind of server configuration panel should be used current with EAP as option or just a list of servers to configure in JBDS +The attached patch fixes the issue +Did you mean LuceneIf you never saw the OOME then thats really a bug +I have merged this in onto the dev are also img guys can you answer the above questions +HTTPS NIO is not within the Axis project anymore +Enables parallel updates by default the more I have thought about this the more Ive started feeling we should change this default + +Committed revision This fixes the and Grouping tests. +I think we have an experimental Java implementation here at Facebook +David J maybe you could have a look at the patch for trunk? I think this would be a good feature for +Problem fixedpatch attached +This should have been resolved then +Git commit dacbbafafddcfbed. +Hadoop QA is not functioning due to svn versionRunning tests locally +For example would be another example +Oh +Ahh of course +Alan quick sanity check +Im reopening the issue +Thanks Amareshwari! +Thanks I committed your patch and the testcase to our CVS +btw to be clear Im fine either way I just dont like spurious maven excludes if theres no good reason for them to be there +In particular the tops of paragraphs These two paragraphs may appear the same but they are not +Chris how could we include the other project by reference +The fix was not in was reverted. +I just committed a fix for mdcreplication since it wasnt compiling anymore. + actually I think we should deprecate this ctor in nobody should use that fix version for the deprecated +Assigning to Mike we can now update the SPI +bq +I will have an updated patch soon +It will definitely fix the issues that are linked to this oneSince this is a refactor the side effects will be potentially breaking something during the refactor +Please file such requests in the INFRA tracker in future. +So a big bad reload button on the resource edit page that the configuration from disk? However there are sort of some issues with this right because now the feature type config actually holds on to a list of attributes +verified on +Some I can sort of make sense of but others would take me a while to figure +Ah interesting finding +So is there any news about if this is going to be fixed any time soon especially that its one of the really popular issues +Hopefully others will find it here but maybe you can add a FAQ to the wiki? It has come up before albeit rarely. +patch the patch and should be fixed in the svn head now +failure is unrelated here +Since JIRA and are marked as duplicates of this and since this is described as a potential configuration issue are there any configuration files that you would like me to attach to this JIRA +This actually seems like an appropriate solution for some of the scenarios that you outlined +Lets try this again +Status update Still collecting logs hope to have policy completed in next days or so that will only leave packaging +I am not sure how this can be fixed +Covering the corner cases extreme situations we find is important. +Just moves secondTracker registration ahead of zkListener +Please verify +This patch fixes all templates used by as suggested in issue description +for patch +There is no bug in Xerces there is for that +Ill have a go at this one +The table was a great idea. +Cannot reproduce the issue +its what Id have done no tests are needed as this is an artifact update +Looks like the work has been done for this thanks for that can we close this issue +It would be possible to mount a denial of service attack using multiple clients each extending a parameterHowever if we can set a size limit constraint on the uploaded archive and place it in a protection domain with no permissions then it might be workableFor Jinis current market that of predominately trusted networks with a handful of developers this has no use case for an internet domain untrusted network where there are multiple unknown developers thats a different storyThere is also the case for evolution of Service API with a client submitting later version parameters to an earlier version service + +I will apply this today +When using NetHTTP directly I couldnt induce the error + +Found some rogue +In addition even if we used Lifts Textile parser this bug would still be open so certainly we cant fix it before the release using Lift +Verifed on JBT +If this still doesnt look correct feel free to wordsmith it further + changes to the v of the patch +Maybe the reference output is just wrong? If its sorted alphabetically then d should probably be before f The output has been last updated July and is from August must have been generated before the sorting +Verified in JBDS +Fixed on trunk of. +Basically do svn mv then apply this patch +Brought that up to ms for now will see if this pops up again +Committed to and trunk +I had this large amount of dependencies when using complete meclipse +From a UNION ALL queryFAILED Error in semantic analysis Schema of both sides of union should match destinationidcol col +Maybe we have a record min bug fix +Ill commit it after youve tested it +Updated tests moved to single patch +verify +Can somebody check to verify if its not an issue with my browser +IMHO we should not add support of JSR directly to the Struts Core we have plugins for that +Seam does not support being on the shared classpath +Other than that questioncomment +What exception did you see +As far as I understand you want to let your search for objects +I agree DW is an impl detail and naming is hard what about +We dont explicitly set that value so were not setting it wrongly either +I guess this is really an error trying to delete a project not a project group +I received an error if the app was not previously deployed before running mvn test + ve committed this to trunk and. + +The slaves file is currently only used by the startstop scripts so it wont help herePerhaps the jobtracker and namenode should have a public API that permits particular hosts to be banned +Looks good to me +One tooltip could be defined and just called with component control from any components. +I might port this to the main branch in future of if someone else is interested please ping me. +Scott this is absolutely awesome +Applied patch with thanks to William +Two changes were required on in and the other in itself +Also this was one of many many many incompatible changes that happened over last year +Feel free to reopen if needed. +seems like a good fix +What we currently offer in M was done in a very limited time slot it was the best we could offer in that limited time +Im pretty sure this was still an issue with Groovy + was marked +I was too busy looking for a test case to parse the logs closely enough I had put in an invalid for my and so I was getting a List of Elements rather than what I was expectingI hope this still shows some useful aspects of the plumbing that may need to be looked at +So it doesnt sound like the workaround in is applicable in this case +in the example a User is always an and an is always a User +Issue fix verified by Jozef Hartinger Ales has confirmed that this wont have other side effects +The second run did include the patch +In the absense of configuration for this property is sets the value of the property to the value of the expression +Its current state will be preserved in the tag +Please test the latest snapshot should be solved +Lance there are still unresolved issues in this patch +Patch rebased against latest state of the branch +To implement this we can simply iterate over the collection invoking the equals method on the individual objects +bulk defer to +That did fix the problem but do you have an idea why it was working for one config but not the other even though neither of them had the HEADERSEARCHPATHS set +You need to include the in your as +Extra change to in the patch but was not committedThis has been committed to trunk +As those values are created in the component +The patch was applied at rWould you please verify whether it was applied as you expected +I have not added test cases for this case +Then we can use a simple Combiner to do local blocking +Perhaps this is a seperate issue for the webservice team to decide +Perhaps it is a regional thing. +Not sure if I find the time todayThanks tried to modify your patch to eliminate the one issue +Final patch for commit +pushed to verified this after fixing but could not +Ill leave this issue open for now until an updated version of Atmosphere is available +Need to ensure CREATE ROLE PUBLIC is not allowed before any release +No functional change +This is an attempt to address Knuts concerns about exposing dangerous behavior to applications +Is this not already committed in? We have these patches applied internally and have been running the tool without problem. +However this might not be the final solution +We may not currently report an error or throw an exception when XSLTC cant compile a stylesheet but pretty soon that null bytecodes has to get youI agree this could have a better error closer to the source of the problem. +Committed in revision +You could of course manually scale this having one multiple sources per NASSAN interface +Can I get some more details for documentation? +If the developer wants to use one of the credential handlers he need the anyway rightRegards +But thats going to be a high bar for other things +Did you include the Struts JSON plugin when youre using version? The was moved recently from the core to the plugin +The same patch as an attachment +Hi DomienThe patch looks good in general on the other hand Architecturally daemon belongs to the extensions and not to core As a result new properties the client API and the implementation should be in extensionsFrom a style perspective The header is missing in some files Some files have separators like constants which are not in line with CARGOs styleI guess once the first two points are solved you could check in to the sandbox well then better evaluate and add automated testsThank you for your S +The object inspector itself had no use for the null sequence +is anyone working on this +What do you think should we treat the return string as a in key have the method return a +I just attach a updated patch in case if anybody has interest +Patch attached +Feel free to commit as you go too +Moving this out to M since isnt available for M +previous patch produced +My general view as I gave a while back is that its better to have no logo than a bad one +Take the example of a job that writes a file in one step then reads it in the next +Committed to trunk and +The attachment with the most recent timestamp does away with the serial +However I get an OOME after a couple of commits to a branch +The inequality issue is fixed +Fixed +Thanks Mark for the change +Hi TorstenThanks for the patchAnd your patch left some new added filesfolder like subdir subdirYou can usesvn add newaddfileorfolderbefore you runsvn diff to generate the patch which should contain new added filesfoldersThanks I indeed forgot to check in these files to svn +Heres a shot at the reopen the log approach for +With the iterators I got around this by directly referencing the iterator in code +Thanks Trejkaz! +Actually now that I think of it we never had anything exposed via jmx for the RM? Looks like this is the first bean we are adding +Patch Attached +Fixed render module locally +eh +this is posted pdf provided is incomplete in its content +The contract with Result API is now inconsistentPatch coming +Is there a workaround? Is there anything I can do here to help you find a workaround +Regards Dave +Uh oh Im afraid I dont see any Subversion changes in Jira now +What makes you think that this is the DNS lookup which is slowing down Ivy +Committed to trunk +Agree Ming. +fix small corrections on the presence check for facelet view handler + + +Ilya modify code of demo site application please +Maybe have this in a shutdownHook as wellOf course that does not help if the is kill d or the s machine just dies or theres a network partition etc in which case wed need to rely on the ZK timeout +if we can rule that out then it looks good to me +git diff for for version +from fixed version +Sorry but there is no install stage of Messaging the base zip used to create the already contains to the paths used in this drive our productQA tests +Java has the convention that variables are started in lower case and classes in upper case +and not put all of them in +I have not attached any tests as this default resolver would only be loaded at startup due to the new configuration +I looked into this issue +Fixed in trunk r with the following changes update to use toolchain if toolchain is found added dependency to also be updated as well +can you confirm if this is fixed or not please +How is this tied to? If Camel has a tight integration with then we have failed +So I think the term will be understood by Java usersAnd yes we wont need to worry about changing the WWD example after all +Verified in this to Toni since he has fixed this bug and may answer these are still awaiting the outstanding information for the Release Notes on this one + +But my issue is probably already solved +Move back if you disagree andor have time to tackle the issue +Closing issues +That makes sense I will modify test and commit +That would be really appreciatedMeanwhile Ill carry on refactoring authentication logic on CVS HEAD just feel so dumb for asking this but +I agree that it may be very hard to completely fix this but an option to at least avoid the most common frustrations around this would be welcome even if it is only fixing the symptoms +Wont the server need to present a valid principal in the KDC for whom it will purport to be? Wont the server be unable to recover the TGT from the service ticket if the server lacks the keytab entry? If true isnt the system already very compromised? +patch looks good +As much as I would like to see those long running tests optimised I think that is not going to happen any time soon +Thanks for your useful comments Robert +This is fine +Create patch for I also did a doc build for it was successful no problems +I managed to get this again on maven with release plugin +recent test results +We have the same problem whenever we search for a word which has synonyms defined. +Sure as long as it worksIf you could please create a new test and attach a patch to the issue +Cross check with our its good now as I see currently on stage. +Meanwhile I found the API replaced by in eclipse is not applicable in eclipse +debug isnt debug option and also added custom goal to resume in debug a previously started instance allowing to debug an instance being restarted +On memory usage screenshot daily tests wtih given libs were going yesterday from regular daily tests were started today at and till now they are runningAs you see no swap was usedOne more remark memory usage described for everything on tornado system oracle tests etc +patch committed +Reconsider for M. +StephenNice work on the code review +Thanks for the reminder Willem. +I dont think the second payoff justifies the maintenance cost of a custom caching layer for ivy +Use of the deployment repository is optional +will chew it over a bit and then commit in the morning +Updated bin refs and changed several dependent types in tests. +The intent of this fixed size buffer is to limit memory usage +It should be fixed inI dont see any easy solution for step would be to return the attributes the user is waiting using the names he gave to the server in the attributes list to be returned +If you could try and reduce it to a script we can run on our machine it will be extremely helpful +Case can be closed +Current camel will enable the jmx support by default it need the +Thank you for your instructions +Ill take a look at it tonight +should we create another jira to implement it firstlet this jira depend on it +Thread adds a nested pipeline that yet again adds a DLC +No point making anything extra for out helperspoint out factoriesspecify which rspec formatters to of these configuration options are now available in the configuration file. +Committed also to and. +I would like to cluster them in a tree and use the model to answer the near neighborhood type queries +Scheduled for release +Ok +HiYou are the king WernerI dont have any objections that you include the sample XML +Should I commit +get? to commit +Good jobNow my process diagram will be more readable and simpleThank you +If we change the format we can do PREFIXCOMPRESSIONV +I just Differentiate the Font Color Bold which are the Changes i have added for this issue +Fixed by a javascript solution +You should be all set +So it could perhaps be made publicBill Barker also said he would review the patch on the dev list +For example see Michael Courcy +We currently ran into this as well +Its just not worth the pain to fix it in when it doesnt buy us any real advantagesI have to disagree +Brian Please test it in CR if it pass lets change this issues state to be resolved and fix version to be its a little better but still weird the way it works +I attached patch for as i was working on that +Since posting the original patch Ive received excellent code suggestions and have incorporated those suggestions into this updated patch +Flash player on BETA but not on current at +We now need to wait for a separate release to be cut so that it can be updated within the AS source code +We tried to make adding new features as painless as possible +Doesnt happen anymore +Hi SaschaHave you a clear OOTB use case to test quickly test the issue and the fix? Jacques when you set a breakpoint in the Method you will see that the debuger stops twiceCheers +Thanks it Ill change the ticket status to minorThanksIsaac +VERIFIED +The warnings generated by the script will be annoying to developers like me who will use these scripts a lot +Putting in +sorry Randall didnt see your comment before I posted mine +ChunhuiCan you take a look at this? This is in relation to +Is it not better to use some Camel type converter or the transformer to avoid creating a new instance all the time +Using javap it looks like at least the DS classes are compiled for java +the change above will sink both master and admin in unassign case +The implementation for Tibco will be done in a separate issue. +I suggest the former so policy routing only needs to do integer comparison or bit operations not operations on arbitrary strings +I understand that work needs to be prioritized but please dont plan to just drop this on the floor +I noted that most of the tests from the project got removed with this patch +A missing step in user gestures not a bug is in play on this one. +and now works fine. +Fixed in in JBT +So we could sort memory replicas first then disk replicas +Verified in +Here is our thnks. +thanks for templating the version as well +Resolved reviews still welcome +Karel ok so if I understand correctly removing the dependency from the bom was the correct solution right +We are not supporting alternate character sets in the read and write methods +You are right the probably should remain ignored +This issue may have been fixed and code may already be present in the old trunk +out files now? That way once I get an issue filed and fixed for just that last part of avropartitioned support someone just has to get test outputs +What do you think could they be included + +But this is simple and could be put into as well +Hi Arsen to be completely clear do you want to be able to limit the maximum number of concurrent mapreduce tasks from a single job that run on a given node? Or the maximum number of concurrent mapreduce tasks that run on a given node across all jobs +Closing the issue +The is already coded but the classes will be implemented as part of +I guess Ill have to learn how to use the advanced one +Is that always the same one? Which logicsheets are used in the failing XSPNB we are heavy users of on Cocoon Tomcat Java Solaris and have no problems anymore since the fix mentioned above +properly builds etc +I fixed the table +closing this following comments from reporter +I have a directory that is being used to read files and send web services on the fileWhen the web services times out the file is placed in another directoryWhen it does the file is recopied into the directory at some other timeIt will still have the same timestamp so camel doesnt read it alreadyBut we still need to resend the file +I would like this to be am willing to fix the issue if needed +Committed to trunk +My point was to have all constant strings in one place instead of each class having some of them It could be either interface or class +Hard to verify anything without further information. +No test yet +I only skimmed the patch but it seems to me a much simpler approach would be to add a new overloaded method that takes a instead of a Key +We should have a test that actually runs a child JVM that tries to log beyond the limit and then we should verify that only the last N kb of log file is retained +small unit test +That way I can use that as a demarcation to ensure that the script is not accidentally executed as is in another tool +My latest testing is against +all issues to the CR +Also Im a little wary of moving to XSD especially if it replaces the DTD as Ive had problems with Spring and some servlet containers Resin specifically +You have to register it twice but at least youre not duplicating more than the registration should look in the future at all formatters with the conversion service as well for stringobject conversions +Thanks for looking at the patch Kathey +Or one might use a filesystem with a global namespace like AFS +Also if we are able to reproduce the TT returning port more often I think we should get the Jetty folks involved again +This should have been fixed by the modules jline terminal adapters +The API changes is only related to some unit test with auth as a class is moved a package between and trunk to use to Jetty seems to require a release of CXF that supports this releaseMaybe that is only feasible with CXF so pushing this ticket to future +This violates immutability but the reseed methods currently provided arguably do this already with essentially the same practical consequences +looks revision . +Also had the problem with the ported to windows and proton is not using boostReported it to Norton +Okay Ill look into the timeout on unix +right click my computer then go to properties Advanced System Settings +solrwebappwebimgfiletypes +Thanks +Although I am not entirely sure how to thoroughly test it. +the maven plugin is up to date with cobertura. +Used against Dojo dojo provider for the to return response otherwise its errorupdated to use dojo version use the version no need to override interzs dojo provider coz I already override it and put a quick fix +Patch applied at r thanks for your contribution please verify +Would be even better +We cannot currently provide a way to import and run on server OOTBInstead deployment instructions have been added in the projects completely agree this is too complex however I tried adding the compile goal for builds and that led to OOME after or builds +Thanks for the update Narayanan v committed as revision +Since it would definitely be overkill to make this a configurable option Ill go ahead and close this bug out +now we decided to remove it from main menu this is solution for if it possible we should find better workaround for it +It works like a champion after runs +I will close this jira if you are not able to produce this on Apache releases +Patch looks good to me +I just saw it when some integration test too late for +They seemed to be there when I tested the patch +usually collections are iterable +I think the copy from one dir to another of a GB file is slowing down the machineIt should be a move +Thanks Vinod! +What I am hoping to accomplish is create a territory based database with say Norwegian locale +Adding extra logic to elimnate duplicates silently may cause more problems and isnt as correct as the callers doing the right things. +I would appreciate a review of this patch +I think I fixed it but I dont remember what I did +makes sense because the current geospan might not make sense for all gazateers structuresI agree I will disregard User Story make the interface generic and go for User stories and Once I submit those we can take another look and refactor as the group desires +present in in BRMS +I can add the jars in question +Nit Below looks like it should be class comment rather than internal implementation comment or do you think otherwiseIts actually an implementation comment since it describes the implementation Class and method javadoc describe the external stuffbq +Original posting of the take patch contained incorrect path information on the files +Moved to future release to coincide with more granular project structure +Reopening to port to. +I also added the new velocity version to the Maven poms +All the methods in throw exception if the resource is used before delegate is set the reasoning I see is that you are in trouble if you are trying to use a proxy that proxies on a second look I agree toString shouldnt really be throwing exceptions reopening +Checks for inputs and patch this time +Limitation of RHEV +this +EWS team confirms that the fix should make it into the next release +Then I copy the deployer to the deployers directory of both serverdefault and servercargoconfig +CVS frozen via CVSROOTavail +Fixed on trunk and branch. +Its up to you whether to raise the limit by default or not but the reason Id suggest doing it is to avoid confusing new users +Jimmy this patch look trivial do you want to commit this the branch? Other branches +hi Andreas I have applied your initial patch +The problem was that the pattern for exporting packages excluded the root package +Please test this behavior in the most recent nightly build +Verified on. +This should be applied from the modulesbeanssrc level +Ive fixed it so the build defaults to +is compiled against Java so we can not use generics in test cases +I had no IDEA that wouldnt encode the patch correctly +If the changes are in an release please open a new issue and link to this one + +Fixed in and checks that SLD file exists otherwise prints warning and drops style like aaime closed wrong JIRA issue +Please provide a download or were shut out +If you have long lived temporary destinations this may not be a big issue for you +Thats the way you have to such thingsI added some additional checks to avoid malformed streams in revision +The tone in one of my previous replies wasnt right +closing it +that make this a less straightforward decisionBut I agree our JS tests need love too +Nice so can we get that bug fixed? One should never be redirected to an AJAX call +Also updated clearsnapshot functionality to be similar +Sorry I misread your other commentAgreed +Oops +Can you document how a QE person should test this? Is it just a visual test? What are the steps? Can it be automated? Etc +I skimmed +The patch of Daniel Lees works naturly still not extensive tested but it built it with Compass +Can you take a look at this one? It passes tests and I think its a little easier to understand +Thanks Sumit Pandit Rishi Solanki Parimal Gain and Chirag Manocha to review the patchArpit Singh to r +Patch for use patch file instead +Could we put this in an new JIRAyeah it is an extra feature +Ive documented this limitation and the workaround mark this as wont fix. +BTW quickfix would work I think +Updated affects version to integrate if there is no objection +No longer relevant connector has been replaced. +This version first replaces characters which are represented in hexadecimal format in the urls +There nothing wrong with trying to improve a patch after it has just been committed +First attempt at a patch +Not sure if thats much easier though because what you said is true the realtime branch currently is basically the DWPT branch +Is that necessary? I dont see how the new condition would exercise a different selection than the previous logicOn the kerberos principal shortname and uid I dont know what to make out of it I dont think the current behavior is entirely correct but I see you point +All unit tests including the new ones are passing +Thanks! That is an odd bug in File +Using all and are replaced with and whose iteration order is determined by the order of insertions +r +All resources referenced in solr config files should be relative to the core conf directory +button? still why would you want to have the launch configs regenereated when you can just calculate it and it just stays the same +very weirdWhere the vars comes from? They are all start with I tried to set these vars on my laptop but failed on The syntax of the command is incorrectCan anyone explain the vars? especially the +So it should workThe test fails with and runs successful with against latest SNAPSHOT and the test case passes without error. +Thanks for pointing out +Not beauty but avoids the need to manually parse the URL and unpack the nested archives +It would be also very useful to explicitly define the return type of +After is resolved JDWP agent implementation should be changed to use appropriate VM property to report bootclasspath to debugger +This issue should now be fixed with being resolved +It probably should not fail silently but this is a great work around +Another version refers to vIt also changes the behavior when user interrupts the sleep +The workaround is to add the parameter when executing Sonar +This would leave room for webdav ui soap or anything else in the future +I have to touch the text control once again for it to get the proper focus +Alexey please verify +This is now reflected in as of rev +I am not checking the Existing Application Impact box as I dont know what the rules are for regressions +This removal is not effected by the patch +Moving out of +You need a unique package for each WSDL otherwise the annotation is overwritten see +Attaching a proposal for a fix +to close this issue + +The later fix is more general +Available profiles are now inferred from the projects parent hierarchy +Mathias could you please not set fix version in future when creating new issue in Sonar core +Closing since there have been no other comments since this was marked resolved. +I will also go ahead and enter new jira entry for doing a better job of handling the exception during server shutdown +replace the blog portlet +bulk defer to defer to. +If that is the desire for the packing list to collate then I can imagine a scenario where by if I do a splitmultiple shipment for an order and each shipment contains one of the products in question Im thinking with your patch applied we will get a short count on the requested? But I need some time to apply your patch locally and test that a little furtherThanks around to testing it and found a few more issues +Turns out to be a validation best practices question for now. +Added additional tests to get coverage +Whoa thats a lot of scaffolding for a simple testId rather implement the test case either as a unit test with a mock or an integration test that uses the normal Oak or JCR API to set the permissions to be evaluated +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Works like charm +ASF registered the grant on +IMO with release the API is fairly stable +No s si creena s una bona traducci +Thanks for bringing this up Jacques +Integrated to trunkWill resolve after at least trunk builds where this test passesThanks for the review Stack +Include the deps download within the individual module +This is when running in jruby mode +may i ask why you are using it in your class names +The knob to control job failure due to task failures are and +Finally frameworks can still disable urlhandlers per instance +Fixed in trunk this was due to the name change in FilterMappertransform +What happened to +As I said on IRC Id really like to not change to an interface at least not in +Back on this JIRA I think patch can fix this issue the user faced on the mailing list +As we get the patch created for the release we can also talk to support and the product team what is the best route of pushing the fix upstream for upcoming releasesThanks Jason thats great thanks for sharingAs an update on my earlier sidenote the icon that usually indicates attachments have been granted for inclusion appears to have disappeared from I know I ticked the box myself on previously so please disregard my comment if you actually did tick the box Jesse +r provide System property to set maximum wait time change of resolved shipped +to the trunk as revision . +That sha looks fixed now! Thanks for your help! +This patch can not go through hudson +Your is missing the module entry for +Renamed packages to replace ee with as +The patch also include the patch for converter message and EL of issue +Just noticed that this patch was for and not trunk as such i wont be able to look into it myself +Bulk import compaction and gc must all be coordinated such that the file isnt inadvertently deleted while still being imported into other tablets +bulk defer open issues to to +And Im starting to see opportunity for some API enhancements to support this better +Verified in +I thought that it might have been something to do with having hadoop enabled so i disabled it again but I am still able to reproduce the issue +Bulk close for +Kerneltap has been added on trunk +Merged patch to doc branch at revision Merged patch to doc branch at revision . +Should we change the title of this ticket if its for and not AMQP? Also now that we have the CLA is this patch ready to go in +This bug has been marked as a duplicate of +All the unit tests but skewed join passed which timed out +Bump on Yuriys last question Are performance issues around the number of documents matched the size of the index or both +Vladimir Klimontovich please assign this issue to yourself +FAIL is related to the runtimes the user added when executed the JBDS installer +So if a commit operation takes a very very long time because say one of the web servers is slow to respond or because maybe there are a lot of web service participants to contact the client may give up and claim that the commit request timed out +issue still a problem +The reported problem is fixed with the latest code +It was my mistake sorry for this of course it can be closed +The merge to is busted +rendered should only encode children with renderer types and +I will try to find some time to think about it and Ill let you knowKeep hackingthanks +Sorry Gavin but I thought Hibernate bugs should be posted here +this wont be good enough to use with +It can go as a separate section either under Deploying Derby applications or as a separate section after Localizing Derby and at the same level as Localizing Derby +Will somebody try the patch +I have updated dev guide as well. +I think there are some Android plugins that depend on synch +Patch applied thanks +So if nobodyapplies the patch I can do it at the beginning of the next week +Some of the uses would be related to issue +Interesting +Possible Having a instance participate a transaction with XA resources +Thanks! +Im thinking that the socket is closed right before an attempt by to enter another blocking read and since the code closes the socketHandle in an attempt unblock the readers that the recv call ends up using a bad value in aprsocketrecv +Hmm +for the patch +Deletes in the source readers should be handled correctly +My last comment really should be handled more comprehensively as part ofThanks for the update +the proposed patch +Please go ahead to upload a patch +Can you share it with us +Ive committed this to trunk and +We may need to design a new class to handle null values from hive stores +c Return the return value of my answers +It gave me some ideas which I have integrated into the existing code +Here is a first pass at something more generic it attempts to forward for all appropriate requests +With cores Ive been running without XX +Using the fixed ordering I am also able to reproduce on the trunk not just head +I was learning Persistence using SQL and I got the same error mentioned +Juergen I understand that the Seam BOM version of has not been updated recently but that doesnt prevent an application developer from overriding that settingI still believe my statement holds true as has been available since July and the latest trunk of AS still refers to now I understand you mean the Seam Intl users should have the update possibility they have itIf you add the simple Dependencies line to than some users can just use the version provided by and other users can update to an older or a newer version by just adding the JAR file to lib +This was not part of the original request but we can certainly add it +In order to preserve backwards compatibility a configuration setting could be implemented +If we have all the components in one node then its fine +Maybe we should try to achieve spacing uniformity using a maven plugin +Hi RobertDoes the test results look goodOn your question Do you have plans to port this to trunk? what would be the process to do thatThanks +dont think i want to tackle this at this dont think I should change anything here +I missed a broken test on the first per Teds suggestion added vector size and element index to the seems some tests fail with this patch +Thanks Nicholas +Merged to doc branch at revision +at i in dCch replacement amp + +It might take a couple of minutes to run the test +How the heck could it be used? The purpose of the warinplace goal is to build the WAR file directly from the source tree +Thanks Nigel! +This is rooted at branchesvfinalnetui +Patch m withdrawing this unapplied patch +please look at the home page for instructions on how to replicate the bug Thanks for your time and help and also for the advice about when I move over to version CheersSimon +should not require a to show help messages +The serialisation is now called before marshalling which will only be invoked for non vm transports +Forgot to mention that Online is present only on +Plans are being discussed to greatly reduce that but Im not sure well ever get away from it entirely +We cannnot see any fprintf messages once the java processes die +I extract my code and made a new projectIn this project I built parser for Office and PDF and copy things from the original parserIn order to avoid parsing file content I use an argument in the contextMore over I fixed custom properties parsing on docX xlsX and pptX as it is done on doc xls and pptRegardsEtienne Jouvin +Its incomplete it needs to get the from somewhere +Not sure whether we wantneed this but if so heres the patch +All good points Ralph +If not then we must err on the side of proper error reporting +Good stuff +Wanna try? Please realize that were already short on resources and you shouldnt count on anybody fixing this soon enough for you +I like the simplified logic you describe +Please verify links are as youd expect in Google Analytics for the GA downloads and close this issue if satisfied +For what purpose? We use for importing and creating connections in Teiid +bulk defer issues to +The platform is Windows +Even it doesnt approach something better than cycles will be nice +This lock is simply for auto layout of the cluster if you are going to manually specific the layout it wouldnt be used +So purge only works when all revisions of a document are purged right? Otherwise it must crash +Just had the same problem on an application using ntlm authentication and I can confirm this fixed it +Please dont close this out without addressing ve been trying to do this as part of a build validation process +Fixed in the SVN +But I have not used your Ant build file +RPM works out of the box without additional config tweaks +Dhruba Didnt we take care of that in? +But what you do looks fine by me too +Verified on +Not sure I understand this issue +Committed the patch into SVN trunk with r and into the branch with r +You can either write a filter which sets it up and clears it after the request and put that at the start of your filter chain or submit a new feature request to the main Spring project to create one there +drop this in the root jetspeed source directory replacing existing this file in srcsql this file in srcsql this file in srcsql to mention you will need to edit the to set the path to your posgresql jdbc under several different environments +I would prefer not to move the extended test to +I was thinking about storing special state in zookeeper as well +But that seems ok +Once an EOL is encountered the lookahead is not required +UndoRedo now work fine. +Patch attached +Ive committed this to and trunk. +Since does not exist this pom is useless +Is that a request for a new panel or is that something that you developed and would like to submit +Mark any idea +Julien Im reopening this ticket because a warning message should be logged when a property value is incorrect +Moving to and +SorryLeon not Chris +resolved issues for already released versions +It should be +Regarding the proxy annotation problem +In Opera I see a kind of flickering for half a second + +Fixed +Thanks Zheng! +Try with the build that chandan has created and see if that helps +can you please attach a confirmed that this issue exists in the Railo beta +Could you include the endpoint interface so that I can try to reproduce the problem +There have been a few situations Ive encountered with custom iterators where it would be SOOOOO much more convenient so emit a few extra keys then to ensure the reseek initialized itself correctly +Your comment should be enough for now +Test case maven on master +Oops +We could then fail the job if a particular map or reduce violates such limits assuming that reexecuting the mapreduce would lead to the same kind of violation +Modified permission checking in Rev +kindly review +Linked with +You are not responsible for the antiquated infrastructure I have to deal with +Issue may already be resolved may be invalid or may never be fixed +If you need it sooner patches are encouragedAs to my feeling about log levels see +EAP Beta not ready for inclusion yet +Please delete +OK but if you dont use XML youll need to submit a small test case which I can run to reproduce the issueOK Ill take a look at but at some point Ill have to fix this issue too if I want to release +Works fine without knox in the the middleAdditional informationIt is broken not only for data upload but also data readLooks like any call to datanode would failYou can not upload or open a fileThis has nothing todo with the fileYou should be able to try this with any file +so apparently fedora supports it +Andrei Im not happy about this but after flying over the link I think that we have to add this +You are rightI have a new version of the patch now that only does that when minimum versions is enabled for the CF +But it was when using endpointSo I will fix thisThen you separate args using space just as you would type on the cmd line in the shell +Committed at subversion revision +Hany you have remaining minor tasks for + to remove +Thanks Mark. +closing resolved issue +Porting the integration tests is +Id consider this behavior to be far more intuitive and closer to the idealistic BPMN definition +For now I was content with the smaller goal of generating the site with mvn +for review rajesh +for the patch +Mass closing all issues that have been in resolved state for months or more without any feedback or update +It seems it has been fixed inCan somebody confirm? +This also doesnt work because it is hardcoded to accept only the item id pathSeems to me id mapping should be an optional step and shouldnt be inherent to the generation of recommendations +From what Im able to determine everything seems to be parsing ok +New patch adding dynamic field wildcard support to uf +I would love to see support for the far more advanced +Ill need a little more time to get the sample working using it +This patch removes those checks +Few comments on the patchThere are lot of javac warnings +I borrowed the class from Ians sample code +Released in. +Thanks for this contribution +Gurkan I will try to combine both might take me some time +Ive removed the from the hadoop build and will create a shim for this +This class would need to be created for every serviceBean but may be its not that big a deal +You can then goto Administration Language Editor and customize that text +No this isnt clear in the spec +To be applied under java branch To be applied under java branchThanks and Regards applied the java patch at r but I removed the two tests for expected that the patch would add to as these tests failed on the RI +This is the output I get from mvn clean install site site filecTemp on the attached project. +fixed and tested +and are now much simpler because all the pooling code is gone after was committed +If the problem persists after that one is fixed reopen this one +This is the simple fix +Awesome! A patch against the current trunk would be great +Here is the patch for describe functionalityI have added data for product idThanks to Rishi Solanki +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +bulk close of all resolved issues. +That downloads all required dependencies into my local repo and does use the proxy settings in +Ill paste another shortly that will apply to is just the logging data but Ive taken one of the weirdo cases and formatted it so its easier to think about it +Thanks! +Since there fluido evolve and remove all refs to a bootstrap reponsive cssI will make a pull request with updated info on githubapache Emmanuelthe patch is here as a proposition to hide non sensible information when using a phone tabletmaybe not the right implementation but provided as ideaeg preserve link to the main project but hide other reference to Apache Fondation Project. +This issue has been fixed and released as part of release + +bq +Your patch is in trunk revision +No need for release notes for fixes to features new in this release. +However in general there is no way to detect that another configuration class defines a option with the same name +Now it is +Need to research the schemas better to determine if this would be needed or not. +The feature works but the technical implementation is perfectible +see comment for like this narrowly missed +But I feel that design is a little awkward and not too elegant +problem was wrong DTD declaration +I wrote the source code from the JSR it is licensed under the ASL but the Validation class has its implementation dotted +Could probably use better testing +added accessor methods so that applications need no longer directly set or get +so I think what we can do is require since that seems to be the only place to get SIGKILL and SIGINT on Windows and I will them in for release with +I am also not too happy with having Random used in tests +If one server dies the next request will fail over to the replicated session on another box +See the discussion in for some additional commentsregarding the reasoning behind this proposal +Remove support for specifying a bundle name with a messageerror keyAlso one of the users who downloaded my patch asked for support for specifying a property name to be associated with each messageerror key sent with the redirect +Attaching diff which applies changes to the Camel parent POM and adds an Apache Karaf feature called Currently using snapshot versions for jsonlib and ezmorph bundles as was only closed yesterday +Thanks! +this looks good to me +The includes and excludes do have two stars the editor stripped them out and made the text bold must be some sort of wiki markup processing +Update m pending build in trunk before I apply it there +I got Kafka to compile and run with Scala +Was a dupe not wont fix. +Excellent +Id just like to make sure everyone is on the same page before doing a bunch of migration work +We will track progress there separatelythanks a lot +Looks up to me now +Thinking about this again if we commit this we have to make sure that this would work with serialization as well +Hi TobiasPlease can you explain what Immutant is +is the patch which fixes the issueThanks +regions calculation for CDI validation is improved +for now we can go with this +committed +here is a preliminary methods were already there but they were storing the information in the UGI itself rather than the JAAS Subject +This change has introduced a testsuite failure when a datasource is packaged in an ear with an ejb jar +Sorry for the confusion +edited this comment was talking about file permissions for chmod for ssh files not related to this issue +Anil its all yourPS to all all the tasks in this list are unassigned +Please update the attached quickstart application so that it fails +I still dont like the thing with toString in but as long as the build works I can life with it +Updated without Destination with a delayed check on set +Just updated the patch with clear message saying includeSnapshotsinclude +packagingtomcatpkgsrcmainresourcestomcatconf we must not use these lineportalportalrealmName +The field addition is synchronized to avoid concurrency problems +BTW when is JDO scheduled for GA? +I tried to break it down as best I couldI wrapped up my eclipse project +Available in +I tried to get grammar into one lexer but Antlr would not have it +No demand for such component at the momentClosing as Later. +Apart from being a bit clunky I made them separate so that each class can evolve without masterregion usages steppingconfusing on each other +I just committed this. +ring is balanced and datanode is the same +What is the algorithm used to identify the written bytes from chunk from server a as oppose to chunk from server b if the data are stacked together? There doesnt seem to be an easy way to identify if one of the chunk is missingOur deployment has collector mapped to exact number of datanode deployed +Closing +a release date would be really handy to see how old each release is at a glance +In absence of sample code demonstrating the problem I assume it is one of thetwo problems causing text to be dropped in certain circumstances which havealready been fixed in CVS maintenance branchPlease supply a code sample or test yourself +New patch addresses your feedback above +Patch looks good +if not bound then +I just pushed this to and trunk +Uber is not supported +This change is needed also for Instead of saying Check for +Is it possible for the DSL to also allow you to pass in the Class as I had it originally as an overload to the String version? Then we get the option of type checking in the IDE + as I forgot to add a pull request for switchyard core +This bug prevents me from using because Weblogic cannot compile a JSP if the tld it uses is out of synch with the classes in +The build fails if there is a problemOn a side note the tests generate a bsp directory in the cwd which really should be inside the target directory +The ant build script also added so the tests can be started via antThe ant build script installs compiled jbossts bits into the local repository and maven tests then run against them +Ive not heard a contradictory opinion +The cplist code needs to be cleaned upas it is not clear which counts are counting bytesstore blocksdisk volume blocksWARNING this fix required changing the disk structure which will result in a cache WIPE +Let me it +this removes all the files that are svn ignored +Committed +Ok sure Im alittle bit sad we wont all be seeing this feature in tomorows build witha bright red bow wrapped around it but heyAny idea when this might get revisitedAgain thanks for giving this so much consideration +Tried it this morningYour solution solved the problemThanks +The current solution does not completely work due to the incorrect use of relative +Well see whether theres enough stuff to justify a release but generally speaking we recommend to upgrade to RC straight away once it is +Others will fill this module and refactor their own code later +defer all issues to +Incorporated Raghus comment +Ooops close by error it is resolved unit tests are good +Please let me knowThanks for your time Doug +Why comply only with the minimum requirement when it would be easy to encrypt the other sensitive dataThe guidelines state These data elements must be protected if stored in conjunction with the PANCardholder NameService CodeExpiration DateHave I misinterpreted the PCI document +I do confirm that it works with rev +shipped in +Changeset edfedccebbbbdffAuthor marcosjunqueiraDate Message Updating to new sonar version updated +OK heres the first cut of this +This also depends on else people will have to modify things in order to just enter proxy server settings +Yes I totally agree to The trunk should be our unstable development code base and may even sometimes be broken that is the reason why we should get a release out and then create a branch for bug fixes etc +Looks like the following shims methods can be removed fromusesisfileinputsetgetcompareTextsetget +Looks good +New version of patch with of the scan of EJB annotations s requested by Denis I applied the patch +corrected misspelling clarified fix in the trunk +It will always generate a warning about the task tracker since the old value was set by the codeThis is not true +Latest commit is on ropalka branch to allow Richard going on with his branch work +The patch remove valid code calling +works for EAP GA +Glad it worked so easily +Rather there is not any selection in combobox +I made it volatile +In other words except for the Java language override case where obviously the subclass position of a method is significant it shouldnt matter whether a method is placed on a superclass or a subclass +Committed on trunk as well on revision +In the mean time Im resolving this to LATER +If a sufficiently large independent community of record users develops then it might someday make sense to split it into a separate project with its own releases committers etc +Thanks Chinna Rao Lalam +this cannot be done by default i think because mappers can have side effects +But this is very common that team A is also adding new feature and creating SNAPSHOT version before release +close it. +I guess it expose Pig to system environment settings +Make it trace +Since both the binding and the cloning are done by jQuery it is possible but Im not sure about itPlease provide a quickstart app to be able to debug it +I am trying a fresh checkout to see if I can build Jims current state of the project +No hurry +However I frankly have no clue what we would need to do differently there those errors rather indicate internal state problems within Jencks +It sets permissions on first install or failing that when upgrading from earlier than per second patch looks good but maybe we should change that to +yes is true +So possibly you need to revert the changeThere is no testcase in Hive covering it +i committed this patch to as well. +Defer all unresolved issues from to +FYI the Geronimo tx manager is being consumed as a subcomponent of the Aries tx module +I cannot access that SVN location +Emmanuel may have fixed this so Ill try to confirm the bug is present or not +has a unit test +Hence an attribute langen should be added to certain tags +Im so +You are perfectly right +It is injected on all managed classes inherited or not +Ill try changing the handlers in the security module to do what dims suggested +I could just always grab and use the second if the first one is offline +no more feedback regarding the remaining accts and they have not created rb accts so will assume those dont use it. +Your browser and proxy caches have probably expired by now + trim option to language expression in the DSL +Verified with JBT. +You are correct +Needs more javadocunit test of the new enum but I wanted to get feedback before finishing it offGarys repliesAgreed that Android could be there as some kind of String description however I dont think its a backwards compat issue and would prefer to keep it simple +Logged In YES useridNow with a fresh checkout things are wel again +So from security point of view obfuscation may be evenmore dangerous than storing passwords in clear because then everybodyknows that the settings file has to be protected in some other about using real encryption? For example with a password basedencryption scheme where a master password that has to be enteredinteractively is used to decrypt all the other passwords +added to all relevant tokens +I will try to add a unit test for this +Others might +FWIW running the job as the yarn user work for meTo me it looks like there is some issue with user resolution +Toby or Przemo are you planning to act on this in near future? If not Ill take over and commit the patch in a few days +Its up to you but unit tests extensions and examples may have to be fixed if you also take the Reworked link hierarchy patchIm sorry if the issue title is not very meaningful +Patch attached to remove the fat +The good thing about removing the method is that it improves compatibility +Can you provide a reproducible test case +Thanks Knut! Ill file an updated patch +Then we almost wouldnt need a shell script since it would just be java jar +to trunk +Pull it back in if you think different +Havent verified but should be fixed now +In the meantime I think the biggest problem is that there are no test cases +On commit Id added missing files and then afterward backed out +Created individual parsers for each Teiid DDL statement +Thats definitely something we are interested in +This is somewhat fixed +Some discussion about this can be found in +On the one hand I really would like this feature but on the other hand I do not like the idea of having one of these created with no children and then floating out there for some indefinite period of time until someone finally decides to create children under it +for the patch +note just to be clear my confusion is about who maintains these going forward and how these quickstarts are to used by user is a new pluginfeature required or will they just add this for now and thus we leave the full integrated solution for JBDS or how +Quick fix to handle only jar files revision +Also used it for the mail component and all the test are passingIf it gets accepted I can convert also the other components to use this new class +Verified +Maybe we should just do this as the tracing sampling I proposed for +This is by no means enough but its a start +I could have gone either way but thats what I did +I dont think we need a new issue +Target tables with beforeafter rowstatement UPDATE triggers +DTP fix targeted for DTP June +Thank you DagCommitted patch to documentation trunk at revision +Best regardsve also investigated this issue and agree with Andrew and Ilya RIs logic seems very weird and its hard to reproduce it +Can you add permissions for mattmann to do this too +This is the correct one +Moreover if we keep it in the document it keeps parsing and processing +At the minimum it sems to be inconistent given theres some outstanding bugs in this area. +Were in agreement my bad +The patch as it is now would be uselessI will admit that I dont completely understand the issue description or Uris first comment but I did understand the bit about a dropdown with all valid fieldtype values +Please provide a minimal schema document that illustrates the bug along with the error message from the parser +Looking at the branch and current trunk I dont see the accessors for the proxy property in the current source for +attached patch to fix and some in trunk in rev and +Thanks for your email +Excellent work thanks Aaron! +This is a known issue in tuscay +Committed +Query Plan and sandbox and Hive driver in the Teiid Server and similar creation of the connection profile in the Designer I did not see the above error +It looks like the VC projects are still generating the default symbol files instead of +These are finished and passed all tests +As long as you cant login with incorrect credentials its fine +needs returns false in the Java codegen now so no place checks are generated +Also I was just wondering if we could have the hostnames printed rather than the IP addresses +The searcher interface makes kind ofsuperfluous +in your application do you provide branding for orgnetbeanscorestartuporgnetbeanscorestartup possibly also orgnetbeanscorestartuporgnetbeanscorestartup +By not yet supported in my previous comment I meant not yet working +Bulk closing issues resolved in JEXL. +I managed to NOT get an error after cleaning up EAP and doing a new deployment +This is confirmed as a blocker for BRMS wrong issue This is NOT a blocker for BRMS but it should be resolved as part of resolving else needs to be done with this? Can it be marked as Resolved +Id appreciate any suggestions +postponing to JoramThe last part of your comment is exactly as what Peter Horvath has explained inIs there anything else in this issue? Or shall one of them be duplicate of the other +DUPLICATE with am having the same issue but the workaround works only for but I have installed +Thanks! +Including a regression test. +I was just looking to take care of some of the modeler cleanupOtherwise Ill let you know if I need anything as I dig into it more +Ill put this in the queue for now +merged. +of mails I get currently with question marks instead of text but the PROBLEM REMAINSI changed the description of this issue +Having said that Im not one of the XML gurus here but to me it sounds to be something different then the original problem Bill had. +I was not looking at this correctly +Ive committed this. +I am not sure if we can achieve serializability and portability still +More work on this is needed +Section of says clobber remove all built objects except jars +moving to BETA are cleaner. +is the sld supposed to work at all? i cant seem to get it to do anything thus far +Patch applied on +There is no high CPU during the debug session +Changed consistency level from ZERO to against +we could also fail the test here if the random generator is out of date +we get an error but laterin another method call than on embedded +The improvements weve done seems to have worked we have fast feedback for all jobs and the jobs runs weekly and ensure API compatibility between components. +Now it works for me both with and on Windows XP +Looks as expected +lib jdench binsh echo CLASSPATH +Hi Mark Ive marked this as fixed for as well was it just the tags that made you think it was only on x or is it giving you trouble merging +Attaching +The patch Ive submitted adds the version information so the the stack trace can be more useful than and layout versions must be the sameHowever one thing to consider is whether this should be checked every time the datanode starts up rather than skipping it if ea is disabled + +should I submit a patch or do you prefer to do it yourself? I dont mind either way just wasnt sure what the process isThanksIf you have a patch ready feel free to attach it +I am using Maven on OSX +I think the content +Hi Yasin thanks for this +I know that is not a simple Task and the reason for this feature request is not the gui in my caseYou are right that the performance in Portletserves are always poor but in our company we must support a PortletserverThe main reason for portlet support is in my case the authenticationauthorization access not the guiIt would be a big feature if click supports a alternate super class page like +Between the two options the former is often prefered +Here actually i should have same cluster ID for this two namenodes as they are belongs to the same cluster +patch for OK +should we look at disabling these tests until someone has time and interest to fix this issue +The error is from an old project I no longer have access error code to throw was not done before when translating exception +gmailed to config to scott via gmail +Any plans to fix this? Otherwise i must temporally do a workaround inAFAICS the none handling was removed in Impl +updated moving current into while me know if that works +If you are doing this do you want to add a alter database also +The fixes also clean up the error message when the property has invalid values to conform to the documented settings +heres the fix forTests are passing I will commit soon +HiCould I get an indication ifwhen this issue will be addressed please? Is this seen as a priority fixAgain thanks in advance +fixed in and +Only when autoc commit is false +I thought that I have to have the permission to browse through the svn from the top down to the foundationofficers directory but obviously that doesnt work + we should really have left IA to only +This decreased his level of urgency below the critical level required to get a patch into good shape and he is reluctant to give you guys an done patch since it would just increase the work for you +After all Ifeel that this is the most direct method without any side effects +Since this is relatively low frequency use case with file and http connectors the workaround is fine +So it informs you that there is something that wont happen +Lets continue the discussion and work there +Any reason why Version cant live inEven if I group infomonitor into one bundle there will still be cyclical dependency between this and mbeans unless we move Version out of infoAt present I am simply placing each package in its own bundle +Just wanted to check in on this issue is it possible to get this in by the next release? Let me know if there are any questions or concerns +issue was fixed by new errata respin. +Thanks Aaron! +Oleg could you attach your error log +These blockers and critical issues are resolved but not verified +Alexey please verify +Running on Windows . +trust feature is no longer available in the source codeMarking resolved as the better solution that was discussed on the mailing list was implemented. +Was unable to accept the fix resulted in test failures. +The web interface for Solr is pretty nice +But the shutdown order does seem to be problematic in this case +Whats the status here? Have you tried Vladimirs fixes Vivek +Well we need to wait to add this one +Sorry missed your log attachment +Hi SaschaI tested your patch and its cool tree is working fineGood startingRegardsAnkit Sascha Ankit ErwanI will wait the new version before commiting + +Ping +Nice response timeJust to be clear I just svn upd which put me in rev but the only actual file that came down to me was the one relevant to this change +I am working on the and and +You sir deserve a beer and a cookie for that one +Fix pushed in svn revision this has been addressed right? It doesnt quite look like what you implemented in your commit r is what you initially proposed but all seems well anyway +I believe the build order used by Continuum is slightly different than Maven command line as Maven starts with the order of modules in the pom while Continuum starts with an alphabetical list +Im not completely satisfied with it but at the very least it works very well +Ive attached a WAR that reproduces the problem in trunk +I swear it was real +applied at +Have removed the changes for CLASSPATH +argh that is correct +If not any idea when will be released +The GSSAPI mechanism uses that information to get a service ticket + of fixed for a long time. +To the question does an appdef event get triggered if the agent is down and you add a resource to inventory +Emitter Id better say +The attached patch makes the file compatible with Java by removing the code that handled methods with parameterized return types a new feature of Java +I just committed the above patch +the real issue was in the result mapping prob another jira issue for that. +woops sorry to have missed that +Should we just disable that check? should have the results soon +No need for a patch +But thats probably an unreasonable fear and maybe is just a minority I happen to be part of +If people feel the additional tests I have added are not necessary I will remove themI have run only the upgrade suite and it passed +Please use the mailing list for general questions about how to use Jackrabbit +The reason of the issue is style which is not more supported by Gecko +It would be nice to have them back too +Thanks Brian! +Cloning this because I cant reopen since I am not the Reporter +The fix is in the current snapshot version. +The exceptions being thrown are inconsistent between authentication and authorization failures +Looking at the commits it looks like we are disabling the JSF validation instead of just telling it about the variables we know +Any comments on the threading? Does it need fixing for concurrent accessConfiguration should indeed be +The attached patch fixes that one tooCommitted revision +Then what will be returned for ls on har paths +I would like to rename it so that it wont be mixed up with the existing file +I like the fact that I can add quorum servers simply +Closing all issues +Okay +SvanteCan you help to review it? Import this patch and run command mvn P to start the integration test to see if it can build successAre there any output files in target? svante to review +Aint gonna kill no body if we get caught at the gym wearing black socks in shorts +Pushing back to +changing fix version to +Its assigned to PiergiorgioIf elastic search test fails it may be because of changes I have recently committed +Same for current Chrome that there are many examples of ajax working in this cannot be a systematic problemUnder what circumstances are you seeing this error? With what component and with what data? This error is usually indicative of jQuery having trouble parsing the string returned from the ajax callIf you could include any sorces files and steps to reproduce that would be greatAlternatively you could wait for the upcoming M release where weve update the jQuery version and see if the error persists +Patch to fix findbugs warnings +I would add a new +But these arent C functions they are C x ones +Some minor update to the script +Closing all resolved issues due to a successful release +Existing tests now run with latest patch +however the ant commands which copy the dependant jar files are commented +The root cause was the cellar hazelcast bundle did not had an header generatedI added the service that are exported manually and the problem is now fixed +Okie so I guess we do not handle rethrowing of a checked by an unchecked exception specially +Verified. +And it occurs to me that if extjs were to change their license in a way that makes it incompatible then worst case this component could move out toEven so itd be nice to have a view on whether legally we can use extjs as currently licensed if we choose to do so +This patch upgraded Pig to try and clean up the dependency tree +Juergen I can assume that you have any huge process in your listener that takes a lot of time +But I believe the statement +Referencing a class which doesnt exist at runtime is not really working +In the meanwhile can you upload the patch on jira and grant ASF perms +Still looking for comments on the addremove by category +So by design it is not implemented for concurrency access to invoke startMonitorMulti web application in EAR is in AS initialized in parallel threads and it results in race condition while Seam is starting +Im interested about this project idea for the this year +Nicolas Want this in? Is it ready +Tx for the updateVerified that tests are passing on my local box +on which version? im using postgresql i have done some more research my case seems to be a special case +Im on Solaris +Hi He Yongqiang I feel this validation is not possible on client side because the client side validation cannot ensure the jar presence in task tracker classpath +All unit tests are passed in my machine +Looks like both issues are addressed +This is by design + +Svante Added another revision +This probably isnt going to be changed for. +It updates the name of the property from currentVersion to patch looks goodI changed the Fix version field and set it to the next release +ClosingNo problems reported. +This takes away the simple option of creating a Command directly and passing in an instance of and moves further towards a model where users have to specify a complex config file before they can accomplish simple tasks +I will file a bug in and follow it up there +The default max value is MB +Robert committed to. +Closed with the release of. +Exactly what I was looking for +The files are diff patch with revision with a couple of minor changes fixed a build error in and reformatted a couple of comments. +Theres an issue with deleted user if a user has shared a filter and then this user has been deleted then its impossible to go on the issue filter management page +So this does not impact the Facility +No way to disable the popup +And they want this in a single data format +Do we need to remove it there tooI will also remove from trunk but this does not lead to deadlock because weve simplified how IW RAM so this particular deadlock case cant occurStill we should remove it so I will +This will require allow instantiation of Page Fragments declared as abstract classes +Thank youBest regards at r +Also you can include Luke in the conversation +Looked at the codeprivate class extends implements should all extends Serializable +Perfect thanks! +Updated patch attached + the way to define the th submenu as of the provided demo gsp +I dont see specific references to Vmware +That said though I dont at all understand how a change to could affect client Blobs +classpath +MergedThe tests depend on corrections yet to be made to depending on. +We should prominently link to it from the About page thoughI agree this change should go in +Page flow and jsp demonstrating override of expansionAction being ignored +As part of trying to understand deferred constraints I did a bit of searching on the Web +Proposed patch +Released to the production system +Why not just set those properties to false on the command line during the build? Im guessing the upstream build is setting those properties somewhere external to the build itself so we should be able to do the same thing +fixed in released! +Revision . +Thanks Jothi! +Can you please check your artifact or group ids for dashes +Reopening to edit release notes text for release notes +Will fixretestrepatch ASAP +Fixed patch +Deferred to +This broke backwards compatibility in a big way as many jars compiled against the latest no longer work with weld and visa versa which has been a big problem for extension developers +Does anyone anticipate having time to review this patch proposalIll try to look at this patch this afternoon +I think youre right that it would be better to use a in this caseI like your idea of having implement Callback +Sorry for being stupid +Committed removal of in revision +Its nice to have a name per process for situations where we have multiple HDFS Clients per machine +We should consider adding tests for this +Please if it remains an issue. +Could you please give us more info? What SCM? Error Log? etc? by now I cant say whether it is related to for the workaround SudharmaSo we have to run that for every project we idea when we might get a fix to this one in a releaseThanks againDuncan +Applied to trunkSending srcmainjavaorgapachegeronimoaxisserverSending srcmainjavaorgapachegeronimoaxisbuilderTransmitting file dataCommitted revision . +before this issue can be resolved we need to get the actually released jars from b right now we just got a remoting snapshot +if it is you can check it easily by giving the script any extension like +shipped with +Multiuser mode will be carried forward in a fork up on github. +Committed rThanks Matthias for pointing this out +Thanks Fredrik! +Am I right that this one is dead? +closed +I think that since karaf uses jaas there is no point in using the user admin service +Drz are you planning to work on this +I applied them but still get the errors reported above +The reason is the inconsistent of the order of columns in reduceSinks rowschema and the order of columns in reduce sinks output column arrayrowschema is used to create a file output op when splitting mr tasks and the output column names are used to init the inspectorwill upload a patch soon +This solves the parsing problem but not the blocking problem +Okay Id like to hear more +Have fixed the auditor but the exchange details do not carry the correct information they are returning the message type as being for a normal response so will need to create a testcasebug for leave this issue open for now until the switchyard issue is fixed +Thanks it looks like this indeed is caused by many deletes which end up reducing the size of a subsection of the HAMT in a way that the current code doesnt expect +Copy the to see what Im talking about +Please try this one +The javadoc for setget should describe what the session id is used for as should the description in +All tests including enterprise tests passed with this +Yeah looks like this has been upgraded to so Im happy to close this +I guess were not doing regular snapshot deploys +There is always b. +its fine to also backport we should really do this as we nned this for further improvements +bq +the visibility of the getter is not included in determining if something is a property or not +Im going to mark this as not a problem since its not on the Galaxy Nexus or the Nexus S +The original append in was quite buggy but has been improved to support hbase +I believe whats happening here though is that hte displayed value is an eclipse classpath container path +Submitting another patch +That is what I tried before but because the parentQueue doesnt know which queue made the call we dont know what to reinsert +Adding a would be fine as well of course +patch can be applied at it +Ive added that but I still cannot reproduce the issue can be reproduced in the following environmentSun JDK first commandLink works as expected only the first time it is clicked +Below is the define of Stringpublic final class Stringextends Objectimplements Serializable Comparable fact also should implment part or all these interfaces +test passes on local findbugs warning is probably bogus all patches seem to have warning even if theres no code to trunk +So a couple of regressions check +Also the use of described in was repeated. +Adding the upload bundle jar this time + +I did +Arent they required any more +This looks a lot like +Jeff would like to have an additional field of the old bugzilla number that this issues was migrated from +Ken that is waaaayyy too much wrote a snippet for you copying the usage of the block +Not clear if this is an ibm jdk specific failure though it could be just a test that occasionally fails with the sun jdk also have to investigate some the cxf binding component with the ibm jdk SR on linux +This patch allows to build xercesc libraries with native CC compiler on +Anita this problem should now be fixed +Can you run one of your night long tests David without SEQUENCER in stack +UPDATE The problem here is that I had done a fabriccreate first +Alternatively can all the messages be retrieved on application and cached? havent been many inquiries for a generalized solution here hence Im closing this issue +Related fine to do the rest + +Attach a draft test case +CVS diff of with a is the patch to add htmlunitjar in the differants classpath will commit the patch this week that class has no external dependency looks like a classloader issue then +Why not an option to keep inputs for failed reducesAlso recall +Thanks Robert! +Rajesh could you update your patch since is in and put the diff on RB +bq +FYI +I filed +Ive marked for per Daniels suggestion +Here is the patchAshish Ashishwhat about using javascript in to get the current year? This way next year we will have less work to do he he +MarkApplied in at repo revision Please check it was applied as you expected +The third one is for a server that I previously configured then deleted +Justin please review the change to wfs +My preference is to either stick to get or getAlso get or get is also okay +This is the ftl wih my thanks for the patch! +I suppose that was my own fault Looking at the comment I was well aware of this change +There are appropriate unit tests which run as intended +Have you got any further with these errors? If not please can you send me details as I cannot seem to reproduce +Thanks for the advice well try this on monday and Ill let you know if the fixes the problem +In commit bbaaeefabc the logs are being automated +It might be due to a lot of factors + +But as they are in one long running conversation Factory executed for one time currentUser is filled with data but I dont know why the currentUser data lost and becomes the newly created User +A separate implementation in the same project is probably the easiest since it will have very different properties than the existing source class +Ill wait for that new release then. +But its needed a clear statement that they should update them util the version or is released +Committed revision . +Will in a minute +The current patch includes a test for this in which I realize may not be the ideal place +No +We need to change the group id +Defer all unresolved issues from to +resolved since TRUNK too +I have some development confs which use xinclude with paths and it used to successfully pick up the included files from my conf directory +file mentioned in JIRA +This patch resolves this issue +There are problems the ROOT context is and modcluster expects the REMOVE needs to be sent first +It doesnt specifically do any harm but want to discuss with you first +Just thinking about this some more it would be easy to write an around advice that accepts a list of domain object superclasses such as Entity +Sandy Can you please fix all findbugs warnings in one patch instead of opening too many jiras? Thanks +but not the reverseBut here is something from section page Domainvalue OPTIONAL +Ruwan Thanks for your prompt response +Because all amount values will be pushed to a map the map value will be overwritten if the payment method Id are the same +Have edited out all references to idlj from the various files to see if this helps build +This is how I solve this problemVery simple solutionIts already used in my project and works fine almost +The interpreter shown on the diagram in are wrong +It would be nice if those are the standard translations +I am very curious about their solution! +This shouldnt get in the way of real modifications of the code +patch fixes the problem +Ported to +bq +So I hit the long resend email verification but no mail is sent +If for no other reason than it makes bisecting easierDIf a couple other committers sign off on it then Id feel fine committing it +Will commit shortly +Looks like Ant Javadoc supports PrivatePublicProtectedPackage as well as access +Committed to and trunk +Sorry LukeBut Im unable to read the PDF file sounds like its not complete or something +Ill wait a few hours to see if this gets fixed +We probably should +Later we can reduce the priority again +Coming up a unit test is hard because the situation rarely happen and I have to simulate a flush only partially complete +Removed the test from both. +Running through hudson again +Thanks for all the feedback Seth +Yet another update with a cleaner code +Might be worth a visit here to explore? Im happy to set something up. +I will investigate this matter +This is at least about to be resolved one way or the other +I dont think it make any difference as the attrinutes name is converted later to an OIDIs there an issue with the existing code +Presuming that you intend to comply with the restrictions I see nothing in the license that would disallow this +I have known what to do +on theBut I think youre right that theres something else going on +Do you have time to verify in real clusterWe can discuss names of variables later +It isnt but it is far safer than assuming all uses in output will be caught +Done +Will check it in to trunk now and later to branch when all the noise resolves to something branches +The patch looks good to me for commit +But we could probably still achieve these two benefits while using a single class for looking up everything cached about a field? Ie the could return bits but null valuesExactlyAnd with NRT and increasing number of segments the number of synchronized lookups per segment could really start to matter +Attaching a fix +I figured out what is going on +Patch for reference demo code is not yet committed +I had this issue in +added support along with unit tests for it +In otherwords from a normal install +fixed in +Fixed in revision for in revision for +As an alternative we can Action and register assignment as a new action type in +Deleting line in seems to help +So I can coordinate this with you +Ive set it to yeah you cant set it yourself youre not thick +The only difference is that the column indexes are in the index file instead of the data file. +Ditto code patch attached with following changes good to me +Latest email was a few minutes you still not seeing anything at your end +weve decided to remove libbfd usage + copyBlock use to get a Socket so that it has an associated as required by if there is an error performing the actions that the user asked for we should return failure status +I should try it first sorry +You may want to if it builds if you have the jlib property defined +This exists in the undocumented configuration right +Please provide it as soon as possible +Subin Can you attach it as a patch? I think best would be to create a separate module for this something like and put the code in srcmainpython +Closed as a duplicate of this one +My only question is in regard to the formerly deprecatedmethods +Closing all resolved tickets from or older +Looking for feedback +Reopen to merge fix to +This weekend Im out if you can wait the first days of the next week Ill rebase and clean everything +There were a few minor grammatical problems which we can fix once its checked in +The issue is do we currently now when a bug is going out of scope +suresorry to make you this is my first time submit a patch to the communityI should have known this before i submitthanks for reminding me +Ah right +Regression tests added +Your patch has been applied to all branches +on and transition all resolved issue that did not see any further comment in the last month to closed status +I havent looked in trunk yet +I was getting there! applied the second patch in! wagon to version seems to have solved this issue for us +Another note this started to happen with virtually same setup only after we moved to much faster server from Win IIS where it never happened to Win E IIS +This is completely duplicated by another issue so Im resolving as a duplicate. +Once a skin has been requested by the user we parse the stylesheet and store the stylesheetnodes with the skin +Release audit warning can be ignored +However given number of users me already has I believe we cannot just change default behaviour + and the other ones no longer. +Currently the solution will be to properly escape the filter path with +could you tree diff those wars +Solving this issue is not trivial and will most probably require a big refactoring +I was concerned about the lack of a last location +I have just added a similar bug which has the code fix +Cheers for the delayed response yes this works nicely +The sense in the exampleA hores dara ja deuen ser a la fronteracertainly seems to be by now hores dara ja deu haver arribatthe sense by now might be inferred from the construction involving deu haver arribat +for the to trunk +I have attached a patch. +Please see my comments on for details +minsAnyways good point +The Mongo users of the API would expect findOne to return one object as the Mongo API does +Latest patch from exact maven commands did you run +Having done a quick diff on that file vs the version that I am working with they are identical +Thanks to infra this issue has been fixed now +AFAIK it has never been a problemPlus feature speaking we should merge SETPROPERTY and ADDNODE as one permission in ECMS +Patch looks good to me +One more detail the callbacks are testsorgjbosstestremotingtransportsockettimeout +Patch for trunk using we agree that this patch should wait for as the parser is already ok and backwards compatible in so no need to do anything? In together with the other unicode changes we will update with +Sometimes this corruption looks like described in +Then you need to file a bug report for container +to trunk. +obsolete +Also include fix for +I have attached two patch filesThe first is for commons validator +Attaching patch to include the improved wording for JDBC Java +Heres a patch that does a performance test of the reference counting mechanism +Any progress on this? I am running into this as well after having just upgraded from to +Therefore it +Actually part of this was completed as part of Roller scheduled tasks ping systemBoth of these items are now better suited for a clustered environment due to the new task locking code that went into RollerSupport for clustered search is still unresolved. +Somehow Im not able to successfully write a unit test demonstrating the issue +moving to to be handled by the new +Thanks Chad +No longer seeing this contention. +Thanks Sid! +Nick please verify and close this Jira. +Connections view with some configured +Agreed a test to see that our does not work when use thru hive would be good +Applied in svn revision +The existing filtering code applies to files only I presume this would have to change to operate on streams or the dependencySet assembly would have to go through an interim file stage + for patch +What if Consumes had a special parameter for this purpose if we can get a good name +For Reference this is Rev The part regarding the shopping cart was for a custom app and thats been taken care of for a while now and not really of much use back in the community projectThanks for finding the time to look into this +Hmm not yet Im on Windows machine and checked it on all possible paths there +do you have a use case for that or only for properties +Thanks for the quick reply Ryan! I have the inclusion disabled but is there a way to skip the generation process either via a maven plugin option or in the? Also thanks a bunch of your hard work on this product +Thanks. +I realized that we rolled back to jdocbook from due slow with setting of apply to false in options works for me +Looks fine but for one comment that Integer rowKey row line looks redundant yes but without it the code is actually boxing ints millions of times creating Integers each time +Release Notes marked Not Required as this affects an internal Engineering release and is resolved in an Engineering release so no customer could have been affected by this issue +Also realised we need to do the same for sending doubles floats from NMS +Once the ALTER INDEX support is added this one should be good to go +Agreed Ill open a new JIRA for this and schedule it for +recut with the latest version of the original code in ve just committed this +HiAs I told previously this problem doesnt relate to Hudson plugin +But may be its with rename that throws not creating a role which already exists. +included a backport for setting table properties +If you want this use applyLayout instead +Thanks you Venkat for all your effort on this feature +I only ever use Dojo for AJAX +However only the GUI shows duplicate entries +Ill take a look in to that we do this as an private to then the only way I can see to tie the created session back to its Service is the local socket addressHowever I dont think this is all that easy The sessions local socket address is given by the socket itself but its most likely that the acceptor bind address is just a wildcard address on some port +Moving issues that arent going to make it into to +Has been fixed by wrapping the into an. +However Im not sure we should fix cqlsh but maybe rather not send the bits +possibly related to +Palo can you post the details of your finding to Tellurium projects page? +foreachentityas described in +And timeout exceptions from this deadlock arent logged as from run where all timeouts were set to min +Thanks Shai +This would allow users to define what they want loaded first on a per application basis +cannot reproduce on windows so I would say it is fixed +This patch is definitely not going in as standard behaviour for copyFile +But I found that Im not comfortable with it as I was with the axis tcpmon +Thanks! +It could be an environment problemWhy is the encoding assumed to be ASCII? Its right that those are definitely not ASCII charactersIn emacs I have a problem with the visualization of the uppercase gamma greek letter but it appears in geditIn any case compilation and HTML generation from the Javadoc proceed fine +CXF has a copyable scheme for using the +I will inform you when the issue is resolvedBest in Trunk Revisionfor your good workHow and from were can I download version of Geoserver with Fixed in Revision ? see this is the reason why I comment you that the geoserver Nightly is using +Hi VladimirI think we both understand the spec Here I propose two solutions +file a bug against? +They seems to be fake because is alwasy set from BTI environmentIm going to test this patch this night and retrun to this JIRA only if any issue will be detectedThe patching itself was testedThanksVladimir Beliaev +That is after Im supposed to have fixed it btw +Ill shut up now +Hadoop is open source community project +One can only add spring security to MVC apps at present +Pls check if needed in also +You need to add the service class dependency jars to the created service archive at the third party dependency page +Closed. +Distributed Search is still relatively rare though it is certainly picking up the pace +will commit this evening since trivial unless any dissent +reopen to set proper fix proper fix version +Cool this would be a nice feature +Reviewed all these issues and marking them as closed. +More later +Lines with errors are commented as following error all resolvedfixed issues of already released versions of Roller. +Patch attached +Can you please rebase it +bulk close of all resolved issues. +We are putting our main development effort into JAXWS +Thats inaccurate we dont override the renderkit for all components only a few and as can be seen from the stacktrace theres relatively little Trinidad in here +at the same time was curious since Jason proposed a fix as the solution and was wondering how it apparently solved the issue for himmany thanks didnt see the problem you described Asankha +Does this mean we need more than one cur to support the right behavior on WindowsNo there is still one and only one current working directory for the running process so there is no ambiguity about a relative path like dirdirfile +So any specific build configurations by default have to go outside the cvs sources structure +The cleanup depends on some basic operations working correctly including delete of an individual instance and delete by extent +Im not sure if there is an easy way to test this +Random binary noise would be expected to produce a ratio of +Add more detailed description on configuration xml fileRevert the since there should be no change on this file +HiSorry for the awfully late reply +My original workaround was to simply pad the column nameYeah thats what I meant so ok better formatting will help +The build process succeeded after I deleted the +Is this in trunk as well +Its definitely either the connection pool or Hibernate taking that way to avoid loading all error codes is to extract the file from the spring distribution jars and edit this file removing any beans for databases that you dont want to include +Here is the patchRegardsMridul is the screenshotI have also added a link in Component for thisRegardsMridul sounds like a good idea to me +The reason I raise that here is because it seemed very similar to what is being discussed here +bulk move of all unresolved issues from to +Also the connection cache in RPC Client uses the address as the key +Did Shai already fix his test problem +I usually work on and off with the project but I havent worked on it atall for the past few months +Is it right +I dont know why MINA and react so differently but after resolving the issues and giving it a shot on the extra large boxes everything went very smoothly. +I renamed the new files that remained +Thank you Tian +Updated patch previous one had a few holes +Any updates here? Did this solve your problem? Can we close this ticket +The patch doesnt appear to include any new or modified testsbq +reopen if there still is a problem +no newnoteworthy items for Beta in to correct +actually have no interest in the orclrevpwd attribute and dont use it +and +Improving the error the patch to improve the error unit tests +fix committed to trunk in revision the changes to branch At revision +Mark it as wont fixas it doesnt seem to be an issue for someone else +Tests ran cleanly for me +I am fine with these changes. +Testing demo with myfaces I have found two problems +Added logic to remove the model object annotation when a property value is being set back to the default value and no other extension properties exist in the annotation +Its hard for me to get a clean ant test these days but I dont think any of the failures Im seeing are related to my latest patch here +Verified on the latest build +Hi Subho I am working on thisThanksShameera +as I said the test didint seem to change much in the factores between your and the actual implementation when I use a pre allocated buffer or not +Now that work is well underway with AS all previous community releases are +Decided this wasnt worth doing +Karel any? Else well this issue with the explicit transitive dep removed +Page has been fixed +I have some concerns with this snippet of code we have no way to tell if the original author allows us to use this code if so then I think we need to add some attribution in the headercan we send a mail to the author requesting for an official authorization to use this code +Anyway the aim to have it in also +Marking the initial problem as fixedThe second problem showed aprthreadexit +please categorize this is issue into one of the components +Do you think I need to add something to the events returned to the clientI just checked you changes automatically return correct to the clients also +Patch for various components into specialpurpose folder +I plan on tackling this in the new year after I can complete work for some year end deadlines +Something weird happened when I updated from SVN I think because of all the shuffling of package names +Thanks for the quick review +Would be better to understand what causes it to spin and to send notifications in normal rate like for participants +Just committed this +I will rewrite this jira +They are both loading schema through spring Resource interface so reusing one loaded schema instance is impossibe +Kabir I made the comment after the merge. +Rebased patch and improved error message for the +This is not supported +well I just figured that for our AS view we could hook in before the normal WTP dnd API and ask the user without triggering all that magic +To work around the AS issue I think we can copy over existing files with usernamepasswords in them +DavidIm attaching all sources in a zip file just put them into the opensearch extension directoryThanks for your interest +BTW it probably is a good idea to split this into separate common and HDFS patch +Reenabled the test +I encountered the same issue as well +nevermindAnyway the right place is to fix poms and notIm reassigning this issue and will fix it asap +Thoughts +Dan I just got back from vacation and saw your request +Sorry for breaking the test +Feel free to pull back later +Im ok to continue on the dev list but please do not take anything personally +I think in current trunk youd get Unknown datum type M which is somewhat more informative than the message +I submit that there are multiple problems inside the job and somehow there is a problem where the main job configurations do not get transmitted to a subsidiary executor +Gail Badner No In these cases we have used HQL +FYI this eliminates of the CPU burn in my app +Yes they should and if possible make them pass as well. +I am just observing that removing author tags is not as innocuous as it seems and would propose that the issue be reconsidered +The reason for the request was that the way that generated were written a java compile error was occuring because there was no declaration in the abstract base class I cannot remember the exact details but will post more fully with the example +However we are still debating if server should also support these methods +Assaf said there was no good reason +This is being worked on by Alena for this specific user. +Could you kindly provide me an example of X code that results in incorrect Java codeMany thanks +Thanks +Attaching which contains a test program for various rpath scenarios +Only formatting changes in description +By the time ships the version where we drop support for I suspect theyll be ready to add to their repos as well +But if I try to access the directory directly with the complete url everythings works fine +Looped times using the patch and they all passed +does this actually set the external variable +It takes a whole lot longer time to wait for a dependency to get released before you can consume it so then its ok to add it higher up as a first step +The was an easy fix so thats what my patches job not provided case is more difficult since thats generally handled long before the logic to classify exceptions for an exit code is applicable +bulk close of resolved issues. +Verified in IR + +This is the aar MTOM optimization works only in case Changing from Document to and adding node programmatically using the ADB generated bean from wsdlJava tool +Everything works fine +fixes this problem output is reduced fsck runs faster and its much easier for ops teams to build tools around +passes on local +Fixed starting by we reopen this one? From the discussion of seems has not completely resolve the issue +Cancelling for the moment. +Im going to let him or someone else check that this doesnt cause ws problems +Ive set this to be included in JBDS since is already done and accepted as part of the PRDERD and this is just more on that same issue +and commited +I go through the open issue list like weekly +It has a mixture of and issue has been fixed in newer dont remember the original issue in JIRA but this issue is duplicate +Jason If you mention in the bugs field andor summary one of them triggers the review board appending JIRANicolas Any chance of the girls pasting an update on the issueGood stuff +I just came to realize that we have some silent listeners here + +dimsBut Dims isnt that just another argument to break up the eclipse specific stuff into a separate directory? At least I find it quite inconvenient to work with compiler excludes +This is still a problem +This lets me focus on the functional changes under review +No details +Ok now thanks to some suggestions from Karl we have a new in our code +tmp +How can I enable there? Is it a system property? Thank thanks for the info Charles +These new tests are a superset +Thanks for this patch John. +Suggested patch against the current trunk +So sleep to dali in eclipse does not allow it then we cant fix this in anymoving to. +service restart +The auto message says that there was a test failure +No objection! Did you already make changes +Changing the on the file system didnt result in a change on the browser +Im not sure how that happened. +This will be a configuration option for the +Moving back into this will be in the first dev release +Icant see that being a problem +As proposed let us handle the bulk of this on +yes after GAOpen a jira for it and ill track down the magic for doing it +we should also have one server side to manage common abstractions like userids and such +Committed trunk revision Committed revision close for +Jacob your use case makes sense for sure though we dont have such a contrib in Solrs codebase currently and if we did wed just hardcode wire it into just like we do the dataimport JSP pages currently +There is no HTTP response in it no exception of any sort no handshake errors + +Attaching a patch +Marking it major this patch also gives a fix to the number of reducers in sort example which was changed due to +I will go ahead and commit as soon as you make the changethanks +But I think there are a couple other issues with that process so I want to defer it to a later JIRA +the accounting folder the test branch is missing the org folder +Please review +If you need to use a topic can you please describe how youd like the concurrency setting to work with topics +Shridhar thanks a ton +Id really like to see this addressed +Needs one change in the +Rob does this fall into your chunk of functionality? If you tell me where to update it I can create the patch at least +push out to +This version uses the right bean +Can you pls regenerate it? Thanks +Does this still affect you on Flume +Would error results accomplishthe same thing? I dont really see how but I could be mistaken +I was using release plugin version +Can do +create a new Map +Complex issue +testcase and patch are welcome +Alexisthanks for your supportPlease rsync times a day two times a day is fine tooYour mirror will be added to the list soonGroetenHenk HenkThe modification is doneCheersAlexis +Juergen your quick response is much appreciated +Looks like includeBundles entry in +Sorry for the delayYes I think this would make perfect senseAnd if we point the localization attribute of the metatype descriptor to the web console in files we should even be able to reuse the existing in filesWill you do the patch? Thanks +This patch solves the problem +Still what do you think of NOCONDVARS removal +Bottom line is if you need this code is a fairly focused API from which to have to get it +Generated code uses the identifier rec several times +Linking to which hopefully is the only issue stopping this from fully working +Not only the syntax is dont believe the mechanism has been tested as good as regular tables +So is the problem only with anonymous checkouts? Is that why I can work with CVS within Eclipse but not using the instructions on the web site +Could you elaborate on your meaning? If you mean fix the constructor call then as I wrote before I cant see how to fix this issue as it compiles as Java fine and since groovyc should compile Java it points to a bug internal to GroovyMany thanks for your assistance +maven directory and running the build again from scratch +r on Ubuntu Linux x +Patch looks good to me just waiting for performance testing now +oops didnt mean to Resolve w last comment +Thanks for the ample explanation of the changes +this per my previous comment +Ive committed this! +Any comments? In the current patch addIndexes goes to the update handler directly so not through the update processor chain and commit is not called at the end +Is there a reason why this isnt using merges +Patch applied on and all issues that have been in resolved state for more than one month without further comments to closed status +I just committed this +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Ill try to confirm tomorrow with a more taxing example +Trunk patch which contains the for branch test test resulting avro for adding default values for merged schemas. +Try to undeploy it first or use the redeploy command +You could use defaults and just get the corename from the discovered instance dir +I agree that at the moment perhaps none of the available enhancement engines are really stableFor Metaxa and I havent tried them or seen examples of their usage +However when I tested with JDK too and it worked +I see that there is yet a patch available +Either Im wrong or well encounter new troubles in this area soon enough +We also need to figure out where the raster merge fits into our plans if were going to try that before embarking on feature validation +Please find attached fixes to eliminate them +Correction we need to wait for a Jenkins first +looks good +Should we not just pass the bytes to a Detector if we have a but nothing else looks promising +I havent tested topN and dump jobsBy the way having a SUCCESS file in the current crawl db will also throw errors for the url job +You attached the log of the groom +Thanks for the good debugging work +One correction +The patch for the ref manual looks goodCommitted revision I think this is the last one please confirm +the patch for Luo NingSo I understand correctly you are saying that rather than a timeout of for as we suggest elsewhere you say that it should be set say to the current default of minutesAlso rather than keep raising the xceiverCount as more regions are added youre saying that it should be bounded at the current default of ? otherwise the OOME about cant create thread etcHow many regions per server with your patch in place Luo Ning? understandsxceiverCount must be set to avoiding OOME with Xss option and memory size a java process can use it can be set more larger than im using in my DNhow may concurrent open mapfiles in hbase how may xceiver thread in hadoop +Yes please open a new issue if you see it again. +going to commit + I think this is a good idea +I am not sure whether these will be fixed on or just the trunkAlso I have reported multiple issues on this page itself +No such constraint exists for any other like +The patch is sane and looks good to me overall +The same fix should be applied to on trunk. +then this is not a realistic scenario for updatesdeletes right? So it can hardly verify scanning from multiple files compaction effect reading old data etcIt will always update data that was just written before its ever readI wonder what is the use case +This has been changed in CVS already for RC We just log the exception message at debug level Juergen. +Ive just committed the license patch +based on Samants patch on new trunk code base. +So I will deploy by Monday and test againRegarding bits vs bits its because I have many different kinds of servers on this cluster +Closing stale resolved issues +Heres a doc patch that adds this list of client dependency jars +Hi BobOK I will take your advice do you want to upgrade this to versionregards Malcolm to scriptaculous which includes prototype MalcolmI have not yet updated to the new add script which fixes the original issue listed in the first commentI will wait for to fix this issue +I think it might have to do with the allow dynamic method invocation property but I dont have time to check into it tonight +studioeclipsefeaturesmaven +I dont think that will be enough if the version is different maven will use the version definied in the filename not in the POM +Do you have idea why this pass in +The comments regarding missing clustered testing were largely out of date +Whats the advantage of cutting a corner now +When our deps make breaking changes when we upgrade them it can cause havoc in what our users thought was a simple patch bump +Note that the proposed fix uses a method introduced in Java SE +Its not totally the case today because we needed to support the old hardened stuff while was built but not moving towards harmony or unification is not an option IMO +It is also needed to monitor the health of the JT +Or throw an exception and not use the configuration at all but applying some best effort approach is causing for trouble +A first attempt at some fixes for this issue have been committed at trunk revision The commit note is as followsinitial fix for and which aim to improve and consistify the in situation in Roller +It works. +Closing resolved issues for already released versions +already has a few places where the response format is still marked as experimental and as subject to changes in the futureMarking the output format as experimental is just trying to be safe +stack Melbourne +converted Mark Schnitziuss example from the ML to a failing unit reports that he can reproduce using Thrift rc too +new patch looks good +If it was the fix could we close this jira now +Dont volunteer people +Which clients are you thinking of here? What we need to avoid is is overload on any failover restart operations at v +Need to clean up client handling of this case before can commit id suggest +Thanks Allan +Actually I directly close them without passing by resolved statusHappy to see you handling issues +Need to add more unit tests for diff report after renaming operations +Closing this out for now. +At this point the is true +This will be using a new compiler that is based on JDT +Nevertheless it can effectively emulate data corruption during transmission and verify the patch works +Hi Benwhen will the migration to SVN take place? I need to commit changes to JANINO and Im unsure whether i should commit them to CVS or wait +bq +Thanks stack! +Or if this becomes an issue that is frequently asked for by the community we can revisit it at that time. + +bulk defer of unresolved bugs to +This patch fixes VALIDVALUES generation to include all identifiers and formats the output a little +Is it possible to fix the pomThanks +Logged In YES useridNever mind +It looks like the wiki page needs some love +Should be done as part of the same issue as +have always been special cases so if that is the case with those tables special case them +No new failures +Thats the same proposal here we error out with Sorry you cannot use the OEV on an edit log from this version of Hadoop +The problem here is that changing beta will change Y which will correspondingly change XY +Added the test contained in the patch +In the meantime I will move this one to resolved. +If one thread creates a new collection whilst another is iterating it then in the absence of synchronisation there is no guarantee what state the other thread will next see for the collection +because a string that has one or more trailing blank spaces is treated as identical to the same string that does not have trailing blank spaces +Clients do not have synchronizedclocks so they need to contact the to get a proper timestamp to stamp files being deletedSince it has to send a request to the any way why not just ask the to perform the operationThat kind of logicWhat is good in Wendys implementation +Here is the rebased patch +I have addressed this for Spring M as part of a general overhaul of s path +I committed the Spanish and Finnish updates +I think Arne resolved all of them finally txs +We just need to document it as well as how to handle scenarios that are described in this issue +I dont know if its still an issue but it still exists yes +Committed to and trunk +get seems to work +Please help me finished this issue +Derbys commitrollback behavior in procedures has not changed since +Without this namenode does not know the datanodes that could connect to it and hence cannot show dead node list if the datanode never registered with itSimilarly the nodes in decom list are in include also +libcppsrcwindowscompilercppsrcwindows compilercppsrcwindows to a template and automates uses of version rather than a random define version string +unless the property is a huge binary i would argue that reading the propertyInfo is not more expensive that having an SPI call to check the existance and the validity of the new ValueNo that was not the proposalI would propose that JCRSPI just calls SPI to set the value and the responsibility for checking lies inside the SPI implementationIn general JCRSPI should not do things that can be done more efficiently inside the SPI implementation +initial commit never were in branch so this is for I reopened the issue only to let you check it one more time and confirm that it is fixed even though one commit is reverted +Creating experimental branch for sqoop is fine with me +Let me know if youd like me to fix this or if you are anxious to get to it +This works for me in both Safari and Firefox on Mac +It sort of violates encapsulation +That could be a basis for an immutable Schema data structure +Go to Field Configurations select the configuration that youre using and select Hide on the Priority field +Thank you Shaz +It would make brms engine cleaner and simpler +Marked forFinnish localization is correct +contains both previous patches fixes fPIC issue with +I checked it manually +Closing all resolved tickets from or older +The problem was fixed finally with. +I thinkthat without SQL Auth in place the system is the old prepared statementwithout it but in this case there are no errors that resultThat is SQL Auth converts this scenario from harmless incorrect of an invalidprepared statement to NPE in permission descriptor handling due to incorrect of an invalid prepared statementI will work up a complete patch with regression tests comments etcThanks again for the feedback and extra testing +I integrated and tested the solution proposed by Ben +It looks like with this patch globStatus will not deref symlinks but globStatus will +Feel free to reopen if it is still not tall enough +Since the Affects version was set to an earlier version I just thought of calling this out so you are aware +Ive gone ahead and committed this change +Personally I think newer is a horribly bad default for overwriteMode +done + +Moved to fix in only +Ill leave this open for a little while +They are the patches we were discussing yesterday +Vladimir Im not against doing anything sooner +If this problem happens during hot deploy then its likely a general AS issue +Using the latest JPOX build test successfully passes for datastore identity and application identity +You may need also +Not that I have a problem with how the password is stored but if this can be fixed with a Railo update that would be good to knowThank you +Ok I think your point is more solid +Looks to be doing right thing +This was a trivial fix to make the mkdir url in sync with the others +Hm not everybody uses the and people do use the same package as for example their annotator on purpose +Please review +Presumably the single Place handles URL encoding and decoding so screens are still bookmarkable +region historian might get in the way since its bunch of data so skews things in that +Merged to doc branch at revision +Does anyone have any comments on this +Without batch commit log we cannot guarantee that an acknowledged write wont be lost by a nodeRight + +Legacy Embedded issues closed out as unsupported clearing the project JIRA. +I think it would be nice to enable this with an option strict +stub docbook guide exists moving this subtask to to track to completion +If it is a JDK problem that hive can not work around why not close the issue +However the plugablity of metadata repositories doesnt match the declaration of data roles which are vdb scoped +fixed. +The user has the choice of keeping track of the lease usage or not +Or the +It is a single write multiple read lockYou cant do that +Ive just added myself to the and files without any problems +Im not following +Im digging further +Did anyone look into how this affects jobs where a user modifies the counter limit to be higher than the cluster configured value and what happens in the case where the jobhistory server is configured with a limit less than the user supplied limit +bulk close of all resolved issues. +Perhaps a better solution is simply not removing the node until replication and done +I have heeded the note on Tuning Compaction and not run a major compactionAlso once you run a major compaction automatic minor compactions are no longer triggered frequently forcing you to manually run major compactions on a routine basis +Ill file a JIRA to consider your recommended API changes OK +I seems to have fixed that but not sure about summary coverage report is going to be generated at the end +The purpose here was to see how examples would get implemented with the new API version +A pass on the port +I only have access to commit changes to the code that is checked out by the build slave +All pass +Because of the rollback the out parameters are set to null +patchNow command default def testfailed on test execution instead of loading exclude file +Thanks +The only real difference I see is that you added xerces to the classpath and changed to quotes in the eval command +I think the complexity of the locality group model is not increased +Sorry have been away from the internet for the past week +Ouch +Please vote if you want to see progress on Ill have a look +Anyone else have an opinion +Here are the files from my project +RTC close of all resolved issues prior to release. +WDYT +The ability todo rolling restarts and have processes spill to disk when there is outages so that we can in withstand hours in stead of minutes in the case of a failure of the cluster or a strange network outage. +WPS request for gsNearest generated by WPS Builder in +No longer relevant in with NBST +The formula is correct I verified it again +Carl the comments have been addressed +Its failure of master joining running cluster +Over to you Justin all my code should be committed +I get test failures with this patch +It seems the xpath xsdelement does not work in Jaxb +Any update on thisfebbbcaaacaedcadecdfbcdfc +Right this is the place I changed and the specific problem is fixedMy concern has to do with the fact that I am now basically hiding a logical property in builder but at least I added a call so builder can handle this in some other way in futureBut as to your case problem should be fixed in trunk. +bq +Done in rev +You could use the minimized from the stub compilation to connect the classes or look at what is done there and do that for super classes only +Do you know why it happened? It doesnt sound right to add code to deal with out of memory exceptions because it is really an anomalous situation +The interface takes a as a parameter and returns one as the result +Will only need this by EWS release so still a bit of time yet +Is it possible to also reuse buffer across multiple file copy operation if you dont plan to copy file? It should reduce pressure on GC which may be the reason why its slower to copy a lot of files. +Jake has access to this node to fix +This issue should be reopen +Updated to as previous version but changed filename to for clarity +Reading and writing accross namenodes +Thus we should make the m build working prior to add jbossx support for RAR +I have reviewed Brians patch and it looks correct to me. +has some related discussion about how to do certain types of upgrades without requiring a the namedireditdir snapshot stuff that we currently do +Should we do message boundary events in this issue as well or create a seperate issue for that +Cant reproduce the bug in the latest build eacdeeaacfbfcbeSteps taken set Being a regular user created a vpc +We are restructuring the samplesIll do my best to commit the supplied in the scratchpad in the next days if nobody else steps with otherThank you +Rather than using dot files just append +To get the rdf despriction of an one might just add description to the URI +However we have been busy with other tasks such as AS integration and our way towards GA +srcorgjbosscacheaopcollection +Committed at subversion revision Touches the following fileM javaengineorgapachederbyiapireference +Everybody will be happy +Reopening to write release note text +can be solved by existing remoting plugins on both frameworks rest rmi hessian xmlrpc wsclient +Just ran it on Ubuntu +Not sure if I will fix this for final or not but if some kind soul creates a patch I will apply it posthaste +Without significant code refactoring redeployment and downtime its not possible to remove the overloaded methods from those services +I dont think I will be able to get to this for M but I can probably address this for RC +Hmm youre right +As I mentioned in and the security mailing list this feature and its equivalent for YARN do NOT really weakens security on a properly configured cluster +Closing due to release of. +In the meantime Hadoop has released version so it makes sense to upgrade to this version instead of +Hi RobbAre you running behind a proxyI think I already saw this exception when youre running in a controlled environment +Closing issues. +I am attaching a patch which adds the ability to specify name and scheme while creating the component using the archetype +ChunhuiPlease let us know testing result on your QA environment +dbafdcbcbeec SSL wildcard lookup doesnt find the longest match +Please reopen if weve missed something +Apparently the start Stop script was fixed for short cut may still be a problem. +I committed the patch to and trunk +This is obscure behavior even though we know at least one person who is taking advantage of it +Issue from my list above remains but this doesnt cause any problems and is a little trickier to solve +as T is not allowed in X generics so that settles that oneAfter discussion with Dave GThe original case is settled by the unconstrained T manifesting as an Any for the purposes of the body of f and Any is valid and if T is a struct will result in the c backend boxing the struct at the point of the +yes of cause for not for implementation for tobago +patch v fixed unit test failure updated error to trunk +patch to add version hd indicating row tombstones have been computed can reproduce upgrading from to and running upgradesstables afterward +Can you please report the steps in all details? I assign it back to Libor to get it back afterwards +Update to critical since it prevents usage of the deployment interface in domain mode +The concern that I have is that Thrift adds its own layer of RPC and may prevent us from doing things like transfer on the consumer side +Thanks Shai +For me this goal is supposed to be used in a release profile or something similar so that it is only run prior to a release being madeWhat is your use case +The original patch was contributed by Simon Wilkinson +One of my pages loaded seconds before index optimization and seconds after index optimization +This second rev of the patch adds some additional tests including an upgrade testAdds the following tests Verifies that you cant use views to subvert the authorization checks on SYSUSERS +With your decision to silently drop what you supported up to Sonar and to only provide an for the NPE you are simply complicating the of large scale projects +Committed and available in CP branch +large spaces between list items in IE done comments in the linked issue + +Dyre if you post your script Ill look at committing the changes +Thats crazy! I must be digging the hole in the wrong spot +Hmm Ive searched the Tomcat configuration directory for any and just cannot find anything in that respect and I havent really changed that much its really a plain vanilla install with Railo configuration on top +A few more fixes. +This could be done by moving the code that tests the document to a custom flow controller +and statement +We have docs too now +Ok Im migrating this to a new host +I do not believe that you cant even look on thatReason why Ive submitted the patch without granting Apache license is to put stress that is not ready to be committed +I thought our distributed tests are broken + +Hi MaximI have created this patch with changes in IDEA and code +Hi RobertIve committed a change can you give it a try? ll give Harmony a shot once a new version exists where is fixed. +merged Thanks +Sorry for the long delay still busy at the weekend but ill have a look next week and see what i can do +Many thanks Darran +But that is all very loose right now when the caching mechanism is fixed up I think the whole DEFAULTVALUESOURCE will be cleaned right up especially having to set it there on theI think we have to support remote for but I guess wouldnt have to for +Socket file names on OS X had the last character missing ggreercarbonusrlocalvartrafficserver lsTSversion cl eventapisocke mgmtapisocke processserve This patch fixes that +The test in triggerGeneral has been written for a specific purpose and that is why I didnt want the test to fail with select +jar file +Should also be decided on by +Yes it seems like eager delete may break some guarantee +There are still more updates I would like to do in particular storing all queue exchange binding data as serialized binary so that we dont need to change the DDL every time we add new properties to a queueexchangewhatever +Also we will need to migrate the to the universal +For review +Thanks for accepting the patch +Some task attempts are killed +If you think they are critical for CR please them +Trivial patch attached +Created to track the proposed ij improvement +Nice Thanks Ben +I shouldnt be writing comments at am +In JDT extends +I found it +When I open file in Linux there is no selection in table so that only Add button is enabled +sent feedback. +bq +seems to be a problem with inside the plugin +ThanksClosing issue. +Thanks Srimanth I have applied the patch but I noticed that +I guess that wasnt a very big change then +in trunk and +So other placement policy can override the default deletion policyDone +WFS allows for shorter locks WFS is just broken this way no sense trying to fix transition all resolved issue that did not see any further comment in the last month to closed status +The javadoc does cover the information that it is client side translator +Fix segfault in second vsnprintf is still needed in the case where the realloc happens since if there is not enough space in the buffer the vsnprintf needs to be replayed +In the meantime Ill have Bruno add it for me +This only applies to iOS as the same event on Android behaves as it was onThe plugin was installed when I tested these but I doubt it is related as its not mentioned in the documentation anyway +without scoring or ranking documentsWith scoring or ranking would be great too Im sure it can wait though +Maybe that is not so bad because it is only changes on itself we would have to watch out for +using the error does not as per reporter. +Somehow I was under the impression that this is specific +I created a branch out of the tag have updated both the trunk as well as this branch +was not reset when changing the archetype version +thus it is impossible with this rule to sort them across profiles +This patch already supports this +GPD is part of the distro +I took a quick look at the import code and cant see how the statement could return before it was finished +Because the WST adds the modules that will be removed to a deletedList so the error dialog pops up actually the modules will be removed successfully after the WST execute the next removed loop. + +Will take a look at this since Im heavily updating the +I will try to understand +Bulk closing stale resolved issues +It is against trunk however. +But anyway +Here is the new applicationscontentdata file that declares the pages as SUBCONTENT of CONTENTERROR that is itself a TREECHILD of TREEROOTLike this they apear in thehttpslocalhostcontentcontrolnavigateContent you Patrickyour patch is in trunk rev +You can apply use the patch utility to apply the changes in the +Would you mind opening a jira for the tests that are failing anyways? We should track them and disable if needed for a clean build +If you want this patch you can build from branch or I can send to you the fix. +Please make it possible to disable this filter if you plan to add it to the Seam filter chain. +Thanks for the review Hitesh +It was just used in the original patch +Hi WillemI think we can close thisGreetingsChristian +Here is patch after removing the changes in and +The fix makes Derby balk if someone tries to do something thats not in accordance with the JDBC specification +But IMHO a CP release is ideal time to fix this inconsistency +Comments were added to the code and documentation that version of the API should be phased out by VCL +trace log can this be deferred as far as is concerned +Changes look fine to me +My project looks like thisapp root pomappjava also a pomappjavaproject a jar projectappjavaproject also jarIf I build from app the properties are defined +It was a weird one for me to figure out on RHEL myself which is why it sticks in my mind and I chimed in when I saw your difficulties +thanks for the it works with the version from +Committed to trunk. +Wrong product +No changes since the commit that removed version from +The former talks about allocated local storage and that shows correct values when I checked on my setup here +Hi I wish we could do this +Is there any hope of getting a small example that demonstrates this? It seems like we should be able to fix it +Thanks Ramya! +s with the reported error s working with a clean +Since this is a new feature we dont have to get it done in the very short timeframe so I postponed it to +Resolving since all platforms have signed of on testing and adding the sample app +The Mapbuilder toolbar is now mapped to and all Mapbuilder buttons as well as all map and vector rendering are completely based on MSN and Yahoo layers still need to be mapped in +Whats the status on this development? Is work still happening in the branch mentioned or has it moved back to trunkIve seen similar memory issues on Windows platforms and Id like to offer some assistance in this realm if possible. +Need to verify unit tests to see what is the problem +If you cant get it to produce validation errors then please follow up to the list +Camel netty has pooled producer since or thereabouts +go to +Instead of a hang it should throw an exception +I dont understand why it says both core tests and contrib tests failed the results dont show that +Stand corrected revoking spec violation linked Forum topic +Doable but very low priority atm +patch applied thanks! +retrospectively setting fix version in preparation for SDO issues fixed in SDO Java M in preparation for M release +I do not think that this is a seam issue +Moving out to due to note of additional work required +Thanks Ollie +The currently has those to specify the JDK versionsMy project is currently using JDK so I have tweaked it to use and +Are you really using such an old JDK? Actually Sun JDK has already reached its and is not even covered by Sun support ve just verified that this works on JDK for me +Jives with the one posted to +Closing all tickets +We now keep track of the size of the buffer after the most recent successfully serialized record and use this as the block size when we write a block to disk +me checks the WAGON jira +Applied as rdfc + +bulk close of resolved issues as part of JIRA housekeeping for release. +Attaching an updated diff file +are now green as well +Works fine. +Theres already a here +reopen to record that Im backporting this to +Testing things out after I wrote this I noticed it is bound to the debug level so ignore my question +Rather than catching Exception surely the code should catch +Please try this one not sure what happened with the last file +To be franck I only reviewed it but had not enough time to test +I have had a brief look and it looks very promising +So for the plan is drop patch make the other tweaks I mentioned and then make the CLH changes as part of +So the producer doesnt need to know ZK +This is fixed by separate portlet bridge project and code in the jira has been closed as part of a review and archiving of the aj project +Well never support this +The top button worked fine just the bottom one was missing the orderId parameter +In addition the web console is accessible using telnet +I notice Harmony has different behavior with RI when path name contains Harmony convert it to and treat it as path seperator while RI doesnt do that +I agree it would be nice to have a package for localized messages in generalIn the future this leaves open the door for localization of other parts of lucenesince it uses etc I think it would be very reusable +Fixed under revision was also fixed in trunk under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This will be looked at after the branch since this is a regression and we dont have time to do it now +Per Marcuss note following the documentation appears to have resolved this issue +One might be able to abort a deletion in flight but I dont believe that a user could react quickly enough to a mistaken deletion +this test has never failed in CI +Increase thread sleep time by ms +Ive fixed that and excluded as well because it causes regressions. +Flavio that is not scheduled no. +mrepositorycomgooglegwtCProgram FilesJavajrebinjava Xmxm cp CLASSPATH gen Djavatargetittarget +wontfix for issue apart from that the problems indicated in this issue are fixed. +Our facilities could make good use of this +We can look at this reorganization laterThoughts +I havent done much testing of external ZK ensembles +Jake and I discussed the reasons for removing the generic from the its not actually exposed to clients so shouldnt be a breaking change +It actually bypasses the properties and looks up javaglobal on the local serverAs I mentioned last night I see a fix for it one that wasnt possible before due to JNDI issues in the VM +Thanks for the update Michael +Got interesting things that we shuold look into though +ok remove loadSchemacheck for keyspace on client move the check to make it throw IRE handle the IRE in the just like change all the name to describeRing to be consistentstill have problem if running simply via looks much better +Closing this ticket. +one major functionality is add few selenium case for your app as well +Resolutions other than done should not have a fix version set. +Thanks for checking it out StefanIt is a bit confusing as it is +Now we need to make sure that all hadoop MR benchmarks on big clusters run well before we commit this patch +no worries Andrew +there is a series of things to be done +Please move issue there +Yes this is an issue in and +On the plus side I believe it handles the socket types better +Im not sure if thats a bug or by designThis patch now includes a regression test that simulates closely +I think having cacheable views is pretty important +Just create byte files in the conf directory with these namesThe unix touch command works great for this +Sounds good +I checked out a clean copy to verify +In the referenced thread the user is specifying an alias rather than an ID so that doesnt create a separate instance +Minor changes to a single class +Its just the opposite +closed +what about spilling to disk only when a worker is low on memory during INPUTSUPERSTEP and at the end of it reloading them when the pressure of processing splits is done and the pressure of mutating the graph and message passing begins +I think this is the same this is a to clarify when using the distro then creating a child container say works fine however creating a child container with profile seems to cause the issue +Just the info in the comment applies +is it able to reproduce this now +Could it be tha the Web Console bundle is not bound to the Configuration Admin Service APIWhat happens when you visit Configuration page in the Web ConsoleThe web console has an optional import on the CM service API package and happily starts without it but without registering a +Committed to trunk +Starting you please add a testcase? Thanks +The change looks goodThis sure had me puzzled for quite a while +I particularly dont think this belongs in Utils because I think the basicconcept should be integral to usage of Chain +Sigh +Dave I am trying to where bufferSize could be applied +Are you sure it is memory leak instead of just GC not kick in yet? If you are certain can you attach a junit test case to illustrate this +Added a comment explaining why an interrupt is used instead of a shutdown flag and removed a spurious sync patch +Hi Michelle please consider adding the screenshot as Burr suggested +Tisham I believe this one can be closed as you have your community module doing this work rightStill hoping to see that module become and extension btw +pending on followup item per review board discussion + +Removed the tabs +This should fix the problem +Im looking into this and I dont think the issue is to do with static or singleton properties +Unfortunately it was not so easy +This project demonstrates the works for me +The JON team think this will be fixed much later. +I just committed this +I admit to not doing a whole lot of testing but the patch seems to be fairly trivial +if this flag is setted to true this feature is there or there is not this feature +So when did it start? Is it the naming reloaded merge +assert o! null o! null null pair assert! null ! null null row +the idea makes sense +There is that danger +hmm dont know how to apply this patch syntax +Thanks Todd for fixing this! To make it simple you may do it without deprecation +Your patch looks good +to master as beec +Thanks! That stack trace shows the problem +If you dont plan to add the necessary documentation unassign or assign to somebody else +I fixed it under a new issue. +If none exists it creates one +Thereis however a problem with the recording of the test results which is preventing the build from stabilising at the moment +Shreyas can you resolve and close this bug +This stage detects when a tokenizer has skipped over punctuation chars and inserts a deleted token representing the punctuation +Grepping through is really only for developers and that should be possible on the documentation as well +Minor tweaks for style +With no comments from anyone I think this issue should be closed until someone has specific UI suggestions or defects +Affects as well +Noticed a minor regression from previous skinwhen a TI is being edited and you click on another one the soft keyboard will slightly lower then raise againfixing it now +I think your patch looks good +Fixed with Revision backport maybe +I note that the reason we need to support and replace is purely counters +This is the same bug as +Any thoughts +Fixed with patch version +Levels are in a priority list +Refer to and its related issues +Maven feature contains the plugin that isnt compatible with Eclipse +work completed +processes passed IO at initialize +jars that are built separately from the Lucene release +Please note that in this case default would also process that bridge and creates documents in the ontology in current implementation +the transformer code definitely can be invoked by more than one thread but based on the log output excerpt above I dont know if it ever was during this execution +Thanks for the clarification +Ive confirmed this issue exists in the branch and I bet it exists in as well but havent checked yet +This is the folder flume looks for and also it is the default location for the flume file +Ok should be all set now +Code has been deleted so no longer an issue. +Wed like to keep high test coverage for this project +were backported to as well adding the latter fix version +There are concret recomendations in SWT documentation about it +Ok closing as duplicate of. +I dont remember but I think was used in signatures because of methods unique to that class +The same could be said for the data source initializer maybe we could put them into the infrastructure of now within a test package and potentially more them out later + +So restoring the PATH setting +Fixed cleaned up the client REST API a bit +Maybe simply just testing the performance for longer time before making conclusion? Will that work? I am trying to see if there is any other better ways to identify AB problem and use gracefulstop to help improve hbase cluster performance +Integrated to trunkThanks for the review Jeff +Flipped the build to SP leave this open until we verify its ok +I have to discuss this with David today +There are plenty of competing dependency management tools available that Im sure wont try to impose such authoritarianben +Hm that was actually by design +Ive cleaned up some of these plugins got rid of some unused classes externalized some strings and basically added a few methods that will allow for either workspace settings or settings on what to commit during a publish +To fully use the advantages of BITMIME we need to use the EHLO protocol when sending the messages +I understand the desire to avoid wire incompat and I would agree if this was or +Jason most of the things on your list specifically are not You are simply lumping requirements together +verified +Its entirely possible to run a perfectly fine Linux system with close to MB of RAM free +Will this add another bit of complexity? The single node accumulo instanceis intended to what amount of entries? At what size does the having thenative map truly affect performance? And to what degree? One billionentries +Adding self to cc +or where you talking about actually editing ruleflows in the BRMS? Not sure how we would do that dhtml doesnt handle graphs well +I guess thats why we ask all folks to provide all relevant files +Im not sure if you need that the patch of is made on top of this patch +It works well with HA mode. +Although the branch does not have all the files that the head branch does for those that are there the changes were also committed to that branch +So the problem here is something data on the wire is already processed by the crypto subsystem and in s buffers? Not sure how to make select see that data +a client doing get on usermahadevtest and a get on usermahadevtestusually any filesystem like namespace handles this and would return the same thingan error is fine +This bug makes it difficult to debug jobs after the JT has been taken down +Why is name and redirect commented out +Sure +patch for this patch has been applied thanksD +srcjavaorgapachecommonslangtext +There are a client and server bundleThe other bundles in use when this example was created are +Could you extend the patch in the way that the polling can be disabled too? As example when the content of are no polling will be used +I understand your concern regarding lack of reproduceable test case but IMHO this should be adequate due diligence before accepting the patchWont you please consider making an exception given these circumstances +That should be scheduled for Camel +Ive got what I need so Im not in a rush for anything else +Last i heard from him was when he was upgrading the Jira instance before doing another test +i have to correct my commentwe get this exception on executing a not on update a entity +rebased on current does this patch preserve the old cachedirectory format when the feature is not enabled? I agree with John that doing these invasive changes particularly changing the directory size and the coupling between disks is potentially bad +However my intuition tells me it is unrelated So far I tried running without cql and then dropping from cqlsh cqlsh running with and then dropping from cqlsh cqlsh +Can this one be filed forFor my purposes it is neither an enhancement nor a minor issue +So to fully understand the problem it would be nice to give a sample +Thanks Jie and Elliott Ill commit patch v +The other is a DLL project to generate the library of native I invoke maven on the parent POM project the build works and it runs javah just fine +Derby provides an embedded JDBC driver that lets you embed Derby in any solution +I have a fix to and a corresponding unit test for it +Sorry I messed up the commit message for r which is also relevant to this issue +Actually I have worked a little on the which is a port from Java concurrent API +Is there a +Thanks +Paul The first set is correct +I can also say that there was a lot of cussing and gnashing of the teeth when I did not get a get type method +Committed to trunk with revision and to with revision JIRA issue to rewrite this test as a junit test has been filed as. +Since you already started working here I will just close the other one +Okay +Clearing proposed Target Release value from to bug was filed against WFKs GWT component although it affects Errai its not directly an Errai bug and its not something we are likely to fix ourselves. +Actually there is same situation at + +Thanks for the patch and tests +Ok +I forgot about your patch you must have tested it on so it must be on the mvn classpath +Moving to +Even a year later your advice is still useful +will be reevaluated would actually be a new feature +Create new file asset guvnor type png and upload image +The master is getting stopped in + +Fixed +Going to close this JIRA sorry +Thanks Stack! I will take a look at this +this is a minor the next that work is well underway with AS all previous community releases are +Manjula has fixed this issue in AxisC. +I dont quite understand this yet + to trunk. +Committed +Patched the code in SVN with a change in the code to compile using call of theexecutable is now guessed using on windows and m on other systems +I have now and it seems the new feature works guys changed things a lot for a +Shouldnt this be a part of the default chainI think soRegarding separate components I think its enough to be separate classes +I also tried to classify various queue configurations a little more clearly like single queue setup multiple single level queue setup and hierarchical queue setup giving descriptions of each +Submitting patch so its picked up in the queue +the hibernate core dialect doesnt control the reverse engineering +does that work for you +I had left this open because some jenkins fails started when it was committed +I attach my test output +Hi BrianThe cli looks greatJust spotted a minor error in the cli menu for filemgr and I checked the for filemgrYou have Workflow Manager URL as description for bean url and similarly propertyNamesThe same goes for +Tested the upgrade of nexus indexer against IAM ver +Logged In YES useridIf i need to backport this fix to how do i do it +Code in patch just catches Exception which is bad practice but makes living with UDIG trunk a bit easier for intend to work on image mosaic pyramid ecw and gdal next +What do you think +Looks like a bug +This patch adds in deletion of logs that are too old +Hmmm +Im ok with different framework +Closing. +Please advise what information I should provide +Why has this issue been rejected? It is clearly a there an existing issue for this problem in Hibernate Tools +Just to clarify +The larger issue here is that we do dispatch from a type that doesnt descend from the abstract type so protected methods cant be called +Do you think that this is acceptable? I think says that if a datanode dies the client should detect it and possibly take appropriate action +I am assuming I can safely ignore all the attachments in for the merge? Just want to confirm things like that Brian attached dont need to make their way to the CP release +Thank you +I am totally fine by this Bilgin +I think there is just not quite a setting for what we want to do in this ticket +I wish I had noticed this gap four weeks ago +Ill try to deliver it for weeks +but the append semantic is not guaranteed rightTheres nothing there that violates the append semantic +This issue seems to be resolved by Dons patch in rev +I just check cxf it is a same issue +bq +This is a duplicate of +It may cause problem to us also so better not well need raster support I imagine we could just use Mapserver for a start but well see +Run HDK +The portion of this will never function correctly in Hadoop +No need to abbreviate so obtusely! newSearcher and firstSearcher pretty pleaseDang! Thats fine by me but there seems to be two camps in those who abbreviate and those who dont +Why do you think they are not consistent? The alternative would be that they contained in one buttonIf we do what you suggest we will have to remove the Add and Edit button and add the AddEdit buttonEven if we do that you still wont be able to test all native dialogs using +I starred them +Create your own service that implements Runnable and takes an of Runnable +Thanks Daryn for the patch +I added a new chek method to the which wraps the exception in an and updated all model classed to call it +Reopened to track the gradual refactoring of the Spring Python core container towards a stable release +Then the name of the Reminders will still be the same so they will all end when completing one of the task instances +It would be nice to get stderr and stdout itnerleaved in one think we could just call setOut and setErr on the same actual object however I dont know if that would literally interleave them down to the level of individual characters +documentation patch and Ulrich +Otherwise it fallsback to the service url property. +Each one is hoping for convergence after only iterations currently but were still talking about jobs launched from junit for that one test +I checked the web page manually +Patch that removes unnecessary reinits of the various TIP holding data structures +Would someone pls try this +Did the patch apply to Lucene correctlyYes that one applies fine +to Knuts suggestion +Rob care to document what the new behavior is +This would not require changes to the system +Simple project which demonstrates catch! last being before last might well be the problem here and certainly explains the observations above +In the past we have run into cases where the default concurrency level led to high memory usage +Its a workaround if you use the id property in the tables you want to display in the same page displaytag gets confused and uses the same name for both +I can remove the rhetorical question from the patch but do you have Andrews suggested CLI example +As for my last comment I cannot reproduce this problem today +It failed five times in a row just before the commit +Thanks Amareshwari! +renamed to add an extension so it doesnt give windows users issues +If you have to pass additional C compiler flags you should use CXXFLAGS. +Most likely problem is in our code +What in the Home page is missing and would require an FAQ +Committed mine and Ryans patches above. +I will post some update soon +created a test case for State Manager by +Committed revision +We are tracking this on for a future Felix release. +Patch attached. +I did a few test and it seems that the bug appears at MB of memory ask precisely +Hi LauraI think in the numeric example of the ORDER BY clause ref page the following statementIn this example the salary and bonus columns must contain DECIMAL data types +Please close this issue if you agree +Sorry for my type mistake in the JIRA subject +Resolving this ticket for issue of live migration not getting triggered due to miscaculation of available capacity which is being addressed inFor other issue of migration with operation timedout error +Now that said none of the implementations actually use it +This is a tricky requirement that I dont think can make it into GA +Waiting for the fix then. +Ok +In the test case all destinations eventually reach their memory limits and stop caching messages so a dispatch needs to read from the store which limits the rate of of note with the test case is the use of strict which means that only one of the consumers gets used +So all issues not directly assigned to an AS release are being closed +This is provided by the Persistence Extension now right +I verified all the documentation links and very thing looks good +already has a bit of code like this to enable servlet injection but theres still a lot of stuff that needs to be done by the users to make a war portable. +Yeah this is going to take some more thought +buckets are removed only if there are more in the current than are required +Ive noticed a significant slowdown in performance in the current CVS over the previous release particularly with pages containing a lot of separate elements this might be the same issue because it sometimes manifests in images taking ages to appear + is created to add PMI link and brief info to the classlib introductory page add RMI link to the sitemap remove disclaimer and legal info from the docThere was a misprint in the xml file name it was instead of +It is also inefficient to insert large data using character constants +Pushing to +There is a compilation failure but the reason for that failure is not being described +After debugging it I found the issue to be more involvedThe issues seems to be more related to then +change fix version +thanks at revision Please verify that the resulting doc is what weve been negotiating over +Resolving issue factored out the due date calculation added test for task due date and description +thanks for your kindly words and for your time regarding thisbest regardsEdi Mitrea +Created a plugin that contributes an xml catalog for resolving switchyard schema references to the Eclipse for best results disable Hounour all XML schema locations in preferences XMLXML FilesValidation +on JBT +Removed the syntax highlighting in the note and added it to the Java code in section in both versions of the user guide +It seems to me that the idea of merging is to support running multiple independently developed things together then perhaps the right answer for abc def is abcdefWhat do others think +Created reviewboard +The code exists in however it hasnt been tested at scale +These would be the classes for all the magic values in +I took a quick look at the patch in and from what I can see it replaces with a new implementation that doesnt use at all so this bug should be fixed ifwhen the patch is accepted and applied +Ive added a few small changes to make the exception googleable +Any chance you can rebase it off trunk +This was fixed in revision +xerces doesnt support schema support yet +Thanks +The BIRT team has changed the BIRT Web project structure and the way it launches +Thank You for the comments david +Its your call I ran into an issue fairly specific to working with the client and since the client symlinks are generated in a unique place this fix applies only to them +If the purpose of is not to prevent verification of certs then this ticket can be closed +Merged to the branch in revision . +Marking as a Wish and leaving it open with no fix versionAlternate implementations are really hard to put into the code base right now since so much calls directly into couchbtree and couchbtree calls directly into couchfile +Trivial patch attached +The will need to use groupIds of the form +Closing JIRA. +Tx for the reviews Zhijie +Its doable but would need to refactor a bit the base class so that the hibernate subclass can do the session setup +This puts us in difficult spot +this looks good +More diffs between revision of the trunk and the jaxws branch +This portion of code has gone away since release +Joerg what is the status of this issue +Sorry +This issues needs an attention +Or you could verify that it does not and attach a project showing that it in fact does not if that is the case +It goes away every time +Marking as m thinking we should just check in the jruby jar and make the shell jirb now in advance of any DSL design +Though this is trivial maybe we can test with this patch? Have a look at patch the code looks good +Happens on my Samsung Galaxy S Tab and Note +Simple implementation provided +Best of both worlds +Great no problem +ok did not notice that it was not in your patch will follow up with rob and revert the change +Many thanks to Carl for submitting the originalpatch and for the thorough testingMike +Sorry it is not quite correct +Gavin King I have to say it is very surprise and regretable to see your comment +Werneryour correct we have that property overwritten custom but iam not sure why we cannot overwriteany ways for i just commented and it works for me knowThanks +Marking Patch Available +committed +data works for both of +you have sample code on how to iterate through JIRA comments? above scenario actually may happen to HivePatch A was tested on day +This is a good pointer +Added in rev Thanks Todd +As of packages detects classes as well and passes them to Hibernates addPackage method +obviously cannot remain the same since it is a static +Slipping due to build changes code changes and the lack of time to keep up +So even using the idea of a depends in the map that is the first parameter to the target closure this would have to be mapped to the depends function +You should specify real file encoding +The popup can be cancelled by clicking the icon again +Willem here is some code I did for experiment +Hope the original problem will be fixed as well +Hence the DN was considered as a binary attribute compared accordingly +HiI only included the Microcontainer text to show what the logo looked like in ve got no opinions on typeface other than maybe trying to find some consistency with other named projects +Could you please attach a repository dump so that I can use that to reproduce this problem? Thanks +This adds a line to the web interface section of the HDFS user guide noting the syntax for the but ignoring any permissions issues +Patch looks good committed to the trunk. +Max could you please review the patch +Management server should print more error message +adds a cyclic dependency when added to +This is a problem with Grails rather than the plugin which is why I have moved it +Adding patch containing solution where we just throw the if we get an from the entrylogger trying to read an offset that doesnt exist +Will make sure it works on Linux and Windowssuggested patch +I dont understand the part much +Closed upon release of Hadoop. +Thanks Simon +If so please close +Looks reasonable to me +Having thought about it some more its probably not worth the effort +This includes Daves deprecation of the retry factory bean norollbackfor and adding coverage of the fatal list +Different feed but probably a related issue +I will provide a that we can use to prioritize work +Just skimmed the patch wo reviewing it +Now that we are caching a single opcode inside the we definitely dont want these instances shared between threads +Ive applied a slightly modified version of this patch +I dont see in the Core API or Core Data specs on how to get the format of the response for REST services if format parameter is not specifiedI believe the core data spec only specify that the default is jsonI propose to add get converter for response and use these steps to determine the right converter +If you can post the code that does the performance checking I can take a look with yourkit +Those three failing tests are known to be broken or flaky +We just have to print aBIIIIIIIIG FAT waring message in case transfer encoding is anything but chunkedand return the input stream as isAm I missing something? OlegI just looked at the patch and was wondering if you should only warn if youre in Strict mode? think this patch should take care of composite Transfer Encoding headers +Using JSF tags like form or all should work at anytime as the documentation does not tell any limitation or constraint in using them +Closed. +The problem here is there is one and only one release specific doc +Im not sure what is kicking off this background thread but whatever it is needs to wait until the service transport is initialised +push to +Thanks Chris. +will generate the default sql statements for jdbc persistence in spring bean format so people can tune the sql so I put in in resolve this for now +Added two tests regarding this issue and corrected the bogus part checking if the path was containing two dots with a segment check +Hi RayHave you a patch? forgot the attachment shame you cant do it on the first jira creation page +At first glance it doesnt look like to be a problem in CamelHowever the interceptor documentation could be wrong how its meant to be used +The first was testing it within your PRs branch many times during weekend and it failed only two times +I will be the first to admit the current code drop needs some further polishing and now I have some ideas on what to do based on your input +Also the dialog itself does not appear avoiding other errors due to null ssh session +This supports fonts correctly positions the text correctly and breaks up long lines as well as correcting the hardcoded Tj Td which would be too small for large fontsThe one problem is that the positioning algorithm is still black magic as I just copied the normal text box algorithm and tweaked it until it looked okay +bulk move of all unresolved issues from to +Simply labeling those as is a dodge +There will be a cost to processing stuffIn any case they will always need to be different in some way +for example one use case is that one project wants to get some data for a reasonably large subset of the users +Thanks +Thanks SumitYour patch is in trunk rev +am attaching a preliminary patch +Were now binding to the mcastaddr in the constructor on Solaris too +I have an application that does container allocations using pure memory or pure CPU +Also could make things even faster +bulk defer to no longer have a platform conf +stack +Ive added an autodetectInterfaces property to for interface detection +Created and an example demonstrating the tool +Same area affected in SY Model Utils I believe +Unsetting Fix Version on unassigned issues. +I may as well fix that as part of this bug rather than hang yet another issue off this oneAnother patch will follow +Thanks +This behavior resulted after +changed component jobName between Q and E so it is treated as a literal value +Ok +Additionally we spoke about this in the IRC channel +bulk close of all resolved issues. +Im not sure +I making this patch available and you can change it to open after the patch is submitted + +Committing +Fix is now committed with regression tests. +Thats no problem at allThanks +This would bring the write size up to by default which performs better in than in the Java implementation +No tests because its a Web UI change +Ive also tried it with and all of the basics there work though again a few quirks due to mapping from the broker management model and QMFI cant speak for as Im not that familiar with that but Ive just tried with a few options and that seems to work tooThis was primarily done to prove that it could be possible to align management across the brokers +The python tests need to use PLAIN before they will run against a broker that requires an implemented mechanism to be specified +Marshal has this been completed? And testedverfiied that it works standalone +Does me working on picking up the logs and you working on the message work for youSure! Ill file a separate task JIRA for it and please feel free to continue working on this one +You actually specify the dependencies inside the plugin element +I was stilllogged in as juha not as admin +bq +Maybe we should make this dependent on the DB version in use +Therefore I used the configuration from functor +Thank you for the commentI updated the patch +i didnt spot an attribute in the m http related there some reason this same pattern wasnt applied to the http inbound gateway +Can you automatically subscribe everyone that is on dev to commits? This would reflect the previous setup where the commit messages came along with + +Ann if you need more information you should contact the list as Jean recommends +I think I should apply this first to see what happensIn my tests I saw two different failures Some parts in the test just took a bit too long +That should be fairly easy to do +I didnt deprecate the elementstringvalue call as per Elliottes request. + that adds more of the dependencies adds some updates to the docs about requiring mvn package to fully compile +Saw you also got rid of some casts +Sort of at a dead end next step is reach out the maven community i guess perhaps with a post on stack exchange +I think I may close this one in favour of finding a solution to reliably detect edits +Jonathan I applied to the branch +the wrap is only used with the tests so it shouldnt be a problem in general +Clicking the reload button doesnt help +For httpd theres a configure command to set where you want your installation since in the case of the zip that is an unknown at build time hence we need to specify the proper config file and log file +So Ashwin whats the status of this ticket +is now out so closing all resolved issues +committed revision close for Solr +ah point taken on that last example +Thanks Keith +instead we could look at how both are used and if theres adjustment to be made +your tool sounds pretty sexy especially if it can perform validwhen validations on the client side + +No problem now +Hoping Hudson QA has no probs with it either +closing +There is no savings of doing this in the same fieldSo to me its more clear to query on fooexactwhatever if you want an exact match versus doing it in a roundabout way with a sloppy phrase query +Android Kitkat is expanding its to Nexus tablets in the next couple of days so the impact of this issue is only going to widen +I fixed the path errors caused by moving from the directory structure in the contrib tree to the trunk tree +Issue may already be resolved may be invalid or may never be fixed +Hi AhsanI think the Eclipse patch tool was confused by the presence of the fail directory in your workspaceCan you please try removing the fail directory and regenerating your patchI think that your patch should be quite small when you attach it just K or K in size rightthanksbryan +Reopening to update developers guide +sorry priority isnt major +It is fixed +Path applied please close this bug if all is ok +I think usually committer resolves the issueThanks for taking the patch! Im going to try and break Giraph in a few more ways this weekend +patch looks good. +Jarek I can not reproduce this problem +And set up could be made easier too +Thank you PaulexBest regards just wanted to say I am looking at this but Im not really sure why its failing at the moment so Im not sure how long its going to take to fix +There is a testcase for this +I think your pom is actually broken at the moment +What is this about? Is it a way to adresses from spammersThis appears to be an Axis issue rather than a WSSJ issue so closing this. +Let me get into the code and see how hard that would beAgreed on the info metadata approach +if you drop it into clientjavasolrjtestorgapachesolrclientsolrjembedded it compilesruns with eclipse +Lets say when we add the entry we have a ratio of and when the entry is removed we will have the ratio of then we will leak some space for cache utilization +Not reproducible on a fresh bit install +The patch has been appliedPlease verify +I reopened this issue because it doesnt work as opposed before +Dhruba mentioned On lease expiry the namenode should check each block for the file to determine if any of them are below their intended replication +Patch has been applied at revision +So we might just always use String in maps especially as keys to avoid comparison problems +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +This is the right one +for the second optionIt would be great if we can complete remove UTF in the near future +I wasnt able to get the pax runner framework to execute using a fresh build of the framework trunk but when I retested this in the original failing environment this same problem still occurs +Problem is that the exception occurs before faces messages are dequeued into the faces context and so that gets skipped +for the patch +Fixed in +When the ofbiz trying to connect to the database the logs says No Managed Db ConnectionsI wonder why not connection are handling by the pool to regain the control on the connections openedAt particular instance it is throwing lot of error stack like Connection reset by peer socket write error problem because of the socker errors Unable to acquire a new connection from the pool the heap memory is growing and geting the system CPU to hangAt present the hang over has happened to occur more frequently like time server downWe are running the application with nysql Also we are using the visit and serverhitbin for statistics point the problem probability is growingI request to give more information on any updates on any of the above sectionsAlso to control the heap memory to be assigned fora production outlineThank youSrinivas +If the URL were stomptcpactivemqhost then the assembly would be loaded and used +I just committed this +I like photography and actively use such products as Flickr and Picasa +Allow to load hidden impl classes using its own classloader +Closing because the sample is not shipping at all in v. +Tx +Adding a comment to a closed bug is not how you report a problem +JVM just exits +My original environment was but this error occurred after I upgraded to the latest dev release of meclipse +Committed as ree +I my with the Hibernate +Note that in Eclipse it doesnt really matter where resources are located as long as they are in the Eclipse buildpath +Worthy of a new bug report? I believe this is fixed now but if not yes. +Seems reasonable to me TedAttached patch doing as you suggested +The should not ignore the I dont think there is any need for a new factory right now +Test case attached +fold include files into +We should take a nd look maybe they are only used for testing +TRUNKCompleted At revision BranchCompleted At revision +I committed patch with revision . +There is the same version of netty in HQ tag ASFinalbuild and EAPClosing jira +Underlying cause is something in what version of Commons CLI did you use? I tried your test case on the trunk and besides a missing space in your test it worked fine +When the components were changed from unwrap to factory components the component name of the classes changed which messed things think we need to revisit these components at some time as coupling the phase listener to them is a bad design +Can this one be closedset to resolved +Thx. +looks like this is committed and thus donePlease reopen if that assumption is incorrect or else close this. +I think this is duplicate of is the split implmentation we need and just needs a commit +I dont have access to Java methods through JVMTI and plus Im already showing the references throught the memory you tried it +xxx +was trying this yesterday +Many thanks. +Jervis is on vacationI guess Joao means that the operation to changeedit repository connection equals Drop the connection and create a new connection with the new changed value so nothing need to be done for the already checked out resources +Is this fix allready in the Jboss NightlyIf it is what correct way of proceeding with the solution +By refusing to give eclipse the resolved artifacts it will be unable to create the references thus creating the reference using the installed packages from the local repository +The big problem that comes into mind in the repository locking +New Patch Fixes test failure by restoring original queries and instead adding new onesThere should now be lines longer than characters +You definitely have green light from me for the turbine tree +How in general do we want to carry the new Admin UI forward? It needs some love if its ever going to replace the old UI meanwhile were stuck with having to either maintain both or have features in one but not the other +Ive incorporated the getInstance change and added Javadoc +But the test case still produce an error because it seems HSQL doesnt support use of tuple in a IN subquery +Or even go straight to +Fixed as suggested in revision +Try it and see if it does actually make peoples lives easier +Could you place this jar under the maven repository also +No but it would be very nice to get it implemented +Found A Way. +I committed the patch and Nigel updated the Hudson jobI filed so this could be done in a single commit wo breaking Hudson +Please do not apply it +FWIW this dependency was added to today and will be fixed in +Currently the behaviour of a Expression is that a warning is logged and the FIRST statement is taken as the valueThe patch changes the behaviour so that there is no warning and the LAST statement is taken as the valueI dont know which is more sensible going forward nor do I mind which approach is taken but it seems to me that the change needs to be agreed +After the of both embedded and client check type compatibility. +Can you see a reason why +Question removed from docbooc files in. +is always in seconds not minutes +added a servlet filter that will send a redirect to the appropriate wsdl ifwsdl is specified +This will also fix current problem with current +Added check at startup to verify if the cleaners are presentIf cleaners are not present everything is fine unless therere snapshot in the +What about simply making the same changes in master and and just make the default false in and true in master +Pass false to or set Java property false +As long as the ifdef is isolated to a single place in qpidsys I think its OKWhy did you call it QPIDTSS? Might QPIDTHREADLOCAL be clearerTed +Damn I am using you please provide some more details on this? the raw stack trace Im afraid doesnt point to any obvious bugs +Should not cater for the lowest common denominator? Entirely a matter of opinion I would +I think you want touse viewBox instead of objectIt is important to realize that objectmaps coorindate in the child to xy of the objectSo in your case the patternUnits already mapped xy +Now we are using from +Brian could you please list all updates in documentation +I did have to change some files besides the POM in order to have the examples run correctlyPlease let me know if there are any tweaks you would like to see +I try to fix the issue contains faild tests list +So it seems the problem is fixed alreadyI am closing the ticket but if you still find any problem with trunk as of revision or later plz the ticket and let me knowThanks for providing a good test case +Screenshot showing how the ide even supports adding href for extensioncomponent jnlps +Fixed by +Awesome thanks Otis! Have a great vacation +So are you suggesting that by doing the sampling correctly we could distribute the data in a such a way that we could throw away all but a few maps output? That seems doable but also like a big change +Fix is in CVS +Added as well as integration tests for Standard Oracle Postgres and Derby ddls +Rebase try +Just an update I have still not been able to recreate this situation again +Bulk move issues to and +The violation errors exist in the controller but are mysteriously lost when rendering the view +Committed +odd? That request should in fact return no results as the way its specified is latlon yet the request returns results as if it was lon maybe the bindings are not passing down the srs to the polygon elements +This patch seems to fix the problem and passes the testsId like a second look however before committing the patch +Ill check to see if its the same verify error or something else +I think that the full packages contain a mixture of good things and bad things in this regard +Should be fixed now +The issue can be closed +modalPanel name is checked +Nice catch Jim +Ignore this trunk and branch +As the new test framework internaly uses with almost all its methods this test isnt required any more +Unfortunately we cannot add a meaningful test for this because the test must fail compilation and it will do that even if the error comes from javac +Waiting for the endpoints to not have pending exchanges will only work if this is done globally for the else some endpoints needed to process some pending exchanges may be unregistered too soonAlso we need a timeout on such a wait +Just great! +Hi OliCould you update the description with a bit more information on this taskColm +Patch for ARQ +Let me know if I can assist in testing +Adrians test case had lots of unecessary stuff like apache commons dependencies validators etc +Then yes +Lexical dictionary files unzip it to somewhere run with this commandjava pathto +This patch fixes misconfiguration of four tests +I am guessing that you must have the same entry point configured for as for +Will add an exploded goal as well +Lowering priority +Or maybe you mean the vote for this Jira issue? Then it seems only Jacopo agreed +I wonder why you need to specify that lib jar path to begin with? Arent files in those jars picked up by the general classpath URL alreadyve added support for such lib +If the retry fails then it just doesnt include the session attributes. +I think it might just be to do with the way that the reader is logging that message +Please let me know if you have any other commentssuggestions +Will go ahead and fix it for +This was on an iPhone too did something similar that might have been part of what helped. +Turns out that the classestypes that are changed are not used in drools or jbpm +Geez Hadoop QA +Have you tried outside cygwin? the priority of this issue because there is a simple workaround which is to edit the script +When compiled with locale zhCN enUS work correctlyWhen compiled with just locale zhCN or locale enUS only one number is formattedIssue here is locale doesnt exist or was not compiled into applicationPerhaps when setting localeChain an error should be thrown for unknown locales +The old datanode will +No it is basically a average of multiple runs +Jason can you rebase +Findbugs warnings in just moved around line numbers in the new class there are bogus warnings around wait and notify in theres bogus warning around writer variable access due to making the method synchrnoized trying to add lock instead +Issue not present in the test Id created though I did duplicate this on Doms branch as provided +seems like a good place to start +Please try this using version of the plugin +On commit any uncommitted takes are removed from the current byte total +Im always impressed by how quickly you guy response +comments on Johns latest patch No change to is required +Please close this bug as Not a bugThanks +Does anyone see a reason why it shouldnt be +Would the NN ever return aninconsistent such that the file size has changed but the last block size hasnt +sorry about that +do not have to be registered and yes I think that could work as a solution. +However I am unable to work on this at present +I will make the change made the change as suggested here accdfbdeccada +Again ive had a patch up for a month and I think it makes our release more clean to depend on this real release than to have code copied from apache labsI dont think Noggit is a good example +So from me +The simple Inject is what binds to the deprecated api +So in future i will try to avoid this mistakeThanks a lot for the help TedAnil GuptaSoftware Engineer II Intuit s get this inAnd lets please not argue about empty lineIll format the patch according to our guidelines and will attach a new version soon +One case Im wondering about is when get is called on a Job with only events +Thanks for catching it Giri +Attached the patch +If you patch fixes that I am sure that this will be a very welcome contribution +Indeed it would fail sorry I uploaded wrong version of the patch +Is this coincidence or is this really the content of the? AFAIK these are illegal characters for referring to tables +As a work around you can change this one line to your naming strategy of choiceprivate static final namingStrategy we change it to the and have had no problemswe also change FOREIGNKEYSUFFIX id to Id this allows us to have fields like some map to the same in the database +Thoughts +I could submit a patch to do that if anyone has a strong preference for it +This PR updates the EMF model to remove the unnecessary bits and updates the binding wizard bits to adjust to those changes +Lets see how many I have to remove to get it not hanging anymore +M released +We just dont want it to be confused by the JVMs GC behavior into paging part of the JVM itself out. +Thanks for trying it out PeterOdd it could fail after passing all the Lucene unit tests +Anyone who wrote say ClosedResolved will now find themselves looking more issues in their announcement. +I dont think using existing helps muchI suspect it would greatly speed the upgrade +Ill take another stab at this +See my comments on about the async routing engine in Camel works nicely with all the and a selected number of components +Maybe it should be set to low until we have more data on it +Just let me know when youre finished and I can take care of the Spring stuff before the merge with trunk +Its probably doable but makes me wonder if keeping the current code path for counter is not way simpler. +this sets the stagethanksRamyaa +few assertions were not done hereSPEC Testabletrue Section following methods of the must be based on the properties provided by the getProperties method of the interface getHeader getHeaders get getCookies get and get +Did anyone try to use single machine but not with local mode but with nutch acting like one node? Maybe this is workaround till bug is fixedI need to recrawl about k urls and Ill report my timing +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +It is difficult to write a testcase for this issue +Verified in. +moved issue to jbosstools so i can link to it +What version were you thinking this should be addressed in +Thanks MikhailFixed in modulesbeanssrcmainjavaorgapacheharmonybeans at repo revision rPlease check that this fully resolves your problem +Well it seems the factory for the launch configuration can now use the standard launching and embedding API to be reasonably generic for most frameworks +Noticed there is a prior Jira issue to this +Ive uploaded a revised version of the patch that incorporates your suggestion +see also for a related task +We should be fast and as accurate as we can get while preserving speedI will fix the constants value as its wrong +For now Im rejecting until I encounter a case where I really do need it. +Fine otherwise +As I said I shall update the Wiki with all this information +In the code you posted more than one thread can enter the if block +srumRAM +Thomas You are confusing two things namely wrapper classes and autoboxing +Removed the if condition and used the existing switch +anotherwords value +trunk Prefer to use as its the best general purpose pool +Please if there is a specific patch. +The entire repo is already local +new classes and most of package +Patch for marking the status of as ReconciledThanks to Rishi Solanki and Sumit applied to r. +Also changed the screen image of in the documentation +The administrative portlets pages do not contain chinese and Japanese +Please provide it as soon as possible +unable to reproduce this on a different machine +Thanks Dag Ill remove the author tags +Would be helpful if you could include a test case in to demonstrate your issue this will ensure that the issue gets fixed and remains fixed +Is there something Im missing here +the commandline params that dont work with this plugin are issues in how axis itself seems to take them I have run across a number of bugs with axis in general in this regard the downloading of the xsd was taken care of I grabbed it like the other one and dropped it in the right place for resolution +Patch to fix the wording issue +I started working on +I filed to discuss this since it seems like a separate issue from this bug +If I focus on instance editor for the association alone it fails to stick there it surrounds the widget but getValuesetValue are being propagated inout of that widget for the association property +Waiting for Android fix +committed the patch to master +I think we should checkin this patch and tune GC and see how that goes +Jackson supports applicationjsoncharset +Not sure if I am meant to or if I have the wrong was my fault +The patch enables the failing tests that were already in the codebaseWe have just ported the selenium project to and I think all sorts of TESTS are the valid use cases for this patch for instance A test starts a process and wants to verify that the process is running +One option is to wait for before assigning focus +Manually test added. +This will be resolved before release +Test report j profiling results +Thanks +Is this not a dupe of +We usually dont appreciate sleeps in unit testsSo I propose that you isolate the issue into a separate jira if it is still current +Raising priority due to user request +Bulk resolved close state issue transition +Attached patch includes specs and catches the same fault in another location +Thank you for the test case! +Rejecting the PR as it is out of date +This probably will add overhead especially if the request is compressed +So you can also use the assertions and data used in which is for the sequential cluster classification +Just ran into this one hangs and leaves a java process running with the test output file locked +If it works fine I will commit this unless there is an objection +Now that Java ME support has been removed from trunk I dont expect this bug will be fixed +The events from the NM looked good +Michael everything should be fine there. +Update the issue summary so it makes more sense +Jar put in dist +Thanks SteveMarking this incompatible as it changes a public API. +yes some unit tests fail with IBMs JDK +Kostisafter having applied and this patch is definitely no longer valid +I dont know who introduced it but isnt it enough by just removing COMPILERFRAGMENTDATADIRECTORY option from xc compiler +Humm +Looks good verified in JBT to write release note now release note text has been written. +Committed to. +Ive created to continue that discussion +It seems like it would be useful to support it no +Ill try provide a log snippet soon +I took a quick look at the migration handler and it seems like its doing what its supposed to do +Some first step changes are commit at Rev to fix the compile errors +trunk +jar to be used to hate to say it +Id like to make a comment and ask a question There are two files missing license headers +This has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Kevin the security issues are all related to in the Java browser plugins +Very much appreciatedI just need to do a bit of testing etc and then Ill commit it +There is no virus scanner and the lock is not held by explorer but by the java process itself +With the approach the changes on MM would be Add compression info into Use instead of Producer in MM Add the batching and retry mechanism around Use a specific to send and maintain the cached metadata structure Reconstruct the Message object from the object and then construct the object and then the and call the would be implemented as a different class which will mimic the Producers behavior but would use a different interface that takes instead of. +The preferred fix will probably be to change the caching from being around the connector to being around the work item +It is not yet understood how to refactor it since it calls C constructors which a component manger is pure C +The two failed tests are asnd is not related to RAID +Thanks +Duplicate +we are using the tranql derby connectors that use various kinds of datasources internally. +Hope to release a more recent patch soon +The magic translation on that boundary is well and good but its a partial solution at best +The following is thrown when both positive as well as negative inputs are given for these typesExceptionReceived content is faulty value must be non positiveThe existing testscases XSDnegative and XSDnon can be used to see this problem + +Bug has been marked as a duplicate of this bug +Patch changed back to original double rounding mechanism +Remy could provide input for this for the next release +I am really not sure at the time of the sort whether both template and row data should be or +Bump until we can solve the two issues cited above +Cheers of the tab was moved but tab was not removed Committed revision . +This was put into the wiki. +If I understand your question correctly as Namit mentioned in his comment the metadata for partitions is whats used not the tables metadata +The server is specifically Enterprise running in a RAC config but I doubt RAC is contributing to the problem +This seems good to meI let Gilles review it too +Shouldnt the CHANGES entry in this patch go into contribCHANGESIt can Ive never quite been sure +Patch looks good but lets get a test in +Removed out of office that ended up in JIRA +Done +An example of an RPM that does add its own modular config to etcI did not even notice etcaccumulo in the summary +The synchronization logic is only present for JDK now for JDK or above no extra synchronization occurs +convertion of productKeyword and screens and rev +Resolved +This is for earlier version of Hadoop not for commit here but illustrative of the changes +I will help out the testing as well +I think this is a workflow problem +This should address this issue. +removed trailing space +Perhaps similar work could be done on Windows Phone +Applied +Fixed some time ago. +Closing old issues +Fixed in Nightly build for Sept Based on patches provided by Greg Ludington Steve Stair Saul Q Yuan Add ability of required to handle checkboxes radioand field types. +As I already mentioned meclipse dont have references asm bundle +It appears that installer embeds the full JBDS site +harness to explore cache and synchronization behaviour +We should update the demux pipeline to work on hadoop and above as soon as possibleWhen hadoop is in maven we will use ivy to pull that jar file in another jira +Thank you Jrn for pointing the problem +Good solution I totally agree with you +Maybe GWe really need a streaming API for large rows but thats a different storyI take it that maybe this is something to consider for +Here is the mend I made on Master to restore what I broke when I applied master rewrite +I added a note to the wiki and also set the target and source javac options to as Antoine suggested. +Thanks. +dont have access to close this but please do so whenever +Submit to hudson +I want such a job too +Builds but is not testedThe samples should move to genesis parent pom for new sample is genesis not the sample parent +The test that fails isnt a regression so +Where is it supposed to be +Added patch to null out HIVEOPTS before calling exec from metatool +Please append follow parameters into +This manifests itself with qpid c failing with could not open null but qpid c +there seem to have been some infrequent changes throughout the year so I will also review the wiki itself after working off the dump then Ill have Ben close it down +It would be hard to reproduce now +Can someone delete the file file called is the file you want +Okay let us stop at +The logger is not visible patch attached to fixes this if you want me to apply the patch to trunk and just say so +Committed with message tweak +Closing all tickets +Bulk closing stale resolved issues +ok fair point Jackson can be configured not to do that thoughTbh Id be surprised if anyone is using this GET operation as its not much use without the start date of the job on it +looks ve committed this and merged to +That seems to me like a nasty thing to do to people. +Thanks Shai +I applied the change for adding the GR option to the compiler options +Please close this bug if satisfied +The super value of the sub process is still intact +Attaching version rebased to trunk but not yet +New version with shorten links +Otherwise if the loaders answer varies then return truefalse selectively in the implementation +Running tests +Rebased and merged into the master branch and then into the branch. +Please review +Apart From these issues I do not think that we are considering the TTL case in +That seems pretty wrong +Im fine with removing the line entirely +Can you please attach your WSDL file? One other option is to try generating the client using ADB +Its also up for discussion whether that kind of class should be in +So all issues not directly assigned to an AS release are being closed +If both are populated WARN in the log that the className is ignored and call the initFunction code +We will need the other developers to have the same permission and it looks like David Welton or someone else with Jira admin permissions will have to add them +I really dont know what might be wrong with your execution +Thanks a lot for the review Eli +Even reading this variable on the needs the class to be loaded which in turn means running user codeSo for the initial proposal for a configuration property +It is enough to store the revision type in the parent table +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +The datasets never get too large but the matching can be quite complex +It is likely that it is still an issue +Just found this old bug report about the same problem handled with the patch in bug +Thanks for the multiple iterations +we can add the multiple RMs check later on +We now have a single extras directory as we have agreed. +The only realy change I made was to call it a instead as thats kind of more appropriate +What file was trying to be edited? Can we get that added as an attachment +This fix was designed to be as performant as possible +If you feel this needs to be backported to as well please let us know by reopening this bug and well start the vote process on the STATUS file for the branchThanks for the patch! +Fixed in geotools +Isaac please fix for and trunk +Andrew did you get a chance to take a look at why may have failed? Would you still be interested in pursuing this patch? +Ciaran I have applied the patch based on your proposal at revision +The mismatches in are because the new code put t on its own line and caused every line after it to fail +The class leads to believe that it only reads incremental imports while the javadoc seems to be more generic +Ok discussion is moved to but this one we can close happily +I think that limiting children to k nodes is artificial +Ill progress on this basis. +Bear in mind my earlier comments about this patch when updating all objects will be loaded even if they were not initially loaded which may be unecessary and inefficient +Check out how the shell works +This is because that approach suffers from +Let me know what you think +Please check now and add me to the listBest Regards +I have committed this +The change has been made for but given the problems with and the refactoring that will go into them in Im marking as fixed and other changes will be dealt with in the refactoring +I couldnt duplicate the problem today +Building with the latest IBM Java explodes the build in a different way +Just executedmvn clean compile package Dhadoopand it was a successful build +The other projects will come in a future patch +Ive just finished doing this for the cattrack instance Im running internally in watson for use by was fairly easy so at some point in the next month or two Id like to try it for jikes rvm +I have reproduced the problem using postion testcase +I dont think that this is an issue +Clearly the two gets are the sameThat way only the needs to have the system directory defined in its configurationThe other piece that would be really good to pass back to the client is the default number of mappers and reducers +This is indeed necessary workThank you +So Im now leaning towards switching to true by default +OK +Its OK using another machine with IE . +Not using the usual mechanisms for finding service providers +Other than Ashutosh I havent heard of interest in reviewing this change therefore I am assuming no else has started reviewing it +Still need to add tests for the new directUpdate functionality +Exactly how it was before in other words +We tried it today it worked fine +We were not pausing between retries +Dont loose time +committed to trunk +Thanks for your message and your solution for interoperability with metro +Uploading a slightly improved version +JoeDid you create a separate ticket for the torquebox rails error? If so what id? request merged +It shouldnt be that hard. +This patch changes charset from to when reading headers from as use exactly whole byte any character will not be modified + +Ill take a look at the updated patch +So this issue was fixed upstream via and the latter I forgot to backport to +Fixing this in the code seems tooAlso the log message that is getting in the way is presumably useful in other contextsWouldnt it be better to simply do something like the following in the affected scripts? mypidqpidd check grep v LoadedI would propose that we leave this small change out +You should now see a Hello World message appear in your application servers console +The fix for in Rev +trivial of these dont make sense urls are always case sensitive regardless of file system +applied the suggested fix +suggest to remove thisMaybe the use case does not convince you but I would leave this in for the moment +issue fixed in iPOJO +Essentially the same API until you put a in there or some other rulesetOnce that is done then would have access to the delimiter that actually matchedBig question is whether that API change to is good +I fixed this in SVN +I am fine either way +Thanks Yan. +core tests +And I expect the bugto have something to do with this logic either when you update the last column with missing columns in the middle or some problem during undo with same logicWhen you do the compress as knut said the system underneath creates a whole new table and inserts new rows into it +I think that if you just inform that this error can be caused by the timeout when using SSL will help a lot when other people get the same error because they will be able to go straight to the point and not loose time searching for networkconnection problems like I didThanks again OlegThats it +Iknow there are plans in talking with Joseph Smith and others for it toperhaps run automatically but these capabilities do not yet existWhats more servermaint is a tool which runs along side something likeMesos +more information about this error based on Venkys testsFor the case where the job succeeds even though the script fails the following logs appear INFO done ERROR Script failed with code INFO done INFO DESERIALIZEERRORS INFO SERIALIZEERRORS INFO Task taskm doneFor the case where the job fails INFO done ERROR Error in writing to script Broken pipe INFO done INFO DESERIALIZEERRORS INFO SERIALIZEERRORS INFO DESERIALIZEERRORS WARN Error running child Broken pipeLooks like exceptions thrown in are being ignored +Dmitry do you reproduce this problem regularly +Please see attached output in html for review and provide either feedback or a thumbs up so we can get the changes commited +This is a problem in that has been fixed in +So what do you think a legitimate use of a warning log is? To me this is the essential use case report something to the user that might be a problem but only the user can determine if it is or not +I am more in favour of abandoning the METHOD based mechanism in favour of enumerating the objects that can be accessed via the management interfaces and adding ACCESS and UPDATE possibly ADMINMANAGE operations to them where possible +Dmitri have you tried to run Eclipse using JRE? You can also try to point m launcher to JRE +committed +Hello Alexandre looks like the problem comes from the value of the status field +Fixed in r. +It is definitely a newline issue +who can shed a light on this +different exception but is also a regression that we encounter when trying to upgrade to m assuming the comment has helped +there isnt usually much point in submitting patches w broken test suite +Could this be the same issue asDave was using as well +Actually it was checking first if the name of the attachment is correct and then if the name of the datahandler is the same as the original file name +Compatibilityis preserved by keeping nutchs around and adding extra code for reading from nutch if version is Also changes toString as hadoops does not have a good toString method +Tests are also included patch file I hope that you will include my changes in official source code +is a misconfiguration in this is unability to parse valid xml +some code to correctly check for duplicate type names after we convert the name to a java name +With this patch if the bad DN also stopped heart beating you will end up avoiding it during recovery at primary DN but if its still heartbeating but bad then you need to tune your timeoutsBut as I said this is beyond the scope of this issue and I want to keep this focused on the namenode side of things +Committed +the size of snapshot is equal to the size of volume in DB doesnt matter its delta snapshot or full snapshot +Hi ChristianCould you give us an update on your findings? Do you think youll have a patch before the RC? No worries if its not just updating JIRACheers JamieI already committed the part I have done for all branches +I actually liked the first base class approach better because that makes it more future proof and dev friendly +Ill commit this patch tomorrow unless I hear that changes are needed +Test code attached +this would resolve +Will be in. + push to +Closing old issues +This is a duplicate of +Note that this issue is still relevant in empty directory filesets are not created when generating an archetype +Many thanks to Tsunayoshi Egawa for providing the original patch +Please ignore it +Im not able to connect to the forge process using the +Sorry some other higher priority tasks delayed my plan on this feature +Close but maybe is actually called by the bootstrapping node so the check needs to see if the current schema version is empty and if so pull +We might want to fix both problems +jar itself +I am not able to replicate this +Im going to commit this momentarily +New patch! This one fixes a few of my mistakes in the previous patch +Do you think that we still need this +Changing from to works fine +Missed while creating the patch +OlegI do in fact have a reproducer but it doesnt appear I have permission to or edit this issue +Already been fxied in the latest GPD. +Looking for reviewersopinions +i found the problem +One implementation I had in mind was to have a kind of replication central controller that resides outside of all the clusters +This sanity check will optimize the cases when Pig guesses the double field as intlong by eliminating the exception handling +So Ill try the JVM switch first I think +Patch to upgrade to Hibernate core and relevant Annotations to the same versions found in AS additionally Hibernate Search which is the latest stable release targetting this core are running fine Ive tried dvdstore and it also looks like did not change the wiki example which is containing its own set of libraries +run command would probably be a more appropriate name +We can default it to ascendingI think without this it is not future proof +Closing +patch applied and reviewed I did a cleanup on your patch and replaced pages automatically generated by pluginsIll update the site tomorrow from the office where I have a faster upload connection. + +Fixed logic in toolscp f etchosts echo ip DOMAIN stratos domain echo ip mb +I dont suppose theres a JIRA for that yet so we could watch it +It does not depend on the +So that looks like +If this is urgent then please call my cell phoneotherwise I will reply to your email when I get backThanks for your patience amr +All are pretty important but all are additions to this API rather than changesFinally I have a little problem in viewing a lucene index itself as a single matrix +Adding document for requirements and st pass general design +And that switch is going to basically require Autobase to run before the Hibernate plugin which Im not sure is supported by Grails in any way +I havent looked at the implementation details but I like the committed in revision . +given a good hashing method you can be faster than n compares even if you have to create n objects first but that might be a big n since creating objects is not cheap at all +Fixed by dependency injection of Settings into Mojo and then parametrizing theResult adding B removes upload progress from the appliedThanks! +Can you try +And where I said that it was more of a generic JMS thing +They both mention tables system tables and views but they have no entries for synonyms +It would be good to have this classpath cached in a similar manner as the is doing +Is it doable to produce a test case out of the scenario that sijie mentioned +Thanks for the feedback Mahadev and NicolasHere is patch vI will provide patch for TRUNK after we finalize patch for hadoop +Maybe the tactic is to warn that the copies dont get copied +I will rev up and figure out where things are going wrong +I just checked in the shared project +Please review this and apply to trunk once considering this is a valid fix +Observer this even for Registered template +It consists in a composite repo scripts to checkout a given branch for all it enough to mark this issue as resolved +Fixes it so that if all fields are blank the context mapper andor message composer will be removed appropriately +It already have many abstract base classes which just needs to get extended +Not a bug. +It is pulled in by a layered product the signed jar is also causing a! +All patches have been reviewed and checked in as part of the types branch rework. +From a users perspective Im very much in favor of implementing methods in DGM style +Assigning this to him +Maybe they should even be target +Hi KenThe rel attribute is extracted in Tika so thats fine +Because as Benjamin said this is not really critical anymore and we are too busy for now +This will solve issues uniqueness and the usability issue of having all the indexes in the you considering fixing this for +Check if you have any old SAAJ jar in classpath and remove it and add the SAAJ produced by Axis to your classpath +On the attached file I propose a possible refactor +Cool! Then I guess lets try it out +we most likely wont make this until what is M and are there any chances that Spring will be available for CF +So in summary this approach can be addressed now its slow it works +Then shall we resolve this as Invalid +In I can add this parameter +MacroScott OthersSince last few days I was looking on the commit notification on Groovy conversionNow I can say that I am comfortable in doing development in Groovy +Note that until and have been resolved the demo can only show a single cache manager and single cache in action +Good catch patch attached +Its the level for something is probably wrong but application will continue working mostly as expected what I feel is the right level for a missing text +This should be the correct patch sorry about the old can definitely use a readlink if available but the problem is the plan B in case readlink is not available +Im not actively working on this right now +But it seems as if this issue is related to only when creating the with the +What is the first error +Jim here is the change +I guess your browser is caching the page +Reopening so that I can update the fix version field. +I could reproduce the problem now +I propose that the s mentioned here are disregarded +and heres the related patch +For example what if you have both annotation and xml mappings? Should XML take precedence over annotation? Should we throw an exception +Thanks Nandana for applying the patch. +Ill set the status of this to ResolvedFixed +Move it back in if you think differently. +Would it be possible to get the server logs for this? Im not sure how to get them from the smartfrog run +There are pros and cons to have packaging outside of the source code +Thank you Jacques +The URL params in the JSP can simply be changed to make it work fine +Shweta can you please attach management server log file and the mysql db dumps of cloudcloudusage dbs to the bug? So whoever picks up the bug will have some information to work with +I looked at the part of the patch where you are setting to jdbcderby +The recently added bundle persistence managers cache +Looks like comment holds only reading partitions with old names but dropping them is a problem +for patch Nitay +Hi AndrzejThanks for your comments +So I can choose to display checkstyle errors only from the checktyle plugin and not all of the checkstyle columns if I want +And hudson caught the test failure but it was still committed +oops wrong issue +The connection pool implementation is responsbile for closing physicalconnections when they become staleI guess your application does not profit of connection pooling as it never hitsthe same host twice +The change in the attached patch closes the session also in the standalone server avoiding the loopEven though closing the session might not be the optimal solution it is much better than an endless loop +This patch allows a more robust host parsing +Look at the content of the nutch script to see how the solrindex command is converted into the generic one +some were lefet out +Martin provide a test case for this or we will push it until after GA + issue is a part of in. +icon +The earlier versions were left behind in the repo and the had the earlier version set to loadfalse +Is correct FrancescoThe patches files are equivalents +strangeThere are some problems with the git version above includingtestcsharpProperties is missing an opening using keywordtestcsharp is doubledIll look into it further and update with a revised patch hopefully tonightId also like to suggest that the event in reference the C compiler via an environment variable +Thanks for making the change it looks good +Patch that fixes broken build and solves the issue +However we already throw for a lot of our API if the argument is unexpectedly null +I dont see how this would the converter worked for all subclasses how would the conversion from String to a targetClass instance know which subclass is intended +Hi GertYoure rightIts working now on version Thank you yes didnt scroll far enoughChanging the mentioned lines in Transfers and upgrading to works great +This is working for me now too. +uhmmit appears that is not passing the parameters for the new jvm please correct me if I am wrong pts Sl usrlibjvmbinjava Xmxm XXm XX XX XX classpath optmavenboot optmavenbin optmaven coberturacobertura pts Sl java cphomesonar +Vinay I think thats a separate issue as its packaging in the tar files +Thanks for testing the new feature Jim +bq +hh +a webapp for doing JNDI lookup and see if an object is bound to a given name +You can also remove the shiftSpecial that is only for our appHowever here is my patched code it is using recursion and if I remember correctly the code got even a bit cleaner because there were two similar code routines that could be unified to one +but for now this patch will do +This patch add the exception message to log infoThe patch must be applied to the trunk +Again applying your patches is joyful +We need to make tests run as fast as possible +Deleted docshtml from CVS +However their platform specific parts are either not processed or fails because we do not have the treatment for that platform in place +just assigning for CR for ensure discussion I dont believe it will be done for to Brian since he got the task of defining the simplest webservice wizard that will make things work without too much work in CR +Which JDK distribution do you use? Sun IBM BEA? Did you override the default parameters from +Thanks for fixingThis problem also occurs the next version will fix this problem it +Oh and I need you to test the bash scripts The client scripts are added but wont run because the required jar dependencies are not in the bin folder +There is a task for invoking however this task requires a installation of +Thanks for thequickly patchJust as Claus Ibsen has said there are still other tweaks in OSGIThe end user must make ensure that the generated JAXB class is visible to the camel routes bundle some times the should manually import the generated JAXB package if the generated JAXB class doesnt exists in the same bundle of camel routes because the cannot recognize the spring configuration automatically +Wondering how this case is handledMaster asks a RS to open a region and I guess adds it to regions in transition +User +in my eclipse i can go to WindowsPreferencesJavaFormatterExport +A more common practice is to use the or me to these settings +Moved it back +Have you tested this manually with debug points +Dont want to commit in a rush +page on Stage cannot see the fixes on stage +Attached a new patch but it cannot be tested while pending patch is not committed +issue resolved in designer +I am not sure this is a bug per se +Moving this out to but we should definitely backport this to when done +Here is a patch that implements the method described earlier +Ive not tested but it seems to me that Adrians last change should have fixed the issue. + push to +My own digging shows that rsync is possible I can use it just fine with my own idpassword but there is no public id that maven could use +Ack I keep forgetting that we do not have target version builds +should not trigger +Walk through to get the problem with the error require the two ruby main ruby file will call the second file with a require testcompile the files keeping the directory structure in the directory structure intact separate and remove all the ruby source that should be remaining is the compiled class the files with the proper manifest to point to the main the jar file with the java jar above error will be shown on the command line +sorry Erik somewhat it was not included automatically it should be working now +From what I can tell this only applies in for this check returns all class methods defined on etc +Please check nowthanksdims +Same behaviour in s still on my todo list +As youll see Ive tried with everything from and up and cannot get it to fail. +Will wait for to commit +Which Commons IO version did you use? surely can not shutdown Reaper thread in some casebut above the Reaper thread source had been rewritedI think it will be fixed +Do I have to upload it again or is it ok with this statementASF Granted License granted +So I think this issue is fixedDonald hoping you can verify? +Reopening for handling +i am closing the issue after compiling source lib with new everything works like a charm +Could you please post the Exception here? I want to try to find out wherehow we could possibly grab the root cause + + +Patch for added +The workaround does require a restart of the jobtracker so I will go ahead with the patch but since theres a workaround Ill only check it into trunk +Are you saying that it is now illegal to have xml comments in the +Reverted r fixed at r +We can copy ideas from there but we do not need to maintain compatibility +Thanks Dawidgood and bad news +Have you setup that final project dependency +We round of to a multiple and should work in most cases +HiClosing this issue attached resolveds the you have further queries with respect to this please the associated DEVSen +Btw +Can you please verify? Thanks +I was warning about from the time we have decided to change the way version is handled +Theres some comment about not using the database name inthe comment to the method causes some problems but Im not sure having the code perform a different action is the correct solutionId also assumed that to have a connect to the system a null databaseName would be used but this test uses a database name +I have updated the sample FO and now it really does demo the bug +Just something to point out +simplified to reproduce the errorAssuming the and tables are created. +Can you do me a favor and try downloading stock Eclipse from and try running it without any extra plugins etc +It looks like there might be some API changes to that upgrade +Arpit should the comment say milliseconds instead of seconds? Otherwise for the patch +Thats a good reason to continue working on that ticket +Ok will commit soon unless objection +So we agree on moving this to +LukasYes necessary changes are described in Part I of froum is not fixed and we do not have a special JSF version of RF you can use common snapshots from Hudson +Test failures are same patch to kick jenkins again +Agreed +to trunk. +Although even with I got like ss since we do have less things to serializedeserialize and we compute only with floats. +Patch for trunk and verified in real clusterI will submit the patch for and after verified in real cluster and done the unit test +I dont believe that we need the relativePath element I just put it in out of habit +The class is now documented hopefully I have made the concept easier to understand +David as mentioned in the call yesterday we do not have access to that code +It should offer a new service and leave the option to override the to the +Thanks Chip now Im able to boot the database and reproduce the problem +The approach must allow for cycles and does this by essentially picking up where it left off on a given revision when it tries to populate a revision that it is already populatingWith this approach it is possible that a given revision can fail at a deeper level in the recursion stack than where it started +The tests also provided +Looks like this got fixed with the commit of. +And if so implementation could be improved + based on Nicks should have been closed a long time ago +Moving it to which we will attempt to certify +Should we really be picking up enough tasks to killYes +Ross can you apply it to your version and see if it fixes the problem youre seeing? Henry can you glance at it and approve or propose changes +This way someone could update the original source file database whatever and have the changes take affect +Closed with the release of. +This patch adds the bundle as a compile time dependency and fixes the compilation issue +Resolved as part of Mandanas work on getting content assist to work for +v committed to +Thanks. +Who will be project lead +Its actually quite common and started before my changes +You could install it with npm install g cordovaIf youre putting symlinks into lib for other reasons could you explain what those reasons areAlso Josh that logic in the PR looks backwards +yup hadoopqa is broken +I was recently able to use this code to create a shapefile +Martin I believe he is talking about the temporary directory created and used by the server adapter not jboss tmp dir it self +thanks Tim +Basically there was a thread safety issue with the sampler logic +So closing the issue +This is only being applied to trunk correctAlso it would be good to verify geoserver passes all the wfs cite tests +This must be a typo +I am not against I would prefer we are more consistent and add the missing headers +Derbyall runs cleanly modulo wisconsin noise +Any chance that this issue will receive any attention in the near future +Sounds good thanks! I committed this to trunk +I also understood the issue behind +hmm +You might find the command helpful and it should be installed by default on your ubuntu system +Could you please review and let me know +Configuration is performed using the standard cxf configuration file format so anything that cxf supports should be possible with riftsawDocumentation to follow +Since JDO does not require binary compatibility I propose to defer this to a maintenance release +I was wrong +untested patch +Thoughts +Yeah they do have something +Most because is really hard to make Java compliant in a clean way +I am going to commit it shortlyI am not clear if needs to be in as well +So the basic idea of having counters in the and and not touching the format remains +Closing this as is sufficiently different from trunk as far test setup is concerned +i would be in favor of that improvement +Also could you comment on the initial thread concurrency from default of to +I put a patch up on which blocks this +Therefore the invalid read cannot come from this library but comes from either axis code or axis generated codeMoreover the notify example doesnt use axissvcclientsendrobustwithopqname but uses axissvcclientsendrobust +updated version to you please review ittnx for have been using the on my local machine and bug is fixed and havent had any other issues with it over the past couple of days. +Cant reproduce with Hi prints fine with me +A stretchable vector is exactly what I need actually +I have theories but without a way to reproduce its hard to say +doing a little testing myself i see that after the Add framework manually then remove using the cli step the Frameworks folder still includes the plugins framework but the Linked Frameworks and Libraries section doesnt have it +Logged In YES useridThis is a conflict between the age and size based cacheeviction policies however this has been fixed as of the thread dump you have for the does not match that version +Then i did another reservation for day another for days and so on +its a wiki you should be able to write there I have to admit its not there because I didnt know +I will do it in a separate jiraThe patch incorporated all Nicholass comments +Seems fine +Hi CurtisCan you test lower cases for all variables in? I got the variables when I use lower casesOur configurations is cat etcvclopenstackostenantname adminosusername adminospassword adminpasswordosauthurl http +It is unlikely but if theres a lot of pressure on the map from other tables it could happen +Ineed to find it and study itThis is why I do Jira every once in a while +Test case for page vs +Makes sense +Filed to add any new tests we may have to. +Thanks for the infos Alexey +Hi NickIn terms of manually adding the update site zip I think this may cause confusion as the page implies that the contents of the individual zips are in the main release +Theres a catch however +Regressions ran ok committed as svn for the first part of fix +By strange chance we need to respin the MCF release too +Thanks +Should be live in mins +Close std err and stout when process detach from parent process +Seems to be resolved in +I have split the patch into section +Do you think its better if we implement a fail fast mechanism +The attached patch works for latest trunk +Reevaluate for to Manik for since the whole configschema systems being overhauled. + I believe fixes Todd +Sal can you provide an example that youve run into? I see code in that skips downloading a file whose namespace is already known even if the location is different so Id like to understand your problem in more detail +Agreed +Paths whose are for different filesystems or are somehow unparseable should cause an exception +Uploading new patch this should address all points now +Hi Ferdy the patches I attached were identical for branch Nutchgora and trunk +AS will have support for the remaining Java EE features that arent in and I guess thats the version you would want to use +FYI on why the JDK specific JMX permissions work in the policy file in JDKAny unresolved permission is automatically resolved to and laterThis is actually needed for application defined permission classes where the system cannot load them until the policy has been created otherwise their static initialization code could take advantage of the lack of security policy at that time +affecting certain versions of I dont think it would be suitable for to incorporate itGlyn I will try to produce a narrowed testcase and file that against Virgo +I found only one scenario where the behavior is a bit unpredictableOpen a project with a single Delete that XML on the filesystemWhen modeler prompts to react to the file change click Save +This is going to need some more investigation but I thought I would post it case it helps someone else experiencing this problemFixed results are now only processed if a result set exists +This needs to be thought through before exposing an implementation detail in the apis +Moving to. +Correct +It does look like the plugins file from the staging site is bad +The problem is gone when I moved to JDK and higher +I dont have the code in front of me but I believe there is a case for this with any of the readerswriters we have +I noticed on my PC the FOP rendering drove CPU utilization up to so its working really hard doing something +Because of there is no reply from reporter other people running Derby do not report such an issue I porpose to get it resolved as Cannot Reproduce +I see the same exception trace that I posted above +Hi BenI tried with lowercase and it works +Its STS originally but upgraded to last week from the s update site +Let me try to do a proof in a unit test and then propose a solution +A small amount of work is still required to automate collection of the coverage data generated during unit interop and recovery testing but generation and analysis of that data is automated +As I can see this error is not caused by Axis one possible reason is your application server is fully utilized does not have connections in the connection pool +Hi SidThanks for reaching out +This is what I meant in +After reading more about the classpath this should be fine +Ive not run into any problems with either or THANKSWhats the current status? What commands were you using? What platform +A patch against that allows for injection into a generic superclass where the field type is a type variable +applied tos patch ported to CVS HEAD +Kristian askeda Is the test all about getting the stream and the length twice? If so maybe the could say thisI went back to and tested the repro it does in fact require you get the stream and length twice +Also a test case would be nice. +Not a problem + +All tests passed with the patch +I have verified I can put in the same exact value for both +I just checked out built and installed CXF on it +This is a duplicate of close of all resolved issues. +Fixed on and work thanks Andrea +What was an example of the whitespace plagued usage +The pattern matching operates on instances of not on strings +Ron this sounds like a good idea +Instead of blocking access why not secure it like the other hadoop admin pagesThis patch doesnt block access to the UI +Actually isnt quite a replacement to mvn archetypegenerate +I think shrinking the list is really bad because the list indices might be significant. +HiIn the earlier patch the service bindings section was overlooked and now it is fixed and all tests pass +Heres a new version that includes double underscores where necessary and the keyword dynamic for C +How do we feel about that +Wait until Geronimo +Lets hear what Jandalf has to say +Definitely with a vote +Please read the comments in the pom +let me know what to do +if this is actually fixed could you close it? +caching the bytecode dynamically for the relevant instanceNote I actually would prefer users had the option of declaring the type of a variable so we could validate a page partly at dev time and fully at runtime but that is not for performance but for correctness and maintainability +released with +Provide a unit test in the patch that tests the positive and negative aspects of the patchThese steps make it easier to find which directory to apply the patch and the unit tests make it easy to see the fix in action +Issue went away when removed the can close this issueThanks so if someone else ends up here or if someone asks me +Any word on whether this solution is getting the thumbs up? I personally need this functionality and would like to proceed in a manner that will ultimately be accepted by the community +So the current discovery solution establishes a new generatedmerged JBDSIS file containing both JBDSIS and JBDS Core content in a single file +Matt No because a live cluster gets its logJ from conf Thats why this should not impact any of the code server or client but the file in the JAR does stop any other client app from controlling their logging +I accept that not everyone would agree +The installer plugin is able to generate both binary an source packages out of any Griffon application. +Note this problem doesnt occur if you use the new Microsoft SQL Server JDBC driver with SQL Server +Sorry should not have checked Ship it it looks good except for the necessary change needed noted in review +bq +Tim your tarball doesnt have any contentsI messed around with the api some yesterday and I think this would work +Hey GuysIs there a maven repo for Kafka at this pointI sure could use it if there is oneCheers you are seriously my hero +scratch that ill do it. +If we transport uncompression data we should do it as same as your great idea +This patch fixes the issue +Created a new ticket to better name the intendJust from a single test I did with the cold restart sec warm restart all resolved tickets from or older +hope this helps +Yes FORWARD is one of the dispatchers for the filter +Commenting out the import in the test project as it doesnt impact the result +I would be okay with unlimited as the default. +So the message is printed out twice to the default appender and PIGCONSOLE +Still getting yellow marks on the reopen solved it +will ping Robert and talk with him if we like to cleanup anything else before going on with a release +Added a patch that changes the logic to a best match instead of a first match +Core will break tempararily until all ofthem are added to the repositoryThere are a few other modules which also belong under the LucyObj umbrella +ok removed from help message +attach a new patch which contains Ivans fix and test case + +The pseudo column containing the source relation proposed in the first comment seems unnecessary +Thats a good idea if is always tied down to JCR but do we really want to tie to JCR only due to illegal property keys +Linking to as is blocked by +The workaround is simple although perhaps not obviousMy concern with changing the default is that it would break all those folks whodepend on the current setting to keep their indexing from blowing up +is the result of a getmap request with transparenttrue + +So should be fixed on trunk +For the simple Acegi contacts demo where there is only one active client thread which switches identity the patch I came up with would be OK +Well we will have to address the issue one way or the other and this patch doesnt help us with it as is +Committed at revision dont have a good insight on the test but I have the impression you might be testing for undesirable behaviorIf a request is asking for less attributes than the required one the WFS is supposed to add all the required ones transparently so that the output is schema valid not to throw an exception +Since has added a few methods and refactored Writers heavily I will close this particular entry +youll have to base it then? Is that so? Whats couchdb do? REST is their client +Im doing some more testing with a small change and we are going to have a hotfix jclouds release tomorrow +Of course I meant to say +I have a very similar issue +Matt Also see which will allow you to change the compaction ratio on a basis for clusters +The directory structure needs checking +It should be at the top level of the svn repo just as the current website is and downloaded into the folder of our demo instance during deployment +Are you building a perl server? We should probably add a test for this +Removing Fix Version since dupesinvalidwontfix bugs clutter the release notes. +has been marked as a duplicate of this bug +Latest patch looks good +Text tests passed with it +jbp is injected to build before runtime compilation but there is no jbp archives to unpack during building JBDS installer +Still reproduces all the time for me Apple JVM OS XI guess this means I have to debug it +Unsetting Fix Version on unassigned issues. +a simple sh is used by sh sh +This contents of this patch have been incorporated. +patch to fix tc owner and permissions while installing hadoop binaries using rpmdeb pack +What if we document it +We will ensure that this is seen to in due course +Hopefully we can make a final push to get stuff all wrapped up in +Committed in revision +The value of select is not NULL when Im selecting the empty itemI have no idea how to prevent this behaviour +Would that be an option +it would have been nice to include the extra information for Thrift clients too even if its just rendered into the error string +Not that its difficult to port code between the two but it would be nice to reuse the exact same components without having to load two rather large js libraries on a T pageIt also feels like jQuery is more widely used with a huge number of plugins availableHowever as Dan points out such a switch could break the promise of backwards compatibility +I just saw the multiple consumer test fail for the first time as wellIve added a patch to the stomp connector to fix a race condition when creating consumers +Committed to trunk! Thanks Gunther +This wont make it into this release +Nicolas the originator of this issue agreed that only Mapper should have the ability to receive the input cache +i also removed cycle from the main pom in the check profile as the continuous integration wasnt runningill see if i can fix it by removing the workspace ll include cycle in the check profile again to check if the force clean build in bamboo fixes deleted the excluded test from the +I will commit this in a few days if there are no objections +This is a Dojo problem +Id think that API change should be done before we make our first release with coprocessors +Im using not but otherwise yes +Sorry I didnt know you had the proper fix ready and I am on a tight schedule to release +I tried you patch on a form with a listener on upload and it worksbut when there isnt any and you submit the form with another input typesubmit the fallowing error occurs Submit widget already set to Upload upload +If I create the shard first and I dont have enough nodes to sign it to it should probably just fai just the way create collection does +This is just such a bad idea +Is it included or are you using +Please be more specific on the components that has this problem +Given the info you tracked down posted already and after reviewing the posted patch the patch looks good to me. +Thanks Kanna +The deadlock will be fixed by the synchronization through the FSN RW lock +With this patch in place the tests passNote that this patch is dependent on the patch getting committed to first +So checks if the error list is empty on an update and if so will disable the layer. +I am getting those log while deploying In UI +Thanks the sent privately fixes the issue + +F still doesnt work with native launcher needs more work +Just cleaning. +The fix is extremely simple so it should be easy to put it to trunk +I do not think this is a problem anymore with the maven build +r for translations +BTW what path does the plugin look for +Note that the patch attached to this issue will need to be merged with the patch attached to assume this havent been applied yet +Thanks +depends on so it cant be moved in its entirety +The hadoop command is still supported and will give users time to move to hdfs over time is that an acceptable solution? What do others think +So all issues not directly assigned to an AS release are being closed +And I contend that just like other caching this is not Hibernates responsibility to manage thisAlso we need to outline invalidity on changes done through Hibernate +I kept the memory settings what they were but the change for adding debug ports should make this ticket much easier to implement +Your ballThe code no longer allows for provision of a default message code in the rejectValue methods which used to be used to resolve against the app bundle +I looked into this a bit and I decided for now to use a modified default configuration given to grape +is an issue for getter accessing to public static fields +If shows up again please reopen and attach the relevant TRACE log. +Alternately we could cheat in the loader and describe the ring ourselves but that would be going outside of CFIFCFRR +One think to check how much Ivy uses the network +This worked fine in Eclipse when the path to the context schema is a file URI but in maven while building the path is a jarfile URI which is opaque and does not resolve against opaque URI contexts so relative includesimports fail + Yes that was a typo Im well aware that distinct has to happen after the project was trying to write this up hurriedly last thing before leaving work sorryI think for our database REDUCED may not be supported so that doesnt help us muchStephen Yes I remember that discussionAs I said in the original description I will run some actual tests and report back with some figures Ill try the various different suggestions and see what a link to the SVN location for the test I will look at including that change as part of my testingHave a prototype transform for the evaluation which I am testing out atm will report figures for that to see how much it improves results from running with my new optimization in placeSummarizing as beforeORDER BY DISTINCT on same level ORDER BY DISTINCT in With LIMIT used results are as beforeI think this applies to ORDER BY REDUCED as well so I will try this now applied the same optimization to REDUCEDORDER BY reduces times to followingORDER BY REDUCED on same level ORDER BY REDUCED in same query Not as big an improvement as the ORDER BY DISTINCT case but still noticeableI have also generalized the optimization to allow ordering on arbitrary conditions provided all variables used correspond to variables in the project this as fixed since we seem to be in agreement that the optimization is valid and it gives a big performance boost for a certain class of queries +Its ok for me +Is this bug going to stay open forever? What should users do if they want to make use of the logj trace method +LDIF Export is correct and doesnt include the name objects dont have attribute name in the schema but it is shown in LDAP. +Since is not released yet its still time to change it + for the change +Sure Ill create one tomorrow first thing when I get back to the office +Good job Oliver! Ill mark this as fixed against the upcoming. +I am no longer working on that project where we needed to do that so it passed it off to someone else and I am not sure if they tried that or not +What to doCommon sense says to me get the best match firsttexthtml is an exact match so I tend to say the client prefers it over the generic +WARBUNDLESJSFIMPL is broken in AS right now but Im working on it as we speak +Spring EJB support Abstract class +patch seems to work so I close it +Thanks +Targeting the release instead +Applied thanks. +Maybe this is related with apache jira issues will try later +actually the only thing that should be done here is to enabledisable an enableddisabled is now there in the UI +GreatUwe can you please update the jar in this issue? I will make sure the test passes with it +Attaching version with fixed test and correction to the one real findbugs warning +lgtm +So Solr is not workingThe trace from Solr indicates that it is the same problem as in alpha despite the code reorg that has taken place since then +This is addressed by the fix in +HiDid anyone get a chance to look into this oneUnless we find a solution we will not be able to release with CXF and well have to rollback to xfirell dig into this a bit + +The space and a group is setup +If I type groovy x y z or x y z runs and gets the arguments +Does this mean that you intend openejb to run on spring only? Otherwise what are you talking about +I feel bad creating yet another but its not obvious what the right way to configure this is since doesnt read in the chukwa configuration files +gitalso upload the entire patch that includes that common change plus adding the license to hdfs +Laura it is the configuration setup required in case of IBM JVM usage for junitreport ant task. +Should be in post ER build +No one has opposed to the idea of changing the references so far +So I guess it wasnt the right place for tried to find where I could hook up some code to verify that before and I guess this is in the class +Yep +Changes implemented and merged. +Thanks for fixing it. +These kind of weirdnesses will be cleaned as we go through OSGI goals +I altered test to that it runs transacted as I think this was the authors intent +patch without int tests that need server +The additional columns will automatically be returned when using the notation but we prefer to be explicit as in the following example for a to a DogNative sql queries which query for entities that is mapped as part of an inheritance must include all properties for the baseclass and all it subclasses +in rev +Pull request merged +you dont use hibernate tools configurationf or this? how come? This will make it necessary to duplicate the logic +Id just unzip it in trunk and hit SVN add and then commit +Each document could have hundreds of changes of which will be lost +Hi KeithThanks for the feedback +This type of query is more suited to the user forum for Portlet Bridge as youre trying to find an answer to a question as opposed to raising a bug or enhancmentHaving said that for your particular case the best option is to perform a redirect as youve been trying but you need to provide an absolute url to the portal and portlet +It is the same +Hi RobertI believe the following patch prevents this from happening however it doesnt attempt to fix existing trees with such valuesIs there a known reproducible case +After add affected product simply click the Save as Draft button which will engage the flows save event as it should and display a validation failure that you need to enter a +Its not the same +Does that really workAbout the Threading issue the must be threadsafe anyway otherwise how would it work with multiple sessions in different threadsAnd it was only a suggestion to useWhen You checkout the forum you will find numberous entries of people that try to exactly do what I want to do here +Ill change this back to fixed since the specific issue that yathish isexperiencing has been filed in a separate bug +Help me understand the first line in the example +Hello DiveshHere is the patch for the given functionalitySharad Gupta +The YY computation is moderately expensive but the subsequent Cholesky decomposition is nearly instant +Sorry about that I did not specify it on all of my questions +Theyre not hence my confusionApplied original patchI still note that although notify and apply treat the generationversion tuple correctly switching to a compareTo would be clearerPerhaps it would be clearest of all to have a single long version instead of ints where the high order bits are current generation? +The key as I see it is simply to avoid the issue going undetected +Change neededUineed to a flag to the UI of the Solr Field Mapping that will produce a boolean variable in the Line add the management of that variable here to allow the correct management of the for the ticketIf you wish you can submit a patch +Agreed can we get some javadoc and source jars published to Maven +Patch attached +Reopening issue to reassign them from version to or before removing version from Jira +Current state +Regions when they get into this state are offline and client accesses get an ISA offline region exception +You should not attach files while creating the tickets as the radiobutton for granting this license is only possible afterwardsSo please use Attach File and upload it again with the licenseWe can only accept patches with this remember to document it on the wiki as looks like camelrun already had this functionality +Migrating old kind of archetypes to new kind is still to be +In this issue Ill add few unit tests for Hama +All known issues fixed +Unix rm does not use the trash nor does the unlink system call nor does DOS command line etc +Thank you Ashley for your clear bug report +Closing. +Is this two issues in one? f so we need a JIRA per issue or a clear understanding if the work around fixes the entire issue +Can you confirm that this issue is still reproducible for you +Ill have a look at this shortlyIn the future the best way to submit patches is to set up an SVN workarea based on trunk and make your changes there including adding new files with svn add +Short of including a binary jar as a test resource I cant see how to build a test case for it +Thanks for your checking Bryan +The fact is though that there are a significant number of people running who would like to use Pig and supporting them out of the box seems worth it +Lets commit it +I am assuming it will run fine under and XP too +The attached patch has an assert which fails when editlog sync is called with either read or write lock held +Moving this from to as theres a work around committed for because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The problem is that itemDescription is always filled when making an order with the description of the locale used by the client operator +It is not behaving as I would expect it I did the same in java it would work fine + +OK +Let me review the case Ive stumbled upon yesterday and update the description accordingly +if someone is going to set a uniqueKey wo knowing what field it is they also dont know the type of field it isif one can assume that the id field is a string why not just also assume that this field is named id +The first argument check was mostly for Windows which had a issue that I couldnt resolve unless I launched the program could try to check for redirect operators on the and launch the shell in that best bet for a workaround is to prefix your command with sh c or bash c in which case it will get launched by the shell +A wait for shutdown to complete would hang now passes with fix to trunk re shutdown thread interrupt. +Maybe John can help here +Also Can you run linux task controller tests and update the results here +When upgrading there are cases in which we can have some downtime until the upgrade completes +Now the tests works as follow holds events tries to receive events duriong default timeout and expects timeout exception releases events tries to receive events agian and checks that expected event is now receivedWith this patch the test stably passed for me on both Windows and Linux against M build +The default location for xdocs in m is still xdocs +If we want use modeshape we will need to download the lucene sources and recompile themIn modeshape we dont need other customizations +I guess when we added the reloading to bincassandra we neglected to also add it to the init +Protobuf utils were split so that replication uses a different util +Seriously? Bleh +And I tested my bulk import changes +Failed trying Hudson would be really good if we can have a test case +Not critical for +The service provider should only be a method on and +has the fix +committed to trunk and +If people want to use the domjjaxen integration they simply deploy a full and as the server configapp deployment level +Thank you for your contributionCommitted revision +If Todd is not around then I will commit the patch so that it wont block Lars work +For bug reporting I added a step about checking to see if a bug already existed but that should probably be better done like your other earlier info boxes +I can look at the patch too if you like +Please check and let me know if that fixes the issue +Can someone try the commands put by Giri? Im getting error when i tried +bulk defer of open issues to +But with raw I mean just the original format ofthe graphic toolSo we can have a folder GRAPHICS in each project and therein we use the same structure like we have in resources. +If so streaming from the node being replaced is not availableHow about force the repair happens if the user specifies he needs the consistency of quorum while the original node has gone +Updated patch +Change execute phase from compile to +I found that the before image was being created prior to populating data from the datastore so it always appears that the data changedTheres no way for a to create a new state image because the method is protectedI havent looked at it any more since I decided to use build enhancement but thought that this information might be useful to whoever looks at the issue. +Attaching a first attempt on a release note for this issue +Thx very much test cases are next on my list + +Added License for SLFJ in the latest patch +good solution +Im sorry for the confusion the documentation should be clearer on this point +I want to put this in also in case it comes to at revision +Using Faceting with the combination of sharding and grouping in our queries an error occurs +similar set of functions performed but focused on testingdemo the application rather than production use +why is using the return value of the insertupdatedelete not ok? And why is the first return parameter better +because its fixed in have the same problem with Eclipse and latest stable jboss IDE plugin version version is still beta and Id like to keep working with a stable version +The issue can be closed +One thing that would help a lot of these cases would be for us to make a version of that works off the Time interface the scheduler is one of the few parts of code that is still susceptable to timing assumptions not controllable from a unit test +I should add testcases for it this patch +I am using firefox and IE +I had to kinda force it by artificially setting closing flag on region and then adding new threads + +Attached is a zip file of the test output and database +thanks! +removed catching the OOM in r +Also youll have to edit to set your connection parametersIf you run it with a purge option it will fix what it can and will purge orphans that cannot be recovered +Revision for on with switch to release a switch to instead of mvn is made +Its been around for quite a while + +We dont need to rush it I think its fairly contained to change we dont even have to deal with this for if we arent happy we can deprecate just have it extend in a future release tooBut I think the name is not really good at all so its good to get all the ideas out here +Ive opened another issue to see if that gets any attention so Id suggest watching. +jitrinosrccodegeneratoria and jitrinosrccodegeneratoriafix of the original problem +you open the jar file you will see the SQL scripts with short column definition +Colins explanation makes sense to me +It seems I forgot to synch but for the issue mentioned it is not necessary + +I wanted to bring up the general issue of capitalization of the name of the network client driver +First version of webpos application to be expanded into folder +If you apply the patch for does the behavior change at all +Nothing else seems changed to meSo basically removing the Override annotation from these method would make this class compatible with both JSF and isntIm asking this because I understand the need to have seam be compatible with previous JSF versions while I would like to have a JSF Seam which is really much easier to deploy then when using FaceletsThank you for your hints +Know youre busy Olivier and I love this project +Committed. +bq +I agree with Jiannans comment about removing the dependency on construction +Theres still some cleaning to do but considering the amount of patches I already submitted for it I think a new task issue to remove deprecated API calls and such would be better +run mvn dependencytree an examine which versions of junit you are using +bundle +The first two of these are resolved for my purposes by +on commit +Applied patch from Erik +Also looking in greater detail at the it appears that even with fixing this issue there will still be bugs +Am I ever this one should be good to go but Im on my way out right now and wont be around to help if it breaks again +Sorry Assignee shouldnt be changed that was by accident +I have also this issue using Compass with Hibernate +lgtm +Thanks Adrian your patch is in trunk revision +Closing with wont fix as there are workarounds and this is just how the lifecycle is +in + to generating release notes +TOC and structure should be reorganized well work on its required to look though GSG guide chapters on JSF and Struts and if necessary move something of it into our new guides on Struts +Bobby could you also add for the Reduce Tasks pageNot necessarily part of this Jira but the can use the same set of columns for reduce tasks +The will have edits from regions not relevant to the tables regionsOver in each one of the multiple can be devised to receive edits from single table +This patch completes encryption in out going message according to policy requirementsplease review and applywork remainingProcessing and Validating the incoming soap message using security policy +In that sense temp exclude list should work for the connect problem on large clustersBut somehow client telling NN what to allocate is less appealing to me +PR for master merged in master branch +Now that contains the requestObj and hence there is a non zero reference to it still +I findthat helpful in these situations +Did I said something wrong fellowsWoody +First draft of Release Note posted please let me know what needs to change +The extracted patch looks good to me for the branch + +Apologies for the spam! +Thats probably JNA or snappy +test failures look like some kind of environmental problem unrelated to the patch +You would hopefully expect a different error when requesting the artifact itself +trunk +if there arent any objections im committing tonight +fixed the problem +Patch with the new and modified tests +However Franks approach and patch is also reasonable it depends on what we want +I need this so I would like to implement this improvement now +Basically it is not possible to add this in general since not all may have a +Without this optimization client probably going to get session expired without be able to restore the session if the root cause of the disconnect is just a temporal network failure or operation timeout on the server side +It is only implemented through DDL +If no rename +From reading I agree with you +If you browse to while youre offline the page should still come up it just wont find the images +As i recall it was quite a bit of manual work to get it to run +Verified as complete. +Does that help? You should see a Resolve Issue link in the upper left +Unless of course you think the existing single test that turns off the phrase highlighter is sufficient +It shouldnt be too hard to use that for +Files are in src! David +Verified in CR +I did a git pull and it works perfectly now +I agree with Tucu that its not reasonable to generate this value randomly given that this cookie will need to be validated by all of the hosts in the cluster +Left couple of comments on +Committed revision . +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Patch applied. +This approach sounds much more friendly and would make it much easier to support a driven approach to entity modelling and also supporting versions +Confirmed +made it so we could pick up multiple files from directory +Sorry I think you can close this as this issue is actually reported in. +Thanks for the quick response +Almost all of the work for the past two years has gone into the v branchMaven does not address any current problem as far as I know +has completely different semantics it determines if the test class or test method should be executed for a given test suite which potentially has nothing to do with which active bean definition profile to m therefore reopening this issue to make sure that both topics are considered in their own right. +The attached file is an alternative approach to a stream based logging interface +For example your have project with jar class which pull in other jar +Why does it jump a static bytes? Is anything else broken +Pull request merged +Please ask questions on the list JIRA is for Struts bug reports +So some kind of profile system would be in order +Probably was resolved at some point. +code change checked into branch and reapply to associated ticket is resolved +Fixed title page to format better +Ran into two issues with the maven ant tasks while working on +I did little research on this problemand it seems that the library provided is not matching with the TLD filesSome attributes defined in HTML TLD are not defined in corresponding tag classSo this is something related to built version of struts +The solution you are proposing use a transparent wrapper that upcasts on access should work +Thanks for your work on this +Attaching a patch that explicitly set the owner for the staging directory +still it is not consistent with my earlier problems when where name LIKE Sting indeed produced a result +I have already begun work on the unit test for this pending some Groovythreading for the patch I will check it out once we have a unit test that fails +I need to add some test cases for specialized stuff +Im unable to reproduce Anirbans problems though but maybe try it again as well. +Allowing the user to upload an untrusted xsl stylesheet and do the transformation on the server introduces lots of nasty security problems. +Not planned for +thanks. +So it looks promising but lets give it a few more days before we declare victory +Integrated to TRUNKThanks for the review Stack. +internal for executors and +r for the documentation update +First rev of release notes checked in at subversion revision +has been marked as a duplicate of this bug +The language has not being designed to support this constructs become hard to read and dont want Groovy to become too perlish +Well shit +The pb is not in but in the startup delay +KnutIve done some research on this issue and the following seems to be a reasonable design for this featureI think the following would work for emulation ofValidate that the name exists on the service use destroyService as a default when the attribute is not suppliedIf the exists on the service enhance the class in the following ways +I dont understand clicking to the View info link of the Plugins page again I now have creation date information available +Ill produce a new patch +Please attach the error logs you had errors shown at wizard start and at the end +r for the Solr changes +Suggested patch for this issue +Similar sql sentences will be run correct on DB Oracle +iBATIS has been pushed to iBiblio for Maven +Thats fun the first times but Im ready for a better solution +Wonder why doesnt work with mvn tomcatrun +Im sure they are interested in fixing any you have a pointer to any more details about the nature of the problem or do you know someone who I can contact to get that info +Thanks! Ive got a ways to go on this one but will start with small steps +scripts to recreate the issue +The file contains a third proposal for updating an in an existing configuration file +This will be fixed very soon I am in the process of adding in more checks so the superclass is not mirrored and removing the check that stops Roo execution when a annotation is detected +For example the deprecated assembly plugin is marked aggregator while the single mojo is not + +You can take a look at the patchAlso I think the lifecycle changes should happen on a separate branch which should be reviewed and later merged back into trunk +However the conversion tofrom gets expensive in the sense that during the sort Bigram is performing multiple decodes per pair that is compared +New patch moves construction and adds back schema mismatch check +If the associated scope is complied before partner links are resolved then it is possible to push the associated scope into the structure stack before resolving. +In the table above there are a few obvious cells and a lot that are unclear +The problem is that the has a and it closes the entity manager +patch for the you for this patch +Thank you! +In case anybody needs a short workaround for this problem +We are testing it but its working as expected +Or I missing something else +Patch for concatws w unittest +I was so concerned have switched from to and am currently getting a new baseline using the larger instance types presumably with better io characteristics with and with false +No tests because its a change in the build only +Different expected exception was happen in negative test in blobclobBLOB +I plan to take a look at this issue once Ive got access to a PPC machine with a fully workingFeel free to work on this issue if you get to it before I do. +dims +This could be very useful to anyone debugging their application that uses the web UI SE looking for patterns or issues with particular type of application tools using the webservices to aggregate info and create their own useful experiences etc +Throughput slightly increased over writes +Since the only way it can happen is by an abort and its possible that an abort may also cause an it didnt make much sense to distinguish the exceptions +Thanks Rushin and welcome to Chukwa! +Although Ive posted the patch for maybe its better not to apply it +Opened for restrictRange event. +reopening for editing Fix Fix Version +I cannot figure out how to make it work Ive spent way too much time trying +Should be against component +Fixed in rNow serialization of Java arrays preserves isomorphism +Thanks for your comments Sid +Using jbossjta as JTA transaction test caseRead to run it. +Web UI doent update superstep count again +Im seeing this some myself +Verified on with JBDS +I dont think theres any need to limit the maximum value for this setting +Once the poll thread decides the server is up it sets the server state to started +This is a known problem +Jonathanbq +Its currently possible to create an event without defining the category which leads to an unexpected behavior those events are not displayed in the Events widget +Probably a good idea to put this on the Embedded options in the owning entity not in the Embeddable annotations +addendum committed +This directory has been deleted from the repository. +But this is not a an issue with not being present +And separately +Thanks +Grant the code thats been committed has references to Lucene version +Partially done in a series of commits on libraries are built but we dont link against them +Lets try it out +Thank you. +I will add an entry in documenting how to turn off this feature +This has been complete in the branch. +Duplicate of close of resolved issues. +Ive fixed this problem and Ive verified mvn clean install and the test on a local cluster +Is there anything else you need from me? Do you want a PMC member to comment on this JIRA? +Cool! Sounds like what I try to do here +i sent you a link in email datastax +I suggest to split the patch in a refactoring and in an actual patch +Idle connection cleanup did not properly account for closed connections +Im also a big fan of your architecture but as Brent we also need the c web code +Integration tests were updated +Im also curious about your thoughts on the possibility of getting this included in the trunk +Uploading a patch for this simplification running regressions +Should be able to close out as well now as this patch fixed that issue as well +old bugs +So all in all according to Brians spreadsheet we wont be allowing expressions for and connector as for load metric type wont either +It looks like there is a tiny window in which its possible that a write check task could be fired after the static has been shutdown because all connections closed and the class attempts to close down that service until its needed again +Part of the issue is that the individual tests cant run independently +OK I think the branch is ready to goThe one thing I dont like is that the gets included twice once in the modules lib directory and again in the solr lib directory +Hooray +I believe r fixed this issue but I still need to verify +Seems to be disabled here +I would vote for fixing this in CP if we dont the user will have to edittoolsconsole they can run the DB set up scripts for the management console +Theres not point in working on this issue until Remoting +When specifyVlantrue check box is selected send specifytrue only when no is selected +This is a stacktrace from a simplified pipeline including LDAP transformer +Which behaviour is that? Server continues to accept connections even when its call queue is full +Just got around to actually reading your suggestionWhy would you remove warnings for invalid levels? I think it would be good to notify the user that the level they entered is invalid and that the value is being ignored +Nice one RenatoThe change as per your suggestion has been committed +One thing I noticed that those exceptions are also throws in the very beginning while I was setting up a SP Repository Connection +If it does work with invalidated key cache or with caches turned off that means it could be a concurrency problem in cache implementation +Please reopen if theres an issue. +Feel free to create subsequent issue for ducmentation and being able to pass the file name as a ve created issue to stop hardcoding this report file name. +Revised patch also changing just committed this. +in Lucene s I shall check that too +Great job with the space! Will commit shortly +I dont see this as a backwards compatibility issue +Im adding the above regression test so we can ensure that it keeps working +Thanks issue is fixed and released as part of release +Thanks for your work +Looks good to me. +I meant binhadoop jarHong I agree that the documentation should reflect paths rather than paths as the documentationis for of Hadoop not Hadoop developersI favour explicitly describing the parameters from in the documentation +Can you elaborate on this cant really see what the issue is as Messages without an expiration are not expected to be removed until they are consumed + +I have cleaned up the example a little more and replaced some println statements with logj ones +Made relevant instance fields volatile or final in r. +Fixed. +The user has no way to choose what arch is installed +Thanks for reporting +This by itself is not a bug +These patches essentally rename newi goals and allow the includes and excludes defined in etc files to work +core tests + to trunk +You need to use the code above on every substring in the convertMulitline method +Attached patch a which removes the code using incorrect synchronization +applied +In there is a folder created called nullzookeeper from what testcase does that come fromI think it should be from the I shall look into it +Now I see the problem is not that easy to solve +And as such it shouldnt emit the Z in date output either +Unsure if this is still needed but heres a rebase +Its running near +Sometimes Eclipse chokes on this especially after its been done a number of times +Here is a POC for camel integration I made a few months back +Yes exactly +have to set the internalname also +The original problem is believed to be fixed as well as all problems related to ast transforms breaking content assist and other UI features +For this particular issue we dont need to change interface +That was a simple but nasty bug +Think yours was fine it was mine there was a problem with but it seems yours were removed by accident +Its done enough to close this one now. +Contains several examples of the UFDF in different fonts +patch committed +Carlos can you confirm that you original checked in file was UTF and didnt use the encoding as shown in the patch +patch +Hope can be ignited by a spark of encouragement +I intend to commit this in a day or two modulo review comments +Version has been releasedClosing this issue. +Moving to the next cycle for reevaluation +has been implemented and the issue can be closed. +This is a patch for +Requested in +Release Notes marked Not Required as this issue relates to Red Hat internal testing +Probably by the end of next week +Working fine now. +I have found a good solution for this issue that passes all tests done +In the meantime you can update manually by downloading the jar from the h database website +Perhaps this can be drafted first on the wiki? Who will write this +plan to fix for to be fixed in should avoid making major changes to the bridge already contains functionality that allowa reconnection if destinations arent available on startup see etc +You probably got a core or +It sounds like you should use typematerializedclob instead of typetext +Prashant do you want to take the honor +We still need to tweak the scripts +Ive fixed the issue and added a test that triggered the exceptionFixed in trunk in revision Fixed in x branch in revision +Closing as could not reproduce +Great thanks +All currently enabled tests in the spec fail should empty array and block yield problems there are separate bugs and separate fixes +I confirm that is not used in +sigh +I am having problems that I need some advice here +onsourceclickontargetclick onblur and onfocus should be IMHO rendered in DOM on the same place as +Patch resolving the issue +It should operate solely by the server instance created by Servers view provided by the does + for the patch +Hi Justin thats great! Sorry for my late reaction Im going to check it out now +Ported the aggregator component to a leveldb aggregrator issue +I hope the will be released soonThanks guys for your work +im using a Set instead of a vector tooDamn +Especially since that leaves more iops to keep the rest of the system healthy and not go down ourselves +Im not at work today so I dont know the exact version but were using some version of right now +Made appropriate modifications so that the Blur shell does not fail in the presence of the new APIAny clients that need this data can use this API by doing the necessary changes on the client sideRegards will close it +I also whould like to know what the commiters think about to integrate it in commit to trunk +I am unassigning from this issue since Weldon asked to allow him to care about threading bugs +I am using Windows XP without Cygwin and I have both Sun JDK and and I tried with both Dframework and Dframework and couldnt get to fail for meMight it be cygwin that is the critical factor? If so how might cygwin be affecting the test harness +Could this be done as a set of chained bindings? So the page has a property foo bound to the component as value parameter +Need to regenerate the patch to reflect latest commits +Merged to the branch in revision . +Is the test change necessary +The issue you are seeing probably have been fixed alreadyIf you still having problem after upgrading please post the herePlease note that JMS sessions should not be used concurrently +The criteria code is pretty complex and this was my first use of Hibernate and being in the middle of a project I only had time enough to spend solving my problem +can leave bad rows in statistics catalog would be good to have a way to manually get rid of this bad row +Hi Aaron the reason why i did not decode is here there is no need of encoding then why to encode and once again why to decode +Please open a NEW ISSUE if you find yet another regression. +Matei On our roadmap +Missed that test comment about no need for random directoryAbout the decRef dup code yeah thats what I meant but okayI think this is ready to commit. +However we need to determine how we can allow users to add more drives for data directories when more disks get added to the nodesIf we support reduction in the data directories then we need to add support for decommissioning a datanode changing the disks and adding it back. +Patch applied thanks Kristen! +Im still working on my slides +I put a reply on +Youre probably right Ill prepare a mail for the team +Ill test +Thanks for the review and comments Suresh and Colin! Update the patch to address Sureshs comments +We could temporarily or even as alternative consider not using anywhere anymore but rather always throw subclasses of Springs own +I agree but my lack of knowledge of regexes was the reason to opt for my approach +Makes sense +AwesomeSent from my iPhone +pretty sure this is already done review +OK I think this patch fixes everythingThe persistence problems were eventually due to the fact that sticks a slash on the end of instancedir but the persistence logic pulls the information from the Config object not the +Calling terminate on the runtime would then scrub the pool closing selectors +please see my comments on has been fixed in +Sorry of course we can change this +Moving out of +Sorry +It was all my +Verified in +You now get compilation errors as there are two copies of the source file one in the and one in the generated files directory +So it seems that the problem isresolved but we still want to do more testing +Currently the last modification time of a bundle is not persisted as for example the start level + +I dont think the email came through. +It is not the type of issues that should be resolved any later +And we send message into job queue +Block replication and block creation should be different block creation allows partial block but block replication should be atomic either replicate the entire block or do nothing +generate segment +As mentioned in the send operation will not work with Fuse ESB or Fuse MQ as they both secure the Broker +Hi AleksThanks for the confirmationDoes it fix both renaming issuesRegarding the jump back and forth between the entry in the search and the entry in the DIT its probably an issue due too many update UI events being sentI guess the behavior is exactly the same whether youre using the LDAP API or JNDI rightI will have a look at that too +Fixed with rev . +Correct theres no point in maintaining two copies of the codebase the old releases will still be available but further development will be done on the OBR goals in the bundleplugin codebase +Added wiki documentation as well +found while converting for +At some point the master will be a version that expects the locks to be honored +OK it does work I tried applying and compiling +The patch looks good but the recent added some lines to the output from the metadata test +any suggestions Simon +Please ask questions like this over at . +bq +Thanks +I cannot accept copyrighted code +Patch checked in +Updated documentation to better explain what the pmd parameter change submitted in r. +Ok +However while testing the argument parsing I found a bug I suspect I inserted into the patch a few versions ago + seems to be application bug +The issue seems to be caused by command forwarding logic while state transfer is ongoing +As this represents a breaking change care should be taken in applications that werent guarding against null being injected when there is no Twitter authorization. +So using recent snapshots you shouldnt see the issue anymore +in order for it to take you back to the last click we would need ajax history support which we do not yet have +Good catch Hari +if there is a partition of GB is reserved and DFS occupies GB GB is available then remaining should be GB +The file is not syncd and the client is never told that the nd set of mutations were saved +At this point I actually upgrade hdfs to and hadoop to +But upping it would not affect your situationThat thing with the client holding of the HRS its very weird +Committed at revision +The HTTP content length in a compressed scenario is generally not considered valid or at least shouldnt be trusted +You are probably right +Instead of using the token following the return expression to mark the end it now uses the last token of the expression +Only the native threads seem to growWhen googling the problem that exception unable to create new native thread seems not uncommon in +Attached more tests +Tsuyoshi the changes in this patch are done to enable parallel execution in only +As part of this should be deprecated in trunk in favour of or an additional note in pointing people to +Please address the hudson failures +I will try it with sane build +Patch addressing review comments +note however that the http service is not used. +Thanks +Duplicates that is targeted for heads up this may end up similar to XML processing. +The number of possible Shapes is just too massive to ever reveal anythingWhat I feel would be better is if we defined Shapes that test particularly troublesome areas +Lets run it by jenkins to see how it does +Also this could happen only if the retry attempt is infinite + +Fixed in commit on bugs marked as resolved before and tagged as version as fixed in RC +bq +rsync supports automatic login using password with the FILE argument that is why I thought it would work +adding jars to solrhomelib must work The request parameter can add optimizefalse to disable last patch wasnt generated correctly +OK I reverted that part of the patch +Any chance this patch will get applied and included in the release +What did the spec look like +This works for now since there is no differentiation in the existing MR runtime +basically need to move srctestjavaorgapachejackrabbitocmrepositorytosrcmainjavaorgapachejackrabbitocmrepository +Currently it might be possible for namenode to ask to invalidate a block twice +applied to trunk as dadfbaeabdbaaaef +The findbugs failure is odd +Thanks for the reviews Nathan +Its leaving artifacts around +Committed to trunk. +you have the correct host name and port number for the Agent ControllerAny ideas on this message? Kind regards its not my expect! To be honest Im not familiar with profiling tool so cant give you any for thatMaybe both of us have to focus on the TPTP project and report problem we metThanks +Committing a fix +Thanks Ill take care of the quickstarts in the jdf repository +Applied Geoffs final patch. +Im not satisfied with the number of unannotated tests and have opened a thread on dev to get some feedback +ok I downloaded the new core deps and built this branchIt appears to be running the number of Active docs quickly shot up to I can only assume those are the users foundI did notice that the status still says Starting Up even though there are active documentsOnce the Active column got above did the status change to RunningIll post again when I see documents in my personal WS in the index +project files have now been deleted +I was a little suspicious when you used the word trunk +node eventually went red while node stayed I dont see this behavior +The web services still dont seem to be communicating though +This looks like sorry I have not seen itThey are the same issue I will watch please close this +Here is a patch I used to simplify the actions and can reproduce the problem +Reopencomment if incorrect. +QE verifies. +Hi ErdemYou can just reuse your own custom Form control until this is supported in Click proper +See open a new issue if you want for integrating it in the lifecycle +orgapacherolleruicoretagsmenu to be fixed in +Applied in +Also committed to. +I still need to pull this up to the release branch +Offline Ive requested the test code that is breaking because perhaps its the test doing something unusual +if we linked the libraries statically we would have had all the dlls containing the mxvc libs + plugin to use a default revision in case scm is not configured or not available +I just committed this +Then H will look like a normal headingPreformatted text have to look the same as it was in the source so the only way to fix is to not indent the preformatted text in XML sources +I think the tests for circular dependency behavior show DS is doing all that is plausible to establish circular dependencies +Im going to assign to Stefan to vetcommit to ensure my first foray into the new UI is on track +Hmmm it works for me sudo optatsbintrafficline r Are you sure you are running both trafficserver and trafficmanager on this box +So I tend to move this to or because of missing support in regex contrib +Id be wary of adding it in srcmain +the patch to for should be the end of this bug +double entries on code assist and incorrect internal for auto indent on pasting committed +We would prefer to start with an empty repository and import the current code fresh without all its history +AndreasPlease check out the latest camel trunk and verify the fix +bq +Will commit when tests pass +Thanks Jakob +application +Patch breaks build as seen in reverting this results in successful build again +Hope this is not too much trouble within your marshaller and does not cause a performance going to close the issue but if something related comes up can either or create a new +Apply patch from heremodulescore +slip +I cant finish +The limitation is Indexed items can be no larger than bytesCouldnt we in addition to the path store a hash of the path and index that instead of the path itself +is not officially released as a jar in Maven Central +I can see what the UI does to report request per interval +Committed to trunk as rCommitted to as rMikeDavid please reopen if you want to apply to other branches. +Committed revision . +But you agree though that the test itself is flawed correct? Im not super familiar with this test and wanted to verify I was on the right track before creating the patch +Attachment has been added with description Sample test case that should work +Usage documentation will be an integrated part of the getting started guide +See the file from or any project where any java file would not compile +Ive attached my revisions in a zip file +Brent Ive downloaded the db extra you shipped with and this issue seems to be fixed can you checkMoreover where are these +Since the combination of coprocessors on a region server is limited I was suggesting that the report of uneven coprocessor presence be embedded in +Removing Fix Version. +Verified + +Logged In YES useridI have updated the resource loading unit test to includeexactly the configuration you indicate and I have no problemloading the dtd resource with +Tim do your use cases for limit involve sort more often than not +This tracing seems very useful too can you include it as part of your patch +Moving to. +Ive moved the issue for RC +is this still an issue +When the online snapshot branch was ported from git to svn this mistake on merge was not made so this patch was not needed +The feature is already present! +thx for reporting +Hence you can build Axis as usualPlease note that once you push this change you need to move to Maven +Is it committable +It didnt run the first time +lets not let this hold up the release +This issue has a workaround in place some more work is required though. +Would you accept a patch thatadds dead code to a projectCurrently the jar command puts the as the last file in thejar +Templates and other files and directories of the website currently online have been moved into directory legacy as a convenience temporary storage while the transition completes after which we consider definitive need to know where in svn you want us to base your sites checkout onThat will become part of our svnpubsub config and will make commits to thatarea appear on the live site nearly instantaneously +apparently changes in and logj confs are junk +Confused as to the remaining open actions on this JIRA +Please keep up your quality submissions +Seems that we can close the issue +Patch applied. +assigning to since theres consensus we should do something here +Thanks Gabriel updated the patch based on your feedback and I added an integration test that fails under the old Sort lib but works under the new one +If paymentid is mandatory then the click on update should display a message saying that paymentid is mandatoryLooking forward to seeing this completion of reconciliation task in the next weekthanks +However I am just curious to know if there is a way to reproduce this +It also has details on how a call will handle data that is smaller than a crc chunk +not fixed in t fix this in because the jaxb limitation +Patch committed +As I already mention on the mailing list please attach test projects to that would allow us to reproduce issue +the last change it includes was before the actual import of the codeThe trunk was created by first copying from incubatornmavenbranches at the point were the project was forked from then importing revisions after thatCloning the repo shows it still at that copy point with nothing since +to get Gregs input on whats left to be done this to GA +Closed with released. +it is working properly. +James got to it first +Set the proper target +Okay the algorithms im using in our version are not actually the same as their old algorithm +SlinglinksSlinks +I didnt get far enough along to figure out how large an impact either would have but that kind of IO variability seemed like a bad thing to mix in with repair +Ceki I thought you said this is what did not happen +captured a thread dump from the hung fix in trunk to prevent the deadlock when the advisory consumer subscription was getting set as sync dispatch +Thanks Kevin for the review +It was released with the fix in +So I think killing the entire process tree did not work +Version has been releasedMoving all related issues to the next version +Shouldnt this be closed now since the Exception code has already been introduced in? I think so +I should probably file a jiraAlso and should not be marked final in typical casesconsider removing google mapreduce paper link as prerequisite since the goal of the tutorial is to provide all the information needed to understand A picture would help in the overviewIn the Input and Output section remove the use of combinerIn the wordcount example simplify it even more by avoiding the use of submission amp submission and finalizationwherever overriding is mentioned also metion the default value +Attached a patch on trunk Adds Javadocs per the comments here and my understanding is a simple unit thanks Shalin! I changed UNTOKENIZED NOTANALYZED in the javadocand switched to in the test +This is the whole patched file from ojb +is the EAP bug report +Doesnt render this patch useless? +My guess is that the next release of Avro will be run more on JDK than on JDK since the free version of JDK is no longer being updatedWhat do others think +broken link fixed +Added to the Reference Guide a placeholder for a chapter about development tools and development environment +Please let me know if you want it to be polished some more +Hello BrandonThanks a lot for responding to the ticket +that would after all be a much cheaper fix for the problemsince we will have to address the childnodeentries loading sooner or later i wanted to state it heredo you have a patch for the straight forward fix at hand? that would be coolangela +Interpreted mode for this giant method isnt great but it wont affect the behavior of the application +I verified that theincludedocstrue endpoint changes its when a document is updated +I confirm the problem is related to grails +This fits with the deployer interceptor chainfilter notion we want to add +Im fixing the other one using this defect. +showing what went wrong would take a bookand that is where I suggest such information is handlelike when you use the script to give admin log init is not associated with a partyID so the myportal does not workthat is why the suggestion to modify the demo data before installation is recommended +I assume that the field to highlight resides in the to document? Please provide more detailed information on how to reproduce including all parameters you pass on the URL your schema and example from and to documentsIm not sure whether highlighting on to docs is implemented or even supposed to work yet +I got the prio wrong sorry its not so important +Would you mind posting an updated patch which moves these to the appropriate file? Thanks a lot +Andreas correction to corrects this +Can you provide more information about this project pleaseThanks +Francois Masurel has verified that the fix has solved the issue +With Farah to will not be a moving to for m no longer seeing this error +One should think that today in the ages of VCS those tools should solve that situationsI think my main concern is the badmisleading error message +The problem on Maven Jar plugin is still existing +Is it possible to get this David made changes to the debianrules and changelog files which seems to have fixed this +Patch attached +Im not addicted to GAE actually +Thanks KnutCommitted patch to documentation trunk at revision +The system is probably as well +I would use Maven for the entire build but it is too inflexible +Guess we need a doc or ref to define what valid char values really are +done +Thanks Devaraj and Forrest. +I think the first part of the patch tries to handle c but is not quite right now that I think more about it because the if test will also succeed if the latch has been removed and mayCreate +Ive also added a unit test for that scenario. +This should now avoid all possible NPE problems even with complex builds +Yes I look at the diff and I see the modifications to do but I dont know where and how can I modify the phonegap source +I second this we see exactly the same behaviour +Im not sure why that would be but I definitely fixed some things in the other two testsAlso this patch only includes the unit tests so you will need to include patch as well +You shouldnt post your user question on the jira please post your question on CXF mailling listThanksWhat I am trying to say is this is a problem of CXF configuration +This is an expert API +Do we want this even when people have set the ignoreXXX option +Yes the commits from are there +We could possibly add in a +Hi ErwanLooks like your Shift button is still broken I know +Patch applied with small modifications +Requesting approval for this sensitive info collection hereI think your suggestion is spot on the info im missing is though if any class will be missed by removing the above jars? Probably none but would like to be sureAlso wondering how this will affect since there arent one source for this client jar is there + for + +Shazron thanks for your quick reply +Hi MatthiewIve committed a fix for this +First I need to deploy additional jars if building my own webapps with cocoon and that would require me to rebuild my own applications with a special maninfest as well +shipped in +I cant find the superclass its not in any of the hibernate jars Im using and I dont have a clue where to download also cant find any javadoc or other documentation on it or on writing and running hibernate test cases in cant expect users to provide test cases for the bugs they submit if there is not at least some basic documentation available +Hi ScottThis is very probably a duplicate of +The lib is used instead which is a super set of. +If I run a test individually it does have all of the requested logging +I committed the patch +IMHO it may make sense to have it in java +x +I think the only detail that might make performance worse is how well the inode map performs as the bucket chains get longer +Verified +Note that the records would go in the edits log which isnt compresThats true! I am still ok with putting the txid on every record + + All of this can be replaced by just HADOOPHOME env variable which is already used by hadoop +I attach a very succinct Eclipse project highlighting the may have to be updated to suit your Eclipse environment +I guess a probe is in order to discover which metaclass is actually the delegate of +As you cant see the whitespace in the comments above Ive attached a screenshot +Thats a valid point indeed +I think James forgot to put this in resolved state but I see his commnets meas this is resolved +It sure hangs because the protocol level from client is accepted by the server but for shutdown existing server expects username and passwordWe may have to revert to another approach to this problem which is to send shutdown command with protocol level whether there is username and password or not +Its roughly equivelant to Version or Author in svncvs speak + +It seems to be immutable +Note We havent had any other performance issue reports with jsf includes etc +Ill take a look at them later today +m open to suggestions on picking the interval between scans +Looks good Mike +Were mixing two things here +yep all of this has been committed +This patch cleans up another mutability problem copying an array when a is created +I supposed i could tweak the css to my liking right +Its not critical to get in this month I think +Doesnt look like anyone is exposing them for monitoring in time for. +Using the transformer in the simplest pipeline +Decisions made in private communication are against the Apache Way +The next step is to make the tests pass on all these environments +It comes with basic matrix and vector operations that should make implementing the algorithm a lot easier +Without primary keys there is no foreign keys so the FOREIGN KEY statment always return an error +This page is the one users initially log into +and the user one +pppttt +a new page? What this new page should be named etc +I made this change throughout the code with revision +Dtestcmppassivation testin the the testsuite directory +I was using the term install loosely not so much the way a typical package has a GUI wizard and lots of stuff going on but more in the sense of raw Solr where you download unzip and files are in sub directories right where they need to be +For now I suggest to have one jar and all dependencies cleanly listed or distributed with itSure we should have as less as possible dependencies +The part will be making sure that this works for +The warning on line is caused by the cadvise bug described in +Will appreciate feedback on this patchSpecifically should I wait for START event or is constructor fine +Fixed in rev dont resolve issues without having the fixfor field set +The fact that you didnt have to change any code that reads the logs and snapshots indicates that we are not checking the magic numbers which is another bug Please commit +And they should be getting cleaned up but there is obviously something wrong +I have been write dummy stub +It will take me a few days to set something up since we are development +fixed typo in summary +in the file has all the alert labelsokLabelyesLabelnoLabelcancelLabelAnd its already translated in all langages so why not use it instead of a the new bundle +Unfortunately there is no meaningful error message +lets expose in nodetool and call this good +Note that it require directive to be Full +This removes the option to pass by value or by reference that was available in old api +The attached class loader seems to work ok now +Which repobranch do I need to look in to find this code +Thanks Knut and Rick for the reviews +A GWT application requires entities to run and the entities werent being proxied due to the reversed model not having a version field +The key is defined with the definition of the implicit map and not with the type itself +Updating Fix Versions as code has been checked into and trunk +That can lead to former running files to fail now +I will give it a shot right now +Current implementation sets two headersServer DateThis should be sufficient for now but later on we probably want to set and comment about in responseClosed by bcceabbbacefdacaebda +Being I couldnt create a branch in ASF git repo hence I was working earlier on fork in github +Good catch I think +Did not take a look in great detail but some high level comments +can you please verify that it works and close the issue +The issues are now closed. +NET or installed a windows update such that this issue was fixed by that? Does the issue come back if you downgrade to +But do we need so complex support like hadoop which use and to provide comprehensive version info and a lot of API +Considered the previous use of structural objectclasses and extentsibleObjects as preventing users of graphical browsers from working with Config Admin +There might be work to be done to improve this modelAs it is today annotations are a type safe model quite close to reality +Overriding the method and detaching the request targets by hand is not possible because requestTargets is private +JNDI services are not always tied to the lifecycle of the underlying component +We cannot use hibernate with mysql with our mappings and hsql +Were talking the singularity and thinking if we are going to switch this is the time to do it +So we are now intitilizing the UDF and passing some context +an octave version of the same problemOctave uses glpk which does not find a solution either +Thanks for testing Sushanth +Tx. +Wouldnt a more appropriate fix be to simply change your repository configuration to include the javacompenv prefix +tested +The one under the jdkjrelib is not sufficient and usually the first one to be found +patch commited in svn +That method should be renamed +If it is the ordering issue well be able to suggest how to reorder your pom so the behavior is the same +Please review and +Improvments to add display the checkout queue add checkboxes to remove entries and not only one add htm auto the project that is rev have created a separate issue concerning xml rpc. +changed for Hairongs requirement +It would be really great if those tutorials would be updated to evaluate Clerezza +resubmitting the file granting a license for inclusion this looks great +Sure Java is out of date but if any framework should be conservative regarding JVM versions it should be logging +can you please upload a diffthanks do you need the wrappers around andthanksIve Heavily tweaked the code that you sent and checked it in +trunk +Will fix this +So Ill assert that its more than a mere documentation issue and would very much appreciate it if you would give further consideration for the addition of an optional strict mode +Will create a new jira for removing the actual dependency on logj +Base UIMA documentation was purposely left off perhaps some of the core documentation must be there +Or did you mean something else +Ravi thanks for taking this up +Another similar question about applicability to different branches +I think we should commit this and close the issue +Its bad at threads but thereafter the patch starts to shine +I attached the xml and xsd files last week +The rest looks good perhaps instead of includeDeletes it should be called keep or something really descriptive like that? Write one read many and all that +Would it be possible to get this into the code base +Thanks +Pushed to master +Could it be done for our project only? I dont want to disturb the other Spring projects +Thanks for the patch with it we brought back the wagon +So the answer is yes it is resolvedGerald JeromeVnet +So I postpone following up on this until then +At the same time lets minimize custom code code that is particular to the addition of this feature only +Hopefully it will work better. +Heidi made lead for the project +I added quite a few more refinements to the chunking approach +either remove the s at variables or change has to the wrapping approach looks okay to me although I havent tested the patch with a Groovlet yet +After we land a lot of the current Cordova will need to updateddeprecated and newer more relevant specs integrated into Cordova +I had no time yet to fix the scripts as these are the cause if it +Latest fix for rollback of breaks normal flow of exchanges +Verified +Im free this weekend and would love to put in the finishing touches +The current patch assumes that the first column is the ugi the second column is the permissions the third column is the command etcetc +If this issue needs to be split into two then please do so +The new only wraps the exceptions from qpidmessaging +Just the proto and for protfile to name request and responses a and good +One more handle styles? see rename the name is completely misleading as it is not a sink and on top contains only apt specific methods it shouldnt be used by anything else maybe put it into apt module +Im not sure you need a separate class cant you s decRefincRefI guess I did not see the simplicity the reader offers blind due to +Thanks Knut for the review +There may be a better way to do this +Submitted patch v adding a unit test that uncovers the problem +I think each case needs to be considered on an individual basis +if its name can be trusted locates the resource for Wicket consumption which is what I have already achieved without itBut it dos not alter the paths of resources that this resource references and that only the browser is interested such as images and style sheets +If you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +implements the JPA spec and should bahave according to it +If the code is changed and the user initialize the class with a buffered stream there would be buffered stream encapsulated +Ive backported the large set of changes that are believed to contain the fix +If JCR team doe not understand the question I guess the question should be removed from their FAQ +But if Seam Facet Install Delegate was invoked by New Project wizard then it should copy all required resources as before +Hopefully you can then the same split logic for both objects +Ill verify that this error still occurs +This patch takes care about the issue and allow a once set property to be changed dynamically +Looks good +If filesystem caching is enabled which is by default the verify uses a cached instance no new connections are being createdExcept idle connections are closed every sec by defaultbq +I see the original bug under the following circumstance Remove the database +The project is focusing on the future versions of and while we will have a concept of maintaining support unless this problem remains in master or branch there isnt really anything the project can do to fix it in something that would be consumable by you +I agree with using standard type +OK that is a good idea +The changes made to the threading have severely impacted thread synchronization +Thanks + +The attached project contains the test case that was originally attached to this issue +I also think it is still important to be able to import log entries to a different table +bq +Not in this case +files when no TEST +I have attached more simple and clear patch +The patch was made against Seam second that feature request I already had two cases for which I require scanning of components for custom metadata +Running regression please review +Ive found that the catalog parameter doesnt really work for the transformgoal of the Maven XML Plugin +Or if you prefer could use in stablebranch job for Beta and in trunk use to pull the latest every time a new upstream build happens +Read the as Resolved instead ReopenedIs this something that can be fixed or is it being pushed out? +This is the last drop wed like to make before working directly in Apache +By the way the reason this doesnt fail always but only for certain codecssome codecs have assertions that get tripped so they fail the testother codecs dont have these asserts so they pass the test and checkindex happens to passbut this is only because checkindex ignores deleted docs in testPostings the index really is corrumpt in those casesattached is a new test demonstrating this for some codecs it triggers an assert for others it makes a corrumpt index +I do not see why you close this issue then as I suggested that it works out of the box what is not the fact with +Fixed in r +Closing issue which has been resolved for some time +But again maybe the people on know more +Shouldnt we loose the timezone information if we convert datetime to longAfter going through the jira for datetime it appears that datetimelong wasnt supported to avoid timezone information loss +Comments are fine by me +Chatting random would have done for our case +Roshan that patch looks very issue has many watchers +Ill investigate more +changing title as different tests in the encryption suite seem to be failing intermittently +But apparently notShould obviously also work for wrappers as well +xlsx file +Style classes and Figure +Those two have only value which doesnt make sense +collapsible template was recreated header and footer were added helper classes were used +If you need me to clarify I can take another stab at it. +Hanswhat does the client gives up +Good idea +fixed in CVS +Hi attaching patch that only displays error message in page and in logs +I will take care for the Workflow Controller next + +A testcase wizard generated plan +Yeah I would not be opposed to that as an alternative +So yes this fix will be in the release when its GA. +Mike did you commit the jar tooWoops forgot and now fixed thanks for catching that +Thanks Senthil +Of course now we have the opposite problem as reported in +The patch also contains few updates to dependencies in the examples mvn module required for running the testsI am still examining the rest of mr test classes and will be updating the patch +I just tried it out +Version where mediated methods are introspected from container using +Well as it does not hurt to be a bit more aggressive from time to time lets mark this as INVALID +Do you have hard numbers for that claim? Are these checks used on a hot path? Just curious +This is almost certainly a bug in arg processing for the batch files but none of us are experts on bat logic +class file is not in a jar already +Committed to trunk and branchxThanks David +Thanks again +FWIW Id simply make the repository writable and let the Blur team push in the rest of the code once the IP is cleared +the METADATA pull was merged so now need a new release of it to create more pulls to integrate it on AS with an integration test specific for Jaikiran can you help with this +Hopefully the port wont be too painful +Ismael thanks for your understanding +So I commented out those tests until they get didnt know how to fix it sorry +has been marked as a duplicate of this bug +This has several benefits +Triaged for +I understand I could change the CHECK to FORCE but it wont actually change much in real cases +The Avro version of the patch is atI read this sorry if I missed answers to my questions thereSo by adding these classes the Avro jar one could run Avro through existing clusters +Will update more as we go alongbq +After looking at the documentation I guess this is the expected behavior +Should be backported to +The relevant code that needs to be removed starts on line of +US to donate a newhard disk to Intalio if required +I still have to address the other comments some cleanup and +Its up to the user to avoid complexities in the finalize target processing +I couldnt find your change in one of the patches for this issue or +The patch that I have just provided also shows get Phils patch resolves the issue +Will commit if tests pass + pending to trunk and +a situation in which you were really counting on distinct Environment instances between parent and child contexts? Thanks! +All of the subtasks have been fixed +I think majority of the failed tests are due to result diff +Is this just a change issue +The former always had the facilityId whereas the latter had a null facilityId +The docs need to say they are the correct version +I would really appreciate to see the new version in the repository since we are waiting for the better calculation of regards Chris is now available at maven central repo +Moving out of +Since NTLM authentication does not use the concept of realms Id previously just used the default credentials +Cool! Would be great to have this in. +How about I remove these as well and file a separate jira for porting Gridmix to MR +David thanks for getting the repos done +Joes does we still need to provide a destination for commit mails as well as commit mail subject line patternBTW we could go with the defaults currently settled otherwise let me know and Ill provide to you +Marking for well that was me +Im happy to announce that we have secured a time slot and dedicated space during Hadoop Summit NA dedicated to forward looking Hadoop security design collaboration +Fixed patch there was an error in the prior +Will these need to be available in order to compilebuild +Patch by Phill PorterSending srcmainjavaorgcodehausmojoexec file data +also theres a config class that as it turns out wont be needed until the policy is implemented +tika cli app has option z that extracts all embedded files to current as Not A Problem as this is handled by supplying a recursing parser on the +I read behind the lines that its vain trying to change that +Well this manages to solve the issue I was experiencing +Im trying to do the full process here now +MichaelI tested the PEAR plugin with Redhat Enterprise Linux Eclipse It works for mePlease make sure that you have added UIMA Nature to the project that you would like to PEAR + +Good catch Milind +Many thanks for contributing the patch AdreaOleg +Wow what a dreadful flaw! I will apply your fix +AFAICT is not in libendorsed +Merged changes to branch as well +This will be easier to do once makes it in +Just make sure that classes from console are still workingA proposed patch for this issue for the coreColmA proposed patch for this issue for the consoleColm +Perhaps we should start to be picky about it and force the latest +a I dont consider this a bug rather a new feature or at minimum an improvementb While the idea sounds really nice it doesnt sound like KISS at all +In addition initally enabled all test except test and test for sapdb +Wed like to see this committed to trunk soon but either way were going to be running this code indefinitely whether or not it ever makes it into an open source release +Committed +is in +So I think Ralf can close this issue after rectification +Ive changed text comparison to use first byte instead of equalsand the boolean text is written as true or false +The stdlib files are instead installed under the root of the jar which would be the following entryfileUsersheadiusprojectsjrubylibAs a result the search is failing to find rdocgenerators as a directory because it doesnt actually exist at any of the paths rdoc m going to experiment with the structure and load path a bit +Revision Author mfriedenhagenDate Message Find defining parent recursively include given message New rule to check that a property in a project diverges from one given in an ancestor project +See +This will hopefully enable the creation of a generic function that can be used to integrate with Drools and allow the pass thru of data without having a new function method to support every data typenumber of column combinations +The +It looks like get already does the sampling but falls back to midpoint when there are less than keys sampled +results contain duplicates Hibernate passes them on +Affects the module +Added shell script to accept parameter for loading sequence file to database manually +Have you tested it on any real views +Hi Andyyes as mentioned under affected versions this happens in and +At this point no one has picked up recently and kept it up to date +If it is working for you at some level then hopefully you can keep going like that until they decide to push this into the project at some point +No its a problem with your patch because does not OOM as written. +Bug has been marked as a duplicate of this bug +This jira will be closed when associated jira is fixed +I guess the best we can do is to strip code that exceeds the USHORT values allowed by the geotiff spec and treat them as custom parametersIn this case we would not get back the CRS with a proper EPSG code but rather as a custom geotiff crs +Were still buying equipment at this point so by the time the OS is set up and Oracle and SDE are installed and were ready to load the data which will be the decision point of BLOB or STGeometry probably months from now +If no one chimes in to the contrary I will commit this patch tomorrow evening +Added transferring cookies from the response to a subsequent request in case of redirectsauth challenges +Situation when theres only one run but its direction differs from the base now handled correctly +the wildcard character +Enabled standard UI for adding sources to java libraries +Just curious why is seconds ok? Is this based on some expected server behavior +Any feedbacks are welcomeThanksBo +Would this create an ASF digital signing certificate for IOS applications? Who would hold this certificate securely? Does this program expand to the full Mac App Store? If so then I know would be interested +Sure I can +I or Karthick should have a patch coming up soon +Roo doesnt offer support for and but it should be straightforward toYikes! I was adding additional fields by hand and assumed that it was OK +I can add another fallback which tries to pickup elements with the same name as the original one if the name is set this hopefully fixes the problem to some degree but not in of all cases of multiple forms +Ive been studying the my performance logs in normal operation over the last few days +Sorry if I sounded like that wasnt the caseHowever I still need a way for my tables to look good when they contain no records +Conclusion to successfully use EJB module within any EAR project in any CDI WAR module in the same EAR project is required to add CDI depedendency to that EJB module and also add EJB module to the classpath of the CDI project +Sorry +Not by me if RAD users want it they should come and contribute it +Closed due to the release of Tiles. +Is there a reason you cant create a as I showed above? That would allow you to create a +For some reason for deprecation is not working for +point is not met due to table and column renaming think it would be best to have separate injected into but to keep interface compatibility I extracted the functionality into and made and delegate to it under the covers +I looked at the code more closely and rollback is not requiredThere are couple of choices on how this can be done If editlog file size then treat it as there are no edits +Great Thanks Andrew +Can we explore the possibility of using a rd party lib like ehcache for cache imlpementation +Could it be related to +It solves this problem but some unit tests failedNeed more investigation +merged to trunk +Bulk move of issues into after vote that we merge branch into TRUNK up on list +We dont have any fixed date for the release +I suspect that the exception is thrown by the same cause +Transmission has been abortedSeems the serializer expects a length namespace +CorrectOh and your kvs were bytes +The problem is more profound than that +I wouldnt worry about this WARN message in particularDo you get the little quare box that shows the stats at the end of a scan for the internal repository in the logs +I feel this bug report should not be considered to be invalidThe creates a thread to perform the transformation while the main thread receives SAX events that it uses to build the DTM +Im going to patch that one as well +attaching a new patch which relies on block recovery to always return a newly generated access token so that there is no need for adding a new method to to get access token +After all some updates are minor enough that they dont even warrant announcements +Closing old issues +RESOLVEDThe project has to be in the workspace to run the mentioned tests without failuresAnyway the problem with conflicting templates still exists +I like that idea but is that doablefeasible Yahorif not push to later guess this is doableAnd there are no much difference between horizontal and vertical splitting so we could implement this feature for both of them +Reproducer is attached +Well commit them to if will be fixed before release +Rebased from branch after is checked in +But do feel free to discard that opinion if its just me being french and if composer sounds perfectly fine to you guys +Please try with the next build and if there is any issue attach the TCP mon any any other information to understand the information are properly displayed now +Ill try a work around like you suggested Ill let you know the outcome. +running in without the order bys did not repro the problem +defer remaining issues to +We need to address this in development drive +the latest patch looks good to me +Does that clear your concerns +This helps in reduction in unnecessary duplication of code more importantly dont need to know about internal Tez code +Manually tested +okay now it all works properly +Unless something was disabled in the tests this looks like the issue could be resolved +As discussed Dinesh thanks to increase the default severity level to Blocker +As I already have the same code applied in my eclipse I just generated again the patch with addressing the above commentThanks a lot It look like I just generated it before the commit of +Fix was added to CVS by tng on +Any news on this? Im keen to hear if its sorted your problem! Thanks +Removed setter for name use set instead +Just discovered that I can no longer commit to SVN as well +I have problems with the above code +OK makes sense +Thanks Sumit and Anil for discussion on it +Removed unnecessary stuff +Is there anything else we need to do for this? Id be happy to help if needed +resource prefix is mandatory +From now on itll happen automaticly +As of we have a new timeout impl that normally will never spin up a thread +nico all you can contact Anilaprodduthe is the Redhat DBA that has been working on this task +That said I still need to look all the changes over in the morning +This is a feature that is supported in ibatis +Do you want to take a stab at this John +I agree that puttingthe complete SQL syntax in IJs help would be problematic in addition to theissues mentioned the IJ client can be used with a range of different versionsof the Derby server each with its own particular dialect ofsupported SQL +The error message you are reporting seems to indicate that the JRE is not installed correctly +Oops! I named it +Seems a little strange but required based on scheduler implementationIf were always doing the rack lookup we should change the MR AM avoid rack lookups if possible may be difficult considering it tracks RACKLOCAL allocations +dims +version of the config I had to copy some code which is lame I put in a comment with a TODO for you to see exactly where it is +Fixed inr of branchr of trunk +did you have a case in mind? i think it is reasonable to turn off the sync for the observer but we probably still want to log to disk so that we can recover quickly +Julien are you ready to commit this patch +Hi What was the fix for this another user is currently having the same issue and I was curious as to what was the cause and fix for the issueThanksSen + for the patch +About the itself methods are modifed all of them are not fully implemented in current Harmony code base and methods are added +I think Id like to take the perfomance hit for now and fix the broken behavior +Would be great if you could attach a test case to reproduce the issue this will allow for quicker resolution and ensure the issue stays fixed into the future +We have multiple implementations of view servers +Im afraid this might be one of those patches welcome moments though +Any comments on last patch? Should be good to go +Appending or indexing on a character instead of a String is about fasterIf performances isnt an issue then it should be left as people see as clearer. +While there may be ways to automate the process the lack of implementation of those ways does not keep a person from using StrutsRecalassifying as Enhancement +The name should describe what the class does ie or +And the file would provide the location of the services and modules to be deployedIn sysnc with the current model if the file do not provide the location of services or modules then it would be assumed they would reside in the folder same as where is +push to +Fixed +But it is still good to keep the feature that allows the collects RSS +When running in a lenient mode the message builder will pick up lenient field parsers per default unless configured to use a custom implementation +Hi Thomas that looks good I think you can commit it and then we start improving testing it +Added to all classes in HB. +Is there a way to describe this task more concretely so that othersmight attempt it? Can the work be phased? Why is this a bug? Can youestimate the difficulty level +Hi Steve Should we apply the patch from to the latest +Works for me I can add these but I need to know for each Special Characters Number of Plurals and Plural Equation any idea where I can get hold of that or perhaps there is a file upgrade I can do to make it easier +I am getting Invalid alter operation partition keys can not be changed +Added a new subsection called sourceModification to xdocsreference +Applied patch in r +Using the parent bean in error messages would be another use case +Otherwise Ill do what I can to make a request that large and try to replicate the problem +Id be happy to take a look at it +Tom started down the rabbit hole but it was too large a change too late in cycle +All of the other features need to work properly +Thanks for the patch Elliott +Which means this isnt a grails issue unless lines are slightly blurred +didnt commit it yet but attached a patch +We should get it goingI think the top patch on that JIRA is ready for review though it may have fallen out of date +To lock the entire table in exclusive mode for a transaction that will update many rows but where no single statement will update enough rows acquire an exclusive table lock on the table +use the form of update where appropriate +testing set umask to and ran with and without the +A test project demonstrating the problem I mentioned in the previous comment +Remove target since no patch yet and need time for review once submitted +release is slated for next week and usually it takes a couple of days for Build to catch up +and ifhow to handle possible semantic changes for previous version files + +bean and method call using ref now check on startup if the bean exists in the registrytrunk +Marking this for as a blocker +So you still get sparse storage +If possible please apply the fix to branches and master +I thought the client ID was only used for durable subscription +Uninstaller creation is now conditional and will be skipped if condition is false. +New version of the patch +Thanks for the feedback +on adding test +The link is broken +A patch will be posted soon +Modified the getting started and tutorial chapters +What does your schemas look like Alex? +Has there been any development on this issue? Im very interested in allowing my slingpostoperations and posts in general to return JSON instead of HTML since I am dealing with a front end +Hi this is the latest patch for events handlingThis additional patch should be submitted in order to fix test case build problems of the previous patchRegardsAndrea +Thanks Toby for pointing this relationship out +Seems to be fixed +Im pretty sure that every one of our use cases that might take advantage of a cache invalidation feature could be implemented as a materialized view with a mechanism to call the stored procedure used to refresh the materialized view so I consider this feature a low priority +To avoid the patch to become to bulky I created a separate issue to address the data node +patch for the index node creation it skips creating the nodes if they already exist +If youre using the resteasy modules wont the resteasy modules classloader be used as it is a parent of the applications classloader +The spec defines that this feed should support GET and POST but not DELETE +Attached a script that I of Date +XML Service model was retired +But currently since the broker uses the same socket server to communicate with the controller this is not an easy fix to put in can probably be fixed as part of +Will apply it shortly and get it included in +Unlike the Excel plugin which uses information from the source document to determine the background color for table cells this patch merely applies configuration settings made in to the transformation processApologies to anyone I may have confused by my patch description +out +RAILSENVproduction jruby true rconfigboot rcommandsrunner e production PhysicalUtilityTook a couple of tries but got a good pair While processing family Expected varwwwappsbpmservicereleaseslibservicebilling to define ServiceBilling loadmissingconstant at varwwwappsbpmservicesharedbundlejrubygemslibactivesupport constmissingwithdependencies at varwwwappsbpmservicesharedbundlejrubygemslibactivesupport loadmissingconstant at varwwwappsbpmservicesharedbundlejrubygemslibactivesupport constmissing at varwwwappsbpmservicesharedbundlejrubygemslibactivesupportFrom a separate thread While processing family is not missing constant! loadmissingconstant at varwwwappsbpmservicesharedbundlejrubygemslibactivesupport constmissingwithdependencies at varwwwappsbpmservicesharedbundlejrubygemslibactivesupport send at orgjruby constmissing at varwwwappsbpmservicesharedbundlejrubygemslibactivesupport constmissing at varwwwappsbpmservicesharedbundlejrubygemslibactivesupportUnless I didnt pass the option in properlyNote adding lib to takes care of the problem +I also refactored the unproxy method and the run method of console to make them simpler better on trunk revision Fixed on branch revision +Thanks I incorporated this code +This way it leave little more screen real estate for the text field. +reopening +Ive implemented these changes in my personal branch and merged them back into the milestones branch +The ha +Lowering priority +i have also updated ITC with the updates to gossiper let me know if this helps +When not defined it defaults to no timeout which is the current experience obviously not a desirable one +testResource settings +Actually Id rather have the config file property but if you want to have both fine with me +Basically it acts as an anonymous function +From talking with others it looks like this may only happen on a single node cluster as well +Committed revision +Checking this in +Committed to and trunk +This should fix your issue as well I hope +The bin distrib already use a pretty similar mechanism just less user friendly in the bin distrib we provide and a using to download the dependencies described in the +By checking that both are the same we are checking that the data that was supposed to be passed has actually been passed to the store and there is no bug in the transfer of that data +Even though this is the earlier issue there is more discussion on the other one. +Notably ants wrapper script uses so if thats portable enough then this patch should also integration tests do pass +Attaching example snapshot since the provided links will likely roll off soon +Apologies +its in +Is there any progress on this to report +cordova directory and trying again +Attachment has been added with description this case shows how is ignored if the previous block has a transition resolvedfixed bugs to closedfixed +A fix is on with +If I put a breakpoint on the on it fires +Could you please elaborate on this change +Thanks +Committed to trunk and +In some cases too it is much easier to depend on an extra repo rather that manually uploading to central every time a rd party lib has a release +I too missed it +snjezana so reading the patch looks like the problem was the refactoring into a default configurator caused this bug and this patch fixes itYou are right +Its there now +After the upgrade I think it is cleaner and simpler to treat this as hard error on the blockIf the only copy of a block has no CRC shouldnt we still permit folks to access the data somehow? I dont think we should just remove the block in that case +Master and Slaves are communicating using JMS FlowMaster uses the to partition the data and sends the step execution requests to the remote slaves using request JMS queue +Fixed via the patch that was already applied for the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Additional fix in rev Its not quite clear how to deal with the heuristic exceptions they may need the same treatment. +If anyone does try to compile with and encounters problems please reopen this bug and provide details +For those words that do fit in the line only one box is generated as before +What is included in this binary file Image files for Chukwa UI Apache Jakarta Taglibs JSP Standard Tag the tarball at the hadoop trunk source code and run svn commandssvn add srccontribchukwasrcwebhiccimagessvn add srccontribchukwalibsvn add srccontribchukwalibsvn add srccontribchukwalib +Thanks Teodoro I applied your mapping property in must be uncommented to use Continuum with has exactly the same problem except that the maximum length is +Dimistris marked the files as checked the compatibility +regardsNandana +The isGroovy change you are suggesting to is the next logical thing to look at +Also note that the already has the functionality to add Jersey resources +Closing all resolved tickets from or older +I will give a try based on your your support and response. +Covered as part of object storage plan has been executed +Here goes the patch for tests +Tomcat does not know about the wrappers +Fixed up the issues with the test pointed out by Dan with revision and to trunk. +I looked at chrisgs entire blog and couldnt find any tags that were improperly formedIll review the code for possible race conditions +and as well as in rev +clearFiltering returns the same JS error as api so increasing to Critical +Broker restart at end clears wrong queue stats +It sounds like youre close to reconciling your expectations though and that you may both be writing to the same feature +Dont forget to set the component +a clean now wipes the output directory if its not blank +Indeed might be relevant to backport too +This is correct so Im going to leave the table field even though its currently not referenced +And if you look at similar patterns in +made new issue for making impl is currently working but impl is forcing users of the to serialize themselves +Not that this helps you muchA workaround is to not use system package extras +No code provided unable to reproduce. +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Maybe it needs some restructuring +I had hoped that I could add more child levels later +I just committed this to both and trunk +this is not specific to JBDS nor a specific JBDS requirement +Note about building demosBuilt by mvn clean install Djsfprofilemyfaces to activate appropriate profile with u post any more details on the error or the config? What was in the lib directory? EtcAlso at this point you can open up a jira with this info please +Never used in demos and Ive look the code and this widget do nothing so can be +It seems we did the same simulations with same conclusions! GoodOK I think I did a lot of noise for nothing it seems and I do apologize for thatI just want to understand though why the fomulas in and NIST Statistical Reference Datasets are different +Perhaps something like getI am just trying to model it on OCF where monitor is a way to see if the service is runningstopped +Ported in rev +Reopened because Ive noticed that changes in class are missing +Yongqiang I think I fixed the double processing by using a different regular expression +ok ive updated the title to reflect reality +The fix in fixed also this issue +The has been used to launch lets say gateways on a single system and have them all pound on the same server testing its performance and stability +Patch to reflect relocated code +Added comments in the review a new patch that address review comments +Please wrap the long lines as ve added the pathc wirh version +So whenever a custom WAR or EAR deployed in the container of your choice brings an own ASM version then you can never be sure that it is compatible with! And I really dont like to render useless only because some customer projects bring an old ASM version as dependencyOf course I can see your point with ASM as primary target thus Id suggest to set the ordering as following asm looks like a pretty heavy bug in ASM so farget always only takes the own +In the process workContext becomes unnecessary for and so its constructor is simplifiedThe patch updates and passes test cases in core and spi as well as the supplychain simplecallback and samples +Revision +what good can come from allowing someone to force lucene to run onThis makes no sense to me +Attached quickstart +Tests with custom build of Shared Test Suite ended as run with online snapshot +bq +Please the the issue comments for more information. +See linked issue for my proposal. +Same patch as before except patch paths have changedCant mark this as Patch Available until the YARN workflow is fixed +Grails doesnt need it as such since we can use the solution I added for +Thanks for the reviews Nicholas! +There can be multiple pools available which gets determined by the planner when the attach volume happens +doing that now +This should be solved for RC +Manually tested! +Regarding that TODO it is because for builder pattern generally we dont have setters but here we need to set based on the Writer V or V +I was only speaking about French +push out to +Most build tools can handle this without any problem but Maven forces you to use their single output workaround above helps but is would be nice to have an option to use use Maven Integration for Eclipse in Dependency Management Only Mode +There may be a better way to do this reusing some of the code +I just ran the TCK against the pluto branch as sanity check to validate if its still compliant before we can start the release process and it failed on two checksIt seems the changes introduced by this issue are not compliant with the specification or the TCK tests are not this first needs further review +I also made protected and used it +As first step I made naming of the productions similar to those of EJB QL as suggested and done by Udai +updated title to match current plans for ArrayRail +We should not have to kill the server on a space exhausted condition +Hi Hiram is the original bug posters test case +Is dblog the best name? I think it is very Derby orientated and might not be that intuitive to new Derby user +Most likely theres some small error in the tests because of that +If you know how to do that with gradle then please jump inThanks +If the cell is null we put an empty byte array in I think it should been the s reponponslibity to handle null cell +We might as well keep it simple for now and extend those as and when required +Updating the summary although this is not a problem +Thanks Jeff and Ted for finding this copying +Ive updated the code base to talk about keyword substitutions and use substitution rather than macro as a variable nameAlso the linked documentation has been updated to talk about keyword substitutions not macros +Volunters appreciated for providing a set of test points +if we dont test to insure the segments with deletes are in fact gone were unfortunately not testing the functionality of the methodDo we need to? We should be testing that the method is called and look for evidence that it has done its work +As this issue is sorted in userguide I am closing this issueThanksChatra +This was caused by omitting the mechanism from the registry of SASL callbackhandlers +I do think something of this form could be a valuable change +it monitors the container process and kills it if its memory consumption goes over the configured limit +Andrew? Gary? Any comments +Ive only tested the seed site so farAfter reviewing Cyriaques comment I tested and its still broken +Looks like warnings are coming from itself +Btw as far as I can tell the services on Equinox add to every bundle +Hello sorry for replying late I was out of work for many times +Sounds like that Protege code is expecting to run on Equinox then +Same for. +Upgraded the test suite to run with and we see the NPE. +But they are not +closing issues for released version +javatestingorgapachederbyTestingfunctionTestsmasterjdkI will fix it soon +Id be really great if you could contribute a short section on the caching extension to the tutorialCheers +fixed test caseboth in and an issue for that in method callhad to change to isInterface since for super class abstract method we need an invoke virtual and not invoke some more tests checks +So I think this is an important performance issue + is not present in any more +Where would that name get injected from +Bulk move of M to Open Ons works fine for EL at least for seam to clarifyWhat were changed here? I see no commits related to it and I dont recall openons for EL being specific enabled for specifc tags? +Pushing to verified +You are right Paul I just my web app and it does list the directorywithout the trailing slash +The reason being what Yonik said in that the core is not ready yet at that pointAlso need to address the possible double loading in +You got alot of other errors in your error log and I think trying to remove all those first would be a good thing +If you can reproduce please reopen the issue. +Ill put together a suggested patch for this and open a issue +It sounds like what you really want is a distribution policy that will add new shards to the collection each month +Thanks for the heads up +Hi Devin I would not be surprised if this is outdated +It is expected that there will be jars in the EAP zip +Fixed in SVN trunk and targeted for inclusion inOleg +now its others estimated to hour and will be done tommorow +I am going to edit docs i have already sent ICLA to apache and send an information to camel dev forum +A in case that the annotation could not be found could also be an option +What happens when WAL compression is enabled +Its bug of WTP +optbin S s servicejmxrmijndirmijmxrmi u admin p admin This is the command used when using a remote jboss as ec ami and this also works properly + +This is a pretty gnarly one to dig into +I think that this feature request has already been addressed +For now I am just assigning this JIRA to myself +Its different with some file of local disk because they are stored in memory +No feedback after about years so closing this as incomplete. +If you compare the layout of and youll see that in SOAHOME the folder src has disappeared +is it to get info from and instatiate new +My current suspicion is that this is due to the fact that we are passing the request to and concurrently in +Thanks NicolasOK I just tested after an hastily review +So its not a show stopper just an annoyance +Id be keen to try it out someday but its a issue for me +Patch applied at revision +marking it a candidate for merge to has now been to +We have similar issues at other places like when you start a bundle we have to wait a little bit just to return the started state of the bundle etcIn all these cases we do a sleep alreadyTherefore I think your suggestion makes sense Ill change the code fix has been released with therefore closing this bug +Committed branchx revision after release. +The code should move the to Caches folder +Raising an exception when theres no loader would make it impossible to simply require jar files to get their contents available which is a very nice feature of +The fix should appear in Roller +j b o s s +for this change +Attached a patch withCould you please review and commit the patch attached +Switched to using compiling srctestjava only and the issue is resolved +We really appreciate the help and understand that this is not the only thing you are working on Collecting notes sounds so familliarWe ran a test with your patch to throw a when an output already existsWe did get a read past EOF but the additional is never thrownWell add the other log points and do another test run with those +On another note Dont waste your time presuming to lecture me on problem solving techniques +Hey Eric coolSure Id love to collaborate with you on this +bq +failed in build It passed smoothly on MacShould we make its timeout longer +Interested in sending a pull request +Looks good to me thanks in in rev and for trunk in rev +Not tested but yes it should +Can you point to the code and svn revision in trunk +Seems to be a Derby problem +so all peers collets and calculates the same result after each peer executes its portion of the loop a number of times +Issue created +Patch applied to both trunk and branch. +When I was doing I did consider that possibility but there were too many cases that it does not handle +bq +Agreed can you provide a patch +can you provide a sample project and explanations on actual vs +I am not seeing the issue anymore on galaxy s but cant confirm those other devices +Always curious please attach it +Network Client cannot be match embedded in this regard as the DRDA spec specifies that USRID has a MINLEN of +Changes need to be made to the source code Create classes for all data types +trunk support for a sources jar in r +Bulk close for +Well look into it and will keep you informed +The patch the original jira is far from being done I think this should target +I find the option to stop reading exceptionally helpfulEdit log toleration is even better since it allows you to specify the toleration length +Sounds good Thomas +Worldship does allow the shipper to choose either business orresidential +These are good points +Thats good +Backported to +I do however worry about the users +Andrea please have a look at this patch which should fix the issue +This appears to be a dup of +Vinod many thanks for the thorough review +I would say that we should not try to integrate such a check in the framework since it would probably make the codeharder to understand and maintainCould we resolve it to Wont fix? this old issue as wont fix +Missing double quotes added +NPE in war bug fixed +Attachment has been added with description patch has been separated since it is purely unit tests +LGTM +Hi Tony is the change already viewable in the odfdomdeveloper source repositoryMeaning If someone clones odfdomdeveloper source repository and create the is it fixedI always thought that the URL to be exchanged would be set as a Java System property in the read by the TagletDoclets and used to replace the URLWhat did you changePlease dont forget to use patchesThanks will take care of thisReview it and resync and push it after integration of +Stephen is that a Ive tested it and it runs OKRob I moved it to test package engine out of coreThere s a small problem about running under Eclipse + +The only one that is causing me difficulty is the editor preferences +Yep my snapshot was before was committed +The Graph event system must be in place before this can be done +Minor update added javadoc to the class itself +Closing old issues +esbtests +. +The process definitions and instances list work well +html in the browsers +Closed since the bug Velocity has been resolved. +Daniele since this is an unsupported module do you want to queue up the change ask Graham and then commit? It sounds like you can do this in a consistent manner to the other geotools plugins with which you have experience +Julien did you make a start with Ill start working on some code to populateupdate the hostDB from an existing webtable? If not Ill start on that one too +This patch didnt apply cleanly to vmvmcoresrcutillinux +refactored patch + +Also attached the patch +I like it! +Hm this is interesting +Vicky diff between and results in tons of diffs +we will discuss in team or better we already do but for me still a request timeout must end the request otherwise it is a setting in admin for this would make sense +is part of the Thrift library not something built from our idl. +Please verify that you dont have network problems at your end + for the patch +Run javac javadoc findbugs and release audit tool and report any new violations +Committed thanks! +But presumably by then you have already shut down the thread +Can you sync with trunk then hopefully a committer familiar with pipes will review and commit +sh files and reverted the csharp files +Yes this is expected and what is the problem +Thanks +Does this occur on? Can you attach a + +Closed issues related to Roller release. +and removed final from the name definition +Ok they are now +Dropping this to urgent priority and moving it out to +That sounds fine to me Andrey +Thanks for the tip EmmanuelI had read a little about those links but they look maybe too secret to be reliable +Do you need me to commit other scriptdoc changes +I moved this from blocker back to minor +This patch is the correct one from this patch already got committed I would recommend opening a new JIRA issue to amend it +Marking the issue as fixed + down but now theres a permission issue when accessing +Unfortunately this old version of axiswssj has a hard dependency to xalan see after the pom change +So we need to check in memory the name of the files we are creating to make sure there are no duplicates and print a fatal error if duplicates exist +Agree with you Christian +The tests run but with failures +You arent specifying the JDBC type for nullable columns +You guys are awesome +Those will be tracked in seperate jiras +Verified by AlexeyVladimir. +Applied most recent patch in revision Ard are you OK with setting this issue to fixed? I think once there are more features available in lucene we can create new jira issue +And also got rid of +Hi JeaglesThanks for your review +Hi ClausI have created the wiki documentation for Lucene and added the link to the component listCheersAshwin + specifies a theme +Also its a blocker +NamitCan you briefly explain why there should be one common entity instead of readwrite entity +GC has less to do then +You need to write some unit tests to ensure the behavior doesnt regress +And we will need to be sure that Ma users know about this issue and that they continue to have access to M until it is resolved +If nobody complains I will soon commit Bilgins patch +I think just issuing delete flags for the Files should delete them when the VFS gets garbage collected wouldnt it +Closing this issue. +Hi Graham Im afraid the only workaround is not remove the annotation +If it is exactly the same work then it is likely it does know +We are working on greclipse now +This is not hard to implement at all +Rerunning testsuite against with the additional artifacts now +Ok so how hard do I have to hit those two files until they stay dead? I remember Ive removed the bin directory once +after rollback my copy to revision all is working fine I know the proposed logic so i wondering about the curious new logic +Patch committed rev cbcdaecacdfb +I differ in keeping the AM alive artificially so that the client can talk to it OR unregistering and staying alive in the force kill grace period +Thats fine for me +Thanks Bhallamudi! +Sorry about the churn btw I was only compiling for which obviously isnt enough +thanks +Sean and Matthieu I encourage you guys to look at to help ease your implementation and to ensure you dont have to be exposed to the details +the latest concern that KB copies of site still remain in certain folders has been resolved +Thanks Chris! +Patch committed +Ravi are you able to reproduce itTried and trunk +This no longer applies +It will still be creating a new consumer every time +The code changes were committed to master in the following commit of work for storage migration feature for resources in cloudstack has been done over this ticket +Should the latest patch be committed to? It sounds like it should so Im marking it but please let me know if you disagree +Corrected to trunk and +Not directly related to this patch I wonder if we should send the exception stack trace also through +As we have changed the whole handling now I think we can set this to fixed as well +The attached patch should solve this +Need to fix the build to remove that +Initially I only really care about supporting and compaction making better decisions +A couple comments on Phabricator +Ive committed this +Also your patch is strictly a bug fix to your refactorNot exactly my fix reduces the time of running FI tests and it does fix the issue rather than your patch which simply swipe it under the proverbial rag +lgtm +I am just getting ready to checkin the changes where while dropping the table we only ignore table not found error and pass the rest of the errors to the test so the failure exception will show us what kind of error drop table received +Could you pls link to the discussion thread in this task +Like I said I have no idea if it works and it looks like it does not so I guess you cannot trust everything you read on the Internet especially in comments of some random tech help forum +Could you patch for too? One thing that we need to be careful of is that special characters in groupid could fail the mbean registration +ie +usernames +Fixed on and issue has been resolved for one month or more without complaints +So I meant that we can do the method refactoring separately +This will contains the required docs tests scripts etc +revision +Thanks issue is fixed and released as part of release +Categorizing bugs related primarily to stylesheet elements etc +It does seem to record the othertransactions in this batch +Perhaps all of should beI agree I marked all of the serializer classes as public evolving in the patch + +If yes can Ruchith please close this +Great! My first commit to the project! +Drain the individual consumer queuesSo if I change the order then I agree we dont need the message +Maybe it would be better to have components +This seems to be fixed now with the latest nightly builds of +Removed from failure is unrelated +studioeclipseplugins +Please provide a example as the title does not convey sufficient information +The naming is deliberate +classes +The test is covered in +can be used to see the side by side comparisons +code looks good to me +Thanks Ben. +OK thanks +Updated Branchx as well as AS trunk. +Admin can modify and delete users. +This issue duplicates so it will be closed as well +Ill file it about the text back and forward? What about changing imagesThis seems to be an eventual plugin issue +Patch against trunk to display in header of JMX against Branch to display in the header of the JMX console +push out to +I assume the patch would be close to identical +Server stack trace when the application is deployed +Added to KiranYour patch is in trunk r r r r +This will always cause an error if you delay your HTML load since the Callback Server will not have been created yet. +not even on mac +The specification itself says plainly Groups are represented by interfaces +Its the same icon +i was thinking the timestamps would be updated only on close or recovery +Committed to trunk and branchx +webProblems Error on startup Mixing nullable and non nullable columns in a property is not allowed SolutionSet nullable false in both Error on list some entities Unable to find Entity with id xxxxxx Failed to convert to internal representation +A google map layer has been created but it is not well aware that putting a WMS layer on top of a Google layer has problems because there is not a SRS EPSG code for the layer which means you cannot ask a WMS for a layer in the same projection as way forward on this is to integrate with the rendering engine and use the Layer +Added an integration test in Rev +I am attaching a zip file that contains XML files the XSLT and a Java program that uses the XSLT to transform each file once for each of threads +The RFC does not allow +I have run this test for an hour +No Im not aware of anything firebug as its highly unlikely this is a Spring Security issue. +This is pretty important for integration with any full text search engine to keep index in sync with database at least remove event. +Patch for trunk +Why is the numFound reported this way +I probably missed something here +Ive found much more issues in the tag handler generator +Addresses encountering +Nice doc +Please administration works correctly after installation of for struts is presented. +With a scanner do we still need +to be the reviewer of the patch per the JIRA meeting Nov +derbyall on Windows with ibm against sane jars ran with no new failures +to see how we deal with the linked +Point is partially taken care of +what is procedure to download the files into the exisiting Apache Solr move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Just to add that retry logic is used to reconnect to newly started AM if the previous one crashes +should be a list of paths +parts were added back to patch does not degrade performance on both server and serveraggressive emconfs as measured on Ghz Gb Gentoo Linux x. +Yeah if client adds new edit w exact same ts and the comparator used by memstore does not take sequenceid into consideration well have issues Todd identifies +Since Groovy we dont always box +This fix needs to be applied to trunk and branch +However I dont see how its possible. +Adding the logic on top of that with whether it is a active session sounds more accurate to determine a Derby crash situation occur or not +This was certainly a problem with the release of but it was not due to changes in our build +Hi Brandon Let me know if the attached patch is sufficient +Needs some in +This patch is an updated version of the patch +Invalid port is flagged by Tuscany +the use case you mentioned firstThanksDenis +logs from the failing test +Maybe I guess it is just me although the beta I used for some Prakash demos had some of the same problems +To be reproduced on a current Spring +Theres a new test that checks this behavior +When there is a patch I can try it out on a test cluster in EC +FYIThe version of is upgraded to in Portal needs to happen for as well +Ill commit the changes for the +There is some identifying information that I will remove promptly which will make the names slightly shorter but still long and ugly +Increasing the sleep delay doesnt completely solve the problem +Just out of curiosity why doesnt releaserollback actually remove the tag? I am very sure this is rather simple to implement but apparently there is a good reason for it not to be +I just committed +Needed for class loading support +Attachment has been added with description This is the test fo file on which the problem can be reproduced +Junping thanks for checking and +You can look at for an example of usage +Run mvn clean install e Deploy the bundle and launch in the SMX issue is resolved and now available for until patch is was brought to my attention that the naming convention with a was improper since the archetype generated a pure CXF Code First Web service with no JBI underpinnings +Strange but follows Eclipse way. +This fix will be for only +Beyond that that is what javadocs are for +This can be ovverridden by replacing schema dependency has been moved to thus this one moves as well +That sounds good to me +Hi John Thanks for digging into +Thanks for the review Hitesh +bq +Nothing open here anymore +This patch was developed against trunk ive tested in local mode and on the deployed gridI deprecated because its odd to have two ways to add script files to the jar and i wanted to have the patch work with and any code that might be written against the public member variable scriptFiles +pat i will review it today sometime +In that case Edison putting this command in post scripts will do the trick no need to even document it thenpost scripts sed i sg etccloudagent we do that lets first cp a etccloud etccloudI am generally against using post but dont think there is a really good alternative in this case +ENOSYS +Edgar the statusIds part of the URL is there to make sure that only issues that have been Closed in JIRA will be returned +It requires a backport of +This write completion event is triggered when data in the kernel write buffer has been successfully sent over the wire thereby making room in the write buffer for more data +The most recent was the one that got committedRaid is implementing simple xor parity right now but we have plans to extend it in the futureSorry no design diagrams that Im aware of +Sorry for pitching in late here +fixed in r +Committed as r +Patches for are ready for review +Potential issues +Ill be looking into this and we can discuss futher on the next the if your customer is still stuck is the way to go let me know if youve been able to workaround or not +one with explicit FALSE +Please go to the user forum for such but this is not a question +Rafi is there anything we canshould do here +Im voting we dont do that in theres already enough in +removing statement from the patch with doc changes and new change which solves the user limit comments on test cases Can we please include a short comment for the test cases describing what they do +The button itself doesnt really offer much on its ownI know that I originally suggested that this behavior belongs in and not but the in actually goes back a really long way back to when Pivot wasnt even called Pivot and was actually implemented in not Java +Im using it but I missed those + move to is what happens if some feature has a dependency to artifacts not installed via updatesite or the feature is the dragged in just because of that plugin +Deleting the patch attached that actually belonged to different JIRA + JDK for now +Im going to diagnose the issue based on the differences between and to see if its actually fixed for good +The difference is in the URL and can be seen using a tooltip +All fixtures pass although testStream depends on the patch for being appliedI think I have addressed the issues pointed out in the previous submissions +Alan Im curious why you think managing our own fixed point implementation is better than just leveraging and and the arbitrary precision they provideI mean I can benchmark things to see how radical the performance difference is but Im just curious +Replaced on the main developer page. +Ive just committed this to trunk and. +I dont know enough about the details to know if that IS possibleIn the short term we probably need to support setting a low limit as mentioned in existing configs and set a high limit if its not configured +wsdl +RHDS is not a is an IDE +From the comments on the mailing list and here it makes sense Jimmy Xiang will list the exceptional case to make sure they are coveredI will try to have things done in two or three steps remove the code from the master remove the code from the region serverI will use the review board for this in a distinct jira +are used in hash maps by and missing hashCodeequals was causing duplicate cache entries. +Support for as codehaus plugin but I will hapilly change back to Apache if it is needed. +I looked through the changes and they look good +Weve thought about that already but the spec does not allows that explicitly it only pass the given entity it would be quite inefficient to to support it since it would require multiple object graph is too tough for the PC farm evenThe problem could appear ONLY when the objects will be modified and there are no side effects when only readingThe most common case is reading the main object lazy properties collection for validation no changes in the objects +I think you mean +My weak mind cannot comprehend all the different instances where you may not want related dont know enough of Eclipse or Maven to fix it myself +Attachment has been added with description fixes NPE in transition resolvedfixed bugs to closedfixed +Nothing else was neededBecause Ive already tried invalidating key and row caches and the problem remains +Now that work is well underway with AS all previous community releases are +Using Sanjays new support for RPC implementations we can continue to keep separate and only mix it in when starting up as the BNGood idea +However paths may be parsed somewhere in classlib +seams to be used nowhere and can directly be removed +on itThe git mirrors still point to the old svn locations and need updating +Or you could even run it again without to see if it happens again and we could get another set of logs +By the way to reproduce it youll have to create a Remote Java Application Debug configuration in Eclipse +Closing as there are other bugs that are adding patches to petstore +RESOLVED has been refactored +New DDL script for with TEXT data type for the CLOB columns +This is the problem +Thanks for the patch Alexey +Apply patch to TRUNK and branch +One more question in the case of these projects will there ever be a need to apply patches to them? If so should we stage the project sourcesproject dir to some temp location under target so we can apply the patches without affecting the main project +The three files that would not update all have Windows line endings once I had removed those the patches apply perfectly OK +Added +And patch for should be applied +This should do it +Committed revision +I moved the error handling code from the platform implementation to Marmotta exception mappers can now be injected and in using CDI injection but they need to be annotated Dependent +It belongs in I think +Thanks for pointing this out +I have ported it to our internal branch +It does not affect if a user switches inside the server editor + +The core changes are committed in revI will resolve this issue and open a new issue to track cleaning up the +Here are some specifics Updated the version in the pom to like the rest of the poms so teh site can be built properly Add info about the general SCM format Removed implementations list from page they are in the menu Add more info about the CVS SCM format and the various methodstransports Added Windows specific example showing the use of as patch for includes the documentation patch that is part of in +has been marked as a duplicate of this bug +Looks like were having classic data loss due to the data type so might not be a simple fixA unit test should be very easy to create from Joss data above +The JMS client sends all the time like it has tourettes +Patch applied. +Ill commit this mid afternoon today unless I hear otherwise +but the tests might pass and they appear to be workingShingle was reviewed and changed by me I think this one is OK +This issue occur with java version +Ross great work so far +Okay Ill be the first to ask it C? Really? Is this the beginning of the end for Java for the world of searchI dont think so +Need to rebase +css files just in case +Or perhaps it really makes sense to push it thanks for the review BTW and follow up with some improvements +Resoled in now since fixed. +Removed compare loops +Doing it in isolation in MR isnt sufficient +I think we should start off on the right foot here even if its just one option for now +Screenshot of the web application before the issue occurs +not sure we can do it cleaner though +it uses Container from rather than allocated Container from +The error message states poller rather than trigger +Please confirm and if needed proceed with their documentation +Ive changed the default bridge to use ONLINEEVENT instead of HANGINGGET +fixed in r +Doc issue closed via +Przemyslaw thanks for the clarification it looks neat yesNote though I find it hard to see how this can be generalized can work at the features level yes but unlikely at the individual provider levelBut looks good nonetheless Id have no problems supporting something like that at level too +Im not at the moment so I doubt anyone is working on it +The major difference is between real +Lets restore the behavior from Maven regardless of how right or wrong anyone feels it is and start with a clean slate + for the patch +Hmm +Rohit Can you work with the testers to get this resolved +Good job on finding this +This zip contains some demos that use the packagingTwo files are plain spring configuration files that can be simply dropped in the deploy folderThey need to be fleshed a bit though +Closed upon release of +It turns out that this patch triggers during the unit tests +r on trunk +Changes look fine +If zlibs crc doesnt perform significantly better using what we have will be simpler for supporting different polynomials +Recommend you create a test case to demonstrate your issue +Luc will you be working on this +You cannot simply update the runtime wo updating the SDK feature +Do you think this patch is unnecessaryincorrect? Or just pointing out another problem in the same area of the code? Should we commit this +Ive opened a new jira. +hadoop compatible releaseTo build with hadoop in specify the followinghadoopversionx +Could you please add my work email to get change notices there it wont allowme to add it in the CC box +I just ran the example and it worked fine with soapui. +Committed to branch and trunkThanks Cheolsoo. +security is designed for the container to do the login not the application so its not too surprising that having your application do the login doesnt workIn your post please describe the jsf bean code whether you wrote it and have control over it and where you are looking in the wiki +If you exclude this as well it should dependencies on this module will hopefully be removed at some point before the Wildfly final release +implemented +here on moving examples like the bulkloader into core +btw +suggested failed another trysuggested patch +Feel free to reopen this bug if you think that its still worth to change the transformers +The overall patch looks good thxI have a minor concern on the patch thoughIt seems like you are changing the default behavior by checking if there is a prefix for the message wrapper element nameAs the default value has no prefix I suppose that it wont be used by defaultSo im removing the lines on setDid I miss something +Eli Reisman mentioned in connection with this they both have to do with waiting and timeouts +Sample request with applied to and looks good to me +Thanx a lot that did it for me +I am closing this issue +There should be small finite possibilities to skipped records +Im not sure this is a scalable solution do we really want each mapreduce task to come up with a server for connecting to particularly for large jobs this makes very little senseOTOH most profilers allow you to dump the profile information to the local disk +While very nice conceptually the split of interfaces into immutable and mutable versions proved too cumbersome in practice +Please see fixed. +Thanks for the patch +Code clean up will be discussed and performed in separate issues everything else is done +As you say the makes sense as well. +The new functionality doesnt need anything from Wicket so it can be applied to Wicket if approved +Makes no sense to publish to Maven central at all when the deps wont get resolved. +Thank you for your improvement +Will look at a character replacement option and open a similar issue against odataj +Correct? In my testing I found that length of each was similar to and once a complete record was read there were some more extra bytes at the end of the record which is causing all this issueWhen you convert a to is it in place replacement or the old record is marked for deletion by just changing the EXPIRINGMASK to DELETEDMASKI will try to produce a junit test case +Hmm +package has been set to abstract +Dont make this an official feature until there is a overdue test has been added to +Unless we are talking about EL variables here it shouldnt matter when the expression is created as long as it is evaluated later +Still need a bit more verification from some consumers that this handles injection of the reader in all needed cases +Patch rebased to latest trunk +I did not commit because Jake is preparing a release and he already made a branch for +Integrated to and trunkThanks for the patch MaryannThanks for the review Himanshu. +Hey Stack +commit ba +It now errors out +reopen for clarificationThe problem is caused by the change of the WTP deployment +As part of this we need to remove Turbine from the Jakarta download page + close of resolved issues. +Leitao and Hari apologies for coming in late +Closing. +I tried XEP and there was a problem too all string was printed tigh on the rest of the line with all characters stuck together or each upon other and was unreadable tooDefault should be to cut the string and to print it on several lines to preserve readabilityIm not sure it was possible to declare as hyphenation characterI dont remember the difference with third point +thanks for the additional information +I am going to close this one as a duplicate of Instrumentation +But shouldnt we do these changes in the codebase of the Modeler itself? Can we do this or is it better to do the changes afterwards +There are other ways +patch for already commit to patch looks ok to me +Set the fix version and resolving since is set +I had a look at the code and saw a few things that I would do differently You can use the instead of manually having to create the and handle it +Now that work is well underway with AS all previous community releases are +looks good +Note this patch also include fix for +I would suggest we leave it as is this functionality exists for almost years and I think users are familiar with it +just run any query that is generated when double clicking a nodeIf you enable the old parser via everything seems to work fine +Thanks to you for helping us on sanitizing the commons deployment! deploy on Nexus is working like a charm thanksHave a nice WE all the bestSimo +Devs can you please review so that I can commit this +the deleted supercolumn w a live subcolumn wasnt being included in the count +Will change it to a released version once hadoop team makes a release and before we release trunk as +Copying styles from parent style manager is not a functional workaround in our case as loading a module containing charts screw up other loaded modules styles temporarely +Any thoughts regarding how to accomplish thisThanks +Some comments +Marking closed. +Closed with the release of. +Field is initialized to which should mean retry once +Im the culprit who mapped class to type +We have over artifacts and we release once a week +Bagsie not please add a more detailed description outlining the potential impact +I also attached the failed Surefire report +The patch from AM has been applied at revision . +Otherwise the Linux client will resend the same request when it sees the write count is different in the response +Fixed in rev +Fixed +Since heartbeat lock is also one of the hot locks most of the time it might wait for current heartbeat processing to complete +Patches is MatthewThanks for you patch the patch license is granted to ASF by default you dont need to click the check box any moreI will take care of your patch today +Verified on Hudson +Feel free to continue +Patch applies compiles and fixes the problem +Out of the box both Jetty and Jetty load fine on their own and I can hit the example applications directly by specifying the port number from the outside +With the recreating of VFS to get around the issues they are not counting From the testing I have seen the files put into the vfs cache directory contain a number in the filename +Are you all over this or do you need anything from me? Sent from my mobile device +Move the majority of and blocker issues to the next point release +Remove expired sstables from the set +Closing resolved issues. +I think the only real place that I saw in the codebase that has the hard coded check for hdfs was in the +Thanks Edward! +Reading through the conversation consensus appears to me to be to make no code change to Lang. +Thanks PaulexThe fix looks fine +The handling of endpoint publishing is quite similar +I dont actually use Java to compile I use Java or so that code compiles fine +Attempting to provide a template for the resulted in ambiguity errors insertion operations that do not have a logstream counterpart will not be when the logger is disabledTo complete a message insert LOGCXXENDMSGSee examples for a trivial example of usage +This fix was required as a result of the changed in resolved issues. +Ok. +Ditto have tried with Pluging Cobertura Plugin still no luck +I think Id like to keep the error to avoid surprisesAnd some more eyes should look at thisjgray stack wanna have a look too +TedI commit all missing files for aries test cases in revision and +And to your comment Leonardo Its true that fajax only works on facelets but isnt it possible to use the javascript api also on +Fixed in +Submitting +I would much rather refactor now than laterI will send either a diff or the modified file when I figure out if I can migrate to the new validator +To Sijie has created JIRA for test failed +the same seems that commonlib is indeed unsigned but seamlib is still signed in +Brian reported it would work with deactivated cache +Closed with. +Here is the patch for this issueThanks Pranay PranayDone at rAshish Vijaywargiya +It has a with some documentation +Thanks for fixing this +Mail Reader EL Example Faces Example Scripting have been removed from +Logged In YES useridNote The way it ships it works +Used method similar to +I had run into that one too +Required typechecking errors are produced for +Also +Attached patch is for the branch +Please reopen if the issue persists. +Didnt wait long for reviews because this is pretty trivial stuff +May I ask you to delete the attachment please? Moreover I believe you should not need to update the repo but rather use a proper package naming +The following patch removes some of the old files of the original +Christoph are the two attached files fixes? If so please do us a favor and submit this in subversion patch formThanks +super constructor invocation now supplies correct oper to get versions sorted now shceudled for to fix +bulk close of resolved issues. + RandyCommitted. +TO EVERYONEPLEASE LETS NOT HAVE ANY MORE DISCUSSIONS RELATED TO THE POSSIBLE ADDITION OFFTPS FUNCTIONALITY TO AS COMMENTS ON THIS BUG +The follow up issue is now in JIRANOTE Ill not be able to spent time implementing GSSAPI or the like at the moment +Verified by Vladimir +Starting pointsThe Cassandra inputformat for Hadoop is in the record reader and input split are in the same package +Really clients should be making this call for themselves based on the of hits and what comes back in for spelling suggestions collationsin my case there are tons of similar words and i get suggested words for almost everything that gets submitted +Hi ColmThanks! The proposed fix should solve the problemThanksChance +Just confirming +These are standard defines so I dont think its all that important to tackle this now + +As I mentioned in there will still be cases where response has already been committed and it wont be possible to create a new session +they extend to infinity +Id certainly like to see this in it would make my life easierIm trying out the code now and hope to feedback in depth soonMeanwhile some initial comments theres inconsistency between space and space tabs in the code and a few that you probably want to remove or replace with proper logging +Thanks for the clarification +I close issuer +Bruno you mentioned in a gtalk conversation that you had a patch for this but I cant find it here in JIRA +But we only expect to make that transition just once so we perhaps neednt build a general mechanism capable of handling such transitionsOnce we transition I think well be able to handle interoperability between adjacent major versions using your rules +Could you please rename the method name to be something like get? The data in the buffer may have already be flushed +Thanks +Edited description for readability +Then we can remove support for old files at a later timeI would not add a property for validation rather than that I would only start validation if the namespace is found +Unfortunately eats the exception +should be renamed alsoprobably +I dont see any compelling reason why SQL Maps should arbitrarily insert whitespace that was not intended +Ive got a simple ejb resource with a single Get method +Hi PierreI see that this issue has been resolved from onwards +The current approach is acceptable but we need to flag this issue for a more comprehensive solution in change of Fix Version for all issues fixed or closed for +Id be willing to rework them to fit the test infrastructure if theres a consensus on how they should be integrated +Committed to master. +for the patchThe branch patch looks good to me +from me if Jenkins comes back OK +If a child is refresh we should refresh the parent pojo +A new set of code in modules which extends the izpack functionality to allow additional control of the installation work flow +will be removed in the final patch +Committed to. +I do not see why this is needed with the findbugsfindbugs and findbugsgui goals +Thanks Nick +Comment out the line below FIXME to see the bug go away. +Steps to reproduce was used from Description +wait for testing increasing for performance decreasing passage time for daily tests have been increased on hHave no results for +Can you attach a screen shot +Dont forget to update Thanks for your help +It should +The warning already exists for ping and statusupdate +Closing due to user retracting issue. +Hi MarkI was also coding up a fix for this JIRA issue +This patch is dependent on which applies against Y! cleanly +Sebb this patch puts it on the front page +I will open a felix jira for that shortly +I think this is the same as for which a patch is available +Is the real fix in a more central location like or Configuration? If we just target then well have to address this behavior in many places and open the door to inconsistencies with missing configuration files based on whos using +I have tested several cases to make sure it works fine +If I omit this option problem test seems to be not compatible with full profile +Is there any status update for this issue +I couldnt replicate the problem using those test classes even before the fix for but my real app showed it within a few seconds +This was fixed by. +I want to keep it completely contained to improving toString to make debugging easier +Ill give a shot. +I think I have a solution which Ill try to apply tomorrow +Use as alternative which is even easier in. +Ive just started looking at the jmx stuff and the things that jumped out at me about the jmx stuff are what Bobby said +I have committed changes to trunk +Adds semi colons to the viewcollation test and etagshead test +The new patch Ive uploaded includes tests +can do tags or not +Yes I stumbled across the though there are two of them but could not figure how to apply that to an Configuration class +So your connection counts might be off +I fthey are not in the patch then I probably did not include the unversioned files +Im surprised you cant get this path since we are using it for the oauth check +It is not clear to me exactly what is being computed in get +Hi JustinIm assigning this to you for review since youre the module this directly into seems wrong to me +I acknowledge that +When a dependency is discarded because it creates a cycle the remaining dependencies are still processed as usual +Thanks for the patch +committed this +Im compiling on Win XP too using eclipse and the javacc plugin +bq +New patch with minor modification +Im going to work on that soon +Compare it against the JSP and what you would expect +Changes ported for Remoting +most recent build which is does not contain duplicated plugins +To me this is an WTP issue but Im not completely sure about it +Thanks guys got it startedJust upgraded +issue is reproducible on latest master +Looks like Vladimir I +Thanks so much for spending time on thisWell be testing too when Im back in the US + +Wow you figured that out very quickly +I forgot the patch needs to be ported to trunk +Patch for for the patch! fixed as suggested and applied same change to as wellfixed in svn r +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +If you set the systemId of that source to be something like filehomevicdtd then the parser will have something to absolutize the relative URI in the external subset too and it should give you a more accurate result +I think pretty much every project I know of that uses task has run into this issue +Patch file for UDF doc +Thanks Jerry +I agree with you +ran derbyall and got failures in these tests which seems to be related to changes in this patchderbyalljdbcapiderbyalljdbcapiderbyalllangderbyallderbynetclientmatsjdbcapiderbyallderbynetclientmatsjdbcapiThe failures seems come from undexpected metadata output due the new procedures +Keep the interfaces in the Core +The Pipeline class will also catch and rethrow the exception to state which stage there was no mapper or to trunk in +Now that legacy servers have been replaced with boot there is no runtime information displayed in any runtime +This commit was recently reverted +Thanks for the analysis Ivan! s state machine is missing a number of events in the FAILED state +Corresponding Guvnor jira has been closed by Toni +Paddy can you show me your URI configuration of the beware the FTP should not use absolute paths for names +All documents were read successfullyI figured you probably had but they wont end up coming after you they will come after me As expected no issues hit yet thoughIll commit this later today +So yes these could certainly be updated to use the screen widget instead of including forms in an FTL file but its not a huge deal +fixed in svn r +Too meaty for a point release and master rewrite is all about fixing this kinda thing +has same issue in Apache Flex develop branch +Where do the docs live? More than happy to contribute +the suggested fix in a patch thank you! +I realize it may be premature to file this there are newer versions of Jetty we could try first +Scott what do you think? Is it relevant anymore? Id prefer not to spend time on this since I dont this is a big deal per se +What is wrong with it anyway? Isnt the important thing that the wicket ajax js reverts it. + to trunk and branches +Closing +doesnt the detectors say from within what detectors they should be used to find possible subparts +Same patch as previous one except that the fix to is removed because that is taken care of by +Downgraded to Minor because there is no response so it cant be very urgent. +Thanks Jordan! +Attached +This creates and +We have to modify init as well so that the instance does not get initialized more than once +Remove old docs incorporate docs into the build. +Will this patch force clean up any tombstoned data or only because of columns that have expired because of their TTL +There is also a similar problem with Mac OS X build where only libcurl is available The problem is that the test for the libcurl presence checks for a header in usrinclude and usrlocalinclude +Usually we name them compile runtime test and there is the default one called default. +Patch committed to trunk +The attachement is the test +Dyre why exactly are you adding a new test script with one simple test? Could it be possible to add this test to existing trigger tests? It is possible to get a connection to someUser in the middle of an existing test +Notification message will be show after +Sorry for the ignore forget about this issue +Attached patch fixes some conversion functions +jar I can upload it to fusesource repoI tried downloading it from the apache site but even the bin zip did not contain the +One more thought would it be possible for this command to support passing as the source argument to read input from stdin? This is something that put already supports so it would be nice to have on append too for symmetry +Rather I think we should consider making again +We triaged the problem by increasing SLEEPTIME and by increasing the number of retries +does the key of the map always requires a String why not an object? +Instead of having clearance types have it point to a clearance type entity +Burt thx for getting back to I dont recall for sure the setting +Hi Ravi Thanks for the compliment +added a small subset of nodes most of them are dangling +Best regards +How I can solve this? +Patch looks good +JDBC tests run cleanly +A similar fix has been made ready for but is not yet committedavailable. +Patch applied in r with thanks to Dan KulpI was kind of thinking a build system component in JIRA would be better of these kinds of things +Dont worry Ill find a solution one way or the other +Adding a patch file for DB +Thanks Burt +If you originally created your database with Cloudscape then it should upgrade but definitely you shouldnt be changing + +Also modified the integration tests to now issue deletes to verify the accepts deletes +Thanks for supplying the original patch Takashi. +to vector branch +Setting Fix Version to future for issues without a set fix version +Note that this issue is wider than Weld Embedded +Assuming that fixed it and resolving as fixed for now +A nicer API for mapreduce +AIUI there is no default +Of course it could still be exploitedMarcNo problem with handling this in SantuarioColm +we can run tests without Byteman in arquillian currently although it has only oneNow Im working on inject Byteman script to +We probably have to remove the action for the Project Explorer fix this issue I would need to know exactly what this action is intended instance the action is called for any collection what I suppose isnt correct +my laptop can mdsum at about MBsec so given that log segments under recovery are likely to be much smaller than M I dont think we should be concerned about that +Sijie has several comments on this quote them as below how to delete a ledger? so you need to take care of telling from a leave znode from a inner znode in your implementation +WDYTFrom user point of view I dont see any reason to change skin of running when unused skin is deleted +Attached normal output for the sample project using Ivy rev debug output for the sample project using Ivy rev is normal Ivy behaviour +Is there a functional problem as well or just unnecessary output to the console? Id like to make sure this isnt hiding something +When is the next release scheduled for +I switched to use which seemed a better way then my first version +How can that be correct? +Hi Paulex Would you please try this patch? ThanksBest Regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expectedBTW I prefered not to remove the old patch when you submit a new one so that the history can be tracked and on the other hand sometimes Im confused if the new patch has same name with the old one +Here is the patch for review +I havent written a migration script yet but Ill try to do so before +I looked into what it would take to make the RPC server support a state wherein it could return one final response to the client who initiated a shutdown from an RPC but cancel all other and not accept any further incoming connections +We can extend it to try to create non existing directories and check the result +After that fix it will never return null +Right or wrong its by all accounts the dominant convention for similar +A couple of months ago we reduced the number of people with job admin privs +Ive done that by cloning this issue now Im this one. +It is not related to the +Patch on svn head +The only problems that could have caused this are no exception thrown at a paramater that could not be set it is not possible to work with JDO configurations according to the use of Unmarshaller to set the parameters at the of this issues has been resolved while working at +On the point of of blurring between INSERT and UPDATE I personally like the separation +Finally decided to create an option called default which can take values or depending on the Task Executor type to use +Is that going to be fixed in +this looks like a recent post in theif i remember correctly i asked for further information that help me understandwhat problem you are struggling with +Ive just noticed that seems to return the same session rather than looking anything up +Have you tried to remove this lineThis plugin has been written one year ago maybe there are a few incompatibilities +Patch with unit test +it and it depend on the +Created for the Decimal type literal +xsd is not changed yet +Hi BabakIm sorry I didnt go though the comments carefully +Oh and one more question! I was looking to change the JIRA notifications so they now go to the new lists but it appeared there were a lot of separate notifications which would each need to be deleted and recreated +What matters is the encoding +I have committed it at subversion revision +The resolution description should really be Not a bug since the Exception was caused by some sort of cluster misconfigurarion on my side that I wasnt able to reproduce when I started all over using a clean portal bundle +Here is the maven project README m going to test the example and add it in the SMX samples setThanks successfully on last SMX to +Thanks IsaacThe new Hibernate Tools runtime jars were created and included because the ones that were included before were build from a snapshot rather than an official build and thus not really safe from a repeatable build perspective +for an optionproperty +Closed issue for version +If that works heres a patch for the NPE for the patch +Hmm +That patch solves this issue too +deferred +A backport of to +Tom Jenkins failure is because testcases are run at MR level and they are picking up a prev version of common JAR from M repo not using the changes in the this patch +The guide is available is version +Thanks JingI now have a better understand of schema and columngroupsWhat the projection and partition are used for +Feel free to add comments here or to attach test cases and Ill be glad to reopen the issue if necessary +See +Rejecting a new member who wants to join under the identity of an existing member solves the reincarnation is resolved +You are right +Attaching new version of previous patch +Generalize this task to check all classes in the jee modulethat are serializable agree with Suns serial from on how we need to fix the ids to be compatible with Suns jeeri +Package level attributes are a convenience when DRL is compiled into a binary package the attributes are inherited by the individual rules +Thanks Suran +Im working on this from a bit of a different angle +Good! Please feel free to pick up conversion of though +Hi Andrew Im pretty confident that this is fixed now + +We should address this issue +Committed. +Thanks Ryan posting the issue so the Build engineers know would be great +Good point +its optional for a reason its valid to want Cassandra to validate values as a specific type without also indexing them +Youre right it should be ID +Had attached a wrong patch earlier +Only when dynamically generating from the wsdl using the +Can we at least publish the pom only then? If I have to prepare a different bundle for this please tell +The watch table has a hash that keeps the alive until explicitly cleared with the clearWatch command +Committed to the trunk +bq +Rerunning through Hudson +It becomes really a big performance problem for us +mbs out of the max mbsIn the case where everything is filtered were churning through gb in the cache in less then s in byte chunks +Planning to write a testcase +You have mentioned a HEAD problem in your comment +Patch provided get into +New patch to fix broken unit test +What exactly is the semantics of a destructured point field? Are the components also val fields of the same object or struct? Are they initialized with the parts of the destructured initializer of the containing field? What about fields initialized in constructorsOnce the above is answered we can implement this +Could I mark it so +applied to branch +merged onto branch + +Part of a bulk update all resolutions changed to done please review history to original resolution type +Not being able to I had to the again +I am a believer in lazy test cases where each problem gets a test case so we can verify when it is fixed +Any word back on the results of setting the limit to be unbounded? If so then can this be included in as the default or are there other issues that might push this from to +I do find a potential defect in this moduleI will investigate it and supply a fix as soon as possible +Seems like thrift generated code changed a bunch Im using the compiler + +Has anyone seen this bug? If its invalid please let us know +This bug can be closed +Ill work on this today and post a patch +If this should be reconsidered please add the label +I have long time been using OSS and id really love to commit something to the community in return even as little as this +The client will cache a list of nodes in cluster +AFAIK the headers are completely ignored +Yeah agreed +bq +Could you provide more details about the code you are trying to make it work +Feel free to change it to ordinary if that wont cause the patch needs verifying for correctness +though it is tantalizing that the glyph for this font are in fact Unicode code pointsI just hope there are not too many buggy fonts out there! +I am interpreting your reference URL link differently I am reading it and seeing that using exceptions for recovery is acceptableThats a separate issue +As a result we can support JDBC and JDBC +After changing all the bundle list and seeing the effects of it I would rather agree with Reto +Up until there was some fixup code which tried to determine the id nevertheless which was not entirely correct so this was dropped +works now +This issue after wrapping some operations as jobs cause jobs does not block the workspace from triggering builds while they are in executionWe should use workspaceproject locks inside these jobs +Im trying to get this testcase done myself +Further remap requires extremely little parsing of a URL so it was easy and faster to handle that directly +Unique dirs per vfs will help with the potential for collision but then we need to make sure that the different cache dirs dont collide +for a more proactive management of excessive sizes +It just doesnt work on ICS +Ruby lib requires rpsec version rspec is not supported +r +What do you think about splitting into and as suggested in? +Thanks Sid! +Assuming the missing values arent very important you can create a +Thank you +Thanks for the feedback Dennis! +Committed with regression tests. +closing it. +Lets wait for QA +Attach my patch +John failed with diffs +Breaks Facets API. +For this some of the utility functions provided by WSSJ is used + +Didnt know you could set the release to rc +looks good will merge if the tests +Sets the query timeout of the validation query to a fix value of seconds +Will get back to you tomorrow with my Ronald just tested with version and the problem still perists +Attached additional patch resolves the array cast issue Amareshwari identified +the default value has been added into the attributes description +Ill check if that is still needed and if so report as a separate on and trunk +Just a quick question Ive setup tox and things how do I get to run the test just for a single provider +This is still an issue with +The new test case in discloses +tested on last local build with WTP +There were also recent reports of failure in this test attached to and +As Lucene is a library it supports a wide range of use cases +Committed this to trunk and +deployed is added later +Verified on EAP. +There still could be racing issue in resetting the counter +netbeans works with and without import of settings +So we can handle specific code to manage continuous replication or just render a change +I will try it on a different machine +The import and purge task could be added later +You said that a couple committers think this could instead be integrated at the lucene level and that some lucene users have voiced an interest in join +Hence and librecordio are not included to lower the barrier of entry +leaving ticket open by added which can be applied directly to the master +to unscheduled +Here is the inheritance page flow +systemjobidoutputs can be directly ttroot as was discussed offline +I think mode sync is better +Thank you! +You get thisMessage Unable to locate constructor with Class parameter for class Caused by Unable to locate constructor with Class parameter for class +I wonder if this kind of optimization is more generally applicable or it is worth pushing down into the regex comparator +Step If the collation derivation for DTD for this node and otherDTS do not match then check if one of them has the collation derivation of NONE +In addition to the above changes its also necessary to do the following +This is fixed with bug +But Ill delay changing the on other issues for now though in case we decide to do a compatible bugfix release +Does anyone konw how could I reproduce thisOr if you have the datanode logs showing this problem please upload them +Bulk close for Solr +Thanks +Looks good +If this is fixed in it doesnt make sense to file it against the same version +I guess if executing the class your launcher starts via java is normally working well then it is most possibly a problem in your launcher +Please update to the latest snapshot and let me know if you are seeing the same +Yes categories was not part of the August WC Contacts API specification so it did not get implemented +If estimate is split in half it will be less inaccurateThen given that all these totals are speculative even in normal case and moreso after split I am adding the artificial reconciliation of compaction progress after compaction is finished to show the final total and current values +I have not had time to look into the proposal and my feeling was is this complexity worth adding +Should be easy to fix +if it is not +Im going to reproduce this issue +Ive removed the useless factory +Thanks that fix is working for me +right these are osgi specific tags that will get ignored normally +Dmitriy Thanks for your suggestion I adopt your idea to let the second argument have all the options and let the first one as the columns argumentPradeepI check the test report It seems that the new generated javac warning is acceptable this warning is the same as the another one warning in +Thanks for adding the testportlets for the testsuite +The problem remains the same +Hi I am a Computer Science student from India +No nothing is missing from the the problem if intercept is true my class needs to have a no arg constructor +But Tim asked me to ask you +Patch that removes the abort statement like the other callback has +It will retry until region is fully onlineIt is crass +Thanks Paul! +I took a look at that class and it does look somewhat similar +I need to think a little how to refactor this +i didnt edit the xml but instead use the tools we have +Dave is rightgetting that correct was a huge headache when we first made an effort to get eclipse running correctly +fixed in trunk and backported to and +Their icons and splash screens are simply bigger +Nothing else happens in main it just returns +Version didnt have compilation problem +bq +Verified on +needs to either generate shard requests or piggyback on existing requests to merge the data in finishStageThis brings up interesting questions How do we perform distributed search for which are not added by default to? Only if the chain has can we modify its requests +Fix broke trace test +what problem are you trying to solve +Hi JochenI tried the code in the groovy consoleCheers problem happens also for normal interfaces +No counter mutations are only the unflushed system changes are replayed +The feature itself is complete and I would like it to be in RC for other users to test +Main cause in this situation is that url to file contains whitespace character +As part of the configId fix the will now be a wrapper around a repository +My point was that we can call put before initializing localEndpoint to so its naturally going to be null +That way we get prefix compression on the transport of a list of within RPC which should improve performance +Let me write the code the way you have it and see +Patch committed to branch Committed revision . +Thanks! Now there is more work to do +Todd any reason for not porting this toI plan to port some of the BKJM improvements +Thanks +Im not sure how but it is on the classpath for javac but not javadoc +My mistake +Change to fixed in +Could you please try out the latest SVN snapshot and let me know if the problem still persists? against trunk and it now returns failed which is what one would expectAny idea when the next beta of will come out? Thanks +If you dont get them all fixed in the next pass just add a step in the README that says which ones to delete +I have never tried to have two versions of on my machine +They are not static though +Also the JPA spec doesnt define extension points so this would be heavily provider dependent +The configuration of the xml parser is something that just should be done externally with the jaxp system properties and endorsed library configuration rather than relying on setting up classpaths via source code +Below is the updated textPlease review and let me know if I have the correctly documented nowException When you create a view trigger or constraint Derby first checks to determine if you have the required privileges at the +So this is not intractablebq +Defer all unresolved issues from to +onkeydown onkeyup onkeypress were hidden +Assigned for making translation +Please verify it was applied as expected +Moved logging utility functions from to +I also kept the groupings the same as the if operation so they may need to be changed to whatever is suitableIve also removed the stuff in the patchRegardsDid you test it? Id appreciate an example of your test in for instance in place of the loop op +Patch appliedThanks againI might be wrong and I will check again but I do not think the same issue happens elsewhere +Nah thats perfect! +This patch fixes so that it properly ignores class path resources that have been registered using register +committed rev . +This whole issue is about correct initialization of and while phase is processed +Have updated the summarysubject accordingly +As per recent change the class in no longer contains the static create and release methods +Can you provide some sample values +Thanks Steve! All done +They seem not to update componet tree if you just move component to ather place without restarting we cant do anything there if AS Adapter publishes the page correctly +Thank you +Thanks for the fix Pradeep. +The sorting is supposed to be inheritance based as some parts of the compiler need that and isnce I introduced that because of those parts +An actual jar is required and Eclipse has nothing to do with it Maven is an Apache project +Hi Christian +This patch adds the static resolve methods to and changes the use of in checked in +the changes to target intentional +Lets plan on this after we are done with this ticket so we know exactly what to displayATM Good point I filed to take care of that +Will follow up with a new test case +Great thanks for verification Ron +I also see only failure with the message Can NOT get breakpoint location! and RI does the sameI still cant find a cause for this +Yaaay! Feels good +Emmanuel can you tell us more about how this came up +Per exempleLi va etzibar que era un va etzibar ets un palls +Let me know if this issue is fixed so it can be closed and a eventually released +First time around I did not understand the prepareModifications method +Merged my pull request into the master branch + +Runtime Detection wasnt checking duplicate names in included runtimes +Its in CVS now +I just uploaded my proposalsWDOT +Its not this patch that was prob +An updated version of Karels patch +The current behavior of the Netty server is neither intuitive nor does it fit Scotts desired semantics +So there is no need to have the JAR file at all +The fraction is alkowed but optionnal +can have ROW and ROWCOL bloom filters and Cassandra Key bloom filtersBeing able to know they exist which points in the schema they pertain to and extract them from the underlying store is very interesting to optimize a series of use cases +I think that updating the Jelly libraries might have been a bad idea +Its already done on trunk +I thought more about Philips suggestion +That might be part of the problem +Checked in to and trunk +Hi MatthewWill when resolved provide what you need +Pushing off need to consider howif we deal with this once dispatcher is out of the whiteboard. +take quite a while +yeah disabled row cache with multiple threads running no assertion error so +properties file +Basically we always know which kind of we are dealing with we then should pass it to the when registering or searching for some name. +Once moving is done Netty will be released immediately +These values are passed as the newReplicants list +Ive asked Chris for the IP address of our old server once I have that Ill point something like or at it so that we can pull the content from the moment I have no way of accessing the old content +The SQL parser converts SQL to Entity Expressions and then uses the Entity Engine to execute the SQL +in the version for nb Ive added the possibility to execute maven build externally +It is a plugin issue +Could you attach a patch instead of the file itself? it shows the changes and helps everyone evaluate it +committed to in Developer StudioVersion Build id Build date +What I meant was that those jars seem internal to the AS and not thirdparty jars +Does infra have another means of getting this to them +Does this broadcast to all masters or just the connected master? What is the idea behind keeping these noncommissioned node around in the status list anyway? As well as could we not extend the unmap command to clear a node from the status list when ran? Since most likely when you remap it it will appear on a different master +Yes we should just enable standard menu for our editor +I tested with another key that I did not import and the connections was terminated +These new error messages have yet to be finalized and are still forthcoming +Fix logic error in server lease expiration +I committed not this one +Lets see what Frank has to say about this and fix this issueThanks FrankWill modify the and same in Guththila as discussedBTW a +Is this issue solved? Would it be possible with some hints for how developer should attach multiple behaviours to the same event +With it some tests still fail but they belong in a more specific ticket +I implemented the optimization that avoids boxingunboxing +Iv played with this with my own load testing but I cant get it fail +trunk +I guess the code for including tabindex is already in place however the attribute for the same is missing from the tags +SKOS shouldnt have any problem with multiple definitions for a concept +I found a cause of memory leakI will illustrate details and show an idea for modification later +If invalid an exception is thrown +And third why does only the serialized field fail? The other mappings of fields match +infrastructure pinged to see if this is possible +On Linux you can delete file even another process has opened it. +But I guess the absolute path thing is also present in +Issue may already be resolved may be invalid or may never be fixed +The interface is not a Controller interface its a utility interface such as or even worse the trait which is automatically attached to all Scala classes +But the reasons arent really that similar +This seems to indicate the feature type schema cache is involved somehow +Since you are in this file could you pls generate serial version ID for the servlet in order to get rid of the warningEclipse will do it for you +Shouldnt we move this issue as will to +is it correct to store custom properties in security doc except admins and members? +Thanks for the patch +Committed to TRUNK +has been marked as a duplicate of this bug +working on right now +The latest snapshot should work well be doing a release in the near future +It looks like our patches crossed in the mail againIm happy with using the Python code that you just committed vs mine since mine included a separate file that isnt necessary +There was a in my patch in the way it detects the IBM VM that Brian noticed +It turns out the problem can be reproduce only when using IBM JDK +If would be happy to find a better solution and will publish it here if i find but feel free to do it beforeBest regards +Can you please use review board this time? Ill try to use Phabricator for future reviews +Going to dupe this to +Thanks again for this excellent patch which I was able to apply with no changes! +Use bamboo +It may take a day or two for me to get around to it +I dont see a tag for but the version from works fine. +Assuming im correct about their being an optimization to not open a new searcher if the commits are identicalI did some more experimenting and confirmed i was wrong about this from Solrs perspective a new searcher is definitely getting opened and warmedIm not sure but skimming the code I think this discrepancy between the commit point in use and the commit point on disk may just be a result of using in +Probably it require some additionalother steps +The latter should be a good workaround for the short term but Ill get a out with a bug fix as quickly as possible +Starting Eclipse with clean solves the issue for a a classpath issue or an interference with another plugin +Attaching an initial patch +I know IDEA +Im happy with that +but your test doesnt use the fact that the schema x has a field f +number of possible parameters passed on registers +Thanks Hari that explains the missing piece +Thanks for your patience! Will do although I might not have the chance to do so until this evening +The patch attached fixes the problem Tom saw and not related to issue in Jira description +I ran the jBPM and samples in release against Jdk +If you have an approach to refactor s AES code path I would be more than happy to adopt your approach +if I add a dependency to it found the cause of the NPE when a created the unit test for it +just marking it as for not since is out before dependenciesbuilds have been aligned to make this happen +Then run the client test cases +I am going to unassign my self for a KVM expert to pick up +Just schedule new blocks to the next scan period +Something is wrong with at the EOF there are multiple id entries +Please reopen if it needs to go the way of the dodo too +Great start +The updated patch is attached +Ive committed something that should fix the small functional differences from Jays patch both standard error output are redirected to the same file I kept the standard maven output instead of using svn commit m support redirect execution output to a file +Ditto for objects that should persist across vertex but not across DAG boundaries +This doesnt look like such a good idea after all +it worked fine +The test case +Thanks very much DagCommitted patch to documentation trunk at revision +Attaching v to do that instead +Is the root cause could be the same for this two issues +Adding dependency tree reduction as image + +One of the page is text editor page +Attachment has been added with description Short with empty wrapper generating P open bugs to P pending further review +I still get the same errors even after cleaning and rebuilding itBy the way I could not find the respository folder you referred toI thought creating a repository is an optional task and hence skipped it soon after downloading the sources +It would also break the current contract +Yunkai lets move the discussion here +Sorry that would be horribly really confusingWrite state are for the data portions of the codec postings vectorsthey have all metadata available at this point so it makes senseHowever these metadata portions are bootstrapped do not +You should dig into what is doing +Good catch about the Portuguese manual Knut +I should have closed this issue after I reverted the original change. +The test still fails on iseries but will not leave the hard to remove directory behindI will commit this patch this afternoon unless anyone has comments +I just noticed your command line +I think this is unlikely and should be addressed as a bug should it ever occur +The content is not intact +Attaching and a corresponding tarball of generated html pages +While indeed the JS container doesnt get a lot of attention that doesnt mean we cant accept a patch for a minor improvementApplied committed. +Done. +My expectation as a user is that I would turn on replication and then do a MR job to transfer already stored data from the source cluster to the new peerGood idea I feel dumb not thinking about it +Committed to trunk. +the schema definitely can work. +so no need to clean them up during the network gc +This patch depends on the patch for +I just ran into this limitation earlier +revision +Thanks for the fix +for the patch +This is a bit curious +Marking as duplicate. +Thanks for the fix +Thanks +Hi Dong ShouyanAre you going to open source this? If yes can you update the recent work? This can help others to better understand +Branch tags created +Ah ok now I get the same error strange that I did not see it beforeIn this case I think it would be fine to change the epsilon to but I would like to understand why we get different results for jdk and here +Thank you for pointing to that +Fixed by +I grepped all of jbosstools code base and I can only find reference to in Arquillian and it is only referenced in an instanceof should be trivial to do dynamic test instead of hardcoded version? or even better find a way to avoid having to rely on super internal class loading semantics +Currently there is no simple API for dealing with all this originally there was just the global variable and that was fairly straightforward both within Gant and from a script + for +This patch moves the mapred classes to a different jar +Verified in JBDS Linux Sun JDK +IT added +examplesbinwork c chunk in reads from local not hdfsand reads from hdfs not localso the second isnt caused the trailing slash but also not sure what is caused by +I think the patch allowed the file descriptor limit problem to become visibleas far as goes I will move forward soon but we have some production stuff running and getting the time to port and move the production app and cluster is difficult. +Just to add it may not be a bug depending on how tolerant we want to be to what browsers send as Cookie +JeffDo you plan to publish your patch in of this JIRA +Patch applied in trunk +Cancelling patch till Cheolsoos comments are addressed. +Filed for the remaining places that use the filename parameter +The issue now is fixed in the latest head of both and the richfaces file upload component now works sufficiently +Updated and retested on Chrome Firefox and Safari on Mac still works was also a slight problem in the javascript which has now been fixedWe Just need someone else to verify it on IE +I believe the issue is that turning a into a representative Span query is only an approximate conversionI will look into whether or not I can improve thisThanks for the feedback +Thank you. +Look at these +we can look into target jsr +I think the basic question is whether these components should be explicitly designed for concurrency or whether statelessness is a nice to have feature that should not harm performance or complicate implementation in the sequential case +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +My preference would be to configure this at a broker level I would imagine that most users would be using the same topic delimiters but would welcome input from others +This needs to go on top of and includes the following since last patch avroj genavro tool implementation fixes to so clean build works properly with javacc stuff test harness and some tests in srctestgenavro added support for namespaces added support for errors added support for using reserved words bare in certain contexts and escaped by backticks in others +Added section on custom caches to srcdoc +We have updated the configuration as noted above to make the SSL configuration be consistently located. +Thank you +patch +thanks for the report +Vinod could you? The test I ran worked for me +Thanks for the patch. +Ive emailed Trustin to double check but Im pretty sure this is only in and will take a while to get released + +The fix here is to unfortunately stop using sendfile for the time being +Fixed by Maksim +Would be great if you could work up a test that can get it into this tangle +on the latest have just committed this to trunk and +In some cases such as a client that connects but does not read at all it will take three heartbeats before the connection is cleared +Committed revision +Ive missed your comments with new patch at all David sorryIll check the modeline because I do hate the comment fix +Thoughts? +To avoid this I changed the tests so that each testrun uses a separate path in the repositoryThere still seems to be a problem though that could be related to a deadlock +current buildcheckoutprepare buildWhat do you think +So out of the box would work with a stable version of Apache Hbase +This ended up being resolved as part of I meant +This feels like the same thing +updated readmesexamples readme updated with announce of GAE buildshowcase readme updated with instructions related to GAE updated libraries list and components list +Andrus hi +Seethe attachment in comment +prior to generating release notes +Resolveded with revision of buildersriftsaw buildersriftsawCommit message Use a fake installation root so that the extra configuration changes can be picked up in the SOA build +The patch file to be applied to the files below A patch file in Git format +Im apologize if this was misinterpreted this as a I didnt get an answer on the forum but I genuinely assumed that this was a bug since there were no replies +The page is up +Patrick Im passing this off to Krista whos seen the same exception with another repro if you feel like verifying that its fixed in your case too that would be swellI verified this fix with my applications and Im getting the behaviour I expect +I run through the tests +Nice guess but no +If something else will appropriately note progress while the is spitting back results then perhaps this isnt an issue for us +Using contrib as component for now but we should create a separate Handler component and move there +Patch committed to to trunk +So with sane builds causing an Assert which will dump info about the current threads to the console and also identify which Thread is hitting the exceptionWith insane rethrow the exception and in the case of an IBM jvm also cause a java dump to be generated which will also contain a list of the current threadsPatch ready for review +sigh +Straight Forward Patch + was a firefox bug with SVG images that have floating point dimensions ie pixels The programmers guide was still pointing at an old version of the ve updated the book to use the correct image +Attila this looks like a server side issue not a jboss tools issue +No its not possible to deploy a war to multiple remote servers because of not yet fixed +Ill commit this tomorrow if no one objects +The user library is now created when its not present. +Im currently working on an favicon for the site as well +The changes are a bit complex for maintenance +Want me to commit this and try it DD? Want me to leave the issue open after commit so we can see how it does over a few builds up on jenkins + +Being able to include the config in the pom itself would be great although this would most likely bloat the pom. +On that noteThis test originally did something unique in the old test harness which always kicked off a separate process for networkserverNow in the junit framework were always starting networkserver within the same jvm so what is the use of this test +resolved by reverting to the old signature +BTW you can see that it didnt duplicate file in the root of the jar and used the correct one from project resources which is good but strange. +But I dont see it being used anywhere in other parts of the patch +Giving it back as you have the code already +Here is the patch for apply the patch +bulk close of resolved issues. +Why not just check uarawbufferreader? It should be set iff the last clauses are set +updated patch after removing a few unused methods +Fixed by Wolfgang Laune +deprecations listed in issue description have been removed in classes should also be removed + +Removed fix version please let the dev team assign the fix versions +That is troublingI agree that we should not commit this until is ready to commitI suspect that putting it into HDFS is harder than it is worth since HDFS doesnt have a native component already +modulesorgapachexercesmain is not privateis actually from slfj +Hi HadrianYes I believe so +Correspondingly Ive added an apply flag to our Portlet which can be set to false to apply any given to the Portlet action phase said the general recommendation remains to use the API for interception needs in particular to differentiate between the action and the render phase +Hi AaronThank you very much for your reply +change status from ASSIGNED to NEW for consistency +rob confirmed and in JBDS Beta + +Pls let me know if I can help +Ill commit the same changes to trunk after the code freeze is lifted for beta +i would say not needed right now +Ikumasa do you have a patch for this +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Thank you for the explanation +Receive this return from Return Header +After the release of well going to dive deep in the trunk development which will be a complete build overhaulSite and subproject documentation also will be tacked then +So we have a dependency from support to util +Fix verified in +Agree and for the rebased version of the patch +Sending this one over to you now +Uri What do we do with this? Should we move it out to for further work +Works now with a single thread +We dont define a fix version until a fix is committed +maven +reopening the bug +I added a comment in that area of the code before committing. +Completed and verified for GA +Will look into this issue as part of +updatedTo avoid the number of threads is limited +Or did I miss the point +Issues resolved in +I renamed a couple of tests to reflect what they really do and found that compressed tests fail once this patch is applied +I guess Ill have to upgrade thenClose this bug if you want +Did you mean to refer to something else +Thanks for the replyThank YouFernando Vega +where we can download jar file with this issue fixed +Postponing to to unblock work on release +Ok fixed +Screenshot of error from admin into the master and branches. +Thanks +Fix applied along with new tests for this case. +More specificallyIf there is a call to a service method proxied by Hibernate transaction manager in a situation as follows save Object A store STATUS that save was sucessful save Object B transaction ends and flush occurs When flush occurs A and B get written to database but the STATUS was already stored so if A now fails STATUS will be incorrect +HTTP support should allow for multiple contexts per port +This is no bug your URL is invalid! See use list archives this was anwered multiple times +It has one test case that can be executed while the application is deployed and running on a tomcat container +Here is the output generated by parsing the same file with the standard SAX parser in Java +changing the ram in solrconfig didnt seem to help +This is to help in releasing unused resources and reclaim them +my final commentreal testing should be done using a payment gateway sandbox +Users found it difficult to manage ship group information using the existing forms +I have enhancement working with no errors or warnings now +Hi BillAre you working on the axiom test branch for this issue? If not please attach a diff to this issue so that we could do some testing tooRegards working on the routine to copy a node tree I found it needed the same worker routines as did axiomnodedetach to manage namespaces declared in a parent of the subtree being copieddetached +applied the patch atRevision Author dantranDate AM Tuesday August Messageset the default izpack install home to staging areaModified trunksrcmainjavaorgizpackmojo test t the snaphots once confirm i will release alpha this Markus +Any further work should be done in new issues +Thanks Kan +So the in bundle A in a way that is an extension to Smooks and could be attached to the bundle as a fragment bundle +Important thought that jbosstools does not require this information to be present nor correct though +Could you post an updated patch and Ill try to look quickly before it gets out of date +I need to clean it up and make it a bit more generic +Assume this should be applicable for so tagging as such +Stay tuned +I can definitely add it to the framework so that its reported in the logs even if its not guaranteed to return the PID +Uploading which should solves the issue +The job works and outputs the correct sum +Any update why this functionality has been removed +because you are using a build that doesnt have the fix yet +Users guide updated thanks for bringing this to our attention +Fixed in rev on trunk. +In this way the test shows that the declarations not bounded in body are automatically copied from the envelope to the attach my version of the testcase +So the value of setOutput has no influence +That is fine for most files but if it is a media file that has an entry in the content store then just deleting the file leaves a black hole in the Gallery app +Im not able to apply your latest patch all files under standart folder are considered updates not new filesIm using eclipse apply patch feature is there special step Im missing +I cant effectively do that in my overridden class unless unless that variable is protected +log file except the +It pains me to see this ticket open for so long +Beans which are added via addBean arent defined as managed beans from A managed bean is a bean that is implemented by a Java class +So this is the best we can do for M and we need to find a better solution for M +Hmmm it appears that the has a shutdown hook as well +I believe the resolution is to move the archetypes into the Create Project section of CentralCreate Project would then contain Java EE WAR Java EE EAR HTML Project Dynamic Web Project potentially EE with Projectand Java EE Quickstarts would disappear as a category under Project Examples +Default server configurations has changed in CR and the perf results can not be compared with the default config in +Why its not a bug? Suggestion Box disappeares when i scroll in horizontal or vertical directions inside suggestion Box +You need to open AdminConfiguration set conf with key to tested and this works +Correction this is the good one +The XML Schema document +I also attached my including the relevant lines and a screenshot which shows the created file when running the Java program +This bug has been fixed some months we stil have problems in this area we have other defects to cover it +The reason is get just call get +Patch against current trunk +The new class is never usedBest Christian Ill update the test cases and submit a patch in another ticket to fix it +Some project do some dont +Thanks for your attention in the details +The results seem to get lost in the outputIve attached a diff showing the changes I made from Revision I am new to all of this so if I am off in the weeds on my solution or not following the correct process for posting it let me know +This looks good +Since other valid may hit the same issue we will remove the restriction in the wizard and allow the user to continue with errors with a warning that results may be unpredictable +Closing this one as well. +So I would say that is likely to be good +Looks OK except that would recommend using string constants rather than having copies of the string in both the code and tests that makes the tests too ll update patch soon +Full exception stack trace +Fixed the buggy test test +Just committed the patch +Same result for Microsoft SQL Server SP Express we meet database deadlock on JCRMVALUE table +This patch is a major refactoring that opens way for other implementations of stored fields and postings pruning +Oh um also . +Your mirror was removed because it failed to do an update in over two weeks +have committed this to and above. +This really needs to be improved to make scripting better to useIntrospection may be ok for some script languages but its not very natural to others +To route Logging to slfj you must be running logback the slfj backend only exists as a means to access logback that is all generalized over slfj is not supported and is not intended to ever be supported +Project completed +txtThen it uses to scan through the filesDuring opening the ZIP file to function to resolve base URI is called and as of generating the wrong base URI as described above the file is not added to the ZIP container during initHope this helps. +It works as expected in trunk but i cant seem to find the issue with +It replaces the asserts with exceptions +I think it fits in nicely with the style of the other icons using similarly sized circles and lines +Burr what version of Seam are you trying to deploy +Thanks atm removed this import +We do use Jolokia now for connecting to a fabric +Fixed in Rev +That is quite a big fix +Certainly! Thanks for pinging this too +I didnt want make you do another change so I made a few modifications myself to see how it looks for you Rakesh +The workaround is to dragdrop the Task from the palette +introduced this new API and I recently work on a stronger semantics of recoverLease at +Next Ill do the same for +This is an interesting new feature +Because no matter how many concurrent updates with optimistic locking only one will succeed +It should allow for simple mechanisms such as using or more complex user written configuration +uops +Arun does the explanations from Scott and Matei make sense to youIf it looks good to you I would like to commit it +to identify which ones ought to be for your hint could you create a property to true or false +Thanks Sean! +It does support keyboard navigation for everything except it seems the arrows are enabled based on mouseclick in the tables instead of selection changes +I plan to commit in a day or two +I have updated the patch to the current trunk versionAny comments on the proposal +has been marked as a duplicate of this bug +It updated the sourceDirectory to include srcmaingroovy which I dont want it to do because it no longer checks the test directory +I added Kristians patch to the original patch +These were likely added by Eclipse but should be removedI would prefer to understand the logic changes which are being suggested first then see a minimal patch to introduce such changes +Let me know if there is some direct download URL I can use else I will try again later +Definitely Id like to see performance metrics of view building on heaviily editted documents before committing this +What I mean is that JBM should tolerate failure conditions like connection failures and server crashes +Thanks Ben for the docu update +I eliminated a couple of unnecessary property assignments in the new classes and added the ability to use the button styles individuallyThese changes shouldnt affect any testing that has already been done +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +Verify that this worked the diff didnt apply properly +If you do ill add is just the POM with the version changed +When the concurrent merge finishes Bs delete file becomes and Cs +in +Sets the mimeBinary property so based on it we know that MTOM is applied into a given message +Hi JacquesPlease close this issue because its is fixed at rRegardsAmit Sharma +Roman thank you for the additional information and code +Some of the crazy tests that are in the Functional Tests should be added to the random walk tests +I dont know why the asterisks appears as rectangles creating a big messJust look at the output message user before instead of looking for asterisks +What sorts of default values should we have in? Access control is not in this version +Yep looks like Stomp has the same issue +now were down to type name only as a String +Is this still a problem if you build master or the branch? I suspect this may have been due to or +The version of it seems to try and use in my case is +This is missing from the previous patch +thanks definitely add a post once you have had a chance to run your tests for awhile on latest branch letting us know new failures or lack of failures +GailI sent an email to HSQL mailing list asking for some tips on this +Assigned you this issue +Complete regression suite was clean for +Thanks it doesnt affect correctness +Hi Robert I presume that when you say we should instead improve standard analyzer you mean the code should work more like the original Javadoc states it should? Or are you suggesting that moving to Jflex The problem I observed was that the current rules dont implement what the Javadoc says is the behavior of the tokenizer +just does not get you have them check the latest nightly before reporting more issues? +Closing all resolved issues due to a successful release +How did you rebuilt system keyspace +A patch implementing option B +Offsite Feb +Another thing to think about how should the injection of helpers into look +It is usual solution +You can set no in to eliminate the need for +well reopen if users say its still broken despite the charset fix +Exactly how that is done will depend on further information from Apache Legal or from the httpclient team +Thanks Ivo +Or does it? Anyway I filed to clarify this +Might be related to but I doesnt look like be related does not fix the problem I am using can see that Arnaud is the assignee so he might be working on it but Ian can you provide details on what you expect to see happenYou have defined resources in your pom so they need to be in the generated eclipse files somewhere or else your eclipse project probably doesnt you can manually configure eclipse how you expect it to work and then attach those files that would be helpful +So in trying to define the requirements for this issue what are we shooting for? Given a PGT the each key will correspond to a different instance of the same target type? Or will we try to support different target types per key +emmanuel i bumped into this in other places too +Personally I was nominating contributors +Defer all open issues to +So I made it a simple class +fixes the latest issue with an extra quote char is fixed +Thanks +Ive commited at r please verify +Just complementing my previous comment I dont know of a better way to implement this that wouldnt be either overly complicated or hard to configure and understand +Catch them and try to continue +This will dramatically improve the response time for obtaining nodes with binary values +Tell us where or the report of pretty much useless +We now report gif on our capabilities and things seem to go along when you call it but instead of an image I getThe image httplocalhostgeoserverwmsbboxstylespopulationFormatimagegifrequestlayersstateswidthheightsrsEPSG cannot be displayed because it contains ideas +It is currently being built by ant into dist +I will give it a try to see if this fixes this bug but it I think it will not be before MondayThanks for your is related to the work that you are currently doing +OK +The fix mentioned previously seems to be working correctly so closing this issue +Eventually we would want to enable users to plug in their codecs into the to test it against all our testcases +hi Jody the patch looks fine to me +Updated a patch to handle in the handler +Hem +Committed to branch and trunk +Thanks guys +I just committed this +Thanks get this fixed shortyJake +Sorry that the library didnt report an error properly +Patch fixing to not again using the new method to roll back the last N bytes +Unfortunately the error node in the view is not showing so Ill open a bug for that +Ah right yeah sorry I missed the array part ok so yes a single string it isI think the default for additive should be true not false as this is the default logback and level are properties of the Logger instances and not of appenders +If you read other comments here an in the mailing archives you might agree with me that there a plenty of users complaining that they are not able to use Continuum because they cannot add pom URLS as project definitions +FWIW I fixed this hopefully for good in fd on the branch and afa on masterThe issue was due to incorrect usage of with file paths containing spaces that caused either or paths +issue in +Its probably better to focus on getting the correct info from there then adding special cases for returning null here +Florian are you working on this +Thanks a lot! +It was time consuming +Ive got it fixed locally +settingsIt should be a well formed xmlFor Tools Version just remove it +Thanks Scott! +I know of at least one person who puts the status icons on other static webpages +It cause skipping it when running whole test suite from stabilizing test suite in IE. +This helps us to explain away not actually doing a full review +Now there are synchronous and asynchronous methods to the entire workspace and to a subgraph below some path to a maximum tests were added to verify the methods work +The issue was a mapping problem +is jars the right name? libs? You have a patch Mr Taylor +Burr updates from branding +Is this issue still open? The associated forum thread seems to indicate that the issue has been solvedIf its still open could you please attach an example of portlet that does not work as expected +Fixed +Thanks Kihwal +That problem is still there in and trunk +reopen this issue to edit Fixed Version. +This shows how it isnt working in Developer Studio +We do not need MR api changes in this case +MauriceYou are correct +Either through hbm files or through annotations +Making changes and uploading patch again +And please let us know if it works for you +Final patch for review +That would be nice Andy +TODOS removed in revision Plugin made optional in revision can be disabled by configurationThe use case for disabling the plugin is when background jobs are operations our plugin is good enough for background jobs but for end users its good to have a UI thats integrated with the rest of your systems UIThat was the whole idea behind the service to be able to create customized consoles in which case it can be good to disable the webconsole plugin to avoid confusion +Fixed in revision +Smooks editor move Input page bits to Process page when Input task this additional changes are needed because the Graph pane of the process tab is already described in Process tab Remove the palette from Smooks editor +The document also include some test results +Verified in the release notes in RC. +We go for +Quick look is good to me +meaning not to leave the values provided by char +Fix attached to +As so many people are watching this issue but the last comment is pre maybe so of them could test this with our new release and inform all if this is still a general problem embedding Groovy into an application deployed in BEA Server SP or SP appears to be working ok in the release +as r +See also in version which is now available on Maven Central +I dont understand why we cant create empty directories +snapshot +Andrea you are absolutely right its not necessary to use insetad of to read property values from for in support +Committed the fix +scheduling for the eclipse plugin already preserves natures and builders in +buildgriddevbin +Closing. +This could possibly be added to the AM REST if its not already there +The Struts site is ready for migration could you make a switch and replace the current site with the version +If it is really a bug is it possible for the community to publish a new? Or is there a workaround that I could do for it? Thanks +Kindly attach the broker and client logs with debug turned on and whatever configuration you used for the broker and client +No need for a new issue I just changed the summary on this one +HiThe Click source code is in a position to be imported to ApacheIll run a svnadmin dump and gzip the result to my account atWe will still maintain a separate branch at for bug fixes only so can someone from Infra please recommend when we should have the svn dump file ready for importkind regardsbob +The original request has been closed with a wont fixThe only thing I am asking for is a reason why this issue has been resolved as wont fix +work in progress +Charles example is now working +The synch +Hi Fabianplease wait with the refactoring until I have the branch as it will cause a lot of changes and about new engines +Perhaps we need boolean keepSmaller somewhere so we can explicitly control the behaviourIm not sure +Dan it seems good +call it RADIX +of Darmstadt and F +I guess they are used to get table nameschemacatalog information into the annotations +If Resin adds a new jar in the future its going to break cargo +Issue no longer occurs when using a newer portlet seam demo from portlet bridge does not have this problem while the originally attached does. +Elena couldnt reproduce the problem and the reporter never responded with a test case or any other comments. +Please try with a nighly build +The test failures were not related to this patchIntegrated to TRUNKThanks for the patch LiyinThanks for the review Karthick Prakash and Mikhail. +The code is nearly done +Time to validateclose this +mmm right its ignored +still held open for a period of time the deletes will eventually work +Are you sure it was the latest trunk? I fixed this error in r I hope +Hints for all preferences provided. +Can you give an example why some any application would handle this exceptionWell for example the currently catches in many places so that it can write to its audit log on access control failures +Hey Snjezana youre always willing to take on a quick task and this ones been bugging me for a long time +catch nested in try +Reverted because of +Patches would be to unless patch +At the end we left the parser in the parser packageMoved Descriptor stuff from main package to the descriptor packageRenamed the stream package to ioAdded an for the javadocs to point developers to the parser and the message classes +Ive my code and the problem has now gone away +Sun Studio installed +Costinhave you noticed that in my example is in the same bundle as and only Car is in a different bundle? How should the test bundles classloader be able to load without being able to load Car +Ill work on getting all the tests to pass + +Sorry about that +Have been pointed at a better and more robust implementation +Shuffling the first time is very important +Any volonters +In order to capture the full thread I will post each individual message four in totalDavidI just committed part of this patch to trunk but made some major modifications to it +Javascript forms of validation depend on it so if you are not properly including it you will have future problems +Moving to +Upated Informations section and removed from my files ve just tried another excel file of that list its another Excel one +Verified with JBT. +Can you check if it is ok for you? Thank +Yes Rob +The current Geronimo build has the same problem +Stack it possible that a file referenced by your table has been deletedYou can search for missing files with +Any updateprogress on Budget functionality implementation +Im envisioning books like the line that associate all of thier bookstechnologies with different animals +Dhruba this means that there is something else running on the same port +Verified +The question on my mind is is it important that new components are added to the composite to service new jobs in parallel or is it instead a matter of ensuring that sufficient instances of an existing component are available to service tasks? answer is yes +Verified on JBDS +It includes the standard Traffic Server copyright at the top If thats not right then let me know +S has a stated requirement of Java Servlet JSPYou may be running into a JSP support issue +I found a usecase that produces the problem +This will give us more time to better engineer a solution that we will be happy to maintain going forward +srcavro would also make sense for the genavro file imoAgreed putting the private protocol definition in srcavro leaving public protocols in interface and generating code for all protocols into src would make sense +Trunkjobvpe +That would be greatAt the moment the tests are run from see and +Patch applied at r please verify +The changes look good to meCommitted revision +This is a critical problem +Once you fix this issue please apply this patch here and to trunks +Technically this was fixed in. +Right these are good additionsWe are definitely looking into supporting as many injections as possible instead of using factories etc +Closing all resolved issues due to a successful release +Attached a new patch which is rebased +Reduced to Critical as this need not block the release since we currently do not link to any plugins with incompatible licenses +Please note this affects both and CP releases +I cant find any callers for +Looks good to me +Patch is ready for review +now support both possibilities in cvs +I agree with Martin +Patch v Todd +id also consider the possibility of keeping the link to the papers page in the General not in the Documentation group on the sitemap +I agree logging the url about to be downloaded should be done in verbose mode +We have moved out the timeframe for fixing this bug +Verified in r +Nit it doesnt seem we use the row mutations saved in hints so maybe we could use a simple rather than a full concurrent mapBut otherwise patch looks good . +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I figure clients should be able to provide a configurable policy to theirI think we also still need the isCached flag for scheduling +API changes can be done as soon asThis change will need a change in too +Supun Your fix makes sense to me +Trivial patch to fix Lucene library path in Eclipse +and return without rollback the previously committed operationsIf this limitation cannot be lifted with reasonable effort I can resolve this JIRA. +Yes it would be one of those features that one would have to use with care +Pedro I fixed the deps with new nonoss build check newer build from Likitha Pedro is unlikely to fix this may be please fix this if possible +Thanks +We renamed to in the release as well as added and a new +And this issue had both and and thus matched the filter +Marking it as closed. +However both algorithms provide a splitting algorithm but do not provide an easy way to normalize your row into this keyspace +No problems to report +and refactored into the Apache Felix domain in Rev +for example if the clean database decorator failed to cleanup the database it could report the failure and then blow away the database +I have tested on Internet Explorer and Firefox and now seems to be okThanks JacquesIm sorry but Im not explain to you very well you have only to apply the patchCould you please try itThanksMarco +Its just more stuff to maintain +Robert has added a service that should address your needs. +But its purely an optimization and on the riskier side as far as these things go +Thanks for the contribution Prasad +Also Im now added the endorsed directory to our tomcat with the spec jarhttplocalhostjuddivserviceswsdlseems to bring up the correct wsdl +Fixed on now porting to Simone interested in trying this out on the ND branch? If so the commit occurred at revision +Which builder are you actually using +Resolving this issue as wont fix may reopenI had the same problem but could solve it by using my own password type +The second complaint is covered by this issue. +I just tried this with Groovy and can no longer reproduce the problem +Therefore the unit tests unload all models between tests and only run +This would also impact fsck as it goes through the same code pathDayrn i will take a look and see what can be done to add the ability to write a test for it +So all issues not directly assigned to an AS release are being closed +On the other hand inline enclosures have only been added recently to +Attaching an attempt at mocking up this template +Closed. +Also along with Override added which interface the method is being overridden from +WOW +Closing all resolved tickets from or older +Integration tests are working as expected on and +Applied patchthanks one additional historical note I was gathering some of the memory usage statistics before and after this patch was created +Merged fix to with latest build fromWhen editing domainaccountuser from a region that is not the owner region we do not see the issue mentioned in the bug anymoreClosing this issue. +java ij version ij connect jdbcderbycdellaterdbTerritorycreatetrueterritoryxxzzcollationTERRITORYBASEDERROR XJ Failed to create database cdellaterdbTerritory see the next exception for detailsERROR XBM Collator support not available from the JVM for the databases locale xxZZij exit +. +Emmanuel +Why doesnt that workA few nits with the code It should probably be named net insead of util +I had to exclude one of the activemq deps that was pulling in this bad dep. +If there are no further comments on this issue I will close it shortly +I could really use this +Committed patch as svn resolving + for going with the proposed solution +I may have been unclear sorry +It looks like you attached the modified file instead of a patch +Avoid reading request inputstream in BEGIN event +bq +I added a comment on phabricator +I have commented the section on State Transfer out of the Cache User Guide for both EWP and EAP +Seems odd to say this is FIXED given that the subject of this issue is for WSDL +Do you have a dual monitor setupI noticed the settings arent always properly remembered all the time and I thought it was in the case of a dual monitor it your case too +Bug fix for release includes fix for fix for release is attached to this JIRA issue named includes fix for release +Discovered a minor issue with the patch attaching an updated version +Whereas the fix I submitted is alright as long as there are no null keys or values +i will do this +Simon could you crosscheck against trunk +Added a test +bulk defer to defer to +For example if the home directory was varhomefoo the code would fail +Scanners should work on all platforms +No answer here +Excellent thanks. +Unsetting Fix Version for unassigned issues. +Attached untested fix for this issue +Attaching rebased patch +Lars Ted I will commit +I have tested it by deploying it as an ear on trunk +Normal distribution case had a bug +I am not able to apply this patch using git apply +thanks +Thought I made a ticket for this but closest I can find is the comment in +Hi ErwanIt works well with IE maybe you can recommend to update IE +bulk defer to +Yes it only applies to KVM host running RHELCentos +Quick fix about pom dependencies +Ive certainly seen the frameworks problem +the form layout was wrong and so the details form was not shown correctly. +Please find the Sample and the integration test from the attachment +to after fixing fix version +Patch looks good +Could you include also the changes in the patch +If we have not seen it since then we can close this issue +The patch was created against but it looks like the filtering code hasnt changed a lot in so it could easily be applied there as well +We need to create on at since is not a stable location +Added to the Derby DOAP file at subversion revision +another task for the new +Changed to Major this is not blocker for the project as is +cancel this patch to add new patch to support support for did more performance tests +Lets reduce the amount of INFO logging +I have resolved this issue and patch attachedThanks and RegardsPranay I dont have commit access to framework componentTransfering the assignment to Anil Patelt it worse to hide the error in the logThe error definitely should be shown to the user +The simple solution would be to package in each web application +you can use the maven plugin from source to get you most of the way there +opens up a can of worms +Now you can successfully deploy can duplicate this issue using version and meclipse plugin version loaded from the meclipse update site. +I think we should just keep the RF icon with the yellow star overlay instead! is much more readable +New patch w the changesI also moved it to and renamed toI think the tool is ready +its just that if i understand it correctly it will change the way component references are resolved +we might want to pick a different default since oozie console also uses port my last comment the default is actually in oozie +Sorry I have changed the status to resolved without confirmation drom theXerces team +If a reproducable example is attached we can probably try to squeeze this into the track +we could split the import job into parts where we import the database to hdfs first and then import from hdfs into the specific hive database that you need +Thanks ChatreeYour patch in trunk at revision Tested with +Looks good +I think this is likely to be an improper optimization of at here in XRX +The patch seems ok now +Then move rosetta directory out of core and to same level as core +Let me know if a spec is required for this I looked through the existing packaging specs but couldnt find anything that tested the existing packaging task + to trunk +bq +While Ive read up on LZW Im nowhere as deep into the implementation as you areIn addition I must admit this is not my top priority right now so I cant promise to have much time to help out +Change the name of splitTime and splitSize to better reflect that they are describing splitting hlogs and not splitting regions + +Max I always thought this is a feature but not a bug +It doesnt work neither on JBDS nor on JBT +Failed to read artifact descriptor forxmltoolingjar Failure to findpom this is also because of it should be fixed in since the opensaml dependencies are now excluded because we dont support opensaml for EAP +Firefox Firefox and Safari and dont give me any issues +This file I attached is basically something I use in leu of a proper Axis POM and so I needed a groupId of my own +This works just fine in BTW +Its unrelated to how HDFS does its access control +Thanks Vinod! +I simply logged the blacklist in different files on the path from AM to RM +dialogsnotifications and everything passed the tests I ran. +Empty result is checked later after values loaded from joinedHeap scanners +The gray portions of the document remain unparsed while the darker black text is extracted fine +Even if the dist cache is not supposed to handle directories it is being used on directories in +you should fix the time in your +Nonsense +Rebased patch to trunk +Is this part of another feature? As you know Im a little leery of adding code that isnt used anywhere but I assume there is a reason behind this +Verified in JBT nightly +I think we can close this one. +My stress tester hits it quite regularly +patch appliedthe patch is not enough to fix for FooBarBaz things unless someone do it rare use case +The Apply Template task doesnt have warningserrors added at all +The key caching probably has the same issue as this which may speed it up if fixed +When and if other issues get merged then we can turn the test back on +I think the intention is to sort by hidden element title and filter based on text data of the cell +That would be a if Id actually had time to follow and do something about it myself +updated patch without Yale license version to as a number of is available from Nexus so if you can test against this version and confirm that your required changes have been applied ok as I made some minor changes +Unfortunately does not pass it seems that get returns something when there are not ledgers which makes test fail +we can show a that no album is found in the repository and redirect to admin page so that user can create album and upload photos +fadvise approach showed a increase TLB flushing and a decrease in context switchingvs stock or vs the direct io approach +Also this does seem specific to as Statement would not have any sql text or parameters directly associated with itI too think it is good to understand what the exact attack scenarios would be +Can we do what we do for HDFS and MR? That is have separate install and configure scripts but share the +All seems ok +Patch similar to Uwe for taking care of this! +I have started to look at the java source code for the stuff inan effort to work out what is going on here +So some of the acegi features cannot be used +Because it uses multiple threads there is currently the possibility that a peer sends inconsistent information to other peers Adds a unit test to based on the one Ben wrote Finally because of the patch of jira it touch some other parts like to comply with my changes +Version has been releasedMoving all related issues to the next version. +Cause from tagsoup +committed to and to. +Unfortunately we just blowout with an error and we do not recover the bytesFortunately it looks like you can recover the log if you make a copy of it and move it into place +Fix committed + +Making Henry Z +Same problems occurs with other libraries I have in local with a custom pom +Ive committed this. +I could be missing something here in the way the pooled connection factory resources are managed with Spring so let me know if there is somethingsomewhere else I should looking +Should we make this a part of +In this case subclassing causes us no harm so we should allow it +ill polish and commit soon barring objections +Mark both newcomer +I am curious how Enis patch above is functionally different? Is there any case where rootDir has non null parent +Ill put my substantive comments on +I checked revision of and the problem is fixed +Janne I can understand your reluctance to offer information about the configuration of the system when it concerns the general principleHowever the Interwiki links are a special case +Fixed +Thanks for your efforts +This wont actually pass mvn test because we arent including the test categories +Agree the current separation is a big but I cant imagine many users wanting to do that + tracks the implementation of the decision made in to deprecate snapshot mode +Once we fixed that we can no longer recreate the issue in CMS +Ive just committed this +its beeing worked on at +batch transition resolvedfixed bugs to closedfixed +The priority for this issue should be changed to minor +The library comes with a number of unit and integration tests recommend you see if you can add test scenarios there that reproduce these sorts of issues +Yes +defer to +IT added +Changes have been checked in in revision which wrongly has in the commit message +No harm rightAny other comments or is this committable +EMT isnt a target platform for performance tasks now but in future it will at r thank you. +It wasnt possible to fix the related issue in Tomcat so I have fixed it in poolThe issue was that the context class loader for the would be set to whatever class laoder happened to be in use when the Thread was created +Ive just committed this +Your participation will be very helpful +Hmm +Yes it seems to better respect the depndency tree +Any comments on this one team +Its used by too many plugins to be folded into one plugin. +Ciao Fabio +If you could point me in the right direction I can certainly take a stab at itThanks move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +As implemented is very specific and limited to the default service scope. +git mirror of giraph seems to be strange +bq +WDYT Olivier +Committed for in rev +Closed with the release of. +this is actually a bug +Therefore the total number of supersteps performed by all the tasks is the sameFor +Committed revision . +slip to next version any updates +If you want an other XML node name you can specify it on the mapping +I cant reproduce the problem at least when restarting a runtime workspace to testdebug +This new patch includes the csv style metaheader with numFoundmaxScorestart +Resolved +Applied patch with thanks to Piotr. +The dependency list requires no additional mention in the NOTICE and LICENSE files as far as I can tell +It cannot work with M since every plugin is loaded once and once only +The patch was committed so is there anything left to do on this issue +I made a patch that updates the test to use a clone of +Included is a simpler patch that does the modification suggested by previous comments +updated to close temporary excessively slow assert statement +Which means +patch to not recalculate cache capacity if its been manually looks good to me +The answer to this issue was not related to the real issue +Attaching diff of changes +This is fixed in SCR trunk +Yes there was discussion of it on the dev mailing list awhile ago +highwatermark file by the checkpointing thread +Its more to the bottom see the attached screenshot +Still testing some additional code changesRegards release will be built soon so hurry up with your testing +Thanks Tom! +Seems to be ready for a commit +I made it a subtask ofI just tried it but I dont get the default help page in French what am I missing? Should I do something special after patchging? no disappointed I get the same after the +Verified by Vladimir. +The underlying cause is not correcting the order by above the union +You could deprecate it for now +I think this error is consistent with the language design choice we made several years back that we werent going to support overloading based solely on method guards +I did several tests no problem now Can you apply it to karaf branch ? Thanks in advance +Graham is already looking into this it should be assigned to him when he gets karmaNote also that this JIRA is releated to and possibly a duplicate of which Arthur is already working on +has never been supported and the main reason for that is was that it is more difficult to assume which class loader is used +webperform eclipseThe images are still valid same thing happens using direct entity mapping +The helloworldws war now includes all the helloworld dependancy jars which doesnt seem right +Reply To Type etc +Correction Both services seem to be published to both contextsSo maybe the contexts share the same cxf bus +For example you could create an app without any web artefacts at all +The Partners file will contain the partner link type whose name is with a role name +Im going to call this done +Thank you comments overlapped +You write one piece of code that reads the message line by line and feeds it to the parser +Why we force user to download this artifact from another place +Defined interface in Sling API and implementation for node based resources in Rev +I tried it with a simple project that depends on a SNAPSHOT version of a separate parent and the release plugin would not allow me to release due to the snapshot dependency +My proposal is that we get rid of the NUTCHlib directory in its entirety I wasnt so much as commenting on the plugin jars though now that you bring it up maybe we could integrate ivy in them as well +Using client excludes the test from running on j are we not expecting to run client testing with j at all +In extended test setup stop and start broker between tests to provide better isolation +Find the location of the tablet at the time of the lossWere you looking at the logs for the timeframe to ? This is the timeframe in which the pointer to the missing the data was written +Mike can you describe the difference between DWs deletes and deletesFlushed? I have some idea however maybe some clarification will help +need to fix this otherwise bugs might be to Prasad +committed w addition of hex for discovered that bytes doesnt work via cluster tool getendpoints +Probably as youve suggested we remove providing the installation path via s call it known limitation of the Cygwin build +For example this forces me to include the command line options that I use with a particular as an xml comment to be able to reproduce the same should opt for a configfile with as little command line options as why do we need cp and classpath what does define do why are ffeatures not supported what is genwsdl needed for what does import dofeatures donotoverride is useful do we even generate stubs? otherwise what does nodatabinding do? what is nomultirefs is duplicated what is serializeinterfaces strict vs +Added check and setting correct exception RAISE state to execute release monitorThis patch also contains several fixes in parts of code unrelated to but necessary for correct prediction of the next bytecode for invokevirtual instruction + currently serves multiple purposes represents a route externally provided routing configuration lookup key in The final purpose needs to be determined firstThis issue assumes that the only purpose will be +so with a spring jar springorm can resolve the dependency and can create an entity manager +This works when modeshape is deployed in the brms kit +Fixed some links that still pointed to removed from the archive. +srcmainjavaorgapacheflumesourcehttp +Issue can be deferred to the next about we just add some documentation suggesting that users use UIMAHOME paths without trailing slash? Since nobodys complained yet I think that would be sufficient +overall +This could be useful in scenarios where you want to use the back end for testing but dont want to change existing code +It doesnt need to be public +In rev I changed this to only try to create one if there is none in the initial services +I dont know what couldhave caused this although I was experiencing problems with the new kernelcrashing when I was attempting to retrieve the entire mirror following a freshOS install +Attaching logs and DB is it constantly reproducible? Looks like your agent was unavailable when the code tried to apply LB rules +Charlie can you confirm that this is fixed now? looks quite different from what the patch suggests + +This patch removes the arrayCopy added in earlier patch and byte array is passed along with its for the split method +The patch seems fine +are you constructing a link to this page yourself? from the stack trace looks like this page is being accessed by a bookmarkable link this page is not bookmarkable and is not meant to be accessed that way. +Ill get a fix in in the next couple of days +I tried this latest version but the problem is still same +yes good to bump the issue with is way more complicated since there seems to be a general problem with getting convention configuration setup in JB even when class reloading is turned off +Ive set up a repository deployed snapshots of the same project then set the Repository Purge By Days Older field to and enabled the +I am very fluid in eclipse subversion and I am a maven continuum and PVCS user +thanks +Sorry for not providing steps to reproduce before this issue still just removed this contribution and pushed a new can you update and report backOdd that I never saw any complaint about this missing class + +Actually let me add it to slideshare and do it that way +Committed at subversion revision +trunk +After that it worked like a charm. +I think this is related to so that we can provide resource in YARN and achieve better resource elasticity +For the events defined now the general approach I would take is to include more information than less so it leaves options open +If you can provide a patch that mimics the code on the JS shell Id give it a serious look to add to but as I recall its a bit convoluted +HiWould you please try my patchThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +It gets quite confusing when using flexible quorums + +Try out the war and go to httplocalhostmailtest to see if this demonstrates the same +enhanced to allow use of abc type expressions +There are many instances when the same piece of data resides on multiple HDFS clusters in different data centers +is not correct in sense of mask calculation +Also filter out the packagefragment jcrroot which gets shown by JDT further common features of the content browser include standard actions on files eg open properties of jcr nodes should be visible in the properties view and editable using appropriate cell was thinking of add new property and rename property name as something essential for after that I think we can close this and do +as we talked about on irc I thought it checked the properly resolved project not the model +Seems to be fixed could someone please verifyCommitted revision . +Thanks Gavin! +I can see how it could be useful in the command line for automating things but in the IDE changes that youll need to execute that goal more then once are much bigger then from the command note that we do provide Maven launch shortcuts in the project popup menu and there is to allow to configure those launch shortcuts +Assume closed as resolved and released +Ive taken the attached example JRE metadata and used it to create the system bundle exports +If Ondrej does not have time I will do it tomorrow +patch containing support for DAV good to me so farMaybe we can for the UUID parsing +I dont think it would break anything +looks fine +Committing shortly +This is not a bug +test work fine +Could someone generate a JIRA list that tracks the changes in the opcodesIt can be done by svn log +Regarding your first comment Axis does support unreferenced MIME parts We have documented the behavior that Axis does not support based MIME part addressingI agree that the APIs we have at the moment for attachment handling are suboptimal specially given that Axiomaxis MIME handling APIs were designed initially to support only MTOM +Punting +Thanks MattiasIll give your project a try as soon as artifacts are in Maven centralCheers +The mechanism used by the async indexer could be used also for ensuring serial updates of the index updates +That lets you compile hive against SThe real fix will be to update the POM file for hadoop core though +This doesnt seem to handle spaces in the file path + is now +Yakov Ive played around with the patch too and it seems good +We would need a complete test case to be sure but the most likely cause based on the snippet you included is that the button still has keyboard focus +Attaching the patch for the axis branchI ran the axis tests against this patch and all the test cases were successful +propose API extension +is changed to make use of it +Hiperlinks +It seems its even coded to do that however as per another WTP bug and partly a bug in our code we are not prepared to handle the removal of objects and our publisher fails to undeploy it properly +Lets tackle it in. +Thank you Jing for the review +Thats why other threads remain locked at this monitor +Im pretty sure thats the case for but we should verify that its actually testing something in case someone changes later +sample committed a fix for the bug to the SVN trunk +Unit test that fails prior to patch and passes after patch +Thank you so much +Junit test class is also included can someone please review the code +Line of is missing the at the front of the variableLine will not work correctly +Im not sure why the patch didnt apply but it looks good to me +Short answerYou are using in some way in your software +Patch adds close method for +It would be gread to reduce this somehow but Im in lack of ideas how to achieve this +Heres the latest patch that includes a compareTo for compatible unions +Ive attached a patch for this issue +Bulk closing issues resolved in JEXL. +I honestly hope that by running from we can take the execution time down. +We should minimize contents of that file and as above ensure that all properties defined in are correctly represented in code too +I just committed this +In a servlet environment it would still be valid for the user to specify a servlet context listener in their would it not? In that case the problem still existsI also have to say that the more I think about this the less comfortable I am in having the initializer run automagically in a servlet container +With tcp there is always an outstanding read that will fail +sorry meant to create on OOZIE not MAPREDUCE +Committed the b version of the refactoring patch at svn +Current implementation with dynamic imports work fine +Logged In YES useridCheck the updated +The WST tool has done it +As Snjezanas the expert here on the need for old versions of examples Ill defer to her and close my own issue here as Rejected. +it s no bug +Are you running on NFS? If so see +Typical principal configurations would make this look like userhdfsDOMAIN +The doc issue that you highlight here is already raised and being discussed on the dev mailing listTo verify the fix per se you can use the snapshots that are available +I add another the output file in frequentpatterns folder a sequencefile or notBecausei want to save it as a table in Hive +eyebrowse is being disabled. +I dont know of a way Flex can provide a solution for your needs at this time but we are open to ideas and you are welcome to contribute code to the effortI believe some people have been able to refactor their UI so that the controls that need softkeyboard support dont scroll and I think some folks have even tried swapping skins on the flyBut AFAIK there is no easy solution at this time +Ran skewed join unit tests against this patch and random sample loader and they passed +There is only onePlease dont waste my time here if you want to waste the legal teams time thats ok +Issue may already be resolved may be invalid or may never be fixed +OK Maxim I will change the title as requested +The same for in trunk +Thank you for the review Thejas! I have committed this to trunk. +We should aim for simplicity rather than performance when things arent in a hot path or in a path +Now the test passes with DBCP andI will also update the user manual +How will you know what the compile dependencies of those projects are in order to raise those jira issues? In fact how do you know what the compile dependencies of your own projects are if you are using mavenMaven always adds transitive dependencies to the compile classpath so you have no way of working out what you actually require to compile +I guess I could read the patch to get the new name regexpath it is +Maybe we should open an umbrella issue +Closing issue since changes appear in Latest Alpha Manuals. +Vlado can you double check it on the latest nightly build +Thanks! +patches for both and nutchgoraThey simply remove all reference to broken PMD test reporting +Isnt it only distributed as a tarball or an +Please could people take a look and commit if they are ChamikaraHmm the test seems to pass cleanly in my environment +VinayI dont see any benefit running a separated thread than checking only exception happens +Sorry I have found the bugThe problem with the SFTP is that we dont have a java SFTP server we can use for automatic unit testing +Patch added +A test on our cluster showed the patch fixed the problem +Or llibertat de moviments or llibertat de maniobra too +Ah ok +As to the method name closeThreads Id rather keep it as it is since it is not a newly introduced method and there is already a method called closeInternal +Applied patch +Good observation about the lack of transitivity but would the potential solution you have in mind resolve that? Can you explain what you have in mind or if its easier just write some code + +It is closely related to the perticular ognl expression used for double +Wrong component +Claus sure ill work on it this week. +In the case of triggers this is a +It looks like patch is committed resolving this issue. +It will never get called and I documented it as such +Tested it on a cluster +But I wont cry if we close it as wont fix +tag for is added but it still generate an exception states failed to resolve channel name and message is not routed to the default channel specified +Reopening to apply a checkstyle exception for the method which has parameters +to after addressing the minor issues. +Thanks you attach the patch to the JIRA and mark it Patch Available if its ready for review +Thanks and sorry for all the fuzz +add ant property to make the setup cleanerAlso I miss Francis comment about +Here is a simple against including Testcase solving this problem for me +Pls ignore v +Your list now matches what I figured over in only I did it w grep sort compare. +Then mark the fix versions accordinglyWasnt Branch a temporary thing for Beta +Its great to have examples of this stuff but our kitchen sink is getting pretty full and dirty +I think FORCEPASSTHROUGHBUILDER is consistent with other constant names defined in class +Thanks +I think tests should be fixed +Fixed on and trunk +Here is full patch +Thanks Amit and Nick! +The old file loading prints a deprecation warning as well + +Committed +The only thing needs to be kept in mind that set of test applications needs to be consistent from release to release +On a related note what happens if you create a managed table and modify its location to point to the storage location of another managed table? Will Hive let you do this and if so can you use this trick to delete someone elses table data +Patch for just committed this to trunk and +bq +Technologies like Maven Spring and JSPJSF do not support this and likewise I vote not to apply this unless it is an option that can be toggled +The fact is that stops depending on other components +Theres two layouts to think about the storage layout and the data layout +Currently has the even worse behavior of silently doing the completely wrong thing +Wow I almost forgot about this bug +I think for booleans there are special cases like is is is is +Spring already has a generic way of injecting beans into other beans why not reusing it +Verified in JBDS L +Corrected to use Windows style environment variable references +I have always known there were issues with this but never understood them well enough to dive in myself and think I could fix it so bear with me while I ask some questions about your fix and also the preexisting codeBEFORE YOUR CHANGE +Changing target version to not going to make wouldnt it be nice bug +I think this is an edge case left over from suspicions were correct +Ive emailed Kelvin and asked for his thoughts +Changing this to Normal Urgency +Great seems to clear up the intermittent failures +Awesome thank you for going through the process of building these +All of the fields are unused for data transfer right? All we really need is here noYou are right +Thanks Rajesh. +One commentIt might be nice to add an option that create https addresses even when auto detecting the hostOtherwise I have not looked closely enough to see how the two approaches differ +The thng is out of source files I only get the error in the that contain the Enum +I had to make some upstream changes in how our works which hasnt been officially released yet +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Thank you for the contribution Xi +Change from Bug to Improvement +Resolving this issue because it has been ported to +It replaces the with an enhanced implementation +Ill create the other one anyways later +This looks like a potential duplicate of two other issues and +New patch makes makes client constructor block until zookeeper has connected +revision +Peter I added the patch hope thats OK +Please do not use this experimental component it is a work in progress and probably will be remodeled to a full seam like conversation system if you work in seam use either seam conversations or use saveState which does the same as scope over will close this since the scope is a component in progress not being expected to work everywhere and it soon will be revamped to something else anywayI dont want to use this component +Thanks for pointing that out Richardcheers be done with the work for. +In this case when one starts the creation I put a lock in memcache with CF name to be the key for example +I dont quite understand what youre suggesting with this +The problem with there is that the spring part of the ocm project relies on the repository util class +Contains a new file srctestapijavaorgapacheharmonytestsjavaawt +I think that warn or extended might be a better default +Do you agree +Can RobertThomas pitch inWill ping one of them to take a look at the proxy just looked at the proxy changes and I am a bit confused why they are needed +Can you take a look at this change please +The rest of looks like validWell I dont know what is wrong +Thanks a ton Ryan! +YourCase id like to hear your counter examplePardon the typing on phone will beOn computer soon +pulsing isnt setup to work with the cache yet Im punting on that for nowOK thats fine for now +However I can find no code for populating county from postal code or anything else unless users are supposed to enter their county +A way could be to try them one by one and pick the moment one fails less burden to search the reason +small enhancement made to not retain consumers that have no dispatched messages +Or jruby S irb also works +Wow +closed for. +Incorporates Konstantins comments +A new version of the patch with a minor bug fix related to complex argument being passed as a parameterThis has negative tests +We could lover the level from DEBUG TRACE +Your call +Thank you Knut I think that must be the reason +Gabriel what do you think? +in SVN +mrepositoryorgapachemavenplugins Ive solved all the next run the application automatically downloaded the last version of which is was my solution way +I will commit this if no ones objectThanks NicolasIts ok to me +funny Im getting that too +After discussing with Lars and Guillaume its better to use mock +Should apply to both trunk and +I think the profile solution is the right way to handle this +But if its One Big Cache how does modifying capacity per CF make sense +Henri I agree that this issue should be closed as wont fix +Some of my other work sneaked in here +It should just use a map from hostname String to to keep track of the lengths on each node +There may be some kind of masking going on with exceptions thrown from open but Ill need to investigate a bit to make sure +Removed call to releaseResources in finalize method +I dont believe we still have access to the tools that were used to produce the UML images and I havent found a free version that matches +Thanks Tsz Wo! +Sending srcmainjavaorgapachecamelcomponentfileSending srcmainjavaorgapachecamelcomponentfileSending srcmainjavaorgapachecamelcomponentfileDeleting srctestjavaorgapachecamelcomponentfileDeleting srctestjavaorgapachecamelcomponentfileAdding srctestjavaorgapachecamelcomponentfileAdding srctestjavaorgapachecamelcomponentfileAdding srctestjavaorgapachecamelcomponentfileSending srctestjavaorgapachecamelcomponentfileDeleting srctestjavaorgapachecamelcomponentfileSending srctestjavaorgapachecamelcomponentfileSending srctestjavaorgapachecamelcomponentfileTransmitting file dataCommitted revision +thanks! +Javadocs only state that if ALL values are the result isCases can be made for both ignoring spurious or for triggering a resultCould this be clarified +Thanks. +Since are used in box settings I assume there is some sort of behaviour but I cant tell how to trigger itFor reference I am using an ivy jar bundled inside a java project +Proposed patch +Thanks for the patch Chenghao and thanks Ted and Sergey for reviews. +Let us know if you can reproduce in +One question have you considered adding a test case that ensures that a JN which experiences this scenario will return to participating in the quorum after the next finalizenew segmentNit looks like the method comment for test got messed up a little bit +Cant figure out what I need to pass as albumId BTW but thats another issue. +I put the save statements into a finally block and extracted some code in tested the fix on trunk with featuresinstall workingfeature failingfeatureThe command show the error that the failingfeature failed to install and featureslist shows that exactly the working feature is installed +Currently the patch only validate the defined by schema +No plans to work on this any time soon but patch would be appreciated +Thanks Ill look at the review when I have some time +Screenshot is attached +Please put the contents of the file in Piers export and press generate site +Looks perfect ! +Thats weird but I can take a look +Committed to trunk for revision +for content not content withinNeed to open other issues for edit and clean up of our wiki moving what belongs in the hbase book there deprecating old content etcWill also be moving some of the javadoc content into the book but thats another JIRA +Comments welcomeI no longer believe this is a codec issue since the codec doesnt know anything about record delimiters +Thanks DeepakYour patch is in trunk at rframeworkimageswebappimages does not exist in I gave up +Manually tested! +Thanks for the help! +Filed and started work on a duplicate issue over at +Do you think a space is most appropriate or would it be better to have a more distinguishable characters like for separation +Ok +Yes also not sure why cloudbridge db is being used +Merged fix toCommitted revision . +Im not explicitly using +Im pretty sure JAI can handle coverage rotation I guess its just a matter of passing down the for the vector path rotation has never been taken into account so yeah I would expect a number of fixes to be necessary +Hi BjAs there is no functionality present so that anonymous user can verify his email address and unsubscribe from subscription so i have set it as new feature than a bug +Hmm this patch does not seem to workTested on Windows Works in FF but neither in Chrome nor IE Tested on Works in FF but not in Safari partially works in ChromeOn the systems no keyboard input is accepted at allIn Chrome on OSX keyboard input works with displaying keys as them being pressed but the backspace key actually goes back in the browser history instead of killing the previous character +I dont really see a good use case for having two different lock levels +If there is a problem please open a new jira and link this one with that. +Is this a duplicate of or should these two issues just be linked together +bq +Discussed issues with Martin we believe the best approach would be to move the integer tag before the date pattern stamp when is used to aid sorting keep it in line with when the is ve or ve and allow easy identification of the previous count values in the face of changing date patterns +Much easier for me if the patch is in the form of +REOPEN the issue is not always reproducible + This is great to have thanks +and http +management server logs from a xenserver those two lines are not exactly exception it is a way to use exception for racing condition that is why log level is INFO +See r +Hi Bassam any chance you could create a patch for all this + as i get this same issue today +Fixed Patch applied +But not all queues are working afterwards +Yep sorry must have been the transient thing! Closing out then. +Can anybody respond here please +verified on nightly because s changes had influeced CtrlSpace shortcut and content assist got broken +The example is in place but still needs to be polished and lacks documentation +namenode it would be great if you can let us know the steps to reproduce this situation +If it is a simpler issue and especially if you are the reporter then it is best to just close it right away rather than coming back to it later to close +tabs are not allowed we use spaces for indentation author is deprecated in ASF code we use other means to credit contributors +I am attaching a patch for this +on commit whether or not we wait on factoring out the hfile stuff +I think it would be great to take this fix one step further and have the javascript tag support the Scriptaculous load attribute +Gavin and I have been trying to connect but keeping missing each other +It is pretty much pointless to include additional features we are not able to adequately support +I looked at applying this but it doesnt compile +There is just two departments persisted so should return only two instances +Here is a dark header option +Its not working +I still dont appear to be able to add other users to the space +BTW append must be change to be a POST operation because it is not idempotentAgree +First JIRA for that is +Understood +Did a skim +thanks Cedric Ill review this patch Ben youre welcome let me know if you have problems with the patchRegards +of course +If you are saying an alternate bundle home is defined in the configuration for the module but is not found within that module then yes wed love to see a patch +This patch actually works albeit a bit sloppy +Bruce have you noticed any regressions? +The release audit warnings are about lack of Apache license on the html doc files that are automatically generated +There are several place where it has not been updated +All the k tests now passes successfully with RCIm going to open a ticket in the Ruby ML to see if someone else noticed this behavior beforeThanks for your support +I just checked back on the patch and I must have applied it to a later version so I saw the changes made at r as part of your change +define new protobuf messages for diff report +It wouldnt be that hard to improve this for subystems that use the view framework +No its fixed in the final release version +Fixed with close of resolved issues in. +You might want to verify that this is still a problem since is now fixed +I have a similar problemI would like to re open this issueDetails of the problemDepending on the OS browser and flash version the button start import doesnt respondI tested for several systemsWindows XP SevenIE Flash Import a document is possibleGoogle Chrome Flash Import a document is possible Mozilla Firefox Flash Does not workDebian LinuxChomium Flash Does not workIceweasel Flash Does not workGoogle Chrome Flash Import a document is possibleMacDoes not work with Safari i dont have the version Number +Hi HenrikCould you please attach the wsdl file that you used to generate codes using WSDLWS? Then i can look in to the issue +In the src distribution there is seambuild with declared dependency onjta version +Fixed on and closing all issues that have been in resolved state for months or more without any feedback or update +Now it doesnt throw a NPE +Its unlikely that a customer will try a similar scenario +See +It uses deltaspike but IMHO does not offer enough customisationThey do not use the deltaspike configuration CDI extension probably because it is another CDI extension and the interface javadoc precises cannot use this configuration inside a CDI Extension before the boot is finished! uses directly the class but do not offer a way to add a it is why the only way to add a customisation is to reference our own to make it available when is called +Yeah this was my second guess applies for all of schema operations can you try to apply before running your test +I that possible to reduce space between method declaration and documentationFor this screenshot space between getter method declaration and Returns twice bigger than it is really needed +I almost sound like Im obsessed with test cases but then I probably am +fixed in +If that still fails would you please paste the output of trying the command as mvn X tomcatrun +Thanks Konstantin! +This is tricky +It is was in because we already had that class custom in our applicationI will test it in +Of course I would also welcome if this line vanished from the Apidocs Only intended for internal use +Dear AwadallahI dont mean to be rude but please be a bit more logical +Mark yes +In case a custom location than you need to specify the correct location through JAVAOPTS +Fixed on trunk +Fixed in trunk with provided patch. +We should choose one of to hex escapes or URL encoding +Sounds like a very reasonable plan +Or perhaps we can add an annotation and doclet which adds Evolving in bold as is done for deprecations +So the UI at least for Regions will be a feature +Seems like you would want an exception for metadata tablets? Reassign those tablets immediatelybq +Simone Daniele does this ring a bell for you +Assigning to but move to if no patch shows up +as of now we no longer test the trunk with IE. +changed all constructor calls to and in either the mrunit or packages to use the shortcut new etc callsupdated instances of these constructor calls in the documentation to the shortcut methodadded shortcut constructor method to include the combiner in theExcellent work! The only comment I have is the method added to should also be added toThoughts? would agree but the currently has no support for are correct sir! I will blame it on being tired +Right an Atomic counter should then be more appropriate to track the current amount of tasks in queue +JIRA is not for support requests it is for seam bug reports +Normally if you get that the exception falls through the request will fail everything is recycled and the user gets disconnected +The problem is that describe ring will still include ranges from all datacenters +CVS subfolders as for +Resolved in revision . +And the only reason that I can tell that the depends on is for using from to load the resource Maybe we can consider to relay on thread context classloader +might have been the day beforeProbably worth trying at the lower perm gen to see if we should reduce the test anymore +Attaching you acknowledging your messagesIf you are using paging there is a maximum amount of unacked messages that can be in memory +Note the missing M +Ive committed a fix for this issue in svn trunkCould you give it a try and post your feedback herethanks! latest snapshot build has fixed the original problem +in trunk and. +We need to provide an alternate way to specify the location for reading the properties file +committed +Please send a new patch for this issue to be applied against the trunks see if its third time luckyCheersChamikara +A Simple snipet of the page flow and validator would be very has been applied with just a couple of tweaks +I just committed this +So it has dependencies on the Eclipse platform including OSGI and SWTId recommend to use the latest Eclipse SDK for viewingElse Icould need a step by step instruction how to add custom appenders during runtime +First we have to generate patch on +Does it match your usecase? It has a different approach and requires to mount a bookmarkable page +I found another misbehavior some lines above +If I start to see it again Ill post a new issue +we should probably start with upgrading to jquery since it removes support for older browsers that we dont support jquery to the bower deps +Do you have an example of the code that is using Spring JDBC where the returned key is NULL? I just ran some tests with the driver and it seems to work fine +bulk move of all unresolved issues from to +Just produced this again +Yes As part of It has handled for all the partitioner implementations but in this case we can avoid the call for the partitioner itself +Built in Brew and added to the thirdpart repository Committed revision +Assigning to Enebo +But the discard oldest is an existing task from the task queue so that is a different storyThen we need to provide this as a callback to the so it can do custom logic when the rejectedExecution is invoked +I dont have the jar file anymore to try now +Started vote on dev list as this requires Maven libraries and will break Maven support +qix file creation suddenly become unavoidable I never had time to insvestigate on why this is +Updates several icons with better images +I dont think the complexity of these algorithms justifies add more dependencies +Fernando Im assigning this to you as per our IRC youll patch JBC for RC + +So you may not need to import lot of schemasthen try to codegen and replace xmlbeans files with the generated files directly with scomp +This looks great Bryan +This issue is resolved by and. +Looking over the code I dont really even see a reason for this function to exist +if you have some kind of design doc for it that would be cool we can see if it lines up with the kind of ideas we were thinking reviewing the latest code drop from our internal team we have decided it has architectural flaws significant enough to remove it from consideration at this time. +Making this a blocker since its making debugging quite difficult by hiding the root cause and giving a potentially misleading error message. +Currently when you do Reset the runtime summary gets printed out but its not complete wrt to the settings you mentioned above +Bind macros now assign a temporary string representation of based on boolean other data type +Ashutosh Chauhan Its good +Will post an addendum fix shortly + + +Is corrected in rev +Typically why wouldnt the parallel case of DJ not use the initRequest method? Yes technically thats because it is used to do send snapshot commands for only in the sequential case but that makes for a poor abstraction imho +Im still working on this +Retries times + +attached is a monster patch for flex with autogenerated Lev and Lev from the i just uploaded +Additionally this is GUI functionality which would be difficult to test +Forgot to add that this was changed in revision . +The clover instrumentation doesnt seem to be designed to naturally cover them +since the test is related to cache eviction so I moved the test case to and changed it as what you suggested +You are right +This is a known issue and the MVC addon needs rewriting to be more ll mark this as a major improvement rather than a major bug +This patch was created for SVN revision +youre right +Release cancelled due to Disclaimer notice in to move to as was issues after move to in +Ill take care of it. +This issue is about text extraction and not about creating pdfs +think Ive solved the issue tooIve integrated a matcher to the and generated a suitable submap by copying the to with an additional transformer to filter the entries from theIve also updated the +Juergen +The Eclipse project has dropped the idea of renaming the service for nowIn case of later standardizing this API it must be repackaged anyway and may then be renamed to a better nameFor now this issue is resolved without fixing it +Yes its what I saw there are several dependencies not used at allI will submit a patch for this Jira +Will not close the issue until confirmation that they all have the accounts operational +Therefore a replacement is being prepared +Undoing the profile turns out we dont need it after all. +Thanks Stack +Patch attached to fix the issue + +GuysThanks for this patch +But yes the work around is good +HiI get this same issue +You are in a much better position to make judgement on this than IIt has been an interesting ride thanks PS +Patches because the same issue has been reintroduced in the build file of the slow consumer plugin +Hathaway +A couple of comments have you really meant to say in the test case in the patch? I think you could have fixed a number of iterations and coordinated with objs instead of relying on a s sleep +source archives for all jbosstools and jbds plugins not sure if they are really needed in installer +Help messages are also updated accordinglyIm attaching a patch for branch for review +is the handler while the other two files are the target service for the the bug is fixed in the current ll try the next nightlythe problem has not been fixed +looks good to version of the diff Im bumping both multicache and hostdb cache just to play it safe +Moved of GA as agreed on call +Thanks +Repeatable with JBDS CR on a Mac but not when running from the IDE +svn ci Sending mainsrcmainjavaorgapachekarafmainSending mainsrcmainjavaorgapachekarafmainSending mainsrctestjavaorgapachekarafmainSending mainsrctestjavaorgapachekarafmainSending mainsrctestjavaorgapachekarafmainTransmitting file dataCommitted revision +But please keep me in the loop too as this system is how we deliver and automatically install the Cumulative Patches through ON +Attaching reproducer archive type ant Drepositoryreponame Dworkspaceworkspacename to run it +All new features need to go through the community +Hey JustinIve been getting some flaky runs on jenkins which probably have nothing to do with this issue but I would like to see some reasonable results before I close the issue +I ran the test with set to something other than buildtest +Release notes are not updated in versions and. +maven should then use its resolver to use the latest will close this issue as it doesnt look like there is anything that can be done atm +There are likely a common set of failures which should return a common error code across all the usage aspects which could be added on by tool specific failure codes partitioned to different rangesI think going to a simpler for all errors may suffice for now +Unmarking patch available as Lily discovered some test failures +Posted an bug in their forum about this. +Connecting to a service generated the expected no TGT exception then the client did a relogin from the keytab and successfully connected +We have added this textThe Enterprise SOA Platform includes a new administration console +This can be resolved for now +Ship it +I never invoked closeregion through hbase shell in my environment though +Attaching patch for +On Windows x and JDK still has this issue with karafBut win bit and JDK no this issueRegion and Language option on my machine please see attachment . +Otherwise the compilation of the test will be broken again and again +I dont get a good read with bisection it may just be flaky +We may also want to include email based notification in the event of a server side issue such as a full disk drive or an offline database +HiI am still having the same problem with the buildCan you please check thisBest Regards +Ah now I get it yepp youre right +Once youre subscribed the mailing list address is Thanks +No I didnt but doing so doesnt change anything +hold on until we get a decision if its needed to rebuild installer to include this fix +In case you do that please make sure that you link this issue to saying that this patch is incorporated in +I have fixed this issue in current SVNThanksDeepal +If a parser is relying on the order of the map keys being the same then they are doing it wrong + +I tried the workaround from the Eclipse bug report on my colleagues system and it worksWe created a Apache Directory file aside the Apache Directory executableIn this file we added the followingvmCProgram FilesJavajrebinThen we launched Studio and waited until it finishedWe pinned it afterwards and it worked +Within the CLI we now support the use of a descriptor to specify the SSL keystore and truststore settings in general if someone wants a remote silent authenticated connection then SSL would probably be the better way to achieve this alternatively its use would be recommended against but a username and password could be added to the same descriptor and we could also consider vault integration with the CLI. +Can you please rebase it on the latest revision? Thanks +Future autogenerated configure files may have the space in the shebang because autoconf generates them this way +Bulk move issues to and is now complete for and is used by to forward around update requests. +package we cant create it +This is patch for this sorry +To my understanding that only concerns the client side isnt itAt the moment it seems that we cant register several repositories on the server side since the contains only one Repository object or are we supposed to instantiate a provider for each repository +The second one needs to introduce a map for recording size for each baseDir an int field in for recording its size and a string field for recording which baseDir this cache lies inSubmitted another patch based on option agree with option +bq +Ted If you update this patch +outputs of processing are no longer buffered they are replaced with a list of skipped items that are removed from outputs before they are passed to the should be added for the described tasklet unit test and integration test within. +thank yousorry I have not really understandnotice all details in the documentation +Removed all my js scripts from and plugins from to test +fixed in trunk for beta as rfixed in branch as r +Except for the double L in in your URL everything else looks Ok +Added missing content styles for select and select +Verified with the fix +Will keep testing +Debugging why that assert is firing will likely take a bit of time to debugIve never built the vm on windows but theres a first for everything +After upgrading to the latest version of Apache we are no longer seeing deadlocks in our production environment +Sounds reasonable although I think it would be better to come up w a different enum for the Paxos phases than CL most of whose options are not appropriateI actually think on commit is fine +This ticket will not show up on the radar because it currently has a status of resolved. +Actually the current heuristic doesnt really work at least not as a DOS prevention +This was introduced by I will have a I committed but svn was unavailable then forgot to commit again +thanks knut if no one else tries this I will give it a go on monday against +We are just using the Cache in buddy mode for session replication +We want what we always for and thats an easy way to reproduce the problem +Patch makes sense +This issue is very closely related to although goes beyond it for the purposes of LDAP searching +I seeThank you for committing +sounds good +Great +Upon further reflection I conclude that its good that it locks all children but I want it to do so only if the set of children names changed +Added check so that blocks that require single replica are not added to neededReplications queue +thanks Tom +This patch was done on +will not use that instance but the one refered to by JBOSSHOMEI just got burnt by this +verified on +btwI wanted to have the srcmainjava and srctestjava dirs show up but you need to have something in them for the archetype plugin to work +problem of old client library issues seen in new is not related +previous version not for commit +from me +Now go to facility component and select Inventory items from the bar select the inventory item id for which the order is placed +One of the RS znode was not yet deleted but the master started coming up +closing +With commit cdeddaeaeeafd we randomly select one of the last five favourite tweets and display the twitter screen name. +I had the error JAVAHOME is set to an invalid directory CJava when I ran the grails command on the command prompt in out my JAVAHOME environment variable was set at System Variables level +I just committed this +Many of the widget style attributes were created before the big UI refactor in where we agreed to style only container elements and the child elements are styled using CSS descendant selectors instead of assigning a style to every child elementI think this issue can be closed +If it isnt there already Ill put in a ticket for the +Some amount of mapping configuration support needs to remain though to support plugging in alternative filesystem implementations that ship outside the Hadoop codeThe HDFS NN just needs to know is this scheme hdfs or not? No configuration is needed +this bug as there hasnt been any reply and the bug seems not to be related to Cocoon but to Tomcat. +The other thing is that I was not planning to do the string replacement so that we keep the old behavior +Here I just attach our report of this performance evaluation thanks for the good comments from JD stack etcPlease have a review and comment +Ill modify later +well I dont have anything at hand +This is seen in our nightlies where we see other services being impacted by namenode being in safemode +Pls +Logged In YES useridSorry +Here is my installation historyI started with Featurein indigo but I wasnt doing much groovy development at that time and I dont recall having corruption like thisNext I went with Feature in december and Feature in march +Fixed in revisions and +But otherwise Id call a fast a feature Now I know very little about jetty configuration so this may be all wet but aging out that information quickly would make the part of the UI quite hard to use +Added Len as a developer on the JBDS project so he can be assigned +Maybe have these methods call a common method +BTW I dont think any of the Commons bundles includes the parent pom +has been marked as a duplicate of this bug +failure is unrelated +Adrian assigning to you so that it gets revisited once NBST is in place +Im going to mark this as incomplete since its too amorphous and well probably work forever to build embedders can use. +This should be resolved in has been fixed in +It looks like the developer started to implement this but never did it +it would be nice to see Accumulo with an asynch client library +requires set to test to ensure all artifacts are available as dependencies when building the deployed +We cant fix all the bugs from previous within those or well have a maintenance nightmare. +thanks +Bit funny about this +Ah yes +I added one test on node name s immutability in MOP project support changing node name now +Either approach is fine as long as we document it clearlyIntuitively returning a long makes sense barring the corner casesBut I am open +What Attributes +Once the region keys get messed up it is very hard to manually fix these issues +Fixed +Alternatively we can place this into cookies but I am not a big fan of that solution a solution I would simply always set the page variable to if the user clicks on the list all menu +They arent particularly noisy tests so Im shelving that issue for now +I guess what Angela is saying here is that we already have some ideas on how to implement this in the context of the existing access control mechanism +Also r +Ive added a separate runner that can be run independently from Ant +Sometimes you have to the tests because the Arquillian adapter is not super reliable yet +If it is easy to fix it should be fixed soon +Documentation patch that does all of the things I just +I do not remember the exact case of using Thread CL but it was solved the problems that was related with javassist class loading issues +For the time being you can use the samples available in svn +crap wrong to release to make sure it is even awesomer +On looking at it it appears this happens because the classes without an annotation are coming in from a jar +Rayees could you try it again +I will change sleepTime from to + +It is failing because the test isnt running with the upgrade of the +Given the date of my last comment the fix should have already been picked up into CP although I cant be sure +Jonathan let me inform you that the Java EE support has been deferred from the WTP plan +Well wind up patching hadoop ourselves +Modifying the title accordingly +Ive always thought it was some kind of classpath issue or something like that but I could for the life of me not figure it out and it seemed to occur when adding as a dependency +Was easier to see when I objdumped the executable instead of the +Which means eclipse update merges and for installed product +So now both and have their own and files which get pulled in when generating the packagesThe permissions problems and junit inclusion has also been fixed +Currently the solution complicates our code a lot +Still occurs with Grails reproducible in git head +Great that eliminated the problem I was able to successfully do a mvn install of svn revision earlier today and then svn revision a few minutes ago without seeing these +Here the policed oneIn my opinion something is better than nothing +changes are atSending coresrcmainjavaorgapachestrutsactioncomponentsSending coresrcmainjavaorgapachestrutsactionviewsjspuiSending coresrcmainresourcestemplatesimpleSending coresrctestresourcesorgapachestrutsactionviewsjspuiSending coresrctestresourcesorgapachestrutsactionviewsjspuiSending coresrctestresourcesorgapachestrutsactionviewsjspuiSending coresrctestresourcesorgapachestrutsactionviewsjspuiSending coresrctestresourcesorgapachestrutsactionviewsjspuiSending coresrctestresourcesorgapachestrutsactionviewsjspuiSending coresrctestresourcesorgapachestrutsactionviewsjspui +lgtm +Tested on Metamer Developer Sample richfaces showcase in Chromium and Firefox. +But when I downloaded everything through CVS and rebuilt it worked fine +This will be done as part of the migration to was done as part of the migration to the new brutus so this issue can be closed +This makes the interface consistent and makes it easier to get a file name for a placement policy like colocation +A minor bug in the controller channel manager that fixes it to read an properly +Reopen to add documentation about the changed behavior +Oh nice! I didnt know you could ref descriptors like that +Leonardo can you comment on the server side check if we need it for textxml +Unless theres something unexpected here I dont see any bug so resolving this issue. +Feel free to add a comment with some updates once you know more in case other people have the same problem and end up here +srctestjavaorgapachehadoopyarnserverresourcemanager +Thanks Santiago +Not ready to be committed but closer +Either JDK or CGLIB and I have used Spring autowiring quite extensively +Changing description to reflect the problem and the corresponding code changes +bq +actually Im not doing the thing for now +Interesting +Is this ready to be close now? +. +The solution provided resolves the same +This issue would be addressed by my anticipated implementation of +Forgot to pull before rebase the new version apply to latest now +Lower priority to keep list t Fix +Removed row in to schema update file +Initial version without any convenience methods committed +This patch adds the new source file to the patch +Just fair warning that I think the patch is kind of ugly +Maybe theres also a possibility to use this switch via language flags or something like thatVincent How did you integrate the languages to the default project site? I created a version with en fr de +Thanks. +Sending modulessrcjavaorgapachegeronimoclientbuilderSending modulessrcjavaorgapachegeronimoconnectordeploymentSending modulessrctestorgapachegeronimoconnectordeploymentSending modulessrcjavaorgapachegeronimojeedeploymentSending modulessrcjavaorgapachegeronimoschemaSending modulessrctestorgapachegeronimoschemaSending modulessrctestorgapachegeronimojettydeploymentSending modulessrcjavaorgapachegeronimodeploymentserviceSending modulessrctestorgapachegeronimodeploymentserviceSending modulessrctestorgapachegeronimowebdeploymentTransmitting file dataCommitted revision +I was about to check this in when I realized that a user might not actually define the in at all +It also implements Ryans idea of moving the to the. +commit cebaafbbcfccfeeb +Sanjay your table summarized all the scenarios +Per defecte serien totes dues +It would be nice if a new version of the Stage plugin would be released so we can actually use that one instead of making our project rely on a SNAPSHOT version +Thanks look forward to doing more +Yeah having a nd though of this and a chat with James as wellWe should ensure producers is within an so that file cache is deleted when its doneSo I am polishing the code a bit and making this work when you use Produce annotation as well + +And then use on page like following llayout +duplicates +has been marked as a duplicate of this bug +Ah i hadnt spotted there was already a patch +Hi +Also lets back port this to +From this information we will write the final release notes for all issues +That is the Maven convention. +Thanks Robert +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Rob I believe this JIRA is complete if so can we close it and track any bugs related in separate +Basically for each group level you need to execute search +close due to all sub tasks have been closed +Thanks for the fix Knut +Got it +I will cut the branch from master on that day +Part needs to be a bit smarter +If you finish it before we release beta please correct the versionve linked the tasks for the remaining work +Thank you for your quick answer! +I am struggling with this issue for long time I tried many things but nothing worked out +committed to trunk +Thanks a lot for the review ToddAny other comments? If not Im going to go ahead and commit this later today +src generated wsdljava +Verified in jbt build. +Francois agreed that this is acceptable since it works fine with to happens with for certain users +Patch applied with many thanks! Good catch +updated the slide share link +closing because this was not followed up on and we got another ticket from Kevin that had a fix +So you just say Too bad! Not enough details and I am not your car! Duh +Nice! fails because I have changed the version to +Im pretty sure the list of OSJDKs did not change for this release so we cant include it unless QE sign off and say they tested with this +general query bit were moved to fixed to. +Will do this for RC +Log from a lightning run attached +New patch maximum queue size is bumped to GB the exception is now sent back to the client +This will be available after the merge of the SASL branch currently targeting +Retested +Hum +All the unit tests still pass and speed seems to have improved tremendously on full builds +Please review this +Thanks Michael +Not sure what we should do at this point since it seems like its more of a Callback Server bug and a nasty one at that +I added some notes to the javadoc of and to the users guide +Therefore all the following processing will use this charset +The problem is repeatedly reproducible in Windows XP Prof Service Pack with Java and A Start run c allMachine B Start run c allDo a Farm deployment one war file in Machine deployment transferred the deployment to down in Machine for some timeStart in Machine forum is required for Critical issue +This change went in as test has not failed in over a year +r rewrites reachableChild to not use mutual recursion or recurse over a tree instead it uses work queues +Should be resolved now +Thanks! +the issue was resolved +to commit just yet +I think I hadnt done a full clean build or something once I tried to debug it I couldnt reproduce +Please reopen if you think that there is still a problem. +Just guessing because it was not me who created the SOA branch +But Do we need four commands? I think one command is only needed +I view the additional config property as just a to enabling a coprocessor on each user tableI do think that too many configuration properties can become confusing but I dont think one more will kill usFor naming I prefer for consistency with terminology throughout other coprocessor codeconfigs +Retried with version works fine +Is the status still Coding in Progress +Heres a patch which addresses the issueThe trouble was that a helper method used by both failing tests had a race condition +Deferred to +you still need to go to each pom and update the parent rev +Too often problems are reported that are caused bya mistake in using the API +Add Pig support for chukwa +Added tags and parsing in line with suggestions in this issue. +Both situations are not allowed for insertBetween +Nicolas when you get some time can you please give version of the patch a go +Was that with +Or one of the two cores is managed but is currently not active +not sure here yet +Im not so sure about command key but Ive read somewhere this is Meta key in Java +So all issues not directly assigned to an AS release are being closed +So the trace tool will have to parse all the header files first and keep a table of all the instance methods so that it only passes this in instance methods +It looks like the apachehandlers test is affecting test execution +Try first to delete the previous syntax then add the new one it should work +If you look in the buildtest dir +Im touching base with each of the leadsDamon +Closing the Jira +This bug is the direct violation if Grails its set aside +Point being I guess Erick if you dont have the time and we need to have this out shortly I can tackle it for you +I might even be a good thing as it keeps network traffic down +should have started the comment with we have two cases in consideration here +Thanks for getting to this Robert its a good improvement +Worked on at least one file +In almost all places in Roller we use to encode but in the weblog calendar model classes we use the encode method in which is the one that adds the JSESSIONID +updated patch adds the jar generation as a prerequisite +Committed +I added the build method back deprecated it and simply let it delegate to the update method +Does use mean we can include it in svn with its runtime libraries +those are all variables that google is asking for in their replaces previous patch with same namecreates service and simple methods to populate template +Yep cool that seems to fix it + looks good to me +Thanks for the feedback +AlexPlease review the issue +I think we need the good servicemix guys to put smack on theirsmack andsmackx have been added to the servicemix repo +The archive contains formatted sources +revision Made cases in switches sequential +If users really need a shared singleton generator they should do as Gilles suggests which is to wrap and then only they will pay the sync overhead cost +Dup of + +use writeFully now that its available write a megabyte at a time to minimize seeks +Hi +Arun I thought Id leave some to not mess with test coverage for that ctor as it has not been deprecatedreplaced by this change +do not see the motivation to override the default commands +Even hacking the db would be most difficult +vmdk files +sounds good +Enhancer does not draw the download links so they are only used within Contenthub currently +This looks like fixed by +I dont think there will be a Geoserver release can you upgrade to +My xdoc dummy would like to update the german translation of the xdocs terms as they look after being revised by for your god work again +Well have to deprecate the class and create a copy of it with a slightly different name +Hmm +The looks good could you please make it writable and put svn in readonly for that one +Id recommend retesting with instance types this will get you a lot closer to real hardware IMHO +opinion on this one +Making it patch available to kick off the tests +I dont have a good grasp of it +Lots more code being removed +Wondering did anybody of you tried OL with and whatnot +I see that this ticket was resolved as fixed but were any code changes required? If not it should be resolved as not a bug and the version should be removed. +This would make it to get started and become familiar with Thrift +The default is already login +look for errors in your error log +I promise yours is the next patch that is going in +This is a documentation issue +I dont think this is realistically possible in order to record the positions we need to know the tf for that termdoc because that tells us how many positions there are +Dup of +jBPM will be supporting BPMNI shall keep this issue open for a while to see if others have any thoughtsThanks againRegardsGary +What I was proposing was merely just putting in the means to the test to accept such a license +Are there any realistic situations in which a scanner is actually unknown or are all those situations just with the wrong name? I ask because I do not want to mask genuine errors with the wrong message +Sounds like a great idea +I get the artifact which works find but no sourcejavadoc attachment as Gradle needs the POM +Ill commit the patch after running and unit tests +now that is closed. +Patch for fix described above +How can you say it is fixed if you cannot verify +Reopened to set the Fix Versions field to JDO beta. +JIRA housekeeping +the +Krik would be great if you can resubmit with the test case +MiteshI guess adding this to might involve a lot of change +Is it not possible for you to only update the dirty region of your frame during your animation +New patch syncd to trunk +First attempt here +I am attaching the missing +Thanks Tim patch applied as expected +Ive Ignored other methods in this class and due to intermittent failuresMy suspicion is the failure are due to messages being removed somehow before the mgmt op to remove or move them comes in +Interesting a bug in versions afterCan you tell that to our colleagues over atThanksregardsMartin +Should be available in next Grails SNAPSHOT and beta +The example that I started with did that +Thanks Willem your patch without tabs is now in trunk rev +Thanks +Nice work Tom +If its deprecated in this release and wont be in the next release Ill put it in +Workflow automation capabilities multiple project phases +The exception in my comment was caused by the lack of disk space +This means modifying the language objects and rewriting the resolver and probably some other stuff as well +I am running tests nowI agree with Dags analysis of the problem +Tim installing packages for your build is up to you +Ive run out of large invoices that are not in ready for posting +Please provide a patch +Forgot to mention that in every case the result of mvn help always returns a result with all the properties successfully expanded +summary changed according to Vladimirs proposal +I have created a test with a longer text from Apache Solr web page +Committed toThanks for the port Sergey. +at least the local overridecustomizations files can surely require sorted order +A row in a table must have at least one column according to the previous section Introduction to tables +Fixed for EAP and EWP +Great find although a bit obscure The issue is that we differentiate between OUT parameters and INOUT parameters by looking at the value suppplied for each call +Ok I probably missed some part in the patch +perform dbcreate and register outputauthrepo my question what does Initialize command do? Im sorry I misspelled +We would need this sorted one way or another until we reach infinispan +commented out in the mean to +The only thing as you mentioned jackrabbit does not implement local txi will have a look at since is fixed +As per previous comment did manual testing on the patch no easy way to write unit test +OK I applied the supplied patch to the sources +I evaluate you first proposal +Committed revision Thanks Noble +This is year old now +This seems to be working for me too +Patch v changes the comments +While the column is in the list nothing showsDo we need to make some other config change to this project +Uploading a new patch +This will make all connections use the connection parameters specified rather than just the initial one +I will commit this to the branch for now we can revise if there are further comments. +Patch +Apologies +This bug only affects what is displayed and also results in filtering on keys not being reliable with instances launched on +Liyin what do you think +This leaves the system in a bad state wherein the host is no longer there but running on it are still present +So it would be good to explicitly set the producer type to syncRemoved the override of as there was no particular reason to not use kafka default +I mean the one used withIs there any bug on this exampleRegardsJim +I had the same d agree that the error message is quite install before webstartjnlp worked around if equivalent Id prefer to runmvn package webstartjnlpYou may want to point out this hint in the error msg. +I dont know that it would fit in this jira issue but I think the ideal solution for password stuff is to have the tomcat connector refer to a keystore gbean similarly to how the jetty connector works +the needed code to implement in RI is already there +In the screenshot the version is But must be just +I will see later +Patch v subclasses in the test so that the waiting for slave to catch up can be done through notification instead of sleeping +First called successfully without error further calls issue exception because resource is already re getting this in our application when we move to +I suppose that could be done on the first time spellings are requested but that seems like it could all of a sudden cause a much longer return +The distcp can explicitly exclude copying tmp files +We need more data and some confirmation tests before we can add this to the guide +Ok I have a workaround Im considering shipping with +Verified the problem has been fixed withThanks Ted +Will commit unless objection +getsetdelete then youll be fine +Im going to rewrite most of it but it shouldnt take to long +It is done +Hello Davesorry I couldnt respond sooner somehow I didnt get an mail about your comment +I think we need to involve the Rake folks on this +Hope you dont mind. +Reused the SOAP service connection pooling logic +Which brings up +Documented in NN need to get this into the BPELJBDS guide. +Ill go ahead with it then +Beautiful! Thanks very much Chris +Thanks DagCommitted patch to documentation trunk at revision +Related bug was resolved +Im wondering about making the rounding method setting a global one vs making it a per instance fact in Java the set method is per instance +In Eclipses problem subwindow there are error messages likeerror Attribute references to layout that does not exist in web content +Parameterizes scanner caching and changes default to lower safer number +Done +patch of unclosed issues. +OK Ive added logging +Stuart has sent a pull request for this. +The easy path is to fix that for Hibernate which will be Java and up +Decoding tests disabled pending submission of missing binary test PDUs +Seems hard to reproduce without full code +Thanks Jason +This issue is more an optimization I guess than a bug dont do wasted work on segments that will be droppedbq +It might also allow access to a file in resources in the root war +My guess is so many batchmutate errors were being logged logging consumed all the cpu before the gossiper timer could run again which would have solved it +I quite like the idea of adding it to the XSD as long as the default is in the Java source code as well as you say +Alex what happens if attribute is added to the project entry? Will that include project properly? If not then it need to be reported and fixed in WTP because project configuration should work without manual editing +Have to discuss the rolling of logs as Im unsure how changes it +Solution in this issue was done. +Thanks Christian +I have not the in depth knowledge in seam to know if this is a seam error or Internet explorer errorThen clearly you need support +This can be considered as closed +This is not a problem anymoreJust to be a bit opinionated for a few moments though I do believe the standards to be wrong on this issue NULL values are an excellent way to force scientists to really think about the query theyre running and implicitly removing them will generally lead to errors and more wasted time than having to call the remove nulls version call it avgrn explicitly. +Here is a patch to fix all the ones from common ve just committed this +In this case the will cause classes of package to be loaded from bundle +A unit testing framework is meant to run tests one after another +Ive converted your diff to a patch file +From what you list above it seems that a lot of the magic needs to be run once per test +Added some gasp unit tests that were gasp written test first! jruby e include Java m puts now includes the aliasescontainskeycontainskey? this is newcontainsvaluecontainsvalue? this is it I forgot the equals overload for setters +Please review the updated patch +This patch looks pretty good +nd patch applied rev +there are no successfully finished tests +Please let me know if it needs any more changes +Please verify +When no text is selected selection will be automatically expanded to select all the entire text under the cursorthumbnail +I opened new issue for continuing the above balancer explorations for review Ted + +Do you want to do the fix Christopher +Will push future changes to github later +committed missing code with Revision +This is an initial set of patches to implement character normalization in +WSDL created by the CXF service attached +Thanks MarkusYour patch is in trunk revision release +sample project to see the issue is still not resolved +Just curious if its possible to enable the combiner for the? I believe its worthy to make such effort to fully take the advantage of against the join in certain casesAnother more general approach might be enriching the optimization such that the common partitioning between two jobs is recognized +if hadoopqa has any attention for if hadoopqa recognizes this nkeywal fix +Marking Fix Version as because the work has been ported to +Fixed as a part of +looks good to me. +Attaching a reproducer +They are not thread safe and are not meant to be used like would recommend the pattern of using an EM per session operation also to serialize the instance from only when needed explicitly and to not rely on the hooks +There is one more thing it should handle if it does not do already +Thanks to Nipun for the contribution and to Brent for the checkingI have checked in minor editions to fix checkstyle and findbugs errors. +Thanks +I just committed this +One additional piece of information for the example where the thread is left spinning at the CLI also does not contain any configuration to allow the servers certificate to be tested we would expect this to result this in the user being prompted to decide if they want to trust the users certificate but in general this shows that the timeout and spinning is at a point where no connection to the server was established rather than a connection being established and the HTTP upgrade step failing +You can get a when building a DOM using a stylesheet if your stylesheet does not generat a DOM +already fixed svn r +Hi +Do you have a Git repository somewhere I could pull from? There is a Maven submodule for izpack projects. +Please verify +I attached the results +So given the current behavior and assuming we wont change it for the release what is the best advice to offer users regarding maintaining statistics +If were really locking up inside Xalan it must be in that codeIt might be worth putting printouts around the filters calls to the f object to track how those are occurring +but if sync thread happens to flush the ledger before journal flushes same problem occurred +Let me try again +Added tests currently depends on for resolution +logj issue title +Verified in ER +Patch to include the following in Hive ApacheThe version is controlled by the version property in which will need to be manually reved whenever a new branchversion is created +Patch with changes proposed in the description +Looking forward to your patch +Once your source tree conforms to the proper layoutthat is what will happen yes +Snapshot related methods should be moved to an inner class or separate class +I am working on a newWork in progresstrunk +The work proposed here can be done quickly and safely +See patch +WDYT +Robert Im agree with you but it was a test to see if it will return everything +SimonYour idea sounds good and should be no problem to implement +OK I think I got all of the deprecation related failures in ant Its tough to tell for sure since there are many other types of failures in +the patch looks good +Thanks Luke +Please do a svn diff and add the patch so I can see the changes against the trunk +Springization is doneStill working on concurrency issues questionable if done by final release Dec +You should like stop accepting writes from clients at that point until the condition expires +I thought about modeling it after JSFs but Chains seems to be closer to what we are looking for +Sorry Ive written wrong commentthe suffix is added in overall premise of this change is just wrong +Good catch! Fixed +Chris I wanted to say that the whole package is not public though some of the classes in the package are public +Here are some screenshots +This patch adds a decode that given the first byte of a vint decodes the total number of bytes +Sorry for the duplicate issueIt seems like too risky a change for the branch to me +Including places I dont think are due to this patchThere is an issue of handling possible xsitype +I see +Manik Can you please post the xml snippet +It looks like its working now thanks +Thanks for looking at this KristianUploading a new version +We had large date skew between servers and we run ntpdate to force the server time +Excellent thanks for fixing it! How about? Will that be pluggable as well? Shall I report a separate issue for that? Also Where can I get nightlystable builds of Grails beta +The DN docs state A is designed to be +That would be a nice improvement I think. +As you see the subject is encoded correctly but the body is with charset instead of +Thanks! +The issue is critical for In Harmony with Eclipse Alexey Ive tried on my desktop windows xp sp with J vm it works fine +That does have some downsides though like checking progress and jmx timeoutsIf we do this we should hopefully make it generic enough to hook all of our long running jmx calls in to +Ram +By the time is released then we will look at working with AS version +This initial patch demonstrates the problem +No matter how many map join operators in one Map Red Task for each map join operator there will be one parent operator from big table branch and other operators from small table branchesFor big table branch just leave it aloneFor small table branch create a new to replace the current Operator +However retreiving documents is still not associate with any cost what so ever so in a sized index that use Lucene for persistency of fields I still see a boost with or so compared to +Ok Ill try this tonight +Also IAC commands are not really negociationsBut I doubt it is worth worrying about that the Javadoc does not explicitly say that there wont be other codes added in future + +We have found an apparent bug in the Apache module and will fix it +So when using state saving they probably should not be destroyed +I have removed author tags for the classes in my experimental branch +Closing as duplicate of +I agree +Does this need a documentation change somewhere aswell? Can do as separate jira if needed +Works a lot better nowThank you guys +Thanks for the feedback +LGTM +Moving all issues to the CR +Lets go with option then +Victor will try to run the test suite against Progress and see whether this makes a difference. +Query needs the entity manager to stay open for its implementation +Just wanted to touch base if an issue I ran into should become a new JIRA item +Describe a new wizard +If anybody else has sometime please validate the error +I dont think this is a change we can silently shove into the CHANGELOG and forget all about it +Any subsequent run will come to a grinding halt at either TC or TC +Why are users still downloading the older releasesbq +When we add any new tests need not change others +Revision . +If there are no objections Ill commit this later today +But it buries validating somewhat which uses to confuse people +Artifact cleaning for ecommerce component +Hey Juraj could you please review the guide from the grammar perspective +Linked issue has some tests that demonstrate the behavior noted in this issue +Alexey any update on this +The absence of that indicates a problem in some upper layerI think we may have to throw some other exception if we detect query string to be null here rather than continue with the call downstreamLet be debug it and provide a more detailed update +but before you commit can you add a javadoc for that method so we can end the cycleThanks +It didnt appear to be necessary because all this is not necessary for getlegendgraphic to work it only needs a dummy feature +Prachilikithacan you review this patch and get it approved +The message does not fail and correct response is given at the client +Attaching the certificates for completeness +It would be nice if the datanode considers all disks the same +I updated the snapshot with the latest code +Only that my approach fails and yours doesntI have no time to track that down so lets go with your approachOnly one commentZNODEPATHSEPARATOR is not correct here since we building a FS path not a Zookeeper path +its a plain jane WFS request has nothing to do with versioning and besides wfs unit tests for it are passing +Will fix the issue later + +closing the issue as duplicate of the newly created one in issuezilla +If one wants to use it then it must be installed +Theres more in MR tests Ill open a separate jira for thisbq +It is going to be system property passed with D +There is no annotation in Seam that Im aware of +Could you please tell us a bit more about what you have in mind +On Windows the class uses an array containing a maximum of channels +The only thing still needed is the python script invocation to publish the built site +bq +Teiid handles similar situation when a large file is being deployed remotely over the socket by customizing the encoder decoder objects of the Netty channel and route the contents of to a temporary file store during the transmission +Code and config files implementation +Fixed by implemented the class but still get the same error do we need to redefine the hibernate dialect property +Current problem is that juddi appears to be trying to configure logj +Busch my friend Ill buy both you and Uwe many beers if you resolve this issue soon +I think this is a valid improvement request which is why I filed it here +I will rebase this after gets committed +This patch passed mvn clean verify successfully +Thanks for the tip r I did a fix which should cover most of the ajax forms in the projectThe bug will continues to appear when there are more than one list or multi forms in a page using ajax pagination +Lets upgrade to blocker when we have a test case and volunteer to work on the issue +This has been resolved +All tests run with this patch putting up for commentStill have to deal with Yoniks and Hosss suggestions about recording core creation errors rather thanBut the major change here is that creating cores via should correctly persist toPersisting is an incredible pain +Will be described in new bug. + +Why didnt the autoReconnect solution work for you +If there is an exterior client making use of Result constructors in MR? Else I do not know of one then Id say lets not change the constructor and add in newResultcreate +is related to issue +Could you try swapping for the system sed and see if it occurs to pinpoint this +the changes were applied in both cases +Radim does this stacktrace appears when you close? Looks like it is +Whet is aDoes it extends Configuration somehow +Less chance for unforeseen consequences and hidden bugs +However I reviewed the patch +If the community wishes to provide some other great tool not necessary related to Apache Hadoop but useful nonetheless why preventing itUnless these are satisfied any changes in the goals and charter of the project cant be put in motion +Hi Martinmy interpretation of the code and the FDF spec this issue is not easily solved as we would need to lessen some of the checks which are currently done by the parser +Commentsedits welcome + +Application Server still presents in JBDS +is there a time limit on when to submitI opened this for comments before final submission +Thanks Kim +Test cases being added +Il will look at them in details thanksFor some time can we just change the code in such a way that TOM thread is not started at all +Bulk close after is released +war plugin build time s avg +Attached latest patch that addresses JDs commentssuggestionsThanks Stack for running the tests and I did it toothanks again +There is nothing called or bit JBDS +Still no closer to figuring it out though +Something like this patch +Honestly in the projects Ive used it on we never had a need to directly wire the action +Ive added aregardsDieter +Thanks sanjay +The is referenced from the installer script though +has been marked as a duplicate of this bug +I think the code and classloaders surrounding this problem have changed enough so that the report is no longer relvant +Go go Rickard! +All I am arguing for is that if somebody wants to use DV it should be simple to do so +Assigned to DOTNET looks like the only place where this still currently applies is in the file for the configuration serviceCorrect me if Im wrong but I believe that only sections within files can be encrypted not sections within files +Both and look like good changes to me +The license is different but trawling on net would indicate its apache compatible +Updating Fix For to Unknown on issues not targeted for classes will be put on the chopping block soon enough so we wont be fixing this resolving. +I would have to say this is for as the does not work without it in most cases +Fixing fields +in this not seem to be causing immediate problems right is now causing problems with RF do you know if we can remove our dependency on this +Your concern is still valid and Ive been thinking on it a wee bit +I just committed this +Version Final is released +As for a this would be okay but also consider that generic forwarding functionality could be used by many existing without major rework if it was included in the APF +Run mvn package in the top level directory +Id love to pass this feedback to the team +it should only require port for +BenCan you please clarify what interface vs implementation do you mean +So you see REST clients as having to do the encryption however is appropriate for each particular clientYes but more like I see REST as really suitable for layering do the basic data access protocol in something like Stargate and then do caching access control and data security on top with specialist components +I actually ran the update manager to download the plugin and install it so I had to exclude from the update in order to go ahead with the plugin installation +Ah sorry +Verified on JBDS. +FYI I branched qpidtrunkqpid to branches to work on this +Heres the current version of the code +The problem comes from the sonar maven plugin so I change the issue title +I came across this issue today and being able to spin new cores at run time each with a dedicated +PatI am going to close this as anymore +This should be handled more gracefully +Tests are not necessarily set in stone I think they can be considered a caution against changes which is what I would consider this test +This issue should be resolved on + several duplicate jar artifacts are bundled in the source distribution under antlib and parserjavaantlib maybe only one antlib folder should suffice +I still think that we should cleanup Castors cluttered package hierarchy and as it seams right now this is the only way to do that. +I tried Ma and found these issues and does this issue still occur for you on Juergen Im not sure howto verify this one +But from QE standpoint of view we only need different platforms binaries to be consistent +Perhaps the line endings are the problem +All tests passed cleanly for my on the open jdk preview for mac os x +Not a big problem probably but we should be aware +I restarted the nodeA server the nodeA can not automatically join the cluster +Therefore the temp patch is still incomplete if we want to have DB on all the Tier platforms +Is there anyone who knows how that code works +I have a memory snapshot after the issue occurred but unfortunately it cannot be compressed under the MB limit Jira imposes on attachements +See +We use RF +I added a delete for this file before running javadocs so its regeneratedNow its final +I am not sure a space is a valid character of a wfs type name given that technically even an underscore is not +add console war to as distributioncustomized to include switchyard specific +has been marked as a duplicate of this bug +No worries I didnt expect it would I assumed you guys are probably busy with the M release +Seems we are missing a null check in get +Test failures are unrelated +This behaviour has been copied to the parse method +Can you say more about this Paul? Were already doing checkpoints every time we flush the view results to disk +This is useful when the client reopens the same file for append +Ive committed this and merged to and +Add proper constructor and it will be fineRemove void +Fixing javadoc +Len can you confirm and update the fix version? The issue can then be marked as resolved again +A mock up test is useless in this case anyway +Can you attach the page and the stacktrace it appears this issue was caused by some extensions we made to Tobago ourselvesIssue closed as invalid. +Can you please take a look at itThanks +The patch I also discovered the class javadoc for the interceptor was wrong for the example +This ensure the database test cases are disabled and can be enabled when there is a database running locally +With some sort of inheritance to handle it +I really think it needs a design document before we start committing patches +As soon as I try to bind a nested path with a null value in any position other than the final object in the graph a is thrown +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Its confusing that stuff goes into the dist folder +thanks Ivans great work +I really appreciate the help +Obviously the makes no difference to the existing oneA for das alte testament +made compatible with commits +It worked on the tip recently +Since we are only testing one version it would be simpler to remove the references to a version number in the classes and properties files +Log mining applications might find it useful to have a raw comparator option if they consider Authorizations objects +Attached what I committed to trunk +is bad term since it is really +Patch updated with release Add documentation for new comparator attribute on field mapping +Could you test in Windows XP? I think it might be a problem that is reproduceable in a certain operating system only. +There is indeed an encoding issue that I reproduced +It seems possible we someday might want to permit binary too so I added the encoding to the version string +I think that why this was not handled +Also would it be possible to accept a null model argument and have it treated as an empty Map +ChunhuiYou can open a new issue for improving the above code +However compared to most other types of error reporting fromTapestry its on the weak side Add in all the class enhancement stuff that Tapestry does along with the factthat the condition didnt generate an exception under b i felt I should filethis +I use and JDK + +Alpha release for review purposes only +First just adds the autoconf checks +see comment in parent openon doesnt work anymore +boolean deployedExploded true means the archive gets unzipped before its deployed +So there would be a change to the definition in partyservicedef as well as partyscriptorgofbizpartypartyThe only thing is there would need to be changes made to any existing code to include passing the product when calling the serviceAny thoughts? If there is a more simple way to pass that parameter without having to modify the existing code in the callers of the service it would be much better + is task is assigned to Sean Rogers as it affects the FUTURE version. +I just took a final pass through v +change committed to trunk files seems OK on alpha branch +Verified on EAP ER +Here is a patch +Additionally Ive added a new option namespaceInfo to allow for namespacepackageprefix options to be specified +fixed in CVS please test +As a forrest committer you do have commit access at Cocoon but would need to show merit there to be invited as a Cocoon committerPMCAnyway perhaps you know how to run services without being under an individual user account. +Attachment has been added with description escposrenderer with apache header and modified package added apache headerill clean it for newer fop versions +I wonder if this is linked to +stack Seems this bug can be closed +I have committed a change for this on trunk which I think should make it into +in and branch +Enjoy +Can you please help fix itthanksdims +Hi ShaneThank it be possible to include the fix in and not first inThanks I snuck it in for you +Thanks +We need to document that when deploying brms into the default configuration to avoid this warning user should copy the from serverproductionlib to serverdefaultlib +I have committed the patch +Jitendra can you please update the description of this jira with the problem being addressed +Is this just for the purposes of? If so why dont we just read the itselfIf were looking for something that refreshes the log displayed on the web console each time a new message is logged then using a might be an approach +Ok +Is this still a problem? I have succeeded in running the TCK on a platform using final without any problems +So please reopen the issue if there are any runtime with my Schema thank you +So even though this is only an issue with the JSF reference implementation shouldnt it be reopened and fixed anyway? am actually having a difficult time reprodusing the problem +Am I missing it? We should have the whole error listed here +Moving to +closing self reported issues resolved on versions already released +What we can try and do now is this stop creating markers and only call validation. +Here is when refactoring for testing would have made sense +I committed attached patch to see if it helps up on hudson +Have sent an to Kevin Conner Tom Fennelly to get feedback +I care not about the so Ill axe it +Thanks for writing this clear spec Rick +Stack thanks for the pointer on To shoot for I should work against the trunk? Or is there a separate branch for what will be +Good to go. +New version Puts BB test in javadoc Cleans up tests with new call Fixes Pig and word count +committed to trunk +All jar files are merged into a single master job jar file that is then submitted as the to hadoop +Triage for +Have you found in a lot of the datasets that you are looking at that is used prevalently +Moved issue into legacy component +Can you please add some more details for your jira? Im not exactly what your setup is or what you are using +On one hand having all the async stuff in one makes sense +Thanks. +I have a patch which I will submit shortly +Forgot to mark the issue +the patch currently adjusts the default based on lucene but this is confusing if we release that disagrees with lucenes actual defaultsI dont think all of Solrs defaults need to match Lucenes defaults in general we can take it on a basis +Running the same test on Fedora MRI Ruby hangs too so the problem might not be related but system setup related +Just added the file. +Can you reproduce the failure locally? If so note the frequency of failure and an example command that triggers failure +Im not sure what that property does +Was apprehensive if this feature would be accepted +This issue is the same with +This may be a while before we can come up with a good fix but in the meantime try making your own +I put a watch on the issue you mentioned so I can commit this change when yours makes it to the repos +Thanks for taking a look. +Congrats Julien that works perfectlyI assign the ticket to Simon for code review +Or maybe only when generating release artifacts +added for +I have no objections to the patch +This is still very much an issue the supplied test case shows a largerapid memory leak under Ant Groovy. +Ok guys Im back +Andrew The fix is in cvs and will be available in the next nightly build please help verify thanksRgds! the fix works +Completed in a squashed commit under +I think Hudson needs to add srctest directory for checking javac errors +Ill be banned from JIRA +Here a examplemx components have such item events including mx +I guess we can close afterwards as it is just a symptom of by junits are failing +Looks like is assuming a byte address during deserialization but actually writes a full byte address during serialization +As we are really close from release I postpone this Jira to +just mark the issue as and fix version is not validated + +need to repro +IT added. +Santiago has sent me the updated file I have recreated the patch using eclispe for orgapachexalanxsltccompilerThis patch contains the changes from Santiago just the patch is in different format +Thank you +Could you at least document how to deploy and start it? I looked everywhere and couldnt find itIm going to resolve this +Is it in yours and is it declared in your settings +Username asomayaj +Patch applied Thanks! +This is disappointing +seems easy +It now appears that something could be done in to insulate against this issue +On its absent it takes epoch Unix timeI thought that this was an issue with the servlet engine but it seems to be a diferent problemWe cana Find out why has not been returnedb Implement server side instead +And even if we use WTP JSF Tools we still have to use our own validator to provide validation in XHTML java xml +Thanks Erlend for the new patch +Please upgrade to the latest code snapshotOleg +bulk close of all resolved issues. +Committed +Where to put it on the other hand is purely a packaging concerned and Im not familiar enough with the project to make that choicesuggestion. +Surely we should do the intelligent thing here and send the anon replyto if the user has added reference parameters +Current patch on implements a map reduce framework in the regionserver that can calculate aggregates +Im against this proposal IMO validator should stick to its primary purpose defining validation rules and not get into the business of page flowMy preference would be to close this as WONTFIX +Alexey can you review v +Castor does attempt to read from the filesystem you might want to try modifying orgexolabcastorutil to see if it fixes your issue +I just committed it +I share the and I delete the parameter it can run the job normallybut I dont why +This is a dup of + +More research about the root cause needs to be done to resolve this blocker +You can manually fix this by removing the tmp from all the files +At least thats what our production stuff is doing +I tried to implement two new modes OPENDIR user can choose one single directory OPENMULTIPLEANY user can choose multiple directories and filesHope this can help. +Suspect its not reflectively accessing the handle field of the class correctly +The patch will fail wo these +See Upgrade to Remoting Ron +Please consider about following points +Also I had forgotten to change the sidebar link to the All javadocs to include the new all path element in srcsitesrcdocumentationcontentxdocsThis version fixes these two issues + to removing the unnecessary check and leaving the assertion as a verification +Patch has been applied the generated bundle looks okay +For tables the maybe the filter could be built in parallel with an MR job +This problem can be revisited in the course of api overhaul +I tested it a little bit and it looked promising +May be the problem is more prominent in our testcase framework +Change in javadoc for patch can be done at time of commit +I didnt remove the level of indirection but simply allowed generator methods to declare their return type as directly +Closing per above comment. +A reason for using the hash function for cookie authentication is that you dont wanna have your password to be persitently stored in cleartextBase +both of these sections refer to using the schema tool database configuration switching databases sections of the administration guide have been merged +No suspension there just a single Connection supposed to be used for multiple subsequent transactions +This was on a cluster large enough and free enough that all of them could be running at the same time +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Thanks Karthik +This is the trace of the maven test execution +There is no reason why these files cant be installed expanded at installation time that I can seeIs there a use case for outside of native libraries +Just a note the version of roo used is from gitroo which I pulled down at GMT this had a severe issue with members in a controller as a result of some refactoring +I have updated the wiki +Doesnt appear to be an attachment on this issue +Also is there a relationship to? Thanks +Uwe you can also remove the writer package in the jar Solr doesnt use it and it has been removed from Commons CSV +to do that +This is likely to require little bits of additional cleanup but I did some looking around in +Setting priority to high +simple enough to fix i suppose but what alarms me is that we have a that explicitly tests if it gets debug output when using the MLT Handler and it seems to be working on the x branchIm hoping Koji can chime in here with some idea as to why the test is working even though the code is clearly a typo +I found the solution +This change still requires a lot of the Expression subclasses to extend so that the annotations can be retrieved during a transformation +Bart +I have often wished I could see a tests output while the test is running +I made some changes mainly caching the file handle in fifh and its performing much better now +So im afraid the short answer is not for this CP release +This issue is more than minor for my current project as native scrolling is the only solution that provides acceptable performance for the complex scrollable view we are implementing +Im sure the Hadoop scripts have this same issue as all of this was copied from them +got a checkbox baby! +fix version needs to be changed +So now uses and a +One last question +So this is a non issue +The JIRA project is now etc +Attaching a patch which adds bundlesbson and +Remarkable point is as next Original code was using variable bytes to handle remaining amount of data sent and remaining roon in DSS segment +I am making this patch available for Hudson blessings +This seems messy to me I would love to keep an invariant that once a file is finalized it is never renamed or changes contents +When JBOSS framework is used the service sends the response to the REPLY URI and it also attempts to send an HTTP response with the full payload back to the client and throws the exception +Ill implement Bens suggestion re warn +seems good +Martin I think just initializing the component might not be enough since the referred component might rely on an parent component or the page ifself that needs proper initialization too +Hi Leon did you have a link example for this error in Ofbiz +META +For further help move to the mailing list Second Im quite sure what you defined are fallback fonts +Line in Xdockicon +Patch added +Could you verify that this is working correctly on your server environment before I commit it +i think it would only cause the nightly job to take an +selection can be aliased such that the alias can be used in Order by clause example select as comp from Emp e order by support for JPA and for JDBCSupport for case expression evaluation will be added later. +Someone posted a very simple patch on google code Ill check that one out. +thats just for now and wont be released that way +Ran ant +New patch looks good +Verified on JBT +Talks of and gets master address using old style zookeeper access +Documented in JBDS User Guide chapter. +Closing per Kevans comment +I put this into trunk and +So we could do something similar w a public update site and a registered user update site using some form of http auth to validate before loading the contents of the latter update site +Potential patch but breaks which uses Delegate with a protected field investigating that now +bq +I think you mean CDI lifecycle events arent really designed for classpath scanning so I dont want to really consider this use case as a driver for the Im not sure how this feature would help you with this anyway +I sgguest that we not create the util module for now and Ill add the code in one of the existing corba modules +It looks like the test was run on GCV +If user does not touch tuning knobs we expect no slow down and on this platform see a slightperformance gain +Im not sure why these tests fail seems like some timing issue +Patch Committed revision thanks Susan +We are pretty close to eliminating modello with woodstox +yes the patch is old and will soon be rendered obsolete by +Its already bad enough that the also contains the same information +I keep meaning to get back to it +Btw thanks for the corrections +Due to this should not be cleaned up during network GC +commit a better version of this patch which also fixes dataloss but also provides a more optimized code path for theGiven that the code does not exist for option above and option causes dataloss regardless of your opinions on HA we should go with option aboveKeep in mind that even outside the scope of HA this fixes a potential bug with restart +So got rid of explicitly passing TESTROOTDIR +Verified the changes are present in the jar +In the description for this JIRA you say that CHECK CONSTRAINTS cannot create a nested context +Marking closed. +IMO the guarantee of delivery should happen within the replication timeout of a SYNC cache without the need of an extra flag +The reason Havard gets this problem is that the project in question only contains interfaces and no classes +For instance Berkeley uses runit +Code based on Dougs comments +there should be no SNAPSHOT version in these bundles when releasedbuilt +This attachment contains the lines that contain? in the rat reportApart from the commonsjson code where we kept the original headers those are small test or utility files with no creative contentThe only files that could arguably be considered creative are IMO launchpadcontentsrcmainresourcescontent launchpadcontentsrcmainresourcescontentTo which I added the Apache License header in revision I dont think thats a blocker for the current release +The second half of contribhadoopstreamingoutputbin is the part that actually interacts with our users create objects and write them to stdout using the Avro API +Even though used through the the question is really about +If you remove all elements from the view scope does not means the view hasnt expired +I closed the pull request without pushing due some issues I had in testing of the commit +Closing per above comment. +There are many more places where we need to qualify the source we generate but the patch attached fixes the minimal set to fix this persons problems +Thereafter it could go into the reference guide? Good on you David +Either way well give it a shot +I didnt add a factory interface because its not necessary yet +Hi Is there anyway to specify not to create a zip file of the signed jars? because i dont need zip file and the webstart will by work fasterIts possibly? +looks like this is fixedthanksdims +A few notes added code used tabs for indentation please ensure your IDE is configured to indent with spaces it did not compile on GCC until sEdgeKindKindEdgeKindg would be nice to diff against svn module root which is workingvm to save committers keystrokes +Piotr I dont see why we shouldnt fix at least some of your issues but Id rather have Janne Dirk or Andrew have a look at it +Further work in this area can be assigned to a new one +Verified in CR +JT does not compute that computation is done on the client side before the job is submitted +suite setup +Can you apply this to too which is what we are using atm? +Closing old issues +Seems that when I fixed I cannot reproduce this problem +Perhaps i ran the test on windowsAlso you will need to redo the patch with latest code thank you for taking time for think I only ran mvn clean install when I did the patch +Tried it once laughed and moved back +Add a interface or something and check whether the channel implements it polling if it exists + are some use case where the current fix will failNeed a cleaner fix the dll that need to be ignored are also present in the bin folder of the webasp project +Cool +I guess we could just use a but then you would be relying on something which is not portable +Yet + +months later Im uncomfortable with changing this without looping back on the subject on the list and possibly getting confirmation of the output +I dont understand why we need to average any idfs? this seems really costly IDF lookups and averaging etc should only be calculated for the top n terms that finally make it into the query +Thats perfect Dart +fixed. +For those interested here is a patch for which includes and the fair scheduler +Ill look into correcting this problem +This is fixed in m +are done Please see the changes in Also a couple of bug fixes with the path +That would be of huge help +Please apply the patch +No lets keep it simple +Please verify and close +Lets move it to +Manually tested! +Still an issue in though hopefully frameless execution will bring about a better way to solve frame performance issues +Perhaps we should consider printing more information during recovery to help an admin understand the state of the editlog +Im using Derby from EBR +Hi AaronDoes the new version generates xdocs or relies on Cobertura for the HTML generationI am asking because I would like to write Dashboard aggregators for Cobertura but I found that it is almost impossible with the current architecture which doesnt generate any intermediate data file that the aggregator can parse to obtain the coverage Nascif +So the correct behavior should be what is proposing to doI think should go into but I am not sure if anyone is looking into it + on patch +This UI needs to be redesigned from scratch +Example now deploys for me successfully +When a User is being deleted also the following trails are deleted Authorities Pages their Regions their and their The owner field in Widgets the User owns is set to issue the native queries fail in due to lowercaseuppercase issues +Has a couple of fixes +However the comments by Chris needs to be addressed seperately +Dont use a thread for every new client Connection By default the max number of threads now used for network io is the number of cpus available +To be addressed along with in a point release subsequent to thataddresses support for multiple simultaneously active dialogs +Should this JIRA be closed +Both type of issue and fix for fields are to be changed +OK if we modify tuplefromarrayheader and tuplefrommethodlist here are the members that would be defined as an empty tuple rather than a None value when the underlying Apache data is not definedI dont know if we should change the behaviour of those members or simply alter the unit tests though +howver there is small problem If timer is deleted CLI retains the resource in view even though it has been removed +ready for code test review should have posted this comment earlier +Depending on API available one or the other class is instantiatedThe and classes are moved to a new configuration package and split +This bug dont make sense now. +Can you confirm that Ramesh +bq +So the difference is that in this version the invokes and in tomcat does this right? yes this observation seems correct. +It is all working thank you and there is no particular jar that needs to be updated for us +Bumping to since we havent investigated ve seen this as well +Sigh +These blockers and critical issues are resolved but not verified +Fins on jo s s exclusiu de Mallorca o potser Menorca per no el fan amb un programa del Karakia parlant de cuina iraquiana amb en Pius Alibek va dir que ells tenen un plat molt similar i que el seu nom vol dir alguna cosa semblant a tombar o girar +Marking as resolved +I agree with marking this as not a bug +fixed in and +for patch if this does not force installation of Google App Engine SDKdependencies +Good catch ToddIve just committed this to trunk +Updated patch for trunk +I will see what I can do about creating a reproducible test case +FYI I started playing with and I was able to create a chart programmatically and update it via DWR call +I have attached a patch to the branch in addition to one that targets the trunkPS If you like I can clean up the trunk patch a little bit I guessCheers cleanup and submit the patch Edward +To use it we must include it as binary into the Cas EditorSure I then have to publish the source code somewhere elseCan we do this +You can reach me by email or mobile +Closing this bug as I believe it is fixed due to recent changes in the update site structure +I think I was planning to clean up the access on password making it private so access could be sychedAgree that it is a little smelly and inefficient to have instances hold factory references +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +Will be done with +will do. +I applied the others to a fresh copy of and it worked for me +I think Ive nailed a fix for this now +Possibly +But provided that it does the job and has no significant downside Id be very just from the mere fact alone that it natively supports exposing metrics through HTTP and JSON while still retaining JMX visibility and from the fact that you avoid the Thing acrobatics +All unit tests pass +Ive made some text constants so the tests can import them the tests will not break if you change the textAgain I am fighting a losing battle with git here and deny all responsibility if the patch does not take +Kind of a dupe but is partially done +Mauro I cant reproduce this with version +reopen to change CLOSED LATER status +if not a fix some small workaround will also do for the time being till its fixed +Duplicate issue +Cannot be s been months and JBDS has been released +N we should make sure your patterns make it out into documentation there is a dev section in the manual that we should add them to +It is jBPM specific it is an error that it is produced when you have a recursive need to understand the implications of releasing the lock early though which is why this is still open +Will post an update soon +Committed to trunk +The stack trace is not to standard out or standard err rather it goes to the log +Since we have to push out I will change the issue type here and delay until +The build is now using Spring in preparation for our milestone release. +Here is a patch that handles the double start and fixes up some exit values +See attached patch for a first cut of the proposed implementation +Assuming proposed solution is suitable. +Logged In YES useridThis very much sonds like a cron job that is removing filesolder than or days +What do you think Wensheng +Done +cool! for the patch then +Just to get the check instead of getting the stream am now getting the block from which i can get the stream again +In other bug tracking systems I have used there has been a category that indicates that it is a real problem but not with the product but instead the OSJVM +very cool +Yes removing does not solve the problem + should not become dependent on Server rather should own the and Server should be a client of that functionality +bq +I also found that the type has a field called type as well I think this would suffer the same problem so I changed that one as well +For example you dont need a redirect after successful authentication +While all of this is new content and news review can you please pay particular attention to Chapter +Hi DushshanthaThanks for the confirmation +Close all issues. +Another remaining issue is scalability of collecting members a group +There are actually issues described hot deployment of extensions is not possibleI cant comment much on the feasability of that one but it looks like a developers candy feature +HiI modified this patch quite a bit so it could go into the extras module +Thanks all for your input Bruno and Sean! One thing that we all agree on is that this JIRA is not the right platform for discussing and solving the larger problem at hand +Please feel free to upload new patch +Hi Eli I think it is a good idea +This is not true +ThanksIm surprised you werent getting the failure before your patch +I think it is less confusing to have the src trees not overlap packages and therefore it would be better to have it in +The new patch is called +Bootstrap shouldnt cause anything to be written to node AHmm but it does cause A to flush +Added a patch against current fulcrum applied to fulcrum trunk and turbine branch +Thanks for the reviews +This method will have to act as if it works outside of the which is different form all other methods +Its saying that the snapshot is being modified while were flushing +In there are many classes should not be exposed +There are plenty of applications out there mixing the two so well probably need a test caseapp that reproduces the problem to investigate it further +sorry Mark ive confused the files i was looking at as being part of the sources tree +After a rather annoying amount of work here is an executable test case attached +Sebb agreed on there possibly being a feature here +The user is automatically added to the default group +Committed revision Committed revision +Only minor changes to the code that reduces findbugs warnings +I am new to JIRA +I have a patch that starts on some of this that I did on the plane the other day but havent had time to write it up +The caller of getProxywaitProxy could cache the server version by itself +Here is the patch +So the usecase that is not handled yet is when you create a key in aws manually +I didnt check out the trunk I checked out the tag because I wanted stability +It seems like the patch file doesnt contain and +I was wondering if someone with more experience on Hedwig can have a look at the scripts being removed to make sure it is ok Utkarsh +For backward compatibility first we will break either Tuple or the impact is equivalentNo +Resolved at close of that have been completed or are now invalid. +Hi Jeff can you give a try this patch and tell whether it works as expected +Please post ant test and results +Ive committed a change to the which gets a free port and configures jetty to use this one instead of the default one even if thats not our problem doing this is better m fine with just keeping the maven based ve removed the code from the it test again +I think the works ok on could you check there please +Did you try my samples +How to trigger onclear on server sideThanks! +Do you see the same amount of failures +In what way is it not passivation capable +I also like to have the option to deactivate the type system map because in the situation above I really like to open the CAS with a different type system and dont want to open it with the smaller type system nor see some errors +Let me know if you have any trouble running the sample or reproducing the issue +Have the Start class implement a state machine +indicates that this test timeout was introduced by and indeed I confirmed that reverting this commit allows the test to pass +on the patch that fixes the tests +cool! +Ok I applied the patch and committed to revision but there is still problems with appclient getting access to the classes so this is still unresolved +Fixed in +fixed rThanks. +After all we dont have to have a ZK that is available just +I think the solution needs to cater for the fact that more than callback may need to be configured +Looks like I could successfully connect maven reporting with the problems mentioned above are not existing +It prevents using lists as a hash tables element +Checked in patch v to branchThanks for the reviews Jon Matteo and JesseWe can use for remaining comments +At least it was understood and usable what we have now in Maven is neither +webservicebpel is showing the same issue thats all I s fine with me now +Committed at subversion revision +Over two years old code example that never worked +Thanks Nigel for the patch! Its applied into the master and the maintenance branch. +So listStatus continues to use add instead of listFiles +What if a k cluster w each server carrying hundreds of regions +should handle the locality so that most data shouldnt go over the wireBTW blocks not used by any file are not known to the name node are they? When theyre reported by a datanode the datanode is told to remove them +Thanks all for the review! Committed the v patch. +should we rename type to value to make it even nicerAt the risk of less evolution +Please feel free to reopen this if it is still a problem. +Yeah maybe deprecate it first and print a WARN message then eventually remove. +patch applied at with. +Thanks for your patience! +to test import a instance in inventory +YCSB workloada was used for all tests which consists of readupdate workload with the total number of operations set to K +the flow we implemented that takes care of davids concerns isif there is an authorization then the hours or business day limit is checked for expiredif the athorization is expired then a is performedif it fails then the order is not set to approved status which means it can not be processedquick ship also runs the SECA so the order can not processwe did add a flag in the product store to capture instead of authorize then capturehowever no order can get to approved status till the capture is completedthis is handled in the entitysince the retries send a message there is update as to the problem of gettng an order authorizedthis may be a bad card or mis information that the Customer service person must call the customer and get striaghten outthis is the process that many online store I consult with handle thier orders +Note that will only track LOB locators in its current stateI think it can be extended to track all but it is not clear to me at this point if that is the right way to proceedAnswering these questions would help determine that a Shall an exception be thrown for any type of column if accessed twice? b If not which columns types does the restriction apply toI think a more general solution would be needed if an exception shall be raised for all column typesIf a solution is needed only for I believe can be extended and used for the purpose +Ah! Thanks +It probably gets activated but the systemProperties property is overridden in our pom file to specify some additional command line arguments +It is a current limitation not a fundamental issue +For Change the way znodes are read in +I dont think I can offer any technical arguments for this preference +The failure of is of course not related +r updates screen shots and documentation. +Committed rev ffddacac +Added a section after the install instructionsOPTIONAL Install Native ComponentsInstalling the Enterprise Application Platform native components can provide faster performance +The issue occurs on all OS +Zero in on precisely that element that creates this errorwhere addition of removal of that object causes this error to occurnot occur +Account created +Heres a fix +Thanks Michael +I think would still be needed for something like for noting the edit source so values are not back to the peerstackbq +It looks to me that it is more straightforward than the name of the name attribute of the fault element +Marking resolved in preparation for the release. +I did a quick check and it seems to be fine +Could you tell how did you verify it is working on tomcat? I dont see it working with tomcat nor tomcat +It will validate that the bean declaration is setup so that spring will inject a reference or value of the correct type when a bean instance is created at a later works for beans inner beans List Map Set and Properties +Change client URL from failoverejbd to ejbd +Addressing an offline comment by Bikas on extraneous logging +I dont find it strictly necessary to have that test case but it would be great if we had a test case or not in the right place +You are mischaracterising my statement +small went stale +I fixed this on both master and branches +Whats the benefit for making all thesemethods in HRX public from the perspective? As far as I am concerned itjust makes public API unnecessary complicated +Same error for meAny chance one of you could create a small sample project that shows the issue and attach itHow about some wiretraces of the messages +As we discussed earlier it only goes one deep caching immediate descendants but not recursing +I get a stream modified exception during the renewal +Fixed a problem where an unknown agent resulted in a null pointer exception +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +This requires more tests so move to +RSA is an asymmetric block cipher +Created wsc wscNonce wscLength wsc new tokens wscPropertieswsc +Closing +I made the setters and also removed a javadoc comment for the new method implementation on since the inherited javadoc from the interface seemed more informative. +Committed thanks +However Im waiting for that investigation to complete and have to cover that issue +new requires a new security permission to read a system property +Set correct fix version +Dont think that the patch has anything to do with the reported unit test failure +And isnt this one of the main goals of import as +Same pain you are experiencing with the Nutch dependencies which are not in the Maven repository will be experienced by Nutch users which might want to dependreuse Nutch code in their project +I am not sure I understand why there is a need for on the Aggregator +Being such a core section of code Id rather make fewer large changes rather than many small changes +Where is the definition in the spec as I cant find it in the dr +Since Id got so far I thought I might as well continue. +Cant figure where it says protected setter method can not inject a property +This may be an issue with the joint compiler +Here is the please review and apply this patch +This version makes two improvements to the previous version of the patch Logs a diagnostic message to if we cant get the locks needed to flush the unused preallocated sequence values +Fixed by +We can have a which can default to theWill do a comprehensive review tomorrow see if you can include the above in the +By definition anything in the internal directory is considered private so its not part of the public APII would defer to Alberto to comment on the change since he made it but it was part of revision and the log message indicates it was to improve performance +h +We could create the intentmap completely in code +right +I did Boolean first then replaced with enum because the null checks looked sketchyI guess I can replace it back if theres strong opinionRest fixed +And we are not sure if there will be ever a +trunk Java +Thanks Forrest for reporting the prob. +Username prudrakshala +I cant reproduce it now +d Run simultaneous jobs and check if all windows are populated with proper values of tasks +I forced a build of nightly but still investigating why the builds stopped +The reasoning here is that if I put in flush into the name it implies that this decision is done at flush time but it might be in the future that we make the decision earlier ie +I dont understand why anyone would use them instead ofSince youve done all the work I suppose these patches will be applied I just am curious why you dont use +Btw when testing this optimisations itd be interesting to quantify improvement if possible and make sure you test one optimisation at the time +This is more recent than releases closing +committed builderseap Rev If we change the build to use the esb install process to install directly into esb this needs works as expected in ER. +Thanks for the additional information +Not finding a word is a better behaviour than maybe finding too much terms? OkHow can I explain a customer that he cant search a word that he can see in the index +This is probably the change which will be most useful to people looking for this materialo Add a little language to the subsection explaining that we are talking about what the SQL Standard calls invoker vs definer rights +I added a instance to forminput so that someone can try to track it implemented a workaround +A preliminary version of the token filter +Lets not sidetrack the issue +Once the release goes out this problem will be closed +Thanks +Some more questions Do you compress database tables? Is the disk enabled or disabled? Does anything unusual happen before the corruptions +We need a newer version of AS than CR to use Final +Ted +Awesome +Can you attach the generated source Hardy? Does the project compile if you dont apply the patches +Sorry I dont quite follow can you explain a bit moreWhat do you mean by lost the row count? Does the paginator not work +Thanks for the patch Shaohui +Set fixed version +This could all be encapsulated as part of the complete call +Now it works on developer machines but is broken in production +Now that work is well underway with AS all previous community releases are +Im not familiar with how to write tests so therell be a bit of learning to do first +Reverted from both trunk and +Can you please check and confirm thatThanksDeepal +Agree the symlinks thing would be a mess and would also cause a problem for Windows users +Has been open for almost years no progress +I wonder if its because there are no values for the attributes? I can take a look tomorrow +That way you will get the stack traces +I committed the really wish you guys would stop using camelCase in command line options +a more questions? b an approach towards a a solution +This keeps the FS implementation trivial +What to do when the problem is detected is unfinished +Thanks Kim +It seems to have fixed the issue +Thanks for the review Sid +Not sure what a smart way to do that would be though + +Sounds good to me +Another approach would be to add a translator property to disable binding string values +but I agree there is something wrong here +Now tests fail with This is not a platform +Cut at documentation for cookies +This appears to only have been an issue on head country and not branch +I imagine these numbers are hardware dependant +Were deprecating the file based server so this wont be fixed +r for API documentation +I am on IRC at fusesource fusecamel and the name there isThe FUSE MR has beenRahul +Just now I ran the tests but in my machine it is not hanging + +I will close this bug as wont fix and look for other ways +Leaving open and marking for so I remember to update our dist to include it +This decreases the number of transactions required to read the status of all of the leases +fixed +Proposed patch +Committed revision +blog good idea gonna try to get a password firstnew feature good idea too +Is there any value in letting exceptions out of a close? What will the application do with an exception thrown on a closeThe application can know that the data it wrote to the file could be bad and trigger appropriate recovery +The loop terminator does not catch this because the loop is in +Hello TimThe fix looks good +See next exception for details +I cannot duplicate this problem +If we fix that we will have to fix it in branch. +Closing issue out as part of JIRA cleanup +I just noticed that your patches added them so I removed them +Shouldnt we add Pauls fix +Does the content in the dirs the looks at actually gets updated +Just want to make sure this doesnt hide a real unknown problemI think was wrong +Makes it cleaner and is also a no op change +Having said that the project charter can be changed and exceptions can be made given enough interest from the community +This seems to be low on large clusters but I dont what the optimum isIf will not solve the problem we can go on with the modifications to the code +Benjamin Bentmann Attached +all issues to the CR +Me too Validating newly compiled units picks up on classes that are not in the include path whether I am running tests individually or as part of a suite +Interesting +run mvn compileand see targetlatexuserguide line to work now thanks +How about I add back in the main and run as deprecated leave it in for developers and change the wiki to use bingiraph for the example with an eye to removing it as soon as the example jar is set up +Have you respun either the JBT or JBDS with cleantrue since this change was implementedI remember I respun TP and I think I set cleantrue +If my theory is correct that should still leave you with some failed tests due to the interrupt but the cascading errors caused by Derby shutdowns should go awayThis fix is part of addressing the issues I mentioned for underThe real issue here is finding out where the interrupt is coming from but I dont know how to do that + uses heartbeat for block meta data request +Bulk closing stale resolved issues +I didnt see an errors when I ran the initial mvn install from the base directory but the examples job files still werent builtAlso note that Ive already altered the wiki instructions to include a mvn install from the examples subdirectory so youll need to leave that step out +Can you confirm if the problem still exists with Grails and separate Grails Quartz plugin +Thanks +Thanks +fixed +The fix is trivial +Still need to change site once release to make API changes available +Please attach the should be support for documentliteral wrapped with attachment issues solved in released versions +Harsh the test should verify if the patch is workingIt makes sense to check different values of packets and their impact but it can be done separately +JustinTed is the intent to deprecate the old API in favor of the new APIThe changes look fine btw just curious as to the intent of the new methodK +fixed at revision +Patch moving the reporter into and moving getCounter methods from to +The same happened to would not be a problem if i would know where should i search for the interesting part is that the class where the error was is not even a ROO generated or annotated file but a spring bean created in xml +Yes Ive verified that Im using version using the class +Bulk close of old resolved issues. +Here is some changes and additional clarification and releases before that have opcode conflicts +Will reopen if seen again +I do not have access to Starteam so I cannot supply a changelog file for my problem is not really my locale its all locales +yes I think so too they moved all the docs to a more central location for the EAP +At the moment I am not able to reproduce this issue but the solution youre suggesting seems reasonable to you confirm that synchronizing that loadClass method is enough to definitively solve the problem for you +Ive slightly changed some of the signatures for example type checks work on classes rather than objects and identity returns a String rather than a +Checked Process ExplorerTrying to uninstalled Intel SDK for +Sorry for the noiseThanks! +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I just realized that the fixes for this were still uncommitted in my checkout +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +That is why I am in favor of option now +derbyall passes +Fixed in HEAD +Functionality is operational this bug can be closed now. +Did you delete the old patch +This contains an example Memcache and two example using previously described keys comparators and deletesThere are wordy implementations of different queries and their execution logic +No its not +Assume closed as resolved and released +Please take a look at this +Some libraries have been moved to buildivylibPig and pig script does not take care of it correctly +This feature has been added intentionally +All tests pass +We would like to avoid the XML document and pass the directly to the Knowledge the moment we are being conservative on what api we expose as changeset is a work in progress +No tests added because this is mainly related +OK Ive reproduced the problem and the fix seems to work +Hi KoonHeres the patch that fixes the issue +dont see any infrastructure in place for testing the contents of the web pages under various conditions +ATMThank you for the recommendationsSecond pass is attached +Its more config for the user with no benefitbq +no but I have aglobal Processor processorand was thinking that the feature would allow me to call methods on my global +Close. +committing +pushed +Improved in rSince it is not something important it is improved only in. +the current build scripts do not built the jar into the dist directoryThey should +Committed to branch with revision +Opening the editors worked as expectedI then quit Eclipse with the editors in place and restarted Eclipse so that the editors would be open on workbench +Id love to see an annotation like would be the local equivalent of the me I need it to work for Maps as well as Lists. +My spring knowledge is limited and I do not keep up on it unless users like you ping me +Does that also drive generation of the Maven? The reason I ask is that you found several other subprojects where this issue existed but I dont see those same dependencies in their published +Committed trunk revision Now backporting +Added preinstantiated lookup into +Ah I see +created +Seen on laguna skin in browsers werent checked +Looks good to me +This hasnt been seen for a long time +bq +Becuase of the way that the existing documentation talked about user authorizations I had a difficult time trying to use the phrase connection access mode +Sounds about right though. +Whats your hadoop wikiid JonSent from my iPhone +AlexeyOleg Thanks for tracking this downOleg Ive applied the patches in r +The patch is mostly for my benefit so we can use the functionaltiy before the next release +for the patch +It is now possible to easy add statistics for other field types if desired in the futureI have also updated the test class to include tests for String and Date fields +Fixed in r +The same as +I found several spaces at the end of lines in the +Then Id like to simplify the trunk integration by having a jira per component with all required changes +Today all new projects should use CRCC it can be done in cpu hardware SSE and polynomial is improved from crc +which is not included in previous diff file +Im reading learning and getting familiar with the classes in +if its not in Java then Im outOr better said Im alone to work on botLet the bot war begin! +The test has been passing regularly under since that time. +I will take a look +Bouncing back to Rob to close as we have clean continuous builds w the latest fixes +Should have been closed a while back +Please integrate the change request +One is not embedded but may show with the acrobat reader adding Asian Font Pack +Also it would be good to create a test module to run an integration test against these JAR filesThere is thissrctestshbut it should be modified to take the mrunit jar and hadoop jars on the command line as opposed to assuming them +Threw a in but the problem remained +A new release of is needed to address and no longer serializes the annotations that look to be causing this problem. +Avoid using environment variables +The comments above are prob +see subtask is over patches were applied as applicable. +First draft of a patchAdded to common package to share field flag information between the and the Added to the to extract the flags information for each field and make it available in the ResponseAlso exposed the indexInfo through a getter since there are still other pieces of info that are not exposed most notably the detailed information on each of the +There was a small bug in the callCounter handling the method which caused the max Call limit to never be reached since the counter was decremented after the call +Ill wait a few minutes just in case you already are +I am uploading a patch that clears for anyone else looking at it +Patch v looks good +You will see COMMA DOT in the URL +I think the synchronization should flush and then close +Bela you can close this one out it should be all fixed now +It does look odd though having the data member passed into functions +Thanks for the help +Support cases +Anyway for a patch implementing +Committed to and trunk +Filtering on message content is supported for expressionsDo you have needs for another dialect +We will have a Fuse Message Broker released today to fix this and we will package it up with Fuse ESB release is being prepared with fixed Fuse MB jars. +Ill commit this tomorrow If nobody objects to how I add weights in and +It doesnt matter if the child contains beans that would override dependencies in the parent bean definition it only matters if the child defines the requested bean seems to me that a change in those semantics would lead to unexpected behavior +Thanks Brian patch appliedProbably should add your second comment above as a new issue. +There is no upgrade schema issueThe setup you are testing on is now partially upgraded due to previous upgrade tries so it wont get upgraded nowPlease retest on clean setup to see if there are more issues +Returns already resolved class on fast pathEliminates excessive call to classloader +r +and well try very hard to create a helper class which does that because now we cannot inherit from two abstract Collectors each achieving part of itI also feel that when Ill upgrade all of to Collector the decision on what to do where will not be that difficult its pretty straightforward I think +You can use a as well +added an improved footer a link to security and a trademark symbol in the text in revision TODO Logos And Graphics Powered By +At end of transaction this memory is freed +Patch moving the tests to +Will pick it up in for sure +I attached an example for OPT mode + +Done in revision +Without reproducible case this will be hard to diagnose +Just trying to get the transfomers between model versions and to work. +should have tried that first +Please note however that the test sets were created in response to a problem originally found in the environment +This patch introduces a local variable local to avoid the race condition +ve made some corrs in diagram +I am talking about the model +bq +Best way to help is to create an test case that demonstrates the issue so it can be added to the set of tests +But in your code you are not dealing with objects +Committed and pushed to master +committed svn which backports the change from the trunk to the branch. +They are unit tests +Resolving as Fixed. +Can you please provide a full that fails? Can you please also precise the JVM and Maven version on which it fails +Ok official Mark Little response full ahead with the original Seam Forge this may change later but it will be quite some time from now and we should get started on this as soon as possible +Any news on plugin release +will fix my spelling +I know that it is not up to the framework to strip that encoded session id but would be nice if it is possible reopening just to attach the quickstart I used to understand the dont see how this can happen at all +George will you be able to apply this patch to trunk andIf you could apply it for the new in trunk that would be nice too was a bit quick +Im happy for this content to be left out +With this optional flag the script will not include Pnative in the mvn callsI tested this on OS X and Ubuntu +I can reproduce the issue now +This issue has not been implemented as it has been superseded by. +Maybe it is for the bestAlthough I would recommend that we discuss this on the mailing list before implementing it as I think there might be more than a few people that will have to make changes if we implement this and they might not be happy about it +I have downloaded and tried the MSSQL JDBC driver with the example and it failed in exactly the same way +Assigning to Paul to pick up the additional commits +Im rejecting pull request just to indicate its not mergeable yet +Bulk defer all frontend issues to +I knew this wasnt quite right but then I didnt think it was quite right before +Please reopen if you have additional input +Thanks Koji +Ive also updated attribution in so that its listed under release. +Fixed with regression tests. +Is there some other package where these exceptions would naturally belong though? Like for I think that fits much more naturally in whether its in orFor all exceptions that dont depend on protobuf classes I think they should be in +Would you need to access this ftp server as well? If so then I have to work with IT on how long to keep files etc +If no one objects Ill and commit this tomorrow +Hi BrunoThanks for the comments I will look into it +Please let me know if i can take this bug up and fix it +the has had major changes to it and this patch needs to at least be upmerged for patch according to like the patch I just have two comments before I can give it a +rconcurrent ok rcrypto ok rimageio ok rOthers are on the way +I may have to figure out how to disable foreign key constraints for this particular test case +Cant wait to see this finally a realityBest DavidThanks good to hearCould please move my whiteboardfthomas to git thenBtw I noticed my whiteboard is already read only is it normalCheers a heads up +COPYING +Ping? +Doing means we need to use multiple objects +Committed to trunk with revision I dont plan any more changes +Dont know will it help you or not +I mean to me everything you describe here points to this as a problem in the JCA implementation being used here +Will create for removing Exception from notifyDone methods + +To set Current to the first element of the collection again you can call Reset followed by +Resolving the ticket. +AronLets go with your existing portlets that renderJIRA project details viewJIRA issue details viewYour questions above Q Ans It is not required to have a single portlet +The NN then dumps the contents of this block report on the floor but returns success to the RPC +ternary operator doesnt work in cases I second that ternary operator is not enough +This fix breaks builds that use Boost as theres no boostchrono +Good eveningMy name is Paul Im having trouble on the set of the Eclipse environment +Could you check the mount point on your kvm host +Cant remember if were meant to be supporting and +Ill add the entry to the FAQ. +Thanks! +If the server supports both the spec defined constant and the old Shindig defined constant then the client should support both for backwards compatibility reasons +Without the grammar pool in the configuration preparsed schemas are not accessible when we parse the instance so the schema document ends up being parsed twice hence two error messages. +I am a beginner with Maven and am getting this issue with Maven +Manually tested +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks +The changes should be in the addImport method +Manual Checkpointing test case. +Will attach patch shortly +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I would be very much against having the Manager exposed to Layouts +This is great news Lorin +Attaching which cleans up another batch of warnings +So lets have a quick fix which will disable WTP Facelet validator and get rid of such an error on a working code +But please note at the moment you need the to compile and the to run +I will have to check this out +Benchmarking is a separate problem +Fixed in the trunk +HiAny news on this change? I too would like to use with Junit +If we can find a quick reproducible case we will fix it in +Please reopen if there is an issue +If nobody wants to tackle this I will close the issueThis comment is part of an effort to close old issues that I have reported +custom not default configuration +We are ramping up a maintenance release +I was using that do to this merge +Applications then can use it with their existing classes without the need for implementing interface +Actually one thought about the directory mod time +It makes sense to actual cutover from to before migration away fromBut you previously indicated there was some issue in Solr that made that problematic +With Maven and and plugin version I still get in scm element properties replaced with their actual values +Sorry again for the path without fileThis patch upgrade the httpcomponents to version stable +Basically I think we need a way to name a vector and have it carried through +It builds you just have to build pig with the test classes first then test piggybank +Retrying attach since Jenkins build failed on ubuntu failure unrelated happening in trunk without patch also +When the schema of each operator is recomputed the computation uses the attributes of the operator along with the information about its inputs +Please check that the patch was applied as you expected +I suppose no new release vote is needed since no bits are changedUnless anyone sees problems with this Ill make the changes tomorrow +Thanks Nick for working on this issue I think this is nice to be able to highlight queries that target fields which have the same content but are analyzed differentlybq +After applying this page nodes listing table will be sortable like other pages +Please commit it and let me do the necessary adjustments on the Jenkins +Applied at r thanks! +bq +Moving to Beta along with issue it depends on +The reason of that failure is that has in its classpathIts not correct if you use as a module ofSo you have to configure your project correctly +dims + +Reviewed for looks OK + + +I could not reproduce this with the current CVS code +Just now I cannot test with other version +Small update I guess for completeness request and response should be passed into the method as well +Clarifying if they are not they wouldnt support prefixed names anyway right +Mark Gavin Can you please confirm if option is possible to implement? If adding fake issues is not acceptable would it be possible to configure JIRA so that the starting issue number begins at a specified high valueThanks +bq +jsfplugins +Oh haha +Is there a reason why we as a client would need explicit synchronization to calls to addSynchronization? Our integration of does have separate threads calling addSynchronization and that may actually be committing +could you please give the full error message? Because if the message you gave is a method selection error +test is a new test case against annotation and test is a variation of validating annotation +It is also common for people to import XML thathas no declared namespaces and then want to go in and add namespaces to those elements after the factSo Id say we need the ability to assign the empty prefix to a namespaceand also the ability to change the namespace of individual names even if they have no current namespace assigned +If we keep it that way we should also take the corresponding line out of the +It works fine for me +However I think it should be rather simple to customize your builder in order to have thread safeness +Also at the end of the second new paragraph procedures should be changed to procedure since youre advising them to run one or the other procedure not bothIn the new topic you might want to add an indexterm or two just in case we are ever able to generate an index +So I had to cast it. +Hi Ayondeep What is the Axis version you are using +This looks good +which I mentioned on the mailing list +For example vs +I will test this asap on my side +KP Trust me I want this too I spent a few days of coding sessions trying to get this fixed +In theory the order in which properties are read shouldnt really matter and in this constrained environment if youre going to bother to clean up unused locals you could save these bytes as well but I could go either way +I just checkout a clean trunk and I was able to successfully apply the patch +Configuration AS RI the cache in Firefox solves it +This adds to a sentence which directs message writers to the website for guidance on how to write messages +I will chancel my pull request and submit another for resource clean up onlyRegards removed thanks for raising the issue +I think I know what is happening here +good catch +Richard regarding your point Do we need different multicast addresses then as you stated above? Why not just assign different ports +Checked in fix to develop branch +Had this not been the case then the API is fine +In case of dynamic partitions you can also have outputsThey will contain the correct Table definitionAre you taking care of them +I dont know much about this particular issue Brock will hopefully have more info but in terms of the release we were almost ready to do one and then a few blocking issues were logged +Its not ready for release postpone it to be automatically resolved through the new Security Model and reimplementing SSO on top of it +all your help is welcome and dont worry reassign them to you if you think to work on them Im not jealousBye +Turns out its only two files actually The Bundle activators of the JNDI Core and JNDI RMI bundlesAttached patch replaces Java Util Logging with SLFJ in these sorry +Hansthank your for the thisI also think the content component would be the right place for thisAnyway in my opinion the configurations shouldnt depend on the product but the webWhat do you think? +The idea was to speed up versioning operations like checkincheckout Ill revert the changes for now +The density estimators both use a set of representative points taken from the clustered points output after clustering +these bugs have been fixed and will be avaliable in CR +All tests run but more polishing and testing is needed to ensure all features are working correctly +The attached simple patch causes splits to happen during my test case +Heres an updated patch which does that +Mikedid you reinstall +To correct had to open the ticket and click edit and remove the fixVersion and then save and click edit again and enter in the fixVersion again +In order to support wrapper style mapping of method parameters in this unwrapping facility in is neededPlease correct me if I am wrong +Should I try an older version of ffmpeg +If a seqid do we need a mvcc or should they be related and if not how should they be related? And so on +Holy crap +Have you had a look at the maven documentation? In particular the Getting Started Guide? The section How do I add resources to my JAR explains how to add additional files to your artifact +Are there any workarounds for this? I am also able to recreate this issue +Hi guys what happened with QE test environment integration JIRAAre CSV prepared for integration to showcase and metamer? +and updated to include test case. +In a first call we request script that will create a dependency for script on itself and Bean as it should +JonCan you add some details to this JIRA +Cant really do anything about this one since it was intended to fix an issue on +! This fixed some workarounds I had in Griffon and +Devaraj! I committed this to trunk and +I think Tom can help us move things forward +I dont like especially for Ganymede +close issues for release. +Seems like itd be better to keep the interface simple and do normalizations on the mahout end +Alternetely could just not define theseYes but Im afraid of what would happen when users dont or use a staleI would rather have those calls fail instead of unintentionally overwriting the files +otherwise we should not add it here and remove it anywhere it is foundits definitely useless if we do not consistently use it for all deprecations +I will send you a link tomorrow +Claudio the reason for closing this bug is more down to the scope +no there are no older jar took an ear that worked in and trying to the debugger is nulli think the difference between the tutorial and my descriptor is that i am using for the default was the only way this worked with +It makes sense to optimize this query could easily be generated by some app +I think we can even get rid of initializing the array on each validation +periodic bulk close of resolved issues. +The recovered node begins receiving writes normally +HiI have the same problemLocally i applied the patch from Mayo Fragoso Jun AMIt is working very wellCould please someone commit this to the repositoryMarkus +For the patchIgnore corruption after the sentinel as long as it takes place in the last megabytes of the logTest this exception +The syntax isnt really important +Garylooks like some classes are missingI get compile errors that there is no replaces the first patch +Part of EJB Core + +Are you saying that maybe we will never need to have a I think it unlikely but not impossible +Thanks Mahadev. +I just tested this and did not see any problems +Im with you in considering this a WONTFIXcheers am not against fixing this bug +Please see my comments at +Are you sure that all other classes which use this class will work correctly in case quoted identifiers? +Alessio could you confirm this is fixed +Thus completely decoupling the old from the new APIRe Adapt existing implementations Absolutely agreed +We had something going on in one of our getters that caused the loop back into the session. +If you believe the issue is not resolved please reopen or open a new issue +We may need to change the platform build in order to generate these jars +I created a features folder which contains a set of pom projects that group modules into functional features +Many thanks for the errors reported are resolved +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Is this a timing thingTried with new projects both created from scratch in UI then following creation with renaming +We could wait +Thanks for the review Kihwal +Hey ChrisNo worries about the time +Note I did not apply the excludes list patch due to your previous comment +Ive just committed this +Which UI scaffolding are you using? MVC GWT or Flex? Which version of Roo is being used +Part of a bulk update all resolutions changed to done please review history to original resolution type +Fixed. +It does make sense to provide a long value for the block report interval for example if you have a lot of trust in the incremental block reports and thus have no need for periodic full block reports +I am really close to raising this to a blocker for Hadoop BUT Im worried this isnt the right ticket +When you attempt to register a driver with registerDriver it uses the callers classloader to find the driver +Lets try that again against trunk rather than this time +Fixed in trunk. +patch to fix double declaration of the in jxls for the patch have applied +trunk +Email are not IDS so two users can have the same email +bulk defer to +New patch Remves the chunking in Lower to bytes This prevents from malloc on every write! Unfuck the read loops in and Use when close test in that ensures if chunking is working correctlyThis is now ready +At revision all old Resolved issues +Logout and login the browser caches the basic auth and sends it with the second get +Especially seeing as components like Messaging have dependencies on client libraries now +fixed a problem when the last block is empty +I forgot that it is officially stated as not recommended +Thanks for taking care of this so quickly. +When more than one Ruby thread communicates with the server Ill sometimes receive the error. +I think these are bad bugs to be fixed as well +OK +What is needed to finish the Route driver that is currently in libcloud but disactivated +bq +The content can be +I applied to trunk +How about we create a package in core and then split to two classes and where is an abstract base class for all hadoop http server +Let me verify that +The implementation is a direct port of PROJ code +Once the method returns the logic bean is no more +I agree with Darran the only way to properly identify a servlet is to check whether the class implements the interface +Thats not valid in any language as far as I were you trying to accomplish? +Cato is a commercial product and it has nothing to do with this issue +Include the new tag into the new logicsheetIs this correct? Can this be posible +Odd that they respond differently +Snjezana lets at least get it documented in the extension point that this is not intended to be extendedimplemented by individual one of these extensions points should be active at any given lets get a warning logged if we actually find more than one extension point registered +I just committed this +NET code can use J implementation of or Still trying to verify if the toString implementation use the same representation +The cause is due to the wrapping on line which marks it as logged by there an easy way we could avoid yet another file? Perhaps making the elevate file a system property in the existing and just change it for the bad test +Thanks for the patch Harsh +Thanks a lot +For I think it is fine +fixed +Thanks Gail +So my query is is this defect applicable only between and versionsYes the servername format has changed between and and they will not be able to communicate +Would it be a temporary workaround to have the just create the same folder structure regardless of whether artifactID matches the of the projectThats exactly equivalent to implementing since the folder structure is determined by bq + +Target to +Applied with minor modifications +Trivial patch +Your committing of Mr +I think as there is a new issue that has the aim to make this work this issue could be closed +A recent commit changed the heartbeat interval for small clusters that looks like it made this test race and fail +I left this unscheduled but at the latest it has to be done by +Thanks for the patch Todd +Thanks for the explanation +Move all implementation of flushCache from to +General changes descibed on JIRA to Rick for applaying the patch +exactly there doesnt need to be a lucene issue to put this code somewhere else +See This has a patch to enable in Batik +You can copy the source and use it in your own package +Rebasing yet again +Fixed on branch a long time ago +Thanks for committing this Jean. + you probably want more moderators you dont make jakartahivemind instead if ust gets svn mvd to hivemind you probably dont want the snapshots unless you had them before +Having the same option on the eclipse plugin will solve the problem +failed tests are not related to this issue +Yup looks good. + +Attaching patchmarking pa to get a full rest run +Thanks for pointing that out +Added as ve placed a link to this page in Labs Administration on the main Wiki page +I am attaching a patch to show the work in progress +Hicould not reproduce the bug +Im going to try if this helps our job +Multiple threads may end up trying to localize the same files +No update. +For example you can implement a new http transport senderreceiver library and give its library name as the class attribute of the transport senderreceiverHowever you cannot add a completely new transport dynamically for example JMS without some code changes to AxisC +Added thanks. +Yeah just did some testing if I do a top level clean and then run tests its not downloading a bunch of jars. +is no more in the geotools codebaes +Fixed in faaddabbaeecf +not really easy to navigate with a combobox +Ill take a look at this for the m not sure I follow actually +The definition tags will need a mutable container +Ive now renamed the repository to +yes +Thanks +why is this really a Major priority + patch forks specializes template for Harmony copies out all from is able to build RVM in generated configuration +Migration document is improved +Also I suggest we instrument the code to measure coverage and run a server for a while to identify dead or seldom touched code so we could review it +Again files are binary identicalIt now loads ldif file at startup and I can connect to it and verify correct behaviourI am totally clueless +The session beans should also behave exactly the same but apparently the types being used in the specialized beans map do not match exactly what is being resolved +apparently not +What were the other issues? similar blockers or +Please vote if you approve these changesIf they are OK I will apply the patch and I will stage the source distribution for voting at the MRQL dev list first on Monday morning +Without the patch the connection is refused because does not see the createtrue property +Matthias what sort of metadata do you mean here? Is there some standard format +Created for clean fix in future blockers and critical issues are resolved but not verified +I think that I will allocate a buffer of say MB and then when the OOM is hit free that buffer so we can cleanly exit and log +Committed to trunk. +I just committed this to the and branches +Element definitions +Yes you are right +Why not just add the seed to the queue instead of its children? That would be simpler +Sean thanks for the patches +then you only see it if you go to the other tabs +The upgrade is documented in the we were thinking along the same way as far as externalizing the configuration +Committed revision . +However a client working with an old server will still see it as it will revert to EBCDIC +Thanks go to Brock for the review. +This rev incorporates feedback on the previous rev from Dan and John +and using just a single anywhere in the template results in a parsing error +It uses the Editor contribution to put the tools on thetoolbar +Yes please do +My debugging shows that the async request from ATS to origin server for the ESI include is kept alive for seconds before returning with a proper success fetch event +Spnego is now also supported using a +Missed the boat +This model has a model object annotation that is empty +Sorry to take a while to get back to you +Still the validationAware would be set from the stack regardless of prototype hate it when I do thatMarking resolved as per reporters comment. +There isnt drools runtime in +Found this one trying to course! The message is a bit confusing though +Looks good +I spent a fair amount of time trying to reproduce this issue +nice test +I dont mind to have new plugin that would be developed from scratch and have code covered by tests and to have it but I do mind to have something like +nightly docs label should be removed from release docs build +Maybe +A few very minor changes comments in comment and logic change in few very minor edits comments in comment and logic change in. +I agree we need to fix this for +Works fine +Remove comment in RPC address I do not very much like that you add as a members +Yes we do have these two rules +And just to be clear trivial is a committer process comment not a characterization of this issue. +Thank you for verifying the patch Dan + +The path is cygdrivedszehadoopDoes this patch fix the problem for you +Is this still the case +exception do you get in case of selects? Maybe we can extract some useful information from the exception message or stacktrace +The signs are not always visible +reverted last update so the content points to the examples page as it did before. +Forget abt my comment +Justin please give us the thumbs up when that is complete +Raw files that have been changed as a result of the patch +Also it seems fairly common that two repositories could be on the same the MD digest could work would the repository id be used in a file path that the special characters would blow up +We committed to implementing the schema validation in line with the processing expected in the pipeline this has been stated earlier the description in the PRD makes no sense within the context of the pipeline +Warnings are about logging potential mistakes that dont actually result in an error +I committed a few more javadoc fixes +I think as noted in this test should run server side anyway +I have pushed a probable fix for this in fafc but I do not have a Windows machine to test on +Thank you Ray! +th +Nice work +Weve already encountered the same issue in the openengsb project +should be ready to rock +specifically does not support this scenario it assumes a bundle will only be in a deployment package once and each new version of a bundle constitutes an updateI would opt for picking the highest version even though I dont like the implicit nature of this. +If no one against inclusion lets include it +Pull back if you feel otherwise +Needed to check for null results after the initial retreival of the count +No but it should be addressed before is released +what else was in there beyond a stock JBDS install? Maybe something you added was causing the problem +Attaching the patch with the change +So might as well nix those plans for now +Patch prevents setting a null association +command on the server contextual is osbsolete +In Branch SFC is also impacted by the issue so the port should correct that +FYI the situation is the same in the just released Developer id assuming that this is still the caseProposed Release NoteIssue with Developer Studio Graphical Process Designer and included examplesWhile editing the file of the websale example program shipped with it is not possible to edit task properties using a and the context workaround for this issue is to navigate to the properties view using the menus and and do the editing from there +In that sort of case should not be fatal but rather indicate a temporary condition cant really remember +I am having the exact same issue on windows with JDK is a normal way since appserverConfig is not a profile defined in your settings +Please check the mailinglist for replies to your inquiry +Thanks Ahmed +I tested in the shell and it fixes the issue +First patch in trunk rev +yea +That didnt show up in test nor in a crawl but im not using anyway +We can change runningJobs to get +I like the idea of making it understand both by prefixing the long with LONG +The mvndeploy is successful and I am able to successfully access the URL httplocalhost +I think this might be a problem for the code sharing proposal which relies on the ability of one jar to be able to load the class from the other jar if the other jar is loaded first +Getting the following errorERROR save Error creating user invitation Cannot make existing permission into pending permissionHave you previously been sent an invitation and accepted it +The attached graph shows the results +Luis can you take your changes and turn them into a patch file? That way we can apply it and test the changes much easier +I just realized that there were really two debug messages still hanging around +Dwayne any update to my last two comments +I hope this isnt patched by the Debian team to work this way +Do you think it makes sense to include the actual above that triggered the bug in the unit +The discussion there may shed light on how to address this issueNote that this vulnerability can be used to copy any database out of directories into public directories +figure shows the new sequence after this changeThe next step is to look into elemenating the deletion of current client engine on every method invocation + +I just filed a bug for BK Bench +submit the formThis should work send a validation email I get redirected to httplocalhostcontinuumerrorCode and the page displaysInternal Error The URL requested results in an internal error +is more clear than this +logXXirdumpBut when running the following error occurs PMF Invalid common pipeline for CSOPT does not have a root path +There are some advantages to the current serde such as being able to read the rcfile with a different schema although you are right that the binary is better choice in generalMaybe the right answer is to make a config knob that sets the default rcfile serde +Deferred to defer of unresolved bugs to +Attaching +This issue seems to be due to a thread deadlock in when it attempts to synchronously execute the Runnable +Jira still messed up the formating here is another attempt to avoid that +contingent upon real requirements. +again Im ready to start +Patch for trunk +It is not a system function so we can not do it +This could be useful for since it allows people to share configuration files without putting credentials in them +Mirror Is Fully Up And RunningVia Both HTTP FTP +Maybe Im missing something but your bigmess patch basically works for me +Looking forward to release. +John Galder We did not add to the infinispan integration code in AS +I will try out the patch +Remove it in v patch +Somehow I put the changes twice in the patch +Id drop the booleans for whether it has been used +I will work on it soon but i am thinking of working on tika parser so that it can get all the attributes by default index them and send it to solr attr dynamic field so that instead of specifying manually any attributes will be accepted +This patch adds the osgi data to the manifest of the maven bin jarThis patch also includes a fix for plz close that one when you close this +Thanks guys +Verified in +Changing the behavior of the formatting as proposed by the fix could break code that relies on retaining the doubled quotes +Are you connecting to a broker that is in use by anyone else? Is this broker local to the machine you are on or in your control? If so can you change the configuration to disable flow control and run some more tests to see if these stalls repeat? Ive had occasional failures on those stomp tests but they are so infrequent that Ive never been able to determine if its the stomp code or the tests that are at fault +Thats a good point against making a x status code the default for nowHowever it might still make sense to make setting a custom response code for validation errors a bit less painful than it is now where you have to copy a lot of code from the processValidate method when overriding it in order to set a status code +We do not assign this issue to one committer only the whole team is responsible for updating the code incrementally +To be honest I dont have a really good use case for more then targetclasses and target? Other then for generated code +I hope this will increase the likelihood of the issue being addressed in the next release +thanks! +I plan to allow balancer to throw away blocks that was moved more than one and half hour ago +I also noted in comment that to reproduce this behavior parametermappingtest need to be enabled for client +applied correctly +Tom can you look at this for Dart +Looks very good to me +Ill revert the patch +Git commit fixed in efedfdfdbabbfebba. + added to the list +HiCna you please let me know what environment this test was run inThanksTony +Blocked by compile error the test failure can be worked around by reading the in that test plugins root folder +I will commit this soon unless there are objections +just parse it once when clientimpl initializes and keep it in variable defaultHost which could be used directly when creating channels +What exactly is that stress test testing? I havent taken a look +See classlibmodulesarchivesrcmainnativearchiveshared line Strange crash in close function +This seems to be specifically related to the use of s lookup there Since the exception thrown from there is a s late startup detection fails because it catches only +Looks OK to me. +The dramatic gains in the indexer is evidence of this as the pipelined processer passes documents and views rows from process to process +I am wondering what are the reasons that people want a longer varchar? Increasing the length may give the user the impression thatit is supported well +IMO this is related to + Created in for internationalizing the text Click here For Field Lookup written in alt attribute of img element +Committed thanks again Maja! +works fine with no other changes +this is a know difference to ACF our implementation was prior to the ACF implementation and follows the implementation of the function listContains that works exact the same way in ACF and Railo +ThanksVersion does indeed work +Hi James Dyer and Mikhail KhludnevI added in of tomcat the line belowFINEAnd ran again in with threadsThe select below was performed timesselect url from video where indiceidindice This select of should have been executed only one time +Ill mark it for and so it wont slip away again +Scanner patch applied at repo revision rI disagree with removing the transient keyword from modCount so that has not been appliedPlease check that the remainder have been applied as you expected +Ok I assumed the timer work was removing the timer implementation Ill look into that bug and set up this one to do the same +RohiniGreat +Im sorry I did not read Jannes suggestion forReverted the modifications delaying until Fix version +Hi nice work it worked fine without the here +Added patch which removes leaves dangling entries in when newer version of module is deployed +Xmxm done +the code where you execute the +Ive got some reservation on throwing an exception see +I am just starting the wfs work so I imagine I may run up against this as well +Of course the Avro project is free to copy any of the ORC code into Trevni but Hive has the need to innovate in this area without asking Avro to make changes and waiting for them to be released +java files +Here it is +So its still open +Need to add changes to couchjs script or the option wont be recognized +automatic associations removal after principal removal would be same bug is still present in JetspeedI noticed that if Jetspeed is restarted you can add the remote association so it seems that shutdown forces data flush or cache clean is issue belongs to for which I created new issue +Theres not really a better solution right now though newer JVM releases do combine the two modes finally +User reported that there is no hang +Please comment and if no concerns on this i will update the wiki +Thanks Raghu! +patch for +Some test cases would be nice +If Im wrong please. +Also set the Lucene default to true as I want to improve sorting and FCRF +The last problem I had fixed in the patch introduced this bug +Fix attached +Ok +Current unit tests in cover the functionality +Just a thought +Anyway the trunk compiled OK for me only after I changed the version of slfj from to +But I guess we need svn repo for the CMS because we are using git for our source control +Change required to line of see this log statement change log level in to info and then connect a client +Some of our tasks can be heavy running with G heap etc +Move it back in if you think differently +Derby client not supported by Derby +added constructor taking a applied to head extra constructor available. +Fixed directory there a place where we can download the nightly sandbox binary jar +Even though I can see the new version you may have an old cache of the page in your browserIt should every time you visit a page IMHO but it doesntHope that solves the problemCheersJared +Note that this tool is specifically intended for migration to HA and not to support change from Haivvreo to +Feel free to do so in the future! +Having the validation in the processor definitely looks more natural. +This wont be fixed in ve updated to use +Further invstigation via simple axisc services shows that this is probably an Axisc issue +I took this up somewhat recently Im hoping to get a patch up in the next week or two +this patch must to be unpacked into root of portal project +Is this resolved now +Hi WendyIs this thing about having a string containing a list of typesan established xmlrules convention or is this patch the first time this hasbeen done +Thanks for the patch as usual Hamed great workCommitted to all major branches. +Leave it to me Vladimir +bulk close of all resolved issues prior to release. + unscheduled issues to the next release +Sorry that I am not being clear +Here is the patch +but that explains perfectly the problem +Fixed with revision +ChrisYour patch is in trunk revision revision +Will deal withweb archives later this week +The container automatically enlists all the resource managers accessedby the business method with the transaction +Added license headers to missing files in the trunk at rev +I just committed this. +Instead they should be able to depend on a pure spec jar +Izpack do not create with this fix. +I am using +Unfortunately there is no way to do a scan of classes and interfaces in a package with the current reflection possibilities AFAIK + for the to trunk. +If I had considered this was solving the current JIRA I would have just closed it and marked it as duplicate + serialize proxySource as only +We currently synchronize on coreStates when we publish the info to zookeeper +reopening so I can modify fix target to +Would you like to submit a patch? We highly appreciate contributions +Committed this to trunk and +Once the test case is working it would be interesting to try this with with drda tracing on to better understand the server round trips with that change +moving out of wont be ready for release timeframe can go into +Also theres a dependency on the for the integration tests +Works for me +As per James suggestion Ive changed it to refer to insteadClosing as fixed. +Ill upload a patch with a proposal +No as in my test things are generated but with some problems +on the above suggestionWe can trade some complexity for better compression rate +Integrated to trunkThanks for the patch WenjianThanks for the review Stack +The exception being thrown upon successful failover in startFailover needs to moved outside the try catch block +Error reporting has been we could actually deal with those files if we had the ability to force the SRS onto a coverage +what point are you trying to make? your comment is completely worthless +Pull request to allow for Solaris override +Some days ago I tried increasing Xmx to m which also did not help +Heres the simple patch to just synchronize looks good +I have already created a task for DROP INDEX in CQL +Ive uploaded a patch but it is dependent on the code I proposed for +And in srccontrib I dont think we need the jar target +does seem to add an entry into history and clicking on the entry in history does run it again +sorry it is my typo error +Then we can document this as a known issuelimitation of this strategy and close the ticket + +Should we undo this as of the recent changes from +no prefix size k around bytes +CXF has been instructed to use logj through flag cxf +Bulk closing stale resolved issues +TimoIm currently working on so Im very interested in files for testing +Oh I should also note that you probably want to turn off DEBUG logs or comment them out before testing this on any real log +About Dialog under Mac OSthumbnail +bulk close of wont fix unreproduced and duplicate issues +Is it okay for you if I supply it ina week from nowRegards Geurt Wisselink +tnx for implmentation to match the ACF behavior not the function use +INFO Outputing statistics for superstep INFO superstep total bytes sent INFO total bytes sent INFO superstep total messages INFO total messages INFO superstep total millis INFO total millis INFO workers INFO Superstep megabytes second INFO Total megabytes second INFO Superstep messages second INFO Total messages second INFO Superstep megaabytes second worker INFO Total megabytes second worker INFO Superstep messages second worker INFO Total messages second worker INFO flush starting for superstep totalMem maxMem freeMem +But it is still reproduceable with Internet Explorer +And thanks for looking into the problem. +Labels like Base Name Property name Resource Bundle and Value are made to be bold +Added the transaction panel to the console +The plugin therefor tries to parse it as an version ve added debug logging for this in r and deployed a new someone with insights into the version format can give a reason why this is happening +AS ships with which is the version were currently using +This issue is out of date and will not be fixed +menuGroup ajaxValidator and graphValidator components have been. +in svn +With these additional tests can we now close this JIRA? +Can you please provide a corresponding stacktrace +Si will you give opinion on this one? Its been months since your last comment +closing it. +is independent of this as it is designed to handle the test failing on an overloaded system where the replication takes longer than the sleep times so they both need to go in though is a hard one to replicate +Right the JIRA is for adding a better error message that doesnt leave a user release will focus on error handling +If we wanted to move it into core at a later date we should put it in its own package but that really needs and to be done first +or change it to timed waits and add assertion as expected after that to ensure loop exited only after counter reaches expected entries otherwise assertion will failBut we have to figure out why there are Bookie failure in your tests +Can you confirm that its actually in there? This may be something really simple +Ive confirmed that fixes this bug. +Examples depends on depends on depends on could move all classes except for the ones above but I think it would be really awkward for someone to figure out why only these classes are not on core. + +Closing. +Moving to CR since we have less than weeks till Beta +Streaming needs to become much more unified with the rest of Hadoop +The Change put the text in the body if there are no valid headers +f +One should use seeds bad I didnt pay close attention when moving from to +Committed I should be more careful. +Patch applied at r thanks for your contribution please verify +However if I slightly change the setup parameters according to more recent innovations it works +Simon when do you think youll be postingmaybe within the next week I have a busy schedule but does this patch keep you from doing any work? You shouldnt just pull out stuff from month old patches especially as you dont even give me time to reply on the orig +If we are not arguing about correctness Id like to commit this and file a follow on for perf concernsI agree with the abovebq +This was a very important fix for us as we have several projects are currently being built using Spring framework and Spring webflow in environment +Reopening to submit patch to trunk +patch adds jdk and ibm to the list of supported jvms. +In switching the and downloads to the archives spotted these missing signature filesdistdbderbydistdbderbydistdbderbycommitted changes for in revision and for in revision +XSL problem is a duplicate of an older bug . +I think that would work for everybody +Lets both ask our ops teams +Thanks Eugene +Closing as this appears to be a Jenkins issue. +That would need to wait for +Thats a good point Claudio thanks. +Here is a I would hold off on this particular patch until we can answer a more general question of how do we want to approach these types of situation +Revision were used rIm agree that we can close this JIRA +ok but surely its a bug still? I mean all the other pagers in admin work with a mouse just fine +In the last trunk it seems to be fixed +and this was the only request for +Thanks! +Ill try and work out the correct way to find the information you need +Work in progress patch +See classes and packages under package +These links should contains the namespace as well +This is also consistent with the of the editLog member thats also done in the +Im inclined to agree considering there are separate options. +for test had a little hiccup during test suite runIt passed when run individually +batch transition resolvedfixed bugs to closedfixed +Here is reattached file +My patch for this patch might help in your case but is not a good solution in general +I havent followed this thread closely but as someone deploying Rails apps on Java appservers frankly the idea of touching context classloader scares me +Maybe the problem is caused by it +Fixed on and idea nice patch +ivy directories and related changes so that jars are matching uptested quick start script and looks good now on checked out this morning and this breaks the Mavenization from The dependencies for is nested in the resulting which is not valid +HiIn the case of bulk transfers please also consider to induce a random delay in the range of specifiable in secondsminutes +This has been fixed in CR and should be ready for verification in the most recent builds +As a subproject its purpose would be to pull down Hadoop jars and instrument them +Linked the same PR as for +it could be just a static util +Note sure Im following your concept of a class +B is stopped and loses all data +So Im unable to use the login module at all +Test failures are unrelated to this patch +User that has already logged in into Guvnor is no longer required to authenticate when accessing secured servlets. +I run with the new patch and everything pass for me with the patch +This is quite annoying. +The reason that it validates every previous page is for security +PS this is a new installation +I changed the default to a while back +It would also be nice to make the time human readable +Bulk move to +The patch does modify the menu model at but that is on purpose the idea being that these external transparently become an intrinsic part of the model being extended +Attaching patch backported to +Thanks Suresh +Ill address it in my client code and +Verified in CR bits in Hudson by provided test cases. +The keyvalue pair is still present in Cordova but is still a plugin +Bump +Thanks for the quick comments Knut +Backport and update the jython wiki page after fine +What the coding style prefered? Only mark them as private or hide them in a impl class +local localDir userNameThe above code can be replaced by calling a method in the tasktracker it already provides such nice utility methods +Thanks Shai +If you want to change the code to use your new library that would be awesome +Im continuing to look +Im just worried about stuff like someone running a Thrift server as root that gets passed an address of etcpasswd +So the job should be removed from the scheduler job list by next polling interval +I doubt its related since it doesnt stress any failure scenariosI actually did open a new patch for this issue on trunk which is linked here +Ping +Thanks! +Help in resolution is appreciated. +Resolving as Wont fix +FYI for those interested I tried updating to but still doesnt workIn my case Im suspecting that the issue is inHowever if AS would give a more detailed logging itd be very helpful +Closing as duplicate of. +You should delete the headings for the optional sections that you have not used +Fix and unit test extended to cover this as related classes +my recollection is that relative urls are relative to the location of the module in the ear +Changed component to bsp documentation +there anything left to do here +fixed. +Candidate for SOA +Thanks for the quick fix RobWorking fine for me here now. +Verified using a policy similar to the policy provided by Christian. +Waiting to see if there are performance problems +Please review the submitted patch +argh I guess that was my fault +I agree we should not infer types for constructs with an Native annotation +Below are some additional comments Can the version link in the reports be removed? I think it would always display the Unable to find +Support +The ZK client is actually very lightweight +The logging needs to be improved to clearly bring out these kind of errors. +Note sure if this applies to the other uses or not yet +Still running test suites +REST had an owner and got in timely and swiftly +changes look great to me +Verified on Jenkins viewTrunkjobvpe +Caption code was added twice without notice it +One bit of reasoning you have in there could be improved if the data ends with a then it cannot be compact protocol since compact protocol will always end on a byte +Mapping these will result in a classcast exception initially +Checked in tests pass. +Before I close the issue however can you please investigate persisting the split pane sizes? Its definitely nice to be able to change the pane sizes on the orderings and prefetches but those values are lost as soon as the Modeler is closed +Ill investigate +Merged changes into with revision . +Emails to jira are still bouncing with out of quota error +Therefore we dont merge regionsFor successive regions R R and R if we delete R we can change the end key of R to be the original end key of R and drop region R directly +Missing labels added +Added fix version back. +Thanks +Triage flag set +Patch attached adding set to and a test case +Ill have to dig into the history here of the older methods and see what is the best solution +Thanks Andy I think we have the same problem for SDB. +What do you think +Hi Oliver Patch committed in revision +defer past +just got the latest trunk +code looks good to me +JuergenI cannot for the life of me reproduce this issue +Thanks Mike! +looks like there are spaces in your source path Documents and Settings +I will also attach my +It sounds like those requests should be using applicationjson then +Thogh I have on more issue in the Layer publishing under the defaultstyle dropdown list normally it displays a legned preview but it is displaying only a image placeholder for dem i noticed this issue is only for dem style and flag style and the style I have created from dem +Snjezana apply this to M if you see this otherwise Denis will carry it over later todaynight +Link patch dependency +AlejandroHow do we drag Snappy code into Hadoop trunkMay we drag the Snappy svn repository of the google code as svn externalsI think that itd make Hadoop trunk more complexIMO is simpler approarch +I didnt update the build to auto generate the servlet examples view source htmls as there are only files and they already exist and provided by tomcat +Yes thats probably better attaching v of the patch which include an adaptor +I havent +Likecoretargetclassescoretargettargettarget directory includes + +Also most connectors will never be dealing with archives +But the build fails when compiling +To check the shared edits we can not use this logic +Here you are I hope one day this code can sit in Cocoon where it belongs. +The potential dead lock described above needs yet to be fixed +Ill run it through the same YCSB tests that show the lock contention + +We have not picked the build up yet because we do not use integration builds only release builds +I think I can close the issue isnt ityes thanks +Verified by Alexey. +So that provides yet another option for those who do in fact care about the noise. +dbf does not serve all points in the WMS +messagesI think Hoss made his point about not leaving this code in queryparser contrib project and I agree +and if compilation changes a bit and there is a regression tomorrow and it starts failing for a reason different than the one desired then it wont be caught by the test +A state where zoolock is uncertain of the status of the lock in zookeeper +linked to this jira +Thanks for the patch Kim +Maybe its stated somewhere in? +This patch implements the test scenario where job killing is done by user explicitly calling the kill job +The instructions to reproduce this are not relevant any more but the bug fix is necessary to display the right sections in the create and edit screens. +code review +Theres actually an issue with the XSLT engine that prevents from actually running under JDK +Reproduced and fixed but not committed +doing social scheduling as you suggest very quickly breaks down not just at the very large scale Im used to +This is not a bug in Railo +iirc normal form submits should have a working progress bars +Updating description as per suggestions +Applied the patch with minor changes +applied patch +Gabriels suggestion was elegant but if dbtype is wrong the user gets nothing at all +The imports added and the blank line wspaces before toString seem unnecessaryIs any sort of compare really used on? Otherwise +Username bcotton +It was with either or +Will be published on next site version. +Closing this issue as Wont Fix +cant reproduce in displaytag from cvs +It looks like the version of Equinox you are using is not +revert the change in rev cf where I change aries to not mark the methods as found a lot more instances where synthetic fields and methods were breaking the tck +but it still uses you very much +Thanks! + +Im kinda surprised that this can happen +The reason for this crasher are the changes made in not sure why yet +I will review my patch once again to avoid this kind of bugs +Will open another JIRA for the same +This patch fixes the warningDont know why Hundson complains about test failures +I believe this has been resolved with +But the common case is much faster then the worst case not that it is much comfort when you hit the worst case +This sounds like a good idea but also something we could roll into a point release +James there is no need to create if the task is already small +Hello JacquesI have fixed the issues for some of these URLs and here is the patch and right now i am working on rest of themPagehttpslocalhostordermgrcontrolordervieworderIdWSCOLinkshttpslocalhostfacilitycontrolcreateShipmentprimaryWSCOprimarystatusIdSHIPMENTINPUToriginexternalELPagehttpslocalhostecommercecontrolmessagelistshowSenttrueLinkshttpslocalhostecommercecontrolreadmessagecommunicationPagehttpslocalhostecommercecontrolorderstatusorderIdWSCOLinkshttpslocalhostecommercecontrolcancelorder Pratik but this patch needs little rework +In Nexus you can configure on a repository basis +Could anyone test with other platform? Then Ill be able to commit the changes +If you decide to go that route it makes little senseto do it all piecemeal please file a ticket to convertall of your existing infra over to the new name +passes locally on all +Submitted a couple of related patches that came out of working with the existing php test code +Thanks StephenYour patch is in trunk at r at r +do have a way to hide the +Updating fix version. +I dont have commit right since we switched to the new SVN repository and I will be away of internet connection for the next days +add drools back to dependencies of runtime pluginNot sure where to do this +Cant reproduce so I assume it must have been fixed previously +We can revisit this at a later time if it becomes an impediment. +This will let sophisticated users who have a need for use them but removes them from the common case where they are simply causing confusion and painFair enough +Why not attach the modified on this JIRA so users can replace it in their local installation if they want the functionality +corrected html page titles and js file names in order it correspond the appropriate demo on ftp in html files of demos +Hence Im marking this issue as resolved for the time specific class is effectively not very useful at present since nobody will use weaving at the EJB JAR level +Brett there are indeed separate problems the need for a tmp location that outlive a build session +Would love to see this fixed for the release +I saw lots of failures too +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian this is fixed in. +Im moving this for for now since its a kitchen sink type bug and doesnt really need a target +do mvn help on your project and attach the output here +Changing Fix Version from the closed release +Excellent +Looks like a bug indeed. +Add to project menu as well +It gives you human readable html and an ugly but serviceable XML based api +Issue may already be resolved may be invalid or may never be fixed +I think that maybe we should be changing the renderer at this point +A couple unit tests are not converted to test yet +Whether is more difficult to install or vice versa is completely and utterly irrelevant with respect toYoure just assuming that were going to make things difficult for you +v +We can consider adding an api to configure this number after we gain experience with how behaves in the wild +We arent overriding a definition from a parent classloader rightWith typical use of filters +And all entries which have categoryhidden arent encryptedIt is possible to force process to be the first component to be called +h file to include and +Screenshot with surefire report pages +Headers should be rendered as for not empty table +Attaching a patch with a shell script for this +Since this was only a case in our own specs and has never been reported as a problem in the wild +patch has been applied +Covered in +It is certainly not a priority at the moment +I believe this patch is good to go +You may reopen it later on should my findings turn out to be wrong +When dates are selected and form isnt submitted valueChange event is not fired +reupload patch to test QA bot +peter I didnt +I believe it was a temporary workaround when we transitioned to +a patch for patch of patch for javatestwsdl directory +The paths in the tests that are using builders do so to test the builder the codepath used in is much more straightforward but it might be worth it to port some of the Counter tests to use streaming rather than the builder directly +Hi Albert thanks for reviewing +In case it does not apply to all modern Unixes we can remove it +Can anybody try this patch +Renamed the method as requested +Thats how we intend to deploy when the server adapter is in Binary mode +adding Release Note Needed flag so this gets picked up by the +It is not like that you can concat on any set of filesI realize that but I dont see that as a good reason to treat them differently in the context of a immutable snapshotbq +I think the patch also fixes a bug where the original code could end up trying to cast a to +This has been resolved +May have to load Regexs from a file +Forgot to add that I ran derbyall and without failures except +This issue is no longer relevant for it is fixed inChanging the affected version to +As connection pooling is almost exclusively your department do you mind taking a look at the patch and checking it in? for me +Not sure whether the JAR can go away in favor of the betaFWIWFeel free to ping the list if you run into any questions +applied thanks +So we get regions +max please review my as critical since this was supposed to be fixed since it is very likely we will bump into nameversion mismatches again +Each broker watches those ZK paths and take actions accordingly +Thanks Jimmy. +I replaced the is flag with an instanceof check on +bulk move of all unresolved issues from to +Close after release +So comparator for and is different +I mislabeled my merge +putting in but if you think its more critical please move it up to +Updated patch for changes to trunk +Personally I think the payload example shows that this is better located in user code +The XML changes are Eclipse jibberish +Seems to only happen when running on a Windows box +looks good conditional if we can add a test case as well +I have run the test for iterations on my machine and not seen the problem +the rpc endpoint will fail if you do not provide json input +Bulk close resolved issues +The part locking does not working well in your comment should be an another issue +Sorry I dont like this change +Proposed fix for missing get under +Rather were planning to run these services in parallel in the same JVM but each service will operate on its own unique port +Hi! Sorry for retrieve this old task but Is there any update on itI am searching some implementation of Spring Javascript and I cannot findsalu +Generates two reportsmvn verify or mvn test P run for all tests +Current contexts being the eclipse compiler and Groovy console +ok thanks +Investigating possible classloading change in that is a missmatch with the Module System + +I put up the for +However it might make sense adding a simple boolean ignoreRequests property +the hang in safe modewe have the two alternative solution you can reference the above comments. +Added instructions to edge docs to warn people of this issue with third party plugins +But any other reasonable place is also welcome +No problem +Thanks Mike +Duh were not resolved yet we need a bit of documentation before +I thought you would like that +Patch looks good Shai +This patch adds a test for replicating databases when NATIVE authentication is onMost of the changes in this patch result from replacing copypasted code with calls to some common methods which create connection and databases +Patch for earlier version of Hadoop +Are we looking at the possibility of removing same images if any exist in all accounts +Fixed in SVN r but it will take a little while to reach the live site +when deselect the check were some problems with the get method I have fixed that. +I have set the max concurrent consumers to and added an info log message on the component to record the change in case there is a component that runs into this ceilingI will commit it to the trunk since it is a rather trivial fixCheersAshwin +No tests because its a pretty small change +In fact it was not a bug but a lack of documentationFixed the documentation +Ted very nice +New patch is good for me +testcase only occur with ls and locale see testcase that was fast! I cant tell from the GIT commits which versions this fix is going towards +sounds like +Please confirm that they are applied as expected +Feel free to punt to another JIRA +I can see some plugins where this would be necessary right now cant include the same Facelet twice because +Resolved issues being marked closed in conjunction with the release of. +If he doesnt have any further information on how to reproduce the problem then yeah we should probably resolve this one as unable to reproduce the problem +a suggested revision +Stacktrace when trying to call Filegetc for a a file object opened on devttyS on Linux +Id just go with the reply count +Attaching the zipd database which Mani posted with his email +OK added in and to the D options +If anything I misundertood or missed feel free to comment on this jira +add statistics for total cluster cache capacity and usage in bytes +Thanks for the ack on Lars +Can you please debug the environment further +The patch also adds quite a few Transformer implementations +The LDAP sample application doesnt use security as the default namespace so this cant be a general case + +Did not include any as this is just extracting to a config item +Patch for Y! to to pathfilename components are safe the original patch lgtm +Additions to and patch was created from wrong revision to trunk +However I am as it happens actively working on this issue now +Attaching a patch for this issue +If a user attempts to use the the generated pom file will not include instructions tags for the maven bundle plugin configuration +after which I had to clear my session cookie toexperiment any further +Hi AronI am connecting to Oracle database +I would have thought creating a Controller with Roo and then changing the form backing object to would reproduce it +If need a particular feature or bugfix to be included in you are welcome to get involved and contribute. +Lets do that for back to implementation was completed and work on unit test cases is in progress +These are the actual test casesBillThis is just an issue with EAP and jbossws +This patch is for the branch +Alejandro is the thing that undid +The need to compact several stripes in parallel can probably be alleviated by just having less stripes? As future improvement it is possible to add +This would be really nice to have in if you can commit the time to it +Are you asking for to have a copy of the errors? Are you asking for details about the validation failures to be in the exception message? BothPlease clarify +See patch +yeah sure I can take a look +New AMQ freeze test from failed some additional work on yesterday the problem seems to be back to not dispatching +change the capital file name +This probably indicates that the NATIVE documentation should say something about the casing of Derby usernamesThanks just changing the names in the call nowhere else causes the program to work finejava cp exporthomechaasejavadbmorecodetrunktrunkjarsinsane +It was me who was confused aboutthe fallbackURL +As for phpize being specific to a particular php installation I cant say for sureI did the following test which makes me think it is not +For example if osgi and messaging arent needed remove themAlso if the primary usage of this file is for testing but its not really a good illustration for users it should probably live in the testsuite and not in the distribution +It turns out the failure to start the server is not exclusive to +can we hadoopQA tests on the new patch +After run a period of timeshutdown send client will not run sometimes and stoped +Probably this is yet another stupid user errorI created with the following contentsNATIVEcredsDBI can run just fine to create the credentials DB +Julien Ive created ticket and Im going to close this ticket. +Best regards YounesHonestly I have no idea what would even trigger it to start off with +The additional javac compiler warnings are a result of the newly moved classes that already had these warnings +The password was provided when authenticating to download the POM from subversion +And yes it would be great to have a contact person on the Google side +Sorry it looks like the cut and paste somehow mangled the m attaching it as a file +If you can provide one is fine if not Ill close the issue +Thank you Giri +We have now retested our tests with the latest code from BranchAS and have found no change in behaviour we are seeing either a null message being returned or the loss of of data +If you finish it before we release beta please correct the version +Adding the exclusion did the trick r +This is to be fixed for + +Could you give a brief overview of those changes +Attached a small patch to return exit code for invalid mapred commands +Fixed on trunk + +Moved to new issue to get rid of dangling subtasks with time remaining +In SVN. +Patch from RB committed rev fadabcdbdcdbff +Version has been releaseClosing all related resolved issues. +To confirm this I suggest to revert the commits from locally and run the performance tests on that +Hi Henrik thanks for the are not maintaining the just providing what is in the community AS +Verified with latest automation runs +However JSR impl doesnt have any workaround so maybe +My fix was a combination of upgrading to cordova and clearing all javascript timers on the pause event +Here is a patch +Need to add some additional info here not sure what you are trying to report +In other words if the generic parameter is specified and matches exactly to the type you want to marshalunmarsahl it will be picked over a more obscure onesIf you followed your above example it could break a lot of code and give inconsistent matchings based on what was registered or not +Were talking about the method call ? I dont see a problem with it all Apache licensed under Apache auspices +Because there may be other implementations that work without maps or have a hardcoded number of attributes with optimized implementations +Thank you. +The usage would be if you need multiple paragraphs of copy that are styled the same way +But anywayIve put in a redirect +In the retry part the goodRows after the if should be removed to avoid counting rows twiceOther than this w added test in like a different problem +This looks good to me +Bulk closing stale resolved issues +Is there a reason why this new config not apply to all goals +Committed revision . +If there were any unlogged operations runningwe throw an exception with the following messageReplication master cannot be started since unlogged operations are in progress unfreeze to allow unlogged operations to complete and restart replication +Hi DavidIm sorry to distrurb you again for in patchesI have translated also errors message into Italian and some missing properties file to be add to the SVN repositoryOFBIZHOMEframeworkminilangconfigOFBIZHOMEframeworkminilangconfigOFBIZHOMEframeworkserviceconfigOFBIZHOMEframeworksharkconfigOFBIZHOMEframeworkwebappconfigOFBIZHOMEframeworkwebtoolsconfigOFBIZHOMEframeworkwebtoolsconfigI hope this is really the last patch for inThanks a lot for you work have included at the begin of every single file the licence ASF headerCan some of you add those files to the SVN repositoryThanks a lot for your work have included at the begin of every single file the licence ASF headerCan some of you add those files to the SVN repositoryThanks a lot for your work have included at the begin of every single file the licence ASF headerCan some of you add those files to the SVN repositoryThanks a lot for your work Marcothe new files are in svn with rev +Fix for a lot +Good one +I cant build a test a test to demonstrate marshalunmarshal because the code it would try to use does not about if my test case calls instead of? If its feasible to do that I think that will work +No the JS method is provided by the framework +However I wanted to avoid changing if we put there it would be a big performance hit +That would be ideal +Could you explain what the problem is and how your patch attempts to solve it +Is there a particular reason why it hasnt been put in yet? We are thinking of rolling updates with the patch for older struts versions but if there is something preventing this from getting into HEAD then wed rather not start off with a faulty fixAny help here would be sincerely appreciated +So I removed that one +I suspect that it is deeper security processing differences or in some details of the older wiki brings up the point of supportingmaintaining the much older branch wiki code and test suite since wiki is only really being maintained in trunk +Committed revision Thanks Jayson +This is in the newest release of CLI which is not tagged as latest on npm yet +Just to copy over discussion with Libor it now says activity in last hrs even tough I had activity somewhere over hr ago +On the Beginning JSP book if you know which authors did the chapters cited letme know +Would suggest that if coprocessors only need three states then just do those +Did some artificial tests and couldnt see any increase in used memory +In both situations delayed or immediate the client thread would put the data in a queue of some sorts and the async thread would take over them +I can help with the testing but got a backlog of other things at the moment +But I want to make that clear because it is a slight change in behaviour +resolve the symbolic link for the initial bundles to start +Hi AdrianDid you see some issue with my commitsCan you give me some detailsThanks a lot are replies to your commit messages +OK then you can ignore my comment too +We want to be able to run reports off our database using native sql without having to provide mappings +is this still an issue +Change pushed to master in JBDS B +You can safely add the updated output js file in WP Cordova projects +the new patch fixes that version of the patch +Ops teams are expected to reconfigure the ports via the configuration files instead of using the defaults +Ive applied your patch as is perfect quality with all redirections really nice! Thanks a lot for this contribution! +manually split data etc +Thanks for reminding me to upload the trunk work +Please do review +But EDS has no issue with using the latest +Ready for me to make the switch now +Before including any additional BC source code in it needs to be reviewed as incoming source code to insure there are no IP issues in those files +In hadoop I think the rpc request size spikes are block reports +It does not work for lines or rasters. +Rename to CHECKSUMTYPEDEFAULT to that it will be consistent with other field names +This really can only be done in a practical way once is maintaining its own indexes. +Maybe I can do that myself +We couldnt find a clean way of separating the patches but we felt it is worth discussing the issues separately +Can you file a ticket +Yup I think that makes sense +appliances +The null check in was intended to handle that +emptying veditor should work +The problem is that the does not exist on Mac as it is part of the core JREFor now i suggest the patch is applied +Fixed in Trunk +Fixed as of rev we now enforce compatibility of throws clauses and return types between handlers and event methods +Bulk closing of issues that are old and have been released or marked invalid. +We have to qualify it with the outer class +Thanks this is looking better +yes this is for DTX in a cluster. +All platforms have now documented the migration. +Instead of pushing the operation to the master +Trace on the server up to the disconnect is attached +TomLet me ask this +Ill look at this +This issue is not closed the points made against it were not even answered +Section +improves to visualize payload savvy tokenizers and tokenfilters +Personally I dont like error messages that try to make the software seem smarter than it is + +the final result of our vote for using the Apache Nexus repositiory is votes votes votes The minimum requirement of votes with no votes has been metThe proposal is hereby accepted +We should at least log a message +Attaching a simple weve used as a reference for this issue +There should be exactly one of those per request +fixed in r using the suggested approach. +TIA +Its builder plugin generated file +Why? So much for back compatibility +However based on what I can see the DS Shell AM at DEBUG level may not be necessary +There is already a logic which checks existing dirs on +The +The patch fixing +Seems fine to me +Id appreciate the feedback +It should be more efficient that we create a directory and then delete itIn this particular jira is it OK that we create the output directory by the job client? So the job submission can fail if the user does not have the permission to create the output directory. +And yes this patch also suffers from not having good support in a similar way +AgreeLets use a modal dialog +Added a unit test +These names are not allowed to be two different symlink names +Just wanted to know what exactly this issue is about +If you use a for it should work without having to change the source code as in JIRARegardsSven as described +I think this general limit extension needs to be addressed as a different ACR +I agree that something less specific which captures the intent rather than the implementation details would be better +Update online release notes svn ciSending downloadTransmitting file dataCommitted revision +Im going to mark this as done since the code was dropped long ago. +We need this annotation implemented for our release +this is the release note following the new release note template +I am reviewing the patch and making some changes +If the exception keeps being thrown the same request is placed over and over again without any other request having a chance to enterMaybe it would be better to just retry once and on the second round log an error message but to not do it again +An example with to reproduce issue +or +Isnt that a bugbq +Yeah I was a bit reluctant to add the extra config file to the assembly since hardly anyone is using the features OBR support +Few days later I cleaned up my project workspace to check out the same project from the CVS on which I was working +modified my patch to use custom which is faster then just fixed side one + +This is on Windows it may behave a bit differently. +seems to relate to compaction behavior transformations +Bulk close after release +the needs to do the file watching when it is a new property called which tells the service factory whether it should restart the broker on xml configuration change or not +I ran and derbyall with weme jdk and jdk +fixed the just committed this to and trunk +lgtm as r +But the Nutch project should not advocate such irresponsible behaviour +I did something wrong here I wanted to review the text above before posting itIm sorry about that Ill just continue here when it gets too messy another jira issue can easily be opened +Attached patch file contains the changes I tried to implement in order to solve the the problems related to deploying rar into Weblogic +A response reminded me that a template matching in a mode could also be triggered by the root of a document opened via the document function so the passage you cite does not necessarily imply that the WG expected that such a template would be applied to the root of the original input documentBut if our competition has this capability maybe we should have it also +In the patch the doc stores are being flushed on get reader and the doc stores read successfully +v seems ok locally +Ive tagged this for investigation for +Why not accept all types by default +updated version of the patch +The newly submit update integrate the backend and UI +It will be optional to generate unigrams as well +Ok I now understand what you guys are getting at +Pretty sure the build failure has to do with the recent dependency issues. +The patch adds Slive to +I have changed the patch to throw an error when this occurs the open flag will not be set and no activity will occur +As a good citizen we have to honour is tracking the service use problematic? Surely we just walkthrough the context using the OSPFBs +I was not aware of the file in before but seems to work fine when this line is uncommented +The second case is implemented in the fix attachedWith respect to netclient codepath only the third case is implemented +Thanks for committing Claus I will add the documentation +Looks good +Committed revision +Also there were failing tests for this which I will look in to soon. + is not necessarily an take for example file +Im still against this Iterable! Iterator and breaks the Iterable contract +now calls the database every time regardless of the status of the cache +Thanks Hans +It looks good for meIve verified mvn clean install +Hi I guess almost all of us actually useJon is using Eclipse AFAIRIf you have a patch to attach we can definitely commit it for you if that makes contribution easier with Eclipse and mE +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +This just renders the iframe and does not take care of any chrome +Because your proposed fix will become a bit ugly for the UPDATE statement and it would be a pain to extend to future new clauses +Wed be monopolizing the whole cluster +It includes a couple of commentwhitespace changes that are accidental +Yes certainly looks better now +Hows this +If no contextPath init parameter has been defined for the it will use contextName as the default which is basically what is happening at the momentThis will effect effect third party implementors who have written their own implementations of +Thanks for pointing this out! Fixed in M would be nice to have a fix for this in the branch +Lets not change that here use to create servers + +Do we have anything that can check that the docs are making their way in +Fixed in Marecor +Dont have to fix everything for MAny ETA? build for QE is supposed to be done today + looks good to me +I am not sure the global setting works either +with +this passesant test Dtestcase DargsXXexcludeorgapacheluceneanalysisenstep +You could ever use multiples databases for differents purpose +This patch also includes addition of a few protected methods in and so that I could subclass them effectively +Please fix it +Patch committed +By the time its gone in it will have been deprecated at least months and two releases +v +That might change how snappy would be setup for +The problem may be resolved if we change all of the rhq +Luke wrote that code and I dont completely understand it +Patch that adds the copy in HBAs constructor and removes the copy in get passes unit tests +Issues went down tested! +No longer blocking bc of fix to +nevermind got the revision from patch +Attached an updated test case that shows the same problem when using the +Can you please check your version and report back +Hello David I did not went trhough the book +Currently each has reference to the it uses +Thanks for the info Mike +Its and if anyone trying it uses either of those build tools otherwise all dependencies are zipped up in targetdependencyIts hooked up to derby so it will create its own DB automatically in the working dir +Johan can you take a look at the suggested solutions and comment on themI believe you are the expert in the area +Ive added the object to the constructor of the object so that it can be synchronised on during readIndex and then protect indexTerms indexInfos +You already filed a JIRA go ahead +Ill commit this after jenkins gets back and create another jira to continue dicussion about codes enums in the error response +it was breaking mine too +The implementation has changed from to +Attached is the unit test I have to recreate the issue + +Shoot this introduced test failures +As it is not working in sure it should not work in +I add the extensios but still cannot see any pack for selection +this is obsolete due to forrestbot +though wary of changing it if others are not seeing the hang +at one point the event wasnt firing +Ill submit a patch for this next +The drawback is that there will be two of them in the list of editors +Thanks Tomaz +I was running on Windows bit with JDK +I was thinking under siteimages but the Wiki is fine too +Theoretically Hadoop should be listing as a dependency in its pom it should come in from ivy via transitive dependencies and automatically get on the classpathIn fact thats happening if I remove this dependency from ivy it still shows up in buildivylibPig when I build for hadoop but not if I build for hadoop Should be fixed upstream +If there is no objections Ill commit soon +has been replaced with the new project builder +I also agree regarding the general rule of trying not to do expensive things at startup time +Thanks for looking into it. +Tried to catch you on irc but you timed out +See any difference? Whats the math like? How large is the Map if M regions in it? We remove when region is bad but ones we dont access could stick around for ever +Marking as closed. +Im am looking at this now +Here is how to apply it +I think that you like declaring validations at UI level instead of data levelNot necessarily +has been marked as a duplicate of this bug +Ive replaced these messages in with use of the message accessor +I used the one from Matt Jensen adding the ability to enabledisable the moduleId generation +One JAR file you dont need +I was waiting for Struts migration to be done and now Im no more using Continuum at my job so I dont have the time to work on it +Perhaps not. +Hence when the Active is starting up it should respond to client requests with startingUp response to indicate to the client should wait +It may also help to find a solution that works for our current tests and this issue especially as we have versions of the code that work in one or the other case +dims +A tablet cannot minor compact more than once at once time so you shouldnt be seeing multiple minor compactions unless the table split +Karam can you please take this patch and your tests +HiI could not find such a project +Making blocker per JD +I think the Spring Security guys would be happy about this as well +The only thing which is not functional right now is setCapacity but its not that crucial anyway to figure if that helps to fix the situation. +Deleted the following files as of revision gtmodulelegacysrcorggeotoolscsmodulelegacysrcorggeotoolscsmodulelegacysrcorggeotoolscsmodulelegacysrcorggeotoolscsmodulelegacysrcorggeotoolscsmodulelegacysrcorggeotoolsgcmodulelegacysrcorggeotoolsgc modules have been gone for a while +Milosz please work with Tim to come to a common resolution +This has already been fixed in trunk some time agoIve now merged this fix into the branch as well. +Would this approach work +This is marked as a bug in but I cant seem to replicate it +Its all piped to devnull +left the name as it is since its shorterAddressed the rest of the comments. +We should either fill in the materials or remove them + +Why was this issue reopened + +Id like to hear another committers opinion on deprecating as well but I guess we will see if we are able to draw the attentionI think we have a lot of latitude with the move but I dont know what the consensus is on the limits of that latitude +attached broker initial observation here is that the configuration has the clientconnections disabling the inactivity monitor +Thanks aaron Your suggestion looks good +Wont close till prove it +I wouldnt call that fixed as its not always desired to deploy exploded +so I havent changed thatWe dont need to discard for error during read since it hasnt been constructed yetAgreeI think this patch is ready for commit +Verified +I think the server should go down similar to single db node case to keep the behaviour unique even db ha enabled to true +Should it be closed +I have only asked you and Daniel to look into these for now so it is up to the two of you how you want to manage it +A directive from the Board that the Legal Affairs committee is responsible to work out a policy proposal +Not pretty but I dont see much of an alternative +Ok got itWe use index with no reverse index +if you have a better idea Id love to hear it +Aleksandras I believe this was fixed in by +Have had to add to the MR so thats its available in the +Moved to CR +Yes Its a bug of I just committed a quick fix for it and updated the to use CXF continuation APIPlease check out the latest Camel for verification +Why not simply moving everything to the session instead +The developerBA can now set an attribute on the questionnaire within the rules which will enable page validation +You can see the logs for jar signing as well +You need to release note changed behavior +Please close if it works for you. +closing since fixes are in released code +Looking forward to reviews of this patch. +Under what URL to you access the console? How does your etc look like +Thats much simpler to handle than filtering contentI may be incorrect but I believe the filter could be as simple as Show Forbidden for conf +No its something else was never touched git log p on the file dates last change on it on Wed Jun Let me take a look at to debug this dont know what causes it exactly but its in hypervisor layer unassigning so anyone else can take a look +files +It would be nice to have this done along with the release though +The general problem here is the admin UI as mentioned by Felix +If this doesnt make thats ok +How is this related to +is a service run at the to facilitate transfers of intermediate Map outputs to requesting Reducers + is a plexus component so use component private projectHelper +Attaching a simple program to reproduce the problem +We need some updates for our other languages + +B +Fix is pushed +May somebody pls try this patch? Thanks +Note that while +Hi TimSorry for my careless mistake +Dave can you please double check and then assign to the docs team? Thanks +Weve had some problems in past with it which were supposed to be solved with the newest JRE +It has patch files the utility the sample changes using this utilityPlease review code readme and let me know the only in for accomodating all tables creation +done in thanks for pointing that out. +Patch was applied and tested +please dont close it +Sorry for the trouble +It was fixed in the runtime and an update of the embedded runtime solved the issue in on. +Everything all set +This does not seem right + +The data type used for the CLOB columns was specified as BYTEA instead of TEXT +Findbugs should be gone and the units tests should hopefully pass +Yes +Also created for this issue. + +No idea how long that will take +Thanks Mark! Ill test ASAP +mark it seems the scenario you described could happen if the value of is less than the number of messages +The one shipped only did async flush meaning it didnt actually flush all the data +Committed revision +Thanks YING I did commit +In fact it is not the correct approach +beans test crashes with the patch +I could have sworn i tried this but i will try it once more and letyou knowThanksilya +Dont touch the current implementatioins but make tools to convert universal input such as csv into the specific file format +Im still looking to see if theres a clean way to extend to better test this in the absence of Kerberos +I opted for lower risk of breaking CDH etc +Similarly in +Issue closed. +Sounds like we decided to leave this as is +Attaching the patch confirmed that it fixes the issue for the sqoop tests +So a search for availability should show a room as available until all rooms of that type are booked + +Is it an issue specific to or do you have the same kind of issue with Gendarme +Im attaching a file with suggestions on how to bring the JDBC Reference section of the manual up to date +new workflow GIT Pull Request workflow created based on Seam Workflow with patched Pull Request Closed transition workflow scheme GIT Pull Request workflow scheme created to use this workflow for issue types where meaningful Git Pull Request custom field reconfigured to be available for all projects for issue types where meaningful new workflow scheme assigned to ARQAJO and ARQRUSH project +Fixed thanks for reporting this. +Looks good to me +So we can decide that having an ability to roll your own is good enough and close the issue or we can leave open in case we want to add something like qw in directly at a later stage +Is there any method that is or can be called after all properties are setIs there a new stack trace that we can look at +get does read from a volatile an action that people say can be realized on SPARC and IxIx architecture by a simple load without memory fences and thus without wait times for the value +Fixed in +Pls review it thanks +have you a stacktrace for this? +The fix is in the trunk you may need to wait for the merging to the branch +Im confused why has this been cloned and why not just reopen the previous issue if it is still a problem? There should only ever be a need for one ticket for a single issue +Merged the test case to trunk in revision . +Can one of you comment on about the changes you needed? Just seeing if we need to fix any incompatibility upstream +Sorry for not getting back to you earlier I recently found out that this is actually an We recently switched to the on the mac then these issues were gone +Heres a patch w simple test +Answers to your questionsAre you properly shutting down Tomcat so that it sends the contextDestroyed event in all of its web applications? I am using to shutdown the a breakpoint in s contextDestroyed method and see whether and when it gets invoked +If I could be assured that all transformer components assumed responsibility for copying the properties from the input message to the output message then I would not need the Pipeline to do this work +Moving out of +so Ive created the patch again and made sure that its got the and Test this time +Done +target and should be set in parent pom to use jbdevstudio +Is this available in any SNAPSHOT build yet? I tried yesterdays build but it still uses the in the Bytecode +Removed some metrics test workaround now that hadoop common in has the right fix from trunk +Hi JesperDo you want me to the flag or is the int enough? I dont think you can OR the flags on delist so the int should be enough? Tom yeah there is no OR so either the int or the string value would be for the report PR waiting for CI +Not sure why it was working before but I assume that the compilation was doing anything +My personal view would be not to permit the usage +Im always a bit afraid about the CPA stuff as I know nothing about it and alsways fear to destroy stuff it relates on +Ok thats more specific. +Also rebased +Shifting to as there are outstanding issues which will not be resolved shortly +I think it would be a different issue in any caseSending moduleskernelsrcjavaorgapachegeronimogbeanAdding moduleskernelsrcjavaorgapachegeronimogbeanruntimeSending moduleskernelsrcjavaorgapachegeronimogbeanruntimeAdding moduleskernelsrctestorgapachegeronimogbeanruntimeSending modulessrcjavaorgapachegeronimodeploymentserviceSending modulessrcjavaorgapachegeronimodeploymentserviceSending modulessrcschemaTransmitting file dataCommitted revision . +It already does this but it depends how you bootstrap resteasy +I am not sure what the original intent was but I think its clearest of all to increment this value in the line before +patch for feature +The version may be superseded by new configuration features later. +Can you clarify which tests do I need to include there + should be made runtime scoped and the api compile m not sure if its really required though is it +So far it looks like were going for the Flying Fish technique where well have on the trunk and a on a branch the use of instead of so maybe we could use the release metadata to help with moving to the next version in the release plugin +In You are manually closing outputstream and socket immediately +The point is as we develop libraries we need to be careful to design them right for X +Thanks a lot Suresh +I will also take a look at your exception changesThanks for your help and review +For now only Flat package presentation will be available +Martin I cant reproduce this issue anymore +Since Metro solves your needs I would just use it +to commit +FWIW i can reproduce this fairly trivially +Patch containing more thorough testYour Greeter and you should use Greeter for the operationName to keep it same as the annotation Then it worksAnyway we should honer the annotation this is only an issue for code first way and specify different operationName in webMethod with the real operationName in interfaceThis error comes from that we choose incorrect the fix is quite simple and Id like to include your testcaseAlso change this issue title accordinglyThanksFreeman +Covered by recent updates. +There is a workaround for this problem. + comments First if you want an XHR request you need the ajax template not the xhtml template +Nevermind I see Hive now has for this purpose +Does this change mean that we cannot package a bunch of python scripts into a zipjar file and let hadoop unpack them and run it +Close out issues. +Note that this patch doesnt add proper Guice integration for the social data servlets +by is a new plugin based url filter functionality so this patch is obsolete. +Completed in branch At revision +I managed to patch it with what I think should have worked but it didnt +I would prefer not to make any change unless necessary +Its unclear to me what Mesos gains by starting a blog there vs +We dont require the Host header for http it is now we never send a Continue so as far as this issue goes we are okIf we in future decides to implement the response we have to make sure that its never sent to clients though. +By this log file messages was stuck +from me +Ill see if I can come up with a working example to attach +Is there any workaround until the fix is available +We need a test case in order to investigate this +A fix was applied in revision which is a bit more broader The most important properties of the current are now inherited by child builders +forgot the resolved in part +Attached some first documentation patch covering things implemented from my side +Merged to with revision . +Ok +looks like fixed and kiran mentioned that it is not an issue closing it +Are you tagging this because you plan to address it relatively quickly or are you just being optimistic +The threshold is too high as double only provides just under significant figures and you will lose some in calculations +But there are other cases where it is useful to map a field to multiple columns +Correct patch +Attached a test program +It is an additional library for integration with AS that gets pulled in by Native the Native distribution bundles the resulting jar which is why it is not visible in the AS +There was also not much response to a dev posting on how well deal with sessioning so Ill make the call there also and create the appropriate JIRA +node is just too generic +This problem happened with the snapshot currently available on the public snapshot repo but seems to be fixed in SVN alreadyWhich means this patch is just fine! + fixedBefore After fixedBeforeAfter +Ill try to find some time to review and apply the patch soon +change for junit test +can you provide an example where this happens? It could be easily solved on the second list by these parameters thereI was adding many viewindexsize parameters in listbuttons to make this work and this change will fix all lists +Fixed in trunk +Blueprint would then search for nonexistent file in the local filesystem +I had the same error message a few times +Injection point metadata applied If a bean that declares any scope other than Dependent has an injection point of type and qualifier Default the container automatically detects the problem and treats it as a definition error + branch requires update to trunk required update to Ideally they should be using the same version but there are some incompatibility issues with cql in branch when using latest dbapi that I hadnt time to fixAlso these changes should probably be included in official repository Ill have to open an issue on google code +Case insensitive seems reasonable though I would implement it by lowercase or upper case tags on arrival and when querying +See attached screenshot +Patch diff +Hi John Patch applied at r +Committed revision of release. +This prevented some of the files under test from being included in release tarballs +released blob draft another blob null yet another want to learn what the term relation means in the relational data model +You added a method to system +committed to as fdfeebfefedbcd +Running the tests +Ill implement this feature for now using the consumers we can just switch over to later if we decide to use it instead +I reopen it and assign to me. +closed for release +how is that fragileFrom issue description Relying on index is fragile because users can be addeddeleted by other requestsI cant guarantee that link with particular index will refer to same user when page is initially rendered and when it is submitted +This is a great addition +Reopening until the patch is reviewed and applied. +Moving this issue back to HDFS since the content type can be cleared in +Sorry for very late answerI dont consider this request as Bug but Improvement requestLet me explain better the scenarioExpectationWhen working with JCR implementation I need to work with paths for everykind of functional implementation and paths I can use withhierarchy manager ar JCR Specification compliant +This fixes the problem +I also added a quick start section at head of the book as well as other clean up +It would create a default implementation with a hello example and expose a web service that too hidden do you think? Were actually talking about putting some better testing facilities into the tooling but are not quite sure how yet +Build for jar files and docs works fine +This fixes the issue with permanently caching the null resource loader if an exception occurs during initial load +They should be independent of each otherThat does not really address the problem this Jira tries to addressIn the reduceNone case the user wants to control two things + +For now direction function is disabled if anything happens in newest versions of tools new issues will be created. +Thanks Brock! +further to this the was returning an incorrect array. +Thanks Leo what process would we have to go through to get it in the Apache maven repository? Then people wont have to install it locally to use it they can download it from within their buildsCheersMarcus +Wayne yes this is on the latest master branch problem is only for currently opened files +after release +Ok cool then the contents of that patch should work though we might consider allowing a configuration object in the driver constructors +This is not the correct way to do this and is an operational error and has been since inception +Sorry for the false alarm +New patch against current in rev deployed. +There is the Test Connection button within the console +I am having the same issue with RC +Can you link to a wiki page or blogs post +I hope this is suitable +Possible patch to solve the problem derived from the tag +reopen if you dislike it +This issue seems to be related to Classworlds will have to investigate more later +however we are currently working on some inspire related fixes for wms and the client would like an official release of it relatively soon +If youre going to make the change then I think that you should spend time developing a tapestry js api into which various framework can be plugged +Merged this change to. +HiId prefer to have something more generic that would work for any issue tracker +r for problem is that embedded Derby does not work in a multiprocess environment +Hi allI applied Martijns last Patch to the trunk and encountered a problem with document countswhenever I set the rows value to the query the numFound result parameter is limited to exactly the value of rowsThe facet counts are also limited to this valueIf I omit the rows parameter everything is fine +The situation in the bug is quite specificI am sorry +When you add in the way documented in Comment you may see a message in the Problem Markers or Markers view that says Classpath entry will not be exported or published +id definitely need some guidance though if you could outline the basics of how this query server would work and how it would plug into couch that would be a big help +I have attached a better fix that take into consideration synchronization +Will gauge feedback to determine whether global Table comparator property is also requiredregards Malcolm Edgar +This said can we please take care of fringe cases after release? Like please? About merging code to trunk Ill do this as soon as we solved the incomprehension about my use of the branch Im sure this wont take too much! As far as I am concerned you should have just committed your changes to trunk and be done with it +This issue was fixed in the jira has been closed as part of a review and archiving of the aj project +the you use for sorting depends on the filter you pass to the search method +The JVM works even with this warning but it sure points to some kind of problem +Thanks alot +Thank you Nicholas for reviewing the patch +I have put back some code that supports this flag +Vinod I really dont understand what is going on here +If a timer is past the time it is due the timer will fire as soon as the Seam jar is deployed but when the rest of the application has not yet been deployed +When one talks about Apache being about community before code that doesnt mean one has a right to do anything in the code +Is the fix version for this JIRA correct +I barely can connect the stack trace with the files you attached the ine number arent the same +There should be no migration issue since these are transient objects +Other implementations seem comparable wrt +Let me know if there is anything else to do for this issue +Hi Sebastianwhat is supposed to be used for in the PHP worldConcerning the build directory Im gonna have a look at it +Thanks for this +project in the packagingpom directories it would not recurse into them +Nice work and thanks for doing that +But really it should not have been since is frozen. +I suspect that this is the real source of this muddle +Closed again now properly recorded against Fix Version as well +Hadoop was just released. + +thanks for catching itits fixed now +Unscheduling due to inactivity +This patch supercedes the previous patch providing the from Handle that were missed the first time through +Already resolved I should have been using rather then am reopening because should be forwarding to when using that backend implementation +An annoying side effect of using the workaround is that custom log handlers no longer work +Before this can get merged into the lines we really need an equivalent patch for trunk and possibly and + +which means the version number should be available through then I would Groovy could do is to use a different way to store andor retrieve its version number +Thanks for the input MarvinI can see the advantages of what youre proposing +Flavio Thanks for your comments +It was overdesign +OK now +Mahadev Nicholas Do you agree with this proposalThanks good rodrigoIs this jira up for review? Seems like its PA +I tried some of the failed tests and they pass locally so I dont think this patch responsible +I was only wondering if it can be somehow a long time I mean seconds +See prior comments there hasnt been any justification for why this is necessary. +Thanks for pointing out this issue +So we made the change in such that the master can be mocked +Lets do that after well have the same problem laterWhat about keeping them but modifying them to print and ERROR message saying should be used instead +Holger For testing it is ok to depend on logj +We could but it is substantially more work +For beta lets focus it down to as followsCreation of a new component and bundle that can be used by Stacks specifically to hold all classes which depend only on platform AND which other bundles depend on +Here are the derivations of the formulae used +It is missing +Due to the dependency I am going to defer the commit to Native +I tried to look for example I did not find any +Yet this looks very good thoughI wonder if we should use rather than the system property +It looks like these changes have been applied. +Thanks Ankur for contributing. +I couldnt reproduce in none of these Chrome IE +Sorry assigned to wrong Pavel +I have not checked the CVS version +Some patches are high in size and thats why I decided to chunk them down to projectsmodules +No new unit tests for this patch because it is a small patch and testing this behavior with unit tests would be challenging +DennisThanks very much for the explanationWill the version of the release plugin in the super POM be automatically in the next minor release of Maven? every release of Maven the super POM is given a look through to see if there are any plugins that needs to be upgraded +Now Weld also supports alternative strategies for providing singletons such as keying on the TCCL +please resolve before moving the class +If someone wants to implement one of the alternatives then a new issue can be created that describes it. +patch applied to. +So the real problem is that the first time it has not picked that up +It doesnt seem to be fixed on trunk +marking as done. +Thanks the docs in really explain things well +Fixed and uploaded the latest patch +I had gaps in knowledge of security Sorry for invalid bug report +lets just commit the files for all third party libs we use into and completely eliminate the need for net when building javadocs +Thanks SaschaYour modified patch is in trunk at r +Ill commit it soon unless someone objects +For backwards compatibility and to handle the case where there is no domain suffix supplied for a user well keep the concept of a primary or default domain +Closing issue fixed in released version. +sure +I confirm the error message on server stop +No warning in place yet +Will commit shortly +This patch fix the Add Tables ColumnsMessage Select tables and columns +YIKES! There are broken links all over the javadocs +The fix is in revision +There should be or blank lines at the end of the test function and no blank lines at the end of the for loop in the test +Missed another prompt message to update in UIMAService +Tkanks! Cristian +BTW it is worth pointing out that all of these options are already in +an example would be to use memcached based implementation of the Hibernate L cache and for any of Rollers presentation cachesthis has been done by various Roller users and works quite well +Conversation with ben on IRC clarified for me I think this bobmcw c b c b a in that order then? +Thanks Brian +Ted +Are you sure? Do you have a link for that? +no you shouldnt add lgpl license to ll ask tom +Can we update both the sbt script and the jar to include the new sbt +I havent seen your stack but maybe you just have STATETRANSFER below FC? It might be interesting to script something to invoke the toString method +So for that reason Ill do Solution B but will not filter at all +has been comment aI will change GCUD so that it will transform those kinds of main methods into something that JDT knows to lookThis is what I was saying about Groovy creating a Default script when the module node cant be parsed +I yet have to move the sources from to Codehaus so I might as well do it in one go +Testing with Im seeing both profiles active which is what you asked for but not how it has ever worked in the someone confirm +Comments made earlier today suggest that a deployer was never implemented is this trueIf this task did not cover the deployer then what did it cover? +Incorporates Xins suggestions +Since the current fix touches rarely modified code Id like to keep it in trunk until this is resolved +Heres my latest +project buildcurrent bb implementation requires configuration of a bash file before cordova tooling can work +As to how robust the buffered socket is going to be we cant really tell unless we try +By the time child kicks is the for taskScheduler is no longer available +will be in the code you can confirm the fix using IIS manager to set the Host Header NameStart Administrative Tools Internet Information click on one of the Web Sites select the Web Site tab click Edit to change the Host Header the proper Host header is not sent with an HTTP request response is Object Not Found +can you port the actual revision that was committed to so i dont have to redo all the edits +removing the xmlbeans generated classes and regenerating them +switch to IP everywhere +THX Josef! +not being able to keep up with even medium insert levels makes unusable +Applied fix based on Sunny IP proposed XSLT +I updated the patch by review board comments as follows +Its in the splitmap package so why not calling it +In the used at the customer here it is definitely not included +My eclipse project tells me that it is missing class so I have errors +At Salesforce we override it to anywayAny comments from anybody else +Now everything should be committed +I cant reproduce the issue for the file in my environment +Seems ok to meI have a theoretical question +Yeah I believe it was added for the xml user group and role services locking files when adding users etc +All of our exception classes do derive from a base exception and it would work in our case to specify rollbackFor +works fine +Yes and power +As implemented the only improves performance ofthreads that use multiple +on moving bootstrap logic to its own package +I did the changes you suggested +Correct this is fixed closing. +just yet +Sorry Karsten must have missed your pull request +To be clear I think well see a net speedup Doug pointed out some things in the current code that are less than efficientAs for NN operation well it and shoot for no more than slowdown thereI think the discussion has shown that theres interest in this so long as we can make it performant + +Hi OliverI will try to investigate and check the whether the XML passed is validI will check as well which parser is in use when the error happensJust want to check with you? Do we use the Apache for the parsingThanks +patch is only currently for test comments +And leave aside the classes in Probably SnapshooterSnappuller can use +Also want to test abort probably need a testing codepoint to cause the failure at the right time +I had an idea for the issue of needing JVM restart to duplicate the bug reliably if I define test cases calling a common method in the same file itll launch in a new JVM each time under Ant right? I didnt see a way to run just test under ant and the tests take all day to run on my machine can anyone confirmAnyways heres a new patch that attempts to run the test times in different JVM instances +We had this assumption to help us map tasks to slots easily + +Attachment has been added with description patch file for committed to cvs +Patch v fixes the test issue by adding Writable to as per the previous comment that makes it run as expected +Does anyone know of a better GUI Git tool for windows +I added a footnote to book in delete section pointing at your blog Lars. +Attaching an updated version of Phils patch that applies cleanly with is good to have I will take aThanks for the review +EDG Alpha SNAPSHOT revinfinispan last commit baeedceaadbcedff Mon Jul AS last commit eebafaeaeddfadb Mon Jul +The javadocs for have been updated with some examples why updates on such a configuration can be problematicClosing this issue as Wont fix because no code is altered. +However when reading from stdin and writing to stdout we should definitely suppress the copyright and help info +Dave can you please look at the suggested patch +Looking for from passed +Mind adding javadoc to your new method Sishen? Thanks +Frans seems not to have time right am looking now into this again freshly +If present witha boolean value of false an exception caused by loading the correspondingconfiguration source will be ignored +Note this doesnt allow you to add new query parameters to the URL but it does carry along the query parameters from the previous request +Will run all tests once and check this in if they are good +Patch looks like it covers both cases to m not sure I agree with this patch +For the release the database will need to be updated to point the provisioning table and computers using the xCAT +The api changes slightly so the server code also needs to be upgraded +The attached patch enables idltowsdl tool to generate wrapped wsdl and fixes the reported part name issue +It appears I was getting this message due to my Maven proxy server being down +Should be easy to add and makes sense +I wouldnt call a release where this can occur stable +I do however have a slightly modified version of the ant script that does not try to instrument classes +that is the code actually used so somehow I managed to put the same code in two places +Thanks catching latest changes in file structureI starting to test this patch and going to commit it in a day if no problems found +release bulk close +In revision I updated the code that displays that warning to execute a little bit later after a deployed was found for the module +In order the mavenized booking example and Seam Web Project wizard to work you need to use meclipse should be updated in releng now it uses +Otherwise someone could send some giant string +tools core docs site categories +doesnt explain why the needs a session reference it just states that is does then it follows with details on how that can happenI couldnt figure out from the code why that is the case and after removing it everything still worksBut I may have missed something if that is the case please shed some light on the matterI also agree about the name of the class it probably need a new one now +We should have a release note describing the cases that change +Thus if we only import the plugins we should only allow the specification of an existing location where the server is already installed +Ive added some checks into the Multipart services so that the Upload component can correctly detect whether or not the value has already been set and ignore its own parameter if that is true +Patch reusing Exec and Teds comments +Also cleanup logic in the connectormanager for dealing with the state. +It seems an encoding problem and Ill try to set up my local win xp to replicate the error +Thanks Uma + +There really isnt anything in that particular chunk of code that can leak memory +Thank you for the fix Akash +it seems to be a generic problem with nested named complexTypelukasz +If yes care sharing the results + +Wed expect to throw an in such a case +This patch was created based on the current trunk + May take up to hours for the fix to reach central s a new update mechanism that was put in place latelyBrett +Ive been working a bit on Modello to apply patches that are required by Denis patch for the changes plugin +in simple wordssend does communicate with a other running thread with a other sessionits like you are calling a hotline the person on the other site of the phone is still the same person and is not converted to you +Ill try to look in the code to figure out +Same as above patch and removed author tag +An error is thrown on input besides TFIDF or TF +Ive just implemented this now m modules expose both a sources and javadoc configuration containing the corresponding artifact if it existsI agree that this is not enough to work with sources and javadocs easily +Currently theres only a snapshot build available +Im still not completely happy with that no tests as yet I need to get more familiar with the code base to work out the best way to capture this change in a test and work out how to to TDD for dev +Just look at the report generated by any of the examples that use java bindings +I think as long as this feature is documented it should be fine +Related issue is cannot install Tools from Nightly Update Site missing any way to monitor when the replace old site does not happen +The patch does not as it synchronizes after checking whether tehre was anything in the map and so multiple threads waiting on the lock could then add to the map +Closing for now doesnt seem to reproduce? +This leaves extra map files waiting for the next memcache flush to add a new compaction to the queue after the compaction completes on the region +Please attach data and style to reproduce the spot on +The only new command functionality is the part that reads and registers the connectors +Please refer to for the patch for thrift +It does work if activated in the Guice modules +patch that moves the protWords check below the posInc calculation and sets it +Im not sure if i agree with this concept + +Patch file +I agree I want to make it easy for people to submit translations as well + and classes do not extend Junit Test class neither are marked with Test annotations +o Commented out cookieDomain property configurationo updated Plexus Container snapshot version to +by the way the most confusing example of this is the queryparser which takes a Locale parameter but doesnt even use it for this operation instead using the default +Patch accepted into trunk added test caseThanks for the great work! +Alan Can we get this in for Friday? It seems like an easy fix with low risk impact +Ah Monday morning brain farts +When I set rule to false I cen still set rule in editor and it apperas in xmlThese two are minors though +It also allows nice packaging of flow components into jar archives for in multiple web apps +Pushing this in +Committed. +Thanks Sijie. +Allowing it to go through some reason Hudsons couldnt report which tests failed on its web portal even though the console output is still there +It is correct that these files are missing +Sorry I meant to say +For the record my remark was addressed to everyone involved +Thanks +Thank you Tim +Sorry it has nothing to do with correctness +buildivylibcommongriddevbin +If you the fixes are in the svn are they in trunkI just tried to do another update of trunk but besides the one a few days ago that changed the version number to there has been nothing new +So when ibm releases SR the skipping code can be removed from both and +Triaged for +Wei +Ok I understand nowThe difference was happening before when we used calls which returned different results for different jdk versionsNow when using the result is of course consistent with all jdks as is a pure java implementation +The number was picked heuristically +For the sake of two copy methods that are very standardjava I would rather keep complexity inhouse away from the user there are manypeople who feel strongly about this +The latest version is defined in section asThe version that has the most recent is called the Latest Version of the series or equivalently the latest version of any Document object in the series +Dmitriy I think the deprecation idea has its merits +Inheriting is difficult because of the code +Redback trunk builds successfully for me r +Now working with ivy maven +projectM toolsidenetbeansnbprojectM toolsidenetbeansnbprojectM toolsantpropertiesM toolsantpropertiesM toolsjavadocM toolsjavadocM toolsjavadocM toolsjavadocM toolsjavadoc at subversion revision +Shawn Here is the patch for branch +I did not test with CR. +Dirk +Applied thank you +Sorry I forgot to do it yesterday +i have tested and on windows client against RHEL FTP server +Viacheslav Daniel I have some questions In my case when two observer parameters are in one method it is considered as error no warning +Merged to Fuse Camel stable and patch branches. +This is a very desired feature for me +Patch is for trunk +This patch synced with the trunk and fixed a bug in +clients can use bytes and text messages so long as they follow AMQs scheme for making the distinction if is specified its a bytes message otherwise its textSo I dont think any action needs to be taken on this issue other than updating AMQs stomp page to make it clear how to make the distinction between bytes and text using the stomp protocol. +The new patch looks solid to me I think committing it is a good idea +I just committed this to trunk and +Great! After Emmanuels comment I abandoned hope this issue will be fixed for +Not sure about how are trickling out at the end +Still cant get the repos right +It has generated files with all the TYPES but not with the actual namesof the element. +Escaping issue addressedUndeployment file locking issue located and fixed. +that often involve subclasses of common base types + +It uses Andreas version of the pom fix and also patches to support building with Dall +nkeywal Thatd be sweet undoing one of our custom is annoying times the same errors +Interesting +The notification for fileupload was turned off so Ive turned it back on and am making sure this issue hits the list +Fix should appear in and subsequent releases. +I highly suspect this just being the issue we had yesterday about file being deleted for war in the error console it probably say something about datasource not found +Is that so +nsis uses lzma so it should also compress nicely +Ill rebase and put out a new patch version +This will certainly mean that the primitive values have to be cached in the parser +Ok closingVersion Build id +As Im not familiar with DERBY could you tell me how to call new on your version please? That is how to install my mocking subclass of in my testingThanks very much +Im not suggesting that the language support be removed from + +Uploading another patch to limit sharing of state between classes not in the same package by cloning arrays not considered performance sensitive +Patch applied. +in trunk +Added lines of code to check for null values before adding to list +movie is remade +Adding the label to that effect and the ticket +Regarding the versions of the other files I could see that in rare cases it might be relevant if versions are mixed but in those cases the client or tools or whatever would typically have a separate running environment where sysinfo output would be more relevant +DavidStack Please take a look at it +looks goodThanks + +We have only two and in Solr +I had to configure memory settings +The command in has the subscription name spelled subcriptionName and since the code uses reflection to find the property with this name if you set it as it wouldnt work +Will be in +It seems to contain only deleted lines forthe whole fileCould you please double checkThanks +After all the time I spent figuring out what was going on I did not really spend any time on the patch +Introducing a new method on Persistent and interfaces doesnt seam to be a big issue to me as I think those interfaces are not used very offen +client JME appSafehaus Server server side ADS customized for safehausVirtual Directory Penrose for identity federationSaml Project for Saml integration and Liberty integrationThanks again +If youcan match our Java convention for the next patch thats of course a good thingsince it will save me time to go over itNo need to send a patch just to change one line +Marking fixed. +Resolfed and tested +Ive been debugging this with help from Arpit +I dont think there is a memory leak I just fear that too much is being buffered even on console output though it could be hidden by Ant playing games underneath +changed the to add missing module jars +log file to see full java DenisThank you for responding my comment +Would content assist able to resolve the Seam components from both projects +I am closing this issue because either there is not enough information to move any further on this +This was committed to and trunk but was perhaps mistakenly not marked resolved +Do you have any thoughts on key management? I was thinking that each tablet could have a key +I conclude that we have to modify the pdf generation configuration file so that it specifies a different font for Japanese vs +A few comments on custom contextual informations +This needs some cleaning up but unit tests pass +This looks good now +We dont have any plans for a new release but if there will be one will be part of it +Ive opened to track this +If you can provide any evidence to the contrary please this report. +It stems from the fact that creates a new rather than use the Configuration from +The files that I temporary deleted and made the problem go away do not contain anything that looks are interfaces with some extends statements and no additional methods +This may be related to +Patch looks fine +Converting to a is not good enough since the caller expects to get something that casts to the requested target class +bq +havent had any update for a while so closing please if you are able to provide more info. +I reviewed the changes and would like to request the inclusion of them into for +Attached is the patch for SSH provider for release +is this still an issue in +Committed amendment that backports additional waits and state checks from branch +Fixed revision +narrowed it down to tests +or that stop the cluster in the middle of the test or that use the Incommon interface so can do tests against orUnit tests up on hudson used to take minutes to run tests +Sebastian cant reproduce +will take care of thatThanks Rohit for feedback +Guillaume fine +Bulk closing stale resolved issues +Package names changed to and in SVN revision +Bulk move of M to m not able to reproduce it now +I just committed this +So far it has notWe havent seen any either +Without this patch when user clicks browse File system on the webui and gets redirected to datanodeipaddress and hence no cookies will be sent +Which API is this for btw? +Then in the reprojection stepyou do the same thing +Thanks I will ensure the bit JDK exceptions are noted in the OSJDK combination table +Resolving for verification by QE and anyone else who wants to critique the categorization description icons etc +I just tried your suggestion of renaming and the bottom line was Its not working properly +Thanks Innar that is fine now. +Reopened to set Release Notes Docs Status field +See for further details. +Will try again in morningI tried to make a Serialization that would do for all the hbase objects in but messier than just have three SerializersThis patch also makes Result take pbd instead of just byte arrays +Brunothanks +Against trunkmaster would be best +Where should I point maven to get a build with the fix + +required a slight code change to fix the change and regenerated the patch to include the minimr can you update the review board with your latest patch as well? Otherwise its hard to track the additional changes +Likewise some javadocs have been cleaned up +To get around this I was executing my optional before now that I look at having setVersion be executed every time it doesnt make sense to me to have validate check and enforce versions at validate time +seems like pom declaration is whats needed for. +Is there any alternative way to achieve this +Looking forward to +Craig identified a problem with the patch after talking with him on the phone Ive updated it to address his concern +There are no logicalfunctional changes +I just committed this to branchThank you Mayank +Resolved by implementing a proxy servlet that lives in and proxies requests to servicedependencyoverview while setting appropriate authentication +Ramdisk means data is not persistent after a reboot and this JIRA shows that losing data in one node when the quorum was in minority may lead to permanent inconsistencies in the quorum +This is no longer relevant now that is almost done +Attached views from before and after the test +Thanks Mike such patches go out of date quickly +bq +Paulex test ris behaviors must through jndi public API must need a ldap server becaues ri will create socket with server then send and receive messages its hard to write as unit test we are planning to add such tests to BTI this is in progress +This was resolved on by David Adler already althought the issue is still not marked as issue has been resolved for one month or more without complaints +Go ahead I dont think this is a big evil and if people do mind well have to fix it properly down the line. +AS probably still needs the container contracts which would be done here I guess +Cant see any attached patch +Running through hudson +For every TODO it would be best to have a jira created +Wait the device unlock and access the app again + does not work its something weve been thinking about to do a upgrade eventually +The threshold is used to determine whether the content of the item once uploaded should be retained in memory or written to disk +If you can get this bug to rear its ugly head in the trunk please reopen with specific instructions on how to duplicate it +patch more goals shows the same copypaste errors +Junit added fix fixes crashing junit +Doc archetype renaming is possible Ill assign the issue back when it gets the final name +Committed working prototype of new in r +Sounds like the mar is not being recognized on the classpath +There are a lot of formatting changes +I just fixed the demo test servlet so that it reports a decent error message when it gets a +I created branch lucenedevbranchessolr in buggy patch but shown just as a difference to Uwes +Using the example if you have ten seconds of idle time it should commit +fixed stale resolved issues +Thanks Amareshwari! +Dont worry about future changes to the schemaSee my recent changes to and +In half of all cases it uses autodetection from parent class in the other half it decides between random access and iterator access +Here is a patch to that adds a test case for pass false +Heres a proposed patch with no additional tests +As long as the current tests are successful there is no need to wait any longer +Thx for contributing +Noticed this was the wrong for this issue +Net framework profilePlease make an optional component so that apps dont need the additional overhead of an unneeded reference +Stuart agreed to take care of this +with Constructs a server that will run on the named port +Abstract beans are very useful for defining common attributes that other beans use as a parent +Bulk closing for +partly +Fixed by. +Not sure how thought +expanding this issue from to review the whole test support that we offer +Logged In YES useridThe tests run fine in against java So it looks like it has been fixed in more recent versionsof jacorbor jbossiiop +I have verified by inspection that the change is in both codelines and that there were no new failures in the nightly tests +in +Comparator cache entries couldnt ever go away because objects usedas weak referenced keys were by the very sameEntry object +Patch to remove class +A conversion of filters to queries throughout the code without any other mods results in a test failure because no results are ever returned +is possibly a duplicate although the stack traces are different +workaround set defaultcatalog and defaultschema to correct values +The replication message will restore the JBC node for the session but it will be partial data missing the session manager includes logic to defer expiring sessions that arent active locally to give time for replication messages to arrive so its not clear to me why this would fail +on to branch and TRUNKThanks for the review Stack. +It is hibernate issue +Hi SiI ve attached the patch during issue creation and Grant license radio button has not been presented to me so I had not been able to select itHave I missed somethingNext time I will attach patch after issue creationIve take a look at your patch and Im not sure it is a good solution to remove the global transaction because it will change behavior in case of errors with ofbiz original code if a row service call fails all next row service calls will be rollbacked even in case of success +Im closing this one for now Im sure it can be improved but we can open other JIRAs for that. +Attached is the patch and in r +As far as I can tell were doing the same thing except that Mediawiki is apparently using plenty of Javascript tricks to work around some rendering issuesIf you turn off Javascript does it work stillBTW my name is Janne not Janna +Trust your Its not my its a huge shard and I dont blindly trust it +If you want to implement this I suggest having a look at how is implemented and deals with TTL +It was my fault why this was not working +In either case it should not be visible to the backends +Im currently waiting for some word from Jason or Brian to get a new template +Creating a junit for attributes was a good idea I was able to find many problems +If you agree and assuming nothing more should be done here we can close this one +settings file to svnignore on the main Nutch folder this is Eclipses project settings file +We should perhaps put some more effort in thinking about a longer term solution for the next generation Nutch +There are a few answers to this problem which would be new types of indexes +If you think they should file a bug against Xerces. +Ilya please review the attached PDF with the changes and let me know if they are all OK + done. +I think there may be some issues here as you note +Attaching patch for mapred changes done for Thanks patchThanks patch for reviewThanks comment on the patch should instead by MRAPPLICATIONTYPE +I agree with Adrian that the visual presentation is highly subjective +Thats a cool technique but I believe one of the value adds of the Tools isthat the developers dont need to know any of the magic words +Unassigning until revisit to jBPM. +Thanks Patrick good to know +Hey it works! Thanks Lukas! +Removed fix version. +but like milos it may be in something like one month because my priority right now is to rewrite the to unmess eclipse plugin build +For complex constants the schema computation is a bit complex and involves type promotions null introductions etc +But worked fine with a newly introduced Security phase +Has some questions for Clint to answer first though +I think LONG means RDBMS LONG not Java Long +If this property is not set then it works as earlier +Then I upgraded to when I used show keyspaces I saw index name usernameidx appears for different columns families + looks good! Excited for the release +I will not mark this since it will cause tests to fail unless is committed +ll commit this one in the next few days unless there are objections or improvements +Why do we need to manually splitI think we should add check for the state of zk node and decide whether to continue or not +theyre only related in that there is a cql statement to query it +Rebecca is doing most of the release notes so I am assigning this issue to herBec when youre done can you assign this issue back to Rostislav so that he can check the document +The same issue is for service +But we do need a baselineupload example so that we can tell both that it is broken and that it is fixed +Add in joined versions and update the docs +We have this fixed in Seam context lookup for javajboss as is already released +We dont need this fixed for feel free to push back +The NTLM proxyissue should be addressed separately I suppose along with the related we believe that the bug has been fixed +Many places the the access is under global lock +Thanks for the info +Attaching patches rebased against +Bringing into +somebody help me or give me some suggestionor is it easy to patch itthanks very much +If all buttons look the same then the distinction is lost and the UI suffersYou can see the difference by viewing the Layout Demo using one of the older themes or by using the Flat Grey theme on the demohttpslocalhostwebtoolscontrol +But I do think EL should at least behave identical so maybe it should also return null +stack it seems the patch could be committed +Commit at your leisure +I cant think of a compelling for supporting multiple key references on the client side. +This issue is a duplicate of because the original issue was closed by the reporter due to lack of interest +my framework will execute a couple of c programs where the output of one or more programs will be needed by the other programs think of it as a complex query each operator of the query will be a c program in its own and each one needs the output of the previous one and in the end i need the final output of this complex query +Yes will do +looks reasonable +bulk defer minor issues to +Patch applied +Now that I know that some do I can try to move this issue forward +Renaming files in svn repo is always fraught with troubles +So we definitely can not ignore it +Devaraj The server to server protocol is very different from client to server which makes it harder to implement kerberos in the quorom peer protocols +I am guessing this is due to the fact that we are using regular expression for load +Most recent version is now so Ive updated this in the title. +we should keep in mind that someday we may have a partitioned namespace +But Norman should make the final say on whether it has been fixed or not. +It ends with errors +Marking this is a duplicate of the fixes thats already been done for Range requests. +Worth considering checking the mount status instead of waiting s as a improvementThis is fixed in +Thank you! +so im reopening the issue +Would it make more sense just to document to a user how they could use ij to print out these messages from the VTI? This way you wouldnt have to continue to maintain this list and try and make sure its consistent with reality which seems like a task meant more for Sisyphus than anyone else Alternately could you write a script that runs the VTI and generates the DITA for you? I could probably write one pretty quick if youd like +Ive checked r and r the initial failures were introduced by r commit +Although I havent checked these patches yet it would be good to have a consistent way of specifying segments such as by using a dir directory option +Thank you +Null and are not allowed as the username or passwordI am running the full regression tests now +Patch will be obsoleted by patch +Hi Slava +Closing old resolved issues + +encode the characters in bytes using as the target encoding +Still good work +Johan you are a committer now +patch applied in SVN revision the embedded derby jdbc store used in the unit tests does not support setWe need to use it in an oracle specific lock implementation of make its use configurable demonstrates +Added a test case to fail with this incorrect behavior and then corrected the method that was determining the JCR property type from the string name +OracleSQL Server stay in the same datatype +Ive got some specs for it +Todd Given suggested Interface how we map from an hbase session expiration to a Replica? What if the DN died but RS didnt? Wont the fact that under the wraps is banging its head timingout against a dead DN once per be hidden from the RS since its being handled down in? Dont we need more knowledge on workings than suggested API exposes if we are to avoid dead? If we do figure we have a bad DN do we then per open iterate updating priorities +However perhaps you can still try to recover from this +Each registeration puts the datanodeDescriptor in the heartbeat queue +This issue can be resolved by using an LRU Cache with a fixed instead of the two currently used in +I changed the distribution file to be a zip instead of a jar and used the naming as proposed +will submit the changed patch after the modification +ok for now I updated the workaround to Define as system property false + +I committed an initial work in progress in revision For now I left the JCR version support repository descriptor to false +I havent tried to whack the ant scripts to do this but maybe someone could take a look +Accidentally I have commented here +resolve this issue since this is not an axis issue +Did that not work out +I would be totally unsurprised if it came along with something backported from +I realize two minor problem in this patch +We do the same work to reproduce the same the only difference is that some times it appears quickly and some times it needs some restarts +thanks David +If you still see something can you please be more specific +This comes straight from JAXB +Id love for this to get fixed +but ill try looking into doing it with for Criteria to begin with +ok +h files you have a qualified member which should not be qualified you need to include when you use memset and the resulting makefile does not include as a required library +Hi BobOK that sounds like a fair callregards Malcolm Edgar +sounds good to me +Only if a view is set and that view declares that it should be sanitized should we sanitize the thanks a lot! +I just spent an hour to write a Java programe to generate the shells which could use wget to download doc jars and use jar to extract the files for me. +the generated doc for reviewing +Looking into this further only a properly implemented theme inheritance is going to resolve this +NET doesnt even have it in its Framing namespace +Here is the patch for the same +I definitely want to revisit this soon but there are a lot of issues that will need to be worked out to get things going +I will have a look at the systests laterThanks Regards Jakub +That way we can keep svn history for those files +One further question do you intend to fix this bug properly later on so that people like me that use single quotes in their comments can still expect their sql to be formatted nicely +Thank you Julien for your feedback +One example from mailHiIve noticed that we go a couple of questions the last weeks on the forums that relates to how our IDE cope with shared projects +sitting there just for legacy may be +Committed revision to trunk + +Thanks Sanjay +Since the initial focus will be to to just make it feature and backward compatible we probably will not be focusing on optimizing the performance until thats done +Seems like hadoopqa hasnt been reporting in the past few days while weve been having jira stability issues +Heres a new patch with the suggested fixes let me know if that worksThe only reason I noticed the discrepancy in the documentation is SSVD in spectral was originally tested in standalone mode and obviously the is not available in that caseUpdated the patch actually used the new parameters +Kurt can you attach the configuration file that has the initial content specification in itScratch that +AXISCONFIGURATIONRESOURCE definition is not used in axisengine and therefor AXISCONFIGURATIONRESOURCE is not set in nandika + +I am not sure however if the ssh accesses using the correct user come from the scpexe or the svnssh URL +The only thing that is still missing is a unit test for +Thanks Ivan Vasiliev for the patch. +I saved off the log for this +Reverted from trunk and branch since this wasnt really a regression and the patch was incompatible +allows anyone to implement its own plugin to provide notification dispatchers that tell who should get which notification +sample t this complete now? +This would be a nice API for JavaHow would we implement this? Would we serialize these to the splits file? To a new file? In a parameter to the RPC? it would be nice if could be easily made by applications +Any help in how to write test cases for this would be appreciated +Peter I tried to Generate Domain code with your reveng properties and all worksMay be you expect the function in some other exporterBTW +I added this framework in +Code Delivered. +Before putting this issue aside I want to record a couple notes There is a workaround for this issue viz +Thanks for reporting and thanks for the explanations guys +I reproduce the issue in Tomcat when deploying foo hope this issue will be fixed when upgrading to +If you added a new table for example and speciified true fo active record then the new object will be generated as an active record entity otherwise it will leave existing entities unchanged +We are just asking for tablescolumns limited by the given patterns +I took a look at the test case and indeed it does look like the situation I describe is tested in the test case +Would it be possible to support large ranges? If not could you suggest a workaround? If I divided all the numbers by a known factor and then multiplied them in my app could that tie me over until the behaviour is actually fixed? Or is it a limitation with sig figs and not the absolute value +Jeff can you attach the whole of the dtd for Nano the diff is huge and unapplicable on my Mac +Small testcase that reproduces the sql error +One clarification question +Fixed in r +My mistake +The solution is still the same +The short answer is yes committing this now would make it much easier for us to run mainline Cassandra +Reviewing the patch have some comments +Component a gets value a and will be submitted +Group sholud be has been updated accordingly. +Attaching process involves creation of instance which is patched by that time +Here is the patch for the fix +Sure thing +Finally managed to reproduce this problem under OSX by setting the paint delay to a very small value +Yeah works for me too +It no longer causes the system to hang does asnychronous writes to the cache and takes more care on closing file streams +Look at the and the life cycles there and wheres its called from to understand the wiring life cycle +Yeah the spec specifically says should so its really up to us +Try it and let me know if there are any problemsI think you may have found a bugThe idea of billing accounts is that with paymentId and no invoiceId increase the value of the billing account where as paymentapplications with invoiceId and paymentId decrease its valueHowever I think that now that we have the Customer return invoice this model may need to be a bit more sophisticated so that we need to say with paymentId and invoiceId where the invoice is a SALESINVOICE decreases its value so the to the customer return invoice would not cause its value to be consumed incorrectlyFortunately this should be pretty easy to fix +Formatting tags +Is the change to get in just for the sake of the tests +Please be more specific +In distributed search you dont apply the filters when it is a shard +I see +I tried to with winxp with your path no error you try mvn clean site U to update to the latest javadoc pluginand send use the log file mvn clean site X e am able to reproduce it by setting your CLASSPATH entry in my shell +What if another message arrives in the next millisecond? Wont it get the same ID +Better version with factored testsIt still needs some more documentation and Im hoping to add more cost functions +bq +So closing this issue. +Will test it ASAP +The interface mentioned here is a lot more versatile +Well give it a try +assertReferences needs an implementation although it should be noted that the previous version of the test did not have an implementation for references checksQuestions Is clientserver necessary here? I guess verifying the metadata methods work properly is worth running on the client although perhaps the clientserver tests should restrict themselves to the metadata test fixtures +Hi I will incorporate that +I didnt convert all the yet I think we can do that later in a followup too +or just store all the created Dependent contextual instances within the of its parentGoing down this route now +So Im on this patch and will commit it later today unless anyone shouts at this breach of protocol +I have the same problem in ajax components when is trying to solve beansCristi +Updating the PR with support for tail options in the dialog and more appropriate wizard description +The driver is modified to decrease the number of reducers in each iteration finally to +marking as resolved. +Links now point to cwiki. +I never got back to implementing this behavior and the locking involved is a little complicated +Problematic HTML test Hannes +The tests failures were not related to the patch +Now we by defualt FORCEMIME if MTOM is enabled +Most unix commands behave like groovysh does +Alternatively we could provide a new Tuscany interface to hold the options +ve attached to this ticket the rules to be activated by default +Im not sure how +Is it late at night where you are +At this moment i dont feel quite right to commit them in our svnThanks patch is to enable the twitter test to run in the offline mode and replace the remote repository xml with a local repository which will avoid the potential test failures as a consequence of the remote repository changes in the futureThe patch was taken from trunk dir +Then well decide to create or not a JIRA ticket +Im investigating +When you import these example projects into your workspace these projects will get their server and ESB runtime from your workspace preference +the qalab guys sent me the bundles for the qalab core and the m plugin to be deployed as I am familiar with the process for m +I attached new patch including +all issues to the CR +Consistency is just a tool for reaching an objective it is a means to an end +Are you sure you dont have an application override on? Id like to see your +The maven build is good much better than maven +PrakashPlease can you update to the version of drools since we fixed a number of bugs +Is the scenario having containers from multiple users asking for resources within their quota and then delegating them to a shared service to use on their behalf +This seems to work but its kind of a workaround +I added some commits +Uploaded a new patch with the formatting changes +This required some cleanup of references in several places +We will submit a series of patches about port in next . +However we didnt have any information then that this was a limitation on our user clusters as HOD essentially followed the same approach +That should clear the problem I hope +Not a bug since this is obviously due to a version incomatibility and the fact that the kerberos code was not updated to compile againstIve applied a modified version of Robs patch to the master branch +jar and +was called twiceThe first problem was fixed byThe second problem was fixed by +Verified the Job metrics using property enabled +Solrs replication test failing or taking long time on macosx only +Ah HA! Found is caused by attached artifacts having their own metadata which is written when the artifact is deployed over the top of the main artifact +Can you check the code? I made the following modifications Review the method lock +Nothing more to do on this +I thought I replied to this but it seems to have eaten my comment +Maybe we have what we need already? Let me look +ClausI dont think the varargs is a good idea +fixed r +Ill take a look +Verified on +resolved +Committed this change to the trunkmibmsvn commitSending javatestingorgapachederbyTestingfunctionTestsmastertSending javatestingorgapachederbyTestingfunctionTestsmastertClientSending javatestingorgapachederbyTestingfunctionTestsmasteroptimizSending javatestingorgapachederbyTestingfunctionTeststestslangSuSending javatestingorgapachederbyTestingfunctionTeststestslangoptTransmitting file dataCommitted revision +Basically a stray println was causing the regular expressions in to fail +Just having a Java Build Path Reference should do the is working for me in the latest version +qix files are already available on the disk image if they are the datastore should not try to write on them until someones tries out a bug reports coming in on this subjectI notice that when Udig read a shapefile creates two new file with +The original problem statement specified that the classloader lost the reference to the jar +Test output shows it is ok. +The user has the ability to migrate their project sets +Seems like good enough reasons to me for doing nano timeYes for these particular pieces of code not everywhere. +Maybe something a little more? Like ALL USERS or something like that? There is a similar but more pertinent display issue with empty string +Im not experiencing this same problem +I just committed this +Here is another use case User clicks a shortcut link Open Editor from a dashboard page +Robust alternatives are available in the compositepool package so one resolution is to close this on release of pool but I would like to consider addressing this in the branch as well +I also ran existing HA tests and they passed +This patch +having a corrupted edit log because we tried to roll them when there were no inodes available is a very bad thingCertainly true though it seems unlikely that this would result in a truly corrupt edit log that the NN couldnt restart from since running out of inodes would presumably result in no file being created at all which isnt an invalid edit logbq +Serves me right for not checking latest version +But many stack trace falls in console about class not foundAlso I turned off beans for logging in +I hope to have a patch available soon +Colin I have already commented to this effect on +From a practicality standpoint not as enthusiastic. + +QA verified +is a bug in deb target attach + +Test plan update with latest state of and perhaps final update of the test plan fixing block recovery FI test descriptions and synching the document with actual status of pipeline tests developed +So i guess we need a better definition of reliable +I agreed either using put or replace works well in my previous comment +But current directories are passed into the server with a system proproties +Thanks KnutRegressions rans fine in insane mode +Please have a review on it if you have some timeI finally did NOT delete all the methods in the patch initially proposed because of the discussion here +According to the time when the test case have been attached Im quite sure it is one of these we lost +The fact that the counter only got to means that something locked up +Youme is the selected value but the converter is caled with You and a second time with me. +Hi are you sure you are on? I am asking because I had exact same errors when I built Guvnor with JDK +Cant make this change before becoming that removes +contains an alternative patch should you wish to keep the alphanumericrestriction on service names +The patch has been applied and undergone sustained testing +Thanks. +Hi AndyThanks for dropping by +Caused by special configuration file jrelib in IBM JDK +There seems to be a lot of places an empty string could also show up +Recommend we put this in JBDS CR and get in for release +Viktor why we have this defference between CA with EL prefix and without it +I mean for storing d in you dont need a wrapper +Dozer works without that only one implementation of SPI interface depends on Hibernate +Can you try if this is still an issue +Applied thank you! +Committed at subversion revision Touches the following filesM javaengineorgapachederbyimplsqlexecuteAdded dependencies of aggregates on types similar to the dependencies of routines on typesM javatestingorgapachederbyTestingfunctionTeststestslangTests for GRANTREVOKE USAGE on types which are used by aggregatesM javatestingorgapachederbyTestingfunctionTeststestslangTests for attempting to drop a type which an aggregate uses +But I agree the REST support will not be based on the SOAPPOX part of though we will leverage the XML Marshalling module in there +committed to TRUNK +there are almost no scenarios where one would want to include some JAR dependencies inside the RAR and others at the EAR level +Changes to associated with have resolved this issue. +However that is what the Xalan team published for the release to the Maven repository at Apache and that has been synchronized then to central +the patch readyAh forgot to flag it as patch availablebq +Frames are deprecated and are being phased out I do not feel that they should be supported as a result. +committed +finebackport does it now +I see you added a Xlint option is that our current convention? +We just call slfj functions which themselves call logj functions instead of calling logj directlyIn any way the bug is now a year old +This was allowed in m fine with allowing multibyte character support here +Should I create a new JIRA for that? +Added minimal project needed to replicate the error I am getting +The failed is not related to this issue see +Thanks Colin. +I ran it on current trunk +This issue is dependent on which will provide the necessary JBC event granularity to ascertain which keys have been modified or removed +Thanks for the report and fix Steve +This is a useful feature to have +This is per spec and not much we can do without violating the spec +I cannot replicate thisIf you still notice a problem please reopen this bug and provide more details asto which page and which links are broken + tests this +Therefore the APT output is redirected to and not handled by the implementation +reopened because implementation was reverted due to conflicts with SesameChanges marked in code with FIXME same issues as Andy describes for Jena will occur when upgrading to when it comes out especially the checking of datatype before language patternI will be recommending that users dump their Sesame databases to and reload them using as the simplest way to ensure that all literals look correct inside of their database +The writeObject method now completely validates the serializability of the descriptor field values in the writeObject method to ensure that values do not prevent accessibility of the mbean in some form over connectors +Could you check out the sources again perhaps you had a download problem with that imageThanks you please add a patch +Except one user Waalsvanderwaal You should have your PMC chair check this then ask them to add the user +Add a link to for an admin console bug. +this mean youre giving up on eliminating type tags from nested containers +I updated to Beta and the problem seems to have been resolved + has seriously reworked recently in Xalan +Thanks Kan and Owen +Just to answer my own question Does this fact get logged to the console? is being opened with true so that means the contents of both servers logs go to the same file +Agreed +looks good +In the the android file it uses the network onlineoffline events to start up the communication with the native side +But there is not yet any official releases to useAh interesting +Any objections? error Unknown BYTEORDER endif ifndef ZADDREFPdefine ZADDREFP ZVALADDREFendif +Im all for getting more information and doing this right +Do you have more than billion keys? Or sparse long keys with numeric values above billion +the question was motivated by your +the message Im seeing is the same as the one noted in the referenced conversation that prompted this changeIm running with the M why am I not getting the updated format? Should I be +bq +Fixed in revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Seems flakey in general also failed on Build Build Build Build +Also once this is checked in please update the wiki docs +The approach pool is taking is probably the cleanest change the groupId packages and possibly artifact Id to avoid any classpath conflicts. +OK thanks Dave +test case for reproducing the in svn r +I appreciate your working on this as I wouldnt have much more time until next week +Im going back to my original suggestion of cutting out the localhost special case +If this is still a problem in then we can reopen. +Please note that this change should be made for both and CP releases +Added Dona to the Avro contributors group and assigned this ticket to herI did interop testing with Java and we did not see issues with decoding Java data in C and vice versa +I didnt do enough to document the replacement classes although in most cases they have the same name in the new package +This is a test that proves the the patch is faster than the current implementationIt also includes sanity tests that prove the correctness of the implementationI tested it with and I got the following resultstest millisecondstest millisecondsWith mina patched with the binary search the results aretest mstest msYou can see that this is faster timesThe tests were done on a Laptop with attached an improved that detects the errors from the existing issues +thats whats causing your problemthanksdimsDavanumThanks so muchSwavek +All except the last retry should skip writing history file +So I fixed to install useosgiinstall mvnpingerosgiinstall mvnpongerThe description in the is wrong +so I plan on committing to trunk only took a quick scan but looks nice +FWIW Ive attempted to add jstack to the out of memory error and evenforcing it fails +Not a bug +Thanks +All the subtasks have been resolved +Closing old resolved issues +Heres the example +bq +I have forked the code from github and if I understand the code enough I will try and put together a test case +Change resolution back to fixed. +ah ok now I get that you want to make something more basic +Looks good now and will be in the next update. +for new patch sync warning from concurrency warning reported by findbugsThe core test failures also exists in trunk builds and looks like server configuration problemI am not quite understand the hdfsproxy test failures and I will continue investigating on it +Please check that it has been applied as expected +ok Ill add a mapping to the api for it too +Do we have a conclusion about the current patch so we can commit and work in other issues to improve? I also want to open issues to remove broken from core and move to sandbox +Hello Would you please try my patch? To apply the patch please +Have you tried the user forums or knowledge base? When exactly is this problem arising? Is it a standalone client or a client running in a container a server talking to a server or something else? You say youre using EAP and AS which is it? What are you doing to cause the issue? How could someone else reproduce your problem +Should have tried it out in the others +Can you suggest a better idiom or interface? +If it is in the client side we may not know whether we have received the response in an async requestresponse case. +This patch changes to a simpler more explicit configuration approach +I confirm I am having the exact same problem with Tomcat and Axis +Checked in the patch itself as it basically prevents people from using the newtransaction manager code with global transactions +One pain of oom is that the thing leaking the memory is not necessarily the thing that gets the exception +Has nothing to do with this repost this bug report with groovy. +Eclipse crashed when I tried to disable this But yes Ill do that +Thanks for the patch Olivier +No this issue is not provides a workaround for this bug +Trying to develop Solr as two different systems is too costly in the long term IMO +on pulling the etc +It is then up to the to provide this info I guess +I think adding test methods in makes sense +Spring is to be removed in WFK. +Im setting up all and production to use HSQL Clustered +Dont think we can resolve this problem without major retrofitting of Multiplexer +The patch applies cleanlyHadoop QA only applies patch on trunk +If I hear no objections and no one steps in to fix the connection GC code Ill commit this patch in a few days and close the issue +The default context is what answers when there is no contexts available +Not to be committed +No capitalization anymore +This is written against +thx to Ryan for providing the patch +Will be available in next snapshot build +Committed this to trunk and +Committed to so that we can get the system tests in working. +vpetests +Sorry I cannot find a connections security tabMy apologies the tab name is Access CredentialsThe document status report gives no clue as to the reason for the hang +Willem whats the status of this? Seems like the end user should use that partClass stuff in the dataformatThe user never send any other comments on the JIRA I did some test on the annotated classes withIt looks like the can pick up the rightly from the class rightlySo I close this issue with can not reproduce +In fact No i think i write the image only onceI finish to write and clean my solution and then i will send you a copy +This is not true for modules +From Olivier at lunchtime this is impact the runtime as wellCant be resolved as not a bug +Patch to change the usages to +Dont be afraid it is only the name till we find a better one With this only a subset of Groovy will compile +Just to get clarified does this mean dumping specific column families for specific tables? So the snapshot command will take both table names and column familie names +If thats amenable to you Ill go ahead and file another one +verified in verified in BRMS +According to MSDN and are present since Windows so this should be safe if some child creates a child process with CREATEBREAKAWAYFROMJOB flag this child process wont be terminated +I havent tried this sequence but it looks like a general compile error + +The source if you need it it is actually very similar to this in from the real from the example this time with some container that holds both suggestion box and input +Feel free to comment this ticket if you disagree +this should be already fixedthanksdims +varlog file with the below changes INFO +If the URL you are giving me is manufactured by the portal tell us how it was done with a new JIRA issue for it +Ondrej the EAP default profile does not ship jgroups +for GA detecting qualifier annotations as well as a generic return type declaration on factory methods +Changing fix version. +We should also with the the G garbage collector +committed to support will be a new feature for JBDS we will not document everything it can do in the release notes instead this will occur in a documentation guide and the new component will be mentioned in the overview and new components sections of the release notes + to trunk +I agree with Brian +Thanks for clarifying +Hi Andrea the speedup I get from this patch is twofold and it makes a difference in my testing with jmeter With the createindex flag set before this patch was applied the VM I was testing geoserver in literally locks up after a few minutes of hammering the server with getFeature requests due to an IO backlog caused by geotools attempting to see if the index is up to date +Based on tests in a private email conversation I believe this is fixed now in SVN with the changes I made + +This is obviously broken in concept since you cannot tell an error result from an equal resultI see these options Document it better leave the code as is Change the code to thrown an unchecked exception that wraps the +I never did figure out why my initial patch for caused this problem +You want me to try this and commit it +Im not intentionally trying to duplicate the code I was attemping to entirely decouple the embedder from the CLI notions +Ive committed the move on trunk to contribvelocityI see this issue marked for also though is this really needed? I personally am fine with this change being solely for +The current setup seems to have +Does it work on an hbase server +Addressing Sharads comments and adding Apache license to the test class +Still doesnt work still broken +Before tryingto add the relevant code Id like to hear opinions +I think that Stepan Mishura could help +Btw it is also related to Ians patch about killing links for i believe this is resolvable +I dont see a problem +There are many +I ran netstat and checked to see if there were any zombies none foundThis test just failed again +Closing resolved issue no response from reporter so assuming ok. +Roman I was aware of that thanks for pointing it out +binsqoop and it looks ok +Freddy should we also add average number of builds between fixes longest time taken to fix a failure greatest number of builds taken to fix a failure +Do you have any plan to add an additional method to the so that I can pass in the WSDLJ objects instead of URL? In short term I can create a similar utility from your code +Any chance of adding some kind of test? Maybe a data blob that matches the old format? Looks good otherwise +If I run the standalone test with debug then I get OOM errors from the lock logging +Its not necessarily going to start a trend and we do want this issue fixed so closing it as wont fix seems inappropriate +I have added a test for this problem to the existing +This test case uses which I also extracted from my ro mode patch +This is most likely a case of either attempt to connect while another pooled connection already exists with the same client id or a case of a stomp consumer that has not cleanly disconnected from the broker. +Im gonna shed a tear guys and gals +since this would change behaviour +I took that approach so that the project compiles and runs even if the script was not run to generate the version information +as i understand patch is against hibernate but we already work with this is most voted bug in hibernate +Let me know if anything else is neededCheerss the banner as well. +Weve never had enough information to tell whether there is a problem here or what that problem might be +Could you update the patch if you have time? If not Ill take care of itThanks! is this scheduled for commit? Were about to release a new Solr version which faces issues with the logj dependency +A recent example of this with some Adobe patents was already answered here previously and you might want to refer to Jira for a similar issue +If you have time could you update your patches so that the full list isnt displayed when the window first appears? Once that is done I will commit the remaining patches +Im kind of against adding the rar into the distribution +Changes are trivial +Yes actually this has been verified to be a serious bug +Unfortunately this update also included new schema changes for and these should be ignored +Got Jesses Kindle Fire for some reason the screen is resized before anything else happens +Im assuming the failed test is a flakey +This patch uses the responeWriter to output the error +Fixed in revision +after looking into this one more time I think you simply do your patch +Maybe a new transport flag for the email sender setup command would be most appropriate email sender setup transport DEFAULTSSLTLS +This has been done for Note this does not imply it is done for as that distribution uses a different version of Seam +I think its low priority +Earlier release patch was created using git diff +Once we start returning only active nodes well need to add the filter API or just modify get or to take in targetState +Please apply patch +Sorry for the inconvenience +Reading the code itself is probably the best way to understanding the change + +Also attached the current +If you want different behaviour you would need to use a different request cache implementation +This would be used to determine if the locality groups contains any columns of interest at seek time +Patch will need to be rebased since code has moved +Thanks work +I dont have much experience with git +This seems to be fixed in +we already havemavensynchomegroupssseseleniumdslhtdocsmavenrsyncsshCaue GuerramavensynchomegroupsccahtdocsmavenrsyncsshCaue GuerramavensynchomegroupsttutubainahtdocsmavenrsyncsshLuiz Realdo we need to remove them and just put the one you submitted +Right now when the maxIteration count is reached an exception is thrown by the optimizer and there is no way to retrieve the best solution that has been found so farThis is not acceptable in some +Id suggest you setup exception breakpoint and evaluate classloaders for your class and its ascendent +Looking at the patch for reviewcommit + +Youre right I misread +Hi Weinan please do this upgrade for us +the goals of the implementation have beenas its name suggests being simple having zero deployment requiremnts and minimal dependencies +Its a wrapregardsAndrew Dinn +I currently use a constant instead +Hmm +Closing the defect as per Saksham comments +So please reopen with a demo project that reproduces the issue +The issue does not seem to be public so I cannot access it +I also run for release and can not reproduce this problem +I just downloaded and I still have this bug +This is Eclipse Web Tools JSF validator +The updated patch has the Assume statement in Before +issue still happens in +Another potential issue is that the RM and AM each have their own interpretation of a jobs result so the State and reported in the Cluster Metrics display may not align with the defined states in the Cluster pulldownThis is the main point +ant found the issue its because a got into a bad state where omitTF was true but storesPayloads was also true +Also weve moved to using so for future patches please post thereThanksJon +Whichto be honest I really appreciate +Kathey I will go ahead and close the jira for now since we dont have a repro +I wasnt able to replicate this issue on branch using the poi dataset thats shipped with geoserver. +Thanks for your help! +We will not make fairness decisions based on it so it will not be similar to bringing slots back and will not be a substitute for using resources that already existWhat are your concerns? Is there undesirable behavior that you think it would encourage? +Ill implement this enhancement next week +In general the changes look good +Im sorrythe test file is not +Hello Would you please try my patch? To apply the patch please +Im thinking of hardwiring to after release. +I think that opening this new jira issue is more appropriate than reopening older would be also great if you could provide exact description of what is broken and how to reproduce esp +add wiki tags so the show you have a xerces or jar in your usaaraojar directory +This method assumes that it will always get hostnames and splits the string by +This should probably come from the system classpath +It was done as expected +You might not have known about the mailing list +patch and editsStored after merging latest changes from trunk +Thanks Arpit +Good thing is that ajlog console shows the correct behaviour and it works perfectly +Please confirm this so we can close the issue +Issue was unassigned +In response to an email request for impact assessment Excerpt +This patch adds a little message under the results table with the following formatView had a request duration of +Luckily this only affected +No this is an execution time cost for every statement +JFYI would also for tomcat we always store the beans in the session than we should also remove the failover stuff +My apologies for misreporting +also svn rm lib +Now that work is well underway with AS all previous community releases are +to backport for +Ok i understand the property and indexproperty +Resolving + has been marked as a duplicate of this bug +that should be our own response which does not wrap the RIs response +Ill get a release note up in a few + I will make some commits on this issue without first submitting patches since it involves a lot of subversion commands that are not amenable to patches +I just committed this +NoBTW my problem at this point is solved as I put on our local repo but I am hoping the above will provide a longer term solution +In v and deployed to production +There were a lot of configuration problems in your app +In our case I think the issue is with internal java classls vobnmacsacsinstalltargetstagemgmtwebappsclassesorgapachejspjsp shufflejsp the files actually unpacked? I wonder if catching this exception in the dependency plugin is a good workaround for now +As soon as roberts automation patch is in core we wont have this problem anymore anywayUsers should add trivial extensions like on their own +Additionally storing column statistics as a keyvalue pair in the tableparams table is not space efficient +Make an addPlan method that takes a Map of plans in for Teds review +i ended up wrapping the model parameter in an which uses the renderer if available +OK a more general fix and tests to go with test file also includes a test for +This uses the module as a platform and additionally provides Jetty as the web application container for launches and runs Selenium driven tests against the test site +Attached you will find the corrected implementation +Could you provide a testcase or at least a model that when stored is able to reproduce this +Mahadev shouldnt MR always be able to survive a task failure if there are sufficient resources? The client shouldnt have to distinguish between types of task failures the AM should just the task on another node +Brian just as an FYI you diffed again rather than trunk +Thanks Rob +scheduling for at this moment only osgi dependencies in manifest are for the convenience of finding particular dependencies in eclipse build path more quickly Id suggest the following SORT OPTIONS by artifactId by group I consider it more important to have the option to control CLASSPATH ORDER by sorting classpath entriesaccording to the order of dependencies as written in the POM +I may end up hosting this as a separate open source project that can be linked to from the clay documentation +Checked in the documents in docsreleases and xdocsreleases for testing +Thanks Yuri +Daniels increasing time out worked fine for Lee +indicates that hes using the excelExporter component tried with that one +Add a field to Product entity and go from that way +Whether or not we like the behavior its necessary to be JSF spec compliantSee the FAQ for information on how to submit a change request to the JSF spec. +Fixed +Thats certainly true but the Teiid instances are still not expected to show up in the Server view +uploaded a simple patch that changes localhost to that usecase works with this patch from me as will have to wait till to be merged before we commit this + +We have the same problem which is also becoming a production issue +At least I have the source codes loaded so I can start looking whats happening +This fix is in version and not in alpha as mentioned above. +Is the use of the constants DRDATYPENLOBLOC and DRDATYPELOBLOC for clobs in correct? I would have thought that DRDATYPENCLOBLOC and DRDATYPECLOBLOC should have been used instead +So let use commit it only to the master branch +Linking to the Tech Review JIRA for this the tech review process this issue has been and close +but we need to integrate get with termstate to fix the nocommits inThis should also finally allow us to fix the cost of that extra docFreq +so I added a unit test for that +see previous comment added backported to not sure about a resolved issue so created +it doesnt make sense to me to have a surface form at all onThis is an iterator of terms freqs +Do you really want to ship and deploy test code on your production platforms? Really reallyI encourage developers to run tests +So to simplify I would rather the current option support two different persistent formats +committed in rev strings were not fully translated so I am reopening this Patch Attached! Issue May be closed +Will commit once builds test all passes and if there are no objectionsand after the patch gets a of course +Could you explain why this fixes the issue +You have to autowire in the converter to your test +Committed as rev . +Agree with Vinod we cant have zero logs makes it impossible to get an operational view +Have we been passing all the existing tests like impincl? If so then theperson investigating this bug should determine what loophole was not coveredin the existing tests +btw if you run long lasting tests then consider removing the mocks because they store the received exchanges in memory so you will eat up memory over time +New SNAPSHOT published. +Filed. +It should be uBuI attach new file +so it will try to undo the previous copies and then delete the dir +Should I close this one and open a backport patch +This is probably one step +pushed +Adding dependency on artifacts Changing jackson artifact versions toAlso updated patch on reviewboard +Committed revision +Surely you can take a stab at it N +I was seeing the too many store files issue recently too but that was even without this patch +Also took a pass over the reference to better consolidate the sql support. +Is there a validation test that goes with it? to trunk at +Hi Jrgenyou are totally right the version spec is required at the beginning of an LDIF file so it should always be addedWhat I meant with no sense was the following If you on an entry and select AdvancedCopy Entry as LDIF only one LDIF record is copied +Done in r. +rev +XufengYes you are right +To recreate the bugant deployant runtestant undeployant is a new problem in CP +Moving this out to for now +This time it is right +Changed committed in Git ID OK so closing. +Could you please submit a patch for the branch +createUri is a pretty generic method +Kihwal we removed reference of in a while back to cleanly separate these layers +Cleaning up old issues +Hopefully we can release this code in the not too distant future so it will gain visibility. +Update icons for component servicesreferences and address gradients for composite servicesreferences +applied patch thanks. +I would not like to destabilize Roo in the release +Thanks Divesh Sumit for the contribution Done at rAshish Vijaywargiya +Ali Just opened and for refactoring probably for Cargo +Thanks +We should test with current CVS code to ensure that this is not a duplicate of bug +both of them failed +Do you have an HDFS jira? We are attempting a similar thing with but requiring an online schema upgrade so cached values can be refreshed +I am not at all convinced that running an export would be inherently any more unsafe than viewing a page +Alexander is working on this problem for us +version . +has been added with description additional has been added with description my output for additional sample +since were using the same problem exists in +Dont close the issue as some others will come +Havent tested it but the idea is that any insertion of a triggers a memtable flush +Ive added new checks in the unit tests to verify that the current parser continues to parse correctly the new statsquota nodesIve also changed the parser so it can handle parsing unrecognized fields which should make it easier to add new quota type in the future without having these backwards compatibility issues +So yes the issue with the scratch file exists and there has to be a solution +unit in r +thanks for testing this +Lists createdmoved + to reproduce the think your workaround is the solution +Could you try reducing the test case +Reopen for backport to the fix to with revision . +Very usefull! Verified on Tools +This patch allows maven and builds Giraph with Hadoop +Can you elaborate a bit? Perhaps attach the Controller code that illustrates this issue +We should either swallow the exceptionor throw it as some kind of +First shot contains basic tree will continue tomorrow showing the current content of the selected file and sending modified content back +and branchx rev +No rush but when can we have this issue resolved? It would be nice to be able to run testsuite as well +I am leaving the bug report open until javadocs for new classes are provided andreferences to deprecated methods removed +I will update the patch +Vladimir has excluded but we was talking about +A call to tailSet does not make use of the current position and require relocate to there over the skip list +added for adding header released +orgapachehadoophbaseioencoding +It will remain dead until administrator manually intervenesI prefer option +Oh the new File expressions are very cool +using the p parameter will launch the partial implementation and without it will use the in memory implementation +As visualization is also part of my research interests I am excited to collaborate with this idea this year +Diff for files in modulesxmlattachments folder for files in modulesxmlom file for include folder client sample have commited in the patches +For MIN and MAX works because the seed for MIN Double and MAX Double could be clearly defined as and +Do you have a test case that illustrates this problem +My work is not finished but I do have something deliverable that is internally consistent +Can you please so me where you think it is not possibleYour suggestion of using custom transports seems to me ugly and somewhat not clear +Willem a million thanks for the above configuration +The benchmark starts a new mutation as soon as control is returned to the main program +yes I wthink thats good idea to move code into now using you can creat a service for client side or you can create a service for a given java class as well + on RB +simplification done some time ago +md and sha the rest as WONTFIX and someday we can kick this out of the repository when we clean things up. +That entry has now been removed from the initialisation data will also check if it is created dynamically +The XML file selectionI will remove the Load button and make it to load the file itself when user click finish button +I just committed this +I couldnt regenerate the problem in my machine should try with valgrind to identify the corruption +test failures seem unrelated +Ive updated existing PR with another commit +defer to +Is this ready to be closed +Maybe thereis already something available in the classpath project which you mightreuse +There was likely an issue with how they were attached that caused them to be out of sync +Apparently a null value is being passed to the regexp match value +Verified in JBDS Linux b JDK +Attaching a test case +Removing it should not have an affect on the +merge to to docs as svn . +Any help appreciated +. +doesnt need to import +Why does maintain in springframework core API? +An advantage of this approach is that it allows blacklisting +This would simulate disk Io errors and network partitions thereby triggering code paths in HDFS +According to the forums thread this is a task for Dimitris not a bug report +Im going to look at the foreach slowness since its much more drastic +It causes other issuesI am now using a workaround that adds an exception in similar to the others there already to recognize the commentsURI as being a local refererA longer term fix should probably the flow around the submission and previewing of comments +We should get this into the codebase as this ticket is popular +Regression test has been integratedin the test suite. +I think you forgot to commit on EAP branch as I didnt see the fix in the beta release available on websiteIs it possible to fix it on that branch before the Final release +I dont remember exactly why we did this but IIRC it was something to do with the inability to always and unambiguously determine the right setter for the propertyOn hindsight we should be collecting those exceptions and reporting them at least as warnings in the Problems object +it looks good to me applied to HEAD +My only suggestion would be to rename deadCount to successCount or something like that +The trunk patch may not work for multiple namenode installation in a federated cluster but I think that can be fixed in a separate jira +Fixed javadoc and findbugs warnings +Direct instances of the various collections that have their types directly declared there should be fine +Although they use a very similar mechanism the seperation is important +Close! +that is not the case if the index is on a key column +Also further modified and to use coreName in creating core +reopening this +Heres a patch implementing your change +provided patch based on Horias code sample and suggestion +neither null nor its length equals to +Im asking a friend of mine who is something of an expert with Active Directory how this all works to see if we can come up with a resolution for +Also added and a goal that should get everything +Instead were looking into having a counter which accomplishes this. +can confirm that its the version of the rest plugin +Martin can you verify the instructions +I wouldnt want to have the version printed extremely often during normal operation though +I have implemented this in vbranch +Yes I just tried your test with ICU and I get the error +bulk close for the release +Propose this be moved over to Commons Math +More is attached to patch +So the behavior that youre seeing is exactly what you are supposed to a for a collection plus element type is nevertheless a valid use case in particular for testing +If then will find the user making the request in the RPC request context +Add a test of the desired behaviour and you can be confident that it doesnt break in future +Whereas in case of it was having an impact because code uses the source position from the itself which is by comparison one potential solution could be to have a wrapped inside a and make assert code use the source position from instead of you think it is a worthwhile thing to try that change? Or do you see a simpler solution +A more pluggable execution engine enables a Storm backend more easily + +I could not make a good example for because is an invisible controlSo I committed a unit test for control instead of an example. +added license header +This bug has been marked as a duplicate of +I wonder if the starting point was the tax law had a minimum item not purchase priceIf the tax calc service has access to both the list and adjusted price for an item we could even add an attribute to to say which one should be used +Yes you are correct +Actually I wrote this part it is independent with Yings Accumulator code +Ok Im wrong +I will revert some changes of this patch as a test to make sure the patch to is good +There is little difference between the two since is a conglomerate of the three changing the package name to versus If using the RPM distro edit etctomcatX +Maybe my workspace was out of sync or something like that +patch looks good +I too have had problems like this over timeAnother way to deal with this might be to hand out a new random ID to each instance on startup +so closing this issue and anything else on this topic can be filed under the project grouping component +Another solution might beright +I started on taking the approach of moving deletes to a classthats a member of IW +fix pushed to with the release of + +operator cant have more than one child or rather any children of cant have more than one parent +The test case only applicable to pig trunk +i had started the server the EPR is right too but i cannt connect with the server Connection refusedconnectI use java whats wrong? the structure? the commando which is to run the client? or the codescan anyone tell? +need to email the gossip folks to import the userdevcommits mailing list archives for +Then we also have time to commit your great patch as well +This change is reverted because transferring POST parameters into page parameters leads to all sort of problems when the POST parameters leak into generated urls. +Yes I have changed the file and testsuite ran without in classpath +This is needed when using the splitter aggregatorNote that other components may need to be enhanced too +When you put them in the middle as you have done here you need to up the rpc version +If someone can come up with a clever hack patches are definitely welcome! +I guess Im +Thanks JuergenIf you have any questions to help you understand and diagnose the problem please do not hesitate to send me a message +Marking this for to see if we want this included in +That change does indeed apply to the GB version too +Kurtits a long time ago but can we still expect patchesIf not Id like to close this issue +Merged to the branch in revision . +So how do we get a copy of a good sample tutorial between now and the next release? + +Maybe the patch is reversedYou should also change the ASL header to the ASF one not the Zenika oneAlso dont forget to check the ASF grant when you upload your patch +rather than the one we care about +Ive committed a potential fix to trunk +For any further insight its up to Thomas as the original author of that code +So this can probebly be closed +Includes patches and +As always there is risk to bringing new code and we have to asses risk vs gain +The expanding of empty elements is a domj issue I have opened a bug report there +Includes fix to detect when a successfully authenticated user fails login +javadoc cp errors are fixed +Thanks! These shouldnt take too much time to work through +Thanks! It should be fixed now +my apologiesI will see if I can find my info and update this ticket shortly +Just because your IDE handles it doesnt mean that it should workPlease provide the full stack trace so I can at least see which part of the testing stack is rejecting it +HiCan we have a svn diff +Since we cant demonstrate a consistent win with this patch I think we either need to close it as wontfix or go back to the drawing board +Hope that helps as well +See +was a hack for the old swt versions problem on Mac OS +Using new jar now +Generated new patches using so robot can test +Or you can add the code base to one of the security policy files that would allow it to run as priviledged +In the mean time index page stealth forced the index file be flushed +Patch applied except removing from the parent POM which broke my build. +When the was written it was meant to be a static reference to the list of the jobs which are currently scheduled the job queue +Thanks Amareshwari! +Now we use the s bus instead of creating a new oneTODO We could add a busId option in the cxf uri to let the cxf endpoint to chose the bus from registry +Applied the patch into trunk and branch with thanks to Antonie +Thanks Yi Xiao for the patch. +Im a little confused +Fergus seems to duplicate s functionality instead of using it +Use one of the existing tests and modify it +I guess well have to max out compiler configuration and take into account settings declared in executions +we should fail way earlierIMO this AM needs a rewrite if only to make its various transitions testable +since not all of the node types and configurations are available in the xml and database persistence +The test case should be able to handle this +This patch can be used to fix the context example +Merged good job ivan! +Assigning to Eddie for review +Hi GuillaumeObviously Id like this fixed but Im happy for it to go into a release rather than a release of bundlerepository +The tests PASSED for me both release and debug DRLVM on ia +Closing as this is an ESB issue +is opening session per request not per http session +We do so for SET TRANSACTION ISOLATION +That sentence was intended to warn users that behavior was undefined over streams with position increment of zero +This should be fixed handling of operational attributes has been changed see also +This looks good to me thanks for fixing +We will try to improve the state of things in +Please review and lets get this party started +If this is not the case then we are barking up the wrong tree +Thanks +Patch applied in r with this change linkOnly param is always true in offline thanksAny plans for the next release +still missing info license scm. +Doesnt COMPACT STORAGE provide the ability to model thrift tablesIt does for some +Mircea Im fine with your comment but could you update the code to add a note that we decided to leave the validation where it is for the reasons you said? This way someone that looks at the code wont be wondering the same thing. +It reports to the name node and aborts +svn merge c +Fixed. +I will post the results once done +Is this fix ready for download and use +If no secret is set a random key is generated +Please post a patch there and Ill commit it +Please find attached the modified and classes for heres the patch file as per the Math developers contribution have just a comment on this proposal but beware I have almost no knowledge in the field of statisticsIn your implementation you seem to use as the mean +And it is very difficult if not impossible to ensure capacity before queuing +IIUC xmlbeans will be added to the surefire classpath but never the compile cp through using +I cannot reproduce while editing srcmainwebappthumbnailMy environment Windows XP JBDS Java what is your OS build etc +Fixing a typo in an assert message +The new client will get from the old server in this case +Committed trivial change to trunk and branch +Cool thanks for the benchmark +using refresh option does the refresh option actually do? a complete fresh cvs checkout +But we might as well update these versions anyway +I should mention that the retro approach has the drawback that we do not get compile time safety as with the comment based approach +which node owns IP provided by EC? i think you have to check your vpn settings +for commit with a wish for code review first +Unfortunately the current descriptor code doesnt make this easy since it keeps no record of whether a descriptor is inherited or not +Thanks for pointing that out Zheng +Patch and committed +It runs deeper +Can I take it for GSOC +There may be more needs outside of Embedder but I think it would be a mistake to remove streams altogether +Therefore will receive a Session within a transasction allowing for write you experienced a concrete failure or do you suspect a problem at a conceptual basis? I wouldnt be surprised about an issue in the details but at a conceptual level should work nicely with or even plain JTA EJB CMT Juergen i have just looked at the latest from CVS and see now how this has now been implemented we are still using version which we chose because it had been around for a while when our project allow me to save some face it was documented as working in this version +Output interceptors are moved to packageModifications of JSON schema are merged into the test sources +Chris Since is imminent I doubt were going to do an release but let me know if you still want this backported +tried against top of branch failed with different error message and worked directly against tableselect name from myview where name like ColERROR Comparisons between CLOB and CHAR are not supportedij passesselect name from mytbl where name like ColNAME rows selected +No need to iterate the list you are fetching for itemIssuance can be used +Especially examples that work and easily understandable +It may be better to add support for the function + for the patch +The hjoin work that Kranti has worked on has not yet included +Then youll have a greater chance of someone taking a look and being able to resolve the issue +Hey Ron were looking to get this completed asap +I can just pull the field from the first search N results and hunt for repeatsDoes this patch do that efficiently? Would a custom or be the right way to do that +Patch for trunk +Whereas formulated questions have been answered and checked I think the issue can be resolved +It does work when you explicitly demarcate transactionsThe default is to rollback so unless you commit you lose your changes +So Ive just committed this to trunkIt doesnt make sense to commit this to until and friends are to so leaving this JIRA open +Hive already has sort by I can not see a reason why we would need order by on anything but the final result +Thank you Tom +Added a test case +Renaming testing directory to test broke things python had a test module +We will break layout of page if we will make Seam Mail tags visible +Moving out of +This shouldnt be allowed +Hi Atulthat looks good at a first glace +Not sure yet if we still need asynchronous recovery but the original issue should be you confirm that the original issue is fixed now in? +this is a patch against the hadoop Bradley I dont necessarily think this is the right fix since it only works in the case that the job submitter shares a group with the user running the JT which shouldnt have to be the case +The fix is simple but I am not sure the way it should be should the fieldName have the in all cases or should it be removed +I also agree with Gavin this should be provided by the dispatcher probably as a Seam component +Will not implement until we acknowledge real needs +We also should learn from the hadoop folks +I ran the example several times and it seems to clean up +We now call both commit and rollback but only if autocommit is falseI wanted to note that I am now quite convinced that the only correct behavior is with autocommit false and calling commit and rollbackIn the case with autocommit false the transactions are reproducibly not getting rolled back properly. +Consistently use path and pathString +I dont think that this will ever be fully closed but we should try to keep things down and file individual issues when they come up such as the weird Dead Code warnings. +for the patch +This is the expected behavior and I removed the state column some time ago. +Please check that it was applied as expectedIve had to add a symlink to in file just like youve done to +Given that the authorization popup should be just a bit of explanatory text maybe a logo and a usernamepassword combo what about just rewriting the content of the gadget so that it renders the popup inline +Just one comment apply this checkstyle will make harder keep in sync branch because it mess up almost all files +The fix is quite straight forward +The bug is in +Then I compared the ivy library dependencies that we are declaring in the sources of the three packages and the libraries in the lib folder of the built packages +Thanks Chiahung Lin. +Furthermore I did a complete refactoringreplacingremoving all deprecated stuff and updating httpclient to beta Can you please open a new bug for the non set the severity toenhancement and create a new patch for this enhancement using the current codeFurthermore please test the recent version of the ifstill everything works okJoerg +Or am I missing somethingSame applies to and its jdoorm representationNot sure what is testing its a concrete PC has jdoorm info +Weve just started having some issues with this bug as well trunk patch is attached +Maybe try a git pull again to ensure you have a full copy +This way its easier to track progress for each specific for your help Eirik. +An idea Ive been thinking about is to change the add to also include solr paramsfor each shard +Could this be some sort of concurrency issue where does some class compilationreloading after some given time? +I hope Ive fixed the bug +This has been fixed in how we organize our repos +Outputting results on the console is good enough to parse and store in whatever database you want +reopen it to deal with followup for the follow up patch most of the changes are good +created to track this +ls liblib standalone version based on eap installs class occurs in thirdparty does not exist in EAP beta and therefore does not make into either SOA Ive assigned it to you as the EAP Jira was assigned to you if its fixed in EAP we will automatically pick it ll look into it +Im looking into this issue +If I dont do that urls with parameters are never matched + +Looking at the latest code +thanks! +I closed it because there was no comment anymore +But you need to remember that this is a community project and the KISS philosophy should prevail unless there is a real reason to complicate things +it was almost too easy Im thinking something just isnt going to m going to assign this to Charles Lee Charles can you tell me what some of the issues were that caused you to pause wrt this issue? What are some of the things you can see going wrong? It compiles fine for me now but I havent run with it yet im scared What things would you expect to break + should be applied to the we rm the developers section from too? I dont want pom to be cannonical for anything and I dont want to have to maintain it in addition to the cannonical place either +How does user verify whether the above commands returned inaccurate results +Updated and installation guides and in the upgrade guide for to clarify that these jars must be in the commonlib directory and removed from the distribution lib directoryWe expect eventually to remove these from the distribution +Please post the changes in unified diff format against trunk so I can review them +It is a common cause for corruption and should not be done +Would this really stop execution of all queries that can be executed in derbyo no tests written for this mode with a TODO in the only test file for the feature +Convert the test to be more of a unit test of +Is this possible to reproduce in a testcase you can add here? should never be null AFAIK so Id rather get to the root cause of the problem rather than covering it up +Will wait on hadoopqa +Heavily dependent on +The driver to use is now explicitly provided. +Looking at Garys comment from Aug can priority of this JIRA be lowered now that we have htrace in trunk +Marking closed. +Were not seeing this right now though I suspect the issue still exists somewhereWe determined that the problem was caused because Cassandra corrupted its data files in some way which caused it to leak when calling getslice +If I try the latter I get ij connect jdbcderbyfirstdbdroptrueERROR XBMI Directory firstdb cannot be removedThat seems like a useful safeguard +Mahout does not use the Metadata feature anywhere +trunk +looks fine to me. +Hemanth we rm each file identified by svn stat before we call svn update +try addressing inner classes using interactive mode it seems to workWhen running a script the preprocesser catches it and tries to do parameter substitution +Just my fault +Can we close this +Adding scopeprotected to the Javadoc modules is a good I dont think we should add the allow options +What happens to those newly submitted Work units that the restarted DMLC sends to the? Why are they and hence their message consumers not executedHow can I helpWould it be useful for you if I develop a small headless web application consuming JMS messages to showcase the problem? since the problem since to occur only under specific setup and lifecycle conditions I would rather suggest to report this to BEA and see what their analysis is +Thanks Arpit for reviewing the patchI have committed this +If I have some spare time Ill give it a try +reopen for proper relevant so rejected. +Another deficiency is absence of support for faultReason locales which also would get resolved if was used. +Brian Nick I am still facing problem +And it was filed for master +If someone is able to find such a mechanism we can reopen and resolve this +We actually have a need for exactly feature on our end in that pages need to be able to indicate which layout to use or delegate the layout rendering without knowing a head of time which layout will be used or even allowing swapping out the implementation to a different layout +The related issues have been fixed. +Heres a rather old and POC implementation I once more investigation +in your file for the user starting the server add falseand restart the server and that should get you going +Im pretty sure this JIRA i sout of date now and these problems have already been fixed +As I said in the last comment +Remember the philosophy that Forge is supposed to support round tripping and is based on the idea of project comprehension without making assumptions +I think there is a similarrelated ticket about this reported as well +I did a complete rebuild of seam and the issue disappeared +Crap ignore that first line that changes the include from the +but I have a workaround and dont expect the world to conform to my wishes +This patch is a straight port from the security branch to trunk for the corresponding commit in the security branch +I dont know but its fixed in current configuration file isnt the issue are two problems the wagons were not being meaning the first used remote proxy stuck forever the being used did not reset the system properties after useThe first is fixed +Please let me knowImplementationI have kept it independent of the ecommerce moduleA new module has been created vatecommerce +Doh youre right +Is there a way to set up the regular with class to call up the numeric soft key pad +Reopen this issue until is fixed +Im pretty sure the is not picked up so a better test is required +Keeping default value for +You have every right to choose to wait for JTA standard support for nested transactions if you wish +Hi Hansshow the stacktrace generated in debug mode from guysI think the best option is remove all RCI categories related stuff as it has been removed from SonarI will commit changes and release +Example HTTP IPC server and client for ruby + for the patch +This way you can use a different set of regex rules during fetchupdate and another during indexing +Verified that the exception is handled correctly and neither a nor an is not thrown +I mostly submitted this to document +Hi AndyIve tried a couple of days ago and it wasnt working for me +nothing critical enough for +This behavior wasnt changed by so it has been there for a long time +I think its better we commit it to also before Lars could take the Rc +Issue with HDFS writers is more complecated thant it looks like since everybody can see the same file at any given time +to Darrin +Please let me know if you will be willing to defer it for later +I committed the patch to trunk and +thanx Stephen! +dhruba do you plan to run this on your warehouse cluster or just scribe tiers? If so it would be very interesting to find out whether it affects throughput +Thanks for the patch Florentin +The test is a new version of the +Implemented in +Borrowed code from +rename to warn if there is more than one kerberos principal associated with a ticket cache and patch look good to mehdfs is dead code is it used later in the tests +I also had to adjust the security properties in and +Querying list of apps based on etc +Probably indicates an issues with the script +I regenerated my classes with the Tuscany jars and see the same exception +I managed to patch it myself a while back using the as a guideline +Ming Ma Described functionality is essential to make or to process multiple regions from the same tableWhile trying to merge datasets together you better be aware what structures you are processing +will commit if the tests +The namenode has two limits the soft limit of minute and the hard limit of hour +This change is fixing what you are specifically objecting to +Abe I have checked in your fix +It comes with crimson not xerces so the JVM option should be this helps someoneGareth. +So I have excluded it from the for the test failureThat is exactly the kind of tool I was looking for +agreed easy label at the Project Explorer in JBDS +Reverting version until we agree +Or maybe a combination of bothSorry to say that we thought the problem was solved +to trunk as bcaeddcdeffbf +Slipping until is fixed +Why not make the debs consistent with the rpms in this respect? Eg should cover debs as well right +I did try with the example layers you attached when I said it was working for me +defer all issues to +Also please note that is about ! So the code above is effectively tries to print out MB of data +This looks like a recent addition +The same problem may apply elsewhere in IO +Im not comfortable changing the move code in a stable release but this sounds like a good change to make in +Nice +I think that what Bernard is trying to request and what you are offering as a solution is both valid but does not address exactly the same problem +Will close this issue after feedback +This issue was fixed however results in a new one see. +Im optimistically marking this fixed since it mirrors other issues that have been fixed and Im not sure it was retested against current ARJDBC and trunk. +As it is now Im not sure anyone has the same picture and were just going in volunteering +support the approach +Ramnivas JuergenWhats the best way to get hold of the nightly builds +some documentation +thanks for raising the issue It was fixed yesterday fix will be available in too +Any reason to limit this to upgrade? Why not just have the capability to change territory andor collation at any timeSeems to be independent of upgrade +it looks like actually loops over each entry and calls put one at a time which is the scenario for binary tree rebalancing quadratic time +Excellent guide +jar file produced by +on latest patch +Some tests were catching only FNFE I fixed those like replicatorI also made randomly choose one or the other exception if a missing file is to be emulated + +I think it wouldbe reasonable to mark this issue resolved and file Dagssuggestions for further improvements as a separate issue +I will take this on and document the new parameters that could be configured for noticed that this was already updated in I will just add a few comments to what is already thereIn the patch I have also added some configuration settings that were absent for the sample agent using MikeIt makes sense and will be less confusing to the end usersI will update the patch and the new version based on your recommendations +Thanks for the explanationI realized that is the case after putting in my comment +Select the Create tools with c shortcut +Moving these out to which probably means they wont go into +Instead if we had a list of required codecs we could abort the regionserver during startup + +Patch to allow an empty password +The comment is added in the the format of the Jarek +Not everything is implemented but I have been able to preparepreform releases from Continuum and the Release Plugin +I take a look later today +No I guess I found that one in a spring forum when I was playing with SpringI guess they got real busy with their platform and the bundle so the real repo to use it the bundlesSo I will close this one +Im not sure what John did to fix this in the other issue but it looks to me like a Hudson issue per the commented issue on the other ticket + now delegates to issues resolved in released versions +I think if you start parameterising goals to death they become unmaintainable and unusable +and sure enough removing test makes it fast again +I certainly agree some comments would clear it up but I think its a useful example +harportdirfileinhar is still opaque +closed with release +Im not sure if its due to SVN or Eclipse or the +diff file for the other changed files +The errors on checkout are because we dont commit the dir but yet is is expected on the Java build path +If all information is included youll be on the lucky side but if not thats bad luckIn the given case you are able to provide a suitable mapping because you as a human can simply compare the given pdf and the extracted text +branch didnt need it. +Total column contains +Yonikin the original version the maps the boolean values compressed tokenize binary store store and store to the Parameters and +Please review the submitted patch +Would be good to give it a try on a snapshot release +Check out the new test case test in which tests exactly the case I discovered +desktop file name +If it doesnt please let us know and well investigate further +Its working as you expected to +and create a directory and write to it caused the permission denied error +v addressed Teds +Well have a look at integrating it for the release. +Verified on +Fixed in CVS. +Close. + is intended to solve this +Just to be careful I need two clarifications +how long trunk should be frozen so that this renaming can be completed without many rebasingsIf we have clear plan I am fine with reverting to single namespace for now +Patch for for example application which demonstrants usage of patched in CVS thanks for the patch. +I have tried listing and not list the types in the persistence unit with the same result +We still have some problems with dependent CDI projects in case of excluding java classes from +Committed revision +The design is that all search operations that can be feasablyimplemented remotely should be performable on any Searchable and other searchoperations that can only be implemented locally are Searcher methods +Fixed in svn revision thanksdims +Looking forward for your patch +I guess this is a correct behaviour for but the should go on and simply omit the empty attribute +Kathey i will take care of NPE as part of +Yes it will thats why it was fixed +There the message will only be logged when a binary is actually added to the data storeComitted in revision because the log messages are only written on the cluster node where the changes are written +bq +Everything else like updates and closures should be filtered. +Im fine NOT doing this for the reasons listed above +This was updated in the ESB docs for verify in the SOA Platform docs +bulk move of all unresolved issues from to +Jimmy found an unnecessary synchronized keyword in the code v removes it +Ive committed this to trunk and. +May be a good solution would be to covert the distributed upgrade into local upgradeIt will solve both of the problems above plus eliminate the warning message reported inThe only disadvantage of this approach I can see is that will take rather long time to startup around minutes each on a large cluster +Appears to have been fixed some time in the past +Thanks Fabien and ChrisI commited a refactored patch from Fabien and Chriss patch in trunk revision +Thanks I dont apply this patch then I get exceptions and Messaging wont work correctly +Ive committed this +passed on my box +The container attempts to commit the transactionwhen the business method has completed +Hithe Windows installer should register as the executable for +Added fileDir to the list of implicit fields listed on the Wiki +Tried applying against current trunk r and got the following patching file libdelphisrcpatching file libdelphitestpatching file libdelphitestHunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at +I tested it out on this end is there a corresponding fork of I should be looking at as well +Before the patch in this issue you are right that the number of threads increases as more and more users submitting jobs to since the threads live forever even if some users stop submitting jobs + +This should be a much more straightforward fix for a problem that doesnt relate to any licensing issues for cppunit for example +If youd like once you have the hook in place I could take a stab at doing the jacksonjsr one +Resolved by Noel +Resolved in rev +Should allow each method to throw handle them leave it up to subclasses to catch them in each individual method +In CVS so I assume itll make +Thanks Ron +Failing tests in this suite are test could not activate failed to restore message received can you comment on this behavior + +Ive verified mvn clean install +The Filesets section in the Server Manager guide has been updated. +I think we just can use the qtime from the first query and subtract that from the specified timeAllowed +HelloWould you please try the patch +looks pretty good to me +Can you please record the status here? Thanks +The reviewer has become the reviewedI agree with Rohini that we need to be vigilant about this get situation +LGTM and thanks for the fixes incorporated +Hopefully is part of is the correct link type +Version RC is also affected by this issue +Class has been deleted in trunk +Done + +problem is the local timezone definition of the a lot for the input +In revision I added support for package requirement search +Most people can use it out of the box as is +New patch adds a method to instead of duplicating +Could you please try out the latest SVN snapshot and let me know the resultsOleg +Dierk Knig is working on the with the unit tests and also functionally against thelatest grails and Reporting was all a lot to Marc for the great work on the patchthat again cuts down on complexity and LOC! +This patch undoes that part +svn moved. +You have raised this NPE issue as critical so it must be reproducible in some way but I want to understand whether there is some other root problem +Before I dive in too deep here can others comment +The converted test changes the default locale and doesnt reset it when its done which leads to many failures in subsequent tests +Will be available in next snapshot build in about an hour +talk about poor naming +This is also the reason decided to split into separate interfaces like etc +It doesnt make sense to install it again if it is in AS already +We just put a date on our logs +OK then I am a assuming that you also are going to come up with a patch to put them in as part of their build +So we only ever use IP Addresses and never hostnames +Then there is no point at all in providing read only beans and methods! Now I will follow your advice and transfer this whole thing to the forum +Keeping the window thats wide open today open a bit longer doesnt significantly compromise anything +patch against just tried the unit test and it seems this now works +The split between emf and runtime was implemented by an additional project rather than a division at the java package level +Maybe this is just not a popular fix +Made a pull request on +Please tell me if you have a different result +I was just trying to arrange them like the mutateRowmutateRows functions are arrangedAre you also suggesting that I should the gernerated code? Or do you have a pointer to the exact version of thrift that was used to generate the current filesIndentation in should be like the rest of the code in that file except for the tabs +So you have to figure out the right strategies for a bunch of different edge cases for each datastore for each possible way to configure primary keys for each of the WFS insert options +push out to +This looks like in the gt code the filter was looking like bbox and include +with split info +Please could you verify that the patch has been applied to your satisfaction +This is not just to guarantee sourcebinary correspondence the artifact currently in the repo was built with Java +A fully integrated layer is now available . +Provide a test case for this look at comments in +ThanksTransfer request initiated +Committed to trunk. +See instance hasnt been installed yet +Thankssuersh +So for this pass i am doing the easy way to get the obvious onesThe main thing that i want to do for this release is the first sentence of the issue Description +Please apply this patch also for and branch. +Too bad we cant use positionfixed in IE +New Jira issueThank you but I think there is some confusion here +JakeYes +Are you able to try on a different JRE level just to see if the error changesI dont suppose you could make the failing grails projects accessible to me could you? All the sample ones I have build fine +Any news on that bugIs it likely to be fixed withThanks for you help. +Attached server logs +Added a new testcase to test the scenario +We provide it for users who repair in order to avoid reappearing tombstones rather that to reduce entropy +The double OOM error is a hunch I will try to find more evidence of it +Is this a duplicate of? Its blocked on a patch to trunkThe behavior responsible for the test failures is an incompatible change that probably needs to be reverted anyway +Fixed in the SVN trunk and branchOleg +In performing an analysis of XML Data Binding frameworks we found that Castor was the only implementation at the time that met our needs in terms of its support for WC XML Schema and its ability to integrate with our Web infrastructure +The unit tests pass when run manually. +For the most part it seems to be working ok now. +What version of Continuum are you usingOne thing that comes to mind is the Build Fresh option which will make Continuum delete and out the project +mrepositoryorgglassfishgrizzlyUserstucu +Committed trivial patch to branch and trunk +for +This transaction is aimed for better performance with the same semanticsFor with has been doing the same thing for a long timeLets open another jira if this is neededFor +If you could please it makes it easier to manage +quick comments for now remove calls to and use logging insteadMetadataNutch MIMETYPEKEY duplicates the one in Metadatall change the legacy to logging +That makes sense and it did the trick +However I think it would be more robust in the as KILLWAIT seems like a state primed for hanging if we dont really need it +al let me know if you approve of this approach and I can commit it + + for pelase and resolve to both versions +Only preserve comments +added missing files +Does the server side continue to have a timeoutSeems like it should badly behaved clients could cause a server brownout by sending never reading the responses and so occupying all the handlers +When we built before this change did we build example code as part of the general build? Do you know SergeyWe were not building examples under parent examples dir +Assigned back to the author for computing stats in the future +Committed to trunk and +Wooo hooo! Thanks RobIll update and provide feedback asap +Connecting to a normal sshd rather than a should be easier +So giving the builds a bigger heap should solve these problems +Logged In YES useridI believe I have fixed this and all related problems in jboss +Thank you +You out there Joseph +so it looks like the complete was OK and the unexpected fault in the cxf interceptor looks strange + +This doesnt seem right as a long term solution since we should segregate the runtimes usage of from application code since it is a common case where the runtime uses a different versionimplementation than application code +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Long running stream operations are better not dying even if they have to wait while S fixes hardware shortagesfailuresNot sure if its a separate issue but failed writes arent very well right now +This is a patch that fixes the bug +Adam if you want to hold off on this for a week or so Ill take care of it and submit the patch to you via JiraMike good to me. +test open on ASM + line the debug message says this is instead of On line Im wondering if it would be faster to create one list of null records and use it as a template rather than creating a new null list one record at a time each time deserialize is calledline ff +Im a confused young manIt does however make sense to cover all four while Im at it +Patch Applied +It would be nice though if the archetype would work instead of having to create it and then import it +json and has no in files that dont fit the pattern and log with warningthrow exception with file information for files that should have been migrated +That is a good idea v implements that +Fixes NPE fixes problem caused by suggests coding pattern and may not conform to scout coding formatting guidelines +Simple patch +The patch is in CVS +Slava please add steps to verify to description +Still waiting for testnoextension +Jon can you please post the relevant mapping file fragments? What is and what properties does it declare that are protected? Itlooks like you are trying to use lazy loading on a relation correct +bq +Thanks Hairong. +Could someone check this +With the trunk code changing in an hourly basis it can be frustrating at times for contributors submitting patch +Would you object if I simply make all of these new classes? Then if we need we can reveal more later as needed +Removes the previous confusing example policy for network security +I renamed to though +The client has one shot to do an INITIATE before failureIf the client sends INITIATE and guesses wrong the server responds with NEGOTIATE +Thanks for fixing this! Maybe it would be a good idea to let the code reference the web site so we remember to update the web site description ifwhen we change to code going forward +In fact we had this conversation months for us to do this for you wed need a copy of the repository archive +Sorry about that +Ridiculous +Juergen +Because Wont Fix thus I close it. +Note that the branch example above is not valid anymore since branch are now supported as a standard attribute +we are going to need this feature ASAP as well +Just because most of the public methods in dont have javadoc is no excuse not to add javadoc to new methods there + +I find it quite unusual for a web application to write to a file anywhere outside the containers file spaceBut now I realize there is a note about read and write permissions on the installation page which I stubbornly kept overlooking +Fixed in rev +I have addressed all theissues pointed out in this patchI have used the handledExceptions method and have followed Jorgenssuggestion in making handle the exceptionsto the master controller without wrapping themI have however made handleExceptions carry no implementation in for now +Ok heres the updates for the vectors +Thanks +It has to be decided if this option is ever supported +I checked with modded with Weld yes its still present everything is the same as originallyBut OK I have seen some discussions on the list that lead me to believe solving this is not trivial and will only happen in the next version of Weld. +Give it a little nudge +bq +It does work on windows set JAVAHOME manually to a u home and add the JVM option to +I think the one we have right now was written long time back + +Attaching a new patch +I updated the manifold check method using the status command of this latest patch +Minor update to last patch +Ok got it +Simplified patch +Change Priority from the default to Minor +Ive been working on a couple cases with index growth and was wondering if the output might provide some insight +Fixed in nightly build and will also be included in +promise +bulk cleanup of Jira versioning +I tested the latest branch +Back to you +Can you wire the monitoring heartbeats to a JMX operationJMX mbean server is started before we start the persistence adapter so you can query the brokerView Mbean +Yes this makes more sense for indeed +Its been a known issue for many months just have not had a chance to look into it closely yet +It may be a problem in ESB Core +Ivan is this patch for the pluinsmejb directoryAlso does this patch need to be applied to branches +Why not but how it is related to this issue? +Thanks for reviewing MikeNo I didnt perf the perf task change +is usedexclusively internallyOnce starts there needs to bea Factory method +Sounds like a pleasant afternoons work +Its the anyway that contains all the resources so reusing a single wo reusing the saves nothing +I dont like dealing with the type attribute of the SAX implementation since it forces us to those I tried to reproduce the error but I couldn you please post the exact code you are using to demonstrate that problem the parser name version you are bug should probably be closed false alarm on my part methinks +In my opinion it makes no sense to add yet another documentation of a Visual Studio bug since this issue is not limited to the usage of lognet in development using VS +This seems to be happening intermittently when the CLI does not contain the servers certificate in the local trust store +Thanks for putting that in +Assign to Svante for review +It was only used by subclasses +It is now correctly looking for imports before trying to apply the Immutable annotation to the source +But please confirmprove me wrong +Havent had to do much patch based development in some time +this uses the new rpm setup to fix this +I have changed the resolution to not a problem +fixed as side effect of by tcunningham +this has nothing to do with the maven ear plugin I would say +Patch combining the previous patches and fixing tests +It is presumably an issue with the plugins and not Roller. +Term lookup misses can be alleviated by a simple Bloom FilterNo caching misses required helps both PK and queries +My guess is that the reason why the submitted value is used is so if there is a validation error the incorrect value can be rendered without changing the model +The test that fails is not a new test its test +bulk close of resolved issues as part of JIRA housekeeping for release. +Kabir is doing stuff on managing the remoting subsystem so hed be in the best position to know +Assignee Please remove the netstat output +bulk defer of open issues to +Already included in documentation +I didnt have any problems at all with Maven when I installed it +test will commit this patch in two days if there is no objection. +so its really not used at all todayIn my opinion this tokenizer just causes us to scratch our heads all the time and thats it no benefits just wasting of committers timeI am happy to also only remove it in trunk i dont care +When NM dies the on it also would have died +run +Any chance that this functionality will be released soon? Maybe in an alpha version of the plugin +The patch looks good to me +After that we will only make minor changes till we move to Avro +Create new Work Item Definition Select icon to add URL entered was httplocalhostrestpackagesmortgagesassetsguvnorbinaryIf the steps to replicate are different to these above please advise +Patch applied. +committed to trunk +Do you think you could provide a testcase for itAlso +Thus even if we use the fielduserprincipal attribute name instead of the hardcoded uid this will work with previous ldap configurations that were working in the pastI will provide a patch if you want me to +I think it is sufficient to just remove HADOOPCLIENTOPTS from HADOOPOPTS to make it workThat is what I am doing in the patch. +Is this issue fixed now +I am going to proposethe easiest ones first and then check what remains +Well theres a tiny shred of code that might imply this but the rest of the implementation is just missing + +RegardsAshish I think we can now shut the coffin on this oneI applied this patch in SVN rev with just a few small changes to make the optional +Committed to TRUNK +Closing issue Oozie is released +Starting the process or completing the task without variables and posting form properties through two different REST provided wont developer needs to know the association between form property and engine variable +This issue has been fixed for all characters that are illegal in a Java identifier not just dashes +Do you mean that when using this kind of errors are ok and when I move to production thell be automagically goneI think this should work on Dummy too +The discovery plugins dont require any Java source +This is because you are not a member of this project +Is it ok here as comment or is there a button for doing this +I think disagree Im afraid +I did some tests +Thanks for looking at it +Well that I was afraid of +Moving out of +JohnI dont see how Dynamic Proxy will be better then current approach +Is there any chance that you could update the patch so that it applies cleanly to trunk +dims +could explain what is happening to your file and how it is wrong +Marking as closed. +Hey BrianYour patch looks great + just committed this +rev +It is simple to use and Apache licensed +For other requirements please create a new issue. +No worries Andrew +applied patch in r +thanks +Call outs should be placed it does now partially support them not very pretty but ok +Thanks for keeping up on this +Thanks Akira and Andrew +Well done! +The was applied in Qpid trunk at We could now add a set of new properties that follows the naming convention agreed on the list while providing support for old property namesNext step is to dig the code for all the properties and populate the correct tables in a bit of work has been done on this JIRA but a bit more needs to be done to achieve the above goalTherefore moving this to the next release +Can you try with this snapshot and if it all goes well I will tag another reference is still up after undeployment of the. +The fix for is the same but that line is now in +the change +In this case I thought this should be okay +Removed from exclude list with commit r. +copy lists as well +Happening to me too on Vista +Needs more tests coming soon +Is this still openWorks for me with groovy from trunk +Sync up after checked in and fix compile issues +classpath and +Uwe i totally disagree Im sorryInstead all these stupid ctors and should be removed and this class should be simplified to be a simple Ref APIThen should be removed from the class because its totally broken to call something internal but pass it around in every API +This turned out to be easy +I might need some help to test these modules +going to resolve this one since hbck has been substantially improved since it was filed +That was a problem with the preferences page +Im attaching diff files for the changes I made to CXF and dosgi to make them compatible with dmServer +They are often abusing strings or other objects to pass on the calculation of the hash +I forgot to say that problems with the ticket also occurred in version +Since I have written the code anyway and it might be generally useful for spi connector implementations I provided it hereFor example usage see the test cases +This should be wrapped up early in the M cycle +Done +I want to know when any object that implements the Lifecycle interface is initializedCheers +Hey Nicholas Ive been running trunk tests the last few weeks without seeing this +Id love to get this in as having in the central maven repo is really going to help our users outGiraph folks great that youre interested in using! Please let us know when you run into issues and well take a look +Are there incremental changes that need documenting? Or is the doc bug simply out of sync +Doing this will simply make it easier to use with JDK logging +Screenshot showing first iteration of the console plugin UI component +Ok adding notes here for what I ran intoThe problem with adding the methods to the proxy modules is fairly simple they interfere with dispatch +Lars There are new additions to the shell isenabled and isdisabled +A possible workaround for that part of the problem is to create an index that contains all the columns referenced in the query so that it doesnt need to scan the heap +If it fails there as well I am not sure what we can do about this problem +Thanks to Ingo Hofmann for provide this ones + update eclipse classpath +it is still possible we should update the to do it you need to edit the layout of the you try +Thanks for raising the issue Karl. +My preference isto include it with a test case! with the existing behavior as the default +fix in Note not required as this only affects internal Red Hat testing +Oops there was a line that got caught up in that patch that shouldnt have been removed +This will permanently hold filesystem objects but no longer creates a FS per operation which caused the original issue +Chris LaprunI have reproduced this issuePlease dont take care my previous comment +I changed my app to use for android and kept input file on happens to me too on Cordova +Needs short descriptions on each example +since I am not familiar with code submit but this small changes can enhance the performance a lot hope to see this new feature thanks +I believe the code in which prefers the context over the current is the correct behavior +Those are good is a patch that implements lock expiration based on the algorithm above modified by Randalls comments +Tese things can often be discussed by more than a single personNot ramrodded by someone being a bit of an asshole +Writing all the tests with special characters in file names is a bit tedious and redudant maybe the current tests simply should use special characters themselves +Run tests to VPE to long operation +What is the patch you are trying to apply? I assume it is the mega security oneIf a patch is generated relative to repo root then will not be able to apply while running from common hdfs or mapreduce +I will do another sanity check with some more elaborate tests before applying the patch +Should be fixed in latest SVNthanks RobertI was able to code generate adn run the client succesfully against a more or less identical version of your code +Extract over remotinghelloworld example +see also +It also makes aware and the component now propagates headers like the +A quickstart demo RemyThanks for your responseCan you give me an example of properly configured application? Just snipped is didnt get what you mean with The request has to map to an event Servlet to go into event mode +Attaching the compatible patch +patch for including unit tests and upgraded for the patch +added to livesites thanks +In SVNAdd comment change constructor and applyThanks +Thanks for lookig at it DavidNo we havent grown that much +The issue is with the order of your arguments +Gunther Thanks for providing a patch! I verified that it fixes the problemIf you add and to ivy I will and commit this patchbq +And unit test in revision . +If you get some clue please let me know +Did not even think about it earlier +Its really great work + +Addressed all commentsCreated as there are some scheduler config constants that have different prefixes +Aha found it +Andrew m using the assembly descriptor +In trunk the flow of execution does not go through but rather through +bulk close of all resolved issues. +Decreasing priority to major because remaining change causes just warnings to log +lgtm +Wouldnt be a good idea in the case of Seam adding conversationId to this key +Any comments +It might fix your problem but it changes the behaviour of the flag +Please verify and close if satisfied + +Triggering the hudson build to see if the tests passes +Carlos any update on this? I havent been able to find where is the problem by looking at the source code and dont have a Windows machine handy +The spec states at the end of that the preference is for resolved exporters over unresolved ones +Great issue stackAlso we need to consider how to best add additional threading to +I believe that the improved connection shutdown code in the current trunk should resolve this issue +Note that all unit tests for me except for which fails fairly consistently with or without this change +I updated the patch by incorporating the above comments +One more test we should do is to check that the FQCN is equal to the instance of +I will start working on it +Hes also posted a patch over on that JIRA +thanks for the update! +Assigning back +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I guess though in solving the issue we will see if this issue is also solved. +I started a discussion on this bug in the dev user list as I do not feel the fix is appropriate for the problem and this problem should be addressed for all tomahawk and trinidad components that may set local component values from other components or renderers. +In my current configuration +Nothing has been discussed +An additional note to the first patch I removed from the Manifest files if the Castor so anyone who dont wants to use the will have to add manually to his classpath in order to use the Xerces parser +It was released with the fix in +Wherever it gets fixed I just wanted to record the bit that still does not look correct +deferred +Pull request cannot be merged till tck updated and a new release published +Essentially the dateFormat attribute is not honored at all +The product store is currently fthe focus for products like pricing where the product is located +Thanks +The out handler that checks the policy and the mtom policy persistence +could be classified as a feature but there are still big cleanups in the PTF work +I am closing this issue because can address this issue +Updates readme files +issues to the latest version. +trunkCommitted revision +Avish Im closing this issue if you have any comments type them here and well discuss them SergeyOK from my side to close the case +seems this is related is it right +If Glenns indexing fails again Ill try seek through that file to figure out whats going on +Ill see what I can do with this +Manually tested! +I also the metrics tests to verify the is still reporting JVM metricsNot sure why Jenkins didnt catch this failure back on +Heres the current screen shot in case anyones +does not sound right that if the log event is added to the manager not matter what is isFiltered returns +So I would be on this as it is generally useful in speeding up the build a bit +FranckLook at where I posted a patch. +is duplicate of +The changes look good will commit if the tests +I reverted the previous patch on and +Since Jetty still doesnt appear to be stable enough weve decided to investigate Grizzly for the same +attaching a project that failed to reproduce the to reproduce but could not +Author certainly is optional in the spec and throwing Null Pointer isnt correct +You will have to wait for the release +But the is possibly corruptedCraig +So we could check for the return type but dont have to +Would be good to have preferences to let it be controlled +There is some interesting stuff in which extendsRather simply you can just insert a bunch of sequential rows and run manual splits to create multiple regions +bulk close of all resolved issues. +It works fine now! I am closing the bug +uploaded svn stat result by mistake +Although its hard to assure there were no other side effects I was able to submit jobs +Alexey thank you very much the patch is fineAll the blocked issues are now resolvedClosing. +Please create a new issue with steps to reproduce +is about the deprecation of line already tests that default is consulted +Same with labelpositionBut labelposition ls marked as deprecated in andShouldnt we just remove the deprecated attribute and just use labelPosition insteadWhat is the preferred attribute name for max length? maxLength or maxlengthComments +Added typeName equals toString hashCode functionalityAugmented existing test case to include code from this bug for the close of resolved issues. +I was having a similar problem at work on my windows box but it looks like we have a solution +Sorry I should clarify when I say possible I mean that I will be able to implement it not that it does it now. +Moreover its a runtime one which changes after maven start so no static value will work maven interpolates defaults user global maven ve added settings and session to the execute mojos default namespace. +Thanks JessePatch applied to module at repo revision rPlease check it was applied as you expected +the settings for both dataDirectory and brokerName get ignored regardless of whether I configure them on the broker or on the persistanceAdapter +Thanks Zheng + +Patch appears to be working so I applied it. +Im closing this +hey want to break the benchmark out into a separate jira issue for simplicityits independent of the improvements to shingles and we can use this for other purposes like realistic benchmarking of saverestoreState across analyzers with varying amounts of boolean Attributes +Patch looks goodAlso tested with both trunk and earlier version and passing single and multiple values in the HADOOPOPTS through the envs optionAlso tested that invalid key value pairs now fail with a proper error +Verified on JBT. +We cannot have a required dependency on LGPL +fedbbafaaf +The Accept header is actually set by the JDK so as a workaround you can inject the into the template instead of the default +Marking as critical +Firefox Fedora in jira +I just browsed quickly the WSSJ codebase I may have missed other places where is used +Use to enable this release +This is the updated version +What is the marker +When running from a portlet this happens in the processAction method of the which creates an with the Portals Response object +I just moved felixs svn to tlp +Ill commit it asap +Also the special group may now contain other members like admin users besides TT unlike before +Verified very thanks Tony. +Anyway try it out. +but are you planning to use the right object or simply find the right action and copy it to the? The former would be a lot better as it would bring all the other values associated with the correct as well +so in reality we arent talking about a TB log storage requirement +Committed to all branches +Need to backport to the branch when its ready +I di this via grep so an extra space here or there might mean I missed one or weve got one contributor still to contact +This issue relates to +And well that was many years ago about years ago or so +If this looks fine this would be final copy which will be checked into trunk +Test cases are not required because it is a trivial change +will commit after tests thanks Xiao Li! +Thanks Ivan I will have a try anyway +Yes they can possible be removed just make sure the gets closed as otherwise it can lock file handles on Windows. +Fixed at revision related some of the tests in are specific but a lot of them can be generalized to run against any MK +Thanks for the feedback +I fully agree with all the points I just dont think it should be the default to only log the exception +Bulk defer all frontend issues to +Will piecemeal it +This radically cuts down on traffic to namenode +WDTY? Regardless I dont think is a blocker to so Im pushing it out +More precisely we may someday add options to i that it would make sense to expose to the user but we dont have such options currently and letting the user change the ones currently used internally would be a bad idea so its probably simpler to just refuse options altogether for i for the time being +Max What are the chances we can squeeze this in to m? I think its pretty important +Im curious as to whether anyone has just thought of using the Clustering component for this? If your collapse field was a single token I wonder if you would get the results youre looking for +Right? Otherwise a fix might be has been closed so this issue has been resolved. +see subversion tab for revisionsThe view source htmls for are during build +Change checked in as part of at m going to take that back +Ben what you ask was not part of the original spec of this jira +I like it +According to this and the error stacktrace I think the problem is related to changes at said that I as committer frequently build Castor in the same way on different maschines +Im adding debugging to this issue where if debugQuerytrue ordebugShardstrue then the results will basically list thereturned response headers from each shard +Zebras the implementation of PIGs does support glob by design +Are you basically talking about why does not work now? I am unable to read your there is no error at all +it would be nice to have an option to put a custom caching strategy in place +a and fa +we still need to add error handler to so postExecutor does not get blockedIm not sure we want to unblock it if the flush errors out then we definitely dont want commitlog segments getting cleaned up +Feel free to reopen this issue Gareth if you know how to reproduce this issue +Aaron can you please commit this change if you are happy with the latest patch +Thats not very defensive +Looks like ill have to modifying the import wizard State to pass the creation of Services functionality to the Catalog +This feature will not be implemented +and committed +number consider it a timestamp and if a map treat it as attributesI did this +Fixing resolves this issue +its not ideal the proliferation of calls is not all that nice either +Are you sure you want those new lists named like that? Doesnt seem normal to mewhy notmerge into devand then into userhaving juddi is unneccessary imhoLet me know so I can continue with the move +As mentioned this is a platform configuration issue. +link to to keep track of clientembedded differences +new attribute on statements to expose the variability and acks attribute on posgres adapter in case generated name of pk constraint changes +I agreed grep is problematic +This has ended up being a duplicate of where the ASM framework has been employed to annotate classes at build time. +If a connection from a leader to a follower times out then the follower goes back to leader election and a new leader is supposed to arise in a few seconds +The bitmap files appear to be hosed as well +Change to tools folder and run mvn for building +A release with this fixed would be great it would make Commons CLI usable for my project +yup pleaseMerci +If you think it merits a separate issue then I can rework both patchesbq +Hi WernerI see this problem both with the original release and with the current CVS code +Attaching a patch for branch +works ok for now. +Closing on behalf of Andre. +Thats a rather interesting one +Bumping up to to havent run into any practical issues with the existing implementation and we have been carrying this issue along for several releases now +There were couple of formatting problems in the patch which I will fix and upload a new one soon +be on the server or not +Summarizing the comments over the last week no one has expressed a strong opinion either way regarding bundling the packaging and deployment functionality into a single jboss more I think about the issue it seems acceptable to have the two plugins as they represent different build phases and contain drastically different that seem a good rule in the case of multiple foo plugins consolidation should be considered for goals that are either attached to or typically run within the same build phaseAt this point Im just looking for a way to make some progress on this +Hmm maybe add an explicit test for the no terms provided caseI will add one before I commit +Yes but how can I achieve thatThe only thing I can change at is the two DNS entries +Why has this patch not been applied +Carol As before this is a weekend activity and I hope to get to this then +Jim is complaining about the listdetails paradigm that we are using all over the console +Updated Simplified Chinese translation files +Thanks Mamta for the fix +Best you join me on IRC Harry when you can from where Im sitting the apache config is untouchedr root root Feb default +not really a backward compatibility more a bug based databases that dont support it you can do order by colum DESC when you mean order by column ASCThis will reverse the order +The changes to shuffle dont apply cleanly on on any branch +you dont need to delete old patches already attached +MITRE has no copyright on this code Ive marked it all as ASF +Assigning to you for answer on the above Jozef +The code inquestion was added months days agoThere may be other formatting problems still however if that is the caseplease reopen the bug or create new ones as appropriate +paulPS this comment was first done on +Kate I think the patch was created with an older revision and does not seem to apply cleanly +Does anyone have any idea when this featurefix might see the light of day in a stableofficial DBCP release +Ive been able to reproduce it quite easily on Windows JDK sure you run it on something with package and compare the output with JDK details were originally posted to a little while back +Sorry I didnt understand behaviour of your application try to download latest richfaces this happen all the time or just with manual input? Have you tried with different localsWe are just trying to narrow down where the issue may be +we already have one such auth system using subnet authorization in srcjavamainorgapachezookeeperserverauth +This patch removes the ij EXPECT command +Wordcount should now be part of hadoop which is now part of the mavenized build +Committed to master. +I realized that the Hibernate mapping file generation would also need the prefixsuffix support so I added it in +to Dimitris to properly resolveassign this issue +Change execute phase from compile to +I now see that the build has jena in it +Committed trunk and +Havent seen it beforeThe eclipseeclipse issue is a restriction of the eclipse plugin +and thanks for being patient and trying these things Ive absolutely no clue why you are seeing this problem right now so it may take a while to home in on it +Reworked pacth to avoid in unit tests +Handle files and directories in a manner consistent with other commands instead of silently skipping files +So update testsuite poms and committed revision Now these suites has failures +When running this I saw the out of memory exception was during optimization +I added a throw +In the loop over the termEnums had been changed to reuse the and only they are not reused + +Juven I had some other complaints about slow sync times last night by the time I asked the user about it the sync had occurredWhat sync schedule are we on and what causes the sync failures +Thanks Omkar! +Lets discuss on the list +If you know of further work that needs to be done for this requirement please feel free to this issue specifying the necessary content additions +Still working on this +Do mind take a look at it? +This patch depends on a previous patch where Cale made plugin changes +By framework I meant downstream metrics related code in common hdfs and mapreduce as per discussion in. +ThomasI took a look at +Referenced in the settings with filetmprepo +TFFT failure corrected +Get this trivial fix in for fixed in rResource keys are and . +Moving out a that is not being worked on +Jack please reopen if this is still an issue +Feh +million bytessec ms million entriessec +Since this is an upstream issue and its inconsistent Im going to close this ticket as a Unresolved since we cant actually resolve this one +Contains changes to all tools projects that could create the directory but dontM toolsjavatowssrcmainjavaorgapachecxftoolsjavajsM toolsjavatowssrcmainjavaorgapachecxftoolsjavawsM toolscommonsrcmainjavaorgapachecxftoolscommonM toolswsdltofrontendjavascriptsrcmainjavaorgapachecxftoolswsdltojavascriptM toolswsdltocoresrcmainjavaorgapachecxftoolswsdltoM toolswsdltocoresrcmainjavaorgapachecxftoolswsdltoM toolswsdltomiscsrcmainjavaorgapachecxftoolsmiscM toolswsdltomiscsrcmainjavaorgapachecxftoolsmiscM toolswsdltomiscsrcmainjavaorgapachecxftoolsmiscM toolswsdltomiscsrcmainjavaorgapachecxftoolsmiscM toolswsdltomiscsrcmainjavaorgapachecxftoolsmisc +before install upper example bundle +Select the JDT Core Patch for Plugin +I tracked the problem down +The patch looks good to me +Patrick I believe the jenkins job runs on Linux +Since when should someone always expect a? This represents some error not usually recoverable by the application or at least expected to be handled by itThe API uses sure because were opening streams and thats necessary +this could be the mechanism for creating a new template for a specific view the sample application though we need to decouple the console into a separate war +Yup the test for case insensitive dictionary assumes that entries are merged +We are hitting this a lot in the recent Hudson tests +It only occurred when the table user existed prior to teardown of the first security walk and then an operation against the table user occurred on a subsequent security walk +As noted in previous comment all of the relevant attached patches seem to have been committed +What isinteresting me is the root exception of the +This is a duplicate of +Since it already has ability to limit the number of files that each reader can access it might be better to make it work on all N files by default +export LANG not only before building also before doing the svn checkout cause subversion also uses the platforms default encoding I think +I hope it is thanks! One last note You should try to avoid reformatting of code that you otherwise did not modify +You are trading off your desire to not introduce a dependency with the consumers of the library being forced to use as part of their logging system that they must monitor +Checking now if this is still an issue on Amazon +So that method should just be removed noAlso youve added braces to all of the log statements +BTW Author tag should be removed +Im pretty sure this is not the same should be fixed now see +Hi Frank +Any use case for this? +A quick look makes it seem likely that this is coming from the WAL file handling +thus can be problematic thing even with the attached WARN I was able to access actions defined in a jar outside war actions are mapped Ive added screenshot from my browser +When using the I cant reproduce the problem +Ive made a minor adjustment to link to a video which is more of a Hadoop tutorial to address Chriss comment +Yes Ill try to submit a patch in the next week or two +Hello Kokov and sorry if I closed the issue too quickly +The attack seems to depend on the DBAs ethics and her judgment about delegating responsibility. +Swimlanes visualization from the Chukwa project that visualizes job history data +Fixed to attempt to find the right destination with spaces padding trimmed before falling back to the configured fallback handler +Updating the copyright year +Do +Besides it would require JMX remote access to be available to this node +I suppose my point is that we need to define the relevant properties for the project +Transitioning to closed. +I also have a patch for this if it is deemed appropriate +Join is still fails intermittently but thats expected and not scope of this fix +after setting new Affects Testing Regression Blocks Testing Fields +Committed revision . +Please reopen or create new bug if you continue to have problems. +Youve already asked Rob Stryker and Andre to chime in so Ill leave it to them +At least not if the interface implementing class is also implementing the method +In addition of the documentation we can add a shellvariables command to display all known variables session variables system variablesWDYT +Seems like I dont have admin access in Hudson itself +Okay? +I tried on an mdo Im working on and found the following unused importsimport import import import import import import import As far as I can see none of these Classes are actually used in the generated +Is it the right behavior to keep the file directory entry even if the creator calls abandon +OK now that test related failures have been nailed Im back to looking at thisbq +Looks like the icons were fixed previously please verify in the latest build +Applied time please make sure that you follow our codestyle and conventions. +IMHO this should be a major bug the way it currently fails +Odi If you know how to simulate NTLM server using an compatible java libraryIll be happy to learn exactly Oleg I rather thought of a replay attack Capture the protocoltraffic of a successful authentication session and replay the server side withour test server class +Possibly but wasnt sure abt what our plans for were +Personally Ive given up attempting to use version ranges in Maven due to the numerous bugs surrounding them +I think the comments in and the reply in suns bug report explain this The property is not required by the JSE platform specification its an internal detail of Suns implementations and should not be examined or modified by user code +This one has it +This would still allow the content type to be set explicitly correct? I post JSON in some of my gadgets +I tried what you suggested on and it seems to work except that it wanted instead of commas as a delimiter since the commas seem to upset the CQL it looks like we dont have a bug on our hands +This will enable the installer to configure other runtimes included in the Platform such as Seam or Drools +Rob or Brian will have a look +Thanks Mike. +The bug was not reproducable on other boxes. +the issue for further attachments +When it does it updates the list of servers with a new shuffled list and then uses new logic to choose the first server to connect to from this list +Thanks for the opinion its not a general problem of webapp but a missing solution in CIConsider Jira itself if you click on a link from an and have to login you are transfered to the requested page after a successful login +If we remove the avro implementation then this is moot +And the tests passed +running is getting worked upon at through an alternative approach. +attached patch should resolve test with attached back to Marios to commit the attached patch to git master branch +Vamsi thanks! Could you please add SUID to Holder class in and retestCaused by lder local class incompatible stream classdesc serial local class serial at +Need to verify with Sybase whether there is a known issue andor a new driver. +Yes I have explicitly stated that it works with +obvious typo +When it does this some event completion updates are lost and the reducer just hangs +As of revision can create a typical database deployment of revision used by the console to create database deployment plans +Thanks. +Filtered out client components for Status command generation +the only problem seems to be aggregate goal which is not capable of retrieving the configuration from the reactor projects. +There is an additional build environment field in the release prepare page +thanks much lots of work needs to be done on this plugin before Ill make another release +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Prepare another patch to raise an error when compiling the additional query families you have discoveredThanks table functions should those queries be rejected according to the standard +SELECT queries ran then ran separate UPDATE queries updating each of the different column types +push out to +I wonder if that wont complicate IW code in return? Perhaps well gain a lot of simplification on DW so a bit of complexity on IW will be okThat will get rid of all that insanity for each DW component if Im getting it right +Reassign to Karel for keeping track of release in release notes. +Victor I committed the fix to trunk but please review this code and make sure we have junit tests for this case +Thanks for the benchmark and the reviews +in bug reports +When web monitoring is launched is there any specific metrics required to display? At the moment Hama metrics system similar to Hadoops is under construction +Ill check in with Mike to see if the problem persists with him +Id consider that being a bug too +Just in case this helps others +If you want to do more play with the docs ant target and add transform that will dump the product into the right place in builddocs +Because weve said that we can never overwrite the files that exist to get proper layering this means that any files in B wont be recopied again to the target because they already exist +When adding entrys to corrupt it might be good to get a object and use it as a key +Closing old issues +For java now forks the two ls forks in parallel which should help some on the observed slowdown +I dont think we should add the jar in lib or liboptional +I will attach the complete module when I for your attention Antonio +Is this what youre looking for +Resources have been updated but we have an issue around etc file not copiedI fix that +Id like to close this issue by documenting that requires SP in order to work properly. +Thanks Owen! +Cheers +bulk defer to +Ready to to in trunk and in branch. +Added RHN chapter back to Install Guide and updated RHEL and commands with information from this JIRA description +Additional comments should be attached to. +databinderfcst +We will need to enumerate the list of column families submitted for bulk loading and insure the requester has WRITE permission for all families in the request +Is there any reason why the base URI cant be corrected by the application so that it is a valid URI +I also discovered one of the unit tests was failing and the fix slowed down the new specific writer slightlyHere are the new results for batch size +Im now wondering if theres a way to go around this problem +Is there an overview of how this functionality behavesI couldnt see anything in this issue and JDBCJTA specs dont seem to specifiy how the timeout is used +What do people think? This is a fairly large source code reorg thats potentially disruptive +The person working on the issue for the jvm notes the following workaround +Where should I find the property team if I dont know the targetEntitytargetEntity is redundant in of collection generics not regular collections +Closing +attached document on how archives is implemented currently and what improvements need to be made +Still broken on branch removing marker because is dead +Ive added a junit test but do not manage to reproduce the problem +just copy the WAR under webapps and start the serverthen call httplocalhostmp in IE and click button modal panel +OmarHave you had a chance to look at this? Should be pretty simple to fix +this is at eclipse. +I can see helper methods that the shell or tools could use but dont see these being run on every tablename we are passedbq +bulk defer of unresolved bugs to +thanks Harsh I edited description and attached a patch +works in M but not in at revision. +Thanks Max! +Instead this should be an hbck feature that automatically patches holes in a table if this issue here is only to be able to share those scripts until we have this all covered in hbck +Any update for this patch? +Its not because you want to return the response as a string that you are creating it yourself +all issues to the CR +Please use english to communicateThis response is by understanding your query by Spring or lower release document says its resolved but it doesnt seems to be resolved +Well I guess for the newer endpoints this is wrong because they are asynchronous butBtw I think the total number of connections should be set very high in any case as this is more a system limitationAnd we need to move the thread pool configuration on a per endpoint basis instead of per component imhoSo what if we put both limits to a very high number in the default configuration? Is there any drawbacks +Looking good Todd +Now that work is well underway with AS all previous community releases are +Thanks TedIll try to provide a patch following your suggestion +I think requires more investigation to determine how effectively we can get it into or punt to +should have been +Closing +Cleans up some issues adds license headers +It saved me a lot of work just to be cancledI do appreciate you support. +I reverted the original patch +Ive tried to compile trunk with protobuf and succeeded to pass though I needed to run mvn clean command explicitly under directory +Ive asked for this multiple times +Also the parser in jruby came into being around ruby conf so it is possible things have changed in current parser as well +from the ejb packs selection to further trim it down +In the interim Ive created some custom builds but thats not ideal +is sent when SSVM connecting to MS it happens when MS is started you didnt see this issue before probably because struct is not changed before +This is not a broker issue cant be resolved at this level. +Good thing it was a dev cluster +We should make sure the old property names still work though for compatibility reasons +Does the system need a mechanism to addremove audit loggers? When a failed logger is fixed do we need a way to refresh the audit logger so it is picked up by the Namenode againPlease address this previous comment as well +Merged to as svn +Harsh This will really help in other systems also as you mention above +One thing Ive not included is the documentation on the RSS based monitoring introduced inI am working on a patch to include the RSS based monitoring changes into the documentation as well +Attached is my DEMO note that you need to configure a Database +BTW is there a way to revert it back? From the script I can see that you are just updating the status to U +What do you mean? Is there something we should change to make it easier +I tested a lot of scenarios locally to find out why the is not but I did not find a resultI locally removed all instances from the code and also fixed the problem with the mentioned by Stan but the still isntSuggestions are welcome +The is not relevant anymore I can regenerate one if you want but I suspect there will be a lot of rejects to take care of +Latest patch from RB +Thanks for the test patch +We just need in the test project a jar file with a +What is the intended lifecycle of these hooks? Created on call or created on RepositoryOakMK start? What if a class implements both interfacesBackground I could imagine an implementation where beforeCommit and afterCommit cooperate in that the final afterCommit might be interested in stuff done for beforeCommit +The configuration includes the velocity response writer +Proposed fix attached +Currently this just makes deleting a cache pool faster +Mass closing all issues that have been in resolved state for months or more without any feedback or update +from trunkextracted the standalone build at moduledistribution build the sample jaxwsdeployed it in axis and start the server +Wed only know with testing but Vasiliys comment that starting and committing the JPA transaction through an interceptor provides an adequate solution for now +Trunk does this already. +In our scenario we have got a lot of write operations and more than nodes +I will add a comment to this issue once thats available +Sorry I have thinking of things like convert which is just a stateless utility method +Hi AllIve attached a patch that will emit query plans to a +Ive applied your patch in rev Thanks +It reads the beginning of the file to read the sync marker and other info +For a new htable that is not passed a configuration can we not make a new instance internal to the htable constrictor as we currently do +Hi Jingthanks for reporting this issue +Although I stupidly had to pick a test which will not run in this particular example the patch does work +Translation of the javascript error windowA runtime error you want to start the debugging modeLine Error is Null or no object +Until then should work +Any objections to committing this +Can someone give me permissionThanks +no more NPE +It was running OK +Ill take ownership of this +Fixed in trunk rev +were using baseBinary on the Teiid side +I know that we have a switch to bind the EMF into JNDI and Im assuming that there is some way to do the same with a SF instance +The unit test still shows that both success callbacks get called +Use Axis or dont understanding your commentI was asking you to confirm which version that you where using when testingI was using as that is what I raised the ticket against +I didnt understand +Tag youre it +Youll get the downloaded data plus garbage data at the end of thebuffer since the read size from the input stream is never used tovalidate the amount of data to write to the output streamScottScott ObergSenior Software EngineerEngineering Systems DivSRI International +It could very well push out of our technology stack +Could you answer the question above? This is one of the last issues remaining before we release RC +Committed trunk revision revision +Thanks Andy +Fixed in HEAD and Branch +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +When you document it please assign back to Marek +Venkat could you explain why this is necessary as I dont think we should be adding any Tuscany extensions into the osoa namespaceJim +I did that myself earlier but then after changing the two instances it was failing later where a key was reported not present when it should have been +NP Ill close the case config helped +all quickstarts on both as and as after and Kevins changes everything now passes. +Fixed with revision +This is too to be useful. +Were going to get to it dont worry Bill +Cannot reproduce this +Also just wanted to let you know that this is the version of java I am using on Spade +Either the th oct build ran to early to get this or the nightly build for the th have somehow not been published way something needs fixing +Do you also want to address that +The latest looks big but fair amount of changes were trivial amendments due to the timeout requirement in unit tests +Yes this is what I wanted to suggest +Because having such a file in the every project metadate wont look a good solution at the moment +Just to clarify Im testing with r on the branch +If others feel its better removed Im ok with that +Verified on Build with System VM Template for. +The names or both have the problem that they only refer to the indexing side +This is odd though because Hibernates provider implementation should ignore such exceptions +In . +Thank you +Ok Lars +Thanks Nandana +why not use AS if you are doing this? I believe dmloyds bot uses that +Yes I referred to left join as just an example of what direction to goWhat you mention makes sense since left and right are really doing the exact same thing but just from another direciton +AlexLatest patch here doesnt have the last few fixups +I submitted the review request +Moving to CR for triage +Every time someone says oh you can use a separate config file I need to start thanking them +Patch that fixes the add affects version +Rolled back most changes as discussed at last ff +BrunoTo be clear the error only happened for me when specifying an updated since or before date +Finally we got a cut in Spring Framework master now based on Jeffs latest pull request but heavily refactored and restructuredThe public entry points are and now closely modelled after Spring s and its relationship withJuergen +This is actually a bug in JSF which Stan has raised with Mojarra +Under the current contract it is the cilents responsibility to ensure that objects are not returned to the pool twice in sequence without a borrow in between and that only objects borrowed from the pool can be returned or invalidated +It should solve your for the report! +The patch enables falling back to the default handler if the URL to parse is null +Sorry for trouble you may close this issue. +Hi Nathan The testcase should be excluded out on all platform with IBM VME +I have added correctness tests as part of the nightlies +Retry +The mailing list address is To subscribe send mail to It is already mentioned in the xdocsI think Noel J +which I believe is wrong it should just be equal to brandingToken +I dont know if that would be a part of this issue or that would use the API in this issue to make that happen. +This is a small but important bug +Id recommend moving to Eclipse +We have added this then removed it in favor of our own schema +Why did it work again once I recreated the keyspace and column family then? Did you at least try to repro it +Just in case render nothing means not render that attribute and not the whole component +the federationLink bool member of broker is only ever set and never queried or otherwise used so can be removed +Thanks Peter! Ive updated the notes to include Java Security Manager as a factor +Bulk close for release +PR wrong issue +I see this issue as invalid but if you still see it as a workflow error please suggest a better workflow +Found small changes that will remove an unneeded file from being added to the +I should have mentioned that we are using the release +I think the best path would be to create a OSgetcurrentimageimagerevisionid routine which parses the contents and returns the imagerevisionid on the loaded node +I think it would still be good to have it available to improve coverage when it is not built in to the jdkjre distribution +Do we really want to make this data publicly available +as per reporter +The remaining AMQP work is then essentially additive optional and therefore low risk +Once the job completes there is no tracking url since the MR goes awayThus we can use the same field +It doesnt make any sense to me that SSL would work everywhere else except with +As for the fix for this issue its planned for AS in of m getting the similar errors as given above in Jboss FinalJira says Fix Versions Can you please help me where from I download how to apply this fix to the server +I am about to merge to and all of that code is about to changeThanks! +GuysThis is Shital from Yahoo! We have been struggling with this issue of join not happening since jBPM release and see numerous complaints on forums regarding this without any resolution +trunk +Really agree I end up doing some similar hack all the time +I just did an svnadmin recover and it me know if that seems sufficiently fixed +Lets see what Jenkins says +seems like the issue is fixed +Actually it looks like the message was in for checking the region location cache +Im inclined to agree with the originator hereIf it helps some users and does not cause problems for others why not set a fixed EOL style +I have come across this issue for another customer +The persistance for Calendar type should be consistent +Tested on a local cluster to verify localization is not waiting after a download completesDont think this requires explicit unit tests functionality is being verified by existing ones +Or just email me. +I will look at it right the regenerated patch with lastest trunk image +Same issue of +fixed in Thanks Lars +I dont think you have solved the low coupling on +no +And coincidence +Again how would you inject the broadcaster when one isnt given? Or should I remove the null check and risk an NPE? same applies to risk NPE? this can be piped in a bit further +Following is my understanding about auto implementation +Manually tested +A lot of the lock management here is duplicated for every layer access +You could close this +I am not sure that the current implemented cache policy would definitively fix the issue IIUC in that way even if XSL is invoked with different parameters it returns the same contentIs there something I missTIA reasonable could you add more test cases to make sure it doesnt cache the same contentbest with different parameters always get back the same content from cache +To persist the you need to manually copy them into the writer +This would be useful even without since typically we want to write out both recordkeyvalue lengths first before the actual data ala +Thanks Wang! +In the case of a dispute PMC votes trump contributors but everyone is encouraged to vote and their opinions are taken seriously +I think we should close this one out as Wont fix +IMO the refactoring should be considered in two parts Refactor struts initialization code out of the servlet environment altogether so its not tied either to either a Servlet or +Attached file that applies to trunk +Bulk closing stale resolved issues +if youre working with svn and you want to make sure you have a fresh start use svn revert R and ant cleanOk but that doesnt solve the problem if youre not using a checkout +I can do this this week +I looked at them a little bit and found some lines that indicate disonnection issues inShould I send a copy them to youHow would I now up or downgrade red? revision . +I think Kojis point is should we have something like healthchecker in Datanode similar to Mapreduce? If so periodically Datanode launches this healthcheck to determine its health against disks nics etc +README file added to the for the patchintegrated in revision Boris please check taht it was applied as expected +This issue has probably existed a long time since the original refactoring work to create problem has been fixed. +Improved patch adding additional tests to TC to check whether mode is being dealt with in a correct way +Thanks Bryant +Wil see afterI have no IE +H code for regression testsuite based on Alexanders as MikhailVerified at r. +Attaching a first version for attached patch supports for both the old and new API +Hi have amended the Javadoc for the to be explicit about the view name determination in Spring and Spring and I have added a section to the breaking changes section of the Spring reference is indeed a breaking change please do accept my apologies for not making this explicit in the was no big issue to me and in the end it forced me to refactor some code that is now better than hope it may help others to avoid troubles migrating from to +I just committed this. +Yes this is a problem on the artifact plugin and there is an issue for it already +This seems to be the case even when you go to Settings ASUS customized setting Battery Saving and uncheck Disconnect network during sleep +However users need to realize the running the agent as a user different than the one the JBAS instance is running in may result in decreased functionality +Heres the updated patch +Just have to implement some policy to handle timeouted paged search +Is this still being worked on? +I know it is still a work in progress but I think it will be easier for you and me to work incrementally of course feel free to send more patches as they are readyI did a small change instead of using lowagie I have used the methodsAlso please move the class from the screen to the form package I didnt do this myself to avoid to you headaches because I know that you are still modifying the class +Minor comments Typos in the description of the deleteSegment API in +When you post the issue you dont always remember about it +This is an important issue because according to the WFS spec the committing the transaction should return the final fids +A local connection is just specific use case of remote connection +This issue will be fixed as part of defining the behavior of storage for the late reply +Hi MyrnaThanks for bringing up this topic +But occurs while shutting down the application in TC serverAlso is an unreleased version and we cant use it in production can submit a better unit test that can be added in we can make sure things get fixed up for the release + +Reproduced here with and the version of surefire +bulk defer of featurestaskswishes from to +The idea with the JSF Deployer is that any WAR should be allowed to choose its JSF implementation or to choose no implementation at all +Again Im able to pull it up easily within the edit dialog repeatedly but the delete dialog cant see it for some reason and they exist in the same view and are being called in the same place they are both within the same scope +Attempting to upgrade to though it appears the issue has been exacerbated +The current Doxygen build of the API shows the proper prefix include paths for file references +I am actually done with the changes needed to process filters for subqueries on HQL for the short term solution +Thanks for the great feedback +We will investigate more about this during the CR test this to TBD bin as this was not fixed in release +Getting the object via and performing the test suite sequence executes after which invokes the JPA remove operations +Its a bug +This is to allow using the authorization id of user performing upgrade as the grantor +Also if we say that its dynamic but its not a good idea to set it dynamically what is the explanation for why its not a good idea +Why do you need an other way to configure it +defer past +attach the theCould you include a unit test that has a load func that fetcher the logger +sorry my windows rights management is broken. +Note Disabled vi and tested invoking the again +use the growing lru map functionality used by michael in the ac evaluationhowever i didnt add additional debug information since the node identifier didntseem to be useful from my point of view. +Must have been fixed inadvertently +I havent yet uploaded the patch that doesnt throw an error if it is not configured and havent come to passing it through the analyzer yet or ignoring special characters +Looks like the worker log got cut off? Also what version of Hadoop is thisDoes it work with different numbers of workers +Thanks to Simone to look at my patch and cook up a better one that has been then propagated to the other branches +We normally avoid putting anything other than bug fixes in patch releases so this wont be in unless there is some major reason why this should be included. +Should this problem fixed in the codebase +looks good to me +Thanks a lotBest regards by Richard. +Attached screen works with latest versions. +bulk defer of open issues to +Updated source file to exclude problem code in windows +Similarly would have to implement this in java +Opening BARRIER after pending thread have been flushed and the digest has been retrieved is OK but we cannot resume stability +Adds quite a few things and more tests +So keeping this in for folks who dont want to worry about embedding the web view kinda makes sense +Never mind I found itThanks +Should be careful while doing this +i think this would also help with some deployment errors by tightly tying the data with the cluster config +This was fixed by but Ive added these tests anyway. +Upgrades Avro to +Change the wstx jar to version +Robert +All of them would be is that we need to define exactly how the scenario and thus also the Extension scenario looks like +Thanks for the patch Boris +You had this issue implying you had more work to be done +It may be that generatesa sequence of VM instructions that Jrockit tries to optimize but fails to do soin some cases +I am just looking for a small patch compared to the extant that youare proposing +After pressing Back clicking Appearance in the navigation menu results to the error as well +In addition the pom also needs to be manually installed if it is not present in a repository on the web iecd stanbolparentmvn clean dependency to is required by the unit tests of the bundle +One more thing the default is to sort by name but the javadoc says the default is system order +A which I should have mentioned earlier would have been to do a full republish every time by on the module in the Servers view and selecting fix turned out to be very simple +Classpath bug +Why would replication performance testing be undeploying applications +So I am pretty sure we cant use Trove if its license is not compatible in any formColt appears to license its code in two parts and the part we need is licensed compatibly +Do we need to add a new serde for this? can you add more in the description +bq +My hope is that this helps the team identify where the crash is happening +Hi JeffHave you verified that deployment is the cause of the problem? A normal deployment of the same artifact works properly? I kind of doubt that will be at fault if thats the case +You can imagine refactoring the Derby type system so that the serialized objects are lightweight enough that their classes could be included in the client jar +I just wanted to point out that my initial intent for using the alldocsbyseq view is to know which documents have changedSo I know for example what action to execute with a changed document +ivycachejarsUsersmumrah +Attached patch has a unit test and fix +The default implementation simply throws a +GA or later +Thanks Ari I just committed this. + is for a build problem that is keeping the mailcap file from getting included in the javamail jar file +I would comment though that the indent style is somewhat off when drawing comparison with the rest of the nuctchgora codebase would it be possible for you to edit this please and resubmit a patch? If you could get this sorted then it would be great if some other folks could have a look and pass comments +Sure is a useful bug fix +Actually that last link was for JBT not fwiw +Believe it or not these x images are the only and icons I can locate +This wasnt covered by +eclipse and uses it for next starts +Im getting the same error +Thanks Sanjay! +Hi JordanI havent had a look at the zip package yet but I am curious to see which are the reasons behind handling more events in order to provide to the injectee more entry points with Guice we should be already able to handle all injection situations but I could have missed somethingMany thanks in advance +Yes +Please stay tuned +ahhh +Moving code cleanup tasks back to works on streaming renderer by default +patch for this +Nirs point is a good one we dont want to have the cache polluted with translations +This does address the problem that you outlined at least to some degree +I upgraded the patch to the latest Protocol interface using and fixed some stuff here and there and added ASL headers +In revision Ive changed so that ujaxredirect means redirect to created or modified node +We have a functional test that covers the case were fine for now +Also if you could post the DDL for one of the tables that cant be reverse mapped that might help us to narrow down the problem +Pull request submitted with unit test +There are race conditions in the proposed patch and I never found time to look at them very closely +Im closing this ticket as it duplicates + +I just committed thisthanks pat. +Tried out the patch +Fixed failing unit tests +We use Maven you facing any issues with Maven +As a programmer you should know what information must be given to debug a problem + test mentioned in the for the test I have committed a slightly modified version in revision The test is currently ignored +Yep youve got as much info as Ive got about it ll take a look at what youve done this evening and might make some tweaks +Im still waiting to get access to some big logs +Attaching +I have never seen any reference to IDEA classes too and googling on the error is not really resulting in anything do you get a fuller stacktrace of the exception? May be that will give some hint on the origin when I googled a bit for a similar error the suggestion was to open a ticket with Apple for it +I think I wont guard for this +Verified on revison +Can you add a testI mean create a table like another which contains data and then describe extended the new code changes look a test and fixed some existing tests +Sorry I sent the wrong patch +Patch for +If there are issues with the update please create subtasks on this case. +For +Ok I am fine with the patch then +This happens in branch HEAD and too +and implement parts of the same operationprotocol +I cannot reproduce this +Reuse makes sense because original images are heavily reused +Fixed by ensuring a JBC Option is set to force data gravitation whenever the cache is accessed. +Is that the optimization we are talking about? +Bulk close for release +Ill run some tests and then commit +Thanks +patch is for orgjpoxstorequery +ivycachelogjlogjjars configHere are the jars in my project directory find +ok my bad +Closing the ticket. +Not impossible and in fact they may already have an adapter for it but will need more investigation +What is condidered Main in those situations where there are only artifacts with classifierstestng for example has a jdk jdk classifiers but no default artifact +This is for retrying +Proposed patch +The half file was renamed from to +Two new configuration attributes were added to the module application contextThese should give users the flexibility they need +Resolve as cannot reproduce bug +Sure I havent had this problem in a while +The idea is to avoid making promises to users that we may not be able to keep because adding support for more values might get pushed to a later time no one may have the resources to change the code etc +Sorry I didnt notice that an is already present btw we should study a Monitor that lives in the thats able to reload entries +btw on this new patch after youve with your logs and run it through the full suite +Yunkai Lets move this to since this is a duplicate of that +In addition I have added a test case in. +The underlying cause was that someone forgot to set a boolean variable in I could write a test for that but it seems kind of trivial +You need to fix your classloader issues first +Case closed. +Right now I was rebasing from the trunk to rebuild the patch and when I was running the integration tests for Hadoop it fail in the same way as for Andrei +Umm but well still need to treat these as Object parameters and do a toString on them to get a log right? I see that as no different from passing a stringbq +Attaching +DRLVM regression test attached +Requeriments hibernate with their libraries and logj library if you want logging +This problem only occurs in the datasource subsystem +All XMLCTF tests pass successfully. +Was looking over this patch and noticed it only does ODIRECT on writes +Hi Ben did you have any news from Altassian +Using avoids the effort involved in using SSL +looks good to me +OpinionsObjections +Please provide update or instructions +solved the major issue I was running into but I think this JIRA is still important as a general topic we should expose timeouts not retry counts as the user visible idea +What exactly needs to happen here? Should they all be +Hi Richard If you setclasspaththen it fails +I was only guessing +Yes user documentation on case insensitivity will be needed but my patch has yet to be accepted by the Derby developers +Ive looked through and everything seems okay. +New patch fixing issues raised by John +I dont think its users will want to play with loggingour commandline tools theycan create it on their own +Pushed commits to address above thanks +Its just a requestview sequence which the XML validator doesnt like but which still runs fine +What should we do with those patches? Should we apply them on or wait for +I would like to revisit this when Apache Mina is GA and the component is updated to use it +logsIm not blaming your patch its actually much better than hacking up the scripts to see what went wrong +Thank you for the commit knut +After discussion with Mohammad I think we should allow both kill and end +This class howeverdoesnt have the ability to set the anchor for the new url +I think that this proposal is to support a generic relatively powerful agent +a that is independently maintained and used by a larger community than just the mapreduce community then that might be a good thing +If the output file for native MR exist the query does not fail at compile time it fails only at runtime + +removes most of the calls to concurrent hash maps +Verified fixes this issue changes in didnt introduce tmp files are kept on file system because of their filename is constant so the number of wsdl files do not increase over text should be changed +Would it make more sense to use the SQL practice here? Im not sure that we can use PUBLIC as a username wildcard +Sorry the attached previously has a little problem that Disabled doesnt change to Enabled on an enable eventNow attach the new fixed one +It allows you to set a default fileset for each server type +I the example into a document describing the difference with original protocol and applying this patch +There is no connector to support old models and there will be no visual indication that the models are unsupported +an interface and have different implementations for CS FS etc +Even if we cant think of any reason for an application to rely on the scheme handler being reset we must not change the behavior between RC and final unless we can tell people how to get back to the original one +Havent looked at the code and am just waving my hands here +Anyone may reopen this issue when it happens again +Extract the lines into a new class named which can only be used on the client side of a hadoop cluster +Use released today +Please review +Great DagI shall make it into a patch then with the change that you suggested +I fixed the bug you discovered but the example still does produce meeting annotations +hi ruchithThink about this scenario as well +We deploy anywhere from to times a week least once every few you verify if this actually worked +Those deps are effectively a part of our API +Gant doesnt behave exactly like Ant without a complete reimplementation that is not looking possible +You can add it to page +Moving to and rolling back to these as out of date as we will soon be looking at integrating a new version of wise. +Wow thanks for your excellent comment +untar in quickstarts dir configure server to use oracle db for juddi and run issue is not fixed in our current version of jUDDI and we do not expect to get another update prior to releasing SOA +Jorge I apologize that my fix was not entirely correctThanks + +Fixed in Branch BranchSPCP Branch and trunk + +Cannot reproduce +Adding to JIRA like you have done is enough thanks +Thanks Deepesh for reporting this issue +slip to the possibility be added as a property which will be recognized by? +Consequently documents with a rotation arent displayed correct +Someone who never built richfaces from the source before can just replace orgajaxjsfjavascriptscripts from with the attached changed filePatched version of contains the same changes from from richfaces ve signed the problem under Windows IE Patch of branch same changes as in attachment but in patch format +Though that would be less commonly used it would be nice to have the standard On Delete actions available. +Closing this +batch update of issues with fix instead of the correct +if its or something else doesnt to the reference documentation explain the limitations of client side validation and why we chose not to support it within validation framework +Please +It only ever gets +The default for undefined properties is to leave them unexpandedThis may be considered a bug in the upstream build files +The cause was a final class variable was modified by another class and in local mode all the mappers are running in the same JVM that resulted in the dead lock +Maybe we can use millies and fill the lower bits from nano time +Given that we can specifically bind the NN and JT to specific addressport combos the admin has control over what is actually valid +the makefilescriptant changes are relatively easy but need someone to do the IA inline asm for atomicops in order for it to be actually useful +I will make the suggested changes and attach a patch to the JIRA +Hopefully the rd time is the charm +Will have a patch tomorrowlater this week unless something intervenes +That is the change that was reverted +Patch checked in +Dhruba could you please clarify thisIf I dont want to play tricks with replication policy can I consider your patch as a mere refactoring? I think this is important because we can talk about consistency of different replication policies mixed together but this is more an experimental exercise +Something that has to be done eg X weeks before a release? Or how does it was waiting till a bit closer to the release so I wont have to do this twice +I see mentions of Solr so perhaps we can grab the version of that jar and add it to Solr starting with releaseQuestion How will the spellchecker you are writing or considering writing going to be differentbetter than the one in contribspellchecker +The using directives for the +And Ant version? That may be the problem +FWIW I looked at this and it seems like there are no messages in the archives on minotaur so something is probably misconfigured for the mailing list itself +Ive added some code for mapping processing +Reporters of these issues please verify the fixes and help close these with latest build fromThere are no exceptions seen when executing the Closing this issue. +Ive begun the horrible slave work partHope to have a patch soon +It seems confusing to accept different terminology for the reduce sidebq +Unable to commit changes +Although the output folder will now be added as part of classpath container the output folder still needs to be removed from the container entries so that it is not included in the is now controlled by this feature is a requirement for you then I believe that there is a JDT option to set so that class files will not be reproduced. +All the issues related to this task is in trunk now and working properly +Fixed +buildDocument is made aware of start and rows and it stops producing documents once it reaches the limit is no longer aware of start and rows and it does not call abort anymoreIll commit this shortly +Ideally wed like to get this into the release +If you wanted to have this included in the upcoming release a working test case incl +In any case Ill it for was actually removed for Spring already since it isnt used within the framework exactly does your code use that constant? use it to access commandName in our custom tags helper classes +Hanging at mo +Ive made a patch that fixed can use it with the patch command like this +Is there any progress on this? We may need this or something like it at my current company + +later well have authentication. +Dont know if this is how send a patch +Looks fine by Boris. +committed +CDW currently defaults to first unreleased version which is in JBDS +Revisit open bugs older than a year as Out of Date +Id prefer that we not use wildcards +I dont know what this means +It has been a while but I am working on adding this to Pool now +Tested +all the other projects can use a variable based version number +Thank you Ryan and DavidI did not have time to look at this until nowUnfortunately I tried it on IE and does not work +Two are requirements for and there are two corresponding issues in libhdfs +Makes sense to me. +Unfortunately since that is an if it ever does do that it will end up catching that in the outer block in copy and a map attempt wil be blamed for itAlso now that I think of it we arguably should be incrementing the ioErrs counter before calling report since this is an IO error during the shuffle that prevented a successful map output transfer +for the to trunk. +Assign to during tidy up prior to the issue now that it has been resolved. +I committed the patch to trunk +Looks very good AliI just wondered if there is any way to determine if the app is running in google app engine without setting a config param +already fixed with + is copied into the war? Looks like the is somehow the reason you get the same errors is probably because you still have the war deployed +Ive moved all of those methods to Springs respectively unifying them with existing Spring utility methods where appropriate +This turned out to be a bug in the most specific method check that we had there anyway Due to bad ordering the most specific method only got checked for the annotation but not for +Thanks Namit! +The fix has been working for us in production for about weeks now no mail anymoreYou should do a jstack and ensure that the Camel and spooler threads are not blocked anywhere theyre not supposed to beYou can use JMX and in check out component mailetcontainer mailspooler threadpool +is to fix the security unit test failuresWhen I ran all tests I forgot the security part which has a different RPC engine +but I dont think we can make it work with thrift. +Maybe some wierdness in the quoting your post to the eye looks like it has wrong single quotes +it is used in JBDS this done? Can we try it out with a nightly build +Run +The changes from to are vast +Given dans observation will this be backed out of the branches until the correct fix is available +Make it auto flush +I fixed all svn props and committed the check for now so we arent fighting an uphill battlewe can refactorsimplify later in other issues +This is awesome! I will review the patch carefully but I have a couple of questions for now +Im pretty sure this is dependent on the media device screen feature +Sorry about that +Michael could you please upload this as a patch rather than a tarball +All basic tests are not having any issues not sure of any cases that I have missed +In any case we cant persue this unless the Apache code is usedMichael G +Paulex I have tested them on windows xp with ms visual studio redhat linux with gcc and ubuntu with gcc +After I removed build started to complain that is undefined during the apr component compliation +on v +Would you then please paste the full exception message and stack trace that follows? Thanks very much +with this patch the adapter adds the author feed to each entry that doesnt have the author element +The workaround doesnt Sun Java b Windows Server archiva tomcat context to deploy just wanted to comment that this works right now using jetty with mvn jettyrun on have the same problem with Archiva on Tomcat on Debian Etch and Mac OS X +Indeed +The list just holds statement looking again at I realized that Spring closes automatically the Statement +This one replaces the results with strings +Sorry for thatIts sometimes hard for us to describe a usecase that is so clear for us but indeed not always for somebody elseThanks for the time and well try harder to explain in the futureAbout the jira issue well try with full urls instead +Mark this bug as resolved +Will fix it later +This has been fixed on trunk by reimplementing the core functionality to use static methods instead. +I only deployed by selecting the ear file in the package view and doing Depoy To Server +Keep them as we dont have a better alternative right now +Yes +the fact that the Deployment is now being handled by means that the deploy and undeploy is happening Before and After each test method +I have fixed this. +Im more worried about safety issues after a mapped buffer like Jakes gets unmappedEdit but if hes using safely then I guess the clean out row cache after compaction code is working pretty well +Yes +The other issues have been addressed as other bugs +It seems not inconsistent provides some of the benefits and is considerably simpler in short a good thing +patch applied under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +first cut an ugly patch +submitting +But this is very limited comparing to the in general +thanks for the fast response +Hi HowardYoure right +Logged In YES useridIt does not look like a problem +Should we take a crack at this before committing this +Please pull back if this is not what you had in mind +For Alpha we prioritized the proper handling of SSH keys which caused a few annoyances and needed to get fixed properly +Heres a basic patch with a test case showing delete by term not working +In this particular case even after this is addressed you will still receive an exception as does not have a unique key defined in the metadata +Infact some of my previous patch has the similar implementation +Thanks for closing the loop on this +If people use XML then I think there are only two types of invalid data character and invalid surrogate pairs +Either the comment should be removed or the throw +Nit Please reorder the method location to its original place +We can always catch and swallow the exception +The original patch was incompatible since it changed the Scanserialization without versioning +However fsync does not throw an exception +Seems fine to just remove it +Is more than that required to satisfy this +sorry i am having trouble understanding +I will think about it a little more and look a little more at this db +As I see a connector may need two levels of cache one sessionconnection level and request level +We just need a clearly defined process +Are you able to or can you assign this to someone who would be able toThanks have gone back and added in the screenshot for the JBDS release and this will go live shortly +I instantly thought Provider as well +Concatination optimization for LOG calls not needed because times concatination of strings takes ms in IE and ms in of a bulk update all resolutions changed to done please review history to original resolution type +I can use the snapshot I created so theres no rush thanks +I added the import but still a bit confused by the presence of two files in that look like manifests +So the latest patch is ready to go in? I guess I could take this one then +Most of tests are now cutover except those that require refactoring of and Solr +As this is a component migration that has been postponed past Im shifting this to Major instead of news on this component +Please file a different jira for that issue +I feel we are lacking a way to definitively test thisOn the whole I think this is good code and worth further review +Its that the second array should be full of references to the first not new references to the interned string literals +the new column already exist +Agreed +The thing is I dont want my application becomes dependent from the way other websites deploy their images because currently every time they decide to change the domain to server their images then I have to update my application which is not so convenientFor sure I can override the should and remove the check you have added but then what will be the impact on the security? Do you have any advice to resolve my problem +Seems the problem as been fixed in for resolving this issue guys +Investing developer days to increase the coverage could be expected to save of those hours each week etci guess in short i think its reasonable to assume that the numbers provided are estimates and not to expect them to be perfect +RSA and +for a release unless the is imminent +to commit if the tests pass +If desirable i can then also commit +Jean does it mean that you no longer experience issues with resource optimization other than using custom skinWhen you switch to one of skins everything works fine +Actually that does not mean the CRM application is broken +Patch for closing this issue the patch need to be reviewed +Bringing into +Applied +For that would be the ER build +Bump +Thanks Justin and my apologies for the spam in your mailboxGS TrunkCompleted At revision +I will look at and add some tests there +Ravi can you address the test failures introduced by this patch +So after more carefully reading your PDF here Mike I think this ticket should get a better response than wont fix regardless of what code is in the project +Given the current patch is already K I would suggest handling only single hostname in this JIRA +brew install swear I saw that documented in one of our INSTALLREADME files but I cand seem to find itNoah Dave do you remember where you put itBenoit This is a new dependency from the docs branch afaik +Lukasz Lenart Any updates +However in any application not all pages need to be stateful even more commonly there are more no of applications with pages demanding stateless behavior +Tez does not handle the splits as such +The one I am setting it on begins with inputFormatnull +Its simpler if the deserializers can assume they will get a valid version +well have to see what happens after serialized inventory is received and is actually correct but inventory balancing needs to be done for serialized inventory which Ill create a separate issue for. +I and Suman verified the patch with and it is working fine. +Although I am accustomed to using the +Unfortunately the exception has to be caught in handleSecurity so were not sure if the exception really comes from the post processing of +Going to abandon my attempt for now suggest codegen for static fields gets cleaned up first then we look to do this +will take care of integrating this to Ambari Web +fixed in svn rthanks for reporting this issue +However ant reports javadoc warnings +This was fixed +The Test results showed all tests are passed why Hadoop QA says its failing contrib tests +The corresponding jira in is +Patch contains the following fixes Addresses the current presumed test failure by moving to track taxonomy epoch rather than creation time +Charlie explained this to me +This appears to be the result of a change for Brian +cherrypicking this to +It works good +New version with +Thanks! +that is ve at least one user who uses and as far as I know Im the only one that has run into this particular problem I think that setting true would be optimizing for the wrong case +I dont submit because the test modifies the integration tests only +Same code used to work fine in wicket +We still see in user mail list people saying Regions are in transition and they are not sure how to get around itAtleast this patch solves the obvious cases if region in OPENING state +please dont make anyassumption and just keep it backwards compatibible +Issue closed with release. +Please grab the source or the latest build snapshot and as I believe this has been fixed in +Incorporate thread safety fix +I think with the other Solr issues that are looking at and the ability for components to add results that we could rework these thingsAlso why dont the aggregate functions just work w all the existing functions +Can this assumption considered correctShould be the case I will also attach a patch containing more junit test for test covering the map casesFor sure though having as result testAliastestAlias sounds wrong +Craigs suggestion would solve this +I added a new selection in the pulldown called DEPLOYABLE +Ken would you please attach a simple Roo project to demonstrate the issue? Where do your unit tests differ from integration tests +Posts from the forums would be bridged back into the mailinglist +Last warning +The code creates a local instance of and uses for loop to iterate over it +patch looks good +heres a patch +Ratnesh Ratnesh and Ravindra for your patchIts in trunk rev Special thanks to Rishi Solanki for helping the guys +I got this wrong while talking to you about this earlier +This patch should release references to frameworks when they are stopped +Ill be happy to work on a better solution if someone can help me in reproducing the issue +This patch should properly apply to trunk +Committed +Thats awesome! Yes for sure it would be a big deal for performance especially it will allow us to cache the majority of the join query which could be common for all the cases +for xx patch +Could you mail it to me? gareth +Use tests namespace and namespace if necessary +The problem can be solved with a different plugin configuration. +They would be counted as output from the new task instead of the old oneIf we implement hadoop then jvm resuse becomes a less urgent issue +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content issue is out of date and will not be fixed +Sure sounds like a classic my memtables andor caches are too large symptom to me +on local cluster +Artifact cleaning for marketing component +A bit modified with respect to previous fix +Patch for trunk which uses the Atlassian repo for the JAI artifacts +Change to major as its a serrious problem if this doesnt on RC and it still doesnt work for my particular test +TomI added this feature in our consdata branch +Aslak please use the unified diff format for a patch +Thats only a broad principleThe tradeoff is indeed between paying to serialize store transmit and deserialize this value versus recomputing once in a while during deserialization +Each html file needs the new ASF header in comments +where not necessaryI will put that KILLJOB thing in a separate issue which helps cleaning out the local dirs of the job in corner cases +I did a very quick glance at this and it looks good +push out to +I only really focused on the dynamic validation feature. +The question is if we want to allow null or throw an NPE +It should go in srctestjdo +It is being reviewed right now but I can post it anyway +David thanks for trying it out again +I was pointing to the wrong H package +For most cases I imagine you could just strip them out or you could round to the nearest utfAll we want this ticket to do is expose the sampling Cassandra has so we can do the rest externally +Recorded in metadatadatabaseSending metadatadatabaseTransmitting file dataCommitted revision +But if we do go that way we should throw an exception and explicitly disallow partial update of signature generating fields +Well of course I changed my to use the back to the then + +attach a patch to fix it +Slight correctionit appears that its actually that saves the instance to the session not the +However if you then select a purpose the error persistsColm +we should declare that its for internal use by Lucene onlybq +great workcan you explain how to configure the datastore or is this feature not yet implementedi mean will the datastore be configureable in the +trying to read the DTD? At least it should be made configurable if doctype will be addedIn order for easier XML processing afterwards I would propose to change HTML doctype to XHTML +Fetching the whole row is fineWe should have LIMIT support like in the list command or equivalent defaulting to something sane like +In fact you are more than welcome to commit the changes yourself if you like +HA! good to see you here! have fun with the mergeHe is working hard its am in California +the multiValued stats stuff is buggy and wouldnt work with sortable fields either the tests only covered plain int +Could I already had used simple jndi to verify datasource lookup for the jdbc referencing test notice the logger still is using the class rather than just the package but looking at other code it seems all the loggers dont follow this convention so forget note about parameters being marked as optional this can potentially break clients that rely on certain parameters being mandatory no? For instance geoserver will not now longer give an error message when the user does not specify the host for it make sense to put all the specific JNDI specific stuff in a subclass of? subclass could define the parameters as well as override the mandatory flag on the existing ones +Yea I think this cleanup should wait as there might be some cases where we want to read a hlog in +This patch is a very dirty code and builds hang forever at examples but allows to display Tasks error logsIll finish this tomorrow +patch available would you please try it thanks a lot +Patch for capped collection support tests +Attached patch that changes the default prefix for the namespace from xsd to xs +set Josh do you have a patch for this issue that you could attach? Thanks! +Its a bug with TSEVENTIMMEDIATE will cause a segfault because the event count goes negative +Also upgraded the +Heres a patch thatll get the plugin working in and continue to work in +Also there are problems with pagination if it is involved in long running conversation +I have just noticed that clean build thing +Initial patch +having derived elements that are valid substitutions for bases elements in inheritance by restriction +I followed your steps and tried the attached application and cannot reproduce this see my comments +But this problem is not occurred again in deploys +In some cases it seems like its not even adding a runtime at all +Patch reusing the Set in in the append a small patch for another small performance improvement when applying a new rule do not create a new just update the internal phoneme setAdditionally some tabs have been replaced by spaces after applying the previous patches +Two of them could be related the last is during DFS cluster shutdown +I dont have a Windows box to test it on but it looks like Windows isnt too happy with double quotes in file names +for the addendum patch looks pretty good +I just wont check them in anymore unless they are a result of changes ive made +It would also only report every th failure +Patch v for trunk +Thankshow do you make progress? Did you find some time for testing the suggested bugfixThanks in advance for the info +There is a chance of this problem rarely when you have only one ledger to replicate and also that eligible target Bookie is slow when compared to others and all other Bookies are not eligible to replicate +Verified in +You can handle this in your event +If you see fit please assign to yourself and schedule for an upcoming release +Could you please remove each of the bundles under the groupId and then my newly created bundles back to the repo +Sorry for not going through all the correct hoops + includes the whole project +Using such a synonym analyzer for indexing and a non synonym analyzer for searching will work fine +Some more improvements please applied +Gagan thanks for going through all of these files +It would be possible to set up passive monitoring whereby a process on the monitored box reports back to the monitoring host +If possible could you add a test case to that breaks with the above error +Seems like this was committed +As Andrew said +line coverage +The Journal holds the +Changing the code generator will fix this problem but will introduce backward incompatibility for previously generated records +Thanks EricA has been deployed +Updating Fix For to Unknown on issues not targeted for issue out as part of JIRA cleanup +total r Howard admin Feb Howard admin Feb libr Howard admin Feb +Need to add OSGI tools as optional bundles +Patch to check to see if file is a directory or a file +I honestly think this is a bad idea its not normal to require the same dependency twice in your classpath and it shouldnt be allowed encouraged by maven +Please try again +I just committed this +This has been solved in. +I pulled the latest mvn built and installed +Trivial patch attached +I created a to work on the patch for initializing the artifact map ve posted a patch there that takes care of it for me +These fields should include a primary key in the mapping +Silly meIn that case it shouldnt be a problem +Thank you for your reply BrianWhat I want is to generate client code on folder in eclipse user interface mode like previous versionPlease keep on solving the issue +Im not clear on what Heinz is doing or if it is the same as +target file contains WTP feature which includes the plugin on which you depend but the mirror I have contains not So I can either update the mirror and update the +looks good but the multiple files in dir affect +there is not really anything to wait for other than the passing of time +I sent an email about the projectartifact naming in eclipse +Started to do this as part of but that was a mess +QA Verified these can now be created. +Perhaps the length of retrying can be a little short or something +Finally it seems that the solution would require too much and maintaince due to the integration so we have implemented a spiderGood luck of. +Got it +If we can get it into it would make some people happy +Hopefully you cancode it to handle it do processing get to safe spot and then throw an error indicating a real user interrupt during a wait +As promised a quick example of what the API would be like with the policy Ive suggested +Kristian I dont see where the is flag is set to true in this patch +Thanks for taking an initial look at the problem +I think that would be a more cleaner way to solve this +adding an updated patch that includes a fix for one of the new test idea when this will get officially releasedwould come quite handy. +Is it permitted in the latest language spec +Cool that is what I figured but i do think it would be nice to integrate it in as its a pretty common use case i think and would be an awesome feature to have +It also adjusts the logging level of some logs +Done. +Attached is a patch with changes to the appropriate namespaces +Any reason not to update to as well +Hi Yannickthanks for pointing out how to fix the datarace! For the record This Bug still exists inThanks +Sorry I made a mistake +here is a suggested fix for the issue in case it is in +Jiri please a project with RF and enhanced sample +These resources are registered by +Thanks very much! +Added selenium webapp test for appearance page +bq +Then I tried to erase +sorted +The main idea of separate VPE from model its to give possibility for generation preview not only from jsp xhtml but from other files too +Flex is winding down +We should make sure we have tests that would have caught these as well +Yes this does indeed resolve my issue +move resolved issues to so we can generate release notes +Thanks Bogdan that would be great +Thanks +But it would be nice to have all of it! Navigation for tagx file is a one more separate taskI am ready to collaborate to provide a end user callback and viewpoint how to implement this functionality in a better manner +Resolved by removing Richfaces from the project +I did not mean to sumbit the issue without finishing the title +We should always run ant package before running tests or anything +Will file a separate bug against spillgen +Not sure what the original problem was +A feature of the API is it mirrors so the userimplementor has precise options for overriding or augmenting client actions +This patch does not register a listener in the tld but uses to delegate the methods to +In addition to Lukaszs requests could you please attach the whole stacktrace +The action must depend on the app +and are updated +Now that work is well underway with AS all previous community releases are +OK sure +My Bad +The third is likely to take more time +This one is pretty straight forwardWill be doing a bit more testing +This error is because Googles title is not Google Search like the test is searching forThe title of the returned search results page is actually just Google +Fixed in trunk +All tests also passing for me +But the changes were localized driven by specific use cases +Added a test case which is essentially same as the attached in r. +in order to check just put JSR validator and fvalidateRegexp to the same work. +castor appears to be gone. +Here are some newly created files to be added into trunkmodulespacksrcmainjavaorgapacheharmonypacksrcmainjavaorgapacheharmonypacksrctestjavaorgapacheharmonypacktests +is now a valid super class even if is not on the class +Please also add the NOTICE file created by the patch to svnThanks! +This may be ok initially as the examples will probably only use the current set of bindings but thinking ahead any thoughts +I ran ant test again and only test that failed was +WDYT +I want to see if this triggers JIRA mail +Also whoever implemented is my hero for making CQL testing easy +If an annotation type is selected as enclosing type then subsequently it appears in history section of Implemented Interfaces Selection dialogI will report this issue to Eclipse Bugzilla +This also means we dont need the code to parse octal escapes +It needs to be suppported and I would prefer not to add another option to the database reverse engineer command to suppress schema m open to suggestions but this is a small feature addition to and there are much bigger features to devote our time to +Thanks +Copy the javagroups jar from the all configuration intothe and +If you find further issues please create a new jira and link it to this one. + +So heres the patch containing unit test as well as fix +The destructor of the consumer should not throw an exception if it is this is an error +Closing. +Pls provide your failure seems unrelated to this fix +Please give the next nightly snapshot a try and let us know whether it works for you +This is in some ways related to +Test is from reliability test suite +Extremely sorry for inconvenience. +Move Protobuf files into moduleThis new module should make it easy for others to interact with without depending on much from hbase +Before making the next patch will wait for others commentsStackAny Comments +Thanks for the patch Berin Ive committed it +I cant say for sure but perhaps you could wrap or subclass the introspector to add additional behaviour? +Thanks RobertI also added the test I missed from fix is good but it broke the system tests +The suggested workaround to use a package phase doesnt work +It should be unreachable +My bad for not updating the scripts that run the build +Chaitanya It is +Heres the correct had still posted the wrong patch +Is this still something you need Ben +This patch adds a new method to the Renderer interface that takes an extra String parameter +Sorry I made a mistake looks like currently RPC response doesnt have byte total length and you plan to add byte total length to rpc response right? If thats the case I think you are right it is fine for IO although the prefix in rpc response body seams redundant +A workaround for this today is to Store the Referer in Session before doing the Auth Redirect +Added headers +Yes Lukas IMO it makes sense to rephrase the label of the checkbox +DoneUsed the MTOM data binding part as the template +However this is not an issue about the as all the other text does not render correctly etherAll help appreciated ads this is a critical fault on our system +But as I mentioned earlier I believe that option A is less oftenImplementing option B is more tricky +Nothing appears when browsing the repository though +Thanks! Im really excited about having a reliable test suite so I have more confidence changes were making dont break things +This will default to false for backwards compatibility and since this has yet to be requested as functionality Ill leave it as effectively undocumented and unreported by the driver +Fixed in r +I had to use an optimizer override to get it to use the index +Good point Ive updated the patch to do that too +Do you have a cluster on which you can test both +Unit test to trigger the NPE +Simply throwing an exception in init tasks should do the cleanup and killfail the job +Ransom can you regenerate the patch +I will update to account for the same +Quoted strings not handled by original fix. +I just encountered this error in our own QA environment +Continue to leave this one for the moment pending having hooked up for automatic syncs +Tomi the patch looks fine +In any case the proper way to implement it in the xsl would be to check if the format parameter is specified and if not omit it +I believe this is pretty much a duplicate of the work Eddie is doing to add a control test container for controls testing +Note that s fix with hashing makes this work with bridges +Id better ask this explicitly Can the property be set on a database or onlySo far all properties that are only are also static +Looks good modulo an unneeded import in +It is broken on even the simplest mapping file or annotation +The git log also reports what I expected +Im pretty sure I know what the problem is here +Sweet +I do not have the intention to look at these now so Im unassigning myself +Feel free to reopen if you think there is still something to fix here. +in your case you should have usedreferencedinspectorid ie the PK column of the joined tableBut there were a bug in this area also so this does not work in betaIve fixed that in CVS +Given that there is this work around and we are not going to change the expression langauge i vote to close the issue and leave it as wont fix + +Thats correct after some testing I found a very simple POM which shows the error +Fixed +Ive deprecated that page so that it will no longer appear in search results +Thanks a lot +Have fun and I hope someone else but me finds this useful +I know I spoke out in favor of Java only a couple of months ago but I have since changed my mind +If you could perhaps even share some of the code from your that would be helpful +JD will no longer be usedThe situation here is at the beginning of a segment for example the very first transaction +thanks for merging streaming from trunk to types streaming patch +I was thinking so far at calculating what I call the interest automatically but did not manage a way of doing it +please verify the fix in the next build Stan. +just for understanding the issue it may help to try with all transactions read committed +Fixed by Gauravs with latest automation runs +Applying your patch fixed it +Would you be OK if I sent you a couple of fixes in the interim via pull requests into your branch? Id like to help out however I can but the last thing Id want to do is step on your shoes +This doesnt make validating the jars in a distribution any easier +I suggest Roo shell to be more verbose by first checking that mvn command in available on PATH. +Corrected the aggregate pushdown logic to use the proper projected names +batch transition to closed for remaining resolved bugs +Committed to and trunk thanks Mingjie +It should also support public key does this not work for you +First code drop of for the new email addon as per revision +These blockers and critical issues are resolved but not verified +It must be already fixed +Once all delegates in the parallel step respond the aggregate checks if there is a single step list attached to an entry in the for a given CAS id +We basically ignore the boundaries and split the files based on the chunks +the filename alone is not enoughOtherwise sounds good Jimmy +So I have to query an environment variable to find out if the default is being overriden +Still needs checking by some client code to see that this is really what is needed. +Not sure but will try my best +So sorry for the problems so farDo you think you could test your aggregating datastore functionality against my githubs geotoolswfsdatastore branch? Or I can rather send you a couple jars +have seen a few complaints about this error in EclipseSTS and I really try not to have any red Xs in my code +Will this fix still be in? +Needs more testing and also with files with old formatWill find out if something better is there +At the moment and seem most in need of javadoc +A solution with one zookeeper at each DC would work great except for possible latency problems +This may have been introduced in in Castor +init start stop and restart are only function calls on the same instance what the instance does in this case is up to the instance but somehow destroy the instance and load a new one would give ugly resultson the gateway page you have a list of gateway INSTANCES not gatewayswhen you change to code create a new gateway instance and delete the old one +Ill try tomorrow +patch to fix this thanks. +Thanks Uma for signing this up +Looks good +Or we could call next again +Thanks +If nothing else it could make it real hard to tell what error caused a build to fail. +Verified in +On the its funny but stuff in here could be unassigned for a short period of time but usually its transitioning to OPEN or CLOSED or its SPLITTING +The tests for the Standby mode should be on a +The images seem rather small for px width +on those blocks become less useful as you said but it may still cover certain rare failure modes +I agree with Gregory if youre looking for a performance boost then simply switch to use and instead of Hashtable Vector +Be sure to use the attached for any testing or merging +r +In both cases the are sorted +Major goal of is to make this change so marking it as critical +With the Acegi Security it was is instead of if +Addendum for trunkHadoop QA couldnt work when compilation is broken +This appears to be a typo on my behalf f is an instance variable and outEncoding is a method parameter +has been marked as a duplicate of this bug +Has been fixed. +organized buildtools hierarchy according to Maven conventions +Thanks it looks more cleaner now +But then I did it several times in various configs +To keep this on top fix version is set to Beta +Attaching a new file that is needed with the patch andshould go in xdocssourcesxalan directory +The user must hand model or import a schema from some other database and then edit to add additional metadata extensions to make it work with +No it is not invalid +This issue is fixed and released as part of release +Committed to x on +The issue described here seems to have been fully resolved +Maybe maven has changed +Added few more fixes +guess that content should be configurable as it can be sensitive +Feel free to push the updates and release a new version Im buried in work for the next weeks +Yes you are right +I think you meant to intersect the level sstables with the ones +run before Tokenizerbq +Attaching to add a NEWS entry for this changeCan a committer double check this and get it in with +The addSource method appears never to be used sources are created in their entirety in the init method +Bulk close for +Reference this JIRA in those comments +Work reviewed and committed by Martin +Thank you for your reply +Copying is the method of file creation +heres the patch +Im going to wait for Hudson to run on this and then commit it if it looks okay +No need for hudson process +wasnt as hard to fix as i thought +Thanks Dag! +A JIRA of type Support Patch would be required so that the patch workflow is followed +Olle do you have the error message you were getting when IO was not present? If so then I could also make a FAQ entry +Im working within the bounds of the Shell and Commons CLI +Jetty supplies its own and and it seems that we are getting classloader problems +It works as intendedThanksDumindu +Its just a matter of getting ibiblio to do an rsync pull from beaver as required. +Dan you can already use it in continuum if you put starteam provider in continuum lib doxia and +Committed to trunk. +Changes merged into codeline with revision +patch applied in rev +Personally I would prefer to abort to be sure I dont have clients trying to use this dead serverNote that for questions or discussions its better to use the user mailing list +There are no applicable XML files in the styles directory either +Closing it againFeel free to reopen if its still broken. +Im giving it a try +Will be reflected in the RC. +RI treats such invalid time as time without DST offset so we should follow this +Very interesting i will try to verify this with my setupIt would e very helpful for sysadmins to get a warning message which provides information that a queue is blocked for enqueuedequeue because of the fact that there is no space for potentially submitting messages to the dead letter queue. +No jar is correct +This needs to work nice with the configuration file as thats the way everything should be doneTo create a permanently datasets you configure in only servicesEither this feature must know which services are write so as to set some flag as its one servlet instance per service per dataset or the config needs modifying carefullyMaybe instantiate a wrapper service that has the old service and is then when the reverse is done get only service and put back in placeOr a wrapper for the dataset and change the dataset acted onThe wrapper is only present while in a temporary mode not permanently there although that could be done as well +orgapache +TamAh that is very good way to get more feedback +Thatsit +Going to dig deeper to find what else has changedJoseph for branch +Attached patch has the updates as above +Yes I am thinking it will be an alternate codepath for Solaris only keeping cp l as the default + +I added a boolean to to track whether or not the cluster had ever consisted of multiple racks +There is a slight complication now that deaulttreestate holds reference to tree so Im not sure how much is this relevant at allThe other thing Im not sure about is removal of listeners +mrepository file +New fix has so far only made it into trunk +Can you use this and check on your system +I see this as a request handler +All outstanding work for the release is complete I believe. +Pushing to since they are not critical for and because were approaching the end of +I leave that to problemIf you really want to stick to component dependencies first then configuration parameters you could use constructor arguments for the former and bean properties for the latter +I can confirm as well that is issue still exists +testcase ignores the Timezone definition in this case and takes the local timezone! +Closed as requested + +But it passes all the tests we had and the new ones I added +Actually I do so its not this patch but it makes testing this one difficult +In my mind the factory should create every single instance of the objects needed by iBatis +Read and found grammar and spelling errors in Chapter +Extension Metamodel retired in +Need JDK minimally for building and that check is already in place +You need to expose the base class anyway right +Is there anywhere else documentation should go? +Ive went ahead and tackled this one making sure that the code accurately tests for multiple Destroy methods and making sure that if an Remove method is marked as Destroy that is that is the default remove method +Thanks! + +It adds checkout and commit functionality to the web ui and cleans up the code a bit +If implementation with the current patch normalizes a URL containing it gets replaced with which is another problem to fixThis is the cause of spaces not handled correctly +This will probably need to be updated to include descriptions ofyour classesThanksMichael +We could have a master image file that stores a snapshot of the current state of a running job +Not sure how we could solve it schemacatalog has nothing to do with package names so they are not influencing it +Jiri could you please clarify where guards arent enough in your use case +As temporary workaround Im using manifest goal to create manifest file for assemblies nasty but works for me for now +Do you think we want to keep it or we should remove all supports for DDL queries + +No changes needed for +Given that Pig compiles against this bug can be closed with will not fix as far as I am concerned. +I have some tests for searching but I do want to add a few more tests fir the actual refactoring before I commit. +The ability to set separate principal for each namenode in HA config can be misleading but does not hurt +Dont think its related filed with the rationale +yes +Aggregator is the only stateful action that we deliver out of the box +I am a bit confused about how tmp is created by a user that is not jenkins? unless the other person explicitly set to be jenkins or they manually created that directory before running +I think the new functions should return the number of bytes actually written otherwise its impossible to use this as part of a serialization routine where you want to continue writing more stuff into an array after the vlongvint right +Patch applied in fedaeebeeacadebdebdAll tests passedThanks Darrell. +Let me look into all this crap +I am surprised it isnt bug as opposed to unimplemented +Sorry the patch has gone stale +Increase timeout for test to applied +Also added simple tests. +fixed in Revision +Under Spring I saw no evidence that this method was ever called at all +You should be able to use the class and add the necessary jsp files for the name node +Im not working on this +Small patch that reduces the you send me the java files +Thanks +Added a check in the jboss deploy plugin to run the verification. +Looks good +Done. +ThanksEDIT actually looking into a side question that carl raised about OOM problem +Resolved for a while +Since will remove this part of the code there is no more need to have this fixed. +I will try with Maven later and look for an workaround so that it works on both versions +The fix looks fine! Thank You +I have committed the patch +It will sync on CR leaving it open as a reminder though +updating Lucene in Action to explain all the recent changes to contribbenchmark and explaining the tiny differences between these was awkward They are nearly the sameIll work up a patch. +In the comparator can be instantiated either via a call to instantiateFunc or getComparator +make sure attributes descriptions are present all over the guide +There is just one subtask. +Thanks Esteve +We have also produce a workbench with the same functionalities program structure and complexity and this workbench does not have this kind of error +Will doI have opened another JIRA on the subject that there is no way to find this out short of trying +This also assumes that you have the dependencies already downloaded +Great to see all the recent activity +Long term I hope to see something like emerge for hadoop +Good idea KarlYeah lets have better error reporting +Bulk defer all frontend issues to +Bulk close for +Its often been challenging toexplain why NOTICE exists and understanding its origins is helpful. +New patch takes care of implementations +Please check that the patch was applied as you at by Tony. +Hairong mentioned that is relatively slow and more expensive that calling +fixed the issue and update the sample for check the issue +I went ahead and enhanced the example so this is easier to test. +In the standard implementation getProperty delegates the query to the metaclass +I am hoping to do my final testing this week +Well my understanding of if no source if found then returns true is not really ambiguousIIUC this is the case when no provider wire is found so we should just return true +it complains if is set to true +YonikPlease find a new patch +A little more information please? Whats the error +xiezhi thanks for your contributionI think this patch needs some update cause we cannot assume the port number is it needs to be dynamic as well +passed on my box +Probably the functionality is already there somewhere but I didnt find it +Konstantin could you take another look at the patch? thanks +I have checked on trunk and and both work fine +New features go into is a big feature and I dont know exactly when we can get to it +project +OTOH forcing the XSLT files to be under confxslt would avoid cluttering the conf directory I think its a good idea +But I have actually tried something equal with bspRootjobidsuperstepBut lets see +If youre using or higher you could try using it as workaroundBy the way Ive changed the bugs summary to something more descriptive +In case of not to modify should be create a different classpath for? attached with changes +I reverted the above changes in revision since Justin wants a full quorum for a PRC VOTE to pass. +Thanks Ivan for the review +Thanks. +replaced TAB spaces +Thanks Sanjay +and assigning back to myself so I dont forget about it +Hmmh +Just hit the same bug Works in Singapore Works in Bali Does not work in JakartaNow I know the workaround it is not such a big deal but for such a supposedly bullet proof and telkoms grade platform this is an embarrassing weakness from Erlang +worked when run on a standalone basis +Thanks for the commit Daniel! +Im compiling C binding for Hive +But itd be nice to get something before +Forgot to write with unit the situations the same +Thanks +page file and was producing source is nullThanks for your patience +Then my computer went to sleep and interrupted the upload +I guess the code to determine the correct URI must already be in there somewhere because test interaction using the servlet protocol works correctly +Is this using Cordovas API or the Geolocation API +Please verify with xalan classifying bugs as Keywords bugs that pertain to specific API calls or advanced usage of Xalan. +How can an NPE in an optional feature be a blocker +Its not something Im intersted in pursuing personally at this point but if anyone else is than please do +There are no immediate plans for this hence Im closing this +please test with these +This patch is for assemble model stest in source +Yahor the +I added a new integer text field to specify the SortFilter Limit in the Search Result Editor preference pageThe limit has been raised to by default. +Designer includes them simply for parsing the +I was seeing +Is this reproducible +Test cases are also updated to use this new version +Only when images are replaced by Page Segments are they not rotated according to +The other case is while the table creation was going on theere was a failure in table creation and so we go and try to delete the status and znode and also the table znode +Note that some unit tests in the branch may fail +interrupt aspect moved to closed as resolved and released +Can not reproduce +Additional steps required More tiles to be defined in hyperichqwebtilesi +Then in we put the username into the after authentication +No I just assumed it is related because its a test and you earlier wrote that data store tests are now enabled +What problem is this solving +This is provided now in Roller and above +Yes sounds good +Ive just committed this +I followed the procedure for converting a git repo into an SVN project with history preserved +but its easy to solve I can also helpAnd this is helpful +Sounds good +try patching am sorry to attach a wrong patch beforetry this +transferring Bobs issues to Mandana +Are there any more questions that need to be answered at this point? Is the solution mentioned by Greg back in August likely to be implementedCheers raising the priority as it appears to be too low +I removed the Assertion since I think Assertions are only appropriate for errors which should not occur in production +While its dead ugly it appears to be not causing any issues so Im just going to leave it alone for now. +I have so far identified code that does deletes followed by updates in the same mutation but I havent yet found any updates followed by deletes +Please confirm they have been applied as expected +No memory leaks found +is not the cause +This is in the way of making further progress with this and rpcThinking some more rather than thinking in terms of blocks when writing and reading Cells on the wire Id been thinking Id be reading individual +If you need additional from the JEE spec you need to install only the libraries you need +The shortcut mvn sonarsonar works now + is too broken for the moment +Is there any idea when we can expect Hibernate +I hope to Hbase can automatically recover service +Great that they do this sort of randomly +Closing issues which have been resolved for some time +I cant see the benefit of this +Will test it out. +testing is for stuff related to arquillian testng junit etc +My patch is based on Sebastians patchI removed the balancing script and added handling of region split and merge +On that new Jenkins I worked around this issue by first invoking a mvn install to populate the local Maven cache with build artifacts +Patch has fixes for both and was getting released twice +Works for me on master +Ill be able to start working on next week +That assignment was done accidentally with keyboard shortcuts +If we pass Scan and Delete to endpoint we can handle arbitrary deletion requestsAnoopI clicked on the link above it is not obvious which comment you were referring toPlease refer to comment by its timeSome drawbacks due to not taking Delete object +Thanks Stephen +Ive disabled the pool completly in Revision I found the hack and removed ve removed the pooling code now completly in revision +So isnt there planned a OCR extension for? +hmm +ssh client is not installed on the KVM host +sample sent from the sent to client +Please let us know whether this bug will be solved in the final release on November thThanks +targetclasses +I attached the Flex Builder project I created to test this bug but you can also right click view source from the link above +Pranav I think I am fine with that +I want to keep this on our radar so Im moving this to +Ive commit a fix please close this issue if you are satisfied +Thanks Anoop for review +IMO the original priority for this ticket was accurate +Simpler that way since the cores need to be physically present anyway in order to be referenced from a pluggable architecture. +Would be nice to solve this for CR +Salikh please take a look at +Committed to trunk +My reasoning here is that weve not talked about how to actually deal with this so I think it would be best to roll it back +I know this race condition is very hard to reproduce let alone debug +ok I will try to patch +When I added FOR UPDATE OF X I got NPEassert failure +I am using this Ant build script to run +Any thoughts +Thanks Prasanth! +Thanks +Thanks a bunch Jan +Patch to some of the base projects that address the nasty error appearing it does not address the problem of clearing the is relatively ugly because it is being done without changing! RespectDo you mind if we use the same jar in +Here is a version with tests +Resolving as fixed. +Thank you Laukik. +I synchronise before releases +Heres a patch that implements this idea +Closing this issue. +If I google for information on this problem this bug is the first hit +I have a concern about where the encryption information is located in the Developers GuideIt appears in sections and in some cases there are topics with different content but with the same titlesThere is this sectionJDBC applications and Derby basics Derby embedded basics Working with the database connection URL attributes that has these topic titlesUsing the databaseName attribute Shutting down Derby or an individual database Creating and accessing a database Providing a user name and password Encrypting a database when you create it Booting an encrypted database Specifying attributes in a properties object Then there is this sectionDerby and Security Encrypting databases on disk Working with encryption that has these topic titlesEncrypting databases on creation Creating the boot password Specifying an alternate encryption provider Specifying an alternate encryption algorithm Booting an encrypted database Changing the boot password I have created these new topicsEncrypting an existing databaseEncrypting databases with a new boot passwordEncrypting databases with a new external keyand have them located under the section Derby and Security Encrypting databases on disk Working with encryptionbut we should take a hard look this information and the links that exist between these sections and also avoid any duplication of information +Patch applied as revision closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Scheduling for AS as this is a core requirement +However you wont see exceptions from failed recovery attempts +But making the code simple sounds even better +the fix would be quite simple we just need to compute the right start seq id of the topic +Fantastic +checked in. +Grid can you look into breaking up these tests +where I meant to say JIRA is used to submit issues about +Here is the patch for this ArunYour patch is in trunk at rev +The page specifically lists the sequencer as a technology preview and nodes that the default output and associated are subject to change in future versions to better align with the Java sequencer +This was done really for interceptors throwing such exception +This is the reason the setter method got deprecated +part of the import line +The pool and the connection will both be flagged as invalid and will not allow client use +is the method also called often? Because if it is called very often then caching at that level wont get you much +That would be great +Jason on your comment on it is again a matter of proper configuration +Reopening to write release note Maven has not been a component of JBDS before this feature does not need to be noted +Attaching +The class already implements Serializable interface +I first commited your patch as is in r +also +I agree it has sense that if the webapp doesnt start it it shouldnt shut it down +Great nice extensionsIve added some comments about test to. +Im not sure how we can test the JDBC object store without either a mock JDBC driver or an embeddable db +We havent deployed to our Namenode yet since we havent done a DFS restart recentlyIts a one line patch but the area of code its in cant readily be unit tested other than generically by existing tests which instantiate a cluster and hit it with RPCs +sorry ignore my last comment i didnt see the first isnt this a major regression to JBDS +A test has been written for this case +it simplifies the code +Using default namespace in the restconfig unfortunately wont work so i would have to work around it +Now deploy a faulty service +Ill wait a day or two +Otherwise youll end up with tag spagetti imo +Updated title set component to SQL reset Urgency field since that is reserved by dev team at triage time +Patch adding max argument to Solr factories for and along with simple tests for this functionalityAll tests pass +removed protected attribute for the configurations storage since there is currently no other way to remove a configuration +What we need to do to get this in Mr +Agreed but this is caused by the client not properly handling an interrupted thread +Duplicate of. +can we move the from mojo project to surefire plugin +But the original content type header is still presentSo now the poor browser is confused and gives up by showing bad unicodechars +This patch should not have removed binmapred and related stuff +Any comments regarding the second enhancement hadoop jnipath. +Hi Zoe I think you may need to manually start the bundle +Do you want to make a new patch Ning that refactors the example so rather than a list of column names it instead takes an index file full of configuration that it then parses and inserts into the job Configuration +Probably needs to be propagated to master and branches in +Christian I reviewed your changes + +It should be possible to take these from the process definition through the action delegate previously mentioned +The new patch looks pretty good to me +I have a standalone Java application that is uploading files to a Tomcat container via HTTPSSSL +Fixed and committed +Its not practical to assume everyone who wants to submit a patch has this infrastructure setup +Now I see they revert the changes +Theres little reason to have the reflection wrappers anymore so I think what we have will be fine until we make that larger move +The patch is updated +I write an application that spawns threads each of which lists a directory of size for times +I guess the doesnt have a class attribute right +Linked some jiras related to the backup masters and the web ui +Working on a commercial project integrating SolrLack of an XSD is making integration with our business service back end a royal painWe do have from all other rd parties we are integrating with +Edited for formatting +Taps may use metricsanalytics modules but they arent the same +Can you please also attach the so we add it to the integration testsThank requested my file +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content EAP +Anyway sorry for the noise. +No adding entries to existing collections through XML bean definitions is not supported +Possiblly I have if you do come up with an alternative implementation that you want to share with others then by all means open a Jira ticket and submit it back. +The patch has been applied to trunk +It works I can SUM my counter columns using CQL but this is of course not a great way to implement itAny pointers for the right way to implement this? How do we want to define generic aggregate functions in the grammar and what would be the best way to handle them +Thanks Kousuke for picking this up +Ok that sounds good +I propose this convention IDL files should be under srcresources by package plus protobuf +Closing all resolved issues due to a successful release +Hi Hiram I saw you postings in some other issuesCan you kindly help in this one +Yes +is there any reasons to adopt different policies for allocation and deletion? For allocation where to place a replica has an effect on performance +New patch attached +Attaching benchmark patch brought with Roberts last patchHere are the current results on my machineOperationrecsrecselapsedSec +I dont immediately see how SWF can have an effect on this +There is one issue with this removal eclipse pluggin is also built by DRLVM build system +Uploading a README that explains how to install ANT to compile TuscanyAdding the ability to compile in debug mode +I am agreeable to a just warning being issued and then passing the PDF as a verified PDFAI have only run across this error on PDFAs generated by a very non standard method using GhostscriptI would not wish to change the jar more than necessary +Some quick comments on the patch testThe test breaks the pattern for tests in a few ways the suite method performs test setup +May be easier to code tooFor now only delayed containers +The class is not invoked during a Maven build BTW +Uploading a new patch that is simply a porting of Jacobs patch for +Wouldnt you agree with this +I get failures from this set of tests however I didnt look closely to see what might be wrong +New Patch I have made many of the changes you have asked for +So you can use UIMA to get all primitives from the aggregate AE to get all primitive class namesBut note! If you use this you will have to use directly the Java logging to manipulate the logging output +Reopen if need be +I am also experiencing this issue and would be glad about a solutionwe have to download quite large files and therefore i want to use client side caching +In case we do do this a patch with the changes to +I would try define using as example +Since Ive had no luck convincing the powers that be to reopen perhaps we can move forward with this oneIt sounds like there are at least two situations that can expose this deadlock having more threads than database connections and having deep object graphsI think that both situations are resolved if it is possible to configure hibernate to use a separate data source for id generation + +Close out issues. +That way your github work gets visibility inside in the project you can iterate at whatever rate you want and you dont need to go via committer +Hmmm there is not patchHowever I fixed this for both options to be safe. +OK the log is now anonymized +bulk close of all resolved issues. +No need to run the partition reassignment tool +Sorry +Id commit it but Im not setup to build the doc +reopening for possible backport +Applied this change in revision +Please reopen if you believe this merits further discussion. +The new license does indeed permit redistribution and modification +Closing all resolved issues due to a successful release +Note alterations to the CF schema can be made online +bq +Will post for comments in a couple of days +Will post a review board link +committed fix to trunkmsvn commitSending javaengineorgapachederbyimplstorerawlogTransmitting file dataCommitted revision +All accurate debug output is helpful for troubleshooting but reasonable people will differ in their tolerance for spurious messages +Nice Matthieu will try it this week +Duplicate of +to trunk and Thanks Gunther! +I should be able to get this setup today +Just a quick comment about the entity +after sometime a core is loaded and the version is x +FYIReceived a confirmation request from Dotster todayI just confirmed it Your confirmation has been recorded +Regressed with latest master +Youll get self explain error response if you try to violate these rules +However my desktop which is running Windows XP and JDK has not exhibited the same problem +iii You can create your own SQL objects and write data to your own tablesprovided that your connection authorization has not be set to read +Backported to with revision +Looks fine! to commit +Holes will appear when the engine is bounced and some number of identity values are thrown away +Thanks Jeff for the patchSeems it is not compatible with branch. +We need a test that doesnt depend on snmp gem ideally something simple that just uses +Previously we we only pulled the successful ones any others would remain and the next snapshot request would enqueue more snapshot region tasks but fail right a way pulling out the old cancelled tasks +We no longer force user to use mysql driver +we added warnings for package names and if you fix those manually the deployment to get stacktrace or more to know what fails. +I renamed the setenv script to bmsetenv and adjusted the call command in all the other scripts to call it by that name +The main reason I didnt apply the patch yet is merely because I need to learn how to patch the code and integrate the new build first +Another application not related to geotools at all +Im assuming that if it doesnt work in without SP applied it probably doesnt work in either which is a bigger problem if the goal is to support all the there a way for someone to test this using +Would you test my suggested implementation if I wrote it +All methods and variables need proper javadocs +An invocation of nonimplemented method via interface throws expectedPlease do not commit the patch yet one of kernel tests currently fail +Commit the patch to trunk at +compared to serializing through byteswritable intermediary we would save on bytes per serialized record as well as less memory copies +that will work nicely. +Ive committed the fix for this issue but left it open as I need to add a unit test for this case +I have a grid computing application where my objects are returned from the grid and they contain a that is a critical part of the calculated result +Tested with JBDS since JBDS is released +Check out the Alpha branch or tag to get something that builds +I have a patch for this I think +All are bit versions running on Fedora +bq +Attached patch also defines new constructor in to pass the Instance to the as is. +Sorry wrong file +Try that version when its out next week. +I updated the WSDL and applied the patch locally +Problem is that when forcing to execute in complete mode onewill have to alter the OQL query to return full object instances +Simple still check for null just using tail if now which is more in accordance with the rest of the codeName to trunk +we will need to modify to let queries with clause pass validationIs that just updating? I also see EQ referenced in updateRestriction but Im not really sure whats going on therebq +Should be Filetruncate instead of Filetruncate +If there are any more test or anything else that you guys want me to perform to test or please tell me + +Two questions I dont see a way to edit this how can I edit it through the console? If there are only two attributes in advanced do we really need advanced +So now instances are skipped +Fixed in +updated that with the new attachment +A minor comment on the diff otherwise looks good +These types are object mappable to Integer Integer Long Float and Double as well as simply mappable to short int long float and doubleIn addition I added tests toRan Derbyall and with no new errors +trunk +Great unfortunately I dont have time to do this in seam as I am trying to get a drools release out +I just codegened from the given wsdl deployed skeleton and invoked using the stub +committed in r thanks Robert! +Described the User Site feature showed an example of the projects XML file with a project example + for the patch +Removed monohtml references from subversion revision . +Im implementing this now +Regarding the packaging I guess this is a hot topic but I always found it rather strange that clustering is in the stat package +Please also remove the words Incubator Disclaimer at the beginning +That can be used as a workaround but it should not be needed version store is an implementation details and user should not need to know about that +I got to many thing going on right now to make a new patch fill free to mod my patch to work the way you want and use it to build a patch for trunk I would like to see this fixed in if at all possible +Heres a few things I dont understand how removing LOCKDIRvarlocksubsys from Sqoop Hive and would work? why the asymmetric patch for all the hadoop daemons +bug can be reproduced by defining the base type in a schema A and importing that schema A in a schema B and then extend the base solution is to create a class info for each imported types and simply not output them +Returning is good here because it lets the caller make gymnastics based on the order of the keys which is sometimes useful +Since this was being called after the was constructed it was initially trying to connect to a default URL which was incorrectIll upload a patch once I come up with a test case +Getting something committed to Apache SVN sooner might make it easier for folks to collaborate on extension points layout etc +Any plans to implement a fix for this it seems to keep being pushed back? +Opted not to make validate private or final for the sake of testability +Previous changes reverted. +Works as expected and fix the issue +Odd +For those watching this issue it has ben back ported to see +Yep in fact theres already a test case for in the source treejruby testPasses for me on Mac but fails on Windows under in eaeafa +Closing issues resolved in released versions +And there is a workaround +I believe Justin already added the ability to add extra classloaders somewhere for the sake of working with dynamic scripting languagesJustin could you have a look at it +HelloI would very willingly provide more information about this exception if only I knew how to turn on the debug on the Roo Shell +Most web service frameworks have a way to programmatically define a service +I think we need some sort of answer for this for other than users control your writers +The fix involved a change to how merge is done so that the order of property cascades no longer affects merged resultsThe other option was to document how the cascade order was determined and how the order affected merging a detached entity +So maximum you can have is maybe bits smile +btw Flavio there are a few lines with trailing whitespaces in your patch fix a new patch right after +Prototyped but either not quite right yet or exposing a bug in jythonOSR handling +Could you please create a new jira if you want to show compatible runtimes +bulk defer of bugsminor issues from to +Committed v to. +Probably a good time to do that would bewhen we introduce periodic checksum verificationId like to underline some advantages of using a client program to perform the checksum upgrade +Maybe Iwill wait until next though Im not getting the same error Libor is Im still getting the Selection Must be WSDL error dialog even with your patch +This would at least let me know that the problem occurred between then and the build +Does thing think of it +Closing all resolved tickets from or older +its more of a bridge +It instruments the Transaction class to illustrate that there is unexpected concurrent debugging code shows something going wrong exactly what isnt completely clear to me yetI get output from T but in every case the transactions detected are READ transactions acting over a WRITE transaction that has not been flushed to the main DB yetI only see errors after several iterations of the tests I think this is because the JIT runs and because the code has no locking the issue of where the concurrency barriers mean fields are updated matters +This means when a branch is merged its commits look as if they happened in the past and suddenly appeared +Please feel free to any of those when required. +Limit input values to and +committed to trunk +I also attached to this back porting task ported fixes from and +It is easy to create a database whoseowner is APP which can not be authenticated either at the system ordatabase level and the developer would not presently notice since noenforcement is in place +You misunderstand my previous comment but nevermind I understand what you mean here +Still present in +I love Seams use it for many purpose +This was one of the main reasons why people have been asking us for it they want to be sure that their cluster will share resources the same way as before +How to propagate it in the UGI? Use delegation tokens since the remote UGI is constructed via the token +Part of a bulk update all resolutions changed to done please review history to original resolution type +Bulk close for Solr +we could have the RM renew tokens but not cancel them +It worked fine after that though in the same session and after a restart of Eclipse it was slow the first time but Eclipse didnt end up complaining +Do you want to address Rams comment about patchThanks for the review Andy and Ram +I did find someone else who was using the same version and couldnt reproduce this error +But youll be able to tell the various handlers apart by accessing the data through the existing SOLR JMX beans +This is not a native launcher issue +Doh! See previous comment +Please this when such a patch is available +Jim this is committed but can you helping out with the documentation effort? Basically anything that is new can you make sure it is in the docs for this interceptor +Once thats complete Ill commit this enhancement +IIRC HQL is parsed into an AST then a visitor for the respective database is applied to the AST to generate the SQL +butIll submit just after reported bug fixes +The separate treatment of free memory for maps and for reduces makes it easy for us to extend the job configuration to be different for maps and reduces +jBPM integration has been reviewed and improved over the course of many specific issue reports +This could enable a more smooth migration path +This really is an issue here as everything is monitored with JMX and if there a chance that the monitoring is somehow stoping messaging then we need to figure that out +Closed after release. +Ill do it but it implies changes across more files +Any objections to creating a jar that contains service UI and API? dont understand what the Service API is +This is occurring AFTER it has already been put in and being deserialized +Passed all tests +What browser are you using +This issue isnt linked wthe issue above +Hi Christian I cant seem to get the patch file from this bug for some reason +Thanks Chatra. +Why do they need to share a interface? But Ill bring that up on that cdi spec mailing list as a separate issue +Please review your items if not ready for please move them out to future release +This has happened to me on artifacts that have no overlays defined and on artifacts that do have overlays defined +The could also be used in a divisive hierarchical cluster algorithm to choose the cluster which needs to get split next + +commonplugins +This NPE still happens for system scoped dependencies in +But Android Studio allows to create new projects then and there and so if there is no command line I believe that is acceptable +Thanks for fixing it directly +I personally would prefer that and if I were implementing it would do it that way but I am not implementing this and would not block this patch over that decision +Tablets can point to files in other table dirs as a result of cloningThe above procedure does not copy existing table configuration stored in zookeeper or the tables split points +Ah I havent read your patch correctly +It simply delegatesthe call to +I dont get but instead my test is missing ESB plugins +I have committed this +A more clear comment here would be usefulClear meaning what is the use case that you are talking about here? A small example of the behaviour with expcted output vs +is always called +I used jconsole and figured out what problem the JVM has +Add part in +Correct comments about +Moving this issue to the s call this or something like that +Sure! I can do that +Producer flow control in action +I removed the option spoken however it did not workI created a new installation and copied the data and yet was not +Feel free to put it in +just ran on Rails edge and the useless void context issues are gone +bq + for patch +Here is patch +You are right that setting under JDBC should cause autoloading too +Implementation strategyWhen a servers lease expires remove the server from servers put the serverInfo into a new Map servers and put a into a with an expiration of of a server lease timeoutIf the server reports in in that period the is removed from the and the server is reinstated by removing it from the servers Map and putting it back in the servers mapIf the server does not report in it is removed from the servers Map and the is processed +Would it be possible to include tests in Sonar test statistics even if Sonar cant analyze their source code? This would go a long way towards supporting tests which is becoming a popular practice in todays Java projects +RomanPlease do the same or modules do for compiling native code +So I created some new projects in the mq project for the NMS CMS and AMQP JMS clients but Im not entirely clear on how we want these packaged +I am not sure why it cores when ssl shutdown because ssl shutdown just send a shutdown notify request to server +This will need to tested on OS X Snow Leopard which is both a JVM and doesnt use jrelibserver +This is possibly a bug in the tomcat plugin since theres no problem with the jetty plugin +shows distribution of sessions for the moment modcluster doesnt care about sessions counting sessions would mean storing sessionid and time stamps in a file and processing it later +This would cause unneeded retransmissions of the message by the sender until the ack was received +have you tried with the cvs version +This has now been documented on the Transactions wiki. +However the should check for null pointer + This bug has been marked as a duplicate of +Hmm +Thanks Scott Karthik +For terminal vs CLI IBM writing style states do not use terminal where you mean CLI Terminal should only be used in mainframe context +Okay it was the CXF test that was wrong +bq +Can I close this issue +thanks. +So I am uploading a patch that should be applied over the latest one atAlso thanks to Jothi I am adding a small change to suppress the unwarranted findBugs warning earlier reported by Hudson +Verified during EAP closing +I will be committing the changes soon +I dont think there is a single component that depends on Java +To the lognet project it is invalid to you it has been resolved by the package maintainer +Theres also relating this not sure which one is better a standalone tool or a plugin +classify the some notes I made on different attributes used for based on element nameattribute elementshref base link a areasrc script img inputcite blockquote q ins deldata objectlongdesc imgusemap img input +I create a patch which modifies files under the pluginssrcmainwebappview directoryThis will bring some style changes on the tables in the DB Info and DB Manager portletIf this patch is reasonable I will provide more about the rest pages of the admin console +Sorry wrong version of the patch +The two configurations are nowBoth give the size over all threads so the size is divided byNeha I think this addresses all your concerns +Almost +We already had the strict property to do this check +I infer from your previous comment that you guys are doing xpdl on top of jbpm? Can you elaborate on this? Are you planning to contribute something +I was also able to run your without issue. +Sorry about the delay Im in a different time zone so only recently got your request for a testcase and it took me a while to isolate the problem +Any bugs or feature improvement requests should be placed in a new Jira ticket. +Is this still an issue in +and facescomponentsrich facescomponentsrich +We already add a lot of overhead to all data stored in Cassandra we should be careful about adding more +If name and alias are same no need to repeat it +there is no patch associated with this since im removing a file +This is a candidate but not committed for EAP for +to the next patches to Branch and trunk +If you want to know which files are being overwritten take a look at the files in we should add a warning in the wiki page +Unable to repro with r but I did make some gossiper changes there so maybe I struck gold +Screenshot showing the Swimlanes +Will have to wait for nevermind managed to get this in +Sorry wrong patch +I guess it is rather a client error +Good to go Noble +The code with Random data have been wrote early +I have started to modify the semantic of this class but much should be done +It may be more tractable to have a cache expiration policy in based on the semantics of this particular use case +can see the fix already in the code +I will review and commit the asked Aniket to add a test to make sure that we dont regress when we switch parsers +This patch for simply resolves some discrepancies between the src name of the language identifier plugin and the name of plugin within runtime directories after it was compiledThe consistent name is now languageidentifier this means that the plugin jar file is also now consistently named +When non resource is selected by default that means that the project resource is selected + +I will add javadoc and a state diagram in the design doc +Fixed in +I will correct that soon +lzo files on reads it respects and confirms any data specified in the header +I included both in case it might should have been more concise +Can you please make it public + +Transitioning to closed. +I hope someone familiar with this code can evaluate my patch +In other words it fixes the bugApologies for the confusing description earlier +already fixed +It looks to me like this patch is safe to backport. +I have some updates too please dont resolve until I have added them. +Maybe this should be included in the enumeration in chapter? +Most of this is already done but changing the rest would be a binary incompatible change +Getting close Yonik noticed I missed merging his page so that will be in next patch +After the snapshot is available please verify that this is working for you now +I suppose we could lobby to have that decision revisited but I doubt we would make much headway +My client was way out of date because I had reverted to a very old version to track down a regression +updated to fix my use of the incorrect version +I have committed a fix which closes the stream and deletes the temp fileCan you test it out on your systemYou can use a as the logic to retry in case the poll failed +Verified +Also there is a spec clarification that needs to be made when a service name is not providedMissing this feature will mean that we cannot show async from code but we can from managed code +Testedwith derbyall on Solaris x Sun JVM with these errors on both and frameworks believed to be to thispatchI would appreciate if someone can start looking at it it is much smaller than the patch for embedded SUR +Verified upgrade in ER reopened only to correct Release Notes field +return number of bytes been read when reach the end of file +Working on trunk as it as renamed patch for trunk +The sample Aaron attached is incomplete and doesnt compile +Does that make sense? Thats the way weve done this in Lucene and Nutch +This happens for row triggers if the WHEN clause contains a subquery +Q apart from checking with litmus is this the only issue that keeps us from making absolute paths the default +Thanks Owen +The struts request wrapper is mostly just which delegates almost all calls to the +Thanks +the patch for Jenkins sanity +bq +Thanks for clarification Mario +Fixed on trunk. +My error is and I agree the error handling needs to be fixed in still need the steps to reproduce the original problem report +Then i deleted the eclipse folder and reinstalled in the same place +Before the authSupplier was also asked for premptiveAuthentication +I think has its own lookup on keyboard activity so calling may be part of the problem +Patch to ensure we always set the finished boolean in the log aggregation threadOn a side note we havent seen a reoccurrence of the OOM condition on the nodemanager so we havent been able to track down what caused it + +The same for +If not its the client code freaking out and creating connections until the OS wont let the server create more threads for themAlternately you could try it with the hsha server so its not +I have run the patch on my linuxsystem without any failuresI request for this patch to be considered for a commit +I just checked out a clean source tree from mojo on down +Cant we just move that to Igors bug then? Or do we need both bugs? +I have attached a second rev of the functional spec for system privileges +The page Ludovic gave in his description states that the repository is on and you are referring to +Apparently Equinox marks everything as deprecated in this case but even then it is still a gray area that is now completely avoided +It came back after a while +Modifying vswitch type is not supported in cloudstack versions prior to +I also made some changes to the packaging structure +that last comment is confusing to me +Just to track so it doesnt get lostThe official Solr release is a bit delayed there were a couple of bugs that had to be fixed so itll be shortly SOLRSITE needs to be updated to the official release when its released which should be mid month or soFWIW +We however need to revisit this when is released and we need to install a newer subsystem to it +These issues are no longer regressing EJB +I cannot reproduce this bug on both and latest GO with mono since the message is Missing method its probably you have a conflict on your setup +good work thanks Vinay +You are right this is a bug +Yes I agree it is a duplicate +Okay new cleanup stuff should be in new JIRA tickets as this one is a big one now +Sigh +After Container +Right now your guess is as good as mine because it does not happenyour friendly Qa friend +The fix for us is to drop the trailing slash for the command +Sorry I hit enter by mistake too soon +I have testet it in version and it works have testet the issue and the stacktrace does not come any longer. +It would be nice to have a way to dump out all configs used their defaults and their documentation strings as a way to automatically build the site documentation and keep it in sync +updated version +Answer Change to maven and invoke mvn install +In fact we have other components that hang on older +objects sounds like a really good idea +A couple of examples in chapter were highlighted as XML even though they were Java and I changed those too +I shall modify that +Issue solved +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +The component resolves this issue. +Well take a look at this one +Disregard previous comment it is peculiarity of linking +you need to be using spaces instead of tabs +This is with the latest snapshot +Transaction and visibility are orthogonal conceptions and I do not see any reason to mix and match them +In this way user could quickly find one containers log of one application +Should we close this +rejecting this for now since nothing in log nor info about the project to reproduce or explore possible causes. +Thanks Will! +This is a really old JIRA +I have also built modcluster on rhel and and updated to include the new version of +Algorithm handler now detects alternate algorithms and passes them into the RSA routines +Thanks Vinod! + +Recursive references should be resolved just must be missing a unit test for such cases. +I committed the fix +its not a duplicate of itself +Assign unscheduled issues with Blocker or Critical priority to +This should be done in the context of Maven instead of the old the JEMS installer should be able to resolve dependencies at install time from the Maven repository. +Switch to using beforeafter class +Can somebody please fill in Affects and Fix versions? Now I dont know which release this is for +Im afraid my company abandoned Open for Business in January of and my efforts have been focused elsewhere +So the recovery adds duplicate replicas of the same block id to a datanode diskMaybe what you claimed is true but the current trunk does not treat the temporary detached blocks as you explained +Do I need to have some kind of custom serializer? Do I need to modify my JAXB annotations? Thanks +Ive just committed this +Besides often high traffic we sites treat different user agents differently +am i understanding it right +updating with Aarons recommended changes for using link tags to the ClassMember +We could also make the dispose method available in addition but as long as there is no real need for that we shouldnt do it +so the new writer need to write a new document catalog and overwrite the old one +Verified Betae +Fixed thanks for spotting this +In fact I was expecting this to be the default behaviour in Cassandra until it bit me +So I would prefer a scriptSorry just my cents +It defaults to true +Patch looks good +Erwin Bolwidt agrees to this change +the patch resolving this the path into master and branches with thanks to Alessandro. +A more formal solution will come in +It is not a blocker for but it neesd to be documented in the release notes +Hence updating the priority to Critical + I am not too concerned about since were due to fix that up soon is a bit harder to accept since I would hazard that the majority of m builds involve submodules of that configurationIll take a look through this though thanks for the patch +That is at best feasible for test environments but not intended for production environments +Is it really that critical because at the moment the jBPM work has my almost eclusive focus +for the patch +Will this cause the deadlock reportedOne of the necessary conditions for a deadlock is +That is the commandreload when using bin had an similar issue with an application which is using corba in AS +I have opened the separate JIRA for multiple runs of FI tests +there +I dont think that it is a bug that DROP TABLE does not attempt garbage collection but do think it would be a nice improvement +Sorry I dont really see any way without some serious changes to the event dispatcher API +I created a interface and and implementations for the operations to work through and added some checking to see that all the requested columns and rows are read +The v of this patch completely builds both jetty and tomcat configs and enables them in the of the jetty assemblyThis patch also enables the activemq and configs in the jetty assembly +I finally decided to backport r in done at r +I think we have hit a bug in the JDK on Vista +Done for Portal Tools Birt Plugin Integration Drools Tools and Eclipse Guvnor Tools Ref guides +error window for debug version of problem appears to be in throwing exception from native method which was called from another method +The is not installed to include folder +We discussed the current handling at the Hamburg Oakathon in theory its desirable to have identifiers that are more stable +I have dealt with the multiple source directory issue and I can get tests to run in the same project +Here is a patch for this issue actually this is a single line modification +You also should keep in mind that these classes are included with Java so we can throw them out in about a year +Thanks Luke. +The problem is indeed in +Fixed in revision . +This patch also contains several fixes in verifier graph dump parts of code unrelated to the issue but needed to debug verifier +The fact about the is a good point +We have not moved to hibernate for the same reason +I cannot see the issue on of a bulk update all resolutions changed to done please review history to original resolution type +defer all X issues to +is now working as expected +Give me till end of the week +this is a databasegrant principal fredfredsDB +I think this is a duplicate of +is not includedWould you please upload it +Thanks +You might find that you get off the ground faster while you wait for me or whomever to get to the bottom of this charming muddle +Yes you did say you were giving WGS and EPSG bounding boxes and I asked what coordinate system your gridset is using because you did not tell that and it is an essential thing to knowI do not suffer from Cant seed tiles with bounding box issue myself and I am doing it right now with our EPSG gridset +I did not get catalog mixed mode I added the data rule + +Maybe we can try to keep the API the same or at least similar +Thanks for spotting the incorrect message +Attaching v which adds the unit test and fixed IOOBE from Ivans comments +I have resolved this issue as the latest set of changes have been made and the revised document is now available on docstage +The Reality Check is also presented incorrectly +I have removed your logging system and will write my own since your stupid reply this morning +The functional spec says that this is the expected behavior but subsequent discussion on has failed to find good motivation for this harsh behavior and in fact this is not how Derby behaves right now +Here is second part of review +This patch also removes a couple of filesqlsrctestqueriesclientnegativeqlsrctestresultsclientnegative +Thanks +Nothing to do with who is managing it +Alas I was wrong sends the SIGTERM pretty much instantly +I wrote it pretty quicklyYou cant possibly have written it any more quickly then i did the original dismax parser +Fixed in revision +Based on the previous patch this could contain a pattern to be applied as the filter +If any user defined externally in LDAP is allowed to register with JBNEM without admin intervention would it not create a potential security breach? It could be dangerous if the user can assign himself a role but if thats not permitted a malicious user could create some confusion about the JBNEM user base +Here is my proposalInstead of extending the this patch introduces a conveying the resultIn addition a new namespace +Enjoy +It was removed as it caused many other issues +Thats great to hear! Is there a anticipated release date yet? +Since the initial approach is not required I reverted it +I added the df parameter to the select request handler +Hey Stefano description fixed +The optimization in the future would be to use scoring if random access bits are used +applied patch +Skipping Hudson since it wont detect thisActually Hudson shouldve detected it because system framework compilation is a part of. +without the patch the embed test will always return as a ejb modulewhich in turn will cause some errors later +to commit more Javadoc and unit tests +I have to run in virtualbox to even test this stuff since there is no available for my platform +Each lease record has a pointer to the inode +c gui itll work +Please verify the fix and close +I just committed this to trunk and +This is a bug in both column operations and table deletionIf they think a region is being served that region is not added to the list of regions to processUnfortunately determination of whether a region is being served is imprecise because itis based on infoserver and infostartcode matching a server that the master knows aboutWhen a region is offlined it should delete the server and startcode fields +PSI wanted to see how Hudson system reacts to my test code with my submissionsI have gotten used to the system and I will be more careful next time +close off release +I hope to do that by next week +Thanks +Let Victor to verify it +There wont be a release +project you attach +I will do this if this is possible and ifpeople think this is okOtherwise I will raise seperate JIRA issues +So after second run we dont have anymore SCM information available in the Sonar code tested +There was a temporary issue on with an empty javascript file being deployed but now search is working properly +Moving it out of release +Applied original patch better unit tests +Just tested it and all works perfectly! Many thanks for such a swift resolution. +Fixed in geotools added an extra test to make sure it does not come back in transition all resolved issue that did not see any further comment in the last month to closed status +It turns out that the replica created in destination is a temporary but not a RBW +Thanks +either of those should work actually but I think option would be the recommended way +Could you wait for Sonar? +added both SCHEMA and TABLE aliases modified tests to check if SCHEMATABLE is legal +just saying that we might need it in the future +Installing into to the GAC solves this issue for me +No code changes from previous patch +So maybe we can add a new Ajax service similar to searching that will create a new wave on the server side and will just return the wave id that will be opened on the client +Other than kludgy stuff like running commands in the background and sleeping then grepping ps for stragglers I dont know any way to do this in bash + AS Admin Console displays subscriber count metric for eachMessaging +Created for a related issue. +So it should be valid for all platforms. +Ive applied the patch and deployed a snapshot +The Tuscany Java runtime does not have much of a model anymore except for plain beans that are used to load the XML contentof SCDL files +The summary should be changed though since you cant really open any dav directory in IE whether an Archiva repository or not. +Looks good +Thats why I believe it is something with your setupWe cannot help you if we cannot reproduce the problemIf you are able to create a quickstart with AntGradleBuildr +Resolved by adding apecial parent class that is +Git format is fine acceptable by Hadoop QA +I like prepare +We use the priority to coordinate the tasks developers are working on +Verified that test fails without the fix +It works correctly +Complete item in revision and the refactoring work mentioned in initial comment +The second one would be used for this callI think thats closer to a solution +if i use the RI i dont need to add the navigation case so from my point of view the behaviour is not the sameprobably the patch was not the right solution +I also manually ensured that the test passes without timing out after this changeThanks a lot for figuring this out and fixing it Chris +Happens whether one runs the mvn grails within the app or not +Caused by agree but Im also worried about what happens when some of the other predicates are true +What do you think about title for Step then in +Could you give it a try +imo further discussions are needed +Thank you Santhosh for the review! +Committed! +There is no redeploy in Tomcat Manager installed on the server +We could look at making them injectable into extensions at boot time if you think its a better api +I may have opened this defect prematurely +How about changing the mechanisim to not be done by the but make binding impls do it explicitly when required +We shipped a config file for the first release of the platform so is this issue really serious enough to commit past code freeze +Merged fix to and committed revision +My guess is that you didnt author these in a paint program but in a drawing program +Note the File Contents text area is empty but the Finish button is enabled +Added test for unique index fix +This does sound more like a bug +Closing all resolved issues for. +Patch appliedIt would be even better if we had a special area at the top of each day where these events would be rendered separately like in Outlook and Evolution. +Every assertion is performed using the assertTrue +GregoryI would suggest referencing other by their names only such asThis way we would easily see whether the JIRA has been resolved +I just committed this +Enabling verbose GC output shows that there are during a benchmark so I think its running long enough +StackIf patch is fine can we commit it +setting a fix rev +Please verifyAlberto +from So all ambiguities are not runtime fail as Mark describesHowever yes Im not sure why War isolation is not good enough hereJozef could you test in Weld quickly +Its much easier to use than to switch out your JVM +A patch that caps at max rows +So i removed all references +Thank you for the tip +Max should this really be fixed in M build? If not please change fix for +But some like Jim may have a need to override. +this is not simple +Thanks Jonathan and Rohini +Im entering a separate bug report with my suggestions for other improvements to the Mail package. +The DTD hasnt changed for a while +This patch contains tests of changing a columns default value +Will set status to Resolved for consistency. +Is this an issue within Struts or Commons Validator? If this issue is with the generated for error reporting you might have better luck opening a ticket against Commons ValidatorIf you opt to file against Commons Validator and there is no problem with the way Commons Validator is integrated within Struts please close this ticket when you open the other oneThanks +Youre right we should really refactor the API to be get +on rev for +with no Spring code +Im attaching an attempt at a release noteIts a little contrived maybe itd be clear enough with just the summary and application change headings? +I think the changes are ok but I think there is a few more classes that implement the interface and need to be updated as well +out of and into +Its not reasonable to pump ALL the docs from SOLR into Hadoop to dedup and Id rather have map reduce jobs to find the duplicates based on the crawldb and send the deletion commands to SOLR +There is a patch under that fixes this issue as well marking patch available +This is attach SRX configuration +I am experiencing errors in the code and cant locate where they are so I tried excluding the test classes fromanalysis +please contact the mailing DanielIt looks like is heavily broken and already a step child of the folks +Also moved setting thereThird times a charm? crosses fingers +Perhaps we could do the same hereWhat do our linear algebra gurus think + +But I don t think thats a guaranteed thing and I have no idea ifhow the one is dependent on the other +Well +This ticket is requesting additional token information bet passed to the authorizor and permission handler interfaces which is a significant limitation on those interfaces which makes me wonder if thats something we want to address in +When will it fixed +We have not just tarballs but packages which has to be installed with yum orI am reasoning to the idea that having our own well tested and persisted image would mean everything preinstalled but not yet activated to start automatically +It is in fact a URL but is it incorrect to think that xerces should interpret the space as ? It makes sense that if it comes across a space in the URL that it would add the instead of throwing +The parameters were actually only used once when the descriptor got created +Role names and user names live in a single conflated namespace called authorization ids +Changed +Argh! Cant edit my own issue! If someone can fix the code block in the description that would be great +It is very important with a huge ontology +I am keeping a list of all the nodes that are completereviewed +Right I misunderstood what you wroteAnyway on extra complexity this is not a common requirement +Committed to master. +if log type is not a job do not set jobfinalstatus +Ill take this. +Has this been fixed for +Remove my email from the CC list not because Im not interested but because Ialready receive all the Bugzilla updates via +I have mostly been running on hadoop lately though +I think we should close this issue +It adds property to the titles and index entries for the dozen Network Server properties +Addressing latest for the quick responseI will try to go over the new testsThis is mostly ready to go +Target version no longer exists do changing to unknown fix version no longer exists do changing to unknown fix version +I will check with him to be sureThanks thanks Sande! Please reopen asap if you need us to something for this issue. +Converting to int throws conversion exception but converting to Integer gives +note is not important if ensures we wont fail +We need this +Ok so thats the real problem. +unscheduled issues to the next release +I notice that in the nightlies area there is a release of but notIt would be nice if I could download a copy ofAndy. +This is a duplicate of. +Patches attached +Created a new patch with seperate new source files without any changes to existing files +Patch applied on and trunk +Installed only IDE related software from the STS site installation history compare shows the addition of Spring IDE Core and Extensions AOP Autowire Batch Integration Flex and Web Services Security Web Flow +I tested it for trunk and it seemed to work +thanks committed to as commit acadcebcaebebce +As regards to local and Hadoop the are the same +Further investigation revealed from the logs that cannot be foundDuring the build and are created however the +Attaching which demonstrates a related problem for BUILTIN authentication you can create BUILTIN users whose names are too long to be used by Derby authentication +If we want the new API to work at the level though we can never really achieve that +So I described the problem with as much details as possible to allow better tracing +chris could you supply a minimal test project that had this error? We cant reproduce it otherwise +Thanks +but please feel free to if you replicate it again +In other words the source tab is not the SVN Aries in fisheye would be great but thats not what Im after +which we dont want either +Looks like something was up with openshift app +Tika was recently released and were not using this latest version +I reread that thread and also revisited the discussion in regarding the correctness of LENGTH returning characters for aa +patch added +Bug has been marked as a duplicate of this bug +Are there any plans to investigate what caused the performance regression Olav reported in? +That adds an extra step of cleaning them up +Update page adding a link to the API docs for the strategy + needs done first +Can one or two of the clustering experts at Taobao review and commit this please +retrospectively setting fix version in preparation for SDO issues fixed in SDO Java M in preparation for M release +The root cause of this issue is a fault in the Axis service population logic which ignored the parts list provided by the bindingbody element +BrunoYour patch is in trunk revision +Granted this was mostly just annoying but it made it difficult to kill the application in a friendly way +It would be good to have laid out the same for buildjars as it is for the distribution. +Committed to upon release of Hadoop. + +If youre not willing to make a patch this isnt very likely to happen +Thanks RobertPatch applied to LUNI module at repo revision rPlease check that this was applied as you expected +Tim I dont think it is related but if you are convinced otherwise please create separate issue and make sure to attach configuration info from Help About Eclipse SDK Configuration Details +I was just looking at the diff and it must have been wishful thinking on my behalf +The reason was that although Idea knows to copy resources to the target directory a la Maven we had thwarted it by putting in a different directory and using a copy directive to put it in the right placeI guess I hadnt experienced this before because I had done a mvn compiletestinstall before going into my IDEThis is exactly the kind of thing I was referring to when I suggested that we follow the Maven convention in this case +Minor change from the patch was to reorder the constructor parameters of to match the other existing constructor +I am closing the issue +The real problem is that creates temporary files in the source directorySure they need to be cleaned but they should not be created there in the first placeI think this have already been discussed dont remember where +Verified in JBDS B +To fix that you have to add the following property in etctrueIts now the default starting from next Karaf and but it wasnt the case in previous versionsIt means that previously some blueprint container were able to start before some others +Sorry for not being more clear on this +to commit +Thanks for your comment +Thank you Mark +Scott can you elaborate on this? Is this for checking the jsr names follow the spec +Housekeeping due to oddities in the migration +Since my comments dont seem to be getting posted to this incident here is a sample configuration file +Attaching two version of patches +Applied patch +each handler could check if the connection has a error +It would be good to get this info into the documentation so Im reopening the issue to make that change +Good on you +I will postpone this to Camel +Looks like it belongs in Category A to me +Im going to commit this by the end of the day unless someone has onjections +Can you attach a simple sample app that demonstrates the behavior +Closed due to the release of Tiles +Integrated to as wellThanks for the patch FrancisThanks for the review Andy +I committed the patch to +What is the url for the nightly builds? +Steve was doing it at the same time +fixed inI did not mention it in the original comment but the web page also needs to be updated to mention that you can now create a keystore that contains a two level security accessNote that I was not able to determine how to use this typo of keystore setup using the standard ssl properties. +Will factor it out and put it here prolly this weekend +fixed with svn revision zip specification Use of the term MUST or SHALL indicates a required element +Hi i want to add it to githow can i get permissions to push? test still fails for me +The is primarily for interactive browser based authentications and not the best solution for authentications +So after setting up a user name and password using the installer the user should now be able to login to the Web Console +New JIRA issues can be opened for any further improvements or bug fixes. +Manually tested +Maybe they were only in jBPM +This is because the decommission task is not processed properly resulting in agent sending back a heartbeat response that is not parsable by the server +It might not be worth fixing the current code + +If youve chosen to abandon the use of Avro in this project then perhaps you should make that clear and explain the rationale? Thanks +A client may want to update the column value only if the value hasnt changed since he last read it +Well get it when we get it +Alexander To be clear did you try one of the suggested workarounds and did it work for you +Atmos returns HTTP for successful creation and HTTP code for already exists +to branch not trunk +OK thanks for clearing that up. +is the rd patch of logical plan generation +Committed the patch with additional changes in revision +James it will be useful if yoy can put up the SLD draft document in codehaus +Already solved as per are in. +I just committed this to trunk and branch +I think its a bug in it shouldnt skip terms that are of zero length +Thanks Manuel. +Did I miss anything +Address issues raised by Tom +Now Im confused +added it to the link to underlying richfaces issue +Verified in Developer StudioVersion Build id Build date +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +Bulk move a zip file is a very narrow focused operation +Attaching the notes +Good changeComments Add annotation +In that case what is the need to reuse an existing configurations? Jene thank you for your patch +The addendum fixes the issue reported by Ted +commitlogrotationthresholdinmb will not be affected in this change +Bug has been marked as a duplicate of this bug +the contract of is pretty clear +push out to +I have also installed the AJDT extension +Maybe something has changed but I never found the api to be simple or friendly +added any chance of runnig above scenario on MS SQL Server and reporting yourfindings +Printing the seed of the pseudo random number generator should be enough to make any failures repeatable +This time another issue with a wrong order offootnotes +Thanks for the review Todd +I wouldnt be surprised if it was related to as you mentioned +Thanks FelixThats awesome +Same problem here with hsqldb +This would allows us to be more scalable as no of bad records could potentially be highI think the number of bad records supposed to be low +Yes +But everything looks fine to me +Can you make this an svn diff + +It uses the instead of the false positive I fear +Thanks guys for the positive feedback and the heads up on the patching! +This is an issue only when no key is being passed in +Also ignores when checking for local modifications +I just committed +there is no one separate validation phase that can be turned of +new patch is to try and trigger on jenkins with extra logging +ProbablyIMO the right way to think about this is how can we expose the data we have in and not how can we make this look like Oracle +The attached patch is a testcase that currently fails but passes with the patch for +fix small comment will take a look +Update to address Aruns comment +It will need to be updated to work with Java +to trunk +We have to check whether it is nil by checking xsinil attribute of the element +It is null for one artifact but not the other yet these are in fact the same artifact +Thank you very much! +it is fixed with upgrading of Richfaces to +I am of the camp where I want to just get rid of the +I dislike installing to the local repo unless Im developing one project against an unreleased version of another + +Stan I do not think this is a major bug as its a configuration error +live and you learn +unit test failures during build +you can see the effects on Accounting General Ledger Accounting Reports we created the view entity with the required fields as given in the issue if any one who have the extensive knowledge of accounting component please suggest us Are any more fields required for the above view entity? Submitted byRatnesh UpadhyayRishi SolankiSumit all the conflicts of previous patch and here the updated patch as per the requirmentsSubmitted ByRatnesh UpadhyayRishi Solanki Sumit Pandit Ratnesh UpadhyayYour patch is in rev ThanksVikas + +Thanks Todd Ive committed this. +I was indeed thinking of the sending side the patch looks reasonable for the receiving side +I have tested with different values of sample points showing always the same results loop variant slightly faster +If he prefers to continue to work with instruction then he must parameter it like that resolution dont recall if you are creating this bundle +for including in +Committed the patch to trunk +I am against adding more welcome contexts resources and all sort of cosmetic things that get in the way of running a production server +is there more context to the exception than that +After studying this to see if it looked I checked this with an extensive gridmix run +The suggested work around works for me +Attaching a patch that improves the logging +Verified and closing this issue +At this point I dont have time to work on that +slip all minor and move from M to a sence to update the patch and apply it? +It may be perfectly valid to appy a version undo record to a version page due to row level locking where changes have committed but the th change needs to be undone +I havent seen evidence of spurious class loader errors like you describe though +The session isnt destroyed until the end of the request so you do have time to do your logging +This is necessary since updated the eclipse classpath with connection pool libraries which will be outdated with the application of this patch +This functionality is available under Deployment section +Alexander please confirm? Doug is this acceptableWe have internal customer testing blocked by this jira so we please need to find an acceptable closure today +has a proposal for making these components pluggable which might make the work of this JIRA simpler +What is EBR? +Close after release +Bulk close for +The ping is a blind operation meaning that whichever program binds the port being pinged the ping will pass + +nor do they know what particular needs to get propagated forward +Also since this is an API behaviour change I think a flag to explicitly enable this makes sense +I dont mind sure +Wont fix as theres a available User might decide to create a issue though asking for a new feature to be implemented +I would like to remove Avro tests with HDFS and write a separate test common that tests as done today with HDFS + does support mod calls but only through XOpen function stale resolved issues +Arquillian updateuse and instead of the use in tests so they pass on AS nowother changes in persistence +Thanks Ratha for the contribution. +One observations was that I can not test the schema that I generate with the original schema +Is this still a problem? +x +Any reason not to +Additionally I will doublecheck the used xslfiles to check if they arent using something which isnt implemented in xsltcas described under docsxsltcconf +Can you attach the Eclipse configuration details to this issue please +Closing because bug could not be reproduced in +Relative to aI improved some comments and improved some code formatting thats allRegressions ran okAgain the corrected behavior can be seen by running the new test case testDerby without and with the patch applied +Bad idea +Hi Regispatch has been integrated at r +Test failures are not related to the changes look fine to me +Let me know if this patch works for you +The rack represented a group of machines that are close to each other and also within a fault domain +Merged into the master branch. +I have not added support for that but assuming the file is kept from each run a separate mergereport target can be usedPatch ready for review +Im reopening because the committed patch while ok is only a partial fix + for the patch +This analysis and the error reported are stale. +Have we determined that this issue should be closed as Wont Fix +awf andrew incubator Nov +Which of course also breaks lazy load but thats another problemMerge init changes set more than once but we handle that by allowing setting to the same value as many times as necessaryWell look at Tuplizer +Since was for we still need to decide how we should pull it into hbase +However if the property is set before the Interfaces property then it might break the code again +Because streaming can use binaries that the user doesnt control they dont always have the option to add printing bytes to stderr +This file contains a small sample project which reproduces the problemSteps to import create new project extract the src folder from the attached file into the project folder add Hibernate to the classpath of the new project open Hibernate console and create a new JPA based configuration for the project try to open the Session Factory +This version makes the changes which Mike suggested +set can be made public +What is the suggested ant for building the dist now +Greg mind attaching a patch as you are using ehcache +Thanks for the help! +add null pointer checkbug with code assist is not reproducible neither linux no windostested at least of linux workstations +We should not run this task for any archive that we release +Niall I agree with the example you provided +But XML is valid +Two slight problems +There are down sides to using so this should really be a deliberate choice especially as this exception is only raised if the user has decided not to enable as the global default +Hi HanineWhich Embedded ID bug are you referring to? Could you please link to this issueThanksGeorge please test against to confirm this wasnt already fixed +Given that this patch is not a straight forward port I will not commit this to merged this change to. +Rich Newcomb is working on this +Looks fine thank you George +table name modified from PARTITION to to PARTITIONS +Titles and subtitles verified in released beta build. +We also need to be able to bind the to JNDI on AS and EAP for extensions that rely on the lookup of the in JNDI to work +Ive tested the too and I guess its ok +updated with compilable version +push out to +However we counting different things +This is hopefully something simple but needs to be investigated as it may be an issue with AS integrationDoes this happen with other browsers or just FF +I assumed they were generic error messages but I guess I filed the JIRA a little too quickly +The plugin was added to avoid having to code addtional bits and pieces into the main class if we can have the logout codeui + +Yeah the jndi port isnt used it needs to be removed +Created to more slot millis logic +commited this +Of course you are right +Attachment has been added with description This file contains sample +If you need more help let meus knowTed reproduce this in +Did some simple test on a cluster as well +Integration test is a side effect with the Developer Cockpit plugin so we must keep a backward compatibility. +Apache Felix Web Console has been released +So assign it back to Amila +Is all the code already contributed? I thought it was in Google code +in the description +This is not a blocker so should not go into the local mode we should disable the confusing message Cannot get jobid +htaccess to redirect to the Confluence site +Ill make sure to submit future patches as pull requests +I see that is said to supercede this issue but I cant see whether it has been released as the new archetype plugin +Using the manual way is easy to overcome this barrier but it is very stupid I also used the similar solution in my project temporarily +It is scanned when it is needed +I think you can close the issue +The old API will work +Log with a transaction +Any ideas when will be released officially +This sounds like a good thing +Since the passed into passivate contains an instance of the nobody has an objection Id like to whip up a patch that does this? second thought although this would potentially work it would also put the system in a wierd state since all the components would have been Destoryed +Good idea +Could anyone clarify? Thanks +I think the newly added incremental time should not be on by default as well as logging per id for deletes and adds +Attached +Absolutely agreed Jason +I would think if they were still using Java and arent able to use JNA for some reason? Or are we requiring Java for vWe are requiring Java and this patch gets rid of entirely +Fixed in trunk for these featuresMflowfeaturesMjbpmfeaturesMjbpmfeaturesMjbpmfeaturesMjbpmfeaturesMjbpmfeaturesMjbpmfeatures +Eric I think modifying someone elses patch that was contributed to the project with the license box checked is fineThe license box isnt strictly even required +It may conflict with original intention of missing timestamp +Load this is node is defuncLoad this is node is super OKThat is correct +But its running tests now and should be landing soon. +I dont think the patch needs to be updated +I think I have a patch for it just need to run all tests +bq +The default is session scope which is why you are seeing the problem. +The simple solution would be to add a special value to the repositorys properties +Arghh +Please be more careful next time +If you dont want the test classes in the jar then thats fine why force others to feel as you doIm not forcing anyone but Maven simply cant be all configurable itd be a mess +I have already detected that whenever the list is loaded or changed multiple list changed events are raised however we only need just have no time enough to participate in the project but I can share with you my current implementation that does the trick +So eventually this issue wasnt solved? Were experiencing this brutally in production with adding the comments made in to this issue for who ever read this issue firstIn Jonathan you can see that is used +this code is free and clear I wrote the code I own the company the code is ASFs under whatever license they want to use +sure +Not sure if we should add parameter utf to etc file +As my resolution comment states though I can back my fix out if we decide we dont like it. +The font for version number is Arial Black size px +I already have a patch to fix this for the trunk +Thanks for the clarification. +I made a minor change to the initialization of the ivy object with a new property called Build Property Files + +The relationship type is CONTEXTMEMBER +fixed it +t +What is really a workaround at the moment +Isnt this a dup of? I was tinkering with that a bit but feel free to close it since theres a patch already posted here +The warnings about the Invalid length of the css and js stuff are still there and sometimes the page is not rendered Ricardodo you mean you could not withe current patch or with the previous coderegards Malcolm check into trunk +Thats pretty damn cool +mostly done IMHO +Committed. +patch file and committedThanks MariettaChatra +moved to sandbox +Fixing this also impacts following test cases +Its a feature +I was just my original comment and I hope I did not create any artificial rush on this issue with my comment +Hopefully third time lucky +I got rid of the problem by opening the spreadsheet and in the edit of the cell place the cursor at the apparent start of the text +Nope not about +I though about describing the shortcut way around adding the Eclipse site to the list of available sources but decided to describe the long way for a few reasons it is the standard way to do it newbies should not be using short cuts if a user has already added the Eclipse site to the list they can skip over the install guide does use the suggested pattern for adding an update site +Thats how I realized we should use a different delimiter +My intention was the same Ashish BTW last updated patch comes first in the row +it generates the report in target a lot Arne +So we can just change them anywayPlease let me know if anybody has any objections to my slightly modified patch +I still think we should document it though +I reproduce exactly your results on a Linux computer with openJDK on an AMD processorIll have a look at this issue +A possible problem is that there is a single record that does not fit into combiner buffer +my naive question first why do we need Externalizable at all? In any other case a simple Serializable works well +We need an alternative solution to logging and I guess the JMX api gives it to us atm +Can you tell us what we can provide in terms of trace or a possible debug module to help diagnose this issue +The only disadvantage is that the user must wait an extra s or so for the login page to load for the first request +do both at once +Added a JDBC datasource to the AS kit and an integration test which verifies that the DS is correctly setup and can correctly access the repositoryThe documentation has been updated at the above mentioned links +use to from testlibs +reveng strategy name is validated to be Java valid identifier. +Thank you +BTW yeah it will get new pair but it will receive them from cache and those instances are no good. +also change the release process +My current plan is Under this issue change network client emptyArray and nullParam classes to return the generated key +Done +Mike I think this use case is more than just a classified use case its something that developersmodelers would run into in any organization that develops on a different database than the production database +patch puts the same compartor in the quantile job than in the sort job +It still happens with version of the plugin which uses is not an issue with the version of findbugs being used +There should be a property page on the jbossws enabled project to control this setting +This is WTP we do not contribute into formatting +Please customize the script once you installed the attached couch for testing +The default namespace usage was causing issues in this scenario and always qualifying namespaces with prefixes avoids the problem +Patch looks just committed this. +To use such a timestamp my POM needs to force all plugin SNAPSHOT dependencies the enforcer and release plugins should also check for this when they accept a timestamp for a plugin or m having trouble with the idea to use a timestamped snapshot to ensure reproducibility since snapshots are always the first targets for cleanup +Seems like if you add the new methods expected by and have them throw you should be ok for both and +It seems to work fine + +And for now such tables can only be created programmatically +Maybe this is not defined in the specification explicitly but if you look specification as a whole it may reflected +Then jar it up after aspectj compilation +After looking into this again I think it is not worth to complicate endpoint syntax for this use case +CS will treat the VMs which are not created by cloudstack as rouge vm and will delete it +Sounds bad +That class does much of the implementation +Even though some critical issues were fixed in to make it possible this jira cannot be adequately tested until the makes it into the distribution. +The protocol implementations will not be written in Java +We had about messages in queue +Many thanks RolandI still would like to see if the new functionality can be made optional throughthe use of a marker interface +The inner class is probably big enough to deserve a separate source file +The approach could fit because an event and a thread are quite similar but there are many possibilities for doing it context could be enhanced by some after destruction phase +Committed revision +Thank you Andy +bq +Patch for the conf file +Closing this record since is now released. +Cleaning a few of these up +Need to push the site still +When youre linking youll therefore need to be linking against ICU +The system generates a delivery count record every time a message is deliveredif a persistent message can be redelivered infinite times +Yeah i wasnt sure what that line was doing in there either +This issue needs to be resolved at the compiler level +I then looked at the source code and the CVS messages and it seems that the bug was only fixed for +For the discussion with Max doesnt have support for jdk yet +I have been using the beta in my sandbox to date +Hi AndyCould you please give me the stacktrace you mentioned? The Controller class would be useful to have as well +verified thanks +Reproducible on occurs in Internet Explorer +so well revisit the issue later +In we made it possible to avoid the use of but even our own hadnt broken the coupling +Reopen issue since it failed to move to Resolved when asked to yet no longer shows the Resolve Issue option +This is for only Im assuming as trunk already has this +Thanks very much Knut Anders +Thank you Lohit. +Ill resolve this issue as fixed once its fixed also in EAP code on EAPEWP CR. +thanks pat. +Thanks for commiting the patch so quickly +Just to resolve this issueDont forget that you can still disable the archives builder in the preference pages as described above +I dont think theres much we can do about arbitrary bitflip failures like this except depend on the fact that we have a quorum which makes the probability of such a failure much much lower +Resolved +Yes why not +So all issues not directly assigned to an AS release are being closed +Do we expect different serialization implementations to share code? If not this should probably be in contrib +Attaching a patch which adds a test on +deprecates C extension support so this is now a Wont Fix. +Please create an appropriate issue if none exists for your problem +This seems like a required behavior +Can you attach a full stack trace and indicate what version of OSJVM you are using +I think handling Boolean and Comparable pretty much covers of the cases. +See with JBDS +All people of should remain admins and certainly be contributors no other changes +You can also set your input attribute to a servlet or another Struts command that does nothing but return the status code you want +deprecates version a which due to a last minute change broke quite a few testsIm still asking for comments on the approach as there are other ways to solve it as wellIm rerunning all tests and expect these to be done by the end of the day +Fancy writing a patch for thatElse I guess we need to include the xalan jar in the sar but id rather be able to remove thisdependency + is the API that developers and the ESB will use to communicate with the registry +For a multi module build this is almost certain to fail on a Continuous Integration server like Jenkins. +Ive suddenly run into this too as part of my object structure gets more heavily populated +I just committed this +Ok +A would cleanly do the trickIf this includes shuffle in trunk youll have to do it in nettyI assume you already thought of this but there should be an ONOFF switch in for this +Attaching a screenshot of the path file tag to always render empty value +Not a big deal +Merged into master branch. +How can a test suite create a problem +Thanks for your contribution +I changed the output connector tab names because users were having problems figuring out what tabs came from the output connectors +The test checks that main threads thread group is not a system thread group +Committed patch to documentation trunk at revision +so we need to fix this issue for release +Seems like a reasonable ll try this one too +Mike are you going to commit thisBTW shouldnt we have handled it as part of? One of them is a duplicate of the other I think +Otherwise Ill commit this and this weekend +Can try committing it via svn if we think it should be there as it is +What do you think if we support the same behavior in HDFS +There are several query in this test caseThe idea is just set the +That makes even that mechanism pluggable +Ive created because I think these are really two different issues although closely related +My compliments to all contributorsRecognizing sparse files and reading their sparse blocks was not that hard but correctly writing such during exporting will be more complicated to toThus unfortunately I wont find the time to implement correct sparse file export and Im more than happy that I can at least kind of skip them +Changing the Fix Version to TBDEAP as this issue has been fixed in CP and not in CP which is relevant for EAP CR +The show provides the Folder Bookmarking feature +Your patch does more than just make the header extensible +Now would be a good time for a review +The maxThreads setting is applied to all services and only one of them using a JMS listener +My understanding is that Torben was able to figure out classpath issues and solve this Torben if Im mistaken reopen this +Ive located the offending code and have a fix lined up. +for scratchpad for discussion on mailing list +ChrisThe last date for feature proposal for was and that needs to be done on dev mailing list +First issue resolved in +Great thanks for the confirmation! +Looks like the Derby requirement of FOR UPDATE on updatable resultsets is causing quite a bit of headache +If its normal maybe you could catch it and ignore? Its just confusing now. +I have added this feature for the following containers we need to add it for any other containers +Very cool Frank but I wonder if the best home for this is the Struts core codebase +Patch looks good +all set +attach same patch to kick jenkins +Jon I started out that way but I felt that all the try +If I missed your idea please correct me +If you finish it before we release beta please correct the version +Thanks Mengwei +please apply CSS styles to fix the. +I havent looked at the patch but Im sure can be used in conjunction with to further boost performance for repeated range queries +Thanks Vikram. +The other issue with the table bounds remains +just checked the changes in CVS +Thankjs for the feedbacks Andreas can you provide please a patch with a testcase? It will be much more than appreciatedMany thanks in advance all the best +Note too DR is out now as well so we should use that instead of DR +Looking in to if this behaviour changed to confirm +I mean the servers should be configured to be aware of where all of them are +studioeclipseplugins +I think its unlikely Ill get approval from my employer to sign a CLA in a reasonable amount of time +Interestingly I cannot reproduce this in a test +The current patch only has the fix +Thanks for reviewing Flavio. +Sorry David I thought the html pages were being dynamically generated on the Apache serverIt seems its not +Im not sure what you mean +These patches have been committed to trunk and branch with some minor modifications +OK! I think were nearly there +The fix you propose is indeed very simple +Ill try to merge the best parts of both patches +BTW I dislike hardcoded name like this but unsure if we can fix without adding extra depsThough is still mentioned in the classlib porting guide available on Harmony site and maybe in other docs +Verified +Looks like my SNAPSHOTS for both were fetched at PDT +We should do those first and gain some experience operating them before radically departing from itCan you give us some data on the workload challenges you are facing? In our current benchmark we are not to a place where these issues are the logical ones to tackle + there doesnt seem to be a way of downloading the patch from phabricator +will indicate to stop processing resolution +ie +If you get it please copy it to jira and attach the jar +Attaching patch +Patch to fix the issue as well as tests that demonstrate it +sample project to validate transport functionality +Having them still around causes anycall to or to try executing the java tool +Handle icu driver in its own server so we can reload it when config change without reloading all the supervisionedited patch to handler driver reload +I am in progress working on this feature update and expecting to commit something soon +I will also investigate using faciliities in for good measure +I found an issue with the previous patch + +To be included in release +doesnt bother checking versions any +Could you enable modrewrite for so we can put our rewrite rules in the +Can you explain why publishing of TXT and CSV files works and publishing of XMI files fails +Because if no find the system it should not be this issue +Can we explain when one option is preferred over another? And then hand over to the docs group please +the rule a test case can only be run if it contains one or more Test annotations seems to the correct and simple solution to the problem +I dont think this is necessary but it looks cleanerM javaengineorgapachederbyimplservicesreflectAdd the Aggregator package to the list of Derby packages which the database class loader is willing to loadM javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangTest for this functionality +This seems like a pretty bad bug that causes data corruption +I expected that this problem is caused by in that this problem occurs when a bzip compressed file is splits +Its been rejected in the past for backwards compatibility concerns however those who need backwards compatibility can use the is far superior to +Thanks Paul! I was looking for a good piece of referenceJoe Leong +only use one lock + made it possible to run the in SASL mode +It is not critical moving to to to next in this pull request there are some UI fixes most valuable of them is the following When we openchange skin skin was shown first and then changes +Thanks Thejas for review! +This attachedfile is my testprogramBest regard +I hope this is the worst part of the mess +BTW with your edits the javadoc reads much better +Thanks Ari +I have found the caused of the problem +Will commit once Jenkins gives its +Changing back to since changes have been applied to trunk +Is that right +Added cleanup code to and uncommented listener registration in. +loadContext has been turned to in the meantime already available in recent nightly snapshots +Work around implemented +Do we needed that is placed to dist folder for out of the box multiprocess workingYes +I have additional though on this I know at least one good example of API where even the marshaller is the too low level API that the user isnt much concerned about +Committed to trunk and branchxThanks Nils +The character case will be fixed in another issue. +If we silently disassociated the transaction from the thread the app would fail +The +I need to think it through +and it is pgsql +Please dont add this to hibernate since it drags jboss common code which makes hibernate tools standalone m also not sure that the build system handles this yet +Any objectionsNone from me +The nightly snapshot section clearly has this text Nightly Snapshots These distributions are built and deployed nightly and contain fixes and improvements +bq +Suresh! +The concept of copyField is implicitly a judgment that a query of the merged fields is significantly better than the dismax query of the separate fields +Assigned for release +updated patches for trunk and for code looks OK for the most part +Left Open. +Id love to skip this altogether if there is a way +Jira still open +Actually I think the real problem is that it retrieves a list of all the products then calls the service get for each product which then goes and does pretty complicated stuffSo yes it does take a long time but it also does a lot of work for you +surefire happens in issues likely caused this occasional failures does not identify itself as a test is actually a test that writes to the repository and needs a different test root in the configurationboth will remove the test content in their tearDown methodFixed in revision to the branch in revision . +We cant have any dependency on JB since this needs to be used in the standalone distro should be part of our normal test suite not integration tests +Committed trunk r r rI keep this issue open for Mike to add the solr checks too +I need this fix in +Is the Resolution Wont Fix finalIf so I have to work my way around +Very nice to not have to impl twice everywhere +No problem +Attached patch representing minimally necessary code for respective processingUnit testing and according handling up next +Have you updated to the latest snapshot? It is likely fixed in the snapshotAs for the first exception thats worth looking into as well +Running patch through you for the commentIm modifing my source code +patch the thread can wake up when an exception is thrown and the code shouldnt mask that with are you saying that the real issue has been masked by the exception handlerThat makes sense since in this case the timings didnt imply a timeoutWhat was the real exception for this issue +I dont know if anyone of them already have Confluence account yet +The newer versions of Xalan catch more mistakesin stylesheet syntax so you may have encountered such a case +ACK patch works for me Running testsFinished tests in testss assertionss +The only change is that the access modifier on the test classes have been changed from package to public +Attach simple. +Patch for using Colt in is an important result although as Sean mentioned the current class is a so the results are not all that surprisingOn the other hand Colt has license issuesIn contrast MTJ does not provides even higher performance and is in the process of being integrated into Commons Math +When tasks are scheduled out of order then the whole testcase gets stuck as the task that is scheduled out of order waits forever blocking the cluster +How does this fix it Jimmy? I dont see your remove of our interrupting a syncing thread +now the TM will be able to take action only after xDeltaWe use updatetimers to prevent unnecessary multi assign region it just updates the time as the related serverComparing to previous logic here Delta is at most sAnyway multi assign is much more serious than delayed assign byAnd we should try our best to ensure that there is no case to need use assign regionCorrect me if wrong thanks +NB Changing the delimiter may be an issue for some apps but the +not +made this change on both and after release. +Patch to fix the issues raised in +Given that all a handler is doing is just printing a log using a single instance of handler should be sufficient and need for separate instance does not seem necessary unless different actions need to be taken for each signal +Translation of to +Thats why the import of the is supplied + Arjuna for fixing this +Ive confirmed that this is fixed in the RC after building it from source +LGTM +Should I possibly the issue when Ive added questions or commentsPlease see my previous comment regarding what information to add +I do not expect that its difficult to find consensus on these questions but I feel it would be good to have it written down somewhere. +ScottMy work on and is almost finished +is still partitioned by rowYes +One of the failures is resolved by someone else +Updated version addresses tlipcons review comments +Its not controlled by JBM but you will find more details with AS guysHoward +Could you please provide more details on how to reproduce this if youre still seing it +You may feel free to adminstratively move the thread if you feel this is a critical issue +Looks like the works been done thanks Danno. +This patch illustrates my idea +Will open a separate ticket +Also added similar logic for naming child handlers in a chain. +I got a bit further by separating and. +I cant open a review request is there a tutorial how to use itI get can upload the diff and give the project url as absolute path from apache svn root +Same issue +Patch that moves the code up to the new +verified in JBDS +Ive attempted to track down what needs to change to make this run properly and have been unsuccessful so until this is fixed Mesos is useless to me +Note I think we can still run all of this from the driver +I agree I even stumbled upon the same problem just a couple of days agoWould you consider submitting a patch we could then apply +New section similar to the system tables section in the reference manualThere is no public javadoc for these classes only internal +Is that the expected behavior or should the www be the one being rendered +Replaced to patch +Dhruv can you update by chance? Otherwise will remove this from +This patch adds a unit test specifically for reading zero length bzip files and well as another testThe patch also fixes an error related to figuring out the name of the output directory when determining whether this is a bzip file +ohh I just read through the javac options and saw the option to disable those path warnings +Inclusion of requires that the module become a dependency +Essentially it is octalsymbolic umask settings in a universe where sticky bits dont existComments Line remove reference to t in comment Line Add back reference to X being prohibited Line Extraneous import of These are minor and otherwise +Marking Release Notes field +Please file a new issue and attach your patch there and I will look at it and commit it +this replaces the static file with an automatically version of the same fileupdated room pom version all the Action filesadded call to pomaligned the file with the generated file so role hints where consistent across the board +Just had to wait few minutes +This makes much more sense I think +from p +Committed revision +Patch is draft +yes the actual injection is handled by the cdi container +I am using to do remoting but to configure it I need + +Thanks Hitesh and Sid for the prompt review. +Im willing to do the commit but just need the contrib test suite to pass on a machine first +Is my a defined property of class or is it open content? It looks like it didnt find it in the Types property list so it assumed it must be open content in the but didnt find it there either +Well spotted! I will investigate. +Then no matter which bundle is activated JAXB in GEP will be loaded firstI attach for trunk and for branchThanks +Great to see somebody picking this upThe place that I expect to see problems are any use of the legacy math classes +But I have concerns about the option +Feel free to review the code and let me know if you find any improper use of JSSE API +It uses for temporary storage of uploaded files for base decoding of attachments and saves the attachment map in a request attribute called The comes right from the specification and can be tested using the commandcurl headerapplicationxml httplocalhostformdata +It works well and all unit tests are passedThe patch includes some commented out code +This feature now implemented against CVS head +Good to know +to and trunk thanks! +I just committed this. + +Thank you! is the example login war for Flex and is the source that compiles and works against +Please contact your administrator to be granted the appropriate permissions +Better than GMT but not a global solution +d is doneAll done. +Id like to see this added in a future release +Thanks Runping! +Given this is incompatible the target version for this cannot be +Patches are welcomeI touched Ill hopefully send a bunch of patches later for those findings here which will stay without someones fixed note +Given that this only affects should this JIRA be closed +Committed revision . +I apologize for this +LGTM +Could we put it in a separate configuration that we load before but that does not depend on +perhaps geronimoopt? IMO plugin is packaging terminology and not indicative of whether or not a module belongs in servertrunk +We should at least implement all PHPMD rules They have really good background +I dont think this issue should have been closed +This version uses the Jdbm store +The word count matrix is extremely sparse +PR pushed to for setting proper fix proper fix version +I took lsof output and a binary heap dumpContacting you directly on how to get the data to you +I am concerned about making this part of the release that is coming up +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Apache Karaf is now availableWeb page is updatedUser and dev lists notifiedJIRA is updated +If you get started though I probably wont be able to resist but join in at some point +Removed conversion from +Would you mind posting a reference to more information about +Geronimo changes in rev . +This is on my agenda for first this Monday morning +Id like to get to know nutch and lucene code base better for my project +I havent had time to check on their log files to see if there are similar numbers of quorum events in the past and that the issue doesnt reproduce at all +Lincoln if forge wont detect these things automatically could we add some markers that it would pickup +Hansthank you for testing the patchThe edit button is commented because the portlet parametrization is not yet implemented and not even designedThe edit button should show a form that should be specified with the portlet screen +I investigated a lot about it before applying it and I combined the two patches provided theres no better way than using the documentLevel variable see first submitted patch and tested with test provided in the second patchThanks Charles +add one or we will have to reject it +Ill try to get JSF updated before the end of the week +Blackberry How can I get the fixed version? Im new here. +Sorry GeorgeI cant figure out what you want to make +So all issues not directly assigned to an AS release are being closed +Resolving. +Can you resubmitpatch p i +I believe this was fixed in the current build it came up when trying to debug a problem with WSSG +Please find the Log you use the default snmp marshaler class or have you done your own marshaler? If you did your own marshaler please let us see what you are doing with exchange and inMsg objects +For a query that has terms it saves Kquery plus its also quicker by about Dont double buffer +There are too many variables +Please attach a sample attached PDF causes the problem. +Jiri do you need anything else +verified in JBDS +If I would have known a solution for this I already would have fixedthe bug +Filesystem JAR permission processing is being tracked by +So to add a new edits log dir the dir needs to be added to the object and also add an edits journal to the to +Tests ran fine except for jdbcapi which needed to be updated for the new behaviorIll attach an updated patch +if you move the check to setup method rather than in the bsp method I would be +due some oddness in the servlet spec when doing a query against the request parameter map you cant specify the encoding afterwardsSo put the filter behind the trinidad attached a ready to run WAR and the source containing a very simple sample application reproduce the problem +We simply need to get a quorum of loggers to agree not to accept writes from the old master at which point it is effectively fenced for writes +I agree that this use of xsitype should be eliminated from castor xsitype should specify an XML schema type not a java type +Obviously plan creation is fast by does it take so long to get the iterator object first? looking at the code path I cant see where the time is the queryThe one at the start of this JIRA is not respecting timeouts for a different reason +Also the redelivery boolean flag was always TRUE even though there was no attempt of redelivery +This is designed to be as simple as possible the same service listener is used for the global listeners and the notifications are blocking on state changes +JAXP is already included in JDK so I think there is some clash going on with the JDK and the JAXP API jar +Youre right Knut those methods are just for DRDA arguments +I agree that it would be beneficial for performance to have iBatis refactor the way Eduardo suggested +So I will leave this open for a while +it seems that your patch is out dated +Access control can be enabled by Acls as in case of +We have effectively made these processes daemons +So what do we do in that case? +This could be done in itself and is pretty straight forward to do +CVS is missing the from libendorsed this bug should not be closed. +See diff where I am doing this insteadThanks +Checked fix into develop mouseChildrenfalse on the itemRenderer seems to fix this issue also +and then rerunning the install management server script +its intermittent for me +Try to refresh the project +Sorry my fault +Oops! Minor fix forgot to adjust a couple of references to the constructors for that I changed to support overloading as an optional feature +Having weird problems with cvsssh via the level proto +Closing this issue. +This should solve it +Karl it doesnt seem so clerezza uses mvn can work around the problem by adding to not very elegant but a temporary reto for the patchI can confirm the performance improvement +What is not required is dynamic installation during deployment or at runtime +I saw this jira notification and thought although error manifests in portal its caused by Webservices meaning that if a webservice sarwar is deployed in EAP it would cause the same error +Is it possible for you to test with a pure and make sure it works? I tried Maven but it broke so many of my plugins that I gave up on the build runs multiproject completes but no dashboard html output is generated +Fixed by updating to. +But looks like in the example Satheesh gave select from is getting cached and that is why it does not detect that needs to be mapped to temporary table when it is executed the nd time in the example +bq +I used the archetype +Witness your Ant patch is not very longWhat would be complicated is supporting all the possible variations which is why I suggested a template approach might be the best +It could look for reference from tablets and delete entriesIn addition to not handling bulk imports the tool does not handle the case of tablet dirs that are not referenced + reopen to be able to target correct fix version. +I had a go at implementing this but ran into a problem +Similar results with trunk on Windows XP +I was not able to find the fixes on the stage +Error output get using commandmvn test Dtest +I just committed this +Let me know what you think +Indeed you can easily get sources andor javadoc of one dependency but it doesnt work transitively +Requires the framework to be present +This means that the ICU collation stuff is currently not useable within OSGI +it is lacking some formatting updates +We should probably try to detect the endless loop +Reverted changes to string concatenation +We may not need to add a helper class for the user +I dont know if this is the same problem as jlages and to me it looks as a different issue +Sorry for the confusion Jim +back port the ported the fix from +I was having a look at this maybe its time to finally move this contribution to the sandbox +It is a minor interface change that will hopefully become the standard +Also modified to xml format to ease its use with accented chars +should be correct conditional task is not a single job but n +patch commited +If we could make the threads better we could commit this patch +in r +and also back ported to all our release branches. +Perhaps we do have another limitation but in that case you must provide such kind of reproductible scenario +Should be faster now and no longer so high invocation counts +Version of my compiler is +good points I didnt look at deleteXxx since I dont really need it +I am using OS X cannot reproduce this with cocoon winxp sp latest Firefox winzip beta +Yes i did +Thanks for the patch +Id prefer if you could raise this issue in the mailing list so that answering for each and every point would have been easierAbout SOAP classes packagingLet me explain the rationale behind the current Fault implementation +At the very least it doesnt break any of the existing functionality +As long as users understand that when JBT Core drops the SOA Tooling stuff may or may not be availableAdded complication is that when I move the development site from say M to M Ill have to remove the previous SOA Tooling stuff if its not compatible with the current Core stuff + +I guess create a fake user with that role? +Could you upload explain result for that query? It seemed to be a genMR order issue which was thought to be resolved already +The patch looks good to me + I like it +Those methods were needed +Fixed on closing all issues in resolved state that have not been reopened nor commented over in the last month +Here is the patch i applied a bit reworked to factor out some duplication +No and no documentation +Closing this ticket as duplicate of because it has attached patch. +to trunk. +moved of because of cleaner patch for to patch attached +This is typically a long operation that you dont want to do too oftenThe can be used to store the mined collocations in an indexPossible uses for collocations are automatically identifying candidate terms in a query that can be turned into a phrase query better spelling correction by using all terms in query as context to pick the most likely spelling variation Havent done too much with this code but Ive added it here because it sounds like it could be relevantCheersMark +of controllers +Will be fixed when using same version of rule engines as the one on Sonar Server +Correction on option Implement a Joda date formatter using the spec in the wc profile so only that format is supported +Thanks Brajesh Done at rAshish Vijaywargiya +Added a code tag +I want to be able to access beans from a new thread. +Tested along with updated unit test +Hi Eva one more for the Install Guide +That is the reason why I did not comment to that effect in +successfully deployed a version of the with the release of +One of these bookies will have to create a ledger descriptor when the client tries to replicate the single entry +Can you go a bit more in details as I do not really understand +Thanks Ram +I have set this issue to block further progression on the module upgrading to writing tests for the full module documentation etc +First everything is reformatted so we cant readily tell what actually changed in the files with svn diff +Sergey am I right in my understanding that the original issue is not reproducible on SLES and we may close this JIRA as fixed +But no testcase yet +Thanks +FYI If S keys with are stored in etchadoopconf there is no problem +Running this through Hudson +They previously added a MS Chinese font in their repo but later removed due to copyright issue +Hi John after having deleted the project do you confirm that this error vanished? after I delete the project I can rebuild the project this error is gone but all history data are gone too +I would use it in some other places +AntonioI will be taking a look at this +Looks good to me +I will move to a native driver eventually. +After I successfully ran wordcount and terasort examples +I am not sure this is a good idea for the last table unless you stop streaming you never know the contains a fix for both and +At a glance it looked like it continued reading at wherever it left off in the stream but I havent looked at it as closely as the contributor and wanted to ask after its behavior +But offhand why cant node in examine the znodes and perform the right actionIn any case a design that relies on state transitions is bad +Im going to be putting up new patch shortly to resolve this +Ill start the IP clearance process so we can get this in ASAP +Patch that maintains backwards compatibilitybut allows members to be added by their principal name rather than the path to the principal +Whoops +we now implement the and CSS propertiesThe attached document specifiesgeometricPrecisionwhich implies that the text is aliasedRemove this property or use another value to have antialised text. +Thank will try itI need to explain the binary package is build using mvn +Do you want to contribute a patch as well? Its not hard the file is sitesrcdocumentationcontentxdocs +We dont know of any compelling reason why we might want to move +Someone trying to change the history might know about its effect on or +Are you able to deploy using GAE SDK then +Thanks Jeremy +If in the future someone would like to tackle this feel free to open is and attach a patch. +Do all test have to sub class a DBMS specific Base class likeIt is hard to imagine a test that does not need database functionality but if you have a unit test that is standalone and does not then no you do not need to subclass anything +there were to be another piece of code that inadvertently calls this piece of code once again by creating additional Pattern objects would be suicidalthere is no other piece of code currently calling it inside SolrThere is a tradeoff of complexity Vs +Im working on this +bulk close of all resolved issues prior to release. +See with a related change +Is there a reason why this is not the default behaviourAlso I did a search on google for this property and only found similar articles and forum threads about people having the same problem +So would be abandonedI think more effort should be put on that JIRA +I think this is a duplicate of which has been resolved in +Youre need to be a little more specific +maybe it could be in test code to be moved when needed but Im ambivalent on have committed this +you cant really distribute this entirely +tgz +Linking this bug to the jira which tracks the rewriting of the parsers GROUP BYHAVING logic +Thank youissue closed +svn statM javaengineorgapachederbyimplstorerawdataM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTestsutil +Unsetting Fix Version on unassigned issues + Such a implementation could also be used to better isolate the current caching logic behind uncommitted changeswouldnt that ideally be the transient space +I already have a fix for it as part of my patch for and after some polishing and a regression test it should be good to goI expect to get this done during the coming week +comments on you update the documentation +This results in that during execution the parameter types for the last statement will be used for all statements of the batch +Works fine now +The dist target now invokes the dist target in langcsharp +To clarify a little the plugin thinks the server has started and displays this fact about seconds before the server finishes loading and displays the Geronimo server started message in the console +Generating schemas on the fly when writing data is a fine practice too +AFAIU one aim of close is to release eg the instantiated instances via releaseInstance +All of the metrics templates look good and are well aligned +There are spacing diffs +The image mosaic needed to reproduce the be a bug in then + does not support testing in a clustered environment +We are not accepting any further QE feedback unless they are issues that would result in data loss +The only additional feature I would like to see would be the ability to somehow embed more complex data types inside a row +Im okay with that consistency seems like a good idea and as long as the general functionality is still there we shouldnt have any issues although it still depends a bit on the expression language stuff +Upload a new patch with some code cleanup +reworked file with refactorings reworked patch makes Expression extend but only populates and visits reworked version to trunk + fix generatestore jobToken at fileFixed one other bug alongside file shouldnt be written to HDFS under JT lock +Thank you for the nice new feature Riley +Closing resolved issues. +Patch committed to SVN +Robert could you please clarifyI just checked out the tag looked up srccouchdb and found a filterview function in line authored by Benoit and Filipe introduced with fAre we talking about two different things? If yes this should be made more clear +The original data goes through a processing stage and the datamodel only saves the output +The disappears but the created willbe wrong +But locally on Windows it runs +rebasing patch +I added this feature but according to schema only users should be returned thus groups are not put to response +But I want to see the devpmc support for a change in a very long standing policy to stay out of security code +Driver Driver and Driver are removed +Does that scenario work for youChris Mair +Where is Junit tests for this issue +This is not true +Fresh screens for JBDS Features on are made and sent to Svetlana +here are the log events of the last hours type of thing +The second patch on indeed solves my problems partly +By the way I can also build a private without the XERCES and XALAN XML components +mvn sitestage tmp wasnt me And I want to have the full site with all the files but I thing it can wait till next resolved +Now the test has no data loss +Franz thanks for the contributionCan you help with documentation also +bulk close of all resolved issues prior to release. +Thanks Knut for that suggestion +I will attach my unittests for others to try in other environments +Committed thanks Michael +Meanwhile we will stick to as an embedded serviceFyi below are the changes we detected so farThese are the classes in that we usedimport import import import import import import import import import These are the new classes found in the newldap packageimport import import import import import import import import +Committed to trunk +Fixed assigning to Eddie for review +If any one of them are not true the AST based approach is used +I can successfully add custom repos to the profiles are all processed after the project is loaded so there is no opportunity to add repositories for parent pom resolution +Ive added into SVN trunk support for the branch attribute on ivyresolve and all like ivyretrieveCould you give it a try to check if it solves your issue +Ill try to develop a patch that uses subprocess only on Windows and popen otherwise +So that may be the reason that Hive implemented its own failover +It is not enough to protect from all known scenarios +needs a small fix that will be included together by Namit +Committed in abafacebfaebbdee +Yes I tried with a sleep in the test and it helps Attached is the patch +Now added to X branch and tweaked as per recent comment +Hadrian I will hold it until we get the conculsion about the DSL +I committed the patch to andThank you Brandon +timestamps Proposal remove timestamp file when corresponding file removed from deploy on trunk revision +Committedx branch in revision trunk in revision close for +CXF does that +Fixing checkstyle complaintsI will let this patch sit for another day and commit it if no objections +committed minus the unnecessary changes to worked before and I dont think its unreasonable to seed a single node with keyspaces before bringing other nodes online +Any reason why the following attached patch wont work? This pattern seem to be used in the other js files inOne of the issues is that inside there are references to which it expects to be jQuery +patch looks good +Could you reformat your sources to eliminate whitespace differences and regenerate the patch? I cant figure out what was changed from your diff against CVS HEAD output seems to be better now +So perhaps for DB just set it to validate by runs on an IBM System i and the way the majority of the tablesfiles are created is using a source file that gets compiled into the tablefile +Am i required to work on this and How can i help further +Sounds good +Expect a Context that supports creating missing subcontexts in bindDoes not include davids proposal for the fix +You should run this target if you have source code upgrade or you changed the source code and you are getting some errorsEvery time I do a svn update its a source code upgrade and I should do a clobber +Thanks for the feedback +cache hit or miss +Unassign issues assigned to Nate +It will be available with the Roo release. +all subparts merged need to setup and fix navigation pointers +You can tell the two apart because CNFE uses a +I remove my reservations +Im testing it right now +Changes to pigContext to support querying and instantiating functions registered using in +HiI dont know how much it is used +Yes I ran large jobs with maps and reducers on nodes and saw no performance issues +I hope Ive given you all the information you need +If keys were and URL included any of below characters then we had same problemTrue but the problem didnt really rear its ugly head until the keys were completely binary data +How was the developer defined in the pom +If need I create a topic in jboss ws forum please let me know +Committed to trunk with revision I dont expect more work to be done on this issue +This change should have been fixed in though itd be best to test on and then let us know if youre still having trouble in a new JIRA issue. +In fact the result should improve with more accurate versionsDo you have something in mind where this assumption breaks? +Needs a merge of one of the methods with that in +Please ensure that the conf file parsing works even in face of invalid inputs in the conf file +swondered this is certainly something you want to have we have different plans +use properties view or wait for the Details + This looks good to me +for some you can recover for others you cannot +Fixed on revision +Here are the files reflecting tutorial fixes source XML and output PDF +Assigning to Ondra as he is author of this testcase +New patch +pushed for QE to verify since this is only about the client library +please go ahead and commit the codetiaRainer +Uploaded patch +Hi Arun Ankit AnilId like to close this issue if nothing is still pending do you agree +Moved from to just after merge of old branch into TRUNK +Any plan to backport that in +bq +I have something such in the migration tool +Hopefully will remove the need to do this +Will try to get to this next week +I dont know what has changed between and +Dont quote me out of context +If we make all this throwing exceptions we almost introduce a schema here which makes lucene very different in terms of RT behavior compared to previous versions +And also improve to return meaningful return code rather than default return code +I dont undestand what unscoped is really + +Whats the advantage of doing this server side? filtering and aggregation make sense because it reduces the amount of data that has to be read over the network +yep looks fix +It was something i introduced in as an extra check even though it was only called by + +defer all issues to +Not sure what went wrong but the patch I uploaded was slightly outdated +As far as I can tell know well be shipping this milestone build before xMas +Any errors or exceptions from the Eclipse log? It sounds like a duplicate of Callisto WK +Radim youre absolutely right + is now completed OK rev +Alright I went with the straight forward for resource requests extracting a doDispatch method that can be overridden if somebody really insists +I personally think forbidding this also improves the code quality since it makes accidental shadowing more difficult +Updated the pom to exclude on Solaris and also adds a couple of tests weve been excluding on the Fuse CI +We have had support for since before so I am a little confused on this one +Is there a strong need for this feature +Im good with the fixes and explanation. +check if this ok now? I added the documentation to the component list +If there are any changes that have not been applied please attach a patch that fixes just the items necessary to complete this issue +Done is using the API +Rev ports rev and to trunk to form the foundation of using the release plugin there too. +Surely the jBPM plugin should follow the standard Eclipse behaviour? In which case the files are made writable +Thats related to WTP update site +It still seems easier to me to throw a no work found exception and have one point to say means the job was successful via configuration +Attaching diff file for packages +patch attached to match behavior of sequential cluster this definitely gets rid of the inconsistency in the outputs of sequential and mapreduce versionsI will commit it once the build becomes stableThanks for your contribution +I dont quite see it as an enterprise feature with the strict sense +When I checked this morning it was still in couchfile as a byte offset instead of +applied in revision Anton please check that it was done as is OK thank you Mikhail. +You dont know without additional history that is basically equivalent of +Tests failed because region servers did not shut down +I dont think this jira requires all classes to have the annotations just the important user facing classes +Same preliminary version rebased against trunk and without generated for atomicbatchmutate method +We should fix this issue before but to include rspec in jruby have two options update to latest rspec for or dont ship rspec at all +thanks +I also ran about wordcount tests after restarting the RM +Applied changes +Feel free to if the problem comes up again +However all requests from a web app will have gone through the web app and it will have assigned a default identity +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The issue is valid for ajax sorting and grouping of table columns +Catching NPE is problematic and it looks like is avoiding the problem. +I dont care either +please add clustering configuration and specify building from in the ref doc add a topic where you can add the helloportlet with the jboss eclipse ide to use it as a portlet dev tool for jboss +Merged Pull Request +This is waiting for Dalis annotaion support. +Im more concerned about these tests being racy and flakey due to the noWait variable in the shutdown method +Yes +They should really by part of the QA setup +Id suggest that you file a separate bug report for the issue youre seeing +I am not able to reproduce this with the same config as you have provided and the latest marvin build +But according to Intel and Architectures Software Developers Manual Volume B Instruction Set Reference opcode used in this patch also works on +Are you running in a environment +So as a workaround I have moved them to within a unix script +Unless there is a strong objection against RST given its lets implement the documentation format for Sqoop on RST +Yongqiang I dont quite understand your comment +Changed the test datasource to be unique across the JDF quickstarts +Please upgrade or patch you code base. +But that takes much more work to fix +the reference doc is incomplete compared to the feature options we offerCould you list what is missing +Paulex verified +If programm is launched via shell then there will be no error +Added it r +Yes Evgeny that looks really good +The changes went in as revision for backport analysis +Ok I will provide a test web app as soon as possibleYes I was talking about +updated to include test for parsingsetting workspaceManager +NS will take care of deleting them when Autoscale policy is deletedWhen you delete the network of testb account is the LB server rule servicegroup which got created by autoscale policy created from testa account are present +Two arrays would indeed model all the C sparse vectors I have met +Thanks a lot Todd for the commentsI will check and update accordingly +Actually compareTo doesnt care if they are decorated or undecorated +unscheduled issues to the next release +Is this true +I reviewed the documents and added text +for different in general +Can you please provide more info +Hi Rotislav can you please direct this to someone from the ejb team? The WS test here basically relies on the EJB for the appclient testing +But obiously there are other ways of doing it which are not covered in your book +seems like we have a good bit of information about sizing already +verified in current stage of installation guide +Ill keep trying +Have fun great +Temporary fix is to replace in and with the version of this class in +markus +Property is properly set but properties and are not setValues when property is set via mavenValues when property is NOT set via mavenError displayed in JBT +Thank you. +Unfortunately there is the possibility that the previous VFS is still in use by a running classloader so running close on the old before the new can potentially cause some devastating issues +Also this file has been checked into svn so that the next build will pull the update +I thought there was a clear selection in the edit menu +Few fixes for compile revised patch +Trying to help get the unit tests working +The sample has the same issueFix committed in trunk and branch +it many examples with proper closing of resources +So there is really nothing to doYour initial use case was sending an event from a page in a to the owner page which in my opinion is custom use case +A more generic framework is useful in addressing encryption in components other than MR +Refactored and Dirichlet to use +Instead we can decrement in the try block after the merge completes +The plugin is working great just a little bit different then before +I not keen on that as they are really old now +Perhaps its best if we push this one off until or later +fixed in rnote bad log line was actually in not after release +The c client updates it for all operations and java not +Marking this bug as resolved as other bugs are tracking the real issues +otherwise it is totally okay to have classifiernull and append there goes two days Ill never get back +Thanks guys! I figured that out too +bulk close of all resolved issues. +It turns out the default value for ignoreParameters is true +patch applied +If you want to use datasource as the name then you would manually delete the property +I think I applied all the suggested unit tests if not please add a patch +Interesting +compress ratio total block before size after size +be done in the newer fop block see +Leo patch applied at revision r with some changes thanks a lot for this enhancement I add some error handling codes to it so please verify everything is OK +Galder the should detect that the prepare call thread has a Transaction context attached to it and should be able to look up the appropriate from the +As we emerge from the panic of our first release well pay you back with more contributions to the community +why this issue is not detected by our test suite +Please let me know if you experience this again and we can look a little deeper. +Ill work on getting this polished up and committedWith regard to the numbers I would think that this is simply going to be a bit worse because it is now dechunking and chunking for transforms +Derby should not be printing stack traces to it shouldjust throw an exception +Committed revision +That type of information needs to be at the beginning +Note that the patch introduces a dependency on will be JDK only so we target this for Camel Camel is JDK now so can we integrate this patch + +Jobs that shouldnt be recovered should set to false +Patch attached +Suppressing findbugs warnings +It is corrected now +A previous lack of test coverage is not a good argument for changing default behavior that late in the release cycleBefore I agree to this change I would like to see discussion of two questions What and how many applications might be affected by this change? Over the years weve told a lot of people that it will only work for relative +I noticed the exact same bug about a month ago null getting injected instead of a CDE thrown +Thanks +Actually if theres objections of any kind please feel free to reopen this issue +If there were a race condition in I would haveknown +Can anyone tell me if this only happen if you are accessing the same object in two different threads at the same timesession is of course not threadsafe but invocation of proxied methods should be correct +Nice catch +I have implemented a which XML elements before returning response to the client +Ive the same configuration OSX and it works for me with with Seam Booking Example WAR to reproduce this issue one more time cannt reproduce so close as cannt reproduce issue +We could have rdebug launch with something like jrubyd instead of jruby which would imply debug gets passed +Thank you +Verified at +Thanks! +Thanks Mahadev! Yes and all tests ran successfully +corrected names of attributes and components for toolTip and names of attributes for aj name of attributes for aj is in progress because of the enough amount of RF attributes names for support and dropsupport components +I dont see other reasonable way implementing permissions setting +Please verify again as it could be fixed +Rather than committing this patch you should revert and commit it once it passes testsWe should either maintain the Avro RPC code or remove it if we dont intend to maintain it +This would allow s ls to check isLink and issue get if necessary +Add that contains a query specific configuration used in and Task +I dont know anyone that runs sets of tests inside their IDE using the junit category feature +I think that the interfaces to the speculative execution are integrated enough that this would be difficult to get right +Well need to determine impact on the docs this can be postponed if needed +Applied svn revision thanks to add the right Fix Version +Nobody likes that rule much but its written into the SAX specIf you want to know about errors you must register an error handlerI dont know whether Xerces comes with a prewritten error handler that you can use for this purpose +patch contains fixed typos and added clearifications moved filling of hashmaps to init method updated request +in JBDS h Fedora +Im going to upload a patch with all of the admin handlers changed to turn off http cachingAny objections +Would it make more sense to have one truncate command and have an optional parameter which preserves the truncate? Rather than having a separate command for truncating and truncating while preserving region keys +I adde a few Socket constants in so please retest with to see if its still a would also be helpful to list exactly which constants arent supported +implementation introduced a bug in the minmax behavior +The patch has been committed to the branch too. + +Subin Still not done with the review +The reason I unassigned myself is to signal to the other committers that Im not actively working on this and let others step in if they wishThis is the way Apache works we are all volunteers and I am sure that with some patience this will make it through in time for final +Also for my husbands account could you email the reset password to his username at gmaildotcom? Thanks in advance for your help +I found a couple typos in The default implemeantion retruns Default ratio for compound file system useage +HiCould you test your server via remote desktop as localhostMay be problem in the internet connectionThanksVasiliy +You are right. +Committed patch in trunk at revision Thanks Shrey for the awesome work again! I am just loving those DOJO pages +I think the bundle info table should be pure html table instead of clickable dojo table widget +Highlighter has already a test for and the build dependencies are already in the ANT file +ECMS is different than a portal + +objections +None in CDH either nor in Yahoos one atTheres a available in however +I get two web services defined in the same project +Documented as a known issue in the EAP release notesLaura I checked the release notes of eap but found this issue is documented in the fixed issue sectionthe fix for this issue isTo find the join column hibernate compared the referenced in the with the referenced tables columns name but as the referenced column name is quoted so actually it should be compared with the columns quote name this has been corrected +I was able to change the locale browse to a folder select optionalrequired components and finish the installation without having to use the mouse +Description of this issue updated +Those threads are created by the default task runner factory and this factory is created once and then never shutdown throwout the life time of the process +I am trying to see if i can mock the integration test once i take care of it so i need to know what he scenario is +Looks good +Commit fbdffaeabbca in issue is fixed and working fine without any issue + +HiIm resolving it as wont fix die to the fact that the issue youre seeing is most likely related to What do I need to do to get this issue fixed? How can I use servicemix and deploy bundles that need to have different resource rootThanks. +I just made a second variant of my original patch +Attachment has been added with description Test file showing that the removal of the forced breaks leads to normal P open bugs to P pending further status from ASSIGNED to NEW for consistency +Previously I hacked apart the existing web ui to output information in XML or JSON this would make much more sense +Then tried the same segments that didnt work and now it work automagicallyThank you for response and a great work! +Hi Lily Thanks for the patch! is marked fixed but I dont see how your change could have impacted that test +This issue has been fixed and released as part of release + does not exist in current tools relase +Closing since theres a workaround. +Id say so! Talk about a quick implementation +Im building on +A user accesses an application for the first time at path xyz +I can confirm that this is indeed the no longer running when the clock is pushed backward far enough +You got farther than I did I couldnt even make the app generated by bb compile +With this there is high chance of getting notifications in the same thread and invoke the pending operations in the caller thread itself +I have two concerns +I have logged as a vehicle for carrying a release note describing the change to the CARDINALITY and PAGES columns +Its a big progressI fixed a small java doc warning and some small java doc flawsReviewed and pushed +I think you may be seeing we dont do sampling +The problem with is sometimes we dont specify a key since we have have Min token and an intrinsic Max token +Im told the patch did not pick up the registry or key file so I am attaching them separately +This is already fully functional patch +The mapred package is left unchanged +Pls review the patch to check if it feasible +bulk close for the release +I suppose Li is back to course work and nobody will pick this up in time for + makes a lot of sense +I have not tested the actual branch creation +Moving to +Any thoughts on fixing this for series of Hibernate? +This issue just states the file size reported is wrong IMHO. +Other potential candidates for this wizardization areStepping users through engineering scaffolding and other high level concepts. +Adds deprecation hooks +Thanks VasilyPatch applied to SWING module at repo revision rPlease verify it was applied as you expected +Would it be possible to have the build process output a file in the packaged jar that includes the version in it? This could then be loaded from the classpath? If the maven was included in the jar it could easily be looked up from the folder since maven includes this by default +Are we sure those fall back in the same category as the layer preview the demo requests and some random request builder like the wps one? Tools sounds to me a new category not the new name of the demos +Patch will follow once I get some papers published for my employer Have a great weekend. +ID like to point out that the stacktrace does not seem to point to the exception ocurring within application code +testsuitesrcmainorgjbosstestmdb +Changing to bug and fix for resize problem +Thanks Govind! I committed this to trunk and. +fixed +Good to know there arent any objections +What do you thing about this approach +jar +As far as I see it whatever was enabled in JBDS GA GA was doing the correct thing +Can you PLEASE split that up intoThanks +Or perhaps we have an instance of EAP somewhere thats public enough I could ping it and try to reproduce what youre seeing +It is default mechanism +its a massive gap in the GWT platform not having a great JSON library marshalling efficiently to beansIll give you a shout ifwhen I get chance to hack up a little prototype we can tinker with +Ive attached a patch which insures it doesnt happen on the public methods ofIn fact I think could go even further +Many thanks JohannesOleg +the main issue is that EE deployments of JMS do not support message listeners +Thanks Anilyour patch is in rev +Perhaps just DB then? the context will always be clear since this isnt a component being mixed in with others +Do and pass with this change? They are live tests run against the real S service and are not run by default +Im trying to get this resolved in with Denis Golovin and I would change default url back to httplocalhostwebdav since the jboss tools should target to this url and for JBDS it should be override it in product config files or some other ways the preference default value can be configured in file but I didnt find it JBDS install directory do you know where to configure default preference value so I can test it? if you look at you can see my proposal and patch that does that for JBT +oh and would you know if we need to keep the wsdl file in wsdl where it was put initiallyIve looked at the original day trader sources in geronimo and they do include it although I have no idea if this is used at runtime or again +Initial values are set +Looks like is a better place for that larger fixI would agree and changing path mapping can be done independent of any newbetter representations though that makes sense to do too especially if we can cut the cord to the old types given the singularity rationale +Actually quite a few tests in the mapred package are functional tests than unit tests they do some sort of validation in the sense that they all launch a job and verfify some functionality on completion +In that case could you tell us the way to do it? will you integrate the patch in the trunk +change default log level to INFO for this my quick review the categories below set to WARN removed most of the noise for each client request +regenerated the patch tests added as this is a change to shell scripts +So please start with a simple patch first and get it working +seems like a good idea +since you only installed the seed data in the base delegator when you specify the DEMO delegator it cant find the correct tenant recordThis is not a bugThe command you are trying to run will populate the tenant database but it will do nothing to add a reference in the base +Apparently switching to the Griffon way of doing docs the problem disappeared. +The patch has been appliedPlease verify +Can you say a bit more about thisIs this something you intend to doOr you would like somebody to doDid you want the result to fit into the Taste architectureor what +patch applies to clientjavasolrjsrcorgapachesolrclientsolrjimpl! this is great +which is +CDI integration is designed and implemented both for the RI and for Infinispan +It would be great if we could enjoy both worlds +Unit guess I can acknowledge what you are seeing +The Hibernate exception classes need to be available at the client not because of the handling but because of the handling of all other which are included as nested +Ditto for Scenario diagram. +Ill try to get to it soon +Pilot error +The features are still from +I realize you are following conventions already in place from the existing patch but it is better not to place resources in the root of the package structure instead place it parallel to the class that requires them +public static final int TYPELOG If any other action was expected of me here I sincerely apologize and request further elucidation of the action suggested +Please reopen if problems persists after using it from cygwin environment. +is it a configuration parameter +I tried both and Tomcat +Theres a timeout on the latter behavior +Why is hdfs block distribution method removed +MDT creation time is effectively identical +We should use the from instead and do we really needI think we can go without an completly and just really on +I committed a fix for this +Patch looks good! Will commit shortly +bin reproduced the same issue without any commands the console did not even appeared +for using an interface type from a dependency package +So what is the explanation +That was my fault +Thanks +Based on conversations on IRC with Mark Offset is used for nested patterns and index is the index of the pattern within the tuple +Thanks! Daniel can you elaborate what you mean about this being an IO issue and maybe open a bug for that if this fix did not resolve it? +Also unit tests dont use hod +And finally Patrick where did you get the file from +Im currently trying to find the answer to these questions Is it correct to store these objects with Java serialization instead of using Derbys Formatable logic? s serial has changed many times before +This should go in trunk as well right +Hey YongqiangWhats the status of this patchThanks +Attached the patch in fileAttached the steps to reproduce this bug in Gettting am really concerned because as a new user i am unable to download it +Yes youre right +These changes have been remerged +masterae and Charlie +Since we have switched to +judging the riskof unexpected regressions is very difficultmaybe we should target instead +Rebase on trunk patch was stale +Im ignorant +But it hasnt happened for quite some timebq +For stateful restarts of RM this will be used to build retry cache for requestWould it make sense then to generalize to a common file rather than an inner class of +Ill see what I can do about that +Thanks +Rev for das tools maven plugin and some modifications to tool it resolved and further enhancements based on future requirements +In order for the client to work you should have logged in as the target user who is doing the authentication I have not figured out how I can do that with my test machineI see it makes about calls and they seemed slow comparatively +Hi Jacques Hi Sam i modified the get method +Is that correct? Does the cache thevalid returned credentials +It was only when I used an explicit port number that things went +Default config now just stored the events using the +Hi Bartek patch looks good to me thanks +Jeff has not started on this one yet postpone to next release +Steven is working on this +The output of the new testcase appears to be nondeterministic +Manually tested +HansMridul and Jyotsna have worked on other jira issues of SFA +Shouldnt take more than minutes +Please could you verify that it has been applied as expected +Reopened so correct version can be assigned +I just tried to clone the url and got the following errorCloning into ravewarning remote HEAD refers to nonexistent ref unable to checkoutAny ideas +Lukas how about attaching the patch you have so far and I will make up my mind whether to commit anything in to trunk or not? +If you now refresh the browser you get doubled ugly charsIf you change the encoding to then Click the Show button +Will be fixed in next and jettison dependencies have been declared as provided or test +Maybe at the end of the next week +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +is working and the deployundeploy readme instructions are consistent with the other deployundeploy now consistent but the example message is missing at present and readme is required +I revired the code and Im very sure its relatet to it +Leo Ive modified your patch a littlePlease verify. +I took a look +Updated +Should be a small change +We used to have an html output one which was useful when you could type in shell commands on ui and get results as an html page +the green dot highlight the extra memory allocationsprofiling done on looks good just testing +Partially completed +Thanks +Thanks Rich for the patchI added an unit test to as we tend to like stuff being tested +One line change that produces lines +This provides a complete set of files for all the subdirectories under the travelsampleservices and travelsamplelaunchers directoriesWith these commits the compile and run targets of the files under contributions services and launchers all run to completion +Fixing it should be trivial if someone wants to give it a shot +Will commit it after running tests +yes sorry I left instead of +Change the addResource order Use HADOOP env to load some Job parameters +Where would you suggest I plugin to exploit my extra index information at Read timeSorry if this need seems exotic but I think we can get away with casting this need under the Extensibility Design Principal +Will investigate +gif used in the creation of percentage bar charts +Cant reproduce +We can make the current code less complex wo a flat layout +Applied patch to resolve issue +Checked into develop branch +ER verified +Checked in my first version of the +Its rather difficult determining which input elements use src and which dont another wonderful decision from the wcWhy is this marked invalid? I see the src attribute defined in the html spec but not for the struts tags +Furthermore userscan now use exnodeaction to specify more actions on a node other thanjust rebootnode +As Ted suggested can you measure the performance gain in a real clusterespecially when therere multiple scan request on one and more tableshaving a table like numnumTables with the performance numbers it will be nice +Fixed clustering configuration of stratos controller +Can you give more context here? I can tell from the stack trace that this happens during signin with Facebook upon handling the callback and while exchanging the code for an access token +was not found in the mirror when Ambari automatically install nagiosBut i didnt install nagios at all before +I also had this error +INFO Running for rows took ms +Im sure we could wrestle up some interested contributors form the +I can add this functionality tomorrow if the idea is accepted. +implementConfig +I will use svn diff in the futureSo what is the time frame of? Like I commented in the other issue when will get out of incubator +Only resources stay in SurvivorFor the sorting issue If the bag sort is done correctly that should be good enoughIn this JIRA I just want to experiment something +Moving out to move back to if this will be work on soon +I tried very hard to ensure that there would be no crashes if the something was not as expected +A patch for this JIRA + +FYI TC is the unique identifier for one of the Castor JDO CTF test cases +Any idea when this will be done +This would surely be a problem +The commit log code faithfully checksumd and wrote the data out as if it was valid + i still need to add lucene based filemgr query action +Testing +Reopening to consider changing this functionality in +If we allow a universal configuration to skip project references why not just allow the possibility that someone would want to selectively exclude a project reference +Your request is a good example +These changes should be merged back to added the patch to the trunk yesterday its now working again +Fixed with formal types to constructor instances to match the arguments of the node at the pass of local class remove +Committed to branch in revision +Thanks Inder! +I just committed this. +I hope to be wrong +bat resolved issues for already released versions +It sounds like you are defining an attribute as a URI which isnt a simple type so Axis doesnt allow it to be an attribute +As Heiko says it would if the id is changed +End Original Email Then David Jencks replied Start Reply Thats a bug +Hi KnutWhat about something like this for replacing +Even using an emptyMaybe it is something related to your training dataIf you like adventures you should try to debug the code and check where exactly is the problem +Ive modified RTRIM +I thought this was because my repo was stale and tried again editing the patch by handStill failed at revision I edited it again and now its at revision Needless to say it wont ever upload the patch and it seems to keep increasing the revision number +You would want a deployment exception since it looks like a name is required +h and the code in to remember the cookie state in the curl structure and to call curleasysetopt as needed +I will fix this before GA +Ill add a guard to handle this more gracefully but in the meantime you should be able to work around the issue by just deleting the concat style +This was fixed some time ago +this makes it impossible to even try using seam could we please downgrade the error about version number mismatch to a warning until we get proper support +Please tell me if I am wrong I want to see how this makes a formsecure from unintended input +Report was based on a snapshotAll is fine in the official release. +Very Best Regards George +Closing as a application can start in AS if a datasource is present +Updates to all shared composites to better handle duplicated code and per focuskeystroke updates +I will leave the issue open until we reach consensus on the rest part +Lucene is keeping everything in memory and adds new document to it +xml using is the using everyMy fault on the config + +Simple test app and its generated I was also quite surprised that there were no default +Sure I will do as suggested so that we can minimize the changes +Hum thats what you get for not even compiling before attaching a patch I suppose + was also cause problem properly shutting down a to port all the previously posted revisions to trunk +The full suite ran cleanly for me +But seems like Jenkins is not coming back soon +Any progress on this? Patch had been available for almost two years +The message labels are optional from the XML point of view but not from the Component Model point of view +By default it is using separate folder but it can be configured to use same output folder as Maven uses from properties on the Maven Dependencies classpaht container that you can reach from the Libraries tab on Java Build Path properties page in the project preferences dialog or directly from Configuration or Properties action in popup menu on Maven Dependencies container. +Any ideas here +nightly builds reflects all of the changes in scope of this issue from what I originally proposed and passes SVN trunk of JDO TCK +Which release will contain this fixThanks again +Not sure if we can do this +So we shouldnt change to much in freemarker just because +Upgraded from beta to final in revision . +I didnt took the time to test it under Windows +My for commit +Mojarra isn t that restrictive here +This fixes the subject part of you regenerate your patch from trunk I cant apply it +This allow subsequent constructions of physical plans to use the original eval spec again + Bug has been marked as a duplicate of this bug +The stack trace seems to indicate a severely mismatched state +Updated patch to generate the id on the datanode +I wonder if Dibyendu plans to work on this any further +That is we shouldnt bother archving anything that says nighlies unstable or RC none of them are obviously releases of any kind but rather snapshots for use in testing or voting onlyThe rest we can move +Compiler should check the constraint self n and realize that it is not implemented and give an error +The format issue has been brought up enough times that it is clear we need to provide a means to override the JDBC escape format +Usually you want all the columns in a single update to have the same TTL if you dont you can always break it into a batch +The patch looks good +Moreover it is not backcompat +Cassandra is throwing a which is bad because this is a totally recoverable situation +I misread the original description +I was thinking of thread local as well +Perhaps the only solution to this is a shared memory segment but that reduces the benefitisolation of the query server approach +an increase of +I will try to have a look at it in a few days if I have a chance +Shared variables removed from X. +thanks a lot request sent! +It would be great to get this code in +on this issue and if not are there any alternatives to axis for consuming web services +Lets commit +The process is nice and clearThe only key trick that may confuse people is that each step is a sampling +All tests pass locally +quickstart to reproduce the ofMust be fixed in trunk +Pull request merged +Is there anything which would stop us from adding this to BV +Closing this issue as resolvedOleg +Dup of +Thanks PatrickRev than Ignore the test on commit I posted a small patch to to disable the unit test for now +I forgot to mention that I also added +For example are output written to a staging directory then moved into place? Can a worker task that crashes during writes restart and rewrite the data without getting duplicates? Is there a commit or close type operation to signal all writers have completed writing +Thanks! +Add new function parseurltuple for HIVE +Thanks Enis for the nice new functionality +Same symptom as which was fixed today +The apache common logging API does not let me set logging levels +Embedder is not able to resolve module projects so me does not know parent GAV and does not know to refresh module projects when parent changes +Here is some info on UPDATE privilege in general though +the deadlock can still occurr +But we should do the right thing if they just update location and the browser doesnt actually reload the page. +about the double attachment of things +Thanks for testing trunk and reporting the bug +in case I forget it I think we really need to pull out the AXISALL dependency from the module it means we might need to put those specific utilities classes into axis module otherwise other module are all depending on those AXIS jars +Talked to James +This issue is fixed and released as part of release +Jacques Christian I would like to review this part of the system because I am not sure if this issue has been addressed properly +Looks like we added a dupe link but failed to close +Ok I understand the code better now the scores array keeps scores only for docs that were visited by Scorer +It is a sizable effortSecond we should create a branch for the screen widget refactoring +This is fixable if Partition takes a Progressable input and calls progress from there rather than using the vertex progress +attached review and commit if youre happy +Supplying srcmainjavaorgsafehaustriplesecguardian as a separate file since svn bugs seem to prevent making a diff +thanks for trying +Have a look for instance to line and following +We should use instead +So this isnt going into because it didnt apply cleanly to trunk? Im confused why the status is set to open and the target version is now set to +Ive located the security issue with AOP stack and created related issue +Works fine on Mac and unfortunately the reported didnt include any details about on. +on the patches +I changed it to There is still a big performance problem numTerms builds the entire list of facets and then reports the length of the list +This is the same patch as but applies cleanly to +Yet Id like my to be able to rebind to a JT that has just been brought up on a different addressAt the very least all the loops where workers spin waiting for their masters should reread their configuration values every iteration +Thanks for committing the patch KristianWith regards to Craigs concern about introducing a problem on Mac OS X the answer is that I am not aware of anyone who have tested the patch on OS X +Sometimes it can take a long while for all the files to be synchronisedI suggest waiting a day and checking again +The test passes consistently for me on Mac and Windows +With all these compliant complaints bandied about its amazing we passed the JPA TCK +Moving all issues to the CR +bbfec on master branch +Can we move these methods up in Class +are the refined propertiesget derived from the blocks used which is completely refined and renamed get however this is an intentional over estimate and can easily report a value larger than the vm size +Therefore I think this change it ready to go +in is missing for error messages like Error getting next orderId while creating order and Error creating order while getting orderId which you have followed in case of +HiCan Sqoop run a import on kerberos enabled hadooplooking at the source code I found that sqoop obtains usinga new empty configuration +whether to leave off the +xls +Thanks Tim! I just committed this. +That means we end up sharing a single reference to the same open descriptor and therefore we only have one position to work ll see if theres any magic cloning option possible but we may be SOL on this one +I cant reproduce this error +That was a mistakeleftover from my workspaceI did check the box to allow ASF inclusion so Im not sure why that wouldnt show up +The files are gzip compressed +That allow us to use the concurrent map with an atomic operation using the putThen we can have access to the Exchange during shutdown +Hi vinodBy enabling in the RM retries specified number of times before failing the job +I would really get a separate service for property code completion +It must just have been pure luck that the past builds seemed to work with the above close at your convenience. +Thanks guys for finding and fixing this issue +Im getting the following swallowed exception system clipboard data unavailableI noticed this in getState but likely is a problem for cut and copy as wellEnv +The behaviors of CRT functions are more unpredictable in my experience +I am still in landscape orientation and i have checked +These are still of value and should be considered and having open issues doesnt kill us +Why dont you remove it from endorsed directory and have a try +deleted the part it worked what reason has this part been added? Perhaps it just can be deleted +I agree with maybe not logging it at ERRORWARNING level but we should still log it at TRACEDEBUG level so that we know what caused it shouldnt weI dont think we should be completely swallowing exceptions. +Is there any for doing this via annotations prior to a dedicated Any annotation +ThanksNew functionality looks good but it needs test cases before it can be merged +Those two attributes are alternatives singletonfalse is equivalent to scopeprototype +Also adding update checks that route down to the fs resolverThis allows the local mvn cache resolvers to pick up local changes instead of reading the stale +Thanks for your work +Flushing and loading of doc stores and term vectors Were only making a fully copy per reader of the term freqs in +We are doing something similar here where we compute the value of SKIPSONAR variable depending on a number of different criteria +I dont think it makes sense to inflict that kind of performance degredation on developers just using Ant directly when a clean alternative exists for those using Maven to drive their build processIf you could try some of the other settings to see if they can fix your issue without necessitating the need to fork the compilation task then we could consider applying that to the Ant build files but if extJavac is the only solution then I think the proper direction is just for downstream POMs to control this setting themselvesIll leave this issue open for now in case youd like to test the alternative options +The basic idea is to trim the dependency list of to remove anything that may have UI elements +For this is defined by this is still being defined in +Please provide a stack +Andrey do you care to work on a patch and test it to see if works on +has been marked as a duplicate of this bug +As a temporary workaround you can set accepted to a regular expression that accepts spaces +all the test reported by Hudson as failures +Heres a patch that applies to hbase +I think Install and README should go to the docs +bq +This feature is yet to be added. +Ill have a patch ready soon with the test and fixes +Are you running Java out of necessity or because the installation instructions for Continuum mention it? Using a more recent Java VM may solve your immediate problem +Sorting prefix markers first has another problemPlacing a prefix marker for at T and then another for at T will ignore the nd one even though it is for newer timestamp +This was another regression introduced during the transition to iterative stylesheet execution +patches are ready +Not sure if this should be resolved though +I think we dont need to fix this issue currently +Move issues to next release +It is +If we can get it in soon I would be willing to try putting this in to so this major improvement can manifest in a release +Did you experience thisThanks in advance +But it wont be bad will it +end of optionallycd +Hi The fix is fine Ive committed at r please verify +looks very same +Flavio brings up a great point +User indicated we could mark this resolved as it was user error but it might be good to open a doc issue for this +Moving out of +But it does allow to plugin iteratorsIt may not be as efficient as a framework backed solution +Removes any previous changes toUpdates entry size according to research by erik +Sounds good +Maybe it will change in the future. +Not tested will try to +Alex you forgot to add to you classpath +Committed to trunk. +The fix is incorrect as it adds many svn ignore entries to directories like src main java etcPlease revert and if the command line client does not let us easily set this Ill do with be fixed in r +New patch that implements Johns suggestions about adding the configuration variableThis patch fixes the issues with column numbers in explains +I dont think you can assume that the process has Kerberos tickets for the real user that was pull out of the ugi +Your patch has been applied to all branches. +I suggest not including this JIRA into the trunk before we will have at least tests +From John G to Brian issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +Designer is adding support for varargs in the function definition +Resolving as fixed +This may break applications that depend on the exception name in the error string +Hey Maciej no problem and thank you great job! Ive ran my tests and everything is good +The java version was and ant worked just a little while ago future testing I will try to find ant in the rhn channels but did not find it this time +bulk defer to +Change the patch a little to make it more efficient +See +It leads to lots of issues +But I really dont understand why we use to display attributes description instead as we made for tag description +Attach first patch +for conditional calls +We check if sought key is inside current block only in +Also provide a mechanism to register exceptions as per categories via different annotationsLet me know your thoughtsIf no one is working on this +Ahh my mistake +fixed in branch and trunk thanks for the another jira closing stale resolved issues +Examples can be fixed here or separately +This is nothing to do with dist +If you could do it by regionserver youd have some means of doing other than a flush and so on +The trick is that the JDK itself ever since some JDK update back then and definitely since JDK ignores annotations that are not present in the classpath at class load time at least for standard class loading if you got a where specifically did that come from? Our component scanning maybe or some algorithm? We should be able to ignore such annotations there as I remember it correctly we had the same exception trace +I ran into this issue as wellWhy would be such a widely needed feature constrained only to inverse relations? Inverse relations have an other set of annoying limitationsIt is becoming impossible to navigate between Hibernate limitations to do a decent mappingWould be nice to have it incorporated in the next Hibernate release +Patch for this +I should have had someone review this firstCould someone please review this nowShould I revert it until someone does? +This looks pretty good to me +Thanks +Does this help +The instance variable f is initialised to create a which is not Serializable +This is how it is currently implemented +So that test was passing before the change? It looks like this new solver thinks that the vectors are bigger than they should be? But then how could the be passing in the other unit tests? Hmmmm +Move issues to next release +Thanks Donald and Jarek for your help +The usually flushes the session before closing it + thanks +After failover is portlet in mode view on second node +doh missed the correct package names +OK this patch works with both the branch and the trunkCommitters we really need this reviewed so I can commit itThanks +Thanks Sreekanth and Amareshwari +for orderby and for skewjoin for skew join do not load partition file inWe then return the sample job as a whose status is set to successful so directly puts it in the successful job queue without submitting it +committed to trunk and branch to trunk is sufficient. +do take a while to start do you mean the service will take a while in STARTING or in INITIALIZING? I think in general a component should do the minimum to set up itself to a consistent state in STARTING stage and defer length initialization in the INITIALIZING stage +After looking at the jencks example Id like to clarify that the point of this jira entry is very different from the goal of the jencks project + +In the meantime sorry for the inconvenience +The problem is that we still need to be compatible with Java and therefore can not use that +Assigning to ADK for final review of commits post discussions with +Another approach might be to add a flag to would be one way to address this +Shortcoming is that we need to introduce new and register parser for Spring configuration files just to register few wrappersIs there any simpler method? Carsten could you help us again +A patch for is at +But if I run into a problem Ill probably ask for some helpThanks +Committed this +Yes Ive never really been too keen on it +Mark should we push to +Perhaps its checking for the srcmainjava dir instead of checking for actual +Thanks Xuan for posting the patch +becomes +Im on getting this feature into in some fashion +Included in release. +can we use a tableThat assumes would be backported to right +I dont think its the same thing +Ive committed this +MatthewDid you sign the contributors agreement? It would be better to do this to avoid later trouble +Mass transition all resolved issue that did not see any further comment in the last month to closed in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +Attaching with the change to the line in +Tested State Service WSDL import and the generated procedures looked fine and were error free +The specified flag qaliasnoansi does not help in my case unfortunatelyexport TRANSCODERNATIVEexport MESSAGELOADERINMEMexport NETACCESSORexport THREADSpthreadexport BITSTOBUILDexport CCxlcexport CXXxlCexport CXXFLAGS qaliasnoansi w O DNDEBUG DPROJXMLPARSER DPROJXMLUTIL DPROJPARSERS DPROJSAXC DPROJSAX DPROJDOM DPROJDEPRECATEDDOM DPROJVALIDATORS DXMLUSENATIVETRANSCODER DXMLUSEINMEMMESSAGELOADER DXMLUSEPTHREADS export CFLAGS qaliasnoansi w O DNDEBUG DPROJXMLPARSER DPROJXMLUTIL DPROJPARSERS DPROJSAXC DPROJSAX DPROJDOM DPROJDEPRECATEDDOM DPROJVALIDATORS DXMLUSENATIVETRANSCODER DXMLUSEINMEMMESSAGELOADER DXMLUSEPTHREADS export LDFLAGS export LIBS +xsd validation was not performed properly when it resulted in an exception during parsing +Youre right the code snippet above worked fine +If someone has a patch please feel free to reopen. +I wonder why we were waiting to merge patches meant for until and were releasedThe problem is not a lack of branches but a lack of committers to maintain them +I have committed this +Though if we have the Shingle analyzer enabled for that field that might work for +I just made a build from the head of trunk and it no longer happens + +Committed revision +Flapping occurs with vnodes or without +If you have a available but the checkout fails there is an network error or something else +No this is mainly a issue so unless its causing a merge problem for other patches I would say its not worth porting it +What is the benefit of having TDC wo PQ just so that we can have a PQTFCCollector is the base class for all and if someone wants to create a TDC he can do so +Seems like these didnt rewrite history they just were another thread of history merged in +so we can get this one resolved for good this time +I tend to use the version with which I also do work but I agree that it should work with +latest patch looks goodbq +My app does this about million times a day so this log is misleading and if its debug it should not be an INFO message that looks like an error +Drawback this will amount to adding more allowed dependencies upon ie +Committed PM Simon Willnauer kB At revision Thanks Simon in benchmark spellcheck spatial snowball and queriesi didnt finish in contribqueries so it still needs to be done +In this case the application should continue to work except that the faulty service provider will not provide any serviceAn ideal solution would be to make the behavior configurable so that programmers can decide what to do depending on the service +There is no in the Installation Guide +Just a side note here +committed as r +So I delete the first and attached a new that superseded itOn the other hand the two spec patches were complementary +This is a reminder for the Hadoop committers that is ready to be reviewed +Please use xc and not xc +Will do so in the next day or so +failing test works for me when retried +Merged from fuse rev in rev . +CAS for +BULK EDIT These issues are open to be picked up +It is fixed in +I dont think pulls in a conflicting dependency from mvn dependencytree so not sure if that is needed +Updating patch +Any guidance on a shorter workaround until we can upgrade to would be appreciatedThe test case attached represents a use pattern we need to support but has been narrowed down to detect a lost message in the timeframe it takes to trigger this issue +Ok this setting will also be relevant for the parser thenRegarding the beans I was planning an annotation property telling if empty values must be turned into nullI suggest to report this after the release it can be added later without breaking the compatiblity +Not to be implemented. +You should be able to simply update the under lib to get SBT working on a fresh kafka checkout without having to use s script +With factory methods the framework is free to give you anything that satisfies the basic contract +passes with fails without it +Reported fixed by aerogear resolving for QA to verify or reopen if are still present. +Downgrading to major +Can you elaborate on what is needed beyond this or if you hadnt found it how the usability might be improved +WeirdWell its fixed in but we want to remove it for future releases as were getting rid of the whole services concept +Committed revision +doc will be included as part of next release +Thanks for the summary file +Sorry for not understanding what would you like to be done +XSD Namespace and basics are completed +Heres that dialog +check richfaces showcase +This is not actually a bug at all and in fact WTP is referring to something completely different than deploy location of the modules +you want to ignore the error when you apply delete looked more closely and it looks like before the code that was throwing an exception was protected by a check to +No just for local issue keys +Committed +If you have concurrent access to a database with castor and such a database application it may be possible to get problems with the different representations of true +Integrated to trunkThanks for the suggestionThanks for the review Sergey and StackWe can revisit this matter if we face OOME issue in the future +The stubs were culled from Apache Harmony +Assigning back to Rob for him to comment +Patch was created against branches unassigned issues over to the pool where they can be worked on for the in all probability patch +Cant upload the test log +Lets discuss the headers on the list there are many ideas to explore +ATM aside from Todds comments Id pull the copying from initialize out to a separate method Nit would say populate instead of initialize since we format the shared dir above Feel free to punt this to another jir how about an test that Starts w non HA NN Shutdown enable HA on NN add SBN NN Run initialize Start and transition to active NN Run bootstrapStandby Confirm NN and NN are up and HAOtherwise looks a lot for the review Eli +Frank do you want to finish this patch? I want to do it but I wont have enough time until later next week +committed to the snapshots closed. +Closing as rejected. +Can someone fix this patch for me? I got the thing so it tries to throw up a link in the table after replication starts but the damn thing fails when loading the local doc regardless of the URL I use +This was the original infra request to create the zone the zone was subsequently created this issue can be closed +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +fine with me +agree with Niklas et al that its best to regard any line prefixed by the boundary string as a boundary +the enum value of the attribute should be generated with alphabetical order otherwise the enum values have no change the order might be changed when we generate code +Thanks for your correctionsCan you solve the ticket after code review and commit or shall I do that +This issue was reported for the previous design of the. +Setting rpcmaxthreads worked +The only difference is the file name and a line offset to accommodate the different code baseIm not changing the workflow state because cant run on anywayPlease review and commit to +The servletPath is required for applications that use whatever mapping +patch with the variable name fix +However it appears that there are currently no known cases that would benefit from such a change. +There are two advantages for kinit +I have removed the JSF experimental admin portlets code as well as the JSF dependencies from the security portlet +I failed to set Deploy as ear on the Seam Runtime page of the wizard the error caused by a missing mvel dependency in parent occurs only when you leave the default selection as war I have a complete set of projects that shows no errorsFurthermore I noticed that eclipse does not show the ear and ejb icons on the projects and several other things but I wanted to let you know this asap so that you can stop looking in the wrong Ill try the seam support on my newly created you very much! have removedmvel from the war configuration because it is a transitive dependency for for your help Dietrich. +Perhaps that issue was alsoprematurely resolved +Attached a patch generated from the trunk +Thanks +Nice thanks for the tip David +Please review the patch on the top of +Also had that problem with git +All done now +Thats quick work I attached my and can confirm that I dont get it on the following patch should fix that particular problemIts the only putpost request that is using chunked encodingLet me know if it works for you +Thank you all for the ideas +Comment from yonik inbq +Patch updated with +One minor nit pls use junit in the newly introduced unit test +The assumption here seems to be they can use environment variables with the command for expansion and that the onus of correctly escaping the command escaping any special chars is on the users themselves +I suspected mode since the code works fine in mode +then you can secure it based on the originating IP and not worry about messing with dav them in +We had logic to handle pg style set statements but the regex needed to be more general and the driver handling needed to account for the empty string. +This is a major drawback for me as Im hosting many applications on one instance +The first in correct for that URI +It needs tests to cover the new metrics +Looks good to me. +Fixed by fef +This is what you get from and I think we should not behave differently +See the attached zip file +What I try to optimize is the luni path defined in +Is this Issue a Blocker or Critical priority for EAP GA? If yes +moving to XRX as I think the current blocking issue is related to clocked asyncs +I think these warnings are new and they are caused by failmon +Tests in progress with Cellar Cave Camel CXF and on Windows XP Ubuntu Debian Fedora Solaris AIX AIX in kit release notes svn ciSending Transmitting file dataCommitted revision +I have verified that the fix is included into the CP codebase +I would guess one of two things are true There was a reason to do it this way and it was never documented Some old code did it this way and nobody has bothered to change itEither way the patch seems pretty harmless and it sounds like the default collation is what we actually want +Thanks +Committed revision +Ryan Your observation seems quite correct! So the outstanding puts case I was worried about shouldnt happen because of the updatesLockSo the mystery thickens Ill need to dig further as to why these edits got skipped during replay +A new patch containing test cases +That is I didnt find a reference that said get only works on element that can be navigated from the root element. +As an alternative to providing a complete test case can you explain the nature of the error and possible conditions that could cause the error message? The error message appears to be somewhat specific in nature and is obviously related to the processing of a mapping file +Im resolving the issue +The creates a which natively supports nested exceptions regardless of JDK +Thanks Zheng and caring to apply this minor fix before next version +JacopoI cant see how this relates to commit +The attached file is a patch with listpeers support +New patch onto and test pass there as well +for on the side as well +I need some time for this as I am not very familiar with junit fixtures +Through some additional debugs I have files containing stacktraces from every time is called so I know it is still being called times during startupExamining those stacktraces I have determined that each of my cores has request handlers in common with two additional handlers and for a total of +Commit dfceaaafdccbdbI really wish git would post this automatically +leads to a few response times that are over second but the average is way down closer to msyeah you are rightthere is one more feature in whcih enables cleanups to be done in a new thread +bulk defer of featurestaskswishes from to +May still be issues here but nothing critical +Committed revision w some test cases +lgtm +Looks like the reason it was not included in the EAP repo is because its excluded from the EAPAS build +Regression tests added +Thanks for the patch Ken. +Also for setting the schedule when the discovery should reindex. +The last comment means that we can close this issue now +Ah well. +Hey Matthias I totally missed this one and now the patch doesnt apply cleanly +The attribute map I get from the event does not have expanded properties in the attribute values +Questions about the point of such a query aside +The file system connector needs the ability to create Binary objects from Files without having to load the contents of the file into memory +Perhaps picking any type in the AUTO sorting code would work since it wont yield any hitsOops thats not true since there could be plenty of hits in an index but none of them may have that particular field indexed +Temporarily reopening to update release note status. +was committed and is working +Ill check what does exception or empty result and fix FC accordingly +Please advise what exactly should I do +Whilst its true that Java changed the implementation and thats what causes the slowdown FOP still elects to use that particular JDK library +Either if we use the original logj logger or a new one we can improve the performance using the same mechanisms that are used to make the WAL efficientIn relation to Aspects I didnt understand what do you mean +I dont understand the rationale these methods are here to create Readers and let them open for use by the cant close s callers responsibility to close the reader after use +Good news you find a way to use itIll check +This is not a forum +Note that this solution is aimed to avoid a dependency on which Struts solved through the class +According to comments all known method issues have been addressed in other issues over various versions + +Updated patch with proposed changes +Please open a new issue for the abbdict contribution +I rarely make use of CtrlC and CtrlVSo yes CtrlC and CtrlV still works in the scenario Ive presentedI also understand that the automatic highlighting lets you use CtrlV to paste over the original http text +Very you please attach separately patches cant include binary files +Your right about the implementation of the different JDBC apis +Revision as a workaround the cost for index scans can be set explicitly nowAlso the cost for the primaryType index is set to +Ill investigate +Left as implemented. +Will post a new patch then which enables the remaining tests in client mode +I keep this issue open to fix this problem +If it would make git users happier we could remove it in instead of +Fixed in revision +If we still see this problem we could create the JMS receivers programmatically upon seeing a JMS port rather than configuring the receivers inthanks sounds like a plan +Thus the negative impact on readers +having the JMS dests deployed automatically have each process check for the existence of the JMS config before runningThese are all minor but would help the OOB experience +I will open another JIRA to remove these references +Simple fix +I will add the javadoc comment as soon as possibleOne more thing i would like to change the name of the reader from toThanks RegardsGaurav +supports full incremental and clean builds in eclipse +this is by designsigil repository config does not have to be loaded via AntIvy so an AntIvy properties context may not be availableThis allows the sigil repository config to be used from Eclipse where Ant properties dont existHowever we could pass some additional override propererties to use for variable expansion which would be Ant properties when running under Ivy and Workspace properties when running under Eclipse +New patch that adds set method +If you give this group a standard name this will work for an existing domain model as well as we can apply the group there +Thanks for this +This is across all jars in the lib and serverall directories so thirdparty changes are also serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for Type current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current WARNING serial error for current ve fixed a couple of classes +Marking as a duplicate of which tracks publishing in mavenThe functionality requested here publishing to the local repo is already happening actually + +The decorator class should implement the converter interface and will delegate the actual conversion operation to the decorated hibernate entity +See s use this one to track the issue for EAP as is a support patch issue and has a different workflow +trunk + +Thanks SteveI guess you have noticed the explanation about the problems i faced while using and +HIIve sent you by email +Lars are you working on a patch? I could give it a try next week +I cant reproduce this issue +I have committed this. +Mark could you indicate some way to release new indexes in for example I made some changes in and need to reindex collection +Looks like I ninjad this exact patch in right after committing the original to fix the test. +No +Id be on committing this so it gets in for then opening new for subsequent issues +thus I am changing this +Hi DanielA small message to up my problemDid you take a look at my curious caseHope it doesnt bother you too muchBy the way thanks in advanceRegardsTonyThis was fixed in + +This issue is reported against a version which is no longer maintainedwhich means that it will not receive a fix +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Why? The sync point would only happen when the instance variable changes value not when the object is accessed +Thank you Tanping. +See also +So the problem seems to be that it is trying to write the file to a directory that has not yet been created +Mike sorry to flip back and forth on this since last night +Vijay please check if this is still an issue? Seems to have been open for a long look into this +Hi LesI understand your problem +class does not exists +fixed how in trunk? with orand I really dont think there is a need to disable these when they are third part installs +Good point +Can you test it out with mainline and see if it still is possible to useI am still seeing the issue present +No but equally Id like to hear when issues are found rather than hide them before hand +With a I can catch the exception when I try to set it and of course I have access to the instance +This patch allows a new to be added to the file +This example illustrates the problem +I will update the patch ASAPThe is needed to change active to active in the nav also it needs to inherit from not in order for it to get the new Nav +It was a mistake to add a call to releaseResources from the finalize method in +Jdbm is not widely use so Id leave it out. +Canceling patch until Hairongs comments are addressed. +Chris can you please take a look at thisThanks thanks +package is not enabled as identifier in trunk yet +Maybe the solution is to change the toString so it doesnt try to get the count of the object if not comparableHow would that beCould just be Not a Comparableb or something +Hi Scott Commit abfdeeaadbadeaeade now includes the content type directive for all jspx pages +When writing the patch I thought about the possible cases for which components would be affected and I came up with the following channel Sourcesink which buffers data SourceSink which has a bug in it about cleaning up resources seems incredibly unlikely given is an and also seems unlikely since the channel is a buffer is a bug and I dont think we should worry about breaking theoretical bugsThat is how I came to the conclusion that only would require this behavior off the bat +Shouldnt a container like Geronimo filter out everything except for the classes that implement the specifications +I was about to your latest patchThanksC +All relate to the doc for application integrators +The new patch acquires the switchlock for the entire cleanup so no flushes can happen +Hello RuthThe patch has been applied at revision r thank you for this enhancement +We will have to request permission from Oracle for its useThis file shows the UML diagram of Wongs Developing Jini Applications Using JME Technology ISBN Chapter page The surrogate architecture was conceived by Jim Waldo an engineer at Sun Microsystems and is an open project at the Web site +This has been incorporated into CR build once I can upload to Ill mark it as resolved +the namespaces will cause the ref counts in the namespaces to be incremented and thus the namespace structures will be retained until both the former parent tree and the detached node tree are released +Hi Andrew I still cannot get the patch file to automatically delete +I tried creating hourly snapshot schedule with a keep value of +No source in the archive +Need to discuss what to do +I would like the inner structures and methods of those classes remain private +seen before +It works for me but Im not sure that it will work correct in all cases +Anyhow Ill debug that tonight and hopefully find whats wrong so that the interceptor can make it to M +Suggestions on solving this +Verified on EAP release note fields. +I will take this issue and try to resolve +Btw the tracing level tables are duplicated in the reference manual and the admin manual as tables Table +The gain is quite build with and without debug the sizes are vs M for standalone compile i think is is acceptable also allows us to debug the installer produced by izpack and given the high speed connection we have these days I think it is worth to turn it we may want to have images and have user decide on which one they pick +Additionally we identified a few different areas where the data mixup could potentially occur but those were easily tested under heavy load with no negative results +Thanks. +Also thanks for the liking of my username +Converter class is the main guy +If you for instance have a database created with Derby you cant create the procedures when you do a soft upgrade but all the lob functionality that was present in should still work +I still think we should do it though +This is a trivial fixI just committed +The binary flag has been set manually as ask by Simon for fixing this issue +Its probably not what most people would expect if you said rest +Since in is only designed to resolve bean definition properties the only way to get access to that functionality was to subclass it with the subclass calling parse passing in the string to be resolved and a Properties s good that the code has refactored this stuff since the odds itll become easier to reuse the recursive property resolution logic within application code but Im trying to find a way of doing this that works under both and so far without success +The failure was due to the Jenkins host running out of resources again +Its working as expected +It opens the door to cache synchronization issues and is not a perfect solution but it does give the developer the necessary control over query cache invalidation. +I guess they might need some help here +Garys patch +for the link back to the edit product screen I meant to Anil your patch is in with rev +Only if their names are known +So even if we have the only bean which will be injected we will have two options +Looks good to me +Do you have it since my changes on fetch groups +Thanks Binlijin +I just did some quick tests and it seems that the showing and hiding the with displaynone with the setTab is responsible ie +An extra flag to relax JMS semantics with the warning that it would make user code that relies on that behaviour non portable +Hello vikasHere is the updated patch with changes you suggested +anyone please also rename the title to support for nested headers in datatable including inline cell editing nested headers +to commit +Even better than the first which was also good +Apologies Arun d patch attached +Patch to fix this issue +Now with tests +Will commit when tests pass +Thansk! +Merged to the branch in revision and to the branch in revision . +Committed revision +In JIT mode the throw location is incorrect +Also it was time of IE on Win +Issue is resolved +a is doneb is donec is done +According to the forum thread it seems that at least one issue you were having had been solved +to trunk +opening this since this just seems wrongWhy are these merged and not split correctly instead +Im glad its now fixed +testcase patch +Not patching due to I can acceptRequiring the client to configure a custom protocol interceptor in FAIL +Hi DanThanks for looking into this patch +Yes I did not think about the bubbling up part +Thanks for your help BrianIt is not necessary to mark this WONTFIX +See also for eCloud which now appears to be a different API +I have some minor comments about the test case in reviewboard +Ive committed this +I find myself leaning more towards the name but I can be pursuaded if the community thinks otherwise +If you wait for st server to finish startup before starting the nd then the mentioned issue occurs +XNTHREADS XRESILIENTPLACEZERO XNPLACES x Sleep secMain Main Woken commented out the above line in and by rHowever it needs to be verified if removing the line is the correct solution. +current value of the loop counter +It seams to me that both issues are about the same problem +About the point on read I dont think thats a bug +The algorithms are too simple to have such a dramatic effect on performanceIt is possible that the test that I provided is flawed +Hi OmichActually the plugin is not complete so you will have to wait for a few days to use it +Alexei Fedotov mentioned Actually it was Alexey Zakharov who had problems running smoke tests with Ant +Ive fixed this for the branch by simply reverting the old weak map logic +No response assuming ok. +Thanks John +Thanks! +Based on reporters feedback Im closing this. +Like the original reporter I think its important to fix this issue because it will prevent plugins like Cobertura and Clover to run in a multi project build if the modules have interdependencies +Cannot reproduce m guessing that it depends on the hardware +Will likely need to generate Java from WSDL in that case +anyonei have configured it to run every hours +Hey Ken are you OK with this going into? Im going to try and cut an RC within the next few hours +No response from user could not reproduce secondary problems primary issues are expected. +scheduled in a job + +All tests pass locally for me now +Need to test +Yep Thilo that can be done it via s argLine parameter +bulk move of all unresolved issues from to +Trunk with test of excludePath and release errors fixed +The pleasure was mine +Appears that the jar downloaded to my bundlepool was the culprit in that it was devoid of classes +Issue for +Integrated to trunk and branchThanks Ted for your review. +bumped to +So instead of having a Related Concepts link for the whole group that only applies to two of the topics Ive got two similar links in the Connection and Statement topics +have sent you email +sorry for feedback so late seems it had been fixed already for this new version +Binil confirmed and tracked down problem +This means that local changes will over ride anything that has been installed into the local repo previously +The jira can be added as a subtask of that! Ill link the new JIRA to that one then +Also replaced foinline with foblock as foinline cant render directlyA contribution from patch in trunk and all the branches +The new files for unit closed and comitted to CVS +Agreed The fact that the stream is created in the constructor is a bad idea as well +There is not much enthusiasm in this proposal and in fact no requirement has ever been raised in this respectSo I suggest we write this issue off and close it wont fix +DmitriyI see no reason not to commit this given the test report looks good +fixed in r +If you think thats the way to go let me know and I can file a separate JIRA +just commited about itcould you give it a try with next snapshot please +Moreover I do not even see Commons Lang declared as one of the Synapse dependenciesI totally agree with you that it will be good to not depend on Commons lang just for two methods +Anyone want to turn Marks solution into a patch with a test? Or even just contribute a test? It would be nice to get this into +Resolved with revision +Related to +The TCK unit tests that use order by in queries now pass and as such are included in all runs +Thanks Rick Ive got the jar nowWhat should I expect to be different about my build? I didnt get any build errors is thereanything new that I should expect to see in my build outputs if the patch was successful +Go commit +Its actually identical to one of the java puzzlers in the puzzlers book by joshua bloch and neal well it still seems pretty damn stupid to meobviously doesnt have any negative valuesif i am testing for something that is odd it is natural to test if not if ! in fact it made me stare at it for a while +This was fixed as part of the work of implementing the Databinding annotation +Phase part patch has been committed +Just one nit you missed the Test class +Which is why in they added the annotation +I see that the current code sanity checks that each supplied is in fact a of the original rangeshould we also check that the supplied actually covers the entire range of the shard +Arron the test case works for me +closing resolved issue for release +Hey Guys can we cook up a free MP file with similarly corrupted tags? How hard would it be to do that? Id be happy to try but need a pointer of where to start +Im aiming to replace my first design for the JAXRS service with a new version based on our discussions posted to +JPDL and BPMN process will start on this soon. +Fixed some issues with onselectionchange behavious +Permissions is a structure internal to the +Please consider continuing this contribution in +Also Im interested how you arrived at this test case +Because i asked about backporting to +Ill submit a new patch after is working for me +The Web UI can be at configure time via the configure flag +Because has been releasedif there is still some problem on this please it +Need to leave as is. +Sun JDK and above ship with a complete implementationof the which is also provided in the mxj having the mxj JAR with duplicates of those classes breaks this test I remove the mxj JAR from the Groovy lib directory this test case runs there any reason why Groovy needs to have the mxj JAR when running on or above +I believe paths as expressed within Jar files are platform independent while the file system path of the Jar file itself will be platform dependent +I would like it committed asap +They are in a sense that they dont prevent you from using something and arent even discoverable when you are adding an import in Eclipse or vim +Its not a jdk issues because they fail on jdk here as well +Moving to + +Like Oozie +As far as I can see in bits weve got there is no clerk configured in clerk for riftsaw is still not configured in are a number of files that are overlayed by the riftsaw installation however not all are applied to the SOA build +The proposed patch does not work +Part of a bulk update all resolutions changed to done please review history to original resolution type +Sure I could do another archetype with openwebbeans +Thanks +ooops the fix introduced a regression as it used wrong index to build binary valuesof a property +Marius could you look at this asap +Let me know if youd like any of these processed outCheersc +this is a DB admin problem +Removed the host name to ip address mapping +But this change seems fine in that regardBensonI am completely down with you modifying rather than building beside Colt +Thank you very much for your contribution! +Update including a minimal Testcase +Dhruba is that a documented exception +It seems to take about microseconds to handle an IPV full address a bit more for an address with elisions a bit less for an IPV address and about microseconds for a addressIll open the regular expression for comment before I fold it in with the patch and with other changes Ive suggested in the main patch +New patch corresponding to updates on patch +close region by method It can clear record in and close region in regionserver +properties of the client specified URI +Edell would you mind committing the required changes to the features project and close this issue +The system implementation would stay as it is an access path for bytes +Hi Rex +The lack of logsBasically when the split fails you have single line in debug with in progress waiting each secondIts better to have the details of the task especially with the server that runs it in info instead of debug + to any change here or shortcut +If it hangs for you or mbairnes I would suggest to try it out also in your machines +Ill fix the document problem tomorrow its too badly written +When the patch applied the plugin assumes all grammar files except the one being processed as candidates for super grammar source files and passes their paths to the antlr tool for the actual processed +But not in maven +Although the error is correctly recorded against the type it is also recorded against the type +I tested out and it works perfectly fine and load goes back up to on the restarted node +Thanks for the examples now it works indeed as one would expect so on change as wellI still havent tested change but I expect thatll just work tooSeems were getting very close now to finalizing the patch and committing it to the trunkAgain thanks a lot for your help and time Thijs +In addition allow the destruction of contexts after each request to be delayed so that in testing one can query the values of wicket components that have injections after a request in order to determine if they received their proper patch replaces the previous one attached to this report +Currently to add to a collection you have to get the collection then call the add method +Hardcoding is done in Regards DamithaWe already have this information in the +Perfect +The patch looks good +The patch causes regression which is reported by rolled back at r +I will check it in +Got it +only supplies the right index dir if the directory is an instance of since we now used a Directory wrapper the index dir was not properly coming from the +Let me think more about it +OK resolving this ticket then +Logged In YES useridRedeploying only the Session Bean +Attaching and with a change to M srcref +I first tried with maven and saw those failures but it sure compiles with +Applied although the build XML is only available to enable users to build a special JDK version +I believe this was fixed by the changes to for +Ticket for Eng Ops created +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Hi JarekI think the root cause is is it necessary to start? For your solution there is another problem is port confliction if user runs offline deployment when the server is started +I openeded +was my bad +SaschaYour patch is in trunk at r +Marked as resolved +I myself always use theendorsed dirs method to use newer versions of the Xerces and Xalan jar filesthan come with the SDKYour suggestion to describe how users can download Xerces and Xalan themselvesand use them effectively is fine for development but no so fine for stabledistributions for usersI found today that the tests run well with the jar files from the SDK butI have problems when I substitute newer jar files for them +Username rickardoberg +The Forge Script you can use to get the Forge Console up inside of ToolsJBDS and then run after the plugin has been you have two options +Weve been using that pair constructor to get around the exception but now that youve pointed it out why it wont work it makes sense +same as fixed bug +While I believe it is a net positive despite the downsides it isnt always cut and dried +The element should contain a whose value is the name of the application security role that the user may be mapped into +r dont end the conversation until after the redirect +i didnt know this s a good idea and i will use itthis issue will be fixed for the next release this evening in released the with Issue fixed Thanks. + +We can achieve authenticationauthorization independency +Then to Renato +This is intended allows you to return null when the return type of the method is a primative and Groovy returns null when there is no expilicit return s arguable that the correct behaviour in this case is for the null to be interpreted as false which is the default behaviour when null is treated as a think you have raised an interesting point +I will cut the branch from master on that day +Ill try to do that from within eclipse +In this architecture does a given socket get bound to a particular select thread for the lifetime of the connection or do the sockets move around between selectors opportunistically +New version of the patch with + +As would I +Eliminated the need for object id disambiguation by reordering the mgmt periodic update to send deleted objects first before status of active objects is sent. +SaliyaYour patch is not clean +counter send should be called with remotes not targetsBut other than that it lgtm +Please let me know if any changes are needed +Fixed in Axis trunk revision +Actual creation is delayed until a method is called on the proxy +This replaces the previous dependency on Google Collections is being removed +I am impatient to see if you can reproduce with +I will update the docs and resubmit +But i thought clover had in the reports a true count and false count at least? In this case at least the branch can be evaluated True or False +. +Attached latest context Changed name of issue to Improve presentation of usage errors to reflect the hierarchy in the codebase +Only works in the nested created by the +The patch attached contains saving messages to hdfs before sending to other peers +After putting in a fix for the verification job ran in hm a nice little improvement over hm. + +Ill look into that and report back +Seems this bug is invalid marking it that way. +Right Stefan is mostly hitting the du problem +Attaching straightforward port +Trunk sorting is working fine for me +You only get warnings to make sure next releases use the latest conventions. +The calls on the stack above that are still part of HDFS are known to HDFS are under HDFSs control and are modified in a coordinated manner as a unit +The patch applied cleanly but It doesnt fix the issue +This code is a lot cleaner there +Moved to JBTIS though I suspect this is out of date since JBTIS will ship with SY or and AFAIK theres no plan to include SY in the JBT SOA site. +committed thanks! +resubmitting just checked the repository and the patch has been committed to trunk +for the patch +Thanks. +Marking as resolved since all the points were addressed +This chapter includes a description of the Logisticx scenario and the client server and supplier applications that are built and used in the draft of this book will be available for review by March +Yes but ODE had quartz scheduler some time ago +Jayson I see your point Instead of changing the current isnt it better to add a new one for this purpose +Erik Holstad added a comment Mar PM +For assistance have a look at how security module have been written +I believe this affects most european locales. +Your patch has been applied thanks for your thank you +I havnt been able to get the deployment to work since there are many relationsships that contain a many multiplicity so I cant say if my testcases works in the one multiplicity case sorry +Im seeing this reliably locally on the latest change as of today and one one of my jenkins jobs running an older build +simply because the input of a hbase mapreduce job is split based on the regions +takes care of all the details +I will check it out and mark the box accordingly +OK Ive got passing +Please open with a failing test for trunk if you find any more issues on this. +You can create multiple conditional packs instead of one pack with conditional filesets then the size would be have to check if it were possible to calculate the pack size for each condition during compilation so that we have these values correctly at install time +Its probably better to leave things the way they are. +provider has been committed +All tests pass +Kicked off another Jenkins build to see if repeatable there +Also it doesnt break older implementations +Resolving applied to trunk and. +DavidThanks for the quick responseIm attaching the CEEDUMPIve tried to build and run with and without XPLINK and got the same resultsI can send the compilation and linkage output as well if necessary +Can you validate against your use case to ensure it still does what you want +Actually in my case I refer to the prefix to check whether it is the containerappappattempt id string which is already written in the file +Resolving this issue as a duplicate. +Hadoop was released today. +Was fixed for +Uploaded a patch that changes the methods to public and marks them as Private +looks like it was just a test bug. +The problem has to do with the coercion engine and defensively passing the value to the engine to ensure a safe value writeThe new performance spread of get versus set in trunk is about +This would also not be supported by the current implementationIn addition the old implementation was based on the assumption upon the actual path used by the resource resolver to access the resource +To solve the issue above we need to prune the dangling branch before proceed +Not sure this is really necessary +Is there something else that needs to be done to get poms actually uploaded to ibiblio? If there is something else please let me know Ill be happy to do it +Now its way better testable and the effort for test writing decreased enormously +I think this issue is not needed +Test failure is unrelated +This works fine in . + +committed to trunk +Note just seen this is fixed in the CVS HEAD +The proposed solution here is to extend the configuration of the to beviewNamelog will be defaulted to NONE but you can easily configure your own level Juergen all in all I think this decision of yours is a wise you for your attention and please go on caring about the quality of this framework the way you have done so Fabio. +With the commented out this bug is irrelevant but that doesnt make it invalid +Added Apache license header to the new test Uses Junit annotationsIll commit this shortly +My view of Sonar is not an indicator of whether or not there was a problem with the build but rather a place to store metrics related to quality +Curious how do you know that andor measure that +Patch committed to trunk +I dont see anyone that prefers rn trying to make this consistent or making noise about wanting consistency +And + +I forgot to mention I used java in all the tests +Testcase I wrote in patch makes this happen so Im looking into it now +This patch is made to support proxy object as keys for so as to follow RIs behavior +does have tests for almost all jsps but mostly concentrates on if authorization works as expected or not +I too have a two or three searches that I need on a regular basis and setting them up from scratch every day is very annoying +jb are you still working on that patch? +I havent been paying close enough attention but I missed the part where deletedtrue revisions were kept in the history +Version is what will beStarted in on Migrate +Take it! I havent started +Ive since updated and havent run into the issue described again +On closer inspection this eliminates of the CPU burn +Please find a patch which fix the JUNIT test +And I dont think the replace we have is any problem for the flyweight pattern since all we support is replacing a column by another with the same name +In addition the source folder should probably include src so when both packages explode there is not a conflict in naming +One solution is just skip this test because we cant do anything from myfaces side or copy it from myfaces impl project and reference the tomahawk inputHidden to myfaces rendererThe more easy is solution so Ill commit it if no objections +Applied revised patch +derbyall still running +bulk close of resolved issues. +Can you attach a stack trace? A? Anything that shows the aspectsof the failure more preciselyAssuming the problem is reproducible a file from a run with the property set to true would give us a lot more to go on +Verified +I have a fix for which I think will fix this also +Ok +Please validate issues like this against the original showcase using a URL like showcaserichfacesdemoattachQueueskinblueSky on your mobile device +Ive committed this and merged to. +I agree with your positions +About the virtual memory limiting we have tried to use it in our cluster +is attached correcting the Hibernate Spring tests +committedOleg +The ones generated by the m build dont work since they have the wrong groupId for geronimo artifacts +I meant Teiid is included +I am a Pig developer and I need to query exceptions that are thrown by Pig in the runtime envrionment +I cant get any generated classes at all +Wow +Not actually fixed until we have updated to docs completed the tasks etc +Can I have you check it out to see if it fixes your problem +Thanks for all the hard work Alfonso for pushing this through +Do you want to declare it a known limitation that clocks generally dont work with STATICTHREADS? Is it possible for us to detect tha bad situation dynamically and exit with an exception instead of hanging +Glyphs ignore from all but layers view now. +It would be nice if such a display were hierarchical when multiple versions of a database are supported but what were describing here is having specific translation logic for version and of mysql +Thanks for your answer +irrelevant until we implement the various Region algerbra operations +I was just curious why one cannot just add these annotations outside of ROO and it would work because JPA is in use with the project anyways? I can understand that it would perhaps confuse ROO until its implemented and would not be recommended thing to do but I get compiler errors saying Embedded cannot be resolved to a type if I experiment like mentioned +I think I got the solution for this issue +The resolvers are walked to build up a list of possible Urls and then each one is tried until the Pom is found essentially findFirsttrueAlso note that a parsing failure on the parent does not fail the original parsing +Added synchronization for class stage change into class initialization phase +Should I create a new one +Hmmm I actually did an svn status to see if there was anything to add tooProblem is my current tree is too messy and I missed itThanks for the +Invalidating as the number of fds doesnt increment after reaching open +I needed to do the following in the correct orderStart cd apimvn installcd +dlls to create one for NH one for NH and then one each for a seperate OSIV module +Previous wasnt applying correctly +The benefit of switching from last returned as head revision to last accessedseen for figuring out when a revision is still needed is that we can allow unused revisions expire much faster +A smiley destroyed my regular expression +We would like to look at this in the future but right now were not sure +A more extensive version of the release notes +Closing Out of Date issue. +In Server a calls time stamp is updated when it is received and when its reponse is sent to the Responder +Attaching updated updated html files for this patchKathey I tried to fix the appearance of the table but there must be somethingwrong with the style sheet +Im not yet convinced this is a bug that needs fixing +Changed to Major because its a PITA today +playing doesnt really help in understanding what you are doing +I will commit this tomorrow if nobody objects +Ok further investigation the partial response writer is not hit as it seems I set a breakpoint on startUpdate which is started at the beginning of the partial tree processing and it is never hitSo must fail earlier than that the update is pretty much almost at the beginning of the ppr response before the components themselves are processed for rendering +Thank you Benoy. +An encoder that merge all those little chunks could help to improve this behaviorOf course for very large PDU like Search results we will have to deal with the memory footprint in the same way we should do it for the decoding process +Hi Sebastian Could you enhance the patch to make the model threadsafeSounds good +Attached mainly for approach review +another way to address this issue specifically for CQ was to fix as we currently have to create separate ACL nodes for every single user +I think is OK +Fixed in New Revision +I do feel that this is important enough that we should get a fix for this issue inThoughts +Thanks for the contribution Wang. +Ill check the log. +If the problem persists please reopen this issue we can continue fixing this by using our own. +Should be verified in gatein master and reopened if still valid +Picket Link has been released containing this fix +Delayed entry parsing is not enabled with this patch +The same problem is there in handling MTOM in Axis +Took me hours to figure out where the hell the exception was disappearing to +closing for now +I added and to the summary because it is very similar problem +The value depends on particular block being read or written +To add a bit more potentially relevant information I am running these unit tests under Vista JDK on a duo core machine +Started on a test that may be able to recreate this but no luck so far +Attached patch fixes the issue by +Anybody able to translate that last comment please? + +Adding as an Affects Version since I can reproduce this there +I experience the similar problems but with Ivy +modified when the warn is performed to be based upon time in queue to eliminate spurious warnings. +Thanks anyway Rene I wasnt even able to realize what you did looks like graphic design and I live on parellel worlds +sounds good to me. +Please keep us posted +Patch looks really good ChrisJust one minor comment you are missing Apache headers in the newly added cmd scripts otherwise from me + fixes PDM errors for fixes PDM errors for fixes PDM errors for for andthe http patch doesnt work +I am starting to think of is as Wont FIX since RLR would work fine with +This is an appropriate test +I dont see where this is limited to pagesI need to list out all the jspservlet paths and check whether they are user facing or not +Look forward to collaborating on this projectHi Arun definitely Id love to contribute to this and make it into the production. +Can you please let me know if this is really an issue or I am doing something wrong please? I use with Java and Spring Framework +Which part do you think hacky? Also parseLong is there for a long time and currently the only problem we observe is is null +Thanks for valuable input +Patches welcome +I hope the rules are fixed now but we will see +Ive had one other user report something similar but we thought it was because the build was broken +Bulk close for release +Thanks +Could you try on trunk? It seems to work for me +OMG +While I think it is good to have WEAKREFERENCE support in the Sling Post Servlet it must be made sure that the Sling Post Servlet does not require JCR +Did it make it to the latest nightly snapshot or do I have to build from SVN +Additional changes to be completed for are described in so closing this issue. +Fixed in revision +But it occurs after import for site classic +and from gradle build scripttargetCompatibility sourceCompatibility +Can you try to run in binlinux directory this command +Proper fix version +this one has the patch now has +What use cases do you have in mind for concurrent access to Config classes +This is now fixed as there is no anymore WSDL files are not handled by the Contribution service +More details about the failed job +First version of patch with create and remove extracted + thanks for the status please update the Percent Complete in Description section +true max connections to and bumping ulimit from to client threads just tried this patch against the tip of +I ran the failed tests manually and they passedIntegrated to trunkThanks for the patch GregoryThanks for the review Stack +if it doesnt help try keeping the default location for the repository +There are uses in the JNI functions in particular the critical variants +You would likely get it fixed sooner. +Thanks +now load using bundle! +But this seems not the reason of nd tests failure +Id considered this and have been going back and forth on the idea itd be useful for other reasons tooI wouldnt want to do this as a partition parameter itd have to be a new column in the SDS tableIve mainly been concerned about whether or not its worth it given that we can approximate it by checking if the partition is bucketedsorted and the table is not +reopened to be verified it good on Ubuntu with JBDS in Developer StudioVersion Build id Build date +John can you handle this in M +Running Graffito with HSQL required HSQL version +Of course you must also put and into lib testcase was run against groovy the test case with Groovy and Spring +RTC move of all unresolved issues from to +Can you point out what you needed to do to make this happen in the Camel build case? The current source builds as a result of the source plugin in the Apache parent pom from what I can tell and takes the name from the root pom which is the oddly named one +I did some regression testing +Oops forgot to put the issue ID in commit msg +This version of is dependent on WPF which conflicts with our desire for this to be cross platform +I suppose a secondary page such as the one employed for choosing feature variations for virtual products could be used for handling assembled products which could provide the type of explosion etc +Thus it is necessary but not sufficient +But I think if we are going to use this static imports we should import specific members so we can find out more easily where the imported methods are coming from +Thats clearly wrong +taking this off the list because nobody is signed up to fix this. +Merged in the branch in revision +Exported journal to simulate OOM if increasing the on the server doesnt resolve the issue +Please let me know if I changed the intent of this JIRAJacob any chance you have any doc updates for this JIRA? ThanksNo I do not have any updates to the docs +Im kind of with Todd +How about removing it instead then +Could be relatedsimilar to +unit test making sure the old path still works has been added +Uploading new one with comments from dhruba +This is high on my list of nice features to implement +bq +less warnings than trunk +We tried the fix on our end and no more errorsThanks +The Maven plugin should rely on the log API provided by MavenMy main concern now is to find a way to keep the with previous versions of Maven +bulk defer of issues to +This will ensure that consumers of Avro C code dont need to have autotools installed to build itFor pure autotools projects a simple make distcheck runs unit tests for the package and if they pass creates a tarball ready for distribution +Hey Renat +Does it solve the problem +Let me know +It is widely acknowledged that the responsibility driven design and the separation of the orthogonal functionality as good design principles +Mike write From the stack it looks like this is a hard upgrade issue +I just wanted to confirm that this has been fixed in trunk +I have committed this +Definitively not urgent + +Attached a version of the patch that applies cleanly but doesnt clearly show the changes to +Allowing the webconsole to just import from the container instead of bundling it itself seems to resolve the issue +Im starting a Windows vm +Thats quite certainly not what the user intended when specifying a and also deciding to specify some custom editors which is why custom editors always have highest priority +yeah when working on integration tests I usually run mvn jettyrun from a shell and then run individual unit tests in my IDEIll fix this as soon as Im done with the integration test Im troubleshooting +Same patch with trailing white spaces removed + +The open source Xalan does not have localized messages +I have added this info to thanks for this be added to the download page? Or install page +files from Copy the files and the jBPM libraries from the jBPM distribution to the platform + +Im fine with Allens suggestion from Feb Conceptually it is not different from what I proposed earlier and looks good and consistent +It seems this is an intermittent issue +Editor is actually that is Java Editor +the way I have fixed this issue is wrong +Rob Does the +It even lists junit beanshell and yaml among its mandatory dependencies in its pom! We can work around this broken behavior with some gradle hackery but alas jcommander is in even though we only use testng in an embedded fashion in our testng module +Siying the new patch looks good in general but can you open a different JIRA and upload your patch to that? +Is there anything I can do to help move this along +Attaching junit test case to demonstrate the issue +Committed to x and trunk +fixed +I have committed this to trunk and +We usually close all fixed issues on a release date +Aggresive inlining parametes can make OPT compile more methods then JETWhat are the timings for release builds +yes reproducible and deployement successful without data disk. +File with stacktraces +Let me know applied in SVN revision by SVN revision +Create a new realm with the same name as another realm which is NOT loaded +shipped in +Filename was not set appropriately for jenkins to pick it up +Heres a patch that fixes this +We could introduce a separate transaction controller per sequence and synchronize on those transaction controllers +Defer all open issues to +This is already done for XML and JSON responses but not for all Clerezza based serializers because the interface of Clerezza does not support the definition of a charset +If folks like you use it and give us feedback raise bugs if necessary that would be great +Thanks for the solution +This would make testing easier as well as isolating the retry code +Thanks for the quickstart +Code checked in but could do with a unit test +Ill give it a try and submit a revised patch +BrianI used Internet Explorer to test the transformation result +But that isnt right +SDB could need to be optimized with but I have not enough knowledge on this database system to do that +Let me test this a little more and commit the fix +I have reviewed this patch changes look good to to trunk with revision +the bundled Snowball C source code +I am using from May PM + +Once the closest instance is found I look for the closest node in the chain of parents towards root +Moving to I dont think were ready here +Name of all the required Jars present in the patch +The first patch adds SSL support to and is ready to be committed second contains a extra test and enhanced pump method this cant be commit until the it Java SSL implementation to make it more SSL implementation extracted most byte buffer manipulation to improve readability +Changing assignee from Kristian to myself for backport +The properties file does have the right licenses in the right places +I wonder if this is because powermock +but you just did so you get mad until you realize you have to go back and save the server editor and THEN open the launch config +Update for tests +Patch applied +Will look at this first thing tomorrow +The issue here isnt what were outputting its how were handling http responses from remote servers +Can you please make sure it works now? I will investigate why the classloader was not gced in the first place +Let me know if there is any way I can assist +The mirror is now available at gitRead only for now though you can use git svn dcommit for limited support +Here is the JCA component +Updated description to reflect the above this is a general issue not related to + +I think this should be able to help clarify +The rest of the changes are still fine +Maybe you can paste this in for meIt happens in the Grails OOTB scaffolding template +Example added to wiki under Streams +ideally the would have these semantics in order to handle any configured that may perform background syncsI agree +It might also help solve the problem of determining if just didnt do its thing +thanks Flavio for those corrections +You can first review the code if any other questions i will upload a new patch with removing commented lines +Committed at revision +Some classes are not thread safe +Closing the bug. +Sure I can do that in committing the patch +Tested and applied +I hope Scotts can help us continue with maintaining and updating the solr store +Move all issues to waiting for this to be fixed in +Corrected URI resolved Release +Can somebody help review this +I believe this is the basic test case that distills the issue +s pull in Joe Bowser since hes the owner of the Android component +I will check in after rerunning tests +So I guess the resolution here is WONTFIX? +scanning is not atomic +however cant seem to see the plugin development section appear after I generate the docs +this is a surefire problem if it works with forkModenone but not forkMode has been fixed in the latest surefire plugin. +pretty cool Alexanderflexible includes the missing documentation update +If we need to switch back to Jetty in the future for some reason it shouldnt be hard +Any chance this patch could be in the next alpha release? Thanks +Maybe that was an intended change to enforce consistency +From the stack trace it is suggested that NPE happens after failures so that might be the root cause +If there is no way to solve it at least we can document it as a known issue +Thanks Henryk for the patch + can not in SMX bundles because its LGPL license which is not Apache license compliantI will add this bundle in Extra +Regards the image quality problems differing DPI in the source images to that of the rendered PDF cause a lot of the problems this will be fixed to some degree in future versions +Shall we put setting up permission directory and files in a static block to achieve the above instead of doing it in a method +Fix confirmed. +Fixed r Just needed to replace the references of with +It seems like my proposal would be more trouble than its worth +rebased should be working now +Theres been a lot of back and forth with this one and honestly I wasnt sure what the outcome was +had an extraneous file in the plan to commit this afternoonPlease let me know if you have any concerns +could we discuss that there So this does not solve that issue +Just one comment Can we please change fails to failed +Can you use LOAD or FOREACH as column names +minimize diagram coverage area reduce empty +If this is the case then we can treat this ephemeral node as written by the broker itself but from a previous session hence backoff for it to be deleted on session timeout and retry creating the ephemeral node +Should not it be set to be true whenever it seeks to a new node +That is probably because I did not look in the code to start you have an understanding of the issue would it be quick for you to cook up a patch +In fact I understood your example even before And yes it was correctly interpreted by Jira so from my point of view this notation already reserved for anchor links on current page like Lxxx but I agree that first proposal is too generic and we must find a better notation. +In the schema is different enough that people will migrate to an empty database when moving from so we can fix it there +Himanshu Are you planning on creating a so that can test +push out to +Are you in a position to test it? I wasnt able to find anyone who could provide much testing before releasing the rarHow should the URL be constructed from the database name +It is possible to write one feature at a time and basically write huge shapefiles without any memory issue +Sorted bugs in the Junit test and added a few other improvements to the test +I think this is ready to go +Added issue to Known issues section of docs +committed +Ugh +I see the changes in the was there a class? I dont see it in the patch +My question was more about the use of +There seems to be an issue with the transformation of the result let me see if I can get some details +I committed the patch to +Committed +What we have is a stack trace sanatizing tool that is package based +with moving to the axis in Version +Chose to call the new parameter editTime + +Max do you want me to remove all values for the archetypes +Closing issue Oozie is released +Ill put together a new patch for this +What does a compatibility centered report look like? takes two sets of source generates one XML file for each set and then uses the two XML files to generate the report +Sorry I wasnt trying to say the two were related +Two files were slightly modified see the diff +patch i added a test for this +If ommissions are found or modification to existing help files are needed it can be reopenend againThanks JacquesRegards backported r in r r r +All tests pass again with weme +Patch committed as is. +Bravo Shai IBM! This sounds awesome and I hope will prove how modules will help lucene and contribution Shai +This depends on +secs are tolerable but sec delay makes it much harder not to notice the delays between executions +I post on dev about removing the error handler mbeans from JMX as it would be easier than trying to keep properly track of which error handler is context vs route scoped +Patch which logs a message about class not being found in the annotation and doesnt add the null class to the xmlschemas Set to avoid errors starting up the endpointPatch applied although I moved the log string out to the clue what that mysterious generated package is all about? +Had I to do it over again I probablywould have argued against including even in the basic packageBesides all you have to do is redirect where writes to and youveaccomplished what this bug report wants +Tech Pubs +The with a change for used +Thats silly +Linking +These blockers and critical issues are resolved but not verified +Reported sys its working in RC +Yes Ill check that out +That was fast! Thanks +I saw it but as I said on the downside the fix carried a significant price in terms of performance +Need to attach a file so I can click the Grant license button +Closing old date issues. +The scheduling is split up between the resource manager and the application master +closing it +Closing the ticket +Thanks Kihwal +How about this change +Mi MaximWe usually dont let translators know directly about releases however if you watch the developers list at all you will get notified +If you have a better nameregionLocker +Part section says that NATIONAL CHARACTER is implementation defined +The updated grammar landed in d +Could you try with the head of the source tree +A patch is available +The core parts are in place by no +Passes locally for me now +Fix for for +CXF also supports the schema validation +The workaround works for me but there is a chance that it may be permission related +I dont think the failed tests are related to this patch at all there seem to be instances of similar failures in the previous builds +Mirek is not targeted for EAP +Just try and imagine a life without grep +We need skipping in our in order to be able to pass them as filters and improve search performanceOne could Sort a every time its used +attached as you for your +The availability of such an option needs to be an controlled at the administrator level +Sorry I forget to say that error was dtected on Windows Vista with Visual Studio Express Edition C compiler +This is an awesome idea +Hi allCreated a simple logJ logo +I agree that passing this in via the constructor is better +Create a wave +It was JRE inconsistency problem +Patch committed to trunk +Enis Soztutar you want me to commit your windowing patch at same timeThat would be great +Thanks for running these separately I just looked in which is where the opHooks list is being initialized it looks like the opHooks list is always being initialized even if there are no installed +Really nice job on the code very clean +what if you add getter in in and later add the corresponding getter infor i think you can just add getter in +This works very wellThanks a lot Stefan! +This will be used for all JTA operations besides suspendresume in any case +i would start by checking what version of the tools you are really using and otherwise in ant you can just put a in the classpath to tune logging +Fix for annoying problem +Splitting the project into has probably made patches trickier +Resolving this since I am not able to reproduce and the resolver change for framework are intended to address this issue. +Revision +Thanks Vinod! +Committed to trunk +This patch includes Template Code addition as a result of new design patch contains both and changes with a test for +is not so general thing +I read the original issue from which this was cloned and looked at the code diff but I still cant be surePlease confirm if the description is accurate and if not what needs to be changed +tiny patch to bump to jclouds now that its in Maven Central +Though it would be trivial to fix with the current setup +Didnt see any objections tests represent this behavior now. +Igor sorry it took me so long to get back to this +Sorry Brian Im not entirely sure what youre looking for +Please dont forget to check Digest Authentication on Jonas server after deployment problem is fixed +Ive revised this patch as follows +Sorry Im having a hard time finding the commit anywhere +Francis if you are agreeable to my modifications to your patch we can commit this +It would be a bug if XSD would allow but it was not working as designed +The RPC layer can be smarter about passing only declared exceptions and converting others into the remote exception in the RPC layer I recall doing something like that in previous lifetimeIf the is the wrappercontainer then yes client side should unwrap it but only unwrap what is declared since these are the only classes available on the client side +Committed revision . +Added patch to fix the test case +But the important part is that it works now +Have started using this target as the foundation for the and works nicely for theblobmaster want to give the thumbs up from our side and subject to approval it would be good for it to be merged into As of Teiid Designer we no longer depend on UML +Updated design hadoop QA would treat pdf files as patches +Rename that file to JBMVERSION or some other name that is less likely being used by some other +I tested this and this seems to be fixed +Thanks OliverPatch applied to at repo revision rPlease check that the patch was applied as you expected +for the changes in the patch +It seems this part of the release cycle is a bit neglected +BTW sometimes the behavior you describe here is desirable see and. +BR code should deal with inactive regions but NOT propagate this to the recipient since with BR the data owner is the only instance capable of activating the region in any meaningful way +Patch to not bail if is looks goodWe should also update the comments in the to state that CC is also acceptable +Move uses a method balance +If thats true we can document the ability to the projects would be a new feature well keep this issue open to discuss that. +Related to to reproduce this issue on APIPROVIDER and Which was working fine previously +brunch w takes more time during the initial compile +Well it wasnt mine I just saw it posted in the cfwheels forum but I missed the extrenal +Seems behavior is different across here I was half expecting issues with this test since Iinterrupt a thread +Havent tested it but it appears to be reasonableIs there a reason for not having a timeout configured by defaultCheersFelix +Regarding the test result testmapreducestacktracehadoop is a flaky test tracked inReviewboard has the latest patch +i think it becomes null which is local file system i think +The release will not be in time for the release so I added labeled jars until we can get an official release from jUDDI +Yes I like injection at boot time much better than adding methods to the for each and everything +New issue submitted for this +actually get access to the test file list? If so can we could create a utility class to register this file list in either a static variable or a but this will require the tests to depend on m to register these elements in the first alternative might be to supply a utility to parse a POM and retrieve the test from that +v upload a wrong file +Alexey notes that this seems to have similar symptoms +committed as r +bq +Quick question why should this be namenode metrics +While working on some other issues on branchx I was nearly bitten by this +aborting the creation of the proxy if it contains operations is very restrictive +The tests passes on Xen and are failing only on KVMI think it would make sense check if in prior tests nics are not removed and causing the problem +What is the actual database type for the boolean column? The translator is assuming that teradata uses a numerical boolean +Thanks Harsh mind updating with your report? is kind of alarming +Committed in revision +I created a new directory and copied over the directory containing the bean files into the new directory then I navigated in the IDE into the new directory and it was able to see the Spring bean files +The patch changes several of the existing canopy files formatting unnecessarily the patch introduces author tags which are not according to ASF policy +In a few places I made use of the new PARSER idiom +Hi KevinI added server start and stop stuff in hudson itself so you dont have to do +Moving items not being worked on afaik out of M Fix issue out as part of JIRA cleanup +update was backported to just not correctly +Only the tnstances in other ec regions will route to its public IP +We are using a single shared for cache mode and for transaction mode so im going to have to create separate definitions +Note that if the has a child that is an but not a then this may cause other issues +Heres a patch which addresses the issue +Ive committed the new gold files. +But i guess it cant hurt to force all the deps to be the same so ignore the scope issue the validation has revealed valid issues +That being said it appears that DATAURL isnt working on the Camera API which isnt a huge loss because sending Base data in the DOM is poor practice +Tomaz could you ping me about this when you get a chance +Fixed reserved keyword issue + This fixed the hanging problem that we saw while testing the patch to + refactored and renamed to created a new class Cli for the balancer command added a new option policy +Refactored your test into srcbugs directory +So I agree with your assertion +awesome Chris can you assign me this ticket and the tickets +What kind of errors does it produce? can you give a stack tracce +install is filled in trunkCommitted revision . +Oh i missed that you just added the files +We dont have any explicit support for siteminder using it is simply a matter of looking for a particular header name in the request +Im documenting the Display outputs and it would be nice to have all of them +This doesnt sound right + +Forgot to mention that all unit and tests committed +compact files +Updated patch to include a fix with issue found when values cannot be compared +The attached zip contains an example +thanks so much for your interest and for explanations +Resubmitting the patch +Whether Z is here or not doesnt change that at allBut if you drop Z then Y bumps really can only mean some change that dont break backward compatibility and you dont convey anything moreNow different people have different expectations concerning version numbers but Im pretty sure that for most people communicate the idea of a minor increment over and so probably additionschanges that they can check when they have time while communicate the idea of more substantial change something that is probably worth checking out +The four lists are operational but we do need the additional moderatorsThanks +Koen I remember you saying this were supposed to be fixed months back +waiting for build to finish to dont see this anymore assumed fixed. +Thanks guys +It is a superfluous namespace declaration +Committed as followstrunk Touches the following fileM javatestingorgapachederbyTestingfunctionTeststestsjdbc +Response times dropped back to normal and throughput went back to normal +patch for proper buildbot submission +I assume this is a simple review and close +Could you provide us a new version +Ill add the full stack trace when this appears again +And found that WMS tests were still failing because they use a different setup +We didnt wipe it out +Maybe the problem has to do something with path separators +It he header is Connection close the keepAlive method what should return true or falseThanks Selva hope this could be the patch desired to fix thisi think to do this one possible way is to add an headerIterator in the +Ok heres the patch for normalization +Patch looks good other then few tab keys instead of spaces +The clientppid is simply reported as a way to help identify the process running as the client +Fixed inLucene upgraded to current version +I think we should be more explicit about this +Hum Ive tried all of the above with but the parser does not get generated only the nodes and the +as mentioned above I think wed be better off distinguishing between subscriber types via fields on some new Subscriber object which could additionally allow for the namespacing needed for +Im currently working on a reworked patch for that +For normal they should still inherite fromDoes that sound OK +Attachment has been added with description Full and Subset embedding patch replaces the previous patch that I uploaded and contains further improvements tests and more bugfixes in and fonts +I clicked on the wrong doc and wasnt thinkingI see the problem has been corrected in the docThanks for your patience with me and polite responses. +To the best of my recollection this defect was fixed +if I can get some help reproducing it. +DBRE is incremental in its operation +rev for branch submit the new codes and delete the from serverRex +Unless this happened again it should work this time +You are right +Let me know if you need some other visual aid to explain what I am doing hereKarthik +Say I was to write some up for Javascript what should it look like for submitting a patch? Just an hmtl doc +Note to the committer the following files are new with this patchsrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdataThe following are removedsrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdatasrcorgapachebeehivesamplespetstorecontrolsdata jeremiah +getter and setter should really only do as much work as possible +Amila Lets try to fix this for this compiles a wrong in svn revision thanksdims +Refresh cloud via +On the other hand there are other major open issues +hope you are talking about the first part of the JIRA code here +I see that dfs count of tmp does take into account the quotas contributed by its snapshots +Could you provide an example of the kind of extension hook you have in mind +Thanks +Later in the same test run we see there is no space left on the log device so thats probably the +I thought updating all the resources to add in the additional update method would be cumbersome but as you said it hasnt been a big deal on your other project +r ldimaggi ldimaggi Mar +commit for will use cxf to CXF revision . +Tests failed and seem to be real bugs +They will be in the which is the central location for the +Good point I was not thinking about performance but more about completeness +Reviewed and tested everything is ok +I am on this patch too +there is no community release +For full details see the documentation in Teiid Documents +It will be nice if there was an even partial dump of what ever code has been generated for the query even if it runs into problem because that would help with debugging +I would just like the components to work out of the box with no manual steps involvedThoughts +the Interceptor interface seems to be there as well in the that I uploaded +First download the file with the hcommandLink and then reload the content with the works the file with the richmenuItem and again reload the content with the the download is executed again +Changing to an enhancement request +Added the clone fix to +In this case if we receive an empty ledger flag as a hint we have to switch the hint to zero which means that the client will start recovery from entry zero +Thanks for the report and the patch! Patch in CVS would be great if one of you could verify +Patch applied as expected +Right! Nice catch +I attach a small patch to For war project depedences create a linked resource to webapp of war project to maingen +Test case includes all modified files +but yeah its on my pet project list +Should i make a new patch with this metod or this one can be commited? +Patch Applied. +now the git version compiles +I should get to it next week +If issues w this patch we can make new issues to address +Fixed at revision +Please oh please rid the world of expert properties +The archive contains a parent POM and a POM for a jar +This isnt a tool for testing correctness so generating new cases isnt of value +Phils suggested fix for will also address this problem +. +The findbugs are from and unrelated to this patch +I may have done this incorrectly? Could you please help out with a working test case that implements what I tried to testWas it the quotes which were in error? +The first rerun above OOMEd +AFAIK thiswas fixed a while back +still wondering why the property was not found +throws as is or log the depends on the value of the strictIm not sure if this is a good feature as it only make sense to developersRegards +Pete on mailing list That does the trick! Thank you +Ill assume this was a bug in and is now fixed +I will modify the build system accordingly to use the new plugins and make any layoutconfig changes when I figure the difference out +solr would not be supported within a client app because there is no Solr instance in the appIn short the solr URL refers to the Solr instance available within the current JVM via the current classloader +Part of the fix for JIRA +ok +can you please try itthanks tried and your example but the results were same +to all involved in this issue the comments here have been exactly what Ive needed for Release Notes +Not related to ll need to check but perhaps this issue was fixed as part of +Please note that target release date too +Need to check if this issue happen again after we removed the subclassing +And the easiest way to get started it to write an integration test under srcit +title to match after release +moving to released +You have no right to get annoyed about anything +Sorry this is not a major bug + +The public beta release works fine with Spring GA as does any case Ive added adapting through a proxy when the JPA API is present +Can you attach +Thanks for including a test case +CheersAnjana +There are failures in the auth and indexing tests +let him know I havent forgotten about the documentation of +Some classes which indicate the problem +Isnt there any way to have onlymvn clean cocoonbuild jettyrunI do not want to have cocoon wared or installed into local repository also +Fixed w Joels patches. +Have we thought of features like hive and hbase import? We would need to have hudson integration for windows related builds +BULK EDIT These issues are open to be picked up +This might be related to some optimizations that were made as part of the tweetstream example +is is done +Specifying an should work without of course +Attachment with Jigneshs consolidated changes +removing them from the code base + adds section Initialization Module to and replaces all Destroying the VM occurrences with VM Shutdown +I would be to adding as a dependency for for the feedbackI dont know if its the right place but my opinion is that DBCP is widely used and it seemed to me more relevant to add it in DBCPNever mind here are some answers +Personally Im a little bit confused about the whole concept of deleting the latest put would you do this to get access the the second to last entry or when and how would this be usedWhen it comes to finding the timestamp for a put maybe we can return that from the server so that if you wanted to keep track of latest you have that option even if you dont set an explicit ts client side +should be on the mirrors now in the obvious location +Are there other datasets I can try +This is now working in the latest svn + just committed this. +You can just describe the documentation changes that are required +I make ear last and war second is a simple junit test that show the doesnt always return the correct order +Ping Lucene and Solr RC is being voted on right now and Id like to be able to distribute it at the beginning of next week assuming the vote passes +when attaching the patch otherwise we cant use it +M will be fine +In fact my matrix package uses the Colt Functions object for my own matrix implementations which is one reason I cant distribute my own matrices very easilyAny comments on this general structure? For machine learning I have had very little call for complex numbers +Some hints on the mysql documentation site seems to think mysql will accept most anything +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +Attached +I placed to screenshots one is displayed missing Seam Tools when BIRT and TPTP is not installed other displaying the Seam Tools when BIRT and TPTP is thereIs visible at your please try to reproduce it +Potential let me know if there are any issues with patch otherwise I will apply in a few days +server down s weird CARGO normally does not use proxy settings when deployingCan you please mvn with the X option and attach the output as a log file +Whileits starting the status shows Started +Your point about the difference in behavior between and is very important but we still need to lock down where the difference is + is the right one please provide a relocation a relocation to pointing at +Hi guysBased on the latest codes I think we can realize what the issue sayFor example using a wsdl file to generate a web serviceThe first time it runs okThe second time it will ask if you want to overwrite the impl class and the +Fixed with IT in if you take a look am planning a release in a few days so let me know how it works out +Is there any easy way to reproduce these clients getting stuck +As an update an implementation option will be given on the jsf all command to choose between Mojarra and news! Good job +So youre likely to see fragments with high discriminatory value +For now id say lets bring in this performance improvement first and we can deal with the customization of the comparator that is used for sorting in another Jira issue that i would appreciate if you would createRegards do some testing on your application and let me know if works now as expected +Youre of course right +Im hitting the same issue at Clebert I dont think its a bug +Actually I had misread this +Consider how compression codecs plug in today at the file level and in as options to etc +Also renamed the to I do not see the file being removed in the patch +Two patches one for Trunk and one for the branch are attached to fix the problem with this testcase hanging on databases that do not support a locking timout. +What if HBASEHOME is not set? I dont think HBASEHOME or HADOOPHOME is required to run a hbase or hadoop client +Note that chars are still a problem this way snakeyaml will complain when you add them as the initialtoken +Thanks Siying! +Should we call it instead +trunk +Chunhui on patch Chunhui +I tried to keep the supported productions that made sense and could be translated to valid SQL but had to remove some syntax elements +Ugh +Mahadev you need to rebase after +Thanks for great comments Nicholas! I addressed them in v patch +this bug is introduced in since it change parsing packet in a different way +Apache Chemistry has a couple I believe and we have a CMIS connector +Another option for this problem is to concatenate the thread id which might be a safer +Is there any reason why they cant use the solution proposed there +This one job blocks gridmix whenever it happens +The above comment is based on my master branch setup +DavidIt is difficult to diagnose without knowing exactly what you are doing but if you have multiple all running and all going to the same Solr instance that might be where the problem is +I forgot this to mention in my previous comment Please make sure that all files that you add have the Apache License header +In the actual plugin i get no NPE but it is a good idea to recreate the deleted file with default values +The issues for the wizard have been reported in separate +If multiple threads access a format concurrently itmust be synchronized externally +per the added of the linked issue modifies the SSL configuration to allow defining multiple trust stores and assigning them +This issue doesnt exist in trunk +Bulk change of resolved shipped +It should have challenged me for the password first? I think it also should have diverted to my netnewswire app maybe the needs to be sent too? it actually should have skipped these as you said it should work with the guest account +ping for the info Rudolf +That way the other modules could simply use this dep to pull in all shims +No exceptions on client or serverWhat do you mean with used the same client scannerLastly does this happen every timeAny extra info you can provide might be helpful +Kristian you are right I labelled the heap reports wrong +We are trying to work through some camera issues +We can implement it by checking for config variable SCRIPTFILENAMEKEY in block function +At least then I for one would label it a pure SQL solutionWhat do you think Saurabh +Just changing the text to be the same as the other improvement it looks like Tim has outlined the features you plan to implement in the thread do you have an outline of the changes to your documentation with these updates? +So I dont close now but will close later with or wout Sergeys files + +Thinking further on this making this completely automatic wont work because theres no good way in ant to make one dependency of a target fire before another dependency of the same target without linking the two dependencies together +He has written a paper on when to parallelize based on pset frontend experience +Thanks for fixing! Just curious what was the fix applied? +postpone this to post. +Tom can I close this you thing? +Thanks for the test case +Ive committed the simple fix for the flush issue and added a test +Please verify +Committed to trunk +fixed in revision +Old smooks editor was discontinued +Also my tests confirm it. +This will not detach the faultdetail element as it doesnt have a parentFor the second request when we set the detail element it will detach the fault detail element from the parent +Marked for patch might need not +It looked scary if we ended up doing anything more complicated than just trying to send a single email per commit or some variation there of and even then abstracting out the source generation from the formatting aspects did not look like buckets of funEitherway Id appreciate any input people have on what they think would be the best general approach +Can you provide a a test to demonstrate the fix please +link to Release Procedure is added to getting started chapter relevant resources section All Cookbook articles are looked through links are created to them from dev guide where its is done +Unfortunately I put the wrong bug number for the patch that fixed +It made the Hortonworks teams morning +The problem is the conflicting constraints but there are no error messages being shown +Thanks for humoring my inner pedant +Not sure where this issue is +A maven project with a replacementextension for the actual ant tasks +Slip out of issue +There are few tests which I cant fix now All my tries have no is causing their failure +I would argue that missing symlinks suggest about the quality of the installed packageGood point +Assigned normal urgency as that seemed to be the intent from the triage comments +This will fix the install updated Designer use caseWe could try and discover the latest version of the Designer jar but seems like a lot of work for a resource +Many thanks Mladen! +Trivial change to add one +Patch applied to trunk and branch. +in many environments there are very long sessions its not forever but it approximates forever +Heres a patch for review + +for message change in to trunk and. +I dont suppose anyone knows of most places the zip is in use? Thatd help to test this +This change was made in the merged PR for with JBDS L +However maybe you will have to click a link in the TOC first for the show to begin +I will look further but from my quick look at the code I do not think server sends the protocol version as part of the handshake for a ping +Patch attached +If it turns out its not used that often we can yank it +Addendum for which allows better grouping of stack trace by thread +for review Nitay. +If TLSSSL is enabled then the default SASL parameters should be relaxed to allow for the fact that the channel is encrypted +Okay this patch has single threaded support as wellThere are two tests in unit test that fail quite often in both single threaded and multi threaded mode but that class is very opaque and I cant tell if it is a bug in the unit test or if it is exposing a bug +I made a move on this one and it fairly easy butThe systemaddkeyspace? What keyspace should one be checked against? +When app finishes event could be sent to the expire monitor and it handles it +Jeff this was committed right +We kept the XSD files for future reference only they are not used during build time or during runtime +Thinking about this a bit more + This bug has been marked as a duplicate of +The same issue occurs if I generate the WSDL using +The patch has been appliedPlease verify +to me that sounds to have a similar reason as +Is that possiblebq +And take are of your back +I only had time to try the EAR +Drew my concern with is that we have a property which is initialized in the properties file but then overwritten by the xml +added we discussed this as part of the priorities for the console in some email threads and phone call please update the fix version of this issue to the CR if you didnt anticipate to get this done in this release. +Since this patch does not modifed Ive filed +Meng interesting analysis thanksTo be perfectly honest Im surprised you guys are seeing this many memory issues with the JT +Hi I updated full launcher so that it can include bundles of stable launcher as default +Exported from Word however use rotate leaving height width +Ok +Thanks Stepan the patch applied as expected +Just setting forkTimeout is actually enough to trigger this it would happen even with forkModenever as long as forkTimeout is setActually the issue would probably happen quite frequently as long as the process outputs to stderrstdout at the exact time it times out with forkTimeout +Id suggest keeping this patch minimal and creating another jira for enhanced option handling +Attaching a patch against secure In testing this on a cluster on this patch avoided the Text file busy issue I was seeing before +Thats very interesting about the driver though Ill have to remember that +updates patch +I reviewed the html files and found that the eg for joinOrder incorrectly has the table names on the same line as optimizer overridesSELECT FROM joinOrderFIXED t tWHERE The table names should not follow joinOrderFIXED instead they should be on the next line as shown belowSELECT FROM joinOrderFIXED t t WHERE +Will run tests and check this in +Lets get a screen shot up in lieu of a test and if it looks sane Ill +Note that we got upcoming in early March as well as M in late would be great if you could submit a pull request on since thats much easier to consume for us these days +Please check whether neethi jar is there in the class pathThanks is in the classpath +Verified closing. +But trying this on DB that I have access to also fails +Seb could you re attach your patch and grant the ASF license? We cannot apply any patches that dont have this granted +Great +Back to the drawing board switching off patch availableNote I also added some carriage returns in the description and env just cosmetically so this bug hopefully needs less horizontal space +Did you click on Publish on review board? I dont see your responses there +Thank you Jitendra +Applied patch in r +Patch for attached +So we should wait until that release is out of the door before we can add the component +Can you try with and see whether it works +Sorry for wasting your time with this I will look more closely at the documentation in future! Thanks for your speedy response anyway +not enough time to test it by profiling the Maven side so I reopen the ticket. +and one more with osgi tried to prepare testcase for this error but had difficulties with it. +The patch also fails if the number of regions that need to be moved in one table is larger than the limit +to fully load the resources on instantiation and remove all loading support and functionality. +I just committed this. +However I still think this is an important request to consider adding to Sonar at some point in the future +small diff format +Need confirmation whether it is the right projectYes +What changed between and that triggered this behavior? +Ok thanks +JSF component depends on CommonWe have changed some API in Common +Patch applied at r thanks for your contribution please check if it is fixed as you expected +Lets implement for but given low priority +would look at change. +API is updated now propagating change to TCK and +Looks good to me +This patch depends on the patch for already being in place +Completed need to add more info +That would certainly help +bq +Let getGroups be the first one +What happens with topics is that some node stops receiving messages +No blocks seconds Worst case all secondsI made the initial size configurable and also disabled automatic growing of the array +On we will create the tests discussed here + looks good but I assume that all the calls to fail were meant to be replaced by a call to set e instead of x +As for the server side thank you for your intention to fix it in a fashion applied in your new patch if I can be of any help I encourage you to contact me at any time +Thanks! +Slip to fixing issues in JBDS +The only applications of observations are in where the weights across all the clusters sum to and in where the weights are the sums of all the points ever agglomerated by a cluster +The one thing we should do is to write a guide about value generation +Commited on trunk with revision accidentally broke a unit test in the examples +SSHprocess and mergingRegionsThanks +regex matching field names eg +He is going to fix the problem and remove the logging in +also fixed Praveens tests +I had thought that I heard that was going to be pretty compatible thoughAnybody have any thoughts on this +Requires patch +I copied the ones from the rc jar and they worked. +You will get the same issue if you try to create a Seam project without Seam Portlet facet using the New Dynamic Project wizardI think a Seam project can only be created using the Seam Web Project wizardAlexeycould you please review this +This cannot be reproduced with the release. +Fixed in ef in repository +Are you running under JDK? thanks Im looking into it. +Also added this falsebut not sure if it had any effect on the last delete of the child object actually successfully based on the exception I just posted + +If the constructor is used and a valueComparatorvalue is specified then a filtering by value is done additionally +Do you mean to say that for any job that is currently running and for future jobs this should be doneI meant for jobs which currently have tasks scheduled on the faulty tasktracker +Done please populate with your latest release and Ill continue with the setup +This now works in trunk. +reopen and closed with right resolution. +I would not like that does not care which order the Ftplets is invoked that is up to the API user +So your patch will not work for test source files that belong to a package +Would an for the entire tree be good enough or do you want to actually log which ranges had which key counts +A long time there was for example a problem with invokedynamic and BND because it didnt know the invokedynamic problem is what is wrong here +Okay can reproduce that with a test case and have a fix locally waiting for dont see the link between this issue and the commit you refer Bernd +Fixed with commit +This is the most effective solution because rename does not require that all leases be searched +After discussing with Brian we need to consider how much to add to this operation and how much could be accessed using authorization impl specific operations +Has any consideration been made to looking into this issue +The described answer therefore is incorrect and not satisfactory +this hits the same limitation of this described at +But I am not sure on the best way to realize this +You meant true otherwise false right? Thats fine to me I prefer to stick to a common idiom +That API isnt exposed the same way JDBC is so any notion of Data Preview isnt available even though you can open a SQL Scrapbook on the CPThe Flat file content structure expected for DTP is different than that for Teiid so we would need to investigate the differences and try to mergeconnect them in some wayIn the meantime users can test their FlatText File data sources by creating their Relational models adding to VDB setting Translator names to file and JNDI names to the names of the text file data sources you createdeploy to your Teiid instance +This is what happens when you are writing to the edit log and you get cut off in the middle by a disk full conditionWhen its done properly preallocation solves this problem by making sure that we have enough space before starting the write +I ran the attached and unless my test is flawed it appears is just as fast or faster than +harddeploy also worked but for some reason hardundeploy did not +Fixable for this release +It seems to be a hack against a AS designWhat do we need to do that on and not on? Whats the difference +I really appreciate your comments and I really enjoy our discussions +I changed in Scanner to Queue +leaderIds is not used anymore +I have gone through other issues and yes there are already similar ones with the +damn too many issues to cover so this slipped +The new queue instances are discarded due to the bad configuration but they are left registered with the metrics system +On retest the output looks beautiful now. +Im just still finishing up some of the tests +must be the one of the record resolution time +Thanks Ate for the patch this makes life much easier for Eclipse users wanting to get started with the code +Need to backport the latest JMS rar outbound as well +Need to make sure which branch is the depending it is from trunk then this issue should already been fixed by Dan +I use it only in client side +Closing this as cannot reproduce but please if you still do have the problem. +Patch updated with thatHowever the generated file shouldnt be in the source control +This issue contains verifying test cases +It really comes down to what does the name node consider live +As there will be no more connections than the number of threads at any time +Forgot to close + +for now perhaps it is enough to just let too new clients retry with lower version if they get an on connectionAgreed +Change has been commited although not using the suggested method +Phil you have any updatescomments +By the same token we should make all thrift getRow callsWe should have getcolumn getkeyrange and a bunch of getslice overloads +txt file similar to all other packages +The jstack indicates that the former call causes the problem +Im curious as to what you came up with so Ill be keeping an eye out +Same problem with Maven and Eclipse has been retired so wont fix +I updated to Version and found everything working what I reported not working in +The thought however was that the two systems while similiar address slightly different concerns +Amars patch for MR incorporates the fix for this issue also. +need a test fine with Hibernate rc +Thanks for fixing this Robert +I was trying to back port for our version +I would expect that such problems are logged at info if not warning level. +Bringing into +Its fixed in the nightly site +Thus you have find +Please share any feedback you have via this JIRA +Committed to branch +Please verify +Committed to and trunk. +I dont think that M has a java source artifact right? Even if the patch sounds like a workaround its really needed +Changes look complete now +Also listening for viewDeactivate and Deactivate events and doing an is an oversimplified approach to working around this problem +Not clear what this means can you give an example +reopening so I can modify fix target to +I guess the collection contains nulls +reduce the priority since this works fine with the axis think this is resolved but marking it as a blocker for so Amila can and close out the bug if so +LGTM to trunk and branch. +Do you see other already checked in tests fail? you might be able to force the issue by running with Cp but I thought that the tests were careful to use a consistent encoding +I can discuss with you what I see if the current problem if that helps +bulk close of all resolved issues. +Hi LukeThe Metrics is very excited +Fixed findbugs and javadoc warnings +Slava dont forget to add JIRA issue link into a comment when you commit a fix +and also will give a patch for TRUNK +I am assuming it is not yet time to use JDK as the base JDK to run derbyall on +Does introducing mean that there is no need for the vhost code to createNo I dont think so as is used in srccouchdb and will only be set when a request is rewritten +However im still getting the same stacktrace +Switched to Cassandra branch and dropped support for +Ill be updating the geronimo build very soon +See the patch for for a fix for the from thrown from +I think it is more painful than just that because users now have to manage the password adding some sort of ui to accept the password and possibly making scripts that could once run unattended now need an attendant +I will also update SCM +I think that the default for this should be onRather than add HADOOPCLIENTCLASSPATH lets make a new variable HADOOPUSERCLASSPATHLAST +Example project for the test +Being able to will require providing a list of scoped to the as well +Fixed with revision +Out of time for Beta +Hello GeorgeThe fix looks good +Attached patch that adds before semantics to graph moves and makes the required changes to implement them for the existing writable ll leave this out for a day or two for comment since it impacts the graph layer +A test case for chunked encoding +Because we wouldnt want to introduce new constructors for any new configurable value and because the Config objects are made exactly for the purpose of configuration +Is there another version +The default lucene range query has a limitation regarding the number of terms in the range +q test file? I look at the other +They do not have any Test Annotations and so the code will only be executed when an explicit call was made +For reasons that I dont at all understand this is caused by the inoffensive looking change made in +thinks he has a better solution involving a working OPHF +Next release +In order to check this open your JPA API jar and find the package that the class is in Is it in or in Note that Springs JPA support is capable of working with both JPA and We use reflection to support the full JPA API and SPI when available +Can you provide those detail +on commit +Several of these have been eliminated. +In my case I would select off of +Adds a rubygems option to the gen rb compiler target +Hi antHeres the updated patch on top of your changes please review and applyThankss been so many changes this patch isnt even close anymore +Remove the filter definition at it should work +Should this go into? The patch seems contained to me +Media Capture plugin uses now +Read it wrong +reopen since its relevant to fix once there will be an bugfix update. +backingType +Unfortunately I did not have the time to try but I do think it will work +Some want to treat Groovy in the same way Eclipse treats adding Java to a project +Would it be better perhaps to throttle by the amount of live nodes? It wont always be accurate but if you lost a WAN link between youd be throttling a lot more than needed at the time +Okay theres a new patch attached that uses a style consistent with what Kevin used in +new patch syncd with trunk +Ive deleted the files from the staging svn and Ive created as requested by rmed from the staging svn and filed to track the noted +Subsystem is not working with Beta +also look the behavior in java via the api +Services for checking permission are already in trunk +This has been fixed by rebuilding the packs on panel activation. +It turned out that the client was reading the data too slowly hence not a Netty issue. +I tested the patch manually by installing a secure cluster +Well announce the release on Monday +This second attachment extends leaving the base impl untouched +Are there corresponding tests for linux to those for windows +cosmetic fixes such as tooltips detail panel can be added in +setting prio to minor because there is a workaround +Especially today is Sunday. +In most other places if I could fix the code by using information documented on the web site I did +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +reopen for back port to fix version to match the version on the branch +Ill probably put the idtoken in here some day soon as well but that can be a different issue. +Yeah I would just put the impl with the spec module +Currently each docId pair is stored in a Map which itself is stored in the +But this option is only working if the file is used +I updated the testFibonacci and the test tests as suggestedWhat issue does the test test detect? I guess if the issue occurs an exception will be thrown because this test does not assert anything +However when I look at the public repos is at the level +gives a different score than with these queries +Added some logging too +Please see new patch +css +Ok that patch actually worked! My test failures were due to a locked file and not due to my changeIve tested this and my MDB now runs with the default subject and thus the works properly. +Hong I agree that changing the default behavior to include other contrib projects into findbugs etc automatically should be a separate JIRA +No more redeploy +Thanks for checking +I just pushed this +I didnt try Metro over the network as it showed such an improvement on localhost + Bug has been marked as a duplicate of this bug +In the case of the first code snippet it appears to be a client issue the broker is registering the disconnect but the client doesnt seem to reset itself to a usable state as the second connect call never reaches the brokerIn the second snippet there appears to be a bug on the Broker side +We are running +Design doc cant help but notice you refer to a variable doc in your validation function that doesnt exist +And if there is an empty XML document etc +class file will not be added to the dependencies +has been marked as a duplicate of this bug +But would still prefer sorting the test results when comparing instead of changing code to use or +I dont think it would be that hard to implement a duration data type +Just correctly check whether a node has been removed from the parent builder by checking whether it is present in the base state but not present in the write state +HB uses the timezone attached to the Calendar +confirmed working on trunk +It is not committing the transactions so I guess this could be the reason of the described behavior +Both work fineThanks for the speedy fix it is much regards might also add that this is quite a major issue for us +But then theres the problem that we say refer to the release note although it wont be mentioned in the release notes for that future release so I agree it may be somewhat confusingPerhaps we should keep the sentence for now and file a JIRA issue to get it removed from trunk once the branch has been cut? Then all releases will give this hint hopefully helping most of those wholl be hit by the issue while future feature releases wont have a reference to the release notes for an ancient release +Security holes should be plugged not ignored +unsign homejcostatmp can this task be resolved +The user just needs to know about the other issue +I can confirm this but after minutes of it worked yesterday and nothing changed it has never been a requirement I dont know why it still is +Removed the tab gremlin +When I run these tests manually they seem to pass as well +So I put in a flag which marks whether there are deletes pending so when a fields update is added and the flag is set there is an automatic commit +as per how the JMM does things and the described locking and lack of evidence to why what we have now is bad I am closing this +I dont think patches and have been applied +In addition to not wanting to have dependencies on openjpa I really doubt you want dependencies on the geronimo spec implementation jars +As soon as you can identify a test which has a localeproblem see if you can fix it using the techniques youve identified andsubmit a patch for that test +In most cases its your own patch in the projects base directory +and seem to be lingering in JIRAIs the plan to merge them and get the code into Solr +Committed the patch to TRUNK +Hi Tomasz I checked a fix in +Initial patch which depends on +Reopening this with more description. +Currently I am writing them as a standalone java program rather than a junit test +On second thought I think I will address this a different way +The feature is marked for +Attaching an example of noop compressionThe general idea is that would implement LZOLZMA by default and subclasses could override in the future to add actual type specificity +Applied to branch and it fixed the build +bumping down this priority because the deadlock was potentially caused by starting the workspace with an old installation of Eclipse and an installation that uses Cocoa instead of Carbon +Baruch my pleasure! Please provide any feedback about how the feature works out in practice +Lets see what hadoopqa says and will then commit +Each GETMBRSREQ triggers a round of GETMBRSRSP which increase the initialmember count up to its limit in the Promise blocking the discovery + resolved issues for already released versions +release +However I figured it was better to get it in the public domain sooner and let the iterative process do its work +Hi Anilhttpslocalhostassetmaintcontrolmain now loads the main view from workeffort I assume this is unintended? Also why have you commented sections out rather than removing them? after comments by Scott +bq +I think that is a bug but maybe Im overlooking somethingI think the bug is in your test case +I think this should be dealt with in or at least considered and then set for +Thanks Ill do these tomorrow since it is in Shanghai +It maks my proposal less appealing +I am inclined to slip this one except it has two votes +I guess I assumed you had Timers going to the Job queue +Thanks Jitendra for the review +Duplicate of. +Pls retest against latest EAP ER release +Bulk close for Solr + +Closing this bug as invalid can open a seperate bug for building Xalan onJDK +All done just need to start advertising it. +issue is fixed + patch looks good +So I dont think this is a major issue +Trivial but see attached +Stefan also is not keen on parameterizing examples +When the is first created let it have a private boolean optimized true + to close this as WONTFIX +lib filexdb I started HSQLDB as same as you before but currently Im not able to replicate that alias error even with new database +See and +class or use jrebel to have actual more real class reloadingOr are you reporting something else +I think this also affects every release after beta including. +is indeed the NNs view of the length which is why I think it should beconsistent with the NNs view of the last block +Can this be used as a recommender or is more work required +In order for this to work we needed to be able to order interceptors created from the same factory +Yea none of the tests run at all for me +Here is the patch for this issue +The scope has been changed from consuming just DB data +Looks good +does the cleanup which has the test case for this jira +patch to allow other qualified beans in check will now only be done for +Also DGM references are markedI need more testing before I can determine if there are any problems +See attachment for more output related to these crashes +Patch against +Verified on JBT +Doyou see that as a problem? Sent from my androido + Cactus EJB test web app eclipse EJB eclipse enterprise app have the same that work is well underway with AS all previous community releases are +Thanks for spotting and sharing thatFrom a first glance select from history for runtime behavior looks indeed weird +The initial SOAP action URI is now set to the request not the response +Hi Bruno thanks for the detailed explanation since I was not aware of the to Albrecht Grge and the first Vienna Hackergarden +context +If creating endpoints based on entitiesannotations would break the paradigm as far as I understand +The patch can be checked in nowThe patch does not apply anymore due to +even though the spinner control is released under Creative Commons if this was wrong i will upload the patch again and delete the old one +If a locationmap is missing a later processing will throw an error about unfound resources +The test used +Im not sure if we can do a particular item to get it fixed for the release +This is better than C s generic type inference for method invocations but not as good as C s +Brent a CLONE of what issue is this one +hrm +This is resolved along with the database schema upgrade fixes in +btw +Thanks a lot +im hoping to complete it by this weekend. +The patch Ted adds should be factored into any balancer improvements +Affects the snapshot build as well +please review it +Is that relevant +PR needs approvals +Disregard the issue was that libvirtd on the Target host was not listening on the correct port as configured in etclibvirt as per comment +Sanne doesnt limit the number of files but it adds a mask to the hashcode ints so that you have far fewer buckets +Fix unit test issues +Im deferring the implmenetation to to move it out of the critical path for +after has been committed +Since jspc plugin could expect to see different version of Eclipse classes +an elegant solution is necessary too +Are we planning to release soon +Didnt make initial investigation this seems like an issue with the that is calculating different dependency set then regular maven build and I dont feel all that confortable by changing that right before the release might cause other side effects so Im going to move back to +The new query engine in includes the system content in the indexes and are included by default. +What do you suggest +see is not resolved and should be reopened +thanks on the numbers nicholas +Yes that works for me +root root Jul tmplIn your set up you are running as user cloud which does not have permission to write in that dir +Marking Release Notes field so that this gets documented in the Release Notes +Also Id say we should just log it at DEBUG level for the Not checking caseGood thinking +I had just cut and pasted from below but youre right +Fixed on the branch of on +We should increase the prio of this +It seems that recently the code got even more broken +Here is the patch +Should we have a priority mode between queues +However I do not feel strongly about this +It may be true that this code block is causing memory leaks +Triage for +Use the ANT biuld would be the recommendation. +Lets discuss this in todays call +Ill post a patch after I burn through my email +Its definitely not suitable for committing at the momentI would appreciate any type of feedback +Thanks for the patch but since we include the generated code with release tarballs I dont think these tools are actually necessary anymore +Maybe this is expected behavior I cant tell from the my app I had a miscellaneous Factories component named factories and I wanted to change one of the factories in it from uscmRegions to which seems like a reasonable thing to me +Pls run the benchmarks at least times +I see something more like a new execute method that gets called during to allow some interaction or some other way to map methods to a super phase +At this point we lost the ability to configure it support was simply removedMuch later I needed to set some timeouts for freebsd blackhole reasons and I found I had no way to do this so I shoved the shardhandler into simply so I could set these timeouts in tests +Please attach the proposed patch to patch on behalf of Xiao Feng Yu r +Setting as a feature since this is not a container but this means it should actually discover the extension within the war and enable for that scope no +the modified patch I tested with +Hi JuergenCan you please review the patch +Flume has a couple of wrapper classes on top of mainly to simulate flush semantics in code that was originally developed with hadoop +New patch fails to apply too +Please let me know when you hit again so that we can confirm +I can reproduce this by following your link but can you please describe how you got to it as I suspect that would be more useful to us +I have committed this +I just tried with and I can see the same error as before +Hi AviSorry for the delayCan you please submit the patch so others can have a lookOur resources are also very limited but we can see what we can do Cheers +Does the current fix need to go into too +And its as he points out not invasive towards the developerSo something like ats at the end if the file was last modified on February th at PMThe same would be useful for CSS files by the way +I didnt do anything about that similarity can be less than issue +A patch which fixes the bug +Flash +Will update the summary once I find a good one +Ok I misspoke WSRPJ doesnt actually compile if you just change the jars + +I think were actually OK but Ill doublecheck +i think it is ok to generate errors randomly as we are doing but we should have periods of error free operation so that things can settle down +Specify the URL whose requests should be forwarded +No worries good to hear that it works for you nowJuergen +On the advisory in an the advisory is only used for temp deletion detection as an optimization to save on a round trip to the broker +This makes it hard to enumerate the true dependencies I have +Guys before it is fixed the current workaround is to not put new projects under Eclipse workspace folder +eclipseeclipse +Marc is there any chance that we could get some of your help +I dont think we should invest time in restoring the old behaviour +I was wrong about the parsing +I believe that is the best place for this discussion +says to use dots for separating all portions of the version +Patch which adds formId attribute to xjscookMenu of an optional form which contains this menu +pluginsRemove the server xml item which you find offending you +r +Attaching test log output from build affects versions as for now +Thanks Owen! + and completed +Maybe we should escape characters for query classes that the query parser can handle not for other query classes. +Sorry that was my mistake +Sure +Sorry for the troublesBut cant it try to resolve it in more intuitive way if only action is specifiedaction is also a standard form attribute this is why I have falled into the trap here +Patch applied +Changing Release Notes status to Not Required as already included in Release Notes. +If you attempt to use the cxf subsystem it will fail with due to the bundle not finding the installed cxf provider +Hi Simothanks for the update +I know for sure our data binding framework properly handles abstract types +So that we reduce the number ofbug reports thank you + for the lengths the lads had to go to to avoid a migration and hack in a fix for fact we couldnt do binary keys +Lets do this +default is the current behavior we will always do as we have done unless you make your own class to do something else +Hi Kathy I am attempting to fix this bug just realized I may not be able to add the repro in jira into the functional tests +This jira has been closed as part of a review and archiving of the aj project +Hopefully I can get the commit in yet today and you can take a look +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Applied patches to +Thanks Emmanuel for this notification +Im using for the proxy +Great +If not Ill submit it later this week +Been traveling most of January and falling way behind unfortunatelyThanks for all the contributions +Awesome! Thanks Andrew +This isnt configurable in but will be allowed to be configurable in. +Resolved bug fixes and new features are delivered via the update site +Depending on where the bug is that might still trigger it +That said we dont use the feature very often +The beanNames property of does not just accept full bean names but also bean name patterns like tx +Regarding duplicate connections it makes sense because the inputstream cache is per file and it is quite possible that the clients read blocks belonging to two files that are on the same DN within the window of reads +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Provide timely statestatusstatistics updates to subscribed users +This could easily be added what do you think +I am agreeing with the first attempt I think there should be a solution like generating GWT apps +And if that does not help then state what files you are opening and if there anything in the error it does not fail on some specific file? Just in generalMaybe you are not running with enough memoryTry and goto Help Report problem and attach the generated file +The component description was added to the guide +After loading a particular version of Derby I get a connection and check the product version in the connection metadata +Added class as it was missing in the last patch +Closing. +Similar changes will be made to jetty and minimal assembliesIf the server is started using java jar bin or binstartup the default instance will be run from geronimovar +My attached script DHESPO works if you use code instead of icode on line and following names works iCode code idcodeI have only seen the bug if I use icode as the element name +Is this still relevant? now only propagates simple types jms headers +Looks like this causes a real pain for people +That would argue for permissions rather than +Closing as unreproducable +FYI This should probably now be upgraded to the client +Johan this error isnt about the Mac issue is itYura what version of JBDS and what platform is thisAssigning to Maxim +I have seen this request many times in recent past from users this in my opinion is a MUST haveFrom my first comment we can not keep making different importers for different source types that generate same kind of data +The order in which the JAR files in thelib directory are scanned is undefined +More description in and thats the issue Im working against as I resolve this. +Correction it does reproduce inside a while forever loop +It seemed more risky to try and pull out DFSNAMENODEDELEGATIONTOKENALWAYSUSEKEY support then to simply remove the test +I accept that sometimes things are kept in the bad shape intentionally for example due to insufficient resources +For the identity transform they may be invoked and we dont see any bugs there with the latest code. +Attaching a test file and extracted text file of it +Merged to. +Are you doing that or are you using the binary distribution from the Apache web site? The distribution was built with VC so it wont be compatible with VC +Logback can do what youre asking +False alarm +Committed v patch to trunk +this is superseded by and +Attached a debug log +We probably need some minor documentation changes. +Yes please send a patch +So all intermediate changes have to be rolled back +Cant reproduce issue was certainly fixed while working on another issue +Dan buildnumber plugin is really very unfortunate because it will execute in any case a shell script calling subversion and would increase out build time with artifacts tremendously I am not tied to a name of the goal +So add this to all projectsAnd which notifications send to this role all only issues related or work and worklog related too +In the install folder runant deploy nativeit will replace the if you specify the cxf then it will use the you dont specify the it wont replace the web service stack it just uses what it comes with the AS. + was added in that should solve the example you give to large extent get and getQuery are protected methods you have more examples to point out? I think this needs to be handled on basis rather than adding protected getters as a rule of thumb +What more needs to be done here? The checklist looks complete. +Making this optional also has the advantage that we could change block size etc if we want to +I checked the code updated code in CVS head please check out the last codes and run command ant deployablewars you will get deployable portlet war in webworkdistwebappsdist directory +Please attach sample WSDL file to help fix the problem +Ive fixed it and the fix will appear in Remoting +Example of CA window with a type and a declaring type specified for each doesnt work if comes from a project source folder +Ill try to get to the screen shot asap but Im completely snowed under with work at the moment +We also need to account for things like JME where you really need to use the tools provided there so things will work at runtime +I think that if you want to use the transfer methods of it use should be extended in the flush method of but it seems to create an break in the design +trunk John can you test from your windows environment again with the latest commit I did +No just one image would be fine +JamesFor your information I have created junit tests for the component based on +Is this something or yours that is missing from the patch +Implementation for the +Some comments Lets move the tests under hod directly instead of under hodhodlib +somewhere +Currently the JBDS and JBT docs are built from the same source +Sample multiproject +Sometimes I even start by writing out the query as a sentence and then translate it bit by bit +I am easily convinced by numbers +Public key replaced. +how come it is this big +Thanks Andrew for pointing me to this and as something to look atI thought to also put the useful pointers you gave in here for easy reference record keepingAndrews email Both deal with the new groovy test case wizard +Tawfiks configuration is right please check CXF version is really updatedWillem +When setting the timeout I do see that the is still there after undeploying but there are no strong references to the short I do not see an issue so long as the timeout is specified +Verified by Tony. +Its true the flushes the response +Kevin is right +Thanks thats the kind of thing I had in mind +Committed +This problem is now resolved +If this happens maybe they fix it +Given its clear this isnt going anywhere. +Its Oracle version +This is the TRACE enabled logs of for you +Thanks. +Looks to me it is safe to remove the lines Eli proposed +closed +previously sent data is available locally on startup +Looking closer it seems to be that the majority of data comes from still has a reference to it it should still be in the queue and show up in jmx +IMHO it is indispensable to write tests for odd behavior or whenever a failure arises with the permission evaluation +Patch which includes the class in the embedded resources +Verified on ER the exception does not occurSo you can resolve it if you want to +Keith please commit the patch +Feel free to switch reschedule if you feel appropriate. +There are a lot of references to the Scheuler safer to defer this to by SVN revision +YAY! I think Ive found a reproducible test case! Ill attach a zip file with the details +I am unable to reproduce the reported behaviour now +Pushed to branch. +Great +Minor commentsWould be better if we rename updateReservations to incrReservationsIn the jsp all slot info should be together +Lets leave it open +Development for the release with SSL support is complete +The body contents are not Serializable so cannot be marshalled by the ESBJBESB issue has been rejected +This issue is fixed and released as part of release +Hi FolksI have updated the patch for removing the promoCode from cart +Groovy doesnt have to leave the building just dont use the windows installer if it causes said well investigate the issue +We will likely use add SASL thrift interface that is used in Hue +should work with JSF if you are experiencing an issue with that please file issue with described test case +should take care of this +Could you make a little step by step as I dont know what to do with the patch +MartinJust specify in environment under environment base testing is done under Mojarra Tomcat JSP is not supported +Kelvin yes that was my intention +Since the old method has been deprecated and the new one should be used Ill be fixing this +Testcases patchedtestorghibernatetestidenhancedtestorghibernatetestqueryplantestorghibernatetestqueryplantestorghibernatetestcachetestorghibernatetesthqlfetchscroll +Updated patch that tests Hoop interop with webhdfs file systemStill issues described at are pending +done +Previous behavior can be enabled by setting property force to true on the. +Feel free to add appropriate labels + +Thanks for testing Ral +I think can be committed +With bundle to the deployments folder their is no notion of set +svnadmin dump pathtofsfs svnadmin load pathtofsfs +Deleted minor typo convienence +Making sure every colspec element has a colwidth attribute makes these errors go away +rev for +Attachment has been added with description patch for has been added with description patch to disallow multiple image has been added with description update to previous test has been added with description update to previous the idea was never to make these part of the source +My previous fix doesnt resolve the problem completelyIll paste another one +Thanks Henryk. +Attached is a patch to fix this +Thanks for reporting. +Patch includes modified tests including small fix for +Mass closing all issues that have been in resolved state for months or more without any feedback or update +To fix this we would need for the range in the Merkle tree to share the node range boundariescouldnt we just take the interesection of the computed ranges w the range actually being repaired +Right now its too hard to work on because of the massive changes being made to the trunk +Fix capitalization of table names in schema dumpsJohn I created these schemas using s schematool utility +It works as expected in embedded mode but it is more important to behave accordingly in the case of Network Server where multiple users are connected rather than throw an obscure error of ERROR A network protocol error was encountered and the connection +Those still retain their marked and local after compaction even though it could be way past gcgrace +OK I tried +Problem messages are queued when member is not the keyserver this is changed when we have the key +The original query now works in hibernate after fixing my regression +Lol +Verified by Paulex. +The UI issue +I justed tested it on freebsd and bison flex from ports are needed +Up to you to verify +I guess Im the only person that has encountered this problem of mixing IPhostname with mapreduce when connecting via IP address +the actual has been added with description patch to support classpath resources +Thanks a lotIris FreemanI have revised the test patch according to your comments +So yes there are neededAlso the more extensive the coverage the better +Added module +It makes sense that need some canonical key they can always be known by +already implemented +defer to + +This should help but I think there another failure type in yet to nail +They take a and a Path as input parameters +Changed them to thanks! +Now known keywords are escaped in the namespace. +Try to update it to latest version +Even build fails without for +The crisp definition I had in mind was if disableAllocationtrue for a dont use that by itself for an allocation +Its really quite a minor change just a spelling issue that happens to span a few files +remove lease during delete instead of checking existence before writing to image +Fixed with all issues that have been in resolved state for more than one month without further comments to closed status +Thanks for trying the patch Ahmed +So would work unless I got lostMy question is then should this dynamic analyzer override field level analyzersI guess no it would only override the entity level analyzer +bq +I think this is a better approach than that in the Docbook case because it keeps the responsibilities separate +rThe messages in the look like theyve been updated but the ones in have not +Pots canviar lexemple si et passs +Completed At revision +Youre welcome to give it a try yourself I dont see how we could get any further otherwise +I see Knuts point about prefix SYSCS being redundant in the name but I think in order to be consistent with the naming convention for all the existing stored procedures it may be better to continuee the tradition of starting the procedure name with SYSCS +I have applied it +we just have to make sure ant example takes care of putting all those jars where they need to be +It has nothing to do with modern applications cores clusters or anything the like +The object has the proper notification on setStatus however that is missing from the +Sure here or separate JIRA doesnt matter to me +Cennot be verified due to issue the dependent issue is resolved in CR resolving this accordingly +I will fix thisbq +Thanks Devaraj! +Ive added a unit test and fixed the bug. +Ran out of time for full resolution of this +I guess we could write redirect status as a temporary status somewhere but it would be too complex IMHOJulien any ideas on how to set a redirect status without overwriting the current one +I was able to reproduce the hanging test and was investigating through the root cause +Should remove +The current patch would provide value in that you could convert an array to a Collection then flatten it then convert back +Committed at subversion revision +this patch should cover all the changes in the code to be able to run the plugin with groovyI dont see an updated grammar of updated generated parser files +HansCould you sum up what is meant by getting the SFA usable +Jimmy if youve addressed Rams minor comments commit Id say +Just forgot to close I guess +Now it is fixed in the trunk and the given schema is added as a test case. +Can you retest this one with a clean WAR file and Tomcat work dir +In theory there is no reason why this could not be added +This is great work +Ive committed the patch but we still need German localization of the following propertyError while executing bundle operation +If she wishes she can put in a define right after the iterate tag and push in this bean into request scope +The patch also contains the fix forThanks applied +WW Update fix version for +Ending tasks was added to end method of instead of join activity to have a complete prevention of leaving tasks detached from executionTest case added as well to verify that tasks are finished with correct Maciej I have tested the build against my multi instance scenario and detected an issue +I just updated and Im still getting this error on head +Thanks for finding the bug! By the way any time you have issues with Camel feel free to create a JIRA or post a message to the user list were usually very responsiveNot sure whats wrong with my comments to your blog either +has fixed the error +Marking this as resolved +This Infrastructure page UI is only available to cloud admin who is the person who set up S secondary storage so heshe already knows S secret key +Committed patchThanksChatra +Now it is set to something explicit + +Norman is also working with HQ which is current version of HQ in AS master +example of resulting in svn r +Added patch +Moved issue with commands being echoed to a separate JIRA +This allows alternate implementations to be defined +Lets see whether security guys will come up with it later +Ive done some work on this in the past +confirmed that the fixes works on both linux and windows. +The unit tests failed but most of the failures are unrelated +Ive just committed this to trunk and. +The reason is that thiseffort has been superseded by the other frequent itemset software that Robinimplemented +Just to maintain coding style consistency +I reorganized and edited the section on in documentation +to crate a database for you to play with on oraclei instance we have +Indeed its all fitting on one screen for me now +I dont think we need it in it is a pretty esoteric caseDid you already merge it in trunk +bulk cleanup of Jira versioning +Steven Ive committed a fix for me know how it works for you +batch transition to closed remaining resolved bugs +Only if no working constructor will be found at all the exception will be propagated to the +By rethrowing the exception we dont see that the requirement came from a fragment which means we dont detach it in some cases where there is a conflict +I think its kinda crazy to force one or the other +Thanks Jarek! Im attaching a simple README here per your advice +Using in place of where its used for clusteringGot both and passing let me know if Ive missed somethingThe plan is to ship this regression fix in and do the more expansive refactoring for +Whats with adding removing adding the tests here? Was the previous omission intentional +Patch file to update function tests +It is strange since it works in my computer +Closing issue due to no response +A couple other ways to do this A custom interceptor either configured once and used multiple times or configured per action +Hi JimThanks for your investigationCould you please supply a testcase in your patch as wellSo we can refactor in the future without regrssion +Is it not being enabled or is it that the comment is inconsistent +Patch to trunk +The report refers to the a test named conn but I am unable to locate any test with name +It may be better served by users just implementing a custom extension to satisfy their needs rather than providing endless configuration options +Thanks for the clarification +Sorry I spaced out once again + +I would however argue that the should not be part of itself and rather a separate component +Sorry to come in late the patch has gone stale +EDUMP IN XSL Should return java cp +After first use if no changes occur the following message should be displayed No changes has occurred since first use of this Quality Profile +HiGreat thanksThe files to pay attention to are this is the core of the additional startup scanning stuff is just a helper to print out a proposal for the configuration by scanning the whole classpath its optional you do not need to add it if you just would like to add the startup scanning stuff +This allows users to extend the and override the as shown by the accompanied test case +Fair enough +I will file a separate jira to fix implementation +Could be handy if we ever added direct support for +Someone from the Domain Management team will pick this up +This is kind of embarrassing +trunk +Ill give that a shot +fix committed to HEAD +updated javadoc +It caused bythe wsdl has a portypeaccessDB then we can get a generate web service interface I use the accessDB to find the java class so cant find I am confused why the user can get a correct result at original but then get the errorDid he change the porttype name in the wsdl? from to accessDB? for the error that cant generate impl bugfix release +is removed +This really explains a lot +That way I ended up with about the right relations for optimistic locking +The actual PPMC ATM are the mentors and all official committer together +Could we get feedback regarding this defect +Am planning to commit thisReviews welcome +Duplicate of +Its perfect +the step about enter a subject but do not put anything in the Message Body text +I did not know much about the patches before but i will try to follow the rules next time i do this +Sure a note is added to buildREADME. +asking jim if he has logs +Part of a bulk update all resolutions changed to done please review history to original resolution type +thanks +Attachment has been added with description transition to closed if someone wishes to restore one of these to resolved in order to perform a verification step then feel free to do so +The jhighlight jar file is included in the war +Usually takes about minutes for our test to fail +This is not an easy problem to solve +the problem is that a component has a reference to itself in the variables scope and this produce a infinity version had no information to the private scope in the serialized string and because that this only happens with version +Im new to the static metamodel functionality so please correct me if Im wrong. +Updated patch +Its new issue Property Sheet Page always have low performance but user thread doesnt interrupted early +Updated patch and changed to Datanucleus +Implemented for +I have the same problem but for me it shows up every time I do a show history +But we have it! Isnt it just the instantiation order list +We should actually consider requesting its removal from eclipse release we should include all in ext but just not the jboss related ones +henry did some work on user pages and managing roles that I am still trying to get a look at so I can factor that into the white site and for the group management bit I am close of having the summary page worked once I iron out why the project summary tables are not rendering +Closed upon release of Hadoop. +looks good +Marking resolved merged to with revision +Flavio that suggestion worked perfectly! Simply restarting the leader caused a new round of leader election and things sorted themselves out within a few seconds +for this patch +Its called resourcecalculator plugin +Oh yeah has the same problem +Whoever fixes them can extend the storage engine interface if they see fit +Linda +Check out the comments +Closing the issue +This seems orthogonal to the other issues here +log +Issue is also present on build from repository todayUbuntu JDK EL test cases. +I think it should be a purple elephant skeleton +I dont know if this helps you or not Loc but I thought Id post it + built the new plugin content assist is starting to get really good +it is already working well +Being able to control the structure better is definitely a nice feature +Reopened to correct Release Notes field. +Not to ask the obvious but did you upgrade all your dependencies +Fixed in snapshots +Users can now add their own seeders +We should be able to consume it locally +Yeah I would prefer this to be an error but Gavin was against that on the discussion. +If thats true maybe we can restrict this further to continuous replication only and have it use the changes feed with includedocstrue? That should make it even easier and faster to do this +used here is with updated patch mentioned in issue I might be wrong giving this advice but is it possible to change the so that we can generate sstables that are already compressed? This will also help in making loading process even faster as the data that need to be streamed will be very less and save the resources required by cassandra nodes for compressing +I do not object to the change +ThanksI noticed that I didnt closed the that I used to read the XML content +so it really seems like an odd solution for nodetool. +Will be available after the next snapshot build completes. +reopening this because I dont agree that this is a configuration plugins in dropins is the way many will do when they use Eclipse hence we cannot just close initial thoughtquestion is why we need to use reflection to load the VPE class +OK here is v which sends PBSD in and uses the extra to avoid clearing the cached attributes in every XA state change +covers this anyway +what is the corresponding JSR bug +Verified as working with with Hibernate when above property is set to create or you say it works +The examples are now consistent. +The stale connection check can be VERY VERY taxing in terms ofperformance for relatively small payload requestresponse pairs +to reproduce hit build a couple of times while still checking out was just querying in the wrong order + fail to kill roo and take me back to the windows command prompt +Because this thread runs only every half second app start can be delayed by up to a half second +Thanks XutingDaniel +Too bad we cant vote on families and then the winner from that family +Thanks +Fixed in Rev + Looks good +Closing +Will need to add this property to Hudson instead +XYZ in my job xml config file +After LDAP Ill decide whether to look into this or my Enkive archive mailet. +I just committed this. +Initially it was suggested by the ODE dev list that the faultresponse messages should be wrapped in objects +Jaroslav thats right +Thanks Dave well be looking into this +Strictly regarding the originally filed issue you made the exact same change I did in my locally patched version so I feel confident that youve fixed the for the bigger change of refactoring code up to a parent class +The patch also reenables All cunit and smoke tests pass +It is just a convience class allowing me to switch from commit to rollback modes +Patch applied at + the URLs in the description +The problem has been repaired now however the first problem still remains +can you split the scanner change into a nd patch please +I guess he answer is Yes +I made the always return a new connection factory instead of always returning the same proxy +In my case your patch from didnt help to solve this problem +This issue is not a blocker nor a critical for EAP +Attached patch removes trailing to trunk and +What I have to figure though is whether it should +Its in Robs code but that doesnt really mean its his problem +So in those cases reported numbers may confuse users +I love the solution you have come up with but would suggest that it is moved to instead of +Inline functions should be fine with the caveat that if they change between versions the user might have issues +If it turns out to be an problem or something can easily code foragainst this JIRA will be reopened +The patch looks good +Attaching the logs +hm jep jacques i performed an update before i tested your patch +The changed fixes the bug and I have testedit with the wsdl that I submitted in the bug as thetest case +Hi StuartThat are good news! Im waiting impatient for +Any update here? This ticket is more than years old and effectively a violation of the spec +reviewed fixed issue for + +Hi Karthik the patch seems fine to me and I agree that the patch is so simple and writing a test sufficiently difficult that it seems unnecessary to write a test for thisOne question can you comment on the ramifications of this issue? How does it manifest itself? And what triggers it +Much of Metrics such as the visible methods of and much of the can be reused +That is because I havent published a new version of the tomcat plugin yet but the fix in in the issue was marked as fixed I downloaded snapshot build and tested +Committed +These issues are addressed since the version of the cant quite follow is this feature now implemented or not? If yes whats the final syntax? +go ahead +Do you see any issues with that? Is there a similar function for? I couldnt find any +Closing based on Ondras comment. +After my changes to the this should not happen anymore in Karaf JB I Will also take care of this in Karaf if that is ok for you +Tks a lot JacquesI create an addon to wait your review I need to this functionnality for customer +Patch looks good to meRather than synchronize why not use a concurrentskiplistmap? Also for sure we have synchronized all accesses to +Ill commit this +Thanks That definitely solves the Serialization issueSo to use Singletons Stateless Statful Jee objects with to avoid Serialization issues create a or CDI Bean that wraps the EJB +gitignore to make it work +MR part of the patch +Hi Ian this looks brilliant to me! So fat from me +Patch applied +But at least I now know why we get that message. +All tests pass on my machine. +The latter +All of these poms are already committed to svn and your patch appends another copy to the existing file +Can you perform the same test as your first comment to see if it helps? +Chef uses to store states +What is the status of this? Wanna finish for or should I move this to +It will be good if this patch is applied to also +Changed to for Got broken +m so after each upgrade we didnt have to remember to hack just upgraded to running under Tomcat from Tomcat +If a load task was assigned to this thread an infinite wait could ensue if the assigned thread never loaded another class. +A few other minor changes such as README etc +Just to be clear here +to trigger QA run +I feel that this is something that and should expose +Closing resolved issues. +It is not resolved +The fields seem to always return the correct element +In the subsequent composite op the add step can then just provide the content hash +There are and basic builds but no release build will add are now online +run jaxpcrimson breaks under jdk but boots unders jdk although I doubt is used because its doesnt even exist in the distribution +dims +Can you try examining the local repositorys jar content if its really a jar file and contains classes +Fixed all Wizard Page issuesPage issues fixed Made appropriate fields Added to Root Path label Added a Tooltip to Path column header stating that its value is relative to the Root PathFixed all Wizard Page issues +Updates smooshed into this patch +dont see us working on collections before at the earliest +I think this should be OK now +Took some time to update this to newest trunk based on the fixes in the patch +Maybe I make sure to clarify some definitions Inner when join condition left record doesnt match any right record exclude left record +Works fine for me on +try Hudson once that is committed +This should fix the problem by adding a call to next on each iteration +rebasing +Changed to add the intermediate directory +The class name is not passed down +Upgrade to or sp +This makes implementing just about impossible +As weve had confirmation this is working wih the Axis C server I think we can mark this as not being a problem. +Failed on both iPhone iOS and iPad iOS +This should address this issue when combined with other related patches from other issues to address lazy loading +sorry for the inconvince +at Minors request +trunkCommitted revision +Put all of the files in a package called aThen simply compile and run Testy classI expect that you will see the message FAILED TO AUGMENT which to me proves that the default values arent being augmented +Hi FreddyI understand your argumentation +Yes I can use list from some external property it is list in codeI can use list from some external property file But we need list for current do not understandIm asking what you use the version number for? Is it used when setting up the AS adapter or how +Should be fixed when using at least Sonar +Known cases were resolved in ER +I did not get to run them but reviewed the code +The fileUpload stack is already configued to work even when the request isnt multipart so now you can just use the completeStack which has most of the features in place. +Also sent out the API for review +NickI still would like to see this patch applied as it would lighten the contract Jetspeed has to fulfill forembedding PlutoAnd as Ive said before this patch has no on or loss of behavior for Pluto or the PortalRight now Im required to provide a mutable interface for the Navigational State component in Jetspeed just forthis +Maybe Ive just been lucky The BPMN request is more something that I would to follow so I wont bring it up again +But supported! encouraged for all new installs +Improve Normal scoped managed bean implementation class has a public field Named +thanks for the new report +Committed to branch. +This patch moves the to from AllI opted for Knuts first approach of adding the class directly +Hi Christhis patch will add at the end of the line a link for downloading the files added +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +it is not fixed yet as my root partition has space in GBs devmapper G G G +Already fixed in the new version. +Personally Id prefer to commit this +Redmine itself does not provide Mailing lists +Thanks for looking at this cosmin +In the process of writing it has become clear that the instances themselves are responsible for validating that the data they are given to serialize matches their metadata rather than relying on client side checks that they are passing in values of the expected classes +Integrated. +Marking this issue as resolved because Maven Hibernate JPA library is created and some other libraries can be added using the extension Snapshot Repository we can create a new jira. +branches trunk trunk +paul did you figure it out +End streclaimlongcol storetestsstoretests +See attachment +Looks good +Please use this patch +you sureOur current API does not make it easy to determin when the user just wants to read as opposed to readwrite and it makes quite a few things harder both architecture wise performance wise and security wise +This subsumes the earlier patch +Adding to Affects Versions and Target Versions +I cant reproduce +Marking this wont fix. +Im not sure I understand the use of theI think it would make more sense for to implement and do the archive extraction in the install methodThoughts +The test using springs. +Once a method is made protected it can no longer be changed without breaking the binary compatibility +Versions of jbossws not yet supported will add a feature request to upgrade for the next version of ESB +However we lose the specific NOTEQUAL with empty string IMO an uncommon test especially in a +That way if others add similar bindings for other languages we can accomodate them as well +stefan I tried to apply your combined patch but it seems that the test case does not compile +Could you show the code you are usingIm not entirely sure I think that a property file is easier to update than java code +Hello any info on this issueThanks +Thats a very good news +New version which allows extra dependencies to be added to the generated pom file +require more more information for this to be a valid bug report +Fixed in prod +YaytayNot sure what issue you are having +We dont want the demo or contrib code to become neglected +Closing all resolved tickets from or older +Its a nasty one as its hard to understand that its a bug in the plugin +if the current active does not respond to the graceful request throws an exception indicating the reason for failure +HiCould you please append a testcase similar as I added for this issueThats the most fast way I can have a look at your test and figure out if its a pure cxf side issue I dont wanna setup any JEE container to run your testcaseThanksFreeman +this seems a bit scary to me +bq +A patch adding the new method to makes sense +Logged In YES useridHow do I turn on debugtrace info for the deployer +If not it may be illegal +This problem is now partially fixed +fixed rev close of all resolved issues as part of closing items. +Mirceas PR for should solve that +Now in CVS +Wont fix issues should not be assigned a release as it makes for a confusing release notes. +Thanks Xuefu +Thx for the review RickI too think further improvements shouldwill be done as with all other wikis so Im resolving this issueI updated and deleted the wiki page +I dont think thats the feature you want theres a better way to do it +If the connection between the client and server breaks the currently doesnt not reconnect in any meaningful way +issue now release note status has been set +could be used toaggregate pages from A +Could you rebase since trunk has moved +The connector has a hard dependency on xalan which is pulling in transitively +Interesting that you have to make these changes Jimmy +So I think we are not already there +The tests are a bit prone to timing issues but I think Ive got enough time in there to be sure the consumers get all their messages but the tests wont take too long about seconds for me here +Isnt this fixed +Hi GiladThanks for the swift and very useful reply it has given me good insight in the sampling process +Once the database exists concurrent builds succeed +The only difference between them is just that +Attaching simple bash script which executes a series of SVN commands to clean out the existing module and rename most of the files to what they will be in the next patch +Changed the compiler version from to in the parent POM and in the file +Thanks for the review Christian I got around to testing this too +One could probably try to adjust the line length in the description to avoid this but that way it gets really fragile against changes +TODO redefine an approach that introduces the overridemapping from a JMX mbean to a either at the bean layer or as mcbeans in the deployment rather than a more general mc beanfactory approach. +This should fix the issue +Since were calling chmod directly we might as well roll it into a single mkdir m root +I have fixed the issue with however the problem is deeper than just a sorting issue at one deployment level +Pulled this into so that resource monitoring can be controlled by downstream projects. +I felt the bullet items were a bit awkward so tookyour advice to simplify that portion of the page and committed the changeto the docs trunk as revision . +at least for the first version +the has already been removed +Im going to have to report a bug against the leveldb c impl to see if they can figure out why the repair did not work +Minor changes +It is better to remove from the fsimage and store only that much information that is required to generate a new storageid for new datanodes +Patch to fix the in ceddaaceffecdde +Make sure you are using a current jboss version and either provide a testcase or clearly explain how your app is different from all the testcases that close connections without problems +Thanks SparkPatch applied to LUNI module at repo revision rPlease check that the patch was applied as you expected +Cool thats great +Giri any updates on this patch? It would be really nice to get people to resolverlocal to do development across the common hdfs and mapreduce +The create call cannot be made which makes sense given the stack trace produced by. +I am in an environment where my major central mirror goes down somewhat often +It is available +Fix in progress +backported the fix from trunk to branchsibmsvn commitSending javaengineorgapachederbyimplstoreaccessbtreeSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawdataTransmitting file dataCommitted revision +Ok thanks Erwan will check this +Patch applied at r +Here is a fix for this issue +Username igorfie +Also FWIW this approach let us use other connectors and point it to the same letting us use logstash as the only place to do content processing +Section added as discussed. +Hi Aidan can you take a look at this change for me please +At present it is text boxes but the hope is a fairly tech person will be setting this up esp since its in the package editor so I think this can be tolerated for now +It wont always generate a conflict only when a conflict happens which is fairly rare I agree that forcing devs to handle this edge case in seperate ways is frustrating and leading to people not actually dealing with performance impact is real and severe it must be resolved before this could possibly happen +for itPublic methods are javadoc distribution patch without tabs +I will vote to drop those tests altogether +Closing this one +to commit to trunk and branch +In other words there is no extra scoped proxy in front Its the itself being serializable in conjunction with its backing relies on Spring s serialization feature just like and do! It is working perfectly in Spring. + +Someone has already raised a bug for but we havent been able to solve this yet +We shall put it into scriptJars if jruby is used +Yikes I think youve uncovered a major gap here! It doesnt look like we can do any of the chained expression stuff in the Spring XML +Hi Donald Please help to review and commit the patch and close this jira +I think this should be committed to and as wellSureReopen for committing it to and +also works with xerces +Marking this as resolved as this was for tracking of the problem in the s already another JIRA to track the including of the srpm of +AFAICS it works as long as you use the latest binaries. +I suspect this may simply be an overlooked corner in the SQL spec since none of the major vendors support functions in check constraints and functions never need a schema qualifier +Issue rejected +sorryCould you patch the current trunk so we can work towards a fix in your next releaseThanks checked in the candidate for fixing this issue +Expect something to be committed soon +Nice work Jon Auditor Hsieh +in +Verified via unit test as well as via squirrel +Thanks for following up on that. +Is this to go into the for each platform or into seperate guides in the docs +Reopen with new arguments if you feel strongly about it and well revisit the debate. +Therefore the removal of the file is not starting the js communication with the native side which is a good cause for the device ready event not firing +Of course it helps thumbs up +Ive committed this to trunk and merged to +I tried the following attach volume in a XS setup +This is really good find! Not sure we can fix this in time for +See the changelog and +Sorry I hit the return key too early +The RFA rewrite should fix this issue +Rohini Point taken +I think this one has been taken care off with work done for +Thanks for the patches! +Pay attention to the parent and url fields +Fixes some warnings in the RMRemoved the slfj pom changes which were fixed in just pushed this +Just sounded similar but I could be wrong +Committed to branch and trunk +We will also need to take into account the capabilities of all the clients in the ecosystem for supporting it +I will ask him about it +Lets close this +I cant verify it before is fixed +Manually updating the ruleset refs appears to work ok so this is an acceptable workaround in the meantime +And it also batched hardlinking to make upgrading fast +When moving to page I got no results when moving to a later page I got exceptions +Running run Failures Errors Time elapsed affect version now. +Attaching that fixes combined javadoc classpath and externalizes list of files from directories to include in src and bin distributions +Thank you Brett I tested it right now and the memory retention problem arises no more +Closing all tickets +I can confirm that this issue occurs on Seam with facelets and goes away when using the facelets jar from Seam dist +Marek Im assigning it back to you for further action in solving it for +When using Ivy from the trunk the error no longer occurs. +that reproduced it for me +Is it still an issue + +Have to understand what is happening there +This should be fixed when is moved to pool +Feedback label for com styles is ready +I deploy brokers clustering topology as pure masterslave +Thanks for comment on +Attach initial patch +The install doesnt throw the errors as described above which is good but membase doesnt show up in the dropdown list when creating a new cache connection +This is not desirable +We should probably redo those examples +updated patch for the disabling table regions will assign it with however will unassign it again after it is openedBut why should an assign happen and then go with reassign when we already know it is a disabling table? Anyway we will try to come up with a wholesome solution taking your suggestion alsoGood on you Chunhui +Its often the cause of problems +Attaching patch to codeThanks patch appliedColm. +xml import code +Table decorators only iterates on row currently displayed so thats the expected you want to calculate totals on the full list you can do that by accessing the table model available in table decorators from displaytag. +Ill look into this issue +Added the Time and classes to help with testing +. +What version of Cordova was your project originally created in? Sounds like does not exist +Dont know what to do to get proper orientation in both patches welcomed +Blank passwords will be always refused without any configuration option +Relate this issue to that issue asks how we force the agent to perform a new runtime scan immediately to help detect those missed by the initial import that aborted +The error will appear immediately if there is JPA project in workspace +Link to related issue +Added column attribute for consistency +Much appreciated +Anyway I didnt use any newer features but I wouldnt argue against using more a more recent version if it brings a feature that we find useful +minmax constraints make no sense for String Ill check it for Float as you described in dev maillist +This test keeps failing on and off on Mac boxes as well +Select Contenty Type as portlet you will see list of portlet instances but no vertical scroll bar even though list is longer than that could fin in the window +done +attached a patch for trunk +Transferring ownership of this issue so Guillaume can do a release +The reasons +I can close it +Thanks for the report and the sample +Looks more important than I excludedParams committed thanks for help! +It was easy to reproduce and after a quick investigation I could see the key up method was returning true to say the menu button key had been handled +This enabled me to simplify the special casing in a bit +but thats for another patch +Will I just have to handle the renaming myself +Most of the links on the mentioned page had already been fixed although I generalized them a bit today +With XalanI got the expected results for lang and lang +solr doesnt support highlighting trie fields +Passes unit tests on my end so Im committing this +See linked JIRA to track the problem on the branch +I think its fine to acknowledge that key cache load will negate the advantages of enabling saved indexsummaries for this ticket and open another one to improve the key cache design +This causes the drivers to silently fail to autoload +Not sure about fb since its not as easily available but would be good for and secure at the least +I think you mean r which is odd as that commit reduces the amount of stack trace elimination that occurs which is possibly responsible for the regression +Fixed in branch and trunk +The converters are now active trunkMost of the unit tests passThe XSD in the spec is clearly not complete and there appear to be a number of missing fields and missmatches to the text of the spec so I am working with a modified XSD that matches the text of the spec document and should be closer to +Thanks +Patch updated with patch because undoing api changes missed a change in is patch after adding the change to do with get in and incorporating Devarajs commentsbq +committed to trunk in revision Simon! +One can also use and together which would also work +Excel files really really munch memory +Thanks for following up Knut AndersI have started a for trunk and will do at least a few runsWhen RC is out Ill do some testing of that one too +As a workaround you can disable the JPA configurator in preferences Maven Java EE issue cant be reproduced using the latest jpa. +it currently works with but not so some minor bug +fixed in rev Thanks! +Patch committed to both SVN and branch +successfully tested with release +It seems like the only way to prevent this is authentication by way of internode encryption +I dont have a test case handy and Im not actively using ODE any more +Until then this isnt an actionable item +Ive also tried from one Datamap to another Datamap +Can you please provide a simple war or maven based project +Sorry correction from last responseSo why am I not seeing being loaded in the verbose output earlier on? And it really looks like has not been loaded at that time +I looked at the diff +Attachment has been added with description A test font to reproduce said the file in Google Chrome to see the missing glyph in has been added with description A test case to be used with Google Chrome +Nope not + +This issue was opened to exchange info with the translators +Apply this patch to Pig TRUNK Pig Note No new test code required changes to documentation only +Patch against snapshot +Ive been looking at your report and I think the problem is not in Spring but rather in the way you are trying to use JPA +I actually have code for this which I am using in an internal benchmarking tool which well be open sourcing at some stage anyway will convert this code into a patch and contribute when I have chance +There is still something unfinished when these global app events are triggered on IIS side class constructor works and we have been now using that successfully for yearsDebugging just showed that it stops processing these configurations and thinks that is has already done it +Just a heads up +I faced that few times too with +Since I am working on other projects I dont want to miss updates andor block the issue on something which is pending on me +Have those changes been moved out to another patch +network accessible resources will be handled as global agree with you that we should start off with something simple +It seems to me it might be better to try to repair the block if possible rather then just delete it +It is a much cleaner class now +KoenKris could you tell us which preferencessettingscode we should call to allow for an setup of jbpmdrools for a installation +Thanks Yongqiang! +I think the issue may be a spelling issue +Fix has been tested and verified. +scenario is is my scenario which fails Client ajax event Server Client submit Server Server error if submit happens before ajax response the error happens +Several useful tools in the past has been floating around +Can you describe a flow of actions when this happens? Please tell me more about the invalid XML issues +Go for it! Luke youre a committer on Jelly right? +Historically it wasnt possible to have two P in the command line +Problem here is that not all server versions may be supporting remote configuration restart +And with revision I adjusted a link to suns web pages in the ptBR ref manual similar to how it was changed with revision for English +i havent seen this in a while resovling for now as cannot reproduce +So I cant run the application on Tomcat +mavencacheIf that doesnt help check to see what files it contains +identify the offset table by checking the relationship of alternates key and the earliestkey +bq +Keeping sandbox components inheriting from a release +Nice find. +Someone with a disability would have a serious issue clicking on that line to see its properties vs +The beauty of the new Controller architected is that you can make one Controller class responsible for a particular constraint through the is destroying the architecture and thus sounds like a design flaw +I am not even sure its necessary for the DTX package +Committed w below message +Simon the file is now like but the file now has the graphic at the wrong place +A fixed asset cant be transfered without a financial transaction +Committed revision Thanks Walter and Noble! +Added simple unit tests to verify that functionality is working +Does it mean slice on R or slice on the latest proposal for R +Closed by submitter request. +Unsupported version +Attaching a simple patch to simply limit the queue size for the replicateonwrite stage +This problem will cause document embedding fail +Fixing Affects and Fix in Version +It should be there now. +Ill dig into it further +Jason thanks for handling this one +Thanks Gregory +Will commit if tests pass +Sounds good we are doing this generically for the openon question were more about picking up the ids to be able to navigate should be simple but I guess that is done generically today too so all is good +This adds the property setting to the buildjars target +The API does appear to be a match for +BPEL Service and Reference available in trunk +However as for Continue action it works like that right now in case target iteration is started the story and tasks dispositions will be set to Added but in case the target iteration is not started the story and tasks dispositions will be set to Carried Over +and share a lot of code and should be combined into a single class with differentThe class level timeout is min +Duplicate of resolved bug. +Move the fix to +Regenerated patch +This sounds like it might be that +Current work in progress +If someone wants to enhance heher should find a sound base to do so nowEnhancements could beas suggested in point A for POS in my comment above to show dynamically the agreggated price in the item config windowas suggested in point B in my comment above to show the base priceas suggested in point C in my comment above to show items prices. +Ah right I see where it was +bulk close of all resolved issues. +Was it committed by mistake +I think it is not user friendly and a little bit confusingI think YN is the right way dont know +Attachment has been added with description row formating patch fix for transition resolvedfixed bugs to closedfixed +Hi Jarek thanks for reviewing it and you are right it is best to use the getName method +I was sure before your comment but OK the only way to be sure is to test it +I wont go any further than that as it is just a testing issue +This patch adds logic to prevent the creation of a credentials db with vacuous username or password +If scopes are used and SCOPE is not in the stack config we should throw an exception +trunk could you please provide any suitable example for this feature verification? Many thanks +you mentioned that you still had some issues under tomcat such as not being able to create objects? Could you be more specific +srcmainjavaorgapachehadoopyarnapiimplpbservice +As a result debugging andor reproducing a problem is impossibleI dont know what the support resources are for the zOS XML toolkit but here is the home page which should help you find outhttpserverseserverzseriessoftwarexml +Unit test to demonstrate this bug +from me on this issueIm migrating Metawidget to MavenJenkinsSonar +st version +from T where partCol value and col col will be null even for value partition therefore value partition will also be unknownThe basic problem is that there is a implicit int to string conversion I dont think this canshould be fixedwill upload a new patch undoing the patch +The spec does not define any rules for how inheritance among node types works because it was judged to be part of node type creation which is outside the scope of the specOf course maybe this judgement was wrong but that was the original reason +Passed tests locally try hudson +The sample app from attachment uses Grails +Well is another option I will prove your proposal +WIP pathThis one has testing for the read and write paths started +I confirm I also ran into this issue but could not reproduce it in an isolated work around has been to define loggers the java way using SLJ API and everything works correctly +The initial release of this will only return post collapse facet counts which is often what people wantThis ticket though is part of a larger design that includes which is designed to have facets and other aggregate functions calculated as Aggregators +This is just a build inconvenience I think and wont make it for +MDL Dictionary also mapped to create columns for both log file for hadoop and job history log file for hadoop +The utils jar however has a bunch of dependenciesI would deprecate the existing classes and either +I will probably need help getting an adequate dataset to test with +I am that Hibernate Validator is the RI and that its developers include many members of the EG including the spec lead but my evaluation is that this is an issue +Also of course we must always set a start time before starting a taskThis happensbq +Verified + +Patch looks goodThank you very much Eugene and Hitesh +heres a patch +Many thanks for the patch. +It seems like if I created a table with one shard count +This patch implements the above logic and changes all detach methods to be unlink methods for the readability purpose +Okay I changed my mind to move the addition of a convertor to a new bug as theres problems with time zonesand conversion to and from long values +Ill make it protected on commit and then if someone else makes good arg +Your first observation is particularly egregious +This is the layer above the ipc +This was done as part of. +I have just migrated to Thunderbird and are fine either way +Removes stray and files +This has been fixed by the refactorings and other changes. +Thanks Charles! +Sorry Dennis for the wrong information +It is but the AS default host is and not localhostSo you need to useROOT to exclude +The data set that theDFSMSMVS Network File System feature uses can be any type of MVS data set +Use it like thisfill in content at +Sorry this is too big a conceptual change +Committed this +The url is as followsscmstarteamUsernamePasswordProject. +This should live as a separate project hosted somewhere else +Schema has been set up and configured according to the instructions in servesupport. +I think doing this for one codec will be supremely useful as I have found the files format page in the past +In you can have many roots +This is on the other hand a misuse of locales since I would retrieve the locale via the country code and not from the user localeWhat do you think +It looks like they were just trying to improve performance when the counter was being used in a single thread +I apologize but I dont understand your comment either Ted +Fixed in trunk let me know if there are any problems +Pls find attached a patch +what version of the surefire plugin do you have? worth checking it is the latest as I had some similar issues with the first snapshot jason deployed but its fine now +I need to update to say that varargs are also allowed for parameter style DERBYJDBCRESULTSET +Bikas when you say creating an API for blacklisting a set of nodes are you basically referring to or something else +Fixed at revision at revision +thanks Rob +I just committed this to trunk and +trunk +Also add a test case +Yes +Profiles work much better in Maven +Ive added a test case to demostrate the issue and its now resolved in consequence of this seems to be that the nice feature of Composite Destinations will not allow a temporary queue in the list +It looks as if the Problem has something to do with the service Name being the same as the type Name of one of the arguments +Logged In YES useridAlready fixed +If it doesnt then we need to fix that +This feature will really help us move forward performance wise so it would be nice if it would be available in our next major release in the not too distant future +seems like it is fixed in Thanks! +Any update +Andras I guess this is a JDK issue +I tried it with both and of ffmpeg +Closing old issues +you are right I just didnt have trunk in front of me +is not instrumented whilethe file srcmainjava +trunk Improve to properly shutdown +Ran unit tests locally +Then removes itself from the Map +Committed to trunk +Now it works for me +Marking for so that it can tracked accordingly. +Deleted patch needs comparables +The Java Reflection API is useful we use it throughout Castor already +Patch look good Yonik +Just got to that in my inbox +What information does your last comment add to the discussion +The patches was reverted due to a known JAXB marshaller issue +Do not add multiple concerns into this one requirement jira. +using GC and hardcoding minimal width at it as resolved +Please remove the reference to Sybase it is NOT supported in thx +In several lines are much wider than charactersPlease use the formatter from on the new files +I my last comment I mean You of course instead of I. +Thanks for the quick response! +Another approach here would be to use a large circular byte buffer andadminister space in it yourself +Using xxx helps to reduce the n problem but isnt as good as subselect for my case +If I were to contribute it now knowing it wouldnt be accepted it would be tantamount to sending it to you dont get the wrong impression I didnt mention I had a working version locally to try and lure you in to a false sense of hope or anythingI was just trying to explain that not having something written in Java was not really the issue its the promise and the software support responsibilities of Spring Source +After that nobody resets to back so it gets +a maven reposistory or similar +Its exactly the same issue +Currently if we fail at the bad time we could end up without a hwFile +Added repro steps based on Cassandra already committed a fix for +patch for trunk all supported support symlinks Windows doesnt +Ron please confirm it has been applied as expected +also attachedI modified code in to handle this case thought it should be reportedI am eager if this can be handled through +Add unit tests to +What Stuart is asking is that AS could provide its own class table for those classes whose class name and module info is known +Also this applies to other versions also right? Not only trunk +What version of Groovy are you using? This looks like a bug we fixed a little while back but only in +Thanks a lot Sijie for reviewing the patchI noticed there is on unused import +Hi ThomasWhen you mean irreproducible are you confirming that there is a problem with multiple auto elements or are you saying that you dont see this problem? Thanks +Most changes are about relative path handling +Patch is committed into and +Resolution Rejected. +I think we will know a lot more as people start reporting in on their results now. +Im a little bit dismayed that our eventual upgrade for this point release is not going to go smoothly +It should list out which files are missing licenses +It simply uses the to test the +I think my preferred option would be to snapshot then rewrite metadata and remove on startup +Only the XML tags do thisOf course its totaly bogus to do this and it should be explicitly disallowed +Changed summary of issue as outlined in previous comment +Untyped listener fixed in revision . +Niall if patch v is obsolete perhaps it should be deleted +Hi Andrew The bug has been fixed on r +Thanks for the feedback everybody I would like to commit this sometime soon so if anybody has any objections or would like some time to try it out first please let me know +contains a comprehensive and well tested address parser +I have applied the patch in and will update the docsI will leave this Issue open so that you and others can add patches wo opening a new Issue for every patch at least until this document becomes official +Hi +Enjoy! Closing as fixed +Thanks Hairong +The patch currently is not complied on since mockito is missing +OK I understand why my original solution isnt sufficient and why may be needed in clusters with heterogeneous hardware +Right now I have been trying to use Rudis code based on Mojarra impl +Since this has happened in the past and we have no cleat path to reproduce Im inclined to push +Done in rev +Another Dojo bug setting fox version to Future until a decision is taken on what to do with the ajax are moving away from the Dojo tags in Struts so this issue is not likely to get fixed without a patch +Tentatively scheduling for and leaving open for now but might just close this +group ids in maven as this is what Maven recommends +Its probably not a great idea to change the order of ingest operations after there are many systems deployed that already account for this behavior +Thanks Jonathan +It will be be per VM not distributed +The improvement is about how multiple stores can be configured following the fluent API semantics +This really only needs to be done for the class +I reproduced the issue it requires a dedicated test case since it is tied to the initialization of a static field in +I fixed the above issue and benchmarked again on the same machinesetup +Alternatively if you have a hex editor you could edit some binary file to insert those bytes and then verify that the uploaded file matches the originalSince Scott mentioned that the line ends may differ across Mac it would be particularly helpful to test with a file that contains all of the CR LF and CRLF patterns to cover all bases +I think if we were to bundle an implementation of with it should handle all supported object models including DOM JDOM maps etc +Rest assured that your dedication and the time you have put in wont be lost in anyway +Thanks. +This is already committed to trunk. +I only talk about those files +This also applies to Token +Openning new interfaces is costly and something we would need to maintain for a long time +BTW I found a funny bug I think it is in Eclipse +I think I can get DFT working too +I will upload the patch to review board after more testsBased on stacks original design is a configurable class thats why we use the refrection to call the method +Please clone this for and keep this for +Please try deploying again +The underlying problem is that dateFormat does not have a timezone argument and so CQL expressions that use this function to format Dates use the local timezone of the JVM +Works locally is well known as flaky +svn ci m Adding buildersjbpmAdding buildersjbpm buildersjbpm buildersjbpm buildersjbpm buildersjbpm buildersproduct buildersproductbuildResults buildersproductbuildResults buildersproductbuildResults buildersproductupdateSitejbds buildersproductupdateSitejbosstools buildersproductversionTagsjbds buildersproductversionTagsjbosstools now see jbpm in the build but where is jbpm ? Nothing available on the update site afaics +Revision broke the Jackrabbit build since the query engine rejects two jcrpath constraints in one queryIt sounds like we should rather fix Jackrabbit than this test case but until someone has time to look at the details its best to avoid failing the build so I reverted the change for now in revision +Perhaps Mamtas additional investigation could happen in another JIRA +Chatted with Johnny and it looks like isnt a good place to put a new test case for this +Woops sorry missed a bit of the change when cut pasting attaching new patch applied +Im unable to deploy any rails app onIm trying to swap and jars versions to bypass this error +And it looks like removed hbase +The regression tests pass for me on this patchThe existing NATIVE authentication tests cover the following operations Database creation +The userids in the repo will not match the ASF user ids +I had these attached but my previous attempt to submit the issue timed out and I forgot to add them again +add links to demos into to doc version for less overlap with engineering tasks +Moving them to CR +Given the migration to Maven this may not be an issue +If it is possible we should make the path relative to project class path +Currently if we fail at the bad time we could end up without a hwFile. +Jacques I will close this issue for the moment and if see that the issue is still not solved you can reopen itThanks in rev +I have reassigned to you Todd +bulk close of all resolved issues. +The timestamps in the exception above are JST which map to the same time in UTC so that looks ok +We would need to provide a mechanism to ensure that magic numbers dont overlap though. + +it looks good to meRegd What about STORAGEJSPOOLDIR which points to a different directory from currentJSPOOLFILE also happens to be same but intention of original author might have been to be able change +Reopening to investigate I believe a dependency to xmlsec has to be added +Thanks that did the trickKen +In case it is intended it should be documents in the code +However a empty string is invalid thats why the validator complains +New patches for and trunk that pass and I also added for +George Aroush +But since Im working on the bootstrap code now anyway I might add the flexible mapping if it turns out to be really an improvement +Then take a look at target to see what needs to be done +This is clearly incorrect though it would not cause major problems as most deployments would overwrite this inCan you run the tests with these changesant lib testlibs testand submit the outputAlso please do a submit on the patch and we will pick it upThanks +The patch seems not in text mode looks like a binary fileAnd the javawsdl will be replaced by the javaws soon so the patch should go into the javaws module and make sure all the tests are passed +ah +looking at the code history it appears as if this has been toyed with but not fully implemented +Is there a reason to use over here +Odd +They are junit is a package for jboss esb +As child element of the element +Hmm did you attach the patch to instead +I am sorry then am fine with doing that so long as the default is not suppressive behavior +defer all issues to +Thanks for the screen shotPerhaps replace the links in the columns with NA and a little question mark icon? Or instead of the question mark add the Need help? thing above the box +See. +Added implementations for HSQL and H sequences +The test I was suggesting is what the code is effectively doing and with no performance impact since the formatting is happening anyway. +Now its updated completely. +Submitting patch for trunk +Also the reverting of to +the latest patch looks good to me +I will do up an example that we did for our distribution of camel in servicemix +failed tests were not related to latest patchI ran them manually and they passedPatch integrated to and trunkThanks for the patch GregoryThanks for the review Anoop. +Could you remove aj log from page? it hide output of rich messages its happens because of form AntonI did not is the reason? How can I fix itThanks and Regards you code in with ajoutpuPanel after h AntonI tried it but still it is the same problem and is not and Regards Amit Could you attach simple war there this bug reproduced +Yes all looks fineThanks Mark! +So I need to investigate just the +What was actually done to fix this issue? I am still encountering this problem in version +Im looking at implementing the of the Thrift API +links to apidoc for Script Style Page Columns progressBar links to apidoc for pickList is reopened some links for column columns dataTable inplaceInput messages panel panelMenu panel panel progressBar scrollable is done. +It would be nice if the flex plugin for sonar would do code coverage without having to be configured to reuse the existing report +wait few min and see that there are still CLOSEWAIT sockets +Anyway it will be nice to have same wiki syntax in Magnolia and Clearspace too +Thats the same proposal hereSorry that was not clear to me by the title or description +appears to be working right on my branch but Im not sure what commit fixed it +More details as following Adding support to fast master up Adding a test case for new code Fix test function makehlog to creat rows with keys within assigned regions key JeffreyIs this for? Or for trunk? If its for trunk I think the patch not done on the right directory level + SergeiYour patch is in trunk r r r r +The observed consequence of that problem was that the transactions involved in the deadlock never agreed on which transaction to abort because none of the transactions thought they were involved in the deadlock themselves +Thanks to Val Blant. +build the latest Harmony jre +This changes only metrics JMX and jsp pages remain same +Verified fixed at svn revision +Hi Mirko and VikasShould this be backported to earlier versionsCheersPaul Foxworthy +Applied to branch At revision +Since I cannot reproduce I have no way to verify this +It puts elements in the proper order and adds a name attribute +Cakewalk s sens dubte angls americ +Right Forge runtime should depend on the runtime the plugin provides so I think there is no need to configure additional Forge runtimes +Didnt see your comment we are at a commenting race here +Sorry for that. +To propagate that information from head to body requests Head renderer has to save these attributes in the and restore before body request +I suppose this is probably a task for somebody in the EJB team +I guess one has to use but I didnt yet figured out howDo the exception occur when you try to convert the attached example? The stack trace indicates that probably something is wrong with one of the streams within the pdf +I agree it is a concern but it will be a little less of concernI can already see the emails titled Upgraded to but cluster doesnt start help! in which we tell them they have to change their configs and they have to pick which one of the they want less ofbq +Ah I didnt see that +I tried writing to and immediately reading it back which worked fine but that didnt exactly inform me as to why it cant be read within the Mapper +a new contributor looking for an easy starter could pick it up +Fixed patches on trunk awaiting Jesses review +Reverted previous fix as this doesnt solve anything when testing with RF +So this issue is a bit more problematic than I thought on the beginningMaybe solving it by listening to scroll events and adjust the item editor position accordingly +Enums have now attribute support +thanks +defer all issues to. +I have converted Ricks test into junit test with revision +I dont like this approach it creates a silent performance trap since on every search for every segment well convert the incoming bit set to aI think its best if we continue to require a FBS so users get a clear exception instead of silent performance hit +Hi GraemeI dont know yet +done finally +I havent tried that though. +Attached rebased v patch +The function returns theconnection access permission that is currently set for the user that you specifyand factors in the default connection modeI still dont understand the last clause and factors in the default connection mode +Do you have a reason why they need to be initialized in the constructor +Or these two jiras for a to +Its not a logj problem +Timezone for the server and the application are set at UTCI cant see any user preference settings to localize the time per person can you investigate reviewboard application and let me know where you think we can set it thanks +Ed If you dont use the insert update and delete commands they wont impact your use of Hive +Patch for the srcmainjavaorgapacheservicemixeipsupportTransmitting file dataCommitted revision . +Updated the patch as discussed in previous comments Included a test case +Thanks Earwin! +So it seems that the user name was addedchanged in Guvnor +And does the existing cargoundeploy satisfy this requirement? If not why +Thanks +Is it still valid? if yes please attach the wab sample used for testing here +Added the jacorb jobs to the maven repo and updated the as component matrix pom in trunk and branchx +They may as well use xsstring there +AmosCan you try reproducing this on your own machine? You will need to build Narayana and AS from source in the same way that the CI script does +There is a bug in the way that the test portlet determines whether or a not a batch of tests fails or passes that cropped up when I added the extra portlet lifecycle phases +everything else is generated code +We should either entirely fail the task regardless of exit code or we should consume the rest of its output +The latest attachment shows your failure mode a the deletion failure rather than b the propagation failure +This has been pending for a while +The patch removes unnecessary code and has been manually tested thus tests have not been added +uses a to as pool +Patch added so this bug can either be fixed or rejected + +This looks strikingly similar to errors from FFI users trying to load the c library and not finding it on Windows +It makes a lot of sense to have that on the SOLR side +Hi Can you provide a runnalbe test case +Setting the Authorization Bearer header also requires access to the request headers to upload files to an secured endpoint +retrospectively setting fix version in preparation for SDO issues fixed in SDO Java M in preparation for M release +Its in less than great shape for users and maintenance I thinkYou can look at the tests for a little help though +Bouncing to Snjezana for now +Its now fixed +Does it make a difference if you specify usefalse next to your cacheSeconds setting? Finally I got time to throughly test this +euh whats this cli checkout +On the other hand vpc offerings are independently created +Right +Ive tested M on IE and this behaviour occurred +Is it worth discussing adding a lenient vs +Changed classifier to backportConfigured the assemblies so they do not include backported packages. +Ideally should through an error during connection +Andrew +If you do not provide a label the default will be used better than nothing I guessAlso due to recent commiters policy I can only commit applications filesSo now I would recommend to check those empty labels and if it seems to you that its ok like that I will commit accounting file and let Si or another PMC members commit core files +This pull request is against your own repo +This is OKThe fix just disable the corresponding fields no to confuse users +I am able to reproduce this on both trunk and if I run the tests with pointing to a directory containing the Xalan jars +However I cant possiblysee how say test parametrization might be solved from this side +Sure! Ill start digging into the source and see +These are probably all obsolete +r on trunk +You should really be in the hausmates group which would have allowed comments already +Better use of object inspectors +It would be preferable to be able to support functions for a particular clause +I spent some time putting this onto trunk +A proper test case +ThanksI left it as warn since this is a problem its just a problem with our build process rather than code +Bulk close for Solr +Excuse my ignorance but why is this needed? Cant you just code them in your javascript directlyI guess its the cross domain bit thats tripping me time frame for this one when do you need it +I am for moving those out to be part of a tools or examples package somewhereWe also need to look at cleaning up our classpaths +Yet Im moving the fix version to since I wont have time to get this in for and the geotools improvement is on one has been rescheduled a number of times +Its not worth the cost of making merging a bitchWere already there +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Otherwise out of memory may occur when the message generating rate are higher than network bandwidth +Now it works fine +Fix has been committed. +MINA does not magically transform a WAN address into a LAN address +This is apparently caused by r which may have exposed an earlier bug in the Lowerer +tested on kvm working fine +Looks good to me except for a small conflict while patching +Renamed patch and removed extraneous Findbug warnings linked on the build result are not related to this patch +So this breaks on Hadoop +The reporter has not been able to repeat these problems when the client code was changed to not refresh the session from the thread on which the observer is called +So if I understand you correctly you are saying that somewhere in the xmlbeans configuration one is able to say relates to XSDA releates to XSDB Ive got it here are a few questionsIs this for two and only two XSDsIs it possible and perhaps less of a hack to have two separate projects producing two separate libraries in order to use two separate defined namespacesDo you have a you could attach to the Jira issue so I can better understand the use case +Patch looks fine to me +In this case the SPI is the service id at the moment +As soon as we have specifics we can let you know +Thank you PhilipIll take a look in upcoming days +I have resolved this in the sense of the distributed JTA will detect these types of transaction but the JCA wont as JCA will rely on the timeout value +Fix the help message for df h option +Krasig you shouldnt set your archive destination to your production server directly that is dangerous Instead put the jar somewhere in your workspace right click the file and click Make deployable then say which server you want it deployed too +Thats fine +Verified and closing. +What is your OSJDK +So when you change sub to inherit directly from basebase it will pick up the property OK but if you then moved the property to basebase then it wouldnt be for your replies Im glad you can reproduce the problem +I forgot this was already discussed in an issue and opened another with my proposed implementation +I didnt want to blindly commit it without someone else checking it over however +apachecassandraio + +What if you dont +Updated and its occurrences in mapping documents +Brill Could you retest this and close the issue if it is indeed fixed? Thanks +Thanks for the report fixed in am currently out of the office i will be returning Friday December nd +I think this may have been caused by move issues to and +Hi Michael why do you suspect an error relating to indexation of resources +bq +Note that the proper or JAVAHOME setting is for the jre not the jdk +A patch for review +Worth someone looking +If Derby had simply blown it away such recovery would not have been possibleBut yes it looks like for the problem is the new error message has been thrown where it shouldnt be so fixing that should resolve the issue for the user expecting the old error in that case. +Im lowering the priority +There is no logging happening that would provide any sort of clue as to what is actually being loaded etc +I can understand YARN not knowing about lost nodes as it doesnt have a list of all NM hostsHowever I think atleast the decommissioned count should be set based on the exclude file information +I will get rid of the unsupported option in due time +Looks like these messages Could not determine the size of come from the vfs library +We have opened another issue for improving the messages +Committed to master. +Struts sometimes as is the case with other projects has to addidiosyncratic code to meet browser failures to follow specs but doing so tomeet developer controlled relations to frameworks etc +This portlet was also contributed by Chris Mills and will be available on portletswap for now +the non standalone version has different version of hibernate several with different checksums see atched resolved In order to ensure we dont have multiple versions of hibernate removethe from both jbpm and management console and use the in productlibext to compile against +Yes better with a new one but then we gotta come up with a good name and parameter or just nameThe purpose is to quickly being able to mark a method as the one Camel should invoke in a class that might have many methods and thus it could be hard to pick the best oneAnd to use annotations for meta data instead of specifying methodxxx in the Camel routeIs Handler the best name? Isnt it associated with all the web frameworks +Now that I have just committed it should be a lot easier to fix this issue + +This patch is not applicable to SVN yet since its not generated based upon the current SVN image +Shindig is not a project +for the patch +Thanks everyone +I submitted for this +Fixed both in M and +Is there a fix or a workaround for this currently? +This is a website issue or an issue specific to subprojectsI dont understand where we should add issues related to mimej or to the websiteUntil now we used JAMES for everything not specifically related to JSPF JSIEVE or POSTAGE that have their own projectsWebsiteJames ProjectServer issuesother subproject issues have always been listed here Noel where should we reopen this one? +This version of the patch builds on what Mithun and Thejas started mainly updating tests to buildIn addition to a couple minor build changes we ignore which hangs indefinitely +Only if the installer explicitly added a specific named server do we do autosetupRight +Scott can you explain in more detail why you are not going with the option to not generate the stylesheet when suppressStylesheet is true +My patch only contains the Java version of the store +Its working but I need to clean up some of the flags on the server side and add more testing +Fixed +Can we add a hard redirect from the old Incubator site to the top level site? This could be done by creating a separate empty directory in the source for an Incubator site and then placing an +Ill leave this open until were sure that recompiling will fix your project but it looks like this is pretty much all wrapped up +Along with the scenario mentioned there if we take the current patch here I feel the following problem can come Because as per the current patch attached herewe force the znode to OFFLINE state if there was any current assignment going on due to SSH in one of the RS that will get stopped because of znode version mismatchInternally that failure to OPEN regin will make the znode to FAILEDOPEN +Glen Ive built and run the demo without problems first I built all the demos in samples next in oauthserver oauthclient +Since there is a workaround can I put this for consideration for the next release +Ill wait for next failure and recheck +Reopening per Rohinis request +Ugh that was sloppy! Uploaded a new patch that shows that member vars that are null indeed can be supported if they are defined as a union of their type and Avros NULL type +Thanks for the patch Gregory. +Need to update the docs at some point +I like what this does but not particularly the way it is done +This would also include deprecating some of the constructors +Lives again on +Thanks for your fast feedbackSet to resolved +This issue is fixed and released as part of release +I suggest closing the issue as I have not been able to reproduce it in other projectsjobs +Some unfinished code as a start to support for generic types has been added. +Hmmm good catchI will see if I can fix the code asapThanks for the report +The problem is that even with this patch the handler API is active and serving wrong info +Simon do you have a test for this? If not Ill try to make one +One additional question on the new analysis page Does your code also support? I just ask because I had no time to try it out it just came into my mind when i worked on +This change and the others in the set just basically override the defaults in case they are not set +and are almost same with this so this issue will handle them all +Thanks for the help and apologies for the misunderstanding +Proposed solution would be serialize the blockSize with in SOn the same lines while reading we can accordinglyBut this solution leads to backward compatibility problem +Applies to branch and trunk +Finished. +This is just a minimal change to make something which wasnt configurable before configurable +the previous patch was checked in with tabs +Note that has just patched the bug on their branch and we have requested it be backported to +Put into PIGHOMEsrcorgapachehadoopmapreduce then compile +Following the terminology in that thread this implementation follows the approach where there is a single merged big idf map at the master and its sent out to slaves on each query +Think this minor change should fix the the error +Yes +Patch applied +Ack! Apologies I was in forum mindset please forgive my markup faux still dont see any major advantage to this +Implemented at Apache waiting for the code to sync to now for and code is synced to FUSE trunk. +Any objections against committing this as is? And mark some convenience methods as experimental? I think that such a class will ease everybodys experience with Castor +I havent heard back from the review process +verified with configuration set to true the fetch failure is notified after +After the job client acquired tokens it flipped the value to false +A role can be turn on by installing role config rpm by the current implementationThe role based package would be decision of the community +Until then +CheersScott +moving this to since it doesnt appear to be critical for wont object to someone renaming the context roots and testing all the portlets but I sill feel strongly that knowing all the apps web and portlet is important. +So moved to view in in +Because in case he wants to use a configuration option that is not yet supported by the plugin then he can switch to the uncheck mode + +f on master branch +Im going to try to track down what they do +Which I think pretty much limits you to referencing filebasedresolvers +IMO favorednodes should be aspect of stochastic rather than distinct balancer +Wow that was quick! Thanks Dan. +Not fixed in +Converts files from DTD format to DTD format +If everyone can verify it works for them Ill close it out +is the latest but will be out very soon +It seems the IDE erroneously puts the older JAR in front of the newer jar in the class path for the project inside +a reduce then the reduce task does not have to restart +This doesnt seem to work +I will propose a proper solution for this type of issue shortly +I think I forgot to edit the file when doing the release +Colin for review and commit this! +Hi Luke I need to have Ganglia working with Metrics +Ran all the tests again +PR for master merged in master +Regardless of whether the instance is on an intranet or not the fact is that without SSL credentials travel in the clear +is derived from +Sigh +It also matters that the sql started with a malformed keyword +there is hometrafficserver dir hard coded in many place we need to fix this solutions create hometrafficserver dir +sorry guys for not having responded to this issue +This includes the patch for +I should have a patch to get the branch in a working condition Monday +As long as the required files exist in the final build source and binaryIf you build with the profilewhich is the case for a release build the source distro is created with the correct files +I am hoping that you dont mind this +Those screens dont explain that the AIR SKD and are also downloaded. +Ough forgot to attach new if somebody is interesting in ve tested this again after a cup of coffee and everything works as expected +updated affected that happens the first time I keep seeing this every second or so about times and after that AMQ is dead +Attached a simple patch that Im currently still testing +this issue is resolved DBRE is the think was an example to illustrate the use case anyway we are finishing migration of gvNIX addons to +Or is there some other back compat issueWe add this param to a protected method signature so it would affect subclasses of +Thanks! +If they come back clean Ill commit first thing tommorrow am +I checked patch and the existing code in the trunk seems pretty ok to me since it is checking whether a service is invoked to determine whether the returned fault should be Sender or Receiver +Step If the collation derivation for DTD for this node and otherDTS match then check if the collation types match too +If SQL Server supports it then Sybase probably does too +The move touches enough stuff that Im going to move it to another JIRA for clarity +is basically the same feature request and it is in the roadmap for +Bulk resolved close state issue transition +How about we put writeread lock inside method like other methods currently inbq +another option is to kill and restart the master and let the cluster inspection take over +Or we can simply skip the check we are doing before adding the endpoint in to wsdl so that any transport endpoint will be added to the wsdl no matter whether it is configured and loaded in to axisConfiguration or notThanksLahiru +Thanks +If you do not want messages removed after access do not include the appropriate command. + +I believe we are going to remove the ivy configuration for Gora for pure Maven approach? If thats the case its time to clean up the build wasnt sure but going for pure Maven is fine with me +Im just trying to fix bugs +Thanks a lot for making these changes Eli +Is it correct or should it be bundleBy the way it should be interesting to add directly the plugin in the pom because the code generated will be deployed as a bundle in SMX +This is a duplicate of the other issue you filed Mike +The issue will be closed +One of the reasons having this feature is to avoid writing to a single replica but the statement above allow it +Attaching here to let qa run +Tom do you consider this proposal as setting variables inside the expression? To me it looks simpler evaluate the expression and set the result as the value of the named variable +an endless story r +It just adds a random number to the directory name the caller requested +The pom now has a placeholder which can be modified as per the builds requirements +I guess I was thinking that we wanted to make native users more solid than the present mechanisms +Marking closed. +The big problem I still have is that I need one URL for internal requests and another for External +Please apply additional patch to fix it +Hey your patch looks fineIn my opinion you can just +I have forced a build and I think it should fix the report generation issue +I will remove these scenarios and add the modified unit tests +I believe the following changes are still required add licenses for castor and possibly antlr add notices for castor and xpp +This is very cool +Hi Bob +will seeWhat you think +Triaged for +For information there are no problems with maven and lower versions. +at least not for now +What is the motivation for changeThe reason was that there is currently no way to retrieve the result table name in the server +Since well have hint future only for node that are known dead waiting on only those doesnt really give any special new guarantee to the user +Editorial and committed editorial corrections +Is that the proper option to set for starting with security +Hi DmitryI already tried it but the current project is so big and everything seems to be so nested that Im unable to make a small project +bq +I think otherwise we are looking good here for a stab at a patch with just some minor cleanupsI like how this format only described its blocktree metadata section etc I think this is easier to maintain so I did the same with the current postings format and factored the guts of blocktree into s docs +Raghu I think this one needs regenerating too due to conflicts +Done +I just committed this. +I had same problem in VistaWI attached a patch to delete the directory content before savinguncompressing the zip file +I run on local PC with the patch and returns OK +The napkinlaf artifact is missing due to + +when turning off advisory support the issue disappearedYet I cant turn off advisory support since I need brokers networks +The blank screen in the UI might be caused by an exception but if that happened you should either see it in the log or written to standard out +The idea sounds sound but is blocksize the best unit? Theres a certain overhead for each additional task added to a job for jobs with really large input this could cause really large task lists +the way the date examples I am referring to are using and not +Sample requires SP configured for NTLM users +Will commit if no errors according to same rationale that accepted +Yeah wasnt overridden so you just get the standard WFS one +LGTM +Your patch depends on etcfunctionsWe cannot have that because this is not neither present nor the content of that file is the same on all OSsIf you can update the patch that would simply use the localized versions of successfailure functions Ill review the patch +So we are left with little choiceApache Mina looks more complex to me than using Java NIO and we might end up writing more code than ideal to workaround its various limitations which would ultimately make our code unnecessarily more complex +committed my changes +RobCould you please tell where is placed +I think plugin + +Thanks for catching these jeff thanks for review stack. +These two together take sec on my laptop after my change faster +Fixed at with. +For a future version you may want to offer the ability to configure where the user lands on signout or provide a hook to let the another controller take over after the session is invalidated +the wsdl and a sample messagethanks +data file is accessed as byte buffers which is subject to when a GC happens +This patch still applies to the latest head revision and still tests out do you think you will have a chance to review it soon +Committed to trunk and branchx +Please confirm that commit mails go to the right listAlso I have set rdonkin as the only moderator at present +It works for both Hadoop and Thanks for the work Cheolsoo +The snapshot of Web that is used at the moment has included this change for a while. +Marking trivial +See bug this helps for me +Moving MR part of the patch from HDFS ticket +batch transition resolvedfixed bugs to closedfixed +It took a while to figure out what was broken +Perhaps we should focus on that +Attached is a boilerplate patch for the analysis moduleIt contains README describing filenames etc LICENSE NOTICE describing all poached analyzers licenses etc CHANGES describing all analyzers changesI yanked these CHANGES from corecontribs CHANGES file +Bumping to +Thanks for getting these fixed quickly +closing issues for released version +Thanks Mingjie! +all I see is that generated +I understand that this is not common use case but there are management ops in AS which requires server reload and it would be nice to have the possibility to add relevant tests to the testsuite +dstat is also +Yes this is exactly what I have done +It now first tries to use the official java API and only if not on Java falls back to the Java thing +Arnaud do you think you can take a quick look into the and handle this issue +Im intending for the protoserverid to be a general way for the client to determine if it has the required credentials to even initiate that particular auth type +Thankssvn stat qM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileD javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlcatalogM javaengineorgapachederbyimplsqlcatalogM javaengineorgapachederbyiapisqldictionaryM javaengineorgapachederbyiapisqldictionaryM javaengineorgapachederbycatalogM javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsA javatestingorgapachederbyTestingfunctionTeststestsupgradeTests I have understood correctly unique indexes always have up to date cardinality statistics because cardinality row count + +Rick should I put this into the branch or wait until after you cut the new release candidate +For this patch works if Hadoop permissions are disabled or if the user starting is the Hadoop SuperuserHadoop will contain a fix that allows querying the state of the dfs without being superuser +Full Import and Delta Import are functionally different things +so my best idea in getting around it is adding flags similar to the ones proposed the dialect to modify a query and pass back a new one would be a new notion introduced +If you really need manageability you should be using a real application server +The error strangely persists again even when using the latest Fuseki development build. +If you have any questions feel free to ask +README to the list +Defer all open issues to +Chris to get better bang for our buck one way would be to look at the maven bugs with previous releasesCan we create checks that would have detected any of the issues that went wrong? Later we could add some of the more thorough checks you have suggested here +I think in this case the performance impact is small enough +How can a long running client force the removal the temp files prior to exitSorry that I forgot to answer this question earlier +As you mentioned the original notes for this JIRA describe two bugs that stem from this root issue +Just added a forum link to discuss this +That is great news! Im all for a tmprestart solution +If that lock has a active writer then your code uses versioning +The behavior I see indicates that jar classes are available on the classpath at the point at which the client creates a job +Indentation at some places in can be fixed +Tested +Thanks for a fast turnaround +First after Ive killed the process in the Task Manager the service has been stopped +This will be fixed under +trying this with the latest code Im getting dwr qdox etc first when searching for junit +Can you create a new issue for this and let me know what it is? You should be able to attach a patch in a new +You wouldnt by chance have a sample XML document matching this schema attached +Plus had made mistake in localizejob in corrected it +Final patch +Original WSDL WSDL file +Facebook job hang from problem is that when the its own data structures it interrupts the fetcher thread but does not wait for the fetcher thread to exit +Well I added a test to and I cant reproduce your problem +Id think the database version should be the version of Hive returned from the Hive Server +I mean the option +Just an update Ive take a look into this +Checked for row presence they are missing so please add them. +It is not that we will be implementing other languages but it distinguishes it from artifacts +Patch applied to CDK generator +Not to be committed. +Its important to be consistent with others. +The fact that registration properties are needed is only known after a refresh and the error message is to alert the users that there are indeed registration properties for which values are needed +PS I think teds suggestion that we need stretchable and versions of implementations perhaps some boolean flag that causes it to expand versus error +Fixed location +The issue at Jackrabbit is closed so I guess this can be closed too? Ill do so unless someone objects +I will install these shortly +If the throws an the should unwrap it and throw the original directly rather than wrapped into the +Uploaded new patch incorporating Cos comments and some minor nits +Thanks Koji! Can you open a Solr issue work out a patch so Solr can cutover to this? Thanks +Just need to update the path a one line change + +Just to prevent confusions the code Dirk quotes is in the not in the +His review comments were Looks good to me +Local test passed +Sorry i didnt see your comment i was writing this message when you closed the issue. +Globally it would be ideal if HDFS settings could be under settings +Im not sure how much of it can be accomodated out of the box +It should be solved now with the upgrade to there is still a problem then a solution should be found in the of some vfsdeployment logic to reduce the usage of . +I reviewed and applied itbest regardsChristoph +Ill close the report +Forgot to include in screen in previous patch +Thanks! +Needless to say it hasnt worked very well +quickstart it is shows the problem +Which version of the plugin are you using +Closing stale resolved issues +This jira is incorporated by the patches in and meant +You pick a jar file that isnt in the local repository but is instead in the remote repository where the searcher window is +going to commit +Committed revision Added ability to specify dataDir in as well as ability to specify dataDir in s core create commandThanks Noble +If you are interested in the distance you would likely order by distance rather than use a score that would be compounded with other variables like query matchingIm thinking out loud here +bulk close of all resolved issues. +But that whole renaming look quite odd and does not match original Maven packaging +I like that name +work differently on RI HarmonyIBMVM and HarmonyDRLVM the outputs are similar but all three are differentSo though is now resolved the problem still exists. +makes sense that RS could get same treatment +Workaround +I tought we were already using filter encoding this gets me quite by surprise +On platforms that do support its that will be picked if Ive understood correctlySo perhaps we should rename to? dont think the name is so bad is the driver which is cited in the jar manifestthat gives it a good claim to the name +Sorry +We also want to update to have that config in it +Which brings up +Is that too tight +bq +A ticket has been created for to update the Apache configuration +Raising the max of the varchar willinsure that we can never index CLOBS are optimized for long strings +Thanks for verifying that it works in CR and M however +for the patch +I followed JDKs New Annotation wizard as sample in wizards list it is labeled just Annotation but in description it says Create an annotation type +So it would benefit +Schema files with version and have been that test that were updated areorgjbosstestsecuritytest jbosstestxmljbosstestxml fails with errors +Closing as this feature is already available in g. +this should be fixed since we introduced a module that aggregates all deps into a single jar +Every time I launched mvn clean eclipse recognizes that the folder was deleted and derived flag is unset againSimilar occurrences could happen with other tools which validate a certain folder structure +The query part is missing as the url encoded characters are somehow converted back to special characters and so the query is considered as a different it fixed in my local by escaping the ? and in the +As such when all the were purged from the pool the application context was been cleaned up but the resources pinned by were not +Its also unclear if theres outer join support for yet but INNER JOIN and ON is supported +I have found that our email server is being listed as a spammer by UCEPROTECT +Addes intructions on how to build and run samples and unit tests +the two samples on top are updated +First patch for the new style to the LookupBom forms +We use and its existing filtermechanism in many ways in our project +bulk defer to defer to. +Adding a comment after assigning the bug report +This is not the desired result we want a nice Bad Request as the return code with an explanation in the body that the name was too long and is restricted to back to Michal for further looks to be fixed in deltacloud git commit version afcbbdfaeaccebccfc Bad RequestParameter name must be characters or less Server message Deltacloud Parameter name must be characters or lessJIRA assigned to Ramesh +Would you be able to make these changes and the patchThanks +data and scripts referenced by commands +IIUC the values of the test runs are cached when the RS first starts up so retrying wont help +i quess the logging is now handled differently so old logs sre not overwritten anymore +How can I test on XP? What can you get from view +Ill be submitting a patch for this issue +Webwork just switched from velocity to freemarker +Thanks! +Any inputsuggestion is highly an and and looking into for a more optimal approach on distributed searching also emailed Lucene dev mailing list for problems with the checksum when trying to get an already open +I deprecated UDPNIO and will remove it in one of the future was removed as it is not supported and has been deprecated for a while now +close issues for. +fixed in rev Thanks +Hmm +Karthik +to get latest patch version or jdk + for the patch +Java Version +Another potential fix is to actually allow to work with a zero prefetch +I guess i should modify the patch to iterate through all files encountered in case more than one jar or directory contains function definitionsThis essentially allows users to mix and match what they include by altering the HIVEAUXJARSPATH +Some progressPreviously hot rod and memcached endpoint settings were read from the +Heh JRUBYOPTS should already be implemented and workable solution +Will make necessary changes to requirements doc FS and associated bugs +Yes I will close when I resolve this issue after tests complete +Set it to true if all sockets successfully started +I think its just bad initialization +Backported to closing after release +changed toString and updated impacted test cases in r +We will create test cases to recreate this situation and fix it before +Once again perhaps youd like to make concrete suggestions once we have something we can all look at +because this jira describes the problem to be solved vs is a possible solution to the problem +This issue is fixed in we can close this issue when camel upgrades to new released CXF version +I think you would need to extract anything out of the request within the adapter +Active reviews per developer widget is now deleted and is replaced by Unresolved issues per assignee +Verified on latest buildFixed +is still failing +I think that we should create documentation for in +Updated latest patch by fixing the Uma for the review +Committed as rAfter making an additional change to Clownfish as r to add typemapsfor the types to the Perl bindings I was able to perform a sitinttg for and have all tests pass. +This should have returned the same number as before but it did not +My project need can change the height size of try configurer in all property of component and CSS complete the event on selectionchange dont work properly is equal than must think to incorporate keyboard eventsif +Added to print all the splits +Works fine in Firefox +Patch committed in r extra support will be necessary for non address based hash codes as used in GC spys object model. +There is still empty line in patch vI use the following alias to obtain list of files modified locallyalias slstsvn diff grep IndexOnce you have the list you can use the following to generate patchsvn diff svn diff +NOTE The main reason for disabling this feature is that it is very JMS specificand confusing for users that dont use our embedded jms understanding what the mean can vary between jms providers +here is a patch that makes pruning the ticket queue ie +Any ideas +Double checking with the original reporter +Hi Mark Is this feature available in if please provide me one desperately need this because I can only assert one fact to rule at runtime and in the rule I will get appropriate objects and execute my am writing generic rules implementation accross as part of +I think weve already used the repository dont we +cannot reproduce at this moment +Its not enforced +The border can be enabled or disabled by a panel attribute. +This is awesome +Please verify and close +Did you do a full recompile after switching from to? being a class wont be found on but it shouldnt be getting looked for if the compilation has been done by +revert the commit waiting for repository open again +The a patch looks good to me +I expanded around levels and everything looked good to me +The workaround should be enough for now +Attach for reference +A better separation is good for everyone +Anyone want to take a stab at a patch +It seems to me that the problem is that is asking for the address to bind to and get is returning an address when has not been initialized +I just committed this +reopenStill wondering about though +Hi committersI will create an issue which is Japanese initial translations of admin console as new feature before longWhen would you commit the posted patches to branches +I can now confirm that this workaround allows the parse to completeSorry it took so long to get back to you +It is not intended to be an umbrella POM but rather an artifact that is released with the Spring distribution +I see for for +moved to M as the archetype sync task has you think this is still needed? +yes sir +I dont think comparing implementations of individual operators is going to lead to anything useful one because they work in the larger context of their respective frameworks and two because the real value is in converging on a single optimizer not on best implementations of filters or joins +Please dont hijack the topic and use the mailing lists +oh I see you use some other stuff please recheck +They all pass now +gcc compile LadislavPatch applied at repo revision rPlease verify the fix is good +Is it a legacy server that is going away? A legacy api that is going away? Is it the thowaway social data part of shindig? A permanent piece of functionality from another part of shindig that is now in land? Maybe you could explain the inheritance used here? It doesnt seem useful to have so much but Im also not sure where shindig starts and ends +Linux considers the requested listen backlog as advisory +We need to force a gc to start the cleaner ci m Fixing in JDK the gc call does not seem to be giving us the garbage collection so Stephen came up with this to force that to happen +New patch with better test +Reopening to remove Affects resolution after reopening to remove Affects flag +This issue is about the case where you do not want to use it +Its a serious serious bugThanks! updates on this issue? When people omit arguments to my service and have o debug to find out that arguments are assigned in the wrong order it requires time +Forgot to mention jira for other places where we need to use read range checkingI think takes care of all the other cases + makes the test call a command that does not exist +Still dont have the resources to get this fix into the current version +The component project name is. +Im wondering if somebody could review this patch and eventually commit it + +Tom looked into it today and its more than hes comfortable doing for + for the patch +The attached patch fixes it +If they want control the access they will leave the visibility to true and manage the data roles +I really appreciate your help and patience for this benh and I will have the patch committed this afternoon along with a fix for the error you just providedOnce the patch is committed there may still be additional issues on some linux systems but we will be able to resolve them fairly quickly rather than with a giant patch +ddfdaffcaefbeac +For mostly services such as the this mechanism alone isnt that desirable +I think this could produce some strange UI behavior when the buttons are not all children of the same container +Failing to take that advice is considered impolite +After a couple minutes I restarted the one node and it came up fine +Logged In YES useridCan you provide simple example that demonstrates the problem? Something like a stateless bean talking oracle and a client of that bean +Im not happy with the design of the patch +Cool +It supports the basic build phases clean install build test +here is a better patch +Are you using the multithreading in the DIH? I have the same problem but when I remove the maxthreads attribute the indexing completes successfully +Applied the nd patch at r +Bulk move to we can close the issue as because the Doxia feature is not available since me +bulk close of resolved issues. +My vote for resolving this bug decide which patch to committed +In general its safest to just find a committer whos willing to review it though +If lots of spew probably something wrong +Patch addresses the symptom and gets the UI working again +We talked with Emanuel and decided to fix the model description +Added locking to setReplication datanodeReport just committed this +Over time these unused threads build up until some limit is reached and broker becomes unresponsive +Please verify. +applied thanks +Added revised sample implementation to pool Fixed wrong name of classes fixed memory leak and renamed internal fields +I dont have a real idea on how to activate it just for some hadoop versions so I will do a last on the logs and propose a final version +Committed to trunk +This was committed in r +I dont think you should need to release this +I swear Im not able to replicate I have I should haveEclipse looks exactly the same but seems to work perfectly for meI really hate to do this but Im resolving it as works for me for now and going to let QA tell me if Im insane or not +Please list all of the views and editors that are currently open in your workbench including fast views and which perspective was currently displayed +And if you have troubles with plugin then please use Users Mailing List to ask questions. +Please verify +Committed revision Thanks to Gordon Kuo for finding this problem and helping with the change +Committed +Current version of the jspf sources +Thanks! Committed. +Thanks for the second patch and the additional JIRA Kim +JacopoAs this issue seems to be here for a long time I propose to delete the patches and related comments that are already commitedWhat do you think? oppose deleting both comments and attachments because it kills and history which is a big part of the point of an issue trackerIf attachments are replaced the replaced file can be deletedOther than that deleting such things should never be used as a substitute for closing issues and opening new ones +at some point well need to make a release +hbck right? This way they wont clutter the hbase when many of them get created during +Have a look at codegensrctestjavaxmlgeneratetemplate for a sample test case +Moving Clear transformation into that category and disallowing its undoabilityAdded to the Confirm Clear Transformation dialog message to warn user the operation is not undoable +Thanks Russell +I applied your patch Alexis because it solves problem in tests +patch for this jira to add resolver interface and implinclude unit and system test +Tested with Spring release fails about rd of the Juergen +No fun +Is there any way to configure hudson to not build dependent jobs when an upstream job is broken I wonder +Can you comment on what fields we should keep and what should be removed? We can then assign it back to Dart to work on +Since all existing tests are passing with this changes no need to have special tests for this changes +Keep the lease section at the end of the fsimage but do not save any records in this section +PreciselyI think there are a couple of benefits to the stream approach versus having many calls +Im assuming that there is something weird going on in that the test are using but Ill need some time to investigate this assumption +Can you please document this asap or point me to any place which has an example of the solution mentioned +testtake time +I looked around at the first level of upgrade code and nothing jumped out at me in between and trunk +I guess we could add a flag like supplementtrue onto svalidateAll +Shouldnt the expansion be independent of the presence of dynamic sqlI obtained the sql fragment by deliberately introducing an error in the statement and observing the exception message from the JDBC driver which includes the sql +This is a sample PDF with three text fields named Text Text and Text +More version of the backported patch +The tests no longer use raw keys which should catch this from occurring again +with empty meclipse goal just replace MREPO string with the actual local repository also read the referenced eclipse projects to detect use of classes in replacement to the standard targetclasses +The file has been renamed to and the associated references and reference docs have also been updated. +I think you should still try and use the first patch as I had some weirdness creating this oneSome problems you may have hit with the first one im using eclipse i checked out the module as its own project called because i already had another copy of it for another patch +Just noticed about the ciphersuites myself +The test in does this ok. +The links to nightly in should be removed +We could make it so that code can read classes generated withIf the method that takes the split strings and merges them into one with the string buffer before parsing is inside the generated class rather than the change would be compatible +What are thoseI think theres a bug to set some values since three system properties and are never used +This seems to be caused by an empty +Thanks +This works for me thanks! Im not sure I have rights to close the issue though +Will this one be fixedI want to submit patches but feel safer if the tests are working correctly +Any chance that some or all of these issues will be implemented? +not a dupe its at the other end of the character range +Joep Any chance of seeing the compile error? How comes trunk compiles against hadoop wo need of annotations? Is it because of the profile? If we do a profile option +Is the solution to this just to make sure that JBOSSHOMEmodulesorgjbossloggingmain +bulk defer of features from to +Please review that and other related issues and confirm whether you still have an issue with the latest snapshot version +Tim Fulmer I guess we are having this error too but our error isnt the +Preview of pig rpmdeb package +Nicely done +Resolving as duplicate of. +stack trace after deploying app and it looks like we made a bad assumption that Bean Validation would always be available as we didnt consider jar deployments +A dedicated logger has been added +looks issue as it has been released as part of release +Hi Claus I have started making the necessary changes to the Netty component based on your feedback +Im aware of these problems but it shouldnt be that big anymore like a couple of years ago when did not know what refactoring is +There is more work to be done here regarding distribution sites webpage etc however this cannot be added until weve added the issues of getting them up and runningAs I assumed this breaks my build locally so there is some work to be done before this can be patched against trunk +I did a search for classcastexception but didnt see it +I also dont see a reason to rollback the change +Fixed in . +Attachment has been added with description Example has been added with description Opening patch +We will close this issue upon the first release of Crunch +Do you describe your component using XML and annotations? I mean a part in the and some annotations in the codeIf so you have to choose where you describe your component either in the or using annotations but not bothThe iPOJO manipulation process does not merge XML metadata and annotations for a given component typeIn any case I agree that the NPE is not really nice +Is there an outstanding issue with the ear plugin to resolve hereI dont think so. +So is the JBEDSP being worked as a GSS case or will someone need to create the relevant Teiid issue for the costing problemYes there should be a designer issue on the dont think there is a JIRA yet on the cardinality +Here is one that hopefully works better +Fixed +bq +Sorry for the last comment assistant replied to Germns comment +This could go into the code review checklist as a reminder +Further refinement of the patch based on feedback from Abe White. +We wouldnt want to punish the algorithm for failing to recommend something I have rated but didnt like over something I havent rated but indeed would likeOne very crude way to do this is remove all ratings in the data set and see how many of those actually come back in the recommendations +Already integrated in the SVN HEAD +Look at to set up the proof of concept +Resolving. +This looks good except its not tested in the code +The hashCode is fine as the only requirement for hashCode is if two objects are equal then their hashCode must be the same which this implementation provideshowever as stated feel free to ignore +for the updated patch + I mentioned in the email accompanying so I think its not a Ok we need to drop this instruction from the ref doc this is a development convenience We need to mention that you add this! +I am attaching another patchwhich I think addresses your review comment +Please review this change +Thankshz +componentbehavior can be added to the page beforeThis is even how trunk works at the moment + and probably also need to be merged into a single goal oriented processorBoth dd and annotation should be processed to form the injections +The bug is actually blocked by because it requires aligned fields in the objects +Thanks issue has been fixed and released as part of release +And it is not used by OPIC +Hi Jerome Yep its user error +The issue is that the server that contains the repository will not let me do a chgrp on a file I do not own +When i now import the model jar it will import correctly but expose the mentioned +Committed after testing. +This patch is not as the changes to are addition the original code recursively deleted all records of the extend hierarchy while the patch does only delete the record of the class directly extended +The current is much too specific to an implementation of file based logging +Added a minor comment to the test Annotator +This basedir property should then point to the directory of where the imported file is allocated +Hi Dustine I saw the logsI see some issues +alright thats a fair pointIts a fairly small cutpaste change so Ill take care of that tonight or tomorrow +Set and for Committed r +I think we should move this to the backlog as the UI works fine without it and weve never had this feature before +Thanks Pete +Probably the problem is solved +If you pick this rule then you should read andunderstand the code in that deals with this case in theold indexes +enabling debug logging on would show if that last is the there likely to be any issues if we switched back to size tier compaction +Does that mean that old properties are deprecated or no more supported +Hi AtsushiI dont think you can do it just like in Java because C has no concept of inheritance +SHORT DESCRIPTION This patch add an Interceptor to besides that is responsable for prevent remote clients to launch exception +The change may be required in Trinidad +New Maven Indexes view allows to add custom repositories +Can you please explain more where these labels for the messages come? In the OMG Incident Management example pdf the message links have labels but in reality these labels cannot be found from the XMLInvestigating the figure that you pointed out it looks like its the name of the message flow itself +Please wait a couple days ill put the code on github or something like that. +So Ive fixed that +otherways i will try tomorrow +But if your columns are named pw instead of password arent you more optimized then a bit or bit integer anyway +I had some feedback from the Maven guys and I will push out a new snapshot release for in a few minutes +Also if we want to change pushdown to a boolean pushdownallowed metadata property that would be fine by me as there is little value in having the pushdown optional vs +That said there were some WARNING level logs that are completely ignorable that should be flipped to FINE level so Ill do that. +By default they shouldnt be escaped since they are already escaped +Committed +The Start operation was hit before previous Stop operation finished +Mark Woon has fixed most of this for the release +Attaching a test project that shows that there is no problem using a classifier for the WAR plugin +Good idea Jaco +The test is still disabled since it requires a serverclient scenario which for such as a simple case seems overkill but we might address it in the future. +Discussion on how to remove cited problem of MR jars needing to be in RMs CLASSPATH +And since connection with credentials still works we will stick with that for now +Code changes from Stack Todds peer review +ClausIve just uncommented out the failing test and have been running without any failuresCan you try this yourself and if you see no issues can you do the same on the trunkThanksMarat +Thanks Clebert +I am removing the fix version for +Use cases is described on the page +Thanks for your review Karthik +Bring down the priority of the bug as I dont think this is a critical feature for Axis +Both screenshots display Connection profile this is not a part of Hibernate toolsThe issue about error message in console configuration + +Thanks for the patch Jacob its great to be able to run these offline +I shouldve deprecated it insteadNow I dont really know how to proceed +Looks reasonable to do for the next my comment do you remember what the reason was +Patch for submittedCommitted by Ryan +Another concern is if a user can run the same application code againstie if I wanted to scan scan this would scan table +Patch attached fixes the same in a test while were at it +So youre saying that there is some sort of readahead here that is taking messages out of the view of a browser even if they havent yet been receivedI didnt see anything in the JMS Javadoc about something like have to keep this code compatible with any JMS server that follows the spec +Running it through Hudson +Ive opened a new jira for it +Invalid bug +you are not a fool +Thoughts +Before mvn site you may have to run mvn install true because the version numbers have changed +In the second patch that I sent for review some of the tests did not have the Test junit tag +Thanks +Most people will be best either with normal analysis or with creating multiple fields themselves if thats what they really desire +Thanks Sreekanth! +In fact the results look so good that I believe the rule should be enabled by default +I have asked Scott to know if we should back port to we will see. +I found a memory leak in store plugin during verification that I will log as independent JIRA. +Attach the superinterface +When I started to develop this simple I just downloaded the source code of and worked with it so I cant provide a svn diff file for the current repository of I dont think it is much of a problem +This looks like it requires some investigation still +It does not occur in either nor trunk +defer all X issues to +Please test with +We can open up a bug report in for this but it could just be by design +Rejected as Juergen reported this is going to be fixed in upcoming release. +This issue seems to be about issue where the filesystems are the same only the one used by the regionserver is wrapped in an +Started work on +The job failed eventually +Ill let you both know once the bug is fixed and the dependency is removed +If there is an existing parity file the code can identify that by counting the number of parity blocks with the expected number of parity files +Ill have a look asap +Committed +Sending message to multiple files using eip works +Having engaged my brain it is impossible to implement this as slink and sbutton do not produce postbacks therefore there is no component tree built to find the correct component for the action event +Fixed rev +Do you want to put together a patch which addresses this +Patch updates AND and OR tokens to now accept the lowercase versions. +Minimal patch for branch with tests attached + +looks reasonable +Sorry I forgot to add that this obviously changes the metrics that are logged and I am not sure if that is OK +I think maybe you should leave Patch available checked as the patch is still available though its clearly not ready for commit as isYour observations about the behavior changes that are caused by the patch are important to record can you attach any information about the test failures? Can you briefly summarize the behavior that was altered by the patchDo you have any theories about how the patch caused these changes in behavior +In general there are cases where its probably fastest to create instances without defaults set +I set to depend on this as we cant start supporting Ganglia until these basic Ganglia issues are resolvedI like stacks refinements +Ran on MAC OS X using Java +Attaching a few resources the server log serverfcfc more detailed exception dumps Jackrabit repository configuration should contain the issue was migrated to +Hope thats okayCommitted patch to documentation trunk at revision +In the diagram i can see this has classes for Php mysql etc +Oh I just found out you could also add an import statement of import to the Classes +Removed in rev +Sorry for your inconvenience +We rarely ever to releases before the one currently under maintenanceStep appears to be burdensome on contributors but in practice isnt so great an issue +All the spill files together could be larger so we could open one HDFS spill file and keep appending +I see those initial names plus many more in the deltacloud group + +which OS youre usingwindows? linux? Mac osx +The disadvantage is that is another component to deploy of course +Id rather get a release out and work on another than delay any longer. +Tom and I are arguing about how to fix the problem +The check is on every Append call +hmm +One of those is yours changing links so that it works in the zone +If you add this as a library to the EAR it should be able to load the class +Does the attached patch look OK to you +The first patch for WS entrypoints +Petfield string fieldName name notNullcontroller scaffold class +bulk defer issues to +I already applied the fix as proposed so lets hope it solves the issue +Would be sweet if we could get parallel into stock +Although Bigtop has not really enforced them that much but that could change if we start going that way +k It would be better to write a separate API that just fetches the actions of a workflow and use that for pagination +Thanks much for reviewing +No tests because this removes untested code +But dont just this JIRA without coming to some type of agreement +Ill make sure its included in the Release Notes +I think propagating the exception to the top of the jsp engine is fine this avoids cluttering too much in the jasper codeI think if we restrict the errorPage to a jsp page you should be able to get the jsp page by path if you look at the method get the jsps is createdfetched by resource path so using this approach it should be easy to call any other jsp from within the jsp scripting without going through the resource resolver or the script resolver +Amit picked this up. +This issue is occurring in current AS SNAPSHOTS and I have also filed Brian thanks for checkingI have read the record and I feel we are safe with triggering topics deployment and applications waits for topics to be started +will allow one thread to go past alwaysThe patch looks fine could you comment on the bug you saw with a deadlocked shuffle you saw once since I do see youve reverted back to original patch wrt the fetcher threads and stalling +and also hold connection parameters to the LDAP server +I tried to recreate a clear hierarchy similar to the real one but only with one problematic property +Seems to wok on our tests +We are done here. +Daisy Hi Weihuacould on of you review the patch of this issue and push it to the repositoryThis was really tricky one especially the test involving XSLTBernd Eilers was so kind to fix theWhen I wrote the test I found the class not intuitive and did some refactoring +In fact i believe the behaviour you are seeing is mandated by the spec +OK I confirmed that this report fixed in +This new patch gets rid of FOLLOWERINFO and only uses LASTZXID and it seems to work fine +Just attached a patch addressing most of the issues thanks in advance for considering it +As Mark says the tests need converting to Junit +works fine with +Opened issue for publishing the artifacts +Looks like I have to take this on myself +Definitely worth doing Im putting this against a footer always appear at the bottom of the viewport is a problem that is very difficult to resolve in CSS alone Ive usually resorted to using JS to reset the position see for an example of this sort of layout +I recently posted a couple of questions that I would like to see addressed before we put this Issue to bed +but of course if the underlying selection is not mappable to something the wizard can handle it should just do its default initialization +Attaching new patch incorporating fix for bug found by Hemanth allocating extra byte while doing strncpyTODOs change the configuration reading method to use fgets by allocating large buffer +Some of them cannot support this effectively and other would need to keep a cache as they are populated +looks good to me +tomcatx configuration is affected too +DavidSorry for the delay +Theres no urgency at all +In particular a change to conf that defines the filesystem module to be invoked for kfs urlsLet me know you have questions +Rob can you get these tests to run wo timing out? Or should we allocate more time for them to run in Jenkins? Maybe something is stuck blocking +At revision +If someone else beats me to it that would be great +When I have a project with java source folder that is an immediate child of the project the editor opens with all tabs for file projectNamesrcName +Sametime we are working on building the interop war +Adapted to new update handler frameworkI think this is about ready to commit +The problem is in changed JSF processing wrt RESTORE VIEW +Thanks for the review Cos! +Fixed in git repo +If we have attributesoperations available for an offline server thats nice but thats not a requirement for dealing with this specific issue +So please await further patches +Copied the magic gball files to their locations in the big and add the poms +One point to raise it would appear that this commit includes changes made for another purpose and is not restricted to the current issue. +the commit the precommit env bacame flaky and it seems that surefire cannot parse the tests results +I have it on Fedora distribution linked to GTK wont work with Eclipse JBT uses our own distribution +Build log and a possible patch for the problem +pushed +So all issues not directly assigned to an AS release are being closed +Should we change this too +Could you please attach or show me how to build itWhen you are done pls assign back to me +When I get some time Ill investigate further. +Just so that I make sure I understand the discussion about netty is to justify the solution of the patch right +I attached a wrong patch +attaching a sample project might help +The bundle does now contain javadocs and sources +However releases are packagings of multiple files and these are separable from releases and independently repurposeable +That would be great +Changed to which uses +Move to fixed in +If you build your index with you would need to reindex if you switch to unless we add the version so we can fallback on the old behaviour +I am facing one more issue regarding cursor only +Erick I am looking at on trunk and see you added this as a bug fix under +I agree +I did a bit of work but theres probably more that could be done +What attribute would you suggest other than the connection ID and its timestamp +Very similar to global config update? Thanks we just have to document this fact +Assigning to Snjezana +First we normally provide typing information but there are a few special cases where we need to deserialize an embedded object that we have no typing information for but are willing to inspect the raw information after deserialization +However well need a set of jars on maven so that we can start testing +Ive attached the patch + has been closed as a duplicate of this issue but still contains some valid info on the bug +bq +Until is finished could catch this and avoid the stack trace +I modified the to not install the plan file and let the deployer do it +So it seems to have been released already +Good points thx for the clarification Stan +Im running it through some of the ee tests +should not contain invalid keyvalue pairs +also move of M to to M because of no updates from +git is all setup now we just need the post commit hooks +Good idea Bruno +Put the version we want to fix this in to +Renamed the service to create and moved it to +Participants page e permananent instead of permanent chapter +Code delivered. +I realised that the could not be serializable + +I can change the existing serdes in hive codebase but there may be external serdes out there which I haveno control overWe have to take this hit sometime +Marking closed. +classes also retain the original charset of the parser +is this a regression in or was it also the case in +Enis I was thinking of getting things more likeI moved metrics about getputdeleteicv into a helper class that mutates data in +Added filter to all places version property might be displayed +FACE PALMoh crap + +Fixed in svn revision thanksdims +I will move this to hadoop + should be fixed now. +Would anyone want to try this +This means however that the constant that names the subdirectory name should be owned by and not by +Thanks Gunther! +The initial features I intended are complete +Erin I was wondering if we could either make the fonts larger to fill the extra space +Closed Issue +Here you can Ignore leading path name segments pick the number in that that causes the below Patch Contents window to not contain File does not exist errors +Ether hudson is still down or its IP has been changed so I cannot access it to verify this issue +The build runs fine without any test failures. +Updated for new svn +I think originally I had code in there to with a DC during a decoding +Patch for this is attached to +The problem is that dependencies from the custom provided scope are not added to eclipses compile classpath. +Do this to verify Window Preferences Tools Tools Runtimes click Download +There are still more places that probably need to be updated as well like and anything else that does content type negotiation will need to be updated as of your existing unit tests pass with no change +Taking the first answer as the correct one might work for some but not for everyone and that is why I believe we should choose an implementation where this can be done independently of the name finderI posted on the dev list about name merging please continue the discussion about it thereShould be easy to store the name in the dictionary have a look at the implementation and our xml parsing code +a Maven issue +No I dont have free cycles +It may help you to check out the tag of the surefire version you are running +Is this still a known issue? +I did a test with a windows path and it escaped the back slashes and the colon for some reason +Its fixed. +Patch committed to trunk +Applied apache profile +Thanks Edson + is available within maven central +instead as The umask used when creating files and directories +Going to deprecate them just because they need loving and because they are expensive now to runNeed to make sure this patch fixes shell STOPROW before committing +You are rightThank you Scott for this efficient answer +Looking at the exception has jogged a memory I believe there is an installation order dependency with UDDI +I dont think it would be a major loss now with the conversion of this test as we are currrently deleting both these lines when run in the old harness +I have made changes for this as well as persistent inner classes +It should clear state that the suspend is done for a tasks process +anyway given the patch is almost the same for and trunk I will take the above as extending from backport to actual patch +I have checked Cordova Feature Discovery dialog one more time and scrollbar is there so either I was seeing it incorrectlly or scrollbar randomly disappeared I cannot reproduce this issue anymore +Hmm +Actually they are a part of the patch +Closing. +A tool that adapts JSON data to schemas could be generally useful if it permitted enough control of how the adaptation is done but might also be rather + Felix +Adds timeouts for the connectread in getFile +This patch only contains proto files change which does not affect generated java code +I think failed testcase is invovled with +fixed already? If not release blocking +Rebased corrected compiler warnings renamed the package to and added an alias for the in +Hi DavidWould it be possible to submit a small unit test that demonstrates this problem then we can take a look at it in more depthRegards to Daves comment will the help in diagnosing the problem? Not too sure what kind of unit tests required for this kind of informationThanks Can anyone look at the issue again? This defect blocks me +Thanks to Balaji for pointing it out +Keep this open until the test gets fixed +Please post the WAR and plan youre using to this JIRA so we can try to produce a better error message like deployment failed due to invalid deployment plan format or whatever +Is your manual execution a real execution or a dry run execution +So removing the link issue +Found I had to deal with Session instead of SessionBuffer because uses them explicitly +Using without the Hibernate Console +Either part of or dependent uponany reason that my pull request has not been accepted? +As suggested above I moved the detector classes from to subpackages in revision That should complete the last remaining open issue with this feature so resolving as fixed. +have just committed this. +Could you also attach the new test case +Is there a way around this? It seems like a nifty feature to have. +I doesnt manage to reproduce this behavior locally +Thanks for the review +This makes it unnecessary to transfer the record count to the reduce lets the format remain exactly as it is and should be fairly easy to implementThoughts +Adding corollary to other implementations + +There are still a few spots that still use the modello reader that we are going to clean out. +There have been several recent class loader changes introduced into Tuscany +You quoted an earlier comment where I hadnt done as much investigation as the one where I start with Heres what I know about the different problems where I aimed at dumping my whole understanding of the problembq +This had to be related to datasource issue in because after fixing that issue I cant reproduce it +To be really sure you usually have to close it in a finally block. +ok this is now done in EAP seam too. +Once it becomes stable we follow the standard deprecation rules for API that protects the policies +It gives the same error Object id must be set +The text of your comment is sufficient to confirm the patch license. +Agreed +I dont remember why our own standard is maybe brett does +Added a test to to test that a collapses into a single MR job +The first stacktrace mentioned in my previous post +release notes attached to this status to resolved since the doc is ready for review Darrin does this have the final bug lists +Thanks Gang Tim Liu +Theyll show on site next time we push it up +Actually this Jira is +Hi attaching patch for fix on the issue thanks +Defaults can be get by +Fixed in revision this issue +This could be done inside or in the caller the first is easier but would end up masking the lack of a native bbox to the UI and whoever needs to know it has never really been declared +I am facing a similar situation +Yes that would be fixed in DR +httplocalhostservicescheckInv httplocalhostwarehousecheckInvmethodcan As I mentioned to Ross Im not sure where Mule stops and starts on this issue +I would really like to see people creating an for each +I think that we cant really help the enums in containers problem +Thanks Matteo for the review. +Tricky because there are quite a few moving parts +Im thinking this has to do with the bonddevice since its normally supposed to look in procnetvlan to see ifthe parent interface is tagged and then look up the parent of THATinterface +If you have ideas of more tests for this feature feel free to suggest themRerunning regressions +Closing stale resolved issues +nkeywal i change this +Max count of fat locks is equal to bits used in lockword for FATLOCKIDIve checked this patch on IA WinIA LinuxIA Linuxemt +This technique is documented as part of the highlighter package +Can you include the test case also in the patch +As was discussed with Freddy we are not going to add additional validations because Sonar Runner provides meaningful messages. +Thanks +ERRORRepository read access deniedfatal The remote end hung up unexpectedlyI can clone other repos +org version Tools Eclipse I indeed worked on this for the GA +Appreciate any comments before +METALooking +Will this change be compatible with older point releases? An old client and new version of RSMy bad I have not done this the st time +Reopeing for family as m closing this as is now a factory for Forms so a subclasscould create forms however it need to. +Assign to Devin to review it +We just switched permission schemes and you werent assigned anything +I agree we need to run the schema validate task in the CM interchange format ANT script +I have got the problem with deploying of this app so I created the jira +This is the second suggestion of a +Updated patch file against geotoolstags and a sample data file with duplicate column +Im in the middle of working on fixes to correctly support in my fork +Deleted and recreated as requested +Submitter mistake +Log output from my local build which failes same stage as jenkins build +Could the fixer of this JIRA either create a new or incorporate into an existing test a test that would create a for this function so that we do not regress + +I was fairly certain there are no masters that log the info from sysinfo in that much detailBut just to be on the safe side I did run derbyall and suitesall with ibms but saw only failures related to the new errorStream testThe only relevant junit test I could think of is which doesnt run in the nightlies +Done +Ok it makes sense +Closing as done +Thanks Regards +Patch to fix ISO image magic and extended the buffer size so that the magic can match +integrated in revision thanks for the improvementRustem please check it was applied as complaints assuming its OK +I dont think we need a use +another documentation please add your FAQ so I could apply your patch +Ive created the symbolic links for the version less references +Patch for patch with unit test +is it crazy to suggest Instantiated +Ah I misread +This code has been used to remove applications and manage multiple applications in a single runtime +Providing patch +Sorry my first answer was wrong +The patch does not seem to have the new file and its Suite change is out of date +I created to cover the task of fixing the test failures caused by this patch +The log line string was not evaluated when no the correctly returned false +Now that you mention it Yuki already added that in +Will throw an exception +Committed to trunk +As Dave said this is unlikely to be a Struts bug with a test case if you still think it is. +GSG is checked +jiri is there any particular section of an ee specification associated with this tck failure +With the fix reverted performance returned to normal. +I buy that +Yes will take a look +I had assumed that this was cause by the sender assuming that the receiver had cached the needed values but looking at the extracts closer I think that wrong instance being accessed might be the cause + creates a new wires iTest that wires two components together +The patch file is now attached +I made a simpler change +What about the final release though? +I like it +Pardon the above thrash +Should be fine +Changed priority from critical to major +So this issue might need to be reopened +Jimmy Wilson and I have been doing some testing of SESSION granularity with CCL in the config +Thanks SaschaYour patch is in trunk at rBTW I had any problems with Subclipse +Hi can u check the createProfile +this patch removes the mkdir p chmod and chown commands on the datanode dir from the script and letting the dn to create the datanode hadoop with this patch and deployed a secure cluster with the above script and the cluster came up successfully +Everything still seems to work OK for me after applying so Im thinking Ill commit the patch +No new tests because Ive only seen this issue manifest in interacting with postgresql +They are not related to this patch +Remove references from +Ill get the test done asap still committed to it Im not sure it is needed though for this issue propagating HTTP Servlet references is needed for some of our demos to work but it is useful irrespectively of that please feel free to close this issue. +Verified on still exists in +The only test that fails for me is test which is fixed as part of merging from trunk to +Its a bug of deployment +fixed bug in delta cas serialization when the CAS has not been modified in any way +Can we close this issue? It looks to be fixed already in the trunk for all pages referenced in the patch except for panel stack which isnt working at all +That means is not a replacement for in such cases when using configuration +In the new version it is cleaned +Fixed in Rev . +i am happy +Whats the problem here? I would guess that its the dependency on a C compiler +Patch to directory +One thing I didnt know how to work around with direct ivy caches was jar checksums I think it is a good idea to keep these but then I dont know how they could be verified if we use ivy caches directly +Duh wrong patch +Pushed back to CRM is about AOP integration +Ill take a look at this today +Can you be a bit more specific and define invalid? Can you show the problem based on a sample request +Then B acquires write lock on R and tries to lock R however A has acquired write lock on R at this moment +Almost done +I will file another jira for investigating how to get the web page execution out of the namesystem lock +The handling from the original patch for cygwin gcc build on win is worth keeping if this is going to evolve to something that is as cross platform as possible +Quartz job bean this is not fixed +After that I plan on focusing on fixing refactoring bugs +thanks Erik! +Disallow to upgrade by checking for LV version with an error Upgrade not supported +any workaround for this issue +My last comment was wrong the impl was changed before commit so it reuses RAF +I build with the patch for bug and still see differences in some of the output files +Of course the Executor interface should be a backported version of Java +Not idea how to reproduce it +Results with topNnumSegindexsortBymethodquerytopNhitswarmqpswarmnewqpsnewpctgsimplecountryvaltext simplecountryordtext simplecountryordvaltext simplecountryorddemtext wikititleval wikititleord wikititleordval wikititleorddem wikititlevaltext wikititleordtext wikititleordvaltext wikititleorddemtext wikititleval wikititleord wikititleordval wikititleorddem numSegindexsortBymethodquerytopNhitswarmqpswarmnewqpsnewpctgsimplecountryvaltext simplecountryordtext simplecountryordvaltext simplecountryorddemtext wikititleval wikititleord wikititleordval wikititleorddem wikititlevaltext wikititleordtext wikititleordvaltext wikititleorddemtext wikititleval wikititleord wikititleordval wikititleorddem numSegindexsortBymethodquerytopNhitswarmqpswarmnewqpsnewpctgsimplecountryvaltext simplecountryordtext simplecountryordvaltext simplecountryorddemtext wikititleval wikititleord wikititleordval wikititleorddem wikititlevaltext wikititleordtext wikititleordvaltext wikititleorddemtext wikititleval wikititleord wikititleordval wikititleorddem +They do not seem to be used anywhere as a dependency +the Responsibility Type functionality +No such module +Since before app summary log may be missing if RM crashes after app finished but before the app summary is logged +Mathieu explained it in textHowever see his subsequent comments too +The nd iterator is to be a subiterator to print out all annotations starting within the aaa token after the aaa token itself +Final RPM needs to end up in target and be signed when ant sign is invoked. +We can look at moving this into at some point +There were entries that dealt with escaping special characters and such +Apparently the JIT youre using isnt +Its for document bodies but attachments themselves are not exposed to a doc update handler only the related stubs. +Patch Im committing +We do not generally clear such lists on shutdown since they are part of the original configuration state not of the internal runtime are you holding on to the instance after closing? Otherwise the garbage collector would perform full cleanup anyway including nested state of the instance +release no +In PDF versions the numeration in footers and titles in headers are situated variously according to the page its necessary to dispose numeration to the right bottom corner and page titles to the left upper corner of the page +Merged into piper branch commit cbaeffdabbfbThanks Daniel +patch for you provide an example of what isnt working? And what you expect it to do +Sample xml document works with current code. +I have fixed but I can not commit now since build is broken due to some other reason +Dave when you say a pull request with a test case what do you mean? Unit tests or something else +Closing. +This patch generates a bit integer instead of bit +At r +This may be related to because it is the only issue committed between these two builds +The same issue applies to itestdatabindings and samples +With all the new security stuff in Jackrabbit and later I think this whole issue is already obsolete. +That is users causing a conflict might not have sufficient rights to see the information the patch includes in the error message +Isnt this issue the same as? It is better to have patches in one place than two +Experiencing the same issue +The spec says nothing about timer persistence on a controlled shutdown +This patch fixes an NPE in optimizer tracing +Since houtputScript and houtputStylesheed will be relocated it is responsibility from the listener that relocates check for that condition and set the id properly +See the linked issue for more details about this feature +I guess well have to add a service property which allows services to use the Reader method instead of the new method +Comments in code speculated that we should probably tend to the more conservative +what is the HDFS location for tables in session level metastore? is it the same location as regular table +Karol Ok thank you for confirming that its still an issue +This is likely an area Ill be putting up some patches on to break all this stuff out for pure YARN preparations maybe Ill take this one +You bet happy to review +Thanks Bernt for your info about SQL standards +Can you check if that works for youThe reason i wasnt using the logger inside security was that it was doing security sensitive calls during that time which created all sorts of problems +I can send settings files for using external database. +Hope the jira project setting is project and not global +visha flavio If there is just one thread running at one point in time then its ok +So the number of codecs created will be in the order of +In particular the imagejpm format is declared to have an alias of videojpm +Im not getting on this url +This issue was fixed on by way of and on trunk via +Absolutely Ill take a stab at it. +on Shutdown server it will failover to server Run on single point server failoverejbdsusesp on trueAdd Ping Sum So close it. +So the solution would be to pass this pointer to the transport dll when it is loaded with the state of object preserved +Is lowering the clients idle timeout for reads a workaround +Reader ReaderAgain what about bq +Verified on JBT +Hi Nicholas I agree with you completely +Had to get myself more familiar with Mockito but I think this patch should do the trickTested it on my end and it passed +It would be better if users see hdfs s and file urls +for what is this needed what else I could to with this implementationhow the calls look like +I have the most recent release and I am still receiving the error when I try to use an object with more than fields. +This is no longer an issue with the new component +Closing as suggestedKarthik Do you want to attach the github link you mentioned? +If there is uneven distribution of consumers across the cluster then you must have a to prevent starvation regardless of how is setThis appears to be a configuration issue to me +This is fixed in way it now throws an exception saying the eager load group is not defined instead of returning null +might have something to do with the way in which Java is installed on Mac +Ill put it in when I commit. +The default hook is to print the message that has been printed in the last couple of releases of Gant done by explicit writes to the check whether the variable is null is made before calling it so setting the variable to null is a very quick way of switching it off +Thanks +Closing. +Need to do a similar fix in the for the current snapshot versionThanks David for the fix +For Too many open filesTotal time Tests run Failures Errors Skipped Many tests hung or are not in the report so the final result is not meaningfullBut it will make more sense to try when the ulimit settings are updated +Solr should always use the same simple locale +See screenshot where also shows a negative count for this test case +I ran all unit tests manually +This has been re introduced into esb kits can you take it out again? +Source for scripts EAPSRCmainsrcbin EAPSRCserversrcbin EAPSRCconsolesrcbin EAPSRCmainsrcbin JBOSSWSNATIVERESOURCES JBOSSWSNATIVERESOURCES JBOSSWSNATIVERESOURCES JBOSSWSNATIVERESOURCESSo going to have to modify the WS scripts as and when they are installed into the build +Committed patch to resolve issue +Could you please provide a DEBUG log of the SSL filter +Ate I just committed a a fix for this +Users can consult the doc if they need further clarification and typically people picking up fixes on the branches dont rebuild the doc anyway so wouldnt benefit +QE complete +Im no expert in how this is supposed to work but the comment in the trunk code makes me wonder if the patch is correctWithout the patch a URI withdiscoveryURIfoo will result in setting the discoveryURI and other will use the defaultWith the patch IIUC the default will never be used +I reckon is the cause +Thanks Sijie +guarding against a null after release +In particular they are using HTML quoting and not URL quoting +After looking at it a bit more I think I found where the problem is +If everything looks good Ill go ahead and commit +preliminary patches +Thanks E +Shall this be marked as fixed or there is some work left here +is not yet started at the time we need to register the +Also switch to use debug build for testing +Xavier +Also if you just upload the patch again with the same name JIRA handles things very nicely +Tomcat +Step two could be to actually look for database artifacts that should get cleaned up like log seg tmpThere may be smaller fixes that could be made if whatever if causing the db to notbe created is something other than the process going away +bq +Thank you very much +So all issues not directly assigned to an AS release are being closed +Please let me know if something is wrong +The review is needed also for Panel Menu Group and Panel Menu Item I will not open new issues but will recheck after completion of this task. +Added statefulness some convenience methods to support transition +Ive also changed the default deploy directory to applications for wl and autodeploy for wl +Thank you for the bug report +See previous commit note breaking test projects into smaller files for JIRA +Please select a method of payment +Fix by r in need a fix for trunk +should be solved by +Feel free to provide input there as well +Thanks very much for reviewing this Knut Anders and for catching that issue +This is an issue with the examples rather than the clients updating component accordingly +BETA is almost out the door +cant test if the dependency is resolved properly when our tools are installed into a clean eclipse +Attaching a repro that exposes the leakThe repro repeatedly boot a database performs some db operations and shuts down the database +first installment of help files for full documents +I tested folder names with spaces and file names with spaces tooThanks so much for your effort today +Once you enter safemode manually it makes good sense for it to require a manual command to leave +Resolved by commit on resolved +mavencache file updates making sure we have the current should fix the problem you had +These tests used absoluted path to get local wsdl file +Unless you think these DDL constructs change often I am ok with maintaining them manually until we find an alternative +Ok +Together with this would allow Lucene indexes to be efficiently served right out of hdfs rather than having to create yet another copy on the local disk +I think we should also include this one in the it is even more important than the issue we committed for the +Whats next +Taking a closer look now +Maybe there should be another way to handle this? Maybe a command to strip down the file installation size given requirements such as English only +After more thinking and discussion with Stefan over IRC we think that leaving the table display as it is right now is probably the best solutionAdding the fingerprint we make the label way too long to be displayed correctly without having to resize the window and this kind of situation where you have multiple certificates with same subject DN is not going to happen that oftenAdding such information might pollute the display for most users +Leo patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Maybe the is left over from the early days of Jena before that was the caseI will go ahead and remove the cache entirely +It seems to be broken +Ditto get numbers before commit and yes a write up would be helpful +I was not sure if that was handled by the first case +This one fell off the radar +This is not a pretty fix too many references exist to the shared properties object in +I think these notes are sufficient to document the behavior +Implemented added unit tests to verify +Loooks like fuser does not work as published on rhel +Initial code using Maven for build management +is meant to be more to ensure that the application submission should fail if the AM memory request exceeds max allowed +resolving issue for next release +Added apt file +After all the main idea of is having reusable modules which you can just pick install and use preventing reinventing the wheel over and over. +Patch that fixes all the checkstyle violations +Not sure when Ill have time to work on it though +The pull request in that JIRA has the updated schema +It never fails that until you open your mouth about something you figure out you were probably wrong in the first place +issues +Im attaching a design document for establishing authentication of the datanode to the client +I think Hibernate will still generate SQL that has an in it since that is what is in your HQL +Thus it was looked into as the best stopgap for MIt though turned out that the operationscode needed to do that is more or less exactly the same to be able to provide the workflow I added above for fresh new eclipse projects so that is now what we are trying to get at since that usecase is more important in merging from existing git projects +We currently dont have any way to pass the list in +Refusing to fix it is irresponsible +hmm +one direction Im thinking is that the job executor kicks in before the process engine actually is ready. +Abhishek I get a abhishekkapoor cannot be assigned issues +Requires build against +I do not have time to pursue it tonight +The current tests sometimes check for exact error messages which gives me a little heartburn around maintainability of the testsOf course there wasnt an easy way to do anything else besides check for an exact match so I whipped up this patch A special expected string of regexmyregex will match the encountered value string with the regexThoughts +Im thinking we should expand the scope of this issue and also include Results +i had never ran into it because i always hit Enter to exit the cell +Closing resolved issues. +Ive seen someone complaining. +Should we therefore not postpone this issue until were ready to release? That would minimize impact on open issues and it wouldnt make this task any different +Just modify the update the provider from DATANUCLEUS to HIBERNATE after generating the source do mvn test +Using static typing should restrict you to the methodsproperties available on the static type otherwise theres not much point using static typing +improvments +Were not supposed to link directly to the wiki due to the load it causes +Here is our exact scenarioSome bad code resulted in a Mb row in one of our +Okay I was playing it safe +This will improve performance for user deployments not so much the overall since we will likely not have annotation driven deployments for core services +Not much different from your patch Ted + for separate map reduce cleanup and setup +Yes definitely +Canceling and resubmitting patch to trigger the QA build +Cayenne itself refers to a few other classes missing in Android I wouldnt mind if we internalize the alternative introspectors +Until then Im sorry to say its proprietary +Take a look Ill commit this in next couple days if no complaints +But it doesnt replace it since we still have the namespaceID inPerhaps a nice compromise would be the following hadoop namenode format gains a required argument for cluster ID +Changed to make a copy of the eval spec before altering it and pushing it into the combiner + +why not collect disk io and bandwidth for scheduling as well +Looks good Paul I will commit it if the tests is failing I havent debugged it yet +It seems we need to have two classloading methods +No information has been provided to duplicate or fix this issue +Marking this as an improvement since the current setup works in most situations +Feel free to close this issue or leave it open as a reminder to update to the latest once it gets released +Fixed with commit +bit vm on windows wont use client +Gary did you select the correct version of Seam Facet in project facet configuration? You will only be able to choose from Seam runtimes which matches with seam version tou selected in Project Facets configuration +Fixed by r on the trunk and r on the branch +It seems to work +Browsing the scroller is acting in the same way as clicking a commandLink so I presume this is related to this same issueregards there is no more autoform in the current release this is no longer an issue. +There is a Dparameter you can pass to package a different version of jsvc +Suggested workaround is to uncoditionally reset the event after each interruption +This is fixed in Geotidy +Ill assume the maven central orgosgi version is the official one +Plan of attack is +But as I saw also in README of branch I finally thought it was that +UDDI section max for lang is actually characters +Updated empty input failing test case +Sergey could you please check in your linux environment? Seems cant be reproduced under windows +Created. +the patch looks good to me +Wait sec for Jboss to undeploy it +Let me talk to Max Pete perhaps CDI Web Project offers real value + fo printRendering should fail +Attached patch +Steps shown above dont seem to work +Added test + you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Let me expose a new method as a fix +We can handle this on a basis sessions are explicitly not thread safe so users should only have one thread in a given session at any time +I just committed this to branch +Fixed thanks for the patch +Attached the patch +Thanks +bq +I uploaded a simple patch for +Consistently adding tests when changes are made is by far the best way to have a large deep regression test suite which helps future development +built a patch testing s not right to have unicode in +Unfortunately my editor reformats the includes and tabs +Doc type attribute added to dragListener dropListener change at snapshot +thx that did it +refactored the codes +After it you have to kill the Node and restart it as Standby +Closing resolved issues. +Im sorry we cant change this. +Committed trunk revision Committed revision close for +Error message still misleading as of +Can you please explain how this issue is related to Axis + +But these tests cant reproduce current bug now on my machine +The default buffer size of was chosen because it gives good performanceHave you any performance tests that show otherwiseIf so we can consider implementing this for all the copyLarge methods see bq +If we ever decide to implement truncates which is a reasonable extension of append this whole logic will have to be reconsideredI think the decision should rather be based on generation stamps etc +We should be able to reuse the Property pages from the Properties view and pop up a dialog with those in it +I committed this. +Part of a bulk update all resolutions changed to done please review history to original resolution type +You did misunderstand me a little +BULK EDIT These issues are open to be picked up +It sounds like that is your opinion that is not correctbq +If people are using the store to build map with overlapping features having a predictable iteration order is important to get the right outputAgain a solution were a flag controls what kind of map to use to privilege scalability or iteration order guarantee is likely the best we should probably get discussion back to your actual bug report +This made the change more straight forward in that we didnt have to change the args passed to +Could you please fix the warning in compilation that was added by your patch? Ill commit it if tests pass but new warning is not good +even the old code also remove the scheme and authorities part when try to match partitionDescno offense the old code is not very clear and not efficient +Mahadev Nicholas can one of you take a quick look at this patch? Its a simple one and Ive already tested it with + +Honestly I doubted whether it had much chance of workingSomeone had proposed pulling out the anonymous pipe transport into its own file which has its merits +Screenshot for the updated instructions attached +Thanks Craig! +Crawl completed successfully here +Move this to +Manually done. +Running maven build with creates an Apache ready source release by using the predefined project assembly +Also access checks while traversing nested directories seem to work a little differently on Windows and Unix +JeffThose are excellent suggestions +i would need similar instructions in a environment +Related to +however be warned that there still might be an issue +bulk close of resolved issues. +dkulp Are you saying this is done and ready for updating on the webserversYep +bq +cvs dpserveranonymouscvsrootjavagroups logincvs z dpserveranonymouscvsrootjavagroups co P this problem be linked to could not flush the cluster for state retrieval problem on other node joining to the cluster +This may not be easy to spot when looking at the code either since the fields may be in different classes within a class hierarchy +MaxCould you please review the patch +Committed the test case into camel trunk will revisit it when we have time to rewrite the +According to discussion with Freddy current behaviour is correct +I agree the new results are now more ambiguous +I am handling the design documents which include the answers to your questions and i will upload them as soon as possible +Actually I read the report as the current lang code relies on a Sun feature and should be changed +I could imagine some scenarios that would make sense if we have callbacks or multithreaded clients sharing the request the client code would send a stream of gets puts and writes +Kurt is going to try to get it working within Guvnor +Marking as closed. +Ive got this error with default configuration of hibernate and j +Id imagine most of the of the other TZs should be AmericaLosAngeles since they probably assumed Pacific time zone +Eclipse project that reproduces the for reporting and providing the patch. +The links have been updated for the next ReleaseSee attached patch for change details + +closing since is now released +Could it be some app engine issue? Our apps that deploy to regular environments arent showing any problems +David how about marshalingunmarshaling a sample document +I was the one who pick this option and wrote the xdoc reporter for Findbugs and added this feature to Maven Findbugs original Maven author decided to create a report directly in HTML +Yes you are right +but neither seem to other information can I provide to help diagnose? in the meantime are you sure that this is actually caused by some XML files not being loaded? That should become obvious when looking at the debug log for the package for I took out my controllers by commenting out the component scan for them and now Im getting other issues for in my it appears that anytime I define messageSource in my applicationContext I cant autowire it +Accelerator keys and context menu are all working +xhtml files in normal java projects +Many of the tests that are really tools are benchmarks so we could call them that +Is this still something that you guys are planning to change at some point on Android? Main reason being when uninstalling an Android app the files remain on the sd card which could cause issues if the user the app in the future +Its a local variable so shouldnt affect subsequent requests. +Yes the incubatorbigtop directory is now empty. +Even if I know that a particular process in the Job Object is a batch script I cant avoid sending the console event to it +this change might break existing code which may assume large precision +As always I welcome being educated about what else might be relevant in the specOracle supports the following coercions but not the inverse coercions and Oracle documentation does not address localization issues DatetimeInterval NCHARNVARCHAR Number NCHARNVARCHAR does not advertise any ability to cast tofrom national stringsDB and Postgres do not support national stringsIn summary I do not see much guidance here +I think we should keep and add exit status as well for clarity +well the problem remains independent of eclipse using the ant project build +KevinI will take a look at it +Committed patch to trunk with revision ran fine with sun on xp with patch the comments as suggected by Mike and committed patch to trunk with revision +Would you suggest a way forward? +The best solution I have found so far is to implement an analog to Criteria API which translates queries formulated with objects into HQL queriesAlmoust with Criteria API but produces HQL rather than SQL +I also ran both and with JDK but without Xalan and the tests quietly did nothingIn particular when the tests passed I did not see the diff that I had been seeing in my environment before +When it didnt work I had to follow the tutorialThis issue makes the process of deploying jetspeed with mysql impossible +Still happens in Axis +Added a small modification in query parser to treat as regular joins since the algorithm is redundant in the local mode +Another option could be to use JCR for storingversioningaccessing the configuration +Some updates +Added thanks will appear soon +I dont see this going into +So Im reopening it +fixed at with the release of. +As for generating from the POM can the POM express ? I didnt think it could +Obviously JDK code uses JNI and loads in bootclasspath just fine +Already updated examples but need to update the others too +This is an inputCalender +It timed out +Yes I added parameters to turn each of the goals on and off and I turned off for testing +Thanks Graeme for your quick that did fix it +Hi TatuI also tested it with a clean simple Glassfish app server with a single servlet from where Im calling +After completing the aggregation for that job it contains the logs for the tasks which executed after restart +You just need to use DJBOSSHOME +for the patch +This improves the previous patch as followso Renames the webpage to trunk and removes it from the publishedapi directory +So I am assigning this to you +This error indicates to me that the read request itself is timing out which means the client is making a request and of receiving a response for the length of the readTimeout +And this works on both versionsWhy is not supported in thisCould you summarize the issues foundsolved +Marking as resolved since several new test cases verify that the behavior is now corrected +I agree in hindsight it would have been better to have the alternative default but for backwards compatibility I think the current coding will be to put another way if we were to change this default we should also change the default of to have trimWhitespace it would be best to bite the bullet and do both of these as it would better align them with +Ben do you mind exporting the Spring Batch namespace as pdf for those of us who arent using a Mac or Safari +once addressed +Went to review the patch and found it no longer applied +Doesnt the edit when option is being added to the jointOptions list only take account of X options shouldnt all options be prefixed with F +This is the condition the backup has and then on primarys failure its watch on the primary is triggered and above wait wakes up +I tried running trunk stomp tests on windows x all passed +Please let me know if you have objections to this approach +Someone didnt remove the comments and then handed off their program to someone else and it lead to confusion +verified the reported warning no longer appears after kens fix. +Shouldnt these tests be included in a normal buildI have had a quick look at it and it seems to me that these tests are targeted to existing running applications and not to applications being developed +Integrated to trunk. +A that is wrapped to support functionality would also work +This is the best type of resolution nothing to do +Ill create another issue for that +Moving to to fixing issues in JBDS +Its still a valid Infinispan demo showing the shared cacheI definitely should have realized this solution sooner +Checked that there is no javadoc warnings and the link works correct +Same goes for employee in role can only view partiesparty groups that have the role Dealer or ConsumerWe had to use custom data entry screens and a small java service to implement thisIn addition we implemented a concept called an Organization Context +The latest patch doesnt include any tests but and verify it +Alternatively we could set up a parallel jenkins build and commit this on a branch to see if everything is all right +Please cross check and then close this bug +Please raise new tickets for bugs +Id think that its reasonable to assume that server mechanisms is typically a superset of that of a client which typically contains only one mechanismprotocolbq +Integrated to trunkThanks for the reviews Andy Matt and Stack +We need to make sure this will work on Jenkins builds +This issue uses a new entity defined in so it needs to be committed after Pranay Chirag your patch is in rev +new CTF test case +Thanks for the fix! +Sorry to bother Fabrizio +The C version compiles regardless of whether or not the return types differ +I had always seen this issue as a diagnostic improvement to help users track down deadlocks in their applications rather than a diagnostic to track down Derby bugs +More work on inflow contexts in the txmanager component rev +Attached is a patch to include fix for +KelvinI agree with this approach +Antoine No I wonder if it is simpler to shade the bundle though it mixes tests and classes in the same packages so that should probably be filtered in the build as well +The same issue in JBT Reopening +I do believe this would be a nice addition to the DIH product +Sure will remove the extraneous import asap +dependend on having patch applied. +what is this about? important for +Tried both import by location an name +Hey Grid +hopefully we can use this ticket thread to raise issues and answer questions that can be incorporated into the documentation AndrewDo you know what the license is for the and thedropbox SDK jarKarlSent from my Windows Phone +bulk defer to defer to +change permission checking for overwriting a would be good to at least run namenode benchmarks before we commit this and perhaps even a big sort +Thank you for your improvement +Some minor changes setClassloader was removed in favor of making the implement +Ill try to write a unit test shortly +Or did you implement it separate to not allocate the extra array if only quicksort is calledExactly +feel free to reopen with details on how to reproduce it +Cleaned up Xies patch +doc +Another concern is handling of the available resources for the scheduler and reported headroom to +adds the +May be it can be helpful +In case of closing and reopening iteration in the same day the closing data sample should be overwritten by starting data sample +You still havent explained why lifecycle is needed for comparators? It seems to me that you only need task level lifecycle which should work with JVM reuse as well. +Tihomir do you know where you are getting that funny syntax from? Is that just how Maven wants it +A bug has been filed with Sun which has been assigned an internal ID for review of the issue +Thanks release note for publication. +On IRC was asking if could add history for +I verified that kill QUIT works on Geronimo running on Ubuntu +Thanks +Java code for curve fitting and a spreadsheet with results +This was easier to do because of how the plugins are done in Maven vs Maven being said I do have a need for xdoc output and am working on do have a version that will produce either an xdoc output or HTML but not just did an upgrade to use Findbugs v which required a number of changes required in our plugin to are a few enhancements to allow us to create a plugin that can produce both xdoc and HTML am looking at letting you pick any Findbugs reporter output format along with am trying to get some time between work to finish this up this week and get some testing before committing it next this is not quick enough turn around I can give you a version that uses Findbugs and give the option of xdoc OR html output +for Ivans demandBTW thanks Ivan +Its probably not useful to add the offset to the shuffle trace but its available if you wanted to be consistentOut of curiosity how are you using the offsets in your performance analysis +When I wrote this extension in Seam it felt like this would more often be the case +startSlave vc changes of the new exceptions to reflect changes in other patches +This is something we can do in open source development across teams and organisations +Yes alias name for map value is meaningless +It is the job of logj in this case to tell you that it cannot load its configuration and from my experience it usually does +Ive checked Netbeans implementation and it seems there is a bug there +This issue isnt reproducible with clean installation and target +Not as urgent as bsh though +So I do not think CN should validate it at download time +Thank you Ivan +done! +Committed revision . +At some point in our code were copying the info over to the +Use a property to track clustered setting check defaultminimal or ask +Nesting borders are supported partially onlytwo levels of nesting +Did you mean the class or the Context interfaceThe interface should not change in the series for compatibility reasons +Moving to +The approach Ive attached will circumvent the issue by simply removing the use of the EMF list causing the problem +Added junit test and fixed the NPE by removing the from +Ugh your patch is against the old implementation of +Perhaps it could be made to check for the request header +The jobs are now registered and are running tests thanks for using! +srcjavaorgapachecommonslang +In trunk there no ms annotation on that test +Finally found the use case that turns my headAttachment contains a simple project for identifying it +If my approach to initCore is acceptable then its ready for committing +from EmmanuelBecause core does not have support for injecting custom constraints on generators today I find this feature useless and this has been confirmed by no real user asking for it for the last years +Lets say a user includes a to be used as a template for an editor for example +Absolutely +As not to double up on the exact same information within the clustering section I am removing the old data of the Quick Tutorial from the beginning of the clustering section and leaving the upadted material that is within the clustering section +never mind +This patch fixes the remaining memory leaks and avoids stability is applied in revision +Patch committed to fix this issue still need a test case be verified in gatein master and reopened if still valid +Committed trunk revision revision original unrolling here was to dodge a JVM bug possiblein all java versions from +That would be neat indeed because we could unify the split for both classes of evaluators +s attempt +So this issue needs a short revisit +The new patch is attached +No one else appears to be interested at this point in time. +Everything works fine I openclose files in editors modify files cleanbuild workspace many times Mb is enough for thatBy the attached stacktrace I assume that in your case exception happens soon after Eclipse start not after long work is it so +Any word on the final release of +Simply because we do much more than the standard +Hi Jesse please review and merge + +Not sure what CQL syntax for this is +I have patched to solve this +It is because of the check +I added a review Sergey +Can you please try with +Tested it out in Oracle with both varchar and number columns backing an enum and it works fine +Great! Thanks David! well apply that patch to our build and create a to be used for EAP. +I set it to for now can you please fix it? Thanks! +Definitively it is clear that the assumption that query params are the same from start of the request is wrong because the user can use jspinclude jspparam to add query params +attached a version of that fixes the problem for me +Moving these issues to as part of the release process. +Depending on severity we maymay not get it in +Mithun Thanks for pointing this out +weird +If you dont want two ways there is no need for the Startup in the first place your argument doesnt make much sense to me Im sorry +Not a regression but we can still do with a better error message +There is an class whichis used for classloading inside xsltc +I will see why and strongly suspect that an incorrect classloader is being used. +possibly a dupe of needs to be dont exclude a dependnecy if after you are including it again +I believe we updated out rexml as well +Sure just create a new issue and attach the dumps +Instead users should be calling Pathget to get the proper FS associated with that PathRegardless this issue is moot because the change was an inadvertent backward incompatible one which I think we all agree should be fixed +Probably not relevant but running on Windows bit +Patch is now uploaded let me know in case it need to be modifiedd +And no need for manual tests! +I wonder if the race condition is really possible with a configuration in theory the writing node owns a lock on the written guess its more tricky for async cacheloader its unlikely because of the of writes on the same key but still for Oracle you need a stored procedure to implement it properly but whats wrong with suggesting documenting that? Especially as many other databases do have a solution like UPSERT or dont like the Transactions approach as it could interfere with other transactions integration this doesnt look like the proper layer to manage transactions +Is this request still standingI have committed a change of the policy files to but neither the fix nor the documentation changes have been backported +If the process requires a response to be sent to a specific service then it can easily specify that service categoryname as part of the mapping created at process instance startup +When would the be selected? On all unknown typesTika can report the types it has registered to handle IIRC +Using protected attributes would work +Ok I got the point and I like your approach +up +Updated Installation BPEL Runtime section and added the info about the prerequisites +Here is what I have +We can combine this logic and something else for example to detect if a regionserver is deadWe used to have the owernership issue as Stack mentioned +since we the classloader now this appears to now work as to verify this fix but at Step I get the following error when accessing httplocalhostgrailsauthorcreate +Reopened after I posted the sample file that repros the bug +to trunk and +in svn is still not functional becuase its using +Dimitris can you provide a patch +Its looking good to me and if there is still a problem it digs up thats got to be a replication bug we want to find +Logged In YES useridDoh! Yeah Jetty uses this directory by default! Thanks Ifeel like an idiot now +For this issue depends on is no plan to release +I just committed this +Would be great to use mockito to add test coverage +Thx again Ulrich! +Besides would you please to give some tips to survive in application with ASF? For example the criteria to evaluate a student +with JBT. +There are just a few details that I need to fix +Well it is certainly one for us to check out but it is unusual that it works on windows and not linux. +If the classpath contains a directory and that directory has a database as a subdirectory then the database can be accessed via the classpath subprotocol +Will commit it once Jenkins gives its blessings +Oops somehow assigned this to +Done. +in +I also think Its a test related issue not HQ +continous build s commit this almost as is with the single change that we handle inclusionexclusion of tests only through or test framework +An improved patch to address performance concerns on reattaching the resize event +The problem was never in any released version of Cache not even a beta + +Thanks +Calculate merkle trees anti compacting each sstable into data being repaired and data not being repaired tmp sstables during the process +Here is the test case +Not an issue for trunk +Sure Yonik we shall separate the changes in core classes into separate issues +So imho the best option here is to be defensive and check and make a and log a warning or error at least it should not fail here +Added as a Resource Change listener just like class and performed same close editor steps +What you want to override is the method which derives the methodCall +This means by default Roo will perform a scan every ms but if the last execution time took over ms it will wait that duration until the next scan +Rajesh since this cannot be included into EAP CP because of the docs freeze +It looks good +This was solved by. +It is now +no problem to have this file generated old way it was just an restoredPlease open a or new issue if you have more things that need to be changed to improve usability ease of access +Thanks Philipp +I think the example setlocal should use to ensure the passing char is UTF encoded string +Could you help me with that +Updated patch for portion of the patch is actually for trunk only +lets hope future releases really the mess +BikasThanks for creating separate JIRAI would like to work on this and will upload the patch soon +Thanks for spotting this +Because i always notice that invocation count also counts his own time so if a method is hit often then you will see thatBut you are right localization seems to be always a bit of a problem +Simon at least with Nexus its easy to rebuild metadata +Here is a patchPersonally I find this feature quite useful especially when dealing with issues relating to lost trackers +Thanks Brian you should have highlighted that the patch includes tests I would have taken a look at it earlier +Please provide steps how to verify this Jira +I can submit a patch for that too if you like +Looks fixed. +Session ownership logic is obsolete so this error will no longer happen. +Therefore edit in can be set to null after these calls +in +if this parameter set to accept this plugin allow only entries which must be language codes and remove outlinks to pages in other languages +Hi Joerg thank you for your suggestion +The default generated are now HTML compliant +Need updated to get new secondary licenses visible to user for acceptance +Rick! Ive committed the patch with revision I plan to post a follow up patch which adds more details to the comments about strong password substitution +directly. +output of nodetool repair on one of the nodeoutput of nodetool Jonathan I read through this ticket and trying to understand the context + +In iocorenet line sreadreadylist +As for the missing localeswe should not put too much effort in there now because we might need to remove jscalendar shipped with WW again due to LGPL issues +It does not fail but Im not sure whether it even covers the issue +Sorry for the delay Owen +Thisis a different issue +I can make a patch for the code but cant test it on my system +anyone up for pushing this through the opensocial spec process +So Im not opening another JIRA ticket +I found this problem when I tried to compile our complex application with +Since reading from a socket isnt idempotent what I called redundant is actually harmful +selecting the ledger to recover should lock the znode so that other worker do not try to use it +But can you help w a testcasebq +Its got a patch +for the doc +typoCannot determine how to convert the bytearray to int +For case after the snapshot yes you will only see the +Just FYI wsconsume and perhaps wsprovide scripts are also affected by that +Patched and trunk +Thanks Bill for the s my idea on your commentFirstThis is very similar thing what i have done and only difference is theres no memory allocation in this if its better to avoid another memory allocation i will do it in that waySecond This is not much clear for you can explain some more thatll be greatThird Forth Fifth for all three ideas +Maybe someone more familiar with the code can comment on this +You may take a look also at bug It reports basically the same problem but with a comment instead of CDATAThanks +The patch also removes the old test +and laterTherefore we will have to go back to Mojarra which seems to work fine +We dont currently have a plan for as we are still working towards +I send email to apache lists from machines in my domain +CS is on Central and working Verified JBDS CR L + +Theres no way for me to add a listener to either the combo box the editor or the section +Actually I think the patch didnt apply cleanly on your side because its supposed to rename to VertexAnyway heres an updated one just in case +Awaiting fix to be pushed. +Another option is You could configure in your root pom to set use this way you dont have to enter this param from commandline +Can you run the query directly against the repo and see what results you get? I didnt add any functional tests to the patch just a dummy one to make sure that joins over davex dont break anymore +I was the original commenter on the blog about this issue and have previously experienced the deletion of all files on a drive because of the exact same restriction the fallout from this is massiveThe issue here is that many people who use lucene will not realise that this can happen and this situation will occur sooner or later +the cron jobs are controlled by cfengine +Put in scratchpad with minor changes and refactoring of the exampleTo test it easily do build installscratchpadwarSamples are under httplocalhostcocoonsamples Please cross checkThank you for the contributionNext one? +As modalPanel is always rendered closed by default it will be hidden automatically when upload will be finished and navigation occurs +I have attached the trunk patch with few test cases +Thats all Im saying +Sorry the new one extends that but ignores it so we have to add it into both +Perhaps not in the first release but eventually +User guide references to the job executor servlet were moved to the servlet context also updated a section of the asynchronous continuations chapter that explained how the job executor worked in terms of old stuff such as the context builder the message instance and the job interceptor +bq +OK. +It is ok after clean and you redeploydb +This is fine as this change is more for scalability than performance. +Obviously the included page is not able to have another resourcebundle loaded with the same name +updated workaround from ticket the fact that there is a workaround mean that this will never be fixed? The issue has been there since possibly earlier and still exists in current fact that in the web admin it already knows the value of the webroot path surely this is a relatively easy anyone likely to pick this issue up as its a deployment nightmare +Ive just committed this +If it isnt a problem rethrow the +I also included two other types +This does resolve the issue. +The patch makes use a helper method from to delete the database directory +Edit release notes for publication. +All tests passed for me +Maybe you should try tried Christians suggestion about loading XSD and from the classpath +These tests are under IP clearance and arent part of the repo yet +committed +Heres a patch against current trunk +TLDR its harmless +Hello LaurentDo you know when this exception is raised web request on project dashboard coverage cloudsthanks FreddyIts in rules breakdown page +I believe this can be fixed without changes to the base tag or adding another Globals constant. +this patch add some debug messgaes +Im not positive the consequenceneed of this +Changing the priority from Critical to Medium +patch reworked +Im against just dumping an impementation from elsewhere that duplicates much of the code thats already in IO +see also for to trunk because this need many core changes and should be delayed to a non backward compatible release. +It would be really helpful to add a log line where we break the s loop on detecting the killed conditionCan you please make these changes and make the patch run through Hudson +IT is passing for this now. +I think wed have to use a weighted cache with the weight being an increasing sequence number so it removes entries based on cache size but in insertion order rather than LRU +HiI think they should +Verified fixed in BRMS ER build +It will probably come in after +Which section of the specs claims that +is throwing a ensureexception +The solution to this is to remove the seam dependencies in +Assigning to you for now feel free to dupe this if you work on it in the reopened Jira +Is there an update of this plugin available for the current trunk? Or is this kind of functionality implemented elsewhereThanks comments related to in this issues comments section +Now I rememeber this was the root cause. +Hi Shyam thank you for your interest to FVHThis issue is a duplicate of and its been already fixed +Ahmed +The proper jars are included in the lib directory so I have attached a patch with the updated references +Not really a problem but I added an small fix for planetstruts +However I do agree that it will be low priority given the simplicity of the workaround +With the latest patchit allowccccdbut it does not allowaabcIf that is needed by PIG then we need to checked in +I have overridden the method and included space in the acceptable parameter pattern +Welcome to the Fellowship by the way please find the first draft of the SSL guide attached above +On order for the compaction to make progress we could assign scanners to epochs where everytime we change the readers for a store we go to a new epoch +It is for building openejb from the top most directory and removes ve looked at the and most of it looks good +This issue seems to be verified but not solved +Anyway if you have a patch please consider putting it up so we can take a look at it +Hmm not pretty +When things in the message passing plumbing settle down I will come back to this +Thanks a lotBest regards by Richard. +Thanks Jason +Attached jar might help +Another side effect was that the signature were not generatedThere was such a thing also in s pom before the module was created but it seems that it was never usedDo you think that we should remove the assembly from the maven repository +Committed fix to the trunk +Oops! Good catch +As the referenced doucment is no longer available it is impossible to verify this problemI suggest to close this issue +I dont see how cleaning up the variable names and other inconsistencies depends on flattening the layout +Hi SimonI agree with you all Java project should compile with Java so the problem is not in +GailIll get back to you on this next week +Just wanted to verify that we are seeing this same exception under under JDK on Linux +Using mocks to simulate component type without multiple files +A fix for this would be to place the registration of the archives before the addition of the nature +So Id like to extend with a find method that would allow the to be more efficient and put the current s implementation into +copyrightlicense notices removed +The patch is in as of revision and I believe it should fix the issue +i change in cronjob the line now rsync is using now should be work +Your a force +If Affects Version and Component are not set the issue will be rejected +In the meantime was released fixing and several other bugs +Should be pretty easy to copy it over and update as needed +It doesnt implement readFields or write and just inherits the ones from which will ignore the in theIn my experience using inheritance for implementation is extremely brittle +brett workflow seems to be broken on this issue I cant close it can you nuke itthanks! +This is not an issue that would have any measurable effect on performance +No problem if you gather these together over a few weeks andsubmit them together +Do you want me to create a textual version of the CPL as +Can we avoid computing checksum on the fly? I linked this issue to +HiPls find a patch attached to this issue and the related explanation doc +Scott if you got a bit of time then a new patch would be nice +Blacklists and wildcards can be implemented as a custom so we should be able to leave that up to users now +The BOM has been updated so it should be fixed nowThe jar is present in the repository under the name orgjbossas on EAP ER +As per Lars suggestion may be wont be giving a client class also +Hithis issue will be fixed when the patch that is attached to is applied +Bulk closing for +This patch fixes the bug +My understanding is that the query engine should parse the query and this would in turn translate in a pair of fulltext conditionsCurrently the Lucene index does implement its own parser +bq +I would say that it is good to fix this problem for those that are running in a container that has a enabled and desire not to grant additional permissions to a webapp +Thanks Hugo for pointing this out Edison can you please check this as well and cherry pick on as neededWe need it because this change should be done as part of db upgrade +Never mind +Resolving +I want to fix this for so we dont have to maintain deprecated implementations +Resubmitting the patch +Closed upon release of Hadoop. +I was trying with version +sql files and to erase and in the +more tests more polishing +getting the switch from DOM to DOC state rightAn issue that is RESOLVED FIXED without patch can never become VERFIED otherwise it is duplicate which does not fit to the titleWhat do you think? and pushed the revisionThanks as integrated to masterchangeset cdbduser DeveloperDevinWei HuaReviewerWei Huasvanteschubertdate Fri May summary bug Fix method when rowlabel and columnlabel are null +DOM Level should address this problem +Agree +patch +Ashutosh deals with restricting location only to external tables +Attaching patch forUpdated ivy settingsproperties fixed compilation issuesSome did change and I adjusted code accordinglyThis builds and packages fine for usWill be doing additional testing on our cluster +I was wondering whether you could please try this again from Roo master or alternately provide a small code fragment I could use to reproduce this report? I have read through the code and dont see any immediate reason for the NPE but the line numbers have also changed since the original report and so too has many of the module internals +Ill push the postDefine hook upstream for now and include it in the CR release +How can I attach my testcase? +bug can be reproduced using testXSLT command with the DOM flag specified +That would be helpful i think than theI am not sure if this feature is already present in NutchAny suggestions are appreciatedThank youKiran +Updating +So we still have a ambiguity that needs to be resolved +Holding off on reattaching that patch until tests pass with the trunk patch +implements Chriss suggestion and removes Date Author Id Revision from most in +Applying that to the branch causes the tests to pass +Fixed +I think it also occurs only with the Main page. +Exactly same scenario as described in this bug for property am attaching an updated testcase that covers the reported issue for both property access and method edited the generated code example because the page width was totally the xml shows that the implicitThis is set correctly +I fixed the code that the original stacktrace referenced but it looks like the problem is much more widespread +foggy +Thanks Dan for your time in looking at this +I dont think they are useful but if so we can improve the code as suiggested. +so no further workso it can be closed +This was in IE btw +Fell free to just go ahead and commit really we should have made sure this went to trunk at the same time. +Updated the description to reflect the changes in the implementation after the integration with patch to incorporate the suggested right I should be using HMAC but I am finding it very slow ms compared to ns for a SHA hash +Reind reports can you open a new issue for this otherwise it screws up tracking the versions +I think should go in before making this change here +And I will try to update my version of works as expected just docs were improved +Please go ahead and create itthanks far as I can tell this has all been done +Also its not such a big thing that ORM would add additional boot options not worth it +MISRA +There is now an in which handles this for us +Comment race +Perharps should I clarify this thing +We have a cluster with m and the write performance is improved a lot with m the cell is big and most is K +I created the file but dont know the means by which I create such a signature +Managed to get back to thisRebased onFixed resource leak issue that apurtell identifiedRuns through on my workstation completely with default ulimit of no crashesIm going to try and benchmark this on a real cluster this weekend and report results +thanks viola +Creating the session in turn makes Jetty add the expires header thats causing troubles in on and trunk +Hi Nigel would you please elaborate a little on your second bullet? The Pig coding standard requires using spaces for all tabs in java code which is what I have been following in the code +Im not sure why +Only fixed in for the time being +Oh no I am not that extreme I will update the description to make sure it reflect the approach correctly +Should we admend the message to include Check the master is running? Otherwise I dont think the message bad +is hard coded in the script +What encoding was used that led to NPESorry i dont have the original accident log right now it was fixed in Aug +Heh +I would imagine almost all clients just ignore the MIME type here anyway as they are definitely expecting XML or JSON but I do like doing it by the book +If you have ABC then upgrading A will force B or C to lead +committed to both branch and trunk. +If provided a row with a page size of the current row is set to so that the row appears on the page and the page size is maintained but no rows are skipped +Thank you Doug that would be brilliant +Here they user err +Itd be better if you use a static boolean has instead of having to do Stringcontains calls in the finish method +NicolasWithout news from you soon I will close this Problem NicolasThanks for your answer +Theres a link to your suggestion here. +Agreed resolving as duplicate. +Of course I could be reading too much into this +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +CP fixed in revision fixed in revision +thanks to james for his help +We want to allow a to come into play here too so looks like we want the security filter generation to return a Query not a Filterbqbq +That patch is way to complicated for finding duplicates +Im targeting it to but that can change depending on when is released +In the near future wewill be moving to a newer version of maven +I will attach it laterHere are the result codes for an operation as described in the issue for different serversSun Java System Directory Server LDAP error code Object Class ViolationIBM Tivoli Directory Server LDAP error code Object Class Violation LDAP error code displayname multiple values providedPlease note that for Tivoli no error occurs with displayName because it is not single valued in the schema of the server +But then dont we have the issues of the tests polluting the other tools or am I missing something? When TOOLPATH is added to the classpath as is done when running tools like archive and distcp the tool could potentially encounter the same kinds of issues with tests in the classpath that were trying to avoid +Attaching a patch for this +Is this stuff that could be done with an +Yeah zk should be persistent maybe we have all the znode as ephemeral because we dont want hbase depend on zookeeper state +Ive just checked +That seems painful +Thank you very much Ted +Lets keep it the user decides to uncheck the file then I guess we should error or at least warn the user that he still needs it for whatever goes into the file +Thank you Rishi +Committed with as tolerance +I can take a look at this one +If you think they are critical for CR please them +Fixed offending classes +Ive moved the stack to using version of including the fix +This has a chance of making things work better while reactivating the old build config just perpetuates disfunction +CarlBrokenness in building grammar was introduced by and we are discussing to improve the situation on Though it seems like you are getting different errors +Patch attached +Applying review feedback I chose to remove misbehaving audit loggers on the first exception instead of keeping track of how many exceptions or the rate of exceptions being thrown +Add to mvn command line +On ant this is handled differently and therefore does not occurRelated to fix in patchRemaining testingrun build with ant to confirm issue +data tmpdummy +Its unfortunately not enough to check for illegal characters +Moving out minor issue +So please use the Pull request method making sure you have rebased to latest code +Patch to update Hibernate source code in order to allow nested instantiation using select new +If it is not the order you prefer it please let me know and I will gladly +I just reverted the change. +Added DSYRK with offsets in . +Pavel can you see if it reproduces for you? +lolSorry thought it was obvious move a lot of the background on SOA etc +I believe we should make the request of the Commons Validator team first +FYI we are currently refactoring the overlay mechansim +Excellent thanks. +It would make things much clearer in the more crowded forms +Closed after release. +But if unchecked the application name is disabled +So every none submitted value will not be checked if it is required +if thats the case this will have to move to be an blocker +We can wait for check in and close the ticket +Includes comments and javadoc fixes on Interface no code changesIncludes fixup in so keeps going if available returns negative numbers +The other one was too big and unrelated +The commands to startstop httpd are service httpd startstop +However the existing files use instead of the Ant plugin +I dont think this is related to +It just a twitterj library setup issue. +I can see the intention but Cell is a small part of the overall client API +We were discussing this internally and thered be interest for us to have working with +Sorry didnt mean to close this to a feature as affects proposed over email I would like that that the changes made in the from the j module under be adapted to the JCA inflow messaging service +I thought Id reverted all the changes Id made to the core library +This is fixed simply by overriding equals and hashcode methods in Ill run some tests to check it doesnt break anything then commit a fix +We should run the tests again on such hardware and seems reasonable to turn off debugging for the comparison +Just attempted to reproduce using master of AS and no issue was present jconsole was working as expected. +Duplicate of resolved issues for already released versions +Thanks Amareshwari! + +No exception caused by but there is still problem with cached fragment +As a result the code in class is now much cleaner and more maintainablePlease review and let me know what you think forgot +Resolving. +here is a patch that adds a to replace those two booleans +Here is a copy of the pom I am using +If it was not already added should be backported +This is now done +But maybe there is a need to ensure the thread locals is cleared when the exchange is done +RCsvn vs +I have to sync the latest code check but youre most likely rightLet me check if it still applies resubmit a new patch +This bug didnt fixed in just released +I hope the tone of my request is more clear now +on getting async IODo folks believe that this will improve IO latency? Can be used to demonstrate performance improvement +isnt this done +This is one of the things that changed for the final release in +plan to add the when the basic pluming is all in place and can resolve this issue +All subtasks are resolved please reopen if additional tag lib issues are found +I would have done it myself but Im not a PMC member. +Attachment has been added with description Example drawing to reproduce the the resolution given to this is being marked as duplicate of +reopening to mark it is fixed for as fixed for RC +Ill open a new Jira for this with a test my issues related toTx. +This patch is cleaner +Attached for JSF +If your use case does not require it then you may choose not to implement it +More work in r subtyping and LUB computation implementationFixes +We get errors where all dependencies are found but the checksums are incorrect +The app would have to ensure this in order to use a collector that takes advantage of it +This could is looking for one +Thanks for pointing out the fix +Also not sure if I get it +Ill have a look +Its ibiblio policy not to change already deployed artifacts but Ill add sources and javadocs for the next releases. +I like better although it adds one more looks good! Any chance you can resubmit it in git format so it is easier for me to apply +How can it be LGPLI actually asked Sun to create a project on month ago where I will place the Official Sun servlet which is under a BSD am still waiting +Thanks! +But at least it worked +This explains why no block deletion was observed at all +productsitePlus these jobs +Changing flattened attribute to a different path results in the old path stuck in the dropdown and the new selection being reset to empty string if a dropdown is opened for a given patch +A test has been added to the to validate the serial between jboss and the jee reference implementation binary values. +We need to look at how the xsl transformer factory is being located because there should be an ability to override the xslt factory even in +It feels like a lower level issue +Does this still happen with version +The message store configuration was already updated as a consequence of changes made to the branch the only outstanding issue appears to be am separating the issue out from the jUDDI issue +Marking closed. +forgot to remove public keyword +First i will upload patch to revert +I think we should let the user have the control of the setting I dont think we need to filter the in POJO CXFMESSAGE PAYLOAD by default if the user doesnt want to use the default one or copy from the source he can override it by setting the header in his route. +We have to add better validation to the group validator page so that you cannot create an empty one +is now reverted and the patch can be compiled +Attaching a patch with initial implementation to REST service +Add DHCP ServerResultUnable to execute API command addexternaldhcp due to missing parameter podid +Thanks Stepan +In release release +NB It seems to only fail when passed through the proxy +Quickstart project demonstrating the behavior fixing the issue for for the patch Maxim! +Confirmed fixed +Patch against am interested in configuring the cookie name +i remember fixing something like this recently +On the session request that fails is never called +If we change the public api signaturethat will be compatibility changeSo to avoid this can we provide overload apis for this change? which will not take default value +It seems like the is called and not just y as we are used to the of were using as could take a look at the the sourcecode to see if theres something we have to make contact with the developers of yacc +I see literal sequences here instead of newlines +Correctness should always come first IMO +Vadim youre keeping us busy +Hi Rustemdo you have some tests showing performance improvementsThanks +I am using as our STS +for the green oneDirk could you send me the source files or can you attach it to the issues +I had proposed final draft of spec where this chapter was completely missing +Contrary to the comments on the mailing list thread this isnt an openejb problem +Its fixed now however it would be nice to see if there is any real mail server that doesnt implement something like RECENT and other flags +Even with the previous versions of surefire it was possible have mixed lines when using any kind of parallel test +inconsistency in these priorities is my work +When a replica is detected to be corrupt it is deleted +As you point out the client probably have done that already +That can happen now if the mutations are not on the same rowThat is true but Ive never seen that happen in reality +Your explanation sounds reasonable and the patch looks safe to me +This is required in future if someone uses same utility and wants to handle its own events +This issue is to allow a special patten like Join means any number of connected to Join +Thanks for pointing that out Dan +Thanks as always for the superb documentation + +I tested problem in my local instance of JIRA and user browser works correctly +Is this still planned +Test code had been kept out of width limit traditionally +fixed bug in be addressed as part of. +And there I thought Id know that codeOK so these need to be top of row +They may have some good ideas that we can reuse for these types of things +And as for trade secrets as long as you keep them secret we wont pry +Alright thanks will look into this again when I get into the office tomorrow +bulk close of resolved issues. +documentation updated +Tims provided upload scripts + released +to v for commit to trunk +vmdk is consolidated into base vmdk +bulk defer of features from to +I remember thinking long and hard about this but forgot why I picked this model it still seems slightly more powerful because the newline acts as a with a higher priority than the but I am not opposed to switching the priority +BTW The patch uses avoidtrue as default as this is the safer approach +I will change the dist postGoal and remove the code creating the zip with the date in the name +Looping and adding a timeout to the test revealed that blocks iterator doesnt support remove +for caching on application level in deployment also reworked the logic of for retrieving last modified timestamps it should be noticeably faster r it should be the same in DEV and DEPLOY modes. +Such an application will also load the ejb home and remote interfaces from the same classloader and these interfaces will be equally permanent even if the ejb is only used once during startup +I havent taken a look at your reimplementation yet but Id really like to see the conversational attribute and the immediate redirects from actions. +I did some more debugging and the conflicting jar is the usrlibhadooplibAny idea how to work around this +They shouldnt know about the business rules +Please be sure to do an svn up +The merit goes to Simone Giannecchini that provided the actual patch thanks a lot from my side as well. +An integration test using projecthelp to see if the central repository has been overridden +Looks good to me +No followup with recent or so marking as incomplete +I agree that only using request parameter when the is read later makes a difference +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Since we do not have now shall we put it on command line +That could take some discussionfwiw there is no Software Grant on file right now +When a polygon with a single shell provides vertices in counter clockwise order the polygon failed to load +The all match at the end +Roo only creates an Version field if it detects from the database metadata a column names version +Looks good! I will give it a try a bit later today +Just checking in on this issue +Can attach this patch or at least some code sample? This will make our work easier +Danushka I hope you are a comitter +Please you report the message interpolation bug clearly please I fixed that I got the conditional logic test +Usually we dont remove tests unless your new test cover all the old tests +This lazy evaluation gains you nothing but makes your code significantly more complex +The patches are a fair bit different in what they accomplish +Not necessarily if you reuse them +From my perspective you have a degenerate case since you have a requester with its own private copy of the interface it is querying +Thanks a lot for the contribution Colin. +Hence we had to build our own mechanism for post processing +Thanks for Todd +Added fix version. +Thanks +Patch with unit tests included +ok checked +Verified this in the latest build and the issue is not observed hence closing the issue. +No not at the moment since you cannot replace bundle content with a fragment just ammend +Ping! Can a committer please take a look +To prove it add a trace statement to +Refactored the unittests going back to junit +investigating +Hi Commiters my patch is ready +The correct time is actually in the raw data stringActually this is the typically case of shit in shit out +Abhijit think When a new Employee is created from Party Manager using seca we should create employment +please remember to attach the patch to JIRA in addition to posting to Review Board +The manual contains a workaround but hopefully we can get a proper fix soon enough. +Hi Kim The documentation looks really great! A few minor commentssuggestionsAdmin guide Replicating databases Suggestion In Derby you start replicating databases by +Ill give it a look +Ah yes I see +Bozhidar I agree with your over to you +I am currently doing several changes to component and will take care for a fix hereKeep you posted hereRegardsLars +That was supposed to have been subsumed by xlib +This patch lacks unit test code just because trunk uses some new features that are not available in Cloudera distribution release +should be the last version in the series due to the binary compatibility problems +We have not been able to replicate this issue since the first discovery once we do we will send all relevant files +s to put on +So I changed that to be no IOYou might wonder why you get into this trouble +Thanks Uma for review and commit +fixed in r +If we do anything to customize the presentation it should be internationalized +Could the problem be that commit is not done +Solution delete existing shortcuts with the same target before creating a new shortcut +This is fixed as of r and r by an update to JNA. +this resulted in compilation errors +please move back if they will be worked on asap for +Can you post your modified patch or an incremental patch on top of the latest one +My guess is that this didnt matter in testing so far because it is rare for an async socket to connect instantlyThis addresses Camilles commentsI have added a few javadoc fixes and one weakening of a catch from Exception to Throwable in the general spirit of making things better when I see them +Added a test case +and look fine +Now using copy by value +and ran cleanly +I think Lars is referring to by Prakash +I am not sure if the EAP testsuite is dealing with large messages but I dont expect so +After implementing the workaround using object array I was still getting the same exception +By the way op team reported the cpu usage got a little higheri think thats because of maxoverage is too small +Also the cost of tuple reconstruction accounts for a large proportion of the whole execution time +GaryYou were right the problem was the static variables +At least there should be a very obvious warning telling me that the process contains unsupported elements +Is the write cache on the SATA drives turned off? If no durability of the file system is not assuredK flush operations on SATA disk without NVRAM is amazing for me +Small patch so that if the Throwable is an Error throw it as an Error +Missing a bunch of unit tests and not all unit tests are passing yetDefines interface and uses it from the spooling source +Patched applied to Ode branchboisvertboogsvnode svn commit m Response and WSDL are invalid Sending axissrcmainjavaorgapacheodeaxisserviceSending axissrcmainwsdlTransmitting file dataCommitted revision +and +After patching will also need to propset ignore on compilercppsrcwindows +If the new patch is uploaded I will commit this change to trunk branch and as well +But with the latest fixes all three platforms behaves well with this branchId suggest moving back the feature to land it asap and get more testing done there +I just started working with +Sounds good + +Ive added those ids to the file in svn +Brings bug count from For what its worth about of the remaining are out of our control through ANTLR or Thrift +I think we should close this issue as WONT FIX +Thanks for the help Sharad and Amareshwari. +The same functionality is available in most if not all web frameworks +Due to this uploaded all contributions again new SHA hashes +Closing as not a Commons issue +As a result only message makes it to the +The fix was therefore backed out in revision but the tests are still part of the regression test suite +is missing from the list +Can you attach a sample so that we may try to reproduce +committedOleg +For the time being the metadata structure only contains the field name and is an inner class just like Isset but unlike Isset this one will be the same for all classes so may be it should be placed somewhere else +Kevin wouldnt this be fixed already with the new gwc url mangler that delegates to s +Right? Added fragmentation details to UIAdded a total fragmentation precentage to in top table for example Regions On FS Number of regions on +What does your cpio say +i dont think its fixed in +am i not allowed to have those tags? it worked in to define header stuff in the template which was then copied over to the final place +If Geoserver can access the data on both the Postgis and Oracle database and serve them as WMS layers mapbuilder can display them +is updated because of recent commits in the trunk +Christian Making connect and read timeout configurable should address this issue right +closing since is now released +I will take care of the above but I was not sure of the charset names +Actually i guess we dont need a full blown proposal just need to ask permission for a new unsupported module +Merged to the branch in revision +Propose up on dev list Id say Jon +That is why no header is sent at all with responseOne way to overcome this issue is by calling resetBuffer instead of reset +Added default editors for Boolean and Number objects basically using special incarnations of and +Heres my patch for this issue +Good point on breaking existing usage if we drop the syncs + can also be used to implement meaningful equalshashCode for is a slight bug in the original patch where method parameter information was not added to the id I have attached an updated patch that also allows you to build ids for members directly without having to wrap them in annotated have run the in container tests and it was all good +It appears that the SNMP module build script uses a copy of libtool that is bundled with the JBN EM distribution at +Closing the issue +committed +I will have the permission to commit on the SMX svn repository soon +OK thanks it was tough for me to tell by just visual inspection. +if not +I tried r and still see issues +Perhaps someone else can figure out how to fix this without adding methods to Navigator. +also for tccell +This in turn of coarsemeans that if Tomcat and depend on a library artifact but dependon different versions mightdecide that its OK to use version of the artifact forboth and Tomcatwhen the reality is that Tomcat was built with version and +Consolidated in +Closing the issue because the consensus is that this behaviour is corrent. +i attached file +The generated code does not encode those special characters correctly Hence serialized message has wrong values +Evan Looks like the actual spec didnt make it into that tar only the PublicPrivate classes did +The change looks good +Thanks Aleksey +Use a if you need some strange functionality +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Julien Andrzej you have my to take your axe to the old one +is defined since jsse is provider based and that the chosen parameter may not be relevant for other providers than the one included in the JDK +Hibernate needs to be JDK compatible +is executing before the scheduled Activation which asserts the new fact is executed +Here is a Test case to reproduce itI launch it on a and it deadlocks each to run +Otherwise it would have been mutable and therefore I would have preferred to just return the prefix lengthI added a new version of the patch +This is a patch against the subversion sources as of applied and tested +Im going to implement this feature together with independent of to be able to get it into release + +Eclipseeclipse has been failing today in several other jobs +No activity for a while +That could be made a little clearer +Ill update this thread once I have more info +Here goes the patches for issue including ones for and +Thanks +Thanks for the patch! For future reference bugspatchesetc for proton should be filed under the PROTON Jira instance +Yes please! This shouldve been a part of the Solr schema issue but i overlooked it +so this really needs to be listed as a backwards break since it will directly affect usersI will note this in the CHANGES once we are patch with davids with trunkI added a to contrib and fixed the issues david raised +if check always has a null first parameter should we just take it out +You up for adding this test here? If not I can give it a shot +This bug was caused by an unrelated NPE thrown when pressing the apply button. + ran the reporducible test case with couple different releases of Derby and noticed followingon trunk if I quit out of the ij session which does the setup and runs the query with incorrect results and then start a new session and run the query again it returns correct result +Top level further looking into manager for the actual app just wont deploy +Here are some tests +What front end? or Simple +Do we implement a or try to leverage one that already exists +Some time this month someone fixed what was causing the exception + looks to branch thanks Eli +I will take a look +Patch provided by Eric Andersen +Any update on thisAndrew could you provide us a patch to fix the issueI have no websphere and jboss setup arround to address this one thoroughlythanks still waiting for the patch +Voor urgente zaken kunt u zich richten tot het secretariaat van GX. +Edited issue summary and description changed Maestro to Continuum instead +All plans will have to change regardless to update version numbers and will have to be adjusted to reflect the changes in and groupId as they move to Apache +Is that intentional +Here is the list ESB Tools BPEL Editor Maven CDI Configurator Maven Hibernate Configurator Maven JPA Configurator Maven Project Examples Runtime Drools Detector Runtime ESB Detector Runtime ESB DetectorThe Tools for Savara +I will investigate later +The XML attribute to set the debug is now either debugGraphics or debug no longer debug or debug +Arvin working on value numbering pass which will replace polyglot constant propfolder +These tests are now working properly tests are now working properly tests work for both platforms +would you please revert that change +As I see it huge number of persistent messages not delivered could be the issue +Hadoop was released. +I fixed the issue by changing the rule in both grammars to require at least one character for a match to succeed +Thanks Daryn +names iterates over all attributes and you modify this list ofattributes by doing the final setAttribute +I just confirmed the Server Activity page is working with this change +Fixed in trunk +So it will go in instead +bq +Thanks much for the suggestions Jarek. +I dont see this issue as providing a complete AntMaven conversion system +I hope that would download the latest snapshot for meOk if you didnt want to add the local m repo as the last resort when nothing defined +snjezana can you take a look +Any chance this will be scheduled sometime soon? +no longer reproducible with this patch +Copy over to +I dont think we have to implement something +the patch to include the border problem in Linux for the Env Variable in later patches +Attached a quickstart +Im setting aside day a week to work on so Ill schedule this ticket for investigation next FridayCheers +If you want behavior then dont set timestamps and will use behavior +we lost one more which is +Maybe the recovery is taking time +Committed to trunk as revision . +I understand this might be confusing but there are other feature requests open for improved skipping of phases that would accommodate this +Having to insert tags by hand and not being able to find descriptions for tags easily made it hard for me to like Forrest +This patch fixes the c vint serializationIn addition it fixes issues in generated C codeGetters and setters for map were not correctly generated +Im gonna try +This is especially true for pig jobs to monitor its +committed +A more aggressive refactoring separated the cache config from the translator functionality +I believe I traced the issue to authenticated users +Reopening to remove other update in revision . +Specifically we want to include +Give me a few hours to think and talk to guys +Any comment? If no objection I will commit it after testing some cases on Geronimo sidethanks +Hi guyswhat is the status of thischeersMuki +I missed your update for some reason sorry about that just saw it when I was searching for something elseThe release vote for started now on the dev list if it goes through I will commit your patch next week +Closing all tickets +Applied to branch +bq +Closing as invalid. +Correct repository picture has been attached +Fixed both on and closing all issues that have been in resolved state for months or more without any feedback or update +I saw that an insert statement was executed not an delete +Catalina this looks like like it resolved my testcase for so can we get this back ported to the branchIll create a patch and attach it as I need to apply this to my local code and rerun the failing EJB TCK testcases for Geronimo to verify that this did fix our problems +Sorry I am gettingacademic +Assign for test I forgot to close +seems still present +The current situation is rather problematic keeping opened connection per workspace doesnt scale well at all +We already have a and supported string serialization for schemas +This issue should be fixed with the patch attached to +Resolving the bug as fix not required as support is already in place for requested functionality +So this bug is about FULLXA mode +Thanks Arvind! +Cant actually add a user with incorrect credentials +of screens in were updated according to the new view of VPE +Had an offline review with Namit +Linking This issue will be very important users trying to determine what application changes are required after that fix +I dont think ordering is the correct solution How about first installing all bundles and in a second pass starting all bundles +The second EAMRESOURCE is used to for the permission look up +Im starting by clicking on the icons as a normal user +Store config templates in PREFIXsharehadooptemplates and change related script to use the new location +Cant test as I cant build Axis under maven at the momentIf this is included in the nightly build I can run it through a websphereMQ testCurrent functionality remains unchanged there is no logical or API modificationThe new functionality is only activated if the new parameters are present in the JNDI tree +Can you verify that this issue is still a problem +Bulk closing of issues that are old and have been released or marked invalid. +Done. +Once the Urgency field appears we can adjust the severity of this bug +Thanks Dag! I changed the wording to this in revision return a node representing a Boolean constant if the result of the operator is known otherwise this operator node +I will tomorrow +Sorry for the long silence +Sorry for the delay one of the interop servers was down +Also I am curious why the BAC does not throw an exception if the configuration passed into the BAC is not found +This opens the door for SQL injection attacks +The testing was successful swt suite did not hang for me for runsSo I suggest committing these script and patched to eutextra and using patched EUT for testing +Lars not yet needs some more testing and validation on a real cluster +patch for avoid memory leak +Sorry forgot to mention this +Yeah we had a babysitter on our cluster +At least in my mind +It seems the comment Property values have been moved +I am now going through my backlog been quite a while +Committing +Confirmed +I agree +v is about trying surefire with the latest trunk + enabled to call arbitrary delegates enabled to call custom collection processors may be used with an expression a comparing lambda function an arbitrary callback delegate and an improved to allow for ascdesc sorting and removed restriction to input +Seems OK with. +I just submitted a patch on YARN Jira to kick Jenkins will let you know if all is well +Ive fixed the corruption issueThe idea of being able to run in parallel is a whole different feature I will open a dedicated jira for it +Better to use the workaround or file a bug with the appassembler plugin if still needing to be addressed. +It works allright with previous versions ofI will go ahead with the separate changes. +Asked if its still an issue with newer versions of CXF and no response +Adrian is this in a scenario where you have configured the framework to not rollback? That can cause some exception swalling since the framework assumes that if it got to a certain point there wasnt an existing exception +moving to Alpha since we ran out of time for to since we ran out of time in in Developer Studio +This issue was really fixed for and documented in CHANGES under that release +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian in latest release of XALAN +We have row level atomicity +These tests pass on opencmis inmemory so it seems that this is a problem with Alfresco specifically +If youmodify castorbug to turn off validation you will see that the errordisappears +The overall point is how does this feature request fit into an overall supportable management contract for this subsystem +Bulk move issues to and re right that the patch is huge and would be better broken up in smaller steps +The content negotiation algorithm needs to select a concrete type before interrogating all message converter +I just pushed this to. +Even if we gave a constructor with the URL in it this would still be impossible to test for several reasons +DevarajIf its just for tasks of the same job then yes life is simple no classloader problems and when the job is finished it will ultimately die +There are no plans to migrate to maven it doesnt generally solve any problem we are having and is complex +Apart from the documented contrib tests the core test failure is +Heres a patch that removes libhdfs +fancy having a go at this one +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I checked next tests on hudson All of them say Duration min sec Fail Skip Total +With this patch I got through runs of derbynet on AIX +Username lqd + is no longer supported +I am consistently getting this on +Attached is an example WSDL with a qname element that will show this bug occurring +I started to work though some issues a while ago but did not have the time to get into a releasable shape +Ted sounds good to me + +Various tricks are currently done to get around limited interfaces exported by Hadoop +Closing the issue as fixed +Logged In YES useridHello ThomasThank you for you effort +Hi Gary I originally raised this issue on the FUSE forumI dont mind if theyre removed but the counts on the connections tab should also be removed rather than showing +This doesnt appear to be fixed in the BRANCHSOA CVS branch and were past the SOA CP deadline now +it should be fixed looks like it didnt make it into build +We will have to port this later if necessary +Alan As a newbie to this architecture I have made the assumption that you wanted to files +Colins the discussion about resolved paths versus unresolved ones belongs on not hereDaryns comment does belong to this Jira +Tried it on my mac and I still get the three failures +For example what is used for? Why is it needed? What is the design of +The only part I do not like is the change to the sorting inPlease let me know what you think +The patch has been applied with some FAQ has been site has been deployed. +Updated patch which enables us to run test from the commandlineThis has NOT been tested merely updated against branch +Revision is . +Using bloomfilter is not always a win especially when the columns are already cached in the file buffer Maybe we should add a per CF option that enabledisable checking the column bloomfilter +Not quite +Closing. +Seems like a trivial fix and this affects our container so if nobody has any objections Ill apply and commit this +dedup logic would still be required +Be my guest +It may be better to have a get operation rather that a one in orphanedMemory methodIll create a patch soon and thank you so much for the reminder Tim +Resolution will take too much time to make it for the seems that a solution to this issue was close at hand in Dec + is also failing because of changes in +Going half way with many static method signatures complicates the API more than it helps +My bad +This is already configurable through properties of the respective implementations +Please give more specific explanation of where the problem the services page Im seeingAvailable servicesEndpoint address httplocalhostservicespersonserviceWADL httplocalhostservicespersonservicewadlhttplocalhostservicespersonservicewadl is a link with the underlying value being the same httplocalhostservicespersonservicewadl +Im switching this to you since its the only item marked for a release +Perhaps for everything memory related I can make web services all common with MBIll break the version info out more similar to actual classMy understanding was that jerseyjaxb handles escaping but I will test to verify +patch with updated properties file +As for the license we made a mistake assuming it was LGPL based on the first page that Max linked to +Patch helps but doesnt address an underlying race condition that is a cause of a number of of this change should be possible getting the client to perform a synchronous opertation but modify the request to be asynchrounous +There are Objects in the with mbOur project has probably around filesPerhaps it has something to do with subclipse +Have you made any progress with that? We can tag team if you +Attachment has been added with description Soft mask dump from the has been added with description Test case with changing has been added with description PDF render with different backgrounds +modifies and to allow dumping of top terms in the topics +I found sereval licence headers in some ASF projects +JoeyBigtop looks cool Ill check it out +Thanks +bad the previous patch for the Java file was not the good one +The information and examples provided by Jack in this issue have been placed in the DATE TIME and TIMESTAMP files accordingly +wasnt meant for M +It was exposed via our MMX welcome page which is gone +Still an issue with trunk and with the release +Gerolf can you look at this if it is fixable for? else change it to with keyupkeydown scrolls the selected list item to the top of the suggestion listalso when the user clicks on the scrollbar focus is set on the textfield to enable further keyboard navigation +Fixed but the whole ECS is facing problems with Brew builds of books +Updated the patch for trunk since it became stale +So i am thinking of backuping up everytime the localStorage is modified +Patch attached splits the Rakefile into smaller parts +Will get the patch uploaded by for fully qualified HDFS path in addition to simple unqualified pathThe qualified path indicates that the path is accessible on the specific HDFS +is anyone able to replicate this issueI sincerely doubt this is replicatable still or it would have been solved before GA +Second patch for the new style to Lookup forms +I have attached patches one just shows the main code change and one shows the full change that will be applied with changes made to the tests in you so much Andrea +Thanks KelvinPatch applied to LUNI module at repo revision rI removed the duplicate block comment and replaced tabs with spacesPlease check it was applied as you expected +Is this solution sufficent +This patch fixes bugs in the number of results returned sorting order errors on edge conditions among others +That is instead of setting locking to unknown set it to empty +I will rebase it +Will run ant test and testpatch for it +Hmm +Actually binaries are called which AFAIK is so that these can be used for EAP EWP and whatever else we have +st Patch will be thrown if we try to setTimes on directories Moved the atime field to i obj +This needs to be done in M or it needs to get pushed to +I committed this +Checked into trunk +Checking fix into SVN on Helens behalf +FixCreated in filters to seq scan was done in +Everything works as aspected here +Screen shot looks good to me +Marking patch available +updated and srcetcthen executedmvnremovemvnformatmvn clean the following commands on branchgit ace ab bfvi elbgit add elbgit commit mfoogit rebase imvnremovemvnformatmvn clean installgrep R Licensed to jclouds Inc +Hi HadrianNot sure how to avoid this +It should mark these props as mandatory since they obviously are and it should also throw a friendlier exception when a mandatory prop has a null valueFailed to update Queue Destination managed component cause the property pageSize is mandatory a value must be specified +Committed to trunk +If this is still an issue against the latest release version or the nightly build please provide a relevant application which reproduces it +I want to merge the java and jsp properties files into one and take out the duplicates during the processIll commit the patch shortly +Working as designed +say ls is aliased to something in some environment this test would break again +If possible if Nutch PMCers could check it out next week Id love to get this release out the door and focus on +Dear allI was just checking on the progress of the component integration and could not discover anything un Did I miss anything? It would be really great if the component could be pulled into one of the next releasesMany greets and kind regards for the reminder +You try to run class main method +Great guys +with different deployment the classifier should be also included in the ID if it is present +Please feel free to share your commentsquestionsthoughts on this +We just discussed on the call that we are not doing a dbupgrade step for this but reset the schema version to with no remains when do we start using dbupgrade scripts again +so I leave it up to you +what are the big dipsthe figure is formed by several benchmark runs +But for deletion the cost of deleting any replica is the same +Do we even need this? It works from if data in is world readable +This makes the s lookup of grandparent directories more portable +It passes my unit testsThe new version was merged with the current trunk +Patch for port mismatch issue +Verified successful +Username hansd +Committed to with revision . +Thats right +I looked at this again and reduced the number of duplicates a bit +I am good with the name of the procedure and with respect to roles and permissions I think it should just function the same way as update statistics routine does +The current UI allow you to specify as many type as you want there is no feature missing here +As soon as I submitted this I realized it was because I was missing the Realm information in the header sorry +AmilaCan you please check for other situations where we can avoid calls or even completely eliminate itthanks added the synchronised part to method name +I need to figure out how integrate this with the Solr Cloud distributed search stuff +Theres no features branch +It is pretty Ruboto specific but I suspect you can generalize it to cover all JAR filesIll do a pull request when I can sort out our fork +Anyway its going to be listed in the examples section +In negative case there are deployed two of them +You need to add testds jndi entry in to use +One acceptable alternative for now is to drop the jars and add a note to end users that they should explicitly get and add the JAI libraries if they want support for PDF documents with rotated pages or embedded TIFF images +The issue is that the versions that populate the are all Strings so we dont necessarily know the best way to sort them +Thanks +This is fixed as part of LSR and will be released as part of Pig +Thanks for feadback +We should bring discussion of annotations over to the list +map in reduce out Should by default just consume bytes and produce bytes +Its very nice to see the source code will full Javadoc +Ill write the remaining tests for the standard message body readers to make sure we keep behavior consistent +If manipulating the verbosity level is what this issue is about then the job of work is actually to make the setting of the verbosity level easier in both in Gant scripts and in the Gant source itself +I believe the ideal solution would be to find out why the label is needed and fix the code that depends on it so the label is not needed any more +Do you think these sources are compatible with? How much effort would be needed make it work with the autotools scripts +I know Brett has set up similar builds before +Countersincr was not preserving the display name +Shawn dont worry we can iterate over you patches! I am optimistic! Welcome to solr lucene +verified thanks Tony. +To clarify moving people away from Thrift is a Good Thing so my position is we should support batched counters so cql is a realistic option for counter users +This should be a nice improvement to the to approximationPatch with fix and new junit test attachedpatch little addition courtesy of Michael Goddard +we find the nice n where this is almost as bad as brute force and cut her off thereyeah that sounds good if there is a nice transition at a relatively lower nbq +looks good +It seems that we do not get any clear solution for this bug for version. +This should not happen in vanilla hiveWe uses which does not use hadoop cluster running locally in hiveserver for small dataset +This adds a dummy script class which is useful for reproducing instances of this bug +looks good. +I agree that form in not needed and this form is now +Snjezana do you have any ideas on this one +Making the register and unregister methods synchronized +Unless someones been paying attention to this ticket Id doubt that theyre going to come up with its staleok but starts a pass +Committed revision Thanks Jayson +if i input invalid character like to define a jms resouce name there is no error message in conosle but error message in +Im struggling to understand all the implications of this +already just downloaded the last successful build from hudson and looked at the relevant code in the file scripts and it is still you please tell me in which commit this was fixed and what the duplicate areThank you +If it lasts open in Ill definitely get back to it with a great pleasure +It is a bug +since servlet requires java and wicket requires java we cannot properly support this yet +Perhaps the circular reference is to be avoided? In general though not using the constants exported by another class has a bit of to itPlease advise +Moves the entry to and makes a minor clarification in the updateCommits log message +Ill still keep the ticket open because of +Who exactly are the initial committers? Is it thosefolks currently authorized to commit to cayenne or those listed as being on the pmc according to the board resolution +This is a good learning experience using system properties to get rid of the equivalent of defines in Java +Then it does not compile alsoCan you please update the patch to trunk +this is a unless someone chooses to override the default directories which could be very helpful at some sites. +If it is referred elsewhere then I would check the classpathIDE RAD betaApp +Ive changed my mind and removed the port property from the sample in documentation butwhere have any of you seen documented that property its only on a wrong sample its not anywhere else +To me interface just feels like extra baggage +Given that is about parallelism as well as concurrency and parallelism is solely about performance improvement benchmarking has to be very important +bulk close for the release +I will create a patch and attach hereThank you! has fixed the issue on the axis branch and will be available to us with the release +I used gitapache which is lagging behindI pulled in the latest changes and now I can run the specsThanks for your work! +QE please confirm +Screenshot of diff between old file and new file new file and stacktrace from Eclipse +have enough memory on the clientLet me do the math test it howeverWe can as well do something in the middle do a LRU on the first map the one on the tableName +Nit would be good use the method where it is typestring is accessed since it gets cached anyway +I hope I will not need it somehow +Hello JacopoThanks for your comment on thisI agree on all of themHello Abhinav Can you please take the recent patch uploaded by me and start working on Jacopos commentPlease ask question if you need help so that we all could discuss the things on the mailing listThanksAshish AbhinavAre you working on the points suggested by JacopoDidnt see any response from since long so was curious to know how is it going? AshishWe are now working on the th point related to entity and fixed other comments as suggested by Jacopo dated AprWe will upload the patch in few daysThanks RegardsAbhinav VaidiLabs L T Infotech LtdMumbai +The intent of this loop is to operate like the standard group fetch for normal scans +Right thats a good point with later availability of endpoint +This will be part of the jbpm integration +If I do so is it overriden by the context variableIf however I call this method from JSF expressions well I could simply reference the user context variable in the call I am missing the point +That is what most users would end up using +You could also try grabbing the latest xerces jar and adding that in to use that instead of a buggy schema implementation in the JDK +I think S is just for running scripts in the bin it work if you just do jruby scriptserver and jruby scriptconsole +Further investigation showed that the roots of the problem are in BC provider implementation +We have it applied internally after doing a perf comparison seeing positive results +current code decides the DN should shutdown while your patch decides to keep DN runningThe behavior is actually the reverse currently the code will tolerate failures as long as there is one volume still working my change makes that number configurable and restores the old behavior which was that any failure causes the datanode to stop running +There are that reexport which is not recommendedIt is safer to fix it after branching for M +mvn coberturacoberturaThis project does use both hibernate and spring which another commenter you using Maven from SVN and Cobertura from SVN? It was only just fixed +committedThanks Byron. +Im hoping to do this without having to manually copy any jars into the lib directoryIll create a patch and upload for folks to comment on +No code change needed. +Vesa Im sure theres something we can do to help fix this but would it be possible for you to narrow down the bugs youre seeing into a few simple test cases? The sessions shown here help us understand that theres a problem but theyre somewhat cumbersome for us to use in assembling a fix +This modified versionalso sets the default location of keyStore and trustStore passwordKindly review and applyThanks ShawnSince the patch has been modified so that it no longer throws Null pointer exception therefore I do not see where I can add the new message +So Id assume this failure is due to being unable to replicate that view +Filed to fix the documentation +What should I do to set this parameter globally? grails war works now if I do grails D +Adrian thanks for the clarification +Also applied Teds comment +There is a new class which along with class is a subtype of a new class +Next step would be to remove the dummy +But there are other architecturaltechnical reasons which I cant disclose which require that this arrangement has to be in placeIn production this will probably become a as all pairs will likely be brought up and down by a single user and in most scenarios there will only actually be one pair +If has no default goal nothing happens and this is expected behavior. +for urgent issuerelated to my jobemail eugeneyumobile no RegardsHao +Sorry for not making patch name intuitive patch also changes the existing testcase according to the new changes in capacity some more testcase to test multiple assignement +bulk defer to defer to +Thanks Micha! Now it looks much better +Ive not added a similar method to the interface which is intended to be simpler +or AS +so the Eclipse SSH preferences seem to store the keys with erroneous permissions +If you have any other approach you would like to follow I am all ears +It seems fixed in and do we have any plan to patch it in the +Weird. +I cannot explain this +Thanks Sergey +in a cron job +IainI think that we can close the issue now for that screen the facility id is mandatory because it is used to quickly create a purchase order from a list of requirements and so we have to use requirements for a specific facility +Correct version of trunk patch this time +CheersVivek +Which if I have well understood is not a best practice at all +Unit test is passing on cruisecontrol +Zone id is optional part of any address so there is no meaningful reason for disallowing it in loopback network interface +Thanks you very much +Adjustment for fix made in +Yes I understand doing the right thing is never an option +I mention Grails purely to give a reference to the Groovy flavour being is a Groovy problem +Looks good +Please feel free to review comment critiqueOleg +I saw this failure too +A test case that requires the multischema reader +If I use Xerces I get this same error +The documentation shoulddescribe it you described does not make much sense to me +Also has access to salary for EmployeeAlso has access to team when using a candidate class of Person +I think this should now be fixed +According to the EE specs I am pretty sure we are supposed to be adding this anyway +The purpose of this is to make each command leaner and the replication stream smaller even if this does involve two separate commands and hence command +I was annoyed by the syntax of some methods that had a throws on one line and the exception on the next +I already configured the job it includes RHQ Jopr tests trunkThe tests run for this configuration but the whole testsuite will ask Shelly and Ondra Zizka for checking the build +We can add the pieces suggested here +Done +The fix +Actually I think itll make more sense to have instead of new encoding value that makes functionality more explicit and can also be set as default for factory +This seems like the best and most compatible and Hpricot are good examples of this approach +Maybe its not related as exception stacktrace is different but scenario to reproduce is quite similar +bulk move of all unresolved issues from to +thanks stale resolved issues +Great! Ill take a look +Pleas advise +Here is the patch for this issue +Just delegating to the inherited hashCode and equals from Object satisfies the contract and poses no synchronization issues +If any key works not as it should have to please change the binding yourself in class +You should be able to unzip this on top of your existing perforce project add the files and check them in +Please provide a test case for the issue or help us diagnose it further please +Attached is a patch proposal that I believe corrects the problemThe resolve method of has been modified to properly validate return types SMALLINT INTEGER BIGINT REAL and DOUBLE +We should consider creating a new RC for with this fix. +Try using maven working with trunk you help getting this do you think might be the problem at my help +Please see the attached unit tests +seems to be a duplicate +Finally the new patch is fine I would just revert the changes in the XSL file this is unrelated to this issue +false prevent the switching to cookie VersionIt should in +banners should have beta as fixed the two that had that listed and +Bumping issues to except those that involve with Struts or +could return without confirming meta is assigned +As Sergey Gluschenko mentioned this will be a good option to improvise. +This patches puts the checking of meta location inside the loop instead of outside of it this way if youre in the thread that comes from ZK then you wont be waiting on yourself +There are created jobs for running on JBP and +Ok +a longer timeout cant hurt here +To be able to do something similar with Infinispan yeah wed need to implement a change similar to what you suggest where each query is stored under a cache instance with that region name +It has been committed to trunkIve created branches which rebases the branch and moves the functionality to it +It would be helpful to see what your code is doing +Applied James patch and all tests pass +Oh yeah thanks for reminding me +Thanks Karel! Integration tests pass on ec +It disables the UI +fixed in r +Following this idea the solution would be to add imports to per method +An icon file is addedNotesa +One step at a time +Attach an initial test implementation +Rescheduled for extra hrs for the next week need to discuss in what context queue events are called +Id say make it configurable it shouldnt be a lot of work to support both protocolsLog warnings if correlationId is set when not expected etc +I also removed nocommits from indexcodecsindexvaluesindexvaluesindexcodecsappending +have to read through bat scripting to fix it +Thanks Ajay! +I think as we noodle on this we should probably add a Runtime Settings that would allow people to get a summary of the modes in effect +Closed due to the release of Tiles. +Hi NickIn the case of my attached file to this bug I get junk files none of which is the file I originally embedded +I changed the ctor according to Andreas suggestion. +We have a task called which may or may not submit a mapreduce task at runtime + +war file under MB attached seperately to get the +Sorry I must have had a stale view. +Committing now +Done please check the latest snapshot and report back +For the record I wasnt the one discovering this issue! +I would make it DR only for now? And only when we come across a demonstrated need should we generalize it +Don Im pushing this out to bc I would like to see it done but know that you wont be able to get it going for a while +Add code believe this has been fixed in +Thanks +Thanks Uma. +Attaching +Thanks a lot for the reviewFixed the nitsI will post the patch for trunk as well +thanks for participation +So have you been able to reproduce it? I also think that it is a really bad bug that I want to fix but I dont know where to start debugging it without more information about how it happened +Add a note to clarify that the PCs should propagate as defined by JPA specsStuart if I do this is it sufficient from your understanding +HenryThank you for your comment +after release. +The code has been changed to output a dummy which should resolve the issue. + +I see you added an include ofimport I see in this directory there are a couple of other examples each with a comment about we really should not do it but went ahead and did it Strictly speaking we shouldnt import classes under the impl hierarchy hereimport import import Strictly speaking we shouldnt import classes under the impl hierarchy hereimport When I looked at Dans suggestion I also ran into this issue +Moved to be fixed in versions from fixVersion to targetVersionAlso since has been released without this change set target to instead +As a workaround you can remove all the data from the version database or simply create a new empty repository to import into +login as admin +Closing issues resolved in released versions +to resolve issues +Thanks cmlenz +This one is probably for the FAQ The environment variable is named XERCESCROOT not XERCESROOT +Sharing the between projects would require to have some file location configured in your and could be probably be ll try to patch this +Not supporting convert and the use of a scalar subquery compare with an aggregate caused that exception +Fix applied in fffeabedcdcdfb Testcase applied in bfaffcfcfdcbcaecd +I have updated the check and ran tests on to verify +Letting Hudson give one final shot at it +No response assuming ok. +Thanks! +So basically there is no fallbackIn the patch we have fixed the missing function in the native jni code +The documentation reads thusbq +The error is in VMArray +Basically this patch moves the logic for comparable method from various implementations into DTD +Ive just committed this. +The correction to the AS kits configuration file has been correct to reference the correct sequencer class +The current version of Activiti depends on Mybatis +Really they are just setting up the defaults for the new java project wizard +We upgraded to and the error disappeared +categorizing +Another question would be how far the Manipulator interface would go +In general getting some visibility ideally though tooling or more pragmatically though a generated report would be useful +Looks good to me +Tested your files with but couldnt reproduce your problem +Close all resolved issues for Engine release. +Marking as out of date since axis is no longer used for web service calls. +Having seen a significant negative impact of that approach in production I remain skeptical that its a viable approach +Username gohde +Switching all issues that have been in resolved state for more than one month without further comments to closed status +I opened for catching accidental DV fields +ted stackCan be committed imho it seems we all agree that the issues are not linked to this about Stacks proposed changes Dec +Jonathan Halliday suggested that the EAP build had been pulling in a bad version of jboss logging or its not getting added to the compiler path and hence the annotation processor is not runningLet me know if you need more help with this to related Tomno I think once the source of this problem is known we only need to have a release where correct version is pulled during build process +we do have a thread pool we could use for this +Jconsole output. +This issue will fix this with the new report +Looks like it is already filed as is a duplicate of +Created custom implementation of to be able to convert the new lines inside the attribute values +Perhaps we should add unique columns that are not Combined so we can quickly identify missing values +Ren unfortunately we cant assist you with issues in please report issues with corresponding project +Anyway she came up with the attached designs which I believe are very good +tidying to do still on documentation +Hi Scott Thanks for your quick replyYes I have instantiated the widget and is using the correct userID and shareddatakeyIn fact I have two machines on one the API request is working fine but on other its giving issue +This is fixed as part of +For the development mode once Jetty configuration is figured out I will try to make the update ViolaI have fixed your problem with revision Thanks for you verification +to change validation should that be allowed to be used arbitrarily during processingOr only between usages of the instanceIn either case additonal testing is needed to esnure that the new settings are handled correctlyIf the setter is only allowed before intial processing how is that enforced? Or is it purely a documentation issue +Part of a bulk update all resolutions changed to done please review history to original resolution type +If it works we can pursue the Struts upgrade +Or maybe I fixed this already without knowing +In the patch the doc stores are being flushed on get reader and the doc stores read successfullyFlushed by not closed? And that works fine on Windows? Thats fascinating to meIe that our dir impls are able to get an when theres still an writing to itBut for RT this will be necessary +Tx Mayank too for the reviews! +The SIGSEGV is caused by the linked issue +Sonds like a plan! +committed thanks! +Also this would require lots of changes in the Yarn as well as an additional complexity for app writers +Thoughts +Attached patch for improved algorithm +But this is the problem of application that gets exceptionsWe have conversation cleaner thread in +so if we similarly dont count a couple storefiles we are just extending the time window not changing the behaviorAlthough I agree practically this should stay +I ended up completing this work +Committed a fix +I totally agree with Stephan this would be a very elegant solution +Fixed in SVN trunk and branch +Fixed under revision r in the trunk and revision r in the branch +Final solution is now on trunk and branch +How to start up Katta after solr patched? a start the master and the nodes of Katta from the commond line or the Katta as a component in solr +Thanks for the help +Left it as is for now Ive deleted them and it looks OK to me but this was part of the original submission so only Daniel can really answer if there was a need for themIm away for a few days so wont pick up any queries and appologies now if Ive missed something tends to happen when rushed +One requirement here is Axis branch use Axiom not version hence it may require some effort to merge few revisions from Axis trunk to branch in order to work Axis with Axiom +If nobody has time for a review I think we can do that as part of the RC vote if thats ok +Observed on as well +made implement added lots of ASF headers and comment stubsYonik andor Hoss +Update hftp to use secure port to find tokens +Thanks Suresh so in this proposal you means we just make shared by several and in client side view its totally dependent different HDFS clusteres +Test page for drag drop tree functionality extension avalable overhttp +server toI hope now is everything okBest regardsTim Tim added +Sample +jenkins is not working +Removed The following manual tests were performed +Getting the log right now + Added missing static identifier for the cache map +Reopening to get Bjorns test merged into Nialls test +This issue also applies to Hibernate and +It was released with the fix in +It changesadds hostname but this patch addresses portnumber related issue +It doesnt work too +The pom described in this issue will work as for more details +including main page update +Ok thanksDo you have an idea on how log it could takeThanks again This one was a bit tougher to support as well. +Sorry we do not provide any such redirects as it is technically infeasible to do sowithout ballooning the size of our config files +I would like to verify the compatibility of this patch with older clientsOnce this new behavior is committed we may want to correct some metadata +Marking wontfix seems like IPC andor HTTPFS is probably the better way forward here. +Considering this Resolved +You dont note it hear but weve talked previously about providing a HDFS URI for buffering purposes +I have also created some tests to prove the problem and test the resolution. +is not descriptive as a name Id rather have I do agree +I apologize for not responding soonerRegards thanks for letting us know. +The name are only created lazily by wrapping the name in a Hits instance +doing more IO +resolved as part of release +A study in single line outline drawing +Btw another challenge we need to solve is multiple plugin includes on one page +A little frustrating taking some time converting to storing in session as well as recalling data from a session +It could be possible to split the current request id bytes into a version and an id field one byte long each +Actually one without any Roles is I am working on is achieving the shiro style permission in So one without any Permissions is normal I think I could launch a new discussion after finishing my you for your reply + The attach contains the readme file to instruct how to compress dojo files in Directory The dojo package is the compressed dojo files +It need to update hbase +We still need to check compression type metadata before applying dictionary compressionIn this regard adding WALVERSION seems to be redundant +My proposal doesnt help your early termination query +This bug has been fixed as it is a clone of which has been fixed +This must be the same issue described in please go ahead and commit this Matt +Gavin could you explain what does your phrase The Seam Facet page is completely broken on Mac mean +It looks like only unpacks static files and but not classesApplicationsserverdefaulttmpdeploy ls lRtotal Howard admin Feb Howard admin Feb r Howard admin Feb +The piece of cake was bigger than my mouthPostponed +I have identified and fixed the CSS bug however there are some paging issues I need to resolve before submitting this patch +Adding a new module or to an existing Apache project creates as much work as doing it in the incubator +Are you saying it is caused by this patchOkay I saw that there was a previous comment +After a certain amount of nodes are created the triggers add node events for nodes that already existed +Yes I thought that was odd but If I remove the css plugin jars it runs fine +Have you fixed the cause you found on trunk? Ive written an IT but cant get it to fail with trunk +patch looks good +New patch with cleanups and minor to trunk Ill let it bake there for a while before backporting to is nice Sami great work +will be attaching a patch to fix these shortly. +IMO can be marked deprecated +Thats a good question +Should be fixed by Jons latest commit that makes us using Commons Logging again. +Confirmed also +This is the job of the application server +For we will go with the more limited approach implemented today +all questions answered +Fixed by Roberto +ETA to be finished by the end of today Mar +I moved this issue to the release plugin because I think its related to pb must be fix in of send the user tagname to the tagname sended must be tagnamepathtothe in parent pom scm url is like this scmsvnprotocolpathtoparentand in module scmsvnprotocolpathtoparentpathtomodulethe tagname must be for the module tagname is a possible fix +So the client still needs to be able to skip unreachable endpoints itself so describealive seems like gratuitous complexity +I prefer the default list +ping + +Some more fixes +Perhaps this should be made into a configurable option? I agree that ms is a bit excessive on a small cluster where the is generally bored and weve also seen big throughput improvements especially when users submit jobs with tasks +Ready to use compiled jar +Note This applies to inflight unprepared dont need recovery by the parent +Hey Vaibhav SandyDid you guys pursue this any further +That was a typo +And in fact the resource managerapplication master are completely ignoring disk utilization at this pointThe plan is to add in disk utilization to the resources that the RM uses and then have AM request both disk and RAM space for reduces with disk space based off of the size estimate currently used +Thanks Sid! +Mass closing all resolved issues that have been closed for more than one month without further comments +When a patch is approved please request it to be committed it to +Any progress on this guys +I am hoping to do that here today +this patch now applies to good +checkstyle file to point to somewhere meaningful under UNIX +Okay now its as fast if not a bit faster +I will devote some cycles to that end +Code delivered +Okay +The test failure is unrelated +I see +I would rather click more and get all and have less just go back to +Replaced the use of mknod with mkfifo in to improve compatibility with Mac OS X +Siying I dont see the new you cant see trunkconf is already included in the diff of the review board +Dennis can you please this so the issues can be addressed? Thanks +In PTF handling we do keep a copy of schema in Desc object to be used later at runtime +Sending toolsjarTransmitting file dataCommitted revision Only xaHelper is required to be added to. +We should close this after the review of. +Hi JochenIm sorry I was wrong it should indeed compile and fail at seems to be a general problem with y where y is any operator +I could add a test for that but is that really worth itAnother area that could potentialy throw an error is in implementation of the +David could you please try the attached with your test caseIt goes back to your original diagnosis that the issue is in my other fix took care of the broker cases but left a leak with the client where the operations are done in reverse order usually +So it is even more broken +New patch rebased on the latest trunk including all the files that moved +I see this as a better solution because ultimately we would be building the entire engine into a single binary so it wont mind in what source file was the header implemented +Thanks! +Could you please clarify what filefiles do you mean +i think this is plenty reasonable for most systems +It should not be done on parse +See the attached for a quick and dirty experiment of what we could do if we did want to directly use the log serviceIf the runtimes already unpack the bundle jar then I have no problem with the embedded jars +I may be missing something here but browsers have their own lists of which may not be the same as the used by the Java installationIn which case maybe the problem is that Java does not have the required CASimilarly for wget and curl. +No I agree just checking So a patch to the docs is what youd like correct? +The Thrift compiler project is still VC only +I had to modify the wrap the JAVA and CP arguments on the command with quotes since JDK onwindows now installs by default to the program files folder +Here is a simple test case +In general modern machines are good enough that megabyte to gigabyte sized files can be read and written reliably enough that CRC errors are not a dominate concernWhere they are a concern I suggest we make the CRC files visible +Currently was testing with the netty server +If none of the name nodes are available exceptions encountered are also listed on the web page +After expanding the test cases it was clear that the patch didnt work +How about merging this to + must be loaded by web application class loader each application must own their own class so there must not be any should also get refactored to be picked up with rather than using static methods on it +Thank you for testing Katrin +removed the section on how to compile libhdfs as doug pointed out it comes with a release +Cool Ill try to get a patch put together today and post it for review +There is no in is a in the edsmodeshapeclient directory + +just the that installed with geoserver +If someone later comes and implements then there is nothing referring to the fact that the object snapshot returns is thread safe +Another possibility would be to just create a configuration parameter to tell whether lib should be populated or not +Test log of that already have a patch ready +Attaching +Yes Springs XSD defines a default for those attributes +Patch to switch highlighting from bold to italics in would this be better? With your patch you wont be able to see the highlighting for italic text +Various pom fixes include removal of references of guice +To workaround the problem I used sed in the test scriptHopefully by next releases EAP test suite will use the correct artifacts thus work with a correct maven repoIt seems that we dont any modifications to the repo you have provided nowThank you for looking into this. +How about Windows Scripting Host? Javascript syntax can use COM +Well never delivered anything on Hibernate Search before +to the fragmentshost way +Not for sure this patch fixes this fail but it didnt fail in patch build and its been failing all up to this in patch build tests +Is this problem specific to updating a CLOB columm using a stream? See +adding as known as known issueVersion history is not cleared when importingWhen performing an import the existing version data is not being removed +Jason +Ah paying closer attention to Kims comment Its only the version that was attached after the issue was reopened +The LFU implementation has the same problem as the Expiration implementation +Would you please try this patchBest regardsSpark patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Will update the ticket as soon as I have it +That way the external process wouldnt need to continually poll the job tracker to see if it was done yet +Fixed issue added for the problem mentioned in the comment above +fixes to return on service start if service is already running +And I dont think its bad SRS contains fieldInfos we can just assert in that they are null +Thank you for your improvement +David has been given link to dump file. +nit bq +Patch with to remove public methods iteratorEntry iteratorExit and merged with iterator +Thank you Alejandro for the patch +tld files within the folders of the bundles. +Migrating bugs marked as resolved before and tagged as version as fixed in RC +Ill leave this open because it needs some browser testing and because Im hoping someone else might take the opportunity to spice up the javascript a bit +If thats what youre looking for then maybe we should spec that out in a separate issue as this one is tailored to +patch attached +Please update if you disagree. +Please test it out on tonights snapshot and let us against an server should work thats what I used when developing the patch. +It is doing an operation that is specific to the internals of a core +The two might be quite complementary +the example uses hibenate jpa +I just tried this again in a workspace launched from trunk and I cannot reproduce the problem +Im not sure I signed that but Ill have a look +svn upAt revision +Thanks a lot for the reviews Todd. +Sorry that I cant post the exact extract of the code as there changed too much +Also if you need any help or feedback on that second issue please let me know +Thanks +The log level for JDBC driver was set to +I have installed jdk on my machine and removed the code +Some questions Suppose my application has two threads T and T each with its own Connection +Fixed +Should not have been code now acks in batches when possible +Its been like this for a long time so maybe the change looks good at first but maybe notLet me post on dev to get their attention +Bulk closing stale resolved you remember to reverse to condition when trying isAnyway like i said before i think checking against the array ckmpinent type is the more accurate way to should read component +Attached screenshot of effective roles page +Please find the patch for attached please suggest corrections and improvements +Aaron do you mind committing this into all those versions or at least some of them +The tests all pass when I manually run them on both and trunk +Moving to is fixed since we consolidated on. +I start the eclipse and wait it totally initialize then I start the MVN import and wait at least minutes to be sure that is a deadlock +Thank you for taking time to do this +The patch looks good but Im now confused by a different thing +Why the cloneduplication tim +Thanks ArunYour patch is in trunk at revision Not sure its the best fix as maybe it fixes only the effect and not the cause but at least for now it works! +Applied the patch in fb thanks +This patch doesnt change anything with regards to insertions past deletions +Attaching a screenshot of how the admin console looks in IE +We are about year later and they keep releasing Zookeeper with flawed despite a fix is fairly easy +Wont doing validation in parser will result in two network calls first to fetch the current column info and then update +You should have access to one +This is because jboss messaging on the client has its own into which it deploys an that it downloads from the server meaning that most of the aop classes are in fact being used. +keep the old api alterpartition and add the new api alterrenamepartition +However by inspection you can see the obvious concurrency bug that exists in the current code +Can you confirm +I am currently using +The In usage of createString modifies any thrown exception to include information identifying where the exception occurred +Well for starters you should amend the test to check not just the length but the contents +I seem to keep failing to understand your pointI guess I should clarify +It is not the same as sending a single header with a comma separated list of values +Would like to see this committed asap +I just committed this. +I believe there are existing holder objects that will take a stream andbuffer it to disk if need be so that you can read it again +So if the problem persists Ill need a quickstart to reproduce it +I have committed this. +This should fix it +This aspect no longer exists. +Thanks for the earlier explanations Cos +It has been sitting for a while looks like is no longer a problem +Just tried with and works well +It is difficult to have a testcase for this issue since the number of spilled records depend on a number of factors like the value of etc +The original patch would not apply cleanly for me against without fuzz and whitespace options and when those are used it applies incorrectly +Ok I find that I didnt annotae the method with so You can close this issue +Patch looks good Koji! I will commit shortly +Closing as wont fix +Looks like this is fixed in BPEL editor version +Gave the bench a try here on linux with and I did one throw away run with the new patch version and then recorded the times for both got drop with the new patchReran both many times after and got approx the same times for both every attempt using both JREs mentioned aboveDunno know what that implies but another data point +Yes +HTML file that will reproduce the bug +bq +I merged the patch to. +Verified in JBT +Damn second time today I collided with Lars. +Ill do that next but will take some time +How about we create a based on what Tom has done here +You should first analyze the whole line as text doing all kinds of substitutions synonymizing colocation detection whatever you fancy and only then try to build a query out of it. +Isnt the default s? s is long indeedThe trick is to find a timeout that catches actual machine failures quickly but does not trip on temporary network glitches or GCetc +I am fine with any solution on it + +One day it would be nice to have have some respect for memory usage and performance as well so that it can be used directly instead of needing this alternate implementation for dealing with non trivial amounts of data +Both are not HQL functions +Hi jingthanks a lot for reporting this too +Actaully shoud make sure return the raw on get rather than a wrapped object +Fixed in the version. +Closing all resolvedfixed issues of already released versions of Roller. +Duplicate of +Yes what I was saying is that we should whether using extensions is the right way of registering beans etc +Superseded by. +This patch implements the proposed solution using Records Config variables +buildtestclassesPIGHOMEbin +Also server support rsync protocol +As i understand the types are inorgcodehausxfireaegisll try to the fault serializer but if needed ill make orgcodehausxfireaegistypesoap and put a there +This patch includes some new tests in to highlight this problem and fixes in to take care of the issues +Some day we maywant to store blobclobs differently that would perform better +This scenario is generally avoided by making the dependencies optional but making the mongo driver optional in would obviously be nutsI dont know if theres another existing artifact they can reside in which avoids this +svn commit m Add properties and methods for apisrcjavajavaxjdo Sending apisrcjavajavaxjdoTransmitting file dataCommitted revision svn commit m Add properties and methods for apisrcjavajavaxjdo apisrcjavajavaxjdo srcjavajavaxjdo srcjavajavaxjdo srcjavajavaxjdoSending apisrcjavajavaxjdoSending apisrcjavajavaxjdoSending srcjavajavaxjdoSending srcjavajavaxjdoSending srcjavajavaxjdoTransmitting file dataCommitted revision . +Chris thanks for letting us know this duplicates +First set of tests committedCompleted At revision +What do you think of the minversions ideaI think a new minversions in core is a fine idea +Nothing incompatible either +Thanks for catching this Rick +No where in there does it say the new mappings do not support sequences +Updated to include signature fileThe RI passes the TCK with these changes +Thanks! +Thomas if you have not seen any issues with the patch I would like to commit and start backporting back to +More research on this grails war deploy in tomcat running with doesnt TOMCATHOMEconf adding to each connector the attribute solves the problem when deploying a war generated in any tomcat plugin and installing jetty works but when trying to deploy war generated in tomcat does not +Related to +Are there any errors recorded in the broker logs? +Added new selectedIndex enhancement to the +Of course this should somehow point to an existing +Actually its not YAMLdump but YAMLload that has the problem + for the fix applied at r +I ran this against my local jenkins and all tests passed +I recommend backing out this fix until synchronizationthreading issues have been resolved +Going to do a little cleanup of the generated +This is fixed in the CVS. +Verified by Leo. +required for the demo +Yeah ICLA is patch Yeah the docs looks good +Hello StevePlease try to find the message so we can have a test case for it +Forgot to switch it to three in the patch and forgot that it could be before a compaction on trunk +No bundle is available under specified location +I believe those files are required to let the Guvnor plugin know which files are associated with Guvnor +Sure we can reopen when we can reproduce. +Committed revision +Still need to resolve downstream hdfs dependency issue first +The pb with the Excel export is that we need to hold all data in memory +You need to upload first the parent pom please didnt releaze the parent pom would be that the upload bundle doesnt include it or that it doesnt become merged into the you have your reasons +There are currently no plans to proceed with this since JPA support works fine for transaction boundaries and there havent been many inquiries for this to work with raw EJB CMT +SURE! That is absolutly ok for meThanks a lot Piotr +Going to mark it as resolved +This is a pretty bad one +It looks that there are some confusions regarding the the pipeline diagram on page +Sorry guys this patch isnt right for +i followed the convention that was in there for the orders button +push to +I tidied up how update events are encoded into JSON incorporating this new pattern +SanneI changed the method name and updated the documentation +I think when Shai Erera brought the idea up before he proposed or something like that +Or use a commandline switch sys prop so you can use any env target +Thanks Vinay! +I just committed this +Unfortunately no I tested it afterwards +We run into this error about once a day +Please see did not find the definition of rp +and sorry about that +Search tag elyogragspringclean +At any rate we do have such a library it is on its way and it is moving forward +I have no patch as Im not confident on the various shellBut you should check not to use bash only instructions asthe old tcsh default shell is still there for people thatdont switched to the default bash shellIn my case I need to explicitly specify that I want bashshell as during the upgrade my login was kept to default tschI also know some mac programers who prefer zsh so youdbetter keep your global path setting using only sh instructionsor check the current shell and add the path accordingly +Are you comfortable adding a test case to the test in Futon +Weijin do you have time to take a look at this +No problem my original code also was very +Unfortunately were too far into the release process to introduce additional new components and maintain an appropriate level of QE +HiIve completed the conversion of the rest of compile targets +I verified this against SVN +The data dir now exists but the build should stop in this case +major oopsI set ackclient in the stomp subscription header and acknowlegded incorrect message ids +Username javajedi +ill commit now +I am currently looking into test failures first reported by Jon inI dont know how long this will take +There are two choices to add to the index the offset of the wrapper message which is the offset of the last message in the compressed set or the offset of the first message in the compressed set +ooh i was reading this as it where happing when you automatically chekked for other sites should definitly work +It can be reopened if BOOLEAN is reenabled and still has this problem. +I applied to TRUNK +Im still surprised though by the fact that this was only detected on RHELCount on me if you ever need some more testingThanks again for you effort +As part of my current professional services for a user of Castor I do have access to some quite sophisticated XML schemas. +Thanks for that Tom +Patch updated to fix path of NOTICE file in! +The error is caused by the inconsistency between real exception msg and expected exception msg +Order of the columns is now saved when the asset is saved. +I havent found other code path leading to block cache creation on master +Some comments For the usage line I think the convention is to use surround optional arguments with square brackets +I loop through a collection and try to back each element of the collection onto a checkbox +Comments +Temporary fix using jndiPattern combined with but this is a mavenvelocity parsing issue and should be reported on their forums +We need to have a very mathy committer review the patch or alternately put up some unit tests for it etc +It has different locality properties than the objects in the data segment +So it should be easy to move them to separate component later +Done thanks to Desire Atanga! Thanks Desire. +Were excited to get to Groovy so please let us know if you have any requests or suggestions + +Doesnt ICU gracefully handle the case where it runs out of room in the output buffer? and close per submitter. +Can we open a separate bug report for that oneI am not sure if this version of the parser can handle attributes we will need to check. +Added the step for GA. +All done and pushed +I took the relevent part from and put it in that is automatically added to if scratchpad is used +Server log file an info log when the server cancels a query due to timeout. +There is no way the export could end up producing invalid data if we do just locking which is already possible +Closing it. +Updated in revision . +Its the same as in Hibernate Core +nstable also fits into cfcolumn naming as well +Thanks for catching this +another user made a similar comment yesterday which made me open speaking before working on something we cannot write to we need to be the issues in the code added a new test class to deal with this along with its setup +Are you seeing any exceptions +committed! +I will gladly post patches as necessary +hope it is alright with you guys +Committed and opened for follow up. +A warning Trying to crawl using the web connector is not going to work for quite a number of reasons +If I remember correctly the javadoc was generated at the very last mvn releaseperform it doesnt produce the javadocs now then we are fine and you can close the issue +See very nice +How the is exploited is binding specific +A would seem to be more appropriate here +Fix the NPE in your own code before commenting again I doubt anybody on the dev team will spend time on this issue otherwise +Any comments on this one team +Corrected in what are the impacts of this issue on the release? Does it mean there is no callback registered at the moment in the installer +I meant that I have added throws tags to all the javadocs corresponding to the am running tests on v and shall revert back with the results as soon as the tests finish +Depending on how you intend to use Hadoop on Demand please be aware of some of its main limitations +Im confident that the test failure is unrelated +Committed to trunk +I may have implemented it through Postgresql client tools directly rather than through RailsJDBC +Updated Removed the extra wsdl only test case may have spoken too soon Im now testing with my other xsds there seems to be an issue +One of the component libraries jline used a only native library in windows +Attachment has been added with description has been added with description result with latest night used to convert +comitted including phpunit test suite! see also +It was a devenv issue and wasnt related to ToolsNow it worksThis issue is closed +All tests in solr pass with this patch but I dont know how might be affected +conformed the issue and applied the patch +If we need it outside there may be a way to run SBT to interrogate the classpath for a given project +It seems mainly to hang when connecting to Yahoo imap server which unfortunately i cannot see the logs forShouldnt the gets timeout after a while if the server is not respondingRegarding that patch was way back in +Working on fix +Other changes are harder to know how to handle but in this case a warning seems most consistent with this compatibility policy no +This issue and and likely more can be ported to together with +just unpack the tarball and run the test with mvn clean test as proof of the fix version to in preparation for release. +NowWill run the whole suite and then post the new patch +We just need to release note this +So we should add after the optimization of new logical plan +So Rado if you want you can try a build from that revision plus using the ejb core jar I sent you last night +Only the latest will be remembered as the successful event for a task +Actually +Tired of moving this around +Link relevant Jenkins my unfinished fork of the Jenkins priority sorter plugin where I have been working on implementing dynamic build priority boosting where builds are boosted in priority based on their classification as fastslow and their current wait times +No need for a detailed review it still needs a lot of cleanup but I wanted to show people the idea in more detail +This patch adds a listimages unit test as well as changes the to be just like other drivers do +That would be great thanks for working on thisIf you have any trouble is seeing reports in CI feel free to ask me +spec and debian control files in awsapipackages as have removed all these files since they are the pieces from the cloudbridge standalone project. +Patch to add javadoc to classes that lack it +I used previous versions support jars to register commonj jndi in Jboss already created other to ask that question +John can you close this if the stuff you created for the cbuilds does this closely enough. +accepts java system properties for defining the keystore and truststore type +Assigned to Denis as an owner to linked since this is now discussing and the incompatibleclasschangerror this issue is about is no longer reported im marking this as resolved +Hi AmosDo you have any update on this one? There are very few existing CI issues but this one has been sat in the queue for a long time for thanks a lot TomI will take a look and add the debug code. +No flame here but on all the decent apps Ive done so far +All unit test cases passed +Not sure if its that wacky yet I dont strong opinions either way +Thanks Eugene for the contribution +Hello Please put in your test project and put in run classpath for testThanksBest regardsWould you please try my patchThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +This issue is now resolvedrevision . +All the dirty flag issues actually boiled down to the main property sheet and an errant command being set +Attached the patch for the same +Same thing happens on tomcat +Thank you again Daniel for your work +would you guys mind testing the next nightly build +I will try to add a test case shortlyHowever you must still specify a class for serviceClass +So dont support it at all and revert the last commit clearly saying this +out files I wonder if the above files need the header? Please advise +But if you want to use it as a convenient provisioning tool to bring up a single Hadoop cluster that can then be shared among multiple users as a static cluster then it might still be useful +I have some new TC and a patch to change the module location but it doesnt remove the module folders I think there is one issue with svn diff commandMaybe you can remove those folders directly on svn can you do thischeers +Updating the summary to indicate that this is a code improvement but not a bug +For the record I have implemented Asynchronous in a practically identical way in Seam Cron +That is correct +there will be a version for trunk as well +Im not able to reproduce it successfullySee attached error log +trunk would be the only other place which would get pulled into the zip and andor think the Swing code should be pulled out of kernel and included into a new trunk submodule +Initial proposal for pluggable application context determination + Use directlyDan is this sufficient for your needs +Lets also make changes to add namespace scope for C to +Any chance of a review of this approach? Thanks lads +defer all issues to + for the to trunk. +I have attached an example file that demonstrates the described behavior when extracted +This has been corrected +We will limit transmission speed and compress transmission compression to resolve the problem but not to change format of the fsimage +I was told that it was OK to copy from the spec doc +Reran the test case and verified that it works +Could you please recheck with the latest version of the script and +I thought this was what +its pretty deep Seems to occur when the xml text is loaded into a DOM tree +I think this can be related to +Norman you reopened the wrong bug is the one still being reproducible +This wont be in +This is because in the file of the Tutorials module we haveimagemosaicplugin ofimagemosaicplugin attach the new patch for the file of the tutorials module as I stated in previous comment +Though I still have on IE but thats a different bug +This is something we patch jbossws for during ESB install on +May I close the bug for you? +Assume its OK +Closing issue Oozie is released +The motivation for this issue is avoiding misconfiguration of users upgraded from previous versions who is tending to set HADOOPIDENTSTRING in +So +Can you please attach agent logs to investigate further +Moving to +So close this issue. +Added areas for links to relevant docs +Applied patches in r +We have more mirrors and everything is registered with current address +The attached patch weve been using in Nexus for months +so I assume were waiting for an update to confirm this issue is fixed +New patch also looks good to me +As features use from nmr parent so this issue already get fixed by checked on a local build Karaf is used and starts correctly. +Thus we can limit the to the resource in question via the on the special queue for the resource +I test on current SVN and it works for me can you please confirm meI am attaching both client and server code for ur refernce +patch should be against solr trunk +Looks like was applied after all +Moved to future completion +Patch committed +If it is hard to read we can just go with explicit selection based on a counter although IMO that is less clean +merged in +Verified on and also in svn. +This jira was logged in January and we havent gotten additional info from the reporter Should this be marked closed for now? It can be reopened when we have more information available. +For each field on which we want to make a facet I get the field content from the lucene document in the search results and I check if the Map key is equals to the field content if no I add it as a key in the Hashmap else I get the Map content wish is an Integer and I increment it +Thanks for the pull request! This has been merged to master and also for the upcoming release. +to commitI suppose this fix should be backported toRegarding Jason and the developer list I can add him if he requests to be added +BayesCbayes job now uses the String Tuple Writable +I wish I fully understood the scope solutions and so on +Nick Max Inc +Changed Target Version to upon release of +Turned off blooms on +Ive been working with computers and software since before Al Gore invented the internet if on the other hand you feel compelled to vent because you dont agree with me well then by all means go for it +applied patch in rcreated new issue for the injection strategy question +It actually first appears on +committed. +I would rather see it done correctly all the way instead of just shutting down the node manager +Committed to both trunk and branch. +To run my app I do grails +I hope it was clear in the description but to get the memory spike you only need to create one of these objects not actually use them for any work +youre right thanks +are now managed by the container and not through Teiid. +This version service invoke the version service in Axis +Comment was not for you it was for the patch authorbq +I have reproduced this problem +Which one is the correct result +Update Make test case pass if GBK charset is not available +Might be good toupdate the patch comment with this new info so it can beremoved ifwhen this is fixed in +This is fixed in the latest release. +This is a nice catch Diogo and the patch looks good to me +Integrated to and trunkThanks for the reviews Jon Matteo Lars and Andy +thanks patch applied +Patch for the patch to reflect the changes to due to is already committed to and should go into a separate jiraCreated to commit the patch. +Lots of random reads +Looks like theres an update that could haveshould have been merged +Thanks a lot. +Fixed with with recent automation runs +I think it is worthwhile creating a jira for it +Thanks NathanThe fix looks fine +Those upgrading from will need to change the interface anyway +Its just a pluggable code that should decide on what goes next +Ill have a look for alternatives but as far as I know the only other way would be to lie about the axis orientation and just slam in a value that does not represent reality +IMHO adding a mojo called is the best option here +Arghh +mingzym you see any problems removing this related to clustering? Or anyone has any concerns applying this code please voice it now +rb file out of a jar +Documenting what the patch does +It was released with the fix in +Ive committed the test case to trunk with revision +Currently deployer script only generates and copies data to hdfs it needs to be enhanced wherein it can do creation of tables in metastore also +What makes you belive they should be there? The AS build does not install those anywhere actually asm is not even a dependency of the AS +That might argue for making a more general interface instead of being a member ofI thought about that and decided to just keep it in until we discover the need for it elsewhere +I know that we may potentially be acquiring the same lock twice but these are locks so I took the point of view that the code was simpler if we just retook the lock rather than adding more conditional logic +We wouldnt implement that in and if youre upgrading to you get as Chris mentioned which is a better fix +I also added the synchronized because it was strange to have all other public methods synchronized except this one but I didnt check if some test case were actually calling them in +Not sure how to reproduce +thanks +What is the error that you get on the server? Can you look in +Would you like to format this as a patch and submit for review +r +It seems strange to have to generate the server code to get all of the schema properly mapping through adb but maybe its just my axis naivetyIm using Axis +Michael any chance you could test and then commit your unit test if all looks fine for you tooExcellent thanks Nick! Ill test commit +I tend to like Or because for a given you cannot use bothOK go for the patch that way we will make a small warn in the javadoc about a potential change +This issue has been fixed and released as part of release +Release Notes text deleted as per latest comments that this is not relevant to EAP +Thanks Carlos I missed that part. +The errors will get logged to +no updated patch I will change issue status to WONTFIX and close itIf you would like to work on this issue please note soBest attach a new patch and reopen this issueIf there is any question please be welcome to ask either on the list or myself directlyBest regards issue +Also it looks like this approach should result in a shorter KML file which should help display large datasets from that are refreshed based on the viewpoint +closing stale resolved issues +Please disregard previous comment +What if we just specified the jar locations statically? Given the location of JBOSSHOME we know exactly where these jars are and even what they contain +it only occured in the first time if you click the node Beans Support +The additional weakness only comes in when the safe is opened +Please also see related issue. +I agree with Rohinis opinion completely +It appears Suzan was correct +To be reviewed to see if feasible in timeframe +Oh looks like is now used for app state queries +This is moot since weve jettisoned storage drivers in. +GaryI am a very very very happy here +The patch didnt so I made the change manually +committed to trunk resolving revision +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +But youve convinced me +Fixed +extarct solr to solrbase +Just to compare with EE deployments +The suggested form for contributions is through a patch though +Perhaps this was resolved by another ticket on counters? Can you still reproduce +Is someone currently working on the architectural overview outlined in this issue? As part of preparing for the final release we definitely need to consolidate documentation +Its pretty obvious to me that we should not expect to find any automated solution using SYST to the date problem +bulk close of resolved issues. +I use the same project as you attached with +Thanks Amareshwari! +I think a good resolution would be to validate the configuration against such case throw a meaningfull error when indexing is enabled on an invalidating cache +Attached patch a with a small set of regression tests one of them demonstrates the bugThe tests are added toI also documented a utility method in the test class +ChunhuiVery good catch +BTW all other calls to first call that is responsible for initialising the local +Instead of an empty map I get no map at all a null pointer! Ill post the in case theres something else stupid about this + The patch fixes the issue +We also want to allow multiple clients to append to a wide row so the write lock design probably doesnt work for us either +It is not worth pursuing a fix for releases. +Theres a good reason its protected at the moment changing the is not something that wed encourage userapplication code to do in an uncontrolled manner for a number of reasons including thread safety and understanding of precisely when such a change might take affect within the lifecycle of a transaction +No messages are delivered anymore +It seems we forgot this issue for +Will commit soon +Not actually a bug per se but the error message is not exacly the most helpful but I am not sure whether it is worth chasing just now + patch commited to fix applied to + +I looked at the additional section and it looks really good +Fixed in r +I Decided it was worth it when I was fixing my th typo in the beans config +After some testing I notice null in javascript is an object so typeof ret returns object + +I suppose the author means a click application +Thrift build the problem is that by default sudo discards the environment variable ANTHOMEsudo ANTHOMEANTHOME make install resolves the issuePerhaps the generated Makefile should include an ANTHOME definition I believe that would avoid the issue. +thanks Alex for digging into it I think youre right and Ill have a look as soon as possible +If any problem is coming with large jobs they can use this feature to find out the problem in the user code +Werner I just noticed your comment which came in after mine about the new test case +It sets collation in for string types to that of the current schema +I didnt say there should not be a useful message +Hi SebbI havent seen any activity on this issue in a whileAre there further improvements or changes you would like me to make or is the patch ready to be committedThanksMirko +Its not in the releases. +Tested +Works for meThank you Tim +Thank you +To be also have this problem ini cant get started with ACIi defined prescriptiveACI as the official manual demo one +i think i have fixed it all introduced a unset on the class and called it in all the places i could think offAte its part of the Portlet stuff so could you give it a quick look an then close this issue +Arun I think you forgot to svn add the new files +This is the result picture +No Apache member should believe thats the professional way to protect the legal rights and obligations of The Apache Software Foundation +but this one is still resolved. +Adding specific targets treatment for ESB editor is a matter of a few lines in +Patches are welcome though I am not sure if embedder is currently support this +Brandon should we add Yarn to the list as well +Verified Diagnostics now tries to create the folder if it doesnt exist +See my latest comment in why I am now in favour of adding this attribute to the artifact element +Gerhard no Singleton is defined by the Inject spec not CDI at all +we were discussing something pretty similar to the patch +Is this naked Windows or using Cygwin or MSYS? This is important since it is a question of whether it is the DOS batch file or the Posix shell had a very similar problem some months ago with the DOS batch script +closing as a dup of of +Out of curiosity is there something blocking this issue? Should I ask jetbrains to bless this issue? +The slow dist build is bad but it does not look likely we can find a way to speed it up anytime soon +I was planning to look into it myself at one point but it is not my top priority right now +Thanks +Included in. +Or do you see duplicate entries in vmreservation table for same uuid +sounds like a good improvement I will give it a closer look tomorrow +Not sure what the best course of action is yet Ill ask the list and see if we can judge the JME compliance rather quickly +As I can see this is not at the system shutdown this has happen when processing a reqThanks I dont know why but Im sure it happens at system shutdown +I guess our datanodes arent that busy then +Cant reproduce it either +Do you want to commit that patch +For example if you know your usecase does NOT include Mac you may want to pass in BYTEWIN BYTELINUX or if you want to validate against all BYTEALL BYTEWIN BYTEMAC BYTELINUX +Doug feels like process has been violated by this patch so Ill roll it back + patch after all I had to do was throw a new from and let ipc retry deal with it +Cool Jacques thank youHave a good day. +But as I said more a matter of taster your changes are perfectly ok I think +Faster version of with minimal added some improvements in revision based on Philips proposalThanks for the contribution +Does that example make sense +This is the forrest site for a comment on the content there is a reference to a team member from Rumania if this refers to Vassil then the country should be Bulgaria +log file and here it isSESSION Apple constants OSmacosx ARCHx WScarbon NLen arguments keyring Usersthebugslayer +Scan copy constructor doing the same way +This bug has been marked as a duplicate of +This has been committed so Im closing as fixed. +can be unclear to some people so it might be better to say as specified by instead? Typo Change Within routines the execute to Within routines that execute +Thanks for the patch and test it all looks good and I have applied it +The bpel will contain a reference to the partner link type whose name is +Any followup work needed here for +I believe the Priority of this should be bumped up significantly from Minor as it affects a large number of developers +Patch for TRUNK has been integratedOriginal patch integrated to branchThanks for the patch Jinchao +There is an +We were locking before so were just adding back the lock +Hi JodyI just applied the patch locally +Also do you have any benchmarks for how much this improves things +Moving to logcxx which is an apache project +Note that is only meaningful if the corresponding JSP pages are also included in the WAR +value is now returned correctly +Possibly a logging based approach might be a better way to go +That copy is used to expose the data to jmx and other consumersThere is no remove function +Please note the patch in is buggy and cannot be used +I suppose that makes sense since doesnt have to do with parsing +Neither Jenkins nor my machines hit any compilation errors +Closing + + +Im reviewing this one +Well just have to agree to disagree then +Tests too +Some tests are failing for patch fixes all test failures +Please can you attach the Java code that you used to generate the failing WSDL +The query abort code is just changing the timing +Local pig is mostly for coming up to speed on the system not to run large scale processes +Could you provide some performance numbers to quantify the before and after improvement +If you have A depends on B and B depends on C and C depends on Athen we had excludes that did not work +Thank works fine +use a object only when initializing theWe shouldnt need to load within the app master itself since it wouldve been loaded in the client but Id prefer being certain of this +The project is a useful stopgap until we get regular releases going again and it allows other projects like to use Snappy in the meantime +or or orThe lexical analyzer in XSLTC is using the same token to represent an asterisk that is used as a wildcard as it does to represent an asterisk that is used as a multiplication operator +Add a check that the driver directory exists +This seems consistent with the JNDI docs but inconsistent with the usage in the rest of the class +Happens using Hibernate ehcache as well +You can drag and drop a number of user tasks on the embedded sub process so theres no real need for this functionality +If you happen to have examples of the SQL being executed could you send them to me? If not let me know and Ill dig through the code to piece it together and run the different SQL calls on both SQL and +I think +Please comment on what testing youve done with this patch and what the impact of it is +So is this actually resolvedI added true to the vm properties but my pages are not being compressed +They are not for this issue but for a different one. +attaching a patch for this issue +Proposed patch +Ahh I see the trouble +shouldnt you test using in the test +Design doc on doing hash joins using the tez framework +Could you try the latest snapshot? The issue should have been resolved with +With the default values this happens after segments exist + +Can you just modify that to mean that the is failing? This is another general suggestion File tickets describing the problem and then propose the solution as a commentbq +The Tokens are not +Closing as these have been committed. +latest patch looks good I tested it against a tarball install +I put a build from HEAD into my serverdefaultlib overwriting the original one and I now see occur +Also patch which includes antlr in the template +An interceptor can take care of this too +You can ask for an upgrade on the ml though +Right now it is geared to publish whatever is in testartifacts +rm CM and +It would be nice to have a error if the user makes a mistake like a space after or or writing field you may grab it +Partial thread dump including the two threads responsible of the ve also managed to reproduce the bug +Sandy thank you for help with code reviews +Attached the output +Thx Michael +Thanks Kenji! And congratulations on your first patch! +This happens before the Camel Context is fully started +should allow to manipulate the context and more. +Ill port later this week. +I also removed the old file as it doesnt appear to be used anywhere else +The logging message shows the class name as with a lower case l in taglib +Had to modify the patch a bit as the client may not be created yet at that point but basic idea applied +I was able to reproduce it with Tomcat but NOT with any lower +Updated the outputs for unit testsFixed the error message to include the line number and position in the query of the invalid column +JBDS tends to work with the projects by default +I would target this issue only in the reimplementation of the RFA +Its already on phabricator by Namit +Oops the updated didnt handle features with the same name in different repositories +This change has been rolled back as it caused several our test suite to start failing +The classwhichhas been added as part of is also present in this patch which as indicated by a class level comment in this patch is a dummy implementation which was to be removed after the commit of +Are we also going to ship with an impl that uses? Or is this something that is going to be impld and shipped with AS +Attached is my file which show the secinaro where old request object is required to display the values back to if validations fails +JozefDid you want me to do this one or are you planning to take care of it +please ignore. +But if you immediately shut down the ZK before snapshotting again you will lose any record of these transactions on the ZK in question +But Hive stil unable to execute query above +Are we waiting for more code review before this goes into the tree +I agree with Bikas and Arun to a point +Released so closed. +So I take my proposed fix is a real fix? I need a local working version and will do the change manually until is out +Note this is work in progress +All cases behave identically with would you please build a metamer test incorporating the provided scenario code snippets? +If I would start off with a new database design I probably will not do this again +So it seems like the problem is deeper than just not being able to run any scripts +Im closing this there will always be more screencasts that could be added and Im satisfied that weve made a start. +So the good news is I dont think its incorrect but the bad news is we broke +Agreed looks like a bug +Fixed by Ben in r +IIRC the reason this new behavior wasnt simply enabled in regardless of persistence version is that it caused issues when running some JPA TCK variations +support generic +So we will be able to drive some improvement by our own requirements +Remove the role account cure the problemLet me know if you wish me to bulk and then remove this user or close this issue if you will solve this yourself +Its probably reasonable to assume in this case that the one on the test classpath should override the one on the main classpath for classes on the test classpath +Christian Can you generate a set of task files for the reduce with the failure with set to true so that we can run it in the isolation runner and a profiler +I added the and tools folders to the etcpaths file and I can execute any of the Android commands from the terminal successfully +Just this morning +was committed to recently +This should be hooked into the that gets configured when the use flag is turned on +Merged PR. +Heres another performance test +Looks ok to me. +If you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +I dont know whether its a bug or a new feature +However using and attached configuration you can still reproduce it +Now that work is well underway with AS all previous community releases are +Looks like the python stuff is now calculating them on the fly +Thanks +resolveInterceptors should resolve transitive interceptors if that is what you are asking + assumes that there are RS available and adds other afterwards +get in the way of solving problems that DOI would prefer to keep it open +Moving items not being worked on afaik out of M Fix Arnauds comments Im not sure that they dont miss the key point of writing to disk +on caches retrieval +batch transition to closed for remaining resolved bugs +I will do that. +While running this suite I need to pass DjvmflagsmxM msM so that the values mxM msM are passed while running individual test cases and while starting network server in case of framework +Marking for though I dont have a solution at hand at the for +We need to sort this out with logjIm therefore changing the issue type to Task and reducing the priority to minorRegards issue out as part of JIRA cleanup +Please use your original WSDL and try latest nightly CVSthanks in CVS fix in todays daily build +This appears again in EAP +Before applying this patch lets make sure that this definately doesnt go against the SOAP or SOAP spec +hmm +Hmm so this can happen +Thank you Dmitriy! I will commit it today +I checked out the latest source and ran a test and it looks like the issue is ok but with the latest code base I am seeing new compile first issue is what appears to be reported as and from there it goes down hill +On that metamer page table sorting works fine for me +Were now using re instead of regexp which means that or higher is a requirement +So the answer to your first question is no this is not the same there is no way to add invidivial entriesI dont quite understand your second bullet +Can someone verify this behavior +Hi ZhanweiGood catch +In the mean times they are all included in could be mergeddiscussed with +should not fail unless all of the instances have failed +Now in this patch changes is merged +Thanks Espen! +quickstart +Ill be temporary excluding the jaxbri module from the javadocs till we get this resolved +Im using the source code without problem on windows +Looks superb +When you just assert retract and reassert it works fine +On the otherhand the Synchronization will only allow at most one thread to execute the code at any particular time +Moreover the default severity of this rule should be blocker +Guys what about adding one in the trunk for as wellThanks Julien Ill get it committed deemed suitable I could commit the trunk file to svn against trunk directory +Where do you think the rules should be processed +if dragged and dropped everything works fine +this is the default implementation since r +Patch committed to the trunk +In the future whenever it is to judge whether the hostname is of any location is should be called +The RFA rewrite should fix this issue +Instead we should look at +I think we should to maintain consistency across all the property files +Patrick and I talked about having a simplified more way to accomplish this +Ill try to separate the patches for each half of this idea +corpus file looks fine +Committed with revision . +Thanks to John George. +hmm would it help if we set the server dir as the current working directoryWhat is the value today? About the specific option how portable is this between various version of ASIs it the exact same jar for every version etc +Im resolving for since this was fixed upstream already an will be caught by the rebase +Uploaded which included the document validator fix in the original patch plus updates to the assertion number and message in +What if you could just say reload the file? Tying it to the commit operation is the core problemAs it sounds reasonable to me I wrote the attached patch to reset readerCache in +I have a fix for this but and its thrown up a few more issues so Im just investigating a little more before I check this in +updated namespaces but tests are modelemf refuses to openResource emf does not exsist +All patches need to be granted apache license on the upload page before they can be added to the codebase +Please provide a test case that we can use to reproduce the issue +Having applicaitons automagically update sql schemas or push avro schemas can lead to accidents unless the security model is implemented properly. +A game which uses the data and which runs on a handsetA javademotearoffstoolkitA toolkit for declaring parameterized subscriptions of data +I will give you and update and leave a comment if i have any questionsThankspulasthi April Suresh Thanks I am looking forward for the opportunity to contribute to such an interesting project +We will think about creating our own TP +Hi Yip thanks for the review! My derbyall and runs were cleanso Ive committed to the trunk as revision I intend to merge this change to and commit it there as well +patch Erik for providing the patch +This however is inconsistent with the rest of the software where text file logging is used +my comment When I said I was just trying to hedge my bets +Before I apply the patch which I am sure is a valid solution I want to verify the behavior you have reported +Whether the Firewall plugin is required is another question entirely +Patch to insert the IRC notification config into the plexus config file +I will create a sub issue to fix this and take care of it The Archetype Catalog doesnt play nice +Hey StackExcellent feedback thanks +From looking at the code and testing your submission with testXSLT it looks like the exception thrown should be a not +MladenPlease take adev is sparc not xAre you sure you are using the correct package +Instead I moved the libhdfs code by doing a svn move +should we put them all in the operation metadata or let it as when I say in my mind +Reclosing the issue after changing the fix version from to. +Steve please take a look thanks +Passes test on trunk and brnach. +Changes look good +Sam My argument is that concludes that global changes feed in not the right way to think about this problem and that a handler the dbchahges are the right way to go +Suggest new name for unDetached variable +Dependency added Sending modulesassemblysrcplanTransmitting file dataCommitted revision +Was a regression +Hi Sergey I understand the problem its trying to address but I think the only sensible solution is to do what I did with email and make other constraints ignore blank and null values +for the patch +As the issue reporter please confirm so that we can close this issue. +suggestion from frederikalter table add column col +Otherwise the is not registered automatically +Completions on the DGM +Can we generalize this to dont bother doing validation compactions for RF +see comments above +In addition isnt RPC supposed to use persistent connections? From the stats above seem to using new connection for every RPC call +Attribute Started +Sure Im working on that now. +updated to see change current changes +I have to think that the added transmission time of all the redundant data in aggregate is at least as much as the added time it would take to do translation and possibly more +Committed to trunk +wouter is this still happening with the current trunk +Submitted new PR to change the name of the config option +Need ability to manually t really apply for JBM jbm already has extensive management features for copying messages between queues +LGTM +At the very least we should put this note into the headers of all the source files +Kjell any news on this issueThanks +AndyI think this also needs a change in and set +Probably it is not a problem for most of the users but this should be documented as an additional requirement for the plugin +Colm. +I have in the past done tests under load that should reproduce what you are seeing but I havent encountered such a problem +lets not add extra logic to the for finding out if requester is authorized Instead we could quite easily derive this info using the +This has been already reported and fixed +Plugin removed +Each single deploy comes with a runtime cost +Its still pretty alpha but Click core and extra tests do run successfullyThere are also a few new tests under the mock package trunkmocktest that can be used as examplesRicardo if you have time can you take a look and see what is missing for testing PagesWe also have to look at how to package the mock +Resolving this JIRA +Thanks! +I have recently made some modifications to the xsdbaseBinary and xsdhexBinary classes to resolve memory leaks within the client can you try the latest code to determine if this has also resolved the problem for the server +Gotcha +doesnt directly depends form I mean that in manifest doesnt configured that its depend from dont think that its could explain such I unpack eclipse and put jdbs into dropins folder its was working fine for me +This does not happen every however happens when I trying to merge a particular package +adds filename argument to acceptFile method and adds allowedExtensions adds filename argument to our app I basically made a copy of similar to the one included in to add this behaviourThe other option would have been to check for the extension in the action itself after bindingI attached patches merely adds an filename argument to the acceptFile method +The correct Avro type for such fields would be a union of the subclasses but the reflect API has no way of knowing that +Looks good otherwise much easier to read +Its part of the site +May they have improved the ScannerI will skip this test if we run on Java +Adding documentation for debugging +And even worse it does it in the phase +If we change the spec of occurrenceAnnotation in show these as squiggly instead of underline the problem goes away for some reason +I believe Ive also backported to the branch +Can you create a quickstart that demonstrates the problem? That will tremendously help the developers in having the time to fix this +still needs the parser but this is a trivial addition +I did an even simpler one and was able to reproduce this +Can we have an update on when this bug will be fixed +Regarding hz vs +Tagging for for now please add or a later if and when this fix gets backported +If you click the icon you get more info +I apologize I seem to have made a mistake here +fixing only in because we do not want to change existing behavior +Possible loss of precision +I looked for it in js libraries but i couldt found that. +Sorry about that +FYI Ive created to fix the Squid bug in Sonar +In the GSS meeting last Friday it was decided that we would let the email notifications run their course accordingly without changing the systems behavior +Perhaps itd need to be based on the iterative index size changes? Ie if the index has grown by since the last save? Id like to design the recovery logic now +I wonder if anyone else in the community thinks strongly one way or the other +fixed in trunk +one tiny change +Its a very simple style attribute +Im willing to know if the issue was accepted or rejected as our development will react consequently to this choice eventually applying an alternate solution replicating some of your code in our framework to circumvent the method visibility problemPlease let me know what your decision will be as in case Ill not receive an answer we will be forced to apply the secondo solution not the best one I can imagine +I understand what you mean but its not necessary to break the partitioning it could be handled as mutationsresolutions at the barrier of synchronization before vertices are exchanged +I dont understand why the change proposed here doesnt match in this Because the proposed change would create an ambiguity as to what exactly constitutes a cookie value +This should make Plexus super fast as wont do the entire discoverty which can take some time. + to Eddie +In EAP theres a proper dialect for MSSQL in Hibernate so this workaround is not needed anymore +closing issue. +Scott Im not sure what the final build will contain +The viewName property now resolves to the view name. +should be nonblocking as the old version is like this +Verified in Developer StudioVersion Build id Build date +After updating code to it now compiles without a problem +As you said it working the problem is with Jboss App Server GA Deployment Ordering +That is the way the Servlt spec works and one of the reasons chaining actions as you are doing is generally a bad idea +bq +bq +Committed to trunk +bq +Test case that cause the server error +Please verify and close +It was released with the fix in +Committed to issue as it has been released as part of release +patch adds Key attribute discussed in just committed this +Karthik +That output might help us nail this down +And a private getter makes a property +regards already confirmed that is misbehaving here but considering that the was not designed for generating XML output in the first place I cannot agree this beeing a Blocker issue +GregAre you asking for the ability to use full sort spec as the collapse criteria? I believe you are but I just want to clarifyYou can currently use the full sort spec now to sort the collasped result set +Following Franks suggestion Ive modified the INTERNAL prefix to be INTERNAL +no issue there that needs a lot of improvement +Dan do you have another recommendation for removing these references so we can drop the table +Pull back if you feel differently +Could you follow up with another patch to get this working +I dont know +Upgrading to a blocker for Hadoop +Line added back in Git ID a patch be done to roo +Finished +My guess is users will really be concerned about the stats for their queue +Is this version information or object information +Hi RoyI think a better solution would be to use an ordered map +The five patches in the zip apply to the following classesfiles javascakernelspisrcmainjavaorgapachetuscanyspicomponent javascakernelspisrcmainjavaorgapachetuscanyspiextension javascakernelspisrcmainjavaorgapachetuscanyspideployer javascakernelcoresrcmainjavaorgapachetuscanycoredeployer javascakernelcoresrcmainjavaorgapachetuscanycorelauncherm going to start looking into this patch +The next one will be and is expected to be out on October +Are there any following steps after rebuild the jansi native? Thanks in advance +Thanks for your feedback about the Java compatibility +Sorry should have done that first +Ran Notifications and GK overnight and was able to duplicate the failure +This patch should apply to both trunk and branch +Thanks Jervis +Im resolving it since you said you reported it separately forCommitted thanks for the patch anyway + on trunkPossible to switch to tunk and verify +I have left these tests as ignored and the restI could not get a hang over runs with both and +Thanks +Trying and asked Kabir to test it too +It points to the position after the delimiter if any found +The main issue is how to handle multiple definitions from different interfaces inherited into a single class the multiple inheritance problem +Consequently the only way the logs will get cleaned up is if the region server shuts down cleanly the region server crashes and the master splits the logs for itit appears that what we need to do is when a region splits we need to remove it from the map of region names last sequence number +Obviously part of what Maven helps with is finding where you are using inconsistent versions of libraries +I believe you are proposing the client always sends two messages upon connect the connect header and a SASL INITIATE +Fixed by logging the mismatch at debug level and returning null from the mapRequest method to let another mapper try fixed. +This issue is out of date and will not be fixed +I guess using test works +My position is that it makes more sense to use +Yea Im aware JMX is a superset in functionality compared to Metrics in that its notBut metrics is a superset to JMX in that it already has a bunch of different pluggable output libraries eg Ganglia +Fixed upstream +There are differences between them +VenketashYou can just use git diff +Release +Let me look at this for +Stack brought this to my attention late last night only +And the simplest solution would be to specify that all types should have a Type suffix appended at generation +So care must be taken that a predictable naming strategy is used for our scanners and that such changes are carefully tracked and maintained by the tools +Please reopen with concise instructions as to how to reproduce if this bug still exists +They should be moved to the utility commands and they are not depricated +trunkorb now contains a module that builds with maven in which all of Orbacus is in the package +Fixed compilation problem on windows in latest patch +Let me know if you spot anything else in there +We are getting rid of the multiple classes +In my SPI my propertyDefs were not being populatedFixing this bug is not a pressing issue for me now and could probably be a low priority since it is now only a matter of defensive programming and preventing bad data from crashing the JVM and also an issue of the stack overflow obscuring and complicating the problem solving processThx again Angela and Julian +Just make the constructor which doesnt take in a appType to call the one which takes in appType + +I didnt see any counterexample for not getting rid of it +is an incompatible change it will change the output of jobs +Pull back if you feel otherwise +Btw I still have not found any time to step through your problem +Initial commit in Git ID bfdbcdcdfbdadcd for dependency adding changed in Git ID eefccccacd +Im concerned we do not understand the others enough to design the interfaces well enoughThats a risk +Removes and from the exclude listThis patch applies successfully after the patch in +So with every restart you get a new entry and actually I think you also have those services configured and running +problem with the use a firewall solution is that sometimes firewalls fail +Cassandra support is seeing increasing demand +New patch removed rmaddress as a parameter from the above test failures is not related to this patch is also failing on trunk other test failures are related to which should be already fixed by to trunk and +We accept patches you can dive in and implement this and attach the contribution to this issue +But I dont know how to submit a patch +We are relying on file renaming being an atomic operation +Yes this has been corrected in recent versions of the Camel docs. +ThanksGreg +commons dos no longer depend on other Stanbol modules +We wont find the same group in a different segment +Ill take a better look when i have time but if anyone else would like to do it i wont fell bad +So the with exception of overlayable would have hurt +This means that and fail due to ordering constraints that are not specified in the grouping +Just one minor comment +just fix it +Alexthe patch was applied in revision missing copyrights were added to the new filesplease let me know if its OK for its OK +From our point of view it passes all possible +Unless a new issue is found I am closing this as complete. +I will try the next nightly build +When I released the oracle wrapper the release plugin didnt work with svn +both applied +Using require in Utils looks good to me +Corey Nolet was having the same problem for a while +Tested ok +Fixes and the test case added on trunk +I tested Linotype with Mozilla on Linux Fedora Core and its works +Other noteFrom our experience so far application that use local session will see higher rate of session expire +the Psecurity profile so I cant add the run all tests profile to these it seems +Committed for in rev +Resolving as such +It should be stripped from old api stuff and is not even part of the equation of the new apiThis does fix the original bug +Does anyone have any opinions on where this ticket is going? Also as far as I can see the mulithreaded issue should now be resolved as Alexis commit still stands strong +Waiting on your feedback Keith +The attached document introduces token definition interfaces and low level design. +bq +Please verify that you are still see the issue in the current co r Branch buildoutput assume this is fixed in do I get is it released it +testfilesystem failure I saw earlier was because of my config +Ah I missed those good catch +attaching the +Communication with the endpoint brings out the bug that this report is dedicated to +Thanks Ed! +thanks +postponed to CR since we to since this is a +and sorry for the off topic think we can now commit the patch and close this bug since the test was successfull +If a client attempts to connect to the slave database before startSlave has been invoked the slave db will be booted and get out of sync with the master +Someone needs to spend a little bit time with this before we lower the criticality +Would be possible in the docs to make it more explicit that the behavior is an OR of all the from the root to the leaf queue +Would like to get that in as subtask on this issueI looked at what else is involved converting +Attachment has been added with description XML has been added with description XSL is a document and an external image missing +Thanks +Isnt that concurrency option +One is working for the more deployment the other just for the OSGISince we could tell if the camel context is for the OSGI environment the camel context can delegate the creation of theI just checked the s construction caller code they all can access the camel context +OK but can you create a mock on the fly that emulates a file of a specified size such as a GB file +Ive realized that this problem occurs only with dynamic endpoints used in a proxy service +Any other ideas +This has been applied to trunk in r. +Would you please try this new oneBest would any commiter please change the component to classlib +It might make a good demo app though would you consider submitting it as such? We could add it to the demos project +Yes I am using CamelThe snapshot is working fine +This has been fixed already we can probably resolve the issue +The problem is that there is no stack trace +Closing issues for there be added any doc to the reference manual for this? There doesnt seem to exist any. +I committed your patch + add static methods +Patch to fix deprecationsThis cannot be applied to +Furthermore it took about minutes to define a Hibernate Console configuration for +will cover most cases +Agreed +Ill see if I can get this fixed for +Hyperheuristics is a selector thing and independent of acceptors +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +The accurate number does not need to be exact +Fixed comments where it talked about the force paramter instead of the new majorCompaction parameter +Ok Ill take a look again +Good catch. +I want the ability to filter threads by version of seam ie if I know my problem has something to do with version I dont want any threads about version in my search +Committed for and trunk +So all issues not directly assigned to an AS release are being closed +Lowering the priority +However this is a recurring issue for lots of components and the patch could fix it forever +At some point I added some code to try to avoid the jdbc classloader weirdness +likely one of the newer files in the patch is missing the Apache License header +This is a working example of what I am exploring +Working through this test case may help to provide the real culprit that is responsible for the introduction ofPs Benson I dont mean to hijack this bug report I just want to focus the issue +Can you close the bug if my guessis correct or give some more hints? using the export CATALINAOPTS +Ah my bad +Fixed for upcoming release. +lib +Profiling should be done using jvm toolkits like yourkit jvisualvm etc +Ill take care of this +Thanks again Hamed +corrected for How to highlight rows in a dataTable when the mouse is over? is done. +Heres the patch +This issue is not technically about the raise but the discrepancy between pl with MRI +All right lets just run with part for now +ooooh +Feel free not to apply that change ifwhen committing +Attached patch makes the following changes Adds clear to and the implementations of +I already did +Retry to see if the unit test is because of this patch +No need for IT +I tested this case with trunk after fix and freeze no longer occursPS +Verified +Added as part of authorization example. +Fixed the test case to create the statement CLOSECURSORSATCOMMIT +Specifically there are some issues with the IN operator in the jdo stuff in this patch. +bq +It can work with the legacy code with no change +The same project I used yesterday works with VS +But would you add this feature to the project in the future releasesby the way I would like to congradulate you guys for this project because its really easy to start working with today is my second day and I am seriously considering to build my project on Click for user interface parts +Closing it +My guess would be too +Closing all resolved tickets from or older +V patch is attached which fix the unit tests +An output list is maintained along with the sorted list. +committed you issues for released version +The attached patch integrate username token and timestamp addition to a soap message accoarding to the security policy +The integer value for both of them is now +OK then eithera It is a parse errorOrb is equivalent with except if it is the first in a setAfter some thinking I say that b is better +Duplicate of. +So did this get completed for? If so whats the on master? If not can we move it to +The default build definition already has the the version since nothing was agree the actual behavior you describe seems wrong +Closing all resolved tickets from or older +Change bars would be cool if supported but that is icing +Ok I see the problem +Simple to reproduce just keep hitting the server with or some tool +Lets target the version for this. +To put this into terms that many might have an easier time understanding KSSL is roughly equivalent to WEP in terms of its vulnerabilityId also like to point out what our spread looks like and lower insecure only so irrelevant through only had KSSLhftp and up is availableSo were looking at a window of releases of about months +Can this go in for MR +Doesnt still stand which requires SWIG ? While it may not leak memory it corrupts messages containing +I made a build from the cvs sources and the bug really seems to be fixed This bug has been marked as a duplicate of +The signal event was registering a listener that was not Serializable resulting in this exception. +If you could supply a patch that would be great +commentable PDF of deployment guide +The attached patch shows that changing referenced by works +The client does not know the new server but the new server should be able to tell if the client is allowed to talk to the new server or not +I was able to reproduce the problem with the given details +Why is this necessary? +This will help admins +Ouch really? MB? Most people will use ATS in reverse proxy mode why do we need MB hostDB? This makes no sense please explain andor revert +I have few more questions +If I change it to use the standard LRU implementation all tests pass +bq +I have updated the description to say that the jar files are NOT split +But isnt Find Seam not only relevant for a very limited set of editors? So only show it in that specific case instead of only hiding them for Java editors +My exact steps were +Moved to can easily be addressed by any application by its lowered to module file format is preferable +Is this a Drools project issue or a BRMS product issue? I believe if for BRMS you need to raise issues on Bugzilla now +Thanks +After solving this failure there are still four testcase failures for +Dominik +v adds an onRemove callback to the gossiper interface to do the remove from +An example portlet with can be used to recreate the problem +Thank you for the comments Knut +Fix the remaining unit test error under Daniel +However as in my previous comment you can avoid broker shutdown a good solution is to set a queue limit policy on youre queues with a limit that is lower than the size of your store +This issue is closed now +PROPMAXQUEUESIZE PROPMAXQUEUECOUNT queue depth on queue creation via ACL +Jason if you can fix this easily fine +Ill let it run overnight and see what it looks like in the morningIts easy to avoid this problem by modifying the query but so far Ive only avoided the need to time out +It needs at least one method +Believe we have fixed this close it +Started with no +this jira for cloudstack hasnt been done yet you guys ready for it +Please let me know if there are any problems to incorporate this feature +I downloaded the and applied it the way the README describes. + is already adressed with a patch in is already adressed in Add URL where? you mean next to each app repeat the url? you app does not necessarily have just one URL and is Open With not enough here? The Openshift Console has no notion about when you last deployed to the application not unless has that available in their API window is not connected to the server nor the project in eclipse that is why the eclipse project and server adapter exist to create such connection +bq +Will attach a new PR and push it in +Joins are only supported on JDBC databases there is no support for any other store type +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Guess I jumped the gun a little bit taking it on faith that our pom was configured for the latest groovy compiler +Can we move up to and ditch testng +Raised to version +The test is now fixed by saving all synthetic methods and all setter methods and verifying in the end that all of them were received from IWC too +for once we would have to use JPA and its annotations and it would have to be a java class instead of say entitiesSo Id leave this one open but in low keyOf course if anyone out there feels like he must solve that issue to save the world we would oblige and help him in his contribution +Resolving this issue for now waiting for feedback from the intiial reporting user +Ive added a delete button in each specific database +The first part of the warning is spot on you are pointing to a JRE thus you should be warned since it most likely will not launch EAP SOA for youThe message and detection in itself have been changed in and requested initially inThis same message should also occur in Tools if you pointed it to an JRE instead of a JDKMarking this as resolved since the message should now be better +What version of Maven is this targeted for inclusion? +Brian Jessica can you review this issue +A bit cleaner sample of the problem + add avro ivy settings in sqoop and locally +closing issue. +it works this is temporary solution which suitable for HibernateJPA generation and i create subtask to enhance the solution +Sorry the past part of the comment applies to another issue on TEIID I opened I think that the same compatibility problem apply since libs should be mostly the same and maybe eclipse complains in a different point than TEIID on a similar subject. +Closing for release +It may be one of those cases where each use of this is somewhat customized and the surrounding code knows the matrix implementation +See preceding comment +Please attach a quickstart and we will reopen the ticket +I agree with Jon +I was afraid that this might happen +of myfaces encode characters as into amp +It does seem that theres a runtime issue in the core App Server where its hard to restart a module without a descriptor +Other versions either doesnt seem to have any dependencies declared or the dependencies are not the same as submitted by the issue reporter +Could you please describe the problem more clearly so we can reproduce example would be even nicer +Here is the fix +The biggest issue with this ticket is that apparently for some people the exclude in the pom doesnt appear to be excluding the jar +Committed patch in revision Thank you all for reviewing. +Thank you for checking +As we are near the release date please review +Refactor function +Nutchgora currently ships with iirc +Does this work for you? What do I need to do to be able to use this ODT plugin again? It worked before but now its broken +Thanks Manuel +Perhaps the plugin was disabled during a time when some projects exported +here is the patch +Is this going to get a fix for +but perhaps that is the same thing as returning null +Id suggest we get a known config working first then look into the matrix builds +One more thing I didnt like with is that modules are not alphabetized when they are put into the Ivy container so it is very hard to find if a module got resolved or not +Verified and tested +Interesting did not know that +I think Stan Silvert has a fix for it +Werner its not new but assigned to Bruce +I dont see a reason why this should not be added +I have this need OFTEN +Committed datestamp only +With your replications dying often and restarting automatically it could be that youre often up to date when the task is startedAre your replications behind? Are there documents on the source that havent made it to the target even while it still says starting +Wont be addressed in time for the release + +Its a file for seam move from M to M +Added strategies to the exporter +I am trying to find out how we can avoid this but the problem that this exception happens even before initializing of servlet context or before the first line of our code +Im not sure I really understand this issue +would definitely make senseActually we were discussing about that with Stefan just one minute ago on IRCIll prepare that for next version +The patch modifies execution plan calculation algorithm to eliminate plugins their executions that need to be skipped +The attached zip contains +The failure in is not directly related to my change + +It was pointed out that these names should go to README not deleted altogether +Release closing all resolved issues. +hbasesnapshot to HBASENONTABLEDIRSThoughts? Too much effort +committed +for the to trunk. +Moving to think Commons TX is a dead projectI dont think we should work on integrating it with Camel +This minimum would have a role similar to the minimum required overlap in the here when disregarding the required and prohibited clauses +closing all resolved SOC issues +Why? If I break the build Id expect someone to be chasing me and the same thing Im doing for whoever introduced this + goes a long way to limiting the ugliness of this bug +Rajesh after clean EAP installation and using target this test case passes +I have never understood what the intended meaning of these serverwide or labels might be +Fixed into trunk rev +I hope this covers your needs +all issues to the CR +NP Ill do that right away within the context of this said you should really be considering separate with individual bean there +fix as a note the issue was a side effect of separating this patch from +Context description was a problem with tcpIf problems exist enabling tcp on the transport +Ok must be something with my test programI guess i am also a little bit confused about this whole positive definite thing +Fixed thanks for the patch +That JIRA suggests not to wait for agent daemon threads to finish but let them finish during VM shutdown +OopsI think is precise in that its not an expression but an expression plus orderby extension +Also to me such a policy seems to be awkward +it doesnt require id but it does declare id as the uniqueKey field +even if we think it might be used down the line +Ive added it to the plan for +The problem did not specifically involve clustered environments +It should be updated in the repository in a few minutes. +Yarns already passes on Windows. +Minor changes to and to a new test case for this issue +Attached the patch with the new test +Makes perfect sense to me +Added patch with sort as a an excludeFilter to patch +why would we want to put global actions into the toolbar? why not just remove them thenCould we at least for M get the search defaults fixed +Since AxisC fully supports I dont think this is a big issue +Thanks I didnt hit the submit button +This is the kind of thing that Lang exists for inmy opinion +Javadoc is for users. +push out to +Oshani I think this JIRA can be closed now +Absolutely no reason other than I was trying to get it working and learning maven at the same time I can add that to the list of +For example cxf it has not theI have created a topic in jboss ws forum to ask a method to get the version information but for now no answer +They have to be listed in Is there no solution for the jetty classes? Can we add easymocks to work around the outdated jetty version if its not actually used just referred to? The existence of jetty is horrible I would have that if I would have seen earlier +The patch looks very good +Thanks also from me for the kudos +I think a factor that weve overlooked is data locality +Can we get the old behavior restored? Thx +Is that correctExtensions should only be used in one direction +As part of the bundle repository page has been migrated to using UI +Thanks for the review TomUpdated patch to address Toms comments +I want to backport this to releases to facilitate things on demo has finally been done by +We have to check for class names +Eugene the Maven command line instruction that I use ismvn archetypegenerate localThis will return an interactive prompt from which to select the! +All these informations are only needed for versions using HAVEPTHREADH so please expand ifdefIll attach my proposal as patch +Also the package exports are refreshed for the deployed bundle only so the performance impact should be minimal. +for all the three +How is this issue related to? I dont see the connection +Bug fixed in revision The patch removing redundant checking and the backport are still not committed +We going to get this in for +Please verify +This looks good to meTo be clear do you mean the benchmark results or the merge +release? Or do we need a new patch where the issue is solved with parameters +I forgot to mention that the patch is based on the release +Is anything going to be done about this or should I be digging into the Mina code trying to fix it myself +Hi Dag thanks for filing +Ill add it +duplicate fields! +It looks like the answer to my question is yes +It looks like this is a regression introduced in abaedefffebcfc. +Updated patch +I think we should try to make sure we use generics for Range object now as this would typically have caught this one +Check the recent commit which also updated the examples. +Good chance for people to stop readingonce they see optionalFix the issue I was having though thanks +I can also confirm that mongrel shows no problems +So I dont think this is really necessary +Thanks +Ill try submitting a patch in the coming days +Summary After fixes and improvements in the JBDS sources some failures appeared +Patch applied and checked inI skipped the change to as the code uses +I wonder if this is related to? It is not the same error message but the symptoms appear to be similar +has anyone brought up the idea of creating and scripts that use this java class instead of the curl example that currently ships in exampleexampledocs? it would be one less thing for people to figure out and possibly screw up +Pete thinks it might be an Seam issue on to him +Yes the trigger is mostly the dev time behavior where you start karaf and child containers then do a clean build and you end up having staled karaf instances +Is this one still live +adding code to could generate slower or wrong answers from work which can lead to subtle data corruption but its not the thing that sets off Allens pager at night +Verified in +Im assuming you mean it will provide a warning and error out when the user is in local mode and or are found in the classpath +Thanks Eli I was trying to compile on Mac and realised later that hdfs native does not support compiling native code on mac +Is this meant to be relative to your home directorydrive maven home? I mean what would the final absolute path look like? Will it work if you specify this final string directly in the +Making critical +I switched it to look at how many storage dirs got removed by asking the storage insteadUnfortunately its difficult to write a unit test as was observed in the original JIRA +Graeme why do you just postpone this trivial patch instead of simply applying it? +This logic must not be working correctly +It should go into +I merged this into +Hey Bruno +Added unit test +Im attaching a patch that improves the callout mediator to use Endpoints +Committed +Looks good Paul thanks. +Updated patch fixes the test failures +Have you executed +So you can use same jndi name for and +Ive changed the character encoding for could you please try it again +See the attached patch +The patch looks good to me but I have one question +If we go from new to old leaveing out the oldest map files that are the largest and take the longest to include in compactionThe other regions will still tie up the compaction thread for say mins each on regions that only has map files becuase it will include the larger map files for compactionIn that time the two region that are getting lots of updates will be flushing more often meaning they will have many map files +Integrated into trunk and +This issue is DONE +The alternative is to double or more the number of per file operation which is unacceptableIf were using a different style then it should look different +Was wondering about the test +If I dont hear anything negative Ill commit these changes later todayI also updated and created a new testcase so that we dont accidentally this problemThe other change to was to avoid an obscure NPE when parsing someThanks just committed the changes posted in my patch from last week +Thank you for the reviews Rick and MikeRick you are right of course +Will have to check later +Hi Arturcan you please provide us with a test case that showcases the problem or at minimum your mapping and the class that holds the collection +To get this root cause mentioned above I had to patch and I think this small patch is good to be applied permanently as it improves the diagnostics quality +HiI have a table where de ID is generate using sequenceThe table is inside schema newschema +bq +Hi Stepan That is OK +Actually it was me who requested more information +If the test for node complete is done immediately after storing the element pointer from deserialization it can fit into the normal error flow of testing the status freeing the qname and returning the failure without risking a memory leak from building the node and not storing it +The IP clearance is now complete. +I updated Spring Build and grabbed the master branch but now theres no ANT target so I dont know how to generate the to get a look at them +I have been able to reproduce this locally it only happens with abstract decorators +Sorry for the previous postThere are a couple use cases that this patch affects NN DN JN should listen on the ports with the right protocols +But in any case namenode will fail stop if it faces some IO errorwhen reading edits log at boot time +This patch +Is that intentional +The situation should be considered when multiple gadgets are specified to be displayed in canvas mode +Here is a snippet from the logstestBUILD FAILEDhomejenkinsworkspacetrunk The following error occurred while executing this linehomejenkinsworkspacetrunk The following error occurred while executing this linehomejenkinsworkspacetrunksrccontrib The following error occurred while executing this linehomejenkinsworkspacetrunksrccontribstreaming The following error occurred while executing this linehomejenkinsworkspacetrunksrccontrib Source resource does not exist homejenkinsworkspacetrunksrccontribivyThis happens only when running ant when basedir property is propagated +I can attach it here if youd like +Committed to branch and to trunk +Alexey Im afraid the task you mentioned should be discussed before filing the bug or RFEI mean that does crash if is invalid +Ill update it soon and commit those unit tests too +There is only one way to restart an osgi framework and it is to update the system bundle +The was only failing on our build server Ill try to get some debug from that shortly +Assigning to myself to get this committed will reassign to Hairong upon commit since shes the contributor but this way itll stay on my queue +committed to trunk and +Test succeeded +Stefan may already have a ticket for this so this ticket may be marked as a duplicate +I should set this value otherwise sessions of will not be closed +The patch builds upon trunk and does not include the previous patch contents which are already committed + +Wrong button patch convert line I missed in +Attached testcase which shows the issueTestes against Camel and also fails when simply throwing an exception multiple times in the same route +Coz we specified the sizelimit before the users upload we told them its not going to pass and I canbarely imagine some will upload a super huge file by eating hisher own uploadbandwidth considering currently most CableDSL still dont have a good uploadspeed +I think pollers for internal queues are different they need to be shutdown later than ones that create new messages +Its just as preposterous to block a lucene improvement because solr cant use it yet +The problem with your suggestion is that you might be sending it out fast to the follower but the follower might be buffering all of it and not doing anything with it +I might make minor changes like not making an inner class of an interface and just making it Writable rather than separate the Writable wrapper +Could you please that +If anyone sees an issue that appears to be related to this potential thread unsafey we will bump the priority up to blocker for the next upcoming release. +The second argument indicates whether leading or trailing or both leading and trailing pad characters should be removed +to go through one more update now referencing the in the schemaLocation in the xmls. +cool +In this case test works fine +We are going to go with in production because at least that hasnt failed yet +The only issue is that oozie still reports the job as running although the launcher and action job complete +But this issue is fixed closing +prefix as well calling forrestWhen we did in main +Attaching a patch for version +Mac JDK user might need a copy of a jar that contains +This one seems pretty important +resolved in trunk and CR on. +Committed revision +Updated the patch for version. +core runtime ui and templating are all required +Unable to reproduce this on the Geronimo adapter +use this patch applied +Ill continue to investigate it +Please feel free to reopen if neededJarcec +Thanks so much! Let me know if you would like to review the final MEAP +update membership close for the release + +There are no performance implications for the more common case of no repeating terms in the queryI plan to commit this in a day or two +Previous comment incorrect it was broken in and fixed in. +I dont see this in the hudson queue +Patch committed to branchr +Suggest testing against a newer release. +You may not want to pollute your repo with this artifact that you may regard as unnecessary +New patch supercedes bug +See the STATE OF TESTING section of my long comment on above +Verified in CR build of docs + versions removed +The smooks editor cant generateparse XSLFreemarker file contents currently so if user want to use the graphical to generate the contents there will open a dialog to tell them its nonsupport +Close out until reproducible +This issue is better handled by existing aspects at this point +The problem here is that the Kernel uses the latest release of which was changed since the last release and this uses the apache commons vfs component for its file support but I am not sure that the vfs component supports query parameters as its from using the resolve in code that it returns with the URL without the extension so it then cant determine the file type +Thanks Nathan +Quick fix for Ubuntu +then a classifier would be a be if they need to write unit test using Pig +Patch applied to trunk +Closing old issues +Hmm my understanding was that this test just needs a fieldcache instance for testing purpose but I may be wrong +Fix in place +SiI went back and looked at the modification I have made +This issue belongs to jboss bpel editor +This issue cant be commited into branch because its caused in modifications from which are planned for of a bulk update all resolutions changed to done please review history to original resolution type +The ranking config would include a collector and other parameters that would be needed to make it work +table is a wrong name should be CLASSB +Please committed to trunk. +The Fix Version we are using in jira is trunk anyways. +Cobertura plugin triggers the WAR build that fails to unpack a dependency +I think I have a patch for this +I have tested this against IE and IE against https seems working +is a good alternative for that +a tough one +However it breaks when you specify a different directory +Note that Im using and the issue is still open. +Adds class +I committed the patch with revision there are no more change sets kept around in the propertyindex +verified on +oh the title of this ticket confused me +Moving back to for now +It would be nice if we can get a unit test that tests the scenario when the replication factor is decreased +I didnt realize we tried to patch the Erlang VM to support arbitrarily large revision trees +the real bug is in see +unified the properties into a single +Make sure index is based for this to work +I just pushed this but I havent had a chance to test in AS yet +We can also extend entity by adding two fields namely approverid and job +needs for now +We could create a document per field and nest them together under a row level parent issue we need to work around is Expiring columns +Committed +Wouldnt want to run a M video through it though +Solrs build is broken by these changes it seems +bq +In which context do you see the issue stated in the Jira? Seems order by with set the right reduce even before the patch +State of container has nothing to do with localized resources map +Also the bundled binary Ruby on OS X works as expected +I believe OS X should have offered you the upgrade long time agoBut anyway I am still unable to make the server fail to startI was using JBDS B EAP LibraryInternet ContentsHomebinjava for JBDS installation java version And then I added the EAP server using the default Java on my mac again +Bug reports are not for questions they are for bugsYou are trying to build Xalan using GCC with a binary distribution of Xerces that was built with the Intel compiler +It looks okay at the moment +I meant move list to core +Havent seen the above failure yet will keep an eye out for it +please adjust the priority to indicate that the Jira should be considered for CR +Adding Audited solved the problem +The following files are missing headers package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package package +Feel free to provide the patch for possible inclusion in +Attached a first revision of a regression test as patch a +Wont Fix. + did some updates in the description +Do you see what I mean? Of course it would be prudent for any developer to try out multiple test files not just but I still think it would be useful for you to provide a testThanks transition resolvedwontfix to closedwontfix +non of resteasy modules should have private +I can now see it happening +That implies calling as one option +Can you verify issue still exists and how to replicate +Ill try to mimic that behavior in Avro as much as possible although bolting an autotools project to an ant project isnt optimal +You have to specify it with camel case like portletsPluto and its not changeable right now as other plugins depend on this have the same there other way exclude camel naming? +The afterfix is applied at +I think appapp should also be fixedYes the upper case is pretty standard for naming the final static variables +So +Let me see what could be done here in connection to the UDP notification +Assigning to me for verification during final doc review for rel notes +It seemed to be related with +Patch for you verify whether this happens with +The order of patching initial contribution with additional tests is the following +This means including both versions in a classpath could cause problems and including only might also be problematic if something was depending on some of the classes included in and removed in +This would require some thought but perhaps ultimately what would be the best approach is to simply treat them like normal bundles and dont attach the fragment to the system bundle until the resolver indicates it is needed +r +Ill try a clean install and make sure I didnt mess something up when installing the all should be included in the bpelContent directory? +Id vote for doing the refactoring to make +Checkpointing my current state here +Pushing to LATER but hoping that ifwhen we redo this editor in Graphiti or Sapphire well be able to handle it +Revision . +If the directories are not the same the worst thing that happens is that some resources are wasted +I am providing some test cases for assertions +In my example mini app what if I want the output xml to have the complete content of all Paragraphs and only the id of the Tables? Some tables might be very big and they are used again and again in the whole document +OK I misread the patch +Still seeing it on logcat cut f d sort uniq c sort rn head TCPHIT TCPIMSHIT ERRUNKNOWN TCPMISS TCPIMSHIT ERRCONNECTFAIL TCPREFRESHHIT TCPHIT TCPHIT TCPHIT +Where are we at with this issue +Implemented +Thank for the help +Changes to +HiCan you reproduce this warningCan you confirm that this is a regression from? I didnt manage to look into it yet +So itd appear that failover in the Hot Rod client is not working as expected? Ill check the test in the EDG source code and try to replicate it in Infinispan +bq +I will post again once I have the answers to the above questionsThanks for the help +success with the test +I think the reluctance to do this before was based upon also needing Designer changes so that the packaged vdb would be appropriately designed against other vdbs +bq +I was thanking on this there is some other thing to consider like table splits will the regions be the same on both because there is no guarantee the compactions will happen at the same time or the split will find the same mid keyI would thank the master would be the idea process to pull logs a pass to peer master then it can split the logs in to regions and pass the edits on to the servers hosting the regionsI would like to see Sequential process of the edits to the peer so everything is in the same order and thats the way we store the wals nowI am not sure what the current status of appends on hdfs right now but if we had that working the master could just remember where in the wal it read up to and pull every x secs to see if there are any updates then we would not have to worry about waiting for a log to roll which could be a while in some cases +Ah I didnt notice that +Thanks +Proposed patch for the preparing support for Key files files with certificates or Java keystores +Fixed r +Patch w failing test +fyi is caused by out of memory +When I will have time I can rise that issue in other forum group but I guess you can do that faster +It should only serve for presentation purposes +Uploading rebased patch for Vinay for rebasing the patch +Patch adds a testcase and fixes the bug +Its not actually The User Guide docs need to be updated +Well revisit the use of placeholders and expressions in. +So I think a can be just about anything +We must think of moving that to the and provide a common syntax for allowdisallow requests +Submitted patch v for trunk +I understand you got busy with other things but I wasnt going to stop working on the project in the meantime +Great work +Did you? +Added a test case both of the DSL and the XML versions which are now working great +This is a better fix than +Running Eclipse classpath verification +Hi Martin I used the Eclipse Indigo SR and Spring IDE I also meet this error can I fix it without update the Eclipse version? Seems the latest version of Spring IDE must use Eclipse Juno +Should I create a separate Solr issue for Solr updated files or can I upload a patch for all updated files +Resolving as fixed +Fixed in CVS +Corrected version attached +bulk defer to defer to +committed +Added patch +It does square of the number of regions flushsize +TimothyI have similar questions to yours and would be interested in helping you with this investigation +Monitoring think this is pretty good but we should probably favor filters for actual monitoring value changes so that people who have their own monitoring infrastructure already can use that +Attachment has been added with description The fop revision . +Transitioning to closed +The patch has already committed to the branch +JAXB +to work +It also shows off the two options + Username mtkopone +Resolving as such. + +the file with links to fix is attached +Squashed your patch into Estimated keys in based on the first keys or megabytes of data whichever comes first Fixed the test in to catch the longint discrepancy only tested recovering a single row Removed get in +Andrew did you see the other languages? I translated the new message into German Spanish French and Italian +Now that the mochiweb branch has been merged back into trunk this is fixed +Slight update to previous patch calling deleteMeta during the map phase stopped the scanner from getting anymore results +If it works for you I can add it to Wicket +Im not ready to reject this issue yet that is why I didnt +Attaching +client codeThis code shows described errorTo server sent value but server received value this is bug in thrift php code +I have the same problem +Nevermind not worth the closing stale resolved issues +Im just saying that you dont have to show explicitly setting the default +Should you? You used to set it if unset in previous patch +If a token is available the RPC client will attempt SASL regardless of the clients security conf +copied from +added already Known Issue Documentation text from Bz +Another one +However its not correct IMO and leads to the NPE described above +Attach a fix +Without its very ugly but it should fix the rows that are not getting committed +Rohini did you apply the patch and test it works +exactly +With flushtrue when is called do we still wantto not have concurrent merges execute synchronously? Or onlywait for the merges to complete that are from ramDir toprimaryDir +I just committed this +I reworked both target platform definitions and the product definition to fit the SP version +Already applied +I subscribed to the dev list +Ofbiz site and dist both moved to svnpubsubEnjoy. +I should mention that this happens easily with clients each doing requests sequentially with a delay in between so there is no way there could be more than in the queue already +Oh Owen is right of course + +I cant explain it but this is the behavior I seeMy supposition that this was related to a custom parser configuraiton did notbear out +Added a helper program that prints sample text to stderr and changed the test to call this helper program instead +tested and works +just make it last thing we close out +these are all unimplemented m features +I increased the max from to +I agree with Anders too many apps depend on this variable simple renaming will break backward m just not sure it worth it +Rather than do the calculations every time wouldnt it be better wait until you get to tuples then calculate the average? This would miss the case where you can store less than tuples but that seems unlikelySome of the comments in that were copied over from the previous code such as dealing with spills in the midst of reading are no longer true +Oops the patch was in the Fisheye tab right in front of me +It now seems to work out of the box with and the latest servicemix deployer +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I just committed this +Fix for failed test case +Fixed from the perspective +Uploading patch that addresses Hiteshs comments +take mematzew AT apache DOT orgIll check for one one of the mentorshenning AT apache DOT. +Rev attempts to restore parameterized inclusion of the tm +This is a major change +However putting a link from emma to the mojo site would suffice for this as well and may be better +If pure reflection isnt showing the field as you would expect what else could Struts do +should be fixed now +This problem turns out to be the reincarnation of that has never been fixed. +The original issue still has not been fixed as the givenName is still displayedI suggest the following Display the cn as and provide details of how to request a change if the value is incorrect +Everything goes fine with revision of +Net bindings for the C messaging API. +Oops +I have given your wiki account karma to assign tickets +This example is rendered as correctly as could be expectedIt contains effects which fop and pdf do not handle very well +Viraj can you try adding hadoop properties you wish to override in and then add that xml in the classpath +Hows this sound FitzDefine the URL and authentication details for this server instance +I cannot reproduce it now either +could you please apply this to +Thanks Arun +Thank you +Patch is in already resolving to make sure we have this in the all issues that have been in resolved state for more than one month without further comments to closed status +Missing a test +Same thing goes for andMaybe this should be implemented on so the behaviour can beconfigured on each instance +Thanks +An udpated patch not to the comments given by Lars +See an example +I also applied to in r +If the FCQ is known the client will send the Stop request +ASAP +Either has pros and cons +This has been inactive for about months since the patch was applied so marking it as resolved +Committed to trunk. +It would also be helpful to be aware of what the other envisioned subtasks are for this effort +If you find further issues please create a new jira and link it to this one. +If you dont install rails as a gem but put it into bundler dependencies you will have to use scriptrails or bundle exec rails +I will attach the file in second +Yes I know all this but did you actually try and run this inside eclipseJust because we stick some declarations into does not make it actually work inside eclipseHow did you test this? Did you osgify all the other dependenciesDid classloading of user classes work correctly for you +One other note +Heres what I have so farYou will find a new folder called ConfigService which hs the same structure as the Service it should have a Also there are some addtions made to the common folder an interface and its implementationI am sticking with the format of the php +zipunzip preserves permissions and remotely the command line unzip is used +Committed +TedStackthanks a lot for your valuable review commentsI dont think I understand when you say you want to port it to? What does that require exactly? will be more than happy to address that +Heres a new version of this patch that includes tests for each of the new toolsIll commit this soon unless there are objections +solves this problem using milliseconds +I dont actually know how much time this saves us on large files +Your patch was applied at r +I have a nit we seem to create a new each time wouldnt it be easier to the object? Create it once and it each time +Thanks a lot +Thanks you Boris +Good fix this +Nikolay Sorry if the following sounds tough but It looks really strange how much consideration has been put in the patch if you did not even tried to run unit tests? It is hard to follow how the relates to initial issue report +Cancelling patch to incorporate Devarajs review comments offline +The lexical error occurs when the first escaped is parsed +I did not get the error +Applies on top of to branch +We could have a thread pool on the side just to deal with such requests but this would be effective only if not all events require blocking on IOIf anyone has more thoughts on how to deal with such cases I would be interested in hearing +Hey Mithun your patch applies to both and trunk for me as well +It isnt my decision as to whether will support the EE level default datasource +Kicking Jenkins +Committed patch to trunk +This is bad user data in prod +but wait must have fixed something +Thanks for the reminder +Do you happen to have a testcase for this already available? If not I can write oneThanks +Cool I can do that +I was assuming work would continue incrementally to add authorization +I have tested this patch manually with its deps in a secured HA cluster and it works fine +patch was fixed by +The latest version has a fix. +I think this no longer happens may be you check? It may be fixed but needs to be verified the script is obsolete due to changes in the arguments to tools such as producer performance +I committed this to trunk +If you find further issues please create a new jira and link it to this one. +For example a router handler no longer sends messages to resolved channels directly +histProxy is actually a which contains the actual proxy +We dont know and there are no test casescheers Roland +Let me file a JIRA +but wouldnt it be nice tohave one +Ill check it +The code is in but it relies on to be fixed before the code can work +Can someone sort out the licensing with this code +JoshuaYou can create another issue for it incase we release camelI will go through your new patch when I get some time today. +Hi Franciscothe error you are seeing is due to STS shipping the Spring Integration jars +Patch applied! Thanks Kevin! +Note all nodes under jcrcontent are ntunstructured +Groovy Eclipse works much better for us today and I havent heard any complaints lately +Is this behavior by designIf you do an add followed by a delete of the same document one would normally expect that the document would be deleted +This is a very simple test program that tests that after applying a empty Clob and a empty Blob are properly inserted into the database +Can you please post the full schema and stack trace? Thanks +I use git hopefully that doesnt matter +Great +The attached patch is for consideration and discussion not immediate submission +Havent done a make install yet with zookeeper so that part should be ok Ill try applying the patch to a fresh checkout Mahadev +If you want to find us on IRC perhaps we could work on it together +I wasnt sure about it so Ive rather created this issue +We should definitely keep in mind the Cyrus solution +I agree with Sylvain +Example that you can use to reproduce this +to the patch +matt please copy the bundle urls in the same issue to avoid overheadI uploaded the jars to allow maven users to use them its definitively in my roadmap to have a stable m build of spring and try to convince spring guys to take ownership myabe javaone will be a good poms most of these depend upon + +How can I disable a rule by default +Please see my comment +narrow bit in the middleThats by design +I dont think that jar is intended to be released with the distribution +mf files in jarssanelists +Ill mentally ad this to my backlog +Thanks for the feedback Hitesh +And the header without the copyright should be the verbatim notice as in the license directive +Fix to patch to return null when no stats were calculated +Probably a session timeout! error displayed on webpage +Keep us posted +I dont see why it would not +StackCan have a look at thisMay be needed in the that we will be taking either for or +It is OK now +Verifying build on Jenkins +Taking a look +fix by r +thanks the patch Granting license to ASF for inclusion in ASF works +Could you make a patch that includes just the header change? Please make the header identical to other scala files +note will have less overhead when the records with same key values appear next to each other for example when the data is sorted +Lowering priority since its a feature not officially at the merging of all functionality. +Closing +ToddCan you validate if this addresses your issueThanksSent from my iPhone +There must be something wrong with your maven setup such as including a repository or using a broken repo managerIf you want help figuring out what is wrong the user list or irc might be more appropriateIf you mentioned this problem somewhere before I certainly missed it +I would suggest you get a version of GCC in the series and go with that +Fixed so that whitelisting and params are read by when it is operating standalone. +Added checker using equivalent select from dual resolved issues. +Added one more chart where the increasing time of each operation is more test was ran with derby but the results are same as with oracle g +I do know somethingabout and the ftp parsersIs this external link bit e something that is zOS specific or do otherflavors of unix support it? Would this break the parser on OS systemsAre there other differences between zOS and regular unix FTP servers that weneed to take account ofWhere Im going with this isWould it be better to accept this patch or to create an independent parser forzOS? To help answer this question heres anotherWhen connecting to a zOS FTP server what does the SYST ftp command returnSorry to take so long in answering but would like to have these questionsanswered before I jump one way or the other +is used by region server to report status +Please do svn revert R +Martin commited the above noted fix previously this should be resolved. +Fix job conf parameter only +Ill assume this is correct behavior and my fault for specifying a conflicting dependency +The comment was that the extended command expressions are available from onwards allowing the approach mentioned above +Ill be making noise again in a few weeks then +as this hasnt had action in a while i thought id comment to see where the status isthis should be a very simple fix and its still not resolvedthanks would be nice to haveThanxwas active +Additionally talks about accepting HTTP with no content which is already implemented in CXFSo the patch for this which suppresses the transmission of HTTP empty SOAP envelope responses I suppose these both issues can be resolved as wellI will also create a separate jira ticket to make the client side change that is required for letting the CXF client handle HTTP for the call when the is sent back to the decoupled endpointregards aki +fixed the javadoc warning +It needs which seems to be on both branches but this never seems to have made it +Considering that that file is a few bytes long it will be cached and this will happen only when js validation is enabled and dojo is used I dont think it is worth the trouble. +Is this test case checking for empty host in second zone +ToddAaron Can you review the patch? I have tested the patch with the failing config and it works +Anyone looking at this one +I wonder if you have a jar in the classpath +dbentity dbentity etc +Jimmytry myproxy sniff with Harmony M and Reference implementation +Thanks for the patch Ted +It should say that specific config is missingCan you please paste the error message with your latest updated patch +Investigate further for if reproducable steps can be established +And my position remains the same Sonar core team shouldnt do it right now. +This is the same problem as +This is extremely important as layout is very complexAlso is this Knuth algorithm copyrighted? Where did you get it from? It is rare that we have classes named after authorsm going to attach the corrected patch sorry againKnuths algorithm is described in the essay D +I did it anyways given you are finding it hard to get some time +Thanks Bryan for committing the patch +People that dont want to just dump the logging jars we give into their container will need to learn about same as before +Closing. +We might also have collectors start spitting back HTTP errors if they cross some specified load average +Please reopen this issue if you can provide more information on how to recreate it. +This issue has been resolved for over a year with no further movement +duplicate of +But not force this on Designer to the translators have multiple import properties that can be used to customize this importer but so far Teiid does not have smarts in the system to expose what these properties are so that Designer can provide a intelligent GUI on top unless they hard code for each of the translator +The attached patch makes Metal LookFeel of similar to RIs +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Well I suppose backing out deletes will throw off codes so yeah lets not do that +Fixed this issue with revisions Just switched from developing on Windows to a Pro +Ok great +Oracle should already be set up to get the struct +I will add more detail about this issue late +We need concrete specification for JSON requestsresponses with exact formats headers and descriptions for all feature the service will provide +as suggested by Michael Enke I am ableto view the svg files and my own sampleThanks a lot and apologies for my insufficient information in the first placergds +Yes it is easy we can remove that logic +ah darn does the query server spec run without your change without errors? that file is a bit +is still needed to cover recipient with upper cases +This change also fixes a somewhat severe issues on the protocol path caused by the changes in which meant the client would not recover messages as requested and instead could consume the resent messages during the recover operation +In this way the Run As Maven install action on the parent project works correctly and the application is deployed As Maven install on the EAR project still doesnt work unless we disable workspace resolution +Your testing scenario was correct +Anton the exception is correct +The patch puts in loops around those calls +I have fixed the issue following the examples of the other required jars +No need for another patch Ill merge it inKind regards +Thanks Namit. +We configure these operations with the appropriate role attribute but chooses to ignore this value itWhat is the role attributeIs there any other JMX Console which not chooses to ignore this valueEugene +You can also say that this change is only supported starting with a version such as since that was a major point in the API change +However in more recent builds several classes changed and there was an exception added if you tried to write a stats query using a multivalued facet +I agree that encryption is part of the solution not the solution for this Im sorry to say unless you choose to become your own CAUsing encryption requires that you issue certs so inherently you are becoming a CA when you use itbq +Actually I think this should belong in its own example class because it does not allocate a taxoDir as well as cannot demo hierarchical facets +Will update the patch to add operations in create +TRACE log +Sample +You mentioned that you added a patch and testcase but i cant find it +The problem is grails upgrade and etc even if there is already one or an old one with the naming +It uses a timed expiry thread set at low priority +GunnarIm not sure what you want to achieve with but if you say it might cover this issue so how about do not do any changes now the more it still looks like a workaround +Patch for +and the difficulty of adding new methods to and making sure that all implementations correctly propagate them through to sub +Copied into another context problem was gone umlauts displayed correctlyBack to context in commented out problem was gone umlauts displayed correctly +and are different casesReal problem is in invalid java code with incorrect or deprecated login algorithm to ssl site +There is a problem applying this patch cannot find qlsrcjavaorgapachehadoophiveqlplan to patchCan you take a look +Thanks a +This bug caused users hard drives to be completely filled with these huge erroneous attachment files +This issue has been fixed and released as part of release +Now Im not against your idea I just like the proper authority to implement such a drastic change +Merged a pull request. +Matt is this still happeningSounds like you installed additonal plugins into your system? tc server meclipse and moreCould you list them all and also try increasing your permgen space since that is what for some reason eclipsejbds is running out ofWe dont see this is our testruns so my guess you either you either have some plugin or project contents that in combination causes the problem so well need some more info +I looked at this +Weiyan if we would use GIT I would have said nothing but Apache Hadoop uses SVN GIT is just a mirror +If reproduces change type to bug +All of these are related to bugs in integration all classes do exist in you need to alter the exports to cope with this please open a new JIRA issue for that +The issue is that while we have general faith in processscripts for migration we have had occurrences where branchestags have been missing in certain cases +What is the internal error that gets logged on the server? is a lot of keys +I was getting some errors or so so I wonder what changed +orig +A way would be to rely on to determine the actual output urls but its too complicated with the current systemAnother way would be to keep track of the files that change +I have implemented this extension and checked in as Rev +By map task creation do you mean task startup costs? We have not yet explicitly measured that but it does involve starting a JVM +It is in and is actually a problem in the last test in the base +regression test drlvm at r +No problems have been reported with the patch that was committed so closing. +Keep this JIRE open for more refactor +Does that sound right +Attaching the patch to this jira +This patch doesnt do anything about those are all untouched still. +If someone tries to run the examples jar the classes from the jar will end up being picked up before the jar because the Hadoop framework jars appear before the users stuff in the classpath by default +committed +this is the attached file that is throwing the at r +Im using attached an example illustrating the problem. +I wanted to close this out as the initial porting is done and now its just improvements from here +BTW using the double mapping trick work by pure luck +Thanks Mike works like a charm. +Ive begin to work on this only to add publish for the moment +While transitive dependencies is what maven is all about and makes it the wonderful tool it is there are times when those dependencies need a finer grain control +I am really not sure in which version this was fixed +I suspect this was fixed by one of them +I tested the build using my patched and it worked OK +avoid lineEnding and filtering when handling binary files +Basically implement the EJB version of this JIRA +Take a look at the content of our existing distributions to see what is included +Tests that failed are not related to this bug +The other option would be to create different domains that are configured however to use the same password +Doubtful on renaming + to the proposal +I already followed this link but this link does not show a specific manifest to application clientsI think the fact that Java EE spec requires that app clients specify the main class should should be considered and at least a simple specific manifest for an application client example that specifies should be providedAlso the doc may instruct users to specify manifest and stating it is required by Java EE specThis doc would improve user experience reducing risks of problems +bq +Patch against x adds asserts to SM test to ensure they are not allowed +Colin thanks for the hint +Supposedly but not actually +I meant to say that I like Eugenes suggestion but agree that this single method is not really enough to start up a new class especially given the kind of String manipulation being done I think its good for now +Now I am out to walk the dog and get some fresh air +Net code quality projectsTechnical Debt Calculates in US Dollars the work required to fix all quality issues in source codeTimeline Advanced time machine chartTotal Quality Provides an overall measure of the quality of the project linking code quality design architecture and unit testingUseless Code Tracker Find duplicated source code within projectSystem pluginsCheckstyleAnalyze Java code with CheckstyleCloverGet code coverage with Atlassian CloverCoberturaGet code coverage with CoberturaCoreProvides all common components required to cover all languagesDesignAnalyze Java bytecode to compute +For now I have verified this test passes on an upgraded database with this patch applied +If regressions pass Ill commit +Need this in since it preserves the old behavior of isDir +Also includes tests +So cant find any images +If there were an issue with setting features on the parser several others would have reported it by now +Sorry if my comment wasnt clear enough +While porting the integration tests I found no more references to or other Felix internals +I will try this out and get to the bottom of this +Please check that it was applied as expected + Yeah I wondered why that starting coming back +Please verify it fixes your issue +Closed upon release of. +Any luck finding the code Jeff? +Atomic increments are a lot more expensive than a simple increment since it has to push the value to the processor cache and notify any other processors that may have copies of the cache line +a FLOAT column with values and will be serialized in the column value buffer as +The file is handled by the but the file is handled directly by the re right if +No update for ages assumed slipped to next please set the appropriate fix version +Thanks Deepak your patch is in trunk rev Bilgin for committing the patch. +The current design doesnt handle in this case M will have to wait for M although M was sent by a different sender +looks good to me +Ran the tests again got the same results tests failed due to the hostName problem +Im going to start messing withCascading right now so if you want my input just let me know +In the process I also modified the spellcheckscr example to no longer use the generic activator since it no longer exists +Upgrading on another platform and moving back to AS doesnt work +I implemented a new fix for the fix +Thanks a lot for the review Todd +Attached a issue +Thanks Dinesh. +Possibly? Are you thinking of adding andor annotations via? That could work to achieve desired main reason for distinct way to denote views would be adding bit more semantics within classes +Issue addressed by. +The file to be patched is usrlocalwwwapache on this file is currently not in svn and need therefore to be patched local on the vm. +Merged changes into codeline using revision +ad +I should have searched the JIRA first +Closing as Fixed because it worked with last time I triedPlease comment here if it doesnt and we will reopen it. +It has been the architectural policy of Xalan to always opt to continue to runin the cases where such discretion is granted +Test +Hi David +Hi JesseI am considering the solution which dont use Hlog +why not adding him as a committer +Thanks for spotting this +The stacktrace shows that the hudson runs tests against JBT +Thanks for the details Shawn +I will submit the patch again with +this change goes for all methods that generates an Fieldvector as output +I thought I just saw you say you arent planning anymore releases +But nothing strange there +It exists in the same way in as well as in the trunkI accept if you decide to not fix it but claiming this fixed is just so Adrian +for the patch +Is the GUI hanging trying to add the new layer or when you have compute the native bounding boxWith layers Oracle that is going to take forever better fill the values would be fixed by the fix to +Stay tuned +Should we consider this for backport +We commit these to SVN to simplify eclipse usage and rebuilding them on different machines generates spurious diffs +Yes decompose was added back to after I rebased those patch and I didnt rebase them back +Interesting I was wondering whether people actually need to compress fromto byte buffers +Diwaker thank you for the work on this guide merging this with the new site now +Has there been any progress on this during last years? +Ill go ahead and do this now +the separate log that it generates can be periodically merged with the file system image +This ought to fix the extra javac warning +It took a while but I think we are close to getting a resolution to this Issue +in the scr project in Rev +Also it might evolve and perhaps should be versioned with releases +It works but it might not in the future or may work differently with a different UI +I remember rave was displaying a error message A Widget already exists for this URLBut i am not getting this message now +Patch committed. +The tiger doesnt show up using theJeremias provided me with that example long time ago +Dave Im too lazy to look around and see if your hack is in place +Datascrollers nested inside two or more iteration components wont work +also not included and Store changes as they already handled +Nirmal mentioned that he was going to do a bit more refactoring of this one before calling it reviewable again +It is much clean and safe to use hbck to handle the assignmentregion overlaps and so on +Sorry I didnt thoroughly test the first one +Addressed Nicholass most of the comments except removal of soft reference +One missing instruction from the previous comment the plugin on top of the IDEA JPA facet +Sure +Does not appear to be a problem with +This is probably hard to do as well +Patch committed revision +This affects +It invigorates greatly when someone is interested in your work +Clearly waits are not a good synchronization mechanism +Issues with patches have more change to be fixed +And there is a better alternativeThe better alternative being just use for the and a separate component property mapped for the child parent relationship with insert and update false? +for branch. +looks good +It is not as nice as abstract bean but it works +Some comments on the code Code in runtaskasuser can reuse code from initializetask I think can be moved to a test folder following convention from other C modules +Sorry it should have been assigned to the components component +Attached patch w test showing the issue and fix that fixes the testI plan to commit in a day or two. +So overall I do like the last patch attached by Yuki +Upgrading to Teiid CR then Teiid FINAL fixed this issue +Mostly looks good some minor comments +The fix has been verified +Project dependencies by design never contribute to the plugin class loader +my patch over rode all the changes +We only made one modification to the Master as a suggestion for one design option +Im assuming a classpath issue with some Hadoop jars so more research is needed but wanted to add to the discussion above +Any sequence of one or more blanks in the patch file matches any sequence in the original file and sequences of blanks at the ends of lines are ignored +However all jars in ER have different checksums due to the unsigning so they now fail diff but look the same using +Closing issue fixed in released version. +in branch +Now you just click on the profile name to edit it. +Just for completeness Continuum version is now build +Log files reports etc +I specified the oracle in server class path but still it is using the ibm classesThank youSathiyaTheres not much we can do with this unfortunately +Closing applied BJs patch with revision . +It leads to confusing code that is easy to break +Thanks for the bug report however the codes not complete +I think the problem is in the existing API though +I will check out its complexity during the weekend and let you know if I go for it or not +avroAnybody else have an opinionI know I probably should have added a test but since the other tests didnt do anything with testing the file namepath I was just taking the quick way out +Ill look into it now that I can reproduce it +So this is a regression on Sonar +from the standpoint of external error codes would seem to be much better suited +rewrite the changes to handle the HA merge +This should be fixed as of commit fbbdcfdfdeaafbdddefbdcf +added a deleteMessage process to the and action. +is this still relevant +I just committed this. +For added a configurable property +This can not be resolved without a bigger refactoring of the test framework therefore this need to be a standalone test for the There is another standanlone test at that tests to load a driver configuration with +I love these ideas +hi Flume may in the plan when does it can be released? +Does it make any sense to mark things like this which should only be used by the serialization mechanism inside hadoop as deprecated so that people dont use them in regular code +The trunk patch went into +This is great for folks trying to build on WindowsPerhaps we should have left behind the shell scripts for folks on Linux who are doing just fine currently without python +The only thing the exception says is that the code was in Socket IO when the JVM ran out of memoryThe memory leak could be anywhere +Can you include the plugin generated spec as well for comparison +Im hoping to resolve this tomorrow +This seems to be not only a db issue test cases using limit and offset do not make sense when limitoffset is not supported +Ive been working with a client that has this exact problem and this patch cures it on last nights x code +Hello LeonardoThats probably right +applies on this problem and the better is let is as is +The reason why I thing it is related to is because the problem started when I put the in my lib and took out the sunsWhen I took out the and returned the it worked without me doing anything else +AA your is out of date you are taking us back a version for the surefire plugin +I was concerned but we are pretty close to a release that will hopefully go GA +If it gets an errordoing a piece of work it should catch it and try to figure out what to do +However in it is random order +That explains it +b We should leave it up to the user to create a storage system that hooks into the callbackYep that would work too +But we could and our clusters do have blocks that have two replicas and all its sources are in this state +Reschedule if completed sooner. +ClausI added a example with comments in the unit test and its already a part of the wiki page +wrong tla there dyslexics of the world untie +We can certainly do some set stuff there triggering of the registers itself as for automatic Connection recovery now +The feedback I heard from James was that WTP is currently very flakey and it wasnt worth productizing that stuff for this release +If this issue reappears I will do a +Second if you put your handlers after the dispatch phase then that module will be engaged to the service that you are trying to engage +Then when creating altering a table immediately complain if the compression isnt in the supported set +I made a change to the so that it will always wrap the request even if it is already wrapped and this seems to get to the same goal +So the only option for us will be to reload a default configuration +This should be discussed on the user mailing list not treated as a bug. +On the other hand we perform the du calculation for every directory statd always returning it with the other file metadata +Here is the but in this case the echo off should be the echo line up +Fixed in rev +I think I might know have the impression that the problem does not come from the fact that it is above but more from the fact that it tries to consolidate measures in which some are above and some are do you think +A common token format with variable identity attributes to support access controlCan you expand on this and also give an example +would be glad to helpCheersSveta +If there are no obligations I will commit it within this week +Native SOAP stack is not supporting JAXWS any more +oh more thingsI thought you found your way around the y argument? I still see that in thereDont comment out blocks of code just delete themAdd some documentation about creating new Exec Engines to the docs or at least post it here +Completed At revision o Removed the file from as the module does not provide the implementation classes +Sorry I fell off for a bit been swamped with a deadline and the holiday weekend here +Thanks! +bq +Ill commit this if Hudson comes back OKThanksEli +base file from asfWas imported in two steps +Implemented can someone else test this and give me some feedback +All tier databases tested to not produce any errors in runs +Version Released +Excluding netty might is a build system dependency at the moment +to cloudlocalhost identified by cloudGRANT ALL ON cloudbridge +Done. +Camel consumes only once so you need to convert to other format if you want to read more than once +Sorry for that +Hi Todd! Ive tested it with Cassandra using the for the normal Thrift API but not with hive or hadoop +If you look at the image in the watch window you will see connection uri contents and it was tcpThis failed during the connection initialization. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +I was thinking maybe we could add some logic but this is ok +Reopening this JIRA as Michal Fojtik sent a patch for deltacloudc usernamepassword issue +But if theyre minimal like you say the easiest thing do Im sure would be to just change them to be full instead of the shortened links +The code is just a little trickier than what I had originally imagined so I tried to add some more unit test coverage to make up for italso JIRA notifications seem to be back! +Hi +thanks Patrick for working with me on this +committed to trunk branch. +I tweaked a unit test to cause this behavior and it worked normally with the inserted into the file so unless we get more feedback I amd going to just go ahead and mark this as cannot reproduce +Start from a remote server perspective means just connecting to it +Ok I see +Marking this as Wont Fix since provides an alternative. +Extend the SQS component with a property like amazon and when the queue is created set the policy on the queue that allows message postingI dont know if thats enough since the example also shows policy setup on the SNS topic +Functionally we could just change the webView to be a and take away +We have the same problem +I still believe that we should use for our code maintainance +mentions this approach as an option +Whats the URL of the SVN repo for it? There is a readme file in the root of the plugin project that has some info about building is no readme in either nor in the is an old version of the embedder in the central maven repo but it also doesnt specify the finds a lot of pages when I look for but no hint about the source please just send me the URL of the readme or the URL of the embedder project page +Fixed on closing all issues in resolved state that have not been reopened nor commented over in the last month +fixed in free to reopen if need be. +Patch to escape +Very annoying +What is the use case for this +Fixed in . +By then we may have much more experience with generics and be able to make a better choice +This seems specific +Ok I actually did fix this with some other changes I made +For federate purpose the required class is overwritten to fix the issue +Ill run again +in is missing this jira +Yes there are some scary dependencies on stuff that would have required refactoring the world +just ugly +Implemented support in the relevant objects +Im thinking that this has to be a strange JDK quirk as I can vouch for the fact that the Sun and IBM and JDKs do not have any problem with this whatsoever +Files used to create Web Service The Web Service The object returned by the Web Service +Yuanyuan I have committed on your behalf and am closing. +Patch does not apply to +I havent yet tested the patch by setting to the latest version if you are still having issues open a support case +Yes this is a lucene index format issue +Interestingly the POM inside this jar has the configuration but it does not appear to have been executedIs there another version I need to pick up with the bundle informationIf not can you roll a new release with this bundle information in the Manifest +This can be closed however maybe documentation should mention this as known issue or similar +Could use to create an upper bound on the memory size of the puts mapNick can you implement this in the next patch? could but what would that maximum number be? Should it be fixed at some number? Should it sniff the Runtime for heap size? This seems like a pretty arcane configuration pointId prefer to other parts of the MR plumbing such that buffering the puts is unnecessary +Tagging this for as it seems a bit complex for inclusion in +Chad from the repro application what is both the expected and observed behavior here +Closed after release. +I updated the website to give visibility to the. +I sent Arneke a dataset including slds last week could you get it from him? Or would you like me to email it to you +I guess there are also additional references to the block indexes that would need to be removed in order to actually evictOne way to do this would be to have a special eviction method called when evicting block index blocks +Not possible for series +Verify that the case was in QA +Committed is passing in +Can you upload the line of data that caused the trouble +bulk defer issues to +Its a known bug in our build chain only one pass is made over programlistings and the like so either callouts or highlighting can be performed +Did that thx +Also I have no problem making it painful for people who do this knowing they are doing it wrong to give a nudge to people who dont realize they are doing it wrong +I will do submit patch after that +Fixed the javadoc and findbugs warnings +bq +Fixed we close the issue +Thanks Adrian for the fix +This should probably be handled in +Created as this is a problem with the app servers that bundle weld rather than a bug in weld itself. +And BAM youve got your I dont like It doesnt use any of s service model or servicefactory stuff +Maybe Im trying to be too protective but being able to convert a toOne to a toMany is simplistic to the point of being hazardous IMHO +Ah that makes sense. +Thanks for fixing this guys. +I implemented a change as it was in first proposal +Need to look into the snapshot server +Does the zip file that is causing the problem when loaded from an HTTP server contain any composites or just utility jar files +there is no time between CR and GA for addingfixing features +Im going to commit this momentarily +Moved to jbide since not jbds marking as critical for CR not for fixing but at least for understanding what is going on here +Did it? What kind of problems iima +You should be able to use the PDF to get the printing done +Parser will throw an exception when it detects a name conflict in macroThis is not a perfect solution +The only difference is fixed javadoc between v from RB and v +At the time no one else had additional ideas so I just went with it +for the patch +Now I dont have to explicitly turn off html escaping on all the radio buttons throughout the apps +the skip is fixedthe scope of this issue I just checked +I plan to commence our new developments with release in order to leverage the latest features +Verified on. +I followed these steps but didnt get the same results +Yes this would be quite useful +Please reopen if more problems are found +private static final String WORKINGDIR userhamabspBut we can add a configuration property for it +I am waiting for the approval or other instruction about the conceptThanks AndunThe message is okYou should add this message to an use to add for other languagesThanksVasiliy +In fact I reallydislike it not the least of which because of the security issues ofletting people specify what method to call in the URLI thought the concept of wildcard alias mappings was going to makethis unneeded +The failure I mentioned in the previous comment was a duplicate of +The test case class and corresponding application context file in Eclipse patch format created against HEAD +Committing +you can get rid of that issue by adding the EA repo to your as most of us did +One more thing at least on linux I have verified with lsof the tcnative library is loaded by the jboss server +This is a problem +Committed a preliminary fix +It is handle as well in spring bean injection +Assuming I just need to validate the demos Im think that is an os specific issue +Alex andor Toby any update on thisDoes the latest patch Fabio added help resolving the issue +Could you post what you did please +I moved the feature under a separate cetegory on th update site so that users without installed will not get an error when the click on the whole category +As javadoc I think +The attached patch contains a possible implementation for a completeBefore I commit can you please check it to verify eventually legal issues we could have with OracleImplementations are different but patterns have extracted from the Inflector I dont think they could claim TM on regexp patterns can theyThanks in advance +You may be right +Youre missing the row attribute on your dataTableHowever since I hit the same error the first time I used dataScroller I think its reasonable to report a better error than +Defer all unresolved issues from to +cannot use members of because it is static +The cache is a cache by Primary Key +The first step is to reduce the number of methods in +This also fixes the highlighter problem with +Ive made implement Serializable as I believe this to be the right thing to do regardless of whether it helps address the Google App Engine issue +Below is a summary of the changes made in the patch in order to enable priority based queueing Added a use option to the Broker Added the use parameter to the class Updated code which was creating instances to set the use Converted into an Interface Created a new class which implements and contains the same implementation as was previously in Created a new class which contains the newly created +INFO Running for rows took msjmspaggihbasetesthbase cat grep took grep INFO Running for rows took ms +A Junit test case is attached. +Thanks JonathanCan you reattach the last patch with ASF grant? +A solution which does not requires fixed namespace or context names +That could be addressed in a seperate jira which moves the commit logic to +Hm +Just pushed this +I havent heard of efforts to add these types seems that strings have been working well +It is not a blocker so if you cannot do this now it is OK but if it is a quick thing we will rebuild the plugin with the changes +It includes the test harness that I accidentally left out of the previous patch +It could be that the other end actually does close in some violent fashion if this all takes too long. +This is a plugin which does the reflection to make the actual method calls +And if it cannot determine a type that matches the object it will fallback to a generic typeAnd this generic type is not a that is the type you need to use +Closing as fixed please reopen if needed. +Ready for review +All functional tests involving the use of fireEvent addObserver resolveObservers succeed +Fixed heading sizes. +Id be interested in the test code +Another thought +Thanks for the suggestion! However I would recommend to explicitly check whether there are any field errors for the given field via the method +Well add it to the queue and get to it eventually +closed after releasing version +Please recheck your configuration for that +adamberry can you please verify it on your box +This way the log would only be written once per timer and you would get almost as much information +Bufferring the items could add extra footprint depending on the situation +This patch adds in UGI information to Job for the user that launched the job +but step issue remainsIt is working fine if you log into in different browsers at the same timeFor web applications like console which is heavily I dont think it is a good idea to have more than one within a browser things like Session could be disruptedWould it be ok that if we have this Warning info like You should only open one window within your browser otherwise you may see blank window once you logged in or couldnt log in in your second window +Could you confirm that you have finished the review? I would like to deliver the final patch +Or am I misunderstanding this +Redesigned this to work in environments +certainly looks like metadata corruption that can be resolved by cleaning the plugins from the local repository +RuchithThis is fixed now right? dims +Im not sure using the WSDL location is a very common pattern. +An element may be valid without content if it has the attribute xsinil with the value true +I implemented an and it enforces lower case attribute and table names so the tests failed because of it +Created that gets called by the JDBC importer when the driver is used +Can you try an update please +This will not be done before jbds ga and it is not a crit so i reduced it to minor status +Can this be scheduled? Even if just placed in future +added a comment in pom. +Oh as for the dependency itll take another few years before we can assume java were still not assuming java +Do you have an suggestions as to an existing class into which they could be incorporated andor good examples to copy +Is this still an issue? I havent seen it in some time +No it has been synced but the part you are referring to has not been updated yetWork in progress +Attaching diff to fiximprove namespace handling of xmldiff +So we will need to investigate the impact of returning the code to its original state +If theres something I can do to help out with or any work in this area necessary beyond that patch please let me know +Post a patch +rebased this time committing the encapsulation s a clue the bad magic being passed is always zero +question is why return false +This change alone wouldnt solve since that issue is about a field being initialized too late not about being initialized to the wrong value +I just could not understand given the current code and its age how this could ever have worked +I tend to commit that soon if I didnt get other opinons +resolving confirmed passed rate a while pack issue put to rest +if context root in preferences is set then this is what happens +Good stuff found by the long but useful precommit ant targetI also added in the second wait for nowIll commit very shortly +Linking to related issue +Correct dependencies are downloaded from m repository +I think we really just need to update can understand your concern about performance bottlenecks there although I worry more about people putting any too many values or something too big than I do about the schema itself +svn ciSending mainsrcmainjavaorgapachekarafmainSending mainsrcmainjavaorgapachekarafmainSending mainsrctestjavaorgapachekarafmainSending mainsrctestjavaorgapachekarafmainSending mainsrctestjavaorgapachekarafmainTransmitting file dataCommitted revision +How exactly do you try to execute this script? What menu item are you refering to +Setting Fix Version to future for issues without a set fix version +Remaining issues w JBDS feature version being produced as instead of is in of update site w http authentication is in. +I just would like to known if the Whitespace Tokenizer behaves as expected when a t character have to be annotated as a or not following a character? And then I would like to know if a patch could be applied? I dont have examples where the get method returns a length string +I didnt want to turn those back on without more assurance they might passSo I installed freebsd in virtualbox and started playing there within the confines of the black hole behavior +This looks good +If anyone can cause an error reported in after the patch for please post +Hi +If you can regenerate I can run a number of tests on it that would be good to do before we check it in +Ill remove srcgetstart too +fix version needs to be changed +and Europe +Over a minute? OuchIs the time from the setup or the getcount calls? Suspect doing the inserts in a batch will help a lot. +Also possible all that was needed was a restart of the tablet server that the map size was incidental +But it must be by you very much guysDemetrio +Change reviewed ok perils of +It looks like we just use type error if the type isnt recognized +I have a patch almost ready for submission that would fix most of these issues +Patch version which supports async batch +changes to jcrspi and the spi for SPI and replacing the previous patches for SPI and +Please help +hi Manohar what do you exactly mean by how to verify the values set by these And also you say However there is no usage information +I have some tests that are not production quality because I still need to play with filter parameters +Fix the javac state to invalid may break applications as that was not present beforeWe should to probably set to running by defaultThanks is removed from the lates patch +Sorry for the inconvenience +ThomasDo you really want us to check in the code with your copyright headers? Is it a bid deal to remove them? i think id prefer a clean patch wo the copyright headersthanks policy is to require the copyright headers as otherwise people wont know that we have copyright in the files +Yes but we need to modify API for add a generic multicast could map join operation to session creation and leave operation to session close +Dmitry Is there a nightly build somewhere or do I have to put these changes somewhere of course the next question is where +Im still a little confused so produces classes that are needed to be used by the compiler? +otherwise it is need latest version of jpox to change locations easily +Dependency on URL string File has beeneliminated +dont depend on any other +Im fixing it finally. +If you are not interested please just mark the bug as wont fix for lack of time + +Will reopen the original Jira and attach the compatible backport patch. +configurable memtable flushing period and the initial token of a node +Can you please attach your file that causes this issue +The changes are in VPE H build from March th but for some reason it didnt get into aggregated update site from March th +This will let you replace the block with a switch block if you prefer that sort of thing also +applicationsproductconfig +What does this mean? Is this related to shades only? or regular features +rebased +did not run as part of the precommit build +biggest issue is maven taking too long +The compnent reference says External links take you outside the scope of Wicket +I suspect somehow the alignment calculation for the second span ends up being applied to the first tspan as wellBTW I see the same error in Batik as in the rasterizer. +Since I wasnt doing a good job of explaining my idea I just wrote a quick patch which should illustrate what I meant +Pat and myself are taking a look at it +And if I install neither I get to sort it out myself +You dont think itll break stuff in trunk? If not for sure +Using thread local Macs means that the servers dont need to hold a global lock while they compute the +in JBDS h Linux +An XML entity body is definitely NOT +Other things to add are maximum regions to assign per balance +Thanks +The red line has none enabled +The language spec has to be updated to reflect the renaming of loclocation to home +I will use clouscape migration tool. +Im beginning to think running from might not be the best idea to have a quick dev cycle +Containment is a nice thing to have but yes single containment is quite limiting +Given previous comments I am just resolving this +Verified in jbdevstudio linux installer find +replicated with Windows and JDK final +Moved the test to srctestunit and also rearranged the test methods a little bit added some javadoc +Wolfgang for the byte vs +I might have some existing documentation for this that was written before the feature proposal to +Works with the latest EJB plugin against fix version. +Id be struggling to come up with many situations where remember me authentication is used with something other than a login form and the current framework is pretty much designed with that in mind +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Not really critical for is resolved the continuum builds are running. +Thats just not needed anymore. +Ive renamed the property to give SNAPSHOTS a try in a day or two thanks! Ill test this asap. +I sent a PR for so you could have a look if you want. +Ive resolved +On the other hand if someone unplugged a machine you were using youll be able to start over without any operator intervention. +ShingoAre you still working on this +So if the iframes on browse the main windows url would be utilsbrowse +did you run all the tests? I think there is a test which does show functions which will break and needs to be updated +Thanks Thomas +Ill keep you updated +to get the details of the input splits of the which failed? Anything elseThoughts +so I guess you are asking for better documentation right +Im okay with both approaches +I believe the fix there was to enable Manage Group Filter Subscriptions + currently returns a +Thanks Patrick! +Ill take a look into it a little further and see what can be refactored +The attached files and list all the methods that were missing or incorrectIn the embedded driver methods are missing and one method hasincorrect return value +No feedback from user so closing itAlso this looks more like an old Maven core issue rather then something caused by this plugin. +Checked in the fix for trunk now param is long gone +We dont want to have to check released over and over again for changes the checking should be made upfront checking every time is bad but we should be dealing with malformed better +Worthy of a release note? Or specify that must be set manually in the file in documentation +pending Hudson +fixed in svn r +Fixed one unit test for TOKENIZE in and added another unit test using the query in the descriptionPassed all the other unit tests of +I buy Bikass argument that blacklist should be separate from whitelist and it shouldnt be per +After more days of testing I think I can confirm your assumption +committed revision +Reopen with more details if problem still occurring. +Thats why it cost me so much time to solve the trival mistakeBut on the other hand you have you hands full of more important bugsimprovements +Great! Thanks. +Fixes the default log level and applies changes to the default config +on copyadd keeping the sequence of jars within the container and in context of the other dependencies will be important to avoid suprised shadowing effects +and as clockfly said splitting an empty region with no midkey specifies still behaves as before +passes system test +Verified on ER +We need to update parent pom accordingly since artifactId for targets have changed +Kenny has been like this for a LONG time so lets talk about what the problem is that you are having. +This provides the API to fetch the total no +I ran locally and it passed +Ive seen odd socket errors twice in the last few days while running the unit tests I think in +Thanks for your test case azgard but I had no time to take a look at it yet +Currently when I catch it I throw an exception +However this decision is not really a decision +The code generation code ingenerateExpression might be a good resource for us to make make the code generation changes for NATURAL JOINs join column in the SELECT query +What scheduler are you using? If you are using the default youll run into which isnt committed to +refactored make all tests Im testing on hadoop now +Most of the test cases in the first section Easily Refactored have been refactored and committed to the testsuite +Anyways I think the fix here is to add session to the list of parameters that arent accepted by the even though this may mean breaking some existing applications +must use http client as restlet uses restlet has upgraded to HTTP Client so they are just keeping restlet on the Thanks for the info +At least in the Maven SCM sense of it +Yes you are right +I decided to leave the class as it is and added a new class which can recreate a view when the input or CAS view was changedIt still needs a bit fine tuning but it looks like it works much better than everything else we tried +Please ask this kind of question to the user ML or otherwise provide more details to make the issue reproducible +Fixed +See my most recent all resolvedfixed issues of already released versions of Roller. +It seems to work for me what do others think? In case we can agree on this patch I could still make it part of Felix +If so MUC should be able to run as both a component and as an external component and it should be a module since it relies heavily on discoCould you elaborate a bit on the last part of this sentence they both live on their own separate component subdomain and have their own set of stanza handlers both of what modules cant do +HUB failed +Yes you both were rightFixed in CVS HEAD +Hello GeorgeThe fix looks good +Test cases for Streaming in Local execution mode +Thats what it should be used for but not for the +There is a hard copy so any changes in should be done in shared I think the solution applied in is better so Ill close this one as duplicate +ping! bug! blam +Fixed +Rob could you review this change pleaseThanks good to because I forgot to update the file for Windows to also pick up the plugins +This patch just refactors the code and reduces the amount of synchronization +Also Id like to add that this issue affects not just autocompleters but I believe all Dojo components and possibly other inputs +Send read to all replicas +If theres anything that misses there or if I misunderstood your issue please correct me +theres an empty tearDown method is used +sorry for wasting your time +bug at +Reopening to move to as was issues after move to in +Yes I am able to run ee tests +I think this should be done before +Hm actually Ignore works for me on hdfs tests using the same version of ant wonder whats different about MR here +The other RMI samples and tests in the Tuscany build dont seem to have this problemFurther investigation is needed to establish why this problem only occurs in the travel sample scenario +java files are coming out of it +Seems a bit risky +Thanks +Committed to trunk and +No it isntIf I define a webapp listener the current context will refer to the applications class loader and so the fix wont workThis patch need to be run outside the webapp +Ill definitely get it fixed ASAP +In general this sort of use of service groups is a bad practice +looks like it is Option but instead of updating the znode periodically the new reader will request the from a set of quorum bookies is that right +We unified properties with other logic we have to relativize paths so Ill call this fixed. +Hey RudolfAgreed your right about is not getting a config object +Normally this is fine but in fstab it is a bit weird to have ousetrashoprivateo +I just not want to break build in case of future js release in short time +Does it repro in as well +A Java applet would be a similar class of consideration as the currently implemented Flash capability since we are trying to avoid deployed code over and above XHTML for production support impact reasons +older issues for Apache since were no longer actively working on these at the moment. + applied a similar patch in revision based on Adrians proposalThanks for the contribution! +Work is complete +supersedes the JAXR release issue +We can expect people to have the XSL on their system as a +The links are illustrative only per the Jan doc review meeting new examples of how to use the links via soapUI for example will be added to the user guide +Committed revision +I may need help from someone to verify the fixes +Yes is not empty if it closed succesfully +part of release. +Thanks for the update Ill try to reproduce and if I can Ill the issue + +WDYTthanksnandana +I have completed the Initial work for this task and will attach the patch for this task for initial reviewThanks and Regards AllSince Eclipse Tycho is now officially available and stable we can go ahead with that approach +I cannot reproduce this problem with the snapshot build of Apr +It should only change selectionWidth attribute couldnt done because inplaceSelect width changes according to the valuedefaultLabel attributes +is respected on custom Service class by sunri while jbws native stack respects it only on SEI +Well honestly I opened the issue so I think it was not completely nonsense that I closed it +Thanks Russell + +Thanks Leo the modified patch was applied to SECURITY module at r Ive removed testing of serializationdeserialization for object with certificate +Looks like Hudson had some trouble running tests +If you find further issues please create a new jira and link it to this one. +This patch builds on the prototypev patch by adding code to to ensure that ORDER BY triggers a sort when necessary adding tests to and to demonstrate the behavior of NULLS FIRSTLAST adding a few more comments to the code changesId be most grateful if you could have a look at this patch proposal and tell me what you think +Should be fixed now in +I just committed this to trunk +This is the logj wirecontext can confirm the problem is caused by a bug in +Yes it is at critical +I see +I think assign should be changed to be assign and you should not use this in your createTable method +I think I dont have permissionsHow can I close this BugRegards or RainerCould you please review the documentation changes that were made for this issue? If there are no further comments I will close this by Friday +A support for authentication will improve its applicability to our other usecases alsoThanks is not on the radar at the moment +dead simple +page demonstrating the will deploy new snapshot later today +Feedback welcome +Okay +Ive seen other forum posts with the same question so if this the way youd like to go with the adpter as well maybe an example in the documentation would help +YesWith the snapshot version the tracker always sees the service being added +In general it seems like is one of the only few deprecated classes that is being used by the classes in the new mapreduce package +Added javadoc comments. +missing header added by jdeoliveCheck Copyright is in all files +Simpler way to run director +With a runnable it holds a thread too in waiting for the region to be closed right? Holding a thread should be fine since this is some corner case isnt it +Yes agree +mvn takes care of this. +Stay tuned +Sorry that I didnt post this earlier +Thanks for the patch Lars. +All outstanding issues on this appear to be resolved closing. +You can make the test easily by using which is a dedicated env +Thanks for taking the first step and filing this detailed report +This is fixed by now isnt it +Testcase that reproduces this issue +Start Broker +Working on a backport for +Relying on the order of things is probably not very good +Implemented in AS +And yes surefire has pretty good test coverage +By the way can I have the honor to be the assignee to fix this issue? Thanks +Originally I used common client ejb etc as artifactId +Of course that should be done in any caseAbout the name of the optional keyword something like plain ALL might also do +Actually I think consensus now is to support only jetty in trunk and jetty in +updating subject to reflect the more generalized work done as part of this issue +Please go ahead and commitOleg +Meh Im fine with either +I have removed a few of the attachments to this issue as they have already been applied to the latest codebase +Ok so if we always prefer to bootstrap from the correct token then I still think we should combine get and get +I forgot to update in the previous patch so I included the update in a new patchI also discovered that didnt update and included the change in the new patch +Thinking about this a bit more the when to execute a health check issue is not related to JMX only its the exact same problem when you look at them via the webconsole or programatically +Bulk closing stale resolved issues +Thanks KonstantinPushed to master +Hi Oliver I have retried it on a different machine and now it works +Im putting this on my low priority todo list +Patch to add more logging around split +It is a great things we have regression tests to catch for things like thisI will run only Sysinfo to see which policy file change causes such severe behavior +Fixed. +I think we better not to change the behavior of ant test Dtestcase +I run the tests from within Eclipse and it said Provider not foundWhen running the tests with maven there arent any errors or failuresWeirdpublic abstract class is in both and +Builds on the changes to the shim classes from this patch +Cannot find any documentation on how to add the RAT excludes file + to work properly in +It showed as its IP rather than name because reverse lookup was broke but there was only one server in the list as opposed to the two we used listThis patch should help our usability in general w its fix of the ubuntu exception +fixed code ed to Trunk +Do we need a separate upper level JIRA to trace all the hadoop integration bugs +I am going to close this as WONT FIX +Thank you Chuan. +I dont think it is critical that we throttle for recovery so I was wondering if it makes sense that the calls to read entries from do not try to acquire permits +I think this is a clone of +Closing this record since is now released. +Thanks so much for all your work on the patch +Just return SSTR instead of storing to local var +I actually made it an instance var of the because some code paths through the stub objects dont use the instead going direct to the +Applied patch with thanks to Perter +The marker file is here just so that the deployment poller can now a bundle has been processed instead of retrying deployment continuouslyAn enhancement could be to have a +committed +is updated to compile +I need some pointers to where I should look in the code though +its my bad I should run the test before committed the codeNow I just committed a quick fix for it the tests should be OK now +Chunhui do you have benchmarksmicrobenchmarks from when this was developed or from your app +Further doc coming soon on wiki +Let me know if thats acceptable +Thanks for the patch! Applied with an added integration test in r +A cancel button can be accomplished as just an +Patch implementing the in Rev +It also assigns a style to the h a selector so that it doesnt appear as a link +Tnx +This is also n the WAR file +Issue fixed included in the iPOJO release. +Thx Thomas +Ive submitted a patch for this issue in +Some featuresbugs were implementedfixed for others we have separate issues. +Might be worth trying to do if feasible. +Or add a new type value like PUTWITHTAGS or some such +Grant I committed his patch a while ago +For test could you remove the timeout and run it again +This is now fixed +That is necessary because it is required to parse some files scan for annotations and merge all that information +The optimization in the general case would be to clear this set every time the variable bindings used in the Order By change +I agree that moving synch to the synch classes is better +Hi BikasAn isolated container is one of the requirements for a secure cluster +Any new on this? will import it in codehaus repo asapBut i d need a patch with Apache license header instead of LGPL older issues for Apache since were no longer actively working on these at the moment. +This patch removes all instances of and replaces them with +And by the way the staleFile argument does not put the stalefile in the target area and thus is not deleted on a clean +Verified with JBT. +Cancelling until we sort out comments +OK thanks for the explanation! I have a few other remarksquestions on the patch why do you substract to to compute the maximum number of bytesords? sum looks unused in addOtherwise to commit +The isActive is controlled by which should give when instantiate +no point in uplaoding different patches +bq +Some more questions Are you running any custom AST transforms that might have some affect on type hierarchies or generics? May any of your Groovy Monkey scripts make any changes to the Groovy in the editor +Minor issue in test case causing test failure +I still like this idea +This can be achieved safely using the zookeeper compare and set mechanism for setting the count iterators +Thanks me +above by respect the CRS dimension by presenting as many input fields as dimensions the CRS the component is basically made already thought it has layout issues when mixed with other components +This appears to be a general problem If I type random text in the rules editor and then attempt to switch to the RETE tree view the same error appears +If you think it should be reopened you need to provide us with a sample project that can be used to reproduce the problems. +Go ahead and commit it! It works fine tested and verified +See last comment this is a documentation issue +Ive updated the document and deployed the new site +Can you supply a patch +I see Ill change V as a +If I understand it correctly if some artifact is added to svnignore then it should also appear on Modeshape ignore pattern list +knew the debug mode thing would make it difficult +It should read and not +I recall changing how this code path worked +This one is a diff from trunk as of now +Any comment please let me know +HDFS and YARN use different mechanism inheritance vs translation in protobuf implementation +Can you check this +I always thought that the underlying cache provider should take care of this +Yes this did work for me +Certainly it could have been publicised more +The synchronisation between the worker thread and a calling thread has been reworked in revision . +implements the suggested optimizations +This does not require AJAX for failure in my case +It happen onle times during the month testing cycleThanks for attention! +Im mostly now running tests using groovy junit tests +However next I stopped and failover DID NOT HAPPEN +Ill let this bake for a while on trunk before backporting toThanks Earwin +This patch conflicts with +Thanks Laura for the updates +I dont have that deployment available anymore +This has been fixed for JBAS For earlier versions please replace the attached jar in your JBAS instance. +I also updated the release guide +I updated the site build to match svnpubsub requirements and filed for the required infra changes +However this might not work on Windows because on Windows a native error may appear if different are opened in the same SWT Browser +yes weve got that implemented +Ive simply added an explicit check for the latter +seems upstream fixed +Or should we invalidate the bundle context after removing listeners? +regards in CVS will be available in release This bug is actually due to amp in generated URL +Hi Jeff +If you need a new custom one you are into some serious stuff already +All thanks for your feedback +Same issue in EAP. +HostsServer Configuration it it was hidden by the advanced switch +The other issue with sending events is that the final jobfinish event is handled by the same async dispatcher so it will be busy finishingshutting down and wont process any further events +i deliberately used a different name that we are having in jackrabbit core to avoid confusions as the default login module in jr was doing all of the stuff that is now split into different login modules +maybe im just overcomplicating things Anyone with a suggestion on how it should lookwork for an decent sizedcomplex collection is very welcome to attach a drawingproposal to this jira +Please reopen if you disagree. +It has to do with sub shard replication +Im on this kind of hack going into before we add the feature to HDFS +That is half of the problem +Eric Woods committed this patch in r so I will close this issue +The example is e very simple JEE transaction one +onif you want you can open a new ticket for retrying Initiate if the message gets swallowed as well as recovering if a source is correctlyincorrectly marked dead but IMO those have very small benefiteffort ratios. +Not going to lie that was a pain in the ass +Thanks for the linksNot really clear to me if we need to do all this counting in all the cases but it seems likely that even if we could fix them all now wed find another one next weekWe should at least add a comment in the source where were monkey patching as to whyThere is a bit of silver lining + +Vaidya uses the history files and this issue is blocking it being used +Nice all tests passing +Committed to SVN trunk and +Please review and close. +The attached patch calls getPM to retrieve the PM and closes the PM +It now adds the module to the server if the same application is checked out rather than create a new server +Ideally for a REST service we should not maintain a cookie +has been marked as a duplicate of this bug +Compilation failure should not be underlined at all +Ive a look at the forrest build logs and it seems to be a problem with my changes to the forrest documentation concerning the wholesite feature +Checking equality is much efficient than compare operationI am going to open separate JIRA for this possible improvement +Theres not a hard deadline that Im aware of +Im going to mark this as wont fix +Please attach the generated spec file +Im not sure if this is an specific issue or a broader Roo issue +Right now I consider this issue to be fixed and verified on CR +Ive updated the wiki page +think your earlier suggestion to reconfigure if the has a null configLocation is the correct one +I think this is a specific enough problem that comment or javadoc could explain it in enough detail +A better tester with more test the xpath implementation attached by varuna. +Please mention that this test tests creation or multiple snapshots +I committed this +Fixed at revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The wizard needs to answer three questions Is the provided url ok? Does the sonar server asks for a loginpassword? Are the loginpassword okWe currently need three queries to check that +ZOMG I LOVE THE NEW SITEI also think its interesting that both of you came up with a logodesign idea for OODT that was in nature +Thanks Ted that is useful +a first stab at a patch i got the core tests passing but it still be better and there are problems in contrib +hello +Also should mention that the test environment is JDK +Fixed in r +if this doesnt work then unroll the changeif a project depends on wicket i want to be able to check it out run mvn eclipseeclipse and have everything work correctly once the projects are importedperhaps what we can do is put the bundling into a release profile so its triggered only when we are building a release +Problem still persist even with the patch +The restructuring proposed in will help here as well +Woops I caused this when I added double as a parameter type in Config +Mark after my last comment I added a workaround for this issue by adding johanwalles to the bzr URL in the POM file +pending jenkins +Please retest +Verified in ER +For the changes to take effect job must be rerun +Thanks Carsten for the fix +I dont see any good reason this has to be in a new module +It may take a while given large number of items to work on but it sounds like one of more important features to add +This makes me nervous about rushing in a change in +This syntax conflicts with exploded formal parameter to +Getting a decent JSON parsing library in there would clean up a lot of that response code +please provide a full project structure that exhibits the ve been seeing this problem +I dont understand the description +These attempts should be logged as KILLED attempts in instead of FAILED +Thanks for reviewing guys. +This is s default specific annotations at the method level always override such a default mapping so this should still work fine with a global of course you can always use some good old with specific +doesnt have this issueIn there is file to record which need to check when PS is put into maintenance mode then the SR is removed from this file hearbeat process will not check hearbeat on this PS. +linux is a command +Simple perf improvements should go intoI have to digest the patch but it looks good upon coarse inspection +Can you please test it +Will share it as soon as I move my env to a new box +So it depends on the flexibility of deploymentinstallation method user is using +Hi Tomohiko I gave you perms to edit the wikiThanksEli +Patch applied thanks Rudi +So if list empty and item added we should start our immediate upload +We will handle this through the customer support portal +Confirmed Enrique Ruiz executed CLA with reference number +If you worked on this in the mean time just attach a patch to get the trunk code up to date +To get the code apply the discussions are right here attached to the JIRA one of the principles of open source is that decisions are reached in publicThe general process is that someone has a suggestion andor code and makes a JIRA +This removes uninstalled bundles from the set of bundles to handleAnother cause for an thrown from acquiring the bundle lock is failure to get the lock +Then I run +Delivered to +Verified in JBDS +Just for info Ive been getting this from my JNDI connection with geoserver since upgrading to and cant seem to find a way around it +httplocalhostdocs +Wondering if after all the changes in security land this report is still valid +I can confirm that this remaining issue is related to selecting the media type that has not writer available i have a partial fix that resolves the scenarios described above but break few unit tests +We wont include this code as part of Buildr. +Might this be better modelled as a geoId of type GPS with a fk for this geoId value in the entity +This is fixed after updating trunk and building +I think this should be a blocker for +Still the provided scripts are buggy +What properties should be exported? What did R expose before +Let me see over the next few days whether I can get my head around this +make it on whole cocoon and you will have half of the classes locally modified +release audit +This patch now also contains unit tests +note the missing leading slashquick testing seems to indicate that adding the slash fixes the problem +Example program with endless loop. +Resolving +See for the connector SPI. +Fixed in trunk via. +bq +Would it also be possible to add a general serial field? or is that something that Eclipse should handle separately +Did you do an ant yesterday for the first time in a whileYesWhat Im doing right now is to run another job which fulfilled for two weeks ago +It looks reasonable and I have applied it +This is being fixed as part of. +Testcase works running it integrated does not +Instead of using Solr I believe it is more elegant to add that filtering in the lucene mechanism to add a custom syntax and deal with Scorer Collector via it +Seems to have been fixed +Hi Patrick what kind of DB driver info we should put into the config? DB connection pool driver class nameIt depends on how you want the configuration to be structured +Sticking in in case such a thing ever happens +Please provide an example project to reproduce this issue +Raised another issue for getting this done at WSDLC level +Read the documentation use the forum +Regards. +I fixed the problem in the SVN trunk and the branch +Thanks again for digging in on this +Modularization would be pretty disruptive +If instances are escaping we will need to catch them and make sure they dontDoes that make sense +The patches for the m builds are a bit nastier since there are no transitive dependencies +With this known we were able to gather all the properties file in one place and declare the messages component in EJB package +Comments re Cascading arent strictly true but this isnt the forum to clarify +So that means we still need to take a closer look at this with respect to performanceI am going to go ahead and commit this patch since we still pass the same TCK tests and now also pass the singleton test too +So properly closing this issue is just extending Harmonys test suite and than say the bar is green + +Ive changed the iconDaniel please verify +Any change its possible to fix this tonight +bq +GaryNo code duplication is necessary as long as you use the deprecated Scheme constructor with the deprecated interface or the new Scheme constructor with the new +Ok I figured it out +Thanks John and Todd. +when i run with defTypecomplexphrase i get below errorHTTP Status lucene lucene at how to solve this +But you didnt change the old JMS MTOM test so it was failed if you applied patchI changed the old JMS MTOM test to catch that JMS configuration exception and also updated the exceptions message for itPlease check them out +all the validators included there works. +verified in JBDS +Marking this as resolved on trunk see for the actual fix +I agree and came to the same conclusionIm currently working on a fix for which does +batch transition to closed remaining resolved bugs +Presumably this will cause a problem when the JMX monitor tries to monitor the CMs but that will have to be fixed another way +Cool thanks Jing +bq +I tried to debug this problem and I have a fix for it which I have attached along with this commentThe was not implemented in a way to show the line number of the invalid referenceI did not want to touch the existing code without fully analyzing the implications +Could you make the necessary changes to revive that build target? Im busy tracking down issues with +Hi guysWhat do you think about generated JSF files could be customized using any visual designer like Tools +The framework tests are organized as follows the core test package consists of tests and a class the agents test package consists of tests and a class that extends the core class the test package consists of tests and a class that extends the agents classThe question is where should the classes live? Will maven do the right thing if these are all moved to coresrctestjava agentssrctestjava and srctestjava respectively? Or do we need specific and subprojects for the test base classes? How is this usually handled in Maven +This Jira was about making changes that were mostly already made by me in December +Since I dont know any way in hibernate to get around this issue we are creating a stored procedure as work around to execute the statement +Patch applied as rev +though I wonder how this was working before +work to add a simple connection the to add the right Fix Version +Not that it matters +to apply am OK with this compatibility break given the advantages +No particular thinking except other converters had the same default agree to Markus that the behaviour of this method should not be changed suddenly +Thanks +discussed with Namit +Evidentment la qesti no t gaire importncia +easy workaround dont use th fully qualified class was just a warn +Maybe we should add some flag to an endpoint whether it supports same endpoint but with different parametersIn the seda case it should not do this as its confusing so we could do sedafooAnd when you do a getEndpoint with sedafooconcurrentConsumersYou will get an exception that sedafoo is already registered +Any commentssuggestionsThankssuersh +Awaiting user verification of beat me to it +classify and set fix close of issues +i think is best if someone with write access to svn simply gets the files from and checks it in +then realized tests has not run +Yes but I think you mean in place of the second on the bottom line and in place of the first +Were effectively working around a potential in persistence providers. +It compiles without error against as well but my entries are so I left it at that +I attach new diff with only this task modifications last patch has other modifications +in top level cross is not a problem since will take care of the uid reassign +Todds patch for trunk +Do you have the property set to true in your config? Yes intentionally +rebased merged patches +Yes I can get this done in that time +to +Im in favour of +James is on a roll now so he is also doing the dataformat as well so the flatpack gets the whole deal now +Chris good stuff so farI have added a couple of final keywords and prevent some autoboxing in as the char int values should be mainly cached anyway +Thanks for the feedbackWhich form doesnt display the validation errors for the email addresses +The new pom xml configurable parameters are groups groups to run test in excludeGroups groups to specifically include from run suite List of optional testng suite xml definitions to run threadCount Number of threads to run for tests parallel When using threads whether or not to run them in parallel +running the CodyI havent been able to reproduce this +I also agree with Niall +If someone cares I can revive it but Im moving on to. +But we still need to update the code before we can commit +Fixed in CVS please reviewed the code fix this looks like a working solution +Heres a patch for +AgreedI wonder whether we might be able to simply the implementation by using only for attributes rather than the complete hash map +Will do next patch you added is missing the new test case + +The idea that multi transforms ops into results is nice and simpleIt is true that it is hard to imagine that the Stat structure that comes back has any new information in it +Ill be monitoring this bug let me know if you have any questions +Presently users can only be disabled by the administrator not deleted +c sI am not be able to reproduce the failureCan the originator or any interested party update the test case with the necessary use case to reproduce the problemThanksAlbert Lee +mostly need to write some edge cases but the intent of this task is enough for +for example consider the following directory docsprojectsimages and the fo file docsprojects which contains a relative at imagesif you run the ant task and set basedirdocs on the fileset of +Im not adding any code that is specific to Haivvreo to Hive nor is this patch supporting the change from Haivvreo to +Carl any progress on this +Ya I thought thats why the patch was full was from the formatted code +With pleasure +Attaching preliminary patch +Is it a correct assessment that this issues affects begins with the initial implementation of i in? If not when +The changes required to get rid of the old running in +Do you want a test caseI can only do my own validation if the validate method is called +Punting issues from RC to final +Thanks for the unit test +Added the security settings and did some testing +The example in the documentation is actually the preferred syntax for specifying that injection should only be performed into a method with a given Java return type +Then it should be easy to upgrade to this new clover keep you in Freddy SimonPlease let me know if you need any more information regarding upgrading to use the +Changes in patch original patch compared the SOAP action with the operation name wich is wrong +That seems reasonable +Is there an easy way to integrate this with the release pluginA system property seems fine for using the scm plugin directly +Docs Q I am assuming this will not change what users have to do +Time to start pushing for beta +The problem is additional sync between every vertex computing supersteps +thanks mark I was goig to ffix this yesterday but got distracted by a geotools meeting of doom! Ill try to get it done m thinking the most efficient way would to have locks +some services like create dont have any permission check in place at this time + +Im going to merge these issues since they affect the same code on similar lines +ill add your process to the test suite and check if it works +MartinThis sounds like a much better way of solving thisThanks +I suggest we do not flag this for the M releaseIts necessary work but independent of M +Yes this does not happen with maven any more so it seems that this is fixed now +Fix integrated in master. +Also I will backport it to through codelines since it has failed intermittently there too +deploy resources under So if a developer wants to override a resource they can simply create a custom version and put it in their webapp + +I dont see why listeners are some special case that require more isolation than all of the other state we expose +Sounds like we need some kind of cause method that returns the wrappered problem +Committed to as well +Im closing it +Rename README to be Both these changes make Hive compatible with Hadoop Core build environment +fixed in latest version +comes to mind as similar problem but I have also not been able to duplicate the error in there +Cant we just have an scommandButton that drops the useless immediate flag and allows me to control the processing phases more +This is a quick poorly documented no testing implementation that mimics the current Standard queryIt is mostly intended to flush out design issues +for splitting btw +Thanks for the patchesJMS is fully backwards compatible with JMS so this should be okay +Attached a trivial patch that applies cleanly against both the branch and trunk +I was unable to recreate the problem in testing but there is a fix in trunk that should work +targeted for bug fix release +Im not up to date on FBs distribution +Too late for the th might be able to get this in for the th announcement though +It seems that this issue should be closed as it isnt a bug +can someone point one out to meKim thanks for opening up for this issue +Only difference for me is that it throws a with a different class in linux an additional information I was able to build Castor in the morning of +Regressions ran fine without failures +No because theres a chance the table is in the process of being created +Milinda has fixed this +Looks like a duplicate of is checked for both types of users but due to the password check is skipped when the username is weak itself +If we could just exclude classes and lib this problem would be history +Im setting priorities to reflect what we think we have to fix for Suns JAXP Beta release high being a must fix +Duplicate see all old Resolved issues +I could add an option int he to let it return the value of so you dont do additional hacking on the +r which became r initially failed but succeeded ona nd run. +Committed +Updated docscomments and checked into develop branch. +Please see for a followup to this issue. +Per Dans suggestion I moved the stored procedure methods to the new test +This is a bit tricky actually +added in revision refactored to use it. +Committed +Your setup is simply ambiguous +Do you have a test case it +Mahadev helped us update our POM +Setting the target version instead of the fix version +I added based on the test case provided in this bugThe attached patch includes both tests and fixes for and +thanks +The second one is what uses to break priority ties or in cases where its just going to batch +If you have this you dont need the separate Fairshare and Capacity Scheduler projects +Is it compatible with ASF license +In your example youre mentioning as the only possible way to send messages to multiple recipients but in fact there are other such as and so on which offer a more granular control over who shall receive the message or not than a channel +This looks like a case of slow consumer and producer flowcontrol +Web site content isnt versioned and the eclipse pages now accurately reflect so am marking as the fix version +MattIve committed an update to CVS with a property hide which should work for both JDK and CGLIB proxy targets +Once I have something I will go ahead and the Piyush +Lukas your fix doesnt work as expected and probably it broke something + +If the property value points to a file or to an HTTP URL it is silently ignored and the default is used insteadThe added log messages should help troubleshoot this. +We will accept this as part of ER +regexp permitted zero or one whitespace characters now permits zero one or many. +Yes that is the correct url for jffi now +A disadvantage is that it does not allow some requests at the same priority to be and others notIm not convinced yet on which approach to take +thanks +Regarding patch this patch was done against the tagged version +What i missed is that the above happens with Maven as well as Maven +This is a fairly disruptive incompatible change but can be done at the same time as the package change +assigning to Andy since this seems to be Spring Loaded this is already addressed in latest version of we just need to upgrade +Both times it passed +I agree with Ulrich +was working with the snapshot a months ago +Log into CSP with the user +I dont see it as an issue to have to name fields uniquely on the form in fact it seems like a good thing to meYou could raise a bugzilla for this after this enhancement suggestion is accepted it might anoy some people if you raise one against a feature that isnt yet in Validator Also Im still not sure how Id integrate this into the javascript that this extension is generating but if it does get into validator you could submit a patch with your bugzillaThis is nice workThe version TLD lists Multibox among its tags but the is not included in the downloadHubert +to the patch +Is this resolved +Would you have the time to get this fix in +My first idea is to add another preference checkbox to enable this functionality +Instead use methodsand schema with id attributes so the Xml Security library can use methodThanks for reporting it +I was wondering if I should commit the patch but that doesnt sound like a good ideaCertainly there are other issues with higher priority currently +only with allowFlashtrue +Also once imported into the nd project and after review a bulk move should be enough to then move them all internally to the current XW jiraGo ahead +And im fine with this so long as its configurable and works consistently for macros directives and references +If you find further issues please create a new jira and link it to this one. +Changing the priority of these issues which do not seem to be major issues +So this can probably resolved fixed then +Applied the same fix as for static references as did for dynamic references in Rev +Sorry +Hi Marek Im interpreting your consequence answer as The transaction object remained in memory resulting in performance and resource issues +actually i might change it to always throw the exception again but add an error code to specify failover or not +Just for the record I think I found a workaround for this issue +Ive committed your patch +PROPAGATIONNESTED provides exactly those semantics however it will only work when nested transaction support is available +Of course there is +Looks like nobody else got confused for a while resolving. +The existing only inspects the property restrictions of the Filter +Link the gossip and messaging service so that invalid connection pools can be shutdown when a node goes offline + +Are you implying this was traced to Log messages on? Of course there are a lot improvements to other parts all the time +Resolving this issue. +Attaching here since it was done for reference +the behaviour seems consistent between the two versions for me now +Unfortunately would this be a backwards incompatible changeAlso we need to make the web UI handle thisProbably throwing an exception is the most feasible yet correct solution +I now understand what is happening but this is unfortunately confusing. +You can see the Related Issues on top right hand side corner of the thread +Strange +I have implemented the second solution suggested by Frank +Thanks! +Lets please do this using a helper +Add cleanup action to each catch block as Ivan suggestedCommitted to trunk at . +just wanted to add a JIRA for it though anyway +No it is a signed int so it should be fine for it to roll over every billion requests per connection that will take a while +This change now works for me when building via maven as well as STS +Closing issue which has been resolved for some time +I will upload it to the new JIRAAlejandro check for the doc patch in the new JIRA +Did you experience this behaviour yourself? Ill also try to test this soon as I get some cycles +As a bigger solution to having to ask the users to drop and recreate the triggers during upgrade it will be ideal if we could internally do this ourselves +Thats an error + This looks good to me +Mark I dont have strong feelings one way or the other +If we make that work we may also earn a version number +That avoids the slash issuebqbq +It is easy to reproduce in our environment we have several failover stress tests with heavy message traffic that we run in a loop. +Then check log for failures +This is not a duplicate but the reporter did give away a valid fix to it +Currently you need to have commit rights to be assigned issues +Also things like and are positive steps in this direction +submitting patch + +Ive implemented native support for H and Oracle +Yeah thats correct Oleg +Replaced cxf spring jars with version +Some changes +If your test fails on but passes on then the problem is if you test passed in then it means something wrong with the test +Attached a patch to prevent unnecessary WARN messages +tests and TCK passed on all supported databases successfuly +Great idea slating for +Im testing on a Windows VM +Also adds some extra junit testing for scenario where the Impl class has a namespace but it is not specified in the or +Patch meant to be applied to the branch +Or I can assign to me +Update mapred package as part of input tools compilation +Install JBDS from +the patch looks good to me +so next time it return null +Almost everything has a default and this will be reflected in the updated XSD +The new passed to get are exactly made for this problem. +Added the as explained above in Rev +uuidtools gem updated to now works with trunk +Also how does the stuff factor into this +I assume that for unknown reasons is not available and thus the compilation fails here +it flushes as much as necessary to stay within the prescribed bounds +It have to be there because it contains implementation of their RM client side +I then triedsetting Startup Poller back to JMX Poller and found that if I also set Start Timeout to zero then I am able to start +The class is cleaned up a little +When the timestamp on this file changes the file is validated against the +Ive attached further patches fixes compilation errors from +So if Shane wants to maintain it that would be great +Thanks Ashish +Thanks +Yeah I can imagine having a billion distinct but they may not be the logical entities you recommend on +backported to is this one closed for or not +In case people wished to use engine they could even use REPLACE DELAYED which seems to nicely fit the requirements for thisTo use these I had to implement a new custom extending I could contribute this if theres interest but I think there would be more value in fixing the existing and letting users override the full SQL statement or plug this logic in the existing database autodetect I guess other RDBMS could have similar tricks to use +Please review and give feedback +This may not be a bug in Wicket did you ever consider that the new JS implementation of Google is broken +Improved documentation really should be included for lets try to keep this issue on the todo list. +That will break code anyway +Unable to confirm on Pro and +Making pool volatile should take care of it but only on as before that volatile didnt prevent reordering of readswrites around it +When it gets a null simply continues +Also does a bunch of renaming in filter dirs +As I look at the code it doesnt look like it would be too difficult +is very different in compared to +Thanks Ilya great reviewI have a question In the Compiling source code section you say developer settings arent required +Agreed that it doesnt matter for power user +fixing would also fix this issue +Rebased and merged into the federation branch as part of +Ive added first implementation in revision for the jetty implementation. +Check configuration revision counter before providing configuration +It happens for of my test runs +This patch adds a unit test and updates v of the patch to match whats now in branch +validation strings for simple chinese +Im good with splitting out the WS Client bits into a separate wizard +Attached and modeshape configuration else I found is that it appears the database is being setup twice +Can we rewrite the patch this wayYeah I think this worksbq +This should have been fixed by commit blockers and critical issues are resolved but not verified +for the extended patch +It assumes userhost as part of the command and I dont need that but if the command is a scipt I can always ignore that argument In case youre wondering what Im trying to do is to call a command called clearfsimport to publish artifacts to a repository located in +Hello I have this problem still in version could not be possible to solve the conflict in the second way? I think is strange to mix in the same context user application and framework specific variables +The keyprovider specification has an additional superfluous yes we did change that interface +You could also extend from the base class of existing authorization tests inAFAICS it provides a number of utility methods as well that should allow you to apply access control on test nodes +I have never heard about PIFMT before so I used classic ifdef here +How we resolve udf and udaf evaluation functions or evaluators given the argument types +I changed the offending s to require instead of including itInstalled a update zip into Eclipse and Designer worked ok +Also disables automatic dependency loading after boot and disables reloading code on every request as these are fundamentally incompatible with thread safety +I think this already exists in need to check the was fixed in. +Another problem of character escaping for you Jochen +Its now in trunk called +But there is one more condition where the RIT was getting a null server destinationThat is what i wanted to address +Priority neednt be major though since theres a workaround +Applied to trunk and branch +If anything is wrong the operation is retried +I will add the to the templating API and will then reassign to Dart where he can integrate to the IDE +Updated patch to set the max cached instances to I gave it a bit more thought and Im not sure what the best way would look like to track and support different connection sets is so I propose we go with this patch as a definite improvement over the current set up and leave further improvement to future work +From what I have read online it seems like git is the rightonly way to generate patches should we choose to preserve commit information +Introduced option to control which parser to run +Yeah heres hoping that it does in fact just work +Yes this is the simplestbest way to do this +I forgot to resolve it. +Applied the patch +Ready for review +Looks like this is fixed in bndlib version updated BND to so please close this if it is fixed +Committed revision +Makes sense we should always abort if a flush fails + The motivation for the feature is to provide a more comprehensive materialization m not sure if this effort is necessarily tied to materialization as the current mechanism for bringing in source tables as temp tables is sufficient since the they can be pulled in at a global scope +I agree +it would also be vary helpful for us to have the FOREIGNKEYSUFFIX would love to see this too +Im sorry but Im not sure what you want us to do +committed +Looks like has been converted to just use implementation so it has exactly the samePlease open a new jira linked to this oneThere does not seem to be an option of reopening this one +Im only going to suggest that we use tmpDir instead of rootDir to be consistent but Ill make this change when committing +Thanks ArdPlease post a mail to the developers list to discuss optimization of +That should be good enough for Grails since youre subclassing anyway right? And youre explicitly depending on Spring Framework not requiring backwards compatibility of your newly refined subclass with Spring Framework or ? the Grails needs through a build template method extracted fromAlso our for Servlet async processing now obtains the from its locally request handle and delegates to build as well now +creating the extra html pages which could be many therefore If both approaches were available I would probably prefer the abstract technique anyway! Plus content areascomponents can be made optional to override if the getComponent method is not abstract whereas with the multiple childextend this would require extra tag semantics +is it submitted to maven repository or not yet +Marking it resolved I just dont know when it was actually resolved. +I have attached a patch that resolves this issue +Patch guards in process against regions that temporarily go missing perhaps while in transition states +Furthermore Hudsons delta calculation says it fails before applying this patch +Can you reviewthanks is the same issue with +see subversion commits +Scott do you have some code that triggers this by just running it? Such code could be a useful test case +Added description of changes needed for the Admin Guide +Parent issue has pointers to tests listing where started going bad + +HiI am also getting same error this issue is still there jasitha +if you create a patch Ill apply it +The reason that I said the nocache header was set was that I read the catalina log and saw the nocache header got set +want to change the fixin version so reopen +IMHO we can discuss all the possibilities and choose better approachThanks Sijie for the detailed analysis +Im not even sure if this issue should be kept around given the responses I got to it +there is too much base on that kick and wont cut well through bad computer speakers +The go renderer does not exist anymore in the code base +I introduced this bug in my recent container to screenlet change. +Ive committed this +Darran do we have a test for this +Closing this JIRA. +Thanks thanks a lotSent from my iPhone +I am providing a patch file for the proposed implementation +Assigning to Joe as this is very closely related to +I approve that +Thanks for the reviews +policy file for itHowever due to the current behavior regarding error for or XL A lock could not be obtained +This patch was actually applied by dave kerber on and is available in the latest on the updateDev site. +Fix two small bugs Not closing the plan tag when the logical plan is empty The was closing the output streams even when it wasnt the one that opened them which meant the closing plan tag wasnt being written out +By the way does productization use beaker or not +Thanks Jing! +Also before committing I will need to resolve +Im concerned that there is a more fundamental flaw in what Tapestry is doing in terms of encoding or even encoding the HTML itself +Nothing came up when I googled this issue your link is either missing documentation or an issue that was created during development +Moving this optimization to from to just after merge of old branch into TRUNK +This query and similar ones worked correctlyprior to my work on ROLLUP +I wrote a test and it failed for a different reason see +I have just tried it and it appears to work ok for me +Update the ftp wiki page for the new option of fast. +to trunk +Thanks Xuan! +Ill have a could you do the patch on since that is where the tree is actually at right now +BTW There is also a minor bug in the layout appearing in Opera and IE The position of the vertical seperators in the navigation should be a bit higher +Byteman based unit test is +Im not sure if class ordering is preserved on different platforms afaik Surefire uses filesystem ordering as default +Fixes fact that can return negative value if hash returned is +This configuration is invalid +Im not positive if the patch will delete and or just leave them as empty files +The disk binary is used in the rev id on the COPY destination but the ejson body is used when the doc is first createdA related behavior was noticed and brought up by Kxepal +we should make another jira issue and follow on solrcloud host is trouble spot it will generally be different for each node if this goes in zk that type of config is still a problem you have to use sys prop substitution and require system properties for setting these particular settings +Thanks +Please review and provide your comments cluster online has encounter the bugAfter disable and delete then create the same table We found the region in deleted table is still on the meta with inforegioninfo missed +However I will follow whatever strategy makes sense +deleted suffix +Verified duplicated by +Here is a patch that fixed the problem for me + +I know that jackrabbit and JCR exists but thats nearly my complete expertise in this area currently +Will we see any functional impact of this change? For example in the JON console + Add new +is capable to unmarshall Table and Column annotations and add it to a newly created jdoContent field +Thanks AdrianYour patch is in trunk revision +jprofiler cpu chart when should already be fixed in undertow upstream which will make its way into WF this week. +I think this reply should make it into the XML +The Req class includes a URL list +Updated version of the patch that addresses my earlier comments +LGTM thanks Akira and Colin +I vote to remove this legacy piece of UI +Upload screenshot +Could I ask you run the ear out to disk on your setup and double check it when unzipping +Better to return module references at all these cases + +It needs more work to be done mainly to let configs to be propagated to +Sorry to be a noob +no feedback provided +Closing old tickets which has no activity +Sincerely by Attempt to add id property but already have property private registered as id +you just lose your patience a little +Thanks. +What does this mean is there a ANSI standard that implements a pretend ANSI standard something else? Whats the advantage of implementing these utility operations as DDL? The same functionality can be acheived probably at a lower footprint by using procedures +Thanks for the time spent Grant! +I want to work on this one +truncate does seem like a file operation to put in +No Ill admit in British English it probably isnt +I think the reason that you are getting the access denied is that there is no permission for the Tuscany code base on your app server +However these rules must be kept in mind when doing future development. +Hi Aaron thanks for the patch. +Repackaged patch to include test to +Why do we think master log splitting high priority when it could very well be slower +Just committed this to trunk and +Hibernate should be more verbose IMHO +Note that the href attribute is only valid for the style RPCencoded +So this issue indicates that clients keep a dense list of regions and locations +I have a heap dump and lsof output running which I will contact you to see if you can utilize +Please add new Jira for specific requirements. +They were inspired by the failure of a stress test but that test now passes without the changes +For authentication use Spnego when talking to the NN which then redirects with a delegation token +Inserting fixed it but I dont think should produce whole as tokens to begin with +Applied patches but the last one that doesnt apply correctly +But the number of documents returned is redundant data if you have the number of matches so it doesnt seem like we want that extra stuff in the log for every query +Yes they are blocked +Wow +Tim Alright thank you +had and and recovery of lease seems to run nice and promptly +Ofcourse it should only be done when internet explorer is detected +Could you tell my why you think its incorrect +This is an edge case but I have had it happen in production with a database that had crawled to very slow writes and pull replication +Its because of Embedded and testsuite +One option is to use for locking we dont need to worry about leases since supports ephemeral nodesThe zookeeper quorum can be specified via some configuration parameters and they need to be specified for concurrency to be enabled +Fernando just reassign to me if theres anything else I need to do here +Works fine under +I will add time limit properties as per your suggestions +I have another JIRA which states how a new connection is created for each sstable transferred +Looks good to me +Hi JaikiranThanks for your commentsI think we should use the asynchronous invocation from the web container to the EJB we have multiple tiers defined we may have to define how many system threads for JSPServlet and how many for EJBMDB JMS and +Committed this to trunk and +Hence can not make +Integrated to trunkThanks for the review Stack. +Hi Dan thanks for opening this issue and for the patch +Iow feedbak will be limited +So the server must cope with this situation appropriately namely the default port number if required +Sorry to come in late the patch has gone stale +filipe Awesomejanl Id agree with Simon that staleonce is too clever and could be misleadingChris okgo makes me laughSimon So far I think this is the best proposal +resolving as wontfix lets move the memory optimization into another ticket +Sure Vinod Try to look if i find anything else and will add into thisThanksMayank +Attached sample app browse to httplocalhosttest and to be already fixed cannot reproduce +We should probably just do it ourselves and not rely on one of the slightly more esoteric bits of the spec for it +Tested +John thanks for jumping in here +I also thought it wasnt a legal layer name but server seems to allow it +Didnt make it into EAP or +Committed revision +Thanks Christian +Changes were committed to svn. +I have just committed this to Trunk and +It will affect all the EL resolvers whish work with java methods JSFSeam enables the static methods to be present as EL operands for Seam projects but disables them for JSF projects +Many linux distros already package it +bugs fixed in the last version +in the jar into tomcatserverlib in the is doing the same for the Hudson as resolved +I would suggest that num is used as a parameter for readField +I would probably need to copy and adapt large code snippets of JVM which is legally not an option +But rather than removing the I would turn them into comments since they document confusing behaviour +Could this have an impact on +Verified this bug with latest build from ACS branch +It does seem to solve the problem +Could you provide a sample testalso that patch is not good enough since we dont want to creat a temp dir on user directory since it polutes user directory when fails +If applying through eclipse select the reverse option +When I initially tried this I was unsure as to whether or not it would work but as mentioned it did +just committed this. +Sorry I forgot to mention which I believe would also need to be closed if this syntax is forbidden +Moved to upon release of +Oops +I assume the only reason I havent hit these yet is that Im not having dynamic resource creationcheers ChrisI tested the patch with my message broker +Add printout of the timestamp +Is already below +And added the build version number to the jbossas kit +if I only include the jar to the WAR +You have to run the script first it creates users +Higher risk higher effort and not for GA +Add clone method back to Option +This issue has been fixed and released as part of release +Sorry it it has been a while +Prints Accuracy and Reliability stats plus standard deviation of reliabilityAccuracy Producer Accuracy includes unclassified resultsReliability User Accuracy does not include unclassified results. +Trying v against closed. +is there any way we can get access to this documentation? I can not seem to access +Could you confirm that one of your file in this project has no comment and no line of code + +Looking forward to getting my Mahout on +This has shown up an issue related to that is still being worked +Ive found that in even though implements the interface the receiver instanceof evaluation returns false! Thats beyond my understanding + Changes reviewed and it looks good to me +Im sorry for not describing detailsI this issue to request a hudson account +Regardless it makes the most sense to figure out how to run this through the resolver to determine if attaching the fragment is possible since trying to detect it requires knowing something about resolving anyway +Thanks for reminding me +created new issue +Includes changes to POM and related to package refactoring but not the package refactoring of existing classes +Performance looks much better in my testing tooNow with all these changes its probably time for a beta release to encourage people to start kicking the tires and make sure we didnt change something important. +fixing for +I am already considering changing to for the name +Uncomment all of save +Hi What is the dummy iv used by MRI? Unfortunately due to legacy issues we need the exact same cryptic being generated for things to work +Fixed as of revision +Patch adding comments as useful chris! I will commit it thanks! +Please add a test caseThanks +I got a bit deep into the code today in the and have added options for idempotent so you can keep track of already consumed filesSee my new post on the nabble forum with this ideaJon do you mind review the commited source code on trunk +Thank you for updating tests ArcadiusIll commit tomorrow if no one objects +I dont mind adding +We should only use from command line programs for output dataAlso should we add a test or two for the new functionality +Restart doesnt work for me either +Salut Arnaudcan you please open a new JIRA ticket for this new bug? We need to change the API used in WTP integration plugin to make it backward compatible w eclipseregardsFred Bricon +In this time period it could end up picking a broker partition that is unavailable to complete the send operation +Looks like this has already been fixed in master +but the general gist as it related to this jira is that if things are streamed across not using protobuf for anything but then we wont have any copies in the rpc response layer +I have change all the import statements in the code to reference the new location of +This appears to be a bigger issue in that we dont seem to to have the equivalent of the that we had from previous AS releases not only did that valve clear the thread association it also set the new association on subsequent requests that did not require +how does the CMS portlet generate the next primary key valueAny clue for a manual fix would be appreciated its a thanks +something which would be very appealing to users for obvious reasonsThere is a separate Jira issue on this topic altogether +Lets open a new issue for the equals +Linking to which is the last time this stuff was touched many moons ago +This zip file is a complete replacement for the sample +fixed for for +you arent going to handle a full disk very well as at that point your constraints dont get satisfied +Strangely enough Solr logs said nothing +commons vfs core in trunk now provides webdav support using Jackrabbits client +How would you use these cue wordsdictionary wordsafter youve already done your NER or attribute discoveryI think this is a little too similar to either using rulesfor everything or basically including whole feature vectors for eachdecision madestephen +Sounds goodLets get a green light from the buildbot + +Closing the bug as this issue is not seen anymore +Addendum is fineDo you want to default this off or have it under status +Please use mlreq to request that mailing list and well get it sorted +I still need to add tests to this so the patch shouldnt be considered ready for submission yet +Most of the other pb serializations to pb seem to have it though +Try it now +I got it working now with binary files +I have been thinking about this more and Ive realized that perhaps the easiestbest way to implement this would be to add a request to the clients +I am also glad that I am not THE ONLY PERSON WHO IS EXPERIENCING THIS ISSUEll again apologize for the fact that there is a bug that recently appeared in the driver that have now acknowledged and fixed +log and LDAP Konrad for the logs +Oversimplifying just a little it counts Thrift columns which are CQL rows. +WhtI will cleanup the fix and commit later +BTW congrats on the issue number +How would that work out HeathDo you need to explicit enable this functionality by setting some endpoint option firstAs people may as currently have some data in the body already when they invoke the sql endpoint +Going to commit in a bit unless there are objections +No with the current algorithm we avoid keeping all of the groups in memory at once so we never know exactly how many unique ones we hit +Thank you for the catch Laura! A slip of the finger +How does this relate to +I am currently traveling without email accessand will be able to read your email by th JanRegardsPranav +get seems to be the principal factor behind the poor performance for operations involving mailboxes +Patch committed to branch. +However only the lib is needed if its being pulled in to the bootclasspath +So all issues not directly assigned to an AS release are being closed +Works fine with patch doesnt seem to work for me did you try it +I think youre right about the timestamps not necessarily being useful to a client +I need to look at the linked jiras to get the big picture +Straightforward refactor +For example changing something in the management security configuration etc. +Done for Migration guide and GSG +thanks John +Of course after splitting up the maps we dont have to add the static constants to the map anymore +Closing. +Please if this is still important to you reopen it and upload the whole project with all the entities. +Fixed in changes to +thanks +Christina could you check that you dont have any file in the target directory of your Maven projet? Could you also take a look to your file and search for any configuration? Freddythanks for tha fast have checked the POM and there is no configured +Would be great if you can provide your thoughts there +It may be easier to guarantee increasing keys within a single statement but not across threads inserting into the same table +Please try with and attach a quickstart +However note that this does not qualify as a bug since s resource is not a direct equivalent of a Velocity config property but rather has its own semantics +That way the delta between trunk and is small +I will close this JIRA +Much appreciated to fix this one! Thanks +I just wrote how JBT works and why +Steps to test apply patch run +Good catch Kathey +Im on it +Although I was unable to reproduce this kind of error on the client I did manage to generate a similar exception message on the service side if the client dies SAX ExceptionSocket write errorI believe Ive fixed the code so that such cases of an wraps a wraps an can no longer happen +The skus have been added to +This issue has been fixed and released as part of release +Yet the GB was still transferred fully to the client and then discarded since there was no waiting Call object +Thanks for your reminder +Thats not good and leads to a misleading status like KILLED +Removing the Caching and the releasing of the connection +I can confirm that the issue is resolved in the +I dont have answers but more questions +It could have but nowhere that Im aware of +Thats the smallest download possible +A better design is to write all the locations in parallel and then make a policy decision during recovery as to go with higher txid or the lower oneExcept that the remote NN cant see the local edit dirs so no such policy exists +I tried out the solution and it works beautifully +I see many methods use arrays as their input or output parameters which can substantially benefit from this optimizationSee discussion started in +JDK was +This issue also includes Issue but more +Yes its the same behaviourBoth renames work +Or we would activate only the totally clean rules +However logging did not work for me at once so I decided to evade this problemHope someone finds this useful +I will file another JIRA for it +Ill also try and see if those logs that you pointed to have any useful information +Does it show the same behavior using the Apache Directory LDAP API and JNDI +And in addition we have a button on external tool bar panel for CSS dialog and it disappears if we are not on +Patch with unlock key option submitted +By the way +I was asking about dropping the key length +Template file does work but you have to set some values in the pom and some in the template file +SET CONNECTION statements are uncommented as well +The Hadoop TEST are being published to the Maven repos thus easily available +Using JSF standard request scope for on. +Test docx file and the output produced by is still a problem with Tika and the latest trunk +OK thanks for testing! Resolving this as a duplicate of some other change we did for. +Setting the audit context to a rolling file appender will achieve the intent of this issue +Ran jdbcapi test under JDK and JDK as well as derbyentclientmats +Derby is used in embedded mode within our applicationIf this additional information is still not sufficient let me know and Ill look into creating a standalone app that demonstrates the problem +But I still think the default option should be to install OUR VERSION of the rd party. +Option looks good to me if its feasible +One example is the followingimages pelt skin is ready +Fix UPDATE sentence based on wrong understanding of. +This is most probably a dupe already been fixed +I reverted my changes from revision with revision with a FAQ entry as of revision +no +I do not think that it has something to do with IE or FF +So essentially this is an iOS issue not Cordova? Probably need to confirm in a app with just a for iOS Cordova doesnt do anything fancy really +Will think a bit more about the two issues and maybe do some changes later +Right now errors only have request scope +to trunk +thanks Harsh +Perhaps +HiEricwhen I read the source code ofI found it was still exist in JackThe JIRA was designed to remove jdbc dependency from hicc as the first step +And Can you more detail explain what exactly the html option mean +I have used the range for defining dll names and the block is reserved for the core +Naturally I can store the information elsewhere as well but didnt know whether it made sense to keep the info w the matrix +We arent in a position to make up arbitrary language files ourselves without the appropriate knowledge +Couldnt we simply replace the composite site for requirements by a set of sites in +I tried to figure out what goes in top level and the ones need Organization Party in context +Mmm Im afraid were seeing the behaviour originally described in this issue again about minutes after a restart +jar tf works fine and the jar is exactly the same as distibuted by I tried three times and it didnt work +Not all listener threads shutdown causing a memory leak while running extended tests +Patch updated +This patch resolve some tests on JDK +We have nowno intention to unblock the IPRegardsHenk HenkI do not by any means want to seem argumentative with you over this matter again I respect the ip ban and am not going to ask you to remove it +Taking the as is not a good idea anyway +I have modified doc +The class name change didnt get make into test +Thanks Thejas for review! +This looked like an issue when the in trunk was cached +fix for this issue and also some formatting changes +All posts of the RF forum had been reviewed +The link is dead again. +And yes these patches will be ported to +Attached is a patchI have tested importing and deploying the application with AS and ASI havent tested the application with AS because it uses Seam which isnt compatible with AS upstream issue +I will add the commentsI will wait till this evening to commit in case there are any objections +When you added this did it successfully upgrade an existing database? That would be my only concern to test before applying +The code appears to only take the context path up to the first The service path is correct +show partitions after uploaded binary filesWhoops seems is loaded as data not ascii files +I too would love to see this functionality +Especially Notes had some issuesI do think Camel should reuse as much Spring stuff as possible instead of rolling our own +Hi sureshI think due to the zero byte file index file got corrupted can you please apply to correct this + +Update not closing it as of yet +testcase now works like arrayFind for none strings but still search for substrings if the value is a string +Also the could implement a reasonable default when passes a null session argument +I think a new exception related to the class loading issues is appropriate butonly after solving the first issue on which I commented above +The problem is that we are bumping up against the MB limit for unit tests and there is some sort of XSD information leak +form +Tomcat editor has Modules tab where context path can bi viewedmodified +Fortunately I am using Directory Server +This version has fixes to work on windows +Changing property name from distance to +Any updates so far on when do you think we can have at least the Boxgrinder part in a reasonable enough shape +Milton I believe Ive applied a very similar patch during the last two weeks +I can confirm this a blocker there is a viable description and workaround added to the Fuse IDE Release Notes +No complaints so I think this is fixed +I can imagine people might want to modify the snippets before adding them to the string builder +However Im not sure we need to move them to srcit +We shouldnt beadding any new features at this point thoughJust mark any issues you find like this as blocker and give extra time to review since we are in freezeIts definitely helpful to be finding and fixing this stuff before the release candidatebecause it saves us all timeThanks +This enables speculative execution to work +This issue has had no progress and is out of date +How will this be integrated with Hadoop? As a contrib module +Probably fixed now +Did the same for the Fetchers too which can cause a similar race condition and hangNo tests only fix is to makes the threads shutdown properly +But unfortunately we cant do anything about that scenario +Committed first cut of new Async APITODO needs to be migrated DONETODO Thread DSL removed +Yeah I think it should be generic +Please if you disagreeThanksRick +Pushed to master +Attaching v patches +Verified at rev +Fixed in revision +Alan promises to fix this any minute now +I can confirm that the test case attached to this issue works fine now with Apache DS RC +It could be good if someone could test it on liferay portal so we can commit the code there and close this issue too. +a patch to let xbean build with java patch from Romain applied +reviewed and pushed +This implements the described functionality +Patched custom struts Dojo widgets by Musachy and applied to SVN +Perhaps this can be improved to dump just the HTML content? Perhaps thats tricky since the formatters run outside of the outputI have no problem pulling the commit to master so you can rebase from there more easily +When do you expect to be readyIn the meantime should we move to in AS Branchx +Looks to me like it isnt in there +Is this resolved +It is really more I need that data source instead of being able to define my plink is pretty tied to last I checked +It really seems to be some strange timing issue. +And thanks Arun for looking at it +Like said started with non root user accountSo +The EWS release will include Tomcat Native +The linked issue already fixes an issue with empty titles +This patch enables a bit of code to do dynamic resolution +I have no idea where this magic should come from that you expect +Anyone interested in digging in on this test more to see why it went awry +I dont get it where does this README file goAnd how do you point to the licenseI can see the file once I unpack the compiled owlapi artifact +Added script for patch for branch +The patch has been appliedPlease verifyNB The only possible problem is that the marked as synchronized now +I put as an attempt i doubt well have the time to fix it cant i didWent on the admin portalChose the user called user clicked on Disable +screenshots +A name node ships a snapshot of all data node descriptors all blocks to the process in the begining +Ive change to check for the existence of the session +The new code is about times faster in this case. +Rejecting +Changing fix to catch looks good +Moving issue back to longer an issue in F. +hrm i already made this changeryan start a vote on the list to have it undone +We have a need to more tightly control the flushing so we use manual flush but we do NOT use which I agree is evil +This bundle is as suggestion here +Thanks Kristian +We have our own New Project wizard so if the link could kick that off that would be question in terms of how dependencies are handled +I committed this to trunk and. +Is there any chance to get this code instrumentation +That warning will disappear when the updated version of the patch is applied. +It does not honor the P you remove active then snapshot will win as expected +In that case we can keep it in the server package but I dont think it should be a pb implementationThinking beyond may be we should ditch etc completely and simply use etc in the Reader +All relevant changes to the +Please let me know if it does not apply correctly +Hard code like this is actually fun toobq +Yay! +I have used the unique index behavior since ages on Oracle and PG and now I dont know how to support it on Derby +Benjamin did not show up anything new +Change has also been made for the EWP guide +Based on all of this we know that the problem is somewhere between rows and +I did a quick verification of this bug +Note that for Managed X projects that use the current behavior makes sensePerhaps this should only be changed for Native backend projectsAlthough even for Managed X projects it would be nice to differentiate java and X classesinterfaces in some way +working test and service +I will leave that to you forCheers +I will do a second pass while going through the actual code. +Thanks Arun and Zhijie! +HiHave you built the libraries on NT? It shoulld be fairly simple? We have no plans to build on NT on a regular basis but we can certainly help you build on NT using the ANT scripts + +Moving out of +Hi Aaron and ToddThe changes probably is not intentional but it is a good change since the class name should not be in the message +The proper fix of course is tobubble up until you find the correct parent but the easiest fix at the time wasto simply use since I didnt have much time to spend on it +And I also saw that there was a bigger problem than I first suspected +Does the appear in your JMX console +Ive fixed some things since I only had time to focus on the leader not going down case for I spent a bunch more time on this case after was released +In my earlier changes I had moved setting blockStream to null to the end and later forgot to move it back +Could you confirm +also the should not be used but acts as a guide for how to configure it to use the new exsd +This issue has been fixed and released as part of release +The icuc and zlib jars will now be unzipped under the text and archive modules +Attached patch file with srcmainwebapp added to the Introduction to the Standard Directory Layout document +You need to follow s manner like etcsslcerts youre removing nowDo you have a pem format CA file? Would you please try pathto +Reviewed by Robbie +But please make sure that I have done the right thing if not please the issueThanksDeepal +I wondering about x +The amount of work is the overall design of Venus is to fetch to a cache and then apply templates specifying an output directory +I will generate a patch for this +Had missed adding comments +removed because it is included in the more recent BrunoNot sure why but there are instances of in you patch testing +Thanks I knew I was misreading it +Ive just started seeing this same problem +We will rather make an alternative which really uses the Servlet Session for such cases +Option Xfuture is added +if we dont remove them the table will grow and encounter the same issue as SCANremoving the item should need the between bookie servers who own the deleted ledgers +The issue not happen with RC and only started with RC and then final releaseThen had look on change log of RC and copied all gwc +For example sleep jobs dont just work anymore because they depend upon test code in the common test jar +Public SSH key +I think that applying the test changes to x branch isnt much effortAs far as I know there will be a release so this bug fix will also be included in the this release +So can this bug affect data integrity +It can screw dependency structure of a user project if and module are mixed up +Similar to Aris where we have s of queues doing simple qq processing and message volume isnt a huge issue +Ive already been testing my fixes on the release +remove debug attributes from configs +Bulk closing stale resolved issues +editor reseted after switching between modes +We let TT dont ask for new task when report a failed job setupcleanup task in heartbeat to fix this issue +So all issues not directly assigned to an AS release are being closed +It feels like it is something they forgot to police when adding inner type support to groovy and are just discarding the static modifierIn case Im missing something can I ask what you are expecting to be different when you make your outer class static +OK from the DB front +But dont worry! Just email any Mailing List grab us on IRC or submit a ticket to JIRA if you want to be added +Remy this is not an issue for you +Found a bug in the original prettyPrint code while testingRemoved the str repr functions as these were causing the script to behave differently as previous versions +But it strikes me as way overkill for a select list that isnt going to be that long particularly once its paired down to the search oriented ones +Fix applied to both trunk and for reporting and providing test case. +Verified on vmware +Committed to trunk. +The reason for giving up lease recovery was because it required a whole bunch of timeouts to fit in nicely or else you could have infinite race conditions and a super long recovery from region server failures +Punting to as its been implemented this way since the time where multiple networks per vm was implemented +All are being set as when the source code gets generated +Do you also have youre example +What I applied +Verified in jbt +I confirm that following the Upgrading to Roo instruction did solve this issue +Initial patch commited trunk rev +there are some online sources that recommend doing this for a +Is anyone aware of a workaround for this issue other then removal by cron? Also can these files be configured to be placed into another directoryThanksBrian +Hi I had not included the new attributes in the file +However I wonder whether were this +Thanks Knut that was fastThe results look great and Ive incorporated the masterfile update into the patchDid you have any other comments or feedback on the patchAs I looked at it again I see now that the and JDK result files are identical so I dont think this test needs a separate result file +I hope that when we talk about v we are for now speaking about v for now +Fixed in revision close of all resolved issues. +Closed upon release of. +I wonder if you could retry sometime on a newer version of IDE we inherit the version now from the parent so even if a version isnt specified we should find trying now with beta refresh and latest fine now. +I dont know if I must raise a bug on their site but this Plugin is very popular i think so also other people must have this please check back with this special Firefox Plugin +Looks like you used the wrong issue butI get it +Changes the build description to reflect my environmentTouchesM M Committed at subversion revision +the scope of this bug is that there was NO progress AT ALL as the wrong progress monitor was passed in +The patch for has been added +Triage for +You can always open a new enhancement request thenI guess we shouldnt change much about the current implementation of null handling If the Oracle i drivers do work with out current strategy a strategy looks like a quite ugly workaround for an old Oracle bug to of course you can always pass a specific SQL type to +Once this is reviewedcommitted I would like to merge this to branch too +Also can you point me to the specific branch and path in your repo to the LSMR implementation? I poked around but couldnt readily find it +This is a blocker for Qi Ming for the port +On large volumes the copy could take significant amount of time +Ill close this issue now. +Hrm +It is working for nowBest regards for the pointer +Patch applied thanks! + + logging the seed in the tests +Could we move the testcase to already existing test file instead of creating a new one +ive reproduced similar not exactly the same results on tomcat control +We took some time to solve this issue but could not produce anything more than a testcase +Thanks +I noticed for instance one example where in the old translation Network Server was inappropriately translated and this is now betterIll fix up the j test canons in the next few days if need be +Same is true of several test cases in attached +Doing this stages might smooth out the process +I have modified your fix slightly so that the createAdapter method is nevertheless called again after the creation of the adapter +Patch applied but I did not change to because the originalfile name is already tested oncePlease and close the bug afterwardsThanks for your contribution +A patch which contains the server detection for and +Closing this as I cannot reproduce the problem +According to Ubuntus documentation Bash is still the default login shell its just not the target of binsh +This patch fixes the problem +Addressed some of the comments +Very nice! It appears to be OK +Patch committed as issues associated with existing releases +Run mvn install on felixhttp Check out Sling trunk and apply the patch +This is somewhat related to +I took an initial look at this last week and was trying to see if I could increase the heap size just for the AIX boxes +this jira is about working on Juno moving that to we just do which is about making it compile as much as possible and allow to install which we do but hibernatetools JPA integration will fail because of Dali changing their api alot +Thanks Alan +Added missing line break in test project +Attached patch calls is to determine if security is on +Lets go ahead and do this on monday next week +Here are graphs for the current packages to better understand their As far as I understand the cyclic reference issue has been resolved +Still need a bit of thought how it all works with power iterations there need to be changes there as well +Benchmark results +Modified messages to display seconds instead of milliseconds +Ill look at this +Feel free to go ahead +I think the problem is select uses the wrong names +I didnt took time to merge it properly +Same thing like in a good restaurantThe question is now WHAT the question is IN WHICH ORDERSo I prefer you do both Would be cool if Petter could also check it on his repo txs! +Thanks Benoit for the patch. +And it works for me under windows XP and jdk +Ask your system administrator for access +Therefore some refactoring should applied and also some documentation should be added about this AE +Moving out to until we get a confirmation if this can be closed or needs more work +Only pages with my own composite component fail with the error +Note that there is no implementation in and the way I found to get it is to use a method from Collections +Tony patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +The application should provide a sensible. +And it looks like could use a small tweak to call instead of deciding on its own what the LIBS flags should beIt also looks like a similar usage of PREFIX may be needed in +Looks like we actually missed a use case. +This is correct it is overridden to support different server bindings +You cannot modify a deploy location once the file is added to the server +patch looks good to me +Adding a test to reproduce the issue +I had a look at the sample folder and with my inexperience with Maven it is not at all clear to me what I need to add to where +It adds some HTML elements to s schema in our constructor +Great +Yes not about the spreadsheet file +bq +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Ill see if I can fix it +Thanks Alejandro +Can you try to run a memory profiler and see whether you can spot what is holding all the memory +OK +I have added both the files +Cant reproduce anymore +AshishIlya or any one else whoever is interested feel free to pick it up if interested. +Not for commit +Sorry I think I need some help herethx I think the problem is that you have listed as a module in the pom but there is no in the subdirectoryTo try again you can delete the parent project from the group by clicking the red X out to the right of the text jUDDIv Parent and and then it +Can I modify its fixed version to add +So far it looks good +bq +Will test and commit +Its entirely possible that your device doesnt detect connectivity properly. +Testfailures are not related to the patch +It serializes and unserializes a knowledge base that contains functions +Understood re Jira +did you have a version of mevenide installed and upgraded or you installed directly the version? did you directly install into or you had it already in +in devstudio this has been reverted +The downside is that the less reliable service will take down the more reliable service with it +wadi went to WADI is at Codehaus +There are probably more improvements possible in controller failover but at this point the aim is to focus on the low hanging fruits and avoid affecting stabilityThis patch optimizes the replica state machine startup by not reading the partition information for all topics from zookeeper again optimizes the partition state machine startup by not reading the leadership information for all partitions from zookeeper again +The comment above is not meant to be +I think it should be acted on quickly either to accept it or reject it at the very least +Could you please update patch after commit of and +If Lucene ever decided to change this type of thing it would happen on the index side +This stuff looks to be working fine in latest builds. +The generator storerestore score stuff seems ugly +regarding Andys comment on foo bar not sure that will every happen so I wouldnt worry about it +If you install the bundle manually it works just that if its in a features them we get that error abovekarafroot install s mvnecjBundle ID +I dont understand why this JIRA is closed +OliverIts my code i want to specify any possible xml parsing option i need +Any docs links etc will help +I simply max out the del counts between the current SI and the pooled SR since deletes in a segment never shrink +and Browse button doesnt work if no deploy project then shouldnt be enabled +is for branch +We probably should do the collation a bit differently to make sure the words fit right +This was tested on a known failing system and prevents the stack think the fixes is already there in the code please check and confirmThank you! of +Gregor can you please test this on our finalif you reproduce please add exact steps to reproduce +But still with your approach needs to know if the block reading is for block transfer or not by checking if the client name before throwing +However is fast becoming a requirement for us +replace the patch i was incorrectly at rev +Ran derbyall with Sun +Ive altered the versions of and to be the same as in ajaxjsf +I merged this change into to be picked up for. +I never ever thought it was something to be forced on the primary implementation +Please specify how you put the service under https with config files if possible +In particular the move of from Class to Interface +HiCould you please give me more details on this +Changing the defaults in a bug fix release could break an existing systemGood pointIll go ahead and close out this ticket then +Committed revision Thanks Noble +Its complicated because tests are downloading many different artefacts so actually its test dependent +This Bug also occurs with Hibernate Final +I am fine for the patch + +And recently in our customers UAT environment we encountered issues where the stack trace were indicating the above behavior of instead of calling method A of Ruby Class X the method A of Ruby Class Y was have tested and confirmed the above behavior in all versions of and as well +Because of the same reason by where I skipped the clean up step Usage of the Deprecated Override annotations where required Ill also skip the step Remove the trailing whitespace on all lines even on the empty onesBy this ticket as othwise I would touch too manys sources making backporting of the other fixes much difficult +NicholasI am canceling the patch as the latest patch unnecessarily logs the task and diagnostic information which might affect the heartbeat +docs will be handled under I added a comment +Going to push this to M +I think thats because I was running this test by itselfI get it even if I run the test by itself +Hi Arun I have similar setting for and with and same trace is Hadoop AndHadoop reducers does not goes OOM I have seen this in onlyAnd even in Hadoop the behavior is not consistent random number of job used fail due to this +I have an idea +Sam testedit with XSTLC and it works +Do we need to do it for the specs as well +Thanks! +I found this in a different test case that seems similar enough to optimistically hope they are related +for updating fixVersion +What I did not have time to finish was should be caching home directory from the time of the creation and not use remote connection to get that information at allThoughts +No changes needed there so wont be adding changes for todo +I can help with you in this regard in the next patch review +ThanksArunI updated the patch with new directory structure. +I think now that Roshans idea is the way to go +hmm at first look isnt in the mvn here are the changes I made contains the changes I made to and the other two files are newThe Acceptor and Connector require the latest version of as and even early had a bunch of typos +In the spirit of which fixes the race condition between put and check for both and I think we should at least fix the race condition for for Append and Increment operations like we did for check +Cannot reproduce please provide steps to I meant in the development SNAPSHOT and not in will try it once more and post an update here later +fixed in trunk +While Im glad a workaround has been figured out I would like Oracle to fix the problem has anyone filed a bug against the JVM +The problem is now no longer reproducible +Please note that I request the following variances The javac warnings refer to deprecated interfaces that will require a lot of study to remove and are being used ubiquitously +Somehow I actually do believe you +Please ignore the there the new file in question is missingIve directly added now +Missed test class has been added +I havent included any new tests which are the next thing I will do +Addressed the issues raised by Uwe about the in +After that everything worked just fine for Bob +Either way this is a bug that needs to be addressed soon +actually even when trying to login with the credentials in the url it still doesnt allow you to login +moved this to common at ATMs suggestion +Having a customer make that change restart the server to get what is pretty basic info creates a lot of friction in a support case hence this jira +Closed upon release of. +The state maintained in the union operator context can be moved to the to be consistent with the model that we have today +attached is the patch which will work without changes in m pretty sure this issue applies to the latest version of Nexus as well +The contains patch file for Writer and new sets of extension classes for the enhancement listed in the is a simple protocol file for testing +I just noticed that in get there is a chance to generate thesame transaction base dir for different transaction ids +The business use case here is the ability to gather statistics against a development database and then update those statistics against a production database that has the same structure but different content +There must be an individual String for each mime typeBesides applicationmsexcel is not a valid IANA mime type whereas application +There has been a number of fixes and whatnot in +Here is a patch fixing the NPECurrently NPE can be see from logs +We fixed a bunch of bugs for the same issue a few months back +Bulk resolution of old JMX Management Console issues +The user cannot reproduce the report after downloading a fresh copy of our jar. +Attaching review commit +Jitendra as part of this bug please commit the unit test for NPE on local file system to trunk so well know if it ever recurs there +Wont apply anymore +We will also need to make sure the XTS client handlers tolerate the scenario where there is no transaction running +Patch to prefer reuse over creation of have you tried this against or? Thanks +New patch that applicable for current branch +If the patches are acceptable please add them to the upcoming release +Hi Jacques Thanks to look into itIf we cancel a picklist from Manage Picklist screen then this ECA will fire and cancel all the +Thank you SteffenTrunk rx rx after release. +This makes the emails that JIRA sends out easier to read +This is actually a bug in Sitemesh that has been fixed in their repository but they have not released a new version yet with the fix in it +Lets try to avoid it +Added two more test cases +I think it is because JCS sets the timeout for all RMI connections +The patch looks fine! Thank You! by Alexander. +Im not what Im going to do on the next pass but on the list are reuse Im seeing a fair amount of repeats in curator creation that Id like to reduce Use curator namespaces pretty much everything Im doing is in the scope of the root +Waiting on to do this for trunk +Heh it seems less will run nroff on the page +Apart from the whitespace changes that looks pretty good +Attached here insteadI did not attempt this in Mono + +I agree the current case is wrong but Id much rather see a special case where if leftright is a then the other side is coerced into a and calledWDYT +The way to solve this is to pass more memory to the javac process itself using the mx option to scomp like soscomp mx m +v. +Its currently possible to inject extensions at runtime we cant remove that behavior without breaking backwards compat +Have you checked the logs? I remember seeing something like this once but it was so long ago +Dont you have that? You must be logged in to see it +If youre willing to undertake that work please file a separate JIRA and provide details there +Please provide some further information to the Jira as its currently a bit vague +This runs fine on my Win XP system with the Sun can be tested as a standalone test by executingbuild Dtestorgjbosstestclustertest at the attached log the exception connecting to host messages are an expected result of the router stopped line +Merged thanks! +Looks like Eclipse wasnt failing on the asserts +Closing resolved issues for already released versions +it was not ready at that time now it ready + +Any suggestion? Regards for the follow resolved issues for already released versions +By default the recovery manager thread kicks in every seconds +Little cleanup made that ZK starts first and ends last +I did another review here I think its ready to land on trunk! Nice work Simon +Remove it +Any further thoughts on this +Still no big following in Tycho team on this issuepatch +Marking resolved as we did everything we agreed upon. +It seems that in distribution rake gem is not installed at all but it was in +Paul proposed the same on IRC +Related like the commit comment for mistakenly referenced this Jira instead +Also I recall that you seems preferring communication on JIRA or public mailing list +It applies thanks +For the moment I dont see any more significant leaks in the Aries stack +The problem appears to be in which depends on +Done. +Are you sure that the fragments are being attached in the case where you see failuresCertainly a reproducible setup would be good. +Ill remember next time +I will add a changes entry and commit this stuff +In your example you only ever define one method that is visible from a static context so the behavior calling from a static context doesnt change +Should fix the +I think he must face the same problem I didSome of my servers are shared with other service +Well lock into this but I would not expect that to be part of d like to see this fixed also +Yes this covers all the cases +i will try to analyze the rest of my with this versionthank you very much for your help. +I hope to fix this soonish though it should work to use the above snips to create the classpath directly when loading the custom task. +Since this ran in a VM as opposed to real hardware on my previous reports I this on a Lion VM with RB and got the hangUnless theres reports to the contrary this would narrow the problem to RBLion +There is one thing we should discuss about the messages on console I did +Committed thanks Mark! +The issue I was having was different +I believe this has the same root cause as described in but this section of the tool doesnt have the special case code +pushed +Right ok just wanted to make sure we are om the same as part of as part of +So no need to remove the workaround in the in JBDS betaL +You are right +Zhijie +And Models +It looks like at least in my case the mock driver is exposing our poor handling of errors when trying to start an instanceIf the credentials passed to deltacloud are incorrect instances get stuck in the Pending state +When that gets fixed this should fall in line with it + +dims +We didnt make use of that because we wanted to break up the old into smaller more focused libs and ended never doing it +Fixed. +I might have given you a hard time although it was not my intention +I verified fix +And shiro does not force a configuration today +DefaultThree one dimensional points with value of eachWeight +This just needs to be committed no +The assertions fail when running against or trunk but pass when running against orThe pycassa test suite covers these areas a bit more thoroughly which may be useful to check a patch against +id accountid domainid poolid NULL lastpoolid NULL instanceid NULL deviceid NULL name uuid size folder NULL path NULL podid NULL datacenterid iscsiname NULL hostip NULL volumetype DATADISK pooltype NULL diskofferingid templateid NULLfirstsnapshotbackupuuid NULL recreatable created attached NULL updated removed NULL state chaininfo NULL updatecount disktype NULLState is in only even after upload operation is completed +Patch available +Id tend to agree with Joern that the changes should be done after adding the code to the Apache repository +So why did you reopen the bug? +Instead it is stored in + +Thanks Jakob +Can you double check +This is not a bug this is something that should be added into Release Notes or Installation could you add it into doc if its not added to getting started guide installation section and release notes for this to set release note note text has been added and marked correctly +revising this broke the build +Thanks +Weston volunteered for that +This patch fixes the bug +bulk close of resolved issues. +Please verify that this corrects the issueThanks +Attaching new patch incorporating Tucus last comments +Im tempted to mark this as Wont Fix because this would be incredibly difficult to fix especially since saveState and restoreState dont read the state correctly +In which case issues such like this makes Joe Average complain about how cant deploy basic apps which other AS can +Tests passed cleanly for me on the previous rev +to Eddie for review in SVN by adding a null check around the lookup of the usages of xmlbeans found for anything related to the ejb control +Thanks +Hi SteveThe String that the method takes represents the String representation of an Object in the +Possibly its best to avoid the whole nth party concept and find other languageI think directory approaches are much more manageable +During normal operation follower vote by writing to txnlog +Given the state of this test might not make it +All of this works now +Awesome! Glad you got this figured out + +So this one is a backport +Fixed on and closing all issues that have been in resolved state for more than a month without further comments +With two name nodes +Closing the issue as a duplicate of +Comments +I like your changes Todd +If files are identified with their md checksums we no longer need to verify file modification time in the TT +This can probably be closed as long as EAP is packaged into +Verified on +It checks for the status of the AJAX response which seems to be +For a short term fix its probably better to roll back some of the code changes for +But something that bothers me with the patch is that it creates a hard dependency on the annotation class so the jar will need to be on the classpath always whether or not the annotation is being used +And under RHEL I used all the time +binmahout script +sry forgot about this +The code as is does not build +applied thanks teody! +A plain old test case is fine. +Digging out the sports equipment now +HiMy name is Malith Dhanushka and I am a final year undergraduate student from University of Moratuwa Sri Lanka +The changes to and introduced a race condition +Verbose +You might want to look into while you are working on this one +Does this make it unsuitableLooking at integrating jirb now +stack You mean theres something wrong with this patch? So what should we do if we want to enable this option? Use the code from trunk? IMO a correct patch against will be very useful until releases as this feature is so critical for online service. +Committed to trunk with revision . +Colin could you please have a look at this? I wont have the chance to address this documentation gap before Ive added an actual chapter just on the topic of Resources +Multiline strings are handled. +I just committed this +Are you sure you updated +Upgrading versions by substituting individual jars will not work +So the solution would be to have a key the resource manager instance the class path and the data path. +So this seems to indicate that locale of the database will determine the language of the error messages +It just checks if the operand is a column reference and if not skips the pushdown like other tests in surrounding code +Added +Please on EAP CP CR +In many cases you have no control over the object model being stored in the cache either because its not your source code or its a generated object model +Here the patch to the JIRA This has additional support for holders as listed in JIRA +They are not really beans as they do not have a default constructor +definitely possible would also be nice get code completion for the issue as duplicate of the newly created one in issuezilla +Patches are committed. +A testcase to demonstrate this +is that right or should one of them have been mintf +Hi Sanjay can you please set the target versions field appropriately? For what release do you consider this a blocker +Since Nexus is neither worse nor better than the previous process there is no appetite to revert the change now it has been made +Committed to trunk and +Im not familiar with this code path myself but Im pretty sure some other folks will be able to helpThanksve added a test to +This array is sorted by the term value +Cannot close since FF its fine +This is still an issue in ER we are not utilizing a sizeable of the screen and requiring users to scroll horizontally in the part of the screen that we are using +What feature request are you referring to? A view of the form SELECT FROM tbl is a perfectly valid target for materialization +bq +Nice thanks Chris and Aaron for the review and commit +Ok Thats clear +Shall we punt this to +It always responds then throws the relevent exceptionLater in +Updated the description with the URL to the Ant manual that was cited +Since my questions werent addressed let me ask them again more directly Does anybody believe that the code in question is qualifies as open source? Does anybody believe that apache maven repositories should contain non open source codeIf the answer to both questions are no then asking Oracle to make this code available under an open source license may be in order +Editorial pass over all release notes prior to publication of +Definitely need to know how youre doing this +Hi Matthias +Although it doesnt appear to be used currently it might someday +Its a darned if to you darned if you dont +Attachment has been added with description a resulting pdf dont just attach the resulting PDF +Anis and I looked into Android yesterday +I think this is the same issue as which I just fixed +Fixed to also take into account inherited groups. +I could not reproduce it +The mispelling is just an annoyance and should be cleared up +Can you please make this change +At this time it does not support it for the primarily because it isnt clear how to do the rollover safely +Hi Sergey The has been modified and your patch cannot be applied now +committed thanks +downgrading the bug. +Looking at what representation of the unique id to hash on by hashing based off of the string value rather than the indexed value solrj and any other smart client can hash it and submit requests directly to the proper shard +For sure input validation can be improved and more explicit runtime exceptions can be thrown +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +This is a consequence of the changes which were required to minimise the impact of issues discoverd by Kurt on Thursday +I mean that we hope to provide more complicated expression support in a future version which will supercede this issue +Another solution I am thinking is have the and keep the information about what privileges they require +It is a leftover from the version where I used strchr + +Hope derby can provide an improvement for stream operation +Chris yes I think at a glance this is where I got stuckRelated to this there is duplication in resource loading code already that would be nice to clean upFor example Lucene and Solr have their own separate code etc +checked in r. +Merged the second into the master branch. +The are subclassed and store their values as explicit fields which makes using their values much easier +and +Ah right +He may have responded since the end of last week but I have no clue as my VPN has been AWOL +The difference Im running of the within GAE SDK +This scenario is covered by the Stargate test suite which passes in the release and on trunk +Not an issue in AS +stackyes I will commit some fixes soonI dont think all of then are real leaks I believe there are some pool effects +seems easy and supported in rLeave it open until I have review documentation +Fixed again temporarily +I think the issue deserves some time in the faq on the website or in documentation in the code since it isnt trivial and the justification for the decisions may be forgottenChris +Jakob you can either do them in this Jira or file a followup jira for same +this is a maven issue see a workaround you can specify a full path to a parent pom like +add from the EAR project to the classpath of the EJB project This is necessary because of a bug in the WTPs EAR classpath container +Commenting so I get notified of updates +all issues to the CR +closure arguments now start at +Also the copy on is up to date. +Right now it sends out a PROPERTYREMOVED and a PROPERTYADDED event +Ive added your patch. +Ill start in this direction +We also have for generic type converters that is tried when the regular converters fail +Fixed createOutput and listAll as Yonik suggested +And the script component could be automatically added using the renderer decorator approach on the form renderer +the header is indented but the file is notThis was using version in r feel free to testcheck +Lets not close things for the sake of cleaning up JIRA +And the number of block replication is also set to be +reopening to target for theres more work to be done for this issue please or open a sub task for the remaining work +It might be possible to copy some of their implementation from MRI or Rubinius +just uploaded a patch with an class for Iterator support +Scheduled for. +Hello I checked the latest codebase if the error was there according to my code it should not happen anymore in the latest codebase so give the svn trunk a chance +Ok is running now too so both demos up both with https SSL connections working +Overall despite all the confusion regarding the app token leaking I made the minimal change to satisfy this jira by removing the appTokens fileWorking on this issue is fine +What exactly is incorrect about it? +In TimeoutError is not a but a Interrupt which is a subclass of +Attached is a brief document that contains a diagram of the new class structure for this refactorThanks +Fixed in r +We have strong indications that nothing important is missing. +Thanks for your supportThere seems to be a minor problem with the server configurationWe require the index pages to show folders first but they are currently intermixed with the filesPlease could you your server settings +Im closing this +Thanks for this +Note usersdesignauth does exist and seem to be valid when I evaluate it +bulk close of all resolved issues as part of closing items. +Yes I was able to reproduce the bug I tried also other size near but just exploits the bugI moved my attention away and wanted to return back after some time +Manually tested! +I was about to add release notes to this JIRACan we come up with a better name for the new qualifier in +Early patch that at least builds +I have no idea why this is the case +Sorry +Unsquashed will happen as soon as works for me +Committed to branch +Its notA nice to have would be an unauthorize error message but I doubt thats easily doable +This is a very shy bug doesnt express itself very often +Duplicates +If we build the main project the sub project is built first and during the build of the sub project is when a truncated pom file is installed into the local repository +Sorry its my mistake if we select any table and press Select all Children it works. +In other words users that are creating a URL with ant patterns with the namespace configuration will still get a case comparison +Thanks HyunsikI agree with you +and dus +thanks for this one! this has been a major pain for me. +Krishnadasan T Syou can use some tools to analyze why your broker hangs suck as GCjvm stackthread +Case c would be more challenging but I do have a very good case for it at the moment so maybe Im going to have a look at it this weekend. +is this still a problem +Attached test case as file +Now I understand what the problem is The request is retried instead of failingafter the first time +I spoke to Kathey about how to pass null +Adding a test module in a future version will not change that +Unfortunately the particular html code that builds the page from those toolbar links is within the codebase. +The DMG likely shouldnt have and definitely doesnt need the classifier so that needs to be thought through and handled somehow too +The updated patch for sitemap and DRLVM index improved and merged with last changes +Merged to master +The only cause I can think of is that the compression ratio is too sensitive to the data to or evenlzo is under GPL +Sorry for some reason no one picked up and applied you patch +Thanks it seems that the late hour has shaded my sightIve implemented the stuff by just following the maybe this issue should be changed to make the better +New Jersey version is +Im not getting the problem +Is the xew plugin available in a maven repo someplace? Could you post the stack trace? Maybe even a simple hello world style test caseMore of a jaxbxjc issue for which we have no control and no response for a couple weeks. +DoneBuilding dependency tree Reading state information +latest the interface has changed in this new patch should we revert the first patch for +Attached are the patch files of all components that have been let me know if they are OK. +agreedthis will come for free with its scheduled for the release hope to get that feature in soon +Added the two sticker issue for me please close. +why we need a full parser? what exactly feature were talking about? has some features likes Temporary tablescall level interfaceScrolling cursors Maybe we should discuss individual feature instead of discussing whether we need a new parser or not +Approved for. +Im sorry I didnt have time to complete the project +In previous companies Ive always had a separate test configuration in the srctestresources tree which is used for unit tests and usually points to a different database whereas the productionUAT schemas are accessed using the srcmainresources tree +It seems it was not working for the OP who opened the ticket +Yup we can look at it +Test case modification is necessary to one of the behaviors described above +Corrected spelling error explicitly stated the fallback mechanism moved the description to with a reference in the to eliminate multiple maintenance points +How about closing one of the as duplicated +I am pretty sure if we half the number of threads and double the max q length we get exactly same benifit +I have committed this to and. +Bulk close resolved issues +I am not expressly +The latest woodstox release is the +sh +This issue needs to be patched +The utilities only need it for portability with the libc locales on the system so when theres no libiconv we dont need to worry about it +The discussion up there strongly suggested compile although personally I was also leaning towards provided +I thought that the best idea was to have a label for each localeI tried to test it for the label you indicated but I couldnt + +Please submit a patch that works for you +No such error found maybe this was fixed by Hugo and my latest fix Pl +It is not a bug and I think improving this might introduce all kinds or unwanted behaviour +If the offset is then ajp would be which is does not identify however if I set the ajp port in the to with no offset it works +Thank you for your careful attention in what must be the wee hours of the morning for you +it should be enabled already +Ill hopefully have a PR soon. +Thanks Pontus for reporting this this issue has been icking me for days without any further reactionWill take a stab on it and see whats required +Also add equals and hashCode for +VERIFIED +New patch updated to based on change in the patch for +Hello you are welcome to submit a proposal if you are interested in this project. +This would also not make finishStage be so dependent on what is happening in the engines when they create their cluster responseIm still trying to wrap my head around so see how I can provide test methodsIf option that I laid out is preferred I should be able to provide a patch for that as well +Its not specific to only task processes +I think this limitation does not exist anymore +bq +put on branch in bbf +JonathanMany thanks for contributing the patch +Since someone changed all the method signatures the old patches wont apply any more so Ive updated and combined the two patches +yes quiet funny considering atlassian makes a wiki +Yup fixed thanks! +Data nodes are throttled to prevent rebalancing using too much network bandwidth +Looks like I already fixed this for +That spec chapter describes the classloading strategy Tomcat used to use when it was considered the RI +bq +just attached properties file +Latest version of archetype and a sample of generated application +Marking closed. +I would like accessBe sure to leave your Google address when requesting access +I need an original file that follow to Without sample file we cant reproduce and thus not fix +patch seems reasonable. +Those optimizations have been written after some profiling sessions on Karaf default distribution +Thank you its done +tested! +This is also true for Dependent scoped beansThe respective paragraph from states Unproxyable bean typesA bean type must be proxyable if an injection point resolves to a bean that requires a client proxy or that has an associated decorator or that has a bound interceptor +Hi Weinanif this doesnt require a release note Id like to close this +I guess its just a coding standard Ive got into and it certainy doesnt harm +Why not be consistent here too +a lot for the test case +But I understand what you mean with the main method implementationIll add it back and provide a new patchI also tried the sample with the new classes and it all seems to work +One more thing that need to be addressed is that I like to attach pictures to each one of my entries and not necessarly for the whole blogI am making this my first extension for the ROLLER +Patch applied and tested. +That should mean that something has been fixed between and +I noticed this as well +Example for Willem you set the version to +couple of comments We need to add a section on describing the and its motivations etc +deleted the newLabel method all together so that it defaults to the +Test and verify sending Longs as floats in Ganglia actually works +The patch looks ok to me +Well keep an eye on weird behaviors and let you know if theres any +I ran into this with a project that had dependencies with a type of neither jar nor +Users can now set whenRequired attribute on the installuninstall to configure when this install is executed at which state +No IT +Thanks Christoph HartmannI have upgraded the jetty +Closing. +On Linux with Karaf updown arrow keys work fine I can navigate in the history leftright arrow keys work fine to edit the current command backspacedelete keys work fine tooThe only HomeEnd keys dont work as expected +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau + WilliamIts a small patch I currently use +is the latest version now +Oh right +From your latest patch shims works this way +Hi AndrzejAre there instructions on how to add this plugin to NutchPlease let me knowthanksilango +Right one of the issues that comes with trying to keep workspace and namespaces in sync without having a transactional api for the catalog +Based on comments closing. +Updated documentation and tests. +I agree it should be a higher which workaround did you found +work. +Theseelements are the only elements in my serialized state that have the namespaceattribute +Henryks patch improved the performanceHowever not using any regexp is even faster so I added a simple replaceAll method to in that is based on a logic +Thanks AkashYour patch is in trunk at r at r +If I figure it out today Ill commit and otherwise push it off +Are you sure is running? I am not having any problem with any of my project with or without the xrefLocation +Having talked with Arun and Aaron about this a bit we came up with a few candidate solutions several of which are described above +Closing. +Well the entire Apache Solr is still green with my fix not that it makes any guarantee that the simplification does not matter +My priority at this point is code generation if you want you can work on other missing functionalities like data file RPC etc +Is this committed in TRUNK or +Instead of throwing the sending the decommisioned signal which will shutdown the NodemanagerAdditionally one more problem has been addressed in this patch when some exception occurs in communication with resourcemanagerthe thread in the will get exited and not started again in this case nodemanager will not send the heartbeat +So the segregation is actually intended for optimizing of tasks by not downloading cache files that are not needed by themFor +Heres an updated patch +With afternoon commits I get the message right away +Data files that trigger this problem +Hi Dan +Likely still needs some workAs for the Lucene error I thought I had updated the Lucene version to be which I believe makes this all right +If it has we throw an exception and will close the key We do not write the fetch response for all the topics to the channel in a blocking manner +Improved patch +for previous version not for commit +When I unzip the zip I am not seeing the same directory structure that I see when I unzip EAP zip +Retitling this to cover the entire underlying problem you cannot interrupt a Client if the far end isnt responding +Its a lot of edits but they have a familiar rhythm to them +So I think this jira can be closed +Id actually be happier with a separate test project +I think with my changes this should now be fixed in CVS +this time it can pass all you regenerate the patch I am getting some +This has been fixed by. +I could not reproduce this issue on using both themesimple and xhtml setted on the form +This isnt hard to do +Thats some good news +fixed in as issues are fixed in +please update to version and test have the following in mindYou can define or a can not be converted to a and visa versa +Or I guess on every request build in a timed retryMy main concern at the moment is coming up with a plan for these timeouts though +The same patch is applicable to +A big for this +I will commit it to trunk +I also can confirm that as Matthew said setting false in the files does make the problem am running the Rails app from my Windows machine using Jruby +And then talk about a solution for Camel where we develop a dedicated polling consumer for fileftp that isnt scheduled based +Fixed in revision by me +Why not move the check under the sync too +Spoke too soon +It will take automatically new routes during process execution and will determine what kind of invokes to do blank execution or synchronization with other processes +v attached +And there is a typo it should be Tomcat Status and not Stauts +The tablet server only has one node to read from zookeeper so there is no consistency issue +Moved dependency of pom to scope providedAdded dependency to and to compensate +we are writing to the WAL after the increment this may be the only way to do it effectively but we should also think about this carefully +Wayne can you make sure I the right one +I didnt commit it yet because I had a format problem and to perform some additional you reopen the task so I can refer it in pull request + +Committed revision +If I were a DBA bent on increasing my salary I dont think I would use importexport to do this +So we have a fork in the Id be inclined to say too bad dont use arrays of symbols but I certainly dont want to break your code that depends on this if possible + +This has a lot of consequencesI mean we should not wait for years like with No heavy committing please g +Hi Stephandoes this patch break any backward compatibilityGreets I havnt noticed anything +Neither do the similar codes EPSG or EPSG AmosI have had to patch these projection is for my own use it isnt really a bug in Geoserver it is simply that they havent updated to the most revccent version of the EPSG database +Will complete this soon. +Thanks a lot for commoting my test +It is +Not sure what an accessibility screen reader will do in that case +Perhaps a site backup script is running or rsync or something placing siblings alongside the + thanks for the review Stack! +Setting Fix Version to future for issues without a set fix version +I wonder why this issue is open since +Fixed in trunk. +Removed dead code i forgot +LGTM dont like wiring this stuff for all testsmost tests dont dont think my comment was ambiguousmost tests dont care about this close for +Thanks for reporting but this not a problem +Thanks a lot Tom +Theyve asked for +Not really possible I fear +The test case along with my and persistent entities person and Address +you need to provide some more info here like the route descriptions on and In any event that does a dequeue and enqueue needs to use a dequeue transaction that spans the enqueue such that there is no inflight message lossAlso in the event of B failure the failover transport on the jms connection will do an auto reconnectHow are the activemq components configured to access brokers are the brokers persistentAgain attaching a test case or some variant of your routes would help + +I had posted my nd comment about my misunderstanding of the code I put in my first comment which is the same thing that you have told is being fixed is the issue reported in original JIRA +patch available would you please try it +Committed to trunk. +Hi Werner okay +Hi DanielWhat platform are you on? Can you post a screen shot ideally one that includes a comparison between Pivot and native textThanksGreg +I believe this was because the old version of hbase was bring in hadoop +Testing this feature requires a shared library which has to be copied to +The same version? Or should I increment the version to get it processed +Adding a feature to the management agent to uninstall the currently installed deployment package should be enough to achieve that +Thanks Harish! +So just go ahead and do it Thanks +Great job! +Provide the unit test which reports the bug +We will just continue to use SVN for the website. +Use cached keys whenever theyre asked for +Marking closed. +A couple of notes Renamed the to +Clearly bugs regardless +I have followed the profiler closely and I believe that the correct behaviour would be if I set the session timeout to minutes after minutes the number of instances should be around the same value as long as the load on the web page is about the same +is scheduled for this week and I think researching this will take a bit more time than or days +Patch attachedNote that this only affects nodetool cfstats as well as calling the affected method directly no internal code calls this method +Thus my previous comments +committed to both trunk and committed to the trank and branch +Sure there is always the way with Acegi but in a simple command line client +Removing bad version +I agree +I would recommend using static importsWhat we need is to come with a list of the imports that the user needs to add on his bundle to be able to make the discovery process work easilyI may be wrong but for annotated converters i think a static import of package should be sufficientI agree the discovery process in could be a bit smarter and look for some camel specific MANIFEST entry but this would not change the fact that the bundle has to import the needed packages +This is just to prevent the unit test from exiting before the async reply is I wonder when you are going to use this for real use casesHaving async behavior flying around at will makes it hard to track down what happens +bq +For would remove support for +Id happy to close this issue but sylvek saw this issue too though I cant confirm it I do hope that may be Ill be able to close this issue soon enough +Defer all open issues to +So here I guess this has to be enable by th program +this begs the question can a zk cluster be shared across multiple hbase clusters +Ive added the three database reconnection parameters to each of the andIve added this issue has been confirmed as correct the Clone ticket for EAP can also be completed +all tests still passI hope this helps to keep the PQP +fixed in rev shipped in +Committed to sqoop +If you really want to start a network server from the command line inside the test you could have a look at derbynet +Cleaned up patch still without test cases +Ive commit the change the the test passes +Changing O to proudly called a patch +No objections to closing +One thing for m devs the lag time involved in this process may be too much +Even accidently loading them two times because of is too expensive +I needed to add a quick hack to add a when the tokenstream doesnt have one so that doesnt complain that it cant find this attribute when indexingBut this is a bug +I guess youre talking about extra coordination to make sure a new node doesnt start serving reads until it gets the data from all its bootstrap buddies +Hi Jaikiranwhich browser do you useI tried this issue in Firefox and Chromium on Ubuntu and I was not able to reproduce itThanks +Looks like the axis handlerchain support is incomplete and Ive marked handlerchain as incomplete Axis function in the web services support table +Ok! I think that will be automatically solved by my solution to point +I pulled this into also. + +Fixed in r +If it does then great youre done +Sorry for waisting your time this was obviously never the intention of +Users do not set priorities +We have retired Woden code +GordonAre you suggesting that options like etc +This has come up in the past and Ive gone back and forth on it +I am trying to get it tested on secure cluster will update on how that goes +Maybe there is a method that does both these things which we just need to call +The idea is to show a client which eventually gets shipped in an application its not a test case and should not be complicated by test considerations a Junit test case showing how app developers can use a similar component pattern to develop test cases. +Verified on. +Are you building with a patched pom file? The official source tree compiles against the Maven API and the Maven runtime has a compatibility layer to handle this +A simple zip is handled by WTPs utility classes but a recursive zip with nested modules is part of the archives integration +Attached is and that you have to set true and true + have performed another test using the same repetitive actions but with the following memory sizes Xmsm managed to do cycles +Everything turned out to work fine no strange behaviourI tried to put some trace code in and I even got by simply dereferencing the this pointer so when the happens a method is called on an instance either never allocated or already deallocatedMaybe its ibm jvm that breaks the codeCarlo +A clients lease goes stale if it fails to renew the lease for minute + for the patch +I cleaned up the test +It wasnt relevant for my needs since estimates are manually entered for the most part +Attachment has been added with description test has been added with description test case andDid you miss my comment about ICLA? Luis really needs to submit an ICLA because he has submitted a few patches now +eclipsekeyring arguments os macosx ws carbon arch x keyring Usersthebugslayer +Ive moved the commit toIn IRC Martijn asked whether this commit should be part of and I expressed my opinion that I prefer to not include it because it was checked in in the very last minuteIve also added that we can always release few days later if this breaks something major +Yes it could be improved but Bugzilla is the place to fix it +Attached is the version of the patch file that I was able to applyWith this applied SDB builds so this doesnt break anything however I dont have a SAP install of any sort to test the contribution againstI have committed the patch to SVN so please could you check out the latest trunk verify that the patch looks to have been applied successfully and test against your SAP install +No activity since October +I have changed the title of this issue linked and left a comment there +bq +bulk move of all unresolved issues from to +Please see attached. +The fix can be verified by new unittest class + +Thanks Kavita +Dup of +for no need of wait in after the queue is filled up +The other important point is about how the health checker stops +Right click on a groovy or java file in a groovy project +Thanks Ted +Initially I had problem pin point the data because I forgot my time zone isnt UTC and the time selection widget needs to be offset by my time zoneThere will be a follow up patch to include the backend analysis part in another JIRAMy fault for not closing this bug +The preferred fix is to get it to be more precise since the LS liveness calculation is actually correct for this code +Do you mean that I cant have jsut a return on simple methods or am I using half sentences all over the placeReview from the backend guys would be appreciated +Test case +Attached is the patch for this bug +Its a bit more explicit but wont hurt anybody I think +For the time being I am just going to this bug to and add a check for hostname value being blank to +This enhancement will be considered for a release after Struts +Hi Christoswere sorry that you have to rebuild with a new modified file every time you upgrade +see previous ll check it again this evening +Yes Im seeing this exception with the today pm svn after a clean and build I ran the following commands on my terminal svn up +Are you still planning on first backporting the fix made to into the trunk line? Have you done any more work to determine if the out of memory is this change or something in your environment? Given the issues we have had with triggers it may make more sense to go with the safe change for and then later after is cut you can check in your new fix to trunk and get testing there before putting it into a released codelineMy interpretation of reviewing the change is that it should be safe and not cause unexpected out of memory +Hey RadouaneIll have a try of this at the weekend and hopefully have a solution for you + to trunk. +Most of a fix is in geronimo rev for trunk +On the suspicious deprecation showing as an ERROR all compiler errors and warnings show as errors +Eclipse classpath is caused by not by this patch +attached +Great job! +For traceability it is good to know versions of the compile dependencies for Spring +NET +Can you please help removing the jbossweb examples tjhat will not work on the EWP? I guess the whole jms dir can go what ellse +I dont get this one +Just another short comment as I have amended the JIRA notification scheme in use for Castor +Thanks Sumit +Hi RamiThe problems with the BOOLEAN datatype center on concerns +Looks like now that the line moved into an if block the corresponding should move in there as wellThe description in this JIRA says it should try and reuse the hbase delegation token in if the kerberos token is not available +Now the documentation reads The Executor is a web listener and needs to be started by a servlet in a WAR file +This is an outline version that can be used in environments +Sorry but theres nothing can really do for you +So I am going to reverse validatornormalizer order in your patch and commit it soon +This is ready for review +bq +This would require to do some more work to ensure a property is in a reply +Ive committed it and fixed it as resolved. +Is that by design +Please find new version +And that should be fixed too nowLets hope the results will look better +on commit +Merged in master +This version is not able to read the partition size but only the free space +sorry all +It throws in case of failures +Help in resolution is appreciated +Thanks Shai this looks great +sum looks unused in addWoops Ill remove +The patch throws an error if the skip was short it might have been better to loop until the required number of bytes had been read or EOF was reachedBut if you are sure that the input will never arrive in chunks then that is not necessary. +I am using jruby and and I am seeing this issue when I run dbmigrate rake task on an oracle g database where I am trying to load data into legacy tables in a rails app +Progress is persisted as bytes are read from the underlying file +When is going to use that version + attached is a fix to the bug +Committed this to trunk and +OK assuming the section link works and what is described there matches what we have I agree it is easier +Current disposition for this JIRA and and Mark these as resolved reopen if made necessary unlikely by jBPM rework leave these open as they are jBPM issues New JIRA to require to document in Release Notes that when jBPM is used then it is not safe to use on Linux The jBPM does not work with Linux combo mark as fix in Project bug for +Confirmed configuration text updated in HQ drop +unassigning from doesnt look like anyones working m looking at this but atm +There was an error inIts now fixedThanks for reporting this. +Have you startedfinished work on the Extractor factory patterns? I would like to get this issue integrated into trunk +We have the same problem with the menus in tobago so they dont work in operaMaybe someone else with more opera experience can take a look at thisregardsthe problem is that you are fixing this is not written by us but by the prototype guys +I can use there is access restriction in the JDK on this class +test suites passed +For the apache incubator position it depends on your point of view +Thanks +Fixes the bug as well as adding rack information +Thank you very much for the commit +Why even expose the class in that case? It cant do anything more than a can so as a client of the library I dont need to know about that class at all if I can access all its functionality through its base class +This needs to be fixed as many of our current taglib docs are broken because the import snippets from the docs +Bikas my mistake missed the change +Is there any reason why youre unable to upgrade to a more recent version +There is another error in the that was highlighted because of this fix is not issues solved in released versions +Fixed +Adjust paths of the JAR files so the debian package will build +The binpig script points to when we dont supply it anymore +Opening a separate issue in the SCM plugin for that +Doug theres not enough information here for me to duplicate the error +I think we should migrate to java for though +Seems to build perfectly with that configuration. +This file is notated with instructions for running the TCK +Committed revision Geoserver Committed revision +Updated patch fixes a problem where the primary keys of the joined tables have different types +Setting the priority low because the use of a document function within a document function with paths isnt too likely and can be worked around by putting the files in one directory +Since we do not have dtd for xml files other than this makes sense +One commentIn in handle its better to log only filtered events +Sorry about this +Looks good +I am going to define additional constants JOBCLEANUP JOBSETUP TASKCLEANUP +will revert back soon +Ive also disabled on our JIRA so I dont think there are a ton of plugins left. +The corporate security policies at work are so fantastically paranoid that I continue to use Maven on the job instead of upgrading to a more recent version +Sorry Brian I tried to create Simple Web Service against Java Project and NPE is Sample and Sample Web Service I can see NPE only in log which is i guess ok +This means a lot more is getting executed as root than just the java process +I also need to think about how to test this +The previous fix was under the assumption that we are packaging the sample app generated by the create scripts +Kudos +I just noticed a of setting validate to true is that it also bypasses property annotations made to methods in a base class +Having said that there might be a where +Thanks Sujay! +Thilo PaulWith two independent reports on this issue it certainly sounds like a genuine problem +Attaching +Good lord +I am unable to recall the context on this right now +This change seems like breaking Mac i do ant binary with this patch i am running into this issueBUILD FAILEDhdfs Usersbharathmworkprojectshdfsbuildclib does not exist +Ive opened an issue with Infra to cover most of the remaining itemsWriting a remains +The tests in depend on the filters package having been updated with the patches in and +I will attempt it again today +Is this a reasonable approach +For order manager also a product store for sales order entry but an internal organization for reporting purposes and for purchasing +Thanks a lotBest regardsRichard +Some handlers do not store the configuration +The replication factor for the sample file in HDFS is not changed with this patch +I committed this to trunk +I understand rc is due soon +Should be fixed in the latest RC. +Not the actual filter yet +Since there isnt much that you can do about this that is just pain for no gain +Thanks Dave Im a big fan of Unit Testsone issue +Im not familiar with the new JEE Model can someone explain to me why the return type of a method needs to be serializable in order for the whole object to be serialized? +dIon since you are releasing them can I leave you in charge of thatThanks. +Problems renewing dfs leases are now logged +Here is the update +Didnt make it into EAP or +You can look at the to figure out whats the requestresponse from the faultTo Endpoint +I was on the call as well +Updated patch generated on Sept +I think you are right the error message should be more clear +But I ran all the tests before commiting so what was wrong +Can you try one of them with your data and see the result +Assigning to Rob for review please? Thanks +moving website +Make memory usage more intuitive now that there is a read and write buffer for each cleaner thread +Stage and on as it uses the new base class. +A patch suggesting a solution is attached +Ive asked Fernando about which way to go +I have the same problem using Sun JSDK! Of course there are! Even times +On my side active rules from inherited profiles are correctly exported +Thanks everyone +In the Add Library wizard dialog that appears select +This is a slightly modified standard example file that comes with Netscape DS +I have attached the output fileIll commit this Monday +After that GEP gives me the very first panel to select a local repository without any promptIts kinda confusing +Try to reproduce it for file +We need a feature for this new component so its easy to install in do you mind get started on wiki documentation for this new component +Filed a ticket to review this and other code paths forGeorge +This patch addresses the issueIt reads multiple hostnames or addresses separated by commas and resolves all addresses for each hostname +Revision +meta data is very small compared to large content data +This is getting closer but there are still several fails initest +Yes a limit on the number of concurrently running tasks cluster wide specified by and enforced against given job would be greatBut also having the option to set the number of concurrently running tasks node wide for a give job is also important +Maybe change to private on commit? Should be public? on commit +So Im triaging this to. +Closing based on comments. +plugins dependency to Beans Support Feature Integration Mylyn Integration all +I think its ok now +Hi to your advice I open the debug flag while compiling the project it works and will not throw such runtime exception great thank youAnd I suppose to know weather it means I must compile my project with debug flag from now on +As I understand the project only intends to use GIT so it will be great to have a direct setup without having the need for svn and then mirroring it to git +I am going to clean them with the other changes I did to task names etc +You do know that doesnt deserialize private field data right? It only works with getters and setters +Ahh good point +Fixed in revision +There is one for byon as well +Was not been able to reproduce +Shutdown of the base classes is properly performed just the do nothing on thanks for the quick fix! +The point is the encoding of Bad t encoding t testcase fails the verifyUTF check +Committed revision . +Revision Expanding a row now triggers extending the resizer line down to the new bottom +submitting new for the resubmission trying to trigger a new Hudson build +yep youre rightbut the source features are created automatically in the build process +I appear to be incorrect so Im reopening the issue I tried deploying against a enabled share directly served by Apache and have a similar problem +Its late now so I will leave it for another night to look further into +And as well AFAIRChristian good work we will get that into +This is +Closed. +Thanks Wei +Would you mind doing a quick review and test in your environment to ensure the patch took care of all issues? +Alessandrowhats the plan wrt this issuemeanwhile would u please also summary this issue and provide a release note for it? see +Single model defined +Targeting version +The Temp File will not be deleted after reading from +I just committed this +Once this patch is in youll see it for both +Not sure what the best way to make a spec for this one is +Hi Joedefiningselectorxxx on a queue name doesnt have an effect youre hoping for +Hmm +I added the preference for this issue now to the annotation editor preference page +Ie +PR sent +No wonder it would failI have renamed the task here as its really an activity someone needs to doI will keep it unscheduled as we dont have resources now +Imo the whole underlying workaround is not needed +applied to branch and trunk. +Thanks for integrating Ted. +I am not still convinced with your argument and now believe that those assumption should be removed from API +If it is then I would concedeMy suspicion though is that it isnt although I can certainly envision where it would be useful +point me at an example and Ill rewrite the Im stuck in land when its apache +bq +Hi PaulexThe fix looks fine thanksBest regardsAndrew +running +Completed in the SIXPT branch +For this issue I have created a which can be used to store an object from an action to a single render +I will fix this for the next release but for now I am leaving it as it is +The patch in just adds validation for theservice name length and improves the error message +Any updates on when is this being aimed for reality would be helpful to plan for +Patch looks goodUpdated patch removes the unnecessary stacktraces and updated the messages +This issue has been fixed by Michael Glavassevich and commited in SVN revision File But for some reason this fix is again missing from the src +Java EE runtime doesnt require Java sources thats true +A cluster with replication enabled will most likely be reading old logs +Attached patch fixes as requested and html file is included for review +Thanks for the extra info JustinHave added a Release Notes draft +Check to make sure you really are using the version of +Patch that appears to address the issue +Its impossible to know if you found all occurences that need to be changed but your patch seams to cover quite a lot of them +Was just asking for clarification to see if I was miunderstanding. +These blockers and critical issues are resolved but not verified +So its more a matter of looking at and documenting it accordingly I dont think any change is required +Ive just uploaded based on your implementation of +junit is included transitively from jdrom +here ist the stacktrace unfortunately appears to hide whats actually causing the error so Im still lost in the woods +Yeah solves the same root problem. +That induces a test failure +LGTM . + for the patch +exampleslogs +We also have a team that has this Id say half done +Thanks +We need a set of Camel JMX annotations we can use instead of the Spring JMX annotations +You should upgrade your versions of both Xerces and Xalan +Can you check it is it still ok or can you fix the translations there and send it back please +looked into the tomcatcatalina sourcecode and found out that between and the order of the steps inside was changed stop children contexts stop filters stop listeners stop stop children contexts stop filters stop sessionmanager stop listenersSo in my case will be stopped before session serialization but with tomcat the listener will be stopped after session you know something about this change in the tomcat lifecycle +Attaching the patch using the name as it feel more natural to me +Keeping this one for +Your comments please +Is this resolved +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Followup for + +This doesnt actually hold true in the case of jvm reuse +To the best of my knowledge JSTL only works on JSP +Not sure if Nick is monitoring so a comment here should email Nick on it +Added a test what do you think +Tuomas initially asked for another interface to our mailing lists but apparently nobody else thought this would be useful so Im closing the issue now. +Thanks +Please help review +That ends up being quite easy to iterate over in Ant +Yeah sure +Resolving +In other words it should not be done just to deliver more semantically named objects to SOAP clientsCheersPaul agree also agree with what if I try to implement a different service so that it can deliver strongly typed entity objects to soap clients without affecting the ofbiz architecture +dims +from an API point of view i would rather prefer ifthere was just a single way of getting the status etc +Personally I like this approach better may be we can augment the one in YARN where applicable +r was used to execute this test +I think the treatment of fields makes sense. +The source assembly should exactly match a checkout of the source tree +We need to understand the issue and decide whether this can be fixed in the WSEE layer without breaking existing deployments +at least as it is described here +Discussed with Devaraj +This will need to be investigated +incudes both the System and the Device extrinsic operaitons and the use of for handling complex properties +A CME is still possible though much less likely if another thread modifies the original Map in the middle of the iteration needed to make the copy +Would it be possible to create dynamic RF values based on live endpoints that would work sort of similar to how consistency specifications work? ie ALL equal to the number of live endpointsQUORUM equal to more than have of the live endpoints perhaps other constantsthese constants could be mapped to negative value constants i suppose but would be calculated at the time they were needed and not at startup +thanks stan +It was from apparently applied on the wrong dir in the source tree +binhbase hql to get old deprecated and beanshell have forbidding licenses theyre out of the running +Patch attached +Not going to have time to focus on it +Werner freezes on load after I installed thats all + attached patch +The url should have database namealias like mydbin the url jdbchsqldbhsqllocalhostmydbor it should be specified when server starts java cp +Closing this for now as the javadoc has been released closing. +This is the main issue in the current implementation +I checked out your topic branch and ran the automated storage tests fine on iPhone iOS and iPad iOS +To put a number on that I see test suites that currently exhibit this problem on Windows +Attached files to demonstrate the scenario +That just leads to a lot of emails to the mailing list or frustrationI think our default options should be and not something where you immediately need to read the docs because the application hangsSo maybe we can patch the to make those options available by defaultOr we supply our own custom shell +Ive also hit this identical issue on +If someone wants to toss a monkey wrench in the lane is wide open +Nor do my actually require the Lucene DocumentIm not picky on the details but if you see a clear way to achieve whats needed for all the use cases you can think of at the moment limit the exposure of the Document internals so that s delegation to the doesnt let it muck with other fields +Related issues and is lacking any unit tests which would prevent the fixes from being protected into the future plus serve as a demonstration of the possible problem +Should I download the source from SVN or the last nightly build +First patch had mistake in it +not needed +The patch looks great for me +Will fix it when merging to trunk +This is hard to do on a generic way +Please grant me the commit access +NOTE the groupid has changed for ant from ant to +Different race conditions seem to apply in each class of runs +its my bad though so if no one can make time Ill figure something out to unblock the release +Does it also work if we make? If so I think we should do that since we probably dont want to be used directly +Therefore it does not make sense to determine the time of day from the position of the mouse +I understand if we want to steer clear of the OS logos but I would like to use the HTML logos if possible +Meant to say feather logo +could you resubmit with a diff format +batch transition to closed remaining resolved bugs +Part of the fix for was to reorder things like the callbacks those now occur dead last after all rendering has completed +I think conversion to the same value with zero scale is the correct behaviour for SQL. +Simon looks like we were working on it in parallelThe patch you added only takes care of that particular test case +Perhaps the fix is to make nodes ignore gossip schema from nodes +so columnClasses should onlyapply to instances +on patch +Expected Behavior Upgrade should stall because at least one replica of every block would not be available +Yes ive seen this +My initial feeling would be to add a attribute to the link element +Now each Ajax call has its own instance of attrs. +Moving to post release and marking it as so +Indeed its working for a long time now I thinkAt least it works completely fine when combined with Studio +I do not know the details Jesse let me try are correct it opens up the document and shows Pictures Thumbnails Configurations mimetypeSo how could we accomplish this on the practical side of things? Some script you call to explode an odp file and another to gather it back up againCan I assume you want to explode the as well +Submitting +I think its controlled by a different subsystem which means in both ISPN subsystems could run +defer past +Take a look at the patch and let me know if this is what you had in mindAs for the correct me if I am wrong but I thought that this was needed to lookup the BS in the CLIENTTOBS table +Ive had a look at the code and it appears that the jvmRoute processing is done as the request comes in both the in the cached session is modified as well as any information in the Cookie in the response +EWP uses the same codebase with EAP rightis so this should be document as a known issue the fix for is not in the current used by eap IMO this is not just a change in the way the code works it changed the contract of as a fixed issue in the EAP release notes +Current VPE Templates Preferences Page screenshot is attached +It makes sense to me to clean up temporary staging dir when the job was successful because data is safely moved to hiveSo I think that it is better to handle your suggestion in a separate jiraThoughts +We decided that we should try to provide some verification that the user would trigger by hitting some button Test +Ive set up a temporary mail forwarding configuration that should allow issues to be sent +Also is there a race condition which we are missing? If the last mapper throws a fatal error can we miss it since rj is complete by then +No worries! Thanks for the submission +Super! Thanks a lot Ognen +Sorry what I meant were the Unlimited Strength Jurisdiction Policy Files not the JCE of courseAre you sure this is an issue with a standard JVM +Added ignores to project +The index by NORDERNUM was added. +From this the implementation can grab the last product and use information from it +Heres a patch with a unit test for this case +committed +I should have pointed that out +Please tell me exactly the package and file name of the code that needs to be changed +After those changes this issue seem went away. +And I named it I went to ShowHide Drawers I saw that was already there just not I clicked the Palette Editor navigated to JSF selected Tomahawk right mouse clicked requested delete +My patch is attached +could you include tests that produce the problem that this fixes too pleaseAlso Id prefer one diff using svn diff that marks out each file that gets patched +I am thinking this can be entries in the slice. +You cant use external properties files nor external XML ended up describing my properties in the file which is the proper way to do my case I have unit tests that need to have access dynamically to some of those properties +Rescue was not the for that expression and essentially never got hooked up +Yeah the callback must be invoked in the async method +pending jenkins +Unfortunately recipientlist only works for one way calls +I like your cut at a better algorithm +Can you also provide a patch for trunk +Please use only spaces +Apologies for the false alarm +thanks. +Thanks to Ivan for reviewing the patch +That can cause badness if you dont have something watching space in tmp and you are in a shared environmentAlso note that you can easily specify it viaGiven those things can you provide further arguments for making tmp or some other location thats not cwd the default log location +Ran a test job which required two lib jars and it worked properly +I have my m repository in CUsersJukka Zitting +Here comes the complete +As far as I know nobody is working on it at the moment so its all yoursAlso keep in mind that the driver shouldnt depend on external libraries such as boto +Thanks +. +Yes +Hmm +for this suggestion +I guess the UI breaks down there but I cant think of any way to handle all use cases and still solve the major one +Tested against hdfs with new common jar and no problems +Luckily we had someone who could wrap his mind around it +Please verify +and committed to trunk +no no my patch is invalid so you did all the work thank you Ian +Thanks Ill take a look +Thanks Oli the patch was applied as expected +However for updates to be persistent a is required +the two patches all resolve the http errorAnd the contains the properties file for globalization usageAnyone can help commit itthanks the changes to snapshot at revision and trunk at revision +Why has nobody responded to this case yet? It seems like a pretty major problem to me +Thanks to both of you for the patch +Resolved issue awaits confirmation +When is the release for version planned? +The issue reported by Ignacio Coloma is also reported in +KasatkinOne quick question will this work for other NTLM authenticationRegardsReflex +HiIm seeing two test failing at the momentCan you please on your sideGreat patch by the wayCheers Sergey +Im giving up for the nonce +I kinda figured you might conclude that some refactoring was in order after looking at my patch +patch looks good +Committed +these templates work rich rich rich aj still doesnt work in rich. +Moreover KDD at and map tasks take exactly the same amount of timeOuch Its a copy and paste! Ok Ill be more careful with the next test bq +The easy workaround is to use tar files to preserve x bits +Dont know what caused this issue +was thrown while evaluating an expression +These strings will come with the translated +I looked at +I want to thank all the reviewers for their helpful feedback! Those are very good suggestionsI am hoping to return to this bug and continue working on it incorporating the various ideasI was dismayed by how hard it seemed it was going to be to make a variation of the Network Server ping API which was silent +There shouldnt be any differences between in JBDS +bq +For instance they might have to change their exclusion patterns etc +This is fine +Thanks +Now truncate means changing browserSim measurements saving aspect ratio and changing webpage request is rejected because we cannot provide adequate s why we doesnt add zoom +Addressed the comments above +Updated to latest trunk changes +txs to Btt for the contains proposed solution from revision and updated branch +for the patchThanks for your contributionIve just committed now. +Is useable for some of the other maven goals as well camelrun etc +I can see how there will be incompatibility issues with and any other system that uses jars that contain different versions of the same classes +Yes multiline attributes are possible continuing lines must start with a single when I posted the bug Maven was not obeying the rules if I remember right it also even inserted a fully blank line which is not that helps +The change in API to is required because it is not an inner class of anymore +And if s finalizer executed first it closes the handler and then the tempfile module executes its code but the handler is already closed hence the by adding the call at the end of the tests eliminates the m not sure what to do in case when close is not invoked directly and executed from shutdown hooks +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Confirmed this does not happen on tomcat only tomcat +PR! +been solved by. +Renamed the interface to +Can you provide the information I asked for +works with hibernate +The tables were a little large for a phone +Provide DDL request was accepted +bq + for the patch +I agree with Alexei +I just committed this to trunk and +Yeah it looks good to me +We need a way to unit test the Ruby shell code easily +Committed with revision +Thanks. +Surely having a stdout stderr policy is not too much to ask +To do with the intermittent failures on and +Logged In YES useridSamples have been updated +Seems to be now fixed in meantime by believe this already be fixed by reopen if you found anything wrong thanks +Edit this issues subject so its about adding an option where we stop processing if data file goes missing rather than just log it and press on as we currently do +It hink in retrospect the fix is actually wrong and we should implement this behaviour differentlyConsidering this a regression of my Rev +log has details without Pnative and with Dskiptesttrue failures. +Ill keep an eye out for CR in the Maven repo as well as an AS Hudson build that uses it +This is really bizarre +Should move too +Yeah I did some testing +Jesse nope because that test depends on a bunch of classes that are only in +So Im closing it as fixed. +Thanks Keith +Thanks +I changed the cookie time span from year to days +Check that theapplication server and JIRA have permissions to write to it Cannot write to attachmentdirectory +mvn Prc package deployBTW unless you have the need to run your build twice mvn deploy is sufficient +might be removed before we release because well simplify the API so we dont need any frontend API anymore so it makes sense considering current situation +reviewed the patch +I think there are many separate issues +Attachment has been added with description A sample grayscale JPEG has been added with description Gray to CVS +So we can revisit the selector approach later +I wrote a test and sent it to them +Duplicate of +set ZOOCFG in the environment and it will be picked up +we dont have remote locking detection in current versions and it would require all deployers to honour it anyway. +By the way how often do you guys commit to svn? Its kind of painful and to keep passing around patches for a long time +It would for example be able to create a certificate with a webid that has not associated Profile and see if it can still loginSuch a test suite can then be deployed against all webid enabled servers in order to build a profile of the capabilities of each +See my previous comment +Ive also noticed that the current domain match implementation matches with +In the latter case I use the same namespace for both reason I had tried to use xsimport instead of xsinclude in the first place was that I was trying to get rid of the schemaLocation attribute from the generated WSDL +Also as far as the reformatting goes please send that in as a totally seperatebug +made the edit to the attached the script I used for doesnt have the issue with using shows errors for and HEAD for me +Moving message transformer work to M +Maven cannot find the artifact if it hasnt been installed + Possibly related when you first mark the jar as deployable and deploy to a server the context menu item changes to Unmark as Deployable +The filenames for these topics areDROP INDEX DROP TRIGGER DROP VIEW It would be good to add a section with the title Examples and use the examples from the CREATE statements for examplesunless someone has examples that they want to provide + +Checksum will be calculated and verified per transaction level +Ted the is parallel +Just to avoid any confusion the actual artifact along with sources javadocs and all checksums is inside the jar as documented on the maven web site +I think it would be good to implement CORS support independenty of the implementation and +But anyway I upgraded to RC with aspectj jars and everything seems to work fine now +Shows the structure of the code but comments arent yet included +Im sure theres some holes in my notes +Correct patch this time +postponed for a pitty. +The rest works fine though +Closing it. +Change affects version and fix version to +Would be really nice I agree +This mean one could use exclusively the url httpservermountpoint instead of httpservermount can be reproduced by setting the port to in then running the integration tests +A new JIRA will be created for the new design and implementation. +shipped with +Thanks for catching my mistakes Jason +Thanks Kristian +Not sure whats the best way to divide these into a sensible package layout +Here is the actual patch +Thanks +Rich please justassign this back to me after the patch is appliedWe can keep this bug open and I can try to createa fix for this +Regardless I will try it first +When I find it you may base some tests on it but perhaps you mind start with drafting what you would like to test from a bird perspective How you assure that you have all XML elements attributes belonging to the table feature covered by the convenient API and tested to be altered +not to talk about and +Works for JTA and JTS when using +Thanks TomI put this into trunk and +Upgraded to POI Final in r. +TMS per entry point. +Therefore its probably not worth the risk of fixing in since we would like to release that ASAP + +Keith Id really appreciate your comments re my work +Attached an updated patch that takes maximum shares into account +hiI have the same problem mentioned here about the loading resource depending on the component order in the page +Revised scripting for your feedback AntonioFor the exception I see three solutionsa we find the place where it gets wrapped and avoid itb we unwrap exception and look if one of the wrapped exceptions is a c we use the request attribute approachI would like to avoid c and also b if this happens which exception type is wrapping the +The ejb jar file +Are you proposing passing a command line within the command line? likeoozie sqoop command import connect jdbc query select from T m config default +Test case that illustrates the issue +I think this is likely to be extremely rare since get resolves symlinks fully and symlinks are mostly transparent to the application +The problem seems to be caused by an error in the underlying image pyramidWrite protected the properties file from geoserver causes the exception from the underlying error to appear in the log +And even then it is questionable what kind of changes we could make to prevent this in the future. +Patch to fix the issue +According to the discussion I have removed the module in r +and should not be in the API jar +Sorry Eugene I didnt realize about that checkbox my fault +Lets try this again +I think you can also enable the debug in Archiva to verify whether the download went through the proxy +As far as I am concerned this resolves the issue +let me know if you need more info +I can workaround the problem by extending the class as attached +Thats a remarkable improvement thanks NikolayPatch applied to REGEX module at repo revision rPlease check that the patch was applied as you expected +I think the behavior you are seeing in a Page is Auto binding adding the links to the Page object which in turn processes them +This additional patch fixes a bug with boost being lost +Can you try this out +Now reimplemented in Java and can no longer see the deadlock +We need to be able to overload for methods with no parameters +Robert will upload a test patch soon +Ill check it in +Looks good to me. +The tests for CPD doesnt seem to workIf anyone has any ideas of what might be wrong Id appreciate it +Indeed otherwise it requires another configuration step for the basic user +I am not sure which way the coin should drop in this case +Ill link it to this one and close this ticket as duplicate. +I will commit this patch Thanks committed +Please note that as this is only a partial change it currently breaks a whole bunch of unit test cases +For reference I attached to a document outlining my approaches and why I took themDiscussion is welcome +Im glad to see that I wasnt crazyAs for the part about it ignoring the copyField line that seems to be my mistake +I can do a with just this patch +generates error when parsing time with comma +It looks like this issue was fixed by reports issue fixed no test case given to test the issue. +Thus when the client code using the cache type casts the key to a the server throws aWorkaround is Turn OFF result caching in the server +The issue happens because JPA simply casts to the transformed classThis is a possible patch +Since only on win and only in very specific case not critical for GA + has been marked as a duplicate of this bug +Rather dont start anything this should be elaborated +Trivial patch for trunk +Added a first proposal how to begin the migration to Tomcat with integration tests +Sorry wrong issue eek. +The goal is certainly a good one +will need to be updated as well +But it is less clear when you generate a pull request on so it is up to the request handling committer to IP clear it and possibly with the pull requestor +Is any work being done on this issue? We have just created a web services application that passes maps around and are having big problems with this. +If its required Ill add the necessary hidden element instead +Yes thats the basic idea +Cool youve marked this for releases +well for now how about someone commit this little tiny inzy binzy patch file to make us eclipse users happy +Doh forgot to add the exception edited in nowThis test is a running in eclipse +Thanks XuanTx for the early reviews and the Jian! +Id be tempted to close as wont fix +Fixed several issues with adding and removing variables in the input editor +My Bad +suggestion component really should just do that +Hi Andreiok thanks looks good now your mirrorwill appear in the list in an hour or soRegards Sounds good +Did you test this patch on WindowsYes and both the failing tests passed as I mentioned in my previous comment +was serializing tokens using toString rather than +Id like to make sure I understand a couple details +Lets use as the all basic tag attributes +Revision Jarcec +Updated patch with review feedback +I did not touch any other java source files only created the two news files can I just attach the new files +has always been broken this way it is broken by design +uh? What is this test tagAttach a test project that reproduces your problem +dblook test redisabled with security manager so this failure no longer exists. +updated the limit check to include this based largely on the passing unit tests +Please also attach the Just uploaded the project +I meant that localhostNormal otherBootstrap does not exercise include pending ranges code since local node ranges do not change during BS +The server will try to delete the message from DB +It is unfortunate that the CXF community is leaving some users out in the breeze +See +SCR has been released +This is a major show stopper for us +I have reviewed your patch and patch is ok. +This should be corrected with the latest commit for +Bulk move issues to and applied this patch on branch +jar from build path and add project as a dependent project of +properties and activation package from the in Rev +bulk close of resolved issues. +ZK callbacked is delayed with ZK state OPENING +The problem there is that are processed in a phase with bean bootstrapping and hence bootstrapping kicking in one phase later +Ill tests with latest +Ive just applied your patch to the maint branch +It attempts to create not only the inbound but also an outbound wire and then run in it through the wiring connecting stage allowing for an needed interceptors such data binding +Ill double check the patch +Patch removing Xerces dependency from +passed for Y! dist patch +Im sure that there will be some sort of issue with them on the DLL nothing ever seems to work right with +Try tracing CREATE KEYSPACE or CREATE TABLE there is too much information all in inconsistent formats +this code doesnt even compile +There was a problem in and some of these issues relate to that edition +I did and it works. +Usability and customer sat is important thus this is made a blocker for SOA +Gary if you still have problems please reopen and provide more details for us +Use the root user on the box is running on as the super user +We havent had any luck with the proposed workaround either +Ive seen another use case mentioned of a dependency on a datasource which is not available at shutdown for would be a bug +Resolved tested with point features with default style with personalized style in mac in Linux + test case java class and mapping file +I have tested it and it works fine +done for the first task +load balance the first and third nodesI could reproduce the error exactly when one of my bootstrapping nodes did not contain the keyspace definition for the ranges the existing nodes were trying to send it +This issue was resolved in +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +I think the proper way to do this is to run the class from the command line +Why is it taking hours do we know +The fix I created for only fix bean +Could somebody review this pull request so that I can make any changes if necessary and get it merged? Thanks +Just committed this. +The proposed fix is easy enough that I dont think we need to restrict the use of +class file +pending jenkins +You dont care to fix it we dont care to use it anymore +Please cross check and close. +database tables +Ive also updated the code to use the new +Thanks for the patch Jimmy +Committed this +So it will not make it to initial platform release +As another option we can have a config parameter for disabling updates through this UI +DavidJozef can you confirm +Patch using the described approach +Hi allI am new to the Tajo project +Basically is just plus a handful of bug fixes plus the compiled JDBC drivers +Props to Xin Huang for finding the JIRA defects and Rich Scheuerle for suggesting the fixthanks revision +I noticed a reference to in the wiki +Great work but why dont we link that from the Thrift Wiki? I remember that it took me a while to find out how to build Thrift on Windows even with the Wiki info but without this stuff here +I hate to bikeshed here but I have a small problem with the use of forced removal in NEWS because the hierarchy of node removal from least forceful to most forceful is decommission removetoken removetoken force assassinate +The ddl for the dynamic serde that is used to serialize data at this boundary is double long where as the output generated is long double as seen in the plan +Also some more specs added to rubyspecs. +Theres an even simpler approach +a bare bones pom that displays how the problem occurs +btw the new findbugs warnings in the results above seem to be truthful +bq +Patch committed to both trunk and branch. +Example of how this might be used for to follow +Wed better update the documentation besides from the release to provide releases without bugs sooner +I think we can do this more simply by prohibiting the creation of s whose valueClass is null +report plugins like clover can handle empty src dirs I think +Any objections to closing this as wont fix +Tested and both patches work for me +This is the that I used for this crawl +Is caused by the different structure of maven select image dialog expects the images under src not srcmainresources +Code based on Mitkos work in Apache thing that I noticed was that Argument inner class trims the value data therefore if a command line has ABC then the space is removed +Change the value of CHUKWAIDENTSTRING from demo to in +upload a revision Thanks Jianhan +Roger is there any chance you can add specs for this on the rubyspec project +Anyway I have run all tests and every one that performed successfully before my changes is working now also +Thanks again Vinay Uma and Brandon! Ive committed this to trunk and. +Can you please send the logs when you execute mvn with the X switch so we see the detailsAlso youre using the same CATALINAHOME and CATALINABASE which is a bad idea +Ugh ok +Until then for performance tuning it will be very convenient to have it configurable +In the particular case of the queue scenario it seems like the head and the tail of the queue would be forced to be synchronized +If I use then the +Thanks Shawn Jiang for the patch +After first showing the correct splash image this smaller version Brion! Ill take a look at it soon. +The test should cover all scripts +I believe all this is done to avoid requiring the fetch of hive jars for the basic piggybank build +demo with patch for tabPanel project with added solution work only in client mode I dont know whats wrong in the others +If we add them here expectation is that they should be every where and they are not and its too late to add that +The design to retain user modifications will be made post while the warning message for now shoud be put in has been no movement on this for some timeAnd it would appear that either everyone has resigned themselvesto the fact that you cannot simply edit the or few peoplefeel that there is a good reason to do soIf there is sufficient interest a new JIRA should be opened on a morecurrentactive branch. +For the record the second Olas commit bae fixed the problem. +In any case early cleanup is of course desirable and there may be scenarios where it really matters which is why were glad to have heard about this problem and are fixing it in both of our active branches nowJuergen +Looks good +Thanks +Attached the new start menu icon +Minor update parameterize types and use the new close paramter from +This patch fixes the case mentioned +That would be fantastic especially until we get bootstrap am taking the issue in order to apply the solution from Nov AM +Thanks MatthieuThe next time please create patches against SVN repository +some kind of reserved flag might be optimal here instead of relying on id values andor ordering +Thanks for pointing it out do you mind to attach a patch file to this issue +Changed the common log to slfj which is used in camel by default +One way out is to put our properties in System property at bootup +That looks nice +bulk defer of featurestaskswishes from to +I have one quick clarification and one request +Hi ArnoI still havent isolated a repro but I did try reapplying the old attached to this bug on and it fixed my problem! So maybe this regressed somewhere along the lineIn on line I changed to +Also fails in Eclipse +Please try to provide a minimal test case to reproduce this +Focussing on the class we already acquire name and content attributes from meta tags +Verified this in the offline document and the changes are made accordinglyThe docs need to be added to the Apache documentation for the +Was that done +Patch integrated to trunk. +We can not disable Facelet HTML Validator not for all kinds of problem so we should show this quick fix for all kindsIm not following what you mean here +from me to proceed with committing your patchLets get it into the codeline and start getting some wider experience with it +I decided to put the plugin in as a concrete plugin with its version number +Patch v from review board. +Thanks for the heads up and if you dont mind please keep me in the loop +Any inflight messages with id that that sequence are deemed to be the even of a consumer failure the close info from the consumer is lost all unconsumed are deemed redelivered and there can be false the event of a broker failure all redelivery information is lost +Just to be clear my contention is that when parsing M and converting them to M you should convert from the syntax to the new improved artifactIdgroupId format +fixed in trunk as of rpatch applied with a slight change to compile with Java sorry for the delay +Moving function queries is going to take some time since they are very coupled to Solr +We have bigger fish to fry with Network Server +For hanging threads this will avoid the seconds delay which I confirmed with the test +Ill take this one +I guess you could also have different hadoop home directories in the tarball with symlinks between themBuilding two entirely parallel HADOOPHOMES for different architectures out of a series of symlinks? That sounds like a disaster waiting to happen +Cannot reproduce the problem using Win Chrome will not hold up the M release +Ive got another idea This one might be better anyway +Introduced a capability to handle unhandled flow exceptions on a per flow basis in a custom manner including displaying an error page with the option to start a capability to start a new instance of a flow that has expired or ended automatically. +I recommend we keep the code for the longer term planning process and refine into smaller package sets until packages ready for Debian distribution +This issue is tracked in +i think i was wrong +wasnt fixed there +Ill be happy if we can find a solution that is closer to the XML Schema spec than the one we have now even it is not perfectThank you very much for your time and efforts +Appears to be the same issue I just fixed the exception in the description +fix typo scale fixes typos and are you satisfied with the multiple invitations of mxj case +Daniel can you please rebase the patch for trunk +Tagging as waiting for user input +It does not process the request and response elements in imported files it only allows you to refer to elements found in imported files +Ok I know that works so whats the DEBUG level in forIs it a joke? +Do you think its really needed? Manual replications tests worked +This is consistent with how row reads are handled I thinkFor the quorum case I added a check that repairs rows that are missing during collation and reduction in the response resolverAn open issue remains as in namely that replicas not involved in the quorum are not repairedPatch against tip +One detail for maximum flexibility the nd argument to the protected method doc might be replaced by the indexreader +antcopyLine +added in the test application +Just copy a jar to the deployments folder start server and watch it crash +Also all those export packages are added to the meta data of the application +Can you confirm thisIf this is the case then please rerun your tests using the external tomcat serverDeploying opensso within the server was never supported in SOA platform and has now been removed from the quickstart +Just a couple more comments +The patch looks good to me +Why a class instead of a struct? It would be faster for stack allocation. +Only to the builds +If theres no objection Id like to close it while Im working on the release notes +Sorry +Yes it is a hack but the hack circumvents the OS issue. +alex do you plan to upload a patch for this +Were using Springs in the clients and it looks like Camel use Springs to send a reply on a request reply +That parameter would be problematic as this is only valid for strings all other types cannot have no value +The typechecker seems confused about formals in constraints +My edits attached +If those test changes are uncommented the test will fail +I will have a look when I will get a chance + +Patches to be applied to enable building in mead +Bumping the urgency of this issue +Thanks +We have a feature of displaying errors in formtree editor in Tools XML editor +For this can we just do shutdown for and lets admins restart the? Cleaning up the data structures to do a clean restart might be a little hard to do +Im not ready to check this change in yet +Closing resolved Release +Issue may already be resolved may be invalid or may never be fixed +Hi AmosI am assigning this to you while Mike is under load +Updated with comment on why is passed as readpoint in certain placesAlso replaced smallest in tests +Done +diff as submitted by r all drt tests pass +The installer is long gone we arent likely to try to resurrect it in anything like its original form. +once this tool is usable you should discuss with Koen how it can be integrated in the process good base conversion is work will continue in the next release +Container! slot nor A fixed amount of RAM +I just committed this. +I refactored my code and everything works like a charm now +leaving issue open as something to handle better +In addition to variable support a has been added to provide functionality equivalent to Logj +Closing stale resolved issues +Just tried with OO and it fails for me +Testing using HTTP request +reopened for fixVersion +This is easily done with jconsole jmxterm etc +In which revision has it been fixed +Sorry about that +Given the concerns of some people around automatically registering a dynamic field perhaps we should optionally allow a sub to be passed instead of sub +Post logging configuration questions to the mailing list. +Thanks for the work around +I discovered after adding my comment that awt stuff was already broken +I made a slight change which made it friendlier +Im not exactly sure why this would happen unless there were a bad plugin installed somehow +Do you foresee a case where the header passed to a builder who be modified by before the format is built? It would seem odd. +Without this code behaviors like ajajax and richvalidator cannot be generated by CDK +I added some assertions to ensure that a concrete ast is never requested when the system property is left unset +fix testalso convert test to use Sami! Committed. +Is this also a ofYeah I couldnt find this ticket beforeBut my guess is most of those are related and most likely will be fixed in one shot +Im currently working on the NTLM auth as it can be usefull for me to use on other protocolsIll mail you James for some questions and maybe api improvements +test CXF bc work with CXF se to handle external request return the correct Fault type defined by wsdl +aj and ui repeat isnt applicable in your case +Its not the right solutionThe questions for this one is If the user adds relevance as a sort but then also asks to track doc scoresmax scores how should the collector workI definitely dont like the idea of more specialized collectors god knows there are already too many but maybe we can avoid this +Thanks Isuru for the contribution +Just found out that the thing I mentioned already exists so moved the fixing logic to the appropriate placeThe test is still failing though due to problem in instead now +And since the actual processing of the stream into Comment format is pretty unique to the file I really didnt see a need to generalize it for Files +fixed +Looks like IBM issue +Thisll be a patch on the cloud branch to reuse whats started I dont see any core management code in there yet so this looks complimentary +During that patch the was split into a and a with a super class called +Just to mention it but there is nothing that prevents users from doing it them selves in their own configure in current wicket +My instinct tells me youve made it +Im going to try to combine the two +Lets fix this for Spring which is scheduled for release within the next two weeks anyway including not only manifest fixes but also fixes for regressions in the framework give a snapshot a try once Costin fixed this so that we know that it actually works for you was not a problem for however I improved the manifest so now they are updated on each version wo manual intervention + +moving this over to you james +Ruby and have the same problems displaying unicode characters in IRB +If the id exists then we consider it a modification +Im testing a fix using right now +Hi JacquesMistakenly I forgot to add ASL header +Makes sense +does this also apply to +Security Roles +There is no file datafiles on trunk at the moment +Vladcan you attach a diff? id like to see what you changed +Ill check that again in M how about we close this JIRA as Wont fix? With the universal installer we cannot change the jar into a app +it appears to be a significant bug to me +I think that perhaps an exception is being thrown so that a lock is not being released +And the from Maven Central matches what is available from the site as the current downloadYou said you viewed the code of +Can you tell me why you are trying to clone the configuration? This behavior was added by +defer all issues to +Fixed with commit aceccbaedefdfeaa +Student asked for it in class +Tested works fine with the newly rewritten KML module shipping with +I missed that part +bin localhost messages threads topics sequentialbin zookeeper localhost topic sequential python B f Output from script Checking ordering for thread All messages are orderedChecking ordering for thread All messages are orderedSo far using this test I couldnt reproduce the reordering problem on the draft patch but I ran a short test +Thanks Jeremy I backported also in at for fixing this bug actually crashed our system consistently because we have been running for quite some time and have hundreds of thousands of accounting entries! +ready for reviewcommit +The suggested patch fixes the problem for meAdded global lock to prevent other threads from starting and terminatingGregory please check the patch +Take a look at the patches I plopped in +Good news Dan! Can you share which hardware this runs on? My running theory is still that faster will make this show up more +Patch applied to trunk and +stack No problem +In reality doesnt this ultimately mean that we should not be performing upgrades if an editslog existsNo +yes its possible +When combined with this will actually be more efficient and also avoid the bug described in this JIRA +I thought about that and figured we could recursively make the name longer until there were no conflicts +You cant view the Help page without being already logged in +Well just remembered to provide an explicit model to the grid and it works just fineEnough for me. +So again for the Hibernate LC case adding this API method doesnt guarantee correctness +Obviously that cant be implemented in the parser +The current structure is a wireframe of the directory path layout by Owen for the greater hadoop stack +If the above patch is correctaccepted then the same logic can easily be applied to as well +Ive been trying to reproduce this in a sample application with similar code to what you posted and I dont see this leaking +Hand editing anything for a build is bad all build scripts info must be checked in for a repeatable buildIll get rid of the copyright entirelyIm rethinking the one RC file approach and tending toward one RC file for each object +Fixed under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Its hard to know what to clean up especially in the face of overlap with other plugins +In my understanding this particular test is manual right +Xalan caused the problem stated in we need to know the list of things to put in the repo fop already has a pom in the repo that can not be changed the patch is against and old revision and doesnt apply cleanly patch says revision latest is like I added the wrong patch file +the specific problem should be fixed since also in there are new tags to allow for more fine grained control over HTML layout +Moving to is still more that can be done but I am resolving this issue since we now have the basic monitoring capabilities and its a good opportunity to get some are now providing and names might change +This is no longer valid as we ended up removing Tomcat from the izpack installer for M. +Thanks +Contributions are most certainly welcomed Venkat! Feel free to assign this jira to yourself and start working on it Ill be more than happy to review it + issues for released versions +code changes look good +And now with class trunksrcmainjavaorgapachesshdserverchannelSending trunksrcmainjavaorgapachesshdserversessionAdding trunksrcmainjavaorgapachesshdserverxAdding trunksrcmainjavaorgapachesshdserverxTransmitting file dataCommitted revision +Ive verified that the jdbc suite still runs cleanly with these changes +Any reason not to update all the stemmers with this fix +Thanks Josh! +Thanks Ashutosh +Are you saying the error is just that this caused an invalid error marker? Im not sure that is the whole issueI cant seem to have my filesystem resolver work unless I only have the property +I am closing this bug but the work is not really finished +Verified on. +Will be covered by solution to so assigning this JIRA to myself +Maybe . +Pushing to to to multisource planning to allow for an explicit multisource procedure parameter more validation to both the procedure and insert the handling of insertupdate when no sources add docs of multisource in general and the procedure scenario in particular +Also renaming the option to contentCache is indeed more consistent +all issues to the CR +Is one of the items pending in the backlog to change the region servers to push more information about their state to Zookeeper? If that is the case pulling it into the master from there and exposing it via another bean should be trivial +is specified by API +I forgot to run the previous test in mode +Go commit +Brett any updates? Its been days or so now +Yes +I get a on that think this is Emanuel Muckenhubers test +patch looks good +In this strategy used new which can speed up implementation in oracle +attached is a patch that changes the javadocs for the spatial class to will be removed in a future release since it is experimentalif no one objects i would like to resolve this issue soon it would be nice if someone could at least give a quick review make me feel more comfortable +Thanks Robert +Error appears to be in jbossvfs file srcmainjavaorgjbossvirtual line and is marked with a TODO comment +I just committed this +I have raised this issue to Mojarra team but it seems they can fix only in release +please comment on it +Feel free to provide a patch based on trunk code +Ive committed the additional jsonDecoded! null checkHowever I am curious where what platform php version your seeing this behavior? While the documentation states that jsondecode will return false on a json structure that is deeper then nestings on plain failed decoding it always returns the original string to me both on a mac and a linux host +Heres a patch that fixes this +mrepositorylogjlogjCDocumeUSERNAME +Merged into the master branch. +Can this be closed +Please attach a simple test radiobuttons in this example are in the same document and have the same so tabbing should not go to the second group since it really is part of the same groupI thought your problem was specifying in different MXML files and combining them into a single UI +I think searches occur much less frequently than a literal access and the cost of a more expensive literal access may not be recouped by a faster search. +relativePath entries were added to the poms. +Like I said before I am not that familiar with the IPC code and all I did was a quick pass through the patch without digging to understand deeply how each part of the code that was changed was used by the client +We tried hard but we cantWe know this feature is quite young and thus can lead to problems like that +The check performed is actually for the current authentication having an Administration ACE for the ACL being modified or being the think the proposed solution is still pertinent because requiring an authentication to have an Administration Ace on and ACL to be able to modify it is not very flexible +This code isnt mineTry using the code from the spatial branch instead this is what Im using +Patch to correct the issue +For example +Do these patches need to svn patches +SDK came with a pretty early version of Xerces + are not promised to be threadsafe +Looks like this is duplicative as this +Heres the latest patches +But why do I need it in the first place? As I said I already specified that I want only the by using it in the relation. +Can you try mvn clean install +Applied fix from Nathan Bubna in +My advice would be to avoid deleting a queue in this state if at all possible +is submitted for hudson tests +So of course I will make the patch for Windows +Barry can you look into this +Thanks for the details +bulk defer of issues to +Michael. +Agreed that it would be a good change to include in all schedulers +OK I see that makes sense +Assign to myself for backport +According to Bela this message is harmless +Indeed config files are accessible through permalinks in that respond to your need +I wanted to remove the final keywords that I had added myself and did not notice that I removed some of the existing ones +Field name is pubdateiso +CMS default please. +There will be a cpptasks release before the logcxx release but holding off till both are ready +Dan please realize if does jboss tools should also be changed +to the +I think its pretty much finished now but maybe you have some smart ideascheers +The issue is fixed indeed +since it introduced overhead to maintain confliction cases that we need to handle in resolveConflict +Opened new issue for that +Here is a patch that adds a test case and applies the fix suggested by Travis +Yes thanks +Supplied patch seems to work nicely +also the mutation null line is dead code +BTW we use logj with DEBUG so we should normally have seen the error. +Otherwise I will run it under Unix tomorrowPS If you want him to recheck on his issue pls change the owner +Im cleaning it up as I use it to test Also I was going to add support to the algorithm format for setting max field length +Patch for trunk with the fixed Lucene trunk javadocs URLLocally ant javadoc under solr succeedsCommitting shortly +It is is a known bug in WTP and in WTP this option does not exist anymore thus you should just use Run As Run in Server +Updates MOTD so there is a out of the box working example ready to go +The upstream manifest file srcmain does not contain a terminating newline which breaks the signing script +Will do that before I commit +If we use a custom wrapper class to fix the equality problem then we can make them complain about nulls too +We may want to look more carefully into it before rushing into a fix +Here is the new patch for the same functionality with some improvements after discussing it with Pranay +This work has been idle since Jun Comments say trunk doesnt need but no test case so cant verify without effort +Can you split the patch and upload the MR test changes to? Thanks +See Junit test in to reproduce the issueYou will need to adjust the +Do you want to address the openclose throttling while I redo the monitoring part? Let me know +video from something we can fixcontrol +on v +just to set Release Notes flag +thanks robert! +you need to put the war in deploy of portal not in +indexing as UTC will be a little more efficient than with local offset etcIf you use a and store the long epoch in it it is UTC as no localization involved +Using in and as as variable names also causes somewhat confusing error messages! usrbinenv groovydef as println asresults in startup failed +Finally we configure it to are so impressed to Spring Frameworks extensibility and your kind advice +This merged into master +Hi LahiruPlease review and apply the patchThanks in feefdbbdedfcdbff +A trivial patch included in the code in current trunk +However you are the defect orginator so it would be best you you verified that too +I have removed the with a compile time error saying that the multiple assignment isnt currently supported +Resubmitting an updated patch which also addresses Umas comments above +theres javadoc a basic integration test in and an implementation in resolving as fixed +agreed that this sucks +Rob are you able to reproduce the NPE from the file I posted earlierI managed to get the test to fail finally +TedWhere the log file I can get +Please verify +Wait until we get the OK from the Infinispan folks +Those fragments were applied and committed under and. +Handles EOFE and the IOE that was mentioned by stack +Its ve looked into the early access cvs +PATH before calling mvn +Lianhui Wangyes this must be optimizer problem +noone working on this currently +spec The SELECTPRIV column has value Y if the SELECT permission was granted or N if the select permission was not granted +I forgot about this +Verified by Ruth. +See +This is a current source code of hama +I saw that several times +Jasondid you get a chanse to try this out? It seems to work fine for me and I plan to pop it in the trunk in a few days +bq +Cloned from with Description modified to contain some snippets from the discussion including a stack trace of the NPE in questionProbably not a super high priority issue but a bug none the less +How do you tell if there is a problem with itAnother valid question +Although is upgraded to use and test with the actual change does not affect core. +To be more clearIn Eclipse click right in the project explorer ask for new ask for other ask for Java Project from Existing Ant BuildfileTry this with a project that uses ivy. +Transactional annotations on methods are ignored by Spring using the annotations on types define default transaction semantics for all public methods in the type annotations are permitted on any methods in addition and override the specification if given. +You still need to get a file manager for your stuff but you can pick from gallery and audio player which is what you want in most cases +etc +Change the Apply Features From Category form at the bottom of the screen to also use the to apply all the selected features in one hitThanksI vote for +Now its documented on this JIRA +Command line with error message. +Thomas can you please open a jira with the issues that were pointed out here? inconsistency can be addressed along with unit tests +Clarification Im not proposing elimination of the ability for users to use logical uris only providing supplementary logic +Is there some reason why you made the socket address instead of the original private +Otherwise we should consider deprecating the current methods and then introduce new ones +The description is also empty +Ah man +Asokan +committed to trunk and branch +Please correct it. +Hows this +It should work +Added standard messages +But I would really like to come to a conclusion that satisfies the main use cases +Move all of the above to a perf sub + applicationI hava tried for Joe Bohns sugesstion it works well +Great idea +Thoughts +Can you regenerate the patch +And if does a log should? What info is useful to log in the master at steady state +As soon as we have UUID thats the one to use +looks good will commit if the tests +Logged In YES useridHas been fixed a while back +Dont you have anything extra to be compiled +Nevertheless if there is an emerging consensus on this issue then I will not be standing on your way +will close theconnection ONLY if the server erroneously sends content body with responsewhich is a gross violation of the HTTP spec thus rendering the underlyingconnection unreliableAs you say the response does not have a content body +Wed better use same way do solve them +Is it worth excluding attached files that dont have an ASF license granted tooOr if theyre patches adding for that reasonAlso rather than checking for other file types couldnt we instead check that somethings indeed a patch file and ignore all attachments? The unix file command does a good job of identifying patches as RCSCVS diff output text. + +Fixed spelling as well +Lets fix this in a +An issue was raised that may or may not require an RC be produced +Giving me negative width in latest version works in but image is different than the pdf version +Any reason why these methods are designed static methods +Wondering if you had a chance to cut? +Closed again now properly recorded against Fix Version as well +This is fine and should be left in place for backwards compatibility but an additional alias should also be added named enumeration +And add unit tests documentation et allThen when its mature and ready then we can take a look at donating it to ASF for inclusion in Apache Camel +Marking closed. +Nigel if you are not familiar with Maven and how to build Castor and put it into your local Maven repo I will provide detailed instructions on how to go about this. +more test case for this issue Sending srctestcomopensymphonyxworkconfigprovidersSending srctestcomopensymphonyxworkconfigprovidersTransmitting file dataCommitted revision need to put this into xwork as well +srcmainresources line dont have an Oracle instance to test a patch with +Verified with JBT. +Len I have created to handle the build side of this equation to have it taken care of for the next milestone +ok just received one from dev that took about minutes +Take for example the abdera project +I can reproduce the problem now +Ask the developers of geronimo why they dont like maven +Also? Oy +Patch is updated to trunk and incorporates review comments +I concur with you PDF if saved as text in Acrobat does not display fields +patch looks good go ahead! good to see you around +Diagnostics are also gotten from directly which gets it from the AMRM +Would need to be careful of performance regressions +Id assign this task to myself if only I could find out how to do that +Limit call queue length +Thomas this issue to yourself due to your kind offer +We believe this is a bug and should be fixedThanks has been added with description patch of proposed a grayscale color conversion utility that is used when setting the color value on the GOCA stream whilst rendering SVG to AFP in grayscaleWARNING This patch will depend upon an update to that includes the patch defined in Attachment of has been added with description patch of proposed has been added with description patch of proposed patchMoved some files to xmlgraphics for the patch Peter +Provide the patch +Updated the patch with new directory structure +I understand that spam is an issue but there needs to be an easier way of fixing this when nothing else is obvious about the spam +Thanks Sohan +Assigning back to Barry as this is a Teiid Designer Guide re not sure if this should be exposed as an eventual product doc +This bug should have gone +Firmly convinced that trying to load tmp is misguided +Need to verify if this is fix in trunk since were using of release manager +Attaching a proposal +status file needs status file in r should be syncd in a few hours +Should get rid of the socket timeout +And yes this only happens with Enums not +Ill add a javadoc comment to to this effectDone in +BTW these VPC test really consume a lot of resource +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +The patch uses assert +Heres a simple patch that removes the max clauses limitation from SolrI considered going the deprecation route but in this instance I believe most users would be best served by outright removal of this trap since maxClauses was set by default in the example +Its not completely obvious what the boolean return value is supposed to mean +Wont be fixing this in +It looks like Sami was storing the electionContexts by core +Ive applied the patch for +Into trunk it to M +For reference in the attached is the patch to just the problems described in this issue +Is that intentional? Its a bit jarringYup intended to highlight it +Thank Ralf it could be very great if you find a solution +I will need to spend more time on this +In top level transformation queries with literal project columns typed as biginteger float short etc +According to the JSF RI bug the underling issue has been fixed and should be in JSF should be verified when that version of JSF RI is with listed version of JSF the issue we should review all components and add autocompleteoff to affected hidden inputs +Im going to mark this JIRA as invalid +The cause for each failure was a timeout +Having cassandra classes that duplicates the avrothrift generated classes is duplication that wed end up paying for with increased serialization costs +Committed revision backported to branch for +Attached a plain source of the test also +Only remove CQL after the CLI no longer depends on it +Thanks very much Yip for the reviewIve committed the patch to subversion as revision +Committed to. +Hey ChrisOnce again thank you very much for the effort +Currently it sets a property in +Ive backed out the upgrade until I get the repository situation sorted +I was looking at the which doesnt actually use that ensure the client doesnt lie to me about the length I now check the number of bytes copied from the stream to the length given by the client +However the important thing is that the debian directory should not be shipped with the release tarball but in a separate diffSo does that mean we should not commit this patch? Or should we add the patch as a standalone file and commit that? Or should we do something else +Created a map nnc which has as key and as value +Tested with cvs and can you test it with your cvsnt? a new snapshot is migrated to in new code of changelog plugin and actually only native cvs provider is full implementedtested +Based on the last plugin version I anticipate this fix will be in the splashscreen plugin +I will use that to wire up the API flow for custom command +just about to add the Add context menu to the Outline need Delete on the editor pull down menu the Outline context menu +HM server now supports login Gumtree +If the cancel action of an event happened between the event dequeued from the protectQueue and inserted into the how to free it as soon as possibleshould we check the cancel flag before put it into the +Noted as incompatible in +The only thing I did a little differently was I added an extra test case for testing that NN is capable of rolling the edits of NN rather than triggering a failover in the same test case +Hi Davidmy testcase is still failing with the same error +Screenshot shows nicely the patchy rivers at a smaller scale +hehe guess that was only done in H +Though having themselves as other members will compromise security +Suggestions are welcome +It should definitely be used as it represents a way for developers to control how many records are allowed to be skipped within their individual batch job while still allowing for one way of skipping to be declared in the execution environment +At the very least I think an option should be added to the configuration that allows parameter handlers to override constructorsstatic methods +Ill post the details on when I understand it better. +This has gone stale +I dont see any patches +Or the other classes that derive fromThe Parameterized and only needed to be in and all the inheriting classes just rely on the base class to collect the different localesAnyway this one should be much better +Mmm there are conflicts will see later +I looked through the server start code thinking that it had done a quick recovery I saw it show soon after lease expiration in the master log thinking that server start up removed logs but thats not it +If this overhead were discussing is a simple conditional check then this is since as it is I need these null checks on my end of the API due to my FST being potentially null. +Hi Johngood catch the object is only kept internally but as you say this would prevent it from being thus we should change as you propose +Patch v addresses review comments by introducing enum Also added more comments for the new test +A new cartridge was created using the latest Cartridge Agent implementation +I reopen this issue as I still have some problems in production with that +I committed the patch +I included unit tests for new cast classes and did testing for all the different casts in vectorized mode +Thanks Kim +If one of them gets removed the PERSISTENTBUNDLESHANDLER goes down which FRAMEWORKINIT depends on +If not JSON +Also tested on AS nightly build +I really dont see why it is so difficult to manage the logic yourself and throw an exception to force a rollback +We have an internal patch +In some instances the target webservice may return a response which is not valid within the invoking context +I might have a go with it +Help in resolution is appreciated +Sorry about that +Sorry about that Kathey the last rev +Id just remove the blank throws Exception that got added to a javadoc statementPS Ill let you do the honors of committing as and when you have access so as not to take away the fun +That said both approaches would help us out over what currently exists +Attachment has been added with description Patch to update license Ive applied the patchI wonder what tripped up the script on these files? +So cores could be sharing the same object but then the schema changes and we create a new one +StackIts better i handle it in testcase +Ok +For now I think we can do without but with automatic load balancing coordination issues will become more complexIm going to have a look at this pending range issue tomorrow +Take a look at the schema and give it a shot +Resolving. +So right now EDG should be configured with this in mind and to that extend we should play with and elements in in order to allow retries +Thanks to all who contributed to this issue +Some code would be nice +Theres a lot more to this than just making an HTTP client and it doesnt require using Nodecouchjs implements the view server on an embedded +Please apply patch at root level of airavata source directoryRegards Thanks for the contribution I was just reviewing this patch it looks good and now aligns well with modules Saminda suggested +Adding JAVAOPTS in etcsysconfigtomcat works correctly adding it to Dave +m sorry I just had a look at rc and noted that it will not work as expected +Not yet ready to commit +If no manifest is specified a default will be created for the assembly +Fix has been checked in and integration tested with modjk +The note will be beside of reason field in Kryten to direct users this way +Shouldnt Symbol spill be in ARQWhat should go in +AliIt doesnt work for me even after a log out ctrl F +trunk rev branch x rev Thanks David and Peter +Joe I am not seeing any ICLA on file for David +Only the problem appears possible in trunk +And if you do decide to add the new property sure I can make the full regards ShinobuThe feature of choosing the group name is already implemented in the plugin +where are my notification mails +Verified with JBT. +File +will fix this +Our supported version of jsf is JSF RI of a bulk update all resolutions changed to done please review history to original resolution type +Forgot to add that I ran into this while working on +probably need to augment testWrites to do a get and check the the created file has the right group +Verified on JBDS. +Anyway I attach the file for completeness +Firefox works fine the from Jan it seens to be better but i think the problem still in advance please vote +lgtm +My test was wrong this problem seems to be in place even in. +Is it ok to just abandon the index on A or should we recognize the removal and clean it up +attemp fails with TOOMANYFETCHFAILURESThe job will effectively hang since a new task attempt isnt startedTom this is another case which leads to the same situation of a stuck job +but on the other hand its not that difficult we just have to find the time to do it and create that extra convenience +The NIO connector has been removed thats it +Without more information we will have to resolve your issue as Cannot Reproduce +Fixed for but occurs on and fixed in snapshot tree +Philip please see for the issue about the empty directory case +A solution of this problem can be found in bug which unfortunately has not been checked in yet +I have not had a chance to check this out on my end yet but it does make sense to me and if it resolves the issue then well its no longer an issueI didnt know about registering for the event in that fashion could you possibly point me towards any documentation on this or is it just something to knowRegardsMatt is something I know because I tried something similar long time ago +Users have choices to choose a REST service provider from those two different implementations +Attached patch will be in the code drop +With this now closed it would be great to revisit to allow compressed Avro output. +Ive committed it to master and. +If those semantics dont work for you then youll need to use some other mechanism not implementing +Removing also breaks the testsThanks for the heads up +I did a quick review of the licensing situation +Any chance this could be executed after running the tests by mvn install + +After adding inner class. +Len any ideas on these? These are not tests Im familiar with +Pull request client version contains a fix where if the user configuration passes a username or a callbackhandler class then the SASLPROPERTY will not be added by the EJB client library +Makes sense to provide this as an option +changes were reverted by dgolovin so have to and revisit +passes all tests as refactored +No I havent tested it +Is this or? Ive got a Ubuntu box and dont get any segfault on that +reopen as have active request for being able to build on linux +Exceptions are properly caught and displayed as per normal when thrown as +Fixed in deeaddcaadbcf thanks for the patchSorry about the delay. +I also dont think its using any threading +Now with reverted this no longer applies +You seem to assume that applications execute gets without setting watches for znodes they access frequently +User did not respond and the patch looks to be an incorrect approach to whatever bug is supposed to be fixed. +submitted in r +I will get back to you later this week after I have tested your changes properly on Resin +get does not include offset +Alternatively we could also set the thread context Were doing that in our for Hibernate and it seems that Hibernate picks up the thread context by default as +It appears to have been suggest this be closed unless someone else can reproduce this bug +in a few testSeems this issue was fixed already +Looks like the same H brain damage that caused +change applied +Recoverable is a good example of where the framework can provide an obvious strategy for state management instead of using instance variables internally +Downgrading cos it is now just a question of sensible defaulting. +My derbyall run was cleanso I committed to subversion as revision +The only legitimate reason I can see for thinking this is ok is that cryptic error messages make the product harder to use. +Looks okay to me +fixes this is the updated patch which fixes also the assertion problem +Thanks Mahadev +Tom pointed out that with the current patch yarn jar will not show the necessary INFO messages. +see comment +bq +i had to teach myself a few new skills but im finally getting better oriented in the Camel ecosystem +If the stream is really closed instead of NPE we will get some instead +Returns what you expect +I added two enums for tunnelled and layered +pushed +Ive now read your LIA contribution and will implement the optional XSLing and caching as you suggest +Theres no provision for defining any other encodingIf you want to use a file format with some other encoding it is possible to use XML as a source of Message Resource Bundles +Pls share your thoughts of the over all idea + +Transitioning to waiting for update on this? It becomes more and more important and we notice more and more volunteers who want help us with translations +Is there a ticket already for this +By having them static you are allowing sample information to persistbetween invocations of Cocoon +So yes Im seeing what Ted reports above +send via email if needed since I have a project that is at a stop because of this issue +Simple task of instantiating the yarn client am secret manager in the rpc server and initializing its master adding the client side changes. +Fixed an issue Daniel found with the new Util function used +I will follow your suggestions and create a patch for the patches for trunk and +This patch fixes two bugs mentioned aboove and adds readOutput method +also since it was default in the change effectively lowered the region split max value so it probably made the number of splits go up significantly +Just guessing as i do not know much about +So this issue could be closedI have reproduced the issue with not validated that are pasted into Java Editor is created for the issue +Jeanne please review this patch svn Trinidad trunkFor Piyush Hari +proxy object take the role for actual at NN RPC server +I was using +Thanks for the review mbedward +Copyright field still not yet +You can then reference the jars from the regular +Thanks very much for the patch Fernando +I didnt add any new tests yet but I fixed up the code as Christopher described and changed the behavior of creating and deleting namespaces as Keith suggested +Do we still need to send a partial list of blocks from the in the? This was the reason we did it in the first place right? So may be now we can send all blocks at once while controls how fast it removes them +Im not sure if there are any other table properties that need the same treatment applied nor have I run the unit tests to see if anything broke +integration to pushpull was done in + +patch attached for future reference contains Michaels fix new test in that exposes the bug some debugging improvements in +It happens when a node is trying to tell a dead node to send more messages over +Im happy to commit this since it looks fine suits your purposes as an active user and remains reusable +In addition we made a lot of changes to codegen since the last recreation of mapping source +We had all our releases up to version in the Maven repository +Verified +I moved the svn treeI believe this was the last thing for which infra help was needed so Im resolving this as Fixed +Includes additional method in to replicate the bug and the change to to fix the problem +For bundles this means you should probably create file names based on its symbolic name version for other types of artifacts you should use some kind of unique name some kind of version numberThat does not invalidate your comment about the cache not using weak references though so thats something we need to look into +Great +Anyway Rajesh has explained with more details +Basically the family fieldname and we need to deal with when the family is null +Checked in a fix +Applied +This was a temporary connectivity issue +BTW what version of Tools are you using +Currently the additional empty folder in the directory is now gone but there is still the empty antlr folder hierarchy at the root and an unneeded maven folder with contents within the directory. +What with all the link rewriting logic and the ability to decorate the service you could probably fake this starting in +We are not doing anything special to columns that are typed as time so either the native type is wrong or well have to KIworkaround the native issue +Or more precisely the tests are uncommented for the embedded case +But those were using a normal LAN connection so it was important to encode fast even if the image returned was btw in the compression is the opposite of the value one passes down to the encoder saying means the encoder works with and so on +Thanks for the excellent test case +hope thats gonna be soon cause bubbling ve just created a new JIRA for the complete replacement of the bubbling tooltip with a standard YUI one. +a patch with a couple of unit any feedback on this +The option is already there as a TRACE level message +Committed revision i went ahead and fixed this using the display the bucket approach +actually fixed in revision . +Thanks +If they are used only for the lifecycle of reading a single block id say size it to the compressed Size +Raised to critical as it seems to be a important security VM control is allowed without credentials +Adding in a first patch for this issue +Was wondering would only a dba have permissions to invalidate the prepared statements +We just need to add a meta info for schemas and since it is not big one I believe Slava can do it for couple days + +Trying to avoid database hits lead to the problem in the first place +Attaching fix along with a unit testProblem was we do some paging on the index using as the page size +redoing the link with relates to instead of blocks + +The intension is to let Hama have ideas of current system related statusI agree what Thomas mentioned that Hama needs to have its own metrics package because IMO Hama and Hadoops are not completely the same + +The only difference is version numbers for both hive and hbase +A quick update as to where Im atOver the past two weeks I Tycho and Maven have not been friends +changes checked references have been replaced by as we were already using +Ive verified that it passes existing unit tests and also that it does indeed use a single mapper on a cluster +Looks good to me +Thanks for your help I am waiting for your solution! On the other hand when do you think the next version will be released +The real problem is axis does not have the message qname and it keeps only the message name +I will try and review this over the next few days +to the move with some caveats hudson has to move up first to make sure it is happy backporting becomes tricky and that includes toStuff built with Java doesnt necessarily load in Java Ive encountered problems with libraries I built myself not loading in Java +Yeah lets see how far we can get to reduce the core dependency and also offer a client JAR with minimimal support for transports such as JMS +No that looks correct +Changed version attribute of schema element to +The error messages are the same which were shown aboveI run version on sun JDK +I dont know how hard it will be but it might be easy +Bug is in LIL for x +Whats the redirect forI dont know how get differs between application servers so this will need to be tested on multiple servers to ensure we dont regress +So this is not an issue at all +I think this ticket is doable and I especially like the ability to provide etagsHowever I think it will require a change to the database file format so if this is going to happen it should happen forThis would be a patch to pull off so if someone has done some internals work and wants to take it to the next level this is a good opportunityYoull want to see couchdbdocflushbinaries for one place the internal attachment metadata is manipulated +there is one to one between attempt and containercompleted is currently completed map attempts percentage +Thank Enis can you please comment on what type of issues you consider as over and above what this patch addresses +Im not quiet sure but you could perhaps set the thread context class loader to the class loader you need +Hi JefSorry to reply to this a bit late +I wonder if it might be useful to add copy throttle support to the replication +updated with Toms feedback +URI cannot be subclassed and we want to add behavior so we wrap our in Paths +for names of hidden states which I didnt want to serializedeserialize so I wrote lossy serializer to be honest since I dont serialize all information from the model. +See screenshots for example +Also as you noted this only works for DFS and not for all the +Im inclined to state that we dont want to support connector bindings from designer +My point was does the failed merge transaction restart again on master restartBecause the resubmit queue is going to be empty right? Should we forcefully read the merge znode and completed the failed merge as do for SPLITThanks for your answers +The current implementation doesnt support the fields though +Lukas can you confirm before I close this issue again? it works with latest snapshots +Sorry here is the patch for applied in r +success +Allen Im sorry I missed this ticketAs we briefly spoke over IM previously the CS in is designed to not allow a single user to go over the natural limit of the queue +Applied for now to prevent CI builds from hanging +According to review I made two patchesPlease check Thanks +bulk close of all resolved issues. +Hi Morten even if I can understand your point of view this requirement is definitely not a bug but an improvement +Attached +Given that NFS itself already has built in retries and can be configured to arbitrary timeouts it doesnt seem like we should worry about short hiccups any outage that makes it past the configured NFS retrytimeouts is likely to be worth causing a failover IMO +What do we do when a number is so big or so small itll be encoded in scientific form that is +of course is still is more of what he is talking about. +A few comments I think launch would better reflect what the new method is doing +Sorry guys +NamitI thought we were getting rid of negative tests +just committed this. +Will work with Tom to make sure we get the combinations right and update the tests accordingly +Accidental duplicate opened because of Jira system problems +Hi Do you want to add this to release +on commit thenIm happy happy to do that unless you guys would like to +Someone working with this? +the command names are already updated in the is not yet renamed +Committed to. +Once you create the patch files please verify your detailsMany patch created using git + +Going forward all the vectorization work will happen on trunk and vectorization branch will be obsolete +Just the repo +The first two properties are for and dont think theyre important +The fix that is commited already also handles this scenario. +The test runs successfullyI will submit the change in with the fix +If we decide to go down that approach we may be able to subclass the log appender used by that framework and override some methods +Since this patch gone stale I rebased your patch +Doesnt seem to be an issue anymore cannot reproduce the error or situation +is created by the +Use cases make sense +But since we always delete the tomcat tarball after untar it it has to get downloaded againAttached is the patch that removes the delete task from the dist profile +I am digging into this right now and will use this issue to commit the test and fix +Ive edited the commit message so that it has the correct issue number +the reason we dont bundle help is that it would add about mb of extra downloadWe are in talks with doc team to see if there is some remote help we can includereference +To SnjezanaThanks I modified the file as you said the file can work now +RM ends up removing references to the container +Derbyall and ran cleanly +Something screwed up Eclipse completely +I might be wrong but the major difference between FB cache and bucket cache is the former caches blocks encoded and compressed +This feature has been implemented inPete if its still doesnt work then please reopen it as a bug with step to reproduce the issue +The featuresuninstall xxx should automatically select the version of the installed features else throw an error +Currently Jive provides only github UsernameThese new profile fields would be neededGoogle usernameTwitter usernameFacebook usernameLinkedin username not needed because doesnt return username during social profile fieldsGoogle usernameTwitter usernameFacebook usernamehas been created in PRODSTGDEV +And there was a bug related to thatThat improper support isnt ideal but it may have worked for a lot of use cases +No +BTW you also have a point that the seems to indicate that in your case the flow id would be +Setting fix version to next release +What I am building is indeed state machine for HA +I like the idea of adding the ability to specify the collection this should be optional in case the user simply wants to make use of thedefault +bq +could you attach a project with a +The current behavior is consistent in that you will receive the same view on each refresh which is expected from current redirectrefresh metaphor +Is it reproduced always or only sometimes +Bobby I too have seen in large clustersjobs the law of large numbers We dont see the random number generator +Why would we not implement the ordering as a linked list +eugene enter a doc jira for your suggestion as this issue is already closed. +In your case the value of the parameter format is given as a RTF +There is an existing Derby entry to correlate the client and server connections +However they were able to reproduce it even after priming the default page +Attachment has been added with description Concurrent map based implementation of property cache supporting hashCode has been added with description propery cache thanks for looking over it I agree that TODO shouldnt have gotten throughThanks again for taking the time to check through it +New patch capitalize instead of Fortests The magic numbers and are pretty messy and hard to follow +If you can get those things we can bring it into hive contrib +We were able to produce this scenario while testing +Lets go with your suggestion and see how it goes we can always adjust in the future +We will keep this issue open to track the idea of optimizing that model to minimize iterations. +The pessimistic write lock can still be a potential problem for other databases where select for update is only effective in certain isolation levels +Please check that the patch was applied as you expected +Sorry for the late i was busy with other tasksThanksUma +Thanks Tim +applied the patch +How do we involve a wide group of committers given that? Suggestions welcomeA lot of times people dont weigh in until something is about to be committed this never popped up on my radar beforeEssentially either my comment will pop up some other comments from other committers or lazy consensus will take hold +We do need design for it +Attaching patch for solr could be changed in similar way +Can you try again with the current BranchAS from github please? +Are you sure youre following all of the steps here +if not fit final change link type +I may be crazy to harm on this but still reads as very generic to me +I have reviewed Christines patch and believe that it correctly resolves the problem +Fixed all seds checked updated +Resolving for now +Verified in JBDS to write release note issue now the release note text has been written +yeah this is not fixed yet if someone could create a pull request I would like take a look and merge it created a pull request for master and dont see the patch wouldnt work however the next release is we really dont add new feature to a minor release so I will leave this to the next major release like thanks for the request merged into master think its great that this issue is getting fixed however is there any way you would roll this change into? Im not sure when will get released and you mentioned above it should wait for a more major release like +After adding a file theres no need to manually synchronize some indexpage but that is done automatically by rebuilding the site +The Ant build script is provided as a convenience for the user only and validated for release builds only +instead of the file written to by the main test processI created the test reports by running EMMA manually on the generated ec files when I ran these experiments not by using the or ant targets +in constructor with passed zk instance you assumed that the passed zk instance is already connected +Thanks for the fix. +gen will issue warning messages when type merging that extends features happensCDE will pop up an OKCancel dialog if this happens prior to calling +Problems is I think there are so many people out there have the same needs like me they will have to do the same work like me right? Even myself have to do the same work in a different project like copy the classes I write before to our new porject or something like that +And by deprecated we really mean deleted +How about The data files stored in HDFS are from an earlier version of +The Detector contract specifies that detectors should accept null streams so even if a can do nothing useful in such cases its still best to degrade gracefully to avoid unexpected problems when combining or decorating detectors from various different sources +Removed the EXPLAIN query that causes the problem +I think I can contribute a plugin like thisWould it make sense to implement it as part of the bundle from or a separate bundleIf a separate bundle then would need to be registered as a service +Ill take it out from the dropdown +TomaszSame error in Application Version version JDK +By the time its youre already in trouble under heavy load + to +This didnt seem to get any takers or a patch +plugins +Hi Leonardo thanks very much for the comments +What specifically do you think is wrong with this XML +After that the follower will receive a broken pipe exception and it will retry after that +Special mention to this issue as it was our longest standing issue +Confirmed Matthiass observation at high concurrencies +Karthik +I just wonder why it works for admin and not for flexadmin have you an idea +Hi Dan Im not sure I understand the issue +Thanks for the feedback everyone +If you wait for finishing of server starting and refresh browser then it works +Will take of this via a delegating ctorA process question I guess I should wait for more comments from other reviewers before I create another patch in case if others are reviewing the current patch +Verified with JBT. +Thanks for the patch! +The seek in old stuff was an actual stream seekNew stuff should have difficultly timing out though I suppose it could happenId suggest we close this issue because so much has changed since its been reported +That is great! You dont got the! I will check this after lunch. +Check to make sure the address in the wsdl is correct and the service is actually running at that URLThis is just a normal when the service URL is wrong. +You need only change the versions of plugins in the and between maven and maven +Nevertheless would be nice to have an example of JPAHBN working in an SMX environment +Have fun +Thanks Mark +To illustrate another side effect of this problem consider try following You have a View Action which displays all of the data of a specific type which uses the in retreive in order to show confirmation messages from an Update action +Thanks Andrew patch applied at repo revision r +in db dir theres a script to create the schema +Closing this bug as WONTFIX per Craigs comments below. +The previous spec file would remove the cassandra user and the configuration alternative when upgrading the package +Lewis after applying this patch I still see some references to in ivy and some files too +IAC should we commit this to or revert +It would be great if the process of making an index file included on the fly the creation of a solid checksum that is then baked into the file at the last moment +Applied your patch in Rev +I am concerned users will be confused between analyzerscjk analyzerscn and all of which are different +Thread dumps were taken and show On April th locks on On April th locks on On April th locks on and Cache configuration file +Go figure +Please change to if you intend to submit a fix for +modeshaperesourcesruntimedeploy +I receive enough of these supposedly helpful about failures that turn out to be transient that I tend to ignore from an admin saying Im killing this sync wouldve gotten some attention version of rsync does the synchronization happen to run? My current hypothesis is that its which isnt compatible with rssh +Attached path adds new statistics +Version of the plugin works with Grails +It was really difficult to identify since tests all ran locally but failed with maven + is fixed consider merging the change into of should be merged into AS to reduce memory consumed by logger instances in +Stu the idea was to use the passivationId internally in the Beans and not to serialize the all the fields in those beans +Im afraid this is not the right callback for your purposes then post is not expected to perform any tracking or the like in particular since it is not even operating against bean instances +We cant move to it because it breaks old snapshotsmodule versions +If there is no update Ill close the ticket by the end of the week +The priority could also be lifted +For this case I have used at the serverJAVAOPTSXmsm XmxmAnd I have used in the client threads have ran this test and discovered some errors along with will be in charge of running this from now +Backported the fix and test to with revision and the fix only to with revision +Do you think I should enumerate all cases here? It would help me in the future I think +Deleting the +Looking for review +Tests are passing +Sorry! edited a wrong issue +Ill tag this for to see if we can get better error logs +Sure done killed the overwriting your ones +but why should none actually process this? I think this should really be discussed in the EG how process none should actually work. +Alternatively instead of adding the three methods to the resolver interface we could make it optional by adding the three methods in a subinterface of resolver or a to separated interface +I encountered the same excpetion while working with bounded +This should be helpful for HSSO where we need to add a construct like which determines groups and attributes by extracting them from the wrapped identityaccess token +In default there are only the parent calls for start and stop +Its a fairly trivial patch and improves the user experience if somebody needs to configure DIH with relative basedirpathsIll commit to trunk if there are no objections could you to the release branch Steve +Files branchesBranchxconnectorsrcmainorgjbossresourcedeployersmanagement you sure this is going to fix the issue for connection factories too +and methods returns the same results for this string on RI and HarmonySo the only way to fix this bug is to fix method +First buildI flubbed but will keep an eye on it over the next few attempts +Just to make it clear +Update the according + +Currently DN threads block more often on network IO which is the core of the problem +Documented as a known issue in the and release notes +Unless performance is a problem I think the meaning of the code is more clear as it is now +Update schemaRun your app and access this class hibernate says the column is but the db returns a null value for the column +Propagated changes to all Trunk job producing files +seems ok for the reviews! If there is no objections Ill commit the v patch tomorrow afternoon +Just FYI Im have started implementing this +A last thing for this to be interesting for Buildbot tests I will need to backport it to releases +Can you please confirm if this has been fixed and which version we can download that has the fix applied? Thanks in Advance. + test and implementation +Yes and I backed it out in favour of removing all our own format and instead using the containers own set after checking the impls +Thanks Arun Pranay for the contribution Committed at rAshish Vijaywargiya +Resolving +What is an example client context that is exposing this info? The behavior of the login module itself is a feature request to use an alternate msg +ah ok I will check first thing tomorrow morning just off to a meeting now till end of playI just downloaded a brand new Eclipse Build on my macbook pro and installed JPDL Designer CR started a new jbpm project created a new process definition and straight away got the same error so this is happening out of the box effectively on you attach the Eclipse log? Make sure you first remove everything that is not related to this issue +This seems to already work as expected against. +the method transformDocument creates a new Transformer for every call +jelly also as +The failure is not related to changes in the patch +Seems not possible for now +If no objections Ill commit this in a few hours +There are at leas two classes in the standard library that does HTML escaping +The underlying behaviour is wont fix +Fix the description of +Keeping it brief is great but we lose we have to manually fix the imports +This would avoid the problem for my case +Descoping items not being worked on for M into Unknown Fix Version for Marnie can you review the documentation for me pleaseThe broker has no properties to document +has traditionally not specified error behavior either but Im on the fence whether this should be specified behavior or not +Sorry +Tests fixed. +The fix code has been added to Brach Branch and Trunk. +Jie if the fix option is not set it will not attempt to fix the problem but still will report it +It works for mePatch changes two files set portal scope as protected instead of private so that inherited classes can access to this attribute Add portal in list of roles forI hope this will help +by Alexander. +Fix requires newer version of Hudson OlivierThanks for feedback +Great HenriqueI just committed your patch! +In later releases prototype does not extend the function anymoreSo my conclusion is that prototype has a bug and we should upgrade +like adding some clause with more prompts +jar files that I have in the bin of xalan +This is the responsibility of the deployment part +Executed the cases related reset ssh key to VM feature and verified the resoved closing this feature bug. +Also would it possible for you to give a bit more info about your setup? Arquillian Core version container you are using etc +The last line before BUILD FAILED is the problem +that sounds like a nice improvement to enhance the equals operator to be able to use the Object equals method to compare objects of the exact same type +This will demonstrate the get failure +Mass closing all resolved issues that have been closed for more than one month without further comments +A servlet the should provide a very comparable end result to that of a filter mapped to at least when it comes to serving resources +I have created a sample project that replicates the original project +I think we should have a to keep connector to balancer instance then have singleton for each connect +Just to recap some of the points I made in my previous response to Mikeo A separate credentials database is only necessary for systems which already manage multiple Derby databases +My username is mheath +removes some crud that slipped into accidentally and should pass all tests +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +bulk close of resolved issues as part of JIRA housekeeping for release. +This patch fixes the indention problem and returns null if the target directory is deleted before the full listing has been fetched +lgtm +Id like to see blog posts increase our overall web traffic +You need to edit the wsdl and append the entity type +Thanks Kim +Please revive this effortThanks! +Every request needs to be decodedThis is on JDK +Added the class to hold the collector http parameters +We may need more iterations to clean it we look at backporting this one to given is a little ways off +But from a tracing point of view it would maybe be nice if it walks all the nodes without skipping any +I am seeing the problem with the latest snapshot and jdbc +Hopefully it will be synced to maven soon. +Patch for the Felix parent POM as described +we should probably bump the dtd versions for both and before release +Test project which fails when running grails + what its fixed? Its normal? I get the same problem this is a totally unacceptable behavior when fetching thousands of records + docos need to be refreshed for accuracy too +fixed conflicts +This seems to be fixed with the latest upgrade. +Fix is in SVN +I am still producing the same Could not create entity Unknown column test in field list +Ive made the attributes but Im not sure about the Lists inside these does not seem to be guarded either +Its those damned mode tests that seem to be so buggy that are failing +Some of them are marked as Stable and some are marked as Unstable +I havent tried to run this patch so this is just from looking at the code but you might want to check to make sure it runs correctly with foobar foo and barfoo foo while searching for foo as Im not sure those cases are correctly covered Maurice I was away last week and now Im sick but Im considering what you did and gonna take care of it its unfortunatly just a matter of time I think about next week end to keep you on touch and give you some feedbacks sorry for the waste of time +I committed this to trunk and branches +Closing tickets +Thanks Chris done in trunk at and in at I just modified your patch to add a test if logInfo is active +I think the test separation is good and there are two items of work required to complete this Change the harness and policy file to handle separate locations Have the buildjars target create a directory structure that mimics the layout in the distribution ie +bq +Ahh right we should performance of this after is done +Thanks Tim +Also I believe the direction the design is heading there may not even be a profile service anymore +I should have said so on the list right away sorry +New listeners are are able to also receive the text selection information and also the annotation selections +Done. +minimum testcase to reproduce to minor as this affects development on Resin only seemingly +And one month is not worth such tricks +Thank you the worked without Exception + on adding restriction to names +all issues assigned to +Committed in revision updated docs in revision +Patch attached all tests seem to pass +Thanks for getting me back on track +THAT would protect you from a wide variety of problems from network disk ram or bugs +We werent validating document ids that get pulled from URLs to be valid +Jonathan sorry for the delay +Fixed in trunk +In total about methods were implemented by this JIRA +Thanks for your reply +HowardI updated the file +I played around with this and sketched out how this could look like +I added utility methods to to do doublefloat string and vv and then did a straight replacement everywhere was used and removed. +Committed implementation of on all components including velocity template junit tests +The main reason I didnt is that I couldnt see where to fit them in +I upgraded all dependencies for maven beta +I dont really see any harm in it at all +This patch adds the start script +Attached patch +The indexing benchmarking apps I wrote take command line arguments for how many docs and how many reps +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Results are included in parts are not completed +Thanks for fixing +Resolved waiting for pull request to be accepted +Fixed in Hibernate but didnt check +No feedback from community on point so Im leaving it as it is. +classpath file confirms that the noworkspace option is added to the MAVENCLASSPATHCONTAINER about that config page makes me question the function of the Enable Workspace Resolution and Enable Nested Modules menu times in the Maven project context menu +Theres a portManager which has a counter +Arnaud has improved the documentation for this will file this for a future enhancement +This is not a bug its a request for an enhancement +It will only compile the flex apps that are configured in the enunciate configuration file so if you dont want Enunciate to compile anything then just dont configure any flex apps +Patch submitted +And I am almost done +The test is not for testing the console outputI think that it would be sufficient just start with an alternate console +Tried which is still green +tests passed patch committed to trunk and branch. +So sorry about that +Defaults for strict readwrite by languagelanguagestrictreadstrictwritecocoafalsetruecppfalsetruecsharpfalsetrueerlangtruetruehsjavafalsetrueperltruephpfalsetruepythonfalsetruerubytruetrueBlanks indicate that strict readwrite are not optional +Will submit a patch to update it later at some point. +Hi DeepuCan you please attach the used at the clients endThanks the has solved the shoudl be noted that this issue is not solved in the version Some clustering error was showingOn using axis and rampart we are able to resolve the with version +There is a added in that is not needed +depends on that is specified in the its +Maybe use a different classpath scanning algorithm since this implementation leads to additional dependencies including ASM which often leads to classloading issues especially when combining with other libraries like Spring and Hibernate +netbeans directorythanks for your wonder if it works with the build and the old userdir + +Jopr r. +Rich Heres another patch for reviewThanks the fix with revision +I will try the current version in the jbpm distribution instead +patch vs tag Fixed thanksColm. +file while validating with Xerces and it worked +Could you hold for a few days +The docs JIRA is for will be in extras and is already in tomcatserverlib +added enumlabel and enumchoicerenderer +regioninfo is a big improvement over what we currently do +We should investigate this for these to please move any bugs back that will be worked on before the November release of. +Run the expired code before removing from queue since its presence in queue is used to see if lease exists still +I would like to decide which would be appropriate to append image stream to the output based on a variety of conditions such as transparency color palette etc +Its apparent there wont be a release soon enough to get updates into so well go with what we have and try to match them for +context was not considered in when testing if a resource exists +Hi Grant thanks for your comments +This is not a bug but an enhancement thats nice to have +Work in progress +I removed the zip archive that I added because its no longer necessary all changes should now be distributed among the subtasks and the issues that this issue and the subtasks depend upon. +JaypalPlease can you check why upgrade fails at upgradeYou can use the schema attached byany update on thisthanksJayapalWhy is our management server logs not showing the SQL exception that Prachi is seeing? We see a rollback happening but not a SQL exceptionAlso why does the management server start even after upgrdae fails? We should not see the components getting loaded and starting the management server when upgrdae scripts fail correct +Commit ddccbbecdaec in committed to jacorb upgrade component jira opened to include in EAP +Anna can you run the test with trace logs enabled please +The lazy situation ALWAYS occursIs that correct +Your rational makes sense to meDo I need to make a new patch or can the commiter just remove the comments from after applying the current patch +I have fixed that issueThe search control wont be shown if the editor area is too narrow because there is not enough space to be displayedThe user has to expand the editor to dislay it +Now that work is well underway with AS all previous community releases are +However I believe I used the string to tag the commit rather than hence the commit does not appear in the history for this JIRA. +Jan is that the case here +There is also a more configurable solution under way for TCP see +So should I apply all three patches one after anotherOk the patch already included the modifications to I will review it tomorrow hopefully you will include and problem Ikumasa hopefully I will finish reviewing the current patch and get it committed by then +Done the ll do others only if they there is shown to be an issue. +I edited my previous comment +This has been resolved already. +Adding the Fix Version to +Management servers IP should be routable from private network +Has there been any activity workaround or resolution to this issue? I am experiencing the same issue with Geoserver +I may come to do it eventually but anyone wanting to do it before me is most welcome +Considering that Teiid Modeshape BPEL jBPM etc +v +I think is is fixed now please check +Pushed to as its not required for +But it just make sense to me that once I saw it there it seemed to me thats how it should work everywhere +Hm my Japanese is a little weak so Im not sure what exactly that means and what exactlydifferent spelling means in this contextGoogle says means How dopiece +test failure is unrelated +Eric Norman shared an approach on the mail list that outlines getting the user path more cleanly +Extensible Object Class The extensibleObject object class if present in an entry permits that entry to optionally hold any attribute +Aaron unknown exactly when it fell out +Suresh could you please take a look on patch? so that we can resolve this issueSorry I had not seen your comment Uma +Once that happens we will change the admin console to display the new statistics provided +What should this thread do if it gets a run time exception like this? Should that be part of the patch for this +Damn missed svn add on the new file also added license header to the new file +A conversion exists but isnt used for some didnt see the failing test but I did update the code to work with sets +The delete issue is possibly caused by your theme +But perhaps we should add an option to let admins force a compaction +What is the status on the Call for Logo +Attach the patch for fix +Im reopening this until the solution is integrated into master +weve become more dependent on zk and an the get in rather than less +Instead I moved the log statement to the class +Hi Ill be starting working on this +If moving your code into Apache is a goal you could also start the project in the apache incubator +Marking closed. +Confirmed Frameworkssection +Issue closed with release. +this is a preview patch +Have upgraded to in head +Hi Joern Do you have a decision on how we should solve this one? is also pending on And there is a high chance that I will submit some more patches for in the near future So I would like to close the older ones so that I can provide clean patchesThank youBR for the delay +As its reproducible under certain circumstances I reopen +Closed the issue as requested by Luke. +What about returning Ambari APIserver information on a GET request to api +I needed it for release plugin +The PGP keys can be found in the CVS in javaKEYS I think putting a link to them on the website would be an excellent idea +HiI think it would be better to avoid this relationship +Ooops +Are you sure there isnt another dependency bringing in these dependencies? Do you have a test project that displays the error +Thanks Jasper! I added your test case and I also added a bug fix for it +This will wake me up tonight +Version doesnt have this issue +The teamcity cert build was updated to the tag +It may be because jdbc doesnt list common as dependency but I didnt investigate a lot +Santosh it seems an unnecessary complexity to give the user the option to failcontinue +To be clear I am running against that is already shipping +Yes Bryan I agree with you +So as long as the session id matches starting from whichever recordedseq should be fine +investigate some way to make the jspc maven plugin generate fewer servlet would check whether the delay is caused by logging +Breaking change shifting to +Merged to in revision . +Applied to complete defect resolution +I committed a very small fix to get that test finally working but it seems that this issue was targeted for and wasnt committed to branch +bq +Im just wondering how many times GWA is going to have to update his patch before its applied +Does this sound fine? +Rename Slice to Working on this now +Added all the other dependencies for ideally we should do this in a maven build +Please read . +committed to svn repository at r +Committed revision +checked in to mojos sandbox +I think this is ok considering that if name is null were injecting by type so bean name will have no effect for cache purposesIm attaching our profile result based on and the modified equalshash code +Thanks for your review LeonardoThe latest patch does not require the init param +As per user tried using Hibernate cr and still getting problems in getting it work with merge and composite idI get the composite ids attributes as null +I agree with you +One possible usecase is with systems that deal with very low loads +my JVM has this in its lsofjava hadoop mem REG homehadoophdfsand that build is most definitely from branch with the aforementioned revision +If given a choice between the semantics of a reader that sees changes as they are made versus having to do the whole reopen thing Im betting most users will say duh I want to see my changes right away and choose the IR that is synced w the IW bc that is what people think is the logical thing to happen and it is how work which many devs +Since we have two readers and we didnt want the layer to be loaded by both uDig is making a choice on the presence of the tfw file alone +Some more detail the way our data is aggregated is controlled by the we currently have a situation where several types are aggregated once then that aggregated data is itself aggregated again +I still need fisheye and mailing list migrations but I think they should probably go in separate jiras. +perfect thanks +This has already been in IR +This will evenly distribute each tableThat plus the change to do on table create should probably cover distribution fairly wellI think the next step in the load balancer is a major effort to switch to something with more of a approach +The patch is attached +Pending resolution +It also appears that this problem is not restricted to mode +fixes only javadoc + +Sorry for going back and forth between the proposed solutions +Thank you the pluggable interface is exactly what I had in mind +In fact the memory given in configuration is little above GB +I dont think we use it anywhere else +Yes Ive reproduced the crash +Thanks for the patch StefanI must admit that there are so many corner cases and variations of this exact issue and I would really like to be able to reproduce each variety before committing any fixes +Weird +One more thing +Fix committed into cvs will be in CR +Thanks Alexander Vladimir the patch was applied to BTI branch at rPlease check that the patch was applied as you expected +the released version the code is not catching but an +Leo I wouldnt enable the old page +I think I found the issue +BTW Id like to reopen but I have to check the situation before and if this is really interesting for the project +OCD alert attached fixes the style violations I noticed +The servlet works outside of faces so our normal exception handling code doesnt work. +Committed patch a to trunk with revision Backported to with revision Updated fix versions and fixed typo in the summaryThe addition of the test concludes the work on this issue +RichWhich part of the MIME spec actually defines as a continuation character? +Hi Steffen +This has to be to somebody from the doc team +Provide additional information on the like username and uri +It doesnt seem to break looks like a minor issue of WTP +unit on the dev list that this wont be fixed. +javadoc +Thanks for the tip though! Let me see if I can get it to work. +Yep +As much as I like the Comprehensive in the name of this issue its rather difficult to track when its done +I wouldnt attach any great priority to this especially if the solution has severe performance problems +So I think the issue is about +same error with Eclipse RCProblems occurred when invoking code from +Since its already hosted on sourceforge I think Ill leave it there for the time beingI should figure out how to mavenize it though +This has been fixed in r +This works in as well as in but is broken in +I think it is OK to commit this one +Ok changed it back +A lot of renames and refactorings +Please note that this also happens when times out +Thanks Xuefu +I found out the root cause the hostname change is greenTed could you please commit it? Thanks +Rebased for trunk +Most likely it should be raising some sort of marshaling exception or similar in this case. +Ill leave the ticket open for a week or so and left in this state +I will change the map to store Container directly +The commit was missing the two newly added files and they didnt have a license +Fixed in trunk and in tag +How about renaming to +bulk defer to defer to +Its critical as otherwise the discriminator column is never set to its correct value +If it is okay to publicly share please attach you test data schemas and unit tests to this JIRA +we have a project where we wanted the user to enter their height and weight +Need a decision from Vijay +Ugh +Still after uncommenting the same failure +defer all issues to +But if you could not reproduce it on windowze i think we can close this +Here are the tests that the deadlock issue +Trivial patch to fix the reported issue +I think that the problem is that fields are not enclosed in double quotes when creating the schema +I propose that we backport this to? This is a very simple and row risk fix and helps applications that depend on the correct offset being recorded inside Checksumexception +Added more tests and strict error handling +Using Spring This is a serious production issue as data is being polluted in different sessions +I wonder if thats the connectionCan you try with invoking nodetool flush against A instead of doing a bootstrap +I also changed the refactoring slightly +Therefore the eclipse service plugin fails to start +Just need to enable the ability to upload binary licences into Xircles +Thanks for looking at it anyway +as a contrib package +the patch just provides a partial view context using the factory finder +really good to see this +should beWhen you encrypt an existing unencrypted database you can specify whether the database should be encrypted using a boot password or an encryption key +I just committed this. +Works well JulienIm reopening the ticket just because we havent decided yet the final names for the properties +I also expect that the two patches will have similar performance characteristics +A directory url has to end in a to signify this as documented in the attribute comment. +My thinking here is that its fine if one committer does some minor fixup or adds test cases to a patch that another authoredWhen the change is minor I think the committer should not be counted as one of the authors +tests included +Attached patch that implements the new fields and convenience methods for node thanks to Randall who implemented the CND parser support for the new node and property options a few months ago and then waited patiently for me to catch up +Assume the situation that communication doesnt include any data should be secured +This will let people set resource limits environment variables etc +Sorry to chime in late but I dont think you should shy away from removing that line. +In the current is used by the test target and the ant task replaces out the tokens for the deployed tokens this isnt easiest thing to understand IMO so we should change it +Building the existing source should require you try again and let me know if you are still hitting the issue +I dont think its invalid for an AM to exit without launching a containerSandy I am this jira so that we can fix the test +Adrian I had a long conversation with Marc on the forums where I proposed many possible tricks you can apply to minimize the problem and he kindly posted a recap of new ideas on the mailing list +Maybe we should just simply drop it here and care to have it recreated in the register method +My bad +Wrong test was failing because of I have applied in my local branch +One would have to think this through first +Forgot to close this home is working fabulously users are happy and I have absolute no complaints! +Niklas thanks for the test project +We should change method name for the component or in the component documentation +Applied Alexey Pretrenkos patch to update the copyrights in the core modules +I think it does that for HADOOPHOME now +passivateObject fix committed +I have not yet determined where exactly in the code this occurs +My network is isolated and cannot access the update site +push out to +Please close this issue without applying the patch +Ok +We dont want code in core thats not both widely usable and actually used +When I first wrote the code I wrote them outside the geotools code so altering the existing classes wasnt possible and subclassing didnt work +This flag should be added in the Query in order to make it available to the and ultimately to the is another andrea rendering optimization Jesse have you done this alreadySince it is an optimization I have knocked it back to minor as it does not prevent function only performance +bq +Please note that I added the CORS filter without updating to your new code and my issue went away +Rename action has been removed +Why didnt the compiler catch these thingsNot sure I follow +Patch applied +targetclassesunder the directory +What problem are you having that needs solving +Then maybe reopen it and real fix for so it wont be forgotten? +Now that its showing up in hopefully hell look over it for us +dont know how i could help but ill try to find a wayinstalled jruby through rvm +I just updated the wiki git diff git diff and svn diff are all supportedAwesome! thanks again for all this work +And it is even no longer confused by running under cygwin +Thanks very much for yourexplanation of the issues +Replaced file contents with those of the +If so then a patch is needed to remove this from the current codebase because its in there +Adding a patch that adds a section for the manuals to the page +I think there are a couple issues to address first at least in the lucene partCan we please not have the ObjectObject stuff in? This is wrong already has a generic type so I dont understand the need to discard type safetyThe unicode conversion for StringStringVAL is incorrect and should not exist despite the name these types can be any bytes +How about just giving any final review comments and let Amita make updates and submit another final patch herself if necessary +fix has made it to trunk branch and issue has been resolved for over a year with no further movement +the breakpoint in ejb jar was submitted but the war file one was not +now try and figure this out in a hand written javascript application +I disabled the other Mono unit tests until I get a chance to test out the build on Linux +As per Davids comment +However since whats passed in there is the servlet containers raw I cant see how those parameters disappear from that Map +also this happens in grails i run the controller view in the browser i get a correct response which way should i be testing this in my controllers? unit or to run this via integration tests to get the response working +Its +and then Ill commit the whole shebang + +Im marking this as cannot reproduce because its only happened once and Ive run this test a bunch of times. +We have to let Jason have a view on thisHe use in other products than combined withBut yes Im in favor of having throwing exceptionsIsnt most of these comming from the OGNL conversion? +For the record and to establish authorship and transfer of rights to ASF Ive attached which contains the changes that Brian checked in for me on Feb +I am going to do more testing including setting up pooling. +Should we bypass the fifo queue and deliver OOB directly to +trunk +Still waiting to hear from you on this one +Im not sure if its correct but it works. +Not sure what really happen but Ive seen in the log that the CM mean was being registered twice which obviously lead to this exception +Does that mean I checkin the code +Chuan thank you for the contribution. +In response to Otis I dont think each word should be returned individually +Im not sure about the testcases included in the patch they didnt merge cleanly for me +Are there any other changes needed in the Work Effort component that you are aware of +Add new attributes to the tabbedpanel tagbefore topics published when use click on the tab the tab widget and a cancel object are passed as arguments setting true will prevent the tab from being selectedafter topics published when use click on the tab after the tab is selected the tab widget is passed as argument +from the logs I saw that some works are applied to the index by a thread which shouldnt normally write to the index the thread which is doing this work is the caller to the Executor so it means the queue filled up as the strategy applied for is caller runs +And confirmed this gets the tests compiling again +Can I commit this +Please review my attached patch +I dont think that is really the fault of the groovy compiler itself +c file in MRI but a lot of that appears to be +But really I would check with them since these are features provided by a third party +I also add testcases +Assigning to Brian since hes got commit access and hes probably the only one who can really judge the stuff that does +Do you have any D CSS methods on the input or the parent element? I removed those on my app and it worked +The error occurs with the query commands like query objname TypeConnectionlocalhost +Heres patch to fix the issueShindig seems to forget passing the variable to output after parsing the feed +Added some copyrights and cleaned up a bit of code + +Active as opposed to the lazy approach in. +Thanks Enis for prompting +Kindly help us to resolve this issueThanks and regards look inyo this cant diagnose this issue even with the stack trace +I think we should then leave the patch as is +Attached patch from the Github pull request +Looks good +Your patch will be applied against trunk when its committed so in some cases you will need to make changes to merge with trunk and theres never a downside to refreshing your patch against trunkFrom the patch file formatting it looks like youre using svn so you can probably just svn up and resolve any merge conflicts +I see this issue with jruby on tomcat +Verified in BRMS as a Resolved Issue in using Internet Explorer the Refreshing item dialog did not automatically close when archiving assets +Either we should put this in or move these flags from a global config level to the zone level where explicit code checks can be put in to disallow both flags to be set +Fixed in rev +I completely understand I just wanted to make sure we dont have a functional bug here +So you probably indeed encountered a bug in older JDK lucky prize winner +Ill add a note to the scaffolding docs. +Thanks +The tests pass on trunk and +Work should be finished now +Hi Frantiekthanks for the bug report +Status update Sept Progress On Due Date Fuse releaseDetailsAll examples in feature kit now have a CPI test and corrections to these test cases will continue as the java clients included for web testing are included in the final fit finish work will be required to allow tests to pass on slower build machines +Do we have some Arquillian test for this +As Im creating and testing the retro poms Ive found several missing features in retro that will need to be implemented to get all the tests working +It currently passes of the test in +Tested and verified +yes I youre absolutely right ShayThat was reason for choosing the silly generated alias we have people can obfuscate the alias themselves it does make more sense to use the basic class name as the default alias personally I prefer to have on by default and provide an option to turn it off so for the sake of consistency in plugin search results I would prefer to implement the two together +I think the problem is that MR hardcodes the +Claus James could we move this to so I can start building the kit or youre almost done with it +It should also consider get to solve this issue +Committed to branch and trunk +linking this issue since the main goal was achieved as a side affect of Hoss I personally would not close this issue since we have completed one point but not the second oneWe have decoupled the XML parsing from but not moved to based parsing +Also the updated plugin introduces a new error when creating the Java EE ProjectThis error is related to the KB builder not to the project examples +I checked in the performance test with revision +Or should it delete itThis should be handled as above +the download page in to the FAQ entry. +Fix the javadoc and refactor the test +Please post your thoughts here +I really like the title of the jiraChanging the parameter name is backward incompatibleCan you support both the current parameter and the proposed parameter for nowDocument it clearly and say that the current parameter will not be supportedfor this from or something like that +all righty trying it out with defaulted to false +Verified in dist build from SVN Since I had previously tested the patch I removed the file completely and then syncd SVN +Oh I thought this was but its actually a duplicate +Patch applied in +This has been committed to trunk resolving +Updated the tutorial with patch from Pablo Hoertner Konrad can you have a look at it +patch for the issue although its almost not worth the trouble +This should reduce the uneccessary waiting time +When moving to the next node you want to be careful not to pick up duplicated keys +Hiwe are currently evaluating a switch from Ant to Maven +It should instead be migrated or something Migration should be idempotent ie even if the migrated flag didnt get updated migration should be able to without crashing Duplicated code between update and update Each region that is processed results in a call to create which results in calls to the NN this will take a long time on a big cluster and is unnecessary No sanity checking that all of the for a table are equal Migration code should ideally be moved to a separate class instead of mixed with the code this major rather than critical since a good few of the listed items have been addressed by +Default value will be usedWarning Max heap size you set is too small reset to MB +In a large grid based search system that requires uptime this feature can be particularly useful +One question are we guaranteed return from for all errors or should we be testing for return instead +Setting the backgroundColor however shows that the parent div of the element is still px wide +Strange I was quite sure that I posted that one already but I cannot find it anymoreAnyway what version of Firefox are you using? I started getting it with FirefoxYoull also note that if you want a mousedown event on the tab but not a click the tab will show correctly +I suppose each time you add a new public method you have to add it to the gold file too +I just committed the changes +Ive committed the patch +But it happened! Ill give it a try in the QA lab +I also took the liberty of changing the extension from +What are your proposals +I can confirm that the is working for AS remote +Sounds good Xiaoming +Can not find the error but trying with the provided WMS nothing will be displayed because of a rendering problem. + Sandy +Tomas what version of Tools are you using +Hopefully I will have something to submit by end of January to complete Phase I and Phase IIAlso need to support upgrade and migration of legacy databases and update JDBC metadataLet me know if I missed anything else +The log issue is caused by API bundle which is used in Servicemix doent has the class of managed to walk around the issue of by changing the velocitys to use common log in downside of this change is we need to add a import of +The Log file containing both the Java stack trace as well as the SSL dump collected using all option +It works in the same way like maximum and message +The Details of Usage section written in manner is created. +Because were leaking a zk cluster in the tests somewhere some test is starting it up but not shutting it down all subsequent tests fail to start an ensemble on the default and resort to unorthodox portsIm looking for the culprit but havent figured it yet +add changed the status to patch available for triggering the hadoopqa for and apply the trunk version later +These blockers and critical issues are resolved but not verified +Confirmed that the plugins load and work fine thanksIn terms of the categories sorry I dont think I thought it through carefully enough based on the current JBT categories +Depending if we can get devent automated access to upload and cleanup then I would say we should use this as replacement since the atl based address a d machine is quirky maintained at best +Yes the fix for this issue will also cover user types. +If no appropriate appender currently exists then one is created +I saw the string like this GET test? HTTP in the nginx access log +Sounds good +Change directories to the service directory cd TUSCANYSCACPPsamples +Even when such a mechanism is provided it takes someone with knowledge of the application to turn it on again a supportability issue +To run the tests you need to execute the sql scripts that I refactored according to my knowledge of this databases +So unless theres a good reason Id rather not support versions especially since I doubt anyone is using it in earlier versions +One moment +However it hasnt passed yet +This could be easily consumed by the mahout algorithms +Closed due to the release of Tiles +Thanks Thijs +This also prepares for using Filters as clauses in reason to make things package private is to have more freedom to change things lateron that are not really part of the public api now +Fixed in trunk r +thankscommitted to as commit cbcebeaeeaab +Actually to begin with maybe itd be enough to just add general onoff Feature first and see if that might be enough +I resolved it as code fixes have been backported and assigned to QA for testing and verification +But its using the same return type and its confusing +once the build passes this can be committed +What is really needed is a redeploy notion that would recycle the dependent deployments but such a notion does not exist +Alessio I remember you did some progress here +Schema browser page Yep after selecting the geohash type the URL changed to httplocalhostsolradminsinglecoretypegeohashAnalysis page I was thinking of two horizontal scrollbars one for each pane not vertical ones +But I cant find any naming generation rules when the is being represented by a Map +Now we need to load against other patterns since there doesnt seem to be a way to create definitions with the old naming conventionsMy main point with this Jira is that someone changed the naming conventions for a mature tool without anything obvious showing up in the release notes Jira list +Most developers are probably not going to have debug logging running against the servlet resolver and I would expect that when I register a servlet under a path where it will never work I would see something more prominent than a debug message +Im adding directly default socket factory for the test +Ok +OK +Verified this on a single node secure setup +Thomas can you take a look at this? I am trying to eliminate failures in the testsuite in prep for release of +Brandon +x and marked as upstream +I leave it to you or somebody else to implement it +But I couldt confirm its jamm not specified as java agent result the heap problem +It updates the db with correct public and private IP of GSLB in the DB +Thanks Kan +I think the real answer there is that when B joins it should wipe any old state it had from A before applying new state +Attaching patch for +As per the patch the online regions will be added only after the region is fully happened +Can you give me the Link for the Forum Please! +I think these two are related +Currently the applications definitions will require editing for each specific locale +Follow up patch committed. +Sorry the above comment is not mean for this patch +I think the syncd flush is a big bottleneckIs this because indexing stops while the DWPT segment is being flushed to disk or are you referring to a different sync +Hi I just created that follows an ee failure caused due to this patchI want to make sure that any fixes we make there does not cause too significant a performance degradation here so could you please track that as we solve it? Also if you have any solutions Im all ears. +I have not included other reports as I feel the two are good enough but maybe thats just my way +Please contact him first about this we will need a list of the javadocs that should be includedI dont think this will be done on the EAP stream so dropping that from the Fix version list +Please remove the referencesThis is an issue with the relationship even though the error reports a +Im uploading version of the patch with one more small change to add file name and size to the JSON output +Very hard to script this in terms of javadoc groups +I have addressed the latest round of review comments and rebased on latest trunk + added + +Thanks! +Linking to the other half of this issue for deployed artifacts +If you are still concerned about updating timestamps and version ids then you may want to revive part of your patch +Once this fix is made to the trunk I hope that the reporter can rerun their tests to see if it is also fixed +Actually I just found I was working on after I switched to the issue is gone +Incorporated Erics comments +As always this is not set in stone +I believe that a startup error occured and the was not created +This is currently not supported +Fixed in revision to the branch in revision . +OK Ill add a null check in s ctor +Thanks for noticing that Dag +Which brings up a good point We need to agree on which way the association goes country to state or +Maybe you can mark it as deprecated +This patch allows under replicated blocks to be added to neededReplication when data nodes get started +We cant reproduce this problem. +If you want it in thats fine tooLike I feel that without this is a recipe for disasterPersonally I would love to get into that would allow for safe and +Thanks very much for the helpful spec and trace of the issue fix +But the exception was not clear to me +There is a bug but added sparse matrixvector classes +Max allowed age I Current disk usage +Please create a new issue thanks +I agree we should test the common drivers before changing this +Should we consider implementing in filters? We already have a which does something similar to what we are discussing here +Thanks Ivan Stud for the participation +Question how many of us actually do need to be tied to the web of trust anyway +Sounds good to me +The OS would be made available the same as any other by uploading an ISO or a template +Trunk rBranchx r +So once it has a Metaclass the property accessor wont change it +Ive modified the integration test for to inject the configuration from just to be sure it works +class files from being filtered +In certainly not because we cannot modify released versions +Ruby master does not appear to fail your script +Applied thank you +Ive not been able to recreate this yet can you provide any more details or a testcaseThere is a unit test for reading and writing properties which works ok +I need to work on distribution in the near future so I can revisit classpath management when there isnt the focus on getting out the door +I think covers this case much easier to contribute new for different types and anything related to happens much later after is fully up and running. +Lists are fundamentally ordered making it obvious how to compose their ordering into a larger orderingIm not completely against adding this feature but it does seem a bit unnatural +great its working fine +This will be fixed when jetty reloading is done +How did you manage to fix this issue without running this testAnyway Ill test next which is probably a related problem. +Note I changed the debug trace flag to RAF to accommodate an exisiting trace not related to recovery +thanks +So all issues not directly assigned to an AS release are being closed +lgtm assuming is clean +We should decide whats going to happen to g and m by +Bumping to Struts +Relevant callers now pass the total number of input threads instead of just workers +I am not sure which one +You will probably get an error otherwise because once used transactionally it must always be used transactionallyA transaction is one that has finished +The problem was fixed in and there wont be any so were all set with this. +Keep us posted on the perf number after the fixes we discussed +We should push this fix out to other lock tests +Resolving as invalid since we already have this functionality. +Unlinking from since not necessarily required for believe all of the issues in this JIRA have now been addressed +Could you please provide some more information about how to reproduce this problem? Or even better write some small test case that exhibits this problem? That would be very helpfulKind regards to reproduce issue and no feedback received. +I would also like to be able to limit the total number of entries that can be returned but we have the issue with ordering here like Sid said +I took a quick look at your patch it looks reasonable +Well I can write a junit test yes +Adding a simple cache dropped the CPU time for my workload on from ms to ms and improved average latency by a few as well +TYPO Forgot to ant clean fixed and completed attrunk r r r r +Merged PR +I am tests now +I also tried the fix and I no longer got the error message +bulk move of all unresolved issues from to +Attaching initial attempt at this which logs the query highlight and facet parameter for every query on INFO level and the whole query string on FINE level to reduce log noiseIn the long term the parameters logged on INFO for each request should probably be configurable in but this should probably be tackled as part of a separate issue +around this system +We should be able to get this into +I set the property ignoreerrors to true on the get task +Im wondering if we could close this issue since this test now is run in the harness +Please can you give me the entire stack trace +Here since its its basic usage of guava I could remove the dependency +Unfortunately to my knowledge these are not fixable these classes are the only way to access this functionality and the warnings are not suppressible +Fixed in trunk +Thanks for the prompt response Ryan Ill send further questions to the mailing list +Was done by patch for +In addition to the fixes committed in and this patch should likely address the timeout issues in the listed tests +CommittedSeems fine to meI just hate the bit above it where we create an on serializationdeserialization but its rare enough in the scheme of thingsThanks for the patch Clint. +I just took a peek at your other images especially the and it looks fine +Im pretty sure that such construction is not required and can be resolved with good refactoring +Implemented in r +Perhaps Im mistaken but I dont see the downsides to providing a option to the configure script +Im going to close this for now I cant reproduce it and havent had any new information + is in contrib now +Now it goes to extract and asks me which program do I want to use to open this file +Done. +This patch supports EQUALS with null which the previous code did not support +Merged with the Daniel code of with richard with richard change +Id like to tackle this after we get working +bq +to trunk thanks for looking at it Stack! +Linking blocking common ticket +please commit +Thanks Daryn! +Can you submit a unit test that demonstrates the problem? Basically getList should work just like getString does +Please open a separate Issue for that +can we limit number of digits for the attempt ID +JBT uses Eclipse P that doesnt update plugin unless the user required thatThis is not a JBT issue. +That is why I reassign the issue to you +We will use sequenceId to sort like in compactionWhen merging there is no overlap between these two files from different regions I think its ok even if two files have the same sequence ID since we dont use it to distingush the files in Store any more Patch looks good for me +It will take a bit longer but should be ok +I wish I could have done that sorry +Also cleans up some fields that dont appear to serve any real purpose +I dont seem to be getting this error when running FUSE from Eclipse Im using the steps belowWindow Show View others Server IONA IONA FUSE ESB was wondering if this exception is being reproduced consistently +Linking to which has the same cause but different symptoms +Poke +bq +We will not fix this for +Well BJ do you have any suggestions +In case it is not ready yet update the Jira with justification comment and preferably with date when it will be resolved +This value is probably being fetched from the CPUs cache and since no locking is enforced it is getting a stale value +Internally ODE stores values associated with message types as messages +Is there something we can change to make lease recovery fasterconfig parameterblock sizemax file sizeThanks +Issue may already be resolved needs to be testedconfirmed with an uploaded example +Probably fails some tests +We thought the health service should not mix with child tasks reporting status and hence kept it different +Marking closed. +The reason why the app failed to start is now included in the error message r +In the reproducible testbase autoCommit is set to FALSE and more important the resultSet returned by select query is not closed +Tested on Ubuntu with JDK +In this case we should use two instance of VPE +I think this has been added a while ago. +v +if you want the heartbeat feature youll still need a dummy record +I uploaded a new Quickstart that uses and displays a blank subpage when the user selects a list item +This should be resolved in the trunk revision Please confirm that this is resolved for you +Id add that this makes a big difference with turned on since that creates many small sstables +I was misleading myself the M size was relatively early in the run +Thanks Jesse +Can you make parameterized test so that being true can be exercised +Maybe the easiest thing to do is just change the error message to Error starting transient name service on port port +Well if its been there since and this is the first time its been reported it hasnt caused the world to stop yet +duplicate to is the reason I was not able to repro it again. +Jan have you had a chance to look into this? If so can we get a status update +The should be in oagsecurityrealmproviders with the other login modules +Because pages is application scope this causes all sorts of problems in a concurrent environment +applied marking resolved. +Are you checking these out with a unix cvs? If so then cvs will do thisautomatically for you +The fine grained approach doesnt make sense to me +They look great to me Laura +javadoc failures are ve committed this and merged to +Showing the source code now needsexplicite pipelines the direct mapping from request param to filesystem isremoved +Uploading patch for Yahoo! internal distribution with this issue resolved. +Its already waiting for s there +Hi parses the TEST +Trunk target files should not work with Indigo only Juno +This is important for the stuffer query +Sorry but why not ask such questions on the mailing list rather than opening an issue here? Just for completeness sake this is a bugissue tracking tool. +Fixed with +You mean just as a javadoc commentYes +Added example patch forContains a ugly hack for state saving in the and for showDetail and Table sorting + +tis doneI think you know there were lots of failures that Ive fixed in this diffSorry I didnt pick up on that earlier +The correct commit id is in abfadadeedebeeca closing the bug. + itself seemed to crash in an unrelated part of the code recycling Hudson status +r spelling correctionrenaming from Authenication to Authentication +Wayne is this ok to close or were you leaving open for a test? +Please get back to me if you see any problems +Okay Ill make those changes next time I have a free block of time +Can I get this issue assigned to me +Hey man commit already +What if you need one account to have a different number of networks? I know you have the setting globally but I think it would be better to add some granularity and permit setting. +And WTP team has fixed some related bugs +I will tell you the branch etc next +Im not sure if sorting the set is the right solution +verfied at rlooks good +Please reopen if you see the failure again +I agree that the trailing is a little subtle +So its caused by the patch +But we can add support for the hadoop metrics later +The stuff looks great but it wont make it for unless its completed in the next couple days +Plus I believe the Forrest logo is a forest of feathers no? I just thought itd be nice to have a theme for the skins sanctioned by the Forrest project and the phonetic sense of the name just came to mind +Updated patch to deal with paths containing a mix of slash and + was updated to deal with empty elements +Plus that the sep codecs skip data reader was tricked by this bad state +Thanks to for pointing me in the right direction +The fact that is not a bean type of Office is implied by the fact that is not a local interface of Office +Thanks Drew! Good to see you jumping in +Hmmm +As I wrote Rhino in is source and binary compatible with Chrisversion from its usage in Cocoon point of view +Trunk code is changed +CheersAshwin +Community doesnt seem to have anything to say on the matter so Ive gone with the historical approach of just requesting a license ourselves +Thanks Yip and Deepa for the info on the upgrade testing infrastructure +Yahoo! doesnt appear to publish anything outside of Apache JIRA +I think Seam your try to integrate with the JSR Swing Application Framework it is already based on annotations and it really makes it a lot easier to develop Swing based applications with UI databinding to JPA Pojos I think Seam support would be a natural extension for it +FYI there already is componentscommonrulessrcmainjavaorgswitchyardcomponentcommonrulesutildrools so we need to figure out why this isnt doing what its supposed to +The was also updated to add the ehcache java interface and its dependency slfj +Adding to the latest +Well so we can create a new tool called that will handle all non job and non admin related stuff like getting clustertracker information etcWe can move the existing commands thereIf this is the plan I am willing to work on a patch for it +It was a bug that it handled stopPropagationMaybe we should add a separate setting for stop propagation in +Mosca +Ive added a test case for it +I merged the patch to. +do you have any measurements of tons of megabytes +This sounds natural to we do plan on providing an option for customizing this behavior +its forced for now so dont hesitate to rename it if your war files are loaded following to the alphabetic order +This actually can happen several times depending on the number of optimization levels available in the compilerSo what is different with the client VM? First we have interpreter to compiler transition which may be one place where the JVM bug exists +Pull it back in if you think different. +This is the abstraction taken from the June version of the entire patch +Sorry I changed my dialect from to and now it works fine +Please review. +The first patch is not committable +corrected the npe. +Ok I left the new mechanism in but set everything to synchronous to avoid problems seems to work now and still is faster than xhr get manual eval +Committed to branch and trunk. +As a practice I rewrote improve listStatus performance in a backward compatibility way +bq +its not properly this way committers are invited but anyway would be really nice to know Igor is interested to joinLets move the discussion on the dev ML. +I agree we dont need an empty interface but maybe even better than remove it could to move all methods we need to an interface and avoid keeping dependency to theThere are many changes scheduled for the maybe at some point we will need multiple alternative implementations +Thanks for the reviews +I am not sure if I am following you +If this isnt confirmed soon then we should mark it as later and not worry about it forIt would also be good for a Mac person to provide a patch if this is still broken because I suspect most of the developers dont have access to one +I will commit this if everybody is happy with the changes +should be fixed with webtest plugin version that calls test in a separate VM with its own verify + Mice do you have any updates to share with community +Yep I think we still need this to go in +That comes from the nodexxx line in testsuite +Im not sure we need to do this +The code rethrows unhandled exceptions which still will result in the conversation lock not being released +Thanks for reporting +Tests pass for RC +see docs for latest release and examples directory +True +Ediiting adding or deleting files in this directory will cause the Derby recovery system to fail leading to unrecoverable corrupt databasesFor the file in the seg directoryDo not touch files in this directoryFiles in this directory are used by the Derby database to store user and system data +Based on discussion with Olga we need to limit that each command option can be specified at most once +Not fixed in final +This bug has been marked as a duplicate of +Applied nd patchtrunk +Move the fix to +Patch rebased to trunk +Attached application which demonstrates the issue +This is needed to enable management for other resource types +Sorry to bounce this back but Ricks latest jdbc test suite checkin has thrown a spanner in the worksI am a little unclear how Ricks checkin impacts your patch here +I get the exception too in the head svn +permlink example breaks because of stating that is the web directory where it actually is webapp +Support for other candidate rec +And if we forget to do this anywhere we still get deadlocks that are very hard to debugDo you anticipate that this issue could be fixed in a way that wont require a proliferation of extra connections +Esper is following SQL standards use property is null to check for you mean the Statement Object Model API? Is there an is null expression missing in that API +Maybe the problem is that were using latency instead ofWhat does cfhistograms give for read latency +alright Paul thanks to your whining ive finally gotten around to whipping the into shape +In my setup the patch does not cleanly apply on trunk so I have generated incremental patches for each of the review changes +move to doc version for less overlap with engineering tasks +Is still a work in progress or should we close this issue +Forgot to add a file +With that addresses for the patch +The problem was that there is a XASX and a SQL State in +Net clients which were removed from the repository months ago and not the current +recently made this more complicated and Im not sure I fully understand +Just do it +We introduced this behaviour in. +enabling slfj configuration watching sounds good to mealso need some working examples at the bottom of the configuration for enabling debug selectively not the stuff that never patch enables watching on the logj properties file every secondsSo you can alter the config and application will reflect changes without restartAlso added proper examples of enabling package specific logging +This works now +Just another short comment as I have amended the JIRA notification scheme in use for Castor. +Use fast algorithm for solving Vandermonde matrix +for the patch +See +So the other nodes marks the replacing node as dead before the token bootstrap starts hence should be storing hints to the replacing node from that point +And sorry HowardRelated to your svn account I thought Howard was your last from problem Markus +FYI Im pretty sure this is already in JIRA somewhere but I cant be bothered looking for it right now +sighOleg +What is the classpath value? More specifically do you have more than one JAR begining with +Added entries for the additional bug fixes mentioned by Rajesh +Will run all tests later on +will probably come in handy when trying to figure out a solution for this +The plugin seems to be correctly built but I get an exception when opening the editorI remember something that if your plugin contains jars then you may not deploy it as a jar but have to deploy it as a folder +What you mean by +Verified at revision +I dont think soObviously as a JCR implementation JCRSPI needs to provide a to the JCR clientHowever right now an SPI implementation needs to provide implementations of both and +Heres the test case its a junit test so you can just run mvn test to run it +Some of the stuff is not implemented +Theyll be archived all over the web in addition to at CodehausIve removed them from our archive anyway +Thanks Jarek and Namit +When the client comes to DN all DN needs to know is that the requested operation has been authorized by NN +What are the subtasks for this effort? I see it is configured in the branch but what is left +done +Fix is pushed into +The installed POM does not have the groupId set in it so that projects transitive deps fail to resolve +Ill try to figure out how to fix that and possiblyprint the stack trace +Once all are found and thus removed the next compareTo call will see that there is nothing left to match on and return a +This needs further thought +This is not a blocker for AxisC and so Im downgrading the priority +No it wasnt overlooked +Suggested patch attached +I would just like to confirm so I know for next time the problem was not actually caused by my patches +Dont hints timeoutYes the timeout is set to the of the CF the hint is for +Offering the registration point at the db seems pointless theres little value in registering at DB as its got a short short lifecycle +Im using final version thats why looks quite old +Indirect jars are not found by in external library directories only when packed into the war +bulk close of all resolved issues. +Added test source code +For the security conf properties we should do the same as other HDFS security propertiesJetty caught the exception and simply printed in the log +I think just keep it simple? User sets RAM buffer size and we compute fixed highlow watermarks from there +I dont understand at all +updated patch distributed tests still s my prototype parsing patch that adds support for multiple fl params field globbing and function queries +The vast majority of the code is disabled by default +Looked at the patch +can confirm this is fixed in the nightly builds +This doesnt have to be a special test case or anything like that just a failing project build that you could zip up and attach and maybe some instructions for what options to use to build it +bq +Old bug +I did not realize I made a mistake when doing the release +Just put it in the request parameters +No response assuming ok. +Then test again +zip file containing Eclipse to be fixed with that work is well underway with AS all previous community releases are +Since it is a clear abstraction to logging frameworks it looks really nice +So we can implement that +thanks for the least need to correct exception handling and formatting +Then probably as the error message indicates a declaration cant be found for the root element +It looks like this functionality is available as optionmapper Needs to be tested added to regression pack closed +See the attached scenario log +Just some notes after looking more at this stuffWe have two different assigns one that assigns a single region and then another that assigns multiple regions in the one go +The that generates the report is not compatible with Jira . +Thanks for applying the patch Paul and spotting this error +We do not use the resolved status. +This so called feature is now supported in through Yecht +This specific routine might work for some subset of situation but the policy to disable new syntax features in soft upgrade seems clean and leads to easily understood behavior rather than allow some and not others +There is no pull request attached thus Im rejecting back to Open state +Shouldnt the also be part of the plugin? Also since it is a larger contribution and not just a small bug fix it would be great to have a junit test within the patch +Added schema and parser support for TCPIP gateway +I had made changes to the code in the past month and maybe that would resolve your problems +Patch applied +Where Ive seen it on images the machines were visible on the subnet publishing an would have worked +Alsodid not fail when run individually +defer to +Verified by Tony. +Which only happens when running in debug mode +would be nice to measure the effects of it +Patch already outdated +I reject the documentation fix and have supplied a patch to allow to report back correct details with axis order matching the dataA second fix to adjust request generation will be needed +Moved back to +a comment as requested +It allows the user to generate different mixes of read write and list requests by specifying the probabilities of read and write +Can you provide us with the document triggering the exception +This was necessary because the security around the Administration operations is broken +It it primarily a upgrade effort for the componentCheersAshwin +It doesnt make any sense to force the user to use a node and onclick event in order to get the desired behaviour +So I dont think this is a askedClebert I think his point might be that the results of the initial security check may be cached indefinitely because of the assignment to lastCheck +I see +This is required after +If you update to the latest version of the you will get this configuration change automatically. +Also present in ER +cancelling the patch +ah you are right +Now let me try and use the mvn release plugin +to you Paul +Simple and straightforward patch +Committed thanks! +Keep on the most common methods in and leave the advanced options as fluent builders on the xxxTypeWe gotta get the methods listed in the code completion list from the java editor when end users use the Java DSL + still doesnt exist and this JIRA must be reopened +The hope is that I can find something that compacts well without having to be run a lot so I dont want to bias towards things that compact too much +But I used in Seam application it can prevent jsf navigationIs there something like this in Seam or JSF +We want the Thrift API to be +Make and share the same base class and also corrected some problem in the framework preventing us using inherited classes for UDAFAlso fixed most of the checkstyle errors +Currently NeethiC security policy extension does not support this +In the former case if macro foo is defined in a parse file then foo renders to foo but if foo is defined in a library then foo renders to fooI general it is very difficult to know when text that appears to be a macro needs to be escape and what the resulting render will look like +Thanks Liyin +Committed patch to documentation trunk at revision +Sounds good! Let me get together a new patch for this and work on the documentation piece as a separate patch as per your suggestion +I forgot to mention the user session attribute defaults to the servers time zone if no user preference is setAfter reviewing Vinays patch and giving this a LOT of thought Im considering creating a separate class for localized datetime utilities +However this seems to be more of technical debt rather than design choiceShutting down of components on configuration change is needed in order to make sure that there are no resource leaks in the system and that these components can be dynamically reconfigured during runtime +I cant say it was easy +Make Questionnaire remember the +This is an update on the from JacopoAdds another main menu item to access selected warehouse so you dont have to go back through the main menu every timeAlso includes new screens for stock take allows you to look up products based on SKU or PLU and view their ATPQOH and bin locations +Committed +Everything looks good at fist glance. +Feel free to close if so Ill try out the recommended fix in that JIRA in my setup +My bad just need to make it +The change can be described as adding a configuration option to control the loggers log level and the default being LOGWARNING +anybody care to reviewcommit? this ones trivial +meclipse configurators dont configure JSF facet that adds JSF nature +unscheduled issues to the next release +Save the file and all should be goodCheersIan is very likely caused by the input data +My only concern with using value is that a term like target is more descriptive of what were describing who is the target of this section +path for files under +Im working in a patch +reopened since the fix has been reverted +Committed at subversion revision +Theres still more that could be done but well do that as ongoing performance tuning work for and. +m file? If so you might want to track this bug +seem to work on alpha +So all issues not directly assigned to an AS release are being closed +trunk +After restarting the machine its working fine +Includes an intellij project file and a few jars that are currently being used +The string is not in anymore +Closing for now as the included patch still passes the TCKPlease open a new issue if you want to us to try and challenge the TCK or create an updated patch that passes with the current TCK +The cumulative pull request for RF was openedbut it does not works as expectedIm rejecting this pull request as more investigation is neededPlease the pull request when its ready to be merged. +Closing issue now that Daniel has verified that its fixed. +Worth the tradeoff for now though we need to be able to scale the huge number of shards. +The patch adds extern C to interfaces +Changes to any is resolving the issue but based on an offline discussion with Tucu he advised to use the pom in similar to the root pom any variables set there are inherited by all modules +configures Sling to use Eclipse JSDT +We should handle both but not necessarily in this JIRA +I feel that this is duplicate of the bug fixed recently +I had just copped out on the static configuration part and was using a system property +The tracking URL is correct and not unit tests +Also the best solution is probably to store the length of the posting slice into the skip list byte pool +I guess this is what Oracle is working on +fixed in revision The tests were not integrated sinse similar tests were already there binded existing tests to the JIRA good +See also the thread running at the user list +You can also do jmsObject to force Camel to use a. +Done +I honestly have no idea you seem to have declared the view metadata properly for both columns so should be able to pick it you tried with more recent versions of is the problem still reproducableIf so I guess Ill need a sql script allowing me to setup the same situation on my Oracle XE install so that I can debug it step by step or else a VPN connection to your database +Please find attached a patch that inserts codes for properly calling full GC into Grays +Ok problem was related to solved it +This API is by now several years old +I do however still prefer using Intellij over eclipse though +Since we will be supporting both istaton and istatoff seems like we should test both modes +Should be fixed in next nightly. +and only but perhaps no other Subclass of Searcher might work with a docBase +Im thinking of switching from cruisecontrol to continuum but stumbled onto this issue as well as trying to get continuum to ignore a file +can be +Pedro Can you try to check and set the gateway of your computermgmt server +The attached tarball contains sources to two Java classesLicense granted for inclusion in ASF works by Team Gigabyte IncNote that there is one known logic bug in which Ill be fixing in my own codebase shortly +To clarify both client and server dont care about chances of eavesdropping or tampering of communication +read committed +Contributed by Todd Lipcon Aaron T +Thanks for patch! I have looked over it and it looks pretty good +Tomorrow I will start working in real examples and continue adding functionalities +Oops left out an if statement +What errors do you see exactly? From time to time Im seeing A lock could not be obtained within the time requested error but otherwise everything works as expected +I think the problem is really limited to +did I commit this? If I did it was an accident +This adds another layer of locking to the Request queue which in itself is a Blocking queue implementation +bq +Yes additionally it is impossible to scroll to infinity and loose documentsimages put on WB +BTW we notice these on a RHEL platform with Oracle JDK update +Looks like there are conflicting changes +Has this problem ever shown up again? Its pretty old +I did understand that it was not new However it was the action of removing the Delegating which gave rise to the issue +Committed Thanks +i dont have the cycles to review that docs myself so they would have to be removed if micah isnt able to update it +Addressed comments +The properties null shows up for one message out of tens of thousands +Ill work on that nextHonestly fuck it! and are so exotic Id leave it out and move it into a sep +Looks like I hadnt rebased like I thought I had +Brock +Maybe its the container managed part Im missing then +please try latest svnthanksdims +Could you attach a diff as well? Otherwise I will diff it myself +I have made some modifications to the page +Later versions work mostly so I m not making a for this +I dont have on the code any longer +Will add additional comments as we resolve this exact connection problem +Good one Tim +Both issues are with the same object trying to be stored in the session +This patch seems reasonable to me +rev to generating release notes +It is very complicated application to reproduce all steps +How much of the stuff do you want to keep? Or are you just popping it in for IPs sake +I agree with that +See comments in that issue for details. +Hmmm +just changed the script to gracefully error out in case its not able to find the pid file +It would be good to get in looks fine to me +committed to trunk. +merge with branch and change resolution version +still doesnt do fields as requested in the description +When I change the port in Eclipse it should propagate that change to the server +Closing it. +The revision that claims to fix this r broke a whole bunch of other code including type inference on methods and constructors +Updated to mask these warnings +It removed all of the ant code +No no no big reorgIt was only meant as a cleanup task +This is fixed in we are not planning patch releases for branches for apache so an upgrade to is to and problem does not exist there. +Marius A coworker had different suggestions for the API +This ok as default but notin completed +I also added in some checks to make sure it wasnt just the first run of a new virgin rulebase but to no avail +What are on the names of the files associated with it +Attached my patch which is an ugly HACK as it suppresses beeing thrown for both duplicate imports and duplicate macros +If you have any issues or questions please let us know +Please provide more background +hey Jordan I too experienced the issue on a machine +I understand why you did it but in case someoneasks I just want you to be aware that this bug is in and itrenders most of the infinispan configs useless +By forcing all keys to strings were eliminating this step thus speeding the serialization significantly +Or +on the patch +But yeah thats for another JIRA +Closing all tickets +helps me understand that the exception was generated on the remote server and is useful as a piece of debug informationWondering if theres a better way to convey this information maybe via the message string +file +Pi I dont follow your last comment +Somehow the system picked it up as m +Ok Jeffrey and have you created and declared any new homemade Flex rules +I will commit this once our internal QA tests are run and verify that this does not cause any regressions +regards kaspar +Sending scoutsrcmainjavaorgapachewsscoututilTransmitting file dataCommitted revision . +it breaks the gaurenteed session scope identity not good! +Final patch +Omit the Stacktrace +Changes associated with previous comment +Which DB are you using Sylvain +Kihwal I just put this into trunk and +m adding as explained should fixBtw I will change the and add automatically the in the classpath when running testsSounds a nice behaviour +does not exist +We will make the change for though +This patch has been available for quite sometime and also passes all tests on Hive QA +A couple of simple logos to get things rolling +Logged In YES useridIt is jboss version +I guess this can be closed as works as designed +Im going to make patches for this so you can review itIve managed to get the module dependencies running it is not as beautiful as I would like to have it but we can refactor it later +Thats the solution I was thinking of do not know if there is a test in our old test suite that covers this +Maybe Thrift Developers and ? Also what would you think about having it read the version from so we dont have duplication of that information? + +I provide an as payload to putBlob the file size of the input file is bytes +Here is the first patch from my side +Closing as this is all resolved now +In the meantime use which does work at least with the snapshots of Weld and the +Yura I have reverted in trunk and it seems that it was not committed to branch +Thanks thats a lot better indeed good job! +csv file and return the correct orgid for the +Back ported it I like it as it clears in the case an exception during the login process. +Please do +Thanks Lin +Egads this is sneaky! Nice test caseCan we move the fix up? Ie leave as returning boolean indicating if a new reader was really reopened but then in always carry the gen forward once weve called maybeReopen +What do you think about using one of the existing solutions rather than write your own? +Setting to will prevent mmap from happeningbq +Id give it a try tomorrow evening +I like the idea however I think that this is out of Karaf scope since the target is to have Glassfish use Karaf and not the other way aroundBy that I mean that it should be implemented either at the Glassfish side or as an external project +Removed legacy layout of running from source code +Attached the patchRight now this was changed so that the existing code paths would still be followed +svg svnnative +VERIFIED +Can you reproduce it anyway? +If these limits are met or all jobs in a queue have been looked at and no task has been found the scheduler looks at the next queue +And it should actually be pretty easy to write too +patch that changes this field type to use stemming +I modified the extension logic to not bind private superclass constructors at all since under no circumstances would they be invocable +One less bug to worry about! +Merged to doc branch at revision +Regardless of clover running the issue always occurs +Closed. + to show the problem it should be fixed to expect the proper enhanced and added to SVN and problem resolved +Automatic cross reference if managed at repository level will gradually simplify its a suggestion and all at your disposal. +Sorry for not responding many requestsqmx sent us several and Ill pull those and do more if needed within a week +Itd also react to config changes to the parsed state on configuration change +See +The colon is needed to separate the message from the timestamp header +Patch tested with Firefox and the check for last is required on Ajax POST too not just Ajax to trunk and branch +I believe its included in +It is memstore buffer +Renamed and done recommended changes +Patch applied +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Not good to hear it +Thanks good to learn +Thanks for reporting Ill check +The slightly sophistication is that the centroid being added can be added in pieces to several other centroids if they are the closest ones +Please attach current exception even if it looks the same as exception +is blacklisted +Going to commit this small patch its just testing first if the file exists instead of letting the exception happen +in +In that case not being able to disable dynamic would be bad +by default this is disabled. + +I want to study this issuefix a little closer before committing anything so Ill push this through once I get time to sit down properly with it which will stil be a few days from now +I will commit the current patch to +Ok I figured out a hack that seems to fix the Chrome issue I can do a deep clone of the pdType node and then replace it with the clone +Does this sound good to you +Add Apache license to the header of the file +In future could you use instead of Issue in the commit comments? If you do that JIRA should be able to link commits to issues and display them in the web interface +Patch committed in r. +Why would that work but the constructor parameter wouldnt + +Ill look at it again on Monday but its probably safe to call this a false alarm +I think it was some template grabbed from somewhereYou propose another variant of what I was saying about moving the links up +tests passAdditional work and testing in progress +Could someone please try this patch +Thats reasonable but really this could be optimized as well for selects with no transform necessaryBut at least when any mapper has printed rows Hive should print those rows and kill the rest of the job +As long as we keep the workaround we might as well make it more convenient for testing purposes +We were experiencing issues with the file system at the time so have since moved to a different architecture + +Dirk do you have a schedule for this feature? If you think its easily added then we can take it in otherwise we should probably postpone this +How about we resolve this one for now and you when it would happen again +This is a follow up to targeted at new release +This will prevent the reverse dns lookup in +Changed the patch as per Thejass suggestion +This issue is marked as resolved +Ill split patch into two or more parts to ease reviews +It would only take a small amount of code and would save the users from having to use rd party since its use in this case seems quite trivial +Good catch I didnt realize that other commands were doing the same thing Ill create a new patch shortly +Im leaning towards doing the merge with both the offlineonline code and it is easier to deal with one instead of two +Issue fixed and tested. +A subject resource and action are required as well +Yes it is +This looks good +Since we can detect that this is a self join we should only be reading the table once not twice +Silver circle is very classy +Seems like a simple enough change for too +I encourage you to try it again while I look into the other issues +committed to trunk + And also did I understand correctly that you suggest to use default lower case names for dbentity objentity procedure query etc? Yeah sorry overlooked the specifics of these objects +bulk close for +We could also consider using an AOP proxy to wrap the +Heres a first stab at implementing this +project with class to test error using danielim using activemq so ive attached a project to show you the error please have a look +avsc files so people at least be aware that we rely some how on Avro for persisting data +API is available to listdelete affinity groups of users +Thanks +I think this is fixed please reopen if there are still problems. +that wouldnt work for my use case since it wouldnt allow for the response to be processed by another thread +Match Rating Approach encoder tests I think +It doesnt appear there was any data loss. +This design deployment scenario will need some testing on separate hosts to see if its feasible +commit m OPEN issue Migrate build from Maven to Spring Build in order to work with Spring Build all abstract test case classes are now named with the prefix Abstract subclasses of the above abstract classes have been broken out from inner classes to classes UserscbeamsWorkspringsourcetrunkcoresrcmainjavaorgspringframeworkconfigjavainternalparsing UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues UserscbeamsWorkspringsourcetrunkcoresrctestjavaissuessjc UserscbeamsWorkspringsourcetrunkcoresrctestjavaissuessjc UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Sending UserscbeamsWorkspringsourcetrunkcoresrcmainjavaorgspringframeworkconfigjavainternalparsing Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues Deleting UserscbeamsWorkspringsourcetrunkcoresrctestjavaissuessjc Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaissuessjc Replacing UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues Deleting UserscbeamsWorkspringsourcetrunkcoresrctestjavaissues Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Sending UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Deleting UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Adding UserscbeamsWorkspringsourcetrunkcoresrctestjavaorgspringframeworkconfigjavainternalmodel Transmitting file data +Finished backporting to +NInstead of starting msging between master and RS why dont we make the master update the status in the znode itself so that based on that let the RS carry on with further processAnother way could be make the RS create the znode in SPLITTING state +Fixed build to restart Hudson patch again in attempt to restart Hudson attempt at patch current patch to upload new with Aarons review comments with review comments patch looks good +yes a good idea also should match on type automatically +Which is what this task is all does that affect everyone else +Patch looks good +This patch adds additional simon metrics to record block validation statistics +So we need to make sure we get the AOP in build. +Could you please submit a reproducible test case possibly using an embedded HSQLDB or the like? just hit the same or a similar issue using tests +objects are not required to be interoperable across vendors. +I see was used by old forget my advice +Ill try again using the current svn code tomorrow +No its not in integration tests only +Suggest retrying with client and server from or not really enough data here to investigate. +fixes for SCR release +Schema changes made as documented in response from Tom +Yes the following test reproduces it when is run with the affected IBM JVM +Ill start with a short analysis on how to best implement this +bulk close of all resolved issues as part of closing items. +Ok all is now implemented. +Add a check to see specifying COLUMN in the alter statement is optionalI tried the above tests with patch +Picking this up now +Is this incorrectly assigned to Prasanna instead of Parth +If this situation occurs now an alert will notify the user that the ajax request cant be completed since it cant find the component to update +How does it manifest itself? You have a log +As for the other issue open up a new ticket regarding its type conversion support. +hans +Close out issues. +That patch looks like it should work +It describes this in the header of its javadoc +It is working fine nowI also ran the test target which passed +I didnt spot that when I first diagnosed the other problem +But some countries dont have any stateProvince inmaking it optional at both placesSpecial thanks to Rishi SolankiREGARDSONIMA Onima RishiYour patch is in trunk rev +Hi AmilaI would like to help and take over with resolving this JIRA as it is already staying too long in the queue and we urgently need it +Thanks Eric +A printscreen image with the text rendering issue +for example is making clients other than snacc based ones hang +Pull request merged could we close that issue +The Archiva is running under as a service +I disturb your vocationDaisy told me she found some repeated work between this patch and your newest patch after she reviewed the two patches +yes i know this paragraph and there is nothing in there which prevents a redirect afterwards +I think so +This can be added in think this fits in better with the enforcer as a new standard rule +Those may need to hide behind some kind of compile time version check. +oops I hadnt run reconf +This is the new interfaces of LUNI and are patch files for and unit test forCould anyone verify my fix? Thanks a lot +We could accept a URI and check that it is of the form we require +I just run into this issue Im using +Patch is ready for review +Closing all old Resolved issues +It would be possible to teach Qpid about kill HUP to its config files although this raises questions about what to do with things like the store directory if that changes +That being said a solid third of my Android apps totally crap the bed when I do this so its tricky to see if its us +Fixed by using +JasonI guess add the two files in this zip to the root folder of and then let me know and I can do the rest of the guys +Let me know what you think +Ive started by copying the over to +In Oracle the client used to define the schema +maybe this is the same problem +That said I think recovery mode may be useful if a cluster admin choses to make it manualFrom what I understand based on previous comments it allows an operator to continue with corrupt editlog or abort +By the way ARAT had a bug +JIRA cleanup +I havent read the thread carefully but Im guessing that this is due to losing data in the disk write cache upon a poweroff +Committed to trunk and +Where are you looking for this list of jira projects you can adminYou are listed in Thrifts Adminstrator and committer roles so you can still admin the project you can see this for yourself by going into the Administer Project linkNote that Jira Groups would be listed beside a Jira account but Jira Roles are not listed +The modules that contain the pluggable transport would be usable as on the broker or just as plain loaded via a simple mechanism on the client +Oliver could you prototype this feature on an own branch first? +Works like a charmThanks for the contribution +It happens on all groovy projects on the one machine new existing generated in eclipse generated from grails anything that has the groovy in this is the first install of the plugin on both boxes both were installed yesterday so there are no older plugins have tried removing the groovy plugin totally from the installed plugins folder then didnt seem to help +Fixed for EAP and EWP +Hibernate doesnt seem to have this problem +This is a bug in the unit test +Looks good to me +Output executing mvn package with +Solution posted as comment to +Closing all resolved tickets from or older +Closing all issues +I dont think we should block other commits on +or perhaps setting the release metadata on plugin installation since it is only local? risky for deploy though only want that if they are actually releasing. +Oh no do not do that pleaseIn fact the SCM plugin is important for me respectively if it would show what exact commit has been analysed not looking for what is after it +I cant replicate this so I need a test was due to lazy associations in Hibernate being done with lazy thanks for the testcase I was barking up the wrong tree without itPlease test for me works! thanks +Applied in r. +Thanks Eugene I can confirm that with the build from the stack trace no longer occurs. +Seems not to break anything also added more unit ll leave the issue open some time though until I got the latest improvements ported to version of the patch fixes the previous problems with not releasing the connection if an exception happens during the instantiation +I just reran my search and the only one left is the one in Configuration that is supposed to be there +Just did a git bisect and this patch seems to make hang +Committed latest changes will update wiki shortly +I suggest adding this as a listed requirement for Perlator +Since the issue is in the PDF it is most likely a problem with the XSLT +Thanks Bryan! Committed revision +Let me know how that works outConfirmed fixed in. +Feel free to reopen the issue if needed. + rtrunk closing all issues that have been in resolved state for months or more without any feedback or update +for branch and for trunkThanks Viola for the patch! +Thanks Todd +Sounds like a good idea to have a patch for +Thanks Jacqueshm every time when i make patches for labels i have trouble with the encoding of the file +I tried to reproduce this issue on the latest code and was unable to cause the failure +This is the zip of all the mapping files +Has anyone found a workaround to this issue? I was trying to think of a way to intercept these to rearrange the query in a hacky way but I havent found anything yet +I have cleared logs then start tomcat tried to close iteration then shutdown the tomcat +I have your test suite read to go in my local checkout +Ill resolve this issue for now though since the fix went in some months ago and we havent seen issues with it AFAIK +Generalizing the Axis IL isnt really the best option were probably going to work on a generic embeddable IL for the next release. +ive uploaded a too +I already notice a problem with this proposal the elements of the pair might well be mutable and since we store references not copies of the objects passed to the constructor we cannot ensure that equals will stay consistent with the cached hashCodeDoes someone see a way to make Pair truly immutable +I think its ready +Correct affected version should take in in +Can you update the patch for the other branches +Cocoon is build right now with debugon please this if you can and report back +Delaying to we need to decide how we will introduce regex in to xwork +is a bug +doesnt apply cleanly on trunk +This version works with Lucene +Another workaround for this would be to index the attributes that youre searching on using the! operator +So the synchronous execution of the task may not existI just checked because of some indirect relation to this problemIm just not clear if all the activationdeactivation tasks are queued to run why would there be a race condition for the serviceRegistration? If we gonna take the calling unregister services out of the Disable task then that would be a race condition for that field +I wonder why these two plugins are not children of one common abstract parent a lot of logic is just duplicated and issues are the same. +Thanks Sushanth +There were two entries +The discussion on didnt make clear if this one is a blocker or not +does it fail with the supplied test case? Or do you have an alternate test case that exhibits this +Can you expand on your comment a bit? Are you saying that this issue is therefore invalid works as designed and that the problem is wholly on the sourceforge library +xsd or +Not sure if it would be possible to leave the situation entries expanded? Call trace opened the config section and selected the refresh cycle to cause the drop down menu to display kept this displayed long enough for atleast one refresh but again it didnt seem to cause any at the moment it does look like it is specific to the svg rendering in the service overview +It would be good to include comments in both explaining that any bug fixes will likely have to be fixed in both placesHi LilyI see what you mean about +Thanks for the plugin. +Not again DamienSimple criteria the size of the view file should be proportionate to the data in a view on initial generation +v adds a debug line +But Len can still have a say +bulk close of all resolved issues. +Yes we need to do more work to test hdfs without mapreduce +Currently there is no API to specify custom artifact resolver in Maven embedder component +I had checked the book but somehow I missed it apparently +Perhaps this is the way that uses a single named index for all repository content but there are reportedly also problems with Hibernate Search using Infinispan Directory inWe cannot hold up the release for these issues so retargeting to +Seems to me that this ticket duplicates +Tests are fixed +Unfortunately this has been left unresolved for two months now and we have since moved to +Ah ok +Enforcing the match with existing fields also helps to avoid typos in field that arbitrary global errors can be registered already without any association with a particular field +I believe that we have found the reproducible case +Updated the patch +Fixed a test error +verified on +I am not doing anything special for ssl traffic +IT added +Do you still need files? It will be incorporated into? Do we still need cygwin after that patch? If not shouldnt all cygpath occurrences be removed +That way we do not break anybodies code get rid of the shape subpackage in the next release and everybody has a heads up +I seePedro apply the patch. +Because it turns of merging using configs and never turns it back on it remains off for future queries run in the same session +Feel free to reopen +I dont use and I dont see a lot of value in it but if its important to you it would be nice to get some help to fix this problem +This is not applicable only to GAE but also other solutions where JMS cant be run. +Checks that the domain name lengh is longer than charactersThis patch overlaps patch for both happening on the same java class +It appears to be a paid for product which is a big hinerance to implementing it +Best regardsPutthibong important note that I think should be made is that when using this with Mapper interfaces without transactions do not get closed was running into issues where I would get IO errors from the database because my connections werent being closed +The new patch includes the file +Thrift execution failed on my local machine looking for usrlocalbinthrift +Nick seems like Ive already mention this +Minor update removing a spurious println +Sorry bout that. +I managed to get a good glassfish run on trunk jboss and tomcat fails with the attached errors +If language to this extent is not included we should file an issuebest regards think we should also implement this in and branches +testcase was not able to reproduce the issue perhaps we miss something hereplease modifie the attached testcase that it shows the issue and AFTER THAT reopen the priority to minor issue is to specific to be a critical or major issue +This SHOULD be handled by the doclitininterceptor +I had nothing in +If something changes in FSHDFS it looks very odd to go and change the corresponding tests in mapredWhy shouldnt HDFS tests and benchmarks just run with a specified version of mapreduce instead of the latest version +There was NPE in testPlease fix +Testing around the Hadoop related bits and class +Hmm +If justify then it is perfectly reasonable that word spacing is adjusted and if the user has specified a wordspacing minmax with no leeway then justification may not workThis seems reasonable as is an advanced setting and so users who change it should be aware that it may have ramifications on justificationAnother way of looking at this is justify is the case where wordspacingpt! normal Again this seems reasonableI am inclined to reject this bug but perhaps we should wait for opinions from other committers first +Issue is solved with last fix. +Wrong also +Kris let me know if it needs tweaking for +Ill treat that as a +Im going to discuss with the OPSJ guys about that +So all issues not directly assigned to an AS release are being closed +From Tom F via email already has my feedback on it +Reason Package uses conflict versionRegardsGert +Thanks again TedBut as you point out it would be nice if the core feature did not show up at all in the list to install +They do already if forkModenever +From my understanding EL is in a state of flux at the moment I believe Norman is waiting on IT to set up a new SVN repository to hold the code which is what I think we need to use for this +the patch +Change made in r +Fixed in trunk +You can see all the debug steps +Thanks for this perfect set of patches David +oops +In I changed the code so the order doesnt really matter much +But while processing SSHafter calling assign to root there is no waiting mechanism to make sure root assignment completed +Thx +If this looks ok then I will break it up into addition of use of in as requested by Sid +rendered obsolete by the decision to remove from the language +Pete can you point me to where we actually get this? Latest webbeans download and as does not list that deployer the supported way for AS + +thanks! +Confirmed closing. +Could you test with version +Got it thanks. +We really need this +Can I run multiple Solr servers in psudo distributed modeSOLR is completely separated from Hadoop and has nothing to do with local vs distrib +Suresh I didnt hear back from you +It is best to fix this issue by creating a link from to +Brilliant! Authentication is ok and works as expected. + +the fix is quite simple just new an instead of using from a static instance +Did you just tell me to go fuck myself +Currently it is expected that no messages will reach +If thats the case than this issue can be closed as invalid. +but I still havent been able to crack the nut of getting tableNames from +If you still encounter problem please reopen +I did a rewrite of the test today and to what I can see it works just fine + can you please look into this? Thanks +Ok this patch has a big problemWell not the patchThe crux is that we shouldnt have added the new config knobs to +So the problem is exactly the same as described in the linked issue +a sample of the report that cannot be test project with a report and a VincentI have answer you on the mailing listIt looks like a bad usage of the reuse report featureLet me know if you still encounter any trouble +Fixed inwbuildrsvn ci CHANGELOG libbuildrjava spec m follow up excluding libraries from war is confusingSending CHANGELOGSending libbuildrjavaSending specjavaTransmitting file dataCommitted revision . +Moving to because this wont get done for +It should be Why it is related to? Typo? +Can you post the stacktrace and some sample code +Affected componentsattributes collapsible expandMode collapsible event column sortOrder extended frozenColumns tree toggleType selectionType treeNode immediate dataTable extended dataGrid etc columns +Here you can see I changed the ruleflowpgroup to name X and ruleflowGroup X but its still saying Guess and Guess which was its previous value prior to when I deleted the original +Unhandled event loop exception +Can you post your replication handler config +I committed a fix for this via +Committed in r. +I missed something Before the whole schedule starts the cache has to be invalid so A actually does some cache writing +If it is problematic or leads to bugs elsewhere then we should reopen this bug and change the behavior +Each digit maps to a code that has a special mening so there must be digitsInternally it will offcause be represented as an int but any presentation must show digitsYes it probably would be possible to use a string but the problem is that the XSD is supplied by the customer so I cant just change itI know it sucks +This issue is missing Affects Version Component or both +So far results of testing with PDF rendering have been positive +The patch is simple and is exercised in other tests so I havent felt the need of another test case +Quite some of them treat all lib +Fixed in r thanks for the patch! +I got a different impression from Jochens comments +Agreed thats better +Chat is implemented +Thank you for the comment +If you can create a patch against there is a reasonable possibility that I could create a Lucene jar and in our test environment that was the source of the graph I provided earlier +In this jira we can probably also log the client id when hit unexpected exceptions in the server +I used a JDK compatible implementation though Thanks +pretty simple so I plan on committing shortly +Any active progress on this? I might want to take a shot at fixing this +Fixed the datetime ones by specifying specific dateformatsOne still errors for me which is bizarre as I dont see how its turning into +As stated in the list discussion about this subclasses of these classes are not supported meaning that this is not a published API and is subject to change at any time +fixed and verified +Hi PiOn hadoop side I am hoping that by using tasks directory we can get the distribution and cleaning for freeFor local pig I think that the idea is good but I question the priority of this feature +Only a minor comment here I see EMPTYARRAY defined twice here once in and the other in can we leave only? I have no further comments any more +Each test has a setup phase which loads the initial UDDI data +notes at top +closing this as duplicate +Ive submitted a JIRA about a fair scheduler now +working on the lists done +It calls on at a time +Another possibility is to just call directly from that would simplify the architecure and improve perf of DDL ops dramaticallyOne possible issue here is concurrency hive code is not completely thread safe +The fix for this has been pulled up to the SOA branch by Tom Baeyens +Inline templates are not meant to be available to other mail activities +Its an issue request to the RP STS +This looks fine to me +And Im pretty sure we have a previous ticket dealing with that problem + +We are now on JDK +In case the upgrade does not resolve your issue feel free to this issue. +Attached patch addresses the minor concerns +I think this is a bad direction +Removing Solr from the module names would give Perhaps should be? We have always made cell extraction confusing +Would the use of ORDER BY guarantee anything or is it just something that might workI dont think it can guarantee contiguous values since other INSERTS can occur from other transactions that would use generated identity values +Looks like these documentation tickets go to you now Laura +only seems to support NTLM v which is not enabled by default on newer Windows servers and administrators dont usually allow using itNTLM autentication works well with JCIFS though +Fixed in be merged with +At least the vanished in EclipseRan the Unit tests and everything seems fine now +Fixed as per jpeachs wishes +backported to with revision +If you have a dev environment handy though it would be nice to verify +uploaded the patch as file for +will submit a new +New Patch Version with Shard Support +The UID problem I have not seen yet. +This patch modified to load immediately rather than lazily so that the features listed in featuresBoot will load when Karaf is started +Thanks and Regard +Clebert can you please comment on the above +pulls down runtime dependencies as well at test timeWhat I tried was confcompilemastertestdefault but apparently it doesnt work +Saluti. +The problem does not exist in CLI +This would have the symptom of connection timeouts lasting roughly double what was specified +submitted patch Vishal Good job thanks! The patch is pretty much good for me +A second option for a version for to Laura Manzanilla +Luke can i know what you mean by support for more sophisticated expressions +This is the svn diff for updating the file and all its dependents vcproj +The booking example is compiling fine for me and its Authenticator still uses +Committed to branch and trunk. +Bumping to chance you could provide a bit more depth in your description? I wouldnt mind trying to tackle this if Tom hasnt finished it +Please see in zip file +Ill close the ticket. +Any chance of a review +So all issues not directly assigned to an AS release are being closed +The tree sounds a lot better to me +Projects using SWF can map their test configuration to our testing configuration. +on patches +I think the fix should be for tapestry to ignore abstract spring beans +I have added the git repository for mahout now so you now have both svn and git +I agree thats the best approachComments on the patch Use methods +This problem should be fixed by now +Can this now be closed +See that will take care of the core cache part of things +As a workaround try making your eclipses workspace directory name without any spaces. +I have attached a that demonstrates this +Program will exitOn ARM with u java XsskThe stack size specified is too small Specify at least kError Could not create the Java Virtual MachineError A fatal exception has occurred +resolve this as duplicated since has a pull request attached +I got diverted temporarily to a different project but now im back dealing with this issue again. +Attaching +So I personally dont mind if we pursue implementation of copying in only and then investigate the current threaded indexingsearch tests and add some if we think somethings missing +I already reviewed this as part of and this looks good +I believe this is complete now +This leaves the demo web application descriptor parsing issue to be resolved +While for I think it very important and need more work +This is being addressed in JSF +Thank you for reporting it. +Significant changes would need to be implemented to even read local port bindings much less remote +Users believes that all attributes present should works +dont forget to add a test in for title +Ive managed to solve the issue by upgrading the jetty components to version So in libweb I updated to and toProbably the websocket specification has changed between jetty releasesAlso some extra documentation on how to use the client libraries specifically for would be handyIt took some time for me to find out that the destination should be in the form topic or queue +They should not have to extract jar files and reverse engineer them +Ive just patched the pom for to see if that helps any I wonder could you try give it another go please and see if this fixes it for you +Nothing can be done on our end until theres a conforming tree in svn for us to work with +Already fixed in Cactus SVN HEAD +javac +I have opened for setting default overwritetrue in. +Thanks Cheolsoo +At least reporting this as an error in revision +Its very difficult and time consuming for someone to review your efforts if were chasing down how a particular change plays into it and then to find out after a half hour of chasing that it doesnt +I am not trying to be a pain in the neck but test HAVE to be considered as citizen by dev +hi thanks for the feedback +They are both request an improved entries speak about bootclasspath +Rebased and merged into the master branch and then into the branch. +to handle the cancel case +In this mode it would do the same thing QE does simply create a Maven Dependency Library that you can use in your project +looks pretty good now so resolving as fixed. +Id say this is an inconsistency issue +Izpack dont support create anymore +There is no redirect to a Hub and back architecture +Plexus is no longer being used on trunk +Tom any update on this for +The log has us just sitting around doing nothing dont know if a flush request has come in +Ill close the issue as invalid +This way service access tokens can be provided to clients that have been authenticated with specific types of authentication event requirements like a strength level of the authentication performed +Thanks for collecting this data Ill take a look +lgtm +I was going to wait until I had something more substantial but Im a bit stuck on the removal of and I didnt want to lose these changes +For example the JSP EL expression language always requires a bean instance it might be useful to configure a utils object as a managed bean in a JSF app for example but this only works if there is a constructor available +reducing priority since so far we have not been able to figure a way to make this code run efficiently +No I am referring to and its confusing because none of this is clear in the comments associated with the code +Im going to close this issue since this isnt really a bug +ear and have them loaded in my container by the EJB +It seems Spring DMs use of a separate thread to handle bundle event is causing acquire to be a problem +The doc for this was already in it just hadnt been generated for the site +I seems that its working with selectors +dims +thanks for the explanation +was a dumb issue anyway +patch to fix in english. +The deal on committing this patch is that no bugs will be reported in lock reservation code until after Thread Manager code cleanup +Thanks +Great stuff. +Was checking just in case +thanks for testing +This looks like it could be moved to the HORNETQ project in JIRA +Reopen if you have problems +Stevesince Juca talked to you this issue please take a little time review this patch if it ok then assign this back to me +One drawback of is that it may increase the load of NN +bq +The patch has been tested without any regressionsmdsumscdebffbfbebfbbb bfada +I have to click on the button to fire the event manually +trunkCommitted revision +thank you for reporting this +Looks good to meI have same fix with more comments inShall we use that +applied the issue now that it has been resolved. +Kelven What are the missing files? Is it fixed nowPedro Can you check in your global settings host key is set to correct IP then kill SSVM and retryAbhinav Can you confirm with new packages on a fresh installation +display the html included in the is an Workaround It could be an issue with the types of these gif images +I did some testing and seem working ok +I will commit it shortly unless someone has an objection +the unit tests are working now. +This is causing tests to erroneously fail +meantfalsethe value is the default +One question though Is it impossible or inefficient to include more details of the changes? I would really like to at least optionally have the same information available as the per db changes feed exposesUpdated patch never exits but does send type and db +is now against trunk rather than is that right? +Thanks for the report I could reproduce it nowShould be fixed with r +There is another issue for that is always the for clients who want more control +and Delete button is not working for the report but its a duplicate all old Resolved issues +I will happily accept patches to make the new collections decorators serializable so long as they come with working tests +but changing the process signal mask after any qpid api has been called will have an undefined effect +Move out to as this is at best a potential performance winfor certain cases +Test case for reporting and providing test case +Thanks for the review Luke +Check state in CLI before we send the kill command and will print out different message if the application is already in terminated state +It would be a couple lines of change but that would also mean that it would disable the solution for and anyone visiting a web page on the AM would not be redirected to the history server once the AM goes away +Let me know if there is any better information I can gathert really offer a lot of help since I dont have HPUX testing resources so its mostly going to be up to you here +That sounds fine Sijie +At least this fixed for me +Did you actually mean changes in +Probably wont make or releases. +I see that is pretty easy +The problem is that IBMs somewhat esoteric rules for translation of Java message files currently requires the files to either be properties files or to extend directlyIBM developers working on Xalan would like to continue to contribute translated messages to Apache but in order to do that we need to have the messages in a form that complies with IBMs rules for message translationAs an alternative Id like to suggest you take a look at the way the definitions of message services and their keys are separated from the localized messages themselves in the and classes +Uploading just removed some commented out code from +This patch needs more work +Another option would be to write a logj that also has a JMX bean +Rave now does not allow to add the gadget +I love this process +The correct and info are being displayed in the monitor page despite the Name Node is Down message. + +Moving it out of release. +no dependencies? no even in servlet api +Just another rebase +I have added a test case that shows the problem to srctestjavaorghibernatetestentitymodemapbasic fix to the previous attachment +Wow that was fast +As I said you are running something that is mot in release +Found no definitive answer me neither so they did not bother telling the world about a fixed axis ordering yet at page CRS defintions there is a table with a few well known CRSurnogcdefcrsOGCCRS WGS ogcdefcrsOGCCRS NADogcdefcrsOGCCRS NAD are all +The build runtime deployment are fixed sorry for the inconvenience. +So for finish to be pressed I think must only be due to the incorrectly created Combo and its weird +Thanks! +Contribution committed +The namenode format operation could be moved out but its only called once so is not really a problem in practice +Perhaps should also throw if theArgument was null +Arent we writing tests to catch such bugs +Updated patch to merge recent changes on trunk +There is no plan for any more releases of AS and no development is going on related to that +I will need to make time to tidy up again +for the patch +Yes the idea is to add the SOA version to the server agent and plugin names +It was impossible for anyone outside of Y! to review this before it got committed to trunk +Now the JSP pages do not create sessions +Main things to note are that a default locationmap was added and an image was moved to get around an odd issue with Cocoon where otherwise Cocoon would report the file as missing even though it was clearly present in the filesystem +Should be fixed by now +Are you trying this on trunk? Because specifically patch on addresses this problem +I also stumbled on this issue +for the second to trunk and branch +I also did some testing and I saw how is slower than approachAlso I didnt find any other Java open source project that uses pooling for +If the caller is making this kind of mistake we should probably just let him know + +Done +The main test class is +alas. +Filed. +Fixes have already been committed and should work in the next release. +Can you please attached your Designer log? also on Windows the stuff has been generated by TEIID and it should be easy enough but it stop during WSDL generation with no other error and the only option is to kill eclipse +When you require we load that into a separate thats a child of the one that bootedNormal class accesses as in your workaround will go through that child classloader and find those required jars +It might be nice to test the revised config rules afterthe patch and also in the future should we change any of the config stuff +We need to mark only because readLine tries to let r alone to be a valid end of line character +The webserver did not take the configuration option but once I set it as a global system variable in Windows it workedNo changes necessary so this bug can be closed +Yoko did a pretty good job of explaining this issue and theres others that track startup time in general so Im marking this Not A Bug. +Please try current trunk and reopen if the problem still exists +Same problem +if this is decided upon might it also be better to normalize the into a join table between Geo and + Id like this too +I looked at the patch and I think it looks very good except a couple of stale comments +It just struck me as something that should probably be addressed at the framework level for the benefit of all. +This is by running mvn jetty as mvn tomcatrun doesnt work with have been thinking about this and I think it would be relatively easy to have gwt and mvc method to access both views or set a default in the initial frontend technology used +Good catch EricWe should fix to set this for +Woops I accidentally uploaded the same patch again heres the one with the fix described above passes +It is still an open question how best to implement atomic and when inXThe spec allows weak atomicity so many avenues are openThe current implementation is lock per place +Sending the patch Sanchit for the contributions +This is all fairly advanced functionality so the user doesnt need to mess with it unless they want to read in messages a particular way +Otherwise all other subclasses should be factored out tooYes all other subclasses should be factored out too +Ive just committed this +Hey ToddLooks like the test added in this patch is racy it failed for me after I looped it for a couple minutes +If there are unresolved bundles present and a new bundle is installed that provides the necessary packages it will automatically refresh those dependent may not be required by the specification but its actually really handy for development +Both problems fixed but no tests yetSending modulessrcjavaorgapachegeronimoschemaCommitted revision +In the meantime lets go for option +Closing this issue because we dont intend on making subclassing usable in the near future +Sry for confusion. +fyi Using attrMap works on Harmony +bulk defer of open issues to + +committed! +Please see logs for more details +I will commit this unless someone objects +If you use Ivy from svn trunk do an update and try again and keep me informed +Ive committed this +Looks good other than the slight niggle we discussed with the filter clear prompt +I think that for to go final it must at least support the format attribute +trunk Now its fixed +what feature? The any content? The issues is closed so can be retrieved by reading the from the context parsing it to a Dom andor String and getting the data out of it +precision scale shoud be added to precision for sign and decimalIn the embedded codepath Only first and third is implemented +changed to closed +My app uses autoload and used to sporadically encounters method not defined errors which presumably has something to do with concurrent autoload during startup +Thanks Ted and Nick for helping me. +Alright one by one +I think the assembly plugin is a bit heavy and that the dependency plugin would be the ideal place for this kind of functionality +On answering the first comment on this patch as this can fix the issue as well +Looks like a duplicate to me but for now I am just linking the issues for information sake +rebased with latest master and updated my github branch +Just a note that I was able to reproduce this +Erwan not sure what you are talking about here +Im sure you intended to include the headers but just for the record could you please indicate that this was so +Read you should know where you are connecting! we do know what server we are going for but in high performance environments a dns may be resolved to multiple ip addresses and we would like to know which one we are currently using since even if there is DNS round robin or alike this doesnt mean all physical instances of an identically named service behave the same way. +for this important patch +The code looks much cleaner now +contains +The instances will share the a single instance which should result in proper connection managementPlease test this out and let me know if everything still works ok +Per the comments this appears to be a bug rather than a problem withStruts. +Updating name to actually +Shall we get a shot to this by + +A question regarding the addition to this basically means even with an updated API any user wanting to make their own custom filter would need to not only add their new filter to the CLASSPATH but also to patch and rebuild their distro to add it? My original intention +more variations from juanof them i like c g or hthough i dont think any is a real winner yet +So with your and my the average is +Im thinking each edit needs to carry two timestamps to solve this issue the user designated one and another that records actual insert time +on Garys patch +I dont like the idea of having these conversion artifacts hanging around when we havent even released the first version of pluginsI waffled as to wheather we should break backwards compatabilityAfter Craig explained it was because of our Designation of Betathere better be a pretty good reason for breaking compatabilityWe promised backwards compatability and I agree that for that is the way it needs to beOnce we go to the etc like Ant then we can adoptan new policy +Test case should be rewritten either using logcxxs Thread or APR thread or fork calls +thanks +Ill take a look +Corrected some minor errors +Aha! OK got it +Groovy formatting now available from the editor. + issue comment +Here is a patch for the problem +I had to add the license header to to make RAT happy +That is how we have significant functionality such quorum journal manager Automatic failover and many other features +Thiruvel this is a forThe schema for describe and describe extended cannot be changedYou can add at the beginningend but the number of columns have to be issue is that now desc extended displays tablepartition parameters in different lines +This is going to be the new behaviour? It can be defined as a best practice to explicitly specify in the definition whether to bind or not although right now it means some definition revising for me +It fixed the problem +Thanks +Reproducible test case attached to of which this appears to be a duplicate. +you mean to know how they are mapped to tables +Thank you. +Also updates mapred protobuf thrift code to take advantage of these changes +Contains a few bugfixes and makes the code less dependent on +go for it Ill be tagging tomorrow so there is time +mnt unzip l and found out that the content inside the ISO does not contain files for The problem seems to come from the RPM build that is for packages RPM build should issue maven command with following optionsmvn P systemvm DnonossCould someone who is familar with RPM build to take a look +So it is meaningless to include a unit test +but what was the reason to exclude arrays? With the current patch one can only use an exception for a throwable +Its on my queue although at the moment I dont have a reasonable estimate for when it will be done +Request for review for subtask available for Please review +This isnt really an issue against Isis Id rather that there was just the parent ticket for now +Need sudo chgrp bval homeapmail run by root on minotaur this depends on the bval group being created +Many thanks for the report +It would be good if you could update the example to use Derby +in one pig script one macro definition which mask a builtin UDF COUNT +Thanks a lot Brian! +Thanks for the patch +Also tested with version and fails to but does work but the conversion is to mention I already did that I tried update in and bits did you tried to process the attachments? As I told you version does not crash but image transformation is low qualitybut versions and just crash the a question why this works with pdfbox versions and below? At first I thought the same things you mentioned but then I noticed that older versions does workAlso I have test this in at least different computers with different OS win and win server Hopes this helps +Patch applied at +The important things Chapter needs to go away +If we anyway wants to do two implementations then the better choice would be root interface derived by both interfaces and have their own implementation +I used to have access but then the machines got reformatted and no longer do +See new issue for a suggestion on how to tackle this +I just committed this +Not really a dupe in the purest sense but Ive broken this down into two separate issues and +Im thinking perhaps do both options +The uses the CXF Endpoint URI cxfrshttphostportpathresourceClassesInterface which is actually working for interfacesBut the error occurs when declaring a cxfrsServer as described is this issueThe logic behind these two declarations is probably different. + +Hi GianmarcoGood to see that you are trying out the aspects of the code +Removed all references to brand in sample job +Sorry if I came across the wrong way but I honestly cant think of any more testing you can do when the original bug was dependant on the name of the class and the producer field +attached same patch with Grant license to ASF for inclusion in ASF works +Just committed it to branch +makes sense to meCheersChris +uniqueVersion trips back to true on child projects in the hierarchy but only when a profile is enabled +Apply attached patch to changeadd links +Thanks for you comments Jiri and Burr +Fix coming in amomentOleg +It seems like reasonable pricing and I hope it becomes successful. +Stu are you planning to multithreadify this then +Generated client stub class +for Rob to decide on +Thanks Junping! I committed the fix to branch. +Please check that the patch was applied as you expected +Closing since the changes have appeared in the Latest Alpha Manuals. +F +I think the patch might need some minor changes +Besides a note in the release note this needs to get into a permanent section in the docs which talks about using EAP with ipv +So if you specify readOnly this may or may not prevent you from modifying data depending on the underlying persistence provider +Added a new unit test in order to test priority fixed the patch file to include the added files instead of including them in a separate zip +This only affects testing +Fix checked in for please test whether it solves this problem too +Number of tests in increased from to +If it can be done with aegis maybe that should be configure to be the default +Nice investigation applied the patch and an updated version of the test +A patch to fix this +This needs to be and closed as Not a Problem +So if possible we should support these constructions +Please check here +One of our users on IRC had this issue only when BUNDLEPATH was set and said it looks like was loading its gems from two different places with that set +Attachment has been added with description Patch for FOP srcjavaorgapachefopimageanalyser +The patch was applied too lateIm leaning towards + +Farah is their a workaround for this +Matthieu I ll generate and a single patch and upload +probably more sensible to do for do agree with Max +Applied at r thanks! +If there are errors in the Producer please do open a new issue as we want to get this all sorted before the release if we can. +except for the aboveAdded username check for the other cases +Note that a working solution for is present in this patch tooI think it is better to submit the solution of as part of that JIRA rather than including it in this patch +CLA send via facsimile +Then if the active action queue is empty we sleep for s and then re process +Jonathon or Hans have you a precise view of how to do it +I have now confirmed that the problem exists with packages created with Drools +I can work on this in the next few days +Is fixed now that is fixed. +defer all issues to +Polish patch to meet patching requirement +This was fixed in JPOX last week +Ben is leading this effort +Patch is against trunk but applies cleanly to branchx as well +Is that failure because of your patch Jimmy +Also thanks to Vinod for review and modifications on the last patch +I may not have access to the for a day or two but when I do Ill let you know how it went +The point is that the groovy class says extends instead of implements +Marking priority as critical so that were sure to fix +I finally got a chance to read this doc nice work +Yeah probably not a big deal +I reviewed the first patchWell need to fix facet modules dynamic range faceting to skip missing values I can do this after you commit this patch +Pure conjecture says that this is associated with use of the native connector which uses apr to bind the socket +Ill publish the patch just after +I will take a look +for V +sorry wrong comment +Is there any chance we can hit that mapserver directly to reproduce the issue +Patch that has this fixedI think I should do a bit more before committing +common component +This also explains why it does not happen after every test restart but only after a number of index segments have been created that trigger the index merger +Also backported to branchx and committed +Thanks Stu! +Thanks for helping out +Note that this is just a debug log message Its expected to catch and log such exceptions but nevertheless proceed without interruption +Without this change the history server takes over seconds to process the existing jobs and after the patch it processes them in around seconds +At the moment the JDK properties support is pretty nasty in that it shuffles around the properties definitions and removes all comments +closed as per Santiagos request. +All Ok Alan +Hello TonyThe patch has been applied at revision r +If the map output data is compressed you dont need compute the checksum at leat not needed per bytesYou can use the uncompressed size as a kind of validationYou may also need to compute two checksums per map output segment one for compressed data and one for uncompressed dataThe reducers can use them to validate the data +Categorizing bugs related primarily to stylesheet elements etc +Not being able to have more than one process in a package is a separate bug than refactoring please create separate issue for that +Mapside cogroup does require all files being grouped to be processed in an ordered fashion +Can you please provide page code +We havent had a deadlock since the st +were on different servers +FYI not so long ago ICNE was changed to subclass there should be no retrying going on anymore +Closing old issues +Lukas please make this a priority and see if you can give me something runnable by Thursday +small quirk +I just committed this +Please go ahead and attach the file +Changes related to this JIRA lead to this failures +It can be replaced or removed and a newer versionadded to the server level lib directory +sorry netbeans formatted the structure +Empty DN in are now handled correctly in Twix decoder +This could include comments on appropriate content presentation and style +We are using the appengine plugin and so the bug you referred to above will most likely resolve this issue when fixed +I say resolve it. +I have basic working patch +Maybe the test can be written without sleeps using +This patch here makes everything happy again +Its a problem with all query cannot have method calls inside the criteria closure you must assign values to parameters outside the closure then use parameters in the closure +Thats a different issue though +Since we are opening a new connection for each call this fix should do it +Adds a target that runs tests contained in testlongMoves to testlongEnables longer tests in test Hanna in IRC the patches look fine +Btw when are you planning on releasing the? And again thanks a lot +Committed a simple change to enclose the in a Debug checkand confirmed that the SANITY warning now disappears from an insane jar buildCommitted to the trunk as revision . +I am not sure whether we should iterate over all properties and trim them before putting them back in the mapAdditionally the term I dont think we need sounds suspiciously like no application will ever use more than KB +sorry about that +Im not sure what can be done here +No entry was added by this commit btw +this one should fix thatadded some garbage entries into the properties file to verify that broken entries get ignored +My Query looks like node +Resolved in trunk rev +Link to related issue in are now complete +Could someone please try this patch? Thanks a lot +the reference was updated to reflect this +Moving to dependencies +Fix a noisey log message in case where recoverlease method is missingTested on cluster wo the api and seems to work fineWill apply this patch to branch and trunk unless objection +However as of revision it looks like youve switched to GMT offset +The Fix that I did on the trunk was to throw a detailed error as to why it failed rather then throwing a null pointer +For example see problem in described +this is definitely a new feature nothing in geoserver or geotools takes care of units for distance filters +Verified on + +In this case I didnt even add any new artifacts I simply fixed a typo in the name of existing artifactsIn retrospect I probably should have gotten consensus before doing this +change tableDesc in the get method to added in two +This has been fixed alongside with the recent changes in close for the release +Patch that updates Avro Javas dependency and plugin versions +The new TAG will be used for next release of EAP +It was identical in behavior apart from this one point +Fixed under r now by Simon Laws +Thanks for catching thisIf it doesnt work for you please attach your sample so I can try to reproduce it +There are remaining issues Solr does not currently do everything at a per segment level as you know but per segment use cases will be very fast and non per segment use cases will be much more NRT tooYou are questioning my whole patch Which I am happy about but Im wondering if you have actually reviewed it if not I wish you would +maven and Continuum dont use the order in the modules list to found the build order all is based on dependencies you attach a sample project that reproduce your issue +Excuse my ignorance for the last two weeks or so but I am working towards a release start of November and I am not in a position to clean up patches that are not unified against the latest code in CVS especially when so many changes have been checked in +aiirc this was fixed in earlier version. +Thanks Vamsi for the patch +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +submitting to jenkinshudson whatever this guy is called did you attach your new patch +Would I deploy straight to it + for the patch +I cant reproduce this you have a deployment the following remove the testseamportlet project reinstall the EPP server import the testseamportlet project example deploy the testseamservlet datasource and the project start the server click the tabI have tested the issue using EPP and EPP +MarkIm trying to reproduce the error with duplicates of the file I uploaded +You are welcome to submit a complete patch with accompanying unit tests +Yes this should be implemented in Java for backend processing of Templates as well. +Hi John Wineclipse SRInstall GEP and using geronimo server +just tried to repro and could not +It is not applicable to Commons as Slide used to have their own fork of which was notsupported by projectI believe this bug no longer relevant as Slide should now be using a stockversion of +Unless its from PM team I wont commit to it +Matt Having the libraries in the and i is okay +The patch looks pretty good to me +build failure is related to not to the patchI just committed this +Cleanup is necessaryDid the original snowball authors agree to license the software under the? Thats what says now +Attached the patch just a one liner really +I still have the exact same issue +Stripped down partially tested +In that case Id be afraid that wed incur more actions on the concurrent map and the random number generator which actually would be slower +Note that is OK in this regard because it only accesses minIdle via the synch +The problem arises from the type alias for boolean which maps to the object class +and since tests pass for me with this Ill commit it in a bit unless I hear an objection + for the updated patch +But yeah something with perhaps Delete it in seems a more appropriate name +re classloader issue is probably worth doing great thank youre what problem eliminating the use of static fine Ill be back +Hi HimanshuI dont have the Hbase source on my eclipse yet so it might take me little more time than usual to upload the batchThanksAnil GuptaSoftware Engineer II Intuit Patch for +built the site and the link works +Ive done several tests and all is looking OKIm resolving this issue +After one action fails the group action can verify this by running a scheduled control action and setting the execution order to have the first in the list +A patch which adds the class to the util package and replaces the serializer with using JAXP Transform +locks have to be acquired for store to do the right thing with rows marked deleted can not do read uncommittedo in immediate case do the lock wait read scanand in deferred case do the lock wait read scan at commitIn both scan cases it is possible that while we are scanning someone might add a duplicate rowthat the can will miss but I think that is ok +Then thats a different issue see needs to be addressed in Maven not in Wagon +The fix might be due to +The formatting here is a little betterCheng wroteSo the connection string looks like jdbcderbymemory or jdbcderbylocalhostmemoryRegarding the clientserver URL should we change the client driver to understand the following URL insteadjdbcderbymemorylocalhostIt looks cleaner to me +Did another test with messages at a time and then it happened at exactly the same total number of messages Does that ring a bell? Is there some list that gets fullMaybe I should try with a much larger message +The class type returned from one of the configuration points is which extends +Moving to +Created factory bean to create and populate a dbcp based connection pool +Wrote some gadgets that use pubsub +set to java and not rely on am now resolving this as I tried war and ear and the entityManager not found error didnt show am closing the issue + thanks +BTW I am not sure to understand what youre saying I found out that work great request and app contexts are active during message delivery and request context is destroyed after message delivery completes +Website up. +Yes it does read continuously during I will continue working on this to get on compaction reads as fast as they should trunk last commit is thing that you should do here is make the can option be configurable at runtime via a +How do you have projects you can in the addremove project wizard? There should only be one ear that you can addremove + +my properties file this should already work! The only requirement is that the nested property path does not contain any null references +I guess our use case is kind of different from yours +Thank you Jing for the review +What is the use case for these exact counts? We do have progress for maps and reducesThis is mainly for automating large scale testing +Thanks to pushing shoving and helping from Devaraj Ive got Grizzly to serve map outputs +fixed +On one node with full drive it showed something like dev Total used and available spaceIt doesnt add up because os reserve some space for diskCould this be a reason +moved to target definitely work for request in +Attachment has been added with description Patch revision with enum and other changes from back to NEW since info was provided and I didnt change back to NEW +Also addresses by using a instead of a List to avoid creating duplicate names +Seems to be okay so far +Patch attached for commentIf desired I can file the HDFS and MAPREDUCE to get the corresponding changes into those projects +iocorenet +We could make the modules completely independent but this would mean that we have some duplicates +Looks like it not taking into effect the padding or edge metrics like the parent class + for patch +Its great +Thanks Alexey! +is created only from and it is used as a wrapper from the real rightOther alternative could be do something on so it uses another class like but that one override get passing the call to the original one +sorry for wasting your time. +An upgrade path also needs to be provided +Please verify and close +Some minor cleanup +You can disable schema validation by setting the system propertyfalseFor the webapp samples running in Tomcat this can be done by adding that property line to the Tomcat conf is a duplicate of because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +We may not find this issue until you have some unuseable space in the partition right +Nope I think you are good +nothing prevents you from controlling what currentSession will return at any given time so control that if you wantin any case use the forums +Fixed in the release. +Rebased merged into the branch and into the master branch. +More testingThese tests run compactions fewer times +Here is the updated patch fixed graph size to maximum px +Best regards GeorgeThe fix looks good +It seems the created from the type defined by the still has EMF behavior instead of SDO spec +According to comments I assume it is IE specificJan could you please update us on what environments is this issue reproducible +One was due to ql testlibs being needed +Part for branchx + Add unit tests for metastore you update the description with the tasks since that gets sent along with emails +Id try delete the pom that refers to it as it might have been changed since you first downloaded it +These are test to reproduce phenomena found at commitedSending javaengineorgapachederbyimpljdbcSending javaengineorgapachederbyimpljdbcSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiTransmitting file dataCommitted revision +Fixed in trunk under revision +That should be trivial to do once we have these changes for the AMF module +After you click save you should see your profile created and a better error message You cannot set your login name because your credentials are managed by the web container not the wiki +Cancelling patch until RB item is addressed +Must be fixed +A short reversion should be fine Ill try to let everyone working on this in geronimo know to expect this +Fix will be there in next releaseThanks +Postponing Sorry +Even though it is a mobile project I am getting this while running in an AIR simulator on my Mac so are problems definitely could be related +I was surprised to see that fails due to this +Thats why strategy and dont the way if you are defining a template they appears in the possible completions as conditions are not evaluated +Patch is ready to commit +Apache Directory studio version has been released. +About null entries we dont store them ATM in any of the backend +One reason for VMEM spikes are forks are u seeing any +updated the authentication settings on for hudson user +Attached is an updated patch proposalThank you very much to all the reviewers the comments were very helpfulThe major difference between this patch and the previous one is that this patchincludes only the RENAME COLUMN statement +Thanks Jonathon Ive applied your fix both to the release and trunk +Benjamin Did the plugin work for you? Can we mark this as fixed or is there still a problem +sh file to each of the brokers bin directories Change activemqhome and activemqbase to valid directory on your system Run start script at top level to launch all brokers +Sample configuration in JAVAOPTS false falseNote security is disabled in this configuration +Can you explain what you mean by dead gossip list and how this prevents truncatebq +This seemed pretty impressive to me +why dont you cache the class? There wouldnt be a leak in that case +Thank Tom for making the change +The fix was not backported. +Patch available +Preparation for Gora release candidate +a jar with the namefurthermore you would have to open a jar file to see to which lib it belongs. +testcase pending +As discussed with Ashish offline +Anyway I think the patch is right because its only explicativeOFFTOPIC Is there any way to discuss issues more dynamically other than the list? +Thanks for your help and clarifications much clearer now +For example JAVAHOME reference can be used in and and several dependent components +Fix is to remove the reference to group in the empty groupingSet alternativeThis doesnt fix the warnings because of the multiple alternatives +Are both apps going to enter the default timezone to the db? apps are going to be entering the preferredtimezone information +The store doesnt disappear on close only connection is closed and the store is reopened on the next poll run all listeners remain +The remain failures I couldnt figure out after a quick am pushing this to the next release because the templates need updating and its also unclear if this prototype will be further maintained since equivalent functionality iswill be offered by the profile service is still in docsexamples so we can see if there is still interest for it +I looked into this before and it must have something to do with the rendering engine +Upping priority to Critical +Currently there is support for the old Tiles FrameworkProviding a patch to add Tiles support would speed that up + for the latest version of this patch +You can see some html and javascript for render reportregardsChatree are lots of examples nowMany thanks Hans +I suppose that in case of few occurences all the terms can be loaded in a single read but when the index contains more terms the enumerator probably comes back to the reader to get data progressively +I accidentally made a instead of a. +AddingThanks +closing as issues are fixed in +As each retry was on the order of an hour the application didnt completely fail until all objects had reported the same error hours later +This happens when server is shutdown or killed and the reconnecting proxy that Teiid uses for admin connections tries to to server +Removed the test from and trunk. +backporting change from trunk to branchmjdksvn commitSendingSending javaengineorgapachederbyiapitypesSending javaengineorgapachederbyiapitypesSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +It is done once after modifying the archives rather than times in teh middle +This looks like a different concurrentmodificationexception error than before +I think this was because I failed to use the proper JDBC URL +Lukas can this issue be closed of the work required is covered by the new jira you opened +Is it possible to expand on this? I Dont understand what it adds +Thanks Rajesh +If so we should move this out +Gregory can you double check this works for you now +Any update on this issue? We are experiencing this problem with and it would be great to have it fixed. +Im going to need a test case with current CVS please or close the bug if this is no longer a seam pay example sconvertEnum is fine +I was not sure if the new code broke the test itself or the feature it is testing +John is right +This script has been updated to point to axis version. +This would create an odd use case +Though I may have done it in one of the other iterations and gotten fed up and did things poorly until we were well cemented +This would have the advantage over the shell R commands in that name nodes syncs from multiple clients are effectively batchedFor this advantage does the program need to be mapreduce? Client just needs to invoke multiple threads +Codec methods are currently applied to not +tried again with latest and chrome beta everything worked fine +bq +PHP version still needs to be fixed +Actually my comment about the naming collision might be slightly wrong +Also as I am new to GIT the upload taking longer for me to figure out than I expected +bulk defer to +Changed component +most stemmers use length of the string syllable count +I had to upgrade the os version to add install any package +I am thinking about sync delete for the failure case +The generated client initializers for declarative control clients now enforce on control fields runtime +Committed revision +Final patch. +fixed in revision Vladimir please verify that it resolves yoursorry for my confusing comments in the issue +I dont think that using peekgetDirect is a good idea it could cause cache consistency issue going through the interceptor chain is needed most of the time +patch including unit test which should fix the issue +I found to do an automated check of this and I have attached what I have found +Am slowly working on the stack of issues +Heres the patch to tomcat with the following fixes Adjust to the new sample structure copy stax and wstx jars to commonlib per JIRA +Committed revision . +Its in Im not wrong only color is different +please ping us and we can reopen it if you see it in or above +If in different shared network might be accessible then admin needs to make sure subnets in different shared network are not overlappedin future release CS needs to warn admin the subnets are overlapped admin can make decision if they want to continueNote if two shared networks are accessible and they use the overlapped subnet it is very hard to recover some need to be stopped +Im not sure how JON would handle this but I would have expected it to make a query as soon as you had inventoried it +Ill do another test on F shortly. +It feels similar to the Session situation we have in JackrabbitIll follow up on about some design ideas for the refactoring that would be needed for this +The intention is to provide an additional safeguard against administrative error +If I were to use mySQL Idsimply go and clean the database after running the test and Id have a freshsetup +BTW my findings match Garys +Attached a snap of Xcode +It would be possible to do this if you did the redirect on the client side with +one more unique application of Similarity component to the problem of semantic filtering of speech recognition and tests passed +Sorry for late reply I took a short Christmas breakI hope you had a good Christmas tooFew things I metioned incorrectly areI am using SP of Weblogic and I am using StrutsIn some forms which extend the I am submitting the forms using where I am appending the URL such as methodSubmitAfter submitting the form as mentioned above it displays Page can not be displayed and then if I refresh the browser it is going to the actual pageWe never had this problem on physical machinesRecently we started testing the application on virtual machines and getting this problemI believe I provided the required informationWaiting for your responseThanks in advanceJani +Closing since this is an old issue and we havent seen it recently. +Yongqiang Can you please hold off on committing while I take another look? Thanks +See there for the real resolution +committed to head and soa branch. +Denis please read the component reference for Zone +This fix will actually reject case like import location +Unless the jobToken is required how about sending an empty byte buffer to ensure the additional service is initializedConsidering how the is tied into the as well as for the port information I think this should just be left as it is +Seems to be specific issue +I added a initial section discussing the migration and pointing the the wiki for information on running and writing testsMay not be the final solution but while things are in flux I think its easier to point to the wiki +Screenshot of the schema schema updated the cache +We need to introduce policy for this +Perhaps check into runtimedata contamination in the clean build state +I could start the system on the command line and then use jps v to verify that the right permissions were to open while this is verified manuallyWhat would need to be done to allow Hadoop to work with +Looks reasonableI committed this +Do you have an example of one of the functions we are using that is not found in another JVM +Indeed if the primary key is set up to be then it just wont work +The problem is that you will always find use cases not covered in the Query API and if we add each and every of it it bloats our Query API since we somehow JPA Criteria API +However were not all the way yetSome files dont have any license information +Patch address most of the comments +The Jackson update has been pulled from upstream +I assume this is a bug in the driver as it works in other RDBMSs +You forgot to document these properties checkstyle checkstyle checkstyle checkstyle I think well release all these plugins before the beta and we are likely to forgot to update these properties before we release the dashboard dont find that its a proplem to anticipate the changesVincent you are the RM WDYT +Had the same problem with jruby and JTDS MS SQL Server JDBC Driver on Rails seems to work for me +ThilinaCan we close this one nowthanks like this is fixed alreadythansk +Will remove it in a jiffy +Can you confirm if you had to add that single line of change in Oozie for a lot of tests to pass? Also could you educate me on what line this was so that I could try to run these +For now you can see if this is sufficient for you +Reopened for documentation +approved +Please attach sample project that you have problem with +Rebuilt patch after reverting and which are not relevant to this patch and should not have been inconsistent with the trunk +Bob do you have that sitting around somewhere +Really strangeThe code that I downloaded and built works greatI just downloaded the release it has the same problemAt least for me recreating this is pretty easy +All resources from Ajax block are loaded directly +By now its quite out of date so the work would probably need to be redone +This test has been added to the exclude list + +really dont see why we have to make this so ut realize that there are people out there with different think letting the user decide using a configurationoption like the one i have proposed in the issue is the best solution +I tested this patch by applying this with the patch posted on and confirmed manually that the NN protocols which still use work again with these patches +Actually the spec should said that this annotation is dedicated to EJBQL queriesWell have an other annotation for SQL curious what core movement are we waiting foris and not enough? +fieldName causer type +then the queues were routed to wrong what caused the message lossI dont think this was on any released version hence I dont think its needed to have this on release notes +But Id start with a discussion on the ml +My EAR file is more than M! Cant attach it in JIRA and may be too much by email +Testing now with +So well have to find a different solution +So if we call some java service and we get a lets say containing the real cause in some deeper level of the exception cause stack then we are unlikely to appropriately react on thatSo we needed to store the exception object first including all the causes to then use +But I guess either Ive completely misunderstood how to use an appserver or its another one of those areas where the Sun spec mostly solves theoretical issues and leaves the actual implementation and real world issues to the well +bulk close of resolved issues as part of JIRA housekeeping for release. +Now that I released AMQ twice it wouldnt hurt to align the two build system +Looks good to me +there may be other patches to port to make this happen +Ive also added pace to the by getting ensuring jobs do not as aggressively as they make progress +That would be nice +Looks like the change fell through the documentation cracks back in +Added ganglia for your detailed commentBut I have a question for the issue +Yes I think that would be a considerable improvement over the current behaviorand would certainly address this issue appropriatelyCan you attach your proposed patch which creates the message as you suggest aboveAlso you should mark this issue as assigned to you +I vote for being able to run GWT and Spring MVC in the same app +It works so far +Maybe just leave alone on this issue! I dont think it should be mixed in with this editing stuffI dont see a need for a base class shared between the two anyway +With the new annotation Im testing now we should be able to make this test a little easier to handle we can keep the same space goals with without assertion checking. +Can you upload the patch which reverts changes that will also trigger Hive QA run +Yeah and this morning i made the same change to Nightly build as well +Ive attached the patch +Logged In YES useridI think it is a misuse of the IP example the RMI server started bu needs to determine the IP address from the hostname +The motivation is to keep unit tests focused on testing one thing at a time so it will be easier to find what failed when there is a failure +Maybe something in the neighborhood of minutes would make more sense? Thoughts? In the test after restarting the two recommend adding a waitActive to make sure that the have finished restartingregistered with the NN before proceeding on with the rest of the test +Hi I have reverted the patch as the test is not passing +Archetype that that includes hibernate security user management file uploading etc +Here is an example of trace I get when trying to build Apache Directory Server +Private list removed +a note to RFC and may This issue is VSLDAP relevant +Ah thats good info +Simon this was done for a purpose if my understanding correct then those libraries required only for analysis but not as a direct dependencies of a pluginAlexandre is it possible to provide minimalistic example of project to reproduce this problem +But yes the basic problem seems to exist +See for details on global blacklisting of tasktrackers. +Patch Apply +Committed to trunk and +The way we package rpm will work on way we build rpm today works for +Itll become a circular dependency +just committed your patch Thanks Suraj! +True but unless this is a tight inner loop the performance isnt worth the increase in mental complexity of having a loop body +It probably makes sense to start from one of Hosss original patches or even from scratchThat was said before a lot more work was done +I think and resolved this in +java and + +Thank you. +Seems like svn update +the figurative expression for this in English is a hornets nest +Attaching svn patch +Deferring post console work +Mass transition all resolved issue that did not see any further comment in the last month to closed status +All Designer prefs should be able to be exported and imported now +Will take a for parser error issues and touching up of error message strings +I will let you know my findings +Will see with Jackrabbit the reasons for this. +link to the amq issue that resolves applied in r ready for +From your view using the tool namespace incorrectly caused the problem +Obviously I didnt try everything but NEXT was important the structure of the FST doesnt yield to easy local deltas its not easily separable and pointers typically jump all overbq +Iowno more endless +Sonar was basically running fine after the upgrade except for the delta feature +groupId is whilst it was statcvs in previous far as I know the id or groupId has never been changed +But when I add this attribute the exception is not thrownIs this just a documentation Bug +Jaikiran yes correct +fixed in r snapshot deployed +once Andrews comment is for the review Andrew and ColinNew patch addressing the comments +Jerry do you have time to do this as part of the conversion of DRM to use Cache +Just looked at the MH code +This improvement request has been open for over a year and has stepped up to implement it +verified on +Ill resubmit once Ive fixed it +At present I have taken the simplest route of just hard coding the feature into the +A CRUD framework that the end user is neither going to be expecting nor familiar with +er works with zero modifications. +Good point about cloning making the change on the curried closure rather than touching the parent. +Attaching simple test case +just hide it as an implementation +Eliminates assumption in tracing that no two attempts of the same task can run on the same node even if we allow two attempts of the same task to run on the same node we need globally synchronized clocks to disambiguate them +Looks like this was committed +Triaged for assigned normal urgency. +yeah Ill generate a patch but its not necessary to apply it until after the import +The reason that I want to see a statement id is to cover the case when the output contains plans for multiple queries +The whole chain of all nodes from child up to the rootAh right +Thanks Grant +Actually I did test it against the latest Maven snapshots +Somewhere along the line it seems fixed so I agree close it. +Hope this time the changes make it to a bigger platform +The existing mappingschema is clearly incorrect +Patch v + +I can agree that the class is just experience the same issue as +Previous test fail is caused by my wrong configuration +Patch applied +I take it back this reproduces reliably +Sure no have included a second patch which is related but exposes the configuration property of maximum to the connection factory +The backend for objects is called quaereobjects using that naming the should be called quaerejpa +Few interesting things in this patchAdd Transaction form uses JSON for responseJavascript specific to form is loaded at run timeAbove things may seen un necessary But they are good POCThis also acts as example for calling service from a Javascript +If there is a decommission command there should be a commissionactivate as well +now contains some new testcases to test the behavior of listiterators of sublists +Tenir sexe sembla un calc de langls +has been extended to include role and enabled as required by spring security +Does the above workI thought about this solution too +for the patch +Lets call it fixed and deal with any additional bugs as separate reports. +More specifically the problem is that the selection is calculated differently based on whether CTRLClick is used vs F +Should be verified in gatein master and reopened if still valid +Screenshot from JBT + +That said I would settle for a simple configuration option which enables the replacement of invalid characters by white spaces if they cannot be replaced by XML would only be a last resort +Unfortunately I am not familiar with Cassandras needs and its code so the pool was implemented based on a brief description of the problem and ideal behaviorIt was a fun exercise for a long weekend +For the sake of completeness if date fields in db have time values the between clause will exclude the To date being it something like as upper limit +Usually in usage descriptions square brackets mean an optional argument and the diamond symbols mean something that is mandatory +I guess we can create a new tracker for that and close this one +Not yet but Im working on it and soon as its ready I will attach it here +i found an issue with Scheduler that prevents it from removing tasks correctly which might be causing these memory leakstake a look at and give the patch a try +It failed the first time but after replacing the +Thoughts +I wont be back porting this thoughYou can try out trunk or wait till I release +Now with the fix volume copied to staging secondary storage during migration gets deleted from staging storage after successful volume migrationSo closing the bug. +Regarding an automated tool The javadoc tool catches these kinds of problems I think +Testing along with that Ran all the HA tests and all are passing +C generator doesnt create constructor right nowAnd I dont know much about other languages +This part is now obsolete given Kurts actual deployment is made through the deploy task which includes everything within the processDefinition directory. +In FOP we normally use dpi +Updated patch should fix the test failures +Ill use on of those in the future +Thanks Ivan +My suggestion is to make the test passed in case of RETURNACCESSDENIED and print a warning message +This would be better solved with but maybe we also want a quick fix just in case +The docs are pretty limited to the exposed lib API only +Sent a mail to their mailing list a while ago in reference to a similar request +Resolving since this must have been fixed months ago now. + +Instead I could add show tables where name like testhiwtable +Patch adds one line change to to make the heartbeat interval a smooth function instead of a step function +In my situation I do actually have to use this feature plenty of times because Im using my own Axis deployer and deploy a service type that needs to be updated frequently +Expression values can be injected directly into Java delegates less code +Thanks to Hitesh +Would need to create a new method name and deprecate the old one +You have to follow the tutorial from the beginning +Refer to the Enterprise Application Platform Installation Guide for instructionsThe Installation Guide instructs users to install the files into the directory +is this still an issue? +Do we want to fix them as well +probably +I dont see any reason to remove typeName +Also it is indeed possible to get with a tiny heap +My own initial focus is going to be and +Im using the jar which are given in above im trying to run the same application in struts its giving problem for small file uploadIf file size is small it not uploading +The commit was done to by mistakeNow the connection profile properties are automatically added to when new project is created +Wonderful thanks for taking the time to do it LarsIm not too worried about it seems like a weird use case but it looks like its easily fixedAbout thats more important +Attaching patch +Suggestion Make as a static inner class of the test +We should potentially have the ability to start with a given default config this issue needs JIRAd and ideally the option for taking a config file to start the embedded broker +In issue page someone tells it does not prevent James to come upCan you tell us more why the above WARN message is preventing James launching? Maybe its not coming up for some other reason? You have more logs for us to look at? Ask on dev list and add log there rather than here +It seems we also need a unit test for testing fencing for log split renaming +I also ran a selection of general HDFS tests and they passThe patch does introduce some more in places like the balancer where we are currently doing some custom RPC connections to will file a followup JIRA for that when this is committed +Complete patch +looks good +since yesterday evening +It almost sounds like you want to turn this into something like rsync +Thanks +defer all X issues to +Markus definitely +If its in a different one then you do indeed have to specify it +Looks like RI contradicts with specification +Verified on EAP checked lib directories and all were removed. +This patch fixes the to handle both the old style and new style names +We want to use the exact same +Attached patch v which fixes Joels comments Formatting issue Scaladoc issue Changed shutdownNow to shutdown But leaves daemon as the defualt +In this image Rob suggested we move the Package Name field to the bottom with a separator +Fixed +Actually just thought since theres a special in Castor JDO already why not use itSo debugging some more now +Thanks +Oops Ive missed your patch sorryPlease untabify the patch +Not sure what i was seeing before. +Closing it as Not a Problem +HiI tried to compress this image by jpeg compression and it was reduced to kB instead of kB which reasonable size +Allan it comes down to pestering a committer to reviewing and committing or just hoping they look at itWhich I just did +How are you passing the document to the parser? In particular are you using an with a valid system identifier? A relative URI is relative to a base URI so if you havent properly established one then that could be the problem +regards. +good +The internal links should be reviewedProbably the introduction can be left of as it repeats the overview document + thanks +But for libjars every time hadoop will ship jars and treat them as new distributed entriesSo we can use to avoid exhausting distributed cache +on fixes the double assignment issues +Of course the button should be Browse +Your comments are very helpful +I think you will have better results when we fix +I do plan to add user to all namespace functions but the current task is simply getting a user associated with a file +At present the only way to do this in is to embed the URL into the jboss dd inside the EAR +I just have a large one moreover full offoI presumed that by reading the code pinpointed by the stack trace you couldguess in which conditions such occurs +Uh oh +Sorry about the delay it completely slipped my mind +Changed s to instead of for the input Nicholas +I think this is only going to work for reading but not writingWe dont have any unit tests for CS but there are tests in examplespigtest where we usually try to test the copy CF scenario for various types +Together this gives the major status of the clusterSo for a cluster that had a with one collection shards and nodes for each shard you might imagine seeing circles arranged in a square under the heading collection +Moved to MAPREDUCE since this is a problem with the MR and not YARN +on formar +Thanks Wei +No longer an issue. +add EAP runtime +BertrandThe class java doc specifies what is been done in this case +Seems similar to issue we looked at last week. +rubuntu smoke just to fix the again needed to set the component to DRLVM +Nope this still seems to work +The warning is generated from a method that is deprecated but left in for legacy support +Does it apply for flush as well +I mean both +I gonna add another patch to include commands into the camel feature +If the diff is empty it can skip uploading the patch and creating the rb +Thanks Shai +Hey Gary +Reporters of these issues please verify the fixes and help close these the submitted fix address the same issue in? If not then I dont think this can be flagged as fixed +Is there anything at the moment you need help with +can you send me a working example iplemented +This seems to be across the board with your install +We suspected that the console was affecting the behavior in a manner similar to your assessment +Whoops some tests fail +The test is necessary by itself +Okay sorry to keep spamming this issue but Im debugging through stuff as I go +Hi Geoffrey can you also translate may not be null or would bemag niet null of leeg zijnFeel free to send me a mail if new messages would be added AT +The reason you see these at the end is because retries documents that fail if there is any possibility that it could succeed when tried again +For some reason it no longer knows the changenumber +Ok thanks for confirming that +False alarm! We found the source of the problem and it was our fault +New version of the replacing the old version +If my machine has GB I might want to allow a MB container to fit in between a GB one and a GB one but might not want to allow MB containers +On iOS we could ship a Cagetory that wires up the Client SDK with the iOS hooks for REAL work is tracekd in here +slight change to include supporting the hostname tested fine +Though did the ability to deploy using a symlink pathtomyRAILSROOT to work again +Please only open issues in JIRA if you are sure that there is actually a bug +it is definitely prefs moduletest was excluded while it may fail in the default configuration +Using means no parallelization +We simply cant apply it because you didnt grant ASF the license +I am open to improving this but I think the status quo is Good Enough for a start +We cannot perfectly impersonate old pulsing format yet the old format divided metadata block as inlined bytes and wrapped bytes so when the term dict reader reads the length of metadata block it is actually the length of inlined block +Test renamed from test to by RobbieChanges made as discussed cheers Keith + by apache whirr +Also it leaves room for lots of comments for further development down the road +Attached a patch to wait for memstore to be flushed before we move on +Done in revision +Committed to trunk +you are kidding right? Take a look at every Java API out there they all make heavy use of interfaces because it makes API maintenance easierNever mind +thanks for commentsfixed and attached +The piece I have left to do is a capture of an example shell session +Thanks Tim +Im hoping this will make it into that release +This adds a method to an interface which is an incompatible change and will break implementers +I dont think that a complete list of inconsistencies ever made it to the ML please correct me if Im mistakenShould this really be resolved before? If it is Id propose to postpone it to +Thanks Koji! +This issue has been fixed and released as part of release +I have also attached which is the output of the following large data run with debugging changes time java true true The large data suite is not finished yet and hence is not the complete run but it has debugging info to get an idea of when setup and teardown get run +The same policy used in should be applied to features. +It was that removed those properties from +The target is validly unresolved in this case as the service is not available in the composite that holds c +attached draft of release notes with upgrade is now in the sites migration directory not yet linked from the closed as resolved and released +No need for an emergency release if nobody is experiencing problemsI havent used volatile since the times when I was programming in C and assemblerand I dont remember having Java threading issues because of thatStill Ill keep it in mind and silently put some volatile declarations intocheers Roland +Owen thinks we should consider it for as well +Tested with and latest +Alan can you take a look at this? is this by any chance alleviate by your other buffering changes +This is also a bug in reopening +here are patches for each of the s a quickstart for demonstrating the look ok +New test for the specific fix +Its wrong at the moment because when we perform the match in we first the first endpoint that matches based on interface and policy +Thx for the report Andrzej +But the chances of ever having to try more than twice are even smaller +Thanks +I am afraid this is getting confusing +What doyou think guys? Am I being too lazy and unimaginativeOleg +Closing all tickets +There are parties here +Agreed +Correct resolution. +Required values are in the constructor +David and Daniel did you agree on the proper solution for this? This is marked as a major priority newcomer bug but this newcomer is unsure about the proper solution +What about pages like Who is using Apache Foo? Apache Foo in the press Related projects and other similar pages with links outside? Not many projects are doing those or have them on a project wiki instead of the official site but I think we should allow such cases as they can be quite valuable to the project and the communityOr perhaps we should limit the scope of this issue to just Thank you pages and deal with the above cases in a separate issue +Checking +Pull it back in if you think different. +I was behind a proxy but rather than tell me it can not access the network it tells me Failed to load Wizard. +Posthumous yes please. +The debug comment actually says the wrong thing +Should I include this in upcoming version? What do you think +Ive attached the patch +The problem is solved with snapshotThanks! + +In general processing will not be as efficient as it could be since we are not saving intermediate results for example if one side of a join is and requires a Teiid sort we will need to perform the sort with each processing run +I just found using the validation with the xsd parser to be a bit cleaner but if its not as functional lets get rid of it +I do not think this patch has a fix for that +I just tested the tests on Xcode I do not see this error +I think test inheritance should be made illegal makes it very hard to debug test failures like this + demonstrating the issue +We can open a new issue for the older branches. +The patch looks good +Found a problem in artifact distribution logic in and fixed in the latest EC image +Ill look into it today but if we cant fix it quickly then I guess we can disable and file a followup jira +This will avoid adding me specific settings to the project Eclipse should show you a quick fix for that +Hotdog +studiopcachebinary +It seems that only checks the first warnings since the javac args include Xmaxwarns +trying hudson again +There are also more tests included +OK this might be misguided on my part as the logic does actually differ +Please check it applied as expected +Is there a particular reason why META has the region as key instead of tablehaving the regions inside the table row should give the transaction on the table modification for free right +Hi NicolasI submitted a patch for reviewCould you verify please +This is actually the same NPE i got in trying to build project and throw NPE but works fineIm a little stuck here without being able to build the jar I wanna be testing +Create a publisher which auto creates the destination +I note that there is no change to do to the binary protocol since it has always allowed nulls they were just not correctly handled server side +It seems to be a general enough problem +parsePoint in why use as the separatorA comma is more user friendly spaces are often already used as delimiters in quite a few placesWhy did you replace more optimized code that was already written in parsePoint with less optimized code +However I am concerned that performance will be so slow as to make it difficult to use +Attach a simple patch follow Sijies comment +Committing. +One more update coming +Demo works on Apache +A fix has also been made to get to deal with events which begin before the last event but finish after the last event +Closing this issue since it would only couple the and instead to decouple this for the new direction that is being explored in. +Postponed we wont need those classes in shared until the server does not use them +We cannot safely get rid of mBytes until the setter is completely taken out of this class +Clint Please review +Fixed on trunk +Im not referring to removing D CSS but I recommend removing the D CSS +I didnt consider the fact that it would still not work when connecting to a JMX server using the standard RMI connector +Emitter doesnt duplicate prefix for segment registers access any more +This version of the patch addresses a Hive issue with picking up the right patch includes fixes for JSON tests +OK to add + has been excluded from the first resolved issues for already released versions +Patch committed revf +If I execute the command manually from the command line it executes correctly +Why is that there +in rev +I mistakenly requested to be applied against the +Looks like youre trying to use the hash syntax on a runtime +classify and set fix close of issues +Hitting this issue again +Package name will be changed as appropriate for tests excluded as FIXME The following tests require JSSE provider may be commented outAt least it passed on me winXP. +Apologies for having missed this +Thank you for your answers Amareshwari +I havent committed that fix to CVS yet as I have done no extensive testing about possible and dont have time for that right now +A new property is available if the user wants to generate a script compatible with ant. +isdeployed is missing at server boot and process it as an undeploy +The is not used by anything so I agree that it should be removed at this time +Benoit thanks for reporting +Fixed in trunk +POSTPONED SPEC Testabletrue Section following methods of the must provide the functionality defined by the Servlet Specification get getMethod is getSession is is and is +Its bigger than I thought +seems to be a good option as default delimiter +I dont use the New bean Definition wizard much and you shouldnt really have to create a new bean definition just to add a file to the IDE right? The main use case is copying a file not creating a bean +Thats why I ask you to fill description because I dont want to search for it it should be on top of everything +I dont think this is a problem with XercesRegardsDavid +After applying the patch you must chmod x srccontrib before building the first time +Thanks JuhaniRev dcfacffbcaacbNote I am going to leave it up to RM whether to backport this to +There is the same problem with inplace input +analyzing compaction performance the compaction algorithm maybe some block cache settings etc +Oy +Perhaps you have a slightly different set up than I do +So if you need to test both patches for and this issue then you can apply both the patches +Move startstop scripts from PREFIXbin to PREFIXsbin +It would be great if somebody else couldconfirm that reproduces the problem for themIll try to convert into a form which issuitable for inclusion as a new test case in oneof our junit test suites perhaps +Im staying in touch with them and will relay back any new info on this +Actually probably shouldnt preserve it but should +Yep agreed +Should everyone like these comps I will put the new LF on the page in and we can make adjustments from there +I forgot to mention the following +What if we had Cheyenne shrink them just a touch so that they fit into the concentric circles from ? Cheyenne would you be willing to experiment with that combination for us if thats cool with James? Hopefully we can produce something that is a favorite of both ALR and AslakWell eyes in rv are only very slightly different from rv +I just committed this +Committed into SVN trunk with r and into the branch with r +Coming out of that either AS will be fairly soon on the roadmap or there should be more releases in JIRA to which I can assign this +Bard this is done and in trunk right +Filed a jira +Such custom properties should also have the ability to be overridden +this should only take ms +I see the job in question now running +Would you be able to do something similar +Surefire wants it renamed to Test +Right? +Patches for some tests are attached +Searching through the Broker code doesnt reveal any printlns that are active so maybe check on your client side +If any of these need fixing urgently feel free to keep bugging me until theyre fixed +open a new jira to address following commentsCreated jiras +Yeah good idea Christian +This patch and HTML file are based on Sateeshs list +In fact we recommend that the the underlying cache not be configured for expiry at all +Applied modified version of patch in rev There are several problems +James per our conversation if you see other issues please or file another issue with more details +Committed revision . +While running the build against Apache on Solaris one of our IT guys found that aclocal is looking for perl at usrlocalbinperl +Ignore the strikeout +I reopened issue to track this change +regularperiodicalSorry translation mistake. +Ill try it again +putting as blocker for beta since if the plugin gets renamed it should be done before the first release of it +Dont know how to patchSo I just attach for has been fixed in the course of thank you for your contribution anyways! + +Thanks Jozef! I must have been blind +Verified with Developer StudioVersion Build id Build date +bulk close of all resolved issues prior to release. +is resovled a while back and Pig starts shipping pigunit maven artifacts +Comparing prior art shows that both and send type information for each execution but I cannot find a reason for this other than historical accident +Change made as suggested. +why dont you explain what we get from compiling private as package private? As it is atm the class is at last correct from a Java currently destroys the Java class model by ignoring private scope +Thanks Lewis maybe you can assign this issues to me +not sure to upload patches so i added this patch revision +AS the is deployed on the server we have access to the system property +Add test cases +Done +trying to find out when this broke I just tried with EA and the problem did not occur +I agree is the preferred solution IMO +Hi MaxYou are right the failure was not because of the patch +You can easily reproduce from the command line by trying something like +Released. +It just doesnt work properly +The params are passed to a controller action as a url request from the openid provider +This patch converts the following packages to Java typesmost of the packages under The patch introduces a temporary lint warning in the build output +Likewise and +Assume closed as resolved and released +Rework doneUnit Test AddedUpmerged to last commit of Mar th I decided not handle the case where corruptfiles is combined with another options +Why not just setting the timeout to ? It will make the thread to exit immediately when theres no connection to handle +Im attaching the sld and a your images were not internet accessible? Thats the only case where I can imagine it not working since will give GE the link to the image as is +Can this be achieved by using provider selection mechanism +JarekWith regards to the attached patch +Its harmful and I can certainly think of alternate taxonomy implementations +since test still this is not the correct issue for this +Perhaps sayPresuming the database owner granted the task role to a user that user be allowed to set the role as shown and would have all the privileges granted to the task updater and reader roles +Attached a patch for review that implements this by modifying to take an additional optional parameter before +Also it gave me time to write a unit test for +After investigated a bit it does not make sense to package system dependencies +Added top of O +Might be a side affect from m going to chalk this up to the lack of a monitor logger configuration when I ran that test. +Hi Len I dont really understand the title +Sorry AntonyI cannot match your patch to the code I really need a more context aware patch based on svn not git +Tim thanks for the thoughtful and detailed test cases they are much appreciated +Done in branches at +I did some tests with Netty and I dont get at all the Kmsgs you getRunning the exact same client with a ms delay waiting for incoming messages to be there threads I top at msgs more than MINA +Thanks for all your added your patch in so this NPE wont appear in future even if the property isnt set. +Created already examples for level security external vs internal +With you change the thread will block until another thread calls the resume stuff +Proposed patch is accepted should be released is June +I dont think it was ever expected that this would make it into or +Looks good Yonik +I dont think theres a quick fix here pushing to +Closed upon release of. +Which version contains support? Which parameter? + +these error are not shown on JBDS file a JBDS Jira for that +In order to optimize the index size I analyzed the payloads and came to the conclution that or so had the boost f a clear majority of the boosts +Note that this problem is related to + +This fixes the described problem and also one other +The files need in the configuration +viewStableBranchjobseamtestReport +My opinion cluster without DNSIP address should be used without the hostname cluster on xenvmware without DNS Same as above +On Linux we should be able to do this as part of etcsecurity +While the sleep itself is one part of the problem the more compelling case for having a separate thread would be the case where the is slow and the shuffle scheduler is stuck because the actual call to get is taking a long time to return +Details fixed number of cleared violations the widget will show the On new code and On old code lines only if the number if different from +This issue is fixed and released as part of release +So even isolating the queries to a single DC you can still get crazy split sizes based on the ranges from the other datacenters +The log output is fine as there is an issue with the password used +If we decide to introduce a constants interfaceclass I think it should be named and be located under +and get DAV client stuff at the same time +Marking this as closed for now +You are asked which connection you want to use to follow a specific referral URL this way you have full control regarding encryption and authentication options when following referrals +Just asserting whether it is null doesnt make senseEspecially for async publish youd better concurrently async publish several messages and check their message seq id are right and same as the message seq id received by subscriber +Dan Allen on twitterI think you can leverage the Seam Faces navigation event +Yes +This is looking good BrunoIve applied your patches in SVN rev In rev I also added a new themes directory for theme components and added your bluelight theme to it +The big downside to then sendingreceiving when they should send or vice versa is that it breaks unicode support in subtle waysSo I dont really want to see come back +Similar to trunk +Looks good to me +SQL allows this construct which is still not accepted by Derby I think ALTER TABLE t ALTER COLUMN c DROP Knut +isnt this a duplicate of Update SPI locking to match JCR from my point of view it doesnt make too much sense to create bugs for all the failing tests as long as jcrspi the spi andits implementation are not adjusted to work with JCR tests listed in this issue no longer fail. +Tested thanks +It might be cleaner if we could somehow recognize the problem and possible fork off to a one off set of code +This issue is resolved from camel side. +Thank you for additional information +I think we really need to take a hard look at this mbeanregistry stuff anyway it will probably go away in the end +For we only provide the ability to sort by a field out of the box because numbers feel more natural to define a static rankMaybe another case where could help is if almost all documents have the same value +This patch looks to be mostly a rename of existing classes +Set extract to false for Jetty configuration +Looks good +Closing all issues +Adjusted version number to Beta +sounds like a bug to me but since it is not JBDS specific i moved it to jboss couldnt reproduce it so would be good if you could say exactly what version of JBDSJBT you used plus a screenshot of where you see this since there are more than one way to start editing a runtime +WernerI appreciate your help in looking in to issue +Patch applied. +bq +Currently looking for a Remoting contact as ESB problem is fixed and the tomcat problem is not fixable +This issue become more about commonresources usage model +Im not sure what other context info would be helpful +Emmanuel this class is abstract for easily firing up the server and running unit tests +This is fixed for RC for fixing this so quickly! +finally block although if we ever update to JDK we could change that to a block since is a Closeable +Updated patch removes some unused imports and in order not to change existing behavior does not override inputoutput keyvalue classes if the mapperreducer is a regular java one +The build infrastructure expects there to be tarballs to blow up +Resolving. +Patch committed with revision completed. +Will add in new patches incorporating new to exit the Task Tracker when an unexpected exception is thrown +This is fixed now +fixed checkttroot +And I also noticed that s was deprecated for so I made that simple change +I understand Pedros use case but applying his fxipatch does only create new problemsI was thinking along the same line as Igor +Let me take some time this week to look over your suggestion +Fixed svn rev Verified as fixed +It does not do any harm and it provides me with a lot of info when there are problems +Completed At revision +That completes the reorganization except for the party component putting as the first styling so that a search and replace can be done replacesnow that you did all this exercise with patch attachments would it be possible to attach also one big aggregated patchSeriously I think it will be much easier to commit it in that way +I have added the two complete lines form the console for the above mentioned issue +If problem still exists we should fix this in +Same patch for trunk +minValue and maxValue must be in the format specified by the format attributeI have resolved this issue by added more detailed error messages which explain which date was not in the right format +Cannot reproduce in FOP +The MR app has its own development legacy and it may not be possible to a client onto it without major surgery on both sides +Looking at the code youd expect the event handlers to be called for the test to pass +The existing build targets work but now the build only produces jars and warsI have moved repos dependencymanagement into the base pom along with the repos lists committers etc +Im happy to do a quick patch +Only if the window is closed a new one will be opened again. +Do you really need another test case or the one attached in the other JIRA is enough +I can confirm that the codes changes suggested by Tobias are working +What does your implementation of look like +Patch against r +Yes i have tried both +verified on JBDS M H +from me +The difference of the is the column number +Unit tests +This proposal sounds good to me +Thank you Kathey +Hi do you have a view on how the api should be modified to achieve this? The reason I ask is so can expose it in a similar way +I have committed this +I think Thrift supports API evolution via default values for missing parameters but setting a default value in this case may be a little trickyAgreedbq +Cancelling patch since this doesnt work on cygwin and MAC +I havent dug in whats happening behind the scenes yet but as far as the user is concerned all the applications are being treated as a single deployment that appears to either succeed or ve only verified this behavior when deploying multiple applications via +Please check this with linux this works fineRegs the Windows build now runs correctly again with the binary libxml +If no one object I will assign this to myself +All subtasks are completed and Im resolving this as fixed +Same fix as for the workaround may get you a bit further but I think I have it knocked out for the next version +looks doable +I remember though that some tests on failed before and after the patch as noted earlier +code imported. +Hi Laxman not yet +Opened to track the Transformed issue the other classes are generic at least as of svn rev . +Will do what you suggest on commit +The issue seems to be fixed in Branch +Resolving as a duplicate of the more detailed issue. + +The checkbox tag hasnt been extensively tested though +We can not reliably reproduce this but at least two times we saw a node replicating the entire index after restart while indexing to the cluster +Otherwise dynamic discovery doesnt work +please tell me Id like to help it done coz it is a blocker for methx in Lars in fact this requires to patch the Sonar core and more precisely the class in the Java Squid Plugin to be able to acceptinject some +So really this type of feature as it would require engine changes would be better recorded in the ODE jira +Pushing to later release as it needs the coordinator interposition support and thats not ready yet +Committed to branch as well +Thanks Regis The fix is fine +This fixes a broken patch previously unit tests are failing with this fix wont work with the unit tests which is run an embedded mapreduce on a separate thread +For trunk it is time to change it to because changes in jpt hibernatetools brancheshibernatetools why did we drop the +BTW is a subclass of +I am wait your ok before close issue in jiraThanks in good applied to in r many thanks for the fix RodolfoShould we merge this into the release +I am moving the vectorizer to core +IT added +I kicked off a build since Jenkins had not picked up the new patch +Okay I updated the wiki let me know if thats what you had in mind +Seems I was a bit hasty +I only changed the composite to let some components use and modified the Interface of Gamma slightly +Disregard the I uploaded at this time +I agree that the implementations belong in the infrastructure project +Of course Compression and decompression of will be provided with different handlers so stay tuned +The code generated processors also arent in a position to do any kind of clever to the right handle type +However it is possible to used bean that will contain method set in createContent attribute and inject different bean holding data +Also added the new test case to the class +Is that WONTFIX because there is no contrib and thus the provider is now at super cool URL or WONTFIX because there is a technical reason this is no longer feasible +There were two rename participants +If youre committer then its implied you review patches and interact with the community nicely +Here is patch version to clean that upI didnt see any remaining imports of the stuff so patch version doesnt change anything related to that +Nick I am not sure what is your problem with Archiva +Committed to trunk +Duplicate indeed thanks! +testing +The intention is that they would only be sent to Derby client drivers version and newerThanks for verifying Dyre +The Derby JDBC driver cannot be loaded +Glad to see it fixed +It may indicate that the client does not renew the lease and so the lease is really expired +The general idea looks right but it doesnt address the core problem in this ticket namely that a crash can reset the view index +Removes dfs scheme +Willem is right we cant do the locator thing with this jar file because it is intrusive and replaces and adds to the contents of the jar +Here is another patch that ignores white space +Thanks for figuring out what the issue was Billy. +Are we making an exception here +sorry you are right the problem seems to be in wiquerys + +DavidI am sorry I took me a while to make time for this issue +Should that be the case +This problem happened on the JIRA upgrade at Apache as well +This was minor classpath dependent jar files needed to be in the compile time library. +Ive seen some tweets about people having difficulties and Ive had some problems when I tried stuff myselfAnyway Im going to try and clean up some of this stuff over the next few days and make the wiki a bit more user oriented an extra pair of eyeballs would be welcome +is a simple implementation and complements and +I would hold up Tomcat as an example of an Apache project redistributing Suns JEE API +Thanks very much Knut for the quick reply! Im attaching and with changes as followsM srcrefA srcrefA srcrefM srcrefM srcdevguideM srcdevguideIn addition to adding topics for the two new properties I added them to the table of properties in the Reference Manual and in the Developers Guide added mentions of them to the NATIVE authentication topic and the list of propertiesPlease let me know of any changes that are needed +I didnt know where CLI stopped and Eclipse began +Its done sorry if this was an inconvenient time Id finished before I realised that I hadnt asked if you were can use svn switch relocate to switch their working copies to the new +What tables did Jesse add if any in his patch? System tables being with a dot +This patch adds the related packages to the and jars and is added to the import search path +Its a documentation issue +Thanks RickRC for openejbbranchesvopenejbRC for openejbtrunkopenejb +Its actually a problem that Hadoop is so dependent on at it offers little extra over +I dont see a reason for sticking with +PMC voted the attached designs winner as new hbase logo +Mydb is a cocuchdb database that is being accessed by the cocuchrest gem +Thanks Amareshwari! +Personally Id not look into Attachement functionality for the time being as Id consider it to be the least important +this is normal under the Visual Age C Professional compiler +The following is what I used +Thanks for the heads up +Hudson failure looks unrelated to this patch +I went with the majority If you think it belongs in I am uploading a new patch that deletes the one +added +This also got clarified in the spec +Can we reach out explicitly to folks working on the fair scheduler and dynamic scheduler maybe add them to the watch list of this JIRA or something +Thanks Richard and Gunther for contributing this complex and very useful feature! +It works +Description Key Features Component identification parameters Customization were API was added +This one has been committed all issues that have been in resolved state for more than one month without further comments to closed status +There could be a build process that creates both and debs and just makes it an option for installation. +Im interested to see what you end up doing +But fix is good +If a user hasnt enabled it use the constructorsgetters that dont have that param if they enable it use the constructorsgetters that have that paramNot sold +It can take arbitrarily long between the time that an object becomes unreachable and the time that its finalizer is executed +FYI the hanging only appears to happen on Mac OS which does not provide production support for so we are putting this one aside for now +Checked the two openjpa jiras and they applied to and trunk of Geronimo trunk now uses and branch use + solution but requiring a new method to public APIPlease try and tell me something +Bulk move issues to and +I havent been out of that project whereI was trying to use Axis JAXWS +Need to possibly clean up the required css and add to docs +But they show a few ran into a column actually smaller than the previous column of the reducers were restarted due to timeout being reached when attempting to contact region servers the batch therefore ran for more than hoursWill rerun it once more on an empty table to have a double test but for now it seems dataloss still occur +Then please make a getter in the to pull this value out +Test class a lot for the patch Madhukara +Whatever Max That was a goal for the User Team and you know it! +Converted to of though its technically as much of a of +We opted for simplicity in the initial implementation +Fixed +Sorry to come in late the patch has gone stale +Creating a single patch for skewed join +I think t might be better than unsigned int so thats what I put in the code. +CVS is also reporting escape +Im still curious why a lowercase string wont retrieve the table name this seems to me to be a bug in hsqldb +You can make such changes only in major releases and not minor +Feel free to grab me on IRC if you have specific questions +Moving to. +Patch looks fine +In fact its not working with primitive int eitherBUT its working fine with primitive float +will handle these one at a time +First apply the patchand then copy the files to the documentationresourcesimages folder +This bug has been marked as a duplicate of +This patch fixes the issue +The new patch worked for me thanks +I just reviewed the code in and it looks like the major piece to bring over is the idea of the converter and its ability to transform the raw data and provide a schema for the outputting formatThis would add a lot of power to the existing implementationIll start on this tonight. +User can modify these operations manually as a workaround +More logging closing open file handles still works for me +Fixed in. +I had a really vanilla checkout even no maven repository and the test now things are problem +Which mojo are you using? Which version of the plugin? Can you send a test case +But if you put javascript code in template file of a class this bug dont appear +Fixed lint issues and merged patch into trunkNext time please make sure you run tox and lint step passes before submitting a final patchThanks! +Tried the patch with no improvement +As for the branch Ill leave it for the core devs to decide on what to do +Merge updates from classlib trunk since rIn addition to concurrent failures we find many failuresNeed further investigation. +Pete will you agree +You cant access a field thats not declared in a schema +Also pg uses different syntax for client and stored procedure cursors +I dont see how we can fix the context paths without removing the seam prefix which risks collisions with user apps +i have the same issue here +In each stage I tried to import an X project from this git repository +i would appreciate it if a developer can confirm this if it is indeed a bug and nobody has the time to fix it for the next release then an indication of where in the code i can start looking to eventually fix it and submit a you in advance cheers you Andrea for looking into this +obviously when compiled with wasnt fixed by r +This Test case is to test the receive serialized and inventory +Then basically my options are to setup a more responsive ZK cluster or maybe raise the ZK timeout on Solr sideThats all I can suggest +I think this would be an improvement +Im going to apply the remaining changes from your patch to with some modifications to apply the m std layout and to use the updated artifactIds for the deployment modules +When will this failDo you have such a +Manish it would be great supporting positional parameters +pls try to reproduce an issue +Attaching and which differ from the previous versions only by including changes to the Admin Guide topic on replication and security and on adding changes to a Getting Started guide file that describes exceptions +Patch submitted +I know this is really easy in Jersey with the annotationsIm also not sure if guice and jersey integrate nicely to allow you to use the same uri and just specify the type of output you want +We should let the portlets override the SSO credentials if credentials are defined in the pass these to JIRA or SVN else if user is signed in using SSO pass the SSO credentials to JIRA or SVN else use public access to JIRA or SVN +Using port in your provider url will use hajndi for the lookup which should fix the connection factory lookup +Where are theseNothing in yet +Furthermore the driving force for us comes from how Oozie constructs input paths +Well get this prioritized +ok I see how this will work but it is a bit of a hack all be it one that is of valueI would expect some sort of generic RAR impl that is aware of a resource adapter in jndi and can make calls on get to do recoveryIt looks like no such thing exists in jboss or arjuna atm Having an option to publish an XA connection factory in jndi looks like a reasonable workaround +Down the line we can probably change this but unless this is breaking anything beyond your sensibilities I dont have a good way to solve this for this release +Either would be fine but Ive certainly found a lot of utility in historic docs from other projects and this would provide maximum clarity +The other parameters are configured later which are the period etc +The case with producer threads is pretty interesting +This will cause it to overwrite the Errors instance in the request scope +implicit router would look into that field for routing any value of the field will be the name of the shard where it belongs toIf the collection is created with routeField other routing params are not honoredThis deprecates the parameter for routing queries in favor of a parameter called route +Thanks Ted +yea cool patch txsHmm David since is reasonably stable may we ask you to become an OWB committer tooIt would be a huge honour for us and hacking on seems to be one of your favourite areas anyway +Reviewers please also check adherence to naming conventions among other things +Please merge both the patch and into one patch for trunk today as thats supported to do now and makes more sense. +I have just encountered this bug +lets move this to the hibernate tools forum for discussion instead of in jira. +Labels of assetmaint checked in Label Manager missing +TB full data set +adding docs +Isnt it that you just have to uncomment the Basic realmadministrator line in? +Please reopen with a testcase if this is still failing. +this test was putting the archiving dir beside the test log file +ThanksHeshan +Could you provide a link to place in the file +Consolidating to so we can then rename to +Hi Mahadev any suggestions on the reworked patch +Thanks for the patch Devaraj. +Note that combined with other issues this nearly completes too +So neither solution and solution solves the fundamental problem and it does not matter which solution we use +The new patch contains the fix the testsBTW All the other formatter test also seem to pass +I forgot to mention that the test failed is nothing to do with this issue +Any news Zhu + +Compiled against hcatcommon +Tom fix for this issue requires an update to the README in quickstartsdemocluster as well right +Thanks. +I guess you mean an exception stack trace in? The log was deleted when I restored the database so I will send it the next time the problem occursIf you need something else than output from the please specify +I do not see why I need the manager running in a RS? Why the RS need to be able to list namespacesget is the wrong name for a method that returns a Thanks +If the input format and input path are not given they will work as the current implementation using +Ok this looks like a softball +I checked a few things out and this looks good +is tested by so no new test is needed +Ive spent a bit of time attempting to work my way around the various classes and methods but I am not getting accurate values for the map input and output counters +That is the change I tested yesterday and it solves the problem +The machine would go readonly +AndreyIt works fine +For all intents and purposes unknown is the same as null here +Hi Andreathe EHDR is internally supported by GDAL +Further progress +Of course the implementation is free from choosing how to handle SQL +In preparing the patch I updated and then discovered it had already been recently fixed at revision . +I will check this and update you +Brian does this need to be fixed in a cumulative patch +If there are many curves connecting two roles then we cannot predict which arc will appear in the list +warn before terminating any rogue build processes +David try AXIS and OAGIS BOD Flattener +So when cache will be replicated and deserialized in another node SF wont be deserialized but taken from NAMEDINSTANCES map +A Host Controller does not install runtime services for subsystems +This issue is fixed and released as part of release +This patch will return all topic metadata if none are specified in the +Assessing for this test in svn revision +must already be fixed. +All looks goodPlease help to review and commit +During this period I do not want some production projects to start relying on the sync features then find some regressiondifference on blocking our upgrade scheduleThere may be a misunderstanding the flag never controlled whether sync was enabled + +idea as code +This implements choice A from the appspot code review using as the injectable factory object +Any ideas on this one? If its can you follow up with Bob +The test case is disabled for now +standard wont work since each source node is doing potentially multiple copies +FWIW I think we ultimately need to add support for an element +We need to get a SGA submitted from as well +I tried out some simple examples and the results matched my expections so I think that my expectations must not be the same as yoursPlease attach a sample stylesheet input document and the results that you expect to see +For example look for and options which hasnt been ported to the scala DSL yet +Ok +What you pointed at is really a problem and I have another solution for this +Fixed thanks for testing Stefano. +Hello AlbinYou should rather use the user mailing list +Failing tests pass locally +despite the compiling process ends correctly with the attached patches theres no way to let the example test program to run it segfaults + + for reporting this Tomasz +css files are applied to the leftAccordian elements +The patch is attached +Otherwise we can make a mistake and modify the wrong div +Ive attached a patch and an archive containing all the files in the patch some screenshots +Patch for this patch fixed the issue for +As I believe the collation code to be mostly the same between the two versions the problem may not be specific to CollationBut that needs further research +Paul which branch did you diff this against? I tried applying to trunk but +Please review +Setting to future pending a patch +sorry its late already +deferred pending a decision on the future of the transactional driver vs +The map of standard attributes is never read. +I like this because it makes the users intent for a table very clear on a system with many users +could you explain this a bit more? why are these libraries needed +jpeg of the mind we close this issue and consolidate it with? Is there some reason to have two threads +For basic questions like Maven dependency handling setup and other kinds of support questions please consult the forums +EDITAfter looking futher into this i think the issue is that we do not look if the hypervisor itself is pingable +Not sure when Ill get a chance to take a look but could you raise a separate issue for it and assign it to me? Thanks. +Forgot to check in in changes described previously. +you use other container like karaf ensure youve edited etc to commented out package to ensure it will use +I agree with Leonardo +Hi in fact the URL we used is like thishttplocalhostecmdemojcrrepositorycollaborationjcrsystemjcrversionStoragecfcabaaacbdadjcrfrozenNodethere are no problem with httplocalhostecmdemojcrrepositorycollaboration but if we use it to display the version node we have the exception +After talking to Lily on IRC I made the folllowing changes Improved the comment for get Changed the upgrade tests to just always run with create for and removed the extra run +updated more test was committed by Kevin last week I think he could not update thejira since it was down +thanks +Attaching a draft patch for the first part of this Upgrading a database from to should automatically change database owner recorded as owner of system schemas in sysschemas from pseudo user DBA to authorizationID of the user attempting upgradeThis was being done in method +Updating to newer Xerces version solves the validation errors +Your tweaks are a big improvement +I ran both tests w and they passed +Sorry I come late to the subject but I wonder if the core code rather than the test should be fixed here +The test queue definition +Patch from Paulo Poiati attached +Currently I know no other way than recognizing catching the exception and skip the contents in that case +Resolving as fixed +However for we dont want to change the API especially to replace it with a potentially less stable APIGoing forward we will undoubtedly have a discussion on the proper API to use +The first one was already covered in the mock test +This way we would keep the crash from happening ever again while still it is reportable when it occurs +Implmentet natural to Eclipse behaviour see last checkbox +Resolving +Your suggessted fix is exactly the one we have made. +Resolving as fixedThis was committed to and had remained open in order to backport to but because this is a new feature and the branch is open only for bugfixes it will never be released in any version +One of us will review it ASAP +value attribute description was rewritten +Im still waiting to migrate this to a Tomcat server instead of the standalone +now contains a layer which creates an havent tested it but by looking at the code it should work +Since this is breaking change for systems that relayed on this feature it would be good to see some backward compatibility option +I am sorry and next timeNo problem! Thanks for +I dont know enough about FOP to say whether this is the best approachLooking through the code it looks like it would be possible to get rid of +I dont think there are any others like that yet +Moving as there is a workaround for this +Though we dont have to support binary compatibility but a best effort would be good +I tried to reproduce this naively but not able to +You are right +It would help if that comment were added to the script for future readersIt looks to me as though the perms entries that cause problems are the ones with gs so it might be sufficient to usechmodor maybechmodstto fix the specific issue as stated in the descriptionThe other options fix a different issue which is that the source permissions on minotaur are sometimes set incorrectly +This patch should make this test run on all platformsKindly use the resources from dont see how this patch addresses my previous comment is windows only isnt it +Hi SteveI think it should be an issue of adding the style sheet and images to replace the youd like to send me a link Ill take a lookCheers youd like to ping me directlyemail cweaver I actually meant in terms of what I needed to do with the j styling +I just got fooled by thi +Next chapter The workaround in the last comment is only valid for the current process +Yes Vincent it was basically that although you did have a setter with a single parameter too so it was a little more complicated and buried deep within drools compiler. +here you are +Added a patch for the modules listed below core demo wmsHopefully the patch can be applied this time +Patch for earlier version of hadoop +It is too bad I cant think of a way to change Jackson to just work here but it is rather subtle combination of things +Surely this should be raised against HAMA not INCUBATOR +I backed out the automake setting so this should be working again. +move to as we create new close all issues. +So u dont want to use the compression based on the config param +failing until test passes with the release so this is a regression wrt +verified all generated files have the correct version +Answers for the Questions asked above +In CVS. +That offset will be used to truncate the followers log and for refetching from the leader +Thanks for review +Let me know if the patch works +remove tag from Fix Versions +The problem is that Rave simply crashes with the wrong settings and I did not really know what to do +Makes sense +Re static caches this should probably be addressed in a separate issue +Resolving +Using a specific version is soRather than upgrading the version Id like to try using s git tip will change the descriptor +ok +Let the create return the that is acceptable +Even if eclipse configuration generation is skipped for mac installer it is not gonna work under bit arch because it contains eclipse for bit platform +Catch the abstractmethodexception and keep going +Any dependent projects may not end up with everything they need in their The build scripts will need to exclude the runtime scoped dependencies when generating the lib dir which would normally be This solution will not work if the project has real runtime dependencies +Having the option available to the user sounds reasonableInteresting question arrises is it the responsibility of a networked component to be a good citizen in the network or is it up to the network to manage its citizens for concerns +who knows? +Hiall I will provide a patch for this JIRA to show better file upload support in servletAlsothis example will demonstrate annotations for serveltfilter and listener +The improvement has been checked in +Hi MarkIt should have read to reinstate using the version of Emma that we previously were using in +Things are starting to pop out a bit too much in places +Can you attach patch for trunk so that Hadoop QA can test it +pushed to in Developer StudioVersion Build id Build date +Sebastian I have seen your commit +See the work log for details +Zip contains Patch to to handle nested messages used for +The only place the splits would be instantiated would be the submitting program and the task jvm +Can you verify that +and trunk require that this patch requires for and trunk and forPlease apply this patch to all three versions +In order to keep things simple we provide an operation to add at the end of the list only +In we keep the resulting tuples in a Map and then when the from is executed again we diff so we know what previously existed and what is new this is necessary to keep intermediate results in situations like accumulateThis level of state management has a cost +stated rightfully that there was inconsistency in this class since it didnt actually used the services provides +Please post link to JIRA item for your idea +Patch removing This is just to get Hive QA run +As stated all feedback welcome +bq +It should readbq +I have also added another test which runs the protocol requirement tests against a using the Ehcache backendSomething that should be discussed is behavior on a failed update +Ive committed to trunk +Its my first patch for community +Thanks Mukul Probably Oxygen has not set the required feature I will report it to themRegarding to your developer support I will change my code accordinglyMany thanksBert the Oxygen PreferencesXML XML Parser XML Schema page there is a checkbox called Enable XSD CTA full checking which is unchecked by default. +The syntax block now refers to the and topics which dont appear to be a perfect match but are pretty close to itAs Knut Anders observes this should help to describe the various combinationsof schema name table name and column name more rigorouslyPlease have a look +Very usefulcompiled package binary packagewould be clearerI dont think its advisable to incorporate the list of license types just link to where the official list is found +What version are you using +Bug has been marked as a duplicate of this bug +Committed. +Came across a situation during testing where the regionserver waited indefinitely to flush because no compaction was scheduled +Otherwise the changes look is updated to trunk +It would make some sense to close it and open bugs for each item still left to do +Unless one of the committers wants to take care of this Im happy to commit it +Tool should be able to deal Id say. + +I will post the test results tommorrow along with the tests that need to be run for the cases mentioned in the comments +Please please add support for these types and include xsdunsignedLong +I dont know about using hibernate in a full app server but what happens if you want to use jta transactions in a web container like tomcat +It is still critical and should be addressed but their is a is a workaround yes but lets leave it as Critical please +Will take a look +split classes into separate patches and fixed bug in estimated tables +After applying my other patch this problem is fixed and the test passes +Starting listener threadTRACE High volume detailed loggingIll change it to resolved issues for already released versions +I see +Are you talking about the hadoop tree +But if someone offers to set up such a for it Ill put it there and maintain it +The Marker created prior to receiving a complete CAS in the reply is therefore invalid for future delta CAS serialization +I guess the should be refactored to handle the subtleties of creating appropriate application context +Heres the patch +I think your problem may lay in your way to use threads +batch transition to closed remaining resolved bugs +Sigh +all issues not touched in a month to and it looks like these are not checked on a basisSo if a user sets different limits for different fields and wants combined limiting that is not possibleFor example a user might setcombined And the combined method will not be usedIf the user sets combined it looks like the same limit will be used for all fields? Whatever the global is set to +Removes the repository readwrite test so Geronimo will run with the repo on a filesystem +Did not keep the expired certificate as the only purpose is to serve as an example which must always be replaced. +The revision is The result will be improved performance and reduced memory footprint when there are large attachments and a large fileThresholdThanks! +patch because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Maybe returning a empty string or parsing as much bytes as possible will be a better way. +Gosh these are all fantastic! Any of the ones youve made would fly with me +While there may be a workaround by insuring all statements are closed the reality is that many users programs have this problem and a deadlock in the field is going to be hard to diagnose and tie to an unclosed statement +Removing the fix version since there is nothing to fix forFurther improvements should be provided in a patch and can be added later. +Patch applying the chainsaw to the forrest The documentation root is now builddocs which is created by XSL transformation Query Parser syntax was moved to the javadocs of Scoring howto is still outside javadocs but by the carpenter +The new class checks whether the object under consideration had been deleted before or had been modified +The only thing is that the module is not under optxensourcesm but instead is under usrlibxcpsmrootdevcloud dpkg L etcetcxcpetcxcpetcxcpusrusrlibusrlibxcpusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmmpathHBAusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmsnapwatchdusrlibxcpsmsnapwatchdusrlibxcpsmsnapwatchdusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpsmusrlibxcpbinusrlibxcpdebugusrlibxcpdebugtpusrlibxcplibusrlibxcplibusrlibxcplibdcopyusrlibxcplibusrlibxcppluginsusrlibxcppluginsusrlibxcppluginsvsscontrolusrlibxcppluginsusrlibxcppluginsusrlibxcppluginsusrlibxcppluginsusrlibxcppluginsusrshareusrsharepythonusrsharepythonusrsharepythonusrsharedocusrsharedocusrsharedocusrsharedoccopyrightusrlibxcpsmusrlibxcpsmvsscontrolusrlibxcpsmusrlibxcpsmEXTSRusrlibxcpsmISOSRusrlibxcpsmNFSSRusrlibxcpbinblktapusrlibxcpSo Ive just created a symlink as John suggested as follows and will try this againrootdevcloud ln s usrlibxcpsm optxensourcesm +Use whichever you like but note that is a clone of svn and is a clone of +Fixes Multiple PK Entities handling in Tree sounds good Daniel have you an use case to test +If we find any user without a SID the user does not exist +See first comment in this issue point No I need the new Eclipse Code formatter for that purpose +we should have fixed metric name regardless code refactoring +Id prefer merges where aborted usually if you are going to listen to an interrupt at all you try to do it in a timely manner +I just talked to Todd and he agrees that this code will be superseded in by the work thats going on +Changes are done Im now testing the changes +If I disable the implementation the internal codec becomes active and fails on the default configuration settings with said exception +Branch was completed and now entering. +My password hadnt sync with LDAP non related issue +This issue has been implemented. +Submitting an updated patch +errorTextFailed to authenticate user in domain please provide valid credentialsdomain is for root user +James would you please look at this? It probably works as designed and needs to be changed to an improvement ticket +Updated in revision . +Am not sure if i have enough time to squeeze a bit more performance +do you know what part of that controls thisI couldnt find any ref to or is it just the presence of an that enables it +The chosen port is however registered with the AS socket binding manager so at least we can identify which port was ultimately used +I just committed this +Adds methods to decodeencode paths in Http requests +update set to Harmony jre home +YesAs you can see in the forum thread Hibernate should choose to return date and time information like JDBC does +If I can detect any difference at k then Ill stop there +Ive tested this patch against our test suite +Ning I will test this patch with our usual set up and let you know how it goes +If not will try Resin and will commit if Resin too is happy +Thanks to Frank for the fix proposalThe attached patch has passed all build tests and Davids Test Case +An example application is to offer users the option to search further afield having read through initial batch of search results +iPOJO Core was released. +Ill have a word. +Updated patch uses property in +Geoserver part to come separately +Thanks Devaraj! +We should then ship the redis source +So these two tests have not be run for a long timeI ran these two tests manually +I ran the junit tests in sanefalse mode and the prepare gets XARDONLY which I think is what is expected I think the test is wrong and should be looking for that return +I think this fails because enum is a reserved word. +Merged to the branch in revision . +Thanks guys +bat file specified by an absolute path +The state machine expects the first element within the soapBody while it is feeded with the original soapEnvelope document +Sebastien TardifI am facing the same issue while doing push in with STS have tried all the options mentioned in the help is appreciated on resolving this issue +canot renderer wms and am getting this while attempting to render layers popplace on httpmswfsgmapversionrequestgetcapabilitiesservicewfsif I zoom into BC it displays properly +Need to change version from to +needed in can be null when column pruning is disabled +Will do +Gregory patch verified at r +Either that or we might specify a file on the command line where the port should be written +Any news on this issue? Has it been resolved or any workaround been identified? +Committed at subversion revision Touches the following filesM javastubsfelixorgosgiframeworkM javastubsfelixorgosgiframeworkM javastubsfelixorgosgiframeworkM javastubsfelixorgosgiframeworkM javastubsfelixorgosgiframeworkM +More detail is necessary +looks to me that a fix has been applied and this is therefore resolved in M +If you take a look at the patch Ive created I think you can have an idea on whats going onFeel free to continue from my branch or start a new one if you prefer and dont hesitate to ask if you have more questionsThanks again Im just running your github version against my production code +This issue has been marked WONTFIX which has been accepted by the reporterShouldnt it be closed in bugzilla? +The reply to EPR in the two channel cases is special becouse here user ask axis to start Axis server and do the two channel invocation +The NN will also stop sending the machine name in the param string by virtue of the fact that it uses get to form the param stringI also tested this patch with an NNNN and it works just finebq +There was some failure last night which hasnt been recovered yet Ill post more info here when I have it +Ran a couple hundred times without a crash on a couple different platforms +I was planning to file a feature request for something that the Gluster project calls +Cool +So for now we need at least continuous smoke testing of all features including HWI imho +path to fix add the requested these items are enhanced +Im going to resolve this one for now. +In order to handle the language code the database schema for the table was updated +So if you specify the epoch you get everything +Didnt lex get handled by? +That said I dont know if AND that behavior will very often be needed at the same time +In any event I suppose well stick with our opinions until a better word comes along +It uses the Plexus compiler utils to do dependency checking so only report files are compiled +adds a +Committed trunk revision Committed revision +Thanks! +However the Java add import declaration doesnt know that +the required will only be used if the wizardId of the proxy wizard isnt found in the current installation +Then we should make a change to the trunk to see if we can run against BOTH and +Committed in r a slightly bigger changeThe tests changed to comply with AS Arquillian and Maven requirements +IMO is worth the move +Thanks Rob +Hiya release is the correct one for SOA +this issue can probably be closed. +Why is glob going through path? It seems to me that a glob string is not a path string and shouldnt be processed as such +Im willing to apply the patch but I never used JSF so its hard for me to check if its valid +Confirmed works but development looks pretty hopeless until this nasty bug is fixed am voting for some Graham magic its so weird that this plugin works in a milestone version instead of a stable version +Moving to in preparation for release. +Hey Matt +jBPM also gave the lanes semantic meaning and it not even is a big change if a user task is in a lane and the user task does not have an assigment use the one from the lane +I wrote some code for this issue and now Im testing it with +Even hardcore developers patience has its limits +Rescheduling to . +Now the system is merging the configuration information from the xml file and the class annotations +However each of the targets to build one individual artifact get meaningless autogenerated means the ant script guy has an choice +thats just the case in the in jackrabbitthe suggestion here as far as i am concerned was to be allow to indicate that given session would never make any attempt to write it was a which isnt the same thing as creating asession for a user that doesnt have any nowhere at all +Ive tested this patch on with HP aCC compiler we use on this platform in production +From my understanding for Nutch and Giraph to be compatible in a useful manner Giraph needs to be used as a dependency in Nutch not the other way around as seems to be the case hereDoes this make sense +Do we also need to make some changes to the for that to work +I think the restriction can be removed and I will do the fix by tommorow +This should do it here +Instead MC should use the property names from Attributes like the does +Why doesnt that static code copy the list and then add the few extra terms that are needed? If there is a reason place it in a commentBut as I said by all means proceed to a normal patch file now that the tar contribution is legal +In the case this is necessary please reopen this ticket and comment +Glad to hear that the problem seems to be solved +It was decided to let both convergence methods availableThe reason there are two different way is that the implementation originally came from Netlib and I kept the way it behaved +What is the issue here? What you mean by AWS + +No followup with recent and so Im marking this incomplete. +No I didnt touch that one at all +i thought we had agreed that this wasnt an issue and it was just log played a little with it and you seem to be right that this is just logging issue +Runtime bindings also a big win +Issue also only occurs in and above +I will think about PREFETCHNTA +The path didnt start with a so contextRelative should have no effectI like the idea of passing the url straight through if it doesnt start with a slash +Note that this is due to so this affects the branch but not any released version +The only way to fix this is to restart whenever a consumer restartsRegards need this +much better idea IMHO and youll have to mantain and manage only one index for your applicationIf it gets too big or if you have law regulations you can then use sharding to split the index and store them separatelyHibernate Search in Action has quite some tips about this situations +I was looking for the Lifecycle interface discussion so that I could just append to that +Now I am much more confused why it doesnt work on +Fixed the issue with spaces in file names failing to be selected. +The close got an NSRE because region had been offlined by time close came in +Is this a duplicate of which details and encompasses the differences between file system implementations +We had an issue where two apps were sharing the first apps rmi registry which doesnt work well when you shutdown one of commented that out because unitils doesnt shutdown your app context so living with it for now +Thank you Haohui +I see what you mean +I only had to do some minor changes to remove all artefacts from old test framework and to get derby scripts to run. +So I hope to have this issue resolved this weekReally? thats awesomeWorst case some of those targets could be literally put back probably with minimal modificationsMy idea of temporary nuking was to try to start over extending lucenes build system as otherwise i got lost in all the xml +I actually just did all these same things from your latest patch over the weekend +Closed upon release of. +I will try to address this with an external file later todayAnyway this works that is really great +Agree with Jon that should subsume this especially considering my latest proposal there +Thanks! +Juergen You want me to fix this? Unfortunately I have the pleasure of being familiar with Oracle ve played with OCJ a while ago too but havent currently set it up on my machine +You are correct the region level seems to be the correct location +Correct Ted +Lets open new issues if there is anything that I forgot +I dont see any reason why I should have any more namespaces than you have if indeed you have no +Thats pretty valuable even if constant table updates are impactedIs there a particular piece of code using removeconst heavily +Yes +Max please review and accept approved +Problematic lines seem to be in +All new implementation contain a consistent Config Service and Config tab on the UI +Ideally all matters is the interface and Pig shouldnt care how the underlying layer implements an interface or instantiate an object +Please verify if the new patch is a better solutionThanks +act like and just ignore the Symbolizer +Yes we can simplify the if we get rid of the YARN one +So youll probably want to add it to your path in order to test your changesG +coresrctestjavatesttypesruntest + +There is no update on these items and are in open state if they are not ready for please move them out to Future +I verified this happens with the current cvs code +I previously stated that I would prefer changing how the locking is done so we could backport to but that is no longer a consideration +Attachment has been added with description has been added with description an example of a PDF with full +Moving to +Weve seen this already in with the step issue where the uses the system property named profile +This issue is caused by the type converter does not acquire the when it coverts it into +Thanksderek +Reporters of these issues please verify the fixes and help close these issues +The first build produces regular artifacts while the second one produces the indy artifacts +Caused by special configuration file jrelib in IBM JDK +LarswowHi Gertyes! Thats it! I was looking for the schemas for my forThanksCheers! was removed after changing all incoming links to the new XML schemas page +I just finished testing this change on +Also how does this issue relate to? They seem similar but in a is expected yet here a is expected +Its not just a Pig issue its an issue for any code written for that expects to walk a directory tree +this wouldnt be a safe change +Thanks Lohit! +M language pack is approx +Related to regex in v +I must have been dreaming +Verified thanks a lot! +close off release +Alex this is blocking a lot of tests for QE and should be a simple fix +My time is limited but my pc and servers have time enough to build our application +Pushing to after all +and this one kills any task and cleans it up other one I mentioned is the one when the decides that a job has finishedfailed and wants all the tasktrackers to cleanup +I removed link to as Kenney found out it wasnt dependent on it +I dont mind thank you +Looks like a problem but not sure how it could handle spring removal nicelyOne thing we could do is warn if you uninstall a feature which is a dependency of another installed feature similar to the force flag we have on osgi commands +Hi Patrick the c clients all run in a Linux environment the kernels are +Thanks Uma +As far as duplicate jars it may be in the deployers directory +Passing this one to you Shelly I think it just got assigned to me as a default responsible person during the month I was driving dont know if at this point theres actually anything to do here +The test I used to create this exception and the two repository configuration files I used to configure the two nodes +Ill review the patch +After further testing and evaluation I intend to pull this up into the branch for the release +The value of any properties will depend on the solution as the service name is compared against the value of from the native SPIWe will have to wait and see what is discovered before proposing a solution +Thanks! Ive applied your patch on servers where we were experiencing this problem +I dont know which tests use it but if you run directly with you should see the effect +Correction the infrastructure is there +Is that the case with the snapshot you mention +Potential ve committed a different version than the patch instead of merging the maps during an update of the auth handlers a fixed array of possible candidates is returned and this array is queriedIve replaced the hard synchronization with a read write lock as after release +looks good to me +This is definitely going to break a few existing configurations +Id actually prefer to keep that comment +Also a couple of import optimizations are done +Both Images and Constants classes in core were in the same package with the same name as in drools and soa. +Attaching a patch for this issue +should have fixed in axis trunkPlease test with a latest build and reopen the issue if it is still exists with a sample wsdl to reproduce the issue. +I committed this to and +That said I still think theres something weird going on in the JDBC +with the thought that it should not be handled as part of patch with the above comments implemented +Applied patch to trunk to branch in applied much appreciated! +Updated test file that attempts to bind the three DBIxxx time types +SF definitely prefers commiters with ssh access +it doesnt belong in runtime data also makes sense for the gpd to support all the description tags +Returning decimal would be in line with this +All the other items mentioned in are not required. +Patch applied with minor corrections +I suggest you bring this up on + +After experimenting with the service configuration for and some of the duplicate entries are gone +This exists in as well +I find it more likely that we just werent testing something in got this message from the hosting companywhen doing Run All in the test suite the first test that fails is compact the error log is attachedfrom time to time I can see quie a few +How is it possible for a cyclic dep +Ive verified that the patch works as well +Try to connect to it using client script +Telling me the URL is wrong and needs to be fixed but not providing the URL youd like there instead means Im forced to guess and sometimes I guess wronghow would relative work? point at filewhereyouinstalledjbds +BULK EDIT These issues are open to be picked up +This really is an issue not a question +We definitely do not want to maintain job tracker url construction logic in Pig code +Theres a way to go still because of and other complications but this is a good first step +pushed both PRs upstream +Shows how the looksWhen run on RI try to make frame even smaller and see how arrow buttons get smaller. +Anybody cares? Is anybody using this in production? Else I will simply change the fields names this is not secure +Will post another patch when thats cleaned up +It would also be nice for other attributes like title to display a hint when moving with the mouse over +There still needs to be some more polish though +I see very little point in commenting on resolvedclosed issues + change the code to try to parse the uri with class anyway but if it fails silently fall back to I am in favor of this option +I had to get off my companys VPN to do this and I had to adjust the previous in +When i tried this with the producer turned off it worked perfectlyThanks +I thought this would be useful in supporting backward compatible with aliases down the road so client code doesnt need to reference generated members directlyI have an ICLA on file with the ASF and assign all rights to the ASF under the same license as Avro +Its done +If you look at the code the only references it outside the Strategy classes themselves are a couple assert statements that could easily be cleaned up +Maven builds were failing and when I checked Archiva I found the same Error there any info I can capture that would help track it down +This change will require an update to the tooling module in Solder to use the new annotation locations though Ken you are likely aware of that already +It is now yes +Is this a resolution to our problem? If so then please can someone take the decisions to resolve this otherwise can someone rnu with it? I dont understand the issues enough +I can fix this at the same time as if we also want to trash the Abridged bucket in favour of a better name +Stefan +Bug has been marked as a duplicate of this bug +The changes should be tested also +It sounds like this is the time when we should change the Lucene indexer from a commit hook to an observer +Committed to trunk to resolve hudson trunk test. +Fixed as of +batch transition to closed remaining resolved bugs +Any suggestion for how we should structure this in the repo? I have never set up a project with manpages before +I might be mis reading but it looks like the fails were added to the wrong location +Maybe thats the problem? I am using Mac OS X +HTML file to trigger issue +Many thanks! +Can you please fix also? Thanks +Not impossible +Once this is there we can get rid of it from the main and all should be good! projects are officially moving to a maven build system but some projects take time eg Seam has not officially been done yet so I had to use that one + design team is helping with some graphics here +Thanks for great patch Ryan. +Initial rights are set on WB constructing Then there should be the method to limitelevate rights from java is implemented +Have added a test case for pointing to problem +bq +Actually the code already reads GNU long names just fine BSD is not supported at all neither is writing the GNU format which is difficult to do in a streaming fashionI havent managed to create a BSD formatted archive as the BSD ar command on minotaur creates a GNU format archive strange +Leaving open for backport +I dont see the first field saved when I go from to +Log Manager will in the next release and logj are the only two that do. +We cant really do much with cases +The remaining class is +That was exactly my question why cant the VM see the circular reference? Is this specced behaviour or a bug in the Sun VM +bq +Just like Todd said we dont want to change anything else besides the shims layer +And thats exactly the reason why I want to be able to replay your problem as otherwise Id be making a change based upon reading source code only +But from my point if view every code blocktag body should be able to have its own local scope +The isset should only be true if you set that fields value to +will be stored as because in GMT time it is GMTAs the timezone difference to GMT becomes bigger the unwanted side effects may also increase +Actually is not designed to work correctly when the source directory is being modified +Flag me and Ill remove it again +Logged In YES useridWere getting the same error with under heavy load where heavy load is stress test with threads +As long as the bitmap is painted in a transparent way the normal background painting would not be erasedWould you mind listing the remaining issues on the Wiki so we can track progress and problems +do not need to it is so You have to fix +XSLTC can insert line breaks anywhere inside and betweenelements. +Do we by default change all occurrences? Maybe when the db schema tool is run and modeshape is determined to be deployed we add a prompt telling the user that they will need to update the config file with this DIALECT +Under jcarder this shows the lock inversion in the previous release and shows that it is gone in trunk +Im going to commit this today unless there is an objection +So having float without a definition in a data file wouldnt let us later remove this as a primitve and replace it in runtimes by a standard library +I have added option to set wait time for schema migration in seconds from command line if it was not set system will use nodecount secs or secs if couldnt connect to JMXCLI checks node agreement on every keyspace createdrop and if nodes were not able to agree in given wait time system will print error message The schema has not settled in d seconds and further migrations are until it does +Also +Hi Howardyes first it was commented and from logs i could see that this property was to false even when i use +JohnI have attached a patch file containing an initial attempt at honouring the war setting on the based on something done elsewhere in the also contains a modification to the to include referenced projects within the same RAD workspace they were not getting loaded when launching the web app in a testi try to use this plugin but i have no clue in which order i have to use the packages above +Dont we want it to be in the jar so you could build from the source in the jar? I did also find an issue that you could only build from mapreduce top level directory so will fix that +Alessio checked on Sun JDKTested on EAP CR and CXF installer for EAP CR Hudson jobs are available if you need +sorry doron i actually reviewed these patches the other day but aparently i got side tracked and never commentedi think you made the right choice with the +that means he just should have used the exactly SAME pom to execute like he does without problems for the and just fails to resolve the relative pathI cannot imagine that this is such a great deal to fix +Patch applied +Which method of service deployment are you using? I see both a and spring files in the sample +We need to avoid any indexing operations wherever we can +All right got some statistics data +Uma How about making it as an option to let user decide when shutting down a bookie +Documentation and sample application also tests remain +Attached patch is intended to fix the problem reportedAlthough the patch scope is wider than reported by this test it removes cache of classes which loading or initialization had failed +This issue does not occur with Maven . +Reran removenode as node was now in a DOWN state +I do not want multiple connection pools across multiple Rest +Youre right that the test could do a simple validation query +Thank you very muchNow and all build successfully without encoding error. +will commit if tests without this patch the test failed on +i think this patch will fix it but my build is broken for unrelated reasons +The correction of the configuration problem on the other Server is still left but this would be an easy oneThanks for the moral support +The command is mvn clean install PcodegenKeep opening just because you said there is a small issue need to fixed +no params and no answer +If it is the case you may also fix itThanks Nicholas +because in it assumes there is a string in the userdata section +This patch patches only one file and should be used on top of am fine with the feature but I have problems with the diff +I have moved the issue to the correct jira project you wont be able to use it for pom interpolation +The error is happening with default DB and with standalone run target doesnt have access to the java namespace +Trying to dig in more +very goodI tested the DOTNET Stack before building the patch and by applying it to a clean folder and using the ST scriptCan someone else review and test this patch before commiting +This makes these methods to evolve independently +Hithis is my first version of a patch to solve the problems using the approachdescribed above +After your all should be good again +Fixed at close of all resolved issues as part of closing items. +Both should be removed and a period should end the messagetextThe fact that a warning is returned is sufficient the exclamationpoint is unnecessaryThe current text isXJ The scale supplied by the register methoddoes not match with the setter method +Closing it. +I took a look at the patch +Open the project in Intellij and attach source to the spring jar fileThe second attachment shows the state of the project after running mvn ideamodule in the project + +This problem was there with the Axis but I fixed for axis +Also are we certain that this fixes the issue? Do we know why this fails on Ubuntu +A quick recompilation sorted out the be of help to someone at some point +As usual thanks Gianmarco for the hint! I tested the code against Windows Xubuntu and OSX Lion and works properly +Committed to both and errors seem unrelated but it is odd to see so many +This is a proposed implementation of the described feature +This will be all new code other than minor changes to include javadocs +The correct syntax for describing a column isDESCRIBE table columnorDESCRIBE column +Minor patch to rethrow +It is not possible to deploy more than one processdefinition at once +While implementing it I thought that serialization support out of the box would be nice +The suggested changes in had to be rewritten as the code has changed significantly compared to the patch reference +pushed +Set a flag in the data being repaired sstables to show them as repaired +I will now create a version of the patch that does not move rename or delete any classes as that dilutes the patches +Where should the file go +Please if still a problem +Although the fs itself has no limitations there is still value in having the flexibility to establish limits +Swapping version and the Muse service around results in the correct path and the following log message loadINFO Two services can not have same name a service with already exists in the systemThis appears to relate to the service group handling but I havent dwelved any deeper as of yet +If projects A B are and project C is a war project +part of release. +Andreas re Whats special about the setupnothing +Committed patch revision +Yes I checked in an update that I think will work +Ill let you know when the patches are applied +Yes this is a known problem and I have code to fix it but it cannot be checked in until all the broken conf test cases are updated to reflect the fact we are not serializing the carriage return after the xml declaration +It is meant to call those methods on the which is aBut is also an and thats why should not only call methods on the but also call methods on a SAX +Im assuming this is related to which like shouldnt useAs written an instance cant be reused between streams +When the application terminates iterate the properties of localStorage and persist at this point +Looks good to me +release done it is in deployment annoucement on dev and user mailing lists will xml seems to be missing from update site +Attached the patch file +Closed. +This should do the trickant Let me know if you need any more help +The only reason to keep it separate is we dont want these jar files already in classpath on all nodes +Delete portal that has just created on step +Your copy must have failed there is no template in your publichtml that is done let us know and we can enable the custom LF +is opened. +Forgot to specify component and versions +Dhruba says I vote for using exceptions internal to the to keep internal interfaces clean and elegant +Thanks Shay +Content Assist +Also makes it easy for others to plug in their output mechanismId like to commit this as an interim today so that Jeff can sync up and work on his side +Can this be doneAlso it would be nice to have the command as follows +It looks like is moving around +and am very happy to confirm that the attached jar fixed it Thanks +Verified JBDS +I think you might be right +Added to download page. +MingThe expected version of an znode will not increase once the node gets deletedLike if the region gets balanced then a new znode wil be created then we will have the expected version as again +Thanks Timintegrated as expected +Ive traced the code to the is method which fails with an overflowunderflow exception +I doubt this isnt fixed already +Ah sorry for not updating the issue after committing the patch +Great! Issue has been fixedVerified on. +Made the following changes renamed method get to get added new package element and moved utility classes to this package the following elements were added Element +Id like to push it in for you ok? Thanks +This not a bug! This is how to do it in case anyone reads this +Simple patch +Create another Volume from snapshot Note failed to create volume +Now the all components have tables with facets example of usage +Patch adds a Screen to receive Inventory By Just Browsing Products from the Supplier +Here is the application that i referred to in the ticket +with a high priority +See my comment in the code review +The will move to the public space after minor and. +The docs have been updated +The pull request was merged to AS master +Fix npe in mini hbase passed and no more NPEs +It doesnt seem like that you can cast it directly to +Perhaps adding optionalsuggestion for such comments would be good instead of cycles of nits and time wasted on addressing them +Is there any workaround I can do to still use encruption but not have the error we were getting? the very least it seems like we should document the incompatibility between IBMJCECCAencryption provider and Derbys encryption requestsIdeally if we could understand the incompatibility more clearly it would be nice tomodify Derbys encryption code to be compatible with the IBMJCECCA provider +So restoring patches in JIRA actually works with help of infra but this is useless here as we would not use it in our source tree +the tests addedRunning regressions + +Committed revision +Must be . +The upcoming File related changes you mention are there another issue describing thisI can confirm this problem is present in cordova aswell. +We could have global XSRF protection on by default and allow application developers to selectively disable it using a annotation +Committed revision +This problem is limited to the latest x compiler version +Huned could you please post the results of ant test and +Retrying +We dont only work for eXo Cloud we also work for the support and we work also on some features for JCR. +Modified the syntax +Thanks Xuan! +Thanks for those initial comments +Please reopen if this was not the problem +I havent looked into the logic change but it seems with this change the following are the two valid options mergeIndexes is an update command and the target core should be online mergeIndexes is an admin command and the target core should be offlineThe first option is close to what we have now +Ive changed affects version as and removed the log from +Thats also why I opened the thread as a blocker the compressed Gb file is Mb large +Verified on. +Hi FreemanIs there a way to this bug until a patch for this issue can be incorporated into JDK ? An SMX bus hang on SA redeployment is a serious problem from my perspective +I think there are two stages to fixing this problem +and unit tests pass +The does not have to be modified in order to use Axiom it should work +The docbook has been checked into the webservices module in HEAD under webservicesdocs is an ongoing task +applicationidentitycls +JohnnyHave you added the following jars to the lib directory of both MQ installationsAlso check out where boolean expressions are not evaluated correctly +Therefore it would make sense I think to go ahead and make a new request type that doesnt yet exist and then the current fetch request remains the same on the wire and the behavior of it is just a degenerate case of this new use case with delay and bytes set to I think you might consider how useful is it to worry about user specified timebytes? It will add a lot of complexity to your implementation and frankly if I have just the ability to do a that will wait until something has arrived and send me whatever it has at the current moment that will be good enough +Is there a way that we can make webapp resources available in the test classpath and not necessarily on the runtime classpath? I would be happy to work on a better patch if you could point me in the right direction +Test passing in hudson. +Change ij to correct undocumented functionality to behave according to specification +Back to the original scope of this one just namespacing the properties for clarity +Oops missed your last comment where it sounds like you already did this +on the patch +This how globStatus is implemented but not for listStatusHere is a patch that fixed the problem +I will upload a patch with the just committed this +Can you check whether your cache bean in turn has a reference back to the lookupMapper bean? you are quite right +This should lead to a retry to that the number of rows inserted should end up being correct +Attaching the patch which includesAbove mentioned changesalong with and addition of HADOOPZKFCOPTS from templatesconf the correct patch to remove duplicate VM could you regenerate the patch it does not apply on trunkAlso some commentsquestions based on your patch fileYou have added an option to the script to skip hadoop opts skiphadoopopts and you are passing that in all the various places is called and this skipping the setting of HADOOPOPTSI dont think we should make the assumption that people will have the appropriate values set in the env by the config file +! Filipe +unrelated to the bug but you are not supposed to have the MAVENHOME environment variable set to Maven location +As result +committed w minor changes to COPP to use collation key as to use DK even more. +Can you check it +I just committed this +As far as I can see you will need to close the session before destroying any consumers using it +One could see it as a bug too I suppose since throttling is desirable to keep the flow of requests under controlIt is fine to have it marked for for me +Charles Thanks for the contributions +and it seems like having KV and Cell for every method other than those that take paramsWhich methods we talking about here? There a list +Authenticated urls are only used for token ops since all others are secured by a delegation token +This can be solved by removing the observer add in the start method +Passed for me +Changing a xml is always more convenient that changing and recompiling your code +for anything that helps end users properly wire up their +so the dialog is very misleading as it is now +Below test cases also failed with same error is because of incorrect marvin package +No it is NOT +Is there way to use DBCP in Geronimo +We have some stress tests that QE runs that will test that +The other idea is a Web Coordinate Transformation Service which is currently an opengis discussion paper that might be able to handle the transformation +Hello Michael thank you for updates +It is now faster and uses less resources than previous version +Ill target for +Thanks +Same merge policy would normally end up giving different results since the data is differentIf you have the primary index and are building an aux index what you want is a policy that wont merge at all but that you can manually flush at the end of every segment in the primary index. +The reason we backed away from vector clocks is that they can tell you you had a conflict but not how to resolve it +My past experience it is crucial to get those out first and give community time to test the snaphost before pushing out the karaf release pom still contains snapshot versions we need to update these entries to release versions +Generally framework will call +MuahahaUnity has its config reloaded every hour at +It will allow to generate snapshot download URL easily without need of analyzing a rest service to do that with query parameter v +isnt in source repo +Probably a regex matcher +net plugin perhaps it is a stupid question do you clean up the webappsonar directory before each deploymentIf you are using tomcat when you copy a war file in the webapp directory the first thing tomcat will perform at start up is to unzip this war file +Attaching patch +Just wanted to make the code more readable Mark +Rebased against master +Thanks for paying attention for me and catching my mistakes +Hi LauraThey are fine +We did this and it work Thanks Eric +This would be a fantastic idea +Thanks Jing for reviewing the patchI have committed this. +Yes its was the problem pooled actors are not yet supported by the webappthank you. +Another option is create a boolean property to set if the interceptor will be used in the client with token profile +This was done so that we could change to the new APIs in the branch while still keeping the same base implementationIn the branch I took the modifications made to and implemented and extended +Making critical so it gets a bit of loving +This will take care of both create trigger and trigger action recompile because of invalidity +Added a label to the UI above the table viewer +Fixed with revision +Looking at the versions being proposed for upgrading too I found them in the central maven repo +Add an additional fix for in v patch +Without the patch the test fails with the same conditions +Thanks Ben! Works like a charm. +The javadoc tool appears to have generated warning messagesjavadoc comments are unrelated to this patchThanks a lot Ashish for the patch +Made changes to avoiddelay need to change hdfs tests +apply in svn rev +I told you that on the mailing list already +i will fix these later +I do this right now in a copy I maintain for usage within the company I work for but it would be useful for everyone using giraph +If has a better suggestion Ill do thatGood point on the downcasts Ill remove the unnecessary ones +I spent a bit of time trying to repro this with master but failed +The tests are run by the target in other necessary step is to set the values in the file in the main directory +Final version as got jms replication working with osgi testPlease note that this patch do not contains any jms replication functionality +From a users point of view and remember users should not be expected to be ODF spec experts text is always aligned +The lt is important in that sentence! It would read better if the hypen didnt cause a line break +Please confirm that the live site when the update appears is as you expected +Setting the fixed in value to ER as this is the next build +yes its committed and now closed +Caused by the compiler bug described in +Patch to add support for new method Patch with by license +At least we should be able to see which files it cannot delete +I can help translating to esES +My thinking is that it is preferable that one can specify the subset of fields of the buildVersion as an argument to a task trackerThe task tracker uses that info to check whether it is compatible with the job trackerThis way we dont have to customize the +Thanks +I have tested Final release binaries for this issue and the problem seem to be solvedNeed to Linux binaries once the size problem is solved +Fixed with snapshot is deployed please testFor the error please file it as a separate issue with a concrete example +We should also avoid adding to many pollEnrich methods to the DSL +But yeah I think autowiring by type is bad practice +committed +Once a thread dies all values should be collectable +I verified that it fixes the problem +Fixed. +The example you gave works with CVS FOPBatik. +That sounds wrong +Lets wait for it +All of our providers see the operation fine. +But I didnt add that in the patch because that change is not backward compatible +committed this +The current branch does support nested snapshots +Thanks +bulk assign all unscheduled issues to next release +Maybe theres a reason why clobber leaves out trunkjavadoc +I am also going to update svnexternals for srctestbin in for it is pointing out to the commons trunk rather than to the appropriate commons branches +Also are you planning on updating the patch? Is it only the dir name rootDir tmpDir +Updated patch fixing a small problem with class path issues after I decided to move the new caches to. +HiIm having the same issue +I firmly believe that the problem with this test is a different one now +This is an updated patch some bugs were fixedPatch is ready to commit all build test was passed on platforms +I have incorporated them all +Committed to sqoop +Still will see issues with your first case since its on the client end but the second case should be better now +programs are indetified through their names and each program corresponds to a Solr indexcore +Please investigate cases described in forum YuriiThe bug is reproducible sometimes upon two conditions ajax attribute is set to true and ajsupport added for event onselected +bq +Scratch that about the missing changes +pending JenkinsColin one little thing I think you couldve picked a better name for the attachment +Thank you Lars +it doesnt look like the got changed back +Committed revision . +Now its really clear in the docs +We dont have our production case anymore but I can still poke it with a stick a bit and get back to you tonight +Test is temporarily think that the best decision is to get rid of mozilla simulation at all +for the change +The provides localized strings for the templateIt should be placed in the directory srcjavaorgapachejetspeedportletslayout where it is referenced by the patch +I have patch for this but your IT does not quite address the problemSimply placing an annotation on a class does not make all types in the class belong to that namespace +As for using undocumented features or future enhancements regarding how the admin and jdbc ports are used please log an jira specific to the enhancement you would like to see +Patch file for adding locale aware import to that code to create patch was copied to Clay from a external class so is untested and not release ready in this context! +That is by design because we provide automatic dependency management that should work the same way between Maven command line and the IDE +When using replay detection with apache server second request per each thread caused the thread to crash +Is it per application or only one thread in RMI think it should be one thread in RMbq +i made the patch from x +Reopening this this issue as I need this for the C bindings build system. +Ideas patches to achieve the design goal with less hostility would be appreciatedI would have to but I dont think that there is any test instance state used by the methods in this class +It just makes the list of methods too big when people press ctrl space in their route builder +After sniffing I obtain exactly the same results but the last call failsSNIFFERGET secureviewrsspidstatusIdsresolutionIdssorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornone HTTPLocationviewrsspidstatusIdsresolutionsorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornoneGET secureviewrsspidstatusIdsresolutionsorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornone HTTPLocationviewrsspidstatusresolutionsorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornoneGET secureviewrsspidstatusresolutionsorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornone HTTPLocation +Were looking at pushing out RC this week +Closing resolved issues. +Currently this patch can be used for resolving exception while changing order status and generating pdf +we will see how things go with the rest of the tests +Ivan can you add a test file to reproduce those leaks? +It adds an installed couchctl command that you can use to print the port +Hi Wang Xu +this be done in? can support this in one of two ways Store very large message as file on disk outside journal and just store a pointer to that Split large message into fragments then can be stored in journal +In fact I even consider this an unexpected side effect +Im not sure but I think this is fine since the only risk is that we miss that someone else has recompiledinvalidated the plan for which there are plenty of tests when we start executing the plan +So I wantopen from Open with menu +Changed subject to refer to version at Moazs prompting +The patch has been appliedPlease verify +Fixed in ER +still works in my environment +I can test this but will either need a project snapshot or instructions how to build Ivy +Since is closed this issue was created and linked to seems like this is because the request contains maybe we should trim leading before evaluating the bw t reproduce this on trunk +Thanks created a new bug report for the next issue. +At the end it was indeed a very difficult problem to solve +same comment as Matt Hogstrom +For the redirect I think we are using which is included in the redirect URL? Thus we do not need to worry about hostnameip address here I guess +Using a solves the problem that i was encountering +In this case this would be just minor + +Attachment has been added with description diff for has been added with description diff for +Thanks Pat! +Default eclipse editor +Added fix to and trunk +Your patch created a second Server Overview page and we couldnt find a way to get rid of the first pageThere were a few things I noticed in the files you created +I got the correct resultI will attach the code I used for this together with request and response messagesBTW please only raise issues using JIRA and please use mailing list to ask questions +Note a few Javadoc comments need tidying +Tested on all supported DB. +refactor and api though is more important for the current one +Updated patch that uses the spath could you have a quick look at this one please +JurajWill you be able to apply it in your environment? Do you see any further issue relate to this topic +However as a solution Im willing to commit it +These issues should be fixed in +ThanksIve attached a patch for the configuration example +I see the same problem when I look into +Im using Flume on Ubuntu Maverick +Caused by but the question is why the code fails with an NPEI havent been able to reproduce this on my own PC so far so this seems to be timing dependent +Thanks Jacques. +Please review +We need to start the listeners at last so that in dynamic LB the LB can start sending messages to a node only after its listeners have addresses the listener starting order the patch for azeez +Should we provide both enums and strings as keys for countersAt the end they are going to be both in a map as a string I guess so we should support both of them +Let me know what needs clarifying and Ill gladly do itIve been on IRC and the mailing list and weve asked questions about these things and tried to figure it out on our own thats why I decided to open the ticketI didnt assign the issue to myself because I dont feel like I have any idea about whats going on there +Ill attach my Eclipse +this would generate the Integer instead of int +The commands work fine by Apache Derby Run SQL script using ijMy problem is the execution in SQL editor + on this refactor +RegardsUlhas Bhole +it would be cool if we can add unit tests for is a dup and was fixed per +However there might be users out there that are already using sqoop and are fine with this strange mapping and we definitely dont want to break it for them +Thanks +Can your provide a toy input example and what you want in hbase for that row after it is importtsvedCan you be specific with this feature requests scope? Do you just plan to include intlongfloat encodings? Is this in values col names or +NET CF framework and adds +Still open and perhaps good intro bug for a new hacker to work on files examples expectations failures bugs that are real bugs have been fixed +I dont know what to do about this though +Therefore lowering priority +Before i could start further discussion what type of changes does one foresee in the current structure to support Tags and how does it work in tandem with different Encoders available including NONE type +Ok here it isRenamed User entity to +I think I might be misunderstanding the help +I think jenkins itself should detect whether its running under a hypervisor or on bare metal and adjust accordingly +Committed revision +Fix and a test case attached +bq +It is unlikely that and will ever get merged to +Thanks for your comments John +Trying to operate in these situations is hard +Three proposals a monogram of LJ a stack forming the letters L and J and a illustration with steam forming a log text filescript +MaxI have coded the functionality and attached to this JIRA +Can you rebase or publish github branch +i was hoping my interaction with the ofbiz jira would be a learning expierience and it has been +But lets commit this meantime +Link to issue which is really an appserver task not an MC task +During split it didnt so issue another execCoprocessor with new adjusted startKey endKey +Without it I dont see an easy way to supply custom to an if you want to use XML files instead of annotationsFurthermore how does the look up in the local namespace work? Do I need to prefix my entity declaration with a or not? This needs documentation +Please try it fire up again with removing the presist data in that plugin named directory which was under the ECLIPSEHOME +encompasses a deeper refactoring. +I have found the cause of the problem to be due to there being two different versions of class in two separate jars +It seems there is no easy way to access resolved pom properties from a Jenkins plugin +The current behavior is When you start server externally and then startstop in eclipse nothing happens to the running server it is still running +Packages are on Brewrhq +Thanks Mark +Any idea how we can enable analysis for those processes as well? +I did not attempt to make any unit tests +studioeclipsefeatures +Yes we could certainly attach a priority level to each journal +Is this just on the emulator +Please review this one +Oh actually I see the repository was hit thereI think the problem may be the HTTPS URL for that repositoryForget it I figured out the problem here +This is essentially how I found this bug and I was rather surprised to find that my first environment block was ignored it is not at all intuitiveIf there are really a number of defects open around this behaviour perhaps that is an argument for rewriting +Added missed files +Do we still need this functionality in GEP +Unfortunately were still seeing the busy loop +Should I apply this to or HEAD first? Username airsquig +Will check in the addendum soon +Simply put we are looking at the initial in a path component and failing to append at appropriate places +Thank you much +This should probably have some testing around it +Tested +Will commit if tests pass +Ill have to think about what to do here +I was able to create my review +any kind of suggestion about this will be appreciated +agreedalthough standardAttributes of doesnt need atomicityI design it just as a cache +I have just committed it. +This one seems to be caused by the same problem as +yes this issue should be prosponed to wicket or or what ever the next release is that builds on +Thanks +Should be filed under Sandbox +The problem is that the object in the serialized continuation snapshot still refers to the old class before it was updated +this jira fixes all the edge cases that could not address without an HDFS Nicolas I seem to have let this patch rot +David even when using Ant for triggering source code generation theres the option to pass in +Vinod makes sense +Do you have a test case a unit test patch preferably that shows this bug before the patch is appliedThank you reason for the file here is because the file length is shorter than previously so how can there be more data left to read? Weve already read beyond the reported length of the file. +and of course a thorough review of all contentNote that antlr contains two licenses + This builds and passes tests for me on Ubuntu +I tried to start on this however nothing can be deleted without the terms dictionary and the terms docs working in order to obtain the doc ids to delete +This info would not be used to define the client type system +Delay this I want to test such a change on various crappy drivers before releasing it +Patch with test case and fixes +New patch including Toms failures are still unrelated to the patch +Realized I did not make it clear adding the Rack gem explicitly provided the same results on +Not a supported ORB and because all of the old Orbix classes have been removed from the code for over years this is unlikely to work anyway +looks like +Gaetan would you be able to check in the patch for this? If not update defect so jessica can pick it up for am OOO from to +HeyI just posted the info you requested +Can someone download and verify the patch +This way you avoid the side effects you can get with i let a stateful matching a wizard but depend on your architecture replace session by passivable scope in my previous sentence +Removing from patch queue as this Jira can be achieved through linking between wikis +This is a patch fro hadoop +Many analyzers can use a stopfilter and one of the stopfilters params is to enable or disable this setting +Verified in +So at the cell level either no data or probably lots of varying +This patch fixes current Turbine Trunk which has the same problem +Does this mean that the all the committers have voted and the decision is a No for this to be included as a news item? Thanks I dont think we have voted but it seems that books would find a better place in the books page +Thats good to since isnt scheduled until Im going to have to figure out some other way to use Seam Managed in my environment where I cannot get a through JNDI +on adding though dont forget that has to be added as a master coprocessor as well +Im not sure how you could backport this to if you want to keep the bytearray encoder separated from the outputstream wrapping itPlease note that the timings above compare our current with the one I attacched here +is your repository with id internal a group +Wdyt +To deploy update the with and by putting them in the lib directory of the war +Ah thats why it didnt work and thats why I didnt find the documentation for that prefixThanks Im using +I guess we have to roll it back +Exact steps +Adobe AIR issue not an Apache Flex issue. +Ill take a final walkthrough tomorrow along with any additional suggestions and submit another final patch for review + +As indicated its possible to quote a reply by clicking on the Quote button in the toolbar when replying +Actually I still consider this a bugThe interface should ensure that if you call suspend then resumethings will eventually resumeSo can you still provide a test case +So all issues not directly assigned to an AS release are being closed +That is not realistic especially if we want to implement which does the Puts last in batch +In case of necessity of running emma on jdk please use workaround of activating old verifier. +Patch applied at +Fixes the described in pipes a public method in in pipes that allows background submission of a job. +It seems to work +Havent been able to reproduce in unit tests +Closing wo a response from the submitter +Performance looks unchanged but code size should be fractionally better +bq +again and the node is unfrozen and the node border returns to normal +Fix works +looking forward to all this good stuff +The only way this could be addressed is to have a flag to validate that the entire project is checked out +OK but it would be good to remove once were sure that its no longer a problem +Drop in a python script with a deployment descriptor and start using it in your pipelines +A puristic approach would be do have an Java solution +A minor correction to and a merge with current trunk +Deferred to defer of unresolved bugs to +Actually new lines in an expression dont have all that much a meaning to the parser in Java +Right any deref node should lead to an inclusion of the system tree index +Arun the idea is to have the stats being updated in real time while the app is running +If you need to redo something this is perfect time as time invested so far is minimal +At least hereCould you please let us know if it fix it also on your side +SVN move sub tasks complete +Sorry for the long wait +Tried to reproduce this and Im having a hard time +This is just some work a contributor did feel free to submit a patch +No this fix will not be applied to the branch but you can always backport the fix to your own copy of Wicket. +is now used and I dont have that access restriction +Sorry just found a minor problem with my patch will resubmit shortly +is checkstyle compliant +Two quickstarts to reproduce one for each version we tested +The best way is to add an index to the signature instead of fail out +This test fails demonstrating the existence of the negative wildcard query on bug +Anders why are there Java files in here as well +This should cause a loop if authentication were to fail +Please close or assing it back to me close the issue since it did never happen to me any more +In versions of all ws plugins are correct and features as well +Thanks for the quick review. +patch looks good +If either phase is to be executed how can or should the portal communicate the fact that access is not granted to the portletWhat are the ramifications of stripping a portlet out of a layout? Do we need to consider having a Access Forbidden portlet placeholder and if so how would it be specifiedFinally does this feature need to be bundled with or related to the proposed Portlet MenuJ Fragment functionality that allows portlets to appear conditionally within pages based on menu selections or other url criteria +Fixed with resolved issues. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Attached a short example class which reads a mapping file and extracts some of the informations required for in ways +Defer all open issues to +META +Thanks for the review Arpit! Ive committed this to trunk and. +Please retest and close if you cant reproduce the problem +As a Fix for this i have introduced new nhttp properties to be defined at file Socket timeout for Listner Socket timeout for Sender +Thanks Gopal +But Im fine to fix it in trunk too I cannot imagine an application will be break if our are more defensive here will commit a fix soon +Dear Justindoes this mean it will not be resolved at all? our product depends on this issue greatly +I also like the approach to use an exclusive file lockAs a consequence it seems that creating and removing the +This issue should also include any changes to the responsive portlets outside of pure css modifications +Please take a look and see what else needs to be fixed +Thanks its helpful to really understand the issue +I have verified this issue is fixed in revision +Initial implementation committed in HansThe current implementation uses the Payflow Pro SDK so you wont be able to use it unless your processing credit cards with Payflow +fix version +Only fixes for +Ah thanks +That seems wrong +I saw this again while testing and again with but both times only with ibm not ibm +Feel free to include if useful +if youre going to patch for removaladdition of bundles please do so there +At the moment were using the same key for the NN and ndNN but this may change at some point in the future +As example in the examples components there are features The test feature shall not be tracked of course +google code github +So in future we can consider two options update this rule to also cover declarations or create a new rule +Is progress still being made? Does it need to have the fixin for +current log file number +Wouldnt the python analog be qpiddispatch? For me that also more sensibly matches the pattern with qpidmessagingBy the way you left a bunch of empty directories in the subversion tree +The routing strategy should be external as making the splitter decide the channels it has to send stuff seems a bit too much and hard to configure +Hi CosminI think the most recent patch is good though it will need some munging to apply toIf you can apply it and test it out for us that would be great +Work largely done by Ivo under. +undef GLIBCXXHAVESETENV It seems to be connected to that error +I guess a look at ant and how they the forking would be a good idea +closed +Ive requested a new wsdlj service release from the wsdlj project admins +Im not sure if I understand your questionThe difference with the previous situation is that I first retrieve a reference to focusManager and then check if its null that check wasnt present in the previous version of the code and sometimes caused an RTE +It should be enough for the first iteration +This was fixed by +here is a in rev +Most probably there is a misconfiguration on your side so please use users mailing list to get help + rev trunk rev thanks siqi for the patch! I click disable the graphs under the view cant be seen +my initial comments and thoughts + needs to be split into and m fine with moving factconstraints into Guvnor +I think this needs broader +Added code guard to prevent error to pop up. +I attached a patch on Please review. +New patch created with latest source code from +I think is related or duplicate +Good this will be independend from the occurence of an element +One interesting idea from the block replication code is having priorities for replication work based on the current and expected replication factor +This patch adds the file which is a conversion of the jdbcapitestRelative test to the same package as the latter +Hiram I get the same results with both of thesea b SVN checkout ofRepository UUID Revision Can I give you any other information +AST prostheses +I have added some functionality for retiring process but the request failed at server side +We can revisit this later +Verified with JDK +Regression test committed in SWING module at repo revision r +Looks good to me +Thoughts +Command line tools would be an easy way to standardize yes +To work around the Mb upload limit on JIRA I had to split the zip into two parts + the modified test manually +During compaction we try to keep the column index in memory +Can we close this issue now? +Hello GurusAny update on thisThanksPranabI believe this is now fixed +I have been using this without any issues for the past few weeks +I reverted this patch on local secured cluster and it works +Will switch to a snapshot when it becomes available +Rails startup time still a major impediment for me and I consider it far from fixedrake environment in my Rails app takes seconds with +Please review and give comments +Are you looking for something specific? Its still running smoothly parsing documents per second +modified to use Exterlizer and Factory annotations to make calls out to methods in a feature +fixed rev can be tested using trueI wonder if having a mojo parameter to activate this could be betterWDYT +and +Yeah someone first needs to commit +Still broken in for the patch +Please let me know your ideas about the service implementation +I had already started committing +I first assumed it would be related to the fact that Im using but I tried with a brand new repository on and had the same resultI create a folder with documents using +Added a missing space + +Please the tag +what error are you getting +Do we need to shutdown the master in this case? or we can retry +Things may have changed in the codebase in the last couple of weeks leading to some test failures +I agree this definitely needs fixed +Seems the new jar wasnt in the repository when I did the check +No test provided to try to reproduce this issue recommend upgrades to the latest release to capture all the current fixes. +Im not sure but I know the big change there was to have the save the in the Request and have the UI tags use the from there +A quick vote on the today resulted in The import can be realized during the upcoming weekendThanks a lot in advance! +New patch that requires and that adds fixes for locking in and better handling of +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Apache Giraph would like to use for interacting with Hive +No it was supposed to be backported +fixed in if the group attr is present to support testng +This patch fixes the issue +stays in other modules? I am confused because from the above description commons module is a separated module that doesnt include etc +Proposed fix +Is there any change solution proposed by Marcel Overdijk could be bundled in springmvc? I am talking about possibility to additionaly annotate method with which limit usage of such method only for specific actions +Its been a long time ago lets forget about this issue unless it pops up again +That would be what I suggest to do as wellRunning tests now + +Oh and of course +I guess I personally always like unit tests because they give me great sleep at night +Any recommendations? Maybe only convert to if the OS is windowsItd be nice to fix the problem for Windows users tooWhy not try fixing it inThinking further perhaps this is not a bug +Added Professional to also need Enterprise +should be fixed with rev can a windauze user confirms +The is just to make the compile errors disappear rightThe wizard will run without it on the classpath correct +Hi EvgenySorry for that +Thanks Charles that fixes the tests Except for the odd failing locales that you noted as +I verified on both Mac and Windows +Does the C library build on Solaris or do we have to add this there also +batch transition resolvedfixed bugs to closedfixed +Updated patch for current trunk +That was the only information I got from the plugin user interface +Defer to all open issues to close all resolved issues in preparation for and releases. +Also we should add a unit test to detect this going forward something to the effect of actually cooking up a base conf file with fields of the same name from different tables and then testing that the SQL produced is correct and that there is no exception when trying to read the field names +Under Jiras project Mail Configuration for Pluto someone changed the from address for notifications to note the extra j +Can anyone answer my question? Thanks in advance +How can it not be a bug that the is thrown when navigating assocations on an entity that is attached to a Session? If thats really not a bug then I am seriously misunderstanding the issue and that explains why I am not good enough to submit a patch for itRegarding your frustrations I do understand how you feel that way and it is surprising that nobody has contributed a patch in all this time +Added live cells to Histogram patch for already adds tombstoned to Histogram +Im testing a nightly build today +lzo patch for review + +openjpa plugin does not depend only any version ofit just needs on during also have no problem run the build with openjpa and the immenent version of java are you using +what version of jars are you using for the client +Do we need to consider this JIRA for as well and apply the patch to trunk +Mahadev Flavio Thanks for the commentsI agree about renaming from Peer Learner isnt a bad name + +Actually I assumed that the test focused on the patch in the changes section of the jenkins result of build +I have applied them in Rev +Sorry it is the fix to the problem in the FIELD granularity repl tests is fixed. +I would move the to the top +JakeThank you for recommendationAveryThank you for informing meI post my progress of this issueRecently I have implemented and tested a lightweight RPC implementation based on netty and which resembles to + first comment is wrong +Implemented iyappans comment +The test should use the class and run with client too +Created Need a distributed file checksum algorithm for fixed a bugAll the patches up to now implements option which is our usual approach +Oracle Email Signature LogoAndrew Rist Interoperability ArchitectOracle Corporate Architecture GroupRedwood Shores CA +. +Hi Uma Sorry for a delay in response +I think eventually we should only allow the users to setup a so they can only set classes using the existing hadoop api and cannot create any of these objects +I was thinking that maybe taking the locale from the machine isnt the best idea always +this patch implements +Closing as dupe of. +bq +if This is ensured by fencingand the standby become if the server of active Namenode is shutdownfecing active failthe standby is not become hdfs is not used + +Updated to address Sureshs comments +This tough introduces a kind of inconsistency since you still can add the threshold attribute to the metric element of the event element +Talked with Prasad and he mentioned its possible to do on the metastore side with runtime translationBut I dont feel strong to simplify the name of the serde at the cost of adding additional logic to complicate the metastore logic. +Which you can set to false to allow duplicates and then filter those in the route and handle differently +I reverted the new version in as this broke logging into ap or ar the front page had a bad exception after logging inSince this is a new feature it was better to turn it off until it can be investigated further +Therefore you can create more validators for your beans +Verified. +Taking this a option to allows specifying one or more directories to be used as a the test plugins and conf file I used for manual for the commit Jarcec +etcsecondlyshould I call include times in the same script +I created Ln version and affected the ticket to it +Please this and Ill get on it this week +In fact after changing the documentation I started to add a param to the tag +The fix for the canonical file name use on open is in not sure what else can be done here if anything +and should have the same change. +usrlocalbintrafficcop +I checked in the last two patches and this appears to be newBR into in revision after releasing +When will you have a struct name and not the type of its fields? Is there some use case youre thinking of that needs this +Thanks +Changes in trunk r set response header to authenticate if repository is think everything is done here +The problem is that Axiom is not supposed to invoke a custom builder here +Thanks Kan +I do have the wrong file +This issue is also affecting me +The max doesnt seem to make any difference whether I set it or not with it set at or not set at all I can query rows but not rows +Hi Ive tried the test on Linux and Linux and Harmony passed both tests +thanks. +Is it necessary to have a single file with exactly the same information +A patch would be great or else someone else could pick it up too. +We are also only using currently +Well include this bug and its test cases in that work +Granted we do log a little too much but the correct fix should be to selectively downgrade so that important transitions like nodecontainertask failures are not missed +this is the right BJ rev +Closing this issue because the remaining failing test case is not related to Java under WindowsThe new issue has been opened. +Most have historically failed and the rest failed due to too many open fds +updated patch IO directory is now ioalso having trouble with test being off by and failing see that theres a JIRA up for this and the test is at fault? so Im not worried about it +which will create instances +It seems that Axis cant get scoped proxies from Spring without adding the service class in the +I thought it best to keep it consistent with the contract of Complex divideAre we ready to resolve this issue now? If so lets remember to update with a complete summary of what has changed +Will you revert it or should I do it +Patch. +It cant be that bad +Please consider apply attached patch +Implemented semaphore solution and added test case which triggers the problem without the flush +Yes the fix is OK it was a typo when I wrote extra character in at r. +The writer doesnt commit anything the transaction is managed by the Step +is now out so closing all resolved issues +Of course in the case where two datanode instances are somehow servicing from the same directory the reaction would kick off for all blocks that get stored into that space so locking would definitely be necessary to make this work +ThanksI added the patch to the contrib directory as I dont have a way to test it +Arent you worried about your files being magically changed without you being aware of it? Yes a backup is made locally but personally I prefer to be explicit and know when I made a change that the poms and then let an external tool try to them +Thanks a lot +Can you then run the crash recovery tests with just and see if you get this same error +PR applied in Beta you trying JBDS or JBT +Nevertheless the patch is ready for review +if temperature sensors exists and SICS is configured display normally +I committed your patch to the branch just now except the jars so you should see it if you sync up +You never know when tickets might come in for them but they wouldnt be priority anymore for a release +Weve resolved the documentation issue on FFOS side I will update the code to remove the getAll hack +Earlier it kicked in only for IBM no it will kick in regardless of the JDK vendor as long as a Java compiler is used with Java librariesI also added some extra debugging printsPatch ready for reviewThe approach in patch looks reasonable to me +Committed to tez branch +What is the status of integrating async triggers in Cassandra +The remove method in the interface is poorly defined not currently used and in my opinion superfluous +You on the other hand can use thatPessimist though I am I think that if CXF was adding junk all the time wed have heard about it from other people +Part of a bulk update all resolutions changed to done please review history to original resolution type +HiCould you post the source ofBy the way please post in one of the mailing lists before posting a JIRACheerss a page in shouldnt be a problem with the Totals class being a nested class +Can the patch get applied +Need to do some lookups to see if the avro side has similar facilities +bug fix version to trunk +downgrading as only happened but the server can not get the response with tpcall to advertise the admin service. +I didnt think that this was true after bdceccddaaeefefcae Hugo changed this and AFAIK it wasnt reverted +If you wish for this issue to be supported you will either need to create your own custom avi parser within Nutch contribute to the Tika project or else wait for the tika developerscommitters to support this format +Ive heart Martijn is working on the test case too +Retrieving document at httplocalhostwsdl +And of course performance these can be subtasks for BETA +Closing +I have looked through the implementation of and I didnt find the place which could cause the issue on this specific platformTony could you please look into this as you clearly have more expertise in ICU? Thanks +Please also confirm that you are using the latest version of +checks for unused imports etc +The attached patch makes minor editorial modifications to the Vinods changes and adds one point about incorrect configuration in the hadoop tarball +Jeff can you point us to such tools +Can take the union of addons in and and remove anything else? Either that or perhaps we just find a way to not delete the addons automatically at all +Since switching to maven and using filtering for the properties files it was able to use maven profiles to switch the database to run against +Objections to committing to +from unpack to here +My bad forgot to fix as well +Thanks Avery +Attaching a simple patch which will set a Locale object in DVF +Fixed as part of issue has been fixed and released as part of release +Its now optional +I guess the simplest solution is to change s populate method to ignore parameters starting with multipart +Is there an actual problem? Ive been using Maven to build CXF for a while now and havent seen a problem +Will refresh patch up in RB +this is derived from Im posting it so we have something to vote +Committed to both trunk and closed with release. +The error code returned when you remove a cluster directory create it again and then use it for another allocation is +I have adjusted the to reflect this +Maybe we can add a +It is unlikely to make it into ESB though +Given the anticipated amount of work that has to go into this would make sense to track these changes in a separate branch to avoid the one huge patch that touches everything at once problem +timeout was reported in +Note that Maven SCM does not know anything about the context in which it is running +sigh +Never mind figured it out ant compilejute +Closing resolved issues for already released versions +I think thats what the escapeAmp attribute is for +Thanks Pradeeps diagnosis +Its a project with EAR containing WAR and a set of utility +Hum yes thats possible +I introduced the boolean so as not the change the behavior of thecall to fill from since that call was not involvedin solving this issue + committed +Thanks Prasad +have you done any benchmarks of the two options +This patch is for branch only +I just committed this +Thank you Raja for the review. +We dont commit to implementing enhancements for a certain release anyways so only open bugs are relevant to the release +This would allow for reuing the same content +Ill try to see whats the exact trace +class files in the physical path +Oooooops i meantIt seems that when yourre using an enum in your command object the default nullable constraint is true +My first change to open source project +I think this needs to be looked at againJeremy +Works for me +complete after rev . +Needs review for locking holes and some of the TODO comments answered still +open httplocalhostmetamerfacescomponentsrich +I cant submit as it is work code and not complex enough to be worth running the legal gauntletI have just implemented a that calls flatpack directly and knows about the special header and footer namesThis is a low priority issue but I think it is significant when you find the limitations of the camel components mean you have to write your own with slight customisations +I added your proposed messages as detail and set the normal message to the standard validation error message used in the whole properties file. +attached patch there some solid documentation about thiswhat is the diff between VS? +As devs we would prefer to run off the source I think and it seems best to keep the distrib build setup as close to how the solr checkout works as possible doesnt it +to be clear if this issue was causing larger issues with deployments being unable to deploy i would mark it as critical but not a blocker +Also we need to make sure that reversion of the above wont affect any tests +not currentlypatches welcome +Attachment has been added with description Example has been added with description PDF built from example has been added with description Patch for orgapachefoprenderpdf seem to have added the entire Java file instead of just the patch! HopeI get it right this time +svn didnt capture the new files Liyin added +Lets leave it at and +Committed +This patch file fixes both issues +Emmanuel this is a patch which adds Stefans tests to the SVN repository in branch +I agree this makes sense to enable efficient bulk operations and also fits in with a past idea I had about mapping foo to paramfoo during aIll give it a couple of days and commit if there are no objections +This now happens following the remove call +This part is fixed in trunkAbout this function do we want that upstream or only couch +Does this sound ok +Doesnt Michael Neale have the AJAX toolkit development skills I remember reading a blog about his uni project or something +This makes it disappear from Release but its not lost +Is it not possible to force the constraints on LFS as well? I understand that it would be best NOT to add constraints if possible but just curiousPerhaps this needs to be a special case in webHDFS that specifically uses DFS and if not then we throw an error or something? I feel like we should be able to access the DFS concat API at the least +Assertion seems to be untestable and thus I drop it from +WhenIf yonik finally pops up here he will have some good info to add I think +Unfortunately I removed the svn from my machine as soon as the git mirror is updated I will try itthanks + problems in problem has now been fixed by deleting the storage when the shared library is unloaded. +Vitalythe patch attached to this bug cannot be applied unless you sign the Apache license +svn moved. +for the patch. +encodes additional information to allow readers to skip large arrays and maps faster +This should go to groupId unless you control the whole for the other Carlos +already fixed +rev page controls are always shown on all pages the list will always consistently show +Can see ERROR status and no dropped object info bellow the drop area +Todd could you try your runs again with the latest code +Example schema exhibiting this have used the base type to convert the lexical value to the actual value then the pattern facet would be used +Hi Would you please try my patch? Please decompress to module and run modulepatchThanksBest regards Paulex The fix looks good +The same problem is in JSF Tools Reference Guide JBDSguidesversionsenjsftoolsrefguidehtmlsingle Figure +Fixed in revision +Must be validated +Good point roman yes its certainly okay for small clusters teragen rows of bytes bytes KBteragen rows of bytes KB And the default value is so we are in good shape ls altrh datateragentotal K +Verified. + +This is not a problem in +This is what are users are used to doing +Attached failing has been commented out in Revision +The docs recommend reusing a single to avoid some of the overhead of opening new every time +Attached original version +This applies in fact for Java DSL as well +Its not really an Inflater either it may be worth either supporting that interface or using an Inflater member rather than inheritance since it only calls public methodsGood point +However I did make a change anyway to similar to you patch expect that I just used the s readFully method to ensure all the data gets pulled off the wire +though this is applied when I set the jar plugin to use the archiver it still generates the wont fixing as will remove it by default +Patch in sync with rThanks Akshay +bq +This upgrade should also convertmigrate templates to RF is still issue with custom converters and validator as is already released months. +This included new JMS destination queues packaged with the application to replace the test queues removed from the server configuration fileTested the changes against ER and both quickstarts build and deploy with no errors +v patchset guards against a possible NPE per +And if it will we can just include that fileI agree it was a bad argument heres what I was trying to expressMy concerns rotate around slowing down the development of the tests themselves which are not sexy to work onand always need more improvements serious cleanup work complete rewrites etc +Fixed in changes to +Looks like it was fixed by the patch from +Note that the dialog is not cancelable at this point as I havent found a good way to interrupt a HTTP request but its a step in the right direction anyway +get it +Seems simple doesnt it +I think it is an automated trigger process from svn dist +Ill study more +Updated description and summary to reflect new understanding of this +Please comment +The doc is ok I have fixed the xml files +I changed the dependencies but I actually wasnt using solr at all +I think wed have to do something where shutting down the NN with would result in and the NN would reopen that file as upon restart +Mass close all issues in resolved state that have not received further comments in the last month +I just did a quick test but I wasnt able to reproduce itCan you give more information like your settings file your ivy file the part of your file where you create the fileset the output on the console much easier to diagnose once I had a working cachefileset to compare to the one +Yes the evil behavior Id seen when I wrote this up has been corrected +The basic rule is that INFO level will just be used during startup for logging during operation only DEBUG level is used +This mojo offers far fewer options than some other plugins and they still manage to keep their code remarkably clean +IMO instantiating a full blown server for the factory test would make the test run way to slow +In now the that stores multiple values for a key is initialized with the value tuple in the hashmap is initialized with expected number of fields +In what comment? Would be helpful to have it in a comment above getEnumWill do! Its in the log message not commentbq +Replaces wrong used constructors of +Since there will be updates in web services and hornetq and this issue wont be affected +Here is the patch to usingperl e print timeto get the timestamp in seconds on SolarisI have added a new function in to set the start time and updated all the scripts to use that +jBPM identity is still very basic but has the delegation methods that are useful +I dont know what use case there could possibly be for having multiple bookmarks with the same name in the same folder +It seems to be backwards compatible too +Although I still believe that having them added as fields to the Documents properly whether it be at the level or somewhere higher means we can more easily use the values in other areas without having to write custom code in each place +Setting worked for +the parameter gets split up +testPlugin installPlugin ccygwinhomealextestPlugin installPlugin homealextestPlugin you send me a full patched startGrails script as Im having a little trouble applying your applied +Committed to theWould you like me to commit the fix to the +Attachment has been added with description Test case to showcase an SVG using lines starting around has been added with description Animated SVG document used for the test order to try to resolve this issue I wrote a very simple animation enginewhich changes the DOM directly and modified the test application to use thissimple engine +Thanks +Patch restores rake test functionality in lib definitely need the dup effect +Please provide a small program showing me this behaviour it should be quite simple to isolate +It appears that the master does not clear its return message queue after it sends the close message or it keeps finding the close request in some other data structure that is not getting cleared +Proposed patch +As for whos who in the testing business usually defines the standard and the others follow suite +Closing CNR. +Ive committed this +might be that a recent merge commit reverted the actual fix +In the meantime any feedback about the attached code change is most welcome +Daniel and Davide thanks for your very helpful posts +Attaching the patch for +How do I read the above? Its same amount of kvs in each of the files +I opened a specific issue for this +thanks for the patch! +I think you meant to open this against Xalan +Here is a html version of the same document +It looks like Nick added SHA in fe so well leave it at that +please open new issue I think we now mixing two problems into one JIRA am closing the issue +Now I get Error File filehometamirobugdrop not foundGood message fix verified. +code review +If I am outputting to a DB using a DB Output Format I now have to set two configs instead of just one to make this work +Ive now updated the WP part to work with the same Code as WP can you please verify +I applied the patch +Reopen to change the resolution. +Screenshot of wrong output in safari on german mac and patch with supposed change +Put it into Id say +Any ideas whats going on +Updating the Apache Felix dependencies and plugins appears to have fixed the issue for JDK +Done in branch. +Is this JIRA still relevantSeems not +Initial cut at is now out much of the implementation details above were not necessary is just another Plugin! Very simple. +Any insights or helpThanksAre you actually running or are you running something straight out of the repo +Perhaps a string should also be added indicating why all results were not able to be returned +There very little difference in bench mark results +I am sorry it isnt a maven project +richdoug is this issue on track for R? If not please push out to R +Thanks Gunther! +Applied Thanks for the patch! +The assemblyassembly behavior that uses the timestamped JAR name causes a disagreement with jarjar thus resulting in a zip package that cant be executed since the jar names in the manifest are simple names and the jars actually delivered in the ZIP assembly are the ones +The second stacktrace mentioned in my original post +Furthermore with native code it is much much easier to have code +to branch and trunk thanks for the review Stack. +Let me know if there is anything I can do to help resolve this issue +Finally I got it and it really shows me that with a transaction the data are not Username labo +And some more questions +if it is EPL licensed it needs to be in an xpl folder +Oh wellAnyway both jars go in contribbenchmarklib +IMO document thoroughly the lack of isolation when paging and punt + +This issue has been fixed and released as part of release +From reading some of the source code comments it seems that at one stage we didnt generate the fully qualified names all the time and the imports were required +I think is required as well +From a software engineering standpoint that class is the one responsible for raw comparisons of tuplesOnce you factor improvements in that class all the other places that use it get the improvements for freeEven if we go for a special case I would inside rather than outside itThe fewer the places that make assumptions about the serialization format the easier we can change it +Let us know if you need help with writing them +If there is a problem please open a new jira and link this one with that. +However body responses are still not cacheable +Thanks +Any other objections to allowing the following short form alternative license header? Licensed to the public under the Apache LicenseFor the record I continue to object +Its hard to think an already started job is changed dynamically +lgtm as well +I strongly believe there should be a readme even if it only contains a link to the wiiki +I cant reproduce it +These can only be stack allocated and have identical semantics similar to java weak reference intrinsics weakRef softRef and phantomRef create the appropriate type of reference to a heap object +The file link creation seems to increment a reference counter on the source file thus you need write permission on the hbase data files +talk about consistency between wars ears rars and xars +Id like to try and Postgres before we make the change +Hi Hairong is in the patch but missing from the commitThanksEli +NET +still you need a logger impl in tests like logj +cant you leave this issue open? i can appreciate that you dont have the time to work on it now +Looks goot to me along with +Would be nice to be able to add text to the methods +Patch looks good nice work +This is to make the token more opaque as it was originally designed to be +Since this is not core functionality and more of an addition to the module i am ok with committing the code as is +Im going to add some additional INFO logging to make it more clear what the RA is doing when it cant find the destination in JNDI and automatically creates one to match the MDB. +Here is the patch with proper NPE handling +Detailed description of the patch + +. +If the problem persists please the issue and attach a demo project to demonstrate this +What will happen in a Tomcat environment since we changed the entry on services +A flag in sounds like a good option +but it seems like a Maven issue any ideaI mean i have to declare the arquillian core and other indirect dependencies like direct dependenciesWell i having right now an issue with Thread already associated with a transaction! in this same code i will try to declare transaction extension explicitGreets i can run my code without transaction and it works but i cant catch the cause the transaction is handled by Arquillian so in what way can i catch a if i wanna test that +Wow you can do THAT? Amazing! DohYeah thatd be excellent thanks +Kishanthan Your last patch seems invalid it seems your patch contains code formating changes can you resubmit your patch +Or maybe which Im using requires more memory +But I think this illustrates my point in about the relative usefulness of unit tests and system tests nicely. +not if there is a copy constructor which I added as you suggested +Definitely need to improve it +Hi KatheyI believe there is more work to be done on this issue +xmlThe only reason I asked was that there was previously an issue where tests were timed out if they ran longer than a setting in the configuration +Also their timing is indeterminate which might bother someoneIf you look at the patch for a schema the patch sets off cleanup when a core is closed +Including the individual failures as they happened at the DEBUG level seems reasonable and useful +Just ignore me +Sounds good +OK Im resolving as optimistically invalid +Unsetting Fix Version for unassigned issues. +There might be cases where multiple nics on one machine are load balanced that way but I would think bonding does the better job since the RPC connections are relatively compared to things like http requests +Or is there an existing junit test the repro should go into +Were doing something similar where our convention is where is the build +Hi ToddI have updated the patch with your you have any comments with this new patchThanks ToddEli Do you have any comments on this patch +The fix would probably not imply API change so it can be delayed to +I tried to make it not sort the import files and not very successful +When generating CSV it is common to get blank columns such asdatadataThere are columns here and only two are empty +Hi DagThe second patch looks simpler and more elegant +Which classmethod do you mean? In general I am thinking of having client allow users to launch apps in managed or unmanaged modeI think that is a good goal but this change doesnt preclude that at all +I hope Im wrong +Marking this issue a duplicate of because the discussion is more active there. +I marked this jira as a blocker for to ensure it gets picked up for the release +I think this is actually fixed but doubt Ill have time to prove it before to be fixed and no updates to this JIRA have been reported +What you are saying is that if I put two portlets on a page then I have to know every request attribute that might be set in each portlet +i execute and not the that what you meanmaybe its not possible for a java to use the path file +backend for some applicationpresentation logic and you can whack the expiry up high? Lee can you comment more on your usage scenariosWhen updates get involved things start to get a lot hairier though maybe if you want to turn caching on your understand the associated risks of seeing stale data and can tolerate that? Again Lee do you have comments on this? +This patch works +I think this is ready to commit we can address the variableconfigdata issue in a different issue or smaller patches +Have we verified what other dependent jars this zkClient pulls in? Weve hit the multiple jar versions problem before +I tested with on Ubuntu as well and I think it should fix this bug +True this does not implement a true cache but it will provide significant performance improvements in the common caseI just dont like seeing the SQL generation process taking over of my programs execution time when I simply execute the SAME QUERY OBJECT over and over againI would have loved to fix this issue myself Im just not yet familiar enough with the inner workings of +reopened because solution not optimalreverted above version of antlr to assembly and +ps I was entering your jira using username test test dont you want to protect it better +thanks suresh. +is still honored but deprecated +Claus youre absolutely right my bad +Commited to trunk with revision . +im still seeing the old behavior in trunkbeta +Thanks DavidYes that works on Unix systemBut that doesnt work with Turtoise SVN on Windows XP even if I set Proxy settingsI will try at home Friday Night and give report. +The only thing that remains to properly handle ordering of duplicate versions isOne solution is to include sequence numbers for every KV in the +That way we would do much more frequent deletion of archives but we would delete a lot less each time +Thats the whole point of the analysis +Modified title to make it more visible in Craigthe isnt bound to a window so its right to add the null checks +As such it will be possible to delete the classes in the +The contents of libruby are moved out of the root of the jar into lib this allows the existing load paths to work correctly A pair of targets for and produce a complete jars with only the lib files for that specific versionWith this change rdoc works correctlyBut because of the fairly major change in structure Id like others to test this out with rails monkeybars and other embedding cases +Accidentally pressed submit too early +Nathan Roberts requested that I include it in as its absence is causing ops problems with +Hi AndreaWhat you are describing dramatically changes the semantic of the MEP based on whether sync or async is being used +Can I reopen the issue? I dont think a message like not a number or not a date is very usable since I cant change it in the xml file +I am planning to release version soon please try it and if you will have issues please comment or reopen +Drop this into libbuildr +on the changes for +The other questions are still relevant +However if something canshould be addressed in feel free +Should work out of the box +Committed revision Committed changes merge method renaming to trunk revision investigation showed that supporting open on clonereadonly fcks up refcounting +Would simply reducing the FP default be a Good Enough solutionCreated to do this in the meantime +Addresses all the code review comments +Its also a feature Ive never needed. +My previous comment applies only to being rendered in a portlet. +To deploy as a singleton simply declare the as a Spring bean +It looks like we are detaching the returned entities before they have been fetched +Im having the exact same problem while using hibernateDid this fix get properly promotedmerged to version +Alexey please verify +Is this ticket still relevant Simon +I think the reason export to ant doesnt work for the ear project is because it is not a java project it has no source files and nothing is on its d even go so far as to say the export to ant feature is not a action but rather an action for regular java projects +Triggering Hadoop QA again +The comes with JCS persistentstore +The Flow execution history subsystem proposed in could help to solve one of the major drawbacks of the current namely the fact that iscurrently useless if a default is not added ability to query stats on expired or completed flow executions wouldallow and related implementations to determinethe flowId of an expired or completed flow for which no default is defined +Closing tickets +This made it +How much would break if we applied this patch now? I think we need to line up the osgi and geronimo states better before we can proceed much further removing gbeans +bq +But I got theerror as TNSlistener does not currently know of SID given in connectdescriptorThe Connection descriptor used by the client wasorcltrue +Its not possible to do this with Hibernate unless you use something like Hibernate shards which is still in as of Burts plugin which solves this issue +Yes a workaround would be nice +The attached a patch removes the remaining Vectors and replaces them with +For Id like to see decom work like rt and take a token parameter but I dont want to break existing jmx tools for so lets just error on extra params there +We have systestsjaxws test module so you can add a test like there +All relevant patches have been committed +Patch for this feature +I will confirm soon + This is really needed +Apparently I failed to commit the modified in my last checkin +The parser will no longer try to compare headerelements with week days or anything of the sort +RTC task move of all unresolved issues from to +Using above patch and running with threads was able to read at second +A lot of people are forced to use scribe since theres no other option and for a lot of them this will be the only part of their system in C +A lot of changes but now is as it should be +But you do have the local cluster running on your local box right? Jarek Yes +If I have Test methods to one of the test classes I can clearly see the methods being executed in parallelI am running against a Selenium grid environment so I can see when tests are running in parallel +Verified in +bq +Its the first time anyone reports this and its not like it corrupt data or anything bad so Ive committed to only. +Adding lineage service extension to hive binary +heres the patch for +Thank you Eranda for the fix! +FO files with exception of main report produced by Dump +Phew it works just fine now +it isnt satisfying +set behaviour has been fixed +ok moving to so i remember to bug you in that time frame +Ok thats much better +Namely allow project names in p switch to use file separator rather than a colon for the project patch is to list projects in helpprojects using file separator +field reference jpa class +Thanks for taking care of this David +Fix is in +Leave the issue open meantimeSure +Closing +It currently returns but it should return + +Additional InformationODBC connection used SLQServer driverJava Environment m tempted to close this bug as an environment issue +However when adding two identical keys to the same mutation or when two mutations with identical keys show up in the same commit message we get a different behavior +cool +However adding a line in the reference manual would be beneficial since I had to figure out the missing dependencies myself +Resolving as a duplicate ofthanks Isabel. +Committed to andThanks for the patch Jeffrey. +Broken with the new gshell assemblysrcmaindistributiontextetcSending gshellsrcmainjavaorgapacheservicemixkernelgshelladmininternalSending gshellsrcmainresourcesorgapacheservicemixkernelgshelladminetcSending gshellsrcmainjavaorgapacheservicemixkernelgshellcoreSending gshellsrcmainjavaorgapacheservicemixkernelgshellcoreSending gshellsrcmainresourcesspringSending gshellsrcmainresourcesspringTransmitting file dataCommitted revision . +This patch only has the fix for the issue that was originally raised doesnt clean its temp directory and fails when seeing it again +Would you mind checking this in the recent snapshots as well? It seems the fix for schould fix this one as well +Comments are definitely welcome +Will do +What JDBC driver are you using? The trigger should select out of the original table if clob or blob types are detected +Doesnt require to be run as root but will change users properly if you do If you run it as yourself and also run the datanode and accumulo processes as yourself it wont bother changing users If the user you run the agitator as isnt the same as the user for the process being agitated the script will try to sudo +Screenshot showing the memory patch +Fixed in trunk. +This looks nice and simple to me +old issue +fwiw I dont think moving and to oac is a good ideaMy S shows deps from implutil and from utilimpl so impl should depend on util I guess +The release we internal +Pass mode usergroup to rather than rely on the newly introduced which is then unnecessary +I found a few more places where bindir was being hard coded instead of using a macro +Detailed analysis on +Check for null should be fulfilled before invoking relevance check object but it should be in the array even if it is null +adds storageType to adds storageID to updates storage when processing block reports +Sean I just updated and cant see such section +Its only when i moved to smx that i started to notice it + +Thanks Hari +bug at + +Note conformance test attribset fails for the same reason copy with multipleattribute sets having same name should merge the attributes +Patch committed thank you Kate! +I have made the following changes Merged in Johns changes which escape angle brackets in text +Ok I distinctly remember one of those projects was the reason reviewboard was chosen to be supported at the ASF now lots of projects use itSo just to make it clear we will not support another review tool at the ASF if you choose Phabricator it will always be supported by yourselves externally to the ASF +Which i thought would make the readwrite path to have additional code incase mvcc is not handled inside the codec +You could use something simpler for that Id suggest a stdvector +Cos It was there in patch available state quite a long commit it +A slightly improved version of the last minor tweak to the patch +I load now the coreplugin with the detectors from the local repository +Need more info +Reducing to minor until a test case is provided to validate +Patch submitted by dafu +Could you boil down which piece of ruby code is generating this warning? I suspect in an earlier version of Ruby a particular syntax used to fire this warning and no longer does +Committed r +Please reopen if it still happens for s reproducable if something is printed onto the console like for example the devwatch command was used last and the bundle in question is updated this update is printed to the shellSo every time something is printed to the shell via the the last command does get removed from the history +May have to recreate this one since so much time has passed +This was resolved in m when the last execution id was put into the instance tables +Is a machine you are building from set to be able to reach reddeer bundles +I will close the issue +It is not obvious to me that the third example should succeed +I think I get it +Hi DeepalCould you provide a link to an example? Thanks! +initial update of in kit release notes svn ci Sending Transmitting file dataCommitted revision +working nicely again +Verified on. +Geoserver may do reprojection in this case your data has a projection yet it does not match any of the EPSG ones according to the geotools CRS can check if this is the case if doing a LOOKUP CRS in Geoserver returns UNKNOWN data is in postgis so it has a projection for sure if its not know then a reprojectionwill particular Geoserver will assume your data is in the projection it cannot associated an EPSG code to and it will reproject towards the EPSG projection you declare +Instead of using the patch one can do this in insteadTrue That would also solve the problem +Added fix version affects version. +As cited by Jason return null for queries that filter out all results +To Artifacts in are already shared across jobs no changes needed +Critical on this alpha blocker on alpha +Made first release candidate available. +the leak occurs in RA want to fix this but not right now +Added in Git commit cfcabcdfdbde +Great work Cedric! Thanks! +Could you give some more details about ? What exactly do you mean by component in this context +The fix looks good +No test case provided to help reproduce +The diff cached doesnt include the license so I put the files and the patch +Ill check if I can find where the scheduled task is being created exactly +This appears to already be done +The items above got addressed by patch b underIm not sure if should be wired in to the suite yet although it would be good to run the test on a variety of machines platforms +Does it fail with an assertion or an exception? It runs OK to me with or without the patch +but i dont know to what extend that matches or mismatches with the alfresco api +I now thought a bit longer about the lazyInit stuff +the else return was a bug +We are one step ahead of the spec +It seams I have been a bit to fast this time Will also add the other changes you suggested +It is probably fair to expect that values could be Strings +Note that any script demos other than will most likely not be included in the Pivot distribution due to licensing issues +The issue still exists in the latest snapshot containing Undertow +are always five characters so Im not sure why the method needs to truncate the expected valueFrom the comments in the description it seems as though you are expecting tests should use constants from the enginesThat is not good practice it is not part of the public api and it can lead to hidden bugsIf a test is checking for in a test then if a error in coding changes that say from a SQL Standard errorto E then the test will continue to pass even though a bug was introduced. +Reopening bug +Note that it will conflict with M but I wanted to have the patches stand by themselves. +I am struggling to recreate this issue with sample company has a fairly large product that weve set up like this libraries library library library libraryN webmodules webmodule webmodule webmodule webmoduleN initially adding the root and letting continuum load and build the rest adding a notifications section to the root will place continuum in a state where none of the projects in the tree can be saved to the database +AlanThank you for your review +Now that work is well underway with AS all previous community releases are +This is still an unresolved issue and I think you are right that the filter must also respect mixin types +We wont have the problem of Outlinks on that page getting queued in but the original URL will still be there and will be queued after an interval for repeated crawling +About at line They removed the constant in eclipse I think +Were in the process of getting a better handle on some of the tools again now that we have a few more folks to work on it +Running jenkins with v +LGTM +Thread dump attached +Just a few things contains a number of public inner classes +Just +Some more informationWe had an JDK in place +Im working on this the patch will be available in a couple of days +looks fine for both branch to branch +Even fixing things like our existing duplication of junit version numbers etc +Is anyone there +Note that this is not a cache saving issue but a cache population oneOf course if you dont write enough to cause a flush no sstable positions will be cached +The patch to change the comments format in js file to fix this problem +If in some case this error message is not necessary than set the level of logging for the class to ERROR +had a similar result but the bug fixed here were that the root metas were not mutated by doing a copy where the right fix would have been to override the metattributes completely +JPOX maps to cols +While adding tests REPLSYNC cache mode does work properly with mapreduce however LOCAL cache mode is not currently supported at all for mapreduce so this would need to be added +For the service archive all the files generated by the code generator is required +Docs are in the comments not its up to you to move it to docbook and close this issue +Thanks Toshiya for the tips on testsPlease verify +setup +In the KILLABORT state we dont need to worry about them because the KILLWAIT state should make sure all tasks have completed before going on but there is no corresponding FAILWAIT state +Im moving them to so they can be checked that they no longer cause a problem there and fixed if they do +. +Hey Brianthanks for your feedback +Alexey please do it and close it +This can be turned off though viaColm +I agree they should be separate rows +Bulk close for + +You request is invalidWe cant use standard beans class for holding configuration data because then wed need to know in advance what parameters we need to support +I think I was trying this on an Android emulator +Assert the final tree is an actual tree +In light of this I would the change on the patch branches as people in production shouldnt be affected +The result was still a failing test due to the change from returning null to an empty list causing a different policy alternative to be selectedIn anycase things are updated now +Did you look at the JPG screen looks mangled up +merged +Would have been done sooner but we were having some difficulty updating the site which now seem to be cleared +I have started the tests and will commit when they have finished +no objections I am going to commit it in a few daysDIGY +Tested the attached patch +The failure of Phabricator attaching patches might have something to do with the fact that Apache Jenkins has been down for roughly a whole day +Will fix this once Im able toAnd fixed. +bq +In terms of then the idea would be that when you use handleFaulttrue then Camel turns SOAP Faults into exceptions which try + +Issue to watch +Excellent +Updated patch to usebq +As a half measure we can stream from the right node very easily if we continue to make the simplifying assumption that no other node movement happens in overlapping ranges during the operation. +What are the issues? +how about we create a tutorial dedicated to demoing namespace usage? In the past Ive made the mistake of trying to illustrate too many things in a single sample resulting in confusion for the trouble shooting guide sounds like a great idea +If that is a better solution then I am all for it +This has been addressed as part of patch for part of +this is a patch for +Perhaps this calls for a config switch +Thanks for clarifying Ravi +It throws the error properly +I havent changed them in this patch because changing groupid might have impact on any toolstests that are creating consumers +Stein am I right that the only difference between your final patch and the patch is with the use of vs +Adam please assign as appropriateThanks you maybe try to create an Envers test case for this? As the whole testsuite passes +Were waiting on Javassist to fix an issue they caused in their latest release before we do +Hi SiI guess its because when Andy designed it there were no needs to ship +Simple patch that implements that I have been using since April +reopen since it is proven that this plugin hardcoded the main class +aside from proliferation concerns +Theres still some minor cleanup documentation and testing to be done but overall the implementation should now be ready for release +I think this isbetter than dealing with the problem at the test or distribution level as Isuggeseted +Can somebody please mark it as Resolved? Or is there some step that needs to be done because of which the JIRA is kept open? +Dominik could you add screnshots or video to clarify the steps +is there any way to avoid forking whole definition? You know its not good to do is how extension mechanism work +I just submitted another patch that includes a new tool for hadoop file handler +It expects that all dependencies are already resolved which is an invalid assumption +Issue with referencing service and port specific properties should now be resolved +added the testcase +BTW Ive also added modjk in there so that we are consistent with modjk and modcluster +Okay then +Other than that it is ready to be checked in +Well need to add a check and see if EXTENTOF is available if not well have to use SDOAGGRMBR and lets hope Oracle does something to speed it up. +it is will go out with this weekends promote of Labs. +This is not a Gridgain issue I commented all the gridgain related still getting the same error +Looks good +I havent had a chance to try and reproduce the problem myself but my reading of the javadocs for says that if no classloader is passed then the classloader of the current class will be used +Works on my unit tests +I have a G heap and running through M row cycles I OOME at the th or th run w faked k servers and k regions +Yes you are correct +Now you can manage the policy from JMXtrunk +Thanks Andrea +So the commands would give me a clue of what you are attempting to added the dependency to the initial pom file being generated +With the current behaviour it may not compete for taking any new splilog workI feel we can retry for some times and then we can shutdown regionServeror other option is to retry forever on any ZK exception +needs to cache the results in +Freshers version that also handles the Page itself if it implements the required receiver is a great patchThis would enable a more elegant way to handle addComponentIf possible please make this feature available for the next wicket release +I will hold commit for just a while +MR job with tweakable parameters that bulk uploads rows with and columns and valuesCurrently requires a single line HDFS file as input and an table that exists as output though this uses the API directly and not TOF +Note that the properties ended with +Logj change was not intended +dupe of +Added additional tests to both x and after release. +Note that this issues is closely related to the changes made in conjunction with. +Content was added from wiki docs and the reference as well +Build it and use it in your environment +java repro will make it more likely that someone will look at this one +Stage fix for quad transformation. +However does thatIn this case maven simply failed to update parent project and deleting from local repository resolved the issue +Mainly because I am behind on another project and will not be able to work on a junit test for a while +this patch have race and may lead ts crash +Fixed part in rNow exceptions in constant initialization expressions should be handled correctly +We might have tight gun controls here in the UK but Id still not recommend breaking changes +is the correct port +The patch Ive attached has put them back in and the separatejars distribution should no longer pick up the XSLTC dependencies in the once this patch is applied +Moving out until someone smarter figures whats up here. +Problem occurs only when additionaldata is set otherwise it works fine +Ah there it is just with some delay +Please try it and reopen this bug if you still see the same issue with new build +The documentation update are spun off into following separate issues and for this issue it should be documented that plain web app deployments dont have to do anything +Done in updated patch +If we want to make the producer then the shared work queue approach can violate the ordering +It is a pain +bq +Please verify in EAP documentation on stage +If you dont want to reindex just merge to the default codec then upgrade +I do see one difference I did not catch till nowthe link instructions sayjava Xmxm XXm jar install readersseed delegatordefaultjava Xmxm XXm jar install readersseed delegatordefaultDEMOwhere the instruction in the says java Xmxm XXm jar install delegatordefault are multiple databases now in ofbiz +But this feature may not work in and requires you to have s URI in the write interface +Straightforward patch +OK to close it with Wont Fix +I tried it just now and was able to download ok +We want to update to latest for final but we should also allow this process to work +The existing mechanism for creating a validator is not as user friendly as it is for a dynamic WSDL +Our prod environment is running but Im not getting a repro in my local installation either +I will upload a patch shortly +What is your feeling about stability of the API at this point? How likely is it that you may want to make sweeping changes to the classes and interfaces that currently reside in? If you anticipate major changes Ill apply the patch and start working on the first BETA release +When the form is not versioned the page works correctly. +Maybe in doc reference html Then weve somewhere we can put the site javadoc in there later updates to use the first hh find as the page heading change the plugin phase to site handle duplicate anchorbridgehead ids this was causing the generator to break on and resourcesimages are now handled nicely in html but still needs work when to support pdf output +Ill look into how to do the Maven property thingWing Yew where did you find the +I have also upgraded to Sonar and it still exhibits this behavior +The given patches seem to only fix the examples +Sorry it also removes the PUT method too so you are left with no way to create or update the entity from the controller +Thanks +I am guessing we are still not quite right +The stream should be stored in memory for a small documents see how it is done in +Verified using attached sample bean fixed closing. +Ill hunt done some diffs forthe bug +May be we should try the as well to see if one of our recent fix broke something +probably failed because the first patch is already applied so this patch doesnt exactly apply to your trunk +This is a simple test case that reproduces the issue +We should not use unless we allocated and created the memory ourselves +If there were a patch Id schedule it for instead +If its useless dont feel bad scrapping it +do not repeat the object unnecessarily +It doesnt just break in IE its a violation of the JS standard which doesnt allow for trailing commas +Property message is not included into the list of unrecognized properties instead field editor Message Flow Priority is added to the end of the form + +EdwardGot it thank you so much for your help +Acl and are default acl lists +applied +Close after release +Sorry I see the extract that is required by the JCP licence +minor s a little snipet of Matlab code which evaluates the performance of SSVD with and without on AAt first glance it seems that Q is relatively insensitive to the mean of A so that reasonable performance can be achieved even if A is not normalizedIm not sure if there are corner cases where this may not hold +I updated description of this issue in order to make it easy to understand what is the problem +Removed dependency on junit for branch uploaded +bq + +Stacktrace is leading me to think the LOG variable used by jetty is the one which is null +I can confirm that this is a sorter issue sorting to be more specific +I would prefer to have a test case that reproduces the error and becomes green by a fix +Any ETA for a fix? This is a big blocker for us at Travis CI atmFYI This issue surfaces when using +I am facing exceptionerror on my client application which is java based using CXF for webservice soap call and deployed on Jboss javajEE based server application is deployed on Websphere thought I moved from to Websphere for my CXF based client application exception is same +I guess that is a better solution since tranql already knows about the tm +All AS integration code is doing is pass to configuration the names of the class to server is started it will uses its own classloader to load the interceptor classes +Sounds like an issue with the top level chain worker +Please verify. +Decreasing the priority as has been implemented and will go into! I meant that makes perfect sense +Attached patch for this issue +Heres a patch that should fix this including a unit testCan you please tell me if this fixes your example tooThanks +Resolve by the changes to issue has been resolved for over a year with no further movement +Can you remove the included Dojo artifact and use the version shipped with Geronimo +There is a section with one paragraph already but its insufficient +Whoa youre right! I missed this dupThough shouldnt we add the doc method +RI also behaves the same way +Sample code that helps reproduce the errorThe Source File contains a servlet that runs in a application server and is therfore able to create a user transactionWithin that user transaction we use two XA Resources The Jackrabbit JCA Adapter and a Database for which we emulate commitThe overall behaviour is somewhat strangeFirst deploy the Servlet with the update statement active +all the have been resolved just close it +No IT +Pls help me to unsubscribe from this list +If someone has a template that calls methods of a object it will no longer work unless a utility class is created to call methods of +This works for me +Verified can still test and build after this +Ashutosh Do you have time to take a look at this? Thanks +Nope its still an Scala issue +Lars You want to review and it +Does Axis not automatically support WSDL with multiple portsThat seems pretty important +committed to trunk and +It turns a MAPREDUCE MAPREDUCE pair of jobs into a single MAPREDUCE jobPlease consider prioritizing fixing this idiom to work above some other issuesThanks +It is just an additional database option for the wizard +heres a patch +The Control Bus will also enable monitoring of Message Channels +Will post another comment after the suite finishes +new bundle jar created with artifact +Sorry for the delay +By default its set correctly +Its intended for inclusion +now stores the references under wikireferences in the repository. +I dont think that this is +can you provide info on source and line numbers for those code bits? If you can reproduce in a test case its also helpful +The one youve been using has been deprecated in +Should be fixed in the nightly build starting +Committed in revision . +Though its related its not really a DroolsGuvnor issue +Preliminary patch +Hi PhilIm sorry for late replyingI ran the latest test cases and passed them with my patchI added some comments on javadocMy patch is bit difficult because programming is difficult to comprehendBut I think its not so long then you can understand my patch easily with additional comment +Patch is missing test +Unit test which reproduces bad behavior and small fix which seems to work fixes test +The archetype problem is really something similar to the staging plugin problem where the wagons being setup arent getting populated with repository proxy or authentication information from the +This is a patch against the current trunk to bring the generated Thrift classes up to version of Thrift +I got it working finally through the proxyThanks +This also happens in standard derby install +Attached patch corrects mailing lists +here we go againHi Jens File does not exist or the prohibited access to ithave you checked whether this file exists and is readable to the userwho runs TomcatThe file is accessed via the file system not the web! So unless youhave a directory called in the root directory of yourfile system this is a misconfigurationRegards no its the other way round +Your comment is actually a nice piece of documentation. +Other tests that may be failing because of the same bug arenumberingnumberingnumbering +Closing again. +Eli Do you have any more concerns? I intend to commit this patch by tomorrow +Currently if I make a request against the pages api without logging in I get a redirect to the login page +Canceling patch as package cleanup will invalidate all current patches and will require patch regeneration +This overlaps with the of this issue might even be closed as duplicate of that one +Correct +This is why you couldnt use an anyway even if it wrote its javascript event handler into the header it wouldnt be there any more when the request comes in +I committed code just a little while ago that opened up some of the visibilities +Yep sorry about that I unintentionally closed it. +correction it is k memory per GB MB for TB +Glad it worked thanks Mark +for TRUNK AND branch +Am having the same problem with maven see an offline issue in as with plugins but with regular to work in mentioned above I cannot reproduce this error +I think you might encounter problems with the cumulative count and percentages if there is no sorting +just committed this. +This issue will be fixed there +Also when you check in code please make sure that the issue numbers are in the checkin comment. +which means we wont acked an add before the metadata persisted inDo you encounter such issue? Is there any log about it +The newest snapshot resolves this problem. +The values that gointo javascript variables are being quoted so this problem is avoided +Second fails and shows callbacks dont work correctlyYou can paste these in +Please try again I spotted a flaw in buildbot and so mails were not getting sent unless they were manually triggered as you mentionTested with another project and all is working fine for them now +Thanks for all of your help. +Hi Ondra +Hey Ashutosh this can be marked as fixed +I fixed that now in a different way than the patch suggested but it should be fixed now +I think it might be good to explicit mention what attributes are needed when encrypting an unencrypted database and what is needed for encrypting an encrypted database +We cant assume that the user code uses logging at all +Simple patch skipping rb +If youd prefer to just prevent the JT from running if theres no config file we can throw a +QE doc reviews are completed +Problem was in cache +I dont remember what my use case was back then but I think it could now be replaced by the assembly plugin +I looked at the patch for a while the zipped results look ok except the is blankBut the patch does not apply cleanly for me and looking at the diff it is a little odd with the and in the new fileWhy the renaming? srcadminguide +bq +Allocation of a new block for a file +Assigning this to you Li not because Im saying you should fix this but just so you know this thing exists you are messing in this area and you might trip over it or know whats up here +See. +ignore my last statement there in that case +I might adjust the doc some after playing with it +I think the attached patch should fix the issue +This is a more complicated issue and will require some time to fix test properly +The list of svn cp s to run before applying the patch is getting very messy! If there are a few places where an entire package can move save for one or two classes it might be worthwhile to move them +It appears that the two Item instances that are defined in the test data are not persisted to the database +The mistake in the model is thinking that you want version range or pinned version +Hi Carl I forgot to say thanks in previous comment +Cleaning up old if necessary +the replication step are above +this patch will validate the user given against the schema which is also included in the patch +Just didnt realize it at the timeAnyway I just want to put this issue to bed once and for all by encapsolating the file openreadingand etc into a class and refactor +The patch missed an import statementPlease fix it with +If you want to integrate Solr into a Katta type world you are going to be writing some serious code so dont just budget a week to hook Katta into SolrWe ended up going down the Solr Cloud route on a recent project and hooking in some distributed support for durable indexes +The patch changes the behaviour to suppress out optional fields that are notpresent +The remaining comments have open Jira issuesTherefore this issue can be closed. +I rename the title of this issue to be more precise +The cancelshutdownerror idea is goodAlso gives warnings +If there is a performance issue and people have some use of the lists or if it makes the lists more useful then we should work on it +Given that most of the folks run hadoop on unix platforms it would be great to have rpm builds +Not sure what happened to this +We need to narrow the steps to reproduce this +Pull request sent +Obtaining an for a class and +fieldName commentText notNull sizeMax add field reference jpa class +hdfssrcjavaorgapachehadoophdfsservernamenode All specified directories are not accessible or do not exist +Note also which tracks the optimization task + +Does this affect only embedded container or even managedremote ones +I think I found the cause +I believe that is much clearer than otherwise we might want to have bu is it really a problem if we have an exceptionFor the reengineering component I think would be perfect +Ok Ive tested the patch and it looks all good +Committed to trunk +Hi AlexI am able to confirm that this issue is in the Android version of AIR but I havent been able to test iOS or Blackberry yet +m happy to report the fix works +Reporduce this bug in Hibernate +Please reopen if you see it again. +Attached patch that loosens the checking of required primary types against ancestor required primary types for residual child node definitions +I have a proper fix for both now +The previous patch disabled the case when is not set +I will attach a new patch +All the above mentioned testcases corrected in trunk! Closing this issues as Not A Problem! +stored in sessionState variable +Ive added a unit test to for those dates but it works as expected +I think what might have been going on is that while the importer was being run the decorator thread started running +If there arent enough nodes the command would fail +Pages and Numbers format support will follow shortly +Hi Mark! I saw those options however IMO the clone concept is completely and utterly broken +And for form name what if my login page doesnt have a named form? In the case of the site Im trying to spider when your session expires you manually go back to an https page and supply your username and password as CGI parameters +Attached configuration I dont really know Timothy +I get the feeling that Benard wanted to have maven not try to write to the plugins directory +for the patch +Nice catch Todd was over zealous in my uniformity +ondateselected event is fired after the data selected but before the input updated +Read the comment by Steve on the linked issue +Moving out of +You get the correct line number info but the is not reported +Committed to trunk and. +This might be a point of confusion for usersI think thats fine as long as it is documented in the to branches +Thats actually also true for the +Patch for fixing compile error with Java in applied +Can you attach some of your test scripts I will try to debug and see if there is some thing I can quickly fix +bulk close of wont fix unreproduced and duplicate issues +Hi Tom Thanks for the feedback I will try to scrounge up some time to do the tests and report back the results + for the patch +on the patchIve run and and ant and have gotten working Eclipse project +Fix for +Looking forward to more Netty optimizations +This issue is not specific to Id +Title changed to reflect broadening of issue +Ill vote for it. +I tested this patch with a forced failure +I suspect I will find the issue is still present as nothing has changedI am not a JAVA guy however I wil try and execute the test under JAVAI suggest that you try with the C CPP code as you will see the issue I am talking about +or something like that and every file hat matches the pattern will be copied +I need to make sure that an artifact with overrides is loaded first +Ah those warn level logs were there for debug purpose only +If theres still an issue here please file a new bug. +upgraded trunk to +implementation would allow to make use of as well +Closing issue Oozie is released +This seem to only show up with snapshots of plugin artifacts that use a date suffix +Without further input and help from the community Im now dropping the anticipated fix version +Actually we are not invoking the jobs we only register the jobs within the QuartzSpring I dont see where we can catch something +The last patch trigger a for meIve tried to correct it but I really dont know if what I do is correctAnyway I attach my patch please remove it if it is a misunderstanding +Reverted which caused this +notNull install finder class +The findbugs warning is unrelated and Im confident that the test failures are unrelated as wellIm going to commit this patch momentarily +Optimizing the case when it fills just as a task is completed but before it logs that its completed doesnt seem worth the effort to me +Hopefully didnt do anything bad +Verified in JBDS Beta +I dont have any problem with lsgreps color control but I would tend to agree that its probably not worth breaking any backward compatibility and a would be good enough +Would you be willing to attach this WSDL to the JIRA as a file with the license to ASF box checked +Fixes issue within my environment +seems the best approach to mecheers agree with Roland about the Singleton issue +Laura let me know if you want me to rewrite it again or if you got all the comments +for +I just committed this to trunka nd please set Fix Version and also commit to the appropriate release section in in all branches at the end +We will need them as soon as multiple Web Services are used in one process definition +Marking as a blocker for this issue as we are optimizing the calculation of number of waiting job count in and then fixing running job count in this JIRA +So this fixes my xml element issue and my custom serializer issue related to this. + Username dsundstrom +Whats not clear to me at all is whether the object is shareable but Im looking at thatYonikAll things being equal all cores sharing the config set would get the updates to both schema and core but probably not until they were reloaded if they were currently loaded when the change was madeShawnReally good question +The probem was in logic for our source menu items visibility items were shown in WTPs source menu while focus was in Package Explorer and thus another one source menu appeared +yeah thats true i checked with the patch and it doesnt cause any memory leak and solve the problemAnd i checked the performance as wellRegs chasing another guththila parser issue I discovered that my original source fix was not quite ideal in handling the case of a memory allocation failure for the namespace block +If you want feature parity on Windows that is a different discussion altogether +Resolved +For +Raaah was simply missing the flagFixed in trunk +Agreed +Issues found by are fixed +There are many way to have something sortedAlphanumeric String length date entered Date removedso possibly have a parm passed that determines this in the widget +Thanks Ben. +Unfortunately I will not have full network access until this SaturdayThanksIanSent from my Pearl sorry about the briefness and spelling +I recently did some performance tests with Rampart trunk and noticed the same kind of issue +Konstantin I think the difference between manual or automatic is not only the operating procedure but also the information consistency +problem is in the monitor marking as services component +Is there a way for us to work around this issue till Camel is realeased? +I dont think there is problem for concurrent creating the same file +Im also running some cleanup tasks to get back more disk space +From there build a reference implementation in the execref module +Sorry for late reply +mrepositoryjunitjunit deewu +This issue is cloned from it was considered for but since we were not going to ship an update to the JON plugin pack for we will not be fixing it there and will fix it in the plugin pack update for +We only have the options CASCADE NO ACTION SET NULL and SET DEFAULT +We should move the master version of to src +Fine for now but for future referenceThanks +This is rather an issue that rcomponentControl doesnt work with pickList +I think this feature is worthwhile I dont see a reason to limit map keys to just constants +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Fixed +You are the release manager do what you prefer +Thanks Vishal +HiWhen i created the setup with Patch A I have created with newly installed host and then upgraded my setup to +Slowly getting there +this improvement would solve this issue as generators are an extension point + +This is done +And I will check if it has same issue as your expection I will raise another JIRA +Can this issue be closedThanks for the reminder MarcoI have committed a such thing recently +It shouldnt really +Vikas +Close after resolve. +Should probably also add KI comments on DB +good point +Wont fix this right now +Why not call it toAfter all that is what it does. +Thanks DanWell youre right +Closing off issues prior to release going to see if the sub tasks stay open. +In my experimental branch is used quite a bit +On it works partially +The default is clean but you can change this to the goals you want +Using the same entities and call patterns might make sense though +also exported +Get all the snapshottable directories where the current user has permission to take snapshtos +But even if you select it by hand it simply appends the extensions a second timeWhen uninstalling the installer does not change any variables because I deemed it to dangerous +Getting a parallel version in place for comparison is very valuableI wont be able to look at your code due to my current constraints but I think your general direction is very good +Verified thanks +MichaelIts a great point about a contrast between for query time and index time joins +Checked into SVN as rev . +Thanks a lot Andy for your very thorough reviewOne question before I commit this patch though can you please describe what sort of testing you did to verify this change? Are the warnings about pending exceptions now gone from the logs? Were you able to ensure that memory is no longer leaked when exceptions are thrown +I have made some really minorchanges extracting constant definitions from all the RDBMS names and movingthem to etc +Ross you totally ROCK! This documentation is very well organized and extremely helpful +Hey Jimbotron is there anything else we need to do here +Just committed the fix to the problem and it works fine +Since it sounds like were not all going to agree on the best approach I vote for applying Jonathans patch and opening a new issue where we can debate the right way +Setting fix version to ER to on ER. +This issue can be closed +Has anything moved with this? I ask because I know some recent patches for xlink encoding recently went in +The example needs to be fixed not the functional test +Derby sends lengthless LOB information in Layer B streaming +Please verify by marking CLOSED +Thanks Arpit. +Hopefully this test should pass after fixing this issue +NilsYou can specify a value for the for attribute statically rather than relying on the path binding +Ok as it seems the current alpha stage javascript maven plugin provides what we need regarding concatenationand compresion +This patch changes the test to keep track of the datanodes that have been decommissioned and keeps them in the exclude file +Setting this to fixed for now +Please assign to me and i will take a look +do you have seen a difference? +Are your using? we having threads running at What do you mean by the above? Are you setting to true in which case I can see the problem happening +This patch adds GET and POST support for against a WFS data source +This may only ever be an issue with FTP but we need to harden this code if it is a bug or make it explicit about configuration requirements if not +Not reproducible now so lets close it as. +DVR is name of the workspace VERSION is schema im using behaviour without transactions is as expected even though it results in more store operations than would be optimal +Ive edited the JIRA and shortened the description +Agreed +Attached is the Patch for the fix and TestcaseAlso attached is the Test zip file +Issue resolved by applying the patch +But only with path +My prototype bundle with a new pattern resolver copied classes to fix publicvisibility issues and a new application context and context creator does not really work because of the access issue to the Activator I can get the build problem solved then I can also run the profiler to better provide visibility into the issue to you and confirm it finally for myself. +As soon as your work will be completed please create a new issue and attach a patch to it +I didnt realise there was a JIRA issue for it +Sounds like a doesnt Checkpoint Signature prevent this problem from happening +Fixed in r. +That still wasnt quite right +It would be nice if the logic for finalizing a block is at finalizeBlock +All test cases pass. +Karl asks what do you plan to do for the list and execute verbsList would be a GET and execute would be PUT +As far as I know the latest revision resolve the problem for stderr output thread but not for stdout output threadThanks +I suppose the writable could derive from Text +Committed with small changesr +Should be ready to go +Added a testcase +Added new document to describe the SVN access. +Stack I dont object to having whatever running wherever +Will commit to trunk and nutchgora on no objection +Fine by me if you have a better way of doing itWho would create the object? The +accumulo minicluster command is not working in master I had forgot about testing the proxy +Remove create new button from the launch configuration properties standart browse dialog with smart create new existing browse dialogs with new one where it have sence +Actually theres no need to stoprestart the cache just the fact that a Store is configured screws the data live +Isnt it best to keep something like this out of the core +Edmund wants to release it in a separate maven there been any luck getting junrar into Maven Central yet +I think we can just remove that line away +attached is a patch cutting over all uses of whitespacemocktokenizer in solr test configs and fixing the errors this foundid like to commit shortly we can improve the coverage more by looking at other configs and cutting over things like have a new patch i made on the airplane but it found more bugs +Implement the Expression Rules defined in spec additional logic and unit cases. +Seems to only happen when security is involved +Part of a bulk update all resolutions changed to done please review history to original resolution type + +Thanks Jacques hava a good Weekend +If we intend to support the jobhistory format as a public format then I think whats more critical than making this constant public is to fully document the format and in that document declare the stability of this format +Also submit can be renamed to the patch changing the setup order and removing asserting lines from because wicket tester do not maintain request parameter between requests even if they would be originated by browser to form components on page +So I believe Google also fully compliant on iCalendar Object Spec +Support for registering and resolving nested services ie +Maybe we should close the other issues and leave one issue open to this issue +That is mostly not correct +Im a jar shortMissingjar Try downloading the file manually from the project website +Ill wait for the release and check that the issue is resolved +The problem is what prefix to use for the new attribute +It will still be done by tomorrow but API changes in the fork will be necessary. +Attached the patch +Thus it seems to me that the responsibility of specifying what the requirements are of the UDF should be the onus of the UDF developer not the cluster administrator or the end userquery writer +Thanks for catching that Todd +For local transactions we can rollback but we need to be more careful to do the right thing if a global transaction is in progress +Bug has been marked as a duplicate of this bug +Moving to there was no release +Inside the get method a query is executed that returns a group count +in the README should be +Hi allI confirmed this issue with the latest code in CVSAs the result the latest version still include this issueI confirmed these three points +Bulk close resolved issues +If this does not give us an acceptable level of reliabilitywe might have to design something better +bulk close of resolved issues in. +If there is anything else that should be done for this issue please open a new ticket. +This is not invalid +as it seems to have gotten fixed over time +Attached a patch to remove section from main site docs +We could even randomly assign a different delay value per flush +Looks like this test conversion was committed. +no I think that wont work if I think about subclasses +I committed the patch +moved this to to keep this ticket open for full fixing of RR on strongRead path +Dear Henk PenningHmmm thats a bummerWell pick another project thenThank you for your timeJeroen Wouda +Reviewed with minor tweaks checked in. +Woonsan could you see if this problem was addressed in your new work on the Wicket portlets +Unless someone objects if the tests succeed I intend to commit this patch tomorrow +Reassigning all XJava debugger issues +author Hugues m closing this one off for now as done as much as I can for +bin system did not rebalance +yeah this is a bug +Struggling with CSS +Thanks for the help. +Attaching the correct patch this time +The tests can be when the driver gets some work. +overwrite the Cordova created image with the metadata of the originalI believe Android saves the metadata and restores it after it has been strippedThe simplest way is but this changes the expected return value of the data field in the success callback which is the images pathBefore I implement and test anyone have any suggestions on which method would be best +to trunk +I tried the same scenario and the host was marked as down after thre ping timeouts +Is there a reason for making them separate? Could we do inline editing of profile information? +Patch applied in trunk and fixes +Is this capability actually implemented? In which release? How does it work? Through the Admin Import Export +Ousp +Please any one respond for my is impossiblePlease use mailing lists for queries. +we need to fix the path for downloading a release +I mean that we can hold our filter execution performance hostage over this internal API since you have clearly voiced you will revert my commit if i commit the patchAre you saying you wouldnt support an updated version of your patch that went back to builder style +Well it seemed safer to verify that Sections arent leaked even when relying on implicit close +Thanks AshishYour patch is commited at revision +Having the request object in validatedocupdate would solve my immediate issueI thought the context solution would be more genericIll leave the decision to the experts +Now Interfaces and Classes will be proposed +Fixed in. +Nevertheless i just created +coding changes look good at first glance +How would failures be related to this patch? I dont see it +If this aggregate is nested in another the same conditions hold at the final step of the outer aggregateBut with this new processtrue option the parent must be held after the CM until all of its children have completed processing in all aggregates +Not so sure about as the build is brkd +In my case it is not that easy but it could work +so not sure +Hope you guys have cont +looks good will commit if the tests +While working on I have a need to know if JDBC Resultset is closed but not just by checking the but also the underlying language resultset +When many threads we get a lot of messages in the log about the failure to connect +I did not change to limit to one API yet since I dont know if thats the decisionThe tests require ant clean or youll get strange errorsThis is my only patch that I have yet to import into +Right actually doesnt make sense there so attached a patch fixing that +SVN Commit +Users can define one or more security configurations and reference them on the Security Policy page for a component service or referenceAny feedback +The use case is a Facebook app where any image needs to have an absolute URL +Thanks David sorry for the delay +Only the AM does +Logged a separate issue for the temporary fix and tracking progress on the same +Add tests +When the release is rolled out resolved issues should be marked as closed since there is no point in reopening an issue which has already been released +Please check that it has been applied as expected +We are using xlC Vitaly +How is this expected to work when having a relation? If you only delete the orphan you will trigger the foreign key constraint. +Yeah it has in our caseIf the keymap is not found then DEFAULTKEYMAP from is used +Could you attach to the drlvm process with GDB and send me all stack tracesI will try to find Ubuntu linux from my side +This will enable to enlist multiple WASMQ resources in the same transaction +new version of patch adds a second check for hints during hint delivery +Ashutosh I agree with this is redundantDo you want me to remove that codeconfig or leave it around for now I cant think of a good usecase why would be useful other than lockingbut intuitively it seems to be good to allow insert to finish atomically +Jimsola I really prefer to track these two issues separately +You wrote you can not reproduce it using XSLTC while this bug is about Xalan +You can close the issue +So changing this just introduces test failures and additional work +I managed to have my PNG chunk filtering working on a jar project I just need to have it working for a webapps files +So we are far away from a from protected to is a breaking change in Java +This is occurring in version +We can wait +Jonathan push up a new snapshot to the repository yesterday you should now have access to it. +Fixed in revision +I believe this would be functionally equivalent and still be cleaner than the old do you think Andrea +The image is added but is still present in the upload list +I describe above how to fix the synchronization +This is pretty wacky and I have no clue either +But Seems jenkins build is hangedwaiting for somethingYeah +Thanks Henry +Attachment has been added with description Patch applied to revision . +Hi Prasenjit your problem is not nd level cache related if you have not configured a nd level cache since Hibernate uses no nd level cache unless provided with one +Hi Keith could you review this change please + +I also change and add some documentation +And what is a main difference between soTimeout and +Duplicated with the Please close this Thanks a lot +Yes was an odd time to change this but true! +Thanks Dag +Closing as WONTFIX +We can talk about relaxing checks on the dev list if we think they are too onerous or picky +Thanks Suresh for the commit and Arpit for the review +Distributed tasks were allowed to execute on a joining node before it was finished starting +on Majas fix +Please correct Fix Version to one without x like depending on where it was version was updated +Committed revision +Felix why did you feel the need to create these classesregards there a plan of adding OSGI bundle information for +batch transition to closed remaining resolved bugs +Thanks Rob +Im going to look into getting WSDLJ fixed or possibly writing our own extensibility elements. +This is the nature of swinglabs projects prepare things to be put on jdk in future versions +Thank you Hiro! I apologize that I am coming at this as a fairly new participant to and Ruby in generalGiven that a number of changes were incorporated to support perhaps a special check that generates a warning might be helpful to catch this situation +Both jclouds and are now available from maven central +Were currently discovering Oltu but as soon as well be more comfortable with and knowledgeable about the codebase well start submitting patches for sure +Ok I see +We also should know start location for each vector +The job stuff has needed a little bit of love since Michal stopped working on Grails +Starting with a configuration will starts all as a single operation +It uses to load +You can build your own strongly typed object from the result if you wantCheersPaul Foxworthy +i dont think we want to support this +Michael could you try testing against? I can reproduce the issue against Spring with but upgrading to solves the issue +attached patch pushed with commit hash defceebaceaa +Plus looking at how long this jira has been open it is no blocker +Ill get back on this issue when Ive spoken with Sebastiano thank you for providing the details Paolo +Thanks Mike! This is a great improvement Ive already made use of it. +Assigning all documentation related issues to final since they dont represent a change to the t really apply given the new DSL and new docs. +So however it works in it needs to keep working in the same with with +Fixed +If you provide a patch Ill apply characters fixed +The patch to fix the issue +So all issues not directly assigned to an AS release are being closed +Seems you are back from vacationI solved it the other way round Adapting the imported Ivy version from to you are correct Removing the Groovy container is more generic in the way thatlater updates to either the Ivy container or the GRECLIPSE plugin would break it asking This mismatch in the rd position of the version number is really an issue to be checked for +yes it runs under Java +We could perhaps pass all the configuration values as command line parameters to the script which would then look for one which sets cassandras replication level +I am not yet sure what causes that need to dig deeper +The exclusion of files based on absolut paths works well with exclusion patterns but not with inclusion patternsMoreover the following items must be added to the Wilcards documentation In Following rules are applied file to define a pattern based on the absolut path In Examples orgmycompanymatches all files in the orgmycompany directory +Jimmy this ready for reviewcommit? If so please submit +A bug in I could follow your links and find out how to rework my MVC configuration +Consider moving directly to +Test case demonstrating the issueSuccesses point! resolves and ReaderWriter subclasses as well +Committed this to trunk and +cd AVROlangjava +Unknown LicensesUnapproved licenses srcmainresourcesorgapachestanbolcommonswebviestatic srcmainresourcesorgapachestanbolcommonswebviestatic srcmainresourcesorgapachestanbolcommonswebviestaticlibjquery srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlibSmoothness srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlibvie srcmainresourcesorgapachestanbolcommonswebviestaticlibvie srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlibAristo srcmainresourcesorgapachestanbolcommonswebviestaticlibAristo srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlib srcmainresourcesorgapachestanbolcommonswebviestaticlibhallo srcmainresourcesorgapachestanbolcommonswebviestaticlibhalloFor each file we have to document the license +I have removed specific address parsing code from the Java BrokerPreviously if you didnt want to bind to the wildcard address you had to provide an literal address now the code allows you to use an or literal or a dns name +an solution to fix this like the circular symlinks are caused by and which are both ignorable tests for now +features this version is newerrm ln s +I have just encountered this issue in IE FF Chrome on windows XP +Well it stopped receiving updates cause the SVN was reorganized and we dont actually need this mirror anymore it was replaced by the mirror +A was missing on the +Update the fix version to include and +I committed common and mapreduce jars to HDFS thus completing the fix +Im on providing the new shuffle as an option in if it provides better performance +The main reason for doing this is to avoid creating a new message every time we have to send a new piece of information +but i leave that to you +I plan to commit these patches Friday if I dont hear of any problems +Can anybody provide some references for the to say now +If the user enables parallel commit using this property then it implicitly implies parallel rollback +We will take care of this in since it is simple and is a great test for the subentry mechanism +pushed a change if you can test it please. +Thanks Suresh +Patch looks good ben +the problem was that it was trying to use the string null as I see the database received a connection request from null when it shouldve been +When sort used queues sorted queues are not sent to think assigned queues couldt be empty +Youd want to log an enhancement to have the max value use a limit on the source query if possible +Can you please provide your Eclipse version and can you add if your quartz library is configured using the Add Jar or the Add External Jar button in the Java Build Path properties page +Given we have a JBDS coming in June would it make sense to switch the tests to use this ER CP of NOW rather than waiting until ER CR CR etc +patch for trunk is also applicable to branch +The part of locating the resources is still in work +Its what I did in startup scriptEasy to do +Its the that is building up the root map around the components and using the name of the component as the key +fixes the build process in jdk and resolves the issue of the build not failing when a bug is introduced in specific classes and proceed is set to +Lets check the patch +out it was probably fingers +Upgraded to for fix. +I like being a crash test dummy +While fixing findbug I overlooked Base behavior +Can this issue be closed +This is an interesting one +The producer window size stuff isnt completely implemented in NMS yet as far as I know its a bit tricky because of the issue with matching the reported message size on the client with that of the broker theres usually a small difference if I remember correctly so it needs some more work +Fixed in SVN . +Closing this ticket because Closing this ticket because bug is a duplicate and has patches available +Patches welcomeIt looks like one could add PREFIX in the search path for curl inside +Doesnt do the right thing? If not why cant it +I found a workaround for this issue by specifying the port type definition completely into the main WSDLHowever I think it might be useful to have the error checks in the source code as proposed +Thanks on r thanks for the feedbacks +closing issues for released version +Thanks issue is closed now +Committed +No I dont think you have to rename the classes +Could you send us all the changes that are applicable to this fix so that we dont have to wait for the code drop +attach my patch +Thanks Gunther +Sorry for the confusion. +This was applied +Id be fine closing this as Wont fix if you dont think itll help +You can find that dicussion somewhere in the mail archives +Is this fixed +Then all properties prefixed by myproperties +The attachment is a complete eclipse project without the required jars in the lib directoryThe lib directory should contain the jars from the Axis june th build and for JibxCould a fix for this bug make it to RC? Finalthanksm looking into it and hope to get a fix within the next few days +Jacques thats really strange i tested with different Browser but ahd no problemes using the button +Is there still some work to be done here or can this issue be closed +Markers are created in asyncExec after validation is finished and that I suspect is reason why these tests fail sometimes. +There are more broken tests in too +Changed the logging to trace level and committed to. +good points Im starting to get the fact the s main point is mastermaster replication which we are also using it for on the more moderatly sized dbs +Attaching an implementation that is described above +Closing tickets + +bugs in JDK nothing to hold up a release on our side +Net servers +But if we do not want to do it and want to fix this bug we must syncronized all accsess including READ accsses +Havent heard anything from Michael in a while +Missed the MiniCluster committed this +For example are you assuming the presence of the selector header lines or will the user need to specify the relevant schema for the parent columns? Is only level of relationship possible or can these files describe hierarchiesAlternatively is there a standard or RFC that would be a good guide +Anybody have any objection when we create a branch for this +Updated to ER in revision +Maybe not urgent +Do you want to use to encode filename so that binhadoop fs put foobar baz will work +fixed rev +I dont see us ever using a Hadoop Configuration in the client except the default one to be able to access HDFS +There may be others I havent checkedSo Ill leave it to you guys to sort out which one you want to keep then bin the other let me know when this is done so this issue can be closed +Yes the new patch will work well +Applications can avoid calling for getting the skip behaviorAlso the mechanism of notifying the framework by throwing exception is not ideal as it doesnt work for streamingpipes. +trying to trigger the patch process to pick it up as i dont see it in the hudson +Apparently this patch will break solaris and maybe other old +tests included +Once he is in an API package he will not be confronted with internal types againIn other words with the split on the API and impl are IMO basically interweaved while they are cleanly separated when doing the split on the +I dont think theres actually a correctness issue here but does seem that the regions availability is faster if the RS does the edit +We try to include as much as we can in the release but not everything makes it +I havent checked the patch file yet Werner but I think this is a good idea +This patch will only keep when theres slaves +Yes I need virtual dispatch +So i prefer the maven eclipse plugin as it works and i can publish to several servers +RHEL ships a much older version +pls reviewThere is one more place where there is a spelling mistake Node Manger instead of Node ManagerI checked for similar mistakes in the UI +sampleswsdlfirstwsdl it runs fine and propertly generats the toString methods on the beansIt seems to be workng fine with and request for more information has gone unanswered. +Im going to do it tomorrow and keep you informed! +Ill commit this momentarily. +the grand total composed by the sum of abIn the above a is partially done need to enhance the functionalityThe above work is remain patch will coming soon after that this issue can be closed +I have an interest on working on this issue +QE please check if the issue is still like this is not yet implemented unless we call suggestionBox something else in RF +Tim Sell is running a test that avoids TOF +BTW I noticed has the same bug. +Now portal can have a dependency defined on the plugin which I committed as well +was not in +Added workaround +Attachment has been added with description my proposed has been added with description same patch but including another change in has been added with description correction added missing to include the affected in the previous attachmentNo feedback in the meantime so Ive committed the changes see rThanks for the feedback +The new patch replaces the old one in this ticket +Forum this is not a support system +I will commit this +This actually works but the messaging could be a bit clearer +I checked frank already added this artifact with new build +Uwe thanks for your help with verifying thisIm going to close this now. +I agree that the vast majority of repair issues are actually caused by a streaming failures which is why i thought would fix itim thinking that might have created another exception which once again appeared to break repairs + +Patch for upgrade to Karaf the title to reflect real problem +I will take a look today or tomorrow +Thanks for contributing the logj logger implementation +Hence marking this bug as closed +implemented this in my fork of the samples +The state of the connection is not changed at all +That sounds a bit like option +Im not expert of Spring but it seems to me that Spring did cleaningThe generated by PC is XML escape +Both ways have their pros and cons +Disable any other options that might prevent a proper sync to disk when Derby is writing its transaction logs or other dataIn general it would be good in the developers guide to point to the maintaining integrity doc as when Derby is embedded in an application it is really the embedding product and its users that have to take on this role of maintaining database integrity +great! do you have snapshots in maven repo somewhere or do I have to build from trunk +Samuel Yes I agree about the first gen iPad but anyway who would be updating to on it would have done so months ago right +I actually got pretty far with this simply duplicating the logic in the Hbase Storage handler +Frederic Thank you for your work on this +Can this issue be resolved? Was it only committed to trunk or also to +As mentioned before these methods arent used for DRDA commands but for the arguments of the DRDA instructions for this reason they shouldnt be handled by the CCSID managerIve already ran and derbyall on this patch with no failures +Assigned components and version +Bug has been marked as a duplicate of this bug +Just saying your workaround may not work unless this is not the case +Closed upon successful release of. +Go ahead and is long dead and afaik this one was solved by closing all issues that have been in resolved state for more than one month without further comments +Unscheduling due to inactivity +Philip youre a genius! Removed backslashes indented appropriately +now it passes for me too +release bulk close +Please check if the mgmt server is restarted +Closed. +I think it would be a bit harsh to fail the build though because its an error that we can cope with quite easily +After updated and on hdfs some tests are failing with orgapacheavroio +Perhaps the mappers can keep a count of how many times each item has been seen and favor among items that have not been seen +Then I guess my only concern is what is mentioned in my comments on one can probably be resolved but I would prefer not to close any of these until the underlying issue is actually merged into master +to to trunk and +Is s automatic error managment not good enough? If so we should fix that rather than work around it +Strange stuff it appears they have the groupId in this location set to the other path +Can any one help me +looks good to me just a bit confusing calling it a header when it is essentially a handshake packetcould we negotiate SSL the way we do compression? nothing secret is shared during the handshake phase anyway after that we could upgrade +moves sstable level into sstablemetadatamakes files mutable to be able to send files back to L and to be able to migrate information from old json file into the metadata file without a full compactionscrub of the sends back files to L when load is calledalso realized that people might drop in sstables manually and then restart cassandra this could make sstables within levels overlapguess we might need to check for overlapping sstables on checks if the metadatafile actually exists before trying to change fixes overlapping sstables within a level on solutionI think the last thing we need is to force everything to L when moving from STCS to LCS unless I missed where thats happening +All my apps are polish and using woody some persistence +Yes sounds good to me +With development mode enabled the message remains the same. +Closing as discussed. +is only a conceptional one so maybe there is still room for diverse notice that this impl +Sure it lets you quickly swap in a stub class +Please implement Pretty please +I still have failing tests +The dimension tables can be shared across cubesStorage Fact or dimension table can have storages associated with itStorage ModelA physical table will be created in hive metastore for each fact per storage per rollup +Patch is good +I couldnt reproduce the error but added some defensive code that notices this scenario and cancels the job +Fixed in r by fixing inspection in to reintegrate results into the main inspection +Otherwise their slice of the global repository will not be authoritative. +I migrated the the to the local test suite and added test methods for marshallingunmarshalling simple xsdbaseBinary declarations +Hi Saeid my experience is that the labels of gateways can help to increase readability of BPMN diagrams +Yes cookies are enabled allowed until they expire +the ticket is valid +There is no semantic change to but the patch is big just to move to the new API adds which contains several new stress tests for failover session expiration etc adds a which keeps track of its owner +The implementation feels a little hacky this is arguably better handled in itself +Brian negative +Heres a simple patchI ran with it a few timesPlease have a look +Tested it with and it works fine +It has been fixed as has been can confirm that the error is no longer present +The regressions ran fine so Im closing the issue. +im therefore tempted to resolve this issue as Invalid +This issue had already been fixed in previous versions of Sonar when cleaning and refactoring code of the WS API. +Can you please attach a new patchwith just the hashcode fixes +Why not download it from SVN and give it a try +There isnt currently a good relationship between the properties table and all of the object type tables +The unit test file with copyright +This will be resolved as the individual issues that its spawned. +DNS configured +I experimented with M blocks +In general it just seems like a bad idea to me +Done +Doug sorry I had not attached the latest patch +Committing +I committed this +yup was fine +Patch attached to Jira +Karma integration was deprecated. +Tried again +I thought a would be stable but it apparently isnt +I also started using it for +Fix for this issue +I guess I would need to include doc verification from CSP as part of sanity check +Ill make sure these get fixed via +No I didnt +Resubmitting patch with a minor delta change in testorgapachepigtestutil to make the random null data generation more correct +in +Pinging +This is working fine for us too +windows is a lot fussier about open files than unix which can unlink a file from the path while still keeping it open +What do you think the behaviour should be? The server is up but the management port cannot be connceted to +Rebased with latest passes for me at local also its cousin passes on this qa run +Ill review it this weekend +Then I fixed the bundle as described and everything resolved and the web app worked +Moving out to we can add these later +applied in revision +However moving a block to be appended to a different volume when the current volume does not have enough space for the estimated new bytes is an optimization which probably we should work on it in a different jira +Fixed in +bulk close of all resolved issues prior to release. +WIPIssues Menu wont render sub commandNavigation items No exampleIf somebody could provide feedback for this it would be greatThe work is based on the package +Committed revision I guess I assumed Solrs default init magic worked for components too but now looking at them more deeply I see most dont have defaults +Applied to trunk in rev +We should also store the Sonar bootstrapper used to launch the analysis Maven Ant or the Sonar Runner. +Attached test project +My question isregarding the patch and reference to preflightsrcmainjavaorgapachepdfboxpreflight +The layer assigning this address id to the hostport could be switchable so the application or user could introduce easily its own routing policy +On top of that check the returned collection when DISCARD is on +I havent implemented http client yet +zip file later if you rather would like this instead of having to checkout the code from svn +Is it to only support multiple file extensions or to have some kind of moduleextension in general? The first surely could be part of the later but the later has much higher requirements than the first +Attached contains consecutive java dumps taken when the tests were hanging in the hope that comparing the thread states in the files would reveal the problem +So indeed we should just update the rule implementation to prevent firing a violation when the last clause doesnt contain the break statement +Verified in Tools +I am using default Aegis bindings in other words I have not defined acustom Aegies bindings file +my is perfectly fine +he means the Jboss server remember this issue but I also remember fixing it +Just tested on and youre right it isnt reproducable +about moving the focal point to follow the logo a little more closely and rotating it degrees so that the axis goes from the top left to bottom right instead? Just toying +Note that previous releases used so to be interoperable with those in we should continue to do the same +I was leaving the top level out of the change for now until theres something to build +If you used the same x in all cases I believe there is a bug in your exact and not exact codes because you getonly decimal digits of precision +It will be visible for every selection +Heres the skip listNeed to Add reusable starts int array Runs tests randomly and perhaps improve them somehow Execute a benchmark vs +You should be able to stash your local changes before switching the should also note that the problem is because in is called and no type for the newly created type is returned +Jens IMO we should remove as much overlap between the specs as only possible Required Mandatory etc +nope not a priority but fairly easy to add so will be adding it +We have a container struct that has a map of structs as well maps of other structs without doubles +Then calling multiple times will not be a problem at all. +lib pattern works for me in the POM file but NOT in the sonar exclusions UI +I just committed this. +from +It was trying to test for the difference between S and struct S when we were thinking that a single X class definition would generate both a struct and a class type +In my understanding that patch does not solve the issue at all +And none should scale as well as. +Well I agree with keeping things simple +for +It would be nice if this work around was easier to find without having to debug into PG source code +If this is not expecteddesired the following stacktrace may help identify why the lock could not be obtained or something like that +And later someone can work on porting the patch to branch +Please if the problem disapearsGerhard +As we have moved filtering jobs to an other component +Reopening related case now that is fixed. +done some time ago +Marko this is all well and good +Id be inclined to call this an improvement and target it for trunk +Please vote for this feature +The adapter workaround was moved to where we already do specify logj as the selected logging framework +Build is ported tests are +I made it so that we call getParent in a secure action +Verified at +Attaching v patch matching the issue description +Fixed on closing all issues in resolved state that have not been reopened nor commented over in the last month +The issue can be closed +its out of date nowIll be submitting a combined patch to that fixes Half issues and compaction speed +Addressed in new patch +cp overwrite files by defaultId rather not make it the default +This however predates jBPM so I assume youre still using Drools Flow? Or did you download jBPM separately? This would mean you also need to use Drools Guvnor +But this smells pretty bad +That is for any client wants to gain a lease of any file it could just rename it and then open it +Issue on which work was done + +Attachment has been added with description Fop font loading flag as importance P open bugs to P pending further a bug use font configuration features to select desired transition resolvedwontfix to closedwontfix +javadoc warnings are due to +Ill close it when I see it doesnt +Tracy please update the wiki documentation +Thanks for the pointer to the TRANSLATE and TRANSLATEREGEX functions Lukas +Interesting I had not seen SORTBEFOREDIFF as well + svn ci Sending apiSending apiSending assemblySending assemblySending assemblySending assemblysrcmainjavaorgapacheopenejbtomcatcatalinaSending assemblySending assemblySending assemblySending assemblySending assemblySending containerSending containersrcmainjavaorgapacheopenejbcliSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigrulesSending containersrcmainjavaorgapacheopenejbconfigrulesSending containersrcmainjavaorgapacheopenejbconfigsysSending containersrcmainjavaorgapacheopenejbcoreSending containersrcmainjavaorgapacheopenejbcorecmpcmpSending containersrcmainjavaorgapacheopenejbcorecmpcmpSending containersrcmainjavaorgapacheopenejbcorecmpcmpSending containersrcmainjavaorgapacheopenejbcorecmpcmpSending containersrcmainjavaorgapacheopenejbcorecmpcmpSending containersrcmainjavaorgapacheopenejbcorecmpcmpSending containersrcmainjavaorgapacheopenejbcoresingletonSending containersrcmainjavaorgapacheopenejbcorestatelessSending containersrcmainjavaorgapacheopenejbresourceSending containersrcmainjavaorgapacheopenejbresourcejdbcSending containersrcmainjavaorgapacheopenejbresourcejdbcSending containersrcmainjavaorgapacheopenejbutilSending containersrcmainjavaorgapacheopenejbutilSending containersrctestjavaorgapacheopenejbSending containersrctestjavaorgapacheopenejbSending containersrctestjavaorgapacheopenejbcorecmpjpaSending containerSending containerSending containerSending containerSending depsSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesinterceptorsSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending examplesSending exampleswebappsSending exampleswebappsSending itestsSending itestsSending itestsSending itestssrcmainjavaorgapacheopenejbtestSending itestsSending itestsSending itestsSending itestsSending Sending serverSending serverSending serverSending serverSending serverSending serverSending serverSending serverSending serverSending serverSending serverSending serverSending serversrcmainjavaorgapacheopenejbserverSending serversrcmainjavaorgapacheopenejbserverSending serverSending serverSending serverSending serverTransmitting file dataCommitted revision +Definitively a bug introduced in one of the commits since thenIm currently reverting the commits one by one until I see the test running to find which one introduced the problem +I view primarily as a CPU optimization and tuning parameter +Fixed in rev . +Rebuilt local +Ok can you merge trunk into that +This is not an issue in +Thanks John! +added code does this work? In short because there is a design error in the Groovy runtime which allows access to private fields in many cases +Individual Ack now supported in svn rev we document somewhere what this feature does? What does it do? +Linking this to and adding a patch that incorporates both +to the general idea of logging the data and reporting from the logIt would be good to expand a bit on your idea so it is clear what you are suggesting +Will have to look into it after +Not able to found with latest runs +But this can work along with splitting the cluster state in +Yeah Id say remove it for WFS to have a no nonsense output for the most used protocol version keep it for WFS and if anybody else complains forward them to OGC. +I have assigned the JIRA to you so you can do with it as you see fit +in +the font loading issue which cause this bug in my case was quickly fixed +As agreed during team meeting all these attributes should be hidden. +Sorry I continue to be such a stranger a new role at my company is taking almost all my available time. +Which directories specifically are being affected? The clean should just wipe out build folders which you wouldnt have checked into CVS anyway +Launching a jar from hadoop has similar constraints so this should not be an issue +maven find +Otherwise this patch looks good +Funnily enough it seems to be our own not adhering to that rule of ignoring annotation types +Peter please reassign this to me when the hudson setup is ready and you have something for me seework on +We need to make sure all the error cases will fail +There should be a standardized method of expressing dates as strings which is consistent across all implementations VS just using whatever the various platforms support via their datetime parser +after release +This makes finding a bug in FOP highly ambiguous since this could be an issue in the XSLT youve used to transform the data XML to FOAlso what version of FOP are you using and could you also post your config file if youre using any custom fonts or settingsThanks has been added with description the generated fo has been added with description the fop has been added with description the pdf ok well this looks like a kerning issue +Ill ping Martin about it +What people think about this kind of solution +We probably still need the configuration since operators often like to control which ports daemons are using +Some javadoc could be added in the class but other than that I think it is ready +Heres the +The attached ears deploy fine with the default configuration in +This also affects the thrift metastore server +Claus I cant find that context entry any longer +Any feedback +All the code to unzip the AIR Integration Kit has been removed so this is now a +Thank you very much for your detailed response! comment is not good +does swallow +Patch committed + +Thanks Harsh +I dont think that this is webconsole issue +Patch applied +This means there might be more than one place to change in the code. +Lets stick to what people know +Just tried this on the lab cluster with nodes in the cluster with the latest srcs from HEAD and didnt see any +Yes thats meant to work today +Thanks Sanjay +You can only inject beans that actually exist not environment entries or beans that dont exist +verified +for this release +Waiting for that Kafka release +thanks added you as a issue as it has been released as part of release +I can understand that there is a desire for persistent job ids but that is a big enough feature with enough downsides that it should be addressed in a separate jira + +How about and for the names +BTWThis can be fixed through local changes to the mapping files created to explicitly call the appropriate type closing this bug as invalid +It is true by default. +Please open a separate issue for the data dir problem if you want to +Thanks Arun! +In revision I added a test case for testing runtime JPA enhancement for applications +There is in DIH which uses but it looks like is loaded at runtime +I ran a MR job went to the successful map attempts page and I still get hits that are not found if searching via the browsers function when searching for strings that are in the title attribute text of table entries +Thanks again Andrea! +But we need an early implementation of it in M covered by JIRA +I have filled issue which covers problem as it is something more general and I think it should be separated +Projects imported in workspace wo any errrors when the right Runtime exists! Cool. +Confirmed +Ive committed my latest work to x Joel can you do a bit more testing with a recent checkout +Have you tried the latest version of moodle pluginHow can I download Moodle Plugin to check +the diagram +The patch also contains the necessary update to the examples and shows how simple it isMay be we should simply apply the CVP patch to as well +QED The spec is the written document not the test + +done +PSWhile having a closer look at the diffs I noticed different line endings in the files +This synchronized some state changes in Componentmanager and uses an for +Indeed this problem is fixed with +There will likely have to be some work after doing so +Committed the latest patch +Plexus updated to allow the operation of the. +The reason is that may not installed for all hive usersdev +Adding Jody to the discussion since hes the module maintainer I just did occasional patches on did not do a review but checked where the client dependency is and +to completely disable the inactivity monitor you can set false on the connection URI +Enabling the windowId handling by default is the option which makes the most sense in the long run +upgraded to we could be smarter and use the jdk classifier only for our build but for simplicity well likely leave as is and move drop the jdk classifier altogether for Groovy +This isnt yet a patch +Nothing specialThere is no RS logs as Im in standalone on this one +Patch for which includes the missed handling of CLEANUPAPP and CLEANUPCONTAINER which was already in trunk +im still in doubt wether i should delay the release for this +There seems to be a configuration error somewhere +So you get your exception during construction +Also we are providing an specifically designed for CDI to deal with that logic so in the long term we will be able to update the integration codeI think it is important to keep the responsibilities in the right place +Ok I got it Im fixing it now should be checked in within the +Thanks Nicholas! +Oracle graph shows that average execution time of new get query do not grows +So why not include in the javaee chunk and GWT in the as or common oneThat would reduce the number of repos from to and greatly simplify the overall view of the components and their interdependencies +Abiword now can read document +Hey Nicholas patch looks pretty good to me +The problem that I see here is that is actually created for a specific and operates on that so we cant really cache it in request scope Im afraid +That also involves looking one byte ahead and rewind if the character that follows CRLF is not a LWSP characterIll explore possible fixes but Id like to work closely with you on this +I will commit this tomorrow unless I hear otherwise +Downgraded to minor +Upstream issue attached to description +HiIll look into thisThanks patch against trunksrcsite +I rechecked it againI found one inconsistency in RI behaviourIf you get fields of calendar before time zone changing then setting new time zone has effect as here describedBUT! If set time fields and change time zone before any getting setting time zone has no effectIve checked that after harmony has the same behaviour as RIIve modified regression test patch +by the value currently +committed revision +Please open another issue as an improvement for the other problem of the not being generated for the qname of the binding in the port. +Thanks for your suggestionsRegardsPranav +attach the test case is much much better than put them in a doc you can see the Creating and Patches link above the highlight REMINDER section above +Will this interfereWould not want test to pick config from ACCUMULOHOMEconf where is defined and use the production accumulo dir +If under CC you want to perform a build and then deploy to weblogic isnt that process all custom stuff you are doing in? Cant you supply your own properties as part of? You dont actually need as the name of the file in the project correct +A patch for has been submitted which addresses this problem by updating the datanucleus plugin as well as dependent libraries for Hive +Thanks for the patch +It was reported against where it works due to +I suggest we verify first whether the total sync time drops when moving from interceptor to loader otherwise theres no point in doing so +plus make the code completely unmaintainable +The reason exceptions are discouraged for normal control flow is that they hide control flow +Well the interface name depends a bit on the strategy we follow with JMX +Again if we decide that the group page shouldnt be accessible to unauthenticated users then we can explicitly specify a URL that points to a minimal group page with just enough metadata to make Facebook for now the group page doesnt contain any functionality or sensitive info that requires a user to be authenticated so theres no reason to lock it down so tight that FB cant see When I first tested this fix I didnt think it was working because the little dialog that pops up under the Like button for adding a comment still had the old message in it +Using the latest version schema the generated classes looks good and without whitespace changes like you said but only the changes with the schema changes +This issue is fixed and released as part of release +The findbugs warnings have to do with naming convention +Patch applied with svn commit r +I didnt see the mark as duplicate option before This bug has been marked as a duplicate of +This may already be fixed need to check +Normally the block is streamed to the datanode as it is written +Please add the fix to those too +JrgenYour suggestion is helpful I added the extra predicatessince I didnt know all code paths through this code the simplification is way better Illmake that change once I have run the regressions again +ChristianYour patch is in revision +Hi Jeff can you look at this +It doesnt address the question of what to do about the external interruption but thats more of a question of ifhow we want to make an API change +What if srcFs is an and fs isif fs is then the does not support v yes +Travis Im not very familiar with so Im trying to understand what the report generated is it the case that required the older so wed bring that down every time to satisfy dependencies? However given that were not writing testcases and thatd never come up in our unit tests and our tests succeed with we ask to exclude junit as part of resolving s dependencies +Closing this issue for now +Im able to redefine distributionManagement in my child module and it overrides correctly the entry declared in the parent POM +Let me know if you need any assistance +The JPA spec also doesnt define any need for using instead of the class annotations itself +Just take the settings that doing a png does and make it so the same happens when formatpng and palette this one is not so easy either +Keith after some investigation weve determined that CLIENTACK is not correct and AUTOACK is the easiest way to handle the general caseFor the specific JCA case we are exploring the possibility of using CLIENTACK by modifying the ack mode runtime + to trunk only + contains another issue +Its possible that purpose of those keys have been misunderstood +I just noticed that the fix works when running command line but not with the ANT task +To be clear scando +By RC the jacorb versions will be the xmlshproperties files all are okay except for the last variasrcresourcesservicesinvokerhttpsrcresourcesservicesinvokerhttpha What differs in theseregarding source zip vs +Ill fix that too +I think we should state this clearly in +One more comment Does is make sense to support multiple listeners? If yes we should implement the observer pattern + +Screenshots of Connections for the and File Owner +Haha I just write almost the exact same implementation for newtrunk then remembered that there might already be an issue open + +Ok that would be awesomeWhat can I do to help you with this? Provide hib mapping XML that simulates the problem +Fixes the intermittent failures in and for your help with all of these Luc +Also seeing the same result with the notifications quickstart +Please confirm +Hey guys this has been fixed in commit defcafdcafcd +But I really dont know why I used a one relationship from Party to its geo location history +Thanks +Id toyed with the idea that its just a simple case that needs its own transaction and therefore a simple REQUIRESNEW instead of default REQUIRED +this patch moves all the explicit bytes to char conversions into protected methods adds a surfaceForm method and opens up for subclassing +Sorry I meant +if you have time to research i would love to know the details +As soon as I do it Ill repost the patch +Thanks KiranYour patch is in trunk at r +Im interested in the exact way you get this NPE apparently I am missing one area where I dont close the lock or something like thatCan you pastle your NPE stacktraceOf course a testcase would be fantastic +in r +Running derbyall now +If we distribute the jar it has to go in lib with a license no Maven shortcuts +You favorites pages renders perfectly in FF or SafariI noticed you are using a different skin on your FF browser +Attaching new patch handling this case +We didnt do it when introducing Maven or protoc +I tried the script it connected to metastore times set check and commitJob +firefox seems to render them correctly in edit mode but it is not valid HTML +So no mvn install hasnt been run before +Should we leave the intents to be lower case for now while an issue is raised with OASIS to make the WS Binding spec and Policy spec consistent +The patch includes patch files for classes +Please note that I changed the hadoop version manually to Hadoop for my own testing purposes and that doesnt related to the problem +If we cant split the property on spaces then itll make the code more complicated +Note that the rewriting needs to focus on I think the test jdbcapi already includes the details of but doesnt run with which is why was kept alive +This has some impact on the amount of memory needed by the NNIdeally if we can compute the UUID in runtime no additional memory is needed +Less hassle heck yes +Youre right I skipped the fact that gcBefore is different from gcgrace +YonikGood catchIm gonna give it a try +Thanks for the patch Mikhail +Closing since this was previously closed in. +closing off issues prior to release +To be confirmed +I am subscribed to the following mailing lists and moderator for all listsuserdevprivateuserdevissuescommits used to be olgan but that address is no longer valid +Pull it back in if you think different +just for provide the data for config in another patchonce this is put in will do a update to current usage. +I havent started on the forward port +I agree with Todd on this +Patch fixes the problem with CC test executionWith this patch BTI prevents the test suite from being executed if its superior suite failed +All necessary patches submitted upstream +It might be worthwhile to add tests using the workaround to ensure that the dialect supports the functionality +Initial patch for review +Verified +Please let me explain why i think this is a necessary for increasing namenode transaction performanceThe namenode has threads serving client requests +Ciao MelodyI am not sure you went through the mailing list with this as that would be the right place for this kind of the messages above it looks like the process cannot access the pyramid directories see for exampleexistsfalseisfalseDid you check if the process is able to access the pyramid directoriesI am going to resolve this for now as it is not really a bug. +Also added a entry +Without disk quota we dont have any way to compute the disk usage +Hi Im interested in the same patch +Will be fixed next time the rsync updates the website +lgtm patch is failing to apply to trunk however +The question in my mind is whether custom attributes represent HTML or not +Id like to avoid that as far as possible though +oh yes +Antonios Please provide the URL of the documentation page that needs the correction you recommend +Thanks David and RobertWhat an incredible step forward now you can easily try out all sorts of scoring models or make your own +I have reviewed Michaels patch and I approve it +I believe all the jars the app server uses are now in. +oops sorry is about exclusive gateway +It seems from the service side we should never let an unchecked exception be thrown back to the client +Dont bother doing a time limit on the status reporting +This is obviously an area for lots of exploration and experimentation +These artifacts are supposedly useful for system intergration tests which are not yet really officially implemented anywhere AFAIKFor the sake of existing CI build systems we can have doing the current thing which can depend on two new targets and +If you extract the contents of the war into a new folder and run command javap s c verbose on the class that is being warned about what is the output +Tests run Failures Errors Skipped Time elapsed resolved issues for already released versions +My apologies for losing track of this patch +bq +Postpone toWe might want to by the way open this one to the public as its not really a security vulnerability as such + Affects Versions this is incorrect affects version Jacek currently have been released so please specify correct affects version +Itd be a milestone +Im not sure this belongs as a JIRA but lets discuss this on the next EAP call +We can always reopen if not enough. +First time digging into the code and missed that +Did some checks with javap to ensure that line number etc +Hmm there are entries with namenone which I didnt consider in my patch +I have added a jira to cover a potential refactoring to split out of few of the responsibilities currently carried out by the class +Anyone disagree +all issues have been resolved! +bulk move from M to are done so issue is resolved +Please ask the question in before you raise a JIRA issue +Here an improvement of the MTQ only patchThe auto MTQ rewrite mode now collects all terms into a until cutoff +Sorry I missed Bills commentbq +patch attached +fixed + +Thanks everyoneRobert can you open followon issues for your other ideas? I think they are important +i dont want to block the cq release if there is no other solution right now +The mapping is quite small so hopefully will be helpful in running down the issue +The patch is just a quick hack from the comments of the jarjar bug that fixes this particular case not a real solution though +Updated patch fixing the findBugs warning including Amareshwaris suggestions and a minor tweak to shorten the +We should probably looking at running TestCLI with both trash on and trash off +Thanks Lin +With respect to option I think that invisible columns are best handled by the language layer in the existingcatalogs which describe columns in a table +Having not purge the messages seems sensible to me +Thanks Si and Leon for improvements! I checked only INVSALESTAX because this was initial purpose +Resolved long time ago closing +The patch adds one condition check +NET matrix build we should enable the Pnet profile to run some additional project importer tests +Thanks Amar +Thanks Amareshwari! +will take a look +Sun has relicensed from SBCL to CDDL so the switch may be avoided +If we add the change logging that will come first in the log and will give sufficient information as to what the problem isI think this will not affect existing read only users because their database will have been set earlier when we were accessing the lock +eventually we also have to handle SSL transport andSAML tokens +It sounds good +what a delightful solution +If it still happens after has been fixed we can reopen. +Any benchmark +Ideally wed like to remove the hashmaps and just use an efficient search of the JTOC +I am sure there are plenty of docs in the incubator that explains how this works. +Sorry just run any ruby subprocess like system jruby +Patch committed. +Nice support guys thanks +I applied your patched and ran a trunk build with and it looks good +For now removing supports from +Are schemacatalog not used in? Asking for specific tables in a specified schema +A version that doesnt delegate to infrastructure can also be done as there is very little dependency on +Custom interfaces arent supported in RC but Im hoping to get that support added in the next couple of days so it will be there before GA +svn ci m archiving legacy artifactsAdding archiveAdding archiveAdding archiveAdding archiveapiAdding archiveapiAdding archiveAdding archivebtreeAdding archiveenhancerAdding archivefostoreAdding archivemodelAdding archivequeryAdding archiveriAdding archiveruntimeAdding archivetckAdding archivetckAdding archiveAdding archiveutilDeleting trunkDeleting trunkDeleting trunkapiDeleting trunkapiDeleting trunkDeleting trunkbtreeDeleting trunkenhancerDeleting trunkfostoreDeleting trunkmodelDeleting trunkqueryDeleting trunkriDeleting trunkruntimeDeleting trunktckDeleting trunktckDeleting trunkDeleting trunkutilCommitted revision . +Did the same and it works like charm just deployed +Add it went smoothlyAdd waiting for bits being published to Maven from Maven central went smoothly. +Yea I was gonna ask you about that. +account created +Hi Amita I got a problem with the file +We have a maintenance window this afternoon +Generally you will use the same one on all of your nodes and you can use system prop notation to override it per node if necessary +Resolved as per David comments +Link to the related Nutch issue +Any comments on this +Added licenses and notice lines in all the binary distributions. +Please attach a sample app that reproduces the I agree with Christoph here +Not sure where exactly this must be fixed +This should be enough for now. +has been fixed so close it. +Thanks for the patch nkeyway +If only issues could be editable. +I dont think we ever had any insight into what was wrong but at least its not still happening +I still like this so Ill write up a proposal for Maven. +Thanks +Thoughts +Pages documents with a few comments +Patch by Kelvin Kakugawa +This will make it a lot easier to test with mock objects too +issue with aspect container beeing stored based on the aspect container class the aspect when app redeployed does not change but actual CL is swapped thus referenced aspect will be the old one class cast +Thats why I will reinclude the correct name in the next release. +Updated patch for changes that were committed since patch was uploaded +was missing though and I added it the ssl test was failing without itThanks for pointing it out +To be able to produce global events we must improve current batch architecture must probably divide extensions on two groups with different lifecycle global and module +It just kills the build +PR merged the installer will incorporate it for. +this patch for the jetty maven plugin allows for jsp compilation +The ORM info is only appended on to the JDO info and so at the end of the day it has persistent properties defined in and these match exactly what is in the class so no exception is thrown +I test with the Jetty starter class that we provide with the project +Time for some more coffee i think this patch will fix the its a little verbose but i couldnt get eclipse to make it some smaller +If everything goes ok I should have this committed later tonight +Thanks NathanAccording to you test case it will try to find the classes by the in MANIFEST as belowIt will find the class in absolute path of a directory but jar file +note we do not want to remove directory which has files +looked at code does not have this bug +please commit and publish to website. +Here the JDBC URL jdbcmysqllocalhostawnautoReconnecttrueAnd i have the same schema name in the Entity Defaults DB Schema +It is intentional! I think we can learn from Woden on how to do this +I like the patch tests but it no longer applies to trunk +Uwe also just checking i dont know javacc at all does it use unicode properties? We have a lot of queryparsers out there +New version of the patch resolves conflicts with recent changes and cleanly applies to review the problem is very visible and causes significant problems when dealing with unit test failures +Resolving the issue +Its possible that we have a race condition somewhere but all the more reason that we need a case that has been known to fail at least once + + +v is a simpler solution if thats the only problem +That would be an alternative but as mentioned before there are use cases where synchronization is not needed +Contributed a patch to webtools which traverses a and zips it up +Great! Thanks David +Added corresponding unit tests for testing without reporting events on move is note implemented in since observation behavior in cluster is changed there is no need in +This has been ordered and I believe is awaiting install and config +Let me add some checking to make sure the file count is decreased +Would be a good idea to check if it works in +But since move of Hudson to new machines this is an issue again +Since is a public class we should deprecate it in this patch and only remove it in the next release +Fixed as part of +changing version number to see if we can fix before so we dont would possibly also solve +This is a duplicate of +The decaf code has not been removed and placed in its own library at this time in order to expedite the next release of as we dont have time right now to work through all the DLL issues that we were encountering +A test stack verifyfix and close +Patch instead uses get so symlinks under data dir works as expected +for reviews +gitignore file it adds +Ok done +Can you supply me with a sample XML payload which triggers this behavior? That would make it a lot easier to solve +Hey DavidIm not sure whats going on here because that value being null means the term is not the index yet is in the Term Vector for that doc +Thanks for replyingDoes the status open means that you have reproduced the pb +Fixed in in +I thought we got rid of +Ill try to replace htmlparser with another solution +Attaching what is applied to +We have fixed the problem in our internal version of hive by simply create the during the semantic analysis phase +new features +Those are from? Dont we have an axis site underThe Axis stuff has not been migrated yet +Patch applied on SMX trunk. +There is no protocol in the specification to close a prepared statement but if the client reuses section numbers in the PKGNAMCSN for prepared statements that are no longer referenced by the user this leak could be mitigated +See my patch in which addresses this issue +K +Hi Tiese do you plan to finish this before November th? If not I suggest to move this issues to Sooner +maybe we could improve the documentation somehow? +Attaching patch againstr trunk +Sorry please remove the print statements from the patch +Scheduled for +Integrated addendum v to and trunkThanks for the reviews Matt and Stack +Work was not done against CP +New patch which addresses many issues with handling nulls and also adds many unit test cases dealing with null data +Carmine unfortunately this is not supported at the moment +The xmpbox module is more or less untouched I just removed an unused declaration +Thanks +I will do so shortlyCheersAshwin +Attaching +This sounds good +Then we get the exact occurrence of every match for free +Youll need the system adminstrator for those machines to configure the or just to talk to helpdesk to help with this. +The diffs issues to next release. +unapplying patch due to conflict with +Awesome thanks Thomas! +Can you explain a little more what you mean with default values +The user management would be nice because many apps are going to need some form of user management anyway +Committed revision . +Yes I can reproduce the issueThis looks like the Null Message will make the cxfConverter treat the to Node be impossible if there is a null message passed inClaus fix resolved this issue and I just added a test to verify it +The default lock implementation is fine for most DB is only for output on console +also unused variable startColumn in to as commit aceadbabfdacbdcb +derby triage +Deadlock still occurs in in behalf of a new internal resources that spans the entire trransaction and locks it on a workspace scope +I think Mr +Given the complicated nature of this bug I would agree with Stack that we not plan to fix it in it be possible changing memcache scanner so it didnt set timestamp to LATEST but instead to the actual cell value? Would this be a simple fix for at least the case where users ask explicitly for a single cell +OTOH building umbrella projects is far worse there are innumerable examples of that this is something Chris is trying to stress +Thanks for quick fix and release! +Further research shows that this has been the case since the code was added byThe current behaviour needs to be documented +Many thanks Sam +Applied +Ill let you know if I have any other similar problemsMany thanksDavid +We should be able to do basic spellchecking for Japanese with this +committed the patch but Ill leave this open until we can hear back from Tim or Gili that this has resolved the issue +For example you can test an RPM jbossas installation without altering the installation and without needing workaround problems with chown and such +The testcase now fails +Ignore my earlier comment error resolution is not Im now checking the response so see whether its XML +This is filed and tracked in as a blocker for Hadoop. +The patch is a simple update to give someone a hint about where their jars need to go so they dont get the original +Id like to release this week. +Is this feature still on track for merging? If not is there an alternative that doesnt require a +why +It is possible for jndi to return different objects to thosetwo lookups +Ive emailed asking a jira admin to create that +Applied mvn issues resolved +We will not fix this for. +Behaves correctly. +Deneche do you think this issue is still live? Is it possible to read any input in general into matrix form +Some users reported that eclipseeclipse worked with rad +Or are you asking for two things here? Only JIRA admins can create new workflows +Missed few files +Patch looks good overallLets correct the related config items to all follow the same naming patternThe test is good BTW +Not sure if I am the right person to take this one +So a slice is seek to the closest column not greater than the start and read until you hit the finish or count + +This was rejected by the reporter he determined his problem in the testing code +this patch introduces some tabs in otherwise seems good +Assigned thank you ll review your PR today +Ive successfully tested this on a secure cluster +Ive heard from others in the company and all I hear is that this is a free product but ignore the real issue +This doesnt require a strict mode being enabled to allow it +Not sure why this issue was marked issue is caused by the fact that we use string literal for both string constant and regular expression +Ive fixed one or two of the easy failures but still working +Kurt all of the existing stuff was transfer over to the github wiki except for an image or two +Didnt added unit tests because we dont have a test plugin for Project Templates yet +Ill send you two pictures with the dimensions tomorrow when I have the time +Thanks James +Ill take a look these are usually pretty straightforward to resolve +Attached full stack dumps from master and all HRS +self Review +Requires for Betaa build +To this end the Map itself is not accessible as a first class object like a context context objects the this keyword will return the context object. +Closing JIRA +I just added a control test to test to verify this as a bug but am not able to reproduce it +The user didnt pass the negative length so is not so relevant +if you want you can get access to wicketstuff svn and make the change in revision +There is no bug and your report makes no sense toaccesssomeusertype is wrongtypesomeusertype is what should be used and of course if the someusertype is not available on the classpath to return the proper classname we cant do much else than assume the type is actually someusertypePlease use the user forum before posting +This patch appears to have been incorrectly applied +The user count of widgets is now displayed in the widget store and widget detail page +I experienced the issue on Windows bit for reference +You specified this against +If anyone wants a build of boo with this or any other patch enabled lemme know +I have to update to the new Axis and then try to debug this further +Have you tried the trunk version of the code +lets put it in warresources for both WAR and EAR projects +Please instead verify that all of the jobs are marked as failed at the end +We can always move them later should the need arise +Minor changes plus some more checks added in the new test caseRandall can you please with your implementation +Changed the to measure the full map rather than enforcing restrictions on individual prepared statements +Stelios I sent a email to you but it seems that the email address is incorrect please tell me how to contact you +I just committed thisCommitted revision +In this case the best course of action I can see right now is to have an operator removing the bookie from the pool and executing a bookie recovery procedure for the ledger fragments on the dead disk +We could I am against I love +Lawrence as outlined above there is no commits list at present +Thanks very can comment out then you shouldnt be needing all these extra classes +Probably a incorrect inline usage change the extern inline to static inline looks fix this see the patch +It works with tomcat +ThanksInteresting point about ejb +Re this is actually testing a class in +general concensus that we should stop adding new to the admin interfaceMakes sense +Understood +Ill execute the test case and let u know the added support for decryption with an embedded symmetric key name +Changed Target Version to upon release of +I did not look into code but I cant find a better idea Scott +Attaching new patch with the changes +cleared the patch available flag since the patch needs to be patch against Patch Available back to true since there is patchv committed +Cool thank you guys +There will be a lot of longer tutorials and text blocks +Applied to branch in SVN rev +Please update if you can +Otherwise we will be back here again as Daryn tries yet again to fixDaryn why is it that after this patch is applied that no unit tests fail on our side in common? If it really is critical behavior and is something that we want to be part of the contract between filter file system and the file systems it wraps then we need a unit test to enforce it +Any comments? Weve had this fix running in production for quite some time now +We are discussing whether the discriminator actually violates the spec in any way as you suggest on the developer mailing list +It will probably be an incompatible change so not really recommended for +Im going to lower the priority on this and shelve it +Apply patch provided by Costin +Actually I applied this fix along with fix for on branch. +NET use cases that leaving it for a release would take too long +Agreed +committed patch +Just client connection unlike kurtsorry +Change toretrieve the autoincrementValue from the SYSCOLUMNS row +I guess you could say that if you have a composite comparator and some columnmetadata then you are likely a CQL table but again not very clean imoAt least internally I would be in favor of keeping the componentIndex as it is cleaner +It properly defines a dependency on whereas junitjunit bundles the classes inside of itself +Dealt with in the close of all resolved issues prior to release. +The patch has been applied please verify +Issue closed with release. +I will personally both of you if this gets in for +Text for jBPM guide and release notesWarning The following is an important note relating to the security of your files are shipped with the platform +I havent looked in detail yet actually +Done but Im not able to reproduce this issue with your project and rulesAnd thats why Ive removed this ticket from roadmap forFreddy could you also please give a try +Set and patch does the following adds a test for extracting outlinks from embedded in HTML +Need some cleanup on the text here +Reopening looks like I committed the wrong patch +The site will be published when version of this plugin is released +That does not seem right the shapefile reader is supposed to check the filesize before turning on memory mapping or not +I dont think were going to move everyone over to multiple Hlogs immediatelyWill take a look at those tests +So I would rather have it there than on theThen end users knows where to look for themIn these places xxxConstants +Im unable to repro this +like so +Reverting because looks to cause failures +So you have one node? And the region servers session expired? If so then thats the expected behavior because the master cannot assign those regions +The patch is ready for commit +Patch is now committed +Personally I was nominated contributorsMy typo +There is now a variable in the binding called terminateHook +Adds new class which fails wo this change passes with it +Fixed thanks for the test case. +contains a patch against trunk r that adds a wicketized version of the Portlet Tracking portletI only tested with mocked objects since I do not know how to force portlets to be taken out of service +Ill try it on the latest +The draft looks good to me +The spec doesnt run with the rest of the specs +Suggest to closedirk +The +Im updating for tokens and keys and to handle startup discarding expired ones and properly using the interrupts lock +Once we will have written this tool we will be able to fix corrupted be fixed with to has been releasedMoving all related issues to the next version +Eclipse Tools cant reproduce it either +It is also worth noting that updates can be either full or partial updates +Hi OndrejI just did the following logged onto dev and dev in created a directory cJBPAPP downloaded ER and unzipped into the directory used ipconfig to find the IP addresses and started up EAP on dev c all b u started up EAP on dev c all b u and they clustered togetherSo you should be able to stop the interface on one of the machines and reproduce the error. +Basically Im defining a custom guice module in the file +I have fixed the bug and I have tested the plugin with your data sample +Thanks Vinod. +It seems that this is a simple backwards compat issue +Committers can you please review ASAP as this patch fixes a critical bug +Looks nasty do we have a testcase for this one +I guess it just take a little while to get used to People are not good at reading the! in if statementsBTW How many of the components are still to be of the same standard +However we should be able to support implicit conversion the comparison expressions +reopened for fixVersion +The affected file has not yet been migrated out to subversion +Since no part of the filters was working correctly with I backed out those changes from the trunk so we could proceed to work release It may be prudent to work this upgrade from a branch to avoid having this issue again on the attached the changed files that were attempts at upgrade to this ticket +We need to present a layout OOTB that reflects todays web standards and techniques +The patch looks fine to me. +This change among others are JSF issues which will be addressed in the Trinidad release which will be enhanced to support JSF. +So that the changes either made manually on the file system or admin api result in the same artifacts always +Posting on the AVRO list shows it is better than them all somehow +Bulk close for +The attachment is a gzipped tar archive +This Jira is no longer needed. +per com a la Gill ha assenyalat un snippet no s pas un objecte fsic i per tant podem parlar dun snippet of conversation que hem sentit. +I did the trivial cleanup items but not the change on the protected field nor anything related this patch of course +Relative minor issue +OK done +r is review board review It has r in url +So this operation looks for a user exactly the same as if hes chosen some perspective to like a good fix makes the links work more consistently +the internal variable to the smaller of the two takes this a step further without any harm +When multicore first became a reality it was multiple Solr instances in one webapp +I was not posting a request for help with my codeHowever I do expect the parser to produce error messages +Considering this resolved with the recent Web update +Marquis and I have prepared a preliminary patch +This sounds messy +Now there is one more tab for Return History with Return Header and Return Items by clicking on it you can see changes for a specific return and related entities +All of the incorporated issues are fixed by. +Maybe it can but with a method invocation result as part of the concatenation Im not so sureIn any case its a simple change +Fixed in trunk r remove second invocation of complete scan in which causes the m getting this same error in Archiva +Yep this was restarting a master around the same time as restarting the regionservers if I recall correctly +Looking for the root +lester I am fine with where it this is heading but the assignments jsp on the integration is referencing continuum taglibs +And here is a patch against from +Exactly subflow will find exception handler and during handling which includes transition to end state and reactivation of event in outer flow can happen exception +Assigning to Sameeer since apparently he cant assign to himself +Attaching patch +Running the source generator with different options fixes the problem so marking this fixed. +Does that mean running the command line wsdljave is my only choice? ThanksRobert +You can see in Googles mapreduce paper that even their C word count example is serializing and deserializing the values +Do you really want the block to be set to times the number of milliseconds in a heartbeat interval? That seems to highoops! good that you caught it +depending on the discussion we might pick that up later again +No other changes to were made beyond what Ive described +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +Closing old issues +No you cant really +It is consistent and convenient to keep a reference to the schema providing full type information at runtime +It does not do this via a job and it does not do this with a progress monitor so ti cannot be canceled +Niclas please to me when the patch is available and I will apply it +Ill ensure that the code gracefully handles the case where the +We dont have plan to update andor rewrite the tutorial for the latest Camel versionA lot of the information is still valid todayThe author of this tutorial is busy on his new project the in Action book +Yes this is caused by our separation of JDK compatible sources versus JDK dependent sources +can anyone find any others in the sources +Applied patch in rThanks for reporting +Using cast to cast from or to short gives a null pointer exception +just applied the patch on +This is simply not supported yet hence Ill classify this as improvement +When a new was added and no inner constraint set a NPE was thown by because it was trying to iterate over the s constraints. +But Ill hold off on commit till we have all the pieces working together and we have tests showing the and foot prints of the the completed jobs +Committed as r +They call it montor topicsSo it should be possible to write a special for Tibco ems +Done +I applied the patch in rev with some text changes in the poms and README +Just to made it clear for me your preferred way is to have pluggable entire classJarcec +We just want to set log level to WARN + +IN PROGRESS hudson job should be set up for running this modules test suite +Fixed in CVS +On the other hand this also means a big refactor to the IO formats since they depend on various objects handed off to us by the MapperContext our Configuration is the easy one to deal withAlessandro I like this idea and I like the simplification +Doing a quicksearch around google returns a large number of people complaing about commonsloggings complex classloading stuff +v +I think most webapp deployments would want some sort of customization the most common of which is logging but there are others +Maybe test? The message thrown should be correct now Client is currently executing another method +Current release is and neither do we change released versions nor are we able to change anything on Maven central that repository is for any published stuff +Heres a preliminary patch +Closing all resolved tickets from or older +The heuristic used to determine whether its worth it to add L to stripe to drop deletes is really stupid what ends up happening is L gets compacted with priority when theres a lot of it and when theres like one L file and lots of data overall this L file always gets split into tiny files when some stripe is compacting +Attached is my updated templatesimple file with two changes Added disableddisabled to double select tag if doubleDisabledtrue Changed to for input tags created if doubleMultipletrue +I have a few years of expertise with JSF and recently switched from to as my favourite component library +Fop should not enter in an infinite loop but display an error +Maybe next week +Need more context around the strategy of allowing configuration parameters to come wit the message headers in the first if we can avoid the prefix its prettier +I am going to disable test case decorators for throttling HTTP protocol handlers +Cant replicate +This is a nice little fix +the same is true for get get might be more accurate but also a bit cumbersome +ThanksInitial import in rev version changed to remove use of to slfjI leave it open until we have some unit tests +Allen system call to GPL software does not automatically make caller GPLIm fully aware of that +Move it back in if you think differently +Looks like a problem of your project configurationsMake sure you really dont use the same JNDI names in your applications +Bulk move of M to from M to moved to are cases where is not enough just use style display +Committed. +Hey Chris have you managed to think this through +Ill upload a new patch once that one is submittedRegarding the docs Bookkeeper was the only other contrib documented and that was in the main srcdoc +Yes +A second plugin could then act as an adapter between JDT projects and the CAS Editor by hooking up to that extension point +Also adding in a check to see if there is already a +Not really +The fact is that the device is a Motorola device and the device has Motoblur installed as does the Backflip and the Droid +now everything works. +Thanks +This will be supported with all releases of and later +patch reviewed and committed to trunk and branc +Regarding referencing a temporary function from a view I guess we could just allow it and leave it up to the user to reregister the function in a new Hive instance before referencing the view +My apologies +Seems this is generic issue. +The preferred approach would be create better extension points in the jdbc connector logic to perform pre and post statement options + +Patch has been can you mark this bug as Fixed once the patch has been pushed to branch +The name is supposed to be but also +I take this to mean that you would like your groovy files to be able to look exactly like that java file +bq +I can offload some of this work + +Thanks for your help +Maybe today people dont email the MD sum of the pom alongside the binary but when you think about it they ought to +Transferring back to Michael Baessler as requested. +The ultimate line of my first description above was wrong +bq +resolved proposed for add check in a unit revision . +I dont see any problems in number gaps +Could you please explain +Github default changed to master +bug demo +Perhaps we should wait toupdate this config until the name change on the mirrors? Or we can go with for now? Either way Whirr seems to run well with so itsa viable fix +Move open issues to CR +Max this is the context menu appear at my workspace the Guvnor is at the bottom did you mean Guvnor should not be appear there +Good point after reading the description again I must confess I dont have an idea what is being asked here +Thanx for the helpDependencies also cleaned up a bitThe patch will be part of next release hopefully next week. +I am not against adding such capability +Thank Hiro for the fix +Changes have appeard in Latest Alpha Manuals. +According to previous comment this is not a bug. + +should also write a paths info back to the client +The has the version number incorrect set to rather than Also there the and versions are also have the same problem but I believe there needs to be a release of those components to fix this see and +We can probably resume progress here now that is in although I hope to get in first +Eric the latest changes look good to me +I have added a bunch of for what still needs to be but not with this patch use css to show errors show all errors not only one +Im closing this up to get ready for the release +I am closing this speculatively since we have had many ARJDBC and releases since last comment and that comment implied things had gotten better + from meNote that a user can change the section he is editing during editDo you expect the user to be returned to the last edited section or the section he was looking at when pressing the linkdirk +This issue is still existent in as we officially support Java +Thanks +I note that we dont do any effort to preserve the streamId if you screw up the frame format but imo its fair game +Created reviewboard +from Danit looks very goodthanksregards aki +Currently using to start a test server +The is changedThat would then need to be +Please post your additional comments to the new issue. +OK thanks +Ive committed fixes for this +thanks pat. +This JIRA may have broken backwards compatibility as testcases that were passing before are failing now +Im attaching a patch that removes handling of JCC and Java in the compatibility test +That seems to insert the license header at the correct place in the html files +Marking JIRA as done. +I buy that the patch should fix the problem and dont think we need to have a test here +I just committed this +bq +Oh +Attached patch but first I need to commit +A minimaltrivial resolution is to make the actual exception object available as a member of the structure this would be less than ideal because it would require updates to code which already works with ACF but it would allow us to work around an issue which is currently a complete for us +good catch +done thanks for your patience +Ive tried it with both of those values and receive the same error message +Would be good if someone can review it +I would find it much more confusing to have Glob be a subclass of Path because now how do I indicate that this API will not do expansion or I have to update all to optionally do expansion when a is passed in which seems like a much bigger change +But we could as you suggest instead make s buffer larger with a policy that data is checksummed as late as possible +Although we have a BPEL Editor component in the jira I think this is really a Tools issueWhat do you think Bob? If so we need to get it moved over and remove the BPEL editor component from riftsaw jira +This page is no longer available +recently the class path for running the junit suite was changed to not include as it should be automatically brought in as part of other jars +Resolving as laterSuch feature would fit into for which my plan is to unify the codebasefor jsvc and procrun into oneSome work has already been done as part of Commons Runtime sandbox projectto allow cross platform JVM extensions with features missing from the JVM andwhich are needed to run a highly stuff like daemons are +The main issue is not whether or not it is part of the first but whether or not it is part of +Can you please point me to the right place? I may have missed it +Resolving and resetting original owner back. +boost May Darwin Darwin Kernel Version Thu Jan PST rootRELEASEX x +The fix is incorrect +Added reporters types into SVN trunk and reproduced error +Bump any more reviews on this one +I will add a test and resubmit thepatch +Create a new jira for anything closing stale resolved issues +done +I am unable to reproduce the compile problem I removed the csc from the PATH and everything still compiled fine showing that there is not a PATH dependency +Thanks +Im OK with get though if you are +In order to add this feature I did the following changed the plugin added the profile action to the the extension point changed the class so that it handles profile modeThese changes dont introduce any dependency on created the pluginIt defines the extension point that makes the associations between the launch configurations and the TPTP data plugin depends on plugin wont be loaded and the profile action wont be available if a user doesnt have TPTP installed +See the nabble as the is a code snippet that shows how to do you want to take a stab at this one +Ignore my previous comment I missed the part that you already added it +pull request to master. +The idea was that the tree could be persisted between repair sessions and ranges would be invalidated as writes arrived then the validation compaction would only need to compact invalid ranges of the treeIn the current implementation the tree will only contain invalid ranges since it is being created from scratch for every repair +I think youve ufficiently described the problem but Id prefer to be able to replay things +First I think I need to remove the archives from minotaur +The testsuite progressed up to which failed and encountered an heap space have attached the TEST file the and the from the server +Suggested patch changing the text for the checked in +You are right I changed the build profiles and now they are happy! Well at least some of them +needs similar logic for processing normal commits whose base revision is older than the latest oneI dont think the specification currently requires each implementation to do something as complex as this +Release +Can you show your current code so I can get an idea of what you want to has already been implemented +Thanks +Applied +project and +Closing as wont fix +Im going to commit this momentarily +Will check the Javadoc warning +This had been fixed with EAP ER. +Native SOAP stack doesnt support JAXWS any more +exception attached +Im told others have access to as well +Ill post back with a reference if this is related +The one caveat with this is that we will need to in the future when the design docs live on different nodes +Alls fineThank you Mark. +It appears this is due to an incomplete implementation of APR threads on the platform at this time and not something that should be worked around in logcxx +Limiting task diagnostic info and status are done in +Rajesh please verify these changes have been applied on +Will file another issue to investigate this I had to add SCM handling for files so they are checked in and removed properly +Verified by Nathan +The reason I did not want to overload the behavior on the existing property was to maintain backaward compatibility +Ill fix +attach new patch from Narayan applied thanks +Maybe something like this Your IP address is not allowed to console access please contact the system administrator + +Thanks for your contributions Madhukara! +Agreed +settings the users use case the proper thing to do is import the project observe the errors on the project select properties Go to the targeted runtimes section check the box next to the runtime the user would like to changes both files AND the classpath container +this may be a dangerous thing using a custom annotation has its advantagesfor example if we use the standard Inject and the spring project is using aspect then wicket components will be instrumented by aspectj instead of wicket +of a bulk update all resolutions changed to done please review history to original resolution type +We removed all mentioned unused mesages from these modules +Now that work is well underway with AS all previous community releases are +Thanks for reviews ted and stack +He is still working with the customer for more technically if the binary content is really really long it could be a memory issue +This fix is broken in AG as well as AG as requested by Ashish +FYI theres a Vote link in the top right of JIRA +Be my guest +Committed to trunk +We should see if we can squeeze this into +I also wrote that I reproduce this problem using the procedure that is considered a workaround for the classes of the stored objects are in jBPM classpath not in the process +Sorry I missed this commitThanks Stefan +Who comes first the null or the not null valueTo get around this I wrote a that allows you to specify who comes first the null or the not null value +project and +class +Merged to branch also +Assigning to Kelvin hell have more idea on why the component fails to get injected and processed in needs to be configured with proper adapter name +Ive also built and which includes the above +Im closing this ticket as the title of ticket is more suitable. +This occurs only for jdbcmysql adapter works fine for jdbcmysqlite adapter +Runtime detection is extended to Tomcat servers +Based on the code the javadoc for Springs the keys should already be in a deterministic order +Its verified during EAP verification. +EdCan you please zip up a complete sample project that shows off this issueWithout one it is very difficult for us to verify it +do you have latest chrome as in latest on the channel + is this tickets parent combining +Verified for +Erlang lgtm though we should make sure this makes it into the release notes +Grid were probably not going to get a runtime fix this late in the process +This patch fixes my problem and passes all existing tests +Thanks +True +You might add a comment saying something like Nested class used so that table is not built unless its used +Keep them coming +Ive fixed this in the scripting java stuff +Committed revision Thanks Sangjin +If there is a problem please open a new jira and link this one with that. +All right was able to apply the patch from Jerome to finally +Thanks Luke! +Attached Added is to so you can get it from get +Ive also included the response for the attached will in the future want to support WFS in and this may require us to support predicates +In any case there should be a timeout for Jackrabbit locks +fail with the patch +If our dependence on the jars is incorrect and we required code changes to fix this we should schedule accordingly +Expand set command so that set can take arbitrary pairs and pass it on to the +Ive now managed to get working as well +The tests for this issue have been added to the existing test case +Moving it to +To test httphostport then click link test fileupload +I agree that BeforeAfter is a better choice than since saving a millisecond isnt worth the fragilityMy point was just that based on your code comment it seemed like the filesystem created by setUp would interfere with the filesystems the nd half tests were creating +They can change it back by copying toolsresources to serverdefaultdeploy + includes a Set Map test that ensures that Enumerated works with +reviewing +The following is the current status of the codeCan you attach the patch +Ive got a honking demobuild box for this +This does not make a difference for youBasically what we need is an image on the left and link on the right +If we do then well need to add copyrights to all the test files in the impl project as well +Fixes compile errors caused by missing file in utils folder in after review. +Also I do not see any references to set in the code either +Hi Guillaume can you specify the exact version you are having this issue with +dont forget that the user is actually the system administrator +Thanks. +Yes there is a file called VERSION in think that the test is enough to check that the is Messaging because that means the correct VERSION file is found and server returns version +Please upload a sample and we will reopen this bugthanksdims +javadoc supports linking to other javadocs so +There are some additional related informations in this JIRA +Ill see what I can do on the side +Here is a patch that can be applied on top of +Yes I found the first case you mentioned straight after posting the previous patch +we did a repackaging of because it uses jdom +mark as +But in these ambiguous cases I am more for guessing or throwing an error +The patched works like a charm with many thanks to Nocolas DanielsAll one need to do is to include this JS file as the last within HEAD and it will take control over the bundled version of +Since what I wanted was just a small extension to I started by subclassing that store and registering it as a service +Instructions for how to run Quick Start with are in the FAQ + per above comment. +Added check for the most common users of it and just checking the empty objectsand their sizes for now +I answered the question and I dont believe that this is a bug +Closing. +The patch doesnt seem compatible with current code base as there are compilation errors +bq +I tried search for related issues before I registered this one +. +Committed revision Committed revision Committed revision +Closing now that is released. +This is a non issue +Committed to trunk +Some more infoThe meclipse plugin is configured to use an external maven output is enabled all other checkboxes are disabled +patch from tip of trunk +I just committed this +I have always used for graph results +Closing all resolved tickets from or older +This is the best case usage of Derby +Just checking in +This is better than the current implementation where any three datanode failures in the entire cluster can cause one block to be lost +WMS would not have been affected +NetIts lacking things that I think are better handled as part of other tickets responses SASLSSL support +It looks like is not in the patch +Committed to trunk and +To be honest Im leaning towards the latter +This bug reportis almost a year old now +To be honest I was confused about this too +After update Ive been logout and tried to access again and this error havent tested the plugin with Grails yet +Code for this implemented patches attached highlighting changes docs updated. +NPE in comment fixed by r +corrupt sstable +Hey EliI filed a jira on infra for it I will with them and update the status here +Committed to trunk +The client is capable to deal with an Axis Java server that requires basic authentication +What you think of this patch AndrewIts your patch plus removal of the blocksize fromIt adds logging on config +done + +Changed the callback interface to reference processing rows rather than the full resultset +decides on gcBefore when sending out it looks goodBut is it better to calculate gcbefore just before sending rather than doing it in constructor? for all given are created at once but those are executed one by one +Was just thinking about test cases + +As a result they may cluster with each other leading to strange behavior in tests that rely on their only being a or node cluster +If there are no empirical data showing that mismatched checksums are either more common or more likely to succeed with more than one attempt why include a separate retry mechanism covering that particular case +got done sooner than expected enabling this to be completed in. +On other weve havent seen this row locking issue so itd be nice to see consistent locking across all the we run +This makes it easy to flip it from true to false +Does that mean that we have no chance to fix the issue in Cordova and shall I report it to Apple or shall we still wait for a fix in CordovaThanksjo +Personally I dont mind which name space my code sits in as long as its being put to best use without being reinvented all the time +Build was failing as this commit didnt include new files +Once a datanode hits sends an empty replication request to the data node on every reply to a heartbeat thus not a single scheduled block deletion request can be sent to the data node +Will commit tomorrow +Thanks again +Sometimes something that existed in the code eg a bug or a duplicate function was being changed if there was a sufficient reason for that change +Closing this JIRA. +Do you want me to add not contains as a synonym for excludes +Attaching fix for branch +Thanks Tom that did the trick nicely +Would have been more appropriate for the resolving for now but you can if you decide. +it might take a while if its a large repo +Committing this +Will commit if tests pass +Some directories has mode permission +The related are still open but known +Thanks +and you are using a hard class reference on the should use reflection to instantiante the factorys implementation +This has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Marcin we should pull the updated tests in the suite already is set up to run them +Administration and configuration guide work +As far as I can see this is a normal uses constraint violation +Im sure there are similarities since were trying to achieve the same goal I already took a look at the in Hedwig but I need some more time to fully understand its design +Should once be since +Thanks Jasha to pointing out i need not clicked grant button +Moving new changes to. +Unless I am mistaken project trademark issues are pretty narrow there are either existing trademarks for in various jurisdictions or there are not +I waited to check it in until I could run one more test to measure how much memory savings this had on my original test case +bq +There is a wrapping of or by an interceptor to map it to an ejb security exception to make it more precise what type of ejb exception occurred +How should we address these individual projects though? Via or one global patch? I can do most of them and compile a list wed have to discuss about +Committed to trunk Ill let this age for a while before back portingOn the backport Ill leave the default contiguous merges +Assigned to you for next steps +Pull it back in if you think different +Postponed +Marking this as resolved +Got answer from Vojtech +Considering this Resolved +This patch disables renaming of submailboxes for INBOX and the is fixed +Paul Can you provide a simple dummy application that exhibits this? Were going to have a new release soon but since youre still seeing the issue in we should try to investigate +does that exactly as Vitali proposed +Or do you mean sources for the actual JPA provider such as Hibernate or? Of course we dont ship those youd have to download the respective product distributions for those to get the have downloaded and extracted +Still need to do the same for at least the following shared datasource metrics shared datasource operations shared connection factory props shared connection factory metrics shared connection factory of r I consider this done +It would be a spring solution which is not but it would also remove all the jaxwsendpoint xmlTorben this really needs your comments before we proceed as we are all arguing about what you want without your input which is kind of silly +Ill check it out carefully +Do you have any other possible methods to do this via OODT Crawler? Id be very curious to hear +A realistic query parser is not likely to be any simpler than this so why not call it simple +I yet have to enhance them to verify the Java model same as WSDL tool tests do but you can use either of those tests to get the breakpoints in +to DOS format +thats right I meant the code that only executes on the masterDo you require a full design of the algorithm in the application +by default the registered prefixes are unregistered once the factoryBean is destroyed +Thats why the opendapps module which often deals with metadata fields including a comma overrides the default delimiter of to be a which is a far less frequently used character in metadata fields +is a good workaround code + please disregard the first attachment +Mongo Connection Profile note that currently translator is NOT capable of returning metadata so using with Teiid Connection importer will not work +I guess MINA might have resolved this problem +I applied the patch but there are still broken tests in semion see attached file forgot to add for integration tests Bertrand started some work in the fise subfolder +Added the description of use cases in +Reporters of these issues please verify the fixes and help close these now when scaling VM on unsupported Hypervisor +Also if I Run As Java Application it also works with configuration showing error that the original attached launch configuration works in Helios with verified the original attached launch configuration works in Helios with +Does the test case help? Can this issue be solved? What do you think? +The files generated in tests differ in two fields testFeature expected testFeature but found actual testFeature expected testFeature but found actual So far I could not manage to find the reason for such a strange run now on and Windows +New featuresplugins included in openshift site +I need to look more closely at and because much of the work for this has probably been done there alreadyThose discussions seem pretty future tense and very much not something that will work in v +I still keep track of the lists even though I havent been as active with Nutch because I have been doing more real time search work Zoie +Ill raise a JIRA to coordinate the work unless there are any objections Ill get started on it once closes out +Thank you. +As for the extension I think a new issue should be filed +still refers to if LOG object could be passed as a parameter we could remove all references to from +Thanks +Fixed in Rev +Is this a basic zone setup +Actually given the code that knut posted it seems to only matter if you change the number of unique values or go from to a number of rows +I think I know what might be the root cause here +That field is present in the the revalidate request to the origin server or it is notIf present then the reply must not be cached +I have run the tests that failed on a Fedora VM and the tests all pass but it does bother me that failed on the patch build +Fix confirmed +They must have been available at some time as the Roo team all have those files in their local repositories. +Changed attribute mapping in and added test case + +I have made a minor update to a comment in the patch +Attaching my reproduction +Test case attachedTo reproduce failure described in the issue run java verbosegc Xmxm Xmsm Ubuntu smoke and kernel +Updated getStream comments per Sanjay Renamed classes for wrapped streams per Luke Removed default buf size per me I was using the RFC max for which is different than gssapi It was the wrong value +Improved patch after writing small test case to verify assumption +Duplicate of +approach seems as elegant as we could shoot for right now +update which manage in a better way exception it be possible to have this patch in the next release +The Namenode uses the name of the configuration parameter called to find the name of the class that chooses block locations +A simple app with a simple controller compiled only methods which includes the full dispatch chain routing rendering mongrel and so on +Thus the actual concurrency you could achieve is still limited to the number of threadsworkload the JVM can handle +Patch to add a test for this patch as the test doesnt need to be isolated from the rest +Thank for the patch +Ask if the user wants to import global assets into the global area or into the new package If there already is a global asset in the new repository with the same name ask if the user wants to overwrite rename or discard +Thanks for the patch Aleks +Please provide acl items from the DB for this tier +Ok thanks for the configurationApparently this is a eclipse running a sun JVM on windowsI would suspect a problem with Eclipse Java at this point +Thanks a lot Luke with all your help and advice +has been released closing issues. +Christian Can you check the behavior with the latest patch once again +No need of regionserver context especially not live regionserver context +to KristianCommitted revision . +Prakash Should I open a new JIRA to schedule all logs for all regionservers? Good stuff +Even though its a complete rewrite you prefer I submit a patch anyhow? From a prior commentI looked at the plugin code for webwork that existed before and couldnt patch it to do certain things so I wound up rewriting it from scratch +now i only suffer from the incapability of eclipse to edit files +Hi Venkateshthank you very much for your patch +Issue is stale anyway. +I edited the subject of the issue for the clearer description +Review comments welcome +It does but v has better do not like that it binds to cassandra storage IP +Quaerendo invenietis +The passing along of tokens should be an invisible service provide by the hadoop security framework +Fixed calculation Allow disabling compaction throttling Add JMX method to adjust throttling Added protection for target excessBytes being negative doesnt matter because we check for a positive value before sleepingStill applies atop +Removing trailing I apologize if any of this submission is out of normal operating procedure for hadoop this is my first submission to this project +We can make the log messages pretty print and live with it see you added partitionerName to but I think better would be to add it to which would allow us to use it in and well as Gossiper +I was trying the nightly build but I cant find the credentials to log in +Hmmm +This likely points to needing to have a common Eclipse code style errorwarning profile as well +Thanks +We could roll a +That new scope is the dependent scope which Web Beans am in complete agreement with you that I dont want to introduce a new scope in Seam right now +Not sure why? I will download official debug version and try it but I dont understand why it will fail for revision if it works with +Try also including the config +Contains one file that configures the database. +the issue about the Google Plugin for Eclipse not finding the module seeIll fix the doc re +Closed due to the release of Tiles. +I tested the patch +Just wondering if this has been fixed in any recent versions of cordova +In pseudo code setting your classpath from within a shell script might look as followsset CLASSPATHetc. +JT can poll the waiting jobs to see if setup is complete for them but this will be done in heartbeat which becomes expensive +Thank you Matt +This will result in a split package in world and the files in bundle wont be found from bundleOtherwise I dont know of a way of getting the without a +It has the serialized participant data which is binary +I downloaded it just now and confirmed it is there +Yea all of the in the constructor is what worries me +Im still looking into the error that we get we a property is set to null by a java clientUpdate both and Decaf class and test +Hi Samwhat you suggest custom Spring annotation is exactly what I mean +Then at we call archetype to actually dynamically create the archetype +But it is not easy to decide what can betolerated +Take module foo +It has been doing for openejb release and release +Thanks TomIve updated the TLPs site and will update the others once the privacy policy page is live +That sounds like an bug. +Maybe it is specified JDBC type but there is no sense in column of type NULLSame for object of type NULLAs I remember it was add for binding when we were passing Java null to so type of Java null was mapped to JDBC NULL typeThat is only case it was used +looking at the current state transitions it looks like it is not possible for to get an APPACCEPTED event and also it is not logical to get one +uploading zip from URL +batch update of issues with fix instead of the correct +Move the fix to +Can we close this issue? It seems like the discussed concerns have been addressed +If having a JSON string is often useful then a returntype argument here would be a convenient neither of you have explained the benefit of doing a struct for this if it was summarising the details of the errors in some way then it might make sense but only having errorsisvalid is unnecessary complexity + +I am really notvery knowledgeable in the library in the RTF spec or in the usage ofRTF in the wild +I think we need to run it again with +I have attached the patch file with previous suggested code changes +Thanks for the pull request Andreas well take a look. +Al youre referring to fixed +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +When a cache archivecache file is stored in or removed from this baseDir update the corresponding baseDir record +The t failure is a big one whats happening is that the we have calls in a BB the first should throw an exception following this theres a flag set operation +What am I missingAlso a simple unit test that demonstrates the problem would be very helpful +Enjoy +Hmmm so looking at the logs +Needs fixing +Bulk closing stale resolved issues +Does Torquebox have any for or other that I could hook into to implement this within our app +we may be able to handle this in the installer since it depends on modification of the file +I agree this could easily be a threading problem +Guys any chance you could actually hook it up back to our Jenkins so we can make sure it doesnt bitrot +Looks like this was assigned to the wrong project and closed by them +Tnx +a static set skippedAtributes has been added to class +patch from committed rev efcaafdbcdcc +Some preliminary comments We can define SERVICECHECK as a command +Im pretty sure that we can implement everything we need by plugging into the interface +If thats what winds up being done Id like a pointer to an existing example so I can stay consistent +Heres the list of that reference that library +is is added +Thank you Deepesh +Would you be able to build in VS which is already available on the Windows slave +At this stage any suggestion is better than nothing + +Based on what John has provided Ive made some adjustments to XMLRPC by adding two services one for retrieving all dependencies of a given artifact and the other one the dependees +Paulthe way people make dinner is Peel the potatoes Salt the potatoes Boil the potatoes Fry some steak Put it on the plateSo if I want to make a tea I should Peel the potatoes Salt the potatoes Boil the potatoes Pour the water to a cup with a tea bagbecause thats the way how potatoes are prepared +still interested in you get the that represents your special form or whatever using this method how do you attach it to a view +Bugfix attached +It breaks abstraction that makes certain later optimizations impossible +The first review version knowsabout a small number of standard SQL functions and user defined functions withfree name and parameter count +Attaching +Can we have a better vetting process in future to either prevent these things creeping in or at least document them such that they dont come as a surprise +Attaching a patch to accomplish the custom error processing requirementAn Interface is defined +and are both implementing and +Providing Patch +Committed to master. +valid values are json or cfmlthis was added because cfml supports a lot more types +lets do this after verify and close this Jira or provide steps how verify it +Whitespace problem is also resovled +Further debugging revealed that putting the jar file on the Java endorsed path resolved the issue +Pushing issues to the next CR +Here are the results comparing for andForbuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimeForbuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimebuildTimelookupTimeThe lookup time in second case seems to have further reduced with a slight overhead for build time +Please apply fix on branch alsoThanks +Assigning to Jon so that he takes a look +The maxconn feature exists in and then in +I do not understand the reference to class value and perhaps this is the issue Knut refers to +Nick this issue is closed so you cannot attach to it any moreHowever looks similar and a sample project would be helpful when I Ralphs solution +I am attaching a trivially simple examplePlease let me know if I should go into a use case based on this example that I consider to be reasonableThank you +Verified with EAP any progress on this please +Hi AbhijitI added a log message to show that the container is being downloadedIn your specific case it somehow gets corrupted during the download and while extracting the ZIP your CPU hits not much we can do about thatIm closing your case as being not a bug nowThank s not a reasonable explanation that the download got corrupt every single time I probably ran it more than times +Make as Incomplete +Bug has been marked as a duplicate of this bug +I will apply the patch soon + +Throwing a exception instead of silently ignoring it is definitely a better solution +bulk close of resolved issues as part of JIRA housekeeping for release. +thanks Mark and Bobby +If the header specification in is not removed then both dates appear separated by commasFYI The headers emitted by uuids now have +fixed in trunk +The Lucene stuff uses a javascript charting lib but I dont really want to deal with javascript would rather stick to java when I can +Assigned the issue to myself +ll clean this up during the next release of Logj +There will be no release +Okay I added that too +the Gabor this is very much needed indeedWelcome to Mahout +So make it so if www exist in dir we use thatnote im just trying to find a default that works for most cases at launch time instead requiring logic when creating the config and customization is still allowed to set specific values +Thanks a lot +Ken its up to you whether this gets done before or after beta +People can pick and chose which Spring version they want to use +okay due to the number of votes and since we seems to have a working solution that fit Ill review and apply the patches in the upcoming version +H or just a specific one +Refactored and to use Shrinkwrap to build the packages needed for the tests +This patch fixes the typos +Thats a big dealBTW are the semantics different for localfs and hdfs +Since its not intended to be ran as part of what do you think about removing the hooks in? +Marked this as fixed by +You are right here this is output as an error and it shouldnt be as it isnt +Just to clarify you are talking about the Client and the Processor the for the Clients Iface makes sense but the for the Processors Iface doesnt +Do you have plans to fully solve this problem in future so that its compatible with ACF? or is this something Railo wont be supportingThanks +But it is not the case here. +Committed thanks Trejkaz! +See the below linkhttpslocalhostcatalogcontrolproductOpen the Order Status block in the bottomHere you can specify the default order status for any sales orderGive it a try and let us know what you see Divesh DeepeshAshish Hello Ashish I did as per your suggestion +The patch fixes the issue and at the same time increases test coverage +If it was then we should mark it as deprecated for and remove it after thats released +Went back to JBDS and with EDS +Fixed in xbean in revision I also added a test case for in a WAB to Geronimo trunk in revision +We will revisit the case after Apache Dir becomes a SAR and their is some community validation. +ack +Thank you Jing for the code and Nicolas for helping in reporting solving the issue and testing the solution +Sorry for the confusion +Thank you Benoy. +Rev +Done now +Do you think you could open a RB request +Issue appears to be between jbossesb and jbossws +too assumes an model for the most part where all of the things that are in cache are already local to the machine in caching makes this a had I wanted it all in my application in the first place Id have used the cache to start with +It looks like port to is all that is needed +Since ganglia Host struct contains more ganglia specific attributes I think its better to keep away from +Even if there are no benchmarks any guesses as to order of improvement in performance if we got rid of of messages +This feature exists for both onException and dead +You can use double quotes but no spacesThis needs to be address in the shell +a make target to run pylint on python sources with just tab related warnings enabled +Please post the results once you have finished running the tests + +No response assuming ok. +Thanks Alan +Use the standard Hudson colors Blue Success Yellow Test suite error Red Build brokenI like the of the bird head used in the banner so maybe that can be used as the base +lets you RELOAD a core +I tried it with the new snapshot version and it did indeed work fine +This patch add the new Version ctors and deprecates the defaiult ctor +I am doing a similar thing in where you can obtain the metrics via scanner not the scan +Thanks +I did something similar in my project extending but your solution is more complete as it implements a new parameter to make the filtering a choice + issues resolved in released versions +thanks +One more comment related to what I mentioned above wrt +Thanks! +The request reaches the successfully but does not have the form suggestions +committed to backported to released issues. +Of course I did not build the site and did no release +Marking that the server is externally managed means we will do no polling or verification of the servers state +If we want to keep them as part of version control we can check in a as well +Commit to branch and trunk. +Applied the file as all unit test and integration tests pass with these changes +The network Im on worries me +Please punt any of these that you dont plan on doing to by +The reason to keep the empty files is to avoid picking properties from confThat sounds reasonable +Attaching an updated for posting an updated patch Andrey +im agree to update groupId of old artifact but i think we dont need to use +Seems I need to install Fedora Linux x first since Ubuntu Linux x has no problems with XPCOM Welcome +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +OK Ill assume it was fixed as a side effect of +Basically pretty much any JSF web app out of the box that has HTTP session replication on will get the issue unless client side state is maintained for JSF +bq +Thanks +In my experience when developers see these problems they start doing clean builds every time making the build time several times slower + resolved issues +Assigned back to patch triage team +we need did you say that you no longer need a JDK build? If so then this issue can be marked as Done +I forgot to say that I did not add new tests since the changes are obvious. +Hello Sumiti have added drop down as Invoice Type in Ap formAttaching the patchRegardsSurya applied to r +ExportImport as in the case of Load operates at file level rather than at record level +As I said plugin is trying to initialize embedder on startup and if that fail plugin is not initialized +You can insert the pull request there +Without this patch operation will be blocked +The problem is still there even with MOJO and MOJO applied +So please attach both new patches to +Andrew can you review these changes please thanks +Stupid me +Waddya thinkI thought of doing that already but really I considered the patch I attached as something to evaluate firstAn om should also contain the om classes already now in commons +No it would not +We should check it after commit and close this one as duplicate +I reran the project on my Mac using Java as well as and had no problems either +Is the motivation so you can use D? None of the other generic options seem like they apply to the DN or NN +richmessages component have been modified to keep messages list and update that with messages from the current component only +Can you provide the content for the guide at all? If not then this will miss this releaseThanks information from the BRMS team I am unable to fix this issue +Are you sure you need them +for prefixing control parameters with just a could you give some examples of how the sling and prefixes will be used? Just to clarify +Is there a specific use case to reproduce +Updated snapshot pushed. +However as you did many changes to javadocs I decided to run ant javadocs and it fails +what do you mean about mysterious behavior +I forgot the two new java classes +Looked harmless just a check of rather than +You can probably close this only thing I would be concerned about is that even with the fault surely the message should still be put on the local queue? If the bridge is having problems that should not effect a local send? +An upgrade from to should be ok for a minor release +What about secondary indexes? I guess youd like to be able to cache hot columns in secondary indexes CFIts a good question +From the comments it appears this is not a Test issue +Will take a look at test and test with subtasks of +It became a documentation issue +bulk defer of open issues to +What does the result if user calls context methods directly? This is undefined and not explained in specification +I added one to set pos back to +Ill continue with axisunsignedbytet then +If its new we can probably go back +It might be a good idea to add this to Pig +After importing injectgizmos dont forget to delete categories description else you will see nothingPatch tested on a fresh check out todayIf nobody complains I wish to apply this patch in some daysThanks +Should be easy enough to fix either through the config or in the filter code itself +As you can see relates to but ok Ive added one more link +this is not a bug +inspiring +Marking as Wont Fix with +Actually looking at your email address I have a good idea which scheduler you might be usingP + is not planned to be included into EAP is for EE so it will be available in EAPBackporting support for CXF Stack in EAP is not possible due to some technical issues discussed with ws guys +It still complains that it cannot find the file to patch coresrcmainjavaorgapachemahoutcommon +As explain by Alex this is the intended bad moveI thought it was a bug but its a feature +Closing as Cannot Reproduce for now might reopen later when we have more info +I have attached the latest code. +This was a trivial fix I had placed this hook therebefore but forgot about it Use the jsFunctionattribute to specify the class path to the jsfunction. +Any chance of meeting such a schedule for this issue so we have time to get the changes made and into a release for people to test with +A new patch updates some code in the +Fixed upstream +issue is fixed by wtp team test is passing so issue can be resolved +Navigating through the website is wasting my time + +Committed bare bones version of the web extender in revision +for now nothing beyond servers view makes sense. +But my experience is that most users are ok with a couple of jars in a lib folder as long they dont need to maintain those jarsThe user experience should be Download zip fileextract filestart shell scripstart using pig localchange some config valuesuse pig distributedAs long getting started with pig is as easy as possible we can grow our user basis fast more contributors getting things stabile sooner +Why does the connection timing out cause a problem though? It should reconnect the next time a request comes in +This looks good +and wrapper class got missing equals will be funny but doable +Please review. +With a build including the changes for I do not see ERROR XY anymore in the nstestHow to go on do I close? Or? Or both +The xsdchoice element was not handled in the XSLTAttached is the patch for this issue +Forgot to mention that the revision for this patch is +Good idea Bryan +If we were to remove the plugin architecture we could use dependency injection in its place which would make it easier to change out components and to test various component piecesRight now most of the objects are created through +If label occurs in input field richmessage will use this label in validation message instead of full elements ID +Im attempting to get some heap dumps so I can compare them +The critical piece is figuring out when it is appropriate to give the counter to an or the merger +Resolved in Git ID fcbbdbeeecbbdedecdba +If youve got some spare time feel free to help me out +ThanksCould I contribute with a patch for review +Filed for the test will split that out of this patch +Adding tests especially around edge cases could accommodate Teds concerns and it doesnt really hurt to be extra defensive when coding on sensitive code +Going to commit this patch +That was surprisingly easy to build locally +the okcancel page is an editor you cannot remove anything from there +You will see that it is possible to apply value using immediate. +I put the changes in to the final release of +I only show the output of Firefox for comparison with ChromeChrome doesnt load the projectWith SDK I stop the output because it loops the progressHandler in an endless loop +If the call is too long then it should be processed asynchronously imo. +Yes I can reproduce this failure too +read the document +Thanks +A fix was submitted as part of to bring the trunk fix for this issue to parity with the fix +build looked fine so Ill just assume this was some sort of build glitch. +But it wont promote the slave to master so metastore can proceed with read only queries on DB +I agree with all of your points and I am working on a new approach that closes these gaps +YeahClosing this issueTom. +Empty in the meaning there are two examples but they are not explained +This patch is for the branch so the above results are not applicable +When the fix went in we noticed a few queries in wisconsin were not doing sort avoidance any more +It would require our users to start a particular version of thrift depending on which one they wanted and would make it difficult for users to migrate unless you also came up with a way to run both +Updated version of the patch posted to +I had something like Csomename and the n got interpreted as a new line no matter how many backslashes I stuck in +The schema or WSDL imports will always point to the original Webservices location only as done by JBR gateway +Moving out of after chatting with Ryan +Thanks for the comments Nicholas! Update the patch to address your comments +committed +ChunhuiHave you verified patch v in real decent sized clusterMy concern is that the safe mode would make cluster startup longer especially after a critical issue caused cluster shutdown +Fixed by +NOTE I see the correct behavior in the build +I just verified it +We really need a single source of truth +Attaching +License information is missing in the APT file +This code should pass type checking +ok +Could you please elaborate this a bitRegards a look at the class +The client works with both jdk and but the server only works with +Committed to as well. +ScottI cant say Im convinced and am in fact more concerned from your example given that this approach essentially builds dependencies on all of those projects into Avro +Tests work when cassandra jars are in the lib directory +Part of a bulk update all resolutions changed to done please review history to original resolution type +Move to us be issue fixed in released version. +Make sure that you compile this program at bytecode level or earlier so that you can run it on Java +Thanks Thejas! + hasnt a central system to describe admin roles +Tried the version version upgrade as Alex suggested test is still failing +uid is what it uses now +check to disallow nested is missingWill add these and more egs in the next patch +Hence we do not allow a local disk to be placed on a shared pool and vice versaMoreover if there is only one local storage pool on first host we cannot migrate just the volumedisk to another local storage pool on second host as the vm is running on the first host +Hi Rahul this is ready to go ahead as soon as Camel builds complete +If you have a servlet transport please add it to the then only it will add to the endpoint list in wsdl +John think we can get this fixed for in the next day or two? If not lets move this out to +just committed little bit late on this +Open a new bug for new featuressuggestions. +Ive created a minimal project that recreates the problem +fixed in latest in trunk +Initial patch attachedI intend also fix all places where we use the older names to use the new names on the next rev of this patch but wanted to get the basic change out first in case anyone has violent reactions to the new proposed names Naming is always the hardest part +So what is the scope of the permissionThe vdb property enables the use of the given languages in the vdb +In this case we will have as FINISHED with the auxiliary information of to be KILLED +The first subflow called works second subflow called works final subflow called completes successfully before main process fails with constraint slimmed down workflows +The class comment in could be updated to reflect the new check +ok so this work but only for JSF projects which I cant see why that should be limited +Did you see the comment in the wiki pages about rebasing when in EU +I now have cloverclover running but sonar does not for the same reason in the description +If it is sure that we do not cause a NPE in our catch block commit +Internal patch that we had ported to the latest branch +Now that work is well underway with AS all previous community releases are +If you could attach the entire trace log when this happens I may be able to see something +James seems missing in the changeset +To fix it consistently we would have to implement working copies in model which is a very complicated task +Works on my machineI still have mixed feelings about this issue its adds a layer of complexity that doesnt seem to balance what it aims to help at +After upgrading the unit tests to use Derby instead of HSQL the tests pass again. +This is to document the config more then to set some sort of a default value +Im almost done getting an initial patch for this just one issue remaining +I dont think this is concern +Looks like some value in our standard screwed it +But how would you like to register it? A config property or something else +please cross check. +It would be great if it could be fixed again +Patch referred to in previous description +Attached surefire report for +Ill close this jira since was and work on remote jndi is in progress +Some operators do not have any internal state that requires rewiring +gmcdonald thanks for the comment +Eek how awful! Ill fix +This is in modulessrcjavaorgapachegeronimodeploymentcli to wishlist +Yes it would thanks +SuccessFUSE HQ downloaded this morning and installed autodetected FUSE ESB +Something that come up in a conversation with another developer was the idea ofusing or Struts to make the checks +Regardless the value of resolve when I start an exclipse workspace I see that Ivy classpath containers have been emptied +This is a regression and should be a blocker +jllesieur jllesieur lrwxrwxrwx jllesieur jllesieur homejllesieurSVNROOTdistDebuglrwxrwxrwx jllesieur jllesieur homejllesieurSVNROOTdistDebuglrwxrwxrwx jllesieur jllesieur homejllesieurSVNROOTdistDebuglrwxrwxrwx jllesieur jllesieur homejllesieurSVNROOTdistDebuglrwxrwxrwx jllesieur jllesieur homejllesieurSVNROOTdistDebugNote Ive checked that alls of these links points on reals +The jar is included in the distribution +I cant recreate this +This call returns +Do you have an idea about the lifetime of a bug +I like it +Committed to trunk with revision Three different algorithms are currently tested AES DES and BlowfishStreaming ciphers do not currently work with Derby +see for completeness if you write a servlet and register it with the http service you can get the resource resolver and get nodes from the repository so can be used +So only things that did work in but didnt work in should go into +People with more needs can do their own mapping +Need to convert to XML and commit to repo +I think we are both of the same opinion +i appreciate your giving it a good look and completely understand yourreasoning +Im this pending clarification as to whether it will end up in or +fyi there are some things i dont like about it and i didnt look at the directories in miscbut its worth thinking about i guess +Ive made a commit but Im leaving this open until I can verify with team what happened to the mbeans and if theyll be back any time +Extractof proposed patch attached +Linking to which contains the fix for this issue +Can you checkout build verify and close please +SHAccfaecaafecac +should actually be done for I fix the Welcome Screen with the proper Getting Started link forAnd what shall I do with Viewlets and Tutorials and News links +Thank you very much for your contribution BilungJarcec +I thought Id searched the programmers guide but obviously not +Ive gone ahead and committed your patch in the meanwhile since Ive verified that it works for the problem in this jira +I have resolved this issue as all requested changes have been made and the draft document on +Should I trigger Submit Patch? Or no since its not for trunk +Im going to try more databases just to make sure its not a jdbc driver +Attaching patch to fix the warning +Pavel take a look at this and see if its worth pursuing +I do not understand the problem behind the issue +Fixed. +Could you post your java problem along with the xmldtdxsd files causing the problem? Thx +Uploading a +Closing issues which have been released. +Or Boolean for that matter +we just have to make sure on is only executed once per session. +Hi Howardthank you for this +I especially like the looks of + +Yuki I think everything is ok +It is an expected behavior for visioned pages +Hi Eranda thanks for attaching the patch proposalYour diff file also includes the changes from +So were going to close this bug and open another one and the new specs +As far as I can see we never implemented this +already fixed in +Indeed see test added +orgapachejuddi is not the signed for your great support. +If I move that test app to a directory sibling of the plugin and reference the plugin using +Attaching test cases for chemistry with federated modeshape +All others pass for me and seems to work fine +votes and watchers so far tells me people want this in Lucene + will become part of core with the next major version +Thanks Mahadev +Thank you for pointing this there is indeed an inconsistency +I would argue the should be the default as it will work in most cases and make Caspers issue the exception. +Heshan I agree this is a needed capability +Once its tested and ready to go +should be tested with to test with JDK s tool to localize the exception ve tested with Jconsole but didnt occur any my last comment my statements hold for exo should be fixed by be verified in gatein master and reopened if necessary +seems like a very good improvement whats the holdup? I guess well have to wait for to be out no holdup other than we didnt want to keep adding stuff before was out and this could wait a bit +See if site updates are working again +oh no! you are right this is the wrong patch sorryIll add the correct patch this evening its on my other laptop +I will test this solution today and give feedback +So apparently the issue is caused because the registration div is absolutely positioned +Split out comments into informational table COMMENTS ON statements and column COMMENTS ON statements. +Also this could be fixed alreadyHave you tried the latest version of Crimson +This way the performance difference could be less of a serious issue since youre paying for it a fewer number of times +Remove adaptor reference from adaptorPositions hashMap when adaptor is removed through +The arbitrary topology ticket wants to address that and this ticket is intended to be prep work that is legitimate in its own right +closing issue +The patch lets the recommissioning skip the check for dead nodes and logs the total number overreplicated blocks per nodeI expect the future lock improvement will reduce the duration of namespace write locking +Merged to doc branch at revision . +Not a bug as its how its designed +I took a look at the patch +so +Moving off indefinitely +additional javac warnings are due to as expected +So it would be good to finish this sooner than later +Incorporating it +Yea if its that generic I suppose it makes sense +is it configured to be run on jenkins? not sure about thatas far as i am concerned i dont have time to run it +Would it be possible to have some idea of how long this bug will take to be fixed +Matteo Ive done that already looks like my got messed up Working on pushing up a new patch +Something like +Closing. +Reopening due to issues discussed on which resulted from the changes sin this JIRA +Hi Ive moved this to JBPAPP as its not something specific to jbossws especially the upstream said since you already looked at the various jaxb lib used can you please write here which change is actually causing the problem if you already found that? Generally speaking all changes besides the major version updates are due to patches from us because of JBPAPP issues +I then went through and removed all references +No +Actually I think that Groovy does make a static type available but other script languages might not so it doesnt make sense necessarily to make Groovy a special case +Committed thanks for testing +Please have a look if does what you need +bulk close of resolved issues. +The enhancements to the core extension classes look like theyve made the container impl a lot simpler Ill look at what can be done with them and the container +If you are not using them then use the standalone examples or compile the native map +If we start caching +SSL are enough +There is no business case with x null in SQLThe fix only allows a better error handling. +Are you in a position to supply me with something reproducible +The patch looks good to me +I think its redundant +Can you check +Its disruptive but having worms or similar constantly taking down your site would be a lot worseIs it possible to add a wrapper over freemarker or jsp el or an interceptor that is only run when freemaker or jsp EL has been running +John see my comments in +The wiki page seems pretty thorough but I cant verify if absolutely everything there is correct and up to date without running the Spellchecker component myself which Im currently unable to doId mark this as resolved and update the Wiki whenif weothers find errors or omissionsMike +And I am soon off for a one week break as well +Adding maven tried to use jira markup like underscore around words to get italic +However if I comment out the log statement that causes the NPE and repeat the experiment I do end up with a row in SYSSTATISTICTS describing an index that no longer exists +off topic patches moved to issues applied fixed in revision +Attaching failing test code to reopen issue +I uploaded a new patch here and on review board +I think your comment should be a new JIRA moving from file per plugin configuration to one config file would provide more readable access to itSpeaking of this issue we should be able to read everything from CLI without even touching any configuration file like this p vmware o thindisk true type personal d local o path homeoddthesisbuilds overwrite true package falseor similar wdyt +There was also one other original which had remained open which I closed today +Ruby anyone +that gives a out of heap memory with me l fix that +Duplicate of an already fixed bug +Committed a patch +Do you want to do that? +During the simple test of redeploying a minimal web app these classes leak one instance each time for WELD just fixed the leak that I was talking about yesterday +Committed revision All tests passed intck failures down to standalone and incontainer +Moving all issues to the CR +works logically exactly the same way as and and it would be illogical to have it function differently +So please check with SNAPSHOT and let us know the statusThanksAny updates on this? Shall we mark this issue as resolved now +Thanks will get this into the release +I prefer not to tweak myself as I am no Derby guru and cant go too far in Derby internals +My fonts are OK +Oops missed to close this issueReason is when I came to close this issue then I noticed that changes are requested port into so forgot to close itThanks Jacques for your quick turn around on thisAs mentioned by Jacques commited in trunk r and rAshish Vijaywargiya +works but the source distribution now includes jars that we dont want to distribute. +Marking Patch available +A new patchMoved the shell execution code to a new package protected class +Closed per above comment. +Attachment has been added with description has been added with description Result of proposed Using always FOP will try to keep the complete table on the same page resulting in the attached PDF +Reopening Im going to take a close pass over this +Karl Can u post a simple test client that u used? I would like to look into this issue +Tested for several project and well workedThis implemention is derived from C code +Patch fixing the testcase +Similarly for getset +In if fs is null and cache is empty a shutdown hook is added to close all filesystem +The will be rejected with respect to Geronimo since that is a Tomcat issue +Gotcha thanks for explaining +Please see the user mailng list archives Ive just answered a similar question Andrea it works perfectly. +I made my changes +Yes even Hadoop agree to fix its a long time away +In the last step of the code generation with the plugin I got that exception and so the code wasnt generated at all +Cant reproduce so on later +Will work on this bug once is with latest build fromWe do not see Rolling back the transaction messages in when soap calls are made. +Time permitting I will attempt to get this in for M +Some people use JIRA some issuezilla +Thanks a lot +The javadoc plugin in the reporting section seems to be the trigger +It is a mistake that write operation still remain after patch +Would be a cool feature for the security module had exception handling integration enabled +Justin Daniel just committed support for the new option in and develop See +Here is the regression test for the issue + to patch if similar problem does not exist during reconfigure + branch no more under active development +I checked the jboss admin console + +For an example of a test where this is used see javatestingorgapachederbyTestingfunctionTeststeststools and the corresponding files associated with it that makes use of the supportfiles attribute +On a second thought the is created during the bootstrap when not all dependencies have been loaded +But anyway we can do a check for both senderand receiver and take the transport name from one of them as appropriateYes we store the transport senders and receivers in the axis configuration +writing a single writer than encapsulates how filtering works +Unfortunately the permissionsprincipals model does not fit perfectly to be configured in the property based configuration files so I guess we have to include a new format +Separating the concerns in the repository by using different workspaces is IMHO a good approach +I will attach the latest patch +Thanks milinda. +Formatter does not work any more with codejava code +Committed wnit fixed +This should ensure that WTP is used in not just JBDS but JBT too. +I have to figure out whats going wrongHave a good daySascha +PRP works fine as standalone from one appportlet to another +The issue may prevent the Test to be executed fixed all JSF Wizard tests on. +The tooltip is still confined to the datacell therefore being hidden +Let me know if you find more issues +If you want to keep pnsslt opaque then I think that the signature on pnsslinit needs to change to pass the hostname thats being connected to +If you finish it before we release beta please correct the version +Are you interested in having same content in a couple of places or just a paragraph of same type to edit afterwards +Actually DIH started as a standalone webapp inside AOL +Hi Ole Sorry for not promptly reviewing your revised patch +Its fixed now anyways +Verified on EAPEWP ER +I will commit it shortly +So lets go with cdh cdhmr cdhyarn + +OK Ive significantly reworked this code to workaround the bad lifecycle in you please give this latest change a real hammering? ve recompiled with the same result judging by the exceptionmy log messages were gone +Committed to trunk as of rev +what version of Maven is being used here? The external wagon is bundled in Maven so the version in your repository or the like Dennis suggested generally wont be used +In revision I committed the new classes from the latest patch +In the pipeline rework pig will use hadoops local job runner to execute local jobs +fix has been checked into HEAD +I would assume that the majority of users are NOT using the IBM VM why should they be affected? Even more though there is the aspect of correctness equals is not a method which does need security permissions +Some very large performance gains can be acheived there +bq +Have the agent parse the iptable rules and have it find out which are not applied anymore Have a buttonAPI call rulesThe first one could disrupt traffic for a short moment probably not desirableThe second one is however not that easy to implement and could potentially hurt stuff when parsing goes wrongI think option is more doable +It might be time to pull in Hadoops interface classifications and mark that method as private unstable +chapter and should become sections of some other chapters or something else as they dont look very logical on the place where they are now +As it is this is something that was mostly out of a project I am working on +Product management has decided they dont want this code in the project +I havent looked at Libbys sample yet but I would also like to take a look at updating the sample as well +Ill add this one to the wish list for the swing module +Integration tests were updated +Actually the error seems to be in +The bug was resolved but the general approach to setting the needs to be simple and consistent across all components +Plain and simple +Here is the fix Mamta requested +Bulk closing stale resolved issues +Why did the issue drop away and what do you think is required to get it back on track and possibly in the codebase +I am reopening this enhancement request and has uploaded an updated version of +Verified in the JBDS build. +removed erroneous comment +wrong on the codeFixed but you dont need a release just for saw a hang on my environmentBut this is really about the test so its minor +lgtm +The changes Sebastien made looked to me as if they actually introduced errors +Appriciate your help if you could move this issue to relevant groupAs this is not any Apache Any specific problem I will duly close +I have an updated test that uses a latch to wait for all threads to terminate before letting the test complete +Sherry That sounds good to me! Please resolve this as WONT FIX +I think this is a regression in that has already been fixed in the trunk and will appear in for more details +I think my only remaining concern is the default interval which youve said youd address +The change was applied at r +patch based on feedback +Yes every fluent method creates a new instance +Attaching stylesheet input document and expected output +Patch applied with thanks +Hey EricHas there been any progress on the ODBC driverThanks allThe ODBC driver that I am now working on supports most of the basic ODBC functionality for connecting executing queries and fetching results +seems OK do you know who we can assign this to +The same failure occurs to me also on plain trunk +Is there supposed to be a patch attached to this? I dont see any files attached +Is this simple features or application schema +To keep back compat we would need to add a magic name to the map that we know isnt a shardproperties +It will need to be done to implement so maybe just do so over there +pre patch and patch applied to the snapshots branch +Evan If you poke around you can find the logs of your test run up on the apache build server +So to make all this work I had to use Spring Data JPA again and disable Mongo from my app +Im pretty sure that doesnt really deal with the concept of extensionsendorsed but rather just deals with what packages from the system classpath will be wired up to bundles +Ive uploaded a patch for this +In our tests we are killing the active namenode every mins and some times we see that after the transition from standby to active the namenode goes into safemodeIn this case since failovers are frequent NN entering into safemode is possible +Looks correct now +Ok thank you for clarifying +It can be applied to jBPM. +Alternately maybe configure LOGLOCAL dir permissions? which could also be used to ensure the dirs have the correct permissions +Aron shouldnt have been assigned it in the first place +Since I had no luck using the propety settings suggested in the Oracle FAQ I think a quick fix is in order +Shouldnt s timeout param timeout and not block forever waiting for FC? And if it does is this by design that the timeout param is only about comms timeout and not for the to return on an invocation? Will ping Bela and check on this +Agreed there isnt any checking the currently for these conditions +Resolving as Wont fix +It looks for it in the accelerometer directory +ChrisI tried to go with but I ended with a whole lot of mess after this fix Ill try that again and report if I find problems +The current code changed so much that the patch is no longer applicable +The ptBR files and with the patches applied +I really like this approach although I would choose a different name than hint but nothing comes to mind at the moment +Do you want all the testsRandom readwriteSequence readwritescansOr just the scans? The more tests the longer +I swear it is the last time +Cleaning up old if necessary. +I just tested it again and it worked +We wont implement lazy loading for this task since the time frame is too short +Almost as if there should bes a patch for the Javadoc it adds a hint for those who are in my situation +If you have time for it and do not use advanced features its a good option untill the Richfaces datatables are updated +It is possible to specify policies in the Integrated Registry +Of course you are right +I will have a look at and come back to you if I find something. +It modifies to use the new listStatus API implemented in +not have to import generated pb but rather the or some such similar class? Just say +The patch is definitely needed since the outpout clearly shows that the old deployment method does not work with AS +There is still some pending details related to fview transient attribute +and are still valid +Youre good to go +Keeping this jira open until we can tell for sure that the original issue is fixed +Looks goodShould and be combined? If not is usful outside of analysis? Should it be in +If a tarball is being used the user has to untar the tarball somewhere and set the path correctlyAlternatively she can always define hod script d n nodes foobarhadoop jar wordcount input outputMore generically we dont interpret anything inside but pass it as such to the shell +The proposed patch does the following Adds a utility routine called check to +Will backport shortly +HiIve added you and Tony +It compiles fine for me using JDK +I just committed this +Are you ing this Ted +Reopening this because the description sounds extremely badWe should not randomly search the whole project for these files there could be multiple ones and some might be temporary files unless ive totally misunderstood what this issue is about please rollback this change +Id suggest removing or adding an assert test to it +Compile once test everywhere I guess +How do we want to proceed with this? As we threw out lots of the experimental code maybe the annotation is not necessary anymore +I have a interest on working on this project +This occurs with or without a custom also occurs with version when the version of Ivy is +can have values that must be converted to two java characters +Hessian and Burlap are also affected in the same an embedded Jetty solves the problem and httpinvoker calls are answered immediately with JDK havent heard about that before + evanescent relates are deprecated and should be removed completely in can I find more documentation on ignore pleaseIs that deprecated too? +In this case need not extend the +Thanks Dushan for the contribution. +This issue is similar to +Hi devsI am considering to add this theme into the trunk +Instead of reading every byte we read block header and aggregate them for a tablepartition +Please tell what you would have happen if not an error telling you what is missing +Im currently using seam snapshot and manually applying this patch myself +Is there something I should do to trigger again the Hudson build with the new patch +if it had stuff in it +This patch adds a new section to the Developers Guide nested under the existing Setting the SQL standard authorization mode topic +the metrics changes look good thanks for catching that +oh well not a first case like is developed currently against the release of nb jee team is adding methods into interface which is a not compatible change +Quartz contains a that is a poller so possibly it might be better to point people to that +I will fix it +Thanks +Tested the new version and it appears corrected. +I checkd all logsbut there is no error +Hi Andrew sounds good! lets collaborate and we can see how to work together once we get more info +Bugzilla cleanupI suppose this is a bit now +I forgot what I have done coding change for this in the patch +I just made it a bit more Eclipse standard +Voldemort does have an rpcish remote API like we do and then they have a thick client +The service descriptor is in that module +Thanks for your contribution but it sounds like being an apache mirror wouldnt be the best use of your extra disk +I filed bugs with to fix it. +Main page and daytemplate pages should be allowed to be any page +Hi Markthanks for the helpDo you think you could write some tests for these scenarios and maybe propose a patchThe classes involved are and the problem is that when we get the id we dont mark the paths in the object +Also do you know if this is documented? I couldnt find it in the documentationthanks +See the resource decoration example on how this works. +Thanks for the reminder +Just drop the min user id to in the configuration +We should reopen +Hopefully Jenkins will catch this sort of thing once ZK is out +In your case when conf did not exist were there queues configured in +Let me know if you need help styling +Cant reproduce integration checked in revision +I just committed +Sending a message incl + +PeterThanks it now as before with the paintComponent instead of +Ah I misremembered I guess we are in the same boat Hopefully someone else will pick this up and commit +Fixed in master and +thanks for your feedback +Thx for the patch Leo. +Did nothing but set but the problem remained +Kostis please submit a single patch that solves a single jira issue on the issue youre solving +So here is the patch +Attaching another example where at least line breaks should be preserved +Well commit them to if will be fixed before release +I am uploading the patch +Forget about as renderer type +Adding style to the div solves the issue +Fixed by applying simplest patch in rev +Thanks +In C the and are the hook points for this +Tests are working fine for me +Vinay thanks for looking at the patchIt looks to me like there is a JIRA limitation that must be in the same project as their parent issue +It is based on the patch v for +After the release of we need to understand these app and make them more stable +Could you please handle it? CarstenYour patches for and are committed respectively at r and rAt least your effort to build separate patches have been useful in this peculiar case +Submitted fix is only a partial fix +forgeplugins directory and starting againLet me know if that worksThanks +I was not expecting this given receiver queue is synchronized +Looks like this is another special case of and +Life keeps getting in the way I have crafted two on this I would be interested in working through this with you +But for the time beingI assume that the debugger falsely works only on one branch thus if you dont get into that branch the debugger will not work +Yeah absolutely! I wasnt sure what to put for versions so I just picked something afterThanks for reviewing Ill go ahead and submit the patch now +Obtained usingsvn diff +It seems to me it was originally intendedto cache a small number of subquery result rows but the code didntseem to implement it that way +Im afraid this can potentially disrupt existing applications that rely on the current behavior +Pull request attached +Reopening issue so it can be affected can be changed to a different version +We dont use EJB timers anymore. +It is meant to expose the most commonly used properties only as you can see by looking at the other elements in the schema which have one or two attributes such I wont apply this patch +Got it +I realized that the above fix was not enough name mangling scheme of fieldInitializers doesnt work with separate compilation +The style for Hadoop is +Hi Sergey the pull request linked in my comment above addresses this issue +Ill also want to fix as part of this issue fish parses to should be a parse exception +Could you please try again? Thanks +This issue seems to imply that project references will always be used if version matches +In my TODO list +Currently testing precommit +Ive omitted autodetection of the exception class for the time being to be able to determine and cache the exception constructor on startup +the best strategy is to create a CLI script that deploys your application to all your server nodes +This test passes and Im running the rest of the suite right now +I m not sure if JON based tool is of any help here. +I totally agree that this should be supported for compatiblity reasons only +Submit patch give some javadoc warnings This patch remove that warnings +Indicates the device is ready on Galaxy Nexus rock JoeD +Ill close that one as a dup +Ive just committed this to trunk +Can anyone suggest how to determine whether we have a active session or not +I fixed them and committed was easy fixes +Updated title +Im not clear on the need for the typedComparator logic in +This patch is for problems occuring in HRGlobar HR SettingsPay Grades section + some reason the regionLocation isnt set in +Would you please help to try it +It does not handle desc indexes or desc ordery by properly and can incorrectly assume that a query ordered on a b c satisfies and order by c +Hi ViniciusYour proposal looks good details everything you intend to do and the proposed solutions looks good to me include what the community has previously discussed +Fixed. +Patch attached +Attaching since JIRA obfuscated the output +this was simply just a case of lacking any examples that added whitespaceAs for your other problem it seems like we also need a skip option? +The underlying causing exception was displayed within the stack trace +oops correctionIf the layout was called it would make sense to be applied to all views of that CONTROLLER +Changing the random test was a bit difficult for me +Is this good enough for committersPS thanks for the patience throughout this patch +and this is the diffpatch I meant to upload in the first place +Gentlemenare we any closer to a workable solution for multiple Castor configurations for Castor JDO? We are also using multiple Castor configuration files representing multiple data sources databases +A Saschas patch commited in trunk at r almost fixed this issue but in the case of lookups rendered from FTL you can see in the issue from FTL there is still an error there +Im going to push this to +Sorry wrong class +Just for documentation purposes this is the version we are now using +Tests pass +Heres a patch +Its not very good and is really only there to run examples and test code +We can go ahead as soon as I can get the webwork data +Hum +Also see exceptions which may be related to that on. +That was the site problem +Can you do one thing if you have no data in your cluster cleanup the namenode and data node directories +Mukund how many IPC threads did you use for the test? Default is +Included a test case +Are we following a new process here? Why create jiras for wiki edits? We dont follow the review process for it and wiki already have a change comment log +This avoids the null pointer exception +Let me know if you have any questions +Ill have a look at committing this change over the weekend +Committed revision +Create jar file with composite component in resources +Ok argument accepted +So Id suggest that rather than adding a buffer can just be lazy about reporting statistics +This issue was a false alarm +use version and struts to error in new version +will commit if tests pass +Rats trying to do too much at this one v +I think it should be I started by merging your patch here and then ripped things out until it was the minimal change to just refactor the code out of as much as possible +The patch is relative to the assemblies directoryIve also modified the startup and shutdown batch files to perform the same directory checks as geronimo and deploy for consistencyTODO It would be nice to have these scripts invoke a so they dont have to duplicate logic +Maybe we can assign an initial random value for current +marking critical since we will provide just a basic blip about SOA as long as we get some idea on what exactly +I am definitely a on exposing this unless we can understand the necessity of doing this +So hereby I am letting you know that there is interest +It doesnt fit with the way we allocate objects in a space and I dont see any compelling reason to constrain all super spaces this way +Im attaching a screen shot showing this bug +Preliminary patch merged with trunk +should probably be something likeIf there is no active context for the scope to which the bean declaring this observer method belongs then the observer method is not called +The first block needs to get fsynced to disk as wellAn alternative proposal is to provide a filesystem create API that takes a fsync option +I think your problem is related to that is resolved for the next release. +For loading bz Pig implements a separate and this was silently ignoring the concatenated data when the input to be loaded is a concatenated bz file +actual behavior +But now some of the partitions could be under replicated +The patch applies cleanly for me and builds without problem +Patch applied +Let me know if you have any issue with it +An example of an page using a highly customizeable grid layout +true Much bigger Allow pool to shrink faster false discard +Its time for opposition to be since Struts doesnt adhere to this anyway +This also affects the RPM maven plugin see +With this mail I join a summary of the modification and the code of the modified classes +Maybe there is an int vs +git +Absolute paths would still be stored in metadata table and used if the GC actually needs to delete something. +This patch can be ported to or trunk when contribs will appearThis patch is for Lucene branch only +Yes the sentence detector training behaves as described +A tweaked version of your patch is in svn trunk nowCould you please also provide an update for the docs +On trunk one additional test failsBut it fails before my patch was applied +If comments are supported in the editor then they can be used both to break up the lexing of the individual statements as well as form the name of the tabs in the Hibernate Query Result view +This leaves the test archive deployed +The script sent to the client on a dialog launch for a PPR is not getting written as a valid partial response +to both trunk and branch +Here is a patch that should address the above issuesOnly odd bit I made one method of static to avoid duplicating virtual table structure lookup code +Closing the issue now that it has been resolved. +I noticed that the estimated cardinality was being set for some parts of the plan but not others +I am in the middle of reviewing both of these patches and noticed that they are related and wont mix well +Oh my just saw this +So if this issue can be resolved soon it will be very helpfulRegards +I think I will start HICC document in wiki then move to xml document when API is ready for public +But Ive noticed that SOA has some problems with their web services which use AUTOACK and has timeout to ms +CONFIG INT Finally note that for the Cachepart regression test to pass a cache storage has to be allocated MB +I found that to be very cumbersome +Changing this to an improvement request for +this has been applied to the brand documentation +if no runtime team guys fix it I will try to create a patch for it +However the current doesnt really have equivalent token types to in order for to actually do something useful +although the Summary for this JIRA is now incorrect +Removed author +Out of date +I had a feeling this might be the case +Operators contains the desc the DAG and the transient variables for only +I hope that is acceptable +Since a node cant change names as soon as you acquire the lock on the parent no children can be added or removedI think youd have to acquire a lock on the child in order to guarantee that the stat of the child would be processed atomically given you want to include that in the response +New version of with amended copyright is now in trunk in separate module +One other place where bulk import can be expensive if we bulk import all into a single region it would likely incur a compactionsplit storm +im curious what does both versions mean? Isnt there just one platform +svn ciSending shellTransmitting file dataCommitted revision +Skipping Reviewboard as this is simple patch +Please examine the initParams attribute in the etc +Attached is the patch to fix this issue +Hi Soundararajan Do you mean you are trying to replace or extend it to support? I am not actively working on this JIRA now +The Version description says this Match settings and bugs in Lucenes releaseI guess we should at least try I think we can do it +I dont think Maven uses by default +This is of course unless you already know what classes are influenced by this bug +i dont have a largish project that is using spring integration using wickets ioc module so its hard for me to test and make sure its working correctly +I would be willing to apply patches to CVS to solve this +Why inner classes? Also maybe a unit test so that we can see how they are meant to be used +Thomas the test case is wrong +I cant think of reason to maintain order off the top of my head but I would be surprised if something did not break due to the changeIll open a separate issue to generally improve concurrency in moving it to use and +My project structure is identical to the one described in this issue +Gregory you are right it should be freed +Xpp is and you may succeed with decoding but I expect decoding errors using a parser +What about adding some property injection? This could be used to enable an easter egg +Merged up to trunk revision missing javadoc descriptions very basic may need further elaboration +We have a workarond for now +Heres a new patch which does a sort on filenames +Thats great thanks for the quick response + to trunk. +wsdl with just a types section +Looks good. +I should have checked that option +Moved to workaround applied +Trivial implementation +This adds a test case to Bretts patch and touches up the javadoc for +In order to verify this feature we have to test manually this feature +Moving to next release +Will has added support for escaping markup in references to the core +Attaching patch to clean up comments and javadoc +I was just joshin about the author weighting anyway I think the standard is that contributors are never included in the author tags. +Please confirm that it has been applied as expected +By AM Recovery I mean that when the AM dies +I have created to remove these exception messagesThere is however an issue that causes state transfer to sometimes take a lot longer than necessary +DhrubaMost of the times the test cases keep failing +The request would never reach the hidden I had to place the hidden before url and set as url pattern instead of mapping the filter to the spring Juergen +looks good to me +Fix committed to and trunk +adrian a data preparation is indeed a solution however i intent to not only use this for export invoices but also for parties products and possibly others +If there is anything I can do to helpplease let me knowBen +The change is intentional see +Running the test in a spinner might be the only way to make it happen enough to catch +We have seen some activity in accounting component in last few weeks +Applied patch into the trunk +Time to point this out in the FAQ I think + +I had missed tagging public interfaces +Mavens dependency resolution is complex and sometimes doesnt do what you think its doing. +Canceling patch +A token would only need to contain a base uri for the REST callIt should be an abstract service and I think we should for now push it to and then move it to when it arrives +I cant SELECT on a view and get data immediately in the same way that I would upon writing the same query +It ignores dictionary content that cannot be parsed and makes getString in more ve added Seans patch with version +Hi GeorgeThe fix looks good +HiCan you retry this on the latest code to see if the problem is still there? I tried a remote schema and didnt see any leaks using purify on Windows David A +jbellis gdusbabek for im thinking that we dont want to allow deleting ranges of rows +True no danger of being bitten at the moment Ill have a bit of time this weekend so will look into it at the end +Thanks a lot for fixing it +Fortunately seems to build a dependency graph when shutting down the components +Could you please provide details how to import into keystore +I am attaching first draft of two versions of patch +Merged to and committed revision . +Trying again +And by that I mean that CQL uses microseconds timestamp by default so it is in a way part of the API +For details look into the Specification in the table of the chapter pleaseIt also contains an fixe fordefaultNamespaceThis is needed for namespacing events and public render parameter without +Maybe interesstingerror goes away when the constraint on the name column is removed +Revision . +Closing. +Hi there +I just committed this. +One minor optimization is that you could use for shell commands +Please colin Thanks for taking a look at patch and sorry for confusing youYes +Good stuff +If this is going to be more of a core command the resolutionoptional could be dropped and we will just have to add josql as a prerequisite system bundle in +Thanks for reviewing +Fixed findbugs warnings +It should not rely on MACHINEARCH environment variable which is available on linux +The get patch had a sideeffect that getHints started returning IP addresses instead of hostnames +I dont understand why we cant just commit a quick fix? minute is a way over estimation of the effort to fix it +The issue can be the bug +This is actually a bug with itself as the plugin gets the version out of the file +Oops +The buffer will always be initialized in only +Ill look at this when I get a few minutes +By the way the Sun JDKs version of Xalan does not appear to suffer from this issue so the version in might be used as inspiration +Hmm true the prebuilt distribution works ok +Thanks. +Agree with all your points +One small optimization to have only one update if the checkout is available +But we will support it in for DOM L load and save +Just ran a test outside of Jetspeed on another Spring OJB project hereResult transactional commits and rollbacks work as expectedSo Im left wondering what weve broken here as I was sure transactions were operational once upon a timeIm summing up the differences Datasource other project Jetspeed other project Jetspeed other project Jetspeed All other configuration files seem to be the same +but I have no sample at hand with generated classes +patch looks great one question should we keep the reset can return false and we do not reusethis seems like it might be obselete though it would introduce an API break I think maybe we should change it to void +There are some bugs in the patch of st version fixed in nd have refined the patch would please try it? Thank you +Dont we already do this with HTTP transports? HTTPS transports are always configured in the and services get HTTP endpoints like httphostportservicePath +which usage is pervasive in the scripts of Hadoop ecosystem projects +Having said that all synchronized blocks are useless in the distributed environment but until we address the we should keep them where they are +This however should be applied to Branch +Thanks Alexandre. +punted to from as invalid its old and stale +Attached Patch +Use socket to get the socket transport +Confirmed +stackin our clusterit does work +Thanks Knut for fixing this up +The fix has introduced a regression the own installer crashes with a NPE +I cleaned out my data directory and ran again log attachedThe st warning I get is this one +OK thank youi will close this jiar +bulk defer to move away from Regionregionarray I believe the commondesired loop idioms now all generate proper code +I will look into why mvn tomcatrun doesnt work and mvn jetty and comment here on my findings but in the interim please just use the latter command. +For forcing could cause +Thanks Jacques I appreciate your helpRegards forgot to mention in the last comment another reason to address this ASAP is that once something gets into the and is used it is difficult and annoying to remove +The order is the order in which the various files appear in the of the HTML Pages versions of the manuals +Committed r to trunk +Thanks! Ill try it tomorrow +Hey guys I was just wondering if there is a way to get the suggestions not to echo the query if there are no suggestions available +Please give the next nightly snapshot a try and check whether it works for you it works +Ive updated the parsers to use the new metadata entries in r. +please ignore that change +So the same thing happened at this time +sampled +It looks like something we can fix but I dont know where to start +Looks good thanks Tim +verified in CR +Adding tests to the api and projects should be evaluated to make sure that the code paths are all exercised +Admittedly it was never properly defined whether objects would be candidates for JMX exposure in the first place +Shortened cygcheck s output to know the cygwin versions +logI dont think I need to do anything specific here +Attach a trunk patch +If you use overlays wont the SCM metadata filesdirectories be corrupted when the overlays are applied +Any chance you could run through it in a runtime workbench and find which calls take the longest +Thanks Sandy! +Please test +Glen please use a patch file to post the patch instead of description +I agree +Bulk close for +The mechanism isnt designed to handle all possible scenarios +If you need any help with getting files out of Justins home directory let us know +ill trry to reduce the date label width but i think that the day timeunit must be keep as be more readable change the timeUnit in the dashboard config fileChristophe +I really think that it is a good idea to be able to save namespace without entering safemode +Ill cook up a new patch after my commute home +I feel very sorry that I cant test this on windows +I want to bring this to core tooI understand! Lets get this into contrib with either one name +Two things here Eviction manager should wait for eviction to be fully finished before moving on to stopping the other components and secondly could not be responding to interruption properly? Ill add more logging to confirm what Im saying and then need to see where exactly interruption might be swallowed and that it doesnt allow the task to be cancelled properlyThis is definitely a timing issue and its taking me a while to replicate it each time +Ive attached a Wireshark capture of where I believe the failure occurs +v +There are places where we use add as a true add not set +with Owen and Arun and we dont want to make this API public +Its really broken that things like grouping and faceting are coded to a separate API +java file in trunk youll only find it in the clustering contrib module +That being said I dont see much reason to have most of the other methods in there +is the related JIRA I said I would file in my last comment +I set the Analyzer used by to null and tests pass +Fixed as of again for that unnecessary noiseI just messed up my tabs and attached the patch to the wrong issue. +It will not be back ported to unless we have a release which is currently not planned +Sorry a mistake on my side The NPE remains even with the proposed change +Changed to handle and return from the read patch for EOF looks fine +The JCR jar was recently dropped from the war because ofThe recommended approach is to make available in the shared classpath of your servlet container so that you can make a single repository available to multiple webapps +There are some related issues that I will be fixing +Fixed in aa +Thank you Nikita for pointing this out +I didnt cycle through and stop start each node +Documentation updated. +In in the spans package there is also a that mightneed to be closedThis means that closing would have to be done via Spans in the whole spans packageBut as long as there is no real a memory leak what is the point ofadding this close functionality +Test failures are unrelated +The slow query I mentioned in description went from ms to ms! Thanks again +Patch that cleanly applies to trunk +A general fix would be a bigger project. +Because the classes using this modules classes might be outside the module itself and if the afferent coupling is high this package must be the most +Actually never mind I did this myself +OlegYou can try nightly build when this issue will be fixed +Youre absolutely right +It also has test to check support for compiling Java statement in an aspect +Updated files +Could you please let me know how you go reproducing it with +Committed to trunk rev +Julien propose the idea of pakage templates that would be the bare minimum to build on still need the default portal as people built on top +We can always jack up sleepforretries +Do you by any chance have an AD server with the relevant test data that we can connect to and debug from our site +Can you open a bug andor submit a patch +oh and perhaps its not even needed in that case cos rows are already sortedOK cool +Also developers are more than welcome to raise chores directly if they if I was wrong but your problem looks to be almost definitely with the use of http to commit. +Rossen Please can you verify this is true The above would also work before Spring as long as entityId is in the model +Reopening to set proper resolution. +It would help if Put and Delete had an ancestor or even if it was just an interface such as Row with a getRow method in it or +Asks hadoop for version info instead of looking for version info in library file names +No the does not +Thanks a lotRichard +Here is a new patch that fixes the problem the original fixed and another problem in this file +Verified on my EcplipseCommitted to trunk +This ZIP is to replace the previous archive which was missing the archive +Username zds +Wating for next release +Split this into and to track DAG API changes and Runtime changes separately +should not break at all +I have created website pages for the two mentioned links and attached it to JIRA issue +This can also cause problems on token changes since the status event will fire with the old tokens and never update tMD with the new onesPatch to make two loops instead one to set states and the other to handle notifications +Maybe we should just show agenda group activation groups as attributes and allow sort by andor group by in the view +Sheesh Im having a dumb day +No chance this will ever be fixed the focus for galleries is on the websites providing support for all issues that have been in resolved state for more than one month without further comments to closed status +Could we move these wrapper classes to or somewhere else that already has a dependency for other purposes +This is the Rule using the nested clause without the parent +Patch v moves BLOOMFILTERDATAKEY to +Not a fan of the API +I also noticed that if I create it locally then copy it to the network drive it still has the same issue + discusses some simliar problems and includes a workaround patch for this issue +Error message in logs is the only problem I see code is executed ok +Thanks for fixing it +Attached includes an initial version of with unit tests +Can we duplicate this with? +were added +We also need to get committed it generates very realistic graph data for Giraph that possesses many of the properties of a social graph +The original code kept a reference to the roots and that is why we failed to report acceptNode for elements that had the same name was the rootThis patch changes this behaviour so that we hold a reference to the root Node itself +Closing this as the problem is reported as fixed in websphere. +Weird the password did not change as far as I know +Let me know if this is enough +Hmmm rigth I forgot about this one +Simple is to reproduce +Note there are likely multiple connections in the client some of which are guaranteed not to have been active when the client went away +Hmm +is now able to deal with a literal the style configurator will need to be checked over next +I will post it soon +You give me too much credit I have no idea at all if those keytool and jarsigner commands are correctYou say it only gives that error sometimes? Is there any pattern to it +Committed. +believe that base and really critical functionality +The attached is a modified version of the composite key example to illustrate the problem +Now that work is underway with AS all previous community releases are +Closing all resolvedfixed issues of already released versions of Roller. +You configure this using the new fulldroppolicy attribute on the queue element. +I know Ive read with the Java API is possible +Havent resolved those issues yet + this an issue with the wsdl pointing to http? ive just successfully made SOAP calls after encountering the same and then changing the URL to https instead +I have opened new tickets for extra functionality to be added for the or soon after composite primary key support dynamic finder support lazy data loading will be updated to GA when new Roo test script can be used to test the +i found a pure codec approach for providing all the functionality i require here and more requiring no patchesif any committer has interest in pushing this ticket forward i can clean up patchadd suggestions etc otherwise this ticket can be is the codec approach you used generalizable such that we can benefit from it? Do you wanna open another issue for that? +Sorry I missed updating this +issues should get JIRA issuesModeration email can be karpet mailing lists have been migrated to their TLP addressesWe also have the old addresses forwarding to the new ones with a return bounce informing the sender of the change +Merged to and committed revision +Since ads is quite a common pattern maybe events can be used instead of ads +Separate items +This is proposal for VM property module redevelopment +Please free to submit the patch +Hi Sean I have already got a patch for this issue and will apply that soonThanks +So it is possible to have two containers in a project on the same +If user do not use then we fall back to the deserialize version +Fix looks fine to me +some unrelated stuff got mixed into the last patch +Commit the patch to trunk at thanks for the patch Shawn +The cookie save or retrieval is somehow brokenThe code i was using before is no longer working after an upgrade to a never wicket version +Please check if modification is correct for using tree without custom component. +Confirmed that this issue does not exist in +Looks greatOne question +I am not against changing our mock setup but i would discourage any further work like what is going on in WMS right until there is a general consensus among the dev team on what the best way forward is +So closing the bug. +When you scroll down on any browser and launch the panel it will be shifted +Considering snapshot is the bright spot for this issue is enough to sink the RC +That what you do anyway but it could be done directly in the updater the manager should take care of such event +Some tweak is necessary in configureSpecializations +So probably not worth tryingIMO the best option is probably the original clrus fixed versionIll dust off my code so we can do some benchmarking +The patch for trial and the resultAdditional information +Can you please investigate if this is a simple fix? otherwise well just document the limitation with the incorrect endpoint +BTW there is no information about reversed type on datastaxs comprehensive documentation or on +Older one is obsolete +patch for branches which includes original patch for trunk plus fixes in Rev +Since we dont want the framework to modify parameters this has to go into the contextIm leaving this issue open until youve reviewed the changescheers Roland +bq + work on hold through at least August unless someone else wants to pick up the work +If there are no objections Ill commit this shortly +The github repo has been deleted +requalified the issue a Task +should it then not update all the other config files in the dir like etc +Bulk closing stale resolved issues +Is it not possible to just load a native library off of the java classpath +Final patch for review where I have added a new property to the XML code generator configuration +I just committed this. +Can we get this into? +Fixed in revision The test was modified to follow conventionsVladimir please check if it resolves the Mikhailfix applied as by Vladimir +Hopefully putting them in conf will resolve it. +Closing all issues +eps formatThanks Chris +Technically from this resolvers perspective returning a resolved host capability is invalid +To name a few +Ashutosh Is this patch ready for testing +minor correction First a datanode checks its block data with CRCI meant +here are the output formats for the resource manager and the node manager +The spec is measuring acceleration in gs +aries trunk diff forM blueprintsrcmainjavaorgapacheariesblueprintcontainer +Because of these three Im marking it WONTFIXStill I think the issue of external invalidation of cached items is a need in Cocoon and would love to see the problem solved +Can you create a phabricator entry for review +Fixed in and above +. +App committed in rev +It turns out that this is covered in the docs we just havent done a release since the feature was implemented. +This can be two fields today but going forward you may need more fields and this would need to be constantly maintained +First a dumpfile for the openejb repository +Alternately we can create the with a new as the parameter +Here the patch for the groovy scriptcool looks good thanks +Maybe we could have a default upper bound of something like bytes and then truncate with +Thanks for the offer for help +Maybe you can add some logging to lookup process to see where its going wrong +CarlosTo let you sync from our repo do we need rsync or can we just provide an ssh accessBased on the available scripts we couldnt determine if ssh alone was +In general this looks good but why two methods +Still TBD how to deal with infinite loop of aliasing +We need something a like You may look at which as a set of JMX annotations as well +Ive experienced occasional clashes between Spring OXM and OXM in particular when using the Spring OXM with MVC and the appserver happens to pick the OXM classes in preference + will help with this and it will somehow related to is a sneak peak. +Much of the discussion under is about the code that will eventually be part of this issue +see lets existing for recording a +initrd tmp +Go ahead and load it into the main repoThanksT + with the apped eclipse applied thanks +How about this +Failure was just because the random IWC was causing too many segmentsAh damned what a miss Thanks mikesimon +Tom +for +Its possible now to run multiple datanodestasktrackers on the same node simply by writing separate configuration files for each +Niktia why is that you want this? Why do you consider current functionality in this area incorrectRegardless Im not sure this is really feasible +Updated doc +If anyone can think of a better option place let me know +Verifying +showed up today in a form well add it back to the blocker list is it becomes a real problem +ok +Hi Any progress on this issue? I found sync call was removed from the source code +Thanks Xiaoshuang Wang! +Any reason we want to upgrade right now? Id say we can wait until +push out to +With all the changes in this and a few related classes I think this problem is no longer reproducible +In this patch we keep that but for findbugs we now only run findbugs for projects that have changed so the acceptable number of findbugs warnings could be different for different patches +I suspect you have a large maxOccurs value in your schema +i agree this should not be added directly to the zone component +yes exactly module its dependencies allows it +I also tried compiling without and it compiles fine +The path where these XML files should be created will be stored in a that way the data will be available for other coverage tools such as EMMA and CLOVER to pick up and aggregate to the overall coverage +Im setting this jira as resolved +Nicolas any plans for doing this soon +Actually Im not quite clear about how to easily maintain GLB it would be good if you could explain more about this I mean the detailed algorithm to maintain this value +I created a Roo project in STS added an entity deleted the project from the workspace physically moved the project directory to another location and then imported it as a Maven project +Committed to and trunk +thanks Ivan. +I dont see the new version of the patch here +as dabcafcbceeffacf +Unique Diff of the Tree class +This looks like a useful change +This is an artifact of or something dumb we did when hooking up JAXB +Is this an improvement or a bug? The summary sounds kind of buggishIll try to restrain myself but I think we should have fixed unicode support in lucene because then it matches the unicode version of java the problem is we could not do any of this in because you need java to actually implement most of the support so it was a chicken and egg problemimho its all bugs but ill list these issues as improvements +globStatus try each component using get status without suffix which leads to the exception thrownI think this behavior is really confusing in many ways +Pretty trivial patch +I think this is the best we can do +These are the only changes +I guess this should be renamed if it is still planed to change the error message error thrown or else it should be closed. +Please close this JIRA. +I checked the Win and we set max Java memory to m by default +When the enumerator is at this position subsequent calls to also return false +although was not designed to be a general database or large object store +What does exist however is srcorgjbosstoolssmooksconfigurationeditorscsv +Agreed though the awesome unix support will be great for folks using linuxsolarisos x. +Are these the specific concerns for this docAny other items +Thoughts +It also slightly improves logging one can now see how many docs were added or removedAny comments +thanks for the very quick turn around! +Gary shouting needs to be fixed before GA when you say you got it to work doesnt really help me knowing the priorityfix Put on unless someone tells me why this is release blocker with a reason +patch applied with modification to get it working with at least +I hate the messages from review boardThis is a completely incompatible change +The barrier fix in the test looks more promising +Whats the rationale of? Is this being keptaround for backwards compatibility? After you commit this change Im going to move all the tests I can find out ofthe srcmain tree into the new srctestsorg +Scot Joydeep +It simply creates an using the system id +AFAIK the provided scope equals compile scope its expected to be part of the environment at runtime so for example war files wont include provided dependencies but only the compile ones +Let me know what I need to get this into PIG +branches fixed in rev in rev . +Mark ProctorHere is another one with The removed less text the better I think +Non Portal issue its resolution does not depend on a Portal version but on the application server. +So it looks like indeed a test issue and not an actual dual boot +That random stuff looks dodgy but I was going to leave that for another dayHmm +So without the binary Cloudstack will not be able to spawn system +Needs test shouldnt be a problem post can you review this form me +Closing all resolved tickets from or older +not sure +It currently just discards the jars due to the that occursAlso it has facilities for deferring to a protocol different than http so that codebase downloads could be vhttphttpsor vhttphttpmdand other combinations so that it would be possible for various URL strategies to be managed to some extent +Test with Chrome also +Unfortunately I have not had the time to reproduce this in a simple unit test outside of our environment +This is a new issue +I meant the last sentence as an additional fix itemThe main issue is that the doc says that one has to use the internal ID but AFAICT thats not necessary or even true anymoreThe documentation needs to be updated to make this clear +Sorry just changed my Eclipse settings +Am I correct on thatI made all my changes on +The difference between toOne and toMany is blurred to the point of being potentially confusing +Ill open follow up issue to remove unused files from repository. +But is is not persistent across restartsI meant to say with this patch +Actually it adds hours because it must observe DSTSimilarly dateFields value should beZ but just like date it subtracts an hour to keep DST into accountHope this could help you +Hi again +Updated patch to include setting of SOAP Headers and well as updating how they are retrieved +I just upload the polished patch hopefully it address the concerns +Applied to SVN trunk as rev and for xwork +I suppose that through configuration we learn what needs to be loaded and the bookie initialization loads the appropriate modules +Marhsall yes I also tried something like this without success but I do not yet know a solution. +Patch applied +The same with and TeeSink filertokenizer +Patch to address the issue +Similarly the url for search should be DomainNamesearchkeyword so these urls can be configured +Thanks Rick +If you insist +Topics might not break it because you arent sending them as persistent messages or you have no durable consumers +This enhancement will be considered for a release after Struts +geotools seems to be using with very similar packages to jscience +Just a misconfiguration +Thanks for the explanation Leonardo +to commit + Will not contain new look and feel +Just add the convenience methods toAlso it would be nice if we create some Maps during construction so the convenience methods dont need to do so much looping +The tests from the patch were useful thanks. +It means maintaining a per topic counter on the client side +and root file had to be manually merged because of previous merges of this branch with the trunk +I cant reproduce thisIf you are using solrexample perhaps try removing the contents of the work directory +Im going to try to update a bunch of these now +The cygpath code for AUXPARAMS needs to go at all the places where aux params are getting created from file names as opposed to at the end +The problem is that GF also uses SLFJ but different versionTry to deploy the +If so than shall we remove the earlier framework at the same time we adopt this? Having two frameworks doing same thing will be confusing +Patch for +I went into and removed from the search +Jira titles should be descriptive of the issue in a short sentence +Im actualy at but that looks almost exactly the same asI did take a quick look at and would really rather use the Posix file manager code +comment questionComment Forge should ask the user if they would like to install before going ahead +Thanks Jon! +lists should be done too. +fink or macports Ive always gone with just compiling from Build Servers use macports could someone please describe the procedure to install Thrift dependencies with macports? +Regression on the regression. +The problem was that didnt even acknowledge the possibility of a mapping +Its possible we can get this help from Oracle if not Anil and team have offered to help out +I can confirm that the error is no longer present +Anyone out there +It no only happened it happened s not expose customers to a broken EE API +Please reopen with more info. +jar in the lib directory is loaded +Personally Id prefer to just work on one version at a time +Carl most of the development in hive so far has been done in a iterative mannerThere are so many examples of features that have been checked in multiple patches indexes views to name a fewThis patch is not breaking anything existing but is not ready for final consumption +However the orphans will still not be removed from the DB during a save +Closing! +Added support to upload war files from the GUI in Rev +Hmm +Can you verify that your myid file contains a small numberConfirmed! Namenode is data nodes are and +Clearly is someone is running cygwin they are running windows and we have perfectly valid +Can we attach a small test case that +I wonder why there are all the other null checks though +patches the poms to use spring instead of +can you try a current release or +Thanks for the quick update Marjus closing this issue +I have a few minor changes to make and I plan on trying to get the new beta out this week +Attaching a patch to do both of these +Especially if the thread performs processing on data that should only be done once +Were you able to find a solution to this problemThanksAparna +This is not the correct way to make instances so this bug will be rarely seen +After we branch for then ALL work and builds will be done in the stablebranch jobs and trunk will become JBT JBDS +I committed the patch to branch +I suppose I could commit this fix but since Im not the author I wanted to provide the patch and make sure it doesnt have any that I havent +now ran ok on phoneME and +I spoke too soon it looks like this issue is still biting us on CI +Should I create a new issue for this and attach separately? Please advise Im a little new at this +Oops! and are added +Taking a look +As the given binding file fragment demonstrates how to achieve the desired functionality I am closing this as Wont fix +only ony JAR and remove all other ones from the classpath +Jim please also check the same issue upstream +Items contained in a batch from the server are set to StatusMODIFIED without establishing whether they differ form their cached values +Ill build a tonight +for the patch +nope +Patch refactored WriterAction to be able to supply custom data generation +BTW I filed we continue the discussion on +Thanks +see comment there a work around for this without patching the Spring code +Ill change the fix versionWhen we DO get to this well hit the question of how exactly will be able to commit to svn +would like to add an profile for easy debug purpose so that anyone who want to contribute more testing could easily debug testing without startstop server frequently +Hi instead of throwing an exception wouldnt it be nice to support child axis in order by clauseGood idea! However lets leave this issue as the bug report for not throwing the appropriate exception and you can file a separate feature request issue for actually supporting that feature. +This patch tries to reuse the include guard if the output file already exists. +Throwing into the bucket since thats where all the dynamic runtime control of bits will be implemented. +Can we have the MS logs for all these failures? Without logs it is very difficult to debug these failures +I think these values came from clientrepo because seems get that value from a parameter +This is a untested implement of how a users groups could be used to select a fair scheduler queue +That way we dont declare yet another conf object the local should be used +This was fixed in a wiki page so not directly tied to a release. +I agree entirely +Preliminary patch for review +We tried to set the threshold higher but it cannot be too high because it hurts the memory utilizationYes we will need to make +In the attached screenshot Clicking on any file name should open a resource viewer on this file +Im surprised that you are so confident that this is notCan you check your on one of the environments I mentioned? is the one I have tested against the most +I was thinking supports applies for aggregates too. +Thanks +for this proposal +Yes Ill commit it this afternoon +There are certain cases where the param contains single quotes and double quotes +Thanks for the patch Sergey +It seems to be OK in +MattThanks for sharing the codeDoes the patch support Kerberos credential delegation +Thanks this an incompatible change +for trunkThanks Gabriel +Its much easier to deal with context and symantics when you have forward and reverse state +does not process lists of addresses correctly +your store the bodies on the filesystem cache +uses map to map its row +Question looking at I need to add some examples etc for these components to make it easier for users +Or maybe I dont understand how works +Once I get fedora I will have more info +Does it works with ms +Added standaloneEnabled flag to +So if the plugin ignores the JDKs that are registered in Preferences Java Installed why did you tell me that I should check these settingsI was already wondering where you read these in the code because I couldnt find see how this can be fixed I need to have a look at the maven embedder project +Change spec to say the x is only imported for +I am not certain which version of GANT is being used there I would be using the default GANT that is shipped with grails think the problem here then is that Grails is using Gant which probably has these problems +Sounds like an interesting feature if anyone is looking for a fruit to contribute +Here is hothing to is a way to hide Welcome Screen on open perspective operation in case we need to open the perspective that is already active +In other projects I frequently have integrationacceptance tests which extends from but which I dont want to execute with the unit tests +Will certainly reopen if needed +Better to change REGEXEXTRACTALL to make it symmetric +The websphere container implementation +See +Set context class loader for common am trying to reproduce the bug but with no successIve build a new plugin with a fix to set the before using httclientCould you test it please +Patch looks good resubmitting to Jenkins just to be sure been a while the last time it was blessed +The data structures threads and blockingQueues may not be thread safe +Considering the visibility expression part of the KV was the cause of the data inflation discussed in Himanshus proposal onbq + +minor changes mostly fixesUpdated patch +This has been completed as part of and when that merges back into trunk this will be done. +Its something Ill look into while settingup the releaseIvan +Thanks Braden +Im working through the setValue method again to see if I can figure it out a bit better I dont understand why the first if tests true if field is not a collection OR there is no add method +Using search for that can be processed by something elseI think it would be great if that sub selection could come out ranked though I think that is also valuable for and while the other system could somehow rank it would have to dupe the lucene logic to do it as well +Even with a the project is failing with the same error +What JDK did you use +Verified that using property foo the server name can be changed +If it doesnt there will be a pretty rapid follow up release that does include it +What is this task about? code refactor is not very informative +Updated the example in the file +Just want to add that Derby was +Minor update to merge with some recent changesPatch looks good +Should belong toI think still belongs below as its part of the image on disk in a sense and their lifecycle is tightly relatedMoving up into does make a certain amount of sense to me but I didnt want to bite off too much in one patchDo you think the changes in the attached patch are good even if theyre not a complete cleanup of everything? If so lets commit this and then do some more work in another JIRA just so each small improvement is easy to review +FWIW I confirm that the XSD in is now valid +Closed after release. +This patch fails when running the changelog command in standalone mode +fine! all right nowI use true and is oktoo many thanks +Also if there are any YARN only changes that can fix we should get them in here in the addendum patch +do you have any in your gwt module declaration +We should probably still consider this a blocker issue if there could be unforeseen side effects from the quick fix applied yesterday +I would also like to see a higher priority + +Thats great newsJust a thought on your proposed solution +Details in sounds as a good solution +Adding concurrency to the JITGC enumeration interface introduces a bunch of new synchronization issues +Anyway there seems to be another issue has +Bulk move issues to and +With further changes under r and r the ant scripts can now be used to build and run the complete travel sample and they can also be used to run the jars produced by the maven build +Attaching documents is a much better option +About yes I needed the full stacktrace to diagnose classloading related problemsAnd about the class name pattern matching Id prefer doing it programmatically because they will be most probably static settings +Okay thats more in line with what I was originally thinking +there any updates planed fixes to this issue +The patch is tested with simple c program which loads a shared library +Cool +Was fixed for +There isnt any error log because this is a normal behavior +Found and fixed problem with JNDI references in caused by the first patch +Issue closed with release. +Thanks Tim +Im not saying Hibernate should support this Im just throwing that out there +ill have the fix shortly +Hi KimThat sounds like a good starting point +It appears that my logic of determining whether we are inside UPDATEDELETE etc statement is not as as I thought +Well lets discuss this over at the committers list +bq +thats because of the user module that should be merged into +Recompiled mvn clean site install assemblysingle Prelease agree that looks like a setup problem rather than an problem +but meanwhile the other part is to pass down the replication number into the writer. +Theres also the issue that you may not want to expose the internal to the remote users any more than you want the IPC ports to be visible +In most cases this was ok loglevel would be and the log function pointer NULL but as I could see using the core file loglevel was and the log function pointer was an invalid address which caused the segfault +So no problem. +It is not easy to create new unit tests +I tested this on S with the tests from and it passes +Yes means global limit not the limit for a key +As I was looking at the logo I felt the gradient was too distracting from the form of the cathedral so I removed it +I see no reason not to add this date format to our growing list +There are still some post streaming actions like building indexes that wouldnt be included with that information but that might not be a huge deal +Im on board with not harding code defaults +I will try to improve their speed later +Yes +Ive updated the patch and created a more generic mechanism for registering more examples later +same patch try to trunk +Olivier we dont do authorization here anymore I need a list of JIRA usernames from this JIRA instance that will be administrators for the new project +Thanks Zheng +Thanks Doug +bulk defer of issues to +Gone through the patch again nothing more to cleanup +BULK EDIT These issues are open to be picked up +If you run a command line forrest you would find this error at the end of the processLogging Error Writing event to closed stream. +This Jira should be either rejected or amended +Patch to better implement the nesting of modal windows +Default sleeping time is seconds but because before going to sleep thread is acquiring the journal lock it blocks all other threads from doing update operations for seconds despite of the node they are updating +UG +You can also disregard the trunk patch on this issue if not attempting to preserve the old stuff +If the provided order id is unique than it will create the PO with the provided orderid +Committing shortly +That is still applicable as far as I can tell +Should I make the change in get functions in Utils +Committed to trunk +sounds worse than having to make this workWhich plugins has this problem +Look like this issue doesnt appear on Tools nightly build with Eclipse M +I have asked the JDBC expert group advice on this one +Close all resolved issues for older releases. +No longer an issue for jBPM stuff since they now have their own variable resolver we still need to solve this ASAP + +Ive encountered some issues trying to apply your patch +Does it work with Java on Tiger? This works with on Java on SL +Rearranged the constructors and providers for orthographic and stereographic projection on trunk as of revision +bq +Looks like this is a duplicate of which has been fixed +Or if in some cases the matching is guaranteed to be correct due to how the map is used +Triaged July Assigned normal urgency +has test for proxy +Tagging meaning commit to trunk for now and backport after +Fixed kerning works in the current transition resolvedfixed bugs to closedfixed +It seems to work nicely nowSending srcmainjavaorgapacheservicemixhttpendpointsSending srcmainjavaorgapacheservicemixhttpendpointsSending srctestjavaorgapacheservicemixhttpTransmitting file dataCommitted revision . +Please provide you fo document as attachment not as plain text within adescription +Patch for uploaded +m not following what was wrong with the perf tools or what changed to make them unreliable +thanks for the adding documentation Ivan +This issue appears to be related to on the remote repository containing a metadataversioningsnapshottimestamp node +Transitioning to waiting for main issue here is transferring the domain names from Oracle to ASF +This problem affects other as well +Anyway I think the deactivation chunk mode will be way to go to get interoperability +Perhaps this issue should be rejected +Uh +Fix thank you +The general approach to the fix would be to make the script parsing logic from the odbc client generic and use it from the JDBC client to issue individual queries +The future effort will be on and thats the best place for this patch to be in. + +Unless anybody objects Ill submit this relatively soon +Attaching repro +I dont know that we should replace the JIRA entry in the left navbar +Patch for trunk version of issue type to be Improvement which seems more suitable here +Still I dont think well fix it +The buffer to scan content is also used to scan comments and a few other structures +This should be same as octal umask +SPRINGCLEANING +Maybe the second one offer the possibilityto be changed to include only that xml parts necessary for reading the mappinglater opinions +Quick fixes are not often the best quicker fix would be for you to postGoal the jar goal and add any test classes into the jar +Best regards by George +Thanks for trying that +So the cluster is for data storage and access only + +Ill take whatever makes sense +now it works +What is a best time to catch you on IRC +Perfect thank you +The cycle continued and I couldnt find out why +extracting values from result set into sqlinfo +Thanks for explaining +Should we add that to all requests sent by the controller such as +Anyt ideas +While the folder target has now been created there is no folder under it +The spreading of dependency definitions in sub projectsmodules are unmanageable and requires much more time to customize the properselected dependencyI would say that it is not only productization requirement to have proper dependency management but really project requirement to have controlled dependency set IMHO +Trivial patch to fix issue +Mistaken problem +See screen shots for in +Ups I didnt notice it +bq +Can you point me to the testcase where this is happening? +Committed to and trunk +If theres a problem with string parsingescapingetc in the shell it should be fixed there not in the Java API that can deal with bytesIn fact I think we should revert this from and +Thanks +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +This issue is still not resolved +I ended up using Joes suggestion and changed the target to and the min to so I could add hardwareAccelerated and it works perfect +Is there any chance this patch could be incorporated? It seems to solve all the issues with a nice design and has been around for years +Please experiment with that patch to see if it meets your needs and does not have any problems when using your regular expressions. +But in that case the probably knows that it can ignore it +My gut feeling is that the parallel upload feature adds a complexity where its absolutely not clear it will be useful or desirable at all +Updated to latest Groovy applied thanks Justin +I plan to finish this either tonight or tomorrow nightI will put up the patch for review +Attached is Sergios patch for patch for latest branch +I fixed it by removing the error marker on the classpath container it will be now set on the parent project +for the patch +If you cannot reproduce it then Ill get around to getting you this trace later on today +I dont think its an lucky approach to have it in two categories +We lose creena +Well done the most straightforward patch Ive seen in a while +I think this was fixed a while did not commit anything? How is it fixed +HiCan you run svn diff command and apply the patch hereThanks is updated scamodules directory in my the patch available checkbox +One question do you have a proxy base url set on the server +is the current batching model so it should be correct +We should add some more out of the box type converters for Nettys so you can convert to XML etc +bq +Again this is for code reuse purposesAll Java tests pass and the patch applies cleanly for me against trunk +Good point Ive updated the reference documentation accordingly +looks good to Scott for reviewing thisI checked it manually +I cant find it in the spec now though +can you please submit a tiny test casethanks of the changes as suggested by Nishant to resolve the problemAshutosh +Test cases fix +Moving out of +switching to is considered for future versions but will have significant impact and is not an appropriate solution to this specific issue + +Java file for test +If works for you but not it seems to be more an issue of youreditor +The javadoc tool did not generate any warning messages +Changes from previous patch Added a representative set of observability to and +The initial fix committed by Paul in was there when was released if Im not mistaken hence why Ive set the fix for and +There isnt a way to show this in JIRA but we should fix previously released versions of EAP as well as the two versions listed in the fix version of this JIRA +I still havent been able to find any information that says the custom classloader in this case is doing anything wrong +Yes its normal +Now wrapped all lines to charsThanks for the the last nit as wellremoved the duplicate code in places and moved to a method +In second way I improve the model with Bruno suggest that a user can by set his default choiceFor I understand the idea but as entity Is better to use aI resynchronised addon with Jira +And I guess thats also the reason for? Since the initial replication request is ignored the replication on D can only be done after the pending replication timeout +I would love to see this issue fixed so I dont have to change the template every time I recompile the source code +You could try to insert by hand a reference to your custom fonts into the fo file generated by maven and just run pdfpdf you should get the fonts embedded will close this issue now and open a separate one with more specific explanations of what has to be done. +I suggest to close the issueAnton. + +The problem is similar to +pluginsextensions folder +I have also tried with immediatefalse but this makes no difference +Thanks for the comment DavidI have been attending to other business for the past few daysI recently added support for +Working on it +Version Released +You are asked which connection you want to use to follow a specific referral URL this way you have full control regarding encryption and authentication options when following referrals +Please give initial feedback on the jobhistory storage approach +Kevan pulled it into so updating Fix versions to txmanager branch at Rev Once I have the snapshot artifacts deployed Ill update the server release to use them +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Shervin I have rewrite all examples which use the jboss cache and yes the blog example uses scache so it works now. +Closing all resolved tickets from or older +Doneroot cron backs up database at UTCbackup cron backs up the backups dir offsite at UTC +that worked netstat a n grep LISTEN grep tcp LISTENtcp LISTEN +The changed I made in the +I think these regressions came with the introduction of read barriers +This documents in the release notes some additional bug fixes which will be swept up in the impending +test case demonstrating the case demonstrating the patch yes that new bug I found was actually issue we previously discussedRegarding my comments I only asked to make sure it wasnt against some Wicket policy to include themThanks again +Nothing more to do here +i tried windows bit bit linux bit all with iterations this test never failsweird thing I was just gonna dig deeper but you fixed it already thanks mike robert +We have issues with EL performance but I dont think this fix here should be the issue. +I just think we need to tease apart any tangled requirements +Kay That would be great +Now that work is underway with AS all previous community releases are +Fixed on trunk. +And after creating the feature source the next set of steps are same as of drawing a map from shapefile +My ids are and the break happens around rows +It might make sense to create a replication benchmark so that the performance gain using the new algorithm can be measured +New constructor for +Alternately could the hardcoding just be removed? I would go with either but still not sure why its a requirement +See +Attached the patch with my initial idea in implementation hereReady for approach review +I dont see why finding an optimal heartbeat time is critical +I fixed a related issue recently +Reopening to label as not appropriate for backport to +Adrian says it is since it shouldnt be a part of the key +This should already be fixed in RC snapshots since we removed all declarations right away +Patch submitted +committed thanks! +Does this fix also resolves the problem with web servicesNo +Well I suppose I could try harder at least for things like constants. +Same in by me +The was something I made to install a mojo into the local repo because I needed it promptly +The cache should never be unable to load the column vector since it should have generated a column vector for all +Resolved pending feedback +Thanks +It was due to an erroneous assumption that connection abort would not automatically release the connection back to the manager +about the package? It already contains and +is required to copy from secondary storage SR to primary storage SR +in that case oops! my mistakeI was looking at svns apwiki which says Disallow +please keep in mind that a lot of code all over jackrabbits core would probably need to be touched with a high risk of introducing new bugs +I think Avery already suggested something about modifying to keep and send messages per worker not per partition +above should read +I guess if it is fixed in a later revision of APR we can close this Jira as Out of Date +The mail api was not able to send the binary attachment for some reasonJust changing the code now back to usage will solve the ftp problem but will cause another problem to occur +sandbox will not be using prototype library shortly +I created a new patch that simply deletes and add these files again which works as expected. +In order for all RMs to be able to do this all RMs needs ACLs to be able to perform ACL changes on the root znode +I like your solution +I think they are cornerstone for everything else and for custom pipelining +Building from sources can be tried just once per platformJDK combination IMO +The case at my hand is a bit different +Thanks Van its been applied +I guess the framework controls are going to be more important to document than the visual controls anyway and I cant see them changing much with the CDK port +I was under the impression that all of this stuff was only going to trunk +If Im using the Derby persistence manager I got some exceptions +Please set fix version accordingly +It seems tantamount to extending the specification with a new type +Thanks for the feedback hence we should not deprecate Jtidy because we dont have a repleacement for it. +EAP youll have to use native APR plugin to use it youll have to show me the stack trace as we simulate it if async HTTP isnt available. +Perhaps the problem is jQuery Mobile itself +Verified on EAP ER +PDF output resulting from the given FO input +Looking at tattletale next +for patch +How should the overlap with the current lifecycle code be handled +This pushed the throughput up to txsec +updated to this one over to Ryan as I dont have cycles to work on it anytime soon +select as c from t group by c doesnt work +Attached proposed fix and related test updates +Commit the changes to at based on the patch from thanks. + +I am having trouble replicating this issue to try and trace example above does not yeild any such messages other examples of how to replicate it +URI variables are not supported in redirect before +No junit test required +Here is relevant except from the patch build console log +orgapachehadoophbaseclient +If possible we should avoid that +Still that might be the best thing to do since using getpageslices is currently so hairy +This could be extended so that its exposed via the instance +Removing the reference of in +Ive added an option to disable this feature +Would setting required to false make a difference + for the patch +Committed TRUNK +I dont like how this has introduced a new set of for subscribing +It was released with the fix in +Since we didnt change that I went with what had been started earlier with the release +Of course this can be done but needs extra care +Works like a charm thanks again Brandon +For the This module depends on the former one and the build of the former one is failed this one will fail tooSolutionSee the attached patch +Committed wminor cleanups and logj pushed to ERROR + +if possible could you post the XML snippet that is causing the error +which means that there is no +Gerhard could you please retest this if possible? Wed like to ship it with txs +The scope should be provided as well +No problem there +Please resolve and close this issue if there is no further work to be done +one can be resolved +Committing this to +Im not sure what Xen does but I think most hypervisors enforce such rules that cpus have to have the same capabilities to live migrate +once these are addressed +This patch binds the signum function to a method in so that it should work on all platforms +to Jacks comment closing this as invalid +Closing all bugs. +Optimization for the simple LOADSTREAM and STREAMLOAD cases where the serializationdeserialization isnt necessary +Setting fix version so that issue shows up on applied! Thanks! +Please close or cancel this one +Closing all resolved tickets from or older +Heres a possible solution that disables the schema checks if you specify the java option with value false +bq +Can you go a little bit farther? Im afraid I dont get what you really want +gercan has started some work around this in Tools The Aerogear Eclipse Plugin can now create the structure of Cordova project which is BTW Cordoca CLI compliant +Hi HowardIve just tried to run only test test +I have the same problem with connecting to WFS service +Let me explore the option to use code based archiving too +when use connector caching is being used the next time server loads it loads the metadata from cache not from source +IsaacIm assigning you this issue to let you submit the sample projects as archives instead of war filesBest regards XavierI dont have those exact projects any more but am trying to replicate the issue by performing the same steps using the latest JBT build +Avro was not intended to be optional at compile time only at run time +Could you please submit an example we can use in a unit test or better yet aunit test Thanks +Should we create new JIRA to do that +Thank you AaronIf someone would like to we can commit +Moving this to +please adjust the priority to indicate that the Jira should be considered for CR +JBT is not supposed to offer installation of something that has the potential to conflict with it +Any workaround? +all no good +Fix by Philipp Ahlner +Thanks. +I noticed this independently and fixed it in my patch for +Did not add any tests bc it is a trivial typo fix +David do you have a junit testcase that demonstrates this failure +This is a zip containing all the new v dtds andJust one thing +to the branchI had to make a couple last minute changes to the two new Entities that were added with this JIRA to get all of the tests to pass +Wraps which is well tested and optimized already +Fixed in r Added that triggers a to flush its buffer when a certain Exception is detected. +I suspect that this may be device specific +Thank You! It works with batik from fop distribution +Its more intuitive +This solution worked fine until Seam CR with CR and GA its working only if I put these seam jars back into my there any way to work with these versions of Seam by keeping these jars in my s lib directory +It wasnt intended to be commited as it is +S SNAME pass the node name into erlang shellUse the ERLFLAGS environment variable and remove this option +Ok is wrong in that it doesnt require +Hi Sergiothanks for accepting +Solved as fixes this. +Ive got a patch that keeps it +Mainly because the strings are listed in the order of appearance in the file +No updates on this ticket since November . +Im getting build errors in a couple test Id forgotten about more tests being added due to s problems +the ordering has changed in +It is surprising to me that a flow definition would be Mb +never closed the issue +Hi JonathanI was actually talking about +Modified code to add parent CAS id to an outgoing message containing exception that occur while processing a child CAS +So all issues not directly assigned to an AS release are being closed +Ill close this till theres a benchmark to back the benefit of considering an alternative +See for further details. +The script is already running when it applies itself +This problem will be fixed by the compiler team +Thanks for the help. +Prakash you control that site +Moving to because beta will just be an update of and this change involves building new platform +Attaching screenshot +Can you please verify whether this is still an issue +We will not wrap code in empty try catch blocks because that is a terrible coding practice +Yea I should have closed this issue with a Will not FixNo this behavior will not be added to foreach The direction is. +Ok I can reproduce problem without patch so I will try to first resolve whatever the underlying bug is. +And this discussion is not yet finished. +Thanks TimThe fix looks goodPlease close this issue +Thanks for the fast replyDid you use YCSB to test at all? Or did you just use the stress tool +Getting hadoopqa to look at it +BTW it may be much more effective to post to dev rather than to JIRA +Running into this too +The first two items in Miloszs list are a bit more work +the filename is better stored in S as a metadata the less information stored in the custom tables required by the plugin the that I am referring to the S service being used in a web application serving multiple users where the filename cannot simply be be used as the key. +I used te next command for executionjava showversion Xemserver cp +reverted this +I meant I recognising that the amount of concurrent clients the system can handle is dependant on your test hardware +One issue howeverhsqldb will not work with +Patch attached +This patch fixes the NPE seen in the experiments when the statement is TRUNCATE TABLE or DROP TABLE +also said it has a hash based reduce faster then hadoop sort on hadoop summit presentation +Can you also change the other places default ctor is being used +Please assign it to whoever is working on to Edison +Attaching second rev of the patch +Please close this issue as resolvedVladimir + +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +Assign to helpdesk maybe +Do people think this is small and important enough to apply toGuozhang do you mind submitting a reviewboard +When you unzipped the test case did you remove the target directory before running the test case? The zipped up test case has a target folder that contains a hand copied componentType file +Can you please review +Config entries containing the have been added to SQL scripts +I see +This has all the functionality I wantThe new TTL test is subject to races if the test env is really slowWill play with to control time for these test to avoid these racesPlease review the other parts of the codeIs there some better refactoring I can do to in order to make this a bit easier on anybody who wants override some functionality +Fixed overlapping problem between and associations +However I believe we need to support guaranteed event delivery with TCP +I tried several incarnations of reproducing this but Im apparently doing something stupid because I cant seem to get to the same state where it was breaking in my local dev environment +The correct way should be to return the result for the found rows and ignore the onesThat sounds good to me its consistent with the shell in so much as requesting a row which doesnt exist returns empty results rather than throwing +Yes now its return to JDK and Compiler compliance level and tomcat is you very much. +So lets bite the bullet and update the golden files +Changing the name as Bryan suggested to and making it the first one in the file so any properties defined in will have precedence over any found in Ill commit this shortly +So this is working +Thats just plain stupid +But on the other side its not breaking any compatibility so it might not be an issue to backport it +Looks like Jim took care of this Ill create an iTest scenario for this and close afterwards +Done and done +but the use case is still valid and needs to be addressed users shouldnt have to set up their individual configurations to point to wherever the compression libraries are on the server +true but changing the way the default constructor works results in indirect changes to the code that uses that constructor +Also I left out the fact that I had to catch various exceptions most importantly +it makes sense that we support just type changes for secondary artifacts and should not require a d need to check if this is already possible with Maven and is only a plugin doesnt have a problem with this at least I was able to use the plugin to attach an artifact with no classifier and a different type and install to the repository together +Its possible to do this within common but if were going to do all that work add test cases for every class separate it into plugins based on deps etc why not just give it a new component name also? I think it will definitely help keep things separate +One place that we could add this information is in the Developers Guide the SQL tips sectionAlthough we might want to change the title of the section to Using SQL with Derby +Thanks +t instead +Bug has been marked as a duplicate of this bug +Nothing it seems +No URL qualifier was needed there either +I agree with Myrna that this looks like the same issue as + +fixed as a side effect of improvement in +Committed to trunk. +Sorry if I got it wrong this time +The production clusters Ive seen keep it at ERROR +I just committed this +I have now closed this issue as I am satisfied with the deploymenttests as mentioned in the group discussion. +It has nothing to do with failure detection and has really nothing to do with this issue +Yes but it contacts several machines and takes a couple minutes to throw the +investigated nothing to update in guide basing on dev task +Can an agent detect its own timezone? +Ok I understand better the issue now thanks for taking time to explain it Alan +Marking bugs with invalid fixed for release as fixed in. +contrib tests +is no orgjbossjsr folder mentioned in Jira descritpion +Its meant for a single user to use for local development +Patch applied as suggest by me and confirmed by others. +When reporting bugs we need more to go on than what youve given us +I actually thinking about committing this is as +Applied the latest updates lets see how the tests fair with these changes +Mostly just test code so I will take silence for consent by Tuesday or so +Salut Sylvainyoure absolutely right! can you suggest us please the oldest classes that can be reusable in this context? Im new in Cocoon and Im not an Apache Committer but Im interested in helping you +I committed the classloading part of Davids patch and provided a new patch for the bundle The patch includes exports of and as discussedIf no objections I will commit this patch soon +Thanks for bringing this up +from me too +This needs to be picked up by somebody else since Ill be on holiday until after the release date +The control is resetted to nice +marking fields final in path in the file in question was missing in the patch +nope this is fine + is no longer available as an extended metaclass for new +Another benefit of the stage model you can just stack stages onto theend of an existing pipeline to change things upIe you dont need to own the of the whole chain eg tomake sure certain specific impls are used for certain attrsIf you wantneed a different attr impl the stage just removes thelast one and binds its own impl every stage has full freedom toalter the attr bindings visible to stages after it +I am also on this is already complicated enough with the toXX methods in it. +Thanks Raghu +Is this parameter used by geoserver +In Rev +I can help with that +However r now adds support for instanceof and checkcast elimination by both aggregate replacement compiler phases +Ive renamed decorator to and made extend +Fixed bugs and code cleanup +Modify the two test cases Enable the table priority function by default +on this toobq +Yes I agree +bq +the thread local that held the fieldsStream was static and hence not bound to any particularEssentially that means that the offset and size were read from a random segment not necessarily the segment the field value was stored in + +This PATCH has the new sections +Does anyone have a sample project that we can use to test this +I looped twice with patch v on and they passed + +Thank you Lufeng +Thanks for the patchHave to check the codebase +listsforeach returns ok and also the return value of this function is in fact irrelevantThanks again Andrey +Remove new version of patch +Each socket connection will be bound to an io service in way +Can you use a new error code +Looking at our code writing goes generally together with reading +The problem was caused by the incorrect use of inFixed in the branch under revision r the trunk under revision r and the trunk under revision r +this makes my personal upgrade path more painful from release to releasefor you particular points i dont think changing the name would introduce that much pain +With disabled most of tests passed well but also there are another errors +commit the first patch of this issuePlease let me know if there are something wrong because I would like to continue to the others product application formsThanks Adrian for the documentation link I did not remember of it it can be very helpfull to meJacopo I agree with you but Im not sure if apply those default value can impact on the screen widgets that are not more converted to the new layout in any case I will let you know something moreDid you think that the layout is correct or you have seen something strange to correctThanks deprecated CSS style +Disabled formatting to make stack trace more readable +Of if your solution with only works for jvms that are visible in the Java Preferences app that should be fine +Patch also adds more assertions to +Committed to trunk. +This is a from it is a stop gap until is addressed +Tested with on Windows +I tried to find one but was not able to +The jar has been uploaded in the JIRA +Again we could provide a protected factory method thatusers on could override to return a +Thanks Ye for providding patches for Provenance Registry implementationLahiru +Rescheduling for s a shame you didnt end up implementing this using interceptors +Wido has pushed the changes only in master it seemsWido could you please the same in +Should I tackle this issue? You probably have more important things to worry about +Closing all resolved tickets from or older +There are these possibilities Reporter Assignee Created comment Attached fileIf we define this I may create simple jira report which gives you email addresses of users who do something we defined for selected versions +Marking this JIRA as closed since the immediate fixable issues have been addressed and tested +One more thing that is still not clear to me +A patch to get keygeneration to work with Firefoxes keygen not working in xhtml mode +Snowdrop needs to be added to the Programming Model category in the primary navigation +Has the diff on reviewboard been updated? Trying to review but looks like its still the old one +my time budget for investigative tasks is pretty well spent at this point +It probably still doesnt work but at least it doesnt cause an NPE now. +OwenAshutosh Does this change the format of the file at allAre there any backward compatibility issues +If it uses an inappropriate method that is not really a problem for LANGHowever it is important that the LANG Javadoc is clear on what each escape method does so the application designer can choose the appropriate methodThe downside of changing the encoding is that the generated output is longer and harder to readAlso it may be overkill for some purposesMaybe there should be a stricter version of the escape method that doesAlternatively the Javadoc could point users to the ESAPI library if they require more than just backslash escaping. +Currently this interface notifies about VDB loadunload and ACTIVE +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Update patch to remove the withIds filter method added in +Youre right +A quickstart that shows the behavior in +Thank you very much for fixing thisDan +Added security for upcoming ports and kernel looks to be a red herring +I am pretty interested in fixing this issue +Document containing screen shot of total memory metric you ask the customer to go to the full page chart for Total Memory? It should have what the baseline values are set to at the bottom of that page under Metric Baseline Expected Range +Sounds like a good solution to me +The attached patch is a possible solution here but Im still trying to find out exactly where in Hadoop this gets used so I can verify that the functionality still worksThe libzip and libzip packages appear to represent versions and of the same library +Please update the patch incl RB and resubmit +I guess that oneslipped +The rest of the cell doesnt seem to come intoconsideration at all +Verified on JBT. +It could be better for the componentized fields i think List of List will not work with my patch but other cases should be ok +m and now I can continue +Job configuration about how users specify the resource usage +It uses the regular rather +Sorry for the delayed reply +The error occurs in plugin +This patch adds versions starttime and endtime optional arguments to the Exporter MR job +There is no harm in including the token in these requests since it is only included in POST PUT etc + with its invocation can be extended with appropriate support +If overridden method is not found then OPT does not optimize a call and use the original methodIve set up the VTS VM test suite on a Linux machine and will report the results of testing the patch later today +I raised to track this at Apache +Thanks for the useful hints +retiresjobs immediately +duplicated issue is different from +Thanks Torsten! +Whats the status +Patch checked into trunk and branch +The test integration framework and related tests have been completed in Teiid +Passes unit tests though TTI was giving me a little trouble +bq +Review please +Is that something that someone else has to do? I dont see an option that allows me to do it myself +Please confirm the current status of this issue and if its incorrect change it as required +This will hopefully solve the issue +Im going to run the tests again to confirm the results +After SBS restart system jobs started to work properlySBS currently sending all those not delivered notifications from the last week +Bleh I gave this another minutes and decided it was too irritating to do right before release +Closing old date issues. +Please let me know if there is anything I can help do +I reopened the issue because the problem is still reproduciable in that you try and reproduce and capture a complete backtrace of all threads and attach a report in a platform independent format so I can review it +Thoughts +almost no changes from original to branch +bq +Prefixes come only into play when we have to process or create a serialized form of XML +Of course the behavior I described previously was tested against Hibernate when the Schema was asked to be generated +will check for TRUE and intercept the source file being zipped loaded into a dom document have the tags with connection type namespaced keys removed +maybe it will happen someday soon. +I guess in your posted issue you have already fixed the problem but there is an error in console +One of our users is able to get this to happen reliably +closing since fixes are in released code +Dmitriy have you found the list of valid characters for column descriptors? I feel like Ive seen it before somewhere but I cant find it +Will doJonathan you mean in the book? SureIll add that and post a new patch +Currently the ones that are not defined in that are hidden configs that used only by internal implementation +Thanks for fixing this +It seems that in the prefs inserted are simply the ones passed on the request +Depending on your application you could enable or disable this filter +Bulk cleanse of EMB issues moved to legacy version +diff file +for inclusion in +Our IDEs dont know how to do what is coded in our DescriptorIm thinking this issue should be rejected +bq +Yes it should be calculated by default +Revision . +As noted above in the latest version of Safari it works fine. +I copied it by hand from fabricator for now +Same test +With these changes the attached test document is correctly recognized as HTML and parsed with the lenient HTML parser. +im setting the ticket to Cannot reproduce Bug please feel free to reopen after you have tested and you have more detail to replay the issue andor you can answer my last question +Would it be possible to replace the core templates with them +with do this mvnsonar +does not support anything but right now +Is there any maintainer who can review and commit this trivial patch? Or the project is already dead +Other than that it seems to be working +But if you call it as simple auth instead of security off it becomes meaningful +were you able to generate the any of the stubs +Closed as Rejected +Verified fixed in CP CR the quickstarts and mgmt consoles use of ant is working after the UELA file name change. +Next release should have it +Ah Jesse since were in a hurry with geoserver can you commit my original patch anyways? It fixes the OOM problem were facing in geoserver well have time for better solutions you allow me to I can commit the patches myself +These new match functions should use the methods in the class instead of +attached a screencast which demonstrates the new behavior +Thanks ATM +For In Portuguese standard and default correctly translates to padrao but localization team should come up with a way to differentiate one from the other given that we have both configurations and they are not the same the word web is correct it should not be translated +Title updated +Steve Hathaway +There is a patch attached to it you can try while we wait for SVN access to be restored +No need to apply the open a new issue for any more changes that you need. +Some sysadmin had changed some registry value for Visual Studio which made the compiler compile header files as well +I included a new patch to see if that helps +The hours are passed you can check in and if you likeThank Ill proceed with committing those fixes shortly +This is implemented locally +The seems correct though just the are missing +Would be great if Camel is spring compatible right out of the box +Build works with it for me on both windows +from me +Well the issue seems partially fixed but not for all cases +Verified fixed in id date +Salut Olivierhave a look at the patch +bq +I have not looked at the symlink patch yet to see how it handles this but it might be useful to support atomic replacement when both the old and the new files are symlinks +OK Ill make it DATETIME instead of TIMESTAMP +Okie doke are we doing unit tests with thrift yet +I dont think this is practical to do and I dont think Ill have time to think about it in the forseeable future. +I am hoping that as soon as the logical stuff is done we should have access to all the data classes at the least +I think it is a valuable enhancement to allow people to determine which fields can at the client but others for processing +After applying this patch the problem disappeared +The patch is applicable to all the affected branches +There is a similar situation with Lucene when people import from indexes they make in some other program +This should already work +Many thanks for all the work +fixed by Jose +I am facing same problem in Websphere +We may be able to work around this by adding a plugin execution which runs after the test compiler during the phase but Im not sure if that can be guaranteed +Okay added a test case and committed the fix +If you want to include chars in the XML file then you need to change the specified XML encoding value firstIf changing the encoding to your locale charset doesnt work then reopen but please remember that we have never tested or claimed that Geronimo supports charsets. +Attaching a branchx version of Erics latest patch +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +As far as I can see all lines on Windows from a generated +Given that its a static class anyways why not just call it Castor? Other than that I think JDO is a fine name as is +This issue has been originally reported by Tom Burns. +so basically its all the versions which are affected +Not found this issue in latest runs +but Javascript is not Java +I have added it to our source with the exception of the sample script + for the patch +or FIXME check if +Updating patch including the limit on the number of threadsWith this patch in most cases the should scale according to the need +for prefixing everything with plexus +Confirming the snapshot doesnt fix this +Well technically speaking it does affect HS even if its not getting fixed for HSThanks for responding so quickly +Nicely done +Good point +But the guilty for the flaky test is found its +Nothing changed in the principlePlease note Im signing the CLA in case you decide to apply this patch. +patch looks good +The problem is worse when I begin to use spring distributed transaction +you have to change mvn at the momentWe are getting away with it by mavenizing trunk. +Consider only the patch for +It looks as though we have to use APR or higher as the branch doesnt seem to allow any sort of detached pool +Some of that part was refactored by some defect by MingAlso in trunk as is there we have disablingTables and also enablingTables +Just to be perfectly clear can I take it that The project can implement a RTMP plugin based on the specificationand There is no requirement for the project to add anything to the NOTICE fileOKWhat about the LICENSE file? Does anything need to be added to that? It would seem odd to me to add the spec license as would not be providing the spec but an implementation of the spec +Sounds good then +review in queue +What results did you find with yourpatchSandy +Its easy to reproduce and I have a good idea where the problem is +Merged +Im not sure but I think I might have had a problem with subclipse +Perhaps the layer list is cached in the catalogue or anything +Could you explain how to apply this patch to jBoss? +I dont see the improvement +the lack of something between weak and soft is an unfortunate complication +I dont think we should do it +Youre right theres no support for the combination of derby with JCC now +Could you provide your settings and an example of directory with your revisions? Ivy uses directory listing to list revisions and depending on your settings it may confuse artifacts with module revisions +These features below still have wrong year in copyright information fieldNo date in copyright in +No idea if its related but we might want to shoot an email over to to see if theres maybe something internal that we should be paying attention to +Rich Could u please take a look at thisthanks test providedDims Please test and reopen if a problem is found. +Are you able to reproduce this with a standalone test +add test case in +Any comments on this +Patch applied on and trunk thanks! +it was working +After fixing the test takes over seconds to complete +Should we always throw an exception or should we not an assume anyone that wants to catch it will handle it themselves? Checking the item count seems possible but it feels like a work around +Resolving deleted terms doc doesnt require a sorted terms dict right? Ie a simple hash lookup suffices +By carefully listing in artifacts first we can alter the list of jars reported by so that the newer ones are on the top +We will do the review once we are ready to merge the changes into the main branch +Thanks Andrew for covering me. +Hi Mark +I havent tested it much thought +Committed to as well +To make sure ask that he includes the hserrpid log if the agent crashes again +The advantage of providing it in would be that we have more time to get feedback on the API and apply changes by so that the API is stable in is unit test is +Also verified that the maxmax is computed dynamically based on cluster resources +If you beleive that should be please and contact the program team. +the patch files are confusing and I got no rights to delete attached files Icreated a clone +This patch now fixes thatNote This is a temporary fix until a more global name solution is implemented +Probably also should have associated token if it exist +The docs were done as part of closing this as a duplicate. +Id love to get mapinfo MIF Datastore support in geotools +It is when the scheduled build kicks in that the ticks become errors even though the build was have about projects in continuum if that is any help +I also think its a memory problem +It would be impossible to be consistent until allows always prefixing those commands m not against the idea of adding j versions of those remaining commands but I dont see a way to make things completely consistent with involved +It was last commit on of a bulk update all resolutions changed to done please review history to original resolution type +Closing old resolved issues +Checked in a patch +Something like a sha of all mac addresses in sorted order? Then again Im not sure how hard itd be to get that list from Erlang in a portable fashion +So this bug report is about misleading log message? Or does something really fail here +Nice +John added some review board comments to my latest patch and I have been occupied by other things +When I edit a file in emacs I dont get the symlink +Closing issue Oozie is released +It took me some time to recover and get back to this jira +Cant fix the description formatting sorry +This issue is fixed and released as part of release +just committed this. + +Hi Hoss will do all javadocs and the code is revision Thanks Noble and Hoss +Changing in mapreduce should be good enough +but basically Im seeing the flash scope having values from the previous request even though it wasnt a redirect +Committed with revision +In this patch Ive improved unit tests and cleanup some parts of physical operatorsAlso I vote for Camelias workCamelias patch contains and +committed with the equals changed to test the text and locale + +Proposed patch but not tested very much so far +Jopr rev For authenticated TC if the user does not have control perms ensure no part of an app delete succeeds +Is the changed plan really a product regression or just a? Can the regression checkbox be unchecked +Thank you for trying to standardize that will be is chronologically sortable by default by programs like ls and dir whereas is not +Resolving. +Added Fileinstall to build dependencies due to fix for +push to +The testcase provided by Andreas Felix may yet be useful in the future +This is a container specific issue that can likely be addressed in that forum +Yes but if you remove Grails will no longer work on jdk as this class doesnt exist there +Using LDC is for Java standard since at least +Perhaps some day in the future I will attempt it again but not now. +This bug has the same issue with odfdom bug This bug has the same issue with odfdom bug Bug Table DOC API not usable for Footer HeaderI think its also an opportunity to refector Table API to give it a more reasonable designWe will also resolve the similar Bug needs helper functions to get to Table in this issue to avoid conflictPatch will be submitted laterDevin Table DOC API not usable for Footer HeaderI think its also an opportunity to refector Table API to give it a more reasonable designWe will also resolve the similar needs helper functions to get to Table in this issue to avoid conflictPatch will be submitted later +Was this intentionalMark +Thanks Ian! +And the previous comment should have referred to comment the patch just needs the comment +This replaces the one in to run against of GuyJust to be sure are you saying the bug doesnt happen for you on OSXUnix +I think it couldve been using some descriptive message to indicate the cause of the failure. +Hi FranckI think I got the problem +Fixed in JPOX CVS builds dated or later. +Manually tested! +My name is Bob and I approve this message +Andre Im not able to reproduce any more in Developer id date could reproduce this in MASTER thus not resolving it +Whereas the eviction thread would have done it itself and so it wouldnt have impacted the main thread? Can you confirm this Trustin +This might more be an issue of which Windows environment you are on +We still need to find a way to handle this for +hopefully there are not too many people running the last good version +This patch adds date support to the and the JSON reader +Just upgraded to Maven and java +har correct hot deploy throws error above + +I used Acrobat Pro and Reader X with the same results in bothSo do you think this is a bug in Adobes implementation of the dictionaries? +As Im testing it right now Ill fix all the places which have problems caused by this and will backport the changes to other branches when needed +QE please verify this was able to reproduce in Metamer on this page httplocalhostmetamerfacescomponentsrichTreetemplatesrichIt works fine in Metamer because of popup panels attribute dom set to parent +Perfect thanks +It would be no great loss to just remove that sentence since its just an introduction and not a reference +Does anyone have any comments re relevance of this with regards to the way Nutch is moving? I think it is out with the scope of the release +If the user is written as an OR such as the above then the rewriter will change it into an IN predicate +Im on x system with maven +Thanx for finding itThis patch has to get in +Actually I believe that David is right about the reversal +Committed to TRUNK +Added a simple check to the +Im marking this as Wont Fix since this change only provides minor improvements in very specific scenarios and seems to show degradations in other scenarios. +I think its just per timestamp +bulk defer to +Out of curiosity how big is your typical Configuration object? Ie how much memory will a typical client save once Configuration becomes a class +Would be nice to have that patch on a release +Reconsidering this since it would be a breaking change +install a libvirt hook c +I deployed continuum into the same tomcat instance and even this web application has the same issue +patched for the patch +I am going to fix this together with +Just tested it +What do you think? People might find the docs useful +The tinderbox test indicates that something has been broken by this fixSeveral tests now fail with ERROR XCY Required property has not been set +It sounds ok +Feel free to do a review for this patch and I will then push it into SVN. +Done in and trunks +bq +I chnaged the theme from Roller Homepage to Brushed metal and I can see all my entries I played around more and noticed that I ONLY get this error if I fill something into the field Tags strange +Logged In YES useridAfter getting the iiop tests to run under java Imerged the iiop and jacorb jars back to and all iioptests are running under java +Attaching +It will break any code that uses the existing Admin connection and will not prepare us for the next Teiid release as the operations of the Admin are expected to change +Ive noticed this behaviour too and have verified this behaviour in a debugger +No feature changes +Editing title to be more appropriate +ThanksLeo +You may not move the block from a overloaded node to a node +Furthermore we need to add some license headers +Release Notes Docs Status set to Not Required as all documentation issues and requests are being handled via a separate JIRA issue +We observed that this happens only for very short jobsTo validate the patch we ran a workload where we were seeing this in out of jobs on an average +conf test variable has in it and may be failing for the samereason +Closing +The problem was the checking if a class is audited it simply checked if there are any audited propertiesThis in fact was changed recently so it may have worked before +this fix until after the release +from remote address from client +Im suspecting that the parent class is not there because its expected to be the generic class in jdk dont know what to tell you +But by casting the to a is available which exposes the get function +does anyone else want to comment on this patch +Retargeting to +Thank you very much for your patience. +You only fixed the in javaapi +Correct +Added another check in the test case to show the property affecting only one service does not affect another scenario that was asked about +No it does not provide +or maybe peraps subclassing the relevant bindings to do the ids +Anyways lots of HTTP servers out there in the wild do send a response and immediately drop the connection before the request body is fully read +Im committing this +Thanks Raghu! +Dupes that Alena fixed +I corrected the status typo and did a minor refactor on Guos work +Matt I cant reproduce this problem in CVS anymore +Did you forget to hit Clowncopterize? I cant see them either +Well have to just live with documenting where the files came from so the past history can be looked up within the originating repos +So it is acceptable for +I like because that leaves us room to add useful things to it that dont make sense in +But now that this is happening in my project I want it fixed Ill attempt to reproduce the issue this weekend and upload a zip of the that I dont have time is a work around thoughIn your repository where the released pom was installed copy and paste the missing XMLto complete the that everything works fine +id starttime variant name perf was actually two bugs +See for further details. +The patch seem a little bit dangerous to me +Addressed Teds comments + for the patch +As Christian already mentioned in there are obviously other places to be fixed too +Quick question Your patch does not change runConfigure so whos going to define the symbol GXX and GCC that are used in your modified +Thanks Vinod for the reviewUpdate all your commentsThanks latest patch looks good +is stricter when validating the URL but this specific URL is a special one that works for all versions of IE +Thank you Chris for fixing this issue. +Ok patch has the code for main and part of patch contains the unit tests +bulk close of resolved issues. +In the code contributed by Kishore the type translation was done the same regardless of the bindings used +Yes now its clearer +and we really like that toucheffect know Dan mentioned it on our behalf but you guys are just great +I could not reproduce this problem but opportunity for exception is evident it was previously hidden by swallowing exception +An update of the tutorial +I was not able to repro the other failures +Ill update the apparently out of date FAQ. +I can get my hands on a nexus oneWhat do you mean by Original Galaxy? The Samsung Galaxy S? Are these the only devices you see this on +bulk defer of unresolved bugs to + for the patch +I just took a quick peek and it seems to be in +I think its possible because the c field is not final seeUpdating to version should fix the problem +Which as a user I would read as not closed either explicitly or by garbage collection +We have to wait on other stuff before this can go in +code freeze +But thats because the method is private and MVEL only looks for properly accessible methods +Describering is needed even if topology isnt changing +One of the comments for that issue includes the following quote It turns out that the introduction of these invalid predicates stems from a problem that is almost identical to the one described in +Sure +The new patch mainly fixes the bug that we need to update the blocksMap after replacing an with a corresponding +Was fixed to behave as it used to +Let me describe the change in few words before the change APRRING and global pool were used to keep VM instances +Thanks Brian +for the patch +All the tests passed +Works well Julien but could you also hide the Obeo property? tested! +preferred user id jdeoliveforwarding mail jdeolive +Im sure we have lots of places where we use warn and warning incorrectly +You can create and add patches as necessary and well apply them +Moved test code into and added Schema mapping for string datatypes to STARTSWITH UDFThanks for the great hackday +With you draft you did only move static methods from to its factory +I mean properly speaking having a default correlation id doesnt really make sense does it? Anything other than a counter is basically a bug +Jira was triaged and was decided to just keep warning in documentation +Would anyone please try it +Thanks Vinod +Setting the version number is a bit primitive it has to be set locally to the right thing using a file +Wont fix there are no security issues relating to this +Do not delete but deprecate public methods otherwise itd be an incompatible change +Can you please explain the issue in more detailAre you saying the id field value is copied to the name field? I cant reproduce that +Committed as revision all resolved tickets from or older +Jmeter test plan used to load test application +The state of each proxy will be kept in sync on each node in the +I would phrase it that the purge policy determines how long to retain the files online and what to do with them after you no longer want to keep them online archive delete or whateverSo calling it Retention Policy and would be fine +Or does it breaks your usecase +OK Ive added the following to Volume Up and Down should work Legacy Back Button should work Method that takes in an Android Key Code and overrides it and sends something backThe question is what should that be? We need to be able to fire a generic keyEvent with the key code attached + line in the replaceDates closure +I have to check it a little bit laterAnd yes please assign it to yourself and apply it +What was observed was that the number of minor increased a lot during the test +changed the default for and updated the release notes +One thing to note is that just because a host is not known initially it doesnt preclude a DNS update from making that host resolve at a later time or if running on an isolated system the DNS could be down for some reason +Perhaps we should add a utility method that checks the writability of a directory by creating and removing an empty file +In case of it doesnt find whitespace and deletes the whole URL from rawexcerptI see two approaches to fix this Since the word boundaries are already computed during analysis we could try to reuse this data +Would we be porting this to all other libs? If this is just for java then we should make it optional rather than the I this as a language option? From my perspective this is an inconsistency that Java has when compared to the other languages constants cant be defined in different programs having the same namespace + +By the way is there a place to get the FINAL binary zip? Looking on the infinispan website I only see FINAL binary zip +Its fine by me to support folks who already have a keytab which contains only the HTTP principal I just want to make sure we continue to support folks who are used to putting entries for both principals in the same keytab +Good Message thanks Knut! +Pushed the private changes in acc +wait for at least this minimum time before sending another heartbeat irrespective of whether a resource request is pending or not +The entire URL is configured in as bug can be override the url through defining it in +Has this been fixed +Fixed on. +It would be good to understand what exactly is going wrong do you have some documentation to point at +Done +Updated overlook in the which is going to be fixed with Sonar this ticket is no more valuable. +I have applied the patch to and branch +Ive only encountered this in. +Both patches are for applied +A few of these properties need to be evaulated in the current implementation +Worst case if no one is interested in doing that work we couldput examples of the problems someone would get if they hit this and workarounds +Thanks +Here is the fix turn on WARN log level for the test case +Hanks JimPatch is applied at r please verify it was applied as you expected +This is fixed with caa. +In addition to that I see that for the join order in case we indeed mark sort avoidance for Table m I definitely need to do more debugging to see why we think for this join order we can avoid sorting for Table m +Yes you are right was also added previously +Hadrian could you take a look at this? If this is a component problem let me know and I will look in more detail +merged +It also enables tausimp optimization pass to reduce extra type and null checks +Now we shall wait for people to vote +Corinne wil be working on fixing the Matthiasworking on cleaned and sample code repo updated +Oops meant to say you can check to see if APPLE is set but that covers Mac OS classic too +patch to log port opening exceptions and send a page that tells users that the applet wont work +OK its possible to reopen ifwhen theres a submittable patch to consider +Unfortunately this didnt solve the problemRegards in jboss reflect should fix can you please try this new snapshots? +This is back in web is being released now so should be able to kick a fixed build later today. +If you put it that way yes its a dependency issue with the staxx it affects the xmlbeans m not the reporter of the bug Ive just been watching the issue because its close to my problem + +Thanks for the explanation Andy +No set yet +indicator n +Looking now +svn diff for examples per to fix the example applied but leave this issue openned until we update the site +The fix has been properly applied Thanks a lotAnd I have attached my unit test cases to JIRA. +removed an unnecessary synchronization +will commit once tests + +it shouldnt cause an AIOOB like this +committed revision close for Solr +User created email address of scm +Thanks for the quick feedback DanI guess I was a little surprised when the base test classes didnt load the driver in their setUp but I suppose that makes sense if you want to test different driver loading behaviors and still use the base classes +In this patch I can see AHS is going to be an completely independent process +Javahelp needs the paths to be relative +Maybe it does support excluding in some other way +Scan count calculations are already performed so we can use this to implement a count control on search instead of performing the search +There we resolved to have an all option but not allow specifying multiple states +I had forgotten one comment and thought I could just add it directly into the patch fileLooks like I was wrong +pending Jenkins +Reviews appreciated +Retrying +STEP Code Generation Apply the patch inside QPIDCPPROOTrubygen +Marking issue as still open since patch didnt really address underlying issue just clarified status quo +Would anyone mind if I just inline the current released jar +Please do not waste any more of our time submitting your bugs to JIRA +Manually tested! +Let me know if there are any problems. +Per the spec your expression is at least syntactically correct +Coda Ryan not sure if itd be easy for you to verify but if so Id appreciate independent verification as far as I can see this should plug it but these are nasty enough failures that Id want to +I had only been running the unit tests but Ill make sure to run the system tests as well for changes I make in the future +bulk defer issues to +Open in error +I did some looking through the JCC driver code and the problem is in how JCC class handles resultset type downgrading +slip to next m resolving this now as the primary use cases have been fixed +The comment from Emmanuel is over one year old +Removed the folders +behavior would be to simply error if start or rows +Maintaining something like that would currently be pretty hard or impractical +Addressed Sijies comments +Sincerely. +Reliable sequence for +These appear to be the same issue +Thanks I got it and ran it +There hasnt been recent activity in this jira so I propose to move it to +Worked like a charm +This enables the existing HQ portal architecture including declaritive struts validation to be reused. +on patch was manually tested on a secure cluster inThis issue showed up only in a secure cluster and the code change in the patch takes effect only if security is enabled +Release note flag is already added thankyou +be ready to pay the tab at javaone! +Teiid Execution Plan View text and tree tabs shown +closing stale resolved issues +Again Throw an error on collision +Just one observation +I dont think this will be a problem +Is there any discussion what needs to be done for this task +Thats correct +Committed this to trunk iirc it doesnt happen in the branch +This issue will be easier to address when were on spring security +Keeping issue open as it would be nice to update the docs as well +Since GWT and JAXB have different rules and constraints on the objects and service interface Enunciate had to provide a mapping layer between the endpoints and the GWT endpoints +Anyhow how would one go about representing this data? If you want to represent durations in the d h +m and compile withmvn package Ptmpzk +The GML data store was removed from the sources after years of being closing all issues that have been in resolved state for months or more without any feedback or update +HiWould you please try my patch? ThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I just committed this. +I will provide an example of shipment gateway that it will use the new shipment gateway config entitiesThanks that all the shipment gateway supports the new Config entities I could close the main issueThanksMarco +Elena can you try this on build ? Im not able to reproduce this on because of a fabric issue +Hope its not a nuisance to do thisAbout the shutdown message Is there a way to put a space after the colon to make the punctuation correct at least? Shutting instead of ShuttingThanks +Done +Mark the BDA concept might be broken +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +pushed to master and in JBDS B +I just ran into this myself +I thought about that question quite a bit +Do you have a test case for this? Preferably an isolated unit test that does not rely on Hibernate but one that uses the cache directly +I think throwing exception immediately is betterPlease also run through test suite using Psecurity since Hadoop QA doesnt test security profile +changes include Added an updateToken method to the state store interface and filesystem store uses rename to try to make this atomic +The output of using this should be +David took care of this already it seems. +I know you got it already +Please refer to as +Fixed in r +Can you easily back port it or do you want me to resubmit +The problem with using a label attribute with this tag is that say you have items in a list all would get the same label +The problme was probably just a shortterm inconsitency of the repository +But it may not completely solve for this use case + +Im trying to use the plugin on project but the memory limit make builds to in . +It is wrong to assume empty string when there is no httpLocation value is given +The patch includes the fix and a test case +You should not mark the issue as resolved +Thanks +This is essentially the same issue as +After looking at patch all it does is test for null +That is how it is supposed to be +bq +If so could you please add this empty directory to the repository +Alternate row styles are typically used for long lists but they are not used for every form +I actually checked the change into trunk +Did this used to work in +the new uses now the compiler directly and for multiple classes at the same time which avoids this proble +The real problem is that the requirement for servlet context attributes to beSerializable is totally bogus +We dont have data related security control at the my bad was trying to go quickly and didnt read the full thing. +According to the specIf a bean that declares any scope other than Dependent has an injection point of type and qualifierDefault the container automatically detects the problem and treats it as a definition errorIn our implementation we treat disposal methods getBean as a Managed Bean and it has not scoped as Dependent +Closing under the assumption that it has been fixed. +need to find the time +We should also move the code from dAH to dHTE and give it as gcBefore since there is no need to age out tombstones gracefully +Oops the fix to will come out with Groovy +We have to enable setting SOAP as SOAP version to test the changes related to this of +Also note that the code where this is throwing is actually within a catch block presumably trying to get the logger +Ive attached a patch which offers a possible solution which initializes the with the bind variable values and therefore does not require a change of the interface +This is the latest patch works well on my networks and caches the local hostname and only +Ive verified console system VM is up and running but I just noticed the of the console system VM isAlso Public IP Address is and Private IP Address is which doesnt make sense to me +Here is a really crazy issue Thanks for your report +Socould we have a similar issue there I wonder +SadeepIs there any reason why this issue is still openBTW has the EJB sample been checked into the trunk +Are you proposing that the exception should be clearer +Patch checked into branch and closed with release. +Apply to root of saw the account request go out yesterday +Not to mention we can start to build out the wiki a bit more +Test cases still need to be written +This is really just a JAXB bug. +Ca it be related to +It catches all native memory in its native part and doesnt free it +Patch which uses as a prefix for the directories +Looks like the code originally posted by Andre made its way in for one revision then got lost +For a user that should not be recommended an item we join an artificial summand into the estimation of hisher preference towards that item so that this item will automatically be excludedPatch is coming soon +Also in master. +The current attached files have merges from other ticketsThen its easier to review your patch +trigger the a different note was this schema generated using Apache Directory Studio? I know Studio sometimes generatesa Base encoded value in LDIF for a human readable attribute the Schema project is created with Apache Directory Studio and also exported to LDIF by it +You are right +A new patch based on trunk and integrated Nings last comments +Thats a much better solutionIll get working on this and submit a patch +yeah i will post a patch +hi Deepak Its almost the same of what to patch +However there are other operations we may want to soft fail like getting the current position from streams that dont support positioningNot sure of the best way to proceed here +I made some changes particularly extruding cert files in the form they went in +Should be resolved +Parameters are now stored in a as well. +Hum yes that sucks and Im pretty sure this is tested in the TCK +It has been doing for openejb release and release +im running the tcpmon for all ports from to +I will create another JIRA onBest regardsSpark patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I applied the attached patch to +Fixed and pushed +Increase your JVM memory by using the java commands Xmx option or equivalent +Applied +All suggested changes but and have now been applied +I can work on this in the next few daysThat would be great +will do thanks Suresh +Nick could you please review again? Thanks +If you are unable to provide such a reproducer I recommend taking the issue up with IBMWebsphere. +Yes Thanks a lot! I did not notice that can have none values +Patch applied under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I should have checked the ticket first +Thanks for nice patch Subbu +Duplicate fixed at with the release of. +I will add another JIRA issue to have Hazelcast support +I just committed this +Lewis Both git repos are writable now +Please see explanation in srcmainjavaweb +Should I or others check test and commit the patches which include modified data convertion for different parts? If no objection I will try to commit them several hours later +Ill modify the header for the log so that theres a timestamp written. +I just removed Eclipses workspace to start with a clean environment using Tools I create an EAR project create a SAR file to be located within EARs then invoke Add and remove projects then the result is entries for one SAR observation in the past when I remove my EAR project without removng the generated SAR file and later create a project with the same name create a SAR file for this project with the same name as used before then I get additional entries within the list for Add and remove projects +explicit backup calls after setItem solved all my problemsshazron have a nice day off and thanks to plugin +Brett I openned the bug for the wrong version +So by using the latter instead of contains the condition should lead to a consistent result no matter whether another thread already managed to get to register or point this was originally meant to help with the interaction between multiple +I decided to rework the SQL sample code so it would use Derby both as an effort to promote other Apache products and so that the SQL extensions could be demonstrated without the need to perform complex setups +OK +all tests passed apply it on behalf of Jeff Yu +Reload the same patch to kick off test +feature was pushed alive on Oct +Shouldve named it v or test or something +no +applied to master and +This is something Id implemented previously and probably had in mind when implementing +Is there any work around for this issue +Thanks Carlin for your comments +I was able to reproduce this with Derby +Question how are components that the user can update highlighted in Central? Or must the user use HelpUpdate Software +Regards applied +I dunno I think of it as exactly that flag doesnt seem bad to me +Hi KimI would consider this to be behavior +Finally updated the site the logo which most people liked. +However it appears now that this entry gets reset every time the container starts up at least it was for me +The question now remains does direction provide any information about the relationship to warrant the confusion +Keith any news about this +Is this in the ESB docs or just the docs +Also wordcount cant find the schema because of the dynamic schema changes +Dear allFirst the workaround works fine +I wont broke anything it will just make meclipse work as Maven and Eclipse by default +But just casually looking at the logotype right off I thought it looked good +It is a violation of the bean spec +fixes a typo causing the test failure +Updated patch the specialization of broke this navigation API as well because it doesnt have subscorers instead accessing the docsenums and shit directlyBut there is a way worse bug in this if even one term is omitTF in your term conjunction it treats the whole conjunction as which is wrong scoringI fixed both of these in the patch +asmx files +Yes it does but it assumes that you have schema in agreement across the nodes before it tries to fix timestamps + +Depending on the design of JDOs query api it may or may not be verbose +Thanks for the alert Ted. +Attaching the patch +I noticed this issue does not exist on +Your static map approach will cause a class loader leak and will not work if the objects are not deserialized on the same JVM +Wow thats something I never thought anyone would doIt looks reasonable +Thats weird that the icon wasnt attached and the patch didnt import +Thanks DyreI would be very interested to see your work on this issueI too am working on a prototype to negotiate the encoding in the MGRLVLS during EXCSAT and am having good results connecting old server with new client and new server with old client +I did a fix on that did not involve API changes +Patch attached to fix problem no tests found in didnt realize the test exclusions would add that much time +Instead of fixing async invokers should be removed all together and replaced by NIO invoker +There should be a version already in the root pom +Narrow down the input type from to +Test case included +Is this still a problem +Content of varconfig dir when boot is finishedgnodetguillaumes cygdrivectmpvarconfig lltotal Krwx gnodet None Jun rwx gnodet None Jun rwxrwxrwx gnodet None Jun rwx gnodet None Jun +JacopoCould we not have all in patch to facilitate testing? JacopoI will give it a try tomorrow +Okay to close +after cleaning the relevant cookies I cant reproduce it either. +Are you planning on working on this? Happy to review +We have index that does incremental updates and another that rebuilds nightly and its a good technique in both situations +I will fix first before fixing this one because the newline in the data caused the file format treats the data as having an additional line. +Likewise I feel we are not gaining that much by making immutable per default +Looks good to me +This should really be scheduled for inclusion into EAP +Thanks Steve applied +This is an area I would really like Joerns and Scotts feedback on as they have been dealing with Lillith and Chainsaw and presumably understand the issues better than meEven adding a new field that allows arbitrary serializable objects to be added to the event could have the same impact as item so before anything is done with this I would really like to get more feedback +Looks wait for feedback from the user for a little bit before closing am facing the same problem with +She had missed mentioning it in her description +Update I found that if I specify the WFS version as it works fine but does not +Ive debugged the time it turns out that the jQuery remove takes very long time +It only applies for the case where no term is found +Further more this problem only occur if we enabled Quartz clustering +No but I can reproduce +This issue is because of using wrong schema when process the splitIn getSchema function of if a partition is the prefix of a split it will use schema of this partition to parse the split but this is not always correctFor example partition userhadoopdatapart is the prefix of userhadoopdatapartsupp but we can not use the schema of userhadoopdatapart to parse userhadoopdatapartsuppIn my opinion if a partition is the path parent of a split we can use this partitions schema to parse the split +If were unreferencing x ISTM we should have the opposite problem of premature freeSo maybe we have two bugs +verified and fixed as of revision +Ive to make some slight changes concerning checkstyle complianceThanks to Sean for his contribution. +I only care that it compresses well is reasonably fast and doesnt corruptRight the issue is probably badly named +If Marshaller is bigger the test far all test fail except when marshalling immutable objects +The current code relies on the fact that the events passed to the query handler reflect a correct state change on the workspace +This issue seems to have been addressed with code improvements +Patch applied with minor logging additions. +Fixed in +I will do some tests +We should also only bind localhost. +wont fix +I have not verified in anything but +Heres a patch that weve tested for a long time in an build +Closed after release. +Moved fix described above into shared dojo code that does positioning of other floating things like popups as well +JeremyThanks for getting back to me so quickly Ill make these changes next week and make sure to hit the ASF License button on the attachments +this looks promising but its not passing the tests +In this case a data node may have to verify block data first and then regenerates the checksums +Committed to. +In the metadata seems to be inside docProps +but i actually saw a handful of compile errors +I have been a bit lazy with the guide and unit tests but please find attached an update of the unit tests for the Kalman Filter and a first draft of a user guide to be included into the siteThe guide contains a lot of code but I wanted to give a working example of how to use the Kalman filter +The first patch I am not able to test the SSL but it was just a path change +The error still exists there +Ive used it a couple of times and so far its been a pleasure to work with +I have cleaned up and from above +Attaching patch incorporating the review comments and fixing the issues +Attachment has been added with description A revised fro mprevious patchNo for a YCBCR color model is required for bitpixel bw imagesThe Subtactive should only be set on subtractive images with bitpixel has been added with description patch against trunk +WindowsPreferencesValidationHTML Syntax sure what we can do to make this better. +Using I see that these selectors have assigned ids already and I find them consistent +The site SVN was always done Gav finally got the push to the www servers to work so it should be live now +I think this should be fixed as invalid because its not something that can be fixed in Derby +Please report back here if that is not the case +Ive detected additional problems with GCCC Looks like there is a mess in using static libraries I had errors for not resolved symbols from port logcxx APR +Just to clarify Create a Hello Cordova sample web application that can be used really means we should add it to all the platforms right? Wasnt sure so I just tested it without adding it in +DavidNo we dont need to worry about localized documentation +The issue was happening when the plugin didnt exist +STARTTLS has also become synonymous with enabling TLS over an existing IP connectionSTARTTLS is used by SMTP to do the above +I get a little wary of all the changes happening in this area given the testing situation and the outstanding bugs +Otherwise we simply return the image itself +The client abandons the entire block fetches a new blockid and block locations from the namenode and restarts flushing the current block +It seems like Castor XML silently ignores any parameters that it cannot match to a Java property +defer issues to +maybe that fake protocol handler is actually a sensible idea for such tests +This makes it difficult to implement because we use a SAX parser for parsing the poms +Bulk close for Solr +An admin could select images to permanently delete +Most fixes committed in Git ID further fixes with commit dadecfddde. +cant parse erroneous date Thu Okt GMT +lgtmI just committed this +Yes correct +This brings me back to a question I have posed to the User form in the past but have not noticed an answer +Having recently updated geometryless to use the new interface recommend updating postgis now +I happily assign away all rights to it to the project for them to do with it as they wish +That keeps the underlying architecture correct just with a different face on it +See for further details. +Oleg +Im not familiar with the implementation but I would think that the actual method call can be marked at compile time to limit the resolution to the given way the variables wont need to carry the extra info with them and the impact on the runtime would be minimal +Looks like this patch has already been m closing this issue now because the functionality has gone in and its not going to make it into the list of fixed bugs. +the problem is that wsdljava mapping process removes the char from the generated interface name and were looking for this interface so we can generate the impl class from it on the tooling side +Implemented in r +The patch should be applied on the project have not included any version number in the name of the descriptor +Try also Updating to the latest available versions of Facelets JSF libraries Check if there are any parallel AJAX requests +For if the message cannot be removed the message may be redelivered next time the server is up resulting in duplicated messages +This bug still affects encoding of application schema responses +I will a also look into enabling the tests for both CPDS and XADSI will see if there is a better place for the testsCommitted patch c to trunk with revision +Submitting patch +Is that all that need to be done or are there other considerations +Maybe you could give the latest build a try +Looking at the trace it seems related +the build succeeded in the release +Though the can be solved in other ways the order of creating Listener and Emitter instances can solve the problemWe want the to be created first so that the can connect to it +Adding a single method is a much better pay back in terms of complexity +As requested JIRA to track down change in CP branch already uses that includes the fix. +Ill say thats the technical point of view hmmm and maybe you can use it to veto this particular change +But you are welcome to upload the patch for our reference +bq +I vote for using exceptions internal to the to keep internal interfaces clean and elegant +This is the agent is the JBN server believe Im seeing something related to this problem as well +patch looks goodOn a second thought throwing an exception affects the authentication protocol. +It would be nice to get clean public regression runs before considering making a release +Its also possible that this issue has already been addressed in subsequent versions +tried CSA function and seems the plugins were properly categorized and I can use key words to filter expected plugins +Is the basic scheduler node there and only the backend not implemented +Proposed patch +Correction for contentxdocstomahawkAlready corrected in wiki +Hmm that exception doesnt say anything about the actual root cause +Unless we wrap every write to HDFS inside a retry loop we have not gained anything anywayPlease speak up if you disagree +Hack not needed +Unfortunately Im not familiar with maven compiler plugins +I also saw no other option than to doan automatic code format in my eclipse to get rid of all kinds of tab and and organized the imports +The is instantiated the is added but onInitialize is never called on this panel +The CLI only should do it explicitly +Hi TomA couple questions we usually ask about data corruptions Do you compress database tables? Is the disk enabled or disabled? Did anything unusual happen before the corruption +Defer to defer open issues to defer to not essential to address for +Do you need command line access on or is a Hudson web account enough? Normally command line access is only needed if your build needs custom tools that need to be explicitly installed on the build server +needs to be applied with git apply directory +What would my Action class need to look like to access the properties that were configured from +Committed patch to trunk with revision +i tested it here and it seems to work fine +The upstream issues was fixed and will be in EAP ER release. +You can try it with Q +Cant reproduce me MRI and behavior is note that on Windows temp is treated as a directory relative to current you start the script from different disks for and MRI you might see different results depending whether temp directory exist in the root of current paths on Wisdows are discouraged since they dont mean the same thing as on you please your environment and comment +Closing all issues created in and before which are marked resolved +Archiva RHEL Tomcat Oracle patch that adds filepattern to the DEFAULTEXCLUSIONS so that +Can you please post both files as attachments here +Verified on JBT. +This issues has been addressed and should be fixed in the latest nightly build +Actually its a new daemon Thread for each ping +I guess if the asset is archived and deleted from a package other than Global we should leave the instance in Global intact +Theres no CAPTCHA not even an email verification +ben Can you review this +Building Geronimo after having patched it +Right Im still convinced the trap only impacts committers writing unit tests that compare againstThe patch seems to have a very large amount of changes for such a small thing +Thanks Krista nice job +If you like open a new JIRA to support a Problem view for this problem. +Thanks +New patch adding a Pig property that allows user to turn off this feature +I committed the patch to +NET projects +New option non controls this +Theres periodic compaction checker and theres compaction when region is opened + +Thanks! +BTW the reason for locking problems without the changes into test test is that the test failed because it didnt get the SQL State it was expecting but as a result of failure the locks held by the test remained on the tables and later when we go through the tear down process and try to drop the tables we ran into locking issuesIn any case this issue of locking is not a problem anymore because I have changed the test in codeline in my clientAs i said earlier I will soon merge the changes into and then close this issue +May I base my testcase on example? May I create a small class that use the defined entities +Because we depends on the fix will be available for AS and fix is available for AS and above +domain is added as per +I think theres value in having a way to stand up a subscription to a topic from camel +Can you check if you have false? If youre having ACL problems you might want to try and turn it to true +A task timing out is a bad thing +Yup Ive changed the compile phase and noticed things were fine tooI havent added tests yet thoughIm wondering if there were some reason for the canonicalization phase for being chosen though +Scott what version is this fix in? Also please mark this issue as addressed rather than rejected +Thanks! +If a style is being specified in a widget XML file it is because the default style is being replaced with something else +On it works partially. +Remove the test scenarios for testing ignoring APPNEWSAVED at FailedKilled state as that should not happen +I do think the exception isnt very appropriate for an issue +Verified by Mikhail. +Committed to trunk. +Good to know the background +for maven directory convention and no maven build files and using ant +Im going to bundle one up as soon as I hear back +Otherwise Ill do it +I am assuming since you are running of Tomcat you didnt have to endorse anything that I have to with WAS! Yup looks like you have everything in there +Thanks Harsh! +please review +Added an ending code tag +Seems to be a different issue +Moving them all to the logs directory +how to mock the graphics object +Based on the previous discussion Id like to propose that for now we expose a call that allows clients to open a ledger for reading unsafely in the sense that I explained above +After patch is applied tree should look likethumbnail +Set fix priority to fp +Minor refactoring of storage and integration tests for feed entity parser with table defined +Would be nice if anyone will tell me where to find templates for this tag in other template languages such as Velocity if they are exist +These fixes also address +Hi Gregexcuse me even the selection handling is solved or should I open a ticket for itThanksSandro +I can see that parkunpark would be another solution and possibly nicer to implement with native code +If the user declines to download RJB is Buildr still useful +Closing. +I have run all the tests and they pass without any issues +etcI chose to fix this bug mainly as an exercise to familiarize myself with the work process of submitting a patch. +Fixing this in time for is riskyDefer to +Patch that corrects the functionality +Your proposal sounds correct to me +No test case to allow for reproduction of this issue +Thanks Justin +I separated the base bundle list into two partial bundle list First one contains bundles for framework and the second one contains common Stanbol bundles +Added a test in r. +Please dont commit to trunk as we are in the process of releasing Camel so the trunk code must be stable +Thanks Koji. +are you able to run jconsole to connect when it happens and report on whether memory is an issue in one of the heaps? Are there any other exceptions in the Archiva logsIm not entirely sure if it is actually a memory issue I have a larger repository the default memory settings and run for much longer periods of time without could be related to being a service though that doesnt explain the issues seen by Wendy and you experience this with a previous version +Part of a bulk update all resolutions changed to done please review history to original resolution type +Connections are stored by factory. +Addendum integrated to trunkThanks Anoop +It could be used where we want to take snapshot of znodes etc +Committed revision . +For example the neural net in can be used for it +Absolutely +Im not at work where I have the testcase and its basically just connecting to setting a small area as spatial filter and calling table statistics on one layer +The patch fixes the synchronization for the renewer thread and has the test case shut down the threads before the namenode +Instead I found a probably related problem with the Pascal distributionThe integer distribution patch for this issue still isnt ready +I tried to select the JDK really I did and JIRA doesnt give me permission to fix it up now +I will be reviewing this patch +Anyone picking up on this can just ignore it if they want +When example is selected button Finish is disabled within is not installed within JBT s possible to download file from URL defined for this example and content of this file looks correct +Talking about minutes to assign k regions across servers currenlty vs to minutes for this patchPatch has concurrency issue and will play some more with it but seems like zk is bottleneck all the state changes that happen for a region assignment +Issue resolved! +I will give up this issue to someone who has more experience with Apache Mina because I couldnt solve the issue +I forgot to mention the deadlock occurred in +Is there any chance Derby can choose to optimize a unique index over a one if there is a choice? That would seem to be a valid optimization no +Patch should help thatThe existing code simply allows fake to take precedence over h and h over z + +Ill look into it now +Treating required as having a value of true doesnt sound correct to meIf you consider the error message that should be used for the case of forcing the user to check the box it should say you must check xxx whichis quite different from the error message for required you must providexxxFor the former use case I think an validator should be used whichcan provide a much better default error message +Ill bug him accordingly +Shouldnt the Ui be essentially the same between each of the generate entities +Testcases for ldaps and including a nasty hack to access the received certificate +It would be nice to have my records serializable without doing anything specific for it with AvroAnd havent looked at other record types but if making records serializable is as easy as letting them implement serializable wouldnt that be a good idea +Username bstansberry +Running test +Either way metrics are collected as expected in HQ. +The best thing would be to have a cascading configuration system if exclusion is configured at project level it is automatically ported to each tool but you can still make exceptions for an individual fully agree Fabrice this will be a mistake to handle such exclusions by tool. +Existing tests still pass +Ken would you mind giving me a hint? +I just discussed with Andrei +The same is true for +I dont want to write basic sql steps to fetch data from database +I also added a unit test to make sure that the comment field is getting set to the MIMEs description. +Not sure why this would be happening +What Im looking for here is some action to take if a user wants to run a +But for that a new issue should be opened +Im going to create a separate for the javadoc and close this issue the most important aspect was to create user documentation. +Something with jffi I suppose perhaps to do with the actual error Unsupported platform sounds like it could cause problems +This is the same issue recently reported on the list +The new patch makes them package private +All builds updated marking as resolved +Hi MarkusDid you have time to look into this in order to come with a patch propositionThank revision . +BruceI will encourage you to ask this question on Hortonworks mailing list which is the best place to ask questions about Hortonworks distribution +Feel free to correct it further if you think I garbled itIt may be that your classpath did not include the class tree generated by the Derby build +Ok Ill add these +Committed revision +further investigation should the dependencySet be added much like the assembly plugin +The patch is ready to be committed +bulk close of resolved issues from releases prior to +Nadyanew doc from the DRLVM developers guide and the DRLVM index page underJIT docs +No the unit test just needs to be updated +This is because my customized settings. +Because the run method will access these datastructures and modify themWe actually dont need these checks because no where is a task duplicately added to the data structures in our code +I think this was a valid problem with the documentation so I dont think invalid is quite right +Created by IDEA +However core tests are broken and no Herriot artifacts are being created +Liktha do you plan to work on this Wei let me verify this in my setup +Let me knowI also fixed an unrelated doc comment if thats ok +So for that browser I have to delete the http part before pasting +The artist is making some final edits before we commit +What is needed here seems to be an list of permissions +Is your logging at FINER or FINEST? FINER should be fine though I imagine it will still produce huge amounts of logs +For example what form will the interview take what area of computer programming are you interested in how long will the interview take? It might also make sense to provide a little background about yourself and your goals with respect to career +Hi HowardIll create a new jira for this and try to identify the messagesThank youMirek +Real implementations are delayed to a further release +Rebased once again +I think even SVN modules that dont follow the convention are already broken in this regard +Fixed by +if you fixed it how come it still happens +Here you are clashing with yourself in the same file a very different code committed +Additional tests committed at r. +Please check your error log +It is a workaround because this is not a solution since it only works in one specific scenario +This is first initial patch for people to review and play withit was done against the trunk as of March into the mainWe tried to document the code as much as possible Adriano and me will try answer any questions as soon as we canSamples using the new APIsOld testcases running againstthe Compatible Wrappers using the new Flexible Query Parser as backendThe build and targets are working fineThe test target is failing in the I still need to verify whyIm also working on javadocs warnings in the new classes +to facilitate generation of admin hash +sorry i should have scoped my question better +yes you reasoned correct about what was the documentation bug Thanks for responding on the other side as well +Modified according to comments from Jukka +I dont understand what youre saying thoughIf those methods arent deprecated then developers will continue to use them +Dupe of +Because returning gcable tombstones during read should never be a problem since well skip any tombstone before returning to the clientNit might be worst adding as a comment than skipping collate is only ok because we have a so we dont inadvertently copy the idea when a slice filter is involved +Hadoop QA is not starting +really this should be a capability of the new reporting for m not fixing for m +I dont believe that we can use regexpattern if we want to stay compatible with the Java ME profilesSo we can improve the logic there but well need to resort to cruder mechanisms +This has been fixed as part of revision r +Koen can you provide me an information where I can find the reference guide for these components +Galder Galder Im glad you found out +Demonstration in and +There is much work to be done before there is working code +yep my intention behind support package is to indicate users that those classes are for internal use only +We should try and hit with this too since we are doing alias work anywayWe either need to drop too many closes as a logging error in or set things up so that we only close primary core names and not aliases +I should also have said that it is a only dependencyThe jar is only needed in order to process the annotations once compiled the annotations are made available in the resulting class files for products such as Findbugs to use +Thanks for the suggestion I have now added an Add as Bean Configuation File action to the context menu and it is only enable if the selected file has the Spring Config content type and is not already a spring config file +Most of what I added is about parameters +Heres the log from a session that shows the problem +Product constraintnullwhere Num HGNum rows selectedij select Num from app +Galder so your push already fixed the infinispan LC issueif so please assign this to me and I will fix the Infinispan bits are done Ill assign it to you now. +Tho I dont want namespace A to permit namespace metadata manipulation as thats where well store quota information +Variables in the binding are here to be seen as dynamic properties +We could solve this by copying the required property files into the project source folder but this is not what we want as it can be multiple property files and they are shared between projects +as per mail discussion details on how this was reproduced eventually to come have marked this issues Release Notes status as Not Yet Documented +John ImranI have just committed some changes that are supposed to fix this +were my intentions as well but I used the filenames to do that +This task is specifically for you have any other requirements for the SOA integration then these should also be reflected in other JIRA issues +On the other hand it might make things simpler if we told people that they should never modify transactional resources in the Processor or in other words all writing must be done in the Writer +Why no further releases? Has this published officially on the website? Lucene has a completely new API +ping +Patch integrated to trunk and +Sorry addition +Fixed +Why is relevant to this ticket +Delegate as needed +Well Ill be +What is going wrong with this setting? It was originally put in to keep the task tracker connection to the job tracker from timing out +minor update to the sql in the zip i didnt test the inserts before uploading +We had agreed that the would check and enforce this restriction via special case logic that knows the semantics of +ok I investigate +Can we use the anonymous channel to realize our usecaseThanks adds the option of setting an external replyChannel or if not set instantiates a new one +Reopening to close Invalid which is the norm for java bugs these days +Let me know if you need any help with the release process there +Also reproduced for all xhtml and jsp pages. +Whats the best approach to solving this +Meanwhile Ill exclude the failing tests separately +Attaching patch +File type value for sockets +Thanks Amar! +I guess that would be best done after the release of the next beta +If you do it via the New Wizard instead that is the default as far as I can seeThis is the default behavior of the library configuration in Eclipse facets +What I can do is to send you the corresponding code and you could add it if you think it could be usefull to otherss it +first commit on master I would like to have the same on and +But I am unable to reproduce the problem in a vanilla EclipseThanks project that can reproduce the exception but with some yet unknown change needed in Eclipse for it to do so +Fabrice is there a new version we should use +Axis pass successfully JEE jaxrpc testsI have not tried with standalone axis tckSome problems are remaining in interopNote this patch may be improved its just a first try +Instead store your big data in the body +Fixed in the trunk +I am working on Hudson setup so you can take aPeter +I dont care either way +This will allow different to be released for the two products and allow the version to be used to distinguish which product the CP applies thanks for the info on the default configs +Actually I took a look at the reference documentation +The best solution would be to make the maps somehow so that we can avoid that stage +The factory class creates another object that consults the model +This is lot for now but shouldnt this be dynamic? We dont need to malloc again for the configuration keys and values +Does this resolution mean that only is going to be supported and not ? We need to get a list of namevalues for undefined elements +New patch adding Version to as wellThanks for reviewing Ill commit in a bit +I did notice one very peculiar behavior though +Closing all tickets +Heres a patch that sets retry to in and lets to the retry based on its count +But I didnt check your scenario +Verified +Much of what it does either is or could be part of the debug interceptor +I indicated on the mockup the hex colors that need to be changes to modify the site +Then we can reopen this bug report. +sending messages +to as svn +just realize this is off branch never mind +Ive checked in a fix for this problem the provided report is now transformed successfullyIf you look at your reports in your cache delete the xsl in the cache and run a new resolve to get the updated versionPlease tell me if its ok for you so that I close the issue +I couldnt find where this happened +It is much cleaner +ThejasWas wondering if you might be able to get this a look to see if it seems reasonable +Perhaps the highlevel doc upgrading +No further commment assuming ok. +will into the dashboard plugin might be useful from their documentation it looks like this is a summery report +Thanks for pointing this out Andreas! issue is present multiple times in the reference documentation. +Have removed the test as this requires some changes in the samsa launcher +Ill investigate +Not sure what has happened here Ted it was waiting on you? +Theconnector will provide one extra argument option additionalexternaltableoptionsIm not sure that this a good idea +And of course Ill add the documentation +Ran all ant unit tests to completion +Accumulator and Algebraic are planned but not there yetInclude should work as long as the included scripts are in the class path but I have not tried it + +Hi Knut can this jira be closed +This is at best a feature request in terms of making the names more consistent but there is no blocking issue regarding being able to use the managed properties. +Test results with a prefilled cache of the test script was broken with respect to the no parameter so previous test results related to that one need to ignored +Attached with a stack trace showing that this issue is reproducible in AS . +That saves me or someone else wasting minutes trying to figure out which of the users you are referring toIt looks like Bugzillas user renaming code does not updated saved searches +Junping thanks for taking a look +Just moving out of fixed the problem +Pat Gabe could you have a look at it? Thanks +In fact extending the is actually what I would propose +Could no add reference +for patchWorks for me for windows x and Linux applied in trunk and as fixed. +it is not in the lineup for verification during it is failing on my machine +Id say that this bug is critical for the Checkstyle plugin as having the report in the site is its major feature +Applied at r thanks! +Its in the sandbox +took too long on JenkinsWe should add timeout annotation to it +Thanks! +Disagree with comment above regarding fixing in a more pleasant way + +remaining tasks are assembled in the related and dependent issues +The JSESSIONID cookie changes when the flash is modified which then messes up Shiros native session management +check retrieved service for bindings before querying for bindingsThe process to query current list of involves two steps query for the service information and then search for associated bindings +The reason for this is that the assignment of a start level is an administrative task and start levels of existing bundles may already be set according to administrators decision +Updating every hour puts too high a load on our server +Moisha Breit. +I include the patch with this bug resolved +PDF doesnt show the arrows that show up in tabs when all the tabs cant fit in the top area +Were currently working on the release which has been delayed for a while pending new release of both CXF and Camel and Id rather not do any drastic changes any more and make sure we can get the release out +Please check it has been applied as expected +Not sure what your last comment mean +Logged In YES useridWell then this is a tools issue that we can do nothingabout +Added new component to experimental project and added a few mustella tests forThank you Marcus +I shall rebase the patch with latest code as well as incorporate the changes recommended +That could be a red herring of course Ill try to narrow it down a bit +java files +Ok after reading I reluctantly agree +Im trying to reproduce this on a single node cluster but havent been successful so farIm surprised cluster size would be a factor given that this is an NN issueIdeally you would create a test case using the unit test framework in the project +Added a chained attribute renderer and moved pattern recognition to single renderers. +Its problematic to do throttling using a configuration +v patch will work with commits made by Bryan in +So every warning seems like a security breach to some of them +I will wait for the official version and now I useleveldb store cant cleans up all as the data is consumed from the store when send messages faster than the rate of sending messages the message is consumed when finished there are many not cleared +Donated JPOX annotations in the hope that they will be usefull in developing JDO following recent discussions on mailing listIncludes +I checked the code and connections associated with a global XA transaction can still be closed without being committed or rolled back first +Please verify +Its actually rendering the correct CSS class names +Hey Vinod have you had a chance to look at the latest patch +Any hereThanks +Even ive written some code using prototypei have to include jQuery for using a lot of UI s hundreds kb of javascript for a pageI think the javascript lib tapestry include must be small and a lot of UI can be like jQuerys Write Less Do MoreTo use a small oneand never change it to another dont want to leanuse so many javascript frameworks +Looks liked a duplicate for t the core with the higher version number have to be the leader? Im also experiencing this issue but the replica is only about versions behind +I will try to provide a patch in the near future and we can update the code as needed later +Reporters of these issues please verify the fixes and help close these both the scenarios mentioned in the bug on the build with commit ddfceadeb +Ill try that +Attachment has been added with description Missing +Simple patch avoiding the log for servlet nullOne could also log something different if no servlet was found but that is done by both calling methods anyway +If the problem persists please reopen +I am trying to get something for this out for the timing is tight but hopefully it is will come down to a weblogic deployment option Thanks have found the cause and solution of the lifecycle errors that users are seeing on weblogic +committed +If nothing comes up Id suggest we close it in a few weeks and we can reopen and create one again ifwhen more info is available +Contains goals for installing schema and enhancing classes +Does this mean if we apply this patch we should remove the LICENSE and NOTICE files from the various subprojects nowDoesnt the NOTICE file include stuff about licensescopyrights of dependencies? So it seems like NOTICE files cannot be generated can it +rolled into dev. +Logged In YES useridcheck marked for attachments and submitting In YES useridStill no attachment +On the Administration panel choose Custom Fields then Configure the Geronimo Info custom field +we need a test suite for +should use + +but the option of supporting a union sounds better than making a list contain objects of different types +I keep threatening occasionslly my to do list with this +bq +A good deal of experimentation andor potential schema changes will be needed to solve this issue or possibly using outside locks to prevent more than one stuffing operation from proceeding at the same time +If it helps I ran mvn clean verify on the source and the build was successful +These register a message selector that sets a filter on the request sent to the consumers +Reopen for committing to trunk and other branches +Is there anything planned on when this issue is going to be fixed? I guess this issue is a bit underrated in terms of severity at least in respect to the reporting date and activity +The patch for this issue will provide enough synchronization to stop the CMEs +We need this working for Dinard +Aurelein I dont see the Java changes that you wrote here +I thought maybe it is better to keep the analogy +All set +mrepositorycDocuments and Settingset +This was cloned to represent the fix for CP +There are lots of reasons that a government work cannot be included in Apache not just copyrightPlease explain this +I just committed this +The patch looks fine although Id like to get rid of the sample +Welly welly well that looks easy to fix +is this piece of code missing? Can you provide RCA for this? Given the timelines we did not regress several areas where impact is less and there are no code changes for a while +The difference in size with the prev patch is mostly because of using git to generate the patch +Looks like were looking for a version of rspec greater than which is pretty oldspecgem rspec Roger can you run gem list rspec and paste the results back? If you dont see any output then you may not have it installed at all +screenshots for and chapters are added for chapter three screenshots are screeshots should be were screenshots in chapters and are updated +Verified and for other usersAvoid the use of class navigation for the surrounding DIV or you have to fight against CSS +I think this is a blocker for a release as a TLP +If one of them gets deleted it would not be able to return a valid responseAhh sorry quite right +Ive created a sample application that demonstrates the issue +So it was changed to return a list of exceptions +i dont know if we are using that or not +No point in keeping this open even though it was sort of a bad idea. +Patch attached +Resolve as duplicate of +Iow votes please +I think this is the only reasonable way todo itLet me run some tests and Ill commit +We probably need to extends some of the types to pass along the taxonomy information so that we can enhancerebuild the Document in the using the +knutthanx again for the commentsI see ur point +uses +without the Makefile how do we regenerate the protobuf files +Would be nice to get the integration going at the earliest +Updated with the reader +cluster power restart event where all started nodesclients start hitting the booting services in sync +I agree no separate JDBC implementation should be necessary +What do we have to do to add it to the plugin list of +As you see they have ve default value now so it will not be correct to call them REPLICATIONOPSAny method with CUSTOMOPS attributed will be handled with it +OK thanks +bq +Thanks Tahir +The web interface does not track renames though the actual archive files will be moved at the server +attatched simple maven project +The exception was seen at server startup not sure how much sooner I could have seen something +test to set the Fix Versions field to JDO beta. +I looked at the site briefly what I could see is a download for a server side software +Closing. +Closing this issue. +Please update it and I will apply +This issue was discovered while investigating +This patch closes out this issue +works on stagging so will be rolled out in next +I will make the changes to the version you have and post it backOtherwise I will shortly be committing this fix to trunk and x so you could replace your version with the latest version +Im all for fixing the cause by serializing the operations as well but I see downside to this more robust fixOh and btw shouldnt the get copied in Garys patch since its being processed asynchronously and changes to the object could lead to Null Pointer ExceptionsRobust please +So shall we revert the behaviour to +This shows the help icon in the upper right +what i menat was +then you have the potential situation of n keyspaces with a cf nameNot sure I follow could you elaborate +Is this still an issue? Here arent any details so Im not sure how to reproduce +Now I see the could be global since the instance is an unique key Ill have to look into another way of making sure we target the correct Bean +Final patch +Simple quickstart reproducing the issue +I created a local build this morning and it works fine with that +Thanks Aaron! +The fix for dual on the classpath smells bad though +Tests are not needed +But for now is not an independent component and renders it so couldnt be inside iterable components. +Fixed in and inWill be released with +sorry for not checking +If no one else complains it is fine with me to have it as you suggest Im uploading a new patch +Is this a Windows problem does it still happen +Hi guysId like to go ahead and fix this +Add a to params only if is not empty +Thanks +I am happy that my suggestion was considered significant and was accepted for implementation +Problem resolved for WindowsUbuntu using fails still since its an unrelated problem ill open a new ticket +I think its a good idea to have this feature but constrained it to the dev environment +I realized that the patch was not a unified diff +The likely culprit is the move of the database to a dedicated machine +Ive added a patch with unit testsPlease you handle the improvement of outside the scope of this ticket +Why cant you install the Activiti Designer via normal online installationCan you check which version of Graphiti your Eclipse has installed +I attach a first version of my patch +I can test for end of next week at earliest +Attachment has been added with description patch file for combined +This was a lot more complicated than I expected and I found several bugs in the original implementation please try now +Attaching a patch for this extracted from Aruns patch on +You need to store the epoch persistently on disk to handle the case of journal daemon restarts I think +I cant see that it is used anywhere and any objects initalized with it will throw as soon as anyone tries to use them so Id suggest removing it +Yes I have the bulk of the code done +It would be nice to get some dialog started on how to fix these clear function problems and make it all cross implementation compatible +BTW Im using Hibernate Core Hibernate Annotations Hibernate Validator Hibernate SP JDK u Tomcat will probably address that in Hibernate or someone please link this issue to the bug that it duplicates +Please punt any of these that you dont plan on doing to could I get some estimate when this is going to be fixed? This problem is blocking testsuite development +Mridul for the contributionYesterday I had put your changes in trunk and today it is back ported to release branch at rAlso closing this issue +It is not yet in new logical framework. +I dont know of any wayto cause this in a correctly running current system +Id like to commit this shortly +Fix to +Fixed at ad thanks for the workaround suggestion. +Why not introduce a new config which defaults to something like minute? In the test case shouldnt you somehow notify the servlet to exit? Currently it waits on itself but nothing notifies it + +Ive changed the logic of my program and I check the consistency of the view query result and if in one row the attribute doc is null I repeat the request again up to timesAfter doing several tests I found out that at most I need two retries to get a complete valid result +This has been fixed already. +So permitting deletion is not the bug +We want to do same for? Does version of hadoop even matter for this issueUp on dev list discussion had it that only can require hadoop as its minimum version +here is stack trace +This xsd generates a structure that the test uses +We have several components we share and using imports allow each of those components to contain their own settings +DOCS ZIP examples removed from docs zip and emailed to Permaine and FernandoThe examples will need to be returned to the docs zip at a later date +Can you try the attached patch +Patch checked into trunk +This may cause issues if instances are shared between multiple threads +done +But for null value we do not have and hence no such conversion of data type happens +This flag is set from the out handler when it gets rampart as a engaed module from service +Presumably so does the REST api +A potential error will be caught anyway +Fixed +Beta is gonna go in about two days +Removing the fix version tag for as it fails the same way on +so do is to avoid close ledger until all previous requests are finished +Youll have to build the latest HC snapshot from sourceOleg +If you find further issues please create a new jira and link it to this one. +These are all committed in +This feature can be implemented in ve written some merging code in but having some problems with closing regions and regions in transition +That also doesnt include job overhead +Aidan Ive made those few changes + +Posted recommendations to user on the forum +Any ideas +Thanks for keeping an eye out on this Prasanth +We dont commit during the day +Browser dependency maybe? Im using FF maybe the server was having troublesbeing upgradedrestartedwhatever +Test and solution in different files +Yep works great +The sample case You dont know if JPA specification allows mixing Embedable annotations but Ive found Hibernate allows and support it in any way its useful behaviour +I think David wants to wait on that stuff and do a big refactoring all at once after the geocallaborator stuff +after release. +Thanks. +I guess this is caused by RMI different +At least in theory +Correction the justification is for decimal char varchar so for a combination of stringnumeric types +After updating and rebuilding the problem does not happen on Jetty any more only Tomcat +Hi Ken this is exactly what flashed through my mind when i opened the ticket +DaveIm no longer working on the line +Ill try a different way on checking the repository if it has been scanned or +I think a few small modifications will be required but I should be able to handle those +Currently looking into this with the bundlor kindly upgraded the version which picked up the new bundlor which picked up my manifest parser that dm server is now using +I stand corrected +bq +This is the patch that adds the key attribute to the sslContext element that can be used to pass the keystore key password +Thanks Aaron +but itd take a bit of work to pull it around +If you find further issues please create a new jira and link it to this one. +I have a fix for this bug which I could check in immediately if desiredThe fix generates a new federation tag at startup to be used if management is disabledThe current code declares a stdstring but never gives it a value if management is disabled +Committed revision +verified on deltacloud +I had some trouble with my test config the other day +Note Attaching a debugger to has been known to make the problem appear to disappear +If you have insight into how to hack the discovery site so it can display something other than Certified feel free to put it here +step of Udais reduction where BNF gets reduced by updatestatement and deletestatement +Do you see these locally Ted +This patch has been used to build the following configurationsshutdown The modules are not fully transported + +The use has been removed. +Adds a and a fix +Dave were leaving this one to be documented right? If so please update the JIRA +Another observation it works with STATETRANSFER if we use the TCP transport +I ran the program with +Attached is a patch to the current trunk implementing the change that Oystein suggested placing the trace message under the control of property +Hey Guillaume is there a snapshot of pax logging we can test out? Thanks +I understand its not easy to sort this all out but uploading Camel artifacts to Maven Central that cant be used without configuring additional Maven repositories is pretty annoying for users +In further experimenting it has nothing to do with special characters as I can name the host without special characters +So if the fix makes it into the next normally scheduled release Im happy +I just committed this +When HBASEHOME is not set the test works fine +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +In this case will throw but our implementation will still throw hiding original cause of the failureI was always curious why javalang has ambiguous contract for user class loader and for bootstrap class loader +We may need to add a layer to transform this OID to use the first alias or may be we can use the upName if it is stored +Updated new parameters. +I will see in the next version +Chris Glad I caught it +The will contain the should solve this issue +Committed at subversion revision +Simple text patch please apply Thanks +Hey Gary as a BPMN nb can you put together a sample or more concrete reproductive steps so I can better understand this one +After reviewing the patch the overall codebase and the abstraction in question I think we need another JIRA specifically to refactor temporary file usage with abstract contextsThis particular patch applies to a number of different temporary file allocation patterns from pathname only to filecontent +Spec Section JSF has the concept of an active Locale which is used to look up all localized resources +I think it was taken out because checking the whole stack every time we called another template was computationally expensive and was costing us a significant chunk of performanceMight be possible to reintroduce it under a conditional to be turned on if and only if you think youve got a recursion poblem +Ill tweak the patch and submit an updated version this weekend + +See Bernts response on October +render the unicode text as jsp means jsp will render unicode text such Simple Chinese as it is which can read by human even it is encoded in property file by +Logged In YES useridsorry that i didnt note the version of jboss we are use jboss and there is no classpath tab +We will test more +Verified by Sian. +Already fixed in a later version of. +It works! Thx for your great work +Patch committed to trunkOpen to track hbase hadoop maven artifacts +Getting pull request together for. +Sorry for the inconvenience that caused +I want to be able to easily run some tests on a real cluster hence exposure via Tool interface +Since it was already on my machine I added and committed the missing portlet in rev +If it is shutdown is aborted +Originally the output level is INFO then the output level will be set to DEBUG in the custom logj property checked the the DEBUG level message did show another simple way to do the test +See and Struts Demo +Would it be possible to slim them down so more of the content is above the fold? +Then the start command does nothing because the code is already running +Therefore Ive opened a new issue I am resolving this blocker issue for since the samples are working now. +When in and fixed this at the same time as because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It passed +Sorry I pointed you to wrong build +What we see is that spi references builder and impl also refrences builder +Alternatively you can modify your copy to request just a few items at a time there is a variable in at line you would need to change to try thisIf something doesnt work it is likely an exception will be thrown +bq +Currently it is not. +Removed the logic to override the accept mode value based on the Link Reliability value and added tests to test the rollback and recover casesAlso removed the UNSPECIFIED Reliability enum from the Link class as its no longer requiredFinally removed part of the that checked for the exception being thrown when the a Topic with is specified for reliability +Too late moving to unresolved +If there is a problem please open a new jira and link this one with that. +Attached new patch that addresses problems in the previous one when running from a directory containing spaces +Any suggestions? What Ted says on delete cfTBD +Yes +Im going to comment out the optimization until better solution is found +Can you check my proposed fixI couldnt really make the binary compatible +Built in brew add this patch to native daemon fix and will be part of any idea when will be tagged? Just wondering if I should wait for the tag or rebuild with the patch as I have to rebuild anyway +Fixed in trunk. +Marking against and moving comment I put in duplicate issue here Made this a blocker +This enables S file and foo all workThe outstanding issue is that out of tests fail +It might relate to sort stability issue +Closing is great but what about people who really do need a suffix? Ive got an external system that needs me to write with a +That is texthtml comes before text +See closeInternalIve wrote a test for the NPE +Changed to send the kill event to app and make app in turn send the kill event to attempt +Thanks for the patch Knut +Delay proper fix to priority +This alert should be also available at project level Rename Notifications into Overall notifications Email option should be selected by default +Applied to TRUNK + I know that Ivy attempts MD SHA verification by default but it does that verification against checksum files located on the server so it only offers protection against corruption in transit not against files deliberately modified on the server +Better patch more dynamic Version map creation improved warning message copied from Lucenes Javadocs on +But maybe the data set is so huge with the blobsthat there is so much work to be done that it does not allow the drop to happen +Any comments? Is the ideia good? If yes I guess it is simple to implement it +Isnt this a duplicate of +both relate to the same was a duplicate +I see some issue with test but not with concurrent read operation +Jason I know this book is enormous but please see what you can do +This is a problem that i could not solve and is also a problem with the standard HTML parser iirc +However it would only work for declarative transaction demarcation not for programmatic demarcation but that would probably be good main problem is that wed need a way to find out whether a given supports transaction suspension +Is it worth creating a low priority JIRA issue to add a summary of this command to the help text +I applied Jacobs old resolved issues + for the patch +I am not able to apply the patch myself Forbidden +thats what I just wanted to suggest I will come back to you later +Similar to s the patch +This is already on the sources +These same failures are showing up in a number of the tests +Added cast that I hope will resolve those last compilation errors +Wouldnt be acceptable if the streamed logs start and end with a AN OOZIE INSTANCE IS UNAVAILABLE AT THE MOMENT LOGS MAY BE PARTIAL +to save the file names read by this class it takes them all and puts them in a string separated by a separator in this case to restart it then needs to turn that back into a terms of the resource pattern resolver Ill look into it and repost +Id just like to minimize the actual need for that ve added public marshalunmarshal methods with Input and ReaderWriters with and without a parameter +See +not ideal but should unblock for now +It is very old and the code base is no longer maintained actively. +the ant release notes state Removed it is now merged intoSo Im updating to just use plain ant +regeneration failed when testing this feature +Also will disallow deleting of a build definition that is in use by a template +Please rebase the patch to latest trunk code and resubmit to QAnit Remaining +This is an implementation of the issueDeallocation of C structure was temporary commented and can be uncommented after State Transition refactoring completeBesides some synchronization and shutdown problems are fixed in itDont forget to delete vmincludeopen from SVN repository +Changed Target Version to upon release of +can probably be after release +Just an update to my previous comments +Hence svn stat does not list these files and hence they are not removed +try actually seems to work with. +This is a change in the public API +Ill commit shortly +Builder changes applied +Implemented the fix described above +Looks good +Pending resolution +I dont wanna refactor in here +Running through Hudson +Creating a reproduceable test case helps focussing +It is now on maven central +I just committed this +This patch is not ready for commit but I wanted to put it down for review +Also submit at trunk revision . +Include logic for local mode +So if we have to go that route then no big deal +Pushed to master in ad and in fb. +Please confirm the current status of this issue and if its incorrect change it as required +String SOCKETEXCEPTION String COMMUNICATIONERROR String CONNECTIONFAILEDONDEFERREDRESET String NETINSUFFICIENTDATA String NETLOBDATATOOLARGEFORJVM The following would remain as they are DRDA protocol exceptions String NETSQLCDTAINVALIDFORRDBCOLID String NETSQLCDTAINVALIDFORPKGID String NETPGNAMCSNINVALIDATSQLAM String NETVCMVCSLENGTHSINVALID String NETENCODINGNOTSUPPORTED String NETNOTEXPECTEDCODEPOINT String NETDDMCOLLECTIONTOOSMALL String NETCOLLECTIONSTACKNOTEMPTY String NETDSSNOTZERO String NETDSSCHAINEDWITHSAMEID String NETPREMATUREEOSDISCONNECT String NETINVALIDFDOCAID String NETSECTKNNOTRETURNED String NETNVCMNVCSBOTHNONNULL String NETSQLCDTAINVALIDFORRDBNAM Does that sound correct +Jesus its one big guess gameIn SMX the have selected packages commented out +But it is a problem of the user as I see it +Closing the issue now that it has been resolved. +But I think that would pay changed the code with the help of Ants comments +Vinaya is this still an issue at all +The only thing which might make sense at file level would be to check the size of the file but not the complexity +Unfortunately I can not fix the appearance of a second window with an unknown error +Absolutely no fickin way are we going back to that broken stuff! There was a reason we got rid of it +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +If it doesnt then please reject this subtaskServers View is already included to Perspective by default + has the same situation +Attaching still have a rejectionpolicy of run As I mentioned earlier this is not good and can lead to deadlocks like yours +As suspected the failure was result of a partial shutdown of region server expired in the previous testThe region server thread of this expired server was still alive when the next test started which lead the test to believe that enough number of are availableThe new patch take care of the problem +Reopening since the problem has been highlighted now +Thanks BenWe need to update the doc per tomcat native clustering +Did you mean to remove the patch again Sameer +Patch attached +apparently there is no commandline maven on every ubuntu machine for Jenkins builds +There is also no way in which the user can workaround the problem of bumping the priority of his jobs and getting them to run short of killing other jobs ahead in the queueGiven these points we think it will be a good idea to sort the jobs in the running queue as part of this jira +Thats why we ask the affected users for feedback +Patch added +AI as print shop is not able to well print it as it looks different as the one in pngs +Also I renamed to be more descriptive it is a +It much help me to inlining +The rest looks good +Tied up with at the moment but I will have a lookthink when I get a chance unless someone else jumps in first +aj file should not be edited by a user +shouldnt be placed in the instead ofat least these was processed in I just missed that is placed in for and mainly because I see in +Thanks Todd +Hi Keith +If all tests pass will convert submitted test to jUnit and commit code changesIn summary embeddables were not being added to the L cache as the result of a query so they were not processed as part of the detach operation +So looks like thing +una la que vulgueu +Important but not a bug +I only ran into this because of some of the trickery Im trying to get working +doc review has been marked Submitted so OK to proceed on this bug +svn ciSending bundleTransmitting file dataCommitted revision . +See the mail I just sent to lets continue the discussion there +Maybe it was since very old days +Could you provide a patch for this one? Then I will commit it shortly Thanks +Marking these as fixed against rather than against +Will have a patch up in the next couple days +really +problem is still there +Thanks! Ill take a look through this +Working ranking neural net less the sparsity enforcing part +tdataList iterate over a list without check for rendered property on the component +I am out of the office until Thursday July th +If there is a way to do that somewhere in the build process that would be ideal so this issue wont creep back in on us +Escaped back to the computer +If others feel this makes sense I am okay with it +Since youre the mantainer the rules say I have to provide you a patch +What is the purpose of src if youre supposed to use d and srconly +for all values of attachTiming attribute +Koencould you please review this patch and check whether everything still work fine +Can someone help me to create a test case to simulate the process of the reducer by perhaps writing a MR job with an identity mapper against the data files attached? Thanks +For I believe we seek to the beginning of the start column +Can you give me a sound answer why it should not be added as this is general information that should already be available anyway +jBPM BPEL beta is proven to work in GA which includes. +Closing issue as it has been released as part of release +However I believe those classes should be in reality abstract classesAnyway I looking forward to hear someone elses opinion here tooCheers be useful in the multiplexed services but this seem to add an unnecessary abstraction +I dont think we should deprecate the group declarations +Do you think you have time to backport this to +move to doc version for less overlap with engineering tasks +repo Committed revision Uploaded to repo Committed revision +recap all PDF rendering is broken in the branch and this is not related to invalid XML characters but has something to do with the macro screen rendering +Hello Dejan +Added Matt to watchlist +retrying hudson +Im attaching a patch for so it throws as the patch has been appliedMiguel please verify. +If you think the above is a good next step Im happy to start working on this either later this week or next week +Incorporated the release note text into the Hibernate Known Issues section of the Release to be as resolved in RN +product or patchI guess to have it apply for BYOE too +HiIs the fixpatch available somewhereBrYes the fix was committed and is available at the SVN head +fyi filename doesnt end with +The checkings have been done in process before add is called +Thanks for reporting this Michael +But I dont understand why it would work for some other projects? Maybe coverage is simply disabled for themI should try to reproduce it locally and see whether I can easily see it failing +most likeley latest changes are not pulled in for testingResource Does not exits error when adding GSLB monitor is handled as part of commit dcafecdbbeccecfa +We also dont open it until few cache misses occur +The default value of the internal event mask of newly created IO was not correctly initialized which caused the READ interest bit to get cleared in the interest op queuing mode unless the event mask is explicitly reset +It might be better to add an abstraction layer +Is that ok with you +Merged to the branch in revision . +Ran for sanity +Patch which addresses the issue +I think we should not deprecate but break here this is experimental and super expert +I have run five times against the sane jars +Thanks Martin this setup and cleanup content must be copied to the then +The tearDown method in is setting to null +Ill incorporate your changes when I get around to the full patch +Email from Jonathan reOn PM Misty wrote However what would be IMMENSELY helpful is for someone to go through the Installation and Administration guides and highlight differences for mewell the JTA ones are simpleInstall Guide identical to core except for using instead of as the name throughout + +Sure we should move to you propose to incorporate the WS tests in AS releasesWouldnt the easiest thing be to replicate your hudson configuration in the QA environment +Bulk move issues to and +I wasnt talking about the name of the parameter but about the comment in the javadoc +for Jakes patch +I like this idea very much have been wanting to have it in tapestry for a while +Includes a brief discussion on encoding issues in systemsbonus patch mismatched closing tag on srcdocumentationcontentxdocsdocs +has been marked as a duplicate of this bug +I rewrote the jmx registration a health check is now only registered if it sets the property +conflictres and node now pass +Additionally on Windows program need to be in PROGRAMFILES by product according to the Designed for Windows XP Logo specification +The problem occurs in MVN +Excellent Andy +Final version of patch committed. +Random thought Could the be an actual implementation backed by a resource bundle? We could provide the default English phrases for the catalog and users would be able to their implementations +Deployed to on +But after refreshing to latest build the same fsimage and edits seem to bootup the namenode without the exception +Are there any chances that this bug fix will be backported to the series? Im currently tinkering with this problem in a project where we have Spring with Spring Security integrated +If we didnt have this limitation then would this change still be suggested +defer remaining issues to + introduces a wrapper class which shall be usedto return a subset of the underlying Blob object to the user +Post the exception and stack trace +Can you provide a PATCH for that? +However sqoop is not database engine so this fact is unknown to sqoop +Faranaks issues are gone with +Link to the dependent issue +release bulk close +We could do what Chad suggested and handle interfaces within Thrift that return null with a boolean plus optional valueRe +Same problem as with Spring +Use of to get connection works fine +What values did you use when filling in the create fabric formAlso I cant recall if I had this by build did the page transition to the container log page? I think typically once the fabric is finished being created you should see the app refresh +If all of the goals are showing deprecated its probably a bug in the help plugin +If we wanted to make it I would be okay with drawing a line that says example code can go in contrib everything else should be in core or spun off elsewhere as its own project +Thanks Dave +with mmap mode i dont see thing that I cant reproduce this behavior and dont see in the code what can cause it can you please provide steps to reproduce +Committed to branch and trunk +Why cant you configure preview to set translator properties? This goes way beyond just this specific property +they do not fire any property change event +flushed the browser and still i have the same issue +We need advice from experts +That second scheme is not to use failover +Demoted priority +if I use methods like dont work into jars +See comments in unit test for more detailed information +Verified on EAP ER +Are you seeing this behavior consistently? Can you try reinstalling the openshift plugin? Are you able to create a application directly from rhc outside of forge +Good catch Jun! on for patch v +Works on trunk for me +I can take this one if you should be an easy one +I believe this is related to the fix for +Thank you for pointing this out + in r. +but I just dont see itMaybe sample solution code +Am I allowed to vote on it if I helped push the patch along? I guess probably not +Heres a test case +Fixed visibility for any authenticated tested +Otherwise last msg still applies +Any way to make it work without switching to remote interfaces? +The real issue should have been opened in the JBMESSAGING project +I perform Signature Encrypt Timestamp I alradey checked it is not a problem of encryption or decripton +See is there any way I can get you to use the mailing lists in the future for things like this? Its easier for me to manage that way. +First this really isnt a problem with Commons Configuration so you should close this issueAs to how to solve it there are a number of ways depending on what you are trying to do +Click Update button project created +But the jdkType Type argument is a Class and not a and the has no type parameters so there is no way to construct type with the element type defined +Ashish please find the updated patch with use of variable name modifiers in UEL arithmetic as suggested by Adrian +Can someone verify if I have covered all licenses +Committed the changes +Patch applied at applying this patch the testing results of perf test is given as followsINDEXwriting of integersencode time flush timewriting of integersencode time flush timewriting of integersencode time flush timewriting of integersencode time flush timewriting of integersencode time flush timeThe time reduces the time from seconds to seconds +bulk defer of open issues to +Corrected patch accidentally included Ignore in +Id rename seq to updateseq to maintain consistency +Most of the time that the properties are used they are used by people using binding +You cant +It it either plain text csv or xml +I added it to Orbit dependencies +Adding Sending srctestorgapachecommonsdbcpTransmitting file dataCommitted revision . +This is mostly done though the feature seems to have some minor issues and will remain experimental for v. +Its a one or two line fix +Its not part ofOnly in Faceletssrcjavacomsunfaceletsutil +Hi a few points in advance +Attached CatherinePatch applied in r. +Make bean containing SDT selection of session scope +However I no longer see any failures on a vanilla build of Harmony +Yes document update is in tests passed but checkstyle failed because it does not match the new indentation used by hcat +I lamented its absence in +The new file should solve them +OK I will move it back to +I think that a file named is in a conflicted state +if you dont have both the beginning and end of a the block in the index it doesnt work since you have a question mark at one side or the other +Thanks +Attached is a war which reproduces the issue +This is a dup of +savana which was quickly supplanted in my mind with why wouldnt someone want to set their own properties on an ARBITRARY file of which +Added a test case to ensure for that deprecated cleanupJob is called from +A different task might be fine too as long as the ability to remove artifacts from a local resolver exists +update the patch to include Yoniks fix +Moved that test in its JIRA +branch master this record since is now released. +As for the error not being reset I can also fix that here +Updated patch to use the new binding serialization from +Changing core back to enables me to runSo maybe the problem is in core? +Seam debug page is not available for this particular conversation which is in jbpm +We need to provide additionalchanged methods in the client for this in to B since we already have plenty of high prio items for for old api not in new shiny rest. +Updated codetest to make bad state throw. +Let me fix the broken tests later today +The enum is there for logging purposes onlyId say lets just log the exception object and give it a decent toStringbq +Thank you and sorry it took so long +from manifest will not be possible to guess which is jmock version we are usingNo +What is the server managerYes getting someone from openjdk to look at it would be helpful +Groovy from zip distribution +EdThere is a unit test in the patch which fails without the fix +See +isAnonymous Y +I am increasing retry count less aggressively than original this should be more than enough to ride over server failure given the default negotiated ZK timeout of s + +Attached the source of the that uses rudimentary negative caching using a of +I definitely dont think we need more synchronization +Len +In production since +Just note that the snapshot I released a few minutes ago does not include this patch +Can you check with trunk or? Id like to do a resources plugin release soon can you provide a test case or test with the currrentThanks +The entire contents of the lock table could also hinder debugging by providing too much information making it hard to find the actual problem +We should probably look for another way to prevent the replica verification tool from giving false negatives +It is rather a test repository I was creating for a demo +link to RF project page should be given everywhere its mentioned in guides +Patch will allow for a graceful overflow +This would fail for schemes that have a dash +for the patch +Will reinforce that ZK is cluster mediator +Verified in +Attached patch that removes the offending println +corrected logic that sets initial value of widgets from saved preferences +Why dont we also check the VMs actual RAM and CPU? RegardsGirish +just committed this +I received a response +Perhaps instead of moving within the dna repo you should move these into the jbosstools repo? See for further discussion +Thanks again Michael +implementsI renamed to +It would be much better if setter methods were avoided entirely and the values were required to be passed to the constructor instead +netcom are now complete and whois shows them pointing to Dotster and they show in the account +Move the cloneForms method to class Add unit and integration testsThere are probably a bunch of other things I missed but these seem to be minimally required +I still think that it might be more convenient to upload the keystore in a jar file as maven artifact and reference this artifact like the assembly plugin +The code was blatantly borrowed from elsewhere in the codebase. +verified in Developer Studio to set release note required in release notes as Seam has been issue now that release note status has been set +now displaying the targetObjects class issue in core Spring would resolve the original problem +The script now creates the binary test files on the fly +All changes related to this issue have been done under which is very similar + +Lefebvre The patch is submitted as a pull request but hasnt been merged in yet. +Everything integrated as expected thanks Tim +Call New Project wizard using right click on Pacage Explorer New JSFStruts Project or File New JSFStruts JSFStruts Project +I added test service for this issueAfter drop you can check response via REST access +Unit test +patches can be used by anyone +are still created at bytes. +Just adding that if condition makes it your solution works. +bq +The thing I dont understand is how its different for CQL comparing to ThriftCQL in either case if we to remove the check users wont be able to run any operations safely except to the localhost +surely it still implements that was a typo it no longer implements +do you recall what the changes were to? are you using the scheduler functionality? are you using anything other than the default prefectch or batching settingsThanks +Hopefully we will get localized messages for the release + +I spoke too soon +Nick Rich faces Team Any idea what i am doing wrong +Committed this patch. +Once new topology id is received in response from server it is read and assigned to that reference from Codec got you +Anyway we already had a long discussion on the wiki and there is no point repeating it. +A bit tricky at the moment since regions arent a top level construct +Thanks. +But that doesnt sound like something we need to guard against at this pointIf nobody objects Ill commit this over the weekend or early next week +Thank youBest regardsSpark by Spark. +Kris correct me if Im wrong but isnt always used for rule validation inside the editor etc +Thanks David +Did anybody have the time to review this issue? Best regards Wolfgang +just attaching the same patch file again so that hudson can run findbugs and others on it +Thanks Sanne +I wonder if this could be related to as well although that issue is more about schema locations it does reference configuration validation a management only mode may be a step closer to being able to validate configuration and show the resulting model for that configuration before an administrator applies it to their server +Forgot to mention that the problem didnt happen with the previous beta release +With PBR and RF I cant reproduce the error using the Showcase example portlet +Patch applied to and trunk. +I have attached revised diff and zip files with one additional change I removed the phrase per the JDBC standard from the rrefjavcsti topic +Unfortunately we still maintain mojority of our developer testsunit tests in that v directory +I am working through the comments +The end result is some entry logs are created without the header and that screws the entry log ledger extractionAll new methods should have proper javadoc comments before them now too +yes indeed +I think the test failures are unrelated +not fixed in this last code drop +Fixed in Branch in both FC and SFC +We cannot fix this problem until we have access to the real epsg database +Attachment has been added with description startColumn and endColumn has been added with description startColumn and endColumn has been added with description startColumn and endColumn has been added with description startColumn and endColumn have committed this change +Closing the issue +If we want to improve the message when a reserved word is encountered we should open a separate issue for that. +The solution I am providing is to wrap the key inside a tuple whenever a user defined comparison func is used +I ll see if I can isolate what the difference is and create a new issue +Ive created versions with varied weights for the subset words +It intentionally excludes all of the indexes etc +Fixed r +Setting doesnt prevent the lookup though so not exactly the same as old importer. +Fixed as suggested +I either need to mandate to the user that for deletes to function the field say file has to be defined in the and it must be equal to the filename returned by the entity +Isaac any idea of a timeframe for this +bq +At the create trigger time we add the dependency between trigger action and trigger table +Derby installs the security manager +Patch for release +Instead the is improved with one more attribute components and three more methods that operate on this list add remove is +Andrew shall we include this with or as a separate JIRA? +Hi Stefan Thanks for the reply +In this second rev I offer recommendations for additional of attributes +Let me look into that +You are encouraged to make a list of bad names to speed up the the way each datamap can be used with different databases so Id prefer just global list of you create a new table with a reserved word failureI can imagine creating a column with a reserved word would fail toThe problem is that one has to know that all reserved keywords are safe in all contexts +the comments before I commit this? I have no clue here +Hi BrunoI like it and I commited in r +I found this jira issue unassigned but I think you have fixed this issue already on +Gavin Scott and I cannot get in with the keys that we posted up to and since we dont know the passwords please let us know next steps +I attached the unit tests which shows the issue +We can always tune the split size to tune the performance +s +patch looks good +This is a duplicate of +has properties stockQuote etc +Depending on the type of corruption the metadata may load ok but be invalid +Not fixed either in or installer +Changed to Labs category. +jar is used for building the HTML docs only? And how about the +looks great +Is it always reproducible? Does it happen immediately after startup or is it after a build or two? If you close that project causing the error does the workspace behave as expected? If you remove that single file from the build path do you still have the problem? What if you remove the file from the build path and then restartIf the core types are somehow being damaged then you will need a restart before things will start behaving as normal againFeel free to send me sample code to my private email if there is anything proprietary in it +Thoughts +I agree if it is correct that MS Excel files are not in cp encoding by default +r. +FF started failing around K and I didnt bother checking curls limitAttachments would obviously have similar issuesSo the question is if should enforce a maximum and if so what should it be +What is the size of small table? and the number of rows in small table +Ideally the table name and row key can be specified by the demux parser class via annotation +Did u try running the example by increasing the JVM memory settings +Yonik thanks I am glad it is not a serious bugbq +Could you provide a sample Maven project to help us reproduce this issue +Fixed in kudos to Daniel. +I would have to agree with Patrick on this +Resolving as WONTFIX. +my branch has this now though it is called to clarify its grand ideas as it allows us to have has happened +Around the time when i noticed it first we were shutting down and restarting nodes its likely that at that time some documents got indexed as well +Alsoit should require less characters to specify +I have started working on roles for Derby +same here +it works! Great work thank you +If you want to make a proposal in a pull request feel free +yes the reactor can only build one of each project by design +Id like to know their opinion because the weblogic implementation seems to be the root cause of the problem +Attaching the last file related to this have obviously lost all credibility by now but I notice that the getuseraccess file still has the extra verbiage that was removed from the setuseraccess file the first sentence says permission that is for the user specified instead of just permission for the user specified +I was proposing that we remove that flexibility +Candidate for Release Notes Flag please remove if not reqd +Where are you seeing the out of memory exceptions? You might want to run under java to get stack traces +bq +I tried adding a metadata while enabling model level metadata but it looks like it isnt t understand why +The way we handled those special attributes for lookup was different than fo searches. +Heres a patch to disable the raid build +pdf svnnative +The fixes involve a single file +Thanks what do you think of moving this logic once the tasks have been finalized +Restart the Enterprise Web Server service +Based on Nicholas recommendation not going to fix +This is a sample implementation I use currently instead of theIt is only a quick implementation but maybe appropriate for starting +Boris Im curious what error did you see? Was is a complaint about Kerberos service principal name wasnt configured properly +Attaching the jar file containing the patch on a users request +Set to camera plugin +Libvirt is mentioned in the install guide in the KVM hypervisor section +I am getting this exception while stopping my doesnt affects the flow of my the logs collected as the result of thiss huge +guvnorinfo naming so from that standpoint maybe splitting them up might be better but still link them together +Youll find in the attachments of this JIRA +Updated patch +It would be the end of prevent dirty flag we can simple remove it and the solution to this problem +Undefined variable handling detection fixed in +The wording nowis really akward +and now it turns out that the type attribute is deprecated +There was a typo +I tested the RI with my application and it worked perfectly +trunk +Some things have changed so its not exactly controlled +Thanks Andrew! BTW I am sewing up the chapter right now so your responsiveness is a huge help +Downgrading to critical as this seems to be specific to the users environment and has not been reported by a number of other users +So commit this to trunk for now? Open a critical issue against to get this test back or at least add another test form for rolling restart of under hbase? Id be up for doing this +Please verify and close +Thats what Bigram is doing +fixed by commit fcadcceeccaf for issue just saw this message in a log yesterday +Thank yous no need to add this to release notes +This should definitely fix recursion +OK thanks Ill apply the same fix to Tailer +slip John let me know if this is not true +Friendly ping +And people may consider this as an incompatible change since it can affect particularly their cluster management softwareOther than that it is just a matter of porting the patch. +update at SVN +Camel +ping +How about providing a patch +Moved updated and tested +This appears to no longer be an issue. +Committed revision in +push to +Thanks for fixing this bug so quickly! Is there a way I can download a nightly build somewhere with this fix already included +This reduces GC efficiency and performance +As far as I see every feature has a and every plugin has a proper +Ive not tested the patch but Id be surprised if a change like this didnt reveal some more invalid tests +Looks good to me! As I was not the reporter of this issue I cannot close it +I have checked in version of the dcoument that have a correction of the above quoted statement +Now ejbds is enabled in geronimoWhen I started two geronimo instance on one machine port confict on exeption will occurService Start Failed ejbds +Note that was resolved and will be in Teiid Designer and which version of JBDS is that targeted for +This issue is related to are a few issues at play here +committed +Most probably Ill be providing patched jars with added logging statements at TRACE level +This solution is aligned with what Srinath proposed above +Thank you for your time and efforts +verified build +On my Mac it also works fine now. +Comments The class name is +Please find the patch to do this migration from the attachment +Your patch seems to fix the hedwig issue without changing the flow of regular async +GreatI fixed a couple failing tests they now pass for me w multiplier +Getters and Setters for Name attributePlease upload a patch +Mass closing all issues that have been in resolved state for months or more without any feedback or update +If we can get the Hotel Booking example working that would be great! I see that is closed now as well +Rev removed the mvcinterceptors element from new mvc namespace +Use this issue to cover clean up of hbase to sit nicely in a hadoop package only only now align with of patch that works w the version of that finally got committed +all passing for me +Using as suggested by Lars +I havent had any problems with that for a long time +Can you share your load test? Are there any special OS or Apollo config changes you made +Thanks. +Ive fixed that bug but still unable to replicate this one +Part of a bulk update all resolutions changed to done please review history to original resolution type +This patch makes package fully signature compatible with Java This patch can be applied to core and modules +Can we stick with just a File argument no String +Better bug categorization APIprogrammatic problems versus stylesheet executionoutput creation problems +I just tried out the latest patch and Im seeing a bug +I committed this +I dont personally see a problem with it nor do I think it is really that big a deal +No I dontJames +Alternatively could you use the authentication event mechanisms to achieve the same initializationWe could add an interface that allows a shared callback so you need to register it in the container as opposed to subclassing either or +HeyWe are running into this issue too and would really appreciate it if someone could find the time to fix thisThanks +One needs to make sure that if application resolver obr bundle is to be used it needs to be processed by Blueprint extender before the application runtime or the application management bundle +Dropping this back into the CVS repository should do the trick +xhtml page +I cannot reproduce the issue neither under Vista x nor under Ubuntu Linux cant reproduce it anymore either +We should use the for the UDDI v communication +Ive found a better way to fix that issue +Can you explain how shared mapper comes aboutI believe this is a hadoop vs hadoop unit test problem + +has a patch that handles exceptions better and ignores the exception show in the description above +Well get the docs updated! +I misunderstood before closing it +dont use an indexed collection while marking it as mappedBy +I have applied the changes to trunk at svn revision +Is there a consensus to commit this or is someone working on an improved version +bq +Thanks Juergen for these two jira issuesFYI this limitations iswas documented in the sonar dotnet wikiThanks a lot for the fix +Thanks! +I am not sure how they cant conflict even when going under separate folders? They will end up in the same classpath anyways +All we are proposing here is make choice of Borat pluggable so you can replace with your own version +Thanks for doing the patch +Added in r. +Ive been able to reproduce and fix the bug +the disc is the sales field in the in revision +since in Hub server there is only one ledger handle opened for a specific ledger it is not critical for pubsubif you have idea on this issue you could fix it in this jira and marked as duplicated + patch + +null is better than +Its been some time since I last tested this so I decided to give it another go +Fix provided by Shreyas and committed to trunk version + +Done! +next iteration +In this instance the mapper records can be keyed by the canopyId alone since they do share a common +Fixed deserializer and now it does not crash on invalid XML +Completed At revision +Missed Ivys files for HDFS artifacts publishing +Thanks Raghu. +Im not an expert on the slice code but it looks like we should iterate on each target and return the first slice +But if it is indeed that Compaction just stops running at some point then there is something that we need to fix +Otherwise I would recommend removing it +And make it clear in the release notes that we removed bespoke methodsfunctionality +Patch +Our class was never really meant to be that extensive +It would be nice if somebody can help with implementing proper handling for zypper as well +Closing issues before next release. +Ravi and I could reproduce the problem locally on a Windows box +Ill upload the patch soon need some modify at Query because that tool couldnt access private fields +Similar bug for Turkish ValentinaYour patch is in trunk rev +Sounds like its only really a problem if you are storing large blobs over Thrift and maxing out that frame buffer per connection +So Content Assist performs apply on that single solution may be to disable automatic insertion of single proposals by switching of the WebJSP FilesEditorContent AssistInsert single proposals automatically checkbox +Hi Gert That seems to do the trick and it finds my repo that is not in the default location +That said im not sure how to solve it +Why do you require a spec for an issue thats supposed to work by common sense? Just ignore the JPA part in the description then +By the time you pull all the VM specific stuff out of there there would be very little of interest its easy it would be useful to do a JAPI run against Jikes RVM to catch places that we need to fix +Staged areEAP EWP These should be online in half an hour +changes produce failures of jcrspi tests +I only started looking through this a nit we shouldnt do ps grep +So I cant type in comments correctly +I should check that +This fix is for branch +Mark can you please reupload the latest patch? Im having issues using the last patch you uploaded +trygvis might be able to give you some advice on how to do this change +I did mvn r clean install in kernel extensions and runtime directories and the tests were clean with there anything else that you want me to doThanks Prasad +shipped with +Username djencks +Supported by and +It looks like this is no longer a problem +I am planning to change the computation so that the list will satisfy the filesize limit constraints +looks good +Heres Andys code in patch form +doesnt seem like theres anything that would prevent it +Patch for the issue fixes this issue as well +Nicolas I believe what you want is to set RESTARTTIMEOUT to something +If the file extension is changed to +moving all these to which is in line with BRMS +If I couldnt get that working thered be no point trying any web it happens I did update source folders and that brought things into focus Im still in an eclipse developer mindset doing things I always do without thinking about them +Do I need to have sudo perms to now run? Or the tests just fail if I dont have sudo +patch looks good to meFuture workWe should create a test as following create a file write some bytes but not close it +upgraded to in CR +Ive fixed this for will be available in tonights snapshot +i think this is already working since at least +Committed to trunk and branch +The client cant make a determination until he knows the server id for a mechanism +What about just creating a single threaded queueIve done something along these lines yes +to indicate that this was actually not done. +that was incorrectBTW implicit enum value assignment is basically the same in all generators +could you explain what you mean by streamable infoset abstraction that is independant from strongly typed object binding framework in XML Beans? I looked long and hard into this in context of XML beans and as far as i can tell XML Beans has in its core one particular implementation of XML store that is not streamable and it is not possible to swap it +bug can be closed +Im going to close this ticket with resolution wont fix feel free to reopen it if you face performance issue +I totally misunderstood +updated patch without the bug +These data points will aid in SSL diagnostics +done +Thanks a lot +This issues does not occurr any longer in it +The upgrade should contain fix for for late answer +fix +Sure +Changes New patch rebased against I also changed method names standby and active to more generic names transition and transition +Nice update +There is no way we could safely use the standard versions of these it was the I was thinking about if they dont match any Apache release you can close this issue as WONT FIX. +It worked on integration? Should I do something +This caused some adaptations of other configuration classes tooFixed in revision . +How do we know the body is being displayed in a site in which the linked index page is actually usedThe navigation links should appear in the navigation document for the specific rendering format +In my final version I am planning to put more unit test cases along with some fixes for issues I am encountering on my cluster +The new patch should be correct and works great for me +The formatting also is +I included the library directory in the ant task itselfAwesome that is greathere are some comments can you make sure you include the version name of cpptasks in the jar file just like Steven proposed above? something like depending on from what sources you build it and reattach it to the issue? can you please replace all tabs with spaces in the patch and make sure the alignment is ok I think we should name the task so we can add a lateroverall we are very close good worksimon +Its not a system error like so it should be handled gracefully like any other bad requestOpinions +Also added a test cased based on the JIRA report +Rebased patchDepreated some methods and added prominent notes to the that the passed Scan object will be modified during scanning +If we are ok with using Properties object then I will go ahead and make similar change to the code for readme files at the database level and at log directory +tgz must be built with a dependency that excludes all hadoop libs +Why are the response documents different lengthsCould it be due to redirectionPerhaps try with your local google site instead +As said you have an option to turn off estimated extents. +Does that sound ok? +Sure will review your patch +Ive just noticed which I think is related to this issue but I dont understand enough to work out if this issue is a duplicate of that one or not +I feel like setting the status of this issue as wont fix +If thats going to change then this feature is going to break +It works fine now. +Thanks Gustavo Ill have a look at this over the next few days. +The icons can also be managed using facets +Would like some proof +this is the changed you please attach a patch +I see the same problem with on Ubuntu +Tiny patch for this one case created to follow up +Here is an updated patch that applies cleanly to trunk +Resubmitting to try to get Hudson to pick it up +Could someone please review the patch so wecould move on with the release? fine to me +I think you need to use Byteman to make sure the readOnly message is processed before the producer of the message proceeds +Vijay please take a look +Please review and commit +The is only there to please the compiler +I think I did it because the new method was available in the code line only +Description of This patch fixes backing index related issues described in +Lets wait for Julienss tests and adjust the javadoc accordinglyWe may end up having some classes supporting fields and other requiring commutative fields and just state that in the doc or we may add a marker interface which would extend Field or we could add a isCommutative method in the Field interface and check that were needed +Please comment on that hours downtime for the last add dependency indexing error loop +Also added to control the max keys allowed in any +I can try to reinstall it and see if the error still occurs +I noticed Ranier committed this fix in this week +Now the code handles the following annotations Pending Add option processing only and on Scan jars with files ending with +Attached reproducer +However IMO the AM should only do checkpointing operations but not kill the containers +Since its not really clear what you are doing here all I can do is make sure the exception isnt thrown here +If all these attributes are meant to be after initialization how about we just make the all final +This issue should solve Arquillian Seam adapter in near future +Mark is the version attached to this issue a newest patch or maybe you have something newer +ok thanks +Not sure at it is now gone with Grails as the condition under which the problem occurs are not known precisely but I did manage to reproduce the problem at one point running Grails with unpatched I then changed my GRAILSHOME to point to did a grails upgrade and grails and the was gone +Then it may be a generic fix for all I committed fixes on r and r on +Is this assumption correctIn addition I fail to see why we need this overridable mechanism for such a basic featureSo I think we should rather define a service in the web console that gets called to deliver links for help texts or maybe the help text itself +Fixed by SVN revision +We may add a new jira to look into this problem and adding the batching warm upHandling the error code in the topicBasically act according to the last review comments +I should add that if you are running an older client version the pingkeepalive may cause this warning to be logged +I wonder what camel component that uses +Im on MBP i w GB of RAM +Wrong issue closed +Being verified at +Please disregard +I just committed this +I shouldnt just look at the trunk code when this was reported against +Eclipse shows this when I do teamsynchronize with repository +I havent seen it since that single incidentIs there a codepath that could lead to this condition if some previous operation failed with timeout or +svn rev fix version to +It looks like is also pulled in so I guess even that will be needed on the classpath +forgot to mention one must build blur this way for Hadoop mvn clean package install P! Garrett for looking into this I will apply this patch once we focus on +for setting properties andeverything else are special parameters +Thanks. +For trusted programs this approach works fine +patch applied +I think were at cross purposes here +In under you write If the scheduler does not support queues the value of this parameter should be left as default +slightly different patch exec libtool directly +I didnt consider the idea of including the package but it sounds good to meLet me give a a patch for the Apache Felix Shell to implement the log commandThis patch adopt Stuarts suggestion to importexport in the shell bundleThe usage of the log command is explained in a previous comment if no log service is present in the platform an error message is displayed +Ok Cheolsoo +Should be fixed now. +The doc updates were included in as I had other dbdriver updates +It works for me with doesnt work with Web Services binding because only binding has been fixed by this patch +This would just be a runtime property +should have better support for the various options at least. +Does anyone know if a bzip file can be splitted with the patch or we probably have to implement our own split method to really execute splitting? Thanks +Sorry its my fault +As I said the problem arises in a third party API which is supposed to be SOAP implementation independent +I am trying to see if theres a viable way to test the code in this area +Im curious if this is fixed +Im trying to build it up. +Have not seen build failures since making change + +for info I am now working to include a table of contents number pages a title page and metrics tendency +a win native apache license updated comments +Actually it has been fixed in. +Here is a new version that should address both of the points mentioned +Jon as to the actual bulk load command line it is a good idea +Yeah I know +and all seem related +After deleting the above foreign key constraints F and F I can successfully persist instances of C +Sorry for that I was new and thought that it was the right status +Actually I am already on Galileo SR ganymede was a typo +There are various deprecations planned to either retire or replace legacy functionality but the bulk of Designer will still available +The test output recorded earlier this morning +Juergen +So we decided to run joinringfalse on bootup and let compaction run without taking on traffic +Some of these methods are working only if the class file bytecode version is really or higher +Does the servlet not set a HTTP header? Since were using HTTP we might as well include it there +A simple patch +Issue is unclear and there hasnt been a reaction from the reporter for more than one year thus closing. +There is no need to create the +Before post a comment I talked to Olga her idea was to use single key combination to switch between views for example the first key press should expand source view the second one should expand visual view the third key press should return back to sourcevisual view +I think both will result in no out of memory +Can you please provide a patch obtained using svn diff against a recent checkout of branch? That will help to fix this bug in a timely manner +Fixed inNever was an issue for +fixed already? worked around +Output the value of the rowClasses perthe attribute description below +Patch looks good to me has been committed thanks! +Ok +I was surprised to see this issue still getting so much commentary +the patch looks good to me aside from the lack of unit tests +Thats much cleaner using the new clone method +Please find the attached file which contais a port of the tools from M +just working with somebody who wanted to use very long ID sequence numbers by defaultBut theres always hope! +My one concern was that the build might break if native JAI was not installed +To caputure the log messages a log handler with the name of the implementing class of the AE hasto registerdWhats the suggested way to retrieve the name of the implementing class or namesof the implementing classes in the case of an AAE +patch +Thanks for this +Please deal with it accordingly thanks +Should the user select multiple class locations or not +if property is not specified it will use +not to have Junit on the classpathYou need to edit file run targets deploy test for your advice +video showing am able to replicate the scenario and what happens is the item waiting for a transfer is reserved against a new order +There and of users are eagerly waiting for AS to up and running wrt it nothing i have mentioned impolite its only desperation +maybe I fix itwhen open the editor it will look for the xml file path from +It will create a appliance with Apache Hadoop set up in a pseudo conf mode along with its Apache Hadoop namenode formattedSo right after the first boot the appliance will start with a working Apache Hadoop clusterI put it in contribboxgrinder and it is based on Apache Bigtop +Also in the sample jar filesjar depends on wsdljwsdljjarI couldnt find this jar anywhere in ibiblio but I did find axisjarI repored this on versions have all been aligned for the release but the dependency still needs to be investigated + for the patch +committed as part of +This should be easy to +Correction of type my good to tie together with a better approach to the class +Committed in revision +AFSget will never thrown an because it resolves the first symlink in the path +I dont think we can cleanly fix this wo breaking backwards compatibilityIf its really important you can always create a custom which does respect max during optimize. +You can use egit historyrepository viewer to see what is agree the dialog is weird though and why I suggested that we actually show what changes is being detected since there actually are changes made of course should also check if those changes are necessary but that is a separate issue +looks good to me too +From this information we will write the final release notes for all issues +If there is still something of value in this issuepatch we can reopen. +Punt to +Nice and clean implementation +Thank for checking it up though. +It would be helpful to record the section titles before then to be sure the writer has enough context to find the material even if it moves around in the book +seems like you dont need new Container Allocator types +assume this was fixed in +That would be correct but still would getJBOSSHOME URL filehomealrubingerbusinessjbosswcjbossasbranchesBranchxbuildoutputBootstrap JBOSSHOME serverdefaultconf Base JBOSSHOME commonCommon Library JBOSSHOME commonlibServer Name JBOSSHOME defaultServer Base JBOSSHOME serverServer Library JBOSSHOME serverdefaultlibServer Config JBOSSHOME defaultconfServer Home JBOSSHOME serverdefaultServer Data JBOSSHOME serverdefaultdataServer Log JBOSSHOME defaultlogServer Temp JBOSSHOME defaulttmp If someone overrides any of the locations the JBOSSHOME is automatically omitted and the attention is drawn to that entry as could have a helper class that scans the for some variables and replaces the URL part with that variable name +Please let me know if I am not getting something here +Ive attached a possible patch to improve inferring mechanism +is incorporated here +Now with syncWith sync sBoth WAL and sync sWith sync enabled I seek occasional IO spikes on top of the constant WAL IO +There seems to be some some issue with the order of suspension there +As far as I can tell this is what the old behavior was so I think this patch should be reverted +I have looked at more +is slated for end of the week so should be first week of June +Btw the patch is for +I implemented this in CVS +Committed revision in branches +The RPM used was created from Accumulo but I had to slightly modify the POM to create an x RPM instead of an amd RPM +I still believe the Whirr in minutes guide should be on the homepage +Should we also add the advice You should issue either a COMMIT or ROLLBACK statement to complete all transactions and release all locks before you invoke an import or export procedure +But once I have things functional again I can do this +Lets move the discussion to the forum +Just encountered this in production using version Is this being addressed on the latest releaseThanks +The was now it is +Thanks Luc +Implementations can register and will be called when we query some information +patch looks good +I havent investigated yet but I believe there is an issue with copy fields in both the schemafields and schemadynamicfields responses probably related to since I copied the functionality from +The issue is and should be resolved already. +Which dependencies are not resolved correctlyDependencies external to the reactor or internal to it +Juraj so are you saying this is still a problem with AS? I thought from the linked AS issue that it had been resolved with AS +If the instances are created the only item I think well need to figure out is how to store the EC API key +Nevertheless some time in the near future this should be resolved +Can you please check it again? Please note that you need to apply the patch from before applying the patch from this one since this jira is dependent on that one +Could you provide a svn patch +I have added support for https client authentication with this patchplease review and commit documentation updates for ssl the soapoverhttpsender to load the passphrase as a property +Tomcat crash just because of Java Virtual Machine crash +Hi Yes your right apologies I did misunderstand you +Domains can either be filtered by top level domains ignoring subdomains or by hostnames through configuration +The plugin should without any configuration find these scoped dependencies and add them as a classpath entry to the manifest +I meant would be good as long as there are no adverse +Of course streaming would be fixed to use this +It is not related to as originally reported I recreated the original file and its still happening +But I have test failures in firefox +It passed twice over on the issue +I would like to make some contribution and think this would be a good task for me to start as i am fairly interested in the analysis part +We probably need a separate JIRA for it +So yeah I think we can make +Maybe it is better to validate the patch with compiled jars +I have segcodePERAnd then it does match the following segmentN N ST UPPER +committed both to trunk and to as mem is not coming thru for when I tested on acbdddebbcabfbcb +I agree I was being cowardly when I wrote it because I am not a committer I heard you intend to add a small patch to flip Solr s default on this feature? I was picking Erick since forever and he pawned it off on you. +Would appreciate it if someone could review the commit in +Ill remove the Set thing and upload a new patch +Still have a long ways to go but a plot came out the other end +The inline patch came out garbled +Ping +This patch makes the class thread safe +verified on +Do you have ANTHOME environment variable? Could you try latest from the CVS +Committing Thanks Alan +The fix for this is committed to trunk moving this out to target for documentation updates +Is the a good place to start +but it doesnt seem like you use itdid you see the patch i added to? I think its a good fix to this situation to cleanup leftover files but ill try your patch anyway and ignore that for now +I think will server the purpose because does not mind what is the internal byte structure so with or without tag is should be able to handle +very much RickCommitted patch to documentation trunk at revision +Attaching a patch which implements CtrlC and uses it instead of on Windows +class should now be removed +Unfortunately the documentation is in error and I have fixed this in Git commit ID edcbebdbcefdafd +Release Notes Docs Status marked as Not Required as this issue did not affect any versionrelease of available to customers and it has been fixed in an Engineering Release. +I say no on because Connector doesnt have references to the needed to make the code path make sense +test failure should be fixed by addendum toPatch v is ready for review allows all mapreduce jobs to pass using hadoop dont follow at all what you are saying above +It looks perfect now +is used to delineate messages so having them mismatched would be badThough this does mean that youll need to store the number of bytes before and after compression when you stream the attachment to diskAlso the compression algorithm must be specified by the header not a URL parameterThe tresholdforchunkingcompresponses is a bit of a weird threshold +This is resolved +Cannot reproduced as well JBDS installed from Marketplace +This is because there could be local commits into the git clone +Donald I think David is using this JIRA to track the patches and discussion around this topic instead of an issue in bugzilla +That fixed the hang running the tests +Great patch stack +I find this useful sinceit allows me to specify Dynamic importsHowever Ive found that if Include resources is used this doesnt workany morereplyI think this is related to at the moment the bundle plugin addsan entry for srcmainresources which may bepicking up your manifest file adding your own setting probably overrides this so it doesnt see it anymoreThis can be addressed in +Thanks Sebb! +For anyone who wants to use this validator dont download it here get it from the wiki example +take another look +is working just fine +Example of using the tar file and getting the error +Currently the does only return a warningBut in the current code of the each exception stops the process +I tested this and it also works with no changes to module dependencies +Since this is not very critical we probably should only fix it in +There have also been a number of other fixes and improvements in NIO +Good point in which case Mevenide should simply display it in the synchronizer as a dependency so it is up to the user to add it to the project classpath +Does any one knows a work around to this issue +This is strange everything builds fine on my machine +Will add a new patch with it +The only thing I can suggest is that you try doing a clean rebuild of both Xerces and Xalan with optimizations turned off to see if the optimizer is doing something evilOtherwise I have no idea what we can do to fix this since its blowing up deep in exception handling code +So if we consider it as our internal file and are not going to modify names then the issue is fixed but if it is better to improve them that part of job can be moved to +The doc for the setting should point out the down side and we can let people choose for themselves +Oops never mind just realized I added the projects to the pom a while ago +I attached a patch Disable hotswap classes in geronimo server just copy updated class files from eclipse to server repository let eclipse itself do the sample application to reproduce and verify it +This same application was running on jRuby and Rails without this problem +Patch applied thanksYou can close this issue if youre satisfied +Related to the work done for this JIRA +Small difference in +To my knowledge there is no other open source implementation +Let me know if you need anything from me +Again I can make this changes only after the files are committed +I think since this is such a central part of the code we should make sure to run the HDFS and MR tests against the patched Common jar +Yes this will make it in +I dont think it should be marked resolved either until it is clear how exactly the issue was resolved +We could also strip down some of the portlet wars we package in this configuration +I dont see how including the current seqnr in the notification would help at all in this case +Splitting it is probably a bit difficult since for all changes for the other issues the header introduced another you have a hint for me concerning that problem? I thought about splitting it but then I would have added patches that depend on each other and not on an actual revision +quickstart demonstrating the cannot reproduce the issueHere are my steps enter a select any of the options +Im OK with your solution +I misread your problem +I confirm that there is a bug +commentActually in the original certificate there are no special chars + +Ive tested it with all the other jiras in progress yourkit does not show any blocked thread anymore with two or twenty clientsThe difference seems to come from all the valueOf that check the cache content before doing any to byte array creation +Hi if someone has a couple moments to this patch I appreciate itI know I need to add a stopwords reference to since its BSDI think everything is in order but maybe I forgot something it would make me feel better before committing +I recently upgraded Norton from a weekly scan to a continuous one +Cant we use lock prefix for the atomic access? I am confused +An exception will fail fast +I also like using the messageId for correaltion +passes +I see that this bugs status is still open I wonder whether I can make use of patch attached here for logcxx source +The UI can be improved +Updated patch for so that it applies cleanly against the Spring maintenance fixed a null pointer exception in the retry patch and patch that applies against will be delivered shortly. +That sounds perfect Christian +Couldnt find particular Components value for GPD issue +the patch looks good to me +Of course that would only apply to synchronized Connections that is running within transactions driven by not to Connections managed by our JDBC which absolutely needs to use the same Connections for every level of make the latter tradeoff work with Hibernate the Hibernate connection release mode would have to be set to afterstatement which they recommend for Hibernate in a JTA environment anyway +Thanks Colin! +The simplest thing to do is repair all when only a KS is specified +Good to hear you liked it. +Steve if you can resolve this problem I can submit this patch +if you have really large messages they the destination pageSize can be reduced and there is also lazyDispatch that with a prefetch of would ensure that there are only as many messages as concurrent consumers in memory at a time + +dodeploy have not always been removed by a +Not a bug issue closed see comments +Closing since Christian committed +Without the patch corruption is detected much later after the data is written and acked which makes recovery impossibleThe patch for requires a slightly different patch +Bundle JAR updated. +Ive tried to work a bit more on this one but Im not being able to fix it +My task is to generate the same schema that JAXB schema generates so comparing it with it +We have started this week to review the options to enhance the authentication across Remoting including trust within an established Remoting connection +See Attachement this is what I mean +Committed to trunk now. +the patch still seems to have NUMRETRIESWhy should NUMRETRIES not be there +removing label since upstream issue +release bulk close +Support for inclusive gateway is added to Engine and Designer and documented in the userguide. +The location of the logs are detailed in +I still need to fix shim tests and the missing in +Ive added a method for converting name to canonical host name +Revised according to review +This patch basically creates a new thread on handling shutdown event from for for patch for latest patch isnt applying any more +But if you provide the location where I can get the code I can build and test it +You should raise this question on the user list not jira this is not a bug. +bq +So perhaps we shouldnt rely on id only +This would solveI like the algorithm Doug outlined aboveA few thoughts for refinementProcessing N input splits per mapper task may result in map output spills depending on the value of N the split sizes the value of and the nature of the map functionThus N should be configured by the user on a per job basis +I have attached a test that shows the problem +You might want to check if you see anything similar in Enterprise Jopr +Kind regards dj +So before sending done we make sure counters are updated by sending a last update +However Im waiting for an access to the Red Hat Issue Tracker to be able to see support cases. +In the case of a null value or key we simple treat it as an empty string +This seems like a use case that would be pretty common and wondering what other people are doing as a workaround for dealing with whitespace in the paths +Committed r. +Patch was reviewed and applied to trunk with svn + +bulk defer of open issues to +in maybe thats overkill but you decide +I will investigate it more deeply +Bug has been marked as a duplicate of this bug + +ok great +Patch st per value too long +Patch applied. +It is occuring in trunkI have checked by accident then I couldnt change this +Please confirm that they have been applied as expected +The plan is to no longer use the repackaging of cglib and to migrate to now that it has been released +Editorial pass over all release notes prior to publication of. +Searching for primera pedra I think I found the answeracte de col locaci de la primera crec que si diem primera pedra a seques tots entenem que es tracta dun acte cerimonis +I guess it would be possible to find the class for collections using reflection if your classes are using parameterized interest is mainly in building graphs of Java objects that are fully typed as +Only we do the batching at the level not in an data structureI think its worth reevaluating the batch API for Robert Newson pointed out that the batch writer API doesnt minimize writes like the batch api did which ends up resulting in larger database files due to more writes +I have meant we can use our internal classes +v +If offline flag is on in settings then skip the goalIs it enough? I did not see any point to run the goal in offline mode +At the moment I cant see the difference between what is happening in this example and the which currently pass +Either rejected or a dupe +This is required to autoclose session which was open with Transaction Extended Manager +Thats correct. +Refer to Section Scenario Referenceable onpage for details on the JNDI reference mechanism +Can someone whos had issues with this test it out +Attaching and adding a couple of sentences to the CREATE SCHEMA topic +Okay if you get something that pukes most of the time I can probably fix it +Ok +Look like this issue already have been fixed in +I can remove that part of the patch again if it is too intrusiveI dont think I follow why should be set to null +reopened just to set the resolution to with resolution fixed +I havent been using the test case +I am seeing this in development mode also but only when the database is loaded within the context of a web request using webrickIf I load it on server startup it works fineI cannot reproduce using thinStill trying to cut down a minimal test case +That code unfortunately had the effect that it ignored bean definitions with a because those dont define a bean class +However my recent testing has shown several failures +Reopening just to link the other JIRA and close this one correctly as a duplicate. +Should be easy to write +Were using this functionality to launch our fit tests and for starting jetty within eclipse +It looks like the last step execution might be null there as well so I worry that there is something wrong about that +Thanks a million +Patch in has been applied and a snapshot version deployed +I ran forrest war and then manually moved the buildplugins directory into the root directory of each of our forrest webappsRestarted Tomcat and +It is true however that the use of is somewhat orthogonal and is not required for writing XML JSON etc +just committed this +Thanks Hairong! +Thats pretty cool +Not sure about the parent archetype but agreed that the archetype will need to be synced up I just created to cover this +Addressed review comments +Fixed on the wrong ticket +This code has been added last year and it seems to break the contract of the get method +Not seen since CEST +Cant reproduce +Here is the snippet from Adv +I like the new patch from Justin but I clearly want to be able to resolve requests without an extension as well +Gets in the way of flushes triggered by sizesMake so flushing gets a review +I wonder if this is a dup of +Until I start seeing a large number of GEP users asking for a Geronimo Studio bundle I still feel like we would be wasting our time building testing and supporting such a bundle across at least Windows Linux and Mac systems to cover the platforms that Geronimo developers are using today +Raised as this is the minimal functionality required to support the case +Ensuring messages are not lost requires some form of acknowledgement which might be better implemented at a higher level where more information on the structure of the application is available. +the archive for and make a decision accordingly +So all issues not directly assigned to an AS release are being closed +The compensations will be done by analysts directly into the databases or will be ignored depending on each caseFor this kind of migration to work an important assomption need to always be true that we can at any time compensate a previous activity +One reason is that it doesnt work with virtual repositories. +If I save and reopen editor after Step +We need an methodYes I just committed thanks + for the patch +In other words we have a load factor ofThat clears it right up for me +Created new issues to implement improvementsTenant has already been released with the patches applied +Please let me know if you have any opinion about the design note +Ill start on the offsets some relatively mindless coding is probably about where Im at today and the brief look I had at scared me a bit +confirmed certainly a nasty with +On hold until I get apache set up with a cert to test this stuff +How is this Benot? It implements your second suggestionYou need more though dont you to make TRUNK work w opentsdb +I will commit this patch after I run some tests +For the PartyRole problem I agree that its not strictly related to this problem and that we have to find a more general solution for it +agreed +If this logic is removed this may be resolved. +In SVN +Can you pl tell me where to mark it as suchThis is a committer requirement Dhruba dont worry about it unless you disagree about the characterization +Resolving this as I see commits to this bug +just one suggestion when a patch is ready for review could you also attach the patch to this jira and change the status to patch available? since sometimes the reviewers might miss the activities of a jira when he looks all the patch available jiras he could find your patch to review +Yes it does work for meThanks for committing the patch in +is not included anymore. +Lily may want to turn the flag back on if she wants to continue working on themThere is a small amount of work remaining to be done on JDBC support now that JDK has gone GA and the spec has been published +Patch for problems file +see +Thats not the style for wizards in Eclipse +They will be useful +looks good to me as well +Ok I actually cant reproduce in the case of the jboss server not running are changes not deployed which is what WTP does it doesnt publish anything if ther server is not please reopen this if you can reproduce the situation both touch descriptors etc +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +No response assuming ok. +Probably it missed some classes during packaging the starts working if is placed into the lib this is all information I can provide by the moment. +With pig builds taking hrs it wastes a lot of time +Stay tuned +Note that some information typically displayed on is missing RPC server address cluster ID and block pool ID +I am using the g you could mark them as optional so they are only imported if the packages are available +This issue gets my +This along with making the embedded engine not decrypt the password property correctly thus failing to make connector connection +There is sth +Give the recent commit a runthrough to verify +There are two implementations currently this is the default which matches the current behaviour this one allows you to have an index that is completely held in memory +We have determined this to be the complete setrichfaceslibliblibliblibliblibliblibliblibliblib let me know if anything should be changed. +Currently this logic works fine with the Portlet Bridge +If they are still valuable for the earlier version sub task will be created for each. +Im having an unrelated compile problem ATM. +when an upload exceeds its size it is always the root form that is marked in error +I found more bugs with my code +Patch looks good! Will commit soon +Hibernate does not read the value of the element just check if it is present +Matter of fact TC TC and TC all fail with the same error message. +added browse applied +Please close if you are satisfied thanks! +Oops uploaded incorrect file previously +Thanks Ashish for taking this work to another step +Good catch +The current code is derived from WCS but it has proven to be completely inadequate for the different requirements sported by the wcs request +The patch has been updated to apply to didnt realize this issue before the release so I created the feature request per it was already released and release notes published +Apache commons is used instead of because the latter is not accessible due to package private +all but the change for POM locations and this is fixed in repoclean and awaiting the changes in +Rysiekis that issue resolved or what? i remember you were talking sth +Other parts of patch are as before +Pushed to. +Logged In YES useridOrdering should be fixed +It exists only in version and doesnt exist in any other versions +It has the same Database for the project +Fix in the revision Improves immediate component detection in order to avoid to a component as immediate when it is already immediate +Setting fix version to indicate I want to get this in for +The button bars need to be refactored to wrap around links +Sounds OK to me +The latest patches have been installed in revision +Theres another bug with core reload that I found while running Alexeys test +This solves the issue +I agree with you will make the changes as per your comment +Part of a bulk update all resolutions changed to done please review history to original resolution type +That should still happen at the RM as the very last option at the last moment +If there is a problem please open a new jira and link this one with that. +Ted how do we get for this? Maven doesnt find it anywhere +Id assumed there was a better more general fix but what about the near term? Does it make any sense to this enough to put the parameter in the update handler and commit? I think theres value in the patch that would provide help to users at least in the XML case rather than defer any improvement at all into the indefinite futureAs always though up to you +I did have a quick snoop around the relevant code for issue +In that case MRU is not useful and wasting resources +A patch for the proposed applied +thanks +Needless to say its a mayor annoyance not to be able to have parallel deployments so even a quick patch would be appreciated + +Setting +This looks like a clear follow up from that we never pursuedOTOH though I didnt always feel comforable with the FINISHING state things like these prove its practical winsRe the patch The java comments needs to be slightly improved to reflect what we are really doing Test Rename cleanup to staging +Having a TASKSTARTED event which fires once the task is available would solve a lot of our issuesIm fairly sure that if this was done you could then remove the START event entirely since I cant think of a scenario where you would need to distinguish between START and STARTEDWould it be possible to this Jira? +Just a documentaion problem +If you think that the output to STDOUT and STDERR is too verbose or that it should be sent to the log file only please open another bug for that +looks good to me +Committed to branch +resetting P open bugs to P pending further review +and then periodically take incremental backup that capture the changes since the full baseline backup +ok thanks +I believe that at is quite different issue +We need to fix that +Those attachments are patches for trinidad one is based on trunk one is based on branch and the third one is based on branch. +blah +Damn dyslexia! I can hardly see the difference even when its pointed outthanks eric +Thanks Mark Ill ping another committer to quickly review this +This issue is also another occurence of +However that might be the most useful environment to be able to have the stack trace available for tracking down issues from an enterprise support perspectiveAs an alternate solution I would suggest that we always have the stack trace always available on the error page hidden by default with a see more details buttonlink that could render itThoughts by the community +If clients need more details they need to look at the logs +Do you see any of these subtasks addingremoving from the current User Guide +ultimately I still think case folding is the right way for this across the board but you cant do that without a rd party libraryI think that any project that is serious about texts in various languages will already be using ICUJ +Is the folder the tests are being run being cleared out before the each runIdeally that should not matter but I dont think we have yet reached the point with the junit tests where they all can be in the same folder +Can you share a sample script and more detailed instructions how to replicate this issue? is Domain class plus roo generated aspectj file as well as controllers aspectj domain password adminadmin +file and an empty package will normally trigger the creation of the entities when Roo starts attached +In the mean while can you suggest me some good tutorial for this online step by step tutorial +Ive retested this issue using RC the changes to the WSDL now can be successfully imported by wsimport however using wsdljava I still see an error which has slightly changedType ws is referenced but not defined +I will updated the code by using the endpoint uri which can reflects the config setting +Ill change the label to be more consistent +Would you be able to provide a test so we can reproduce there error you are seeing? That way we can also verify a possible fix +In fact both issues are the same +will make an patch +We havent migrated to the yet because the has served us well so far Did a quick test and it seems that the doesnt support this convention either +i was looking for something similar and was interested in using your +It is not logical to have such workaroundThe most important fact is to have a fix for because you should be able to change the scope for allany transitive dependenciesThe Maven is declarative and thus making NO such assumptions about this kind of scope of changes should be made unless directly specified in POM xml. +Youre right of course +Im not entirely sure why but it might be that things were adjusted in the build to work with debianubuntu but broke packaging in the process +This plugin errors with with maven as well +I dont know if that issue affects a version that we released +The patch is is applied for branch. +orgapacheshindiggadgets +There are nodes in our cluster +Sure I can reuse the +Im evaluating the patch now for final evaluation week and will report back ASAPThanks be placed in the toolslib be placed in the toolslib be placed in the toolslib be placed in the tools subversion diff cannot include binary files Ive attached all the neccesary for the v patch and jars Tatyana +The forms their website currently uses utilizes the input field namedestResident with the following potential values Residential Address Commercial Address Have UPS Determine My Destination TypeHowever I do realize the current API spec is using input field nameresidential with the only potential values of Residential CommercialIs it actually a required fieldI suppose my series of questions was misunderstood +New patch created without using SVN moves. + +I wonder if the spec should mention thisBTW removing the read lock produces a dramatic speedup +just to keep it on the radar as suchAs far as how it should work +Currently not working on sorting works the sorting by Admin State does not work a mix of Decommission In Progress and In Service do not sort as expected. +It seems that there is no agreement at the end +net and VS is way faster for dev in general +I found one more missing ID For XA +The task alive and stuck +In doing all this version files of datanode and namenode got messed up and n NN had different set of information in their version files +It would be much more flexible and would only be a few K dependency for the NOP version +Thanks. +the second problem was the noninitilized register +Fixed in resolved issues for already released versions +mvn install or assuming youve set up ANTOPTS to workmvn install workaroundalias msucksmvn ANTOPTSalias minstallmsucks install alias mtestmsucks test +Thanks Ted. +I just dont want the added overhead of any kind of sort +It is also an indirect persistence representation of blocks that need recovery if the datanode restarts +All makes sense +You need to create a confluence account so I can give you access to edit your space +I also forgot to add that REST API documentation needs to be updated for this change. +The context menu actions now only show whenever at least one server has been registered in the view +Which version of Camel are you using +This seems reasonable to meChris are you vetoing this or can I commit it +Committed revision +We can also isolate those classes on the teiid side for teiid final +We probably should update our defaults to turn off +Hey Zied you raise a good point +attached patch upgrades to +BULK EDIT These issues are open to be picked up +Patch attached +I cannot get the collectors to run for more than mins +the ant task definition you are using etc +Update to latest header design which is under development just now +Attached is a simple patch that adds the passed keyvalues in the jobconf +it can hit namenode and the network pretty hardThis probably is caused by the call getBlocks +LOL Helps if I check them in +Would be cool to support this in udig +I think this is a or never change +Problem was that I didnt declare plugin version explicite and my maven tried to download instead of +Committed to the trunk as revision . +menu item? Or did you Run on server the web projectI cant get much further here without additional information +bq +I was suprised to see that getResource was called on pages where there were no instances of my resource reference +log with full exception trace +I also fixed the issue in +Thanks! +The only provides functionality to manage existing caches +I was able to make the proxy work for ssl repos by adding myproxyhost and myproxyport to the fileIt works but I consider this a workaround +Committed +Marek no its not ear its a war project and there is no +request to be added to the geotools project +Hi Nirmal +Thanks! +Let me ask if therere guys from Cloudera out here any plans to backport it in current CDH +The patch is actually not directly related with but it enables the Fedora to display Chinese by println +Junping Id suggest splitting the YARN and MAPREDUCE docs +Alexander and SianIm looking at the latest patch but had some problems to apply the after applying I believe this is caused by confliction between this two patches on so I wonder whats the relationship between them? Does one of them include another on the test part? If not how to merge them easily? Thanks a lot +I succeeded to bypass the error +Will leave issue open in case amendment neededFor javadocs warnings would it make sense to fix all of them in trunkThen the count would be zero and it any error would be a real oneI know its a boring job but may be its the easiest path +Patch that checks for the occurrences of two commas where they dont have to be adjacentLooping now +I doubt this is going to break anything but lets get a hadoop qa run +Not enough information here to reproduce. +bulk close of all resolved issues. +This is therefore workingPlease reopen this with further details if you are still having this issue including which application server you are using and which version of Wicket you can make this fail on. +Please take a look at pictures +Nope just forgot to set the issue type +Im appreciated if someone review itThanks ve written updated patch sets and attached them to each subtasksNow Ive started to write test case +Committed to master. +They have an issue for this MOJO +This issue has been fixed and released as part of release +Not sure what the status is on this one +which is so called legacy plugin mode will share all static fields between applet invocationsThere is no preferred way to start Pivot applets +closed +Ive attached my failing tests along with the two HTML documents which the tests currently fail on +More tests added also. +Because users might want to just copy the schema over without bringing over all parameters +However flatpack is also a project which seems not maintained anymore. +Thus a configuration switch is pretty tough +ThanksFYI the bug would show itself if you had a PPR request that updated something other than the table but then model state changed out from under you +Ill have a look at them. +o +We must take account of course these other users but they are lower than other considerations like how CDI works in Java EE and above backwards compatibility with Java EE application servers +Progress bar upload component during upload +Wouldnt it be better though if or were used to store parametersI suppose in that way parameter would be easier to deal withImagine if one wanted to add an additional to the standard set of +At Chris Douglas request I gave this a read +Due to your answer I could localize the problem myselfIt seems that it leads back to WSDL files in the same ressource folder +This is kinda hard to test with out a filter that exploits it +FolksHeres the first draft version of the exception handling guide +I have no way of explicitly clearing the so its a bit of a show stopper for me at the moment +This makes it hard to set an isolation level at the JDBC Connection level and particularly hard to reset the original isolation level before returning the Connection to the poolIn any case this should be better documented +Its possible to use expressions in this path but by the time its validated in this step those expressions will have been change was to add the same validation to managed dependencies. +bulk defer of open issues to +This is a configuration issue on the local WAS installation of the user. +Thanks Navis +Hey Tom +Ifgenerating HTML on the fly is too troublesome a note in the buildbotoutput warning about the latency would mitigate the problem +Please note that the standard plugins still dont use boostsIf no objections Ill commit this shortly +this is an improvement not a bug so even if someone provides a patch were not going to try and squeeze it in for +I was guessing that the notification did happen although the server didnt return any content +Removed for loop to construct CLASSPATH and integrate as part of this patch +Java reproducer for the problem +I also took the opportunity to clean up the package a forward port to trunk +Redirects should really not be followed immediately anyway +At the very least I dont think well do it for +Committing to +I did not match generator map tasks to verify reducers +This makes the progress meter show upload of twice the file size +I pulled of course +Resolved issues being marked closed in conjunction with the release of. +Anyway many of my java developed flows were rewritten in xml because the swf api for developing java flows was changed to much to refactoring it in acceptable time +I dont know whether there is a bug hiding here or no and would appreciate someone who might know to take a lookAttaching a patch that shows the new output +I dont really care which way thoughI disagree the new UUID would be more unique than counter approach +What is the difference between Joes v patch and the v patch? Can you please describe the changes when uploading a new patch +Confirmed issue please exclude +any clues? Ive checked the qs maven deps didnt see any deps on xalan nor xerces +Reassigning versions to since a workaround is available which is basically to paste another dependency of the logj exclusion in the pom along with zookeeper +usage as EUR and USD +Looks like previous next links were added to the bottom of the page +Throws an exception if the path is null +Committed to trunk and merged to merged this to too. +Rebased and merged Horias test case into the master branch but resolving as Cannot Reproduce Bug per his previous comment. +Refreshed patch + I guess the mvc namespacehandler doesnt register this class with the bean registry by defaultI meant top level bean that is visible to others +Thanks Mounir +Surprisingly passes without issue +bq +For now Im only fixing this in +The drawback is that it doesnt support rappid development the webapp must be packed before it can be tried +I do agree with your idea to improve the logging API +Kristian reports this as fixed in. +I dont mind either way +ant test is for running the unit testsant is checking other stuffs like findbugs javadoc forrest etc +Hi XutingCould you please take a look +The one that introduces the extra warning had no failing tests +This exacerbates the above rare issue and makes it possible to repro it more often +I am working on the importing phase now that the UI is there +Adding more comments makes a lot of sense +Thanks Norman I downloaded the SNAPSHOT version from nov th and still has the same issue +Next Ill try to understand why this is happening +Also is audit logging enabled when hudson runs testsNot sure of this +Fixed so that alldoc uses the same comparison as doc insertion to handle endkeys +Reproduced on the Windows version of in Firefox +Release notes checked Chapter +Hmm should I update the plugin codebase on svn? I can take care of explicitly including the INCLUDE dispatcher in the specifying the shiro jar dependency in. +I think with the system property we can close this issue as fixed +The only sane way Ive ever managed a cluster is minimally using shell scripts and ssh +If we configure to load flows from classpath +We run this patch in production and use for indexing only +Thanks for the JSW pointer though +Committed revision . +Seems like a good long term goalWhen the package dependencies have been inverted the implementations can go into child class loaders while the high abstraction packages become the common classes that proxys applications and services use to communicate with each other belong in the parent application class loaderImplementation classes belong in child classloaders with application proxy and service implementations +It is instantiated in +This is a dummy cert that is useful only for testingThe use of localhost as the OU and CN allows the cert to pass spoofing validation as implemented by the jdk jsse implementation +Wait and youre saying that the same data when going through but from oracle spatial instead of this shapefile actually works? That strikes me as rather odd +Should be fixed by +Correct +Fixed in rev +But it is going to be a very huge effort for us to fix the database design to make it work +Its always startcode and server cells that are missing their deletes +Post hooks allow introduction of postprocessing and modification of return values +It throws which indicates that perhaps its better to follow this behavior +Ok Ill go ahead and get this going then +Fixed on from the svn branch and this is fixed +The attached project demonstrates the problem when running mvn validate on the project. +within a singe read there are retries for each blockThat would be better in that with the current patch on a block file if we hiccupped the first read on each block wed trip the failures count though if wed been counting on a block basis the read would have gone throughThat said Id be fine with Todds patch its nice and clean semantically and going in and in a new issue working on the Tsz suggested improvement +These problems should have been fixed in release +That is what I saw too and I thought it would be a simple fix +a customer is asking for something that we envisioned in the project but for which we are not comfortable with the level of QA just running the test suite with this configuration definitely does not give sufficient we risk that the test succeeds we say to the client to go ahead with this and then the client potentially bumps into one problem after the other +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +Should be folded in here +Issue a debug message for rather than a warning message +Im using with do you think its possible to have any workaround for this version +This patch makes the replication placement strategy and endpoint snitch configurable via +I changed the Component to use a and it seems xwork now uses a linked hashmap as well +What I did was install this application as +I suspect the most challenging aspect would be connecting to https targets via proxy +I have not used these settings on at allSo you should remove those settings on and try to run with false and then I think you will see that it works fine +Thanks Ill check new version +Milosz Tylendai have to apologize your patch works great +This is the with trace enabled for the clusterd post office package +Sure if they are no longer in ISO specification +I will take a look on your patches +that way it should be easy to delete from Geronimo +Ive just verified the bug the problem is solved this issue may be closed +the call for this seems to have evaporated over the past couple of years as the EE plugins have become capable at naming their dependencies +because of this patch Quit button doesnt ask for confirmation anymore it just kills the installer without warning +Fixed with patch and improvements from Robbie +To generate report use mvn site or mvn coberturacoberturaRonald thanks +I looked into +We have an issue for adding icon support +Anyway the calling Tx is ineffect and this method call is only to get the objectcorresponding to the foreign key +So does that mean its an issue with Tools maven integration +The patch changes that toInstallingSetup completed +Ive just committed this to and +Closing this bug +Can you please upload a test file which shows the problem +Well yes i am using castor for the first time +Are you sure that you did a clean +If its not going to work no point in having it enabled +Part of a bulk update all resolutions changed to done please review history to original resolution type +Also any thoughts on! This tool is coolFrom the code it appears that the tool parses the image file one record at a time instead of reading the entire file into memory +Hi JacquesThanks +The plugin contains a patch that forced the project to be considered a Java project if the PDE mode was in theory should solve this problem +Is anybody already working on this issue +Committed +On other hand there was no huge amount of changes in Maven Eclipse Plugin +This depends on which is targeted for +However it may not fully complete +bq +The shell scripts need a global fixup +Eric you say you want the constructor in IDL order but your original patch used kindabyidlessthan +I added a TRACE log from my machine. +OK great +Nicolas Can you please attach the upgrade log as wellAccording to the code you mentioned in all config items with category Hidden will be encrypted +This has gone stale +Ill close appropriately when Im done +Up Lets remember about this issue when moving flags to the command in +The find call on may take too long and result in task time out +although it is probably wise to test upgrades via the update manager in the future to see if it behaves the correct way. +Either that or the fix would break other stuff +Attaching the patch for java and java branch with the implementation of above comments +Thanks +A null value check should be done before looking up the namespace uri +Verified by Richard +Attached patch a which simply adds the number of leading columns to the toString output +But having an easy way to prevent forced resolvers from being forced during install could be an interesting improvement +This is only the common part of the patch +The unit test passes. +Once this has been done we can make a decision +Ill go ahead and commit this to trunkThanks Matthew +Ive reviewed the patch and think its very good +I would say it is broken without even checking hudsonwhere is it supposed to take target platform now if tagte platform dependency vanished +Release of TIKA contained a version of the patch for this issue which had the described problem +I noticed that this behavior had changed but from a quick look couldnt find any obviously relevant changes +Should we just handle the eof by no longer replaying edits of a busted log? It seems like itd be the easiest thing to do +lets push it for issue with the current approach is that it has the potential to explode the bean mapping view with one click of the button by the user +If that is indeed what the original patch does I support itWhile the bias here is an issue I think it is a larger task to change it across the code base +But I wasnt able to get the tests running correctly without removing the default profile flags +fixed the issue with revision added the test cases +But would it do harm to run the TCK with the latest version? Anyways we can think about it once AS is out +Tentative patch that uses the new API being developed in +Therefore it seems totally valid to make a copy of the Configuration +See last comments on +This was fixed when was merged with which is now called closed as resolved and released +The attached patch addresses the issue by changing checkPath to check if the path does not contain a port but this FS URI does contain a port +bulk close of all resolved issues prior to release. +Patch against trunkcomponents + for the patch +Fixed in trunk +patch which fix this had been reviewed and ready to check in +Maybe windows sockets do not in fact block or it is a non issue on these am therefore closing as fixed on linux and not applicable on Windows. +Try to use attachTo attribute instead +After discussion with alan and matt we agreed to put this into as wellG +Thanks for the issue report +Tested with WAS runtime. +I have a junit test case that reproduce the issue so im trying to find a fix +alternative solution as discussed on the mailing patch has been applied many thanks +Indeed the function returns before savansubprocessorfree get the chance to be called +Maybe someone should let them knowCheers other comment just for clarity the root of the problem isnt in Commons Digester its Commons Digesters use of a JDK implementation that ultimately causes a file to remain locked under Windows +I would like to keep it in +Ill sort this +bulk defer minor issues to +That was the original decision yeah but since there are newer fix versions for Camel Karaf and Spring available we might as well use those instead of releasing a new with version of those dependenciesAnyway Ill gladly revert this change if anyone sees a problem with upgrading these things +The issues pointed have been mostly addressed by other patches +When you create a keystoreits key is locked +As fragments cant be started fileinstall ignore those and they stay in the installed state +jrubyrc file or similar too +Is there info on next expected release +Netty JBM remoting implementation passes all integration tests now +Thanks for the patch Lars +There was a commit not in Jirasee svn commit r and commentsthen another commit after some discussionfinally did a patch that covered the discussion. +The fixes for also eliminates this +Should return? That way caller doesnt have to cast +It is possible to export some repositories to another proxy via a file copy +The only thing that is additionally checked for in the is whether the pid passed is the process group leader +I kinda wish that we could do something better than that or a missing file is kinda strange isnt it? Why is it missing? Just like the code thats printing the exception were missing contextAnyways Ill commit this +We found an easy solution always remove the serverall folder +The vast majority will probably be in the latter case +Btw will a new snapshot be published automatically? +The functional spec lists BOOLEAN to and from CLOB as being a valid cast +th code base +Likely something will be skipped but the rest of the PDF presumably could deliver its text +This is not an exhaustive list and there are more than thismvn clean test Dtesttestmvn clean test Dtesttestmvn clean test Dtest when running the above three tests with jdk I am able to make the tests fail +But yeah harmless right now +Can a BZ dumpfile be loaded into JIRA? If so well have a quick discussion on the list on this +Hey RaphaelThank you very much for the update +If you find further issues please create a new jira and link it to this one. +Note that emptydir doesnt get included in the target zip file +On the other hand already has the dependency on resourcemanager for the test phase +Test can not be executed on mysql as it uses sequence key generator which is not supported by mysql +Pls review the patch +NET Framework has changed the exception message generated by the class +Ill update the JIRA description with errorThanks +Unassigning so that committer can pickup +for attaching the patch Alex +Are there any errors recorded in the broker logs +I can see your point about keeping old ones but still think it is best not to have beta versions in your product that is considered a final release version +What is strange about this is that both link have the same id +I will soon check from a licencing point of vIew +If not pls feel free to move to +The error being logged seems to happen for a lot of the interaction by the console +Although the race conditions that are fixed by this patch were never observed in our development environment they were lurking there waiting to be foundOur test cases still run in our environment with this change +I attached my you will see I would rather have only overriden get instead of overriding do +This has nothing to do with fop but with the xslt transformation that is used to generate the fo file +That would seem to be a misconfiguration somewhere in your might try reliniking to to ensure it works +Note editor widgets become enabled +We dont want things from suddenly break for users because they upgraded and were using a as a configured iteratorOh right I totally spaced on that +See for further details + +java files and that eclipse is aware of them +In order to solve this might I suggest removing the idempotent options entirely and instead implementing an Idempotent which can then be injected into the ftp component +Maybe we should update the project description in trunk to avoid this +Great but can you attach a test case in form of a Maven project or a WAR file with sourceThank you +Removed DCL idiom from. +De logging in debug mode is added in attachment +This defect applies to POM interpolation as this fix supposed to be in RC? I find my properties are not read by maven +retrospectively setting fix version in preparation for SDO issues fixed in SDO Java M in preparation for M release +Unfortunately just adding new constructors for the classes quickly became a big mess so instead Im using an Args class the same way its done for the Server classes and deprecated other constructorsPasses all the java unit tests +I am unaware of any automated jira patch jenkins commit patch plugins or flows that are available +This is still affecting Grails +No action no stack no INYou shouldnt be accessing a JSP page directory anyway +for the patchHan +Can you please shed some light on the part of the design which alows for data not to be available during view change and thus why this issue was rejected? Thanks +More than a few hours have passed but the typo is still there +Mike M docs +Nick +Can you give me a link where I can read how to do this? +Nick the only immediate reason I can think of where wed need these features in the TP is ifwhen we have tests depending on them +It looks like Cygwin Java is a difficult combination +Brian Thanks thats what I thought but I wasnt sure because the error message was listed in the description and I was confused as to why it would be there if it was truly just an upgrade +Issue closed again after Release Notes field was updated. +addressing most of Nings comments except the two places I commented on review board +This appears to often result in sporadic regression failures in other components as a +So we can target then with this change Click +Added as Cannot Reproduce for now +noreply implemented and added clustered line starting options need implementing plus re also some small lying around. +So it should be compatible with that framework +As there is restrictions in licenses we cant of course attach it are still looking for a simple EARWAR application isolating the behavior that makes Tomcats thread to still holding a reference +Before few months ago we were changing their names into names for developers understanding +For me to rollback my fix please give me one example that was working before and is not working now due to current JIRAs all you say makes sense but the improvements need their new JIRA +Thanks dude +I think weve just been talking across purposes? Try adding my test to and run the full test suite youll see the problem +If you find further issues please create a new jira and link it to this one. +Gerhard did you replace only the old wfs jar with +I doubt it +Is our generated website in SVN +Well be forking after release soon and these are the kinds of recommendations were looking for in the next version +Fixing Affects and Fix in Version +It will only be deleted or if the table is deleted +inside component there can be only one Web Service Node +I used scenario described on related project jiratested with EAP ER tested with EAP GA +Will add the header and do a unit test +Patch looks good except for the point that you unnecessarily made an abstract class +I sent one message to the less active topic and it was received by the consumer +Will have to check if WAS is JDK compliant or not +Only after the lock is found will the owner of the session be checked and Brians change makes the code look for the correct lock +Logged In YES useridYes there is a problem +Other than that the patch looks good +So these kind of features must be enabled at the host environment +It has to be applied after +make bugs state match its outcome. +Let me double check +I committed the fix for the concurrency bug and the incorrect reload starting with empty map instead of old after release. +Yes this seems to occur with all projectsYes No other services seem to be could you provide the steps or the project to help us reproduce this issueThanks Luke so if my understanding is correct if you launch a new analysis on each project this should fix the issue +I dont think we are committing to more support for piggybank +When this is copying will stop as soon as the named file exists +The old Select Pro is no more supported and it has been replaced by the new Select JuniorI have integrated it into this new patch and it supports also the new Payment Gatetway Config entitiesIts very similar to the payment gateway and now its no more necessary any jars from RBS so I have removed it from of accounting componentThanks into rev +Selection bar visibility has been fixedResolved +More specifically when unpacking on linux and the files have specific permissions for the group in the archive they are incorrect after the unpackI believe the reason is that class does not support Posix style permissions and obviously this is what is used in the dependency pluginThere are two ways of fixing itFirst way is to use package that supports Posix permissions but this is available only since Java and so this fix will not work with Java +on linux as isues related to new p enabled product build are moved to is a problem about updating file under mac if altervative jvm selected but it is a different on JBDS. +bq +If some of them can be done automatically maybe we should do this in two steps +resolved in wildcard address is now used for the tcp transport +fixed in svn r +performRelaxed or just perform is probablymore appropriate WDYT +last stack trace looks like a issue +Okay all my Samsung devices work correctly +That gives best of two worlds that for some reason is not possible then yes it should be showing a you wrote you did some easy fix what is that? Looks like it is now saying not correct project correctcould you create patch for and attach +Is this on track to merge to? It is currently the only diff btw the two versions +Closing resolved issues for already released versions +Oh and more important is this only if they switch to use JDK? Or not +The functionality has changed for the searches that include wildcards so one unit test needed a fixIve also submitted the unit test from the first Mikes patchHope everyone is happy now +Upload patch v that uses a shim to get +Karl if youre using you can simply add your user to the admin group in there +I appears no header was ever on the file when we pulled it so just listing the source license and leaving the header out +Looks like I havent found all situations where this exception is occurring +Are there any exceptionsmessages in the log file to help us troubleshoot this one? What do you mean by first dynamic deployment? That seems to imply it works for subsequent starts of the bundle +I can live without that +Many thanks to all +It was just a temp workaround for duplicate messages +And there is no specific task at he moment +This change has modified examples to bring in dependency +Verified by Leo. +Arpit thank you for the code review. +The current fix helpedin this caseI think it makes sense to create a new issue if you only had an empty indexesX fileThat shouldnt happen because the old indexes generation file is only deleted afterthe new generation is written. +batch update of issues with fix instead of the correct +This is an issue with how we wrote the mobile showcase not the component support itself +tests passed on my local machine for this Y! patch +If Maven were the only one executing these statements Id be done + unscheduled issues to the next release +son of a biscuit +Im also curious about the use case +TOOLPATH is not defined yet +That would keep it binary incompatible +is not of my making +It could indicate a race in the client or application you were using +transferBlocks could take multiple targets +Will try fine when I run the provided code in the version that I used Ubuntu linux with testing on Firefox +So it might result in TIMEWAIT or FINWAIT sockets on the client but shouldnt impact anymore +Ok I just added the round robin details to the doc source. +was committed with revision +Attaching for codeline +It seems like possibly changing to use a pool of Packet objects for writing might address this but Im not sure I fully grasp the full problem yet +Please review. +Floarian tnx for the quick replyIs the problem based on property not being understand by other WS implementationsCan you point me to the code with RI specific API +Gregory your evaluation is correct +Sorry ScottIm not agains the patches idea I just dont understand why the other solution would only work for a while +Please let me know if you notice any problems. +Also if you arent planning to submit unit tests perhaps this should go in the srccontrib directory +Closing again since the new content has appeared in the Latest Alpha Manuals. +This will result in more data transfer than required as each entry will be read from at least bookies +Additionally its important to annotate the implementing class with not the interface +It will pick up any conflicts where two different similarity implementations are used on the same directory at runtime if a mismatch occursThe solution also feeds into in that as part of the patch it makes it possible to close searchers rather than readers and also to retrieve all the readers for a searcher no matter how it is constructed +Ive modified the bean to inject the Messages bean in the event Ive already verified that Weld fixes the issue +I didnt have a chance to look at the implementation in detail but seems right. +I saw these by just redeploying daytrader a couple times with IIRC not doing any work in daytrader +Some minor changes may be needed in this patch +mvn clean verify tested +this was only introduced in a recent commit on thanks +Comimtied. +i had not run the tests enough we intentionally return less often +cs files are broken up into their individual counterparts +Files containing +This was due to an int overflow in Packeds bulk get and set methods +Fixed with svn revision a small fix for the test. +Thanks for pointing this out! I guess Mark and I oversaw this fact while refactoring +All sub tasks thanks a lot! +Ideally this could be done via the invoker result but the s handlers dont have access to the invoker result in order to set it +for the patch +I will check in a change in abot minutes +After discussing with Tom F +Ill set minor priority +Treat it as a normal exception retry it every once in a while +Committed to +Trivial patch attached to symlink HCATPREFIXsharehcatalogconf to etchcatalog behaviour now mirrors installation +Issue resolved mark it up to user inexperience. +I have ant on all the machines at IBM +Fixed. +I just committed this. +I see this on trunk with a fresh build +Revision +revision . +If its ok Ill work on ftl file for this and submit the patch laterIs it possible to get the current patch go into svn and deal with Overview screen as a saperate Issue? order to fix invoiceItems form I think should write it in ftl instead of form widgets +For examples see classpath of +It seems that your main problem is that the service activator that is renaming the file is invoked before the message reaches the jms could use a Channel Interceptor and rename the file in the postSend method +I can also say that custom types now are saved as embedded attributes so this is a real fix this problem but I have some problems in Inspector I think I will fix it in one or two hours +I guess its speed am I rightYes the whole point is to avoid full scans when possible regardless of whether were targeting a real table or a memory backed construct like the system tables +Thanks Tim +Do you plan to add any test cases for this issue to the replication test suite +Patch from Ate applied +Marking it wont fix for the reasons stated in last few comments. +I used as local database and adjusted the parameters accordinglyWhen I create tables I can see them in the metastore via so it definitely is working + without integration tests. +I think youve rejected this prematurely +You could obviously make it smarter by parsing the header and only refreshing the corresponding hosts +Ive commented out the part of the test that continues to fail its assertions in both trunk and branch because frankly Im sick of seeing this test fail unexpectedly in Hudson and waste my time. +Ah +Review available +Fixed in rApplied a patch by etc +Lets move this to +fixes the mentioned issuesalso slipped in a small addJobs method to add more then one job at once +Isnt this issue related to +Backport to revision +But I Dont see where to place the hook +Ruel can you give us a pointer to an example of a script that builds a server configuration in the jboss test environment? We have such a script but dont know how this should be integrated into the jboss test i think that we should create jbpm server configuration which is based on the all configuration jBPM jPDL jBPM we should be able to run the whole jboss test suite to make sure that our jbpm deployments dont break we should start building our own test tom +HTH please lt us know what happensThanks +I recall isolating and Mario fixing at least +It would also make sense to do the same for javadocs +Works fine now. +this is full but draft Portal coreCMS translation path is version is out of date and target cmsforums +I never succeeded to run the installer and login using Apache Derby with the current revision +patch to demonstrating patch showing same bug when exclude is a containing a +Sorry +However you should be able to look at it and determine what should be done +Enis should be cleaned up on cluster shutdown? I thought I fixed that a day or so ago. +What is your database platform +In the Java editor in the menu it correctly shows the keyboard shortcut as +Well we load scripts off the classpath already so why not off of jars added to the classpath dynamically +Patch for second option +Sounds ok to me +Thanks! +The next level would be and probably a higher level which gives all access which a java gets asynchronous version of I see it +Ralf theres a thread in the user mailing list archive that discusses the use of JDK vesions +Thanks +I also made the two href links external so that in the HTML pages version the link opens in a new window instead of in the same frame as the topicThe updated diffs file is and the new HTML file is +Attached new files with the Classifier filter added +if it happens again ill report it again +Well look at later +Since its so close to the end of the month Im going to let it do its thing then split the old mail archive files later +User assigned to on jira +Both versions behave as if the default operator is OR +Report level parameter at the for your patch Christian +Reopening to bind this JIRA with Bugzilla. +Thanks DenisIve applied the two patches in r +bq +This patch modified javalangreflect to make it RI compatible +On mailing list some discussion happened +I wasnt sure how much of an overhead these comparisons would result in +These are probably all obsolete +Ive committed the changes in the updated patch +Do we still have duplicate issue with Checkpoint file +Fixed in trunk and BranchFrancesco thanks for the fix and test case +Im marking this one duplicate of as its been fixed in for that bug +comitted a fixThanks to close to confirmpaul +My best guess is its something to do with permissions +verbose has been added and for most of the commands the response headers are printed if they are present in the operation result. +If we show the inherited interface as perhaps we decorate it with a lock to indicate that its not editable +Hey Eric this wasnt fixed in +Thanks for looking into this Kim +Throwing from is not perfect but okay for me +Adding new version of the patch since the first contained an unintended have also seen much variation when running the test client but I dont think this is specific to the client from +It is because I create different class loader in one JVM to load derby driver so cause this error +Duplicates +Committed revision . +The same code used to work with tomcat +Fix and small optimization +A file manager in Solr is way too much for a stable branch especially if it has no security at all +My bad +Thanks Bhallamudi +Removing Fix Version + +Note that this change doesnt introduce any new semantics per se +And thats why you shouldnt use work email accounts for OSS participation +Hi Darynbq +IMHO a better approach is without AJAX perhaps a new component similar to the radiogroup all you really need is a way to link the lists and be able to filter on them +This is what I observed when running Hive testcases on both MR and MR +Probably worth a mention on the Getting Started page +Which would take precedence over the transitive Groovy dependency in +Lets also make +Submodules were removed so this is no longer an issue. +for the patchI committed the patch to trunk and +Changing the code to make it work correctly was not the a big problem +I think is orthogonal to this JIRA +Id suggest removing the Groovy Templates preference screen until this feature is implemented if possible +It compiles against trunk as well +Yes thats what I assumed +LGTM + for next release +I improved the concerning the +Something like this worked well for me +I dont think we need to revertAs I showed in the check for valid META table row should be changed to account for two commas which are not adjacent +Verified on EAP that the integration Seam with Hornetq works fine without any issues +Heres a slightly cleaner version of the patch against the latest trunkbtw this does not require changes to the generated code at all. +Also I have tried this with the HSQL and Postgres databases behind jBPM +Yes it appears to be the same Jetty bug and seems fixed now. +Thats a start on this but we still need to update the deploy tool to use it +This patch addresses Davids issues and as a side effect cleans up some confusing stuff in +Ill talk to him and we will coordinate further action +Im not sure where this bug belongs at this point +because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +So please discard the two patches I submitted earlier +Im developing this patch against the branch that were running +The most important to me is that great contributions like this are being welcomed and brought out to Solr users somehow whether the code lives here or there from the start +The current script is a dead end it is too slow and a way overcomplicated +minor comments on phabricator +I think REQUIRES is probably the way to go +Im assuming Infinispan is the second level cache or are you using Infinispan Query as well +The intent was to include all the libjee jars in the same version as the ones provided by upstream if not exactly the ones included by upstream +Whenever echo service not available it sends Service Not found error with a soap faultHowever I didnt come across the problem that you have mentioned +This is in reference to the graphs from +Both me and Andre verified that without it all is fineBut for now I cannot say why IE doesnt like the random parameter +Versions should be changed from to for trunk usage +The reason to do this is that rebooting the whole piece may be an overhead work +all necessary screenshots are updated +Yes +Is REST secured enough for such feature +This sort of fixes the highlighting issue +Thank you Sandy and Tom +Ive tested and committed the workaround thanks this while waiting for Sun to apply you patch +Thanks Jerome. +Assume closed as resolved and released +You can most definitely not put this into Shindig until the spec is changed and sending an email to the spec mailing list does not mean that the spec has been updated +Since we have a separate protocol for haadmin Acls are easy to configure +Is the information currently in the nightly build version of the docs under the area erroneously labeled Messages correct as related to this issue +Id take a wiredump with something like tcpdump to see what the actual size of the request is +Would be helpful if you could create a test case that demonstrates the issue +The same goes for Remove Remove is that addingremoving mime type at would really complicate implementationaddremove applies new mime type to container but container only propagates that data to deployed application at deploy time. +Code looks good +ThanksNo problems I will simply integrate the important lines +Right they should have been initialized but somewhere it seems they were not +Should be in generic only +The code that Jian wrote is working on the RM as well as the AM I tested it +roo database introspect schema devdbSpring Roo automatic discovery service currently unavailableJDBC driver not available for roo +I just corrected the gradients +As you mentioned this was only a work around in the db driverThe confusing part is JDBC Client tools like Squirrel showed the Column Name as the header in their UI not sure if that a mistake or deliberate +Frederik fixed this last week on trunk +Once I tell it where my JVM is installed set the project to use it then it compiles although with warnings +Essentially we are waiting there for the next jbossts upgrade. +Taking a look at this now +This will build locally and install it in your local repoYes right +patch +See But I see that is also required to be supported +Is there any chance you can make your WFS instance publicly available so we can test against itIn the mean time I will try and find another Geomedia Web Map instance that I can try +This is the file for which i am unable to get the problem was caused by a rare MP byte pattern that happened to also match the UCS LE byte order mark +anyone +This patch is with the fix for the javadoc to get the new patch through +Hi KevinFrom the specs of java and java it seems no much difference so I suppose its bug of RI and I will apply it to java trunk is it ok for you +Im logged in as sebb but I can still only see the projects +Thanks Claus! Looking forward to. +Thanks Chris +I had something else running on my box on port preventing to start for mumak +Patrick just because jira creator requested this it may or may not be the correct solution +It will touch actually unchanged poms which we are trying to avoid +Are you sure you updated your +And so this returns the wrong project when a resource from outside the desired project is selected +Test case for simulating the is failing with both and kiranAs you know this issue is fixed last week +The fixes will be checked in for follow on release documents but gonna close this one for JBDS release +In upstream these concerns about official support levels dont apply +Thanks I will check now +Any chance of testing with CXF? If its still a problem there any chance of getting a small test project that shows the issueCXF really isnt tested at all with Spring +I think this problem has its root in the +missing X code to reproduce so keeping issue open not useful. +I missed to disable the check in the test +The artifacts have the so renamed to what is expected and a symlink to the old name alsoThis required some changes to and the libpthread and libstdc to the linker path +Test +Moving items not being worked on for M afaik out of issue out as part of JIRA cleanup +Will try and figure out what Acegi is doing because it seems to remove the headerWe should probably use Spring security instead as that seems to be the recommended library +Also you will have problems with pooling for other field injected things. +you are right null code completion is silly Ill get rid of if +At this point I am not desperate I do plan to do a new release of my app at the end of the next week but if it is not fixed at this point I can always revert to my old which I in Compass and Lucene are very useful for me though so I really need to update to final at some me know how your investigations turn out +Hi KarlI just double checked my build config and all deps are there and I always copy the lapi jar to both foldersI am having trouble syncing with the SVN repository for some reason +Itwould also be fine if somebody got the WIN to work I couldnt +Maybe we could continue to show short messages for a default timeout tough? And then having a parameter for this timeout somewhere could help +why do you need the C headers for thrift here? without all of thrift they cannot compilerun anyway +Aleks could you post the patch here +I found this useful when narrowing down the cause of the fetch failures in my problematic cluster +So then coretests can be replaced with integrationtests in the parent pom? Shall I ask mistria for that +Maybe this issue could be resolved by simply warning the user when the is not going to be created because no classes exist in the. +The output looked better in the than it did in the PDF +Denis Malarevich will apply this patch and fix slink and houtputText +Yes we have tried that and so far it seems OKHowever we are not sure what the perf implications areOur app has very strict response time +Any status on the bug? Its been hanging around for a while +In the Ive made begins with and ignore case options the default valuesIf you all think its ok I can apply this patch too +Hiding in the UI is fine. +Did you actually try a version? I wonder if the heap is filled also +Confirmed that this is still an issue on Android +If I look at the connections on one router I see a connection to the other +Dag I think we should go ahead and commit your patch into trunk +cdi jars are listed and correctly taken into accountThis is tested under linux windows and both works fine +libgriddevbin +Im not very clear on how sessions work but I dont think invalidating a session logs you out +This change didnt go into the release. +First off this is a highly suspect bit of code its making expectations about how the call stack is structured and what format it will take +Not going to backport job changes and code changes into the stream +A bit roundabout but at least this would only happen on cache startup +Thanks Jason +Set Fix version to +In addition I have not been able to understand the defect the patch is trying to solve and that I cant reproduce itRegards patch invalidates the previous one and has a localized test data +Sorry to join the conversation so late +Asokan +Ah indeed +UghLooks good Uwe +Sending trunksrcmainjavaorgapacheservicemixmavenpluginjbiSending trunksrcmainresourcesplexusTransmitting file dataCommitted revision +Faint memory of the whole thing returning though +Please be sure to subscribe to repository to follow any discussion regarding the repository such as upgrades etc. +The steps for Required are described in section The container must invoke an enterprise bean method whose transaction attribute is set to Requiredwith a valid transaction a client invokes the enterprise beans method while the client is associated with a transaction contextthe container invokes the enterprise beans method in the clients transaction the client invokes the enterprise beans method while the client is not associated with a transactioncontext the container automatically starts a new transaction before delegating a method call to theenterprise bean business method +Thanks Lohit! +Thanks for comment Chris. +verified in Developer Studio to set release note notes are not required for as release note status has been set +After some testing look like the sessions arent flushed correctly if a buffer is too big its flushed few time correctly and then its flushed at the speed of the select timeout +Not to be committed +Is it an invariant that will always be in boundsYes in this caseThe reason I did this is for is etc is because those methods are public +Here is a patch that resolves this +AgreedThis initiative is on the right track +Sure no problem +PR for master merged in master branch +I suppose you are going to subsume tag within the row as that makes serialization and deserialization of tag easy +The only guaranteed way to achieve the desired behaviour is not to allow users to configure it they cant be trusted to ensure all server instances use a consistent value +only a few are and require the back compat code +OK +Using the Dynamic bloom filter seems like a reasonable way to go the only thing I can see is that we are still going to have an overhead even though it is smaller than nowSo if possible wait until we know the exact number and then create the filter +I think its better to replace to and use single quote as guardYour string should be escaped like Cant touch this +Rebasing +Sample test case using Camel Nmr in combination with Camel FTPYou need to do unzip and enter the folder mvn install copy the +Ill let you know when Ive done them +Closing this issue. +Hadoop code needs a lot of improvement in this area +I tried on mine and build failed yesterday +The hard coding is just coming from the fact that by introducing Product not listed as a systematic item in the products of a SKU you introduce an anomaly that the dynamic cannot deal with except if you hard code I agree +I have also run all the hadoop junit tests against my patched +Ive changed to a much much better solution that is a little more effort +Resolving as it is not the code but JDK bizarre! Anyway given it appears to be an isolated issue there is no need to mount a full scale investigation at this point unless we start getting similar reports +OK thanks Uwe +underlying issue fixed with seekTo issue fixed with patch neglected to close +Any objections to applying these changes +Patch looks goods another loop above this one which is to weed out that dont need or cannot be speculated +trunk fix coming in next day or so +This is a key enabler for providing client authentication and secure connections to title to reflect discussed approach +fixed in headSending configsSending configsclientSending configsSending configsSending configsSending configsSending configsSending configsSending configsdirectorySending configsSending configsSending configsSending configssrcplanSending configsSending configsSending configsjettySending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configsSending configstomcatSending configsSending configsSending configsSending configsSending configsTransmitting file dataCommitted revision +New patch improved documentation I added to the list of automatically tested codecs in a few bug fixesPlease let me know if you would like to review this patch before I commit +. +We have Seam booking after the patch and it worked fine +Thanks Arun for contributing and Xu for testing +Matthew that path of code that appears in the posted stracktrace is only executed if you keep JDT weaving disabled +Eliminating the searcher field will elimninate the need for to be Serializable therefore retracting that changeI think thats right and that the best fix is too conform the new Weights tothe new protocol for compatibility +Unresolved as of brought forward to visibility +Should probably be fixed for as a release cannot go on with failed tests +Everything is fine if downloaded and opened in Adobe Reader +I am on vacation from to If you have any question on deployment and JEE bugs please contact Saurabh Arora or my manager Maruthi NuthikattuFor emergency contact me at +I saw that too +Note that this issue is intermittent and only show up on my machine perhaps half the time I run the test program +verified by Svetlana +Hi As you are based in Europe you should try and rsync from +Still having now +Im not sure if its related to this issue +A small program demonstrating the bug and its associated stack trace +Or riposte as the case may beIf the replies are general rather than line by line you can also just put them here +Im going to defer this JIRA for a later release when we do have a defined runtime +This feels very much like apologies +thank you Tim +Thanks Corinne! +Thank you Clinton. +Robert +serverproductiondeploymanagementimages +I should be more specific here as to what the current problems are methods not proposed sometimes fields or properties from incorrect types are proposed a getter or isser variant of a field is proposed but is proposed as a local variable not as a method +Nice +I would drop most of the properties that specify versions +Simple application that demonstrates the issue +Yes there is a good chanceAlright I hear you guys +Perhaps you can suggest a work around +Sorry for the noise +Vijay didnt you run into similar tar installed on the effected system +Graeme are you consider that FROM Order is a bad queryWhy I should use executeQuery method insteadChange regex to ignore case and search entity class name between FROM and WHERE clause +Im closing this as fixed +I agree this is a important feature and we will start working into it +A different approach is used instead + +why deprecation since it is introduced early this cycle why can not we remove it +to cache archive and run job +The code in that saves projects is skipping over any resource that is +But we may end up with this if for example the regions are +Client will use MD +Fixed in trunk and branch +Ok fine +The intention was to set up a global property resolver +However users may not always want to know all the info of an application such as the embedded right? Sometimes users just want to fetch partial information of an application to speed up the response +Ill ping you ifwhen I can start working on customer that needs thisInstall a generic remote instance on a host with the setting pointing to a generic Domain from the CLI configure the remote instance to either become a domain controller for a new domain have its set to point to the DC of the new domain +recommit it +need translation +Comments on JIRA arent sufficient for something so is a secure server +and what about if any component threw an exception when pico tryied to stopdispose it? A PIE with multiple causes or throw just the last one +Due to constraints at work I dont think I will get to test for at least a week or +Apparently you cant reuse filenames trying again +This will most likely need to be addressed in when we develop the service module dependency framework +Agree that this is mainly caused by the improper implementation from but is there anything Hive can do to better cope with this kind of unexpected behavior or prevent it from happening again? To provide clear documentations for its related like that in +The patch looks good to me also +please can you check the jar you have attachedI have replaced this jar into lib of my war and also in my incubating modules +That leaves and MTOM +Im not sure if it was meant to be there or not +Im trying this now and will update on. +Had to add a few types to the defaultI realize that the tests specific to are redundant theyre caught by the double test + +I already attached the server log as well that was obtained while running the single test which should provide deployment information +We discussed it while we are implementing in BKJM +Trejkaz if you think theres more that needs to be done for this feel free to reopen this bug or open a new one with more specifics +My patch is very straightforward and simple so feel free to integratemodify it with yours +mrepo +We need to have the skins step instead of setting a default value when an user is created +works perfectly fine on other small projects or samples on this machine it just starting having problems with the real project +please attach the patch if you already have fixAll tests passed Local +Hi Spark Patch applied at +I dont think thats a big deal and I think it can be retained in a way that still brings some overall code clarity security and even reduction +Patch looks good to me too overallHave very minor edits +Some of the avro stuff looks like it could be with some refactoring +Concurrent readwrite to a is synchronized here +Well we should allow to get converted to a Float +Smoketest is fine alsoI have also verified that the patch resolve the orginal reported problemI approve this patchThanks +possibly a thing of the window manager or the font settings? Also you have a monitors setup +Thanks Todd +I think this could be best solved by factoring out a interface +The env is setup and a process is forked just like it is inside the cluster +The above stack trace is from trying to mount the repository from to islolate any potential weirdness introduced in our own code +Acceptable imho as the time in the console is ok +I dont think we need to run more testsCan one of the committer please review +We need to resolve this issue ASAP +This patch includes a working script +This patch correct it + +Patch based on Gasols which addresses the missing in and also inits the defaultReady to go + +in case nobodys noticed the build is currently broken because of this +Moving out of +sorry to hijack the thread +some file formats which will be used in joins could benefit from HDFS trying to place the replicas of a few separate files on the same set of +The test fails without your change and succeeds with your changeTouches the following filesM javaengineorgapachederbyimplservicesreflectBretts fixM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangAdds a restricted table function to and verifies that it loads and runs +Virtual svn add +Sorry +Can you please stop marking the Jira as resolved +Committed revision +While the block receiver is blocked the packet responder is still alive and sends heartbeats back periodically +Looks like timed out +Attaching test file same as pasted in the issue description +API code and response looks ok +fixed in commit frank suggests is on commit on master bccffbafbdcabThanks Frank for fixing the issue +working on a interactive shell with bash like scripting support +Moved to next release. +Pushing to +Test data which was included in the patch was not applied to committed code +This may seem a little lame but we have resolved this issue by removing the section from the docs in the upcoming release +I just added the ASF licence to class +all issues to the CR +To be more specific my latest attachments were +It is being run on a machine and not a VMI am not understanding why you are ready to give up already +I guess what you would like is to hande the Type in Find Example by replacing the dropdown box by checkboxes +Original third point implements the JPA Spec API while implements the JPA Spec APIFor you can also use the JPA Spec API jar from the Apache Geronimo project whereas weve already switched to the Geronimo version for +I think this was changed in another issue because there is no zookeeper in and command line shows localhost starting zookeeper +Regression test is committed at +This causes my dependencies not to be injected +Can someone review the code in the current state and comment +edit updated Summary and Description to indicate this only happens when HA is enabled on the secure cluster +Any updates on this issue +At that point I dont think the characters saved justify the ambiguity over whats happening +You could check in but the regionserver might be added to the deadservers between your check and the add of the regionserver to onlineservers +For the record embedding private functionality is semantically very different from importing public functionality +YongxingCould you please test this patch and let me know if that fixes the problem? it works +Yes +Robbie Im okay with that +Forgot to update it +There is a that can create them but curiously there seems to be no code to unpack them +after adding alternate solution +Heres the result of netstat an just after the crashActive Connections Proto Local Address Foreign Address State TCP LISTENING TCP LISTENING TCP LISTENING TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP LISTENING TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP LISTENING TCP CLOSEWAIT TCP LISTENING TCP TIMEWAIT TCP ESTABLISHED TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT TCP TIMEWAIT UDP UDP UDP UDP UDP UDP UDP UDP UDP +The current way confuses users +Perhaps this issue should be closed then if the problem is located entirely within and the patch attached here does not relate to the problem described +Did you do something different this timeDo we have a flaky migration process +Bumped off +Whats the status of dumping the code into the apache svn +Lets do the ones in this +Thanks for moving the issue JukkaMario can you add the option in your file please Ill then add the mirror everything else looking good +Thanks for catching this so quickly. +If the problem should swing back around to the Network Server side I am glad to continue contributing there +I manually added the srcmaingroovy folder did a refresh of the project config and now it adds targetmain as a source folder as well which casuses problems when a maven clean is run +Yes everything is now but there is time to shift over +I dont know their names so I cant configure it using file +Apparently this goes all the way back to but occurs very rarely +Does this issue still apply to Struts with Guice? Im getting an NPE in the there but Im not entirely sure if it is this issue +Ill take care of it +Thanks Will! +Added synchronization to make thread safe. +added cfdbinfo to restart container and retry once or twiceit does not fail every time for me but nearly every time +Dartifactsvarsecurity the value must be a absolute path +RightThe decision to enable or disable dynamic reduce concurrency will be set by the user +Committed to trunk and. +Needs to be applied to the codebaseBill +As per the cache must be disabled by default and I dont see why you would get this unless you have explicitly enabled it +Please try a simpler filename as requested +When should we stop optimizingThis is a relevant question +Updated the under the plugins project +Issue is still relevant +The smtp gmail is just for the illustrationI dont use it at allIm using a entreprise smtp host +Rev applied the changes to the branch +Also I noticed that in trunk a node wont broadcast a tree to neighbors unless they requested it which means that when a major compaction happens naturally the tree is wasted +is certainly appears to be a tradeoff of memory vs performance +If there is a migration tool should remove configuration from JCRCONFIG instead of remove workspace from repository because the workspace will be till used if wsrp is enable in product +Lance your patch misses the all important changes that both Noble and I contributed in our patches +All active user code should be working against the new usergroup servicestore interface +Confirmed after building GWC locally the GS build works fine +I dont think this update is critical unless it introduces an API improvement or bug fix we need +Otherwise the test would become flaky again +Closing all old Resolved issues +If it did we store the first evaluation and then jump on subsequent evaluations so the closure is in effect only evaluated once at runtime +Also the two Agenda Groups should be B to D and B to C +thats excellent +will explain more in next comment +Closing resolved issues. +Attaching the patch for release With the patch test cases fail +Can we get rid of get and get from the client side translator and server side implementation +change compaction logging to log all files unless it was explicitly called major by systemInterestingly we actually dont log major when we just have all files except after split +We fixed strlcpy by including it with an ifdef previously +The poor naming is one of things that I find a block to using this facility for real +tables not used by are tables shared by both andOne more favor from you please +Setting to two after a reboot wont recover informationOk for the textOk for the split with the address Ills see what I can do +First few steps in revision has been implemented over time +Patch into before it was officially releasedCommitted for Kamran. +When is this printed +Throws if invocation handler is not closeable +This is a patch that adds the functionality for LIMIT and OFFSET calls to the derby database instead of performing the limiting and offsetting in requires that you put the file from the latest release of derby in the driversderbylib folder as for the tests to pass +Please review the attached patch +Closing as duplicate of +Perhaps just a wording issue more clearly indicating the content under the template is what is copied +I removed this dependency and I dont get the error message anymore +After thinking on this aspect it seemed to make sense to not cancel the tokens that the didnt own in the first place +Move it back in if you think differently. +Like before my patch is against branchxYonik you may be right about lastAccessed +We need to resolve this before release +So all issues not directly assigned to an AS release are being closed +Im registered username is seelmann +Do you have a list of serialzables which would be affected by this change +Looks like Pinaki beat me to the commit +correction sAaronrAlexOther thing these entry points effectively become the way to startstop Hadoop clusters +Job history server cannot give real time MR metrics +Both this patch and are needed to make mapreduce compile offline against local version of +In reality nothing prevents us from releasing next week if there is a strong need for this bugfix is already deployed as a +Hi Jaredall examples are working properly in EAP CP +I wonder if a possible workaround would be to derive a class from which has a default constructor then bind to this instance +Resolving this issue because is over and the integration is done. +Patch from Devaraj which fixes tests on windows and MAC +Patch fixing the issue +May be another one for easily creating should be done +davex is an extension to webdav and i doubt that any application in the browser will use this + +Okay the patch has been comitted. +The reason this isnt an issue in EAPASAS is that modcluster listens for JMX notifications emitted by the server before it shuts down +But if we want to enable all of the tests to run we first need to classify the tests between true unit tests that should be run as part of and integration tests that take longer but should not be run as part of the +Your lookupupdate wont be calling the createInvoiceitem to benefit +Needs to be fixed for SAX +Eduardo Millan Yes! We are very interesting in implementation of J LDAP Security working against Lotus Domino! Can you share you sourcesThanks +Just asking +If they build it we will sign. +ThanksI am working on a bigger patch that avoids all the casting and instanceof checking +for the patch +Proposing to anybody against this +To reproduce the issue Goto Add any product to cart click on Quick finalize order now select shipping address shipping method but dont select any payment option and now click on Continue to Final Order Review link +I think the bug is with usage of for the comparison +However subsystems will eventually need to implement some of it in order to make use of OBR +Need to see if MD parser issue has been solved +bulk defer of unresolved bugs to +Patch for for Ive attached a patch to include this functionality into it possible to include this in the next versionThanks +Ill go take it up with them +Additionaly I checked it also on RC and it looks exactly the same +i would suggest moving this to post and also provide a c counterpart to it +I just committed this +Applied to trunk as Rev. +This patch adds an extra fixed with revision +This inconsistency is often seen in a case where you have a line containing multiple italic words mixed with normal textWhat is the strategy for the subscripts and superscriptsIs there any update on this issue? Is this part of +bulk close of issues as part of making release notes. +Agree the problem of duplicate doc in the README may be a problem but the mismatched types make the problem hard to resolve +The numbers were almost as close to logging to disk +Patches supplied for the in r r r +as long as the user either specifies a Comparator instance that would be supplied to the or makes the domain object in question implement Comparable. +The code does not scan but it still need the unqualified jar name as per the specAnd yes the spec allows to mandate in +if Farons evaluation is correct then the bug is fixed in the current cvs trunk +bq +Merged to the branch in revision +i assume you meant and not +Closing issues +The check was from an earlier version of the patch +Can this issue be closed +That issue has been fixed in Bridge but That fix not yet backported to Portlet branch +I can see being stuck for some reason +Patch committed thans mojo was moved to the dependency plugin instead +Dont see those messages too +Simple patch +Zack The churn in the is a major problem in the Lucene code management +I didnt put in a timeout but it can be added later if needed +Thanks ortontrunk all Mark a test case and a patch is much welcome + +to commit +Denny can you look into this? Seems a bit odd +I think we can close this one since we decided to remove sitedocs subproject from the. +I am catching up with a lot of things so wont be able to review Cristis changes for next days but I am hoping that any further changes would be minor +Okay but even in that case you are adding classes not hiding them +The only thing which remains is to handle the timeoutAssigning back +Added some javadocs to suggesting the use of to add business processing in the case of an exception being thrown +The only solution for us was to serialize Workflow access +Fix system failure in the demo setup now has an diagram image attached to it. +OK I will make some tests and review this in the next release +Just to get this installed Camel in MQ usingfeaturesinstall camelfeaturesinstall then I copied the cbr JAR example from fuse enterprise to the deploy directory +Proposed thanks for the report your analysis and the patch +Would be nice to replace the hostnames with something like but I dont think it is criticalThougths +Thanks Dmitry! +FYI I am in the midst of an entity model overhaul that I hope to commit soon +I give another simple use case i want to import historical versions of the project +Attached some slides with a cartoon showing how this implementation is different from pc +I noticed sometimes similar problem with Eclipse on Linux +Only if expose management is checked will the management ports be used +Aah I see +Is there anywhere else in the API you think we should add this +But in the meanwhile we also should clear these block since if the corresponding file is only in snapshot no one will finalize the block I guess +Took me a minute but its jobs vs job +Patch applied with many thanks to Jon +So a first suggestion for one thing you could start with is factoring out the fix for the issue into an own patch and get that committed since it is a +Hopefully there is a standard approach so that we can implement it in Derby. +Can use the for the +Im unable to reproduce the problem using or and the testcase listed is anyone still experiencing this issue +If you expect gays to work on it I wont but you probably mean guys +Currently class is marked Stable +If you try this out Id appreciate your this bug since the initial problem you describe is fixed +Committed changes for AS and AS within the container integration code +Ivo that changes a lot of paths throughout the documentation +Ive committed based on the feedback here +bulk close of resolved issues. +Im beginning to look at it Ill post here as I make any progress so we should be able to work on it in parallel +just added a debug line +it doesnt test at all +Any good log messages you can suggest adding for this test +Bigger fish to fry elsewhere. +The problem here is that PS Level supports LZW compression which brings up licence issues but the compression is free +attached a status report for the st phase +Im confused why Jenkins is saying this is integrated theres no patch on this code out yet +Resolved +Please do not move in a supported module its just a attempt at caching if you do please add test and check the code really works add tests really it loads everything in memory dont use for anything but very small amounts of by using that class in uDig you literally planted a bomb inside it +is committed a fix for that moved the relevant entries to the section +If you could point me out to that I would appreciateFlavio +This is the WS stack supporting and do not support MTOM you looking for. +SergioAny thoughts on this? Sounds sensible and fairly straight forward +made the needed changes +Ok it has been close to a month and we havent seen the error again on any environment +Ported on +Done r +Fixed on trunk r +The implication is that during a hard crash if the high watermark file is missing one has to manually rename it from the temporary high watermark file +for st and rd hunks in the patch +Think internals needs to be changed to have various partitionsSupports multiple queries +This is only happening for those users who are accessing from a secure corporate network +We have a solid test case setup for stress testing NMSWell stress test it once the new Unit tests are added +Appears to be good enough. +Here is another way to do this enhancement by adding a Format processor after the Format processor will go through the nodes and take the namespaces to the parent dont know if there is any out of box tool class can do thought? it just a xml format enhancement I lower the priory +But the clients in use fixed port specifications and so are not aware of any ranging which had to go on to find an available that the test should be modified so that clirents are able to obtain the actual port obtained by the +All your points make sense for me +Its important update the download file remove the jdbc conector and add the problem was the jdbc connector +Thank you Nikolay TimHere is the regression tests for and +this is just a change to the file and doesnt require any tests +Sure I canSo I run the last branch and branch this patch and compare +Adding Ramnivas to Watch List for comment +Please provide the CCFR text and resolve when done +It improves test cases by adding a couple of more tests synchronizing this with trunks patch +Could I ask you to attach a complete stylesheet and input document the expected HTML document result and the output you actually see +how are you handling that may point at symbolsSo far I see no general execution performance hit from this though I wonder if perhaps having the threadlocal flag be compiled in might be the best route to take allowing embedders who want behavior to just do a build with the flag set to true and those others not interested to allow compiler peephole optimizations to eliminate all the conditional logic theres a severe performance hit to the during the test run it goes from a few seconds up to a couple minutes in run time and I have no explanation for it +Everything is working ok please take a look let me know please list as a mirror siteThanks +Yes I understand this +Is there really a memory leak issue here? I dont recall all the details of but I think the problem was not with +Committed revisions r and r +Could be considered as part of this? It seems like JVM command line parameter handling in general needs a bit of TLC +Here is the patch. +the integration test subproject and device access is okay will do so tomorrow +I would say that cloning adding is also a valid use case but the most important thing to me is that this is a public API so just looking at how is using it is not enough and you cant really ask all the users either +This bug has been fixed in CVS a few days agoAlberto +PR for is over to this applied in master branch too? coming this weekend could pick it up if its ready +New patch fixes the failure due to a null pointer exception +Messaging wise its clear only +I checked to see that getClauses was copying the array still and that this wasnt already added but I missed the clauses method has been added +Bulk closing stale resolved issues +Not sure why it is not working in your caseDo you use the standard sequence validator for the pos tagger +We have dozens of projects and it does not happen with every build but happens sporadically across most of our projects roughly once or twice a day +contributed as part of Felix pluginFor testing use case I think it should be filed as separate ticket. +Macro defenitions and +classfiles are added on metrics with slashes. +Patch applied in svn revision the port number causes a port clash when a shared fs masterslave pair are started on the same node +Please if you do not agree. +It will work fine for a week or so then we start noticing messages are missing +Kathey I have verified this is resolved with and +FROM +If it does not the cluster is sort of hosed anyways? All regionservers shouldnt go down thoughWhat you think of the rationale in that moves the PONR before the meta parent offline editYou have tried your patch in? It makes things better for you? Maybe we should apply for +Awesome! Thanks. +Make the Secondary info port settings roll from one number to another +Using of Thread local is related with current context of the current thread +There is a fair use case for allowing multiple completions +Ive actually used the same mechanism than for servicemix means you can deploy a bundle with a services containing the class name and it will used instead of the default one +I have made the according change +I have deleted it in this commit. +Shell log can you try estebans the suggested fixEsteban is that the only instance? Can you check for others that should be escaped with to and then create a patch file? +as long as it says checksum is determined by the content AND other parameters +I ran all of the unit tests and also checked that the example workflows for each of the affected action types still work +Anyway I willhave an attempt and let you know how it turns out +All the beans do implement serializable + +seems not a perfect choice to get the above job done +I know this isnt helpful but to this request! about pushing the progress back via a websocket? In the other case the uploaded client would need to poll for the progress +can we fix this for? if not mark it as FUTURE +The both configuration files are from server side not client side +Other configurations in and are also included +Yes but I was unable to make the work +Why not check against lid being null to prevent NPE +Web Fixed +for the trunk version of the patch as well +Thanks Thomas for taking this up +Is there any way to access the and have it dump its caches explicitly +no longer seems to +Nigel Im changing the fixfor to as the release lifecycle for the m plugin is different from the corer release lifecycle +Thanks Jarek I found that my case is caused by other reasons the codes here should be correct +Jiri you are correct +For example for DB Pools it should allow users to create jdbcDaytrader just like they can create from a plan file +For the other test failures it seems ulimit on the machine performing tests has tobe increased +defer to +Thanks. +FYI the proxy wont inspect the fields only getters can you add them and see what happens +I can work on making the change this week +added code close this as duplicated of +Fetching the error log is only for you to lookup if the exception would divulge you more detailsI changed the title of this issue to reflect the real problem you were to CR since this is not a vital +Robert As far as I know project entries are added to the maven classpath container the same way that the PDE does it +That will make this bug aI will change this bug to resolved once we get the patchprovided for into svn +No I dont have a user account and sure kstam is ok +This is a regression since M +I think that the label there doesnt affect anything user visible which is how it slipped by but here is a revised version +I think thats it +There is one now in a new module +That is the namenode will receive callbacks in the same order that have been submitted so if the namenode processes return values using the same order of callbacks then the execution order will be respected +In it wont target to am going to resolve this as Deferred since the changes and compatibility issues have been addressed in +Test case to use with previously attached pdf that causes jvm to crash +testAutomaticallyfield number fieldName area type class +Yes +Chris could you provide some more info +I dont see a connection to though +Would it be possible to add a default security domain on the as well? Since the ear usually represents an entire application with several modules this would be very useful if possible +New Document and Sentence wasnt aware of the fact that multiple entries can be placed in the properties file +Thats what Id like to do unless you have some compelling reason that it wont work +not been produced on other setups than mine so not blocking now assumed solved +added workaround suggested by MSsee commit log for more details +Created to provide a principle of least surprise solution for people restarting the server w periodic flushing configured. +The test fails on trunk as well +The current one works +This seems valid a test case and a supplied patch to resolved the issue would be a nice contribution +Also there is a redundant call to in that information is already available at we should just pass it down +Thanks thats good! The test will only fail in the first place if you exclude the DOM part of the Woden implementation from the class path +Added mapreduce related dirs to t this being fixed in? I think the idea is that YARN shouldnt be managing MR itself +bulk close of all resolved issues. +relates to improperly written readObjectwriteObject methods +I apologize this issue is a duplicate of and I didnt catch that before creating this issue. +There is not need to deploy this POM as its only use is aggregation of the Hadoop modulesDeployment should be skipped for this module by setting the skip flag in the deploy plugin +Add final keyword to method we fix one thing at a time? The test still fails for me with the latest patch +Please verify +Yes +Can you try it +Hi Arvindthank you for your comment from code perspective +Tested on version with jsf and the problem is still happening +Mine is derick +Too trunk removes the lazy load payload issue which has been resolved and committed in a separate issueStill have to consider what to do about the back compat tests +I guess this means I need to press the ruby people to add the fix +Good to be consistent even within your own patches rightOtherwise looks enough +I added an attachment that uses the directory explicitly if the system properties are null and if the LOCKDIR is relative the lockDir will be based on the PWD. +In the future well allow this to be disabled as well as possibly making the the default mode. +I have confirmed this on Windows enterprise +We had installations of websphere on different machines +Is the inclusion of an error? Its not being usedI wonder what changing the SPLITDIR does in the context of a rolling restart +Maxim looks good +Not sure though and the test coverage in this area isnt great so for the test case +Proposed patch +The fix complete unit tests for +It is supposed to go into the release +See Rameshs last comment +The assumption is correct the deployment scenario consists of about EAR files and each of them contains several JARWAR files which are CDI enabled +One more informationI have created my project using in STS after i removed it from original location and i created a new project by importing a existing maven project +I took the liberty now and let me just inspire by the patch and Ive added the functionality in a similar fashion as variation of the implicit collection handling +Fixed the directory creation issue. +Assume closed as resolved and released +I used the command line tool ldapmodify to import and ldapsearch to compare the results +The fix is actually up in it now ignores any filedirectory that begins with a +Does the Transaction Manager set an active mark? The two howl log files in vartxlog are of bytes size +New patch just about ready +I started work on that the plan is simple first step is to change to be Locale aware second step is to move all conversions of objects to dedicated classes and stick with Xwork conversion configuration support and left as converter for primitivesAny suggestionsRegardsLukasz +In this way it is like a war file which always gets servlet runtime from the servlet runnerA job jar is not a war file it doesnt specify servlets include jsp pages etc +Thanks KevinPatch applied to BEANS module at repo revision rPlease check it was applied as you expected +One is to just delete the bundle directory under datacache for the demo +The delete method in the should either increment the commitSize counter or a new counter should be created for deletingAnother thought +We have to have a better test story +I mean thisbq +Dynamic attributes are already support by all tags +The fixes remove the attribute readOnly +Yes I know that at least projects that would like to have a CLI for blur +Check the name of the repository the workspace name under which name repository was stored etc +Same errors same workaround +It does not address the TODO with parameterized types however +The scan included the both zip file with the problematic DLL files in them and directories of unzipped filesNote that recovering unzipped files from quarantine is really easy but theres no recovery from a threat removal process that deletes the files from zip archives +Fixed errors in sql for providing patch applied and build works fine +Brett I think it is not needed at the moment +Theres a larger cleanup necessary to handle all these really rough cases where we pull off synthetic methods to avoid saving stack state +Ill open a new JIRA for a permanent fix +rebased fixed code style and committed +This is fixed in trunk and now. +We only need to enable the checking in the when loading dataThis will reduce the impact of the problem a lot since its a big waste of the time if the query takes a long time but we are not able to get the query result +looks good +Are you the one who wrote the post on the forumsIn looking at your solution I noticed you created a new instance variable ve +Updated parent to and toIntegration tests were run on a local WLS instance +Fixed in r +Thanks John! +bz files attached to this issue should be put in testorgapachepigtestdata for this patch to pass unit tests +is not a problem anymore +Moved from to just after merge of old branch into TRUNK +Merged to the branch in revision . +Please raise a separate JIRA issue for changing the exception type +Second thoughts the want is outweighed by the resources required +My rationale behind this is that when the count is equal to zero the minmaxmean values reported are garbage and should not be relied on as valid statistics for the field +I did it based on some experience with the chaos monkey I did not consider this past issue though so Ill have to revisit this +Lists createdmoved +Thus in b above the nd is the rd in the source code +patches are welcome +Depends on patch that fixes code formatting and also compiles against java. +This looks good although it makes an incompatible changes to a public used by Nutch +I think that has a lot of merit and in fact it is related to the reasons that I created my vhttp protocol handler which downloads and caches all codebase jars as they are requested by class loading +The meta section is not a clean idea +Thanks for providing the patch + is not blocked by this issue quite opposite unit tests on Linuxx depend on +changed the methods to protected +Amareshwari its a good catch +I am interested in seeing more on practical use cases for these things +It was the class loader problem +The raster data is a bunch of PNG files so I guess not +thanks +Beta? The last beta was released like six months ago? Do you mean nightly build +Duplicate of +I got it +Attached is a change to the warning message to include the name of the class of the bean instance passed to the configureBean method +we will keep netty for upgrade to also requires to upgrade Netty to in the same PR than merged in master branch +Attaching management server to take at your network configuration make sure inside ssvm you can connect to mgt server +bq +It fails all over the place in and Im not sure Im putting it back together correctly +Verified in +Wasnt aware of those additional flags. +I set the jira status to critical as it should make it to the +Thanks for reporting the bug and uploading a testcase +contains patches for java code +This threw an which would have cleared the cache +The summary field can be reduced down to a think sliver but not completely hidden. +Considering it fixed +Not really sure though +VerifiedThanks Sian. +please check the approach. +The patch of the described changes +Error occurs with +Maybe this material belongs on the wiki? There is already some duplication of material there +testnativebasic fails occasionally on a Windows server multiprocessor with all test cases passed but with status code Sometimes a debug window is shown either with a segfault or attempt to free invalid pointerThe test also fails before the patch however it looks like this patch somehow increases the probability of the failure +FixedThe exit code has changed with version. +The IO is likely default internal and is emitting characters in default internal +Committed to trunk and. +agree that it needs refactoring and this looks nice adding extra real components might be cleaner but at times i hear people complain about the amount of files in the cassandra datadir with LCS +Ok +for another extension point for keeping it the way it is for renaming it and mention it in the release note with hyperlink to wiki page which describes the actual steps to migrate +I just committed this +Added more info about how to enable GDAL when running Geoserver as a Windows discard the and use the last onesince the creation of this JIRA some users still asked for help on during these monthsTherefore I hope nobody will disagree if I have personally modified that part of doc containing some additionals clarifications on that topicPlease feel free to improve the documentation if neededChanges are available asrevision on revision on trunk closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +removing would break backward compat +This patch only modifies the creation wizard and related classes +If you can verify for me that your patches were applied as you hoped then we can close out this issue +It seems that since java it is allowed to use a negative scale for and it seems that nobody told about that +That way if we need we can make it pluggable perhaps in the future +Annotated the dag submit API which accepts an as Private should only be used by Tez for MR Added javadoc to indicating that a single client is only meant for a single instance of YARN service This should fix as well +Applied to branch revision +When can we expect this issue to be resolved? This is really needed for legacy projects +all of which will botch up the index directory state by leaving a dead lock +I disabled all the tests in for now +modeshaperesourcesruntimedeploylib +I wiped Hibernate from my local repository and my project built without errors +have committed this +But it looks like some of them have been compiled manually from the source distribution with debugtrue and sanefalse +Did you say that Ted? It sounds awfully familiar +Cant you just merge the commitnevermind I see +Id suggest appending it to the IE jiraChange it to a feature request to add a note that Firefox is the only supported +requires randexp which requires which in turn requires rubyinline +So we let you set a pattern where if any file that matches it is changed we force the server to redeploy the app +The patch is attached along with test source javasdoimplsrctestjavaorgapachetuscanysdotestand test resource javasdoimplsrctestresourcesto optionally add into issue was originally created forSince we have a resolution already its optional to either commit change right hereor create a new issue to specifically track +Seems like a new feature without patch so moving out for now. +All we can get is the not yet run tasks which is not so important +fixed in rev +See also +thanks! +Committed at subversion revision +Missing is a story on +I see +groupId is not rigth theres no proof of controlling is open sourced by IBM +It seems like having a global setting for injection probability will mostly lead to test results that are hard to grok +I have patch the to update CQL tables in +Willem could you add documentation on the Jetty wiki page how to set this temp directory with Jetty +reports the remaining memory leak +While framework isnt an ideal environment in any sense it gives a certain level of automation which is still useful for most of our activities +committed in a fix in order to unregister the component factory if the required configuration is removed +Added special case to looking for UNKNOWN status +Applied thanks! +Inside the EAR I have being declared inHi do you mean you are trying to deploy the destinations in your EAR application +It works there +one about bugzilla migration and one about creating a new project +Thanks to Charles Aznavour who found this issue +I sorry I have some mistake +Committed to x as well though +Following is the suggest fix for the issue described above +Jason will revisit for to ensure it does the trick +Basically the same reason why we pushed so I definitely dont want to tie it together +Id be OK with wontfix +Ted thanks for your commentIf the metadata file is optional all fields will be treated as numerical when it is not provided +Committed revision +Please verify +Perhaps we could add a test to see if the underlying connection is closed in set but Im not sure the exception in that case is incorrectIm guessing that for some reason the underlying Connection is already closed when it is returned to the pool +looks good will commit if the tests +I can move a +I strongly recommend Friedls Mastering Regular Expressions book the most valuable computer book Ive ever read. +if that helps +Duplicate issue has been already verified +This is a duplicate of +Thanks a lot Jacques I will try to do work on it by this Sunday +Thanks JeremyYour patch is in trunk rev +Reopen to change to and branchesKathey +Updated the patch and Added FSYNCWAL to enum +webworkcommand name converter param success error input webwork name value webwork name value webwork name value webwork name value type +Verified in JBDS L +I believe the notifications from peers that are FOLLOWINGLEADING should be treated separately from the the ones of peers that are LOOKING +The type could not be computed because Vector is ambiguous +It can already do index size but it just needs a bit better management and automation +applied in revision Alexei please check that it was done as is fine thanks Mikhail +Not sure what I should be looking for +It wont always be as simple as a direct call to an entity finder and I wouldnt want to put all the logic into static methods on those +Issue fixed included in the iPOJO release. +suggested patch file intended for inclusion CLA on file of. +Does this mean when using hibernate the only way to delete objects that use an annotation must be deleted one by one +As discussed Ill attach an explain plan during this week +But how cool would that have been if it had worked! +It should not throw a exception when it cant decode a single row +This changes the semantics of what was going on +Thanks! +I dont like the heavy dependency on however but isnt something similar possible for Forge? For more advanced scaffolding etc +Cannot reproduce here neither on Linux nor on Windows +The problem is in how provides files to the parser +If you think the is a problem then drop it in your configuration for benchmarks but as I did not hear any actual user notice this it does not looks like a real world issue +how about a build all failed +the expression is not optimized and could be to enterpriseApp with page as root object. +Going to close the generic issue in favor of creating new issues specific to modulespages +If there is region disabled but still online we will wait for the region offline after that we can delete table and this case will never appear again +Should we limit the length of status updates? Twitter is but Facebook doesnt have a short limit like that +since you are inlining your css in the page images need to live in a fixed relative location next to the page but because you are using a coding strategy that generates viretual folders by encoding parameters into the url path your page can shift locations and images can get out of sync +The root cause of the warnings was that Collections methods were being called on Collections whose parameter types were compiled at level +We shouldnt need to add special directories or other features to the filesystem to support trash +This can be done in release +Take twoThe issue was that we actually depended on to mutate the block object +is great +Initial C Messaging API binding for PHP +Pushed a regression test as well +The Antlr grammar failing with the version proposed patch fixed the problem for me. +It just pulls out everything that has result or or ABORTED in it +Resolved in trunk with revision +This is more a part for the wiki best practice nor a patch +you Oleg +Better to have a single copy +Updated patch that applies cleanly to current SVN for the patch +Ive submitted a renamed patch +Thank you for reporting and providing the use case +This is another common misconfiguration with Acegi +Fixed in Revision +Committed revision . +this is actually and in relation to alias generation which apparently is not deterministicone issue i know is that Table has a static variable counter that affects the alias genreation +Second attachment mimics the issue with respect to my application usage +So maybe the first time was some weird stuff? My computer had been standing still for min during lunch. +Also added code to notify listeners of initialization failure via a callback +Thanks very much Knut! While I was in the middle of merging the patch to the branch I did a final check and realized that the documentation of the startSlave attribute has another reference to freezing the database so Ill need to do another patch after this one +arirang is the name of traditional korean song +I was not well communicated until now but it will start to be an issue with the coming C that will support named parameters +before we remove a jar we should at least check if there is a dependency on this jar and if so we can not remove it +look at the project and make sure its build status is up to date +Fixed on trunk +yes i am using marvin and test a new build and tested +Released in +I just committed this +Thanks for checking the release notes Knut +No problems with making the change although providing the patch was unnecessary +Hi Maxime youve to upgrade on Sonar or on a greater version +Yessir +done +resolved in rmade test case a little more general and added explicit shutdown rather than depending on text execution ordering +Reopening the same jira causes a lot of confusion +thanks for this nice feedback +Ill commit +Is there any update on this issue? Ive got the same behaviour on think this is an major issue and should be fixed until shouldnt this be reviewed +reopening for possible backport +Lars Hofhansl Whats url of this patch in review board +Bugfix! +Italiano EAN invalidoEAN invlido +defer all issues to +Thats interesting +Im still trying to get a meaningful backtrace from gdb but the device hasnt got enough memory to load all symbol files involved and Im having issues with remote GDB and library resolution +Performance hit in case? Portability iOS Android impactAlso the integration here doesnt build any code it just uses the headers Im guessing thats okayAnd this piles on top of some other stuff that uses C struct initializers +This JIRA was Closed Done +Given that the doesnt extend the snapshot version of the sandbox parent pom I think this is taken care of. +Ill add the dependency if annotations are present or webservices annotations are used together with transaction annotations. +Nice +does this apply to trunk as well? if so we need a patch for it first +Why are we not using object versioning? If the client just transmits the last version number of the object it had to the server the server need only compare that to the version number in the datastore in order to know whether or not to send the complete object down +If comma separated paths arent supported in this way as is cant be used with Oozie +Sorry it is an EclipseMaven project so you will need to install either the maven plugin for Eclipse or install maven fromyou would then run Maven goal package and it will generate the java filesI will go also attach another project that is based off ant +So it would be nice to do a checkAhad would you be interested in providing a patch or perhaps share with us a sample program with customized comparator that fails +The counters code is already set up to do it We just need to set up the links to be able to point to the correct thingI filed to handle the +Checked in to trunk and appreciated Jan +nice to have but not critical for +Thereafter we will see how to deal with Adrian suggestion +Hopefully fixes other fails +I close this as we do not consider Derby as a production DB unlike Mysql Oracle or Postgresql +Some small comments followBehavioro Some typos to Derby will retain so Derby will retain notwithstanding and lack notwithstanding a lacko I did not understand the red paragraph +Temporarily commented out target till the root cause is identifed +moving it this patch guarantee that the only rows passed to the mapper are the ones that fall between the start and stop rows specified in the Scan? Or will I get rows that fall before the start row within the first split and some rows that fall after the end row within the last split +Please close this issue if this suits your needs +Observe there are now products in categorynd graphic screen shot of the Promotions category +Until will be done the only way to deactivate Roo Standart Menu Service is deactivate Menu this modification we miss all MVC functionality +Then its consistent instead of two different formats +methods can either be public or package scoped +I suggest the following improvements Change Alert events on my favourite projects into New alerts +Closing for Beta +review attached +KoushikThe steps mentioned in the comments is what I have done +Jim can this be marked as solved +Also Patricks Able built onto a similar capability but without the classloader scanning +bq +Reviewed all these issues and marking them as closed. +I will try to integrate via the maven build processit sounds reasonable +The context menu of the VPE will be fixed automatically when we well use the source editor context menu implementation instead of the current implementation +New patch which incorporates the creation of the sandbox as well as cleaning up more references to the XML queryparser +Then tell what kind of error you have for me the build is okay +Currently our users have to click twice on the same columns to get the desired sort order +I think this is a one which should be taken care of in jBPM +A patch to fix the tomcat issue +When HDFS is started up it should be possible to select the mode that HDFS comes up in +Theres a field for this +If theres anything else I can do let me know +The patch has been appliedPlease verify. +Patch ported to +Wont that confuse folksIt would indeed be nice to have the code in a single place that can easily be deprecated but we could instead add a consistent comment to each of these places and file an issue now to remove all places that are +Yes I think Jira took the spaces indentation in the example +not a regression pushing to remember this being bounced around on list we need to define the canonical way to get the maven version and make sure it works in and dont use a system property +WFS wise not sure +Several other attempts to force the view to update itself did not work sadly +Pushing this to M as we reconsider this approach vs +What were the file names of the you set aside +Previous commit was broken revision works +Your new tests failed as expected without the code changesand passed as expected with the code changes +Im sure Im the one missing something not you +Whats the plan of attack here port the code and remove dependence on Gora +There is an outstanding bug on navigational state not being replicated in cluster setup +the native code does not ensure the JVM threads terminate before X main return or +Im exploring all the possibilities of securing a method +Brian did it get into +Or should as in What a good idea we should implement that +applied the filename mapping part already +bq +Has anyone used this patch with success? Is there an expected release that this patch will be included in +patch to resolve check that omit the addMixin if the given node already is of that nodetyperegards with rev +I was told I was authenticated through LDAP but now after chatting with someone with more information on the topic it turns out I gotta go through SSO +Im sure on modern hardware round trips for smaller messages is well under ms +on Raghus proposal +Changing it to timed weak references I was able to speed up the loading of multiple projects significantly while cleaning up in the long run +Moved and to Changed all methods in to be abstract methods +If the feed is updated you get all entries not just the deltas +So no chance to get this done for +Hi GeoffMany thanks for your contribution +And didnt we say we wanted to use only something which is clearly definedGeographicHave you found any files or file types which are actually using the WC approach to store geolocation data? All I have seen until today are using Exif to store it +lgtm +The reason seems to be in patch for ve committed the fix at r +Hi AndersFor now codehaus nexus users dont have delete privileges on the Codehaus Snapshots proxy repo there is a scheduled task running daily for cleaning up snapshots older than days and keep a minimum of I can help you delete some proxied snapshot if you like +is this fixedout of date or +Hadoop version bump +I think its clear no major ui transition will occur now +I will upload a new patch with those changes in the morning +eclipseeclipse is working locally +offset? rows fetch first? rows only +After multiple attemps and review some stuff in myfaces core I added a new component called tauto +Icon changes in browsers but VPE doesnt support this of attribute first doesnt take any affect in horizontalOffset doesnt work in more list of down attributesinputClass inputSize inputStyle zindex jointPoint show seems monthLabels week dont react too +I just pushed this to trunk and branches +I just upgraded to TC +I try it asap +I waited for the git sync and refreshed by git repos +a reproducible sample Graeme was I successful with providing you what you need to troubleshoot this issue +you mean right? +I just successfully put solr in a C app without any webserver running using JNIAfter I clean up my JNI calling code I can post an example app here to show how its done on the client side if anyone is interested +Ill look at it tomorrow when Im up in portland. +Here the patch +I feel this should also be done in because if we want to do it right we need to be thorough. +and for a next rev we should maybe have a warning on the console config if that classpath is not in there +tables were added for the following components datascroller dataTable pickList +This is an error in the suite and in particular in an extension that is not part of the standard the community does not support it please ask all issues that have been in resolved state for more than one month without further comments to closed status +It happens on my side with clustered and configuration +See attached +It is not another kind of HTTP authentication +zookeeper jar is needed as well +Added test case to help reproduce the problem +Enrique that issue is closed as I missing some existing feature I dont know about +As agreed with Mircea Ill take a look to understand whether this is test thing or is related to infinispan +Moved to because for a complete fix it is necessary to modify the Partition think we fixed this overhead. +This is how browsers work +MichaelThanks for the comments +Revised description Use noformat +Brians comment was made on the nd of September +Getting booleans for booleans the getters are usually isVariable rather than getVariable +So I was wondering if the degradation seen with is noise or something that needs further investigation +need more time for benchmarking. +Remove I from the front of interface names +Thanks Amareshwari! +This seems to be a duplicate of when the patch in is applied. +This is an issue for the branch as well +Josh its an old patch so naturally it didnt apply +Robert thank you for this grand effort +Google also doesnt help finding the WikiWhy not strip the main site and link to the Wiki +Ive commented out since it relies on logic that will need to change +verified in JBDS +I will raise a JIRA issue and fix it after this patch has been mergedThis patch also includes changes to allow the to visit more than one idl operation in an idl interface +This is due to my carelessThis was used for load wsdl of proxy service wsdl of wsdl endpoint etc from Https URI Https URL connection +I am not sure if it is accurate but close enough for now. +Nice workThank you very much +Reopening to correctly close it as a duplicate. +Seems like people like this +Hi I am a graduate student majored in data mining I am very interested in this projectI have used some experiences on distributed data mining using hadoop so I believe I can handle this projectIn order to work on this project is it necessary for me to join the GSOC programAs the GSOC requires the international student who studies in the US to apply for the CPT and I almost used up the CPT due to previous internships so I am not be able to apply CPT for the GSOC +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the codebase +Im seeing a similar fd leak issues especially with huge amount of connections +Find a way to represent +I thought I saw this one go in so marking it resolved +The timeout for waiting on a stream to complete An arbitrary timeout here is not the right way to do this +Ive received some good feedback from selected companies I had contacted who were using arithmetics heavily and they reported no particular issue with the improvements. +It is unfortunate that Java does not provide a way to send signals to processes or to get pids +They can be in a separate section say scheduling related configuration +We will take care of this issue out of band +Uninstalling Indexes depends on the implementation uninstalling Resources installed from a Bundle +Fixed in Metro integration resolved issues for already released versions +Copy generated files to appropriate folder in current niochar module +Im seeing fail quite often nowI think I have taken care of this we now wait longer for a leader to show up when registering compared to the short wait we do for updates +Someone just needs to deploy the site for +The standard practice of rolling encryption keys might be difficult as wellIf the machine is trusted and compromised and wants to be malicious its game over rolling keys or not +I still consider modules first classId also have argued that in we should rename the function class to something else +Sorry againYou take of the modification or do I have to submit the patch somewhere else +I just committed this. +The attached ZIP archive contains the latest code which according to me should not have major problems +enabling message delivery for subscription from rA publishing a move message to T in rA +Whats the status of that? It seems to be fixed +updated patch to match changes in a typo in for the update +Fixed with SVN +Ilya why did you add px style for toolbar images? It moves icons down under IE +Make sure this check exists at a high enough level so that we dont prevent it from working for app client +Trc depends on and wsdlj +Ltd +I dont see any issues while building +I am not sure if I can provide a very educated code review for the patch but the changes look fine to me +Followup logged as. +Would need some work to serve as a sample +I just committed this +Could we +I am just trying to reapply the patch as I might have hit some problems +Think this one is already fixed by latest latest build contains README file that contains +Can confirm after using for a long time in production seems the problem went away +pushed in Developer StudioVersion Build id Build date to add screeshot of current s how the UIs looking currently +I am looking at the patchIn the unit test the way it sets the transformation the scan should do is a bit odd +I can confirm that using a prepared statement rather than Active Record works around the inserted the problematic record by creating a record via AR and then settingupdating its CLOB column value via a prepared statement + of Scenario and sysoutputinput from screenIt is not a bugwhen the file path contains spaces you must add the quotation marks around itthanksIn this case I guess a liitle bit more clear description would clear up the confusion upon command syntax of deploy in users guideAnyway thanks for the info +But Im still fuzzy as to how that would work +Got it +Is there away to not generating the top level jar at all? If not I guess this patch is also ok and we can tak it in +Thanks a lot Lars the changes are fine to me +I see +Thanks for the suggestion +Is there any suggestion or comment on my proposalThere is little to say +I did some tests and this seems to be fixed +Evgunei is this issue still valid +please create a new issue +its really great stuff RyanRandom questions Whats the funny chars at the start of? Why is there a MB upload limit and why does it only apply to uploads? I think Im missing some background in this areaIm not going to try and pick out little bugs or suggest little changes to a patch of this scope and sizeI think we should have more frequent Solr releases and that the Solr trunk can have change from one day to the next wo having to maintainingSo since we agree on the direction I think this patch should be committed and we should work from thereId like to hear from hoss though since he was following along more than I was esp at the start of that marathon thread +What do you mean by a deadlock? Do you mean a database deadlock? If so I wouldnt expect it to behave like this you would get an exception from the database telling you there was a deadlock +This is probably why they didnt do it in the ebaapplication tests either +Which one +Id go one step further than Deepal If the requirement is to support then its simply not a requirement for us because were not implementing +Ive changed the new file headerPlease check that the patch was applied as you expected +Not sure if there is another way to do this +Sorry for the false alarm so I should raise this as an issue in Tomcat then. +These issues are now addressed changes are checked into svn +As a workaround developerConnection in the generated should show where the branch will be created +Sorry +I also checked the following files and none them referenced the tlds eithersamplessamplessamplespetstoreWebwebsamplessamplessamplessamples +I have committed a fix and added a stress test to validate this didnt want to change any ordering on the journal so I added a cleanup on every X records as discussed on the forums +Quickstart showing there is no problem +Looks like a sweet piece of code but I would need itto work on other types of phones tooRob +Moved from unknown component to has been retired +Alan agrees that this is a reasonable direction to go in to fix this problemG trunk rev openejb trunk rev Merging with required some manual work much of it from the TCM removal and openejb configuration work dain did +Nick in general I think that my patch from yesterday should resolve all your issues +Yin can you take a quick look if its easy to fix that away +Things still work if internal users email tao me know where we are heading regarding this try to explain the bug and why we cant fix it +Thanks Arpit for review! +Feature list for jBPM guide is added +Should we add dsl parameters +Committed revision +CF or CFSC +provide both trunk patches +Logically a source procedure is executing in its own block and thus its somewhat inconsistent to allow temporary tables created by a source procedure to createreference session scoped table. +Hmm that makes it look like the current tests for this must be pretty weak then +I looked into this again and the signing of the JAR file seems to be fine +does that when using clientserver compression for exampleIll assign this to me and start poking around a the version which always sends the smallest message saw great results in compression ratios a standard stress test gave a compression ratio basically all messages where compressed +I have posted a question on the list to see what they say +Can you test with the workaround? is not the same issue however although it is related to what needs to be done to v in order for this to pass +Use an access library based on the prefix to extract the file at the path from within the fetched archive +Not sure this fix is really the way to go on a busy system we could be canceling and rescheduling the timer quite a bit which introduces more overhead than I think is good here +should be master node server or master heartbeat should be or +No need to bother. +Please help us in Building requirements to make this tool effectiveRegardsAnil IdeaIf necessary the Business setup guide into hierarchical content and then build a Ofbiz content data file +Could you add mauro to group? I still cant seem to be able to edit page ie I see no Edit tab in display +java +The profile in now activates? Also it looks like we also need to update the default hadoop dependency from toIt was my understanding that was going to keep the default +Updated patch for latest trunk cleanups +Before starting on this we are blocked on and +As the util is not specific to HA the patch moves it to +The ability to float the scalebar to the top and quickly turn it off and on is required for the uDig release. +Closing all issues created in and before which are marked resolved +We cannot fix it there easily as it uses the old style logic fromWoops right Ill fix to only +Thanks for the contributionPossibly we can choose a queue name that is auto excluded from networks possibly even a temp queueThe issue with the pong working when the queue is removed using the admin console is simply the default activemq on demand destination creation policy it is auto created on the next send attempt +Am going to commit on revision +With missing dependency before trigger action was not getting regenerated and hence trigger didnt detect that it is not valid anymore +This is similar to carrying deletes over at the segment reader level based on acomparison of the current reader and the cloned readerOtherwise theres redundant deletions to the disk index using which can be unnecessarily expensive +Lars yes we can add all useful methods to the interface +Otherwise sphinx files will just replace him +When credit card is selected as payment mode clicking on submit button does not do anything +to Hudson +Please review +It was excessive now that I look back on it +We still have this problem on vmbuild trying to delete the jUDDI projects +pls commit asap as this is a blocker for other components of Hadoop stack +birt ws and portlet components are missing in update site build Mozilla Libraries Mozilla Libraries Closed. +this patch fixes the javac warning +resolving were in a better shape now it seems. +In M you should only experience this if you are very quick finishing the wizard that shall embed the cartridges +I agree that that dependency is not acceptable especially for such an invasive change to +New feature introduced the problem. +How are you setting up and using? If you have any example Id be interested in trying it out. +If so it would hand off to it if not it would log an error +Sure in this code base there is code whose design intent isnt clear initMetadata being an example Oh I make no claims about how well the existing code base adheres to our prefered style or readability in general only the direction wed like to move it in +Attaching a patch +Changed the way isText works to be compatible with CJK characters +The patch for bug fixed this problem. +bq +It work correctly with ChromeThanks for the fixColm +Thanks a lot for saving me some effort +Thanks Kim the patch looks good +Please check if you have a patch for this. +Simple change +During a branch or tag process we can use the to change all the version numbers at the same time that we change +bulk defer of open issues to +Ok I am the one going blind andor stupid +Sorry +it depends on the additional libs andor your pagesleoyes thats one of the problemshowever in any case the debug shouldnt change the behaviour +the patch looks good to me +cassandra +Once the phase ends we process the stack and add the transforms to the correct pahse +Ted As for readpoint +The scripting is updated and verified +Just for curiosity what is the name of this broken server implementation +I believe we need to drop this from fix close it as its done in will open a separate issue concerning site plugin with current trunk. +Worked around this bug for the release by making sure I had installedHowever the build system should be fixed to handle this there a workaround for this issue? I HAVE swig and installed +Import completed +in sync w trunk all downstream tests in hdfs are passing +I just committed thisThanks for the quick reivew Eric. +Compile and run against Xerces +Please using the mailing list first before you fill a JIRAIt could be more effective for us to track the issue if we start a discuss in the mailing list first. +Applied patch into trunk +supersedes this issue making it obsolete. +Forgive my ignorance but this patch looks like its unwrapping and exceptions an extra time +If you look at the Phoenix code youll see where the rest come from +Thanks Prasad +To make matters worse this doesnt always fix the issue +I still get for This needs to work in order to run applications in that would work in +It is not clear to me why client cannot figure out whether it has the token for the service +EAP for buildRam are you ok with proposed version string +Assigning to Andrew as he has done recent this method to the subclasses will ensure it returns the correct type. +Not exactly +Fixed in rev +Now has a public static method which accepts Please review and exact same patch also applies to trunk +Applied in revision revision Thanks Fix version to +Curent problem or challenge is the embedded logger editor only allows for the creationmanaging of one logging results file +Its a cleaner casebq +This and which it is now almost a dup of will require API change and should be handled in am inclined to close this as WONTFIX given that has been addressed +Seems to be fixed by some refactoring but cannot trace the commit or the version. +On looking at the code in the caller I found that the fix I had will not workr is a much better fix +For I fixed the NPE in +Fixed in version +Committed this to trunk and +I tried to reproduce the same in the RC build of +It is not yet showing up on the site. +If the charset is invalid you will get an exception when you add the route into the camel context +Implemented first step in Rev +All is working now except that the login cookie doesnt have the SSL bit set so it isnt secure +Initial Session Conversation and Request scope committed for Container +Icons in lists demo are revision +right but I guess that means to keep too much history behind an clutter the hudson +I have linked a duplicate issue which was raised some time ago against a previous version of hibernate +Bob is currently in the throes of dealing with our SSL provider +Although the specification says the sca is to be passed on service requests and callsbacks it doesnt explicitly state that the sca is not to be passed on the responses +The code I dont believe has anything to do with DRDA but is just the code used by client for disconnect exceptions +The property table is now automatically populated with default values and some basic input validation is performed. +Ive met the Yeti +It fails only if you explicitly select create new project from existing source +all these two proposals handle it +In the? If so need to think about how we will load the and check for the property before weve actually run the test +Also while the transport is performing a blocking opperation it is not considered idle +All unit tests were passed except for which seems not related to this patch +If not that then something else is different by the fact that it works when the page is not a url relative to the rest of your site +YesSimple file to inJust did check against the newst Struts version the problem persistsMarkus +I believe I partially replied to your question while you were posting them +read all properties from custom and reset them for the log Configuration +No activity + +the new tacos notation sounds like it is what is neededthanks +Patches are welcome +for the to trunk. +Starting a started deployment should be a noop not a restart +However if the RS stays in closing the region due to some bug not sure if reissuing closeRegion will help much +Sorry this was already committed in. +Fixes for resolution settings are perfect now +Should also mention that it can significantly impact load balancing since the first request made by a proxy will load balance using an outdated set of targets. +My only suggestion is that it include a reference or link to the main Cache documentation which of course is far more complete +Attached updated patch +Mass close all issues in resolved state that have not received further comments in the last month +Let me run again to make sure +Duplicate of +Sorry for the worries Eberhard +Bug has been marked as a duplicate of this bug +Patch that fixes the issue +This is a real problem which has been repeatedly mentioned in the context of the paginated queries and in fact solved in +Agreed Probably does not need to be fixed in +Rather it provides which do a migration of payload data toThe DOCSONLY case doesnt require migration per se +Thats great thank you +WIP patch uploaded for safe applied +This can overwrite data with the same row key +So if undeployment of EAR is in process and datasource is unbounded from AS undeployment of EAR ended with this exception +Ill reopen and take a look at constructors. +Same setup as the poster Jboss RC JSF reproduce it simply create a new project with +Going to fix bugs and the Normal distribution case has been fixed +bulk close of resolved issues in. +Could you provide your full Maven log file when you launch mvn install +Therefore I believe the answer to this issue should now be yes since the EPL is already documented as being allowedRalph No irony +This works fine for me on Windows +Project created notifications set to commits +IT added +all the issues mentioned seems to be related to that +done +We do not lookup of urls we fetch at this time since these urls already have a ip that was resoved when these urls was first time discovered in a parse process +refactoring is on the trunk main logic on the branch. +Obviously the tokens will need to be rationalized with what is being done for +Putting my object util in the Context worked the fist time +Am changing this to an Enhancement as the current documentation does not claim that the will be SerializableWill recommend adding a default constructor to the class will allow the subclasses to implement Serializable as required +lang and lang failed with trivial diffs on JDK and JDK +Please give us a few more details on your setup +It looks like this patch causes a regression +and list a few scenarios that could avoid calling this method +thanks +Is it possible for you to try that and let us know if it worksWe do intend to expand the use of the Java to other subprocess calls +a and using the identifier and entity arguements to put store it under whatever key you s not up to sselectItems to determine the string written out to the page but up to a converter +Note that the patch included code for a that was not used elsewhere in the patch perhaps this was inadvertantly included in your changes? For the time being I have not committed the code +corrected this typo in svn +Cursor is a high level one +Not yet +Would anyone want to try this +If this patch author needs to read the EUT source code to understand own patch then I can provide these test names manifests were fixed in Eclipses CVS repo +Its tough to review at this point and I still think the stuff is a little incomplete with this patch +For example if the time format is in hours and the interval is then the file should roll at and midnight +Ok + to trunk as commit cafcecdafaacbacfee +This also happens withCould someone try running the client with a file that sets the logging level to FINEST or similarYou can send me the log file directly at dkulp at as it will probably be fairly large +the test unexcluded in revision Denis please confirm that the right test was thanks +So all issues not directly assigned to an AS release are being closed +But if you can not implement it right now as it takes some additional time it is ok to move this to separate issue to be implemented later +Earlier Chris suggested such a thing about Gora +Both using the native library +I committed the patch to +Since it is packaged as optional it will not get pulled into the tarball etc but yes it gets pulled in during the build is that an issue +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +NET while using the library from an older machineIt would be nice if the generated code could compile on both +Thanks for the patch Greg +So all issues not directly assigned to an AS release are being closed +Problem was solved by copying the Weblog file from +If it is not too development overhead then we would like the patch to be made available for +If it can be demonstrated that a previous version of Hadoop was compatible with AIX then the patch can be pushed into previous releases but lacking that it requires a special dispensation +candidate for like fixed in +Ive attached patch that help to advoid NPE when there are stale data on notification message will be show after be verified in gatein master and reopened if necessary +Probably that file can be used as a reference to copy text for the LICENSE and NOTICE filesIt would be good to have a quick reference to know what licenses requires attribution text in NOTICE and what instead simply requires a reference to the license in the LICENSE fileIts weird no one attempted to write a maven plugin to take care of this for a skilled maven plugin developer it should be really easy to do +The review board is use primarily to help reviewing the patch +Closing. +All tests succeeded and Jackrabbit passed the TCK +Two things You need to inline the schema so we can see what that types are +The reason I prefer to ivy when interacting with local maven repos is because you dont need to thingsBefore redoing the patch Id like to reach consensus on what approach to take +unicode block calculations in the CJK code greek diacriticlowercase folding in the greek codeI think we should go this route +Different schemes might require multiple operations against the data store which may lead to higher latency +New patch that DOES NOT include the code for +After the library is updated we can run a service that queries the library and updates the database +Patch looks fine i just compared yours and mine +So setting the delimiter parsing will immediately have effect +If no one responds soon I will commit it +Can we reopen this issue? Its not looking fixed and as it is classified as such I fear it may be moving off the I will try to have a look during the weekend +Just committed this to trunk and. +I think this issue is similar to the issue we discussed with you in other JIRA +Patch When migrator command fails and force argument is false the program exits with an error code greater then +That process will continue tomorrow +Maybe is should be unhandled exception or sth +BTW happy birthday to this issue +Joram did some checkins for forkjoin behaviour in so Id resolve this in one way or another as fixed or whatever +If you find further issues please create a new jira and link it to this one. +Fixed at revision close for the release +Please dont commit this it should be discussed moreI know you hate java etc +See how there is a space between the left and right halves of the progress bar +Retested in +This is one way to set the JAXWS provider as a system properties +Resolved on on +Thanks for the contribution +Yeah MarkSome examples from the unit test is a good idea +an empty one in the case of no results returned +Thanks MattIt would be great if you could update the docs on the wiki too +We dont use annotations for the moment in continuum its one of our tasks but you can add annotations in your classes and they will be used when well active the descriptor generator +It seems that the problem raises when axis client try to parse the response received from the websphere server +This is a patch against the trunk +Bulk cleanse of EMB issues moved to legacy version +can we resolve this before EAP early access is out +The others are now yarn configs and have different names +Please ignore previous s and not +Thanks Jian +Assuming container runtime as a proxy for container work done this data can be used by the RM to preempt in a work preserving manner without having to talk to the AMNotifying the AM has the usefulness of allowing the AM to take actions that preserve work such as checkpointing +I could make that a default option in restore commandIm more concerned if the operator someone messes up and starts the cluster with like half the znodesbq +revised patch should apply properly in HDFS against When playing around with different patched versions version mismatch with the numbers help in making sure the installations are consistent with each other +war to a remote box using this method +have a look at this linked issue make changes in component description if inplaceInput and inplaceSelect were added to the guide +In addition it is only intended for comilation and not for other tasks like SUID generation that we will also need +Main thread must ensure both loading threads are on wait before calling to notifyAll +Looks like this approach probably wont be enough because the clients seem to periodically call SELECT on all known folders to check for messageschanges +Closing per request. +Thanks but I meant if you it explicitly for instance via clicking some other button +We still fallback to calling by reflection if NIO isnt available so this continues to work on environments which have the +For example +This is a huge behaviour change to include at the last minute in the transition from a beta to release +Sorry Jody I though I made it clear in the bug report I was already using the proper JRE and if you look at the error you will see that GDAL itself is there but fails to complete loading because one of its dependencies is the environment statement says with imageio and gdalAlso the error log in the description shows the location of the gdal library in the JRESo it is not a uDIG documentation error it is a missing dependency +Lars +Committed all of what we have so far to trunk r r +This happens every once in a while for me too but its always a consequence of another problem in your case you should see what happened with mapper +This seems like a reasonable thing to do but please review if this logic is valid and wont cause other issues +OK the main example was a bad one but as I said I tried with other patterns including mere I just tried again just in case and it still doesnt me know if you need more info +And from Bikas explanation thatll require rack resolution in any case +I think my design where is the leader is required to have a older history than the learner might be too strict in some cases given that Zab guaranteed continuous history +What do you think +Committed patch a to trunk +However ignored portions as that file exists already in toolsbin + +Thanks Alexei The fix looks fine +Bulk closing for +Dont hardcode a port number it prevents concurrent test execution on the same host Negative cases are only checking for a +Make sure you dont have a nocache entry in your +Third Write a method which checks if the file exists +Please file? the latest patch looks good to me +libsecurity +I think its pretty good now thoughYes youll definitely want to turn off autoflush +Please keep them coming +Mathieu I wasnt referring to your proving the problem I meant how was the module tested if it doesnt work with either CDI or! +I am afraid what you are trying to accomplish is a Legend view something that we should make for a future release +v still not done +Save on calls +I am talking about everything that Gossip does +Sorry dude I missed your proposed solution +uses maxdepth to avoid recursionBill does this look OK +output in but I did put spaces between parts of the example for legibilityIll commit the fix with these changes +Ok Ill make sure its compatible then + +Confirmed The infinite recursion in is still not solved +The locationmap is still under development the error message access is denied occurs when there is no match for a request via the locationmap +I suspect its due to State +We may as well carry on our discussion about the consolidation of these plugins on the Im not wedded to the notion of having a single I wonder what the cons would be to having the deployment and packaging goals within a single ask because there are currently three plugins for axis goals and multiple for hibernate +and obviously if it maps to then there is no need to explicitly reference it as I believe that will be the default +closing resolved issue +Unable to reproduce on F i seems that the problem occurs in environment such as LANG +Patch for x branch with tests +Reviewed and pushedThank you orfjackal and Devin. +Benjamin reported a failure in some Hammer test on his local machine +patch applied tested and committed +It seems to be based on a quite old revision though and it doesnt cleanly apply for me +JandalfOverall the patch looks fine to me +Verified in JBT Version Linux JDK +This is a language design issue +Joe thanks for the patch +This may create wirings that make it impossible for additional deployments to resolve +Could be used for Graph Runtime partitioning could be done in Vertex was made Writable and also given the responsibility to create the V E and M objects that the framework could use +Basic patch +I just committed this. +fixed in revision Vladimir please verify that it fully resolves your Mikhail +Starts up conversation components when a long running conversation is begun +Thanks James +hiforgotin fedora jar isnt availablebecouse there is a very strictpolicy to not duplicate or rebundle classes +Thanks for patch N. +Separate setters might have their own use? I believe I had a pair of filters that set begin and end offset in different parts of the code +Since dbcp doesnt have issues in its code it didnt for updating pool +bp files and autogenerated +is implementing a but s my new patch +Fixed by +We now strip correctly. +I do however wonder about the getLocal methods because we dont seem to be setting local files anywhere +I has started to merge WSRP integration branch into trunk and having functionality tests on that +aha Im mentioning this simply because we can probably use the same or very similar mechanism for keeping meta data about hosts and individualBut it looks like may be the way of the future +So can we remove it then +All other places just invoked the init API from classThanks as Invalid since these static initializers are no long present in the codebase +Keit The current workaround is to use the twiddle command as described in the fault description to manually unregister the of the EAR +I do think the current collection has many obvious gaps and were still a ways from supplying what people would need +However Would it be ok to not update this JIRA until after EAP and EWS releases as its causing some confusion +I believe the topic was discussed on the mailing list and maybe it has been fixed already +I should be able to post comments Monday +Regressed with latest RPMs +Will revisit this concept as part of the new builder for +No more no lessThe advantage of this contribution is that it is very supportable because it leverages the C client and adds very little logic +Fix Version is not correct +but support wtp projects but a cleaner implementation may be needed for vfs for archives +Sorry this was XJava +This will not work for people who are just demoing ofbiz +Thanks for the patch Dmitriy. +Not sure about +The patch has been tested and verified. +and I moved the issue to component mirrorsThanks for your support your mirror will be addedto the list soonGroetenHenkSystem is fixed and stable +Depends on the fix for +Wait until tomorrow and you can download the binary distribution +What I meant is that issuing an insert while compressing shouldnt result in either of the statement failing +Was this completed +Hi Rick I tried applying the patch in Eclipse IDE which says The patch has segments that did not match +thanks James +Added the sync that serialises transaction updated to the store and cursor such that the cursor is always in order to the store and setBatch can revert to just where it needs to so that there are no duplicates replayed to the cursorrThis is the bullet proof approachA variant could acquire per destination locks rather than the transaction store lock that is currently used +The hive metastore code talks SQL to store the data in the db +Support was already added in Activiti Designer version +Forgot to flag the ASF license would seem that agreements are between partiesAgreements are attached to Orders and Purchase ordersAgreements are attached to workflowNot all agreements have financial information +The new check that is being added is for version compatibility +But how does that explain that with refs the array is +Just save it and the edit button will reappear +Is that so +Now has the Data class as a catalog but it has to be generalized with an interface so thatall the mechanism will work inside or outside using whatever catalog would be the Data class seems to do too many different things +Thanks Stack and Prakash +Yep confirmed +It now considers direct and indirect package uses dependencies to analyze package uses conflicts problems. +This was a problem with not updating known entities correctly +Sitemesh filter by the time is done setting the attribute wouldnt sitemesh filter will have been called? Theres a request wrapping to be done as well besides encoding such that tags will have access to value stackI think its easier if we stick with for the time being until we could come up with a better sollution +Thanks +Thanks Kihwal for verification! I have committed it to both and +Change the component to was anyone still care about this issue or can we close it as wont fix? We should look into how many people are still using x with Cargo +Default behavior of the Java sequencer is unchanged from before I waded in there. +See. +yeah im working on it +I already modified the to verify that and Ill check if any other tests are neededIll upload an updated version of the patch +Still appears though it takes a bit more time to show up than before +patch looks good +I dont know where this is coming from +hit test timeouts so I bumped them up to seconds +It appears that this error was some sort of artifact from mine having updated from sts +I think it does exactly what it has to do +We should also add a note about round robin load balancer behavior is changed in will update the FUSE migration tips with a little detail about this +Low level component of a generic transport library ought not make any assumptions about thread management + +We just came to realize that not every committer is a despot and theres a couple of issues we would not want to discuss in public +Committed in rev +The latter never been supported and I think there is already JIRA issue for that +Either way well get this fixed before we push out Alpha +bq +The jira is resolved with a fix version of EAP is this critical to include in EAP +build you can get trace out too if you add a traceFile line to the axiscpp conf file +Here is a junit test case that reproduces this issue + +attached patch for init priority +The build should work now +Of course there is no reason for the two phase approach to not work but there is part of me that dont like that a simple mess up in the cache loading can make some keys unaccessible +The workaround is great but the default value to opencursors is I used and it was not better +We will work on this after core planned tests have been developed +Ill look into it +A few of them use RPC +this is a bug in SVN not a released version +Maybe before you fixed this it worked correctly using the browse button and now there is a regression +Completion proposal computers should check if file can be foundby the available document +Thanks for filing the issue +If using the MVC namespace that should be the case +Committed to trunk and +There was done big refactoring on JBT JST component I was not aware of +contains a zip file with sources for there a chance that this issue gets fixed in the next release? Please see the testcase that is provided +fixed in rev Thanks! +shouldnt this be closed as per the earlier comment +Even I did not run tests in that profile +Users at many different sites have experienced this issue and HDFS administrators are taking steps like Manually rebalancing blocks in storage directories Decomissioning nodes later readding themTheres a tradeoff between making use of all available spindles and filling disks at the sameish rate +Resolving this issue as it has been completed +Rejected after verifying the service works with the correct local connection URL specified +Thanks ErikYour patch is in trunk revision +It would be great if you can improve it +I did some significant rewriting of the JS code to make it work better within Tapestry and to reduce its size by making use of Prototype. +Possible implementation of allowing a component to update its own service checking that its OK I added attribution to the osgi bug that inspired me +Patch committed at r. +I believe the Enterprise spec says should be versioned at +Applied at r thanks! +A few comments at first look new files in derby need to have a license header the file added seems to be missing a license header Can you please assign this issue to yourself Can you please confirm that your icla is it is not can the community advise on whether this patch is small enough to be committed without one +Committed to trunk +sorry I didnt mark this as verified earlier +cool +bq +During token conflict resolution we were protecting the relocating host by making it win however this only occurred when received a normal state from it +I just added some modularity for multiple systems +broke this +Closing this it sounds like the answers are Use under CDDL +Bulk close for Solr +Didnt make so moving to cannot reproduce this error +The docs were misleading as to the formatting of the hint +Thanks for mentioning. +I guess that problem is solved in version +I think this is a brittle user api and a serious issue with the current functional spec +Anyway Id be fine with the bugfix and tests in your branchPlease tell me when youve prepared a patch +Proposal is to create a new type say and have the several list operations return that type instead ofThis will provide compile time error and allow folks porting over to change their code +It occurred to me that I could set my new parameter to deprecated then when someone figures out a proper fix it can replace this hack +I can publish it on github if you want early access +added an initial implementation and test all issues +The patch will probably work for either convention but it makes for slightly messy code and Im not sure its a great model for the future +INSERT INTO is not supported it will get a parse errorLOAD INTO is the only scenario we have to worry about +srctestorgapachehadoophbaseregionserverIntegration tests for and +If anyone knows how to do this please create a new ticket with a patch. +We will close it for now and eventually come up with a new plan for automated ui tests. +Please feel free to commit this +The demo has been modified so that using it should be easier +The only difference I see there is that does not resolve via DNS to an IP +Its quite possibly an accident that the approachdiscovery currently uses works at all +Could you have a look please +platforms to RHEL i x RHEL i x Solaris i x Solaris i Windows i +So for this first release I think we only need systemalthough the user guides should explain the implications of escaping +Thanks for the patch Laxman +Im going to leave this JIRA open for now and will merge this to after is merged +Minor change +Thanks Tom yep looks like you solved it +Id did release and integrate it with branch far it passes our local test suites +Thats why it cannot be fixed by just verifying value is not null +Im going to upgrade Atomikos to and see what happens +Committed with revision Committed with revision Committed revision +What is the for? Is it meant to be? I guess that cant be achived using mgmt API? +Before reading the solution I instantly thought about using the settings file +Fixed by UI redesign no +A sample MavenEclipse project with a test that demonstrates the problem. +Fixed issues found in the review. +Also I have restructured the code and put some comments in it +define rarelyPractically a patch to enhance the error report would be best +Attaching graphs and a diff +This issue has been fixed and released as part of release +Just letting you know I do plan to write up my thoughts on this just been busy with other stuff still +Also added to the debian package. +The length in the log is abnormal +Issue is probably with the publish job that links JBT contents to JBDS +saxon dependent plugins in Teiid now compile correctly. +qpidjavaqpidjavaqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccessconfigqpidjavasrcmainjavaorgapacheqpidserversecurityaccesspluginsqpidjavasrcmainjavaorgapacheqpidserversecurityaccesspluginsqpidjavasrcmainjavaorgapacheqpidserversecurityaccesspluginsqpidjavasrctestjavaorgapacheqpidserversecurityaccesscontrolqpidjavasrctestjavaorgapacheqpidserversecurityaccesspluginsqpidjavasrctestjavaorgapacheqpidserversecurityaccessplugins +There no distinction between Sonar the application and Sonar the thing embedded in other tools +on patch v +See my comments on how to do the conversion +Heres my first draft +Im leaving this issue open for further discussion. +Thanks all good now +bq +Josephs solution unfortunately dont work process only chanels header +Turns out not only default Treemap filter is broken but every filter displayed as a and tested + +thats true +must be excluded from +Thanks Michel. +Updated to use the new reopen method on +exampleslibmanagedcompile +I think the Maryann suggestion of trying a double or triple concurrent assign in a unit test a good start +Looks like its probably a JDK bug so Ive marked it as low priority for now +After using a patched version locally for the past few weeks I can say that the patch I uploaded definitely works and stops the plugin from completely failing when running a parallel build +However I do understand your point about a canned piece of functionality and I think that is useful +The syntax union +can you please provide details where which class would move tocant we move all the stuff tootherwise i support jukkas idea to use a released version of for testing +Closing tickets + +Ill wait for the regressions to be sure though +We changes the names around so depending on the packages installed and the os the name might be or +Fixed in r +One of our libs is trying to import this package unnecessarily +The issue has been resolved. +Assigning to Rajith since the review has found problems +Please fixed at svn rev + +OSX will sometimes still do that kind of absurd and around for ages +Oopsie the list of components didnt match the list +The patch looks fine +This makes the xml a little more compact and readableIf you are using Firefox you make be interested in this feature of the browser By default mysterious xml documents are presented as a directory tree with each element being a separately collapsible level +Works hereWhat is your OSdefault encoding +final patch +Committed patch in revision . +Please with the snapshot and if the behavior is still broken +task is investigated waiting for all dev task sub tasks were closed and verified +Targeting +This issue is th on the list of top voted issues so probably will get attention and implementation one day +I wonder what impls already exist out there in the void +It is easy to reproduce this issueWhile you are transmitting a file to the FTP server disconnect your FTP Server from the networkIt will generate the after around minutes which I believe is the default timeoutAt this moment if you try to logout from the FTP Server your java code will stay in deadlock and will never endI already tried to test if the ftp is connected using the isConnected method but it always returns true +If not user might getdifferent behaviors as I explained earlier +I notice that many tests which ran on Hudson have also got a on contrib +Done! Thanks Chris Beams +Travis can you look at show create table as well +IanI am not sure I agree +Assuming there are no more comments from Scott I just committed this +Moving the rest of the refactorings for main to my other changes the class is now down to about lines and much more readable +Looks ok to me +Assigning to a patch that solves the issue +Sorry to be dense +It was fairly easy to usesetup However I noticed that their latest release have a lot of work geared towards android development +I guess your project a and a has the same configId defined in the +Attaching the patch into master with thanks to JeffI also updated the wiki page of. +no wonder it didnt work then +ant test and ant pass just fineMany thanks to Nahi for the fixes +The local WSDL will be checked for types first before the other documents in the facade +Uploading patch with Daniels review comments incorporated +To add to that if the patches have issues going back and forth and uploading multiple patch files each time leads to more issuesGiven that we have a github account as well what do you think of going back to the github way of collaborating? Each JIRA can be a branch and contributors can issue pull requests +Seems this is not worth putting a lot of energy in it at this stage +It wometimes workds sometimes dont +We are working on an integration between Pangool and but we cant release it under SNAPSHOT dependency +I start servicemix +Can you attach the error message please +Sending assemblysrcmainSending Transmitting file dataCommitted revision . +This exception is not because of our code but because of a problem with tomcat release +Alessandro thank you for the feedback +Can you commit this patch to the branch so we have one working branch with changes +It adds a lot of overhead for future solrcloud work it ties Solr further to a webapp and jetty when we are trying to move away from those ties and it puts us on the line for security something the project has stayed out ofBefore this got committed Id ask for a vote at minimum +Then Im going to point my finger at +Tested +The same problem is for FTP GW but as the processing time is larger than for file it bites much frequently and seriously +In this case the discriminator attribute would specify the discriminator strategysvn +The channel variables are a mandatory part of CDL so they need to be present +There are still some issues listed in etc and there are still the other tasks shown above +Updated title of improvement to better reflect the to Paul Nicolucci for remember me this one. +HiI will look into this failureCheers Matteo +Thank you Norman +I also did similar local changes for plugins and that works fine too +The demo examples war file doesnt include this file +Please closeremove +Hi TrygveI tested this issue month ago and I remembered there was no different between jetty and jetty +I just changed that and uploading updated patch +we have to improve the add element method so that it should avoid list to array conversion +These are the first three example files need the rest have to try to get Jira to take more than restTerryI ran the tests based on the attached files on a Mac OSX with or Safari +I was thinking of taking a look at this but would rather let you get all your stream changes in to see what problems remain +Assigning back to Pradeep +For a large scale search situation an index containing M lucene documents now needs to store over M and thus require x the amount of RAM for filter caches as I require now +Ive updated the code and added Axis Final to buildlib so it will be downloaded by. +Resolved via duplicate as the current Dynamic Layout model doesnt completely support Page templates +I dont think compaction deletion at the same time is overly complex and if it makes the scanget simpler then perhaps it would be bestSo far I dont think the scan delete tracker works can you confirmdeny +Assigning to you Bruno since you are familiar with this one +There is a load method which also examines the time stamp on the topology file and reloads if a newer one exists +Thanks for the review Tom Ill check it in now +bulk close of resolved issues. +There isnt time to do this for +When I separated palette I put ajaxjsf and richfaces palette groups into looks like your installation does not include this now I separated these groups further and moved them into must be in your installation for the groups be of tld into palette requires ui check that your installation includes this plugin +I just added a patch to refactor so that transport and do not depend on any other packages from +I recommend that this bug be closed +bq +lets commit +will submit a patch after we are ok with it +Unless there are users Im for removing db anytimeLast time I fixed something there was for the Java version of db a contribution by someone else I havent heard of in yearsI havent heard from any users with questions or bug reports in a long time either +passes with this patch +As per previous comments +Seems best for me to add to my CDI bean since there is no need for each user to own their own copy of +Am I missing something? I did find some issues with the width of some of the program listings though +Is it possible to fix this issue on Hibermate +I will wait to see the code that generates har files +Verified this was updated in the controller class. +Committed to. +Im going to create a new issue for that this one will be about the used for the cell cache +deployed the as +Im sorryI had old jars in project +h and help options now work for all mentioned operations +I changed my spec file to include and and it fixed the issueMy full build requirements now read automakeautoconfpython gccmakelibtoolautoconf +Dan could you please look after this one +jar is the official javax mail jar and has the appropriate CDDL license and NOTICE info the issue raised in about including javax +This is probably more related to central than hybrid tools +One option could be moving these tests to the examples module but we also thought it is not a clean solution +I found the same code in and +This is confusing as it mixes two concepts switching back and forth +Some design changes according to Marks requests +this wording could be more concise too +updated patch with the changes test cases for both IO models +tnx for input +groupId must be problems in previous upload bundle +And perhaps this will include defining a configurable +This patch adds the ant goal as specifiedAlthough it appears to be a big patch functionally it is very simpleThis patch moves most of the functionality in to so it call be accessed by mulitple moves the remaining functionality in to for consistent naming adds to create the and files adds a slightly modified from since it currently isnt extensible +As previously stated in email we are implementing our own version and deviating from the jBPM behaviour +For me I need to enable the null transform and after that any request for a large file crashes during the fetch +Thanks +The last JAXB comment was added after my post hereI didnt know JAXB only works with DOMThanks for your help +I have added unit tests that runs fineAnd btw the source code for the Camel in Action book has been updated to follow the practice outlined in that +Verified on ER Affects and Fix in Version +And Im not at all familiar with the innards of Falcon +will think more post m wondering if this can be committed +Could you add some more detail here? Cheers +Moving issues that arent going to make it into to appears to have been fixed +rebased with the latest trunk +The Merkle tree is sent and registered to other nodes that share key ranges with N +Ill move all the categories into a new module which we can on +Anil is coming back on the th +Please try this one +Have you googled for the issue +Please check this patch and see if it fixes your issue +yeah Ive wanted that too and this is a step in that direction +to people trying to parse revisions in client libraries +Committed branch revision Thanks to all for performance testing +Yep looks good to all issues that have been in resolved state for more than one month without further comments to closed status +Adding annotation handler that checks for interface. +Darn +Push to next beta release +For those blog users motivated to do add a term to a blacklist I think email to the admin will do just as well if not better as it allows the blog owner to explain why certain words need to be blacklisted +While its not a big issue for its worse when you have a one way association where the target entity is unaware of the association +Please reopen ifthat is not the case. +Ill need to find some time to look into this hopefully this weekend +comment +Removed an unnecessary assertion +Btw big thanks for the unit test was very nice to use +Much of this work has been done by Diwaker just need to do a check to ensure all files have been renamed and all references to them have been updated +SoIm attaching them +i likeCould we make the mobile frame darkerblacker so looks more like an modern phone +I committed the patch to branch +bq +committing +Diff between local trunk and apache please look at my ticket +No one of the developers works on freebsd +Fixed in extension code when generating constructor logic was generating subclass constructors for all parent constructors including private ones +This is a new issue itll be +It occurred with a repository implementation that extends +Following Dons comments I think this would be better implemented as an external plugin +That part of code needs to be reworked +I think wagon needs to have some dependency on slfj +Ill keep this open for the momentThanks Hiranya Seems ok +I agreed your opinion to several different text in the same paragraph in a cellRegardsDaisy +And same issues occurs on both environments +Attached patch addresses the concerns +I think this functionality is very important for a lot of users as there is even a faq entry +Will you be able to get this finished for Drools meaning in the next weeks? With unit tests and documentation +Thanks. +suggested patch against current SVN looks good +Resolved as duplicated for ll check it +bq +a new snapshots of native plugin and its components have been deployed a well +Thats true +Also it looks like well merrily go ahead and compact together a MB file with a G file we need to have in place barriers to this ever happening +good stuff +Released Metamer with respin of and +Heres a rough draft attached for reviewing the approach +Heres another patch +It provides parameterized logging a wider choice of underlying frameworks no classloading issues and is a blessing for those who are unfortunate enough to run and have moved to Spring Framework should follow suit +Does anyone else have an opinionI was actually suggesting a general test for +Thanks Antonio +Hi Amadthis change is already present in the latest releases +add a query file for EdwardI updated the patch file with a simple query fileThe commands are likedfs mkdir hellodfs mkdir worlddfs mkdir bei jingdfs rmr hellodfs rmr worlddfs rmr bei jing +It works in CR. +It uses the XMLDOM APIs and will use whatever implementation that those APIs provide +I will check in the updated code sometime this week +We could have this only used in html rendering? Could we not have rendered type specific options? Of course this wouls need a mean to enforce the use only with the concerned rendering +This change fixes the bugThis PR resolves the issue +tests made independent of internet connection and made some fixes to fixed +I was too quick in resolving this issue +Fix for this JIRA +? +Either way you look at it the functionality is definitely broken on some level and is not delivering on the documented intentI am not sure what the best way to handle that is but I think these questions need to be answered firstI appreciate your help and patientsBest RegardsScott ScottFor the producer it will build up the request URI based on the HTTPURI header or the endpoint URI plus the path info that is the default behaviorI think the document is not rightly and when we enable the match option the bridgeEndpoint just ignore the and use the endopint URI with the path infor to build up the request URI +Old JIRA cleanup +Drools version is now and is included in revision . +This is probably more an issue of the Rewriter for your patch Ive applied it in revision +I am planning to have it fixed in the next release +I did take a look at it a few months back and the interdependencies werent as bad as I had thought at first +After all by default doesnt filter anythingAnd now that we have perhaps we should rename to something else like or +Patch including tests +This is a candidate but not committed for EAP an ISE explaining a web module needs a redeploy +So far I have been really impressed with +Will not check anything else in till it is clear again +see. +Patch updated +Interesting wasnt listed on trunk Jenkins build outputJonathan HThis test was introduced in and modified in +I changed the IT +Username kenyee +Lets keep it open with no fix version. +Catch all the exceptionError in the portlet for if the is threw an tomcat error page will show to the end user + +Checkpointed records would keep a high water mark of the last completed replication and the seq num and that high mark for completed replication would both be sent to the changes handler +If it feels useful Ill complete it. +I do not mind checking in the patch with IO exceptions silently ignored if from at least one committer is securedOlegI would be to silentlly ignoring exceptions +As Vladimir discovered on IRC this is an issue where IntepreterCompiler doesnt create COW aware Strings +The used by Web Flow to internally convert from one type to another is not used for data binding of user form input +this looks good to me. +You can then query the server to know what port it actually bound to +I also removed the test from exclude list +This issue is fixed now +Some more fixes for Opera and Safari +For instance sending only portions of the file depends on a splittable format and performing the compaction on the sending side and then writing a new on the receiving side depends on the format of the serialized that would be sent across +Once we changed the scope the jars were included in the final war output. +It really stops you from being able to work at all sometimes +I also updated s unit test +This needs fixing +thats a different thingSurely this is a case for catching and rethrowing the +Wed need to There is of course an option to move the splitting of to an own function internally and utilize this from the other operations which need it. + is a patch for all other for attached is an update for +If this patch is accepted then I will create a new JIRA for unifying the scripts into the hive binary itself +gem update solved the issue I had +The patch was applied in the revision . +In my opinion +Please see attached patch file and zip file containing the source codeBTW I have run this code through checkstyle to ensure code compliance +Opened to thrift fixed this issue +I agree but if we will need to the whole room layout cause the menu style interacts with the rest of the layout +Thanks Simon Please see +Patch committed to trunk +Verified on openOn on jarecho works properly. +Looks like the is getting treated as a T is the Time separator as specified in XML Schema and it appears when marshalling any +I am looking into it +I wonder how validation plays together with it should first be part of the sandbox and then see if it makes sense to promote it as a top level project +Thanks for the update Sid +Re the dtd approach +Attached test used for solve this one just for historical guysSame problem here I can see the fail in the stack trace but no error messages on +Sorry for the very slow response +BINARYDOUBLE is also an oracle specific type and hence cant be done through the jdbc metadata that just returns a value forcurrent workaround is to add a in a file +good catch and thanks for the patch Daniel. +svn ci m Removing JDK protecting bugfix +I just want to make sure youre not seeing this bug anymore +I fixed that in revision +For attachment names to work this test must be changed or even removed +Ill play around with your suggested approach +The best any of us can do is attempt to understand those issues and offer whatever support and help we can rather than adding a form of criticism that is more likely to reduce enthusiasm for an already difficult taskAre you in a position to help move this issue forwards +Would love some help with this oneJonathan Atlassian + +Thanks Srikanth! +If you upgrade to a recent nightly build hopefully the issue will have done awaySo i am going to close this one as fixed +in our case it doesnt matter that shindigs caching support currently doesnt respect content negotiation for different forms of the same resource +Im now assigning this to you Dushshantha as youre completing the work for rpcencoded and +Wouldnt it make more sense to use fsubview since its designed to be a naming container rather than trying to arbitrarily add naming container logic intrinsiclyId say the RI is doing the right thing and not treating included ids any differently +PR mergedI also included to test released +treesrcmainwebapppages in the on the treesrcmainwebapppages +Path that uses set o errexit +I have also attached which contains the testcase and my customized +I will test all the scenarios including rolling restart tomorrow morning IST and let you knowHave you tried this down in a cluster +no PMD checks are reported +The is a good candidate for refactoring and splitting up in multiple classes +duplicate of +Moving this out to please move back to if this will be worked on soon +Now the ordering of whether the or the Spring listener starts first doesnt matterAlso there was a gremlin in the unit tests which forced the servlet to start first +finished backporting to +If it reoccurs or if I have any additional information Ill reopen the issue. +Hey Carl Can you add javadoc class comment to explain what this class would be used for? Also line length convention is generally chars and no tabs +Go ahead and apply the patch to WookieDid not enable the pre configured widget. +Really I dont think anyone should ever want to use the slower version anyway. +This input has columns label and text which get assigned to the target and predictors arguments respectively +or add region encoded name in the descriptionCan you try this first +As project despot I confirm that Razim is a developer on the Mapbuilder project +Moved to while we figure out how to support this. +I dislike the use of try finally in JavaWhich part of the patch are you referring to? Is it the change in +This also makes work in my environment +Has a scan been performed on the other spec projects for missing serializables +for to JBDS betae. +Jevica would you like to apply this now +Great +Let me make sure we handle this in netty and upload a new jar +doesnt seem like a big deal as it seems more like internal implementation than interface +would be a good time to document the two properties listed in the release notesIts now time to update the Tuning Derby topic When cardinality statistics are automatically updated and correct any other outdated information in this manualIf any users see the need for further documentation of this feature please feel free to comment further on this issue +What you sayRegards XXkill p to both BLURSHARDJVMOPTIONS and BLURCONTROLLERJVMOPTIONS in file. +was likely named that way becuase it isbasicly a pass through that calls of the underlyingsocket object +I see you submitted a patch to include with maven +But making sure it compiles etc +As discussed on the mailinglist the attribute has been renamed to publishPermissions +Closing M issues. +All done +Moving out of with Bens permission +One machinemale num Another machinefemale male num The num that we want is +Having all on the Uri makes it even easier to run into that case +I think that this was long since fixed. + +This class internally opens a and adds to +Whitespace fix +Fixed was a cutn paste problem calling integer parsing code +Also note that there is already version available +I cant promise Ill implement all counters on the first gocheers Roland +To all Andrzej has committed this to as well in r at +Adds to Changes so that the the property files for the old test will still be copied +patch related too +The patch looks good to me +This change in rule has required adding a new class called which gets used by the sqlgrammar to make sure there are no parameters allowed in views and triggers +Better not to confusing things here IMOCan you provide a patch using svn diff please I cannot apply this patch with EclipseSVNThank you think I see why the patch didnt work for you +This can lead to some weird conditions like bytes being so we want to avoid dividing by this number before we check it +For file systems such as S which do not have permission conceptInteresting +Hi DejanI am using AMQ as embedded broker in Tomcat which works fine +And yes Im working on a wrapper to keep track of the locked Nodes that a Session has +Was not aware that existed +with one NIT docs are missing +That would be great I think +Committed. +Updated the abstract +to Trustin +Hi Delos Im dont think we should open a defect against the server as the new namespace didnt exist at the time and we dont typically go back and retrofit new namespaces to old versions of the server. +You should run the surefire build and integration tests with only the central repo enabled +Because I am unable to reproduce the error and would likely close this issue as Cannot Repoduce +Please review +Removes sorryIm fixing that +My bad +Thanks alot can you please discuss this in the forums JIRA is not the place for more to be discussed IMO +Additionally x branches have it in the examples +Attaching screenshots to make this clearer +But the problem is the Updating Index process is taking forever and I have a strong suspicion that it is having a hard time indexing the Archiva have Ganymede and the meclips plug in +For now for all the changes you propose +Stay tuned. +But this is a good change +The patch is working fine for me +This happens for different RPC servers too +I will also have to update all of the iOS file apis to expect that an incoming path may contain filelocalhost since people may use these paths with the file api +In I added a warning that tells you that phpmyadmins not scalable and thus embedding will most likely fail +This patch also sets the defaults back to relative paths +Sorry for the noise! +I would be all for removing it +The is applied only when a file is created with default permission +Any chance this could be backported to be included in as has been? We could but Im not sure why +I believe this is too restrictive and forces us to create useless intermediate solutions +With my changes I do not see the hang in jdbc suite +Yes please stop printing the classpath +Thanks Carl! +However it does make Filters more flexible if they are not intended to be written to disk in a normal format +Please review +Please test the latest things the change to dynamicAttributes can be related to this and the second could you contribute your html theme to Struts +Yes we should avoid returning an array of size +Here is my first attempt at a patch it seems to work OK however the testcase I added fails +Need feedback from user in order to fix +Logged format will be TASKROWSINSERTEDtmpsureshtmpsureshMade changes to semantic analyzer keep tack name map +That would seem a shame but is probably acceptable. +Not complete yet but I want to get it checked in so I can have other people try it on other s try that again with the right box checked! +Attached compressed and uncompressed pdf file as well the pdfs resulting from extracting the first page from both +Will commit it as part of +Changing priority as transaction issue may become unreproducable +Is this truly a blocker for +Assigned to release +This fixes the segfault in Viveks test caseIf I dont hear objections Ill commit this patch to SVN tomorrow so that it makes it in for +I committed the patch to. +My oversight +just to be sure +Triaged for +Except for a couple of things I dont like these new in +Changing this to an enhancement request +Thanks Jay. +I am still looking into it +think the fix is easy +Moving it to Seam +Could you supply some more concrete details please Im not familiar with jUDDI +Yes the doc team should have a build script that pulls down the docbook source from the hibernate tools jbpm tools SVN repo and build them with the RHDS style we need to update the docbook source to make it work with the RHDS style sheet we should make the changes in the hibernate tools jbpm tools SVN +I want to apply to +Why should the binding process care if a nested property was not explicitly instantiated? In my opinion it is VERY annoying to have to manage the creation of nested object graphs +Appears to be because explicit is only set if a propertyNode matches the identifier +Therefore it is possible to get an OOM error if you send persistent messages to a broker that has been configured with persistence would only recommend disabling persistence on broker that is either only receiving messages +Garvinis this a duplicate andor is there a where this is fixed +Also it does not seem like a regression yet +We have to reopen it because the code structure has changed +Thanks for the patch. +Should we close this issue then Zhiyuan as fixed or will be fixed by ? Thanks +Thank you Benoy. +part of release. +Is there any reason that we cant do this automatically without annotations +Will review this again +In this patch I removed interface and renamed to and moved javadocs to the abstract from interfaceIm still working +It works! Thanks so much! I will change my code base that generates the PDF outputThanks again for responding so fastThanksLance CampbellSoftware ArchitectWeb Services at Public Affairs +Closing it. +Thanks +Patch no longer applies. +This version consistently passes the jrub online testingrestoring attached to +Is this related to +Here is a patch +Patch that fixes the problem by adding a new class extending the class that sets the default in the constructor +Fix that first +Since this is blocking the ability of people to build with Maven until this is worked out Im going to disable the alfresco part of the maven build and tests +Jo no crec que sigui especialment formal +This issue appears in Castor as well +tested works on this to trunk and +bulk close of all resolved issues. +You are not describing anything new +Yes there is Searchercache warming but Im not sure if that comes into play any more with NRT and soft commits +Im working on this issueAt first I fixed the calculation of the family class in revision +Whilst youre correct Geb would be better with a in the module directory saying the same or directing the reader to the README at root level. +The NPE is fixed unassigning the bug so KVM engineer can pick it up and fix the backend failure. +Thats good enough to close this issueI had to do some ugly hacking to get it so I could sync the backing stream on a and then in so it has a length which is equal to available bytes +Thanks Ram and Nicolas for reviewing it +I thought about a test case but I didnt find what to do +I havent been able to verify this but from looking at the code it looks like the following changerevision Fri Jun the filesrcmainjavaorgapachemaven causing the problem +Yes a verbose way of getting true or false but a sure way of generating those values through an example +Moved the to clinet packageRaised the separate JIRA for working on sequential entry writes + +Will try v +Martin have you tried the cluster tests for this or pastecode recently +I got three failures in when I ran the All suite on a clean trunk +The other event handlers are ok +I want to show the missing pageno a typical case of +Attachment has been added with description ImplementsSorry for the delay +Ill try to get to this today +Yes +Resolving see for last remaining issue +adds two tests to the spatial contrib test project testing the cache works with multiple readers and the allows access to all indices +Committed at subversion revision +i updated the code to do just thatthe highest in the stack will be the next active one. +Siarhei We cant support arbitrary project layout for GA especially Maven generated projects are tough since Maven regenerates teh eclipse metadata and our info is lot but we have it as a priority for release to handle this Havent we yet added checks for if the actual destination dirs exists before the wizards are run? If not we need that fixed before GA and tell the user that the wizard cant run +code reviewed +Looking at the other methods in and comparing them with the one in I wonder if the right place for this code would haven been rather inAnd the unit test code should use instead of using the JDK method +Buf unfortunately the activation does not support the varible which defined in the POM +logs UsersjoaDocumentsworkspacescompanyworkspacebin +Same patch Rajesh attached last week +This issue not any more relevant for current trunk that was replaced by branch +Dont hurry doesnt included in roadmap for current version +Yes +Sorry +I will investigate further into this +These changes were done from scratch and are not from Andrews branch but they do share a common goal and overall design +A zip archive of the sample project I haveAre you saying that Subversion is complaining about the status of UsersemilesProjects which is a folder that is not under svn control that only contains three folders that are themselves checked out from Subversion +Should we commit the fairness +backported in fddaadeadcfe +See last comment +This bug has been fixed in Xalan Java +If I dont receive sufficient information by the end of Friday October I will exclude it from the Release Notes document +When I clicked the second email I got to see all of your comments +Add How to get row selection in scrollable using one and rows mode? section to the Details of Usage or selection attribute of scrollable +Its more easier in ofbiz to create the shipment with a shipmentId given by the master +Thats actually harder to fix than it thought but I added a newline so it wont saygrunt gruntbutgruntgruntLets just tell everyone that thats because they have extra newlines in their scripts +class files dont even track where constants came from +Most things seem to work well including the but it has exposed some problems in the +Duplicate of issue +This is the pom and the code for the If you take a look at Bretts remarks of Sep youll notice that he mentions jvmargs as a required addition to the supported surefire argument list +Regards Alex +It turns out that the distribute step cheerfully distributes new code and makes a new entry in the config store but without stopping the old code +The patch shows the bug only +quite the description +duplicate of right +This is a very low priority Denny so it can wait unless you have absolutely nothing else to work on +If so please check what exceptions you see there +Thanks Mark +Ah thank you for clarifying that +If you set it to NULL it will recalculate the value and set it for the feature should be available in the issuetracker tarball i checked in under assets +Committed +Would anyone like to take ownership of these fixes? +Also it has a Fireworks PNG option using which we can generate the FXG skins as needed +Sorry I didnt mean that it would fail or not change the setting +This seems sufficient since we are relying on the admin to explicitly enable this property +updates or progress on this issueThanksGC +samisaI can compile the service now +Set the column width to the max column string value width +So i will add some more validations +Use the Workflow Link Pull Request option to specify the URL for your pull request +Sorry forgot to addPlugin after this error reports that everything is ok +Sorry to be picky but can you submit a test with no external dependencies other than LuceneHighlighterI dont want POI versions to be a factor hereCheers problem occurs when using this exact document and other document which is pdf +Ok I will take a look at this +I think it has been constrained so that the log is not flooded but important events are still logged to permit some analysis. +The comments in the patch are there to explain my thinking feel free to remove themThanks +Committed to branch +Hope you get your account set up soon Thilo +Putting the patch on review board would make reviewing easier +bumping to since since the linked issue isnt applied yet +Postponed as we dont have a clear answerI think that both returns are OK +Maybe because I mostly do stuff on DB and never even stumbled into something called SEQUENCESWell +Yup should be +Steve I think you referring to where the git references were fixed +Also you dont want to have jar in your classpath if you are using Spring to inject but not running camel in container +As you know I have studied this patch multiple times I think it is excellentTwo thoughts if you have not yet committed the patch It would be nice if you could add some comments to indicating the purpose of this test +The fix is committed in the SVN +I recently heard of a related case +Trace is a more accurate usage for logging each crud op debug is appropriate for all of the timing statementsand info really should only be used for notification that weve startedstopped +So please be so kind and keep the advertising out of a bugtracker +Otherwise youre adding a pretty big security hole with the table links feature +Please review them thanks +Im not sure if this relates to it does not provide enough details for me +I didnt know about this option +I think we need to remove part of the patch that causes a cheatsheet not to be opened when importing a non maven project +Attached is a patch to break the main for srcgen into a separate class +Hi DanI havent seen any problem but thought it was a general problem and we could update itI can revert it if it is not necessary or it has some negative effect +Move to be in for now? We can move it to public api when neededSure +I verified that the blocker is no longer seen in the latest build from Master branch +This may actually be causing an interrmitent test fail on jenkins +remove patch available until the shipping problem is sorted have a look soon but Im feeling were getting a road block again +Not sure how this got cannot reproduce status +Actually I had hoped t that Lilys work on would resolve this issue but it seems like it is still really hard to get Derby unloaded once loaded in a class loader +Can you reduce the test data size I mean create a new table with fewer rows or you can use src to show the same new patch integrated Namits comments +Is it Okay +Temp patch looks good fix for ignoring mkdir chmod and chown if these operations have been performed already +Appears to be fixed with build jar harness using java xmlxalanjava and test from CVS late on Joseph if you think Xerces has fixed this please close. +It was part of other unrelated benchmark + +It would be nice at some point to allow multiple billing accounts or store credits on the same account +How is this patch? I simply changed the syntax and examples sections of the two function pages to the escape syntax as described +is there another way of avoiding the use of this tag? +No biggie either way! Thanks Oyvind for updating the docs too +It was done for unification yes unrelated +This issue has been in the Patch Available state for several weeks +I looked at missingoverwrite for example +Hi Martin many thanks for the time you spent to review the code +Should this be closedIs this a Cayenne or application problemThe application can implement these sort of access lists +Ill try this afternoon and close this if I cant reproduce it +This causes confusion as to how to configure server +This fixed missed proper computation of projecting back out the tuple from the index which can lead to an exception but in the worst case could lead to inaccurate results + +toggle from sounds like a good idea +Thats the only thing I see +We can create a new jira for to get all the remaining source jars into the Maven repo +If this is not the case open a new more specific issue +An append happens a few days later to extend the file from K to K +No nothing +Run fine for and derbyall for insane build +But in addition it does also mean that the patch should not contain changes of other tasks +not sure if its already fixed but we should definitely be defending against and reporting something properly +Late session save actions can invalidate sessions or you can essentially your audit log saves +I forgot to mention that these cause problems because the code using the typically doesnt know enough about the request lifecycle to know when it can safely remove the +Please note that there are spaces on your JSON values so you have to type those spaces on the textbox to match the options +I will try to use it for testing purposeHadrian as Claus mentioned its should be possible to use advice withI think this issue should be resolved because in the camel will be capable to replace route inputs too and in earlier versions of camel a custom can be developed to replace route inputs +I used Maybe tar or gzip did convert the text files to Unix format + +Did you miss something or can it be removed +The file needs the right name as well. +added unit tests to svn trunk to verify correct behaviour +I know youre using this in the processor but its probably better in the writer +Thanks for the help Zhihong +The problem is fixed in the latest build so close it. +Then the hosts are adding without issues and PV drivers can also be installed successfully +I assume this should be an open issue +Hi Rajesh This has been fixed for EAP and verify in in EAP CP CR +One twits that I am fixing that may be contributing to these errors +The Jenkins server really should use a newer JDK than u. +You are talking to open source developers who spend their free time here to diagnose your problems and to improve software for you for free +Spring Core issue is marked resolved in. +I had this problem with +wrong all minor and move from M to to M because of no updates from +Completed part as described above +Thanks Dan the links are working now +fails on my system probably due to this fix +Sorry I meant the patch for P open bugs to P pending further provide an input FO file for added an example to reproduce the problem +but did you measure the footprint change? If Eclipse startup footprint grows by we need to invent something more intelligent +First of all thanks for the great effort so far! Im running into a merge issue for properties on the root node +Carl good Monday +User could then continue browsing the portal to another locationIn our case we have portlet called Site which displays site information +I looked at the taglib docs and it looks like this is made clear +I think this should be reopended because seems stalled it was bumped to next version again and changed to minor +AB Seems odd to a return the timestamp value as a and b return the timestamp via the received Calendar object +definemethod now registers proper byte symbol so that later reflective methods can find that symbol +There are serious problems with this issue It will break existing applications styling Button is a core control of Click and it wont work on IE without hacks +If you use over cluster you will also use duplicate detection over the bridge when transferring target should ignore the regular duplicate detection if the Bridge duplicate detection is said that this is not a regression +Also I hope to add instance management capabilities in next two daysDeployment process is same as mentioned in previous comment +Fixed in r. +I tried real hard to reproduce this with a couple test cases to no avail +All of the attached test cases no longer throw exceptions +This is the stdout and stderr redirect from my session +Thanks +Removed the selfClause testing in as its no longer a legal value +I have attached new source file no ldap package in it +Patch applied with thanks to Eoghan Glynn. +I will polish this +is abatut +pluginssrcgroovyorgcodehausgroovygrailspluginsspringsecurityldap +The String constructor is there merely for convenience for the most common usecaseWe are not going to add constructors for Integer Long Date etc +We have changed the topology of the backups now +Even if the is some code on the way for a pull request an adaption or merge conflict resolution should be done easily +Test passes in and a much more extensive set of pack tests live in rubyspec +Thanks +I want the ability to specify a min and max for the threads in works for me +We can probably work around this but I am not sure if it is a great idea to change files generated by the archetypes. +I recently upgraded from to and the memory usage of our application increased dramatically +surely there is something too more in the parameters sent through the test case +We dont want to overload it with other functionalityI think we need some helps here +Look at the unit test promote +s strange is that if I remove the EAR and redeploy it again it works +This will have to wait the next gen build system +wed used to have wrong input types for certain management attributes +same issue for me +complete +Im fine with whatever we can make work reasonably well +Jenkins is wrong +Is there a reason that the random UUID is needed on top of that +Made it critical and for +It may be that the changes that occurred with on Apr caused the resolution of this issue . +Rebased in vFYI To enable we need to update withexport LDLIBRARYPATHjemalloclibJVMOPTSjemalloclib withmemoryallocator +The retyping stuff is still quite foreign to me so a review of the patch would be much appreciated +b They should be able to use alternative packages sources for some of these packages +Same problem here I also confirm that it happened when running unde update update is ok. +Are we discussing compression model for storing lucene indexes? We can store indexes in Sequence files and it is scale able to large extent +Glad to see it is already done +I am testing EWS and the problem still remains +It would be very useful for debugging and developing giraph applicationsIn addition to unit tests I tested it with a simple graph generator and it works well +it rocksThanks +I also prefer to have a substainable solution for I never ran a Castor version without applying some patches Ill just continue to do so at least for some further investigation of this new patches will at least help me to do as we discussed in the chat last week I am sure and would bet euros that this will show up after the release +Thanks +This still happens in Grails final +Ran derbyall and committed Myrnas patch +The design discussion should occur in the associated jboss forum reference +Your fix looks good +Should we also remove OPREADMETADATA fromAdded a new patch that deprecates the variable for now +Thanks! +The patch works as expected +would you please have a try again +Theres no default support for it as far as I can tell get does not work and I believe wed have to use a custom format string which then would not be localized. +It contains the where the xml writer is fixed with the geoserver changes and the Applicatoin resources and new page +I also went back and reviewed the other classes and I could not find problems with the exception handlingI will live the JIRA issue open to be used as a discussion point +In XHTML attribute minimisation is not legal and the recommended value to use is this type of scenario is the same as the name of the attribute itself. +incorporated is not getting serialized because it does not have a default constructor and set methods +Need to resolve this in a some way or create workaround and create public explanation +If not how i can download the rc version? +Open pending final edits by David +Preparation for Gora release candidate +Note that items above refer to previously unreleased classes +Suns phoneME platform is JDK based id JSEThis also doesnt make sense because claims it implements CDC Foundation Personal based upon which I assume means which are all subsetted from JDK +oh +There werent any exceptions or anything in the datanode log? That error will typically happen when it tries and fails to read the block from where it should be so hopefully there will be some errors in the DN log +has a test project and patch attached to it +These two components require docList in them instance passed to them while code to do grouping doesnt actually set docList and docSet like the normal +These are generally issues that are felt to be fairly straightforward and do not require a lot of prior knowledge of Derby in order to investigate +Hi JeromeIve applied some of your patches +Looks invalid to me the correct code would bedefine foo do artifactns do ns abjar abjar endbarendEventually change the version only using +Thanks Aaron for the review and commit. + +GeorgeI am not sure about replacing Hadoop with the appends version +Heres the version +Are you saying there should be twoI looked at the transient state issues and and didnt see adding a second notification mentioned there or in the threads linked from the issues +I was thinking about the read case when I wrote the comment I think what raghu stated makes more sense without my additions +For now Im happy with third party athena builds contributing their own categories to the site +Just bumped into this on a fresh install on our test cluster +If youwould like Ill try to get a JSP together +This is a duplicate of a bug fixed back in early February +Siying Looks good for the most part +net configuration service like php +suo file is needed as well. +I dont recall why I opened this bug +bq +Accidently attached the svn help diff output instead of the patchThis is the patch for real +tested patch on linux and windows operating r +Ill take a look at your next patch and see if I can contribute with anything of valueDo you mind filling out the of this issue +If the is there then code should also not work +That consistently gives the correct result for me +Previous project slip reduced our timeframe for these +where to store the namenode transaction logCase +Tested in Cocoon +We really need to fix local test runner to handle multiple reduces +Updated link in description to point to the right project +A workaround for this would be to make your repl large enough that you would never hit it and instead rely on your background flusher thread from the +Added a simple block to the recovery manager startup for test mode only which seems to deal with the problem + +yeah my issue is resolved but see the comments above +A trick to be released in will greatly improve startup time +Zookeeper does pings and deletes the znode when the session expires but those arent updates per sebq +Fixed by Permaine for by the new zips on now has on the top level dir. + to trunk and branch. +Dupe of +I am not going to investigate to look for the deeper reason of this behaviour +A use case would be +In this case we should disable Create Test Project field in Seam Facet PageThat field should be available only for New Seam Web Project wizard +The central border is set on the column ofthe inner table which has no forced mininum height so it grows up to itsnatural height which appears to be less than +looks to mention the test failure is unrelated +The attached patch is a first version which doesnt yet show the problem +Initial version of updated for trunkIf no objections Ill commit the patch soon. +The upgrade test Upgrade is not included in this patch because the patch submitted for first needs to go in and then the upgrade test will give the correct master diff +browsers on Windows dont have correct colors either +Well mostly here I was admitting that I screwed this up for One of the major reasons we respun was to ensure solr used the most recent lucene version configuration so it had features like this +Thanks! +Anyway I check for abstract beans now +I guess upgrading should cause this problem to go away +It includes a new test for the security null in the case of a prefix bug +I cant reproduce the issue +Will keep trying +Its odd that I dont catch the compile time errors +Pending Jenkins results +bq +I know that portability requirements encourage to do this processing first however for cases like these it might be helpful +Could some committer review the attached patch and commit it if considered kosher +are good reminders so Id say leave them +Configure CC SC ELB and Agent for +Not unrelatedWe are processing whirr output with awk +Hi TimothyThank you very much for taking a look at the dump! I only saw the memory corruption in the lab when running our main application that uses CMS +so it is absolutely correct to ignore the order by in the HQL and replace it with a different order by and to not do the same when using left join but only left join fetch? okay. +Do we really need row groups now that we can have arbitrary nesting within a row via composite columnsWhat about secondary indexes? Unless we add composite secondary indexes +Updated tests to patches in caaba and fcdNice job +A brand new more appealing web application is going to be implemented. +What would you suggest? I would rather see monitor connections of just one connection manager and would rather have the stay alive as long as the connection manager itself stays alive +Huy is correct we cant simply mark serializable because it isnt +Attached is a patch. +Go ahead an create one the first steps to make it possible to disable server side are in but the default config still needs a +Vladimir make sure you create a new JIRA for any further work pending. +Is there maybe an issue because Im using Tortoise? +Contains source that fails to compile +Talked to Amareshwari offline the proposal looks goodAs a we can merge consecutive filters in a with following changes creates a filter operator with unpushed predicates as a child of the operator through which the predicates could not be pushed +Thanks for the review and commit Knut +PDF document showing the issue +Bounce to Burr to review then back to QE to close +Youre welcome! Thanks for the notification +Puppet changes +I forgot I was working on this before trying to work around this leak problem before zipping the project up +This fix looks bigger then it really is +Merged thanks. +fixed in svn revision has been checked in and seems to work well. +does anything fail when using those characters +was released and since this is pre this is not a requirement. +typo on line in var name but doesnt hurt typo on line in var name but doesnt hurt line cut is brittle for this purpose +Yes I think they are related +Thanks Knut! I agree its a bit weird that the JRE is a little schizophrenic about the interrupt flag +Jacquesthe warning message from is not related to the new style and it is fixed in rev +Its been tested against the replication unit tests as well as a cluster test to ensure it does what its supposed to +Prior to Java is required to compile sdo impl as illustrated in sdoimpl +This patch replaces the old one because it missed some modification in the junit test helper class +del + and should probably be done together since someone will have to touch every page for re using jekyllrb now +What do you think? my dumb hack with a less dumb one The methods now test whether the class implements a one of the interfaces or in trunkr in branches +Each subcolumn expires in seconds +The append design explicitly states that the system should make every effort to persist data written to a file even is sync is not invoked +Ok I got long as it doesnt cause a regression for then please apply. +Applied +Committed +Well thats trueit does +If someone wants to pursue that I think the effort merits its own JIRA and discussion. +Thanks for the quick reply +yes it looks like i was thinking the jira attachment ing would be chronological +When is expected to be released +Jimmy is invalid +Would any of the mentors care to comment +The CMS client is not maintained on AIX since there are no resources to test it there +Patch is the mockito timeout Xuan for the patch! I dont have anything else to add on top of what Luke already called out +After having chance to have a look at jbpm which is secretly applied to one of our clients +It would seem that the is dependant on the library +We do detect that the transfer failed but we dont properly use a temporary file +Thanks for the patch +In both cases comment annotations would be correct in a way that the street is the name of the street and that the denotes the of the city or area + +Test caseStart debugging +I have posted it twice due to a browser for be marked as duplicate of +The project still works but someone has checked in new compiler tests as it seems which fail please use true until this is resolved I will look into this issue tomorrow and will fix the tests or the underlying cause +Patch checked into branch and trunk +Also if the read lock is obtained first by the thread before the write lock then a dead lock is fixed on trunk revision on revision +Thanks pass over all release notes prior to publication of +Keith could you review this pleaseThanks look good to me. +The only thing Ive noticed is that this branch is using so maybe this patched version fixed the problem +Hi SureshI updated the description with some more details about the enhancements +I would like to see it someday when we are capable of adding the eager delete function in there however I coded around the issue for now +This is really really important! Current is from and uses a completely different options set than current versions +Thanks Knut thats a good suggestion +Fixed +Thanks Dan +I think there is value in covering the entire normalization issue in a single class +The hardest part was documenting it +Attached is the patch with new driver and tests +I have one question about the point where you throw a runtime exception but otherwise this is ready to go +here is a patch that fixes this issue and adds some asserts that make sure we dont wait while holding the IW lock +Sorry if aduplicate pops up in the wrong place somewhereI dont think all of the changes are correct and feel we need Wolf to comment onthis as his message states that he ran the unit tests successfully on his codeas submitted +keyword does not affect binary or source compatibility +Not working on EDIT These issues are open to be picked up +What was the fix? +Other general questions like this please direct to forums +Should we move this out of? Will it be easier and a better fix when ZK is in mix +Havent looked at the hbmseam output in enough detail to comment intelligently +from me thanks again +I should add that there is still potentially a small window of vulnerability in the compiler itself which may not be happy with the filesystem changing underneath it +I got rid of the class finished the and provided skeletons for some support classes such as token stores etc +If you have any further issues on this please feel free to reopen. +Hopefully will fail more gracefully and give some useful informationI believe that was failing for Myrna because the program doesnt know the names of the IBM on Windows +Attached WSDL and along with the generated source from tool to prove that the parameters are not being generated correctly +the default values can be handled there +But we probably should cap this at something very large but not insanely large +Resolving as fixed. +Attached sources +Attached are the patches for trunk +The filter is configured for +Committed to apache trunk +Im a little hesitant to do that +Ill try the snapshot on vmbuild later this week since I get it a lot Brett +Let me know if I need to do anything elsethanks patch is in SVN + of the to handle errors better including checking the and deleting the partial file if something goes wrong +Patch for trunk +Resolving platform issue as the project peer has been resolved +anyways feel free to submit a better version. + +Reopening per Karels comments +Sorry just found which suggests atleast part of this functionality is there I must have consulted an older version of the API beforeHowever in my experiments I had tried FAILSILENT flag but the xa transaction was abortedIll double check again and close the issue if works otherwise I guess it should be a bug rather than feature request +Is this change incompatible for rolling upgrade + does not implement Cloneable in RI +Hopefully fixes hudson breakage +I was thinking about changing it to in the trunk version but probably will do in a separate issue +We can perform distributed collapsing fine when searching on some other nontokenized field the NPE occurs when we perform a search on a tokenized fieldAnyway Ill attach the small patch now which just adds the null check to Solr trunk +Can you make this clear in the script? But when i try to run it is not reading from the file but it is running all the testcases +Attached the patch and just reverted it from trunk. +I am going to conduct quick pull request for this one if you do not mind +Now it aborts +We want only blockercritical fixes into CR +What problem are you trying to solve +Since we use the import feature much less than previous versions I am moving this down a level from major to minor. +It doesnt support sort queries +Looks like it is and does not seem to affect any existing code +Is that enough for you to verify +Would it be helpful to make changes to to be able to support the command line options such as additionalJars at the same time we make introduce the new functionality +Now I can have junit test result when tests failpassCan you please release it to maven central if you can +After spending copious time on trying to figure out how that failure could be related I backed out my changes and tried it again and I still see this failure +There are problems withI tried to fix those but now I have two failures in testModify and testCopy +Id rather find other solutions for each such advanced JPA setup demandJuergen +Each Fuseki instance is running on a different portThis is primarily done so that databases can be managed and brought up and down entirely separately from each other +And we need it for the bootstrap problem +Verified that deletes datanod and tasktrackers in hadoop +Anchors referencing to put and get sections from copy copy is currently not working +Dont know how to fix so revert seems sensible to avoid spamming the mailing list with jenkins failures + +defer remaining issues to +I looked at the listRouters listEvents listEvents and list and in all cases I found that there is no change in the code between and +This file should be downloaded by the user +Please use one of these or wait a few days for the final to appear +the dead letter channel afterwards +This is equal to scenarios with little to no effort what you suggest will neither support simple nor complex scenario as it does not respect the excepted behavior of a that is the base object for the is a good idea to allow programmers to override the principalChanged method but if he decides not to do so it should at least do the right stuff +So making the footnotes in headers repeat give the user the choice of which behaviour they want +Not supposed to be resolved until commited +Just something to keep in mind +file add with includes changes in please note in Environment that Im using a patch applied from too +Please submit either a test case or detalied steps to reproduce the problem +Do you want me to publish numbers for more set of regexsIll create a patch for rest of the comments +Update wiki with this new am behind a poor firewall at the time so cant commit the code will do later when at home +EdwinThank you for your interest in my patches but your proposal is a lot can ignore mine with no regret and apply your patch ASAP +I switched over to Oracle Java and compiled just fine +Investigate security and handling Redesign the tool API for simplicity Allow exporting of multiple query plans in a single operationThese issues seem sufficiently complicated that they would be easierto track in separate Jira entries +We hit the same problem with deflate compression +Or it will be better not to commit this version until matrix package will be tested enough +Sorry for late response +Unfortunately the fixes for cant be merged to the branch because there have been too many other changes on the trunk +Included in release. +Sending the patch againMartin setParameter when decoding an URL at setURL was introducing the described bug also thanks for notice +the wrong generate sld file even if I do not think the problem is the incorrect code in it was returning the fallback color of black if the parsed color was non as of r we can deploy and confirm the fix for uDig to close the bug +I checked in project cvs latest tag is version will be on ibiblio before on sourceforge +Brute force approach have existing impl tests extract their initialization into a template method +Thanks Chris! Patch looks great +Oh you mean cant get the SQL statements +hopefully its easy for you to reapply that piece +This also causes another problem which is that the +Nice one +I will be uploading a new one shortly +No problem +Hi JMIt would be nice to have bash or ruby script that is doing this job +Sorry about that! It probably needs a pass over with the profiler +Sending trunksrcmainjavaorgapachepdfboxutilSending trunksrcmainjavaorgapachepdfboxutilSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputTransmitting file dataCommitted revision . +With patch for it is possible to create different type of indexing service by extending +Thanks for the review Suresh and ChuanI have just committed this to trunk and this go into too +Sorry but I dont understand plugin doesnt work? If so whats the reasonAnd no currently there is no way to excludeinclude plugins from analyzes +I will commit it shortly +One thing i am not sure about is that we seem to have practice of using from apache commons although this does not consistently apply to murmur utils +At the end of the day if hbase is down having a slightly inefficient automatic solution to fix it is better than having no solution to automatically fix it +This appears to be a JPOX bug +We know that the fields are not retrievable but they are searchable in the original db +Since the structure will now refer to property files rather than Ant scripts does this mean that the Ant Task will now require the Sonar Runner in order to perform analysis +Thanks for yet another great contribution DanielIm currently reviewing your patch Just for the record removing previous patches is not required you can keep them the upload date helps us for reviewingStay tuned +Sounds pretty goodThe example can contain the external resource in srcmainwebapp +Here is the thing the action invocation needs to happens at the end because parameters can be passed in the body +I had to change quite a few things around +projectbootsbt +No comments on the edaafbbcdccdfcaee fixes I made? Should we close this again +Fixed in r +I would place the supporting logos just in the footer area +Ideally if you could provide a small test application or testcase reproducing the issue that would be great +reopening and will attach my edits. +The patch was applied +on how did I know the parameters name of each Decoder? can i get them from the Decoder class +That is a good question +This patch adds Socketconnect and SocketconnectnonblockI used the examples in to test +Applied at revision +It just search for CDI beans by given name +I do tend to think this is mostly a sort by score issue as well but I can see it being asked +Verified. +As soon as a developer is using hibernate to persist their object model he would like the pvm transactions to participate into his application go through those different uses cases we could add a simple end user application calling the through three different is one additional execution mode that we should not forget In the enterprise environment we should make sure to test both usage of a Local ejb execution service and a remote ejb execution service +will likely get renamed to something like. +Once we have common interface you would be able to reuse some of my tests and benchmarks +Kristian +I did then dig further and found out that the bundles were activated lazily which is not valid for bundles that have services +Why does call instead of itself +This should be resolved in the cvs version of openejb afaik +Rob it is the default project from openshift and the html app from jboss special about these but you can get them from there if you need them +I guess the problem has been solved in the transition all resolved issue that did not see any further comment in the last month to closed status +Id like to pick this one up planning on doing a bunch of Pig related work tonight anyway +Ill make the change that you indicated for the Ref Manual and draft the text for the Dev Guide +Feel free to reopen it if you can provide a patch. +Yeah that configuration stinks +Jakobi read the code of the haivvreoand i think it can do with protocol buffers like haivvreogoogle s paper tenzing like hive said it support protocol buffers and columnIO +ok Ive hunted around for the missing bits and pieces and mvn build now runs correctly so this issue can be am not a maven user but each time I use it I find it very brittle +Looks like a trivial patch I have tested a running version but have not yet run the tests +Forgot to mention I have not tried this on trunk yet +you will have to fix it yourself for now in your own code +That should be the default behavior? Is this not the case andor did it change in +Much appreciated +Same patch but with the option to git diff +If not thats fine +Seeing as we released using this patch Ill assume it worked. +assign to review later +Now I used the same approach for components and on +It looks like the culprit is in +We generally leave issues resolved but not closed until the release theyre in goes out the door. +This would be nice to add could you update this patch to x trunk Radmin +This test will demonstrate the string tofrom byte array conversion issues +One of my app is having exactly the same problem +Change applied for EAP andThe id must be a valid integer between the range of to +This seems to be cause by failures are due to changing the security managers but not changing access managers and login modules accordingly +Can you point which EE spec document says something about being a location for classes? If its not a standard one then this issue shouldnt be fixedThis issue isnt to allow classes to be loadable from within +Attached patch make cquuh and cquup work if request is valid +Is this the only problem for Linux or other platforms +It might well be that a connection that does not match for one request does match for another one +Please also see the log for test progressHere the test in prose requirebundleB resources requirebundleC resourcesresolve BC requirebundleD usesresources resourcesrequirebundleCresolve Dverify that D wires to C requirebundleE requirebundleD resourcesresolve Everify that E wires to C +Otherwise looks great +We should perhaps also rename the component name in the JIRA and somehow mark Maven as does not exist anymore +With clarifying comment +Daniel please verify +They should probably be mergedconverted to. +Getting the same thing on unpatched +HiAre you sure that this is not due changes in the way dates are stored in vsHave you tried the same test with an index created by? Did you see the same problemIf the answer is yes to both of those questions please provide a sample code to recreate the issueThanks +Adapted scriptingjsp and scripting projects in Rev +Now I fixed that +Unit tests are what are relevant for the individual piece +But looks good to meIll commit this patch days later if there is no objections from other committers +This is why I said user error +The duplicated logic is all fairly simple and some of it was being duplicated anyway already in the old nothing else I definitely want the new comments and documentation that the patch introduces in place of the old ones +When the variable is read no converter is applied to the value because none was saved in the database +bq +No threads are blocking anywhere trying to do the deployment the blocking threads are management layer threads waiting to be released via a notificatopn from an MSC thread that notifies a listener on the that callback isnt happening +After manually deploying the necessary sonar files to our artifactory we realized that temporarly changing the checksum policy for the codehaus repository would have saved us a couple of minutes work +Prints out the results of two convert calls one without ZONEOFFSET set and one with ZONEOFFSET set by forcing it to be calculatedThe second output will be which is correctThe first will depend on your locale but unless youre in a locale should see a different result +It should be noted that the +Barbie says GIS is hard +After serializationdeserialization you will get instead of custom tuple +Emmanuel well I have tried what you proposed but the build failed +Lets review the protocol in detail before committing too much time to implementation +Website up +The idea being that when you update a feature in a versioned dataset its fid will actually change +Should this be included in the release +Fixed in and trunk +Could this be a option when using the Tika JAR +Unit tests pass +A few standard C integer types are supported char short int long sizetMany others are not the types from signed or unsignedanything long long ptrdifft offt etc +I got this in H build from from Critical to Major since we just get those exceptions in log and validation will eventually get the correct javadocument model and finish validation +You might want to check if the new API exhibits the same problem so it can be considered for the contract there tooI dont think this is a blocker +Rajesh is a duplicate of this issue +Great work Hoss! Im sure developing that test was. +Ill attach a patch later for +Looks good! I spotted missing Override in REVERSESORTER and I think that we should make the impl final? Maybe it will encourage JIT +An example will demonstrate the usage of the functions and the new explain functionality +on patch ll fix Ted comments on commit +In general I thinkyou want to set the priority on the and not on individual renderlets + Hight level documentation Terminology Usage for the patchJust applied the +Application specific plugins are way down the line on the roadmap and we need to get the base platform working well first +closed issues for release +Modifications to to hold on to a copy of the column meta data for each entry used in batch updates +Committed to trunk and +As an example hibernate will sometimes return a proxy object instead of the actual user object and if this proxy is attached to the cache POJO Cache will serialize it since its not tagged with Replicable +You are right Nick that issue disappeared with renaming varitem in template to another name. +The first variant of my class doesnt include static fields but in this case it work enough long +It appears to be a metamer issue as this is the only place Ive been able to reproduce +Whats the input for Eclipse APITo what extent should this go in terms of reflecting all those system props by which various subsystems and deps may be affected? Maybe the known issues workarounds could be adressed by the APIOther thing Should this be a separate project aiming to be backward compatible with older s or would it be simply tied to releases and guaranteed to work with just that version +Works using the latest distro now that is fixed +Nice workThis might be my own lack of knowledge is the SBN in charge of the fsimage file? If it starts much later than the NN does it fetch both the fsimage file and the from the NN +I suspect this needs to be in a version as well +Good idea thanks for the submission +ill ping him +Thanks! +Patch to trunk +According to it can be fixed by setting the CLASSPATH env variable instead of putting the path into the command line +Ill try the following instead identify a sampling jobs children and set them aside somewhere then check against the saved list of known jobs and the estimator for them if parallelism is set to +I will look into that at a later stage +Fixed in +I think eclipse does the lot of work and needs complex testing on various platforms rescheduling to +Fixed in SVN with the new reporting API. +settings is also in the is actually on purposeWe have voted this some time back +I think its a great start and we should try to let lots of other stuff follow ASAP +This wont work for appending to but this should be the best solution for the formsRe new object registration +The only way to solve this is to generate for update of col which is probably pretty hard to do +Looks like upgrading to SLFJ causes trouble to deployments so Ive switched back to the latest release +Done in branch +That patch will go into the next release of JANINO +Fixing it. +Committed revision . +But I think the performance penalty of overflowing to a temporary file is so high that the cost of allocating a new is negligible and therefore either approach would be fine +As soon as we saw the match we were skipping over the SELECT item subtree without performing any translations leading to the inconsistency later on when the view was referencedThe fix in this patch is a bit involved since I had to make the capable of copying the translations for the common subexpressionsI have not yet run through all tests yet but the patch is available for review now +I went back and checked my case to see if any exceptions got thrown during finalization and there were none during tear down in this case +If we want semantic analysis to be easier or more intelligent it is a feature request not an error +But needs an update +The issue is updated. +And there you go +That is let FUSE HQ use a separate JRE. +Defer to all open issues to all open issues to +If you find further issues please create a new jira and link it to this one. +This should have already been fixed before the release +This issue applies to codebase prior +the problem here is that these strategies expect pairs +Otherwise Ill close it +Created an issue against for adding the rollback and list still dont think we can do anything for +Hi FreemanYoure right +readObject and writeObject methods are added in class to guide JVM on serializing this class +Closing as the patch has been applied correctly in the java branch. +Resolved some clipping issues added create patch for current sources +The spec says a TTL of equals never expire message but I suspect some users might be using as well and presently see the same behavior of message not expiring +Ive updated this filtering to exclude the dojo jarsI have made this change under revision r +Thanks a lot Todd for the reviewLooks Jenkins has some issue? last build did not run successfully +This has implication on Scan benchmarks runtime +What is the status of this issue? There seems no progress for more than one and a half year now +It has some spellingcomment fixes also +my mistake endpoint the argument to the method cant be null but endpoint can be +I was trying to integrate Sqoop but the API it exposes are tool based +I think the problems you see here are because ofIn any case I have implemented +removed the synchronized to tickle hudson +Fixed in revision +sorry reduce can put together a patch for this but not until early next week +The JIRA was initially to make the a singleton it got morphed over timeFeel free to make the call I dont mind which JIRA I am submitting the patch to +Only two bugs are actually fixed between RC and +I did this by adding options to and then the returned automaton fromI also added a couple new nocommits +thanks for taking them out +From javasamplestranslets ran java Note Im using the version of the that has BCEL regexp Jlex and javacup all rolled up into oneHeres the contents of my +I have created an implementation of my own which is just one class +Oh yeah Im looking at trunk sorry +Sorry I should have been clearer +closed all issues which got resolved in M +Nige would it be possible for you to verify this since you also know how to reproduce the issue +Fixing javac warnings +Added support for JSONP and JSONP error handling +I have fixed the warning +Trunk patch v adds +Killing firefox cross platform turns out to be one of the hardest problems in computer science +There are already some tests showing the incorrect behavior in! Thats a good example and clear +Will upload a patch +Sorry for misleading info +I would just throw instead of plain d be ok with the static factory method +I already submitted a patch for +Derby is not distributed and deployed by default +Fix javadoc +One of the things we learned was that documentation doesnt fix an unclear product +is the surefire plugin being developed by the same crew as m? or is there a better place to post issues relating to that plugin +Some documentation should be added on the Wiki to explain how to use this tool +Thanks Ben this is a good direction for the interface that I have been trying to take it in for a while +This means the original login is successful and access to HDFS is OK but when renewal time comes this exception pops in the logs and HDFS access is no more +That is why we want to let users to decide if they need that protection at all +Backported to with revision Ill let the nightly tests run through before merging this fix back to although I dont expect any problems +Can you extract this out to a separate class? Or maybe roll the remote metastore client into +Your proposal strikes me as less obvious and more opaque than just creating a new expression +I dont have karma to this project so I cant do that +My understanding is that we want to have out soon and my feeling is that this is not a blocker for given Vishals description and our experience with the system so far but it would be good to hear from Vishal +Reviewed all these issues and marking them as closed. +This allows for faster iteration on just the tools and an independent release cycle for newer tools that folks might find useful +Ralf I for myself very often have the log level set to info but still prefer to see the SQL log statements to be output +Closing stale resolved issues +The core problem is that there is pretty much no identified use case for it +we are migrating from a repository and we have inconsistencies in the workspace and i would like to fix that witha consistency there a way to determine the name of the node +bq +Leif this is the patch cleanup for trunk r please test +This issue is no longer relevant for it is fixed inChanging the affected version to. +is the first truly usable version of the plugin and should take care of this issue +If one could say in that SVN should be resolved by SVK instead of native SVN then there is no licence trouble as there is no SVK code included in the SCM plugin then +We could consider moving the random walk framework into a separate module if we think it would be useful for other projects +Thanks +Sure I will take a look +One question I have is this just the way things are or does Spring go some way to alleviate this? +Done. +Dennis Hamilton wrote I believe that the discussion on abbreviated headers is solely for selected files that are included in a release +Ill revisit in a bit when I find a way to tidy up these other issues +Im guessing no one is using that artifact +Updated patch that drops the scr plugin from the in +bulk move ll commit soon +commit in openjpa truck with and in openjpa with. +and committed +Cool thing +Any news on this one? Since Spring MVC supports the combination it would make sense to support it in as well +Only one file needed to be updatedI added the new functions +Eli the authentication feature is there in and in trunk + +Possibly what I do is that one bond for Cloustack and other bond for alien +We could also add that to the main JMX tree for example under Camel and you chose a route then currently you will have to expand each node the right click menu with these menu options would be nice +I will take it +I am sure it no longer happens. +Oh well if you want to write some go ahead and do it in testunit +Yura this issue has been fixed this morning +This file can be used to create a postgresql table that is populated with data that illustrates the issue +But for case insensitive dictionaries well merge them as beforeSounds good? Or would it be better to treat these also as duplicates and use only the last one? If so we should modify the unit test instead +OK the changes to look OK Ill commit shortly +Committed with the fix to the error message +Committed r +The extracted text therefore contains a large number of random looking characters +on this patch until there is a unit test +Directories are missing from the hash lists and therefore not considered as known deploymentsNeeds to be treatedmessagingclustersecurity +Agree + +Failing trash test is unrelated to this patch +Im reopening this ticket Linda To stop supporting the execution of Cobertura To stop supporting the import of Surefire reports But to keep on supporting import of Cobertura +I dont understand the ignoreSync doSync code that you have in the the +build fixed. +Juergen +Hey KeithI added some stuff as part of is that enough to close this guyThanks havent checked component services but this is not present for composite services +Verified in +The memory usage of is shown +Sorry for taking up your time +Version Final is released +thank santhosh +Either way the fix is probably to get a valid cert +Thank you for responding so quickly +Ive just renamed title to some another title then changed back and HF URL is updated to right one +Fixed in trunk and brunch +I was unable to reproduce this on windows we will need to hunt down a windows developer +We see some contention in this Infinispan test especially in and NAKACK +Revised the new test slightly to make it more compact by issuinga COUNT query and using to check the result +RTF features currently lack a maintainer if a patch is submitted it will be considered for a possible fix in the mean time im closing this due to lack of transition resolvedwontfix to closedwontfix +In the second part of the reproduction only the connection is important in +Len we have Julians workaround for this one +Records with will now have a metadata key with the fixed fetch interval +Changes to how we implement to support overloaded +bulk move of all unresolved issues from to + If we couldnt found a complete record ANDwe cannot read more bufferRead MAXREADSIZE its because the record is too BIGSo and drop current buffer so we can keep moving Add a test case Modify to be able to give a port of so the system will can use any free port +Closing old resolved issues +Patch applied. +Will be updating my findings after applying the suggestion + unscheduled issues to the next release +As there is just Central Community to be selected in Available software in eclipse can it be described better than just Central CommunityBecause the description of Central wont apper until next page of installation +We recommend looking at Trinidad for small server purposes and Torquebox for full EE server support. +I dont think this requires a test case or a review board item +We can go back later and refine things if necessary +I will provide a patch for this issue +Ill be happy to update the patch +Convert to junit +s default and most efficient scheduling behavior comes when locality can relaxed from node to rack etc +python c and java worksformeI had no chance to test php perl and ruby they might even have syntactic errors in it + is related to this effort but a specific case +Now it works +I guess the relevant code is the if statement starting at srcmainorgcodehausgroovyant processes some named attributes +If you have further questions please let me know +WSDL Rpc Lit parameterized types and operations names to verify xsd values +Hi ErwanAs this not been fixed at r +bulk close of resolved issues. +I am able to reproduce this issue now +The features of and this one are different +This read was not +Some comments on that patch +Couldnt this be automatedOtherwise it could happen a next time +This does not appear to be a problem in Bigtop right now but will be when the version of HDFS used is upgraded +Firewall security policies prevent Syncing the data from build machines will be possible in a week or two as are moving to globalsync ask again in weeks +Patch attached +I think this has been fixed for a long while +While some greater concision may be realized by allowing for implicit bean declarations this benefit is outweighed by the clarity of being explicit +Hi Id like to work on this +I have added a association in order to have some db operation in the gsp +Is this broken on as well +Hey rathneshI am also refering the same tutorial and also facing same errorI would like to the solution if u have solved it +It loads the addressing module and executed correctlyClient side also module is loading +You can refer to to see how to handle for nested operator +Committed patch to documentation trunk at revision +It may be that we need to revert to the for these cases and have it emulate the same behavior +committed w comment for STREAMREPLY +There was a network infrastructure update now everything is ok +Jarcec typically most Hadoop related projects tend to create the binary distribution as a superset of source distribution +leaving out of release until it can be reproduced had the same problem deleting +Thanks for the patch +proto files to mark them as stable and refer to the documentation of +So I was up all night hacking on this and I think Ive finally got something working and I like it +Could we raise the priority on this fix +resetting P open bugs to P pending further review +sure +is +How is the expected to be set on the then? I cannot see any API on which could have been invoked +distilled from discussion on IRC new database created as compaction starts writing to compaction completes rename to close fd on next write opens the database and looks for the for the highest Nthe code in alldatabases will need to need to return only one item per +Updating with the throttling options +I was jumping through the hoops to allow a partial read of the Binary and then reset it to get the whole streamWould you be amenable to committing some version of a patch that worked like that +Travis thanks for following upIf you removed all the references to it mind removing get method altogether? deadreduceAlso procedural add patch versions when you upload them otherwise it gets tricky to figure out which of the identical file names is the latestThanks now unused method actually is removed in that patch take another lookGoing forward Ill add version numbers to patches +Please review it and let me know your valuable comments +Kevin Just curious but why did you raise a JIRA for this change rather than just making it directly in svn? +Reassigning for Ted to look at as he owns that particular file +It is very surprising for a user that applying blank or nullable results on vetoing the remaining constraints and as I said I dont think it was the behaviour that was agreed on the above was the actual resolution to this issue? There was a lot of discussion as to possibilities but how was the issue solved? +I will be looking at this problem later this week +Jan Jamrich worked last week on adding richmessage to Metamer +You could then specify or override the markup on a page by page basis +Im sorry I didnt follow up on this issue so far Ive been swamped IRLI havent lost hope of being able to provide a patch yet +This patch makes the transformer dump only the non default parts of a stylesymbolizer for much more compact SLD exports useful for both uDig and the CSS module +Add the default servlet gbeandata to the module context commit the changes to trunk At revision +Fixed in +I will commit soonThis issue was caused by so indeed only appeared in + +Patch applied +The Adobe AIR SDK is one of these system dependencies +also has a number of methods which would be unnecessary if a bytebuffer were used instead +Good catch I didnt notice it. +Done for both entity and collection batches +Thanks Vinod for the reviewIncorporated all your comments +I waited till the TRUNK build complied and succeeded but did not check the SECURITY branch +I can modify the test sure +But all the other parts of console would need to be converted by hand to the new paradigm. +Im fine with leaving all the sstables live +Accelerate queries using aggregate index +However I have to get legal clearance from my employer before I can make a submission to Camel +It doesnt seem like a bug but an enhancement +I agree that propagation is more natural and expected +my solution doesnt require any of this so lets see if we can discuss it some more +I created combined patch with Pavel and Gregory changes +But the spec isnt very clear +ScottYoure right +I tried reproducing this but came out short +Needs to add Nested components sample to Metamer +Im anything but an autotools expert so Id be really happy if someone could review the changes +I see that one would rather first experiment with Giraph and run some examples to get familiar with its computational model before deployment on physical machines +Verify and under current verify under Opera reproducible +Comments about how to make it clean are welcome +Ive checked it in +Thanks. +just uploaded the mac dev jars +I have executed Wordcount job from uses isnt something we should change lightly its probably going to break user appsAt least we need a config to turn this on and it should be off by default +After reading the JMS spec we assumed that we could pass Sessions around allowing each component to create the consumers and producers it needs +If you can reproduce commitlog incompatibility between those versions please let us know +Reopening so we dont forget to apply the new and improved patch +Now accepts things that might not work on other. +This way people explicitly know its in their syntax +This one was generated with git but I was able to test apply cleanly with patch pThis one was even easier thanks to in hadoop! I was able to drop the need for separate interfaces and pass through accessors for each metricIll put up some docs on enable JMX monitoring on the wiki +bulk defer to defer to +By the way it has been fixed for password using CRYPT as well as SHA SSHA MD and SMD +Thanks for the review Nigel +I believe that I have addressed all of the comments +Yeah we need to settle on one of those classes first and then remove the duplication on a case by case basis +Ben thanks I tried a search and it seems to work fine +I changed the index from to genderdescription and that did the trick thanks +Patch applied at r please verify +and +I think this jira should be moved to HADOOP +Fixed +Evangelinathere should be an error message happening before the portion of the log attached hereCould you please send that part too? It would be useful to get all the log messages happening after you confirm the order in the POSThanks +UID is a bust as even sun admits does not work well +Ill take a final look and see what needs to be updated in the specification to close this out +I used WAS! With version I have no problemsThe thing about accessing the Servlet response from differentThreads causing a was told me yesterdayby the IBM team +So the assertion of the session in effect done in loop where each logoff results in another check which kicks off another logoffNow it has been corrected such that in the case of lost connections the connection object is properly recycled does not try to log off a invalid session +To be assigned to Yash assigned to Yash Talwar fix priority fp +How about adding the initial contribution with the varargs and post pone the iterable version to +One alternate way to consider handling this issue is to implement logic +listing the file being created in listPaths but not allowing it to be opened for reading while it is being written +INFO is the level normally logged in production FINE FINER etc +That use case doesnt work on my Galaxy Nexus + +Now those have been removed so yes thats what they are now +You can also check out the tag of the version you are using at the moment and apply the changes to and +Have you tried to apply it on a fresh exports Trunk branch on your side? I will try on my side and see +This issue not reproducible for me and Slava on CR studio +I see the same thing if I kill a collector process +Im able to reproduce the problem in a test I have added in +i did review and test the patch before committing it +You can reach me by email or mobile +I havent had much luck reproducing this I registered an in the sample and saw that its fatalError method was being called +Thank you Tim +dll into fake distribution directory works well as long as you dont try to do debugging with the exampleschanging fix version from to RC since it is an old issue and is now cluttering up the final release notes. +It has been doing for openejb release and release +Ilya please add in a little detail on what you would like to add perhaps in dev forums +Is that correct? The Beans Graph shows all beans including those imported from other Spring +Thanks Daniel +This looks like a piece of code that should be easy to write unit tests for +It seems clear that this is unworkable in cassandra core. +I think this solves the issue in a pretty clean patch should fix the +which of these are problematic when source is formatted automatically +Any chance some tests can be added to validate the component +You can discuss project scope and other theological issues in the user or dev mailing lists +And if it for any reason doesnt I would suspect that the user base being affect is very limited as thats a very old version +Calling methods on an should no longer automatically boot the engine +This is solely a BRMS book +I looked through the changes to the branch and nothing looked like it would fix this from the description +Thanks +If we use plain text files end user will easily add new language based on English language files with a text editorAnd database tables make it easy to query if there is any missing for example from to english langauge add items while chinese add items +If this was actually fixed in CR please against CR +James Carman implemented a patch to fix this +So it is just a little more than the usual JAR files that gets created +JBESBGAFP . +Last things to fix for pelt subtabs are not working color profiles do not contain all colors +Damn itShould we change streaming to use a bidirectional socket? It already gets dedicated sockets so not a huge change although I like how clean sockets are +done in +Failed test seems to be completely unrelated to distcp +Thanks for the patch Bruce! Committed to trunk. +These sorts of refactorings dont happen often and I would like to release it clean +Fixed with and +Hrm +Test passes for the tagged version +Another possibility would be to maintain a list of redirect locations in the in addition to the setWould that solve the problem for you? method is not used anywhere in the code +Closed as Cannot reproduce to Group Management add any user to a group Go back to User Management Error occurs Caused by Cannot load resource at index Caused by +Maven does manage to files across reasonably reliably it uses something called Wagon internally I believe +Is this the same issue as? provides misleadingconfusing information for indexes backing constraints +Applied patch +There are many other reasons for SSH failures not just keys that are not listed +I tried to run the the application via webstart and it failed +Note it was tested on win onlyThe interface is simpleant setupant run the builds only under CCant Drunhutant run builds and Harmony unit tests under CCant Drunallant run all modules under to of all I suggest to move current infrastructure to the run directory to store different testsuites in buildtest module tooSo expected structure isbuildtests testsuites releability vmvalidation run current looks OK +Marking this as resolved since the final subtask has been completed. +maven project for reproducing the have tested with last and I have the same problem +mk folder archive +resolve then returns null in that case for subsequent invocations which makes the method invocation code skip the invocation since it is guarded by arguments! null there +Updated with comments incorporated +If people ask again in user alias wed have to figure out exactly where that came from +that includes necessary java classes which needs to be added to srcjavaorgapachehadoopfskfs a file that needs to be in lib directory docs that describes how to use KFS with this requires a new jar which cant be included in a patch file our automated patch check system will always fail on this +Cannot guarantee compilation of AxisC in bit +This item is resolved fixed +Sounds good that will make it more consistent with the Android version +Even without the support for events begin able to use the serveResource provides a lot of benefit +fix in review board +Optional parameters will only work for post procedures which use more work would be required use instead of for optional get parameters. +Its a nightly build from Feb thAttached is the screenshot +The continuous check in the results method is causing the check of the receiver to be skipped +Committed revision Thanks Wojtek and Noble +The poms should contain the needed info +YesAaronI didnt notice the spaces within the method +We should return a for these if the user hasnt configured their twitter account just like the two post tweet urls. +Checking this in +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Closing issue out as part of JIRA cleanup +If ACCUMULOHOME is only needed by a unit test for the old classloader and we are going to get rid of the old class loader then maybe you could just drop that unit test before checking the code in +Thanks Fred +Huh? What are you talking about? Hibernate never creates with no vaues +ProjectClean and build the property file in the targetclasses directory +I just wanted to make sure it is done for +Patch looks good missing some cleanup for and Oracle DDL scripts +Correct CSS goes a long way +Checked for both touchz and ls +Will be needed for various other tasks esp +probably by included the text of the NOTICE file and linking to the local copy of the LICENSE file +means disabled the default is There are some timing issues with the indexing tests on account of this new async text extraction +I removed from the plugin since all classes are already included in But I couldnt test the modification since my cluster is down at the moment +Weve been stuck trying to get those same tests to work in for ages and at least are now aware of the design changes that are causing failures +Jitendra if goes to trunk would it be good to put this change in as well for consistency +Thanks Brian +for all official services +files are up again services are not will fix them we have not tried to upgrade geoserver to see if it goes away we are unlikely to at this point either +Xalan does work withmy examples but I use the TRAX methodThanksB +While we do not sync the transaction log to disk this seems to be reasonable enough since the only problematic case is with the machine crashingThe test creates a file waits for a few blocks to become available then stops the mini cluster without closing the file and starts it againThe length of the file cannot be smaller than before the restart if the blocks were persisted into log +Thanks Clment! +And nice catch with +similar issue with casting between long and also do not need next in +N is the number of values which are preallocated every time the sequenceidentity exhausts its current cache of values +IMO snapshots should go into a different repository than the releases +the wizards and views are now easy available under the new and view just rebuilds the tree from the model mainly there for debugging purposes +Attached a patch to branch for both wicket and projectsIve added the feature in wicket directly but it doesnt require any change in wicket so it could be in or even in wicket stuff +Waiting for to receive the event doesnt necessarily guarantee that all region servers have seen the event yetVery true its also a problem when you addremove a peer but my concern here was more centered on the client so at least if you enable a peer then read the state right after that you see it as enabled +Thanks for the patch Mike +Francis great to hear! Just pinging because we collectively are working in this area again now +The sample portlet application fixed so it will validate +My english is not accurate yetThanks for the commit. +It was very helpful. +I have a patch but it doesnt follow the thread safety comments of the existing patch +However I cant see how forcing people into a style of selecting RH configs is a benefit +Implemented this change in Rev +second PR updates ripple code first PR updates link to ripple in with JBT +Patch applied SVN revision +Id be happy to spend some time on this if I could get some direction on where I should focus +Woops the first screenshot shows that nodes are decommissioned +If they are which seems a good idea then everything will be enableddisabled in the construction of the so we would only need to support property getter methods +Initially you put the change in on Feb th so weve had weeks of soak time and this crash seems to be the last issue to overcome which your earlier change fixed +Test restored to excluded test at +Thanks Mark! +Thats how it is in Word and thats how it is in just about any HTML editor Ive ever usedI hold firm to my belief that a paragraph tag should act like a paragraph tag and display separation between another paragraph +Otherwise defer to where API changes can be done. + +Ill add the ACL docs once the namespace ACL jira is done. +Works fine +PR submitted +If this issue needs revisiting please this JIRA with additional information about the problem +and doesnt break any other image test cases +I have installed several other plugins recently and have not had any problems with any of these other plugins +Writing a testSince the original job was failing at runtime due to invalid bytearray casting I added a ee test +Helmut is this still a problem for you? If so feel free to reopen this issue. +Could you provide sample code for the component? If is itself a component it may also be stored in the session depending on its scope +Hiis there some documentation what bundles Im going to need in order to successfully deploy and use this bundleI could find that out on my own but if someone already has a list could you please be so kind and post the link here +isolation maintained in the clients connection object can be out of sync with the real isolation when in an XA transaction +Patch has been reverted in trunk will do the branch next +The least I can do for this project +Ill write some during this weeks Oakathon. +Good comments all around Uwe Ill work out a new patch! Thanks +Hey Maurice +And there behaviour exact as I described above +Oh I was thinking he had changed AI to run more oftenBTW I have changed the setting to run every minutes But I guess we had him change it back to the default hoping that would prevent the crash +Oh OK got it +tail will keep trying to read more data and will never shut it self downJon +v Made small changes to setup config of unit tests +In trunk the only bloom filter that is used by is +Do we have guidelines when using String and when +Manually tested! +Fixed on and had a look at this and wondered whether we should cache the binding to a converter directly in the feature model having to traverse the lists trying to find a converter that can handle the datatype is not very efficient +bq +Test failure got rewritten as a Unit test +Ill give it a try +That was fixed within. + Username gregwilkins + to the first three patches +from the index page is nice +releases should be backported bugs from trunk only +I stripped down my test case and I cant reproduce either +Fixed but facing problems with staging +Fix the findbug issue +If you dont think the proposed patch is right feel free to disagree I just put this proposal up as something to start discussion +We dont need to have it as a list box with AddRemove buttons here +I also tested by setting the environment variables in the respective scripts +Select Validation and disable both Seam Project Property Validator and Seam Validator manual and build checkboxes +I have patches on my local machine in various phases of flux including XHTML output for scaffolding the tbodyelement patches for scaffolding this patchScaffolding code has changed a good bit and many of my patches now require +Patch file thats working for me +Attached a patchCould someone give me contributor permission and assign to me +Unfortunately I dont quite remember what they were but I do remember that my changes fixed them +Auto Register Converters All Converters will be registered to by placing a text file at services containing a list of fully qualified class name of each Converter on its own line +Here the patch with the +Run these tests with tdvaio cygdrivedsvnjbosswsbranchessrctest ant teststdieslertdvaio cygdrivedsvnjbosswsbranchesoutput ant teststdieslertdvaio cygdrivedcvstestsuite ant DXCJ cygdrivecdevprjjbosswsbranchesoutput ant tests grep FAILED wc DXCJ cygdrivecdevprjjbossastestsuite +I committed this. +But when compiling activeio with JDK it pukes up a bunch more stuffPerhaps activeio needs to be retrotranslated so that activemq can use jdk versions of this +All even though we dont have restricted memoryThis solution does not use skip +If so the question becomes why blocks ended up with all the replicas on one rack +Rejected since no info indicate it is a client side issue +As Travis says the slowdown is caused by +I recognize that this is an issue however does not really depend on this issue +No response assuming ok. +moving to Beta if nothing new heard then we stick to what we got now +In Collex we do on a single object not within search results +docs added with svn revision +All necessary resourcesPUetc are injected +Bulk close resolved issues +bq +Just committed this +applied thanks +See the blue line is ok the red one is incorrect +Heres what we have currently +I have a couple questionsIf the bug is in Lucene shouldnt we write a test at the Lucene levelWhat exactly is the bug? That when you subclass it doesnt return that subclass from reopen? If this is the desired behavior isnt it up to the subclass to override reopenAlso you say the required lucene fix is randomly in the patch but also that you will post another patch showing the lucene fix I dont see it in the patch so I assume its coming but the only change I see is making some Lucene constructors public we shouldnt likely do that just for this Solr test I think +HDFS portion of proposed change. +Vm deployment using vmware template should succeedWorks fine. +Also if you have time you could take tlopez out of the +removing fix version +Committed revision +Committed revision Thanks so much for the review Stack +What is the concernMost tasks adding other functionality is marked for that branch +Bug has been marked as a duplicate of this bug +No there is no need to install Maven + PrashantThis issue is still reproducible + Kalleriplease read the javadoc to Externalizable +Obviously this is just a workaround and it doesnt work well for continuous integration systems that need to run everything automagically but it has helped me a lot +There is no single join operator in local mode +could you please close the issueThanksRegardsMukesh Reddy L +Artifacts can be checked the existence in target folder etc +Ralf can you please add an item to the committer forum and letsget a vote up there +Passed all tests +I just signed the ICLA and will fax it tomorrow +Although all of the shims do get compiled against their respective versions of hadoop the profile is needed by some classes such as that depend on a version of hadoop for the Configuration class +Jun +Also eap jars are signed which means users enabling security manager could run into challenges with mixed set of jars +I was under the impression that if a code was missing youd just get the code +The emf framework UDIG is based on has good support for persistence +Thanks for rewarding my laziness +stack This patch is already part of +Now that we have nested layouts we need to revist the way we are doing thingsRegards am currently working on a possible solution for this right now +Closing old issues +See example posted by Rossen above +Comparison indicated that this was fixed in the mainstream wfs module but not the fork +After I create a new user with the User Admin page then click on the User Admin menu option I see the new user in the list there +A little readme containing make instructions for the patch +The test failure is not related to this patch +Patch applied +Will post results when I do +Can this be done for +What if we support inclusion of snapshot paths without any flags? I mean alwaysAny thoughts +This is the lucene only parts of in this can be backported to when is ifwhen this cache moves to we will still be able to cache arbitrary stuff on the reader +In general there shouldnt be any difference between custom Hibernate sessiontransaction handling and Springs Hibernate for the response +My bad I didnt check that +recently i downloaded the fresh new release which still has this bugif you could tell me how you want this patch sentfiledcomitted i am an absolute fresher regarding filing bugs against opensource software projects +This new patch incorporates his algorithm and a couple of minor changes +Can you describe overall what is the problem that current drop is encountering not just what you have already posted but something higher level +Here is another patch incorporating Owens fba Validates the no +Adding get and getFull methods that take foreground and background caches +It is listed in the stylesheets stylesheet +I finally can see the problem it could be good to have a logic here but it should be solved at spec level +ran ant test tests passed sucessfully +applied patch Henry Story +The original patch puts all messages as GLOBALMESSAGEAside from these I made by best guess for class names attribute names and parameter names that the committers may want to change +Are you saying we need to strip the seam from the examples context path? Or are you saying that the context path of the example should be seam the name of the folderOne way or another we need to align +This error is known to occur when using Hector and also if a monitoring agent connects to the thrift port to see if it is alive + I dont think we can release these separately until we adopt a stable RPC format with compatibility rules +Attached patch adds the comment suggested by Eli above +Stas +Released as part of. +Right but the point is to change the special case of the legend graphc being accessed as a GLG request and to put the request into the query string rather than the request path +I need to think about this more. +Tested this out should be ready +The previous patch was not generated correctly it did not has +I understand the patch a bitSo previously the minorVersion was never persisted +I noticed this test disipeared from trunk so I assume someone else is having problems with it not testing as well +JBI found there is jruby in the servicemix bundle +You should remove the snapshot but you cant release with snapshot plugins or it wont be reproducible +thanks for pointing me to +Your compiler discussion is a bit beyond me but with Groovys promise of Java integration I certainly think the code should work +Closing all resolved tickets from or older +FYI looking at the backtrace I saw and files +Must have been unlucky on my first test runLooks like the previous patch is the way to go +Committing. + excludes test until test excluded at r. +Enjoy +Add the pull think we can consider the issue closed its fixed thanks Paul. +Would love some concrete feedback on whether or not this deserves to be in Pig and how it should be or could be differentI think that more broadly speaking it would be huge to make type support a lot easier to add but thats a separate JIRA +Brian confirmed is not supposed to be public +Running a set of tests is a natural concept of the build processId say that Maven core should be aware of this concept and should be given the information about how the tests ranI dont know how its implemented but Id imagine some broader interface for Surefire with methods to get the results info +This throws an Exception only if there is a fatal error +I failed to login Phabricator because of some weird authentication problem and still working on that +Marking derbybackportreject as this issue required a release note and restricted SQL syntax +is this a duplicate of +This is a rare enough problem that performance will not be a concern so I think the simple solution to this particular problem is the one that is described +Odd considering I was getting the error after only having restarted a few days earlier +Claus just an FYI +We ever did a series of stress test for write throughput and the maximum we ever got is about using YCSB clientWithout patchWrite Threads Write Rows Consume Time sAvg TPS With patchWrite Threads Write Rows Consume Time sAvg TPS +Please on stage closing +for we need to fix this then right? Actually we did not focus on this much as offheapcache was an experimental feature +Fixed +Will this issue be prioritized? What is the procedure for these kinds of requests? I do not want to modify source code and then have the same issue when i upgrade to a newer version + Go ahead! This should make the patch in a lot easier to read +The scenario we tried was to disable and enable a table that had regionsWe did not see much improvement +Updated alert to fadeIn with a fixed bottom position +Because this is my original problem and its good to know that it is reproducible +Only fixed in in JBDS B +Proposed patch + +I have used eclipse to create the me know if you have a problem with the but with typo fixed in errror message must be accessed within a transaction +The CPU on the namenode increased from its normal of to +The updated jar is now attached with the updated source let me know if you need the rest of the jars but this is the only one that contains any changes +This should be addressed in the next JIRA +Re +But however we turn this the fact remains that there are different scopes needed +patch for +This will break the API in the XML do we want to go down that pathMost likely as the old ticket had votes watchers so some interrest from the community +The reason I didnt remove this was because its part of the public API and I thought policy was that we didnt modify it without going through a deprecation cycle first +Ashutosh thanks for reviewingWill add a Jira to remove the having behavior +Neither or String itself is cloneable so were not doing this because its needed +Are you sure that this is the case in RC? My local copy of it certainly has the correct version of in both places +The intent was to expose connection count on so that we dont need to visit IP addresses are listed under Quorum Server Statistics on which arent easy to interpret especially if therere hundreds of clients +Current embedded version is Android Lint +Please reinstate the fix version if the intent is to fix in +on adding null checks before we stopinterrupt stuff +ver was broken +Completed and Release Notes updated +A solution has been found for the problem +Thanks to Michael Schmalle as wellBTW I have fixed the with all the missing semicolons and checked it into SVN +Unfortunately I forgot that I had also patched the jline librarys broken method without this patch jline is unable to load its own dll successfully when subclassed +Modified the issue with the changes in the code now the calculateType method is in +This could be done any number of ways xml file annotations based on requested resource path etc +problem was due to bug in UIMA AS code using delta serialization properly +Indeed Ive done refactorings on the dependency resolution algorithm and I think it should fix the problem +See the attached project which demonstrates a working test. +Im closing this issue because all the work for is is just a subtask remaining to backport some fixes to. +Sorry Mahadev added missing file +Karam can you please verify if this issue still persists +rwinxp smoke until LOS +fixed inr of branchr of issue +This looks nice and useful +What is the likely date for MINA Multicast support? +Moved to core since the sequence in the output from tools is driven by whatever sequence that is returned from the core +It could be s for instance +Thanks Im able to reproduce the OOME nowIm wondering could this be because we have made activations dependent on the current role and the dependency table keeps references to activations even after theyre no longer in use +Stable was broken too +Jozef are there any remaining file differences between the jar and upstreamRajesh this jar is from our most recent richfaces build +Anurag please reassign yourself if you want to do further work on this issue +This prevent dead code eliminator from deleting such instructions. +was implemented that way so that in case it is called in the case in the reduce it would produce the right result +Help pls +usually we dont create a new component until is out or about to be out of sandbox but i will create one per your request. +It illustrates tradeoffs when implementing built in functions +Horia thank you for investing this +This looks like a duplicate of +Please reopen if this is not resolved for you +has been marked as a duplicate of this bug +Applied part of to module in SVN Revs and +Still no ideas? +This patch should work as soon as jclouds are available from Maven Central +Just looking at the JCR API it appears that the size information is available on the child node iterator but looking at the Jackrabbit internals the size is actually on the parent node which includes the list of child node entries +Added Dima as watcher +None of the covered sub issues was resolved thus rescheduling for changed so far rescheduling the check for the next version +ive played around with some strings and it seems to work ok on ie and ie might have been an ie thing +committed to trunk +Ive moved this issue over to the styles because I tend to agree with Geoffrey +I believe this is possible today for a limited by specifying on the font element within the file +I reverted to using because get method uses +Patch integrated to trunk. +Mina versionCommitted revision +Thanks milinda. +We ought also have it in the serveralldeploy +Feel free to ping me if you have any more questions or if theres anything else I should read +created and committed to jira +Thanks BJ +Sorry for taking so long but I started changing the API and it broke a lot of code which took forever to fix +Committed revision + +The exception logged was fixed in +Now that is modified to use new api you can use new Constructors +But they will be here +Hudson results match results on my local machine +done a while back now admin password is no longer in +something likeEnhance list to add support for integral types both for equality and +I just committed this + fixed this issue in AS +Status please +For these plugins the suggested workaround is not really convenient as they are supposed to execute in a phase prior to the install phase. +defer issues to +Tested the fix with Continuum and the +Fixed in commit affects version as this affected that released version. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Then an IPC client caches all connections to a server +This is a proposal for discussion and refinement +for fixing this in +I finished the cleanup got rid of the obsolete stuff thought I got it all before mea typoI took the extra debug code out of verify +LDAP traffic can be sniffed +No more jars in those folders +It is the reason we got segmentation fault when destructing the +No no modifications just usage as is +when committing can you also nuke +And at no time should we accept a change that permanently reduces performance without providing a credible alternative unless there is no such option +The bug is invalidipgateway and ipcidr do show in Network Detail View +For your suggestion please open a separate improvement request for thisregardsMartin +This seems like a Groovy andor bug not a Grails bug +Note not sure how it compares but Lucene has a TST implementation in it already +I dont think you will have to maintain compatibility prior to android as they are mostly all gone at this point +I will attacha the right hand side you see two fields containing the text Oben and Rechts +No we just closed the connection if in this code path +Plus some other changes copied on good to me JacquesHans the formats are currently hardcoded in but that could easily be changed if someone so desired +I dont think this is an issue +The block identifies a storage block pool and a namespace id identifies a namespace +I think there should be a way of getting the stack traces when wanted but currently there isnt +This issue was migrated to +The target system has a variety of diffs to the build server +Could you try the latest code +Needed to do additional refactoring after realization that this flag is only applicable on IMAP so it doesnt make sense to keep it in the base class +The fix may provide useful debug information if this code ever fails so if someone wants to rework the fix so that it works with too feel free to reopen the issue and attach a patch +Test the patch at single running cluster +Implementations are encouraged to mark methods as synchronized to protect themselves from broken clients +VerifiedVM started successfully after a Disk Snapshot revertClosing. +Thank you for your improvement +The promised newly added a parameter to xmlbeans Ewdc with this option I tried and it codegen properlywhat you have to do is to replace the xmlbeans code with the code you generated with the scomp commandplease see whether it works or not +for fixing it the Gizzly way and write another JIRA issue to keep track of JDK fix and roll back this fix once that is think we wont have time to fix it for +Since the ext +Let me know if you prefer to fix them in another patch or whether you prefer that I commit this patch as is and we improve it the copyright year should be I think that we should use Jox in order to be consistent with the other containers +If it is then do another pass and conditionally convert the infinities to NULLThis avoids an ifelse check in the inner loop for the common case +And make sure our tests are up to date +If you dont then thats fine too +I think this issue has solved. +The patch looks good I did a quick check and before the patch there were local sends after the patch there are indeed many local sends that bypass the comm layer good jobI also agree that in case of a connection loss we should kill the node +I think that this mismatch of where classes are loaded causes the I update the package import section from the manifest in to include the problem no longer occurs +WTP creates them +Currently the sql runs successfully and the behaviour is same in local and global transaction +Bulk close for release + +And yes youre right that the core algorithm between and is the same +bq +But yes this seems harmless anyway commited in ver +This is a possible fix for the problem at least for the use case of the Xml formatter +I have assigned this to myself and since I have already identified the code that needs to be changed will post a patch shortly +Closing issues resolved in released versions +Bulk update to fix version +We might go ahead and do it on one machine +Thanks a lot Ivan for your timeIve uploaded new patch by incorporating Ivans point +Also attribute type dropdowns are not refreshed +I think I have fixed this one the change is committed +However the inline example seems to have been fixed by r so Im resolving this as Fixed +This bug is introduced by +david +Is that correctDoesnt the current have the same shortcomings anyway +In the eventually a response is generated +Both and have subclasses that implements these methods +Fixed processResponse to be inside a clause +Committed to tunk with revision r and merged it to the branch with revision I ran the management tests without failuresThe a revision of the patch is equal to the original patch except for removal of two occurrences of trailing whitespaceThanks for the patch John and thanks to Thomas for reviewing it +I wonder if we can go ahead and close this out as a dup? Unless I hear back +Its out of date now. +Reopening to address this +This is good enough to commit now +Error occurs when GUI used as wellNote Also have same result when using default port from SMX Kernel +goes in xercescutil +All appears to be working as designed +I had my gripes with the way credentialsare handled while rewriting the authentication framework for but there havealways been more pressing issues till nowOleg +side note about it it is important to be able to sign out behind a load out is one of the basic functions +Hi AileenI cant reproduce this +The failure is still occuring on EDS VQT queries in +The dedup remover would allow to do one or both of these operations so that people can deactivate what they dont needWere not likely to have the new deduplication any time soon anyway so am definitely OK for adding the remover in provided as you said that is has been tested and we create a new issue to track the progress of solrclean on the trunk? Id like to release soon and this issue will look open until we do it on trunk which might take some! +Due to recent and sustained spam attacks some with virus laden attachments then this ability has been turned off for the whole wikiWell let you know when we can do something about it +kept a inside instead of an IO +TestedCould you just replace the property description in the Settings page byLocale to be used when generating violation messages +Resolved by. +Lets put all important progress in HEAD +I see that his has now been pushed to general backlog +We just need to create a new release to get it incorporated into AS +If you run some tests with then it means that it has some value for the project and in this case it would surely be better to be able to run all tests +Would you please check this inOr are there any objections for including this into the official Click +We should throw NPEs where appropriate +if true phpunit will ignore any present file for launching unit testsif not set or false phpunit will use any existing file in the running directory +Closing all old Resolved issues +Dyres patch revision Sending javaengineorgapachederbyimpldbSending javaengineorgapachederbyimplservicesmonitorSending javaengineorgapachederbyimplsqlconnSending javaengineorgapachederbyimplsqlexecuteSending javaengineorgapachederbyimplstorerawSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawlogSending javaengineorgapachederbyimplstorerawlogSending javatestingorgapachederbyTestingfunctionTestsmasterTransmitting file dataCommitted revision +No its the same infra as for CXF to use the CMSs buildscripts to manage their site but not the CMS proper +Removed and +So this is not actually bug it is problem in version mismatch. +Uploaded a new patch that updates the comments and the test case +I never tried that What is the exception in either or other log files such as catatlina localhost +POP specifications does not impose anything about the order messages are publishedSo I change this to improvement and minor because this is not a bug +which wsdl? dims +Let me know if you think otherwise +Marking this as a blocker since we would have lost half our hdfs only for our having a backup +It doesnt appear that you are using the failover protocol +Within the scope of this issue the error message we get is very appropriate showing the primary cause of failure +For me at least +This patch sets the different port numbers for CI testing in at also seems to me this is a configuration issue and usually we solve configuration issues with patches +If it should be included please mark as a blocker +Fixed as suggested +Does a clean fix this +If you change that you are free to delete this comment +Committed. +This shouldnt happen anymore +environment is young and has some bugs but is relatively stable make an epic refactoring can be worse than the actual scenario +Verified working +Well open a new issue for the RAT analysis. +We just need to set to both the value generated by the managed service factory and the value from the features configuration +All unit tests passed on my local machine +Verified on. +patch looks good +How do I differentiate a refresh request from a regular http request? Http headers +For example if the materialization window is mins coord frequency is mins coord start time is and coord end time is materialisation frequency is mins then the last action is which gets created before +Since ruby is now offered as a gem I dont think we should include it as it has been rather we should either switch to creating a deb or just let gem pick the version up from +I need to use cfimap and cfpop with two different servers +A patch to add tests for and happened here is the patch didnt result in the avn addsvn add srctestjavaorgapacheactivemqconsole +So will it be possible to check in the CAS Editor documentation to the CAS Editor project and build an own documentation based on our docbook build infrastructure? replace the existing file with this one in the projectThe images are still the same just place them at the right location +Thanks for the fix Samar +reopening to take in Williams patch +Some project will group stuff in a directory if there is a specific grouping to the modules +Lets remember though that not all apps are but I think most are so designing for the most is better than making the other few more performing +Donald I agree that your suggestion would work just as well +As per your suggestion I could discuss it on mailservice +This forces some of the old behavior of deferring the deletes until later +Please check that the patch is applied as expected +The program died as it did before +That change has nothing to do with this JIRA issue +I should be pushing this change to a build snapshot today then well have a release ASAP to resolve this +I found a way to display this data that may point to a fixAttempt to open layer youll get the x icon over layer and NPE in status barOpen the Search view Find the connection for this layer connection ResetLayer should automatically show upIm using uDig and and postgis +not fixed +In any case if its just a weird local thing of my machine then I suppose I dont really care Ill just never run the +Half of the deployment service tests seem to fail because of the move from jbossmq to JBM since they are using some mq mbeans as test targest +This feature is done +Though Im still interested in this issueIve barely looked at this in months so if someone else is interested it might be better they take it overAt the least Id appreciate comments from anyone who uses the thrift serverAttached is what I was thinking for puts gets and deletesIll add a few iterations over the next week +Aliaksey make deployable works but this it doesnt change the fact that imported project isnt as prepared as original and there is certain loss of project related information during import +code reviewed +There is some way you can tell ivy to resolve a set of dependencies and fail if there are conflicts which could be used to check for version compatibilities +Passed all tests locally +The above mentioned java warnings are unrelated and seems to belong to a totally different class +Also I just ran into this +Was that intentional +Thanks KevinPatch applied to BEANS module at repo revision rPlease check it was applied as you expected +close it since the testsuite passed finally. +I believe that rejecting the form is a good option for now I cant see any cases where it would be a good idea to try to handle such a malformed form +These annotations are obsolete not used and not implemented +Fixed in revision +Please let me know if Im missing something +bq +Thanks to you +Yes there will be a that doesnt write Avro but happens to require the Avro libraries +Yes with channels throwing exceptions in addition to returning truefalse it could become a bit there seems to be a similar issue with and AggregatorResequencer the methods sending replies dont return a value +? strings dont meet spec +Ramesh having a disable flag would be a low priority enhancement +A thread thats being used to execute prefetch isnt doing anything that an additional listener thread wouldnt also do as far as I understand it +r +For example you have a Quad core cluster and a Dual core cluster +We should re run it through hudson just to make sure that the units tests pass +Versioning was dropped from. +Will improve quality of logging as part of another issue. +Len still wants us to pare down some of the versions the tests are run against +Will this be addresed in the next release? If not what is the road map to address this +We changed the logging so it might be illustrating a deployment issue on your end +was introduced with so linking this issue to that one +BrettIm planing to commit the code which I finished today you can send me the new patches and I will merge them into the trunkWe could keep on enhancing the code in the trunkWillem +How long is it taking for that test case to run for you? I copied it into my environment the way that you have it and it looks like the method thread is crashing bring the whole thing down but reporting that it passed +is great primefaces is great roo is great spring security is greatWhen it will by all in one? In roo +Starting with EAP this documentation has included several guides +By setting locality relaxation to false youre essentially saying I want my container on these specific nodes and racks +Since the IBM docs are the only source of information on these propertiesThats not true the code describes exactly how they work +Just asking to know your ideas before removing them +thanks. +Removed the input trigger for the nested html element +After discovering and fixing this performance improved by +Managed to recreate it +Server Startup Command for running test with Dtestusrjavabinjava cp homejoeJBPAPPERbuildoutputbinusrjavalib Xmsm Xmxm XXm XX true true homejoeJBPAPPERbuildoutputlibendorsed true c all b Message Counter access needed to be synchronized +Attaching patch that fixes the two errors Richard pointed out +committed to trunk +Until is in place we probably wont have automated unit tests in HODAlso same patch should apply for the Hadoop branch as well +Some examples vs vs vs maybe some the metrics framework is first checked into x branch and then some API evolved for good reasons in trunk and then which was explicitly marked as incompatible changes +please +Actually there is a reason UNIX domain socket paths cant be longer than like bytes in total which would cause failures for a lot of people who have long test build directory paths including Jenkins if I remember correctly +Theyhave a dependency issue that is temporarily preventing them from updatingto good versions of YARN like but that will be one ofthe first things I try to fix as we go along + obvious incorporated Chris comments +thanks +for outbound connections it seems that the serverURL can be overridden +Or declare multiple messages component with different is a support issue please discuss on the forum. +Is there any workaround possible now +This is not possible on manymachines where you need admin rights for write access to CProgram files +tnx for date time string produced by function has a different pattern and is already supported by Railo and ACFthis are the pattern for locale supported now by function parseEEE MMM dd HHmmss z yyyyMMM dd yyyy Hmmss aMMMM dd yyyy HHmmss a zzzMMMM d yyyy HHmmssMMMM d yyyy HHmmEEE MMM dd yyyy HHmmssEEEE MMMM dd yyyy hmmss a zzz HHmm a HHmm aEE dd MMM yyyy HHmmss zzEEE d MMM yyyy HHmmss zzMMMM dd yyyy hhmmssyyyyMMdd hhmmss zzdd MMM yyyy hhmmss zz +later +If you are trying to make a release just postpone this one until a future release +META +This is minor +please diff against and trunk needs a patch for trunk and branchwont be part of the next julien ive been quite busy +It was released with the fix in +Its needed to be dropped into Tomcat so the rtfilter logging works +Running and tests simultaneously I get the same Transaction Timeout error for on both sidesI certainly ran with and maybe because tests are far less numerous I did not cross this issue +Yeah i guess youre right +Since there is no existing blog portlet available lets create a static portlet which simply resembles a blog as a placeholder or links to an external blog +Decided to go with Swagger and look into customizations required for better HATEOAS support as our needs grow +I was supposed to get a screencast from Max but I never got around to scheduling a meeting since we were both busy +Yep +Issue from the last comment seems to be regression caused by last fixes +This issue will be resolved as part of +Tested onTo reproduce on the split the table come back on the table details view and refresh again and again +code used for ldl detection is copied from hadoop common cmake +I think this could be done using the compat modules +Kurt is correct the aop version allows ESB to workaround some issues in that version of intent is to remove as many if not all of these workarounds with the upcoming release of juddi at which point this should be called we still need the aop version then we will also have to add the aop jarsconfiguration to the documentation +What is the correct process for updating the list of tests +bulk close of resolved issues as part of JIRA housekeeping for release. +These counters will aid diagnostics significantly +A rather complete and stable concurrent GC is done which can achieve reasonable performance and rather low pause time mostly. +Next time Ill feel freer to do it if its necessary. +Not sure exactlyThe retries are enabled by default unless the content is a stream iircMaybe we should close this one and open new ones when if needed +This enables attachments to be delivered outside of the SOAP in revision . +Patch checked in at revision +We still have a problem if there are ejbs A B and C and B sends a handle for A to C C has no infomation about the CSS settings it should use to contact A +Incorporate test into if its not already there +I left sourcespackages in place in case the plan there is to update the packaging vs from scratch +Moved samples to correct exception is the petclinic +Its not just about conflict resolution but validation of application imposed constraintsYes I included these in my loose definition +If theres more work needed open a new bug. +Once the module is fully deployed the +applying those to fresh trunk works fine for me +and it worked with maven +I did do the tests but very unorthodox +The patch is more complex than it needs to be to fix the problem +That said forking is quite different than using a snapshot build +So either I do not understand how worksthere is bug or we are missing some info from Remko +patch to add optional the compiler works fine outside an osgi environment thus without the prefs the prefs might be required in an osgi environment +Will it be able to handle deploying a new version of an existing vdb +Would be cool to have out in may +The change is logically no longer in +Wait approval from be verified in gatein master and reopened if necessary +Removes the deprecated function and actually works +packages +That way when a critical bug fix needs to be built and tested the developer or build server is not waiting forever for the local repo sweep to finish +Maintaining our own forum needs more of our time then relying on nabbles service +verified in JBT +Alan do you know about this +Dave may have a better chance at this one because hes better with javascript Ill see if he can take a look at it +Len see for more comments +This patch and are two independent patch thus not influencing each other +This Jira doesnt seem to let me change the priority so I resolved it +it works if I specify classes but it does not work if I specify packages +Can someone commit thisRavi just before commit can you run it through Hudson once more so we are sure? Thanks +It should treat it as it would any other informal parameter +reschedule to future was never much content here but what little there was pointed to the problem of the incubator web site versus schema +AdamDid this fix make it in envers? Im still seeing this issue in with HibernateThanksyes it did +Han perhaps the patch didnt apply well on your computer +we might do something more about it in client? +Removed automatic loading of from the constructor +Why would we want to throttle bulk load separate from other streaming in the first place +Issue is fixed +We have some documents with manipulated and the parser will run into the that break the whole parsing process +Will be updated soon +Likely to be logged against CP but was fixed in IR +Please read it over +NET instead of +committed to and trunk +Fix on revision Fix on trunk revision +Instead the test serialized and wrote individual elements of the protocol request +Last comment was in +The were intended to be removed from for but may do that in this release as well +MartinIve testedIt worksBut i noticed one thing that this parent node is expanded by default after adding a child +Reopening until I back out comparators +Ive seen this after upgrading from to +I am curious why if it was working in it only is fixed in? I think this counts as a regression for Paul I think Youre right +It is currently in Resolved status waiting for Mirek to QA verify +first cut of patch for initial testing purpose +It seems we have a workaround for now maybe we can defer +Ill commit it later today +It loses a great deal of the warmth that is a key hallmark of the Hibernate color palette +changed for HQL nested news support +There you can define the and configuration that you want to use for all your actions +Attached is a patch that address port forwarding in both waysHavent done much testing though +I gave up because Ive noticed that for each bundle that uses Logj the plugin manager is invoked which then parses all plugins +Attaching simple patch to resolve a file decompressor on hadoop fs text before continuing with prior behavior +Issue addressed in patch provided to +Committed to +Yes Thats the behavior as per +Will commit it shortly +Due to the GA slip thiese issues have been fixedThanks Miroslav +OK to me apart from minor nitsThe copyright on the new interface file says which is impossible +contrib is needed till it evolves into something sh based that can work across the board +the bundle will be recreated with the details of each of its resources and it will use the ones from the last resource referenceIve added a special to be able to decide when to preserve the details and when not +prj files on shapefiles that are causing warnings so we need to add +AS we can support the exchange of the conversation states using different protocols such as and we probably need to define an extension point here to allow the pluggability. +Ill add that next +So here is the smallest application that we can offer you to help in debugging the case +Every hive session has a current cluster and the read happens from the current cluster +Itd might be better to use the hostport in the request context +Applied a fixed version of the patch in r and +this bug was stupid +Attaching a patch that fixes the issue +Great work but Todd is dead on we need to look at the few hundred but small chunk read scenario +It was not reproducible if properly configured +This shouldnt be as async doesnt have any synchronization +Those errors doesnt affect the indexing of Pdf files +I dont think it is an intrinsic part of the binding that must be provided +Is anyone aware of any UI tests that I could change for this +This way we should have backward compatibility +Hi GavinYep +So as serious as this is the code to deal with this situation is ehm less than perfect +bulk defer minor issues to +Mathias I am assigning this to you as you originally posted thatCare to look at it nowYou would need to make sure though that we dont loose the completelythanks log reduced and are logged by the servlet container now. +Paul do you want a fix for this +Doug thanks for the comment +at least not a CXF yeah you are right I should report it there +The proposal considers more than one HRS via ZK working in a fault tolerant way to forward edits sent by all the other HRS on to the peer cluster +New patch addresses comments +Also running into this +Were currently baselining on JDK which is why we havent yet added this support +Dont throw convert it to +Looks patch only changes javadoc +JAR fileAs it stands since the +Thanks for contributing +will test and push it in +Cancelling patch stale +What I did to solve this was compile this from source and then youll get a jar with all the dependencies included +thanks +Ive not run with this just a temporary suite and the test itself +Heres a patch for this just touched a couple of XML files to fix the classpaths and cause to not show up in the distribution +Will raise separate for a couple of minor points left out. +I have such implementation in C based client appAlso Hadoop is describing ring prior to job execution inWill problem disappear after full upgrade to? Now I have call describering only on my old nodes but after full upgrade if it wont work I will have a big problem +The of the map task sees that the file already exists and skips itMaybe distcp should be copy a file into a temporary filename into the destination folder and then when the entire copy is successful it should rename it to the real filename +I committed this +A duplicate will be suppressed +thank you for pointing this out +if connection is not in a connected state he is throwing exception +Actually the default is indeed localhost and you can omit targetThis is because the default for http source is also to listen on Do we want that removed +Rationale for using a different DTD URL Its better if the version number is included in the file name and not only in the directory +Yes it is right +recreated hurrah +Only two bugs are actually fixed between RC and +Effectively an empty is the same as a null reference +I applied the suggested fix in a slight variation +You can add a thus changing order leading to unexpected failures +Moving missing part eclipse projects to Beta +This is most likely best implemented as an extension to the OPEN and OPENMULTIPLE modes +Are you able to attach an example GSP where this is happening? Ive been unable to reproduce the problem is the in layout directory +This prevents end users to explicit configure a custom executor service on the model +htaccess configured for confluenzaDNS altered but waiting for propagation +so might takeanother day or two +Should some issue exists we will track in a different JIRA +I generalized the Class to reuse the logic in the! please dont use apply insteadcheers! +Though Im still puzzled by the OPTIONS call not returning what I expect Im going to go ahead and bounce it to so we can further define whats going on in the next release +Peter Youre right! All zeroes on the end of my largest bloomfilter fileThe affected CF has billions of keys and the sstable in question is GB so Im guessing it is likely over million +The problem involved an index to the list of parameters not being incremented in specific cases +Patch applied and fixed a NPE when no request logger is used +Well need to integrate the RAM based indexer into to carry over the deletes to the ram index while its copied todisk +thrift file into the generated +I ask again are you sure that its the submit event that doesnt get fired or is it the selected event? If the latter is the case this is a duplicate of +I think the error is caught and handled by Spring and report as Spring does it by standardDo you have any suggestions for a more concise exception +reverted for now +Attached a rudimentary fixes to allowing targets to be deleted from the client API as well as from theThe patch allows you to delete targets in the same manner as you would delete distributionsfeaturesHowever theres one catch as soon as a target has an audit log it will reappear when the audit logs are synchronized +Every job that was pushed on to the machine failed after a bit with this error +You may post it to the developers mailing list if youlike and Ill see that it gets incorporated into SSL guideOleg +Its true that stronger validations should be applied to the Struts config andthese have been slated for the timeframe +In other words I would expect a plain Portlet impl running on Liferay to show the same behavior +Fixed. +We need a new release of webconsole before I can commit and release that plugin tooAnyway that plugin shows the component only in bundles not in services +These two properties are combined into UNICASTTCPUNICASTUDP when the cluster node is added the property is automatically updated to contain the primary servers cluster information however the primary server needs to be manually updated with the newly added server information in its file. +nevermind +Ive updated this to work with Solr the version were using locally +Yes sorry the ordering is the same as in maven itself I didnt mean to imply otherwise +Pls contact me when the Arquillian project is pushed live so that I can add it to the matrix page +Can you try to reproduce +in AM this will help expose them the soonerWhen we return out of zkassign what will the caller do? Stop processing the assignIs this a bug fix? SERVERNAMEA SERVERNAMEBPatch looks good to me otherwise +AndI believe your blob is bigger than mbLooking at test case I see there are no indexes +I had a long look at this and added some integration tests but in the end I think it will be too much work for +Or perhaps it was just global XA transactions that werent affected by it +when I delete the file for the artifact in question it will get resolved locally and the build will fail for the next dependency as described above +If no HADOOPHOME defined Pig will link hadoop jars inside PIGHOMElib +You sure it was +and should be ready to go +Id love to deploy a fix to our system soon to solve that if possible +Exception appears only in Error log +This required a change to the metadataprocessor so named set methods will only work for clientsservers +You have an index name set on at least one CF +Update the patch to the latest revision This is is currently a in only one Ill add the Transport out capability in the next patch +Certain activities like compression seem like trivial things and perform very well when you are on a recent CPU architecture +However against my better judgement it was decided that it would not be a good idea to change the interface of the bpel server +Yup that solved the problem +While a fetch is running an updategenerate cycle must be able to complete +There are quite a number of things that will go wrong when they are notTo some degree this is logical since we need a classpath in order to be able to resolve dependencies for your script +Thanks Lincoln +Ok this one is created and tested on a Linux box +I looked at the source and it seems that the path is always last in the command string and its therefore safe to concatenate all items starting with the first thats not prepended with a dash +I also added the for the IMPORTREADY queueIt shows the correct queue size of but the message ist still stuck in the queue +I think we need a mention of in the release notes +Clebert is the fix for this issue in the tag +Woden has been modified so that for namespace only schema imports it will attempt to resolve the imported namespace url to a schema using the url resolver framework to check first for a catalog mapping for this url to an alternative path before attempting the actual url specified in the namespace attribute +is failing in trunk without my patch also +Im happy for these ignoreCase methods to be added +Oh used for coprocessorSeems this be reopened +Attached contains output with verboseem s normal situation JET is times faster then OPTOther reasons of slow compilation time Debug build turns a lot of additional and complex checking routines in OPT on +Just link to the space directly instead of the exported pages +Ok I did the Ant changes on and once I get Maven working again on my environment I will finish this issue +works for in JBT nightly and EAP ER +Even if you go the padding way you can add operations when you export processed files outside of Hadoop to get rid of things that would make the file inconsistent +Same is true for in labels in never mind +Yes +This is a good way to solve the problem allowing future expansions as well +Apache Directory Studio has now been releasedLets clean Jira and close this issue. +Posted using the under CVS and labeled as updated for +Im not sure if this is being done correctly but Ive provided a patch that seems to do the right thing here +No wondering if you plan to integrate the proposed ignorecasing attribute as well +Updated the summary and the description to better describe the use case driving this JIRAIve closed as wont fix as it is a workaround of the problem this JIRA is trying to addressI dont think there is a need for an umbrella JIRA as this is the only change we need +Done ages ago +Im curios what motivates this change +Defer getParent until we have a demanding use case +Slightly modified patch for +Changing version to trunk enhancements are not applied to releases +I have not tried to do a test setup +distcp is working fine +patch attached to deal with cfm directly instead of looking up by id and create a dummy cfm for cache saving instead of leaving null +Should be done by end of this week. +Committed revision +You can use it but you must understand the consequences +Apologies I think I was barking up the wrong tree here +commit only after +well +Any advice on how we can debug in such a situation +But in just ignored the but rendered the other symbolizers crashes with NPE +Hi MarkI carefully did all the changes you posted +Thats the whole point rightand we cant just try and swallow the exception we need to do a check that will not cause ONFE +I can create a but I know of no way to alter the references that Spring has to the original request object which is used after I return a from my Controller to format my response body. +batch transition to closed remaining resolved bugs +Needs more work on UI though +Suraj Perhaps you mean loaded more than once? requiring more than once should never cause it to the is an interesting case can you try to figure out which file is getting loaded here? We need to narrow down what it is about that file thats breaking +ivy are basically the same +Ill need to make your svn tree read only while I am migrating this content and both the git repo and svn repo will be readonly until the project certifies it is happy with the git repo +Problem fixed +again +the usual limited content thats +Carsten glad to see youve picked this up +Fixing it is as simple is as simple as wrapping the latter part of the ctor in a try +Uploaded the rebased patch for. +Tested it out manually and it works as Lyudmil suggests +Issue may already be resolved may be invalid or may never be fixed +to trunk and changelist to the branch. +is now out so closing all resolved issues +The speed that hardware permits a task to read from disk or over the network has can have just as large an impact on the processing power it consumes as whatever the task is doing +So they are not technically a default but rather a precondition with handler methods just allowed to narrow that restriction even further but not to relax Juergen +Checked in +Throwing runtime exception for illegal parameter may be logical whereas handling any string as a number could be a little kinderThis is more of a preference than a bug +Robert +If needed please reopen with simple test case +If you finish it before we release beta please correct the version +Yes for a single node cluster I tested the following SSL enabled get file via hsftp SSL enabled distcp from hsftp to hdfs SSL disabled get file via hftp SSL disabled distcp from hftp to hdfs +This looks good to me +Thanks for your documentation Brian +what do you mean by In our setup we use an artifact with our netbeans platform suite instead of the netbeans installation of the current workstation +Reversed position on the meaning of depth +Committed and trunk +Bug. +Spec update committed revision +The difference seems to be that messages resolved against the providers are not recursively resolved +The schema tool has been updated but the platform does not include the additional createdrop back to the CP so that we can deal with this aspect +Simple enough stuff just likely to need to touch a fair number of spots where this check is performedSo Im changing the Fix Version to Final AFAICT this isnt a critical issue for CR +That is a much bigger change to the name node thoughId propose that we make this +Is there anything blocking this from happening now +committed fix in aedeacffddee +Could you maybe give an example? Or look at my example and tell me if I did anything completely wrong +Fixed in jBPM. +It would be nice to compare test functionality +New File creation was added table with a content of the properties file was added +Following Tapestrys foot steps Should we move from jUnit to +This issue was fixed in. +I agree with Ted this is material +Also for debugging purposes it is much easier to follow the state machine execution based on your logs if you have a resolved SCXML document +Im not sure how to fix this one should we close this one as wont fix? Otherwise does anyone have an idea for fixing it +Id recommend you try with and use it if it works +The script removes them +get containerState from instead of getting containerStatus from I do the test I manually create some folders under usercache with different owner and proper permission to test if all those folders should be deleted when the nm got reboot +Not all jars have in theversion number +removes AThere is no FAILEDKILLEDSUCCEEDED event for B since it was running when the AM crashed +fixed in +Ive just committed this +replace with + for api and tck +Looking into this more my preference would be not to fix this or do a hack for M since for component type side files arent really useful do to some spec issues +As you already mentioned you have implemented some components in HTMLWickets +Patch committed as revision I think it should thrown an if concurrentConsumers maxLogging at INFO level just get unnoticed +This jira has been closed as part of a review and archiving of the aj project +sure I will create a new file +and the app is not deployed +We wont upgrade for because of known issues with javamail and smimeWell wait for javamail just upgraded to so this is no more valid +Attached patch incorporating aarons is there an ETA for this one? Thanks +Ok so your answer to the question can I add a consumer on a running session? is noAt this point why the stops and restarts the session instead of throwing an exception? If the operation is invalid this behaviour is error prone and inefficient +Descheduling this issue because we will not implement this yet +This wont compile until is in but marking it PA for review +mvn U fixed it +Also the fact that this was not commited to the CVS at all all this time means that it is not being used by anyone +However if the ejb plugin prevents you from building an Ejb if no is found please raise an issue in this plugin +This certainly caused a WTF moment for me having to create a local server then it to remote. +Could someone delete this attachment that I mentioned please? Thanks +Please include the fixed version in +For now the scope is for the immediate problem but again will likely change +Committed patch to revision committed. +I dont find the visual editor for +Im sorry but youve got a few accidental white space diffsI also think it would be a very good idea if your validated that the types of the key and value that were passed in matched the types in the sequence file +A system property seems not flexible enough and might be problematic +patch committed to trunk. +closing the issue as duplicate of the newly created one in issuezilla +After installing on Windows there is a File path error and the file cant be located preventing Designer views and perspectives to fully load +It was the same problem as above in another method of +Patch v compiles under branch Running test suite +Nikolay what was the URL that you used for you test +This patch removes the restriction on when clause + Nicholas could you please upload the logs of the test failure you saw +This solution involves using for the URL piece of chained syntax eg +Added and the following items to the message headersMessage type Subject Thread ID From Packet ID and To +This is a really really really bad behvaiour of +Thats a good question and Im not sure what the answer is yet +After some investigation we should remove the purgeing on Core close again as this holds strong references from the to the +Submitting the failing tests are covered in and arent related to this patch +If I run with a client program executing threads and monitor the number of sessions in SQLPlus then as the test runs the number of sessions eventually reaches thats for and for SQLPlus +Ive applied the patch to trunk +The reason why I havent implemented that annotation is that should not be necessary in Hibernate its more legacy than anything elseIf you use save or merge Hibernate will know if your object needs to be saved or updated even if the strategy is assigned +Thank you Suresh +Form will be automatically submitted if there is at least one parameter in the Stefanadded in r +Cant reproduce with safari +Does not matter +Mark fixed this when did the changes on the +this is a suggestion for a plugin to m +cs something that is included explicitly from the +I was hoping to go that way with pretty simple helper classes that expose update methods +Ben just came about this old issue here +I think this is importantMaybe there is some way to have the best of both. +bq +uncheck Deploy projects as compressed archivesAlthough unchecked is the default +showDelay works for in both of FF and revision +ok it fails with too and Ive confirmed again that it works with +This is strange you cannot reproduce +Thanks +but it would be could to document these changes forthe release + +Could you attach as a diff +Thanks +Committed to trunk +bq +Im reopening because we need to remove references to components we do not distribute in EWS +I think the patch is fine +Hi I also work with shindig in my project but i cant run pubsub functionPlease give me advice any released shindig version that pubsub s in +Affects Array Array issue applies to +It makes sense to have pre triggerIts a MUST HAVE for Karaf. +Unfortunately WTP XML editor proposes several actions for namespace so it is to to +And pls avoid second screenshot relevant information are in attached log file +Hi Dragos sorry for bugging you again +rather than splitting between ISO and Identifier has been split and only ISO is in use. +Julien yep there was an issue with the Jira instance earlier +Can you confirm +As there seems to be whitespace preserved +I agree with option but not with the recursively partSeems to me that should be libIf you want libsatchel to be found you should be referencing Satchel in your code +BULK EDIT These issues are open to be picked up +bulk close of issues as part of making release notes. +at the end of the of the truncated submission +Im debugging the problem now +Added note to head of the REST wik page that REST has been refactored preserving the API +Thanks Benjamin +I think this patch is a nice middle ground +Looks like the tests depend on the last line of the output printing DFS status +Just to clarify this is limited to yes +Minutes to seconds. +I think you unintentionally always enabled rate limiting in the test casewell I intentionally enabled it but i missed the nocommit + fix +Same patch renamed +so scrub currently already has the side effect of upgrading sstables and is already offlinewhat if we modified scrub to be able to work on snapshots as well as live sstables +I wont object to closing the issue +its great they got that as their default handling my usecase is though I want to run a main java class with a specific scope instead of the default one +It may contain bugs +The repo should permit a Jenkins build but currently I am having problems building on Jenkins +done some reading over the weekend and worked it out +Does anyone have any objections for me to hack away at this making commits when I can? The intention is to work my way through the core classes providing a description of each package then get in to more detail with individual classes within the core bunch of classes +Worth a try at least +Table enabledisable acquires a write lock +The gray bars are still there +Ill open another ticket for adding default properties to classes where it makes sense. +OoopsIve mistaken in the version the issue appends on sorry +tying to figure out SPN for multihomed systems is a matter of policy +Please tryI have now removed the password +But first need to do that work on as most of our requests have multiple sort fields +How do you envision that technically? It seems to me you just add a new requirement that I like but Im not sure the new implementation can easily change for that +This requires you to test the model to see if it is inferencing since the standard model interface doesnt support the prepare method +So that means the data that it written on MR has comparable implemented in why remove the minmaxsort from +This is probably a local caching page is rendering properly for me +Its installed in production now +This is because the breakpoint updater validates the source against the Java AST and since there is no containing class type in scripts the AST says that the source is not valid and removes the breakpoint +The corner case should be fixed in +Thanks Lars! +Master goes down +Attached the patch with above change and tests correction +Here is a trivial patch with removing the unused private method from tried to get in touch with the authors of offline but got no response +New patch is updated for fixing test case +Hello LeoThe patch has been applied at revision r thanks a lot for this enhancement +All references to have been removed or replaced with references to +We likely need some refactoring across all of the HDFS proto files to prevent the need for multiple protoc build steps and also prevent duplicate symbols at link time +in +Just need to check if Git checkin is in the code base or into Commit adeedebaccdfccf +rel using jython as myfuncsregister Usersdaijypigrel using jython as myfuncs +Ill check again. +HiWorking on this +should be package private +When event is blur we dont focus the source component by default +so following that it would be a ESB runtime we would is more a overarching version +That way those methods and this technique can more easily be leveraged by other distributionsWhat do others think +QE Is this a regressionAnton Please didnt find this bug in previous release +Hedwig Provide a meta operation interface over hedwig just similar as what does in bookkeeper +open the diagram +Proposed fixAll tests pass including daily integration testsPlease review +I just committed this +Youre right committed to trunk only. +Thanks for tackling this Bryan! Looked at the patch and what you do seems reasonable to me +I wonder how many are like this one +Rob see Nicks comment +But if those tools are going to interoperate they need a common language of types with which to do so +Matthias has something else to fix in the plugins and has agreed to fix this as well +Bulk closing stale resolved issues +done in r +Not sure why I am specified as the current lead but let me see if I can take care of that +Can this also be cleared +Although i believe it would be a good idea to port longs to properly formatted dates for i do think itll be quite a task since its not only reformatting before sending it over +Included updating the client and JCC master files for test +Thanks ShankarAny thoughts on whether the +forgot to increase the pipeline length in the last version +Corrected spelling in title +I dont have anything else available to test with here +Andy can you confirm this will be fixed in +Oops +May you send an updated patch pleaseThanks OlegStill there + should be in srctestjavaorgapachestrutsrestexample is updated to use junit and include which is required for some reason to make this work +Robert I cleaned up the patch a little and did some minor optimizationsI replace the ugly surrogatePair checks with the version +I think that is the point on Oystins point if I may quote him. +Cant we check the number of versions for the entries and make sure they have the same number into the same bucket? A client might die before writing everything +bq +Now I would start changing the current trainers to implement the new interfacesWhat do you think +added bufferization to the inouterr streams of +The basic structure is purpose of locking of hashing modelsWith this approach you can currently build with fat locks thin locks and any combination of hashing model +I have checked in a possible solution where I implemented a custom writeObject that temporarily nulls out the resolvedObj before serialization +bq +Committed revision in +Could you please check this? I will push your patch and my changes to development branchRegards +That does not appear very useful to me +I have update OBR documentationCan you provide more information on your bundle? In order I think that we would likea +So how is this one going? Were about to release no + +how about add it to +other than moving for the sake of moving? Im inclined to leave things be that arent broken +However now this authentication happens irrespective of security and instead of container tokens nmtokens are getting used +Uwe what is Generics Policeman Point of View +This is the number bug that will prevent people from using it +This is definitely needed for C debugging but possibly for Java debugging as well +However my vote is to do the option +The patch for removes the module so if possible apply this patch before +But Im not sure whether it could be a cause of misuse and unexpected bugsIf you need any assistance regarding this please let me know +Youre right +So can both of these actions be mapped into the GPD +This patch is another solution to this problem +Now all resulting bundles use the same version solving this issue once and for all +I have fixed some bugs and attached few in case you want to try it +Patch to fix the thank you +Damn +Test case showing the issue +older issues for Apache since were no longer actively working on these at the moment. +Thank you Brett for adding this useful functionality +please provide a quickstart with that uses and reproduces this problem and then either reopen this issue or open a new one. +We have attempted the first end to end tests of the new pipeline +Not for commit here +We dont release GOAT so this isnt fixed in as that release doesnt include GOAT +Thanks Kevin +Another patch rev basically just doing stylistic cleanups +How about I write out the failed SQL command to the browser screen and mention to the user to look up this SQL command and its associated SQL exception in the logs +Thanks! +Looks like this might be achieved by setting in the but I have not tested +Nice Apos +The reason I tested with persistence option disabled is because of this reason I needed to verify its independent problem +Because the zipping utility used api a refresh of the folder was needed before executing the git add operation +Ok committed the addendum +still running tests. +Ill move it +I think this patch will do the trick +no point in letting this gather dust +HiI am also getting this error but the path for isnt relative +slip to next version any updates +Watch out the pom change to put the new tool in the build is in here and that shouldnt be committed yet +Maybe the right thing to do is to do first then we will have a better idea of the requirements here. +i got the same problem running sonar on maven with jdk any solutions yet +Sorry please use patch ! I forgot to make the regex replacement for greedy +Is it possible to wrap with Part in? Then we could deprecate it there +While I was fixing I had to change dependenciesThe project examples component now depends on the runtime componentThe runtime component doesnt depend on the project examples anymoreNickThe class and the configurators extension point have been moved to the plugin +feedback issue do you want to address with your patch? if theres a performance issue can you please share some benchmark results +so we dont want to specify a new readers schema per instance but rather one for a sequence of instances +I really hope we do support heuristics in XTS or we dont comply with the standard +The new has been committed to the AS trunk +Sorry the same problem that youll find in should read the same problem that used to exist in +Yes sorry this was a firewall issue on my Dojo Compressor has been added to the maven repo which is good news since now we can use a versioned version updating the title of this ticket to reflect that +There are a lot of integrations that rather novice people can accomplish if they look into instead of learning javaOne somewhat related quick question though shouldnt callmethod be catching all exceptions and leaving it to the to check errors afterwards? With the asterisk click to dial integration I was sending an invalid channel and it was failing to a white screen instead of continuing on +I think you are right retries are being done but of them fail in a bunch really as one failure +Please ask questions on the forum JIRA is for bugs and feature requests. +They include Oracle i and g R DB and as well as MS SQL +Woonsan I committed your patch but will keep the ticket open till the documentation is updatedThanks again +Of those million calls how often is this being invoked for the very same method signatures and target interfaces? I suppose you got many objects which implement the same interfaces but never actually contain annotations on any interface method? Are only the interfaces the same or the searched classes as well? of get with Spring for the additional context Stephane that definitely brings us one step change in Spring is that were searching for annotations on implemented interfaces as well +Please open a new issue with a more specific description if you have additional tests or changes to contribute. +ear deployment has been removed from the quickstart +So on recovery of that server the message would either be in the source or the destination never both +Closing as a. +It appears that the bug only occurs in maven +It only exists for one test +Dont forget that in EAP CP is Seam and this Seam instance deploys to default profile +Patch attached +on this simplification +Thats the part Im not understanding +A nice way to do privileged blocks is with anonymous inner classes so your test class doesnt have to implement a run method that does mutiple things +My patch only supports trivial but properly addresses the fact that the URL Resource as one of the JEE resources is missing from JBAS +Not totally certain what to do with this one +I think a better solution would be to move those blocks to other pages that are accessed through the tab bar at the top of the main content area +Heres an interesting though why dont we throw some ergonomics into Hadoops intellect? If heartbeats seem to increase the latencies on small clusters then perhaps Hadoop can lower it dynamically if a small cluster is detected +There should probably be a release script which removes the unnecessary files and compiles the markdowns to HTML +Hudson applied the patch for to trunk hence tests failed +Patch committed revision +Could be the systems weve tested on and our solaris customers have a certain patch that prevents a crash +This is not updated in the source code because that needs to reference the SNAPSHOT release +This fixes issues where the servers are not at the same time +I have the same problem creating the connector manually +bq +fixed in current SVN +The error seems due to an incorrect management of the annotation which may have been fixed as part of other improvements. +This problem is now fixed +oh +it seems incredibly trappy +Still its different from the default behavior since in normal Java classes you dont have to touch any combo but I guess its alright +Attached +The problem with the current code is that those links can point to nested directories +On the need for a directory to be set looks like the intention was for the code to use as a default but the code doesnt match that +Thank you +I am unfortunately not able to install GA on this server because it is not under my control but I guess this should reduce the scope of research +Not only for what the system happens to do today but what it might need to support in the future if we look to do work to make SEQUENCES perform better in the future +Unmarking this as a blocker for +will this suffice? Its not that but Ive been using it for over months now without any problems +Attaching a new patch for the issueAshutosh +Thanks Matt +Or am I missing something regarding your use case +In that case can we consider making that compression only featureFor what its worth that was basically the idea of having a compression +On the next screen add a new value for your Regression flag +You are right interfaces enums and classes are colored +The space in the value is not lost after escaping the before the space +OK now it +Also we support only a limited adddrop of just tables +First and foremost HTML support +Contents of directory are still not included +Do I need to do anything else here +Committed to trunk and branch. +there may be configurations where it makes sense +first try without testing +Im CCing the original author of the patch but he just became Father so it might be a while until you hear from himJasper +As I mentioned under feel free to improve the documentation +the final page says for Eclipse +Resolving this as a dupe. +For the remote access we plan to use the Springs +Removing them from view should be fine too +Canceling patch as we still seem to be debating the best route forward for this. +I have the same problem Im using Sql Server the latest the Sql Servers JDBC can not provide socketTime timeout so I occasionally hang the connection resulting in procedural paralysis I AnxiousmaxActive maxIdle minIdle initialSize maxWait removeAbandoned falselogAbandoned falsevalidationQuery SELECT test truetest falsetime num min +I wonder what the advantage of using quotes would be +The same exception after group by this column action +putting on beta timeframe for for Beta only +The RPC method name goes on the wire +Which is basically where Stu was going in. +Fixed by Bill by making the proxy cache use weak references + +makes a good example for others. +It would also be nice if a fragment identifier was updated depending on the current view to allow bookmarking or permalinking +whether we should release a plugin that only runs against Castor and above +It will make the conrefs SO much easier to maintain +IT added +I believe the only thing we use objectCashe for is to cause the GC to see the stubs as live but I think its preferable to keep as a data structure +In which stable version of hadoop this raid become a part and how can i access the API documents related to raid +I also moved commitlog up so that the CL dir was close to the data dir like it was in since one of my most common operations is changing both of those +this patch is against tapestrytrunk but I would love to see this release before the release +I mentioned in my first comment that parts of it dont work +I committed this to trunk and +Is this an existing feature I missed or should we submit a new jira issue for this +So while it works perfectly well for the it may not work correctly with the because the latter removes jobs it maintains per pool lazily +Hi FlavioCan you describe to me this problem? I will see if the problem still exists if this patch is applied after applying patch for +The overall goal is to make prettyI am envisioning contributions going into this for a while and then eventually well select a different default theme as a community and move forward with thatOne thing Id like to see in these themes that is a bit different from how they are now is to somehow reduce the redundancy between themes +The replace is included in an ifdef MINGW the W standing for Windows I guess well avoir the Linux problem you mention +Moving to since it looks like well have to wait on upstream to get this fixed +The patch has been committed +One of LIs graphic artists is coming up with something in time for Berlin Buzzwords +For example I think supports multiple property values expressed as multiple child elements +A generic approach would support what you already have but it would be a shame to toss the typed stuff and then a month later we decide the S case is one where were OK with having the typed config in our schemas +Thank you David for all your work +Im going to put the fix on for the changes how long did it take? Would you be able to check the changes in on a branch Id love to play with it +In fact only never and finding are allowedI think we should default to never when we get an error and nevertheless start the service +war instructions for importing into Eclipse in projects +That make sense + +bq +Investigating +I think it would be simple to keep it a hive only changeI agree that Hadoop change is more general but it may be painful to backportIn most of the cases k would be small and it makes perfect sense to put a limit on k as you suggested aboveWhat do you thinkShould we go with the change +Hi GertCould you apply this patch if you have timeBTW the issue is fixed on the component branch when doing the refactoringThanks for latest SMX branch. +Committed. +deferring to just committed this +For this round the default balancer could check with in the very beginning if runs the default block placement policy +Is there currently a test for that with listeners +if someone has more than terms in their query more often than not something is wrong and not being done efficientlyThey hit the limit and realize that and its useful for that reason +Wed probably have to do something very similar in other solutions +Are these tests reflecting the latest Contacts Sample in as I cannot see how result counts could be depended on for how many contacts each user should have access to +for the patch +This is being fixed in and current quick resolution is that update on readOnly attribute will just silently fail +I try both but without success +Accepted will fix as proposed +Second patch with the CORBA app client tests moved out of the testset root pom into the pom that creates the ear file +I it just a little bit nasty to have all dependencies twice in the because the main and the sources are needed +I weakened Array map constraints in +There are more pressing issues +patch to fix typos hopefully in the right format +new project no blank screen the alert is investigate for issue only happens on iOS I have tested it on Android and it is OK so the browser for iOS is Safari +Im not aware of any bugs in that are liable to cause a JIT optimizer to crashIf Sun could analyze any debug logs etc +Been almost a year since last update. +Its a handful of lines to implement so I can post a patch with it later today if you can please help me out with updating the hdfs tests since Ive got more demanding priorities. +Closed after release. +Instead of having separate option tags you should consider putting the options into a collection +Rampart specific stuff will not be visible in the WSDL +all setteradd methods do a return this +I would appreciate a wording review +please dont assign any more issues to +Thanks for reporting +a patch to keep a separated copy of ledger metadata for recovery read +here is a patch that implements those two classes and fixes a tiny bug in found in the tests +This distinction seems to be obsolete +Also I dont think a comment in the patch is a good way to express rants youd be better off by opening a jira ticket on that +Take a look at the DB contrib package it downloads +Will try updating +It seems to me we should fix this in unassignment handler if the table is disabled dont assign it again right +This issue is the fix for this problem in +or this is meant to combine the two so the is actually the right behavior for and? or matches the method name. +Do you think itll be too complex for users to specify? On the other hand this is for experts so let experts have all the flexibility +Attaching a patch for it really have to be fixed on? The support for anonymous classes for enum instances is not ported on right +I put them in the implementation test directories because the tests are simple and brief enough to be run as build verification tests +moving to unassigned for now to make room on my dashboard for tasks +I will run ee test to be safe and commit your patch unless I see any failure +But if you use screenletwith attribute set as same form as the same onecontained in screenlet the error occursMay this info help +I am marking this as an enhancement and well try to get it implemented soon +Hi Johngiven that is required in order to run maven isnt it reasonable that be the default java? if not then somewhere in the compiler plugin docs it should be mentioned that the default is s weird but i cant reproduce this error any more +it would be easier if you send me the javascript file after applying the patch +Keep up the good work +This was opened two years ago with no further comments +Theyre useless thereCan you update the patch? great +Thanks to everyone involved. +JimmyCan you put the patch on a real cluster and see if the second timeout is good for zk session recovery? per Larss comment +Everything is there +Personally Im very conservative about putting any changes at all into maintenance releases since there can always be unintended that we should be insulating our from +I think it takes a few days to sync your will try this tomorrow again and then we will see how it goes +Still nothing available to help us with it +Also the job level metrics are nothing other than counters of the running job +It should be fixed as soon as possible +Yes Context hierarchy should do that +I now want the reports +Fixing for the pureenums option +applied the patch +I would rather install it via the Administrator update featureHow long will it take to get to stable release stage although I guess that is the whole point of me testing itThanks +heh ok +In that regard if the user does specify a configuration file then I think it is valid to fail fast in case the file does not exist +I updated the video link. +Thanks Rishi Solanki to review the patchArpit Singh are still many simple improvements that can be done these set of screens +bq +I see the same set of warnings in some HDFS tests too +it fix too +Can post my install files if necessary but should be possible to compare the code with the previous version see what changed and what could be null. +Thanks DarynI put this in trunk and +Now there will be a line break between the startup and shutdown messages +The test changes went in on Tue Oct th as part of revision +Is it possible that the compiler doesnt support code on x and that I need to install another oneThanks! against trunk not an actual release. +However I will warn that this is still a discussion topic +fixes committed into tilessite as well +Hopefully other tests cover that +is a much more maintainable mechanism +Can you then please update +I should have clarified that +My use case does not require this to be configurable on a basis a global setting would suit me fine +Would it be possible to include the same set of unit tests we have in for this one as wellI would like to ensure the component works +No I havent any installations of GDAL but I havent also any installations of GDAL in win and ubuntu +Needs some more testing for sure but I wanted to get something in SVN so as to clarify things +however has some testing been done +If you find further issues please create a new jira and link it to this one. +The XML for Classs default section is wrong it duplicates the default section for Class +Ive committed this +Lets merge this +jar doesnt show up +This is +Adam Id say definitely submit it as a patch if you still have the code +I belive that the correctly stated requirment should be Its important that common version number for multiple modules can be easily changed +We cant relax our check for illegal characters as dashes are not permitted in package names for example and making any change would affect all modules and. +All subtasks are now closed +Changed all occurances of issuence to issuance +r for maven changes +Andrew Zeneski recently made a change to honor the isPhysical flag for inventory stuffsKrzysztof could you update and test this against your scenarioAlso if anyone else has test scenarios for digital goods or services please check those too +Im about to try it on production with a K url segment threads limit by IP crawler nodes +Any idea why viewNarayanajob isnt failing with this issue +fixed in rev Thanks +Create a pull request against the Git repo with the changes +The snapshots get generated at the moment +verified in +Lets wait for more feedback +The main thing is that for each page that is processed its linkmap is made relative so we could trim the leading slash there +Ill try my own tests in the next few days +Thanks +I certainly think it would be good to have a single source distro that contains all the pieces needed to build on our supported platforms +New jar with a fix for duplicate service registration +With Castor believe I think we came across this issue and decided to fix this +Since Groovy was split into several jars +Digging in it looks like our PMF is in fact giving a new PM on each call for a new instance +Could you review the changes Ive added to this page to document the proper order and limitations concerning the download and install of server runtime images +Here it seems that you can still have a design doc to handle some actions which could pretty dangerousSecond I think testing methods based on paths is awkward +Here is a mockup of what we are headed towards for the OCW user interface + +Can a property be specified that forces the deletion of the message from the mailbox when using IMAP +Assigning back to Ondra for clarification +I removed it from my diff but my have uploaded it unsaved +Logs snippets has a very good chance of being caused by. +The newly introduced component does not specify the required attributes to be a Component +Hi AndrewThanks for the research and comments on this +Instead Im going to close this issue and roll it into + +I can confirm it still happensStill I wasnt able to isolate the issue after three straight hours of recompilationIs code generation somewhat cached or dependent by external factorsI think I got both the correct and untyped results at different times with the exact same pom and source files even after calling a clean +If oldest table is older then user configured number of hours then run compaction +Hadoop record IO has had the efficient and proper handling of surrogate pairs since beginning +It seems some misconfiguration at your ASreopen if you replicate the issue on a self contained package. +What is the state of thisWill it be fixed +I have attempted to determine if any others are experiencing this issue but I have not had much success +Here is the patch for Secure now Make Public and Make Private links are working fine +Verified that the tree tag is no longer accessing icon images from the old resource location. +If that is already the case is it possible that the data being logged is very large +Great feedback Andres +We all agreed modules should be treated like lucene coreJust to chime in since this caught my eye I dont necessarily agree with this either I think it depends on the module thats what I thought we had agreed like some modules could be more like lucene core and some modules might be more like benchmark some modules might have really strong back compat and others might declare something more experimental +This patch is just to change visual components +still looking +I applied your patch the the version of ivy we are using and your patch fixes the issue +Committed. +looks good on both my screensRobert what about youL +Thanks Tim patch applied as expected +Attachment has been added with description a has been added with description source file for has been added with description reference file for test +Hum its a patch usually it works on trunk +I cant know what might have of different dependencies etc +to DOS format +For example if the client needs to allocate a new block it sends a request to the namenode +I havent it but Im wondering if getting the totalTF could take a significant additional amount of time for large indexes +One more important noteWith browsers you become dependent on class ordering +I am looking at this +i will mark all of them with a reference to this issue +This observation does make sense and I agree could very well explain the problem +If I dont hear back on this in the next day or two I will assume that it is resolved +The default isnt a major issue for contention for me was just stating a personal preference but it wouldnt bother me a whole lot either way assuming the choices are well documented +I would probably not popularize the use of JMS but having proxies working with JMS makes sense to me +Since the query is in wait mode anyways it did not concern me that we are making another database load and update calls are not part of the user query and matview updates are to the single source XA transactions should not be concern right? Especially when we introduce the separate connection for execution of these calls +The rat check is supposed to be a check for releasing +I also applied the same kind of fix on others bat scripts +Your points are well taken unscheduling +In this manner the option uses the new configurations API +revision +In my case both sides werent running in the same JVM there was a separate client and server +are properly parsed now as far as I closing all issues that have been in resolved state for months or more without any feedback or update +I just committed this +Ive created to record the issue of using CSS to specify HTML tags behavior +Can we add a unit test for this fix +I confirmed with mvn clean installThere is no problem +Nightly test runs in all window environments that were failing in the ibm runs worked last night so I think we can close this as fixed +For blocks that are being actively written to it is better for lease recovery to decide which blocks to delete and which to keep +Though itd be nice to substitute the short n sweet Grid for the longer and compound word +to the SVN trunk with some minor tweaksOleg +Bumping to at this point I dont think this is going to happen in DTP in the near term +hi why do you say its an improvement and not a blocker bugi cannot configure many default plugins in sonar with these releases and cannot add more plugins +Ok I ran the quickstart with a vanilla and got the same problem +Please verify +This means that the is not getting a chance to generate the JACC permissions after the is ll continue my investigation to find the point where the generated metadata is attached to the deployment unit +If not should probably go in a release or LATER +Well introduced some funky modulo calculation which caused a different behavior +Patch to munge the names of the internal builtin src files as well as the loading and test code. +An update for the initial patch +is it possible to backport this to +How are you formatting the code? I set intellij to format to spaces instead of and then I the files + JIRA with an seems to be my fault cant reproduce now everyting works as I get this issue again Ill open new issue. +on the idea +Currently the falcon bin script doesnt work with the mvn jettyrun version +Patch looks good +The exclusion of in the cass within the thrift stanza was due to the thrift initially including this as a dependency in thrift +I certainly dont object +This happens because gridmix rely on the getProgress API to stop threads which hold on to the sort buffer +Verified +Validated bug fix for XML elements for transient fields are now really ignored according to FAQ with an own test case +Oh one note +Not sure if all additional event objects are necessary because osgi core already ships some +Deferring post console work +Let me commit this small change to tests +fix o o o Resolving issue ready for verification. +Can you please update the patch to pull these dependencies down with Ivy +Godspeed. +Investigating further +I dont see regionOffline method being called from any relevant placeNot sure about +Speak to Brian Stansberry about this and create a JBAS Closing. +AnoopThanks for the review +I think these files are used but maybe one of the folks who know the internals better can correct me +I dont think we actually do anything useful with constraining types once they have been loaded but I had expected the resolve to fail in this case +It should definitely get called +Thanks for the tip youre absolutely right the request is dug around the Squid source code a bit and found that its actually denying the request based on the presence of the tried a few experiments with Squid it returns if the request header is present and after removing that it returns is there a way to configure the client to not use chunked and use instead +The progress monitor passed in just goes to the Progress view and maybe the little piece on the bottom right displaying percentages +Pamela Could you upload the current patch you have +The problem is still present in AxisIn my case I was trying to call on a WSDL that is importing an XSD +Bumped prio to Critical as its an integral part of our kit and the REST extension depends on it +Committed branch and trunk +The patch attached to this jira no longer applies cleanly to the trunk +This applies to this does not help too much +removed the old copies of the Struts Taglib Developer guides +Please let me know if youd like to see further changes +Ill try to come up with something today +Thanks in advance +Also the remaining issues. +this is a Struts issue right? Im reaffecting the bug +hiChrisI am so sorrynow this patch work well +Im interested in the patch +If you finish it before we release beta please correct the version +though it is IE afterall so weird becomes the expectation Do you have any ideas on the best approach for fixing this? We have temporarily reverted to to get our site working for IE customers again but obviously some day soon well need to be able to upgrade to or a snapshot +If it did not it would then assume concurrency issues had occurred use secondResult for the new low value and loop back to repeat the high value request process +Thats a good enough start for me +See the attached sample program which generate the error +If not let the event bubble up. +We are already using the fat thrift client on our branch +on commit +Forgot to mention that this version requires the cgexec binary which while not required for cgroups is commonly available +The addon list returns enttries currentlly and the postgres driver can be installed. +This issue is not a bug but rather a lack of integration and lack of firing the event to Catch. +Dont panic when you see all the changes to the parent pom its only just sorting the property names +Tested with response set to and with basic ASCII and for the patch +Feature applied to and master +INFO Running for rows took ms +Patch is a work around +Patch against the patch attached previously +Added the AJAX API extensions described in atteched files have been integrated with the apache repository +done +At first glance a problem in a closure being inserted by the desugarer to support dynamic checks +Hi Grgoire I did not run your test case as I think Entity is lacking etcBut I am almost certain that you are hitting is not a bug per se but rather a misuse of Hibernate +On the module depending on those I can include the test package to the test lifecycle + push to +I guess this is still not committed +Thanks! +As a bonus languages that dont yet generate unions specifically can just generate a struct and everything will continue to work +Available in HEAD. +This issue was fixed for Solder +prefix to pick the config +Ok +Committed revision . +Fixed in r +Have not been sucessful to date +btw what is the correct git url to clone for this project? I cloned the project off gitorious and it built fine it seems that cloning off github results in build errors +I think its a pretty significant bug that weve chosen to ignore out of convenience +Thanks Eugene +It was stored I save it as String now as s get this fixed for M the should already be in M +One suggestion for the test that is in SVN would be to increase the delay while you wait for the server to start up +What JVM version are you using +Further effort to reach a solution acceptable to all before anything is committed seems well warranted +Thanks Navis! +Sorry I missed this was available I will look at committing this +And I think restarting from an state is preferable to fixing it +Example gsp file that breaks the parser attached +The question is what to do right now nothing works anyway since the change of broke it it not compatible with anything right nowOf course i prefer not having to rewrite again to support config option only +New patch to fix the findbugs warnings +after revert Im moving this back to in via INSERT IGNORE +Oscar can you please update your working copy and see if it works for you? If there is still a problem please reopen this issue + looks good to me +If no objections Ill commit shortly +Several of these questions have already been answered by slow migrational changes that have been made to account for stateful decoderencoder usage +I would like to implemet this the problem is with interface because we need to pass more data from Query to process this something like where we can pass selected fields from query +Please find the patch attached +Really niceYou got a bug in the code tho as isnt valid for injection as it is a full list of matches What I mean is that for you to generate that list you have resolve it and that will possibly be time consuming andor result in a long list of possible matches wouldnt it be better simply to have in Open Injected class and then only if there are multiple options when you compute it then show a dialogWith respect to the search I was just wondering if this will fit into the natural flow of things +Fixed in +The file itself has its mime type property set such that it should as in Rev +If is true we use +Reverted back to resolved issues for already released versions +resolved TODO regarding illegal character handling +basically yes but you can do it without the args and always output text +Why not push it for inclusion inThat should be great +The packs the +Is there any reason anyone is still on and will be upgrading +bug. +If you need a persistent configuration open a new issue and let us discuss how to do that the best way +I might give it a try if you provide me as much information as possible +Updated existing with Fix Version set to to +Theres been no response on how to get the manual upload to working to add the missing artifact +Forgot to add this is in the t this be a bug against HDFSs binhdfs script? Having a script in places and picked up from places will be cause of confusion and configuration errors. +Could some one look into it and see if it should be in the branch? Thanks +Do you mean user code has to do thatThanks the priority to critical as the app can run it unfortunately have to be restarted to fix the issue on his end +This Metamer page has nested forms removal of outer form fixes the problem +autowiring by type is inefficient and in my profiling is taking a large percentage of the requestresponse time +I tried to include it in the patch but apparently there were some slight problems with +Ning please think it is fine to allow both formats +Closing issue out as part of JIRA cleanup +See the switch inside +I did not try the app client guess that some configurations are missed for it need to double check it +The common elements of that can be by contrib build files are extracted out into a The targets for clean dist package javadoc have been modified to execute corresponding targets from contrib build files +The changes plugin produces a report based on an issue tracking system +patch one test that needed to disable MDW temporarily was I tried cleaning up for the test it didnt work its still leaving a directory unclosed +build and build test runs on windowsxp and linux w gcc +Unit tests added to exercise code and user confirmation that bug resolved. +Move bulk of performance items to target +Many thanks for finding and fixing this problem +ive updated the whatsnew page with a starting paragraph mentioning the download locations and +I think its fine by default +See the attached screenshot +Thanks SianApplied as expected. +Given the problem with ValidatorvalidateValue we are not going to offer this feature +Here is the patch adding synchronization to +Its actually implemented this way in +Linking to JIRA with change suspected to be the cause +thanks for the quick fix! +What about the back porting side? Is this supposed to be backported +looks good I just committed. +I would like to see more tests though +this looks good +This may be related to is this still an open issue +This was fixed by the superceding issue thats been linked here +patch with the logic i attempted to describe +I Have not tested this since then that build was from late Jan I think +Is this still an issue +mvn install Build Success +svn moved +And the good news if that there is already a patch for. +spent some of time tracking transaction calls before realizing the real error was hidden +thanks for the patch +ve committed this +Just wanted to note here that has a patch that includes and all with some deviations from the proposal due to problems that came up while trying to use the interfaces +Please verify +So you are just doing what was doing internallyYes Stack +Missed files from +On the test with the this seems to bring around a improvement with regionsThe client can send between and million puts per second +If not it will move to Future +No worries Ill make it clear when Im looking for style going forward vs +text You might consider renaming the classes to something like and which are more symmetric with each otherif someone implemented my suggestion above it could be called backquoted or something +Was this fixed in +Sorry Mikhail I didnt notice your comments before checking in the codes +I have the same problemIt only appears when I have multiple Geoserver instances running on tomcat +should be a workaround in this please consider Hibernate. +Can you tell me whether the note above is still valid +I suspect that this content would be in the out of the box actions chapter under see if you can find if it exists +sfrom from setOwner + +This patch also expands and expression within udf in the query plan generation phase +Thanks for your comment +Im updated Spring Integration to version and all is ok +Hi Devinyou should have addressed this in the IRC that might kept us from issueThe customer might get a patched version that is not on the master +I have reproduced HDFS corruption with scenarios using and +Git repo is writableincubatortezsite is also now writable +WHERE INStill left to doWHERE EXISTS +Converter for model class should be used if generified collection is doesnt need of a bulk update all resolutions changed to done please review history to original resolution type +Pull request created +We do have several multiple package tests +During deserialization we dont fail validation in optional fields or collections simply because values in those fields are arent requiredAs a side note something I would consider totally valid is just not setting fields and collection elements when the contents of a Union arent deserializable +I took out the populating of the headers as I thought that it would be more consistent to have one way of writingreading headers +I applied the change in r and fixed up the tests that were testing the broken behaviourCould you verify this works for youThanks again! +Accordingly Im going to make some changes to the way that configuration is loaded so that the plugin does not need to do tricks to ensure that is available it will no longer used until after the time it is definitely compiled +The current patch depends on the fixes in this issue I found during the work on this one +You can see this for yourself in the Metamer sample if you expand the calendar dropdown and click the apply button +See discussion there +Thanks Tony MikhailPatch applied to LUNI module at repo revision rPlease check that it was applied as you expected +Nice find Matt + Meanwhile I will look at writing a plugin so I can get the functionality I need without having to modify the Solr sourceIm confident this will be fixed but in the meantime isnt the simplest solution to lowercase any prefix or wildcard query in the client +I have added the fix on both the and branches +Implemented the changes and added a unit test where the token has been issued by SAML realm A and the issued token is a SAML token of realm B +Will not implement this issue unless it becomes necessary +Just tried this on a Windows VM and it worked fine +settings files being in there suggests to me that removing them isnt going to cause any issues from a runtime perspective +This is not normally expected except in these rare cases fixed by the patch therein +mfojtik mfojtik May is this umask acceptable +a more realistic measure of if a DN is really busy by itselfCan you elaborate on what this means +I think it might actually come from mxml files that are not managed properly but I need to do a test +Id be interested in hearing whether you have identified such a scenario +Got it now +See for some background +Per recent comments on Leaflet is BSD licensed +JeremyWill you mind opening a new Jira issue for? Then I will delete all the deprecated files in this issue to make the things clearer +Ah +Is the problem dependent upon the contents of your policy file +does fail in latest branch with the same exception +The changes in binhodring where umask is set are still not removed +The reason that sucks is we have to run tests twice instead of just once viamvn clean install skipTeststruemvn sonarsonar +Some FTP server may not allow listing using a name parameter +It is not about EL for Java sources only +Remove all modules but admin and server from the root pom +Updated both patchesPlease apply both and and also put to testsgut directory +Haha this makes it even more confusing +Please take a look at this patch +Was bug with selection but it was fixed for in other issue +Maybe this is not correct way to express that +Please check that you have WTP installed and at least have the WTP update site in the update manager +Hope you like it +This is what the attached patch the directory does exist one could check for files +I thinkBut we should be able to fake one with a if we really want +Should we commit this change +So users could register their version of jackson jar in pig with true to avoid hadoops jackson jars being in classpath +also updates the copyright year shown in xdocs footers from to +The patch is missing some stuff but I guess Ill be able to handle it +Can whatever actually got committed be attached here as well? +Seems like we need to move the tests jars along with their specific dependencies out of the classpath along the lines of what was trying to do +I was using IBM JDK SR +Im using in some of my clients configurations now. +i changed eager to lazy in my onetomany while looking for a workaround and didnt notice +I havent studied the results in detail I just wanted to add that the property is not a standard property and it is meant to be a property that tells you the default encoding +We can drop the tag from this issue if you want to make major changes in trunk +Collation does not apply to parameters to SQRT ABSABSVAL because they only take numeric valuesUnary arithmetic operators SQRT ABSABSVAL only take numeric parameters and hence no need to worry about collation information on parameters to these operators +for the to trunk. +The issue is mainly about frequent loggingSince the patch has already been submitted with this description I do not think that changing the title now is a benefitbq +Ignore my last comment +Here are the problems that we need to solve as I see it We need somewhere to store the version supported by a connection +Initial patch for moving to hadoopWith this patch all unit tests with exception of streaming tests pass +We also lack a good central collector to coalesce the trace info and display it usefully +Changed master files submitted +Description was updated to be more JCR wont prevent users to create tenants with the same name since it is hard to implement and it will never happens since a tenant is rarely created +Already done +Document management is AFAIK less important things isnt +Reopening can we get it enabled on snapshot repository too +Is this issue related to +Thanks very much that should clear up the matter! +Apologies for this its half one in the morning here and time for bed +Integration tests already exist +So of is still a performance increase at the end +There seems to be enough dissatisfaction with this issue to warrant reopening it +tie these together another issue makes the change more explicit +If questions will come up Ill ask +bulk defer to +The only solution is buffering the events but this is not good at allRemains the question why is bug not closed? Are we waiting for a brainstorm +Do we want to add this attribute to the functions that we expose to outside world +The returned Filter can be used +menu + all issues not touched in a month to +No changes are being made to code or docs +Did the happen yet? Sorry to be a pest +Wouldnt this just give you one component factory per configurationI havent tried your patch but I dont really understand why you need the new variable +You will need to get your DNS people to fix the DNS entries listed in previous comment. +Performing work inside the segments and then merging the results later is how the Ocean is designed +Change Fix Version to +We could add a boolean to track and only log started if previously delayed +I looked at the source that comes with the distribution and the finally block still has the same code described in the original bug report +This will improve our enterprise solution of youMirek +Because of messages encoding is GBWhere can the encoding be set +Moved to CR for planning purposes +Therefore you will normally already have all the projects generated and imported so it is not necessarily an additional step +NadirWe need a fix in annogen +I have had problems with Perforce and in the past where it wouldnt detect changes that had been made and thus wouldnt trigger a build +Detection of submitted tab panels based on clientId instead of names +I knew there was a memory leak I just couldnt find it +Hope that works for you +I have updated some codes in the patch it will detect whether the clients browser is IE if not it will skip the url cache operation +Overall the patch looks good +verified +I made a mistake when testing as I tested groovyc not groovy +If you could spend some time investigating this and maybe coming up with a patch that would beChristian +Thanks Stack for the review +not required for files shuold not be updated automatically since it affects previous releases. +Confirmed thanks! +To fix this issue we need repair event which now not available +Ill close the issue right away along with +But a refactoring of the webconsole to have more is also a good idea +This is what I see in the Event Log Security Exception Failed to authenticate principaladmin securityDomain in is caught around line +In our case it would mean to format the text in a way that it looks nice in Markdown. +Anupam Can you please investigate the test case failures +Moving them to frameworkconfig seems to fix bootstrap +But I would want to confirm that there are no other unresolved issues remaining before closing this entry as duplicate +can be exported and imported alongside other columns tofrom a file +Got Java heap space with The last version of the offline scrub loads all sstable readers which means in particular that it loads the summary of the key index and the sstable bloom filter +Robert I committed a fix with revision on x I think we should wait a bit and close this issue +Hi Tommaso sorry for the delay +on commit if all tests pass +It happened more often than expected and I was nervous about leaving any assumed direct access if i returned a clone of the set +But Im not sure if this is the right place to rescue those parametersA Patch for branch is attached +I do not know that code well +JMX Tools Reference Guide +Please review and give your comments on this +Problem still present as of June +Random writer should be able to this problemTry a map only job with mappers each writing GB data +From a quick refresh of the Weld code it looks like we have a bug here which means that we are over validating when you have a more complex deployment structure +Filter queries now work +Thanks for reporting this +also not support take the predicate from XMLif I put a member in the interceptType I will create a Cycle dependency between the and +A lost tasktracker is one that doesnt check in with the JT and a tasktracker blacklisted for a job is one that causes tasks to fail for that job and they need to be handled very differentlyWe should move this issue to +pushed to plugins dev branch +One one hand host data should be stored in but that would require additional logic and sending each segment output to the hostdb in case theres a sitemap crawled +Ive committed this +The namenode tells it that those blocks belong to no files so the datanode removes them +Attaching a cleaned up version +HTML CSS and images of the homepage and template +Humm +Great +Closing as invalid +There are a number of methods for stripping the CRLF +yes i think jandex is as same as at least need testCompile no need for classmate or jandex until we activate the new +This should now be fixed. +A unit test is provided that checks a few message filenames considered to be validThe class is modified by adding some null checks +unfortunately both didnt help +After the discussion with Richie Im convinced now that the covers my needs too +I dont have access to the test environment to reproduce the issue +From what I can tell this is expected behavior +If there are any review commentsplease let me know +This is now it worksCan be closed. +Latest patch +Finding error logs on multi servers is a very very very annoying thing +The problem does indeed occur when serializing the DOM to a stream and moreover when using Xalan to perform the serialization by using the identify transform +I see +Because is about utilities helpers and basic enhancements around the JDK functionality +Please try with the latest cvs I cant reproduce the error with current cvs +Removing fix version of for those that are not in branch. +The patch the +Changed affected version would be helpful to know what database your are each using +Are these both needed +As a consequence the endless loop should no longer be able to happen. +Would you be interested in contributing a patch for this issue against the trunk svn repo? Thanks +I just committed this +Attaching a patch to be tested on windows platform +Verified upgrade in ER build +I fixed as well because it needed the same approach as notifyingthe listeners of the start and end of rebalancing at the right timeI added the consistent hash to the which should bebackwards compatible because user code is not supposed to implement it +However the Client gives NO indication to the caller that ALL of the endpoints provided in zookeeperinit are not responding +patch has been committed to the traffic server trunk and dev branches +Nick you can resolve this issue without applying the patch because the fix for also fixes this issue +For forward porting +They suggested setting the system property value specified in MS +Third bullet true just thought it would be nice to have an empty templateYes agree a test case should be added +NOTICE files generated using mknotice script from in revision ready for review +However tuning docs discourage using LVMBesides performance another obvious downside to putting a single fs across all disks is that a single disk failure would destroy the entire node as opposed to a portion of it +I cant imagine how to make this plugin do any less without using any change such as would require a rewrite that would completely destroy backwards compatibility +I did not plan on implementing it right away fro parquet because we likely will not get that great a performance gain from doing the predicates on individual values before coping them into +I also moved distcp and streaming Main to their sharelib to be consistent +Tiles released. +Example removed +Im resolving this WONTFIX +Version merge problem +Hello SumitHere is the patch for the above desired functionalityThanks to Sumit and RishiRegardsSurya Kusumakar +Resolving + +The issue was getting keys mixed up +The second option seems viable to me Cosmin +I took a look at the debuger and the error is in the generated class which is generated with the modified grammar with error recovery which is used in the plugin +I guess itll be ready within the hour. +Pavel youll need installed as well +I dont think we have an API for reloading an ISO if theres an error +Closing +Yup +Andrea That is a very good point +Side effects of copy pasting +IMO uppercase is better Common Sonar +verified at subversion revision +How about this? if saveNamespace fails automatically enters the safe mode +Well give that a shot in a minute +Eric to use it as a sure +Let me check once again +We should add for creating and manipulating cache poolsDo you mean protocol changes +yeah I guess so +The ref feature is still missing it can refer to future propertiesThe other problem is the linked issue about HTML elements and META tags being used in the body by large websites and not being able to read the attributes of the BODY tag +Reopening issue so it can be affected can be changed to a different version +Thank you so close +Maybe in +Thanks +So perhaps it would be good to add that in the test case +But to think it more even if I swap the impl internall the generics declaration from user field should stay the same. +Gerolf do you want to look further at this? Then just assign yourself +A merge can still be concurrent with a ram flushMaking synchronize on will serialize all merges and ram flushes +Its no problem in second level menus +Attached is a patch for version for this issue +The problem disappeared when I use Jakarta commonslang two days agoThanksBest result running is listed below +Assigning to the REST config main normally require a test case to be included in a patch in order to accept it for commit +I will commit it after running test +I appreciate Camel users can avail of but therere requirements to get CXF JAXRS endpoints supported too +Assign back to Abhinav to provide environment details +I didnt removed new comments + is coupled against the implementation of the old transaction engine which is why these tests fail +See linked issue +Were you hoping for it to be populated once and subsequent calls to init would just return +Please resolve that before compounding it +Isnt it more realted to the core of the engine? As if like there are SOAP Messages which the engine alreday deals currrently but now the engine will have to deal with a SOAP message with a Attachement so that we need to buld the mechanism to the engine +this was done +That could be rather useful for consumers and platforms in particular. +I happened to have run across this same problem earlier today so I might as well continue investigating +I would have expected it to test output but instead runs witherr DEwhich makes it revert to English I guess +However it shows that the external web service is being invoked which was the original problem so I think the problem has been resolved. +Hi Rolando I looked around and all table heading strings that Ive seen start with a capital thing that might have to be modified is the few titles made by more than one wordWorkspace name Workspace NameIs this needed? +Thanks Joydeep! +Fixed now +As we cant deprecate the older version of the methodI do not understand this partbq +Ive added a couple of tests to just like this usecase +Looks good +Check if you like that changesOther than that previous Patch looks great to me +The requirement of Labels Internationalization implies that the key attribute is required in the message tag +We more or less decided to stop shipping a styler extension so I am closing this one +Although the latter would suggest another serious bug somewhere else +All changes have been committed to SVN HEAD on the respective repositories +fixed in trunk +This can happen if you have disabled a deployer module or if for example you are trying to deploy an EJB module on a minimal Geronimo server that does not have EJB support installed +We implemented tomahawk +The fix simply make sure that for an index block we always return all the range tombstone first before returning any column +Committed to trunk and. +Do you by any chance have a reproducible test case +thats not implemented but in any case this info needs to be known +However it seems to be the same issue as described in +Please give the latest a try to see if it and everything else still works forMike +yes please do +Committed to JBESBGACP with revision . +Or if you dont mind and the file is not too large could post he entire logIf you could reproduce this please add a message to log the time that recoverLease return true +Well thanks so far +Eventually this could become a option but for now its only meant for testing +Its just the change to not the test change +You will need to configure your site skin to use another encoding +still fails +log file after upgrading to the last stable from October now contains line numbers sorry +passes mvn verify +Actually I retract that patchThe issue is as clearly stated if only I would read clearly that rewrites will also trip this system and thats just wrongErasing the count at the start of each handlerequest call is the right fix +Essentially the implementation is now static and works as a multiplexer for incoming handler requests routing them to the a handler service in the appropriate framework instance +Have IBM once again changed handling of transaction managerIf neither JNDI nor websphere nor websphere does work for Websphere please let me know the to use and how to get an instance from it +Guestmanagerstoragepublic all lables is cloudbrThis config is ok in the cloudstack +Im investigating this +I will let you know when I do +Fixed Shell problem and package new patch for this test +Has this been handled by your changes +Im pretty sure mouseWheel also cancels editing in MXMX DG did not use scrollrect which also made scrolling the editor tricky +The server exhibits the expected behavior to the client therefor I close this issue now +Are you? I think the only tool that I used was ld +Fix on branch revision . +Yeah totally agree sorry I missed that +Verified in +I think the option idea is good for the people who cannot choose what their admin decides to allow and what notI will be interested in the implementation +How do I write Unit Test +Closing as fixed bug Pavels request +We have a release ready to go and need to get this created +patch synced up to Simon after reviewing the patch I have some concerns about +Reopening to write release note issue as release note text has been written +Fix apply to trunk also +Added reply to mailing list thread about initial thoughts +None of them come up but the BPEL examples +Done +I would opt for using different annotationsProperty for string values for int valuesand so on +Why not make it best practice to qualify the component name? So components are and my own objects would be +Its merely replacing the actual default values with properties but if the default values arent used the properties wont be used either +Re under Element +I am going to run some tests and see if I can get this into trunk and +The problem is that with many of the pages content being based on SVN snippets the content can change at anytime +When someone has time to review the issue they will be able to determine the correct course of action +As an Apache project you dont need to use this method +Hey guys someone want to review and commit this? Looks like we got the OK from the multi folks +Thanks fixed +If the order of the rendered fields is changed by the user the order seems to be kept in most case +coretargettestinsruntest +Attached the full stack trace +Snapshot is initiated only If the write lock can be obtained +TrustinPlease close! +ThanksPatch looks good only thing which is missing are testsIm happy to merge it as soon as you add the tests +Fixed +patch as it does not apply +Could you submit the test case? I am not sure whether we receive an expected result set when omitting formulas in the GROUP BY clause +in its not able to find the user with with goes through fine for user +Committed revision . +Havent run tests just compilation +I may have time to look at that tonight +Ill now correct the Release Notes text +Other than that it looks good +There is however a bug because Felix matches at install time rather than resolve time but this will be fixed in trunk shortlyIf this is satisfactory please close this issue +Not targeting to a particular version because were exploring all options first +Does this apply to freelist without the reclaimable feature +Im sorry wrong project +So in this transition period we must authenticate against two domains so will need the feature for multiple LDAP servers +Fix committed to branch Committed revision +Thanks +Committed into. +Sorry to bother you if Im wrongIts a stupid spec compliance thing +Im always impressed how fast these are fixedLooking forward to. +Okay It took me little longer to squeeze in sometime for thisFor the timestamp based row view I have introduced new flag for the same this patch also have a small unit test with a test pig scriptLet me know if this is goodThanks +The only concern i have is to verify that local stuff is not broken when run from command line +We should be monitoring these updates and execute appropriate smoke tests as soon as possible pushing bugs back to the rd party vendor as needed +Can you describe a bit more about whats going? Attaching some of the output youre seeing would be of great help +BULK EDIT These issues are open to be picked up +This may be related to which was fixed in +Im not aware of a Java convention for file globbing behavior +Im not familiar with this code having not worked on it before so to get a resolution quickly where does index need to get reset? Thanks + for this patch +I finally freed up earlier today and will take a look at this issue now +Marking bug as closed as per last comment +Test case is added which is needed to be able to reproduce hanging at the special conditionsUnfortunately it was converted to test because we dont use +Anyway Ill run them on IBM and Windows to ensure they work on it works we just need to track down where the original test certs were generated from I guess they just came from the CXF project Guillaume may know for sure +doc +patch is attached +I added the PORT definition to in geotools and fixed up the rest of to refer to it and redeployed +This problem is believed to be fixed and should have been reported in C Squids Jira instead of here +works! +ill poke around and see if I can figure out why were leaking them +In case this patch needs also to update the spec file and rule new patch version submitted with rules and spec files updatedPlease review +Added license headers +Code is compatible with Hadoop and +tested again virtual services and worksSomething i am missing +I think support for the old lzma format would be beneficial for all people having to deal with legacy archives so at least support would be greatThen again adding write support wont hurt either + +Are you using java +The problem is that Wicket is giving an statefull id tho the hidden field and stateless pages get recreated all the time +So all blocks in the name table are and hence need renaming + +storing output in a csv format is very desirable feature for end users +rvm use +Yeah Ill fix it +I even managed to uncover a small hour to test the datum shifting methods should allow this task to be closed +Ive applied table fix as recommeded by Jonathan. +Hi SvanteI think we need to determine which operation system we choose as default +Original bug fixed and synchronization looks ok to me now + +Would it be sufficient to roll that session data up and expose it on the jmx interface for each stream? It would be pretty easy to expose the current and total rx and tx values as well as a percentage +Ping +Please apologise me for creating this JIRA issue +When parsing a nested schema that specifies no namespace the namespace of the containing schema is passed +I would not expect this to be found under new feature as hibernate did already support it so its a bug a regrression +Committed at subversion revision +Ill take a look at it when Im back home +I wanted to go to that page specifically so I could find out what the purpose of the Axis Service Archiver Eclipse Plugin is +It works finally +using our mandatoring stateful payload the to store this specific contains now a startDance variant with an internal callback parameter. +to nick suggestion to declare all upstream plugins on which it in feature would dragg in jonas and wtp jboss as adapters +Also the upgradeprocessing is not ready yet I cant really finalize the upgrade work until there isan explicit release whose jars I can check into the repositoryBut please if you have a chance have a look at and let me knowyour comments and suggestions + +If I am mistaken in assuming that this is what is need then please enlighten me +Removing the older version jar from these locations should cause ivy to be used +Yes either disable it in quite mode or batch mode Ill adjust. +related issue +already fixed +i think i need to add you into the team +pushed to column title +See the above comment from Keith Wall +build after the C client libs are built and then run ZKTESTHOSTSlocalhost make test after running srcctests startClean +Your points go somewhere along the lines I was thinkingThis first step towards improving the current situation has been done and will be release shortly +Attaching the patch of ENDSWITH the patch for ENDSWITH for the contribution +This makes the flexible enough to work with the and revs of Mac OS X and its vms +bulk fixing the version info for and all affected issues have in +Running it through Hudson +Better rename the method name to be get +Thank you for the review Devaraj and in particular for finding flaw in my test +Reverted backport of listeners in revision Reverted listeners in trunk revision If I have a better idea I will open an issue +Coincidentally this has been fixed for RC and already following a recent discussion triggered by +Anyway it doesnt hurt to have more options +pushed to both and +Im on this as well +Ive put a fix to CVSCan you please verify it +Changed component to doc Martin can you open linked in EAP and projects also in doc component so we can get the information into the relnotes? Thanks! +TODO items code cleanup tdataTable docs tdataTable jsp tags tdataTable tld tags note that tnewspaperTable is obsolete tests to see how newspaper columns work with extended grouping details sorting functionality +fix committed +We now use LoggingI think Commons is out of the dependency list now too so maybe you dont needIll ping Shelly about +How much of a problem it is probably depends on the scale +I went to html shiv manually since I dont have an old ie box handy +How about this Keep custom maven scripts to generate the parsers +Hi JuergenI am unfamiliar with the colon syntax youre trying to use in the expression +Created a branch to work on this branches +It should be rather straightforward to implement this in the class +Renamed to and added implementation Renamed to and improved implementation Renamed to and improved implementation Renamed to and improved implementation Added templates of junit test for above classes Added TODOs at literal and parameter tests where Udai has to add more tests Some minor improvements to javadoc in literal and parameter Tests improvement for latest patch from review and refactoring of main classes of expression and parameter +QA verified EWS Tomcat on Linux discovered and taken into the inventory +Hi BurtFrom the previous version ive understood that would affect only the registry of callback closuresI think the docs are clearer now +moving to M +I am on osx indeed +wed have to make it paint on a black and white image then multiply by the double in question and add to the previously rendered set of bytes +Please review tests working perfectly +This is essential to prevent the need for +I think this is a bad side effect of my change +Code available in the SVN or the next nightly build +AFAIK was added later than the release +Please see discussions on +Use the Roo shells backup archive produced with roos Backup command contains one entity class Dog and one controller class +at which point you truncate after and tack on uDC if its a high surrogateIf its an unpaired low surrogate I am not actually sure what the old API would do? My guess would be to replace with UF but it depends how this was being handled beforethe joys of vs binary orderI didnt do any of this because in my opinion fixing just the trailing lead surrogate case is all we should worry about especially since the lucene core itself does thisIll commit the patch and test we can improve it in the future if you are worried about these cases no problem +This plugin can work on any archive types so it is best not to tight it to any type thru a lifecycle +This basically allows the consumer to always be a slow consumer and messages are stored based on the max +Duh yes +Fixed and committed. +Other issues or potential improvements will be dealt with as they arise. +Initial patch showcasing the approach discussed above +bq +Verified +IMHO it is good to keep the rule of binary compatibility consistent + +Sorry for the endless droningBut for now the plan is to store the latlong and the geohash in two fieldssearch on the geohash and map on the latlong +All generated code has been moved out of src. +Thanks Weinan and Ron +Update wiki formating of description for better patch was not really applicable since synchronization are now more fine gained +Im not sure what the issue was still investigating +changes in API are binary compatible PR pushed to +OK with a sleep in the dispatchMessage method it fails on M too +If not which are you referring toAs for the schema computation Im not comfortable with having two version of it in the code +I do not think we need worry a lot about it +Please bear in mind that ezmlm and our setup is a black box to me and therefore I am probably constantly using incorrect termsWhen somebody commits to ddlutils an email does not appear on the mailing list which appears to be the functionality for all other notifications +Mark good idea but Im not sure how it addresses your previous comments +This one still needs integration tests moving to +Ive increased the priority to Critical as this is a requirement for Seam to be fully supported +Thanks +Also if we go your path it will be hard to send complete replica states as well as we will have to increase the number of arrays every time the number of state changes +Logged In YES useridIve read your CMP docs multiple times +Plus once I have opened the console and even if I close it will continue to open as expected +This is probably caused by my patch for +All of the function names in start with upper case so I am only following the convention there +updates the python cql driver regex to handle multiline SELECT clauses +Updated to include Pnative information without which the binary wont even compileTested this on a single node secure setup cluster +sylvain points out that doing this at the Thrift layer would break the row count contract +David is right should happen etc and thats what the spec says as of revision +Yes +Added the info about supported in like attributes +Verified +Does it make sense to round these? If so it should be added to this patchThanks few comments +See Andys first comment +Patch looks right to me! Thanks for handling this Knut +Can I close this then as it appears to be an issue with your parser not Lucene +Your change does not fix this +Just in case anyone is interested the +In my tests this does not happenBtw I have updated the driver to version Please update and have some test +If it doesnt work then please give a full example of what youre trying to do and well see if theres a way to accomplish itThe custom cannot be added back in because the internals have changed substantially +Bulk closing stale resolved issues +Attaching simple patch to just not wait on the future +I was unable to replicate this in trunk after the patch +Thanks Deepak Rishi for the contribution Done at rAshish you Ashish for committing the patch and Deepak for providing the solution +I was running JPA tutorial project +Patch needs to be updated and should include a test +For example the from Java Connector Architecture throws a which extends directly from Exception +My changes to for AS M yet to be verified +I can live with having the fragment in the same directory for now but not in the long term +Since the october the postgreSql driver is using an +Let me know if not +Would this be a case for checking the version number and either calling a timeout or not? Im also open to other ideas about how to get the timeout behavior Im looking for +And BTW I have ran the full test suite to verify the patch just one specific cluster test +Trivial patch +I guess I could care less what the default is if you care about such things you shouldnt be using the defaults and instead specifying this yourself in the schema and Version has no effect +pushed and deployed +Already fixed +So if I did not wait a while between the close and open the little bump should be normalSo I would say this bug is fixed +committed +Please change to if you intend to submit a fix for. +Sean Actually would you mind making the same change to qlsrcjavaconf? Thanks +Rick The failure you see is not related to neither autoloading nor my patch +Tomcat doesnt have EAR libraries +Fixed +Verified on. +Why do you require this functionality through the Java APIIf you need to make bulk changes to your entire set of ACL data then you can do so directly with SQL +Official template mechanism of grails is used and the reporter is not this the issue will be marked as solved +So I appreciate ifyou could make those changes as GUMP error messagesare killing me nowBut I still hope the BCEL project will address this problem morefundamentally in a near future +is there anything in your Archiva logs at the time indicating what the problem might be +If the version is resolved by doing this no attempt is even made to look for the parent project +I am interested in Konstantins comment for a different reasonI expect the DFS client shows a different semantics if the client dies before closing a file it created the file will stay and will not be removed automatically by hdfs +The code looks fine +I created to track this +Searched through all the source code and found one remaining reference committed a fix for that +Old one is now deprecated. +Better to fail and let the user take the appropriate action than compromising the password. +Well my comments are only questions based on your previous JIRA commentsI dont really know anything about decompounding Japanese and my suggestionsare only intuition mostly based on patterns that have worked for otherlanguages +Commited on trunk as of revision +Please upload your WSDL and soap messagesthanks someone change the affected version to? +Just find good shelter from the wind +Yes exactly for all your points +adding abstracttrue to the superclass table should fix this. +Can you give a URL +Ill add it immediatelyAny other step I could have forgotten? +We dont know the scenario to reproduce itThere were few queues with average throughtput of messages per day. +No I dont believe that the Locator has the information I need +Now that work is underway with AS all previous community releases are +Ideally the test wouldnt modify INITIALRESPBUFSIZE or MAXRESPBUFSIZE but rather grow the buf MAXRESPBUFSIZE itself +Warning is displayed if class is not found in the class path of the project +Important to doPeople complained when they werent +Attaching patch adding more testcases +Itll be great if someone else can test this to see if this bug doesnt affect other parts of pivot +rolled out +All unit tests pass with patch version +Mik can you check after starting the server to see if there isnt any error? Then can you launch the analysis with X option to see a better message? Your looks strange to me but I am not a Oracle expert +Thank you +Good point +Letting any interested maven dev assign himself +Note that we dont want to remove Pnative completely because that was added to fix a bug reported in +Compaction Coordination read performance impact can be mitigated if we make sure replicas of data are not doing compaction at the same time as the other replicas +I dont think Pig contrib had any big issues and in any case it is less troublesome than the alternative +Each patch has name which explains the fixAlso there are files which could be used to run the test +I will open a separate jira issue for that one +Mick was able to reproduce this problem on the newer code. +She may know if there is mention of HTML support anywhere in the documentation. +Unless another commit has happened since the first this issue is not resolved +Do you have time or do you want me to take it +but while I understand youre focusing on mapping now the big picture includes also global settings settings settings +Is a or sstable +Alex Thanks for the updated patch +patch on latest vectorization branch +r +Looks goodI have more items to pursue +I dont think this is verified by build unit tests +Lets go ahead and do that and file a JIRA to back the change out once is available +But I think its ok to have these +I closed this issue because it was a problem due to the UELSo no need to check itThanks +Reopened to change fix version to. +BTW i couldnt find the Test case associated with this issue +The way the platform unavailability gets reported is as followsAn unavailable filler runs every minutes on the server +Changed caching system to ehcache which gives us a configurable api with better caching options +Attached a trivial patch on could you close it? +After cache is cleaned error displayed +Jim thanks for finding and fixing this bug +Looking at the code inside HRS it also make more sense +Add a check to reload reservations to see if its for a block request +Tried both with parallel and serial execution +I am using this code in my local builds +There is a dependency hidden in here too on support Justin has done most of the work but it needs testing +sorry for the delay +Here are as discussed the originalmodified versions of commons vfs contains the original svn checkout including first changes that were made plus the svn directories so that the original version is present +Second will make this change redundant +Really interesting idea +I ran some of the tests and they worked fine +This will break compat +Alessio fixed this a couple days ago +This will make the very modular with clear dependency among various componentsReason for hosting on was for people to download and try as it was too big to attach to a Jira issue +Its probably because the +if tests pass +Note that it works with showcase. +passed all test that you added is failing +These paths should be global since its possible that even after moving a runtime you want your deploy folder and your temporary deploy folder to remain the same +and cant remember why i reopened it +Ran locally on hadoop and it passedIntegrated to trunkOnce the test passes against hadoop will integrate to Thanks for the review +The integration failing is subtle but can be fixed I dont think is a semantic issue +Hi HowardIll recheck it +The only mention of is in the section describing that is replaced by Messaging +Sorry about thatAlso at some point in future Oozie should have a test framework setup for federation +Descoping items not being worked on for M into Unknown Fix Version for issue out as part of JIRA cleanup +m ok with the general idea though Id have to study the code to really comment usefully +If we are using LRU to remove entries the current code will loop through all entries +Thanks +can you please make a small example for that? have made some tests with your example looks more like a side effect in CFMX for example are this cfcs not accseible from remote even you have defined remote access modifier +Delos please verify the fixes and then close the JIRA +When running in a reactor this problem causes maven to infinitely recurse +yeah is needed to remove it from gossip but it will be removed from token metadata even without which is what matters in this case +Closing issue fixed in released version. +This should be all fixed now on trunk not sure about merging back into will have to look at that when time permits +I know they are but that is a bug +Well leave this open and hope for a fix for this behavior in the next release +Do you mean using svn diff instead +Note that all tests including those new ones submitted as a patch here continue to pass successfully +To reproduce this issue? And what osgi container are you using +New patch addressing Brocks review comments +Do we want to consider taking references to Nailgun out? It looks like Nailgun itself hasnt been updated in over years! startup time has drastically been improved as well +Pushing to +bulk close of all resolved issues. +If you are lucky youll get all the items and can read them until you finally encounter the exception +Committed to branch +Thanks for the patch David. +added support for processing file during eclipse builds +The startup is slightly slower but still very fast +I believe spring batch shouldnt delete the file before the execution is completely finished +Confused about this +One more thing I included the corehtml tag classes so weblogic wouldnt complain but I am not actually using them so I dont know if they work +Tell me if this fixes things for you +Okay if you are not going to be using this code then I think its okay to close this bug +Probably def has no representation I need to check that but if it has then the check should be move to the return type processing +Just to add the page cache is larger in size for the BLOB table case because the page size for the table will have increased to k instead of k for the table without BLOB columns +Reopening to address Ryans question above +Hey NormanThanks for the bump +Since is rollback after upgrading to will fail without this fix +I am not sure what you are asking but I can build project youve attached using but it does fail when building plugin project +This patch adds an error message for the case when the database directory exists but it is missing +I just committed thisThanks to all who contributed with ideas and code reviews! +Move the issues to as not critical +The patch is related to build hence no test case +Ive be using as build extension without problems +If you use base +Just having a naked Change SQL Scripts Area Font in preferences tab is not a good idea +Committed to trunk in revision +This work has been started +Please provide a test case which demonstrates this problem +Allow me to disagree +Patch available for patch which is same as the draft patch +i need to path up to the ancestor build mvn N install from there and then go back down to the child +Good idea can do +Sorry about the eclipse files included on patch +To us the didnt seem to be worth it +Please ask questions like this in the forums first +Ill raise the issue within the JSR expert group +In class when sxdimport or wsdlimport is processed a check is performed and if the parameter is not present the default behavior remains as it was before generating relative +Attaching screenshot +These files should be places in srcmain applied the patches but next time please watch out for atlassian customizations in your patch file +seems to be a number of them +Patch applied in revision +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian certain that I verified this a year ago before the from Xalan went out in Xerces +bulk defer to defer to +Anyway alldocs cant be used to make joins it doesnt make any sense at allSo Ill leave unchanged alldocs as is and I will add a new query parameter for the user defined views to return null rows. +So that should not create any issues +Here is the patch with secured urls on this ftl now both the links work properly +Reducing prioriity. +I think thats a great idea! Plugging in different implementations for the Identification and Deployment services is the way to go here and having a UI to enter those manually makes a lot of sense for demoing and deploying them +When you say bumping the testcoverage would be nice as well do you mean rename some of the test methods sothey explicitly match method names in the tested class +I will be opening a separate jira issue for the other stuff Ive been working onA large part of the problem seems to be how hierarchical configurations are represented in memory and how they are being shared in a combined configuration +Bug fixed +Version Released +Can we have confirmation that fixes are in place for +Done please test! +as commented on the fix is for only since the file versions of ACG differ too much +This issue has been fixed and released as part of release +Ive ported this to since I cant access to Job object in class instances. +Just to be clear the qa db change stuff probably doesnt contain the changes we need here +building from src package works here. +I committed the patch with some additional changes simplifications to trunk in revision +Hey Suresh where? I dont see anything which looks like what this JIRA is describing at jmx on the NN +The new patch version has been attached +Yeah we keep readers open so we dont pay the opening price once rather than inline w reads. +It is because the socket read callback blocks infinitely that this happenesI added a timeout for socket read to prevent this +Implemeted with a switch ducky to toggle implicit duck typing behavior +How should this be implemented in a common way +While I was in there I also changed the copyright yearIt would be nice to include this footer on all pages but my knowledge of Forrest isnt up to that taskTouches the following filesX buildsiteM srcdocumentationM srcdocumentationcontentxdocsPerforming status on external item at buildsiteM buildsiteM buildsiteM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteM buildsiteM buildsiteM buildsiteM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsiteM buildsiteM buildsiteM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsiteM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsiteM buildsiteM buildsiteM buildsitedev changes to website committed at subversion revision +Can be used to look into the issueIts either the CSV data format or something you may do in your beans +And no theres no warning +For now the jars included in Pig sharelib will work seamlessly for Pig with +Well set up reports for after is out +Another implementation of works fine. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Fixed +Making this a blocker since it should really go into +am working on including it as soon as we figure out the fix + is a first step in avoiding any allocationcopying at all but will require changes in the scanners and heap to use Cells all the way up +Passing by hadoopqaThanks Lukas +So this is marked as fixed was there a fix put in? previously it was in MVEL itself +I attached the info thread output of the hanging process from gdb on a x machine +Spoke with +But the updatetimer will update all the regions in RIT +Path for release +Hi Thorsten thanks for the patch I applied with minor modificationsThe problem with class is that it does not return an object in removeAttribute so we have to make a duplicate call once getAttribute and then removeAttribute +Thanks for looking after this +Just a quick noteI would be glad if someone would review what Ive written +Alessio can you comment on this issue +It was noted that reparses the schema for each object it serializes this is now memoized in the instance +Would like to commit today +There is also an implementation of this feature request in bug number +Trunk r tickets +Just the following static html files were changedModified oodtdocssiteabouthistory oodtdocssitedevelopment oodtdocssitedownloads oodtdocssiteThese were the ones that needed to be updated with the news about release +Ted or others let me know if there is any particularitems that would help get out the door + +adds validation unit tests for step of to +This was completed as part of +So I looked again at the problem and found out that supports registering +Thanks for figuring out the root cause +That RAT configuration is common to all projects and is inherited from the common +Im fine dropping schema prefix and going with keyspaces columnfamilies but how do we name columns cf then something like columnfamilycolumns +Are you closing your connections everywhereCan you provide a simple test jsp showing the problemThe mailing list is a good place to debug this +Please close this issue if it works for you +I will talk with some about this to see that we can come up with +It needs also two boolean fields added to public boolean force public boolean wherePath +This has been fixed by the changes to the inner class processing code +Ive a commit which fixes this issue +At that time the motivation was better edit log performance rather than preventing partial edit log operations at the end of the file +Now web API ignores the case when matching the applicationType +I wrote it so that the signals to the with an event +Afaik this was all the result of my failure to close things. +Look intoolsr ldimaggi ldimaggi Dec r ldimaggi ldimaggi Dec is documented in the Getting Started by +I dont think we can afford to dilute this with footnotes about how you can escape hatch back to the old world dragging in all our old legacy baggage again +It you have any other examples of this problem it would help to know if the end of the file looks the same as this one +The shell script has a commented part for now +In any case I verified the patch Committing it +SharadThis update to the patch removes the key attribute from the interop map +We just ran into this bug when we upgraded to should be released in the next two days which has a fix for this +Hellocould somebody please push a to the +Verified in +Im currently testing the Plugin in my App to see if it is behaving correctly so far it seems goodIve sent a mail to the Dev List to find out why I can participate in JIRA as a developer hopefully I will get an answer soon! +pushed upstream +RightCan we find out why this target directory was created +Done +Committed to trunk and. +reassigned to myself +Again the notion of overridden by does not make sense for this has been the case for a very long time +In that case the jarfile +The xsd states that the finalName includes the extension yet as the original issue indicates the extension is ignored + is not supported and this change doesnt compile against +Fix up a bug of error code on WindowsPatch containsAdd check on the null into getvmthreadfastselfAdd support of signals shutdown processing for Windows platformsAdd empty stub of signals shutdown for all Linux platforms +It looks like the javascript was firing just fine but in the CSS on line the default style displaynone was overriding my dynamic inline attribute fom the class file +Its workingThank you Andreas +Thanks for javadoc and findbug warnings +This code has been around for a while and it is not clear where the regression is +This is not the kind of feature creep we want to be adding at this stage in the game +Does it mean we need to get this new into and release this plugin after that +Displaying of select controls works fine now under IE +Hey Ashish +The word desertification is something used instead of desertization +Also fixed XML source files so eventual rebuild will hopefully keep the changes +henry can you make this patch relative to? i plan on committing tonight +Thats ABC and you dont need to prove that +It would be be trusted by many peopleI think whatever solution you pick it has to be consistent and preserve the data that is sent in correctly +Ok Ill add the tab So user has both options +Thanks for the help +The reason I made the branch was becausesome were going around about making drastic changes to the JDO code baseand adding new dependencies and I wanted some stability in the code +If not please run with sh xv binjruby +Attaching a simple fix that increase the number of operator counters to +This looks good to me +It looks like this would allow dropping as well as allowing based servers +new version that only starts time is now configurable +At the same time if we use a variable name instead of the configuration name people think less about the name and we end up with bad names which is really irritatingOur approach to this was to get all the committers to call out and ask for review any configuration names in their JIRA and try to get extra review on these and then not change them +and committed! +I forgot to mention that tests are not included because the current ones are sufficient to prove this change does not break anything +So I upload new version for diff file +It looks perfect + +pushed to master and branch as with JBDS +Thanks for revise +Patch applied to get that stops scanning the input when it reaches the beginning of the xml content + +Most of the classes this report refer to are gone but WCS still has these issues +The dependencies internal to it +Committed to trunk and +Accounts have been created. +Fixed one minor compile issue +Or use streams that can be. +If it is fine I can commit this later today +java classpath langjavaipctargetlangjavatoolstarget s +OK so connections using JNDI respect the Socks proxy settings and those using the Apache Directory LDAP Client API dont is that itIf thats the case this is not a normal behavior and we probably need to tell the JVM to use the Socks proxy settings even with the Apache Directory LDAP Client APII will look into it +Will attach a patch after only? and on the parameter value as Doing to escape all special characters does not work +can anyone review it +TimThank you for the explanation and now I got the bigger pictureI am in the same boat that I need the cosmic version change for my current project but without deeper knowledge and experience of the Aries project and resources I am very reluctant committing to do this work now +I will document this +Fine with me +Fixed for INOUT. +I just committed this +Seems reasonable but maybe we should get buy in from all the project leads FIRST rather than just decreeing this is the new way forward +Would you please help to try it +Linking the issue to +I saw the test not running at some point but it ran for me the last few times I checked +OK thats a good clue the Threads are created as anon classes which means they are holding references to lots of extra stuff +Send us a patch and I work on getting it in +Tested! +Done +Try setting that to true to see if the providing the property tester isnt active yet maybe +forgot to update fix for version +Thanks Lakshman! +There is also a section Upgrading from Solr . +The job that does this checks for being unique and accumulates a small delay to avoid multiple updates. +Question answered and opened related issue to finalise solution. +Changes are committed at r + part of part of more is only thing used in publicationKeeping andremoving Removing duplicate part of out why these dirs arent included in the master build and part of the publication +So if I know what to put in their I can try to reproduce a working Arcgis server WFS processed by geoserverKind regardsThomas +These blockers and critical issues are resolved but not verified +Its hard for me to create test suite because I have to read a manual about automake autoconf etcYour decision to make getescapedstring is better than send this to constvalueHere a couple strange things in your patch we dont need tphpgeneratorescapestring anymore why do you escape apos in html generatorSo if escaping is right its lgtm +But as the patch provided there is a clear separation between the two cases and code is easy to understand Also the second case only occurs once +The JAR hasnt moved with GWT has it +Henry your patch looks goodMay I request you to provide a patch via git pleaseThis the workflow I personally follow +Will do +I will commit this patch as is once I have been given a sample XML document that can be usedd within the new CTF test to be unmarshalled against the classes generated from the sample XML Schema +mesossubmit actually has its executor become the new scheduler so that those entities are unified +You can play with in the attachment which should be dropped into srctestresources +The branch is dead +The seems to be corrupt +Ok I committed this to branch too +Its a little scary to expose this as a supported API for directory +Committed to trunk +I think the libs were done this way just as a mimic of what does +Wouldnt be this needed for htmllink too +fixed in svn +If href onClick +It sat there for a number of months in active development before becoming a subproject +This issue doesnt relate to Hudson Sonar plugin +This is wrong. +I just said if you use a LONG docvalues field and store the long epoch its always +It wouldnt be a general purpose helper just one for the specific problem of doing assignments in the runtime so placing it in the same package is fair +That nice +Is this still open? I think the singlebundle tests should cover this +fixed. +Hicould you try using there were a number of fixes for handling fontsBRMaruan resolved this issue. +Should eliminate parsing problems +Updated patch including changes to srcdoc. +Carefully measure the memory usage of map side group byFlush frequently if the left memory is less than a threshold +Thanks Hairong! +Submitting the patch for Ravi to see what Jenkins says about it +Only support new signatures and privatepackage support for bindunbind methods for DS declared components +I do know that Maven was also burned by a collections change so I wouldnt be surprised if its the same thing +I think the homepage url should be on the group summary page not on the project level +Rather than making yet another version of this kind of component is there a way you could make this an option of one of the existing components? I dont have any experience with any of these tags so I dont really know what youre trying to do and if its possible to merge it with the existing code +Jenkins failure caused by upmerge issue +Of course my integration tests are explicitly creating caches so they can make assertions about statistics +Any chance we can remove the ultra long version identifier from the category namesIf too much hassle then not a blocker just look for something in +The new Infinispan Server distribution can do this already +IMHO thats a waste for a large index +Change is committed at revision revision I updated the code slightly to always update the target platform when the server runtime is changed +latest patch lgtm +versions +I just made a patch that adds the Share output folder option as well as a default output folder if its not also noticed the output folder for resources were set to themselves for a reason Ive also restored that behavior when output folders are not shared +I will debug this and see +Sorry if I was unclear +fixed and confirmed +Thank you for your reply +Where should I look for a working example at the server? Or at least a server which works with your implementation? At first glance it still doesnt make sense to youre trying to retrieve any kind of information from a protected resource HTTP makes sense to me +But I am inclined to believe that the keykeyref should work in the example that I post it +reopening since is only related not replacing this upstreammaster +The Axis stacktrace +can you give us a link to the source causing this error +or perhaps Im counting too much on the optimizations + +It will replace all errors in all the files +Test file can be freely distributed +check ample documentation +Closing tickets +Bulk closing all of these resolved issues. +Please verify +What exception did you see without xerces in the CP? Furthermore the quotes seem to be right for me +If you use your patch in a environment what happensBetween the two of us youre the portlet expert so if you say it works better in portlets with it commented out I believe you But we also have to support all of the users too +Do you know if this got incorporated into bootstrapSee +The attached zip contains the source and classes used to run the test +Settings for Hudson are a user setting file homehudson +What if we could turn it into spills one collation and fetch per reducer? That would be a big saving +If there are no objections Ill commit this tomorrow +no other java processes +I assumed it was in yaml because that is what we usually use for but it is actually already a property. + +So is the blocker for this issue the fact that there is a custom zkclientI believe so +Now with only four tests failing +and use it for the source of properties about an +So this can get committed now right? Can we get this in prior to the cut over +Packing process must accommodate signing process once we are ready for that +From Dans comment in this issue I think it is an unlikely source of the problem +Fixed in the trunkCould you test it and verify +It would be nice to get rid of the s of +thanks for the patch marshallCommitted revision +looks good +Verify that move to curator fixes m thinking about scrapping the work Ive done on this +BTW I kept reduction of redundancy +So Im leaning toward pushing this to Future OR rejecting it +Resolving is fine ER tag fixed now in commit +Fixed an issue in Added more documentation toMade writing of skip records optionalIncorporated Devarajs offline comment to remove the flag as it is now redundant after the addition of MAPPERMAXSKIPRECORDSREDUCERMAXSKIPGROUPS +as requested by Benson +Thanks for the pointer I have already taken a look at the code but will need more time to evaluate itbtw +Note the chapters to review are numbered and instead of and since I trimmed out chapters to which have already been reviewed + +Please confirm the current status of this issue and if its incorrect change it as required +Looks like is properly setting the context +This patch helped find the problem in +The version of the patch is needed for the branch +And Weiyans patch should be fine with the synchronization +I think I see why this might be needed nowLooks like the best way to handle this is to allow m to filter these files to pick up the configuration and thus we need a plugin to set these specialized dynamic propertiesNote I had originally thought that antrun would be best for this +An all optional field struct has just as brittle constructors as one with both required fields and optional fields +After some investigation something changed in Weld subsystem that broke some functionalitiesInvestigating. +Robert and me noticed a small issue Javadoc does not regenerate the stylesheet if its already there +Ive applied the patch here but it would be good to know if other environments no longer see this test issue +Moving to +Using the escape character works +In addition I came across an issue that caused the cursors to be refilled a single message at a time which really slowed down consumption +Committed patch b to trunk with revision Ill backport it to after the nightly tests have been run a few times +This works fine for me in CVS. +Indentation in is inconsistent but otherwise everything looks good +So marking them as related. +Thanks Roger +of the speed of the original lrublockcache +to removing and removed and stefan would you still object to removing as well? please let me know +Sure either way is goodBTW thanks for reporting this bug +No work has been done on this in a while and the GEP has been made much thinner than it used to be +This issue is related to. +This class should replace previous class +Please review +includes the same featuresplugns as the feature above examplesmaven plugin does the same as but does not include the community examplesmaven plugins Is that correctIf it is why doesnt and just include their relevant plugins +A new patch for hbasea getUnique getb +Anything missing in the uber build or in the bpel build +Now I have to reproduce the same behavior even on Tables and Trees before closing this issueIn the meantime many thanks to Edvin for the help +Most of the problems described here are fixed +Thanks Gunther! +Out of date no longer exists. +Pushed in ba +cannot reproduce it in current TRUNK code I guess its resolved then +this should now be Infinispan +A component has been provided for which is not included in source tree but which has been added to +Eventually once we have a few releases under our belt and successful deployments etc +I would also request to add javadoc in for the reasoning behind the ordering of logical optimizer rules +Thanks for looking into it Matt +I believe this has to do with stopping the management server during startup +Please note that i ported the js code but had to use to avoid exception The continuation is suspended and will never regain control +I will review it later today +Issue is closedRegardsRobert +Patch from review that we are done with this we need to Add validation to the upgrade method +r Thanks Benson! +Thanks MyrnaI had automatic ordering of imports and indentation formatting turned on in Eclipse I didnt realize this was a problem +Had meant to remove the Zab Test part from this patch as Alex tells me that was already committed to trunk under another Jira +We have the same problem here +javaBut it should be just a short file under directories like this classes This only happens when deploying a cluster with one farm node on windows and the other on linuxunix +This is going to be really useful when implementing new services +Calling conventions are quite different on Winx and Linuxx +The lines of code you have found are not the problem but I am very interested in the actual issue +the new patches lgtm +Attached is a patch that makes offset assignment by the log optional +Added in revision after you please open a new issue for thisThanks +removed +Unable to set change to branch and trunk to change parameters to. +The other pages +agreed that we shouldnt do that +Also the holds a cache so if you typically label a certain set of categories the index is likely to never or very rarely be hitI think that maybe during indexing this might help but not sure +Classes and proper way of contributing some code is by providing a patch +Patch solving issue with file in some behavior in this case is correct +committed to TRUNK +I will suggest to not do major upgrades on our patch branches eg +Defer all unresolved issues from to +Im going to commit this in next few hours unless objection to trunk and +Currently the topology maps would get out of sync if any endpoint changes dcrack +Screenshots +The in eclipsemavenbug is when the test application is executed from within eclipse the maven dependencies of mavenbug are obviously maven dependencies does not solve the problem in my that eclipsemavenbug is already a maven dependency in eclipsemavenbug has changes in subversion which I want to test +amit my java home is set to usrlibjvmbinjavauma Mashwara Rao Sorry i will be careful the next time +Alan I dont think thats what we want +This method also returns false if c and the current Type represent the same class +Ok id like understand better the question related to Licences +This is surprisingly easy to do at least for our module specific extensions. +to attach related files +is the rebase we needed fix for the thing Nitay mentioned on RB and a small change to which does slightly alter logic +We havent tested with solr +Was off the grid while this went in +Attaching patch against trunkThe code apply after because the refactoring of in there allows to do proper validation for this patchAn added benefits is that right now when we receive a we transform it into a Mutation to use generic code +Yes +Ill commit unless I hear objections we can iterate more with a couple of follow on jiras this one is too big to keep +Here is the patch where parameters are send as a part of +Its unfair to put user errors as blocker issues of a project +Its just that after the SCR request we would hang on to the socket and use a different protocol which I call the SCR notification protocol +A if we are having changes to target platform every week something is very wrongB he can easily use a specific target platform locally no need to push that into every other we dont want to add in things that are obviously wrong to do +I think we are seeing a lot of these now in the ibm runs runs because we have switched to running multiple jvms concurrently leaving less spare cpu cycles for background threads to run +now +Thanks Akash Jain for your contributionSpecial thanks to Divesh Pranay Vikas for review and discussing things with Akash +There was indeed a subtle problem in collect and accumulate in trunk and fix will be included in the next milestone let me know if you face any other back to development +This patch renames them to remove the TEST prefix + + seems to have the same problem +is ok too and there arent any visible changes to be seen +You cant apply this patch since it would mean the project doesnt work in Java and Java will be around for years +I created for. +Im not sure whether the underlying technical issues are still open here but I do get much better results now with this sample request +This is fixed in Warbler rev dd and will be in the Warbler release +I think I have the rights +Need more information on this if it is reproducible +When creating applications with an initial set of cartridges timeouts get even more important +MB larger than what is accepted by Jira +Thanks +can you please elaborate +Closing as Sam answered the question. +There is no opened Generate in External Process functionality inHence change fix version. +Yes youre perfectly correct +I just committed this! +Are you using a post Preview release? If not plz update and see if it still persists +Please identify the providers of that plugin and send a request to them to upgrade their plugin to support Axis +OK third times the charm coded this one against trunk and just successfully applied it to a fresh +I mean what is the interpretation of an empty node? There is no sequence in the input that corresponds to this path and it is a prefix of some path in the input that you cant get from this automaton right? This slowly becomes very confusing +MennoI applied a fix for the problem to SVN trunk +I created the issue which refers to this issue but lists as the affected version +There is however a valid workaround as noted in the JIRA so Im closing this as deferred +Therefore I created a hierachy of unchecked exceptions that serve as wrappersTIA for reviewing this huge patch! applied see r thanksI just fixed some code format incongruences and dropped few useless statements ATM but didnt see a reason to keep it a lothow do you make sure the formatting is okay? I just run mvn site and look at the reports +Not always but I can reproduce it +i was made the mistake of using the tag instead of the branch before which was why i still saw the problem +I think we should mark this as a blocker as this can kill HS which can have dozens of jobs running in the cluster for clients +Oh so this is from the old SLA stuff that explains some things +Ok I found out that the mysql connector was too old +Damn sorry I forgot to resolve it again +Is it perhaps under another name or service +done for the second task +Do you need any more info from me +The PR applied to the master and branch +Yeah but so long as new releases arent missing it im not too concerned. +Summary of list discussion For LOCAL or REPL it shows contents in memory +Thanks +From Ivo Studensky Hi Igor Ive just hit the same issue on RHEL I mean this one errorSSL routinesSSLREADBYTESsslv alert handshake failure I googled a bit around and then got to work it with an additional parameter to the wget command +This is not a blocker +Here is a pom that adds +close this issue. +Added patch Fix for sending pending class load events in the beginning of start phase +Should have been resolved. +Manually tested +Run managerjmxproxy and search for roles +Applied patch by Roshni Basu +The bug fixes were for font caching so perhaps the wrong fonts are being loaded which is causing wrong character widths to be used +Ill remove them +bulk defer open issues to defer to not essential to address for +Seems that branch of logic needs to be called twice because we load the plugins with two separate class loaders + includes a patch for has an attach patch for this. +All that needed to be done is implementing the new interfaces as created as part of et voila +So we have a lot of places where we wire them by name instead of type +Committed to trunk +It should be +Hi Nicholas and Suresh I put some additional cleanup work on +Ill try the branch in a bit +This even includes the change +Hi Sebb where exactly is this date and I will get them changed over to ISO standard +thanks +Can you please verify close this +This logic has been preserved and using the archiverConfig element of the plugin configuration you an override it +Can you attach a unit test or provide a test project that can be deployed in order to replicate the issue +Rebased branch patch +examplesbinwork directory and applied on +Closing as nothing to do or see here now +has been marked as a duplicate of this bug +Yes the suggested change corrected the problem +Long is required by the mvc addon as well as a side note this also exposes the fact that Roos are assumed to always be Long when in some cases they may be something else +Will create another ticket for that. +The Alternatives section mentions sun packages which is valid +But I dont know the solution in advance +Thanks a lot David +The trunk patch depends on the patch in +Fix for was not to cache bloom filter blockIs it possible to find out what type of cached block caused the above exception +Just a tiny nit about error messagesJ and J use upper case for system procedures uses lower case for system version of the functional spec +But it shouldnt block GA since F does work +Of the three themes in Roller now none have a popup widget so I dont see that there is anything left to do here. +Make it +Also are you using thrift CQL or the interface +Fix will make next release. +CommittedThanks Roger and Carl! +Id like to deprecate or rewrite this component but instead Id just say that the Tacos implementation is better. +Blocking should cascade all te way to the adaptors +The other takes as an argument and uses the auth token +This morning I updated portal which now copies intoTomcats confCatalinalocalhost directory +I let people cross the issue and see it in their own languages +i am not sure if on and on are being fire when the app is killed by the kernel or manually by the userWe need to have a look at the apple docs +Please try Axis have tried also Axis and it behaves exactly the same as Aixis fix to proposed by Yevgeny doesnt solve this issue +Passing parameters to a servlet or JSP seems like a critical piece of functionality to me +This size is acceptable so I concur this issue may remain closed. +PranavI pushed a patch to correct this issue on Dec and it appears to have been merged on Jan +Improvement done on projectmgr also for the update note +Im also reporting this +I want to fix it like this for now to unblock other issues in semantic cleanup +These limitations need to be tested and the behavior documented as part of picking up this issueI agree that if we lift this limit on the length of VARCHAR then it will look pretty silly when LONG VARCHAR is shorter than VARCHAR +is the get transform still necessary or can we just cast the keys collection? Or could we just make it a List of DK in the first place instead of RPI think it is better just make a list of DK since accepts only DK +Thanks Arun! +Oleg The issue is resolved in the trunk you can always pickup the latest sourcecode there and compile or grab the nightly buildsBut there is a discussion in the mailing list there should be a soon +Could I ask you to provide a patch and some test data that shows the performance increases +Unfortunately I dont have the ability to import files into the private network on which Im working +During the processing the files I did not see big change in memory consumption and CPU utilization +I saw that this patch was already committed as part of the Google App Engine patch so this has to be closed +it isnt allowed to extend and is also an internal interface therefore the internal usage of is you need to write an which changes the behaviour of one of the scopes you have to use the public spi like and everything else is internal and not intended to be used because it can change at any can move setConversation to but it doesnt make any difference because is internal as well and not every needs access to the conversation itselfthis internal interface just exists because the conversation is needed in observe which is a workaround to fix +alternately we need a complex way of keeping the existing classes and services and merging them into the new ones +Great thanks for the info +Huhbq +Do you have a and registered +fixed in +Does anyone know why the has been overriddenSOAP specEach SOAP header block element information item MAY have any number of character information item children +HiWhat about the hierarchy? If we have pageA in libA with pageB in libB with Will the pageB have access to or will the catalog hierarchies be completely disjoinedRegardsChristophe Cordenier. +Please read groupId then I can change the groupId to if youd just thought because sourceforge is not controlled by the author it wouldnt drop me a line if the groupId is ok and Ill upload a modified version +XSLTC will not refuse to compile a stylesheet because the +I dont think that we should enforce a coding style but we should definitely enforce an indent style No tabs +Ok I finally got my account setup +What happens if you try Xint to run eclipse on interpreter? It is going to be very slow but it would let us know whether the bug hapens because of JIT or not +Accidentally resolved +Sigh +Weve had a lot of problems in the past with conflicting versions of +Since there is Jira to fix the link in the rpms i am resolving this as invalid +Updated TitleSummary to reflect the actual issue. +I hope this helps youPlease do not hesitate to contact me if you have any further questions +Though it is still a possibility it is pretty +There are now to users listed +Hi Dimuthu GamageCan you send your sample code to me? Im interested in this bug +Jettison? If so I think this is a known issue with Jettison that I was unable to resolve +Daryn I agree +Cool thanks for fixing this one quickly +Didnt really have a chance to thoroughly test it but just wanted to check it in before going on vacation +All issues has to be filed against from now +We made this assumption but we always retrieve any context we use from the +We have betas so we can issues like this and fix them before we promote the codebase to a release +not blocking label Relative To cannot go into the first column as it is a child of the composite on the right side including the text field the three push buttons radio buttons and label +When I delivered the master test suite works for me so Im hoping the failure you see is also related to J +you can even add any intellij project files +Resolving this for that for allowing servers to deploy over SSH there is already a JIRA for that +I am not sure about since some backward compatibility with respect to Zab Pre is removedThats one reason I think we should be aiming for +I have addressed the issues found in the earlier failover urlPls note that this patch is dependent on the failover implementationthat being done as part of and will not compile withoutthe startFailover method introduced in +Yes this is going into +would reloading the graphic when the config is reloaded be enoughafaik the image is loaded on the Style object and then remains there as a or actually makes sense to me so I guess we should make sure the Style is discarded and recreated when you press the load Im open to suggestions +Eclipse makes it easy to create a local patch just targetting a few files so if you need to create a patch for something else you can create patch and subselect the few files your new changes affect +New markup +This patch properly includes them in both +Not sure about the failing test though +The patch failed these unit tests The faillure of has not relation with my testcase. +Still need a mechanism for automatically turning off CLOSUREINLINING +Not sure we are going to incorporate fix for the other one described in bytes long header ll attach header CRLF problem patch and unit test +Comparators are used in the JDK and has its own comparators package thus its reasonable to think that Comparator implementation should be reusable +Andrew Sorry I dont even know what I was writing there +We still need the Futon UI patch Benoit made in order to close this ticketBenoit is it up to dateCan you commit it +If you set the containment of parentProp to true the test because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Thanks Andreas +committed with the minor modification to javadoc per my issue comment in r +Committed it fixing as resolved. +Closing old bugs +Fixed long ago +New members and members often have no idea of what apmail is never mind where it is +FixedBe aware that we do not create EAREJBTest project for Dynamic Web projects +If you beleive that should be please and contact the program team. +The wire logging should be easy to get to its just that someone has to do it +perhaps we should set timeout higher? It seems dangerous to trust region server forever then any buggy region server will be able to make region unreachable for the clientI will look at findbugs +Thanks +for instance +Try to using the same service clientThanks DeepalThanks for your valuable commentsI achieved the transport session by using same service clientThanksAsit Achieved by using same serviceclient object through out the session +Closing this tracker issue for GA well create new tracking as needed for CP etc. +Thanks for the quick review and turnaround guys +So it takes the whole junit ant task changes some classes and put one more class to it +Thanks. +To overcome this limitation some converters and validators were created in Commons project that evaluates its parameters when the value is checked There is no variant for fvalidate but there is a mccconvertNumber and mccconvertUnfortunately by compatibility with the spec it is not possible to change the current behavior +The new patch should work without any conflicts +We do not want users renaming extension jars to support invalid connector binding + lgtm +The information that allowed direct skipping was already present in the Metadata and the chunk streams the size of a chunk the number of chunks and the total size of the binary +Can you explain what youre attempting to address here and how it benefitsThe only reason I can see moving forward with this is if were really serious about giving other transport and RPC mechanisms the same treatment as thrift +WARP is now implemented +Committed. +Thanks Jeff for contributing! +So its more correct to say that when the Groovy executing window closes the main window doesnt always get the focus back +However right now this looks like a security hole to meTouches the following filesM javaengineorgapachederbycatalogThe new check to let only the DBO turn on NATIVE authenticationM javaengineorgapachederbylocSlight rewording of an existing error message so that is can be for this checkM javatestingorgapachederbyTestingfunctionTeststestslangNew test case to verify the new behavior +Marking closed. +Allow multiple that inherit a common interface +I then moved the var setting logic to an invocation of that instead of inside the ACARGENABLE logic for all of the enable arguments not just WEBUI +No unit test were added for this one line change +Ill try to look more into this over the next few days +Why do you say this? In both cases x should be initialized to zero +Hi IsuruWhat is the advantage of this over what we already have +Patch applied. + +One of the features on the list was called me +toolsconsole revision of buildersesb in IR +I dont know much about DTDs but can we remove the FIXED attribe from the various elements? Is there another reason for it +Yeah probably lets close it for now +It is possible to create JSF portlet that doesnt configure and doesnt include Richfaces in the trunk +This patch reworks to use a stack instead of recursion +I think it gets used only in some internal function tests to make it easy to create really deep trees without having to generate so much data +That would also lead to of slots +For some magical reason i could not trigger hudson for the last attempts +Currently we use the AOP interceptors to implement the event For API like attach and detach we directly intercept and emit the notification in the code For field access we have a modified OberservedOberserver pattern to do this with as the Observer. +Ahhh now i know why host and port are the only protected variables in I think they were changed to protected because somebody wanted to access it from the subclass For reconnect of persistent variables i need persist to be protected +I committed this. +Does a given instance need more than one factory? It could just be stored as a member instead of a MapOn the other hand if the Map is static it should probably be a Concurrent Map and account for garbage collecting undeployed classes +Partial early patch not complete +Nah. +Alone at home so with the ACDC rocking and a couple of beers I wanted to give a cracker as this one +The memory buffer is cleared at the start of every serialize call meaning even if there was garbage in the buffer itd be gone by the next time serialization started. +Updated the patch to to refactor part of the code +The patch resolves this issue in a test cluster but Im going to add a unit test +Verified by Mark +I have no issues with shipping a or even a small number of CDDL xml files in source form as a part of a larger work +Modifications as described above are in r on the branch +Heres a patch with a fix +Yikes forgot to add this patch also includes because it is dependent on it +I think we need to carefully decide if we want to break backwards compatibility in this fashion +was committed to trunk with revision +Wouldnt be better to include at least Gerhards patch into the code? Correct me if Im wrong but its better to have limited functionality of list tables footnotes than noneHaving footnotes in lists is simply necessary thing when writing bigger projects documentationCheers +The default schema name is customer and if its otherwise change the schema name in The sample tables are in +I committed it to x and x +Yes I agree +Id prefer to have a get that was protected and have the data field private +If Hans objects I dont mind putting back status though +A backport dont look like itd be that hard and itd be a nice fix to have +Thanks Owen Ill try and find an apt home for those kinds of detailsAnything else anyone? Please throw them here +Like for example studio might allow for this but its not something that should seep into our API +Deprecated means still works but you should use the new way +And unneccessary codes were removed +See for further details +Closed with. +Added and fixed some style errors +As to where to put it I see many embedded applications are likely to not be able to use it at all no matter what jar without shutting downbooting +For tests the collecting finishoffer is inside s reduce method. +Pull request to update the +we use our own way adding ajax parameter to request URL +You asked what has to be done for webapp are ant tasks for it in netbeans which are used by the web project support +Where existing VLAN ranges are edited. +I think its ready to be committed +I probably should have mentioned this before but I think having accumulo shell fake working should be a requirement for dropping the appendix +so let me clear the Fix Versions and Affects Versions field +Looks like a duplicate except that applied to and this one to +covers both the extension of and removes duplicate code +If this is a confusion we can remove the after compiling and packaging the client jars +For example if a user enables an age off iterator on the master cluster for major compaction it would be nice to have that run on the slave cluster and throw old data away +So they get an own and can even have their own Extensions +resolved issues. +I mean Struts diagram never supported xhtml files +see inplaces page at new components wiki +committed +My original patch for relied on casting the parent item in the insertion event to and then using its child list to distinguish between the cases +Take into account the patch in +Dan Rosens patch applied and available in cvs. +This one works correctly +when they in the right directory you simply have to apply the patch Jacques thanks a ll look at it tonight! Thanks Erwan at +Cant commit the new images are missing from the patch +If there is no operation each will contain two txids which are the start and end of segment +Simple patchThe removed Delete constructor was added in just for this case so its safe to remove it now +I attach the modified source test attach my try to run with any the result is the same +Thanks Awdesh and othersPatch is in trunk rev +Closing resolved issues for already released versions +My project Adepends on an artifact called A +was not returning these cores +Matteo can you take a look at it if its looks ok i will close it +Eliminate this constant am I in ZK or not? Just always use ZK +it would still be possible to blacklist racks by setting the disable flag on a rack and submitting node requests for nodes under itBy which I mean it would still be possible to blacklist racks by setting the disable flag on a rack and submitting no node requests for nodes under it +The deploy destination depends on how the server were installed into are you deciding whether something is published or not? Looking at the actual files? Where did you lookAnything in the error log? anything in the server log + +And it works OK on Adobe not reproduce issue see testcase free to modifie the testcase that is shows the problem but still work outside your environment and then reopen the ticket +This diff adds a remotable to the plugins tuscany runtime so we can satrt tring to wire between applied in rev +Works fine in FOP +The CRUD Application +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +One more nice widget from the dojo foundation as a jsf componentMany thanks to Thomas Thomas for this nice contribution + +Ive refactored the execute method using eclipses extract method to simply split it up into one method per operation +I have doubt on call which we used here to rename the +It has been updated to use PMD which supposedly fixed the bug in PMD. +The patch for trunk applied cleanly to and provided is the slightly modified patch for +The smooks editor is no longer being developerd so moving to LATER as agreed with +so it at least matches up on the first digit +Thanks George! +When do you use parent context in Spring? Do you have using that +Actually JBM testsuite is ok for all tests +The zip file contain the class test file and libraries +and if the test passes with antoher jvm +If its OK with you it and Ill commit itId like to get these REST fixes into the second release candidate if possible +committed in rev +Consider this scenariopatch A was tested on day On day the following attachments are made in this order patchB Retrieving latest attachment wouldnt pass filter to be activated in When we get to patchB we need to know whether it has been tested this is done by searching for attachment Id of patchB in JIRA textDoes this answer your question +That cant be autoprovisioned +Can you elaborate more on this? What is the property you are trying to set up using the database pool wizard +Sorry I should have been a bit more clear it has to be done mvn site APT is not wired in hdfs yet it has to be done +So I consider this as a bug +Part of a bulk update all resolutions changed to done please review history to original resolution type +Ive moved it out of the release plan for now if you want to see it implemented vote for it +Do Java errors prevent service from being deployed? In any case there are preferences so that each rule can be either disabled by user or adjusted to produce warning markers instead of error markers +Is this solved? Im using Weld SE and is still included +Now instead of sharing it here in environment we should send it on as a part of +The packageis generated by ant during the proper build procedureThe files are generated by xml +Actually on a second look I think the original patch is the right thing to do +Agreed with RE the names are going to be in the form EAP version ESB standalone versionmd provided for both files +Attached the patch +on mvn namespacing for the sake of clarity and code of fcbd I think most of these issues are fixed +The best option is to simply get JWNL in a repository which is reliable +m dir and it built fine too +The feature will not be implemented in the GPD stream. +Make a subclass of +Marking closed. +It was released with the fix in +Is this still needed +Converts to useI think this is pretty close to being ready +The reason is the key is the first equality equal instance further instances just use the same key +It clearly makes this case better and thats a good thingI dont think its worth any backporting as nobody is specifically waiting for itthat I know of +I kept the text but removed the link. +I think I simply need to have on the classpath +Thanks Daniel! +Ill post something to the issue hows that? +bq +Thanks Charles +Sounds like we ought to reach out to a Java GC guru with this questionMy philosophy would be that we should pull in whatever configs we think make sense and then thoroughly test +smile +attached a patch with proposed good to me +why not just use foodfruita +Overlay doesnt work for my Layout pages with currently not supportedSo the only problem left is the font color +If you dont change the packages dont change the gav +Fuhwei I too dont have the rights to remove your attachments but that is not a prereq of reattaching +lets not add new client dependenciesI dont think this patch changes anything client sideWill work on the ICV thing this afternoon +closing issues associated with existing releases + patch looks good +Now when a workspace is being referenced the UI says JCR Workspace +patch is against i will handle rebase to removes obsolete assertion and adds call to calculate in onRemove +or master to master +Patch that restores CSS refs + +The pages you point to are versioned because those docs are considered to pertain to that version of struts +Could you be more specific about it +A CR version is forthcoming to match the settings of WFK +Without this change the containers stay around after the AM goes away +Committed revision +See the latest +Sorry mixed it with +Attachment has been added with description Input FO that is being send to has been added with description output after getting warnings on transition resolvedfixed bugs to closedfixed +Multiple screen shots +Still need to patch the windows create script for the same issue +It should have some form of both displays fine on my Fedora Gnome +Closing tickets +My problem is with version control and when I am manually editing the file I want to be sure that what it is in there it is actually beingdo you delete the exporter from or +Maven test case project including built WAR using have switched to and the problem still exists and I have tested all versions tillSomething is really weird register a new issue with your demo app I will try to solve it please check with the latest you able to give an ETA on vote +This is fixed in branch. +And we find duplicate docs across shards in this result set for the page of results being shown +committed. +Can we get this to the series? +Disjunctions return sum over their subscorers +That does not appear to fix the problem +Or use a Sorted collection in the first place +Thanks. +JarekCan you post this patch at for usThanks +bulk move of all unresolved issues from to +I will add this and also add a test for it +Patch committed to trunk. +Besides seems that we also need to change to load correct hash tables because some hash tables may not be needed by a +This may not be a bug but I do not think the message is very userfriendly +Made it a blocker +Please let me know the preferred approach. +Changes applied to both head and branch. +applied to trunk as to trunk as to trunk as patches applied +Please consider +is it or just machine etc +Connect to a server leave the connection open and if it sends a message route the message otherwise keep the socket alive and keep waitingWe know the system can do syncronous +Amar had commented that it is not required to mark the job status as failed before raising the exception +Tested it works +We do not have guarantees for the data to survive crashes if it has not been even if another client has seen it +So all issues not directly assigned to an AS release are being closed +I verified that the compilation errors no longer occur using the attached repro +I think an inline option for gadgets is important for the future of any based application +I cant find in cocoontrunkblocks +Christian it seems that Im only able to add comments to the wiki page +Im quite sure that the qualifier doesnt trigger wire by name or am I wrong? Does the Resource annotation work here +Ideally well be able to share much of the checksum code thats currently used on the client +The current version of JIRA is at build number but the supplied backup file was for build number +Updated patch after commit of said that I spend some more time thinking about the way transaction managers are handled with the conclusion that we need further changes +Patch integrated to trunk and +Are we able to do this easily now +Oops I meant to say referrer tracking instead of category feedsThe bug is preventing referrer tracking from working for blogs that are +Can you be a bit more specific about the problems you are experiencing +I moved Christians issue to and resolved it +We will want to make sure that Jenkins concurs that it works +I think we have to fill fields for the multisearcher as well which is annoying because the multisearcher will have to let the indexsearcher know to do it or we may do it for no reason +Just adjust the classpath to the HEAD of ARQTDBCOREIRI or move the test information earlier it writes WARN TDB Transaction not active +New patch to address the review comments +Look at the number of comments and votes on this issue its causing a lot of users pain and so is a prime candidate for a workaround +Other rpcs should be at a reasonable size +Please find attached a new version of the patch based on this approach +I will upload the patch soon on which fixes this bug +Thanks for taking a look at this Charles I owe you a beer. +entity java files and derby ddl script to use to recreate this in recent code +I just committed this + was released some days ago +I didnt mean is bad its just not up to time and compared for instance to our business solutions hard to handle +Need to use to convert it to global ref +No support for this JDO syntax present +Preliminary patch for review while I continue testing +Programmatic access to the file system is typically done using Hadoop client software +Tested with build and I am still experiencing the same behavior +Early dev builds will be availble in a couple of am certain that the new version will address this problem +patch on branch. +Yeah I see your point +When do you plan to release a version of the Studio that can work with the last version of? Moreover do you know which among the old vesions of works properly with the M? Thanks +Providing a working patch is what matters +I have a fix for this defect +For history I originally ran into that with a couchapp push linked to before couchapp had the atomic write mode +After setting up a in the it looks like failure listeners are being called +Niklas I am about to close this issue but here is the script that I used to verify that the issue has been rectifiedproject top setup provider DATANUCLEUS database GOOGLEAPPENGINE applicationId opentemperamententity class +bq +And Ive decreased the priority of this issue to minor +Committed to trunk with revision +The issue now is that depends on and +Using any of the current versions of CXF Im not seeing a file generated anywhere +Craig thanks for your comments +bq +Im sorry I realized that the attached project contained a confusing error the linked directories also existed as real ones +IIRC something happened with the mini mr dirs between and which broke the test +so does this mean for you will be removing all the special handling all over the code related to borders +Please let me know if anyone has objectionsThanks +Given were in beta its also the right time to attempt such a change +Im kinda of the mind that it should be one per connection +Problem related with corrupted dependencies +From issue seems to have been fixed in the latest AxisJava trunk +It seems related to +The point is not that the tags be able to generate every possibleoption in the spec just that anything that the tags render will pass validationagainst that spec +Mark Id commit this quick fix for now so that we solve the bug and maybe we can start discussing about a new collection management API on a different issue +I guess lets figure out where things are going differently +It is ready for commitThank RavinderI had a look at the patch and in my opinion it is not quite ready for commitMy initial comments The drop table statements in the setUp method fails +This patch corrects that and changes it to not cache response to OPTIONS +Best regards by Richard. +should we put this in as wellThe ship looks sailed unless we want to reroll for this do you meanCommitted to the backport isnt trivial in any case so leaving the ticket open for that +in JBDS H Fedora +In iOS the fullscreen panel does not resize to be fullscreen when the phone is rotated to lanscape orientation +lets integrate that in due to currentI wrote a Mina GPB codec +Heres a patch which addresses the issue +The module has no dependencies besides one on junit marked as provided but its part of a large multimodule build where most of the other modules would enjoy having an for a testcase +So when can we expect the new jQuery library to be embedded +modify to test partial state saving +Same way to handle the unimplemented functions Not supported functions throw exception All methods that report supporting capabilities of database functions in return true or false +Ok OK I got your problem +Adding release note component +The JSP page that crashes is for the zh localeIt seems that svn diff breaks the encodingAny idea how I can generate a patch with svn that preserve encoding +Is having a comma in an object with no attribute like that technically valid json? If so I would say jsonsstable should handle that case and skip it +This fix belongs with the issue that fixed delete and reload Im going to fix it thereYes of course it is just hard for me to split up the patch because it is all needed for the tests to be green and I really want to give you a patch fitting on top of a certain revision where all tests are green if you add the patch +Please post all your comments there too +The bisocket is already documented +I have a similar issue with an OTA schema +Since DCM is always something very commonly used in CMSynergy SCM +There is no technical possibility to put the editing of wiki markup outside the rich text editor +Please reopen if I shouldnt have done this +ThanksWaiting a bit reactions on dev ml for adopting or not a new distribution directory treeI will ping you when its decided +Completed initial implementation. +I think this is releated to the and how it loads the Logging framework to be usedThe best option would be to write a and explicitly set SLFJ but this is not possible with the currently used version as it does not export +Noticed it has been committed to remember to forward port to and to and trunk fixed +but who? and why +Id be curious to see updated said the optimization suggested here may still be a good idea +Patch applied to at revision +However the current Tajo does not support a internal mechanism to submit a distributed query for sampling without disclosing query executions +I think we should disable this test +I see no thanks I will find myself +The entity argument for that call seems to be a plain here the question is why this hasnt been set to by the +then started loading the table again to prove the patch worked where there was a mix of old and new style encodings +Hey Ankit thanks for the update youre code looks good to me +If we change requirerelative in to decode the uri before passing it to the File methods my case seems to work fine +massaging the regex after looking at some real checked into trunk +Wed like to avoid this so itd be great if this is committed soon +updated to stdout updated patch from Chukwa home dir +Latest path has fixed all the whitespace issues and moved completely to a UID and GID based backend +Yes I had that case but somehow accidentally removed in the final patchThe case of a path with a filter is an interesting one +There is a reason why were not doing this by default already Its the contract of the Ordered interface +resource file replaced in revision its OK +I asked Matz and he said that he would like the same behaviour as you consider it +It currently uses a java Tidy impl under the hood but isnt named in case that needs to change in the future +Its a decision nested in pragmatism +I added temporary name set to detect such name conflicts to Table and classes +move resolved issues to so we can generate release notes +Im trying to check in a bunch of eclipse +Forrestbot and forrestdoc are related and do their jobs well +Use batch commitlog mode if you want to be sure not to lose data on unclean shutdown +If an admin is sophisticated enough to enable auth they should also be able to turn on SSL +Venkat +wait ms and run again +But thats another discussion +minor optimize the number of copies when concatenating blocks +It would be good to throw readable error message +You know reading that Google thread link it seems the problem is with gtest and they attempted to patch by including a guest version with the latest protobuf +Test that attempts to reproduce the issue +This has been completed +I think that if we want the timer service to be generic we need to provide some kind of factoryserializer service for each kind of thing we want to time rather than trying to be all things to all people +Guys we really need to address this before shipping +Especially forbrw its strange that it does not impact any unit test +It extracts an id from a FS it indexes the document text and organization entity annotations +Attaching log +I left some detailed comments for you on the PR +I like the general idea here +Fixed in r +I updated all the source code applied the patch and rebuilt GeronimoThen I deploy the +Need more have done a setup on clean host and Management server +I think the root cause of this close sequence of bookkeeperthe correct close sequence should be close the bookie client to error out all pending bookie requests and after bookie client is close all following request would be rejected +The threashold should be a strategy and I think there must be a better one than what is availableI do however think this is the final version of my changes to the ngram spell checker +Scott i agree with you should not have a special numbering +havent heard any more on this support question. +Sorry about that that node isnt required ive taken it out and re uploaded the Cagdas are you still seeing this issue +This is still at the RD stage it should actually be discussed in forums +ClausRevision you submitted only contains the throwing exception change +Seems fine to me +I think the main problem is VARNAME LIKE abcshould not automatically be transformed to use any at all +Sergey and TedI do agree that it is a great idea +I believe Snjezana would know more about this +Chris in the meantime I guess you can simply try out another server +chapter cleanup will have to be done in a future to QE for incorporated with the following exceptions Rules Terms Ill work on this for FP see what comes out of the newer docs port it back in a future +ill keep it in mind +Robert +Ill take a look +Thanks Dejan! +Verified with. +Sounds like +This caused signature validation to fail +What if only optimizetrue? Granted a commit occurs implicitly during an optimize but seems like that should be outside of the if on committrue +Sorry for the patch without fileThis patch changed also the if condition to cover more response codes +Hi ToddJust to clarify my point I dont need to modify current HDFS file system + if datum is null validate should always return falseNo a record cannot be null +I get similar results building from maven on the command line +Ill update the docs also fixes a DOSGI issue where a proxy can not be created when a distribution is loaded +Ive attached the results of running the rubyspec as well. +This probably warrants a security warning bug only affects interceptors defined as maps +It could be implemented using Serializable but for now it is not implemented at all + +With this behavior Im not able to predict the optimal value for this got it +That is we should copy the schemas messages and portTypes of the WSDL and put that into the generated directory as +Heres a fairly straightforward patch which linkifies the in to which in turn fetches the from dfs for the running job runs it through and prints it outIve had to add a to process since the jsp compiler is and doenst have some requisite packages for the same processing + will be automatically to after +I will merge the branches +We cannot change default as we need to maintain compatibility +Not required really +I will update the code accordingly +Ill commit this +Ive committed this +I have included a list of bundles I have installed v is the one that gets the error with diag +has been released closing issues. +Ran the test a few times passes +Patch looks great to me +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +all working now +It is a known issue that criteria queries do not handle collections of values +This PR updates the message. +Very interesting +JSTL functions are not exactly tags but more el to my ming templating them wont be right +I have set this option both VM still having the issue +Postponed to a later version +But separated and not joined back in will break every plugin declaration out there or the cli tool used in both the bootstrap and bootstraps will also have to be altered +try the branchesm solving this as its not reproducable anymore and the current does not seem to have the bug anymore. +Since I only looked at what was in trunk I didnt really see where it was useful but now I understandI do kind of wonder if readOp itself should be doing this just for consistencys sakeC +To be clear I think the first priority should be to detect corruption in the client +Leading to the +Need to confirm that this is in +Tx +Preparation for Gora release candidate +Hi allDo you have some news about this bugThank you +Can a Hibernate dev please look at the patch and either fix or suggest improvements +Yes sorry meant to add that +The reason for the UUID is for serdeser +Hi NicolasThank you very much for the fast fix +Resolved. +Ugh another problem globally available images arent in the google project any more theyre in and +I think so? This issue just moves the code around +Is there a particular reason why you believe it should use urnswitchyard? represents the interface element in the SCA schema so it uses the namespace defined by the SCA schema +Ah cool nice idea Ill add that to the spec +There might be parts that are even more liberal but it is not our responsibility to point those bits out +Sending jdotrunktck Sending jdotrunktck Sending jdotrunktck Sending jdotrunk Sending jdotrunk Sending jdotrunk Transmitting file data +Thanks for the input +There is really a lot of information here +Its strange that those tests only fail for me +With revision I updated one last file I think this fixes the broken links reported in the area of the web site +With a cannot use anything different +Previously the test cases had to perform the same setup +Please answer me negatively or positively +I think the way to deal with this is to have include the manifest info +Thanks for the but I think I should provide a revised patch that provides correct information! I will see what I can do about thatI think shutdowntrue must also be accompanied by a username and password if authentication is enabled correct? The examples in the Dev Guide show this +I did not look at the details I only had the impression that this is something similar +Its not likely a reverse DNS lookup of those addresses will reverse DNS thing is OK if the address is an address +I have been looking at the patch and I think it looks good to me but it is rather large and there are some unanswered questions in the code that I cannot answer so I would feel more comfortable if Vinod or Sid gave it a quick once over before I checked it inAlso it looks like some of the includes in have changed and a quick upmerge would be good for it to apply cleanly +Closing. +And he need to write a file +Attaching patch based on Geronimo branch directory +I dont migrate for the moment +Thanks guys +patch changes the cli script name +We just had a release a few days back +this usecase was taken into account when we were rewriting url handling in +But as some silly philosopher once said perfect is the enemy of the good + +Please remove the print statements in getLogger which display what is installedOther than this +Thanks for trying out the milestone and for your feedback +Have you googled about why this behavior changed from Firefox to +I tried to limit the maintenance for the deprecated methods by ultimately converting the single table objects to objects when they are used by the internal codeAll that being said you have a good point about just implementing a new class and I think that was Williams position when he wrote the initial patch +Jeff any word on a log or any console output that might help us out here +Also almost have the addendum for doing unit tests ready +Some experimental changes I was working on in must have snuck in something I forgot to dump +I just needed to set the GRAILSHOME environment variable and everything worked fine. +For now you can probably still utilize but well make sure that this issue makes it to RC +Thanks Glen +Is this still a problem in the latest release? If so please attach an example project that we can use for an integration test +Joe I tested your earlier fix after doing a ant clean and I can take a picture now without either exception occurring +In patchvwhen doing if metaserver is not ok now doesnt addTherefore rollback wont abort itself +fixed that +I gone this way as it did not required creating any java class so sort of quick and dirty way +I think we should add that resource ref and require Java Mail and Activation +fixed in mSending modulesconnectorsrcjavaorgapachegeronimoconnectoroutboundTransmitting file dataCommitted revision . +I can move El classes to the another package or move all functionality from the package to the another bundle +I looked through awtswing codeThese threads will exit when application receive WMQUIT event on Windows +Can they be more specific about apps in addition to Calendar +As discussed on the mailing list this change makes old metrics files incompatible +I agree with Daves analysisAlso I dont like the fact that the patch has a hardcoded port in the redirect URL or that it wasnt made from the Geronimo root directoryI think we should forget about the console and if the user opts to install the plugin just load up the plugin installer and do the install right then and there and then redirect the user to the example page +Fixed in branch on revision and at about the same time in the trunkFix was in Apache release version all resolvedfixed issues of already released versions of Roller. +Filed to make use once is resolvedPatch also merges with trunk +I dont see a problem with this one getting in but I was wondering about if it needs to get in as well and if this one impacts at all +I have simplified the code while retaining the algorithm used in position calculation +Patch that fixes the build failures in main that fixes the build failure in examples +Is there the same mechanism for +I think we need to make Grails constraints run with so that custom constraints dont cause the session to flush +If you see these errors again please raise a new bug for it because it is something different than the original bug report. +Still works fine for me with no errors on Ubuntu +Looks like this has gone stale cant get it to compile +But given the time I have I decided to leave it to application developersThe initial version is rudimentary +But the protocol itself moved from package to +Committed the patchDeleting javatestingorgapachederbyTestingfunctionTestsmasterDeleting javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestssuitesSending javatestingorgapachederbyTestingfunctionTeststestsderbynetSending javatestingorgapachederbyTestingfunctionTeststestsderbynetAdding javatestingorgapachederbyTestingfunctionTeststestsjunitTestsderbyNetAdding javatestingorgapachederbyTestingfunctionTeststestsjunitTestsderbyNetSending javatestingorgapachederbyTestingfunctionTeststestsjunitTestsderbyNetSending javatestingorgapachederbyTestingfunctionTestsutilTransmitting file dataCommitted revision +Added to release notes as resolved replies not being handled properly by code was superceded by the new transport code. +How do you suggest we know what is being xincluded to know whether to add that or not +either by switching to or simplifying the pattern itself +Part of it involves was the written to follow an existing implementation of any sort +Our development team has run into this issue as well +This patch further simplifies invocation handlers to only use a message id for the callback case note that also plays the role of an invocation handler +This one is a little more as there were two issues +If youre having trouble getting a repro in your own environment then Id be happy to test new patches for you +If they will thow will use local class loader +The documented incompatibility is this JIRA issue +Files generated will now not be supported inside +defer all issues to +patch applied in commit ccdbecadeaddedbThanks +I am not encountering any errors any longer +is fixed +You want to set XERCESCROOT not XERCESROOT +thats why I was torn between it and syncNot in love with that either though +Yeah that looks pretty embarrassing +I captured thread dumps +aka with default behavior +Hi +I think these feature would better based on that piece of work. +At this time it shaves seconds off of the c +Alexei the patches have been applied as expected +sorry should be enabletrue +for patch +It is quite convenient to refresh the page and return that previously selected tab +Christian please verify the patch +when a primary or secondary namenode wants to fork to run bash it requires double memory than what its currently using +Hi Jared Whats the status of this patch? Do you think you can produce a patch for review and submit it +I can apply the patch comfortablelyThank you all. +Attaching the patch i have mentioned changes in my comment earlier +Committed +Should definitely have a better name than HandlerThe Body workaround is maybe not to obvious but will of course work as Camel will prefer annotated methods +This also affects partially +Addendum removes the print +We are not resolving the port correctly though +It may just be sufficient to drop the stored node updates or even let them be processed by the event which is already in the queueAlso within the schedulers the events can be aggregated into a single list and processed instead of processing them per heartbeat +Its not right though +No longer applicable. +And is just doing thatBut I think that the issue is related with Providers in Libraries +It sounds like a fun projectCan you please provide an update regarding how far we are with this and any documentation that you may have? I will start off a separte discussion thread regarding this on the dev mailing list instead of having it over the jiraThanksRegards thanks Im also on this thread +I added test cases for empty arrays +I know when I use a new version of a container there is always some risk of changes causing issues and have to make adjustments accordingly +I open the wizard and this data source is listed +done +OK thats a different question +Fixed in rev +Im attaching a simple patch that fixed the issue for me +Oh I see see above Java on OS X and on Linux +field. +No response so I am closing this bugFeel free to reopen it if you can provide additional information. +But s engine doesnt support the creation of elementsThis patch will create the element with the new viewstate value in the server side and send it along with a PPR response +Why did you choose a SOAP +bq +Also applies to the following classes fields +Well be picking this up for RC and also for the upcoming does the cause any damage to the VM in your scenario? Id expect memory to be made available again right after that exception is being raised +If we also want to add CLI for cacheuncache is a better place for them unless cacheuncache can only be executed by Administrator +Also the add is pretty slow on Windows the following git status after the add takes the time +The reason is a removal of deprecated class that was using +Not sure how to fix this yet +Fixed +Hi Paul sorry if Im being unclear +Delete account is working fine now +Approved for +If you want to use a larger PQ create a separate instance of the giving a queue size +I thought we could use something like in the valdate plugin but obviously we need to find a way to use uiLabel rather +Both and are now redefinable in the theme +Adding patch for as well +But I dont see why that would happenI assume they are using a version from beforebq +Fine by me +I looked at and I didnt spot why we need the rewrite we doing some custom rewrite rules or how +DUP still NEW +If you dont mind me asking what are your thoughts regarding those that support duplicate configurations? If it is a true problem I say ban it outright. +here is a suggested basic patch +Moving issue to a next is ready for +I cannot reproduce this with trunk +get should always be safe to use but it does this extra step of reading a properties file sometimes +Attaching new version of the release note +Each target has a status message behind its identification showing its current state. +So we can test that after a new version of Fuse is built. + adds gettersetter for decoded stream length +Marking closed. +Alright last fix toI have tested this code a number of times including terminating all spot instances during a test and it worked quite well +Patch looks just committed this +Though to the same effect you could instead add a package that doesnt install the system scripts and then the couchdb package does +This issue is marked as a regression and the user indicated the problem case worked with but the repro program throws an NPE all the way back to so doesnt look like a regression at least with regard to the reproduction attached +I thought it was +Attachment has been added with description FO file showing the in rev +Committed revision +Whoops! Never screwed around with before apparently +So its only a matter of color in header I will take care of it and commit after +pushed +A blocker issue blocks a release makes impossible to use for a wide number of one looks specific to SDE thus affects only a minority of users +pushed +Just thought that if we are going to use JAXB instead then this might be acceptable to see the warnings and save some time +This is a reworked patch focusing only on and +So I move this to the next release to not delay did not occur could reproduce the problem with CXF and Tibco EMS +its necessary to uniquely identify rows by the system +I dont think there is a problem with the functionality of the current code just with some of the names used in the xml +I mean none of us was able to offer a prompt response to earlier inquiries and this got me focused on actually trying the feature for the first time +The way to handle this is to only add messages that match the filter to the buffer +yep thanks +all + +So in summary at insert time presume duplicate and check later at commitcheck time report if we cant get the lock +This patch adds master metrics and adds to the regionserver UI publishing of regionserver metricsI need to test moreAlso want to think if a better way of getting store level metrics than getting synchronization lock on the online regions and then iterating over all stores and per store over synchronizing on the regions store Collection +Please dont add more to WORKDIR should belong toWORKDIR is also used in so I think is the right place for the public static constant +Looks great David +Arg I guess changes dont pass +Duping this to. +an example of which part +is it in alpha? +Is it workSent from my iPhone +The +youre right the contract for the cancel listener has always been that it doesnt rewind the form as i now recallill double check the refresh listener and close this. +Wow weird timing Peter +Update attributes files to refer to new faces id and binding attributes +I realized it yesterday evening too will investigate it today +Now I feel good that in gave for Seriously do you think this is severe enough to sink +Marek thanks for the link to +This patch makes sure that the lease renewal thread exit before it shuts down the RPC proxy to +Here is remoteRepository support for pom and applied +Excellent then Im going to resolve +Thanks for this hint which looks very promising +Ok makes sense +Some debugging had slipped into the last one +Proposed changes +Will the final update sites remove the staging from the two site urls +So the only way is to override it with xml! But this did not work +I am very suspicious that this patch caused what I thought was index corruption the other day +When that issue is fixed this one will be fixed as well +Ask sysadmins +Files to remove in svnsecurityjbosssrcjavasecurityjbosssrcjavasecurityjbosssrcjavasecurityjbosssrcjavasecurityjbossetc +Verified. +think the problem is that the default certs dont have a supported cypher suite in their signature for jdk +Test failure is unrelated +Ill see if I can create something reproducible but that might be tricky for this type of problem +I am also facing the same i am trying to display default value on htmleditortext is simply disappearing for the onfocus and mouseover eventsany work around for this issue +Now can be deployed but two new jiras are opened see AND. +I can take this up if its not been paid attention +I am sure that if you workedwith the Struts developers in customizing Struts to your needs they would behappy to accommodate you +I want to get some comment from committers about the proposed change before attaching the next patch +Implies strong flexible a strategic inside part of an application +Fixed in commit +Whos the guru that can make it happen? I think theres consensus itd be fine to publish snapshots like this and sounds easyI thought that Benson and Drew resolved this at least as far as the apache repos are concernedI havent checked that programmatically just by eye. +Updated patch to address Review Board comments +After discussion with Ilya weve decided that JSF standard approach is not very safe and should not be supported +Actually jboss is not that relevant because it is not a production target container for s +Verified on JBT to remove respin label. +Pushing to +is incorrect +If we want to revisit this Id be happy to make the change but I probably wont be able to post a completed patch until late tomorrowYeah I didnt see the point of multiple states at but from I can see how having a set is a more general API +This seemed to work fine al day on windows the BRMSExcel integration is interesting as to how it works +Added default constructor for +However I was considering that fix for not for +Resolved with patch under +Thanks for letting us know +passes with addendum +I couldnt reproduce the issue using normal interfaces classes and the two rules you posted +If you find further issues please create a new jira and link it to this one. +If I hear no objections Ill make similar changes to main later this weekendOleg +getAlign could be marked as deprecated and noted in the docs +Thanks for all the helpAfaik CMS hasnt been fully properly migrated yet and I have been edited raw HTML files todayIn any case I would like to wait with migrating the CMS for a day or two because the official graduation announcement goes live today and I dont want to have any downtime during this period. +I think the documentation issues discussed in this ticket have been resolved. +We have a different coding style and implemented this in our own package so it was easier to paste some stuff inbtw why does myfaces mix coding styles +Subsequent attemptswould now be rejectedGood catches on the controller failover +I totally misunderstood what you were doing +While this is not strictly a fair statement if Roo had support for Ant it would add to the appeal that Roo has as a Spring product in being a good neighbor who always tries to work well with other technologies in a runtime environment AND in the related development support of Ant would only increase the appeal it has and speed up its adoption as an industry standard. +lgtmNit the two assertTrues should be assertEquals so you can see what the value actually was if the test for looking into this Eli +work started on this issue +Anyway I see whats going wrong and Ill take a whack at a fix +java file +Updated patch to include testsFixed issue with extents not being parsed into as resolved close if the fix will do for now reopen if you would like to work on extent +Please add setup directory which contains DB descriptors for several databases +I have concerns about this patch +Ok cool I will start on the filter cleanup +I just wanted to make sure it linked properly into all the issues we have and have had going for this +Attaching a corrected patch +Would you please try the patch? Thanks very much +Verified that tests pass with AS revision Alessio. +so Id rather store it in the crawldatum md for now +Verified in CR +Resolving as fixed +I will update the release docs with Psite once this is committed +Knowledge Agent and project integration has been for +I always run ant clean test before I commit and that works +what resource is it thats failing? In general this means its getting closed too early like perhaps its getting dereferenced before it should be +I wasnt able to reproduce this but I did make a slight change in how the method is resolved that could help +pushed a fix in develop branch. +Indeed thanks for the correction I forgot about the special treatment of volatile values +The fix looks great and the entry of the test should be removed +I believe this is in the current design as part of the crypto info block +bq + +I will create an alternate patch however with the partial results flag in the response header to compare the two methods +Fixed in trunk and Branch. +Deferring all the from interface changes didnt make it into +This is a paradigm that we use throughout the console +Jacob you want to take a look at this one +While it may have been intended for hdfs I do not see a nn limitation that would prevent it from using an alternate filesystem +Thanks. +Also when this is set to uses standard sockets instead of NIO sockets +It works for me +Regenerate patch +Do you have an idea of what is going on? Is it ok to assign to you and target for +I have a much better fix for it thanks to dmorgan and a patch is on its is a patch +Tim Can we get some updates to the demo clients also there is a function in mutateRows and mutate I assume there there to do batch updates to multi rows in one call but can not find an example on how to setup rowBatches +Verified. +I am not going to fix this bug for now +Ah sorry +Excuse my ignorance but what is a DFA +Hi Michael here is a patch for this one which changes the exact match by class to a looser match allowing for subclasses to be passed in +But Im not sure if this other info is really interesting to be shown in JIRA gadget +is very stable +With a few classes that should be doable +classpath if output folder isnt allowed for source folders +fixed +bq +Java replication currently logs its actions quite extensively +I was then able to add my new column +Attached new patch files +And dont forget that thou shall not clean up files +Done for some examples for calendar +I dont think the thread safety of the test code is relevant here as it is never invoked by more than a single thread +Closing all tickets +Ran derbynetclientmats with no errors +Rounding is one issue but the inconsistent precision is another one +Gurinder RandhawaHere is some more information about that bugIt appears randomly on ANY overridden field used in left half of a currently there is only way to avoid potential prod issues related to this problem you must get rid of ALL overriden getterssetters from your domain model +Used a static field need to revisit later for a better fix +The patch stores the registry into after pipeline is invoked +This is very likely a of the artificial separation of generics and annotations in Aegis +trunk This also added the test to and +New patch that corrects typos in some variable names and documentation +On trunk I wasnt able to reproduce +I continue to oppose any NOTICE requirementDo we have reason to believe that there are necessary claims that either the ASF or our licensees will need to license depending on how this protocol is usedThe RTMP license may grant us and our licensees permission that we dont need +If resources are not free then RM will not schedule more work until the NM heartbeat tells the RM about freshly freed up resources +reproducable try to insert space amp and delete it with bakspace take a look at resultamp +I guess this redundantises if each listener gets its own dedicated thread +Closing issue +Andrew is this done? +I believe they all come from the apache pom parent so we should doublecheck that were still in line with themSent from my phone please pardon the typos and brevity +What ever the value is it is made absolute in localizeTask +I meant to check your +sysEOL patch Use system eol string instead of hard coded in +greenbuild commit +I just committed this +Ah right. +As such we can definitely fix it so that it doesnt barf when you try to run it on a project +Sounds good but is it good idea to commit code that fails the features validation +Sorry for the inconvenience! +you dont get a in the classes folderRobAlex and I reproduced so would really like to know what you see +Nothing much found though +If you want this to be fixed sooner than that patches are always resolved issues for already released versions +Please make sure bugs are marked as bugs not feature applied on branch in JBT +Its illegal in Java to change the return type of a method that is inherited +Ill have another look at it before didnt want to start messing around with changing the path of in Jetty at this stage and wasnt able to figure it out by looking at the to self read in from sourceforge and from svn and couldnt reproduce in either the svn one I just changed the context name in to geoservercite and run from Ive renamed to and drop it into tomcats worked with and without the service if you can think of a way to reproduce just tell or may be this just dissapeared +I think we should cut a branch starting w Roberts latest patch +Not everybody which uses Tapestry uses portlets so it doesnt make much sense to put support for it in core +Since these Copyright statements are just for the sample code this should be a benign operationKevin +Only Springs supports this currentlyOGNL and the EL do not +according to comment fixed in +At utilization is x faster than +Moving to CR +Other blocks must be finalized + +Created a patched version of which is now copied into thirdparty before the testsuite is executed along with the other patched jars for Hypersonic and Remoting +Change drools version in the to demonstrate performance difference +It must have been fixed in in Developer StudioVersion Build id Build date +Pushed this change. +Just run mvn install with either or +Other related changes on top of +Applied at r thanks +Did I perhaps check a box when I created my account that says CC license for updates or something? +Is there any other rationale for your suggestion +Hi JesseI am not sure if we can lobby the RI maintainers to fix their codeI think what we can do right now is to keep our behavior as same as RIsHi AllAnyone else have any options about this issue? Can we lobby the RImaintainers to fix their code? RegardsRay Chen +Verified on JBDS. +You can change +I have not moved the part files to specific directories yet +Resolved for a while +that is the problemthe namenode is automaticaly coming out of the safemode without forcing it out +The corresponding JIRA is +Ive checked this with people around and only few have the same I suppose its caused by some FF plugin or s see if deploying JIRA will help this +So if a previous one returned hasMore as false may be the next one do not do any work and just return false +This should be enabled now please verify and report back so we can close this ticket +Removed synchronized and added volatile in. +I suspect they are a result of a theme problemWhen I delete my search index and then run with DEBUG turned on I see this in my logsDEBUG schedule Starting scheduled index operation DEBUG doRun Reindexining entire siteDEBUG doRun Entries to index DEBUG doRun Indexed entry PM newpostwhatrsquosDEBUG doRun Indexed entry PM testpostwithwindowsliveDEBUG doRun Indexed entry PM testingappDEBUG doRun Indexed entry PM anotherINFO doRun Completed rebuilding index for all users in secsYou need to have DEBUG on for to seedoRun messagesDo you ever get that Completed rebuilding index message? Dave +Ah thanks Ben for the patchI have though maybe the remove button should only be visible when the route has been stopped firstThis ensures its a way step to remove a route +Closing resolved issues for already released versions +Whether it generates an empty jar or an emptydirectory is not important but the ant command for making a directorywas shorter +Will set back to Closed Done in a moment +Committed to branch and trunk +patch attached if its not obvious that its a bug i can try to create some test case that will show the bug with the old lucene a lot Robert +Should be assigned to me after fix without resolving +I dont see any errors during startupdeployment in the studio console viewThere must be some errors in the console view +I think were borrowing trouble if we try to implement one for this +Change the configuration key for +Thanks Manula +I got it working on her machine using this attachedHopefully this helps you as wellBen has been committed to the trunk and is included in the tag +If you have specific concerns please report them so that well try to fixSimone +Transitioning to waiting for not already do this? Or am I missing something else? +Heres Ryans patch with a unit test +The full server publish is and defaults to incremental publishing only +sorry +Also this scheme of config prefixing is going to be painful for operatorsI meant suffixing +however substantially less probable than the walog issueI am not sure its less probable I think it depends on how frequently compactions are running +The memory leak occurs even if I remove all entries from +I made this change locally and tigris works fine +defer to +Thanks +These will both only handle groovy code +There is no problem with having an eclipse running in java but an app server running in java or a project using java compliance targets +However if and when we scale to using a larger number of threads having a manage is definitely a better option than managing explicitly +Would be good confirmation to see if that ruby client test is working now +IIOBE thrown when running this test case +So rise new request ni wiki for new components which will be selects +However that is also not the primary use case for us +Still need ted to confirm that soap wars will work with +I also think that it should be a warning +Please come to a concensus and make a decision +Thanks sofar +Thanks Thejas +Should we commit this one? I understand that we have an agreement here is that right +Granted we have been moving more towards static only properties for config so it may not be a big deal now but i dont think it is something to be done for +Can you update your patch to apply without the hunk failures? Tests will not pass for me locally with the current patch +Use and the is correct because on the there is a debug line +Item is closed now that it has passed tests. +I havent had time to look any further at this but from the testing I did I could not find any issues on the broker side +The external script might consume lots of memory and might not get killed even if the is killed +Closing resolved issues. +It definitely seems to be a issue +. +this new module looks very niceI like the new test +Woops sorry wrong bug ignore me on this one than this I think we just need some more documentation around Groovy Mock +This does not build +Removed any overlapping +Thanks Bobby +It changes the Http Status code returned when the host in a http request cannot be found +thanks Ivan! Committed revision . +was it fixed? I still dont see any changes in +resolved in on. +It would be handy to have if it was ever needed for regression or specialized testing of that environment +Would it be feasible to provide a more helpful error message for people that didnt read +I verified mvn clean install +committed into svn seems to me incomplete for some configuration such as a runtime one +It is fixed now +Attached patch uses the method introduced in. +It would be great if this could be tested as it really does encapsulate quite a few issues which are quite appropriately summarized as solrindex issues +Hey Shaz can you explain this one? I would totally kill this one off but am a bit confused as to what needs to be doneThanks +Installed plugins check my maven repository and find under the jars folder and in the velocityjars folder +The one from the parent just needs to order by the joining foreign key +What is exact version of AS ? +Pls see the comment above and verify on that scenario and then assign to Bolek based on what you find +Seperated out patch for wsa end statements updated test cases and added additional test case +I included the test thinking that can also be resolved at the same time +Is some one looking at this +I dont want to start an packaging argument here +is this really a blocker for the release +A timing diagram twas added to hat shows the interaction between a thread requesting a target thread to suspend and the target thread itself +This is an attempt to fix this test bug +If we add another argument we cannot build against so leaving this unfixed for now with the workaround that users manually deploy probably would be the better option +Ill close this now and we can reopen if any similar issues are seen. +Logged In YES useridThis problem also exist when using +Could you do a for tabs and change them to spacesIt would be really great if you could update the tests so that it checks that progress events are firing and also add a test that ensures that abort works +Thank you Jongood call removing the catch blocks from the test. +So do the script in. +Attaching the patch have done manual test for split boundary conditions +Applied patch in rThanks for the report and the patch! +Add changes to folder structure to move files around +Thats also the primary goal for this issue although we should approach it incrementally as you suggested +I believe you need to build it from source +Attached +Waiting on commit of the avro rewrite +In the end the requires a +Nope but given the number of votes thats probably the next one well implement +Is there a chance you try running your code against that version? It seems there were some breaking API changes that require recompilation against the newer version from our side +Can this be closed as a dup of or are the two different issues + for the review Eli +Better +thanks Andy +Notice it has a table to enter the fixed it so that a valid project is created with m going to raise a separate linked JIRA to enhance the wizard dialog to allow the extra archetype properties to be be in build +Kristian and MikeThanks for both of your inputs on this issue +NicolasYour patch is in trunk at rWorks well on forms and menus +perhaps consider generating the hashcode from the token rather than the key +The invoker was in part specified by a annotation which contained an Invoker URL based on localhost +Definitely an example would be goodI am sure it is something stupid on our part +sql I was thinking I will go ahead and create the release note for the current behavior and we can discuss what the error message should look like so it is not confusing +LGTM +Exactly as I mentioned failures are not reproducible +Adding to Portlet Productization will be assigned to new Portal intern when he will start managing portal demo. +So no bug hereIts really no big deal its just that I had to look around a bit to find out why this message was popping up +Validation is stillincomplete validation rules connected to containers will not be used +Lucene Nutch Hadoop already use so its accepted +Patch attached +But in one run hit for the vms and the sunsun vms +Indeed the issue why I bothered raising the JIRA was because the crash was vague and the OOM does not stand out compared to lines of Block Pool shutting down etc that got printed in the logsLook at my log snippet I posted +Merged to. + +I think the intent behind the default codec is to encode the result of any expression embedded in a GSP not to encode everything generated by any tag +Paul can you rename the so it matches the folder nameCurrently it is It should be You can use this to test commit access +I actually follow this instruction to the letter and get no errors so i presume everything is ok +Doesnt do this already in separate log file +The historyLevel property is removed from the DB +Hi Team Have you find any solution for this issue or is there any other way to achieve this aboveThanks think the issue here is he wants to write to a local file and choose the delim +Cheers Chris +Ben that indeed fixes the problem on Windows XP SP in FF IE IE +the patch looks good. +I just dont do enough JSF work any moreIve also attached a CLI script that will set up the domain running the applicationIf you need to use state saving then the issue gets a little thornier but its doable +This issue is pretty old but I remember the fix was done in all branches since so and has the fix +springide not being updated for M +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +But is this possible to rearrange this issue to the different module +inc content type association but Add a content type detector which will be used to detect if the file has JSP content or not due not to conflict with PHP Development Tool Validation adds a simple content describer class to be used with +just need install Compatibility standard C Libraries as shown on the dont remember that we were going to support Fedora I tried it yesterday and it is mozilla issue it just refuse starting for me under RHEL Desktop without Compatibility C package +Attached new draft patches +Commit bbaebcaabbbb and following. +Changed the security level field +Moving responsibility for line breaks to a separate format should remove this inconsistency +SVN diff to apply this ve been working on this somewhat more and Ill be ready to apply another patch over the weekend so hold off applying this and Ill attach another patch shortly +This cannot be implemented and Im really bored with this discussion +NickThe app that I posted was created with a release version of AIR +if we decide hibernate entity manager is a right place for this functionality we just get rid of in other case should be as a first easiest way was +Closed again now properly recorded against Fix Version as well +Async client is no works due to incompatibilities introduced in +Should the setting be enabled on the module or the controller +Left the flush on client side at MB number of edits should be a good bit smaller now we do things like count the BU row and size of BUBO when summing to see if weve hit the flush boundary +Roy Fielding author of the RFC says that GET requests with body can be constructed +It actually doesnt happen on all the login screens +It might be worth creating a base class such as with the necessary setters +Is it possible to make it work with +Tested on too +sh files are packages with dos encoding +No issues seen verified by running sysinfoproceedfalsedeprecationoffprint for the patch Kathey and Dag +Update the the previous patch works just ignore this oneSorry for the inconvenient. +BTW the property is strictly related to jaxws stack implementation and I dont expect any access +trunk Tooling as one of the effected components +Would appreciate a review +WDYT +Could you please elaborate a bit and maybe provide a short test case for itUnfortunately the attached zip file seems to be broken +really REALLY stale comment collision +VanWere the capabilities of Justins work integrated into the current translator +Zip for the html and png files +There are no current replacements for its functionality and implicitly deprecating it would be very expensive +I changed group namespace of my artifacts half through but left one depenency to the old binary which of cause had a POM missing the concerned scope elements +Not a bug in test case would help make this concrete +In my opinion this is unavoidable for large scale systemsBut I agree with Koji that these workarounds should be configurable and probably +Added a fix seems some of the paths had extra spaces in them +This fix will be in ER. +Is there any intention of fixing this in AS +How repeatable is the bug in your test machine +I dont see a JBOSSUDPPORT in the current cvs copies only those JBOSSUDPPORT ones +As an administrator I can not do code audits to ensure that everyone is using the level they are supposed to it worries me that clients may automatically be choosing . +To reproduce go on Sonar Way profile search for rules containing name edit the parameter of Class names should comply with a naming convention rule edit the parameter of Constant names should comply with a naming convention rule you should end up with a messy page +For example the earear goal documentation is not going to list this flag if its nested under a Configuration class +Ted Thanks for running through this +Slava please remember to mark as new and noteworthyGot an Examplescreenshot? +Thsi application is too big RMI is in myself dont just usehere i have read that it is because of active passive will try maybe it can solve my problem but i afraid that it will produce problem with multithread application +Patch applied +Your patch will just cause NPEs when you hit the system wide urls +The website now has a breadcrump navigation +Committed +Fixing an issue with path in the patch +Yes it is AS +I can not find how many MemoryPrimarySecondaryVolumesSnapshots etc +However we have another view called Project Archives which allows fileset matching when creating archives and allows you to pull in from anywhere in the workspace or on the system +Could you repost the patch for the Negative? It looks like a diff but theres nothing to diff +I would also add a test case where the case expression appears in the where clause along with a partition predicate and a subquery and join +Verified with EAP ER. +I think even if someone were to be using continuations it shouldnt break them as there are only a single issue and license change between the old version and this +Roger that Thanks a lot Juergen +on the patch +The same may be done for +Thanks. +Emmanuel had his fingers in that part of Seam already for he probably can do this easily +The patch lowercases scheme and authority in the ctor so the string op only happens once per Key and not once per equalshashcode check +Is this a blocker? I was wondering if I could modify my fix to consider other URI schemas and resubmit the patch in order to move forward +Thanks Cos! +No it is only at compile time we can find the files either when the view is created or when the query referencing the view isexecuted +Please evaluate +Closing all tickets +Another approach is to mark splits as uncombinable only when necessary +Committed revision . + Remove and code that depends on +Unit test illustrating forgot to mention that I am using version +fixed by +Alejandro is for the patch and so am I +TedThanks for the same patch and submit for Hadoop QAAll these testcases are running fine in the latest precommit build +I just committed this +Im on vacation for weeks so giving this to Andrew +A patch that works against trunk +I could possible split the dialog in two to achieve something similar but I need to this a major feature or we can make it optional +Fix for also solves this problem. +If a user knows enough to disable loading the state and that fixes the problem they can clear the peers table manually. +Thanks +Sorry Ive only glanced at the patch but it would be better to make the contrib module vaidya instead of since the hadoop is implied +If nobody has any objections I will go ahead and commit this tonight +this should work with Axis +Any update on this? +In order to use another version of groovy you must explicitly remove the groovy classpath container and add the version of groovy that you now that is being used you will not get any conflicts when trying to use a different version of asm or are still problems here +Rebased +Committed to branch rThanks Mikhail for staying on top of this +contribsurround what to do? has both queryparser and queries maybe should stay as isAny objections to doing the first part which has no pkg naming changes +test application displays items from mysql data base +Updated RELEASE NOTES to reflect first top level project release of KARAF. +Thanks for the extra info +passed +Please note that the indexer checks first whether the artifactmetadata already exists in the index before indexing the given artifactmetadata to assure that there will be no duplicates +Ill address both of these issues and then submit the patch again +Reopening and linking to another path management ticket +I broke indy stuff recently +We have a number of WAR projects that include other war projects as dependencies +release starting so mark this as fixedIf not please open a new issue +It doesnt behave consistently so you have to right click again and again until it works +bin has been updated to match new api in and trunk +Please mark it as resolved if it is resolved +Thanks Owen! +incorporated Zhengs to trunk +These clients are not authenticated in teh connection phase with the stomplets +ear because when I tried again this morning I discovered that there was an error in the file and the +Actually its pretty easy to unit test this bug +Are you sure it stops executing tests? either in the project or in other projects +Dont we need fully qualified to handle multiple namenodes anywayThe GC should create normalized relative paths for comparison +Venkat +Committed +Feel free to if you have additional doc changes to suggestMarking as wont fix as the original requested functional change will not be made. +Thanks for the clarification +Of course dont call ladies with lao prefix +I just committed this +I reopened the ticket with INFRA +Committed the change in empty file symantecs +A tag has been added to the allowing for the association of a custom or to be associated with a flow registry +The attached file contains the OBR description of the Shell TUI bundle +This is because version is null +Patch applied with thanksNo need to post patches for such things +It might be done by marking feature included in Richfaces Feature as optional +Updated annotation for Credentials to include HDFS as well +bulk close of all resolved issues prior to release. +Does anybody else in the community disagree with Todds suggestion +It triggers a build and move the status to patch available +Solr does already have a configurable max and the default is Can anyone supply more information on whether or not this is something that can be enhanced in Lucene for me this is a very important query and since I have well over a million documents I will never be able to issue this query +This is a duplicate of where I proposed another unsatisfactory patch and determined that Richards suggestion wouldnt work +Must items documented there have been applied +Ill commit this momentarily +Trigger happy fingers +Thanks Jim. +We cannot just move the call further down since the dispatcher initialization requires certain other attributes of the transport to be initialized first +Bugittaa Pahasti as Oleg Alexeyev said already before +JeffreyCan you rebase your patch? Thanks for answering +Thanks +The linked issues needs to be backported to the Seam to branch +If you still get a hang could you please submit a reproducible script +Now locking should not occur in that anymore +Intallation document is not proper +heres a slightly better profiling this is a good solution thanks much +OK if youre happy that the other entities work can you Verify this issue is fixed and Ill create a separate issue just for hex entities +Persisting the txid in the DN disks actually has another nice property for clusters if you accidentally restart the NN from an old snapshot of the filesystem state the can refuse to connect or refuse to process deletions +What do others think +the comment on using rather than to pass coprocessor information is deprecated +AG scriptpatch changes for script to work with Mavenized a lot for fixing those issues Alejandro +Updating to mark it duplicate +See the version of + is not binary compatible with the groovy version of your plugin +I couldnt be arsed to code the regexp for but it would be good to have that in there +There is a duration flag iirc passed to Main +retrying hudson +added logic to and implementations to validate that repo ID and URL are present before constructing an instance +When I disable fusionreactor then there is no faulty header and otherwise there is +Here is a patch to add the method and the related unit test +For all other the standard policy of is dir sep and is quote is used +bq +If youre still having problems with declaring XML namespaces or manipulating namespaces with the parser please address them to the user list. +So packet can not be sent before the ack for packet is receivedThis diagram assumes that application writes faster than packets get sent to the network +will be shortly. +Fixed as of revision +But we also have risk of data correctnessWhyWhen region in opening it will replay log edits and flush if two regionservers do opening region together could we ensure data correctness +I just committed the code which fixed the systest errors when turn off the flag by setting the keep alive property and to be falseJust two quick question I did not find the code in the which set the keepalive check interval +Another place where Ive seen this is hadoop logs pointing to initialization timeThe same format is now being generated by HOD not only on JT TT and client side but also on NN and DN which seems to work fine but is that ok Mahadev +Does this warrant a release +htaccess file +Hi ManishPlease configure correctly in and make sure the same configurations are cumming into classpath +Still keep static s for backward compatibility though it is already marked as deprecate in +please provide a runnable test case thanks +committed with minor cleanupI did have to revert the system tests to use COPP so nosetests would pass +Might be that if the is missing and you have an ICC base color space then it is OK +If we had a defaultValue property on the Column annotation hibernate could use the current dialect to add the default value at the database level +Shwetha Can you specify the details of the template you are using +I took a peek at YDH and it looks like the solution used there is to defer the sleepSIGKILL into a new thread +Wont this nuke the svn md for the and merges? svn propget svnmergeinfohadoopcommonbrancheshadoopcommonbranches +I attached the rules file driver code and also I zipped the entire project if that makes it files are and and file captures the issue of having queries in a rule file +will commit if the tests +Hi Harsh I have patch ready for this issue and submitted the same +Ugghhhh +splitlog and then have a filter that skips all dot prefixed dirs and files +Fred can you check that the commit fixes the issue you reported? basically I added a condition to filter changes on Java elements that are not in a jaxrs in trunk +Please reopen if you disagree. +With the recent addition of rich text support to this issue is less critical. +We were under the impression that the mapred were paramount to binary compatibility but since this is not going to be the case for the mapreduce that the mapreduce would be preserved +If a user enters empty filter conditions in a filter dialog those conditions are ignored and not accounted +Agent log attached +Is there an update site I can use to upgrade the plugin to the snapshot version you mentioned +Includes all my proposed changes to make the retrieval of a random mirror URL for the download of the SDK binaries possible +configure prefixhomeuserfolderA JAVAPREFIXhomeuserfolderB PYPREFIXhomeuserfolderCThis JIRA issue needs to be closed +Fix committed to trunk and x branch to suppress the generated list of external references for product and extras sites +Why not Gift +As for revision r this test does not crashes +Is this true for you as well? If so it is a bug +In one of our cluster XXM was set for testing GC OPTS +Change type from improvement to OTHERUNKNOWN agent support mean that there would be single skinning selector for all unsupported browsersIs then there a real need to specify agent unknown selector in CSS +it requires to implement additional netbeans +Currently ive big problem because my project use maven filtering in and other webapp files +distlib +Resolving issue artifacts have versions and you can deploy different versions to different environments and undeploy a version from one environment and deploy a newolder version +The archiver code +I load a collection of job entries form the database +and +That shud take care of the issue that you mention +Thanks! danielsh Ive notified regarding this +This flies in the face of the current lifecycle design +Need a sample project to reproduce this. +Reopened to fix several small issues related to compilation on windows +should probably be or and there needs to be a check on the predicate to see if it covers the entire row and if so suppress the tombstone but if not return the empty slice +There were two problems surrounding this bug Indexing wasnt working properly nothing was ever being indexed Search execution wasnt working properly the code was using a removed lucene method and also wasnt executing properly unless some features were chosenIndexing was fixed in revs Search was fixes in revs Leaving this open because I dont plan on back porting it to hopefully someone else will feel inclined to do so +Hey Lahriu now you have commit rights can you sort this +Tried this on Linux Ubuntu +Marking as wont fix since data source support has been removed from Struts + +Ill try to get some time in but I mightnot make it in time for the release +Arquillian AS issues are tracked in AS JIRA not the ARQ one +Patch contains minor fix allowing spaces in test suite dependencies list +But about DB authorization we should to create a new option to know that information +Everything works for me up to the deploy to repo which I didnt test +Thanks. +bq +Created subtask to add the wiki page for the new component +Amareshwari looked at my job file and noticed that my setting wastrueinstead ofyesMy bad +Im not so great with SVN so Im sure how to merge these into one patch so any branches will have to run first and then +Thanks + +Thank you +I think I now ran into some similar problem with embedded classes +Thats right +its getting late for mecheers Roland +I need to set ACTIVEWQHOME and ACTIVEMQ update by the end of the day +As a first task +the patched version contains an increased version number this shouldnt be wondering why dependendency to is removed wrong license header in new java and js file usage of yui using the deprecated path yui instead of yui artifact description no longer matches the scope of the artifact will fix an commit olis patch some changes s +After installation test case will run in localhost by default +I know this could be done with a separate proxy but in a small node cluster this is one more thing to manage and in the small clusters the NN load is too low to matterIn a remote cluster the local user can access the Namenode via its external name if the cluster security settings are set up to allow this +Hamlet does have some unique capabilities but it seems to me that it need not be and will certainly become a maintenance headache later +Opened to track that +Ive added an exclude to the JCA compliance test runner for JCA and will follow up with OASIS to get the test to revision I updated the JCAA contributions in the Tuscany repo to OASIS svn r +Is it OK to create the patch against a branch or do you want it for trunk +test case case the problems are still present even with enabled +Looks like the affected class has changed slightly in s an updated patch +Moving out of +I was told this was to properly generate the table of contents +when vms are in running state UI should not show scale UP Virtual Machine action button for hypervisor on which scaleup is not supported scale up is not supported on KVM not supported for system vms on xen How about when vms are in stopped state? Should UI show or not show Scale Up action for hypervisor on which scaleup is not supported? Jessica Scale up is supported for all Hypervisors when vm is stopped +Administration Roles +I was thinking that one might want something like how many numbers of success can I expect in at least cases based on a binomial sample which would be an interval of integers instead of doubles +Well if it is of any use for someone +Makes sense +nice catch Kannan +The credentials have to be sent for each the webapp prefers a cookie enbale allow session creation on the chain and you will get back a cookie +Attaching more logging from reconnect attempts +The license is not restricted to personal and home computers so the documentation does not need to be updated. + +for does the tentative fix contain modifications to the in? as far as i saw it was onlychanging the order of the imports thus unrelated to this issue isnt it? if thats the case please revert it to keep the revisions and issue references cleanas far as the modifications to is concernedlooks ok to me at first glace +Committed +Added some additional logging and exception handling in rev +trunkCommitted revision +This issue duplicates +notdef glyph character is authorized by the PDF specification if the CID is +Let me know which ones are meant to be readonly and Ill make the necessary updates +Tests pass locally +We take the picture first time cancel to use it try to take another picture the app crashes +Srcmainjava etc +Added unit test +Updating the release notes fields +does not throw an exception in the simulator which is i +Now they all use +somehow my alpha install was borked dependency wise. +Adding instrumentation to help me narrow in on the issue +Thanks Matt for fixing this and for sharing this very nice revision number with usYou forgot to update the file and if you update that file please add your name to the list of committers + push out to +If we move these up to a higher level there will be a lot of false positivesIt would be better to move these to a new category I think this would allow finer grained control and would allow the AS to enable them by default should it wish +Right now I cant get trunk to compile + st patch +So from my point of view +Patch applied to trunk on March by Fay and I verified that the DB driver could be used with IDS when testing the query timeout featue so marking as resolved +Thats why the error we need to provide a customized ESB that can cope with these but for now its sufficient to work around the NPE +Revisiting this I dont think its actually a valid optimization if we free the backing Memory object before shipping the read buffers off over the network well segfault +another use case for mixins +created applied to a fresh copy and no bulid errors +The negative impact is it would impact the readability of the gadget code in browser +Also added a test case +As written in this issue firstly almost rewrite whole write +Thanks for all your help +Otherwise reopen this issue with a valid testcase so we can have a second look at it. +Im seeing WARN Maximum number of clients reachedinstead of the server crashing +But my current thinking on that capability is that it would probably be useful to have for all schedulers +I can write a more comprehensive test but that would require changing several fields to be more accessible +Oh I see doing this you used the ecommerce theme for the backofficeThis is why I propose a flag in the to distinguish betweenthemes to be used by different applicationsIn this way only the suitable thames will be listedAny suggestion on how this field could be namedBruno +I have committed this +The new parser is in SVNThe format string is a good idea +Updated to use a wildcard for the pig version also attaching build fixes for both branch and trunk +So I reproduced issue on wink machine with IE +Im afraid I cant reproduce the crash here +Vince those are warnings not errors and shouldnt be causing any problems +The page generation error appears to be due to the absence of X libs on the box and this is being looked into +The reflection option seems like it would be a big perf hit +Attached is a patch to perform simple C symbol validation +Mod to the Order class to make it serializable and able to contain multiple elements in the order +Ok I agree with you on the point and open a new issue for this nowSure I still want to work on adding user info to fileThe way you want to implement permissions using and the way I did it seems completely different to my mind +This is in with revision +Committed patch to resolve defect +So I think this is just the one corner case that findsNot correcting offsets keeps things simple so if possible I think we could just not doanything with iteration marks surrogates and leave otherwise to actuallyreplace the iteration mark with those we would need offsets corrections +r prevents this by making all props in the Topic and Queue resource configs required but this is definitely a bug in the Profile Service +Plus I dont have tools to edit pdf files except notepadHowever the files are part of a huge project so I hope someday to find a page that does not have anything confidential on it +to be allowed to insert USD instead of USD +It exists purely to provided a cross platform way for people to index the data necessary for the tutorialim on enhancing it in ways that could encourage people to think of it as a general purpose reusable tool +James perhaps you could provide us with a patch for the approach you think is best + has been marked as a duplicate of this bug +Minor Can you remove the comment saying auto generated stub from getters +I am not saying that it wont be fixed +Looks like was talking about the same thing +Great work Sam +I attach a patch which has only barrier fix. +We were incorrectly marking the statement trigger to fire for all the columns on the table rather than the ones identified by the CREATE TRIGGER statement +submitting patch +A browse is always going to have to be a snapshot in time and since messages only migrate on demand and it takes some time to complete the snapshot will need to be updated +Attaching revision with synthentic names in the new regression test +Can you add a test that catches the bug? This is definitely should be covered by the test suite area +This means that the bundle B will get resolved automatically during resolving A +Closing released issues. +replaced by in the test since the test only checks the logic and does not create files for fixed some comments in the test +Parent issue has been superseded by the AS domain model work +Console configuration related to deleted project should also be deleted Like java does for launch configurations + It has to be a double or float and boo only recognizes a number as a double or float if there is a decimal points out another bug +Pushed in commit eeffddbca after moving the comments about fgcp to the commit message +thanks Colin will bump the jenkins job now too. +Sure worth committing or shelving you think? Just trying to review all the old issues that havent seen activity in a year or so +Moved to please move bugs back to which you will work on in the next weeks +Resolving it as fixed +Deleting that wrong file so it cant do no more harm +Pretty sure Alex was referring to the functionality of my withi +Julien you are right +Scheduling for this upstream jira as a consequence of having moved to cxf +Except for which was expecting the full package name of +Done! Should be good now +duplicate bug. +Seems that this bug has been fixed in the CVS also +As well as other ones that arent needed any could cause some bugs because Im deleting all of them +This however would not help with cyclic redirects +For example in the case I tell my users that their messages will be there for at least a month +As soon as we get a sanity check from Andrea I Will commit the patch +If this solution does not work for you could you explain why? That will help us track down or spec a solution which does work +Well in the short term I think this is a low risk change +Thanks RichardPatch applied to moduleslunisrcmainjavajavanet at repo revision Please check that the patch was applied as you expected +Oh and it is sort of the most critical feature of the damn system +Document types for which extracting attribute values make sense should have their own parser classes. +Correct version of patch +removable dependency on refactoring in is awesome +I added page numbers section numbers name and dateSame document other than that +I have commited at r +Editing to try to get it to appear in the query +I submitted my proposal as you told me but did not get any answer so I continued my research by myself +Please let me know if you get a chance to try it. +The test cases are also checked in +Let me know which one you prefer and Ill process it outCheers CheyenneThose are great logos +Dejan any advice on why I could not get this to work? Thanks +Well plan on it being resolved and included in the next release +This functionality has been added to Python Gump +Is there a reason why you dont just add the protocols fingerprint as a field of Protocolsignature is shipped to the client side and the client does not need to know the servers finger printBut anyway I fixed the code so that the hash codes of the server protocols methods do not get computed twice in case of a cache fault +One of the nodes doesnt see a live node change because he is stuck in a leader election loopGiven that I plan on committing what I have so far so it stops blocking my other two issues +The CDGA comment describes the meaning of the parameters for the program +Axelplease pass these on to whomever requested the designCheers this was reopened by our recent issue ghost Im closing it again. +The isolated runtime is now fully committed +Hi SimonI attached a small eclipse project containing a for the patch +Perhaps you might want to just print out an error from configure and ask them to change the script +reopen to set fix version to again with correct fix version +However it makes it easy to have different users only be concerned with the columns they needAgain this is not really security more configuration +This patch also takes the liberty of improving some relevant log messagesIn addition to the automated test I also tested this manually on a secure pseudocluster and confirmed that it works as expected +This need a rework on the emailType +thanks Dario! I applied it in branch r +Thanks Brock! +Confirmed that when the syslog appender is involved rather than just outputting garbage like the file appenders do the application crashesPriority has therefore been changed from major to critical to reflect this increased impact +It looks good to me +For exampleresourcestrackerstrackerstrackers We remove the property and instead replace it with jira bugzilla sourceforgeInternally the plugin uses the corresponding jsl file to compute the For example for sourceforge the issue would be of the formissueWhat do you think? Would you have time to submit a patch for this? Fabrizio I hadnt tried using just the aid +Committed revision . +I see the last comment was more than a year back +Thanks for looking at RickUploading a refactored version to remove some redundancyMoved the parser method check to and introduced +Period +so it seems like the client is waiting for the server to send some kind of final acknowledgement of transmission +Reopen if it is still a problem +Thanks Saisai I just committed this. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +This bug has been fixed in Xalan Java +I do see that this test has a lot of capability so most of my comments are more related to utilizing that part +I think this feature is meaningless +rev and removed from component where it should never have been +Verified in +sorry for delay +But if all agree i can refactor it +Moving proposed new feature to +Sorry epic fail on my part +Sebastian the OOME you observed with this patch applied was eventually fixed in right +I think we should continue on the path of an XSLT file to help generate a reference but in response to Scotts question we still shouldnt totally eliminate the guide +This can be fixed through java args in file just need to find right system property name +Viktor please check if its regression and fix it +Checkin with following comments was using current compilation schemas collation if there is a parameter involved +I would like to raise the theoretical case that caused some thread to die that was supposed to perform maintenance for the dispatched list +Oops meant to resolve it not just add a comment +This should be fixed in the next successful CI build last time let me know if you still have +The sample project can be deployed to Karaf SMX +My recommendation is that a workaround be found for the problem in It is not acceptable situation that people cannot deploy Groovy applications to or Tomcat cleanly +Ive added a traceback writer that cleans out most of the Jython invocation crud since I initially posted this +Great! Thank you very much Thomas +Removed in the JDK branch as of revision . +James can you take a look at this? It seems like it might be fixed but perhaps you can verify +Feel free to with a patch. +closing with release +Thanks a lot for this enhancement +seems to be verified ligin is fixed revision +The problem is that this wizard is not aware or and it should pick up nearest version from available and form and it would be but not +No behaves as expected +Will commit after test passes +or press Refresh every time get the attach page that you are trying to open +At the same time we can tackle this in another jira when it turns out to actually be a problem +As in the hdfs the client can get the list of bsp nodes and send the messages directly to the bsp nodes avoiding the masterIn either case the client should be notified of any changesAny thoughts. +Usersjessesrc the formatting emmanuel +There are sample using the term watermark but using the latter approach +Since the tests are changing Transport properties and protocol settings the changes can be verified on the wire using these Test files +Minor formatting changes to the description +for +Thanks for making this update +Mark sure +Moving to CP as final fix will require not changes +bq +ok so if it is a glboal setting who is setting it today +Toinght if at all possible +I believe this would extremely useful even for shell script projects +So to apply this patch in full the following patches can be applied in order monolithic patch generated from svn trunk to make applying easier +Since this is an optimization patch Id like to see a performance improvement demonstrated +Original Anakia mode that the original Anakia style used bold for the DT headerThe Browse mode uses an indent for the DD elements which is goodThe Live site does not use either bold or indentation and the resulting lists are quite difficult to scan +Resolved +New patch +test of fixed locally in trunk while we await a new release of the Maven shared component +but the trouble I have is the lack of testCould you provide a unit test or an integration +He might be able to pick it up quite quickly. +link New Server +Ive tested with attached patch and couldnt reproduce this bug. +This works in +Are there any plans for fixing this +Patch looks good +Thanks for the github address +Fix some logic bugs in the previous patch +If it is not any of these then we dont even know how to serialize it +Thanks Knut +Patch applied fine but when I try to compile the change I see that the class is missing +Attaching the modified code to reproduce +As in lets backport +Why not just Why do we have nowWhy is it called nowI am confused about this script +Which version of Maven Release Plugin has this bugYou have indicated as the Affects versions but in the Description you talk about +Tim can you update that news link or tell me how to do itFrom the Geronimo home page choose the Development Tools link from the nav near the bottom +i would rather delete this issue though and instead get the attention. +Rejecting after reading comments commit reverted issue still issue out as part of JIRA cleanup +It may take several daysSo I am waiting for results +Can you review Dave +I changed all of the files to be db +A version with the web services and web services tests +Well commit them to if will be fixed before release +Any objections from the community on using a jQuery UI widget for this feature since Bootstrap cant support it +Agreed is useless and has been for quite some time +Removed the assignment to me until is released because it is the last SNAPSHOT dependency to resolve +We silently log it in the server log view where users can always go have a look if they suspect something is wrong +Dave awesomeIs this fix required in as well +workaround exists apply the patches attached to to this so much sooner as issues that have of date for all Tasks and Feature Requests created before +Both tags are still in one patch because one is derived form the other +ImeshWould you be able to explain what test was failed and the stack trace +I could not reproduce with this test case which I think covers the described situation +Thanks guys +and all seem to do the same thing spit out the vertex id and value in textformat +This actually made it into the release. +Brian can you show me how to recreate this issue on OSX +Ive committed +ve committed this +Yup +The chapter links are opening in the original window now +I had to switch the test back to using a plain proxy so it would pass and I could deploySo that would really explain the issue then is not intended to support connections to SSL proxies? Thats actually pretty problematic for us +Ive attached an updated fix patch to +Out of scope for Maven we have enough to deal with. +The user must be aware of the syntax of CDATA section markers and be prepared to process them +Please let me know if this looks ok +And yeah next on the plate is updating the geoserver docs and throwing together a blog post +Hi ScottFor your case the relative path is if the request is httplocalhostfoo but if you request is httplocalhostfoobar the relative path is bar Camel should not send the request to httplocalhostproxybar any more the right url should be httplocalhostproxybarbar +Hi DanielIs anyone on this bug? We need the fix for our data +I will refactor them to all be in and +no follow up so moving to beta for sounds like it should have been there anywayWhat is the workaroundn for +And due to bad design of the test this causes infinite loop of reading message and test fails with timeoutA short fix would be to change synchronization algorithm in the test so this exception wont cause infinite loop and tests will pass +Is there some sort of specification out there that tells me how to map this tree to specific XHTML elementsMore specifically what XML tags should I use to separate the emails from one another? And should the output be just a linear stream of emails or should the tree structure be included in the output as well +Thanks Owen! +Users dont really need to know which implementation when theyre trying the example and we only have one implementation now. +I got this error sometimes it looks like the s fallback converter method is not be loaded before the common toString converterBut if I run this test separately the test error goes away +If there is something that can be fixed later please just create a Jira issue and let this go +What is the fix for this +If the two continuous partitions have all elements are same then it will take one element from first partition other one will skip because it cannot write the duplicate elementHere taking the last before element in the partition which is not equal to the previously written element +also include entry logger preallocation +The patch looks good +Im happy with how the patch is and believe Ive satisified all other feedback so far so Im submitting it for proper review +introduce a new specific subclass of RTE for these situations and use it where looks good to me +The super keyword was not handled properly +Closing the defect based on the comments above. +The tests pass when I upgraded to +Hai Chuan Olivier With what is the perf of async for loops and array initializers +David welcome to TajoThanks for the comment Dmitriy +Updated the patches with line wrap fixes and import Arrays +Yeah it would be nice to have a stable maintainer for the module +I assumed that it is not supported yet +I assume thats not the case but if it is a much easier issue to solve. +The difference between yours and JBMs test is that you are starting and connect to it where JBM test suite creates standalone messaging servers +The suggestion of using marshallers was to encrypt the body of the post and not for enabling basic was supposed to be an alternative to using SSL +Out of scope for this issueThis reminds me of from Aug th to Feb th the development took monthsThis JIRA doesnt have as many algorithms as those in +fix for the date and timestamp drts +Missed this file in the last patchset +Thank you for the review Shai +Attaching new patch updating Amars offline minor comments +any idea whyor how to reproducewe do have a session listener right +We intend to only send out notification when there are changes just like the findservice notifications behavior you describe +Fixed under revision r +What version of Seam are you using +Once we know where the issues lie we can create specific JIRA tasks to track them. +Verified in beta +Werner I included a two classes and a small test case +Updated patch addressing Todds comments +We can fix stuff based on added the Apache header to the new files +to guarantee consistency btw two clients yes +Oh my +The v patch partially rejects because create was added elsewhere +I wont be able to get to a more thorough version for a couple days +Maybe it can be fixed in or even? +There is a method to check if the distributed configuration differs from the local one but it doesnt detect the changeMore other the filtered properties dont make sense +Need user to investigate and provide a patch as weve never seen this before anywhere else +examples committed to cvs +So all that is needed is the JMX url needs to be added to the etc property +But Im fine with whatever solution you propose +Patch applied + +It isnt looking too hopeful +This API is for integration like S +Committed to the trunk with revision schemas are compatible and migration issues have been resolved. +Alan If you add a method to the interface wont it break existing compiled code at runtime anyways +With HOD the Capacity Scheduler and the Fair Scheduler can we close this ticket now? +If anything the JVM charts exceed requirements. +Ive verified that I can commit to the svn repository so you can go ahead and close this issue +Theyre mostly minor and we now have a CI run against IBM that should help keep it working into the future. +Verified fix locally by running the same command as mentioned in the description +How was the tiff created +first version of the part of the patchTodo in left Column Patch handles Page wo any cores +in case of the properties file existence warning message is enough file content is not rewritten +Adds the required dependency +Fixed in SVN thanks for your input Ive added your name to the project contributors section. +Thanks for the reviews Chuan and Brandon +Added a warning message with a suggestion to use nodetool status next time +Tracing the network calls there are actually two schemas being downloaded and two dtdsI have a packaging so that the schemas can now be resolved locally the dtd resolution is slightly different so continuing to investigate +Derbyall and ran cleanly with that single file patched +Uploading the rat output + for the patch +I will try it tomorrow +The failure will happen only when there is a movement of region takes places from RS having META to another RS +Attached the file used in the above code including the resultFile contains the original attachments +Wizard is replaced by invoking Properties dialog with Project Facets page selected +How far away is the capability for ECQLAs for XCQL that works for me +the that is thrown when parsing the pom is somewhat descriptive saying which line is the culprit and tried to included some error messaging into the querycontext classes but the solution is not good +Ill take a look at it asap +Excepted Behavior The Namenode should be able to finalize the clusterT Same as T +Try it and seeHowever I do not expect the stream to fail immediately that does not seem natural to meI would expect to be able to write an empty byte array to the stream and not trigger an exception +I thought it could be interesting to put some obvious error I found but you are right simply forget it +also uses mock dir but I dont think the safe leader test does because of the hard coding to standard dirAnyway more on what Im seeing from the failswe replicate gen files we reopen the writer and then the searcher using that writer we get an index of gen the files from the searchers directory dont contain the newly replicated files just the gen index files +Tedbq +Linking these issues as the both pertain to the JMS JCA resource adapter rewrite to better handle resource issues +aop modules first in the has the desired effect. +has not failed since most recent change. +Forgot to say I am in mode +Thanks for the patch it has been applied with minor changes. +It wont affect other data types +We dont have per block type metrics in trunk because the overall cache hit percentage is a good proxy for data block cache percent +Well its weird +At that point there were likely more problematic use cases +I agree +is passingI have tried to handle the different scenarios +Fixed in revision +Uploaded a patch that adds a single whitespace character +Jay to confirm is this is a also confirm if this applies to branch as tested this on +Thanks Arup. +More manifest parsing tests including reading and writing of characters encoded with several bytes +It was my mistake by not verifying with antNow its identified that we have an issue with distributionbundle file and a fix is required +If the job submission queues had been kept separate and then a property of the job submission queue described where they attach to the resource hierarchy it might have been clearer +The previous fix caused problem with table customization +either return you to the product page or the shopping cart after the addSolving it would involve finding out why its behaving differently and then finding a fix that doesnt break anything else +This is correct I just updated the documentation +Go Will +Attached a patch for the issue +We need to manually test this patch for deadlocks +for the patch +I would suggest that if the error code passed to sendError is then perhaps the log should be a warning or lower otherwise it should be an error +It would be best to have a unit testProbably this can be done by simply adding a new test or slightly modifying an existing unit test rather than constructing an entirely new test for this bug +Issue to change Release Notes Docs Status field +Committed as r r r and rAs of r it is possible to run on the Lucy source tree withperltidy disabled so that only C code is affected and no files will bechangedOur Perl code will be tidied using ordinary commits outside of this issue. +bq +distcp initiated from and permissions turned ON data copied from to WORKS +Edison this looks like the same thing weve talked about this and working fine +server log has been attached +Theres an open source C implementation of SAML called which might be useful +looks good +Thanks a lot for the review +Or more precisely you modify the file under src and run forrest which modifies the file under build for you +It wont work for checksum +Seem work +Thanks SianPatch applied in SECURITY modulke and third party notices file at repo revision rPlease check it was applied as you expected +Since as far as I know the does not provide a generic mechanism for security and I have to support Rhino Groovy and Jython I need to get specific when it comes to security at solution for Rhino for example is to implement a and extend from for making my blacklist visible and the way I have found with Groovy would be do deal with the AST +Tim do you have some time to work on this +For example one may want to have timeout on to but no timeout on to +Heres a patch +Brian! I will check with sales represntative to find out the release schedule. +The and are now beans annotated with JMX +The verdict of the PM list is that this should not be brought back in since it will be pulled back out at somepoint anyway +Fixed by creating an file in the exports folder and removing the params from instance for now validates against schema verified +The number of region servers is reached +But thanks for the correction +how do we avoid having to pull the cached field values through a sync on every call? The field data has to be cached +We should reduce the severity of the message +Let me open another issue because we have to port at least patches +Committed to branch and the part about the HRI to trunk. +If you are happy a verified is nice +ok +They are registered earlier in the same method that creates and installs the for our application +On my PC errors appers if delay miliseconds or lower +At revision o Support for web applications o to add +Annotations are all well and but they dont cover every situation +Getting the same error +Thanks John +Dave and Filipe thank you for the advice +So basically Hibernate is assuming that it has two unidirectional relationships instead of one bidirectional +Local variables should in general be the most relevant +I did not understand that you wanted to fix the documentation for which is of course the only possibility without breaking everyones codeBest Regards API committed. +unfortunately is give the same exception I attached the project I use for testing its maven project on eclipse +We did some more testing yesterday and were able to replicate the problem fairly consistently +java Xemopt cp classlibmodulesswingbintestclasslibdeploybuildtest problem is known CFG must not merge nodes withFix will be added soon +Great to hear guys! When you are at a point where it makes sense itd be interesting to see another run of the precommit tests +This issue happens when +The and still reference the old castor jar +Correct to fixed state accidently closed as cannot reproduce before. +the other way would be to let the user life with this mismatch and say that it is his responsibility +Hi Robbie I reviewed your change of Jul +the originator +This would then be leveraged for the next stage +Wouldve added if any of these had prior tests unfortunately that isnt the case +however it doesnt seem to be available in my list of More Actions +Canceling patch to incorporate review comments +thanks +I construct a war that does not have anything in lib all the jars are geronimo dependencies in the ear +QA Closing code change per dev. +I just committed this +Sample app for reproducing bug + +sample program and gdb stack traces +You may also want to include in this list which fixes a file handle leak in on trunk +All tests pass +bq +So I will try it out for a while and see what happens +Attached a new patch to warn the user about decommissioning the +wgt resource however Im not sure why youd want to do this seeing as its included by defaultIf you were trying to test the import facility I suggest instead using some of the Widget resources available from the root of the repository +I would expect that something else is going wrong here as inheritance doesnt use a projects URL to define its identity +I wonder how many customers would be helped by just being able to turn off constraint checking initially and enabling later and are ok with taking the hit of dropping and recreating indexes for this at check time +This is likely to be useful for future schedulers well +Good job AnilWith the various ways of implementing this and the various controls one may want to have over the async stuff it seems that there should be some additional elements and attributes regarding this kind of control in the form definition +html than open it with +I assume it is a mistake that this issue was dropped from the CP release schedule as this issue does affect customers. +Excellent! That makes a strong case for this change to be included in if that happens +Youre right +that would explain it +hiWe have to do something more here pleasethanks HarshaPer jaxws spec with HTTP the call from the client side IS blocking until the response comes back +Ill find out the address on Tuesday when Im next in the officeThanks I think the IP address that is blocked is one of and or possibly both +I just committed this +Applied the piece from that deals with +This issue may no longer be relevant either if have given view key creation back to theThank you for commenting on that well check the sequence stored in the session now same as in the JSF patch +Committed. +This issue is resolved on +Find the problem after rereading the javdoc of avalon +The DSL will be improved in Camel where we make it more strict how you configure error handling interceptors etc before routes etc +Isnt obj to be nullified after response is processed so that it becomes eligible for GC? +Xavier can you look into this +The release tar ball does not contains +Is it really useful? I mean do you think final user really cares about it +patch for patch making sure it compiles with trunk +what does pretty sure mean have you verified as such +see +lgtm as well +On looking at it again I think this has been a redundant check in the first place +Setting speculative execution to false also fixes my problem with parser +field reference jpa class +Committed to trunk. +assert that argument max is not null +You should consider reporting the error to the developer communityThus I am closing this ticket as invalid +I appreciate any comments regarding this +Reviewing now +Isnt it related toCheers +Thanks againBut still not able to render the menu properlynot it appears a js errorError cm is not definedAnd looking at the cliend code I do not see any + Bug has been marked as a duplicate of this bug +There is a similar issue for the code to ensure a stream is only fetched once +Attachment has been added with description the zip file contains the SVG and some other you are running you of memoryThe example you sent is referenceing images this is approximately Mb of image data +we have to either remove the synthetic flag from these methods or do that only for abstract classes +If you find new sample problems please open a new JIRA. +such issues with errors in error log is difficult to reproduce now I just can confirm I do not see this exception in error log when run VPE +java jar dblook d jdbcderbymydb o Kubosch wroteDo you still want the database schemaYes pleaseIt would also help to know more about the data you are inserting and the access patternThe more information you provide the better +Ive had the same problem just now +Assuming it is okay. +We should have something similar +Build the SDK +Posting an updated patch +Thank you very much for your understanding Clinton +Oops meant to say branch + +Are this and expressions of the same problem? +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Closing resolved issues for already released is for AS Wildfly? +Hi KimThanks for filing this issue +I am curious how long the period is for rolling new versions of Hive at Facebook +If anyone needs to recreate the original problem but they dont have Oracle you dont need to actually run against Oracle to recreate +See also cannot run build on RHEL to +everything works fine with basic zone +that means we could remove it from futon and just have it in the test suite. +Thanks Joel for the suggestion +maybe a second in the classpath for +let me give it a try + +Ideally this would be a matter of just disabling the tx middleware once that is made configurable +In near future we might include group information too +Can we put a rush on that by applying monitary measures +But if i use enctypemultipart request type portlet not working in provide the alternate for this and +Problems is actuality +q tests pass with this patch +Kudos for solving this Edson +is not an issue +Not at all right now when reading the actual type is taken into account when writing they will be written as the new type +I have verified that tests now run successfully on both Windows and Ubuntu. +Give it a shot +How does this affect the math in the original phi accrual failure detector? Is it worth getting Paul to look into that +but I expect there is a similar solution to this as beforeAnyway this issue may be closed as wont fix +I dont think anybody would have used it so I wouldnt bother deprecating it. +Since we arent using the systemmanager identity for calling the MEJB we shouldnt run into any problems changing this user +In agreement on a general move towards for monitoring purposes but the greater question is likely to be What kind of statistics are required to be produced? Total hits or countsecond figures? What to use as a fundamental monitoring API? Jamon has a nice fast lightweight Do we then record and report on a basis as well as the full cache for a view of cache operation? +Thanks Andy +This is a legitimate use of JWS distribution options +Just checked and the sample itself works fine exceptwhen I press Content View on any URL containing a continuation +Can you check if its still there for you +and +Dinesh Im reopening this ticket to also cover image button +AttilaCan you please inspect your eclipse log to get us the exact detailled error message +Closing issue as it has been released as part of release +Obsoleted by the introduction of vnodes +And please exclude the test updates from the review request to make it easier to read +Are you saying you had this converted and tested? Ignoring the fact that this was never communicated to the rest of us how did you get around the problems below paths incorrect load fails even after basedir creations? If you have some insight here please share so we can get this resolved +It will only work for developer builds +This TCK has been superseded by the TCK. +I would like to add that over a period of time a will eventually occur even if one of the allows ping +Since it is just a deprecation warning I am going to commit the patch +Still think it should go into the debug dump +The later fix patching the client is the correct and sufficient fix for the problem that Joydeep initially reported +Also need to move Traceable to should be kept stable moving to. +close this issueThanks Michel. +Obvious now you point it out +is it possible to use a dns name with? +And even no input is provided the process wont fail +Im using Tiles which is as part of the struts showcase app +Hi JohnathanI dont have a lot of detailed knowledge about the linuxjavaleap second issue but in my case there was no kernel panic or any other kernel issues +it would be better if more javadoccomments can be added to such as the general steps for each check process and what is tested for each step +Ok let me investigate the issue here +Ive commited the change to +resolve the doc issue +I dont have a paper I am suggesting however that such pruning can even improve the result +Marking for ve updated the patch to apply cleanly against trunk +Patches is of course welcomeAnd in the past we got in touch with AHC to have it improved to allow us to configured it more easily from Camel side using bean style configuration +Way too late for +MilosI have attached trimmed source code of the ll also attach pom file try to build the project and let me know if you have problems +But this Jira is to do it in the query parser as I thought Uwe was seeming to indicate would work +We chatted about this via just to make sure it gets covered in the readme +This class does not need to be a singleton and in fact it should not be since the digestor is not thread safe. +Whether or not it makes sense the element is present in the generic web plan schema +Move transitions to +If that isnt acceptable I understand just thought I would offerI would suggest using something like Django with the app sure its a bigger application footprint but flatblocks sets up each piece of information on the page as editable for the admins of the site +lucenesolr +Lets fix it here +Attached a patch containing the file +Only queries dont work with joined subclasses +Minor changes related to JMS selector +After fixing which is marked as relatedthis error goes still get a stacktrace though if a parent pom cant be resolved or if a pom is missing a we create a new issue with a more global description like never print stacktracesif e or X not specified or this one and rename it? Closing for now since the problem this issue describes is fixed +We no longer pass default credentials for as since there are none and setting empty credentials is the same as setting credentials as the server sees it and they will disable local authentication mode if credentials are passed +Movingto Struts to consider adding support for this new feature +If we compare reads per key per second instead of just readssec that case goes away +As long as there are no cloned workspaces this error will not happen +The method signatures where checked in as part of but seem unrelated to that fix +lgtm +We use the same logic Java Persistence uses. +Attaching a patch implementing the suggestion to remove c from the bash command line thereby avoiding the only code path which can return ETXTBSY under semantics +Well probably add more content to this chapter as time goes on. +commented into branch and trunk +xxd comes with vim +So all issues not directly assigned to an AS release are being closed +Ive attached a patch which I believe resolves this issue +I got the build successful +Looks good to me as well +Good point +If no furture comments would this be committed +NikhilWe need this for rightthanks is the patch with all required changes for this JIRA +Some of the tests needed updating for the new behaviour and as part of this I realised that the is no longer addition the was installing the wrong juddi jar so this has been reverted +Can you include a sample project here +Committed for and trunk thanks Eli. +Here it is + is different then as it uses JDK and has some new deployment features such as deployment +So there wont be any more patches for this ticket unless someone requests one +We need to balance the level of info output here but I tend to agree that those should probably remain info messages +I attached a setup of SOLR configuration files to replicate the issueChange the basedir in to import the filesOnce SOLR is started go to the page and run the data handler +Joerg you coulda fixed this the answer was at your fingertips +Both and have been completed +the code was was fast! Thanks Carsten +I took representatively some single pages as documents from our to build a +It has been fixed a long time ago +This is specifically caused by an explicit case to the Logback implementation which is assumed to be present +Oopps thats my fault +Committed to trunk +Im wondering if this has anything to do with the id and scope attributes having special reserved meanings in JSP +I tried to override the render method but Im not sure its the right you have an opinion on this +Tested this patch in conjunction with WorksI would like to see this in and trunk +Bringing into without the pollution to branch and trunk thanks Stack! +Committed to trunk +The patch to fix the problemThis is a regression after commit +this is not their responsibility +It needed to create a new source foldersrctestresourceswith all the files taken from this source foldersrcmain patch add these source foldersrctestresourceswith all the files included as attachments here +Likewise the messaging in should be updated as well to indicate that aggregates might be off +None whatsoever +Felix can you check if it is ok +Logged In YES userid and are redirected totestsuiteoutputreports at the bottom of each is redirected to testsuiteoutputlog is cleared for every new testRegards Username mattmunz +With this fix Sannes original problem is solved. +See what was done to the init scripts in the parent ticket +I have committed the fix with regression tests +Included in the releases +I think this is important to look at +change reporter +No bug in axis +If we assume the Aggregator value is stored in the Zookeeper server reliably for the duration of the Job why persist it in the first place +Fixes for UNION ALL implicit type conversion upstream input file change propagation +I am working on this nowThanks Nikhil for this contribution +See for further details. +Thanks Kim +Username letiemble +might we change it to only take seconds or so? that would require changing both parameters to floats which i dont think is unreasonable +Perhaps if more performance work is needed I could check that in as a separate patch? Since we are moving from not working to working the performance is a lot better than it was before +What you propose matches what does already +Alsologgers and appenders needs to be defined in the correct order are aware of this problem but currently we are pouring all our efforts into launching version +What environment are you running in +Fixed but still needs test test case +Attached a patch forThis patch adds a connect as superuser method which statfs uses +I cannot see a place where the znode can be in another state +Thanks for the comments Suresh! Update the patch based on your comments +Looks good +Attaching patches for these two versions +I was actually waiting for you to give it a final approving look +Eric could you add a test case please +The former can probably be changed in +Thanks George my mistake +I dont see a point to add jira changes report to this plugin site +Consistency guys +Sounds like we have a bug after all. +This is an extended version of the previous patch in addition to the previous changes it includes also changes to the Protocol interface to enable more flexible fetching and status reporting introduce and in a similar model as Parse and +During release process LICENSE and NOTICE file will automatically pulled in MANIFEST directory of final jar file +Test case is attached with instructions in the Redame +Identified new dependencies in Shindig that necessitate changes to LICENSE NOTICE in that war but are used in other locations so combined LICENSE NOTICE for demo binary does not need to be to move to as was issues after move to in +deb? +Logan no worries at all thanks for confirming +Some of the servers are configured to use ports +Thank you Ivan +But I do not accept that as a solution +It also saves us from potentially scanning large libraries unnecessarilyAdmittedly though this is a lot less important than the general resolutionoptional +Sorry Olivier but we dont have any way to reproduce this issue +Change zk from to for now until is out +Russell please assign back if the previous comment is not sufficient for your needs +i reviewed the change and seems a good change to me +looks like perhaps its only hit once though and then reused +Its likely well get this into +I have downloaded the bit JBDS and repeated the same test case +I am unsure when or why this would be a factor as it appears that the utilcache defaults are to NOT use this file at all +Were in a long process of removing that +Please give this a shot again and may be you can try rich commandLink to get started +Its kind of random +Final? Wouldnt removing the SNAPSHOT be enough +This is a patch to update the error message thrown when a non CVS scm is used +bq +My plan for the future is the following Flush out the version to contain all features I originally had in version Refactor the plugin to use the Java Deployment standard interfaces so that it can work with any app Add the Weblogic specific stuff to the plugin Contribute the Java Deployment code to the CARGO Update the plugin to work with all other app servers that support the standard deployment JSR Contact the author of the APPC plugin and merge that code with my current APPC code to allow that functionality to be in one place and add my code to as Questions? You are creating the weblogic pom and I also think I need one for the BEA webservices jar that supports the appc functionality +Here is a new executable which expands on the previouslysupplied +In the meantime my workaround is to create the outside the loop and close it only after going through all the process instancesIve attached the test case the jBPM process definition file and the file +Would it be possible to split them to smaller pieces? For example we could commit the data store implementation independent of the integration to rest of the Jackrabbit core +Tests ran cleanly for me +I believe this has been resolved in the current release +The patch fixing this issue for the branch with the test case to reproduce it +This really helps clarify how Derby uses Java Locales +Same problem with +Hi Tom is in the voting process and should be available next week provided that the vote passesBRMaruan + +I will create a review board request +It should not be possible to modify a column which already has a null value to NOT NULL +This issue is fixed and released as part of release +lets the container configure the default isolation level the JPA spec describes behavior at +Spot on +Checking this in +No real changes updated to work with trunkWithout objection I think this should be added soon +The solution is probably to fix the pretty printing in or to filter whatever formating tools is doing in the first place +The purpose of is to separate these kinds of settings from the more complex logic in +Was this fixed with Lukasz changes +to reviewapply the patch +SachinThanks for having started to look into this +There is no unit test though since I really dont know how this works or why it didnt work before +This is a regression from and thus needs to be fixed before can be released +You should really officially the commit and give your valid technical reason for doing so before being too over zealous about a quick revert +For trunk I would like to address it in +Check whether the server reporting for duty is a dead one coming back alive and if so update the deadserverList and remove this server +Im sure situations around this can get complicated but this case does not seem complicated at all +Solved +Ah that makes sense +Any testcases requirements and etc +verified. +fixed typo in title and added. +Stop blaming Hibernate. +oh yes +If so lets just leave the restriction in for and push the predicate down into the scan for +Part of a bulk update all resolutions changed to done please review history to original resolution type +At the same time it adds complexityOn balance the API change here is nice because it extends something that was too limited to address your use case such that now it works for you and it also admits the possibility of others +the patch looks good +Hermes now added +Oops +Im not sure why you think this is necessary +I dont think the original patch would be easy to understandapply at this stage. +Test case which shows that the words HAMILTON do not show in the SVGcorrectly +Thats bad we need to fix that +Fixed unit test failures +Few other cosmetic changes +Attached the patch with timeout added to the patch incorporating Jing and Umas review comments +It extends +Rayees how did you create the user VM? Did you use any special network creation steps? Basic zone or advanced zone? Im guessing advanced? And how did you create the networks? Did you provide two guest networks in VM creation +Refactored the order by handling logic to ensure that unrelated items will be displayed based upon their full name +Here is the patch with commentsI think it should go in could you please create another Jira for the problems you mentioned here +I just reread the last paragraph that I wrote and feel that I need to clarify it +and this validation would NOT be included in think this should only apply to default no other serverWhat about if when the user adds a model to the VDB we prompt them toAChoose and existing JNDI nameORBCreate a new TranslatorConnector if they dont want to choose AORCBaypass these options and leave JNDI name empty +I dont believe a framework should expliticly stop you from doing something because the developers dont think its a good idea +Ive looked at line error happened in s invokeEvent function +Radim which installer did you use? universal or OS specific one +Attachment think that we cant use attached documents as data because they are not licensed under Apache license + +Attaching the diff file for +You did the Android stream you can do the docs as well +Yep! Thanks. +Ram are you going to handle the fix for this in +Thanks +Attachment has been added with description Fixed to work with latest trunk code and fixed some checkstyle has been added with description Incorrect direction latin text inside arabic transition to closed remaining resolved bugs +Hi is duplicated with this one which is targeted to finally +This is a bit tricky to get right as I ran into while investigating it +An Initial version tests coming +Does the falcon in version now support the same cluster name to be defined in both source and target definition of the feed xmlNope the same validations exist in +I have seenAny finalized for to start working? +Made some updates to the javadocs also +Thanks Alexey +Any other arrangement causes exceptions from Hibernate at runtimeRegardless the tool could use this convention to make an effort to get the mappedBy on the inverse end of the relationship +fixes the file name issuealso takes care of an urgent outstanding issue with not being able to use an alternate file system for hive scratch dir +This feature will be tabled for now +The major one is if we load all plugin classes using our own custom classloader then appserver wont have access to theses classes +I am not actually sure I can explain that right nowGregory My initial code had a bug where it did not remove the old values at all and hence it not only exploded the memstore it also churned through all every single time to see if some of them can be removed +Do you have the problem with reduces also? I reducer class might be a good option depending on what is required +I need to fix my codeBut this case must fail +The elementType is required to construct the but it wont be written to disk twice even though it is part of the key and part of the object +It seems that in the cases where the error occurs the old version is picked up. +Just let me know when to try and which build +Actually it probably makes sense to spend the time on the interactive option rather than a short term fix like Ive suggested. +Could you please the patches and include that in itI dont suppose you feel inclined to also create the same for the socialrest service btw +aar file +We want to keep these out of lib since theyre not a released binary that we want Hadoop users to use +Do a bug report to MRI for this +Thanks Klaas! +Ill address it shortly +Will we need a respin? If not the Fix Versions should be updated accordingly +Ive documented how to run the tests in the site documentation if any reviewer wants to run the tests read that and if its not clear contact me +So we are removing the notion of RS restart anyway Ill just clean that up from and commit my patch above +Thanks senaka for the patch. +the patch looks goodI just committed this +see previous is just a matter of waiting for it to be distributed +I am still not sure if we should break backwards in +I was using Xalan in Cocoon when I encountered this bug +I think that the pattern is the same +No this is not fixing +I would call tell the developer what happened when there is a fatal error a significant improvement in functionalityIm not aware of a way to use logj as a backendMichael Im not sure how that would work here I can think of several ways to do something similar and all of them are clunkier than just shipping +No prob! I will help you I am on heavy committing +it is renamed to locate The instanceDir is same as solrhome for a single core solr +There can be a link alongside the disabled link which can point to the wiki pagewell partly because it makes the more complicated but also because it doesnt seem any more straightforward from a user perspective +I ran a unit test and verified that the ugly stack traces are gone +Closing issue Oozie is released +I noticed there is a test case but it looks pretty minimal and it was unclear to me how I could add a test for this case +Scottin our case the user dn was only a ldap link and you are not allowd to bind using a link name +Remove the entity by deleting the +If you didnt want to explicitly set the charset everywhere you need it you could write a middleware that adjusts the to have a charset when none is provided. +Well you can probably set jar as a default value for null type +Ive checked with Rajesh and he said that was OK since annotations unit test failures will not block the should be something in the release notes about this being an expected failure +This would make more sense when deploying the sample source as part of a binary package or even as part of the install since the current setup requires a srcxalanc directory to contain the configureMakefile +I created an issue and fixed for . +The deployer is not invoked +well I have posted an incomplete patch +Were having some problems using metaClass to override methods in a subclass for unit testing and I think its related to this problemAttached are some unit tests that illustrate the issue +So the use of classes rather than rows was probably intentionalBut I think the flights example is good too and if it is clearer to readers thats betterAttached is and +New patch with linked up tests +Ill take a look at the grails maven plugin to see what could work instead of using grailshome or if that is set for backwards compatibility or not +We are getting this Exception also in javassist +I have implemented a common for both the quote as well as the company header +Putting in so we can discuss whether to change things +I am assigning this to Jothi since I am going to be out for the next couple of days +As always we agree to disagree +Thats a fair point +This could be a JDK version or classpath issue +Release Notes marked Not Required as this affects an internal Engineering release and is resolved in an Engineering release so no customer could have been affected by this issue. +This should not happen +Will add additionnal tracing capabilities when needed upstream +I added some logs to track +I suspect that this is Spring related +Thanks Nathan +I think dodging it by stripping out double quotes on the end would also be possible but then you get the question of what happens if the user wanted the environment variable to have quotes in it +so how about creating a separated for it +so the RM allocates the reducers before the mappers +Ill discuss that on the dev list +Thanks +Integrated on master BTW +could always just writes which the caller then needs to bulk import +Maybe we could have a wrapper shell script that does that for us and then calls the tycho build with all the proper args using upstream repo local site if theres no change and rebuilding the plugin if there is a need for it +However this would add a dependency on modulesqueries that I dont think exists yet that is the only hitch +in +Attaching most recent patch which was just committed +hmm that doesnt apply for me with or patch +If and when there is renewed interest to pick up the AntIvy stuff this can be reopened and completed +Some of the comments suggest that the app may need to be deployed to jboss to demonstrate the problem and that the problem does not occur with +Tried to look at this patch but it could no longer be applied +Then our fix will work +Ill send a message trying to get some Tapestry committers involved +opened for and. +Id rather there was a test especially for the urls Mike was seeing +Fixed +Should prob be closed. +Patch applied and also upgraded POI to latest CVS versionPlease cross check +Hopefully it gets fixed soon +I think theres a lot we can learn but theres at least challenges client supplied timestamps mean that you cant know that newer files supercede older ones the CF data model means that data for a given key in multiple sstables may need to be like the approach in almost every way except that recovery becomes more complicated than our current scenario +Ill think about how best to change the test +I think a copy should be made here too +Todd not hard at all +Patch looks good and wasCommitted revision +I found that reference counting on nodes created by is unnecessary as the document manages the memory for such nodes +No changes to the code itself except to update the paths for files +I had to merge it before I could code review +And so Martins question remains +Please open another issue in case you still see significant problems as weve just released M +that could be it +Also this class could be turned into an enum to make the code more readable +Hmm the tests are failing because of the TCP port is already bound when the is started +I just committed this +messages can be safely ignored if they reference any class in the packages. +Test failures are unrelated +Finishing operation of wizard is put to background job +Good job Lars. +added a few integration tests +Reverted back to DLQ in revision . +Sorry for the late response +Fixed for +Initial set to commit +Other things Im sure I forgot +note the patch doesnt fix the golden files yet. +At some point the remaining resources should be moved or substituted with ones provided by the common jersey moduleThe kresjersey module must be at the end removed +Adding it to the options is what makes sense to me +I will check with Bela as this was his concern +I just downloaded and applied but hadnt looked at it yet +Seems this is solved in the +I committed this +its been a while since this issue was submitted but I believe the problem is that the user doesnt have access to the installation directory of the npm install command +Should we maybe make an option for whether to quote everything or just quote reserved words +I agree with Knuts analysis +Currently used version of Scroll Wiki Exporter doesnt export this attribute +Fixed with revision of productsamplesquickstartsconf +Now up to +this is the generated doc screen shot +Thanks Ravi +Yes the bug is a result of changes in RC +Added to M thing to note althought this does not introduce new dependencies it means that the icons wont show up until after you call the BS +also see. +Thanks Stack for the review! Ive integrated the addendum into and trunk branch +The renaming of the method or introduction of new method could be done in trunk to reduce the diff of this branch with trunk +Is it correctOriginally I have got this error from test +Therefore this version supports two new optional properties in the configuration second property should be currently set to true unless the in the M repo have been fixed +Soon we will post that patch +Closed due to the release of Tiles. +Thank you +Exercising logj directly is only necessary if you want to reproduce the original problem +Heres a patch that fixes this +Ill give it a try and see if its clean and passes tests +file would be wrriten w a codec that included mvcc or it could be written with a codec that did not include mvccHmmm +This change needs to be made to too so have raised +This will speed improvements in the tomahawk TLD docs +Are we confident that the intermittent failures are problems with the tests and not with the implementation? If so just removing the tests sounds OK +In Web Projects View Right click on NewFileCss +Are you volunteering? +Thanks for the tip +Updating Fix For to Unknown on issues not targeted for attached for review +No feedback on my last ping +a should look likeI didnt mean to attach the merged patches here but cant removeem +Other related regressions already fixed by and +Its required for fix must be done for scm connection url scm developer connection url scm url and site url in fixing it we need to access to the absolute path of parent and child so we need to access to Project instead of Model in urls must be calculated with child artifactId relativePath of parent parent url and module path used for the doesnt you are correct about my expectations +And remove the git nonsense from the patch files +Patch contains fixes for and +I think the source files should be within the package and not in +Clob data will have to be placed in files and a file name could be passed as an argument to this import function +a solution can be just to set uce pass right after inline +port patch work with the latest rc patch to work with the latest rc tag +Do you want to display the nonsecure items? should be fixed in tomahawk svn head now +Yes it is +This file provides improvements regarding the naming of days on the timesheets in Project Mgr +That was the right one +It breaks back compat but frankly straddling doesnt seem worth the effort here +see last comment +Marking this issue to blockWith Dans review of the first checkin he noted that there was still work needed in the area of casting and comparisons for +Is it OK to attach that to this bug or should I open a new JIRA + +The first would be to fork the kit so there is both a community and a product kit +Please close if this works for you +It is probably even worse when compared to which should be more thanIt is also possible to walk through elements of both matricesCould you provide a patch for this +Pavel Yaschenko reviewed this issue +We require a certain number of permissions and this is one of them +Uploaded new path to address Sureshs commentsThanks +Passing by hadoopqa to make sure it doesnt break anything unexpectedly +YARN does not use configuration directly for launching an application +The results I got were intriguing in practice but as someone like Sebastian would tell you this does not represent a solution to the message growth problems with triangle closingIf this is a JIRA issue holding up the new release we can mark this wont fix +Paperwork complete reformatted docs now available in docs branch. +Fixed under revision r +I think it is always a good idea to make things more simple for the users +Shows significant test coverageFew general comments on the patch srcmainnative and a few other new files do not have Apache license +The get method thus just needs to check whether the object implements the ID interface and if so call the xml method +Test case for fix for ia similar fixes should be done for emt and agree with the proposed fix +Oops patch was intended as a contribution +Only two bugs are actually fixed between RC and +Adobe Acrobat renders this fine. +Make the patch be restricted in the scope of YARN only +failed because the storage checksum comparison in the test does not take the cache file into account +this is basically fixed +replaced also related with this +But it is a preexisting issueI did not run any manual tests yet with the old API + +Proposed patch and screenshot +The new attachment streaming should have fixed that +Downs the batch write heap default from MB to MB Adds interface with a heapSize member +menus on the context menu from the packagefile explorer and on the main toolbar +But I think this will be a special case where manual recovery and manual of are needed. +Yes AshishThats what I had in mindRewrite system would need metadata and hence it should be invoked after semantic analysis phase which would make metadata available +closed +If possible please give me a sample data set too +I think the names used in the guide are the ones that should be used +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +I cant forsee a reason why this would be a problem but the devil is in the details +Please feel free to raise the question with the mailing list or open a JIRA +Resolved using Tycho not Athena. +Needs to be rebased though +Thats how it should work +Even after adding to classpath +As a workaround Id suggest to add handle RDF files a bit differently that is continue parsing until the namespace of the current element event is different from that of RDF +Patch with changes as described +for +finds out the location of old and new jars and hence these need not be passed into the classI ran the upgrade test with and without setting property and it works as expected +Thank you Jing! +This issue has been fixed and released as part of release +Only after the RS changes to OPENING the znode of the regions belongs to the RSYes this is the case for znode states +Assign to during tidy up prior to the issue now that it has been resolved. +Any chance of getting line numbers in there so can be for sure for sure +Its just another implementation +Should I add this on the Hive wiki? Im in the process of porting the code over to the hive trunk and am planning to follow the same pattern used by the and Cassandra storage handlers +Its never been a requirement and is an unnecessary step as far as I know +Do you foresee any side effects of using instead of Is is actually recommended filter +generate wsdl definition +Closing because this no longer appears to be skipped or failing. +Theres also a fair bit of code duplication among those classes we might consider doing some refactoring there if you think its worthwhile +Hey AngelinaDid you configure the physical network adapter as a ovs bridge portif for example eth is your physical nic you need to setDEVICETYPEovsTYPEOVSBRIDGEovsbrotherwise the physical interface will not be connected to the bridge and you will lose connectivityYou can check with ovsbr that command should list your physical interfaceFeel free to ping me if you see me on IRC so i can help you directly +Added tests for the Factory +So we can determine where exactly the selectionFailure occurAs well please attach the schema for Input and OutputCheersDenis accept my possible ignorance but where would I find the console log +Andrew and Aleksey are you both happy for me to resolve this issue +Thanks Rakesh +WS is checked +It is a bit different from yours as several parts in your patch dont seem to apply to normal Groovy +This is partly caused by the fact that the programmatic API creates proxy annotations on the entity +SeeI should have a fix for that ticket later today +The patch has been appliedPlease verify. +Also no tests included here because tests were already failing which caught this +My point about the test case is that I feel that such a test should be in ZabTest because it touches that part of the code +will work on the jira this week +batch transition resolvedfixed bugs to closedfixed +The +Maybe removing the groups field from and using always the attributes map is the cleaner way so that we dont have to maintain two references for the same data +This is the easy mind you +Where do we stand? Im game to include it but its gotten a little stale me directly if you update +Further improvements can be tracked under their own. +Looks good to by Imran +The plugin needs to be ported to WTP +The version c of the patches some code to avoid merge conflicts with +Hyunsik thanks for your commentI totally agree with youSo Im going to create a new branch and try to test the patch +All tests pass w community validation +I dont think this is about right or wrong but rather about living by means instead of broken bicycle + +Testing and hopefully committing shortly +Please close the bug if alls now well +For me it looks like that the Impl should be thread safe but it isnt +Actually the same hold for the GUI Uninstaller so I reproduced the behaviour +Chris depends on how development on ISPN goesif it will be ready for releases then it could be part of WF otherwise it will slip into +So this is no bug but a problem in your FO +Build is blue +directory is still not exported neither in nor +Is there a timeout setting I can alter to make it more forgiving to installs that take a while? Its not an undue amount of time but it seems to be enough to upset Whirrjclouds +For this patch just introduce the new interface in common and then file a separate HDFS and MR jira to use the new APIThanksEli +Thanks for report SamCorrected in trunk at r +See. +As Fernando noted this one should be marked as resolved accordingly +Could you try out something like the attached patch and let us know if it has any effect on performance +Latest patch +Separate wire data types from the actual data used on clients and servers +Change committed to in JBT nightly +I had gone through the spec myself yesterday to make sure that what this jira wants is indeedKnut it is interesting that INTERSECT does not work same as EXCEPT associativity wise +Same for me go for it +Are you suggesting removing the ability to assign aliases inside a forearch or just to changeassign schemasFor consistency all relational operators should support the AS clause +I dont know what Spring did in this area but I wont use a proprietary API for such a trivial thing And I dont think Spring has a JMS implementation they probably have some wrappers delegating to an implementation +binhbase master start +The writer client writes the first entry to one bookie and crashes +However it did work in my testing +I have no errors in the logs +. +Right now Roller uses the Jakarta Commons Digester RSS classes and only support aggregation of RSS +movesrcmainjavaorgapachewicketextensionsajaxmarkuphtmlformuploadtosrcmainjavaorgapachewicketprotocolhttpservlet +It should be a minor fix to the jamon pages +It is possible to not include the entry by simply not including the key in the properties for example the property in the template +Will be replaced with new Async DSL later DONEtrunk the nabble link as I keep posting there also with updates +I suggest to change component to differenceBest which behavior is reasonable depends on point of view +It is obvious that Groovy which doesnt scale with the number of cores hurts else can we do about it +will be good to put this in the hands of some devs with +Ill review it next week +I will commit it +This is fixed by isnt it +The problem still occursThe missing class is present in lib however Spring classloader in not able to find the definition no matter how the weaverclassloader for Spring is setI tried to force Hibernate to use javassist instead of cglib to see if this is not CGLIB related issue as Spring present in WFK shades ASM inside CGLIB the Spring in EAP based Seam in does not use + patch applied. +I have attached my test case file to the reproduce the issue follow the steps Install EAP copy standalone configuration to cp rp standalone copy the attached file to configuration directory cp configuration The has already configured AMQ RA in server with command cd JBOSSHOMEbin +reopening based on mailing list discussion indicating problem is still in m not seeing this +reopen if it isnt +In my earlier debugging sessions I ignored the setter so thats propably the reason I did not find the method that was crashing until now +But its a thing and there wont be more s. +Patch with some checkstyle fixes +Attached unittest to reproduce based on from +Especially since cutting and pasting these traces into Jira breaks JIRA export +Sorry thought you had commit access +In the latter we flush and then block +There should be additional check boxes for this and debug flags on Main tab of the launch configuration panel +Touches the following filesM toolsjarReplaces with in the list of extra classes which should be pulled into +If we can get Contegix to host it so much the merrier +Thanks a lot for the reviewstesting Chris and Todd. +well if you keep changing the requirements +Learning is appropriate documentation on the with Ant tools chapter in the Hibernate tools guide i +Ill try to provide test within next wednesday +With that change I always get a response when requesting without credentials +In this case external lock manager like a zookeeper could be used but its not required thing +I didnt restart the node however because this is a production cluster +While debugging memory leaks in NN I could not get lot of tools to work with the heap dump of G especially jhat +Obviously broken blocker +If this is successful is would be great to get a separate job running for nutchgora so I will leave this open for the time being +Committed removal of unnecessary patch including test cases +updated to include the Function extensions on assistants now completely control which properties are applicable for a model object +Could be a great improvement at least for security to not have to open mysql access for all the world +Thanks Gav! +According to the most recent public review of the bridge will support custom requestresponse objects used within the provided that setRequest and setResponse are used + +Steve First off sorry for taking so long to look at this patch +Thanks. +Closing as duplicate of +Ah ok this as actually been fixed but the Jira is missing I will create one in GTNPORTAL and link that one to in Portal +Shouldnt this happen by default +bulk close of resolved issues +Let me know if you dont agree with the changes i have renamed the DCEPS to +You can ignore the warning +What is the exact request URIOleg +Moved from to just after merge of old branch into TRUNK +Working perfectlyMany thanks +Looks good thanks +keep Wed Sep UTC webappsyarn +Seems Mockito dev list will not respondreply thus we have to go with our forked packaging. +is also related to this +I am going to commit the patch in +Bobs working on some new Riftsaw examples as well so hopefully the two of you can work on this together +verified of a bulk update all resolutions changed to done please review history to original resolution type +This relative url is then used for determining what the tag and should patch also includes a new unit test to verify the new functionality +A new version of hasnt yet been released +Resubmitting the patch +Upon closer look it loosk like it might be a bug in the serializers which areincluded in the Garbage scratchpad +Close the issue if everything is fine +So even with separate classloaders it can easily mix the versionsSo my proposal is to provide a soap connector bundle that does not contain the libs in addition to the current one +on the new patch except for one small addition I would like to make to the documentation added for the configuration variable +I havent looked at enough to it +Where I disagree is with the assertion that we cannot get to it starting with a simple implementation +What you are proposing is that when using slinginclude the actual request method is always overridden unless the method attribute is setIt gets even worse when we consider limiting the value of the override to only allow GET in this case the attribute is actually unneeded on the tagOn the other hand I agree that most of the time we do an include the method should probably be overridden to be GETFinally thinking about it I wonder whether your application may not be modified such that the POST handling servlet just does what it has to do on POST and the redirects to the rendering servlet with a GET requestAny more opinions? +The gz file is for reviewing +all done. +Note that this patch only includes new tests the currentproduct already supports the functionality to change a columns default value +Simply added synchronized keyword +Spring is so well written and well tested framework that its nd to none +Fixed +This adds more permissions to and to facilitate server tracing and sysinfo when a security manager is installedOne thing that puzzles me is the fact that sysinfo lives in jar files and +is in your classpath it will load the properties file from the but if +May also need reworking when JCA transitions to dynamic registration of recovery modules +If you think this is trivial here you go +Tests failures are not related to this patch +Marking as resolved +And when you are creating continuation this rootcontinuation will be parent for all continuations for which no parent wasexplicitly specifiedI agree that storing continuations manager in each continuation is not a goodidea +So you may have to actually add declaration for into your +The fix should be to copy a line from the Block constructor + +Thanks for your patch Simone it basically looks good to me with one minor request could you make sure that if a java version higher than is used it doesnt revert back to the implementation but uses the same as for java ? This way the plugin should look the same with java regardless whether we enhance it for java or Ill investigate on what you suggested and propose a patch +I cant find a reference to it in our pom +Added in a list of outstanding required fields and renamed Default to Common think its a better term for the tab title. +After restarting the messages that I couldnt see get delivered which probably is why none show in the Queues list after the restartIf I use Time To Wait and Number of Repeats instead of CRON it works fineI dont know how to set up a unit test to test this given the need for a listener and such a long delay +Im going to close this bug as it seems to have been fixed in the latest CVS +As you know its not possible to maven +messageWill continue to investigate +Well I guess we cant do much about this then +The code should be compatible with Hadoop and perhaps even earlier although I have not tested that yet +Fixed in revision +With this patch Im getting a lot farther along in running terasort on nodes with DCE +In any case our very own Andy Clement is the driving force there so Im sure well get something that is good enough for our purposes in +We feel that it would be beneficial to the users to have access to these fixes and functionality +I do not even remember opening it +this patch work but only when you add the header to http requestI dont know yet if it should be handled by mochiweb or not +the patch that have fixed two more tests +Our internal policy is to limit developer choice with regard to project templates +Great idea do you have a patch +The reason the from the pull request is preferable is that its compromise is simply that other convenience subclasses of cannot be used directly +There are only extra options and its all documented in the Release Notes +Vera JVMTI uses its own special EM configuration so if you specify another one there is no surprise that it doesnt work +and for trunk in rev +classpath as project and also for the ivysetting path +This problem is no longer exists +These legacy commands are fading from my memory showing the wisdom of your original position on this +defer to +For Hibernate for that matter but Ill have a go +At least it tests the coordinate reprojection values +addressed comments on RB +This was committed to fix +Chunking fails not after chunk but after chunk +As discussed on the mailing list the latest maintenance release will be the +HiI attached the test codes for and forBest regards +Committed patch a to trunk with revision +When I said I meant +In order to minimize the change I introduced and so that all necessary parts just require to replace with +Patches applied +Hi SateeshI havent observed any error messages with the other template VHRegardsKiran +Fixed +This is only a shallow copy +This third version of the patch adds a bounded thread pool +Commit defebadbffdffcbba from gdchamals Github +However it passes on my local box +Also dont forget to update the regression test with your additions +Thanks +I will look into this +Yep I think the are all mixed up like you saidI did make some progress on G trying to get them to include a couple patches which make it perform somewhat better than CMS at least in my environmentAlso just did a patch tonight which will try to collapse the byte arrays together into large ones will run it overnight and get a graph +Hmm I cant reproduce this +added output obtained with new gradients of type and are supported in with the new code but this should address most of the common cases +Please note that the is scoped test so it should be excluded from the war +Unscheduling due to inactivity +Just from developers point of viewI can change it if you want me to and upload the updated patch +tableinfo for root and meta only after thatWe shall get to the bottom of your issue soonSubbu +Committed revision +If you could comment on about those notes I would appreciate it +Needs to be tested somehow +they actually go fetch exactly the data youre looking for when you instantiate the coverage +Please apply this patch to and as well +I have still not heard anything from regarding my request for a trial version of their EFT server +Agree with Hadrian this ticket should not introduce new in +What you are saying is that the view server spits out a couchdb version that it works on and we store it? Is that correct? Do I need to do anything else with this value? Dip +Will log another JIRA for that work +In the context of HA BN should be using same principal as the primary namenode because on a failover it becomes the primary +No I was thinking of editing the embedded file or add it on deployment thus your war configuration via the cli would be persisted. +i have the same problemand i believe the reason is that the excel just using the system default decoding and i havent find a way to change that +it tests for this and is currently disabled until this issue is resolved +resolved because of it is notit will be greate if you can describe mechanism how such messages are generated by xilrunneras I think should be possibility to suppress such messages +Thank you Ivan +RobertI believe I found a reasonable solution to the problem +Cool +No response for a long time so Im closing the issue +This patch also includes for +Do we have a test case for this? Whats the setup here? How much heap +Has this bug been fixed +When publishing there should be a publish event which should contain information copying the file failed for some reason +I will try to find time to set up some data for this as well but I dont know that I will have time +This workaround helps alot but the should implement +Reopening for reasons in previous comment. +This lock is a bit expensive one to hold as it used for ops like flushes and one can call any arbitrary code in the preXXX hook + with the modifications described above +Abandoned Amazon Linux although Chef recognises it as such java and other cookbooks dont and Chef doesnt have the option to fallback to centos +But we have been telling users that they must do lifecycle management for themselves at the moment based on the fact that we will iterate over has during the next full scan +Verified the problem and fixed in revision +Daniel is there a test for the commands you mention will fail if we fix this issue? Seems like theres a bug there to begin with and we should fix both this classpath thing and the commands that are otherwise inconsistent +If by magic its done for great +I hope to be able to release a beta version within a few weeks +Ive the parsing using rather than +Verified in the Install guide DocIn the following line is removedThis command will build the following debian packages +Thanks! +fixesremove has a typo missing a e +The problem still exists with both Hibernate and Hibernate retested this Bug with Hibernate and now both statements work fine +Forward paths in the source already appear to be rejected +It was originally nested in Condition in +Both of them go up only one level on the stack where the presence of the adapter class would want them to go up two levelsI do not see any current way to break the required dependency on without biting the bullet and switching to JDK logging unconditionally +Yeah its a better idea to do like JMS endpoint +Please see the patch tweaking it just a little to bring it up to date with. +It wouldnt sense to make this change without changing every other method that returns an STL type the library was intended to be used with the same runtime dll so this would be an extremely large change +Right but you dont want to replay the commit log if someone has changed the partitioner do you? That would be Bad +What is f instead of +Sending this back to Xalan land +A rose by any other name +One less thing +The simplest solution to this problem is to enable the caching registry interceptor but have the validity period default to the same as the default registry cache if not specifiedThis will also alleviate issues caused by other circumstances where the SI is not cached +Thanks Josh for your review! I have updated the patch to include comments for the added methods +I am new to this functionality so I will probably have some questions +Your design does not seem to consider the possibility of multiple concurrent logs which you may want to have for federation +Thanks very much for fixing this Laura! +This will make our own debugging life a lot easier if there are ever any problems with this +The rootEntity in this JIRA and System in the related post are both classes +That seems to do the jobThe patch includes changes both for and the strong would be really nice to get rid of this ancient dependency +Subtasks were added so reopening and postponing fix version +Committed to tez branch +My initial impressions are that handles everything right now so the immediate need for isnt great +Only components that are not compatible w both Luna and Kepler should prevent their installation into Luna. +Im closing this as wont fix but please reopen if you feel strongly about it. +However note that IDF would prevent a bunch of hits on the from causing too hot a hotspot in the heat map +Im obviously missing something +Thanks Andrea Ill notify the user +Join requires a schema to be mandatory be present on the right side and it is used to determine the number of null fieldscolumns in nullTupleAs its a two way join we use nullBag instead of an Array of nullBag +Else it didnt invoke indexing +Noi dont have any other property files and i am not setting any thing for symparameterI have only one property file in file under lib folder. +hi thomaswhat do you think about migration steps for existing workspacesgreetsclaus +Sure +non of the comments have been related to this! to the code +The userBasic install is put in the default datadir location already specified in the class. +Looks good no comments! +If it finds an exception occurred it throws that exceptionIf no exceptions occur then the main thread executes the remaining actions such as COMMIT +The problem was related to Apache Tomcat build and deployment procedure requires Apache Tomcat. +Erik +Not that I know what Im doing really +I am not sure how do we resolve the native library dependency for downstream projectsMy observation in Hive seems like this HADOOPHOME environment variable needs to be set on Windows in order to run Hive unit tests because Hadoop needs to find Once this change is in native library will be requiredHive may need to change its to add HADOOPHOMElibnative +I reuploaded the patch file and moved the Effective level out of theThanks Todd for the comment +Committed +Quick reminder The release is just around the corner now scheduled for public release tomorrow +Plus weve had enough job submission issues lately that affect oozie which are painfully hard to debug +Hi Ning i uploaded a patch for +Thanks SaschaYour patch is in trunk at r +Taking this off the table for now +New complete patch with testcase in +Any reason why the connectors module is not part of the build +As a result we are closingthis bugIf you can reproduce this bug against a currently maintained version of EJB please feel free to reopen this bug against that versionThank you for reporting this bug and we are sorry it could not be fixed. +Patch instead of +A logins firstB logins again +Edward if you or someone else wants to create a patch to make exported yaml more thats fine but otherwise Im closing as. +project and checked +And just so that the comments from review board are recorded in the jira please make sure to update the bugs field with the jira numberI have also already reviewed the java part of this patch but I dont see any of my previous comments reflected +I added Doug and Tom to the list of watchers +Until you tell us to replaceyour existing site with the svnpubsubbed version from the CMSnothing will change on your live site +At the moment we just start them if theyre not already started and leave them running if couch stops +A short summary about the quickstarts webservicewsproxybasic is a straightforward invocation of an external webservice via the proxy +is there any you have methods that are mapped to the same endpoint +In discussion we decided to remove the DNS proxy +An issue regarding was identified and JIRA is logged +This is a bit unpleasant but we dont really have a choice if is going to load internal classes from the persistence unit classloader +Will commit momentarily +This isnt the case with the current clear +Mona Current is going to be discarded and trunk is going to be released as +You should replace all instances of searchPaths with searchPaths in your code +renamed the looks good overall just a nit Lets rename +Revised description +I only moved the Term collectors +the patch as well +Tested for bad service category service name and protocol as well as tested for valid EPR with no contract +Sorry Ive attached againPlease review it +Thanks Raghu. +Since the patches committed changes Derby behavior I dont think they are suitable for so closing +We tried and it fixed the shutdown issue +and attach the comments there +is still in the voting phase +Please post your additional comments to the new issue. +I can send a copy of the database if needed but I would rather not attach it to the bug will send in private email +and are now done +Vijay Parthasarathy is writing a hbase data loader +Question is it still afcting or not +When it reads it gets back data with associated checksum +submitting for the quick review +The easiest workaround for this is which is also +Added testing criteria for correctincorrect password formats characters and flow. +Please check if the following labels are still translated correctly after they are merged into Thanks some Common labels in to Ecommerce in the same committed in rev +I apologize for the delay +Ive moved the creation of the image writer outside the page iterationThe result is functionally the same rendering only works first time for thefirst page of the +If the check is disable the result after snapshot restore is still a valid state of the system +Fix Version should be should be resolved with the upgrade to Spring Framework once it has been released +filterINFO The system cannot find any method in the class that supports HEAD +hopefully we can revamp the stats system to not be dependent on any tick value +Trunk code at Apache for currently contains a working SSL transport +Accepting for attaching a unified patch? +Now the CSS resource references are public static finalSee for an explanation how to add a reference to a bundle without having access to the instance +Yes disallowing duplicate annotations only for Runtime retention policy seems correct to me +Was this intentional? Also a test might be nice +Seems the time to wait for the server is too short +Please find the code changes to axis kernel +Especially in the light of the button Revert to parent definition the icon and the message that is send out is wrong +New patches against patch should go into thanks +RamaIn factI run this case many timesif the meta and user regions are in the same server it will ok +Applied in rev and trunk rev +Im gonna send the pull req for that now +on v +Verified by myself. +My fault I failed to change the build process correctly I will replace the current file with one that has the gant jar in it +Not able to apply +just apply all patches and away you go +in the clientid seems very restrictive +Rebased +Because the toUnicode doesnt contain the character code and that the Encoding entry is overrode the returns an invalid CIDI have commented the toUnicode initialization and the PDF is considered as a valid PDFAI need to read the PDF Reference about the CID lookup to propose a fix +The codec uses in the implementation + to trunk. +Do you still need me to spend some time looking into it? please +bq + can be used as for override if required +Is one old +This is quite interesting if you want to expose some administrative information using some kind of webb app or even better implement a DSML gateway inside the server +Ive verified that it is really in the Camel bundle but not in my application bundle to which it needs to be added manually +patch looks good to me thanks for the updating +Just a quick note this change will affect the rpm build as well which I presume is intended since this is linked with BIGTOP +I tried to look for a quiet option but could not find one +I just checked and yes this bug is fixed in +In addition missing artifacts should not be treated as a failure +Either the servlet must be mapped somewhere else in or GPD needs to support authentication +This is not important right now. +Did you try to disable that property +The error you encountered was due to a bug in the patch +AM +Ah cool +But the existing code handled more complex situations so I figured Id preserve that featureThe happens during initialization so there is no real need to make it static +In any case will ask userhive +Issue is not fixed +WAR EJB and EAR is supported out of the box +I also have been silently annoyed by some of the large commits which have made kind of gratuitous changes to code Ive written making methods and fields static is one example I dont favor +I started with the templates from Marcel and improved them in many points in just being one of there any plans to integrate in into the core templates andor to improve the core templates +As such those differences should remain and we will not ship examples that rely on missingunsupported capabilities in the embedded version +ToddI like the idea of using objects to insulate coprocessors from API changes theyre not usingDhrubaAgree that it would be better to tackle that in a separate JIRA and keep all of the in sync +The userguide is seriously outdated after all the recent refactoring +Anyway this should work well with the new interfacesRegarding the proxy I didnt notice this before it but generating only proxies by seems as an elegant solution +suggested patch also removes unused expressionFactory field +Closing as not a bug but a feature +And of course its fine if I close it and relaunch it from the toolbar button +we can move to Apache CXF we shouldnt need Jersey right +Im running on with maven and cmake version +Ok that should make it so this code can just go into the +Two nits sp Synchronziing Recommend replacing the three test methods with a loop from +I dont have time to write up the changes but will revisit this after release +Ive not looked at the code but if it works it should be in SVN +So what I did I creaated subset of +this introduces regressions on WindowsDoug not sure what you mean by this +You always know where to find the jars or the wars +Possibly relatedBut not duplicates +I assumed this was setup with a default of dasras but was also configurable in the +I resolved this issue using a zip file +This depends on this +I hoping to make the above improvements after I get some review comments +Thanks Mike +Reset password via email is implemented need to add option on realm to enabledisable admins to do this they would have to edit the user Add RESETPASSWORD to required actions Change status to ACTIONSREQUIRED Set a temporary password and send to userIf totp is lost the admin should Remove totp configuration Add CONFIGURETOTP to required actions Change status to ACTIONSREQUIREDATM theres certainly more stuff we can do like having security need to add link on the login form but waiting for Villiam to merge template engine work a form for inserting usernameemail to send reset mailAlso the reset password form shouldnt require the original password +So all issues not directly assigned to an AS release are being closed +Also the total number of Services is very small no need to support pagination at list and list current implementation does not handle pagination parameter at all so count returned is correct +Im not sure if I think its a good idea to expose this level of access and I need some more time to think about the best way to do it +If I do this Ill have to also work on migration +The last patch breaks ant test in general in any case +ClausI do have to admit configuration can be a real PITA in +This parent task is just waiting on BB support now so assigning this to your Lorin as the BB subtask is also assigned to you +Please roll back to in the meantimeRegards cannot find a way to fix this Im thinking about recompiling against the latest and release it +Since you have a testcase I assume it passes the test? Case closed thenD +Thanks Kevin +Grant AIUI +That was addressed via +This fixed all the unit test failures +Infact over the week end gave a long run +If the option is set to true lazy decompression is disabled Unexpected errors are treated as corruptions and the reader indicates no more dataThis allows graceful termination of the read when there are corruptionsThe default value of is falseTested this by using rcfilecat on a file with a corrupt block of data +I just committed this +Latest patch add support for passing along connection ssf on shadow connection creation for clusterTBD need to determine how to get ssf value to the replicated connection +FinitoSending coresrcmainjavaorgapachemahoutcftasteimplneighborhoodTransmitting file dataCommitted revision . +b. +Btw tested with H and Oracle the build worked fine +Trunk has been updated with those changes. +IMO the validwhen and requiredif validations could be replaced byadding a javascript validator using Rhino +That should solve this problem. +We have been running this in production for days now and rss increased only insignificantly by MB a day +The https is globally enabled for Tomcat using the tomcat https tricks generate key with keygen and link it up with edits in CATALINAconf +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Tried running ant test with latest patch but I see a lot test timeouts +I am strongly in favor of defaults that are as flexible and stable as possible +Ill test it today +Note the solution is using a specific exception but does not log any errors +If youd like to take a shot and submit it as a patch it would be reviewed +it would be nice to address the exceptions that Rakesh brought up and set the causing exception as the cause rather than embedding in the messageas for the disconnect problem +I recommend adding asserts for all input params that are expected to be then enableassertions for the tests so you can easily see what is expected and what should be handled +All changes are against trunkBest Regards see file in the archive Ive attached to the GuysI am pretty new to this site +Forgot to resolve this. +This also includes providing decent titles and abstract for styles as they are used to make up the WMS file contains the parts that need changing in particular service descriptions have been renamed fictuous contact information filled layer title and abstracts reworked all styles have been reformatted name title and abstract filled and a reference to the proper schema location has been added as that file also contains a sf replacement that is an actual DEM colored using the raster as a result of your work the sfdem layer is gone as well as the associated layer group +add to firewall authorize ingress + I think thats a good cobertura or emma a better solution +You wouldnt be able to boot without it so its hardly missing +If I disable the firewall then all is good +Let me know if you need any more supporting detail +Command line property overrides work in but do NOT work in. +Looks like it in now. +Closing resolved issues for already released versions +Prashant can you please attach the full console output with the verbose flag +Patch taking the nd approach of promoting columns to supercolumns when SC is null in a super CF +BTW you may want to also look into I think were doing some dumb things with the way we use epoll etc +OK this fix is in with revision +Here is a patch which takes care of this for putcopy and Mahadev +Thanks for committing Im glad the patch was useful +Complete replacement for all files in the package +Do you know where it is missingI think they are mostly bogus and should be removed +This solves distributed grouping in steps +Tim applying the patch is tough lots of diffs from trunk +To pass true to Tomcat you need to set the JAVAOPTS environment variable +Norman are you saying this is fixed in? I dont understand +Getting exceptions from BIRT is fun +Needs to be fixed in too +but I didnt see in the commit list the test which tries with sedanametimeout +Added a try finally block in we resolve this issue now Ram +You might want to decorate your one line fix with the comment block which explains why INTERNALRECOMPILEREQUEST is being requested When revoking a privilege from a Table we need to invalidate all refering to it +Im moving this to to allow for more discussion on whether or not it needs to be implemented at all +Agree this would be very useful +Thats fine if youre using the ODMG book as the master +Created to handle that later +Your testcase fails because Boo only recognizes extensions when the type declaring them is imported importing the namespace of that type isnt enough +so regarding this question Is bigtop moving towards using groovy as a core element of the core package build and deploying? I guess the answer is Yes +It just dont understand why jaxb try to get its information on the interface instead of the implmentation of the webservicesI wont use generic interface and put all my method definition service by service +AndreasRotation is off on this one also for meIs it just my problem? Or did we break something somewhere since you fixed thisOnce we get this corrected lets throw it into testinputrendering along with the PNG file it generates so that we catch any such problemsThanks +I havent used at all +kindly review +Any chance you can make a reproducible test project for this +I cannot seem to get the zip attached but if someone is willing to look at this I will send them the attachment +There were numerous changes since including security fixes there are even new features to be added to the editor which we can do later +Your suggestion for how one with a index can stillguarantee uniqueness for keys seems like a very good ideaHowever one will also need to do something to index creation since atraditional index will not flag if there are duplicate keys +I have no complaint about removing +Cleaned up the to produce +Im experiencing exactly the same issue on my side with Seam Final andAnybody got any ideas? +Can someone help me +Thanks Lars +The BLOBCLOBs transaction may be committed or its connection is closed +Need the exact rules for when a cast is impossible +CC hung on on Dec running on SUSEL I still see hangs on linux ia interpreter +Transitioning to closed. +Ive found the problem with the test +I believe theres a permission issue that prevents creating files in the tmp directory and thus failing the Cargo build +incorporated Hairongs comment +A parents profiles are not inherited however the effect of activating that profile is inherited +Here is an example +Had to some javadocs so the imports dont refer to the server classes +I was returning early in if null data inside is when I should have carried on to trip over the get of region location from the AM memory +dfs write packet size is currently a global tunable turn this into a patch due to potential for buffer bloat OOM on the data node end of the systemWill update patch with an as well. +There were some custom jars so I tried it with the official build and the result was the same +I added what page currently looks like +Thanks. +behavior of Java and CPP Testframework should be the same +I marked this derbybackportreject not because of any risk but this issue has just test changes made during the diagnosis of this issue +Thanks Bjrn! +All Id hoped to do was to make fewer jars but thats not necessary for nowI notice you did not supply a change with the patch +The types array in a protocol definition could come textually after the messages but the types must be processed before the messages and +Ah OK closing again +I am running the full regression suite and willcommit this patch soon +I tested the dependency available on the Alfresco Maven repo and integration tests works correctlySo I think that we have solved this last issueIm committing the latest changes +What you found here is actually a missing feature that I didnt get the conversion working with constructors +There are some more fixes for this issue some of them are already in upstream for few pull request was sent and there are still some fixes in work +Ill try to drive this forward +I dont think its ok +If you want to access a PS on an arbitrary connection pulled from a pool then youre going to need some way of dealing with a connection that doesnt have that PS stored +this is obviously duplicate to that is resolved in our latest build +fix a bug in the committed this +I set the component to Unknown and the affected version to +So it might be better just to leave them alone +What is the problem? You still have to walk through all the classes anyway and it should be super easy to iterate all selected classes and check if it is a class or not +Please review this patch +You need to open a new ticket with an description +Applied the patch locally and it built fine +Patch doesnt work though right? And regards config +I think the typical usecase will be to ignore them and just look up properties +Cant it be reworked into separate java program +I think this is much cleaner than before +I have just committed a fix +Also note that if I make the viewScoped bean session scoped then everything also works as expected +Added more comments to pseudo codes +bulk defer to +Planning to make the config property work +isnt available in the standard Eclipse provided update sites? p should figure out to get those +Thanks Chris patch looks good +After changing the settings to asmxml and republishing to Sonar the sources of MXML files still are not shown +This patch is for both branches and trunkIts been tested against bothThis fix selects Tomcat to run by default if Jetty is not selected to install +Yes +bq +Not a bug in Camel +Pavel I did not receive any comments when you the issue +Please also test Hama Pipes examples if possibleSee csrcmainnativeexamplesThanks +Let me know if you need any help testing or closing the issueSending javaengineorgapachederbyimplsqlexecuteTransmitting file dataCommitted revision +Punting to Im pretty sure its in the same class of errors that need some more heavy handed reworks in the master +We have decided to only make the layer view listen to maps +Given that this issue has gone stale now I am resolving it. +I like the idea of using the config +I think the assemblers could work good in combination with the +thanks tested it works Thank you +Frank thanks for doing more work on this +The backport will be figured out later +I dont know how much that applies to other problems +aldermen comes close but as a culturally specific concept paer needs an explanation rather than a gloss. +Thanks Joerg +Please reopenrefile if you encounter this again. +bulk defer of featurestaskswishes from to +Now the examples are part of tomahawk. +Thats on Bradens plate this week as well though +I think this is a duplicate of which is a duplicate of +yes +But as you might be knowing only regression fixes go into release branches +Thanks Lars +Patch for supporting Visibility labels in MR +I dont know why this is so +trunk +please close +Hi Julien can you please close this one off for us +Will go ahead and commit after running through hudson +I will take a look to see how complicated this is too fix but I bet its not too Jonathan Thanks for looking into this +Patch mocks the in the failing test +mypage is already deleted so issue can be closed. +I think that the current norm in Hive is for the user to execute the add file +I just committed this +Why not serialize the kbase with the session if that is the case? Also we can add a callback to the kbase so that it notifies the application every time the internal state changes +If spec doesnt mandate it one way or the other than sticking to what cli already does is best since than hive will be consistent +But I do see two of your screenshots where the trailing strings after comma were not removed +Please note that the protocol change is in along with metadata changesThe actual protocol changes are what are detailed in the review comments so nothing else new there +Lets see if makes this a +Portions also committed to TRUNK. +Messy patch to implement matching schema and data return in +Do you mean the private field parameter is never used? Then yes we can probably remove it +I think this needs a larger refactoring which will likely take some timeTherefore Id also say lets move this to HV +We could also provide a createServer method in for user to override +Im going to commit this soon and then move onto the flexible queryparser +Thanks for your advice JackyWe have solved it with the same way +If having trace messages is an issue reopen the JIRA and Ill remove at least one of them +Added null check for +I solved the problem by converting a normal href to a form and added a javascript call to the submit solved problem for sortingpagingexportLook in the code method write and write method getI hope this is usefull for your bye +This sounds good +is not a browser +please add timeout option +Reproduced on +Not anticipating anything too crazy +So it has a potential risk that may lead to inconsistency +Shane what version of maven are you using? Using I can still see openidjavas transitive dependencies in the dependencytree as mentioned aboveBesides when building the example I can see the following jars being added into the lib folder this issue +That seems like a somewhat bigger change +I am afraid Ill have to set up squidproxy with NTLM authentication at work next week and run a few testsSo stay tuned what do you think about the patch I just attached? I think HTTP versionof the proxy server should not be taken into consideration by theregardless of how we decide to fix the NTLM authetication problemOleg +With your permission Id like to explore further rearrangement of the dist tree to more clearly separate single process from multi process +This time the lines of differences are much fewer than beforetmnksedenya Documentsapachederbywok wc l tmnksedenya Documentsapachederbywok wc l Well +You can apply it to see whether it s right to help repair bug +That is why the code in the does that +Luke yes please do +patch resolving verify that the patch is valid or if a better solution is needed +New patch +I am going to make the oozie SVN repo read only shortly and then migrate the repo +bq +Merged fix for this back to x x +How do other projects that use SLFJ solve this +Sorry this was my bug sends that revisions property as part of the body of the PUT to the local document but it shouldnt be saved in my database +to trunk now +Harsh it sounds like you are happy with the approach so Ill add in those changes and resubmit later today thanks +Is someone already working on a patch to remove the l option or should I grab this jiraPlease take this up Todd +as Devaraj I think the timrout should be configurable from day +done +Thanks for the feedback am going to applyimprove mentioned points and curious about what to do best regarding ordinals support +For instanceset REPOSITORYfileC is the script I tested on Linux +Thanks +Could you clearify the step to reproduce? Does this issue occur in +See +sure how useful the version check is If the node happens just after RS creates the znode we need not do handleRegion because anyway the node event will do itThat is why i suggested itSo Stack this patch will do or should we raise an improvement and work on the new changes mentioned in the above comment +tested the patch and it fixes the issuethxs +Thanks for the quick fix Jeff +This is the content for that run of theJobtracker STARTTIMEJob JOBIDjob JOBNAME USERhadoopqa SUBMITTIME JOBCONFdfstmpmapredsystemsubmitgymtpJob JOBIDjob LAUNCHTIME TOTALMAPS TOTALREDUCESJob JOBIDjob FINISHTIME JOBSTATUSSUCCESS FINISHEDMAPS FINISHEDREDUCESJob JOBIDjob JOBNAMEsorter USERhadoopqa SUBMITTIME JOBCONFdfstmpmapredsystemsubmitxwkdJob JOBIDjob LAUNCHTIME TOTALMAPS TOTALREDUCESJob JOBIDjob FINISHTIME JOBSTATUSSUCCESS FINISHEDMAPS FINISHEDREDUCESJob JOBIDjob JOBNAME USERhadoopqa SUBMITTIME JOBCONFdfstmpmapredsystemsubmitvgoyJob JOBIDjob LAUNCHTIME TOTALMAPS TOTALREDUCESJob JOBIDjob FINISHTIME JOBSTATUSSUCCESS FINISHEDMAPS FINISHEDREDUCES +especially since otherwise youre looking at duplicating that bt standard and deb scripts no +Hi Jaquessorry we have developed away from standard ofbiz long time ago +Hmmm +Unlinking from properly infers type from the compared column +Maybe elaborate your configuration in both cases there also the way how you generate the keystoretruststore which is very importantthere might be something wrong with my client store and root store to hold certification between each such generated under key tool which is not common in real environment as we need to build some certificate chain between client and server so any good idea to generate such in programming way +Commit believe there might be an issue related to this patch in when you have a transformation that produces no data for example a redirect in this case both a header will be sent and Chunked which seems unnecessary +Can you upload the output of ant +Or just put into thatll do +Bulk close of resolved issues for. +We may skip JDK because of time +I am not a JACC expert +I identify this as a VM bug and therefore there is nothing to fix on our side +This is probably very true for translets but it does not make much sense to dothis for +runnings tests right now review coming. +I filed a ticket few days ago +The transient field here doesnt appear to be necessary one might put the validation rules in the userId constraint +DoneIm still wondering if this is the correct licence header +You may want to look at +What is the error you are seeing? Is there a stacktrace? have you tried validating teh instance? Are there errors with the validation? What do the payloadsinstances look like? What should they look likeI noticed in srcs attached to the Axis jira issue that the package name is different from what is generated by xmlbeans by default from running scomp +So I think it should work +The functionality seems incomplete without Marc I agree this should be included +Ive committed thisThanks Todd and atm! +looks good to me +Lets do it even before +This created some slight problems though mostly dealing with root +Integrated to and trunkThanks for reviews Andy and GaryWill integrate to soon +Corrected theNow I could run all tests without exceptionsRegardsJan +However I didnt know that some can optimize away duplicate objects automatically +The CAS Pool will be drained and all will pile up in the pending queue of the delegate that is not availableAlso the assumption is that a delegate joining late has a compatible time system +I agree that it would be more helpful to include messages when throwing the exceptionsThe case you give is a VMJIT issue though I also see lots of places in the class libraries where the exception is being thrown without a useful messageIll start to do some tidy up +First draft of the patch that converts to +Tests succeed +Ill circle back with more info when I have it +Looks good to me +Hey Tom your command workded directory +If some thing happens only the stacktrace is being sent to the receiver in the detail element which is I know not enough at all +Closing as INVALID because no testcase was supplied. +So if you already have the code handy or if it will not take long to implement sure go ahead +For more details see the java doc however I plan to improve the java doc with a native speaker +Please upload your wsdl as wellthanks attached a sample of the wsdl file I am using +Proposed acfdc +You should raise an issue for Grails if you want that fixed too +Reassign to a release that you feel comfortable resolving this issue in or leave as is +is just for the protocol versionI intended to make them resemble to the original bsplib but I tried to adapt them for hadoop +Oh yes +Looking at the patch now +Hi Radocan you describe as best as possible the impact that this open bug will have on a customer in a production env? Do you think this is a blocker and should hold the entire EAP release? if so why +what about isolated networkAFAIK it should be same for sharedisoldate +Well have to use or any other AOP framework that supports Java such as +Shouldnt sane build have similar performance as insane or they are usually having difference +Closed upon release of Hadoop. +Thanks for the tip on the Eclipse +This was because historically groovy had to handle Java as well as +I will apply this changes as soon as I can +Release Notes Text corrected as per JIRA . +hot dog +Well +closed +Attaching an updated patch +I will update here if once I implement something +r for addition of length filtering to documentum filenet jcifs and file system connectors +Ive just committed this +Then server can be started successfullyIs there any other requirement for this bugThanks +has nothing with the error about test +Hi I dont know how to make a patch +Thanks to Vinod for all his help in reviewing and debugging this +so no warning. +It does seem to work from my limited testing on my Ubuntu server +It seems to run fine outside the jar +It should be left for the caller to add a null character if he needs it +Even if you do they dont belong in an issue tracker +Bulk close for +On second thought even though the oasis docs refer to how the incoming xml should be handled when the type isnt present it doesnt require how the handler should work by default +In the version the exception still occurs but the maven builder does not abort +who knows +A new JIRA is created for the above issue +DONE +Stale address WS issues in a more appropriate ticket +Committed! Thanks Namit! +It is not a bug. +hit Enter a couple of timesthen copy paste the clipboard to the jira issue. +The next time this happens Ill grab log files and a copy of the datadirThanks +The attached patch only applies to the security branch and not the trunk +I ran the test derbynetprepStmt with both derbynet and derbynetclient frameworks and test passes with fix and fails without the fix with a disconnect exception +I did not remove the other version constants because then we have them and can use them anywhere else +Thanks! Sorry for taking so long +Any news to this issue? Youre saying that you are working on this issue but Fix version is not set status is still Open and it hasnt been assigned to anybody +Yes I agree that its fine to do these cleanups as a separate issue +patch. +I finally saw this one and fixed it +Thanks Sami its in +Hi PrashantHow to run these test cases for system Vms? As system Vm offerings do not include the CPU and RAM attributes and normal service offerings cant be used for system what is the way to scale up CPU and RAM of SSVM CPVM etc +So this is supposed to pass when the is done +Had a fix for it before the most recent stint of traveling trying to find it +document the yarn cli +The reloading behavior depends on whether you put you jars in the locations specified by the property or the property +Attaching simple patch which uses passed args to build the new arguments to zookeeper +I will fix the space in committing itIf no objection I will commit it tomorrow +There is still and outstanding issue of what happens if the service interface is not Java +I think the test would be useful + +Attaching and with changes as followsA srcrefM srcrefM srcrefPlease let me know what needs fixing Im not at all sure the info about the version values is rightThanks +The interface extending configuration +Thanks for making the changes and sorry for the delay in reviewing these patchesThe latest changes look good to me +Could you please let me know when the next build of version will be available on the maven repository +This issue is something else entirely +Please let me know if we need any changeBut the way we have written code to invoke javascript function there is a restriction that the javascript function name must be initAnalyzer is that fine? or we should need to generalize thisRegards dont know what do you think? For a first cut I think that we could go with a hard coded function name until theres a need for something different +Thanks for the quick response attached a log from a regressionreport +Is there any possibility that you have a second? Perhaps inside the deployment? Thats the only reason I can think of for this happening +example of system property related to hide dialogs arejface enables these by default when running in headless mode not sure how actually setour own usage tracking hasusagereportingenabledand WTP also have one related to approving license of xsddtd downloads +Marking for but should also be included in Werner Ill remove this bug from blocking then and just commit thispatch as a separate issue +Looks like this introduced bug with preDhruba Could you have a look at? +Well commit them to if will be fixed before release +From discussion with Hardy this shouldnt be an issue in community AS +Will commit when tests pass +What does it actually save us? Is that worth it +Thanks did you work on this at some point +Im actually not sure if this is because Seam Servlet is causing a problem or something else +Why wouldnt the archetype build that structure for you as opposed to the develop having to build them manually +Great and thanks +Should I open new JIRA for serValue method? +In fact there are no errors or exceptions in the logs +Thanks +Not sure if this is the right place to make this comment but I was wondering if the auditor should be a package within or a separate module +The code changes work fine for me +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I cant reproduce this inFor some reason the error page mapping from doesnt work hereA request to page never problem in is fixedThere are no in so if there is nothing more to do in then please close this ticket. +It came without source documentation test cases or Forrest documentation +Ive checked the logs and broker does not report anything amiss +Richard I believe this is a duplicate +Addresses Doug comment +for the patch +It would be like Java exposing byte code instructions because thats what Java is doing behind the scenes +Just need to resolve what we should be doing in the former case +What version was this test run on? Trunk? Before or after refactoring? If after refactor then its possible this issue is because of some of the ugliness in +This patch gives a simple example that users can get started with right away +It would clearly be fairly empty +But that is not enough to veto the removal though +But leaving for confirmation +Its much more predictable than relying on the client ordering of the entries +Allow useAjax in some t apply patch to branch not need to apply the patch as the work will be done in navcontroller branch and will merge into trunk be verified in gatein master and reopened if necessary +Test case for will work on this and have a patch very soon +I can use my current project as a guinea pig to test solutions if that would help. +I already submitted a patch for +That is to its work action works as designed when Expand is invoked appears to the left of the node that is action gives ability to browse the content of tld file it should not expand the tree +Ive tested our projects against +It also adds the ability to recognize native test failures to +There appear to be a number of problems with the Commons CLI combination +I also ran and on a Java ME platform and they passed cleanly with this patch +The sample Android application included with the Cordova download had not the problem +I applied it +Ill amend the output to just print the respective section and paragraph number and move the specific conditions back to comments. +My justification for my comment was that Giraph is being used in Nutch +I checked in all but the addition to the suiteThat can come as a follow up patch when the sed issue with the test is fixedSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaSending javaengineorgapachederbyiapireferenceSending javaengineorgapachederbyiapiservicesstreamSending javaengineorgapachederbyimplservicesstreamSending javaengineorgapachederbyimplservicesstreamAdding javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestssuitesSending javatestingorgapachederbyTestingfunctionTestssuitesSending javatestingorgapachederbyTestingfunctionTeststestslangAdding javatestingorgapachederbyTestingfunctionTeststestslangAdding javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +Glad to hear it! Thanks for following through on a bug report either way +If someone could answer anitas question would be wish I understood these statistics +This has been fixed in +I second Daves thought that we probably dont need an additional key in the +So the name should have been If you gunzip it and then gunzip the result again you have a patch file +Ill take a stab at this when I have a it out of release since this is not a critical bugfix. +This patch intersects with the other blocker +Please apply attached file for the fix +A lot of code management tools have trouble with that situation and see it as a merge conflict +A fix for this has been committed at rev in Qpid trunk +Proposed patch to resolve this +Done now +Hope you dont mind Nathan I also came across this and did not want to deal with a patched inline version given the existence of a quick fix +move reporter from Steve to Dave +Backported to +Please open additional bugs if you see other problems +Also we should have a way that the processes started by hive can be easily linked from the Hive Web Interface +Also coordinator page should display parent id in case it is part of a bundle +Bulk move issues to and +Thanks for the patch Spencer +So as I loop through the modifiedBy and modifiedDate fields in Node are getting set and those new values are properly stored in the DB +patch is a little different + +Committed to trunk. +Patch addressing Jenkins concerns +See. +isnt in WRITE so its not really determined where it is sitting in that phase +Ive applied a slightly modifed version of this patch to in g +downgrade from critical to minorplease change priority as you feelthanks +I also did some tests with multiple shards and did not see this problem on trunk +I made one minor change to the patch Removed the host and port parameters +Thanks for the tipUnfortunately the exception stack trace has to come after the thread dump when we do it this way as the message can only be set in the constructor the super constructor has to be called first and methods cant be called in it so we cant get to the stack trace until after the superconstructor +Yes I deleted the +Igor please check if your problem disappeared since r +I dont used Maven +Im not overly familiar with the historical evolution of this patch but is a new really warranted in this caseAdding the lastBottom param to seems like it would make this easy enough to useMy concern is that most novice users doing simple paginated search arent ever going to hit the deep paging problem keeping track of the current page number and multiplying by the number of items per page to determine the numHit value works just fine for most applications +I hope to send you a patch as soon as possible +If you did not put it there then the class is from a library you use +The previous patch contained a small issue that needed to be corrected +BTW that is a hypothetical solution not one that is implementable todayCan you file a CDI issue for this? +For all minor releases I generated it manually and uploaded to the staging site +Results are consistent always s vs +Setting it any higher is silly anyway +Thanks very much Knut AndersCommitted patch to documentation trunk at revision +yeah local mode is weird +Created confbootstrap to start security Note for EAPThe location of and has changed in this release due to a security fix +Mirror deleted. +Updated to CR +This would be a great documentation patch please do consider writing it up and submitting it +I agree about being cautious about creating confusion +libcsharptest The C generator has a warning about an unused variable +So how do we protect this new interface from prying eyes +seems to be a duplicate of +Is the issue still persistent in the upgraded versions +Also did that stress patch work to get you failed request counts? Would be good to get that too if we can show that even ms keeps requests from failing entirely +My understanding is that this patch only fixed d with ac still open +Changing to be independent of would be in incompatible change at a later point so maybe we should consider fixing the RPC layer to allow additional exceptions right now +I was going to move the code to a new class +This issue not reproducible by description +This way I can make some tests against I have guarded against this condition and will commit the fix shortly +This should already work in the latest + please commit if tests pass +OK Ill push it out + +on fixing for will try to do it the build the meta row keys with only the table and the since it is the between the two keys that is problematic +What values can the Resource field take? What fields of a property are required? Should we permit a description field +Ok thanks for this information +I think the mkdir rmr cat commands should be hadoop fs instead of hdfs dfs +One can always put together a custom that parses only select set of fields while others are left unparsed +Adds an additional test of at the bottom +The original patch does include a test +I reopened +Doing some tests test with my not seem to work another is still not this as a bug but we should get to the bottom of it. +Logged In YES useridThat has been added as well. +What about the contradictory error message? Should it not be fixed? Its very confusing. +Are we changing the default behavior of branch ? That would be incompatible +Im suggesting that passing null values when you dont intend to delete is invalid and should be reported as such rather than silently accepting them +Note that another solution would be to have a commandline property you could set when running JBDS to override the value of as passed to the plugin +I also could not reproduce the issue running many times so for closing the issue as cannot reproduce +Includes an HTML UI for testing +DarynI tend to disagree that we dont want to expose the mapping +Can we add it to our current sample set? Do you want to commit it yourself or do you prefer me to do it +Committed to trunk. +I think the constructor needs to be removed and the code in the constructor placed into the convert method +Please wait before closing this issue to early +Committed switch to OPC weak references in +There is a lot of corner cases in synchronization logic but there is a lot of comments in the code +On the patch side I dont have any comments except awesome + +To make error messages more friendly could apply the attached patch tooInline patch doesnt seem to work so I removed +I would prefer to have separate build targets but a shared repo +Do you have a unit test to recreate this bug +fixed on trunk please aaime test and backport to +closing it +We should upgrade to Tika after the existing version is used for the new MIME type detector +Obviously we dont want to flush every time we update compaction logAlternative way is to write compaction log file with inputs each time we start compaction and remove it when completeIf CF sees that compaction log files in its data directory that means there are unfinished compactions and those sstable files with ancestors that are in compaction log files should be deleted +As we find databases which do not support joins or do not support A cross join B specifically we can change their dialects +Without this I dont know what area of the code triggered the exception +That seems messy to me which is why I proposed this +Great workLooks very good +Youre rightIt happened during multiplication +Well obviously it need to be investigated then +I think Ive figured the problem +Initial patch proposition against trunk of +Another patch +Is a null schema the same as an implicit schema? Or is there a different concept being discussed here +Ive created to deal with void methods +Thanks Bill. +Apologies for letting this slip through the cracks. +The data given in the description now is more likely to reproduce the bug +Committed second patch to TRUNK only +I assume before I do that I need to get more info from Cargo What API is used and what parameters are passed by Cargo +Do we really need to support all the configuration options as command line arguments +In this scenario updatemodelvalue is not called after process partial +Patch integrated to trunk and. +that should be +And it can prevent the error occured in this jira +The fix is you sure this patch worksI added Java Platform API Specificationto the luni manifest and it still doest seem to workIve also tested w the j vm and it doesnt work there either so Im guessing we have a classlib bug +Hey Brandon just FYI per the discussion on just updating the expected CLI output will be sufficient to address this issue +Phelen The image you have attached does not help a muchAs Brian said this issue is probably caused by custom CSS which the look of the editorCould you please dive into this issue in more detailLets use Firebug HTML view when you select affected element you can see on the right in the Style view what CSS styles are applied to given elementLets play with them and find one which causes the issue +Great! The patch looks good +on the current patch +Maybe update the Description part of the JIRA +Thanks for the feedback + has been marked as a duplicate of this bug +Not sure if applications doing lots of batch writes will see any difference I plan to try running continuous ingest with and without group commit +Committed +Started +Patch v deprecates a few more classesPlease review +Jiras that are specific to the seam distribution should be assigned to Marek for triage +Uwe yeah that is what I was thinking +Thanks for doing all this work Im sure it was a pain. +rAlso applied patch from ubuntu will need to be tested on x +I had no trouble importing the example as an existing Maven project +Another is to add an optional isOwner or isHost property for Participant +At the moment only the review part is relevant +from me to commit is anyone else reviewing this change +Thanks Lewis +Corrected the pick and order buttons text +Thanks +So this is not the issue +am just thumbing through the code one last time +Can any one please review this patch +Of course we should keep the around instead of calling methods over and over again +Ill try that in some of our tests and see what happens w regards to time vs space. +after a updatehttplocalhostthingshowjsessionidgpvpfmugafter clicking edithttplocalhostthingindexjsessionidgpvpfmugafter clicking update againhttplocalhostthingshowjsessionidgpvpfmug +Im closing this bug since I couldnt see the problem in and the bugreporter didnt indicate if my tests are valid or not I assume they are Please reopen if you think the problem is still present with an example basedon the Cocoon SOAP samples so that we can easily reproduce it. +If it affects more than one EAP major release please make sure it is cloned so that there is one JIRA per major release +rejected or a dupe +Let me know if you still have problems. +Please be active and so we can fix this issue +As there are a miriads of issues on doing that +Fixed in SVN revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Similar defect has been raised against SWF null event now returned for empty string result instead of exception. +I think we should close this as Will Not Fix +Pull requests sent +Also it would be nice to include the activemq bundle in the final distribution along with the property file so that users can customize those easily +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Please could you check that it has been applied as expectedThank you for these new tests +ZK I did a test today by removing prependIdfalse from my form but it does not seem to fix the issue withThis issue cannot be reproduced so maybe the reason is another +Suggestions welcome +Due to localization is the German notation for the english notation of exists for to be already fixed or not reproducible +here are some of my contributions +Depends on and +The problem happens when you have on the invoking class the bug the comment from the GEP is very old and wrong +lets keep it in +By default header is null and one can always override the content type set at the construction time with set method +I am not sure about it but I think that CHM is the first one that doesnt grant a fully reliabilityOn the other hand I think that we dont need a approach in this specific context reliability is an option +Good point that sounds like the best way to go + +ask any developer on the street what that is and youll probably see a question mark on hisher face +A notification is never sent to the author of the event +Adding a call to provides info so the mr job can include the proper jars and fixes the problem +Committed. +It works as expected +hmm for me its not so easy to run the litmus test on my local pcmaybe someone could run it? or i could commit it and let jenkins test it +DavidSorry for the header +Perhaps people changed their minds in the mean time +Here is patch that adds closing state to Regions and then in the HRS it doesnt offer regions that are closed or closing as part of the list of most loaded regions to the master for it to later try close +Vincent is on vacation this week so he wont be looking at this in the short term. +This is the ant file I used to develop and test activemq based program +What about the long lines in code generated by thrift or protobuf +r +Closed after release. +Thanks Amila for the contribution + allows users to leave off the percent and currency symbols when entering input +A patch for branch +Ive hardened the code in revision but do not have JON to test will attach a copy of the new plugin +Patch for attached +by removing those features we canprevent the user from seeing the questionable behaviour +As for the I dont think it is used for anything else than tests +Fixed old problems +Added thanks +Added canceled check to prevent DS from being created. +Deals with the time element +Ram yes its an open defect in the sense that is disabled +on patch v +Minor cosmetic issue would be more consistent IMO to call it as it builds +Hey Mark to this +moving out + +The problem when using Spring together with is that the correct working then depends on the classpath +Doc needed for AS but will not hold the release +Closed after release. +Closed after release. +Would this work for a directory as well +I will still be using for the next few months so I need my status page to work with both and servers +So + +sonar +For verification we need additional info +committed +From my experience you can only use alphanumeric characters in the password +When you add quotes around classpath it doesnt help because command line is broken into separate pieces with default that doesnt care about quotes so it breaks classpath argument if it contains spaces into separate argumentsThe reason why the patch works on windows is that on windows builds command line back into single command line string so quotes that it contains help to protect one argument from being split again by windows exec code +AS just needs to be released so EAP can use the new version of the Adaptor +Actually under Safari once that typo is corrected I can no longer reproduce the starvationtimeout condition +CXF is already cut and voted the servlet patch is not catch up this releaseI think we should add this issue as a known issue when we release Camel +Very good idea +I love updating the docs so I snagged this one to crank out +You are missing a resource in youresource bundle +This has been superceded by the previous commentThis has been completed. +Checking the validity inside the service registry is complicated by the fact that it still needs to hold the bundle lock to do so for the entire until it is done registering the service which is why the call to it was wrapped with a bundle lock in the first place +Disregard +And if youre clever enough you can manage it +So Im closing this it isnt an OS X issue. +Main change is renaming WAL to +thats what I get for not testing an svn export to run all the tests before building a diff w svn diffIve attached a corrected patchThanksJess +ASF Infrastructure are working on this issue +I think adding this parameter is to support Click applications build against older versions of the frameworkCurrently autobinding performs add any public controls to the page after the page constructor has been invoked bind any request parameters to public fields after page constructor has been invoked add any public fields to the page model before renderingBy having autobindingfalse these features will be disabledregards Malcolm checked into SVN will be available in releaseregards Malcolm Edgar +It is common for a new ESB developer who deploy undeploy dozens of +Rerender button is not supposed to send data it should only rerender the page with data which already are on server. +Could you just wait on the watcher triggering before proceeding? Add that as a TODO when you commit +provide a quick run down on the issue summarizing all of the above +In responce to your previous comment Ive created a new util class that handle the opening of hyperlink in a browser +New patch that also eliminates the three warnings wrt unused variables +Oops sorry + was also changed there was this DELETE FROM EAMCONFIGPROPS WHERE KEY CAMSMTPHOSTnow there is this in and thereafter DELETE FROM EAMCPROPKEY WHERE propkeyconfigText +See +Will there be a need for a trust all connections checkbox in the UI? Or do you think its better to require the inclusion of all trusted certificates +The problem is that the test expects a json response by the elasticsearch connector plugin but no request is done on it +bq +is there a reason this file was removed? I cant build Railo from source now +Closed. +Will show next time we push website +This eliminates the need to synchronize but it wont get rid of the Eclipse warnings. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Best regards try the last patch for this issue the Selectors implementation +committed +Verified it All the problems have been resolvedso close it. +It works after clearing the browser cache. +Also the JEE Module Dependencies workaround from the Wiki does not work for Eclipse features that do not use WTP +Mind if I close this +Was verified by Mikhail. +Replacing local modernizr with applied Kristens patch to the bootstrap branch +I also have some problems with the +Patch is applied at +Then upload a test case +Also includes line length and javadoc fix +Mass closing all issues that have been in resolved state for more than one month without further comments +So if you take a look at memory it is natural to see this class around +Code changed on branch and doesnt seem to be an issue on trunk. +Its only if the MP registered more than merge for the optimize and youre using SMSBut really if your app has risk of opening a index it should be prepared for this excbq +Let me know if you need help on Georgegood news the bug is fixed inThank you very much +I doubt will ever exist a day since brought a bunch of regressions due to the fact we released for users in a hurry and it would need more work we can provide today +ScenarioTwo nodes A and B with two services and that open channels on top of a shared discovery +Something like an icon Maven would be nice +The plugin currently works with and hasnt been tested against you have a small sample file for us to look at? +Closing issue as it was included in version which has been released +First draftNote that for testing purposes markup is set to true by default +I found an explanation to the behaviour Moving items between the left and right hand side are client side and therefore the event is not triggered +will be returned as abc +Yep the speed improvement one +Thanks for pointing that out +So its up to you guys if you wanted one that changes just what we talked about or one that changes what we talked about and removes some dead code +Patch applied in r +Patch to fix the issue with +IT added +Screenshot before saving the bpmn file and the output of view source before saving +committed +Should it not be before HIVELIB is used +I am in the next tutorial now with the being added and you have the same problem with that tutorial only it is more complex to describe how to reproduce +It should be possible to add this to FUSE SF which will allow it to be included in FUSE ESB +Although there is missing deployment configuration for AS +I meant RC +your ve traced the Tomcat source codes about the JSP reading and seems that Tomcat detects the JSP encoding first which includes the pageEncoding tag and the actually file the file encoding part Tomcat detects the JSP files are ASCII or or orand then creates the corresponding with encoding argument to read the JSP the encoding cannot be determined then the default ISO is used to process the JSP this help +I dont know with which software the original was created with but I know that it was reprojected with +Ideally this should go into GA but setting the target version to +However that agent is the least common denominator of all possible agents and therefore is too weak +I cheked in the fix and it seems to work great +is there any indication that this is a fileupload bug? its possible that its Jettys fault +The schema located at contains the purchaseReport element as was described in the xsischemaLocation attribute +If it is not resolved by due date its out +committed in SVN +As you can see extra variable gets lost after the reject operation +Overall appears to be quite well keep looking +Setting to TBD since was the patch to m +Can we get a unit test for this one? Seems easy enough conjuring Charless context +you need to paste the whole URL instead of clicking the link. +If you are up for itLars Thanks for doing the research +So if you want to prevent deployement you have to bind this mojo between install and as we dont have any phase between install and deploy +HiWe never used to get NPE in the previous releases when tried to access MS before it gets loaded completely +You should be able to test the minion program standalone to see what happens cf class Javadoc inThat said I have yet to test the new patch on Windowscygwin Ill get to that today +Here is an early patch for review while I continue testingId really appreciate fb since this introduces a fairly large no +should be fixed Jochen will give it a try +I will log an Emb Jopr Jira +I get that what Im saying is that I dont see the rationale for disabling the durable sync code paths +Passes all local tests +kinda sucks to make them create List objects for each callof course internally it should just translate that to a multiget in and throw the exception if not found +compaction triggered when needed would indeed be perfect +With committed I verified this test is now passing +Are you running the Sonar batch and Sonar server on the same machine +Interestingly I no longer see the process in the Process Explorer when I fire up a programAnyway I think we can close this one. +having local indexes makes applying AND clauses touching multiple indexes easyI dont think that is a road we should necessarily be going down a compound index describing the AND is much more performant and fits the view model much better +Linking this issue to in which a query generated by Hibernate fails on Derby due to the way queries that include GROUP BY w HAVING are rewritten by the parser +With deploy do you mean boot or before it actually starts bootingWhen using NFS for both Primary and Secondary a simple QCOW file copy is required but with RBD you have to go from QCOW RAW on the first copy +If both are local addresses no changes are made +More and more of these happened in Geronimo HaddoopI ended up restarting and things and it seemed to get betterAny thoughts +Resolved by +Git repository changed for the schema applied on branch Alexis +So the idea is that if the NN is an old server the client sends the old RPC getListing or get which send back in full path name +should contain all of the metadata from and If you can tell me what elements you cannot see in I can make the necessary changes in the metamodel package for your use +If the problem still reproducible it is in threading otherwise its in GC +So this is definitely a bug +All set let me know if you have any trouble +The only other solution I can come up right now is to have some shared that would be used by both classes but thats probably overkill for this simple case +You can also add it directly as attribute collectionType to hselect + +I tried backporting to but it doesnt seem to fix it +Thanks Myrna +Broke before the patch +My first suggestion would be to swap this order and see if this improves things +Fixed in branch under svn revision in branch under svn revision and in trunk under svn revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Can you provide any more details at all +Linking to the related issue in range increased +NET so I dont know if using these are cumbersome or has major drawbacksFor other scenarios there are bridges +ensure the commands work or am to update outdated references to svn STS etc as well +the new tests are now part of the patch +upping included in build from +Changed priority to minor +Now it is being removed if it exists +Ari which version do you need? Enterprise or Developer +You are importing the package rather than exporting the package +I agree with this so I committed the patch but with slight changes mainly that the verification that a method has either Path or an http method designator occurs in create instead of parseMethods and a warning is issued in such a case +upload a patch +changes to soa clarify which version of the features you want to include +Also the system will spend time on other operations which should not be affected by this patch +Maybe Bob forgot to set this issue as resolved but this error is not present from SOA tooling +And the patch looks good for me am cleaning up this patch a little bit to get rid of formatting changes that arent relevant to the issue at hands and committing it shortly +simplest thing is to leave the node in gossip but remove it from tokenmetadata +JochenAm looking at breaking up general +ok Ill try to reproduce it in a standalone simple and run with mvn jettyrunNavigate to httplocalhostwicket afterwards +This issue is a duplicate of which has been resolved +This issue related to +Thanks Jonathan +Remove the extra lines +I can run the same app in AS with no problems +and let you know +Applied a few days ago +For a start there should not be the assumption that all hosts will want to deploy a Registry and a Message Store +So using jobtrackertasktracker group is almost equivalent to readable +Tests are passing Ive also modified our own build to use ivysettings instead of ivyconfigure everything seems to work properlyIll let you mark the issue resolved when you and Maarten will agree to say soNow that youre an official member of the team I dont thank you explicitly for your contribution any more but you know what I think +Updated patch for accounting component +Comment Were shipping version tasks completed apart from the issue with the graphics which we are investigating +I do not think it is a good idea to move the class since there is a possibility that we will have several implementations of the interface in the future +Once more as I was able to do in previous versions ofIf after that I commit SESSIONA both message will be removed +I have verified this by rebuilding stubs with the latest nightly build +please move back if they will be worked on asap for +Its not just +Ive opened for the missing piece +Thanks for sorting this out for Qpid +I dont see anywhere we arent maintaining unreachable members correctly so we dont need to do this. +My fault for not updating the JIRA +works for me +mysterious +I changed the global to reflect my ubuntu fs value but no luck +Would you like to contribute a patch to for pom files? We currently release three jar files maxent tools and the uima integration +Your OS will cache the file in memory after its first read so subsequent reading should be really fast +Verified by Roberto. +Dean I am sorry it didnt work for youYes by all means if you have time later and still have issues then open a new ticket +Thanks PaulexAdded a modified version of your test patch to ARCHIVE modulePlease check that the tests are acceptable to you +rev Added results declarations for the script based server operations +Can someone confirm if this fix applies to the branch too? The details say Fix VersionsI encounter the same in the source code release but currently I dont have a git environment to verify it +The fix will be in Beta +I think we could add final to for fixing the findbugs warnings +All the best effort support language has to be taken out +Kihwal adding this line was clearly a mistake and so it should be removed +Sorry I wasnt arguing that job jars conform to the war file spec rather that if were looking for sucessful methods of including libraries in jar files the war file provides a good analogy one that is more widely used than executable jarsA job jar is not an executable jar +Resolved and committed. +This has been corrected +The majority of plugins work just fine across releases +I use the grid component extensively with a checkbox as well an I do not have any problems +Truebut at least i have a placeholder for section +Heres a patch implementing Alejandros suggestion of enforcing the limits at the time of aggregation in the AMI did a test on a single node cluster where I set to be a lower value in the client config than the server config +logo has been donated +Punyades is too formal for me +whats wrong with using +Logs in have deployed a web application on OCJ Oracle Application Server +Then I would chose JPA Entities from Tables or JPA Entity +But by default Cobertura or Clover dont instrument java unit tests so I dont understand why you need to exclude those files +For the latter one has to further worry about what if only part of but not all messages are postedTo be consistent I think all administrative commands simply create some data structures in ZK and should complete very quickly +Let me fix that. +I was going to err on the high side and go with something like k for and GB for but I dont have a firm basis for either of those choices +This patch will fix many broken commands inside the grunt shell +For the moment Im just creating a virtual NIC eth and configuring a static IP on it which works +Probably it could be useful but anyway you can see the log via weinrefirebug lite +Permaine to talk to Mladen +I confess that the fact that Colt hasnt been changed since discouraged me from bothering but Im willing to fire off email and see +I think the main part of the bundles is developing renderlets for the various classes defined by foaf +Fixed in r +Its actually greater than after additional runs +test passes into trunk and +Both and patch complain that the file is malformed because there is an empty patch for around line +Guessing it was an intermediate EMF issue since resolved by upgrading EclipseEMF +If there are some unresolved documentation issues please file another JIRA for those. +The classes included could definitely be trimmed down +Committed to trunk but not yet to branch +So Ive updated the map and psf files accordinglyNote that weve moved to TychoMaven which isnt as kludgey as AthenaPDE and therefore the aforementioned doc needs a rewrite +I dont think asking people to install a later version of findbugs is a big deal +Given these infra changes should we consider moving to nexus? Is there any real difference btw moving to nexus vs moving to +instead of +For which of is the empty implementationof acceptChildren correct as is +the correct diff to development trunk for in SVN revision . +duplicate of +in case of a misconfiguration of a datanode namenode can tell which datanode should have a certain storageid You might be right here +This is not a usable Nutch plugin patch +Hey KonstIll this jira to just remove the +Note that is and should not be used +Will commit later this evening +Do some perhaps not move system clock when a full GC occurs +pushed to with JBDS +Chunk boundary data should only be visible to the HTTP layer +Logs and to previous comment + +Thanks for explainingI dont know whether any test failures were caused by this probably not +Please download and experiment with a trunk snapshot. +Fixed by applying Chucks two issues are related +Let get someone with proper hours of sleep check it before closing +Refactored patch that removes network calls using and simulates better a row matrix though inverting and merging inlink scores +Yes it is +This is just a dummy return value to satisfy the compilerPlease confirm that this is just a warning and not an error +HiYou are correct this should be alleviated by changing the order of the comparisons +Then BA will not be done using new fireandForget invocation but as respose on AB in +this resulted in adding a method to the interface +Thanks Snjezana in JBDS Beta +Thanks +Change look reasonable to me. +Where is that coming from +integrated in revision Mark please verify if it resolves your good +The patch applied to trunk and branch +Hiya First Im really sorry I havent had time to dig into this yet +Is this patch committed to the branch? +Leaving this issue open to track similar IBM VME issue the regression test should be integrated upon closing the issue. +fix checked into head +closed shipped in +Its a lot one for each article +This is the clustering section of our +Good catch +What do you think +I verified the build on both Linux and Windows +issue has been fixed as part of +Consider following two casesset highset highIn both the case job priority will now get set to high +Compacted patch shoud still do the same added too. +Made the function editor similar to +Thus the situation is slightly different from I described above +Sorry +Added new patch after fixing lines greater than From the other lines are in xml and is so not changing to trunk +Thanks +An efficient version would involve modifying the +Closing all resolved tickets from or older +if we dont provide something that can be easily tweaked to run in production then we shouldSolr is supposed to be a server not just a war file +I ran with threads and saw NUMALLOCATEDPAGES which is expected I think for rows + +FYI the symptom that brought me here is if one tries to use with then you cant instantiate HDFS filesystems with a cryptic No filesystem for scheme hdfs +Details of Usage +However this ticket doesnt indicate what environments this is failing in +Alexey can you put here what validation option the user can disable +If none I plan to commit itI havent dealt with the temporary tables code before so if someone familiar with that part of the code wouldhave a look that would be great +Maybe store it in the file +Fixed in new code +Up until now I had merely tried to find a setting in the schemas +In we removed code related to in an effort to simplify development of the capacity scheduler +I fixed the main cause I think but it seems there are more things hidden in this +I think this is as good as we can get in +HelloWould you please try my patchThanksBest regards Andrew Patch committed in revision +Thats a good idea +Unless this is a user error this issue should be considered as critical +Commited to Branch as well. +Look its obviously up to yall if you choose to implement something or not +Attaching the input PDF and output jpg get the same results with Open Office +Thanks on trunk +Ill try to play with the patch a little when I get a free moment +Ted I dont think this diff is related to in any way +Thanks a lot Flavio for patch +war with the issue +when the log system works right the session closed function will be called +the relationship between groups and table think the isolation and allocation tool is very compactthere are only two lines code used to change the and in andIf coprocessor can provide the entries which we can use to manipulate this two class our development would be very convenientBecause our implementation has to change some private sub classes ofwe have to rewrite them to suppress the original onesIn this version there are two functions which have little relationsh in betweenThe function of group involved with a JSP management portalthe table priority function is more light weightusers can set priority with shell or apiWe dont know whether one or both are convenient to add to core +New version that actually compiles +It didnt work for me when I tried the same approach with empty while with read call in condition but now it works for me for some reason. +oops sorry about that with the release of. +It was actually a bug in the code generator that arose due to a missing service +Closing JIRAs that are productization tasks that QE does not have anything to verify on the actual builds so these does not show up in our filter of issues to verify and close +Sorry but Im unclear on what you mean here +should work as well +Prevent refreshing +Verified and closing issue. +I could not reproduce the issue +Can you attach the project including all the config files + Great! Ive verified mvn clean install +Unsetting Fix Version for unassigned issues. +That feature is only there for people who were using the script replications snapshot directories for backup purposes +I tried embedding Hibernate Search but found some issues +Thanks for putting up with all my comments Rick +HOD brings up the DFS and web UI on dynamically allocated cluster nodes using the older configuration items +Cases two and four are fixedFor scenario one and three you have to override the observer methods without the Observer annotation to get the behavior you want +Move the fix to +Where in the ejb spec does it talk about the remote client having a view of the component? It seems clear that this is the case for what would be access through a local interface in but this is not clear for remote access and goes against in the opposite direction of RMI in location transparency issue. +Fixed RAT warning +Im raising the priority to Critical because +This is one example where I include Diverts section +Include a fix for merge and a unit test using the for reproducing the errors +That bundle is with Tools classpath was am closing the issue +So as per the implementation logic only if mBytesnull and numRecordsnull it will return the size in MB else it will return whatever it contains which in my case works fine +In the end it boils down to a typo in pycassas init code specifically readconsitencylevel its using ONE instead of QUORUM +Ill add that I ran all the changed tests successfully and weve also got a nightly Jenkins run for the branch set up on to catch anything else +At least warn against potential overwrites sounds like a good you said you could take a lookMake sure whatever you do it continues to work with what seam generate entities does etc +Thanks Nathan +Closing it. +We shouldnt limit the number of clauses in general either it made sense in the past but it no longer does +Fine by me +Its now possible for me to run my builds in eclipse with the embedder whereas before it wasnt +I ran through several terasorts on a small test cluster with this patch and verified that it helps performance and didnt cause stability issues +Committed patch a to trunk with revision Awaiting test results from nightlies before backporting +Thanks +Sorry doing too much at one +Skipping as the various cases are covered by unit tests +Here is a couple of tests checking the behavior specified in the api and the trivial fix to make it +My Bad this regression was actually caused by not +Diwaker youre a commiter now is this patch still relevant +As indicated above we recommend switching to target class proxies in such a +Ill try it tomorrow and commit ASAP +This is a patch Ill add etc +Bulk close for release + have fixed in the issue in the maven codebase will appear in maven support with the upgrade to later binaries. +Thats because of issue +The loadstore redesign obviously wont be going into +BrunoYes I agree and then make more sense than +Im ready to commit this is the patch to Chukwa trunk or hadoop trunk +Whats the issue here? The warning you see below might be caused by a node being killed so thats not necessarily a create a standalone issue if you want us to look at this +It could run in memory but that would not allow for a very big tree +It compiles +It was a very cheap way to build a classpath with source jars automatically attached Now we need to just push more source jars into maven +This kind of theoretical issues would be best discussed entirely to the forum and left out of JIRA. +Patch for review +WIP Patch for the credentials to be used by Oozie +Patch for the current trunk +Thoughts? +But I refined some aspects of the class a bit further +memory consumption jumps high when a consumer is created and drops when the consumer process is optional Store cursor to page messages into the broker from the persistent storeSVN revision . +Patch with proposed changes for the issue +Im currently running into this issue and am also using so another option would be very helpful +It is best to synch up with Grails events and fix to comment out event processing for the time being +Seems like we only need a single Hadoop jar in libsvn mv hadoopcorelib libsvn rm hadoopcore +If any arise then we might revisit this issue otherwise I think we should resolve this as wont fix +on patch +Please RESOLVE and close accordingly +HTTPD does not seem to exhibit it +per the JIRA meeting July John will have a look at this one +So which types of Node are needed to support? Element attribute document text +I extended for the new class and uploaded the updated patch +changed component affecation +I temporary switched it off for code freeze reasonI will continue working on it in trunk after branching +The hadoop dependencies should be in the profiles of the various projects +We just had a user wipe out tmpWhen are we getting this capability +Can we fix the thread safety issue for +One guess is that the JIT is changing the stack depth by inlining methodsI think in some ways the test is invalid I dont think there are no guarantees about the stack depth for subsequent similar operations +Changes private methods that are accessed to protected +why call cleanupRender if setupRender has not been called? if you return false on beginRender the before wont be called eitheri dont think components should be aware of +v combines dynamic snitch info into +Did you perhaps forget to click the radio button granting rights to the ASF +advanced graphics was disabled in that workspace +Bug is only existent in the version of Maven Eclipse as closed. +This module can now be considered migrated +The verifier however detects first keyA and verifies and you return the information of keyA +This prevents a single page history from being corruptedI dont think we need locking +I would modify the patch according to the document ASAP +No new tests required modified functionality is checked by the existing tests and +I just added an example to show how to set up a load balancer camel context for the camel transport through the spring configuration +Attached a Eclipse project with my sample code +Merged to the branch in revision . +INFO Counters INFO File Systems INFO Local bytes read INFO Local bytes written INFO Job Counters INFO Launched reduce tasks INFO Launched map tasks INFO Framework INFO Reduce input groups INFO Combine output records INFO Map input records INFO Reduce output records INFO Map output bytes INFO Map input bytes INFO Combine input records INFO Map output records INFO Reduce input records above issue belongs to performance issueI just committed this +The web interface is good but the Win UI will allow our developers to take xplanner offline and synchronise any changes +The dynamic metrics dont show up unless is used as the hadoop metrics context +Manually tests already exist +I think creating a setter if the property is final is wrong but creating no getter is wrong too +The default version of that plugin is whereas we run +Are you sure you actually did the upgrade? Also please post a fuller stack trace or attach a full thread dump +Opened +New version prefers to optimize for NTS but keeps the mode for ONTS behind an optionMy branch is updated at github new tag is pending +Though I do think itd be nice for the to appear a little nicer in all other JMX tools but personally am less bothered now weve a nice hawtio plugin +Rob do you mind looking at this while in youre in the neighborhood with the other admin stuff? I think its just missing the dependency on orgswitchyardapiextensionswsdl +Moving it to. +Test case Planning to just add a random switch in the current test to split into X +Seems that for now we can use I cant use the current war plugin as it stands I have to hard code the version to s my issueI have modules they are both packaging type warFramework declares a dependecny on I build application if there is a file that exists in both war files the newer file will win +Duplicate of Thanks James. +Thanks Frdric! +No idea what happened +Are these header and body facets +Thanks Dan! +Backported fix from trunk with the following revisions o o o o o o Resolving issue. +SHOULD NOT rely on cause intermediate s should be undone when the user still decides to CANCEL the editing functionThe approach to internally persist the s is probably the way to go +the current implementations out there seem to just inflate the entire stream to into memory without regard for size which makes them useless to me professionally so we need to make sure that doesnt occur +That jar contains a few classesjavaxpersistencejavaxpersistencejavaxpersistencejavaxpersistencejavaxpersistencejavaxpersistenceDepending on the classloader configuration adding the JPA spec API to the application may foul up the environment +Are you using the Windows color extensions as per or just a vanilla Roo install +George why would one have Created and Expires time set to same value? Doesnt it mean timestamp expires straight way +Hi KatherineWhat is the maven version you are using these depends on the JDK version and maven version you are usingve tried the following JDK and maven combinations and I get the same heap space error in the jaxws component every time using the latest build without changesMaven Sun JDK Maven Sun JDK Maven Sun JDK Maven Sun JDK Im using Windows XP Pro SP on a machine with GB of RAM and Ive tried expanding the max heap size to GB usingMVNOPTIONSXmxmWhen I montior task manager I can still see memory available so Im not sure if the heap size option is actually workingIm not sure the problem is my environment either because I can build r without any problemsIf anyone can recommend a better combination of maven and JDK Im happy to give it a try +Does system module also depend on this pluginThis is good! After resolving required artifacts missing xmlbeansjar in directory and a test in derby all the converted modules can be built +I am not sure any of the developers reading this are using DB on a regular basis +That will make sure we dont have to keep changing console consumer as well addremove config options on the consumer +And the content is not on tools the content is on tools just have a single link that does not need to be updated +Any thoughts ideas +When I try to login with everything is not able to reproduce this in windows with exactly the same setupinstallation changed the security policy on IE to accept all cookies regardless of their compact privicy policy +Committed revision +Im marking this resolved as I havent seen any feedback on my fix +Initial patch for review +Fixed thanks +AlexeyI think my patch is quite easier to read but anyway my solution isnt coincide with other methods code styleAll correctThank you +Seems like there could be a prettier way but it works and performance impact should be negligibleThanks +The builders are a nice convenience syntactically but are not an abstraction hiding the details +proposed patchchanges include the ability to specify a property name that will be indexed for the property index fallback behavior when that property does not exist the index name will be I renamed the property as per your suggestionfixed with revision +When opt exists package opt with the tarball +Im absolutely stuck +Looking good +great thanks. +Just adding debugtrue to didnt give me any additional information. +This patch should fix the issues I mentioned above +Back to you its a vmware issue +junitperf has the limitation of junit that I mentioned in my previous comment +and adapted the to not use it anymore in Rev +Ill update the patch tomorrow with Andrews suggestions +As stated in the original bug report we set a timeout value of seconds in the administrator setting page for the job so its seconds +Simply shows the number of tables left to flush and the number of left to flush in the current table +Patrick can this be closed +As of when you coded it there were parameters in the extension point to control order +Trivial patch skipping rb +then my suggestion is to use my comment from above to solve the issue +patch for test to test for custom line separators in Socket for passes unsigned byte as separator to and spec update pushed in aff +Perhaps there has been some repo confusion +I wonder if this should be closed esp +This was addressed by commit of. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Fixes issue with working against +In theory yes but I dont see a way to do it with the Servlet Perhaps a viable solution is to ignore the reason code when also using the need to consider how all other return values combine as well +Since we cant trust him with issue closing powers. +I could rename them shell shell call it something else or is the intention that modulesshell replaces the command line launcher +Thanks. +Enis maybe not since unless we have an ldap server or something similar weve no way to decide which user are available +Max refused to have the patch backported +Added the comments in as Namit suggestedYes this only needs to go to trunk +We may have to do duplicate detection at adaptor level because the parameters passed to the adaptor are +even though we want to add dependencies on jdk we need to add Tiles backported artifacts instead of the Java ones +I know several portlet frameworks use servlets to render the results so its probably safe +Apache Directory studio version has been released. +Theres multiple aspect to the situation here SD JPA was released in July and we dont bump dependency version on SD Commons for bug fix releases +Since Mike implemented shall we close this issue for now? The annotation solution also applies to the of EJB we might add a separate issue for this feature +Added second version of patchMajor changes include +And Ive added them to the Seam distribution also. +I see Jays point about making this API general enough to avoid API upgrades later +Thank you Graeme for pointing this out. +CREATE TABLE blabha +I think the dummy argument is a good least we have moved on from out process being called Main +etc. +Ive updated the quizcard with the content provided by Rayme +Why does this make idbags unworkable in +Committed revision Robert Its your turn now for you must maybe move classpath before the core classpath to override the default x codec to be description to work guys! to merging it to trunk + +I think you cannot simply omit the name attribute from the fieldelement +the scenario is quite simpleIt looks like the +both and trunk +This JIRA is related to the Issue + tested! Could you just move the Code Coverage widget on the right in the default Dashboard to get approximately the same amount of widgets on both sides +i think the problem is the nesting of locks with the same name +Rene I will look out for you attaching something that shows the problem +I think the latest version that does not cause any problem is is a Tomcat issue not a I wanted to open a bug report against Tomcat but to do so I first had to try the latest stable which is +add compile line. +first submission for streams an alternative logo idea + +Done If there is no parameter argument the table and column pararms will not be generated thus Hibernate will use its default values. +Committed revision +As with I agree it is important but I havent been able to locate a free sample and think that the quality of the algorithm is more important than whether it is freely available or not +We dont want to do this since we dont want the collections settable by client code +The one change I think I need to make is to ensure the sessions creation in getAttribute +Good catch +DRLVM changes to eclipse plugin should be applied to eclipse plugin in tools directory +Lets do that in another jiraWill some tests to and have a final version for review +Patch for branch +The following patches have been included on top of the tagThese ones are included in the native +Today in a large cluster you can easily have choke points with mbit aggregate all to all +Can I reassign to myself +Could you provide an example of your project +not copying was a way I figured we could enforce the user needing to take this action +done in forgot this one thanks for the reminderdone in r +No other tests fail at this time but I think I had more failures when buildingCocoon with all blocks disabled + unscheduled issues to the next release +Thanks Yongqiang +Ariel the patch you show contains a lot of text and not all of that is really part of the patch but of stuff idea added +All formatters can deal with null values but they are uncapable of parsing a new issue for this +buildivylibHadoopcommongriddevbin +In our case the FTP Server is Vsftpd version +What happened with this +Im afraid this patch has not entirely fixed the problem +But only during the render process the component associated to the markup tag is dynamically identified and thus only than the behavior you attached to the can be handled +The Tricky part is tracking the hotness of these chunks +Carl and AshutoshThe same bug appears when running in on +I already have some code for Jetty so if you want I can take the jetty side and you can do the agent side? +I will backport to after +In the latest nightly builds the test runs successfully to completion. +Fixed +The test passes for me now +One performance impact is that the memory allocations are the same for production and default and less for performance than in a patch for standalone version just to get the work done in ER +Im also reassigning this issue to myself at least till we can reliably reproduce it and figure out what to do next +If its easy enough please give it a quick try +I just committed this! +Patches were applied at the revision +That is true for indeed +Committed javadrda changes with revision javaclient changes with +That being the case could the commons math library define a constant such that its inverse it equal toI could then use this as a limit. +otherwise +Any idea what can cause that? I have plenty of disk WARN eMQ Broker Store limit is mb whilst the data directory CFusedatakahadb only has mb of usable space +Appears that either the client disconnect or something in your network shutdown that connection +q test format +In short this starts a GR then clients +It ensures only the is pulled when checking timestamps to make the process more efficient +The PR seems to be correct but is created only for the branch +Instead of using HRI I could just convert the reference that has to the region to and fix up the affected code +I totally agree +There is no new test included with this submission +createLedger openLedger and closeLedger are operations that do not involve talking to bookies so they dont go into the same pipeline as addEntry and readEntry operations +why does the test patch add to forceRather than implementing the neighbor calculation in two places I wanted to reuse the logic in AES +Please your application against the latest SVN snapshot off branchThe bug affects ALL versions of prior toIt also affects deprecated code in branch. +In fact why is PE in the test jar anyway? It seems that should be something to be included in the jar proper +Marc I started testing this with the EJB feature pack +Hi PaulexThanks I made some changes in new patch plus I changed the get because it did something worng when crafting the Jar URL +Would you please help to try this +That would at least not be polite and at most crash other applications running in the same JVM +Awaiting run +Checked exceptions were removed in close of all resolved issues prior to release. +The number of message waiting to be written is stored within each session and as the session does not know that the client has closed until the receive the closing notification the value may be different from Once the event has been processed those information will be discarded +This is the vman patch to the existing pom files +I just committed this +This is the my file after running this task +But well be using the Groovy jar that comes with in the app so wed have a different in the IDE than the runtime +Can we reviewcommit the S patch before that because we are a bit under pressure on deadline +comes to mind if we are looking at top N query +This is indeed a problem in XML Graphics Commons but it was transported over from FOP when the image loading framework was extracted from the old image loading code in FOP +code wins +But I see now that other Avro C tests do the same thing so whatever +That might cut down the size of the API which in turn makes it easy to test and tune +currently generates a new schema event for each new file +ok that covers my problem i cant think very many reasons other than the parent dir issue on why things would fail perhaps things like permissions if the hadoop instance is using one +Closing as rejected workaround exists. +I fixed the violations for now +Patch looks good +What component would do that with the new designAbout dependencies I would suggest to create a separate ticket addressing it +I attached a so you can see the stack +If you look up info on yum when exactarch is yum update only update the architectures of packages that you have installed so if you have the base packages installed you need to comment that out in order to update the +It seems to be a short test so I added it +This checkbox will be enabled whenever a supporting publisher is present +This is for a checkpoint +Thank you thrift does not support the paradigm union is a good idea. +But if you still have a reproducible use case on your Mac would mind running the ls command on the afflicted project directory? No need to post the listing what Im most interesting in knowing is if ls hangs on the Mac OS terminal when run on that directory +Lets move the CHANGES entry to other and change it to adding a test for thisDone +Im stuck here +Hello Martin +Applies on top of +Thank you for reporting. +I was about to do that as well +Wsdl and xsd files to generate the java it be a problem processing xsincludes? dims +Ive reset the permissions and added project members to the appropriate groups +Its need in both situations and thats why Id like to see it being consistentRegarding variable naming My first thought was to introduce some reserved variable name prefix like springxxxxxx +Adam You have this possibility already just not by using annotations +Whats up with skipping the first character +Please help to review the patch thanks +help with that is tom +This seem to only affect seam not seam cant see this as being something tools is causing so not targeting for GA and I reaxched out to the seam dev team to hear if they can explain it +could you please rebase this to masterthanksroger +We need to reach agreement about what needs to be done and how to do it and were some way off this at the moment +The problem is that each took too much memory caused by commit edfbcccffcfabbbcc in changing from static variable to local variable +Verified on jbdevstudio. +Thanks Jacopo +If contains a fix for this as a byproduct we should resolve once its committed +Actually I would be all for moving the logic from get to create +but this is in DRL not GUI? so we can change how it renders to DRL right +Sorry attached the patch before reading your comment Marshall +Which version of mantis shall i use to get into this +for the patch once test pass. +bq +Thanks Joy +Then it would be easy enough to add right +Right on +Updated to for one new issue introduced by the new commit +Juergen +Solution as discussed with developers after the last community call +Verified by Leo. +The upcomming Spring OXM will have all the properties presentResolved through and +Thanks for opening this bug report +Same applies to extra spaces are shortened but backslashes are preserved +added config parameters for JNDI registration of the Publish and added servlet which can be enabled in theThe build now also produced a which contains the interfaces and stubs needed by the caller +Other subsystems must honor this regardless of if theyve already halted other normal requests +netAn early response is highly appreciatedRegards is the newest stable version that runs on the +All unit test cases passed +When I said I fixed it I meant that Axiom support now works +Currently the sharelib is only replaced if the admin goes and does it manually or if the admin goes and runs the command to uploadupgrade the sharelib the only time this should be needed is when the jars in the sharelib need to be updated +The background thread is only started after leaving safemode +KyoI think this issue sounds like an internal JPL task especially since itincludes server file paths on ArgoCheersCameron +Max please review the patch to apply it to the branch +Will continue to look +Feel free to report them as minor issues and at least theyll be in the system. +Ive added an example to the Pelt ThemeLet me know if this is ok or if more is required +Still need to add tests of the explanationscoring for cases no norms and no deltaIf no delta parameter is supplied the math works out to the equivalent of the regular BM formula as far as the score but I think there is an extra step or two to get there +I think this may need to get filed under HDFS instead of Hadoop Common +Are there any further details available such as what the customer should do whenif this happens +Id suggest you install the Forge Tools from the latest nightly build of Tools +I lied +Marking as invalid per the previous comment and due to no feedback confirm this is Invalid +Complete +bq +Otherwise no looks good to me +This is a decision taken way before Weld gets involved always made by container starting Weld +Also good to know some more detail about your case is the spacebeing used in seg or logSomething that I think could be done pretty easily in Derby is to add compression at the datatype level rather than page level +we could fake this out by creating new entries in the read path that skip the value part and youd end up with a bunch of Results that have Keys but the value length +I am assuming this JIRA was for get if so yes we could close this +Committed at subversion revision I saw failures and errors in the junit tests but they were identical to the problems I saw when running without the patch +Thanks + +There is an issue with WAB and dependent bundle on +Commented on the reviewboard diff with some suggested test case improvements but mostly looks good +m not sure that in a pp design we can posit an omniscient the system +Uhm shouldnt the patch be committed in the branch too as it affects +Earlier this week we changed the Network Status plugin to as plugins on Windows Phone cant have spaces in them so if we have the opportunity lets standardize all plugins to be spacelessWhat is the purpose of the plugin? Does it only control showinghiding the splash screen? If so we should start a thread about this plugin on the mailing list as Android has capabilities to pluginize this same behaviour +Any update or can we drive this to conclusion +Thats only about their location in browser DOM tree and on the page not behavior +Hi I tested the above issue on a build from the Synapse trunk +See linked issue +Bug has been marked as a duplicate of this bug +Thank you I agree that these settings are pretty annoying to do +Cannot reproduce in the latest snapshot build so close it as invalid. +Same as which is fixed. +It would be best to not use the JDK as all as the may introduce side effects like that above +Since region historian isnt needed in some tests would it work to simply add a new boolean property called enabled to the class and a gate in the add method +Patch added Popup positioningsizing methods changed from private to protected to facilitate subclassing +Also attached unit applied many thanks for the fix Nathan +or other we need some example of how to use the way ie using the integration +Content of section updated asThe jBPM ESB service that is provided in the Enterprise SOA Platform handles databaseupgrades automatically from any earlier version of Enterprise SOA you have made any changes to the jBPM database schema or wish to use a jBPM databasewhich was not created by the Enterprise SOA Platform then the upgrade process cannotbe guaranteed to work +It hasnt failed since +Marking major rather than critical +Thats right +SAX isnt doing this the Avalon framework is +Indeed the merging has been quite challenging its a bit unfair really one of these days we will have to switch Ill write the flexible indexing stuff and you start doing the hard tasks Ill commit the merge in a bit when the tests finish might not get to the back compat branch if its needed till tomorrow night though +Andreas +Attaching web service and extra class used to trigger wsdl file generated with +did the svn up and saw the Downloads +other test changes look fine to me +Should we add a unit test for that? Otherwise Ill submit this to hudson +what would be the purpose of this? I guess forensics for the adminCause if youre doing the request interactively you should know what you were requesting problem I ran up against was a map that had been set up to request multiple layers like or +example gone from spec. +That will be great +bulk close of resolved issues as part of JIRA housekeeping for release. +I only brought up getters as a way ofhiding the attributes +Removed a couple of variables which are no longer useful in +Still working on the custom loader and so not included as a part of this patch +I added a note on this config +pending Jenkins +No value may be recognized as value which may be ignored in English I thinkBut in sence of Japanese word usage no value is just nullThen the sentence can be directly translated as because a blank space is null +Feel free to attach a patch and I will review it +example b b works best as it can complete a test run in all Camel code with no problemsexample a currently have issues with the component +This is indeed an Didnt even thought about the actually broken WHERE clause sorryIs there a list of known broken drivers? +Sorry request was too long submitted +The attached patch adds more test cases and fixes the bugs +Verified on ER Affects and Fix in Version +Therefore resolving the issue +But although I didnt test isnt the works the same way to handle multiple section names if only one location is specified instead of using locationS +This patch includes javadocs for the changes I did so farI am working right now on fixing open range bug on flexible standard query parser +i dont see as a huge issue +Green block represents Java heap +Hello PierreOur way of thinking is not to avoid the redundancy of the data but really to have efficient user interfaceFor exemple if your customer ask you to know the production status of his orderYou can go to manufacturing interface found the production run linked to the order and get his statusBut why sellers have to go in manufacturing component just to have the production status of one orderIf production run is waiting instead of ready as expected sellers must know easily this kind of informationI hope this explanation could help you to understand the goal of this kind of dashboard that is mixing data from several componentsKind regardsJulien +Sometimes the footnote from the second column precedes thefootnote from the first one +my database Oracle gsome sample data SMININFOIDSMINCOOPIDFKSMINCODESMINCONTENTSMINWHITELISTSMININFOTYPEFKSMINPROVINCESMINCOUNTRYSMINVALUESMINCHECKRATESMINCONFIRMKEYWORDSMINNEEDCONFIRMSMININFONAMESMINSECONDPROVINCESMINSECONDCODESMINSECONDCONTENTSMINCUTYIDSMINCANCELNUMBERSMINCANCELCONTENTSMINQINCOINVALUESMINSTATUSSMINMOMTSENTSMINUSERNAMESMINPASSWORDSMINPRESENTCOINSMINLIMITDAYVALUESMINLIMITMONTHVALUESMINMCCSMINDAILYMTLIMITSMINGROUPIDSMINCOMPANYNAMESMINBUSINAMESMINSERVICEPHONESMINCONTENTPRECISESMINSEPERATORSMINOSUCODETEMPLATESMINIMPORTANTINFOabcCxxxaaaaaaaaaaabbbbCMKxxxbbbbbbbbUnknownY will change all null string +Other than that +set is not having an effect because later calls to getUGI information use cached info instead of new +They are tipically located in the componentconfig folderThe schema is really simple +wiki news on thisIf things are too complicated I can copy the pages from the old to the new wiki no need to dig to deepThanks for delaymy userid is leshik +Please post the exact details of your scenario to the above forum thread and well take a look +The root of the problem is that client data objects use lazy faulting so a new query is generated when trying to resolve relationship +It is pretty straightforward +Hi AndrewWhat I meant was how to get hold of something that maps to access tokens for a document and for a user so that JIRA security can be enforced at search time +Is this still an issue using +patch applied in +Hmm couldnt we tranpose like would allow? Then when we switch to CQL in the hadoop support its seamless +I think the code will avoid this by overwriting the old permission settings with new permissions +Get output schema from generic type is already part of besides can also handle annotation now do we still need +New failure indicates problem is likely. +Added doc +BenoitWhat do you think of the patchI havent found a way of recording the ipport of the client +I actually just realized that the other services have additional and classes underneath so splitting the configuration into etc and keeping binaries in lib is extremely cumbersome +Isnt this fixed already? Please reopen if it is notthanks dims +Upmerged the code to the latest and trunk +This is fine in master +Last I checked the Xerces DOM code like most was minimally reentrant +I can confirm that the test still fails on trunk and BRMS branch +I think we would still need a configuration option that does not involve Java +patch looks good +Cheers I understand now cheers Martin for the explanation +Integrated to and trunkThanks for the patch LiangThanks for the review Ram. +Ok +IO itself comes from read calls which return an int +Tar is the only task that accepts an explicitmode argument +The amount of state is small enough to keep in a hastable noDo we ever need to checkpoint datanodes? Seems like that is a separable discussion +patch renames UIE to and moves log special casing into extracts as to me that DPTE is a very thin layer on top of TPE and Ive mostly been more confused by TPE default behavior than by what DPTE adds on top of that +to TRUNK +Use Firebug or a similar tool on the browser side to verify what request is actually being sent +It returned the correct value of +and seem like the best option to me +This is fixed by +remove behaviour seem to inconsistent with add +i guess that is a separate issue +Minor comment weird formatting in +So you need to start the DN as root when enabling security and running with jsvc no way around that +If you can prove with factual profiling information that your application would benefit from this you can subclass and introduce the performance optimizationPooling logic is not as simple as using a Stack +Maybe change the exception message to Failed to initialize edits log in any storage directory +Almost all attributes descriptions are ready and saved locally +Thanks for the confirmation! +Just verified the same problem occurs for me with AS +I think we need to add unit tests to catch this things in the future +I have not marked this as but I would really like to not expose key value pairs as the Tez engine API data format when we do a release +I have added the abstract with very slight modifications +It might be a common complaint from Windows users for the but its not something experienced by Windows users in generalIm using Windows pretty much all the time and SMX is the only program that Im aware of currently that does this +Applies cleanly to and trunk +Its now I added a tooltip which displays the type of the +Thanks a lot Sijie for taking a lookI also convened with you about negative check +Then I got this error + careful to make sure that it doesnt break anything in the past it has caused us headachesthings likedoes +Overall this looks great JD +I just rolled back the second commit +For now lets leave clientid as it is +The NPE was introduced by recent work on removing the +Attached patch makes it so that the ant invocation requires the user to specify the location of forrest +Ive committed a fix into SVN trunkCould you give it a try +lzo has been around for almost years nowIm okay with lzo +So this is not an issue +Could you be so kind an upload a +Oops this one contains the changes +Bringing in to so gets reviewed and considered for commit in +If you are waiting only for me you can close this issue +Please check the bug is fixed for you +Hi Kiran this patch works for any meta tag that has a property and a content attribute +Can you attach the patch so we can push this to version control +That way I didnt see the error you attached patch inserts the new JAR into ant +Problem is that derby currently does not allow to use parameters at OFFSET and FETCH while Castor requires we have to wait for derby to support parameters or change Castor logic to not require parameters. +addressed the findbugs issue +Sailaja can you post the logs to allow debugging the MS setup consist of nodes load balancer DB server separate DB server servicing clustered MS and MS MS clustered with MS MS clustered with MS hosts XEN XS hosts in cluster +is undefinedbq +But no memory leaks +As the project is for our customer I am very interested to know when a solution or work around for this issue will be availableThanks for your help in advanceRegards anybody react on this +Bulk closing for +jbpm projects are not importable so no need to have this continue to be in jira. +This is the aar for them service +This change touches the same classes that reflection changes though +Talked with Christian and he will hopefully work on this for +that makes them just like interceptor bindings so that should be easier for people to understand in the long run +That way the timestamp is not important and the object doesnt change making ehcache send the modified object throught the my case the application is mostlly read only but with a lot of users doing queries this makes a lot of unecesary network trafic between application instances for ehcache +Just throwing it out there +The popup now displays about an inch below the +no problem at all glad to helpwere any of the modifications serious? let me know if I got something wrong so I can make patches better in the futureocheers serious +People may want leading indentation to be reformatted except around such edges cases which can get +However there is no longer the liblib double nesting +A sample output of a SOAP request created using the attached patchThere are now two different ways of using a derived key to sign a message without sending the password as the old UTSIGN functionality does so Im closing this issue +No problem glad to hear you were able to track down the problem +What do you suggest how we should proceed here? +Does anyone have any thoughts on thatTwo more little comments on the patch I shouldve noticed this before but this patch actually introduces a second instance of in even though its already present +As long as we standardized on one or the other as long as we set the svn to native things should resolve themselves right? Windows users see Windows line endings and X users see X line endings +Rebased and merged into the master branch +This bug seems related to XSLT processor +This is awesome +Alex Can you provide a patch that can be applied to branch? I would like to test it +No response +Also and failed but this is already being taken care of by +The above test failures are not related to the patch +Are you using a mapping file at all or are you relying on reflection +is sufficient? If so its considerably simpler +Now a distribution of can be created by just a simple command ant +Added a first simple test Fixed a bunch of problems I had introduced when I ported the patch over +Im wondering if this issue can be closed and resolved as Wont fix now that is downloaded automatically if it cant be found where its supposed to be +bq +fo filename has been added with description A testcase driver program filename has been added with description Result of rendering the first time +Thanks Pat for the contribution. +Please give it a try +Attachment has been added with description Another error output when there are graphic references +The depchain should be used in any example that previously included the atmosphere version of the artifacts in the depchain are managed by the build bom via the + is an AS artifact and would introduce a cyclic dependency to EJB +Able to reproduce with Java Windows +Avro tests pass with this and applied +I agree with JM that returning false is weird and gives the wrong impression +Did you try the patched version that I sent earlier +Bulk defer all frontend issues to +Only one copy need to be there in svnI think the +Im on OS X +In the original version of the JDBC test there was no felix fileinstall +but if you do not replicate your conversation across the cluster you are okif that doesnt work for you you can store them in some other custom store +verified by Anton +bq +I tried to reproduce this bug However I was unable to do it +I am not sure that this is an problem I lack evidenceThe bug does not affect sfbugsites a standard sample data set backed by a shapefile +Contrib builds still have it but I believe it has to be addressed by a separate JIRA if so desired + visit comes last hence too lateneeds to refactor line number gathering with label +Verified the repro attached still reproduces the error against trunk and +In Admin UI all you see is core +Proposed patch for also resolves +It seems the whole paradigm was invented exactly to make it easy to run concurrently +This is regarding Axis Java +Uploaded the whole patch to show the idea and the HDFS changes will be moved to before commit + +The HTTP response happens when click in Edit Help View links to change mode for example +I could certainly use it +Marking the issue as resolved +Normally I dont like making this kind of incompatible change but its really now or never for a Shale release and this one is important +Other than that it looks good + +Please provide a unit test that demonstrates this bug +Thanks. +Either this was fixed for or this is out of date +In we can use to get such info instead of using +Brilliant +There was already one bug found fixed in the loop unroller of this kind +Gordan No +Nevertheless with alpha there really is something bad + +Ok with the first request gets a response second time we try axis tries to listen to the same port again and croaksthanks in svn revision . +Torquebox gets intotrouble with request handlingthat would normally bypass layouts like senddata +I didnt change the API dependencies to be the latest since we should use the lowest version necessary +Under Architectures Valid Architectures I removed armv +Committed to and trunkThanks for the patch Nate +Fixed in trunk +Hudson builds again +Also any manual tests to validate it works? Or atleast run where hostname! resolved hostname +An initial patch covering all the methods except saveNamespace and updatePipeline +There are two files under services +Can someone on Windows check this issue +Verified in +HiI need to ask why you have the in the URL in the first placeThe reason why the colon is swallowed is that a colon indicates that a port will follow after it but in this case none was specified in the URL +I dont think I will try to backport this. +Again the same basic changes will need to be made on the async side +Omitting the version also WMS and WCS no versions work +There does not seem to be a clear path to proceed on this one so i am pushing it back. +name but its a start + +This seems to me theoretically impossibleCheck out transferable capability tokens +Seems fine +Try playing around with Carrot demo and see if the results satisfy your needs +Found a different way to solve my problem and I am not interested in maintaining the patch set for ongoing versions of the plugin. +Minor fix and improved robustness in path Patch has been added with description Patch Patch has been added with description Patch would be nice if the jar for the Ant Rasterizer Task were included in the binary distribution +It is still expensive +Thanks Kathey + on all +and those seem to work fine +To run see file +didnt work at all when we started but afterward it would successfully run in devcloud +Jukka taking action first didnt allow this to happen or me to make the points I just made +had a different fix for than the older branches +Am I correct that this happens when connecting to the database? My guess is that the database has become corrupted +So I changed finalize to not send any network commands + +Thanks for your patch it has been applied! +Sometimes we are seeing sized sequencefiles as well which is invalid +please let us know if it reappears in a nightly +Attachment has been added with description Same test case FOP patched to add AIS has been added with description Same test case FOP patched to add AIS dull colors are visible in Adobe Reader but not in Evince! Could the issue be with the viewer +Ive been meaning to document the security machinery for a long time and even submitted an abstract to the summit for this topic but it wasnt accepted +Rather than using a file as input we can feed it a string of characters written into the test code directlyIve attached an updated patch with a testcase +HelloI attach the patch which attach resources from smack and smacx jars +line should update the comment for the new layout version +enable gdi graphics usage for but unfortunately broke support of +Im marking this bug as closed. +Reopening as it might be rolled back after some discussion on the dev is not a good idea as there are a few tags using id and only set using name +Thanks manFor the record I did install it into my personal repo but my company has a mirror of the official maven repo and Im not too comfortable messing with it even to the benefit of my coworkersAlso in general I think maven dependency management tends to make you lazy in that you dont bother to check for the most recent versions of libraries in the respective project homes you just fetch the latest one in the repo if it seems recent enough However I concur the Apache way of deploying to iBiblio does seem pretty tedious +That way if a version fails validation then the user is prompted to accept the version change it or cancel the upload entirely +Duplicate of +Its a runtime dependency that should affect noone so why not +Pull request merged with the upstream. +TLD API docs for UI components API docs for framework devguide FAQ were updated to API docs for UI components API docs for framework devguide FAQ were updated to and FAQ were updated to screenshots were deleted and missing were added to devguide Migration Guide was added at API docs for UI components API docs for framework devguide FAQ migrationguide were updated to FAQ were updated +This issue was reported to me on the Flexmojos Mailinglist too +Thank you Jonathan +Updated patch with the config key in to all resolved issues for. +also need to change the line infilehomehughw patch that fixes the problem +Brett CarlosIs this still a valid jira +Bulk closing stale resolved issues +Fixed +for the new changes ivan +Should that be tracked here or in a separate issue? +Extract the zip somewhere run Now it should have created a If you want to change the to run it with change the npver in and run it to copy the dependent filesNext step would be to build the copied resources from trunk and then use +manybe its not +Assume closed as resolved and released +I finally learned to do a findreplace on tabs and convert to spaces +Resolved with revision . +sorry for jira post +Is this the intended behavior or a bug? If it is intended what is the utility? If not can it be prevented? intended for plugins that provide jsstatic resources but for those that dont it shouldnt fixed +So the ZK based solution seems more appealing +But that issue had been reported in +Navigation pages in Confluence have been updated lets see how it looks on the live pages in a few hours +Tested + +Thanks for the coverage fixes + for consistency and wider support for date formats +I think this is a slight improvement over the previous one +This patch was committed marking as resolved. +Print ready file is set up to be a one color print job using PMS . +My application uses all over the place changing all these links to the pretty printed version would mean ive to change my whole app codeIf its just a configuration value I get the pretty printed json without any further changes and can switch back later on easily +If you have any questions I might check this ticket for the next days +Ive committed this +In there is still an import for and which could be removed +threw this class into my environment and it seems to be working +Making a tracker as this is being handled at Java EE level +We will move to maven thats one of our next tasks +Weve been trying to get the config easier and I think mixing them would be more confusing in the long run +Or that its impossible to create a general purpose checker thats easy for you to use? That seems unlikely but well only find out by trying no +if you have any ideas on how to fix it while still supporting im all ears. +The test has been applied with the patch of course +Please use Japanese encoding. +The sooner we get done with thesooner well be able to start the much needed API redesign +The patch does not apply because it is trying to apply it under common +If all of these are fixed then can we get updated with the latest Tomcat ? Then Ill add tests in the testsuite for each that sound OK +OK I can see where youre coming from we dont need to be Tufty +Logged In YES useridThe new line kind of problem is already fixed +Other than that all other dialects appear to work with the constraints +One last question Should be a? It runs for s on my machine +The first alpha might break things so I prefer we test right now +Please dont create duplicate issues +Yes Pig now requires antlr jars and the dependency is set on ivy +I just committed this +Im not sure how its possible but my version is solwer in the performance test I came up withVery odd +Also affects FP +Have you validated the serialization compatibility +has applied the fix to the branch and should be in ER +Overriding one does not override the other and they have their own a more positive note implementing interfaces now works with both versions which I think is correct behavior +I will check it out +My initial attempt to fix it failed as did Rolfs patch in +Aah! But it does! Sorry +Closing. +Ive applied the patch. +In addition I do not see this behaviour with or but have no at hand +Thanks +The only references seems to be in hyperichqwebtiles +Basicly I create using Persistence static method +What exactly prevents you from performing an adaptive monitoring of the clients? If you can use regular intervals then I would assume that you can also use variable intervals +bq +Hi Yip thanks for the explanation +Thanks JasonI pulled this into trunk and +Is this a feature? Im not sure how often this use case would occur +True on the regionserver contentYes on how host identifies itself +How do you use the new builds? Does it wipe out everything in the directory you specify and drop in the new files so that you need to reconfigure the dev instance when you build +I get An unknown line type was found in line ! when I try to apply the patch via +We can close this if that vote in fact does fail +javadoc patch to was applied to HEAD branch +So I guess ignore this until I can reliably reproduce the problem +Its also working fine +Ping this one Ferdy +I just tried to use the url encoding scheme as mentioed above but when simplychanging the value of the name attribute from field to field I getthe following exceptionunable to add attribute name to due to the following error is not a valid +Triaged for checked repro attached and setting normal urgency +Sorry disregard those results above +I guess Im out of nitpicks for this part Well maybe just one last nit for the road the equals and hashcode in Commit are missing Override +But one I cant understand right now is that the first time no exception werethrown +This behaviour was not desirable for us as these are portlet resources so we need to have them and their urls to be controlled by the portlet container +Indeed at the very least it could be nice to have you guys working directly on a branch in geoserver like the wcs work so that its easier to keep the fork in line with head so we can merge it back in directly instead of having to look over all the code and figure out where it fits and what has s your timeline on this stuff? TOPP is hiring more people and our first project with that team will be versioningvalidationattribution ect +With the uploaded patch there is bug with cartesian products and right outer joins when there are filters +Deleting and Archiving Events and section +We had toclose your rsync accessWe send notification to your contact addressapache +Added patch to support more exceptional because I think its present there too +The problem here is that output to stdout and stderr gets merge indifferent ways in different test runs +Thanks Knut Anders for reviewing the patchHere is a patch updated with Knuts comments +Pls change the component to core +We also should make the Serializable if possible +for my use case you can ignore itEverything is working as soon as I remove the testablefalse in the Deployment which is useless since I also annotated my method with Sorry for polluting this issue +Upgrading to svn fixes the I upgraded to and the problem was not solved +Restarting the master seems to have resolved the issue +Sorry +trunk revision revision revision +Looks like there is NO way to maintain an state that is global to JVM across class loader other than using a system property +It should write to an xml file and use the existing method for it +Maybe we could even add an option to the login module configuration which says use JAASIf this use JAAS option is not set and configuration missing this would be a configuration errorIf this use JAAS options is set JAAS is used and any existing configuration in would be ignored +Changing to EAP +Committed for in rev +Sorry StefanIve just sorted it just came on to write comment to ignore entry +Since that would involve a db update I wanted that to be okayed from community +These blockers and critical issues are resolved but not verified +At the starting of the test the test framework will validate the kafkahome and javahome specified in each host specified in to make sure that the user has already updated their environment accordingly +should extend and not the regular Session +Would revisit after messaging scalability work +stagingjobtmpstagingdaijy +What Im looking at is installing our loader repository under a different name if its not registered since we cannot install content into the domain +Just wondering what would be the next logical step +Running tests to see if they run ok +I just committed this thank you Yu +The namenode wont know about it till the next block report +Attaching a patch which would kind of fix this issue +Hopefully we will eventually be able to solve the license issue with github pull requests +In fact it looks like you have annotations there that sound identical on the surface its going to be confusing but again its your decision +Fixed Updated in stratos controller dashboard module and service stub Attached patch file with commit edfaecdacefaacadfeecea +You are right lets leave it for now +For the pb related I think we dont need the enum +Looks like Astyanax is setting cql version to yet is using executecqlquery instead of executecqlquery method +Note patch also contains fixes for jiraFix for this bug was to not rely on storing the Lifecycle object in a data member but rather to recreate for each request +Chris havent had a chance to work on yours but am going to look today as well as move over to the new trunk +I have tried this using Maven and Maven WAR Plugin and I cannot reproduce this issue. +Note that the particular test case which fails may vary but its always the assert at line +commit to trunk at commit to at Mar buidl of geronimo server java ee web application cant be uninstall via OSGI Bundle porlet. +must have been the heat or in short for to configuration admin rev +Hi DavidResolving as requested but note that we havent made any changes to the framework regarding this problem +Any news on the tests? Does swingx solve the problem +Can someone please review and commit +I thought about this issue again +This created all the new methods too +Improved patch +Against svn +looks like what this issue needs +So if throttling is required for deletions we should implement it generically either on the namenode or in +left open as a reminder to double check on first +Thanks Arjen I did try with the latest snapshot and can still reprom attaching I used the for all of the following maven artefacts +Looks good +This is not the latest CVS as the class has been replaced with theSo are you using a different version than or did you not update +And as we know in the laggage check we can only claim once and as soon as the laggage has been retrieved the claim check is no longer valid +Dont need to reassign +These are the review comments for the Guide for the actual content from the ESB guide ie its the same in both and I need someone to verify either that it is correct or provide from info on what it should be +I wonder if that could be expanded to having the SLD param be able to take multiple files? Ive never liked that you can define a bunch of styles in one file in my mind an SLD definition should contain a single style. +It seems to occur whenever a spammer fails to modify a page perhaps the is messing something? It also seems to affect all pages at once so its probably a caching issue +This ticket is pull replication is correct +Nicholas if you are going to check for all states in you might as well move that code into +will be rebooted as JDK eill be the minimal required JDK version in order to take full advantage of modern UI features. +bz file rather than the xml +Hey Nicolasthanks for the feedback and for the patch +However there is some uncertainly around that cost which is mitigated by to the branch +Therefore I dont know +Revision +can you adjust the title and expand on what this is? BTW it has no estimate compared to the others created +mvn X clean Install on rev for webui there someone on windows who can try to apply the attached patch on gatein portal trunk to see if fixes the issue +Complete maven script to reproduce the issue +Browser compatible redirect strategy for HTTPI agree to include this file into the ASF projects and distribute it under the license +I just committed this +Sorry I meant +The file is available for review +I also tweaked IRBs input handling in conjunction with so maybe this is less of a problem now +So there isnt much reason to use them +Row keys are fully optional and backwards compatibleWhat do you mean by thisFWIW I think this feature can be emulated today using configuration and rowkey prefix management +Fix the warnings +Attached example enhanced jpdl file. +Close out issues. +I dont have a full understanding of request handlings internals but I only experienced this when it takes a bit longer for the backend to answer and the user gets angry and tries to click on something else on the screen +Olga also pointed out that we are getting rid of some optimizations in streaming with these changes and Ive added a section on that as well +This should be resolved but is left open in case there are final tweaks with the TeiidModeshape runtimes for final +The build is unstable since this commit. +I only added support for the mapreduce plan since that was all that we needed +Logger is added +Add javadoc and clean the the patch and fix another bug +after release +Can you Click on Submit Path button +Hmm there was still a hidden copy of the Yamaha manual in the test file +Attached is a patch that migrates a version of the from the FTP project into the SSH SFTP server and provides hooks for virtual file system usage +Looks like vfs related +More infoI left broken instances in the cluster and the Brno lab admin noticed that there has been going mbps traffic to everyone in the same virtual LAN +JPA is to a large degree a functional subset of JDO +Hmm ok this is another place where consistency is failing us +jar files in lib + is now out +Fixed additional compiler warnings +Ill fix it for Beta +Reverted the type declaration to allow only tuples inside bagsTODO +and report back the length of the ELEMENT output +For example if the selector were to identify a sink and the runner finds that the selected sinks process method has failed then the notion of failover policy will leak into the runner implementation +Hi Lorenzo have you really tested this with? With this extend mapping is clearly not possible as we added the support for inheritance with you please test what happens with the OQL querySELECT FROM e WHERE like +The use case for this is as follows +If we cant run it at Codehaus thats fine and no problem whatsoever +caching is not necessary sorry not having removed it would be great if you could make the example work on ACF this makes easier for us to compare with expected behavior +small changes suggested by Corinne +Thanks Pradeep for contributing! +Probably this one +Before dereferencing the bit pointer I added the line addp ri ro where ri is a register for storing the converted pointer and ro is the register the bit pointer is stored in +We cannot do anything with this issue because this is system behaviour +Long Objects Long objects can sometimes fit in smaller boxes if they are placed in the box at an angle +Can you explain why you are using this layout? I am always a bit concerned about making everything a property +This is actually consistent with all of the other DFS methods that accept a PathAlright that sounds reasonable +Ive confirmed with my other source that the service does require recompilation against the +The difference here is that when I tried to repro I started with data in the cluster +Failures are not related to this patch +Give me a sec +Ill need to get someone from performance team to take a look at this change but should hopefully have an answer soon +Steve can you assign this to me? I have some patches Id like to submit for the lib and example +thanks! lets resolve the issue for now +Bulk change resolved issues before JEXL. +I have committed itSending Sending srcjavaorgapachesolrrequestTransmitting file dataCommitted revision Thanks Jun +Thanks for the patch Dag +Ill be a JAOO so it has to wait till Thursday though +Closing old issues +This will probably be set at the application level and apply to all Component within an app +I the test after adding the close calls both before and after applying the patch to and confirmed that the test still demonstrates the bug by failing prior to the fix and succeeding afterwards +havent we fixed that already +added a delta patch to do cleanups mentioned by Bikas removed unnecessary logging comment. +Also tested on using cmake version +But they wont work until UTC +or torquebox deploy myrubyappWe obviously need to improve the error message and will use this JIRA as a chance to do so +I will deploy juddi in the default config and get another run going on hudson. +Patches were applied at the revision +patch containing the required changes and the appropriate tests +I the repository and after a while I was able to search for these artifacts +So that we have tow solutions change the configuration of SMX patch Xalan that he releases its readersShould I make a subtask for the remaining issue or do you have a solution for that so that we can close this issue +I assign copyright for this patch to the Apache Foundation +duplicates +If this is the ticket to track the fact that debian packaging is broken in master perhaps the subject could be updated +Mind to open a jira for that +I just committed this +Please check that the patch was applied as you expected +Fixed file attached +Forgot to remove the commented out lines +heres a patch for all the copyXXXXIll look at append now +Actually used retrotranslator and mimicked struts. +Let me know how we can work togetherThank you +Ill attach the fix patch for now as this is a pretty severe bug + +Target +I think this is the test that is causing this port collision issueregards thanks for the patch. +I didnt manage to reproduce this with source from cvs can the if this is still a problem +And instead of you could just use a precompiled regular expression +Youre completely right now that I think more about it +The fix resolves this issue +Thanks for pulling all relavent checkins +Applied to revision +Resolved fixed created on JBT. +implement hashCode better require equals and hashcode as part of the interface same as about a version where the tests actually pass? Will commit shortly +Is there planned to add a menu entry like Maven Build +Would you please try itBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Committed to trunk +Thanks for committing the patch +Thanks Kihwal! +patch includes a unit test +No Im starting with server side Java classes carefully written to be exposed over WSServer Java Classes Client Java the bug as there is a work around +OK Ill give it a try on and open a new JIRA if any changesenhancements are needed +I intended to test this feature with online snapshot to see if it can reduce the flush time +My bad. +as its not a blocker +The most likely reason for that is that the key used in the lookup is not equal to the key used when the connection pool was created +We are using version and we have seen this issue several times recently when we are preparing a new release for our product +have been improved for CamelAnd if you use decodersencoders that cannot be shared with multiple consumers then you need to enlist in the registry etcThis is not a bug but how its designed +gitkeep whatever ever works +Yes the work is already done for such features but Im waiting of Jedis to push the changes out +the same issue although it is showed in todo example +Unfortunately a new version of doesnt help +No warning or error message but the starting process of modules seemes blocked +I backed it out +Verified by Ilya. +Closed +has always included the wrapping tuple as part of the schema +Also fails when clicking on the pencil to edit +Ill run it thru Hudson again right now +Bulk close resolved issues. +No that wont be run Failures Errors Skipped Time elapsed +Im not exactly sure what you mean by operatorI mean the person starting the new nodes +duplicate on for all these urlsdtds +Added a version number to the extension namespace. +Theres a small window of time in which neither old nor new data is present +Closing this because it doesnt belong in Jira +Resolved in commit on branch and on trunk +In computers can be added in the vmhostinuse state but there is no way to put one in that state later +has been added with description Test case for auto +Fixed. +At least they do for me +Ben could you please update all on after the upcoming RC release +Move the config key to we dont spread them around nowOn further thought doesnt seem like the right place for it this is an MR config rather than YARN given its part of the project +Thanks Gustavo! +In items and selection is put to components that may have their own scrolling +Committed to sqoop +I wonder if already covers the simple case. +Please verify it. +This allows functions to return BOOLEAN return values +Closing as this is already fixed in master +Wouldnt it make sense to create a Maven multi module project that builds and installs the projects? Not sure how to go about here +I dont know ifwhen the original hyperic product will include them but we have control over what we distribute +Thanks for the report! +I havent racked down the exact cause of the problem but full support for bridge methods would seem to be important +IMO this issue relates to a big issue with Message conversions and its payload tofrom +has been marked as a duplicate of this bug +Changed CHANGELOG and release notes. +Details can be found in the updated users guideCan we close this ticket now +This also sidesteps questions of file handles breakingHow common are SSN restores with data loss? I had assumed that after a file was closed file data was as durable as NN metadata on disk +Makes perfect sense updated the call as you suggested. +You can explain the implementation better than I can for this +No you just need to register a provider which is capable of mapping all the objects you are dealing with tofrom format which needs to be supported +Thanks +I will create an issue for which is more detailed. +Hi Robbie would you mind reviewing this commit +Patch applied to trunk and SCA branch +Committed at subversion revision I dont know when this class ceased being public +The benefits of the url base are the ability to use back buttons and bookmarks +A nested +RobThanks +I applied above fix in revision +I hope this time it is really left close +Should we reopen this JIRA or create another JIRA for this +Further testing shows the improvement is marginal since the patch makes the code more complicated Im just letting it go the way of the dodo +each directory additionally need to store directory sizeNodes are stored in the depth first order +Marcin Ive already told you to NOT create JIRA ticket on your ownHmm Ive read about that on wiki but I dont remember when you told me about it directly +build test +But also it is displayed on the console +New tables required grants in postgresHowever the build is now failing for what appears to be a different reason +If we set proxy target to true to makes the jersey resources register them properly +The patch has new classes only +Post here if you find anything useful +Unlike dont maintain any kind of selection state so what would clicking on the just content area mean? To a large extent exist to provide access to a menu popup +Thank you for the clarification +Use this instead maintains formatting with current srcAlso includes in with svn revision . +I attached the patch to the list +Reading through I notice that you may have had a WIP patch that parallelizes the region creation for performance +will break things too +Fixed. +This will be a killer regexp I can feel it +If you need assistance ping me and Ill help explain what needs to be done here +workthanks meant and not problem occurs with most of the Axis types in +Master is fixed in abfb +Wow +heres the for the patch +r merging branch into trunk +complete different implementation against trunk +Also when we are using email validation with regular expression error message looks very wired +I think that moving this logicfrom to is the right way to go I had triedseveral variants that were quite similar to this as part of working onThe code is simpler and cleaner after this changeMy only substantive comment is that I think with this change you can nowencapsulate the thread list entirely within becauseno other class now needs access to these server control internals +I have not looked at the patch yet but in general I like restlet +I know I have used Struts forms in the past that allowed the user to enter commas in form fields +This looks like a reasonable change +looks goodNits All the startstop methods should be private test could use a small javadoc + Id rather not do that and leave to the search front end to decide on what do display when a proper title is missing +bummer +it could this automatically expose this in JMX? Looks like it does but if not would be good to have these numbers there too! I have done monitoring of segment sizes with polling and polling of the dataindex directory +Hi MarcoIts just to know if you got a chance to work on this issue because I expected to put calendar in in release which should come soon +Having the mixture of styles makes the code less understandable IMO +Also what url needs to go in the file +Note that the Subversion Commits tab for this JIRA shows a Dec commit for that used the wrong JIRA in the commit message +Bcause we do not update header and footer elements but just table body in order not to loose focus. +Im gonna close since something similar can be done with JSTL tags and Struts istrying to avoid overlap with JSTL +diff as promised +And there is a subclass of the class that makes instances of this new Node class and adds the location also included in the Node class a setAttributes to allow groovy statements to set the value of the attributes +so for example to implement this in fuse I would actually have to write logic in c to decide if a uidgid combination are allowed to access a file +Yes it would certain seem a bug +Yes very likely +withdraw +Thanks ChrisUpdated patch attached +Fixed under Revision to merge to + +Hiranya This seem to be an excellent piece of work +Process and WSDL that when compiled demonstrate the incorrect error +No objections to trunk orI dont think we need it in but Alberts the judge jury etc +I will dig deeper to fixed it completely +I understand problems which the disallowing could causeWhether you think that the current state is OK please reject this JIRA +We will try Solaris ZFS when the disk quotas become reality +I will fix the spacing in and move it to +The is meant for simple types with a single string representation nothing more +Also when modeshape is installed the predefined supported repositories will be automatically generated +Good jobIt would be great if you can change to run the test +If I use existing workspace I change path to the server from the latest build and have the same importing and correcting I renamed my project and all related from Seam to newseam and deploying was succesful wo any errors +This was mostly fixed by moving the facets module from payloads to doc values +Can you check on that +for the to trunk. +I am experiencing this problem and it is a major pain point as I cant share any projects with the Derby nature without manually removing all the classpath entries +Portlet library can be either user librariy or provided by server runtime. +The schema files only contain the and which depend on these primitives +shouldnt the if be a while +The logger was not returning these +The way the CS is setup this is an extremely rare corner case which is hard to fixThe essential problem here is that in your setup the queue has so little capacity that it cant get more than one reduce slotFor now Ill downgrade as this is a corner case as I think about ways to fix this. +The new ones match the modular toolkitThe evaluator can be run on a datamodel +Attaching a new tarball with the monolithic spring jar replaced with the module jars required for the application +I added new type directly to +This one properly transfers configuration properties to service properties +Recreate the patch based on the latest trunk +A long in binary form makes sense but in JSON an ISO string might be more useful +Should be marked as duplicate of +are and design decisions for a software system directly dictated by the end users of your application? If I am going to add a new feature I would notWe really dont know the impact +Sorry about that +Ill went through code and it should not be difficult to fix this +This is superseded by +ThanksI see +Thank you +Theres no way to invalidate constants on a more local basis +the dba is the user who is attempting the upgrade +See. +Jason is this still valid and needed +I wasnt sure if it was just meant to be something where you could see the sources file in your ide if you needed to debug something rather than an artifact like the separate src build we already have +If a class has its own deserialization constructor it must have its own custom serializer too +Dan please let us know what company you work for so we can avoid that place like the plagueYou are obviously having a bad day year lifeIf you took the time to actually READ the comments you would realize that for MOST users of Lucene the performance overhead that would be required in EVERY CASE in order to allow index recovery IN THE RARE CASE is not worth it +But if so we have no DBO role at the systems level and anybody authenticated would be allowed to perform system level operations +The attached test case is working in and +Remember to update the version in the for CXF to be as minimum range +We can not code specifically around json and json and simply making a subtype compatible with a produces unexpected results specifically applicationxhtmlxml becomes compatible with applicationxml and I think it is not quite rightFor the moment you can make such compatibility checks work by setting a contextual property to true for a given endpoint +heres the patchfirst you must svn move lucenecontrib lucene +Thanks Xuan! +We have decided against renaming it +is not suited for long running ve done like in Android IJ plugin . +The include path prefix is excluded from the class references +It works +Can you please test parsing the file using Xerces without Xalan being involved +Attached is the log file +Thanks Colin +Ashish Vikas Pranay and DiveshThanks and Regards Akash AkashI have reviewed your patch +For the jmsReply +Or if I wait for seconds before starting the nd AS m using nightly build +This actually resolves part of the reported issue requesting that the service and port be listed at startupBut as this suggests though I have set to in varconfig the service still binds to the default port and all other dependent services are attempting to bind to at the configured port So s configuration is not unified into yetAlso the etc file still exists in the latest snapshot which contains supposed configuration for +i committed the visibility cache change but not the test +My doesnt have any conflicts when I svn update and it builds fine on forrest svn diff generates this that to me looks totally messed up and out sync +When such a situation happens it will be very hard to debugIs it possible to address that problem? I understand it is not straightforward because of input streams +Reopen for fix in to +This would point to a tagged release +Added wrapper for the injected that handles fatal exceptions. +And Nailgun issue could be caused by JRUBYOPTS? Can you show me the result of echo JRUBYOPTS and an typescript of jruby startup +Yes the Web UI was used for the domain name the first application at least that is how it appeared lots of people scrambling around +Did I understand this right +Can anyone suggest a for this in the meantime +revision Maximthe second part of the issue here was to implementtest the for Screensharing +Will commit if tests pass +Any errors with it should be addressed as new bugs +With that I was able to get the test go past line +applied get that same error in windows +The reason is is used by a ton of people and its been a source of bugs in the past +I dont think we want to do more for +Openejb may need to do some more deployment activity to pick up for mdbs +Updated the engine code to use annotations and query debug instead of logging but also added logic to log the annotations and query debug when at the detail and trace levels further cleanup to +I have left the behavior of with reguard to LOCKTIMEOUTLOG the same so it will still immediately throw an exception in this case rather than retryA patch very similar to this has passed all tests except for failures in which consistently fail on my machine against a clean codeline on the trunk i will run a full set of tests on this patch tonight +Ill attach a log file excerpt +Rough count +Will get this done over the weekend if not tonight +using Data Tools clientAlso tested with and errors out with same message Error Incorrect syntax near intkey +So I suppose that you guys will be happy even if I close this ticket as wont fix. +Aiming to get this into M +Documentation patch for the new properties +reuploading patch to trigger build +If the recent fixes made are going to fix this problem then please mark the defect as resolvedfixed +I am pretty close to finishing the last round of changes for release +Bob bumped to for now +hi michael i would prefer if you would provide a patch first +Both pom and readme files can be filtered in place might make for easier maintenance then a separate file just for the readme +And when the patch is committed please put into the commit comments so svn will back link to this page +To rephrase its the case also for the UPDATE statement we should do something for both or none +What version is this fix applied? Was it just closed and not fixed? Even though there were only a couple of votes for this issue it seems to be a needed feature +A proper fix for this problem would probably have to include one of the suggestions in. +Both builds have been appreciated! Thanks for the speedy fix. +Your guess is as good as mine +Verified that now works for both formatter and +I found out that while this patch was on review weve already have via Analytics component + be more specific I need the rights to CREATEEDITDELETE all Jetty Bamboo Plans. +Trejkaz Actually the JI specs should be run with rake as in rake specji +Updating README to reflect the current build instructions +may be confusion since the same http url path represents two different file system paths +Checked Julien thats perfect +I think that if we do it we should combine multiple booleans into a string instead of an integer +Im checking is a it the randomness set max too small so smaller segments were created than the test anticipated + +Do you have some micro benchmark data to show the performance improvement +and then they get rendered as links which makes the comment more useful. +Defer all unresolved issues from to +Need a bit more time to figure out why that is +This was probably fixed during the refactoring of the receive inventory process if the issue appears again please create a new task. + for the patch +bq +seems like a duplicate of this +Closing as requested. +It does not have effect +Slawomir Thank you for the excellent reproduction! I have successfully reproduced the issue locally and will test my classloader fix now. +If the calculation was done carelessly you may end up not being able to reach the desired goal of +However one thing I realized that I missed is to also capture max timestamp of counter data being streamed over from the other nodes +moved get to +Recently Tomcat did the same for their project +Bulk close all issues. +Closed as fixed for +Giving this a bit more thought +Patch against trunk decided to just log a warning and return default values +This was highlighted as a BLOCKER for ER +Still have problem If I set set in web interface works but command line not If I set set in command line works but web interface notI have to set in both file to make both cmd line and web interface work +Here s what i have configured in my crontab homelresendetuscanysitebinsync homelresendetuscanysitebinsync homelresendetuscanysitebinsync homelresendewookiesitebinsync homelresendewookiesitebinsyncIt does sync Tuscany and Wookie on different times +The new work with the joint compiler removal that I am working on takes special care for these situations +Manager already renamed to policy Utils seems to be a better name for this class seems to imply external usage +This problem brings up a number of questions +Hows this? Not a gitster so not I got all the whitespace +Also we cannot properly collect db creation scripts for our application since our db scripts are distributed among functional modules in all three sub projects +My fault +N did this no? I think we can close this one +seems to be missing as are the IIOP CORBA classes + +svn commit m fixing problem that was avoiding the identity facts removal srcmainjavaorgdroolsreteoo srctestjavaorgdroolsintegrationtests srctestresourcesorgdroolsintegrationtests srctestjavaorgdroolsintegrationtests srctestresourcesorgdroolsintegrationtests srcmainjavaorgdroolsreteoo file data +Closing the bugmysql select from usageevent where id id type accountid created zoneid resourceid resourcename offeringid templateid size resourcetype processed virtualsize NULL NULL NULL +very much RickCommitted to trunk at revision +Currently distributed log splitting and replication do notwork yet +Thanks for the patch +bq +Hello TonyThe patch has been applied at revision r +Updated test with the correct package +WorkaroundAdding the contents of the Fuse Mediator Routerlib directory to the camel component and updating the manually can be used a temporary workaround +A few other unneeded imports have also been cleaned up +I prefer optimizing ma fields than adding new field +Ill manage internally this +I am also confused by this limitation of spring +I used and Eclipse Debugger +I proposed it because I didnt find a manner to create a node through webdav wih a type declared by me +This may not be an issue with the new design in +Can you get a heap dump? The option XX might was able to generate the zip from live instance that just now threw memory error but upload is just taking forever +I will keep this open to address issue with static typing +Make it so we can make use of the conf passed on the command line +This still seems to be an issue +rev +Okay sorry about that +There will not be a moving to for which exposes the new operation and what is the name of the operationAlso did you expose the functionality via the Profile Service as well or only JMXThanksIan +There was a problem with instructions order upon adding a new category we mark that the reader should be refreshed after we add the category to the cache +Since direct invocation of Hibernate Validator leads to the same result Id suggest raising this at the Hibernate Validator forum +Another idea would be to improve validation in the toolEither way this should not be classified as a blocker +The FATE operation would lock the table take tablets offline and then directly modify the metadata table +Please give this issue the priority it deserves +Are you sure this problem occurs when you have installed +Im currently doing a lot of work in the transport area so I will add this as something to fix +Patch to remove the old instance of at srccommonsgrailsutil looks like it was copied to srcwebgrailsutil during Graemes refactoring for and the the old version was not tests pass after this copy is removed +Nevermind +Could you paste the rules? or the offending rules? There is nothing in that stack trace that can help us you need to unwrap it to see what he root exception is +There is something called unique snapshots in Maven +Patch applied thanks for the great test case. +Having a test for would also be useful +Ok I have deactivated the test since the code which is triggered is not yet enabled in the core I guess disabling it is valid! I will notfiy the author of the code about it +Assigning all documentation related issues to final since they dont represent a change to the current documentation should be updated in Default Sort Order adding this odditiy about how to specify the sort order on associatings + +Nevermind +There are now two tests that fail all of the time one is the one thats giving us the problem and another is in the other test fileShould give people a good place to start finding the real problem +Im working on a patch for the first approach and should provide it soonWhat do you think about generality for this patch should it be as general as allowing cross rack dependencies or to be restriced to only cross node group dependencies? Thanks +No worries +HiThis is a more karaf side issueBtw if you dont wanna pick up proxy configuration from mvn you can editetc addfalseAnyway this isnt a servicemix issue +For what its worth I think its a bug alright It looks like when a required field has not been the error degrades to a primary key violation +It only logs the warning if such a message is found after reset +Isnt better if we enhance the replicationhandler so that one instance should be good enough for all usecases and disallow multiple instancesThats like saying should we enhance so one instance is good enough for all usecases why try to shoehorn thins like this? why not let multiple instances exist? thats a major feature of the API that you can have multiple instances and configure them differently +I believe this is a dupe of fixed in Surefire +The iWorks parser will return an empty document for them and set the content type to application +Well see about getting it fixed for and perhaps if its released before then +committed the patch to trunk +Ive just committed this +for now the username will simply just be replaced when logging in +Were only creating one cassandra service +I guess we should forbid users from doing soThe solution could be disallowing users from registering multiple single handler can be configured to be both a master an a slave right? yes +Related Incomplete feature for deploy command in +I understand theres no unit or integration test yet for jelly coreThis is annoying but maybe too hard to solve for now +All occurrences should be server +This is the fun part +A move operation will just impact the moved entry +Added a default jdkName based on the system property. +is one of them is a second manifestation. +I have committed a fix to the regression in this JIRA causing +Since the two concurrent operations are entirely symmetric there is no correct way to merge them +OK +Steps to reproduce added +To my thinking we should use project to refer to community activity the PMC mailing lists subversion Jira etc +Contribution send to me by James Adams +Lets try to complete the list and do these deprecations in +Can you describe a scenarioRegarding the LWT solution +bootstrap +Attached is a UBL codelist schema part of a much larger schema hierarchy that will reproduce the problem +thanks +Moving to +The relevant info posted into the comments will pass to the official doc? Some time read a thread with comments to find the right answer is annoying +We should be migrating towards the Collections where you start a bunch of nodes and then call a create command with the collections API as your first order of business + race condition DEAD +Looking at this onThe issue remains and Ive confirmed that it fails with or without readline loaded +Added deprecation warnings for dus rmr lsr +Incidently +Its ambiguous in spec +Thanks Kristian +thanks for the patch Jason +Hmm it seems to me that the documentation was modified within the last two weeks +Can you switch out the stack traces etc +Thats a difficult invariant to maintain on every OS every libc etc. +I think the message HTTP status code must be enough for AeolusConductor guys that something is wrong with oVirt installation +Im going to add the needed info to the bundlediag command the listener should already be in place +DavidThanks and it is ready to be writable +We added some text extraction improvements including some fixes for the handling of +sorrythanksdims +Reason Connection refused nested exception is Connection refusedI wonder due to pooling becomes slower than Ajax client is it true or my misconception? it on purpose that starts with a poll even before you have the chance of setting up a listenerThis usually means a client initially has to wait timeout seconds without having a chance of receiving something +Im closing this issue +Closed as duplicate issue. +Cool +Only for cant be used in FF +looks simple and safe +Sounds right Scott that if you specific at the target then we should be trying to match with at the service +close off release +I just reenabled the test cases to run against all jvms in the trunkmibmsvn commitSending javatestingorgapachederbyTestingfunctionTeststestslangColTransmitting file dataCommitted revision +. +Please close if satisfied +I consider the feature difference to be trivial unless someone has a concrete case thats breaking because of it +I added warnings about this in methodsI think thats good +The intention is that with this patch the basic role functionality as specified in should be operational although I plan some patchesRegressions ran ok please review +Fixed in r. +Note that such check must span multiple repositories An organization may publish their artifacts to a their private repository but the artifacts could have dependencies both in the private and public the scan were performed on a single repository only dependencies from public repository would have to be replicated in private repository to pass the check which would a waste of bandwidth and resources +excellent point +Can you verify? Thanks +I see +Great work +removed extra this +See comment in attached patch +I have made some progress +The list of globals mentioned above which will be saved cannot be retrieved from config properties and so we need to communicate this to agent via global parameters. +Now that I understand the use of those in test packages I do not think they are very important for Sonar analysis +Ported from trunk to branch at subversion revision +In any case the XSLT is not the appropriate place to fix this. +I reopen the issue cause should be roadmap how to fix the bug with exporterThe issue should be adjusted according the way to fix the problem +verified for for +According to section of the XSLT specIt is an error if there is no namespace bound to the prefix on the element bearing the or xsl attribute +Clarify where copyright notices need to be +I was expecting this unfortunatelyI suspect the same it is unlikely that standard JMX will be supported in any case I have switched my efforts to Hyperic forumI have the suspicion that their JMX plugin is quite customizable and supports classpath augmentationI have read some related articles but havent so far reached the endI will keep trying. +Yes the bean is +I have read the IRC It seems the assert was triggered on Solaris platform? +Everything works fine +I canot reproduce have fixed another problem with maybe they are related +was broken by thisThe new patch should fix these problems +Fixed. +This exception is reliable only if the same file channel instance is used +Set priority to critical since changes made in conjunction with this issue broke regression tests +Thanks for the patch +Marking closed. +Its possible that a delete of a table could be interrupted with regions and the table dir left imcompletely removed +fix is good +This issue will now causes problems in the ability for JBN EM to services note did not exhibit this problem remained constant across app server restarts +Of course it should not drop columns +Nope as i discovered the the reason is different +i am attaching a sample synapse library and configs for reference herewith +can you please take a look into this? Thanks +Thanks Amila for the patch. +list file +need more info on this +yes it looks like this parameter is not properly merged +null is the smallest value so having it at the beginning is ok I is a little bit more we could do to get rid of a NPE here +The check for a directory is required otherwise every subdirectory without a known filename would be classified as a nested I might make this clear in the docs as a known issue +Attaching patch +Simplifies a lot of things +No point creating the auth area until we have accounts to go in +Changes the +Thanks Bobby +This is a build problemI have been trying to create a local buildWhich versions of Maven and JDT are used on Hudson +Was this patch applied to the line +You had a problem in the classpath an old class was still loaded in memory +Linking issue to it be OK to commit the attached patch and address as a +The original poster could verify this using the nightly build. +project so that eclipse works out of the box +Basically just build your full jar and if you see the slowness youve got a bad commitI can try to do that later but Im helping iwth a production issue right now +bq +Fixed in duplicate jira. +Closing old issues +Sorry. +Slice sampling pulls a subset of the original samples whose density matches the area under the second functionThis is interesting because it lets you shape a set of samples according to a fixed curve +Thus this issue is not a bug +Needs to get merged with the branch and to head +Committed this to trunk and +So this is resolved good workI can also run the test many times without a you have the time you could backport this to the branch and mark this solved for too +thanks hong. +Rebased and merged into the master branch +Not for commit here +Im not able to test the C extension library on my laptop let me know if this test would be valuable and Ill perform it on a Linux boxThanks much for the contribution! +Feel free to reopen it if you can explain your idea more detailed. +Attaching updated patch and so that precommit tests pick it up +Im running derbyall now +Thanks Sture I just committed the test document fix! +Incorrectly placed model package. +Thanks Ted +Otherwise you are better off saying the behavior of the and all Validator instances created by this factory is undefined after close is called +committed +I dont think this is any better than the directory I had already submittedIt not really approved because the community felt that it did not offer much over the standard OS file system cacheMy tests showed it was better but I think this would fall into the same problem +Looks good +ve just committed thisand also Ive just removed counter commandBecuase we dont have a counter. +We should add some tests before closing this perhaps +Closing all resolvedfixed issues of already released versions of Roller. +You shouldnt be able to + +fix version needs to be changed +Hi Patrick BarryI do have same issue +How about start deprecating it inI disagree + suggest you open a new issue and request to backport the change. +Patch attached +Marking closed. +Please have resolved this issue as the requested changes have been completed and the modified document is now available for review +Ive gone with what I suggested rather than introducing a special overwrite flag for this operation +Recommend moving to as it incorporates some threading fixes +I do nothing special but follow the directions in the docs the screen just arent the same +to trunk +The only reason to change the code is to remove the variable from the method signature making the method view neutral +I committed this +Once it can be reproduced it can be solved eventually +Committed to. +Pushed to master +If an application requires a file have all its blocks local it sounds incredibly flawed or it shouldnt be using HDFS +Ah fair enough +Resolved as part of +Its really a problem withour layout approach +They have to build it themselves anyway so they should build the dll with the same environment as their applications +Part of the problem here is that were updating ledger metadata on recovery before we actually write anything to the new bookie +Add the support for Asynchronous on the classb +The proxy must implement Closeable or have associated invoker with Closeable is the new requirement +We originally were building it and then ran into problems where bit jvms were being used against packages built on bit machines +Because the solution of these exception is not in our hands meldware groupware applications are thirdparty libraries I created description of required setup in +Thanks Michael! +Hope it worksThanks for all your help Rick +The should also keep track of the number of events delivered to each listener and the time they take to process those events +We should bump all of them +This isnt common at allOn the other hand theres also this large rearrangement to fix the detection of SM which didnt have the anonymous function option +Ill look at this today +The changes look good to me +Resolving this issue cannot reproduce for now as we havent gotten information since August +I missed the aspect +I agree with your assessment +Shall we reschedule it for +Marking as Fix Version because this has been ported to. +though I am more than willing to postulate myself as the official wfs datastore maintainer my spare time these days is almost null +after what Ted mentioned is fixed +Committed support for date type in corba binding generation and idl generation and testsEdell +tgz +Thank you Amareshwari. +Hi LasanthaThe prob I am having is that in it sets the as the value of portInfo to the targetI could overwrite it in by providing an Axis version of the configureEJB method but I had a hard time to get the from there +Further plan reading property INI and XML files from first step file jar +Part of parent ve decided to run the tests as part of third party set until will be resolved +But hopefully with caching authentication traffic can be made moderate even on large clusters and who knows it might be feasable to include it in the namenodeMostly what I meant to say is that folks are used to having a user database and authentication system bundled with their filesystem and that the Apache Directory Server may be a way for us to ship a secure user database and authentication system with Hadoop +for the code changes +AFAIK most of the scripts are executed through jclouds during node creation +All modifications are commented with wf +HiI committed some small cleanups and dead code removal after Clover analysis this morningOne thing we have thread locals for and +Its not the same issue see the thread name +Have I mentioned that ivy is a big problem +version for trunk +Please close the issue +Thanks Cyrus +bq +I upgraded to Castor and found the problem to still exist +Digestfile and support for and +Well as its client project and sonar server is also on clinet side so I cant share it +There hasnt been consensus as to which version would require zk as the minimum supported version +Can we add some special keyword before this data type definition so that parser knows that this is designer or enterprise type? Also should we pull in data types into schema +For our use I will be subclassing the class and putting null checks there +Chris Orders have no impact on financial information +bulk defer of open issues to +I know +Hi RonWe still encountered the same issue after upgrading to +I personally can live with the workarounds or patch on my own without an official release as long as I know its a priority for the next release +Ill to see if is indeed the culprit and if I extracted the test case correctly. +Ive implemented one option for making the type system more flexible and the changes are attached as a patchBuilt and tested on XP with and RHELThis version allows brand new types to be added after a data object has been created however existing types cannot be modified +Corrected the documentation +Changed the package name from to since the code was updated to work on Struts instead of +Please verify thisIs your intention now to submit a patch to correct the broken implementation? Will you do it on this issue? I dont want to resolve this issue since the bug is not really fixed just identifiedThanks +Is it ok if they post the patch here and I commit it? Santosh Malviya has that patch ready +Test APP with IssueTest Image which fails to load +Ive also written up some documentation describing the new LDAP configuration and the properties inSorry for the MS Word format +The attached zip contains classesTwo are analogous classes for generating SOAP wsdl files essentiallycopied from the existing SOAP classes and updated for SOAP the third classis a new utility class derived from for generating WSDL messages for only a single root element rather than allelements in the given schema +Thanks! Jess +Its working for me all the tests are passing with this applied +But the JCR Resource Resolver does and that one can provide an which can inject the workspace property into the based on the request attribute or parameter or in the case of even based on the request URL +Same problem applies toThis detects the copyright header and then buffers everything until it matches the rest of the licenseThis could use the same approach as the case. +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +testsuite results when using standardFor compilers Check output of all modules Compile all testsuites Startup server and ensure no errors Run Testsuite and verify results +Attached a patch +Removed last comment it was meant for another Jira applied +Id be happy to work on this. +Would a unified diff be OK instead +Is this fixed in some other issue or its still in progress +Also renamed column name values to be more indicative as to what they represent +It was slow to the point of being dead though ping was responding +Will have a chance this week though +Pending resolution +lgtm +Which caches the now value +Committed revision +So using preftech is just a +Youre only gonna chose one of them +Hi BrettForgive my ignorance I was always doing a prepare in advance of the perform and the perform read from the properties +And a link in +Release NoteTo allow greater load tolerance in the default configuration the attributes ofvalidator and validator are now defined as and milliseconds respectively and register is set to are defined in in CR Doc is OK The issues was not observed in CP yet +Initial cut at this ticket +Bah yes of course +you mind trying sAWSECECg and see if it still works +ova file to get file size after untar ova package +As an aside I noted that when searching for the string in the PDF version of the documents I didnt find it +Action set not a clean build all my groovy files are marked with error in the first line package +If we actually distribute this plugin we will have to include in our NOTICE file the exact notices required by the Maven Plugin license +but they are all marked optional therefore including as a mandatory dependency does not represent a significant addition to the footprint of any application +The example bundle using the package name +Close issues after release + are now being deployed to the snapshots repo and will be published with Seam was fixed a while ago in CVS and will be correct for the GA release +Its possible to do this in a way that works back to Java by using the String class instead +there are a couple of really trivial things i would change remove returns the result but the return value is never used +Committed to branch and trunk thanks for the review Stack +Going to move to the next release +rebased mdenniss patch and committed w. +This problem is fixed with these two adds +fixed the bit that we can influence and changed string as per jsants email +Closing +Patch to fix port number parsing orgapachemavenscmprovidersvnrepositoryapachemavenscmprovidersvnrepository +HTD and HCD dont need to be Writables anymore now they are to be serialized instead as JSON into znode +It seems that a big part of the plexus code is actually coming from the Apache Avalon project and only got copied overSome classes like are almost copies of ASF originated sourcesTo reduce the amount of work it is suggested to take the ASF original code copy it over to our new sandbox project and apply the maven codestyleThe code needs proper testing of course to ensure that there has no incompatibility been introduced +Allowing a plugin to bind to a phase is certainly helpful but I think some way to allow a plugin to bind to a normal phase and participate in the is needed +Also reproduced with +Attaching referred to in the error message +The current patch doesnt break up the sending and receiving any so the progress will jump randomly +Since this is against an old version of ISPN Im also comfortable with shelving this until MS and ISPN leaving the issue open and then the test to see if the issue still exists +lameattached Bounds and as they should look +But thanks again for taking the time +Added a few more that had slipped through the netAdded a note to the jdbc plugin pages emphasizing the different groupIdDo the XSD pages need anything? I assume not but I dont know anything about them +Deploy as usually +Let me take a look and see how much work it is and if it can be done real quick in Bigtop time frame + +Here is a consolidated patch to show the diff across all the projects looks +Disadvantage is that we need to define document and maintain a custom xml format +Christian could I ask you to verify this error still exists in a recent release? I also note that the generic geotools datastore wizard should allow you to connect to DB +I plan on doing for but jmx will be broken in trunk until I get that done unless that specific patch is reverted +Anyone have an intuitionmeasure of the typical and container wasted while hoarding containers before the useful part of the computation starts? Thanks +Thanks Michael. +Looks good to me +Thanks for making the link +Here I only apply in tuple and now we always use to write the size we use to read +other system tests pass +For those who have been running hadoop for years Namenode being able to write back updated fsimage back saved us during upgrades +Due to the max wait time out elapse when we drop the split attempt we just shutdown the thread pool which do the tasks +What do you think about thisHave you made any comparison with the existing in regards to speedCheers Mikkel +Checking for ticket for Cordova might affect others! It affects and at the very least +Sounds reasonable to me +Im working on fixing this in +I would suggest holding off until are completed +Add REEST for enabledisable connection and job. +I am looking at the patch but I am curious dont we need to add the converted test to some +Issue may already be resolved may be invalid or may never be fixed +Thanks Tim +Attaching a patch which fixes thatAs for failures on Mac env I see that all MR jobs failed there +extend Pythons to use socket client server when uri is avro +Thanks for reviewing my posts + +This bug has been marked as a duplicate of +Patch with basic implementation of the feature +It also fixes a bug regarding absolute paths +Patch for the patch after removing the changes for mapreduce +Perhaps you can get back to me when this moves into focus +Sure +Moved to target +Groovy doesnt make it easy to support multiple versionsGus Is there something specific you need from or are you just wanting to stay current +I think that point looks like more generic +adds the two tests to +patch looks goodOnce you have fixed the typo not not in messages this can be committed +Of course we could make that an option but my gut reaction is weed be adding a lot of code for not much added value +after installation with one ant script run all the examples should be deployed to the to the db without going over the http was thinking of leaving out the http upload in the need it for the config updates +Currently planned tests have been committed to trunk and branch so I am resolving this issue +Benjamin you were right +If WTP ever fixes this requirement then this runtime will disappear +Updated patch to reflect recent changes to trunk +to carry that logic forwardsomeone may not use Shark component but it goes with the framework +Works like a charm thanks +Everything runs fine +Fixed as a result of the fixes for. +Im also attaching a patch that shows the current workI still have some seriously weird layout issues Im trying to track down +Add some missing classes +After fixing the path issue the example would still be broken if it is one of the following dvdstore todoIn that case we should replace it with another working example or remove the example +The latter does not store velues in memory +But I need to double check the spec. +Thanks Thejas. +Committed +Fabian While implementing this Component you should move the Specification over to the Stanbol webpage +Everything else would be more complicated than the functionality added +Although opening again the old discussion Im not sure its the right reading of the spec +First this bug must be was committed +Greppingfor formBean returns only one result the one Im looking for. +Patch is applied at +Im not sure I got all the relevant cdi annotations and it doesnt handle but does show how to use xbean effectively +Perhaps within couplefew weeks after we stabilize and finish up some hanging workI think we are pretty close to this! There are only a few more nocommits to work down +What Slava suggested above sounds correct to me +The three strikes doesnt work given the number of times this is raised by customersHaving an alternate implementation which can react to the topology changes is preferableKev +I just reverted the patch from trunk and branch. +Navis the code changes look goodCan you add the test that I provided also +I had to make a change in the Build Settings +Cant reproduce it here +So yes it doesnt make collisions any less likely in theory but it might in practice +Well take your fixes for the rest +However some changes may be necessary to get the unit tests running especially which does not alter the default value for edit log toleration and will not work if there are prompts during the startup process +Lets get this in first since it fixes the bugThe JIRA is +Attachment has been added with description Testcase showing the problem +One of the things we learned was that documentation doesnt fix an unclear product +Added dangling node contribution fixed testcase however still it is not converging as fast as beforeI will investigate tomorrow a bit moreBut you can still review +I have line patch that adds google collections in all the right places and includes the changes in the patch you submittedI will apply this if no one objects +META +This is a problem in trunk and +If this did turn off nio and it was the only one inyour path then the issue is in some other part of code than I have be guessing at +just copy the into my project somewhereAny help on this would be great as this stopping me from launching my app +I uploaded the replacement files that fix this problem +Ive updated the Bindings property page to take better use of the space with hooks for adding AddRemove buttons if we see the need +Figured out the issue +Patch for to trunk and Thanks for the review Lars Elliot and Chunhui. +fo and +If you set keep to true the result will be keptOr did I miss something here +That means the normal is the model reflects whats on the source +Joerg its chicken and the egg +Is that not OK? It is easy to set up rsync to use a password +What I meant was that these jars are not part of at all and that eventually they could be removed from the package +Believe the code is correct now +Ran ant test on this patch and no tests that were passing in trunk failed on this patch either +since is still +Probably just missed +Now the plugin works fine even with Maven +I wonder what Hudson tested I accidentally attached an old version of the unit test which doesnt even compile with the new Top +According to Aidan this issue should have been fixed with M so hopefully we should revert this change once the merge is done +Suggest removal of +Move blocking form to +its happening inconsistently +Can you try generating a new patch with the svn client? Thanks +I have tried to reproduce this but cant seem to trigger it +Thanks +Closing resolved issues for already released versions +If we find a good candidate feel free to reopen this ticket +I may try to do a release at some point +First of all pages are not and they dont close the scratchfileYour change simply drops the problematic code and therefore doesnt qualify as workaroundKurts comment and the output messages lead to the following conclusions close is called more than once as the closed value is still false something went wrong when closing the the first time if the behaviour depends on the environment than most likely isnt the problemSo the question is do you swallow some exceptions when closing the instance? +Do we still need this now that we have the master rewrite and HBCK tool? Maybe punt to +I got this working now +Intuitively here is what I think we want the config to mean Keep the least amount of data possible to guarantee a buffer for the consumer of at least N bytes +Ok Ahmed I think I understand your proposalI am for but on I think developers in this scenario can simple copy these page templates into their web projectregards Malcolm Ahmeddo you have something of this already implemented to share with usthank youRicardo +I did lots of testing which is why it uses the Reflection class first if it can then the +Checked in! +I guess we can reopen this and try to reproduce again +Someone has to also add tests to the junit test harness before this one can be committed +Samuel have you been using Chrome when this exception occurred +So it might be best to put this code somewhere like instead +Thanks for pointing it out Daryn +Milinda patch applied! Many thanks +So we must include this patch in it +Since r is reported before die in defer minor issues to + contains a and a for updates to the source and test cases respectively +Actually does the current implementation of the LIMIT clause support sqlparameters at all +i am gonna reopen it to see if theres a way to set the TCL and keep servicemodule isolation +for the patch +Tested on Samsung Galagy Tab inchesUpdate +While the logger in the resolver is public and therefore global its not accessible outside the bundleSo in the end its a matter of taste I guess For the additional log message this is already logged in sure thats an advantage if youre trying to determine which class is doing something but if it works then fineRegarding the message in the problem I see is that its at the debug level +Bulk closing stale resolved issues +Closing the issue. +DoneThe new build system uses a common root build script in the projectAll other projects import this script and inherits tasks defined in that oneA lot of common tasks have been added in the root build script compile jar plugin feature javadoc clean +Any chance this can be committed soon? Thanks +Closing old resolved issues +attached new eddie for commit placement in doc this is a great start +Current lots of camel tests are relay on if the is return false the test method should start the camel context itself so I just add a comments on the is method to tell the user to start the in their test method manually. +A sample program showing how itables would work for X +closed reusing old upgrade +Its public +I have reproduced this +Sorry I missed the from Arvind +In trunk of piggybank it just redirects to builtins where the problem is fixedWe can backport to which is what I think Raghu is suggesting in this ticket +It is visible in each perspective like the New +Would you please double check +Dont know for sure but I guess we should not update of jobs that are already marked succeededfailedkilled +Junit diff of changes to exercise the fix +And until I am done with it Nexus staging is broken +NO incubating release please +Bear in mind that we hope to start replacing all of the m plugins with functionality equivalent wrappers around their m counterparts real soon now so there is a risk this will disappear again unless you are prepared to make the change there tooinstead +added global gitconfig for jenkins +Thanks Chris +Nevermind just noticed that x is fairly new thats why its not part of jruby +Also would you mind to point us to the actual changes that you made and believe that it will fix this particular issueThanks an issue when determine if the entity reference is nullable +Main contributor is the Segment cache +Verified fix is okay. +Yes I was intending to create the tld first to check that the mailer is working OK and then either delete and move all or move the subdirs +Im not sure if both are in use but previously replication attributes have been added to both files +Patch to remove the check for tables +Cant +There is no visible error I will need to dig deeperAssigning to CR since its affecting IE and thus Showcase on IE in compatibility mode +I am writing several plugins that provide CMS functionality +Its only at agentresource layer it health monotor policy translates to specific configuration as per the device +PR applied in in JBDS +Is there any update on this issue? Because it really make Maven unuseable with JDK +Yup the proposed improvements by to remove duplication of encoding logic among the prepared vs path was incomplete +it could also be related to memory feature we recently added inTry to use a smaller service offering or add new host to the cluster +Merged the trunk change to the branch as revision +Better to do that at the time of course or when we update each +Slava +So is true then to say that for system table queries or using system functions or system table functions we recommend switching to the SYS schema? I ask because in I went through and changed all the system queries to use casts to avoid conversion errors +Yes agreed this would be nice and easy to implement +Of course anyone can apply a patch but someone with relevant knowledge should verify it first +Why is it called +What Thomas did could be replaced withmvn clean install pl containercontainers amThe same goes for the ppl that are interested in only one containermvn clean install pl containersjunit that Im against a container free build just saying +Updated section to include reference to JDK in Release content is already updated +Here is a preliminary patch with proposed modifications for the review +maybe you should move the export package to a different jar so people not interested in it do not need to include it and do have some dependecies less during the build +It seemed to fix it and so I let it stay there +when using a Palette you also need to have a +With this one only dependencies which are not present in the obr file will be installedve got the same problem +I would assume there are different locks for RAM and disk cache? Also it is my understanding that if Traffic Server cant get a lock it will retry a configurable number of times and go to the origin server +I believe you are right about GEP calling Deployer +now the change never makes it to the cache +Attachment has been added with description Jason +Broke trunk and branch +putting LATER upto koen to tell problem here Fedora Eclipse Version Build id Tools the preferred way of attaching the actions is by using the properties view +Patch file to resolve I dont know the code that well I dont think the patch is correct +As of Tomahawk the dummy form stuff no longer exists +I had a quick look at and I guess the problem could be the parameter true Im not sure if the Tdecode algo takes this into account. +fixes the JMS Server pages +Note added to section page in the release notesThe confusion here was over the tomcat install approach being recommended and the +restart itself should be removed from the product in my opinion with no replacementThe other ticket about the restart problem is the one to fixrestart is useful during the execution of the test suite though. +Im not actively working on it +Also FYI setting as a List also works for Again actually you may be right there was an issue resolved in CXF to do with container filters not able to set or Accept +tested works +I thought that you were disagreeing with the changes to the more specific in the implementing classes also +Regards +Need a new version of jbossweb to fix AS +Marked for +For instance you could protect the users database so that +Also if this is completely custom would you expect this in Teiid or would it make more sense in a message bus layer +Please close the review board entry when you get a chance +I can see the same thing before applying the patch running standalone passed but failed in batch run so its not caused by this patch will apply soon +The patch on might be a more complete fix but I also dont know if there is just one instance of and it might cause some other issues +which patch do you want? one or both +scm info version numbers and legal headers still need to be updated to be included in wink +replaces tabs with spaces in newly added codeCommitted to trunk with revision +Was this the intent of the design? seems like there needs to be additional for these filters to participate in the command line +The word I heard is that the geometry columns dont use the innodb stuff that actually does acid and all that +This all sounds good guys +never mind +nice catch ted +Having this system level logic controlled by this utility plugin may not be a good designWe can instead monitor bundle activation by listener and stop the system when activator throws exception +Issue is fixed +Hope these reports help improving please let me know if I should perform any further testing +I think this is a good way to solve this +from trunk is with gem nowjruby S gem +Thanks SANNI Lookman +The previous patch fixes the warnings but not the problem +As a work around I used to add the necessary header in my manifest +verified build +Thanks! Thats the way I love it. +Minor update fixes +Committed revision . +Is is possible to get a more recent SNAPSHOT deployed +If thats the issue weve got a real problem because were going to have to find an appropriate fix for and build a patched version ourselvesLet me do some more research to confirm this picture +This is a bug indeed +This wrapper is meant to provide a simple wrapper for the data structure +Correcting the assignment +This patch updated the minor subjectverb issue that Kim noticed +Improved XML file support +Resurrecting this threadI wish we had in Android since Id use that to communicate with the Callback Server instead of the XHR +Looking into it +Added the product component +This problem was introduced by Dhruba in +Also the test in the deployer is invalid as it refers to the mep rather that the service info +Ive made some progress preparing for release but there are still a few things I need to finish first +ok Ill do it +for the patch +It would be good to print the block information as well in this patch +Actually it wasnt resolved +While copying files with a few large ones can result in long tails +Should the bundled mochiweb be updated to remove all the deprecation warnings? +Considering that the issue was just opened yesterday +The commits may need to be reverted but well work out that process on the dev list +Like this Jack? If the functionality is okay that can be applied to other fields as well Only have to know which ones +I think this needs to be done in a separate branch so that changes do not effect release +it is totally unclear what this method should do to a person writing her own adapter +There was an issue with the soap content model happening when due to the log level some message logs were not procuded preventing the model from transiting to the right representation before the signature is computed +Frank this will require input from you as to what to do next +Thanks I think that will prevent some people from pulling out their hairs +you need to update the jars manually to get it work with Railo looks like it is resolved? Awesome! How do I manually update +Closing this issueWork on has been postponed. +Oops +Yes we want cqlsh or a similar tool to be a full replacement for the existing cli +Its an abstract class and Lucene is known for breaking the hell out of everything between releases +Mounting is especially useful for some Windows NFS clients which cant mount root export possibly due to its own bug +In your case you are invoking an operation and therefore there is no way to get the service group id back to the clientI do not think that something we can fix to have SOAP session support first you need to invoke operation and then invoke whatever MEP you want +Picking up only one task may or may not bring down the usage to within the configured limits +Fixed +I will make sure that there are tests to verify that the pages load correctly +This does appear to be the same problem in sent you the mapping files +It is not a requirement for EAP JMS clients to work with +Asigning to Thejas as he will be working on porting the typechecker to the new logical to address the case when column is of type bytearray +But if you kill the browser and come back the users accounts are still linked in the DB but the cookie no longer exists +I dont think any ordering precedence has been defined for these three parameters so surely the above example is an ambiguous handler mapping for the requestAlternatively can the ordering precedence be explicitly defined for these three properties +Hi KalleWhere do we stand on this issue? Are you still awaiting the correct grouppermissions from infra to be able to create the target directory +I agree with Todd that the default value of is somewhat in the order of to minutes +Now the question is why the hack do we need that? Just use an Singleton EJB if you like to share state between different WAR files MDB and stuffThe problem with this is that this only works for EE applications +This patch applies to both andThis is a regression I introduced in fix for. +Has this issue been resolved in Final +This has fortunately already been fixed for released in early January! greatSorry for the dupe I actually missed the release which I will be getting right nowThank youLars +Works for me with snapshot and Spring +But it does not always happen for me +This fix changes the following behavior after someone declines a shared page +This and could well be related cos they both link to the same problem +It is completely expected an for all of devstudio is bigger than the m a bit confused by all these comments +Help in resolution is appreciated +Good catch I just noticed the type changed recently and updated the comment naively +Sorry for the all the confusion +I though I tested all combination apparently not dirk +coresrctestjavatestiioppluginruntest +Sure but that should be done before the rat plugin was added that breaks the release build +IMO the webapps created with Hamlet is much more maintainable than any existing web framework because of JavaIDErefactor friendlinessbq +Enjoy +My own testcase is working again +Are there any unit test cases for the scenario that you are mentioning about +Sound good to me +Please attach the patch to the jira too +In countries where all network traffic is monitored for example connecting to an HTTPS proxy outside of the use of HTTP CONNECT is very useful +Oops didnt rebase correctly against trunk +Treat unscheduled issues as please have a look if this is a valid change +Thats why this patch uses the CSLM constructor which takes a +Upgraded rails to and it was causing the problem + +Im guessing this issue can be closed with having been resolved and the docs amended. +Commit changes to trunk At revision At revision At revision +Well the typo is definitely wrong but the package name used to be correct at one time but then we moved the OBR interfaces and apparently forgot to update it here +They do not fail any more +This issue is resolved +Given its size can I get you to comment that its intended to be licensed to the ASF +Updated title as Voytek the dangling link +I dont personnaly see why we should wrap these exceptions +DaveIt seems like this issue should be resolved in the branch as well? The primary use case for the Job level was to use a file created in one step as the input for another step +Jenkins is passing again. +Create another issue if you like to address the JNDI integration aspects. +Try and get this in to +Set JAVA LIBRARY PATH for native compression codec +Sorry again +NEW in dictionary +Committed this to trunk +Nicolas de Loof went and applied this patch before but at that time it was a bit premature +The block points to a directory that doesnt exist in my implementation +How many flushes were there at any one time? Was compaction keeping up +In my previous comment I am referring to the NPE mentioned in the description of this issue. +Ive attached a patch here +Turned off Patch Available flag +Changing priority to minor because there is a simple workaround. +Can you test it out +the returnType isnt VOID +Ill take a look at the logs +I think we should keep the old setting as it was changing it was a feeble attempt to improve performance but there seems to be potentially dangerous side effects +packages +Hi Amine which Sonar version are you using? which DB are you using? +We do not make projects that do magic and include only certain files without providing the user a way to do it +For I think the design should be a prereq to launching the registry but the implementation only needs to be done enough to be putting the correct fields into the couchdb +Adds a new utility method to to take a resource name and execute it using runScriptSome things to note This test is not enabled to run currently as part of any other suite +See also +thanks Brock +Besides that +It was fixed in Alpha +I just tested it with and I have the same problem +must not have closed it out. +Please provide a test case preferably in the form of a Maven project +Im not sure what Jena version is used on this project +If you have used Clover and have defined an exclusion filter this bug would duplicate +In general it might be helpful to issue an INFO level log msg whenever a configuration attribute is overridden +Hi KristianThe doesnt rely on those particular defaults +OK tests pass I will commit this +I think it would need a bit of coding to pull out some test data run the job compare the resultsI have only one little tweak to make to the procedure you mention here +Igor can you verify this JIRA on CR +thanks for the patch +If it cannot then we can write in the documentation that we will make it easier to set the weights in Camel +Create custom +Why does need to accept the rmaddress as a parameter? The whole point of this jira is to not require anyone to supply the rm connection parameters +were removed in Bela +its just a possible candidate when builidng the all javadocsthe best solution seems to be moving that content into contribqueryParsersrcjava +The finalremotingsrcmainorgjbossremotingtransportmocksrcmainorgjbossremotingtransporttestsrcmaintestdetectionsrcmaintestsrcmaintestperformancesrcmainorgjbosswebtomcattcauthenticatorthirdparty somehow is not downloaded in so I copied it manually from the branch to run the utility +John The that maps properties to elements did not include the shorthand properties +Test failures are not related and being addressed in a separate jira +Setting fix version so we can try to get to the bottom of this before our next ve seen this problem before +As an intermediate step one possible way around is to change to relay the from the java Child to the C reducerIn terms of development you could start developing on a svn branch of hadoop pipesThoughts +is about the same issue +Actually fixed a while ago by changes to the to duplicate the finally block for normal control flow paths +Closing as Later +This could be partially addressed with more documentation I did that in my own code but that didnt stop another bug +I have already starting working on a patch still needs a little testing +Moving to as its not critical to the December release +Yes agreed I think we can definitely change this but the should be considered an internal class now +I locally upgraded to our new snapshot and it works fine with mvn so far +well always have a volunteer with hbaseaccumulo knowledge to keep the code up to dateI will gladly do that for the foreseeable future should this patch get accepted into Giraph +I fixed the problem the test now passes sucessfuly +It will open an eclipse instance +Hence it should be ok to reopen this JIRA +is redeployed after minutes in defaultso in step the retry time at least minutesotherwise server will abortHowerver in because of no retry server must abort itselfThe PONR is used to prevent data loss when failed updating meta +Increase the priority so that it shows up in the list +Thanks WilliamTim can you confirm that this fix solves your problem +bq +This has been committed and I updated to reflect that. +Filed for the common parts +But like you said would require updating the s send method as it currently only checks executes these hooks for view requestsI too have not used either of these hooks +Hi JerrySorry for not replying earlier but yes this would introduce a bug into Ivy +Not a bug in particular not a Groovy Eclipse bug +If you are putting components in here with this feature then that is bad +The proposal here conflicts with the proposal for maintaining recipes under contrib +Youre absolutely correct and that will make a great start to a doc explaining it +Works for meThank you Alexey. + +Every issue is work in progress +I get the same error as Matthew +Hope the issues are fixed in this patch +fixes some target dependencies in the ant scriptThe original patch fixes the bug in but introduces unnecessarydependencies on the compile targets +and +This would help me see If what Ive done has any impact +I reimplemented this in a simpler way +Thanks Frank! +Werner you missed some mojo parameters that should be changed from String to File too schemaDirectory dest tstamp +I also think the way Damitha suggests is more suitable and makes users lives more easier +It is pretty clear that previous failure is unrelated already filed to track this +bq +Thanks! +The soup took longer than the bug fix +I think this is the good way to exclude your generated classes from quality analysis +Does it still happen after upgrade to SBSThanks +Ashish for commitRegardsChirag Manocha +Fixed a findbugs warning +pom needs to be updated dependencies are bundle has just been should be stated in the documentation for uploads I wont upload this tojxta wrong groupIds now we use fully I dont see any proof that they control the whole the same as if I want to upload something to just because I have a project hosted thereI havent looked at the sources and I won after some brainstorming on the dev mailing list the following groupId has been chosen for this artifact bundle is up to classifier by hand +I think we can close this now +This issue has a critical impact on Seam as any Seam project fails to deploy when bundled in an EAR +This is what I currently looking at +Defer to +It works in This seems to be a serious regression with castor +Patch with fix +Please if you continue to find this issue in versions . +But you cant see WTP markers in problem view +bq +sorry i should have mentioned this earlier +Closing. +In fact not only are these parameters parsed by Sling also the request line parameters are parsed by SlingProbably it would be better if Sling would only parse the multipart requests and take other parameters from the servlet containerFYI Sling parses request line and parameters to implement the parameter encoding scheme using the parameter +to work as it just complicates buildtest cycles to SLFJ +I find the reason of exceptions + +Maybe inYeah still tests raw mlock. +I have committed your translation to the Roo trunk and it will be made available in the next release Roo I have adjusted the translation for themes slightly to align it more with the other languagesstandardCheers thanks for the offer this will be very helpful given that our Italian is somewhat limited +Converted CSV to ODS for readability +Oh I forgot to mention that it is possible to selectcopy in the Entry Editor when browsing the Root DSE. +Committed to trunk +Same issue reportet as and due to an error in JIRA as again +Thanks Jesse! +on patch +This ticket is becoming long and confusing. + to trunk and +I created a for this one for collecting the resource inI will factor out the codes for resource collecting there and leave the scheduling related codes here +All initial code setup tasks have been issues associated with existing releases +That test fails the same way +Currently hadoop contains is that can useShould be covered +groovy works with since of +Sounds good to me as well +Already made some progress so I think its worth targeting after all +Fixes temporarily committed to annotationspatches and corepatches +Attaching and with the added informationI also took advantage of the opportunity to fix the indentation of the code examplePlease let me know if changes are needed +This appears to have been pushed to already +Can you try against the release Ive run the apps locally using that release and cant seem to reproduce the problem +Committed revision +clss is also cached for at r. +attached is one using the DSL and one using Spring XML Extensions +Use the next best thing +Currently note that we are not syncing clients polling interval with MINTIMEBEFORERETIRE that you added so if later clients polling interval becomes configurableincreases beyond MINTIMEBEFORERETIRE this problem surfaces again +Gre aus Wiesbaden +Im on Ill commit tomorrow. +Did you import all Maven modules as single Eclipse project? Then I can reproduce actually happen because entire project is compiled with jsr mode so junit cant read test annotations +solution found +Earlier Maven dependency upgraded to is looking goodIIUC this will be a parallel build system to ant +But will leave it up to the maven experts to tell +the access to the methods is actually not necessary outside of as first thought +I am not sure we should disable this due to concern of how expensive sampling is +Leonid no solid ETA but not longYou can employ the workaround above in the meantime + +Interesting though the proposals there look rather clumsy and do not appear to implement the fail on warnings aspect + +I have just started ofbiz and have not done anythingI am guessing it has something to do with thealso note this only has and seed dataI have removed the website in the I have put in a Webstie and Productstorethis will cause this error +closing it +after that I may go ahead and try this version anyway even if there are still failures just to see what it does to the index size + +Tested in node mini cluster test passed +It knows of all the configured stores the replication policy and the serialization used for keys and values +But pretty busy today +Applied Thanks for the patch! +Or even a Rewindable interface that can implemented if a supports RewindThen if you really needed it you could use to check for support +If so this is not code will track the snapshot on primary storage if there is no snapshot created on primary storage then the next snapshot will be full snapshot +dir created in r +on patch version +preliminary patch to remedy the patch is useless it has too many white space changes +BTW this jira should have been a and not jira +I wish I could help out but its not possible in the immediate future but Im also very interested in this +Snjezana Is there a previous release of Galileo that doesnt exhibit this problem? Customer is currently on SR +Nit You should probably revert the changes made to recipes before committing +Im curious why the changes to GPFS are needed +I do not understand what you mean by files are not in the correct pathCan you start the WAR on +Need to fix this +The tutorial is finished +However this listener was never removed from the list when the reply frame arrived +Tested with both and from the repo +Im not sure on this because one could be writing in English and Spanish and hence have the enable blogging enabled but later switch to English only and disable the enable blogging setting +Looks good to me +Applied the patch with revision dcfccbfbacaabcfdcddfaaThanks! +Hi Keith I did some work over the weekend trying to see if running the migration again would result in the same problem and it seems that it does +Ive done this work in a separate library. +binstopping hbase +I did not exactly understood what was vetoedI think a nicer way to handle this is thisLets upgrade all four branches at once +I dont believe the changes made regarding filterApplied have any bearing on this issue +on patch +Same behavior as Nathan and others reported query gets truncated at the union keyword +This has been fixed as of revision +That would complete the sync between and +The changes look good to me +Im not sure that wed gain much by adding an Order t you just add a qualifier in the second Given step +They want to cooperate with development team +Its working fine indeedIt solves suffix context mapping problem I have encoutered by pastReally great stuffIm not HTTPServletJSP expert but its looking to be promoted +yup +Just was waiting for unfreezing HEAD +We should iterate through all irrespective of status thus enabling and other tokens to be explicitly cleared and invalidated + +I think the rational is that you never talked to a HTTP server so there never was a status code +It was basically a renaming of your with a few trivial cosmetic changesGlen +Fixed in revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It has been working fine for me but I have not been using it in production or under heavy load +Applied but without the requires test in because it was removed in +What do you say Andrew? May be we can handle that in another JIRA as that will be of wider scopeNo this instance should be dealt with in the context of this review +no more kaboom so i would be fine with closing this one +The Servlet API is added as a dependency for +And you are sure you are using exactly? In this case I guess we cant helpyou because we see the same like you do and solving it without any hint is a bitdifficult +After some discussion with Guillaume on irc I am going to look into an approach where we can import an optional package in the gshell core that includes a resource with the embedding applications details +Sorry +Applied on master +For exampleP Index srcmainjavaorgapachesqoopjsonWould you mind regenerating the patch with svn diff so that I can apply it cleanly? JarcecA new patch is submitted via review board +Tried both Pig and I didnt see error +made sure that we now have just one +However the issue mentioned by Pjotr Andreev is still occurring +! Very exciting to finally have stored fields under codec control +Am I missing something +patch for in the branch and the trunk. +like we do for +Id recommend that you give that a try once its built +Does using the Grab work? +I hope the default is increased too +Ill also do some tests with +Can you provide a sample project that I can use to reproduce this error? I cant reproduce it +Im assuming I need to udate usrlocallibnodemodulescordova +Hi AkiIMHO that would go against the policy spec +For fx okBut for tx the page will not be rendered completely +This is nothing to do with identity management roles can come from any authentication process. +If you would rather not have a bug there you can attach your version of and I can modify it and resubmit +Bad credentials is error of when the user err the sign and not of when the configuration is with acegi has many configurations +BTW why did you decided to backport it for but not for +forward. +Im closing this bug +Can you check that your app runs with mvn gwtrun? Try with a snapshot version as well +Now I see that it surfaces more often if I increase the number of partitions andor the number of messages +Could this happen because the service is deployed in application scope +Changed the to +ok let me know how I can help with refactoring to align the other tools to produce this as output +Help in resolution is appreciated +Ill try to track down which Rack versions expect a nil vs which Rack versions expect an empty string when the client doesnt specify one +I have tried with all these versions it is giving same problem +I also ran the tests locally and they passed just fine +Good to knowRegenerating patch from top level so Hudson can check it +So I need to fork github repo first and work there? +The tag needs to be deleted and +Withdawn the code change in multivalidate +please check with a nightly build +Ok so this is only to set the bean name not the actual reference +A bit but it could be specified that way +Fixed on trunk revision . +Sorry sholdnt have closed this moving it to because its not complete +Renaming interfaces once more +Has any work already in progress for creating kafka adaptersI would be interested in collaborating and contributing as well +Ive merged your changes to the mapred branchThanksDoug +How do you measure only processing time? Yes I modified code and add logging lines printing processing times +I applied the patch. +Maybe it is in with +Did you verify that the buffer increased using strace +If it had less than two the code would fail at runtime +Do you mean that include apache commons http jar +Added a null check to I ran into this while writing a new test +I have been unable to set up an environment to build and test source changesAdam +The browser autofills some fields in the form in JDK the database name is autofilled with system while in JDK the timeout field is filled with systemI guess there maybe two issues one is the display order another is how to ask the browser not to the fields in this formAny idea +Basically the query used to retrieve objects with a null path in results in Oracle returning the root for dashboard which results in the query for dashboard never returning null thus dashboard is never created but aliases to root. +Thanks to you for sharing your problem and the small sample which reproduces the bug! Creating a good to the point sample is almost always harder than fixing the underlying issue +Hi AlexeyProbably not + +The Not a hostport pair is probably because of move away from HSA toward +Roboert ok Ill try to find some time to do it this week +and modified to use new api +Closing as not a bug sounds more like a in the geonetwork own closing all resolved issue that have not gotten any more feedback or comment in the last month +Didnt make it into EAP or +If you find further issues please create a new jira and link it to this one. +this is working fine nowGreat! How about putting a patch up which can be checked in then the follow on work can go into separate +Let me update the patch +Content updated in Getting Started Guide Release available on in GA +removes the code that created the dependency and to the best of my knowledge isnt being used for anythingOf the two solutions I prefer this one be applied to the branch +Ill ask Zhijie who wrote it up to have a look at itCommitting this +Ive add a little test to do so +Thanks! +HiJust checking to see if there is anything I can do to help you certifiy your project information pages on Labs +Thanks Enis! +Resolving to be retested +Ive looked at the Java Ant task doc and Im not sure its possible +PR is pushed to with JBDSiPhone skin is changed to first skin in available skins list after first item is removedSecond window skin is not s not expected behavior +worked for michi and i cant reproduce the failurecommitted fix at revision close for the release +Please try using the from ticket This bug has been marked as a duplicate of +Or maybe its ok to just pick the up off the classpath as is done now +This happens when running jruby scriptserverin the rails dir +This will give lots of unix tools grief +This will be trunk only unless there is interested to commit on +Module deprecated +Guess this needs to be raised as an ESB issue +Ill see if I can have a look this evening I know some of you are in rough shape +This patch switches things to logjIm not wild about adding logj to the +Interesting case +This time Im really not sure we can refactor to avoid it. +The message now saysNo packages have been defined for the project xxxAs far as putting the selected project in the tree +Bulk update fixversion to +Hi Michael I cant manage to reproduce this behavior +Renamed issue to indicate what is being fixed makes for better to match the changes in geoapi + +Also can you point me at a REST reference which states that the key for the object being operated upon must appear in the path +I was close to name it I stated above i see this as a neat workaround until we fix this class contrib eventually +Actually it does look like the export in is necessary this is for the case where people dont have in their conf dir +Richard appears to be set to resolve in the next versions of the +patch fixing javadoc warning +Isnt it better to store the compiled jar file? Eventually this file will be distributed to each node so we can save even more time if we keep it in the node to run several times +As of revision Ive committed support for z archives to SVN trunk +It seems to me that they refer to the same thing +I will look into fixing that +I figured that there was a good reason that SDOFILTER wasnt being used but it unfortunately makes unusable for us +just got to that conclusion too +We need a hudson job that tests suite created in trunk +reopen for no code sets this attribute property so the value is null +I would appreciate If someone with commit rights could apply the patchGarvin +suggested patch demo have checked the patch and it is fine so Ill commit itThanks to Christian Kaltepoth for provide this patch +i like it just pushed this +I found out that the annotation is directly available from the package by default +We should generate a unique internal column name and set the alias as alias in +About this functionallity would never have anything to do with any SCM its just what users would use to have a shared shared location could just as well be inside a project +Ive updated the Uberspect code with the latest from Velocity which brings varargs has been brought across from JEXL stream and a test added +Yes this is very true +I did not disable servlet version for jsf and struts I added servlet version support to in +From the documentation it does seem that the store clears out a log file only if it finds that there is no reference to a single message in the log file +Patch that Mark Miller proposed if it turns out that his analysis is correct +Can you change it to true +if tests passGood work Lars +Thanks Alejandro for the review. +actually it some HTML body coming back as part of the standard SCUNAUTHORIZED t webhdfs verify the responses content type is json before trying to parse as such? Has one bug exposed another +Updated patch +Ok + +If were not willing to pay its probably still important but not critical + to trunkThanks for the review Anoop. +Hi Cody thank you very much for reporting this +Also fixes a couple minor javadoc errors in and Command +The problem with ORDER BY columns can be addressed whenif we allow ORDER BY in CREATE VIEW +But looking at this document Id rather extract the information available as part of this doument and make it available as part of the HTML docs +Thanks Suresh for the reviewI have committed this +I am attaching a patch to correct the character checking issues seen in testcase +At the moment the response of the has as textplain +The pdf plugin was released recently +This isnt exactly what I had in mind I prefer to not use instanceof to determine if we are running in myFaces I prefer to determine the marker that should be used by actually finding the marker in the response +can someone close? I am on mobile browser and page renders terribly! +I agree marking it as a fix for the Hadoop version +While undefined this property defaults to destDir +IT added +Can you commit if its successful +Commit it soon +After we get primary requirements for completed we can look at optimizations in this are +There are a number of errors introduced with the previous +Not sure Ill get to it today +oops needed to rebase +Hi IvanI did not realize that my patch triggered the findbugs warnings since it did not touch that part of the code +Thanks! +This seems a lot cleaner than customizing the current code to meet needs +This should an issue with any ODBC client not just perl right +Lets get an example working from the code out then well figure out how to make it easy to run +patch to address Henrys comments plus the reporting directory has also been moved to a new committers directory in the SVN area +Lots of questions and nocommit in the code +In othe words the server can open from the client IP but can not receive the message from the client +Andrew in now my favorite saintbut St +This contains all the source files affected +First step is to get some solid performance testing. +Thank you Zheng +The need is not that highly common although it exists for certain problems +Copy the code of and name it and use that +Thomas can you please have a look at this? COMMITSHUTDOWN to some tasks +This patch ensures the method names are pulled in from the config and not specified inline +I can rephrase it +This brings s path in line with +The idea was that I felt this was a prerequisite for a release. +I think Seams component model does a good job of facilitating this type customization where necessary approach +Use plain JDBC transactions to perform the schema migrations +The scorer caching is not enabled +properties +Can anyone with build conf access post what the exact command is being executed by jenkins? It seems that another occurrence of this issue in auth is masking the error in mapreduce +perhaps a new jira to update the makefile to include valgrindhelgrind support? it can load the exceptions that we know about explicitly +Its a huge pain for our users +to and trunk. +This patch changes all task to todo +Not close enough to satisfy Solrs pace of releases I think +The bottom line for me is that Im open to any one of the solutions or even both +Not a bad idea Werner +I cant reproduce it in Linux +If time permits Im planning to do a backport of some issues for a Struts maintainance release +For example your first plugin could turn off the insert method completely and then the second plugin could query the rules object to see that insert is disabled +Vasily Whenever a resource created by client code is passed to Xerces by adopted true it implies that the client has agreed to relinquish the particular resources ownership to Xerces which will delete the resource when it is done with it +I think the test should be removed its broken by design mathematically this will always have sporatic fails +Maybe this log could be down outside the lockI think running integration tests a good idea +Thanks it works fine with that property set +Unfortunately this variable effects the job planning for most of the unit tests as turning it on and off adds more stages +Basically the method body cant introduce any inclusion dependencies +Thank you for your followup +The compaction is not only requested when its necessary +commit that sucka! +We plan on picking up the same version of jBPM as in SOA +Its that cluster setupteardown that causes a lot of the tests to be slow if teardown could be asynchronous maybe even with queued creation of things could be faster +A good way to fix this without compatibility issues has not presented itself. +After deleting the directory for this library I still get this error + This bug has been marked as a duplicate of +You should try to get test data instead +As per comment a novel text only version uses Stack as more efficient version can now process any number of documents and remove boilerplate text tokens such as copyright notices etcNew version automatically maintains only a sliding window of content in which it searches for duplicate paragraphs enabling it to process unlimited numbers of documents +Hi Michael I had a closer look at this issue and while writing a unit test for demonstrating this issue I noticed that the davex is actually overriding the move operation thus it shouldnt be related to the comment you mentionned aboveI also wrote the test that basically reproduce the scenario I described in the bug description except that the two processes are simulated as threads and +The original patch applied to but not so I rebased on trunk +Okay +I reverted from and trunk +I have pasted the log above in the comments trail which enables the recovery to happen within seconds by bypassing the stale node +You can use or +Ulrich is right that these changes are needed +The feature have been implemented in MSee +good to me +includes the change in and adds a fix to the test of if the tuscany context exists the context is now null whereas it must have used to be a context with an empty name +yep now it looks correct thanks David +We dont have control over the names of the generated files +More details would be appreciated +Committed to trunk. +bq +if successful then the state will automatically change to started +Thanks for working on this! I didnt feel too well with the singleton either but I simply overlooked the. +At level we have only and +windows supports such use cases fine +see subversion commit tab for locally before committing code +Thanks for all your join the chorus in saying that I do not understand why this issue I reported has been dropped to minor priority supporting entity associations is a core feature of Roo and this prevents the whole application from compiling! Not sure how much more of a showstopper youd need for a higher priority +Yes this is correct +Thanks StefanWe will try to prepare you the sample +Should it just be in? +Uploaded patch +looks great stephen +How do you propose to get the file name from the urlAlso where is this important? What output connector are you usingIn general I think creating metadata such as the file name is OK even when the repository doesnt have one but Id like to understand how it is meant to be used first +instead of for +Good ideaHow about provide pagination on this page +Existing security handles that case just fine +Thanks Sanjay! +However as for the second test case maybe its more reasonable to throw instead ofBesides I have tested the test case on RI +We are experimenting with a change in kerberos auth handler which will require no change in HDFS +Yes I intend to develop a mini cluster like mini to test most of all components running on bspd +Hi RonThanks for you comments +within the same suite? If its a different suite then this shouldnt be possible +Can some one please look into this? Its a critical issue for us right now +Looks good so far +I dont care if the exception gets renamed but its text Table table does not exist in this schema needs to change +Thanks. +Hi DarekThis is a legitimate request we will look into itA potential solution would be to maintain a branch that removes not GWT compatible classes +OK after much research and testing the latest info is that if I add my own entity manager to the generated empty shell class file that DBRE generates then the Entity regenerates the +any plans on fixing this soon +I just used to to be fixed now weve the new archetypes +Ok +Moreover does not work with cmd +Lets start over and try to fix that download page +Maybe only one is +Attachment has been added with description Test case +Yes I have added redundancy Just to prove my good intentions I have also created a Jira Issue to discuss the problem and find a better solution to fix the problemProcessing Payment using async call is lot friendly then doing a full page submit +Yes I should have documented this in my previous email +You can export the java code from the selenium recorderthanks +I probably wont get to it until the end of the month though +Im just holding a commit until Tim Fox merge a large piece of work he is have attached the patch here for backup and documentation uploading the correct file +To clarify what I meant sayYes we dont use syncronized maps any more but a concurrent which help with all the multi reading and writingBut we do have a sorted structure still in memcache but that should be the only place where we actually need that structure. +Can you try jruby d rfcntl e on the target machine +I will add it +The SSL related files are being used where as the files are not used +Found a small problem with a shutdown to to close the thread pool on exit +patch for trunk is also applicable to branch +This is something Ill look into if I get the time but the basic upgrade is there +This should also solve the problem you describe +ok after chatting with doug here is a different versionIt leaves the class loader in the Configuration but it exposes a get method that allows a class lookup by class name rather than by attribute nameIf the thread does not have a class loader it will use the default which fixes the precise bug that Christian was hittingIt also pulls all of the other places that were using the threads class loader to find classes into the same code based on the config +hi Abhi Need to add this to the overcommit section +Not completely sure if this is indeed the problem but for as far as I can see with the little knowledge on Spring Roo I have this might indeed be the case +Fixed error in the first patch +Wrong issue though a related one. +If you choose NOTANALYZED that should mean the Analyzer is not invokedIf you want to do expert stuff control the offset gaps between values for NOTANALYZED fields then just analyze it instead with keyword tokenizer +I think this means that its possible that one of the threads has performed both the select and the delete before the other thread has done the select and then we may see intermittent test failures because there is no deadlock +Currently it is done only forwsdl +Hi LinI notice that when admin console shows Chinese the section Plugin Groups in local server disappears +Attaching +That explains why more than connect requests got served when I set the backlog length to be +This has been the default tomcat behavior for years so there is no reason why it should have been requested +Created a new issue to track this +Can you please check itthanks with svn changes and pointed the svn to the new mailing list as well +Pull it back in if you think different. +I dont think Fortune example is a real argument +Ive changed the static field to long similar to + change to allow to default to the value specified inl still overridesNot specifying in results in the same behavior as beforeNo unit tests checked manually +class s signer information does not match signer information of other classes in the same packageI have seen this before when and are both in the missed one +Do the Jenkins boxes even allow access to the internet? Interactions with +Now that the issue is closed I can safely remove the debug codeThe fact that the debug code triggered a stack overflow implies that there may be something incorrect in the Scala code because the debug code is quite benign +Is this issue still present or can I close this case +Verified by Andrey. +Any pointers to a plugin which handles for nested XML configuration like this +The idea was to do a reasonable effort in trying to schedule the job before giving up +Its extension is +This has been seen originally in EAP but also tested with the aesh standalone +Ideally DN may not be able to find that stored block if genstamp is different from his volumeMaps block +Then each will containkeyvalueThe benefit is that we strictly limit the file per application to +Testing in secure environment is pendingRequires to compile specifically changes in +This patch only does this when Symmetric Binding is used +Note the needs a new sax directory in the test source +New patch that disallows caching of replicas and uncaches on delete append +I backed those changes out and everything worked again +Disallow more than one balancer running in an HDFS +So it seems a direct collateral damage ofSeems the fix was too aggressive and should leave exploded in instead +Updated upon discussion on inlined updateReference cleaned up the second patch replace the first +and i think metadatas should be stored in a path inside the shard instead in the crawldb +OK thats good to hear +related to this when one a setter of a property throws an exception nothing is reported anywhere its just silently ignored and execution continues like if nothing ever this is working the way it was intended Im marking this as an improvement +Not sure whether this issue is a intallation bug +This was separated from the other JIRA +So I am closing this issue. +Added class implements the default mapping used in +According to jsp specs instance variables set from tag attributes should only be cleaned in the release method in order to make a tag reusable when parameter dont table tag allow directly setting the list parameter the tag cant clean up the can only clean it if the users used the name parameter and didnt set directly the list object +Sure Ive reproduced insessionCreated is called when the user logs insessionClosed is called when the user exits normallysessionClosed is not called when the idle timeout disconnects the userIve reproduced with different sftp clientsve added a unit test but it seems this has been solved as part of the channel closing rework of +Correct SCM +Is this still an issue +Hi Rick just read the readme +If anyone has any other ideas please let me know +Then I have to do this check after complete compiling a operator if it contains a limit and requestedParallelism then add a singler reducer after that +Patch attached +Let the receiver initiate the block move The sender rejects the request when the maximum number has already reached +So I think that the thread you cite may refer to a different issue +bq +Please verify it was applied as expected +Serp is a fun oneIll try to get some time tomorrow to take a peek at this one +Changed the setter in trunk to have no effect +Of course to test it you also need to apply patch since test wont pass till changes is integrated to the was wondering what the status of this patch is +Appears to be a heisenberg bug requiring a redeploy of artifact to repo +If your patch or their changes to added back the option that should fix it +Any interest in reviving this and work towards committing a first version +The patch is missing the required modifications too +Put another way if you havent seen an diff at all in millions and millions of users its unlikely to be of consequence +This is sort of analogous to range vs frange +Missing patch was in ML by mfojtik +Minor refactoring +That aa is equal to in comparisons does not imply that it should regarded as a single character +Will create a issue to resolve this anomaly. +I certainly am and I do run the tool over the codebase on a semi regular basis +Ill see if I can get a replacement +slip +Its caused by a Hadoop limitation +Renamed the issue more properly +In order to make calls from Netbeans compatible what is needed to be doneI am using +This looks greatIf we can get some metrics out from the Master and through this interface then its a much easier way to develop a nagiosnrpe plugin +If not when do you expect to have completed the review? Thanks +My following comments are related to the accuracy of the understanding +Thanks for taking a look +do you own +I can not reproduce the error +Ive applied Jasons patch to and +Thanks Colin committed to trunk. +It must have been a regression +realized we had some duplicate issuesnote that the descriptioncomments in have some good examples of usecases we should definitely make sure we support test move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues all has this been integrated yet into any current versions of SOLR or is there a current way in a newer version then SOLR to perform multiple facet prefix queries on the same field? Any Help on this would be great as I am currently stuck from upgrading our system until this feature is available or unless I perform many queries per page like we used to before applying Uris patchThanks Robert +A major change in this patch is that adaptor names all start with adaptor +I definitely do not agree with the statement Nothing in Derby should break if we restore the original limit +Thanks for fixing this RahulDo you still need a junit testcaseThanksChris +Screenshot when viewing the layer in layer looks like there is a problem with teh type of the geometry attribute +Attach patch +In the future MS will support direct remote JDBC connections +I am working on a few improvements of the Plugin and should be able to submit a patch in about a week +Make sure you note it in the bw break section in CHANGES +verified that this does not happen in ER +The difference may be the use of a small journal data file length such that the files are quickly reclaimed when the contain unreferenced data +So from the discussion the test is only a problem in? Do we know why this isAsking because were close to another RC attempt for +Additionally need to check do all tests are included or not +Ill update with test today +Closing as trunk was updated to JAXB +committed to trunk +Please push it to both and +This is even before it reachesUsing cc Sun update C m fairly certain that the code wont compile with a that old SUN compiler I think you need something thats at least from the SUN Workshop line +I think the attached patch file is for a different problem +Thanks! +These are serious problems in the basic functioning of the framework as are many of the other bugs still being reported +Internally creates a special class that carries those settings which in turn will be by +The task tracker should create the task directory the job tracker should create the job directory user code should just create files +It still applies for branch +Simple patch to add back indexinterval to Config and use that as default when setAlso added lines to and for upgrading +Attached patch fixes the problem +Finally apply Needed changes for build and tests to run successfully +I wont create a for this +Chris fixed this in r +No thanks i dont want to read my entire documents into RAM and have massive gcing going onThis is obvious +Now just a single autowired service two tests using the reference second one with a sleep and then doing assert on service reference +My vote would be to leave them returning void and have the default implementation do nothingA few nits all public and protected elements need javadoc +This looks like +Actually correct axiom jars should be delivered with the build i thinkI would prefer to give you a more focused example that would be much better please send it soon so that I can fix it before Axis +What you think Mubarak? Could you imagine to mark this as documentation? +count had it right +Will fail until class is used +Even if that turns out not to be the case theres obviously something broken so we need to find it +Thanks Jonathan! +such as Distinct +IMO Vaclav hasnt given us enough information to reproduce his problem which probably isnt due to a bug in setVaclav Ive posted a C program to the mailing list which demonstrates the correct behaviour of set +but there are a lot of hard disk is this still an issue is more of a wrapper on Java File IO Please let us know if you are still facing this +started to port model repository over to trunkworkflow +But its minor and rather obvious to fix +Patch to fix rangeslice on large datasets patch to allow unlimited columnssubcolumns +Got no more feedback so assuming the problem is solved. +Thank youIf you want I can build version from svn and test it before final +Attaching a tiny patch to document why this test is getting skipped with jdk ibm +Suggest following version info design in base see impl methods get get and get and putting the check somewhere in the overall startup where it could be done only once if possible +There must be update to latest version as well +we finished the translation of all message resource bundles into JapaneseThe resource bundle for which had posted at March was also updated +Ill take a look at the differences between Morfologik and Morfeusz right now actually +Were your intention to use new one but mixed it with old one Ive surely done it once or t know if we should support the old one because technically speaking it is deprecated api +Sounds like this is a safe backport to +Moreover I had to specify the LIBXML variables anyway because our libxml libraries are in a locationLastly all of this was despite building with to check in a RH platform +It looks like the twiki has been changed to mandate the use of the internal maven repository approach even though weve not yet reached a consensus on the correction solution +I will try with nightly build later tried with nightly build and got the exact same behaviour +Failed a second time because of +bq +WSDL to reproduce the issue +Most recent module is marked version but XBRANCH seems to be and the trunk is even olderIve reopened as the bug was apparently fixed in an abandoned branch or even repository. +Marek are you referring to the of the showcase war? Its the only file I can find that mentions did the distribution build and it contains the files you are referring toI tracked the file to originate frombuilddistand the file to originate frombuilddistIll update these files as appropriate +Reattaching diff +According to Andres last comment the groovy console on systems other than Windows so Im closing this. +Wasnt really a problem with the YAML itself just an implicit assumption that value nodes never are null which in this case it actually was +I dont think so +Fixed the version reported against toMy goal was to read the status line and figure out from the response that Im talking with a server that only supports HTTP based on the status lineof the response +This ones against trunk +how bizarre +The problem is only observed on the very first request after the bundle has been installed +Bulk defer all frontend issues to +Jeff Im hoping you have a bit more detail on this one +Thank you! +that doesnt mean that all possible combinations have to be tested +The patch to fix the issue +I think its too risky to take more component updates for this test failures +Andres thats fantastic +Not really a task required for incubation. +Hi allThe issue is open for yearAFAIK most people are satisfied by currently used solutionsProbably the issue can be closed as WONTFIX what do you think +those are on the list +Ill investigate +Mmmm true +Took a look at this belatedly +needs a check before removing the from its context map to avoid an NPE at runtime +A CI server deploy the artifact and in some case no one seem to get the updated artifactThanks for pointing me to other bug report if my problem is out of scope here +Reverted for need to think more carefully about how to do this +Would it make since for me to file a separate JIRA to add some to expose this information +So that we getsrccontribzooinspectorsrcmainresourcesiconsTango +It could still be broken since the default bind URL may have changed due to other fixes +It isnt clear what the right API for this would be +Okay glad that it is plannedWell need to drill down into the challenges of multiple kerberos protectedclusters a bit more there is likely a way to make that work +It passes unit testsI just committed this +Ok to close this one +Because that is the package structure that is used +Closing issues +Im using to verify this issue Developer id date Cannot reproduce exactly the same scenario because there is no such configuration in Seam Web Project Wizard but switching between others configurations doesnt correlate with others new wizardsWhat JBDS version are you runningDo you have any exceptions in Errors ViewIs that clear JBDS installation or something else was installed at the same eclipse instance +Knowing that would help a lot to track down the root good to hear that you narrowed it down already +Now Ive got the idea of the issue the description was quite messy and incomprehensibleTitle and descriptipon was updated +Probably outofdate but assigning to Nick to verify +Unfortunately I have Windows and doesnt seen to like your archive file +Safe mode must be turned off manuallyToResources are low on NN +got +I ll definitely add Semantic compatibility as soon as the wiki comes back + +Jon Stack No we dont have to mimic the Accumulo API though if the goal here is to be an alternative it must be possible to build a direct API translation shim that provides the same labelling and visibility semantisc +Yes this is already known from +Great work +I ran the latest patch on a way machine with threads +Latest EAP seems to fix this at least to a way where I can continue to work +I accept the and withdraw my changes +Do you have a patch to fix it? +Nitay yes that would work if we were checking sizes of Objects but now we are just dealing with classes so it is very hard to take that approachStack yup it is GPL I just wasnt sure how you would add specific jvm arguments to Hudson been trying to get it to work from within eclipse without setting the arguments but without any luck so far +should be a quick one to do +This is fixed by introducing new alternative class for Java and use it at already have this fix localy if you need it I may upload it +There is in attachment the Dashboard plugin with the correction await your opinions +I have not tried killing just the Ganglia server +I think it is a bug I couldnt find similar bugs after searching and hence logging it +just doesnt error well +Mate you submit multiple trivial timewasting bug reports every week +Its quite unfortunate that nowadays this seems the only way to deal with those moronsWell at least there is a way left +trygvis IMO next to edit but it might fit now button added beside the edit. +Thanks for the patch and your patience +This issue has not seen any progress for a long time and it may even be worth raising the priority even higher +I cant think of a way of making SOAP WSDL work without having a targetnamespace for your +In any event the NMS API dll has no client logic in it its just an API +bq +Jseper As this issue will be resolved in EAPEWP is it possible if we can get a correct DTD for EAP version +Was able to load a table delete then refill into table of same name and schema three times which is much better than I could do previously +That was a silly last minute change that went with the wrong polarityCommitted another fix +tomcat log file. +Merged patch with latest trunk +The is without the NPE handling +Sorry but I had clicked the Assign To Me button justas I noticed this was already assigned to Michal FojtikAssigning it back to himIf you would like me to look into this let me know +You need a way to measure distance and a good definition for closest +I am not sure which one is correct +Case and Case below have identical textdo you miss something thereThanksGlen +Do you mind working on the documentation +One run had failures and one did notI dont usually run the junitreport target so Im not sure if its intermittant or happens always +Test case for a namespace node that is not for the default namespace +Thanks! +Patch approvedffafedeffe. +Appreciated Phil +This does the following creates a number of roles links them together in a grant chain drops the first linkMysteriously this also seems to drop the last link but not the intervening second link +Hi we are a group of CMU master students that is thinking of contributing to Avro +Ok it should redirect so I guess that is ok +Suppose it also need to be applied to the master +and I dont foresee the reutilization of setup routineHudson is currently not publishing the build results +I cant duplicate this with a simple project setup that I have +The close operation doesnt seem to work though +David can you investigate if you can quickly fix this +this simple patch test cases under search pass +Agreed +I should not trust in my mind +I think that the regression shouldnt be hidden in the exclude list and should be reflected on testing pageSo I rolled back my commit for exclude list +Old JIRA cleanup +Issue also existed on set and set fixed also. +Cannot cast from to Nevertheless it is a good idea to support a wide range of versions +Based on the email thread I gather Geoserver support depends on GEOS support + checked into trunk and branch. +Thanks +Looks good. +Hello everybodyanother scenario that shows how important it is to be able to configure seperate output folders are web I develope web apps or portlets I use the exploded mode of tomcat for testing +hope to have good news this time +Tests are fun so I added some! Given the code in question ee tests seemed most appropriate +Ill resubmit it if m happy to make a patch +Here is the patch file that resolves the issue + +So to clarify would the logic look like this? For each pool initialize and schedule at least its max number of jobs +Committed thanks +harsh would you have a time to review my patch +This is getting quite annoying + +The rebalancing logic comprised of the following steps release partition ownership commit offsets assign partitions own partitions restart fetchers +Then people who just want to track the status files dont have to check out each project trunk individuallyWhile it would also be possible to use svnexternals in incubatortrunk to link to all the status files in the project trunks that would create yet another metadata locationcheers Roland +Absolutely fantastic and incredible job! Thanks a lot Thomasrevision +It appears this is the same issue as +Thanks Stack +It does not inherited it may be refilled +All these issues seem to that the special char problem in am attaching the wsdl file +issues +httpshudsonviewjob +All that cell handling is is most of the work that hbase is doing and sometimes its just doing it to rebuild a near replica of what was on disk anyway +does that mean we need to do a better job buffering writes ourselves +Fixes missing var +Can we add support for SLJ instead or as well as commons loggingIMHO SLJ is a better solution for providing a common logging API and is gaining community adoption across main large OS projects +This script will display the current values of those variables +Removing the default them was probably a crutch during the development of falcon but I think its time to remove it and see what happens +fix testcases + +I attached patch and please review +missed a part of the fix in the apply +Maybe we can improve the error message for trunk +is now resolved which covers the possibility of creating custom nodes and custom name classes +Weijin are you going to work on this for ? If not lets move this out to +bq +Updated version of the patch dated +failures are because of my bad commit earlier today Elliott ClarkYep. +Yes that JIRA is the final piece in the puzzle +Oleg has included this fix in the current release candidate. +Nothing never hangs with though any exceptions on client side during deserializationsIt is better because with or it hangs indefinitely nevertheless client jvm kills +patch resolution wrapping unchecked to another +Attachment has been added with description the screenshoot showing the document generated by FOP in AcrobatReader and the generated document in xpdf +Jira is good for task tracking not design discussions +Anyone have an answer? +Note that the logic for checking results is not in the process operation but rather in the check operationThe is not designed to raise exceptions on missing cryptographic actionsNote also that a fix was made in the results checking logic for so you should check against the latest verstion of WSSJ as Colm suggests +Thanks! it work correctly +bq +Its enough that it has performed an update without committing or aborting the transaction after on +You should be able to edit comments +Adding or updating test cases for both HTTP and SMTP deliveryfindservicefindbusinessfindtmodel +Im not even sure if thats documented and its confusing +Its working well enough. +Three of the four requests have been completed +added raw methods to +to be a in interpreter +looks like a good candidate to be moved into the platform perhaps we can discuss it at length later + +Cordova plugin filter does filter on the ids +Also the final goal to have all test run with only +proposed in rev snapshot deployed. +Can you attach your wsdl? in Callback handler it should have the method signature your are propossing +I have concerns on adding failover retry etc +What practical costs are there to using the hadoop jar as it is today? Dont we have bigger fish to fry +redirect method handles the null value +I have a Pentium III with MB memory and when I start the process I use java Xmsm XmxmThanks +Is there anything I can do to make this happen +This has been sitting idle for a while but the patch looks fine +Removed the unused import +updated modified from modified from Your patches will be more likely to be accepted if you submit these as a unified diff patch against the latest cvs code rather than as the whole file from an older version +Only considers a connection refused exception as security disabled since an insecure cluster does not listen on the secure port +Submitting patch again with suggested change +Chris the code where this is coming from is relatively complex and so its not a straightforward thing to figure out what the cause is and whether the cause is a valid one and should protect against or whether instead is a bug somewhere else in the codeUnfortunately theres not much logging in this area so I dont think itd help much unless it can clarify the following doubt I have Are you accessing Hibernate within transactions or not? If youre using transactions what type of transactions and which is the transaction manager that youre using? Can you attach your hibernate configuration file? Also if youre using transactions Id like to understand the flow of transactions and whether a particular transaction can be accessed by different threadsBased on what Ive seen a particular key can either be cleaned upa This is materially impossible but by the same thread twice a row +heh i fought that guy last night for quite some time + +I think is a better approach and would prefer to see this used as sparingly as possible +Docbook editor required that can also insert the relevant freemarker template code +Has the problem been corrected with +It is the expected behavior +bq +If we are going to pursue maybe we shouldnt try to hide the logs especially if were going to provide an option to reuse the same directory across runs +I think we can keep the option turned off in the mvc namespace maintaining backwards compatibility along with a flag to change it +Applied patch to build +gitignoreShouldnt need testing only relevant for eclipse configuration +I accidentally closed the has been applied to the brand documentation +Lets close this unless Aaron has a workload where hes seeing something different. +I am hitting the same stack +The playback of recordings also does not work anymoreSame if you upload a video to the import works but playback on the whiteboard failsIt seems to me like Red has a bug in the video streaming +Im not sure how to go about doing this but Id like to have the patch I just attached code reviewed +I now use an Eclipse plugin to edit the bundle files +As soon as i set my timezone to CEST EuropeBerlin the hyperlinks in the calendar disappear +is s elephant is wrapping everything within a binary block which is basically a protobuf message with additional info order to search those block from a stream +I removed the ref counting code that I added in the previous patchLemme know if you hit any snags +Integrated inital patch into castor source tree +Logistics for remote attendance will also be announce publicly when we have that figured out +Thanks again! Was kinda awesome that I could get this done so quick +OK nevermind then +Does it sound right to you +Rob can you provide your specific requirements so I can be sure they are addressed as we continue on work +maybe we are not sending the correct content type? maybe the abort is actually from that resource loader you put in place for css files +Fixed in the latest trunk +with more comments on top of on top of these review once the patch adding enters svn +The issue ask that npanday project deploy the sources artifacts when the main artifacts are deployed and let those artifacts available from npanday repo +The implication of the incorrect progress affect scheduling of speculative tasks for the pipes jobs +Can you confirm that refers to a or association? In the future the appropriate mappings are usually needed to correctly diagnose problems +Thanks Knut for look into this issue and provide the solution +It is skipped for S and passes on +But we could instead +Our artifacts are also a mess Using an from a different release does not seem kosher at all +launches successfully but doesnt work because of incorrect configIt seems to work for me but maybe Im just not trying the right commands +I have confirmed that the bug exists in the latest version of WS +We are not using any of its methodsAnd is not an interface so now you are enforcing a strict class hierarchy where that is not necessary +Sorry but cachingEnabled is true by default so statement on line is nottrue by default so line is always reached by default +Patch for trunk +TPS seems intent on comparing a fictional prior release of EAP with what is in the errata +Build switched to use in trunk +Thanks Steve! +Just to confirm this bug affects the newer API +JDWP test on field watches works now. +I expect something like more reasonable initial size +Not seeing the problem if I revert CPUTF so it looks like we might need the encoding after all +Seam Core Manager dialog has the same fix is not so trivial as I fought +do the opto we do today +This patch requires that be applied first +The patch also contains implementation for that can be used with to read vectorized data from RC file +I had a user defined in with userroller and host which I though meant any host but apparently not +I like +I would assume that the is triggered by the parsing of some large Word document especially since you reported that the problem does not occur if you disable the Word document filterThus if we catch the caused by a single document it will should not interrupt the whole indexing process +Patch for the does not work under Win x so I made some modificationsPlease check that everything is ok and close the bugThanks Vadim +The original file contains a p and the textsubstitution done by the schema tool should not alter this +Changing status of the JIRA to Patch submitted so that it will show up in the review queue +I just committed this +Hi just find that for there is no resolve action +Odd +bq +Someone should probably check the command line on Solaris before we submit the patch again +Messed up with issue number in patch filename sorry +All the generated code is under +As suggested has been opened for this +When I wrote that paragraph I had ICV in mind +My comments were mainly in reaction to Mathias observation that should also provide support for resource injection when running in a JEE container +Need to document the following properties +Hi Gunnlauger we will be fixing the current jiras and releasing in the next few days +You can make it generate the request XML and change it transition all resolved issue that did not see any further comment in the last month to closed status + to the trunk and the branch. +looks good alex +Fixed forgot to mark this as resolved. +Fixed in trunk and will be delivered in release. +Ive this fix to. +A start I called that mode alwaysRedirect but later I realized alwaysRecompile is more meaningful because the idea is recompile the facelet to rebuild the EL expressions if new params are found +As a product evolves things change +How do you handle identical names with different namespaces when you load data +I am linking this issue to the other issues which prevent you from using SYSCSCOMPRESSTABLE and +Added an example to FAQ to show how to implement a custom listener by extending. +I can agree that it is not something that is hampering development +PR applied on master and in JBDS +also includes changes to test asserts to print more useful message on failure +Now that work is well underway with AS all previous community releases are +Tested it works whether render is true or false +Otherwise I think it seems okay so lets try to get it into patch with thanks to Ben I also updated the spring test for it. +Rostislav Ive updated that version numberA couple of questions before I rebuild +Finally Enis can you please comment on what type of issues you consider as over and above what this patch addresses +A quick fix only changes a few lines in and lets order by clause share the same child clause with sort by clause +Added logic to account for resource resolution on Windows +Committed to branch +I could not reproduce on branch but since I assume youre not mistaken well say it got fixed sometime between and +Erik I admit that your handlers Velocity template presented in was remarkably succinct and readable but it does force the use of Velocity which isnt in core just as arent for similar reasonsIll admit to beating a dead horse here I suppose +Ok lets see if I manage and find the time to fix this one should fix the attached jar has a patch for this issue +If you remove completely pleaese also remove and force users to use with a +Hey RossThis sounds good to me +If someone wants to partition their data into they are free to do so using locality groups +Added file parser and basic pattern matching +psml pages to hidden the portal will still work end users can still customize their own pages and cannot access the administrative portletsUgly but it works +Im now checking that and are only called on numerical in revision +Would it make you happy if we added to the javadoc comments on this function not to use it as a key in the same job its generated in +First thing Id recommend is that you upgrade to NMS as that is the latest stable release +patch looks good +Modifications are marked with I believe it could solve the Bertrandplease find attached the Test class that shows the issueChanges marked with string +The issue is with the kill command in later Ubuntu versions +Could you please provide some details on where your em reference comes from? It would also be great if you could debug into the proxys a bit Line of is where the Query proxy is being built using the interfaces of the target Query object for the proxy as have a suspicion that the from has some special arrangement behind it +. +in fact I couldnt find any d packageIs it released yet +no problem +I extended Craigs patch and added two things Exit the test method if the binary compatibility option is not supported +Fixed in Eclipse will be released for +I made a change to refresh the cache which will be in the this is indeed the problem you are seeing restarting the agent should workaround it +right +It is also failing +bug fixed. +Verified on JBT +It fails with the current based mapper +bulk defer to +Any URL that ends with a is assumed to refer to a directory +javadoc too +As long as there is some reasonable effort to remove rows with bloom filter disabled OR its clearly known that a reasonable FP setting is required to remove tombstones I think we could have avoided a lot of headachesHow does the new tombstone histogram feature in affect this issue? If that feature solves the problem already maybe this fix is irrelevant +reopening +Okay thats sounds like a good idea Ill leave it for after the merge is done though +Kurt could you check to see if jBPM has a way to set the task priority once the task has already been created +If you find further issues please create a new jira and link it to this one. +It turns out that this is not foolproof +Can you please add some more comments to this one +Currently we compare apps in ZK vs apps are deployed to find new apps +also tested that it am going to close the issue for now +But the change needs proper testing +Another thing is that the default table engine is often and this engine is not supported in a cluster config +Omitted the OTC change since in its just shoving a bytebuffer over the wire theres not really any to worry about +When using libcurl connection handling including worksWithout libcurl neither nor works +Willem Ive applied your patch as part of but am wondering why wouldnt we just use instead +Will submit it to hadoopQA +The find on shows how to do it which is For every find the of type OBJECTTYPECOMMENT that precedes it On that get the object ID Find a with a shape ID that matches the object ID just found The holds the row and column detailsNormally the ordering of records in the file is cells so youd need to grab things as they went passed so theyre to hand by the time you get to the cells +I am working on creating a write up explaining the previous behavior of the filters and the behavior of the filters after the fixMeanwhile attaching the patch to let the hadoop qa run the the same patch to try to get the Hadoop QA patch attached changes the operations in the to the followinga Check if the current key value is present in our requested columns +In and above any way we have PB based response. +Basically I need to break up the project into two separate projects so I can load the commands in a separate with all the jars in the lib and libext directories on the classpath +The existing Ant script and file structure works well +bq +bq +Note initially in a fresh app the Webkit db is not created yet until you use localStorage at least once so on startup you might see a file not exists error which is ShazThanks for the plugin +This is a direct consequence of what RFC says about Modify operationA replace with no value will delete the entire attribute if it exists and it is ignored if the attribute does not exist +I dont have the host target is null errors anymoreI dont see anything missing but Ill try downloading and building it again +We need some logic to perform quorum verification aside from values that we load from a separate configuration file +Terminated with extreme prejudice. +ping Jon +Assigned to version +The schema has been moved down into the orgapachebeehivenetuicoreurltemplatesschemaNote the svn move of the schema does not show up in an svn diff command for creating a patch +I knew it that hflush on every call may slow things up but just didnt want to do premature optimization +if you want this to take immediate effect you should rather use volatile here? I doubt that this is necessary in this context Id rather not invalidate a cache line for each creation +The code for this issue has been checked in +Does anybody know which axis version and which databinding is stable now +So in a way tail f reports an illusive data that is not guaranteed to be present in the futureThis patch has a lot of important internal changes like removing pendingCreates etcBut introducing tail f at this point may cause confusion among potential users +I out the createNode method +Having said that I have come to see a few examples that initially want beyond my imagination such as a choice within a complex type definition where one element definition was part of both choice parts once being declared as optional once not +Thanks Mitko for your contribution +Issue closed with release. +Im going to edit in your patch +But let me have a go at this +Option seems like it would result in the following sequence of operations zookeeper server is passed in Use zookeeper server to find location of master Ask master for list of zookeepers Use this list to create new zookeepers connectionThis seems a bit odd +I think that should also happen +To kick this effort off heres what I think should go into the sink drain a channel and load all the nodes into the graph drain a channel and match nodes from a graph and return the matches drain a channel and match and delete matching nodes drain a channel and calculate shortest path through a graphDoes this sound like a good start note that the sink will be built generically to go across Titan and neoj I will be tackling neoj first +Verified. +Ted Agree +The previous patches were +Thanks Stepan the patch is applied as expected. +Hi AmreshwariThanks for reporting the diffI generated a new patchHopefully this fixes the issue +Fixed +Closed issues related to Roller release. +Currently we use warExcludes and redeclare all the deps +Please assign this to someone else if youre not the right person to take care of it +It more like utils library to but i dont think its a good idea to add another hierarchy to the repository +After the container failed to passivate the bean I got this exception by every bug means we have to restart the server frequently during development and means we have to wait extra time for it to come up again +This patch contains the fix for as well +svg and vml are the future so i vote for increase numbers of functionnailties of be deleted +I uploaded the new SNAPSHOT +wont add this feature in the hudson environment. +I decided to go with the properties file approach +Would it be possible for you to try the attached test program against the email attachment? and let me know the results +Master initialize after is created +Petes happy closing. +Please take a look and drop in a comment when you have some time +Once we basically agree on design code reviews would be easier because theyd be focused on the implementation matching the designAgreed that before doing anything we should do a design doc and agree on that +Thats an interesting idea very useful for cached portions of the page +See Wiki page Groovy Method Invokation for details. +But then is very common name as well and it was accepted as suitable +We need to keep up our math +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Patch makes sense +Moved existing code to and added abstract hash method +Dhruba what do you think of returning an extra null entry when the list is known to be incompleteIt looks a little hacky but does the job quite well with minor code changesWe cant just test the size of the output provided at the client side because is a configuration and clients might have the wrong value +Passed Failed Error Not Run resolved issues for already released versions +Now we have the same issue +is renamed to statReal +The ESB is not blocked but it is under a heavy majority of the time spent processing the requests appears to be taken up with the XML serialisationdeserialisation process +shipped with +Does that work? Did I understand your comment correctly +Committed to trunk with revision +That means no change of current behaviour +Charset of the applicationjson changed to +Fixed on M will merge on to trunk before closing +Tested this on a single node and the bugs goneChecking this in + +I am also experimenting it +Selecting two profiles results in dependencies not being resolved correctly and thus in compilation only one profile is specified the dependencies are resolved it appears that having commas in the id is acceptable in that case the two profiles which are separated by the comma are interpreted as one single profile and dependencies are again resolved correctly +Revision . +Straightforward patch +Eck +Both are version +you need a dependency on your checkstyle rules at least under maven +we just need to flag one of these as an incompatible change and throw a release note in there so folks arent surprised +I can see that both describe and dump work just fine +Implementations should restrict names to simple ASCII characters that match the pattern in the spec in almost all cases +guess is that theres a reconnect somehow dropping the message a la +asplugins +It will be better if there is something in the message that can can tell these two apart +Oops Im not sure the patch is applicable it contains my local directories +Fix looks OK to me +AFAICS +The user application has requested that the jar for a class be packaged and sent to the cluster +Thanks Niels. +Changed SSL certs expiration checking from per keystore to only when connection builds +DzmitryCan you please provide more details about your application and how to replicate this issue? Ideally you can provide a short roo script to create the app you are seeing and then the exact manual steps you are taking which lead to this repetition of actions by Roo +The test has problems when run in parallelAlso the code was not merged correctly in branchx +I think we should just keep it simple +Thanks for the clarification Anoop +You are mixing up two things here +I think it is not necessary to deprecate the constructor of +Spring is not really a must as I can use pure Servlets for the backend +FYI is used by get and niceEven if is reached it only fails on corner cases +However the DNS name wont work but instead he must use the IP directlyMy apologies +Thanks for the patch GregCommitted at revision and in +Apparently a +I find them disturbing especially for developers not yet acquainted to it +Have finally figured a way out of checking looking at the rawmessage option on modmboxCant check for as their list is very new +When I ran ant test though one of the Anakia tests failed due to an invalid macro callThe problem is that there are likely thousands of Velocity templates out there with a macro argument problem +Given is fixed jenkins shouldnt complain anything this time +Please teach me url +GL Transactions +is intended to allow writing data but FB never really cleaned it up to make it useful for anyone but them +Reported fixed and confirmed +Theyre enforced at scan time +The example is identical in sources so the technique should still apply +Is there a error in the build process +Verified at r. +This is fixed on trunk in is a fix on trunk to the original change in r +Ill need to do more research to make sure this is safe +In maven terms the best way to do this is to put your scripts into a module generate a jar for them then include them in the you cant really assume with M as it is now that +Committed +Is it a logical plan +Otherwise it is going to fail the applications +Ive mistakenDS component should be reRendered in any way explicitly +There is a subquery evaluation order flag in the configuration have you tried thatFor many use cases it is useful to have the subquery not reflect the current event therefore the configuration flag exists +We could be adding pending deletes to segments thatare not quite current because were not adding them in an IWsync block +The goal of having a release will be addressed in +CPU memory barrier is a bit more since it requires are synchronization of the main memory with the processor cache +done enjoy +The collected works copyright just needs to be the ASF not or its licenceesThe patch changes the licence text in the overview html file I think this is wrongThe previous text seems to be fine +Hi Andyhere is the output of the grep on my installationjar tvf grep Builder Thu Mar CET orgcodehausjdtgroovyinternalcompilerast note the time I do not have any version as you that I have managed to keep the old version aroundr root root Mar root root Mar I wonder how this could happen +Hi Jochen I just committed a fix for this +tiny fixes and renames +Had it occur again also for normally named class. +Thomas Diesler has approved the new functionality has been added to branch remotingx and the old functionality has been preserved +Do you maybe mean that this enabled you to navigate the document and address various parts of it is + was unused property so its was +We should only need an export from the old and import to the new +Some attempt is needed to hide the session secret to ensure that sessions cannot easily be hijacked +I committed the patch to trunk +I think the second patch based on the servlet containers path info is better +Might be worth it to refresh my memory again on the purpose of the module? I always thought it was used as a starting point for users that have a complete configuration with a components wired together +Right you really want to use +is defined in the war +Id forgotten to merge changes in from trunk sorry about that +Yeah youll need that installed +Here is a patch to make noEMF option the revision + results wo results w VFS from March reports youll notice that we do break some File IO and things for paths +Good advices! We will use them +Thanks for pointing +The conf files reference aA lot of this is the pig specific portions of the code +The patched build has passed the added unit test case as well as the standard QE set of functional back to Matt +Pether what else needs to be done on this issue? Are you able to generate the eclipse files now? +May fix the test +This should be fixed before we do a release +We have tried once with the limit +what is in +Although the patch here is not small I think this one is a simple JIRA since all changes are straightforward + +In the end the build file is just a sample and it is easy to change in another environmentI agree that the path would be easier to understand if it was pathto +Reviewed and pushed +You mean get it into the trunk and backport to? ExactlyI will wait until tomorrow before committing this +Perhaps the problem is just that returns instance and not a String +gtr invalid backslash escape at end of stringecho i tr sed s +Now on major compactions deletes and what they overshadow are NOT added to the file. +Hi BobThanks for your commentI updated projects +Close issues after release +Ill post my findings +is the correct table to discover sequences on db. +Stumbled across this ticket and saw its been sitting with a patch +But by default Groovy does not provide any such class +We should continue to include flexible that permit arbitrary data formats +Otis I completely agree with your comment on the naming +I just committed this +This was not the case for the first entry +The code add into was deleted by Trong Tran in revision because its processed in processRender its not good +It seems to be duplicate with you are right +What else needs decoupling like that? +However this leads me to another issueIs it possible to configure the code to sign a placing the above the Signature? If I use actions of first Signature and then Timestamp it throws an error saying it cant find the Timestamp to sign +I am wondering about if it is possible to enable highlighter support for and by using this patchs solution +patch for solution implemented is to have it start a new thread to do the actual localization and task launching for that particular task that way it doesnt block other tasks behind it if the localization of one task is taking a long time +my problem is just that the server does not recognise it unless I specify the type +Latest patch looks just committed this +This turns out to be a bug in storage directory restoration +Andreas we already got a couple of tickets on this matter to have CRON style scheduling for consumers which supports scheduling +This is just an edge case from what I understand +We still got the chance to make the official release work properly with remote transaction fresh from the source +The examples page is pulled from the branch so thats why the blurbs wont appear there +I suspect the two are unrelated +Server side Cookie based session need to be implemented seperately and most probably for in +Thanks Graeme +Committed revision . +Fix help messages +JimmyDo you have a test case which reproduces thisIs hadoop used when this problem happened? am working on the test case now +It follows a slight modification of the second proposal above which isbq +Emily do you still need this patch considered if the safe conversion path is applied +Cmodules this sould be corrected as Cmodulesapache servers default port is +Section + +Should be fixed +I can confirm this behavior also for ordinary Stateless is used only for the renders quite useless because I cannot make calls to secured servicesfrom other beans when using +Yes we dont need any of those for now at least +This also handles the cases where the most recent release is earlier than the most recent snapshot. +please update +Theres an issue with uppercase letters you cant find a component that has an uppercase letter in its key +DuraiI agree that having PMML support on top of Hadoop is a great thing but my concern is with the maintainability and scope of this request +Are you seeing the exception caused by the code interrupting the thread? If yes and it is a persistent issue feel free to submit a patch implementing the workaround. + +Patch committed to trunk +The thing is that that would touchinfluence much more +I dont really know where to go with thisIve never seen this problem and looking at the code it isntclear how it could happen +If you want to go that route I suggest opening a new ticket and closing this oneWhat you think we should do D +Can you a not loginb login and not do anything usefulc login and do some useful things but not manage build plans +If it isnt we drop the connectionIf you want to add further security during the socket formation phase nobody will object though we might put it on a configuration basis so others arent impacted as it will slow down launch times on very large clustersHTHRalph +I found the bug ! characterEncoding In the config file the admin uses a capital C that breaks it +its just when I run everything that the logging is missing +Nobody complained so it has stayed as a low priority for me +Thanks JacquesI will provide a better patch later today +Can confirm that this bug fixed worked for me too +So thats why I decided to go with the jettyrun figuring youd eventually pull it into a war anyway +Unfortunately I have a deadline to fix this so unless there are any other suggestions I need to do this today +You can follow the progress here thanks for taking care of it Lukasz +The contains +Thank you +This issue can happen with the new balancer +However I think that the benfits to testing alone will be worth it +Added the addendum +This is a bit of a usage issue as much as a bug C object lifetime is not like Java or other Garbage collected languages +Sounds good to me +Last variant proposed by Aliaksey is nice +use onchanged instead of ondateselected +Accepted changes from Senaka committed to trunk by Martin and reviewed by Robbie +I think Yonik answered the question +Had not thought on security for this use case +If anybody has any other comments on this JIRA please comment +codegen fixed in r +It belongs in the class +Turns out that the name given to ODE engine as the name of the Service is nothing to do with the SCA service name and is also nothing to do with the WSDL serviceportType name +annotation now supports inverseName to define the foreign key on the side of the relationship +Well evaluate if making changes to those classes is the right approach +Thanks for the answer +Sounds good +And verify +Thanks +After more consideration we can not conclude that the for the EJB MDB is actually UNKNOWN +Cheolsoo it would be extremely appreciated if this can be made part of since then it would make Bigtop use one less workaround +Patch applied thanks! Luke keeps getting handier and handierI also agree that the analyzer output is useless especially with the className attribute now. +Yes thats correct includeDirectory not includeDir +The old jcruuid and dnalockedNode properties on the dnalock node type were combined into a dnalockedUuid property +I tried to apply on trunk but got rejections +Bulk closing stale resolved issues +for Simons patch good catch +Its worse The params are also incomplete in taglibs +Why would one want an empty partition +Load doesnt have to be extreme +Note that the proposed patch has issues if the index already exists Im not sure how we can check for that case inside s been a patch for the better part of six months now +Its not in the b patch and this makes the build burp +AwesomeThanks again Galder. +The user would select one of these and that would determine their event type definition +I cant reproduce your issue with any Roo project +Returning to stock Ubuntu given by jcloudsFixing one bug where the tests run some of the commands using jclouds compute service allNodes which fails when there are unrelated active +Using prasads assembly plugin that I believe he is attaching to this issue and trunk as of rev I got a jetty server with modules to build using this patch and it starts +dir +Finally I got the ok from Nikolaus Hansen the author of the original algo +Needs merge into +Induced by change to in +Fixes tagattribute issues in duplicate +Deprecate this feature because of validators to add handling for this attribute in ajajax +I looked at the history of this file and it appears to have been missing this role from the very beginning +Assigning to the docs team +Is it something in the or as a config info that the rpc layer usesWould we be able to use the major and minor version number to automate thisie if major numbers mismath then incompatible otherwise compatibleClearly the special calls to call older less optimized methods when talking to an old server have to be hard coded in the client side +I just dont have time to convert it +Maxim is this a regression +I scheduled the issue for the Beta of +I think that if you build and run a test case for this code it will be clear what the problem and solution are +sorry but I cant reproduce this in or the given test project running releaseprepare is expected to fail there are no snapshots +I will upload an updated versionv adds those files to the script +I get the same issue +I will be doing this attached +So far so good +Patch is now tested against hsql mysql oracle and postgres +Groovy requires ASM +Opsj provides all required versions of slfj +In all cases it properly retried when the retry values were +bulk close of all resolved issues. +Hmm I am afraid but no port whatshoever would be the best +Ok I seethis would mean it is a problem with dynamically adding new HTML content in firefoxregards Assessment by Sh your last comment + +However I think its reasonable that we enable local address reuse since a common use case for is unit testing where it makes sense to start a gazillion servers quickly after each other. +Thats correct if already has a open on that segment because its pooling readers in turn because an NRT reader was pulled then if that segment is merged it will that already opened rather than open a new one +So the frequency of fetching cpu should be more than memory and memory should be more than disk +Generated results on WC test suite +bq +What I am thinking is perhaps creating a thread local that will use to store any built schemas it creates in it +This is a bug on the datanode side right? The tests from at least ensure that CHECKSUMOK works properly where its supposed to +Patch for attached +Great thanks Josh +Patch that adds capability to disable lowercasing of keys in via a System property. +Resolved by in +Please review the attached applied with thanks! +adds an behavior as the close callback in the main page tree after this page gets serialized with the initial value set in +Deprecated calder schema and support for streaming querie +Battling to recreate this one in an isolated environmentHappens every time on a warm start of our application but cant seem to isolate the combination of bundles bindings that trigger it +This was found by Findbugs +The basic idea is to check if the htd is null +New patch which fixes the logical to physical translator to use the load function interface only in the case of bytearrays to pig types during object construction and serialization +The series is no longer being actively developed and is only being updated to patch serious security vulnerabilities +Ok +Sounds like a good idea so I am putting it in +Many thanks. +It would prevent accidental loss of data and ease the implementationWhile I rewrite this would you agree on which provides the basic data structures for thisThanks for your comments +Small FYI on patch submission No need to name them etc +I couldnt reproduce assert on platform but reproduced it on LinuxAlso I split VM part into separate JIRA +Attachment has been added with description Describes the source changes to enable roles to be defined in the has been added with description A simple fo document to replicate the error with +Great m sorry it means the packages available at +what changed from geronmio to + +I can try wrapping the code between pre and post hooks with trycatch in a way that doesnt change indenting inbetween +Moving out of was against trunk but should apply to both and trunk +Replace the internal collections in with and initialise those when needed +Thanks Russel for it +Lets evaluate if we are able to simply fix this in scope of Alpha +all issues to the CR +thats right +Left panel as the common navigation Tree structure +If the response code is set it makes sense to log it +Dominik +Ive improved code and test case +Radoslav please close this issue if you agree +This patch excludes the basedir from the filtered paths checked in the tests +It seems that I cant update status of this issue since Im not the one who created it +Fixed and committed. +Those occurrences should be replaced with. +Thanks Andrew. +First iteration complete flexible base for search engine implemented +We didnt change any part of code +just committed this +Seems corruptedjar tf in JAR file format +Will you have time to work on this before the release or is it okay to postpone it to +i compile jetspeed with eclipse i attach tomcats log but i dont know where is log of also attach +Yes basically is quite buggy +Personally Id feel more comfortable knowing that we can look at the test results for all of the suitescoreWeb DRTcoreWeb BVT Struts coreWeb BVT Struts urlTemplatesjsfWeband will get this fixed before we start a vote +In this patch Im going to remove readercycle script +CompleteBen +If the ER release is not valid paul or rado should adjust accordingly +The solution to this issue is to include the same changes that were made for the release +This is essentially a duplicate of which has a policy written in a different manner +Heres the s the s a new patch since there was a bug in previous generated by Subclipse in Eclipse +Thanks Martin! +Looks like this is happening +just saw it was fixed while i created the issue +This also means is downgraded +work delivered +no had been migrated to use since re using removed the slfj dependency already at least on the metamodel branch +Rado please resolve this if the patch ended up in the version used in +Thanks +Why would you want to do that? a shell script that parses the output of the dfs ls should work v well +we should first declare a interface for +A custom download page is also an optionIncluded is a small patch to handle including relative files with newer Ruby versions +Also I have the access to the Ipv machines for a limited time +Another server widget issue from Windows pushed compositethumbnail +Fixed in trunk too. +I do not see any pointers to any particular version in this issue. +Could it be that these are duplicates? +Xavier Im reassigning to you so you can review this changeMaarten +If the above correctsensible Id say Im for nd option +Can you give more details pleaseA demo application or a failing test case would be nice +As you can see the has inside it +Im wondering if we want to keep the parsed cv floating around in the object so that way constructing a Key and then getting the visibility doesnt require +Much appreciated. +Svetlana please check that it resolves the issue. +my cents +Thanks Mike! +Thanks guys! This fix worked for heavily on my Spock tests should I discontinue using daily builds until the fix works its way in +Not sure why I reopened it will close it now +Can you elaborate on how you replace the shapefile? Did you overwrite only the +Committed rest of the first patch Adding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilTransmitting file dataCommitted revision +I have gone for unrolling a two iteration loop of the sample and checking the output +Sorry it took a awhile +Reviewed the patch looks good +Committed to branch at subversion revision Touches the following filesM +This should be fixed in SVN rev +Back ported to r +This is a patch against +New patch attached with test fixes from further requirements? If not can this be checked in +A fix is in SVN +I need this fix badly +The pull request is not ready for merging since tests does not pass on buildhiveJuraj could you please take a look into failing tests +in Migration guide remade highlighting for word and phrases and corrected some grammar guide +Ill take care of this issue. +Committed to trunk + pinged about a new! Thanks! +done +As this is now strictly a SOA problem Ive moved it to JBTIS +This is expected behavior +sounds like another good test case add a few thousand tokens to the stream and see what happens +david i remember that you worked on this +we should add a jsf facet then Do I need to create a new jira for this issue? btw +to the target you mean you got it to work or you were just revising your instructions in the description? If its possible with it should be possible with +Done +I also did not code that page looked into it but could not easily guess where the problem was +I have applied it to the mbook daytrader m build branch +Thanks! +Remove breakBefore from +Because of a error in my test cases they behave differently in java and Java +If yes we can always revert to the previous versionClosing this out for nowregardsMartin +Can a committer please take a look at this patch and provide feedback +defer all issues to +I did not fix the threadIll fix this soon so it is in the next releaseAs a workaround is it possible for you to use the instead of +looks good Thanks Kihwal +Thanks Sameera for the contribution. +Please add the definitions to the book in a new one +Thats true we can implement several different approaches and decide which one to use based on the current application needs +Verified on EAP ER +I will commit this patch +should not be using private classes of other +Sample code to reproduce deadlock in Cocoon with two for the hint Vadim +Based on released. +r committed the fix +Unscheduling. +The root cause I guess is geronimo propertyeditor package couldnt be saved +Better to get it done with now when people are having to go to the work of making their code work +I am not sure if that will preserve history or not +Only issues that are actually fixed should have a fix release. +for dfs and for mapred +I looked through test source and I think it is bug in test +Changes have appeared in Latest Alpha Manuals. +Could you post itThanks +Slava this looks great +Moving to CR +Verifien in CR release notes +I would keep red as the color for things that are wrong or have errors yellow for information and notes green for new thingsOf course I have been looking at the grey for several years and gotten used to it +Which branch are you on? +Duplicate ofShort version delete the bad sstable and repair +The patch makes sense to me +We should discuss this first +HiI am going to start looking into this +Or is fine +Thu May CEST deDE + +So I dont know if my patch is a good will attach the patch for and for the trunk later is the patch for the trunk +Committed to trunk and. +It will be about sizable diffs +I blew away +Yes this is one awesome test caseThanks +Hopefully you can contribute back some of the work from your forkbq +I plan on doing this on both branches because I think it would be both unlikely and a bad idea for someone to not have a leading slash on their DIH handler and its trivial for them to fix and if they insist on not fixing it then this piece of the admin UI wont work which is hardly the end of the world its just an html pageThe rest of your comment pertains to which as you surmised is what Ill be doing next +I think it would be nice to have such a feature in +Thanks Alan for reviewWill commit once is committed and there is a maven artifacts to build against. +Docs look good +In any case not enough to pursue +I have the same problem on Win XP with both and trunk +I dont see the abstract implementation in the second patch +Probably fixed already +But on platforms with reliable df output I think we should take advantage of it as a means by which to catch a rare but unpleasant class of failures. +Media API changes are prob not going to happen until after the API fixversion and tagging audit +Unsetting Patch Available since the first patch was committed +Anyway I put a trunk release live today and the CPU usage has been at stable historical levels since I upgraded so this looks resolved +the types of thigns this issue points to are taken care of in +How about storing this information in zookeeper +Further discussion on this subject +This will be resolved in the current work for the parent task which is currently being done +Thanks for your checking Bryan. +Thanks a lot Arun! I am attaching a revised patch for trunk +Punting issues from RC to final +Tested with know locale chains zhCN and enUS +I mentioned HTML because thats the format Im used to seeing. +noted +to book on how to decommission a node and how you could do a rolling restart using this script +the real issue is the problem is people that use it not by itself +I dont believe we should start messing about with encoding options at this stage +tested +It would rather wait for the long error recovery for example than failing the datanode earlyWe all know is different in this regard since its a serving system +I wasnt able to get any more information out of it on first run +Attached example of accessing weblogic in a +We can get rid of this method if we ever switch to a proper inheritance structure or use more generated code +No unless you deploy the war unpacked. +Please debug your maven dependency resolution by running your build with mvn X +However it does raise the question of whether the big overheads of distributing buy you muchFor example using a job to count the number of records in the input is a load of overhead for that task even if you have done it well with a combinerAnd the second job uses reducer anyway the meat of the computation is not distributedI can surely imagine this computation existing as part of another computation but I wonder if you wouldnt already have info at hand like the number of items? does this actually form a reusable component +The second is to add support for multiple Holder parameters and return types neither of which was present in +Im on itIm a little unclear as to the context of the initial Hudson comment the display method is expecting D vectors but getting D ones +In the Groovy editor it only shows the action without kyebinding +If it is not too much of a trouble could you please enlighten me on how to correctly build the project using maven else I will wait for the release of +Just accidentally left from a real pageTomas yes I started a somewhat different discussion but I believe your problem is caused by characters replaced with HTML entities due to charset being UTF not +Cant we do the same for coverages +The now interrogates Jars and indexes their contents properly so as to treat their contents as though they were individual classes located in the classpath +I think it makes sense to have test cases to check whether leak +Subbu Could you please elaborate more in JIRAIRC about the maturity of your patchWe would like to commit first since it is mature and iteratesolidify a little bit before committing +Looks much better thanks. +Test code mentioned in the original bug report +accepted to change from Map Graphics to Map have updated the and to present the extension point to users as Decoration +Please ask next time first on the users list if you have a problem. +See details in how Id like to implement this feature +Patch committed to trunk +Which personally i am fine with i think it is a reasonable thing to ask of users but thought i would mention it since it will be a change in behaviour +I am using sun jdk and tested those programs using sun jre +Please verify +Also the fact that it happens only with allblocks is rather strange +Thanks Yongqiang +The latest version of the jbossas now replaced in the errata properly requires validator are not in the yet. +If we timeout trying to commit round we should throw a timeout exception not proceed to fetching the rowsnit Maybe we could rewrite You must use CAS for serializable writes You must use conditional updates for serializable writes +Hi Jacques could you try to reproduce this again at some point? I recently committed a fix for a very similar search issueI think this one may need a different fix but it is worth trying first +As long as the full client jar is in the classpath then there shouldnt be a problem as the service provider mechanism should trigger the registration +No worries +If it makes stuff easier lets just switch to the model +Should we be talking about creating more isolation for these classesAs for Xalan using manifest entries perhaps its time to start building our own packaging of Xalan an Xerces +I would have prefered that the server be able to handle plain normal DN and normalize it itself +It was me who opened and mentioned my workaround +I actually deleted +Im having this problem too when making a WAR file +Patch fixing the start with the patchIm looking to expand it to support on component event methods as well +The bean arg isnt in that set +I dont see a human task war included in the download. + Bug has been marked as a duplicate of this bug +We usually post patches for different version in a singe issue +Its humble pie for me +add some additional CAS creation tests for an analysis engine +Additionally the Testcases have been improved +The uses a thread local to store the locale and the approach specified here wont remove it when the request completes +If we make the cache field static in this should be fine +Ive applied it at the revision +updated to split into separate issues +Hi Howardok no problem +I will rework this patch but first i need to work on pluggable process you have a jira for the pluggable process tree? It would sound like a dup ofCan you please wait until it gets merged back into mainline from? And then make the changes you still think are necessary after that refactoring +Please reopen if relevant to new focus on Switchyard although I believe this is specific to jbossesb +In they now lookup the dependencies in the maven repo as well +Yes the disableTable is not idempotent operations because the subsequent retry fails while we cant remove the enable state check because we use it to sync operations like one is trying to do schema changes and the other is trying to delete the same table +This does very rudimentary string operation which is sufficient for the purposeMaybe your message should read something like dont add a suffix +Super Olga +Arrghhh +please resubmit with the patch redo in side s eclipses helpful create patch t use p but instead use p or p to strip my leading path I can resubmit but I will need to use cygwin to generate the patch instead and it will be a couple of days before I can do so +bulk close of resolved issues for release. +Only one switchyard consumer may be defined per implementation and it have to contain proper service name +Could you take a look and post a patch forSure +neither for Resource nor for Autowired or +Never mind a user just reported the same on users and I confirmed its not working +fixed has been applied committed to HEAD and deployed as a snapshot +A stop was functionally equivilent to undeploy but he refused to believe it +I am also seeing this issue in +The Client class is only single threaded in but this will be fixed in +Anjin debian package is not maintained by us please report the issue on the google code site +After looking closely at the AS code even with the extensions this does not help Teiid to provide custom appenders + +Maybe Acegis usage of needs to be revisited +We are moving to. +Few minor point against metrics reporting and UI pages updates with different version of yarn daemon should also be included +What version was this reported againstThe current code does not block for unreachable nodes when checking schema agreementbq +Waiting for fix to finalize my patch +Galders attached test is on the branch as +BTW I think this change should be OK +I think we should address this for the new componentNetty is clearly taking over mina as the better asynchronous communication framework +In adding tests I moved the tests to their own file +But Im not really proud to suggest this solution +just enough complexity so I can not get without meno nightmares though +Add the command count in the shell +Great +TCK results looks good with this change so this new improvement could be closed now. +How about To avoid data corruption File Channel stops accepting takeput requests when free space drops below this value +The composite xhtml element. +Is there a better way to override the status code than by overriding the processValidate method +Lukas please assess and schedule as you find appropriate +Good news +Under that page is better documentation on our mailing lists how to subscribe and unsubscribe +Also since it looks like only one of the booleans can meaningfully be true Id rather see a single private db field that will be set to one of several symbolic constant values +This is fixed now +Ah yep thats a pretty big bug +Optional values have setters +We need to see the entries in upload table and correspondingly populate the templatevolume store ref table +OK thanks for the info. +xcoulon i assume this is the NPE you were talking about yesterday +Thank you Brandon. +The same on IE no problems +did an integration test with my app and fixed further more fix +Using an API which expected localized format string but we were not using a localized string +bulk defer of open issues to +Must be related to some of the new organize import support +When you implement this feature you can generate the SMTP message ID in the way you need itWDYT +If youve got other testing tools it might be good to have a single or something which people could use as a single source for these sorts of things +Fix to M +bulk close of all resolved issues. +The documentation suggests to use Springs Timed s also one important change in semantics here Beforesafters are now part of the timeout whereas previously only the actual test method execution counted for the timeout +DA only contains Message defintions Will copy definitionfiles from DB +Still in need of startup +Blanket disabling the log mover feature would affect the existing users would recommend against that +Okay as long as there is no need to regenerate in the future I will close this + is the input is the fixed +isnt a CXF issue anyway so close it +We dont have so far all the relevant information associated with the project in order to perform a clean and easy fill that gap +bq +In the meeting on November ninth people agreed on this HTTP File Server service should be implemented as a framework on top of Mesos which is the third design option in the design document +Can we close these therfore? is no longer an issue afaik +Shouldnt the server editor for deployment location state this important and crucial piece of information +It worked if you click the combo box change from to none you need to on the Inbound interfaces table then these column will be all none +I believe the original code tried to mimic libc behavior returning bytes at EOF +for this one +QE can you please confirm this with a metamer examplePlease also confirm that this is an IE only problem determine what versions of IE are affected and see what effect the compatibility mode has +I also ran into this problem +The problem started when I ran the test phase for a maven project +I created for the build problem +I agree with Gunther +Integrated to and trunkThanks for the reviews Stack and Matteo. +Yes i am sure the deployment was quite ok and was running fine +Im having the exact same problem +Attached updated patch +retrieve the real port from Agent via +Alternate fixes for this issue see my earlier comments for ve committed a possible fix to SVN trunkCould you give it a try? +Resolved in trunk rev +resending patch with correct VERSION now +BTW the patch has been created against R +Freddy could you please suggest description for this rule +Then any time we are checkin if a configuration exists we first check the main geronimo +Perhaps not using an annotation can be considered +Better update directly to however r can still no be reverted +From today nothing special going on on node from the looks of it +mrepositoryorgjtestr with file downloaded from jruby homepage and now everything seems to work just fine. +In RT the searchers are cheap +Minor nit renaming bigTerm to is a bit misleading since its not UTF at all +I think that at this point test cases are more important than samples. +Alexey thank youVerified +on the C client we dont throw exceptions obviously +Sure I will try to create an example project +no option added but do not start a vm in development option is needed since dependencies to maven projects are not correctly resolved when avoiding own vm +Got the same thing on in production +I know about custom date just try to set another locale on your system and you will get exception with date +This regression was introduced by the initial port of Windows functionality into. +FolksI tried to put more details today for different approaches in the mail Ron has initiated Sub Tuscany DAS M Exception with default use truePlease take a look and give your opinionRegards sounds good to me +I believe it is better to modify those classes and add an to get the input and output directories as parameters instead of the current hardcoded directory names +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. + Chanages the generateCSR method to use subject DER sequence from the certificate instead of subjectDN +Ive applied the patch + release bulk close +Defer all open issues to +Approved for +We need to fix this soon +Removed the link as it is not relevant this is a generic issue mapping onto any adapter +Dupe of closing +Please let me know if you have problem applying the patch +Great thanks +I will take care of this issue as soon as Im back from honey everyoneFinally my link to the outside world has been restored and Im back and ready for actionIs there anything I can do to help with this issue? It would be really good to finally get asyncweb to its new homeo +Hi JamesThe code change is not for JDK if we use other JDK version that include different Collection util implement we also could see this test case errorThanks +Good +Then if didnt send the BBW report to NN it will never become out of safemode because NN will expect some threshold of blocks to report to come out of condition also added in tests +Hi Scottis there something I can do to help finalizing thisThank youApart Scott did someone tried the injection stuff? +I guess it amounts to preferences +I have not found the actual part of any spec that explains this but a number of browser discussions interpret the spec as stating that sending credentials with an OPTIONS request is a security threat and they will not do it +Many thanks for posting this Chris! for the patch +If that appears to be the case to you all as well then please overlook that +I do not initialize all swimlanes upfront +moving to M +In addition the source code to rampart is also not found in the source distribution +Do you plan to make support for local params for excluding some tagged filters? See Psaltis I would look at and the field and implement something similar to those +it might affect the supported version of Tomcat though +Thanks on the because everythings look better with a system test adding +Uses which to check if lsof command is present +We decided to use the assembly plugin to change the root directory when the zip is built instead of modifying the path in the build directory so that this change is no longer necessary. +fixed in Revision Hans Bakker +Now Im wondering how the old patch passed? Strange +Using your project I narrowed the problem down more specifically +Only the value configured in the configuration file will be taken into account. + +This has been committed in rIm going to close this ticket but if you can provide a test in test for excreatenode that would be great especially for the following situation If you repeat the request with the same client token but change another request parameter Amazon EC returns an errorWhen testing with the live API though I did not seem to receive this error +Tentatively allocating this to time actually change the fix versionSorry for the noise. +Regardless youll need to attach an example that reproduces the Ive been using this for rendering the whole page into a string and stuff into memcached +Thanks +In particular for the launched debug script processes we will have to do the memory management that we do for the regular java tasks +Your analysis makes a lot of sense to me +It has simple modules +This period saw a great deal of commits +This is the current head +Patches committed as revision in trunk revision in branch +The user should never have to muck with the standalone file +Consider using to transform to XMLIt has an ommitField method for methods to skip +I have split the previous patch in two new ones +Is that you agreeing it is working now +Add a nd one instead that outputs information in the format you want +This should get you started +This patch replaces a set of native method stubs for in DRLVM with real implementation +This gave the impression that the field was focused when really it was not and the carat had stopped blinking because it was now an image +Old test suite shows no failures against mysql +Why is this critical to and what does it have to do with compatibility +Line of reads too many bytes actline is sized to fit a zero pixel followed by the row data whereas an extra bpp bytes are read into it +If we changed the name to addCounter then it should also return void to be consistent +Re The reason we need to load first from cache loader is generally to be able to return previous value particularly in cold caches +Adding patch +I confirm that create tool tries to add feature twice also for Oracle layer +gcc is complaining because signed integer overflow is undefined and it chooses to assume that it cant happen +Please check that the patch was applied as you expectedBTW is it possble to make the build for TPTP ANT friendlyThanksStepan. +or import available in any maven repo anywhere? I havent been able to find it +I patched around this by removing the and returning immediately +Let me know if you have any problems using it +I will have a look tomorrow +Considering this plus the comments Doug made about the distributed fetching Im closing this issue +I had this as a warning but due to the sheer quantity of entities in the jUnit bucket that use default access the added logging was slowing down the test bucket and bloating the logs +I dont see anything in the diff for +Both messages are delivered +Upon receiving an open request the name node searches the datanode map to find the descriptor of the data node that runs on the client machine +The other one I looked at was Pipelined Partitioned Table Functions in Oracle they are bit different the partitioning information is part of the Function definition +Will need to fix some stuff first then I will commit it +I think the option of letting the user specify a suffix in the ini file is still good and supports a use case that occurs a lot for example when some governing body is responsible for the provenance of work that is collaborative and distributed and is responsible for assigning unique suffixes to each instanceIm also not convinced that using the MAC address doesnt have problems so Id see adding that as as additional approach + +Back to is that at test execution time tycho pick ups different plugins from one of many nightlystablecompositewhatever p update sites we have +committed +So wed need to a method to execute a CQL query that may have answers +Its seemingly used inside typedbytes of streaming today +Let me try running on BK code +None of these properties are used for anything +Yeah i dont think we need to solve it before committingI do think maybe this class needs some more warnings to me it seems it will use crazy amounts of RAMI also am not sure I like the name Direct +The issue can be reproduced only in project examples of the archetype type +Thank you Don +Thats the first one since +In most of the cases should be better +attaching new jenkin arun +Doing a clean jar the is no longer in thereIt was the that I did not want +I linked and reopened the relevant issues with this commentReopen for backport consideration +Hmm I will look at all these to pass on will continue to look at the patch +You will have to attach your project for me to see if there is actually an issue +This should load both fields number and number +Very nice +As spec lead Red Hat is rejecting this TCK challenge +fixed affectsfix sys admin changed the network config wo notifying +My new seems to be a duplicate of this +The modified patch has been applied +Sorry for not mentioningBTW the patch is still very rough but it shows the concept of converting the vfs URLS to real locations +Would very much like to get it committed to trunk as well but Id rather wait until someone gives it a run first +Fixed in r +I ran into the same problem myself after a few hours I was able to patch it up +I think this is due to the socket being killed but the server side not noticing +bq +Not if a workflow requires these measurements for automated decisionsWhile a human can read the job history data from JSP pages theres no current means for the app code which calls to obtain that data and use it to alter the workflow +Can we make this just Note the results pane doesnt look as smooth and elegant as Romains one +Thats why we may not be talking about the same thing +contains the auto generated stubs and the client codeThe issue happens when the service class and the bean classes are defined under two different packages +The problem I had trying to create a test is that I couldnt figure out how to create a test that didnt have absolute paths for the dir in question and I wasnt able to create a failure with abs pathsLooking forward to someone showing me how to make that happen +Patch for trunk +I will commit new patchGenerally we will not keep in patches +The tests fail with the same OOM even without this patch so I am going to reapply +Unless you see something fail locally without your changes I wouldnt assume a test fail is unrelated +Assigning to Sneja to verify once its spun in Jenkins +Two things here The first message you are seeing is because has been compiledwith JDK and you are using an earlier version +Not sure if that will break something else +Regards +Fixed in will be released within the next or days. + +Attaching a new patch with the changes +Applied the patch. +It might not satisfy all possible needs but it definitively is useful +All goes well except +So if you can throw some pointers on how to set up a dev env i ll be able to a patchI was able to start James with an updated JAR send a Message flag it with USER flag and than use to filter it out based on the USER flag set +Chris okey that looks very parseable +The root cause is multiple Sinks was trying to create duplicated file +Can you please explane a part about exceptions what should be done +This is the symtoms of the release as experienced by the showcase application I presumed that this application was as the codebehind plugin was deprecated anyways the patch I have developed I think is unsuitable for a trunk commit as it presumes that you are utilizing the classloader of the servlet context and as this class is a utility class this would fix be specific to this issue and not correct for this classes use genericallyI can extract the fix to a more appropriate implementation for my specific issue to a separate class loader forced implementation for this case however if this is codebehind specific then as it has been deprecated my efforts would be in vein I shall take a look at my next available opportunity and verify the issue in greater detailI shall follow with more detail is deprecated but it is still there so it would be nice to fix this problem and yeah we need to update Showcase in the spirit of true open source documentation is the last thing done +I just committed this. +This is an FAQ use the forum for usage questions +If you look at rewriting this is done based on the IS anyway for a query so we only rewrite once +but could not easily find where to set that in eclipse +only if I have not clicked by mouse in vpe +This link doesnt appear to be there anymore. +For the long list I think we should just patch the config parser to do the multiple line thing that most INI style configs have which will alleviate the concerns for lengthIve never seen the list of types use media type parameters so I dont think thats a problem for the INI eitherI dont think the stuff is a blocker for this patch +I will try to implement that and report what I found +xhtml +The API only allows for adding Anchors or parameters +I think we will need torun this by the XSL working group for clarification and then we need to fix FOPaccordingly +Management server log file +Let me track down the jira or get one created +Juergen what version of mysql and JDBC driver are you using +Will commit post testing. +interesting that it still works with different +is the layer on top to manage threads for performance and priority of messagesThis is likely to require you to write a custom version of from which presently built assuming one thread of execution +the raw plugin is not what we recommend using anymore +What about random files being left around? Was the file place somewhere that MR will eventually clean up if the process that created it does not +Granted he could split this into the same four chunks as represented by and +I think this is very interesting for people using Eclipse who want to start some hackingPerhaps we could add this profile and the one for to the git repository +applied thanks. +fixed in rev +myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadutilmyfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadutilsvn content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadutil Completed At revision +This patch changes the exception to one that reports the id +It might be possible to do this without changing by abusing the TYPE system +And this is backward incompatible change. +I can see the dropped regions stuck in the regions map +A beta version of JDK was being used +The parser has been revamped in the upcoming version could you please try again with the instead and let us know if the issue still exists? +We didnt see much change I think applying debug is going to be required +closing to avoid multiple threads for this issue +the text should read relative to conf directory also there is no default SSL configuration +Need to add LOCALDIRS as env for +Fixed in rev because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +In my perception it might make it less probable but it still doesnt get rid of the problem +adds described option +The group field has around unique values +I cuted the text better instead +It looks very interesting but its better to provide a svn diff +Changed Target Version to upon release of +Rich would you mind combining few comments together so that we get less emails? I currently get one email for every comment you make +Unchecking HVF +See +thanks for reporting this issue and providing a test case and patch! +does not throw an exception in this case +Could anybody please tell me when this bug is going to be fixed? I am using CR it is still there +No feedback from OP so assume all is good. +This also means that your desire will fail completely if the exception is raised due to a header +I used the profiler that comes with Java +I dont grok this At first I thought its nice to use params instead of fields just to keep everything in as narrow scope as who is going to call this methodFrom Java code I still need to pass an argument +That stuff about P and NP is just kids stuff compared +antonyCan you explain a bit more what you mean by you couldnt process further? What were you trying to do and what error did you get +on the other hand only applies to a start signal propagated by the and doesnt claim control of a start signal to an individual component in the first agree +The Mahout code base is already uneven and sprawling and I think this would exacerbate that and not generate much synergy worth the effort of integration +Since this is minor well move this to EWS +I added one more test case which reruns other testcases with service port on +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +What Eclipse version are you using +xml files but not the persistence files when this commit was made +I dont know if there are environmental differences but it fails in my unit tests and whenever I try to run even simple flows as the enclosed from the web console +It looks like the does not understand the RTS and needs to add the bridge codes +as well we found in +So I stole much of the code of the regarding descriptor verificationAlso removing the node type stuff from the would remove the dependency on the package which may or may not be available in certain use cases +No unused private methods or variables could be found. +is this only affecting ESB examples +r +This issue requires to simulate lost task trackers +I believe that extra tmpdir is from a time when the command was cd s git checkout s +I think it would but thats only part of the solution +Add it to the list of things to do +Hi AntonioNoted the task +Not like the way it is implemented +In the meantime any analysis is done on the old corpus and no conflict occurs +Has taken a while but now have reproduced the fault in a test attached zip file has tests +Namit yes I can refresh the patch in next few days. +Thanks Andrea Committed revision +It is implemented now but we are still missing the documentation update I will start working on that soon. +And successfully have broken Linux as reported on. +I agree its nicer in a variable but it makes the diff more agree that the fix in Arquillian is preferable +The old WCS module will contain only the common parts and utilities shared between WCS and WCS modules but this is a thing we can discuss further if you think that there is no reason to maintian another WCS module on the source +Mukund can you try this patch out? I cannot get hte tests to fail on my machine +Fixed adapted the existing contentEquals method to create new content +Modified to exec the du shell commandMaybe can just omit the getDU call but as an optimizatione it would be better if we only try to delete cached files when needed and can save a lot time if delete more files at a time +So i would stay with the expression would be more consistent +Test case for +I cannot seem to close the issue +with ASF will try to update the validator today +Sorry forgot to include them +Attached a patch using the implementation in +I will apply your patch without the transacted option and open a new JIRA that camel sql should also support the exchange in body for the sql statements +Agree ClausLet me check on my side if we can release Karaf +The input paths are all directories and there is a single glob or regex filter on filenames +I would much more prefer to see each option that make sense exposed as an extra option that we can verify for correctness prior passing to the underlying create external table statement +Ill try to fix this on Saturday +However if you execute mvn clean install Pjsfri from the top level module the example apps still get packaged with the API and IMPL jar files +Ive added you as a watcher hopefully youll get to know your VM for stats has been completedTony should have sent you access details please ping us again if any problems getting inThanksGav +There is precedent for this throughout Spring +Peter Butkovic Do you have by any chance a patched maven webstart plugin ready for use? Thx for the will try to apply your patch next week and then do a release sound good to you +I ve worked on this a lot of time ago and I dont remember. +You mean numerous test failures from QAEnvironment issue. +Also remapping the term bytes in general on the fly is tricky since the remapping could alter their sort order +Thanks +Attachments latest patch here patch for output of the failed testsYan looks like lzo related errors are fixed with the combined patch +We take new writes at offset T T Tc +Missed two typos +released to Codehaus Maven repo. +Closing +This should fix your I will also raise a bug on grails so that will create the right kind of +The fix attached +Try to reproduce and ping Andy as needed +This request is integrated in and becomes obsolete by issue because it is obsolete and covered by +Thanks Dan attaching v patch +h files were copied to from drlvm sourcesAll tools are built and tests run +New patch attached +Along similar lines as konstantin I propose Currently the LV of trunk is +Emanuel could you have a look and see if this is still valid? +This all seemsIf Hadoops general scheme for file layout is causing problems and needs to be simplified lets discuss that +Uploading a new patch that only affects xaI agree with Mike that a new issue should be filed +I got it to fail after looping times +of course no job is expected to occur if you choose not to report +I add a distribution providing the war +Thanks Amareshwari! +In this should be done with coprocessors right? +Running the quickstart with the last source from branch works fine +Never mind found the setting its provided youve got pax web installed. +Implemented built as a small separate jar for now +Strange thing +Thanks! MichaelThere has been another more general revision of the resource synchronization code in the meantime +It is harmless but annoying because once jPDL Graphical editor is opened on any file in ear project it starts logging errors to Error View during build of projects included in seam project + +Resolved in with a new. +Committed at subversion revision +for the patch to work correctly it should be applied at strutsstrutscoresrcmainjavaorg +Thanks for the patch. +This used to work before but I think it got broken when you made it so the formatindent actions can now apply to portions of a file +Changed from Resolved to Closed since we do not use the Resolved issue status +This issue might be related to our usage of pre versions of for discussion on this +Site will be updated in a few hours +Im not quite sure how it is expected to recover from that level of. +Ive just committed this. +Overall the approach looks fine +I completely agree I spent hours searching for a way to externalize build properties only to find that Maven knowingly restricts this +We now have a strict date parser available in +The current service in place now is for administration namely execute repodb scan enabledisable consumers delete artifact query managed remote repositories query repo database consumers +I see the same tests failing for other patch submissions +Please add tests especially tests showing proper handling of associations wrt ordering +java file on the CLASSPATH but I feel such features should be confined to IDL and that JSON should be primarily used for schemas +Hi Antti I hope you dont mind me closing this ticket given the initial problem with petclinic failing is not the real issue here but rather addon upgrading +Patch looks good V +Your test passes for you Daniel? Its hanging for me +If you add jug and xtream to the plugin classpath you will get conflicts +I finnaly did it! With eclipse as you told me and did not take new files into I think you will have to change to run it in because we use an oracle database +Server runtimes are registered correctly only Server view does not show all installed server instances + for the patch +Thank you +FWIW I also pulled the change into our stdlib fork so it wont go away +Assume already fixed +Added a test case to double check. +Perhaps this should be marked as crash leaving that call to reporterSetting normal urgency +I dont think we can do anything about it +Why is defined in +Woops bad Eclipse! Fixed to junit Assert class +Looks like this makes testing new ops also simpleminor the last entry in the patch is a only diff +Could be fixed at the same time +I want to experiment using the deployment this test used +Are these filed required for compiling RAID or other contrib projects +Patch for +Jaroslav Ive never been able to get this error to reproduce locally on Win +either by a backup node +The ability to manipulate where the COUNT and NAVIGATION of the paging banner are displayed is something that is very important to us +Logged In YES useridRead the context first. +Its always important to have these written down so they are not forgotten during see for the next step. +Despite the error in logcat the image still appears when I modify the test case to use the DATAURL +The attached patch illustrates the issue with additional checks in an existing test +in release +I think we are in pretty good shape now so this generally only leaves adding support for optionally required bundles. +Fixed in CR not GA. +The second is not valid xml so no surprises its not first is valid not sure why its not working as expected thought +reevaluate the problem is most likely caused by users actually doing reverse engineering from inside eclipse where reverse engineering will actually the equalshashcode meta attributes for certain partswe should either a do the injecting in the pojoexporter automatically b make hbmhbmxml generate the meta attributes +Please close this issue +Good can you briefly tell what is left? I see that has an commented assertion and code still have some +Well well +Please keep us posted +meclipse will automatically handle this for you +I will commit the patch shortly +Modified resource lookup to use helper class +Closing issue. + +interesting +Of course those are two parts of the same schema there is no input xml is gone +This patch alters the response only if is instead returning an array with key collations +In regular runs I wouldnt expect it to time out but on a very slow machine it could happen + +Moving to movement here +Im not sure what username and password you need Stein +The changes for the close procedure in affect this patch +Leonardo this sounds even better to me +done +Thanks Rick +latest JPOX SNAPSHOT resolves this issue. +Affects Version must be not +I think we solved everything that was required by this can mark it Out Of Date. +Don what is missing to get this done? This would be very nice to have for seems that the only missing piece is the ability to configure which actions to allow in +closing as cannot reproduce please reopen if you disagree. +Merged a patch which unifies the arguments into trunk. +Add Win version data +I think that the JDBC escapes are a bit obscure users looking for timestamp arithmetic may not think of looking in the JDBC escape sectionThe JDBC escape section should have at least a reference to TIMESTAMPADD and TIMESTAMPDIFF +Rob can you make the Push dialog to be modal? At the first moment I didnt noticed that there is some dialog behind the Publish to Openshift Server dialog and I thought that the process is stucked +It doesnt throw exceptions +Cause +For Seam Tooling it works fine this way +For forget password functionality I add your address as its +understand is hard to unit test but is full of static methods that should be much easier to unit test +Ive got passed this by having the Woden top level declare the Apache parent pom as its parent pom +Class cannot be casted to arrayAny thoughts why this is happening? The file has no overides in it except to filter some tablesThanksTony +Just verified your code and it is get service instead of set +So Im happy if stargate would remain a contrib for a longer while until I manage to get the dependencies into Debian + +Thanks. +For this one needs ahandle to the to be created +Too old an issue most probably fixed when moved from the monolithic configuration to the one +Patching Struts +The tests are not failing anymore +The reason being that i could not go to one class for the logic of encoding an sql statements +Can I take this? I could use this for the work Im doing with at the moment +I agree with Kathey that this error may be fallout from an ungraceful recovery after the error seen in +refresh the the tests are passing we are blocked on the names of the new reserved words we have introducedWe are trying to get it in asap +The tests case locates another bug which is somewhere in classloader IMHO +Good finding Joep +Those values are used as default values for all the properties +this is not critical and may be too late to do the change +Id be okay with renaming optional though +Hi liang xie Patch looks good thanks. +yeah this issue is slightly different from the earlier one +Im closing the issue as fixed +It would be OK if the message is only logged when the core is first loaded not on RELOAD +I will change the backwards tests possibly to use the new style relative svnexternals property to check them out with the rest of the source tree +no test cases needed since existing test cases cover the feature +Committed revision +Patch to fix Set bug +Because we need to reuse the code of which uses join tree and row resolver to generate the new map join operator +Some changes in and are repeated +Im ok with applying it if you make sure it applies cleanly against trunk +Which is required for +Still some work. +Attached file again with flag +Job has maps and reduces +Merged fix to +projectPMD +Thanks! +Is that the case +Fixing up a few small bugs and am going to test using a few different model types +Sorry for being delay +Committed as r +All I can find wrong is that the wrong tab name is used which means that the tab contents do not show up +I just committed this +Not sure why its not appearing in the commits tab +As shown in the debug and tcpdump a request is send to the serverThe tcpdump also shows an answer from the server which is subsequently ignored +And I dont think geir needs it. +The reason is unability to resolve class under cnditions +This was running on EC instances granted its not the best real world test but shows the margin is not very big +Grant have you already had a chance to verify whether this fix works for you? FYI the GA is now scheduled for Monday gave it a try and its working great +Thanks to Andy for his inputs for the prototypes and reviewsPls provide your suggestionfeedback +Closing +project file does have the javabuilder commands +First version of a patch with quite finished refactoring of handling +If two users submit compaction request with iterators this design could run compactions with both users iterators settings +Maybe we should use a data structure like? or something along those lines +Those members should really be private +EWSRCTC +In the older the configs for output format and output committer were separate but the committer is tied quite closely to the output format +In order to get hovers to work properly I had to do some fairly ugly hacking around JDT UI code +Here is the patch for this one +Ok I must admit I didnt try the examples in standalone mode for some time now +Appears to be working so far +I just tried wit and faced the same issue +added as an affected version +Best to create a test case and a patch against trunk if you want this included in a future release cant verify it since theres no test +Ok I try to create a patch but what is latest version? or +Thank you for your contribution! +In i added basic unrelated tests to a lot of thingsthe only reason i did this was to just try to improve the tests as a whole you know to hopefully prevent any changes in lucene contrib from seriously breaking something in Solr nothing elsesorry for introducing the confusionin general there are some minor things left to be done in and then there is +Refresh button in the app wizard is back and works refresh in Explorer works either +Also lets use the wrapper for all calls rather than +Can you please file a follow up ticket to fix it in a more saner way for long term? Tx +The real question is why using the mt compiler option doesnt define this +Further work can be continued in a new issue for post versions +I hope you will eventually consider integrating the change or at least let me know your objections if anyRegards sorry about the delay +Verified link is ok now. +In this patch Ive added a simple boolean to start the job only we have detected a valid segment +This works locally for me for both hadoop and hadoop to trunkThanks for the patch Jesse. +It should only close a after its no longer needed +So the idea here is that instead of setting globally for the entire job we need to be associating it only with the paths that are supposed to have index filtering applied +Instead of handling the shared child node entries in the itself now know about when the internal maps are sharedMost of the patch is simply renaming to so dont be shocked about the size of the patch The memory saving is now a bit lower but still significant +I think we need to provide the full URL to and not rely on a domain redirect which could be pulled by prod ops at any time +Changing to blocker so further investigation can be done +Patch committed with format change only +For instance Id like all of the rules in a decision table to belong to a specific +I will verify it +Okay I see your point +is for the same sort of issue +svn ci m Applying unit test and fix from +Just replace the jar with the oneSo something like this cp After that fire up the are the same in all releases +Fix already included in AS. +I think that scroll is excess and change this on dynamic panel is better and implement the drag and drop as alternative method of adding + for the first approach +Use version like +Does the annotation processor generate the meta classes in JDK or just fail after the warning? I am still at JDK may be time to upgrade +Agreed throwing an exception in the timeout case seems like a more natural solution than returning a special sentinel value +I use maven to open it using and debugI also use java jar distexample +Able to successfully install all of JBT from the site +This jira is in the wrong place +Thank you. +Ah +We should throw a warning when folks try to use it saying its deprecated and will eventually be removed +Jukka for this +Closing per comments of Denis and Max. +I also needed to remove the line for the unnecessary importimport as that class is not in Solr +And as an extension to that all table level properties will now also automagically appear as partition properties which doesnt feel right +Patch attachedAlso generated file attached +The merrits of and being interfaces not withstanding the purpose of is to provide a simpler API for people then they would have if they dealt with both an and +In all my installations it throws NPE in +committed your test case so we can protect your use case it into the future +bq +Tested in Continuum and it does not suggest a tag the other hand releaseprepare at the command line for the same project suggests +Font registration now works correctly. +just the last one as the previous one is already works on my Windows machine now +Please excuse the typo mistakes +chances are high that this issue will hang around for quite some time +However keep in mind that a bundle cache eliminates the possibility of having native libraries and will definitely consume a lot more memory +I just wanted to define the leased possible changes needed to fix the current issue firstBut Im all for doing it all at once +GUI handling of dependencies for selected runtime options and the ability to install a feature but not activate it at runtime effective use of +Issues resolved in +Heres the simple patchThis has only a global setting +Reponening the changes dont compile +I run into this error once or twice as well +r. +Dejan that looks a little like the symptom I got when the xml config watcher thread was over eager but the check should have sorted that +Startup scripts these could delegate to a developer maintained subscript that set variables to be included in D settingsGlobal properties yes but you can namespace them +I cant reproduce this and anyway very likely it was just a test framework trying to interrupt all leftover threads. +I tried to make the tests of from more systematic and checked they cover all exceptional conditions +Could you please try it again with the latest CVS version and if its still happening this issue +I will check the possibilities to solve this in. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Similarly the x bit for directories is not needed because directory lookups do not do access checks +im not sure is this bug related to RF or +Ive a look at the document and there is some text missing +If we can increase the performance of readString in thats great but we cant do it by violating the contract of the interfaces that are implemented +ok I like our detection better then +To declare that if close is not invoked by the code the result is not deterministic and current behaviour is +You may have to mod the mail through. +Previously this was an URL string +I have such code in Java which could be used +X no longer has a throws clause +We want ifs on new lines +Done and done. +Didnt account for the memory maps index cache and data now its that supposed to say MBB +Note that as the lock scope is tied to the current transaction scope if you mark the method that is acquiring the lock as not running in a transaction then the lock will only be held for the duration of the method invocation +Therefore part of this work will be also to move as many of the tests over to using Field and working on the Document API as well +Oooh sorry I didnt see it is same as this one but for EAP. +scripts libraries and classes for testing + +I agree with hc busy that in current state is broken +Before this patch it said for +Shouldnt we be targeting the fix of said bug against all supported versions as we go +Alex Thanks for your quick response OK I agree with you but I dont konw what type it should belong to could you help me change it? Acorrding to the simple theme I think the labelPosition should be or left what do you think Alex +It is the parent class of the new Mojo and the existing +Ill attach myresolution to the RFE again +Now I remember why I originally designed it that way +Bumping to for now. +Tricky one. +Thanks HiroMy bad. +I didnt get his Javadoc changes bc they were too hard to determine the differencesOther notes Removed and just used the for those tests +It should call the showcase action and display +Verified on +In my original project also cobertura reports have been created +yoniks suggestions Noble we had a implementation collisionIm doing some quick performance testing the version I wrote now +all bulk edited issues have in a move issues to and +Is large number times data per core going to be under MBIs large number supposed to be hundreds thousands tens of thousands hundreds of thousands millions? I mean if a web site had millions of users could they have one loadable core per user? The use case should be more specific about the goals +Craig the issue for injection is +Er make that they just changed the +Grant I undestand what you are trying todo I just question if what you are proposing is something that really belongs into the core of if it should be done in a pluggable codec +The RSS feed works exactly as you wished it should it shows diffs in case of old pages and the whole page in case of new pagesI dont know how you even got it to show full pages +equals when comparing Strings not +Andrew is working in the network area right now so hes better placed to commit this +Basically you can do Dtrace enable trace logging +I hope you understand that javaWSDL works of compiled code +Right Im looking at the vectorization branch +Thanks for fixing my breakage of buildName these something else on commitEMPTYRESULTPThisEMPTYRESULTPB +Attaching what has been committed +TomIt would be hard to neglect a such continued effort + this looks good +This should be submitted to the team +TRUNK only +its very likely a Core issue +Test passed +But i dont have a test case to verify it works +This one needs to be against the new Oracle data store shipping with GS thanks closing it closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +this is not decided yet so ive unscheduled this issue. +Also fixed some of the issues found during testing in old create and create servicesThanks to Arun Patidar for testing the patch thoroughly for different scenarios for Anonymous and Registered user +I suspect that theres something that LTF does that I dont need that explains why it is so complexI think its trying to support arbitrary lookahead and insertion of new tokens +Thanks Carl that sounds like a good ideaUnfortunately Thrift doesnt support method overloading so Ill still have to create new methods with new names to provide backwards compatibility +file nice +Verified alson on +Or else fix in +The same message object may be sent multipletimes +Ill do a separate tool called +Move the test from test to look fine to me +Thanks +Be default a user scan will also filter TTLVersions so its one thing to prevent the from being compacted away and another to actually make them visible to user scansA similar approach can be followed in pre as long as the coprocessor has enough access to internal region data structure to rebuild the default scanner +Travis currently I have it working on and ll try to checkout the version to make some additional test +In this case the protected ooxml files need special handling before they turn into normal ooxml files so I dont believe we should be treating them interchangeably +Im not able to reproduce the issue with those plugins you provided so I think it could be considered as resolved +I believe that the presence of in means that is still possibleI would recommend the following improvement to this patch The cloned code warns the user to propagate changes to as well +Can you try with this? Also what version of Java are you using? should be part of the Java runtime which is strange +The average patch sits on the queue for some time and many committers volunteer time ill review ASAP +The version needs to be validated before updating to it again +going to commit this in a bit then +Thanks for the patches IvanSetting Patch Available flag +My cursory search was apparently too cursory. +Issue confirmed with persistent store +Unfortunately not as easy as I would have hoped +all the terms are high frequency or whatever + +I agree +I am working on take based on alternative which I would likeyou to review once it is ready OlegI think I got the drift of your changes +I have successfully run using the Windows versions of Java and Java from Oracle +It was put on the +Tested it locally and also have added a for +At least some of these must be caused by the compiler bug described in +Could you append the corresponding paths to class path in bin? Also please remove corelib +We changed it because we didnt want to duplicate the schema in two places and also because we wanted to have it available by default in Solr installations +Cannot reproduce with JBDS neither with JBDS closing. +We at least get a decent default if the editor forgets +Patch created in from geronimo root directoryI verified that I could start all server configs access the console and deploy a web application with these changes +But as in is too verbose +Here is a patch which adds gzip supportIt was slightly awkward to do in a very clean way +Looked at the patch +Well yeah the whole point was to have the reflector provide sensible defaults but i see the point of it making the implementation pretty complicated +I have attached the new patch +If I run cqlsh cql everything shows upIt could be that the answer is rebuild your schema using CQL +Register a repository on a new serverin the Modeshape view +Ive attached an updated jaxb provider jar +slightly improved version of the patch that actually terminate the build +Will remove the server that had reported in as new ten seconds or so earlier +Turns out most code is already using this pattern +Node goes down stays down longer than hint window Node stops saving hints for Node marks Node as needs repair Node comes back online Node starts repair between Node and Node by forwarding the streaming repair task Node is now up to date and no longer needs repair Node and Node know this from tracking the repair task Node does not discover this is the case continues to see Node as needs repair state can only be updated if Node initiates a repairBecause of this I think that the hint state would need to be gossiped +Adds support for Parascale storage as the backing store for Hadoop instances +Hi PaulexThe fix looks good +like you have written yourself i need a testcase to reproduce the current information does not help to reproduce +Can we fix this by allowing default to be null? That seems logical because when a weblog uses a shared theme it uses the default page of the shared theme and theres no need to keep track of that default pages ID within the weblog +we are working on a rackaware ensemble placement which is similar as the ticket hereattach a patch of this work so we could get some agreement on this part +I fixed this case so that now the content doesnt get used as the label +Assume closed as resolved and released +Creating +Here comes my patch for this issue +Youd better post here the url of the new bundle +The exception appears originating from the nd Reply +Yes can confirm it is still an issue afaik the tests use Jetty transport bundle activator runs via the activators list and it calls but it does not help something does not wire based on the feedback from Tang patch looks fine and thanks sergeys offering. +Cut the trunk code over to a branch today next release will come from there next week +Patch looks fine +reopened for for as well +Fixed by fixing splatting +I am attaching an enhanced fileLinks in different positions with different targets some of them using filter scriptAll these links do work and submit data with FF I tried to use it with IE this time IE does not even make request after clinking on any of these linksIE shows error message about missing tooltip script even just after loading before any action from my sideI am still using snapshot downloaded last weekend +Suddenly the copyProperties method throws +Patch in cvs +I believe it is also patched from the root trunk directory +NET distribution contains VC runtime which needsI also dont think it is a good idea for us to bundle this runtime with binary distributions of +Uploaded a patch that does that +any reason why we dont include this in one of the existing many plugins in AS or at least give it a better name +Yes +Patch could not apply cleanly to file srcjvaorgapachehadoopdfs due to other changes in the file + +deleted erroneous cause identical solution different views to add on +The proposal so far is certainly simpler than the current system but the idea having a single name still bothers me +Descoping items not being worked on for M into Unknown Fix Version for robbie can you review this for me +Hi Jarek Do you mean following classes will be available to the wab bundle if the users only import the standard jsp package for the bundle? Unable to compile class for JSP cannot be resolved to a type cannot be resolved to a type cannot be resolved to a type cannot be resolved to a type cannot be resolved to a in a WAB should be working if the WAB is importing the +I am working on an open source specific remote repository is simply hosted on dont have the luxury at the moment to setup a publicly available java application server for the proxy +Check implemented in and all its descendant and. +So its about money thats why its not done +maybe +Youre welcome and thanks for the kudos! +We already are making some special cases for Tomcat classpath loading I can see making a case for as well. +We have become lax if a file is missing when we go to compact perhaps because region since we made the object changes +Changed to set working directory of the as childUGI +Uploading the very simple patch for this issue +Fixed at revision for further review +The tests are modified since they checked for the NEWLEADER packet to be duplicated +Anton Your simple test now passes on code in repository revision rPlease confirm that this can be closed now +Complete error is likely working as designed +Committed to trunk +attached is my proposal mentioned in the comments above + +So tasks would not block tasktracker and datanode daemons although it may not help on other tasks from a different job +We need to remove the provisional project only if we reach the last wizard page +note on the first file was meant to read the above file is empty +Playing with the menus to take the screen shot it has now created the RHDS group and has one set of menu items in the RHDS group and one set in the Other group +The dependencies should allow Seam to run +Perhaps talk to Gavin about +Im really talking about overloading elements not reallyextending them + I believe Jacopo talked about changing process that did not have user interaction to system permissions +Oh and btw youll need to install the minitest gem to run the provided spec +This will expose you to how our services are enableddisabled and some error reporting stuff +And more importantly this is a pretty serious bug and a release should include a fix at the very least for the logging issue which would at least provide a workaround of downgrading the log level for this interceptor +close this issue. +have someone suggestions +I tried copying my Java colour preferences too to no availLet me know if I can post more detailhelp in any other way +I dont agree +Im not able to reproduce the bug which describes Viacheslav in his comment on Linux +Until then we close +Please use that in the future + +Still cant start AS under Ubuntu +on JBT. +the current reasoning was like thisif people want to distinct between api libs and impl libs they would have to use the maven modules from our public the download contains a simple single jar file containing all people need to get started with it + +html extension the url to have a page from a live Confluence are without +Please confirm the current status of this issue and if its incorrect change it as required +q file should be converted to a sort merge join +passes +thanks for the patch +Committed to trunk. +Either the JNDI lookup is returning an Unshareable connection or setting the isolation level on Shareable connections is allowed +Closing the issue now that it has been resolved. +patch looks good +John Can you take a look? Thanks +We probably need two interfaces one for mutable collections and another one for immutable collections +If and when we actually have a problem with too many trivial being opened that it makes things hard to track then we should do something about it +Great +I agree there is no win here + +I applied your patch but also fixed the toString implementation in geoapi +Sure any time +Dont you think that the status should be Cannot reproduce instead of Fixed that needs a fix version? +Will address it when I add more test cases +My previous experience with JNI makes me cringe at this prospect +Thanks +I tried with normal SOAP messages and it was working fine +The code only looks at the classloader of the type +However Im almost done with porting Avro C to Miscrosoft Windows +Where can I find the references to this variable in the code so I can trace whats going on +So it is actually a duplicate of +I think I tweeted a bunch of profanity related to this. +Thanks again +I am assuming this is no longer a problem +It looks to me that we need to add validation when the user saves the subscription +Patch for adding a new exception type called + +The example shouldnt need to change +The metatype seems to be more of a converterdescription for complex types then a the for an the M refactors the compendium support and replaces osgi with standard Spring construct. +This is not a issue +Heres a patch which adds support for running the Hadoop client on AIX by conditionally loading the appropriate usersecurity classes when running on AIX and by conditionally setting up the JAAS configuration appropriately when using IBM JavaNo tests are included since an AIX environment would be required to do so properly +thanks greg. +attaching updated patch +Made for confusing loggingI have a new patch that Ive been testing +Also I will follow this up using the email address given by jean +I think its better we fix keeping online schema changes also +the correct value to use is webwork +No they dont solve this problem +has been logged to provide a publicly documented property for the database format version +Uploading to fix bugs in the first upload +Sorry to miss your latest comment +we discussed this +Hibernate Console Configuration classpath is fully configurable by default it use your project classpath you can adjust classpath so to avoid logger conflicts +Ok understood. +should be verified in JBDS work item for Daniel waiting to see if I can deploy +However I could not do this check because the underlying jettison JSON library is ignoring the duplicated keys +Nope Im not looking at this right now +Even now we should provide some upgrade support +Can you commit to too +Latest updated patch ready on RB +The difference is only for the rowKey that cant get the value from memstore it need do a more read from memstore compared to the trunk read only from store fileYou must know the origs high performance is just benefit by only read from the memstore +Closing now. +This is not likely to be a problem but if it becomes one we can have the method map entry point to the first line of the class instead +Fix works for me thanks! +In my testing the fix for issue appears to have resolved the in trunk and next release +Thanks kw +If a user could specify the to pass into this faceting issue would be resolved +Second test case +So we should confirm whether the number of arguments is rightAnd I didnt notice about a spelling mistake +This should also be fixed in this patch itself +Id suggest to call it tool for the sake of correct definitions +This issue would fix +I think this is a great incremental improvement and as Eli mentions we have a long way to go in improving performance here further +Proposed patch to resolve this Jira +Looks really good to me +Good idea I will soon code the new mojo and add the other boolean in existing mojo +Please define if this is closed or not or been redirected into a new bug and can be marked as solved. +And I beat you to the rename is opMetrics alright + failed twice as reported by Hadoop QA +Thanks for clarification +However it is impossible to derive much useful distribution information from the stats of a Trie fieldIt comes down to comprehension of what the schema browser displays and what a Trie field adds to the index + now starts life with a null startTime and it is updated by the when the execute method starts its work +This simply makes available as a target for +At this time I dont believe it has anything to do with the +Also I would need to change the unit tests +Im trying out ASI dont use it at all it comes on deployment +It really sucks to be forced to use such a workaround +I have checked the hosts files on both machines and cant see any have also checked the IP setup on these mechines and aside from some differences in subnet mask they areidenticaldev subnet subnet subnet subnet subnet subnet and dns entries on both machines identical +html svnnative +Yep looks like this worked fine +Ok! I havent realized thatS so I think that the best will be to release everything as instead of attach a patch file that I can commit which will update all the bundle version and dependencies as required to fix this issueAfter that we will have to release the following projectI dont think that the other project POM will have to be modified because they usually import the packages exported by the above bundles +forget about the first attached patch +Can you give a try to trunk +Applied Made some minor changes Added missing license header for one file +Havent tested lzo plugin but if it uses Hadoops it should work as well +Now it tests several different chaos monkey actions and also tests master mttr +libs +This could be achieved by a REQUIRESNEW transaction for the get operation +for the patch +Feature has been added +I created a test for this issue but cannot get it to fail using version +Verified in +duplicates +Moved over patch for vfszipspaces +Many thanks MohammedIs there any chance we could avoid buffering the cache entries in memory when saving to loading from the cache storageOleg +Can you compare how is this related to + +Please verify and close the issue +I know Atomikos implemented suspendresume differently and I had my reasons to implement it the way I did + for patch +However I was wondering about releasing after a month or which is in days +I think we should start the native process anyways +Where it differs is exactly the interpretation of the space + +Perhaps worth a separate ticket? +It is harder to count that because it is a very subjective number +It would build on the work I recently did to have better inferencing for lists and maps +I think what your describing is the of a Gateway proxy using an interface that exists on the refed bean of a downstream? Im moving this to M so we can discuss it a bit more +Patch applied. +We can have this scheduled for EAP +Well submit a patch for the branch +however having the branches in the few places to support both methods +The net effect of this change is that the exception will only be thrown if the refencing is SOAP RPC +No I dont think every time you save the version in the process definition should be updated by default +Lisa can you please look at this +That is strange as this jar is in the +Rob can you link it up +What version of JBAS is this a BLOCKER for? The Fix version is TBD +Tom this should be it +I think this issue can be resolved for the time being +The logic that does the actionAliasmethod separation is done in the form tag itselfchanges are atSending coresrcmainjavaorgapachestrutsactioncomponentsSending coresrctestjavaorgapachestrutsactionviewsjspuiAdding coresrctestresourcesorgapachestrutsactionviewsjspuiAdding coresrctestresourcesorgapachestrutsactionviewsjspui! Thanks a lot Toby. + +Hiram decision please +This should be improved when we move to POI beta as that has improved image detection logic +Both of these strategies youve hinted at being unacceptable so Im going to close this issue. +Committed to trunk +Im thinking specifically that the couchhttpd is separate from couch but couchchanges and couchmrview have their httpd code in those apps. +Do not deal with javadoc issue here +This task is the work on the clientwriter side and will not be done under +fixed in in svn +Candidate patch which adds a findbugs target to the ANT build along with related infrastructure +Passing along to Rich to look over +Would you mind rebasing your patch against current trunk and put it on Review board so that we can get it committed? Jarcec for taking a look at thisI rebased the patch and uploaded it to the RB. +It caught a bug that would have gone out the door with +Hope thats what youre looking for +Yep +Bulk closing stale resolved issues +Checked in jar files at subversion revision +Serializable object can be divide into three typesaClass instances representing classes that are not serializablebPrimitive objectscOther serializable objects +I was suprised that this issue is still present in the EAP +Ill be happy to submit a corresponding patch for +They all run those big irons and replaced Spring with Weld or +George can I close this issue +When the test runs it is supposed to delete A first but the logic was broken so it ended up just leaving A on the server +Weve accepted this standard and it is pretty uniformly applied at the moment. +Is Wall really? Do you have some autoconf foo to share to make this more portableAlso given that this is the linux dont you think its fair to assume GCC for now until someone takes up the task of making it run on other systems +This is the associated patch for this improvement +Patches welcome +Test doesnt test that the reserve works right I couldnt find a good clean way to do that without the pause stuff I introduced in the last patch +Patch committed at +More info With the patch applied lots of stuff seems to break +binary packageinclude javascript lib and to webworkfree use for webwork develop team +Chrome +removed as a an SXC like approach to make it scalable. +JoeTony I think were ready to go here Im sorry the svn setup took so long + Username letiemble +the same issue appear in +Why does it come into play in artifact resolution? I cant remember and I will take a look but that cache is to link up projects in a reactor to use compile classes directly so what is happening that makes this required for artifact resolution? This is strictly for in situ work +late please remove this poster indicates this is not a true issue + you ideally attach an Arquillian test to reproduce or failing that a war +I removed the explicit notion of the imported DS API version +Looks cool to me +Thanks Rob for reporting and taking the time to test this and Gilad for the fix +Ill look into it again later thanks +Does that matter +I wanted to try it also but so far havent had time +reopening to update release note release note for +Merged your pull request + patch described patch from Ate DoumaI am opening a new issue for the issue. +I just didnt dare to touch the SLD stroke stuff +Patch integrated to trunk. +Wrong operator was being used +I dont think theres anything we can do about it in the short term +Ive been meaning to test Ceph at for a while +Also in new version with eclipse I dont have Sonar perspective and measure view +Any size could be insufficient +Sim please do not do anything I have not completed half the steps that need to be done be patient and wait until the all clear is given +Isnt better to deal with this condition on instead modify +Sample rar I generated to try to reproduce the problem with a in the jar as well as in the rar +In general it really does boil down to choose one and I think most people end up implicitly choosing CPL since its the most easily compatible with other projects licensesAn update jffi and jaffl will gain EPL license becoming dual and EPL +Typing CtrlC in the console will interrupt the current thread +Im marking this as resolved adding a fix version later when known. +Feel free to test it on your system as well + plugin? I dont find such +Is this what you meanYes +Committed in r +Now that work is underway with AS all previous community releases are +MichaelShinsuke Sugaya has volunteered to do some testing on the file upload part +Fair enoughAbout the patch instead of creating a new attribute in the thread would it be possible to use? Or check if? In any case Im not satisfied by those kind of solutions +Amit thanks for reporting this bug and providing a patch +Introduce a new command that switches keyspace on the cli level +You might also consider XX XX XX if you are doing a lot of redeployment +Includes new message in inIve run simple tests under Geronimo and verified that are no longer being held by +Duplicate of. +Here is a revised patch for this issue +implemented +on my index when the query matches like less than of the docs Query method is fasterwhen it matches something like over the Filter method starts to win +Anyone interested in working on this? Id like to share the sources +Removing the Fix Version +But I will stay with until it happens again because once it starts happening as long as I dont do a clean it is reproducible +One thing less to think aboutUnless you run into the same taskdefclassloader problem we had with clover and the and ivy that have prevented us from doing the same thing with them +Everything still needs to compile with as of now so I guess we should remove these override annotations +for the patch +Thanks +the latest patch looks good to me +Please verify +Issue still present in havent looked at that yet so it will still be present +Done in trunk and merged to the branch +Fixes NPE in if the job is retired +This is duplicated by +The design whos causing the problem +Thanks Russ +Back to normal now. +Committed to trunk and merged to branch. +So by abstracting out the service tracker it will be possible to only report an issue at the end of processing the list of service validators +Ive moved the transaction manager is still to do +To test it out you need to edit and replace withI have run some performance tests with various types of load +This is fixed as of Juno SR +Thats exactly what I was thinking about +Rebased on trunk for v +Here is a trace with an identity column in the history table +maven projects attachedps sorry for that I set version to in since I dont know how to user version which is not release to maven repository formally +You mean +it is used by maven maven continuum so we cant use +In real environment this is a magnified valueSince its a bug we should discuss whether this decline is tolerable +Only replay +Committed without the commented out lines. +am also against any update of components until we get some tests going on. +Patch applied +Verified fixed +Screenshot of richpage with nested richlayout was attached +Nothing planned to support this so closing it. +Multilevel nested braces will lead to the number of scans close to exponential growth +Best solution would be to create a service containing CF reference clients can depend all issues to the CR +fix attached +Please review +Hi Levi these are in a really weird patch format +Anyway this patch should fix the first pain +I did some refactoring of pom earlier and it seems that I have removed a dependency needed by the build +Committed the patch to the new HTTP service project in Rev +As per it could be +I reviewed and applied the patch +bulk close of all resolved issues. +I created the documentation +The developer touch point is normally adding a mapper or something to an anyway so it wont much matter how we do it +And maybe that Plugin can be given a class name so it can create the internal plugin instance on demand +added Small patch solved my problem +Just wondering what would be the most obvious use case +Anyone had time to try this one out +once the implementation is doneShouldnt it be the other way around +So ignore them + +I hope this complies with JSR specs +can we turn on partition pruning by default in the unit tests new tests are getting added all the time and it might be difficult to catch these bugsWe can still turn it off in any production cluster till it gets reasonably tested +Hmm we need to throw an error that we dont support rpcencoded +Thanks RuthPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +Now that the Solr integration in PHP wiki page is updated to reflect the new download location for the client this issue no longer needs to stay open. +ok Ill get that diff U to you asap +project and +Parimal for your patch on Facility componentYour changes are in rev +was motivated because of these issues only +Yes sorry I am planning to add an empty file in srctest and use that as the default instead of devnull +Here is the patch for this issueTested the patch for all three process mentioned earlier +seems to be D Group encoded and cant deal with some code it encountered +A particular server may also have a default locale +Patch worked +Apache Felix Declarative Services has been released +Patch to remove the redundant files in and fix the POM +Could you clarify what you were wondering if Ive tried? The performance testing exposing the issue on? Or implementing this change on? Currently neither +dbd on master branch +Otherwise move to LATER +bq +Not really a or not a problem or record this work item somewhere else +His contribution is already committed to trunk rI see that you were not watching that issue +defanyunknown keyword to indicate variable decls for clear scoping +Apologies for the confusion that has been corrected now. +I am using target as temp dir which is certainly wrongLet me know if you need more help to get a good test case +Yes please go ahead! back to this I believe the process is that David will need to unlock the domain +simply works for me +committed this +This bug has been marked as a duplicate of +Replacement for the originally submitted patch + + +In fact in we even had some markup specified with wicket namespace and others specified with foobar namespace and it worked perfectly even markup inheritance worked with markups in the hierarchy having different namespacesAre you saying that in we need to specify a namespace change in the java code somewhere? General Namespace should not be a constantI had a hunch that some hard coded wicket namespace references in might be related or the cause of the issue +Wasnt that part fixed +hoss The approach is fine and if we need to refine it it can be taken up later +Committed in +If I miss your opinion please explain further +I think we should rollback it and plan a review of the code for Groovy m still getting hit by this in spite of implementing Graemes workaround which effectively makes use of the jar +With this issue can be resolved +Fixed via. +Ive reviewed the changes and they look good to me. +Hibernate provides a default mechanism to obtain the raw connection and provides an extension point for users to provide their own mechanism where the default mechanism doesnt workThe application that is using hibernate needs to provide an appropriate implementation of hibernatesYes this is open source but just because you are using DBCP as embedded in Tomcat that does not make all database problems you come across DBCP or Tomcat bugs +I implemented this feature with a small modification +war deployment has a dependency on and the interceptor has not been installed already it is too late by fix this issue by inventing a new +The general approach of having a common entry point which creates the temp file and adds a reference to a data structure stored in thread local is very sound +Cant see anything in the think that John is correct +This patch depends on the enum class from included here for convenience +This releases the connection for the back into the should have some tests for the connection releasing +It doesnt really have any meaning but I bet returning NULL there would not be acceptable +Moved to the dont know how useful it is since it is released under the is LGPL +btw what do you mean be stream the messages? you cant really start consuming as they go you gotta wait for the next superstep +it is a simple patch and thats what Grant wanted. +Hi Sandy The problem was caused by improper use of a static Configuration object across different tests +It can be interesting in itself if the region moves is used by a lot of clients +Working on this more however the test can fail with but will pass when run again wo a server restart +v is against branch +Lars Todd Yes the column prefix delete will break lazy seek optimization +Release NoteEclipse examples had an incorrect +validating parser. +I could deploy successfully on Apache Felix +However that is not something we can fit in the scope of the pipeline rework at this pointSome quick and dirty tests suggest that doing sums on integers is about an order of magnitude faster than doing it on floats +Resolving +Shows the results of running the test client. + for the have added some more code to make sure that maximization of multiple portlets one after another works properly when done using AJAX code +I have take a look and get back to can you add a test or convert an existing one with so that the persistent code is a new patch +While its been so long since I filed this remdiner issue for myself that I dont recall exactly what prompted it I am sure that it was caused by the presence of in some public API +A lot of nodes were being decommissioned which appears to have delayed replication after the first DN was lost and again after the second DN on the rack was lost +Here you would need to write an own class +Have you tried remove the use of string as I did? Or can you try exactly my posted program and see what you get +Ok Ive made a couple of tweaks to the unit tests +So I add null check code +All it does is count the tokens that go by and add those that fall in the range +Thanks +I committed a change that better optimized for boolean data +Alan this plan looks good for our requirements +Added support for aliases Added routing for delete by ids Changed logic so all router impls but implicit are acceptable for routingThe test error occurs because the clears the documents after it performs the updates +Should be good now and added your suggestions +On trunk and this example deploys without error +I will try the latest dev version anyway and come back to you if I notice the same error. +The scope of this ticket is creeping already as its getting mixed in with +Currently hitting some intermittent failures though +set the dependencies of the bundle to compile and set the requiredpackages to in the manifest +Thanks +Committed this patch +you can define a gateway that starts automatically gateways are executed before railo accept requests +Anyway I try to reproduce this with a fresh installation later +Yes I now see both s but Ripple fires deviceready and my app initializes +It must be in the config directory to ensure they are on the same filesystem so that the rename is atomic. + +We need facility in wal like we have in hfile for printing statistics on load carried +My wiki name is +Its definitely not worth a WARN +I am unable to reproduce this error +Andrea I am assigning this back to you for review because you have already put a great deal of work into the robustness of these tests and I know this failure has bugged you from time to needed to fix the build so I have committed the patch +AFAIK with this patch youd get a problem is in the dependencies report that shouldnt try to call that class when its a system patch should not be applied +Could you tell me if my fix resolved this issueRevision is +Still unclear to me what virtual folders are +New navigation node can be seen on both servers after restart of +Tested with latest Tomcat and on JBDS L +So they are too late for boot logging config setting them in is necessary +A bit improved version of the patch that uses hacks in the project file and still uses sbt to build +A dependency can have one and only one scope the syntax you showed is wrong and the warning is appropriate +I believe Reza on the JMS EG was looking for support for this as well +Fixed in version Committed revision Fixed in trunk Committed revision +Hairong just opened a better Jira +I wouldnt object to it being split into two uses the pattern logic uses the directory name OR logic includes the recursion and the includeHiddenincludeDirectory logicIt just seems like an unnecessary split to me I dont see the harm in having two execution paths in a single artifact but Im not architecting the framework so I leave it to any case the Skippable stuff needs to be fixed to conform to so some work needs to be done either way. +This is due to SPNEGO optionally requiring kerberos and not including the crucialSPNEGO is unexpectedly loaded even if webhdfs is disabled +I attach a patch which reflected comment +I was mistaken about it being yesterday +Implemented in revision there a reason why the collection is a set and not a list? +ill do it asapThanks +The patch works! did not fail +svn diff of the change that fixes the for the patch Poorna +Ah I admit I had not actually read this issue I was going to comment on but then noticed that it had been closed as duplicate of this issue so I followed the duplicates link and put my comment here +too +Yes I do agree sonar must fail if source directory doesnt exists but sonar shouldnt fail if source directory is empty +commit logCommitted revision Committed revision The documentation needs more attention +It is ok to move to +on patch +If no one else has any objections can you commit it +Ok Im fine w the formatting +Furthermore socket and Hdfs client may buffer some data which may will be truncatedWhen I implement the first edition of truncate prototype if the block or data the client required is truncated datanode will throw a exception and client will update the metadata to check if the data is truncated or the real error happened +Ok I have attached the debug log of the Ivy output using the version +xml using is the server + +This is there already +A bunch of Directories doesnt use said bufferSize at all making this parameter not only private Directory affairs but even further So my bet is introduce as a simple Enum change bufferSize parameter on createInputOutput to done +With the new servlet mechanism we no longer provide a static resource servlet. +Marking Out of Date Issues closed +Ive rolled back the Java Frame support for now as it sounds a lot more complex than expectedSo without this we turn around this issue +Patch Ran the file through asciidoc converted to html renamed to in the package source dir +Hence the protocol for journaling should be generic enough +Anyway good catch +New patch that applies to trunk +needs a different jira. +Please use this one +I think bucketCols should be same as groupbyCols +To install libXP on RHEL yum install lib RHEL please verify that libXp is still necessary as it is not in the RHN repository +have committed this +Hey Hari if someone just used an agent with a memory channel and an avro sink wouldnt this enact semantics? If an avro sink has a problem it fails the txn and then attempts to reconnect so this is a retry rightIf the goal is to have partial durability and get better performance one obvious extension would be to add some type of batching to the so that it doesnt sync at every txn but there are sync points periodically +So I think if we do not find at least one big implementation that really streams we should not implement streaming for jms in cxf as it would only increase complexity then +I have posted a question on the forum +By now I am fairly sure that Derby diverges from standard DRDA in many ways +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Sure feel free to appeal on dev +The trunk uses the latest. +Please provide an update on your issue as a comment +Thanks Clebert happy to help! +Fixed with on Build +It means you tried to cast something that is not there +Any objecttion +This issue must be fixed as a result of Axis +The harder problem is to coordinate our work +Move to +ElranI havent forgotten Ive just been swamped +Closing based on Russs comments abovethx russ +Anyway I think the general decision is to not change their hashCode behvaiour o I am fine with it +I forgot to mention that if you do provide such a method we will be more than please to add it to the API +For bit java need bit +Any word on this issue? I am facing the same problem +I was able to reproduce the issue in the trunk in and +I will look at using +using an explicit String property type results in the same error +Hey guys here is my review great job should have an license header in it should be removed seems like jQuery is included multiple times that correct +Thanks +Alessio would this interfere with your work on EMF stuff? Can I go and fix it +I understand embracing legacy systems is a huge issue but that would be with any persistence layer +patch file for to trunk as Rev +Test class will not be able to fix this properly until we implement interceptors via clarify we will be able to fix the however as the final method cannot be overridden an invocation cannot be delegated to be bean instance and will be invoked directly on the interceptor proxy +it should be but I cant figure out how to change it +I can reproduce the problem which is related to visibility of fields and possibly a same test also fails for +Please reviewNote that I have included the Javadoc Ive proposed for +This exists but isnt critical path at this point +Only when using REPLSYNC as cache mode I see this problem +From what I can see here are outstandings Remove schema recompute in when appropriate A lot more unit testsAnybody waiting for something else from me +rather than right click theres a startstop icon on the tool bar and the pull down menu on the properties view +Has been solved with with version +This change changes network server to use getBytesgetString and not hold locks for any of the calls +We should update this and incorporate into Lucene +applied thanks for the patch and for the IT. +Currently if an Error happens in oozie it does not appear anywhere in the logs +I just committed this +I dont see users doing much outside of a projectMy employer calls this info must gather meaning it must be gathered by the user and is required by the support team to start diagnosing the issueWhat specific pieces of data would be helpful? And would there be differences in that list between platforms +Closing all issues that were resolved for. +Committed revision +Pending range collision check is currently too and should be relaxed a bit +A short fix +Thank you +Modified the message prompts and the title pages of Build Environments and Installations +Many thanks for this new functionality +You can see from the dialog that I am trying to modify the constraints for the however you can see in the background that the bladeServer variable has been bound to the +This patch adds support for simple preloading to Shindig +has been marked as a duplicate of this bug +But wow libxml is faster than string concatenation +Creating a on port to get a random avail socket for the RM HS rpc web ports +Here are my two projectsselfregister is the service that disappeared when I deployed taketwo +So we should be able to close it as fixed +There doesnt seem to be a way to a Path but at the same time it doesnt look like you can mutate a Path eitherThis didnt crop up on the other relative path tests because they arent a single +And the no problem with that one +please in Geotools please +Im glad you appreciated! And thanks for the hint +Any update on this issue? Note This only occurs when the insertsave occurs in the +What do you think if we support the same behavior in HDFS? I asked Daryn this question earlier +Because the counters use a selector multiple tables on the same page now all share a saved state +Reverting patch will post alternate approach soon +One can not amend the text on the default buttons +Since that class is also referenced by the Tomcat every Class used in our app is retained in the permgen +Assigning to to get a better fix in to eliminate the error message once and for all +means that hotdeploy of perInstance aspect will break the schema needs to be checked upon deployment time? how to allow to hotdeploy something then I suggest both has advice of a perInstance aspectUNION is took longer to argue than implement +What if you just did a read from the db like in oracle select sysdate from dual after a failure? if that fails then dont you know it was the connection that failed? doesnt that allow you to never have to call the stuff before the transaction starts to get rid of that performance penalty +patch for hdfs trunk +Try again with a protocol plugin that understands https like Julien did and not +Ive just gone ahead and done that for you +since it does not really seem that anything is being converted just that you need to do some custom encoding of the field name +So even this region who should be assinged after x will also get updated to the current timeSo now the TM will be able to take action only after xDelta +Many Thanks. +Thank you for spotting this. +The above problem are only for Unix +This is then used to directly create the authorization header +Okay well use the plexus configuration file for the index maintenance +Just missing some automated integration tests +This work has been completed as part of the recent geoserver demo prep. +For now at least I have given the following as very approximate meanings for the proof of the pudding is in the eatingsha de provar una cosa abans de dir resandel ms important sn els menjar i beure fotre i no creure +Stay tuned +Bravo! Ben +defer issues to +Yes that should be should be fixed in +To work around this issue when using Hibernate make sure the parameter is set to afterstatement +I did not see any deadlocks although I did see lock timeouts. +Will not make +Please dont hesitate to ignore this last comment from Chris +Task is of a bulk update all resolutions changed to done please review history to original resolution type +Candidate for SOA to talk to Kurt +Looks good verified in JBDS request was integrated and applied on with pull request on with pull request on with pull request +Move the fix to +Hackery for Basic Persister ensures that getters setters are taken off columns mashed into SelectInsertUpdate for the collection stuff I actually have no idea whether this works or whether its needed I just added it because it looked like it needed it +It may not all be relevant anymore +Ask and youll receive +Ive just downloaded the file MB thats a big test caseJust kiddin Ill take a look and figure out whats the problem +This patch fixes the problem +I forgot specify whole environment +Ok Rahul that makes lots of sense to me too +Though its not human readable as its just a long number +OK now its clear braces ARE recommended regardless of number of lines in ifelse +Even with prefetchSize you will already loose message +moved to +We should probably just defer to the jboss maven plugin for deployment support +thanx for sticking with this one Alexis +Besides is not about always having the very latest releases its about a coherent stack +We tried with Wicket and the problem still persisted +Yes at least a partial dupe +Verified in reopening to update release note status. +sh file with this line in it then pass the files to this script does the exact same thing as the dosunix script +Thanks for incorporating the fix discussed earlier +Could you add unit tests for the new functionality? At minimum a new test for when FEED would be appropriate +IIRC this must be quite doneBut i print a lot of messages to state that it happened there before. +Please write a test that proves the patch +I still think it would be better to not inlcude any slfj binding in the war but place our chosen default in solrexamplelib so people easily can swap things out without repacking +patch against against needs to be committed on and both the fixed version. +My bad for missing that the point was to test a stable sort +Logged In YES useridOK what about this difference between and? That is I would like to know if possible what direction jboss is going in before I update our install doc +Ill take a look into this. +Freddy I think you reopened the wrong ticket History service is +mail server are trickier to setup than I expected +could it be something that we do on the standard resource pathSomething like addingresettrue to a POST call to that resource or maybe having a custom http verb to do that like RESET instead of POST +Patch attached + +Note that this bug makes AAAs under other AAAs behave like inner administrative areas +Attached patch corrects the documentation +make it possible to empty internal note or shipping +good the patch does not work for nested segment groups and such at least I think I have not considered such a scenario at the time of writing the patch +they can quickly fill up the browser screen completely +The proposed solution provides a facility that can be used by such external components as a implementation but the proposed solution is not a replacement for a implementationRegards is an updated patch file that corresponds to AXIS r +Hi Brandon What about fsynced blocks +Attached the project simply extract and run you will see the failure +Thats what Im trying to do with right nowIm happy with any way forward that is verified to work +The wording about classes is a little confusing +Update patch for review +I guess I missed it +of +This would be a pretty big change to the classification scheme that was argued out quite exhaustively in there should be time for people to if they wish before anything is committed +to trunk +local This setting has to do with connecting to Bonjour services on the local network +Makes it unusable +used the test runner for all of Solr and I see it fails now +Closing issue although Im not in a position to test this. +New version of the patch which catches all patch by Emond is applied. +OK +Committed the change to the trunk as svn revision +I took into account comments of Thomas and created a new version of patch +When I close an I cant seem to delete all its files on Windows so it appears perhaps something is not getting closed +Hi Paul sorry for the late reply +Username mtalamona + on patchChange should to is on commit and same for all other should methods +You can find that information in the manual +currently the gadgets access permission does not prevent an user to add the gadget as a gadget into the dashboard can reproduce it now and this is considered as a cache issue at UI level +It looks ok to me +yes I used the pom the maintainer had started and forgot to pull those out +This is definitely doable without significant effort +In my app struts filter triggers on all pages +It is theoretically possible to have the sum of frequencies from all shards to exceed size of an int but I dont think it is practical right now +to trunk and +I just committed this +Thanks million for the patch! Committed +Done in revision +Can the index be made optional? In our typical the old data is hardly queried so we are willing to trade off cpu and notsupport skipping rows for old data to save some spaceThe way I understand it index creation can be specified during creation so it can be made optional +Yes you are right it actually is +If it was a bypass in the past then it was a bug +So for sure we need to keep both concepts around Can a project have two logos? At least shirts for each logo? Obviously we would choose one as the main logo +Committed +Im backing out the changes from and will use to create a exmaple for integration testing. +Done +I say go another step remove any properties from the Model about the data source if at all only have the reference to the connection profile name +It was released with the fix in +I will add the test results and a configuration file for the new classes as soon as the patch is checked in +The error occurs only on the first line of the block I pasted +I still am not able to help provide tests but this is an important fix as it does occur in production +Setting new Affects Testing Regression Blocks Testing Fields +We now use todo this stuff in MINA +Hi Ivan Chris I think the behavior of winutils is orthogonal to what the Java code invoking it does +I tracked the preferences and i found a different configuration for me +Can you please try without with a clean local have attached the output from running mvn X archetypegenerateBtw we are using an Artifactory server to proxy and cache all repository requests +This issue has now been resolved by revision +The lease expiration handler will queue a operation that must run before the root region is reassigned because the of the dead server must be split before any regions served by the dead server are reassignedWhy werent these seen before? synchronization has changed which may have exposed these conditions which existed before but were never observed +I installed the maven plugin for Eclipse and all goes fine with the actual install +I think this ought to work like a block created with define +Verified on revision +MadhumitaAh correct console mode doesnt not support shortcutsWe had come across this a while back and just dismissed it with the same reasoning you provided above that most users in cli will not avail itHowever now that there is a use case in that they can use that automation install in GUI mode perhaps its worth it to implement this featureWe are looking into it now and trying to implement it +Or having a way to replicate the docs +Verified on JBDS. +I just think we need to strike that delicate balance between benefit of people submitting useful changes and risk from those changes breaking thingsGiven the current ee framework I think we in general benefit from making it easy to run the ee tests on a single dev machine +No idea where it comes from need to analyze further +The patch looks correct to me +Dirk I was merely highlighting that there could be other convenient classes to come in the future +Thanks for resolving this issue I agree it can be closed. +Currently does not which is a fundamentally broken data model +Committed to trunk +PeteI checked the code and it should work as it does +nor is the Suspened status of the channel +This patch attempts to address that problem +Updated patch addresses Owens comments adds datanodenamenode logging on during beginning has raid removed general cleanup +Can we remove the? It only seems to box +I dont think it makes much sense otherwise +Add to the classpath for the eclipse target in ant +it uses Pelops at the moment +already fixedbtw thank for report and patch +Your fix does not pick policies attached at operation level +Hi TedUnfortunately I will be offline for the next couple of weeks and I no longer have the code data sandbox I used when reporting this issue at the first place so that I wont be able to provide a quick fix before leaving +Use Strings instead of in the API +Based on these experiences we could probably provide a common storage layer more widely suitable for multiple Hadoop related projects +this is a good observation diogo but i think you may be characterizing it improperly +Well what if you want to retain the state in the visitor before the other walker is called? You would have to associate a stack for that too then +But outright blocking it just seems silly +The issue has been marked resolved if you have a patch please and provide the patch preferably with a test case that demonstrates the problem +yes I verified we can take the suggested approach. +Rob I assume this is one of our good WTP exception friends +Until this bug is fixed does anyone know how to remove this plugin from eclipse +Hi sorry but I tried to reproduce the problem with the steps I described above and +Ok +New patch after resolving merge conflicts +Next step is moving all the and systemStill need to fix even if now it should be relatively easy +I made one other change with this patch add defaultconf as a valid attribute to the configurations element +No other action will be taken +The example and my code work well now + committed to trunk +Fix from back ported to builderseap resolved in CP CR has been removed in CP +I havnt tested HDFS yet +There are rulesets and over rules and they change in every minor release +I wish this to be backported to since Im just about to use cucumber in my project in which is in use +Also make these object oriented +Not for commit +is the interface of the object which we return which is and therefore is threadsafe +project and +Plugins have been updated but theme work still remains +contextzip +Removed open jdk dependency +of testcase with updates for jetty +Anyway do that seems to be very difficult and inconvenient because you need to store such relationship in the EL expression and the resulting design can be broken easily if the component is relocated or manipulated in some way +Ill run my tests once I get everything back together again +I added some logging about the workaround as discussed on the dev in revision +Resolving issue as we have no feedback +At any rate no need for a patch I will get the merged figured out soon +Moving to since no work has been done and the ticket is ve started working on this JIRA +It just has empty implementations for void methods and return null for methods with return values +For the purpose of debugging that seems to make more sense than relying on the s contents +Reopening since Fil didnt like me resolving it as fixed in master. +In the reference I catch the exception bytes being returned and throw them as a +Does your custom font actually contain the glyphs for the you wantto renderIf not then add those to your custom font and it should display correctlyBTW unless you are really sure its a bug it is generally preferable to askquestions on the mailing list regarding the perceived problem firstThe issue of and custom fonts was previously discussed on that list +Thanks a lot for the reviews Andy. +Running locally against it does not fail for me +This patch accomplishes what I was looking for although it would be nice to expand it as mentioned in the TODONot committing at this moment because it breaks inner which Ill have to look at another day +What is the stack trace on failure? Is it from the conversation context being destroyed twice and the being null? One option would be to move the destruction flag into the conversation manager and have it reset when the conversation is set up +Created patch +Sure I could do a step on the tree walker but It wouldnt be proper +I just commited my first version of the JCR Node Type Management API +Related problem the redirect support currently provided by SWF builds on the detecting the redirect prefix +I changed this to be fixed to the type that we want and it now works fineThe downside now is that we cant use this restlet bit in a generic way but for now we dont need toThanks for your help on this and pointing me in the right directionTom +This item is closed because its not a bug in lognet. +You might want to test with logging set to TRACE +Attaching screenshot what i see +The patch modifies determination of and propertiesWith this patch simpleLaunch works just fine +thanks for the patch Erik. +Pull out all the classes that are specific but dont really touch the rest of the codebase +Ill bug Marc for a real release of +I suspect this is part of the mobile first directionPersonally I prefer the new way and I would call it a feature but you are correct the input fields will take of the width available +Then others might be more likely to add additional cases improving our test coverageAnd if were now go ahead and resolve this bug +Ive made Dans suggested adjustment +ivycachejars and Usersmumrah +Why is this change only for branch? Will the version be removed in +It does not fix theproblem this bug describes +This will make the config +I havent debugged it yet +Thanks +good catch +When making changes here it should remain possible to require global CREATEADMIN rights for those actions by policy +Pending resolution +looks current version will always be null since its initialized after this exception can be throwed +Ok so should say cannot be located at javacompenv +Implementing it the way youre suggesting would force users to either have to update urls every time they change runtime or have urls updated but be unable to pass in custom system properties for their apps +As we get more control over object placement we should also be charging things like code and other VM structures to the VM instead of the application +Each block grants a lump of permissions to a single jar file + on disabling the Chukwa unit tests on the they failing +You mention commenting out a line in some file +I cant fix this in hibernate tools since Hibernate core does not have a way to introspect the resulting returned data +move all files under pluginsderby to pluginderbyderby +Closing resolved issues. +This issue has been opened with the spec group. +Patch addressing comments +Im having trouble reproducing this do you have a code example which fails +We leave all the shard key hashing up to and the Cloud aware client we use see +Please use a static field for such values +Hi BrianThanks for the info +Let me know what you prefer +Manually verified all the links in and work fine +slip to m since no updates done by because of no update + Ive updated the Queue and Topic service the Name is now available and can be set in the template +how you generated those classes whether you used the CLI or aJava program etc +Patch attached + Looks good +David Ive added space admin rights to you so feel free to customize All content can rearrange widgets add plain text and so on and so on +Verified on. +If there is no set the statistics defaults tobut if theres a config in you get +Olgas concern was that changed the semantics of combiners in an incompatible way +doesnt actually pass CPU into the Makefile so without this patch builds on Solarissparc do not work because the library generated has the wrong filenameThe JVM architectures for sparc are sparc for bit and sparcv for bit +Can we close this issue now? Thanks +Ive committed this and merged to +There must be something wrong in the Hudson machine +astests +Dont break tests including those originally written by others +The overall thinking I have in mind is somewhat like this the NN stores the location of the extended attribute +lib in the distribution zipIts located +Instead I think we should have it in the constructors of the appropriate Strategys +Updated patch to add some docs changes too +martijn actually the underscore in the id is not a problem as the first thing the code does is escape an underscore with two underscores +This should solve the problem seen by the userIf several threads use the same connection they will now share the referenced but the API calls are serialized at the JDBC level so thats okIf a thread uses several connections there will now be one map per connection which is ok tooRegressions ran ok +Yes! +Released version isable to recognize but the bug persists +Thanks Ted! +is required only by spring plugin so thats why ws tester works good on JBT even after meclipse or subversive installation +HelionI think this might be a duplicate of issue +Closing once again +This program simply writes a bunch of data first and tries to read it all backIf the write phase spans multiple you will notice that the read phase fails with missing valuesThe numColumns needs to be large enough try something like +OKSo do I need to open an other jira to report this enhancement or does this one is enough +I will be proposing this as a wontfix +After that i will clean it up and close fixed the +oh +Issue may already be resolved may be invalid or may never be fixed +In W in dev mode the request took about a second +test failed in trunk and buildI reproduced the test failure onLooking into the test output +Yes I will commit not exactly as is as as I said Id like to add some documentation in ant messages to help users to complete the task manually +Thanks for review +Michal The was helpful +Your idea is just to break in different tests right +But all of the replicas are at least as long as the last successful fsync which is the only guarantee we have to provideThoughts +When running a build definition containing deploy together with a distributionManagement using scp the deployment via scp sometimes never terminates here also +I think I fixed this in the latest snapshot +Please +This was fixed in Affects and Fix In Versions +is closed closing all dependent issues. +This will take a bit more time though +Uw mail wordt niet doorgestuurd +FOP comes with Xerces transition resolvedfixed bugs to closedfixed +The different versions of in the working and versions +So in order to replace it all you need to do is to rewrite the tagx files which contain Spring JS Dojo and create an which replaces the Roo installed default tagx files with yours +we will update soon +Heres the obvious patch +proper integration with the so that the JDBC authority registration command is included in connectors +That was fast +maybe +Test fails on windows bc of file separator differences + +Argh yet another twist +yep the manifest generation is what is going to be likely used in Maven by default so all jars that are built have a reasonable default latest unified diff between trunk and sandbox +Thank you very much for the patch +Following we are not setting as default anymore but rather set no value at all in such a case +I dont know why this was moved to but the same problem is also present in Glassfish +Bulk close for +Fixed in by directly extending and adding a filter on Group instances when looking for subject user principal +All of the connector code should be better unified at that time making this task a little easier +This is the stuff Grid did for jBPM if Im not mistaken +Patch looks good +Looks like this is not possible to reproduce this issue with latest versions of Sonar and the +The base issue here is what the abstraction or abstractions for controlling tabbing should beboth programmatic and declarative +That is something I think we can sacrifice +The test case in the latest patch seems avoiding the assertion failure +In particular look at my last comment on and Dags response on +ThanksBest regards by Andrew. +keysIf we have to wait till cascading does new MR interface lets punt on these as you suggest but if we really need these things Comparable would be good to make the change before releaseDo they have to be Comparable? Can we not pass Comparator to use +Did anyone found a solution for this issue +OK in any case I will add some method for deletion to the collection +The documentation looks great +I was just trying to get the indentation right +And please forget the proxy for now I want to see it working first then and only then will I start working on the proxy stuff +I just have no idea how many of these that we need +Lohit is right about the logging its redundant sinceCanceling this patch until we decide what to do with the metric +However segment parts related to parsing will NOT be converted and have to be removed prior to converting +Im fine with SYSCSRELOADSECURITYPOLICY +The sun jars will eventually be replaced by our own version of the jars so no more work is necessary on this. +This is fixed in branch now +I just committed this +I would try to summarize the main rafactoring things here +Dont know how to test +Adrian I didnt change the priority you did so I havent disregared any instructions listed in the wiki page +OK +That makes sense its not a big problem your code seems to work with but it takes a bit of care to get this right +hod contrib was removed +Hi SureshI hope to comment out the my proxy credential code! So that when we switch back the developer can do the modifications easilyLahiru +commited to PortalBranch. + +Patch is created against but seems to apply cleanly to trunk as of the datatypes can now be converted into any type by using a Spring +patch jars +Call the servlet on httplocalhostservlet The buildfile will try to copy the generated jarwar file to the deploy but this might fail with a different location of the server +Sorry when I meant adding an attribute I should have said replacing the node by another having different attributes which is what the replaceNode is for +I applied this patch with trunk as of August +I ask you something? Why do not you simply construct requests with the virtual host in the request URI? made a quick check we already construct requests URI with the virtual host but this information is not used by for cookie validation at least on +you dont skip any real seek by lazy seek +This turned out to be the result of a bad PDF file +Im not familiar enough with the parser to estimate how difficult that would be +The streaming renderer does not break at the first error but tries to merrily go on +Tests included as well might help to git diff w the test file since I something +set EXECUTABLEINSTALLSBIN EXECUTABLE +the test failures do not have anything to do with this patch +Thus we are currently releasing dependent beans of +I guess this only makes sense once Groovy is released +We should be referring to it as Negotiation not to avoid confusion +trying file jar trying builtinLib trying builtinLib trying resource UserspchengDocumentslibrubygemsgemsliblibrubysiteruby trying resource UserspchengDocumentslibrubygemsgemsliblibrubysiteruby trying resource UserspchengDocumentslibrubygemsgemsliblibrubysiterubyshared trying resource UserspchengDocumentslibrubygemsgemsliblibrubysiterubyshared found jarfileUserspchengDocumentslibrubygemsgemsliblibrubysiterubyshared me the original bug for me is when both of those +can we please keep it open until I find some timesure +I am not sure if thats the right solution +I have committed this +Please confirm the current status of this issue and if its incorrect change it as required +This makes the changes a little more thorough +This should work in any truely spec compliant environment +This issue looks ready to commit +so to get this straight You want to partial update a repeater and then also do this for new rowsI dont think that is possible at this time because we do component replacement as a wholeSo for a new entry you have to repaint the complete table +But lets do that in another ticket for our purposes here correctness will suffice and we can optimize later +If you want real heap allocation you must use instrumentation +Can someone who does do that pleasePaul only need to rebuild if any of the Java code or libraries change +but we should not hardcode field names anywhere +bin should use bash so it tries to JAVAHOME w each invocation however the difference in run time is still tiny +Feel free to contribute to Roo as well. +Thanks for the patch Michajlo +Closing issues before next release. +Sandy yes thanks closing this out. +Thanks for the earlier reviews and commentsI have separated to handle encrypted file +Thanks for the quick fix! +Moving to next release. +r for armoring the in message provider so that it doesnt care what context it is called in it will send messages to an appropriate place +Closing this since includes more information and the patch +HiIve tested the solution described above and it works +Im moving this out to under the assumption that it will be the last version before +did you try to run an test on a real cluster? Maybe works somewhat different than real +The loop wont stop if there is a bug in modifyTable call +patches are not applyable in this format. +Double click on object in Project Explorer opens editor if it is not open yet and selects the clicked object in the editor +Fix attached +I kind of doubt it is the same thing I saw when I originally logged this issue as I typically set up scripts for testing with absolute paths and would have used the same script when I upgraded the JVM that made the problem go away +This is configurable on the Configuration tab of the web console +Sorry for the confusionWe need a test for this +thanks Namit. +I dont see why headerslock is still needed +take that back +This problem is in the binary distribution +This will help provide a clean view and ease readability +If you want more of the exposed as API you need to file a feature request in to move it to +Fixed problems applying current test package renamed testx package to test and committed patch. +Attaching patch +Thanks for quick response +Let me see what we can do +older issues for Apache since were no longer actively working on these at the moment. +The only existing solution we have is to create the lucene indexes manually much like Marcel and then do a once the indexes are rebuilt +The disk based stuff will be done in +I however did it a bit differently and allowed for exporting the ip connection count map for statistical purposes and of course used as opposed to synchronized blocks +Fixed the module fields in each of the filesI verified that I can publish snapshots to the asf repo with this patch in place +The above files and +Will look at this if there is a complaint +And hes probably not using the ECW toolkit +proposed for reporting the bug and providing the patch +I believe this problem was really an annotations bug +If there is a problem please open a new jira and link this one with that. +but findbugs warnings are still being +Does not need the tool any more! +Can you explore using some kind of static constant for names createSnapshot etc +i think weve fixed all spots that exist. +The changes look fine to me +Moved to upon release of +Attached the patch please review the changes +these class should also ommit the Odf prefix and Element suffix +Assigning back to QE for confirmation +which IMHO violates the principle of least surprise + tested +It should also be possible to remove the groovy JAR dependency completely but keep the library sort of a blank runtime +I just repeated the steps as described by Scott works for me +It is an incompatible change that I wouldnt expect anyone to need +With the patch the results are consistently higher than without the patchNote that the indexes here are for primary key constraints so a fix for would probably also have avoided the background work in this case +Fix spelling in method +All the functions addressed in this patch are vectorized implementations of existing functions +We are already buffering IO operations tofrom files +Yea verily it does work in +Some issues with identified in the stack traces above look like different problems so not resolving as dup +Will commit shortly +Ill leave the issues raised on the comments in Gavins blog and on the forum post to a rainy day when I can write some cogent explanatory material in the Hibernate section of the ORM for raising this in the first +Fixed patch apply issue in previous catch on the break +Actual result Error JSON error in console log +BTW I marked this resolved as a Cannot Reproduce for now if you can figure out how to reproduce we can reopen cool thanks for the analysis Rob +Improvements in this area are already on the TODO list under the tasknamed Improve Error Reporting +The same also happens with xsgroup look for addressType in the same wsdl +Committed the path to the SVN trunk +The work for is done moving to so that we dont forget about the same testing when it happensWe should probably automate it some more though +I have attached the diff but I am afraid I am a raw newbie and am not following the correct conventions +Please assign to Lukas once component is added to Metamer +Less than week to pushing off more items to +It might be good to check with the user list tooWould it be possible to make the enhanced restrictions only occur on new databases and ones that have been created with the restrictions? The thing that makes me most wary about messing with permissions is that the errors that users get with mixed permissions are pretty ugly like container cannot be opened or cant read some specific transaction log file during recovery +Once the line function is usable the module can be moved into or wherever. +Ive fixed this in revision +Patch looks good except for the accidentally leaked message destined for the aliensPFixing it myself +Let cassandra run until you see statements like Deleted files Stop cassandra and try to start and it will give you all the exceptions that I am talking aboutAlso I believe contains some more data compared to +ACS feature freeze date is November th +Im referring to and which are related to the orderly shutdown +You can as on that bug report before final is released +Probably the comment on the merge back in x as well. +Id be happy to help add it +Use a instead +Im pretty sure I can accomplish this by reloading all configurations +All right docs are in about the changes +Closed with. +Handlers params +No longer applies. +Can this be resolved with Priority +Also we canshould move keyaliases and columnaliases to Lists instead of +Well see if it works at least I think building on osx works + change to a catch +Is this really a very unhelpful error message? Seems pretty good to me the cause of the problem is listed pretty high up on the error page and if you look at the javadocs for the answer is right thereThrown when an application tries to create an instance of a class using the newInstance method in class Class but the specified class object cannot be instantiated because it is an interface or is an abstract class +That one feels pretty goodNote that if you find that this trick hurts performance rather than improves it you can disable it by setting the property to true +I have to get the test case working. +Hi Simon Im going to be away for the rest of the month but will hopefully be able to work more on this in a couple of weeks +Please apply close this issue and create new issue +I can think of examples either wayIf a field has a lot of synonyms it will match more termsShouldnt its length be considered longer for the purposes of determining how specific it is to a query +Also look at scenario described here +Since nothing we do with the is multithreaded Im fairly certain this is caused by background tasks left in the AES stage after tests +That does everything I had in my timer code and I think fits James intentOh well I did learn a ton playing with different ways to do it +I think that the reference documentation is almost correct only inaccuracy is when you use Java bean component then you need the annotation on method implementation +Thanks Myroslav! + for the to trunk. +I did not see this behavior in Thanks are you consistently hitting this? Runping mentioned that the client node had no cpu left +I have the same problem with jboss and jboss as +We make libraries plugin a required bundle for other eclipse pluginsHow this solves the concerns you mentionedIn this approach Library bundle is generated using Felix plugin and embed the dependencies in the same manner we do it now +but that all should not stop a contributor from donating a onload attr +committed. +Does it basically mean that each DMLC should get its own connection +You have a problem that you want help solving +Attaching patch file that makes works on both Linux and Solaris +I agree with you then that theres no race conditionLatest patch looks good to me +I think it passed because we have the parameter set up in as wellAll unit tests passed on except + works with graphicImage when you set attachedfalse on the contextMenu +Server A believes the session should be expired because it received an exception +Lets add a method to get the retry count in this Jira and passing it in rpc request header in +So if anyone wanted to use them as a benchmark then I would move those algorithms into the test folders +The algorithm for permutating from one set of potential candidates to another when a constraint violation is detected is exhaustive but not very smart +Why is conversion of fragment name to lowercase neededThis is done to among fragment names +As for when the dependency will get picked up it is at the trigger actions bind phase +Perhaps it would be easier to separate parts by and disallow the separator in parts. +Feel free to reopen this issue should the problem still BelaWe managed to reproduce the very same error on our environment with +The fix has been made to the WSM build filesStill have to remove the doc references to +This is a feature request +This default value is just an example or this should work accross the database? Ive got error full is not maped +You can add fix to a new version of the patch Gregory +This has been broken since the initial open source release +nevermind disregard. +Moving to for now +any idea for thatThanks +smooks configuration file attached +Yes It successfully compiled at command line +as a first fix of the inconsistency that connection to the root works but then results in thus with default installation connection to the non existing root collection is not allowed any moreresolving issue fixedwe may reopen it if we find out that another solution would be much better. +Thanks. +See the +If there is a communication error it would make sense to retry a few times +Ignore attachements +Is it the same w +BTW This is not the first SVN action before Retrieve changelog the source directories are checked for local modifications +Sorry disregard the first patch as this one incorporates it and adds all the necessary logic for the admin screens +I hope thats OK +Incorporated Nicholass comment +Resizing the window fixes this. +Is it the correct skin +When have you planned the release +You could certainly then use this list as the topic list in the to only crawl under a certain category +I adjusted the code accordingly in revision +bq +could you do that or suggest another resolution to the question +Found out that the tests failure were due to +zip file and all will be okI already tried the +Patch looks fine to these scenarios and found them to pass a Start a randomwriter job and check if all windows appear properly in the nodes section +Im going to resolve this as fixed again unless someone submits a script that breaks +Filling in the reviewer field +I think we should just retire the widget and concentrate on maintaining the two other ones Natter has largely the same functionality +There is hardcoded in commandLine so in linux environment it will be problem. +Note that this is one case where it would be more useful to use the Future API rather than the Callback APIFutures are very useful for using an asynchronous API in a synchronous fashion or in hybrid use cases +Updated patch to address most of Elis comments +I am also encountering the same exact issue +Can we expect this to be checked in into mainline soon +Should we merge this changes to CR branch +Maven is dying a quick death +When the known issues are resolved and enough testing has occurred I think the next question is can the fix be back ported +Its easy to deal with but kind of strange that there is no default behavior +Restore system to standard operationAdditionally this contians a fix to the problem from more thing Since were dealing with large volumes of data we changed the logging around a little bit +This is the first patch I send to Shindig or to an open source project in general +I just tried this configuration out and it works +The explanation makes sense +The selected scheduler can still mess with the set maxInterval but with this patch the fetchTime is actually correctly set for redirects. +Ill check it in +Bulk closing issues resolved in JEXL. +The JDK will wrap the runnable using an adapter which does not expose API for you to get access to the ExchangeEven if you create a custom and submit that then its the adapter you get when the task is rejected +I guess that the configure script should probably support either install +An example file and example STX stylesheet +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first. +Is AS still under development? Is it worth changing it +For Musachy BarrosoHe saw that is a problem too +Affected version should read Fix versionJira maintainer needs to update release versions +and not feet +That could simply be a coincidence but I probably restarted it hit it or more times +One of the interesting things I noticed is that I had a npe in +This patch wont quite work +When represented within the detyped model they should be represent by the proper types +We now compare the name to find the matching operation +There is a race condition in the original source so the mpendingevent may not be reset to NULL +Please add comment regarding what is meansIn the test what is the reason behind changing the standard error stream +Higuysone month has been passed but there are not comments yetI just dint konw the you +OK good +Im not really clear on the last comment +As I mentioned the CCSID manager isnt yet being used by live code despite being in place and all going well the EBCDIC will be the default even after the patch without anything breaking +Code for the changes is already committed in past daysAshish Vijaywargiya +Doing so invariably generates an NPE +is there a jck test failing +Could you create a new JIRA project for that and give me some time to move review the imported tickets + +Strange +Ive just discovered it during testing this patch +We have to implement the BPMN DI spec for importing +So now it works but the next step will be to inject the Validator Factory instance that is created by a deployer +Queues can be configured with weights to share the clustersomething like and can be arranged in a hierarchy to subdivide resources within different poolsbq +Im wondering if using the item count is the wrong way to approach this +Just as annotation The pomp POM gets also invalid if it is released with the relase plugin +can you please help me to the issuebetter if you could attach ur service and whatever necessary to the issueThanks try Nightly or RCsthanksdims +Hi GalderI had no idea anyone was using +we can do that but then we need one more api in to get last good +This configuration allows tests to write anywhere currently there are a few that do write outside of target +Thanks Juergen! +Are you using the defaults for the file upload or do you have special storagelimit settingsI was going to look at possibly switching this to using Apache Commons implementations but the new servlet spec will supposedly have form handling built in so I decided to wait +Patch applied +Feel free to pick one that works better for the Hadoop community +Ben is this one covered by already +Approved for. +Here are the logs of my coworker reports that it is working with a recent +But it should still be verified that the JMS Destinations node is not present on an EAP install that has installed +for example does forkexec internally but does not pass the terminal streams along to the subprocess +I fixed and also related to this +Code in trunk now Java compatible +Disable Workspace Resolution finds them just fine out of the +Linking to the overall Better integration between SCA and DAS of like I was too fast closing it as a duplicate +on it AaronCreating the new connection after CREATE COLUMNFAMILY worked +Patch s a possible solution that would work for Curator +I didnt use hadoops as they were missing some of the asserts that I would need +Unfortunately I cant find it at the moment +in ant work as expected +If you dont have a patch ready for submission let me attempt to pull out changes related to this JIRA from +Forgot to add change to the latest patch +If the generic part of moves to JSE and the part like this contribution which focus on implementation specific into the commonsI think the Patricia Tree gives a great value for the +Now attaching updated patch file. +Sounds acceptable feature for think there are a few reporting plugins that have had to rely upon using the default location hardcoded due to the fact that it was not programmably retrievable from the project model or the site plugin configuration +And you are able to update the endpointAddress field now for a SOAP binding on the wizard and in the property pageIt also adds the workaround for to the rest of the binding composites to avoid unexpectedly clicking the Remove button when you hit enter on a binding field + +ClosingI do recommend sticking with the in your config +I did notchange everything to statics +Odd +This patch is dependent on the from +It is just warnings from +During unit testing I noticed that sporadically fails and fails frequently +I think the main issue here is backward compatibility there are zillions of pages still stemming from the days that expect this behavior and changing this would lead to broken links everywhere +Done! please apply the and Committed +Said Oops! Something went wrong but at least it did its job because now I can login +Paul can you help review the whole patch? I am still in middle of sth else +I dont suppose there is a list of those subsystems available? Or a JIRA for them? +Mass close all issues in resolved state that have not received further comments in the last month +It would be nice if this bug will be you +I will run all unit tests see if any regression it brings +reorganized manual. +Build the eclipse project using mvn eclipse +Fixed in the current svn headCheck for the response is received properly by the client +Of course I suppose that is currently possible with a minimum of on a small grid +RS B starts splitting +Thanks all this is committed +security Unicode English text with very long lines +I havent checked if thats a problem in the code thats touched by the patch but thought Id raise the issue +Resolving as Fixed based on the changes in revision +I started all cassandra nodes +Meanwhile the obr adapter is avaliable at aries +Please check if its APR implementation problem +I am running with this patch applied now +since released closing. +We need to fix the source of these asserts rather than just disabling them by turning off NDEBUG +rebase the patch with for adding more descriptions to the tests +Also uploaded on the review board +Sandeep youre right timestamp is better hereJun I dont think we need to do anything special we already have code to handle saying i want to block for N replicas before calling the write a success +But I am not an HDFS committerHey Robert this indeed belongs to COMMON since and are in COMMON +I see two instances of this exception in the Zhenyu logs +moving the code out of the sync block exposes a potential for out of order acks to be sent to the broker which violates that ack model that AMQ uses so this fix cant go into trunk +Jeez yeah I was quite embarassed during the demo because I could not hit the damn placemark +Ill keep you postedCheers it would be very useful if you can share NM logs around the time when container was lost +stay I just added your blog +CP fixed in revision fixed in revision +Attached a server logj messages for the patch +Cool not sure why I didnt spot that big blue attachement yesterdayWDYT about the issue of a possible race condition on this static variable when running concurrent testsIs there no way in Mockito to refer to the current configuration to set it per mockito instance instead of globally +ok if I file a followup Jira so we can get the build unblocked +Committed to trunk. +no frameworks can be written to handle these exceptions they must not occur to have proper function session. +Does it have to be an? I like +Thanks for the review Stack. +Otherwise well get lost in the myriad scheduler features that we can implement but of little use +Toby Jira mangled your inline patch +Testing it actually works. +Good writeup too bad it didnt pan out +always assumed the Context Root was set +defer all open issues to all open issues to +Since your already doing +yes is the same as this one +Not sure how well this works right now but its a good start +Ive fixed the failure case for the time being +BJThat would be great if you could get something implemented using the htmlunit libraryRight now the patch pulls a page from the Hotwax demo site and puts it in the Example component main page +So that is something that should be added to the documentation not new and noteworthy +This issue has been resolved for one month or more without complaints +verified by Andrey +Should have made two failing tests sorry Ill learnAnyway the test now asserts that is returns false but if you start jetty and load the frontpage it displays true +Thank case +First patch updates jdktools dependencies to fetch and extract JDI related jars from Eclipse site +Note that should still allow negative timestamps otherwise a cluster containing negative timestamps will not be able to operate + +I do not know if it is also the case with quotes +Will do this soon +Weve been running this for a couple of weeks and it is essentially stable +See also the recent rebase discussion +We should probably wait until we start releasing on to fix this +Implemented by +For Oak Id like to implement it spec compliant and use the node type of the child node as isThis will have some impact on existing applications migrating to Oak +Thanks Ashutosh +Alternatively should we look at an external tool that can do this analysis and compress the files based on mechanism proposed by Todd to minimize the impact to HDFS core code +Part of the release. +I believe that we should remove the second on from the clause on which the object depends onI agree with your recommendation that we should raise the visibility of the new topic by making it parallel to Using SQL standard authorization +My assumption is that you cant use it within a web environment +Thus X and V are shown in the context menu for the Visual Part +Attached the full build log +Please take a look there +For some strange reason Ma does fix this so for me you can close this with not reproduceable +revision . +bq +Ill create a patch for the patch Shikhar +There is a todo saying handle session disconnects and expires +In the file add defaultcatalog property after that everything works fine +Added pom snippet to a sample project and deployed it in a t seem to replicate the issue using +revision to set the Fix Versions field to JDO beta. +Patch is reviewed and modified by Sarah +As i see in Mapper if a page reference of node was removed reference value of node also set null value regardless of it is null or not +Sorry for moving quick trying to get at least an initial implementation of auto failover committed quickly and we can continue to improve and fix it up +I think this is more straightforward and lets us control whathow we expose +Attached patch +I just committed something to trunk that I hope will work but works quite a bit differently +In and earlier a is thrown with builds +The patch looks mostly good +Were going to use DWR for this type of validation so things have changed a bit +This issue has been fixed and released as part of release +I can remove compilation error yet still make it works in sun JDK Do you know anyway of make it works in both IBM and SUN? I mean are there any library? Thanks Indika +As long as noone provides for example a customized type handler I have no intention to fix this +Alejandro thank you for the quick response +I think it makes sense in loads of ways since a common problem of is that during the initial import the user ends up with thousands of small store files because some family grows faster and triggered the flushes which in turn generates incredible compaction churn +I manually verified that the created directories are cleaned up after a run of this test +I do have a Qt build environment so I will be able to do so but I dont predict it will be for another week or two +At work you have to use a proxy at home you dontIs there any way i can use maven following this pattern without having to edit the file twice a day +Q was the intent for to be a general purpose repo ala piggybank? If so how did we come up with the name? I dont remember the context +Still not seeing exactly how this is happening +A small patch on top of your latest patch to reorganize some of the streamRangesrequestRanges code +BTW we will care checking permission of when user or group names are should create a JIRA for Documentation team about this new this patch We change user group membership name from to +So it dont give any information about for CA and Properties to import TLD into Palette view and drop some tag from it onto jsp editor +I agree with you that we should clearly document its limitations though and fix the javadoc contracts everywhere to be correct and consistent +Set Fix version to +Ill figure it out around today or tmr +I like the idea making the pool not part of the data thread group this can actually be accomplished more easily by just moving the initialization from to which is a more logical place anyway +I was going to start on the using s however because this is heavily cachedependent I think itll have to wait for because weneed to regenerate the cache per reader +It fixes some minor problems encountered when using a newer revision as base +We are asking because this error looks very unusual for normal Groovy +I just committed this. +Verified with beta builds closing. +Can you please try this with the latest snapshot if youre not already doing so and if the problem persists can you please provide a complete project whereby I can reproduce the problemThanks +Updated patch for removing the portion of the code corresponding to from this patch and placing it in its own patch on that jira +We arent planning on moving the unit tests into the various different packages at the moment. +Ive got it working for the default integer comparator as outlined in the previous attachments descriptionIf the approach and test seems to be satisfactory Ill work on building the provision for loading a user defined value comparison class +Lets target insteadhas better decimal parsing as beforeAlso has better error messaging when the exception isFor some reason the code special cased that exception and returned the Command not found message +This change may potentially break existing applications in very nasty ways +However for behaviour there should be an option antlike so that output should be as as it is possible to get including the s q v d behaviours +Close issue +I think that Speculative Execution change to false on hadoop helped as well +I think the added info to diagnostics is important to prevent users tripping themselves +Handling sticky bit with ls output style is tricky +Patch integrated to trunk. +Michael Sorry fo such big delay +looks good +You could try terms against each of segments when segments may have had terms a piece in them +Attached +As described by Kai +The unsigned manifest is intact +Roy there is already a set of context properties available for the WSDL constructs that we should be passing inTake a look at +This will allow people to use alternate impls for a FS schema easily +Thanks Sharad and Arun! +Any ETA for? + +Ive since lost that old config. +May be you did the same +I could put in but that just masks the known problem +K N would be executed +This patch makes them all consistently defined in their own separate class making it easier for extension use and inheritance +okie im secretly happy to drop support +are implemented +And thatnull value can be returned if user chooses to not implement it +Merged to the branch in revision +Thx Stefano for the link +HiI see Ning is not working on this at present +Thats quite strange while debugging it hangs within the tasks in the supersteps +Hello TimThe fix looks good many thanksBest regards by Andrew. +Rohini what if we hide this behind a property something like I see accidental writes to the same output path causing problems all the time +Sam I believe the way it works in trunk is that writers are killed when recovery is started and then the replicas enter the RUR state so no more writes are allowed until a recovery happensIn thats not the case see ah yea i had trunk open and just checkedits exactly that way +OK Ive just attached to the issue the trace file +I missed and +last time i used it it wasnt that nicehowever some people are wondering why we have which also uses hg but we dont offer hg mirrorssince it isnt the first time i hear it i opened this issue +Suppose a request comes in to create a file +But I do not see this change in +Plus it has a flag to revert to using empty methods instead of the Exception +Yes the file magic checks and the are not changed +Following the Law of Demeter its better to extend existing interfaces rather than expose additional objects +Especially since it complicates the simple most common case of the usage and there definitelyis no need for a new namespace since users will not make up arbitrary sling propertiesthis will always be avery defined set +core tests +Server dependencies now always come before classes in the users deployment so API classes will never be picked up from a user deployment unless the user configures an exclusion +to Charles he knows the answer +Im just not seeing who is using it +Ram sure for commit +Many thanks RichieOleg +BTW any other method of doing this would require an upgrade of the jbossorg and seamframework docbook xslt +Ive seen some recent reports of this failure +Patch v changes based on Xuefus v more changes per Xuefus review +Marked for postgreSQL as PASSES and tests of package test for Oracle DB and postgreSQL as PASSES or respectively as FAILS +Its mostly meant for new features or changes an operatoruser should really be aware of +The only issue with that is that you need to have a scheduled executor doing that every n secondsminutes right? Where would this go? in each of the components +This is in trunk now marking it as resolved pending further testing and feedback +We should be able to use the example in to figure out how to specify our own impl of +The only time you will not need a udf model is with dynamic vdbs +Adding Transmitting file dataCommitted revision . +Tag libs that explicitly invoke their body Script dont catch instead they let the Scripts go right out their doTag method +YesThank YouI tried latest snapshot +I guess there might be some limitations when looking at mapping enums as constructor arguments but this will be dealt with elsewhere. +I think as Sergey suggested having something like HBASECLASSPATHOVERRIDE in front would be safer +I would expect any given IDE to do all the formattings automaticallyAFAIK Claus makes use of and maybe he can help you here. +I realize that having them on the front page is useful for quick access to the capabilities document so I would not be against moving the links to the side bar with the link to the server page but the links in the middle of the patch seems out of place and a bit sloppy +The only difference is I am trying to install for the first time not as an upgrade to. +Added to Server Configuration Guide as new seciton + various widget test files +Any thoughts when this might be included in a formal release? We are using it for one of our clients that wants to brand their feed using the channel description +I like how the proxy server main method takes one option the name of a config file +Were going to put some notes in the documentation specifying that users should explode the WAR in +is related to that George! +If you are still having issues please provide a reproduction and reopen. +yes thanks will answer to Nicolas now. +The path for the command I was using was incomplete +Sorry for the delay Nigel +zip +basically i need to use button for the look and feel of a button but am bemused as to why using button means my form will always be submitted +Great! Its been good to be able to contribute and I hope to be able to more for this awesome project when I have the spare time +page and java +It would probably make sense to make this configurable +I tested this issue based on the latest jbds build +I will keep going with this +Yes everything is fine +Newer code uses apr for allocation if there are leaks please open an issue per leak. +This is happening on Springsource Tool Suite which is built on top of eclipse +Thanks Ive fixed the case. +Please suggest any improvements +bq +Its kind of a big change to get into though +press in the window that geronimo was started in to shut it down +A bit offtopic here but I was wondering should I setup a build project that can be used to test Axis against some WSDLs from Amazon and Google +Skipped the lineI fixed it in rev +what i should have said was that i dont advocate adding a mention of it to the docs +New resourcelib directory containing Hcatalog jar and build instructions README +Rather one could use the workaround suggested by Owen or we could add a new Writable class that writes the name of the class in each instance but it would not be to make this change to +If you find further issues please create a new jira and link it to this one. +Fixed +For we need a volunteer to backport it to can take a look at backporting the patch once it is accepted as part of +Im going to resolve which was opened earlier. +Yes I see it +Instead we check if it is null and if so create an empty local map +Simon personally I would prefer the Version argument used for such thingsI know this isnt popular but Id actually be for having say a javacc grammar file that differs from the one with version driving ityeah it would be duplicated code but its mostly code anyway and I think it would be simple to understand what is going on +cannot see any issue in the nbm plugin itself closing as works as intended +Thanks a lot for taking a look Daryn +Hi Nicholas +The attribute is set to true in all config +Heres the errors I see +And even for dbs given my understanding of how the parallelization would occur I do not think there is a concernI am not familiar with details of configuration on demand +I have applied the patch and tested with around lakh blocks in my cluster and scanning is happening as expected +The does not work correct +r mstarets lineChanged paths A myfacestrinidadtrunksrcmainorgapachemyfacestrinidadcomponent M myfacestrinidadtrunksrcmainxrtsorgapachemyfacestrinidadresource A myfacestrinidadtrunksrcmainresourcescomponentstrinidad Add a new Trinidad component +please verify if it applied as you expectedDavanum would you please verify whether Seans patch does work for you +At this point I dont plan to convert this +Patch which uses in to make sure only one import is running at a given timeIll commit this in a day or two +Thanks for pointing this out +Youre trying to instantiate a private internal class from MVEL +I still think its a worse API than the other patch I had attached a while back for the reasons mentioned above but I wont block it +Committed revision . +Hi Arun just had a brief look at the patch +Yeah right missed that +This proposal would simply create a separate classloader for the custom components +Because is checksum file system +Werner you are right +Here is the JSP from the BEA engineer that is used to test serialization +Setting this to fix version as a high value fix candidate +Bretts right but I cant pass on Wendys does Archiva determine a snapshot? Is it because it has found a SNAPSHOT in the path leading up to an artifact? Is it because the pom for that artifact has listed as its versionWeve bent the rules here a bit and the answer will effect how purging is looked at. +Changed to Xnobootclasspath in ed +After discussing this with Ilya on IRC this isnt a bug as much as its a feature requestThe IPs being assigned from the public network portion of the description is an outstanding question but Ilya will raise it on the to undstand if its a bug or not +I am considering fundamental changes to the type binding infrastructure and will fold this into them. +Test case demonstrating package versioning details +I checked a fix in for the parse error in r +step is removed +PeteCan you comment on how exactly chokes? If I do an ls on a NFS folder with many files in it it will likely also take some time to return +POC Screenshot do you think Tyler +My feedbackQuery Who is We? Is this Red Hat? Can We be changed to something more specificrepresentativeGrammar typos +Geir are there any problems with the patch? Looks like you postponed it +Kabir assigning to you as you have done all the work! +Chatting with Andrew about how this xhr might throw an exception and we wonder if perhaps the file url format is perhaps not supported for xhr request on some platform +Thanks for the suggestion! Ive added such a property in a revised form as a work property on point Ive added such an assertion +on patch for trunk and +The patch got attached thrice though I did it only once +the beanwrite timezone attribute or rather obviouslymight look like EuropeBerlinsee also also for enum use JSTL. +Thanks for the bug report and patch. +Was working on some of the and got caught with another snag +Turning gets into scans will cause some minor functional changes +You need to run separate processes and use tcp to connect to not +Hummm +However when I tried to set true it works perfectly +thanks! +if the ledger has been removed we just skip replaying the entry +This is an unfortunate side effect of a refinement in the matching process that final introduced +The file has one small change to fix something that is not in the two earlier files I have posted +Committed to branch +The current warning message isProfile with id explicit has not been activated +Was included in +I believe its not only for BTW but could be used with all versions. +Im just not sure this is actually the best bang for the buck and might make layering less cleanBecause the would all take pbs rather than the GetPutDelete etc? And doing this conversion would be a bunch of work that would be better spent doing other stuffServerside going from pb into GetDeletePut just to get the data into and out of regions seems gratuitous and crud we should purgeYour profiling though would seem to make this a minor issue one I would have thought prviously critical to address +vars available +My recent fix seems to have resolved this problem +The fix was simply to add the class in the during the build. +I think DEBUG might be more appropriate in this case than TRACE +This fix is correct but the script will fail because it uses the shutdown command in a way that is not recommended or intended +Unfortunately the Open With is a whole other ball of wax +I dont know why the access log only has the one request this time without the secondary ones +still requires facts to be serializable +wont fix in HB +Igor I need exactly the same functionality that is requested in this bugAnd Im going to implement it in the same way you suggestedOf course I would like to push my implementation upstreamPlease reopen the bug and assign it to me I will attach a patch in one day after it. +verified on and on and it has both Provider and on JBDS. +IMO the purpose of the facade is to keep the API compatible I would not worry about the Writable part +This leaves us in a position +far as I can tell the issue can be closed as resolved +Throughput is identical and errors are goneIm nervous about error handling though +If it doesnt solve it organically Ive added the test so Ill make sure it passes before I merge the branch +I am backporting this to +META goes offline +Its a bit cleaner that way +on shifting away from using in the JIRA for +Ill test w Gregorys patch to confirm that it causes and then will cancel and then apply Pavels +listing the healthcheck policies +Sorry I missed your patch +I never use bit +I never predict dates though probably more likely to get done sooner if someone from the community contributes a working patch + for the updated patch +JoeySee for audit logging +Im currently writing the patch +I commited the patch including Georges suggectionsDIGY +Thanks JarekJoseph we still have a Jetty issue here for +I have not been able to determine a fix for this yet +Didnt updated page before posting my comment +bulk close of all resolved issues. +this is a duplicate of sorry for creating it twice +Is commenting that assumption adequate +In my case a modal window was created with right size color and transparency +Apparently meta TS is set +Could we get a new patchset rebased working with trunk that just includes the CLHM prod +Changes the scm url format link on the start page so that it goes to the Maven SCM site +fixed by the listed JIRA issues +I can copy code from one place to the next But if someone wants to do something better go ahead + in jclouds has an example of how to do this +But we can increase the threshold even higher if we think that doubling doesnt put us enough on the safe side +ToddThanks! Were using +Yes +The code creates a connection to a JMX on initial request message and caches it for subsequent lookups +Please use tomcat to view the stats +and Ill discuss the feature in the mailing listIm working against the latest version from trunk. +Actually one may even consider doing deadlock detection every time a lock cannot be granted immediately +Im having difficulty in understanding what the issue is +I just tested your page without the and it worked fine +The rest requires agreement among the participants +Patch which disallows partial updates on signature generating with a better test +Moving this to CR since there is a code freeze on Monday +Jostling around the release goals and numbers +Please testreview +I gather that it is part of ICU +This bug has already been reported +As this is mostly the hint that cms content can be moved separately to the rest of the This one is ok and should be ID column +thanks +There are places on the site that mention the bit but putting the dependency at the top of the user guide sounds like a very good ideaThe reason for not using generics is to support pre users there are still a lot around +In tests with a cold cache on K blocks this method of block report generation is approximately twice as fast as compared to find datatoddhdfscurrent name blk a not name +If someone does it this wek we can pull it back into +The average size is the used memory divided by the element count +It appears to only happen in the context of +In SVN and tested with +Patches welcome +Hi Keith my comment was thinking of Bidirectional Links Circular References in the object graph +Tested with the latest code it is fine +You are free to disagree +offline and he agreed to look at the issue +Fixed on trunk r. +Probably something Eclipse related +Im taking a look at this +done +Can we see the relevant part of the Planet model +This essentially reverts the output of main to what was prior to +Patch which allows shards to be set in defaults and not cause an infinite loop if the shard is the same as the host +Updated patch to not set the number of local mappers to +Need to update the spec too +My is a little rusty did some early last year but havent touched it for a while +This happens to be backed by the and can be changed without breaking compatibility +Hiro It looks good to me +If I missed your idea please correct me +modified agree that the solution is simple enough so each application could have its own +For the moment I would like to propose committing this patch but commenting out the last test case +Some code refactorings and added some more jdocs +In other words I didnt test your new code +Matt please let me know if you need help committing it +On Ununtu it is working fine +Im fairly confident that the proposed patch is OK as there can be no conflicting change notifications as the read lock is still held while the item state updates are being delivered to the listenersHowever it would still be nice if someone else could have a look at the patch before I apply it as the SISM locking is pretty critical to Jackrabbit +Storing date values as type date in domain mappings will result in having upper limits included in lists an equals filter on datetime field setting precisionday wont return any record unless values are directly stored as are normal behaviours and theyre not directly related to plugin but maybe its better having them explicitly written out in case of doubt +For examplea deploy inplace start startLevel DThis command records the bundle in the above properties file and start it in osgi frameworkb deploy listThis command can list all the records as followingIndex Installed Started URI true true fileDFelixbundles false false fileDgtrbd true true fileDgtrbdGEP can use the new to persist an bundle installation +Many thanks patch applied with svn revision For future reference patch doesnt like gits binary diff format +HiI was trying cvb with the Reuters collection in local mode and it was hanging everytime +One line patch should fix the bug as makes the test to pass +Rick do you really like to see STMTID in the HTML? Since the query is displayed I feel like its not usefulIf that addition is useful where you like it to be displayed? Is it good to be in the header sectionThanks +Done. +This is indeed annoying +Thanks for fixing it +Modify the smtp settings and email address in start the app and then check the resulting email +Is duplicate of +Ive made the timeout value as milliseconds so we have sufficient granularity to cater for all eventualities +Ok Iwill +i am thinking of having a error console on rave page for such errors +Blank text field error shown +Patch is applied at +Even if from your apppoint of view it is single threaded derby may use multiple threads per db to take care of background things like checkpoint anddelete row processing +yes I looked at it again and notice that the test is unluckily called many +In some of the most commercially popular systems which implement snapshots snapshots do not count against the disk quotasThanks for the comment +The idea is to switch seamlessly between form taglib and standard spring taglib +Its only read and match against column families so as long as we prefix it with something that cannot be contained in a column family name +Thats great +A could then have a Attribute which contains all necessary info including the parser and flags like the one we are talking about +Good to see this ticket reopen +I was thinking of perhaps something a little fancier +Apply the patch to trunk +Unable to duplicate +Let me know if you have any questions about the testCheers problem was with combination of multiple methods masking resolved by adding additional checks for . +This interface will then permit the analyst to supervise manage and pilot the migration processThis Replayer will be able to determine which previous invokes need to be compensated due to data changes or flow changes +For rewrite of queries have a warning and cutoff of the generated terms rather than a hard fail +Thanks Uwe Ive fixed the patch as per your commentsAlso Ive added that implements the interface +Everything works fine now +A warning message is logged when those deprecated properties are used +I will hold off committing this patch however since a more general solution may be possible +Thanks Sven I have already seen the article and the standard +We also need to load the jar after loading the +Ill wait a day to see if there are any objections and otherwise Ill commit it +Please verify it on the next steps +Committed in the branch +Revision +Ill look at it later this week when I get a bit of time on my hands +I simplified previous patch and added a framework for integrated tests similar to +So you got it fixed or do you need helpGreeds +I just committed thisThank you Nicholas and Sameer. +Good catch +One can also create a table with as the rowkey type probably should disallow that +It is definitely a bug to include that jar file +patch looks fine +Thats what I have tried to do with my implementation +Patch is working tested on my notebook with both native and cxf stack and EAP +Ive just committed this +Please let me know if it needs any changes +would be availableThanks and congratulations for the communitys reactivity +The only problem now that previous build is moving when new build is ready to publishWhy dont do it like thisLets assume new build is ready for publishing and we have twooneziro builds published already +It is not only the specification but all software I have seen worked this way even GUI apps +Need to catch exception +Revision +You can have a try or you send your generation ods file to me and i test it in my environmentRegardsDevin Hi Devin As per your previous communication i updated to but still i am facing the same error issue +Peter can you provide any supporting linksinformation on the expected header format when using data? Was the choice of selectorprefix based upon a particular products export? +Dont forget the example shell scripts +Jucapatch applied good workim going to schedule another it hangs on methond DEBUG drop table what causes this? +It should be. +Im sure dblevins will ping on irc when he is ready to get access. +Not really +Could you add some +And then try serializing an instance with enough data to trigger the problem probably just a single array element should be enoughI would really want to resolve this before releasing thank you for all your help +perhaps because it will be recreating an which takes the current System time as part of its name +I will do as you request but theres a chance that by doing so Ill need reapproval here + seems to address a good number of issues as well similar to this ticket +Now I see its not working on Tomcat +Ah I can only reproduce it by writing a file using Java code to ensure there is a trailing space +MarcusIll excellent idea +This is old information anyways and can be updated +fixed up the bundle with a new pom which has optional dependencies according to syslogjs. +Salut Benoitsounds interesting could you share please a patch where showing the complete refactor? Just to have a look at the overall designMerci en avance and theirs integration into work on interface +Would you propose changing it in general or just in the packaging +recover fileAlso at this point theres no a valid way to process invalid in hadoop so paying the price when we know that something could be wrong seems to be better that crashing a MR with input files and then having to parse them all +Several people have tested the build and a build has been through the standard CI problem with WS reported which may be a problem in the CI build investigating +instead of actual characters if that would be easier to readcode +I think the former is better +Yogi added to the damagecontrol project +Def strange +Allans patch actually fixes my issues both with an activeProfile set and without + component also has the same issue +Attaching new patch with minor changes and to make the patch not dependent on the patch of as is resolved as invalid for trunk +Patched filesmodified libbuildrjavamodified specjava applied +Attachment has been added with description WIP patch March th created a simple document to test the feature and it doesnt seem to work +ok +Running through Hudson +Could you please verify it if it works as you expected +Im agree that GC should not move objects during the enumerationThe example if JIT has references to the same object JIT will report both of them and expect GC updates both references +It built but please check I didnt mess upThanks +I think that we already have an IDL compiler from a previous donation +show attribute was added with an errors made by Vladimir Sukhov +Im sorry to not resolve this problem +Ill keep you from CVS on does not seem to fix this problem +committed thanks selckin +Ive created a client side exception mapper facility added interface to extended to process implementations of that are annotated by Provider created a hierarchy of subclasses of which mirrors the hierarchy of exceptions thrown by created a default exception mapper which is installed by altered to use an available client exception mapperAll of the new exceptions are which gives a clean extension of the api since any application code that catches will continue to workIssues still working on unit tests need a client exception mapper for Given that methods throw Exception Im assuming that the client exception mapper isnt necessary there +How many bits do you need? Clint what about you doing your types in same way? ANDing or a new typeIf will do it lets make the changes now before release +But doesnt this lead to a similar situation when trying to open a URI thats not http or file? Not sure thats important though +seems reverse scan degrades read perf about its still acceptable per my viewIIRC cassandra supports scan backwards as wellIt would be better if could provide this feature besides creating another mirror table +Stack traces on startup are not +sshidrsaIll produce a new patch for thisIve been running Hadoop integration tests OK but I havent run these yet +Resolved with r +I agree it seems odd to be dividing by residual weights but I could be missing the intent +Attaching which removes the special treatment of SYNCCTL in the sanity check +If this file cant really have its name changed at all I would go with a sleepretry. +Thanks for spotting this it will be fixed in +This block avoids calling flushInternal if no new data has been written since the last flush call +TedIs the name was okBut the name was given lik that because this change is done for RSZKOPENING state +Latest patch looks good +Moving to CR since we have less than weeks till Beta +It seems outdated +My apologies to all Merry Christmas and a Happy New YearRob +i think this is an extremely strict interpretation of API change +JPP BOM and artifacts are still not available techpreview repo +Thank you for applying this patch +Also this issue is solved by one of these classes should meet your requirements. +Type and custom request conditions can now be provided by overriding protected methods in +ACK +Ive added the comment to this fix +to make sure the UI tests continue to pass? You will first need to install the latest version of python on your system of courseThanks +Thanks a lot +forrest site works fine on my Japanese OSKarl is your system setup to useexport LCALLexport LANGexport LANGUAGE +The number of splits we got was GM but still too many +Invalid GC testing +but the at the plugin level shouldnt be creating actual project artifacts should they? I would reckon that to be a common contract that the actual maven mojo plugins do the actual artifact creation and the just enables as much as possible as they can in the IDE +After that Ill commit this patch +Markus I will provide the patch against trunk +Is it necessary to split the testPing test up as part of this? It seems like it would bea simpler patch if we just addressed the locale issue for now +There certainly are plenty of permissively licensed JDBC drivers that can be used as an alternative +Thanks Thomas. +Please zip them up and upload them to bugzillaThanks ChudnovetsPlease do not change subject of the original bug +Uploaded patch +Its more related to Avro +then you have it already got confused with addition I believe only happen when replicated caches are t reproduce very likely out of date. +Thanks the SSH must go through another IP +This still needs to be added to +It will be handled on should have discussed about closing this issue +Verified in CR + +to commit +Oh and thanks a bunch for this fix these are the ones I like where the problem is already diagnosed for me +Hmmm +I merged this to. +I have described the steps to reproduce the issue in this you apply the patch the issue wont will review and comment your changes later +could we also add a check for the block size when a value that is not a multiple of is sent the user is redirected and then the call fails +So it is not portable +If I cant figure it out Ill split this into two patches one for fadvise and another for syncfilerange which is less obviously useful +New patch attached for review after fixing review comments and my observation +Henryk do you mind working on documentation as well? We always need documentation for any camel component we provide +Its very confusing that you cant reproduce this +It appears these methods exist in the current class +Spencer +Updated PR created +Putting the patch back to the queue +Ive added code to ignore empty actionMapping elemenets and to reduce the chances of one being generatedtrunk rbranch the issue in the current code base +Ill try to make a test casesimpler service reproducing the bug above but that will not be for a week or twook no probs given that works it is not that critical thencheers any type support might not work well when wrapping we will address when needed +Committed at subversion revision +In the patch I remove the incomplete entries before the post execution hooks so that hooks dont have to change +You need to switch to the new runner to run tests that just use annotations and dont follow the naming and subclassing conventions +or would we create a new entry point at some other level? a static method added to? Just wondering +Need to investigate more +This would help both processing multiple versions and easily parsing timestamps which are free inside cells +Already fixed in merged to about test failures running the tests using the layout in the release +changed to fixed +A link to alternative UI at the bottom would be very convenient for devqauser to checkout alternative without having to enabledisable js andor type explicit +t we think about cloning the closure we curry too? If the closure for example is used in different builder then this most probably will cause quite big problems as it is now +updated the patch prints the error message if any exceptions while getting the version info +Now we are writing a l health check module and i agree with your idea +Merged. +HelloWould you please try my patchThanksBest regards Andrew Javadoc patch committed in revision +In Camel the is now pooled which resolves the thread safe issue there as only one instance is using it at any given time +So can you provide a test case for it +Perfect that was the problem now I can work with openshift without problems +Updated definition of in so that overriding injection is disabled +Patch against + Final is released +Added my implementation in revision +Perhaps the JON people want node identity as well thats why I asked on the that Im thinking longer term here +Still to be checked into reopen it because headers in are not and confirmed +Will reopen if necessary +its a great little test +This might be the case if there are insufficient statistics buckets available +Reopening although I am having a hard time believing that the fix was not helpful hoping to confirm that there was a deployment error +I just committed this +Reason is that version actually broke datasource capability of the Maven plugin +I merged this to too +and for large enterprise projects it generates a fair amount of in possession of a Windows x box who can devise a workaround for windows builds? +EugeneAre you saying that the need to import the +bulk defer of unresolved bugs to +I added a public empty constructor and it fixed the exception +Ive done that may be missing something here isnt the +Sorry for any inconvenience +It changes the method name assert to assert to match the semantics of the flagThe method name was confusing in the stack trace +Hi Brandon I have been running this for a while now it is exiting +Does this problem exist in XML Beans as wellColm +I have attached an alternate patch for this issue +thats great to hear thanks for submitting the issue +Im aware that does not work on Servlet containers of course +Moving to since we burned as well +scheduling for for now can be moved off later +heres my patch again for what its worth +I reattched the patch +The former usage is deprecatedThanks for the patch! +Nikolay working well for Queues nice component. +Source included +The simple project to repeat the bug +Verified for +Remove unused imports DONE +Yeah got distracted with some things +Lets not mark this as resolved until all tests passPiotr can you fire up the mobilespec app and run the tests and let us know if they all pass +Pushing for fixVersion to to LATER since it is unlikely that well have the time to implement this in the near future there are many more items with higher priorities. +bug in has been fixed so this can be closed too +I subscribe to the Unix philosophy do one thing and do it right so I think that if we can implement autosuggest that works well from the technical POV then it will become a reliable component that you can combine in many creative ways to satisfy different scenarios of which there are likely many more than what you described +I am happy to help checking the report +Would the logic be better targeted to returning string arrays rather than just strings +The client doesnt really know anything about the broker configuration so detecting this isnt trivial +ignore the comment on n n looks like its already covered + +Another issue I ran into with assertSequence is that one of the table names has an +it also works in H +I agree haha +Does the configuration requires to point to a specific classIf the service cannot be started out of the box we probably should not start it on installation +Please follow the and comments in the code than give it a tryRegardsAli Sadik stuff AliIll review and commit your patch as soon as we are done with the releasecheers update on this samplethanks look into this shortlyThanksSanka +I am uncertain removing JDK dependency is a good idea +correctIts worksThanks! +Any progress on this front +I could define a base form bean that has neither such that a later extends form bean adds the type attribute +Do we have anymore thoughts on this can someone review +The utility tests can be moved to testmapreduce package itself something in line of so that other tests can use them too no? Minor nit Keep the code wrapped at incorporating review comments +know what you mean +There is no new release of yet so you still have to set vnclisten to for now +MarkAssigned to you +On Ramesh Reddy confirmed the fix for fixed the problem +Just found a similar bug +we saw another failure of this test today on our CI server +Committed revision . +Contribution of Net Perl module version +At the same time move the method to be above the ordinal methods and near its overload peers +is this patch still relevant +If you want to avoid this problem in the future you can use me and the configurator for me. +The serialization behavior is not detailed enough in the released JDO specification +Fixed on closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Whats left to get this resolved +committed +Many thanks +Uwe it is up to you +nuspec file updated and committed replacement NUGET package pushed to repository. +Allow for the configuration of shared attributes +Restart TB +patch looks good +Did you get signed up? If not I can file an issue with SF and add you by hand +For Java EE Web Project it opens normally +on applying promo code both ruleswill get apply +This patch has gone stale can you regenerate please +But code seems to be exactly the same in trunk +Added in r after doing some minor changes +Thanks Lohit! +Well you did not say what you were looking for at all that coupled with the fact that you created the issue as a New Feature well. +Did you mean in your write up or are you saying theres also an issue with the subsystem manifest +If have to remove the Guvnor menu from the root context menu I can change it only appear in Navigor view +Unit test for duplicate messages is to minor because we only block on multicast messages but credits are sent as OOB unicasts so an OOB unicast CREDIT message from B will always be able to pass a multicast blocked on B because the latter acquires the lock for B in NAKACK whereas the former acquires the lock for B in UNICASTWe will still have to fix this but it is not that urgent multicasts are passed up immediately but then the thread returns and does NOT try to acquire the lock +So this is quite deliberate +The code throws if the AE descriptors operational properties are configured to prevent but the deployment descriptor scaleout parameter is +Closing it. +The issue doesnt look to come from Felix fileinstall as if I put braces in the deploy dir path fileinstall is able to deploy the bundle jarSo I think its linked to the way that we load the in the Karaf Main +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +they should be moved +Josh has said he will look at a reflection hack for +Whats the concern with doing soI am not resisting it I do not see a need for it +The file which contains the special characters in the field names is attached as I had mentioned earlier when I am trying to read the field names it gets truncated at +Just downloading the class compiling and exchanging it with the version form the beta should do the job +Committed at subversion revision +configure will do that already +On EAP there is actually different error +So perhaps we need a parameter used by the upgrade whose default value is +Logged In YES useridWiki is a good place for this kind of documentation +Just curious whats in this version that youre looking to benefit from +That is exactly the issue this bug can be marked still occurring in graph shows the behaviour I am of +Closing this issue as it appears to be out of this issue is still relevant please reopen the issue. +You should put this in and you should call client if possible +Bulk close of that have been completed or are now invalid. +thanks Richard! +Jacopo Anil I am taking up this issue +In theory I would classify logging in a critical section as a bug +I need some way of specifying which party group is responsible for an incoming type of Customer requestI have no problem changing it so let me nowConcerning the field name I looked at the other entities and it seems to be the pattern to call this kind of field partyId unless there are more than one partyId in an entity then it is more specifically specified +Now they dont destroycreate on each operation +hamlet +Hope to have it resolved and commited soon +Addressed in another issue +All unit tests passed and tests passed as well +Ill sleep over it and see what I can come up withOleg + unscheduled issues to the next release +Solved +Asked by ESB team to get this in as the important features to get in +Ive noticed that there are still some problems when you select a billing account as a payment option during order checkout Ill try to fix them or at least report them in another issue soonHowever I think that there is something we should really try to improve I think we should move the billing field from the to the entity +stack hes just cleaning up from one of the patches I recently put in +Need to make the buttons nonugly +This same problem would happen with the raw handle as well right? If so I view this as beyond the scope of Curator +Luc can you attach the patch to this JIRA ticket otherwise we cannot accept it +Uncommented cleanup code +therefore I need to test itIf the answer to the above is yes are you able to produce a patch which I can apply to the SVN trunk HEAD? Thank also attached the patch here to satisfy anyone that is saying that I am not contributing the patches in the correct legal manner +It appears to be fixed +Resolving as Wont Fix per the discussion +Committed to trunk +I removed the incorrect release date from our Jira settings +For some reason I can still reproduce +Therefore I found this project interesting for me to work +Implemented this additional information in Rev +Does it refer to Derby or to the extensionPlease clarify if you can +I will take a closerlook at this over the weekend unless someone beats me to it +I saw CLS but thought CFS +Hmm would be nice with a new resolution that said rd party problem etc +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Here is a new patch which adds loose bbox as a parameter for Oracle on ll give the code a try as well +applied +bq +I dont care about reports for a release so there should be a configuration or command line parameter to exclude reporting plugins from the no snapshot dependencies verification. +This looks bad +To check in verified at +Attachment has been added with description Patch to fix the problem +Commit it! Thanks Gagan! +Please help me to review +If you could attach a complete project or a patch to an existin project in the future it will take less time to verify the issue you have + committed as ffdfceebbcbdaabdcf +smile +I have uninstalled and reinstalled the maven plugin but same Linux have Maven installed and I am able to build Jetspeed with Maven from the command line and have no problems +I dont get it +The code change itself is minimal +To facilitate bulk delete operations to the API well add a deleteAll +I have committed this. +I tried building the latest trunk and I got thisBUILD FAILEDCDevelopment The following error occurred while executing this lineCDevelopmentconnectorselasticsearchbui CDevelopmentlibelasticsearch does not existTotal time minute seconds +How should we know that the server isntcapable of handling the presumably invalid redirect it sent usThis change might break things that have worked before thoughnot within the specificationscheersThanks for the clarifications +I will go with a new v abstract class approach and deprecate the old one +So whats left here? Do we just need to move the website and were good to go? Who takes care of that +Could someone pls try this patch? Thanks very much +good have committed this +signed with to next Steve very glad to hear thiswould you ask a fix of too +Verified. +Any other suggestions about how to avoid confusing the unwary +After chatting with Colin offline Ive tried changing the link binsh from bindash to binbash and it works +Corresponding Guvnor Jira has been closed by Toni +The stuff is all done by JAXB and is outside of CXFs code generationIn any case this would be an unsupported use case from a CXF standpoint +Strange +Thanks! +I tried it again with the layer Id been using before and got the crash +looks like this is already merged +Thanks for the code review! Im afraid Im not sure how to address the comments +The issues I mentioned are not theoretical +committed to my fork and filed a on github +We may close this if we cant figure out the specific problem with foreach +There are two components richtoggleControl and richcollapsible +SAMPLE CODE TO RECREATE ISSUE +is not an instance ofIll continue to troubleshoot this one on my own +Attach new patch changesWrite a standalone simple AM to be used in test +The crawldb status is more about the physical status of a URL but at the same time having a new status would probably be cleaner and we can always store any other information in the metadata anywayLets see what others think about adding a new status is a new patch which creates a new db status and a deduplication job which sets the status of a crawldatum to duplicate based on the heuristics described above +I consider this one resolved let me know if that is wrong +I will think about some small rewordings later on today +And you could change the pool implementation using set on a right +Committed revision . +Any reason why a plugin fragment was chosen over a traditional plugin +One thing to note is that the list is just a list of issues to consider for backport so if upon analysis it looks like it is not a bug that would likely be hit or is difficult to backport it is sensible to mark it backport reject and move on +Hemanth you are right for streamingpipes stuff a second identity is needed +svn commit m updated test to allow either exception or not for field access after preDelete +is there any chance you could test this under MRI? I know it might be painful to get oracle working with MRI but it almost seems like an bug rather than an bug +Committed to trunk on revision +I believe this has been done +The index is still preferred as it will save one listStatus call by the front end +I agree with Blake something else must be setting this to a String +tested also in same is an and thus has to be added to the form to participate in form processingYou can easily create your own that validates the value against the converted input of another. +This portal requires a RHN login with the correct entitlements We should try to make this as few clicks as possible so the new profile activated by default makes more sense than having the user update hisher I suspect many Maven do not mess around in that fine if they can avoid it +I like the idea +Otherwise well definitely follow up with Atlassian + Username lleungsun +The expected behavior is you should see the applications main page but instead you see the login screenFollow the steps exactly as I described +of course if you are using such a scheme then wicket serialization is not meant to be usedseems to me like this was a jboss configuration issueif you dont mind please create a wiki page that explains this stuff so it might be helpful to others in the future. +Thanks for providing itThe issue is resolved now please test it and then if it works for you close the issue +If there is no such service or proxy those will simply be broken +Im seeing this with AS Beta HEAD of RHQJopr as of m still seeing this issue w JBAS r r and Jopr the way is now fixed +What platforms did you test the patch on +Implementation of proposed testing I discovered TS with this patch segfaults +Accoeding to the discussion it seems the workaround is to add parameters to +verified by Svetlana +Assign to Svante for review +One way to work around is to check the class whether it is and then cast it to but it seems to me not so friendly and extensible +Closing +due to +tested +Nicholas so what happens if a dies and is removed from datanodeMapIt will still be in the queueCan we implement some kind of iterator of nodes in state by just scanning datanodeMap without storing extra references to +Otherwise the pom must be parsed which can lead to all sorts of nasty things +Bounce to Fred to review and then bounce back to me +Can you guys post a patch we might port to trunk +Heres a version of the patch without the changes +If we change it to portalservicesservicesrepo the repository viewer is shown +Fixed in rev +Fix affected in head +I went through all of the findbugs warnings in YARN and none of them have anything to do with the changes I have made +PATCH patch is applied with minor improvementsThanks +What is needed here? Documentation of the correct way of encrypting the password? Or other +I have established number of users displayed before check user info +Each time a new groupId is requested I not only have to allow it in Nexus but check and migrate the data from the old repoIf your project wants everything remaining moved that would be ideal since I could do it all in one shot instead of piecemeal +a method that spells it synchronization in one sentence but synchronisation in the next +Issue probably related to PAM retry enhancements recently implemented +Moving out of +Yes Ive built a with the fix to which is what were currently using to run our tests +txt is empty +How very lame Appendable throws +I reverted this patch committed from as well to fix test failures +The bz was originally filed against +Which Maven version are you using? I guess the default version comes from Maven +client with threads running simultaneously c +The smooks editor is no longer being developerd so moving to LATER as agreed with +I will have a lookI think I have misunderstood what you said at point and I guess its now fixed with the changes I commited JacquesIve just done some tests with SVN and all the things mentioned in this issue are now fixed so it can now be closedMerciSharan +After updating installed JBDS with Generic servers appeares in the list af New Servers Runtimes +Maybe we should rename the to and add new Methodwhich extend on this named +Please open a new ticket if you feel session invalidation should also be addressed. +Ill commit this +Test passes as well when JDK is used +Im blocked on this as well +You might want to provide a working patch +Its the approach I use in my Tapestry Access Logger package to define what URLs are logged are what arent +The fixes retained on the server side after revert are and. +Do you want do just take care of that or should I set up a new issue +Michi fixed these and is adding C tests which hell probably upload this week +We have always had this manual step and reading the server config would be a departure +Hum Lets open a votation use an heuristic as suggested by Paul or a file listing the mime types worth compressing some other alternative? of headers even? That is a good point though +for all annotate the getters as publicstable setters and factory method as privateunstable +I decided to use the vectorized UDF adaptor for both and +the patch uploaded at PM is the latest now +My preference is Scott I can implement your idea if we have a consensus on sounds good to me too flexible and powerfulEven better when suggested by Jacopo thanks +Id like to do the protocol as avro records +As it is now the onlistchanged event is called twice because if you move an item from the source list to the item list both lists have changed. +Parth +Merged into SVN. +Closing tickets +My pleasure +In case of a timeout the TX will never leave the ResourcesBut Im not sure if its legal to remove it from the resources as the Transaction Manager could make a reference about it againI would need a valid test showing the issue so we can research what would be the implicationThe Warning is probably ignorable anyway +Yeah I dont have this figured out +bq +This looks to be a duplicate of this should be fixed in RC already +Looks just committed this. +Ok thanks Ivo could you also raise a jira against riftsaw with this example +Attaching updated patch +The Sun Java bug has been closed as Closed duplicate of However is not visible +So I will go ahead and do the check and not worry about whether it is bootcreate time and whether we are dealing with territory based db or not +Pushed in dd +So the claim is that errors during a dispatch should be handled differently in particular about the possibility to dispatch again the already dispatched async context +It has been introduced inBtw did you purge your cache before deploying the new version? +From my manual tests it looks like this patch does indeed solve the issue at hand +Please if a test case can be provided +I documented it on the method but the assumption was that the injected zkCluster is not startedPardon me +Attraching the modified test prog I usedI modified it to accept separate counts and do separate runs for the different implementationsFor example and This was to avoid any GC effects from one implementation to another and to avoid hotspot optimizing for one path and then having a different implementation switch to a different pathThe builder also needed that change from setField to addField to be equivalent +Invoked as +Then open the first project in eclipse and in my case the plugin cant download the dependent project +Will commit if tests pass +More fixes +wiki page createdadded Jira in +This is closed since it is a generic issue. +with few modificationsVikas +by better I mean that the test is now more Alexander yes I get that +I am sorry but why this is ClosedWont fix? We can discuss how it should be fixed but the issue is really there as I see it +I updated the patch after the recent changes to the files +Alos we can add run queries on different versions to see the performance gain test is not a typical system test and tries is intended only to test the optimizer changes so I agree with Armys comments that a betterplace seem to be the derbyTestingperfbasiclang package +This is the EBA I used +When the user uses this method and pass the filename it would be printed in the EOF exception thrown if I believe the test cases may not be necessary +TedIn trunk if we take up this patch we need to solve +Resolving +In general we should not multiply inner classes but rather create separate onesInner classes are fine so long as they dont get too bigThis seems like were cluttering the package in order to the file not a great tradeoff +defer all issues to +Mike +However when the producer is started broker log shows a null pointer exception for every message forwarded to broker +Theres about exactly same thing +I tried using the embedded driver and got no autobooting even if I set explicitly in the file as far as I could tell +Why not just add that method into some test utility class such as? The rest of the patch looks good +In my case Ive repeatedly killed and restarted the broker with both sender and receiver running and each seems to recover +Then its more clear where the jar came from +Thanks a lot Arup! +Are you using the Shell component around your pages? is working just fine for me using Tapestry +I guess I should open up a new feature +I edited the file in my local repository and made sure to not update it +Added support for cancelling a remote request after it has been sent +Thank you Chandan +This will indeed be ongoing piece of workand wont be done by Whats the thinking on how to include the work? Robin said Butbefore committing I will wait for the full change to all Jobs so thatcode is not in state +both were added in the app server guide +Catalina is the check for a null database sufficient? I mean if the has not been initialized then your changed methods will now be returning false +Also when you generate the patch you need to run svn diff at the root directory of the hive trunk +Thank you Gunturi Does this include the Hortonworks releaseCharlesSent using +Thus the handler might be called several times for the same +Split out the commandsUpdate the usage +There is a bug in xmpbox that does not manage a unique rdfDescription containing both namespace fields and namespace definition +Rakesh The patch looks good to me +I can now reproduce the problem every time with the attached code sample +Finally run another full build from the root of the checkout worked fine again +Thanks Mac! +committed +A version can be released to deprecate the method appropriately +This issue is specific to templateIn the new template devcdrom is created instead of devcdromThe correct template is create by Abhi +If you could shed any light on how thats meant to work Id appreciate it +minor chnages to the code + +At face value its just a warning that comes up +Hallo againWe have just applied the Spring Version and the same problem is found +In functionality is ubiquitous and a major benefit of Cocoon +The underlying task service however exposes additional features like escalation deadlines etc +They passed +I think we should make the behavior consistent +the plugin has been moved to at indeed the maven repo has one is NOT moved it is copied! Wheres my cluebatShall we delete the dependency from the mojo project +It was most impressive that youve got rid of. +fix to to see if doxygen issue can be resolved +Move issues to next release +released contains the fix. +I am submitting a new patch that is fully verified to Addresses it +Attached a patch for fixing the issue +The bothclear CSS class was eliminated a while ago +This happens every time +zip your code from github and attach to this JIRA ticket +Can you add release notes +This patch solves the bug +I have done this in my code and can provide more information if need be +Forgot to mention along with this request there could be asimilar capability in the +Sorry about thatThanks problem at all Suresh thanks a lot to you for this great effortI will apply your patch some day this weekThanks again +Its good for master joining a cluster w concurrent crashing regionservers +When I removed jar problem has gone +Im doing a bit of cleanup on this +I will temporarily remove from the fix versions so that it does not appear as an unresolved blocker for +bq I have removed all test methods in as most of the filter methods are now called in + +Havent finished all the tests cases. +But Steven fixedWell I fixed the problem but not the problemSo while the case order you see in sources generated with current trunk will be different from previously checked in code it should be stable from now onIll work on tabs in generated code today Erick so could you hold off a little before checking in +the same situation in all supported as described in bug I cant reproduce in trunk +Uploading a new patch +What you propose sounds good to meThanks +If so I think that would explain why all of the passivated XA transactions that Arjuna keeps retrying are in the FINISHOK state +Henryyour patch on isnt correct +comments on my comments to Phabricator +So I file a separated jira to track this +The attached version of the converter should correctly generate default flows for Gatways +Does not seem to be fixed in +You are right it should be fixed +attached a first patch to discuss +For example I would like column families that meet a certain criteria +Tried to use getType from which can be obtained from but that shows the transporturi not really the bindingtype I need +I saw this behavior in a stress test just now as well on tip of Since we know meta splitting doesnt work can we put in some code that allows METAs one region to grow without bound regardless of its size +add compile line to description +We need to at least have a direction by then +I think this issue is now resolved but lets wait a day or two before closing it just in case something pops up +This bug is reproduced with any other JSF tag that placed inside specified tags not only with fverbatim +The patch looks great for me +integrity of checkpointed dataSplit the +But in this case does it suppose to return all the applications +Hi gentlemenAs as an experienced Java developer but novice opensource contributor I still have a bit doubt about how to startI picked out this issue in order to familiarize myself with the the process and to understand mechanics of a commitI wonder therefore if the issue needs to be assigned to me before I start working on itCheersMiomir Boljanovic +Sure that would be better +if tests pass +His explanation of why it saved memory also makes sense +I think this makes sense now +now there is some minor memory issues in xatmi tests +Thank you for your checkingOK +After applying this patch and doing an ant clobber followed by an ant all I no longer see the and suspect messages +Attachment has been added with description Partial fix patching the should now be fixed in revision Thanks for signalling this bug +And somewhere needs to be a AS Deployer right? Is it only targeted for jbpm ? Then it couldnt go into ESB already rightCan you quickly explain? Thanks a lot for sheding light on this +Jaikiran some of this stuff we need to expose trough undertow subsystem yeah some of it should be done by overlays +See pictures +Thanks Robert Ill commit once hudson s it +This is completely untested all I know is it compiles for me +One small change to to move import out of the munged imports +After this issue is implemented this issue can be that should implement all previous comments +hi jesse i think i know why you cant reproduce the bugIis it possible you have true set? Caching must be enabled to reproduce this buggreetings caching on did the trick thanks +does not switch to micro instance +I am ok to get it in +But I wonder why the distance is +The key here is not the start level but that the framework state is STOPPING and therefore File Install should do no management at all at that time +Reopened and uploaded a patch for trunk +Not entirely surprising +Here is a patch I created with Eclipse +Convert jersey to and use for +Im about any of the possible approaches to this issue so Im leaving theissue as NEW and deferring to the input other Commons Net developers and users +Guys lets change the tone of this issue OKRadim thanks for your patch +Here is a patch that I think addresses the issueAttaching patch to synchronize on applied in svn revision +I think the remaining issues like handling the Wave feature are too deep to address in the next release so punting to this one further down the road as its not a high to next down to +Shouldnt conceal exceptions impossible to test. +Is this the same issue or should I open a new one +push out to +This behavior is also on iPad and Motorola Xoom devices +My contacts in Beijing also attempted to reproduce the problem and could not +Ive allowed only news to have a in their id field because they may reference classes in libraries directly +When a bean is registered in there is a single instanceof check there and the registration happens in a nice encapsulated way +Tests appear to be okay seems to be flaky it passed when I ran it +Providing patch +Ahmed +job here in Holland after days coming here from Munich +JoeThats a good question +Oh I didnt notice this before but the test file has t tab characters in it +Tests are removed pending another JIRA +for this. +Sorry for invading this ticket but the re factor of the file consumer needed to be done in one go as to have most unit test still passing +Why do we need a Type parameter for the annotation? Isnt that clear from the annotated entity +Maybe it would be best at this stage to get them documented as I believe Ferdy is going away for a while just now +Patch addresses New extractor that leverages Apache Tikas ability to automatically extract metadata from a file based on its mimetype +Patch looks good I havent tested it with any JME vm though +Thanks for the comment BrianReading back since it will be in SOA documentation that should be enough as that is the only place it applies +Committed backport to +Applied both pull requests +Can you verify that the sources in CVS will build using? Thanks +Thanks Stefan for clarification +I know +no processing logic was changed +Modified to use threads as opposed to the +this patch uses sysctl to read the max possible value and sets it +Sorry for not being clear +verified at revision +This patch was made against +we are getting really different behaviors here +According to Ronald Spierenburg this issue doesnt apply anymore +Fix has been applied to branch remotingx and unit test passes on cruisecontrol. +Ok +now limittrue on action component from allows to avoid this +Committed to trunk +I created for the same problem +This method is deprecated in but is highly unlikely to be removed +The for the patch Yang Committed at r +If youd like the repro with the client please email me +We can hack around the issue by creating symlinks to the correct jar but we need to drop the version string anyway to avoid confusion +I tried your changed classes above +Committed +Thanks! +It sounds like end of August or early September is a reasonable estimate for when will be released +Ok great! So lets determine what is the cause of the problem when disabling network connection +Here is a patch to add some deprecation support in and deprecated config +The camel consumer throws an exception as expected when a http endpoint is invoked from http +I dont think it is Bug I tested it also with xalan with the same resultsBut seems to be in version It lies in the fact that I lazily create a dom treeBut Xalan seems to always convert the whole to some internale DTM format +Ill add oneregards Willemto write a test case using I need to add an option in to customize the bundle directives of the generated test bundle +Bulk move issues to and +They were Avro related or failed with +this is something Nathan already fixed +Looks like an byproduct of the conversions as this should not be thedefault behavior +Maybe because noone knows? joking aside moved this issue to jboss tools since it is not JBDS sure if we actually can deduce this hopefully the we look at can tell +can you compare your pom to s pom and see how it differs +So this test case will now work even with foreign key columns +More complicated logic might include considerations about JIRAs states +Hi RossenThank you tremendously for the update and for clarifying it +fixed for javaxxmlstream is the start errorin JDK +Backported to branch +sorry + + +To reproduce Run a JON server with the SOA plugin installed monitoring a SOA can you confirm these changes are in the SOA docs? They made it into the ESB release +First draft of the coversion of these testsPlease reviewUgo +If your allowing the factory to create instances if it wants to then I would think that passing the statement id would be usefull to identify when to instantiate an object and when not toPersonnally I would never have a factory that behaves in this manner +I can tell that the traceDictionarys entry points to the positions in the vector where the words weight is stored but Im not sure where to go from there +Done +Edited +As per previous comment this no longer needs to be solved. +On a big enough table it takes hours to run and by the time its done data might have been inserted already and other flushes might have run +Hmmm sorry it took so long +Ok so +This is duplicate bug +ve tuned down my caches to not deal with and swapping + +I was looking at older branch +Cant yet merge the pisoa features into an existing category but might have a workaround for that using a custom +Hmm you are right +the current code is troublesome in many ways +But null isnt +Hope those are okay to too +I added to the class new test methods called testRfc based on RFC that show our inconsistent handling of CR LF at the end of data +yep +Applied +Manually tested +Working as of rev +Great catch +I think things are better than stated but I need to think about this more +No feedback but issue shouldnt be on hold. +No user method should accept a parameter so we would not need to worry about method signature collisions +Correcting component and reducing priority to commited +Ill ask around and see +First working version of a Maven POM for Commons VFS +i think this is fixed as part another JIRA +Just as a follow up since I saw this issue again due to some JIRA change emails +committed +The issue was caused by the same problem with its a wtp bug if WTP dont want to fix we should find out another way to solve it +im working on it right nowand emmanuel is working on more annotations support so and friends will change somewhat radically the next couple of daysI plan on having the shift from tool class to delegate class committed before tuesday +tag youre it. +To find an optimal setting automatically we need to do experiments with different settings on various clusters +The references to are project references to the trunk projectIf you have the chance check them out and provide feedback + +Also would be a useful for this +That file can be replaced by our own version with multilanguage support the same way I replace the other files in Forrest +The end attribute in supposed to be inclusive but it was behaving as exclusive +Marking as fixed +Ill keep looking +patch v also fix a minor bug for metrics system shutdown in mode +We can run a job to cleanup these +is still bytes after this run so I agree being bytes should not be related to this problem +This is a new implementation of +Issue is a change to that approach +Another would be to have a callback that returns the entire +Oh nevermind maybe what you have there is actually all we need +The repositories should be removed from the model in +Aha problem was an old mvelDrools uses mvel I had mvel in my dependency tree. +What is strange in your log is that last message is The syntax of the command is incorrect which causes me to think that problem not in a Java code +These changes were committed into SVN trunk with r and into the branch with r +Done +In general it should not be too difficult to add the requested behaviour +Provide a stacktrace your configuration environment and anything useful to understand the issue. +I ran the tests that were broken under nunit and like before they were green on my machine but there could be a factor that Im not accounting for or missing +The patched source code has been committed +Would you like to post use case for us to reproduce and provide more numbers +Is it possible we can improve tests somehow in this area? Im open to suggestsion but ive created to track it since weve already good a well vetted test for this specific problem commited and on track for a respinIf other good tests pop up in before final then we can certainly commit them but its not an area im likely to be focusing much attention right now +This patch has added two new files +see also disappeared +CQL is not meant to be a bulk load protocol +Not sure assigning it back to Dominik to verify +validate +If the Hudson test run looks ok Ill close this issue +This does not seem to be the case and we will have to review why this is the case +I will however work on a minimal test case that exposes this issue and post ll update this thread Tomif you use a JVM try with the option noverify +I see that you are using two properties that are known to cause headaches +fix checked issue fixed a long time ago +This is because a downloadable package must not contain path valuesThe same problem affects the resolution of which would need to add path names to the distribution jars to bring in the Tuscany SCA runtime dependencyIn summary it isnt possible to achieve all of the following goals +This class has a class loader of +Like yours there are cases that dont need sharing variables +Fixed the hedwig server compile +Indeed Daniel pointed out the code changes werent necessary as this optimisation is done by a different Rule +So rescheduling from. +After looking a bit more it seems that the Observer doesnt write the last transactions in the transaction log and that is the reason for not being when it restartsIf the Observer is started using the transaction log of a Follower then it worksAnybody can tell why is it that the Observer doesnt write those transactions +I will post a new version with these problems resolved. +Applied in r +In that case this JIRA will help in the testing of methods not included inSounds good also feel free to move and under test if you like but not in we can decorate them as deprecated there +But if the size we choose along with other stuff going on in that particular JVM cause memory to come close to max the application will most decidely not run faster it will run far slower +I recommend closing this bug as fixed +Dhruba The API does not restrict the application to read in a given order +Deferring to +And if there is no file given then let uimaFIT find it in the classpath +Hi Jacques jep thanks +Added documentation for cascaded attributes +Please assign back to me when done for addition of other changes +Anyway I see no reason to specifically remove jonas at this point but we can always revisit it at a later date +Good point +It will also be good to emit JMX about how many tombstone we are going over +The issue is resolved as described if you want to continue the discussion about the overloaded method can we do it offline +Resolving as Wont fix +Sadly Im not adept at hte Maven stuff but if I had a project with easy instructions on how to import it and debug I could figure this out Im sure +I will post this list on Monday +The only reference I could find was contribscriptingrubysrcmainresourcesNOTICEreferring to the licensing of the library +Closing as fixed since no one disagrees. +bq +From the log I can see the server keeps retrying DB operations regardless of the shutdown signal +Very glad to see this request finally goes into release in a month +patch looks good to me +Portal identity should be moved to use JAASFor Identity should use JAAS domain nameFor liferay JAAS realm looks like +Merci +A sample template file called is added to the jca directory. +I stepped through the test and found addapplication wouldnt execute twice +On hold until after translator and grammar changes are complete +I think should fix this +Hi Janplease verify for on windows x x +Do you mean a servlet running on each of the tasktrackers? What I meant in my last comment was running a job that does exec tail or any other functionality in parallel with another job already running would be problematic since we have issues with two jobs running in parallel +Thanks for the help ManfredI think we need now Angelas input on the failing tests to see if there are still issues in the fix or not +Eclipse is by default configured for Java +Ive run valgrind against the decaf unit test code and corrected several memory leaks in decaf and in its tests +Alexei Ive checked patch on Ubuntu Linux and Ive received such result as on RI resizing in all coordinates and in demo cube is drawing in the center of the appletviewer frame +Have you tried the same upload without Knox in the mix +The issue is that switching to Netty is rather wholesale all of the existing code has to change away from using to use the Netty Channels +Im afraid I dont know enough about the Struts architecture to be able to patch the actual problem +Fix committed +So you are thinking this cannot be completed until after we add delayed clean up of files? Only then can we safely remove synchronizationsHow many threads we talking anyways? It should be uncontended +. +Coming up with more test casesPlease Werner or some one review this code and let me know feedbacks +For example in the when the RF per datacenter is specified its easier to enforce that the total RF is the same as the sum of the RF +A couple of examples included here + for this enhancement +Why zip when you can patchPatch is against. +Prashanth this needs more work +I added missing licenses to java and files +are welcome to move this issue where you want +The multithreaded test test fails intermittently allowing the database to be booted twice +But the content of the Libraries service is correct +Please could you verify if this problem is fixed and let me know that I can close this Jira +Is it possible the automatic index statistics update has corrected the issue +This the error message I get when I click the search are trying to move this out of scope to Milestone we no longer have time to finish this given the lack of available WRS services +is working well so I suggest a similar one for +Here is the top object capture after implementing a cache for +Tyvm. +as long as its doable on the command line I am cool with itSo all should be fine. +Each builder needs to set this field for gbeans that run ee appsSo in particular the jsp and the have fields where this artifact can be configured to the appropriate pluginIt might be better to go directly to the bundle with the required classes but I dont know of a good way to get from an geronimo artifactId to an osgi bundleI might not have understood your suggestion +Can you please confirm if this is code issue or not and also let me know when this would be fixed approximately +I think this was a bad case of Monday morning +Ill try to review soon +Thanks for addressing the formatting +Please close it. +Three issues were seen when fixing this issuea Invalid path of primary storage is resulting in NPEb No error dialog is showing up in GUI because the storage management layer isnt propagating the exception upwardc If the KVM cloud agent on the KVM host is down and an attempt is made to add primary storage whether the storage path is valid or not the primary storage is failing to be added as expected but is getting entered into the database in the storagepool table so it shows up in the GUI as wellWorking on the fixes for these +Thought of something else if were storing the full mutation why add the complexity of hint headers and forwarding? Can we just make the coordinator responsible for all hints instead +Is that a valid test? Created a new drive by doing subst q ctemp Copied a Jboss install into q when to qjbossbin and ran Jboss it started Installed JBN JBN other servers I had running including the JBN Server and Agent themselves but not the Jboss server running in Q +And I dont see this work going much further so am closing this for now +Hi Klaus thanks for debugging this +thanks for your help Michael is in staging repo as we speak +Maybe we even make it an alternative for a real deployment as well as tests +This just bit me too +I added limited support for primitive arrays and String arrays in Git ID thanky for the reply +Wrote a patch for this bug as well as what would be a similar bug in the putAll method +has been marked as a duplicate of this bug +corrected in seam guide the guide is checked +This is an artifact of +so need your help +I am unable to reproduce your error +This ticket is stalled because it requires a new Derby feature to resolve +in +If youre not using youll have to build binaries with your version of the compiler +Fixed for IE and Firefox +this was run in version via the CLI I dont think that it makes a diff but I wanted to mention just in case +Good to know let me check it out. +Ivan I also put the message header in when implementing the filter +This is deployed in Yahoo and also CDH releases use this change +makes does not recommend unequal forkjoin pairs +Closing all resolved tickets from or older +Hi Murray I think what needs to be done is to add your persistence file to the source repo run all the tests using Ingres DB +An AS project Blocker means a release or tag should not happen until the issue is fixed so a Blocker must have a Fix Version assigned. +Patch has been committed +The initial reason we have introduced job task attempt ids is that folks should not in any case parse the strings and we can make changes to the format wo breaking compatibility +IT added +Having to deal with multiple threads heartbeating adds a lot of confusion and complexity +OK fine by me +I may be misunderstanding the problem but if this is exclusive to it is correct beahviour +The name of the files to search for can be specified by the user +fixed in r new +Ill make that priority this weekend +thanks but could you generate your patch from the top directoryIts more convenient for me to apply it with a better index +on commit +Clone URL git to +will probably have to be rerolled to work with these +selecting Cancel everything is Restored to the default +incomplete partial solution just fixing some broken links +This is a good idea Eugene and I was unaware of this feature but even after doing this I still only see version +Why would checksum ever be nullWhen reading old versions of fsedits that do not support checksum +At this point in time is latest version +xCommitted revision +Does it work properly in with this patch +Will commit if tests pass +to this idea +Alek Chris I think this bug can be resolved now? +after release +The client side iterator scanner can be used to run iterators on the client sideThis would also be useful for testing changes to iterators +Lars was talking about dependency for tables and system tables in general +we cant expect whether new bugs would be brought +now I can reproduce your problem but Im not sure about the solution explanationIll be back on this issue Joachim +Tomorrow Ill make a license section so we can include all the additional licenses we need +If there is a redirect you wont repaint the form but the result of the redirected page +Will reopen if needed +Which version of ADSPlease give us some information that can help us to determinate if its a real bug or something which is already fixed +Added space as a delimiter character so table product type is reverse engineered as +Assigning to myself +Using it +FYI +ping +Derbyall and ran cleanly on Sun JDK +removed fix I wont test with SeamJust and Trinidad and Facelets test case is not Seam dependend +Sounds good to me Jonathan +Plus one of the two classes needs to go +transition to closed remaining resolved bugs +v fixes above and import cleanup in +And wanted to get some code in place right away so I could get feedback and I did thanks a lot +Binary artifacts are not the same as source artifacts in that there is wide variation in the types of artifacts the environments they are targeted at and at the user expectations of how things are normally done on those environmentsTake the suggestion that a windows +Silly me +Micha can if theres actually an issueFor Michas note This is originally discussing see +Closing all resolved tickets from or older +Note that Update Dependencies is now a separate action that can be called from maven popup menu +bq +I thought I have tested though. +Sure do you know how to figure out what svn that is + +Put a due date on this issue as a benchmark for completion +In this case we should just describe this pattern in the documentation so that people know how to do it the right way +attached is a patch against branch which i just port from a versionI havent done a run against the latest branch yet but it pass the whole case set in a branch +Julien on the order by desc generates an important overhead +fixed for windowing +Seems to be some OS specific problem +Please note though that you may have to run grails clean before the change comes into effect since there might be old class files lingering on +I tried to bisect this issue and went as far back as without finding a successful version +I apologize but i havent been able to reproduce this +Im not convinced about the need for this but Ill wait and see if anyone else votes for it. +Is there a valid reason for the close not to complete cleanly +So it probably should not be closed but progress on this will likely be slow unless other hands help out. +suspect that there may be longer term issues with this general approach but i think we should accept that the current proposal is good enough for this release +Then let it bubble up into Solr and e +I just committed this. +However since there are still a fair number of technical professionals that use IE I figure it is still important to try and offer the same user experience for at least the major how often does Xircles gets deployed? Assuming that the intent for Xircles is to be the replacement for Codehaus if I discover other cross browser issues should I continue to post here? What priority should I use +Looks simple enough and shouldnt be a pain to support +Like hardsoft linking in the archive +refactored versions of the symlink tests to use this interface instead +I think storing it in the splitter classes is a good idea +Introduced a setting to control this +Attaching another update to the patch since the previous didnt handle null classifiers guess its a duplicate of. +I havent used the test case with the patch but I expect that it will be similar to the number of reads for the locking +just verified my last comment on mac on installed JBDS and it works just need to verify that on clean installation to make sure there is no problems with first in trunk and branch +As long as a new dependency is not introduced I think copying reusing and licensing what did is the best solution. +This issue has been inactive for more than years +txt in it +There is no sorting going on here +Is this or +Need to move on +Changed Target Version to upon release of +Hello Eric This doesnt feel right +The codedesign makes sense to me and we know the current path is broken and weve verified the fix +In the code used both Level and PriorityThe reason for this was that it was thought that logj was likely to change Level so it no longer extended PriorityHowever never really happened was only released as alpha code. +I just wanted to highlight that relying on isnt desirable given its limitations so if you were in a position to change your to follow the same patterns as presently seen in Roo master youd have a faster more robust and more reliable model +Patch +This can be done by executingbuild within bin directory of your castor project +this is definitely useful +This requires a way of knowing whether the bean +As I have mentioned earlier I will be discussing this with the team to come to a general consensus on attachment support in the WS layer +If any of the steps of the ingestion process get changed they should by default work as they do now to support backwards compatibility +For those that do the consumer must remain active until the TX is committed or rolled back +bulk defer of unresolved bugs to +You can find the interface and implementation in the attached modelIf you run the test case you will note that properties which are annotated explicitly do not get written +Thanks Stephen for confirming this issue fixed in latest. +There are no configurable log levels though logging always happens at trace level +The resolution has been accepted by the board +bulk close of all resolved issues. +Resolves Janitor race conditions +Committed to branch and trunk. +check if we upgraded to fix this +That might actually work in all cases I guess Ill give that a through a starts +The problem is that the corrupted can not be read +But there are some problems with input elements editing +Patch looks good +Does this break on any other device running +I think its ready for commit +Ill apply this as soon as I can though +Yes there is no state stored on the thread we are not caching any part of the id we use a stringBuilder to create the id and then immediately call toString so this change is just to avoid allocating new instances +OK I was able to blow the log as wellDoes changing the log level to Error with tomcatw helpsI was not able to replicate that again using the error log level so we might have some buffer overrun in log +Verified on SVN Revision . +Patch addressing Sureshs issues +Now we only drain until we have evicted enoughI would say this should be the correct behavior +pushed with commit for this it was missing an end can you see if the attached works +Thanks + +Strongly second comment from Bill Bell +I have a use case for controlling the replication targets of a block based on filename at file creation time is it possible to accommodate this in your proposed API + thanks. +The test functionality for derbyall doesnt complete +Lets reopen itFor this failure the region is closed but not removed from the online region list since something went wrong in updating the zk +Ill just work on the missing dependency tree then after that the used by tab +Done in works fine for me on my sample project +The separate tags for fields approach seems to be implemented now +rather the Spring test suite now passes successfully using both the JDK default and the impementation +Screenshot showing several scroll like an issue +Need to make sure this isnt an issue for in r. +Lets squash this bug! +ETA for dns cutover is in days +All regression tests ran cleanly with the patch +I guess I skipped the part about bugs in the guide +Records can change while querying subsequent pages +to TRUNK and BRANCH +Patch looks good +If a safe and efficient patch is submitted to enable this enhancement we can include it in +Also +I will try to invest some time in this in the near future but I need to upgrade to the latest SVN first and some folks advised leaving this a little bit longer until things settle downCheers m a bit confused by the INCORPORATING ISSUE thing +i have proceeded with the same steps but then i get the same error when running +This is definitely the moment to remove any ambiguity or confusion in this API +batch update of issues with fix instead of the correct +The annotations mentioned above are a result of that process +Attachment has been added with description demonstration has been added with description demonstration file should be fixedThe wrapper whould now be able to contain text directly. +Richard I was just having a chat with Ondrej and he said the test pass Ok now +thanks a lot +Improvements in Euclidean distance measure to use v to iterate and v to one had a correctness problem in manhattan +Looks like a good candidate for as its truly private and only static methods access it +Attachment has been added with description An example of output when list body is blank +The affinity will be maintained by the cookie on the client and modcluster on the proxy will direct client to the correct backend server. +Does anyone willing to provide test project and exact steps to reproduce this issue +bulk close of all resolved issues prior to release. +As the map seems to be used for nothing else I think one magic hash would do the job just as wellIn addition empty directories are not removed making the traversal unnecessarily expensive +Applied with svn +Have you tried to rename it to +were using the Oracle drivers which are present in +Java separates the scopes of anonymous classes instances from the creating scope so Groovy is behaving very differently from Java regarding scoping +Please submit patch again when you finish rebasing +This wont be a part of Hadoop. +Problem was in customized FTL where javascript was is committed in QA and modified in POC +Im not sure this is really an issue + Are unenhanced classes the only trigger for running through this below mentioned code? Yes +Updated patch uses instead of for caching instances renamed methods to make it simpler added Alejandro +Next up refactoring the packaging to avoid such silly names +You can usetiffcp t c lzw to generate a suitable tiled predicted image +Verified struts is removed from +We only need a place to hold the reference to the configured +I was able to run the walkthrough by literally typing backwards but not exactly +Verified by Vladimir and Tony. +I have now incorporated the changes as requested and have attached +bulk defer of featurestaskswishes from to +I agree completely +NOTE The duplicate Release Notes entry was noted in +Fix has been applied to branch remotingx. +Good catch +The scrubs will all complete but all of the info from the scrubs will stick around in compactionstats +Verified with +Henk theres a bit of issue with our website I suppose +Good work +If not please hold off until after M +I will be looking at moving to H as the default in the near futureMeanwhile I committed some changes just yesterday to account for the failures we saw in the existing tests in the hudson runs + +Closing the issue +Not from what I saw skimming the uses +We did not wanted to create that dependency in deployment chain +While Derby already has some java objects in its system tables I think we should discourage adding new java objects to catalogs +No patch unscheduling from +OK the +And the code is very similar to mail component +so i guess you meant writing wiki as an independent servlet +This seems to be a low priority item so it is being put on the back burner +Is this assumption safe enough +The module compilation scripts seem to be working fine now +Here is a candidate implementation of the improvement +Hi SergeyGot it lets start from that +it was done a long time ago +When I read the comments in the code I understand that were supposed to keep the +Hi Would you please apply my patch? ThanksBest regards AndrewRegression test applied to SQL module at repo revision rPlease check that it was applied as you expected +Attachment has been added with description preliminary patch fixing the P open bugs to P pending further status from ASSIGNED to NEW for consistency +I think this kind of functionality is usually referred to as proxy chaining +A merge strategy is highly dependent on the application at hand thats why we dont provide a strategy +The duplicate versions will be removed when we switch to the stable build environment +Thanks Thomas and Andrew +This fix isnt in the release perhaps its on the branch this is in branch but not release. +For your reference Ive deleted xdocs file and xdocsjapanese directoryI delete xdocsjaJP and added xdocsjaJP +This seems to have been resolved by that job didnt have at allIve added it lets see +Could I retry now with trunk or is this only fixed inTrunk should also work fine now +Gump has good locking so its easier to implement there +upgraded to on EAP ER +It makes sense to add this workaround to the buildlist task documentation +Ive attached a screenshot +IMO it would be a mistake to try to keep in sync with milestone builds +Ill change the behaviour of get because DRDA protocol is lacking these features and allows specifying the transaction timeout in SYNCCTL codepoint corresponding to the function call +Seems like a optimization +Thanks +Theres a solution in attached file that doesnt degrade performance and satisfies the feature requestPlease have a look +is part ofI am facing some problems while implementing this patch in +to trunk +Good point +Heres a second version of my previous patch that keeps Collections of documents together instead of wrapping each one in an add document operation +Marking as Open until further patches are created +This patch will therefore delegate the responsibility for handling special request and response objects to the bridge +in this case if the data is not a number a lot of will be thrown +Committed to all branches. +It should be of course +Itd be preferable to allow any class to subscribe to certain message types from and then we could add other cool consumers of the event data +Im reopening this as the work done for actually cover this too +Patch does not include the deletes for the old activities adapter +Thanks JrgFixed. +Ive corrected the test and also provided a fix +I have simplified the code and done some modifications +As a I upgraded my web services from the to and have seen a much better behaved application server +This is not something that was recently introduced +and it would be cool if you can post the patch for the first part soon so that I can try it out +Automatically loading out of libnativemap was provided for but the intent of the separate native module was that it would contain all native code for Accumulo +Thanks Ted for the reviews! +Admin job also figures out the branch name of a patch from the patch name by comparing it against the svn branches +It throws an exception only if the location is empty +That piece of code remained there beucase of recent changes in trunk +Ive the same error +Do you think it is possible for us to get an installation for testing? Or could you provide us an installation available via InternetRegardsStefan +Then its easier to spot if there is many different Thanks for the patch. +Pending resolution +I am also very sure that the current is nowadays different than what the demo returns it is not identical anymore and improved quite a lotWe dont use it BP anymore but im happy to commit whenever is in maven or part of Tika if it gets donated to the ASF +It even works in a fresh application over a second guess I tried to iterate over an old object stored in the session context when I got the this is not the right place but I have to say that I really love working with Grails +to our Eclipse tested it out this was the same issue with Eclipse. +Cool thanks Daniel were actually cool with either updating it to effc or flipping the write bit and well push the delta ourselves whatevers the fastest to get to writeableThanks Ive flipped the write bit on the assumption that where it stands now is acceptable for you to work from +Always glad to see more people helping out with the community +committed +Note that adding a new library like guava here will break usersdownstream components who dont expect guava to be on their classpath eg because theyre already using a different version of guava +Fixed +Thanks Pranay for your contributionYour patch is in trunk rev Ashish +I just committed this +I have used version for the test +Looking at your issue it says only works w websphere jdk stock does not have classCan you be more specific as to what JDK is supported? I will have to tell the customer to move down to and want to make sure I tell him the right one to get +Might make more sense to use the dropdown and remove dependency on Igor Vaynbergs select component. +Proposed SOA TP Release Note now performs a correct lookup by name +That is the case with any other open source projects implementing RWN +Have you run a test which shows this change actually improving RSS usage of the DN? I think most folks are running these days so the address space savings doesnt seem too critical to me +Im pretty busy at work nowadays as you may guessDo you have some planned timelines for Cargo to be ready +website can do both add the issue ID and add the whole link +I have a similar issue I am dragging a value which is dropped on a node +Ok so this issue doesnt exist in +See for other Sportsclub and Snowdrop. +Good grief +This comment is not particularly substantive I just want to express my enthusiasm for this behavior being fixed +Been a couple years since I wrote much ruby so would appreciate a review and pointer to how to write a unit test if applicable but this seems to fix the issue for me +test for new functionality is only difference with previous +Re why are used at all I could be mistaken but I dont think you can publish to a remote maven repo with ivy +Im far from that Oracle instance and cant check right now +Committed to the trunk as Subversion revision +Observed in the AS testsuite +Ok I see but Im not sure I agree with the change +Then if I run the eclipse executable directly it starts +I used rank and select because of the titles of sections and in the article but the method names do less than what is described in these sectionsThe method names should have been bitCount and selectIll add a correction for this at there is no need to reopen this issue. +You still will not be able to change types for specific transformers but most of the refresh issues should be fixed and the less than optimal removeadd workflow should actually work +The reason i havent committed the patch is because there doesnt seem to yet be consensus on what to do for RE there are several s and no s Ryans vote was +This bug has been marked as a duplicate of +What are the gains for complication +Such ifndef declarations are already present in the AXISEXPORT there just lies outside the ifndef +It should just work out of the box +Manually tested +I will update the documentation once the code is committed +Patch committed to trunk. +and its not even clear the queue does anything unless one is somehow producing documents in a very bursty fashionStreaming rather than buffering is definitely the way to go +looks good +Existing tests ensure this change doesnt break anything +OK just ping when the svn history is ready to be mirrored +Fix is merged into branch under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Hey Nick thanks for the commitIn terms of release note here is what I proposeAdded a tool to generate a report of API compatibility between different versions of hbase +Both servers forget the square brackets when inserting the IP address but only one Uses unknown instead of the actual host name it looked upThe header inserted by Verizons mail server mentions SMTP server JY +We should not even allow deleting a zone when there are secondary storage and cache storage existing for the zone this is a regression from our new schema change by removing secondary storage from hosts table +Im going to test how it works with the new infra so I want to keep this issue open for a while + unscheduled issues to the next release +lgtm patch looks correct +The bug remains open until resizable VT pool is implemented +Thanks Vladimir the patch was applied to BTI branch at r +I dont think we should do this +methods and fields are really only required for classes that actually are public but that wish to hide some things from all but others in their packageThis is a fine point and doesnt warrant too much analysis +Ive moved it into the reduce and it works properly now +Including Apache on the above for SalidaApache PiletraApache Kleber +There was some mistake by Galder I believe that he didnt upload the artifacts as they are in source maven it should be fixed also in Seam? but why does EAP contains a seam that is different than seam? For now this issue is fixed but would like to know why such changes even makes it into EAP before aligning with the tools and seam versions +defer to +Hi Jacques I do not think think this is easy as I mentioned before look at the changes Andy did in revision if we could backport that? I am not sure +MaurizioI applied a fix for the problem to the SVN trunk + +Is that an yes for deleting the file? Thanks +I just needed to REMOVE BOOLEAN and BOOLEAN and it figured itout like its supposed too + +Ignore me then +closing for release +Please reopen if it happens again +on patchWas stuck trying to fix the testmultiparallel test +You can have a uniform config but should beable to correct it for one or two nodes that are different from everything elseJohan I think this is goodPlease replace by USABLEDISKPCTDEFAULT after that +When you replaced the mapping for you hit a bug a mapping of a class that inherits from a class ran into troubleSimiliar pickles would result from mapping a class that inherited from say IntegerSee rtdatabindingaegissrctestjavaorgapachecxfaegiscustom for the code that I set up to test this in a CXF contextI continue to suspect that youd be happiest just using the default mapping instead of forcing CXF to treat Calendar as a bean +I dont think It suits +Thanks for the patience +Saves one call in an object that matters +I believe this to be complete at this point. +Merged to the branch in revision . +Since a modification done months ago operation is run out of UI thread which causes the exception +Patch to fix tick timeout +lock file it can issue a warning that the repository was not shutdown properly +I have attached the patch file which is created for the test itestcontribution with comments +Thanks Mahadev! +Where are you committing you changes? If not already please create new location in internal SVN and when I am fully back from sickness we will discuss proper location in main code base. +Instead of interrupt using boolean flags to exit from the threadAlso removed the last inMemory merge from fetchOutputs +No response assuming ok. +Thanks for review Enis. +Another example might be not asserting the exceptions message. +This plan sounds good to me +I would expect thatthese new rows are all full and thus dont go through the tricky code +Dont you think a property should either be accessible both with property and method syntax or with neither of them? Everything else will be very hard to understand +patch was developed on top of Indigo Java EE IDE with installed requirements +This issue has been resolvedPlease open a new JIRA and attach the patch thereThanks +Fixed on trunk. +Maybe somebody else is capable of identifying the problem + should be +Since the version of the current config is now instead of the version of a new config should be something larger than . +Will post the corrected one soon +client log +Some minor nitpickings In variable scheduler is never used +Just let me know +Attached patch solves this issueRegards for review to from M so as we dont hold up the main releaseWill do a MC release soon after M +I applied the patch. +Scheduled for +But in a multimodule build some artifacts such as an attached an attached artifact are not resolved from the wars dependency list +Are you considering lifting this limit for? If not I think it would be good to give an error if an attempt is made to create a too long user id +What should we do about this issue then +Documentation missing +verified by Vladimir. +Reviewed. +Therefore some options for terminology Portlet applications Portal applications prefer as I think that is most accurate to what these quickstarts are showing but Ill let someone with more domain expertise chime in to be sure +Completed At revision Completed At revision +Cool I think were getting at something workable +Hi DanielJust confirming that from the th March appears to have resolved this issueThanks again for your helpRegardsMike +To do a full implementation of your changes which is the way to go IMO requires a change in the thrift API +Thanks Ivan for the hint +It worked great thanks for the fix +Added xsd to the example messages in the purchasing model + if tests pass +The only references I can find to any wsdl files are from code in or +Sure will do +I think the teams would be interested in it if Seam thinks its out of scope since it doesnt integrate with the framework itself +splitted the patch into pieces since has been moved to you please clarify which patches apply and which do here is a new done r for trunk r for branch +can you post the debug logs from a failure as before +Cloudera may support but it has to be able to work on EC +David could you clarrify or answer what Oleg is asking? Is the element not suffient +Youve mixed up +Sorry I got careless and forgot to copy my latest draft into my trunk area with those last two sentences removedIm attaching and with the changes I intended +All I did was apply it to a source file and upload the source file and the jar for others convenience +From our testing using the Default lob handler is working for have seen this +What is your proposal to fix that +Bundle with VPE installed works fineThe only problem is that jsfrf tags are not displayed correctly due to the fact that their plugins are not istalled +And I agree with Daniel and Sylvain that serializing query cache that invalidate on update wont be very useful in most cases +Max Andersen may be able to schedule someone from the JBDS team to make this assessment +Do you pass Tests Directory to Sonar +The other test failure is not related +The uses so presumably it could as well +This should be fixed in maven you try to either put your JDK in a place without space andor upgrade to in your maven install +Parser recovery is now available in this situation. + unscheduled issues to the next release +Merged the patch to +regrettably as youll see in the release notes the new wagon wont work with +Verified on + +Sorry youre right about the not sure what I was thinking +So I propose that for nowhave different versions of Cell and one that keeps track of the size and onethat doesnt so the overhead of keeping the size doesnt slow down other use cases +Hi CarlAs mentioned above this patch originated with some of the early discussions on Howl before we had mailing lists set up +Could you please test with Axis if the bug is fix there then we are fine +Thanks +Thanks +Updating Fix For to Unknown on issues not targeted for attached patch fixes the problem by changing the following of existing subscriptions in sessionClose method transport session closing state is checked in SubscriptionisSuspended methodThe system test case to reproduce the racing conditions and verify fix is attached as a separateCould you please have a look at the attached patch with problem fix +I believe the end of stream conditions are handled correctly +Fix to be included in next EWS release +ER EC testing is finished. +OK so this can be solved through using the original URL rather than the URL with cleaned path +Its a starter project for them so we have had some back and forth about design issues +I dont plan to go further backRegarding a maintenance release I havent seen anyone come forward to be release manager so I guess there are no plans right nowIt appears a maintenance release is being planned although no information has been added to the wiki yet. +Fixed with commit on the latest build +Ill keep this JIRA open to remind +Patch to clone +The issue is fixed +Issue can be reproduced with varying timeout values if AMQ is started before the second timeout but after the first timeout +This appears to be working I received confirmation from Daniele Alessandri over Twitter that x version does not raise Notimplemented anymore. +Thanks Tim +I will submit a fix once I can get a clean full test run probably late night pacific time +I agree Ive used jBPM many times on Oracle +oops it killed my xml code in the message so I uploaded the example with old and new is not an Exec Mojo or Maven issue Maven only allows a single plugin declaration +Other than the rename problem there isnt anything else in Lucene that should lead to corruption +This should fix the tests +Thomas why was not + committed to trank and branch +Thats bound to give us problems +I started looking at this patch +This should be better +The proposal is to use this to build support for heterogeneous tiered storage support in HDFS +Patch attached works for me +I havent looked at the code yet but could it be that those warnings are from anonymous inner classes +Actually this problems is not in all case Im looking for the configuration where this problem arises +The first time there is an underrun its fatal +Doug! +Oh yes it seems like I missed that one we should remove entry from the cache when gets compacted out +This is functionally same as the previous patch +For me the same issue with and jdbcmysql and jdbcpostresql m currently in a three day training talking about so its a bit of a surprise +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Paul no thanks +I just committed +Thanks setting this to fixed +That still happening +I would much prefer to be able to specify the execution id on the command line +I dont get the code if I do a clean buildLogs need an edit but can do that in another issue +inttype is new in +Did not make the cut +Heres a question why on earth does Castor ignore the cache entry for a given OID when access mode is set to +Also I dont believe it is easy to test with an automatic test but I believe Ankit tested it manuallyI will review the code and run tests before committing the patch +So if you keep only unique words then bigramsdont get generated correctly +metadata +These are almost certainly not required +on an empty line I assume nothign will be searched for +Will try later +flush and sync are mainly for DFS +Bulk move to +Milamber can you help with this issue? +disable shallow iteration +Now in the case I did use check the exception being thrown was so the fact that a is thrown there doesnt break the API +Independently though the geohash implementation i put up should be good to go right now +I just verified that it works and am therefore marking the issue resolved +Known to be supported but needs testing. +I made some slight cosmetic changes to your v patch to get this +Good point Were now exposing the default as well as the default as beans with generated names but nevertheless accessible by +We use Luntbuild for build automation +Sorry for not addressing the comments last time +I cant confirm that +Thanks Vinodbq +Thanks +Im using a globally installed node and have never seen this +Final version I think +Thanks for the quick fix! Ill try the snapshot tomorrow then and report back how it went +Is this issue related to in any way? Its marked as fixed +That test has been flaky in the past +bulk move from M to would love to get your feedback on this issue +Could we verify this +I plan on committing this later today +Hi MarcIt is just the start of our testing strategy so currently we have a hello world app and the process is as followsa it is checked outb the web test plugin is installedc the container is startedd web tests are run against itAs for inclusion in the normal build no I dont think this is necessary we have the functional build running once a day and the main build running on every change +Committed to the docs trunk as revision +Richard please check the FIXME and if theres still something to be done here otherwise clear the fix version and resolve this +am too +the point though is mostly about +Issue is closed since separated issue can be addressed autonomously. +And what you end with is the set of complexities that we have in catalog featuring screens for sales persons product managers warehouse staff equipment support people sales staff purchase officers and moreKeep it simple +Would you please try this patch? Thanks very much +Conceptually we didnt set any values automatically except the namespace settings +Fix is wrong errors property should only be created for domain instance not for the whole domain class +For setting TTL we could simply extend or reuse writeTimeout property +Ill try to create a simple testcase. +But Toni doesnt see the problem when unzipping that zip but I do on Linux when I unzip the zip build by him +Created the patch to make RM write the history data +The tools are there to get this output +I can see where in the interceptor this can be an issue but cant for the life of me reproduce this +Remove double writes +Testcase that shows the problem +Thanks. +As it stands this report is unactionable +Matthew can you take care of this this weekend or just push it back +yeah +Resolving platform issue as project peer has been resolved +Is the recommended fix to create the trigger by hand +is coming in late Q so its hopefully not too far awayJuergen +but it didnt make it into the patch apparently +After get checked in the use cases for disabling table would be reducedThis should give more reason for carrying out proposal +For some reason though it was moved to the main code as well +Resources are low on NN +has the same issue +In order to be able to run tests in parallel we need to become independent of any test global temp directories +Closing. +looks good to me +One thing I realized is that these might have made sense as Stax settings could have been easyish to add in Aalto as well +So that means that useLatesttrue introduce a risk to have wrong sync between site and +Error message must contain +Can you clarify why we would need to update in the BRANCH +Looks good Jarek +HiI have the same issue +This issue already fixed along with it was +Also PARTNAME in table PARTITIONS needs to remain as is since it is covered by an index +Doing subtasks is not feasible because subtasks cannot span across projects +I think this issue can be closed the solution proposed seems to be correct I was only point out the solution. +Thats needed to make sure that we accept field names that we can handle within the fl parameterApparently if you use a placeholder as field name you receive on the default value which can be empty +Lets create a separate doc JIRA and link it to this +Per Marek there are backwards compatibility concerns with this change so we wont fix this in EAP. +I think the existing mojos should be refactored to use +Hope this helpsDustin Amrhein +Thats a great idea +I reproduced this locally with the standard EAP CP all then reproduced after adding VIEWSYNC and the view was corrected after seconds +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +The path checked is different +Fixed in svn revision thanksdims +BTW I think the decision of adding the jar into geronimo server or in the app depends on whether this jar will be shared by other apps. +Thank you. +on patch + thanks +I though it could be useful to have the favorites still available at the bottom allthough it can be hidden as well +I am assigning this jira to myself +I did not mean to commit the changes yet and I will gladly roll them back if others dont like the idea or the approach +I tried running the tests on oracle and i found that they ran really slow +Added an UNCLASSIFIED classification if null or is passed in +Clients are updated +Sorry bad subject read database insteaf of datanase French people can be offusced bug affects on Linux +Delete database blob before rerunning the program +I think we should freeze the branch now so robert can create a last final patch +I submitted it +Option is undesirable for compressed dataOption probably is okay +The taglib is generated at compile time so it doesnt exist as a tld in the source tree anymore +Theres no reason to eliminate the ops +Could the description be edited +Ensure you use unix newlines in your editoretc +Seems reasonable just be careful to not the example since it really is intended to be simple so that people can see the core issuesdifferences between the and approaches +Asks for user confirmation before deleting an HTTPHTTPSAJP Listener +See how to reproduce this failure +Would be great to fix +The problem is that if your target is String which is a sort of target you may have both numbers and dates that need to be converted to aString + +This is a duplicate of in +mentions but it was actually included in +I forgot Weight was changed to abstract class +This looks like a Psi glitch but maybe we must change message when echoing to the clientAdditionally I wonder where the xmlns is coming from since it isnt in the original stanza it is not added for the message of the client +fix for this try add check for process being null +ok Ive committed another change which I hope will fix the problemPlease give it a try and post your feedback here +Gunnar can you provide me a complete build log with logger at DEBUG if you can reproduce it with DEBUG +For example in this case adding versionto the manifest of one of the content bundles should get around this issue +Could you open a separate ticket for these +Perhaps it could work to keep track of all clusters and calculate their mean instance and look for the n closest mean cluster instance and in a second iteration look for the closest instances available +Ill update this JIRA once a new EJB release makes it to the AS nightly build so that you can give it a try +even if they dont like that approach +this was solved in Betae +BTW also mentions about +A couple things the help for COPY is wrong you cant use AND in the column list it would be nice if it output the number of rows created and perhaps the time it took at the end just so the user knows something happenedFinally since were targeting just want to play with it users maybe a full example in examples would be useful +background on my interest in this I just finished a three month project implementing search with lucene on documents +Both types concern to concept of and +The patch looks like work in progress and so I am not reviewing it thoroughlyA lot of the complexity and verbosity is coming from the fact that we need to create a new proxy per container rather than per NMWe should file a jira to fix that in the protocol +Thanks Bob this sounds like a great we no close since r +Please have someone knowledgeable about RTF look at this to make sure the list is getting reset in the correct place +No other questions so the issue has been resolved +As noted in one of the comments the previous fix did not work for unprotected pages in a with some secured pages +well this is exactly the problem +functionality unavailable Code generation is impossible in such situation +You could renameit to STILLCONSIDEREDOPENAFTERPERFORMINGASTALECHECK cheersI think this one is ready to be applied +Add a test and fix the visibility of the cannot put the init immediately after forQueue is called because by that time metricsSystem has not been initialized +File contains a test case which can be applied to the source and then run +No obvious reason why it shouldnt be served up +committed to federationtrunk in r +This is bad and indicates that the serverside is hung +and passed on my machine +Thinking about it more Id like to see it in the Security component but Common would be fine +This has fortunately already been discovered and fixed +We can grab them from there +has been the schema this issue still something we need to complete +So I unfortunately still cant tell if its a pid issue a timing issue or something else. +Im torn between A single new option with an argument that specifies the method name and arguments to that method to call the appropriate method with arguments on the underlying factory An option per method on the underlying factory with an argument that specifiec only the arguments to that methodAs Oliver pointed out set might be new to JAXP +The portatomiccas at line provides the necessary synchronzation. +There are some edge cases not being handled properly but well handle them as individual issues. +My issue isnt really with recursivestructures it just happened that the easiest test case i could think todemonstrate this issue was a recursive caseI tested the modified digester with Tomcat and it does indeed break onstartup +Created a new patch file for the changes against the latest source code from git +If we want streaming always fail the task for exit codes we can just remove the configuration property and document to wrap commands as Rick pointed + +Trunk has it and has it though? Thats weird +Is the source for available +Link to related issues from xx to trunk +Bundle file resourceshelp contains entriesGlobal ForwardGlobal ExceptionTo specify other labels for action objects next entries should be addedCreate ForwardCreate ForwardCreate ForwardCreate ExceptionCreate ExceptionCreate Exception +The Requires is there however update is not pulling that in automatically so we need to add that to the list +Great will be waiting then +messed up the going to make +I kinda liked it when you were passing in the conf and all of the mess figuring whether to read from HTD or from conf was contained within HTD but this looks less intrusive of a change +minor updates to Vector mostly javadocs +source file of the pdf. +Yes it does +bulk defer of featurestaskswishes from to +Random test fails for nowI think bytes sorted is the only impl that makes since other implementations dont work with ordinals and thus would be much slower +So waiting for your test app before going furtherNote when you says I had built an app in pure Flash with the same configuration with no problem +Then I removed the CMS components as we discussed and ran the test without first accessing the default page and the test ran without error +There is a unit test for this behavior in the FROM clause but not in the ORDER BY clause +anyway I have a fixed version of this in trunk now +Which class which method +If I return a String containing from onPassivate gets converted to F which is URL encoded version of f +Thanks Mahadev! +When I tried to add the link I got error The JIRA server was contacted but has returned an error response +patch applied and committed +Make sure can be shut down +Move to +I will go ahead and implement the logic that way +standard should be skippedadvanced should be shown as usualtry implementing a new custom panel for the standardadvanced mode selectorwe should try to avoid izpack this +Yes a separate index that is dynamic while the main one is static except deletes + +I have not had a change to work on this +srcmainjavaorgapachemavensettings is the one that needs to be changed +Ill check again +When you have a but youre only allowed throw and the map task should fail create an IOE out of the PE and throw thatOK +WDYT +I forgot to mention +Thanks Jarcec! +Thats the case also for invalid namespace +Done in Weld and TCK +In this implementation a refresh that is to be queued while a refresh is in process gets dropped +Also corrected since it was using the wrong attributes as well +Actually this runtime exception was not logged in my logs but instead I saw it in firebug as a postback response which included only this much detail +First when a tablet is not at an expected location and rack this info needs to propagate back to the source so it can invalidate its metadata cache +Misunderstood Martins comment Reopening +Alejandro Thanks +What do you need field alias for +Bernard the issue happens if you have multiple tables with the same name in different have a store for each schema +See for some you get a chance could you please verify that it works for youThanks fine forMaarten +Hm I cant reproduce this problem anymore +re this sounds good already I do not think I could do it myself I tried to delve into your source thats too muchalso I would assume that resolution is optional? that is if definition is provided used itif not try to go after url if url lookupparse fails then really fail +Im trying to understand how some commented code affects the generated manifestFor example the is now commented so it may affect but I cant even reproduce that one +What query are you issuing in dbvisualizer +Therefore the application contexts reside in srcmainresources by point +Sorry for having missed your reply Olivier! My lack of knowledge doesnt allow me to understand if that code can be merged inside the jjt source +implemented with revision external clients that use Metaxa for text extraction the text will not be visibleaccessible anymore in the metadata graph +Thanks Sameera. +Can file another jira for better timeouts +Corrected and checked +chapter is chapters are reviewed Yauheniya will update the guide according to my remarks +These elements appearnow in a configuration with a value of an empty string +Dave can you please look into the unsigned in tomcat? Thanks +Such a scenario works fine for me on both final and +Vincent what do you think of this patch +adding a method to check for deleted such as if a utility project is no longer a than these two things though I feel this should work +Or you could close it +Transitioning to closed. +No savings right + I have a proposal for some the current code in +Not only changes also change for hadoop + to the idea that VALUES should return the literal strings as VARCHAR in preference to CHAR +Theres no just ignore the file upload and use the rest because the file upload data is interspersed with the data for all the normal fields. +They are to be expected during the killing of serversIm downloading the logs and see if I can find some clueThanksHoward +To reproduce +Ran hod after applying patch with follwoing options hod c conf b o allocate clusterDir HADOOPOPTSHelloWorldTesttesttt Xmsm XX XXHADOOPHEAPSIZEHADOOPROOTLOGGERDEBUGconsole HADOOPOPTSWorldTest hodtestHOD XX XX XXHADOOPCLASSPATHhomeprocHADOOPHEAPSIZEHADOOPROOTLOGGERDEBUGconsoleAlso tried it by setting value in hodrcNote If in hodrc some value of keyvalue params contains then value from afterwards will be ignored + of a bulk update all resolutions changed to done please review history to original resolution type +it trips the assert in that detects a max +One thing for future patches this patch combines two different changes which is not the model we use +transition resolvedfixed bugs to closedfixed +But thats not feasible to implement in the timeframe +Jenkins failure was expected due to changes Ive opened for requesting to fix this +I think its a nice finding Brahma planning to post a patchYou can also add a test which can detect this issue +So as a quick work around Ive commented out usage of the connections list in and onStopped +Please with more details if this is still of interest. +See this appropriately. +this patch sets the failonerror flag to the subant and let the test target continue the build + +Reduced priority of non critical issues which have current done! +unit tests are commented out at the end of the new class +And patch for should be applied +Without that master will get aborted +I also trimmed some trailing lines +Note that Im still waiting for a testcase +The Batik problem is related with an upgrade of batik from to +Edit I have fixed the stack traces for all languages so the options are displayed +I committed the patch on linux +Unzip to webroot +I can see how it would be reasonable to do the first approach +This is not a bug +Removing the above seems to have made the tests fail +To be honest i didnt use all the properties because i didnt need them for the problem at handThe Cookies annotation for mapping multiple cookies to a Map property seems like a nice addition +However after deleting the resource from the AS Inventory page the record in the database was set to inventorystatusDELETED I then created another Topic with the same data same Resource Name became the same resource key +Its just +It caused runtime exceptions and it is hard to tell that the Editor generated initializers are not sufficient +ive updated the summary to be more descriptive for the issueWe probably cannot fix all in here because its such a broad issueconcept but for now ill keep it running and try spawn of concrete issues. +A cleaner patch is on the way but the attachment explains the intended fix and a way to simulate it +My fix does certainly address your test case +Ive now fixed the code order and watched the flag go into the xjc device and I still get the stupid file +This was caused by the attempt to in the PBCD Builder +Here is a patch incorporating things put forth in the last comment +Would you mind updating to the latest trunk and applying a new patchMy apologies +If you have made changes to your existing copy of the etccloudmanagement file in your installation the changes will be preserved in the upgrade +It is a bug +With respect to the bind status Im not sure whether thats really valuable +Eric could you update the test and print out the value of and see what is that set to? If it is not set to the patch attached above should help +Lets leave it as is for now and see if there are any problems with the current patch +This is looking to be a symptom of full GCing +Fine by me +The address to send notifications to is believe Serge is moving it now +However this patch seems to be missing some fields for the class +Id appreciate it if all bug contributorssubmitters could help me on expanding srcdoc to cover any behavioural changes +This is a product on Struts now +Thanks Danushka for fixing this issueLahiru +I dont know what is supposed to happen there +ruby e sleep Looks like this one could use some of the same treatment got +Could also be great to support Easybeans objectwebs Container embedded in Jonas or Standalone +Since you posted things over there Ill respond in kind +What do you think +Also In case If you want to use in a server environment you could use to set based which work for both Axis server and clients +thanks arun for fixing and xu for testing +Fix in modcluster by r +Doesnt seem this problem exists inHow can Dequeuecount Enqueuecount Dispatchcount and! ? +Resolving +I can migrate this to phonegapbuilds issue tracker if youd prefer +Despite my efforts I did not manage to install a proper environnement to check my modifications so I hope everything is OK +Turns out it was during juddi deploy but a problem with the unittests +In other words the server will display the gateway as gone if currentTime lastSaw gateway somedeltaforclockskew +So the old unit tests should be validOne caveat is that we need to mention the number of reducers via the parallel keyword when doing order by +I expect the next one can resolve this issue and it will be the implementation of Selector +By addressing that issue there is no need to spell out how to handle Logback with regards to the LGPLAlso I should point out that Ceki should have updated this issue +Maybe there should even be a preference to disable that behavior completely in case it bothers users +get is not used anymore of a bulk update all resolutions changed to done please review history to original resolution type +Im not sure but +Code is in airavata svn +Patch to add a hashCode of the indexed data to the Lucene document in order to avoid reindexing unchanged documents +Done. +Fixing this within the Castor project and making this available through a new release of this plugin. +Else we have a problem when editing a label of with the Labels Manager +Thanks +Sounds good thxs +I ran a build based on trunk with JDK and all Hive tests passed +Exception can be misleadingGood catch will fix +requires an unwind method which needs to be added +Good call Id forgotten to readjust after I added some additional queues so it was skewing the numbersAfter fixing that I got exactly what I expected +could you maybe on the build check if it is in the checkout queue and if it is wack it out of there and then add into the build queue since we know the build process will check it out anyway +This seems to be the right fix are loaded from jars just filenames are not +It needs strategies for controlling processing time and code cleanups +This is circumvented by changes in to make tags bodies no longer render direct to out they return strings when called from within a tag. +is now obsolete. +There is no precedence issue since and have the same precedence +Yet I would rather see this in the context ofWDYT +Is that what the decision was? Were just ensuring that its active and bombing out if its not? For some reason i was thinking that we would be putting transactions around the repository if you used the and we would then bomb out if you also added your own +I cannot get to work however +Im confident that the test failures are unrelated +I have to output everything into my srcsiteresources directory otherwise the maven site plugin does not copy the jnlp file and the jar files to the you have solutions for either one of these please let me know +I will disable validation in seampay and quartz for now +Partitioning is the key of MR so reducing the partitioning function to implementations will not make sense for everybody +Fixed for Activiti Designer a lot! +Vinod or Giri did you look at the directories before this happenedIf I dont know what caused the bad directory t begin with I really dont have much hope of understanding how to prevent it in the future +The patch works nicely Thanks Jan! Already applied it in trunk r btw. +Ive the same problem! For me the severity may be major or bigger +Patch committed +Very nice so it appears the parser has no kernel or VMI dependency just signature and reflected object as inputs are enough? Thats excellent the st step already brought desired parser API simplification! I suppose we need not commit this patch please keep going +What about extends +to trunk +Unfortunately it still doesnt +like parseDestination parse parse parseResponse +All the relevant bits from this ticket have long been merged into so Im going to close this. +I tend to agree those parts have too much in common that extracting the core then including it in each other would be an overkill +moved to since we ran out of time for is important since it facilitates analysing user issues +I really have to echo Yoniks concerns here +This patch is also not applying cleanlyhivehive svn statM metastoresrctestorgapachehadoophivemetastoreM metastoresrctestorgapachehadoophivemetastoreA metastoresrcjavaorgapachehadoophivemetastoreeventsA metastoresrcjavaorgapachehadoophivemetastoreeventsA metastoresrcjavaorgapachehadoophivemetastoreeventsA metastoresrcjavaorgapachehadoophivemetastoreeventsA metastoresrcjavaorgapachehadoophivemetastoreeventsM metastoresrcjavaorgapachehadoophivemetastoreeventsA metastoresrcjavaorgapachehadoophivemetastoreeventsM metastoresrcjavaorgapachehadoophivemetastoreM metastoresrcjavaorgapachehadoophivemetastoreM qlsrctestorgapachehadoophiveqlM qlsrctestorgapachehadoophiveqlmetadataM qlsrcjavaorgapachehadoophiveqlmetadata? qlsrcjavaorgapachehadoophiveqlexecM qlsrcjavaorgapachehadoophiveqlexec is messing with patch so manually generated. +This is only activated through a profile named ear so to create the ear file you must mvn Pear +Fixed with revision + +We can start this patch in trunk and then merge toJenkins will give a for no new tests +It is static so that multiple different tests all using the same database dont try and initialize it repeatedly +This as well as seem to be good fixes for as well +Andrew did you have a chance to look into this? Would appreciate your input +Tidied up some comments +bq +to and trunk +PerrynThis bug has now been fixed in the CVS +Added Tapestry version being used and other environment comments +Theoriginal problem occurs even if you just start and then stop +Verified in JBDS +Please fix in future versions +I dont even recommend putting as the communication interface for the agent on the master server platform +Committed to +Merged the first into the master branch +Though I think theres a chance we may end up calling wcs stuff instead of +Fixed +bq + +This ought to be a little easier now after if you make the new security realm +Please open up a new issue to address any shortcomings. +Tried this with previous patch will now try with last one though I think it it should pass also with previous one +It worked for me both in iOS and Android +Also the SAAJ API jar must be in front of the geronimo jar in compile order +I will ask Rajesh if it is on the QE plan +I had a feeling it is not so simple as I thoughtI am happy with the solution +Added some code similar to the fix to to +The diff also includes the change from the adds get method impl as that hadnt been implemented in the existing code +This does not include test but consistently failed without the patch and it passed with the patch +Looks good +My brain must not be working this happens on branch and branch +upgrading this all the way to blocker so it shows up high in the release notes and it did in fact block some people from deployinggreat work getting this in btw +When will the next release expected? + +Thanks pass over all release notes prior to publication of +The trunk version applies to and trunk +Thanks +In light of that I would argue that this issue is fixed and your dependency has moved from this jira to +Summary fields are fine dont need enablingdisabling +that Change button should be called Browse +However you may use the goal multiple times with different configuration to get a variable with current year a timestamp a localized long date UTC time etc +That is weird +Here is a patch fileItd be great if someone can review itThats only a starting point but Id like to know if im going wrong +Each will print notifications +Since this is needed before rescheduled for the tentative patch release rather than +With this the server and threadpool server appear to be including the case of a valid frame followed by noise +What is the issue +Applied +Rafa I committed your patch added the cover image and removed the cookbook cover image I can see it on the production site now +Hi KhaledI run the test for the new fix +seems to be a problem with SAX events the XML serializercould handle but PDF serializer aka FOP not +This fix contains an improvement of the TCK test output Exceptions during test execution are logged +Thanks for the suggestion! Ive fixed this for both RC and +refers to using redhat kickstart to INSTALL operating systemThis ticket refers to using linux dd like method to CLONE a precreated disk image to hosts harddisk +This patch is hyper simple so I just committed it +Added extra time window to dfs throughput aggregation +That class was described in the Cloudscape Developers GuideA future enhancement would be to write a more extensive tutorial on how to write table functionsTouches the following filesM javaengineorgapachederbyvtiMinor tweaks to clarify that is useful for writing table functions as well asM toolsjavadocAdded to the list of public Derby classes +Im wondering if this is due to the call of the flow script +Im not able to reproduce this on current master +Stack requested a patch for +I think that the issue is very important and I have the issue as following start karaf shell featuresaddurl mvnxmlfeatures featuresinstall cave exit karaf shell karaf shell karafroot caveCommand not found caveThanks have the commandkarafroot featuresaddurl mvnxmlfeatureskarafroot featuresinstall karafroot cavecave cave cave cave cave cave cave cave karafroot cave Name Location Which Karaf version do you use +thx guys +Test in now works +Added in tests to verify that user name is returned correctly by for and for the History Server +I would like anyway to solve in Embedded so I will solve it by releasing Embedded very soon +bq +But its not too urgent since EAP CP and CP are already out so we have a couple of months of time +continuing thinking out loud +Thanks RobertPatch applied to LUNI module repo revision rPlease check it was applied as expected +patch to upgrade to and jta and a todo for the failing unit I looked into the same thing last night +Without this they would fail with authorization error +Bruce Snyder got enthusiastic about this at apachecon and we got some basic code working to get the broker startedrev introduces amq as the normal amq and gets the console working with reduced functionality +Manik youve been added to the committer this bug appear under sun jdkjdk? Is the fix specific to sun jdk? targets jdk runtimes. +Moving remaining bugs to +Would we want to compute this based off of a recent history of cluster of just a snapshot of its current state? If there is history how far back would we want to go and how would we handle some nodes in more regularly then others +reopen to set fix version to again with correct fix version +The risk of a catastrophic failure is too great +Username milasx +Touches the following filesM javaengineorgapachederbyimplsqlcompileAdds optimizer support for Derby Style Table FunctionsM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceAdds a new error messageM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangAdds test cases for this work +Changed Target Version to upon release of +this is an port of the example to use blueprintFixes in are required for this to work +Beta appears to resolve this issue +The unsignedsigned stuff should not matter for us +Or change the JDWP enabling comments in so that we detect whether JDWP is already enabled in HBASEMASTEROPTS etc +For the record I built sonar from trunk and checked our modules multimodules build it now works completely smootlyThanks for the fix eager to see sonar released. +Lets put this on the back burner +I thought blooms kept in block cache +This should be comunicated to the activemq team they should change it in its own of +Then later we added SAAJ and mail transport on top it making things far more complicated +brokerbin for example is very similar to but requires a QPIDTOOLS be set rather than using QPIDHOME or trying to determin the correct path +I cant recall the first time this happened since it happened to me once and then the next time was much later +Thank you for your review wed hope you can help to update your patch +Patch which modifies the schema for SOLR and generate fields with Date objects as value normalises the String representation of Date objects +I cant tell why and b +No +Patches acccepted from Elias Torres added support for Apache Derby +Now I removed this code and I sincerely apologize for thatFor style I found almost every project in incubator is following different style +Now that you are bringing it up I find myself thinking that a query option to provide valid JSON would be greatI think the general sentiment is that it wouldt hurt to give an option +Fixed +If you dont you supply null instead +However patches are welcome +This is a very early version of the patch +sounds initial. +This is the little program I used to do performance testing. +for the patch +I had done it only one testcase +Yes Hemanth +Yeah the ports should be +the web page was updated to make the fact clear. +I think for the fix it will be enough to just remove the check but perhaps and will have to be changed to differentiate or maybe it is ok even for those to drop all the disposables too +Thanks +trunk all resolved tickets from or older +Hmm +Isnt that a manifestation of +Oh sorry my mistake +Thanks for reportingYeah classloading in is a bit tricky +However I do have a field in this class to avoid newing a new grid each time +Committed in revision fix at trunk revision to branch at revision default fetch type for and is eager +Inclusion of the following packages has been requested layer SLD and filter packages will be included +Would you please attach the patch for this to the ticket +Have you tried with IE compatibility metatag? +batch transition resolvedfixed bugs to closedfixed +Will add more comments if I find more issues +Thanks! +In this case should we just take bitsRequired on both sidesYes this makes sense +Spoke with Mona +separating the anchor indexing from +bq +Is the missing text just the one you added with iText +cant use the parameter name by default because its not available at runtime +JeroenYes this is a bigger issue +In our code we also log the column family name and the column name but that might be up to the user who wants to do that adds a lot of logging +we can close the issue now +terrific! +It still may require extra BC gem maven artifacts. +In the java generated code it masks them as which is good but can you verify that the way it does that is broken? Look at an operation that does not declare an exception +Javadocs added on How To install and Use I am uploading the patch + attached set of patches implements the discussed UPDATE spec +PS I am uploading this jar with license to ASF +Ill attach the diff fileThanksDave +I think I get the change now + +Can you try again +As far as I know Oracle NG ignores the blob columns now +Currently get returns a default of DISABLEDMEMORYLIMIT which is +Can you try that +Can you try bypassing the volatile image rendering in and just use the buffered image to see if that fixes the problem +This is relatively well covered in the main parts of the site and documentation +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +linking jira +But Ive not tried any yet so not sure +I will post to the lists +Works for me now so give it a try +Updated title to more accurately reflect the Rob Id like to understand better the reason why you want to have multiple users run Fuseki on the same machineFuseki is a server as a web server you run it once on a machine and people will interact with it via HTTPIt is the normal behavior for a server to create files which belong to the user the server was running withAre you running multiple instances of Fuseki listening on different ports on the same machine using different users +by updates in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Builds OK Examples and my BSP programs are all OK on nodes clusterLets commit this +According to Hairong Kuang Hadoops behavior for creating a new file is that it will automatically create its parent directory if it doesnt exist +Hello Danielthanks for your reaction on this bug +Needs porting to see the linked forum thread +This doesnt explain how or why they get superimposed in the same filetree though +Dear TeamThank you very much for your fix +The dry parameter doesnt matter +Leave a comment if you need something else +I wont be able to review it this week though +As we can see the result comments shown in and the test works well +Thanks for reviewing the patch Dyre +Thanks. +Applied for all classes mentioned by Hen except for In the case of this would mean to silently increase the capacity which is most likely not what the user would expect +Created for the instancestarttime thing +The patch looks good +Please if you feel otherwise +I dont think its really necessary to merge it with and as long as its in trunk +. +I cant replicate it here but suspect it is due to the fact that the ORBshutdown is currently waitfalse and so the next test may start with an orb that is shutting down +PatchUsing instance of if avaliable and method get for detection of navigationThis patch does not make trinidad child of but that can be another solution of this issue too +test pass on my machine when applying this patch to the latest bits from trunk +Because the parameteris now an element and no more an attribute the order of parameters is alreadytranslated into the order of function we might still use the proposed patch to also preserve order ofattributes +c default is tmpnative in the jboss server directory +This is probably because jars have been included these classes should only be loaded from the container check your archive and exclude the dep +I can confirm the reported behavior no EL code completion +Now with tests. +A more elegant fix should be worked out for trunk +going to add documentation to the admin the documentation into the Admin Guide. +Ok it looks like its a test thats trying to do a UDP recvfrom using a multicast address +Dan I think this might be fixed as a side effect of +Ithought I had ran them all but apparently I did things in a different order thanI though +I find that painful +moving to next CP this is fixed please mark it as resolvedThanks +The test are passing in eclipse +Tim whats your opinion +You did not the remove of the project reportsIm agree that the report is placed deeper in the site +But its hard to differentiate between all modified filesRobert Can you do performance tests with the old and new flex patch I do not want to commit before +This is fixed by +the version where this has been fixed is now +Sorry the formatting in the above is off +Chapter +Added description of another puzzling piece of optimizer documentationthe property used to set the threshold for spilling hashtables to disk +It will download the hadoop version of the world and test against it +I have the same problem +This bug has been marked as a duplicate of +Hi KatheyPerhaps you could remove from the methods throws clause too? I mean from the method signature not only from the javadocAnd one nit It looks like the patch mixes tabs and spaces sometimes even on the same line +I created modules +Thanks Mike +So now if the +Ill get it out ASAP +Oh no +However since references beans then to me the archive must also be a bean archive in which case I would lean towards needing to have a in +mar file was available +Indeed the old implementation of was hard to extend in that respect +Thanks for the patch Thomas +Verified on +I just committeted this. +However some of these services are context sensitive or have state and should not be running concurrently + assigning to myself until we resolve what to do wrt the examples and chapter to myself to verify in the CR docs +Seems like the version of Apache xmlBeans being used in our application was causing deadlocks related to array setters +i expanded the patch to all the sorts just to find all the wierd sortingcomparators going onit also finds some false positives ones that are documented as inconsistent with equals ones in tests etcbut we can at least look into the ones it finds +I noticed that the properties that get reported belong to relationsThe good thing is that they are not reported if the domain object did not change at all +cant parse erroneous date Thu Okt GMT +if there are any run time issue I will check and create a new JIRA +the patch looks good to me +Committed to trunk +Those attributes should remain within the server they should not be returned as part of an HTTP response +Endless loop occurs if I open the Roo Shell on an existing project or try to create a new Roo project +Hi Ben the problem isnt POP its sending via SMTP +Feel free to reopen if the issue is still present with additional version and configuration info as requested as well as detailed steps on how to reproduce it. + hasnt been fixed yet but will be for the next is not the issue I thought it was and has been this is another issue which has just been fixed in the ESB project due to the introduction of a new coded jar in the serialisation code +can you have a try with the last one please +New patch with tests attached +Reproduced on Jikes RVM r Harmony r Gentoo Linux x +I have created a unit test for this issue and opened a pull request for this unit test +I would like to add some scenarios which can be handled by persisting +We are working on reproducing this exception +Wed either need to build protobuf from source or find a more location for deb filesAnd yes JDK +Thank youI didnt had the time since my HDD crashed +Patch applied at +With concurrent branch commits wed indeed need something like the sync collection to coordinate the commits +This is getting out of hand true +An exception +There is a fixed amount memory needed per topicpartition for metrics +Should happen only after trying a safe mode first as part of a ride over HDFS restart capability +Looks good +The bundles will still be loaded first because they are process first +That needs to be changed to the release javadocs +Snjeza can you please take a look at that one? SWT is definitely not my thing Ive already struggled putting the new togetherThanks +Checked +By the way Ive closed so you can try it again +f +Assigning to Pete for now +Another oneRendering HTML on a screen and sending parts of it viabuffer will help with rendering preformance as these parts can be first rendered into bean property and used in places +If you pulldown on the console tab you can see two with the same fix is simply adding a synchronized to the findConsole method to ensure no two threads can enter it at the same time +Would mockito be helpful? +with my proposal a client is minimally guaranteed to be able to read revisions no older than N minutesYes and that guarantee is IMO too weak +I have started a thread about this on infra with the subject Hosting docs in distEven though these hosted doc youve set up look spiffy Id like to know our complete set of options before we proceed +test project works for me +Thanks for that and of course thanks for your example +For now the priority is removing errors and unsupported options +Definitely dig into the classpath and make sure the stuff that is there really is what you want thereThis was resolved in by allowing the httpconduits to have a wildcard URL for an ID and updating the WSDLschema retrieval to use it +I was thinking of the same +As this is a new directory you may find it handy if you just untar the attached rar and add it into svn scabindings directory +I feel that we do not need two locks started and synchornized pendingCreates in +help improve the codebase also by improving the stateof the unit tests and that will be a good thing indeed +Can you try thatCan you please provide a status update on this issue? We will be releasing soon and would like to ensure that this ticket is resolved +messaggi non tradotti +Version was recently released but the threadSafe support is still missing +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. + support is already implemented in Cargo +so may still work after allThe issue actually has more to do with multiValued fields with a real precisionStep +Tried build and found the same as before +Hi Tina the first question is caused by the could not be load in the second one is the +Dominik this issue is abot jk status manager not changing status +This could become very difficult lol +I dont understand why the CAST operator has this strange effect on the code so for the time being this test fails while I research the behavior +I see what youre saying now +I will definitely come back to get the answer to this i will not surrender this as i really want to see my code tested automatically +I have now committed a first pass at completely rewriting the refactoring support +Audit Activity logging completed for. +The result is the WSDL definition doesnt get resolved when using which ends up resulting in theIve not been able to figure it out yet hopefully someone will be able to explain how you do things the new way. +Thanks a lot for the review Chris +Please recheck them +Then if is released then end users can upgrade and use that if they use in +Apache Felix Declarative Services has been released +In the past the decision has been made that for remote servers if we cant verify via a poller we take no action and simply mark the server as started so that run as run on server continues to function +So you could actually extend the Geronimo publish processes by providing a publish task that sets the appropriateI guess everything is possible +Pushing to to initial support in support for xmlpi xmlcomment xmlconcat xmlelement xmlforest xmlattributes xmlnamespaces +The testing didnt get off the ground +Backed out the patch from Ramkrishna doesnt have access when he is in officeTriggered build + the patch looks good +I would guess that is not intentional but a small offset problemI was looking for some sort of FAQ or something on this so just a small side questionShould issues with a patch be reported in the same jira once set to resolved or should I really make a new one in this case +bq +the System Identifier should be valid URI Reference +The demos in the installer are currently snapshots of them in a zip +just example just provide examplemaxcsaucdk an HQL query can result in mutliple selectsmaxcsaucdk you want it to be read consistentmaxcsaucdk and if you run without transactions you are running with autocommit on which causes commits on Oracle even though you did not commit the txmaxcsaucdk +We will wait for release of JBDS as underlying issue is set to be fixed in in the release notes as a known question though EAP Engineering page has Seam as the Seam version to include +So if anyone wants to use the shell to test puts including messing with timestamps and has learned the old way after this change now what they know will not work + +doesnt call shutdown on its server object within its own shutdown method which should be start the shutdown of Jetty in the first place +In other words the principal would never be null +remove some bad dependencies on jsr stuff in the base console +Thanks Jon +bulk close of all resolved issues. +constructor must retry even if table does not exist because the table may not have been reassigned yet +If you are doing your own redirect then you shouldnt invoke any additional handlers +trunk Dan give the component a test with the commit from the trunk + +thanks for following through on this +Trace log for your reference +There is a related issue as well as related use case where we want to rename the it to an it to to align with the related issue +To be confirmedI have tested for one region at a time but if i change the code for multiple regions then it will work in parallel as each RS will do process as long as are part of online regions list +Please do not use domj or earlier +Incomplete testing +Do you have a unit test case for this would help to ensure any fix remains fixed in the future +I subscribed under my gmail account and the reply button goes to for the latest message. +Improve performance tests by performing more reads. +If someone runs into that please open a new issue +I think Ive sketched a good proposal for so assuming well fix that one in future we agree this is to reject right? +There is a need for a way of overriding fieldsproperties in superclasses +Well point fellas here when they are trying to figure how this all +Its in the pull request of yesterdayI added an option to the setup command to specify a provider version and added a method to the interface and all implementations +Seems like all tests passed even though it timed out +we always UDF with pusdowns +So its extremely unlikely that users will notice a problem with other portlets +Ive attached with the mvn X output +for the to trunk and. +Done +Great job to you and whoever contributed +Not only for Facebook but also for the oficial refreshtoken +Committed attached bug and added entry to srcetcCHANLGELOG +throw an exception to indicate the operation is invalid +For +Very nice. +Thanks for the patch Drew. +My next task is to make it possible to have custom fields +This works in some regard but not well enough in your case +Ah gotcha regardless Ive been testing with your interface changes for a week now and things appear to be pretty stable now that weve shaken out a few integration issues with common js +We dont consistently use the configurations from Do we need to assure that the various log configurations actually work together properly +One of my suggestion is that use an security provider in bootstrap that is in we just load the specified provider class and get some security tools just from it and do not use the normal load mechanism in security module in order to avoid the chicken and egg problemGood luckLeo +Verified on revision +Because of buffering just calling skip on the underlying stream is not sufficient +Some minor changes in replication admin page Test case for index and config files replication + +Commit as you please +Committed in r +I just committed this +The attached file must be applied after performing asimilar set of svn move commands + +David thanks for the patch +The error log +Its now fixed in the QE test suoite with the sysproperty workaround test so the WS quickstart tests now pass +This is the only component outside affected by this patch +Fixed the client id for the fetchers inside to have not just the wired in client id but also the fetcher thread name +Guys do you have any objections to this issue +This patch only affects only Local Mode in pig +pushing off remaining issues in order to release version in JIRA and create release notes +dont we in the good position to resolve this heavy no longer slow +Fixed verified by Kris +As of the current there is not way to tell if a property does not exist for a Groovy object or if the property is initialized with null if I recall correctly +This patch handles Trust Deciders to the HTTP Conduit implementation also takes care of and most related related to Basic Auth +Im sure other UNIX and systems have similar things +Thanks for opening the discussion forum thread +Requires API change so moving to + +Update variable in contribqsol to to be compatible with trunk r +Flush at MB +Works great many thanks! +Attach a patch only when done +Are there any method to reproduce them +This has been proposed before and has withered when there wasnt much support in terms of code contributionsWhat sort of contributions can you provide? Problem specification? PMML parser? Model code +Seems that you used JBDS wo EAP package +I think this is another gap in the API like being able to use forwards as the input property +Its very helpful for me to be able to see more than one set of cache statistics at the same time +There however we should also have more semantic data such as the class of the service resource +I definitely open to other visualization options +Moving out to for now +The browser back button is enough else please provide a patch +In addition it is not a hidden change of Castors behaviour and can be fixed very easy by the users +What do you think of the idea of passing liveDocs into +Yes to both +I can easily see how the ApplicationEnd event may end up not cleaning up everything if stopping the appenders takes too longBut so far Ive always trusted the ApplicationStart events +Reopening then closing to deal with migration bug. +I will attempt a fix for my own purposes you are welcome to it if interested +The javadoc doesnt seem to indicate that their is any ordering thats guaranteed +We believe the issues are fixed but want to land the extra tests that prove it +I added a license formatted it to adhere to rough rules used elsewhere in the codebase renamed it and added it to +Hard to tell since they were always working for me. +Also the Vertex compute iteration part of code in could be made better +Since the tests structure has been changed I updated the patchPlease try the updated patch +This is is the junit conversion of +will be fixed in jdk tested and can verify that this is fixed in b +Thanks Adrian for the hint +yours that could cause behavior changeAnything I can do to help with this +Bulk close for Solr +Though I agree that you could fix this particular problem by walling off the systems I dont believe that is a very workable solution for larger environments where all you have is the ability to log in as a named user to a system in the data center and you do not have root access and if you wish to setup firewalls the best that can said to you is good luck with that +We need to check it on default seam runtime for JBDS +Once Jieshan gives patch for trunk will commit to and trunk +We may be able to provide an improved AST Browser soon. +I am reviewing this patch +Simplify buffer management since +Here the patch from +Patch for Yahoo! distribution with comment for fixed. +Maybe Ive got the semantics wrong but I had thought its not the watch that gets called but the session event handler +Something like this would be a useful improvement +Thanks Keith great stuff +thanks! +Updated document for install file layout +I have added annotation for geo fields merged all changes from master +Error message in +Changed this issue to no longer be fixed in the and releases. +should set going forward but more interested why this was changed since M and how QE didnt spot it +My new patch simply makes it simpler by reusing existing code +Running your code does not show the issue +Yes you can mark this issue resolved as fixed once the updated patch that Douglas Creager provided is committed to Subversion +Committed to trunk. +Checking how this algorithm works I found a misunderstanding between two concepts It is necessary to save the object created for actionListener validator and value so we can revert them later when we resolve a nested retarget +I agree that the performance advantages outweight the downside but it should bedocumented +Transitioning to closed state. +was changed to be in version but it throw error when work with +It doesnt have any sideeffects +If the issue is rpc costs and SS resolution optimization then lets fix the jira title and comments +Hi Markus I tested you patch Do you forget to add deploy and test target into srcplugin +Im reducing priority on this issue as we are awaiting more information from either Will or Michael +verified in trunk +I presume this JAR is not released under an open source license compatible with Apache License the above is correct would you kindly resubmit the patch in a manner that does not require a dependency on this JAR +note im also not a fan of adding multiple items to this list so im wondering if these can be dynamically added based on the selection to not show up for everything +I conceptualize Commons projects as being tiered anyway Lang IO Collections Pool +Is there still interest in adding this patch? I know there are a few things that I couldve done better here +Can you please try with and report back whether this is still an issue +The only change made is to compensate for thevariation between and invocations +Thanks +Im going to attahc another patch shortly based upon version branchIn addition to the problem in the did not convert the Zebra schema to Pigs properly if nested types are involved the low level column schemas were simply missingAlso the conversion from Pig to Zebra schema is just missing beyond a hack to work on specially prefixed column names +Patch applied thanksAs this a bit problematic issue it needs extensive testing +Attached is the patch to fix this compile Close for all issues before incubation +However I did notice that the web were not working which I think will be fixed as a part of although Im still investigating that +Can you provide a repeatable scenario of this failure +If they are Ill use them Ive got a license for Tangosol and would like to use the springmodules cache impl but would rather not have to write the integration myself +Shreyas has provided some pointers +I will be adding the beta version of the library to the repository since its a dependency of Aperture which I need to add to the repository as part +Ive seen this on linux boxes that dont know who they are yet are on a network with DNSan nslookup of the gets relayed to the DNS server which then failsnslookup ServerAddress server cant find NXDOMAINPresumably adding to etchosts will make this problem go away on that machine +I removed the incorrect uri in the service callback binding under r in the branch and r in trunk. +But I appear to be in the minority on this issue so Ill contribute my own asset path analyzer ahead of regex analyzer to restrict those paths and make things easier for the majority of users that prefer images and css files under app to be accessible by defaultCheersRobert +By the way it would be nice if this default was configurable from the outside +On the way back to text these special parts of the text disappear since they are not stored as text nodes but rather as node attributes +Fixed in revision +Perhaps we can add an interface that contained that additional information +Further controls defered to RBACACL JIRA close of that have been completed or are now invalid. +for removing them from the messages fileAnother thing The after a field label is locale dependent so we may have to support this too. +Improved with r. +It was inconsistent with how the data was being serialized in the old portal but I can tweak the rave core to match thisAnother issue it looks like the regionWidget instances metadata object is not being serialized into a json object but delivered as a json string. + +User should not need to set env vars to make the configure work and if they did you need to include significant documentation which this patch doesnt do +Thanks +What kind of integration do you have in mindDo you have any use case + +This is why I like using style keys since they are human readable. +The example from the drools svn is like this +Committed patch as shown above +clinging to a seemingly hopeless or outdated cause +I think you tried it and it did not work +Ill keep trying +Sorry +the warmup would have to contain all the make calls you usually do +Thanks Sean and Roman +Any comments would be appreciatedThanks +It could be that Glassfish has left over artifacts from previous install +Editorial pass over all release notes prior to publication of. +fixed in r +This appears to be an Eclipse bug not a Derby bug. +Bulk close for +If I cant use your patch with I would love to get some help to make it compatible +bq +The modcluster user guide is now incorporated into the HTTP Connectors Load Balancing issue is therefore ready for closure in EAP EWP. +Marsha Id really need a sample XML schema that I could use to replay your problem +sorry wrong bug. +Also I would only expect the details to change when the actual state changed +Batik is doing the correct thing + +The above description assumes N is small which is probably trueNormally it is suboptimal for scan to return just one row per callLets see what Jiakai has to say +Chapter does not refer to performance documentation its one of the related to optimistic locking and jBPM +Work has been completed on has been submitted to Carlos Sanchez +Components must obviously be set up the component frameworks way +Adding a dependency to as it would be nice to have permanent available for the VDL doc links. +This function fails if system userLogin is used which has no partyId associated +I dont get the problem +The approach can be easily used for some queries or scorers as it could be +I fixed the links again with revision Backed out changes to as part of r which broke the Japanesetext againI suspect this happens because of the localeencoding being used when running Forrest and I dont think is to blame +Yes it is not portable and expects input to be in a certain format +Changed it to be and Committing it with these changes. +We can do this in a separate patch +It sounds like thats the property its referring toThe tools and utilities guide has an example under which uses a property called but theres no such property either +Scheduling to when we will address as a whole +Lets punt it to then +Please fix them +It hassome nocommits which Ill clean up before committing +Enhancements in general image weve added tiff support using Drews metadata extraction library +Hi TimThanks for the fix +Judging from circumstantial evidence this appears to be entirely down to which is called once for each bundle and then iterates over all the bundles to get their exported packagesI tried this on Equinox and Sun JDK +Thats why we can change itI think the benefit of keeping the files within CM within their logically correct place outweights the convenience of not having to understand the small customization of the POM file to get it to put the files where we want them +is this patch formatted correctly? am using a cloned git repo from githuband generated it with git palso any and all feedback welcome +The Bridge portlet assumes that the current object is the one provided by which exposes functionality to change the +Whatever the case the broker should be able to deal with it though and once the tcp stack reports the real state of the network things should come back to normal +We need a standalone test case that we can provide IBM support with +Sorry Andrea just saw your reply now on the mailing list outlook is not that good at displaying mails sorted by conversationthread so your reply ended somewhere else +Excellent thanks! Patch applied in revision +I will attach a new patch in a few minutes +This has gone stale as well not a problem anymore. +How is that going to work? Do we use all dependencies from maven central now +Attaching patch aRegression tests passed +For minor releases those branches should include this in their build +for your example FooBarTest map to a service and FooBartest map to a operation test in FooBar service +Maybe Nigel can help in shedding some light on the difference between the target the runs and ant javadocs +Ive upgraded the snapshot versions of all experimental modules but this should be done as part of the release +This is the patch requesting permission from requested for was already committed for because the Push dialog is hidden behind the Publish to Openshift Server dialog +Wouldnt it be better to simply make javaExecutable configurable as with the other execution options? That would have less moving parts and you can always derive the configured value from JAVAHOME if needed +We could replace the original schema by our fixed one in the XML Catalog +Do we need one specific location in the PATH or two etcFor reference Android requires two the tools and the directoriesWould love it if we could get this figured out before we tag tomorrow +The test the principles are the same my system is obviously much more complicated with many dependencies and customisations but nothing too I can do is begin ripping out code and dependencies until it works +Snjeza I have some difficulties to verify this +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Especially the need for a handshake phase before the slave accepts incoming log records +The template is parsed with the encoding you set in the property +An option to not generate Java classes for imported type systems would be interesting for us too as we usually import from jars that include the xml descriptors as well as the generated classes +Uploaded as orgbeanshellbsh. +Thats the kind of useless complex overengineering weve had to put up with in other frameworks and why a lot of us are evaluating CXFAnyways Ill take a look at the Aegis stuff I guess but Id much rather have had this just work or at least been prepared for it to happenNot trying to be a jerk just say that this stuff should just work out of the box +Thanks +The test failures are unrelated I verified by running the hcat tests separately +The patch is dont understand the problem +and try to convert to the same rep instead of actual representation +Broken for modules and assets +release bulk close +Thanki have Replaced some of the dependencies with stuff from the Hupa project manually before compiling it + +Also some fixes for displaying the taskAttempts page +What will be a better value? we could increase it to K that will add performance impact to people who dont need that ammount of duplicate detection +I downloaded the zip file and can see what you are are taking about the +Ship it +False alarm I could reproduce this but as soon as I finished my small test app the issue didnt appear anymore +how do we do this? Looking for docs as we speak +Once the samples are executed Synapse will create FIX message stores in these two directories +Patches for and attached +Thanks Pi. +PU root is always the deployment that contains +That component is autodetected +Dont we fear about security issues? Giving an access to all runtime Extensions might be a bit too much +There are two newly events and +Assigning back to you for committing +We also need to add a note to the docs about this. +Yeah we just need a workaround in the meantime +Cool I assume you are doing this Andrew for +I have created a patch which adds suport for starting domain in debug mode +excellent +Daryn just checking do you intend to work on this now that are in trunk? +This also has some dependency on right and that needs to be committed first +Is this really a grape issue or is it related to ivy +This class was not thread safe when running in batch mode since all threads share the batch mode flag +Currently when a tablet is closed it interrupts running scans +Also effects with +Welcome Screen images should be existing seam related welcome screen icons in the Seam Icon is used in the Welcome screen dont close issue that QA needs to verify +No need to open another jira lets just use this one for the remaining issues +scriptsvmhypervisorxenserversxen rootusrlibxcppluginsIm going to update build and run through the tests again +reopening this to revert +Ruby script to replicate reported problem first string is plain ASCII and deserializes cleanly +bq +then makes the change proposed by benjamin and includes a it out since its not a blocker +Hi ParickI was near to commit your patch but then I wondered if we should really expire the browser cache for the cart OOTB any opinionsDid you find a bad behavior for OOTB +So lets fix and it and raise a separate issue for any bottleneck that is exists +It comes up and tells the namenode about the blocks that it has +But I want to get even more parallelism by preading multiple hfiles during a scanget for example +Yes as stated in the description And subselect is at least equally important +From which site did you install the connectors? Did you point JBDS at an older Central site? Because you couldnt have pointed at the JBDS Central site as it does not exist yet +fails after this since the patch breaks +I dont use or know anything about AIX just to make that clear +It seems that the groovy nature is being added to projectsmodules where it shouldnt be +will commit soon +Change to the dir field during the test setupAll tests in now pass when executed together + Thank you for your suggestion +Hopefully Ill have something to contribute tomorrow +We wondered about that too +Thanks Tim. +Works fine nowCould you? +Unless theres a general consensus thatthis should go throughLove the author Administrator javadoc tag in the factory class! The Administrator thing is because I keep forgetting to set my preferences in Eclipse +Now that passes so quickly I wonder if it should be a large testI could change to medium but it does still stand up a so in my mind its still bigger than the small tests + +This patch seems to work for me +If a value is empty we interpret that to mean delete the key from the config. +Yes I also tried this as root and the issue persists +I changed that and the tests now pass +marking urgent since this is a regression +I am not saying we shouldnt make these changes for securitys sake just that we should group them together so that we dont introduce new incompatibilities with every new releaseKathey +I am not able to reproduce this problem right now +But I have my silly hat on currently +Thank you for you report +I was trying to see if it was my issue that was causing the CSS code problem so I tried the RC release because it had worked on TomcatThe RC release was the only one that went into infinite loop +srctestregressionH +Hi RicHas anyone had a chance to look into this issue? Do you have everythingyou need from meThanksBob +May you try the new patch attached +I applied it and confirmed that the tests which were failing now pass. +This is an old bug that was also fixed for +Darn and I thought I could keep a low profile by adding it to jira instead of the mailing listAnd now you are making fun of meEclipse Context Menu of the Editor Team +Very odd +Keep in mind we should I think still apply something like my patch to disk full tests that count as we dont want this kicking in here I thinkHowever i want to be absolutely positive we arent hiding bugs before doing that +In Struts added profiling to With this we could hopefully get a picture of how processing time are spend when a request get handled by Struts +great thanks Nick! +committed +It seems that this exception is just logged and not thrown +This is the second patch Ive uploaded +Talked to releng teams and currently the situation is Pushing data out of boston qa lab in realtime is currently not possible +Any chance you could provide a testcase demonstrating the issue? If its a lot of work to make your test contributable how about you just make the change locally yourself and confirm it fixes the leak in your environment +Thanks Thilinaregards for applying the patch. +this is why at least for now and on flex for different reasons I think we should keep this stupid heuristicOK lets keep it for now? But somehow we need to remember when this gets onto flex branch to put it back +If Understand David instead of creating new subtask I am to make a commentSo this comment is about a Import screen and service that reads an excel file and imports the supplier info and inventory data into ofbizthis is a more user friendly way to import since it is used in many webstoresthe info import is more for party like contactmech stuffthe inventory is what product since there is only one the supplier supports and thier inventory informationit is assumed that each supplier will have one or more facilities so the data can also be done thru the UI +Claus The bundle snapshot was publish to apache maven snapshot maven last month +If you cant log you either suppress or mask the previous exceptionLets mask it? That way the user may get the wrong exception but hes not getting a situation when something failed but looks okay on the surfacebq +Looks good +Packaging is still pom in the file +I didnt know how to reopen original issue +But nevertheless I think that the implementation of OWB is wrong in this respect It starts by evaluating the bean query +I also need couch to work with NFS +Please what is the status of this issue shall we close it? +Maybe some option to turn off the use of the isolated classloader +Dependencies changed by Martijn and verified by and me. +The workaround is to the document type declaration in and to add the line false to doctype now gets stripped before use. +In general I think a thin client is best and the confusion and bugs created by such optimizations is rarely worth it but that is just a general philosophy thing that is not related to this change +There could be others +Good stuff +We have tested the fuction of and found that it chould not satisfy our demand on task level recovery. +for the patch +Ha! That is funnyThanks for the report +had to reopen issue to edit release note text and set the release note documentation fields +fixed a typo and added some comments to the changes of +This is always enabled and not turned on by the newwcf flag +Can you add the wire log section for the GET request sends initially to receive the response? Im curious to see what the wire log has for the path segment of the initial request +Still needs work since dealing with Segments has changed +Triaged forAssigned Normal urgency. +with revision all import related TCK tests pass resolving +fix tests for br de fr and nl analyzers +There is an old link to Exadel download page +If we are discussing an integration test framework that must interact with a cluster perhaps on demand then there is potentially one already under development for that +Fixed the parent issue and now the juint test test for this +Could you tell me how to assign this JIRA to myself? Thanks +Hi Ming Ma it seems like in metrics might be appropriate here +Que existeix un substantiu en catal? que en francs com en angls s sovint al plural +This should handled in add I guess +I think I can make it up +Adding a dependency on causes a cyclic reference and build failure +from me +I believe asyncing the live deletes is a good thing as that waywere taking advantage of concurrency +New Action fill in details run the testng test that is generated +I now implemented it as a real command to have completion on repo uri basis which would not be possible for an alias. +If you want you can write your own controllers and views to display the audit log on your own so I think you have a viable work around +I guess then it would be fine to leave it on this bug since this is where all the comments are and you understand the problem now and it is also on The history is here to show the original thing might have been something else if we happen to ever see it resurface on zos and if we do can file another bug for that one +attach a tried the above steps to reproduce with a locally built with your patch appliedUnfortunately it seems that it does not solve the problem for this caseFrom debugging it seems that this is because classWorld null and thus also container null and so wont be disposed +If I had used a Boolean instead of a boolean then it would never have mentioned that I was loading the wrong type +Closing as there has been no progress on this issue for weeks despite a couple of attempts to contact you +Alejandro the cause is the Logger was loaded from instead of logj jar which was unexpected +Oops enter key slipped before I attached the file +Id rather not have this in as the fix is and theres a reasonably simple workaround forIf theres demand we could include this in a once people have had a chance to try this out in trunk +But that isnt associated with a specific filesystem connection so what are the semantics if you have multiple connectionsIn this case it seems better using UUGI and confThe patch looks good to me but again I dont know much about libhdfs +Instead new importer would be enabled by doing one of the followingTo enable use of new importer for all Python interpreter instances in other words for everything will be necessary to specify To enable use of the new importer for selected Python interpreter instances instead of a comma separated list of interpreter names can be specified +Could you share a test case that demonstrates this issue +In the provider I am using that object to load the records into a staging table +Next we should cache it in request cycle +workingclasslibdependsjarsbuildwiniamsvcdebugsemisclassesthank you for reporting thisCurrently I have problems downloading required staff and will investigate the issue on my return to Moscow +Please review and commentThank you +You can close the issue +Done +Fixed in trunk. +The assembly mojo is a hack is you just want to generate a zip of your project +If it has to be done as one piece I would wait for this to go in before supplying the patch for that issue +This design ends up spewing java types everywhere +Is there any news about the location of this parent POM file? +Attachment has been added with description workaround has been added with description workaround has been added with description workaround has been added with description workaround has been added with description workaround has been added with description workaround has been added with description has been added with description has been added with description has been added with description has been added with description has been added with description has been added with description already mentioned an error reporting facility is included in FOP Trunk +is an updated patch proposalwhich resolves the conflicts from but is otherwiseunchanged from the v version of the patch proposal +That is awesome +thanks Robert +For that reason I am closing this one as it is too broad. +Please feel free to take a look +Applied the patch to the Synapse trunk +You can download it and check it out +Thanks! +bulk fixing the version info for and all affected issues have in propose we do not wait for but use the solution from the proposed patch +Sorry for the delay this is done now. +This also fixes the leakage you describe +Upayavira has notified us it has not affect Wicket as this is not an Apache release +Closed with the release of. +There should be a proposal on exactly this extension should do and how it should workI have started working on a demo on how I would envision this +If thats not too complicated I could try to have a look with some help +I would prefer if you detached from the yui and bubbling plugins as they dont seem to be keeping in sync with the latest versions of yui and that might present an issue with using the latest versions in the plugin +We should rename this to something better +It should be less so now +put or clear commands +After can be marked as resolved +has been modified to not only keep around the current SRID but also the full geometry descriptor so that subclasses can use it to gather metadata other than the SRID +Not sure why you think its fixed +Thats why my name appeared as the assignee +To fix this bug each time a clone of the logical plan is created for regenerating the plan the is being reset +Regards Rasin +I agree that events are an important source of information and we could raise the level at INFO so that they would be found more easily +So refreshing the webapp works around this +no issues for to trunk +Thoughts +This should now be fixed +Ive added a patch for a simpler test case to that ticket +Ok modified DTD handling so that if DOCTYPE declaration is completely missing validation is essentially disabled and a warning is passed via no exception will be thrown +expires that or means positive infinity and that means negative infinity +If there are user defined egress firewall rules or ingress rules when a network is deleted will attempt to remove those rules before deleting the network +The real tough problem just as you have mentioned is that dfs kill our locality of IO access which I must rely on +If anybody considers that a problem please let us know and open a separate issue +Thanks Jitendra! +Dont know when Ill find the time to do this Maybe I can have a look on fridayRegarding the name is is clear if you also have is WDYT +Digging up the issues or mailing lists might give you some more info about this +In those cases when using assertScript you should use triple single quotes not double quotes which create + looks good +This file throws OOM exception with similar stack trace +I expect occasional null pointers when this map isused by multiple threads without synchronization +Sounds like the consensus here is that while needs to be included sources do notEven though the namespace is is treated like an upstream rd party component not a part of JBTJBDS proper so sources are not required to be included in distributions. +Closed after release. +Assuming agree on need to have it inCommitting to and +In I back ported some changes in the Aries application resolver to throw much more nicer and human readable error messages +This is the result of discussion on in the thread on general +A first fix attempt as been commited as of revision but doesnt seems to be suffisient +Scheduled for +I found a couple of errors inPlease use this one +bulk move of all unresolved issues from to +Ok resolved as invalid +Marking it as fixed +I know it produces messy output and a lot of it perhaps we could exclude some stuff from stdout like plugin repository loading config parsing etc +Need to check one more scenario where on master restart what happens to the zk lock +I really like it from my side +Ioannis Im taking care to reproduce this issue with Sonar and H DB +This patch pulls the classloader stuff out of and puts it into +At all +This would also mean that we should really allow to parse Dxxxyyy and pass that right in to the JVM +I think this deserves its own entry in as it affects the API +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Please review the code though this is only initial draft +This quarter is pretty painful so far +Thanks Steve +So I think the important thing here is a recovery plan in case the non serializable exception scenario happens +Execute POM with clean package +revision . + +Finally I have finished a complete installation in my virtual machine based on Windows Server R SQL Server R and SPPThe next steps are configuring a personal site with some documents in the document library download compile and run on this environment execute tests against Ill let you know soon any updates on this +Please provide a test case showing a problem. +Also tries to use is if available +Try this to change mavens environmentfrom mavenbinREM MAVENBATCHPAUSE set to on to wait for a key stroke before endingREM MAVENOPTS parameters passed to the Java VM when running MavenREM +I dont know what changed in my setup but something has +I thought the svn changes and comments were already getting long and tracking further changes including the stuff all under the one issue might have been messy so I split out remaining work into. +The web console must also be deployable for non osgi environments such as a Apache Tomcat Jetty JEE servers and whatnotAlso we have no interrest in creating new s +Seems like a good idea to me +This issue isnt specific to Blazejs application +Reviewed +I dont have a sybase environment t test againts but I will give whatever assistance I can to anyone that does +A possible fix for this issue +Could you please members to functions as stated in the original description + + +It uses on a correct pointer and also frees malloced memory in parent process +Merged into the master branch. +Initial implementation in rev +The other way is to write a serializer which adds this data back into the log messageAs Im new to Java and didnt want to muck with how the Syslog source currently works for existing users I chose the path of a serializer +cjava and tests for both also updated +I think Tims idea is fine and should be committed as soon as there are some small tests for the patch +Committed to trunk and branch +Maybe a clean rebuild can solve the problemYou are right +Then it automatically started to install jenkins server app +COM developers still recarll with terror a batch of winx gateway boxes that could generate the same guids due to a common +I cant currently connect to your mirror +Is the test still failing +The bcbms is whereas for the trunk its +load may request other scripts being loaded +Please look at which is the exact issue and copy of has been closed without originators approval as release has been out +It will be great to get this fixed spell checker has always been responsible for a lot of test directories that the test framework cant remove on windows +the stuff is already in the code with the vhost refactoring +If you want to see this change included in please test and confirm that it works +I need these issues to be solved to build to resolve relative s now fixed +If you see removal of any form of username restriction as a valid step for this then lets also remove it from HTTPFS at the same timeI think that this is a good thing to shoot for +It should +Added warning and error messages is values cannot be met by the systemChanged defaults tomb for memoryGB for temp storeGB for storeThe GB value was being incorrectly calculatedresolved by SVN revision fix to fix failing tests in junit tests fixed warning messages can be generated for not enough disk space if the directory doesnt existSubversion Committed revision . +remove the module from the server and republish it minutes later when you figure out how to fix it +I need to take another look +This is fixed in Weld +I would be in favor of putting this in as early as possible +Regenerating site doc now. +Rudy I am not working on this at the moment +Thanks for the patch Kelvin +No further comments closing. +We need to think about the way of managing the change +bq +Is this a zombie bug +new files not versioned + +Adding a Javadoc per Scotts DougAll of the comments above have been addressed +Exmaple of serialized message have added support for BLOB and CLOB in and helloworldtxsqlaction quickstart was also updated to use CLOB for the listener table as an example of the in revision +This is the mirror of have just committed this both to trunk and branch +attach a patch per my previous comment +Updated version of patch attached +fixing the description field should be easy but we dont currently have control over individual rendering of text in this +Primary storage is NFS +Made as static +This will also affect multi tenancy in version +docbook patch based on himanshus xdoc patch attached to parent w some edits +Digging into the code I think it returns a string containing a newline of some kind to indicate end of line +suitesAll ran cleanly for me too but then it also did earlier +Hi KenCan you clarify the steps you have taken? The following command does successfully install all required dependencies and configures email artefactsinstall email provider hostServer protocol SMTP port username yourgmailuser password yourpasswordI have just tested this and it all works as described by you above +Rather detects the specific object and uses but that detection adds a small cost. +Fwiw I agree with Walt and Bertrand here +You want more reviews Gary +Is your problem that you have you own shutdown hook which may or may not be called before the Sling shutdown hook +Close after release +Bulk closing stale resolved issues +No integration tests +Committed with revision . +Thank you for your time +Added +The persistence adapter used can be specified via config or found based on a lookup using the name of the jdbc driver +Its indeed a bug +People often try to use these with Maps or Bags realize they cant and then write a UDFDo you think INVERSEMAP is something that people need often? I question the general usefulness of that one +was removed as locking the lock on the received hashmap locking is now on the individual Credits +I a bit confused so please bear with me So you upload a file but its not a property on the form +could you point to an issue or tag numberBR may have been fixed by theBR EricThanks I was much out of date with pdfbox +We have to figure out why CDI openOn doesnt work in build from hudson +for the patch +Moreover if you want to have fun on the Quality Profile page feel free to Replace the standard HTML Form components by Select components Use the new dynamic selection component in the Alerts tab to select a I improved the popup menus +Also who is still running that can verify the fix took? I believe this information is classified +Associations work correctly +Attached performance report showing the limitations of the current implementationAlso attached a patch to add a new that improves on scalability +btw tests pass this was my mixup +I need this so I can a single instance during decompress +The easiest way to then debug the server is to cd into the bin directory and start with geronimo jpda run +Either or open a new one +pushing to the popularity of the odbc transport having a approach to query execution would remove the need to adjust the odbc socket threads based upon the expected a very odbc specific api into the connectionstatementpreparedstatementresultset classes +Once that is done I may review this. +Closed after release. +All Done +Patch that validates all the interface end points for cluster entity +verified by Vladimir. +Once again IMHO this is not a bug itself but instead misuse of the output Petar +I added your code to the groovyConsole and I got an exception +Will need some more work +See for further details. +If that doesnt help please post to userSQLNA system temporary table space with sufficient page size does not existExplanationOne of the following conditions could have occurred +The data between jobs is created by the first job and consumed by the next so we dont want them to be deleted when the first job is done +Assigning to Rich to review for V +note that jenkins doesnt test macos so Im assuming youve run everythingthanks Binglin. +After chatting with Max it looks like my key listener listening for EnterCR events may be a bit crazy +Doug I think we are on to something +Since and files are essentially the same for all directory server Ive left my files only for ADS as an example +Im planning to commit this tomorrow +why not just use? otherwise. +This sounds interesting +Patch Attached +I just removed height on the meta tag viewport +Thank you Chris. +Ok I knew youd say that +All the junit test cases run successfully and I have already implemented similar stuff earlier +not a proper patch? WTF? +Hows this +Patch for latest for inclusion in +Jose Could you please upload the wsdlxsd in question +Once I get the closure I will finish the remaining code +Im tempted to get rid of the pooling for objects +Thanks for checking in ArunMarking as resolvedNot a problem anymore in Although Id also like small descriptions on whys if you have the time +I think that the SHOULD support the different Source subclasses +The principle I followed in the changes which I suspect from the code you may disagree with is that if youre sometimes locking access to a field you always need to +bulk close of all resolved issues as part of closing items. +Then it can look at time stamps on the files and possibly checksums as well to pick up where it left off on a failure +It seems to me that this issue is still present in groovy +Verified. +I left some minor comments on review boardNice work Skye +Could you please file a separate JIRA for the failure in +Ill look at netboot as part of work on AS deployment issues and the +This JIRA seems to have all work completed against it so should be marked as resolved in to as I dont think items for which it was by Ted have been completedwill make it into Fix For to Unknown on issues not targeted for issue out as part of JIRA cleanup +Im sorry CarlosWe did not find any solutionWe are looking for an alternative to the problem we had +Please check to see if trunk fixes the issue for you +Dependency for Hadoop project doesnt specify the scope +Its definately something you can look at and figure out in seconds +I fixed it in Windows which broke it on the Mac +is the full patch with tests based on Knuts repro +I have tried the current shapshot and I confirm that the problem is gone now the unmarshalling works as expected +Maven migrated to Eclipse Aether +Interoperability with Hadoop security is a prerequisite for any use of Giraph for our use cases +Its not clear from the javadocs whether any data value which is should result in the minmax values being too +Issue closed with release. +trunkCommitted revision +The reason this worked with before was that it instantiated a ugi depending on whether or not that conf parameter was set at runtime for every single authorization callSo we need to change the initialization order in the hcat commandline to make sure the D parameters are processed before we instantiate as well +We dont ship a or in the serverdrseuss tcunning find +This is my developer dashboard the idea is to populate SCM Commits per Clock and Week plot with only my information +if RSZKREGIONCLOSING is created It says that RS has received the RPC +are done in proper order +Duplicate of JulienThe issue you said was a duplicate of this one is concerned with the whereas this issue is concerned with the not the same thing +Well probably not necessary What Im expecting from the richFaces team because is not really an issue from RF code is that you update the documentation of this tag +Trying looks good +This has been fixed in SVN but we havent released a dev build yet +This can only happen when using on version above Sonar +Most likely as the timeout seems to indicate +Having no reply from the customer Im going to close this. +Therere some minor comments +serverdefaultdeploymanagementclasses +i dont think a security domain is the right access so that unauthorized people cant deploy processes is the right solution +Ok moving to we should probably decide if we want to change the method to be add new functionality +For example in our app we registered a bouncy castle JCE provider with even after it on shutdown the JVM STILL retains a reference to the CLASS +Sorry for my late commitNow Ive just committed this +Fabiencould you please try to rebuild your test instance and db and see if the error disappear +thats the reason for the error +People expect constructors at top of classes not utility methods +The cfg is final and required to be passed in to the constructor I should probably add something that ensures that its not null as well which would stop any further test bugs creeping in +Proposed patch +I ran this on a little cluster test and my results were a little mixed +Leaving this ticket here to remind us to work on it for in +Accumulo eventually has to do compactions which have nothing to do with the native map +Incorporated Ladislavs patch +I accidentally committed an earlier version of that didnt include the support needed here +Attaching a patch that illustrates the problem using the existing testcase +But not the other entities I think +Please confirm and close if satisfied +for urgent issuerelated to my jobemail eugeneyumobile no RegardsHao +No Validator upgrade +As discussed above. +bq +You may look into this issue I dont think that the OOM cannot be seen +This issue is there in public builds also +Please continue on it +SergioAny thoughts on this +This will be in the version +Will do +I mean +in the writer there seems to be a bug in Jackson that emits an extra space +We do not intentionally use shortnames +I just reviewed the patch and as far as I can tell everything appears saneMoreover there is a unit test that tests the fix +Verified at +foobarFdral we rewrite it as foobarFuEduEral that is correct but evidently not everyone likes itShould we never convert characters in the range if original property file contains at least one of them in entries +Sending assemblySending assemblysrcmaindescriptorsSending assemblysrcmaindescriptorsAdding jbiclusterconfigAdding jbiclusterconfigAdding jbiclusterconfigsrcAdding jbiclusterconfigsrcmainAdding jbiclusterconfigsrcmainresourcesAdding jbiclusterconfigsrcmainresourcesAdding jbiclusterconfigsrcmainresourcesspringAdding jbiclusterconfigsrcmainresourcesspringAdding jbiclusterengineAdding jbiclusterengineAdding jbiclusterenginesrcAdding jbiclusterenginesrcmainAdding jbiclusterenginesrcmainjavaAdding jbiclusterenginesrcmainjavaorgAdding jbiclusterenginesrcmainjavaorgapacheAdding jbiclusterenginesrcmainjavaorgapacheservicemixAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestAdding jbiclusterenginesrctestjavaAdding jbiclusterenginesrctestjavaorgAdding jbiclusterenginesrctestjavaorgapacheAdding jbiclusterenginesrctestjavaorgapacheservicemixAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestjavaorgapacheservicemixjbiclusterengineAdding jbiclusterenginesrctestresourcesAdding jbiclusterenginesrctestresourcesSending jbiclusterAdding jbiclusterrequestorAdding jbiclusterrequestorAdding jbiclusterrequestorsrcAdding jbiclusterrequestorsrcmainAdding jbiclusterrequestorsrcmainjavaAdding jbiclusterrequestorsrcmainjavaorgAdding jbiclusterrequestorsrcmainjavaorgapacheAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorAdding jbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorDeleting jbiclustersrcSending jbiSending Transmitting file dataCommitted revision +attachment +Seems I missed a wonderful discussion +However we dont need to rush to Guava just because we would gain a lot from having our own service base class be it our own or descended from the Guava one +Its not top of my list if you wanted to merge our two patches and post a patch against trunk that would be great +Hi Paul can you attach the patch that displays the use case +It now has an s on the end so it is not backwards compatible +Uses contenttypeamqpmap with string content +I like the new semantics I can easily wrap a genserver around the client if I need to and it seems better overall +bulk close of all resolved issues prior to release. +BobThere is a setting available to set the default time zone look for +committed. +is there a way to replicate this? an easy way to replicate it +and encourage vampires and was applied to and trunk +Bug has been marked as a duplicate of this bug +The sample has been added. +Added the test program +If someone really wants to write its own implementation and not depend on camel he could just exclude the dependency +Or we can have some conditional code where we can check which datasource is being used and then generate the sql accordingly +Second implementation +The problem with the ftpserver jars is that theyre not yet a released Apache product +The changes for this jira are made upon the changes for which covers some critical to upload patch +Backed out the patch and reopened +More on this later if Im successful +If it is okay I will add it after finish buddy testing +Ive revised this for Spring as well in line with the generally stricter handling in Spring +After I stopped refreshing Project Explorer from I got Illegal State exception with no Tools in stack trace some resource listener still run tree view refresh and still fails problem +trunk patch conflicted somehow with +As far as I know the main reason is that SI requires Spring and the project that uses it is not likely to migrate to Spring for at least a couple more months. +Everything is committed and merged into so its time to close this out. +That is incorrect +I dont believe theres something the framework can do in this case +Only supported by IE? Or is it a standard tag +Now the script handles compilation of redis source and make the redis executables available under redisbin directory +I tried to modify a static field on RI and got Field is final +Make patch available so get reviewed +duplicates this issue and was where the backport work was handled +code checked into checked into HEAD +Given that the two parts of the query are in conflict Im not sure that it really matters very much which part of the query Derby disobeys as either way is wrong and the only thing that can be done is to reject the query as ambiguousIm attaching a patch proposal with some new test cases for the ORDER BY test suite +dims +Added it to the codebase +There is a problem related to xalan which makes transformer creationreset expensive but there is a simple workaround +Against svn +Let me checkthe patches maybe my irracible capital letters eventually result fromsvn eclipse failing to capture themtheinexplicit +Ive got jars already +It suggests there might be a bug somewhere as hinted by DavidThanks again +Perhaps this is already fixed in branch or there is a specific EAR setup that raises the issue waiting for Dennis clarification about the EAR used +Stefan please feel free to this as required but I think we have provided sufficient information. +Very clever thinking +What about a user list Dave +So it looks ot me like a Websphere bug +ex +This some sort of configuration issue on your sidethanksdims +Folks that also do please have a look at +Bulk closing stale resolved issues +Wording changedClustering wont work with the default uses HSQLDB as the default database which is not a a Messaging cluster to work all nodes must have access to a database +Patch is attached for given description +A few minor comments The annotation interface needs some javadoc describing how its used +The ideal thing would be some fancy refactored approach that didnt duplicate much of anything +Thanks Mikhail +Patch to fix the issue with upgrading to you know if available in Maven Central or will we need to wait for the final release to be able to get it via maven +Looks like Eric has a patch on that jira +Cleared via offline discussion +I dont think Subclipse guys will fix it +I think if we are considering putting out a release under the license we should get approval from the PMC +well I definetly acticated it because i also start a code generator by another profile and both have and the code was generated but the multimodule part was ignored +I am changing my implementation to not test against the Shell API +Im resolving this one as not a problem given the discussion in the jira and please feel free to reopen if needed +committed thanks! +Sending srcmainjavaorgapachegoracassandrastoreSending srcmainjavaorgapachegoracassandrastoreSending srcmainjavaorgapachegoracassandrastoreSending confCommitted revision +If they decide it is okay to use algorithms in the ASF code that may potentially be covered by patents held by Microsoft the matter would be settled +If we solve the problem with perfomance of XHTML Validator then we will able to resolve these issues +Fixed authentication problem in trunk r +So I am not repeating those here under the assumption these will be fixed thereCan you please upload the results of and capacity scheduler testsAlso can you please make sure theres a patch that works for the branch +Try to workaround that +What is the proper way to shut down the server +OK I think all of the packaging issues have been resolved +I am a little uncomfortable about how is handled +Heres a patch +How does it make porting of the feature hard +This is really needed for overweaving +bq +Actually having index readers open at the same time is not as costly as you might think +AzeezPlease click on granting the license to ASF +In S that is easily accomplished via the XHTML theme overriding the file that I created +Pig will not have this issue +Minor tweak to the previous patch to include in the fix +This is a facelets warning and you should ask about it on the facelets user mailing list +Patch for trunk attached +Committed to trunk with revision I also plan to backport it to +Is there a problem with this patch? or was it just unnoticedId be happy to replace it with a more version if its not compliant anymore with the repositorys guess that everything in this patch except maybe the beans adding should be a valuable add to the next version of Spring IDE because the refactorings result in a much cleaner and code +Ill close this issue as invalid +Fixed Updated stratos installer to set keypair path in stratos controller cartridge config properties file + has been tagged and it has passed tag is GA +I update the patch +Sure I didnt mean to waste your time here +This bug has been marked as a duplicate of +wouldnt disabling this automatically upgrade to new releases? That seems undesirable to me +out that needed to be modified +Resolving as duplicate of the issue here using out of date XML parser is handled in Hadoop +ZhongThanks +The screensharing has a fps of +No work has been done to make the btree work under these conditions and there may be hidden problems that dont show themselves even if just removing the assert looks likeit is working +I also ran npm cache clean but that seemed to make no difference +I prefer the manual soln by far +Can you specifically say what sort of an issue that you are facing with implementing the pipeline pattern +Going to work on interop first though +Linking which is somewhat similar +that is not what we needwantWe should be able tot update central without this site +One question I have about the port is whether the removal of is appropriate for the branch +Looks fine +just to set release notes in +In such scenario is the responsibility of the AM to detected the liveliness of the container process and react if it is considered hungLike I said we do have an implicit liveliness report process liveliness +I get the same problem when compiling on the command line usingConverting to a groovy bug +Its pretty close but I can envision some scenarios where it might produce bad results +We simple get a completed EAR file with webservices generated according to specifications but now we simply cant deploy them on +In both the jars against or the jars against you will note a freeze +Attached a patch by properly defining path to the review commitThanksMalith +Decided to keep ALL zips through M so pushing fix target out one milestone as reminder to revisit then +There is a new page I have added for each FAQ +What does it meanSorry I chnaged my comment above +We are using on top of SONAR use the plugin for use with Sonar +This works in trunk now. +Could you please provide more details on this? This seems odd since the inner transaction operates on its own JDBC Connection with a commitrollback there not affecting the state of the outer transactions Connection at all +Commit r inadvertently deleted half of the feature and it was never fixed +Sure I will submit a full patch in a few days +Previously I was running on +I was building projects with maven but they also can be build in eclipse using ADT plugin +Marking for +Or at the least limit the current patch to attempts at changes applied to META +Ok I see the issue autoexport doesnt see the Index page updating at all +This bug is probably already resolved but I need to test it to be sureCould you tell me exactly where you set the IVYSETTINGS variable +Can you provide a standalone test case which duplicates the problem +Contains test guess you should use a descriptor to add the lib folder to the classpath in this case +Ill close this issue now +Input as part of down from blocker and setting to think this is fixed +Committed +This patch also contains the changes for +Nice work +Administrative clustering what operations need to be across the whole cluster at runtime +Slipping Java SE issues we dont have time or scope in CDI to address them. +Very easy to someone has a patch that will accelerate the container ports are checked when shutting down +Until a complete solution is available the quick starts already contain an example showing how interceptors can be used to change the identity used for EJB calls instead of mandating the identity of the connection +We can also remove and since we have now +corrected some bugs in output +Now when the IDE boots it puts up this errorWarning could not install some modulesMevenide XML Grammar The module named was needed and not XML Grammar The module Netbeans Manve project support would also need to be Debugger Bridge The module named was needed and not Maven project support The module named was needed and not Maven project support The module Netbeans Debugger Bridge would also need to be Maven project support The module Netbeans Run Jar Maven Bridge would also need to be Run Jar Maven Bridge The module named was needed and not of the mentioned modules are diabled and cant be turned on and the Maven support is disabled +It supersedes the previous patch +I dont see how the first half of the patch can solve the zk issue +Rolled back the change to Resource and included a FIXME that notes we dont cover all cases +On there is a patch which uses Kryo for serialization +Because of the check the poolSize will be set to minSize regardless of the true setting +This is already done in the webHDFS unit test I added it was just missing fromNew patch shortly +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Ive tested further with LiferayThis seems to work ok now +I tried it +You must either enable user authentication or disable the installation of a security manager +I think it makes sense to apply it also to and release as well +If you beleive that should be please and contact the program team +If folks agree I will make that change in the next patch +So the property set on the fo inside the is ignored +Thank youI know what I did wrong when I submitted this bug +Ship it +Thanks for the patch Aaron! +With this patch the maximum interval between two consecutive checks for new notifications is minute + is be possible to provide a diff file instead of a zip of the sourcesIf makes it easier to review the changes and also more resilient to changes in trunk +If you find further issues please create a new jira and link it to this one. + +Good idea +I have removed these changes fromthe patch previously marked as not for commitI ran the modified and it ran without failuresI have started a junit All run and shall revert back withthe resultsPls consider this patch for reviews and tests I had started on Saturday did not complete I have started a new test run today +Following tests failed on RB +We will consider it +So it needs to be done now because EAP is going GA very soon +I will check if it still makes sense to use the LIRS cache +we now go through for the headers which requires a am concerned with the performance implications of doing this I need to understand whatwhy changed in Spring +Lets get that worked out first before we work on tests + Pig has some issue +for this seems to be designed for inheritance but really isnt since most of its methods are casting their arguments to the private do support inheritance for propperly! +i think it will make the code much more clearer when we get this inim wondering about the asymmetrical nature of the current zk fd that youve already mentioned one side is generating traffic and the other is responding +please disregard previous attachment +It gives at least equal if not better response times and reduces cpu consumption of the datanode process by depending on the system and loadIs this a viable approach? If so Ill forward port it to and I can retest if I can find out which works with hdfs +A recheck in the correct tree provedyou right. +I agree with both of the other fixes you made in response to the failing tests too +I tested that +net version behaves +Therefore the type has to be consulted to determine if it is an Enum +The other two options have to wait though. +I cant test it right now because Im not using Seam from CVS +minor change to make more compliant with Maven previous comment should start I just uploaded a patch with a +Pete ok Ill close this JIRA and reject the pull request then +Fixing this will require some work +please change link type if not fit final +Can you please point me to a link where those directions are + for the patch pending Hudson +Thanks Colin new patch attached +Fixed by manual instrumentation of the model +Thanks Owen! +We currently use SVN on the changelog plugin version how can we retrieve it? Were using maven which is taking care of downloading the appropriate version of the plugin itself +Isnt String dimension limited by the length value? Due to the fact I think its saved as an integer we should have up to unicode chars +As I known we use CRS all the time and it +Public in what way? As for timer +Verified +We will improve the API docI agree its arguable if the response should be buffered at all in get +completed +Wow lots of good content Unassigning from me in case that blocks an actual page being created +This is how for example you would solve this in m in favor of anything that will allow to merge folders without throwing an exceptionIt seems like adding a new action is more intrusive than adding an optional attribute to actioncopy which is already implemented +This code in the local tx demarcation strategy should only be invoked if the custom connection factories defined are not xa connection factories +The is not indexed +And since it depends on API additions which will be made only in master this issue will not be fixed in maintenance without substantial lobbying +We could really use it. +Also added a comment +So we will be able to change things in the future as long as we have some backwards compatibility story here +update and you add some tests too +Its frustrating to see huge legacy projects and those results get rolled up with your own project code on the Sonar DashboardI would hesitate to filter anything out especially whole projects since it could easily be used for evil +Resolved in. +Awesome work Andrew +Thank you for your commentI think that can move to +The component definitely shouldnt need those attributes I might have just inherited from a wrong base component +parameter HADOOPRECtaptag can work nowthe problem is has fourthey displayed the same nametake the current value of and append the sub job s mapper and reducer class name as the sub jobs name +Should I start another JIRA or should this one be openedhave you tested on a machine that does not have MS office on itHave you tested on a machine that only has Open office on itI believe you will find the error under the above conditionsNote My Customers are using MACs with not MS products. +checked in. +on the patches +Gleb plz correct all links within the guide +If it can be made so that all access to the action either via the class name or aliases can only be done through securable paths then that would be acceptable +Rgression tets are successful +I would prefer to leave dialyzer rules not localizable because in long term we should get rid of dependency on tool and replace rules by our own +KeithAmilaI just tried it with the sn option pn sn o src d xmlbeansuri I got the same error Transport error +Open Type is not visible when I open Git perspectiveI havent used selective enablement at vieweditor level how is that done +Ive removed all jar This works perfectly in Tools Alpha Version +The columns are the various output format options from gnu time some more useful than others +Hi PatrickThis patch will have to be reviewed by a committer +But Id like to hear other opinions first +Doesnt turn out to be the cleanest code in the world but it should do the trick +reservedinodeid and it is left unchanged then you are allowing a sticky symlink in the sense that if the normalized target file is deleted and recreated with same name then the symlink will never follow through because it will have a new inodeId such a symlink is not supported in any Posix system today +Yes I can confirm that it is working in our setup using the snapshot regardsJens +This diff fixes the issue +Please see my attachment jessica +Choreography monitor is no longer bundled with the Savara Eclipse tools a server side webapp will be provided instead +Thanks for adding me as a contributor and checking my patch JingIve added a new patch +Forgot to unassign +As a first step I rebased our changes to current trunkThis will require the usual recheckout due to operationsNow to the problems you are seeing +To get around the problem I removed all imports inside my schemas +I have committed this +DavidThanks for your input +less memory +The other two tests dont fail on my local machine +Thanks Mahadev. +I honestly really like the idea of fresh starts with people consciously modifying classes for and putting them somewhere new and shiny +This problem is fixed and that is reflected by this issue being Status to which the comments from Carlos seem to relate is something entirely different a failure to execute mvn eclipseeclipse due to resolving an incorrect version of the plugin in the first place could someone with the appropriate permissions change the summary of this issue to Tests fail due to incorrectly attempting to use released plugin version not locally sandboxed one under test which is a brief description of problem +Closing the issue as it dose not requite any testing +I will test it with once it releases and reopen the issue if need beThanks for the help extendedRegards. +Followed process mention with issue and Implemented test case for issueIn which assert tag is used to test desired outputThanks to Ratnesh Upadhayay for their helpThanks and RegardsAwdesh Parihar +You are right +This issue was reported in October +Simple refactor to move the throttle to compaction controller +Knut please file an issue for that if there should be one +Also as Martin found the contents of under are strange +Minor note there is no need for inner xslchoose element +I cant actually run it wo TREC but the tests pass +Ive added a unit test and fixed the bug +Problem fixed in both SVN trunk and +Not able to build from the maven repository for zookeeper SNAPSHOTCorrect me if am wrong +No problem I would just prefer that one does not need to enter multiple items on the command line to get this to work +Guess what? If it was that important to you implement it and submit it as a patch +Patch committed +Are you aware of this in the course of beta testing? Otherwise I could prepare an according test for this specific ProblemThanks againPhilipp +Rayees can you try this manually to validate if this is a script or product found this issue in latest automation run passrate in volume test in vmware closing this defect +Select boxes of dashboard periods and of violations drilldown have been replaced by select componentOn quality profile bulk change the select box was already a select component +Here is a diff with upgrade to upgrades to upgrades to +Fixing this by having the ant task support security credentials is preferable to replacing the war file Ill keep the ESB task open and fix this in our codebase +Verified with latest builds +Hence the visitor was changed to ensure that is annotated with information from all Local rearranges +Adding to the in the example makes it run just fineThat just leaves support for pipelining of long polling requestsIm not sure how Id expect it to behave +UliI kept digging and found the missing key for my environmenttrueWith this in the run configuration all works without it the validation does not work I was running due to jboss being in the classpath for the http services which should not have been +This is an example patch not to be committed. +Only the st commit made it +RAT should be used to verify the license headers and to create a report +Committed to branch in revision . +Hi LilanthaThis is working for me too +Stored prepared statements are supposed to be thrown away and not when you go across upgradedowngrade boundaries +I agree it is a good feature! It overlaps somewhat with broker replication which makes the partitions themselves more highly available but they solve slightly different problems +Its easy to end up with recursive calls. +I could process a little further but couldnt get the issue resolved completely +readability +We need to carefully test it to make sure unit tests pass +Let me attach it. +It blindly assumes the latter while returns the former +from me as wellbq +Switches tests for service from to fixes Hadoop tarball URL for those tests as well +trunk This was a tough one to crack as onCompletion onException and intercept needed special careBut the tracer is now fine grained and shows much better the traced routing +Hi Craig any progress on testing andor documentation +I am looking into it now +It has been fixed by the refactoring of Cache process done in other issue. +I dont think there is an issue here +Because I dont want strongly depend on meclipse Maven and plugin should work for projects too +The brokers plugin system has changed in and a side effect of this is that the files arent ending up where they were before in the broker release package but also arent ending up where we will probably put them eitherAs a work around copy the and jars from qpidjavabuildlib into qpidjavabrokerreleaselibplugins and try again +verified at revision +This needs to be fixed +Furthermore we can encode counters simply and directly rather than complicated nested encoding schemes +A st pass through raw types usage +Also if you are using ntresource in JCR if you set the mime type it will index it correctly +What is the next step? Should I write a new patch taking into account all the different comment +Fixed on closing all issues in resolved state that have not been reopened nor commented over in the last month +It might provide some insights +It would have the downside to the token being displayed in every DN link on the NNs node list but the token is display in the link when at the DN anyway +Improved patch now accepts property inDavidJust to clarify the original behavior has been preserved +The whole line was an addition just removed it +Can we use semanticsyntactic predicates to support the optional brackets +verified +Tested in Safari and Firefox on Mac OS Ive checked that and its also fixed in the next release +If you do this then you should not see the library problem described aboveIs there any reason why you are not using the standard update site +Calo unless you support evidence that the RA is required to serialize factories this is not a bug +I am trying to wrap a Java application using JVM option and when stopping the service from Services list in Control Panel it eventually stops after a long wait with the following message in a popup dialog boxCound not stop the myService service on local ComputerError The service did not respond to the start or control request in a timely fashionThe OS is Microsoft Windows Server Standard x Edition Service Pack and I am using the AMD binary +I always forget that +Ill check into this +This patch replaces them with functions from the re module +Committed revision on branch. +This appears to be the case as different versions of the are used for compile and runThis is inconvenient as the workaround is to edit the run configuration delete the default classpath and manually add all the libraries except substitute withA simple include groovy config hopefully would do this for you + for the trivial patch +Thanks Kiyan + This is probably not a Major bug +And I am familiar with the java platform in general +For the left outer joins where the parameters are lost is called which constructs a new +Closed as duplicated. +Also I wonder if this functionality could be part of the cas editor plugin itself rather than making a new plugin +I just extended it a bitThe default ports where useful as many mail servers are configured in a standard way regarding theport settings +Submitting to HudsonCan someone look at the code to see that it is OK +It would be ideal to release Derby without any known regressions so we dont have to even think about where to draw the line for acceptable regressions so before I think too much about where to draw that line I ask How hard this is to fix? Does anyone plan to pick this up for +Attaching a patch that adds support for High Ram jobs in Gridmix +Added gzip compression support +Although Tapestry has moved away from blackbird for Ive created a patch against the current trunks HEAD that adds a method to +Ill take care of committing and apply it to at least trunk and +As to specifying the location of I assume this is settable although I do not know how +I agree with Jonathans sentiment that we should try to fit this kind of thing into a framework rather than core if possibleRegarding the use case of it is a requirement for a lot of government users where each piece of information may have a different security clearance and clearance is very granularly controlled +At the moment the task adds the timestamp +Appreciatecould you please upload the final version of the tagsThanks in advance +suggested this on I +Verified in +The code of this dialog is too old +The patch passed the jdbc testPlease review the patch and let me know if you see any problems +If you can make a reproducer feel free to. +I think you are using wagon ftp and not adding the right dependencies or not using the latest wagon you dont give much happened when one of our users tried to build a project using a just installed Maven +batch transition to closed remaining resolved bugs +to trunk +I understand the motivation for this issue but I cant see how we could implement it +Thanks for the review +The headers were the old ones and its easier to have only one patch +The original case was filed in error because I had changes made in my client. +updates might take time to actually appear on the websitethnx nadya +Hi Ivan I dont have commit permits but Im testing some of the bugfixes +Im going to test it under various OS DateTime settings to see it it changes the behavior of the widget +I have reproduced it with Manuels think we shouldnt send out of bounds matches to the formatter? Because all it can do is bounds check and skip itI think maybe we also shouldnt even send the passage if it was truncated even if some matches were before the truncation + moving to +Should commit to branch +I do think Gert did some changes in Camel related to some concurrency issues with a component similar to this one +Problem appears to be deeper than the related bugs +rename the property as requested regenerate the patch with relative paths on a unix based system as commented the exception catching is necessary and is also done in other exsting code that iterates to find a free portTHANKS DOUG FOR TAKING THE TIME to get this into the sources +I will try on working on a testcase +Will backport this to will work on backporting this to a patch which will log the Engine shutdown message in +Everybody hated maven so much that we chucked it for ant +From there I will also check where it hangs on my machine to see the impact of policy file changesThank you so much Kathey for looking into this +Verified on. +Thanks +I will check how much time we have and see if we an do it for this release +We should not be backporting this until its behavior is actually finalized in a release +Shantanu says that this got fixed with his patch for bug please test and let us know if there is a problem with or nightlythanksdims +just based on the description thought I havent tested it +The generated builders by protobuf dont allow creation of incomplete instances +If the file path is within one of the storage dirs it gets marked as failed +Changed to and added a sc test +Youll need to declare your notifier in would be best built on a full plugin system that can contribute data and resources rather than jsps +I have started working on OM implementation still I didnt get any feedback for my last two code contributions Further it is required to apply the patches for JIRA in order to work correctly +Patch verifiedThank you Oli your comments make this patch better. +bq +included in patch sent with +I think this patch is ready for commitI agree that at this stage it makes most sense to leave the issue of validity of current row after an update to a separate Jira issue +The idea is excellent +Passed +Right in time Still I would need a quick we have a quick skype session this afternoonHeiko + +Each of the subprojects need to be maintained from each other +fix for patch +Closing this JIRA out based on test comments above + detects an already modified fix this bug its very release does not cover +already fixed in +Incompatible since this changes a public API +No tests are needed for I believe +Patch committed to trunk. +Attachment has been added with description Sample to demonstrate the noticed a mistake in the description the number is missing from after the word page when describing where the Continued text is expected to appearFixed with rThanks for reporting +reopening to incorporate refactoring to be consistent with executionService operations on variables prevent of ugly the issue due to found bug that history variables are not updated when execution variable is checked in to secure that history variables are updated when execution variables are changed +Removes from the mapred packageIll create a new jira proposing that we use the binary file format instead of the for job submission +I originally saw this problem in eap +Im going to mark this as Cannot Reproduce Bug instead of marking it as Done because I dont really know what fixed this issue. +revision all old Resolved issues +It should be able to figure out dependencies from the EJB injection annotation though right +Will file a separate jira for that and clean the above as part of that jira +The worst that could happen is balancer ran meantime +Test output shows a similar failure to escape special characters +thus not relevant to put on simple enough +Unfortunately flushing the cache could cause existing Java objects with the old references to get confused about who or what they m working on various overhauls of the proxy class stuff in +When VM shutdowns it frees and zeroes a pointer to vmmethods table which contains all compiled java methods information +IT updated. +h have to be removed from svn +Just for clarification +I deleted that file in our repomanager and now it is working +Im going to proactively mark this as resolved +Im thinking that this is really the same problem that I had with where removing xalan solved the problem +Each token also knows how to renew itself so bundling it up with its renewer also seems unnecessaryI think its back to the renewer is a singleton +Is it really required? On a small cluster is there any harm in reporting every second? Id rather try the simple proportional approach first and see if it is not sufficient +I believe it could be easily merged in the rampart head as wellRegardsStefanThanks for the patch Stefan Ive applied it with some minor changes to the branch and trunk +I will like to get this patch commited to in its current formRegardsCompleted At revision Thanks +I just committed this +This should have been closed for I guess that means I missed it when I generated the web site with the report listing the Jira tickets resolved in the releaseThanks for spotting the missing mergeFred +Resolving as it works fine here +I will think of a better way to do this and roll it into the next patch once I hear back about +I could have just used if it could deal with the parts of the output that are variable +You actually have to turn this behaviour off if you dont want it using the use configuration close this issue I believe that it is no longer relevant +Is the previous paragraph about plain beans of nonDependent scope enforced? I think it is checked on some other places in OWB. +Thanks for coming back +Committed to fixed in and trunk +Could you clarify what this is for? The description is not clear to me +fixed in svn rev +The module poms are great and at we ll clearly state to our users not to depend on the full jar. +BTW accessing a property directly with +I will give you a shout once I have it doneThanksBrian +Hi Stack I have rerun the tests and got no timeoutBefore killing the process I should have run a netstat to see open sockets and identify if the process was responsible for running zk or something elseI just will remind when getting timeouts to further check the running processopen ports. +Use whatever approach that you deem the most appropriate for your particular application +Apparently is almost ready for commit as Ive heard from RaviHemanth +Merge the notification of and if there were more than one messages or timers been created in a transaction wont be executed twice +They are currently used by MR apps +I was pretty certain this wouldnt be a problem but I added an integration test to prove it +Search tag elyogragspringclean +If others are OK with this patch I will apply it to both and branches. +Starting work on this now Ill try to combine it in a way that is similar to the component and work out from there +I used the example of Sing Li at IBM developerworks httpdeveloperworksviewscontentidfilenamemethodftplocaleworldwidePlease try the third example with the CMP EJBThe problem should be very obvious if you look at the code section I mentioned above +A connections version is determined outside of the class +Just reporting the error +Weve improved this greatly over the years and aside from a few reported bugs we have not yet fixed I consider this fixed +I have now merged this code from the code line to the codeline so bug is fixed in both versions. +This is a pretty easy one for someone to fix and as more users use more watches it would be good to get this addressed +At least one case that I still remember needing reference count if multiple backends need a Channel you need to make sure we start at most one channel +I noticed the code in question has been rearranged quite a bit +Close this issue duplicate of. +Its great to be able to implement the same features but I wouldnt want to see cloudstacks innovation wane and just become a clone of the big guys +This copying occurs once per statement execution +We are removing all of those spec classes as we are moving to the new pipeline design +add code for spotting that some internal generics information was leaking into the bytecode. +ASF BZ +Whenever you make an official release the corresponding doc can be made available online +does not include added +to create the list Tks +m having trouble applying this +line homeusergitsqoopdisttargetbin No such file or directoryHere path is incorrect +Im closing the issue for now please do not hesitate and reopen it if needed. +I will incorporate all the changes suggested +I am not sure if we support then yet +fixed in the new gui +However for uninstaller it would not check the condition and rather execute in privileged mode always +I might ask for few clarifications if I need help on any of the pointsRegards am attaching an improved patch for inheritable attributes implementation +This can be deferred til the next release this is in progress it has been done but Brian keeps on breaking it +Maybe in the future we can support such functionality as counting the quota usage for snapshots under a specific directory +I think Ive found the reason the Invalid log position warnings were being generatedIt was being caused because log files there were referenced in by recoveryreplay records had been before shutdownThis could also lead to log files which did not get after restart +If any one need some strange things it will be able to use different providers +I revised my properties and was able to get this to buildIll try out the patch and post the results +At least that way the leader selector logic is explicit which was the purpose of the change +Cross posting from as this is related to the patch attached here +Simply extending the regular expression with an OR will not do the job for Apache Harmony +I downloaded the RF source code in order to find what was being executed in Weblogic that was not executed in Tomcat and found that in Weblogic the get method from is executed twice for each JSP that I try to load in the web browser also the getWriter method is executed a few times but in Tomcat this method is never executed so the useStream flag never changes and the exception is never facelets the get method is never executed no matter if it is Tomcat or Weblogic and the getWriter method is only executed once so the flag is always the time being I have commented the line that throws the Exception in the getWriter method and at least it is working now without visible problems but I dont think this is the way to go +This should fix it +Booleans are encoded as constants +Ill take a more detailed look to see if there is anything there that Id recommend changing +Marking closed. +Attached patch that adds update stuff and fix cset and uiparam +Also tested on KVM using primary works as well +mrepository maven would fail the IT tests with works as expectedWhat would it take to propose a release of maven with the newer aether included +prompt ll probably have this finished tomorrow +war plugin could generate empty if there is none in the sources +To the first pointThere was a issue solved in a later release addressing this problem please update to the latest stable release and open a bug report if the problem still exist +resolving the issue since the user reported by moving to SNAPSHOT resolved the issue +Ill try. +Used an existing map processor person instead of map processor of ecommerce in create JyotsnaChanges are in trunk at revision Ashish Vijaywargiya +is this Forge or forge ? Assuming it is Forge and thus if reproducible critical to fix or at least investigate before chance you can get a jstack when it goes infinite loop +What environment do you use? Im on ubuntu linux bit with java +i guess that solves my problemsorry for not doing my homework and needlessly raising a new feature request. +Committed the changes to files relevent to in to branch M metastorescriptsupgradepostgresM metastorescriptsupgradepostgresM metastorescriptsupgradepostgres +I did it just in case and because it is easily thanks for handling the problem and looking forward to release +Gary what is the best way to exclude something based on platform +I think we need to wait for the better plugin to be released +I delted the old test file and canon as they are no longer neededKathey +Are you still on this Nicolas? Whats the current situation like after +Your patch looks good thanks for cleaning up the code tooAfter some initial debugging it seems like the problem is somewhere in the read pipeline +It include following feature and enhancement +Yeah I used Void Sale in a couple of different points through a sale and didnt see any exceptions appearing in the log +It seems that you have made some changes to the model objects and have not sent those as part of the patch +I dont think its very common or recommended to bounce mail +The tests passed on derbyall and LilyCould you make sure something reasonable happens if there is no permission in the policy file for reading the system properties getting the source location etc +Reproducer supplied by Pavel MoravecRun the reproducer in a loop and you should see the deadlock manifesting +I still cannot reproduce above error but I believe this was caused by the change inget is used to determine the sections to transfer inside sstable but the method returns incorrect sections for some casesIn fact file that I got from Brandon showed that the node was trying to stream sections way bigger than actual sstable file size +This is completely should run for integration tests +We would call this person a tour guide or tour rep +I would appreciate having at least some notes pointing to this issue from the testsuite so either it will be checked somewhere in script and printed more helpful information or add some file containing such information +theres no awesome solutions to make geoserver use less memory on big meta Tiles +And more. +for the patch +Hi HadrianAny further advice on the investigation of this issueKind regards failovertcplocalhost as the connection URL is the correct approach for dealing with reconnection to the broker in case of broker failure or network issues causing temporary socket lossin or later of the connection factory and camel component use failover in the default URL by default to avoid folks having to know about the failover configuration +Of course it will work +is a good exampleWithout a test its impossible for us to find the issue and fix it +Committed to trunk +I finally got a chance to test and it worked for me +Then when opening the sstable it could be easy to get all the components irrespective of what really created themWhat do you think about it +Could you please log a bug therethanksdims +I have an issue with the I tried to workaround it without success +I have attached a pom that will work for your situation +Yes this is incompatible in the sense that currently NN and SNN can communicate if SNN uses local host the default value for its http address while after the patch they wont be able to +work thanksis there a special reason why its not implemented for handling sling running on different context than +I had a look at the code +Modified publishing wizard to allow the user to type in a workspace area +bq +This was completed for +I have a need for this too +Daniel can you explain this more concrete? How will the new configuration file look likeWill this introduce new dependencies in minimal deploymentsIts ok to start investigating +Please reviewThank you +This lock is checked each time the process is invoked and the whether to continue processing is will based on obtaining the lock on this may be expanded when shared installations and such is supported until that time this effort is sufficient to satisfy this JIRA +I need to store references to these wrapper objects in a way that wont prevent them from begin GCed and is the best the JDK offers short of rolling our own +As seen the Previous replication is after the config files replicated +what do you mean this is not the case? Adding the made go awayare you saying is not applied when sitemesh renders the layout +So I am here +If you cannot convince me that youre right and Im wrong Im just going to close this issue as not a bug. +Patch to modify addssrc +Seems fine thanks apart from Need to add the attached to toolssrctemplatesdoc should be renamed to +what is the solution? +Here is the signed agreement. +Any opinions about the value of fixVersion? +I havent thought about side of changes though +but overally it is good to know that it improves things. +This is unlikely because the hudson build intalls on a clean build of m pretty astonished too but I tried downloading the binary from the website run the installation from the binary distro and got the error the user says +RI has intermittent failures as well actually +Thanks for pointing that out +also add a will be removed in to the deprecated in do you think it is best to just remove this class instead +This is avoidance not resolution +Ive removed the Override comments but honestly this compiles fine in my workspace +Proposed changes +I am sure many people do so +I will take a stab at reproducing his original problem but if I cant and I dont hear any arguments about why this change was made in the first place then I will change TMSUSPEND back to TMSUCCESS +Actually is using the specification version of the package which is read from the file by the Java Platform when the Package object is created +Thanks RamThat seems like a definite improvement +Sorry revision was it seems the fix wasnt working after all +Verified +Bulk close for Solr +So far I havent seen this with CR but I assume its a matter of time +I will give a try with weme +working as designed see linked issue. +since we dont catch itnew patch is coming +This would be a nice task for the new Load interfaces as it removes the complexity of dealing with something like Owl +I have changed the class directory of parallel algorithms +No problem +Streaming writes file directly under the data directoryPatch attached +As discussed +I agree with you I think that console error messages have to be reviewed in general +Unless we can figure out more specific cases Im thinking that this is fixed since we now support Show logical structure for Reference objects by defaultSince now supports debug evaluation this problem is not really too bad any more. +Colin I think were good to go and you can commit any time +defer to so I can release in JIRA and get the release notes generated +jboss not +Why does IP need to be in and not in metadata? +Philippe +new atomic increment for efficiency only increments up to bits total +about documentation on will do it asapabout removing It is used a lot in current code would you replace it with plain calls to? has the advantage that it always emits remoteAddress regardless of filters in the chainWe could inform people that they will only see the remoteAddress for loggings when they have an in their chain +The French marge de manuvre is marge de maniobra in Catalan +Patch for the improvement +I have commited it to andWas it commited to +JonIn our case the split got completed and the RS went down due to ZK issue and that is why the Master was not able to respond to the split region completion +pending Jenkins +adding the rules that need to be rule to implement for information to remain current namelyo changed names of classes reference to the ones createdo corrected the actual object types pushed and popped off object stack +Dig in why +Merge is being worked on but delayed by issues that should be resolved soon +Ship it +All tests pass +Ok changes committed and closing issue +You can can close it now +Carlo de Wolf reports in I see a hang for about seconds and then cleanup continues +The patch is nearly done Ill submit it later today +Can you provide some sample code showing this? What sort of item renderer are you using and what browserOS was this on +Its extremely hard to get going like this +Ok I understand why it has crumbled slightly but I think if the code is there is would be a huge waster if we didnt try to revive it possibly getting it integrated into the code base and maybe getting it added as a contrib component but not shipping it within the core codebase if the former was not a viable optionIve had a look at but I think we can discard this as it was addressed a very long time ago its also already integrated into the codebase +Someone pointed out to me that my comment is a bit confusing +I know this is going to sound melodramatic but it is exactly how I feelI give upI have spent more than a week trying to fix problems with this code and expand it to support locales and time zones while keeping it backwards compatible and I have come to the conclusion that it is a lost cause +Then flags is probably better than flags AND multiple methods +After a bit more thought the code can not throw an assert during undo based on page versions +The tests pass in because the check that checks that the specializing bean has all the bean types of the specialized bean was added in +Fixed thanks Martin. +The ampersands in the urls are normal delimiters for the request parameters no +Resolved under r +Is this related to +The strange thing is this code was originally developed using that version of gcc so Im surprised it no longer builds +This patch remove the native statements to use the Java driver +Closing issues which have been released. +Sorry about that +My concern is testability and maintainability +Should we revert this patch in until is fixed +Modified to not continue processing the filter chain if the has already handled the request. +srctestjavaorgapachewssecuritydomcommon +Thanks +Does existe any roadmap or workaround about this issue +was chosen to allow you to target weekly repairs plus a safety marginFor every user who has trouble running repair weekly there is at least one user who wishes Cassandra would reclaim his disk space from tombstones faster +I use JDK andToday I send the an email to to my lokal Apache and everything works fine +Further investigation will be done in parent issue. +Where are we with this? Can we close +Browse to httplocalhostwebappbutton and follow the link there +If you need help please let me know +Freddy youll find the screenshots in the attachments +Evgeny you were right +If nobody is going to do anything with this Id like to wont fix this task +You may even make demos by slowing it to max +Default time interval to refresh new content about minutesWe dont want to dig and change too much code of Shindig +If we can find a way to avoid setting them to null in default that will be greatBut the problem seems come from hadoops Configuration or Javas Properties. +So all issues not directly assigned to an AS release are being closed +patch looks good +Are we going to deprecate contribregex with this +Also thanks for the investigation and patch +Todd should we run before committing changes? My preference is to do that +I dont entirely understand what the patch does but it does indeed work +Resolve Out of date +Bulk update to fix version +See for more info +For VTI tables I followed the mode for views and the compress code ignores them +r breaks it to be a class loader issue +It might be best to try to overlap reading from the ledger with processing the edits +I have a patch but will rather take a stab later if there is anything to improve after your commit +Im not able to replicate this in trunk at least +Forgot looks fine to me +tests included +IS not intended for this issue +It was somewhat difficult to add a pure unit test case but it seemed to work with a couple dummy parameters. +HenrykCS Check Style dont worry it just some package import order and some code format thingIf you are using Eclipse it could be easy to tell the error from the IDEAnd you can using mvn clean install Psourcecheck to check if your patch is OK +Resize works wrong +Where do you see this not being merged +Id sooner shoot myself in the foot after downloading Oracles db +bit compilation issues updated fix +mrepositoryorgapachemavensurefireCDocuments and Settingscsanchez +Patch for trunkI think we could do with updating the Hadoop API usage +Here is the refresh patch. +Patch to that resolves correct class loader use avoids loss of qualifiers and qualifier vlaues +For discussion purpose lets take the below two casesCase Field injected before resource initialization Actual instance of the field is injectedCase Field injected after resource initialization Proxy instance of the field is injectedIn case i would be able to type cast the injected field to its actual typeIn case i would get a if attempted to type castCan this be avoided some how +Strange hudson does return +Agree it should be possible to create a schema with only one currency field in it +From the logs it appears that the client is opening a connection for each send and eventually the broker exhausts all available heap space and cannot allocate more resources which is what would be expected in this scenario +You can edit the autoexport template and include any css you like or link to a different set of css files +to doing this as a single is breaking because of an issue with script +I committed this but leaving open for potential discussion +a patchdiff to solve this for providing the patch +to the test case lets ask for it in the original JBCACHE issue as it is a test for the Cache testsuite not the AS testsuite +Should be fixed now +I +also added the bulk of this issue +Thanks also to Michael Busch! +This implements the shortcut approach indicated in the initial descriptionIt also addresses an issue where the Capacity in the method might be set to a negative value or a value less than the current Capacity +if the file name matches a regex for a known password you use that if not you fallback to a default password +Verified present in CP +Committed in r + I just spotted that there are a bunch of emoticon images +Please explain how is that a limitation from technical perspective +Creating an interim solution that will create empty files +Have you tried the official release or how about trunk? Its possible this issue is fixed and if it is Id like to get the new release out ASAP so people can start using it +Its EAP CPThank you Anne +Thx Patrick i see your point +screenshots of both cases to replicate with Fedora KDE +Tell me what you think +The quickstarts should use the community parent POM instead of the product parent POM +Screen cap from my Designer +You could write a plugin that does something similar but now were getting into edge cases and while I like working with Camera Apps on Android its not maintainable since theres too many Android UI assets that need to be dragged along with the distribution +Tests pass all tests in the testsuite +Fixed in SVN +Didnt know what I was colliding with but figured it was irrelevant +A simpler patch to apply after which fixes this to apply after looks good lets apply it and move onIts actually a superset of and thats OK too. +The sequencing is persisting to the HSQL repository because subsequent queries retrieve the content. +I probably should have updated this ticket with as much but I have tried to reproduce this a number of times with an eye toward debugging and fixing jsvc with no luck +Bulk close for +Any reason why this hasnt made it into trunk yet +Allen mentioned these errors are still present +Please provide it as soon as possible +Is there any progress on this issue +So needs to be updated to add this as a separate configuration a pom and source jar need to be generated the and targets need to be updated etcI also dont yet see the functionaly problem that Avros jar referring to Hadoop classes creates for Hadoop +committed this +We can keep track of current offset and the last record and compare itThis should work assuming that the date should not be the same and if it is and the record is the same then for us it will be a duplicate anyway +Gettinglogging from Castor prior to the switch to was painful at best +Currently it doesnt do any checking on whether the consumer is just idle so even if theres never been a dispatch it would still get wacked as slow right now +From the JDBC Blob API An SQL BLOB is a type that stores a Binary Large Object as a column value in a row of a database tableSorry I answered too quickly +Closing all resolved issues due to a successful release +By default booleans are treated as booleanThis is useful for integration with pig environments +So the test will get an exception but not the one it expects +so replying to an email thread doesnt take a conversation off the JIRA +svn commit m add some synchronizations around the requestor to avoid it being used concurrently by multiple threadsSending clusterenginesrcmainjavaorgapacheservicemixjbiclusterengineSending clusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorSending clusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorTransmitting file dataCommitted revision +Im happy to chat over the changes on IRC if thats easier +Diff for for +Geez Im comment happy now +Here is a simple patch to solve this problem +Attaching the patch to improve certain site pages +Fair a test case to see if I can I can deploy a SNAPSHOT and you can pull that down without having to do a m being overzealous as I want to cut a apologies. +So it seams to be quite stable and as I only changed package name updating code should be as easy as updating jar Having said that patch of includes a own implementation of +Thanks for the review Ram +He has informed me that he has signed the CLA and faxed it to the Apache Software FoundationI converted the component to the JSF spec +Added a cache size check in revision +Suhas got committed yesterday +defer all issues to +We may need to get INFRA to do it if nobody else can either +Scoping the is very intrusive instead that will be done in in. +All the links in the reports should work including the links to the previous and current Javadocs +New version posted +Using screws this up somehow in the current frame +The is a deployment script mapping file which currently only has DDL mapped for HSQLDB +Although we are now going through deadlock detection I am still not entirely sure I am testing the scenario Mike was worried about +BTW the target is now but lets see if I can get this earlier. +recompiling is not an option and we should not require people to wait for the next appserver release either +Hey Benjamin Niklas has introduced me to the process by now +Ashutosh sorry about the delayCan you refresh? I am getting some conflicts while applying the patchI will take a look +This is done against +Good suggestion +unassigning doing this work is premature +Seems like a few more classes need patching clean building causes compilation errors with and a few others +Addressed your commentsAll unit tests passedAs a next step we need to get this tested in a production grade cluster and hope Lars can help us with thatthanks +Also I figured we may use something like atomic updates to describe them in the docspresentations and so having atomic in the syntax would avoid having to add every time thats the ones with a IF +Theres also an error at the end where the thinks the file is still being written even though its already been closedBoth of these involve the so I suspect were missing some synchronization +Targeted where the core PID contains a pipe symbol does not the problem with the service PID containing a pipe symbol in Rev +But especially in a like EMC it might have side effects that we cant yet foresee +In this case I would create a issue asking us to improve the documentation. +Hi SanneIts true long chains of can affect us +Yes sure please see new attached filesPlease let me know if you have more requirements + +there are two doCall methods and the second one contains a line number in and +hehe np I dont mind +Fixed on trunk revision . +Patch also committed to. +Seems weird though that this can happen +If we index the files and keep that updated in svn when we change the locations but keep the links updated in the system should be able to create a rewrite that redirects to the new location +Deployment will behave unchanged +Or rely on the fact that any type is an Object +Patch applied Thanks Florian +Currently it looks like thisbytes on wire Thrift Mutation on coordinator node on CN bytes on wire RM on data nodeproposingbytes on wire Thrift Mutation on coordinator node bytes on wire TM on data nodeskipping copyingmorphing the TM into this going to further entrench thrift? Please dont further entrench thrift +So blockreport still blocks heart beat +Define harder? Id say if the templates arent included then the batchshell scripts shouldnt be included either +I started to think about some workaround like I am pretty sure that if the new wave is created via RobotData API it can be viewed without problems +Ack I missed the correlation id in +Resolved +how much time does it take to create splits for million blocks on machines with blocks per split for example +Evaluating the RHS first is the right approach in general in Groovy and other combined operations are a bit different here of course +I remember overhearing a discussion that the internally reuses branch revisions as real revisions after a merge somehow transmogrifying them along the way +attached JSON file used to test that a blocked producer is interrupted when maximum duration is reached +System are working beyond this stage +Not a Scripting language like Ruby but itprovides much the same as Ruby in that it alleviates theburdon of writing clunky code like embeded servlets HTML andto some degree perl +If user cancels will not be run. +Hmm but says fixed +Andrew Good point or we could just upgrade the next master release to Java Are you still running on Java +Revision . +Obviously youd need to test that pretty carefully on your end to see if I managed to fix the earlier problems you were seeing +Is it true that native compression was in use when you saw the problem +Attachment has been added with description Input file causing the problem +Ive used config all +If we run into scenarios want to address we can handle it in a separate JIRA +the hierarchical information defined in javamarkup is very important and i would not want to do lose that +New patch addressing Daryns two points +In other uses the counter used to determine retry time is +Reference +In the patch we check if there is a project in classpath with the same name as jar and dont show problem marker in this case +We cant work with snapshots but i took a look at the updated source code and it seems like the problem is the workaround I was using to overcome this bug is the current solution. +I started to write something asynchronous but it was hard to test and could easily leak threads +Hi MartinIvo can I get access to one of the machines to try it out there? Thanks +In the first case the extracted text contains only junk characters and the latter some glyphs and ligatures are extracted wrongly +Gennadiy would you mind taking care of this issue? This should be pretty straightforward +All these will remain system properties which are configurable in like the rest +We need to clean up our reference to the stuff first +I am still not convinced we do the right thing here but I have not been able to investigate with the current code base +PR for master merged in master branch +Maybe later today or monday +Committed to trunk +yes can generate SQL statements but normally people do not know how to run and having sql scripts to trunk and +I meant exposing it via the Hive QL directlyI dont think there is a way to do that currently +fixed in r +You can use the application from to try this out +My Zip patch contained some external libraries that Im using +This would indeed result in the situation you depictwith the specification of being ignoredA more conservative patch would be to construct a member done in thepropertyList object which is an array of booleans and keeps track of whichattributes have been processed so that reprocessing of an attribute isprevented +Yes have been running this in our deployment +java in the source tree the following files contain the improper line breakssrctestorgapachehadoophbaseclientsrcjavaorgapachehadoophbasesrcjavaorgapachehadoophbaseipcsrcjavaorgapachehadoophbaseclientsrcjavaorgapachehadoophbaseclientsrcjavaorgapachehadoophbaseclient was also wrong before but was fixed by Jim so no problems in trunkPatch to found the same set +Moreover rename as well the to since this module is the only implementation for any jdk which supports pack +These are unfortunate but a release is not a regression so I dont see it as a hard requirement that we dont have another before we add interoperability as a featureIf interoperability is a requirement for future releases then perhaps we should push into +Oops forgot version on is a duplicate of +Indeed I do confirm thats the issue is fixed on Mac +There doesnt seem to be a out of the box way to set service properties +Perhaps then we can get a better idea of how to create a unit test that illustrates it +Since we just load the group information from hdfs +Thanks +However it means not just adding a but also fiddling with the build which is rather the one case I can tell my teammates Hey I got this cool set if tests you can run your implementation against just add this dependency with scope test +Patch implementing optimized chooseTargets method in namenode +The fix should probably also be merged to the and branches +Improved and final patch +Using generatedsource tag with attribute idl we can define the ant task to compile the idls very similiar to how javacc is used +I dont think the overhead is worth the benefit so I wont apply this patch but its here in case people start falling into that concurrency trap. +Heres a new patch that fixes this issues on all the hosts Ive tried it on +Ah gotcha +Once goes GA I think it will be wise to lag a little further behind but for now I think its fine to Im going to go ahead and upgrade right away +I dont think so +Unfortunately I am not having luck reproducing in my environment but I think i understand at least the Test problems so will code and submit and see if nighltys work better +initial stab it seems to work but needs to have the assembly fixed so we dont copy in these dependencies +I dont see any benefit to restricting a users choices +This is Google were talking about here +As an easy fix to make tests succeeding for now you could simply use a trivial character +I have the same issue with Mojarra and Trinidad +This quick fix just adds Serializable interface to the bean without adding serial field +Does this behaviour reproduceble only in visual page editor or for others too? For example wtp editor simple text editor +Can you give us a debug log +If specified this value will be passed to the sort select tag +Great work Renato looking forward to the second report +Can be closed now +bug at +Shouldnt this get closed? see the issue has been resolved I dont think there has been a new release yet +to trunk +patch looks good to commit +Edit release note for publication. +Thanks Jeff! Resolving as fixed. +I have a patch ready that does just what I have proposed above if deemed appropriate +A client sends HQL to a server that parses the HQL to run client operations against hbase +This is from me kreynolds looks great +Please let me know what I can do to help +Had a nice quite time on the h train ride without network access +One additional note as Wolfgang H pointed out this same problem is relevant to the API method that takes Source as argument the source streamreader obviously has to be as well +I am concerned about extending the language for supporting features that can be handled internally +The failures on iOS led me to and the code. +But please note that this is no bug just conflicting jars +please provide a until a quickstart is provided +And yeah the solution i have yet to commit is just to revert back to always using the default xsteram map encoder +my solution isnt quite right +Sorry about that +In that case I will have to oppose this change since it is attempting to introduce a new semantic specific container type +test failure looks unrelated shutdown failure in +Those who were seeing failures earlier can try with this patch and report back +Thanks Scott +faster variant from filter code is using family of +Is there any workaround call the Webservices for which the soapAction is not definedIm using Axis and my generated stob contains the soapAction while they are not defined in WSDL hence getting Service Not Found ExceptionAny help is really appreciatedThanks in Advance and really sorry to treat JIRA as a discussion forum. +Compare +This is a problem for Apache POI +The reason that for the first query the numFound is lower the the second query is that the is higher +Thanks for more plentiful description +Can you try with? I just ran the file tests on an iPhone S running iOS Mountain Lion Xcode and all passed +Reviewing the doc with respect to replication this feature will need to add itself to the chain of in Master +as I pointed out it throws a on line of the JMB +Committed revision +Catherine Ive made a second commit at r in order to make the test pass on the RI for me too +Thank you joesIve been following this BZ but cant get it resolved for qpid doing the chmod +Note that the massive memory use of this benchmark could also be to blame it may simply be more of a benchmark than a splitjoin benchmark. +I did an lsof i n P grep i javaand saw where it grabbed +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +classpath is first meant to be manipulated by Eclipse and its plugins not peopleYour current +Working on a patch +Wow this is interesting +Looks great thanks Prasanth! I just made a few minor formatting tweaks and committed +With up to date not being public and the OracleSun merger going to make for some interesting JDK futures marking this as a LATER + +The solution in my case After creating the MKS integration point and session call si connect autoconnecting doesnt seem to work. +The documentation is somewhat improved in the snapshot but it hasnt been uploaded yet and Im not sure if it would describe your you upload a sample project that has the problem +Forgot to mark this resolved. +bq +Security would need to be enforcedbut if we did that we could simply set the jar ala hadoop mapreduce andhave an iterator hot loaded in a different way for a different scansession +There were quite a few issues in our email subscription ll probably find that youve received quite a few verification emails in the last day responding to any one of them should get youBen +Reinserted the gt prefix on trunk as of revision will revisit this issue if and when the maven issues will be fixed +Ive removed the classes along with associated java types implicit conversions request keys and testsHave a look +I wrote all the code so Im pretty familiar with zookeeper internals +has been marked as a duplicate of this bug +I agree that might be confusing +What is the size of copied file +I jumped the gun earlier saying I had fixed the config +supplied test isnt needed +All fix are in place. +I wouldnt commit this until Yonik reviews and approves +objections to applying this one? from me +Committed to trunk +The static inner classes for are very tightly related toBut various static classes in Utils are indeed for quite disparate purposes +The reason why we ship JSTL there is simply because most of our samples built against JSP and are supposed to run on JSP containers +It might need an opening of a bug in eclipse +This would obviate the need for the useof a content dir for such things which would only be capable of doing a filematch by globbing on the extension +I was waiting to see if thered be any further consensus about the pack module on the mailing list but Im not sure that anything came outIn any case Im unassigning myself as I dont have any time at the moment to move ahead so anyone else is free to take this up +Thanks for all the reviews feedbacks! Ive integrated the patch into and trunk branch. +Thanks Suresh +It takes almost as much developer effort for a point release as a new minor version so the effort usually goes towards the new minor version +I missed a line +Bulk close for release +Perhaps this can serve as the ticket for just this approach. +This is the issue I raised in. +Fixed on and update was not clean for me after the update I have several test failures when I run the tests from eclipse +Looks good to meDBy the way youll use syncServer instead of Zookeeper +So I dont think there will be any significant changes in performance or overheadI think this will work as long as no partial blocks in the middle +If they had been deployed to the Tomcat class loader instead of the webapp classloader then I suspect this error would not have occurredRemoving the shutdown hook is not an appropriate solution +Job only runs when discovery site changes +Reviewed by Ken +ScottYes the submitted patch is from the newest revision and includes your changes +Since bug exists and patch has been confirmed to help in two situations on two different platforms inconsistently Im reopening +Closing as it is addressed in +Bulk move issues to and as part of in Solr and +See attached movie for the full in rev . +The problem is that we have recently changed our internal representation of Generic types to be more aligned with those of +ill retry on hudson again to see if the passes the tests +Obviously I meant Xerces +images load fine throughout the doc including in section of Struts Tools Tutorial where the flowchart appears +should be made as essentially current conf is hierarchy based +In my view its unnecessary to have two properties to separately enabledisable the bootstrap JVM and Jikes RVM unit tests +trunk and several revisions back +Closed +Hey Pete currently the as plugin just adds the to the build section of the project pom +the one that failed with the exception can you verify the JAR for the Javadoc Plugin in your local repository is completeintact +I have removed my comment +enhances the existing container to support service calls over ruby classes provides a baseline implementation of introspection based component type buildingThanks applied +Ill have another one for trunk patch applies to trunk +Hi KristianYou probably know more about this test case than I do +Will report back with test results +These classes are deprecated and will eventually be removed +Dennis seems like you dont like this idea at all +Is this something worth proceeding +install the installer and ignore errors caused by wrong gcj or JDK earlier than cd RHDSbuildsnightly optbinjava jar console cd homenboldtjbdevstudio +So I was suggesting you might want to have a providersetup which then would not require implementations to adapt all of your ideas +thanks. +we should integrate into the tooling though as something you can do on the old files +How to edit this issue? Anybody please change Type to Typo +Get GateineCloud assembly from drawmeacloud +It sounds like Ed and others are starting to use JDK in production based on the needs of their organization so it would be nice to have something documented about support for JDK and tweaking as youve mentioned +A common workflow for me would to be to do a forwardreverse index lookup +Gregory you are right +Which headers should be public? Only +They describe my expectations and the current behaviour +I use Tomcat and build never seen this error +looks good +Ralf feel like posting a short request to the dev mailing list +I think ij should remain the same and strip off the comments for clients even when this is ported to so as not to cause trouble with old clients +Managed to reproduce added it in r +Thanks a lot for this enhancement +Changes look good to me +sorry for the formatting mess I shouldve looked up the formatting help instead of trying to remember the you could do that and youre certainly welcome to do so however were hoping to find a solution that doesnt involve modifying existing gems to make it easier for anyone to package any arbitrary gem that includes java code +Second attempt +Can someone please review and commit it +coretargetcDocuments and Settingset +I believe we can handle both quickly and they are somehow related thats why +This issue is fixed by updating plexus components to their latest versions +Well I still feel that this solution is the problem but it will certainly solve the business need +The problem here is that the task submitted to the is not wrapped in an exception handling runnable as it is when run directly by the +One thing that I am realizing in this process is that a patch should be more singular in nature as in if I submit a patch for the homepage it should just contain the homepage and any related files not a bunch of extra pages etc +Please assign attribution to Alexander and Mike patch +It would be good if you can confirm this. +Duplicates. +At least on my machineHowever I have not fixed the tabspace preferences issue mentioned in I only replaced the incorrect instantiation of with an equivalent instantiation +The cactus project built with md chacksum of the archive +patch +adds the uncaught handler and fixes works around the jvm bug +Bob can you please verify and close it if it is okThanks +Triaged for +Can you elaborate more on why Manuels patch is incomplete? Is there another cleaner way to do thatI am asking because I am not sure if I should try it out out and test on different operating systems + updates the MSVC project files so the spec SCA code and Calculator samples build run again +Thanks a lot for the followup could you please submit a patch? It would be easier for me checking the modifications inTIA +Hi Jason youve come to the right place +I integrated your test case into so that it gets executed for all OM implementations +Also unfortunately Alexis second patch is not ASF licensed +The Attached patch resolves the problem +It was not observed on +fixed on the NBCOMPATIBLE branch works on top of netbeans milestone and possible to backport because the java apis in are different from as resolved +I am experiencing the same thing +Since this is a feature request I tend to postpone this to +Im still open to moving the type cast insertion functionality to the validator as we discussed +In the case of the DS being able to disable it if only for debugging purposes may prove useful +The current implementation should at least be revisited because it is wrong to set sunAvailable to false when in fact I can fix the exception as a user by editing a text file that can be passed in on the JVM CLI +Forget it +yes for inclusion of the patch in if its somehow possible +There is a big improvement in performance of +Solves my purpose well +attaching test at revision +Closed Issue +Both of these seem absolutely horrible + IMHO is that a bug that xml comments are considered by the tobago layout system maybe but it is definitive not a bug that html comments are considered by the tobago layout system if the output language is html +When debug versions of files are missing now using the raw version of the file. +Whoops! Will remove Yes the default should be changed +Thanks Jing! +This also suggests the need of an that does toUnicode when indexing you dont want httpxn URLs in your index +This is the project proposal I have submitted for this project. +Duplicate of in r +Proposed improvement is not real and comes from different config +Patch to include the non overridable class list and hidden class list on the deployment page +I am really surprised that it is not targeted for release as it is a must to have for all JEE developers +It isnt necessary since its test still run against hadoop and it doesnt support the override +Apache redirect HTTP HTTPS rule step based on jive documentation applied as well. +Fixed the clearing of the caches by limiting the records only to particular VDB that is being removed or +Gavinif it is really like you said then it would be maybe best to turn off the attachment function in the create issue screen and only allow that to be done in the update issue screen +committed +Looks Perfect! +I had to turn the value of canUnsign to true in it and it to build them +The patch has been appliedPlease verify +Developper guide updated. +drop +I havent explicitly tested but it is very likely that the same defect exists in +But we can still consider this as solutions to solve the issue later? I just dont want to make suggestions with lots of brainwork and sleepless nights involved if its not considered and just be backed out with None of this is reasonable +Jobs are blue +bq +Comments should be processed much faster now +the hotbundles bases on Felix File Install +Should show its head again if a prob +This overrides the index lookup in +No worries I just pasted the patch into the patch wizard and it worked +This was fixed in ve updated the environment section of this defect to reflect using xbean version +Do you do import node with version history +Fixed in upstream master will be merged to +When do you think this patch be reviewed and applied? The changes are relatively small +Any suggestion +I thought you fixed this +If theyre universal then they shouldnt depend on Writable Protocol Buffers or Avro but rather should be explicitly and fully specified in the protocol +I forgot to say that I have the same exception in a similar situation and Id like not to look for another workaround like in the example above also because I do not know how many similar situations I will find in the future +It looks like you imported uimaFIT in a flat layout +Ah that was due to Knuts patch already containing masters for the broken version +Next time try to open a bug issue in the JIRA so we are aware of the issue and can fix it sooner! +In the meantime I have found what the problem isNamely and include +I have reviewed the patch submitted by BrianThe patch looks good to me +You should also link to this issue and mention it would need a +Can you look into +Added Apache License comment to new files +Fixing Affects and Fix in Version +The tests dont really hit the compression +Looked at the patch +Thanks for catching my mistake +is out! +However in order to avoid backwards compatibility issues the load balance policy for home proxies has been hard coded to round robin +Thankfully +Pretty sure this particular issue was caused by not having enough permissions to create the directories +Patch to add back doc on IT tests +I have removed them and checked in a new + +FYI this is in trunk now together with a basic solution for arrays in +Doh! Sorry I did mean cwiki groups +I will make the title more generalIll remove the unneeded permission from the list and wait to hear from Dag before filing another patch +The test failures have been identified as defects by other tickets looks good to me +Useradd etc +We could I assume +Gabriel Yes +nd Version of the test app demonstrating problems with business conversation +The problem here is that uses GET request to send the new content to the serverInitiating Ajax GET request onwicketinterfacemessageeditorwicketignoretruesavetruemessageeditorIfyo +Verified in ER. +On the point of File Channel I do feel that it is important to have that support to ensure that we do not put excessive strain on memory for the host process and that we do not lose events in the case of host process failureAnother point to consider is whether the source would be any different from a regular source when running in embedded mode +Id be fine if you closed this as wontfix it was depended on by some stuff a while ago but if you have better solutions Im all for them +An improved messaging bus with support for this is planned for Seam as cleanup +But dont know which is the best +Wouldnt say so in general but theres no real world user usage data +The refactoring is in the PutDelete code not in the replication code so I dont think it needs another jira actually +This is a bug in your test no? Shouldnt implement Translator? or extend Translator +whoops pulled the trigger before finished writing +git hash adcaaedacfeecfcreopen if we need to need mention that cquup also logs querystring. +I am having a bit of trouble creating a test case that will fail on any system +But from this curve it looks like there is a sweet spot around maybe N +Files are now removed from plugin packs +Bulk close for +Patch replaces enabling more test cases in +Thanks! +fixed see comments +Trunk patch for this issue andA race condition really +Apologies about the git diff as well +I was looking at and realized this also +Thanks +the old Migration infrastructure feels unnecessarily heavyweight now +Please consider asking your questions on builds youll get far betterfaster help there +I have created separate tools issues for the specific problems that occur +But I agree that we could use the for the signature as wellFor your suggestion to drop to use in Revision +Pi In the past we only had to integrate with the major versions of hadoop +Are you sure you are using the same attribute name everywhereSorry my example was wrong the attribute is in both cases +But will take a bit of time thoughDo you think we can add this later without blocking the whole issues? Otherwise you can take over the issue and extend the patch with the command pattern +Instead of storing the entire block report storing only those blocks that have newer gs +There are many times that a file by its existence implies meaning +I dont know why Heinz and Sydney did that in their initial implementation Ill have to ask them. +I wish to give body and header objects as parameters into wsTemplate method +Here is a patch that the observed error is due to the independent issue. +Exclude local copy of and from MAPREDUCE project + of version can connect to of version of version can connect to of version of version cant connect to of version of version cant connect to of version problem in Unbuntu server support with was broken +The patch looks good for meIt is committed. +I talked to Eric Hodel about this and he says it was fixed in +Unfortunately primitives cannot be used that way so some value has to be picked to represent +On codebase of which branch should I do the patch +If still relevant reopen or create new issues. +Attaching final patch committed to trunk +Leaving the JIRA open as there is a POM fixup to add too +Normally that is via the tx reaper timing out the subordinate tx +I dont have a better solution at this point for handling that issue +Committed to and trunk +the unset variable hack work +This is a great news JeremyI will have a look at it with all the current POS issues on Friday +Thanks. +Great +The patch is fairly simple hence it has not tests +Push the issue forward +To use JAXB on the client you have to Stand up the server with Aegis +Thank you for the alarmThe code in and is a very bad example because it also supports loading from and such stuff solely for testing puposes +Patch was invalid +Should I refactor anything that this is testable? Nevertheless testing if the GC doesnt throw away any Map entries sounds funnyBut I executed the tests nonetheless +Committed to trunk +the complte tree is built and sending the messages +Thanks Andrei! +User can pick based on a config parameter with a note indicating that wont work with multiple HTTP agents +This attachment provides a VS solution that is a more complete example than in the issue fails to compile with the following errorError BCE Internal compiler error Object reference not set to aninstance of an object +It would be best for each translator to implement limit handling since the translator may already be setting batch sizes or other settings based upon the assumption of returning a full batch +Additionally the lucene core has no external dependencies and I dont think we should add anyWhile this is definitely admirable and worth trying to preserve it should be weighed against the duplication it incurs especially within the Apache project itself +You mean POST db some multipart content +KelvenWhen it full resolution expected to be in +Sorry for the late correction +A problem with compiling of plugins java was subsequently fixed +Thanks Mark +There is indeed no column contentsourcetypeid in the content source underlying issue is that there are tables in that sid +attach my patch +Resolving +trunk tested this with download and it does not generate a classe from the imported segment definition it generates a new class for each time the definition is used in the model +Updating patch to be compatible with +Whats the status here? Isnt there an rsync to minotaur and thus the main Apache website that can be used rather than redirecting over to cwiki +Can someone else take a look at this? I had only agreed originally to look at the work around and have not had much of a chance to try to actually fix the problem +Test base on with revision can not reproduce it now +I need to find these jar files which are not in the diowloaded library +What was the version of maven plugin used +But theres still loadfalse dead configuration item in +I have run the same tests and came to the same issue +Theres a virtual Jenkins on the v patch though and the changes in v were minor +This issue found in KVM also +FYIIn testsuite I updated the Seam version in to the release instead of the snapshot +Committed revision +Also note the new Synchronized do it using +for a fix. +Linked distance calculator issue +What does version null mean +Sry guysGimme two hours Ill report the findings and the exact reasons why they are failing +Linking them together. +Patch applied Phil can you review this commit +Ignore my previous comment +I tried to not violate the style conventions +Frameworks like Wicket also use filters to trap they are interested the bug description a possible workaround would be to make sure the URL you are using has a corresponding resource in the WAR file even though it will never actually be accessed +Fixed here +We dont use any binding files +See CTSjaxrpceewsirpcliteralmarshalltest workaround exists and the changes are tied to the next XB release +No explicit batching +bulk close for +Yes is ok to use +It only occurred for directory names with q in them +Yes see attached patch +Installation instructions of the Cocoon updated with the all necessary information to install Cocoon on the Tomcat and up +I just removed this class altogether +This patch seems to contain some modifications of file which are part of resolved +which is what you meant +Actually I should probably do one more thing here wait to start sync until we are sure the leader sees as recovering +If there is a discernable patch or issue logged that someone could follow and do something about it will likely be fixed. +Given that JBM seems to be working in the other testing this is likely to be true +There is a real issue here +Trying to rebase yet again +Once you write yourself an extension to a command you can set the code there +Checking the code it should be showing the image +The reduce tasks times out and gets killed when this happens +Still able to reproduce the problemIn a setup with S storage CS it trying to spin up the system vms even though the zone is not specified with staging storge but in enabled state +Correct +What do you think? and you subclass it to add your own stuff and then you would use this concrete implementation in your operations delegation. +It would call the supers method +This is not possible and thus the user must be informed of that fact +It adds a table page to the UI +I am interested in a new for +Matthew a patch would move this along quite a bit quicker +Resolving as wont fix +Jobs used in the workflow were Gram jobs and there was a For each node used +Its inflictionSent from iPhone. +Looks like in the framework directory was incorrect hence why the generated in the example app was incorrect +I dont care deeply about thrift or supercolumns but if we are going to add this this should be done right when supercolumn is set slice range should refer to the subcolumns its trivial to implement with compositesThe test in the first commit is sort of useless +See JIRA re cannot set a description for a function +It can either be connection or developerConnection +And yes I do use runConfigure script to generate Makefile at the first time I used d option for building debug versionRong +today I learned +Fix the there Unused imports Container containerId hasnt been started should instead be Container containerId is either yet or already stopped There is still one more race condition in A in progress will race with a +Its just a matter of adding deprecated no unit tests +Looked at the new diff +un shouldnt just reply null +I hope this bug gets the right attention +Looks goodJenkins didnt seem to have picked this up +Added in with r and in trunk with r +This issue is addressed in. +So this test is failing because it decides of a sudden to connect to a different zk cluster not the one started by this test +Fixed in snapshot +This allows for isolation of handling of this case if it is to be implemented +submiting patch after incorporating Owens just committed this +It generates a Forrest content document from a given FOAF file +It looks like our logic to determine whether a lob can be fetched inline may need to be reviewed for certain queries +maybe I should create a without parent that is missing junit but will compile a +Here is the missing unit test for this bug +Patch for Ant for Ant build +So we need to keep track of the module names too in case of +Ive updated luc so you should now have access when logged inHowever public access still needs to be resolved +tjake were you planning to implement a file format for use here +But I dont remember there is such a chapter in EAP +stack Dec PM Patch applies to and to TRUNKI hope itll be fixed at the release +what is still open in this issueMaybe you can create subtasks for what is still open so I may help with implementation +I have retested it and I figured out that I have posted the wrong parameters in the bug descriptionIt will not fail if I specify the specVersion on the ant task +I changed the clause in revision Could your verify that it is fixed by the trunk version +Added a to common package +Thanks KihwalUnit test failure was unrelated and fixed by. +It doesnt work in the latest snapshot +Thanks TedIll grant the license in the futureAny objection to also getting this into +However we can still see tasks from the remaining jobs assigned to those globally blacklisted nodes +bulk close of all resolved issues. + on the patchmaybe with There are store file to be copied instead of snapshot file since youre copying the info manifest +Committed to and trunk +HiI am a final year undergraduate of University of Moratuwa Sri Lanka +Its nothing serious my patch only commented out the two lines instead of removing them entirelyThanks all issues fixed previously after a brief review of each. +I believe this was done for alpha but anything that was missed can be made into a more specific issue +looks can you commit this? should this go into as well? +Can you please try latest nightly +router XSD Schema definition will be extending channel +Nice work Jun +The OOM errors shown in the description are not seen in the Hudson test runs since the revert to VFS +Id one more question though +Im inclined to close this as WONTFIX +Is it advisable to close this one and it under the Karaf projectAs for the patch I can create one for as well although I will be unable to test it +So just how realistic is it for customers to tweak the contents of these property files? They are documented and part of the public API but shoud they be +I will correct the last javadoc issue then will try it using the h MR job on which it failsClint I guess further refactoring would make this relatively easy to do +As a result running +Added to the list of affected versions and set initial estimate. +For the time being i just saved the conversaiton Id in session and its working fine + +The test test covers this issue +But I cannot understand what you are trying to do inside Axis with itIn your explanation you have mentioed that Filters can be used to hide elements to intermediaries by not presenting them to +I added appropriate synchronization to +Moving back it should be like and a half +you guys know what version this bug was introduced in? This is affecting us but were hesitant to move to so soon after release +v addresses the above comments +svn commit m Terminated exchanges are sometimes sent when handling an exception leading to errorsSending srcmainjavaorgapacheservicemixcommonTransmitting file dataCommitted revision svn commit m Terminated exchanges are sometimes sent when an handling an exception leading to errorsSending srcmainjavaorgapacheservicemixhttpprocessorsTransmitting file dataCommitted revision . +I have succeeded to reproduce the issueThe default wizard properties were set only when the Portlet Wizard page was shownThey are now set when the project on the first page is changed +Users can use it to specify the date format they want to use +Here is the quick patch for illustration +This issue is a duplicate of +This patch first looks at value first then looks at and +Utility and tests for float and double. +s on the dev mailing list for the removal of lock ids from the coprocessor in addition to Andrews here +I appreciate your patience +Please confirm that this resolves the issue + +Targeting. +Will keep it simple +Have fun +Uploading running regressions +When a project moves to JIRA all its bugs resolutions history and comments are preserved right? Perhaps we should just excise projects that have left from Bugzilla altogether +Sorry about the false alarm + Removed dependencies from main source core +The parser can not create the structure when there are broken label statements like above. +Thank you +Please reopen with more info if you still see logj +Pervov proposed to always return NULL from classloader in case of some exception is raised +Linked this in with the overall netCDF issue +Logged In YES useridThis has been fixed in the branch. + +Run a slow consumer using our sample programs ant consumer Dtopictrue Dmax +Assigning to James as a temporary fix would be to scan for the presence of annotations in +Now I get what youre saying +We may try to figure out all those properties we can get from the +profile successfully merged application deployed and running +Im not talking about that kind of accuracy + +Thanks again +Closing due to lack of feedback +Fix error found by review +Thanks Peter +Will commit now +Verified thanks Tim. +Added to! +I can get compression rates about to MBsec and decompression between and MBsec with itThats awesome! Good performance without native code would be really nice to have +fixed in fixed in do apologise for that +Committed +ah of course +diff that adds the API file a class and some example Java runtime library taken from the facebook SVN repo at revision Thrift API + with java source file is a reserved word in pig +plugins +Android I dont know youll have to track +My preference is to do checkpointing in standby +I tested it locally with the version of and all works as expected +It would be nice to have programatic way to ask about the version so that client can detect a mismatch +As stated in the previous comments a method is in place to construct different types of controls for a commanGroup which should be enough to handle the presented issue. +Again the failure in the core unit tests does not seem to be related to my patch +However to support efforts towards AR I will temporarily be extending my mentor role to cover also simple patches like this. +New exception type for reporting exceptions +After reflexion I see solutions create new converters or apply your patch and set the mapping param to false in my case +It was released with the fix in +Was this an upgrade +Change fixes the position of the classifier in the file name but Im still getting the timestamp s only the transitive snapshot dependencies that are coming up with timestamp versions +The exact behavior with references will have to be clarified in the spec. +Got this done today including a test case to prove it out +to next version as this one is already released fixed in +attached problem svg has been added with description SVG file that demonstrates the problem +so the published messages could be evicted from the cache according to the TTL is not be consumed for a long period +The best we can really do is to add some sort of cache based on bean name to the so that it does not the pointcut each time +How about just leveraging the resources settings from the build section of the POM like the jar plugin does +Its closed as a duplicate of this ll reopen that issue and try to track the problem there. +changed the type A type B syntax into type A B to prevent compatibility issues with existing code modified initialisation of known types list +This is using Build which uses the tagged I believe +Please apply the DineshPatch applied with commit revision Imesh. + +Ill post my first patch soon and then quickly commit it before any conflicts are built upOnce I know exactly what the current persistent issues are Ill likely open a new JIRA issue to track that but fix it here +a fix for this new patch that adds a junit test that parses through the examplesI am not sure if there is a better way to access the conf dir from a test +CHARSMAX so the problem first occurs once the length exceeds CHARSMAX I have reviewed the suggested changes and they fix the problem +We will have a fix for this committed soon +We should leave both the old Type optional field as well as the new one and provide backwards compatibility for clients specifying Type +The test now fails for mefixed test in r +Then create an new Bean I dont see the problem here +think Collection Aliases is the way to deal with this +Logged In YES useridActivation of a context was not restoring the ejb remote andlocal proxies which was resulting in stale references toprevious local proxies with incorrect keys. +Works OK in +is a static property so a reboot of the database is needed for it to take effect +Issue may already be resolved may be invalid or may never be fixed +So that means that they are clickable? Im confused +Yes it probably is better to do an eager conversion +In the next version of the patch a mechanism to subscribe on updates on stored data should be implemented +I voted for it +Yes +Hopefully this should a new patch for with the same change to get rid of the findbugs warning +in fact I realized maven mirrors will make it impossible to account all idea on how others are doing? +This patch takes the previous test and refactors it into an abstract class named +I posted review up on RB Jimmy +So I might migrate the method level stuff to another jira if I dont get it done before is released +Or I just blind and cant find the way how to do it +I guess hans has a new setupif that is the case this should be closed or used to document the new setup +A and B are startingdeploying services concurrently +The patch doesnt have any unit test in it for it is the build patch +I would like to help out with some of these issues +Ok for snapshots yes i agree. +it looks like bnd is getting the manifest correct in that it only shows as using +I applied the patch submitted by Jan but made a few minor changes to it +For deb package lets not put dependency on openjdk because hadoop security depends on itself +The regression tests passed on JDK +The problem is that the master does not know where any regions are located except for root and meta regions +Ill have a further look later when beeing at home againThanks for the contribution +Similarly I have a few linked to the Snapshot more than one occasion I needed to change the remote repo for a given set of artifacts +The problem seems to stem from a timing assumption in that test which no longer holds true +So it might be better to do this optimization at that point +Closing RC issues. +But I think the issue raised in is valid +Sorry Gustavo for the delay but your patch is in good hands nowI think we should remove the test that depends on the cluster if we have the right coverage this will make our tests suite better +Is there any chance you could attach a profiler to see what threads are actually generating the load +bulk close of resolved issues. +Already submitted a patch for +My development environment is not on the SVN checkout +done in trunk +docsupdated +It is obviously xplanner since I mentioned it several times in the body +Ill take a look at this targeting +could you attach the binding file? is the wsdlLocation in binding file an url +i am really interested in walking through your approachJoe +We need new release to resolve this issue +We must decode strings from source consistently using the encoding we decide on for +I modified that method to look forrequired validators and set the associated components required property to true +bulk close of all resolved issues. +Alexey please verify +fix in rev targetPath which contains only +Final patch after incorporating review comments removed wrong updates in finally block Fixed conflict with Store operations in javadoc for new jpa executor classesall tests I run individual tests and they all pass +Looks like this is related to start log and stack trace with DEBUG +Otherwise if things express an injection dependency on a JNDI binding it will not start in the right order and the whole thing falls apart +That is the plugin should at least print a message like unresolved dependency +Thanks for making that changeNow all we need is Dan to review +Patch appliedDIGY +for removal of JCC references +Fixed in svn revision and merged into branch +Have you tried using the standard Kerberos login module in the second domain +Committed to HA branch +Added the patch fileIve confirmed that an equal number of tests run before and after the patch +But the knowledge about the default spec is now in two places +On successfull flush clear the list On flush failure create new entry log and all entries +We just should support this because this was actually working before the patch +Does being to clearly look like a servicebinder issueI was unclear on whether the chain of service listeners should be firing before the registerService call completes +Problem is that the responsibility for extracting the payload from the SI Message is in the so if I provide my own message converter the extractPayload flag is ignored +bundled with jPDL process designerYou mean bundled with JBDS right? Tools does not bundle thirdparty big thing I have against STP is that it seems to not have any GA release nor is scheduled to go out with Ganymede +The lock issue should be fixed +Just committed this to trunk and +The production algorithm will not be able to predict backwards so our test environment should not allow that either +These changes should be tested against a static cluster against a cluster that uses HOD and against a cluster that has been allocated with HOD +I just switch to for the build +Thought I had reproduced but had not yet +Thoughts +Sorry for the we go again +Will work on docs as well +I dont think that I fully understand what you mean here +We hit that with tmp files +million strings in seconds? this has no affectkeep in mind i fixed the analysis in and doubled the speed of the default english indexing in solr so if you want to improve indexing speed i think you will be more successful looking at other parts of the code +I began to review the patch and I did not see any problems so far but I will now wait your update to review it patch +correct M. +I just committed this +What do others think +Sorry forgot to mention in my earlier comment +Till date I have only run sort on nodes with this patch and that worked perfectly +References are only updated on save to circumvent a known bug in Graphiti +Here is improved patch +I see the compilation problem after updating it looks like the patch for has been committed +Ive looked at the code and this is now in place in the release +Revision +There is a better way to do this testing a solution with a new entity will create a new issue +Nothing against though +We also feel like maven is a bit too much to deal. +Removed +type mvn installto reproduce the targeting this for expected end of April +Deployed same war file in both of them with an misspelled message key +Therefore closing this issue. +this should have been resolved with +purged value class from spec earlier tonight but there was already an item for that in the release notes. +Mark said the GA version should go out today +There are commits for this from months ago so im going to assume its done resolving. +please refer toThanks + for the patch +patch for tests added as already has appropriate coverage +Attach a fix that deals with keys in explicitly +Thank you Jakob +I have just contributed a dev environment setup documentation +I was praying that by some magic you would come back to give me advice on this issueI understand the situation +There is a bunch to do on rpc still but the new fashion for serializing cells over rpc is committed as is spec +Hoping that they find out what is wrong and wishing you a quick recovery +Dimitri what do you want to do with this issue +I believe this means adding Henning to the JIRA group +Right wed somehow give the user either option +Upgraded to. +However if any of the attempts to lookup the primary key fails we end up with an exception the other strategies are not tried and the feature type fails to be configuredThis patch rectifies the code so that it logs the failures but allows the other lookup strategies to be attempted and eventually to end up with no cough +Functionally identical patch including changes from the last commit +Whats the reason to cache performance improvementYes +AFAICT this was fixed in r so there is nothing to for us to do here +Norman could you please merge this change into the tag +NOTE The patch is slightly outdated and is based on the current SVN trunk of SDO CWith Axisc the type axisenvt has been renamedn to axutilenvt +I fixed that in revision +Spring Actionscript needs something similar for Autowiring to be usable +Thanks mike for fixing those issues and for the cleanups! I must have missed that sync on swapSearcher +Maybe you can use a profiler to make a more qualified assertion about where the bottlenecks are that you are seeingSorry for the inconvenience is targeted for later this year +OK leave it +This is now merged to master +Further other functions in this class follow similar style +The snapshot shows all jars in libThat all looks OK +if you check the you notice that the keys listed there are listed in a erroneous manner +Add new to exceptions +Marking closed. +Copying a file or a bunch of files by Setup API should be declared explicitely in on the other hand without this declarations all files should be copied in the classic way +This explanation also needs to include something like the you place the +Agreed skip publishing relocation pom then +Opened to look at later +Do you still want me to share it +Getting back the timeout exception I knew it was probably a database indexing issue but had no idea as I did not set up the data +bq +Oops yes good catch! Ill fix this +Here is the list of dependencies and status of the blocker patch which I am involved with blocker issues partially fixed by fixed but depends on +The API promises that applications which attempt to violate permissions will be thrown an +Vertices and objects overhead are acceptable to me +Attaching the patch file that JD already has tested +Can someone please close this issue +This issue is fixed and released as part of release +Im asking because is the exact same bug and we fixed it last week +Committed thanks! +In particular advanced features +IWO only the Please complete a words are note there is a space after a +Is it possible that this ant task modifies the project and builder decides that build must be repeated +This has been implemented and tested. +Im right thinking additional YesNo select should be added to the web installer +Remove and some files that were not seriously modified +This issue is fixed on the perftesting branch +I could do here and in a separate jira +Reporters of these issues please verify the fixes and help close these on upgraded setup +If we end up needing to schedule tasks we can invent a new notification mechanism +Ah shoot +Please file it separately. +Yes +The approach I used in the past is as you suggest to always return trueeither with or patch to return true +In any case I guess I can tweak the build to start xfvb if not running already +Mind having a look at the Ordered types implemented in and the associated encoding strategies in? You should find some useful implementations there. +Marking this issue as resolved +mar file is used anywhere so I didnt want to remove it without being sure +This patch fix the problem by making sure that countdownConstrain with a parameter of N means that isSatisfied returns true when it is called the Nth time +Right +should also be mentioned +our ops didnt want to upgrade to ensemble just yet and so we run w a client against ensembleJesse Sounds fine requiring in +In cases where task failures are ignorable we dont want to block reduces waiting to launch new attempts for those failed tasks +would you rather it with something else +Please confirm and close. +works fine with in the jar however throws te above the thread library is bing compiled +fix in rev +Thanks Tony. +Peter what do you mean by generating VMEntrypoints? You mean switch to using annotations to create it? Im not planning on doing that in the short term +All subtasks are done +I agree with the scope statement and am to closing this as per comments above this feature is considered out of scope of +I backtracked +They may be forked or hand written but we should keep them update and rich +The dots in and as well as the ring in is considered part of the letter not diacritics +I will also look into it +Good find Rick +Anyways my impression is that it cannot be changed without a lot of work no change at this point +I just dont think you should have to do that with the build plugin +Once made were stuck with this decision or at least with a really really high barrier to change itOn the other hand we have the downside of extra complexity for the driver authors +Should be fixed now the option does not disable the tcp transport completely just requires a SASL encryption layer on any connections on it to proceed. +The extractor is basically just using the class to load a properties file +Attaching latest patch from committed +shirts shipped by +was committed as r The problem is fixed on Windows now but I can reproduce it on LinuxContinuing investigationIn this fix I reworked the way monexit is called on exception propagation +Investigate for if they should be in SOA Van will set the Fixed Version +Erick requested assignment via irc channel +Closing the issue as it is not observed in the Latest Build. +And Im on LinuxWindows +Fixed in Weld +An example configuration already exists in confexamples but I think its out of dateI think that is a great place to put it +Would you mind opening up a separate feature request for that ideaThanks for all your feedback +Must be uncommented +Thanks a lot for removing out dated attachmentskind regards the sources and build related stuff has been committed to CVSTesting of the Guththila implementation has to be done and would be addressed by the that I created. +This issue should be closed +Try to change the repository to some other directory +Heres a simple benchmark between the two methods +Linking + +Let me think about it +updated with formattng think I need eclipse code MarkusIs there anything else I need to do on this patch? Do you want me to roll a patch against trunk as per AndrzejRegards it needs to be added to trunk too +Fixed in both and resolved issues. +Committed patch to branch at revision +Committed revision Thanks Peter +The balancer should work even if only the normal NN RPC address is configured +Using the newer routines an email like example is valid but exampleexample isnt so we now follow about the javascript jcvcheckEmail methodI tried with latest SVN code +Its impossible to say without a deep analysis of your use ll keep this bug open as I suspect that there could be a connection leak in the undeferred close mode but you should have first posted a message to the mailing list to discuss this in more details +I think this might have been an incomplete commit that I discovered yesterday and pushed in +Check back and press Apply New project Seam web project +when I add a peer first walk the graph of current masterpeer relationships and refuse to add if I detect a cycle Im not part of? Would require an API to ask that question but thats probably a good thing anyway +If working on the backport for this issue + +Test to reproduce this bug +merged this change to +This patch enables with keroberos setup to run on secure Hadoop with defined host key tabs for master and Region server. +Hey Scott I havent discovered anything +On bookie start up it will delete the bookie registration path on and the new zk client will again creates the path and continue startup +If not please write which version you are using +I am not sure about the Einstein accolade but thank you anyway and thanks for patching it so hope the fix is correct and addresses the problem I saw +Thank you for the great job Juergen +Hi Alejandro I doubt we can find HDFS URI from the Oozie configuration since the file inside oozie conf directory doesnt contain the property +Closing all issues. +Linux superclass was you may have to modify some of the tests that you already wrote after I added explicit ids to ui components +I would be glad if someone could test my changes and provide feedback +Key should be good to go now. +full to trunk +for storing edit logs in +When we add the Ops schedule I would say that this functionality would have to be served by two different alerts with opposite schedules. +Im focusing on style issues mostly +This resulted in the installer not downloading the zips again +Failure seems to be intermittent +Good job +Michael since Im responsible for the incomplete I can save you the trouble and modify them +Attachment has been added with description Example +Babak can you raise another ticket for it and ill take a look at it ubuntu updated with latest jdk of as well +This bug was a bad oversight on my part +If someone attaches a jar with the diff applied I can test it in my app where Im consistently able to it though that probably doesnt help unit tests +Thanks for clarifying +This may help + +I think that you already have a version of the missing jar on your computer but do you know why these files are missing from the repositories reached by Synapse project? Is it possible to consider putting them back where they should be or could you at least send me the missing jar by email for exampleSincerely files are attachedConsider adding them into central maven repository? +Tests finished successfully on the branch +Also it should be something whose update is fairly deterministic andor controllableFor instance if metrics will be updated once every so often it may not really help in automation unless there is a way to control the update interval +Finish button is disabled in two cases +bq +AndreasI just ran a test within in trunk the test is passedBut there is something wrong with the I will dig it more tomorrow +the JSR doesnt really apply here because you really are not controlling the lifecycle of the servlet with this code +Now check if Workitem is in the CASs typesystem before creating an iterator. +Great workaround +still an issue for functor I will review. +They should be releasing in the next few hours at which time Ill import the code into the Incubator svn +osgi headers from module are still not correct because of invalid char in example for following headers version be replaced version version number should be now +I thought others might want to use the wrapper where UTF is still used +Our update was successful with. +I opened to improve so that it visits text annotationsI also worked out a simple patch to PDFXHTML to directly extract the annotations ourselves until is fixed +If there is some other person I should assign this to or a particular workflow for the patch just let me know +After some more thought I really think this is not a critical issue anymore since it is really a corner case and there is a workaround for ejb +Can you supply a patch based on the changes Ive made +Added jms and tomcat profiles +thanks for reporting this +Also has some import reorganizing by accident +I have reviewed the patch and looks good +It has one item on it a Enable Message Trace checkbox +If people want to work on it we can add them as committers +Observed speedup on the suggested microbenchmark +Also working with local image of fsimage might make it difficult to catch issues if there is divergencekeeping it in shared will ensure that second NN does not have to format +Please attach your server and agent logs +Groovyc +I spoke with Brian about these and the patches will go into unless is for some unless the fix version indicated in this jira is wrong I think this will have to wait for s release notes. +I am sorry but we looked also for a new Batik version +I will work on a improved patch to address this issues today +If no one disagrees with this approach please review and commit the patchThanks Heshan +Perhaps more ofthe test code can be sharedTo mark the issue as youd edit it change affected versionto fix version to +In case someone can beat me to it +This is not a problem for applications like the namenode that writes to one ledger at a time but there might be applications that require clients writing to several ledgers concurrently +Closed upon release of Hadoop. +I will accordingly be producing a plugin with these changes as soon as I have the +I think my attached patch and test case fixes this bug but I guess Im meant to leave it to to change the status to FIXED apologies for any confusion +To make it comparable to the you will have to optimize the s code +Hi Bernhard if my understanding is correct youre saying that this regression has been introduced in version by even if this ticket is not closed? guysI want to believe that this will be fixed in Jenkins so I do not expect actions on our side +thanks for the review Stack! +step update screenshot step the screenshot should be between step and Procedure +This is probably not the worst thing assuming our compaction heuristics are good but something to be wary of +EricExcellent suggestions +Xmx JVM parameter? Usually this is specified either in or via vmargs eclipse startup parameter +Ill have a fix forthcoming for the next release +Attaching and +Just let me know if you changed it at all +Happens only in docspace which has older version of SBS. +Please try it out and let me know if you see any problems +The COPYRIGHT file doesnt seem to exist anymore +components +Build you were using for testing probably came from box AFAICT doesnt have the needed build tools so it only gave thepartial build +Initial patch and junit tests +Thanks for the review NicholasIve just committed this +were described these two is done +You can take a look at commit aefbbfaffddddaecbde which removes all the non oss +shows that Bound Variable button is available if using shows that the button is missing when using basic types +I will update the Ant task to compare the Woden results to the expected results +bulk close of resolved issues from releases prior to +I committed this. +If not added Service Entity as a change should be in ER +Should be working now can you give it a try Han Hui +BTW if you work with rsyslog the logical newline escape string is which is compatible with how rsyslog escapes other control characters + uses the API access to the compiler we use the same class in places can you pull that class into a common place? Besides that lgtm +commit to branch Committed revision commit to trunk Committed revision . +One things confuses me however I cant find the method get in any version of Hadoop that Im aware of +We need to refactor everything that interacts with JDBC as it has upgraded to +The entityoptimisticlocking test is + issue +This is done in the latest site in the svn +Sure Fire data files for failed be a timing problem the test code installs and starts additional bundles one after the other without considering their dependencies +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I just tried it again and now I get the correct behavior every time like you do +votes and a patch offered in january! Whos the coconut asleep on the jobPlease patch trunk with this it works great +Done +Archie did you test your resolver on windows? Any feedback on this +If it sits here too long we may have problem in applying the patch +In any project using this parent POM the webdav wagon does not work because that newer version is used instead of the correct dependency of the webdav wagon +WAR +Yes +Why must the messageSource bean be instantiated before the are run? +Committed at subversion revision + for the patch +Test from this bug and listed DRLVM VTS tests pass on in all modes. +initial cut at a patch some bugs in Sep and Pulsing +bq +NOT BUG +Please take a look +As long as my integration tests pass I should be able to have it into within hours +This is not necessary +As far as I can tell after a quick look at your modification you obviously found the involved classmethod +We are definitely not going to fix it for or + thanks +Ah so lower down in the create child form theres a usernamepassword field for the JMX username and password to use plus a checkbox you can use to store these credentials in your browsers local storage +The Jersey artifacts with version should be added to the in the dependency management sectionIn the they should be without version tagBy doing this all dependencies versions are trackedmanaged from one location for the whole project +I close this issue +I still need to do some more debugging to verify what is going on +And also cordova is failing to copy the contents of the www folder of the plugin into the www folder of the project +resovled commits in the square brackets as indicated by Peter +At a minimum wed need to test it on Jetty Tomcat and Glassfish +It would be interesting to know it this issue persists with and if so a stack trace of the would be most useful +Much appreciated. +on +Verified on ER. +Its generally assumed that you have full control of the environment your application is running in and that access to the disc and network traffic is secured +Feel free to commit this any done only on trunk +Can you please enable publishing of the CMS site +Can we just drop the comparator details from totally? Looks like we dont need it in the FFT +Im not sure why it does not work in your case can you please provide a bit more info +So those bits will not be documented on the plugin siteIf you write up the missing pieces I can add them to the plugin site +Still its doing really well with the typical SQL Hibernate generates +Therefore the issue will be closed +Thanks Benjamin +I wonder if we should give a warning to the user about the factwe are changing the name of the compiled class file to be consistentwith Java conventions +I think this means that it is unlikely that the IBM JDK is causing the have also modified the file in the +Note only applicable to ppc baseline compiler because it allocates space for expression stack within the frame +Two tiny comments The license header is missing in the new test class +update holding ctrl doesnt seem to do anything anymorehold alt and a mouse button down will zoom but will also activate the current tools that sounds like a job for you Jesse. +Taking that as a yes for now +Replace them by ! will not work in the aggregation +Please try it out and let me know if you still encounter any issues +We also need to do the same for Exchange +Jacques maybe you can delete the old prototype dojo folders frameworkimageswebappimagescalendar frameworkimageswebappimagesdijit frameworkimageswebappimagesdojo frameworkimageswebappimagesdojox frameworkimageswebappimagesprototypejs SaschaYour patch is committed at folders above have been deleted at revision +committed +the name of the attribute is not debug it is showdebugoutput +Can someone help me +All classes have been made Serializable + +technically a dup of closed that one +Im going to give this one another day or two and then commit if nothing comes up +I will resolve this defect based on your feedback +This will allow others to fix with out having to modify the struts source +Of course it may make sense to move it to but we probably dont want to remove it +currently asking myself if its worth to split the diff in +The attached patch applies cleanly on +Regards Aravind +As was just discussed on another useful application of this would be to allow easier of an existing node without with a different IP address by starting a node without data and having it enter the cluster in mode followed by initiating a repair +Well spotted +So I did not duplicate those tests in testingARQOptimization +This is resolved as of commit fffbdfaccaff. +Every feedback is welcome +Hi Victor +Assigning to keep track of the activities +Closing resolvedwont fix issues. +The blog listing wont show back that far +Olga I marked it as critical since we mention that Pig can eat any type of data and the example script shows that we need data with fixed schemas and to perform a simple join +In this screen shots this is single node cluster GB total memory +The patch will not apply directly to groovy core since uses a patched version +Tried to add an empty line with some spaces and eclipse automatically removes the spaces when I save the file +I afaik dont have the perms to move at least I didnt spot a corresponding field when editing +on having this merged into thanks +servicesservicesinquiryservicespublishservicesreplicationservicessecurityservicessubscriptionservicesservicesvaluesetcachingservicesvaluesetvalidationThe webservices are untested by us however as we do not make use of them +The Wiki is still used and linked to off the main web page but it is no longer the project home page +After an initial look at the current EJB timer code my thought was that commonj timers could be used as a building block and layered underneath the EJB timer service +I will commit shortly to branch if nobody objectscommitted +Im assigning it to Steve for the release +You tried on windows? I was going to take it home and try later +Rather than patch locally can you let us know when theres an update to logcxx and we can use the fix in there +Yes I can confirm that implemented change solved the issue +released closing. +We can create new jiras to track them + You can run this or any other bugzilla testletstylesheet bycd testtestsbugzilla Note Paul or Dave or an axes expert should still look at this +leaving Archiva out as an alternative to migrate from in such environments leaving Archiva out as an alternative to migrate TO in such environments +Also the commit date is way before the date we tagged +lets make it one less hassle +public static final int SYSTEMPROPERTIESMODEFALLBACK Check system properties first before trying the specified properties +Mark will not work here we will not scan myfaces bval openjpa but integrate them for perf avoid issues when using mojarra or hibernate +If you want to also include a RHEL specific command line form that can be done +Thats how I read it too +patch looks good +Done at the Event levelStill has to do it on the Annotation binding level +Hi Gregor Thanks for the patchCan you grant the apache license for your patchI cant apply it unless you grant the license +It would also avoid any cyclic dependencies +Ok I see thanks +seems to be fixed so I close it +Lohit can you take a look at the javac issues +I just committed this +Integration test that fails under Maven but succeeds under Maven files from a successful run of the IT under Maven file from a failed run of the IT under Maven didnt mean the plugin configuration which looks fine but the mere log from Maven +The original patch had the retryyum and retryaptget from which hasnt been to trunk yet +In Rampart in clients incoming path we cant check whether the module is engaged when it is engaged programmatically +Fixes the issue with hadoop where the native libs are spread out +Committed at revision +some thought I would recommend that the aggregator tag would be removed so that all the subprojects would be recursed +performs a fallback check for autowire candidates now which implements to accept raw type matches if the target class has unresolvable type variables +I think the failure as opposed to the warnings are fixed by so in should be fixed in r +Thanks Micha +Actually whenever there is an error in the merge the is left in a state where other threads using the would encounter +This page was displayed on trying to download velocity the screenshots to reproduce this is an ASF infrastructure issue +During the M release we used deployable datasources and I do not believe we had this issue +One is from the listener side and the other one is for the sender side +Ive corrected the in the test +Have tried a few times and can not connect Ill check again in a few days in the meantime please check if all is well +Michal can you please give us that would illustrate this issue +Should we rename anchor point as starting point? +We could try applying it to Seam but I remember reading that Seam doesnt work with +I want to make sure my changes make sense +My trace get set is for true get for is for true set is for true get for is for true set is for true get for is for true set set is for true get for is for true set is for true get for is for true set is for false is for false set get get I see that get for wasnt executed before loading data +The patch doesnt appear to include any new or modified testsThe previous justification also applies to the new patch +The will give you back correctly deserialized opCodes for several different version flavorsgetVersion will let you know what version of the edit log you are readingIf your version is too old youll have to upgrade it with +Seriously this is for or just as soon as possible +Having those come from a distinct bundle might make things easier but according to Chetan its not that easy to implement that separation +You should bring this up with the Jetty group +Rumen consumes job tracker log directories and produces the job traces that mumakil and cosume +BTW thats also how I verified so please correct me if theres something more to check +for checkin its a good solution imo +Belated +However Ive only tried this with EAP ER +However if you move the assertion in the way youre proposing the assertion will trivially hold +IMO copypaste from Hadoop can be a simple solution +looks good +Ill try to fix this for tonights good news this time +The current documentation is almost completely user generated +Just ran the test times locally no failures +Patch updated now +The filed is not in order to avoid an accidental erasure +I did not write the code in the plugin which uses this old library and I do not really understand what it is doingIn the meantime the workaround is straightforward avoid implicit transitive dependencies by fixing up your POM until s output looks the way you want +this is adding logj mbeans into the jmx registry so that logj itself can be configuredmonitored via jmxThere should be something similar for slfj no? this allows us to verify particular messages are indeed being logged +We should not be changingpatching these features provided via central in jboss toolsWhat is the error? Im not seeing any change to welcome screen for me +I am not really happy with how I refer to the JVM documentation suggestions are welcome +Closing. +I think it should retry even for since only indicates the exception originated from the server side but it may still a transient problem +What do you think of the small change I did? Does not break the IPC but the question is if anyone relied on the stack trace being in the message? I doubt it though +Reopening to change the +Nearly All Mirror Owners On List Have Been Contacted. +This issue has not been issue does not covers complexType java class are putting our main development effort into JAXWS +Attachment has been added with description Buggy behaviour on SVG of FOP has been added with description Result using the most recent I also kindly ask for explanation why the st file is cropped on PDFI generated it using Inkscape and both Inkscape and Batik viewer are showingthis file complete +Not easy running copies of droids in different version to get clean patches out of theOh the nd solution needs to be adapted to the droids package naming and also reformatted to match the code guidelines +It seems a lot has changed since last July +Fixed with commit LahiruClosing the resolved issueLahiru +patch in zip format containing das modular distribution +I also made the output directories unique since there were occasional clashes between tests despite the test directory being deleted each timeTests pass for me on Mac and LinuxMattArun can you see if the patch works for you please +I think though that most of our users are programmers of a sort +That will be ported to caliper +However when the enters the stopped state it is no longer available to process invocations which leads to an now will back out the changes while thinking of a way to make sharing safe +Phil any new about this issue +Thats what I was thinking as well +Is this ready for commit or has it actually been committed already +The patch with your corrections committed to CVS HEAD and branchOleg + +Fixes tests and behavior + +So all these tests should pass except for V clients +Mck Right now I cant access to our compilation server +It also appears that the use of the inRefresh flag is insufficient to guard against rapid fire repeat events as seem to be raised in SpringAdded some NPE guards and inRefresh guards to guard against this. +Made the rules for asking version information in less strict this method can be used if there is no repository available yet for a targetSee commit +We can add helper code to that allows a client serialized and deserialized types to and from bytes but I do not think this belongs into the API +The patch depends on previous implementation of +All TCK tests pass so this is not an issue anymore. +remote cache guess Dan forgot to close the JIRA. +Updated patch this recognizes now URL parameters such as fields startend keys batch and crawl id +Committing to and trunkThanks Ravi +OK +Previous comment has a typo meant not + Username loweryr +Same problem here not only with ajpoll request but after any AJAX request the spellchecker content is lost. +Must be fixed in CP +This is an example PDF file +Apparently theres a similar problem with a property called there an estimated time to fix for setting property to String? Thx +Rebased +license header patch in r thanks for the patch +There is possible a couple of things to look at that header interface you mentioned +Eclipse allows to define all required properties that you needOpen you debug configuration and put pos program argument into Program argument field in Arguments tabIssue should be verify that you are using jdk into Ecipse +Im just not sure exactly how to do that since I dont have a current problem that needs to be solved +According to Eugenes evaluation this is a launcher problem +This is just an umbrella issue +I believe I fixed this issue +Hi Abel do you mind checking if this issue still persists on your end +Files to add to the tools +I committed to trunk after trying the test that failed +Seeing it makes me want to work on making Configuration settable in a live server sort of like the log level servlet so that we can set one way for one minute and then come along later and change it +With this checkin I am removing that code +I see this as a request handler +Moving issue w no movement out of +if you file a jira and post a patch ill review it +Could you also test whether the link works and upload? +Were thinking along the same lines +There seem to be no test cases +Here is a patch against the latest svn to fix the angle brackets +Can you do something about the where timeout +If we fix this can be demoted +so what was the scenario hereanyway id like either jeff or corey to retest and sign off on this instead of me just closing it +Setting a quota is an administrative command so if an administrator notices a user directory thats using too much space they might want to set a quota on it which would then force the user to later clean up that dir as they see fit +Maybe it is different from your environments +Weve had lot of deadlock problems caused by sync callbacks to listeners who then executed sync remote method could still keep the sync interface but always call it asynchronously just the opposite of what were doing s run this past Scott though +For some reason the commands IDREMOVESUBSCRIPTIONINFO keeps getting generated as instead of as it should be when I make it everything works just fine +Patches for the javadoc and manual have both been committed +Make that +So it partially fix the problem that German has with the observer +Sounds like a good plan + No enum constant In this attribute incorrectly parsed as catch Brent +Read the top of page Edit Actually read from the bottom of page +How would you do that in this approach +how can i find out witch one it tookgreat plugin btw +mamta did you try a sane server on the codeline? Could you post a complete stack from as the posted stack is incompleteFrom the stack it looks like this is a hard upgrade issue +Thanks Mingjie +Arthur I think Lawrence opened this at your request +You guys should be watching the they are the ones more likley to get comments such as which is awaiting your attention +It looks like the similar with which was using wrong configuration elements in +was MB in the end. +Please revise +Preferably this patch would just work regardless but Im having a hard time to come up with a method that wouldnt require putting guards on each handler function which is obviously less than ideal +Thanks. +Committed +Also since the mailing lists dont exist yet I cant point the commits to the list until that happens +Moving to to ensure this is resolved prior to the the attached file has a very simple web project that generates the error reported in the bug +Resolving. +All the regression tests ran cleanly with the patch +the same problem occurs with multipage online version when youd like to open some particular links in +For new lifecycle all we need is to replace the packaging type which is sustainablesecond the official way to deliver a feature is a zip bundle so I would like not go way from this mode +Should this bubble up to be a column family attribute some day +That is what I was pushing for before YARN was Apache YARN +Added test in revision working on i detected that we prevent hidden propertiesto be exposed by the but we forgot to do the same forhidden nodesadding such a restriction will cause this test case to failIMO this test should rather operate on the node state level +Can you share any more insight? Were considering a CR release this Friday and Id really like to have this fixed in that release if possible +Camel normally also skips the setters but in this case it get more attention as it has a Camel concept parameter +Fix confirmed. +Please remove it +This should have been fixed in are you using the latest build? per reporter confirms he can not reproduce using a closer build +Patch for and +implement a unit test to test and +I am no longer seeing the ASM error and the test appears to be working +Please close this issue +I just made it error thru and try this new logic before failing out +Fixed in r +Juergen +Ive attached modification with my changesIve posted my solution to JIRA not to mailing list because I cant attach file to mailing file with my the quote i vaguely remember saying thatI am also looking into implementingWould this fix be necessary if we implementedAlso please send diffs and not the whole file if possible thanks +It provides a class method setFilter that defines a flltering criteriaThe patch provides three Filters and +Closed upon release of Hadoop. +r +The error message should be a bit better now +Bob did this solve your problemIf so Ill close as now uses Logging +It does not fix the fact that custom tuples whose serialization does not match cant be used across the boundary +To reiterate the point its not domain objects retaining references to the Converter its the held in Flash during a interaction +Tested with success on jetty and Andrea +When done youd have to set the buffer in the message again +So I have made this patch and attach it +Afaik there are only predefined node types with prefixes nt and mixIs this some kind of left over or why is it necessaryOther than that the patch looks fine +Wouldnt it be easier to make avail the Desktop classs browse method rather than all these browsers +No we are not using that plugin +Not sure how to tell Hadoop QA to pickup a certain file to run +If you think this more basic patch viable Ill go ahead and make it work w your unit testsThe remaining items are the is and the return of a Pair out of AMprocess +I dont know if this is the same problem popping up again but it does seem to be something similarYour best bet is to raise a bug with the Ubuntu team that manages the Eclipse package +start from believe this is needed for solving the usecase of allowing to deploy specific jars to processes I believe or + +Still not using batch JDBC drivers do not allow to load generated keys during batch inserts +Committed the patch to the trunk as revision +Thank I agree with this so thanks for opening the issue +Merged the fix to the branch in revision Everything seems good again but Ill schedule another test run over the night to catch any other regressions. +oh sorry I will take care of that next timeThanks Lars +The reason we get tests with no linked exception is down toWe are getting an rather than the expectedBoth exceptions should be thrown depending on the exact threading state of the client however in both situations the correct linked exception should be set +Hmm +It starts to get complicated to fuse the existing behavior of cat into getcopyUnless there is a compelling need for this change Id still prefer wont fix +Where do you think this yaml file and utility should sit +Both try somehow to give its own phase without disturbing too muhc of the overall infrastructure +Similarly the offers a nice alternative to needing a default role so Im reluctant to add a defaultRole property to the implementations +Attached the patch +Reporters of these issues please verify the fixes and help close these with latest build fromWe are able to successfully deploy a Vm in ipv network. +When unwanted code could be executed? when some object that is instantiated by a Factory is created +problem report as filed hereIs this behavior to be expected or maybe a bug? As I mentioned using snapshots solves the problem +Would be interesting to see the stack trace +Instead it should change the copy utility so that it takes an additional flag indicating if the destination file should be overwritten or not +Im new to Hadoop Hive and I would like to learn and contribute to these projects +Now if the timeout monitor is mins +We could provide a utility to make it easier I guess +Updated patch to fix a javac warning +I deleted the older and replaced it with one that address Henrys commentson the part of the NOTICE +Non qualified path is qualified in all clusters +Heres a patch that adds in support for SGD +This means that applying this patch on source will properly result in merge conflicts +Not verified but works for me +The license shown on the web site is secondary and unimportant compared to the one in the shipped product +Jboss is still affected +This patch also contains a few fixes to make valid content changing the case of to Histogram rather than HISTOGRAM in +Please let me know if I understand your comment wrongly +Added a version of +Note Thanks for fixing this bug in an official release as soon as Many many thanks for your great security framework +This is with logcxx on Fedora Core gcc amdThis deadlock occurs quite frequently so if you need backtraces or any other information I can provide thisIt isnt ideal for me to switch back tothanks +Committed revision +Test case as maven ve verified that this lives inside the jaxb ri by running the following and examining the resulting schema which is deposited in the sg subdirectory +Either put the logo at the top of all blog management pages or drop it under the menubar on the front page +This is much like what the new does +I and Jesse would accommodate them +No problem +JD will no longer be usedbq +This attachment includes a test which fails due to the attempt to insert a null value for the erroneously included class column. +It also included a minor fix on HdfslistStatus that is supposed to get into but did not make it by mistake +Can we close this issue? Dont think the logo is even used on the site anywhere anymore. +Dennis Can you try this with the current after revision +Toni need to check this one out something has broken comments recently somehow not tracking user +Put it back later after release +Calendar is not displayed +I tried with the HOME and with call parametrers both give same result +Closed upon release of. +to moving the implementation details out of Vertex +Yes this doesnt make sense +You can say this is an application of the Principle of Least Surprise the resolved captures the result of the resolution process in a given moment in time for a given configurationI see your point about how a change in the conflict manager can affect reproducibility even with static versions I still can have conflict versions if at the time of publishing I used a conflict manager that allowed them to be published like +John +I think this is a which gets thrown went an attempt is made to insert a node where it doesnt belong +The issue is that multicast do not provide any guarantee +I have run junit All and observed no failures +I didnt make any changes though so am suspicious +Ill now check out a fresh copy of the tag and then push the release out to Central +had an obvious error in the NADCON list that Ive fixed yet the +Open specific issues from now on. +StackOk +Markus regardless of what you want to do or dont want to do throwing away the session is the only supported means for exception handling +Hi Ilya thanks for the patch +has some nice properties that can be used for coordination +Provided fix in patch format +Kicking Jenkins problem is that you changed things in both common and mapreduce +branch bug fix complete files attached +Sorry the first and second files contain minor bug was located in applicationproduct directory +Go to the forum +sounds good +Note that the preferred syntax is. +Old DLM renamed under Pauls better some time ago. +Such generated code will replace such hand made classesProbably it has to be a kind of project in order to simplify the way to use the configuration toolThanks any way for your commentsI hope other members see this idea usefull I am any way to go on thinking about how to implement itDavid +Fixing it myself +The intention for this filter was primary to extend the examples section because I think that we need a lot of more stuff in this section for the beginners like me to get an idea of how the single parts work together If you want you can close the issue at this pointRegarding the runtime tests I opened connections on my server in between sec +I also added a test case that tests making it past a failed parse +LGTM +Lets summarize a bit Input and are managed the same way as it is in Hadoop +Coolan inverse of this plugin would be great or enhancement of this for values based on patters as i think lowering score of domains like this has not been touched in a number of years the patch attachments are really quite outdated +Eventually I think we want to have a REST jobs API but for now the proposed patch is quite elegant and sufficient +It looks cleaner to me using withCounter method name +is called for each block read no? Anyway still not super high volume +fixed in +Input from OASIS test suite +Hi Vladimir +uses lz to compress internode trafficpatch is large since it upgrades the lz lib to the latest version +This IS fixed as indicated but is back in versionSomeone should update the Affects Versions toHow did this manage to regress so far backWas a unit test missed +arent always needed anyway and they can be copied under different names +Can you elaborate on the problem with reference to my attached images please? P open bugs to P pending further review +Will let it settle a bit before backporting to x +In summary I think that both are needed to really resolve this issue +New patch concurrency bugs fixed +What part of your build is requiring this? Maybe you can exclude that artifact andor update it to instead +Revisit open this bug still existent in jBPM? +But Ill let you marshallers should be functionally equivalent +Anil I verified that the failure goes away is the line is uncommented +Net fix parameter is set to true +What is the version of your CVS client? Is DC using the cygwin one or do you have a windows one earlier on your PATH +Both min and max are generic by now work with all types and go from Type to same Type. +One good aspect of using the same key for ajax is that it will retain valid views in map than redundant ajax views +Commited +The change makes unregister failure behave like normal AM crash +However I did notice that the still lingers around in the session +Maybe I need to change something elseI restarted continuum after setting the to debug +Server log file containing snippets from running both the Java client and the C client +Hithanks for your support your mirror will be added to the list soonI dont see the customary icons in directory listings I saw themearlier now they appear to be gone can you fix thatRegardsHenk HenkIcons fixed. +Andrew owns the linked issue +This patch moves some of the work from the DMB stream that Tom mentioned into trunk +because the old syntax was broken we never actually supported a list there which is what it implies +Just one really minor comment I think we shouldchange the appearance of a tip slightly to make it more distinguishable from acode panel +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +not committed +We must wait for a later version of AS. +This has been discussed multiple times I reported this in alreadyWe shouldnt be SSHing into but they should have a API or so where you send one single JSON per VM and the daemon inside the VR handles itAnother workaround for now would be that the KVM Agent generates a file scps it into the VR and then executes a command inside the VR with that file as an argument +Thanks for testing to for now +Change the object to clone the Date instance to assure immutable state +anymore details you could give? +Introduces interface for support +Committed fixes and Unit Test to check that this fix is valid for both Local and Remote views +It should be fixed already +Probably simply ok will correct +Its now under collection +Bernd I did previously and I just made another comment on the original +Leaving parallelism to is not a solution here +It seems some refactor can also apply to java I think its better apply them to trunk directly will save merge work in future +String which cures this issue +Committed revision +did you use the DTP tools to perform the edits? if you did it should just work if not then it is kinda expected +tables +First thing that needs to be fixed ASAP is that AS changed their lib locations in the latest GA AS team to get the details and put them here and implement them +For the standard usage with Activiti Explorer and REST we wont support business archive classloading +It indicates that some code is somewhereIll dig up this warning and see if I can come up with a way to investigate it. +This was fixed for via commit ace. +This is more in line with who uses the connection and for what purpose +For convenience here the patched file +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Committed +Can you please verify it againThe other thing you reported about np option is actually true +Hi Mikemaybe not use a boolean for the mode instead an enum +Nothing +Frederic I doubt the issues are the same +also note In the reference guide section on installing embedded it says that it does not support JDK +Fixed with attached patch + This is a useful addition +Ok I committed a fix which sets by default the properties to unknown values this way the returned Agent instance even if it will have some of its properties determined by Trinidad but not all of them at least all of them will have not null values +This one adds the process of copying an entire directory to the compile folder instead of just a file because the +Change looks good to me +I was unable to apply your patch file from Jan th successfully and the older Nov th patch appears to be superceded by that patchCan you please update your working copy to the latest revision and regenerate the patch file againNB Atlassian tools automatically version attachments so uploading with the same file name as a previous attachment will supercede the previous patch file but still preserve the old version for future reference +Also added more documentation +The driver jar will need to depend on the thrift jar +As I mentioned in previous comment Ive already gone with the old fallback which works fine +INVOKEETHREADLOCAL is not meant to be set from the command line + +If you get the chance to give this a try today please let me know whether it works for you! I said +I see it calling the for each of the Sinks and Sources in the new configuration +Ideally it would at least throw out a poor diff in favor of a new one thats better +libstdc +Whos Ivelin IvanovHas anyone heard of himI know of an Ivelin atmaybe you could ask him +Could you please add it to the sandbox and make a snapshot release +You can open the archive and look at the java files contained within to see the script and appender filesNote that this test was done using and RFT but a newer version of RFT should be fine +In consequence well have to poll the SSH port to make sure SSH is available before we try to connect with RSE to it +I got a timeout failure in +But if server is rebooting then client will fall into loop imediately and user unaware that he need relogin again +I am working on fixing those +This would be nice if it doesnt require a major change to the ast creation +Yes I do mean to reopen this ticket from so long agoThe same mochiweb error is getting printed but so are a bunch of erlang dumps +Please commit if tests pass +The icon must be added to a special sprite picture +For other cases we just ignore it +Thanks Vinod +Yes +This set of services should only house stuff that is not applicable to our core component services +I thought the limitations we are talking about was only for this would only be an additional change for session beans which doesnt work anyway if the EJB container is struct +I have implemented more straitforward fixIt was tested on my Ubuntu server and was working for meWhat is wrong on your machine +HiThe methods were dog slow in the past +If absolutely necessary theres a package access class in harmony we can copy that does the encodingdecodingI dont understand yet why this isnt an equinox problem that equinox should fix +Using that is probably the way to go for consistency +Patch that adds a new +On a profile update we have to change all hundreds profilesThats why we are unhappy to resign ideaRegards +Both failedAnyway I will try the user mailing list instead of here +Fred please update the script to pull newer versions of me and Ill run the mirroring script to pull the bits and publish them onto +Somehow it is not clear to me how one does add an through configuration +Updated patch to only separate the flag from major compaction not changing the behavior +Please review +Everyone feel free to contribute your use cases +I agree with Jakob +See the trick is that maven is producing these junk checksums upon upload +I know will need to update not sure what others +However with the validator hooks provided by struts the class is pretty tightly interwoven +This metadata method needs to return new values +so Hyunsik could you assign this issue to meI am willing to do this job +Its ok. +As far as the loadbalancing I think the default will give a similar result to what kernel thread scheduling gives in most cases with an option to do something smarter if neededTheres nothing throttling accepts if the invoker pool is saturated which can make bad situations worse +Add ordering to the interface to enable control over the ordering of the panels + +Currently the web console has a legacy Felix Shell plugin +This is what the data store property page looks to +Fixed with CS revision . +This is on our list +still not quite perfect and i need to do a bunch more testing but im having some issues with my home machine which may result in a complete rebuild so i wanted to upload this version of the patch incase that happens +In addition to does not write setget method for FIXED fieldbecause case statement is missingThis includes this fix too +Please attach the wsdl +And this functionality you mentioned state the war as its seam parent I can realise this via the wiazrd in EJB Module I can do it manually by editing but not via wizardsI had started this Thread +Can you explain it to me like I were a novice why thats OK in this case? Why wouldnt it be better to check for the specific exception +Which would still not work if the table was empty +The book references it in several places +adds the TMPFETCHMARK as before crawls any rows with TMPFETCHMARK set and sets TMPPARSEMARK so the Parser knows to parse the row as before removes the column TMPFETCHMARK so that any other later fetch between now and calling wont the row +Good one Rajesh +bq +I recommend you switch to this version anyways +Verified +VikramCurrent patch also lacks test coverage +Interesting idea +odg nor +Feel free to close this one as well +and release rev +A test to demonstrate map state divergence during a partition +Are we talking about the same thing here? Sometimes it would be nice to be able to obtain a pdf of any whiteboard notes. +does not provide grails support. +Typo new patch addresses Sureshs comment +However I think going forward we should fix the reserved table names issue completely in so that the configuration isnt even needed +I committed this. +server times restarted +We should better put it in +What if we wrapped it in the same way as the Android issue +This has been fixed on the trunk +Verified at r. +For a particular column a index type can be specified that is implemented by a subclass +My guidelines for files that are not java or xml Write statements on separate lines without tabs or spaces Avoid long lines Separate blocks with several new lines Use white space for easier reading and clarity +Thanks Namit! +Actually works fine and dandy +Fix describe function and describe function extended for and +The fact that the plugin is trying to generate an error marker means that there is a compilation error +Test for that makes it currently fix to pass classpath and options to java using classes and you provide a new without all the formatting changes? its impossible to know what are the real changes here and what is just you go +And as the job is now asynchronously launched the eclipse wont freezeBut if that last solution doesnt work either then your solution is the best yes +is required for a skip list solution in because you had to flush the buffered entries to make spaces for newly added entries +class files there is no gain to apply it to sources +No its just that Logging is just a logging facade API +That being said I appreciate Elis testing and actual experience +The failure no longer appears in Hudson we now use Sun JDK where its fixed +rev +Either dont catch everything or rethrow those throwables that should not be caught +Deepalthis is what A receives from me this is very specific requirement therefore how about reducing the priority +Ive created to track the issue so that the fixes here can be committed +Added a unit test it works properly in versions but fails on with the error described +Patch to make implement to do this with a +And I cant seem to duplicate it when I want to only hope that it doesnt the transition text positioning is always a problem +Youre welcome Peter +Are you planning to upload another one soon or should I just review this one +please assign this task to me +It works with CREATE and ALTER tooIt can be useful to know that the keyspace has a replication factor of +After some more testing it seems that the error in the th request is slightly different to the rest +Id say we need the links +Maybe add a special check for not there? On the other hand people might already rely on that too as s JTA does work as long as you dont mark the outer transaction before suspension +Why was Built By changed back to English? Is there no meaningful translation +If you get your initial capabilities document from httplocalhostgeoserverwfs I believe it will return you the servlets that dont use WFS dispatcher +Brent could you take a look in to this +Need to reopen +The JDK already implies that the contract is that if a appears in the min or max then the min or max isWhy not just go with that? Any idea if that is defined by IEEE +I will try tomorrow on Fedora KDE +Description of patchThis patch includes and classes +Revision all resolved issues for Engine release +using the same variable twice will not gonna work out +It is better to modify one helper than two +Verified the upgrade with checked release notes +Purge vestigial remnants of previous API variants +MarkIm sorry but I fail to see the point of this +So suppressing the entire plugin for pom packaging doesnt make sense. +Still seems like more trouble than its worthI agree +A comma is allowed if the whole string is and updated docs + +Thanks Sharad! +Thanks for the explanation Andrew +Hi YokoMy concrete use is with Ruboto and in that case we only create one instance of but still the load path is not addedWe only need it forI will try a Java example +Or move to JDK as a minimum supported version of Java and use standard logging +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Hi BenjaminThis patch looks pretty massive which container do you test withThank far I have only been able to successfully run the sample for jo jonas jetty geronimo and resin +This code path had not been tested in a while +Verified during the EAP CP CR test cycle. +Does it fail when you run your unit tests +Can this bug be modified to be against the documentation or should the component act as its documentation suggests +Just need to know where to start + +will think about tried to rework it a few times but got myself totally confused because some of the local variables have already been modified before the collection is converted to an array so the instance is all messed upI think it would be simpler and just as efficient to clone the collection if it is being added to itself +No one has objected to using instead of on branches +Im cool with using extensions but after we check for SEQ +Also do you know which one produces correct results +Also I have had a total of requests for this source code since posting the details on the forum +is this dependent on +Committed fix for this bug to trunk on revision you for addressing this so quickly Suresh! Im resolving and closing the issue after confirming that the fixture is now running without problemsIf Im closing this prematurely feel free to itbut since it looks like youve solved the problem I was seeing I think its safe to close this one outThanks again! +Reproduced only when show is used only in inline a as part of another bug. +Integrated some of the tests from the into plain junit tests at a lower level to guard against these problems in the future +Fixed in revision +Ok +No way to do this task +Applied patch to CVS +Committed to and trunk +I will need to achieve this in two steps +it appears the metadata is not being read for config properties by the plugin +Committed followup patch with revision +Wouldnt it be better to consolidate and add this functionality to the rather than creating a whole new Interceptor? For example you could add a property valuesBlocked to the that would block values +Editing title as we are doing a little more than adding some missing methods +There isnt much else that makes sense +I could have saved me a hell of timeYeah Serge this is THE commit +So all issues not directly assigned to an AS release are being closed +Hopefully people wont think the is for processing codeI will commit tomorrow if there are no objections +Hi JacquesI made two new screenshotsIf you add more than two products and then change the size of the window you can see that the ratio of the colums is wrongI must admit that my first attached screenshot was pretty useless because it had only two products sorry for that +Merging two regions by moving the hfiles means breaking snapshots and cloned tablessince the snapshot want expect the hfile in a location like tableregionfamily and the region now is changed +I was able to reproduce the problem and fix it. +Many thanks for these new tests +So Im for close this issue + +Lets try to modify the implementation in it would be really cool if we can handle extended operation the way Ersin has implemented SP by using java classes stored in the server This would help to add new operations on the fly without having to modify the package +We would need to support to get to mock out just the private method but leave the public one as is +No the patch is fine I just checked again +Yes this is fixed +You missed the fix change and the scope of provided for is the updated pom file +Suggestions welcome +Just to clarifyService is secured using jms transportwhat I mean wasService is secured +I have just done almost the same test and had the same resultThanks for your help +I think this patch is complete and we can close it out Do you want the honors David +The new patch is rebased on the latest trunk and all the unit tests are passed +I believe that we support the autoloading nowThere is a lot of trickiness in here +Instead it should just rename the existing test from xtest to test +The patch uses a trick by sending a metadata request with ignoreCache parameter to gadget container update content in cache +It would be quite a large change +Is this on the level of the JCR APIWhy would we want to decode entity references here? Entity references need to be considered when the query string is represented in XML but here +I havent looked into +autocomplete in browse is now in the same line as breadcrum and I have made it current groupId aware +moved to since we ran out of time for to +I pulled this into. +I did not see this error with nstest runs on the branch after backing out as per +Sam could you please elaborate on why we are requiring the latest version now? same issue with Spring and the new it does work +keyname was not being used +Yep +Corrected the issue title and added code looks related to +You are right uDig should make a guess and let the user configure are a couple eclipse bugs on this but I am not sure they help at all +peek is still virtual right? I think we should just keep the original version of peek defined in and leave the current version as the default in +The user shouldnt be logging into the token endpoint at all and it should always be authenticated with client secrets +bug on CR on rc +It should work fine with but I just wanted to note that in case any problems The default write of MB came out to about sec between checks on my system +Regarding afaik in the console theres nothing related to what is covered in this jira +My guess is this is related to SWTs browser widget and its integration with mozilla on linux +Testing was performed manually +Final and Im still getting this error because of an unhealthy mixture between slashes and backslashes! I fixed it locally for doing my work +Configuration issue fixed that causes this problem. +I just committed this +So not forcing the loading as a requirement of generating the would be ideal +No fix is forthcoming +This should fix the site generation and release generation may need some help in using the branch +I expect it to come back to me after finish but it also may go to Julie since she opened +For instance Java makes available through reflection a means to remove accessibility restrictions at runtime +Im also fine without using the scripts since after looking more closely at the binstat + behaves the same +Kihwal you are correct +Central should be updated within hours +Therefor I am closing this issue as wont updated +Yep +I never know when things are threadsafe although it does seem to work +Attaching patch that adds the missing information that specifies additionalattributes as being empty for the particular elementsI also put all of the attribute information chunked together by the elementto which it applies and put the chunks in alphabetic order by element nameIts just easer to read +But with the patches I previously linked it makes possible for enunciate AMF module to support this use case. +A default cacheSeconds value of for all form controllers is a good idea I think +Im trying to commit some stuff over the weekend +Tuomas the other approach what you have mentioned requires a lot of effort +Is this issue ONLY being seen with etherpad as an app +Did our chat help or hurt I wonder +I believe this is fixed on both branches. +Sorry I thought this got fixed a long time ago +I know clusters are not a pirority but still +I am facing the same issue on Windows fixed that +I still see failures on hbase but seems it is not related right +Im not sure what this is referring tois it still relevant for Roller trunk +So all issues not directly assigned to an AS release are being closed +After a while well split out some client and some server classesMatt On the mailing list you mentioned you wanted to pull out some core classes like etc +I think you have marked the import as optional but the code appears to still require it to work correctly +on patch +Thanks Arun I verified the fix +So it is one of them +bulk defer of unresolved bugs to +Examples please and what is latest tarball +This patch addresses both issues +I have been able to reproduce this with C +Also no changes were introduced to because it is not intended to be configurable +applied in master and decided to check the existence of instead of trying to lookup for specific CDI classes in the project adding cdi jars to the classpath is not even mandatory to enable in JBT +Or because they may have a project that does not build with the version of maven we include and may need to specify a higher or lower version +If it is present it must match the current updateseq to succeedchanged seq to updateseq and added tests +Ive committed the two extra send methods in the to trunk as part of my checkin for ve changed CMSMessage to match the JMS Message property methods +If some asks me to review code and I make suggestions Id rather not expect people to walk by and say thats just +Finally when the name of the datanode is resolved and if this new bit is set then the NN sends a request as part of the response to a heartbeat from that datanode +Closing as requested by Vladimir +Ive changed the components to indicate that but correct me if Im wrong +I dont agree thats common in some cases and it is in the CDI spirit which allows access to all Resource of the container +Exactly this means that we need to implement writeback and buffering in instead of using kernel implementation +Postponing to version undefined +Added Servlet and Portlet implementations of the along with helper classes copied from Commons Chain. +I have a question With we will get a test class based on canon files why not translate the SQL test into junit based on assertion? Regards JunjieWith the program evaluates the canon files and makes a junit test using assertions +because I find it a bit more difficult since a test does not have an access to the underlying code to close the socketBecause the is a subclass of object the close method throws only and not +For otherswith the same problemFirst add an alias to etcalias dummy dummyAfter that create an interface configuration file asetcsysconfigDEVICEdummyONBOOTyesBOOTPROTOstaticIPADDR you can start the up with ifup dummy. +It now runs mvn clean install on Java using Maven +What is the work flow like? Do I add nodes first or do I create the shardscollection first? I can pass these api params only when the shardcollection is happening +In the primary RPM is noarch +Resubmit if you need resolution. +Updating patch to work with latest trunk since has been committed +It seems that I have fixed the bug can you please check with nighlty build and see whether it works for you +to branch at revision +Documentation will be support for and created for enhancements Composite native query should add one supporting array type based any requests +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Ok +The test button in server editor works fine but I can not find test button in server wizard +Agreed with Gary after reading AS clustering documentation +There may need to be a filter to ensure that the params do not override some critical info +It seems that eclipse looks for library so perhaps a uDig start up script could try and locate this library mentioned that all that might be required to fix this is to move the information tools view to the newer browser API +Please use instead of as a path separator in the build you make the compiler emit a warning when the keyfile is specified in the command line and also appears as an assembly attributeOther than that the patch looks very good! +No the only way this gets thrown is if the region server cannot find the lease for the scanner meaning it has timed out +table priority patch for trunk +Pete this was a comment when I created the link to the related Weld issue +I have told this already in a previous comment. +Ill try to write a test case but not exactly sure yet how to integrate +Not sure if we need yet another test for this +Please verify in ER issue is documented in the Exploded deployments but it actually affects deployments too +Documented as a known issue in EAP release notes +Applied at r thanks! +Marking as closed. +You should use some EE server which supports example AS EAP nor Tomcat supports EEJSF +I like configuration but I even prefer better should be the preferred layout then +I tried using but looks like if there is an error during the code generation and we dont generate any class for the problem query +Ive committed to with the other releases coming when the sanity builds complete +Modified passes locally +I fixed this in the Jakarta build script some while back yes it does take a bit longer to remove the spaces but it sure makes the SVN diff easier to read indeed it eliminates a lot of them. +So the test passes +More will come after some sleep +bq +Why would Derby implement a method we dont want applications to use +No there isn would return the production console log to INFO level and use the logj config to trim down categories with unecessary too verbose logging +Regression test reveals the problem +Feel free to go ahead and commit this +Modifies the compiler and the interpreter to create new hashes +Now that is reversed +Closing issues resolved in released versions +With this work data can be ordered and joined in same script +the current logic does not correctly interpret the intent +none of the alpha or beta nodes call it only eval +When assertions are enabled RI smoothly ignores an incorrect direction while Harmony fails with assertion. +Referencing a whole directory is far too vague IMO +I just dont think there is anything to fix here +Looking back through the repo history I see that this view has existed more or less in its current incarnation for a very long time back into the days +For example if there is RAID and we trust it or if the backend storage system which the is using is doing replication underneath HDFS +Thanks Benoit for reporting and then providing a fix +Changed defaultParser into a private static final variable this should resolve the issue. +Attached a patch that destroys all nodes if any exception is raised inside +Ive verified the behaviour and updated the wiki page accordingly. +Sorry for the incorrect change +Ill play around with it +Fixed in rev +The Server Editor is now the proper place to change those two other issue re having the folder not exist I was unable to replicate +Russell Im not sure what you mean? When it comes to pretty much everything is in terms of Schemas +Alternative is allowing ints and writing varints then eviction logic might not be been a while i went thro this code +thing only +Note that the OCR is frequently quite high but since the page image is what you viewreadprint then it is generally fine +Thank you +I just committed this +No I attached the whole output of the including stack trace and thereis no +We can put a README in the project with standard disclaimers +maven support no longer relevant +Thats why no replication seems a little risky but I want to see what the perf difference is and whether it is worth the risk +Is this the best way to resolve this? I already using this way but i was thinking to use the tx compoment because is more simple +So I propose that the be insensitive to order and the serialization format be +Still working on this Chris +i used the tests provided in the differ and test +It looks like this setting was not applied +Are there any documents on naming conventions etc +Thanks Russ +Unit test passes. +So you should use templates without JSF if you want to use Tomcat for example then you have to use other templates +Unfortunately the zookeeper integration tests are failing for me on ec with this change +No tests because its a change +Perhaps the annotation should be a package level annotation as well +In many cases Cloudscape should be replaced by Derby but not in all cases +Fixed plus update in bde +That means it has to send a bundle to the app to restore state +Demonstrates wrong conversion between servers and clients time zones when displayer a Date +Thats the pom copied from the castor pom both were converted from maven it should be fine the way it seem to have deleted it from their section so I pulled the jar from the release and attached it +Mass close all issues in resolved state that have not received further comments in the last month +Validated +This patch contains the rest of the changes Catalina proposed +treating the byte stream as a character stream just confuses thingsLooks goodA couple things +Michal Im downloading the right AS version +should we kill the session immediately or wait until the sessionTimeout +Done +Yes its too small but some industrial users want tomake sure that this kind of legal issue is clear +Not sure what that last build was about +Seems like I remember reading about a way to put a tarball of javadocs in Aapache CMS when we were first setting up the web site but I do not remember where +The formatting of the warning looks a little odd on a couple of the files + +A patch fixing the working much better now but not perfectly Im afraid +Content is accessed randomly when displaying cached preview +Which can actually be a good thing because we are preventing clients from hitting the namenode when it is starting up +This situation can be handled the same way +the forrest shell script does call ant +This patch kind of breaks backward compatibility so that fact must bedocumented +Very soon it will not and well be working on completing it + and completeBy are printed if only exist! +Ok lemme work on getting everything uploaded +So far so goodJenna +please reopen after providing a quickstart may be something funky in your markupcss +Hi Rostislav Is this issue found from RPM or zip distribution +They look good +to trunk +Committed at subversion revision +Im firing off tests in a few minutes. +Now we have our own that problem is may be a micro optimization which may not be worth it +Here is an updated patch along with the other small edits we discussed logging when handling zkclient exceptions consumer connector moved session expire listener registration prior to rebalance listener registration +Wrong resolution status before +Removed deprecated toInt methods on +What do you think +Ill close this for now but we can reopen it if Trustin you or Stefan still have issues with it +Unless its normal to attach an additional listener everytime ofbiz checks for new messages then theres definitely a bug +bq +Id be more than happy to write one +Burton you provide a patch +Check out the Java specificationIts just not commonly done which is why it is looking strangeIt would allow us to keep the package naming structure intact not interfere with deployment allow us to keep the name at the cost of looking strangeIt may well be the option +Its rather large +It always went away a few seconds after the watch was turned off +Nice catch +Thanks! I have opened to look at the memory consumption stuff we discussed +Refactor the properties of dependencies to Upgrade httpcomponents +updated patch +Thanks +I can still confirm this issue with the SVN release +The temporary one can be promoted to long running in the normal way +Weve already talked about adding a UUID to each database anyway so that seems probably the best routeThetype approach just seems awkward and really not much of a help if you can still accidentally delete things when youre in a hurry and dont bother specifying atype value +Obviously the test wouldnt run if it was based on the java version ofSorry for the confusion. +load maybe? is going to be changing so convert will need to be able to take the old version +It doesnt strictly require a reboot but it does require thatyou apply a network config +May get punted forever if we dont have a good way to identify an unseekable file like dev entries +The warnings are generated because the methods of the class are used by reflection +That would avoid the need to manually generate vcproj files and maintain two sets of build information files +It uses English which is the default locale for resource keys +Worked times locally +Havent heard back and repeated testing and still cant reproduce +I did not backport the test to because some code related to the test framework did not compile +Some moretests needed to validate parameter committed +I dont see the deletes +At least we still need The portalPage manage page where a new portalPage can be created or the name can be edited Service to move the portalPortlet around in the portalPage Permission checking Lot of nice CSSA further step could be to implement Portlet configuration by means of an associated form New Screen Widget widget element attributes a wrong link in the addPortlet Lookup screen +And retesting the code on the various platforms is a major headache +pls reopen if you have any issues installing it +The way we designed our Java binding it binds Schema types to interfaces backed by implementation classes without an empty constructor +Committed the fix to you James just test it and close +Sometime onreadystatechange will loss. +See +for the patch given newResult renamed to create + This bug has been marked as a duplicate of +Hm I might have missed something in the discussion but it seems Prakashs original suggestion should work +bulk assign all unscheduled issues to next release +I have no real preference for which version we use but as HDFS uses i need to keep that one installed +Assigning to Steve as a proxy for work underway at ANUSetting fix target to as we hope to have this available real soon now +I have modified the script to remove are no duplication now after LN file generation kindly continue progress of langpack string unification +Kathy Did you mean to link to Andrew Eisenberg as the contact to ask +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. + +OK nice catch upCould you fix thatYou have got commit permissions now +Closing +Verified that the issue is fixed for +checked in. +Thanks Mike for looking at this +My checkin of a test broke last night +Merged into the master branch. +will package automatically +Thank you for contributing Liang +Gets rid of long long to avoid issues +Ive committed this +Thanks +I see no language that specifically says a Folder cannot be added to another Folder via that method +Thanks a lot for checking on the version +Agree with the rationale +bulk move of all unresolved issues from to +Im using EPP and it is not bundled with JBDS I havent had the issue with EAP but JBDS use to have issues in deploy when was not uncommented it seems that it does not anymore +gives you dozens of ways of accomplishing the same goal so I managed to rewrite the code that it concern me was that the upgrade to Grails required me to make all those changes +Michi is working on the missing C tests and Im meeting him this weekend to try to finish this together +User add child node and then add mixin to this child node but exception thrown +classpath files +Log from region guards against a region server not getting a message telling it to open the regionSo we will keep it +we will keep alive +You dont need svn to download a snapshot +This is the beginning +actually I would say having the text box there even without autocomplete would be a good thing +ie +This is a known issue with the current build +Changes I had to make to get the Console Realm portlet to work in a Tomcat only environmentPlease note that the change to will probably break the normal case of running the Console on Jetty +backported change from trunk to branchsibmsvn commitSending javaengineorgapachederbyimplstoreaccessbtreeSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawdataTransmitting file dataCommitted revision +web containers. +AlanThis patch almost exclusively exists to make the publicly availablePrashantWant to discuss further how you want to handle this? I agree that we dont want to put ourselves into a difficult position when it comes to changing things in the future but maybe theres a compromise +Fixed ORACLE problem with one exception it is not possible to use LIMIT clause at extended patch is tested against mysql and oracle + +All other patches are included + and are authored within the ontonet project I dont see a problem with licensing them Apache +Should off and len be final? Does it make sense to change them? Could they be private +I will verify it later today and send some comment here +As you said that the use of current is not preciseBesides the above problem if someone call load to get a scipt classthe current class implement is not thread safeand so s implements thrown inbut tois not thread safe I suggest its a good idea that the current implements should bedoes the use this classI havent used itThanks +While reviewing I recognized that part of this is already described in another +Closing issue fixed in released version. +Ah I now understand the problem sorry about that +Dont think anything changed there did a quick diff and dont see anything obvious the Uri handling is much improved so if you had a more complicated Uri we were probably screwing up the parsing before now +if the invoked method has N parameters will be bound to an Object array of size N +We should try to make this mavenclasspath functionality more obvious though since I was hung up on this for a long time +I suggest fixing the instructions on said download page to point to the to save yourself some forum posts in the next weeks or better yet fix for your input +libln creating symbolic link +jBPM Tools Reference Guide and AS guide is checked +If someone could help me out with this I would appreciate it +So we could docreate a dbPUT dbdbfragmentation +Regardless of the merit of your observations a change of behavior in something as fundamental as the would have too many unexpected consequences. +Yes looks exactly like what I want for +The now takes the entire deployment down again when the WS endpoint cannot be started +just in this bug or create request in our wiki feature required by users page. +The format from old RI versions and below has been used as a prototype for Harmonys +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This is actually +However i cannot figureout what format the patch is in +Mark are you working this? Otherwise I can commit the fix +I think it is a better idea to make this an extention to the eclipse plugin +And a patch for the inclusion of the extra parameter in the file with comments that include examples +that is completely what it was intended for +Well see the maven surefile process running but that should be clue enough +Should be resolved now Ill keep an eye out +making that instance should be fine +Closed upon release of Hadoop. +project file change that this listener is looking for is bunched up with other changes in a single change event +Script is resulting in class named which the VM seems to be having no problems with this class gets generated and also runs fine +we saw production clusters that were no balancing because the old balancer did not want to move regions around to balance the number of regions per table I think the above was due to lack of per table load balancing in Though Ive been testing this with per table turned off as thats where this loadbalancer does its best when it has the whole knowledge of everything +bob and I have also been chasing antlr guess we must have added it to our local repo and forgotten that it wasnt on the remote one +orgapachehadoophbaseutil +Hello StefanThe patch looks good thanks for including it in +Many thanks MichajloOleg +Closing. +Hi guysWe can control the package name as Lukas said +Id really need to think of a way to do it without breaking its a minor functionality and its not worth breaking the good design +I did not see it there and there is no point in creating a new noop method thanks for catching that +attached patch to fix issue +Attached Eclipse +This seems interesting +Trivial patch attached +you can just add the jar to to repository. +Attached please find the latest patch +Thanks for your patch RobertI have applied it in revision I have also modified the wiki accordingly +Attached are sources which show how implementation could work +I have documented the menus required by a hosting application on the wiki but I admit that is not solution +Additionally I chose an ordered map because I would want the Hibernate mappings considered firstBut as it is currently neither of those is necessary +Can this code be moved into a proper SSE project or put it back in Spring Security +file operations on JDK +Please attach a zip with example projects so I can get this fixed ASAP if its not too much inconvenience for you +Its ignoredI was expecting this too but clarifies that it has a different purpose +Duplicate of +Hey haric thanks for the bug reportCould you post the code that causes this +put as blocker for M either it gets fixed there or we close it +This issue has been resolved in revision +As indicated I suppose the problem is on the users side +Then there is only one single class with compile time dependency to Xerces in +I can give you the account info and you can ask berlios what happened +Checked in the fix to trunk in revision +Ill check if I can do something with it +Verify this section in Admin Guide and close issue +The same node appears multiple times and as a result when you bring up or down some nodes it gives you incorrect information +See for additional information regarding the necessity for a username and some documentation updates +What the report document renderer does is irrelevant to cloverclover type executions +Great stuff I think we should definitely have something like this not sure whether we should start a new module or add it to commons testing or could just test for this +attach updated patch shortly +We have never actually made the decision to really commit to supporting JCA +Sample app to reproduce problem. +Fernando any progress on this +Agree about wontfix we can always reopen if the need it wont fix for now as there is no real requirement +s definition of equality is the most important as if this is wrong it can prevent beans from being installed +Note that the new version of mxj requires Java but if you are on OS X thats not a problem +Affects Version is basically all the releases before AS this has been there from the first revision of the deployment in AS is radically different and doesnt have this fix looks to be along the line of whats discussed on the forum thread +This patch survived an onslaught of disablingenabling mixed with massive imports +I think I see why from the code +I think QFJ defaults this to localhost +I recommend leaving this at a major level for now and pushing towards full where we do not need this module which is really a temporary hack +Patched added +I had reports from Jukka as well that the fixes for this last time around helped a lot but did not totally alleviate the issue +Otherwise I think theyre great and particularly like the examples at the bottom of each one which we must try and write into the Perl program +I had a look but only supports which is quite different from +Running the class initializer for is forcing the loading of a lot java security classes + +Format has to be kept the same as current code this diff contains a huge amount of line where you only changed spaces so its imposible to where is the problem if the old plugin doesnt compileit must be better and can be a we have to review the code that is added if you reformat the code you get that linesxxx xxxare different which makes it mpossible to review the actual needed changes +Please ignore the patch number +Fixed in trunk +This incorporates your suggestion now reports when it sets the compiler properties + +User defined requirements by the configuration files are always improper +Integrated to trunkThanks for the patch NThanks for the review Stack and Ram +I used a because the is stored in a so I wanted to mirror that +This does work out of the gate but it seems to me once you use JBDS to update BRMS via the Guvnor repositories to keep data up to date things go wonky +Fixed a bug +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Otherwise the child id is a useless attribute with only one child component it is easy to detect the visiblity +Please keep us posted we want to help you find it +The Parameters tab also appears to have no Javascript correctness checking whatsoever +Thanks a lot +I would think the simple cure for this is to fully qualify String as This particular jira seems to be good from just reviewing the code but I am hampered by the new new issues where not in version but I did notice the same issues in +Thanks for taking the time to help get this issue resolved +m running into this too +case of magic project being deployed causes unknown state +youre quite right +So I went with Simple agree this is an artifact of my own environment where testng was placed separately +submitting patch +I tested in the Eclipse plugin only +Switching all issues that have been in resolved state for more than one month without further comments to closed status +If there is no issue currently open for adding WSDL support to Axis then Ill open a new one. +also just added the withConfiguration to the mapred classes as well as fixing the conf to get pushed down to the individual mapperreducer in or the s in +Asokan +Can you provide an ETA for this? Code freeze is today however we can wait a couple of days +Anyone else have any thoughts? Any objections to committing this patch as a first step +Can you please attach an XML file which causes the problem? Thanks +Permissions change from to . +Committed revision . +I really think were going just a bit too fast here +Try the nonSeq you attach the pdf file if it is not confidential +true the field is final and it should not have been changed +Changes to thrift in trunk introduced by to contrib per linked issues +Perhaps it would be OK to change the default to a more sensible value in? This would allow us to introduce the optimizations for those who need it in +If i substitute the sample it also works +A little late nowAnyway will these get included? Do we need help getting them includedWell anyway we have another set here if it helps +Closing for now +Patch implementing and Clinton this looks very interesting and given this is your first java experience pretty impressive too +The elements of aJava collection can be of any type so there is no way for beanutils tointrospect that the property type of each element must be a String +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Add for this new feature +Make sure you use the same version of jdk and xerces on both platformsthanksCan you humor me by trying it? Note that the order should be the exact same order as is present in the WSDLXSD +Claus API bundle exports the packages so the feature dont need to wrap for itI will update the feature file for it shortly +Lingering Marshallers are on a peer in which the collaborating application is no longer deployed +The only change from your patch is to avoid between rendering of tabs and choice +I did not search Suns bug database though +I agree with Simon +Hi All any resolution on this? Sam? Comments? This is seriously hindering our Lucy Incubator first release and a resolution would help us not have to go through any more effort to these dependenciesCheersChris +defer to +I believe that it only requires minor changes to the current implementation +Thanks for reporting. +then if you hit you will be automatically redirected to the mount +attached Quickstart and patch is applied with slight improvementsThanks! +Not sure of the implications of just using RPC until matures and how hard it would be to convert back to when ready +bulk close of all resolved issues as part of closing items. +I also validated all profiles on resolved issues. +IBMRational has a lite version of called LT but I couldnt find a trial version anywhere on their web is pretty much the same bug but with Visual Source Safe +I need to set Fix Version to but that isnt listed +It wont create new instances unless its necessary +After making the change suggested in the previous comment all tests pass locally for me +Had some trouble with the project +I assume you tested it with success? +I missed talking about the need to use Throttling with other +I have committed this +Another OS X glitch +Anything piped in via a unix like shell is getting properly converted to a string prior to execution +See previous remarks. +Thanks Ola it seems easy to fix and I was so lost +Changing the title of this issue and marking it as a data corrupting regression +Thanks verified. +Tested it and it fixes the dropping of rows in PE +more french ErwanYour patches are commited at revision +yes i think you are right ivan +fixed in this since has been released +Committed at subversion revision + for the patch +Yeah Im not too sure how youd catch this +A patch was created for this issue +I have attached the patch Stack +Ahmed the patch isnt applyingI closed as duplicate +solved with two new attributesappendMessages comma separated List of messages components to which messages will be appendedreplaceMessages comma separated List of messages components for which messages will be it true that you collect the messages of all configured components and then addreplace them on all the components on the client sideIf you have multiple message components on the page this would mean that each component will receive ALL collected messages no? +Im up for committing this to trunk +Add a thread cleanup listener +Closing the bug based on Kishans comment +Because prevous bookie has the znode ledgersavailable +plugins dir +BANG +This is the patch fileIve committed a modified version of this patch which allows you to plug in a implementation via the class +Dennis I have some ideas on this stuff it might be good if we could chat on the phone at some stage +We will include a note about this warning in the quick guide for uprade we will have for this release +In the mean time if we can get comments from community will help in making changes while we are still working on itRegardsAnilGreat progress +The change in this patch seems to result in a rather pointless test +Committed on trunk +When I run the unit tests in piggybank the new fails +so far this looks sufficient see for a followup +All tests passed on my machine +You dont want to be opening it here just checking the mtime +scm connection is changed the to the svn repository from Connector try it again +Ill have to ask Michael Gentry how he handled image files on the dev list +doesnt passing sstable make the descriptor param redundantThere is another constructor that passes only the descriptor +The is not functionally end activity will trigger the continuation of the subprocess even when multiple parallel paths exist +Hi JBDont worry I dig the code for while and found there are some work on the CXF side and also some work on side I will commit the fixes shortly +Excellent work Sijie + formatting line longer than characters +Converting what the user entered into a type suitable for passing to business logic is not the role of the form bean +small one forFor the license are you referring to the content generated by jbossnetworkdownload? This is the only place I have seen thus far in the CSP code that deals with licenses +Still some TODOs but for the most part this works well +I wouldnt tell you that you need to do this though +Yes it does +This is a complex change and after only saves the time to write one not fourSuspending work on this for now +I dont seem to have the option to reopen this bug but I do think there should be nothing in an XSLT stylesheet that should be able to cause Java to throw an Error +is yours a more version? The one we have in jrubybuildlib is definitely not the complete version though +on stacks comments +Merged pull request in master. + integrated to trunk. +Sorry for the delay again I just forgot about the issue +that seems to fit +We can think a bit more if adding a separate create topic API is a better strategy +product file so I didnt do so much work here just configured all parts of +If isolating the timing to just the percentile calculation shows that is where the latency difference is we should reopen +But I think the approach described by Robert Evans makes the most senseExcept it doesnt really work in the context of +Can you please put together a small test project that we can use to test this issue +new patch fix the appreciate your time to try out the patch +minor modification to patch + fixes some of the fixes the other half of the is the other half of the fix +Need to add tests yet too late tonight + for porting to the trunk +I just committed this +CharithaWill you be able to see whether what actually happeningThanks screen shot of tomcat simishagIm making this bug resolved since we couldnt reproduce the bug by following the specified steps +I pushed a tag that has this fix +Im attaching a test case running a number of identical actions sequentially on an embedded broker with a tcp connector +I have merged this to. +That certainly can only happen when the Project Explorer has at least the same features as the Package there is no technical way of extending the Package Explorer and also that is going to go away anyway Im closing this issue +This will need to wait until Java is no longer our target language +Ive committed this to trunk and +fix committed in r +The test supports new implementation +I can read it but yes its too small +Let me do a final review cycle today again +If nobody objects +Ive set encoding for both cases then Ive try to open VPE does not display symbols +out ResultTime to build schema type system secondsTime to generate code secondsTime to compile code secondsCompiled types to +locally I see thisls l mail justinmclean justinmclean Aug mail homeapmail +Applied +Marking bugs with invalid fixed for release as fixed in. +Then it can allow Format be added more easier +serveralldeploylib +No idea why we trigger a Hive build +Agree could you provide a patch +I tested it with and an old +An updated jar is included in this attachment +The addition of std namespace is for the upcasestring methodShouldnt the fixtures data be regenerated automatically? Or am I missing something +and + +After applying the patch I am getting exceptions in all new integration tests too +It even includes errors messages translationsSo to be really picky it would be even better with a paragraph or two in the user guide and an entry in the file to explain the new featureThanks for the good in revision . +This patch +I linked the issue that made the permissions optional +Note from Julian uses the version from EAP CP namely +Unfortunately this would seem to rule out support for callable types with ref arguments. +He might provide an additional patch containing a documentation update and a test +Indeed committed accidentallyI left it and just updated +Can be reopened if issues found +Attaching preliminary patch +I just committed this +patch which includes a test case for the code for not supplying a test case +It should also be applied to the trunk +We cannot just default to possible data loss out of the box +Cancelling stale patch +Do we know whick state it is in when the target server is killed +thing u please try is to put some log in the and print controller name and see if request is atleast coming till there? is not supposed to cache response model I suppose to directly decide the view +Done. +Thanks for this +is no better than writing an Ant script for leave the issue open for now +groupId must change the groupId I hope its enough to upload beansbinding to the maven repository +Greg if you are able to test this and find this fix insufficient please reopen this issue +The inconvenienceof typing cd solr is trivial it was just a surprise that the behavior changed andI worried that testing was brokenTheres no purpose served by making things complex just so I can do what Ivealways done +I think we should have Administrators group for each partition and also for the whole systemTo have administrators for each partition we can create a container ouAdministrators with each partition created and we can arrange ACI accordingly that group for that partitionTo have system wide administrators there are two choices +Seems to be the same bug in workaround +Further investigation is required. +This file is not ignored for but this targets the trunk and not I guess the issue is resolvedThanks again Andreas Pieber for your contribution +I would be extremely happy to help with writing tests and help fix any server or client side bugs that crop up +for the change +The rubypsecs are updated with guard to skip the instantiate? tests with comment linking to this issue +I always stumble over double negatives like not noShutdown +external srcmainorgexolabcastormappingxml executed test suites of XML and JDO and added and so commit ASAP. +When I updated both to use your tests pass +seems unnecessary +Upstream issue was resolved in AS +Looks good Grant +Got it Ill let you know if I have any questions +So I am committing the patchSending fileinstallsrcmainjavaorgapachefelixfileinstallinternalSending fileinstallsrcmainjavaorgapachefelixfileinstallinternalTransmitting file dataCommitted revision +This version of the patch supports overflow on patterns but it requires the SVN version of Batik as I needed to add access to the overflow member of the class +Yep it is correct that I was using unzip command +Delete Region directory in HDFS +Attached reworked patch which includes and please have a look at it +Ahad if you agree lets close this JIRA as a duplicate +I just went straight to the TLSSSL Support section since that is what I was after +Closing resolved issues for already released versions +Will be available in next dev build +Committers can you please review +Overwriting the version of to in quickstarts pom solved the problem +Checked in documentation fix with revision . +If the fully qualified class name would be used then there is no need for an annotaton +using the time of your post the latest snapshot dated back to April the th I closed this issue May the Snapshots have been deployed. +Should we allow any order +Fixes the javadoc comments in so that the examples provided are correctNo new unit test added since the examples are already present in the existing unit test +Thanks issue is fixed and released as part of release +The code is out to the Smoking Man that wont compile either and he is the release manager +Thanks for the help + +Committed plugins changes with revision +I dont know if it will be much slower than alternatives and if its not fine for eval purposes +Added +Ive spotted this problem during working with this file Ive attached one page with bullet punctuation +Attached +Just tried on GEP same excpetion occurs +shouldnt we remove the constant then +It would be great to get PAX URL upgraded to to solve for the default and fallback repository handling and have the above configuration parameters added to the default etc fileThanks! etc file comes from Karaf its not overrided byMoreover uses Karaf and so PAX URL +thanks again +Attaching screenshot of the scheduling information after the patch has been applied +Fixed in Apache Karaf. +The MVEL author is looking into this +The way it works now I will always get the domain name given in the server constructor while I would like to get the address configured by the user +When a directory exists it returns null +Instead of building everytime we call addEntryreadEntry of its will be better to build inside itself and pass onto +I really need this +Adding to the affected versions since was backported to the branch +You are correct that would be a big departure +looks good lets get this to trunk and branchesThanks for the patch Hari! +another alternative would be checking it against the incoming request in addition to the outgoing request +you many want check for the existence of file instead of directory for HADOOPCOMMONHOME and the if statement to be more consistent in checking for the location of +So removing the injection from wont impact the validity of the test against EAP +tar or + +Hey folks sorry about the breakage +The query appears to return the expected result in Derby and later +uploading the patch +mike thanks for the test! Indeed it takes care of merging though +I will add some support for a property that enables the fail behavior but I think the default value for this property needs to be false +Please let me know if this is wrong +It happens somewhere inside JDT and the reason is empty passed to parser but why is empty +We will have two versions in CP +So we could put that code into package of if you want this code to be included in the next release +Ill just leave this resolved. +Since the groovy console feature is not available in right now this feature should be and moved to the GRECLIPSE project in jira +Modified UIMA AS client side to allow multiple connections to brokers +Fixed and available in trunk. +I was wondering if it is really a good idea to have both options normal and exponential implemented +Doesnt seem to be working at the moment. +Patch that adds a method such that the entire metadata map may be recovered via +fbdfdffaAdded documentation added a bunch of unit tests to test the functionalities that the documentation claims to have +Looks like its deadlocking in? There are some failures a null pointer and a delete count Ill look at +Fixed in Trunk in r +But both runs are broken +Initial beta version +The proxy is not currently optional +Yeah this is a show stopper for me because Ive run into a bug in the javac compiler that eclipse doesnt have +Attached agent log and the corresponding ms log +Fixed in nightly build +We are also interested in thisWorked on a prototype with having one instance but underlying there will be multiple writer instances +Duplicate of closing stale resolved issues +However Im not sure we want to deprecate the tag +This is actually a problem with the +Is it the same as +I ran derbyall and the tests with no failures +rev +Patch available in branch commit eceefabdedadcfdbedfIncludes fix for for the fix MatthieuMerged in dev commit ebbcaffaecfabccd +For Jackson use +Im assigning this issue to myself +Hi ChuckIve done almost the same changes to the template but your template seemed to have some good code which I wanted to incorporate +I opened a new jira to make it work again +the config directory i hope its the right one +Thanks for the report Claus +Thanks Justin +was the issue that i thought fixed this +Especially when dynamically adding loggers +Ill put versions older than in a sub menu +Integrated to trunkThanks for the patch JeffThanks for the review Sergey +I simply cant believe that this hadnt been thought about before +When we hit that point well just move everything to since I dont think it will be possible to select it with changes there are some small but incompatible API changes +I wasnt thinking that wed be doing dynamic weinre injection but dont have a big issue with it +This should now be possible as per this image +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I got rid of it with revision +Just tested it with evaluation version of datadirect and here it works without any problemsThis is a bug in oracle driver and comments in the sqlIts weird since all the the other sql runs fine +Hi Martijnthank you for the reply and the updated test case +When you save the file then the builder validator creates problem markers +Also small typo +Given that were trying to eliminate the timeout in does retrying still make sense +Exhausted scanners are closed automaticallyThis actually relates to +Also changed to cocoa instead of for the fix Eric! +we know what the queue configuration is +passed +The request of the bookmark would be a non faces request generates a faces responseThis must be handled by a special servletBut I have setup a working example how your can handled by jsfPlease look at the web appThe url of the example would be httplocalhostfacesThe name is a link to a detail page. +We experience this error on some installations at our customers +Tests passed cleanly for me +Ill see if I can fix this at the same time +Ill do that + +I am wondering about bundlejob and workflowJob whether to need to be paged likeprivate get in there is no args like OFFSETPARAM or LENPARAMand My advice is YES +This allows the to be set in a manner similar to the messaging API in c and python +Thanks again for the review Suresh! Ive committed this to trunk and +Freeman Im not using Servicemix just the bundles from maven central repository with and Felix +Thanks! +patch applied in rev +I think we can call it doneIdeally we should move from using colors in the dumpable stuff to CSS classes +has been marked as a duplicate of this bug +While youre in this area currently if you have set up security and have enabled Hadoop service authorization there is only one ACL that grants permission to however clients will want to access a subset of it for listing tables +Rescheduling to X +Juergen +Eclipse were used +Can always improve if needed +Thanks Sid +Fix to the issue resolves this issue tooCare must be taken in adding new values to AXISCEXCEPTIONS in the future +Datanode doesnt have namenode field anymore so we need to add ability to update namenode per block pool +Lets resolve this issue and reopen if the issue is still there +Heres a patch that takes a better approachIt uses the version to tell which type of bloomfilter to use +Unsure if i am supposed to Resolve Issue so I will leave it up to the powers that beThanks +I recently saw the same error and confirmed that the host was not listed in +Is there an update on what the status is of this request? I have a similar need for validation to sometimes generate warnings that do not affect process flow but can be displayed +I have no permission to reopen this issue but I have now obtained and made available a bit windows vmware image for you to test with if you me at couldnt install this image said that there was no boot image found in the +We can suppress the error message for now +It differs only in single line +dll +push out to +Youve got listed as Coding in Progress so I didnt take it off you +Reproducing the scenario with a subclass of didnt reproduce any errors +In the mean time I suggest we revert these commits from trunk and until these two tests can be fixedWe usually do not revert patches because of unit test failures +This might be a reasonable patch or you might want to put the null check in the abstract superclass +Attaching corrections to the miscellaneous issues listed here +I tried really hard to implement it as a connector protocol option but ran into all kinds of conversion issues for example when the message gets storedloaded from disk since that was not a global option +The implementation does not provide any caching configuration for translators but it would be to create a custom delegating translator that implements a caching policy +There already is an issue for it is this marked as Fixed? I checked the SVN and it doesnt t this be then marked as Wont Fix or Postponed? +So are you saying that support for should be removed from the GPD for now +Then upgrade the version in and couple of other changesAttaching an example changes to DBCP for think this has been done can this issue be resolvedclosed +no unit test fix could impact other code paths too especially since the DN comparision is used by many code paths +I think in this case it will be flexible enough to cover almost all cases without further complication ofIf such solution is desirable Ill create a separate issue and link this ticket to it since ALLOWMACHINENAMES feature will perfectly cover localhost case and there will be no need for ALLOWLOCALHOST option +Proposed patch +The color used on version and version is the same that appears on the Apache Texen website +eeeh +Mike I am going offline for tonight +I noticed if you show create table on a table that dynamically gets all its columns from the serde the create table statement is not correct +Accidentally closed this issue +Attached updated patches that revert back to using the TCCL for resource resolution instead of and also remove from the hierarchy +Patch appliedTIA for your contributionGreetsGerhard +Also this is a question not a jira bug +I mean the Class and method naming isnt very wellBlacklist or blocklist? Whitelist or positivivelistDoes this answer the question +All issues has to be filed against from now +The codehaus jira displays edited comments containing the previous comment with a + +Currently we are downloading Classpath but compiling with ecj +Should we add code to just to work around another products bugs? My general view is that we shouldnt so Id be for applying this patch +If something bad happens we will check it again +Thank You +Merged patch to latest trunk +The patch is ready +it works with glassfish and +Ashutosh Thanks for the + +RegardsAnil Patel +bq +Not ready yet hadoop integration tests are failing +Patch committed as is follow up task created for test case +Should have closed. +Materializing owning entity has undesired side effects if owning entity has any eager fields additional SQL is generated which bring in data that is not requested by query +At that point the JNDI lookup should also work +I cannot reproduce this JIRA if it is wsdl specific +Patch is submitted +updated module +Patch applied thanks again +Is this still an issue post tar ball cleanup +Unfortunately I can test case is quite complex and requires a large Oracle database +My Googling skills arent up to finding the bigger oranger one today. +Site updated. +arg that static method is nothing but trouble! +So when can I see the glassfish artifact in central repo +This change cannot be tested until change from is available +Id think the line numbers would at least be different? Want to post the stack trace +bq +If we need to keep the on all other branches when a change is merged between two branches thats going to be fun times. +I just committed this +It did work for me later in the day on Saturday and Jacek was able to do a commit later in the day I noticed +ltroobot published gt +Will the templates be back to default after I install the next release +Am I wrong or what? Please advice. + this looks good to me +Closing issues +implement fire and forget implemnt robust send seperate listener case write test sample clients once the issues and are fixed do the necessary work in client API patch fixes the build break and corrects a small typo in listener manager +It appears to be coming from the same source though the patch I had provided does not solve it +As far as I am aware this is fixed. +Thanks +Wrap the long line +If I do I get always a lot of exceptions +Yes +I intend to fix this by pointing to the other locations not by reproducing the text in +to Erics commentsin particular dfs report dfs safemode and dfs decommision should be separated into a dfsadmin commandAnd an API to remove datanodes from the namenodes eternal memory would be good in general and useful in this is the first version of code change +I request for v to be considered for reviews and comments and if everything is OK a commit too +which class should be usedOleg I know its in its beta version but do we have any documentation for the? so i dont have to read the code by myself im lazy theres definetly a lot of concepts in the new components +The are probably stuff inside + +Have the mods in our fork +Fixed in both trunk and there a workaround or a patch I can apply to? Unfortunately I have very strict limits imposed by the Customer who does not want to upgrade its production servers to u have a look at the commits tab you might be able to work that out +Done at r +Ive committed thisUnit tests will be added with cant check codes in detail right nowLets try to fix your algorithm together +would be nice if you can state in prose the intent of your schema and also post a sample XML that you want to validate with the schema youve posted +Deprecated property and method +ear deployment seems to have been something that worked during the AS timeframe but does not work with the AS deployers +Looks good +The other problems in this JIRA are either implemented or there are no plans to implement them. +Can you please open a new bug report for your problemOtherwise I cannot mark the first part as ready for +Ill check out the patch and comment next +In the latter case it should be documented in the Sandesha user guide which Sandesha release works with which Axis release +I have to research a few important aspects like +Why is this marked Internal? We have a request to make it public +Right its just DOM manipulation and certainly could be done without the DOM helper +However I see nothing in the spec that backs up what Im currently doing but at the same time I dont see how the correct behavior is possible with what the spec says +The other test suites appear to be is based on revision of trunk +Attached patch makes this change as well as removing a processing from the log message +I encountered similar behavior as Todd reports using for an implementation class with only a subset of its methods exposed using annotations +Unit test to show the problem +Doc from with added sections on running the tests in minidistributed mode chaos monkey +Changing the title of this Jira to reflect the broader impact from attribute generate an exception when is used within the int elementto Refactor and Standardize possible XSD Schema Attributes across all to scope moving to Excel document Overview of Router configuration parameters available in Spring Integration and the proposed changes for Spring Integration Updated Spreadsheet with the final changes + diffing dgolovins workspace w devstudio trunk I found applied these changes to trunk and ran a local devstudio build +I observed this bug on master observed this too however it seems to be random +If my understanding is correct its a +Thanks +Bamboo has no permissions to do builds run as the same user giving it any permissions thus implies that it could do things to other projects +see comments on +bq +I also took the liberty to introduce +Patch looks ok to me David but it would be good to get Andrea to look it over as I am not that familiar with the various coverage uploading use cases so there could be an implication i am not considering. +Id rather be overly conservative than overly liberal with opening things up +The guys should be able to fix it quickly +And what OS are you using on the Camel side +Bean containers with application level DI should be the first target here +Please with a reproducible test case if the issue persists. +Not looking good on detecting this via HTTP version +does not +The problem is related to federation but others exist in trunk also +Im working on a repro but other stuff keeps pushing to the front so its taking longer than I had hopedAt the same time on Linux where this test passed with IBM jvms and the test left behind some databases with really big names one of them including only +Good catch +ramYes we will test the patch after Chinese spring holidays +Oops sorry +Test case +It does not prevent of double object creation +The version is attached but not committed +is failing +But if the release is already ready to go then it might be easier to wait for the next release +Looks good +Manually created for favourite WS. +I can compile and run unit tests on trunk just fine +Should be fixed. +Per the planning discussion for Sprint the activation piece is off the table for Sprint +A big chunk of the tc folder is clustering related and thats in Dont care anymore mode as well +MaxCould you please take a look at this patch +The project pom is a just dependency that must be available in one of the repositories for the build process to succeedClosing as WONTFIXOleg +in the last path element +apptask tasks reload correctly under rails and An integ exists for rails Ill duplicate it for rails and see what we can do. +This has already been fixed as a part of some other issue +See the top of the page +I am comfortable with that level of coupling +f works even without +Any idea what the reason could be +Himanshu I think you uploaded the wrong patch for v +In summary it is much more stable now after some configuration changes and I believe that the root of the problem is that H is simply not capable of handling the demands of the environment +Fixed in revision and some later revisions finally in revision transition resolvedfixed bugs to closedfixed +Same as before though this time as a standard Maven project +I dont think that there will be a problem with the new title +If the flush fails wed flush these edit into the next log fileFor syncing youre right +for regular releases for snapshotsWell this is only a problem with false I guess +Cool +Ive committed this to trunk +fixed in commit rev +Bruno any updates on this one? Or at least a bit more info so that I can take a look +bulk move of all unresolved issues from to +no its not if I understand this right maven currently allows the same product to obtained from the maven repository under two separate groupIds causing two different version of a jar to be included on the class path and there is no way to solve this +Verified on jbdevstudio. +Currently I think the only way to stop compaction loop is to make sure interested sstable does not have overlap so its tombstones actually drop +Also updated the JSON schema so that the channelConfiguration attribute is optional because Hibernate Search provides one +Actually I sent a note to application teams mailing list to collect their opinions about changing syntax and received endorsement from Vijay +This has raised a number of issues and will be changed to a blocker +Ah so this is more of question for the dev forum rather than an issue at this point +All references to Portal I think would need to be changed Enterprise Portal Platform +Kindly review +Final patch that was committed including rpc version change +The use case isI am a component +Whatever the outcome it will be a useful place to forward future discussions like this that may crop think in an ideal world having only JTA would be fine +thx +Thank to great work of bulk insert +Heres a lot more rigorous testing +Both the table and proc return as connect and query +attached +why not maintain a total order +My ordered list numbering skils are above +Is there any plans to continue looking into this effort? or are the current ant tasks sufficientAlso what is the current status with renaming it to jsfunity +why create strings eitherGood point +If we are going to require +The job queue details page just provides the user the position of his job with respect to the other jobs in the queue +Can you do me a favor and try this with available from the download page +Since it looks like subtask B will not be in R Im moving this task to is a rebranding task so should be completed by is going to review this item +Perhaps Jetty renders JSPs simultaneously using multiple threads instead of rendering them simultaneously using a single thread +John Casey confirmed that seemed to be the case and he said he would be creating a bug report later +If this is not still an issue please close the issue +By the way when will the be released +Full stack you see that behavior in? If I recall properly a similar issue was fixed in +MaxPlease the crap outta this so it can bounce to QE for vetting +That makes the tests pass in my environment +Verified in JBDS h Linux +Ive checked attached patch and it seems that current implementation is trying to implement the logic on sink side +Getting late here +Thomas do you think it is large enough need a CLA to be submitted? +It seems you use some old templates which are incorrect for Eclipse WTP +Thanks for waiting guysRegards a compile bug +Its a know issue that the standalone EJB should export its ejb interface via so that other ejb client can load the ejb interface class from the bundleI think we should add this to doc somewhere before finding a real solution. +Merged into in revision after releasing +I have refactored the test cases so that it uses the verifier through real sample projects. +But eventually if you monitor the usage you will see it starts following a familar saw tooth usage pattern since the store usage employs garbage collection cycles to reclaim disk space +How will it be usedIts a common GIS technique for reducing the number of termsspace to enumerate in a single field by dynamically selecting the appropriate tier based on the latlon input and a distance +apply on luni moduleThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I fixed it under r +from me +I was not running with all the prints as it is likely that was also slowing down the system +Now its your change it was MAX After your change it was MIN because of overflow +I considered only disable table scenario in this patch since table need to be disabled before doing any dropmodify table +So you will have to print out the logs on the console so that you know where it is hanging tryant test Dtestcase also will help to capture the output +contains removals of variables and method calls resulting in dead stores +need to retroffiting all shutdownhooks in fixesthe failure seems unrelated as it is also failing on a fresh trunk checkout +Its certainly a complicated problem specially because of its reproducibility +The fix helps reducing the chance of losing data when hflushingShall we commit it to the append branch +Reason is currently our failover model is based on groups +Wouldnt hex be a little more reasonable particularly for large byte arrays +But only some of the page is being updated +bq +This ones yours patch makes a two Swing tag interface changes Invalid bean properties will now cause the Swing script to fail at runtime +this is patch for avoid call +Well it seems this has nothing to do with changing the compaction strategy +Seems Ok from standpoint +Once completely setup possibly in an initialization phase or in a factory or in any way the main application provides they can be used to repeatedly find roots on a set of functionsIll check in a first try on this so you can provide feedbackThanks for this important +Of course will have a look at it tomorrow. +How can I try this out? Do I need to wait for? +change committed to SVN thanks for the fix +Provided as an extension of which creates a suite containing each story path and resolving the path to a human readable form that plays nicely with any IDE that provides integration +Bulk move of M to actually more a feature request +Ive applied the first part of your patch with modifications +Is the patch backward compatible? Function get seems to be new + + works now thanks for help. +Another small correction +Whenever I put to the webapps lib directory on and try to open a JSP +The patch converts the class to a Java enumeration +javac +I will post comments soon +Im confused +fixed stagegraph for +yes I just didnt want to write a complex title for this issue +Works well Simon but not with IE and IE where the dropdown list can disappear quickly +This has been discussed on the mailing list and the general agreement is that even though the are far from ideal they are still simple enough to understand and illustrate the usage of Nutch API and they provide an acceptable UI for usage. +Not suitable for backport bug introduced post closing +Patch with test case +This was fixed back in +Ill have to verify thatLooking at again actually I cant seem to paste at all +Last final version is do we want to update to a SNAPSHOT version +I see what you meanMy problem is this I would like to include a preview dialog in my applicationand its very unfortunate when the preview looks different than the resulting PDFI would suggest that you either make metrics files available for or make sure that these metrics files are identical to the AWT font metricsRegards Peter +So we will change the API at that time +Should I update the website or has it to be done by someone of the legal PMC +This is now landed on the new incompatible branch +bulk close of all resolved issues as part of closing items. +I have just committed it +bulk defer of unresolved bugs to +Closing as dup. +Attach a new patch address both positions +Would it not make sense for the application in question to put their hostnameport into the znode data for the ephemeral node +this fix does not allow the user to override the pk key name from the +Probably not considered I only took a look at implementation of after this issue was reported +Ready for regression of updated patch passed +Patch for and for t you instead use a generated from your crawl seed +Noted as incompatible in +Cool! Thank you for letting us know +is this still a valid issueimo is populated by resolveand i see Type is already binded in am i missing something +Next issue for reference manual was resolved during translation +This setup too gives asimilar problem +Attaching a functional specification +So my pull request is fine in that it fixes the symptom +Bob bumped to for now +file added with includes changes in and note for bidirectional tests we have some TODO maybe more bugs +Jostling around the release goals and numbers +Close issue in preparation for release. +End testProtocol derbynetmatsderbynetmats Do you think this indicates a build failure? A test setup problem +Patch to disable the dialogs finish button when the validation status is cancel and fixing the validation message review as correct? Posted from Bugbox for applied on applied on branch and on Tools +the javadoc screen shot +Regardless it would be far better to focus on shipping support for more concise formats or ways to compose the model of other external fragments in a generic fashion rather than relying on XML tooling specifically. +First cut of patch +I used partially the code provide in this patch. +i aint gonna fix this +This reinforces the value of the limit +war attached +Thanks! +Those are all tasks I created during the migration +I cribbed part of that function from a different place and then didnt think hard enough +Seems to be a Java compiler bug that me +I just committed this. +It works anyway for absolute references. +I will review and commit the patch +Fixed +EMT build applied at . +rev implements this and uses it in jetty and jasper +So what I have to do to safe use of slfj with +Postponed to to with Thanks Anurag! +Resolving this jira as invalid. +This patch doesnt handle inner beans which are not implementations it should check then whether methodName is set +Go for it +Activity engine redesigned in Jive +This patch is based onThere are instances ofsupports that have been changed to add a case that returns true for Safari +So far as I know the jumping code is working properlyThanks thats good to know +I need to add another test for holder parameter using the attached wsdl +thanks for pointing this out +I also looked at the html version of the patch which saysWhen the Network Server is started by the servlet interface shutting down the Application Server also shuts the Network Server down once both run in the same JVM +its rough! But at least passes wI moved s loaded into RAM wrappers up into this way a codec only must impl the direct source and can impl source if it wants to +I notice this has fix version of + +I would also add here that the same code is working fine if deployed into +The absolut portion of the path is always prepended +This patch has a faster comparison +I will commit it shortly +This is already available as is the set of changes to the are the tomcat changes did not see the link to upload multiple files +If the debugger is running I can see my changes the maven side I am using the to put everything into the target this setup I cannot use meclipse without seperated class folder +Thanks Jon! I committed this to trunk and +If there are no objections from anyone else I will commit tonightHow do you feel about the changes toI think we should keep doAuthentication and followRedirects as they are still in use and do not have a replacement yet +As far as I know is only in the smoke tests and that uses a WAR +Tested by editing the url viewing the list of projects and returning to the group +looks good to me +LGTM +There were empty spaces after the +I will repeat the test with latest Axis +The new method has passed QA closing. +Merged into the master branch of the examples repository. +Also added options from key records and removed the logic to output constraints on columns. +Thanks PhilJust to clarify you get the NPE with the installed? Running maven orIt shouldnt be necessary to comment the assert out in any case +Yep seperating out a Color Ramp from a Category from the current Color Brewer based palette generate would help make things clear. +if there were it would simply skip the generation of the finder which is not we want +Havent had any feedback on the resolution to this one so Ill close this now but if the xmlnsns issue raises its head again we can create a new because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +No Ill fix it only for maven ll add a new goal like ant goal antexecute is added +Just deploy this portlet on Portal and try to register it as a remote Portlet using WSRP +Initial changes with commit fabbaedbbabaf +Therefore I will shift this issue to next CP release +This issue is caused by the btadmin forces to quit in the mode +Implemented +What I called standard validation is another extension to that extension point that checks xml consistency requiring it to be well formed and compatible with declared schema but of course is unaware of any ESB specifics so that by our extension we do not compete with other validators but add to themI will attach two screenshots displaying what I already implemented some time ago as sample rules +The type mismatch can easily be fixed +This may be due to changes in how we handle anonymous type definitions +Ok lets do this release addupdate to our add to JBT and JBDS discovery connectors write blog record screencast to document site built locallyIt currently points at the did you load into the discovery site? Did you then point Eclipse at that same site? +Committed revision Thanks Noble +Installation using the Graphical Installer contains also reference to Federation as Tech PreviewThis should be changed also +The current discussion mainly for +If you copy a file to hdfs and specify that in libjars it should reuse it +Example JMS Client with s disabled +Please verify +This does NOT lead to incorrect behavior but I suggest to set returnentirecachetrue +I discovered this small discrepancy when writing the documentation for this enhancement +In these collections were changed to +It also is consistent with the behavior of the RI as they appear to be doing lazy loading after taking comparative heapdumps and analyzing classes loaded +Its much less work for the user to manually override a field back to default if necessary than to populate the entire w the current values to avoid stomping themIf there are optional fields where null is a valid value then we have a bug +Does not affect appears to be little demand for this so I suggest that you implement a helper class to do this for youIf there is sufficient interest and the property names can be agreed then the helper class could be added to the pool distribution. +Start EAP server is not needed just replace it with +We can still support if the users want to configure APT +Modules that wish to contribute dyanmically loaded extensions can do whats done in the following samples +Andras andor David feel free to attach a patch that deals with the suggestion +OK Talking to people here it looks like there are a significant number of users that I know of that set the timeout to so I am going to come up with a new patch to have the timeout be acceptable too +How do I make sure I dont introduce problems like that in the future +Reopen to update attachment +Marking as closed. +Transitioned from Story to Task in preparation for an update to our JIRA Issue Type anyone using bnd to assist in manifest generation its important to note that youll need to build against the bundles in EBR in order to have accurately versioned package imports +Removing target release from issues that fit any of the following criteria No known way to fix them Java integration enhancements out of scope for release Other out of scope issues for d like to make this one happen in if possible. +And we will fix problem in next version +The job submit directory is an example of such a temporary directory but that wouldnt be created until after the user has called setRanges and submitted the jobMaybe I should revisit deleting the ranges from the job configuration when getSplits is called after the ranges are read +Actually TCPSESSION and REMOTEADDR are not require +Rollback is never going to be a first class feature +apply the patch and mark it as fixed +Fixed in CP branch drseuss tcunning svn commit srcmainwebapp Sending srcmainwebapp file data + +Can status be updated on this task +Ill update to address this +In Hbase each col fam is a locality group +so that might work +The patch only inlcudes implementation fixBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +procmailrc to the reviewboard user home dir whats next +Stuff in rivet is quite safely based on pure +Hi KalleWas this finished? I cant remember +Record which demonstrates the issue. +Copy saying something about Infinispan distributing to the farthest reaches of the cloudOr two peaks of a Mt +Here is the patch which will solve this DiveshYour patch is in rev +Verified with ER build +FurthermoreExamples are EWP are without syntax highlightingDifferent indentation of code in chapter Please resolve the issue and assign back to me when all the changes are applied +where is the problem coming from the java sources have the copyright header and this file does notWhat are the possible course of action to fix this issue +Are we just stuck until version or does someone have an idea how to manage this +To solve this problem we should the way we traverse the configuredCurrently the first is asked for bundle and if it is not found then is usedTo fix it we should ask all for check for the key in them and finally ask them again for +Should be fixed in the branch in revision +bulk move of all unresolved issues from to +Lets try submitting again since it seems my test results disappeared +also requests a jetty upgrade +Aside from the patches attached in the previous file there are no modifications to the stock code +In this case the activation method is completed when the component is loaded and by the time it is need and bound to another component there is no problem +Hey JakobAny chance youll be able to address Alejandros feedback or should we? Would be great to get this in +So this appears to be a corruption caused by soft upgrade +Planning to cut a release of Enunciate +Theres really only one right way to configure it +Committed r +Fix included in +What about calling this file something like plexus? If its called there might be some issues if we later will implement discovery of all plexus files +Fixed on trunk revision +Ill update this issue as soon as the updated mirror has fully synced up with the version history in svn after which also Github should automatically recover +The example file no longer causes problems with the latest trunk so I guess this was fixed along other issues with the new RTF parser code that was added recently. +In no scenario have I observed this to be slower +It will be pushed back up to the project doc at the next merge. +I was not aware of the handling exception bug +whitespace to master. +If improves its detection mechanism there will be a period of time with a better detection time from than +The equality checks on Method have been changed so they no longer check for equality of the declaring annotated type +However the way it works today is that the instances are cached and as such you are the same object this is due to the memory channel +Added the required settings to so at least you have a place to put them +Anil could you take a look on this failing test please +Thanks +Using this version of changed the performance dramaticallyNow the Profiler says the script takes up of the processing time as opposed to to mention The Source file in question is in packageI never call this function myself +Gotcha +Perhaps this could be commented in the POM file as its not at all obviousCheers Chris. +Proposed patch saves current frame information in topmost mn frame while iterating through the unwinding stack +So all issues not directly assigned to an AS release are being closed +this seems pretty nasty to locate and fix +At least that was the original idea +So how do I start to generate legacy joins per this adapter +Looking into as it seems a JIRA touching didnt make it there yet and the merge does not cleanly apply +RyanYou mentioned you had a patch for this issue but I dont see it +Its really just about the functions to resolve a contextual reference +I will test +Instead of changing the constructor you can add a new one that takes the default Once you do the above changes shouldnt really be needed +its very strange but I can not reproduct this at the moment will try clicking my way again later so please ignore for the moment +While it would be possible to structure it this way adding a target to produce instrumented jars is far more coherent and maintainable than maintaining a parallel build system +Slip we need feedback from Servlet guys +the warning am closing the issue +Primarily I tried to make the connection between the queues and schedulers more explicit +its fixed in the branch and the trunk +I did test it against CVS at the time as I was still using the CVS version then + +But I have to manually remove dependency for using Java +There seems to be a perception that it is more difficult to become a committer than it actually is +It is not good to have mayn issues discussed and solved in just one a test case is added we cann put this into the codebase +Similarly for +Test client for reproducing the issueIt will fail the call comment out the line with sets the SOAP version and no exception will be thrown +Note a particular problem here is this depends upon bugfixes in EJB I doubt we will be able to do it in the timeframe +Alex please note that this ugly workaround is a timebomb waiting to hurt you badly in case of session passivationclustering I guess +If not the default implementation from cxf will be calledI will provide a Test case for this hopefully by tomorrow +Committed to trunk +This is a great idea and the patch looks good overall +We certainly wont improve our testing situation by perpetuating it +Committed to trunk. +How do you resolve? Could you provide its +I updated patch +Attachment has been added with description small patch a comment header on +I also made the mistake of putting the set into the screen definition but my analysis still holds for form actions too also uses for the value +the configuration works +The new patch corrects the issues pointed by Thomas +can you add a test for this +Will commit a fix later today +bulk defer of items from to +adds more tests and fixes some problems they identifiedShould I also add increasing decreasing and unordered keywords before I commit this or should we commit this and address ordering in a separate issue +birt ChatreeCould you please re attach the file to this page but setting Grant ASF License checkIt is important because if you dont grant this then we will not be able to commit it to the trunk +If we are to do anything here I think I like Jans proposal best. +Both Paul and Uwe are absolutely correctI had blindly add this without looking around to see other solutions within LuceneThe constructor of using Filter is exactly the answer to Uwes question on segmented readersThanksJohn +Could you grab the latest code and let me know if it works or notColm +HelloWould you please apply the patch soon? Thanks a lotBest regards GeorgeThe default case and the overall fix look fine +attached is the patch the author provided to the moman source code that fixes the bug +Added patch for applied into wci trunk. +The issue seems to be either with the installation of the chef gem or most probably with PATH and the way the commands are ran +Attach a simple patch that add echo cid to the container script +Im going to close this issue +Already had such error +Strictly speaking theyre only required when updating the DB but are also applied during generation to allow for changes to the filters +Even if it is not an exact Matrix structure we can start with d hash tables and proceed later +was preventing us from using and +Thanks +Maybe we can use something like in there +patch for to add german time is in +The worst of them is the on the block surrounding thetable +Hope you dont mind Emmanuel that I transfer this issue to youThanks +Fixed in r +great +The suggested fix is to check ordinary throwable message instead +So it will throw no exception and set it to if the retry fails +Some concerns though We need to find out a good buffer size to use for writing to the history file +regression caused by that fixes this patch in r +Im not sure about the you mentioned the test case passes with the fix +Paul sorry this didnt make into ALPHA this was an oversight on our side +Will if you are working on a docs patch please consider building on top of +The test doesnt appear to be failing in. +Trunk will be doing sometime later next week +In tracing the flow through the constructor during the none of the attempts to load the servicePolicy from different locations results in the policy being loaded +Please verify +Hi BrunoWe could create an open wiki page and link it from the home page +Any further questions please go to the users list +correction JBDS works fine on windows if you run with a needed thing is to get a windows build of xulrunner +Pushed changes to branch too and set new jenkins job in motion. +But this isnt the only service that is affected all services that are configurable had this effect +Alex Newman tested patch and confirmed it fixed his issues +This could potentially allow us to write a small C library that runs inside the Erlang VM but is loaded at runtime +I think I fixed the problems reported by Michael already long time ago. +can we improve the sanity checking to check for lack of existence of the JAR and then skip everything +It seems like Im still doing something wrong +MaxCan you still reproduce the issue using JBT or trunk +Ah ok +Actually it doesnt work in a regular class either +svn status for patch c should help if someone wants to apply itM javaorgapachejasperM javaorgapachejasperruntimeD javaorgapachejasperruntimeM javaorgapachejasperservletA javaorgapachejasperinstanceManagementA javaorgapachejasperinstanceManagementA javaorgapachejasperinstanceManagementA javaorgapachejasperinstanceManagementM javaorgapachejaspercompilerD javaorgapacheM javaorgapachecatalinacoreM javaorgapachecatalinacoreM javaorgapachecatalinacoreM javaorgapachecatalinacoreM javaorgapachecatalinacoreA javaorgapachecatalinadeployM javaorgapachecatalinadeployM javaorgapachecatalinadeployM javaorgapachecatalinadeployA javaorgapachecatalinainstanceManagementA javaorgapachecatalinainstanceManagementA javaorgapachecatalinainstanceManagementA javaorgapachecatalinainstanceManagementM javaorgapachecatalinastartupD javaorgapachecatalinautilX nativeconnectorM webappsexamplesM webappsexamplesclassesA webappsexamplesclasses version coalesces the two interfaces into one +Please test to see if it improves the would you please test your project with the latest? My change is not wiping out your nested annotations any more however I am still looking into Birgittas sample +Any estimation on when this issue will be fixed? Eagerly waiting +Refer to my comment about the addHost behaviourResolving this as fixed since the XS issue is already tracked by a different bug +Changed the to be sort of a blocking queue as per Jays suggestion +That is exactly what would be happening basically the issue is that the main appclient thread can spawn other threads so it is not possible to just shutdown MSC after the main method has finished +Andres did you see what garbles the icon +What abt done containers? killed containers? failed containers +Exception occured only at once in total +I will think about thisThough this is not available in other filesystems it is used by the generic layer as an argument to generic command +java root root +If applied to all tests pass +Im loading into the test repo nowWill update when its finished +Attaching patch file +Compare the explicitness of the these changes the problem described in this issue goes away +The read the homedir from the xml file +fixedon a side note warnings showing up under Code GenerationEntities should also show up under ProjectValidate Project +Have you tried this under or the nightly build? I believe we fixed it already +Patch applied +These should be separate patches but I ran out of energy +The implementation would need to be refactored into two separate implementations one for simple skips and one for stateful retries +Patch which implements event +Ah sorry +I have edited the udig workflow to force users to have the Developers project role to click on any of the new workflow should stop any further vandalism you will need to go through and remove the existing vandalism +Yup found this going to get merged onto master sometime? I would like to test this fix out and verify I dont get the error anymore. +Go ahead and submit a patch to do this +Hi ZhengWhat kernel is the newer box running? Is it a brand new kernel? has significant changes to dirty page writeback A simple program to test whether the box will block continuous writes for a long time +fine with me Michael +I really wanted to give a proper evaluation and thought with the refractions work that was done that was fully functional with the product +Henry just a question out of curiosity how do you think the cluster will be created on day one? Would a node cluster be built initially by powering on the servers using configuration and then nodes are addeddeleted dynamically? Or a cluster will be built incrementally using dynamic configuration change starting with a single node +So are you using WARBUNDLESJSFIMPL? Do you think thats causing the problem? can fix it if I ever get around to it +This is an old issue from the old project and since it was not you that filed it I think its a good idea to run this by the project +The patch in is better since it uses JodaThanksBob of. +FIXED IN Axis RC release +Please see how seam remoting is configured in that example +docFreq may not be a very good decision +another error msg thing. +The cause for this is different handling for source and binary types +User reports this works with new test structure. +Your CLA is in and processed Should be good to pull in now +This patch adds in a unit test that verifies that it still works even without any output +smile +Aliaksey do you save tag file after step ? Do you see atr and atr nodes under Interface node in Web Projects viewUntil tag file is saved kb builder will not update kb model +fixing dependency +Thats the old doing the ve successfully loaded your site in my browser +update doc curious why two official git repos +Note that is required as well +All right I try to deploy hadoop +not reproduced at r +Some events may be generated by the AM +But since this error is frequently thrown in perf test its going to create a problem because server will be busy writing log when number of concurrent users go up +Jason I think is missing from your patch +Closed after release. +The patch is not working +Is it known who will do this? It looks like is marked as blocker but its also not assigned and is not set for +It simply works +Thomas could you please quickly review s latest version just to be on the safe side? perfectly +Actually attach the correct diff +Try out the nightlies and let me know if any other problem +I think you should wrap the error even if is being thrown by a function being called +should I do anything for BPEL project wizard and deployment so that it can use your new deployment way +Do you see a problem in not removing the old constructor? If we dont want it to be used anywhere we could add annotation to it +For very large values continued fractions are diverging to +Sorry that I did not get it committed before my vacation +Could you please test againIf it is still present it will be great if you can come up with a test caseChamikara +Hope this is better +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +JulianCan you make a patch based on above analysis? correction for previous comment about scenario B thats the behavior with the patch if without patch rest will also get Unavailable after getMaster retries finally. +If unsure we can leave it as ERROR and file a jira to investigate +Thanks much GwenPatch applied in rI also added unit test for the fix. +Reopening to add a missing +q files +Maybe it is already fixed +If however you write a script we might consider its implementation +Other detectors will work normally +unless this is critical I am going to move it out to the next alpha for t able to include the other fields +Will have a discussion first and put it as a lower priority in my task list for now +Here is a patch with a proposed change +Patch from N +Having a validator subclass for each type of rule would be more of a JSF way +Seems like you have not added Please add that and create a patch againThanks +Requesting that this patch be applied to branch too +I added the ability to the resource response to +I dont see why we have to go through a complex valve or filter setup. +Im not sure what you mean by that but there should be a single way of distributing the scriptSo this is a non issue +Close for all issues before incubation +I also added some new unit tests to try out the function +Hi JuliusGo for it +HoweverI was about to commit all of this when I noticed containing as property keys +Ideally Id somehow hook into workspace change events and rebuild the list as necessary but I really dont know the Eclipse API that well +Am I looking in the wrong repository? Usingpserveranonymouscvsroot is known issue with project repository changes visible only a few hours later +theres a missing piece in the process +Thank Ari. +I will open a separate ticket about this +bq +That was done earlier +patch +Dag I made the update to the privileges label but I think that the recommendation about the column list is not necessary The syntax shows what is needed +Rado a bit of grepping shows me that the only modules using the module are and + +So you means this is a desired effect here? Yes this is by design +and comment it out in the parent pomNeed to test it in top down build +Mass closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Sounds good to me +The new feature module does not validate by default one has to call validation routines manually +Hi AbhiDont know why you mark this as Critical +If you find any problems feel free to open more issues contribute test cases and fixes. +bpel +but Ill let it beast a while more +I also confirm this bug +This is pending release of remoting +on and transition all resolved issue that did not see any further comment in the last month to closed status +Here are the new ones +Implement commands and +It was such a simple one even I have time to commit itPlease test and close this bug if it works +There are quite a few changes between and +While I was at it I ported the test to trunk where it was missing +I have removed the link to the ESB task as this is apparently unrelated to any planned work within the ESB. + +dummy object used for contributed code is now superceded by the solution implemented for +I intend to submit a new patch which is structured as a query parser shortly intended to reside in contribmcf +I wonder if this is also the underlying cause of I bet youre right +This is expected behaviour why is it raised as a bug +Then I killed java tasks and checked c processes dieing + for code please comment with affected test cases results +Should we include deltaspike plugin into CDI feature or should we create a new one? CDI extensions or something +I have attached a patch which contains the fix to the and a testcase +The has been implemented but I have not seen any use cases for it +It appears this change is going to be much larger than modifying the Findbugs Plugin as the Sonar Plugin API never looks inside the XML results file for priority levelSpecifically plugins that generate XML results after analyzing the source code use the an abstract class which does not dictate or orchestrate a method to retrieve the priority level from the XML file however there are abstract methods to get the message line number rule key etc from the XML fileI just want to make sure I have the right understanding here as a modification to the class will be architectural in nature and is bound to break plugins who extend this class +Applied +Can you look at this? Thanks +Is this possible +If I remove the default key pair the integration tests fail with and without the patch +Most of the code that needs to be touched can be found in the patch for ticket +Thanks +Fixed with the SVN commit +have the checkout in the normal layout instead of isolated directories This makes sense to me since Im using clearcase and each module creates the same clearcase snapshot view and hence the same data is repeated multiple times +One of projects functional aspect is to allow EJB invocations to happen through its API and internally it uses the project API to do that +I welcome any suggestion +But if I select another directory for deploy from another server I get the such path Cserverdefaultdeploy is it right action +Some basic documentation has been added to the site and downloadable zip files have been added. +Reopening to deal with timer on timer exception +I remember seeing one +Testing this now +It is brand new so go for it +Roo provides no Eclipse plugin +Thanks good to know +Is there an open ticket for this on the eclipse project I could go vote for? Thanks +Assigned to nobody so contributors can pick that of Benjamin! +Wow this is a really weird caseWe use +Apparently fully fixed a while back. +ps script to the arguments +Yes +Yes Sankaim facing a similar problem so what should i dowhat have you replied to the other mates facing this bugit isnt clear to meThanksAnna +bulk defer of unresolved bugs to +Its generally the responsibility of the to perform such recovery +Sorry about that +When user enter incorrect URL we not threw warning in eclipse log just ignore it +I have applied this patch as instructed and it seems to have worked +This is part of integration work and lets mark it as the first thing for phase +upgrading doesnt help +Weve come a long way since this issue was filed +I resolved this issue for SQL it was related to thread A closes connection needed by thread B +Run the test goal and check the zip file getting created in the target directory +Please note that has been created specifically to handle plateauxsince Im already working on this package Im happy to commit the patch on behalf of Christian +Ola Bini loves openssl +closed shipped in +kindly check +The change to for MAPFIXED was applied manually +So this can not be solved in Karaf +Though Im not too sure about retroactively adding new classified piecesI didnt use the existing binaries because ironically I felt uncomfortable in signing binaries that I hadnt produced myselfAre you sure that there is a sufficient intersection of needs maven and needs to motivate fixing this? If so the original binaries certainly could be published under a mutant version +So the build should no longer be broken and Hudsons opinions trustworthy once more +This issue does not have component set nor does it have affects version even though request to add this has been sent out +Thanks for the patch +IIRC the reason we recently changed the build was to avoid having the aggregate jar declare dependencies at runtime on the other jars which leads to the situation where a user downloads the aggregate jar which then causes all of the other jars to be downloaded and added to the path as well +For netscape SDK you can get it from mozilla site httpdirectory alternativly you can use I attatched. +This has entire cdi package patch from since thats not in svn +HiI am also seeing this same problem when running pig scripts concurrently in local mode +have java or higher +We need a deprecation period when the syntaxes are supported. +Assigning to you because this issue is related to the maven stuff. +There has to be mapping so jboss servlet container can find EJB and execute method on it +It seems to me that when the web service i served from the using httpThe only true erp address should be taken directly from the request urlhttplocalhostmytomcatserviceswsdl should becomehttplocalhostmytomcatservicesAnyway +Confirmed on my laptop with improvement over runs very low standard deviationGoing to do some microbenchmarks +Discussion moved to +I modified the trunk build to keep the report next timeJimmy I think is better than where it currently is +on generalization only when needed and in another jira +Assigning to Sarah +I have tested and just want to confirm that the proposed patch works fine withme. +So lets dont fix the new xml here and fix them with other xml files together laterSounds good +Update remove distinction between statistics and properties both are now properties remove notion of index property replace with Primary Key in Schema add more map encodingswhy did we remove the distinction between properties and statistics +Maybe something is happening when the prepared statements are closed that initiates a transaction +I upgraded the JDK to and the issue went away so not sure why an older version of java would create this issue but it works now so we can close this +Committed rename of as +Its probably best to start with the more recent release +So you did a test ingest no killing of anything and it just hangs? Is it possible that theres something else going on like the JVM running out of RAM? Attach a jstack of the tserver if you can +Sorry Im just getting to this +I would never expect this sort of tramping to happen +and it still works fine +The drop shadow isnt possible without some ugly css hacks unless you add another text node behind the one you want to shadow + try my patch +It looks to me like both uses of requiredTargets should actually be totalTargets +The file contains a test case that demonstrates what I believe is the same bug still present in svn revision +Updated version of looks good +AaronThank you for your feedback +pushed +Thanks Bryan +Are there values we could use that would make it so we could just turn this on? Or if we are letting go too many Chunks could we start up a pool if only for a short while to get over a write hump? Then turn it off again after? +The ack is really just a confirmation that the file in question rotated OK and was a sufficient length when it rotated +We are going productive in July +Now that is complete what work is left to close this ticket? My understanding of this ticket is that it represents work to ensure all endpoints in all being deployed are registered before any endpoints are started +Thanks for taking on this issue NicolasI think its good to change the default behavior for a synchronous Put to not hang around in the write buffer +Seems this is enough +When he gets back he can confirm whether or not this needs to go +It changed the shutdown logic of the Kafka server to go through following steps +As my states it is Oracle JDBC Driver version +but then we have to do that everywhere we catch a runtime exception +Yeah +Write a unit test to prove that this has been fixed by changes +im running os x ran into this problem and eventually figured out my RUBYOPTrubygems environment variable was the culprit +I didnt test the tag but presumably that will be updated again on the next releaseThanks for fixing this so quickly +Continuing on with simplicityclarity theme at this point of I do not see usefulness of these routines so instead of fixing them I removed them +hold off on until this has tests next week c +But I cant seem to find the email so I will justBasically I am writing a patch in nutch so nutch is compatible with +All existing branches would have to reformat their code before merging +it may catch something sooner. +Then when its time for commit well port to trunkOr as the old saying goes an out of date patch in Jira is better than no patch in Jira +Ive got the same issue +not observed in thats an known issue as discribed in the summary +Hi Daveyour patch is for xalan and not for xerces +CarlYou would notice that of the code deals with +thats because you running bit version of JBDS in bit JVM +In that case kindly point me as to where to look for this functionality. +migrate root volumescenario +New patch addressing the comments +Also you might want to review Dispatcherclean just to name another one test what is this URL split into two lines one now marking a Java labelThat was just a quick first review +Looks good in generalI think the statements should probably be +Newbie mistake +These dense sets can be common in cases where +find the patch attached +The tests still fail on this side of the Atlantic so I took the liberty of adding some code to the tests to calculate the expected timestamps in a way that takes the current time zone into account +I agree with Mathias that Id prefer not to commit a patch that allows the generation of buggy serialization +bugzilla configurator attachments together with the patch attached to makes highlighting bugzilla issue ids in CVS +The optimizers are now on by default and I am continuing to work on improving both their safety checks and the range of queries they can optimizeI have sent an email to the dev list soliciting people to test latest if they have queries which may be affected +Actually I spoke too soon +If you still see this issue when using please reopen and report on how you reproduced it +proposed to the branch in revision . +svn commit m Fixed incorrect metadata for persistent interfaces +What I am seeing is not an assertion error but the test hangs and just stops the whole build. +The path was manually tested on trunk in both secure and insecure mode +One minor nit +Will chat with the JOPR guys to clarify +Fancy taking a stab at this one +Thanks very much DagCommitted patch to documentation trunk at revision +I wasnt thinking about any encoding in particular +Theres nothing we can do to help +Given that we have RF Facelets archetype it should be take care of this when you are done with refactoring +Rossen could you apply this to the petclinic sample and resolve +I agree that this needs unit tests before it can be committed +Guess Ill look into JAXB thenBTW this probably isnt going to happen in the next release +Ive integrated your patch with some slight modifications updated and added a simple junit test. +Ive removed from the and it started workingSo we need this to be removed in pom +We did that connection as a kind of life ray signal +withservice +classpath files +Testing +probably not +Then that means you always operate on existing tables if they have a dot so you can just prevent conflicts at create time +Sorry lost track of this +OK I will test for the version +Ill work on it now +No errors observed +I know this is painful but better to do it now before we ship. +Also applied to branches by Mike so marking as resolved +No I have not used this +Timestamping also doesnt help us for updates after JuneSo were really left with continue to sign updates to existing products with the expired key +defer all issues to. +Ive updated as much as we canWe cannot update to Jetty on as its incompatible with Jetty +Seeing a NPE when trying to run remote jboss +The issue is in the portal code + might have something to do here but I think this comes from earlier +I just needed to clean up my buildThank you for your help +I have modified the errors such as Error should also be handled in Archiva maybe redirecting the error to an Archiva error page so that other links can still be accessed from the page. +Oops I meanErik does this work for you +Thanks John for reviewingthe change and contributing the test program +I will commit it if there is no more comments + +Thanks for the report +The project shows a dependency on in Maven noteI dont think anyone is suggesting you need to look into or fix the examples just post any detail on what the error is +Can you please attach a small failing project so that I can reproduce this problem + +The JMX API is really screwed up +It will be interesting because on RHEL with lboostsytem I am getting usrbinld cannot find please try building with cmake +This signal is sent through the command service which synchronizes and makes sure multiple threads are not executing simultaneously +Im currently working on site you want to use this plugin untar the archive and run mvn install from the root directory +The stopwords file is sent using distributed cache +Logged In YES useridI have found some memory leak caused by the use of and jboss jmx +Ill take it Mark! Fixes a bug and adds tests for Tokenizer where there were none before +The last line is where we do the cloning +Same for UNICAST and SEQUENCER +Using cookies to track the users locale is not industry default +Actually the CAMERA permission is not needed +Hi AaronI attached a patch for the issue +added a test that covers the binary sharing size increase situation with rev +We use aproxy in front and the redirects are causing problems +Still running tests in a loop no failure yetAttaching here so that I can get a run +Do you mean build dependenciesI think Owens plan is to use Ivy for dependencies on but not to bundle the core jar into releases like other jars we get from Ivy rather to force folks to explicitly install a compatible version of core +So I understand this issue is not a bug after all +I see this also +Are you saying that only annotated components are not tracked? I dont see how its possible +pid +Updated patch with deletion of temp files +Investigated is devoted to adding several to the build so it doesnt affect the documentation +This is a patch that Ive made for servicemix +Works as intended +Fixed. +proposed for the patch Felix +IMO it might be better if we tackled one issue of the RPM problem at a time with changes being clearer especially with the patches you provide +Can be commmited to smoke tests +Please reopen if necessary +Hi FreddyIs this anywhere in the pipeline? ThanksAlso is it the only way we can get a new notification +Ill have to have a deeper look +Infinispans also affected by this have now moved from the Sun HTTP server to Undertow so the underlying issue here is no longer applicable. +This has to be reverted in all the branches i feelCorrect me if am wrong +Dan lets do this for. +Bruno since you are watching this issue do you have any ideas on this? I dont really use so I cant say whether or no the immediate attribute is crucial +Fat clients wont have hibernating state at least if they use SS +Anyway wicket should not simply crash or freeze the browser with this number of header contributions components etc +Bulk close resolved issues +next +Thanks Parimal Gain for working on Leave management sectionThanks Harsha Chadhar for your patchYour changes are in rev Ashish bugs related to Employee Screens +This was fixed by in release +add code priority to major as this has been a problem for a long time with a workaround. +heres a diff merged to work with latest and you could even allow people to specify filters +Thanks +Indeed this is implemented for styles but that is all +They currently arentThe downsides of the lucenesolr divide works both waysA solr developer adds something to solr wo putting it in lucene only solr users benefitA lucene developer adds something to lucene wo adding support for that in solr only lucene users benefit +Do your patches build upon mine +Thanks a ton for figuring this out +Exporting services is not ideal as other jars use that as well + +Separating out the specification issues raised above +We just confirmed that a simple copy without any modification also cannot retain the informationI can imagine maybe the information related to fields are not stored within the documents but instead in some global area +patch +Thanks Ashish! +As Maven has introduced groupId to negate artifact collisions it seems very natural to use the very same mechanism for filename you introduce a finalName attribute applicable generically to all included jars or you honor the finalName attribute of each inclusion +The launcher will need to go into its own project for a executable jar then the rest of the classes in will be in another jar within the lib directory of the installation footprint +Id like to get it in +I have to add junit tests for that +was not part of +maybe source and xml or drl as the value? +Fixed in the trunkIt would be nice if you could confirm the fix resolves the issue so we can release +It seems that the has not be pushed to the snapshots this is fixed now +would be nice but I think its low priority given this is an admin UI +Miguel thank you for the jira and patch but the test case you provided still fail on my with your patch did I miss something? or it depends on +FYI were aiming for a Spring RC release on October initial cut of our package is in SVN in the meantime and will be part of Spring RC +I have got compilation errors in classes and +i ll upload another this is ok. +Add sockconntimeout to PING as well! Add lingertimeout to PING TCPGOSSIP Add sockreadtimeout to PING of these props need to call the relevant methods in for on head too +Yonik Please dont commit this now I almost finished up the search and will increase the test coverage until the end of the week so that would be a merge nightmare for methank youregards importsApplied against a clean copyAll test passed on machinesbest regards simon +I have not been able to reproduce but have not spent much time on it +The diff containing a fix for the tests +is the only workaround that we could propose at this time +attach a patch to compute addEntry latency in writecallback +I will email you back as soon as I had a chance to apply the patched versionThanks very muchVenkatt + +Thanks +but making a rd party contrib for everything that can use nutch may be worthwhile +There are still to much static methods +This is same as +Thanks Edward +do you want a link to the latest Tool documentation +Thanks +Similar ideas have been floating around internally and while the dedicated annotation approach will require some changes to the framework in core well see if we can work that in during the timeline +I think you want the XMLBEANS project the xml to object mapping rather than XBEAN reusable server components +This issue is reported against FUSE ESB but the only version of using the async send is the version of the component which I think is used in FUSE ESB and +This is not a bug +bq +Just thought Id put Trevs answer in +Im new to the whole confluence admin thing and I cant seem to assign this ticket to myself but just so you know Im working on getting this set up for you +This issue has now been fixed thanks +As an Ops guy Im basically looking for a least leader algo ie make the winner the node with the least leader roles otherwise random. +Closing this issue as invalid committed test contributed by Myrna with revision . +Theres a check to see if the test classes are in a jar that cant contain the annotations +I was unable to reproduce this problem using the Castor release +Committed a fix to +This is not a problem with rather it is a problem with compilation optimizations +Here is my patch for this issue including a test case in which invoke looks for a Facet of a hcolumn child ofI also tested with my implementation and it worked +Mark that test case passes so I guess the problem is fixed +Then we only need to save oldnew values of those columns referenced +More patches are attached +Support for this is already in resolved issues for already released versions +Fixed so that calling getLine with a null janitor no longer returns null and instead returns the source line +This is just a reality of life when using or which many users are still using +Having a second glace at the reference manual I think the majority of questions come from the first chapters + test case +Simplified code in the new patch +We were using Seam +I tried removing them for testing other bug and this problem was resolved +The latest patch addresses Hiteshs latest set of review comments +We basically just need a profile for example +Please cross checkthanksdims +Applied at r thanks +Thanks for reporting. +I do not have a installed on the and would thus not expect to see any content fromcomments +Thanks SurenkumarI checked this into trunk +Read your code more carefully than before +Committed as svn on trunk +The change looks good +While the row index has one index entry per row the column index has one index entry per group of columns +Fix applied in trunk apache site updated to reflect the change +Should we combine this patch with since theyre so related +Patch and unit test +Im wondering if there is something going on correlated to the DDL statements executed by the repro script which sometimes results in tables that return reasonable estimates and sometimes results in tables that return infinite estimatesI now believe this is not an accurate theory as the does notseem to be correlated to a particular run of the DDL statementsSo at this point I dont think that the depends on whether thedatabase and its tables were created by or by and I dont think thatthe is tied to a particular database and set of tables thesame database and tables can sometimes lead to one prepare time and sometimesto a different prepare time +Yes Gary it passed with your solutionBut why is it passed also before when it is invoked individually? +Using Portal to got the same issue until I updated from to It works fine now +In the documentation the is only meant to show that the value is an IP address and give an example of what it looks likeCheers +Fantastic thanks +Can you create a corresponding JBAS issue which is a blocker for JBAS beta release +I just committed this +Hi JasonComparing the error message logged in and here I dont think they are completely the same but since patch attached here will upgrade the version to I guess the patch might also resolve the issue mentioned inIm sure with patch here it at least wont hit any more +Merged the JDBC sink changes which reintroduce the local database source +Not really sure what youre trying to assert there but I guess what you wanna be comparing is that the thread that the executor thread creates is the same one used to run the async listener. +Patch attached + is still happening on AS master right now +Fixed +Wrong patch version submitted +Attached the WSDL that Axis generates for the service +Change resolution to proper version +We already have mass indexer the programmatic mapping aPI the query DSL APIAnd w eneed to finish them to release +I run your test case and reproduce this issue +All affected tests pass +Ill commit this tomorrow unless there are objections +The tracking url can be removed instead a can be added like we have on the UIAdded host rpcport and logsUrl to application need to add more stuff to +The feedbacl is good + Looks great for me +I added more to the class comment to better explain what class is about +It is not in +This patch seems to solve the problem +should we create a new issue for the verify error? It might just need a added or it might be an issue with the javassist version updateThanksI created for the new issue +Honestly I know in my cases I dont care to use any of the methods on URL object from within java not to mention they just throw checked exceptions anyway +This is a side effect of the automatic thread management feature added recently +Thanks Prasad! +Already the property is a resource or a file +I had misread the documentation to mean that the strategy doesnt come into effect if any Field annotations are used +should probably support Float same as Double and Integer meaning that a new kind of variable is needed in the language +Renaming subject to a more accurate one +I did not test it and it might be a premature optimization but wouldnt it be better to check if the stream is already a? +Attached latest patch +Maybe doing this will let me resolve it myself or determine that it is just my limited knowledge of cglib and is not a bug at all +This appears to have been committed as of Seam +Alexey Ive set this to blocker since its important this works but it might be that the error is just expected if is already thereAliaksey Please create such issues in JBIDE nothing JBDS specific about it +If everyone is fine with both options then I move we leave the final decision to Vijay who is coding it up +Wanted to check if you can post GIT ref here and change status +Moving all unassigned bugs out to to please move bugs back to which you will work on in the next weeks +done thanks +Fixed by upgrading ERLANG and COUCH +However implementing this might be hard. +This turns out to be pretty easy it is just a matter of changing the colwidth attributes in those tables to be relative rather than absolute +r +You may be up to something here +Can we make the heartbeat check the exclude list only if it is the first heartbeat received after the last invocation of dfsadmin refreshNodes command +I love removed code +Instead of using the service RPC address we are using RPC address for the webui access +Rolled out to all applicable and master jobs in the Trunk and views +Oh to be clear when is included then passes +Thanks for your comprehension +Patch has a way to use the R get to eliminate the hole and uses pretty much the same idea as r for R +Basically my point is if renaming revokes a lease the client is overpowered +this patch solves the problem for some extend +Thanks Andrew for the commit +We created a simple test case which can be run with Xalan Xerces +Perhaps I post them later as a diff between the base working copy with only those changes included and the svn tag I ve are would be very interested in the patch +Please check that it was applied as expected +I think we should just make a specialized accumulatoraggregator for the case that could wouldnt need to populate an and then make nd pass over the ords +I had issues with thats the value in the common xml +With new native launcher this is fixed just verified +patch applied on trunk and on branches closing all issues that have been in resolved state for more than one month without further comments +Not sure if Jody has any suggestion +They should all be fine now +Over time we have worked to slowly close those hope is the two projects can work collaboratively on more areas and in particular have common code whilst respecting distinctions between the projects and licensing +Renamed issue summary for better traceability discovery updated description for accuracy +srccontribzkpythonsrctest +And get is synchronized +Have added the append to be valid for var as well +Mmm this was a problem with hsql its all right with +Does this mean that there is a bugEitherWhen a procedure is called from a trigger any s returned are not explictly closed by the trigger calling mechanismorprocedures that return result sers should not be allowed in triggersMight need to look at the standard to see what is expected here +Added whats new pageModified linked pages to allow a better linking +If you just uncomment that example you wont get far because it wont be used by anything they will use the default chain assigned to null +Closing Out of Date issue. +However CLI error messages are a kind of public +because the existing implementation dumps all variables out to the generated javascript instead of magically knowing which ones to generateI would imagine that the taglib version in Struts does a similar thing so thanks for raising the issue over there +Then when writing out the file we can use both those data structures to see if we should use the original raw value a new updated value etc +Proposed updated version of to retain legacy doco of CVS access while adding doco of SVN access +Im fine with the changes +Should I just create srctest +We have a test directory for the native tests +Please assign this jira back to dev +RobI thought I attached the test projects sorry my bad +Iteration state saving works correctly with the fix in the linked issue the problem resolved. +Im still confused on how we end up with an id in there even tho we also remove it at the beginning of start +Mind if I take a crack at this +Curious that only one letter is affected +and +Since this is contrib module I didnt run the other tests + +Looks like we could give a try +Rishi Awdesh Thanks for patch. +This looks more like an improvement than a bug so changing accordingly +I updated the topic with Armys lastest comments and have attached the patch and updated the html file +Will do +Youll probably need to modify the +How this can be verified now +Also we copyfollow fails in this directory it is our DLLWe create follow system variablesPath dworktempxercesbinLDLIBRARYPATH DWorktempxerceslibBut mistake not disappear +The question is whether should make anassumption that a redirect to the same URL will always result in a infiniteloop +The corresponding functional test passes now too +Im running a remote test against it now it should submit to the branch shortlyI had been curious about with existing data sets Srinivas points out that this isnt an issue since the result set will never be empty in the case that there is data and so the existing data will be read +The does not seem to do the job since the listener class is sought after in +As far as I can tell it looks good +Patch applied thank you. +Agreed this issue is not about activation its about defining a System property +We have some maven targets depending on that convention +Sorry for the delay with this Ive just not had the time to take another look at it +Didnt set the layout of the separators correctly so they were defaulting to pixels height. +Cache has to be REPLSYNC for this to apply as Hibernate will reject using INVALIDATION for the timestamps cache +I would reduce it by pixels which miens also reducing the Stanbol logo +The proposed patch +Duplicate of +But if weve already frustrated rolling restarts then do people care? Same patch trivially applies to branch +Assuming that the is going to be changed into a bytebuffer to hide from the AM the duplication of fields will be necessary +resolved this issue. +Thanks +This needs to be documentated in release note +The code added is very generic and I guess it would make sense to refactor it to the service superclass so it can be used elsewhere too +i return but with a different message +Will do +The checkLimit flag makes sense to me except Id prefer force or if youd like it flipped enforce or strict +Fixed in SVN revision . +Gunther this looks good +I logged a separate issue for this +Local and remote replicant maps now use concurrent classes and no longer synchronize in a problematic fashion. +We need to be careful about the size here +For example the class schedule for the school might indicate that the first class period of the day started at AM and ended at AM +No it doesnt overlap +Patch which sets the additivity to false for loggers which have appenders explicitly have applied the patch At revision Thanks Chetan. +Ive added a couple of logging statements to the pop method so we can better visualize what configuration is being installed when we pop the current config. +The workaround is not to use the thread pools created by the AS but make use its own pools +Tim I noticed that you changed the fixed in field so this isnt an problem and you can fix it in JBM + would help reduce a data copy on the client +Agree with Emmanuel so much glue and troubles for masking few of the complexity of a such advanced feature +See my last comment. +bq +Willem thanks a lot yeah would appreciate if CXF did better validation of this and reported a better may likewise put in a wrong endpoint name as well +Not sure its really a step backwards if tests pass without the test close directory hackcouldnt we go with this as an iterative improvement currently no directories are ever getting closed right +Closing. +However if the inverse transform is in geotools then this is irrelevant +Attached a patch +Here is a patch that should fix the bug +The two issues should be coordinated +Reference documentation Transaction chapter. +Im going to unassign myself from this for now and set it the assignee to Automatic +KeithDo we have a good sample for showing off our new codegen support? Can you please take this upthanks DimsWe do not have a sample at the moment +I recreated the index for that month those two messages are now presentVery oddPlease keep an eye out for more wonkiness if you see it +Fixed in the revision Error reporting improvements when an instance creation failed +where in this case the is the servlets response +Caching the TM in a static in breaks that +OK got the attached patch I modified classes and adding the capability of using a percentage value in the autowarmCount parameter +Manually already updated. +something I couldve said +Thanks Bhathiya for reporting this issue and fixing + +Ram could you please take a look +Both issues appear to be fixed in +So how about thisAny further remarks? If nobody objects loudly Ill fix javadocs and commit thepatch within next hours +The issue mentioned in still exists for runtime should I keep the scope of that jira minimal so we fix this issue alone or do we want to track the others as well on +synced with trunk to fix the with for HDFS trunk +Thanks +On interest I might add that one too +Ok I really suck! Here is the final patch. +The pattern package is weird +refactored patch that includes fix from finish the patch now +Hadoop work still needed but thats elsewhere +Would appreciate if you could run this patch with your debugging to verify that we get all rows rather than skipping the remaining A real checkin should include a new test case probably based on the recent case that reproduces the problem +Itll take me a bit to be able to digest it all but I think that I might be able to use it +It is initialized as the nd constructor argument +I dont think theres a regex to distinguish between folder slashes and a slash in a filename is there? For example if we have this path in livelink EnterprisefolderRRSuritext the file name is actually Suritext but the path metadata just returns text as the filename +Verified on CR. +For now it can properly inform the users about its limits and later try to expand its scope properly rather than allowing something at syntax level and then not handling its storage internally. +This is highly trivial +Verified in. +Available in latest snapshot. +We have many many tests that have large numbers of chunks they all work now and did inPerhaps if you could send us the response msg we can see more clearly what the issue is +Thats why it is mappedBy it is not managed by the ORM +is the issue +This is what determines how the plugin is built as part of the fully UIMA SDK distribution +I think that would be really simple to do +It compiles fine against Eclipse +A couple others for MPING which would want to use a different addressport from any channel in the same JVM a transition from to can be done within the AS itself +Having a the code changed to make things work for me is not a prove there is no problem with modeshape +It worked! Thanks! Keep doing this amazing work +The endpoint wsdl is not packaged correctly +I think it is just an example of a broader problem sharing properties between plugins +Should the go back toLooking at the rest of the diff +This issue has been resolved in the recent SVN version I just tried it +Thanks a lotRegadsUrugn. +Disk IO in unit tests seems wrong I always try and avoid it wherever possible! But Im probably just things Jim your suggestion might be the way forward for a first stab at this +al +Definitely not a replacement for it +Since submitted a patch Im closing this as WONT FIX +RTC work item defer minortrivial issues to +StackI found one problem with the patchThe return type has to be added in the map in otherwise IPC error is thrown saying unexpected codeIs there any where else we need to change +Errors generated during semantic analysis and execution touch different bits of code and hence it makes sense to separate those patches as they address different issues +Thanks for making these improvements! +I am a bit puzzled why I cant get the functions executed and namespace matching as well +I am not going to this because I dont think that is good for the community but it will mean that Yahoo! is not going to be able to run on stock for a while +Best regards George +causing the host process to die would be very bad +attach a new patch to stop putting entries to cache until old entries are collected + adds FSIMAGENAMEOPTIMIZATION to RESERVEDREL and a test +Keith can you review pleaseThanksRobbie +See r +for the patch +patch to fix this related to the test case reported in the bug report +If we get rid of recursive deletion then well still have a method somewhere that implements recursive deletion and folks will still call it and things will still get unintentionally deleted + +I have a fix for this +I opened to refactor the RPC implementations to a new can you please run before committing the changes +For example changing the jenkins job name that builds the project changes which jars are missing aspects +In our use case we did not want the split to happen when there are not store filesI think ok to succeed the split may be make this as configuration and decide the behaviour + +From the information I have available following quickstarts target Web Framework Kit Ill be glad to help with this +Can you elaborate on that? The look and feel consists of css files images resourcesI dont really see how this relates to in bundles. +Do we need a patch packages and what name that would be +Even will still be targeted at the EAP community version +We need r for Harmony +Thanks Oleg committed in rev +Thanks for the review +Do you think v would be any different? My jstacking was showing that the time is just spent computing costs +The API were using says it tries to avoid doing that when possible but I guess for large blocks heap fragmentation is quite likely and wed run into that issueIm going on vacation this next week so Im unlikely to work on it but if you have a moment to upload your implementation for comparison Ill be back on this the week after next +Couple of needed tweaks and a test for a customOne more thing here +A subsequent reader will never read data form DNThis is not a bug but the situation could be improvedenhanced by somehow avoiding using stale RPC proxies. +Later A is restarted A still has hinted data for B but after restart its gossiper knows nothing about B +Looks goodThanks Tom Ill take a look at the patch +I attached two projectsApp needs App +If the first request failed theres no connection for it to be holding onto + +Fixed in +example wsdl for this Do you consider this as a blockerThanks the issue with revision +I will provide that +the cdh reference is a typo yes please remove it before committing it +to new committed +A migration script removes comments associated to a deleted user and removes authorassignee on a review if this is a deleted user +We also used to include two versions of guava which I think is unnecessary for the stuff were using it seems is compatible with r +Have you tried with IE too +Existing tests passResultsTests run Failures Errors Skipped +In case of RAID it will work perfectly fine with the change log coming at the time of edits roll +Updated patch to address Daryns comments +At the other end of the spectrum there will be users who do care about Reliable Messaging and who do want the guarantee of surviving a server crashrestart +Thanks Johan! +Its not worth letting someone generate complete garbage answers by changing that config value on a CF that already has data on disk +I plan to commit this soon if nobody objects. +Dont worry about removing the duplicates its not that importantIt would be best to delete unused labels in a separate issue +is a completely Jackrabbit independent library that people might already have in their classpath whereas the SPI will change more often if we need to adapt it for Jackrabbits internal purposes +Any ideasId invite you to try generating the code with the wsdl and xsd I attached to see if you can get it to work without modifying the emitted files. +Created the branch and committed this on the branch +Attachment has been added with description Fix for this to head thanks! +The comments in the other report and the latest one here indicate this has been fixed post +Maybe also as I might do that as it is for a long time in Forge and Forge Alpha +won the bet +Approved for +As the comment says since this is a boolean clause I had heard that there were optimizations for this case +closing this issue. +I hope that all requirements are ok nowRegards +OK let me give a shot +Attaching patch for Amir who is currently out +There are still a few outstanding jars that is pending approval from for uploading rd parties +using some search integration with SolrES etc +The wsdl generated with the versino of axis +moving this issue to CR where the remaining issues will be descriptions are correct +due to replace type with the one +Made obsolete by release of GML library +Data file does not match index even a little +Ive committed this +Hey ChrisShould be ready to do that soon just waiting for Apache to approve my mentor connection +The svn move of to is marking all line of +Bulk close for +Yes I had thought of that previously but this would only work when OBR was installed so that it could register its URL handler +That should not cause problems in the upper layers you just cant combine socks proxies and actual local addresses that waycheers Roland +The patch in fixes this issue as well +Integrated to TRUNKThanks for the patch EnisThanks for the review Stack. +I hope you didnt include the linked license to since! Jamon +We want a new connection string for each +Things should now work fine with failover +Attached logs for both cases where autoscale creation succeeded and it failedI see the above exception mentioned in both cases so that doesnt seem to be the cause +All our identifiers are restricted to not start with so that we can handle the character set names +The outdated method which was causing this issue has been removed. +Duplicate of +Christian please ping me to discuss. +but of the time they will use the war as a self contained unit anyway +log to a script and I am not able to do so +so sorry I am confused +We do have such a control. +This is the actual is the view I course the wish view has been simplified and is more complex in reality +core tests + +Changes in this newer version of patch info about checkout to be in one place way in the prerequisites replaced tabs with spaces for the sake of convenience Ive merged your latest patches to get the complete picture of changes +The release uses a different version so the existing ones were probably good enough. +But here yeah project as should actually be fine +I have applied the patches successfully and Im running the tests +I have checked and CP both exhibit the same behaviour as +Solved in commits and tested on Windows XP and +applied to r +Sequenceids for edit files need to be different because sequenceid is the key in a map of all a Stores edit filesThere is already an issue to deal with the case where there is a sequenceid clash unlikely but a possibilityWhats going on here is that youve probably fixed something that was broken +Lowered the priority on this since the workaround is actually better code +Defer all open issues to +Now it will be one +So for this patch +I am not convinced that is the right way to name it because the class has separate constructor arguments and it does not use a constructor argument +Patch looks good +Includes some small code readability improvements much along the same lines that Ive done for the in +How many tests in each Alex? Sounds fine though? You dont want to put together? Ditto for distributed stuff +Fix looks reasonable to me +Bulk cleanse of EMB issues moved to legacy version +the new one has this well encapsulated so it was on too +I also often have problems with eclipseeclipse +it would be hacky + for the patch +May I be so bold as to move to? Theres no movement since we began the call to converge on and its not obviously on the cusp of resolution +I tried running the test case without the patch applied and the test case succeeded +In addition this patch removes comments of unit tests including minus constant valuesI verified mvn clean install +It basically makes the members public and adds a bulkPostings member for reuseIs this the right approach +For that case we indeed use a fallback strategy to resolve standard placeholders as system properties which only kicks in for resource locations that havent been processed otherwise +compability with JSF implementation +Framework has hit the streets so I updated our launchpadbase module to refer to it officially in Rev +push out to +See my comment on for the resolution. +Tested it a bit +The PB stuff would be better left to you I imagine as are any additional tests you feel are requiredSounds good +Hi Deanthank you for the ticket number Ill update the plugin when the new version will be availableNo problem no inconvenience at all +Have we dropped any info that you know of +Thank you for committing Bryan! The fix is in the trunk so Im closing the issue. +The files inside the searchstrategy package dont have the AL header please add itAnd somehow the schema is missing in the Ide Plugin? Or I dont see it +These are the remaining EXPORT topics that were updated +I am postponing this until somebody needs that functionality +Patch committed in r +I tried changed encoding of but what ever I put there I wasnt able get installation work +We model data representation using JAXB bindings and leverage Jerseys support for wiring JAXB to JSON +seems done +Not sure how to handle this cos users can still configure REPEATABLEREAD in the server +Feel free to give it a under Spring and appears at the early stages to look good +I see this was changed in +Created a module which can have user functionsAdded methods to create for files at a location and for a +For release moved the duplicate classes to the resteasy module out of the examples +Well get started on setting up a local svn mirror +make sense to me. +But to be honest ATM Im quite astonished +For this interface is merged intoCase will cause the bundle to be rendered in the filter specified by the first of the items rendered +Attaching a simple patch for the same +Choosing default value of write formatter based on file type is covered by. +True +The expected number of responses is set based on the number of live nodes identified +Please open a JIRA for this problem thanks for reporting this +Closing out the issus for the release. +I committed this +Please could you answer to my issue? Christophe I can confirm this is a bug +I just committed this +I am experiencing the same issue Spring and Hibernate proxies both throw object is not an instance of declaring class +The improved concurrency is typically worth the extra difficulty of checking for nulls though +Attachment has been added with description Output from the fop processor +Ill put it in for since Im guessing thats the are pretty much stable point. +Actually no ignore that last comment +bq +I tried using the and everything is working +Please confirm the current status of this issue and if its incorrect change it as required +Iusrlocalaxiscinclude Iusrincludelibxml o buildDebug mkdir p distDebuggcc ldl Wlsoname Wl shared o distDebug buildDebug Lusrlocalaxisclib L +Bulk closing stale resolved issues +It looked like this was happening in the apache cli stuff +Test script that exposed the error of have been addressed with the display node refactoring +I think we can close this now as of the latest snapshotIve pumped in about billion messageIve got files open now so its not increasing at all + +And a patch for trunk +Moving something to the trash is not deleting it its moving it +Thursday CET good for everyone +Added tests to Sql closed as resolved and released +Thanks for the patch +As well as doing the procedure described here to make it secure +Looks like a recent change in the Inactivity Monitor was blocking exceptions form the transports to get forwarded you waited long enough eventually the bridge was removed due to an inactivity timeout +I guess we will have to do some transformations for backward compatibility +Step instructs you to update the project so that it points to the shared copy of +Build from the root to create an EJB with the wrong manifest entries +copyUtils in hadoop throws if the copying file disappears after the list status is done on the directory +Alessio assigning to you have a list of the classes involved so that I can review the changes before the merge +Actually the has the same problemso I readjust the codeplease see the attachment +It was row cache that cached rows that were REALLY big +You need to uncomment the adminadmin in confprops comments on +we should clarify our plans and inform the community about feature has been implemented in scope of build time minimization for JSCSS resourcesIt is necessary additional efforts to make to generate compresseduncompressed resources and packages of resources and be able to switch between them in runtime +I will have a look later +Tere is no such problems for example in VPE tests +Wouldnt you have to send the full list of facet terms consolidate them and then loop to get the distinct number? That is why I originally sent the WHOLE list of facets and just added the magic number to the end +By the way how you would differentiate the case where someone really wants to put a as key in a map +Yes it fixed it +Thanks for your hints! Unfortunaltely neither solved the problem +I committed the patch to trunk and +I probably wont have any time to work on the backport though +This way we dont need new command typesI like this idea a lot +But it seems like is being used in a LOT of places anyway so I dont think we mind leaking those semantics in codeIf I was to use then Id have to create a collection +I think that is undesirable +See SVN rev +I did not modify the examples index page +As such it has a bunch of bookkeeping data associated with it which is always loaded into memory unlike field values which are mostly kept to disk +the trunk change looks good +I backported the test case from and sure enough it fails. +I think it would give the capabilities that I was thinking about +because a rd party always runs in addition to the defaultThe class is a config so it is the same for ALL jobs meaning the will use always the same class +Add the implementation and tests + please test with trunk I believe this is fixed +This issue occurs when using shutdown or deploy command with secure and encrypt with configured that there has no other way to add gbean to a global moduleThus I add code to crypto package and involve it inTo fix this issue I also add a system property to support defining a keyfile set this to JAVAOPTS in system variable +war +we dont have any data structure providing the metadata needed to recognize a serial column other than the primary would it be possible to work around this problem using a trigger in the db + +The hard part is Im up against a deadline The bigger issue is I have my own input prep all together and even though the Job is built in theory to handle starting at arbitrary phases it assumes certain things are in specific places +Well that would of course solve the problem +I think that is +Most likely a result of +You just specify required bits +Can someone confirm that +It may not be a global solution but creating a page which allows the user to set the session time zone may be a waste considering the new console is underway +Ill close the ticket. +Version two worked on it w Andrew sitting beside me +OK it looks like the previous test method set up the master to only replicate on startup and hence that would seem to be the reason why the wrong index version was being replicated +the difference between client big deal to us and priority for the vendor +A possible solution would be to always start a transaction in PFER even if autoCommit is false +I might port this to the main branch in future of if someone else is interested please ping me +minor revision from previous in no point in with in now +This seems to be the intent based on the existing usage of get +Looking at this again I cant see the problem that this is trying to solve +Other toolbars and perspectives are also affected +John can you provide more context for this request +Thanks Brandon! Ill commit in a few days if no one objects +HiOk that explains why I was not able to reproduce it +The thought is to have the capability for tests to switch onoff a set of faults +Link to issue for incorporating +Patch committed to with the minor change made in consultation with Thejas returns true for GENERICWRITABLECOMPARABLE and returns false for it +Thanks I just fixed it before I saw this +Alright +we ran into this issue on our platform +Moving back to so Lars sees it +Peter Can you try out the latest master and let me know if this solve your issue +can you paste running mvn e +The compile failed because wasnt throwing an correctly +Committed by dreiss. +Please upload your complete wsdlthanks bug is now fixed in as fixedCan you please add the diff for this fixThanks +Attached a simple SSL test +They do it because that is the only way +Committed to trunk. +Thanks for the project Peter +I may have fooled myself will tomorrow morning +Still need to work out some of the details of this though +Can you see the problem here? Maven will amount to nothing if you dont get out of the way and let us upload libraries to the official repository +Reopening this as per sureshs comments +will mark resolved once I verify its fixed in new since fixes are in released code +No great effort was required in creating this patch if it goes stale it would be easy to reconstruct +dims +patch depends on youve not updated as part of this patch and that fails because now takes a instead of a ConfigurationCould you please fixupdateThanks +Please verify that images work properly in PDF generation and if necssary +Im working on it right now it should be ready by next PSC meeting +Im flagging patch available to solicit some review from folks +People wanting to trace history will have to use the cherry picked from commit +Mass closing all issues that have been in resolved state for more than a month without further comments +Ill look at this later today or tomorrow +Im sorry to come in lateIm a little confused by this approach so please bear with meIt seems to me that should actually go away once we goto cgroups etc +Maybe somehow it logs to another file somewhereI had a look around on google and it does seem like there are cases where people had to change their logging levels in order to see the thread dumps +Is generateSources just a task which depends on the tasks for these different source generations +If you need Japanese translations you can assign the ticket to Shinichiro Abe and maybe he can do that +It blocks our development +Hello World App which will reproduce issue +now passes on my host +May it be possible to issue a warning or error to the log that spring wont be able to set the value because its declared final +bulk defer of unresolved bugs to +My concern is how we implement it +The patch changes to where applicable and updates the golden test filed +bq +Turns out what I understood as jar overrides was wrong +Ok created to proceed +This is a picture of the memory for all of the owned objects for one of our library files +These items are still not resolved for and we are approaching RC soon +The fifth backspace should delete the whitespace before where when was but will remove an instead +changed to relect patch status per Andreas your patch is applied to the HEAD branch only +Also an attempt to replace calls to in other areas of the code bring up the same issues and should NOT be made +Closing all resolved tickets from or older +Couldnt reproduce the failure on +I guess you guys are the ones to fix this problem but I really think that beyond clarifying this section the rest of the change for this jira is minor pri +It will have to be debugged on the source code level +Am I wrong +I ran this test with client and embedded framework successfully using Sun jdk on Windows XP +let me split some more parts +my exception shouldnt occour anyway +ui +Those jvms running on slave nodes will down +cannot reproduce and judging by the date it was before I became more parsimonious with WTP +Done by shane +Thanks Uwe! Keeps testing +Thought the real fix would be to make Crop rotation aware +The problem with these is that they take a individual log file rather than a whole log instanceCan we change this boss and pass in an or whatever instancebq +Maybe better log messaging that these classes are being inserted into the cp would help +I am using JBDS as you can see in my first comment you have to have archives view closed when replicating it +a warning seems like the right way to go to meso do you mean logging a warning and do not do cleanup +My simple attempts to script that problem are not succeeding +Now the derived pk can never be overridden +committed this. +this patch adds background to +so if democraty decided +A squashed version for the above patches +Wish I could change the priority to Major +Attachment has been added with description Sample FO file showing the P open bugs to P pending further longer reproducible in trunk according to comment +during compaction we echo a row if its only in one of the sstable were compacting but it also exists in a sstable we are not compactingIll buy that +This time error has different number but stack trace is the same +It sounds right that often enough would throw an exception and perhaps this is the behavior a typical java developer would expect +right? Are using nodetool h localhost +I traced back to it trying to resolve the address +NPE occured when open VPE close eclipse open eclipse assert VPE should be NPE on. +Attached an additional source file which shows the problem syntax more move of all unresolved issues from to +I think v is missing the new Callback classes +Okay Ill revert both values to false in +Besides this is a operation it does not affect any speed concern is not speed but +This patch needs to be applied from the trunkapplication directory +Resolve this issue as invalid +Shelly thanks for the detailed description of the we have to decide how to solve this +The context here is that Im trying to pass along a delegation token in a http request +committed thanks! +disable method check for configuration might be in two places +Committed revision +As for my opinion anyone that wants to perform fixes in that module is great +The code has been modified to handle the vast majority of the cases +Ive made the changes I proposed earlier +Just added the entry to reopen this issue if the mail isnt showing up for some reason. +I will try to get this done within the next dayRegards tom +Reduced the number of database ON DELETE cascades and more manual querydeletion of dependencies index should be fine now. +RamkrishnaIf you have access to a real cluster can you do the following +Thanks Koichiroo +thanks. +This should do approximately what you want please check it out +Are you submitting a patch for trunk too +reopen for clarificationJBDS must not link to so just reopening for clarification that htis change just affects JBT +Ive put your comment in +this means that people will no longer be able to just drop a file into and place and have it take effect immediately +Create a test case and make sure it does not pop up in +this needs to go to branch as well +So all tests would fail +How is this going +Removed HADOOPHOME and HADOOPPREFIX for branch +The methods have been left untouched +Perhaps for Spring something can be done to make a property placeholder type of mechanism more of a first class citizen with the? I think the concept of a property placeholder has been proven useful and having no way to do a placeholder replacement of an XML element can lead to some confusion and frustration to users as become more and more prevalent in +JonathanHave you still got a copy of the old makefile that ran the qa tests I wonder if theres an option in there to run jtregLooks like jtreg is called with however it can be done with ant also +Also makes it a bit more efficient and I think easier to understand Fixes an issue in when the lastcheckpointed value is the same as the high watermark ie no more messages have been written to that topic since checkpointing and there was no head to pick from +Moved back to +Changes in has broken the interpreter handling of monitors +Then we will see later if someone come with a better answer +most of the contents of xdocletxdoclet and xdocletxjavadoc? +This was a legacy issue ported from Google Code and appears to be fixed now + fixed +Or better yet Erics approach of removing the config entirely and fixing Demux to not use that value should be followed +pushed +I pointed out on the forums that releaseResources should have been catching other exceptions as it was intended to do and that addressing this would remove any temporary inconsistency in the we follow up there if something wasnt clear +Assigning back to Marek as requested +Currently I commented that piece of code +Svn patch format this time +will issue new pull +Patch committed. +I cant reproduce this using the code provided +The does everything almost exactly right and its not an easily extensible class +This is the fix for on +The current deploy target still maps to deployment +To handle deletes all columns in the view cf should be tagged with the columnkey from the base cf that caused their creationAt read time the base cf columnkeys that are tagging the columns in the view need to be queried and the view function +Please let me know if it should be something other than Enhancement. +Im not able to reproduce this locally +I just committed this +In some cases you still get this error +Done +Attached patch with regression test for +Can you ask Emanuel about it +However it would be nice to at least log these at a trace level +fixed in +I committed the version of the patch +I would commit this patch in a while and get the trunk build going +Can we have the JIRA project as well pleaseThanks done. +I suspect that we actually have a higher level vulnerability here if the method isnt protected either +Thanks for the help here LukeIs it safe to say this issue is resolved then +Lets go to apply the patch and launch a vote Julien +Like Lukas I cant reproduce with current EAP it was handled as described and closing +When I go through the other +Does the hibernate team agree that this is worth fixing +please try and let us know if u want us to reopen the bugthanksdims +makes sense to push the functionaly back to the MIME framework and also to target the test cases directly there instead of testing with +The main problem is that its really difficult to know how long to wait until the command is available due to the very asynchronous nature of osgi +so whats the verdict here +Apply the patch seems to not delete the old and create the new +Resolving as Invalid +New technical debt is now well computed when only technical debt of issue has changed +The client wont attempt kerberos if the server doesnt support it. +committed this +This is the latest compatiblestable version +Lame +Ill ammend this patch to add an Ignore and well do a follow on change to remove the Ignore and modify to run the common native build before running the Hdfs tests +This is only a workaround until the missing functions will be added +Namely the docs say that the field value can be null but it is trying to construct a Term which requires a field nameJust give it the name word collate you mean that the SCRH would not only return suggestionscorrections for individual token but it would also try to glue together an already corrected query string based on its suggestionsExampleQuery cogito ega sumSCRH returns this correctionerga ergoBut also tries to give you the whole thing correctedcogito ergo sumThat? Sounds useful less work for the client app should the app developers decide that SCRHs collated suggestions are what they would have to do themselves anyway +looks good +Yes +defer all issues to +Please add any more changes to the Span class documentation under this JIRA or close if everyone believes the changes are adiquate +Includes statistics logging for and +These patches all apply to trunk as Steve Rowe noted +Apache Ant version compiled on June even tried the proxy settings in +Committed to trunk and +I will try to come up with a version for +yeah just set the cluster name to local +Until I find time for the wikiThe mechanism does not just depend on the host per se all the following are different to even though they point to the same system httpfoobarlocalhostcitations httplocalhostcitations RandallWhile youre at it doing for databases do you think it would be possible to do a UUID for a server as well? It could be exposed with the version +Oops my mistake +for have committed this +Its really kind of silly to have all these different tiny topics when what you really do want is just the bullet list in the Derby and standards topic +Using a plugin makes it more consistent and slightly easier to configure but I would not regard it as a necessity in this case +I think Simon stats involved a synchronized section +This implementation is a fairly big hack thus far +not call the statereadclientrequestheader directly or else the httpsm maybe deleted before the function return +As Alexander wrote i give him partly right +It should address this issue +I wonder why this issue is open since and still not resolved or even in progressI mean it should be simple to fix it right? Creating a function that would be specific plus its kind of an important featureWhats the reason for the low priorityThank you +The current proposal is that even when is equal to false setOwner and setPermission should be permission checked +I created a simple gbean for this but ideally it may be better of to be integrated with one of the existing deployers +If all items go through without raising any exceptions then on will not be called +Thanks for the review DDLet me fix the failure and commit +Especially if we want to keep support for paoding as this project is no longer maintainedWe might also need to have a look at the Contenthub as it also heavily uses Solr but AFAIK is missing unit tests that check compatibilityI will definitely help with that as I would love to use the compressed indexes supported in +From the add and to the exported packages in +Attaching version of the release notes which includes the detailed note for +Or better yet if this capability exists and I missed it please someone point it outTIA as a simple patch for the Shekhar and Ron have reported +Well commit them to if will be fixed before release +the current contains varwww as directory and varwwwpootle as directory with different statements that is in my opinion not clean but rather confusing +Looking at the png images created by the further confirms this problem +Commited fix for XML test as part of revision with following commit commentsBefore making call to XMLSERIALIZE we should check if the classpath includes the required XalanJAXP classes +I will run tests +There is no need for a factory object expressions are treated as citizens +in +Nicholas some minor nits Set the given directory path to a snapshottable directory Set the given directory as a snapshottable directory would be more clearer I think +Is this supported in hadoop? As in can you specify the input to be on a different hdfs and get a mapred job to work? IIRC no but I could be missing somethingIf it is no then not sure if pig can support it without an intermediate distcp +thanks Ivan. +If one of the committers opens this issue Ill assign it to myself and check it in when Im is great! How are you handling the distinction between stories and tasks? Make sure you maintain backward compatibility with a story only you are doing this could you take a look at doing at the same timeWhat version are you basing your implementation on? branch or TRUNKThanks again +Thanks very much for the contribution! Pushed to master. +I added a test in in the issue isnt with the in r r of branch r of trunk +So the client still needs to be able to skip unreachable endpoints itself so describealive seems like gratuitous complexityI agree since the view is from the coordinator describealivenodes isnt very helpful and also has to wait on the failure detector to mark the node down anyway +So overall it has always been faster +A modified version of with bits of code pasted in from to handle the return type +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be and above it seems that the setting in is ignored when running grails upgrade +A orders Interceptors according to the phase they participate in and also according to the before after properties on an Interceptor +So if you have a service list to resolve id in your components you need manage it in database +Deleting in finally could be good but I think that delete of all the and at the beginning of the test might be even better +We have to rely on TCK and regression testing for EAPChecked only by the verification of presence of this code in EAP sources. + +Verified at +is fully fixed with this patch? If yes feel free to resolve it as duplicate + +Good idea Ill add that thanks! +That would not reflect well on Maven +We then ask the detector what it is based on thisFor Zip we look at each entry in the zip file in turn +However in the switch to block compression that doesnt make sense +Hi Sunitha thanks for the updatesI agree with your analysis and the conclusions you have reached so far +attach a straightforward patch to move common code to a inner class and partition the requests by topics into different delivery threads +Having a single causes the IPC module to fail with Cant redefine +thanks a lot for the test +Looks good +Added a test case +I hope to have some time tonight to test the fix +I dont think it makes sense for which node is closer to depend on what keyspace youre talking about +Hi thanks for quick be the problem is really the version cause Im using hadoop +Id be willing to offer my services on implementing this if youre interested +That way the source tree isnt cluttered with test output files. +The proposed workaround doesnt work for me +we should get rid of them and change out README to say run autoconf and then do +Parking in the next CP cycle as this one has already been released and will be taken from the was resolved in the repository and in the online documentation and the remaining errors in the packaged docs called out in the release notes + Ryan one thing to try is changing the to use the XSD instead of the DTDno dice +Any more thoughts on this or should I commit the patch? think it is good to go to and HEADMike +I will add fix patch and unit test with example soon +But for the sake of getting release going I guess we can move it to common there for now +Therefore I think we do not need to fix the bug +How about running two versions of Derby in two classloaders in the same VM +were backported to as well adding the latter fix version. +You are saying that property list are totally ignored from the underlying connection +A remove would delete the +Each item has properties but the truth is that on this items we are applying rendering rules. +Actually I take back that I am able to reproduce this +Im getting myself confused +Can you explain why the situation is different for HA +Until then how about a work around set of instructions that preserves the Roo round trip +Any updates on the Boost issue +Better late than never fixed on trunk see rThanks for your contribution +Seems like a dup of. +Thank youHibernate with OracleWe have the same problem Hibernate should indeed delete orphans before trying to update them +I would expect RVMs environment variables if set before invoking Maven should carry through just fine +Anyone working on this? Would like to work on this +I still think the setArray method should go away +Thanks +The original bug report is invalid. +If I cant get a fix soon I will disable test on +Thanks. +If you analyze a heap dump in MAT its not very helpful most of the heap will be used by other and tracing the GC roots of those objects is fruitless +Notice if you never flush the table the query works as you beat me to it +I felt it was a little strange to specify param on the grunt shell but it is easier to remember if your using it outside the shell alreadySo this patch does the same as the last one but the syntax isrun param LIMIT param FILEfoo paramfile tested patch and all is good +The patch was applied almost as is except for some tests that were failing. +Reformatted the code fixed some CS issues remove the author as we dont use it in camel +Yes +With the first one at least it point to me Jira issue where I can read the information easily +Hibernate on master and branch. +Give me a week or two +ve made gfx corrs +HiWhat is the status of this bug? It is still present in struts which I believe was released a year after the patch put forward here +Do we needwant one +How do you guys handle eclipses +The package uses conflict might have been caused by having two versions of xmlsec installed or if one was removed and its dependencies were not refreshed +Yes its that one +Ok sounds great just give us the new issue number when itll be doneIm agree to test new snapshot if you needThanks +ArthurDo you still get this error? i cant recreate this problem in current buildtest harnessthanks seems this problem is solved in the latest code base +The SVN plugin for svn is badly broken and running it against our rev repo crashes JIRA +Tomcat on both Windows is running u +Tested with and it works so please check your Spring Security configuration +Patch committed +The test cases from should be migrated to +Attached patch does what is described in the descriptionI tried to be super explicit in the javadoc about exactly what is happening including adding more info to the existing methodAlso one minor change to existing method +Confirmed to fail inThe issue could be related to +the reason for both problems is the same +Fixed in r +We are able to get test fail consistently on some machines but test is passingOops I missed this +No tests because all the removed methods are private and dont affect any testsRan a sample pi job +Also attaching a utility which I had to use to translate the output into the default encoding +Hmm according to SF the last release of jline was and git has not been touched in almost a year +Thank you for fixing it +The patch with updated marker can be used as the more complete fix +I am sending to agent for large scale deployment to avoid excessive local disk usage and tcp incast problem +Do we have results using prefix with fast diff algorithm for the current hfile vjerry i tried all three delta +its different if something isnt possible just because it wasnt thought of or implemented but here were talking about a conscious decision to not allow something +So far Ive got VDB validation warnings for the following conditions Model file was deleted from workspace or VDB was imported into the workspace so no models exist yet WARNING The model file in the VDB does not exist in the workspace Model file was renamed in the workspace and a VDB was found containing the model with the same UUID WARNING The model file in the VDB has different name than model in workspace WARNING The model file in the VDB exists in your project but at at different location sources NOTE dont know if we need warnings here but not warnings are valid Model was moved in the workspace but VDB was found containing the model at an old location WARNING The model file Employees in the VDB exists in your project but at at different location +Fixed the whitespace changes sorry about thatI didnt want to add a dependency between the tests +We do not have VS +Changed the Ant Troubleshooting wiki page to describe this issue a little bit. +By Feb +Please attach a small testcase which demonstrates the problem since this shouldhave been very visible in our regression tests if it was happening undernormal operation +please close the issue and we hope that cygwin soon takes the fix back into their regular distribution thus the traffic for that FAQ entry hopefully soon levels off. +I took it out mostly to make sure there are no accidental references to the systemSession variable +This used to work in but started to fail since but it seems that QA never report this issue +Why should the numeric stepper allow for? IMHO it should only allow numbers to be entered is there a reason for allowing +Thanks Janet! +Have you found a reference in the spec that covers this? Generally we want to have as much common behavior as possible such that we can really have pluggable soap stacks +Hi Ben the patch is empty except for filenames +Good work Uma +Im attaching the document in question +Closing resolved issues. +That was a problem with which I patched in my JS file using the fix that is now part of +I entered a comment on that may be of interest here too +Many comparators have been added a few are still missing. +just committed this +Yes this is important +People may want to return a Message or Exchange if they want to unmarshal add headers etc +Id rather avoid putting the digest utility code in as its not directly related to JCR +All unit tests passed +For a obsolete core method that should no longer be called by compatible clients incrementing the major would be appropriate +Scott are you going to be able to complete this as a Maven conversion of the Java build or should I tackle it with Ivy Ant +The FST approach given that is way more low level will give us more control of the functionality down the road which definitely will prove benefitial +and me all come from Alibaba +Ivy does log in debug in the Ivy Console doesnt it +Your site needs some adjustment the content needs to residewithin a trunkcontentctakes folder to meet the requirementsfor an incubator site +Anyone else have a chance to try out and review the patch for this issue +Closing +For example I got a ruling for the JSON license which is a derivative of the Berkeley licenseHope this helps +Thats not simply enabling protobuf RPC +These info is extracted from the manifest in each module +Delivered my validation tweaks today +Patch. +I thought it was missing because I read the two articles listed above so I went looking in the tools +Fixed title to match problem text note that it is possible to add the directive in the BND instructions you would need to explicitly do this when you are combining classes with the same package from different locations +Thanks for the reviews Daryn +Patch test +Also Im not sure if openejb is using the geronimo builder this change may not affect openejb +Aliok as you can rely on Google App Engine can you make it a constantI will check in the Trinidad util later +patch looks have committed this +I am rejecting the ESB issue as this is definitely a bug in SOA issue will be updated accordingly. +ok still dont know whats going on since I cant reproduce it here I have more questions does this only happen for instances? can you retrieve specific images realms hardware profiles etc? are you able to stopreboot an instance? patch should resolve the issueLeander andor Christian are you able to apply this patch to test if it resolved the issuethanks patch worked +The seems quite dummy which is introduced for only the testing purposeA side note and collect bytesRead but not bytesWritten +We have that in mind +I tried to remove the Controller annotation after provoking the slow content assist +Note the prices of the components show in the journal but the price of the components are included in the aggregated item line +We cant use the package exclusion method because the packages in both of the source folders are named the same to reflect their affiliation to each other +The code is in master branch but still have bugs to deal with +And no questions asked +This test case demonstrates the problem with last nodes on the rack at the getLeaf level +in JBDS or later timeframe +Closed. +Yeah IIRC Ive used iteration instead of a HEAD request because not all of the attributes are returned when issuing a HEAD requestI dont remember all the details anymore but if createdtime is the only attribute which is missing them Im fine with switching to HEAD method because its way more efficient +Nice patch +I cannot recreate exactly as I reported as the delay no longer occurs +addressed believe the javadoc is not related to this patch +Wow nice detective work! Thanks for posting this as it could help others in the future +Or at least thats what the words should mean Im looking at the scripts now though I very much doubt this is going to be five minutes work if that +It performs multiple rounds of indexing and calculates the percentage difference with fastest of each of the techniques +Since we dont there is no place to hang the tests you describe +Would you say that this issue is still applicable +I have tested the solution only with and Airavata Server modules. +enolan Updating the file to add in wsdlj +Thanx for the fix on the click event attributes +They share the block cache by the static global reference +Take a look at +It should arrive on developers classpath by being copied to the buildclasses directory +since the proxy happens at the class level Id want other references to those classes to do the lazy loading I wouldnt be concerned if the didnt do it and instead tried to fetch themThe problem is that its doing extra queries when lazy loading is on even when explicitly joining the other classes +Can you try attaching the patch as a file to the issue often there are encoding problems when doing a copypaste +Hi MukeshPlease provide some steps to reproduce the issueThanks ScottWe can safely close this issue this is not now this might be fixed at previous revisionsThanksMukesh +It makes me do rebase on upstream againFixed +will be uploading patch in some time +yup and snapshot for me now +As a result the is always being assigned to the regardless of the setting +If we accpt Strings we have to iterate it to convert to Options +I just committed this +Reopening until weve found the right tweak to EDGBASEedgeccpconfig. +Kabir it worked on EAP but it probably never worked on AS +See +It should read And wow you replied within minutes +I am just curious I am now wondering if a better class name is needed likeThanks +Also linking to exception during orderly shutdown of a sequence generator +looking good +Ive made most of the important methods fields protected in a few choice places to make it easier for people to plug in their own persistence strategies without having to everything from scratch +Fixed in trunk vGACP and vGACP +As I have commented this is a bug in the wiki page one that is valid + +Please fix! +This patch is for the branch +Custom serialization relies on preloading to assign globally unique id to each type so we need preloading anyway +As Cyril hasnt shown up for quite a while I remove the fixFor version of this ticket +svn areas are all setup including support for staging builds +As a developer I build standalone browserSim only with bootstrap profile and profile name doesnt matter at all +lgtm +add finderName find add finderName find add finderName find add finderName find class +Verified in Tools +Also minLength is not a valid member for Text member type and it will be ignored if used +has been configured +That will be handled in a separate jira +Fixed +Moreover as described in the comment does not need to be serializable as it will never be serialized +Patch is applied at +This is has no relation to the restart +Patch that clears the originalExpiration property when a message is and a unit test that fails with AMQ and passes once the patch is applied +Whats the status of JIRA +Hi Jake and updates on the current status of this issueI think I saw something on the mailing list regarding contribution andor IP clearance however its slipped my mindThanks in advance +Without getting into too much of an NDA area here lets just say the TCK makes two different requests to test this and that its unaware of our view sequence produced during rendering +I guess we could close this bug and clone it for B to ge proper validation before finishing the wizard +The DSL is the same for XML in springblueprint so this example does not bring much to the table. +Thanks Drew +I am fixing the new merged code infinite loop issue and make the example work +Very soon youll be able to do it yourself +In the local mode shuffle shouldnt be invoked at all +Now it is possible to start an editor for all attributes. +It tries to find the lowest priority request that has been assigned the last and releases its container +I mistookminus is using not x but xFFD by unicode +Here is a proposed patch for trunk Lincoln can you try that +Please check that and let me know if i am wrongThanks +The fix for seems to be the source of this bug +Juergen what do you think about this +The version we were running was prior to +this one actually has shield graphics in it +If you create a new database and run in an IJ session running on default values you will experience the error +This is happening because we dont use the qrttidynamiccast option +Obviously its not really tied to any of these versions of Spring Integration as it represents a much more general issue +setting lineage service to hive binary +See previous reviwed this patch +Attaching a patch to pass the unit testIt seems to be correct for me that the counter is not incrementedThe job is not displayed when executing mapred job list because the job has already completed +push out to +Anyway could you try to cook up a small example that illustrates the problem and attach it to the issue +I redid the request today with http instead of https thanks to Donalds note and the http request was successful +All tests pass +Please add a comment to the copymodified file that it comes with thrift and document the one method that was added as the only change +It cannot be trusted to last especially given the question marks that hang over Java on OS XIm for a variety of reasonsa were getting into platform specific code without a good reasonb this could have significant performance ramifications as we add more and more paths to support more and more c there is a config option to specifically eliminate the need to do m sure that would remain true if Apple was still going to supply the JVM +Provided the following changes +Ill take care of this later this week getting smashed by other projects at the momentIf only it were a one liner +Assigning this to you since youre already on this one +I was at the airport and it filled the w login screens +I agree that we can uncheck patch available +Libor Zoubek afaik this is your swt bot test suite +Build system update for Win is required for this issue +Have you tried running the attached example +You can see each exception and see which server it came from +Experience of past conversions is that git svn always ended up being my route +If any one need it please send mail to me +Youare right if I modify the homeomctmp to tmp it can get the files in the homeomctmp directorythe tmp diretory is a relative directory its absolute directory is homeomctmpIt seems that we only can use relative directory for ftp poller uriCan we use absolute directory? How to use +Thanks a lot Ivan for update on the patchclosed variable can be volatileOther than that patch looks great to me +If there is external documentation that would be of value please consider adding a patch or new ticket under the website component with it in markdown format so the information is not lost or forgotten about +I did not check the NDB cluster setup +committed thanks! +Attaching patchThanks the explicit state check is a little unmaintainable if we have new states in future the current change is less intrusive +Its opssupport nightmare when datanodes report to incorrect namenode and lose millions of blocks at once +We agree that it might be somewhat confusing but the purpose of the class is clearly documented in the reference guide and there is now also a brief statement about this in the javadoc for the is a way of executing any method by using the executeexecute methods in. +Q +Another used for this discussed in a Yarn meetup last year would be a way to get rid of the auxiliary shuffle service from the NM +This patch doesnt remove segments but we dont want to proliferate the use of segments +Ill see how well the current setup is working and then revisit this issue +Just dont work on this issue until we do have that +Attaching and documenting the new comment syntax +I missed the shufflemerger change since its a largish patchCan you pls create some natural Move out as an interface into a new class outside of Introduce a new or some such interface which is sufficient for your purposes Then mark all interfaces you need as I appreciate your patience thanks again +I corrected an error in the body of this JIRAPlease generate a patch file and attach to this issue +Reopened to set Release Notes Docs Status PR applied thanks +The java files and the profiler screenshot are posted already on forum +The plugin ignores the feature added with parent have created a new PR + to trunk +Closed after release. +And then on the side you see the details for the the screen shot you provided what is the Default Selector? What does it do? What is the Default Condition Ref? The Default Target Profile? The user needs more context into how to actually use these bits +Yonik do you have a specific test it doesnt work for? it spews a ton of INFO out for me +Ill release as soon as I get some breathing room in my schedule and my prolonged permissions issues worked out. +KoushikI have mentioned the use case that was tested in previous commentI had few vms deployed in both the hostsHad hourly snapshot scheduled for the ROOT volumes of the vmsWhen snapshots are in progress shutdown the master hostI think having the host shutdown when they are active tasks in progress may be causing this issue +And the maven metadata in the file wasGenerated by MavenThu Aug BST version I see the problem now +Also added a test case that verifies the bug is fixed +Look forward to your comments +We can always revert it if it causes problems to someone +CP branch upgraded to the CP +On construction of the the delay is set +what was the cause? changes in common navigator in or in our api +bulk defer of featurestaskswishes from to +Today I was able to build all repos master branch at once with Pmaximum +Let me do it on commit +duplicate to +Yes adoption is driven by forward thinking persons +Addressed code review comments +please confirm +Thanks Joe +I arrived here with the the exact same problem +Patch looks good +It should be more obvious +preview of the test to show the problemPlease dont apply yet as I dont think the Lucene module should depend on all store modules +Doesnt make closing stale resolved issues +Im concerned that components inside loops will not render properly +I committed this +These API changes were not in the original validation release of the apis so Im going to back it out of our release for now +Per comments this issue should be closed. +for convenience for reviewing ive attached a pdf of the changed pages +The request is over why remove attributes from itFix it in Tomahawk +But we saw this in testing cluster +committed patch to the trunkmsvn commitSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiTransmitting file dataCommitted revision +Committed to branch and trunk thanks Gab! +You need to make sure that your DAO actually executes within a transactional context + +sounds like you might have a plugin loading eclipse with debug and check error log view if there are some plugin loading issues +seems to be this bug is already fixedif not please reopen it with a new patch. +Right thanks for the info +The changes look good Ben over to you +Thanks +Implemented using new JDBC connection pooling no longer need wrappers +I have attached a patch to which also addresses this issue +Yes it works! Could you commit +Ill fix those too as part of this work +If there is something I could help Id be glad to assist +I cannot recreate the issue after I restarted I see it again I will try to see if I can collect more information. +We missed our chance to do this before went out the door sorry guys better luck next grid api +You should also include in your optimizations if possible as its implementing the same functionality +The latest patch looks good to me modulo one nit looks like theres an unintended import of in +BTW I think we should create a branch to continue this work +seems to be incomplete and are empty for instance. +Anyone up for doing this +Remoting is in the brew maven repository but not in the main repository +That is an incorrect assumption +I just committed a fix to the Java serialization framework +Approved for. +See for further details. +If one wants to implement a that only supports filtering by Path or that optimizes that particular call then there are plenty of other ways to effect this without adding a separate method to the core API +Doing a lookup every time might be negligible in terms of performance but one can never know without actually doing some performance tests before and after +Updating the subject and component of this JIRA to more accurately reflect the suggested changes +The other is a production version which secures the console against remote deployment +Thanks Jason! +Logged In YES useridWe are facing the same issue as well and this issue will prevent us from porting to +Well use our own jpox bundle with the patch applied in. +The attached project seems to reproduce your issue but the projects are just malformed It are two aggregator that reference each might want to check your project setup for cycles or attach all your involved for inspection +This would control readwrite access since the class will correctly error out if write access is not the intent of this message is sound extend the fine grained insert delete update access control down to the level +None of these present the debug or run as ant build options +if well use ajax I think it will be with extremecomponent table +I figured that since a sequence was declared it should be available within getSequence but apparently that may not be the case +Great stuff Dejan +So closing the issueThanks Bryan! +Later we can fix the problems and thats all +Perhaps a disable flag or something like that. +Thank you Scott looks good closing. +The WTP XML Editor is extensible allowing us to enable code completions + fix version +Moved to upon release of + since it would break H apps and the wording does not prevent the current behavior in the resolution status +Net define NETNETNET To target +It is very useful for slow appender +I am switching on additional debugging for the unit test +On a modified version of that creates a new message for every send my modifications makes it about faster when using Stomp +Thanks for all your hard work through many iterations Michael! +Maybe these could be implemented in if there is a strong need for these featuresWidothanks for your reminder i was just about to submit the patch to reviewboard it wont compile without replacing the manually +This was only fixed in as far as I recall +Its just as if orphan deletion on isnt working at all +I dont understand I dont think a should be possible improvement should be added in a release +Trying with +Modified fix version toAssigned to pamarcelot +Committed to trunk +I dont see any way of fixing this unless the form tag places something in the stack when validation is enabled so the dojo plugin avoids requesting the same file +Seems like theres a property in a custom file that isnt available this to be caused by an older version of the class beeing loaded somehow +on adding a trim in trunk +I am not against the notion but want to be sure +I will go on testing for it by manual +worries AhmedBTW is your formatting intended +Now that work is well underway with AS all previous community releases are +This is closely related to +bq +Thanks! +Reuploading the patch +Screenshot of the error that is reported to the user when he opens the dialog without ssh key configured +Fixes implemented in scope of +Added a new proposal for the new based on the new hfile and thoughts around the new server implementation +Closing as this issue does not appear to be related to Roller in any way +Simple enough patch +This workflow Services are wrong +Hi Ken i am also facing the same issue +Whats the link to current contributor guide +If youd like to post a patch here youre welcome to +Marking critical +Thanks Uri +failing junit test case demonstrates incompatibility between mcompatibletrue and transactionaltrue +cant reproduce with the latest still exists in current +Doesnt look like a real duplicate but its somehow related +Yes I suppose it very well could be a generic Groovy bug +So I think its not a problem of GEP +I also added a constructor that specifies this. +Should I add the fields of the to the that the other tests use? The test should then succeed + +I dont disagree that the implementation of the checkout process can be better but I disagree with using minilang for it +target directory has been removed. +The solution proposed there bases on the class which has been replaced by and which makes a generic solution more difficult +Thanks a lot for the patch +And he proposed an initial patch +Thanks MikeOne small change I made is moved the class to +Just following up is there a way consumers of the can specify the to use today? If so just let me know and I can update the jenkins plugin +Should we flush before compacting when user triggers a minor or major compactionNo need +I have marked as a duplicated of in bugzilla and tested future it is fixed there. +Thanks Steve for your prompt responseCan you please pass the details of your search on and the exact issue that you found which in turn answers my question. +Thats pretty standard fare for IE +The original error message seen on most platforms contains the following or while the new message string seen on Solaris contains an extra comma or Since neither the pattern nor the substitute part of the sed script are allowed to contain a comma I am unable to get rid of the comma after +Can you not add in codehauss repo into your +Oops the trunk patch does not compile +Assigning this to Store component +replaced by +A few of the lines now or still exceed characters +We stopped deprecating mapred +bq topnI had made changes to to look at each run +EAP CP will use WS that version has the fix backported this issue is resolved +Step of the refactoring has been implemented as aboveIn preparation for step please add additional clean up that should be done +The main question behind this issue which tool generates coverage in Cobertura format but not in LCOV +I am submiting the PATCH resolving this issue +Nothing on the java version +Heres a patch to add the interfaces and categorizations. +that goes out w the release that highlights the major differences +Lets get this into +The modification is around the new configuration property The property may be provided as a single string as an array of strings or as a collection of strings +Thanks again to all that contributed to this patch +Attaching a patch for which takes into account the Optimistic setting when calculating the default lock mode to apply to a named query +and commited! Thanks for the patch! +It may make sense to have a hangoutirc meeting because the delta is so large +This makes the resulting assignment more stable at the cost of making the distribution probabilistic +Only API should in in shared lib folder +Its deprecated now +Yes indeed +I am happy to do that but what should the client be doing? My thought is that if the CDATA isnt in the spec then the client should not be including it +hehe Cool! +batch transition to closed remaining resolved bugs +thanks vishal +This issue has been fixed and released as part of release +PR for master merged in master branch +Well have to find a way to pass the realm m looking at the line right now we should be passing the rd argument of realm to the authenticate method +You may now have to wait for a focusIn event before making the next call but then you cant run a loop in a single frameAnother issue has been filed that might be related to this change +The proposed patch appends the file size to the target filename +Comments about this patch will be posted in the reviewboard +If they do go in then do we just use some dummy sql states for them so we can use the existing mechanics to reference them in the code? is the right place +would be very helpful +Once again thanks a lot your test made a difference! +Added this line toJVMOPTSJVMOPTS XsskSame JRE as above +Please dont resolve unassigned issues that I scheduled for a particular removed +We could support another serialization mechanisms as long as they do not require additional dependenciesOleg +revision . +Hi Charles if the application is running upon dfs its really hard for us to achieve for it +not really sure how to troubleshoot this tbh since I dont know any bit windows machines around to try it oneither its not getting into the commitlog in the first place or it doesnt get replayed or maybe its not commitlog related at allMaybe try forcing a flush before shutdown see if that makes it go away +archive file but here is +bug of a bulk update all resolutions changed to done please review history to original resolution type +for + thanks +So in the third chunk of the commit above seems to be the source of the problem +Protocol negotiation cannot work as the socket has been closed +But the problem is that all other conversion methods left in +no longer an issue see explanation above. +Attached patch +Attach a sample project that exhibits the problem. +please submit your proposal to melange +Besides they make use of native code accessing native Windows somewhere in their NTLM implementation which makes NTLM auth in JRE HTTP Windows supports through JCIFS on all platforms +Whether it is setup or command it is not as relevant +I think it still doesnt quite get the case comparisons right it uppercases the cql input but not the key nameLooks like things would be a lot simpler if we stored the cql input key as String instead of bytebuffer +Will reopen should we ever need to add patches beyond whats in branch. +It is not an incompatible change +occasionally fails +I applied the addendum to trunk and +Ill commit this momentarily +I trited to to start a VM and router started succesfully +latest psychopath jar +The only person who cares about this issue is no longer using +I meant RIT is with the original server name only and not yet updated with the new RS +done and tested looks good +Thanks Navis +Ah I misunderstood +The limiting factor seems to be that we have to treat it as plain textbecause it has already been run through the security stuff +I have produced a fix for this problem +I think we can use as a substitute here will make a patch and test it +A blocking API could be implemented on top of the API if need be +since if it is for relationships and has no relationships? surely it would have to have some relationships to test managed relationships? Maybe Im missing some secret naming schemaThe naming scheme for the conf files is supposed to reflect both the type of test relationship and the test data +We went for the separate document approach. +Reviewed by Mick +Im wondering if roo could just incorporate the standard Maven client code at which point there is a maven plugin goal help that parses the POM hierarchy and produces the resulting effective POM +up there if you wanted +Can you have a lookThanks make use of +Committed to branch and trunk thanks for the good work Gabriel +Patch for +Correction But is not specified and it defaults to FALSE +done +Ive uninstalled everyone of the extensions and themes no difference +Steven you applied the patch as is or with the amendment as well +Actually looks like a duplicate in any case likely same root cause bugged fix for. +Thanks Harsh +Here is an attachmentThis can be safely committed as it is not currently included into compilation and will not break anything +Overall I have the feeling that having a clearer syntactic differentiation is probably goodbq +I think we could close this as a dup +yes +Fixed with attached patch. +Generated applications works on GAE Tomcat and. +This caused a concurrent access error +Thank you. +My bad +I dont have the rights to create the SVN repo +If you can take a jstack dump of the process when it hangs we can do some forensics +ThanksUma +This is how it looks now +However I found this problem also occurs in Chrome +Created new methods split to mirror the existing split methodsadding the functionality of preserving empty tokens indicated by adjacent released closing. +Did anyone this patch yet? The version was looking good but I hadnt yet given a and weve had updates posted since then which I was still in the process of working through +Handles periodic shipping of log chunks from the log buffer to the slave Allows for force flushing of the log Buffer by the when the gets a from theDesign of the classImplements interfaces Serviceable ConstructorAccepts the shippingDaemon and the transmitIntervalSubscribes to theMethodsshipChecks to see if a log record chunk is available in the logBuffer for transmissionIf yes removes this log record chunk and transmits it to the slaveupdates the last to reflect the current time at which the transmit was doneflushedInstance interface methodUsed to update the latest instance of the log record that has been flushed to the diskThis method will not have any implementationforceFlushForces a chunk of the log record from the log buffer to be sent to theslave freeing space in the log buffercalls ship to ship the log chunk from the bufferserviceASAPServiceable interface method +Did you had to spend some time to try my test case +comments on phabricator +Complete in r. +thanks bob +Thanks Ivan +I agree with Henning that relying on finalize is not perfect but it should be a big improvementNote that I added a line to set appender to null allowing shutdown to be called multiple times with no effect +It worked several times in a row so well see how it goes +I filed. +Committed fix to branches Branch and master at gitronsigal +Updated the all resolved tickets from or older +It would probably be interesting to have the possibility to define new trainer types in the future and dont break backward compatibility with thatMaybe we should have a method which can return all the supported trainer types depending on the params object +Fixing Affects and Fix in Version +Please attach a real example of the an example app will take a while +The of pbs I did not do but picked up the creation of using byte array to avoid it internally assigning buffer +Please verify whether the problem is resolved as you expected +Refreshed tests pass +I believe there are two loops which should be merged in the first the priors are transformed with the exponential function and in the second the transformed values are summed up +I just committed the attached patch to trunk +I dont know enough about bsh to debug this a lot furtherAny clues from your end +Yes I agree +It would be extremely helpful if you guys have a good example of test +No +If I get a few s I will check this in +Do we need to reconsider the importance of this issue? +There appears to be no public artifact for accumulo available on maven central +except dont modify the logj config +please update your implementation spec to include your plan for hard upgrade +Yes its a limitation of Hibernate not much we can do since its a Hibernate Graeme but i disagree with you +Committed revision . +For some reason slink does not have the image attribute +Dobes dont set the affects version when filing JIRA +Richfaces presumably could help here +I think so +The severity of this bug is critical because many services rely on the Service Engine to correctly validate required parameters +IzzakPossibly try creating the table on one run then check for the table on a seperate runThats the only other thing i can think of +We need to add some alert for the users in xbaya saying workflow finished +below is the snippet for the samecommandaddipaddressnicIdresponsejsonsessionkeymysql select from nicsecondaryips id uuid vmId nicId ipaddress ipaddress networkid created accountid domainid NULL +Ram Yes sir +Would be possible to but it should work as implemented as well even though going through the naming containers might be more performantregards checking ids with special handling of naming containers is better than using clientIds and closer to the spec +Working as designed. +Do you have a suggested fix or is it more of an observation? I didnt see a change in the patches +I just tested MDB sending thousands of messages the other day +You can see as equivalent to +At the time I needed to test non String data and there was noneSo +Im most likely never even going to publish anywhere +The standard optimistic concurrency mechanism can be used to specify that a document must exist if desired +Could you please retest with the right APR library +Could you please update patch after commit of +IIRC Gavin fixed this +and why would they +this adds overhead maybe jenkins is good enoughit runs many times a day and we dont actually change jars that often most of the timewhen developing we are just changing code +It looks a valid use case + support requires just committed this +Updated patch with verifyRequest added to start method +Nick I will need this bundle in as an din the openshift express client that was moved to github +Outstanding issues are logged separatelyClosing this issue +Ill test the performance of this solution soon +BTW found also one not ignored autogenerated resource in the could not directly I apply the patch +Close issue and create another issue or fix and close again +Thanks Sanjay! +I noticed this yesterday too +I will work on your feedback and will upload a new patch soon +By now I cant reproduce it anymore +Committed in +Once is in we may get a better picture where will be the newbottleneck then and make due decisions on some areasOverall the direction is clearly right +Will take care of this today +Edited comment for clarity +The reasons for this are documented in When running under our test harness from jar files under system properties cant be read +Well itd be best if there was a resolution to so as to avoid needing to workaround the flat hierarchy problem +I have tested the kit against SQL Server using my latest patch +sorry for this +If the grandparent class includes a switch on if nodeRackspace then the idea of abstraction is lost +Tests passing in hudson. +I agree with Mike that there is already exposure for encrypted databases when the data is part of the statement so I think that it should be ok to go ahead and commit this patch +Sorry I forgot to regenerate the testcase outputs +Looks like there is something wrong with the commons io versions probably due to the different versions required by Hadoop +I forgot to run them after making my change and since I added in new columns to the data the number of tds changed + +If you have something more complex than what is currently supported perhaps in general you are better off leveraging a build system like gradle or even considering if you need strict isolation and can justify the added complexitySo today we just merge all Grab related annotations together making the Grapes annotation effectively a at the moment and hence why we dont encourage it much these days +Note that you dont have to implement the Aware interfaces anymore you may also use a private Autowired field +This would be for the quick answer +patch for javatrunk rev applied +I think while running its better to log more info about each server +Raising it to a higher value is fine by me. +What kind of compatibility do you mean? Im not sure where compatibility is required when youre generating a random key +no longer reproducible +Setting the threshold to to does not fix the failure +Merge works on so solution is to run it there. +this is fixed now right +Is this a critical issue +Seems prudent to leave as v +Test failure is not related to this patch + +Is there a separate mailing list for questions about luceneutils or should I write to the list? or directly to Mike or youTom +Preliminary patch +I dont think there should be any expectations that the thing called a list in CQL parlance is the same thing as a list in Python. +Rohit can you please confirm that no documentation is needed for this change +Vitali this is opensource if you need the fix provide a patch or suggest a fix +Banner used on screens within the installer +On closer examination all of the examples I have found use an id rather than a name in the path +Im seeing several compile errors due to changes in the Jsr classes and the removal of the classesCould you fix the build problems run and resubmit the patchThanks +perf +so that theyre closer to the signal based thread dumps that were all used to +updated for and trunkGiven the silly angst with the branch I wont try to get this into unless it is respun againThanks Karl let me know if anything is missing +I think its just a tradition +Basically people are looking to use the Labels to isolate web data based on security labels. +Incorporating the change for because that fix was needed too to test any webhdfs operation in secure mode +Closing all resolved issues for. +i also prefer not to use inner it also possible to support this in JDK using conventions rather than probably wont be an issue if the Scope option sorts it until the serialization code is needed it is probably a bit early to know +Before the form is submitted the call decorator retrieves the value from the editor and sets it to the hidden input field +Committed revision +r +Sometimes I almost think I attended a lawsuit mail list when I saw there are so many letter on legals +I have modified the code so lexical are resolved using the prefixes declared in the stylesheetAlain if possible can you test this patchCommitters can you please reviewThanks +Please test it thanks +Shall we have column for info like app start time? If so we can build filter on it so that we can list jobs running within hour day etcbtw I saw timeout intermittent in some jenkins tests before +Also check out m going to resolve this as a duplicate of since they both cover the same ground +Duplicate of +Thanks for your input HaraldChaging the Format field of the IDE Structure parameter to produced a desirable printed output +please provide a patch if you know all the youll find a patch which is more accurate than the current validator which only tries to apply the Luhn algorithm +GQL is already in so the purity is broken +In the words of Homer Simpson If somethings hard to do then its not worth doing +In the past the plugin pack has been a superset of the EAP plugin pack it contained all of the plugins from EAP pack as well as the ESB plugin +Also updated the wiki build page for it +Id like to see the used in test too +Mostly fixed and issue created for the remaining few warnings. +Should have a patch ready for review tonight +The bug was located in the settings page not in jacoco nor in the new settings api +I updated the wiki to include the wsdlLocation parameter as well as the other missing parameters +Hi PriyankaSure you could +I cant reconstruct it based on attached screenshot. +Testcase failure not related to this patch +Tried with windows IE in compatibility mode and didnt encounter any issue. +Bumping it up to Blocker since it fails builds as spent a couple hours trying to figure this out but no luck +Please correct me if Im wrong +Attached is a simple maven project that can reproduce the building deploy the war created in jcditestwebapptarget directory to the app server with support +We do not add child Web Resources with same content as +JamesWhat would you want me to do about the latest version of the patch? Can you rebuild it against the latest SVN trunk? version in SVN trunk is complete +Thanks for the contribution Mathias +I only know that contains +I thought maven would already do that if you dont supply the clean option? Shouldnt this be handled by the Plexus container itself instead of controlling the build process from the plugin? Seems like the tail wagging the dog to me +To me that says its valid XML and that the patterns should be ORd together +Denny thank you +deletion takes minutes so I must say deletion is time harder operation average anyone can check +Please could we include this +This exposes other issues however and I feel we need to spend more time understanding these before we can move forward +bq +but what is happens in default case? As I think this was a reason to bundle these jars +On the other hand we believe the old behavior was erroneous so by our usual rule we do not need a release noteWhat do you think? I guess I always prefer to err on the side of including too much +Unlinking from since we are about to branch for release +Unassigning myself for now +I think it probably makes sense to use the hardcoded version number in the Tools Guide topic the way the other books do +wrong bug number wrong bugzilla +Done in trunk + +finalized implementation of this issue Refactored resolution of error handler servletsscripts to use new mechanism Removed unused methods and classes Renamed helper classes to reflect their functionalityClosing this issue for now +passed smoke tests on winlnx reported classes are now used as temporary storage for enumerating not yet loaded classes should this collection be renamed +I dont understand why the javadocs cant be expanded to show how to properly implement accessible components. +As Mark indicates this could be something interesting to experiment with. +The two issues are critical and somewhat complicated issuesIm going to write the tests without them we only have your word they are fixed today and a random guess they will still be fixed tomorrow or the next day +Documentation applied to Tuscany website because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Needed to set the on the server formatting provider in the test remoting receiver to +Note that for a normal facet query this could result in waves of requests +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +Bruce this again as described in one of my email to the committers list +Hi KimYa its STMTTEXT row of the SYSXPLAINSTATEMENTS tablevisitedpages from SYSXPLAINSCANPROPS is this NOVISITEDPAGESscannedobject from SYSXPLAINSCANPROPS is this SCANOBJECTNAME? You got these two rightscantype from SYSXPLAINSCANPROPS is this SCANOBJECTTYPE? Its SCANTYPE from SYSXPLAINSCANPROPSsorteroutput from SYSXPLAINSORTPROPS Its NOOUTPUTROWS of SYSXPLAINSORTPROPS Sorry for not being clearThanks +But its at best a good default and needs to be changeableGood news is this is already how works sort of + will commit after tests t you missing db dir in there +If I understand correctly seems to be relating to network monitoring whose responsibility is to monitor network and alert administrators when events raise but the purpose of failure detector is to detect node crashes in a distributed system +Mass closing all issues that have been in resolved state for months or more without any feedback or update +If we start with a local repository that has none of these modules deployed and utter mvn clean install site in the the build fails because it cannot resolve module as a have checked with the recently released and and the problem still exists in both +SNI is a TLS level aspect not HTTP +Hey PeterNicely done! Keep up the great work Peter great work! Can you commit these improvements to SIS? +You can grab the header from any of the other java files +Indeed even in term of remediation action there isnt a big difference between moving or constants from an interface to a class or enumand then the issue message should be Move constants to a class or enum +That error was occured on machines one described above and secondOS Linux SMP Sun Oct GMT xJava Strange think is that i cannot reproduce it on th machineOS Linux dc SMP x x x GNULinuxJava think this is ready to be closed soon +Also remember to update the release notes and that is required due the bug in older Gleen has been released which has the fix +why +Thanks Raghu. +will commit later today +Which does eclipse default to +This patch does not handle nulls in joinsgroupcogroup and patch checked in +We can see on the screenshot that theres no matching rules that was the problem +Reading the comments Im closing this issue. +Is that a problem? I think we could remove the Java stuff quite easily but then we will not get the Spring JMX features at least not without some more modifications +Does it mean that I should find another solution for this problem +Only the version throws the which exposes the problem +sending similar request to webdavpackages HTTP to get children list of packages it works +Added additional exception handling for the cases when nodes are removedhandled from multiple sessionsThe changes required were a lot less complicated then I originally thought it was just a matter of adding some extra validations exceptions +Also changed from ARRAYLIST to LINKEDARRYSET the collection that holds the library paths to ignore duplicate entries +We have noticed the same behavior when unit tests fail think this is a regression in one of the maven components because the same release plugin behaves correctly with maven + at +Misread the license option +Usually this has helped to separate logic regarding computation of the token values from processing logic such as authentication which is happening in the Services class called durring filter processing +Steps to test this functionality Create two or more sales order with products and Go to Facility Picking Click on Print Pick Sheet button Now click on Review Orders Not buttonExpected result It Shows list of orders with pick sheet printed date which are not in picked or packed status yet +May be we should totally remove this code at some point. +Some refactoring pending +I will keep the tabs out next time +Let me state it again Applications with paths using as separator could be used fromon UnixWindows Some times becomes escape character and some times separator +I dont see how another server would be able to grab the thread and continue executing it +So fiddling with it should be OK but I do not recommend it because it is specifically maintained for +switchType is fixed for richaccordionItem richtoggle and richtab +Hi I just tried building the branch locally and it builds successfully +Make public +Turning it on and doing a projectclean in eclipse made all the editor error marks disappear Youre also right about having to do a clean each time youre starting eclipse cause these fake errors do appear each I guess you can sum it up on the Make sure WindowPreferencesJDT Weaving in turned The build path +fixed comment +Maybe we can change the server connection implementation to add the IP address to the? Im not sure how that fits with the NIO stuff thoughcheers Roland +Assign to during tidy up prior to +Do we also plan to backport modules to branch +Committed revision +Second try with correct log level +bq +Could you regenerate it? +I will put up my latest cleanups in a second +to prevent a synonym or phrase query from matching over the punctuation +Verified on EAP. +So were really talking about a relatively small amount of work to build a PS api over a String one +How are the Java contributions handled? They seem to be easy to addremove as wanted and only shows up on Java +I think we can tie that to +Its not picking this up + +Null classloader means system classloader you can use that assumption +As you now your encoders are used in search engines like SOLR in a basic manner +I am implementing changes now to completely support mixing freemarker velocity and JSP decorators as well as making the same contexts available to the decorators as are available to the results +If you think they are critical for CR please them +If needed I can supply a different format +Docs are working towards the docs timeline as outline in the DRD Writer freeze QE freeze Docs Freeze +Fixed in HEAD +Hi this bug is fixed in the webconsole trunk +looks good +Krzysztofthanks for testingPlease post your questions to the user list youll have better chances to get an answerIm sorry I cant help more with this +fixed in r +Im guessing this isnt the proper procedure and I probably didnt create the patch correctly so let me know if theres anything I can do to smooth over the process assuming you want to incorporate the changeThanksLP +We do not set the readseen flag in that case so it is usually not needed to explicitly set the SEEN flag to false as it still is unseen +OK thanks +Thanks i applied your touchup patch +Removal and search over the heap is slow with the Java implementation +Going to commit to to verify this indeed is all that we need to fix to get the bigtop ci working +Finger memory +They may fail due to the bookie being rereplicated being taken down for upgrade but its ok because the admin should run a health check on the cluster after upgrading all bookies anyhow +Hi is there any expected date for this version to be GA? Please let me know if i can help in anything + attached patch +branch has been updated to look like the attached screenshot +it should be a solr module + +This is not a Roo bug as Roo provides no function to refactor +patch looks good +But the major change is made to +Hi David new version of jira is deployed +When a file is no longer published to any server the decoration goes decoration for a server when a connection to it cant be established +A map file seems much more flexible than the case statementsRobert in your comment above you mention Solr includes a mapping based on the deprecatedCould you please point me to where I can find this mapping file for the deprecated in SVNOtherwise Ill just run or adapt the perl code here and run it against the codeWe havent switched to the newer and need to emulate the in some custom code until we make the switch and all million volumesTom think this is ready to go +Ill revisit this if it hurts performance +Thanks again for the patches. +Ill see how to get something better going +I can reproduce it on opensuse boxes but it works fine on the latest Ubuntu +Thanks for the contribution Peter +Tested All issues on and all issues have been addressed and pass +Yonik any more feedback on this patch +Without the new tests the complete CTF runs successfully +The sample in github works perfectly fine for the connect flow because it injects display popup additional parameter using the interceptor framework +The solrj tests log just fine without in so I think that this solution may not be entirely right +However we still wait a little while to see if theres anything coming our way to make sure that we can gracefuly recover in most of the cases +hmm I just compared my test eclipse installation configuration details with yours but cant see any difference +Additionally Ive created another which simply does the same thing but only preloads beans assignable to the interface +Committed at Please check that the bug is fixed for you +move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +content is in the response from not given to UI layer +Testing +In addition there are a lot more moving parts and a lot more code than is needed if more code is moved into the compat modules +Yes +Or do those get cleaned up by something else at some point +Great job and thanks +Modified as SVN revision +I have started a separate discussion thread on the Tuscany dev list to talk about what we should do with the travel sample binary distribution +We really need to have this in preferably sooner +The error message is changed from cannot open a file to be File doesnt exist +However I need to verify that the fix does not break any other database if so we will need an abstraction +Committed to +Moving these all to a Doc release version. +Existing tests for demonstrate the race is fixed wo synchronizing the method +Interestingly if you add DESC to the ORDER BY statements that dont have one they fail with the same type of error +However I thinks its important to be consistent with the way exceptions are handled +Failures and errors reduced down to and respectively +marking resolved now that properly schedule +Splitting out into two categories will affect screen captures a small task to dont think code quality is the right terminology for the first category +This is my patch that appears to fix this problemIts not what I would call perfect since it may be too wide and the patch may not apply as there are some line shifts where I have other logging statements in the files +Ignore the previous comment I posted it to wrong jira +I have not looked at what else need to be changed but those two paragraphs should go +Ill take a deeper look at the changes in a momentTook a lot longer than expected +As a result when either the sender or the receiver does not have resource to handle block move the block content will not get transfered across network +Can you this behviour with a Vanilla JSF application? ie +I will polish this +In order to facilitate reproducing the problem Im attaching my cassandra data directory targzed upThere are rows in the cassandra instance +Thats right Isaac and thats why I started a discussion on the same on the dev ML +Ive been trying to get a secondary cluster setup since the st +I think the event doesnt make sense for controls without value because the criticality is also changeableId propose to replace the with a instead which always creates a new instance of the control +Backed out of Some major effort is needed to make enhancements to load balancer coherent easy to understand and maintainable. +I am not sure if is the correct data type for success failed etc +is there solution or workaround for this issuecould I help with providing sample code +Thanks for the answer! Please explainto me what needs to be done +I think maybe this can be seen as a duplicate of but I still think theres a difference because the socket reset happens on other systems too and it does not always cause a hang as described in +Not sure how to fix this one +It is natural to have a common utility jar in earlib which is accessed by the modules deployed as and wars +I was able to implement Acegi with NTLM following Rory Winstons example but Im running into a problem +What makes you think that the code is treated the same way as data in JVM and that it gets garbage collected as regular data +Would anyone want to try this one +In v patch chang all call of constructor to builder which cross different +bulk fix version change tests needs them +Attaching a test project that reproduces the adapted and applied with can check the dont see which change can produce this another patch I can you open a new issue for ll fix it ASAP. +Fixed in various commits +We have two problems with the initialization of the thread +It works very cannot wait for it +I see the following in the Tomcat consoleMar PM initINFO Tiles definition factory loaded for module Feel free to reopen this if the example apps are still not working for you. +It has been fixed only in with r. +VladimirFixed in NIOCHAR at repo revision Please check this fully resolves your problem +We are trying to read headers and the example does not show what context is or how it was ignore last post +Too bad +Forgot to select which version this fix will be in. +I am hoping the CRC is the very last value in a record and on disk it is aligned to its size so that a partial record ends in zero +Dont all servlet containers already have existing logging mechanisms for this +There is a listener on the that was not being called from the what led to this issue +Maybe we should start the framework in our own classloader that doesnt use the maven realm as parentNote that the karaf is started in the maven vm not a separate vm +Happy holidays! I am out of office from June to July during July th week +A statusCode property has been added. +Verified with. +Addressed Jimmys comments! Thanks Jimmy +This one can be described in words simply enough +In this patch string is used for Timestamp for simplicity reasons +So given that for keep alive that would involve potential code and such lets keep that for later unless someone is willing to actually write that patch in a timely fashion +The problem is still there in IR +I propose a test ratio of as a fair deal +Added getter methods to FSD for the underlying fileExists checking in getDirectory before askingregarding the extension +It will also require some reordering of bind parameters +Any processing we do with this is completely arbitrary and +No follow up when a test case was requested +Looks like a good idea go thing that Im thinking about is where to draw the line between generically useful metadata and stuff that is needed by only one dialect +i think this is a better default and if we need to add priorities we can laterJuergen can you go ahead and enable it and tweak the tests? thanks. +Updated patch with tests generified and also using autoboxing in javadocs and test code +Can you please verify +captueres the overarching issue +Patch Justas +Documented in the EAP release notes +Another reason for this change if there is a stomp connector it will exhaust the producer audit LRU cache with haste because of the burn rate on connections +bq +Original problem is fixed the new issue reported is invalid +Ahh yes +First cut of the clean up +Direct link to mail item didnt work +I thought I share it somehow but maybe its really better if I have a standalone behavior and write a small article on my blog or the wiki about this +A really trivial bugThis error wont occur in IBM JDK or Sun JDK +Timeout etc but also programmatically managed +Talked with Ken the name of the teiid and modeshape zips will be changed to assembly changed to make the directory name for the EDSzip file eds in revision of Commit message The directory name for the EDS zip is eds +Shouldnt this be if it hasnt been resolved? I believe this bug is causing my to run out of permgen space +The pattern is now the same +Uwe why do you see so much variance in your times? EG best for trie is msec but avg is msec +Im still confused about why a hardhatted elephant is in poor taste and not appropriate +With your guidance I can put some time helping to implement something +So I am routing this JIRA to Fred as it will be fixed when he fixes +There is a parameter in the LIMIT clause which is allowed by all versions of I have ever tested +I agree that rank reaching is less unlikely than view ID which are the number of view changesIts quite an edge case tbh but as you said Id swap around the rank and view ids to protect against this just in caseI need to verify how easy is it to keep this value within the unsigned ranges +Please run test javadoc findbugs etc with the patch I will commit once these results are available +If the logger is running within the NN the namespace integrity is potentially compromised +I was confused about the issue with multiple copies of interceptors and alternatives this had happened when I pulled the sources and found that the for the descriptors are generated +Otherwise they dont have access to the protected elements in that package any more +Patch works drops test time down to seconds +Thanks +The component should be updated to include the columnsOrder attribute +Deferring +The best approach is probably a comment on the other issue or on the mailing list with your idea +It is essential to have flexibility as the overhead associated with indexing linkdb data is unnecessary if not required +failed tests are not relatedI have committed this +for the patch +Juergen Thanks for implementing this for a question +Patch applied +This was achieved by combining jxr and into the same directory in in svn and also by using a unified release cycle like in surefire. +Moved toOrder Prosessor and Trade UI uses the copy in +I asked the networkers of our partners already why they dont wanted to use this kind of configuration and still wait for their responseIf I get the good reasons or not I will inform you at once +Sorry for the inconvenience and the work involvedI just remember it was the very first example I tried and failed +Consensus seems to be that Iterators are the best choice +I added some java script that handeled additional links in the result tabelCheers patch was out dated +Committed with minor changes in trunk At revision +to trunk +This patch fixed the little thing as well +All the regression tests passed with the patch +Everything was set so that the url conversion was executed in the background +docbooksink book book sink after some rework +documentation +Submitting patch again hopefully Hudson shows up +I see no real use in it it does not affect query performance only index size +Run on revision on revision if create a page and set is not empty layout containers permission hasnt seted yet and display Protected configure the Access Permission to Everyone by default +Sorry +The bug seems identical execpt that was targeted to the branch +updated patch is attached +and your input format is also thatSo with that at least where we use it there is no need for a special duration data type +The meaning of Reserved that current implements is Space that Datanode should try to keep free for immediate or future use by either Datanode or some other application +These other methods may have a similar problemwrite not synced but called from public run method and other synced methodsclean waited variable set prior to sync +Changed assignment of ret according to s point I didnt look carefully enough +I will remove it and apply the few changes that went into it accordinglyChanging the SPLITDIR is from which went into and +I implemented and checked in scenarios and +Looks goodonly remaining thing is switching the import Ill open a separate ticket for that +Another nice piece of work +Manually tested +Daniel can you please also change the indentation in this patch to two spaces +More on this the Java key actually appears when the analysis starts +Reopening issue as per Eric +Great work +The Derby JAR file is sealed and this means that all the Derby packages specified in the JAR files Manifest are sealed and you cannot add a new Class that shares the Package names to the Classpath +This is done reflectively in order to be compatible with JPA as it is impossible to update our at this point The new methods in that interface refer tp and enums which are only present in the JPA API +I have risen upper version to and fixed error message to mention Java +Based on Matts patch Synced to trunk Uses All tests passI had to change frequency to an int to match the output of distributed and cases +The code may become a bit easier to follow but just a suggestionI kept this +It makes perfect sense to have this configurable +After upgrading to Hudson ver +Attached patch file committed adaedeaddbefcbdefThanks +just committed the patch to master +This problem has been fixed +priority increased because ajax update brokes the component fully in nothing works in FF popup under jspmyfacestimcat +it is reaching that as such array now right? Because that is something we absolutely need ot fix +Ive committed a fix to SVN. +But why not closing the? Anyways I applied the first half of the patchregards good to me +Fixed in +Actually I think I might know what is going on +Can I download the latest build from Hudson? Its the same +Thanks for pushing me on this +I think most developer expect from a database that a safewrite will fail +Im assuming that as the reporter I am also expected to close the issue once I see it resolvedIf that is incorrect workflow I hope to be told what the correct workflow would beIve enjoyed learning how to contribute to this project thanks for providing the git mirror +Note that that includes a change to +Thanks for working on fixing the broken stuffIn addition to use cases I want to investigate the exact nature of the brokenness introduces maybe its fixable? Ill it in and iterate until it breaks +BTW Do you find such use of exceptions to be acceptable inside the name nodeNoI thought I made it clear that I had not reviewed the namenode code changes but only those in the public +Now just need train test per +When the thread dies the node eventually runs out of memory as the queue never gets emptied +Cool sometime soonYep +Whats the type of purchaseDate? Ive searched the svn and cant find such elementattributeAnyway SDO Java spec page and doesnt map anything to although theres one mapping seems close which is from any toHowever I doublt the type of purchaseDate is anyMany datetime types are mapped to Stringif you have other formats may help +Thanks for taking the time to explain this at great length +Now that work is well underway with AS all previous community releases are +against be this merged into +CCCI systems run DRLVM tests in modes +To be able to support create which only return should be able to return as done in append +This will give clients an NSRE +Thanks +Committed +Using this dictionary we can put values into a synthetic feature vector in just the locations specified by a single feature or interaction +fix a few style issues use getmntent rather than reading from procmounts directly +This patch should be applied to the trunk as well +Updated the message in the xmlstax case as well and expanded on streaming in the docs. +From one users point of view I still prefer Serializable Exceptions +It also means we cant upgrade the subversion repos to until this is know the answer fix it yourself +To reload without restaring tomcat +Please let me know if it works for you +files +Note that this problem only seems to occur when using a mapping file if I scrap the mapping file override encode and manually add the three the tags properly nestclose +For the perhaps we just need to provide a single implementation that accepts a List of +I would actually prefer to get chukwa out of the business entirely +I think its ok to apply your patch +NIO based Jetty server can be used +Thats slow though +Thanks Amareshwari! +Giraph users that desire support will need the maingiraph jar the and any hbase dependencies +could be nice if it could access to more stuff than just length +Ok Im closing this then. +but when I upgrade to +Including hcatalog sharelib within pig and hive sharelib will work for pig and hive jobs +Ive committed a slightly improved patch in rev +The bug still requires evaluation to find out which component is to be blamed +I think this can be closed? +So the problem is that +In our contribution the finegrained error propagation is solely used for reporting unique key constraint and versioningerrors but the design is very generic so in the future it can also be used for other situations where you want to report partial errors back to the client +Bulk closing for +Thanks Alan +Attaching a patch + this record since is now released. +Before that Id like visualize each statemachine for good design of killing a query +same problem is also in AS it seems only the property is supported while both properties work properly in AS. +Without answers I will commit as is +Well this JIRA issue would be the first step the next step would be another JIRA issue for Field Cache integration and another for +Thanks Fangfang! +wrong on both and sample +Thanks for taking a stab at thisHere are some comments I would suggest we set the initial version to For compatibility reason I would suggest to just ignore system status if no service is registered I dont like interface methods with throws Exception signature I suggest to define the method such as to return a status info why the service considers the system not ready this might be something like a string or more elaborate +You do not want to skip all fields +So this patch does not break backwards compatibility +for the change +the example have an implementation? How does the user run it + is the only sink in nested plan +bq +I wasnt able to repeat the scenario as above in my IDE +Please verify in EAP documentation on stage +We should change user names in initialization scripts +Also it was after I attached that patch +Ive fixed the testThe patc has been appliedPlease verify +Integrated to Thanks for the patch Gustavo +Is Jenkins having the same issues and if not why not? +Committed this to trunk +I need a little more time to investigate thatI also the method a little bit removing the boolean flags and some of the nesting +Shall we include this in release too as its an improvement and will be helpful in debugging issues +There is still one aspect to handle on this the issue of present systinet only has an HTTP transport which means that the ESB cannot be located within the same server as the registry +I believe this is a dependency in Serlialization +No need for IT +But since the trunk patch would presumably be quite a bit different it wouldnt help withWill have a look at the patch over the weekend +Fix released in. +I dont think that should be the only or default behavior since its not the behavior now or in vanilla implementations +If it is hard for even a relatively small percentage of users to get stable performance with mmap then I would agree that the default should be standard IO +Agree that its wrong that RPC wraps the EOFE with a +The intention is to separate the physical disconnect and session timeout as Carl mentioned +I use the export tab in their web page to get my test bounds +Handling of sDont crash with a NPE if some headers missing +Hi Glenthanks for the feedback +I tested it and applied it at svn rev +what is the significance of INPUTINITIALTHRIFTADDRESS forFor an output format this wont be used its only for input formatsbq +This issue is fixed and released as part of release +should be similar operation to +This issue is closed as out of date as per the middleware process rules for aging of bug reports +Is something wrong with his code +Looks good now and will be included in the next update +committed to as feaadaaeafafc +branch to help out our Mozilla brethren +Should we add a test that uses this so that someone doesnt inadvertently break this +The xmlbeans plugin does not pack the schemas in the jar hence added code to do that +trunk and have diverged a bit at this point too so Ill need to prepare a merge patch to resolve that +SVN +Closed with released. +Problem is described pretty well we should avoid calling the int version of methods when the value exceeds +It would probably be pretty easy to change that but I figure there might be a particular reason the object was chosen +Im failing to see how heartbeat intervals will help +Logged In YES useridAdd the full stack trace as an attachment so I can see wherethe load attempt is originating from +I had not been following this thread as the last note I had read was that solution was not going to involve changing the btree code + +Thanks for the answers. +Why nnId? Since the shared dir is shared it should just be nameserviceId right +Updated patch attachedLast patch was not properly generated +This issue is closed correctly +But when you shutdown the framework and restart it you can resolve both bundles due to the same reasons as explainedAnother point is that the exception thrown is a and when the framework will display the fact that there is a problem resolving the bundle i will show something like Unresolved constraint in bundle nullthis because the resolve exception has a null Requirement +Can I download the patched version somewhere as a JAR? Id like to put it in one of our test systems to check it out +this made the release +File Pom file update to test scalate. +Can you attach a simple project that demonstrates the behavior +No response assuming ok. +Committed to TRUNK. +It also has an enhancement to log an info message when a requested path is not inside the execution path +on this +With this disabled no classes will be compiled and hibernate scanner wont be able to find it +In my case conversationId was correlated with request parameter +I doObviously Id prefer not to but I dont know the correct argument to use to get around that +Hi AdrianMy second example uses the ofbiz properties files for storing the entity field descriptions +Updated the tracking file. +Therefore Im for it +We tried to keep encapsulated in the server what went wrong +Isnt it a problem +But it appears that your latest patch doesnt have any of Yoniks suggestions in it and still has the test in a completely new file rather than just another test case in as my patch had +When a occurs we will lose our ephemeral nodes +Patch applied to r +Updated the Materialization topic to reflect the other type of subquery that can be materialized +I believe this is covered by the linked issues. +Maybe I can find some time to come up with a little prototype but no guarantees +It doesnt affect the compiled code +Need to apply changes from in to javatestingorgapachederbyTestingfunctionTeststestsjdbcapi as a result of and was applied in the oposite order on compared to trunk +Both were generated with the maven one plugin +Let me fix the test +Im sorry I checked bin version of seam +Reproducing example removing the comment slashed in line runs it without a lock on +defer issues to +So maybe it is better to parse id and comment in a seperate function and call it for all possible elements when parsing? the same when exporting +ok +Thanks! +Run Maven Update source folders on both +Gonna commit Monday barring objection +Closing this issue lets open new ones when we want to add new modules +It eliminates the warnings but I suspect we dont really want to depend on for instance +So as agreed we will keep an eye on this +Missed an important detail from the logs the exception happened in this threadSFSB Passivation Thread jarnameRegistrarservice bug reproduced with a simple example with no Seam in it. +Plus no additional errors foundAnything else we can think of for this +DanI dont know whether Daniel runs with security on so I will wait for his response +Actions moved to +on patch +Committed to. +It would seem that this decision was in the original version and was reversed +I just checked size of generated java files for the example schemaMain classes k classes thanks I realize the implication You wrote about but anyway the mentioned points are preferred in our project +derbyall passes successfully on +have in mind that serialize is a build in function in from type bug to task and changed subject of the for input +Fixed +Ill be at the code this weekend and perhaps a few days next week if needed +I have updated and the trunk to now use warn instead of info. +Patch looks good +I am unsure because of the active discussion at the moment +The issue is still reproducing in issue is coming in Spring RELEASE any retract my issue +I am brand new to this discussion +It will still go to +Latest patch is a version of that Does document routingSends requests to each shard in a separate thread Uses javabin transport Is backwards compatible with both the response and exception +Attaching two scripts and which demonstrate these problems +In the description I think you mean fsimage rather than edits then +Changing summary and description since delete via web services was already implemented in +Thanks again for helping with there. +The find bugs warnings are not from this patch and exist in The core test failures are not from this patch and exist in No tests are needed for this build change +Interesting +This is due to the fact that is not backward compatible with +now it Ubuntu test now passes +Id say lets wait for the JBDSIS that has these issue addressed for a more favorable initial impression. +Unless I misunderstood your explanation I believe this is the proper behavior +Why do we need a wrapperWe should not be locking on the app attempt id +anything in the error logslooks like an error occurred in the swt component generation phase and then the thrown exception ignord +Not sure how is different from +solution by +Patch committed to trunk +Excellent thanks Nick +So is this ok? If so we can mark this one as fixed for and close Andrea I think Jesse was hoping for review but time marches on. +Well apply as soon as a patch is availableOh and intdiv is not so I suspect you meant intdiv above +Attachment fixes the whitespace +Please also check houtputLink with JSF and JSF +Davi hiThank you for pointing thisI also noticed this problem with one of the first BETA versions of Apache ANT but I decided to wait till the official release of ANT +The necessary changes are now available in the trunk +batch transition to closed remaining resolved bugs +passes verify and text both local and destributed +I just committed this. +You can submit a feature request for it +Although it is self sufficient imo and will cover most practical cases + + introduces code required to cleanup the blob and clob during commitrollback or when connection is closed +I like that approach Doug +Your comment applies to +Being able to force a single connection to be closed will be useful as connections via JMX will require some changes in the connection pool +This implementation uses the basic naming conventions to map properties defined as xml data to java object data +Syntax error in ELAny suggestions +There are a few other cases where we have dealt with identity caches of Schemas +yes this is critical +But I do not think this should be a blocker for +If theres no other comments from anyone please commit it to the internal Y branch and Ill commit it to the trunk as soon as is ready + +Text added +OlgaIs markup completed for the component? I see some HTML in SVN +Matthias I have attached the patch with the grant +Ill check if schedule deleted in database are stopped too +bq + contains new tests for multiplicity +Bounce to Isaac to see if this change warrants documentation +Oh a whole thread can be had on this Im sureWhy does one turn off speculative execution? Presumably because a MAP has unmanagedBut +Good catch +If you prefer I can send a pull request +Thomas could you please look at this? Let me know if I can help out there +But it could be his distributed cache was not working? Will the cache return a null path under normal circumstances rather than throwing an exception or something? +bq + +should post something this weekend at the latest +Now current Standby not able to load any edits from shared storage as expected edits are not present in shared storage +Closed issues related to Roller release. +for the rest of the centroidsdo compute a few candidate centroids which are far offCandidate centroid computationDivide the data into few parts +All images have now been updated for JBDS what OS are you using? You probably have the icons disabled in your OS +Patch to do so +I would even consider to pass that code all to EPL it might smooth transition to Eclipse Foundation +This introduces a deadlock that I am going to submit a patch +Fencing is the correct answer to prevent split brain in the general caseAsking for confirmation might be a nice improvement as well so long as theres a force option +also changed the exception messages +If its not true could you please provide detailed steps howto verifiy this jira thanks +Peter could you rebase this patch so I can review +No +Move it back in if you think differently. +any additional info here +Duplicate of +If client makes backed by a byte array of their own construction and then reuse then they will have issue +Committed at revision +Ill try to come up with something tonight +You can setup an automatic repo for faster upload +The most recent patch doesnt apply cleanly because of the commit of +But if we keep that capability we need to user friendly and let the user confirm +Moving out of +Added test +Its probably difficult to envision the right way to specify the command line flag anyway since its going to be a property or even expression +It maps to namespaces used in Java and EXSLT extensions +Has a step to count the items if they werent already counted in an earlier phase +Thats again so Ill close this ticket +Updated patch with the new test case moved into instead of inI think that is superseded by this work +Isnt redundant if we make optional +ill take a look tomorrow and hopefully come up with something that would work for as it is now and also work the way jeff thought it should +I had to have something hung up in eclipseThanks for your help +Also the can be better presented +Ill take a shot at this over the weekend +Abhinav Can you configure repo properly and try Assigning this defect to you +I will set up a production area for you to post your existingsite content to +did the test fail when it shouldnt? See my comment regarding +SankaI put the thread for a single reason the policy engine to able to read policy url references that are urls in the same tomcat containerthe problem is that on one hand you read the url but on the other hand tomcat is not yet started waiting for axis to deployAnyway I can easily remove the thread creation but we have to live with the restriction for url references not to be on the same tomcat containerRegards. +Right now Im working completely blind because you cant debug the LHS of a rule properly +No harm done + +this section in Admin Guide and close issue +From the README John left I cant tell whether these failures are expected +I forgot to handle deleted event +Since they are skipped the id will not be correctly replaced +is in place +Yeah you are right that we can keep that tricky part intact +Program will exitOn amd with u and u java XsskThe stack size specified is too small Specify at least kError Could not create the Java Virtual MachineError A fatal exception has occurred + contains the NPE stack trace I get with Jonas and the JMS Control after contains the NPE stack trace I get with and the JMS Control after looked into the NPE a little and attached this resolves the issue for my case in and Jonas + +I have overcomplicated things +Is it possible to backport in such a way that isnt broken in the process? I dont know the well enough +Yeah zk session handling +Assigning to for consideration once weve got the release out +Any chance you could start checking in the code already? Would be nice to get more people working on this +This allows a messageSelector to be set +Attached patch fixes this problem and implements corresponding regression tests +The advantage over a single boolean would be if you wanted to detect vs vs nothing +Thanks Dan +Testcase for the feature +Closing this issue +Im not sure +Gavin I assume this can be closed +just which repository are you referring to? reposasf? I dont see apachecon in there +This ones easy +There were some open issues which we wanted to wrap up before merging this work to +I plan to commit this to trunk and later today after getting on the addendum +One other thing I was thinking about is adding another which subclassesWe should be careful here the thought at one time was to remove soon and replace it with one written for the new + +OWASP Java implementation also uses +Do you have any configuration to go with this? Is this using your SEQUENCER config? Please attach the XML config file used then I can have a lookCheers +Is anyone actively working on this issue? If not Id like to volunteer though I have not worked on spring rich before so I may need some pointers +Seems that atomtitle needs to be set to the new name if the cmisname property is in revision +No real difference between Ceres and Isis though Im trying to spread the loadif everyone moved off of Isis then Ceres would complainBlamelist empty is not normal though there is a big backlog of commits whilst Isis was off and so logs on the machine could not record such information +and in thefor exampletrue +Configured following builds to use Maven the number in the setup to there was another port number collisionMight just be transient with an older execution of the same build didnt check +Preparation for Gora release candidate +bulk close for the release +Looks good! Thanks for catching that situation and patching it. +Sure well investigate this further if it fails again still we have a test that did not fail in the last runs and which does exactly the same of most of the other ws tests in code that led to the exception above +check for a way to construct a from a given class this way the velocity writer can have access to any utility functions a implementation has +This patch is going to clash with the changes at andor +This third rev of the patch changes to use the single rather than +Have to be in the class path or in the lib directory of castor +Thanks +I agree but thats not a question more for your webappportletapp server environmentAlso another key question is why is it working OK in one environment and not in the other? +Committed to branch and branch of branch +Ron thanks for your reply +Cool perhaps a minor issue for some other time then +Will somebody try the patch +Here is a version that passes parameters via the Context object rather than by building maps +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +The patch is in process +If I misunderstood something reopen +close for release +Thanks Gary +may in future resolve the issue but Im not sure it will do so without significant complication +We should just drop the hint in this case see believe the more common situation is youre in the process of upgrading to and have a mixed cluster but need to generate a hint for an older version that hasnt generated a nodeid yet +Here is does If the last job has more than reducer and it contains a limit we will add a final MR job with reducer +Canons that asserted results were always integers +What do you think Jochen? GuillaumeAny hope of seeing this featureIf so any idea when? just forgot realized that I dont need to escae as in Lu it could be just Lu +The component is now in richfaces repository +This is the only workaround I have so far +Hi Tonysorry for the late reply +Duplicate of +What I dislike with that approach is that every now knows about the jcr system tree node types +Nor is that URL very friendly +Amareshwari would it be possible to add a configuration variable to turn this offJust a fallback in case we missed configuration with default value of true +Other related todos Need to pgp and md the files too +no tests required +Patch applied at r +Hopefully will be able to look at this more tonight +Yep! then next i would guess it is a legitimate permissions problem in in revision Thanks you robert! We eventually fixed the SVN issue +Is the server om a separate machine from the client here +When will the artifacts be available in the maven repository +sorry my laptop is slow on tests +On the trunk the issue was already partially fixed by r but the code didnt perform proper cleanup in case of a SOAP fault +We need to get all the messages unified in our products so this task for UIMA is going to have to get done one way or anotherThe trick will be selling management on the idea of spending money on something and then not owning it +Eclipse project containing the test class and everything needed +some issues with ssh authz fixed in unit tests +Reassigning all XJava debugger issues +Please update your config. +I wrote the decoder dedicated for to avoid unnecessary memory as much as possible +Fixed in +Hellowe are using Hibernate but we also experience this problem +Thanks! +Please note that there is already dev build for release and you should really report issues against that feel free to reopen if you can explain why your approach is better and what exactly are the issues you experiencing with current behavior +It looks like it should workThanks Daryn pushing this in +We also had no successful since thenSo while I do not think this patch is incorrect I agree it should be reverted for now to work out why it causes these problems +This enhancement will be considered for a release after Struts +Is this ready to go? If I enable it now the site will be exactly what is in SVN +Marek the link above should contain the section now let me know if you cant see it +fix for bad forgot to set the Priority the JIRA triage meeting on December Kevin Cormier agreed to look at this one checking for references to that might be the same problem in other locations in the code also for +dont you thinkI actually mean per read per block +Since this is my first issue reported in the Drools tracker let me know if I did anything wrong in the submission process or if I can provide any additional information +I like the split between installation user and developer guides youre following the OODT documentation model +I have tried all the suggestions above without success +The goal is that all existing grouping test must pass without any changes to the classThis patch also includes collectors for function queries +I live in Sucha Beskidzka +Place it in or classes we should probably not fail if it is in Pete Ive tried that location and I am still able to reproduce this if the is in my folder. +Actually for the record it really isnt a Subversion issue but a limitation of Windows. +verified in +Sure Ill take a look at the patchbq +I like this approach +And then install the plugin +Still waiting for new memory +There are technical reasons and they arent necessarily bullshit its just that not everyone agrees on them +Deadlock after suggested patch generated by testConcurrency unit generated by test generating deadlock using dbcp looking at the deadlock traces and analyzing the code some more I think that the root of the problem in this case is contention for the +Tommaso thanks for resolving all the items brought up in comments +closed with the release of +I imagine there are a lot of single node cluster and pseudo situations where this will always be the caseThanks again nice save +I think the proper solution would be to have a component that could be placed inside the whose job is to the at submission time +Initial version +Discussed with Jimmy +I have to reopen this +Thanks Martin will look into this +If we could run without requiring any jar installation that would definitely remove one hurdle +Dup of +Committed revision +Jars are signed on solaris also. +Bingo! Thanks a lot! +Its no big deal if this drops but we are going to make some boot changes so might as well try to clean up stuff like this while we are at it +These new builds are in can you verify this on the CR build and mark it closed once the right builds are sorted out +same node rows works as expectedS I tried again and doesnt work +Somehow the true severity of this bug was not discovered from alpha though +Wondering if the stylesheets have provided you with enough informationon what is happening here +I committed an enhanced version of this to CVSIts far from perfect and if Joshua can come up with something based on a grammar well throw it away +I will look into it to be certain +to commitbq +Likewise for as skipping bad records is also not supported +The beauty is that if I dont want to limit anything I dont need to do anything and the code wont need to keep track of how many tokens were indexed for that field +I verified that this question and answer were in the nightly distribution. +Please attach a sample pdf +So which test should I extend? Any help on this would be much appreciated +adding script to support Version Roshan! I committed to trunk and. +As Im sure you figured out the stacktrace thrown by the app was less than helpful in determining the real cause. +I guess the tricky part might be that the entity doesnt have direct access to the loaders for all the hydrated child objects. +better point +Indeed lets follow up there. +The Save button should never be enabled +Committed to trunk +This will be fixed in the next version. +See attached pom files and script to build them in the correct order and simulate two different users +ClausOnce you updated the jar file in the fuse maven repo will update its bundle when the CI build is kicked +Hey Nikolaany progress on this issue already? Id love to see some patches even a starter would be great +You might check case sensitivity +I have committed this +Please let me know what you think of this my comment +Ouch good catch! Fixed in revision along with tests also for some other value typesLuckily this is not a serialization bug so none of the stored data itself is compromized unless used as input for copies or further calculations +All passingcorrection All tests passing +Applies to both and trunk atm +This is a early draft of the Yahoo Hadoop tutorial update and rewrite posted for community review. +But that is to generate the resultset needed for delete +Bumping to Struts +Thanks Pedro closing it +This is again as a result of WST disabled facet or better broken Web Dynamic Project Facets +Seemed odd that this patch was removing bloom filters from core then adding implementations to util +And there is no unit test framework in place for the core webapp functionality +If we have to get the exact results we can accept those overheads +It breaks everytime like described above +Is it a special type construction that is arbitrarily present in or does it demonstrate that it can also handle an +I vote to close this issue +Is there something in particular that makes the openide implementation superior to one already living in the ASF? The PipedStream implementation? I am still scarred from some of the hacking I have had to do to deal with PipedStream interaction and threads in Ant code +yes +Yes its fixed +confirmed still a bug in r +buildivylibHadoopcommongriddevbin +I hope it wont hurt the performance much +Any chance of getting this patch committed +Share the filter with the project committers role is one way then ask them to subscribe themselves +Youre right my bad I looked at the wrong constructor in and though they got a null cfId +One of the two is in and other doesnt appear to be +I dont have an example +Closing resolved issues for already released versions +Eran Ive attached a draft fix that I was working on +If people are seriously using a portion of their cluster for analytics this seems like a good costbenefit type of ticket so that their cluster isnt as saturated just because of inefficiencies in output +I am attaching the upmerged patch for review +thats definitely a bug + +Patch is applied to r +Docs are in and we are on a released version of +Issue the upgrade command on RHEL mentionsupdate JMSCHOICE WSCHOICE resteasy update uBut the first command should also include +The fix is to bind the generation clauses after we look for illegal references +There are two basic concerns I have here What is the logThere can be as many logs as there are appenders people have created in the system including application specific ones +pending a lot for for the quick review Eli +Thanks a lotNils +committed the patch into trunk and branch +The lack of forward movement on this suggests its not +Also I should have mentioned that this issue will not affect trunk or thanks to +If theres not enough information there for you please follow up to the mailing list and not in this enhancement requestJust FYI Ive been writing Struts apps for almost three years now and I can assure you that action chaining is bad for your health. +At least make it configurable so we can set it back to m okay with breaking niche configurations to make life better for the vast majority +Ive just committed this to trunk andThanks a lot for the contribution Andrew. +OK so there is obviously some issue with my test environment +Test case is added to the applied at r +I do believe that such errors arent in the scope of +Obviously the fix it to change the C API but that seems like a bit much within the context of fix HBASEs RPC methodWe absolutely require a RPC protocol +Yes thats itSince it fixes the problem Ill commit the fix. +and committed Thanks! +Im on this unless you can fully support the same functionality that exists in the pipeline as it exists today + +Id close this ticket as a duplicateRegards of +Ok that makes sense +for which partial results are also cached right nowI appreciate the work you all have put into this issue and all Im trying to say is that a feature used very rarely should not be enabled by default +to and is a serious bug in with the current fix +Thats why I wrote my last comment +bq +Committed to trunk and +But no biggieWhat about my question about why we delegate mergedelete out to the regionservers? Why not have them do nothing but the close and then have the master do the remove or merging of fs content and fixup in meta? Would that be less moving partsLet me give some higher level feedback in a secJieshan Yes thatll work +Yes I agree it was user errorkarl +I was only making the point that the fact that there is no dtd or xsd at the URI should not affect your ability to validate +Am I missing something +but if you use the uid property it just doesnt happen anymore +javac +First off it does not check if a page is hotcold before reading +its simon +Bug still present in release tried snapshot with no luck +Very simplistic stylesheet added for the XML excludes list +Need to introduce a new RPC interface? can not be used any more +Patch to do just that +I am hoping your approach has a way around itId like to synthesize one approach to using Hadoop based on our implementations and then look to making the whole project consistent in this regard +The user type the enter key at an form field and the form visitor looking for the submit button reach an invisible button the same button at the hidden modal window component DOM that was listed in the submitted form parameters +Thanks so much for digging into this ThomasWhen the release comes out we can add support and tests for the new encoder too +For one part I was trying to explain why you didnt find generics in the current code base and for the other I wanted to understand your ideas betterFor the first part lognet still largely is the same codebase that was started with +Committed trunk revision revision! I leave this issue open so we can upgrade ASM once a compatible version is released. +It is a little harder to run since istat does not run by default which was why I was looking for a checked in solution to running the tests no matter what the default happened to be +Im referring to thetesttestsconfmath and testtestsconfmath files +Patch attached +Commit changes to trunk At revision +the close event I triggered a few mins ago is still not showing in the Activity tab +I think we tracked this down on our side +Surely JIRA allows it? or is it telling us to file a new ticket? Do what you gotta do to make sure its on the radar and gets done to your satisfaction +Not sure what JBDS that matches to +Will commit this now +If you want to use the sub resource locator it doesnt make sense that the serviceClass is a interface as we need to call the method of resource class to locate the other resource object +Duplicate of. +As other VPE preferences it should be separated from the model for convenience and easy access +Solr now responds nicely to +This patch is not applied to +Ive come across a couple of use cases that require ordering of the messages +Thus other users of will either need to configure and use or to use a similar mechanism with the adapter. +It looks like theres some test failures here +Thanks! +Adjusting the priority to reflect the include the version of trinidad your using and some more description +oops +Ive reviewed the code carefully +Yes I think this is indeed a problem +It uses the expanded license that Vincent pointed outThis is meant to be an example commit so Id like to keep it simple + +This is a test case for +This version of the patch includes a way to define probability levels through a standard Hadoop configuration file +Changes are required to patch the examples +Do you plan on declaring a dependency on or shipping a given version of with a la Xalan with Xerces +Id like to have an account too If norman can also create account Ill ask him otherwise bago apache is my address and Im in James PMC tooThank you +Apache Karaf released. +Sorry to come in late the patch has gone stale +Close all issues. +g rev openejb rev Remaining workrationalize the package names Improve the design of and decide on a good location for the and gbeans that I did not copy from the plugin +Patch committed. +Yes I think the same and to be sure we can comment itBut what I dont understand is that other contracts includes links to other pages +Closing. +Closing all resolved tickets from or older +The calculation box does not take in any alphabetsbut it takes in special characters and numerals as the SI is of float data the variable is of string type it takes all the you please suggest me in how to remove the data type validation in my idea is in the when part we have an option of bound with variable similar to it can we have computation on variables in then functionality will be same just the difference in usage +Comments? +Thxs ve committed this +It would be quite helpful to introduce an optional parameter to the java compiler which allows to set an individual java package name +Or introduce ESCAPEXML and ESCAPEXML since the valid charstet is different between the two specs. +I think is more of what we were thinking in trying to move forward the ESB runtime validation bits +your discretion whether to backport the change +Ah a very similar problem was reported on in that case the user had the V two times instead of V two times +Once difference I noticed was that in compactionstats I would see the cleanup and regular compactions running in concurrently but I never saw that in +None that I know +What is your map doing? Do your keysvalues serialize using a series of small writes? Which JVM are you running +Next task is to fix the auto export template logo should be in the upper left corner editReport is not working properly for the home page +Looks like this a left over from the past +This is the source for the code I changed +Patch for Leonardo please review and let me know if you think the changes are good +Forgot to tell that my configuation is different Jboss and JDK problem also occurs running Seam on AS and JDK +Initial test plan for HA testing done at Hortonworks included +java hashtable is using it have no numbers is very interesting +recursion is not your friend +x faster with I have the same orders of magnitude on my box that doesnt have SSD +Its not sufficient because to use the webdav with maven we have to be in a directory where maven can find a pom with the webdav extension +This may very well something thats application specific for me +This is problem is the implementation puts autocompleteon onto the Spinner tag at all other times +I added onclick on the first commandlink of page in webappssimpleand the problem is there +I dont follow +Committed to trunk and resolved +Reviewed by Alex +Christophe yup found that one but I think there are some other version checks in the compiler that need to be updated as well so I will keep looking +Maybe Im wrong and there is a way to specify another destination directory for external resources if so Ill patch the suite to use infras functionalityRegarding the bug with destination is a directory Ill provide a patch +The test was called Failure becuase the deployment was happening successfully even though the war file was purposefully bad +Unless I hear otherwise within the next couple of days Im going to close this case bc I cannot reproduce the due to inactivity +There is now a new build platform parameter that works with the build configuration parameter +So if we try to do like delete and force it to offline still the nodes will keep exisiting in zkThis change is actually very much needed and important now but just need to make sure we dont create open areas +ThanksClaus +Sure I understand +I would prefer to be also put in the branch and not checked into trunk +I made this major because once the crash happens all keys after this one are not created +New patchfile and test cases attached + +Reopen this issue if needed +Chuan thanks again for contributing this patch. +is this simply a of Kamikazes algorithm rather than integration of some release of Kamikaze +Does this issue still exist +its not a bug because we never change the size of the bitsetif you want to add an assertion to that effect fine but making the serialization handle a situation that would be a horrible bug is bad design. +So this field is transient in its sense +Can you map out the issue +Can you the patch? Its out of sync with head at the moment +The one on the web site is subject to change by the marketing guys +Duplicate +Perhaps the report needs to be able to create sinks so would take a sink factory insteadAll of the above are noted at the relevant places in the code +Attaching patch works locally +Here is a patch to add these jars to the classpathrepositoryorgapachegeronimojavamailrepositoryorgapachemyfacescorefor servers +Attached patch fixes the test failure +Please attach your Oracle DDL +This issue is a relative of which was only fixed symptomatically. +No word or info from userperforce so moving this off the fix list +That is not valid +Committed to trunk +Trying on hudson +Fixed in svn. +I think we should mark the appropriate section in as performance relatedThis is more an issue of hygiene and after your explanations no longer a concern wr to data integrity +Please use the uploaded patch file that is about kb +balaji +Sorry for the confusion +Sounds good to me + +Yes people could use that but we would still need to expose some option to enable it. +Added a patch to fix the unit test regression which was introduced in revision applied in trunk tests that passed before are now passing again +Or did you mean that by global property +Regardless of what you call it Camel cant start up in webstart and thats a problem for us +And it has to work with possibly unknown other code in the same db +Regarding mtom maybe its some problem with permision for httpd to create and write file to directory because it worked for me +But agreed we need a better mttr with synchronous lease recovery +Ive committed this to trunk and. +Trying a patch that ignores noauth errors as the Group znode is built up and then always checks for write perms on the final path by writing a test ephemeral sequential node +Uploading a quickstart for this issue +Integrated to TRUNKThanks for the patch DhrubaThanks for the review Scott +It should be done for GA version only +Hi Blakeare you sure youre processing the requests fast enough so they dont pile up in your application? I think thats what was happening to mine although I havent had time yet to graph the profiling data Ive been collecting over the past few days in order to confirm whether this is really the case +Looks fine to me +That would cause a locked record exception to be thrown I guess not pretty or for my short sited comments earlier +Thx +The previously committed class did cause me some problems +patch looks good. +Attaching patch for branch +The codedecode logic will be added in the of the jaasboot module +It shuffles the datanode array before constructing over under etc +Vladimir patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Resolved by a fix in resolved issues whose fix targets have already been released. +Michael just wanna say thank you for doing this +how about it is kept package private for +People please be more careful +The attached patch removes the NPE by setting in before starting slave replication mode +Moderators added +Thanks for the patch Bryan. +Maybe something like async would be better +Thanks Mike after the bug is fixed is now faster on Windows when more than thread is used +No time to investigate right now but interested in a GET dbnamedocidmetatrue output +Looks to be a reasonable solution to me +thanks this information can help me out at least I can think of a temporary workaroundI would really need concurrent caches startup and have Infinispan serialize the requests for the same cache there are more than caches and they are using several approaches they are quite likely discovered in parallel as all of them are lazily initialized by web requestsI know my design isnt the best out there but I expected this to be handled by the +youve obviously not understood the explanations above or Han Hui Wens we do get the dependencies OKbq +Will be part of next release +Would it make you feel better if the pig was chewing the word PIG vigorously +But this is probably as good as we can do without much overloading the workWDYT +Please can someone test itNice RegardsMartin State Saving was added to JSF ant that feature was implemented in +mikes fix solved the issue +Thanks Peeyush +I think I will commit soon if nobody objects +I just committed this + +responded in email thread waiting on data to come back +What would a good default be for this optionBefore we had append as the defaultBut would it make more sense to change to a new default such as Failor OverrideAny thoughts +Adding updated patch as per comments suggestions via +I will try to create a small example later and give it a try +Thanks. +Do you think we can close it +Attaching patch to support nulls as values for bound variables +So please feel free to commentThanks I have finally done up one round of update to the sites look and feel +not sure if its worth it either +I guess I overlooked something +Please approve patch for Beta +Here is the for the test +Applied v. +When it do redelivery it does not know which route steps last time was successful it always start all over againOkay so there might be a bugissue if it works different depending if you have thread or no thread +I know that I need this flag +Patch to include the core name when loading the schema whether the schema name is found or not +Thanks Werner +The same datasource is used by the following quickstarts helloworldhibernateaction helloworldsqlaction helloworldtxsqlaction and simple solution is to provide each of them with different datasources +The pageSize is the amount of messages that are pulled into memory by a destination prior to a dispatch it needs to be large enough to bridge the gaps between messages with sparse selector distribution otherwise messages of one type will halt the dispatch till they are consumed or expire +We used to build this list inside now its done only for a +As I stated above I found it odd because of my experience with Adobe CF +Ill test the patch tomorrowThanks to Jacopo for his valuable input on this feature +Thanks for identifying the patch as the source of the problem +I merged and pushed these patches to both the master and branches. +Hi Another user had this same strack trace +png to iconsidempotentConsumer +This supports router cache +Thi patch fixes a problem with the default connection pool where we were using an Iterator but in the Iterator look we were trying to remove elements in the Iteratorss collection +Theres actually some other circumstances this occurs in +We would just need to block rendering the first page until we heard that discovery had completed. +However today as we dont take into account the initial ipc call were doing two successive tries to the same suspicious node instead of trying one of the others + status from resolved to closed +This used to result in a could not delete dir error while trying to cleanup the effects of the previous test + + on patch +Ok first thanks Leonardo for finding the root cause of this +Update the test to change the transaction isolation to read committed when it detects the Ingres dialects +The one Pavels been seeing are encoded correctly as complex attributes not simple feature attributes +As long as camel uses activemq as the underlying messaging system then it may necessarily have the same issue +Link infra issue. +Thanks! I see that the patches for both and are pushed +Applied at r with minor modificationWould you please veirfy it whether it works as expectedThanks very much for your patch. +Verified in CR +If it isnt too much trouble reverting to the previous behaviour would be preferred +I modified itThe changed ZIP file is attached +I just committed this +Is that correctdidnt know +Please close this isuue this bug report is completely falseSorryOleksandr indicated by reported +The SASL based realms can call out to a JAAS domain and the JAAS domains can call back to a SASL realm. +Simply add an Experimental one some UI magic for visual hints and that should do it +Patch applied +Can you use this only for YARN changes +Stefan is that a requirement when enabling mongodb it wont work without? Our biggest problem with all these is that there seem to be no metadata no docs about what conflicts and what are requirements +This would reduce complexity altogether +Hi Luca sure you can help out +It will be pushed to the and master branches today +Other will hit the for now +I belive the correct fix version is in the link you are reffering to +Re uploading Vishals patch +quickstarts are correctly displayed +The two jars are in the war in lib and in the binary distro +Will add the the sigs and hashes for the release +Is this still a problem? I dont see this test case anymoreThe testcase is in toolsjavatotestsrctestjavaorgapachecxftoolsjavawsdlprocessorinternaljaxwsThe test does pass now and Ive confirmed the generated wsdl does have the required namespace +Super small fix +bulk defer to +second file +Attaching patch that remove the sharing +Or maybe there already is something like that out there +Closed after no further comment has been given +that could be thrown from these code areas provide a hint on what might the cause of the exception. +Cool +Bulk update to fix version +Did I miss something? I also talked to Marcin about it and he said that the MRI version of Oniguruma does not have the same check at that pointIn general we have tried to keep the codebases similar if not identical so that changes over time can easily be integrated +Stack Completely agree +Moves metadata creation from to +Meanwhile I am wondering if we need to add one more unit test case to verfiy this case +Not too many changesDevin +Actually this issue is original one +With the new Maven based directory structure above patch became obsoleted +without static cache it might lead to incorrect usageThe cache cant grow much +Yes I think that would be OK +Bug or known issue is also present in Alpha release +The support for this is already in + +make the codes intent to incr twice clear current get looks like a bug done this morning as part of the patch series +Thank you for the information about the new directory +Thanks for pointing it out Milosz +It also introduced a new SQLSTATE for object permission in general +Should be now be resolved in upstream rbovirt gem +The indexer can simply ignore such nodes +All test cases are Teds comment of removing the unused vaiable an dmoing the creation of OPENED node to +The new enums should have different names and at least one of the default is changing its value in the patch which I want to examine +no patch achieves a pass rate on ant junitreportSo thats a good check that it doesnt break anything obvious +pathto points to the certificate you save in the file I had verified Sailaja setup cloudbridge db is not created +Patch removes the limit fixes the documentation and the test +So having a fully loaded was a huge win for us +There may be issues one around the k message mark and the other to do with the switch between Large and Standard is still in alpha wed really like to see this fixed and released on the branches +Hi LilyIf you have two equivalent queries that only differ in whether they use correlation names the performance should be identical +If this is a JIT bug until a better workaround is available or is released Ill disable JIT temporarilyThanks for your prompt response +Moving unassigned feature out of + +I disagree this should be put in an optional distribution +assumes that all dependencies are satisfied at time ofOTOH a workflow manager like Oozie really sounds like what you want +Tom The snippet of java code and output wsdl that u showed occurs for wrappedliteral +Sample is something very important for us so please fix this issue before do not have the sample in the source tree +Agreed that we move this to +Please assign it back afterwards +Between major and minor +Attaching patch +Fixed in trunk r with the following changes show shared basic project info in versions list browse fix model not being set in repository update on in r. +Thanks Simon +Test failure is due to time out +That seems to me to be a strictly unnecessary pain for both the admin and user that doesnt seem to further Aruns stated goal +Running regressions +This still can get problematic and needs to be fixed +No I guess when we developed this we assumed that wed have separate instances for each transaction but now that you point it out that doesnt make sense +all issues assigned to +Im sorry if this patch got overlooked +Will commit shortly. +package recompiled for Java as a jar so that people using an JVM can benefit from the JDK scripting of course as John explained isnt the ideal mechanism to integrate Groovy into your applications +Quite old out of date +Ondej the workaround describes the migration of AS Messaging to ASHave you checked if it also applies to AS to AS +I just found out that the patch is incomplete +Latest changes seem to have fixed the outstanding issues. +Jing would this change be needed for as well +I would like to commit functional test migrated to arquillian ajocado and i prefer to runtest it before +Hi Clebertno well get EAP ER build on monday +This issue has been resolved for over a year with no further movement +If we do so we should make sure that the documentation clearly says that up to you get one behavior and after that you get a different one +Seems to work +A new AM will be started the client will not show the final state until the second AM gets finished +Marking for and unassigning +I also added a missing Override and fixed some wording in the good to commit +Who are you waiting for information from +appears to have been fixed in the rewrite of the documentation +Hi ColmThanks for reply +The result is right and the master file needs to be updated +Please test with currentCVS version and let me know how it goesThanksYash issue is fixed in currentCVS. +The new patch uses a simplerclearer way of comparing dates +Lets imagine the following two active consumers and six message groups evenly distributed +moving +Camel and below is targeted for KarafCamel is targeted for KarafSo you cannot expect to work flawless in Karaf which is using a major new release of Felix and Aries +Setting this to fixed now after some further of + for the patch +Although Im not involved with your project Id suggest space indent and chars per line +Makes sense thank Jacopo that is too far back to know exactly where at that time we got it from but sure it was an apache projectSince you did not agree to it anyway replace it with whatever you wantRegardsHans +Some screens documenting the process followed a WSDL and a doc file containing the logged errors +Setting fix version ER to verify in the upcoming now as expected +This bundle has no source code attached +Generate a CRUD Database +Another bug that I fixed is fs get dir should create dir +it added +Contributed by Vuk Ercegovac +It depends on how the bug related to jdo configuration will be solved and is not directly connected with this problem +L can be tricky due to edge cases like crossing following the simplex so Ill enforce sparsity with Andrew Ngs bias tweaking trick first +Joakim applied this already +Deployed mkristiannext SNAPSHOT to s repositoryAdded pluginRepository to build to use the sonatype snapshot repository +This is long since done afaik +Sorry for the confusion caused. +I am in the middle of a large refactor so I should now soon whether or not there is a problem +Commit this and open new issue for lease contention? Should be easy enough to replicate in a little harness +The only small change I did myself was to remove empty lines before javadoc tags as checkstyle was keeping to complain about them so this may have been a mistakeFor the rest I have made mainly parameter name changes and removed unchecked exceptions in the throws definitionIt looks like the formatting style I use wraps lines to quickly sometimes +Daniel I think I have found a simpler solution to this issue see rPlease reopen if I missed somethingBy the way thanks for attaching a patch which was pointing me to the faulty code +This could result in an unrelated param being rewritten for some service +Anyone know when a new version of the maven hibernate plugin will be release including Anthonys version and compatability patch? +patches applied in rev thanks! I finally have java working! +The Main part to remove in the new decorator was the header and footer including they are not needed in the new lookup +Added one in with minor changes over Eugenes version +Thanks for your help Doug +Further to set cache to true would require some effort by the or some other type of customization + +I forgot to add that it doesnt matter if updates are sent to the cluster +Greate! you for your fast review +Yeah I know +Youre right this is needed too +Thanks +Passes of the client to the authorization layer instead of hostname +Nobody cant help me +EDIT Ive done runs that did x the data volume on inserts and x the number of update threads and there have been no issues if proper nodetool flush nodetool compact nodetool keycacheinvalidate is done +Same as previous patch except for PUT item implemented SPI created for repository providers allowing this service to be used with other JCR providers with a minimum of coding Split into two projects provides a JAR with all of the required classes +Please be sure to implement server dispatching using wsaAction as per the specIn Section Programming Model +This would mean that if I submit a request with a node the rack it is on and another rack containers could be given on that node or on any node on the other rack +This is a surprise to me I thought we were modifying this file not blowing it awayWe have some upcoming changes coming to the flow of files and other such updates that should allow putting this setting into your +It appears that the commit for made it into +I have reviewed the code changes in +supposed patch which adds a setter for the property I would not call this a bug especially with the blocker severity +Fixed in SVN . +Fix included in and builds closing. +Lukasz I think theres something missing in your new core patch +A number of TCK tests that only test the implementation are now passing but a few more fixes are still neededWhat I need to do is port the old to Undertow and check the changes like the one pointed by arjan above +Its messy and expensive +I say we can mark it asWONTFIX and move onMike +we should merge fs keys and other common keys into one single file +The same changes will be required in add and get methods also +Fixed by getting rid of because if it was changed to that then we would not get EL +Im adding Arun and Mahadev to this issue since they helped create this mess +Incorporate feedback from jd and elliott do selective thread naming rather than blanket change how works +By the way when looking through the schema code I found out that with Lucene trunk it is now also possible to sort the others using the new ctors that introduced +I agree changing the priority is a more appropriate fix +old wizard not supported. +Unit tests pass +Reopen if Magnolia chokes again + Jake +thanks Lars for catching this out and the patch +Want me to post it? +Coding is complete just needs and completed +The initial implementation has been written +It seems it has been fixed please feel free to close this issue +Thanks this was applied as r. +Modified to use +So I guess issue can be closed +Removal of jackson library in ivy committed revision in as not done in nutchgora revision in nutchgora thanks Julien +For my nd blog entry I didnt change the timedate +As commented the workaround is to remove the extension from poms +As everything is fixed Im resolving the issue +Patch that handled the cache misses correctly +There seem to be exclusions in the patch +this will be a big incompatible change for users +there isnt much formatting on the table +Hi JacquesI have just changed the siteName because there are two website having the same name its not necessary to commit as its a demo data +If get does not report missing jar entries how can one list all the entries of a bundle without having to explode the bundle? findEntries is not good enough as it includes contents from fragments. +Dont worry about the list Bob no problem at all +From what I remember we need it in runtime +you probably have a db with duplicate table names in different + and to and trunk +The Failover Transport is working fairly well now the examples will run now and complete without error using the Failover transport +Theres some additional overhead ifthe message is actually logged but thats not so important as the act of reallylogging a message is likely to be much larger than the overhead introducedThe patch is also reasonably large touching a significant portion of all thelogging classes +Please put these files in jaJPJapanese screenshots +it is not failing it tests for ever +I also cant apply this one with patch +How about the attribute name and the type of the valueThat doesnt help much if its a collection and the problem is in an I have a vague memory of long ago writing some utility to analyze a value and help track this kind of thing down walk through a collection or something +Can create an OR cursor over all partitions in the system to implement this btw +Oh yea and Nice! +Fixed on and +This patch makes sure CPPUNITCFLAGS isnt overwritten if one has cppunit in a location its important to have the I flags passed into the compiler +Issues resolved in were backported to adding the latter fix version. +Polling is not necessarily bad in itself I think +Unfortunately teiid still maps longvarchar to string so part of the issue will still exist until teiid +Is this causing any actual problems +That is Graham. +Yep forgot I fixed this one +A FAILED could carry a message from the master like any other such as a +Story Points enabled for all issue types and all projects +hi Jonathan Hsieh I mean region is moving between region server before regionserver getonlineregion +Will be fixed as part of +Seems like the session has been closed while being processed +Thanks ErikYour patch is in trunk rev +I dont think strict XHTML compliance has ever been a goal +Actually in the case of exceptions the sockets are closed and not returned to the pool +HiAny perspective on this really happening on as tagged +Passing to Bryan once his Apache karma is taken care of am guessing that well get additional container support in the SVN tree +Kind Regards TomI understand I should try your fix in my project logging on I wonder how to download your code fix +No actionable details +The test is able to complete without error some times when the test is run after access the default page +So resolving framework extensions eagerly is a desired thing but it should not always be done +Moved to next release the release focuses on the Presentation DSL. +Server allows concurrent processing of GET requests. +Well the desired patch when we get it working right +But I am not quite understand the mechanism of file lock in Windows OSFor this patch does it mean that it will be accepted after has been fixed? Or should I provide another unit test to avoid the case on Windows. +There are pretty much always good reasons why we do things the way we recommend in the framework but those recommendations have evolved over time and will continue to evolve as well and not allowing things we dont recommend stifles this and limits opportunity to progress and improveThat is of course a generality and there is a clear best practice here that should be documented and it probably wont ever change but Im still against forcing on a matter of principle +Ive committed and merged to +Ive just committed this. +I will respond to your emails as soon as possibleRegardsManan +Hi MarcoI tested with the correctedsimplifiedupdated above and it workso May you confirm this point please +I couldnt find an issue in your code either so i finally assumed it has to be my build messing things up +s map function gets called to load the status for all the services and host components from the polled data +Will it be released soon? If it is going to bea while before is there a chance for a for getting this fix outin a released Tika +It isclear that the time to the timeout is the deadlock time NOT the timeout time +Everything compiles though +Also please add information on what is not current +duplicate roles listed in roles page this view is deleted +From your description and patch I see what you are getting at +I think thisll mean introspecting for existing RT field cache entries and addvalues to them as documents are added +In commercial road network databases these attributes are typically formulated in relation to the direction of digitisation +Currently FFIW only throws exception if the filesize is smaller than expected otherwise the file is truncated +Just one question The could be theoretically moved to the in or am I missing somethingyeah good call +Should the commitpull request reference this issue or should I create a new one +Hi Shazron +We might be able to further improve on this by exposing some of this functionality from +Im barely not good reviewer but trying to be like a committer +is for the changes +Policy Attachment will be provided only on wsdl at Port and Endpoint the linked issue for other attachment point implementation done +I hope this makes sense +Im going to attach an updated version of this patch when the issues and are resolved +Would adding a flush on the command socket when sending the progress address the problem? It is already limited to second so it shouldnt cause excessive load +Looking into it uses a at the moment eliminating all n before compiling +Tomcat works fine with the same setup +It passes now +Thanks for committing the patch! It looks like one file was accidentally left in place +Possible regression +Sharcene Larpen Larcene +Change the second wait to a second wait and add timesout on all tests +That looks ok +jQuery Mobile should be testing their tab bar not us +If you think that the issue still presents then just add a comment + +Cant reproduce using distribution generated from current trunk code +Tag +Due to this could not add back the tmp dir creation in +Im taking a crack at cleaning up some stuff +I incorporate Stacks suggestions hereCorrected the sleeping logic and also add one more test for the static callJenkins is also green +Another way of doing it is to create a file format that contains index along with data +Moved the placement logic over to the skin +I fixed this as part of which required also other changes to the quorum patch is attached to the jira +Closing all resolved issues with a status. +It seems like at least the warning should be issued if a database with no authentication is accessible through the network server regardless of local or remote host +IMO the need of the embedder was a result of trying to hide the previous that exposed too much implementation detail +Site moved to after release +This issue is also like resolved by fix for and can be resolved again like as s duplicate +Lukas dialog in your screen is a multipurpose one + +Sorry it looks like I dropped the ball +It fails in +Needs unit tests +Its still good. +Closing issue. +Uploaded +no feedback cant reproduce +But is unknown to me +The JOIN FETCH expressions ensure that immediate collections are fetched +Ive committed the file in SVN rev Im leaving this issue open for now because I think it would be good to use the example component as the example for this and replace the label properties files there with the XML file +I want to clarify my Issue works with but not to install correctly on we have to dogem uninstall v install v supports on my github fork of it +from the jstack it looks like hanging on release which is the issue fixed in +I would suggest we not engage in API renaming inBy all means proceed in if desired by the communityIm open to other opinions on this but thats mine +Any other comments +Please verify +Happy to provide more info if it helpsThanks for looking into it. +So if you say it fails in both servers what does that mean exactly +With this patch all CTF tests finish fine +trivial free to make some changes I dont intend on doing any more for myself but Im happy to see improvements made +Closing as duplicate +It should be OK. +The RAT plugin is not happy anymore +patch that passes hadoop common and hdfs tests to fix the javadoc warnings before at the javadocs improved but no obvious reason for patch test failures diffs dont show any new warnings +Confirmed +Standalone was meant to allow you to run all roles on a single host without specifying any masters or slavesPatch context and Mikes comments be damned +Ofcourse there is also a place for things that are common to all three +committed to TRUNK patch for MAINTENANCE attached +Patch committed +Maven may be unable to directly produce the modulesdistexample output directory in the same way but thats a different problem +we do need to either remove those messages or change them to debug especially since this is a client library +Will be available during next release. +The last proposal before minProposal might not be minProposal +Going through it to figure that out is part of the ticket +Close all resolved issues for Engine release. +Nick Now that Adobe is heavily investing in the HTML world they are trying to become framework agnostic meaning that Adobe does not endorse any particular framework for building applications +Code you please provide some overall design approach +Did you mean to make similar changes in as well? I see an import of the pooling client in there but no other changes +Never mind resolved +I reproduced it here is the full log is because the projects logo is not defined +Ive got it working without the findClass implementation +Later messages from the sender associated with the blocked thread will not be delivered until the blocked thread returns but the pool threads will be free to do other work. +Cameron I hope youre still aroundNow that we are actually getting a release branch going soon after the branch I would like to get updated +Change the server we use then +The patch is ok +There are some discussions on that user might have existing gsh based customized scriptIf we dont provide gsh in geronimo these script will be broken +Tom by convention changes made by the users are set in whereas is used to list the parameters and their default values +In my case after switching over to the SDK and changing the applicationDPI property to any components that I has not already set a font size for appears in the mobile app too small +I meant to cancel this patch until it logs rather than use a user counter to record the information. +In short it definitely can be done but it requires more than just yours truly +Hi Kishan removing both of them except which you can decide if it is useful? Thanks. +Humm +Committed revision . +Applied patch +oops last line meant to say +Ill look at that later tonight or tomorrow as well as the docs +Test class to reproduce the error +At the time I submitted the patch I moved forward with the version that I submitted because I got little feedback from the original author. +Committed revision +I think its time to admit defeat +actually only expected results +You can see the same problem message +Im told that ruby is different but I dont know enough details to know if thats the whole storyIs there a chance you can drop back to a ruby and retry +bq +Refactor code to add script +Or better yet use the weight feature of the cache to limit the amount of actual memory use instead of just the number of cache entries +Yeah +Fixed the path handling and correctly sync cancelling a job in revision Added a test the above changes introduces an incompatibility as I changed the return types of the cancelJob methods from void to boolean +Pushed to production +Unit tests are in committed the patch by Saeid +The problem here appears to be that FOP sets up a font witha platform specific encoding +So my issue is how does this service know what is authortitative about some resource? I gave a URI about someone on the Web and I get a times more information back than he has in his foaf file +That seems to be the best compromise +if the new btree approach is checked in should add tests to look out for bugs like whichcame about the last time a new but similar btree was checked in + is what is in use currently +Call completeFile before stopping renewing the file lease not after +Sorry for thatAnd for if the failed disk still readable and the node has enouth space it can migrate data on the disks to other disks in the same node +The datanode persistently stayed in the blockmap after it got shuted down +Maybe someone else has some opinions on the other formats +PatchPatch applied +WDYT +Resolved as a duplicate of some other problem as suggested above +I think this ticket is duplicated with +Wonderful +indeed i was testing on trunk but not but i have not tried to test using to see if it fails there +Patch enables me to run the junit tests under JMECDCFoundation again +for you delete the TESTROOTDIR in tearDown instead of deleting output directory in every testcase +This method also checks for TGT refresh time therefore just calling relogin before every connection should be a reasonable fix +Once I updated the CLI to I couldnt reproduce the issue +After more investigation these two may be newly incurred errors +The pull request should be merged into and cherry picked into master +Diffs look good +The signature of would need to be changed +The used in the script +Its submitted now +Manually tested +Does that really workIt investigated the Type or better the possibilities +Its simpler to have all the datanodes in a secure cluster running via jsvc +I have to enable them in the trunk and have yet to figure out how to enable them really +It should fix the issue +Is there a spot thats not happening +Yes has fixed this issue +Tried on both line and line +The initial implementation is online for requests should be handled as new JIRA items. +verified on +I agree it looks like a different problem +If it is set all Javamodules go there +I will create a separate issue for the error message +Committed in r + +The perf test that uses also passed which makes it even stranger +The SOA team wants this to be considered for +Since Spring is final now and just around the corner we wont be backporting any new features anymore +The one you pointed out was a missing log line from the results +I am able to deploy the hibernate example on tomcat with jboss embedded +Not sure if noggit can do this yet or not but looking into it +And this was done +There must be something strange happening under the hood +But I am not sure about the conclusion of that conversation +So the check for getter methods in must use equals +But when I built it I got errors about missing files +This alone will help with the memory issue and speed things up dramatically +Isnt the responsible for URL encodingdecoding? This is a tricky issue I think there are two places where encoding should occur in the and inside the code related to construction +Probably it was my mistake or it was fixed in trunk I cant reproduce it nor manually nor by using integration test +No problem I can do that tomorrow +Thanks for the patch. +I can then swithc on automated building and all is guess this means there is a simple phasing problem +Our hosts are running RHEL with KVM +pushed to master +Nasty hack would be to implement a thread local cache for type names so that at least we would only pay the price once for request +Trunk do not suffer from the same problem due to previous code changes +Command line options get processed beforewithout buildfile so its not practical to have the buildfile expand themConsider these two buildr help buildr helpThe first one cant use the buildfile so it can only show you the same options Buildr supports everywhere +is added which contains info on each stream as opposed to doing all the initiate stuffTransfers Source maintains the list of files to stream in +Replaces writeRaw uses in Test but leaves a usage in +The inheritance from the parent pom bundle plugin config isnt intuitive at first but it works well and is handy in cases like this +Therefore this is not a bug +Thanks Alejandro for catching it completely missed thatUploading a patch that creates only when HA is enabled +You could eagerly exhaust the original iterator at the point where the was created +sorry did not update my copy before changingI dont see the Results per page am I missing something +Wanted to discuss first +A small project in our department is encountering this and I have not had the time to really investigate the real cause of the problem +If you need access to additional methods either use those classes directly or implement your own class that extends +Attached an updated patch +Committed trunk revision branch revision close after release of +Looks they are coming from +It might not be good idea to change fscks files option now +That way we can eliminate the special casesI think that having the check is a really useful piece of functionality +The few that failed passed when running them separately +needs to be reviewed as well +Is there any danger of a memory leak from empty stacks in the threadlocal from discarded threads +Some example would have been more beneficial +Ive released that the attached patch to run tests against EJB needs tweaking as I changed the class names of the providers as discussed on the forum thread. +not fixed +After further testing this issue seems to occur on all + +Thanks for contributing +Please help to review it thanks +Need to test before posting review +Comments from Andrea +Just out of curiosity have you made sure that you arent in debug mode when running these tests +Why is it there anyway as nobody of the developers seems to use it anymore and it may be +ouch good catch you are digging pretty deep I assume this is coming from the original code which was mainly made for CVS back then +At least it is documented here nowDue to my tests however I have noticed a problem with the marshallingunmarshalling in +I still havent finish a complete successful run in yet but I want to share early so folks can help testThere are still couple of holes +Committed both in and trunk +lognet starts reading the stream based on the streams current position +No errors thrown as well +I hope that Andrew or someone familiar with FOP and the way that Derby builds the docs can take a look at this. +Prevent the caret on the menu from being displayed +Thanks Lars for the patch and for the unit test +. +Ioannis can we remove the dependency from karaf itests +Anyone object to the scriptdelta? If not Ill commit this to both trunk and +Updated patch to be applied on top of latest patch at +Is easily recreateable from deployerWhile geronimocar appears to come up after any deployer start command the gbeans inside will only activate if the configuration starts with the serverFor example if they are started with the server a stop followed by a start will not bring them back up +We should set the timeout very short to make sure only very short periods between a service restart are covered +Fixed by Neil Delima +Technically has dependency on in test scope +Can you add to too? +I did no notice that Si proposition was not working this way +If ip alias deletion fails ip alias state would be marked as Revoked in nicipalias table +Im working on it +If the directory arg is not given it uses the current directory +Could the wiki page be updated with this new option +Do we need any consensus to do this? Would we break any existing users? Or isnt that a concern with this young project at this point? adding new methods should not be a problem for users although it may be a problem for people writing products that extendIn general I think that adding new methods to the published interfaces is something that we should do deliberately +I dont know all of the deeper implications but I think Jacks summary and a couple others comments would make the most sense and are already what seem to be described in the current documentation which states that the boolean operators in front of of various terms or phrases define them to be specifically included or excluded while the others are classified as optional +This page should mention that the scripts will set up their own classpath for running the tools but to call the tools directly using Java and not using the scripts they will need to set the classpath +Might get pushed into Vladimirs changes to master and the branch +All good. +After some investigation and consulting the portlet spec I found the root cause for the +Patches forimplsrcmainjavaorgapachemyfacesadfinternaluilafbasexhtmlimplsrcmainjavaorgapachemyfacesadfinternalrenderkitcorexhtml patch. +I set the priority to high +Thanks Andrey +RichardCan you provide some more background information on how your flow interacts with the various portlet modes? Id like to understand the big picture about your requirements here and when edit mode is appropriate and when view mode is can you provide stack traces for the andThanks feedback from reporter. +Escalated to ve tested the new commits and everything is building fine on my machine +Fixed test +If in of the plugin I have the fix for issue MRELEASE then I believe this isnt exactly the same bug as Im still seeing the behaviour others have outlined above and I have a project with inheritance alsoThis originally all worked fine in of thePlease open a new issue +SanjayI havent looked at the changes yet but the features are pretty neat and very useful +As a workaround one could either add the missing information to or process the FDF using an own classprogramHow dependent are you on really usingBRMaruan +It seems like theres still a question about the impl here +This patch is currently tested by the unit tests I dont see a way to trigger the bad case given its +Reopen +It exists in prev releases also + for the patch +Thanks for your contribution +asc files Im working on getting the mirror set up +I think I misunderstood on irc +Can you provide more details what you are doing in the valueThe Input components show the submitted value on the next request +It still blocks sometimes on your computer? I havent been able to reproduce the issue locally on linuxubuntu +Thanks Jonathan +Sean Did you see Pauls and my conversation on the dev list? Actually I dont like to add another specialized facade but improve the runtime behaviour of the facade itself +There shouldnt be a need to order processors any more +I submittedClosing this issue. +If the code itself is confusing perhaps you could use this jira to add commentsfix comments +If there is a problem please open a new jira and link this one with that. +There could be some kind of coupling between the css and the js such as code relying on particular css class names +This is complete as far as I can see +Otherwise Ill go ahead and commit it +I cant tell what failed + creation is removed from build Thanks +Fixed in +thanks Stu. +this fix dont handle the already generated metadata files +Specific issues will be tracked separately. +It appears that even suffers from this but to a lesser degree +People can always do the thing externally if they need to +Adjusted the patch to match the input provided by Daryn Sharp + +Committed as r +Moving to to future and not of JMX is planned for Camel is the use case for thisUsing a local jconsole we could expose sendBody requestBody etc +The frameworkName and framework are used to populate the generatedAs mentioned in the discussion on the mailing list this can be moved into +Man I remember when i could buy your vote for +mahadev pat any more concerns or is this ready to go +Thanks for the patch. +Verified in JBDS betaVersion id v date +Daniel Can you include this in also +Most of the methods have a parameter into which the user is supposed to pass an object of type which is supposed to be the information contained in profile xml files like much of the code was using these methods incorrectly and was passing in null for this argument +but we can turn on the option in the +Currently the code will warn only +Probably should be configurable but I didnt go that route +Patch for RTC +no evidence presented of an actual problem +There was a problem with using ephemeral ports for the property which got fixed after an svn updateNow the task is complete with all features +Closing for now release notes will be checked with ER. +After talking to David Illsley we decided to move the real operation detection into its own handler +this patch replaces the previous one +Here are the background images used in the decks +There is no longer a version attribute in +Unfortunately the moment a server starts the current code attempts to connect to jmx simply as a courtesy to the user +Otherwise it is deferredThis will mingle well later with +To maintain compatibility in the scope of the GSE not in any other compiler related stuff the GSE constructor needs to set on its private +Patch applied thanks +Might be better implemented as an aspect for a service that really needs it. +Yes the documentation wasnt very clear there about the implied extra configuration file that is needed have updated that portion of the documentation to be more explicit now and provided a link to that section of the documentation that describes the contents of this extra configuration RC issues. +If a directory becomes bad then we increment the timestamp and write the new timestamp into all remaining good directories +Namit I added the test cases and related changed +ibrowse does not unzip response bodies but its simple enough to do this ourselves with a call to zlibgunzip now unzips response to self or whoever implements this the should change depending on what is used +Can this be solved by adding a Delete button on the new security page +Screenshot of projects Java build path showing overlapping source folders +Patch available in should fix this issue as well +Your project does not have a publishable folder +Both and passed +The triangle seems to be cuted at the shifted down. +I think were good to go for since this was unintentionally added earlier with. +kont edit +This still occurs with todays SNAPSHOT +not tested on pda in patch thanks! +I believe the test of the previous patch got lost in the queue the build moved to bigtop +Viabhavplease attach your XMLXSL files rather than cutpaste as comments +I dont think either issue or patch is valid +I agree that it will be good to have this extra tip to prevent confusion. +This does not search the output area +So Ill work on the documentation next +You can see this by inserting a small sleep just after sending the second message in this case both messages will be placed on the brokers queue and dispatched in the correct order +Arquillian tests still fail as their jars are not available but rest should compile fine +This will not be a generic API though since optimistic locking will have no such concept +Please refer to the text in sections and for more details. +Hi MaxI did look over the code I checked out and I did see the HBX fix but as I said and realize the CVS build did not work at runtime but knowing the potential of people checking code in and some issues that could result I was not surprised but wanted you to know of what happenedLet me know when a is available +This patch adds one line of xACxDAF the Korean syllables range to the code +Use to track this issue. +Implemented in AS +Sorry Gagan I missed the jira email notice on this issueI havent required java for anything so I would prefer to stay with java +Please provide me the new Dialect implementation or direct me to where I can obtain the samePlease note that I do not need the latest Hibernate jar but only the minimal fix for this error aloneThanks in advanceVenkatesh +What ends up with the most clean controller remarks was obviously on the patch +Trunk r branchx r +Ive added this script as so just use grails to use it +There is no readnonblock onThe version of rbuffill used sysread which is defined +This is quite an important stuff for us becuase all of the parts of our system use grails melody plugin for monitoring and jobs are quite important to be monitored +if saveNamespace fails automatically enters the safe mode +Ive edited the issue a bit updated the issue to reflect that its no longer tied to Google Summer of Code minor cleanup increased priority to Major please increase further if necessary. +Bilgin for the patch +Or one could consider the base class with a complete implementation +I should note that I confirmed the mapping is in fact not duplciated in any way +Feel free to reopen it if you can provide a simple project +Thank you for the input +Theres absolutely a lot of interest in this featureHave you checked out? It seems like a dup of this issue but Mark just uploaded a patch for trunk +So if the field name is name then the getter will be getName this ofcourse isnt true for your case +Committed to trunk and +I was going to post to the list about it shortly +Thanks to Martin Sven Timo +Closing all old Resolved issues +The field still needs to be annotated or tagged embeddable in xml +Oh dear +New patch just cleans up a few minor things +This should fix it + +Teiid has resolved this issue for ER +Please review the patch and let me know the feedbackRegards looks good Gagan Thanks! I will commit it tomorrow evening need to head to bed now +It appears that to get round this problem antgant will have to eschew using the Ant parser and return to doing a full parse of its own +Awesomesauce! Im going to merge this back into master now +built against the latest Dave declaring an interface solved the problem. +This should be pretty low on the priority scale unless Im missing something +Thanks Vikram! +Also there is a difference between cmp and equal +Add a flag autoCreate to know whether or not to create autocreate Node +If it were a test failure Id say otherwise +Im not sure this issue is valid anymore if we changed the behaviour then someone else would raise another saying blank values should be assigned as disagree +Yes is not working with distincts +I cant reproduce it works for me every time +Any way try again with version and if you still get same error send your project to Axis user list +Please reopen or even better create a new JIRA for new contributions based on the latest Tuscany code base +Oh BTW +Id also like to see info on the Maven version being injected as a variable +So what would be the procedure to add some clustering code beyond carrot or other available librariesThanks and Regards GrantTill now I have worked mostly with full document clustering +I can confirm this bug +committed as r +But with this patch any user can run all modes with option ant optjetjitint testand testing will report passed while exclude lists for jit mode was restored +made me want to fa venir plorera noms de pensar en ellaIt makes me want to cry +Needs to be considered how we are messaging and documenting this kind of split before we implement itMaybe wait to do this in +Well Ive just discovered that it works sometimes and sometimes it doesnt +It seems that its not yet possible to place an order as described in Placing an Order in docs +Yeah I created the file manually from xcode so this is an unlikely error yet still possible +No that makes sense if the Thrift server is being singled out for this address and we still want to use the other as the default for servicesShouldnt this include an update to conf and the test conf? Even if the element isnt added a comment should at least be added to conf to direct people how to use the parameter since they will likely not be reading the codeFunctionality works as expected +svn commit m move to the correct package +Can you reproduce it in our example appsIs it a JSF RI issue possibly +Seekable and Syncable I think these are hence Public +Few API changes to improve extensibility +set commonjstcore as component as discussed with pushed the PR to master +Well I reported this issue almost one year ago +Here comes the patch +Patch applied as of nightly build . +bulk defer to +Thanks +I just noticed that gradle still uses ASM should it be updated too +Attached patch for this issue +Im hoping that ifwhen we shift from the existing graphical framework to Graphiti this will be less of an issue +Bumping to as Avro needs to be updated along with Hadoop dependencies therefore if separate work needs to be contributed to Avro this will hold up this issue significantly +bq +have factored out a protected execute method to allow for overriding the handling of failures +I see the same thing +Attached patch that resolves this issue it actually would like to add a simple test case that will test this +I didnt realized the first time around that this are shown during the selection of the project I could see the benefit in being able to addredefine the description field but when would the content of those fields be different between examples? Adding a Fix button would be what the user would be presented for anyway when importing the project rightMaybe also reword the facing problem segment to This example has some requirements that could not be automatically configured +This issue is happening even when the data present for insert record which is in a batch with some update records +then hitting Remove deletes whatever Transforms or Validators are selected +so that users do not get confused by the name +Thanks Myrna +Fixed support of these attributes added to VPE +Can you also try creating a new workspace and place just a single project with this file in it +enough +Please try the updated PR +ChasThe issue is that means about squat in the real life +I just pushed a fix for this to the branch. +For example Class C has a class B that has a class A +I suppose we could change the KV comparator so it looks for I thought you had trick to make +AFAIK the garbage collection is only run when explicitly invoked +Derbyall only has errors which I see in a clean client wisconsin sysinfo sysinfowithproperties xa and a new failure in caused by main Exception in thread mainTest FailedI have committed at subversion revision +TODO Deprecate all classes in package +Hi Jarek +Cannot reproduce the issue +In any case I managed to force it through with this patch +I agree with you +This is fantastic work excellent! +Then it replys back the jira with the test result +Sorry for the inconvenience +Hence this JIRA wont be fixed +Committed +I dont see these errors anymore so I think it was fixed by r. +Build is currently broken +Can you add me as appropriate? So that I can see the applications I mean +perhaps could we reuse mavenzilla when it will allow to browse repos and not just search them. +Here is an updated patch with the test fixed +Duplicate of +I thought it did +Please add a test project so we can incorporate it in the integration tests +Thanks Ning +This doesnt really solve the problem because the second that an AOP logger is used which is when an AOP class is loaded all projects will begin logging to the console +dataTable filtering could be turned of by not specifying filterBy. +INFO Running for rows took ms +Patch to remove the dependency +Hi Alex +Attached patch +Issue closed with release. +Rattaching patch after synching up with Dyres change forFix for Network Client should not print token separators in message when it cannot connect to the server to retrieve the error messageThis patch changes network server to send the complete message in its own locale instead of just the message arguments and delimiters if the exception is severe enough to terminate the connection +Ill go and figure out how to teach that one little test to direct non of its output to tmp tomorrow though +is out of the door and this has as targetMaybe it should be changed? +It may be related to XML config. +New patch to accomodate changes in +The problem started when upgrading from Spring Source IDE to +This causes missing jcrfrozenUUID and others properties in child nodes of frozen node +Hum I just noticed that the issue description is not complete not sure what happened when submitting itIll edit it to include more details +Initial rpc proxyprovider test plus code fixes to get the test to initial patch is committed +Okay fixed in master available as for now +This should be fixed now +To be perfect they should be in but since we cannot touch I would go with +I doubt if the v patch will do what it intends to +We are planing for a revamp soon so dont beat yourself for not understanding it +A project should have its own email lists tests branches etc independent of Hadoop mainlineThe main advantage of putting projects in core seems to be to increase their visibility to the Hadoop community +fixed for branch at rev +Thanks Lahiru for the patch. +I was thinking to emphasize this new beginning I thought it worth setting it apart by making it the first inhabitant of our rightful namespace +This isnt quite fixedFurther testing shows that the widthheight attributes for both are essentially the same +Modified patch with corrected a typo in the unit test +That is confusing Sorry about that +change bookie mbeanss domain to rThanks Sijie +KinichiroThanks for that additional input +Note that setting does not change which compiler we use only which libraries the compiler uses to check that the methodsclasses we use exist +is there anything I can do you require from me or rian to action this +Also supported RHEL or Ubuntu in page we dont test on this OS there is no specific build for these OS also so better to remove them from the docs +Now that work is well underway with AS all previous community releases are +It also fails in Windows with JDK +Would it be more appropriate to have this potentially dangerous method to have a package private visibility and then add a public wrapper in a test utility class +I am closing this ticketRegardsgood luck +There is a problem with opening java lucene indexes not with searching them +The commit for this issue removed the minute timeout and increased the scope of impacted connections from those open for more than four minutes to any open for more than one minute +The other option might be to only enable this by default under Hadoop +Fixes for both +should commit +Just one line change +maybe ndepends cql could help see if this heuristic would work +Query resolving at design time still needs to be aware of all models regardless of visibility +It only affects Namenode front page and dfsadmin report +allownomoduledescriptor is too long IMO so we choose allownomd which is far from perfect +This is a misconfiguration issue +for the patch + failed perhaps due toUpdated patch with a more concise comment +You could also just deploy some dummy release versions and NOT release them +Definitely shouldnt be some random operation from the wsdl +Arno I do not have the ability to close this issue +See The version isnt contained by EAP ER repository butNow the Project Examples checks if there is the artifact with a version containing redhatThis fixes the issue with recognizing the Maven Enterprise repositoryHowever the archetype still contains the version that doesnt exist in the EAP ER repository +Since yesterday one of a shards replica has one document less than it should have +This is consistent with requirement that the be immutable +Yeah youre probably right +ok where we got to with this the last comment still applies we are deferring the magic parent version discussion until later so for now it is still required but the tools will update it version will be inherited to avoid having to specify it twice if you are working in a single version m opening a new issue related to the release plugin refactor. +Shouldnt the code resolve the URI first then check the cache for that URI? If not then any URI which is actually resolved by the will never be cached and that would certainly be a bugWhat to you think +I really think it is not very hard to do inline +please attach the wsdl and reopen the issue +Martins patch for this issue +I inadvertedly supplied the wrong file +The parsing of comparator arguments of the previous patches were too naive and in particular didnt worked with schema migrations +Id prefer that rather than a big bang change we implement this as having an alternative that is not default and work to pull the logic out of the struts actions xmlrpc where needed to be in a single placeWhen we feel the new UI is complete enough we can switch over but I guess therell have to be a call about some features we want to dropWould you agree +fixed in CVS is not a CALL join execution join point if needed +I added a TODO in the moduleBut we may want to consider these API changes again +Please verify whether it has resolved the problem of nio direct byte bufferGood luckLeo +So how do we create something better? My suggestion is to split the current maven into at least three sub repositories that are exposed via a proxy +The patch doesnt actually load the tables +According to the issue there should be an option the for jms backend to participate in XA or have I got it wrong +If we forbid this syntax should I close as invalidThanks +The important thing is that the deadlock problem is solvedIt turned out to be that the only way to get rid of the warning is to remove the verification itselfThe problem here is that people keep using configuration class as a container for passingparameters between methods which is a bad practice +Please close. +It seems that most login modules produce both a user and at least one group principal +Fixes but the core tests for ZK fail I think because other tests are expecting a get call to connect to ZK and it no longer does that +Please first post such errors to user list so that the Jira only focus on confirmed issues +Should I just rename this block to Scheduled poll consumer to make it clearer what I am talking about +I recommend that you land this on master and if we discover no issues in the next couple of weeks we can consider merging on branch for a future release +The better change is to actually instantiate and run within but before running pass its Configuration conf property object to the child job +Patch that removes a lot of meta and root stuff thats not needed when +Should we handle this from the client instead +Thanks Lars! +Any progress +Nathan Dawson from WCU is working on this issue +An XA datasource is required for native XA enlistment +Wonder if that is coming into play. +Rewording my comment from in order to capture it here as wellI dont really see a proxy for shell which would need its own kerberos credentials and would have to perform its own authentication of clients as core functionality +Start eclipse with workspace +The object it returns from lookup is of type Remote while if Searchable extends Remote the object returned from lookup is Searchable +Im glad I didnt start when I said I would or I might have wasted more than half an hour +Thanks! +Whats the advantage of the chart component implementing events instead of simply relying on nested ajax behaviours? Are listeners that common a use case that the user benefits from the cluttered API +mvn compile sourcejar deploy should be used instead of mvn sourcejar is applied but were still to rebuilt release +Please update the jpa spec tree with and push a version jar to the repro +I cannot hit this problem again with CR build +Your instincts about what to do with sound right to me Kim +Created an interface to Demo Streams returned based on tag +The endpoint override is not yet used in WSDL mode as it is assumed that the user will just update the WSDL fileA CXF config can still be specified with the benefit that it can now match the WSDL rather than using the dummy teiid local namesWhen used in a dynamic vdb all of the soap operations on the serviceport will be exposed a procedures. +Dan you can leave the merging to me +To reduce the levels of indirection I made sure that all methods directly call the end creator method that is the method that actually creates and initializes the store +However there is still work to be done in order to be able to map a relationship through a join table with xdoclet +You would expect an Invalid target exception but the response is that therelevant message bundle is not in the class path +Thanks +We can use the existing assert functionality in junit +Or were Both producing the wrong results +I agree with the changeHowever it is backwards compatible it will permit more programs to be legal not reject programs currently legalSo move to to the frontend for implementation +The wordings in the examples seem strange to me To configure an existing unencrypted database for encryption use this JDBC URL jdbcderbysalesdbdataEncryptiontrueencryptionKeyThat doesnt come across to me that its an example of an URL to encrypt an existing databaseit reads more as that is the actual URL to use +Patch applied with revision . +Either way this identifier is usedrequired by modcluster and should be made consistently availableMod cluster node members need to be able to reliably know their identifier to recognize the vhosts and contexts that apache has associated with it as described by get for example +This issue duplicates which is fixed now. +I have created a security manager based on the manager disables deleting a file executing a command and setting a system have added validation could you please review +I was just curious when planning new sprint for Beta +Committed revision +sfoxfix to branchThanks for the patch HimanshuThanks for the review Stack +But then we can report the issues exactly on the line of each We work on the bytecode we know for sure if a class is an exception or not but we cannot log the issues on the lines where the fields are ve opted for the solution for the time well be able to improve this rule one the symbol table will be available from AST checks to get rid off the approximation +I like it too +I still think it is better to have seperate macros for unit and system test instead of clubing everything together +There is an issue with startup that I need to look into. +Another iteration over Akshays patch +One of our tests is failing +Ill see what I can do about setting x permissions +Fixed +This is not a problem on Oneiric +The newandnoteworthy label has applied +Paul thanks for checking that +Chirs has implemented a wonderful exceptions mechanism for so we can now sign this jar in its current position +It fails +Local Guvnor Repository Resource Decoration Preferences and Chapter +Current implementation relies on timeouts and that is a workaround I suggest reRender content of modalPanel instead of component itself +Seems to have passed when I tried to regenerate the it comes up again Ill let you know and be sure to grab the surefire reports +Thanks Good catch! Patch applied +I can reproduce this bug in but not in +Add Add +also without authorization for access to cached layers it possible to grab maps that restricted for public access in geoserver +its close My suggestions would be to tip it degress and not make the center part look like a document it is normally more though of as a box or tube that these providersconsumers send messages through +SVN and notification pointed to the attic +Thanks SumitYour patch is in trunk at revision +Actually I double checked the released version and I see this file being deployed so I suspect its some issue in your environment +As I see it this is an improvement to a feature which makes the feature more usable +So just like this type of filter was not really supported and no longer works this method of relying on the to support one id space across the whole index no longer works as well +superceded by the sdecorate templating +Needs to be fixed at the Portal level. +So will work with by default? Or is it a different mvn profile +JaredIm sorry that Ive missed to reply to Jimmys input +Committed revision +Hardlinks would also be a very easy way to implement copies +You already have it in thereGood stuff G +Its already in status +uuid and its variants are a very bad choice for a surrogate primary key generation +Second patch has been attached after incorporating Donalds comments +Changed the default to current time and no tests break +It works fine for me at r +This file includes a simple test case showing how to use decimal output parameters with iBATIS +Thanks Jose! +The code formatting has changed a bit between the versionsIf you decide to do this use the CVS version of the STRUTSBRANCHafter I did the back porting for a project I was working on I moved some of thefixes from Struts to +Closing xml and console issues. +I am however open to a better way of doing this so i am all ears +I am changing my request explicitly to before I call +Applied patch at r +What other plugins do you have installed? This is probably a jar file conflict +No additional source code is whole slingscriptinggroovy project +Ideally the default rule would be in the database and would show in the UI but be grayed out and not editable by the user +There arent any durable subscribers for Queues but for Topics +We could add a component property to load the class manually but that is getting a bit much at that point +That seems to be an adequate patch for yS +Prepping +What a nice birthday presentI will check out the fix and see how it works locally +Yeah got distracted for a few weeks I will come back to it early next week +It will find the instance and add runtime and Seam runtime +Hi DarynThis was discussed in and +Andrea I will need approval for the geotools main partsI will probably need Justin to approve the WFS partsI am aware that these patches are massive and I expect that it will take you some time to look at them +Patch committed in +Please excuse my brevity. +However the source code is packaged inside the jar files for the release +I have no problem with the entities Rick +Fixing this might require some coding to overcome whatever that limitation was +I dig in Solr code and there is classloader set up in +I just had this error when doing a sitedeploy +We should at some point test and be confident Java works +Timothy could you check and see if the fix for improved the performance +The test failure doesnt seem to have anything to do with this patch +Tested +Have you confirmed that taking out those filter mappings works? How does the servlet container know how to map a request to restmaintenanceclean to the +but rules things out one after one +Added full list of defined attributes to documentation fixed XML typos linked to schema for more details of attributes +Here is a test that follows your recommendations +you need to enable compile of those classes +I just committed this +Thanks for the report +Even shorter version of patch +Closing old issues +I have refactored the test cases into since at times as a whole the test case timesout also since the Role was added to dependent JIRA the test cases where refactored +Ian +At least for +A user wanted it on user so now it is available +I wonder if there are still some issues when using activemq for the queue as Zach is usingAnyway please provide a while you see the problems and attach it here so we can have a deeper look +It relates to the timeout change +Im fine with Owens suggestion of removing Chukwa from +rob forgot to merge second part of fix into deb to more time +Opened new issue for Lars for +If this does come up we always have the workaround of the user supplying their own +I thought above you allowed that values could be byte arraysYou have setMeta and then set? Why not just name the second method same as firstIn your implemenation please keep in mind that others will most likely want to extend +This is then released when we call terminate +Updated patch for current branch state +Your code is violating the design of the interface youre implementing +Meaning to stop the analysis after a certain number of cycles was found and a given maximum amount of time has been spent analyzing +Id like to see support for Groovy in my favorite IDE +Apache mandates a common header that includes a notice about the Apache Software License +No +I think the comments here are a bad sign +Figured you would know enough to comment on this +This way we can make sure it is configurable for different environments +The method deleteUser is also affected by this problem +I think people should create new JIRA issues for other ideas around this changeI shall open an issue for the Solr as a standalone java app +This kind of question will annoy the developers and encourage them to ignore or close this problem ticket which nobody wants +recently Ive updated the test to catch tomcat +I can not reproduce this issue with tools beta it seems that it has been fixed +repeat should indicate a duration like +its not fixed in itself its fixed in JDK which we run the tests against by default +I may have this issue as well +verified closing. +Patch comming. +Thanks Robert! I have added the flexscoring branch to the Affected Versions field as well to indicate that this whole issue belongs there. +Patch applied on the branch +Ill try this version next week and give you an update +yes i am aware of various ways to create a new uri scheme in camelA file in the adding it thourgh the API via a routebuilder as you described andalso by defining a component as a spring bean if I am not mistakingWhether creating your own uri schemes is good or bad depends on the use case I guessIn my case it wouldnt be that much of a problemIn any case starting the endpoint url with a routebox scheme or without to me isnt the biggest issueSomething I am concerned with is the verboseness of the url and setting the dispatch keyThis is a very flexible approach however it might be nice to have some kind of alternative way of specifying the logical endpointin the routebox something which is more intuitive to non technical peopleperhaps something likerouteboxbookstoreroutesaddwhere bookstoreroutes would resolve to a routeBuilder or a camel context located in the bean registryA default dispatchStrategy could then dispatch the logical endpoint name to a corresponding directseda endpoint in the routeboxThis reduces verboseness quite a lot and makes it more managable and productive for the non technical people in my opinionTo me this seems a nice compromise to satisfy the use case I am targetting as well what do you thinkBest regards thoughts on my last comment? Claus KristofI have added a warning to the documentation that the component and options are subject to change in future releasesPlease let me know whether you like my recommendations in my previous comment to eliminate the dispatchMap and Strategy and instead use routeIds to indirect to routesI am also planning to use package scanning to look up Routebuilders +should do its best at helping developers catch configuration errors +Most occurrences of IBM have been replaced +has been marked as a duplicate of this bug +Geert what incorrect comment do you seeIs it around To add repositories to the default ones prepend to the list of repositories to addPlease describe your thoughtRegards for not being specific +modjk is released as stable on March . +That looks like a problem of the JMX console +Sorry Im going to this Issue + +If not a proxy classloader of the supplied classloader all classloaders of the interfaces will be usedAs this is a generic solution I will add it to the +No it is definitely a bug +Currently theres no process in place for JDO to install a snapshot but no objections from me to do soIf the maven scripts andor build process were updated to include How to install a SHAPSHOT release then committers on the project could build and install a SNAPSHOT when neededVolunteer to write or update the scriptsprocess documentation +I dont see the issue +Moreover the method youre referring has been introduced much more recently than what you think in beta I think +thanks mark it works as expected now +So all issues not directly assigned to an AS release are being closed +Trivial patch against trunk +FYI Building under IBM worked for me after I added as indicated in +Siying can you create a review request? Ive a couple of comments as well comment in line is not complete comments in line can you give an example in which case all input alias are sampled by syntax but it actually cannot be sampled? line if we want to mimic the old estimation algorithm it seems we shouldnt bytes here? It is added in line right +I think will be enough for cases +I guess I would tend to sort by the displayed names +bq +The issue has been localized in the code +Resolved in verified +antlibI didnt have this issue with previous versions +Committed to trunk. +Wasnt it just the other way around? In the info bar option is enabled by default +Anyway that should be straightforward to address respect to lazy loading We strived for making use eager loading consistently to avoid the need for an +Thanks for reporting the issue +Closing all resolved tickets from or older +Bumping out to +Thanks +Thanks for the response. +There are at least two minor issues with using Tika in Jackrabbit +I submitted a patchIt passes mvn clean verify +Good catch! It looks like it is also using qualified paths in the tests +Assigned to enhancements release +Scripts were just some hackups I checked into bin until we were able to get functionality into HBA. + +please use this patch insteadRich idea your patch is in rev +Since the backend teams arent going to be able to do it the spec now lists this as a limitation viz +SPI has been released fixing. +Thanks Andrew! Committed revision +Listing this as a blocker for first release +Assigning this to Christopher for additional testing +The problem is gone in for letting us know Ill see if I can track down the change that fixed it + run Failures Errors Skipped Time elapsed Released Closing +Can you get a clean build with this patch? I have fixed most of the licenseinternationalization check issues but some tests are still failing with +clearing should be done by the Tokenizer or other Whew +What do others think? Is it likely to break anythingIf not can we implement this? Test case for the problem +That wont be an easy to fix however there is a workaround that you can attached which declares required depndencies including xerces as jaxb plugin dependencies +Its not needed since it just duplicates the individual +Thanks Aki Yoshida for the Jira and the patch. +There was a simple logic error in figuring out what type of anchor was being created +Without M option Pig converts paths to their absolute locations before passing them to the loadersstorers +Thanks Bertrand interesting stuffSince yahoo calls the parameter callback should we call ours or do people think that name is too long? On the other hand wrapper function is more descriptive about what it does to the actual response +Thanks! +Accounting transaction associated to Payment Application should reverseThanks And RegardsSumit Pandit +i agree with all the above +Seems the CXF SMX ones dont can add descriptions to the ones but I will need someone to commit them + classify +I cannot run against Sybase until is resolved +I attach the entire with resend you the correct it +The patch looks good for meHowever this implementation cannot evlauate characters +Attached patch implements changes +Will integrate the patch to trunk and branch tomorrow AM PST if there is no objection +In case of looking for services like say httpd within the then a different logic needs to be applied which is missing right now in the framework +Its actually +Thanks Ashutosh and Deepesh +Not opening one to track txid through all the commands yet since its not clear its necessary +Work in progress +Thanks +EG any access would have userfoo prepended to all row keys +I was tempted to write the general case because I remembered it once was a problem in dependency management as well which in fact was why we didnt introduce meclipse at that present time with newest plugin version I dont have problems anymore except for the written checkout case +Have +defer all X issues to +If you call Debug AsRun Configurations +ve just committed this +bumped to critical since affecting docs working on this so assigning to in openshift central and tests test referring to in Developer StudioVersion Build id Build date +SAR items and creating a packaging configuration for it where is the OUTPUT of the +It requires the script file to have the +I kept this support for compatibility +FYI this problem was encountered while integrating the latest Harmony into Android +Thank you Jing for the reviewIve committed the patch. +Robin is your new web site ready to go? from what I saw its more than good enough for deployment and I assume it has the latest information which would fix this issue along the way +The reason is OSR bug +This was implemented in and went out with +Sorry about that +fair enough still the question persists is it a regression then or something we need to change in our configuration? +Please provide the output of traceroute to narrow down the location of the problem +But remember that the repository example in the Getting Started project does not currently run +There is no silver bullet to recall or relevancy nor is an advice to use OR +I think Claus saw this one on windows too +Please ignore my previous comment +Fixed long all ResolvedFixed issues lacking a Fixed Version +listed files + +Note the proposed patch removes dependency +Thanks for reporting +Here is what I did Build the axis from source +Closing as per previous comments +These blockers and critical issues are resolved but not verified +Fixed in the latest SVN +casts to any alternate implementation must extend from this class +Tommaso I think there is still a glitch here that we should try to addressIn a single node if you ask for a that matches no documents you get a group back that explicitly says the numFound is But with your patch that group is left completely out of the responsethe attached patch has a test showing the was hoping there would be a quick fix for this but the more i dug into the shard result merging logic for grouping the less i understood itI went ahead and commited my tests with a workaround to ignore the comparison of the missing between the single node and distributed cases so it still demonstrates that the distributed query doesnt fail but it ignores the fact that the responses look diffIve opened to try and deal with it at a later dateCommitted revision Committed revision +I just committed this +Applied Phils sources to CVS thanks for submitting +Committed to and trunk +Another quick commentThe in work looks fine except for one problem which well need to solve somehow +Please check that the problem is fixed for you +replace with +May be you should leave a comment on +HelloWould you please try my patch? ThanksThe patch fixes following problems +Hi JorgeIt seems youre using an old version of Ant +Thanks for review. +Now the test avoids the XML datatype if it is not supported by the test configurationCommitted at subversion revision Touches the following filesM javatestingorgapachederbyTestingfunctionTeststestslang +for trunkId like to have some more discussion beforeif we consider this forAny thoughts about that +Those are on a different repository which should also be in the search pathNevertheless we need more information to troubleshoot the problem +hum a naming strategy is provided by the user through a property + +The changes have been implemented by having in keep track of what events it received +No thanks necessary +I can now report that there are nooutstanding problems against bug in and +It was fixed in the commit for +Fixed configuredatabaseusernamepasswordAssigning filename as password string +A nice enhancement to the HR component where you follow the Datamodel resourcebook closely +Metrics test failure is unrelated to this patch +It not ease replicated and seems related to WST +I need an example to replicate +If you read this please send me an email Ive lost the last one +Committed patch as svn resolving +Without the processor you dont get a? Have you tested if this is model based? What I mean is whether this only occurs with once particular set of domain classes +This patch should fix the problem +Tom a tip is that you can use Help Report a problem to get a file generated with relevant log files which you can attach and if Smooks code is using the error logging correctly we will be able to see the full that file is not appropriate for you go to Error Log View and you should be able to get the specific stacktrace +There are changes to the following filesM srcdevguideM srcdevguideM srcrefM srcrefM srctoolsPlease let me know if more fixes are needed +Ill load this in over the weekend +Regards ScheperDisclaimerThis message contains information that may be privileged or confidential and is the property of Sogeti Nederland +I see your points +upload a am not referring to +I was not able to access SVN contextual menu and in particular Resolve Conflict menu now I am not sure it is due to mevenide integration maybe its due to the SVN integration +Actually I think we should discuss this one with Max Andy and the Mojarra guys before having another proposal to see in which direction we should work +Thanks +The biggest sample set I have run was maps and samples per mapJob Finished in secondsEstimated value of PI is I just committed this. +I created a simple test case using the existing harness and deleting a Widget causes an error for both JPA and JCR +Confirmed with Spark theme issue doesnt occur but with Halo it does +If a failed task is not detected until the reducer fails to fetch data from it that is fine with me but if the AM dies or hangs or if there is some how data corruption I really would like to avoid those +Of course execute mvn sonarsonar +on the change +Thanks for the patch. +Yes i analysed the log and i have two calls again +Web XML Validation looks likeshould look like +is a Maven dependency +This reflects current trunk +Thus to enable for a single interpreter called use Or for both and interpreters useIn all cases the must be in the main Apache configuration files outside of any Directory Files or Location directives +Reviewed by Gordon + +Updated with the correct spec dependency +Bulk close for release +remove the tags from xenserver and restart the mgmt server +Thank you Shawn +this is build issue +I hope you dont mind. +No further information provided +started failing with after this commit +Cheers Toby +It adds a config param +I want to keep them separate to ensure Create snapshot operation can evolve to have more fields +This indicates to me that polling should NOT be attempted +The majority of use cases for this are covered by either EJB or inter deployment dependencies +Join Column Attribute Targets +the looks like a bug in the event system of candi +If worked with just instead of its own Path class then it is absolutely clear on how to construct paths and when extra escaping is necessary because SunsOracles javadocs are very clear +Applied patchthanksdims +Many tests have been improved to prevent timing failures reopen if problem are seen on next release. +It should be out in an hour or so +Maybe Im doing something wrong? Please tell me wheres my mistake +The problem is easily reproduced +Thanks +That is probably the better thing to do it addresses and other costs associated with keeping files open +Can you use examples that follow the actual model structure? That is server is not a child of subsystem and host is not a child of server +commented on issue is improvements post that cause transactions to force to disk +Local file system seems to be good enough + T E S T SRunning run Failures Errors Skipped Time elapsed verify the fix in the latest build +Hello Would you please try my patchThanksBest regardsJimmy +ldapsrcmainjavaorgapachedirectorysharedldapcodecsearchcontrolspagedSearch +I was surprised to see the compression improvement over is marginal +If you create a database of some initial size then insert data around as much as the database size you get the above error +When you work in a team where everybody uses ARC telling them that were going to disable it is not an option +it seems like we need to add some sort explicit of mechanism for the node +ok thanks for the answersWonder if this is something the other WMSs would want to catch up with +I was able to workaround the issue by deleting the partially written snapshot believe the exception is being thrown while reading the snapshot and the partial transaction message is not an indication of what is causing it to crash +entries Stop reading run gc through jconsole +Committed simple patch to override the two methods + +Please close it if you dont see any problems +The ping task is working fine for me +If you examine the contents of the zip file as I did the +Thanks for the feedback +Committed patch with revision +javadoc for pointing this out SethJuergen +To allow empty value for a custom attribute which has Directory String syntax +changes committed belatedly in rev sorry everyone. +Thanks Ryan! +Once we see the real overhead on Replication worker nodes will start work on this +The pure ejb profile doesnt include the packs related to clustering. +venkatplease increase the priority on this onethanks. +The patch doesnt appear to include any new or modified testsPlease justify why no new tests are needed for this patchAlso please list what manual steps were performed to verify this patchThis was tested for both ibm jvm and sun jvm +Please verify +Ill send you something to look at +You think GB would be a better starting point? It would save us x regions on the runaway tables +Modified document to take into account discussion thus far +Bruce had to move the connections part of the emitter to a different thread to allow Listener initialization +So we decode as as +It contained a new line in one of the timestamp literals and since we are no longer automatically removing newlines it was invalid sql +pg The description of the DEFAULT has an extra and in the first sentence and the +Please if you are able to provide further details on reproducing this bug or the solution provided in the comments did not solve this for you. +Should we close this bug as Invalid then +I have compared them and found this one to be superior in terms of match to current +At it has sufficed to just continue with processing the request sending a +Maybe the pattern change should be a separate issue +If you step back the help is wrong how can you check user permissions for a table without the table name? Sure the patch fixes the NPE but it doesnt then error out +Thanks a lot Scott +Any comments or improvements +Thanks Gert I have updated the release notes to include a notice about this rename +PR for master merged in master branch +This will be fixed by. +I just committed this +Maybe the Mapper should be called and this one +I can think of a patch that would suite my case add all requestparameters to the key calculation +Forgot to mention the attached is based on a change to the remotingGA code + +So all issues not directly assigned to an AS release are being closed +It can also be used to add back in classes that would normally be also have a very real need with Scriptom to include stack trace information but the Scriptom packages are beneath so theyll always be omitted +Keith could you review this change pleaseThanks had thought about putting it in Broker so I guess you doing so too means that I should have +tweaking client POM to have and only +reopening due to concerns about workability of solution in the lifecycle +I think I filed this under the wrong component +quick note that we cant use as thats only legal in response to the inability to fulfill a request with an Expect headerI think is the correct code but it does expose the loss of atomicity when using external attachments. +Thanks +It is a minor improvement +very simple fix now passing the test +Thanks for the commit knut +Len I have made these changes as you that the HTTP Gateway information is correct +Thanks Mayank Bansal. + +looks goodI just committed this +You also cant disable the table since it wont offline the regions it thinks are still pending open +I see this regularly too +Thanks +This issue has been fixed in SVN trunk +Committed +defer all issues to +It would certainly be interesting to have some sort of restful login service which could start as a community module for example +Should we just revert the change for now +Postponing again until the maven assembly fix is available +verified in JBDS +While I had not expected that behaviour I now see why its necessaryreasonable +Closing again +Ah yes indeed this is fixed +and zones created +Hi JacobI tried changing the virtual memory size and run scomp util and found that I got the same results +StefanoOT The Geronimo implementation now works pretty well +This is a major change especially in the area of backward compatibility with public API +There was a problem in Metamer +Bulk close all issues. +Consider the issue closed somehow I could not receive yesterday the latest snapshot +The work done regarding to assumes log per file +Clarified the comment about is any connection that is not in a JTA transaction. +Opened to address not passing of options rather than this issue +I guess Felix and Equinox by default expose a little more than strictly necessary +Explore on a remote file opens up the Remote System Details view with the appropriate folder in the view +I this its safe to close this issue. +Link to master JIRA tracking security for java routines +Can this marked as done and closed or do we have any more work to do on this? +Ill have a go at it when I have a spare moment +This issue has been solved and can be closed +THANK YOUBTW the difficulties I had in applying the patch were only due to differences in the whitespaceBest +It completely avoids procesing of because it is not always possible to force XML document authors to use two tags for empty elementsregards Tomas should now be fixed in SVN please verify. +Gang scheduling is an interesting feature but it is not the highest on my priority list +Bolek can i mark resolved or do you still need to look at Murrays questions +Alfresco cron expression for Solr is seconds though +We could also leave the key in console if you prefer and do the same the other way round. +many Thanks. +Skin pregeneration service committed in to Trinidad trunk in r. +I think message is fine +htaccess file with the redirects +Can you fix the test issue withAlso not directly related to this ticket but can be parametrized just like +A is not +Can anyone handle that issue? Id like to add some new content to it +One problem here is that embedded ZK currently expects to start on solrport but we dont know the solr port until weve parsed +Forgot to mention the bug is that a Hadoop Path of form filehomeelimrtaskTrackerelijobcachejobjars becomes a Java File of form homeelisrcmapreducefilehomeelimrtaskTrackerelijobcachejobjars +Correcting effects version since is the current target of master +Netscape is lying to you +Added a test case to the test suite for this. +Ill do some digging into what places we can do caching by default in addition to the section you mentioned +Please review +Attaching +yworks pinged for a v! See Xircles. +A few more comments +Why not just issue a WARN and act as a when there are no conditions presentTwo reasons The users intent was probably to have a condition not having one indicates an error in their code +is the of the clusterWhat I meant was more of a SW organization point of view +dot graphs with corresponding html should be pretty trivial actually +I dont have any strong feelings about whether this should be or not +Thanks Pratik and PranayPatch is in trunk rev +Fixed. +You could also remove NOTE This may be null if the pattern is an empty string +Is the proposal to remove appends from all versions of Hadoop or just the versionsAll versions +Kafka is not in Maven do you think we should upgrade to beta version +The beanPrincipal is now synched with the incoming call principal by calling get so that its correct for the duration of the call. +I would like to receive feedback from Arun Alejandro Sean Deepak or other developers and committers who are watching this Jira so that it can be committedThanks +change applied +wasnt working properly within macros when the local scope was offWILL +Ive committed your updates to SVN +Closing this as Wont fix in favor of andPlease reopen if you disagree. +This is interesting yes sure post a patch and will do a review +OK installing Teiid Beta runtime solves the problem so its no longer a blocker for me +We should be ready to commit thanks everyone +The reason why this is problemmatic is because the will only enable the okay button if the status is non error +Committed fix +Please could you confirm that they have been applied as expected? Best regards George +This seems to work +Its your decision to say its not a bug because we dont care much about the embedded installations so if there is an issue then the users have to solve their problem by unique technics or you can say there was not many requests so far but obviously we want to make it as easy to use as a standalone installation and change it to something nice to have further improvement +Another IT is in order +This is the patch containing the solution +Great! Sorry about being slow on my responses +classpath +OK its coming back to me nowN is a newer addition to MRI and we havent implemented it yet +Do I have to close the issue? What is the process +An alternate resolver based on the ibiblio resolver +I did not see any deadlocks +Ive committed a patch that allows the file not to be read until it is really needed +Removed Items +There are just too many possibilities if someone screws up their subnetting +Ok just to give a short update it looks like we really trigger a bug in the html unit script engine here the good news is it is not the prototype inheritance per se which is present but something else which I have yet to nail downFrom what I can gather is that I am loosing values at a return statement which I should not +Can this be closed +Thanks for patch Ted +Good thank you for verifying this +This issue was solved when was introduced +Realized that as soon as I created the issue but would something like this be a natural fit in the framework be complimentary to the + +I added and +Is high standing in this sense really standard English? I dont believe Ive ever heard it before +Alternatively if you really want to do this you can extend and supply your own version of +But we lose the top level reader test +Hi Richard Refactoring changes committed in revision + +Hi SebastianI would prefer to continue with TSCCM refactoring and bringing it to a somewhat consistent state before starting with the goldplating +An EL resolver has been added for resolving messages from the flows through an implicit variable +lgtm +Problem is relying on the fact that the first query parameters is the one that will correctly resolve the +Daryn I want your opinion on this should we turn off functionality based on useip configuration? That way system that does not use hostport in token service remains largely unaffected +The tasksInited will not have been set to true because the throw happens right before that is set and thus the metrics wont be properly decremented +Latest snapshots enforce this by looking for the JPA annotations on the entities +Seems that setting the max frame size on the proton Transport to much more than mb causes issues either in proton or with the JMS client code +We should also fix in this issue +should work now as youre right the export of deutsch lang works as expected +committed w additional changes to get it to build and breaking out the cleanup pieces into separate commits. +Grails generates a file. +extract the selenium tarball in frameworkimageswebapp will review this before letting it to broken CSS path in js fileThis supercedes the previous just tested +One question is in the air how to split metawidget into frontend and backend jars +to commit +Ive added a second patch that removes that dependency entirely I cant find any reason to believe that this is an actual dependency +Even if you do hit such a situation and norms are loaded twice first for a clone and then for original reader it is transient and norms are really smallish compared to other memory hogs so we can live with itIm nixing this sharing +I think that classes under are meant to be generic and be usable outside of particular WIAB implementation +Trivial patch is trivial +The most recent patch looks great to me I just committed it to CVS +Prevents you from creating these manually +Apparently it has been already merged into earlierBut is hasnt been ported back to +Christian please do not fix issues that I have rejected and closed +Find attach the source code could you answer to my issue? cannot test your Page because of the hibernate dependenciesBut you set your to required so perhaps you do not fill in the value? What error does the form print? Can you test this with +Just pushed the changes to and trunk +Apart from the which should be replaced currently the enhancer is also abusing theWhile some of the are being used correctly +Could this issue be +The upstream fixes will be picked up in the next rebase +I find the idea valuable since these samples are actually integration tests and dont have much value as real samples +One alternative is to check for memory freed after the initial GC and if it does not free more than some X MB then turn off the initial GC globally in the +The changes will better some use cases while worsening others +try to resizeThe tree view resizes and is too big so example details could not be seenThis is reproducible always on linux I cant test it on win machine right now +Either way it is better for Designer give that option to their users to choose streaming or not for performance or choose true by default +The goal of this ticket is precisely to allow to configure it +s additional readwrite paths require more reasoning about how it interacts w the underlying system +Awesome thanks for the quick fix Charles +Patch for trunk +Attached are jars built from JPOX CVS HEAD +Any thoughts on whether this affects + DPI is the one mandated by OGC for historical reasons defaults to the default DPI instead +Steve final generic instance methods work fine in C +This will also affect how plugins invoke facets from other plugins down the line +The existing functional tests generate their own configurations to prevent clobbering known instances +Not tested yet +Getting Hadoop to build and run reliably on Java is a prerequisite to thisMaybe we should create two s build and run on Java everywhere and only after that adopt specific features +These can be removed if the committers would prefer they were not included +patch rotting here +The following counters were added to the UIMA AS Client meta process process meta +Thanks for the problem report! +Intermediate Notice your tasks are running too slow and increase +Ill fix tomorrow +bulk defer of open issues to +Missing type parameter checks in class headers fixed in r +Well address it in +patch committed in +it will create a fake table name and put all of the regions for all tables into that map entry +Here is an updated patch that includes everything from This also introduces in solr so that the use the new API with flags set to generate the valid bitsI still need to get tests included for yeah is only the smallest positive number +If you find further issues please create a new jira and link it to this one. +This would have the additional advantage of showing for protocols other than HTTP +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +PatchThere is a threading problem with the proposal that I definitely need to think more about +This wont show up in the static docs though until they are published again +Hi VladoXvfb metacity did its job but we now have a new issue tracked by +a newer patch for patch +Updated patches should work +NPE is fixed by +If not a patch against the latest SVN trunk will do just fine + for codemirror released! +I havent tested this myself but it looks to me like this is resolved? +Ran some benchmarks +I think the code for Store has changed since then and this issue doesnt exist anymore +Updated patch changes documentation a bit and includes the same in help message as well +cleared patch available flag since all patches have been with types branch +I just think that we keep relying on the crutch of providing a option for things that shouldnt be issues in the first place +That would be the best approach for sure but isnt something we can do overnightLets not do one to the exclusion of the other +you are talking about a user error you should be able to show what the user is doing wrongThe compilation command is exactely following the instructions of the Axis pageSo this solution is really a very weak one and doesnt proof the quality of this serviceThanks a lot for this extensive help +I dont think this is really a bug +Want to take a crack at it +Ill be committing later today +Actually I notice now that in webwork the parameter was actually named loc so this modification of the parmeter is probably not intentional +Knows how to do the call against the server +This issue will crop back up when we do any sort of generated source code like xsd or wsdl +But it may still be worth having the inline comments and produce the HTML outside of the regular build and pt it o the website. +Because it use modal window name as a criterion +Thanks so much! +Please let me know if you have any issues +This is a duplicate of +In fact already takes a boolean to decide if it should rebuild the bloom filter so it seems we only need to pass true if we detect the bloomfilterfpchance has changed +bq +Patch in diff format submitted. +However win looks like having some problem +Thanks for the hard work Brian +I have just changed the priority of log messages which will reduce many log messagesSo not attached any tests +perf advantageMaintaining a large system such as Solr is hard as it is +This is a clone of the original for tidiness one jira per fix version +Its also possible to put the HC itself into state via some operation that affects the HC config +The use case depicted by my is quite differentI dont define dependencies within a profile I define classifiers that are used to point to correct dependencies that in turn are defined globallyOr is it that the two features share common erroneous code fixing of which will fix both issues? Please pardon my question but I simply dont know the maven codebase +Personally havent tested it yet +Yes its running with the insane jars now +what file system loader code +bulk defer to +Anyhow I am not sure why the name is used instead if distributionManagement is not defined in the POM +attach an example that reproduces the can also see this pretty easily if you have this relationship and you have grails generate the controller and views off a domain +Please verify +Fix applied to trunk andThanks Andrew +It seems valid that in the case where the transport is not connected the receive and the timed receive as well should just return null to indicate that theres not currently a message available since in the timed case you could end up waiting much longer than the intended time otherwise +Committed to trunk +Dprojectcamelsvn commit message Fixed aggregator not working with onException in a routeSending srcmainjavaorgapachecamelmodelSending srcmainjavaorgapachecamelutilAdding srctestjavaorgapachecamelprocessoraggregatorTransmitting file dataCommitted revision And for branchCommitted revision +Mike Brocks commentTwo war files are shipped with the platformIn the standalone version we ship with the unsecured uploader console by default +It has to do that to honour the contract which is why there is a zoom in on the failed item +Ok reflections not needed +There was a brief discussion of a a couple weeks back and Id left this open pending that +Heres the patch rebased to latest trunk +Utils has been added to the component. +How are you using it? Seems to work for me +Temporary workaround is to generate the html readmes locally and then check them into source control +Using it is difficult and requires complex handling including exception handlers to handle arbitrary file systems +Removed Android so that this bug tracks one unit of work +Deletion added +What is the status of having this patch reviewed and commited to the next release +Verified working against and +Sorry for the mess but the v had some unintended changes sneak in +Fixed as rev +Text in the bug report gets word wrapped and consequently the xml is garbled +Im seeing problems that dont Castor JDO error messages are correct as they stand in of close this. +oh easy peasy lemon squeezy then +waiting for feedback please reopen if this still persists +This patch fixes the problem and adds a unit test to ensure that it works +We wont be fixing this feature since we are not planning on supporting Groovy +Hello allIm very interested in this feature whats the latest? Has any development taken place towards it yet? have not had any time to look into and will probably not have any in near future either +Merged with latest trunk +Let me commit this +Resolved +Change the pattern in server editor to index +should deploy bin include lib nicely in to deploy directory +To do streaming on inserts would be hard +Ill fix it in +Committed to. +I might be misunderstanding how your packaging works +alg files +The lifecycle processor could handle that +file replaced with changes made per Belas suggestion +So if any of the codec libraries is missing we set the corresponding variable to +agree to closing this issue +Fixed at r +Patch for yahoo not to be committed +I dont understand what exactly you were changing there +Fixed rev . +Defer all open issues to +Now we have concantrated on CSS class dialog +Originally I did this work for the terasort record and was since ported over to +That section of the documentation says You may force Hibernate to quote an identifier in the generated SQL by enclosing the table or column name in backticks in the mapping documentIt actually does not refer to SQL formulas and Hibernate has never provided explicit support for thatNow I can see the argument for allowing backticks but that is a feature requestI take it that I can reject this bug report +Doing that is more to set up a system to ensure that things dont break down the road I dont have any concern at all about the current want to keep this JIRA open until that is done +Im not a DN expert but detecting the incorrectly sized blocks is probably something best left to fsck andor the block scanner +We have debugged the same problem within our internal code and once the throws clauses were fixed that issue has not been seen in over a year + will require something like this +while not blowing them upIs it too crazy +has called get unnecessaryI have removed this code. +but Boos targets file ignores the properties +Bulk close for release +So the Brazilian translation will need to be updated for that fileAlso note that UI label XML files typically have their keys in alphabetical order +This is the version adapted to trunk +Ive reverted r on render module to fix the geoserver build +The only point of the bind user for me was to allow for querying of the roles under a context that the user does not have rights to +This patch is ready for review and implements the goals specified in the desciption +Updated to apply on trunk as has gone in +Thanks Tim. +Cant quite remember why that was done +The idea here is that the default is to create and when this happens the default security group applied to the blocks all inbound traffic +Fixed in trunk and branches + doesnt support position fixed that is required for this to scroll more of a bulk update all resolutions changed to done please review history to original resolution type +The changes in so far introduced this error +shows the result of modified library which adopts the above mentioned modification +Changing signature implies more changes +I have some time to investigate this +See also Also is this a JVM specific issue? It seems you are? Did you see the same problem with the Sun Runtime? I am reluctant to add this type of caching based on problems of a single JVM +Added debug stuff to some test cases to find typos +Im gonna go ahead and apply it based on Erron and Jarkkos reports. +Provided the API to fetch the online controllers and the total controllers in the cluster +I have gone over the docs thanks for the great work + +Looks good to me +is the query plan for the query that returns rows wrongly I believe +Hi heres a updated patch based on Jareks comment +which reduce the likelihood of this race condition between a rollback thread and a worker still isnt a solution though since it relates to some fundamental design flaws in pessimistic locking which we hope to address with MVCC +reopen as the javadocs are to blocker i dont think the prettify directory should be in as it breaks the build +Looks OK from a user point of view +Verified as fixed for EAP noted you verified this as fixed for in the last comment +A possible use case you bootstrap a node it immediately dies before running cleanup +Good catch Carl +Added a note on the fact that these are yet supported to the docs page +Committed. +fo testcase +Committed revision in trunk +I am thinking to use the pure Java CRC implemented in to get the performance benefit +Candidate for +where is the test case for this? Lets not addfix featuresbugs without testcases please +This bug is Firstly the kb deployed jar with the this seems to happen when there is a blank space in the directory where JBDS is installed +Thanks for looking into this though. +on linux fails because it cant access Palatino font +Not sure what to do for the beta but I am reopening this issue to put it back into debate +I first disconnected project from git to verify so I thought that the UI freezing behaviour is because of missing git repo +This is a severe issue +Thanks for tracking down the initial problem Chriscommits r after release. +Will look into it +Cant reproduce with latest Mac OS Java and Groovy +Wont do unless there are more user requests. +Note that both testHammer and test fail in the most recent log +Same patch for trunk +Patch to remove the nonExists cache for gives slightly better guarantees than read committed +This error occours with glassfish runing in port I have the same problem when starts webcontainer in port diferent of +There are at least two ways to serve COMMIT asynchronously +The wrapper log shows JVM appears hung times on June times on June times so far today on June This is with r from the June upgrade +Rough cut version +I plan to commit in a day or to to branch +Couple of dependencies are not yet pulled through ivy they need to be copied in manually after applying the patch +Ifyoure still seeing infinite loops provide a test case andfile a bug report +Should the error happen there too then it is most probably not realted to the gnome window manager. +Closing this issue per the reporters request. +now it works +Than you have to have both credentials to set up the subscription +If you have tested this and it does indeed not pick up the right locale you should raise this issue against Spring Framework not Spring Roo as this would definitely be useful beyond Roo generated applications +If you are interested in sharing the new version Im interested in pursing the licensing questionsIm not sure about the performance +Looks good to me +Thank you very much +Theres certainly options for but I was hoping to get to the bottom of the problem in the first place +Applied to both trunk and branch. +might make sense to test in java land +I agree with Chris and Andreas +does seem more likely to be used by the web dev crowd +Its difficult to test this corner case +removedschedulestartscheduleendapp table group tableimageapplist tabledeptid and dept from following has been removed from code and db schemacurriculum tableschedulestartscheduleendapp table group tableimageapplist tabledept tablemoving xmlrpcKey if no one is using API v any more to +Few bugs for this feature are pending will close them as and when they get GirishThat is the latest oneThanksAbhinav +I reverted change in and changed fixver toHowever we still need to fix conflict in opening CFS with the patch attached beforeShould I create and move to the new ticket for that +A testcase is written to test the exceptionHandler +I dont think we should put this in or target this forThis is a major new feature which can be implemented in a compatible manner lets target this for +If there are data written transfer RBW +replaced original patch with one that lists geotools jars in sort you know which classes are affected by those licencesThe dialog is a nice idea +Looks good to me +The patch at least ensures that only the referenced objects from stream overwrite objects read elsewhere +If this is the case there is a workaround +That needs to be is a separate issue though and broader +use Double instead of Integer for percentile so we can ask for percentile checkstyle fix except State object new test case +is trying to figure out the location of where you are storing the +I forgot to add that in order to invoke the test you need to open servlet link like httplocalhostnloptcservlet +Thanks Raghu +bq +Pinging the PMC now +No followup so Im closing as incomplete. +MartinThe issue hasnt been fixed in community branch so its still unresolved +We do a compare delete in the java code so the race condition is now between the comparison and the delete we fail if and only if the session expires and the master node is deleted and the master backup recreates the node +Just an update +This patch includes a new component that split a String into various paragraph by detecting n and r +Reviewed +Mark three comments Looks like Version was added as another string +This is also why Springs defaults to no caching at all when an external transaction manager is in your scenario simply leave it at the default no caching at all +will commit if tests pass +fixed by to set the Fix Versions field to JDO beta. + +Hats off to Vincent! Good catch +Yes I did +Thanks Eric. +Because of this my inclination was to add a new case and new behavior for that case instead of potentially causing compatibility problems by overriding existing cases +the first one will go into the DIRECTORIES cache and win for all future callsHmm good point +release bulk close +Embedded is currently mandatory +bulk defer of featurestaskswishes from to +after picture of a memory analysis +Thanks Rajesh Ill try +Any comments James +Ram I updated the patch a little +NN needs to persist locations for these meta blocks perhaps +Thanks +Hi VincentThanks for the bug report and the example +Great thanksIll try to find some cycles to help out with creating the stats manager +I must have not run ant clean before doing the previous dist build +And thats what we have in both and today +I guess I need to start looking into doing this +bulk close of all resolved issues prior to release. +Patch for issue +Rest of the items are linked from the downloads page +This means that the Javadoc is always out of date unless you build the same code twiceIt also makes the code impossible for new developers to build without figuring out how to disable the javadoc generation +Can you confirm the problem you found is related to this classMany thanks! Michael I couldnt reproduce it +Dave you dont need to mention the commit id eventually Jenkins will comment +Closing RC issues. +Ah Im sorry for confusing you Karthik +I added type compatibility checking code to classThe is the difference between Embedded and withNow Im examing difference +I was more focused on having documentation to match the branches +The file is now present again and the problem should be resolved +I have submitted the patch please review the code +The approach you propose can be a separate JIRA I believe +I agree the set is needed but the current code clones and then adds every single term to the set regardless of whether its already in the set +Thanks Suresh! Committed to trunk +commit the patch in revision for trunk +I was planning to add this later and have this framework to be in place ASAP but I can surely add this feature right nowOne more thing I believe it would be good to have dynamic configuration +This works for me too! Thanks +create a Seam Web project with Seam Runtime A go to WindowPreferences ToolsWebSeam remove Seam Runtime A verify that the above created project contains the error marker Seam Runtime A specified for Seam project XXX does not exist +It is inherited from interfaceIve attached patch for class only +Hi Nicholas +I am wondering if reflection is the best way to incorporate plugins? Can we replace with a factory pattern or something +A couple of other issues I observed with Should check memory availability periodically not on every add Try to buffer in memory first +bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +The previous patches did not apply because they were based out of trunk not trunk +Phil I was about to apply this patch +verified +Oh sorry +With the fix the total number of temporary files never exceeds +Of course this feature needs to be optional since +Quick review of patch contains a bunch of superfluous whitespace changes the commented out examples can probably be simply included on this issue in the comments changes to the generics dont look like they belong in this patch and are superfluous looks like the main change is to call parse with the map is that right +Closing issue the feature has been enabled and most of the resulting problems have been resolvedI think we have for the remaining bugs that surfaced when the feature was first enabled none of which were considered for the release. +I should note that for me Git fell back to a merge when I rebased +Pawel have you had a chance to take a look at this +Fixed for the next release of the Sling with the release of Sling JSP Taglib +If my memory serves me correctly the one problem Gerhard hit had to do with URI issues and he has already resolved that +Triaged July Assigned normal urgency marked as Known fix Repro attached and Workaround attached +I think the crest in rv is probably fine +Well try to get to this asap +It will cache the failed found resourcesThanks +Could you please produce a wire context log of the session exhibiting the problem? do you happen to know if branch is also affected? in SVN trunk and +Cancelling patch +Thanks Paulex this makes sense +Fixed in rev +Patch that checks for null values of slingkey and skips them + +I found another small bug in +Closing +I needed to make a number of changes to to get it to work in the current trunk environment +This necessitated that all tasks returned by the scheduler are checked even for the overloaded method checkAssignment and correspondingly tests in needed to be changed so that it uses multiple task assignment checks where appropriate +Please check that it was applied as you expected +It would be nice to have the standard Login functionalities for RooGWT integration out of the box just like the provided mvc +See comments for explanations +Does this require a change to JBTM or is it a process that is done by productization? If there is a template for projects to do this I am happy to add it into our build process +i think this is ready to commit +Right we tend to use primitives +does not seem to matter which provider I use seem to be happening for any JPA style project +it doesnt fix all the issues but I think we should commit itI think its really confusing that set sets stored as a side effect +Hold out Tim your likely to get further comments before it goes in +Can you confirm that you have access? Cheers Brian it works nowI have pull a firstI resume my work now +A little more investigation and testing shows that the resolve works even with the Workspace Resolver is enabled if I also remove the revConstraints from the Spring Ivy settings files + Attached Removes reference to to grant license +I think my attached zip is bad +DRLVM contains APR which includes expat XML parser +mauro added the converter to xstreams docs +I would propose that we should extend the jobid by HHMM which will make all job ids where the stays up more than a minute unique +Hi GuysLike and bundle same issue also on bundle it wont get installed because of dependencytrue and no other bundles explicitly import it this need be fixed in same wayThanks for the patch and feedback +Is the summary of this bug report correct? It doesnt seem to match the appended description +The original test was committed for and we didnt have optimizer hints back then +this stream does not support setting the caching strategy +Correction is ok but had to be modified for the serialization test to pass +We may want to reconsider the modality of such dialogs +But I could be convinced either way +feel free to reopen it if its still there +Using price is better than price as it uses double comparison instead of integer comparison +Whoa! Now that Guillaume is a PMC member I guess I now have the s required! Will commit shortly +All right +Cenqua has confirmed this is a bug in and will be fixed for +Argh +I think it is a very common mistake to specify bad class name either because of typo or +Its always possible that hadoop introduces new issues tooIts perfectly fine to hope for making hadoop the default in I just dont want us to jump the gun +Ill open a subtask to test a release with JG and a multiplexer sar +Besides this MR patch doesnt decouple FI tests from target +This patch separates solrj testing under Maven so that it runs on its own rather than withIn order to do this solrj needs to include source as test source and include all dependencies as test dependenciesAll tests pass for me locally +We have updated our Hibernate version to and are still seeing this log at the WARN level in our log files +v appliedAlso removed part of the topic watcher registration comment since I think for now we would register the consumer no matter if its allowed topics becomes available +Im just tagging the release correctly. +Otherwise unchangedCos Im a lazy lazy man +This is already possible through the extraInformation field in the parent class +Its more likely the way the POM is handed to the parser +Sorry it this is just my configuration error +But perhaps both cases could be desirable? The MBR implementation is obviously more expensive and some users might prefer orthogonal labels only +Scott per your request Ive opened +Most of the references in were appropriately updated this one seems to have been forgotten +Let me know if you want a unified diff of changes +Adds Writables classes for queues and stacks +Will be covered by the other inlining work that is going on +Added a section to the migration this actually resolved as I see the same for when connection from a Spring client to a server? +Almost in alphabetical +Configs updated with NNCs address but NNAs ZKFC is not modified Manual failover occurred to NNC Some time later NNC fails and the automatic failover does not succeed because NNAs ZKFC still has NNBs address +is not available + +We currently have a number of acceptable findbugs release audit and javadoc warnings defined at the top of each major project +There is still some stabilization work required in but other views are as they should be. +Reviewed with my python hat on +Closing this as cannot reproduce for now. +Yes definitely get the authentication working first +it is called that way in old PDE based build +In the mean time can you tell me if there is any trick to get Pig running on Windows +Thanks +probably it can piggyback on the pass already done by good +Reviewed AppliedPlease make sure to document the usage of this with regard to in a suitable place +I tested with with a schedule defined to every second to try to reproduce this issue and all works fine without Ive been testing with with a similar setup for a couple of days and was about to comment that the issue seems to be fixed for us in too +Itll be caught higher up +sounds reasonable to fix in so that we do not break existing behavior +Verified by Mark +Just want to qualify that the maps do not need to be deleted The current setup is all saved in the +Well look to get to a release fairly quickly early in +testCompressed port out of range +The path is srcmainjamonorgapachehadoophbasetmplmaster in trunk +Thanks for the update Christian +rb +Please check out the trunk branch ofthe repository run the command mvn compile and tellme what happensDo you think the Hadoop folks should note this issue in their pagelong file or would you advocate that they remove thisfile entirely and deny that it ever existed since it contradicts acentral tenet of Maven dogma namely that you can download a Mavenproject and build it without reading the build file +Please attach any artifacts like VDB per my testing it is working as designed +Ill try to check that +Attached patch for including unit tests +I updated the test case to include the standard copyright notice and fixed it so that it will compile on Windows +ThanksJuergen +I suspect it is a stability related issue I had run lots of automated tests on my previous server so maybe all the alerts events etc caused it to destabilize somehow +Do i need to have some other jars? In just copying an annotated class to Axishomerepositorypojo directory made and exposed it as a web service +Here is the merged patch removing the option of gcignoreweakroots +Doesnt seem any need to merge the tests the ones there look like they cover things. +Actually looking at trunk the issue seems to be there too +Going to close this for now please reopen if its still an issue. +pcap of an error and success packet from maven to are noticing the same issue quite often when we uploaded artifacts It is not on the same artifact everytime +Patch applied with minor modification in rThanks for the contribution! +Ive tested it locally by excluding xulrunner from xulrunner fetaure +Cant believe I made that mistake +Commenting out this line makes our application run OK. +If its ok with you Ill mark this one as a duplicate +already verified the root cause that leads to the scenario detected in this issue in for EAP ER as well asverified code change to prevent NPE during race condition when all clusters are dead +If not found I will create a reproduction path and create an issue. +youve forgotten to request the move of your private list FYI +latest +now passes with trueCompiled tests all pass with tracing patch depends on from +Reject application with empty queue name in as does +also. +First test created capped collection with maxsize +So by default probably we should replace illegal bytes with uFFFD as Java does instead of throwing an exception +This is the correct file as simple text and has optimize set to that is not being will create and post a simple test scenerio to execute that will demonstrate this problem in the next day +LOL I am not using it seriously +it makes us very sad +it also includes VS integration and havent merged this to trunk because of the VS component I think its best the nmaven team revise this as a whole when planning the implementation of the VS addin on trunk. +It is OK thanks Stepan +Submitted patchThis is a preliminary patch containing the value profiling framework in +bq +Attached please find a patch for this bug + +Added the option to suppress GUID generation +Just try to provide another way for the webxml not sure it is proper for +I think the patch looks good +Committed +I just committed this patch +James Where are you on this? We are starting to work on and would need your updates ASAP +This really sucks +What are reasons for having to call isStale from isOpen? I personallysee no sense in performing stale check in the isOpen method +Reverted patch due to a bad commit +Fixed by the latest code change. +Fixed in trunk with an integration test. +So in case weve some document for developers like the runbook we can document this there. +in this case right nowsince the issue is within one we dont have to change the currently supported constellationsok ill fix it as soon as ive time for it today +I am setting to and I have applied +Theres one error in a NPE +Could you provide links for some examples of this +that patch should take care of this issue +Sorry but news are badThe same exception from the command lineI suppose it should be better to reopen this issueAnother thing before i create this issue many messages in the mailing list were around the same or similar problem +If this happens shortly after everyone has left for the day there can be quite a large number of emails in everyones inbox +This will break a lot of our code as we use non values as timestamps quite a bit +More detailed design doc and the first version of the patch is in the works +What is happening very reproducibly now is that I started the node and minutes later the forced compaction check in ACS kicks off and then I have looping compaction on the hints but its only compacting the last sstable over and over +There are at least two advantage compared to the current state It is impossible to accidently create new dependencies from persistence classes to other parts of core +already has support for that it might be better to simply reuse that than default to +release audit warnings are all from html files +regenerated those specific cases last time I ran it against the suite those testcases did not exist +Thanks for this enhancement +Equivalent patch against master +Suppose there were two directories in say d and d +Does your local checkout have changes to that havent been committed? is temporarily fragile while it depends on in the parent directory and not a fixed maven hosted artifact +In generate select the key in a round robin manner within the key range +We note that it can really be simplified as the origional was set up to allow for object attribtues associated with each coordinate +Please reopen if a failing case may be provided. +Did this application take minutes to start on but it just didnt time out? Or is the fact that it took over minutes to start new onI wouldnt expect any application to take that long to start but in it would have been possible to take that long without timing out and in it will time out unless all runtime pools are configured to be lazy +has workaround private functions for the issue mentioned in and this needs to be updated and duplication removed once that is fixed +on that Ill commit it tomorrow as for some reason Clouderas offices are blacklisted from ASF SVN at the moment +fixed on master and +One more thought we can group counters by their enumeration class +Complete with commit edcbdcdfeffa. +Another thread then issues another RPC and it completes successfully even though the first thread has not yet consumed the RPC response +Implemented this change in Rev +It didnt work +Ah didnt notice that label and wasnt aware of its use +not minor that page is supposed to be correct +is as a built in annotation and class resolution is trapped out to use the internal definition and error checking is told not to flag it as an error +Updated version of the patch as requested +This is all well and good for running in a full application container like where you can expect some level of classloader trickery but for other runtimes such as Karaf this bundle is broken +I think I didnt understand the issue +I am for fail fast +Thanks +Bug persists with Hibernate closing stale resolved issues +Requirements in a single html file +umthis feature has exisited for three years +So once these two things are resolved we can apply the update ve applied the patch but we still have the BUILD part referenced in snapshots +processing has been changed +Are you sure you did an ant build in the frameworkwidget componentBy the way there will be other Party Manager patches forthcoming +I have tried Russells solution and it works. +Here is the patch file +Here is the attach with grant license to ASF +Youre right +Set If we type appsapplicationTypes or appsapplicationTypes it will return nothing +Verified that the undefined is no longer appended to the idScope when no separator parameter is sent. +Here are two wars to reproduce the issue +Please updateThanks to BETA on the php side want to eval this +Updated patch incl +BTW Sorry for late reply +Hey Zied I have been meaning to add Locators for a while now but just havent had the chance +The Maven POM at configmaven controls the Ant Maven plugin and all Maven repository related think the attached patch is therefore redundant per se +A quick question about your use case +Im closing this JIRA since I cant set it to Resolve so please if you find the problem is not fully resolved +resolving this. +Dennis can you have a look to? It sounds like a possible regression due to this change +I am not sure whether it is a good idea because it is at least confusing +Now it is time to enable the compression +I can live with the ID though +This error happens if you have incompatible versions of some pluginsI think includes some older versions of the central andor maven pluginJBDS IMO isnt ready for testing yet because there are still components that arent built correctly. +See the other ticket about do not copy exchangesWe will address this in Camel will be on a more general level +Im not certain just what the typo is +Removed unnecessary code +If there is some other issue we should create a separate tucket which you just did +Fixed in. +So I guess the direct answer is I have not been able to get this far +Ill throw up an updated patch on Review Board shortly for your review +Ill look at can you see if you can find a fix for +Jesper in your testing of SR and SR are you saying that with SR you do not get authentication errors but with SR you doOr are you just seeing sudden stop of deploymentsDoes the files actually changes +I do understand that two requests may be different and should be processed differently but that is the responsibility of the Message Router so +We have the same problem in and for any CF with gcgs +patch which changes check from as to a utility method which is aware of as as and eap +Please review +Created for in verify once the rpms become available for CR +Attached is the detailed NFS SR was used for the VDI +Yes thats the problem however it doesnt fail under for me +test is unrelated and Mona is looking into it +I went through this patch and it looks good to me +As of Spring stores a reference to the target in the proxy definition using that target definition for is checks without creating the proxy eagerly +The patch does not introduce new features no new tests required +Sandy +I wanted to make it possible for an external client to use this module +Maybe you attach an example that reproduces the problem that would here anybody to help me? standard grails I begin a new project spoiled a transmission parameters in ginclude +Your PMC chair can do this as neededbq +However I am unable to add the test itself and get errors when I run ant test Dtestcase Dqfile Doverwritetrue falseIt says Exception The table defaultsrcpartsrcpartindexproj is an index table +and wrap the other operations as a spellcheckerrebuild kinda thing and not even necessarily componentize those operations since they dont really necessarily need to be hooked together with other operations as a single requestanyway just the overloading of a component to do managerial operations seems awkward +Yes I wanted to volunteer but you were faster +V of the running the full as expected it breaks some stuff +Adding the dependency to +Slightly different but equivalent solution based on this patch was committed +So please apply that patch and make release asap +Changing status to LATER +Would you please be able to review it +Both plugins should be brought to trunk if they pass the Sync is finished +Also note that current implementation might leave unterminated threads when execution was stopped by timeout +Target version cannot be since it is already released +please go ahead Ivan +When I run mvn package in here is what I get in the WARclassesclassesclassesorgclassesorgapacheclassesorgapachemahoutclassesorgapachemahoutcfclassesorgapachemahoutcftasteclassesorgapachemahoutcftasteweblibclassesclassesclassesclassesorgapachemahoutcftastewebclassesorgapachemahoutcftasteweblibliblibliblibliblibliblibliblibliblibliblibliblibliblibliblibliblibliblibliblibliblibliblibmavenmavenmavenmavenmaven +unscheduling for later investigation however I havent been able to reproduce this +Thanks for input Ryan +I believe the jsecuritycheck does not use any data which can be replicated and I have sent a mail to tomcat mailing list to confirm if we need session affinity when clustering +It happens the groovy nature gets added to a project +The code that do it comes from shared so it is preferred do not change it and suggest use jsf instead. +Serp jar and pom until we figure out how to publish these to the maven repo. +I have updated the patch in the following way I have changed synchronization blocks to reflect the discussion of this issue +this error still exists in hibernate is not intended to be used this wayUsing a is an inefficient way to get distinct results + +see comments on ve been meaning to take a look into this but it should be possible to disable validation of search criteria in the Forge configurationThe JSF views for Forge are generated using Metawidget +In Drools will be and jBPM will be +Patch to allow handlers to be injected as described +Can somebody please do that +Please remove the tabs from your patch our coding standard requires the use of spaces instead and then I am a +I think this bug was fixed long ago. +Insane! Will review soon +we could do it for as well +Will provide unit test case later +Committed fixes to trunk for Axis and CXF Committed revision +Can you do the same change on line alsoOtherwise it looks good +I cant close or delete this issue +No NPE generated fro the attached test caseClosing it at the same time +Recent situations show how valuable warnings can be at identifying potential bugs and can spur questions where people ask is this really the best way of doing such and such +Fixed +Great thx a lot Yonik +Steve I did see those warnings on occasion but didnt really have a good explanation for it +Minor issues are corrected +locking at the code I guess its easier to use the special prefix ujaxAh you are right +Could you please clarify this does it fix the problem or not +Trying again withant used onlyant last timeCould be the reason since I used for other reasons before +This has been committed to trunk and. +testcase must be in the included Jetty version then. +The current implementation makes sure this happens +To all interested parties please verify. +Are there any tests for? If not could you add in a simple one +Just to make it sync i proposed ignoreCrc in alsoyour idea also good +If you look at srcc you see that Im just using sed on the to extract the version now +Ill commit this momentarily +But append call can be even later after some time so there may be a chance in adding the new node in pipeline + Username bartmann +I will commit the changes with these last issues after rebasing on once it is committed to the branch has been committed to the branch +Reporters of these issues please verify the fixes and help close these issues +Thanks Cliff +I sure wish the warpath developer was part of this rather then me the novice +Besides these needs only crop up for special MR optimization cases for which a custom implementation should sufficeAdding in for special things is easy but we also have to consider support for it and usefulness when adding it upstream +has performance improvements related tovisiblehidden selectors which may affect us and it should improve performance generally +In Maven this seems to have regressed as everything is showing up deprecated again +As I now understand it the piece were missing is finding methods to wrap the JSP like we do portably for JSF +Im starting by trying to pull all XML references out of +I was busy with automating the EAP smoke testing for upcoming ER I think the Java ACC should solve this requirement I will resume work on this issue as soon as possible +This has already been fixed in the latest versions in SVN at least +Knowing that one could not use them in some cases is confusing and breaks the rule of least astonishment +Try to set table width +Vikas +Looks liked unifying our versions with the rest of Narayana fixed this. +Im okay with that tooJust adding a simple line before the full output like Parameter not known or correct for usage. +No the solution is to get to match the same as +Please comment if you continue to get this problem with the latest snapshot. +Closing the resolved issue. +Im assuming you are using to overwrite the default behaviour of the submit tag +Attaching patch for new behavior is less surprising than the old +If I see this log message Ill paste here the stacktrace +Please check the guide +Has not been fixed as far as I can tell if it has please point to SVN commit thanks +Some junit tests are broken and I am working on them +Tabs in is removed +So if the task is not found it will be mark as rejected +To my understanding we should have the branch ready already so is it already applied to that branch +zip archive you just just installed Spring IDE into plain Eclipse Ganymede SR AJDT from dev update site and could successfully update to +I will attach full diffs as well to +for someone wants relative path into other jar we can target this on request +This is the full document that contains the edits Im suggesting +How do i handle it +I would personally argue that this is a policy decision that the user of thelibrary would like to be able to make +Is there more or did is this summmary correct +It duplicates some code but I have no better idea +from be clear Im okay with the change in concept but Id prefer to not make this change in behavior +The intersection method was not handling duplicates correctly +Have you tried to your patch locally on a fresh trunk export first? Does it work +Committed trunk and branch +This definitely wont get in due to some unforeseen circumstances +Attached is the patch that highlights the menu item when click on that menuitem +Koji your comment makes sense if all threads happen to have the same seed +moreTest Failed +Please check in the next build if it works for you. +As you know we dont want corrupt edit logs +File for in contained some invalid characters +Can you duplicate this on a regular basis +bq +I will definitely be sure to add a version line to the gem spec before we roll the release +Hi AndrzejDo you have a downloadable generated UML for this? Thx +I personally dont like it so much but yes it will be easier to find a configuration if there are a lot of themBTW is also used in the donated plugin +Because of this the server class cannot be generic because Restlet doesnt know the type only the class name so in our implementation we were using the raw type +That is standard bugzilla functionality to add auto links because the text Attachment nnn appears in the comment text +There will be no release. +Committed this to trunk and +Thats the wrong it should be using +This is a pretty big change to what appears only a small overall performance gain and only on the client +changed the jira component to pretty good to me +review for M +If there is no obvious flaw then I am fine with others judgment +Patch built on XP and LinuxTests run successfully on XP and Linux +I have to admit that I am a bit confused about I havent been in this code a lotIf we find a jar finder class that ships with the version of hadoop that we are compiled against shouldnt we use that +the changes i have applied both to and +infinispan log for core module is attached for detailed logs +prefix is used as an extension mapping +Felix has been released the vote is in progress +Pushing to but maybe we should just close this +Closing all resolved tickets from or older +might decide to leave it out completelyThe behavior should be consistent irrespective of the output format we choose +Closing. +No CDI builder hang. +Is anybody doing anything with this issueOr has it been moved to Richfaces project? +merged into as r +nope go take care of your dad +Thanks Sid! +I will create a separate issue to add support for returning with postgresql to be resolved when refactoring and s +Thanks Owen! +If I have to respect any rules please give me a link with instructions. +Everyone This data has been loaded as per the request +clean test now works for me +As per svn commit history the fix was to enable Pam support in the sshd config +Cant do that kills performance if we query user preferences a dozen times for any list with addresses +Thanks +In fact that is the reason I have added logging dependencies in maven +any chance that it will be fixed soon +It looks like is creating the paritions file on localFs instead of HDFS +Good news thanks for the test +All are known issues +Bulk update to fix version +Attaching patch and review request with fix for issue +Making this a design doc option alleviates any of my concerns about views being different on different nodes +Canceling so that the next patch wouldnt run through QA +Ill verify with scout trunk to see if this had been fixed already and get back to you soonBTW could you deploy a latest scout snasphot for our build system to pick up +Eclipse PlatformVersion Build id might be the issue because in the branch Eclipse switched entirely to using P and I think the old modules are not included anymoreHence the error you getJust close the rcp project and it should be all good +In your code I dont see anything Ajax related +Then the skewed keys will only go to some of the reducers and ideally we want to distribute them across all reducers right? I can fix it in +fixed rev Thanks +It should use the property as the root for its temporary data +If it causes a race condition it must be fixed +Cool looks like you fixed this Ted. +Reading the last comments I guess this should either be closed or renamed +The build is still successfulHowever the NPE comes to standard output which is rather alarmingSee the attachment to +Header moved to perf tools +think I need to attach jar filesnd version jar is coming up as it is more than part of Chiragcould you include your report exampleMoreover some of the libs you added in your archives are already present in would it be possible to make it more accurateCheers +My main worry is what happens on cygwin whether there is a sbinsysctl and what it says that stuff may break +This contribution is simply a +Should this check to see if CLASSPATH has been previously set before including it in the new CLASSPATH +Roberto is the path to your src file in your CLASSPATH +KVM Provisioning Host Setup Script +trunk Committed revision x Committed revision Lets see how hudson likes these changes +Think waiting for the next version is a good idea and makes live easier +Client know the original file size +Ive just found a bug in how this is is supposed to be placed in tools but the Ant copy command in build still uses the old extras tools should be the final location unless someone objects +IMHO the test stuff shouldnt be in the classpath unless youre running tests and tools arent testsSo yes Im proposing a separate place to store the tests and their dependencies +I just committed this +JCAIn both managed and environments registration of a connectionfactory instance in the JNDI namespace must use either the JNDI Reference orSerializable mechanism +Passing over to Julian to work on the actual plan doc though +We will see later if we find enough manpower to extend using a generic API + +Sorry also pull request has been merged in and some basic testing has confirmed this work is complete +Hello Volkerthanks for your answer +The same problem occurs with +Just use the +bq +Heres a first draft the patch uses to ensure that the znode is deleted once the connection is back +Thanks a lot Christian! +Internally the patch has been tested also +I dont have a definite answer but there is currently no clean way to do what clinton wants to do IMO +However we still have things spread out overany necessary transformsthe type classsystem vdb datatypesplus any misc handling in string handling translators and the driver. +Thanks Doug +was written for managing multiple services in AS all running on the same channel +Closing Metro stack related jiras previously marked as for community contribution due to lack of interest no action. +The first problem is the treatment of relative +Perhaps this is no longer an issue +So theres no way to do this inline +I think this is worth addressing simply because the collision probabilities are surprisingly high for anyone working off the documentation. +The original title of this issue was not helpful +If this should be reconsidered please add the label +This is resolved +I thought i gave a aboveOh I thought the was only for not making this class a generic termsenum mergerI just committed the patch thanks Boaz! +Will investigate +The result was provided by +Use cases where lots of headers are applied to events very large batch sizes on source lots of sources are hooked up to a single mem channel or event body size is relatively large are situations where user may need to tweak byte +I may have fixed this problem by first eliminating the second id and also by making two superclasses serializable +a new table directory is assembled from the snapshot. +Yes that is what I had in mind +Applied a fix to the branch. +Thanks for the suggestionIve reworked the mechanism quite a bit now always keeps tracks of which beans a singleton that it creates is dependent on no matter if through bean properties or constructor arguments respectively direct references or autowiring or explicit a side effect of that change the get hook is no longer necessary All that stuff is determined and stored at singleton creation time now +Thanks Pierre your patch is in trunk at rI did not backport yet because its a trivial improvement +Not a real blocker if survived M +The solution for this issue was already released with +HiI think this is not the best way forward +DanJust ran a full distro build with the attachment overlays applied and everything looks under +its part of cdi and therefore users will get familiar with itmaybe we should use the same api and add some we have in + +The needs the fix to branch +This new JIRA is about this particular issue +Seems like such a I cant believe I didnt do it sooner. +Attached is the exported version that includes code +request to please check the capitalisazions in objectfactory of +Do we follow any standard to calculate hashCode +Plus I wouldnt want the object creation +Any other environment that we want to use EL in will require similar hacks +I think itd be good to pass the as the only one parameter what do you think +Another thing to watch out for is people running the default config then writing nasty emailtwitter notes because our out of the box performance isnt good unless you do tweaks XYZI think this is simply a matter of education +i think python will be more useful only because it is a natural scripting language and can develop tools much faster there +The suggestion of compositing might be an option but no matter how a look at it I cant see why the loop shouldnt naturally provide at least access to know this kind of information +Youll notice that it doesnt save its position +txt textplainBut I need more mappings +is no such class thus the stack overflow +Tx for the update ArpitThis looks good +Adding a combined patch with MR changes as there is a change in the api that breaks MR +Exemple les enjeux sont lvs the stakes are high +I just want to fix the issue where you have history none but the engine requires a property in the database +This solution maintains a good property any packet sent to datanodes starts at a checksum chunk boundary +Please reopen if this still is a problem. +Ran mvn test with Maven and it tests successful +I moved the patch here from just committed this +Just realized I didnt put all the info I learnt hereIncreasing the wait time for booting the server makes the test passThe hang seems to be due to the cleanup code when one of the waits for server to start failsThe code then tries to shutdown all possible servers it started the code is around line with this comment If something went wrong make sure all these servers are shutdownNot sure what is causing the hang though +As mentioned the simplest thing would be to declare SP the minimum platform for Ill see about appyling SP to our server so we can use it for development too +This will be a solution for now +Sandy just curious what should the behavior be if a user sets both but to different values? Do we give higher priority to mapredmapreduce +OK on it seems to work +The problem might be that the file is corrupted as it has beenchecked into CVS as a text file +Thanks a lot John for fixing this ugly bug +Then you could specify an expression for the init value and still have a default. +We could always raise another feature request JIRA for the subscription dropping also for nondurable messaging we have the option to drop old messages if consumers slow downSo is this issue resolved now +Well if the goal is just to be able to drop entire sstables when we know everything is expired we could compute and keep in the metadata the min TTL of the sstable +Do you know of any specific client issues which require deviation from the RFCDo you read the RFC differently to me +I xml utility function into a class in HDFS in the latest patch +Cassandra throws a few custom exceptions that are +if two clients are writing to the end of the same block +Reformat description +Hi Stefan I totally agree with you and ben +See the attached image showing the file handles after refreshments of the welcome page of our application +Its not really a new phenomenon though We had support for registered singleton instances which are not backed by bean definitions all the time we just didnt use it much out of the a correct version of the code should check before it calls get +Any chance of a reviewBasically a hole was opened by +on first version of patch +Its also a big task to test but my concern is how fast this runs now +As RB as a hide whitespace changes feature we should just use it more +I moved the addition of those default java opts to until after the script is processed to fix this +Because the test is trying to create dfslocalhostsrcdat +I have noticed that the transaction completion code performs an HTTP HEAD request followed by an HTTP PUT +Verified in + +OLD NEW devSincerelyshathawayXalan Documentation. +Our application doesnt know when its corrupted but as user we can know with functionality failures and logs +I discoverred a flaw in the unit tests along the way +It can be used outside of MINA +Rob any chance you could add some screenshots for this to new and noteworthy? +Merged +Looks like the same NPE as in and issue has been resolved for over a year with no further movement +so on the patch +bulk move of all unresolved issues from to +bq +Applied path into trunk and branch with thanks to William +Take a look at and see how I am determining whether to include a given type in the default fetch group +click the add a component button +That way well know if we do get it fixed +We are way past at this point so I am closing this. +Your change in the file looks fineThe changes in the page have some problems +Thats my guessbq +Martin do you need anything else to set up the it possible to port the test case so that the customer will be assured their use case works as intended with the new messagescheduler service implementations +I think that the patches have been applied? +It is in svn but the script which we use to build the releases is broken as it is ommiting this ll keep this issue open and block the release of until its for the heads up. +Committed trunk revision Now merging +okit works finethank you +summary of the oak planning meeting where we discussed this issuea filter hidden nodes in b define set of hidden node names that can be checked upon Treeget in order not to avoid iterating over all children to make sure no hidden items are presentc possibly make this node names configurable and enforce validity of hidden node namesi addressed a and b at revision and removed the Ignore annotation from the test casesregarding c i added a TODO remark with the constant defining the hidden node close for the release +It may take a more severe overhauling of the filters +Im closing this for now +Aah thanks for the explaination and links +Say one thread is already writing a block Band there here is another request to write the same block then the second thread will see that B still not a valid block and delete the filesI think only the receiver thread that actually the ownercreator of the block should delete it +Will wait for others to chime in as well +There is actually a problem with interfaces and the dynamic Proxy stuff from Java which is what is used here + jars required for what caused this +That worked +I didnt notice that the getPrincipals API was there +As far as I can tell from your error isnt being passed a VFS path just a regular path of Userstsinghzenapsispublicuploadstmp +so it will appear in the release notes + Added CHUKWACONFDIR to class +Yes my bad +seambuildlib is there from the beginning of Seam in EAP + this bug has already been fixed +schedule for next release as that warn was recently introduced +fixes only the second problem +It has recently been fixed in Trunk so the problem will not appear in the upcoming version +Everything looks goodUnless there are objections Ill commit this monday morning in time for the first release candidate +This state of the snapshot prevents any further snapshots to be attempted on this volume +User was correct issue not fixed +Mind making your new test junit sytle? Otherwise looks great + for the patch +But that you could be sneaky and just have an internal xlink reference to a value you return at the very end of the document so that you can just calculate bounds as you produce the features +John is now a committer and he has restarted the work in his sandbox +Delays output from until statistically valid number of samples have been obtained +So cloud user cannot create mount point there to copy template from S to NFS cache +I agree that it makes sense to submit both +If you do find the root cause why it worked in the first place please let me know +I noticed that it is possible to register processors just for test class via extending by using test container application context as parent of a test application context which sole purpose is to instantiate test +i checked it with the domWriter using and it worksCould u try it using Xerces beta +You may save yourself postage costs if ordering from very muchies +Thanks Rod +Im working on improving the class it will be able to parse a string in that format +I suppose that serves us right if we dont access them +This allows for more uniform and consistent usepatch p i Please review and let me know if there are any issuescheers in. +This will ensure a large or a few large fetch requests dont penalize several other smaller requests +They have been committed in revision please could you verify that they have been applied as expected +MikeAppreciate the rapid turnaround on this +I am authenticated as admin for both databases +I fact my real problem was with the feature of Hibernate and I havent noticed the that works fine. +The socket channel read methods never block theyll only read as much as is available. +The shortest path to fixing the issue you mentioned is by making stop +I committed the patch with revision . +The list view on Groups page works this way +Bug has been marked as a duplicate of this bug +Turning it off is half of it +Thanks +This patch should be applied from the +The latest in cvs is and the patch doesnt seem to match +The rerun is to check whether the fix is done or not Correct me if I am wrong Jon +Not possible as this conflicts with to HEM fixes this. +You can use multiple beans! To feed a view with data +The output is written out as it comes in +When I tested it out the standard deployundeploy seemed to work fine for a directory +Transitioning to waiting for fine do I just edit and point to +new snapshots uploaded to codehaus snapshot snapshot version works for me +Im starting to hate this issue Fix on the way +use an utility class if needed +bulk defer of issues to +Still having this issue on Groovy with JDK +Closing all issues created in and before which are marked resolved +the example will be only in EAP +Woden weekly call Jan delayed M release date by week and M content +Attaching which is the patch for Phase as defined in my earlier comment +Attaching the patchA few notes on the patch Introduced Windowsget that provides native implementation for File length that also follows symbolic links +and are done +Updated one more has been committed this should work +So this is what aboutI dont know if JPA requires the select new operator to work transparently for timestampdate so not sure if this is an spec issue +Committed the patch to trunk +Closed after release. +We do encoding everywhere some amp and then urlencode on paramsWe should remove those everywhere +Patches were applied the revision r +Jeff Yu took care of the jars +To reproduce one needs to +The patch is committed now. +In practice this is usually a or very low number +Ill complete this for +So it needs to be done as a part of Avro + +Thanks Bryan +I wonder if that just makes all of this a moot point +Thing that may be handy not to complicate things forThoughts +Yes as usual +Marios fixed this in April with commit acefdaaaebbdba. +There are identical copies of the database in store store and store +I suppose the original idea was for readwrite access to JMX to be a function +i will have a look at it soon + Bug has been marked as a duplicate of this bug +Upon closer inspection found blackberry are listed in Project Explorer tab under plugins folder of guess nothing happens in the Cordova Configuration Editor or file as these are blackberry and not apache cordova apis +Nevertheless I must force the topic +Added sym link but not working +now needs implementations for each of the connection scenarios we want to support +Thank you for the explanation +Yes another jira +Hi ThomasI have looked through both variants but am still undecidedGenerally I think we should keep the inheritance to as the class should remain a decorator +Hi Brandon Will it make sense for the remote node to avoid traffic for a given the start and end? yes start of it is going to be an issue but the end of the major event can be at least conservative on not sending data +In the main time a unit test would be welcome +Checking a few more things and then will commit +Thoughts +Robert and I only want to get the fixes for the already existing directories into the core +Please see for the correct use case. +I have merged this to. +Ive verified mvn clean install +This seems to be a regression therefore Im marking it for rThanks Chantal with the current RC +Moving to ER to wait for ISPN upgrade +I just want to continue to set precident that this is important while notblocking adding production level support for optional toolsWe created for the very reason that some production level tools were not needed by some derby applications and thus we allowed users to decide if they wanted them or not +The idea is that you can program in Pig Latin independently of the backend configurable in. +Looking in HRS I see that we first update the +Instant feedback after user issue the foreach statement in Grunt +The code looks really good and has some great ideas that Camel could useAlso needed Default port numbers for protocols DONE Alternative set of parameter for username in URI DONE +Merged to. +It does not use at all + fixes both problems +Fixed in svn +Please remove the call to remove call from within the report +Anyway thank you +Im sorry Ive been busy with other things and didnt have the time to update the unit tests +I just realize there exists a unique constraint on reference per jira +So see hwta you can do and submit a PR and well take it from there. + +Merged back to +Thanks KiranYour patch is in trunk r r rI choose your way adding would have been an overkill and refactoring risky +The only real issue I have is with the tests +Thanks for your feedback by this +Perhaps they can be put on the site before the changes role out in a distro +Assign please to me Ill do that in next days +seeing Thejas I committed this +Of course deadlocks can be very evasive based on any number of factors but Im still finding this surprising +Rebased patch +Heshan please attach a sample synapse configuration file which explains this situation +To apply on ipojo project +The Teiid Designer DDL importer needs to have the and the jars updated +Could this be fixed by reverting to the separate packages and setting the scope to provided or runtime? Would standalone usage be affected by changing the scope +No more dependencies to the portlet API in the core. +You should probably ask for help on the Roller user list +Thanks Nicholas for reviewing thisJust noticed that this added an Avro Nullable +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Ill move this out of until there is a decision on a subsequent juddi release. +Note only the ibm client master file has changed since only ibm and greater jvms come with jce that can support eusridpwdRan derbyall on ibmlinux okCan someone please commit this change +Comment is a serious mistake +I will close the issue as it sounds like you consider this an acceptable change +I was just teasing you enjoy your weekend +But current runtime holds some instances that cant be shared during concurrent invocations such as topSelf I think threadlocal runtime is a better solution +Fixed by revalAt throws DPE if the result is not set and the target place is close of resolved issues. +Its now using generics with a +Can replicate this issue if hasnt been downloaded issue doesnt occur if the SDK has been correctly installed as per the instructions in the READMERELEASENOTESIf you want to change this behaviour you can either download and place in frameworkslibs OR manually edit and remove the entry for the OSMF runtime shared library +Whatever you do to keep lines of Java code from being too long you should do the same for DITA +As David mentioned fixed this so if you dont pass default the code isnt run. +I am not understanding your problem fully +even when one specified useOriginalwsdl to be set in the as has been previously stated +AS trunk is now at to Paul +Thanks for pointing out the issue +Should I backport +The community has to make this call Bjorn can you send a mail to describing the issueIt would be great to show images of both acceptable behavior and buggy one +Please review this. +Exactly +Patch for attached +Ill add some tests that start a perform an HTTP GET to retrieve the page and check for of XML results +I like overwritetruefalse better but i dont have any legacy solr code +I assume there would be fewer upgrade issues to considerhere since sqlAuthorization is new in +What would be complicated is supporting all the possible variations which is why I suggested a template approach might be the bestThere could be a set of named connection templates as well from which the could chooseDIRECTSITEUSERSITEUSERSITEOPENSITEetc +pid files seem to be not getting created on cygwin +What is the status here? Thanks +I assumed that this will be checked already +Are we really going to apply a patch or are we going to SVN merge? I would rather see a merge so we do not lose the commit history on the branch +But for the other branches we can look into this issue a bit more +looks like isnt even used anywhere +Seems like having a separate thread that watches the open files should do the trick +I have deleted the old category descriptions and they still didnt work for me +Will look into it +Is there any chance this will be included in releaseThank you +Reviewed patch and did trial build +The patch partly fixed problem +suresh Is it something different from what Jakob mentioned above? This patch does not use as suggested by Jakob and the rest were cleaned up in +Use real applicationId instead of mock one in applicationIds compareTo method will do its a patch is ready for review nowThank you +The only difference between commands listAlltrue is not being sent with the first command +Thanks a lot Jing! and many thanks to you allve questioned the necessity of this pattern before because its required multiple tweaks since being added +Andreas is it possible to avoid these exceptions to realise text extraction of these pdfs +Committed as a part of. +which then should not be necessaryThank you for your time +And a as a concept it just isnt obvious how it fits in Spring MVC view rendering and what can be used for +This eliminates warnings about resource conflicts even without adding property to use separate output folders +I dont see any reference to HADOOPTIMEZONE in hadoop code or in hive code +Changes should be propagated to throughWill take a look +Since I wasnt sure if aparam is the same as aparam I didnt commit it but attached the patch +In some cases it is not clear what order to evaluate the window expressionsWe have already removed these features from the Semantic Analyzer + +Attached a patch changing web console to use anonymous subject for a patch adding temporary authorization checks for message copydeletemove in management web a patch adding into to allow set principalsubject from the specified request header +Now +What do you mean exactly under all pages? This button shows Preferences Page for VPE +I have spent hours with content links still return you to the page containing the original cms instance +had some problems in small changesChange a column to be unique +I checked the AP sources and didnt find any remarkable static maps or similar +Postponing feature request. +put to trunk and branch +rvm Xaosfinalreportlevel cp LOCATION trivialTypeand then the problem occurs already +Sylvain did some changes last weekend that may fix this problem +I will be glad to provide a revised or additional patch in response to any comments +Theres a JIRA to fix itI looked at ThanksI uploaded a new patch +Thanks anyways. +We should add some documentation someplace for those making the upgrade but otherwise I think were good +I will stick to PLAIN ANONYMOUS at first to complete initial XMPP handshake soonDo you want to give the other mechanisms a try at some time now or later? If later I would ping you if everything is nicely set up for it. +They were written as part of my Hackathloneffort and must just be ported forward to the current CVSI will put them up on the web shortly +It doesnt change the default behaviour but it does provide a new method which maybe called to not get the stderr output and instead get a checked exception thrown +Our application will get more users in the future so were not lucky with this leaks +LukaszThe code is looking really nice +but after I ran into this issue twice in two application within the last two days I put my patch in an initializer for all my applications since I am using webrick with jruby regularly for development +Committed revision +There has to be a way around this? Didnt the folks at Digg implement with client side resolution not requiring for reads? Server side resolution is definitely the way to go +It does not affect trunk +First prototype +Please refer to the following ckasses and to cope with too fast read to cope with too fast write +ve handled this type of thing in other ways and may revisit at some point. +Patch tested locally and applied in a single commit if it causes problems iI will fix or revert although I felt that perhaps enough people had looked and I should stop asking take a deep breath and commit it +we usually recommend between GB to keep the region count down +So it might also come in handy for other problems further down the lineThe patch is very raw and needs further cleaning up +Moving things around is always a bit tricky to express in patch format +Verify this section in Admin Guide and close issue +You want to control the partitioning as wellFor example lets we have the following map outputPK SK VPK SK VPK SK VPK SK Vwhere are primary keys and are secondary keysWhat you want is that all the tuples with same primary keys will go to the same reducerWithin the same reducer the tuples are sorted by the primarysecondary keysWhen passing to reduce calls you want the primary key wil be the key to reduce call and the values with the same primary key will be passed to the reduce call in the sorted order by the secondary keys +Instead it is since hadoop is default +For the documentation I created started new build for TRUNK where passed but still hung +Ill update or close this ticket once I know more +We can get duplicate messages even with todays design +Refactored and added a default poll interval of seconds. +Document that builds take longer with distributed of branchr of resolved this issue since theres a documentation reopen this if you think it should be fixedD +will solve this issue as it is issue for MC and EJB container +Im leaving this one open because Im still uncertain exactly what this should be +Serializing that data and reading it back in may be faster than whats currently being done but it will still be +I fixed the clipping issue +Missed a from the first one this fixes the first diff. +Im an internal employee so I dont have a Red Hat support manager +same file but now the license is to how to determine the charset it looks as though ASCII or are suitable as the defaultGnu docs mention local variant of ASCII +Simon cant we imagine to define a boolean option on Sonar Chectyle Plugin to handle this feature +This seems like a sane approach and has worked well for us +SeverinFixed +This may be refined later to just get the +Works for me about the the capabilities yes np works for me too as well +Shit im sorry man i totally droped the ball on thisi havent looked at your latest patch but catching up on the comments you general API compatibility approach sounds fine to me you may want to follow up with Adrien or Yonik to sanity check your decay code because i didnt fully understand what you guys were talking about the first time it came up +TedSorry Ted +Done. +A new patch was attached +you say JSF implementation checkboxes but your screenshot is JPA +yeah i think thats the issue +Attaching combined patch for to work +The DBO may need to disable an account until a security threat is cleared +Tried this with current code +The common case is Addressing handlers populate those that dose not mean that only addressing handler can do that someone else can do that too +Erik +I just committed this +Grouped faceting on multivalued fields is now properly testedI think it is ready to be committed +Closing issues +I am unable to had reverse engineered the database and it happened Im also having issues with the +has to do with the Dependency Report in Maven Project Info Reports Plugin +Our run only we want a release well have to make another job which is IMO better for the quality of job feedback JBT install test red issue in JBT JBITS install tests red Issue in JBITS +This issue is part of a bulk update that required reopening the issue editing it and then resolving it +Ill try that. +The documentation for the current release can be found in the docs subdirectory in the distribution +It is making the argument into a qualified version of the argument not overwriting itI am not sure its not specific to spectral clustering +In any case the replica manager does the appropriate check just before executing the leaderfollower state change on its end +I wonder will there be any change in licensing details for Derby when we include the felix sourcesm not sure that I understand the reasoning behind this change +Almost every other member in that class can be set except the description field +Merged into the master branch. +I will report it on the TPTP project and keep you informed on the progress +I think the libs were done this way just as a mimic of what doesHmm so should I go ahead with option +Checking it in +I just checked it with the latestLaunched quickstart clicked open closed the modal clicked open another time bam Wicket Ajax Debug fever is on +Fixed in branch should make resolved issues for already released versions +I can take a closer look but itll have to wait until next week +The attached patch fixes this +I mean you send request A and request B shortly after one another and you dont know which one will receive a response first +Anis are you planning to slip this into +Reopening +I believe that all the state is segmented how it should be but the methods of access are pretty muddy still +This patch fixes the problem +OK I plan to commit this in a day or two +thanks for reporting this issue +through hudson +Committed revison +Thanks for working on this +This looks like a good addition to Nutch couple of commentsThe added comments in should be removedAny change to add some junit test along with this +Merged to +Still crashes +roles are pretty well unspecified so anyone using them doesnt have too much to complain about when something goes wrong +I have use example and example help to show how it works of course after community decision about example and exampleext goals I will move to the right place +We can likely apply this patch but our ultimate goal is to eliminate +Im using plugin and Hudson with a maven build and I have this problem too +Devaraj thanks for capturing! Attaching a fixed version of patch +The read barrier can be too slow +The original files if the tests mess up the original files is the compiled beans is the java source for the beans is the java source for the stub interface and earlier file was generated by axis so it was incorrectThis one is correct. +Certainly thats not how I read the comment stream above +Waiting for their fixThank you Ondrej for filling this issue. +Im working on some node management tools and having an easy way to get these things would make this much easierIn an attempt to make this more focused which settings in particular are direly needed and unreachable otherwise +the shared libraries +I have basic Text option set +How do each of these impact the operations I want to run which omissions cause immediate failure and which wont fail until Im into the rd job in my plan? Seems most people I talk to populate their jars everywhere and cross their fingers +Justins patch committed in r on trunk and r on. +In my opinion Both of them should be closed +Thats weird because for me it always returns page +Applied +Skimmed the patch +Should this be resolved now? +dir of logs cant have anything in it anymore else triggers alarms in new log splitting code +Hello SumitI have modified service as per given description will be subscribed if user pass following extra parameters subscribe Y +Possible duplicate of s definitely a dupe but the problem is in Surefire itself not the maven plugin +Make sure its fixed +This is fixed in SVN +If this is still unresolved why is it marked as fixed for +I think it is ready for commit +Cleaned up patch for a cleaned up patch for this for the review Ill make the code style changes and remove the modified default +to fix this issue we require server server protocol change +Is this fix complete in trunk? If so we should probably get the changes merged into +Thinking more about it it could also be a +Thanks NathanIt seems like there are some instead of spaces in the commit +One triggers the +Was fixed by Stan Silvert in Nightly Build. +Please review and close this issue during CR has been fixed +Theres a balance to be struck between adding too much functionality and obfuscating what in actually going on +get +The ejb client jar is named with an extension but is still referenced without an extension in the ejb jars jee dependency is also run the buildAdd to +That patch also contains a release note for the changed behavior +Can you attach a simple project that I can use to reproduce the problem +You mean sourcepath right +It would need to be restarted in order to rejoin the cluster but this can be addressed in another JIRA +I cant think of a nice solution that fixes this +Should propagate soon +however looking again I see that there was another issue in +Forgot to mention leaving as is means we dont need to do anything for trunk +I think using an algorithm to assign podlings to reporting months would be better than manual assignment +We should look for a more local change I think +If its not being used lets remove it +Marking this as a blocker +The developer should know ahead of time through some means when it is appropriate to alter the direction of the text +I just committed this +Dont have cycles to address this minor nit +Will update this thread once we have more info +Get the and run thebuild from build and then run build bundlefrom jetty and the jettyoutput worksfine under +Hi OleI note that other tests use +I definitely agree that the ideal would be for each issue to be logged once and only once +Hi Janthanks a lot for the test thats very useful in these cases +List and Set have been around for sometime but is quite new and was missing the required conversion during construction +I guess I could make it configurable whether it is a host based service or not? One other issue I had was that the signature verification is failing when the SOAP response from the server is a fault +rar file +Marking Wont fix Rejected and Out of Date items as CLOSED. +Here is the patch to fix the problem observed +Thanks once again for you super support +Can you put such an and corresponding artifact on a public HTTP server somewhereIt would make it easier to reproducefix it +What u say +I had to revert at r because it introduces a bad when using a lookup by name at least as far as I can tell +This makes more sense since as mentioned above the only thing that really needs to be tracked is the key itself +I just committed this. +I definitely think it better we provide the extension points in the you need rather than introduce another layer of complexity for now +The patch is being taken from the oozie top folder +Here is the patch which fixes this spec incompatibility +Patch upmerged to trunkAlso fixed for +Moving to CP as there is a workaround +Resolved +Modifying the applications manifest isnt the solution to the problem +Id be highly concerned about this working in +hum +I added the respository URI to a feature if you have multiple repositories installed how can you tell what repository this feature was installed fromA list of installed repositories was meant to give the user an easy way of just unistalling a particular repository otherwise they have to type the maven url for it +for patchIs it not needed in the to be released version Ted? If not will be an open defect in version right +Do we need to change the name of this deployment descriptor from Apache ODE Deployment Descriptor to something more? Seeing Apache ODE though Riftsaw uses the technology doesnt necessarily connect immediately to BPEL in my mind +The patch perfectly work under Windows bit also +The unregister is not successful because of classloading issues +But this time I will discuss it on the forum first +I think what you need is offt +Updated patch to include rename class cast exception as well +If I understand correctly were talking about a bit of an edge case the case where a user has got traceDirectory set the directory exists theyve got read and write permissions for to the directory but theyve got write but not read permissions in the policy file for to the directoryI think it would be acceptable to introduce the incompatibility of requiring the users to modify the policy files +looks good +Why not let it be and change configurationIn my experience users dont set this until theyve been around the Hadoop block for a while and even then this one is easy to miss +but not the +Im not sure when Apache to remote repositories but please let me knw if it works for you now +thats kind of a dirty hack Is there any reason for the applicationName having to be aNow my wikis name is Wiki +looks like this is present in the latest Tomahawk +One of the requirements for the designer is to be Visio like +It also has worse locality of access causing many more CPU cache misses to traverse overEach node object takes bytes +Forgys paper he described basic nodes root and terminal +Does some code cleanup and avoids exporting unnecessary functions +And as I think design problem. +In my use case right now Im recording results though the API and I dont record the configuration failure +Regarding between SOLR and MiddlewareYou need to deal with additional internal at middleware +See some old + +bq +Rejecting after spending some time with AS Beta and discussion with team membersLatest testsuite for works fine with it AS testsuite has no problem tooCLI works without password on the same machine where the server runs +This should apply to both andThis patch looks good +Verified on JBDS. +Only throws the above exception +Details please +Could you please create the account and install the key +Do not stop online as it leaves the offline deployer with no stores to work with +Be careful on this +So I will make the changes to RM JT NM +from by Jun Rao Its probably worthwhile to make interval in HHM configurable +mkdir p varlibcloudmntVM chown VM Enable Zone Now system VMs got startedThanks because we can proceed further with work around for this issue can we reduce the severity to citical from blocker +It should not have Xerces +Elliott the patch committed is different from what you thought fixed the issue? Remember you thought youd fixed it but when I ran it w Archimedes little test code the still showed +DohI should have looked up compiled in the legal dictionarynever mindCraig +Tika uses maven to build which Im not familiar with how to configure +Actually for me build is failing because of the missing macros +Due to I am not so familar with maven can you please provide the information how to add an profile in +Thanks for working through this with me +Patch fixes all failing tests at mysql +This is the original feature request to add this functionality So it did work at some point +Screenshot of the CSS error +Hey Suresh the attached patch looks unrelated wrong one +This looks good except instead of protected fields why not add protected setter methods +I give brief explanation how the language work +So really the only options for a user are disabled and enabled +But another task that is starting at the same time fails with no logs and no explanation +Will fix +Yep someone else was elaborating on this on irc +That seems most excellentWhat does everyone else think +Lukas I misread this as a notify problem rather than a message problem +So they dont need further recursion and can continue with the next step in a simple loopI know that there was a quickstart to check the recursion depth and stack size allready +Hmm ok but this also applies to the data in the Consider you update the file content or change the last modification time stamp +I can provide a patch if necessaryPatches are always welcome +Bulk close for +I figured that a lot of the Strings being escaped would not need any replacements to be done so I opted for a scenerio where the only thing wasted was CPU cyclesI added yours to my set of benchmark tests and got the following many long The only unexpected result is that of the many replacements test where yours beat mine silly + +Are there any other exceptions loggedNo +What we could do is add a configuration that causes Esper to close the connection and obtain a new one and retrying the JDBC call can also provide a interface that an application can implement instead of going through JNDI +The text box is correct initialized with value hello. +I have added overloaded constructor and necessary methods which will have the file name also added as parameter +Updated patch to fix failing tests +Gabe has a solution for this +Oleg +Not a problem of but a usage question +We already have this annotation its called requiredI disagree +Hi Mario thanks for the example +If there is issue in accessing the dump file please let me know +Not a real issue PBKAC +I gave it a test +The first call triggers JI stuff to load up which insinuates itself into the stack trace +Canceling patch pending changes from Nicholas review. +I do not understand what it is you want here there is no notion of a hibernate runtime +This just became a bigger issue since I upgraded to jspwiki and am trying to use the fck editor which is selected in user preferences +Strong I added some code to an existing test caseIts included in the patch attached to this issue +Committed in r with the following changes Dropped type parameter from Added argument validation Added missing javadoc Added tests for invalid parameters Use final where possible Replaced Math with Thanks for the patch! +Thanks Konstantin +Added namspace support for the default output channel on the remainder of this issue will be addressed through improvement +The same patch for branch +The full source code of the web service and the client as well as the WSDL file are attached to this issueActually I should add I was testing with cxftrunk not the release +Verified. +I will do it tomorrow +Added incremental patch with unit test +Does not look like we would be able to do this before +If we had save classpath you can do clean compile before switching to h +Will rename and integrate it. +Thanks for taking a look +Very raw no tests yet +The problem is probably caused by which did not exist in +added Sauls clarification to at r +to commit if the regression tests still pass +Its reproduced by using wicket and upwards +I imagine thats a very unusual use case which may help explain why weve never seen this reported +Pull request merged +Administration Roles +Somehow I didnt realise that it changed +Patch that resolves the need for returning keys from generated batches of insert statements +So this issue is not related to jsf or maven it is just a case of common web dynamic project without jsfnature +Thanks for investigation JD. +The are now running with against nodes and jboss instances +When the datanodes confirm receipt of a block log this event +Isnt this one solved by the new Google projection code +The scanner created by coproc can hypothetically do anything +Task Completed +Fixed but there is still an improper initialization issue that was already there before after the regression was introducedI might look into that improper initialization issue later +Ok thanks a lot +Ive attached a revised patch and a new version of +oops wrong bug +Jaikiran is this still an issue with +Checked in addendum to trunk +Issue still exists in Apache Flex + +The bug surfaced in the upgrade tests when you start with a database and upgrade it along the way dropping a column from a table which has a trigger +Bulk close for Solr +or leave phrase out entirely. +Whats the chance of a release with this included? Or is the release planned fairly soon +The side affect of this work is that we are depending on the metadata that Axis generates for the types a lot more +just a comment that id love to see this borrowed from s in +It sounds like the issue was about the display of the partymgr which is the name of the dispatcher and there because of why I described in my first comment. + Username rst +Before changing the stuff to the I think we should complete the transition from ServiceInfo +I am just trying to get an idea of the context +Thanks Chinna! +What would you like are java objects enough or do you need build file sql scripts +In some cases it seems to work and in others it does not +Makes sense to fix separately +I have a unit test that I can attach but since this issue is closed JIRA doesnt present the option to upload +But I did not get a copy because I currently cannot receive mails via my joern address +For Q we indeed change the interface of Combiner while the semantics for Combiner is the same with Java +Its true that you need to bump up the xceiver count and ulimit on for but once youve done that it doesnt cause big issues in practiceThere are a number of other already open to work on the general issue of socket efficiency eg etc +No harm moving all these into a subdir though Id say named LOG or LOGS as Samuel suggestsAlso yes highly unlikely that thered be a table with a clashing name but I suppose no harm doing the little work to ensure it doesnt happen +The related issue that you refer to appears to have been resolved should this issue be updated accordinglyWere up to now. +Resolving as Wont Fix for now based on the above discussion. +there is not enough info to reproduce it. +Fixed into trunk rev +Is anyone working on this +The other thing you may want to find out is how many Solr threads there are and make sure you have no more Solr connections than there are Solr threads +Also to Trunk +This is the data class passed as a parameter to my example web service and is referenced by the is the interface that is given to the in current SVN +It doesnt seem very attractive to read and discard that huge frame potentially using up a lot of bandwidth doing only that +Namit I responded to your comment about Exceptions on phabricator +So Maven recognizes plugin names even without the dashes +If you have time could you look at instead problem here is that we retrieve the last response from resource map and then try to grab a lock on it +Output from roo perform assembly compressed with bzip +Move open issues to CR +I cannot build it locally because a failing test in impl submodule +does this but does not +Ah I thought your last patch included all the modifications +A text biased parser with a DSL to handle key fragments and type safe keyvalue handling. +Same here if youd like to submit a patch well be happy to apply it +Also I guess it would be uncommonrare to adjust this. +Will leave it open but move it out of +Curious why the XML reading doesnt fail for an empty file +Done r +Thanks for the analysis PaulI will modify the diff command to ignore the lines starting with invalidscheme +Fixed in other set of changes. +I will commit it tomorrowThx +However stock versions of and still use deprecated classes instead of newer +Thanks for investigating +A misleading level of accuracy I dont think its the right direction for +See chapter +Also this wont port well as we add in new languages +Ah OS X +also can you generate a patch by running git diff +New interface created for with release of +When can we expect this? +As a workaround I changed the pom so that it directly references and it seems to thought about it some I guess this is missing functionality or a bug in the jdocbook plugin +DO we need to fix this for or mark as wont fix +Martin Im really sorry this is indeed working with wicketI guess it was corrected withYesterday I did update my with wicket but forget to regenerate my war! So I was still testing with wicketThank you for your help +I am using the input types from the From and output types from the To objects respectively +I agree with all Jukkas comments except for making collections immutableWhere the Collections are private as in for example making the item immutable improves thread safety because the fields are then properly publishedIf there is a chance for concurrent access then the class can and should protect the caller +Im thinking this one can be resolved as Unable to reproduce and then reopened if we see it happening again +Fixed by Aries patch +Then you should either forbid the usage of Timestamp or enhance the resolvage of constructors to support Timestamp + +Im not familiar with the code base yet so any pointers on where to look in the code would be appreciated +I think the right way to go about this would be to have a method that will change the URL path but just leave out the query parameter entirely but that is a far more extensive change than I feel comfortable attempting +It sounds like is working as expected +If you insist on using solution I am not against it +I believe ulimit n is per user not per process +You are right the correct example would beGRANT SELECT ON TABLE t to mariaharry +So I put hadoop and zk up in my own little repo over on +Also it appears from the exception that this is being reported by the JCA inflow +Ill commit patch . +Fixed on trunk and noticed the expected exception is not ignored in ant test output. +Thanks for committing the patch +Closing issue +We ended up reverting the config and now changing the tmp configuration +Patch for +Even with included I am seeing lzo related failures +GreatManually tested +Closing as this seems not to be an issue with current versions +method refactoring refactor method delete to cleanupOutput in on comments from Dmitry Olga and you above I thought there was agreement that this should be store function specific and not part of the interface +Please wrap debug logging with checksThanks + since theres none in Flume NG +I just choose an existing event time out in case that the developer wants to do something when its being destroyed +Reassigned issues to Nick and had to set resolution of all of them to done +Thank you! +This issue can be fixed by add the dynamic import of and in the CXF bundle +Sure but if Avro are production ready I will probably move to use them tooBut from my understanding is going to be removed from Hadoop common so I will not bet to much on that and Avro integration will not be completed on +Thanks +looks good except Id prefer clearer scoping on custom properties like spot price to use the provider idinstead of m using for consistency with +With the changes for it only returns the databaseName +trunk +I tried to implement an for liblinear the class cant be used for this but it is possible to implement the interfaceThe two train methods both return an we should change that so they return a instead +UI should always pass specifyfalse for Isolated networks +Mass closing all issues that have been in resolved state for months or more without any feedback or update +From the docs it seems like this should just happen +It has apparently been resolved in +Thanks for finding this Knut +Id rather leave it out of with a null default +Initial patch tests pass but I havent run ant precommit yet +patches for java for java for and Patchthanksdims +I hope you made time to complete your paper +This may be a browser issue +The attribute was missing fixed as part of . +Since test failures because of order it may not fail alwaysSo it is better to commit in alsoThanksUma +Indicated that jira should be reopened with a reproducible use case. +Can you try with the latest code pleaseAlso do you need to specify all the providers in the list or just the one that causes the problem +Thanks Uma for checking it +I havent really seen a good example of where entity groups are not just a pain in the users rear end essentially we are breaking down and asking users to perform their own partitioning so that we can give them strong consistency between two dimensions that are normally eventually consistentNot a strong feeling just something thats been nagging at me +Eg its better to have all components aligned and consistentAnd dont make the classes final and whatnot +I have fixed few of those already +But I guess this solution will be enough in most cases +The former behavior seems more correct to me +Okay stuff in the code base need to update the docs with examples +The file in the Strut binary distribution is BAD +Yes maven uses +Fixed +recreating patch against current trunk +V of the patch to address indentation problem +Reopening due to continuation of discussion of this decision +Affects only the module + +Next on the list will be which will be changed to support String Document and Source as inputs +It requires the same changes as in trunk plus a sanity fix to commons converters that has already landed on trunk a few months ago but was never backported to on as fix for this issue only accidentaly made it faster and introduced other bugs the patch should be fixed for good +Add the Transactional annotation to the interface or classSince I am using Hibernate its easy for me to see that this doesnt work +SiOK I will take a look ASAP +This issue has been fixed and released as part of release +Thanks DanYour patch is in trunk revision +you should just use the hibernate builds with the codegen not needed for running your apps +after upgrading to I cant reproduce. +We just finished a huge implementation at work which required a lot of OT so I should have more time available in the near future +Btw +What is the preferred way in giraph +Heh then reverse cycle should be a bit more faster +An application that reproduces the problem currently we have tested Grails beta against a using a specific username and password +Performed the following changes Renamed to Added licensing information Switched Frame for Window renamed and now use the same components as I think the default resource filtering rules are good but change them if you wantCommand to run is nowmvn archetypegenerate myapp falseHope this is OK +There are small cleanups still needed but the big is in place. +I believe this is fixed now as I applied the fixes to and took the WSDL verify tests from and added them to the wsdlgen itest on. +After the patch for I think this should work if we disable the Runtime partitioning +If you can tackle this Ill be standing by +The idea behind is to remove the cache at all probably it isnt needed as a CDI implementation should do it for us +and ant test passed on my machine +is watching the directory fortimestamp changes instead of and even if Ichange timestamp for the directory nothing happens +You dont assign we do +pig secret x localFind hadoop at Usersrjurneybinhadoopdry runHADOOPCLASSPATH Usersrjurneypigbin +This patch addresses this issue +There are still some asymmetries here +This may need to be rebased if is committed first and +Patch generated from the right root dir +The new patch is on top of the previous patch for branch +Patch looks OK +The console output is from a old patch +To be deployed as a separate webapplication? Interesting idea that could work very well in multiple cases and leave options for people to do the skinning the way they prefer either override specific style classes or modify the war or the exploded resources +So all issues not directly assigned to an AS release are being closed +Whats the value of having Context Path in the error message? You can spot it in the address bar of your browser +Your patch achieves the same functionality but wakes the thread more often and needs more variables +Thanks Devaraj and RahulI put this into trunk and +Is there a work around for this issue? Some people are stuck using maven because maven doesnt work with Andromda MDA +Issue confirmed +Thanks for fixing that so quickly. +Ive made the following changes Previous popup menu items were removed along with all the shortcuts +I ran locally with patch v and it passed +I just committed this +Oh i agree I want it is fixed one way or anotherFunny your not seeing all those error that were mentioned but then patching is actually a pain unless everyone doing it is in sync Lets give them a bit of time to review and do their dev thing +Ive committed this +are there any other applications for this exceptAlso is it known how internally the scrolling is implemented +removed in rcorrection +But after a while or after lots of memory usage this can happen yes +Changed from bug to improvement +the modified files from merged +Verified both the code and the doc changesThanks Bryan! +I believe the consensus was that only a beginner would make the mistake anyway but it does no harm to check for it + +The runtime now has an icon +Anyone has a link that work? Or can explain what this is +Justin afaik we have had this support for quite a while +Should be fairly simple to resolve so Im targeting this one to Beta +Kristian Rosenvold Thanks for that +I will make the api more and more similar to the html one +This last release of version happened on Oct +This changes the semantics for now we can define the behavior as a bug but if someone is depending on it for some reason we need to change back and just add a second class deprecating the first +Also r. +IT added +I talked with Aaron Patterson and he doesnt see a reason to reject the tests either +The issue where this was addressed is +I will also have to run the tests +But I do not understand I can access the EJB with test context set from outside the JVM +Yeah I cloned the old one but I left a comment below it stating that current is +For Spring and JSF together I think that Webflow is probably a suitable +commited changes to the +In this code why did you use addmethodsignature and becomejava at all? Im not quite sure that becomejava works with extension yet Nick Sieger was going to add it but its complicated and we havent gotten back around to Think your case should work with just extension and no addmethodsignaturebecomejava stuff +Jieshan So what do we need to do to close this issue out? What do we need to apply? Thanks. +unless different actions need to be taken for each signalat the end of our signal handler we have to call the previous signal handler for that signal +It doesnt need to be a course in statistics the existing code isnt either but even a sentence or two on why is using it and perhaps a couple comments identifying the variables would be helpful +It seems that without user defined compound primary keys its impossible to go to change the affected versions +I guess i am the very first person who get cache cluster to work using component as this bug was not reportedfixed before +Theres one issue with this patch +Could you please reopen with this failure or create a new issue +verified on JBDS Vlado said issue cant be verified for JBT CR because of menu items +Patch wont be ready in time +But there several sites where it is documentedInternal but as I know it is common practice to use internal of Windows to build good Windows apps +However in Solr we DO expose the spellcheck box under select and it does not work so I really think we should spellcheck should be in the defaultIf this is the main concern a better solution is letting the GUI disable the spellcheck checkbox if the selected handler is not configured with the spellcheck componentDont know if it is easy to get to such info through clean REST +found a problem on my previous patch the list of build queues in the edit schedule page cant be added in the Schedule +ruby have not yet fixed at present +However the code was not very transparent so its possible that I made incorrect conclusions +Tested now after recent jar upgrade it is printing a nice standard error no JVM crashes. +The existing functions in the client side will not be touched for now +Protobufed the call Changed some error handling Wrote test code into doesnt actually trigger a major compaction +Any specific reason Jimmy may be am missing what is there in your mindI just meant not to set transition to true after try is called each timeif we do this inside the try method +You reduce the size of source code but it is more complicated to read it +its fixed no +Theres a couple ways we could do this +Committed to trunk +Len can we get some additional information on this one +For whatever its worth this seems to be working for Hui who is posting some issues now in this same forum +This could be an entertaining project and might easily be useful for testing as well +Finally committed these +clearing the sortedTuple before adding the batch row allows the results to be built ve installed the last snapshot of Teiid cr and the problem seems to be resolved. +Looked through code some conflicts with patch for only should be at AMCIMinor note at we already have String role field that stores component name +As we focus on PBR issues related to demo applications will not be resolved +When we optimize a checkcast to a trap should we be modifying it to a new call set caught exception and goto? +I found a problem in the classThere is a double loop which prepare the parameters valuers for the query +The unit test actually fails on anassertion that looks unnecessary and I dont see theOthers above seem to be happy with the validity of the issue at stake here soIll carry on a bit more later +I dont see any problem in the AS configuration +Should now properly handle absolute paths as wellNote patch also contains an update to that excludes +Status please + Zheng! +Closing resolved issues for already released versions +HOD can create a directory thats specific to the user something like mapredsystem where clusterid could be like a torque jobid +It think its adequate to do this only when deserializing the entire adbobject so it is easier to place in the xsl where the adb object deserializing is invoked +This Jira has now been undone +The commits are and theres no need to include anything from another merge +Yes its there +As I side note because you are not at an official release yet it is understood that you may change the interfaces +I dont see any issue with restricting it to Database I look forward to getting rid of my cast +So we should not ignore that encoding +See attached OutputI cant send you a tcpdumpethereal capture at the moment because I dont have the necessary permissions and definitively wont get them to provide you with this type of output. +Maybe we should to use directly +But following negative test cases failedLooks like need to update +Duplicate issue covered in +Then perhaps in require it +I think of a union as a struct with many optional fields with the assumption that only one will be presentYour last paragraph make it seem like adding values to enums is potentially dangerous +good call +or +Attached patch that adds support for registering custom instances through Marshaller and Unmarshaller +Sorry for the delay got behind on my email due to some vacation time +Virag says it has happened before and reattaching might make it pass +Ive fixed it in the source control try downloading the latest from svn and see if this works +bq +With this Hive build it appears the actual build is crashing as it never gets to the point where the build completes and it provides the ant build report +Patch applied in revision +Im thinking about to add some or to throw just new +Done +Refactoring patch aimed at avoiding code duplication +Cesar this isnt about changing an error message +But did you check it? Does it work for you +enResolving as duplicate of youSandy Ryza and Arun C Murthy + +But in the mean time just add receipt request for all your sends or at least the last send for each destination your sending to +Ok Henning just because I think you need some more work I did some more fine tuning of the algorithm and deployed new snapshots +Patch checked into trunk +This was procedure I have used installed Tools nightly builds on fresh copy of Eclipse Indigo created simple JSF project imported released Richfaces libs guava cssarser sac imported new tag lib to the palette added some an panel component and nested calendarSummary everything works as expected +Hmmmm in that case Im not sure how you can resolve it +This is resolved in cvs +fixed with unit tests checked +done +Works fine in +Please see for development mode output +The reason was that was not read from property file but it was read from JNDI context +Is this issue now fixed as a result of closing +Maybe all we need to do is fix the build and the build will come as a byproduct? What do we need to do in so that it exports those jars +Here is some first try for the dictionary and a abbreviations file +manually set the test cases for now +These blockers and critical issues are resolved but not verified +First Patch moved Button on Top between both columns uses darker ColumnIll see if i can integrate the often folks just need to see the tokens emitted not the offsets raw bytes or other attributes +Confirmed and obviously represents a gap in tests +changed to avoid retrieving a property that is not yet connected to its parentthanks a lot for reporting +Fixed default value is removed +Damn +It looks like the composite classes have been removedIf they are not going to be resurrected this issue can be closed +These blockers and critical issues are resolved but not verified +But this issue and fix has nothing to do with how wsdl and xsd imports are handled or resolved +Thanks Tim! Everything is ok +When I set up automatic patch testing itll be required for the patch to apply. +Perhaps the same event types as the event types where a watcher or a project lead is notified on +Ill try this but only if it does not prove time consuming +Sean the implementation is a variation of the one +Done +As soon as I specify a runtime I see a problem where the exploded launch repeated runs with no action from easy to reproduce just create a project using from trunk then import into Eclipse then specify a seam runtime +Both of them have some history of failure +has been added +New Teiid ConnectionDDL Importer should work for this import if is of type JDBCWell need a running instance of Hive to test with +I successfully ran several terasorts with this patch +I think ive found that problem with telnet which is not a very conformant IMAP client +I just committed this +Jacopo this patch cant hurt +Suggest to incorporate once Geronimo server start to make use of. +yes I think it will work out OK to have flush return the number of items that were flushedIll create a JIRA for the issue +Is it suitable to always return the same with no scoping attached to the caching? For example if this code is running within an app server process and there are multiple JEE modules that have services with the exact same name could returning the same cause a problem? Im not sure of the answer just wanted to pose the question +and ant tests passed +The finish buttion is not that should definitly only be a warning +Fortress another Sun project needs them as well +Seems there is still some clean up to +I could not get it working at the first tryThank youM +Done +I refactored to utilise the and optionally the so that the would utilise gzip if the endpoint has been configured with the gzip feature +Not that this means it isnt an issue but it does give me reason to pauseWe can certainly keep this issue open but Ill just say that any solution which involves setting the property is likely doomed to fail +I think Robin was just waiting to try this on a bigger set of data so assigning to dont think this is a release blocker +Patch for Y! distribution +The existing have been updated to cover this new behavior when using Git as the SCM provider +It way too specialized behavior for an shelf component. +suggesting attached patch +of tld into palette is part of Tools +This issue has the same reason withFixed on trunk with r +We do have more flexibility in Designer via extended model properties to inject more metadata into any model +Removed conversion from String to Boolean +pushed to trunk +This work is done the code is checked in on the trunk +being pedantic should be wont fix +Ive switched all mine to volatile and it actually seems to reduce it but I havent been able to say its fixed +Has anyone talked to the beandoc team about maintaining an official POM +Thanks +The patch does not introduce any new Findbugs warnings + +As we now use a single line command +Not sure if I can be of help but seems worthwhile keeping this on the radar for +So Im linking this ticket to +Can clearly reproduce it +has just been run in the master but the outstanding catalog scanner is not seeing new entry either because the next took too long to come across or because were using caching in our +Now with say third time is a charm +Applies to with p +Nope +verified on JBT +I find it strange that a file gets locked within just now discover one more intriguing factor +Committed +Implemented the clone method in the jcrresource bundle along with a unit test in Rev +But since I used the provided under I need to transfer the adjustments to the trunk files +This issue is no longer relevant for it is fixed inChanging the affected version to +Committed revision . +Would it make sense to update the tools +Reopening for review +fix for jdbc in tests and improvements in r additional rows in ack table to track last ack location per priority such that the store can be traversed in both directions to retrieve messages +I am okwith the tradeoff of how it is currently implemented vs +The patch looks good to me but I would like to see a unit test for it before I give a +Given it could live with the current approach it may make a good opportunity for an external +for the new patch +If you are running you need to extend the classpath using the instructions in the example files +Have you tried SLFJ with logback or later? + +OK I think I can do this +We need to setup deployment to Maven is missing distribution management information +as part of this ticket can we rename +I assume youre referring to the JVM in debug mode? In any case can you please try to reproduce this in a unit test? That should be possible with a simple and your location patterns and the test VM run in debug mode +Marking this for M so that i dont forget to fix this one +It seems I cant get it to fail explicitly because of the problem with frame reuse +Marking as critical as EC Image support is not usable right now +JIRA really needs a rename file feature +The same ZK issue can affect the controller and the consumer too +Hi GrantIt looks there is a missing class in your patch the class was in the test tree +Thanks for the patch! +I dont see the reason to have at all +So I did not run this patch on my local machine +But I think it works now that default is the new default in only use the out of the box default config and this quickstart seems to work as expected +Brock This patch is not applying cleanly +hidden text is in the buttom of the text means the text in the pdf is transparent +the loop is exactly at thats why I cut it there the last two lines repeat forever +Uploaded a new patch +This is a perfect example of why I extend the generated impl class +this file can be included in a classpath +This patch and issue are obsolete. +strachan can edit the appropriate files and kick apache. +bulk close of issued resolved in after X release notes were generated but before XDT finalized their release. +OK but why do we have a folder on the repository with a wrong file generated +Roshan you want to take the issue +This would be a conservative approach towards improving the codebases path handling logic +Unfortunately I dont know how to reproduce it so at the moment I cant confirm that it is fixed +Now if an edit log fails to write that doesnt cause the image dir to immediately be marked failed so its more likely that it will be discovered at checkpoint time by the +EAP start and stop works fine without having to pass any parameters +serverstandarddeploy +patch looks good +Trunk patch that compilesI couldnt get it to fail +ESB report is +Usually if a issue appear in the current trunk we first need to apply the patch to the trunk and then merge changes to other branches +updated with trunk +Slip to Beta and assign to Len to review or delegate +we cannot construct an absolute url manually without the users help because we do not know the proxyurl rewriting scheme they use +Thanks everyone for the comments testing and reviewing +Otherwise quite a few tests fail +Attached the patch for what I committed patch to resolve defect +Fix done per work log +I switched on logging when a block is allocated when the namenode receives a from the datanode for a block +fixed with rev +I also saw the failures in tests +Is there a gapI was trying to use http but I ran into issues as detailed in +Just curious is this a real world example +Rebased again after commit +The cardinality restriction of +Local beans are not required to be Serializable per CDI spec +This issue is fixed and released as part of release +In summary this patch should address some of the trigger recompilation issues that I have encountered +Bulk close for Solr +One errant MR task spamming connections to the NN will trigger a for other tasks running on that node +bq +Sorry didnt see this notice. +Wow thats a lot of pain for a small problem +chapter +Verified +What comments are you referring to? The comments on the various field types for int tint and pint are pretty clear to me +and their development stopped some time ago +The problem is as Robert raises deprecating the code without providing an alternative isnt that user friendly +patch containing test case ant runnable extended your files to show the issue with the current code +DavidJozef can you confirm +Marko let me know if you encounter any other issueAlso feel free to create any improvementfeature requestI hope you will be able to use sonar on your dotnet in version +Also Id get rid of shell scripts if possiblePlease check that the patch was applied as you expected +Never mind +It was needed and is very much appreciated +Looks ok builds ok samples and tests build ok as well +Here is the modified class based on the release src download that worked for me +When you say you are calling them sequentially does that mean you are waiting for the success callback of the first to be called before going ahead with the second transfer +to trunk thanks Nicholas +No need to change anything +Im going to try and dig on this and will figure out whether to close this as a dupe or not +Duplicate of this mean that the thing has been solved in and that now the generated code goes to? +Can you test this out and let me know if it worksIf you dont have access to the sources I can build a jar with the patched class in it for testing also +Quality of the completed integration +Characters that cannot be directly represented in this encoding can be written using Unicode escapes only a single u character is allowed in an escape sequence +if it needs the model from jsfnature it should not throw errors but jsut not expand +I have no idea why this was final at all +Closing. +I applied a quick fix please close this issue if you are satisfied +Chad could we get this ticket resolved first eg a component that is similar to the existing +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be as Wont Fix since theres not much that can be done on the Grails side +Ted If you run wo this patch that test passes and then if you run w this patch it always fails? That test seems totally unrelated to whats going on in this patch +Whoops attached the wrong patch this morning +and its VM counterpart are in the RVM codebase and Classpath CVS All appropriate uses of have been replaced by calls to. +Done together with. +So although it was closed I think something was fixed in this issue +It is defined as a static variable +Fully fixed by r and r with some cleanup in r +We will generate the section properly +Sorry I cant respond sooner +Ill see if I can cook something up though that is definitely the first step. +Applied in r +from me +it will alsway be falseRightLooks like patch v is good to go +For SOA this should have been documented as part of need to make sure the documentation clearly states that you have to pick your stack +r +Moving to. +Need this for like deepal added a dlb to javawsdl +Container and container are the main migration scripts +Im happy to do it thoughThere are relatively few direct to the property topics in the Tuning Guide itself only about that are not from other property topics +updated to for catching this Anthony Id upped the maxpermsize in cargo but not in setenv. +Im closing this unless Bill as Fix Version. +Thanks for the patch JD! +I also spotted a couple of fromfor somebody named StuartOh and do you happen to have any unit tests for this +Heres the patch for some minimal reentrant locks for the branch +This has been fixed already. +The steady slowdown with a temporary reply destination was due to Eclipses console window slowing down the logging activities when running the test in Eclipse +revert use of rat plugin and fix occurrences feature updating +HiIn the current implementation there is no method to remove the message store from the synapse configurationthis patch will resolve thatthanks the fix for Sample is long overdue on my end +Test failures are not related to this patch +railo does not remove any character in this case can you attache a example file you have a example +Adding patch for the for the patch and the link to the discussion Matt. +I removed the patch prefixes uploading again to trigger patch is good +Fix is in CVS + may be the same problem +Sorry Michal I didnt refresh the JIRA page before posting my m glad the fix works Ill try to get a unit test working as well before issuing a PR though. +I have a patch let me run it on cluster and post if it works +Verified by Sean at r +I added a component to the JMX support +the idea being that if there are multiple mapred clusters we can have some control over what jobs go wherei dont think this is a dup of +no update assumed fixed +new version with unit tests and other minor included some unit tests and a default implementation that throws a not implemented exception +Im sorry I dont have simple project for it +I have upmerged and fixed the issues +Nice and simple design +Yes I think that this has been fixed in refactored template generatortransformer +The bug is initially described in +the updated patch looks good to me +And feel free to attach patchestests too if you can +Decide whether or not to take this by release of +New connectors and changes will be encouraged to use this new mechanism +That works for me +This issue is a key one for a demo I am trying to put together +AhmedSid Any update on this one +But if it turns out that we need RPC changes in the minimum we need to change the displayed to hostnames on web UI and +I see no relevant changes but it seems odd you did not get a parse error with an empty file previously +Please let me know if you feel strongly that this needs to make into. +Is the write cache on the SATA drives turned off? Oops it wasnt +INSERT and DELETE templates + +I updated the patch to the current trunk and added some final keywords here and thereIMO should be final but that is not much of a deal +Max +Can you provide a patch against? I dont want to introduce new features to the branch at this point we need to prioritize stability there over anything else +Shoudlnt make a difference for the framework except for making it slightly bigger +See previous comment +joram says he thinks he already fixed this +Did the error message you saw come from the local file system? Please try using cat instead of get +No +Donation of code from! Id be glad to see an updated README that would describe how to work with it +well check as it seems identical +I will on that and focus my energy on the Java version because libcloud is way behind jclouds in terms of features and it seems like the development is moving forward slowly +Discussed users need a nice simple way to start up +Zip is attached +You do not have to make any config changesTo work with secure cluster you would have to set the value of to true in and set up approriate knox key tab file andWould like to commit the code so that we do not lose itIt would also be easy to review with phased commitsThe risk is low for any existing functionality +Verified on SOA ER. +Patch for consideration +Warning aggregation is turned off for local mode +theres no jackrabbit anymoreLahiru +patch for the input buffer issue +There we go +Sorry Igor it looks like I really had some build issues +Hi Mark thanks for the review! +I know there is a we can check on the assembly if we want to provide warnings at least +Inject executors to servers needs some further work I dont see being a priority because its not really something users shouldcould control +It is superceded by. +Updated to into the master branch. +Well when I put that back in the inheriting page flow directory it is all working fine +trunkrequirementsHow do test know in which directory to find the requirements? A system property? A convention in tests +Is that your argument for not adding a version? Because if there has been no discussion of change up to this we wouldnt possibly need to change the format in the future +Christian I think you have embarrassed yourself enough for today +committed +Looks like theres still validation issues with choices. +World readable means that anyone on the node will be able to grab the ticket and impersonate that user +I can demonstrate this with a patch if you would likebq +Correct filename +a But even here is spring time +Filed. + +Schubert Nice one +as its possible to a have a null when loading collections hibernate cannot decide wether null is actually changed or an error so we allow itso what is left now is just to make that error message more specificunderstandable +Couldnt reproduce by scrubbing sstables produced by performing more tests I realized why the problem could not be reproduced the scrub process hang on a different CF and afterwards any scrub operation would hang until the server is restartedI was able to narrow down the problem to a specific sstable I will soon post it along the stack trace +Shai looks good to me! to allows covariant return types +Changes made +Patch looks pretty good to me Henry +I appear to be getting blamed for javadoc warnings +Its also actual for version nodes works fine with andCannot reproduce the problem +Pls give your comments +Maybe the additional object creation cause a performance hit but is justified so we can ignore it. + +code delivered. +this is not a blocker for beta imo +Hi LowryDo you need anything else? Or can we close this JIRA? +It seems this issue was fixed +See also +como faco dowload dissohow to download to use this +it only restricts gossip which isnt A Big Deal since its so little trafficso how about this? we try to parse gossip packets in the current version if that fails we parse in the old version +Great that was quick +Better might be to include a that describes how to build the web site +This is simply a fact +are those left over files from another runNo I just copied the entire directory from the live running system +Thanks! +Delegation +Like I said before its not a particularly big deal but it would be nice to have subversion disregard these files +Can you provide a reasonably small example application that reproduces this problem? The description is very vague to really investigateThank you +moved original comments to a separate file as to keep formatting +Thomas do I see it right that introducing hidden nodes would mean only small effort is required to change existing persistence managers? And if we only introducehidden nodes once we have more child nodes than we would put into a node in the there would be no performance impact for deep and thin trees right +Does this require any SPI change +A Patch changing the interfcae to support multiple content parts +Thank you for your improvement +I suspect there is a race with the default services that are installed by the framework + +This issue is not approved for SOA CP +Also added the localization in the appropriate properties file +Is this a hard limitation of cglib or just the due to current choices in the implementation of the pointcut matching algorithm? I am having some difficulty locating exactly where this matching occurs I believe it happens in the method +table? This is hbase? Enable DEBUG log level in case it happens again +But for the case when I would like to erase this info I should be able to correct? Once that is done it should take me back to configuraiton page +Fixed in trunkr +Thanks Vinod! +Part of the envisioned use of this domain is to allow corporate entities involved in Subversion development to communicate with the broader user community in addition to just developers +Can someone clarify which patches need to be applied +Ill take care of the just need to know which component new should reside in +Console log with changes to to correct the message and to convert incoming values to mm types. +I will have a fix will upload a release early next shouldnt resolve this until we have a commit with the upgrade in the AS codebase +It only prevents a previously uncorrupted table from being corrupted this waySo I just want to confirm Did you newly recreate the table using? Or was the table already corrupted and you were hoping that the patch would fix the broken tableThanks a revised version of +Im going to commit this momentarily +Werner Ill run my tests against and verify your results +Some super minor comment for The patch removes and adds +Reopening consider assigned to me while I check if this can be considered obsolete +JarekPlease go ahead and commit to both trunk and branch +Fear that this may have missed the boat +we should close other things and then close the lists +If anyone would like more info on this work around let me know happy to provide details +Closing all ResolvedFixed issues lacking a Fixed Version +Looks like you pulled a more recent Ive been tweaking those +I committed an alternative solution following the change done in add a field on and store there that value +I just committed this +this one is fixed on the CONFIGREWRITE branch +So the patch file currently uploaded for was used +occur even when not in debug mode. +One thing you could try would be to set the http session timeout down to minute in +make the programmatic API use the Trash +core namespace +Thanks for reporting Jun I can reproduce this but Im not sure yet whats happeningWhen persists to disk or it asks parent class to populate a containing all schema elements including the unique key field name and then uses to convert the to XML +Issue is reproducible in +The proposed patches would definitely have broken the tests as they test all combinations and do not expect an exception if both flags are true +Yes +I did not test yet please add your comments if you try it +Added to branch too on second thoughts +Try running without specifyingI have already fixed this in my sandbox and Ill include it in the proper version of f along with one more test and some additional changes to the testsutilities +Added a release page +Thanks for the report +Javadoc comments to my satisfaction +What is an example configuration of an app that does not work +Closing all resolved tickets from or older +Digging +This issue is old and redundant and I believe it can be issue is old and redundant and I believe it can be closed +I will create new ones to finish the steps I outlined previously. +correction to previous users passwords are password +There is a fix scheduled for but it does not work for me +patches for and +We have no way to track this currently and will have to be assessed +And with the pull request I didnt need to modify the +This was already fixed in webwork IIRC +Thanks Jons create a new issue for the new disabling test case with the offlinedisable regions and discuss thereFine +That is a issue of WTP tooling +local storage is enabled during zone creation in the performance test bed +Now you would like to point repository test to a different repository +Certainly too confident in others work +Any suggestions? Jim Starkey Inc +One minor nit it might be nice to add a comment by get and get as to why they were deprecatedOtherwise looks good +for emitting events and letting other systems storerender them +AndyAnything I can do to help move this along? Any of the instructions need clarificationIm here to are having a lets roll +Hmm +Will be much more better to run it via one maven command +Covers my questions nicelybq +Negative test case for a file len can be added +I checked it out locally and it does exactly what I was hoping for +Ill keep this open until I hear back from you +Attached is a minor patch to the command line reference page to add a note on MAVENOPTS +The findbugs are known issue see +So back of the envelope math is that youre in trouble if you havent tuned the operations threshold down a lot +The PDF I used toI can reproduce the problem but Im not sure what you really removed +Manually tested but Due to could use the red color to display the differential measures The link more doesnt allow to keep the selection of the differential testedPerhaps parenthesis around the differential measures are useless +Im sorry but that file cannot be distributed as is +As for catching errors to treat the situation as no cache that is not currently possible +How about we make one exception in the case has only default construct +This seems like a pretty major issue any word on possible resolution +Marking it as fixed +Id say go ahead and commit +Any chance you can port your unit test to Java then it could be included in the test suite to validate any future changes +When you have the sneak preview active every time it refreshes a newline is inserted in your edit window very annoyingMy environment Firefox Ubuntu Linux IntrepidIt seems browser specific if I use Epiphany it works fine +Since the wordcountsetup script uses the thick client it wont work until is resolved +VMI is VM library not classlib +It does not appear to me that the fix has been ported to the branch yet +Hemanth this was marked a blocker but had no release assigned +What do you think about this Alexey +Sorry for the confusion I switched to the git copy of Jena recently +Ok Im feeling quite stupid I havent seen this mail +added this should fix your problem +Made the set and get static and did not use a member variable to match the same pattern in Client for set and get functions +Very handy I hope to get some automated testing for this into +Like you said no one reported it +Thanks +If we do this we should keep the feature inside the core but that is already plannedwanted for the core of the tools anyway +Graphs on ganglia +This will cause a dead loop +Ill volunteer as well we think we have the same issue +Problem with the demo setup installation When the installation failed a first time after the hsql starting the next installations will always failed the target cannot work because the hsql database is still running +Oleg +That ant script can then call our java class if necessary. +So retracted the boolean +I agree its good to cleanup the existing tests without timeout but my concern is that this issue is to mislead people to thinking therere missed timeouts in their newly added tests +VikasOK thanks for comment +This has no relation to prev block report +Am glad you guys figured it out but I have the same question as Pat as to why we dont see this in +There have been some attempts at using a localised format but none succeded yet +check the history log file format is still compatible and make sure the parser can parse the log file properlyYes +The purpose of the patch is to make experientation easier to use another implementation of List +The is removed now and after careful consideration it should be correct that wayNevertheless please crosscheck since there is next to no automated testing possibility for this issue +fixed using a zipfileset to include in the distro zip it also for the file +Or was there more stack trace thats been removed from this issue +Pages still do not persist properly but that no longer kills udig. +For grammar parts that are composed of optional elements well need to check that each element is used only once and for the parts that are composed of elements well need to check that each element is exactly used once +Enjoy +I downloaded and installed the patch into the railo lib directory +The test looks like it is passing but without any exceptions ever being caught in the test +Can we expect a patch before friday +Implemented with filtered content assist and Select class dialog +Fixed in version +From a users perspective it should work just you mean throwing exceptions for channels in +Successfully tested by hand +Im using which needs the MREPO variable according to their installation as I said I have no problems running any maven command in the more ideas +convention objection because multiple persistence units could be defined in +Fixed on trunk r +However using an exception handler in step operations that swallows exceptions unless theyre fatal is an extremely fragile approach +We should institute measures so that the broker monitors the amount of memory it is using and tries to prevent ever getting to the point where it can no longer allocate memory +To clarify the havoc was generally test havoc recovery always moves to a new index folder so issues with the old index folder are not a real problem caused strange things in tests though +renew available +To JinglongI agree completely with you I just wanted to be sure that any final solution provided a generic solution +bq +Travis took care of this back in October +Assigned to blackdrag as it probably needs a code generateor change please see the comment Im about to probalm seems to be that the compiler cant set the owner of the Clodure as the owner is not an instance of a think that the best solution is for us to have a which takes a Class as the owner and for the compiler to use that in this do you think blackdrag +Architecture changes are too great with runtime environment and designer integration for moving to +This patch writes new in fstime file if a directory becomes unavailableIf writing of the fails then this directory is going to be removed as wellwhich makes checking of consistency of the edits files lengths on startup redundantInequality of lengths of two edits files does necessarily mean inconsistency of thestorage directories + +Thanks +Didnt find appropriate version in JIRA for Hibernate Annotations sorry +I had to first understand precisely how it was being used in analysisChinese but now I get itI will reply back soon +And I have tried to use some parametered types to see if it is possible to make index readerswriters typed by the index format +bulk defer of unresolved bugs to +Im marking this wont fix though I could be convinced otherwise +was specfically for online snapshots and was for offline and restore ops +Raised this issue in the depencencies section +However the msv component is a seldom used component +New patch addresses comments +Thanks guys for helping! the issue still exists +Making a block immutable makes appends and truncates very heavyweight +Im going to see if we can do a version of this that does not depend on a write delay tho +Works for me now +Here is a unit test that exposes the problem +on my machine and on the hudson node +I believe that this list is not complete +If however I start couchdb and then make a replication request it always fails with the above error +Committed updated patch as revision all resolved tickets from or older +From what Ive experienced this relates to transitive snapshot dependencies +Combiners code moved back to mapper of the same task +Currently we use the csc recurse option +JacquesPierre was a trainee at Nride and is no more working on this +I assume so the javascript error above has only occured when I am using weinre +Can you review this patch before testing in a real cluster +But this gives you an additional chance to catch your mistake besides just the regex plus its just a null checkoptimization if you want to look at it +Yes I know that the original exception is lost +Upon receiving a failover command the following sequence will be followed Stop the master Flush the buffer Send the failover command to the slave +Since this jira still talks about VIP and and solutions which should be resolved as the implementations a very different for these approachesIn my prototype I did not introduce or change a single config key +why does the metadata have a buildnumber and timestamp if uniqueVersion is falsemixing the two types of deployments will certainly cause problems +This solves the problem of reporting errors on ringmaster side back to the hod client problems are still NOT addressed +But there was one more NPE +Now I cant reproduce this bug anymore No warning for JS in the console the problem with rename tab dashboard actually not relate to JS and its been fixed by +Thanks is now so probably cant check this +I have not spent much time look at why though +After some more experiments and further checking it seems that the only information available to the jvm relating to the codepagechcp of the OS is the IBM jvm specificIm changing this to an enhancement request perhaps we can modify the ij scripts so that we set gets set in accordance with the codepage on the OS by default +In this case it the error is a orgapachestrutsutil and is distributed in struts core + This bug has been marked as a duplicate of +This includes a comment on nested bracketed comments +avoids it in most places at runtime +As it we dont fully support in the connector we support a subset of the syntax of +Lets put the time into instead +Log in the Servers View before starting the log in Servers view after right clicking on project and Run As Run on to blocker since it now is seen by more than one +Ram please review it too +Thanks +bq +code looks good +great now thanks +the new patch makes much sense +I try to add the functionality for curve rules in a day or two +Created reviewboard +I removed the under and deployed again it worked fine for me +Actually cant replicate either on FF on OS X +As Thejas point out the right approach and original design is to fail over in such case +I modified my DNS GSLB +This patch fixes the bug and contains a unit test for the fix +Fixed the numericnumber based strings +The problem comes from the fact that filters loading is delegated to that doesnt use the project basedir but the current directory as of included plus an integration test the current version of Maven WAR Plugin uses not +For what its worth I tried this on some servers that have concurrent connections with at peak requests per second +also moved initial state transfer onto the custom streaming logic. +If you want to restart processing then a throw should be used. +the patch is stale and does not apply cleanly should get be in? should we file another issue to fix the skipping feature to be +Gary is this feature for only or will it also work for JDBC persistence adapter +After working up some test cases and examining the code more closely it appears that we cant optimize the case where reversedtrue since it starts reading at the last block anyway and when it gets to the first block it must start at the beginning to find finishColumnAm I seeing things correctly +Update please +The hadoop tar build spends most of its time building docs +Thoughts +Current bug fix causes SQL Internal Exception Unknown column TIMEZONE in field listError Code Need to remove timezone field in as JPA looks for this field in database +Patch applied +On eclipse and wtp all menu works fine + +Fixed the source of NPE in the class +was already active on live site. +OK I changed the cocommit to a TODO and will commit this soon! Thanks for review +i just tried the same whats given in u please help me Bad format kept beating my head against this problem until I looked at the example for writing a one more time and realized the big difference was I was setting the field type to and not string +Might it be useful to promote to a public class in io package +That means that I also have to manually interpolate the values I find +I managed to reproduce the problem with the following simple WSDL is the abstract WSDL +for the quick looksee andrew +Verified on +This can be done similarly to HTTPS integration tests that are effectively the same as HTTP tests just with a different setup +Hope it would contain some useful information to diagnose the problem +Thats what I did +every now and then +Fixed in both ways discussed on IRC +Closing the issue now that it has been resolved. +gets hammered with connections and edits need to go in somewhat promptly +Any chance this could get changed to instead +The limitation of value types not being stored is due to the type being which is based around how Encodings work not due to some kind of fundamental mismatch with Lucenes business rules +This version revives the summary line for the istat feature +Punting this issue to the release +Thanks! Any documents +Now using needs to stay open because there is code in the for this that needs to be removed +So in the same query you could have some aggregatesfacets calculated and some calculated post collapseOne of the central ideas of this design is to allow for pluggable collapse and aggregating functions through the mechanism +is this ok with you +Agreed +Will fix it +Thanks this to the branchThanks Jay! svn ci Sending containerSending itestsSending Sending serverTransmitting file dataCommitted revision +Any further comments? If not can this be checked in +Ive assigned it to you +It definitively looks like a race condition when deleting either databases or documents +Closing issues in +Would it be better to add a new method signature with this improvement and deprecate the old one +PrakashCan you kindly provide your opinion +If youre indeed on it makes it much easier to understand the problem thats all +I agree with Julian having that functionality in a separate bundle would be much better +Joel patch v works better and I see why v didnt work out +I therefore expect that marking will not create any issues +bulk defer of open issues to +i think you should make those analyzers final too +That field does not exist in +Did you every get it working +Ill mark this JIRA as resolved now and we can either repoen it or raise a new one for further changesRegards SimonIt will probably be a few weeks before we get our stuff working with the latest Tuscany +Should the issue surface please reopen +EAP CR +Ravi I just checked this into and trunk +Suppose appending is not supported can we still try writing file in Fuse as the work around you describedThanksRui +I like this idea especially because it maintains existing failure semantics +Initial Patch added payload support for Improved javadocs made field delimiter configurable added tests forOK I didnt realize that was the case I was guessing that thered be somekind of cutover point probably where we decided to move development prettymuch to trunk and start backporting fewerThat said things like diamonds etc +Will release after final is out +Trace you try the latest to see if it fixes the that I need a log that matches the thread the forum thread for more explanation. +This was the main reason why I asked for a review when I first mentioned it on the mailings list +Then you have some other problemThe ONLY way the issue is an ISSUE is if VERY LARGE objects are referenced by the +The root cause is placed in +Confirmed that the probe multicast address is appearing as required for EAP. +Will open defects if any encountered +Generated files for release with diff markers +Resolving for QEIf anyone wants this added to Central please open a new JIRA +Thanks for the patch +For the purpose and will be packaged +Do you really think it is necessary to compile with always thought compiling with DNDEBUG was standard practice for release builds +Ill be back + This bug has been marked as a duplicate of +This can be improved as mentioned above although yet another solution is we can add a global reclamation thread to periodically purge this table +Fixed in efbaacadbfabd thanks for the patch +There were no exeptions in log at all +The file contains new html ouput files for review +This patch introduces the property enable with the default value of false +Regards security as first class citizen rather than as a CP would be good but probably not in scope for Id say +Mostly likely UI issue +Have you found a decision how to solve this? Or wouldnt it be solved +Bug has been marked as a duplicate of this bug +Expanded the subject to include all cvs lists of the logging project please add all of them to the modmbox archive +It seems its only for web UIbq +It is not clear to me how much sense global requirements make once we get to the point of supporting multiple resource types +On Windows XP it fails with the error you described +I have changed the master file for jfoundation accordingly +I committed this +I think is ready for testing +Thanks +The patch is part of. + +done +Also it is unlikely this is caused by +This is also an improvement for the current and future versions +Ive just committed this +Fixed in the latest release +Right +This is extremely odd +People can still pick them up with an import of JAR but it removes all confusion as to whether these are things youd use in productionIm not sure what tests are needed here to round things out +Will do +Great just to clarify though are those lines added to the cluster properties file +Hi Pedro its gone! That was quick! Thanks a lot. +Moving this issue now to +From EquinoxpMetadataGenerator it seems it was deprecated in in +OK I wasnt aware that the +I think getting a patch and ldap support is a great idea although i disagree with the sentiment if you have an argument against ldap but are making a case for zookeeper every argument you make against ldap could be made against zookeeper +why is it not final +Verified on stage. +Thanks Nick +Closing old issues +From the JDBC spec section we haveIf the type or number of results returned by a object are notknown until run time the object should be executed with themethod execute +However it does say There are no known critical or showstopper issues +I can see how this would potentially be useful for realtime +only two started other two are stuck +However it still might be a reasonable enhancement to VFS. +Thanks for the update Shiva +I am working on a patch for that as well but I would like feedback on this before going too far down that route +No unfortunately I havent found any info about how to do that so you are welcome to test if you can +I missed a couple files in my commit last night +Ive just committed this +The is still implemented on trunk +I agree it makes sense to support both distributed and local mode if we canIve verified that seqdirectory reads from HDFS regardless of the HADOOPCONFHOMEMAHOUTLOCAL setting which seems a little odd +The dictionary attached to supports sequencesgenerators on Firebird versions +Im fine with removing the class too +Remove unnecessary method +I should have looked at the code +I have only committed these for trunk so far will try and submit them for the older branches next. +I will look at the frameworks which you have highlights +I just committed this +Targeting this for Jackrabbit to avoid problems with the Java requirement. +Committed revision +Actualy Im not entirely convinced of this order of operations I think any registered custom property editors registered for a particular property should be used first followed by the then followed by the custom property editor by type and finally the default is going to be much harder to code however so Ill go with what Ive suggested above +The problems with Linux build fixed +Yes you are correct never before seen terms are heavier if you use the default weights and default probability +I think the filter solution would be reasonable if it solves the problem +Nick can you please remove the entries as well +Were you talking about something more routine than that? We should start pushing infra for a space on the Apache Confluence wiki +What is your container configuration? Have you changed the locking policy? Transaction should not be able to access A until the Transaction commits +Hi Paulexthe fix looks fine +I added a dialog that warn the user in case he tries to add a phpmyadmin to a scalable application +And in case of a timestamp version one nearly always wants to have the time of the database server +Can you provide a better name for the variable? We take care about the codequality too +Patch with review comments incorporated +I think I can fix the issue of by upgrading the versionBut for the I think we still need to fix it from CXF sideDo you mind to fill a JIRA in CXF for the xjc pluginI will try to use the to generate the code from which could avoid the CXF sideWillem +We should do a separate jcarder run for trunk though +If left for I need to explain that it changes the serialization format for Migrations in a compatible way which is not desirable +and Nicholas if Todds suggestion is amenable to you Ill go ahead and rework the patch to implement it as such + +fixed all unchecked warning introduced by the previous patch and removed the duplicated test +It seems clear to me that before we get to a node Hadoop cluster we will be using select to manage the incoming connections +Im not sure if you tested with a string but in that case the transcoder does not write to charsRead so it contains a garbage value leading to and endless loop +Perhaps this JIRA should transform into change the default low water mark to be of the high water mark or even higher +bulk close of all resolved issues. +If it is up to me you can close the issue +You still leave advisorySupporttrue for the brokerTo simulate an abortive broker shutdown for clients I use the test class from that easily allows the connection to be aborted between the client and broker +Will integrate in hours if there is no objection +So all issues not directly assigned to an AS release are being closed +Still reported speed while fetching update sites is not high its around s +Thanks Owen! +How about changing so that it can be configured with two sets of one for annotations and one for method return values? The latter can be blank by default and so retains current behaviour but you could wire it up with a message converter for handling return values if you so choose + +Assigning this to you Gary since you are updating some of the other adapters +The plugin doesnt use platform dependent features +These are easier to review and apply but hey this is better than mostI agree about the stack trace however that was for debuging purposes +Then the should request it from its storage object and archive that way +I have removed this change in the branch under r and I have added the fix to the branch under r +I havent added a unit test +The original filter will encode text into byte array and then decode again +If executable isnt found we use it without modification so you can use rake or pathtorake +I would actually the idea of a system property as this seems entirely the wrong granularity for this is a patch that implements the AMQP encoded Map messageThis also contains a test case and I have also tested against the C client and python presume we are going to use a message header or different mime type for the new header formatthat way the receiving client can determine whether it is an old style or new style map message without the application programmer having to know up front what type they should expect to receive +Thanks a lot Igor +Since there would be no handler here the request is already d by the time it hits the interceptor +This issue is not reproducible for now +But that can obviously be changed +This file contains the service that fails +Note that this bug and fix apply only to +If you eventually run w a single Collector no need to have in the middle +Thanks for the careful review! +No please focus on optimize your Cordova code +Jencks is EOL at this point. +fixed in r +Thanks for the review Bobby +As per BJs comment it will be part of +patch committed in r +Patch appliedThanks a lot +Since you have the rights to import the space you also have the rights to add pluginsIm going to avoid doing that wo explicit sign off from infra on the individual plugins since it involves running arbitrary binary code downloaded from a third party on hardware and i dont wnat to me the guy who fucks that up and lets in the trojanbq +Verified in JBDS operating system are you using? Also what version of jbt are you using now +Please test this and let me know about this +The new XML document screenshot looks very good +The commit was reverted and the error is now gone +The stuff going on during the and the seems to be the problem +Ok +Reassign to a release that you feel comfortable resolving this issue in or leave as is +However in Solr you can copyField a source to a destination field and specify how many chars are to be copied over +I am unclear here what the correct result would actually beThe spec states for the value attribute that the method should be used to convert the string to a character +Attached you find a patch for review +the error i caused during the second deny call due to the fact that its not detected that the deny entry for the principal of testUser has moved from position to + +finalName path +Fixed in and closing all issues that have been in resolved state for months or more without any feedback or update +Thanks Raghu +Heavy reflection or test ivy dependency on beanshell or some other simple scripting language +Patches for bug seems to be incomplete +Oh are you generating stubs +Thanks a lot for the detailed information + +patch is applied +We can just ship with fix for +Override the invlidate api such that it removes the session resolved issues. +deploy directory +Andrew hope you are doing you tried creating a Beans Config Set to logically group the two configuration files +Reopening this was only committed to trunk slated for and trunk. +Avoiding the use of and always using fixes this issue +I agree this isnt intuitive not in sync with the other quickstarts +With the recent change to which ones are still open here? Can we split any remaining ones out into a separate issue and close this for? +This patch fixes the issue in the simplest but maybe not the most desirable way +Ashutosh Chauhan added a comment Dec Please provide more information on steps to reproduceI just configure the environment make the HIVE work +I can attach my simple project later +it seems commit for this stuff are landing right now +I find the excess locking a bit pedantic but if its necessary to get to so be it +Matt thanks for finding it +. +Fixing the test for this case seems to make more sense here +Anytime where we have to document two different ways to configure the same thing it hurts our users +No error messages are observed with the latest build while creating the VM +I have experienced the same issue with both the Single and Multi Bundle distribution and want to make sure it gets addressed in the Multi Bundle +At revision and I made some further small improvements +duplicate of fixed in trunk +Yes it is worth to document thisCould some of you create a patch to update this libcppREADMEWINDOWSThanksroger +So I am closing this bug as a duplicate of Alberto +Did you intend to break after the first valueExpecting to get the last version tag id +Some of this could be mitigated by having the serialized form embed the initial configuration file used by and marking the fields transient but if you do that why not just go the whole route and have it build a newAnd what would be the point of storing an XML serialized format of the Configuration when the configuration data itself is an XML document? That is just seems silly +Marking as closed. +Depending on the version of the JDK the returns the appropriate class instance +Please test +Committed revision +appplied thanks +I havent been able to look at these docs for a few months +Please be aware that this patch requires the fix for looks like issue was caused by some crap laying around in my ivy cache after working on Jetty patches +Do what you want +Ive attached a copy of the stdout and stderr from the build run +it would be part of the metadata manager interface in Hedwig +r +Hi HenrikIve tried and tested your patch and it works fineIll apply this in trunkThanks +Lets validate it as usual and proceed with commit +It simply worked +Alright would you mind rebasing once youve committed? And if you could include the update of the specification doc that would be great Id rather keep the spec up to date at all time or things will fall through the cracks +Typically the comments are the place for users to voice opinions and concerns +yes is the latest patch. +This allows subclasses to set the fieldData to their own now extends Field overridding the setValue methods as they are not supported also now extends Field is now no longer used anywhere +Yes it is OK +Thanks for the explanation +bulk close of resolved issues for release. +Applied the fix in Rev +Help in resolution is appreciated. +Then please change the fix version for the JIRA to on the items youve done refactoring for +Thanks Kristian +will commit once tests pass +The essence of my argument is the same we have to load all the classes and that breaks stuff for a lot of people. +This was introduced in WSSJ onwards +Sure were on it +Admin please delete older file +Can you provide us with a sample pdf +The first should cover WS integration whereas the second will address documentation additions and some new features +OK so the particular problem that you are seeing right now is the same as +Committing +Verified in Developer StudioVersion Build id Build date +oops +If this is accepted wed probably want to make similar changes to the other libraries +guest is the default used elsewhere +verified +Unassigning myself from this issue +But if OBR was not installed then there probably are not any obr +Sweet +Add references to docs to prevent dups from being collapsedBulk docs requests may have duplicates or multiple docs with the same idThese were being collapsed in a dictionary as messages are passed frommergerevtrees in couchdbupdater to collectresults in couchdbAttaching a reference allows each to be processed correctly +Can you verify this +My commit to broke the +Using length is not reliable there are a few corner cases where data can be silently lost +Please try this +This issue can be closed now +I updated the patch by the latest repository + thanks +Brian I assume this is just the BPEL plugin provided by our forkIf yes we just need to assign this to nick for exclusion in JBDS build +reopening to fix completely +Applied patch with minor changes in r throw an if the supplied array is null or empty minor javadoc changesThanks for the report and patch +But what the componentIndex give us is which of the composite component is the cql column name +Please look at results +Fixed with upgrade. +The hint passed by the file node converter is used to detect the start level directory of a bundle +This will then cause the async delivery process to loop consuming a large amount of CPU +Corrected path in procedural step as prescribed +Bulk closing stale resolved issues +Any plans on this to be fixed? It prevents us from moving from Axis that has some critical drawbacks +It worked fine on linux builds because the link libraries can refer to each other and be resolved after the libraries are loadedGordon backed the changes out in r +Theres little we can do about this without a concrete test case +Required Geronimo to be merged attached that makes changes to gbean properties instead of the system properties uses dojo +binaries are distributed. +Fixed in renabled inlining constructor in constructor inlining again until NPE in SPEC is a separate issue for missing constructor call node problem +A trivial patch without running testcase i just verified with real test env +Updated patch for review this time the new classes and extend from and in addition I have introduced a new class fro creation of instances +Username cgu +Hi Dennisyes +ha utilizes this logical name concept to reference the rms +patch looks good +The compiler should take care of implementing the jobs associated with sorting +No feedback on my questions +I think that we will leak unused values every time the caches are invalidated and thrown away +The spec does not specifically disallow it but it really is +Hi StephanCurrently the documentation for both containers read are added to the classpath using modules and tested with JDBC drivers onlyFor any use case other than JDBC drivers your contributions are most welcomeThank revision . +If wherePath is false compiler works as +verified by Svetlana +looks good +batch transition to closed remaining resolved bugs +IMHO its a good reason to change because it introduces very subtle bugs which are tremendously difficult to track down +Attached example for earlier version not to be committed +Yes Ive committed the code +Patch looks good thanks for sharing +Lets just wait until goes out then do it straight we ready now +Pushed fixes out to master +So here comes a very simple testcase which showing the problem that Hibernate ignores custom column aliases + + This looks fine to me +This issue makes undeploy target completely unusable with running container in and all the examples +Since this is a private field i think this implementation will only work on Sun JVM and might fail on other JVMs +And forgot to add back in My bad +Verified in beta +Im going to close this for now I think the mechanism in has less things going on and may be easier than learning yet another api that wasnt necessarily designed for this purpose + committed to trunk and branch. +Hi StevoI cant manage to reproduce this issue with the provided project +Perhaps we can use part of to capture FATAL exceptions and somehow expose them regardless of the coreThen the UI behavior could essentially show the logging when solradminxxxx does not exist +Thats fine with me +removing the changes to which is part of to +I making an optional dependency on +Has anyone gone through the JIRA issues for the community version of HibernateIf the JBPAPP issues for DB were created just from unit test failures then there may be more issues in the community version that are not unit tested would be a good idea to check if there is anything important identified in the community version and document it as known for this CP then I can plan to fix it for the next CP +We should try running the same source document through the Xerces saxprintdomprint tests +Many thanks Jacques that was fast +This time with correct license +Maybe Id need to run to see the problem +Works in can you verify against the ER build +Whats the release timeframe +Committed the Management API to Rev +Check out the video I linked to in the previous comment and youll see that theres a load of mathematical theory behind the whole thing so its not trivial to do +Helen It looks like you missed something when you created the patch +This looks fixed now thanks long as we can see its an issue somehow related to the database persistance +I apologize for not being thorough before entering this issue +fixed the javadoc errors in the patch +Thank you +INFO Running for rows took ms +also looking for a solution to the log out issue +Please can you add a component for this +Code is different in trunkFine +Patch committed +The same is true for date fields +ive never run themi only rely on my gem tests right passes unit tests and system tests or I wouldnt have dI cant necessarily vouch for the tests +what exactly does that last request break +Dan got a test case for this +Some application would read the text file or store its text in memory at startup and apply the correct locale and send out an email +Could the code be organized so that ICUJ is only loaded when such normalization is needed +This method was added in december after the alpharelease so we should be free to change it +It is based on an excellent GPL Javascript mode by kland +the default realm +At branch JBESBCP revision +AmandaThere is no specific request to update to +Thanks +Thanks for reminding me +This looks like localization not working at wall could you give more details like a fragment of your jsp etc +This bug fix revealed some test failures in the script +Even a hint at where to look in Wickets code to implement it myself would be very helpful +to Santosh Malivya and Jyotsna Rathore for your valuable patchSpecial thanks to Pranay Pandey for the improvement in the work done by Santosh and JyotsnaPS Jacopo I did extensive testing of this work and it looked good to meCan you please see if its accurate as per as requirementI am thinking to create new file and moving the code from into itPlease let me know what do you think about itChanges are in rev Ashish Vijaywargiya +Given that this is an older release with a workaround we will not be resolving it inWe will however make sure that the RF wiki pages for AS integration have this info. +Thanks +Brian thanks for the detail +True it isnt all that useful of an interfacePerhaps it should be an abstract base class instead of an interface so that the API can more easily evolve without breaking user code? If external implementations are not currently supported this might even be a change +Thanks for the bug report +I have included a patch to extend +trunk +Thank you very much for the patch +As quicksort gets insanely slow when these type of data gets sorted this also explains Otis slowdown +So good so farBut we have to handle it in a different way next time +Deploy the WAR to and and you will see the error +Please vote +Ignore my comment on the null return value apparently the label is submitted when an html has no value attribute +Is it still causing you a problem with? It works for me here +Still have to account for shared caching which is being discussed on dev for release +The yellow hint box shows wrong expectations +on patch +The upgrade appeared to fix my problem with +This has been fixed a while ago what version of Grails are you using. +Moved unit tests to srctestjava according to maven conventions Fixed pom to run tests with more memory I got an without thatSo I was able to run the tests both from Eclipse and the command linePlease close the issue when youre satisfied that everything works +When I add another days worth of data the same search returns results +Ive removed that code earlier today to figure out what dependencies etc there are +I agree with Tims comments on not supporting this use case +I have a number of fields in the fl parameter of my query and it works fine until I add the field name recommended +All I did was merge the patch with the current version of the affected file +Corrected wrong link to mailing would you like to proceed with Groovy configuration support? Do you want to use what I have developed over in the Rio project? If so I will be glad to donate it +rslave dispatch now occurs on process when master has chosen the subscriber +Which version of Wicket are you using +fixes this resolving dup +Please try the updated and reopen if it is still an its not an issue for me because Im not even remotely interested in using it any more Im sure that if you cant reproduce it it must have been fixed +Correction The fields names in last comment should read astasks taskstasks taskstasks am assuming the moving window mechanism would be flexible enough to add new bucket sizes as requiredRegarding having the computation on the tasktracker and reporting the status via status one problem is that if we want to change the bucket size it would involve a change in the status objectAlso one requirement for this is to store this information on the +Committed revision +Created to this facility that is nicer than what we have in trunk. +I love all the minuses +Leaving as open +As a result the condition on line is never evaluated totrue +Ive solved my problemNow Im developing an application that uses esper and Im creating a set of portable output adapters for different kind of outputs +Done +I added a first UIMA pear runtime implementation that use the old +All unit tests were passed +Assigning as a task +You could still use relatives urls with this format but against db they arent contained in the design though but I dont think its a problem since the patch dont break anything +Done +Thus to me it would feel more natural if it was the data source for the LOCALREVISIONS table instead of all the other access thats configured with an extra new parameterAlso it seems that we should have some explicit flag to enforce the requirement of such deploymentsThe attached patch implements these suggestions by introducing a local configuration option and an explicit readOnly flag to accompany it +Ive commited this fix to +Fixed the line length issue +I would appreciate it if the reporter could close this issue if they agree with that assessment +The output of was fixing the testoperators the test needs to change missed that since it does not show up +Thanks Giri +Ive prepared a fix for this issue but I do not know how to write regression test here it depends on in javac classpathDoes anyone knows how to modify javacs classpath for regression tests +It shows the Attachment with a size of no and the mail message is actually patch to trunk +doesnt ask for the files length before closing it +Yep +Reschedule for issues to the next CR +Assign back to Stan I cant fix this in Seam +What I said above was that using the annotation on the object triggers this cant use because of whatevers causing and there are still several outstanding issues that we need addressed for the library to be fully usable +Committed revision merging to for +Sorry there was some issue with the previous attachment +Im currently puzzled why the seemingly useless wrapper makes such a big difference +The solution would be removing the each function from or enclose it to avoid conflict with other libs +Else its hard to track what issues actually have a patch attached to themThanks patch for the new feature. +push out to +This issue has been resolved for over a year with no further movement +Where do we want to put this stuff +updating to be a new module in maven build +The problem is not the overwriting of the platformsioswww +Create JSF Kick Start project +Needs more testing +i am facing the same issue currently +This patch contains the fix for the issue as well as the tests for itThe problem was that the object inspector was initialized and added to the op before the check was done for which table this operation belongs toFor the object with virtual columns fixing up the inspector for virtual columns happens after this check +Remove time normalization for hadoop widgets because time normalization is done in demux now +Curios if there is any interest in trying this one outI am curious but wont have time this weekGetting Thrift into a real maven repository would be a huge boon +I just committed this to both trunk and +Any validation markers fail to resolve to items in the editor +I then overwrite the original files with these raw files and I could do a rebuild successfully +The log and tracer now have a default maxChars at +It seems that there are different slfj versions in blur and CDH +Where were you running this? Sounds like wasnt on the classpath +looks goodOne comment please use in configure to build the settings string +Auxillary suggestion eventually it would be handy to have the host alias stored alongside the rest of the method introspection data that would allow us to simplify some code in and might eventually enable us to improve certain error messages by spelling the method name as documented in the host language binding docs +But lets continue this discussion in and add possible fix there +Hans thanks for the catch +what is the further procedure in your opinion +It works fine until the next reload of you +Thanks for tackling should work now +This is then of course a bug +Ive installed the that is required to allow the property re able to change the script contents and do whatever logging changes you me know if this solves your Ben for getting this installed that quickly +I think this is fairly trivial renaming things like test to testPrefs +Sounds fine +But Jettys als seems to acceptanyway Long is ok for me +Verified on. +This compiles perfectly well for me with SVN HEAD +I think it is correct and can be checked in if anyone reviews it +Attaching a couple jpgs to demonstrate more clearlyThis looks like a simple fix if all thats needed is to swap out the definition to use that linked widget +There was a plugin we wrote for confluence that was a bit flakey +closed all issues which got resolved in M +defer remaining issues to +If shared edits is not available at the time of state transition active will fail within secs +The associated test case as appears to work with r +Could this be due to the switch to Hadoops local mode +Patch adds flex path and version detection to +Sorry for not noticing this before but it seems like the class package can be improved +just committed this +r +The two failing tests look unrelated so Im on the patch +Resolved +We are having exactly the same problem we have multiple builds that uses different versions of Maven +PubSub patch with the missing files +Hi JasonI did exposing and I can consume the service correctly +Bumping to +Update issue subject Id say to talk of is and is +See problems view to install it via reopen the issue if it doesnt work even with KB builder +Makes this test case run a lot faster +Hello Its my fault +is in the root directory and all existing files have been formatted to follow it +Attachment has been added with description sample to reproduce the P open bugs to P pending further provide minimal FO file that demonstrates problem also provide PDF output file showing transition resolvedlater to closedlater if you have new information or a patch that addresses the issue you may reopen +One minor comments about is that could some code be moved to some utils or some base class to share with andorCan you confirm that it is safe to remove nullSequence from and +I was reluctant to move or rename where the data ends up so as to break less stuff this is a smaller change +NamitSiying Whats the followup JIRA ticket ID +Jostling around the release goals and numbers +Now that the changes plugin requires Java can the patch for Bugzilla be applied +When do you think you can get this done? week weeks or weeks +because of the consensus right now is to rollback the change for Groovy reverted the change really m going to this issue since the change was rolled back but the underlying performance problem is still there and we need to find a solution for it since it is a huge performance bottleneck in highly concurrent applications that use EMCI understand there is no time to look at this for but for one of the point releases we should try and find a solution +oops +With missing dependencies from Hadoop us having to support two versions of Hadoop and Mavens limitations when it comes to profiles theres just no clean solution to thisThe patch adds dependencies for to the archetype +Reopening to change the fix for closing all issues that have been in resolved state for more than one month without further comments +which is why I used the url encoder it does the job and its already there +Move reasonershermit and reasonersowllink to contrib as these have license issues and are not actively enhancerstores to contrib as this is not maintained anymore and the is used for this feature +Also if someone later decides that size should be a long long assignment is not atomic on some JVM platforms +First patch for the new style to the Lookup forms +Files being touched aremodified srccoreorgapachehadoophttpmodified srccoreorgapachehadoopsecurityauthenticationclientmodified srchdfsorgapachehadoophdfsserverdatanodemodified srcmapredorgapachehadoopmapredAnd BTW how can I get the unit test result? Copy and paste from terminal output or there is a different way? This is the command I used to run the unit test +The problem can bypassed by either increasing the jvm max memory or running the subsuites as in ant Should anyway understand why we need the large jvm memory running directly +Thank you for this hint +Looks it does +hiAlan M +Duplicate of JacquesThe task was not duplicate it is fine what you did lesser task are more manageableBut we need to add all applications components in the +This is not related to the protocol but it just the implementation of the leader +It works as is for now even with the ancient antipattern +Added to +Do you plan to release a new version of the plugin that will contain this fix +Did you mean to expose it in the base class? +Hi OlegI made a new build of with your changes tested it out and it looks fine +Targeting to the next release +The geotools referencing system has no way to add stuff in the dbs its a query only Desriusseaux the referencing system developer is working on a chaining authority that allows to make a compound of different authorities some db based some file based +in the issue description is to be addressed in documentation. +the problem persists with too +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +committed patch based on this one +I changed the embedded code to not add the warning and saw the same protocol error even though the warning was not sent +This maybe a temporary new home as they could be moving soon to prior to releasing them along with DNA +Your patch looked like it was adding lines whose length was characters +how much time does it take to save k added nodes with without this patch +Pathch available tests ok with classes superClass and commit above reinstating Analysis menu item should have been associated with +Hey Barrydo you guys have release notes I can point to from my JBDS release notes that document these changes or should I incorporate those listed above into mine +Sounds good to meBtw the latest two patches attached dont remove the wrong patch attached +Maybe you can rip out the random number generator in your machine and donate it to Apache! thanks for running tests that is of great value +The file named contains two thread dumps taken about minute apart +Hello Ivan +Some of the original assertions are checking for specific error messages on an exception and that doesnt make sense Ive commented out those assertions with a FIXME flag +I havent seen it in a while and searches seem pretty speedy nowFeel free to close this bug anyway. +Both store have a similar design +I had to exclude x x x and x as these are all possible valid values withI tried the repro with the patch with various version combinations of and trunk with the patchI am running tests now +Right I dont think we have to worry about things much over top And while I dont want to take the time to do top for kicks I tried top over a couple runsIt actually does show a win with the new method once you get up that high Its somethin anyway +Attached a patch that strips out query parameters for with the file protocol prior to attempting to load configuration from that URL +One thought is that we could put loadings and averages into an hbase table +Thanks guys to take this into account +This issue is potentially cascading and cannot reasonably be resolved prior to +what do you rename it to +If there is a problem with this change please reopen this issue +Toby this is done +Patch updated +Davidshould we fix this forthanksI no longer know what the problem is here +Some minor aspects are missing thought This new property should be added to The default value of this property should be set in s constructor using the setPropertyThis way the documentation generator will autogenerate the documentation for this new propertyThank Savasthanks for the quick review +Im not sure it requires API changes but I did fiddle with that stuff once nad it ends up breaking dozens of wtp tests +the localized versions? How to compile with jsse with the IBM JDK is not taken care of not sure how that is done +I can see Libraries now but design is always empty in projects I reanalyzed +Well there is a set on the but Ive just noticed that there is no equivalent on the +Looks good to me thanks. +I can no longer repro the issue on with a few minutes of testing +If its not supported it shouldnt be there +It almost looks as if your is incomplete chance you can make a that reproduces the problem +Ive just committed this to trunkThanks a lot for the contribution Ivan +From talking with Eric since he was the last person to really mess with the there is a good chance that its roots are spreading more than we imagine +I had added this test case for another patch +Patch to add to patch that fixes the miniCluster flag on PE broken since we became a just committed a version of this patch that incorporated stacks suggestions and patch. +Timezone is a rule how a date is converted to a string visa server definition is just a offset in milli seconds to the local time. +Im fairly certain that it is class path related as well since I am loading the newer Logging JAR in my WAR attempting to override whats provided by EAPAre you recommending I updatereplace the logging in the server lib folder? I had assumed indicated that was the fix for it but I suspect it was never tested according to the comments +Yup thats correct +An interrupt in java means please stop what you are doing +Sure +here is a patch for it +To add a little more I think the requirement of committer s from people who didnt work on the branch will make it really hard to ever merge branches +This is becoming the naming strategy there any harm in quoting a tablecolumn name even if not necessary? Otherwise I would suggest a Better safe than sorry strategy +Applied and fixed. +Nice work HectorI have a few commentsFirst doesnt your train method destroy the input vector? That seems like bad manners +I was pretty aggressive putting in apache license headers everywhere I could just to cover us +So we need another solution +I agree with driftx that we should add the WARN statement +need to backport this to ported to to to to change the fix version to include +Mark good question +Actually heres a patch with a more complete test that actually verifies that this works with the +The patch in here is careful about making sure the bulk assign is favored +I made this so i can work on it without breaking the working unsupportedwfs module +Fixed with revision facfdcfaeecaabcbReverted the change given in the Isuru for the correction. +You explain this with remarkable clarity Rick and I dont think we ever document clearly that using bootPassword is incompatible with using encryptionKey +Is this want you mean? first populate temp then put it in bulk to ABSC using addAll +Attached the patch which fixes the issue +to TRUNK +The attachment should fix this problem with the broken link +JDYes the bug is same as +new failures from the test change? I am not closing the but I dont think that can cause the new failures +Here are the few things that were left behind +in general you wont get much compression for chinese +Closing all resolvedfixed issues of already released versions of Roller. +Thanks for the patch N +In both cases the suggestionbox doesnt hide but it is possible to move focus to other fields +Gridmix already work with MR +This makes it more obvious what headers are used by a certain piece of code and also prevents problems like you are seeingMost logical reason for not noticing this before is indirect inclusion via other headersFor this case it is a simple and quick fix Ill try to commit it somewhere today or tomorrow +Lets stop arguing when we are agreeing +Ah sebastian I think you needed to check Grant license to ASF for inclusion in ASF works when you attach your patch +Do not hesitate to request any information +Ive fixed this style for showcase but I still believe that this issue could be fixed for skin in +Diogo I believe you meant to submit this patch for review correct +Mark has already committed this and it worked for me when I tried it using the m plugin +The patch is in the svn repository +Remove copyrights line and add ASF to some filesWe will just add changes directly for missing ASF license header +Much better Ill look at the patch +The reason is that one test had to be run on Java +Well doing unpleasant things on invalid input isnt really a critical bug +Fixed with a cute integration test in you may want to give a spin to verify this works as intended itll update in an hour or fixed for my use case +For the backwards compatible manner I fear that this testing decreases performance for reading as each and every property has to be checked +Initially I wasnt very much sure about it +Making this a blocker since we have the and set to false by default +Some code that was copyrighted by sun was removed and replaced with another implementation +hi there thanks for your suggestion +Attach the quickstart with the workaround +OK Im convinced that I should turn off this warning in eclipse instead of putting all those s in here +Closing this task +VerifiedThank you Alexey. +have exactly the same problem in IE +I tried this multiple times can you verify or update this if you are still having the issue + +Fancypants +Final patch +I am not sure what the right thing to do is here +I realize that I could accomplish the same thing by invoking the undeploy and deploy goals but I think it would be a good addition +It seems that it is triggered when release number changes +I am building the plugins locally to make sure this fixes my problem +This is not a bug may open one in either Solr or Tika +I dont really care what Cloudera does as a business +Is there a separate issue for making Schemas completely immutable +cer +Also note that there is no default value for this property and leaving it undefined or blank is the equivalent of disabling it. +. +Looks like limit issue I was seeing has been addressed in the latest trunk +Updated patch seems ok please go ahead and commit it +This should be ok for all typical use cases but its worth noting +I can see how the bug is occuring very good analysis and thanks for finding it +This issue has been fixed and released as part of release +The database is being hit just fine as the project builds with no errors +Applied patch in revision . +as many times as this issue has been brought up there should be some work on itand it should be ver as well since there was a consensus about security +Unchecking patch available since it seems there are no patches pending commit +The use case it address is unusual but I believe its a good thing to install all the bundles using the URII committed the cumulative patch attached to svn ci m cumulative patch submitted by SahooSending fileinstallsrcmainjavaorgapachefelixfileinstallSending fileinstallsrcmainjavaorgapachefelixfileinstallTransmitting file dataCommitted revision Please close the issue if satisfied. +Yes makes sense +Here are my current thoughts then +Yes sure +Andrew could you please commit this? +The concatenated name is used by the to insure only a single manager will exist for the queue +The branch solves the symptom of the error being unrecoverable but does not address the root cause of the error which is that processes are being treated as recursed rewrites +Closing resolved issues. +It must be deployed in extracted form +thanks that definitely helps +Idea shows following messages in the back groundField not copied JAVAINDENTOPTIONSField not copied JSPINDENTOPTIONSField not copied XMLINDENTOPTIONSField not copied OTHERINDENTOPTIONSField not copied FIELDTYPETONAMEField not copied STATICFIELDTYPETONAMEField not copied PARAMETERTYPETONAMEField not copied LOCALVARIABLETYPETONAMEField not copied PACKAGESTOUSEIMPORTONDEMANDField not copied IMPORTLAYOUTTABLEI am using v Once i set this up I am gonna look at your version and whats there in +Ill try to produce an isolated test case +So lets create new formats outside of Hadoops kernel rather than add new ones inGreat +This seems to be JettyQuoting from the servlet spec When Parameters Are AvailableThe following are the conditions that must be met before post form data willbe populated to the parameter set +I think it can satisfymy customer +Is it going to be released along with +committed +Saving lines of code isnt a good enough reason to do thisMaybe you need to refactor the base class if what you are trying to do does make sense +Modification of the file updatecenter which failed to IT +Who will pick this up? Marshall +Closing the issue. +Private means it can only be used within Hadoop +See revision of . +Reverted the patch +Now in we cant use listeners to take screenshost because of some troubles with Graphene static context +Linked which has some logs which might be +There are a lot of issues surrounding the complexity of building the index itself in the plugin that are outside the scope of this issue +Attach patch v +forgot to specify the component to jboss! Sorry +Since were still discussing +the previous attachment missed some previous patch reversed the order +With this patch a specific LDAP filter is created to only fetch a minimal amount of data +this issue still occur now that are being catched per the fix +There is no branch for stale resolved issues +META +Normally I only do improvements on the trunk and some people from progress merge the fixes back to the branches +Nevermind that solved a blowup on startup but caused problems elsewhere +closing this again as it is not the cause of the problem it just changes the error text +The compilation issues have been resolved in revision +examplesbinSeanShud I change the wiki instructions to use the script rather than the n steps that it has now +As a matter of fact the code as procuced by Castor is +Patch applied together with additional changes in rThanks for it Patrick +So it should be backported to or. +Should be sufficient for now. +Charles Please make sure the patch applies cleanly with patch p +I inlined the schema rather than reading it from a file +we already discussed that for +nobody to answer my request +Thanks for filing a bug because I forgot to do it +seam guide is checked +transaction success +Thanks sebb Ive updated the fileI see that where DOAP files are registered has already been updatedto reflect the new location +Unfinished first attempt +We went for the simplest possible path which was the error dialog +Maybe we can use aliases for column names in such queries on OracleAttaching possible fix for this +Only a subtask was committed but this triggered the integration messages since I put this JIRAs number on it +I updated the behavior on only so when the resource simply s on the remote side it should still be fine. +Changed to Long +Code reviewed by Hairong +So help is welcome for this issue +I have not seen an upstream JIRA for this change so we can use this one +checked in +This is not implemented yet +The merge functionality is not foolproof and may cause generated transforms to display incorrectly if added while the editor is opened +Functions belonging to the classs namespace +What exactly doesnt work for you in Leigh? Please submit a full stacktrace etc +Possible duplicate of resolved with looking closer this issue is not a duplicate of. +I agree it is a good idea tofall back on the outer transaction in case of a timeout +Updated for both and trunk +Thanks Erik and Stack. +Passed +Fix sets first result query parameter to +Hi Aidan can you review this change please thanks +I havent yet checked Docbook +Could you explain in more deep where and how insert the transferToFrom operation please +verified +Patch is working tested on my notebook with both native and cxf stack and EAP +Removed any syntax check just made sure the field is a required one +The user might wonder why the project build when they set always build to falseFYI Continuum first checks the BUILD FRESH before it checks the ALWAYS BUILD +I will add a fix in this patch +Where are we using protobufs in? I see it in the pom but I never had to run protoc to build +I already see you merged into. +Will commit after test +. +Includes job and shuffle metrics +Mimej doesnt currently provide a means to determine the actually start of a body part in the underlying stream +We should add per cache so that we can avoid this issue too +Feels like serdeProperties are not available to serde at initialization time for partitioned tables +Remember the feature freeze date for is and the last day for merge request with hour feedback peiod is +New archetype stuff will set all versions + +This patch adds more BlobClob overloads for existing signatures and adds tests for functions which return Blob and Clob +Assign unscheduled issues with Blocker or Critical priority to +If anybody has any issues with my changes let me know +I wonder if this issue can be closed now that the repro passes without error according to understanding is that performance of memory usage was improved issues including and other issues worked at that timeI think this issue should be closed +Tom this looks ugly +Do two webapps have the sameclassloaderI set this as duplicate and add a message to the other bug +Performance jobs could have a performance impact if the number of replicas are reduced from to +Hmm how do I edit this to correct the wiki formatting mistakes +Used hbase trunk to build this patch +Improved get Method for making it more generalize +Fairly straightforward piece of code based on description by Knuth +I believe this is already fixed in trunk +I set systemenv to point to it inDoes this make sense +reopening so we can add the testcase from BJ that proves this has been testcase showing that this failure no longer exists to and trunkThanks BJ for the testcase patch. +The release is unlikely to work against the final release +This would be a REALLY helpful feature. +Thanks for extending the patch to get and the opt methods +Things get dicey when multiple output formats store funcs go modifying a single shared but the solution in was clearly flawed +PS I am not sure if LGPL is an the poms at contain repositories it from and removed the repositories sections +Yes Sarissa is included but the problem seems to be with another script +This issue is common to other Geronimo releases +Great work! Two comments doesnt seem to be used The tabs are messing up the formatting a little +Perhaps the upgrade to jackrabbit has resolved the problem + +HDFS Mirroring Add Data Set Popup page should be implemented +implements Mamtas comments +Suggest we reject this one +Anyway I am happy to take a look +Lets discuss this further +ill add an assume here for OS XNote another option is to conditionally in an afterclass but these sysprops are initialized by the JDK so that seems a little scary +Unfortunately I had no chance to investigate more and I cannot put more details but I hope it might be helpful clue. +I will take it up from here +I committed this +include some bug fix also change undesired behavior introduced by +As well as adding some documentation +New patch with new test case +fixed in while updating xml files and creating validation am closing the issue +A instance is then used to parse the value +I just took a look at this and it seems to already be fixed in both and TRUNK +Can you please submit a unit test +Ive applied your patch thanks a lot for your valuable contribution +fieldName street notNull sizeMin sizeMax add field string class +Basically the code creates a to handle the bloomfilter logic +I did need to change a bit how import statements are handled by the inferencing visitor so this may cause some other tests to fail + +I removed few redundant lines +I hope the script could be easily translated +Why dont you simply cast the instance to given that it is one +not yet committed so is not something which should be set +This patch actually compiles. +Changing the title to match what we decided to fix here +I have put together a simple web service tester that handles and +This is the updated patch of mail notification +Issue closed with release. +This is exposed when packaging against based jars vs +An Exception should be thrown +fixed +Would this be a better way to deal with balancingfailoverThis has made me curious of exactly what the intended use of is now +and do you mean because I dont see this class of +I will add an IT and close the issue +This is an interesting addition toI think it would be better to put the clustering package under stat +I wonder if this is a good candidate for a interface that extends to add a new method registry +I encountered the very same problem today which took me about and half hours trying to figure out whats going on +Yes you are right +Im abandoning the sub tasks since there are changes in a given class across the subtasks and finding it hard to maintain separate file sets +But if one plugin fails in isnt the next used? I understand that in the the would be used as the last resort fallbackAlso Im not sure where the comes from if I use the patch above to prevent generating an exception but return empty +unsignededsmodeshapeexamplespubunpubusingrest jcoleman soa Dec +Thanks for the contribution +Fixing javadoc errors and test issues +yes works. +Good catchCan you code up your example as a unit testA nice simple thing like this is a great way to get started on contributing +Added documentation for the clustering feature +Good find marking for +Try this one +In response to the comment bear in mind these are performance tests to stress the systems were intentionally closing the connections +bulk defer of featurestaskswishes from to +Tests +Also note that even if would be OFF it would be off in both tests the only difference here is the caching +The used AM resource cannot exceed the per queue cap and the real situation would be that there is be some wasted resource of each queue because it is not enough to accommodate one more AM +and Mark Thomas described in Tomcats Bugzilla he added unit tests for this and they passPlease create a demo app that demonstrates the problem and attach it to their BugzillaAdditionally the you pasted in this ticket are generated by Wicket not by Wicket +Im running into a similar problem with +Isnt that same as UDFUDAFUDTF +I wonder if we just need to write some generic tearDown that on error or failure copies current state think most of those requirements are fulfilled by the ant junit task except for the last one which as you say is a test issue not a test runner issueI dont see any memory issues running I run it as a continuous integration test against Derby + has just been released +Resolving earlier conflicts with and and submitting the patch againthanx revision +I created for this. +Will commit merge to today +current schemaLocation IS valid with thats exactly how redirect worksAnd for smx we always ship local xsd files and customer shouldnt rely on the online xsd accessFor your xml editor I think you can import local xsd files and validate the xml +Are you fixing this test or some of the excluded tests? +I have observed this issue also with DB in the implementation that ships with the IBM EJB Feature Pak +Attaching a patch to handle different types of user storesTypes supported so far LDAP amila for the contributionRegards the resolved issue. +Thanks didnt notice that +Since M seems that this problem no longer occursStill cannot change tab if switchTypeajax but if switchTypeserver then I can change tab and see content of this tab +i agreei think there is no wsdl into jms sample component +That way we can support both the more generic capabilities I added and you can still meet your goal + +Isnt this implemented already? +Sorry +Naievely it seems since I was just exposing anunderlying API and because the rationale behind all this is very well documentedin javadoc I didnt realise Id have to argue worthiness of thefunctionality from first pinciplesI now find myself having taken more time to explain localization and than it took me write the code and javadoc diff and add the patch to bugzillawith an example of use +I can help being a Maven user myself +Yep cleaned up my caches and it worked! Thanks for trying Mahadev and Hitesh. +Does any one have an official description of the possible output of subversion +Ok Ill see if can make some too since my is put on hold for the moment + +If that is just as borked as the attachment stuff then Id agree that we should fix it +I think wed have to add another filter to the filterset in the target of the file as well +So Ill start to build a prototype for thisBTW Whats the recommended way to obtain the from within a CDI extension +The screenshot shows the situation pretty well +I tried the patch and I can confirm that the issue is fixed. +Defer all unresolved issues from to +Includes Marshalls latest updates +Duplicate of. +Ok Ill fix that and check it inI was focusing on the dummy which I still dont understand how that is working +That webapp actually uses a inbuilt Java database HSQL +It is natural to create a test database and do all the tests there +Kelvin patch applied at revision thanks a lot for this enhancement please verify +As an immediate fix lowercased requestURI before comparison +Thanks Alexis for the donation +Making critical since this race can leave random regions hung in splitting state +Patch V I am trying to not rely on the Delete object or any +Another potential workaround is to make the childs default styles differ slightly from the main apps default styles +may be use weak refs instead +But I didnt know that I have to run Directory Studio with administrator privileges! Now everything works like a charme And yes I could obtain the TGT from KDC directlyThank you very much for helpy! +Updated the unit test +its not important and we are trying to make a search engine not a file managertomorrow ill revert my commit and go back to deleting all files as I was totally happy with the previous situation just dont put stuff in the lucene index directory +bq +Your testcase includes only once in the output is this intentional? If so how do you suggest Wicket filters these duplicate references +You are right it would not make sense fixing it at both placesThanks Cheolsoo +Defer all unresolved issues from to +The AM can continue to run those containers until the RM actually preempts them if needed +Patch successfully committed +Right your observations indicate that this is indeed working as expected +Imnot sure why it doesnt match up but that may be a problemw the driver? Im using Db +Both patches have been applied +please take a look at the attachment +a column to be unique requires data changes in the general case and is not dependent on the jdbc metadata api and storing a previous version would also not helpIn any case closing since no real issue reported and the idea of storing a previous schema does not provide anything that cannot be done to day. +For instance AFAICT throws an exception explaining the xml problem if it finds a or file that it cant interpretWhen looking for examples that fail keep in mind that does a lot of namespace substitution and many builders have code to upgrade documents valid for earlier versions of geronimo plan schemas to the latest versions +The mismatches in are all because the new code produces better output +Do you know of a way to reproduce this or does it seem to be purely random +Removed the change to and added comments to to explain the intended purpose of the session identifier +I guess this should have hit the mailing list before JIRA but what do people think +Awesome news +see also m marking this as invalid because I dont see where the illegal stateexception is thrown +Ok so I guess this can be removed now then +obviously more work can go into determining things for next release +My only goal is to remove a redundancy from the Chunk API +Can you try this out with and see if you are still experiencing the problem? +As I understand it we are opting for full incubation which means we dont go through the IP Clearance route +patch to hudson +Bulk update component to Hibernate +resources are not filtering out some of the log events in your dev environment +Im not sure wherehow to best fix this yet +So we have a configurable that will make the callsYou will get a bad cast exception if you use the on the. +new revision also removes the places where we set the lease timeouts in the NN to values +Heres the first patchIt is on top of branchThis doesnt cover coprocessors which I dont think need to be covered some methods in accesscontrol which seem to be internal rest and thrifts do these need to be compat? I assume not filters the filters seem to be ok alreadyI am running tests small ones and some medium ones already passed +Also verified the changed HTML files are fine +committed +Resolving the symlink has issues +new patch attached applied to the branch JBT and JBDS and jboss tools aggregated update site should be build before building JBDS update site and on and it works as expectedresolved +I just committed this +Same behaviour for JSF action with JSF projects parts of Seam project results same i remove seam projects from Server they stop falling +Reopened The above fix doesnt work when the bridge itself adds private render parameters in this case the count of private params is and we choose not to combine the maps +there is a new param ldapuselowercase that you can add in the config fileIf the param is true the username is converted to lowercase before validating the usernameThis has no effect on how the password is verified +I would have thought the code to do this was somewhere in Spring in some form or another given that handles requestresponse +you can borrow object using key and without keyThanks a lot Simone +Invalid as above. +Any progress or review to includeapprove this to include it in the official distribution + +Also the groovy library dependency within the project is is it? and not still? We should probably be forcing a recompile on version switching +In that case we should probably remove that toolbar and dispatch its icons elsewhere link to JBT site could be put in the News toolbar favorite to marketplace is it really necessary? Id remove it link to JBT Central preferences should be deleted as theres already a similar link at the bottom of the page +For one that returns rather than null when the value is null use +CyrusI created a separate JIRA since is about automatically inferring partitions during query time and not updating metadata at all +stop is renamed to close and does not throw an exception +this is a Westport issue. +Patch v corrects a mistake in patch vThe current collection of files should be gathered after log files have been processed +I wont have any time to take this any time soon So if anyone has the itch jump +The smooks editor is no longer being developerd so moving to LATER as agreed with +There is no reason you should not be provide a resettable Stream and an Encoding and perform the same operations but reset the stream between tokenization and value storage stages +ToliI misread your mail as you also add a mvn execexec command line in it +push to +patch changes libhdfs only but the failed unit test was in core mapred package +Just go ahead with this patch +memory limits for queue solved the problem +Probably because the line should be broken before org +Is that dead code +That bug was fixed in the patch think thats reason enough to declare this resolved so long as we get the correct patch of wtp in the release +I think we need another approach because it seems like system properties and release properties are mixed +Something more different than timestampof or ts +The idea is to increase the interval at which a STABLE gossip is sent with increasing cluster size in order to keep the average STABLE gossip rate for the entire cluster about the we didnt increase the interval the gossip rate for the entire cluster would you have a scenario that requires immediate acked delivery the tag the message with does this is some places +Implemented in verifycallback does not work still +Notice I have attached a screenshot of the Fuse IDE on this is related to +On after first deployment it asks me to install is it normal? I guess that its not since on Tomcat you dont ask user to install +Do you mind attaching patch againThe previous QA run didnt go through +Uploaded new patch by addressing review comments from Amar +Fixed in revision +Wellive done some test +Default is seconds try making it deltacloudd i sbc V t it took seconds to load the page so it was really a timeout error +I do such a search before bothering anyone with my question +Srikanth already has a demo and I think he plans to make it available +No action takeninetbridge are months or even years out of date in some cases only updating the main DATE file and hreaderreadmes nothing else +What do you think the behaviour should be +Removed old test files Removed DatabaseSystem property setup for usedefaults +The fix has been applied to all branches. +ran update for users from this file output to bala for steps a and +Moving the Jira target version to M. + +It is really started to shape up nicely +Closing this record since is now released. +Will start the full regression test suite +Fixed +Server cannot be started +We canshould probably guard against this though +Thanks +It would be nice if this is fixed inoHowever I dont have time to look into it now +Updated patch with a fixThere was a bug that would delete directories that are not index segments +or at least not try and do it if it doesnt support the underlying configuration. +Use instead of This patch rearranges get a bit because the original version was unneccessary repetitive +Checkstyle can warn you on missing defaults and fall throughs as currently used +Ok +The generated states for the xsall tags child elements were startingwith +pending JenkinsThanks a lot Todd +Resolved in Branch Branch and HEAD +Apply the patch and create your own version of the plugin +has been marked as a duplicate of this bug +Forgot the URL to the fixed the issue +Can I please ask you to raise this as a problem over at the CASTOR project as it seems to be a genuine problem of code generation related to one specific version of Castor +Im looking forward to starting work on when this patch is commited +Test for reproduce +Once Glen has reviewed the change we may close the issue +OK +Most of the znodes are automigrated +Its not working jboss grails. +Thanks sanka for the update +closing the issue as duplicate of the newly created one in issuezilla +The other possibility is that he means the entirety of the independentdriving side of the join would be shipped to form the temporary table such that the complete results are read by Teiid instead of a secondary join being performed +Hi guysAt first one project is for one web service is a better practiceSecondly Brian is right +Morten You dont want this there anyway so good result But seriously it seems to me then the slowness is coming from the time the JDK takes to scan the classpath for? If we put some timings into the above code we can really see what phase is slow +If that is the case I think this JIRA should not try to change any retry policies at all + on the proposed approach +As Jim said on the mailing list storing the schema in ZK would solve instead of having another catalog table named ZK integration be targeted for or beyond? What say ye? I have some work pending for that would not be that profitable to do if ZK will go into +I was able to run the programand I can see in the that the logged statement text is just as I expected onlya token for the large objects stream is logged not the large object itself +the secondparameter in the constructor +But added a testcase that shows the problem +Fixed with a test case. +I solved this one +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +Theyre not exclusive +Theres a wide row test and a range slice test but not a combination of the two +Not a problem anymore +It would be better if we had a single knob whose meaning was run securely +Please note the generated distribution file is called +generates the code correctly for given wsdl +this patch fixes the issue by adding a thread that does the snapshots in syncrequestprocessor +Stefan I dont follow you completely +So closing this Jira. +Hi Alessio our build will produce a anyway so you can version it any way it is convenient for you + +Will do instead. +bq +For later uses when we have more advanced archives avro would surely be suitable +I believe that xmodel does not lock jars +If this doesnt solve your problems please ask on theusers list. +Looks good to me +The perl scripts will be left in place until after the release and when we can confirm the new mail outs are sufficient. +Thx for the fix +Even a patch was proposed +No more occurrence of this warning in log +The patch looks good to me +Will commit once the tests pass. +I attached a patch to add one test for plaintext +Moving to as it appears no one is currently working on this +It is the same symptom but unfortunately the fix for is not correct +Code complete and tested though we could do with some Spring based tests too +Patch applied! Thanks! +What can be done about this? Whats really needed is a notification from the servlet container that objects are about to be so that we can store a object that links the serialized service tokens back to registry +This may be a duplicate of you run mvn assemblyassembly in the parent pom you are asking to run a plugin outside the normal the the parent pom does not have an assembly descriptor then shouldnt this be an errorPerhaps mvn assemblyassembly should not force a reactor build +I believe another customer Robert Koberg has run into the same problem +We have both tools and extras dirs now would be a good time to set a standard for where we add new toolspwdoptlocalDEVtree L +The cache key used is the url after resolving variables +Triaged for +The server side doesnt parse out correct params which wrongly understand what the client want to do +If you figure it out I would love to know whats going on +This patch fixes the root context resolution under Servlet in Tomcat and fixes the lookup of the root context +META +Im closing this as final for now as its been released to wtp +This was resolved in upstream master and in late May can this be resolved here +I tried with mvn install and the build succedes so consider this as +That is fine with me +Tested locally working will be part of next out into production +Everything seems right + +I did run the job from Hudson and not from command line but I wouldnt think that would make a difference +html files will not there once ant docs is run +breaks the contract of Map The actual behaviors of the methods of are vague +Ive removed the check and all test cases run but it would be nice if someone looked deeper +patch applied closing stale resolved issues +If it indeed makes things considerably more complex to share the code as you fear then we probably should not +This patch contains all necessary changes to the file including fixed header from patch has been with the release of Sling JSP Taglib +I know a bit about OGNL but what i know leads me to see it as more similar to Velocity itself than to the Tools used within Velocity +I notice some changes are in there too not sure if that is intended or not just thought I would mention it +Heres a patch that fixes the import and static field issues I found in the and packages +with javap on some class filesSource and target compatibility has been changed to for the core module. +the current Hibernate GA has still this issueWill it be ever fixedCan some hibernate developer give feedback about this issueThank You +Ah much nicer thanks. +I concur a WONTFIX is a better outcome with +for the patch +Pushed Jeans initial branch to upstream of this Jean +Sound great AndrzejWould suggest that it not replace jenkins but can it be offered as an option? We cant just replace the jenkins hashing since the hash is used in filenames out on the filesystem wed have to run a migration to move from one hash type to the other +Sateesh and PranavAny luck on the GUI part of this patch? I recall it was mentioned to be a separate patchThanksilya +Thanks. +Just committed this +The error level in the post seems more appropriate since the exception does not thrown on +Eugeneunable to use means that is excludes everything in the resources folder be to exported into targetclasses +This demo application shows in Harmony has HTML support switched offRun it on Harmony and youll get the following output Harmony eeWARNING HTMLRTF is not supported yet +to renaming the current minimum to indicate slot size or multiplier +Layout has been changed in +Also that environment didnt use Maven like the one where I observe msI just dont understand why a more complex app would slow down classloading for the same exact jars! This is the scary side of Java too much auto magic going on. +The attached pom has an invalid licenses node +Running the existing tests nowThis patch makes the compare cell types rather than array types when checking to see if a wrapper type is compatible with a primitive type according to the ANSI rulesTouches the following fileM javaengineorgapachederbyiapiservicesloader +happens if you try to send an event in a to a JSF related Context +Patch contains nature specific stuff only +Ported to branch at subversion and to branch at subversion +issues after release +ok i am trying now building ql from scratch and running the joini will comment here the resultsThanks +pushing off remaining issues in order to release version in JIRA and create release notes. +Your feedback is appreciated +Code changes committed as svn revision +The patch resulting is really simple the hardest part being to have the interrupt at the right placeThanks buddy +as per Dims request. +Changed to incompatible to make explicit +Updated the review +CtrlShiftF worksFixed +Committed proposed fix in +Hi Thomas This normally happens due to two reasons +Thanks for your quick response I have simulated the cooccurence and make new one that support long for our projectdo you thinks which one can get better performance? cooccurence or item +Fixed in r +Im pretty sure that the same info can be obtained by accessing the axisservices context. +The next production image that is created will have this included. +This should be resolved in as weve removed ripple as a dependency +Will be an issue when batching is there +Hi Dan I was probably planning to see how Aegis was doing in DOSGI I believe I originally reported the issue after seeing it failing in DOSGI +But even if we fixed wed have this same bug if the master had crashed in the middle of processing the shutdown rightYou mean the fixup needs to be run when master joins a cluster too? That seems reasonable a rare case but yeah somthing we should plug +Closing tickets +Otherwise other tests will fail +moving off the blocker list put on again if you think we should fix this in kill TERM patch should be its own ticket +All of these libraries exhibit security holes in their serialization +While Im mostly sure the former one is not active in this run the latter is on the classpath and needs to be there this is a dependency of xerces when ran on would think some exclusion declarations would help but now Im confused +At home doing first the mvn made it so I could do mvn install wo build failures +Thanks for the patch Ghais. +Thanks for looking it over Ted +In which Tomcat version it has been corrected +was host configuration issue +Inifinispan does to both owners waiting just for the first response +create a patch file with the changes you want and as long as it doesnt cause problems elsewhere it will probably get applied pretty quickly +What do you think of this patch +attributes +The issue happens with +Our old was the same from the wording but I must have ripped up the formatting with some reformat in Eclipse blushI only use instead of because I like to additionally comment the interfaces in the near future +For maven users deployment plan project will produce a jar file with deployment plans +The patch fixes the reproducer +style couldnt reproduce the failure so I filed to look into it +We are seeing thousands of in the path to the classloader as depicted above +Theres something broken + +revision added explicit comments in and specific test in. +Phabricator doesnt like the patch as its too big and arc diff produces an error +Use MAPANON macro to fix build error on Darwin +attached the wrong patch for +A fundamental fix to this problem should let throws a when the input path is a directory or the path does not exist +Main Google number Adrian is the guy who was in touch with the provider lately for fixing this issue +Need the disable so memstore is flushed first. +Thank you +I guess so +I think we should look at adding to the release process script and having it update the dates to months in the future in master +But I still dont see this behaviour +Re javadoc findbugs I dont think it is related and I am seeing these on other patch builds +Im a bit confused about the difference between the three patches is it just the from the th that needs applying or are the other two still relevant as well +about your original codesuggestion box cant works correctly if DOM element to which it was attached just replaced with another +I have run tests today and everything was fine +The nodes will notice that and will close the corresponding indexesIt is possible that you want to have a state for shards +I must be missing something obvious +I cannot repro this with the latest code +should be looked at before CRGACould be done with different catgoriesuncategorized items or by creating proper uber features +This looks to be resolved in the most current versions. +What are the standardsOur messages are already externalized to reference in message files but the message files are littered with entries that are no longer used +bug in the. +patch has been committed please verify. +Attached a Patch for this +Screenshots of mentioned problem were like there is reference to different update site with Spring IDE and p pickups latest from extra update site for is fine as soon as there is no external update site is added that leads to latest Spring IDE +Sorry not +Basically you are not supposed to change order of builders at random +The alignment issue is still there +This was copied as is to PutDelete +This patchset isnt too hard to maintain as long as we dont make any massive changes to the classes in trunk before it goes in +In the supplied test case the message loss is expect since persistence is not enabled and the option to keep durable subs active is not enabled +We should only add something to that config if we know every project we integrate will properly support the given configurationIn this case whats being configured isnt a socket level timeout +You can commit your keys directly to svn reposinfrainfrastructuretrunkpeoplesshkeys is a preferred option now dont worry about attaching them to this issue +I believe that is the new +Thanks! is not a problemHibernate and Spring are optional dependencies +Sorry for the spam +Nithin is working on the same +To throw an exception in the case when poslengthlength of LOB I created a new +Complain with OGC changing names if you want for the future Id advise to also have a look at the relevant OGC specs to confirm what seems to be wrong is actually a problem +Duplicate of but we are trying to collect the issues we got for I updated to link to so we got it all grouped. +As with your other tomcat as service bug I think we need some help reproducing this +Fixed the JAXB TCK backend script to work with productized xjc module. +Minor updates added some javadocs plan to commit soon +Marking resolved as we fixed a bug that impacted memory and improved the default GC settings. +This sounds like a good idea Edward +Ashutosh has discovered a bug with the patch +The attached patch adds this WS classpath container has the same problem +Wonder if this is fixed by +Scott can you please attach Eclipse project that would allow to reproduce this issue +I cleaned up most of the issues but now Im stuck with some builds that depend on JAVAError JAVAHOME is not defined correctly +We agreed with Lukas it needs to be rewritten for. +I think bugs of this magnitude should be marked critical and removed from the main grails page +Hey Neha yes my hope is to get the patch evaluated as is and then take another pass at cleaning up the way we handle the satisfaction action as Jun and you requested and try out other approaches to the purgatory data structure asynchronously +If you still believe there is a problem please provide a sample project that demonstrates the issue with instructions on how to reproduce +Would you care to explain exactly what you mean by this? Thanks +Attachment has been added with description Documentation Class little bit of comments and cleaning as well as implementation of +Same problem here this time a DB installation on Linux on Powercouldnt deduct database type from database product name DBLINUXPPC +That would be great +Changes look fine to me +And then the combo is a mile long. +The problem is and also Oracle JDK do not cast nested Maps value to the defined generic type correctly +Youre rightFeng if youre still working on this can you update it as Steven suggested +Committed to both trunk andThanks Paul +Good find Brandon +Communicating the error state using the data object doesnt sound like a good solution you would have to check the error state everywhere +Indeed it looks like we still depend on symlinks in but no longer on trunk +in in SVN +I think I am through all the items you flagged asyes? or TODO Write test cases on the compliance page any way +bit late +why not just have it as an field +was reported as fixed +Closing issue reopen if there is still a problem. + +Ok that was a quick one my tests had been done last december with a snapshot version of Commons MathNow with the official release Im finally getting performance numbers similar to JNTFFTSo Ill close this ticket thank you guys +Hi thanks for the patch +I think bootstrap is probably already covered both by its natural delay for range announcement andbq +The and is sweet +Eight years since this bug has been filed +Thats why I previously assigned the bug to myself +Patch integrated to trunk and +For example set is probably not something we want to fact I would be inclined to resolve this bug with Yokos landing and just open new issues per configuration that isnt exposed +Though it works in most cases it isnt logically correct to expire old token only if a new container comes in or succeeds +not use system specific stuff like +Thanks for reporting +Do you time to do a benchmark? Im a bit busy right now so itd take a few days for me to get to it +the jdbc store has some issues in this regard also +I could go either way +Good catchWill fix it asap +I confirmed that the Or queries are using in both base and comp so those gains are real +Done. +Patch to add samplesextensionsant to m David +Please see the comments on Review Board +Glen thank you for your patch +bq +Bulk close for +Add something in the migration page after that +Committed small changes to and trunk +Thats elegant +need to update version information in build script of trunk +A similar problem with Firebird has been reported in +Thanks Scott +So with a WCF service hosted in a console application we will not be able to encrypt this information +So itll be a nice example +committed to and trunk with indentation alignment change +Its not just a file its a format +I will review +Please review +BTW emcc is not a generality its a truth +Ability to set the UUID algorithm separately for each database came up a while ago so I assume that would also be included. +Yes I tested it and found that if I change default value WARN to INFO no extra log prints to the log file when server startupBut if our app use log method it will print to log fileSo shall I make a simple patch for this jira +Hi Claus The real challenge with the scheduled route policy support it seems is in the route or a consumer at a given time and the most elegant and only real way is to use a CRON parser +Build Derby +No recent work on this one and we now have asyncweb projectso lets close this one to continue on cleaning +shipped in +Fixed strange SVN status to resolve patch import errors +The code generator this refers to is now out of data +Wow + for this committed to trunk. +I will refine my patch against the trunk +Can you share your dataset and style +I am very impressed with s response time on this +if were going to provide features in the runtime we need to be sure a development enviorment can propertly support it as well +changes all field types of Class to String in Meta +I think if the error message is fixed my app will fail to Start due to missing dependencies +Third version of patch attached +You should be able to drop the absolute directory prefix and load the file directly from the classpath +will take a look +The example that follows should also be inline with this +Thanks! As always dont forget the trunk. +People are using this method using reflection and now breaking example scoobiThanks agree the method was marked as package private so we made no guarantees about its availability +Fixed on the trunk and branch +The datanode was earlier using completeThats radical +Would this be less ambiguous +Attached testing script +The only real reason for supporting the other types of Messages would be users with existing applications that expect to receive messages of type for example +is this latest patch applicable for as wellI do not think that it applies to +Thanks for fixing +We have a fix ready for and onward that I am hoping to push before the end of the week +Ill try to fix issue in Uninstaller source code +Ill give it a whirl running overnight but I dont anticipate anything changing +Will need to ponder the implications for the management interfaces +The patch should allow user to dynamically change group membership for individual server +closing since resolution were set to unresolved. +Alex and I worked on writing C tests for this +Fixed in +I had set myself up for undefined behaviour under certain conditions when processing an Http response +ForIn there is no in the client directory +Ive added a switch property +Because is an interface any change to it breaks backwards compatibility +Its working in my setup too +Committed revision Thanks Simon! +Is there really a reason the key needs to be an object? Why not just a string like scaleInfo +Which version of the is in your pluginlib directory? I encountered several strange errors with the last version checked into the plugin project prior the release +You have a problem now with folder icons not showing let me know if you have trouble curing itI will add you soon I am waiting on a few more to do at the same timeGav +I think the right thing to do is for me to create a JIRA for each of these bugs and attach separate patches to each JIRA +remove files per second or so +Then you can use or if you like with either POST or PUT +Current behavior is preferrable to the old behavior +fixed thanks Roger +Attaching preliminary source code +INFO Running for rows took ms +Where is this file located? I couldnt find it on bad +Start network server and run java is perhaps a more specific case of but maybe there is something specific that can be done for stream reuse if a general solution is too hard +now the test project is added as test to verify wrong ve seen that the bug is fixed within it can be verified by the integration test Ive added as patch +Geotools updated eGeoserver updated adc +the start to a patch but I only did the build piece here just as an experimentSolr is formatted differently than Lucenes +Patch was applied +war works in Jetty fails in Tomcat +There has been no feedback on this issue so I am closing it +Im canceling the patch for a moment +This patch fixes the issue by copying setting from toThe only exception Currency is not copied +Should be fixed in rev +Thanks for your time +Simple fix to to check for patch had incorrect variable name +bugfixed version of bug is still present in rel and have tested several configurations and expressions to see if the proposed solution fails and no errors found +The failure is a known issue see +not yet subscribed +Is there a benefit to using a property like this and not using Javas Deprecated +This could be possible but it requires more thought +This is in master and in JBDS B +If there are any problems please let me know +Can this be closed +Please close if its ok for you. +Any good idea +Trying out the change now will submit a patch if the tests execute and pass +Stack trace is in the bug Kihwal mentioned +Try running QA against this patch as for the latest code +patch applied +I would postpone this changeIt would be great if we could commit this before because I want to add new options to useful when creating theI can commit the patch now if it looks ready to you +With a copy context bool property in maven which could be setinOr change to use a token to set the context path just like db username password in +The issue is fixed in a newer WS version so it could be in hopefully +bq +Leaving the issue unresolved because the original issue was a heisenbug and we dont know yet whether the problem has just moved some place else in the storemore tests. +Hmmm +Thanks Kanna for reviewing +Which version was used + potential suite of such helper methods? has similar stuff some ideas here? will help with efficiency in obtaining locks +Can you fix this upstream in the WS product branch as well? The examples are maintained upstream and synced with the EAP branch for every release +this appears to have been ridiculously easy +In rev I added exception handling to ensure the broker doesnt blow off when an invalid rule is triggered during an acl cases to cover the new validation code +Cassandra is starting well +Thanks Nathan. +The cases where you would no longer need a PS for the life of a connection seems exceptional enough that there would be no harm in keeping itIm not fundamentally opposed to an API call to remove Im just trying to keep things simpler if it doesnt actually provide any valueBut this raises something I hadnt considered +Unit tests pass for me +Closed based on Andys comments +Attached patch with the discussed changesx patch follows soon +more API clustterOk lets keep it open but I take it out +I will update to the newer parent which doesnt pollute the users build with extra dependencyManagement +when user left the first field blank the message for first field shows X +could you attach a screenshot of the new ui +A lot of modules out there use SCR and its like Blueprint part of the spec +Attaching screenshot of ER build with three apps deployed deployment was deployed using console and the other two using filesystem copy +validation removed in branch and. +Thanks for the patch +Weird that you cant delete a connection +This was an oversight We changed back to leniently ignore unresolvable placeholders by default but unfortunately forgot about its sibling +Does it mean that the caller identity propagation across servers is already available +Justin does the new icon service address this issue? I guess so but I may be i believe this should be addressed with the server side icon renderer +Look at RFC and of +Ive forgotten to mention that when using a separate class loader then undeployment doesnt work at all +Closing as no longer applies to master. +The mixed kind of delete only will be not possible in that caseRequesting opinion fromNOOFVERSIOSTODELETE nobeNOOFVERSIONSTODELETE no +How would I fix this +How exactly were those logs generated from an embedded Maven like IDEThey were generated from Beta the only location where I see the mentioned warning is at the very end of the log after Maven itself has finished buildingThis is why I thought this was a bug in maven but I was wrong +I cannot reproduce it on my setup +What would this accomplish +I am positive that a line number can be reported even by the NPE +The change to the import statement in but no other changes to the file +My web research turned up nothing +the GUI only offers the default hardware profile +Thanks Oli the patch was applied as expected +But please peek at the code and the call stacks or method usage and improve it +That sounds similar to something Ive observed after I submitted a patch for but I had only addressed manifests +Applied the workaround in r +Failed test precedes this patch its fixed by +It was introduced around there another slowdown you think? Or did the fix not make it over to the branches +Probably by borrowing the jvm search stuff from debianinit and adjusting the directory list appropriately +Im happy to retire this one unless anyone wants to fix it +Either this Bugzilla diff is not working properly or youve made a number ofother more stylistic changes to which make it very hard to seethe real change you made +Ill be uploading a patch shortly +Yes there is some minor involved +BULK EDIT These issues are open to be picked up +Not going to make +Reopen just to update it for Maven Quickstart archetype and the generated application updated to use the archetype to and update some plugin inside. +First attempt of Bean Validation is committed into SVN + of a bulk update all resolutions changed to done please review history to original resolution type +patch added +This makes our bundle appear to require several extra kitchen sinksI fixed this by adding exclusions to our configuration so all the packages that bnd wants to import that are also in the header are not imported +When will version be released? +Im not planning to verifier right now can we do without the changes for this patch? Does that only affect the IT + add doc for getActivity would IMHO already be the most important Ive taken the example expression from the field injection test case +Bring it on +Looks like it solved it +Jasonyes it does but as Alex mentioned we need to change the default ordering of the interceptors +bq +Each fragment is expected to be unmarshalled in its entirety +Thought about it a bit more and discussed with Hemanth and Devaraj +yes we can iterate over the projects and get it to dump some xml file or file of your choice that details this information +This has been reviewed and merged. +how would we get a into the Host in the first place +Initially I see some things not much efficient like I used to store querySessions in which I feel is not the best solution even though I am thinking to remove this entry from map once job is completed in +Also supplied certs are checked to make sure the subject name matches the host they are connecting to +Already fixed +Could you provide full stacktrace pls +The comment in the stylesheet isnot necessary +Whats your local time right now +Could you please tell us what you used first and which one you used to solve the problem +The problem is probably in but the conversation about the problem is completely in this ticket so I the ticket +Do you have a use case or an example? Ive looked at the patch and it seems interesting but I dont understand what you are trying to accomplish +The problem reported by the assert isnt fatal and if youd used jars the code would just use a heuristic instead of the missing statistics and be able to continue +Both rate objects have an invalid rate property +Testing with on client and one HRS doesnt give bigger or smaller numbers +Hope I didnt miss any more +committed w suggested change to bool block instead of int blockfor +Alan I am currently using Snapshot as my project is effected by if I use +Im against this +This issue is no longer blocked +I assume this bug can be considered done with +I would think a good first step would be a call that dropped all the statistics for a single table +I have been meaning to get back to this issue after Bretts refactoring but havent found the time +milind +Thats too infrequent to diagnose the problem +No change is required to the Maven Launchpad Plugin since this plugin just uses the latest RELEASE version automatically. + has been marked as a duplicate of this bug +If there is or more then throw a failureAlso I suggest to add this logic to the doStart method so its run once upon initialization of the component +I cannot replicate this error +Thanks for your contribution Arvind! Thanks for the review as well Ahmed! +and confirmed that counter still goes upThat is the correct behavior +wcs text +A missing test is on ethat shows that xhtml snippets are in fact converted to html snippets you only test that no doctype is added +The new impl looks good to me +I have just committed this +Verified fixed at rev. +Konst would you mind ing this as is? If you want we can file a JIRA for the overall improvement of removing the support for saving VERSION files from the main source tree +Verified +thats already tracked in +Adding dependency to patch that adds var attribute support to image tags +Thanks for pointing this out. +These seem to work for this shell +This patch looks good to me +producers and consumers on the same broker send and receive messages +Harsh do you have any concrete ideas? So far I think logging a ERROR is more reasonable way +Sorry for the inconvenience +I started to look at the code and I have a couple of issues +I bootstrap the spring context and run it as a java application +Later we can figure out how to add it back on +Thanks for reporting this issue +This implementation works well but isnt as efficient as it could be +looks good Ivan +Maybe one of the other Cordova devs would know +For example view matrices should not return view matrices but should return something of the type of the underlying matrix but in the right sizeThe issue is analogous to the problem with constructors compared to factory methods +And it seems easier to explain +html is now appended to all links +This doesnt seem possible as there are checks for the number of arguments and arguments and their values cannot be null +Hi KnutYes I think that is the case +Sorry +Mark as closed +We only support Node and its definitely defined there +I replaced jar from with from grails and it didnt fixed problem for +Ok thanks for the confirm. + Adding that method to Vector +Can we use in the scheduled case just for consistency with inbound adapters? +RTC work item defer of items from to +Gary Take a look at the th comment +Closing issue it has been resolved for a long time. +If it occurs again we should try to find the and any other relevant files and attach it to the issue. + +Thanks. +I made a pass over the trunk patch +No longer an issue +You can help by taking out the merge and replace the jar with the onein your continuum to see if it still merge param is used also on the first checkout +My project is a GWT project +Good fix Im not sure +Update I realized that I didnt finish merging everythingsmile Once I did that properly the tests passed +Stack can you please include this in RCThanks +r fixes a few of the broken +I am experiencing the same issue +If you disable term freq you also have to disable positions +Its actually the first item returned by google if you search for maven directory structure +Indeed will by default flush the Session although you can turn that off through specifying FLUSHNEVER as flush with out new call on rollback the Session wont contain any unflushed data after a rollback so the flush should not actually recommend against using in that style I rather advise to use an outer transaction and inner transactions with PROPAGATIONREQUIRESNEW with no in the +I will look into it asap + +Closing +looks great to me giver +Verified +Lets after committing so theres only one code path to debug +This can be reproduced by running a dynamic worklfow click stop and then rerun +Attaching the suggested patchThe test modified accordingly +Replaced tab by spaces +Here ya go +Patch submitted +Turns out the version Im working with is +Reverting back to the previous fix version +dont believe this method is actually used in AS +Let me do that right away + for trunk and for the most part everything looks good here +bq +I would think so +Thanks David +Non bugs should not have a fixed version. +do the changes in the ecommercewidget JacquesIMO your guess is right for tel number in same as Shipping Address case and there is no reason to bind them +rejecting since nothing we can fix at this time. +Patch against r +Bulk close for +Credit for this goes to Basab Maulik who did initial patch +The same file could be validated a few times by different build threads if user have dependend projects +for decimal separators +I agree with Martins concerns on mixing threads and signals +Two issues remain integration with handlers and security +I could not run svn command line since seems to use and my client is +Committed to main trunk with revision . +Patch that does basic work + +Does this correct +Thanks Andrew Swan for your instant replyRight now we dont have any ideas on any protocols to achieve suggest me some good protocolsbest practices to split web tier and application tier. +fails all the way back to so does not look like a regression +Id also like to bring the API docsup to date +backported to with revision . +the zk connection was created more than once in the producer +does compiled test stuff need to be segregated in targetNo +SteffenIll review and apply the patch within a few days +though its good for a starting point +Please commit once and unit tests pass. +Sure I will put the patch +Thanks to the fix for Camerons sample project now work correctly with the latest +Yeah +Previously in WAR deployed mode we wrapped the webapp classloader in a +HiIve managed to reproduce the failure Ive denied full control for me for the following registry entryHKEYLOCALMACHINE SOFTWARE PrefsSo from my point of view the problem is in your testing env +Correct serialization to be synchronization in description and change were made a while back but I neglected to close the issue +So I should be looking at node for the corruption right +The static build almost works now +Thanks. +How do you inject the respective HTTP headers +But perhaps you have fixed it since this mornings tried switching out the snapshot jars with the rc jars and the test passes fine. +There is one caveat to this it may be possible to have some constraints on the routes +Thanks Aimee for the patch I merged it in dev commit acecaddebdeafeca +I think it is best to go with Keiths suggestion to introduce a new maven project for the filter implementations +done +its pretty cool +I am no longer getting the NPE with the updated swc +Works well Julien congrats +Corrected one test method as mentioned in my previous comments +DEBUG loading main INFO Replayed operations from the journal in seconds +This task can be closed +Should all this enhancement stuff go into a separate Jira? It feels like the issue reported here has been resolved +defer all issues to +Nicolas Hm dontt understand how to implement approach in job without extra complications +me know if you need more info to debug +Apart from I dont see code specific toShould this test file be attached to an HDFS JIRA +Patch I want to commit to I refreshed it again and fixed one other thing regarding the attribute because it wasnt set properly and refreshedThe patch for trunk enables compression by default and removes the tests +Please close if happy reopen if not +Ralf I just found your forum username so Ive already added you +I resolved this with some additional refactoring and now have the compiler running in force mode for a number of continuous integration builds +add component KojiActually the current code seems right if we dont output subclusters we need to include all documents of the cluster including those from its subclusters otherwise the subclusters documents may not appear in the response at all +What if we made a new overridable method that let you create the Runnable then? I think thats preferable to exposing the executor to subclasses +Adds a maxmemory attribute to the javadoc task inThe maximum memory for javadoc can be set via the property which is set to a reasonable default of m +Can you try this patch +This actually looks like a problem with the remote EJB client not remote naming +Bind paths that hit null values in a path are a completely different affair and are in general not handled by the framework +Heres a patch of the latest discussed approach +Could someone please test +Checking it in +Another potentially useful future improvement would be to check not just the root version but also all other versions in the history +Simple patch thats added a new setValue method +Tweak the location of the default FC inApply from srcmainresources +BnestedexceptionisANoconversationcouldbefoundwithidperhapsthisconversationhasendedFDo you see any workaroundsBest regards support for the mappedHandlers +Krystian Can you attach a simple project with a failing test that demonstrates your concern? Thanks for the help. +Also I would like to say to Ricardo that his problem has been solved because when you installed Maven it created de +It does it now +also it would be handy to have a constant on that contained the name of the config property +The typo was in the first sentence and should have readEditing a file in Eclipse produces a red problem marker in left column of the XML editor whenever the project element has a xmlns change was as to has + for the to trunk +ThanksBest regards by Andrew. +Hi David +So could you give me examples of what to go into the new constants interfaceclass +Im mostly on this since it affects everyone and the only difference is this not call the JBTIS one orbtw +David a few days ago I just checked in a small enhancement to the Ant task for the XML code generator that adds a property that allows you to set the name conflict strategy +patch looks good to me +Link is pointing to correct URL in JBDS +What makes you think the visitor isnt going to do anything with thedata object? It is provided to support flyweight visitors tomanipulate some structure among other potential usesThanksDanielSent from my iPhone +Works nicely and a cute solution it seems to me +Here i am attaching screen shot of that screens + + +Nothing to do for this annotation + should allow a method of escaping the colon character preferablyusing the backspace character in the position before it +is a valid root +Ill post a patch for +Resolved according to the description of the bug. +Please Code Review +Id be more conservative and do instead of + +Here are the logsNumber of sent rows varied quite much on our tests +bq +The problem is if you dont have a secondary namenode checkpoint because for some reason the secondary namenode failed silently +Its up to the user to implement the watching of those interfaces if theyre using instead ofThe goal of changes like this should be to make it cleaner and to make smaller not larger +In the meantime I found several other issues but this time with in combination with so I guess I will not push my investigations in that direction soon. +It appears to be the fault of Xwork +Im a little comfused could anybody explain whats benefit of this issue againt? IMHO MR tasks with security enabled should be in +Done +Much appreciated +Thank you +Yeah at least the wrapper layer would buy us some time and a way to keep karaf branch up to date +Currently a room has been allocated on the th from to PT +heres a test +It could even be included in +Committed revision . +That should work around the problem for now but we should restore the normal settings once this problem has been resolved +If this can go within the libphp folder I think that is fine but this should not be in the root of the project as it does not get utilized by anything else within is useless in that repositoryIf it is not possible we will use another configuration of Composer +That would be an even better contribution. +Why is the change to necessary +unit test showing the usage of the new Array converterwe need toa Put this in place in b look to having it behind the APIand we can resolve this issue +Maybe when a master shuts down cleanly it should clear up the folders in ZK +Changing this is a major change and a breaking change too +We could probably just throw an exception if you try and pass a Get a null row too +If command line options are provided they should take precedence over the script file +Refer for patch that adds support for GET POST and HEAD and support for PUT and DELETERegards all now includes for RESTRegards +Ill look into and fix the script if necessary +Build same problem in +The code comment changes look good to me +Solr does not need to add any intelligence here +Forinstance the user may want to assign a unique id to each and every request tothe server +This issue is not approved for SOA CP +Billy can you test and let me know if this takes care of it +By uncategorizing the two other features I managed to clean up the install UIAnd it still installs fineSo +This needs to address Im solving this jira as the has been pretty much reviewed improved in the latest months and reached a quality level that should be fine with the release +Probably wont change it ever +To make things worse I have actually stumbled over the issue in which uses from ant but copied and locally modified +I tested with the nightly build and it appears to be fixed +Fixed core in rev +After further investigation the most appropriate fix is on the designer side +juni Messageremoved the BAM examples from the distro until we can figure out how to get the example working againModified activemqcameltrunksrcmaindescriptorsModified activemqcameltrunksrcmaindescriptors +account added key installed added as committer to cargossh datanga install a +The upgrade to version fixes this. +Role test for components attached +I like the idea! Go ahead +Sorry +I checked the code and the groupCache is updated when doing a move. +AdrianIm not entirely sure what youre suggesting or more specifically how it differs from whats in my patch +It may be that it doesnt work standalone but we have to get it working in continuum and later in Attached patch makes continuum use +You can use the with the patch in +Typically the number of frequent subtrees increases exponentially with the size of nodes +Should be verified in gatein master and reopened if still valid +I am using Geoserver and +Hmm we did have a JIRA to address that at one pointIt resolved things for quickstarts only AFAICT +All Done. +In fact the will never be called with tapestry +Attaching patch for ydist on behalf of Amareshwari +It is also desirable to support duplicate element only issue that I am aware of is I didnt expect a fix to be honest +Changes committed as revision Handle the exception thrown when accessing a path name such as drive not ready from aDont see how to add a test for this as a invalid name worked fine only the exception due to drive not ready caused the problem. +Ill update the document. +Can we close this bug or are there other concerns +ear AND the +Thanks for review +Log with error messages for whats the output of yum repolist yum list +Passes all local tests +Complete matrix including also older milestones up to cr is attached now as well +Spark also has a maven build that you can trigger with a Phadoop profile however this does not solve the issue belowSpark depends on and +I did not know that patch behaved like this +pending t included any tests because this was just a pom change +So for now please change the. +I will verify this shortlyIf this is true many thanks to whoever thought of this before me +No plans to port to +I mean changing the value of default to blank string is unnecessary +Is there an official Railo solution +Tests appear to run ok so I went ahead with it +If someone wants this functionality and doesnt want to wait for just change the package from to piggybank or equivalent and go generically forth +It seems like this was already fixed in the Release Notes in +This issue may have been fixed and code may already be present in the old trunk +This is mostly a direct port of and +Having said that if I cant find it after the next round of upgrades Ill think about +Attached the latest patch for +Actually this is slightly different since we may add header in the server side +You can check it on line of the patch you could see that method do added to so there should not be such a problem +Thats strange! I had checkedserial right after the commit and see no degradationI will take a look Sergey thanks +Yeah there are some problems with distributionCan you try with that has just been released +and we really shouldcall it Entities view instead of classesLet me know if that is something you wanna work on +The dump is currently being loaded into the test repo +Best regards George +So yeah we need to put that on our site is what Im this even necessary now since theres an Ironbee plugin +It is r in the web svn which is in Web. +bulk defer of featurestaskswishes from to + +Seems comparable within margins of error +But the basic browsing of a local file system works and the package doc and test Ant tasks run cleanly +That would work and be clear +attached bpmn file which has Script Task with multiple connection which is valid but not working in runtime +Integrated into trunk andIts a good idea to add a test +Disregard my last paragraph I didnt have access to the wiki once I did I was able to refresh my memory on how we decided to do that +Applied in r +there isnt a version ID in the token to bump +moving back to next need some idea of what is happening with this JIRA so I can schedule docs work for volume of material +My mistake +If we move them to another package we should turn off the public javadoc from that package +a Unit test for +Could you possibly expand a bit on what kind of setup you were running against? I can see how the XSS can occur without validation of the entered username but some more info would be useful +Committed a fix in rev Please try it out +Without IT so far thats yet to be +Will keep at it +I also could not reproduce the problem on both with IBM VME DRLVM and on Ubuntu with DRLVM +The only solution to my problem is raw SQL which is also the simplest +thanks jukka +Change status from resolved to closed + +This patch only applies to for the moment have to change row locks for +Chris thanks for this patch +And now who in the hell are you? OK +Adapting the testcases with the updated URI +As can be used without a transaction manager the optional import statement is normal. +category would be OK too +Have you been able to reproduce it +Works perfectly now no deadlocks anymoreThank you very much for the quick bugfix. +Migration is complete. +I tested with and the issue was resolved. +yes +Attached is the test case that fails +Test case +Definitely a bug in ACF in my opinion +specifically Im hoping that we can create a well known file in the output directory with a log of all interesting job outputPerhaps we should put it in a subdirectory INFO of the output directory so that when the directory is used for future input it is not confused with data? This will also allow us to add other things later +Im closing this ticket as I agree with Evgeny +The new patch applies cleanlyA few tests are failing though for oneIm looking into it might be that a problem is in the test +I also think we should not make the Camel DSL more complicated if there is already a solution for a given problem. +Hello BSP on YARNSo for me the first snapshot is done +Once Ive finished cleaning up whats already in there Ill take a look at Mahout. +We should also experiment to build Karaf with Java to identify potential issues. +The patch was generated using git diff HEAD +Generally I like what I see +Plus can use not only IPADIC but also NAIST JDIC +finished +OK +If a commit fails then the system is almost by definition in an inconsistent state +ATS cannot reload that currently +bq +Updated patch same as before just fixes a small javadoc no space after a param +Fixed for and +Verified that the fixes have made it to the latest alpha manuals +The refactoring work done in the last patch is much better then what we have though +Cool code +Just to make sure were all on the same page my understanding here is that we want to write a new stress tool not really update the current one +The throw exception was added to + for the patch +Thanks John for bringing this up and also for providing the patch. +Closing as wont fix as this requirement is superseded by the Direct WMS Integration feature. + dont see any need for tests for this +Ok applied in trunks! +Patch with fix for findbugs this patch doesnt apply anymore +Ill try to come up with a boot order unit test shortly +testcase point igal I will review this +I just committed this patch +It comes up every once in a while +Fix is in +This patch adds capture support for There are bugs without a doubt but things seem to be working in a somewhat stable manner +In subsequent changes we should post more details on what has changed and how to configure federationNote that the existing single namenode configuration will continue to work without requiring any changes in federation +The tag is the release tag in the ApacheCommonLoggingBranch branch +Thanks I just commited this. +Patch to add Char +Final one I promise +r +Thanks Stephen! Unfortunately the migration from RF is not as easy as we would have liked +Sorry Antonello this just allows the current uDig automatic load save cycle to function when uDig is a plugin to another RCP +May follow up wother changes to make sense of what role plays now with the CLIs plugin interface +Re not if you fail to call super +Thanksrevision +Sorry about that Yifeng +Just a misunderstanding of how Axis works +it looks quite strange and leads to many questionswe also talked about introducing a spi for not forcing the usage of furthermore in wls just the old version works correctlythe rest works without problems i dont think its a security issue +OK Ill see if I can work on it Im pretty busy too while in a different way o +For an applet this letsthe JVM not have to download a jar file just to search for a jar that may not bein itAdding indextrue for the logging jars wont help the logging jars unless thegenerated index references other jars containing logging dependencies +Works for me die die +Upgrade in progress +pdf Are they auto generated? or Do I need to update these files also +Tested in IE behaviour movie revision +Thank you GilbertoIll take a look at it ASAPAdrian. +There are still some corner cases where it may produce a false but all that I can think of are invalid java and would probably require a lot more complex code to get it right +Lets fix this for final +Thanks havent seen this option in any docs so shall I create a docs bug +Im checking EAP now it might only be set to headless in the init script Since in the the server is by default a Linux service +Exists in version as reported by Freddy so reopened +Of course if bulk of them fail then we know there is a serious problem +Will be fixed in next version +The bind tag may of course participate in a scoped nested path next to other bind tags +Herve just to let you know that Castors XML code generator has been able to generate Java code for at least two years alreay +still need to write the content the syntax errors are fixed +Time to pull +This issue may be the result of but Im unsure at this time +Dear committer please use for patchingThanks patch does not work on the current svn HEAD +Did you mean to commit it to the branch as a bug fix? +If this occurs then the bridge will try Max Retries to reconnect every Failure Retry Interval milliseconds as specified in the JMS Bridge since a third party JNDI is used in this case the naming server it is possible for the JNDI lookup to hang if the network were to disappear during the JNDI lookup +More work remains but we dont need more of these bugs to know that. +Willem +Dont worry I will take care of this Will if you want to dig into this please start by reviewing the patch for as mentioned above +I left as public class since it is used in old api class +For example if afcalendar has a facet contextMenu which expects an afpopup as a child tag then the tag doc on the contextMenu facet section will show Allowed Child Components afpopup with afpopup being a link to the afpopup tag docAlso as part of this patch if a component has a required ancestor tag then that ancestor tag will be listed as a link in the summary section at the top of the tag doc pageFor example if afcarouselItem has a required ancestor tag of afcarousel then in the component summary section of the tag doc it will say Required Ancestor Tag afcarousel with afcarousel being a link to its tag doc pageThe tag doc generation relies on the metadata in the component definition files specifically and +patch for Jitendra Daryn Nathan for reviewing the patch +So all issues not directly assigned to an AS release are being closed +The patch is in +I agree the icon is confusing the image should be changed at the very least +Do you wanna just give me the revision number against trunk? Or are we still wating on a user that actualy has oracle to try this out +Patch looks great +We can this bug if you provide more information +cart and checkout pages look ok here as well. +depotscratchbobaProjectdepotscratchbobaProjectdepotscratchbobaProjectlibs +Could you please try latest SVN? i could not get a stack trace from samratthanks safety on is a known problem in WSSJ and +We have been trying to keep standard untouched as that is what passed the TCK tests +Yes It gives an NPE if numShards is missing +This issue has been resolved for over a year with no further movement +The difficult use case is terms index as it returns the term for a docid froman ord value +Since this information comes from we probably should note that in the LICENSE and NOTICE files +Tom will it be possible for you to try things with Cordova +I used jps to get the pid and jstack to dump the stack for all threads +Tested the latest patch on cluster and ran into the same problem as +I tried it and ended up with similar number of changes +Alsofixes null pointer bug found by aggregate tests +Thanks Amar and Luke! +How are you planning on doing this? We surely do not want to add another configuration for the Cache at the connector level and manage its configuration and dependencies +That would be great thanks +So when majorVersion goes to we shouldnt be checking minorVersion rightThe check is for +Ideally this might not need MR +Modifying the enumerator inside the block seems a bad idea to begin Ford said on mailing list that Matz had stated that under such circumstances the behavior is undefined +Please review +HMACInit seems to be defined in libcrypto and fedora policy regarding DSO linking prevent it from being picked up as explainied a few comments above +There are some startup issues that I missed will look further. +Especially if we end up taking the concurrent name +Stefan +version of hsearch in wfk is changed again now it is bom gav in section it should be available for review at the same URL. +Thanks for your reviews Jarek +The main modification here is that i dont have the task metrics context introduced in my earlier patch and instead do everything in the tasktracker metrics context +Attached patch has a first draft of the code changes required in +Also the should be modified so that abacuss javadoc is included as a contrib Abacus groupA updated patch is availableA new patch with for abacus package and updated including javadoc for abacus +Here updated patch use button instead of buttontext +I dont mind having them duplicated into a second view but its a bad idea to move them away +Bulk closing stale resolved issues +I have used that to implement this improvementtrunk +What suggests you that it is not + +For Intro purposes this could also just be a matter of parceling out the failures that remain today into separate bugs so theyre more approachable +it might even be better to create a new bundle with these shared classesOk Ill create a new jcrbase bundle and put the JNDIRMI access code there +Is this still an issue when using +to apply the groupId knowing that karaf is pushed and as well as am reopening the issue as there are a little more need to be donei Align the featues descriptor with the updated groupIdii Align the itests base with the updated groupId +Actually this is in the task so here is the relevant patch +see error message error downloading +BE sure to check it on JSP page +As such I modified the summary to be a better reflection of the proposed additions +Since our compiler so far requires to load classes using a class loader the parent classes do have to be present too +Therefore it is required that a font config file be providedeither through the or programmaticallyo just use the c switch to specify a fop config file in which a link the file is given +Change from addContext to addWebapp +This is kinda obvious that it needs to work +Thanks Sriram +No answer from dbo +Just pushed this +I respect the communitys opinion on this matter +So we need more of a delay before running the tests maybe +The two request readers will clash despite being associated to two different services and being located into two different namespaces +was preferred by the but the felt this would conflict with DOS path names +After search theres file I am using propagationjoin will causing this kind of situation but while using propagationnone will not can I know why Best regards Anna +Again as they come up I think we should clean them up +The attached pom is not usable as there are missing dependencies +attach a new patch addressed issue +Users might be confused that the setting doesnt apply to the normal edit log loading of the NN and the user shouldnt have to run the edit logs through recovery mode first before starting their NN normally +Size of the stack required for guard page setting was tunedAlso several bugs was fixed to enable SOE support in stress environment and on Linux platform +completed +This is what I propose for the POM file its based on the last one off the repository and has no dependencies + to make that happen both JBT and JBTIS for that data related category +Patch applied at r please verify +I will be adding a wiki page soon to detail how the branch can be built and tested +MAC OSX isue see prev Jasons fix above does work keep in mind that some OS X updates reset MTU settings for lo! Something I just noticed +I didnt have to make any change when deploying in another application server so figured it shouldnt be needed just for Wildfly +Committed on and trunk +Need to fix basic inconsistencies on joining first +Thanks Francois +Patch attached +Closing +Juca close this jira and open a new one for fixing any test cases that may still produce transient results +If not then we should check the correct functionality of the described in the linked issue after fixing this one +This is committed to but need a unit test for trunk +This should be fixed in the Java backend is that possibleThe language semantics is in this case just not implemented correctly in Java + +Hi Jarek +bq +Release published closing. +option is fine with me +Renaming the image avoiding the prefix struts and amending the properties file accordingly allows the image to m not sure when the fix happened but I can no longer reproduce this. +Arg +Verified that this works in configuration but its too late in to introduce the change +How the above was interpreted as an ability for an EJB to expose methods on a proxy +The patch part of the patch in +Committed revision If there are any problems with hudson ill yank it +Is this issue still reproducable +form +I changed the names of the files +we shouldnt allow getkeyrange to DOS ourselves +I will trying building from there +This test had failed before in also on a linux machine under vmware running tests but that stopped in The vmware machine running the linux tests was moved to a more powerful vmware server on Dec So I think this can be closed as cannot reproduce +If done in time we can bring it back. +My assumption is that this has been broken for many releases but have not had a chance to check yet + + +The stub is always generated wrongNote that Intellij and Eclipse both seem to work fine +Lets commit to if you get a chance to merge it in Takeshi Thanks +Renderers use direct read in the bean object making simple accessors override override insufficient to completely hook all accesses to any given property so I guess thats an argument in favor of final to not give false perceptions to users +Thanks for the patch Oleg. +is done +Done ready for you. +Im not sure it is required +After waiting second for each context +JacquesDo you mean to say that the same Ajaxified functionality should be available on the newcustomer page but which running existingIf yes then we were planning to do same as you do not like the information of new style newcustomer pagePlease confirm so we can go and upload the new patch but in that case we need to add the JS code in ftl as currently the JS code is there onlyecommercewebappecommercecustomer +its just that we have to think about this from the start +But I guess I can look into it +Having written the converter code for the Atom protocol in the first place I believe the patch to be good and precise +PDF with fonts embedded +You can also use file protocol to specify remote repositories and provide some jars for those artifical remote that you have to kick Refresh Settings on the Maven preference page after editing +changes all uses of on the client to +I removed the slfj import the jars from the was being extracted in a different directory which is why they were not included in the classpath +Is anyone working on this? Or is this plugin dead +Thank you very much! You seem to have reservations on the fix itself +It seems a random informative fact lost amongst summaries of new features +Pretty sure this was fixed along the way +on patch +What do you think +Heres the fix +I do however see you having a line commented out dealing with transactions +Actually the attribute is called failover +The only way to around this I can see is all users running apps on Unix must use and windows use as separators which is even more confusing +The properties discussed in allow you to limit the total processing power that a process gets without pinning its threads to cores +We could avoid lots of repeated object allocations with this change especially for structs with lots of fields +Ive committed it there too. +TamerFinally I did it at r for trunk and r for +HTML PDF etc seem to work +thanks a lot +We still need to upload it so that we avoid the need for the maven repo and can use only in the future +The above Red Hat script breaks +Looking at what was done under the old splitLock it all looks safe in the face of concurrency +Seeing a summary of files changed at a glance in the Subject helps one to decide whether it interests them or notIf you think this a good change Ill take a look +The new patch looks fineFor now I am removing the patch available state until this patch gets reviewed +Summary of issues related to dynamic checks Not inserting checks for the method guard and method parameters for loops closures have a lot of bugs and should be completely refactored +Here is the patch which would fix the bug +Just to comment we got not specific standard so we should just align with content services for this +fixes this problem by not installing the car files inthe first placeDont apply this patch +This would be cleaner than creating an external process +I think this is a duplicate of +Updated patch to address comments and add error handling tests +Verified. +Adding a dependency on is a solutionOk I understand our plugin needs at least eclipse andthen something in the api changed for versions newer thanThen someone added the flag for the newer version to make itcompatible +Is there a reason this patch is unassigned for +Attached patch makes the following changes Audit log is not logged for invocation of methods internal to the namenode +This is easy to keep track of since the initial leader and isr request has a special init flag turned on +closed +Lets hope Ian can fix them + for this feature +Ive found an issue with current implementation of integration scripts for classlib and drlvm test suites +Need to think about integration with libraries etc which users will run in to +Ill move it +Just thought Id mention that the forum discussion says that the interval snapshot mode is not going to be deprecated after all but this issue was already closed before that decision was made. +patch looks good +There appears to be no patch for is that right +How the hbasehadoop doc +If an element or attribute doesnt exist then just return null +The code would crash without them +Robert I tried to make it consistent to the issues but I would vote for both! This makes it much cleaner but we need to change the one tooI will quickly change my patch +Now the hadoop rpm does have shared libsDid I miss something like code stored in jars +Next time please consider choosing a sensible priority Lazily accepting invalid input is definitely not a blocker +Test case using jdev g +Ahmed can you please address Hiteshs comment? Also the bit about javac warnings +Thanks! As far as I can tell that documentation has been wrong since day one +Paulex it works for me +Rmove tose form PATH and add the to the PATH +Its just a location in dfs it cant be full if the dfs isnt fullwhen taking out the trash the idea is to remove the oldest items only if theyre older than mintime until the trash size is under maxsize and then stop +Hi Emmanuelthe fixes your are trying to do do they concert Acceptor only? or Connector tooI am not sure yet but seems the problem is reproduced in our production system with +Im not well aware of this product +Committed revision +Merged pull in JBDS +and were missing +Fixed suggest we changeto be involvedtoto get involved + Username prabhakar +sicne verything behaves as designed I am closing the issue +As long as the sequence numbering also works when the CP intercepts a scanner and provides its own implementation +Let me put the patch on the subtask so I can leave this issue open +this fixes the issue +stale already +It seems like we would get clashes on the session keysAlso does the code correctly handle the case where the server crashed while writing a temp file to disk +Maybe forcing a dialog up is too much and I should just change the default username +Id suggest two additions +Closed all tickets +A new admin command and corresponding bin script +Or simply use to convert the configuration into a Propertieswhich is Serializable or loadsave the configuration on a streamHowever Im not sure this solution will suit the needs +do you recall which test youve seen this happening? Any test in specific? +I just checked this morning and I dont have any of these files hanging around +Patch which checks for this in for review +patch defines extension applied and committed +The positionning problems are solved +If your talking about the root project I dont really think it matters but we can certainly open it up to the list +Fixed +Could we add the latest patch to commitThanks! just committed this +We need to define license file location in order to build coverage report this be resolved now? +Hi ChristianIm working on it now and I noticed removed the unit tests ofCan you give me the reason of itWillem +Putting the footnote in the header is the perfect way to achieve this +Please attach your JSPXHTML code +Otherwise i can create patch for version directlyThanks +A cyclic EJB dependency leads to non deterministic behavior +in the client jarmodule +I tweaked the code so that you shouldnt get a warning now I hope +I just noticed that the xdocs for are wrong +In fact directing output of project linked source folders to a specific output folder and including it in the magmarun classpath can ease the pain of rebuilding themSince there could be exist a problem with duplicate resources however this solution has to be validated against generated and possible enhancements on that part before +I have no idea why I was assigned this +I will help you test at least the unix scripts once committed +Misty can you make sure that the Clustering Guide is updated with the information Galder mentions here +ok we should probably add a new mojo for copying files that will not be packaged +There are two versions of class +Committed to trunk and. +How much does splitting this apart into its own jar reduce the size of the other jar filesBy about k as thats the size of the new sg file +Do you get a prompt on the screen about a missing class +Thanks Owen! +Make as Incomplete +It is moving in pluginManagenent section not plugins sections and it means it is not going to be applied to every module +unpack archive +Here is a patch for this issue +I havent been able to pinpoint it to one particular thing but I did notice that this issue was happening on one project where people were specifying the items in files for friendly urls +No feedback from user + builds with the code currently at HEAD of mkristian +Heres a fresh patch +Right now is +This will be the starting point for a discussion where attributes should be added and how they should be named +bq +Thanks a lot for the review NicholasThe reason I did it the way I did in the original patch is that the code currently has multiple separate catch blocks and thus some code would need to be repeated in each one +I fixed a bug in the test case +I agreeI am on the a t approach +Thanks +This is quite a limitation but its ASs limitation not BytemansThe default standalone and domain startup scripts add prefix +I have to find the correct version of openejb and patch it and also backport to g +minor comments on phabricator +We have to first fix them if we want to compare performance anyway +Has the listSnap cli call been added? +It appears that the test fails in Suns jdk but passes on Suns jdk +I updated the issue subject to more accurately reflect the underlying issue +The chosen structure is designed to scale up to billions of distinct records + +So it could be used to build a typed json impl +Patch for the problem +Good enough +So it should have a button to create Websites as wellScott had not looked at the code yet Thanks for the directionso the finished page will have both links +I will commit it shortly +Jay can you provide a patch for the branch as well? Thanks +Is there a reason why all of the are not added to srcmainjavaorgapacheratpolicy? +nit its preferred to use with with nit fixed. +He said he had added that at the request of someone +In secured deployment service often needs to authenticate first during its starting +the ping request will time out for up to a few seconds or even longer in case of many firstSearcher eventsHmmm perhaps we could add an option for the new core to have a registered searcher before we swap it in and close the old core +added set +Thanks Amar! +Committed in revision that adds counter and size support +Hi MarkoI think this is a generic problem for all users just looks worse in our environment Im hoping can provide an official pathThanks were using Redhat +The datas been loaded +Manually test added. +I feel we should use retry cache instead of doing a separate retry support for create aloneThis is not what agreed upon at the Summit +fixed the typos reported plus a couple moreplease verify the solution when the changes appear on site +Thanks! Applied patch to resolve defect. +You cant delete an attachment you have to talk to IT +I hoped to fix it with which it does for xsd based docbooks but exposes it to dtd based docbooks +Thanks for reporting and providing the test for the us know please if you face any otherve built and tested the latest code and the bug looks good and dead +Select Package Deploy +Ive implemented a new parser on top of Json parser we already have in Avro C +patch committed to. +I never saw this kind of comment in a real mail +Sample for Repository should facilitate creation of Repository interface and classes as per DDD concepts for domain objects and all the finder methods should be visible as part of that Repository interface +Thanks +Suggestion of fix for this bug +Fix the test failures +Filippo Thank you for your submission +oops assertions should have been left in +Please see my rb comments +This will be staged this afternoon +Cheers +made the batchreadwriter examples return errors +Temporarily disabled failing test from +rebase and add missing classes +Added changes for unit test timeouts to be applied over the initial commit for this issue +The side effect of this is we cant let the resource handler work as expect. +The key condition is to have selected Java Project in Project Explorer +done with in rev +I will double check that were using and that were calling releaseConnection +This is a patch that adds the proposed mechanism +Thanks Otto and Brian and everyone else for this great fix +It might simply happened that the underlying files has been changed and thus the patch is not applying cleanly +ported to in rev assigned the task to AB as a friendly reminder This is valid for Chris as wellGentlemen any chance we could get that out of the way? If you are far too busy I could try and find the time to do it myself +Juraj as I understand it you have tried to run tests on IE and they does proceed smoothly only with registered +Thanks for catching that Myrna +Ok thanks for the additional info looking further I see the DROPINDEX invalidation will be sent and so compiled plans that are dependent on that index will be thats what I was concerned about + the +Ok +np Ill clean up the API +There is already a JIRA for that in relation to use on Linux +Patch there are no objections I would like to close this issue as Invalid. +push out to +Im happy for you to take it up +I cant figure out the content from the patchThe rest is OK might see if you can configure the IDE to not change whitespace and watch out for the copyright headers +Here is the patch to fix the layout product review page +jboss soa p serverdefaultdatah +Proposed solution attached to issue. +Id like to review the whats left outside the routines package first +The filter order should be +It is extremely unlikely that process has a bug that went unnoticed until now. +The changes looks fine +Vincent thanks for having a look and committing this +Jason suggested this is because all CDI beans in the chain are implicitly dependent scoped +Bulk closing stale resolved issues +That should readrunning it starts with about m of memoy consumed +Documented as a Known Issue in the Seam section of the release notes +This task is done and the stuff now has its own issue. +if you in jdk go ahead with a nighly build + +I dont think the application developer cares that the Java class limits have been execeed thats an implementation detail +Hi GuidoI think that I completely understand your use case Im just trying to suggest different solution as it seems that Sqoop is not good fit for you +updated hte PR according to maxs comments +Leaving item open until resolved in classpath +As a guess either Theres a difference between the serial in the compiled class between serialization and deserialization the whole codebase needs to ensure that all serial are manually generated and supplied in classes and doesnt assume that the compiler is generating them at compile Youve serialized something that was generated dynamically which feels more ll see what I can see once I get my current project out of the way also put up the stack trace +Most people familiar with Unix Windows development will have no problems modifying a bash script whereas a Perl script +Reading +If you can provide me with compiled version of changes that you made I will deploy it into my local Artifactory and test it against my application +Updated patch incorporating suggested changes +known issue in our bug db that our agent doesnt work w the ibm jdk on does work with the ibm jdk on aix +See for details +Committed solution for the previous comments but it is still pending test it against weblogic portal using portlet and jsf portlet is a basic incompatibility with weblogic portal and any JSR bridge that falls outside orchestra +Ill ping around to see what happenedFreeman +You can consider this closed. +Resolving for now +think I know the difference between the and my projects config +Thanks Sven your patch is simpler and works in my test cases and my application +This is resolved by the patch supplied for +Geoff was this working in +However now that we have only a single prereq it just seems to cause more confusion relative to any benefits it provides +Please review +as wellI though I would never see this day +Does not break the build if P security is not specified +Ok many thanks +The other two we should address in different jirasOn the changes to recovery to not acquirerelease permits I dont really like the conditional acquirepermit and I was thinking if we could remove it by just acquiring and releasing in different places +Visit httplocalhostThis should result in an exception on the server sideI see there was a similar issue not long ago +This looks like a good one to have +New patch with Amareshwaris change to binhive +Mauro confirmed that the command line switch is still working in the latest snapshot +So it appears as if hbase thinks that the slave cluster intends to replicate as well and doesnt clean the logs +If we go with the proposed approach then we would end up requesting every node on the cluster even though we are really only likely to get a performance speedup if the task gets placed on the node where all the files are togetherA fix that would not have this performance implication would be to create a split with all the nodes that are in the intersection of nodes that blocks in the split reside on +not sure if its possible but can the catalog loading process catch and log at very high level those exceptions and move on to the next item to be loaded +The distcp tests I ran are the followingviewfs hdfshdfs viewfsviewfs viewfs +The new approach just creates a throw away object for each invocation +Definitely mkdirs is an easier solution its being used in other tests with no issues +Comments which have been made about this before Current in Hedwig each topic subscription has its own network channel +Well I made the changeI used parts from the suggested patch especially the unit testsSome concerns regarding the implementation and changes that I did The changes apply for context resolvers message body writers and readers +Coincidently I was running the TCK tests with the current state of the new prototype +This project contains two war with different authenticator methods +Thanks for clarifying Arpit +Yes but why? What is the compelling purpose +Changes were applied at the revision +Maybe the solution is in that code +Hi SlavaCan you elaborate on your use case for this? We were actually considering deprecating and removing support for because in hindsight it didnt seem all that useful +code reviewed +Turns out a Deletes familyMap is encoded differently from a Puts familyMap even though it is the same thingAnyway Ill revert this for now and then we can regroup +with completely new git clone build testPaul could you try to run the step with adding P! or even better with adding X tee log at the end of command line and send me the log pleaseThanks +But Id rather keep this change as simple as possible +Thanks again Kostis applied this patch as r +If there are a lot of unfinished mappers then we should not do aggressive mapperIf reducers have a lot of map outputs they can wait for a longer period time before the map outputs that failed to fetcher previously +In the response from slave seems to have this is not present in master +I only found out about it when googling and finding the changelog +Maybe instead of a heavy validation mode we could provide a message in the log file to report entries in the Config that do not match anything in the +. +SQL script that works on SQL you turn that into DDL template files like the ones you can find in srcmainresourcesorgapachejackrabbitcorepersistencebundle and srcmainresourcesorgapachejackrabbitcorefsdb +I guess I should have kept the two a little better separated but I didnt want to stop once I found the time and got goingPlease advise when you have comments on the patch so that I can either make more updates or check in my changes +I think we are ready now +Im pretty sure that long poll is the standard name for this pattern +Put in SOA +Im still not sold on this these use cases dont work with either right? I think thats part of this not being officially supported before eitherAnd passing the Searcher doesnt quite work right either thats already kind of a bug with explain you get the searcher with the doc not a searcher that covers the whole multisearcher space +If the parent pom itself depends on a snapshot parent you would need to have it available on the local filesystem tooActually what harm would it do to just add the definition for the snapshot repository to every pom as per the attached patch and leave it in for releases? The always reports an error if there is a SNAPSHOT dependency anywhere +I think there is an easy way to fix this wo changing the API +Hi could you tell us in which version this bug is fixed +Required some modifications +I am getting the same problem +committed to TRUNK +Although it turned out that we do not use a core profile of but the full build we have accomplished this task. +I deployed a new snapshot let me know if you run into any other issues +Since the new requirement is actually that user authentication be configured and enabled it would be good to mention it heretadminnetservbasic and tadmincbdjhhfdFor instructions on setting up user authentication this page points the user to Derby Security in the devguide but tadmincbdjhhfd points the user to Working with user authentication in the devguide +I am curious +why? The user only cares about is his revision not themore or less internal counting of plus would be the way to go try r try r try +Merged to +I have added this as admin under scaitests and reorganized the packages and resources a bit. +We should merge both tests +I think it is fixed +We have deprecated the implementation so I suggest we dont fix this issue. +I have decided to close it as not a bug since otherwise I would look stupid. +Lahirus commit r addresses it +Thanks PavelI hope to see beautiful stack dumps tooBut Im afraid this needs DWARFCOFF debug info parsing or at least flow graph and accessibility analysis in unwind heuristics. +The patch addresses the problem and moves the pointcuts binding to a different location which seems to be satisfactory +If in case user did not supply the endpoint then a default endpoint is defined in the will be used +Yes I totally agree we need to update the documentation and make sure no NPE happens +Thanks Tim + +Fixed +Did anybody change something +Editorial pass over all release notes prior to publication of +the patch was applied by Alexey Petrenko on Nov revision +K just read your question hadnt read the patch yet +Now theres only one in that scenario. +Some errors turned up in the tinderbox tests triggered by the checkin of +Marking as resolved. +Something to think about in the error throwing is the background thread +sh +Anyway as Sam notes a database db gives you the createdelete changes for have done such global event on couchzmq on my repo +bq +You got a test case so I can try to debug it + moving the codes +didnt change the Import wizard File import is the same for rasters and vectors supported data are filtered automatically +NOTE Im pretty sure the JVM crash Im seeing is unrelated to this patch it shows up in other places and appears to be a Java problem +think needs to remain final +Closing resolved issues for already released versions +However HOD does generate hdfsnamenode in the in hadoop +Will do it +Any comments on this? Anyone else want to claim it? +That way you would get long stretches of parallel indexing across the clusterI suspect that will eventually get folded into based on the comments in the sourceIll ping Mark and see what he thinks about getting this committed +throwing in another three patches are committed to trunk now. +start a new thread deploy bpel in +I have reviewed and tested this patch with and without UPS integration configuration +Does this happen when using or +At small query all work with ehcache OKOnly at this query NoIn which annotation version this bug is resolved +It can clearly go into but if you want it in let me know +The attached diff contains the fix and the test for the same +As for why its not yet a functionality of the common types Ive been busy and have not gotten to that yet +Please verify +I committed this to trunk and +The extension properties outlined in the description have been created for the appropriate model object types and with the described behavior +Assigned to DOTNETCLIENT that weve started to work on M this patch can be committed +Fixed in the Tools XML Editor +Broken special char on update phase on SP with and mvn tomcatrun +Ivan can you please review this and commit it to trunk and when you get a chanceSure will try to get this done tonight +Ill try out your updates for exporting in a bit +hi erickson +This new compiler keeps track of all dependencies between files and across projects +Moved to because its too late to change the source code and test +Thanks +If there are no more comments Ill commit it later today +Only without the manual synching part +However if that doesnt make sense you probably want to have a class loader in every bundle where there is a factory to be instantiated +Renamed orb module as coreMoved all files from srcmain to srcmainjava and srctest to srctestjava +This will be solved on the branch +bulk move of all unresolved issues from to +The first change is fine saying it was interrupted rather than it was stopped by userThe second change is not so clean says user or system +in rule rrefsqljsqlj +Thanks David for doing this +Unit tests added +What AS version are we talking about? Please show us your configuration +Could you please test with trunk +Its also possible that this issue has already been addressed in subsequent versions +I have that already +You will need to edit your search and modify the user name being usedThis is an issue in the underlying Bugzilla product +hum +We will be releasing for both and by the end of June +Now that work is underway with AS all previous community releases are +part of +I agree we should fix this +Or somewhat more appropriate the link to mailing list becomes part of a new section that highlights all the places to go to get help using couchdb including the documentation the guide irc mailing list is a poor fix +With out the patch accounting opens the old Birt helpAre you OK with accounting nowTom +Pushed a better fix with updated docs. +I have attached patch to address this +can confirm whether this is a regression in M or whether it worked for you in previous found out it is sufficient to have one include in to reproduce the behaviour +Will commit when tests pass +In my development branch I switched compatibility back to with no problems and removed the annotationsThanks for the hintsJens +Updated the with all properties as Ramesh was suggesting. +At least lots of branchy code at the beginning +I upgraded to version as recommended by Jrn Schaible +Version attribute. +Sadly doesnt work for use that expect these in the model in ll try to rework my app differently +given that the code is in how would the Main class create an application context? Are you wanting to movereplicate this namespace handling code in? Or should I try to remove the need tor the namespace handling altogether and just create the manually +lets review this for be put in known limitations as a thing we currently cant disable +Verified +Instead of sconvert use fconvertNext affected tags are convertEntity convertEnum convert convert convert validate is the diff of the file that adds a workaround plugin until we get done +I expect the problem to be outside the plugin anyway +PLEASE CANCEL THIS ISSUE Its a bug in my code ? is a regex character +The two test failures pass locally +Ill add all the changes Ive made as a comment once Ive run through all the tests +And the worst one for searching in history does not work +Not all reporters in the Codeplex issue tracker have accounts in ASF JIRA +Yes it passes here on Linux +nfs files but should pass the tests +To remove unused methods and classes try using a specialized bytecode shrinker like yGuard +This would have to be a cp problem not a HB problem I assume +That typically means that you are sending data at a rate faster than the broker can persist +lgtm +I tried this one times with a local mysql instance to avoid network delays and it didnt occur +have been merged. +Since mFunctions in is a hashmap it does not guarantee any orderSo output of show functions isPlease change mFuntions to a linkedhashmapOtherwise the changes look good +Do you use any special kind of property definitions +Committed trunk r r r branchx we have the jar for too please +I fixed this in SVN +please ensure webResources are documented too +Patches should be applied to each file and then the Java files must be moved into a new package +If there are no benefits to a branch in this regard what is the argument to putting this in trunk for further dev? Might as well just stay in patch form until its ready thenbq +These modifications are also a part of the attached patchHere are some questions thoughts behind this implementation I find that the specs does not say anything specific about composite services MUST having a target defined inline as a reference +Could you please retry with a more recent version of Java? Thanks +I also think that Konstantins proposal is reasonable +Any objections about applying this trivial patch to +The RAID supports mirroring ONLY +I was trying to create an other datasource using the JDBC URL for the drivers +Yeah its for including routes that are defined as templates +in the you can still overwrite these values +This creates a cassandra user and runs cassandra as that userAs far as making it configurable Im not sure theres a precedent in Debian to do that +Would you object to being added as an author for the various classesI will apply this patch tonight unless there are any objections +catch and print out the true error you are getting +is for users to need not run sql scripts when there are changes in role permissions and theres already an existing created another redback issue for this. +It is more important to make the jbpm data available through an easy to use api set then to build a console wich wont satisfy all users +Could we wait for feedback from Przemo re his test using the patched jackrabbitSure +Based on this I think we could just keep going with our own stuff in there as long we keep a record of what we have done in case we want to merge with their trunk + +Especially if it seems to be causing other problems +So this is fixed when running EJB withing jboss +It is a nitpick but one that willapparently take longer to document than to change +Thanks +doesnt this prevent us from having an update site with all the files +properties file in the release in the svn trunki think it was modified by a editor before it was commitedanyway i think we should use utf characters for every non ascii character in the +The new one workes fine for me +I appreciate your effort supporting Java is important but we have to properly understand the work that is needed +Id suggest you make a low priority issue to do it but Id say that it can wait until someone needs it +And having it equal to defeats that goalUnless I hear otherwise Ill commit this in a day or two +These need to be ignored to submit the patch +for the patch +We no longer use logwatcher in gumtree. +Thanks +This patch adds a set of native method stubs for management support in DRLVMAfter patch integration tests pass rate is +The box Automatically detect is checked and yet the text box is empty +an sitting right next to the +JonCan this be closed to and merged into or would you prefer to keep this around +Attached first attempt +What actually requires to be escaped? The shell or the command parser or the http call to oozie server +I merged this to and. +so what alternative is there when a SANshared disk is not available +OK that warning seems good +all issues that have been in resolved state for more than one month without further comments to closed status +Im under the impression that this is actually by design to make things more familiar to a Java programmer I recall it is described this way in book +ales Do you have any timeframe in which you want to push osgi into the master +I agree regarding splitting the project as well +Canceling patch to investigate test failure +Hi Thomas I can see youve cloned this off where David Jencks suggested updating to the Apache pom and using profiles to achieve the source output usingmvn clean deploy Does that not work for JMX? Is the above needed as well instead + +Small test case added to for the issue described. +Since we keep getting questions about this now seems a good time to fix itIm attaching and with modifications to the TIMESTAMP data type topic +Interesting +When I program I would like to know what I am doing and why it is correct at every line +Attachment has been added with description with two pages second page with to swap ipdand has been added with description Testcase result before has been added with description Testcase result after has been added with description An empirical for the bug report YannickIll take ownership of this bug +I will post a better fix tomorrow some time +has a test can find the fix from source tab a pull request would really speed up this back porting or I will try to manage sometime but i cant promise +For example look at the tmp issue I noted earlier +As soon as I change my code to use +bq +Ive seen this idea expressed before on the lists though that only you should revert your own commits and that is of course silly +Usually you use Saxon as this librarySo check what library SMX Karaf uses. +I think this is a good enough reason to drop the from the main search interface actually one must not break the back buttonLets resolve this one before release +this patch introduced a mechanism for RPC server to send status exception class and error message to RPC client during the SASL authentication process +same wsdl with operation name made different with the request elementthis wsdl will generate correct classescorresponding classes are attached in jar generated as correctly complying to the request and response of the wsdl where operation name is not same as the request element name. +Otherwise will implement a growable int array +I quite like having the list segregated by category anyway +And also the inheritance of properties need some more review +Based on the version number in the boot header we can call into an rpc engine that supports the formatSo the single motivating use case for the abstraction would not read the headers +add some href links in the to your content files or remove altogether to get directory listings of your files your choice how was it done before +How do you reproduce when compiling the codeI appropriately enough get BCE Constant value L cannot be converted to a int +Ended up needing to deprecate setIdentitfier resetIdentifier and instantiate on and and adding new versions taking as well as adjusting internal uses of these +Patch to fix this issue +applied +Did not finish creating Japanese properties file +Thanks Bikas +Patch looks good Ill commit shortly +They are removed now +Yes and an index argument for the write method could be helpful + +Yes in order to test this on ER I had to get and for that I had to compile AS from source because the artifact doesnt exist in Repository +Please review this +If you still experience the problem in please reopen the defect with any additional information you can provide. +There is simply no need for it +Log level change No required +Defer all unresolved issues from to +We will take greater care + +Ive just committed this +The error is usually due to the gae maven plugin using a different dependency version than the main DN dependencies +I think resolving the dynamic revisions at first level is a great idea and it might solve other issues as well as I remember seeing references to other problems related to the order of the dependencies in the file +Closing for now. +Wise use an independent classloader to load the generated jaxws classes I suspect this issue is caused by Wises classloader +Fix all findbugs warnings +Committed at revision +So submit your patch +terms like smart auto are very easy to be misinterpreted +Attaching which hopefully commits an instability in which seems to be +Looks good now we will have to be careful about keeping this topic registry cache in sync at all times though + +trying to provide the testcase within the next hours +If these messages are harmful then we need to give a better messageIf these messages arent harmful then we shouldnt be throwing a java exceptionFWIW we see these all the time at +Teiid committed the doc info regarding access +Fixed in SVN trunk +verified that requires compileis in. +Nevermind the previous comment +Beside the features tests are neededThe test could render a line in to a once with a first in RED and a second in GREEN +These are probably all obsolete +we should not delete the VMs created by autoscale policly from CS +It should be pretty straight forward to do thisI am not quite sure with apache infrastructure if they let you run these triggers on their servers etc +Im this because David says its related to the ability to mark a violation wont fix as opposed to marking it false positive +Oooops sorry +richpanelMenu +Uploaded with your suggestion +Please let me know if it passes in your environment too +qpidtrunkqpidjavamanagementtoolsbinset CLASSPATHCLASSPATHQPIDHOMElibThis doesnt seem like a good long term solution for including the jar +Your patch needs the rebase +Made connection meta data connection specific added enum of custom properties supported +I cant reproduce this issueSeems to be an issue with Eclipse WTP Dynamic Web Facet but we need your Eclipse log +Fix if you can +Looks good to me Thanks Charlie for taking the time to walk me through this one +See the Architecture document on the wiki +In general when the rename happens seems we do not need to create reference node if there is no snapshottable directory under the lowest common ancestor of src and dst +BTW I factored out the configuration stuff for framework and left it in the webtools +Committed +And it refers to ajax sort filter groupBy and other Ajax features of look into it +Just had to import the certs for the site deployment and I have just successfully executed a mvn deploy and mvn sitedeploy +Please do not close the issue as I will still be doing additional work to fix the remaining failing tests when a port is indeed specified +Ah I wanted to ask you about that one +See documentation for details. +Nasty +My specific query example might have been bad as it would exceed a MB POST buffer but ignore that and focus on the idea +Committed r. +done svn rev to generating release notes +Minus the ssequpdateseq Id be good applying this as is +And we dont need XSLTC at all +i think there may be ways to do this in the manager without much pain but we may need to wait until +Still can reproduce on s fixed only for on to set Affect Documentation reopen to be able to target correct fix version. +has it been tested and is working well +This is currently done by relativePath is meant only as a hint to the current build +I am not a good person to do this ATM because I am slightly clueless as to how beeline works +Oops +I think we can do this for +exexporthomeuseronepath exporthomeuserone Bengtthanks for the patch and test casesI applied a slightly simpler patchregards a lot for fixing Aki! +Pending jenkins +Im going to file the fix to this problem as a new issue +r to introduce a hierarchy modify connector builds to use it effectively and do some of the test builds +Should at least see if its still an issue in the schemas +Can you add more info here why you think the algorithm is wrong +Fixed adding new properties and. +Meant findbugs warnings +Looks like sonaremma plugin is not compatible with EMMA Can you please make it compatible with EMMA +Create a new Sales Order +Worst case we support and up for starters and I backport all the way to soonNot quite ready for careful review but I think the next patch will be +As for writing new tests with junit a note should be added that describes how to use the test decorator to get the connection for tests +committed to branch on todays JBT trunk JBDS +Builds step on each other and fail due to in user directory +is the consumer started before the producer? If not the topic messages will very quickly get discarded as there is no consumer before they have a chance to expire +The resources directory of any JAR in lib works just like the root directory of the WAR +Committed revision +to so how do people feel about the last version of the patch before Mathiass +All agreed and fixed +Did you see his on more than one occasion Kev? Is it repeatable +Ive emailed the list to see if the current name can be released forA base Jail system is ready +However personally I find the fact that properties of JBI fault messages are lost during traversing the camel component a serious problem +Im not sure how I feel about partial key removal +This is a blocker as it prevents us from having nightly buildsIm investigating +Similarly there is which also extends and adds the functionality that loader can arbitrarily seek near to specified keys +We can get fancier by parsing local config files screening off letter cmds etcLet me know if any of that would be desirable +Looks good overall +Added handling of counters inside. +Please comment if you have any issues with thisI corrected a few minor misspellings and messages where I thought appropriate but I shouldnt have changed any messages + good job +I noticed that Camel offers is this the preferred component to use overCould someone please confirm if this jira is indeed a bug please or am I simply misunderstanding the documentation for the transferExchange optionThanks +The new file is attached +Here is the patch for the same +Fixed of course I was using the Yahoo! libraries ajax code as grails has adaptive ajax tags for different libraries. +I was thinking about this +Workaround use the remote container +Kind of nasty numbering but I suppose it works for now +patch with bugfix and test failure is unrelated +v of the patches added one more missed place in +Sounds good +so something about this repo is messed up +I believe this is WTP +LarsWhy did you ask question you knew the answer +Does not really seem like this could be causing your issue but it might have something to do with it +Pull request for the quick fix +It was a sorting issue in server side +In this case it is wire protocolIncompatible change does not affect any one does not make it compatible change +is what I did in the patch +Implemented andrews suggestion of enahncing the prePut postPut preDelete and postDelete apis to take in the PutDelete object itselfIn the process of running tests +This is fixed in +Your proposed test will pass with out this fix as well since that restarts are from same JVM and we will not exit +Not sure about how jars ended up on ibiblio +Release notes are for issues only so no +Pushed to the trunk +SebastianWould you have a problem if I commented out all JCIP annotations in the source files and provided a script to replace comments with JCIP annotations? would be OK for people maintaining the code but users of the code would no longer see annotations in the Javadoc so I would prefer notI dont understand your issue with the NL files +thanks +As of revision its still a problem +Is this a blocker +Fixed so that format is correctly recognised which was the basic issue +Sorry my fault. +Not only do the headers blink but the scroll bar blinks and the table seems to increase in size by pixels before shrinking back to original size +Good point about making this replicatedpersisted +One of the problems I ran into on local mode was a missing plan +m not sure I understand what youre saying + +If you make another smaller one ill review and commit quicklyThanks for all your work on this +This would be either critical or major depending upon ones interpretation of the severity of the memory leak +Simple patch to ignore TATTEMPTSUCCEEDED TKILL and TATTEMPTCOMMITPENDING at SUCCEEDED and keep the job from abruptly ending in errorIm a bit worried about the bookkeeping wrt +Initial patch skip documents that have any invalid fields improved tests to test it allow suggestionpayload field to accept stringbinary values improved AreekI was asking myself the same question the other dayI felt making payloads binary values was too restrictive +Please attach a small test program and instructions on how to reproduce the problem you reported + +This patch allows the secondary namenode to restart without restarting the primary namenodeIf roll finds that the edits log already exists then it simply returns success +Weld issue +Between the last test report and the new patch very little changed in the patch +should be able to fix this now that is in. +and derbyall ran without failures on Solaris with Sun JDK +It only works that way if true +Reverted in revision Where to go from here +A user may want to output the instrumented class to a different directory +not enough info please reopen if you are still experiencing this +The implementation is changed to a global registry +Bulk close all issues. +HADOOPCLIENTOPTS is added by and then by hadoophdfsmapred again +Hi Luigi please report first any issue on the Sonar user mailing list before creating a JIRA ticket + should not take hit from framework separate thread to handle spill bufferI would think we should have a producer consumer architecture where the consumer should have the logic of spill buffers +Reviewing +This is an error in C as well which is what Thrifts enums are based on +Committing to sameLeaving Jira open for completion of corresponding patch to trunk +ZdenekThanks I applied your patch +I misread the issue +But that results that the example table is getting less readable +Since doesnt have a version attached to it Ill remove this from +Let me know if you want to work on this too and we can find some way to collaborate +PermaineCan you please look into this with MladenBruno +I think youre right a wee bit of AJAX could solve this problem +I did not explicitly try to debug every unittest I changed that was already failing +we dont know what the order is in the select statement we do know the addEntityaddScalar sequence though +We use dynamic parameters all across the language practically anywhere that a can appear +Why have this scheme support not there? And this wouldnt be much work +Yeah I agree with the immutability and shareabilityIll give this a shot with taking the babiest of baby steps +libapacheds folder +The compiler created which I missing and thought that the compilation silently failed + +I tried out your patch but it does not work on +Expect to see this go in sometime in the next day or so +Moving to for now +Incremental patch uploaded +It would be nice if somebody can comment if this affects and as well +I know of no otherpackage that distributes its Jar files and considers them binaries +In fact I found errors first there was a refernece created from a refernece because one closure parameter was stored as field +Users can make use of timestamp range filters to get behavior +Please reopen if the issue is still valid. +I tried to create vault using and use it in with long alias and short alias and both working fine +So shall we leave this as a bug? That ideally wed have some sort of queue on Win apps so that they did run synchronously +Unfortunately my application is not running in a web container so I guess I could not really use the workaround +capacity might diff from different directories mounted on different disks +Technically a JIRA has to be reviewed before the commit could happen +I hope it is be fixed at the earliest +o basically the burden on removing duplicates is passed on to the cThis is correct +the patch applied in revision Alexey please check that it applied. +No the pool is just about not about +Thanks for the explanation +Otherwise why not send mail to the mailing list and explain what you are doing? What is the client? what is the server? How are they configured +We have work item handlers added as dependencies and custom jar files as well in the dependencies +System properties may be placed in the main configuration file +bq +Looks good to me make sure the POM is published to the appropriate Maven repositoriesPatch appliedColm +Only issues that are actually fixed should have a fix release. +Tried with and the exported picture is fineTested both Release x and x exported pictures are like shown in and have been experiencing the same problem +This issue has been resolved for over a year with no further movement +Kan thanks for catching this +This should be pulled into the branch +which version of AS? This logging is done by AS and not the tools so my guess is you are using AS which has a bug with respect to scanning for deployment even when we are in the process of removingundeploying +jmYoniks Law Of Patches readsA patch in Jira with no documentation no testsand no backwards compatibility is better than no patch at all +Does this need another or can i get a commit + +Sounds like v is winning +No problems Roshan picked up the error +Note this issues was to do with however +Is there a timeline where someone will work on or enhancements to use this capabilityI put up some ramblings on +Third line doesnt have any tab +I am talking about adding a method for advanced clients not changing current method +I forgot to say that it is best if this issue is addressed for since I would like to use it for Struts to create a backport assembly that includes the backported jar +bq +Jun before this patch each message will use a different key in so are you suggesting that if message it not specified we should still use a distinct key for each messageCurrently if message is not set it uses a default key range +A couple of tests are fixed +If you agree then what if we simply flush the entry log before we flush the ledger cache in flush? we already write in the opposite order in +Insufficient steps to reproduce no reply from user +All fields should be documented since in certain cases the name might not be available +Rather than delete our doco we hopefully can tweak our build to remove the troublesome classes after the doco is generated +A patch was commit on for with an iPhone able to test this against to see if this is fixed +Patch thanks for the patch and reporting! +In this initial version vectors are used to encapsulate multivalue attributes +Likely fix in rev +It should however be up to the framework to do this optimization +The end of each text chunk should be calculated without including possible spacing values so that the text extraction algorithm is able to detect new wordsThanks for the contribution +is stil broken but Im evensurprised that we have Bizness in itshould we not have only New Smooth Feather there +Chris Were reports on user that this patch fixes the bug for him +Rob please export your classFor process on things like this you should change the file locally in your workspace WHILE you work on your code +This is a simple Wicket page showing the problem +We can fix up the static method vs constructor call in a separate ticket later if needed +Camille Agreed +It did not get added in the Tags patch I think +Could you give some more details +Can you attach a xwf file that may be used for recreation of this +hmm this was fixed for not sure why it is still opened it today +I do have the jmap histlive output from one of the nodemanagers but havent had a chance to go through it yet to see if it helps pinpoint where the leak would be +I dont see a patch anywhere +Ive been using the patch submitted by Jonathan Lee on for replicating a spelling directory in addition to the index and it works very well for that purposeIm attaching a slightly modified patch that allows the snapshooter c option to work with an index or spelling directory that is not named index +But some problems remain +Encountered while testing EAP. +Dependant bug contains additional information and a patch as well +seems that it does not support concurrent access +Hello GuysI am blocked with bug as wellIs there any update progress on this subjectThank you in advance +Fourth patch for the new style to the Requirement forms +nto sure naything has happened with this +Corrected spelling error + implement nextLong NicholasI agree that hours is rather a long interval to set block to +Can any one help on this +Hi Radoslavwe have this scheduled for our next releaseThanks should be releasing theme this weekCheers fixed and verified as of today. +Thanks for the tests and associated Nicholas +Do you plan on starting the implementation soon or would you like me to give it a tryI believe it might take me some time since I cannot commit to it +Added myself to CC +This is done during the mvn site goal and a lot of reports are generated not just javadoc +We are not going to compute metric complexity in classes because in Java there is no way to have a code outside of classes so it will always be equal to overall complexity +I think we can finally close this monster +Unable to replicate. +There is really no difference in how the directories are created +the failing test passes on my machine trying to figure out why fails on Jenkins +I did not test the component in Tomcat +I think i can comment only after seeing a profiled snapshots +You raise a good point +It should be available on once the Jenkins build has completed +I looked at your screenshot and this looks really weird +So every time I get nothing back for a scan I would need to do a second scan with no column filtering to see if the row exist +Let me have a look at your attachements and discuss various questionsissues with peer committers +Sounds good so far +Thanks Suresh +Oh Its an error sorry +If I do anything in that area it will be starting from scratch on top of my own XML code and limited to XML SigIs one possibility just to have a method to call to get the buffer size to pass in based on the key? Two calls but +Sorry Oliver I missed those failures +Merged patch with latest trunkIncorporated Konstantins review comments +Great thanks guys +lgtm +Committed revision . +Yongkun to contributors list and assigned to him +Draft complete ready for review. +At most distractionBut a lot of ZK users will breathe a sigh of relief when this fix gets deployedThanks for your efforts on this +Sorry for the delay +What you are seeking may be solved by have a look +Note that depending on how you log it might appear in a different order in your console of log fileCan you attach a minimal reproducable runnable test case so that i can analyse this +It does beg the question though we sometimes are going to return the last closure result +I have execute the SQL in db returned the same you fix this issue in the next releasethanks. +Does it also happen on XP? If not I cannot diagnose myself +I can assume nightly builds are built with different libraries or build properties from our builds +Couple of bugs verification is blocked due to this +What do you mean by invoke statements work synchronously? We have to keep a thread holding the connection otherwise we lose the transport connection +Thank you Ivan +big step forward +All looks good however it would be nice if there was a way to switch to the Teiid Designer after the project import was done +Added the missing license headers +bq +I should mention that Im vetoing the idea of this just the implementation +Also when submitting patches can you submit them in diff u form this helps us see the changes more easily +Unfortunately before java there were no Atomic class and furthermore the memory model doesnt respect volatile declarationIn order to reduce at minimum the synchronized scope we could synchronize access to your fields with an object but looking at the code there is no benefit to use a lock object instead of synchronized method +Hi ArmyI was successful applying +Dupe of +Assigned to Bram +This will probably fixI hope to be able to test Jenss patch on MondayMain oddity I saw in my quick review of the code was the cleanup of the +Maybe we should just say that it points to a directory that contains a www where the is within the wwwAnd I proposedIde love to see an app folder concept but I suspect best pragmatic approach for now is assume path is the www unless there is a subdirectory called www +The orginal make looks like thisjava i ctestprogramcastor dest ctestprogramcastorcastorSource package generate verbose fail CtestProgramcastor I dont think it should make a me try to reduce it to a simple test case and circle back +My area of code familiarity is not the deploy plugin +Added tests to default build +Please take a look test and let me know if I can close this issue +Updated +so more or less reverse and no Mapreverse ok +Im also on implementing. +Needs to go into M as well +There were some problems related to field initializationsI have updated these points and the attaching the updated patch +Running tests on branch now. +I suspect this is because the executes the query in its open method +First its misspelled +What would be the time gain by using svnexternalsWont it slow the buildsvn update? We only the modeler to be built once a month since we dont do any changes ourselves +Closing all ResolvedFixed issues lacking a Fixed Version +After creating the seed data for Employee Position status it will be needed to check that these status are displayed correctly in the CreateEdit form or not +Definitely a bug +Mike we are currently resorting to custom resolvers for such cases + +Patch split into parts Core HS changes Beelinetests etc Thrift generated offline with Ashutosh and Thejas Going updating the patch to split into separate files for manageable diffs +The complete fix to this bug is part of the cleanup described in jira +Sorry I should have been more clear on my nit with the templates +Attaching a patch with a number of these minor improvementsThe small nstest configuration now runs in about minutes on my slow machineI will run some more experiments and if all OK Ill add a comment about it to the and commit +Fix derived from remarks applies. +Setting to worksforme +The dependency on wadi and spring has been removed from I had to add to to avoid +defer all issues to +This should fix it +see +I think it sorely needs a new pair of eyes +Linking which guts and cleans up the geometry shape classes in Spatial reducing them down to the core of Point Rectangle and +Thanks. +Sounds good David +These attachments are against r. +but it works on my rocket +I tweaked the code to make it send more messages but still no core dumps hereIm on Fedora running client lib and Broker Id recommend submitting an example with a makefile next time makes my life easier +Pradeep Thanks for your review comments +Garbage collector or CleaningLet me know if any of these work for youCheersjust checking in to see if you had a chance to look at the new icons yet and if theyre making thing +Patch with corrected message +have intended to look at this for so thats a good clue to fixing it +I filed to add in task specific counters +bq +which may be deemed a duplicate only if onefix will swat both +One would expect it to use port instead +IMO Validator has too many dependencies and Id like to remove ORO in favor ofJavas standard regex package +Created reviewboard +Jan find out if its Metamer issue +Thanks for Delos Dai for the patch! +bq +HelloI looked into the code You uploaded in the review board and as You asked if I have any suggestions in my humble opinion we could write several classes implementing the rewrite engine interface +Could you look at the checked in test and let me know what Im doing wrong +The changes are restricted to the new HA feature +testcase was not able to reproduce issue with version +I put both the to and the from tools as inner classes of the same outer class +Something bigger than k +Will this provide dbunit functionality like when its merged? If not is there any reason why one would not want dbunit functionality in +I will eventually test this myself but I wont be able to get to it right now +It would just be nice if it fails at some point if a user specified configuration for a table thats not being read +I might leave this open for just a bit longer until I check whether I have the spike around still +Updated file that passes Resins validation +weve been talking about this recently and may have a patch to submit +Should just be a change to the resource properties +Ive made a whole bunch of improvements and bugfixes that ought to be committed before branches. + on the change +After several hours of debugging and trying to narrow this error down I have discovered that this bug report is invalid and was caused by our build system not actually loading SECOAUTH but instead pulling in a version of the + +and committed! Johnny in the future could you please attach patches generated via git ? Thanks! +Daisy Repository Repository Browser with the changes suggested in the last comment +Currently Im using lots of parameters +as zk expects in its connectString +The cleanups you suggested sound good to me +Ive updated the web site and deleted the symlinks. +Removing javadoc warnings +In fact I talked to the Foreground Camera Plugin guys about getting them to contribute their plugin into core +Patch committed thanks Ashutosh! +Large clusters will certainly benefit from running multiple parallel threads since the time to scan full will be in minutes +The general view was all of these checks make sense in +Its fixed +Tomo can you pick this up +Thank you Boris +bq +This patch file is the same as the first one + +I had to go back to there a workaround +How did you generate your IF +It is consistent with last versiosn of hadoop that mv does not create the parent directories +Use of is just an optimization code may need to be split in two +Some of the arguments around I have hard time following may be because I have not seen them +To clarify In YARN we should apply to ALL paths of a new patch to fix the Web Service the original patch on for fixing build failures +Any failure in the past was likely a random port collision that is unavoidable +Therefore we cant generally let Spring proxy the JDO there If possible proxying is recommended but when casting to a class prepareQuery has to be called reason I wanted prepareQuery exposed as a public method is I want to lock multiple rows in a database with +So I did get sometime today to debug thisThey are not tombstones as they are present even for that have never been edited +Its far safer to use so that we can support Portlet Bridges going forward +bug reproduced and fixedthanks for reporting this bug Adam +Thanks guys +I think Manjula did most of this work already and Kathey gave for the remaining adjustments so I marked it Newcomer and known fix although its more of an almost known fix +Hey Ryan I think that fixed this as a side effect of the planner refactoring if you would be up for applying that patch to master and trying it out I would be much obligedAs far as testing it goes it seems like we should be able to take advantage of the object which will indicate how many MR jobs Crunch had to run and can use Counters as a proxy for determining which pipeline tasks were assigned to which MR jobs +Ive added JBPM as another component to the JIRA so you can look at resolving that side of the issue. +Thanks! Updated patch to pretty the output +So it will works this way +Brought into trunk with the merge still need to update the docs. +I am using +Then in lines and we try to get the error messages from the original service +This is a duplicate of. +I see +Fixed with +The code thats deleting the files seems to be incorrectly indented +Yes the first parameter can be null if it is a static method +dproj file +Of course I dont really have a problem with it either +I thought it would be good to see the problems with a stable defintely see the above query getting printed in log more than once but I do not know the frequency +fixed with missing s in s at the end of +I hadnt seen that version was released +Made some changes in the test case +The attached code does not support authentication against a local machine because it always trys to locate a master browser on the network +Thanks for the confirmation. +Those are really just the fields required to create the SASL server or client which is independent of what were actually authenticating +See +The test work fine against +should beThere is no Server Started message shown at the console when the server is started using the production profile in which case this message may be observed in the file located in the serverdefaultlog subdirectory +all steps besides adding accounts to the group should be done now +I suspect that future AS versions will have a start flag on the deploy operation +Profiler without with issue might be related to. +We should add a unit test for this see +I think it should go into trunk for sure +Closing +i think would be a more standardized for the hint +Eclipse project +just committed this to trunk and +Hi SijieThanks for your comment As to As stated in comment Oct it is Ok to introduce breaking change in hedwig per folks in our teamAs to the gap of ledger ranges are Ok as you may observe anytime we delete consumed ledgers we always add in the ledger metadata which carries the in itselfYour patch to fix is due to the usage of only in which lack this kind of informationlet me know if you see any issuesBest Regards +Attaching a revised version of the repro + +Well your changes have not been rejected yetI can appreciate what youre saying about making Ruby and Java meet somewhere in the middle +This looks pretty useful for orchestration of provisioning that Im working on +Weve had a number of reports related to not being able to run an app of one kind or another outside of torquebox +Attached trunk patch +Its worth noting that this patch applies nicely to as well +Im not sure i understand the dimvalue stuff going on inside the single dv fieldwouldnt it be more natural to just use multiple lucene fields +Excellent and good timing this is my priority for the day. +Ive revised to reobtain the underlying Writer from the for each write +Removing the rounding check +A couple of more comments +seems the doesnt fire a when the new key equals the previous key +far as I can remember this was fixed in cvs head just after the alpharelease +Ill fix +Reopened to set the Fix Versions field to JDO beta. +VikramCan you create either a phabricator or review board entry of your patch +Just lazy +And the test now runs as fast for constantsimple language. +But there are some things to iron out +Ive committed this and merged. +Committed to trunk. +is missing a license boilerplate +In current CLDR data SHORT pattern is ddmmyyyy in UK +And by the way combining doGet and doPost into one performexecute does not seem nice simple and to me it appears more like cutting a perfectly functioning leg +Is it not possible for you to transform the null value to another one that would not be meaningful otherwise +you sound like very motivated to me +Our example data +One file is groovy one file is java +More findings fails when run by itself in Eclipse or in Maven +Tested and works fine. +Config file for compile for Apache Flex. +Im working on fixing it +All tests passed +Please use the new patch named It contains implementation of the get and set functions at once +A picture is worth a thousand wordsIndeed! Thanks +sorry no need to +This is a dupe of +The patch also makes Restriction use the helper method in to quote the string and that makes the test case passRunning regression tests now +Ive committed this to trunk and. + +Alex when you get a chance let me know what you think +Ideally we build a bit more around it takes as input the standard format as used by and outputs a file format that can be ready by a new implementation that would read and cache all these results +Thanks Rob for the fix I thought it was very cool! I also added a shutdown hook that flushes the appender on shutdown +Now it makes sense no objection +would need to trick the linker into including it since theres no direct reference to the code in acl so it would normally not get included and therefore wouldnt initialize + +I could have multiple jobs in a pipeline that were all writing to the same output directory for subsequent processing +Yup thanks for clarifying +Thanks TimBTW I notice that newline character of and changed from unix to windows +BrunoOn the eCommerce side I know David Jones had something in mind already +Working on this +I added the profile flag to the build. +Hi Artur Ive it to so far the test was failing on +See attached file +I believe this will be fixed in with the Acegi security changes but only if is retained first in the filter chain +Implemented in and merged into dev branch +The v patch lgtm + tested +Actually after looking over the spec again there may be a spec compliant way to implement this +Thanks +my issue was because of a code issue and I have fixed it +Why was this closed as wont fix? This issue is affecting us and causing all of our builds to fail since we upgraded to Sonar +Awaiting comments +I committed the unit tests demoing the problem +bq +Thats more like it the patch looks good nowTrivial update for better exception messages +I will try get this back into jackrabbit asap +The switch case block shouldnt be any bottleneck since the visitor should not run for each scored doc anyway +IMO we have to have specialimplementations for for output formats which dont use thelayout engine +My other high level comment is +Nice contrib Francis +Can you verify if this is still a problem with the latest codeIf you can reproduce this please attach a test case so I can see whats going on +Closed issues related to Roller release. +Carl got you +Maybe it could be interesting though if you could post your procedure before and after the change as well as the sql statement you were using. +I wonder if this might be an existing bug that was just uncovered by extended testI will try to reproduce this by existing code by adding two separate list fields with differing annotationsIt definitely seems like there is some crossover if signatures of structured fields match +As Camel will use Groovy it makes sense to update components to use this version as wellIm gonna create the Jira corresponding to the Groovy update in each components +The erroreous feature that becomes OFF with this patch wont be supported in first versions of the new ABCD +I really did give this some thing is wed like to start decoupling the core Enunciate engine from the Spring functionality +With the pelt theme the database is hit over and over again for the same information before rendering the page +Teiid projects should use class +Please verify +Does anyone oppose to defaulting the attachWsdl to true +Transitioning to closed. +Thanks for the definitive answer +not blocking that still a problem for me is no one else seeing this +for +Aha nice should we set default buffer size to ? We could also expose the buffer size parameter so a client can change this value if needed +Im unsure of the proposed solution on the mailing listOn a connection event the watch will fire we will skip doing anything but watches are one time events so we will have no watch in place +One possibility how a revamped wizard might work +a suggested applied +Markus took it to the next level by adding more technicality So Highly extensible and scalable web crawler software it is +just missing some documentation added with svn revision the with doesnt work because there is no more available when the renders the message +Patches look good to me I vote we commit them +Thanks for contributing Elijah your patch has been successfully applied see r +To avoid this turn delayedcommits offIn my experience has higher throughput under concurrent load without delayedcommits so they are really only a performance aid in the or use caseJust my opinion but I think this is the correct behavior +Thanks Ted +Added Apache license headers for the source files +Ok the changes were merged to master and +But then again there is a problem with Text nodes +It seems I was mistaken about swapping the lines +This is new behavior in gems +From a tooling sure show the option but from Teiid theres no reason to document it in that manner because it disappears and the user really has to understand what it means for it not to be there which is the default +Please try the new patch +the precommit tests again +bq +Verified with JBT. +Yes thats right +NPE was issue isnt reproducible no more +If it is warn also it might generate lot of log because by default log level is configured +Are you able to build Geronimo from source on your windows machine? Id be curious if you see the same or different behavior +This needs patches from and +Julien these should do it +Thanks Olivier! I can help out as well if you folks are willing to give me an account to log into Sonar and set up the project +In Partners A partner can be created which is a +Screen shot of what the results are spinning Activity Indicator +Ondra are you sure you want this to be AS and AS specific only +sorry should have done that beforeEDIT Confirmed in trunk +Interface is not final just a temp answer to the decoupling +Covered also by and +adding license header to testcaseall checkstyle warnings in the are outside of the diff +Ive attached a new patch which is am not sure what you are trying to add with this patch +Ken Gilmers contributed implementation has been committed. +This looks like the same root cause as the new coordinator taking over installs the old cache view +committed +Updating issue with latest the configuration from Sachin Patel as an example on the plugin site. +See also also The issue of only being able to run tests from the test directory is because the new node is now not going off and searching magically for the composite +Will be available in Click +add corresponding test MaryannIf you think your patch is ready to be tested you should click on submit patch to have Hadoop QA testing it +Not trivial +Thanks Kim! you DagCommitted patch to documentation trunk at revision +Scott if you do have a patch available can you please commit this soon as were nearing the release shortly +Deferring indefinitely whilst we evaluate as alternatives +Thats one reason that Ive been tracking the release +Works well Evgeny but as discussed the ability for a user to log in when the LDAP server is not available should be an option that should not be activated by default +Interesting one +Attached patch includes the introduction of CCNHM as well as related configuration simplifications +Then wed have range of sequenceids per file +I have verified that log files now get garbage collected at slaveMerged to trunk at revision Merged to branch at revision . +abstract element was order to fix css bugs especially in IE we need to separate css rules for single and chunked guide versions +Fixed for dotnet. +Move issues to next release + +If so then she may need to submit the images herself so that we know the licensing is okay +This is for the Android version of Cordova +Fixed with Git HEAD and HEAD +The test should fail with the current code and work with the supplied patch. +Ive just committed a fix for this could you please retry and then close this bug? +This needs to be a documented bug +This is the best option I can think of +Do we really need to code format the generated code +Still needs some config there will be someone complaining that they want the old way. +Anyone with a on Unix would be highly appreciated if you can test it. +I would even sayconsider we should just addshow the egit commit dialog UI to let user choose what gets committedadded +looks a bit hacky doesnt it? Handling this race condition nicer +Yeah it really seems its the same change +Fabrice we have exactly the same limitations in plugins for other languages so in fact I would prefer to postpone this for now and sort out later for all plugins +v + to trunk +negative current design is perfect +We changed that code in both JST and struts +Was this fix committed? I have this problem in an enterprise app for a Fortune companyI tried getting but the issue still isnt fixedThanks for an have just applied the patch in the sources now so if you want to get the fixed version now you should recompile from sourcesThanks for the reminderThanks for the quick update +Just to let you know the agent had not crashed since I upgraded the JRE +And there is no warning to tell the user the truth +Verified the updated patch manually again +The patch seems to be here only +orm files and schema files +We should add jBPM to the tools as well +I also wanted to add my name to the list +Updating the previous patch to fix the rules file so the drools SU works properly add the example back to the features distribution +This was actually on my TODO list +The only deprecated thing remaining is the timestamp stuff in the fileftp component +Is there anyone working on a patch for? Camille Vishal +Finally +file solves the issues found +The later could be done using the node meta data +This actually affects getrangeslices as well querying a key range for startkey and endkey results in an empty bounds +I guess we could add auto auto auto options +Issue closed with release. +But the splitter still uses a bit additional memory due the splitting is based on a copy of the input message for each splitted messageRalf you are welcome to test again and see if you can process a bit more than previously +From this perspective it would be better to have a new module which would introduce a further module however +Added to GNU Classpath . +I brought in line +Marking this as fixed against the RC and upcoming GA release as both releases already fix or will have this issue fixed. +Are filter queries handled with an AST visitor also? It would make sense for this to happen then the same HQL to AST transform can be reused and everything is very orthogonal +Can you rerun the testsI am getting some failures in +This patch explicitly states that crc option is turned off automatically + now explicitly checks for transactions marked for rollback only prior to registering closing stale resolved issues +Reopening +On the WTP side theres simply no API for exclusions and on our code side its possible but would be a very very messy change +ahh thanx +So here is my CLA number + was created for this default it expects the security domain name in the system property +The problem only appeared when all my objects references were null +Just need to finde someone to take a few moments to implement it +only some exceptions are handledAh right +with a minor addition in to enforce when auth is turned off thanks +From my usage I found that I needed the to support the class autodetection property of hibernate +TobiasThis issue is marked as Patch Available yet there is no patch attached +We have equally valid uses of it that do not act as traditional servers nor have stable for which we absolutely need this fix +The Rails app could easily run on the JVM via +Patch applied at +I just configured Hibernate to use as cache Privider +Miguel patch applied at revision r with minor changes thanks a lot for this enhancement I noticed that you implemented three methods of Entity but the JIRA and test only against one so I added some tests as well as merge the that one you provided +that command will be run on hiveserver +Felix I was actually referring to replacing the existing bundle in the repository with an older version +Added some tests which show this is now fixed in idea when this might make it into? I am currently running into this issue +committed to TRUNK and attached patch for MAINTENANCE +Thanks Ashish +I committed the patch to trunk and + +Thanks +Closing as this seems unrelated to the site plugin +Assign to Robin as he is adding progress counter as part of work which can be used to tweak the index flush +Either way you make the call +Fixed in r by making them mode aware. +Patch for trunk attached +Oh no +has been removedTested the change with rev and was able to run the test with the jetty assembly +Looking through who is record it is registered to which doesnt look like is a IT organization +See prior comments for usage +My suggestion is that I just move the initialization onto the workbench startup +deploybut merges the new prototype into +Attaching a patch +CREATE FOREIGN TEMPORARY TABLE was added for +Can you look if this is not already solved in trunkIf you still think theres an issue could you add a unit test that shows the issueThanks you have a Junit test case verifying this? That would make solving this issue rocked fast +Use patch to apply on the directory fromSTRUTSBASEstrutsregards containing all files patched +upping the priority to critical +Ignore the patch Subbu +The is compiled from via plugin in the pom +is it possible to just deactivate the private file storage in OMwe actually dont want to provide a private file storage now +So when the number of datanode is not zero the number of racks should not be zero either +Suggested patches to introduce a parameter forced into the which is configurable through a property +Destruction use case correct separation of concerns has not been established +Setting the Fix Version to +Id like to propose a small change to the name of the test class +Committed to trunk +Can you check that this issue is closed when installing from Market PlaceThanks +Added a patch to fix this issue +But Im tired of this issue and it should do no harm so I committed it. +Resolving as duplicateDick please if you dont think so +Willem You wrote Just a quick note for Jetty upgrading as CXF bundle didnt upgrade to Jetty we still need to import the Jetty bundle in feature +Released. +I tried the following snapshot jar and unfortunately it didnt change things for me +I dont think it is the plugin directory you should point to but your xdoclet can you confirm +I should have a code review shortly +So its safe to mark this issue as fixed +Heres a stab at solving the issueIve tested this locally and it doesnt break anything when space IS available +Committed in revision +A simple testcaseSome code may be shared with +I raised this issue but I didnt get a clear consensus that we should do this since it will require people running Servlet containers to go in and comment them. +A Request type no longer changes when an interesting member on an entity it is watching is pushed in. +I will commit it upon tests pass +I am facing issues in setting the environment to build it with theVishal +Limiting it to the permissions you already have at least means that you wont be able to the revoked permission back to yourself on your own +orError setting property setPrice for object of type com +Paul great job! commit +Thanks for tracking this down Paul +I have now placed the patch with just the changes +It also includes the suggested change +Just a hunch does that happen on processes or also for persistent processes +Committed to and trunk +marking as is done upstream +In a bug was introduced that broke previously working functionality +Ive also incorporated the comments suggested +Appreciate any idea from Dan +Fix committed to X and X +Proposed richtooltip migration to +I always quote the entire URL when using curl +In this case here we index number from and with the randomly set term set to we run into a situation where the indexing code was wrong about the prefix +See I believe this is the correct solution aggregating a common class you create +I think its correct to restore the MR behaviour for this so logging a warning is fine +The sessionMode property of has been deprecated in +The classes and have been moved from to as they are more about errors than about contexts +The source has correct capitalization but the syntax highlighting is losing it in the rendering +Ok so Ill commit this patch in and then well fix things from there on? +It worked for me when I added the built plugin jar of the target folder to the bundle classpath what is not really a solutionMarshall can you give me a hint why the of the runtime plugin does not need to be set to the built plugin jar or how the classloader can find the classes? An empty bundle classpath pointing to the root of the plugin should not be able to resolve the built jar I think +I didnt test previous versions but I think that we can close it +Looks like its worth prototyping should be very straightforwardIf theres no performance advantage though we can leave it is the point of doing that? Making from long is very thing doesnt give more overhead than converting to Long +The modified test passes for me +Document in the Config guide about parameter +Tested! +Committed to by renaming last to last and removing an unused comment +Frank after instantiation I just initialized one of my objects references to be +Actually forcing plugin developers to update their spec is not a big issue because as you mentioned the spec has changed and might change again before +All good but wed not want users to have to java a jar I think +My code coverage project builds just fine +Are we going to support? If so we should not make JDK specific assumption +Thus I cannot see the value in any behaviour other than the original request exactly as it was received +bq + +Im on committing this one +Ok youve discovered a second valid issue +There are manythings that are difficult to capture such as projects that arehighly unstable during development yet stabilize quickly asa release target approaches +Needs as +Integrated to branches and TRUNKThanks for the patches JinchaoThanks for the analysis Ramkrishna +is there a webservice to remotely list deployed webservices or should we start by parsing the html returned by jbosswsservices +Thanks TonyPatch applied to LUNI and NIO modules at repo revision rPlease check that the patch was applied as you expected +That is not really a problem + +I mean no rush here I just wanted to ping so that this issue is not forgotten +Applied the patch inthanks +Yep both is not up to us But due to the nature of brew unlike there might a chance that the xctool will happen shortly after its release +Linking which cleans up the Cartesian Tier code including the +If you have a pressing need to implement this before Ive done that then feel free to apply your patch to your local copy +There is enough standardization to enable the system to detect a server type at least for the cases that use US format dates +ThanksChanges removed added a method which adds the srctestresources folder to the classpath +KarthikI think are in the test package of comm which means these classes dont make it into +Marking for branch and trunk +Items are there. +It sounds like you are getting a request without Host header and it is trying to remap it +If someone thinks that is possible please create a new ticket outlining how the directories need to be structured. +Committed to revision +We do not support nonstandard features of +Marked as a new feature it is different to limit connections per service or per application and the proposed change does not include the current behaviourSo if anyone want to implement this improvement please leave us also the old behaviour +But theyd be able to use the Range class +Such constructions have been explicitely forbiddenThe reason is that there is no way to make sure if the objects in the list are actually compatible with what the generic information providesThe only way to get around such a problem is to register a custom converter to convert to the generic type. +Fixed in Git ID cadcbcaaffaccbe +Actually attach the patch +The server is now running for more than hours with clients searching for random entries +Note the same application works fine on WAS Cluster with the same set of jars +This is the updated oneFor committing this issue we should apply and run +the latest patch that I provided includes the following feature as wellability of assertion expressions to contain a variable value as mentioned in the XML Schema spec +Feel free to reopen if you can still reproduce this on trunk. +check if the validation feature has been enabled on the +Closing this as it is not possible to add array notation for unknown keysMy advice is to adapt your code to use getItemsetItem until a future JS engine which includes ECMA harmony +Wait for release +There will not be a moving to for this wont fix to match Remys resolution +Forgot to grant license to ASF +However since no one can do a full review of this patch Id like another committer to it as well before committing +I think using the variant that Cloudera used in their products is in poor tasteThe hardhatted elephant logo you refer to above was removed from HUEs GUI some time ago as an improper use of an Apache trademark +Final patch +Otherwise the old rules prevailmodulo your suggestion that by default is unset and this prevents all routines from running +cfindex typefile has charset the screenshot to illustrate engoding problems +Currently there is no way to turn off warnings with the Javadoc tool so we willhave to fix each individual file to get rid of the warnings +But it fails on the next class on periodic field initalization that uses similar construct +Im building a unit test to demonstrate the flaw and demonstrate it fixed after the patch is applied +Normal groovyc is not solely for the file usage and uses reflection to inspect the classes +Confirming that failovers could proceed back and forth between the two +can be considered fixed. +Well of course we could fix the example the point of this issue is that something that worked in doesnt work any more in +As stated above the code has changed +I think you are not very clear in your mail +Patch with changes +Am I wrong +The way Ill do it is if the busienss interface has toStringequalshash then go remote otherwise dont +Hmmwed need a dependency from to shared and from both and shared to +We can close this issue when that branch lands in develop +RaghuEdwardDoes this sound reasonable? Does it address your concerns +I think we should take a leaf from Androids book and add a further layer ofThat is we always show errors show warnings and logs unless silent and show verbose only if verbose +My deeper question is why this doesnt work? Is there a real problem naming and defining a maparray without nesting +I like the idea of something like testresources +Reassigning to +Oops sorry. +Assigning Ilene S +Max please review the pull make into M too close +has a library that can do the conversion automatically +Do you see any reason to disable it +If this is the case could you try the latest version of the parser? Some threading problems were fixed for typesIf the new version does work properly could you close this bug? Thx +but I may get it back from you if get to it sooner +VerifiedThanks a lot! +Verified. +No problem. +Review comment by Franz Allan See from the maven dev listAre the ranges inclusive or exclusive? ThanksFranz +If people are okay with this I can just do it before committing +Resolved. +Wouldnt it be better to just ditch the and focus entirely on the XMLRPC library itselfThere are real webservers that are better easy to use dont have a big footprint etc +So I am not sure a unit test case to feed columns out of order is for what purpose? Do you mean we need an assert to guard against this situation? Then it seems s next method already check this +Let me figure it + +bulk defer of open issues to +Hence the main job is to storeload the references and it does not matter if code deals with XML elements or inputoutput streams +trying file trying file +Should be deferred unless this it can be resolved by Augth +I am fine with it for the time being although for completeness it would be nice to have +Thanks! +This is the patch file to fix and Please ignore the previous +Its public +Off hand I cant think of aI shall commit this shortly +Ill update the patch to port the same behavior to +Im using this something that you would be able to make into a small unit tests? +It does log host name and port of the peer endpoint attempts to authenticate against but I do not think this represents a security issue in any form or fashion +Thanks a lotBest regards by Richard +Is any activity on it on +is removed for along with so making this a issue +The checked exceptions only appear when you loadsave the configuration which is quite sensible. +With Kryo it failed early with exit code +on. +Just committed this to trunk and + +Therefore the issue will be closed +In the light of this I close this issue as Wont Fix +Heres a test case which shows the issue +Is this also going to be developed for trunk? I am for committing this to a branch until it is also committed to trunk +Very nice ScottI will handle the backport to Thanks again +At least the file srcxdocs from the branchfcs should be published on the web site +LGTM everybody for helping to get this in. +you can check it now +By definition only one can be authoritative though of course both may be officially sanctioned by the ARQ team +The serialization of uses int + +I think this was fixed by applying Erics patch earlier today +Given the gains for I think this is an OK tradeoff +yay! +Hi Kelvingood you pointed the validation topic out it is a very important area long time overween +Whats the motivation for throwing instead of returning null? Can we instead consider widening the defined scope of and use that? This doesnt seem like a real IO problem to me +This raises the issue of when is a build successful +sorry this got lost in the shuffle +Tests passed +I would like input on whether I should go ahead and check this in and disable the failing tests until is fixed +A consistent analysis skip property should be created +This is my first review so let me know if there are any problems with the patch format etc +attached a patch for +Yeah +Thanks + Ive applied modified patchCheck that it is OK +The bean is now deployed in deploy regardless of the profile service implementation +Ill merge later today when I have time same with +Patch for for to Abhijit S Ghosh for his patch +I feel it is better cached in +Fixed for Apache. +passes +Your intended refactorings sound reasonable +If it shows up again in nightlys after this fix just reopen the bug +Looks ok now +Broken patch that adds boolean type +I plan to later but for now I haven did add a personal event however +Babak do you want to try thisSee course. +Minimal example demo application that fails if you browse it with the cookies patch +This does not appear to be related to subsystems +Perhaps we should make int default to varint and add int and int types for when you really want fixed width? Defaulting to fixed seems archaic +Checked with JBDS +I guess that if your fix really fixes it it should be going into +jcrmixinTypesDelete removes all mixins from test +FOP using distributed with is indeed unable to grok encoded XML +javadoc fix +Not sure what an appropriate solution would be at this point though +Thanks Auke for the contribution. +No +Good point missing properties in the JNDI context shouldnt be logged aserrors but it is desirable to log the other errors +The problem only occurred in one projectjob and not in others so I thought that it could be something corrupted in this have deletedrecreated the job and let you if during the day it happens again +Im about to push this to unless someone speaks t be working on these and theyre old +has been changed in revision to use Flumes +Attached the output of update +fixed in revision to the branch in revision . +Just a warning enabling code formatting could make some existing issues difficult to read when some plain text is suddenly interpreted as formattingSorry dont have any examples to hand but I know there were some Commons issues I had to edit to make readable as the poster had not realised which characters were special +v +This is a negative test case +jar youre using +Many thanks to Raghu for his time on reviewing my patch line by line to make sure that the patch works correctly +thanks fixedthe fix is at bit more tricky since its necessary to invoke gettype to check if its a custom type or not +Dont worry about the branches +Nice fix Justin! +I can revert to not a problem +Which productversion do you use +The changes was in UNICAST +Could some one please review the patch before going into stale +Please review. +Ad +Isnt that the pointSent from my iPhone +Thanks Stack for trying it out +However I have a different problem with the imports because uima runtime stuff cannot be located +Fix is committed and will be available for the release +It is probably due to the problem you were experiencing before with the localization bug +Support for all CCITT compression types added in commit resolving fixed. + +See if you can get the XML parsing with schema validation to work just using the JAXP +The is present in the parent project +Yang does not yet have the fixes +I got to be quite annoyed how jboss init script worked so I modified it to behave nicely on our environment +What do other projects generally do +I guess something similar will be needed for the WCS capabilities and for the Describe responses that allow the description of all the layers published by the server WFS and WCS have those though for the WFS I believe we need something different Justin +who knows what folders are to be excluded from search operationsIf Id know where to get the folders to exclude from search Ill do this. +Still need better instructions but no need for +Removing the from lib also comes into scope of this JIRANote that this applies ONLY to the branch since I expect to have the license issue resolved before we release +The code tries to put directly in hdfs +Going to commit a bit later +This test passes persistently. +Regression test and the patch has been aplied with the modified testPlaese verify +Ive had to manually put together a version of the plugin to allow it to compile our projects +Patch adding close calls without suppressing s +But I may very well be wrong on this +I dont see why we need this variable to compile code that requires Java classes and methods +This constraint has been removed in release +Thanks Pavol the patch has been merged into develop. +. +Try fixing when we do this one too +This minimally invasive patch merely swaps out Charmonizers for the newclass LucyTest +Igor Is this going to go into ? If not move out to +Good catch +This was why I added the word transitively +Maybe do another hadoopqa given the test fails above +Now that vectorization changes are merged into trunk I think we should add of to the jiras +This was implemented in a different service +forgot to mention that Im getting is in +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Attached patch against svn trunk +it looks to me like we will have to move the JARS to support an Ant build into a directory separate from the dependency +just committed this. +The ldap tests failed with because its using a different initial context factory +fixes the older broker issue patches applied to the branch +Is there anything specific youre thinking of +Will be adding more support for M +Its been released. +Can you try and simply the issue down a bit? At the very least I will require an example app and instructions on how to deploy on a single machine +After this he has persistent classes completions in addition +after discussion with Tomas I will remove the whole section Manual Install as that is not supported due not distributing spring framework jars in module and in parallel we have Installer which does everything +Attached simplified test for single cpus +This is pretty serious I fear well have to redo the and gt releases because of this one +I applied the amendment +The goal of the plugin form is to show mapping rules applied +Hard to fix correctly without ability to insert implicit casts at execution time +Please check that mkdistro and starting server puts jar in classpath as and no test cases as its a script changeCommitted to trunk and +already taken care of by seelmanns hooked us up with parsers that take into account extensions. +Verified passedThanks Alexey +vpeplugins +Committed to trunk +removing extra release the site docs for quickstart point to the quickstart which has this information already. +Closed. +Christian Ok either somethings wrong in your environment or theres something terribly wrong elsewhere +So +I will change create and remove temp file code to do it via command that can run locally or remotely depending on the mode +Going to do some more testing and get this fixed in for you +It doesnt look relevant for me +I think BytesInts should have parallel apis and someone can just call unicodeutil in general these are reference classes not stringbuffers and we shouldnt encourage abuse via sugar apis +This is something that should have worked but a couple of minor gaps were present +In the example samsun is incorrectly spelled has freqInfo and zero suggestions so its never set to true +Here is the path of strong references that are keeping the session from being confirm that the problem exists we are using connector too +bq +Have you checked whether it is compatible with that +Need to check +This patch introduces a force option to the saveNamespace command +GA isnt released yet is it +I was using hadoopis the patch for inif so the problem is not solvedThe criteria for starting a speculative execution should also include how muchlittle progress has been made in the past M minutesIf a task is detected to be stalled a speculative execution should be started no matter how close the original one is to completion +Already done that though +Added patch Test that reproduces the issue +For some reason which I cant pinpoint at the moment FUSEKICONF value is always showing up empty +Link to that caused the regression in the first place +I am attaching a test case showing the intended use of HeadergetElements method +It will go into the next plugin release as soon as you sort out all the embedder stuff +We have some tickets on the roadmap for to make it less spring dependent +Ok thanksI updated the release notesI will create the patch after I merged your changes in my local working copy +added some varients of the chunkier text with a green blue or grey Apache feather with the Apache this JIRA its well over years old and the logo in use has been about for just over . +And deployed new bundleSo I mark this resolved +Thanks Micah +Verified on EAP for release as resolved in typo in release removed unnecessary details from release note. +applied patch from to the Seam issues group that deals with the patches for this issue +i think these are the same problem + Samuel for the hard work resolving these duplicate classes. + +Ill look into it +If we went this route itd be a znode per region on the cluster +perflibmanagedcompile +Add the interface for use by connectors +Its sufficient to use a node ccm cluster to test this +These issues have all been resolved and just need to be closed. +Nathan this issue doesnt reproduce +It just says out of date but in what version is it fixed +The classes involved in parsing application policies have been updated to allow the specification of acl configurations and tests have been expanded to guarantee that the expected objects are created and set into the +Closed for +good point but Im not sure if you can automatically accept a license for a project shouldnt the user have to accept itI was just going to start with the simplest scenario +Im not Pete but I agree +uses instead of in probably unnecessary due to the instance being being already shared and protected by a split the patch in two as I believe one deserves inclusion while the other may be irreleveant +I kicked the build +Can you attach the test output +Ill keep working on restructuring the now that we know the fix for works +Ive checked Linuxx failure in jdttext suite showed by CC does not appear for meSo only unexpected issues remain on both Windows and Linux x and they are caused by Eclipse bug Harmony bug is now fixed by +is related to +No I just know Hbase provide API for Thrift but dont know how to use it +So if the node is not present then we have no problems we can just log itIf we are not able to delete the node then? so subsequent disable may cause issue right +Brad do you want to fix this one +Nice work Ram +I thought we could workaround it at least by hands but now I see we cannot +Let me clarify on the use of repetitive parameters will get back to you +Instead of throwing just render the error message directly to the screen +Based on test reults with the it was determined that the LRU setting at the default value provides adequate performance and that changing this for all installations is inadvisable +Verified in to add release note information +There is also an OGM pull request on the subjectBut such change would be of relatively low value compared to feature focused and value add to the user so Im not sure we should prioritize this work today. +The merge to was clean and my testing was successful +SVN was also setup previously +Could you try this one out and let me know if this works +Created for implementation of sequential znode based readwrite locks +Should be fixed by the JBOSSTSSPCP release +we dont really promise extendability promises +The Change timestamp button becomes active as soon as the current selection is from a WTP sure what you are missing +Ill have a deeper look when I get back from holiday +In this case I think it can because it would not change any of the released artifacts its a simple copy of the one file to another one wo the duplicate +Here is a patch then makes the datanode flush and sync all data and metadata of a block file to disk when the block is closed +on patchLastly is this something we should add to +Confirmed fixed by inspection of the page in the ref manual for +Bulk update to fix version +forgot to close +Note that the checkin performance boost will likely be even more impressive with bundle persistence +Looks good to me. +jsp to fail +Committed +Theres no real reason why it needs to take annotations as the arguments +account created youll need an ssh key Ive used your authorizedkeys from +The dictum that deletes only apply to the next file is broke +Thanks. +Many of the steps and decisions are using multiple next have attached the job definition to this issue +Oops wrong plugin should be announcement +v with the correct name +I believe we should target Doxia at have updated the and branches to Doxia why cant we use Doxia in Maven +The amount of hardware required for an entire copy of the index costs too much to have in a lab environment +There have been a number of significant memory improvements since suggested by Jason we tried this on the latest nightly build and it looks like the leak is now gone. +You can test now Maurice +Hey ThiruAwesome thanks for the great work +Bumping up to Blocker per discussion with Ken J +Silly me did a last minute cleanup that doesnt compile +For example to package an RMI client JAR I need to compile all classes in order to run the RMIC to generate client Stubs +I realized the just needed to be type of false the to If running from source youll need to switch this variable to point to +Does the CLI copy the icons to the right spots or is this something that the CLI still has to implement +I have a question on the patchThe main effect of the patch is to bypass thebaseclass call to initializethe question I have is when does that get calledIf it isnt ever called then Animated attributes wont workOn that point Cameron should the initialization of the live attributes be tied to the booting of the SVGand CSS DOM +All issues on this jira have been addressed +I guess I forgot about the request handler defaults when I was creating this patchI have removed the possibility to configure partial results via the property in the latest patch +It would probably be better to provide a consistent default and maybe provide a switch +Committed at subversion revision +trunk +I am surprised to see the amount of overhead with more threads +Reopening the issue since the rwtest library fails to compile with the errors from the issue description +has been passing consistently for me + +Will reopen momentarily to ask for a on the master branch +Patch to add puppet folders containing my manfiests +However there will never be a official release from the HC code line +Applied Thanks for the patch! +This can certainly be done in many cases and is an aid to user friendliness +I remember similar issue +If there is a particular functional change or design decision which youd like to draw my attention to please let me know +on closing too +Link to the issue for upgrading on ve flipped it to for now. +It also has a subset of the changes to make future changes compatible +Log attached +Due to the incompleteness of implementation its unit tests require much work +The patch for was applied at check that the patch was applied as you expected +Applying this patch is better than not applying it + +Someone recently posted some better code this is where it would goAside from the dummy algorithm does the whole thing make sense? If it works would we add it to the canopy jobs +it would need to be changed from using recursion to iteration instead +spoke with jbellis offline and I see the argument for prioritizing the bin file +META +Also there does not appear to be any escaping when you are generating the JSON array +This patch had to change almost all tests Ill prefer making more test changes in another patch +Marking as resolved as this is fixed in. +HBCK now ignores valid splits parents for region boundary calculations +We can set DATATRANSFERVERSION to in and to in +Also fixed the initial warnings in just by removing the unnecessary accumulate call +This patch fixes theA similar patch fixed the issue I found in FlumeIm running on a mac and the tests seem to not terminate this seems to be some issue in the minicluster code and not a compatibility issue +Were using a different convention indeed +We can resolve this. +Well Im not sure +For example if the form has these valuesQ +punt to b +These become fallbacks if round robin is enabled +The incompatibility only crept in I believe because there are no unit tests for enough variations of +All set. +I would say that data this dirty should be cleaned before it is processed +was done months ago in new branch to upstream +Its used for UIDVALIDATY too +I see the same behaviour in both and +All Designer spinner issues will be fixed by +Looks to me like the response gzip that comes in is broken java fails to read the gzip stream +I am testing patch v +I fixed one of the array out of bounds problems + varcachehttpddoes it exist? Is it writable by the httpd user? Any or UAC security like on the box +regression test for the regression test for this issue should be run in OPT mode +One change Id like to see if we go this way is that should be in since it is specific to that not to RPC + create jboss https and exclude there know failures look at the and see what tests can be fixed +Any news on this +Iv just committed thisSorry for my late commit +There are some TODO items marked as TODO in the code +CRC checksum algorithm is needed to decrypt ticket part of first +on one interface only +Whatever we decide about this issue it can wait after think enough of the integration tests use the Wink client now. +united benchmark. +This issue will be resolved in HCI believe the issue has been resolved by the latest changes in the SVN trunk +Attached patch fixes JDWP agent to filter out unprepared classes +Ill rather ping the list for opinions about it. +If its good should know tonight Ill apply it +Logged In YES useridThe zip package has been updated with the correct contents +Ooops looks like I messed with the formatting +I bought a nice tool that cut it down to a few hours work +Its been a long timeWe really want to have that capability +Yes I reviewed your implIt works per spec AND x faster to boot +Didnt realize this patch changed the way allocs would get read or that it was doing more than just changing default AM memory requirements +Hmm maybe that comment seemed a bit dismissive of the making an online Sothebys is probably more fun +It doesnt seem to add much value though client libraries would need to keep track of the current keyspace per connection either way +So this will now be userid +Deferred to defer to defer to +Committed to trunk and I updated the release notes thanks Chris! +This is a patch which works around the immediate problem by returning instead of for the edge case where reads a fragment of a MIME boundary +wiki diff notifications changed forjakarta +Fixed by Git commit closing stale resolved issues +Im wondering if this issue will get fixed? Unfortunately I ran into the same problem and the issue is not really new. +Thanks Uma! +using version just committed this thanks Terence. +This issue is fixed and released as part of release +Or an extension to the deprecation mechanism can also be looked at as a further enhancement +Perhaps the should redirect the user BACK to facebook and ask for authentication again +Strange thing is that its working when deployed on GAE but not when run in local dev +Thanks Jing +Attached patch has a modified test that catches the problem and the fix for the problemIm sorry I couldnt resolve the issue earlier +Actually this was the way it was intended except for the minutes part that is +Secondary index stops working on with cachingALL +Just a bit safer I guess since they should be protected +That would be a new bug I think +Why not keep read statistics and sum them up periodically? That seems better than disabling this entirely +It can then be annotated as agree with KristianCould somebody from the developers state their view +Patch v compilesI reformatted some of the new code +Im asserting that this is due to the combination of s class loader thread instance generation paradigm in combination with the new class loader in Java +Hadoop QA is trying to apply it to trunk +Patch applied at r thanks for your contribution please verify +no response from submitter. +A work +Instead it is just returning immediately that the socket isnt writableEverything after the first is just showing a workaround to coerce jruby into thinking it is connected and should not be necessary and is not necessary under MRI +Do you think it would be useful to add this timeout parameter in the sonar UIIn every instance I add it to the not +I suppose I should change the label to something more suitable such as fetch from xpath descriptor? The basis gist is that there are xpaths saved in a list and the user has the ability to either a choose an appropriate xpath b create a new xpath elsewhere and choose that one if their configuration uses weird jboss options or c just use the value +check for txid highest txid not just equal handle thrown from readOp +Attaching the and are using Drools +I didnt know what the best return value would be for these methods in the case of a missing library so I just returned As of right now with this patch Im able to successfully run the AWT and Swing tests in the classlib over the IBMVM on Windows which is very nice and gets us one step closer +The fix for allows the client to browse and delete scheduled by +Worked fine for me testing on a Mac + +Hi DagYes this bug should be fixed by at subversion revision +If you find further issues please create a new jira and link it to this one. +I tested this and could not reproduce the problem +bq +The unit test performed very well and seemed solid +Please review +Here is something interesting +Typically these applications want the system to run at very high thruput and incase of failures are ok to lose or replay small number of events +Opening separate issue for build but want to close this to get in release notes that it is mostly working now. +Thanks Arnaud then! +IMHO it sounds like more a workaround than a long term I could create a Maven project for doc and apply Erels suggestions +is not whats wanted at all +Verified in DR +So is not found on the classpath +In my opinion we could evaluate the inital counters for each bind when building the SQL statement as these initial indizes will never change for an instance +But this is what I meant to say about the result sets I think there actually is a problem with not sending CLSQRY for finalize +Committed +Being able to change tabs is critical to base functionlaity +OK I replaced catching with in and +messaggi non tradotti +Thanks to Uma Maheswara! +should probably also take a Progressable so this doesnt time out +bulk defer to defer to. +How would people use the smoke test jar +Adding extra DEBUG logs +This is a fairly big deal if people are expecting to leave Fiber instances lying around but if they are running them to completion they should be ok +LikithaIgnore my comment above +looks good +Fix issue with synthetic pk attribute deprectate obj attribute constructor with type parameter replace its on other constructor +rev makes axis deployer play fair +First leave the API as it is and extend a Properties object to overwriting its methods by using behind the sence a +Ive created the Eclipse update site +We are not allowing rsyncs anymore. +Attaching +This is essentially the same as and +After attaching the test case I tried running the test with the bit JVM and the files were all the same so I ran the test again four more times and the files were still all the same +I am open to this patch and making it a default if we get more input from others that this should be the default +henry check out on some feedback from trygve I moved that base class to the can be replaced with from the latest snapshots of that dependency a patch for that is on as well +The first issue is just an UI issue of the aspectj plugin which has issues with introduced for ll look into the second one after we published the milestone the second issueJust tried it you have to create two user accounts +Now it is impossible since it is a compilation error +And again if it works why didnt you ask for release in the Maven dev list +Attaching latest patch with incremental changes from previous patch incorporating Viveks offline comments The cleaning up of the initialized jobs list in poller is done at the top level so code can look more clean +Pavol I believe this issue has been resolved right? +The most elegant way is to add a synchronizer which pull the groups from the distributed map at startup +Doh sorryDI mustve messed something up in the git +Correction got my notes wrong +Having it this way means users can choose whether they want to add a single instance for both stored and indexed or two instances +Tested the patch on a cluster with unhealthy nodes and saw that it returned correct list of nodes +can you post full log and example of urls? I will check later in the coming day +Attaching which is a patch for the changes we made +Session will be cleaned up automatically after user logout +Unfortunately there are already include and exclude options that are interpreted as regex right now +GWT includes those two classes which makes their pages render differently if the request is from a desktop browser or a mobile browser +It seems by extension this would be a problem for other dependencies that use classifiers +Seems to working in Pivot Windows JDK m also unable to reproduce this issue on Mac OS X or Windows XP. +Thanks +I will look into this task we might have to create a new plugin and new model to support bpel versioning without break original Eclipse BPEL this is an Eclipse project without any version modifications +Would be perfect if the CTP is removed from the help screen too + +The mirror is currently slightly more than hours old +Fix merged from branch into trunk under revision r. +The structure of the subdirectories will be created under the destination directory and the newly transformed output files will be placed at the appropriate levels +Test failure is because of +i think we could allow new clients to talk to old servers by letting the server tell the client what version it is +Adding and will resubmit +The same problem appears with and for the same reasons +with junit +Patch committed to branch and trunk +What is the reason for thisI am in the process of testing this now +Fixed in WTP on JBDS +It is really a fatal error though not a regular compiler error so the cause of the error will be cryptic +This has been fixed on EAP as Clebertcould you help to verify this issue pleaes? From description is not clear how this problem could be commit eeddddddc in ASFinal +Best wishes. +HelloWould you please try my patchThanksBest regards you AndrewYou patch was applied to the NIO module at repo revision rPlease will you check that it was applied as you expected +See and +New patch addressing some nocommitsNow you instantiate calls +Yes we said we would need to keep Forge if that was necessary for compatibility with the things that havent moved yet switchyard was just an example and used by upcoming product hence why I was wondering how we couldwould handle Cmd and Cmd +backported fix from trunk to minor conflict resolution required in mergesjdksvn commitSending javaengineorgapachederbyimplstoreaccessbtreeSending javaengineorgapachederbyimplstoreaccessheapSending javaengineorgapachederbyimplstoreaccessheapTransmitting file dataCommitted revision +OK great thanks +shifting priorities to processing the heartbeatsI agree with you on this one +Ideally just get the value of the field if following ther is no difference between and this + +Will update patch shortly +We are only using the lib +I use plain Hibernate annotations there +This has been superseeded by the latest patch to ring javawsdl up to Axis RC in +Thanks Kang Xiao +I said above that the earlier patches were lost but nope there they are +Not sure how to turn source code in description into a patch file but the fixed extractText method was included earlier. +Should this be exposed as a JMX setter? Shell seems a tad too much? + Username letiemble +fixed with rev and rev still want happy with the initial index state and added a few more tricks in rev Basically now we have the indexUpdate which contains the last index update timestamp +well these two jars are hibernate dependency and also seam needs them because hibernate needs asm has been add to commonlib in branch confirm that the and are in commonlib for CR. +Closing. +I can see vpopmail claiming that the last auth time for your account was April +msec buffer +I used numberguess instead of drools but followed the same steps above +this would be highly useful. +No movement in over months +jar in my application only to send very simple email having the basics available in would be great +org somewhere for launch +Resolved some conflicts with the current trunk +Thanks Hitesh. +Oops thanks for spotting that RegisThe JAR file has been committed at rPlease check again +This was on master and fixed on master +Here is a fixed version should this go in for +Looks fine to me +Yeah whoops! Ive digged those out needless to sayIgor can you test this out and get back to me +closing issues associated with existing releases +There already is an admin group sorry +MD sumsbfeddebd +documents have been updated +I will reduce the prioritybut I will leave the ticket and let someone there decide onwhether its worth following up or closing +However in this case I dont believe speculative retry can account for all the difference + contains a patch that fixes this +S would be m not sure that S could be used for a write ahead log +Integrated to trunkThanks for the patch Gregory +Dag I will try the patch for the OFFSET clause query and see how that goes +So all issues not directly assigned to an AS release are being closed + can now specify in the UI if a backup should be created +I was just thinking the common way Alasdair described earlier on +Hi Donaldthanks for the support your mirror willappear in the list in an hour or soRegards icons have been corrected +Scott Can you please add more details to your suggestion +I just committed this to and trunk +Agree that default value could be flipped +For that i have open other jira taskThanks at the latest patches Im not sure what the problem is +If I have any thought on the optimization framework I will comment under +Gavin had already done this. +Thanks Dag for that information! Im attaching and which I hope make appropriate changes +Table +that revision actually modified the simple webdav bundle forThe change in revision was for the davex bundle. +Use Scotts new account +Maybe you need a redirect to a URL instead of to a view id +We should ship the correctcurrent EULA file in CP +Resolving due to lack of demand and the changing landscape caused by Leveled compaction. +I see a commit in the all tab and is resolved +Apparently they have changed the way they use the so i have gotten rid of using the directly used the insteadThanks this patch in revision +BULK EDIT These issues are open to be picked up +Hi KenI agree with your suggestion good ideaPavel +Attaching patch fixing problem of not setting supplementary process group of the process +This broke in r meaning it was broken with the in in r thanks for the report +fix in rev to be checked out in the the server does an svn up so its please sign these artifacts +Test Plan +seems it can be done through or posing the jar in apps no +Funny to see that doesnt have the statements in the MANIFEST but ship a blueprint descriptor +Or dl them +null can never be deepCopied +Does this sound about right +As per discussion on dev and comment from the reporter resolving this issue. + +Could you provide more details? Thanks +is this done or +Here are the updated patches +are you still working on a patch +Funny how a failing build once again makes us happy +Some notes Could you attach so I can see what it looks like? Are you bumping hadoop version on purpose as part of this ticket? Wouldnt a find be more appropriate rather than this forcontinue loop? In a spec file Requires statements can be split on multiple lines +This now seems to work again +The locale configuration option seems to be working correctly. +It works with orI am not really sure it is the good way where to find the good encoding? does it matterEric I really need your opinion on that in revision r +So was not about lost files in the end +Otherwise we could just wheel out PC and go home +I have committed the and patches +Closing old issues +Its already pretty straightforward to extract Records in the collector. +Ive attached the screenshots as a zipfile because the Attach Screenshot functionalityShows the VPE preview for the file +Few tests had failed due to some issues on my machine but on they passed + +Shouldnt be failing that many tests with this patch +This is actually regression caused by rev +This class encapsulates all history related code like whether it is running in local mode or standalone mode and let jobtracker just instantiate itAlso jsps could be made agnostic that those are running in jobtrackers jvm or standalone +Do we need a separate HICC user guide +This issue came up from two different people simultaneously on IRC last week so we should definitely get something into +By the way you can use the latest JDBC drivers for g they also work for i and i as far as I also usually perform much better. +as per all encodings must result in a stream noticed some problems with trailing whitespace and the warning about parsing without encoding feature which have now been fixed in my branch. +Just a quick note Ive been using the update this weekend for some moderate griffon hacking and I didnt see the problem and I havent had any logs reported in the error log +Please let me know if i am missing out do i add apis to check the conformance +So if you provide me some guidelines about that I can build a regression that will help to check listeners behavior between releases +Moved to HDFS project since this is a libhdfs build for the updated patch +Here is a patch that adds support for the sort parameters in +Committed as discussed above +Fixed an NPE bug in where an empty throws NPE when you try to write it +Would you please help to verify whether its still a problem? Thanks a lotBest regards allBoth the tests above have passed on my laptop +It extends +I am concerned that merging another branch could introduce bugs if is being used per the older convention +Fixed in revision on the interop branch +This should be applied to the module as it needs the JDBC is my current log outputtrunk version as of minutes ago At revision o clean test Dtest still get the NPE? +we need at least set of data to test this kind of thing to find out if there is a problem with one vendor or if its related to your particular environment. +If its greater than whatever set is then pop up a alert warning the user +It looks like this can be caused if the first batch to the client contains only a single row and the next batch has the wrong final row +I might be missing the point but what do the patch changes in achieve? They dont seem related to this issue. +I like the embed install button trick +patch to add to of api build +Quick bench with new patchTPT ledgers opssecBookie Latency Bookie TPT +Thanks Kojir +Thanks +Will commit if tests pass +Targeting test issues to test integration framework supports the bulk testing of transactions by defining the logical set of DML statements to execute as a transaction set +If all you are doing is fixing typos then one diff file for the whole manual is OK by me +bulk close of all resolved issues. +The fix in aries works and the NPE does not happen is a patch to switch the branch to Blueprint and utils and proxyIt builds to the point of the iTests +Fix committed +The workaround does the trick so Im closing this and leaving the linked issue for Remy to comment on. +Reopening to add extra information just made available +Fixed in rev +Fix and test patches are attached +Committed revision +If you have subpackages in a different bundle there is no problem +Ryan any hints on how to reproduce and cause this +Included in the Hibernate Known Issues of the Release Notes +New patch with unnecessary debug logs removed +I see you point now +details about why F in the path part is a problem with Tomcat is detailed in. +Committed +It would be great to have a test case to demonstrate the reported issue +Thanks Mark +Sadly it is not an file that is missing +Gary so unless you come up with an alternative I suspect we may have to say no on this occasion +The previous one was using get for em in when in fact it should get a new EM by calling suggested changed the method name from get to get and added a if the test didnt close the EM +So maybe we should test the URL for having the file protocol and if that is the case use the old logic but if not then use the connection to know if the modification time is newer +This is the first step of moving grouping into the main search flow +of the system tomcat and geoserver after crashing. +Otherwise an exception is thrown +added comment and there something else that I should fix before this patch goes inI need this one for +We changed the class to have the error message from the immediate exception rather than having Error invoking API +Updated changes in committed revision +. +Moved back to has been releasedMoving all related issues to the next version +Good work Christian +Merged onto the branch. +How does the ant task get compiledBy using the default compiler for the ant process +did you remember whats trivial change you makecould you attach the smooks config file +Yes if you cant reproduce it then close it with cannot reproduce status. +It would be nice to get this minimally working at least +Thanks t heard any futher commentsreactions to the fix so closing the issue + is for trunk +Ill keep this opened so that we can turn this into a watcher issue that allows us to track the progress on the backend +I modified it by adding a bunch of testfluff files into the and modules +yes and also to have pictures in png css javascript html page example +Yeah I played with this I think when pristine host headers is enabled we generate the wrong in the cache inspector +Seems like your setuptest just doesnt provoke that situation +Uploading the latest patch after making changes on current trunk for your patchIve applied it in commit aadabedefceaf +Finally XML configurations can be changed at runtime and edited by UI tools +Seeding it with the VDB name is good cause it is what they selected to create the datasource for +Thanks Kristian for running the tests +I removed all but one very simple test and it was no problem to genreate the coverage report calling mvn site +We have test our suspicious of points and it doesnt work +In my view the asynchronous API has been designed to address exactly use cases like yours +Will commit when tests pass +But this is not always a good approach +one thing I noticed about the patch is that you consider ords as null values which is not the case in IDVI changed that in the new patchI also included IDV implementation forI think it is time to include these collectors into the already existing tests +We should dupe this to and close +Thanks +Great it seems to be working +This allows the same context to be used with multiple requests without the caller actively resetting the context +Ive patched this locally and want to execute the unit tests using the MSSQL bundle PM +perhaps a imageUnless you want to fix this yourself and send me the patch +It would be great to get fixed for release +I guess there is a misunderstanding +Removed vertical allignment of fieldsoperators +Oh the port number issue is only surfaced on the Bamboo CI server +Can the task list become long as well +Thanks guys! Im adding a release note about this fix. +The delegation should be based on the configuration on the +I was looking at this again and we can be and leave all of that shifting aloneBeing CLS compliant only applies to public and protected members +This is not a bug clustering has a simple requirement A working shared doesnt work well on transactions and its not shared +There are now relevance rules SourceBinary rule type modifiers rule type accessibility rule javagroovygroovyxjavax ordering rule and similar package name rule +hi jukkawhile working on i was not sure about the status of the classes in the package because they were not used anywere +Im working on an approach for this where the testing classes are always loaded from the new class loader and only the database code is loaded from the old loader +Some times after successful hod allocation using dynamic dfs +Documented according to the plan outlined in the Description for EAP and EWP +This patch also contains a resource utilization monitoring daemon which run on each of the and a information collecting daemon called Collector which collects and aggregates the resource utilization information submitted from each monitoring daemons +The issue can be solved by using the page dimension +This is already fixed upstream +Yes i was able to deploy the instance after i did some random operations +Please try latest CVS +So I guess we are working correctly. +David is the version of orI see the current release of is and the trunk version isI would do a release for soon with this snapshot clear first I think +Cheolsoo I have also updated the class and made sure that unit tests succeed +I added an empty test srctestjavaorgapacheservicemixremoveme +This is for branch not for trunk so reviewing patch in that light I dont think it so bad +Unassigning per a discussion with him on IRC +This looks more like the compiler is having an issue with the member template dynamicCast in the Pointer class which is odd since the library uses this extensively +to and trunk +m allowed to mark artifacts that need to be packaged but with m it looks like there stould be special scope or some additional metadata in for dependencies +I was about to post a bug about Overall coverage not being calculated if there is no Unit coverage when I found this +Patch illustrating change in tag +we need instructions to get rcov working first then we can fix issues +sharehadoopcommonlib +This updated patch adapts some of Maxs comments and adds some additional validationname changing to help the user out +Any chance this may be addressed in +How about the followingMentors represent the Sponsor to actively monitor the podling guide thepodling in the Apache Way and report its status to the Sponsor and theIncubator PMC +This should make strings and identifiers work +This is all new territory for me but I think you would fetch a room from the db or the cash iterate over the other users in the room possibly filtering the list because some are hidden or invisible or whatever +Ok I removed the require and committed +This does not apply to a date. +First try not yet tested with Mono +Patch for revision in trunk +FolksIs there anything I could do to help get this one moving forward? JochenAre you going to be able to find time to update and apply your patch or shall we move this to +ThanksLasantha +If having default is not required in I think thats certainly better +Ashish for committing patchThanks Rishi and other for their valuable comments. +That exception stacktrace might even mean a genuine wait where the client is waiting for a result from the server +Test failed on the th of the iterations +Wanna commit this Ted +Only thought is that it might be worth running the HDFS suite against a common built with this patch before committing just as a sanity check that it doesnt break anything unexpectedly +Very sorry found another missing programs directory in a pathname in +Already resolved +Hope this scheme is OK for your not then we need to create some special scheme for you in this case specify which issue types you need +Hi Frank Is the an simple jar file? If it is could you please try to put it in the lib folder of your ear package there should be the right place to hold those util jars for each module in the EAR package +Otherwise we get false positivesI also removed the commented call to update target filters from the verify method because this update is always done before calling the verifyDM methodAfter this fix the CT also passes +I just uploaded a diff that I should solve the immediate problem +Ive attached a new screen shot showing for possibly a real scenarioSorry if the screen shot was not clear in showing that you only can expand or collapse a single node not everything to the left I hope new screen shot is showing this more clearlyIll look into your suggestion on a tree thanks for the ideaPlease note that these details displayed are just rubbish since this is only a test work +on patch +has been marked as a duplicate of this bug +Hi RonnieI just downloaded from and lib includes Remoting version + +Committed +YesAdding a public method to just for this test shouldnt go in adding to a subclass made for the test only would be fine +ivy and +Based on the prototype I have I feel its a bit too risky to switch internals to new configuration so thats better pushed out to +Yes it would be great to have an opportunity to tell Sonar that these or those violations areAnd it would be also great to be able to use the for duplications alsoSometimes at our projects Sonar reacts to files and the duplications ratio rises dramatically +I verified that stored story dispositions are only character of length in the code +I think thats the point SVG could be embedded like that but has been deprecated so its reasonable that fxg should be able to be embedded like that +Should we apply this patch before it is going too old +I will be working on this +At the lowest level of the btree we often ask for nowait locks so that part is not hard +Hi All I encountered the same issue on my applications using fop +Indeed! Here is a complete patch +Teiid has a max IN criteria size of set for the salesforce translator +Its a private function anyway so not a big deal +Eg first leg to put into persistent aggregator +Kay I will have to test that but you can please try to select Maven POM XML editor and Maven POM editor features and select required with Callisto node expanded after that? Alternatively you can select XML editor feature from Callisto and kick select required for seem to have missing required dependency declaration for the core feature +Could track and maintain the info in the system catalogs for each table and index +Closed reopen +I will put some thought into how to package them +I think we should just make sure the EBR artifacts have the correct manifests and then resolve this issue completely for RC +Will fix this +This issue is no longer relevant for it is fixed inChanging the affected version to. +This patch contains the following Checkstyle configuration based on the one already present Remove the current checkstyle file Adds the Checkstyle report and plugin configration to the master pom +Im moving this to M since Robert has already committed a new +Review please lads +The URL may come from any repository connector and they all have different forms +I will post this issue with the user group +Thrift has just been released +Apparently some content is needed +Sorry I dont quite understand the bug report here +Max please review my patch and comment +cannt insert content from description file into description for this issue or comment so have attached as file +The using clause of dynamic sql creates a using scalar group for use within the dynamic sql +Ive found another situation +Will not implement. +Use su command instead of setuidgid +The design is currently being worked on but no we dont want to automatically require two sets of credentials. +Postponing. +impemented option added as configiration parameter or action +as suggested by Tom patch redone handling this at server side returning NULL if job not found +Correct +Marking patch available +Benson has started quite a few threads in the Maven team so he might have run out of time to do the release +by using the node type definition instead of the node type definition +Yes it gets hanged +add the missing print calls for those two help functions +Logged In YES useridIn this is just a packaging issue +Thanks +Checked in develop branch +This allows for a split to be totally contained on the RS without any interaction with the Master so we can have data unavailability really short no need to reassignment +Ideally all those modules should inherit from a few common parents where you would use pluginManagement to lock down the desired plugin versions +Quote from the linked issuebq +All the problems I had are gone now +Thanks for the patch +We should move the method to an util class so we can better reuse Claus explains why I felt dirty afterwardsIll update the patch this evening +I will update it +But having that mix is certainly problematic +If I save the save the jBPM process as a +I am marking it fix for +will track storages and storage types reported by the +See and for some detailsOk awesomebq +I uploaded the result after doing some moving around of the nodes and saving +This implies that search has access to our indexed files as well as the folder they reside in +There used to be a script for it +on commit to and trunk +Feel free to reopen add more detail if the issue is still present. +Fixed as Brett suggested +Ok for initializing the header in the constructorI preferred a null array to save some memoryThe intent is to have the mapping in the parser and share it with the records +Just pushed this to trunk +GAE is working again. +Assigning to Darrin Mison as this is moving to a Doc Jira +I think and need to be brought back up to trunk? A quick patch attempt is giving me problems +no checking js usability +Could you provide a link so we could look at this documentation +I committed the patch +Couldshould this not be folded into the existing Ngram code in contrib + is the target to run in the testsuite +It is possible to force the use of the old mechanism by deleting or commenting out the following lines in javaengineorgapachederby and rebuilding store data using a Enhanced version using NIO API requires Java derbyIt would be great if you could make that change and rerun your test +Youd need to do a git merge and then compare it with the current branch +. +I have committed it in Rev +Maybe you should simply start reading docs for the Maven settings +Set as a response of a inonCommandCan you be more specific here Im not sure what you are trying to say what is set as a responselines of code would help? I know that the Response correlator doesnt clean up futureresponse objects on destruction as it expects the caller to handle this when the response times out +IvanI have a couple of comments about the code is missing an Apache License Header +It seems OK +Issue is fixed +actually this one might warrant major priority since it really bricked the whole of roo +Requires with the release +Why not just run mvn install and let it take care of the dependency management and populating the repository? Your workaround is basically doing the same operations but via manual steps +A fix was committed and will be available in the next version of this issue is still not resolved +Side note doing a get on a disabled table will send back only at the second try +Updated patch +Is it possible to get a stack trace on Windows like you can with gdb on linux? If you can having a stack trace of a couchjs that you then kill to free the VM could possibly be the key to figuring that out +If this solution is indeed an improvement the percentage should be computed depending on the value of the offset +I only changed the poms for the artifacts from spring modules that were necessary for having a successful build process but if the solution is viable also the others canshould be changed +Should be fixed thanks for finding the are still getting our CI servers switched over to git so if you want to test in the very short term you will need to build yourself from source +Committed branch and trunk +I found the same problem when testing the issue +Committed this to trunk and +Cant reproduce it issue as well as and some others have one feature in common they happen when test is started before updatebuild is completed +Theres a few written in the latest change set r order to enable edit all instances of and flip the truefalse flag to true +I dont want to bog down this discussion pertaining to whether or not this is a good idea +Will commit shortly +If you reopen this would you paste in the new stacktrace? The line numbers shouldnt be the same as the ones in the original stacktrace +I fixed the white spaceindentation issues and committed to trunk revision Kristian for bringing the issue up to dateIll leave it open for now as I plan to backport to and +Resolved +not thumbnail or maxHeight or maxWidthIn effect this code is double encoding +Completely removed the monitoring code +Initial implementation is checked in +Madhumita does unlimited crypto policy solve these failuresYou mention these failures are random +I screwed up somewhere along the way doing too many changes at once +expects some userId to be used as author however in supportedFields the same was not applicable +This is the NPE bug reported on the dev list today +I tried to fix the most part of your items in the latest version +Is there a workaround in the meantime? I am simply returning a block from my action handler +All is well in and regarding automated unit test +Ultimately its still sored in zookeeper just in a more systematic way +This is refixed +We again abuse the column names a bit and add a userid column with a secondary indexUnless I completely misunderstood your statement above and these secondary indexes are possible in cql I suspect your answer will be to use a separate CF but that does put a lot more burden on the app the avoidance of which is one the big benefits of secondary indexes +Im not sure what to do in this case as if we record the appliedterms per segment but keep the pending terms in last segment info wereeffectively not gaining anything from using last segment info becausewere then recording all of the terms anyways +And I sincerely apologize +Closing all resolved issues for. +checked under GA +deflate will be preferred when both are accepted by the user agent +But of course +On the HTC EVO G the scrolling was slow and the application was generally unresponsive due to the numerous images and gradients used +Updated summary according to evaluation +I only skimmed the links you provided but enough to see that it seems like there is a lot a varibility based on OS and filesystem +At least I supposeI must have been thinking that I really cant rememberAt any rate I unchecked the Release Note field as I cant recall that this introducedany incompatibilities with existing applications. +Next on the chopping block is then global local and inner supporting code +but seeing as how positions are ints too I think it might be easier to do versus the parameter +PR applied on master also applied on branch entries in BoldtYes I do but I failed at rebasing the commits on the branch as I wanted toshould have +Ive found no reference to a constant in reflection for varargs +APR specific memory leaks should be directed towards the APR bugzilla issue tracker +Synchronized filenames +These are the needed files +is already a blocker for +problem just wanted to make you aware of it +If id is so problematic the why dont you just create another parameter so it works in all containers +Ill take a look at this and start running the tests + +Or is it that you have a cluster with lots of regionservers? Can you not just up the configuration for how long to wait on new regionservers to check in? Also can you say more on why you block regionservers coming in and actually tell them shutdown +I may want to revisit this code later but it can be considered done now +What behaviour do we want then log at severe level and continue +there is something you could try which might help us prove if there is actually a problem in or in +Working on I essentially had to the because of the hardwiring of arrays +It doesnt work with an with JDK beta but this still doesnt work +Contains the hdfs side of the changes for small reworking of permissions for createSymlink +I committed the change to the branchThank you Jing! +Patch rebased for trunk +Do we need an updated patch to capture my latest change then? Also do you want to create a new jira case for the changes Benjamin wanted in the zookeeper programmers guide so that doesnt hold up committing this change +Fixed release audit warnings +This test case exposes oneof the more challenging problems of typesetting footnotes +Test CFM page to replicate the Shot of existing code +Patch looks goodI just committed this +PR passed +Please fix +Tom We should apply this +Resolving issue +v attached +Because session was invalidated access mode would be public instead of private mode as you wantSo should we reject this issue because its proper behaviour +Thats why its manual I guess it should rather be linked to the scm plugin which is used during releases +Closing as Fixed now that the patch has been merged to. +Patch looks fine to me +So is moot +Closing the issue +This issue no longer occurs for +HiWhich runner are you using? The swing runner? is the configuration I use on equinoxosgi ssFramework is launchedidState BundleACTIVE ACTIVE ACTIVE ACTIVE ACTIVE ACTIVE The bundle is optional its just used to debugRegards Im using the swing runner +Ok are currently restructuring the code base so there may be some small lag before we apply pending patches. +Im having the same issue +Move it back in if you think differently. +This bug is caused by changed from it imply works with +Im afraid youll need to do some more digging since I dont have access to your application +its a dog the more copying first the worse it appears to behmmm so transferTo causes the write buffers to be flushed or something +I fixed some left over work refrences that didnt get converted to and add +So we should be able to generate the manual +This review will be closedReactivate the rule +Yoav has I believe greatly improved testing +First part is in rev time metrics are working +Marking as Cannot Reproduce Bug since we couldnt replicate the BRMS issue but even though we did add a test case under this issue. +Thanks Junping for the patch +Attaching patches that adds the features described above +What are others thoughts? Should we commit this +Not even sure why it is in there to be honest it serves no real purpose +I will run tests again +The fix for that is to copy the componentType file to the resources directory +Since we agreed we wanted CONTAINS KEY anyway and since I dont think those bits makes review much harder Id rather left them here than going through the pain of extracting them into a separate patch + is the code I use to wrap the library that all my other software uses +Im not sure if this is a really good idea and Im thinking it is better to have an explicit installation +Patches accepted +This is no longer happening in the active branches +My guess is that because is logging debug events that lots of these are being created and then released as they expire +Committed revision Thanks to S +See +Or maybe it is actually fine to let the store files build up and just be smarter during compaction about trying to do too much and blow out heap +I think that this is dupe actually +Committed at subversion revision +Turning the tests in to normal Java code with our current would likely at least triple or quadruple the amount of lines required drowning the essential bits into boilerplate codeCan we combine the good bits of both approaches +OK excellent can you commit that +Still need to integrate it into if we want to simplify an apps life even more though Id prefer to wait for some feedback from anyone that actually uses it first. +patch for trunk +Patch simply changes the following options to flags useKey printKey csv names nameOnly sortVectors quiet sizeOnly. +This image was rendered after making the changes described in +Since I think there is at least general agreement on the overall change Id like to commit soon this involves a lot of changes over a fair number of files and many moves +will wait to see if this is bottleneck in new SSCA implementation and if so address in new issue +Duplicate of to just for verification reason +Snapshot is deployed please test and confirm +We need to allow hudson to post to the dev lists +Added debug messages to the executions +updated +I wonder if we need to revisit how we buildversion these plugins so that the SNAPSHOT is not in their versions but rather the actual timestampqualifierversion of the artifacts? +Guess I misunderstood what bugzilla meant by accept +ll commit this soon +Test and imported class attached +It has to be fixed ASAP +The service starts without error and the last event log entry says Creating snapshot of projects +this was a temp file issue fixed +Looking through the flow of the implementation Im starting to think the whole problem is related to our use of custom scopes +Current way of using will make it hard to say reserve for use +agudian any progress on your examples page? Struggling the digest the implications of +Thanks! Works +It would seem prudent to keep this behavior consistent +Plumb configuration down to the reader so we can avoid parsing the site file when creating a reader +line nitpick serde Class Can we change to Class of the serialized object +Committed +I am able to mvn clean install from basichelloworld think this is fixed in trunk +maybeThe files are all +this bug may be linked with and seems to be broken +Owen and Arun thinks it is too open for outside of hadoop community to inject code at this level +The upgrade instructions are now complete and a ticket has been submitted to engineering ops to complete the upgrade process +Camel uses slfj as logging +We can submit it today +I misunderstood +to you for updates +I forgot that issue +Committed revision Thanks Stefan! +Ive created to cope with the creation of version archives +It looks good +As for the forrest documentation structure how about the creating a Japanese tab and just creating a japanese subdirectory under xdocs? Ive attached a and patch +Im not sure what semantics a user might expect when update and delete are combined but this seems like the right fix to me +This was fixed at the project. +Action conversion properties enum last note +Would be helpful to have this issue resolved +Sorry the url link in the footer was wrong corrected in the description above +All of them have been updated as well +fixing it +And theres non need to rush +See also for a proposal related to reducing the use of temporary tables +committed to +If no other problems are found with existing sorter implementation I am ok leaving it for now the way it is if there are new problems I think the sort observer solution is cleaner for the futureIf you have some added comments for the sorter file just submit them as a separate patch and I can commit those easily without worrying about test running +Thanks for your insightful reviewsCheers to trunk +This seems to have broken a bunch of tests since get now throws an error on the standby so wont start an HA cluster anymore +Thats not my addition + +false testAnd you should see the errorUnknown argument +Ill ask for permission +The problem is somewhat obvious the column reference in the select list is not resolved properly when it is qualified by a synonym table name at bind time of the column reference +Hence with a message store size of gb and a data file size of MB it takes around messages to kill persistence which is a very realistic scenario + +documentation +HiI have added an it with a multi modules project and I cant reproduce you have a simple project to help to reproduce this issue? my badYou are using checkstylecheck and not checkstylecheckstyle +I updated the test to dont care about processing instructions for response anyway Id like to know too why encoding appears for AS and not for AS in response message +verbose comments tend to be annoying in practice operationallyThats the lesser of evils compared to documentation of the options +Well Ceki is watching this thread so I hope hell give his own comments on this +Id like to use it as a test dependency to get the complete repository stack including remoting options +It will be some final status upon attempt completion but some invalid default value when the attempt is first created and stored +It is definitely an assembly problem the source code is simply not included in the distributions +catch the correct exception in test a few more edge cases like supplying null to native functions when were not supposed to etc +Committed revision for the revert +is already listed so if possible I think it would be good to merge the fix to as well +bq +we need to decide what tool we want to use and port the initial documentation from githubweave to that tool +Youll be in the next update. +What could be the the best place to put these validations? I guess in MR api we can put this in and some where for tableI would like to know one thing that is it possible to add custom validations in thrift service itself? I am trying to google it as wellRegards regex is a good way to go but it shouldnt be user defined +Ill commit this one shortly unless there are objections or improvements +Is corrupted +Thanks Edward! +By default we assume can talk to AM to get full status etcAs described in the description of the jira its for a few special cases +Sorry still working on old issues +Bulk change issues closed by JEXL. +Seems a trivial change in constants file and docs update +Ive got some code that uses the JTS WKB io packages available in JTS to do this +Would you mind stop creating tickets for each and everything here? That creates significant overhead that Id rather like to spend on actually fixing bugs +committed w removal of obsolete code +Thanks a lot +Just a few comments based on my quick read the dialog showing up asking to certify I assume is only shown for your custom key which is not signed by a trusted entity right? shouldnt this be done at JBT component build level and not JBDS level so there is no question about wether need to repack resign etc +Could you describe how you get to this issue and how reproducible it is +The reported error is still valid but the patch here can be ignored as the patch for committed all the test cases together at revision +sound good to me let me try it today +Initial polishing was made test and test config files were moved together some minor refactoring was done to a few components etc +problem doesnt exist on as that has no it right that makes a jar named? Should it be called? Otherwise patch looks good +Right Im not arguing against doing it +Closing stale resolved issues +As a consequence this data was not populating correctly +Repository renamed to Common Sonar +This is still an ongoing issue +ant test in the directory fails for me on both Linux and cygwin +in resolves this problem as does the ability to set prefetch of zero +Create a document with the name Human friendly url test +Hi Stefan My sincere apologize for the file corruptionHurray +Thank you for this patchI have a Pro that dual boots Mac OS X and Fedora Core and have been trying to figure out why this one particular test I was running was always successful under Mac OS but always had timeouts under Linux. +Id say that option listed in the beginning is the Maven way although it means more configuration for the user +What is the likelihood that the above will happen in time such that we can resolve this in time too +I would like to resolve this bug as a duplicate of if you are no longer seeing the odd error +See Deprecate multiplex transport +and derbyall passed except for a known issuePlease review +Great thing to get solved Why not update the documentation now +and + +Add patch for +Hmm are you just using the Thrift api? We probably left that alone for backwards compatibility +we should put forth the best defaults LuceneSolr has to offer here +Uses the interface for bean lifecycle we negotiated with the team and many thanks to them for their cooperationThis also cleans up construction of Modules and removes an unused and apparently unneeded method fromStill need to integrate with tomcat +Look at +It is a nice feature to have for people using but I can easily get by with the command line for now +Went from min sec +Closing all issues created in and before which are marked resolved + Username cptnkirk +Hican you link the RB item to this jira +I agree this looks bad we would need to strip out blank p tags +Please attach example demonstrating the problem and reopen the issue +Great news +Added a better error message in r. +on commit to +Doh yes it is +I guess it is less of a big deal since you only redefine the menu inside of a groovy editor +still got the same problem +Committed thanks Ivy. +Can you confirm this bit of magic has made your builds in the Brno jenkins happy again +The Download button in the Runtime Detection page would enable downloading any runtime +Merged in prod. +StackIs this the patch committed? Lot of other changes also in this +SELINUX can take one of these three values enforcing security policy is enforced +SLD example with octagonal envelope filter function +patch from implemented the patching +This issue is now fixed for the upcoming is OK for both and. +How is this resolved? +Committed patch to branch at revision +Support consists of a default set of formatters for formatting field values installed application wide the ability to register custom formatters per class to format model property values automatic model binding validator invocation including support for invoking the model bean to validate itself and support for invoking model validators ability to supress model binding and validation per transition +Dedup blocks would be stored in a hdfs filesystem with replicas +Verified on EAP ER +Changing to improvement and dropping to trivial +are we still using those dispatchers? I thought a bridge from the new dispatcher to the old code was built +Continuum was installed and later migrated to version +Hi Bryan thanks for your reminding I think the new patch can work well now +It seems in the launch configuration it Equinox as the only possible implementation +We dont handle that setting correctly with we check it by iterating through all of the cores that loaded but a core with a config erorr wouldnt have loaded +We actually do want to support onException in a route and at a global level +Mark had merge the fix to trunkSo close this issue now. +I assume you in the previous comment means me david jencks +One problem I just found is the patch automatically attached by phabricator does not have the ASF license +Revision extends updated +Dan you have fixed this already havent you +Could someone please try this patchThanks +Problem was the profile listener was being notified at the time the JDBC sources were being requested +patch used to rework prior patch for this option using the in +this patch is good for me +Im working to provide a patch for this issue and will post it for review soon +Changes required to make patch plus some changes for generics in as is already released +Benjamin does not support SOCKS proxiesOleg +OK I understand the bug and I think I know of a fix for it but it will need to wait until after this release which is badly needed +There arent many scenarios that fall into that category anyway +Not sure why thats not the case with the sample +bq +Fixed in svn +In my tests the aggregation works fine +As for versionning Ill look at it but cant say whenOther than that Im waiting to hear back from Lars regarding my response to his suggestions on reusingSorry for being slow to respond Im very busy with other things these days so feel free to make any changes you feel are right +reattached from bugzilla +I am using hibernate instead of from the pom give me some time for the stacktrace. +q system +Verified in +Could you describe the situation more in detail? Are there any log exceptions? I cannot reproduce that + issues so im going to exclude the entire commonlibFAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current FAIL serial error for current +That seems like a nice thing but I will leave this open for now so others can chime inCheers if you use I believe it actually does do eclipseeclipse underneath during the import but you can just do an existing project import from the git repo +I attached the patch +is still not implemented +If and when I have it working I will let you knowThanks issues in +I think that the main purpose of this issue is making metrics system and applying to daemons metrics +Bill Could u please help with this onethanks updated patch is based on AXIS revision which contains the original patch already committed +branch +Please upgrade to the most current iBATIS release and see if that resolves it +Please let me know if that fixes the problemThanks for investigation StanI think we can close the issue since workaround is pretty in this case. +Adding test case and closing the ticket +Now it returns a Lock object +closing all SOC fix version +Forgot to mentionHad to fix toString for proper directory layout +What is the SVN branch for this? I need to check if my changes are going to make it in +I applied an addendum for v to TRUNK +Seems reasonable to me from a Flume disagreeCorrect me if I am wrong but return code happens only In case of an error while processing any action except for status the init script shall print an error message and exit with a status code and the program is not runningNote that in the paragraph above the one considered by this ticket running stop on a service already stopped or not running is one of the case where the situation is to be considered as successful +Bulk close of resolved issues of +Attached is a document containing a description of the Filter Language and a patch the implements it for the branchIt would be great to get some feedback +will commit if tests pass +Updated patch includes a file missed in previous patch and fixes a test failure +This specialized proxy would just use the location passed to it by the source it would not try to determine a mutations location +do you have a in mind when such a policy if beneficial and why the current policy is inadequateis a policy that tries to keep all disks at the same usage capacity better then the one? especially when disks can die get repaired and then put back into the same system +Regarding the proposed fix Im afraid its not the best way to do it since it will break the signature generation +Ram +Hmmm infoStream is just for debugging +In this case we should generate errors instead of ignoring the badly configured dirs or count the bad dirs towards unhealthy disks +which is not part of and instructions are fine +If you have any issues with this please let me know +Thank you for your contribution! Resolving FIXED. +This will be implemented with a validator so the name can be changed using config +Instead let the parameter pick up its collation from its context +Sachin upon perusal I see that youre directly accessing Sqoop internals which is not supported at this time +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +I hope the merge for the container is complete the following classes must deleted after merging is completeinternalimplcorespioptionalsrctestjavaorgapachePlutospioptionalthe following files has been addedThe following file has to configurate as it has been done in the trunksrcmainresourcesorgapachepluto +Is there an expected time when this issue will be resolved +wraps a very long line else what Matteo suppied +So QPS compared to QPS really cannot matterIf the benchmark can really satisfy queries in ms thats no QPS! Ill come back with a test in a bit to confirm the functionality but Im not going to worryabout performance here thats crazy +Provide a unit test if you want us to look at it +bulk move of all unresolved issues from to +I will commit it soon +FYI I created an issue for this edismax bug in and committed itHoss I dont agree with your reasoning on the the q parameter +Fixing RN fields +The pointcut is not well you comment on the pointcut definition according to the proposed syntax and based on the actual annotation from my testAnd yes its true I would like to apply the aspect only for the methods thing that strikes me as odd is the fact that the test bean is a singleton and the aspect should be as well so the proxy should be created at startup +One concern is whether its okay for rd party to be another ASF product +This is purely related to JSF we need to investigate an existing JSF token implementation or implement our own until this is supported in the next releases of of date for all Tasks and Feature Requests created before +Just thinking out loudOr perhaps we could make it so you could pick the hash field? Then hash on digest +In limited testing nested converts seem to work for me do we have an example of something that didnt work +Committed to trunk and branchxThanks Tim +The original was No receive timeout set on default JMS template when none provided for JMS inbound channel adapterIf this newly focused issue is not sufficient andor you disagree with my explanation of the default value above then feel free to open a new just wanted to make sure we at least get this much in place since we have to create a quick anyways. +that odd client would have to use a strategy other then redirect to many times does it happens that a node fails and at the same time you would have just between the real and the redirect request you that would have had messagesyou have to use a round robin load balancer but for new sessionsNot for existing sessions +since this wasnt fixed +So I say there is a bugWhen the servlet says length I get this from the client logreadline van sz readline van sz readline van sz readline van sz readline van sz readline van sz and here it gets stuck because no more bytes are available from the inputStreamThanks +Ive fixed this issue on the trunk can you please review and backport if you like it +Im not sure how to model a similar behaviour with the jellythreads taglib. +Verified in the release for assignee change. +That is correct +can you pl repost a new version of the patch +Pete Im assigning this to you +Thread dump attached +resolving this as the issue seems to be related to spring jms template usage +Missing JCA JMS services +However underlying relationship is represented to correlating color treatment +optional please provide a patch not a whole new file so we can easily see what has changed +Heres a modded that removes the use of the Valve class which has some contention issues and replaces it with a +Thoughts +I combined the testsuites and committed the changes in revision . +does explain what is happening +Sweet +I just grabbedgit clone git That refers to groovy is that what you are expecting to use in groovy eclipse? I cant remember the last update I did for the compiler level let me check +Best would be able to be able to change the setting on all projects at once +If you find the test case notcomprehensive enough just let me know where it needs improvementCheers applied commitd site updated +patch may be ignoredAfter discussion over a couple of mail exchanges with you folks I have changed the implementation for +An interesting comment by recently was that maybe we should ship w major compactions off most folks dont deleteHmm +but it would be workable +testing +Using a of the documentation + +Thanks Jarek +Is that possibleNicholas mentioned earlier that it might be helpful to print out how many bytes are left in the log even though this can be computed from the information provided it could be helpful to be more explicit about itThere may be other information that can be printed out too Ill take a look +I dont think there is a reason to buffer streams serving bytes from the pages in the page cachePatch ready for initial review +I think merge can wait until we decide on distributing stable Sqoop package for WindowsI would like PMC take the decision on merge via VOTE after we are readyIm sorry this may be a naive or a mean question but do we have any community obligations with Sqoop being legally tested and supported for a hadoop distro? Do we need to wait until Apache Hadoop announces native windows support? I personally dont careI agree on umbrella JIRA tracking these changes +For now we like to use the fonts in our testsuite +the bug is fixed in +Fixed +Attachment is the patch which resolves this problem +Gunnar Im this to you since you have several issues that might overlap and I know you are planning an overall review +OK finished. +Then please move it to +Will comment back on this issue if it comes back again. +Then to will produce what were expecting to get +Converts various to null return value in line with JDK semantics +It is fixed as of now but I am not sure if the target is CR or GA +This requires to remove the from the final generated url +This problem ie sometimes we can get output and sometimes we dont get it with the same code occurs with both of these outputs +for all annotate all methods with publicstable +any idea why i am not getting those +Is your use case restricted to String values so you can handle empty values in business logic? What would you do with Long values? +Attachment has been added with description Problematic svg has been added with description Screenshot from Internet Explorer has been added with description Screenshot from Batik +By module do you mean full +However it will be sufficient to remove this as a blocking issue if an official at Yahoo! would simply license use of the marks to the ASF with an intent to transfer any remaining live marks after graduation from Incubator +If I run standalone or part of All I always see a security manager installed in its setUp method +The US locale is passed when constructing object or of the US locale whichever applicableAlso I have removed prefix failing from testsettext +is not a good enough reason for me to revert this patch +It is tested successfully +Thanks. +I expect these kinds of error to shake out when I expand the test suite +Ive seen it happen on both Mac and Windows though it seems to be much more prevalent on Windows +Let me push to i commit this Lars +plugins +Oh I should add that once the redeliveries are exhausted it ends up getting a and jumping down to line +Resubmit after bumping up sleep in script for test failures +Upload as + +Commited to trunk and to. +The Workaround is to use instead of null as the column name +Retargeted the issue to since is being released very shortly +Well the old style went away and we completely switched over to the new style wireformat so most of this patch is not applicable anymoreThe good news though is that since last sunday messages is supported by the interfaceThat being said i do like the sample container work to be able to set messages in the xml file the sample gadget so if you feel like porting those to the new social api id be more then happy to apply those! +If the power goes out on a system running lots of compaction then data loss seems highly likely +Understood +lgtm +So doesnt mean the depth of the node in the tree? The ranges dont change after the iterator is generated right? But inside the while loop in there is comment about adding a new range +Much of this is out of date or has been fixed +I have added all of the Galileo SR Java Developers package plus a couple of otherGalileo packakges +Hi JacquesJust to set the record straight +Done. +The itself is held by multiple strong references and only is highlighted in the picture +patch for tar version file patched are the format I cloned with git the trunk repository but the two files Im proposing to change are different +after that I switched back to with that same swc +Only changes to are intentionnals +I just committed this + +I have not encountered that at tested on Windows and with Firefox and IE +Updated the nm and rm auditlogger tests to use the new api +I suspect the patch will be appliedthats a different issueJust be aware that a lot of web app libs are not designed for shared usage like this so be very careful +Closed as a dup of +That makes it extremely hard to see what actually changedSorry maybe I should have provided incremental patches +Ive started work on this already +The only thing missing is a test case +He said he would comment on this JIRA today +I left some things under debug that I thought made sense to stay under debug and added one wire trace to indicate the method and URI of the inbound request currently being processedIm guessing more trace debug will be needed and I dont expect to get it perfect the first time through but would like to come close +Should we add something to the doc here in case others run into this? Good stuff +Well that makes it even simpler +But I understand you feelings about this commit +Verified Penny runs with this patch +also checks the outputspecs which requires hdfs access there is an opinion that check should be reintroduced as optional check during job submission we can make the output specs check as optional as well this way the client will be able to submit jobs even when the HDFS is down +Hi VladAny newsThank I use +I just committed this +Tab completion still to do +As pointed out before this is also in xmlpullxmlpullShouldnt all these be in orgxmlpull? +Use existing convenience method that I forgot I wrote. +We need to isolate the client from directly initiating zk requests +My code was modifying the bounds of the layer rather than a copy of the try a nightly build or RC when it comes out. +Try the patchYup it fixed everything +At this point only the standard connector tests are used and the JCR connector passes all of is still an assumption that all node types in the remote JCR repository are loaded into the repository +In this patch if is present in configuration rather than just ignore it its heap size setting wins out over any setting present in and a warning is emitted that this setting is deprecated +Close all resolved issues for Engine release. +Didnt see this problem anymore since switched to JPA Hibernate. +Stop bundling and with +I havent done a formal trace of the code yet but I think it would make sensethat the offset should only be incremented if the character is pushed into thebuffer +Will mark this as resolved +Ive created some fixes for this and have it working against the latest branch +There are no loops in the hdfsWrite or invokeMethod +Picking XNIO to start with since we are familiar with it and it too is a project other transports may evolve as and when +with langxml you have callouts but no highlighting but with languagexml you have highlighting but no callouts +end I think this is the responsibility of the developer to mark their classes as serializable +OK that militates in favour of special global operations on the the targets registered themselves with a central list of targets then we could have functions that iterated over all the targets doing things to might lead to interesting alternatives to the global hook split as it is now +Rob what is the correct semantics +Look at the attached screenshot +can be edited +And I wouldnt say it is minor as it is pretty much able to destroy a presentation +Also since the tests are anyway working only in hadoop can we also test for that scenario +For some reason Im unable to assign this bug to myself +Also we already print the names of the log segments followed by whether or not the index and data files exist +Integrated to trunkThanks for the patch binlijinThanks for the review +Ill start by reverting I suppose and keep looking for a solutionAt worst we can pass a isFreebsd sys prop or something on our free bsd jenkings machine and then dont use NIO in that case +Update and save +Ive run mutlple builds and didnt hit the timeout again. +If you are happy with it please close this issue +This should be fixed in upstream Undertow +I have attached the new one here +Commited on trunk at r waiting the release before backporting to on as well +I have no idea why part of the above comment has strikethrough formatting +The paths for the tests were also fixed +to Show all eligible beans +for scheme i dont think we should use a prexisting scheme name +its probably a download issue +Bulk update to fix version +After revisiting the ERD the intended use of the jdbc driver was within the VM +Closing. +Done +I also fixed a dumb index entry and a formatting space issue +Delivered fix to both and streams +Fixed thanks to Brians input. +Even Tomcat and up can support hot deployment +I should add that its working if you specify the return type on the interface as the extended typeso I guess this is a polymorphism issue +If you delete a project level Notifierit doesnt ask for would be the behavior? You always have to require confirmation +Thanks ThusharaCommitted to both branch and trunk its a bugfix so its appropriate for a branch +As Lars said as we move forward with and use Maven this should go away +new version wich take into account the comments +Hi ClausOk +There may be a top level application class loader an extension class loader and so on down to a system class loader +I think its ready to commit! Ill wait a day or so +Still points to +Suspect the problem is because of forgetting to add new nodes to JAR build process +Just fyi what they do is not magic we can do the same thing with our own style sheet and a little javascript +For now were thinking about trying to put this in a profile so it is not affecting our normal release and then checking out the tag and pushing the site out after +While there are entities now there is no way to know if they are actually correct +Fortunately there is an ASM utility class that can do the heavy lifting for us +A prototype which can handle and deployment plan using schema. +Thanks for the clarification +Yep +in mavenpathconf can confirm that this is an issue in IDEA +So I guess my vote would be a on getting this in for whether the RC gets rolled today or early next weekI could build a snapshot off trunk and run it through my own suite of Avro Netty unit tests +Angie can you please reproduce the issue with the TRACE log level +Solved in current trunk. +Do you want me to look it up somewhere +installer doesnt require privilege lifting on JBDS +Forrest It would be best if we can get a complete test suite of the different places persistence units can appear in the testsuite +I guess contrary to the javadocs you need to specify either or I had tried that too though I just posted the simplest example that conformed to the javadocThis also gives me an exceptionhttplocalhostsolranalysisfieldhello was missing +Added support for both SVN and Radim +No longer a problem on trunk. +Until the changes are merged we should leave this unresolved +please do! Look forward to checking it out +Thanks Alex! +Resolving this issue as fixed with the open issues being handled by issues +Im testing it here and if its ok I will commit it +The server side properties are cleaned up also +Upload once more this time with licence +Could you please try your application again against the latest from trunkIf it works Ill resolve the issue but if you are still having problems if you could attach a log from the region server that would be helpfulThanks +This patch is great! It adds a class to represent sync operations and it cleans up a bit the logic on the leader side +I didnt test them but I will once they have been committedGreat thanks +To be replaced with accumulating variables in +Please check that it fixes the bug +Nice catch committed. +I assumed that the sole purpose of this queue was to know when a new runnable task has to be created for a session buffer +make cancellable through timeoutReportermake it possible to get the timeout value by verb so that can drop according to different verbs +Corrected formatting. +Eclipse project demoing the issue +The current docs refer to but they do not mention the alternative location +Merged +Align attribute work fine +Closing for now. +Raised again to blocker because its a blocker for RC I also added to exactly its a blocker for a RC or even a GA +Fix the snapshot test to catch the unwrapped exceptions +We fixed the problem in by looking at the UDAF to figure out how to use an array an input to a UDAFThis new patch is a working implementation of bitmap indexing +Similarly you can export ERLMAXPORTS or ERLMAXETSTABLES directly rather than pass them as flagsOn a less straightforwardly supported init setup its on you to create your own wrapper script to do this +The good news is that we dont need to persist all of these tokens for RM restart purposes +Addressed comments +descheduling from current modified patch in revision +Hi Todd I think it is not incompatible since throws previously +You can ignore the changes in system test and bin +Propose to be closed since no further info. +Attaching patch for this message +Overall we are getting at leat a factor of speedup when the list size is significant +Attaching patch to reflect latest changes in trunk +We should probably fix that surefire has been updated to +The console output actually says contrib tests have run successfully +to your current codeClaus youve pretty much completed this task so Im assigning it to you so you get the credit +Maybe one could provide away to release locks when they are no longer neededBy this do you mean to release the locks when the LOB is no longer referenced? That sounds good but may still cause issues if the garbage collector has not kicked in but perhaps would be suitable for backport to +There was no OOME but a severe performance degradation after a couple of commits +One guess the rows you inserted had not yet been flushed +Vinod I could reproduce this issue on jdk when I reversed the order of the tests +When i tried to run the same workflow after fixing the output parameter workflow is not running and i am getting exception mention above +I dont get it +It now uses properties and We now need to provide default exclude list for and and then update the tests for removing version specific tests +Most RPC code must already handle. +do curl and Firefox allow this to matchThe matching could probably be simplified if this is not allowed +Thats a good ideaIt should be pretty easy to do since has all the information needed +HiOK i will try another test asapWere using CS and vSphere Regards +Can you please submit a patch that treats applicationoctetstream the same as application to correct this problem +I should be able to get it into M or the subsequent release +your test case failed on my cocoon + +yes Claudio this is exactly the same use case +throws for +Thank you very much sir. +Changing back to rev does not affect us when using this variable +Likely to be hardly noticeable but still +I dont know what is or where it came from but there are a lot of issues with basic functionality over the last day or two that have been fixed in git branch mean debugging the branch not seeing an issue like this +and if you have a patch why dont you attach it to the jira then? +Now I just have to figure out a way to pursue that theory +Thanks for tracking this down DavidI cant change the files but I modified the permissions on the jars to match the distro +derby triage +This will go into Launchpad Base this property in Rev +The put thread does not run the flush +Make it a subclass of +I can hope anyway +JIRA Issue I know you are doing some work on the Maven build can you ensure that this all whats the statusrelevance of this now? Has this already been addressed? Should we resolvereschedule this issue? this time were not planning on supporting a Maven build of the Spring Framework +Contribution is from Attila Szegedi +For these last three Ive changed the visibility of any public methods that are only used within the package to so we start out with as little exposed as possibleIve also made public and turned it into an abstract classThe attached script should be run before applying the patch +Starting a DN with security enabled all of the env vars set but not configured with low portsThe DN now starts properly in the first three cases +I incorporated the first two comments +fieldName damage notNull sizeMax add field string class +No tests because its a change but I manually verified that it works properly +This would however be far bigger change that what I did +Verified this is a duplicate +The change might impact a lot of EE related DUP code +Checking if password changed to force authenticating again +could you be more specific and show me some screens that are not well renderedThis patch is intended to facilitate the developper work and to make him add less options when creating a form +looks good +Giannywhat do you mean by WSDL operation using a wrapper style mapping can you please give me an example? am ok with the second patch +This feature has made it to the release so unless bugs are found Im taking this issue as resolved and Im closing it. +I think a case could certainly be mounted that Groovy could better distinguish between the static and instance cases and use that knowledge when selecting methods but perhaps an even stronger case could be mounted not to use static and instance methods with the same names and same parameters +Santosh At the moment it is not possible to have a unit test for this because of security +Native builds are only supported on Linux +I think I already know +Interesting I tried uncompressed form of our war file long time ago the result is worse I dont know why and didnt go further on investigation +Andy Did you ever write this article +This is probably a very badidea someone that actually knows why this error is occuring should fix thisHowever the attached patch works for me +Also I have included the in front of like this jsps +Verified on. +Yes this patch fixes the issue was having +I have already made the changes and currently testing it locally +project +Although if there were jdk build profiles that were run by I am sure at least groovy would benifit from it they could then run their GUI tests with more freedom. +Ill remove it from Seam settings page +Even if they did have the poms Id argue that its useful +Thanks for the clarification Brian! I noticed that typo yesterday and am working with a PMC member with access to our CI job to fix the typo +Sounds good to me. +The New Field wizard is not included anymore +Thanks Gregory the patch works fine. +David can you comment +Committed the doc and the patch for drlvm index patch at revision I guess we also need a definition of this component in the developers guide and were done +Doing a codereview I found a bug in Buffer task +Thanks very much for the contribution Maxim! Another language translation is always welcome +Thus the requirement of Example requires a runtime +Otherwise you will be in the next update. +Let me test this evening and will commit if all is good w TOF +Will look into it +Trivial patch +I committed the fix to that same branch +Backported patches b and a to with revision . +Updated the patch +Im giving this a spin too +I tend to think this is best and has the least unintended consequencesAs far as the fingerprint generation what is that exactly? Why do we need to do that in id order +Were convinced for unit tests results but were not sure for checkstyle and pmd results as in that case there is no way to know which rules has been used and if projects are comparable +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +I forgot to add that the issue particularly nasty because it will first fail silently +ve just committed this +There have been some caching changes in recent changes and I think maybe these will fix this as well as some other database issues weve been having +Please find attached a patch which excludes as a resource from any java goalI also added an instruction to the clean goal in +bq +Ill post the changessuggestions in the review +Perhaps we should move to +The is in reference to the fact that many are referenced needlessly inside in the messaging code to match everyone up with a Partition ID +Thanks the proposed fix helped! +Nicolas Thanks +Applied in svn revision +At some point in the Nutch code it isnt doing something the same way as for Chinese and Japanese +Partial fix by normalizing the content of the NOTICE file in repo revision into LICENSE and moved COPYRIGHT into NOTICE at repo revision rPlease verify the fix +I think this is because the is not being locked during update +BINRPM differs only in mdsums and dated fields binjboss differs in JDK items bit zip Server jbosssystemserver probably differs because of different jvm used for compilation client client clientorgjacorbutil client client client client clientorgjbossejbmdbinflow client client client client clientorgjbosssystemserver client client client client client docsexamplesvaria! by difference in jvm used for compilation by javap p command the classes are the differences except domj have been verified and identified as due to using different JDK levels a new issue has been created to address this in a future release. +Issues with a particular vendors distro should be redirected to the vendor in question + rev +What about the version in the? Shouldnt it be deprecated +For consistency should also call passed +Robin Putting my Apache Hat on I know how easy github makes collaboration however it would be nice to keep development inside of our project so until Apache supports for git rw access I was only wondering whether an svn branch would provide any benefit +Address timeout and issue in patch +I think there is more work to do than just apply the patch of September +I will test this and then report back +Revision unit tests and created. +update guavas version to r +So is there any chance to have this patch integrated into one of the next releases? Christian not every Hibernate user concerned about this bug is able to provide a patch let alone a robust fix +Ok so how would use of an interface help here? Or I mean have you been able to test the hypothesis that this would work better +I like Zhijies proposal of tags it is the most general purpose and is no harder or easier than the other approaches +When there is awsdl request irrespective of whether use flag is set or not what you will see on the browser might be wrong orwsdl generation might fail +Adds the new Mail binding +Perhaps you could add a test listener to print out a list of all tests that have a running cache manager at the end of a test run + as a committer needs to review and apply the patch +It changed a lot already +Hopefully this will work +Since this time the channels have been working well +Your comments have helped highlight critical issues that would haverisen in the development processMy proposal to handle the issues pointed out are as follows +Attachment has been added with description a list of files with bad CR cleanupsrc has been added with description List of files cleaned cleanup has been added with description List of files YET TO BE cleaned up source and filesCreated two attachments list of files which HAVE BEEN cleaned up list of files YET TO BE cleaned think this is no longer relevant +I will pay more attention on the document updatingI just Updated the wiki page now the issue can be closed. +patch looks simple enough and looks like it does what we want +No patch unscheduling. +Thanks for pointing this out +we are not able to apply this you please send me patch applied jar s urgent for me +It looks like the error is triggered while attempting to print out a that occurs because of some problem rendering the not rendering properly +Patch removing all nocommits +All Im suggesting is that as we alter job submission here we should keep in mind where wed like to go +To really fix this issue I will upgrade to use +I set that class in the HSCIF bean definition +Here is a patch +Closing jira. +has no meaning in the spec except for when placed on the bundle class path +Ill update the Javabeans Cartridge SNAPSHOT to the latest when I get a Kev +Same as last patch except cleared up the findbugs issue +I also have upgraded to Fedora Core since September with no changes in the resultsI wish I could give you more information +And this patch wont apply to trunk +This was committed yesterday +And theres a reason why we dont support TTLTTI or any other cache features thats outside the cache abstraction scope you need to set this onto the cache provider directly. +Why is this not useful +If you think they are critical for CR please them +The attached patch fixes this +the edited bean points straight to its record +I cant reproduce this issue with latest builds +There will also be a patch for in side +The syntax should be completely to new directives that are added etc +contains source code and class files of the additional synthetic test which demontstrates that is thrown instead of when monitorexit is absent +If buffering is turned on imcomplete data is downloaded to the file +Closing old resolved issues +User per above comments. +Ill open a separate JIRA to fix that before we commit this one +Paul usually handles build related issues +Tsz When I tried it didnt work for me +Marking resolved merged to with revision +have not been able to reproduce in JBDS Beta +Also it allow me to call the method without changing the scripts + +A comment from manstis This will be confusing for the business users +Please note this requires a small configuration change just move the authentication property from to and it should be good to drop is a merge of the and the redirection fixes +It is a very disk intensive test +Thanks! Any way a fix would make it into the next release +But I now have it under Project Reports instead +According to your description above I suspect it was more likely a JDWP problem +FYI The blocker is now resolved thanks to Roshan +as always kurt +I guess we need to bump up the parent pom version to accommodate this change? +Have committed this. +Thanks for the patch +i have found h set up tear down a bit undeterministic at times as well though + to accept years bigger than Author gcrawfordDate Thursday May PMMessage Client date overly strict restricts years to max digitsThanks to Modified myfacestrinidadtrunksrcmainjavascriptadfjsLibs +doing some cleanup +Submitted patch for review +Hooray +a trivial element is missing I guess +Thanks for the pointer +No news on that +I cant tell if this is a Spring bug an bug or a tomcat bug +Otherwise revert to using encoding in config file and if not specified try to obtain appropriate encoding for locale and finally revert to a default +It basically ends up being a snapshot of the tablets files that prevents garbage collection of those files +The next step would then be a dedicated DTD parser though AFAIUI there arent too many standalone DTD parsers +Attaching a fix for the optional scenario. +Is this a website issue? Im not sure what needs to be done with this +So Ive been trying to connect using my SSH key +Well thats the thing there is no ejb client in the test its an HTTP test +Hi AllThanks for your quick finally found which one is disturbing using Groovy + has this same affliction +Planning to work on the porting to trunk for now +Just a quick look at the patch it looks good to me but I really would like to see a unit test that can reproduce the issue +The file is also being used and so is one of the files used by the application +Patch for Reduced Parser with Junit Test +I will try to do it tomorrow +Patch added +I experience the same problemLike Vikash I am wondering whether there is a nicer way to handle this +Ive released +Please review +The on entry and on exit scripts where not parsed for subprocess and thus not executed properly. +Closing this issue if more needs to be done maybe it would be better to open an new issue this one has gotten rather large +Is this a zk conn per thead only for perf code Or do the producersconsumers do this +Oops I meant line endingsSome are CRLF some are CR +Note upgrade to slfj included +bulk close of all resolved issues. +is specified along with table names after which load balancer gets to balance all tables +Note that Im marking this as fixed in but its not a issue so Im changing to not a bug against. +closed with release +Use only with IBM JAVA SDK +Nothing in that directory had any additional information +is fair we can look at it in the next iterationIn all likelihood this got resolved after +Many memory changes went in +Fixed with on the latest code +Was resolved by mvn install for root of problem was in using classes from another modul +Im confused +How should we handle exceptions? Is this even possible +Patch committed +Otherwise I cannot generate a hint +The real issue is if you dont want the client to ever have the password how do you intend to approach authenticating the client at all? Have you had a look at the comments in the reference guide regarding using CAS to authenticate a web service over BASIC authentication for some ideas +That would be fine +IMHO it is good practice to stick on a casing that works on all target platform instead of implementing workarounds for badly written program +it was not my intention to unearth quite this large a task +I put one together and attached it to this JIRA +It outputs sizes +Yes looks tests started failing due to this +JacquesI am not going to add ASF and revision tags to the Docbook content files at this time +I did the same test on the last comment using the correct configuration file and it works ok! The line that creates the Cars node is mandatory +Kindly try out the attached war and see if it works in ur environment +This needs to be ported to as well to keep backward compatibility +First the is put in a folder and secondly maven libraries that are marked as packaging jar or ejb gets put in a jar file of the same name +But thats a minor issue +looking good patch applied in r +rtested and committed on ubuntu +If it its not inside a Form it could fallback to a normal +use aj use ajloadBundle of course +I still think it should be an IT test spinning up two clusters etc +UI has been changed +should clear its checksum maps in init before adding new ones +where is that example coming from? +First of all the goal of this patch +Here is a new patch that incorporates Sanjays first pass of review comments +I was wrong +One is the which it uses for HOD specific temporary work +Unable to reproduce the NPE as per previous comments but fixed line numbers on error messages +Stack trace +Added this case to the switch and broke it out of isempty +THIS NEEDS TO BE FIXED ASAPI changed the priority to BLOCKER the highest I could selectI cant stress enough that this is a SEVERE problem requiring a real fix. +Patches were committed at the revision r +Max Ill describe the changes here when completed +We will not create a fork of AJC. +Thanks for the samples +Regression test addedVladimir please verify +on the changesI think we should compare the performance gain of this patch on a r node clusterI suspect gains will be more visible on larger than just node clusters and it is interesting to have measurementsparticularly to get an idea on what further fine graining could bring +Added to the repository auf and tried to install a pool +Nice feature. +It would also be easy to handle schemes +Given that its very cumbersome to implement a special resolveXXX method for each ref type based on an +I ran my example using Xms and Xmx at MB the application gets stuck +so version work only with java and higher +bq +I wonder where it went +You can do that when libraries are packaged into web application only +No test added since it is impossible to make legacy compile in new code +Anyway should not assume that this should work anyway? Do you mean that projects generated with JSF will not work with Hibernate +I think we should open until theres at least a workaround in place +Can you post a patch for it +Here is a patch that breaks out load balancer and that starts in on writing unit test for the balancer code +Closed. +This is a only patch +If this lib is supposed to be placed in any standard lib folder lib commonlib client server +Thank youBest regardsSpark by Spark. +I would like to defer this until we have someone actually try to use this on zOS +Anyway Ill let it go and just roll my eyes at all the hacks and duplication that building an entirely Ajax UI using pure JSON responses entails +jar in a +whats the point of using contentEquals over equalsformatting cassandra puts curly braces on the next line down no spaces between there appears to be indenting offset order to check in the patch is there anything I need to do such as add Daves changes or will the contributors take care of the rest +Committed to JBESBGACP with revision . +When the application says it can not read something Bob has to spend a lot of time figuring out what it can not read +should be extremely lightweight and thus only requiresI think if possible there should be a option on the multicast and splitter type to reference a +The attached patch makes each instance have its own copy of the for the SYSFUN functions +It got in formally inside from what I understandJitendra Can you confirm from memory? +The test case does but Id be surprised if annotations had anything to do with the bug +Reviewed Monitoring guide for release update version number and app server version numbers. +Some features now have their sources yes +In order to search through the time partition the index of multiple spill files need to be merged to provide a linear view of the time line +Since this is just a doc patch Im skipping review boardThe diff consists of complete files in the directory +Updating to a plugin would be great +Ive been thinking about this +Is this now addressed +Bump down in priority and target +Heres a version that addresses these issues +Please provide your clean management server log and db dump for this +So just a quick update Im double confirmed that if I change to simply return null then everything works ve tried making static final and that didnt seem to help +Extraneous changes in the previous patch file have been removed. +If you can outline the dependencies version changes in this ticket and attach any POM code samples Ill see what I can do Strong +What version of Maven and the Site Plugin are you using +Ive reproduced the issue +If not let me know +Moved leaving ring code to a separate helper function and call it right away if rangesMM is empty +This is a corrected file +SorryI meant to add this comment to +Can do that in new JIRAbq +This deviation is to be documented +wombat +bulk close for release +Regression test checked into revision . +Feel free to start the wiki docs when you get a chanceCheersJon +Can you try installing the server in a folder which doesnt have a space in its path? Currently you have it in EProgram Files +Tom please leave this alone for now +Ok so if there arent objections before tomorrow Id move this in and mark as improvement +I also found a technical problem sorryThe line protected static long responseLifetime L +These blockers and critical issues are resolved but not verified +Patch against r of to allow formatting the timestamp only +Any suggestions? Does it matter +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Verified in +You may provide a proper fix if you want. +That can help us to buildtest every thing together also +We do that in the file channel too if we hit an issue that violates the assumptions of the channel we cause the channel to die rather than skipping some of the data in the channel +Tomcat is right the embeddeed servlet container is wrong +Stian ok if you could please attach an example EJB deployment that we can call I will have a closer look +Committed at subversion revision +Feel free to use modify or tell me to fix it if something looks broken +Please verify these changes and report any issues or close the ticket +I guess one situation where that wouldnt work is if someone who didnt have the tools to compile a thrift binary was working on the lib and needed the fixtures +does provide these and they are quite useful +I will take a look +You have to weigh the gains which seem minimal at best here IMOIm flexible on what we do for this issue overall but Ive stopped thinking about Solr as a webapp like I said I dont see the gains and Ive seen how that thought process has held back a good out of the box experience +But can you do that in another JIRA so that we can use this one for the split effort in future? Thanks +The examples we were running use the and theyre still using normal +Committed to and trunk +L. +if the project at eclipse is the group id should be your suggestion would be to change our groupId to while java packages still remain and our artifactIds all start with mmm? I think only changing groupIds is an acceptable amount of refactoring work +Also can you propose your thoughts on which layers within Airavata or at API level should employ AMQP +Let me know if you would like a patch to Solr back compat is not considered the same as Solr to Solr back compat +externalchildidsthis column currently unused for action and straightforward to use it +Great so all those access properties can be scrapped along with the rest of the builtin stuff +Didnt make it in time +Sorry i should have tested on clean machine +Scheduling this for so we would have time to investigate as we prepare for +Thanks Thejas for review! +Verified in the WFK release notes +I went with Seans approach for readers and writers +Ill add a test for this as part of +Are you saying you believe this is fixed in the master line? This same code appears in as well +I dont seem to have permissions to reopen this issue +Forgot to mention that I committed the patch in trunk +Applying second patch +Attachment has been added with description Example servlet which retrieves images +Patch submitted built against trunk and unit test included +Closing resolved issues. +Here is the fix Brian and I worked on +this is pretty cool! we can generate pdfs by using doxia converter to go from confluence to latex +What kind of breaks did it broughtIt should be safer to use preemptive shutdown so that blueprint bundles shutdown can be properly ordered wrt to bundle service dependenciesAlso note that Ive fixed two other related issues in trunk so youd better try with a recent snapshot and see it it still cause problems before reverting the default value +was committed today +the Mb archetype is now available from relevant anymore +This issue now works for me +This is a potential patch file to be applied in in current trunkSorry for pasting the file I am not used to JIRA +Doh +with +Still docs and UTF tests are missing + for the to trunk. +Slava I guess you meant restriction of JDT not JBT +Looking at the patch now +Mahadev I see that the launch container script has the same order in classpath when I run a wordcount using hadoop script +I have the beginnings of a script that I would be happy to donate to the effort +I believe it would also help others who might come across the bug to understand why its invalid +release notes with detailed jira is are updated. +Fixed classpath problems with S build. +However Ive so far hesitated to do this because I dont like the idea of restarting the module all the time +Issue could be closed +Another minor editorial remark is we tend to omit braces in if statements +Now all will be sorted higher than regular proposals +These blockers and critical issues are resolved but not verified +We have encountered this exact issue in +Before you submit the patch please change the issue title in too +In our organization I have encountered very few cases where they needed to add custom jars to classpath +Added a check to make sure the value is between and +This is a patch for +has already been released. +Opens up further optimization possibility +Fix +Fixed for the nextIssue is closed. +we no longer register the Maven Locator to Dali and instantiate it directly to locate the +pass through hadoop qa +Actually I simply want a solution +After sleeping one more night I added code to not even put the termsinto the PQ when they are not competitive +it also repros in my environment on ibmi am going to work on this to see what point in the code it breaks thinking it must be some wierd jvm bug in comparison +good work +you wont ever be able to send that acknowledge back. +As Ive written in I tried it with the same version underneath with plugin and it was at the same speed as with the bundled versionWe suspected that the degradation has to do with the forking of the VM +Can you factor it out into a method +The last time I checked only part of the Java API was Serializable +I dont know if it is of any help but the error first occurred in the beta release the beta release works fine +But if the partition location is masked the existing test case will not test if the issue of this jira has been fixed the test case would need some changes +Paul Davis fixed this as part of the work to integrate support for +We will test as soon as is out. +Bulk close for release +Ive just committed this +I have raised where this condition has occurred using the release. +I have the tarballs for building with the old bash scripts there +Otherwise write the filter to HDFS and pass along the path in the +Patch including test +uploading the patchRegarding would check if this testcase can be added at this point or else would open a seperate jira for that as mentioned in comment earlier +Before this the index column was NULL permanentlyBut it is not correctly implemented as every indexed association seems to result in a cross reference table which is not necessary and produces unnessecary updates during insertPlease fix this +bulk move of all unresolved issues from to +So there is currently no tooling for this feature + +Andrey do you have a way of testing builds in that env +ant test javadoc javac warnings and findbugs were run manually +The functionality has been turned off yes due to excessive spam +Some minor cleanups of comments and a minor refactoring +last time i tried to apply I had issues with it +Actually it may make sense to change return to next in alter in the attached patch likewise for OWNER +Lets revisit and see if we can get it to run in the local time zone once is fixedCommitted revision . +Verified by Sian. +This error still arises +As you can see in the Issue Links this Jira is already related to +Were already on +Changes have appeared in Latest Alpha Manuals. +The project will accommodate additional parameter types for. +Yes sure +When validating a username token with a digest password Wssj doesnt give the control back +Revision +I will submit this to the JAXP Expert Group for discussion +bulk defer of items from to +The newly attached patch has full path to the file in order to distinguish different files with the same file names +Its possible you may have an older version of the Camera plugin +The compensates for this mismatch in sybase +just to ensure the does not play in here +new librecords usage for configuration variables and statistics vs old way +I agree with your comments +There is something wrong with the geowebcache SNAPSHOT artifacts +The problem is related to using of minus characterThe latest version of struts restricts name pattern access because of security concernMinus character doesnt impact security +Attendees included Aaron Myers Eli Collins Hairong Kuang Jitendra Pandey Hari Mankude Pritam Damania Suresh Srinivas Sanjay Radia Tomasz Nykiel Todd LipconThe following topics were discussedClientside failover failover Currently configuration based failover is available +This bug report is about sorting +Looks like this should be resolved in the JBTJBDS beta drop +I dont think we are far of but I really would like the CLA part cleared +Attached is the error screenshot +Can you provide an example +orA b release had been done by Jody with the errors fixed. +patch from agreed upon on RB +I looked at this patch +A while back before we were aware of this feature it made for some interesting troubleshooting like my server is not running but my client still works +I read blogs a chat with people +I also ran the regression test with the proposed fix and it also worked +I think we can now remove asm dependency from hive build altogether then +is a tool to build applications and server data its not there to impose a restricted view of the world where only certain best practices are supported +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Committing also because it logs when scenario arises so we have some chance confirming indeedi ts because of major compaction that results in no output. +The workaround is easy to implement +Aight then Im updating this issue to remove the fix for and closingAnd you are right about the build +It will reduce some confusion and also is not a huge overhead to release them in group +Error messages need to be translated? I hope you mean error OGC codes? The actual error message comes from the internal exceptions which are all over the place and are not meant to be translatable +however it breaks when using the jhighlight decorations +for the changes +svn ci m Adding Vincent Massols pom file from +The analogy about an application that needs Java to run doesnt hold for this situation +with a asList methodBut for now this is already quite fine imho +as I saw that SGF still builds against spring framework I additionally checked whether it works with which turned out to be YES it works as well +its a windows only bug +Bryan did this make in before you became a committer +Both references are deemed compatible by the frameworkThe reason why this use case could be improved is that both bundles A and B can see the package and the service object is supposed to implement the interface +Closing this issue. +Test case has been committed to SVN +Since was added to the srcmainbundles maybe also should be added to that file +That might get somewhat complicated though +I looked in and didnt see any +Sorry for the inconvenience +Clearly something to fix urgently +Okay I did a build all and looks like the samples were built +I shall fix these formatting issues in a patch +Im put my name on the also having this problem list +Uploaded patch for trunk +Stuart this may be a bug in JIRA +Attaching rev of the functional spec for this feature +Here is the torrent file for our We have been able to disable the preprocessing and get away with Sitemesh and related memory leaks in grails +Very strange there must be some abstraction layers in Java causing significant overhead +reviewed and committed +committed to trunk and branch. +The idea is to allow users to use these to generate permissions or check permissions +At first glance the patch looks like it should solve the problemI am marking the JIRA Unassigned +test test failure is occurring on all platforms with RC +Are you definitely testing? Description and asset is now on a separate tab on the asset editor screen +Since we redid the admin UI. +is there some compatibility problem with Vista bit +Ive verified that the target is indeed encountering the same unexpected exception however is not showing the test as a failurein the junit test reports +Yoon linkid +Ill see if I can mimic that +Thanks. +I have attached my out put file here +Eugene Ive tried the dev build as well +Is that the case? If so we should probably change the summary +somehow i thought we were only outputting key rates +Its an experimental feature seems ok to me alterTargeting for +Im going to provide a patch with this fix for +Ed the code should be the same as you posted in your earlier comment +The only concern is we mark it unsplitable +Thanks Anurag +Mind youthis code is quite ugly +Ive applied this and will commit shortly +This version is compatible with the initial patch from +Thanks Mark. +The frontend has been updated to comply with both of the above lists except for xmlrpcKey +snapshot with your patch thanks for the feedback. +I have to check that at my linux box at are definitly right that it isnt wise to modify generated sources even so it seams someone could not resist to do so +I see your point about easing the loading of configuration files onthe fly and I agree +Does that mean the the issue is decided and that we are not going to reduce the number of jars? If not and we do want to change it we should do it sooner rather then later because it will be a big refactor and disrupt development again +as a quick feature was implemented as part of +No +Also the long sentence starting with Otherwise might better be broken into separate sentences perhaps the UNIX and Windows info can be in separate paragraphsOn UNIX systems you need to use the dot or source command to ensure that the script is run in the calling shells environment +Have more tests to add still +As you can see the build succeeds. +I think you can also find an answer already to it using the Search of the archive but you can also jsut that question to the mailing listThanks all old Resolved issues +just committed this +I would not like this having in mind that sometimes logging configuration is changed on a hot systemThere is a logger hierarhcy available which defines that the logger gets the configuration of a if there is nothing available for +Unfortunately we decided to package Seam in WFK differently to split examples from bin distributions +To top it all off Im not convinced that its really that useful +Setting to +Focus listeners seem to mess with some browsers +that doesnt necessarily mean that the Node gets invalid does iti dont mind if the tree tree gets invalid but i think we should fix it for the Node +As MR users may use a dummy class and the enum are added as wellUNINITIALIZEDMEMORYVALUE was changed to long +This is backwards +Hi JacopoThanks very much for your feedback +weird one +It is not necessary to create an object +Can you please attach a testcase? With HTML output before applying your patch and after applying it? This will help us to qualify this issue +The new and updated classes and the in SVN revision +This is part of the reason the tests fail +I believe this is fixed and now merged onto +Committed to branch and to TRUNK + +Approximate peak is Mb on a gigabit networkh +on commit +Our implementation of Node holds the actual wc node along with its children and our implementation of soapChildren +Any ideas? For the write case the leader is updating the request zxid perhaps we need to do that here +Spring continues to amaze meIll put it on my list to add information about this class to the reference document +This fix passed both and local instance tests with both and non secure Hadoop +I think the proposed API changes will address my concerns +Reviewed the changes +created for these new bugs +I can define css style class for all links in +Closing issue Oozie is released +Please include it it makes really +I think rev fixes the problem but Id appreciate a review +I think well go with the fix I committed but we need a test. +The filter join query does not impact relevance thoughThere is going to be a large set of code changes to this ticket probably next week +Imagine if in the future we want to add a flag so that replication isnt just TRUE or FALSE but also LAGGED with a parameter for how long the lag should be? For the maybe we need to extend it to allow multiple once we have the support +I believe Erlang RA has the fix required for closing this +We can assume this value if the signature attribute is not specified +Try adding the following option to the JAVAOPTS variable in the application server file trueThe default value of this attribute changed from true to false with the release of Java + +It is possible to add at least a new MAY AT to an existing OC +Key initiated from the server is still a todo. +Reopening you please confirm the error Im seeing is related to this issue? Thanks +to close in. +Closing this as said above. +It is not trivial to fix since filter assumes use of which is different from how works +In case its not instantiated yet we dont +Im even invoking methods on some of the collaborating services to see if its some kind of second order effect +Fantastic Thanks guys +had a similar side issue related to cannot stop something +thanks +workaround clear any cookies from the rails server in your browser and reload the URL +Looks like a leaking abstraction but I can live with that. +Trunkjobstruts +But please reopen if you still this issue in AS Final +Of course I understand it will be outdated +Catalag derby have been using a pattern where the derby dbs are not protected by usernamepassword +It works as expected we can know get more information about the to workaround Oracle poor implementationI have tested the ability to specify in configuration a custom implementation for theFind attached my own implementation still focused on Oracle but with few more detailsBut analysing more the source code I think it could worth to extend the mechanism to every location of handlingenlistResource is not the only place where we can get meaningless Oracle XA Exception butdelistResourseRecovererrecoversuspendgetMessagecommitrollback +for appTypes it can not handle appsapplicationTypestypeapplicationTypestype +Only looked over the code in the patch briefly but it looks very very promising +We wont respond to deployment errors by denying that the published artifact matches the workspace for deployment from commandline and other workspaces we would need to decide what exactly to show in the server view +SVN didnt add this to the patch because I ran it from and the enum is on +This issue is incoherent and incompatible with prior webservices requests +is that normalYou are asking the wrong question +The best way to proceed is to provide a unit test that anyone can run +The Foundation does not add the notice the individual author is required to add it +But this is what solved it before so I think were good to goThis passes mvn verify and will fix the problem for all profilesbuilds +Keep watchingDo assure can finish much earlier than the end of this year +How about such suggestion +Thanks NamitPlease take a look at the when you have time +and thanks for the enhancement it is even better to let this method react if the service is registered yet + +updated the version since it was pointing to a wrong one. +are not in committed to branch +Then Giraph devs get blamedMaking the thread smarter doesnt sound like it buys us anything as it still needs to read progress data on regular intervals and do work to measure and evaluate it +QE would you please confirm whether this is a regression introduced with +batch transition to closed remaining resolved bugs +Test failures seem unrelated the findbugs warnings is due to a +However it is likely that we want this to be the same behavior that we will enforce for the returned by table functions +Using this model opens the door to a lot of features such as easier support of the alternate auth methods being proposed IP failover token support for and for servers over a single connection etcSanjay Ill contact you offline to bring you up to speed on this lengthy jira +the latest patch looks good to meUnless anyone has further comments Im going to go ahead and commit this later today +With the new serializer both interpretive and XSLTC use the same serializer +Including a blurb on with +I think I will reopen the issue as soon as I am ready with a test case +Canceling this patch as it is over a year old +The JBTIS target platform already provides graphiti +Few changes missed in. +can you do a simple google docs diagram showing me your ideas on this? You and Burr are the masterminds of this concept and I want to make sure I understand it fully as it plays out in your mind +JAXRPC tools issues may be fixed by community contribution +seems like we have an related failure in +Please if you can respond to me requests for more information. +Let me know when its ok to move svn +So Im not sure what happened +please move back if they will be worked on asap for +Sorry but I still dont reproduce all work smoothly on trunk demo as locally +I applied and This broke the build on both Linux and Windows +PDF supports algorithm for storing images so screenshots of text should not increase file size dramatically +This wiki page is more like next step that I will propose +Youre right both hadoop line pull in which we use transitively +I will mark this as a issue so that we make sure we deprecate it at or before that time +Hi did team fix this bug in JBOSS AS final +Other then that the patch looks good to me +It looks like the most recent bpel bits havent been pulled in to this build +cache invalidation at BMT insertion time as shown in this patch actually wont work a read could still pull the old version into the cache again since BMT contents are not live to readersyoull need to add nodetool invalidate instead +ideally should be fixed since may indicate bug that will confuse users +Committed revision +Then maybe you should at least update the documentation for plugins accordingly so that nobody else runs into that trap. +has been filed +to an HTTP error code which can be intercepted on the and examples to follow +Yes was agreeing that should always be the outcome +In both cases the listener reconnected when I restarted the broker and continued to receive messages +Let me know if there are other dependencies +Since all sub tasks are completed I fix this issue. +bulk defer of featurestaskswishes from to +Takk +Could not create patch so I added the java file +Verified at r +Per an element method has been added + When we get to this spot the Transaction is complete +This would be nice to have as a game like app but is not critical +And I cant inject it in the module descriptor of the dependency because if several modules depends on bar with a different classifier it will lead to unpredictable results +Andreas As Amila has explained currently if we use the axis local transport within Synapses proxy service it fails +add a very simple handlers take a geom factory as a constructor parameter and use it to build geometries now +patch does not deserialize the ack to a heartbeat correctly +Please check that it was applied as expected +Finally for complex situations we might want to consider allowing multiple configurations to live for different emails +Attach the patch for this support. +The instance that declared the queue will hold it in the exclusiveQueues list +And adding possibly maybe methods might become an issue as well +I dont completely understand it but this could also be a solution for our problem +Can you attach the file again and use an unique name so we are sure we pick the right attachment +Currently this can be more or less achieved by in conjunction with pomFile param of but this is very ugly approachBtw the same for +Marking all current RESOLUTION LATER issues to be against the Nightly Build +OK You are correct about that +never rely on queue browsing to accurately reflect what is available for consumption +of unclosed issues. +High level functions are staying into jthread library low level functions are moving to hythread libraryIts necessary steps to merge VMthread and structures +touching this jira so it shows up in your query +bq +This version pulls inDecision of pulling in is pending +click on item +If theres something that shouldnt be there I prefer to show it to the user +Readme doc has been updated. +Added some logging for membership changes +I will upgrade this to critical and well look into it as soon as possible +Hey adam think your installer program is capable of choosing firewall information at installation timeIf not we can set up a preference page or something +Brian I am still able to reproduce the issue +I appreciate your commentsThanks believe this issue has been completed +Hari for starters let us keep things simple Allow the specification of more than one end points in a order +Any feedback on this? Its in the probability of occurrence very low category but increases the delay so the probability so I would prefer to commit the two altogetherA few lines after the one mentioned in the description there is a similar pattern +Charles try applying Doug Steigerwalds latest patch I have not tested it but it does apply without I am new to Solr so forgive me if what I say doesnt make sense +I will take a stab at integrating code I got from Alan to get rid of regular expression and then retest the people are not relying heavily on the regex in the load +changed VARCHAR to BIGINT for LONGVAL column in STEPEXECUTIONCONTEXT closed as resolved and released +class javaioSerializable fast version getstatic +done +already had a description +I have retested with the latest jbossastrunk and am not seeing the NPE +Iam happy to help smooth any rough edgesMike if you are still around any chance you can put up your input generation code at a minimum +New JIRA tracking that +Plateforme de r blocked by +This isnt a bug +Yeah this is possible +Kelvin yes that was my intentionThanks +This is my patched srcCore which resolves the problem +I have just made a release with Maven using the SVN trunk version of the changes plugin +I actually maintain those classes but wanted justins viewpoint as so much of jdbc depends on them +Thanks for reminding Mayank I will update the patch with the changes + to trunk. +trimming summary for this clone +Ill try again and make sure it works this time new insights Jelmer? for the time being +Im not sure if this is a groovyEclipse problem or GGTS problem +xml files are not of the same type +Because both the aforementioned exceptions are checked exceptions removing them may cause a compiler error for newly recompiled code +Verified int JBT +Will be handled as part of the session producer +But Ive already included a temporary fix that should solve the problem +Good ideaHeres the patch against p +If so the directories are attempted to be created first any failures will result in the call had the intention of making this a configurable switch although am not overly familiar with how to supply configuration to wagon providers +Great thanks for the patch! Looks pretty good to me Ive applied it to trunk +The code path for was a little bit more ugly to get working than with method invocations +Even if we can fix this specific problem that does not mean that literals will work as expected +Correction This is NOT the case for all other batch files in the bin directory fo rexample jruby and gem work OK +From your problem description this constraint seems to be satisfied you said that the input has a of textcsv +The use case a tree where expansion is posted back to the server you expand with a different targetThis really doesnt make a lot of senseInstead what we probably want to do is this Create an inheritable attribute selectionTarget that can be defined on the tree and overridden +Committed patch d to trunk with revision Patch revision d removed some unused imports and a blank line +updated patch fixing the other postings readers +No test case I close this issue. +GEP will tell the server what ports to use. +Im the original author of +the maven netbeans integration is discontinued for a long time just closing the issues. +I also note the following curiosityo Now is a vacuous extension ofThis makes me wonder whether we could rework the graph further and eliminate all of the Basic from the public javadoc leaving us with just the Base versions +Will testtomorrow morning unless you have time to test it before +to allow any further contributions. +mrepositorycomjcraftjsch SFTPuserpasswordtmpYou may want to add this code somewhere as an example although it doesnt need to be exactly in that place. +Further information +fixed to not consume in its ctortwo more tokenstreams in solr to fix and we are good and +I am upmerging and will have an updated patch soon +How has this been resolved? I am getting this issue and dont understand why the tool is looking for my schemas two directories back +the patch to Hadoop +Shall I attach a screenshot for the new metrics +Apologies to clarify I meant that was not a hard requirement for an Apache release +On the remote the file never hope this helps +Problem is that the file you provide has many changes to whitespace and its too hard for me to make out the real changes +Attaching the patch that contains the fixes for the issues related to +sorry for the noise again per reporter comment. +In order to maintain binary compatibility for the release Im adding a new constructor the other will simply be deprecated +Finally we really dont want to change the navigator interface unless we absolutely have to since that would break a lot of third party code +Committed to branch and trunk thanks for the review Stack! +A couple of days is fine I think I would rather wait till after is frozenbranched to apply a patch of this impact +On the other hand we should discuss in the EG what the expectations are for the usability of Node objects after a failed save operationagreed +Not sure how to reproduce this bugBrian could you comment more details +Update the patch to fix unit tests +Id say lets leave it for the next release +Are you really replacing it with a pie chart +A more generic implementation of the proposed insertion operator +Test failures are not related +Some are related to packaging and some related to code +A new version doesnt need to be released just because a new castor is released +QA Verified the tomcat connector metrics are now showing up. +I added an RPM spec file that can be used to create an RPM of the project on systems that employ that packaging mechanism +will also use trove +Can you post the full stracktrace +This way a transaction writes to files only +really? like whati think managing acl in content is the about the best solution that i can come up with +Adrian rather than facility I would use a productBut as I said above the requirements mentioned in the description of this jura dont seem complete enough +I have downloaded BRMS will go through this and see if there are other issues to be addressed +You guys might wanna the issue next time +wondering if you had a chance to take a look at the second round of logosLet me know if youd like us to further explore one or more of the versions or if youd like to see some new directions +Ah sorry I guess that you talk about the message displayed +OK +You accomplish this by using the additivity attribute + +Just now chated with Max +Please verifys somewhat thatcreate is defined in but otherwise looks good +If you need the value of thisObject inside the closure then use this which is mapped to we agreed that owner and delegate shouldnt be accessable from inside the closure +BTW we do have URI template variable tests in a separate test fixture. +Results for the remainder of my testingAll my indexes use the compound file formatSelecting random combinations of the test indexes I am able to successfully merge using both method signatures as long as the sum of the inputs is less than GFor all attempts to merge all the indexes using either method signature I get the conditionWhat is the next step +For ConvenienceI have included an executable Jar File which will allow anyone to run the following testthe format iseach of the lines represent Sheet Time Worksheet Save TimeWorksheet save TimeFile Rows Total Time Time Per Row used Mem +May still require a merge to trunk +Integrated into trunk and +Regardless of that however I agree with Todd and Eli that it seems fine to ship without this change as it is a alpha release +The question to ask I guess is whether is important now that is out +In the code currently these two visitors are configured as followsrankingrankingThey are used in descending order so the will always run before +ArulCould you this as a patch? It will help expedite the process of my reviewing note that petclinic has been modified recently after you submitted these changes + adds additional for cp key +Im currently negotiating for some time to spend on overhauling all of since it really needs some love +First part committed +Added timer intermediate catch event to the Matthiasdid you set a timer duration for the intermediate time event? Ive noticed that the BPMN xml is only updated thus contains the event in case it is valid according to BPMN spec +Rolled back the pig upgrade +To get the screenshot you just need to Create a new Eclipse project using the wizard Java EE Web Project on the project Select the menu item Properties Click on Java Compiler Annotation Processing Factory Path on the got it but this content comes from a maven archetype outside JBTs control +Do you want to work on this taskWe can give you all the support for this activity if you want to +Is the configuration for both builds the same? Are you using a private local repository for you build and not? Are these historic times or can you point me at the two build configs +Verified fixed in ER build. +Committed patch revision +I think that not copying the payload would show a bigger gain than in this benchmark. +Fixed +INFOconsoleDRFA classpath UsersjoaDocumentsworkspacescompanyworkspacebin +Can we add some automatic profiling to CI +By the way heres a tip if you want to run the unit tests faster you can run just the fair schedulers unit test with ant Dtestcase test +I saw an offline suggestion to include the name of each file in the protocol the remote deploy servlet use +Hi Steve Ive tried to load your data using sqlloader however the ldr file does not contain the table definition itself so the load task fails with the following errorSQL Error during describe of table OSSTRATEGILINE object OSSTRATEGILINE does not existDo you also have the sql to create the table? I can try to reverse engineer it by looking at the file and the sld but that might create a potentially relevant column type difference in the db closing all resolved issue that have not gotten any more feedback or comment in the last month +More like the test environment isnt fully set up or needs some helper to mock a running cluster +Pach files are always welcome also if it take some time to be commitedHowever just create a patch file like below and attach it to this issue svn diff +In ab. +Thanks for finding this Ram. +dll in the file types? I think it should be the latter +Please fix this problem ASAP +Yep but I dont think svnssh is supporteed is it +So its completely up to the service implementation how to do the mappingIve changed the log messages of the clients using this code to not assume how this is done in the service so they are just outputting the bundle and the sub service name now as separate information. +One more usecase for this enhancement I have about modules to change from to since Id like to skip by releaseclean and releaserollback mojos + Is API? Its what the ant tasks use so the Maven plugin would mirror that +Well address these two together +Closing +but were definitely using beta at the moment ls l lib r tom tom Oct lib Ive got something wrong in the POM or some such +verified that a new build given to me by mladen fixes the bug. +This is fixed in svn head +Patch committed to trunk. +I will investigate more later today +Fixed +It works in windows and fail only in linux systemI will try to figure out this issue +My pleasureAt the same time I learn the inners of XW and WWI will continue to improve some unit tests +Because of that you will see no errors in the log file as the UI cannot log anything to the Solr logThe reason why the does not display those information might be because the Webapp context is restricted and does not provide the +Getting this error in the logs time and againIncreasing size of absolutely no context code configs anything I dont know how we could possibly help or even if its a Struts issue +bq +to commit that for M after cleanup and integration of test into CTF +But I am ok with the current name too + +If you want to vote on an issue do it on this one +Separating this into a different manager will make it easy to incorporate update policies and vs +If the issue initiator is not satisfied with that then feel free to reopen the issue +The log message should never appear +Peter I already explains that elements Groovy considers being the same is not reflected by the hashcode +Committed to trunk. +It makes code much cleaner and simplerThe only concern I can see here is calling vmidentifyeip second time on the same instruction pointer + +I dont really know what he did to fix it though +Please review this to and master with JBDS +Hi JasonI started working on it in is fixed +Looks good the get method looks like it could be reused in sysinfo to take care of among others although I dont know if adding the get permission would be a serious drawback +We need to backport after that +Lins changes have fixed Lin + for the patch +Attaching a diff for only +thanks +Another patch is duplication in the current svnPlease apply it to airavata an old patch issues was immediately addressed with commit . +Whilst it does refresh its own connection when message polling fails still carries on logging exceptions due to the lack of an attached to the Connection +This JIRA can then be dedicated to improvements to the current XML config. +But I guess would be a good addition it is mature well tested and has the Writables we need in addition to sparse and dense implementationsLets try it out +But anyway my proposal is to disallow the across volume rename for now +implements nearly all of your remarksI didnt change the code flow +This is not an actual compatibility issue +Thanks Erik for your feedback I will keep the jQuery documentation structure in mind as we continue with the analysis +Its NOT a reference to the PK that was deleted +The JSON peculiarities have bitten me in the past so Im in favor of this changeRead through the patch things look good to me +Added a new patch +to Todds suggestionThough current code logs errors in debug mode only +Hmm +does not suffer from this closing all issues that have been in resolved state for months or more without any feedback or update +through are now done also. +Docs improvement opened as +Depending on how the trunk job goes it would be nice to get a similar job established for nutchgora +Theres similar changes in other parts of the code as well +Not a big deal just at the moment however +Perhaps this could be improved by buffering entire responses in handler threads then having a separate thread that uses writes to asynchronously write queued responses +Ha sorry Dhirag I did not see your last comment +The primary idea of FI was that you have a directory and what is in there is running in your system +additional patch added +patch looks good +Very good points +Patch needs regenerating following the project split +We also need this feature +fixed in r +The patch given is showing some error. +Moved to Core module +Thats why its working there. +samples +Patch looks good +Why do you want to use the etc to provision your bundlesMaybe using the boot features would work better +I would check with Patrick as he was considering to work on it +or trunk and let problems bubble up but pushing it back into could cause surprises +You can keep the if you want to get rid of or you can remove it and check what is written to the logIf you still get the error it would help to see your file and also what is printed to +the patch for trunk to enable QA +Much like Maven SCM does for version control +Rerunning regressions +Committed to trunk +Emanuel AIUI from this isnt an profile service issue correct +There should be no difference between a direct invocation of a JSP versus rendering that JSP through Spring web MVC +console output its not that big +If you still want this patch to go in you are going to also need a patch that supports trunk eventually +srcjavaorgapachesolrschema +Thanks TimThe fix looks good +Ill accept it into the codeThanks +pushed +If a patch shows up before release well add it +It adds no value to the target audience the corporate developerBtw uses Prettify for syntax highlighting is only used for Velocity +I have described the probable cause in +Thanks Maarten! That works on my quick little test +based on a very old version +Resolving. +tried on trunk as of its working nowBTW I was trying to build it from the root of the project the command needs to be run from the maven submodule +Tapestry now defines a number of events to force a zone to update or to be informed of its progress. +Then the workaround message is simpler also +It uses JNDI +Thats why I had linked this jira to +Ill reopen if there is another sighting. +Daryn I will address most of the issues you mention when I get to next week. +No test included since its so trivial +As far as I know CMS has no code in the API headers so in actuality the use shouldnt really be affected by this to much +The FAA pointed out they were missing from Fuse Enterprise hence they raised the PROD issue to request them to be added. +Hi Rex There were some issues seen with informix db and hibernate +The transaction was commited and next transaction was rollbacked +after no further comment has been given. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +David wrote For this reason it makes much more sense to use the as the owner of a unique identification for the sessionI agree the LCC is the physical connection thus having it define the unique identification matches the definition +Supposedly the support for the will be available in the next Beta of the EJB Feature Pack +the fix is r the CP branch is earlier so it is probably no fixed in the CP branch + +Committed changes that addressed and in revision +See as duplicate +Fixed in SVN +see hibernate docsAnd the schemexport functionallity should be available through the dialog of hibernate code generation as an exporter with then utilize Run as + +new commit added +I added a component to JIRA for Distributed Build and edited the subject +Example code to trigger bug two ways +Tika should extract za gl ja ZA GL JA but I gotza g l ja ZA G L JA +Normallyto rebuild the site youd commit a trivial change toIve reviewed the setup and the only issue was a stale buildbot hook on the svn serverRunning svn up should have it fixed now. +You cannot handle both exceptions in Java AFAIK +Can you provide me with a debug log or vdb details to understand what may be happening +That is not acceptable scenario +pushed Released Closing +I tried to keep loading data to my cluster patched with v and killed regionserverdatanode randomly +It contains two launchers fail and pass where the pass launcher has configured classpath with the fixed class +The fix for is for filesjavaengineorgapachederbyimpljdbcjavaengineorgapachederbyimpljdbcjavaengineorgapachederbyimpljdbc at subversion revision +so this will not get fixed in ver +I think you should go ahead and commit it applied. +No objection here on either point. +For some reason I didnt recall DFSMR having both slaves AND includes but youre absolutely right. +Sure done +Yonik definitely + +When we design a FE for configuring then we can decide what operations we are going to use +The example for plinegraph is still missing the key attribute on the pdata elements +Removes the special exception for +This has nothing to do with jspc plugin +Thanks for pointing +Ah youre right about missing setters just being ignoredI guess the best option would be to make log a warning for all such configuration parameters +That is a good insight Jun +Please verify buildtrunk +For Intro purposes this bug really just needs someone to see if the problem remains and get it to stop raising a Java exception +If you can still reproduce it after a clean build and your ENV looks ok please file a bug or contact us on freenode IRC in jruby. +Fixed by the committed change. +complicated not sure where to look for modified content for instance +You prefixed all the component model names with Editable which makes them long +this should be rolledback or fixed so the runtime and server parts are in sync on what jars is being used +Updated patch includes a quick test +Ok this patch has +Yes this config does help +I no longer get the NPE and scripts are required out of the classpath as expected +I am not certain whether the bug applies to other build management programs but debhelper sits on top of make and exhibits the problem from aboveI have a branch staged up and potentially ready to commit that adds a +Include modification in patch +And everything could be shared between all of process instances +The overall tag length is short stillAnd then we have a byte indicating the tag type +host performance reporting is not easily collected to ganglia +Ah I see that bc the autocomplete does not distinguish between value and method bindings then all these completions are offered +We can probably help Mark Proctor fix the Drools build. +Here you have a friendly version of the plugin with a very simple support for adding a stylesheet to the generated HTML +In the future have more respect +Proper implementation must break such cycles resolving plugins and their dependencies from repository +Im not sure what we should do for installations +Next up. +Thanks the patch file without whitespace characters +a client side? Iam trying to understand if the store module usage from the jdbc layer is appropriate +Indeed +Hadoop already uses Ivy to retrieve its dependencies so using the local Ivy repo imposes little on Hadoop +Latest patch +Patch for trunk +Roman yes linked for this +As an aside one strange aspect is that if the Dependency is not added to the manager due to an existing duplicate then it is still added to the list in the +Ill either identify the difference or upload mine as well +Quick fix adding files a second time makes it you please raise this as a new issue Aaron? + +Might have been my fault +True thanks for the clarification +For license link I think we should clarify this with brandinglegal folks +I will take a look at it soon +I will also address this in Jim +I have mentioned as well the possibility of them owning the integration +They should all be easily fixable +Two translators would end up adding their elements at the same spot in the message file +Please either remove it or bump the version to a much more recent one +Will open separate jiras for the other failure cases that are not related to the timing out issues +How it works if you do nd attempt to stop server? cause it looks like it is stopped to trunk and second attempt tries to cancel the process directly rather than execute a shutdown script of any sort +I suggest to skip the fix for primitives for which coerced values are legal values +If we add more hash algorithm we will start using that column. +So there is currently no way to use our refactoring instead of JDTs +passes locally +Fixed on branches and HEAD +patch uploaded +But the sort benchmark does not have any computation in mappers or reducers +Added null check for getOperation in SVN revision This fixes the problem and the unit test now passes +But also see that suggests that clients should time out +No response assuming ok. +Resolve bugs older than a year as Out of Date +I added it but not having much luck. +Resolving this one +Regarding the theory I can see that there are bigger or smaller attack windows depending on what we doHowever if you give the framework to an untrusted third party which has full access to the framework cache there is no way you can guaranty anything +Attaching merged v and v patch just for convenience. +Fixed on and closing all issues that have been in resolved state for more than a month without further comments +This line would be If false was set on a scriptSorry my description in was not accurate +Perhaps a new discovery folder which would contain both JBT and JBDS discovery plugins and their files and which I could update wo having to worry about and line endingsJarred plugins files would still have to be published to and the devstudio SVN tree but the sources for both JBT and JBDS discovery plugins would make more sense IMHO in the same place than being split into two places central and the devstudio SVN +It would be great to have some of it extracted to classes and shared by both modulesOleg +ok resolving this since its the only viable workaround Emmanuel M +I dont think we need to worry about this issue. +Thanks for reviewing Sijie +Thanks +Perfect +Moving open issues to the M release +Would be pleased to Chris? what is your status +I think it would be best if the clients had no overlapping sets of split points + Any mention of these methods in the thrift interface description file +Pull request merged and also backported to +I believe that this issue is duplicate to +Another point is that doesnt need dependency to library anymore +Did you have any luck knocking up some tests to go with the patch +apparently not fixed waiting for test cases to proceed. +Its Maxs call really +Thanks Bob for the clarification +Ready for commit +Fixing known bugs before we cut releases soon. +The Windows version of the service wrapper is not available in the community edition +It is not an incompatible change because it just makes the hardcoded value as configurable and it works same as before this patch +Closing as Not A Problem. +In general I agree that sending a mapping statement to ES is outside the scope of MCF +The response is complete +trunk +Maven Archetypes Remove section +Two statements issue load into same partition around the same time +Updating patch name from to +Commmitted to to bring this listing up in the shell. +This was committed to trunk +Our Launchpad PPA failed to build the C library +Heres a patch for this +Committed to trunk +Moving to JBAS since it looks like we could have some bad metadata around the Message Counter History Day Limit property for Topics +Thanks for getting this set up +Thanks Roy +We have found that Xmxm works well +I applied itRegards filtering what you think of the the attached patch? I tried it here with simple stop row which makes a stop row inside a whilerow filter and it seems to do the right thing +here is a fix for this +This bug is related with being not thread safeIf the is run without patching It will give the same results like in the image posted previously +In the failing test case we are only dealing with List types that are so the previous comment doesnt offer an answer as to why we saw a failure +Could you please attach glassfish logs +Sorry missed one of the unnecessary this statements onThe methods arent final so I would expect to be able to extend the class and have getMinimum called everywhere instead of +Then the best journal should be chosen just based on the maximum transaction obtained from the image outside this class +and we have plenty of new code to deal with in M already +Given that there had been no response from Bruce about why we would need an additional logging level TRACE Ill close this one as wont fix +Issue may already be resolved may be invalid or may never be fixed +I think I will also have to add code to http and jms to fill the propertyThanks fot the hint about bare style +Mostly indentation changes +This new patch also addresses a race condition in +Will go to the backlog for. +As long as they are in the test directoryAlso make sure to include all required licenses +In testing showed a very large standard deviation in response time with a very large maximum response compared to the minimum +any logs in chrome dev tools or firebug? +Pretty simple +I am planning to backport this to any objections I will look into is a patch created from my merge +In the patch is changed to extends and is changed to make use of to startstop AM containersA number of tests that previously use directly changed to use insteadLast but not least due to the mvn dependency check issue all the tests in has been moved to cleans the dependency on server and then add dependency on +I guess that this might not always be the the script did not have itself as a s not that something is always getting recompiled then the issue would not have been there as you say +If it doesnt it should IMO +This is the document itself with the support and source files + +Is there a jira for that +There should be no code whatsoever entering the code base that is not covered by tests +I recommend using the backport version because of the classes names compatibility with Java +Testing for an empty string appeared more of a dirty hack but would do the trick tooI noticed that they were several patches in the queue for the +Not an urgent task at allWe need a packager specialist to do that +normally +Resolved. + +Done done and done +canceling patch as I have to rebased it due to the yarn move +Is there anything left here? Can I backport the commits to +Just noticed that once you login and go to portalproperties and try to add a predefined property like drag and drop you get an uneeded validation error +When I apply your change the second request caused bouncing messagingIll attach the was added +In we changed the behavior when waiting for the application to become ready +So if one plugin targets Java then its impossible to use Java at all +I went through our components and languages and removed the only documentation pieces +all ResolvedFixed issues lacking a Fixed Version +One of the things that I was starting to look at as I was creating the decaf lib was to pull out the code from the headers more for ease of the user looking at the headers but you cant take it all out template classes for example +Glad if someone can commit thisThanks +Patch looks like a clean merge of +Ive been in death march for the last month at the day job +I am using xmlbean for bindingyzj +In here I see a continual loop on all the flushing workers where they endlessly try to with their destinations +Please pardon me Im an Apache numbie +Test use case which reproduces the problem consistently by using a small memory limit and short time delay between production and consumptionA license header as been included feel free to replace it +I have exactly the same problem +I guess the situation is similar +Also included the diff for the driver file allowing this patch to be applied and the programs to be executed as is +Ill see whats going on +I see the correct mails being sent with your example +Im going to defer that to whomever factors out the common auth pieces into a common location +So all issues not directly assigned to an AS release are being closed +put on GA since important but not blocking code +I do not think its worth wile to try to get this working in conjunction with the other changes +Deferring to release +Seems fine to me +Thanks Robert and MarkIll upload new tarballs where the standard ASF license notice is being used in all Java source files and Ive also removed author tags to comply better with code standards +Giving to Izaak to check tried to replicate the problem in trunk and couldnt +Since I am not very familiar to the various opensource communities out there and further it is your initial contribution I think it is up to you to decide where to move +as some might be interested +POM updated accordingly. +Still have no clue why +numThreads becomes threads +trunk Loading improved but it was likely newer multiple threads loading anyway +and it didnt work +The metastore upgrade scripts are located in metastorescriptsupgradeAlso it seems like automatically dropping the tables is probably a bad idea +is that the expected behaviorI expect the iterator to resolve the null value to +I realise that there is a problem with the original XSLT +What about starting a project onIll have to think about this +What is? AFAIK this is an internal implementation class of the MR App master and not part of the YARN API +RI also doesnt stop after closing the C behaves differently in cygwin console +bq +Retries should be left to the client +Noticed that it can be null so will make a minor mod to handle the null case +Cosult with Pavel about lost styles +Unless of course the project has the facets and user deletes the file manually +some correlated references were not being picked up because the wrong root join was being used. +I have committed this +Because the whole matrix will be transfer to reduce stage it is not efficientWe may need a new to read the matrix +Fixed for for random too. +The naming similarity with the Membase successor Couchbase is coincidental and not affiliated with Apache +Dear MehdiI want to integrate this tool in c +And we do not support it as such +Not the final one +Committed to and trunk. +There were a number of other problems with not escaping certain characters which require it in content and attribute values. +Please review and applyThanks devs With above you will get file not found exception in test as mistakenly i havent add a schema file to the patch +Reduced priority of non critical issues which have current slightly differently to my original now have MD SHA SHA and Hex codecs +That is for any client wants to gain a lease of any file it could just rename it and then open it +Adding a fallback rule so that all outcomes are covered is an easy workaround +Marica can you open a new issue for that? Its a security violation + +The file should be added and deleted +rebased to committed +Vinod We need this asap +removed it from ql which doesnt have any dependencies on it +There is no way to use C namespaces to fix a problem with an errant macro since macro substitution is performed before the C compiler begins processing the source code +Javadoc enhanced in svn We might want to consider deprecating these iterator methods as being a bitrubbish though. +Thanks for the review Adrienbq +Jason is it possible for you to use some Hibernate API to obtain the list of deployed entities and their associated deployments? +Stand by +here is the in rev deployedThanks! +Looks good! Ill commit shortly +The patch attached today appears to be from a different +I couldnt reproduce the issue youve encounteredCould you provide a small war with everything needed and precise steps for reproducing the issueOr have you found out what was causing this in the meantime +Seems simple +Fixed so far. +IMHO a vector should be used +Should we fix for Andrew? Seems like basic obstacle to scaling +Either I did not get it or deleted it but I think it is the first since I rarely delete anything except blatant spam +Closed due to the release of Tiles. +I went ahead and tried adding the properties and ran into an error with +RDMBS or in memoryBasic class structure and file formats +thanks Ivan +Ah yeah skip should only trigger if when is true +Committed to. +I am wondering if this is an issue we need to resolve in Pig at all +This should take care of the issue +I see the package depends on useradd chkconfig and service but I dont see any service scripts or users being createdUpdated patch +Ill spend a little more time on it to see if we are all good with synchronization +And Arabian Standard Time and East Africa Time +The merge code is complicated and your unit test really helped uncover problems +To test unzip and mvn jettyrun go to httplocalhostautocompleteTestfaces +ChrisIt looks like you didnt tick off the License this code to the ASF checkboxCan you confirm it is OK to license this under the Apache License +andy how is your progress on thisIf you have anything pending to commit itd be great if you could do so Id like to take a stab at getting this resolved +If you find out that these two issues are unrelated you can close this issue +Are there any plans for this issue? +Thanks Nick! +The was fixed by cabdfadc on master. +Ive gone ahead and added it there +When this happens the new database process most be monitoredAlso updated the tests to be more comprehensive +Please provide an example that illustrates this problem +If its yes can you send it to me? thanks a lot +srcjavaorgapachecommonsconfigurationbeanutilsimport +These blockers and critical issues are resolved but not verified +To say it implements the spec is to miss the point +I introduced the main code change you wanted to make and changed the specs to match the new APIPlease try it out if you still want to put the method aside to have specs on it specifically we can do that for surewbuildrsvn ci CHANGELOG libbuildrcore speccore m BuildrUtil does not find gems on remote sourcesSending CHANGELOGSending libbuildrcoreSending speccoreTransmitting file dataCommitted revision . +It does not support working directory maintainment +bq +They should have been Serializable in the first place as Hibernate clustered cache implementations require thisAll unit and integration tests pass +defer to +These notifications doesnt have the output but we have node specific outputs in receivedResults notification messages +Guess we should document it as a limitation open a Jira for it and tackle it later +after emptying one cache and restarting geoserver it reads Using B of a maximum datadirgwcdiskquotapagestore directoryh and restarting geoserver per reset the negative usage size to the correct think it was able to delete tiles +Closing this issue. +This issue was reported over a year ago and a workaround seems to be provided +Sorry to bother but wont be better to fix the unlicensed files +Able to tail thousands of files per second using a small fraction of CPU negligible memory +Is tihs a duplicate of +Rejected +Need more information Could you run your test one more time +Running in Hudson +I have removed these scenarios from the tests +Thanks Ricahrd Ill review them now +This problem is fixed with these two adds +Looks like the last patch has compilation errors some things still referring to the old r name + in rev +Also are you linking statistics use to the JMX flag? I think there should be a separated flag like core. +Now while this is still a lot better than having no upper limit at all and risk running out of memory if you have many workspaces it would be nice if this as well as other parameters were configurable +I reviewed and committed this patch to the trunkmsvn commitSending javaengineorgapachederbyimplstorerawlogTransmitting file dataCommitted revision +Besides I have more confidence in understanding the code that I committed vs +patch attached +Updated to incorporate Phus original ask and Amars feedback +Not much interest here plus rather to implement. +for the patch +This has been merged into EAP +The difficulty I have right now is ideally both scenarios would be supported by default +Master file patch for fix to branchmsvn commitSending javatestingorgapachederbyTestingfunctionTestsmasterTransmitting file dataCommitted revision . +Committed +I created a new one so that it could be possible to have a gradual migration of all lookups +i know they are meant to be sent over http with solr specific marshaling but we ended up needing to send some solrj request objects to another box via RMI and it was a bit of a pain +Well Im finally trying to upgrade to trunk and this bug is getting in my way +Bringing in since we voted on it and assigning Ryan + +I will read up on its usage as this is a new tool to me +was created to track the failure. +The lack of proper symlink support on the JVM makes it even harder since we cant even tell whether something is a symlink without making a native call +Furthermore Norman noted that in our rc src distro there are javadocs in the www folderI think this is a lack in our build system I see that some of our src releases include the www folder with some parts of the website other releases dont include this oneShould we try to fix this for? What would we like to include and in what packages +SVN +Patch to complete this improvement +A working with a test case in the main method +SVN revision . +Merged into the branch +jbossportaltoolsrefguidehtmlsingleimagesstartingprojecthtmlsingleimagesstartingproject +The first problem is that the eCommerce site does not filter by catalog +It adds the initialize and seed methods +What do you think +I give up I dont think there is way to meaningfully finish this so having removed the superfluous usages am calling it a day. +Hi JukkaI agree on your approach +Attaching jstack of the dfsclient +I found the changes in the java files in the build +fieldName tournrsped sizeMax add field date jpa class +Please verify +This looks great + +Unassigning due to inactivity. +Something liked s maven argument +Ill also look at the patches attached to since its referenced in this ticket +So you dont need enc ansi fonts anymore +Do we still need to generate a +move to doc version for less overlap with engineering tasks +Thanks Jack +Its been noticed you great work for the community +Thanks +Well theres also incorrect handling on master side but this is trivial to avoid assume this occurs in both and trunk +Hi WilliamCan you attach your testcase to help us reproduce the problemRegardsGary +Works very nice +Committed to trunk. +Stefan those properties are not in the SLD specification I believe they are part of SLD Aslo the rendering code will ignore themThe two patches you attached are HTML files seemingly generated by Trac +Manually tested! +However it is never popped and in places +Corey I think I had previously applied this patch for. +I dont think I have permissions in Jira though to assign andor close the bug +Overlapping is allowed but probability to receive such report was lowered due to which covers examples of this ticket +Hi AndyCould you explain why and are needed? There is no equivalent in the xml metadata +This is a clear bug in the code that will result in this error whenever users run more than one jobs assuming is fixed +Hi would you elaborate the problem a little more please +Verified on and on +Duplicate. +Maybe someday Ill redo this with threadsafe statsSounds good +Its rejected by ISAPI +Please let me know if there is anything else +Attached test file that shows issue +bq +Thanks Claudio +I am getting coverage over the class +Also fixedadded comments to regarding new +Pablo please verify if possible +Actually +Could you take a look +So that youll have a folder wise separation +The operation that synchronizes the stdstring reference counting triggers a memory update just like a cache miss along with the invalidation of the relevant cache lines for all processors etc +accepted and committed to sandbox +I dont think notion of slicer makes sense in local mode and ignoring it might be the right thing +Id say its an issue worth looking into +Default implementation creates a WAL for meta and a WAL for the RS +Ill attach a patch later +Not related to the patch. +as major +This has now disappeared after two weeks of CVS updates + +to just use +looks good +Yes for the mouse problem Ill look into that with the shaking problem I removed the indicator and its fixed the display of the indicator is handle by the scriptaculous component which uses the display property instead of the visibility one to hideshow the indicator +In can reproduce with However master branch do not seem to have the problem so I suspect this has already been fixed as a side effect of another bug +Somehow the change didnt propagate from trunk to branch +Done +Missed cleaning up stuff like getX and the commented code in the original +Have you tried using +Thanks Mark +If no one objects Ill commit and push the attached patch +So basically you would have icons for STRUCTURAL OC and overlays for AUXILLIARY OC +Use Pattern Matcher and to perform simultaneous replacement of two distinct values +Committed to docs trunk at svn will to branch +Any updates on this? Can this be reproduced in? Steps to reproduceIf there isnt any activity on this case Ill close it to +Added optional swap callback for aspect services +Please check that it was applied as expectedI didnt understand however what exactly the attached test actually tests +are provided new interface solves the issue +If thats the case then this is a bug in not in Xalan +Jan that answer did not clarify anything I hope I clearly described two very different interpretations of resolve +Yes I considered that but was afraid any earlier printing might change the timing so I opted for a minimum change which only kicks in after the error has happened Well see if I catch any fish +Marking as critical since there are other critical issues based on this +About to commit changes for mySQL only as Ive posted a request for help looking for users with SAP DB as their RDBMS +Heres a test case which illustrates the was not a problem in release it has only occurred since we moved to trunk +I replaced the clones by a constructor call as you suggested +Thanks Stack for the review +Henri This is awesome! Thanks very much +JSF JAR project with +I wont get to this for order of facetnumTerms and fixed the getShard call to be consistent with SOLR I think this is ready +Im sure someone else will run into this +OmarIf you apply for commit rights at Ill assign them to you and you commit this +If you want to move it thats OK with me but it might just create more confusion at this point since subversion does not allow editing commit messagesSorry for any confusion +Please contribute some. +to commit +I have a large enterprise application which is broken down into modules with well defined dependancies between them +would also need to know what version of you are using since you claimed which doesnt exist yet. +Thanks a lot Paulex +Heres a patch that uses the new approach to package contribs in HDFSThe raid project has a lot of unshipped dependencies which will be fixed by this +Thanks for applying the patch +Im rolling this issue over as I cannot reproducable test case provided closing issue +It will report your error +This should be fixed in the new build +The linkage error is caused by a packaging problem in the test +Applied in r. +Closing. +Thats not what Im seeing with a fresh download of Pig +Yea this is a bit strange +Hi Sijie Im not sure how adding the metadata when creating the znode avoids such cases +committed to trunk and. +If there is a problem please open a new jira and link this one with that. +I tested indexing first M wikipedia KB docs and running various queries and found no real perf change or at least less than the noise in the test +just changing the component the problem is most likely in the drop constraint code in the sql layer of the system +This appears already fixed by please upgradeBeen resolved for a while. +Ill check in the next few days and Ill let you knowThanks let us know if this bug still existsthanksRegardsMichele +Ill try to find a public one +attached w import and code style cleanupThe change in the Filter classes looks gratuitous Im guessing this is bc the patch started against an older version of the code when that is what the signature was +S +bq +I updated the release docs too. +Minor update to the patch +Ted interesting +It is not a line or two but I did it in minutes +He needs to make a patch +Made the change in Git ID edaafefdffdaedbe +Im continuing to work on this with a goal of finishing this week but its a background activity +Duplicates it still open or resolved +Committed at subversion revision +Forge include an example of how to get runtime in the test plugin directly when the runtime is available through Maven +As a matter of priorities Id like to not spend time making sure +The assertions in cycle are testing invariants and are therefor left in place +It just started to reproduce a bit rarely +Are we sending a new each time we hit an arrow key +This came about because I wanted to try out Markus patch for +I will try to come back and make some updates to it as soon as I get a chance +This file is created in sync with the project classpath +Colin I have been meaning to ask this +updated javadocs +applied the patch +OTOH We do need this functionality in EE so I dont really know what to do about it +There two moments When Sonar executed from Maven we will always have directories because for example srctestjava always presented in a Maven model +Thanks Stefan +Spoke to Ryan offline and hes OK reviewing this later this weekCommitted to trunk +I take that back +I agree the package name does not fit anymore with its content +Just for the record I have hit this when deploying a app targetted for Tomcat +I think I goofed up the order of operations in submitting the version of the patch +Should we simply require all containers to be consistent with serialisation +Maybe the best approach would be to externalize that setting in the samples and allow it to be changed with the database platform +Please tell me if one of these solutions are valid I can upload a patch with one of them in a minute XD +We should do this for more of our configs let client set them +ThanksKristian +I dont think hive execution complains that directory doesnt existthe drop table drop partition try to delete after the transaction is committed only because we dont want to delete the data without deleting metadatabut as you say we could try to drop the tablepartitions directory if the transaction fails +Patch applied revision +But yeah sounds like a good plan come up with a good prototype +Long time ago I created a special tab that appears only to jbossorg role users +I havent checked to see if this issue is a duplicate but I would not be surprised if it isThis is part of an effort to close old issues that I have reported +Fix to the build break is fixed now +This one should still stay open +fixed +These methods use the findFile and find methods respectively when it is looking for the file to load +Thanks Im going to work on this issue on Monday +As long as theres still a way to go into librb and build a gem Ill be happy +I think this can be resolved +I finally found the time to refactor the original patch to use an EJB plugin instead of dragging into +Thanks a lot of the explanation +Not being a subversion expert a tar ball was easier +Since we removed all ant related files this issue is no longer applicable. +Attach patch +But I agree incorrect is worse +If I dont hear back Ill commit next week. +Still long to get the war from jenkins to testp. +module version may be null so we have to avoid null pointer exception when compose configId of groupId artifactid version and type +I changed this and it worked for me +Oops there are still other issues +Javadoc style is better +No I am not unfortunately the position I am in at this time is that of having to patch because its our base +This would work for too +The test uses some fields in the thread that are not volatile or synchronized on +If you create the db and then use a copy of it for each iteration then that is ok +applied thanks +No +I improved the text message for the exception to differentiate the st and nd cases under r +Yes unless a performance test is done I also believe there wont be much of overhead +I have also replaced some words like atomic and complete that may cause confusion in the way they were used +Now that work is well underway with AS all previous community releases are +This issue is no longer relevant for it is fixed inChanging the affected version to +New patch attached addressing the schema and javadoc issuesFor the schema i just added a new one +I also think that further discussion about wrappers should continue in because its an issue covering the task of datasources bridging +updated patch added +Hi UweAre you willing to mentor this project on? If you are I will keep it assigned to you otherwise let me know so I assign it to me +did some refactoring that I hope to be able to commit soon integrated the patch in the fix into fixed as per r +It is meant to mimic the node +Here is a patch! It has whitespace changes +Meh dont have time right now to test it out +Built with of +WDYT +announces the last migration which ends up pushing to other nodesadded a logging statement to clarify why an update is ignored +This one is also not used and we do not separate in Stanbol between framework and stack anymore. +but it didnt work for me +Can you provide a description if the expected behavior +This one is pretty complex to solve I need to think about at least if you update the entity itself. +is the right fix here. +Committed thanks for the review Anoop Lars. +Patch did not include treats a sentinel byte of zero as EOF and throws an if the sentinel byte is otherwise not x +Please attach an XSD as an example from which we can work +Added missing methods in revision and renaming this issue as the current usage of the fake node is looking strang and causes for trouble the fake node is used to potentially determine the resource type of a synthetic resourcce by using registered jcr resource type providersHowever as soon as one of these providers calls a method this might result in an NPE or any other stange exception as the fake node only implements two methodsTherefore we should rather remove this Ve removed the fake node and fake node type in revision looks like this broke an integration testhttplocalhost +for some reason Im not receive updates from the cayenne so I didnt see that you had resolved this +Followed Johns advice +I would like to have my PK fields mapped to in the JPA entity and then implement but Im not sure how to hook it all up +Also I saw your earlier comment about the other changes so dont mind me +Commited. +If you cant find one you should ask for it +I know people dont usually use characters in field names but shouldnt we replace the w before the colon too for completeness +Ill have to see if I can track down a way to unit test these against some mock implementation otherwise if the tests require a local xmpp server setup theyll likely be disabled like the original tests for this component and probably never run +The problem was database connection to +A review would be appreciated +Revision +So if you have a good baseline and observe newunexpected test failures after applying my patch please let me know what fails such that I can review the relevant refactoring of the container +Status update patches to this issue no longer apply to the current state of the the needed functionality however still makes sense issue is blocking correct interpretation of what is dirty +The patch is very simple so Im bypassing Review board and giving here +I will pull itpoint seems goodpoint when I stripped it down too much the following tests fail state propagates from one to the other +If we put SNN config in clientPrincipal we will be forced to run Balancer with SNN principal +Committed revision to using the patch from jun which applied cleanly as is in now also. +Bump the Jersey version from to Standardize on Jackson to avoid inconsistency +We should either change the imported data or relax the constraints +Please attach your thoughts then feel free to assign it back at me +Delete a needless main issue with this approach is that in most cases openjdk gets pulled for some dependency +Heres the full thing +I made similar changes to all the WSDL files missing definition names or soap addresses +Hi PatrickSo I guess you tried and its ok yes it would be cool to have it in and update with sure! the bug been confirmed as fixed? It looks like the issue owner was waiting for a response to confirm it works +We need to add a clone operation to +Because of the version naming change and some other reasons regarding the archetype catalog generation we currently have only those outdated archetypes +This is a duplicate of +Seems a dup of +BoilerplaterTypePrimitive serves as an abstract base class for BoilerplaterTypeInteger and BoilerplaterTypeFloat +Also we persist the id of the message and the message itself atomically so if the ID is there then so should the message +All rubyspecs pass. +It would be nice for another committer to check out this nice patch +Im still investigating that +All worked as expected again + has already been addressed +Martin please verify the issue has been fixed +Hmm I think this is done Woo hoo assigning to me and resolving +Then the other basically get stuck +The issue was fixed by just added tests to validate +Thank you Ramesh! +The failed testThere was a created +Excellent idea +Ted Looks good +Example usage is included in the documentation +With just a reference count of scanners Id be worried that compaction would never make any progress +No as can be seen from the example POM I provided +Does this work for you Johan +merged the pull with some minor changes +The png you refer to does not seem to be there anymore +I am currently using and it is STILL not fixed +This is a security risk do to the plain text password being either passed in to the command line or saved in the users +pls attach your wsdlxsd reopen bug with your wsdlxsd +i would like to use the function where i only configure it in hudson and the pom is generated. + +Done for GST guide Migration guide for +Committed to trunk +However the client still never receives the which in this case is JSON +I always wonder of what nightly version are the nightly docsSo Id expect the docs to be found under the section +Brian no time right now to make a patch for this one +Closing all tickets +No rsynccp is necessary simply specify the correct output directory to start with +I confirm that it works small improvement would be to have a better message thanAn unexpected error occured now it updates easy on the error message +Closed. +I will look into SQL spec to see what it says there +That is why I was not seeing it in cvs +fixed in r thanks for reporting this issue +fixed now +I sure wish Id paid more attention to this JIRA before youd have saved me a couple of hours of +Good stuff +I also noticed that ADTSM is taking the mdsum penalty for every token generation regardless of the conf setting +jdk update +This issue requires upgrade to new version of jetty +please commit it if only to be sure that your karma is setup properly +So its ok to close this issue +Corey Can you attach your sample ear to this issue +Change commited in version I ran your test case without the failovertrue option and it passed. +producing the same issues for this after populating the local repo with whatever is taking this functionality down +Attach +What is the ticket no +you try to do that and attach the process definition information if this is the case +Woops I forgot to watch this one and it just came back across my radarI dont see any harm in adding the config option suggested here I was just wondering if there were another way of achieving the same goal without adding Even More Configs +has another small example that trips this +For one thing it allows the client to reliably detect whether all the data has really been sentHaving a like setting one disables the other is bound to surprise someoneMike LooijmansPhilips Natlab Topic Automation +Website updates svn ciSending Sending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexcommunitySending srcmainwebappTransmitting file dataCommitted revision +mrepositoryorgapachehadoop +NicolasDo you know if there is any progress with thatThanks a lot +JonathanI think we could add testing to this patch through doing regions replayrecoverdedit after creating writer when doing splitlog +The user management goes via ldap +I wonder if rather than callbacks this might look something like an RSS feed for changes +Can you please provide a full log of the first glance it seems to have something to do with Maven Project Info Reports Plugin +It was our assumption that in these folders there would be files that would give specific details about the release such as the release name and version +TAB will give a complete function list on whitespace boundaries wont filter unless you type in words to invoke a prefix filter +Did you also try to catch out of bounds exceptions instead of doing the bounds checks in the current patchPaul how can we do this? In the mmap case we can because the mmap getXXX will throw the bufferunderflowand not actually read anything if there isnt enough bytesBut in this case I dont see how we can the code to safely do this? if you know please let us know as I think this would be better too +Ive opened and attached a failing unit test for the problem I described. +Now that theres these alternate login methods coming theres a problem if the user has a TGT its authMethod KERBEROS but then accesses a service requiring HSSO +Looks like the javadoc warnings were caused by removal of the ant jars and fixed by +It incrementally adds reserved space on a node until it can allocate the full resources desired by the container +Note in my integration test which does reproduce this it failed times out of runs +I am still looking into this +Is this a port of to +Ive made a mistake in the test case +If thats not happening then its a classLoading issue on the AS +Approved for SOA CP but depends on inclusion into EAP CP and the ability for SOA CP to consume EAP CP in time for release + +Liyin +Fixed in the candidate. +This patch should fix the problem but please verify before I apply it +message +This will be fixed soon +Thanks Vladimir the test was excluded in DRL VM at r +Hi Adriandont worry it was easier than expected +Committed r +CDI extensions do not have access to at deploy time because CDI starts before the Servlet Context starts +Ive drafted a patch which changes int driver to int driverV and adds void driver that wraps int driverV +This issue has been assigned a developer who has estimated that this should be able to be patched by the end of this week +James feel free to commit any time +Many thanks for these new tests +changed the default and closed. +HiLooks like youre using reuseReports mode so this issue duplicates +Added the commit operation to the close and cleanup methods for resp +Can you ask Emmanuel about this? This was also new to me but thats what he just said to me on IRC +The patch refactors the code to fix incompatibilities and does not addmodify functionalitybehavior +Unit tests test is for current behavior and passes another fails should pass once the ticket is fixed +OK I can see that this makes sense as a configurable feature I will look at adding this functionality +If theres nothing left to split I think Hadoop is complete and we should all move on +Patch for an older version of not for commit +Have you done testing on better values Stu +The image file should go into trunkcontentimages + on the simple build processBut it is still nice to have settings files available to help compliance with coding standards +Review points addressed details in attached +Forgot to mention Im using Spring Framework comment indicates a configuration issue which has been fixed. +The code around does not check if the return type is used +Closing. +The testselfloadspecificationsyntaxerror failure does appear to be gone but I am still seeing the other vartmp path failures here +Done see +Oy I have just been informed that is the Oozie port +patch looks good to me. +This introduces the following possibility +This is a nice idea but accomplishing this would take a lot of additional resources and involve a fair amount of development There are a lot of subsystems in this request that would have to have shared keys and api access in order to work correctly this is a security risk for the type of account that would be needed to be used across each system to be able to have global access to all projects +I checked the buildand this seems to be fixed in it +Also some of the files may have not used the convention you have used of starting the void var names with VOID +Grid would still like to see a patch and maybe a screen shot +I am attaching a new patch which is the same as Allesandros most recent but with this flag removed from because its not used in the source code at all anymore with his patch +Closing since handles this and the old datastore has been deprecated +Unfortunately I do not have the time to properly contribute right now +Can we use review board for the patches +Both issues have been resolved and the service seems to work correctlyI will commit these fixes now and deploy a new snapshot such that regular users may also test +Some additional informationsthe issue is not present with version to reproduce the issue you just need to set your form to multipart and submit it via AJAX +This can be closed by now. +Easier management no need to change +uploaded patch for could you take a look before I commit in r +done +Do you meanbq +Im closing this out +Please confirm naming +To be clear those heap dumps are against +Extracted Text from isThe extracted text does not have field values such as C Box etc +bq +Proposed fixes in but requires configuration setting to activate +But looking at the changes involved Id prefer to remove the test for branch +See the last comment +I can connect to the Derby DB using the cmd window +Moving back to for now +According to the bug in eclipse marked as wont fix we cant have the information when the project is opened that some project preferences are available +Perhaps my Maven inexperience is getting in the way here +Your suggestions will be most helpful in that effortFor the time being Im sort of consumed with but after that work is completeIll revisit this work and see if I can figure out whats left to do +RTC work item defer minortrivial issues to +Fix seems to have nailed it +sorry about thatGreg +Im looking at which led me to +Updated patch to address Todds comment +What I notice was when a element is deleted its dependencies still remains as dependencies +I would first replace your windows xp from equation +Fix commited by David Lutterkort thanks! +Great thanks! Looking forward to another release. +Updated to distinguish between a forward include and chained request. +Hi EriksThe problem seems more complicated than that the work field that you have commented out should override the original one but that isnt happening so I suspect that the problem lies in the widget component rather than the form definitionI dont have time to look deeper so Ill have to leave this to someone else for now +Code looks good +I cant reproduce it with trace while trying to delete project. +We should get this for free once having done +We need it very much +test the tranistion logic +HTML output included for review +Tools now uses new logo with no occurrence of ExpressFlex just plain. +So for me this work around is not working. +Please pull back in if you disagree +It reverts to the desktop variant under the iPad device my opinion we should reword this requirement as a feature request to use mobile styles for tablet sized devices +we have nice comments to this file now +For instance I am unable to upload an artifact with the version which is the version of the artifact from another would be in favor of removing version validation entirely or perhaps making it a warning instead of an error +attached +Could we not allow this in the of the input and stored version but normalize it to whatever other internal format in the indexed version +Can you provide logs from the broker when this occurs +Have you tried it? If it works for you can you post a patch +I dont know the specification of wsdl more clearly +When one region is spliting its usually heavy loaded +Could you specify Fix Versions for this issue +also you could face a change in the physical dir layout + to use for snapshot repository to be recognised by ant maven tasks +Ill update the wiki now as its working again +Fixed. +Changed this to be its own Filter rather then a hacked +Currently i have updated it and pointed to the Wiki link +I checked with some wsdl documents they contain the messageLabel but not the direction +Seems to be a general sentiment that this fix should be included in +Propose to close +right not a big deal in reality but good for code completeness +Thoughts +Replace to s a appearance feature request +And wed flush when that DW hits the RAM limit +in M +Thanks in advance +There may be a related dblook problem with in databases +Is this correct? If so the plugable interface will allow the use of different repositories for the metadata part but we will still rely upon zookeeper to monitor node availabilityI have a few other comments about the design doc In the definition of the operation the comparison is performed using the key and value itself +Brock +Patches to add a junit test to to test for the null mappedBy relations +Ive searched the entire codebase +Patch that rolls in all previous patches plus makes a few more changes +The table schema file could be stored in ZK but having our schema in multiple places in HDFS and in ZK makes me queasy +bq +Duplicate of after release. +I think we can close this item +thought that this is no longer there? could u dbl check and close this issue if it away +Heres a patch to add this documentation +I modified to populate the XML type for the fault I copy the properties using reflection +Thanks RayThe patch was applied to java branch at r please verify +Patch reconnects back if unsafe mode is it added a test case that would send a msg stop the source and start the source again +Are they now viable option +Although I agree it is just as nasty for people that do a describe feature type that does not specify any individual feature types +Asking it in a resolved bug report against is not going to get your question much visibility +Added Impl suffix to metadata implementations. +in JDK I could leverage and avoid the locking probably +Im fine with it thanks +I guess case is what is implemented now +Ram should this go into? Im fine w this in branch. +keeps history inAnother solution is to provide a general framework to support retry in ipc +Can you show us your java version? Our first instinct is that its loading a binary under JVM even though you claim its JVM +read the forum post +Thanks for the tagAssigning to Martha as she will be taking care of EAP +See if helps with +I would personally prefer seeing a dependence on existing channels than another implementation of something like the file channel and something like the memory channel +I have committed to both and trunk so if you want to look it over and make sure it meets yours needs that would be great +I fixed this bug in rev aecd in branch and ea in master +Have you tried a later release such a or a SNAPSHOT of? Theres already bee a lot of MQTT fixes since so this is possibly already taken care of +Assuming it is ok closing +I suggest that you try to fix this the best you can and if it cant be done within a day then just push this off to a future release and move on +Or what about other metadata on the Token object that we might store in an extension? In the advanced cases of Connect were going to have permissions sets that arent expressed by the scopes or other basic parameters that are accessible from the Authentication +Whats the feeling +Lets just mark this as and. +please push it out +Hey Gav if you can make me the admin I can always add peeps later +Comprises of forrest xml changes suggested by Corinne +Patch for trunk +Any news on thisIve put it on the todo list for this weekend to apply and test and probably commit if everything passes +the container and the services live together +Am trying to code something flexible to handle case deprecated to new class does not filter out directories but does give control over. +Higher order logis is needed if Process A generates a Resource in Plugin X +onLoad is depercated and will be most likely dropped for use notifyTopics instead +Marking closed. +Nice addition! +finally i got the this bug in jira be movedreassigned to the maven standard modules? or do i need to open a new issue therethanks +can be will have the same +LuisThanks for the documentation +This resolve some issues that we have on the initial patchPlease apply +If this is a temporary problem please let us know +I think this will cover the biggest chunk of the problem +Commited in Git ID ebeeaeadabbdbebc +OK lets see if this one can pass tests before the next conflict gets committed +Moved to due to lower priority compared to BPMN as a of employ global data source in Tomcat. +We should decide also if links to tables and figures should appear in the bookmarks within the chapter links +Ill update patch based on your comment +I will report if its annoying or occurring right when exiting + This patch fails unit tests + ajaxified make multivalued properties intend to release an initial version this afternoon or I am the Assignee but cant assign myself +I am interested to take up this project for +LDAP is a protocol not a storage system +Runtimes are stored as one property key inside the file +Im having a hard time proving in a test that all references are indeed correctly released upon blacklisting and the entry is removed from the shared +Well for me it was not a bug but feature +But module support I think is required to do the above componentization +It compiles all unit tests pass and continuum works +Yes its the same as +This change modifies many classes in different packages +Thanks PhilippeKarl have you used this solution in production? In case it has been well tested could you create a patch +I had to comment out the test case as when the full test suite ran the Derby db initialization in the integration test was causing a +For Nicolas Peters example setting optimizeLegibility for all text elements produces a better output +httplocalhostbookmarksbookmarkshow +We have decided to hide those attributes since they cant work now and it would be risky to just remove the functionalityWe will open the issues for this functionality given that nobody requests it we will wait for a feature request +one patch with basic Python support for Sling another patch with integration teststo run the testscd launchpadtestingmvn clean test Dtestintegrationtest +Set correct fix version +Great work in all casesUwe +Patch for AS support +What do you think +Actually Im wrong it has the server side too +hopefully my concern is now more clearly +Im not at all suggesting that Cassandra should bend over backward to fit another system but if there is a way to nudge things so as to make technologies work together I think thats to everyones benefit +can we get ETA for all remaining WS issuesIs Aug th still feasible +In my opinion should not dictate to users what they can or cannot store into the systemI did not say anything about empty keys which of course makes no sense +Yes Michael! you got it right +I have not found any other issues yet let me know how your testing goes +So far on three clusters weve tried this on it worked fine in two +patches created and not commited as the components uses moving popups to direct body children before show +Ryan had committed this as part of his recent patch along withI guess this issue is resolved then +Can we please get this merged and backported to and release +and including the fix have been marked in of a bulk update all resolutions changed to done please review history to original resolution type +The only additional changes were to which had to accept a as well as an in the case that is used in the +The calculation if the index of a childnode is valid is only done if the corresponding hit is requestedIn our internal tests the query execution performance didnt suffer but is much better for with a large number of matching child nodes +moved to a unit test to ensure its already in +It would be nice to add links to the page about the or a summary of how it works +But committed in trunk and +Using jaxbproperty is the way to resolve such conflicts that the specification endorses +The fix requires a few changes to the geotools filter and gml bindings since they are not really used for encoding there are a few bugs +Please verify. +The user now gets the errorA value must be provided for member bundlePath. +Brandon can you point out where exactly youre seeing the incorrect documentation? Is it in the XSD attribute? Im not seeing it sorry +So for instance wo Seam Faces people have no choice but to use AccessorLocator to do any sort of CDIThe way I like to put it is you should be using this to provide CDI integration not spreading it all over application code +Maybe this was fixed by the change to make grapesgrab work with jdk. +Hitesh please let me know if you need meRavi help with a share of the remaining tests for fixing +Thanks Varun! +I am still in the middle of finding out the problem with my Vista +Hi KelvinWe have noticed that the left and right value for horizontal text alignment of cell is not work so we add Javadoc like the following words to avoid user using such valuesThe parameter can be center end justify left right or startNoteleft and right are deprecatedI am not clear about your suggestion do you want to remove leftright values from javadoc or something else? Like use the leftright as the value set and convert it to startend as the odf attribute value in the implementation of this method +Trivial patch attached +I just committed this +OK Jeff had made a change to the configuration and I edited an inactive copy of +Thanks to all of you for your efforts +Steve you statedHibernate internals are built to extract values from using names not indexes so thats not an option here +Trivial patch +Is this really a showstopper +Ive reverted the minor changes which for clarity changed only the plugin folder name this was why the tests failed +It looks like its the default value that makes it tested! +Added these bundles to the launchpad builder in Rev +Will commit patch v tomorrow +Resolving as WONTFIX as its really a configuration issue +If you add an item to a map with the wrong key you sure arent going to find it later +One possible improvement might be to use for the toplevel package to keep the prompt length down to the required minimum +Done in trunk +Well be using seam forge for future component creation +Sorry I missed one line to reload the configuration before reading that item +I also amattaching the DIFF files for the modified files +The wiki doc has been corrected to reflect whats accurate for Teiid +Unfortunately this patch no longer applies cleanly +Hum disabling the native acceleration seems to fix the issue +AndreySorry for not getting back soonerYou seems compelling +But when I try to deploy snapshots the assemblies become corrupted +Marking as fixed. +Please fix this! This issue has been a bug since will have a look into it BUT to be honest flex has a very low priority on our list we was even discussing if we should remove flex integration from Railo but we will not +I will find out if the test code can be granted to ASF +I also double checked if i have any references but there were none +The related issue is fixed in XNIO AS +In version there is now a new referral handling method Follow referrals manually which does exactly what you requested +Also added simple tests to verify normal clean flag +When you use Web Flow component tree state is not stored in the session but in view scope +looks good +How bad is including those files? Instead of referencing to them +Thanks Zheng +Do we really want this file in our repo? The same information is available in JIRA release notes then why copy it +Thanks NickI thought you created temp resources through getFile but that was way off baseAttaching another shot let me know your thoughts +I can move this one if necessary +I changed that lately +Theres also the maxdocumentsize config setting +committed to and trunk think you should update JVMOPTS in test too and the comment message in conf tou and greater need need k +I would really like to also run these tests +Im not sure exactly what the hashingscheme column will be used for but the salt will make every user have a unique value in hashingscheme and I think Id expect every user that hashed the password using the same algorithm to have the same value in that column +on v patch and notion of followon jira +The goal would be to be able to choose different object factories different factory implements and different configurations for multiple instances of a filter andor portlet dispatcherUnfortunately chunks of Struts still use statics and therefore are unable to be cleanly separated +There are two test failures looks like an inherited failure its been failing in other builds +Both +Definitely not blocker +for log information see the attachment +Ill work with him when he commitsthe latest changesThanks if I understand correctly you have refactored the WS binding code and either the problem does not occur anymore with the new refactored WS code or youre going to apply Raymonds patch +I also modified Navigate Space on Home page of your space not to be header so not to appear as second book after exporttransformation. +I agree +This is needed for Wicket internal needs. +Thanks DagIll update the comment to make it more accurateCommitted patch a to trunk with revision Ill also look at merging the fix into the +Thanks for the patch! +This patch adds the appropriate test for the behavior and fixes the thriftnative project to do the right thing +This implementation also makes use of the hostrolecommand definition rather than defining a special server action schema +Or satisfied also by join on where it is also known thatA is a single value like +Subclassing seemed like a good idea but even it doesnt appear to have a clear point in its lifecycle at which the success or failure of the connection can be determinedAny tips would be appreciatedWhile it would be possible to create a consumer and producer that use the two different brokers Id rather not add that complexityNote that Im working purely with at this point +This problem is Facelets related and mostly solved by version +Multi selection will be provided as well in Select Interceptor Binding dialog in New Interceptor BindingInterceptorStereotype wizards Select Stereotype dialog in New tereotype wizard +Ah for the ones they depend on at least +verified back to +Updated the implementation to reflect semantics discussed in and requisite tests +Ive implemented some parts of this +I think we should close this as a duplicate of +I will commit next week reviews description from comment in Java RaghavendraYour patch is in at DeepakDid you test it +I believe these items were completed +mark as invalid as per Enricos comments. +We tried again to upload the app made by Cordova and now it got accepted for evaluation without any problem. +Since youre already using this I am assuming it will not impact you that much I hope + +The ideal fix appears to be to remove it from the groovy script +I havent been following all the fencing jiras so qq are tokens andor their services involved +Thanks for the patch Ravi +oh and the method adds to would be named not to follow the naming pattern of current s or +Feel free to ignoreFinally I updated the file to make sure that the compiler treats the source code as since the enum keyword is apparently in the code and the compiler isnt terribly appreciative of that +HiIm trying to use prefix feature but Im getting a prefix is an invalid parameter name its still not in production version how can I get an updated sources or may be one of your nightly buildsThanks a lotMichael. +Ifanyone else runs into this problem feel free to email me and I can walk youthrough the changes I made to the code to get it to work in our environment. +Thanks +If you need anything else let me know +slip +This issue is currently blocking our new release so it would be great if we could resolve it +This was an oversight in creatingCompleted At revision +Fixing Affects and Fix in Version +I gave for this a while ago +Resolved via. +By simply creating a project and running a test this failure occurs +Sorry Jason I got sidetracked with and other things at work +Please do not revert the commit since some contributors have started other works based on the committed patch +Attached a patch with test cases +I believe this is fixed reopen if its an issue. +You dont happen to know which classesjobs do you? Im unable to track which jobs are actually getting run +Im not an emacs user so I may be missing something is there a configuration setting for interlocking? Im running on Ubuntu. +texts changed in src repo and release tag +many thanks! +This issue may deserve a release not by itself or it may be a subcase of some larger release note for JDBC features +for +Hadoop passes them to Mahout but Mahout doesnt care about this value usually +HelloWould you please try this patch? Thanks a lotBest regardsRichard +Thanks +I have attached the patch there so please let me know if you have any addition I think the provided patch could support FORM authentication as well with only a few modifications +Updated version of the patch which supports Summarys as well +John was referencing a problem that I was having but after some more debugging this morning the issue was due to EC configuration and not related to this ticket +But before do that we need to move forward +This is fixed +The failure in profile is valid +The fixe for generational mode will be committed as a part of +AFAIK there is no please try again error inIf you can provide HTTP header and bodies transferred that show that this is a bug in feel free to attach those and reopen the bug +Is this something still validThanks +Is there any value thats guaranteed to be safe? Please provide a constant value or if this value is dependent on the size of the Id love to know how it can be computed +Any plans to fix this please +You can configure this yourself if you like but if youre debugging your own code this is dead simple in the IDE +Second typo is fixed and looks OK +This is the version on the default build of the machines I have access to +See as duplicate of. +davisp mostly what I had in head +I finally got around to running this and taking a look at the actually web UI +Im interested in solving the blockers for and the solution for might be slightly different because of the changes due to so I was thinking of doing it separately and possibly after releasing +Pls see attached file +OK Ive found it +And I dont know about vs +Committed branch and trunk. +This is when the is to start the HC again +Yeah Ill find some time soon to do up trunk patchesCheers are the patches to fix trunk and the branch. +Setting to since we dont know when will fix it +Sorry Martin I dont know what happend but I uploaded the wrong quickstart +The src packaging will be named where DATE is the date of the buildIll keep an eye on it for the next couple of days then close this out +Done the script now lives in the +I do not know what was the problem +I would also like to have HTML report which will contain these informations +Class should be checked for the same mistake +Androids the same story I think. +test project have applied this locally and am committing to a still to resolve it wont work without a testNG dependency at present breaking most junit tests +Descoping items not being worked on for M into Unknown Fix Version for test for this issue needs to be written to ensure it is confined to the M code base +Which may be next weekend +added item buffering to closed as resolved and released +If you can please with more information +Patch works just fine niceThanks a lot for the applied on and thats fast! Thanks Andrea! +I can login to xircles ok +After that I merged the changes from done after creating the TAG for into the patched version +doesnt say to me bug +Yes Windows XP +Hi KarlEven if it could automatically discover trusted domains I think it best to have the option to manually specify the other domains so we can exclude some if desired and so we can supply usernamepassword etc for those domains we do wantI will stay tuned +Sergey could you refresh review board with the new patch? +Do we need the latest version of CXFFUSE SF with the various patches to get this to work? Heres a test that is failing +When the maanger shutdown method is called it will release all the caches un export the listiners and tell the remote cache to deregister the listener +First part of patch is committed to the code base +This single patch is solely for review and is not to be merged +Oh I see +Also we could do without the attribute cause I really dont see use case for different behavior. +wait for at least or db log files appear under dataamq directory +Committed revision +autobootstraptrue for node is a dup of +Ok Im taking this one +Done +ResultsTests run Failures Errors Skipped +Have removed the Jars from the war to reduce the sizePls inculde all the tuscany jars before deploying +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +This issue has been fixed in CR +The jira bvt test tests the inhetance and overriding of the form bean in simple action +Logged In YES useridI tried using both the same and not the same column name for foreign and primary key columns but it still thinks the is a +XOM carries its own navigator that I maintain and that I dont think shares this bug +Ive tested and it works perfectly +I do plan to go over the exception handling a bit more although I have made pains to try to make any exception due to missing security permission propagate all the way up there are still locations in the code where I ignore them due to the if we introduce checked exception in low level utility classes the theory being that the missing permissions would be detected elsewhere +I wasnt using Eclipse I was using the CLI +IO operations are throws and +Committed revision +Bulk close for to address Windows test failure +The changes look great +entries configssrcplan configssrcplan modulesassemblysrcplan modulesassemblysrcplanI hope this info is helpful +Okay this implements this hierarchy base class with two subclasses +That however gives me a problem +bq +Lars we can actually backport this without the dependency on +Make sense Andy +Software grant for this has been received and recorded +Its broken that users have to flip a configuration flag then stopstart to make a basic fs api method work +Applied to branch. +Every time I made a change I used the Eclipse jar exporter to package the classes into a jar in solrlib +Where do I create this bundle? common +It will take a bit before I find the time to test it +With the exception doesnt occur on a restart with a removed index +This seems to be redundant since thats also tied to +trunk only or also? Tempted to do also since its just in scripts +I fought with thistest a little bit the other day and somehow clean seemed to temporarily get thetest passing +I use the following commandMAHOUTHOMEbinmahout It failed and the error messages are the same as above +Hi Patch is welcome and will help a lot to have a quick fixe to this issue +Heres an initial implementation for online merging along with a unit test +I need to figure out whether this is fixable in our process +They tend to expire and may be recycled +So we either introduce a convention or we need a way for the user to pinpoint which attributes are booleans +However this logic is complicated and I would appreciate your advice about whether this is an acceptable solution +With the change set for this issue applied regionservers shut down right away just after coming up +Thanks Bob! Performance data looks promisingBTW what Java implementation had you used to build yourThis is not HarmonyDRLVM right +I took a quick look at it and also tried running the upgrade suite +committed in . +What I applied to trunk +Is it that is bound to? Could you tell which property in that class is supposed to be bound to element + to Thanks for the review +This is still an issue as pointed out by PatrickI updated srcsitexdoc on the X branch in SVN revision but dont know the steps for getting updates published on. +It was some other synchronization issue in the NHTTP transport which has been fixed +It shows the lines almost touch but not quite showing meters +Got it +Ive got this error with JBT on Linux bitI just used JBT for a while and then opened web page from jsf kickstart project +Committed revision Thanks for the patch Keith and especially for including a test! +Closing. +Added some optimizations +somehow I sent a version of this patch with some changes but not others? Let me resend I have no idea how that happened + +Hi LauraYes those still apply to the current release + on going through the node type API as its not optimized for the kinds of access patterns needed here +A comment should really be added under URL Constructors that the because I almost just posted this same bug until I realized what the problem was. +If this is happening on iOS and Android its very likely that you have CSS thats preventing the input from functioning properly on +Ok its not the expected result but Its seems to be in accordance with WFS standardThanks +See linked issue +I would say that we have seen a number of cases where the jar tests broke even though the classpath tests succeeded +weve had requests from users who wish to disable browsing of the filesystem in the web UI completelyHarsh what is the reason for this +Probably by use through some of the plugins +After having a look at the it looks like there is some kind of escaping not sure if it works like expected? Maybe it is just disabled by default +So in terms of that reported here it is as designednot a bugThe issue to solve here is then just the batch file one right +Its just cqlsh assuming a little too much +Can you a this patch under the apache licenseb andor comment on Davids commentc andor let me know when its ok to agree with Davids comment as well and Ill open a jira up for the schema to change. +bq +mavenrepository folder but still the error is thrown +Id add it to the item for. +Bulk closing stale resolved +For the two you didnt fix why dont you open a separate JIRA so that you can resolve this one with the issues you addressed + +added to trunk and x +Therefore Create a service which will verify existing active records before creating a new batch for Payments +We will need to step through a debugger to track where the invalid union index is coming from +So I could also place them in +so pls close it +Thatss a good point +Well I dont agree using your analogy svn is the authors draft manuscript and the bookrelease artifacts +I guess I was it +The discussion on this topic has really shifted since was introduced +I have a snapshot where Im using the for the string replacement can you check if that helps in any way? I dont see how the jxr plugin is connected with this +Patch for Chukwa Admin GuideNote No new test code changes to documentation only +When submitting this issue I thought that the weak object in a was the value of the map +This problem lives in the generated proxy +Changed the configuration defined in the AS kit to reference the same modeshape JAAS application policy already defined in the file +Im quite sure there are still more uncoverdissues in connection to the XSLT processing of unpredictable user files +After someone change the codes for getting wsdl file from path to uri +Be nice if we could tackle this for you ask me this bug isnt a minor bug +Bulk close for Solr +You cant exclude all of target and then include a subdirectory +Moved to due to the fact that we dont have a homogeneous cluster of boxes +Approved for. +made public +Should I treat the unsupported flag the same as private +The application is working like before at least me the testers and the customer did not notice any problems with this application +From the example above as a new user I was also surprised that when constructing a new with appendContent true throws a when no currently existsI expected behavior similar to s appendA note about this in might also be helpful to others. +This is the full patch after adding optimizations in the code that expected to find insteadThis should fix the bug without the loss of performance of a full copy every time +Why is it problem if methods are publicSee the discussion in +These files should help as a starting point + unscheduled issues to the next release +apachecassandraiosstable +Thanks for your efforts. +Its called +Committed +Can we do something about itAt least theres the workaround of using ranges or lists to group certain cases in the same case but I think this should be fixed anyway +Committed to trunk! +Ill look into doing that +Please feel free to install Spring IDE PR from the developer update site +This is fixed there was a logic error which was rolling back the multipart request when there were no errors which caused the files to be cleared. +As this is just a debugging aid I dont want to guarantee that itll be around forever +The patch no longer builds a hadoop bundle instead it copies all the jars with the plugin and refers to them via the inThe jar copying happens during the initialize phase so that the libs are available when compilation kicks in +Hi Henkit is ok Now can you heck pleaseBest regardsMEFTEH Henkit is ok Now can you heck pleaseBest regardsMEFTEH Ramzi +Thanks Ravi! +Not sure if this is the right place to ask this question or not Do we have any expected release schedule date for RC +fixed in Roller trunk revision +the patch +I think I saw the problem with the patch +Glad to be part of that recordGood news is that issue is solved +merge with to Federation branch +trunk also has no dependencies to but itself still remains not to surprise trunk users +Is on staging +HiThe pluginname columns of the rules table PHP LanguageBut the key returns PHPChanging the value of the key fix the bugSee the patch in attachmentI also saw that the Unit test is not working for I fixed it in the patch alsoBut the test is not good it uses the same constant as the class for assertion so the test will pass for any value of the keyInstead the test should verify with the key of the constant The code need a serious cleanup about the constantsWhich one should we useWhy there is a duplicated key in? Why not using +And theres no way to figure out which of those spaces were in the original text and which were added by XML conversion +This has been added to the foundation module pull request +Clearing the negative caches now on bundle resolved events fixed in revision +A job should be able to survive a single machine failure right +Confirmed fixed in automated QA +Please verify that this still works +There is still restart data being stored in the instance however the removal of which will be tracked in a separate issue +Permaine Im pretty sure the EAP guys can handle this added the hornetq same has to be done on JBM assign it back to me once the exception is fixed Ill then add the code +This issue is closed now +Maybe you could use the machinename htttport to uniquely identify a datanode +Claus do you think the Spring guys will fix this? I think introducing binary incompatibility in a minor version is not a no go IMO +snapshot for listing all snapshots of the directory tmp but users will get exception if they treat +Please go ahead and upgrade to use Jetty +Robert I was attempting to reproduce your problem given your description but was not able to +The was lost because I killed it since it was spewing status report every secondSo the problem happend before that +a lot for the reviews Todd and Tom +So we want to limit the scope of the to CQL +This might do the trick +So you are really subtracting a big disk from a smaller one +I can sort of imagine how a race might occur between the error and the case +Yes Im able to reproduce this on trunk +This test is written to test for REST POST as TextXML +Username bcotton +Thank you +Yeah Jonathan that sounds like a really good approach we get all the benefits from splitting deletes and puts and dont have to pay the cost of doing that very niceOn the whole putgetdelete issueWhen looking for a value in we have lists that need to be compared toeach other the list of data da in the storefile the list of keys k to lookfor and the list of deletes deToday we compare every da with every k and every match with every de as far asI can tell +Spring config changes resolved the problem +This is done by design +this looks good +And use a role with this up having that role +I forgot to mention that the object under testing is a Person that has an Address field. +Thanks Cheyenne +Alexander seems this issue is duplicated with would you please verify if this is fully resovled as you expected now +thanks Mridul +At present we need to know how to reproduce the issue as well as steps to test it. +Duplicated all issues that have been in resolved state for more than one month without further comments to closed status +I just committed this. +This is an updated patchIt implements the maven work around that I described in my earlier comment +Closing for it possible to backport this issue on a? +So the file was opened +The reason of problem iswhen we expand addres item the gets items children here all works it finds the item which represents out address component +Granted the newly rewritten title I chose for this issue merge segments to sort them is limited to just making a +The issue is we are inconsistent about synchronizing around the parent usage +Typically when folks have wanted externalized access to nested beans we tell them thats when they should declare it as +Thanks works great. +Well ive tried to make the stylesheet as simple as posible +Reopening to flag the issue as Resolved instead of Closed +Backport to branch in revision +Thanks Tomaz I can confirm that the console works with your changes and allows the editing of the properties now as well. +I tried to reason what would happen if we went in the other direction ignoring mvcc and I could only see issues possibly returning older CellsIf it happens we need to compare minus mvcc lets make a comparator that does that for those cases +project that was added wst nature +If you do get this as an issue can you supply a patch +Since this has been implemented this issue can be closed +Just to clarify we need an attribute in +This was intentionally introduced by from my POV this is a horrible change because If your main artifact IS the one that you are describing with the assembly descriptor this is the only way to create it +I recently extensively tested application deployment and removal from the with no problems. +For this testcase we disable speculative execution and see if all the attempt are finished before finishes +Please assign to me +Yeah I guess you are right with the numbers +Resolved +now has cc by itself +When nodemanager gets below error It is exiting with non zero exit codeIs the NM also dying due to this issue? +The file system connector needs the ability to create Binary objects from Files without having to load the contents of the file into memory +Thats because osxcocoa is not working natively with jrubyOne option is to find out whether there is an additional gem for that in the jruby world +Screenshot showing broken need to install graphviz package and dot program in order to render that image +Changing order will break backward comp +Once Infinispan has upgraded to a version of Netty that fixes this threading issues update both components +Ok next naive question do I need any special permissions to update this confluence page I tried to log in with my asf jira creds and was not able to do so +Please verify +I see it down in localputCommitted to branch and trunk. +Looks goodOne comment rather than referring to the default join as regular I think we should refer to it as hash or symmetric hash since these accurately describe how it works +poms cant be changed well keep this open for future Dan tell JOTM and we close this issue. +The code looks good and the existing tests all seem to pass +Patch for trunk attached +Ill add KEY file to release directory until we add release artifactThanks +Robert do you know why the bug you provided the link does not find anything +currently the exact format is used. +un shouldnt just reply null +I dont see how this can happen +What Data integrity validationCheck block data and meta data +Thanks Daniel +In the future make sure all patches are created from trunk +ntresource is not ntunstructured and does not have any residual property definition hence this exception is correctTo solve your issue you can remove the specific properties before removing the mixin +This does not include a new hadoop jar +Couple of other comments I missedDuring design we wanted to ensure cluster ID is unique to avoid accidentally naming two clusters with the same cluster ID +Sorry the server version not jclouds +will be added as soon as jsf will be releases +The only location of +I would make the default to care about order and allow you to turn it off by setting a boolean in runTestI agree that most mappers wont care about order except for mappers in map only jobs but it would be confusing to have runTest default to ignoring order for mappers and caring about order in the rest of the api +The general steps for this Allow the error statement to specify a sqlstate and chain exceptions add exception handling to compound statements this could be plsql like using an exception clause or like ANSI sql using handler declarations ensure that all Teiid processing exceptions have proper state codesNice to have features add the ability to emit warnings add the ability to check for warnings procedure logic mnemonics for common state codes +Im all set to look at this but Im not sure how to reproduce it +No if there are no exporters that is an error hence it is the Run button that is enabled wrongly +Assigned to footer has been added to the +Fixed on and closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Please help to review it again +So maybe someone can close this bugThanksLofi +see it +Can you try an svn update regenerate the patch and then create the review one more time +Thanks Carl! +Applied as part of +bq +Im afraid of there is not much docs about jboss logging yet but it pretty simplejust drop the logging libraries youd like to use either logj or slfj to the classpath as usual and jboss logging will take care of the restD +This last patch seems to work great but the only odd thing is that it always reassigns ROOT and +The Sanitizer is not and was never intended to provide +Yeah the patch which is based on trunk should be OK +Patch shortly +Next we deployed our useless web service in same procedure as the axis war fileSince we got the wsdl oft the web service displayed we tried to execute attack vector shown above +This was run on a single CPU machine +Attaching a sql script to repro this is definitely a Normal not a regression but ugly +When I make a connection from everything works. +Closing resolvedwont fix issues. +My test files and patchhack +This issue is of the way toward a tuning guide for HDFS installs +Copy pdb files patch attached +Or do you think this could be caused by a certificate change at? Firefox has two of their certs +No actually I think youre probably correct but this code though part of the API seems to be virtually unexercised so I have to create a testcase for and go from there +Which is fine since VM does not make use of itBut VM might require an older field that the new field replaced and VM may then crash in an unpredictable way +get but it get property descriptors +The affected functional test passes tooThanks for the fix! +My diagnosis was wrong +Please apply this patch for inclusion in and the next release +its not a problem with escaping +i cant help on this point +It is true that we dont plan to maintain eyebrowse in parallel with modmbox HOWEVER we are working to transparently support eyebrowse in order to preserve existing links. +Running run Failures Errors Skipped Time elapsed resolved issues for already released versions +Are NULL values allowed for a sorted columnI think that the answer is Yes because insertload does not complain about null values +Then when we get an unknown one called on our filter we have a way to crawl to the top and build up the mappings on our sideIf a segment reader can exist under more than one parent at the same time though it would cause issues for a solution like that +Interesting that the Submit patch dialog doesnt actually support submitting patches +Its possible that this has been resolved already +The requested behaviour actually used to be the default behavour +failed tests manually for multiple times each +Same issue is for dir toolsconsolemanagementsrcmainresources +Well I do see the problem with jetty but doubt the other dependencies are that problematic +Confirmed it against older version of too +bq +I thought I could move this into the correct component by cloning it +Contingent to hudsons this patch is ready to commit +Lets get the journal service implementation done first and then do the refactoring later +See the linked issue for some background +Even though Cassandra is in quite a niche market I think theres a high risk of losing potential users simply because they get stuck on the installation procedure +In more delicate terms I was trying to say that I dont think Diego can help here because a Im not sure how I can reduce this problem to a small standalone app and b the code that I wrote that DOES replicate this is not sitting in an repo that Diego can run +for and +It could sound like a corner case but it could happen on a busy application as well +This file contains and its mapping file +The crude checks which I perform in right now cant be performed in the upgrade tests easily because I get a when I try to retrieve a out of a database +added the prefix to incrementer names and used custom prefix in was some very impressive turnaround +Please can you attach the wsdl that is giving you the problem. +Only need create a not null restriction about group believe that it would be working on this improvement or just perform a single validation so as not to allow the removal of the default project group +Patch applied +reopen if more work is needed +If a client add jar files for initializing hive server makes a new classloader and attach it to thread as a context loader +I upgraded from version to and the TC failure upgraded the ConnectorJ library from version to and the problem described by Ralf man whats going on here? Seems the behaviour is not very stable from version to version +This patch is meant for and previous patch was intended for branches +Then you could dedicate say MB to indexingregardless of document size +I have replaced this call +Derbyall and ran cleanly with the patch +Released tag from branch jaxr +We should also move to the +but on the other hand i wouldnt want to have strange features to be releasedwhich we then have to stick with for compatibility reasonscan you please make sure that this feature is not released next week unless we found a consensusthanks +This is covered by the port of to Labs infrastructure +I also planto bump this issue from being a Minor to atleast a Major or higher +Release Notes for will not be at this time +This never got very far and can be deleted +As part of this issue we should also settle on a wiring format to specify the RPC iface inPersonally Im biased towards Avro which Ive been using for a long time +AFAIU Bad Request would mean the HTTP syntax is broken +added an item for doc addition to the user functions +v vProbably my favorite +i think this issue cannot be reproduced anymore +The proper fix for is for JNDI to use only one channel instead of opening a channel per context. +Also on machines that support it an eventfd would supposedly be faster than a pipe +I can take a look at that once I get this whole committer thing sorted out +Fixes this bug creating sub I dont think you meant to mark this as fixed we do that after the issue has the patches committed +Hi JacopoIt is true I am testing billing accounts and order payments behavior +Im happy with making this a duplicate or dependency of though +Eh never mind +When writing the patch I had in mind the next feature with a very similar implementationI want to implement an database +This problem may be hard in general +Thanks Mithun +At least if the classes were public it would truly reflect how they were being accessed +the ant script to build without using Maven will failWe should probably remove the file +It makes no sense to create an array that large for every multi term query especially when we direct users to possibly use +Sorry duplicated with with +Alex I think we need to investigate whether or not this can be fixed by modifying the file first +However I used for maven and for subversion +Is for you also a static generation possible +They may be in use on xml relational input tables but if those are changed to also be procedures then theres no issue there either. +Now focuses on the text which highlights the region in the source panel of the BPEL editor +test test solution link +Thanks Shai new patch +patch to add extra drools features extra plugins update update x add test plugin update map from MR for missing assemblers in customTargetsbuild tests unpack during test phase and run single test log from latest build failure missing libraries prevent compilation of added to and updated to copy them to lib folder +BULK EDIT These issues are open to be picked up +had bad fix version +documentation fix to comply with the standard documentation standard +Closed after release. +I just committed this +But the code generation engine is not aware of how to generate code for out only +Please find attached the archetype for generating an based CXF WSDL first service +Documentation only change so no tests +Tests are in progress please review +The clearance and clearance fields should be the same data typeIn summary this could be greatly simplified and in the process make it more flexible +ignore my comment about why the static I see why +The implementation should retrieve the and parse the specified version + + pending you sure that the test is not already covered by and +EAP is also protected by default we just unprotect the version bundled with JBDS to allow for startstoppublishing to work out of the box +The JMS Spec does not allow +The Commons Digester folks have been nice enough to consider my patch as a workaround. +Scott can you please verify the petStore build and let me know whether we can close this +Docscleanup patch looks good Ill commit to the branch! Thanks +Probably one of the following fixes the problem or. +It generates an ambiguous reference error +I cant directly do anything about the javadocs so Ill leave that for Mike + +It only concerns versions from toolsusageactionwsstartup to toolsusageactionwsstartupI have completely forgotten that this bug appeared in earlier versions. +This will solve JIRA also +for thesecases the cumPro should just be the probability of this query number +This is however unrelated to both and +I havent spent much time with the SRV code but the patch looks basically OK +Hey I just wanted to help no need to get angry +Have used STARHowever winrar in windows does not support extracting tar file of GB +Patched version that display the proper was actually in the core xhr transferring of incoming node atttributes. +Heshe can write a bot tokeep uploading no matter how big the size limit is out there times of Mbis still Gb +push out to +Added some javadocs toAdded Matcher constructor to and extended the test for this +I think the only rationale is to serve as a backstop protecting against invalid reads +OK do you mean REVALIDATE session will fail before the new leader come up because sessionTracker in is NULL +WFM +I meant the findbugs addon +Ill change atomiccasptrsz implementation in my new patch to use apratomiccasptr +The patch added new files to src instead of to src +Done in revision +patch to replace by aThis should provide better concurrency but Im not sure if the cache eviction policy works the same way as the LRU map +Tested +In my opinion the createInput should get both files then the directory can wrap correctly +See also +Attachment has been added with description Patch against svn HEAD to add regionName to the applied +In the example described above say that is rolled hourly in Cluster c is rolled daily and monthly on Cluster C is rolled daily monthly quarterly and yearly on Cluster C Also Customer table is available in cluster H All the location tables are available in HDFS cluster CForgot to mention that along with timely rolling is rolled at dimension state also +bug in JIRA +See +Sebastian is working on this +Its just easier to place it here as test relies on access to master internal state heavily +Related to however the NPE is now thrown further down the stack +cc files that need to be part of the command because the X source file contains multiple X classes +ATS was then working correctly after I did this +Thanks for reply +Elena Alexei are you ok with such resolution +After all it need synchronize namespace operations to BNAnd I also think the under replicated list may be a trouble if NN and BN are managing blocks seperately +release audit not by this patch +However the shapefile reader should be more forgiving +I am resolving this as Later so that QA can try out the recommendations +is thrown instead of sending an error response. +Andrew rightly pointed out two threads could be interfering with each other in that code +Apart from these modules everything else builds successfully without an internet connection +Ive updated the original issue text because I think it should now refer to Project Visage which is s successor +The patch probably also needs to update the canon for refActions +Some of the arguments might break the behavior and Im not sure that users should have this kind of power +I think this is a fairly change and it would be unfortunate if people hit this in production +Patch integrated in rev +Taro can you post comparison screenshots of what this looks like in Pivot and in Notepad on your PC? So we can compare resultsThanks +Vladimir I think the point is elsewhere none of packages has such information manifested in jar +Ah that might be it then +Addendum comments out the assertion that region server holds more than regionsThe assertion is not essential to commit flushing test +I couldnt find would this be? If so here is a part of that file from the relevant time frame +This patch simply simply creates the AgentPlatform and sets the capabilities for the agent appropriately +This issue is still being seen with AS rev EMBJOPR rev be fixed for real now RHQ r +bulk defer of unresolved bugs to +So please update the same on the wikiI had requested access to confluence to edit the wiki template to match the landing page +returned a new instance each time that could safely be modified +Were going to remove the current implementation altogether +If you have a text element with one child the textPaththen the x attribute from the text will be applied to the first characterof the textPathIf this was not your problem then please either reopen the bug or createa new one with a test case. +I am not sure however how to test setting the executor service on the contain as it wasnt done on the current tests + for applied the patch on trunk but this does not workAttaching another patch that works +Create Dynamic Web Project +The word example in the title of this issue threw me off +Committed +bq +Im in process now should be done in a little bitDone committed the krko switch at r on brancheslucene +Let me know if this works for you andor if there are other review concerns to be addressed +Should be verified in gatein master and reopened if still valid +Removed the contents of +Both several projects at Apache do that as well as the Maven project itself +So fixing jsps only is enough at this time. +should use to create broker properties +Attaching +using the wait option jsvc returns if the class is not found. + has been marked as a duplicate of this bug +Why dont we open a separate issue and track of there +In that case I cant really help because I cant add custom metadata to the Info dictionary with FOP +you Oleg! Much appreciated +bulk close of resolved issues. +Thanx for the fix Regarding the authentication failure how do I specify the password? in the connection url +Attached schema document and perl script that demonstrate the problem +Brian that archetype defines +Committed to branch +We did this change for +Looks like it has to be changed as in the given patch +In M is actually based off the interface but the integration is not in its final form +Folks it would be great if we have a consolidated document that describes the design and some details +OK making all these folks admins is dorked +This needs to be. + +Heres a new version ofThe performance should be better compared to the last versionIm not sure the test case is very representative though +Thank you! +I committed it to branchI agree that the test for this will take some work but we will need it once we start exposing Storage Types to applications +I only have right nowThis newest version should be pretty self explanatory +The link you gave in fact shows that it is there +I would like to go ahead and checkin the change disabling security manager for the jars and file another issue to correct that problem +Yes please delete it I cannot manage it anymore anymway +Moving this issue to as there has been no activity in some time and it is not clearly relevant any more +This breaks index compatibility and query compatibility +This is clever cool +Closing issue fixed in released version. +Fixed in +Attached patch implements a compressHTML property in +First contribution will be this file and then we can go from there + test for this issue will be provided the patch is created as you expected can you review it when you have a chance +try and apply at first then apply other patches to nio luni and patternsets respectively +Also capacity calculation of where is used has no relationship to UUID length +Ive determined that the cause of this problem is the presence of inside the lib directory which is a dependency of Drools +Its a bug +You specifically saidbq +Hi Jan if GB is reasonable to ask we would like to have atleast that much +The new rename database functionality moves the database directory on HDFS and changes its related metadata entities +Currently the code sets this TTL on JMS Producer object +The patch has been committed to trunk need Jervis to take a look and test if it works +As you say the immediate after an looks like a jetst issue +Closing all resolved tickets from or older +Could you get your reviews and s +A Mentor hasresponsibilities toward the Incubator PMC the Sponsor and the community ofthe assigned Podling +The container performs the commit protocol before themethod result is sent to the client +Shiva how does this change handle older Geronimo servers like which do not have this new code? the minimal server assemblies which do not have the console or Plan Creator installed custom assemblies that do not have the console includedIs there a way you can use JMX instead +Committed to trunk +This breaks all existing persistence managers implementations +Patch to fix the OSGI exports +Allright committed an Ant build file for CPAPTF and a Maven POM for CPAPTF as well +If nobody complains I will commit it in some days + +how i can add truetype fonts to a property file like you? +Got enough of a start for this +But you cant create or alter a page until added to a orSo nothing wrong therexmlgraphics is done +Assuming Hudson comes back OK I am a on this +We can add tests and improve it as we they are needed when not executing but using the read in the embedder +I confirm this +Verified changes for + fixed +Sorry about the noise +No sample provided or further detail offered so closing the issue +Binaries are m working on enhancing the build script to provide an option to aclocalautoconfautomakeetc that should cure this type of problem +Of course I just want to know if we should give a message from ui side? If yes I can take a look on it if can be fixed for this release +The is no longer leaked. +Feel free to assign back to me if this form of work gets picked up anytime soon. +If has not been fixedhow can we change the value of this +It sure would make many things easierThe combination of auxpath classpath and jar list makes this stuff pretty opaque +If no objections arrive within hours from now I will commit it to branchand HEAD then +modified +Also note that the Variable name text field is filled in with the bladeServer variable but that is not obviously binding to the +added property and wicketmessage to html transition all resolved issue that did not see any further comment in the last month to closed status +We are removing deadserver nodes to avoid double assignments from SSH and master startup +This guide is being overhauled for inclusion in EAP and this JIRA no longer applies. +Otherwise you need a dynamic partition insert for the merge which may generate multiple small files for a partition again +Im currently toying with but not sure if that will solve my issues +Also are you running these two simple apps on or Portal +stack has been committed +patch v +I think it is safe to commit it to +I simply added entries into the doc files for the modified components +I havnt closed the issue because I hope to keep trying but I dont have anything great at the moment +Edits are propagated from +Patch applied at r +Im running on a Japanese OS +Updated patch from Patch patch from RB +Im marking this as wont fix because the original issue isnt valid +We can only update the boilerpipe dependency once the new version hits Maven Central +attaching a patch to fix the incorrect handling of roles with umlauts in the identity admin the fix to Branch not needed in Branch where the admin portlet is based on the Sun RI +Still holds goodAny further comments on patch anyone? This has fairly good user demand even today given that lacks extensibility due to various static methods +If you want to have these changes in coming release attach a patch for review and we manage it if it is dont set Fix version field it is managed by project lead and developers only +During the for release would be a good time as everybody should have worked their changes into the code by thenIIRC tidy removed the whitespace following an anchor element +resetting P open bugs to P pending further old one but seemingly still valid today +Attaching another patch with the above review commentsDidnt write a separate test it seemed to me that its not adding any value for it is already being indirectly incorporated in the other two independent tests that verify the tasks limits and the TT limits +Will try to decide how to filter the jars for m +Heck make it Depends even after all there is a runtime dependency on those packages and an is surely an experience for someone whod like to simply test Spring Python out +Is this ticket still relevant Simon +committed fix in r +For example Camel Servlet will set Request URI as webapp but also set the base URI as servlet +Ive tried to clean the patch as I had some local changes +but effective +Re joint at the code in HEAD it appears that this option is still enabled for the case where you want jointCompilation but dont want to bother specifying the nested javac debug switch for javacadded that to the example on the dont see the jointCompilation attribute mentioned in the doco any more +AFIK is still almost x the users of +Not really sure how to get a test that will eventually succeed on a connection retry +I dont know what should keep you from doing it +So Im not sure how anyone is running with clover and sonar +The default value of seconds remains the same as before +Scott can provide some guidance also +Attached patch against branch to fix this problem +You then try to fill the database and this fails because you didnt give it enough memory +retrying wont fix anything +Fix for the fix is upstream +but Im not sure what you mean by complicated conditionals +Committed in trunk +Nope +IsaacCan you please make the changes mentioned above to the EWS Release NotesPing me on irc if you have any wasnt able to find a URL for supported JDK for RHEL +Still an issue +It looks good to me it compiles without problems and tests run fine +I thought the upgrade tests currently only tested upgrade from releases on earlier branches +In my humble opinion the impact of samples included with Geronimo would be higher especially to newcomers if they contained less or no external references +I think you are fixing it the wrong way +Using connections worked also as something like a stress test and utilized the database much more +We probably wont ever fix this and I think its a bit as to whether map and collect are the same exact logic or not +Manually integration tests +PR pushed to master and with JBT +Please feel free to reassign +So I dont see any problem in having this feature +That would be good enough of errors +Thanks +blurb and link is some magic that doesnt seem to be easily available fwiw +We should get people from other distributions involved in the first place so we can all agree on a solution and implement it +Sounds like a good idea I think +expected behavior minor now fixed glitch in of livereload thus ersolving this specific issue as rejected. +For example see +Integrated into and trunk +I expect the problem here is that we assume stdout will always be an IO object and so try to cast it to for use +Since an implementation will have to try SSL before it tries plaintext this latency will hit all clients talking to plaintext servers +Now regarding a BPEL developers one simply jars up a bpel project and creates an manifest and deploys the jar to servicemix +This is now in place for should note this is in place for INCOMING message on the provider which is probably the desired result +Thanks Gary then I update this issue to resolved +Just something to think about that the mechanism for sending the resources to the container isnt just dropping an XML file into the deploy directory +Thanks Joydeep +It might be now a different timing of Xalan or whatever or the problem is really fixedIf it occurs again I will reopen this bug +If you are still having issues please create a new JIRA and attach a test case +Ok it seems this bug is gone +link to ODE ticket set +Robbie could you please review and apply an attached patch with fixes +Whatever we do we must ensure that this is an optional and not the default behavior +I was intending this to be a quick and dirty fix that would members of as transient and serialize alladdedloaded objects and the identities of unloaded objects and deleted custom readObject and writeObject is also required +FWIW the test worked for me with the changed order +Simply installs the POM into the local repository +Hey GavinDo you know what ver of Hibernate the fix is in +to trunk +Committers can you please review +MikhailBelow is the command line option Xmsm Xmxm Xemopt XXMINORGENFORWARDPOOL XXmI use this configuration to run it failsThanks +will have a close look at your latest patch soon +Ill create a pull request. +all guide are guides are checked +Additional experimentation showed noproblem in so its probably fixed +I saw a mail stating that lib is in has been implemented +Patch applied +it should be a OR and an AND +What were you thinking about wrt more improvements +test cases covered will continue in analysing and covering new test cases +settings and remove the update the maven configuration of in JBDS +Robert would you mind updating the patch to mask just the password +Ideally standardise this +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Could this issue be closed now as Cannot reproduce? +also what version of the broker are you running against +Sometimes suggestions like this can encourage us to look at things from a different perspective which may often be a valid one and lead to improvements in the design +Protobug codec is available in trunk now +I would expect a poll to only work if there is something pending in the retroactive buffer +Feel free to commit it well review it once it is in +ryan Im trying to understand why you think we might need seqids +Attached created reviewboard entry for review +Issue of found the issue in beta release reported as Affects Versions beta due to the reason there is no option as beta +Assigning to mental +Fixed with s inclusion in the TPrequirements latest in +Bryan I am also inclined towards Option +Patch looks good +Hi Id like to resolve the issue as Wont Fix or Not a Problem given that the way to obtain the typed representations of FIQL expressions does already exist as shown in my previous comment +read previous two comments on moved too quick +Eclipse project containing reproducer +Thanks +srcmainjavaorgapachewicketexamplesajaxbuiltin +Glens new did not produce a different WSDL and so did not help +Scroll lock icon was added to the toolbar +As far as I know this is not a regression +Why move everything out of docstrings? Nice to have it inline IMO +bq +Patch applied. +I have downloaded and will hit my machines hard over the next few days with an older version of our software which had the issue reported in this bug record +HiJust want to confirm that were referring to disabling cookies using the session property of the here and that the application will fall back to URL rewriting only when this property is set to false +but theres not enough output for me to tell what it is +I created the user and have changed its email address using my confluence admin karma. +Mark ok +Given Teiid Designer is building fine this was prolly fixed ages ago. +attaching reproducer +Hey guysAs we understood it Pig wasnt due for quite some time +The problem is that the version history is not exported +Dont misquote me please +Everything is mostly dealt with here Ill raise new issues for the couple of things that Ive found in testing. +Hi Tsuyoshi OZAWA +Bulk close for the release +Here is a patch which loads the value of the format attribute as a plugin if it doesnt match the existing shortcuts of solr or wordnet +renames some methodsvariables +Hi AndriyRe the extra parameter info +The default could be configured per and overridden by specific requests +Some additional info I am working on Lucene Java Core so I would like to have access to the Lucene Java Hudson Solaris node +Closing this +Using sets instead of bags leads to cartesian products being fetchedIve just tried switching to sets instead of bags and although the query now executes its terrible on a medium to large dataset as it creates a cartesian product +Grab is not supported in a Grails application the fact it it worked in was a bug +Sorry I didnt notice the currencytimezone error just the selectselector change. +This type of uncertainty made me slightly hesitant to fire ahead with committing some other issues +Thanks to Simon Kitching. +Filipedecodepacket looks like its what we want +This issue seems to be more of a discussion than an actionable item +Committed to trunk +Sorry about that Mikhail +Just added a dependency on bug as clearly this problem would be resolved without any work if we committed bug +The goal was not to seek and scan terabytes of data and attempt to visualize report at run time +backported to in revision closing for +Thanks for the report! I guess I wont patch Xalan at this point then +No should we disable itsorry not sure what the final result of the massive thread was +I might be able to do it sooner if you need to offload some work +For Buildr tasks that require libraries to be there why not assume that a constant declaring the spec must just be +Arne can you write a patch on the spec? I can review and apply then +Good catch +Need now to update the ticket for Jboss Documentation teamSure the doc needs updatePS Is it normalOK that in a file there is clear text and an other the encrypted form or is it a bug? I think its ok +The original concept was based on providing a wrapper around the JNLP persistence API +Attaching a new patch that fixes a javadoc warning +Sorry about that! +We could only encode filenames and directory names but not the entire string +Should we ever fully automate our test suites then this is worth revisiting then but no such plans exist at present +Fixed in trunk +Bulk closing stale resolved issues +Now instead of the NPE it will blow up with a which will tell us the type declaration and the property name involved +Ok thanks again Michal! Ill apply the patch +Logj formats the stacktrace using either the or +Sscoping items for early release of Java Eclipse Mgt Console +fixed with rev the fixed issues. +Closing since the documentation was improved. +Thanks for the reviews and thanks for the rebase Lars. +Thats why I kept this issue open +After giving it some thought and testing the patch I decided to apply main gripe with it is it opens cheatsheets copied to a project in the workspace +As long as you have spatialite installed on the system it should work +Ran multiple times and did not failThanks for the patch ChinnaThanks for the review Ted Lars and Anoop. +Full build log +There would also have to be a lower bound to prevent compaction from stopping completely +Thanks for your contributionPlease do a submit patch when you attach a patch so that we get it in the patch submitted queue +Unfortunately no I see the same issue in JBT too +This is supported in Spring already with the new MVC support classes specifically the which replaces the +Uploading the updated patch +Kalle has promised us to prepare a new patch which would do this betterSo lets see what he comes up withregards theres some pressure on our side as well to get the patch in into the mainline +Sorry forgot about that issueAttached new Version of Patch against current trunkStill NPE because of missing fieldMappings +Will we get an update in time for Spring RCfinal? to fix version when we will be able to pick up the release. +I think you have a misunderstanding of the reserved parameter +One way to get a valid URI in such cases would be to expand the glob pattern in file name but unfortunately that code is not exposed through a public API +While researching the problem I learned that since Thrift follows the Apple naming convention for its generated gettersetter methods this patch was actually not necessary dot syntax will map to these methods automatically +We should still clarify this in the new docs. +Sucks to be using IE I guess +has a patch attached +Ill attached + +Hi Todd ready to continue this please catch me on irc +These use custom copies of so that node starts the DB and node uses the copy on node +can you attach your wsdl adb infact supports values for the boolean +In this case it appears that the message key will be translated and the message interpolated immediately upon adding the message to the system which would fix the issue I have +If it does happen I think we should just break everything and then the person running things should set the configs correctly to limit the max request size the server will accept and the max fetch size for the client +I never set this to true on any server so Id be very annoyed if it was enabled by default +For some reason I always get this error message when I start the instance of DB +I can see that all specs pass locally with GA +Please verify +Do you want to open a new issue on that as you already want to close this one +We reproduced it once more yesterday with all debug enabled +What kind of Boundaries hasnt been considered in test suite +Patches were too broad and have gone stale +rather thanSet JAVAHOME to the root of your Java installation +Apologies +The repository includes the source as additional test has been added so that the logj factory is validated both without and with the in the war has been tested with the commons logging package filter removed from the so the filter has been dropped as well. +Thanks Knut +Some questions on your comments When you say consistent way do you mean across Hadoop configuration and the Resource manager configuration? That we define a similar format to represent both of them? Likewise can you also explain a bit more about the decoupling of configuration source from the implementation +bulk assign all unscheduled issues to next release +It seems that the query is limited to rows valuebefore collapsing is doneDoes anybody encounter a similar problemCheersclausi +I already explained reason in one of my previous comments! As a way to find from where this problem comes since we dont have a good ideas for now. +If we have any further issue Ill let you knowThanks +Fine with of a bulk update all resolutions changed to done please review history to original resolution type +feel free to bump out. +Committed revision +I dont think there is time to verify whether these work in for +Instead of scanning +Thanks Prasad! +Issues called by dataTable bug not by EL enhancement + +May I be overloading the server with too many scanners by using that multiGet case +This should be resolved with be verified in gatein master and reopened if still valid +Defer all unresolved issues from to +This is fixed already +Verified with Revision OK +Hmm +It will be good to convert all existing Kafka metrics to the new metrics ticket blocks metrics collection and graphing +thanks for adding the checks maybe now it will expose the real problem +Added logic to heuristically select an available index to perform sorting on +Problem solved. +See new Step. +The maven filtering component fixes the issue +Upload patch for +In trunk due to the single line which is removed in the patch logj is only included during the test phase which means that the script is broken when run from the distribution +I would say only allow deleting it if it is empty and there are no other groups +Looks like this was actually fixed in. +HiWhat is the intended funcionality here? The forums have a page of their own so what is the purpose of the CMS menuIf someone has a good idea for the services provided in the CMS I could try to help moving it forwardRegardss the patch to fix the error I have no idea what any of the code does but hey at least we have one less issue to worry aboutRegardsthanks your patch is in rev +This issue on in axisc +I think Ive fixed this +but we have for years trained people that a line like the first screen shot is the last thing people will see +Limited functionality +I dont have a Ubuntu available now is there someone who has a Ubuntu installed who can check what lives in its bin to see if we can figure this out out +Updated readme with QE recommendations +AndrzejJLS Treatment of double and longWrites and reads of volatile long and double values are always atomic +Patch which should fix the issue +Could this be achieved as easily if the dependency is in a different Configuration class +Didnt realize that a week was a actual week and not a work week +Changed to return null for root paths ancestor and updated to account for a path having a null parent +Marked the property as +Reopen if this happens again. +Feel free to suggest why not +Thanks +Its a reasonable approachThe test failure is unrelated and most likely caused by something like +The previous patch missed a scenario in replaceChild +Do you have any ideas +Sample process describing the problem +changes look with ran and passed core tests +I had to add a get to to make it as reusable as possible the change in various validators and converters then become trivial +I am not able to replicate your issue when running the is run successfully and I can use operations which touch classes +re right sorry for its fixed in I apologize and we can close this issue. +Checking in generated code as primary source does not feel right +Committed revision +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I could be wrong but I thought no one can create anything in SYS schema no matter what kind of authorization they might haveps Laura I hope you dont mind me sending review comments per page rather than one compiled big comment +Then Ill mainly usetest +This is now in rev +I did not sure is the issue blocker cause I reproduce it on dev +Of course if one starts to dig deeper norms actually does have a physical meaning in the index so preferring a label with norms in it seems completely reasonableTheres also history to consider when you change the name of something you cut the link to the past in search engines and in the memories of many developersAs it relates to Solr I dont care so much since it makes sense for the Solr schema to isolate these changes and stick with omitNorms regardless +One option would be to change the key to reference the one but that might cause backwards compatibility issues +These address the comments from before +have committed this to trunk and. +Ill retry this JIRA after fixing these test failures +Currently is only a proposal and I didnt start working on it yet +Attached the startup output +Just pointing out possible issuesConfiguration may be just the first step +Why not also add WAB support? At the end of the day users decide which implementation to use and there may be different decision criteria to choose one over the other +And showcase +I agree +Added timeout parameter to tests +Accept zero message from local Scribe client with normal to trunk branches +I have tested +Thus moving this to hibernate conmponent so central doesnt gets filled with issues +The size of the data being transmitted is marshalled first and then the data +It could be an explicit configuration of the war plugin. +When review and refactoring of parser is finished the schema instance variable has to be removed +It doesnt have all of the features mentioned here but contributors are free to submit enhancements. +Are you using an artifact repository as mirror? If yes please clean the artifact in both mirror and local repository +Thanks for fixing. +out file that u get when the test fails in both the cases +it is thereAddremoveupdate sections is rather a maintenance task so I will close this one for now +Now that I think that were ready to go with this one +array error +I so close this issue. +If something else is needed lets create a new issue for it with uptodate info +This will give us a chance to demo the Bean Validation features that are available on both the client and server side +creates explicit close method for history and calls it from points +Available Processors Total Memory Max Memory Used Memory Free Memory +Note that a versioning implementation should also take the somewhat unique access control implications of versioning into account +Defer performance work to +It has to be an appender which exists but causes an Error when created the most common of which is a when dependent classes are missing +local +Can you provide us with the full version name and where you obtained it from +I reverted r at r and fixed again at r by backporting r from Jacques. +Reporters of these issues please verify the fixes and help close these can you review this duplicate +As written in this issue firstly almost rewrite whole write +Heres the new version of the credits page + patch looks good and sqoop tests pass +Too late the release cutoff was already made when I commited +Patch committed to trunk +Note to self on how to run w JTSBefore running testsuite modify AS viacd JBOSSHOMEdocsexamplestransactionsant the name is not actually the PK of the Account class thats being inserted +Closing +From the logs it shows that before nd attempt could get a chance to run third attempt is launched speculatively which leads to job completionSo should be disabled for the test +Just wondering why +It does not always happen from the debugger +Removing the bloom filter would make this a lot simplerNote that it really matters at the end of the day but just to make sure were on the same page now that the filters have been promoted to the index file I dont think removing or keeping will have much impact on this ticket +Bulk defer all frontend issues to +Any ideas for this issue? +I agree that this is a pretty major issueLet us know so we can proceed +This patch just gets rid of the changesto it leaves every thing elseSo far my testing showspostives fixes all run test cases of and negatives Causes some diffs in wisconsin probably the known cases where we dont do sort avoidance anymore but still get correct sortingunknown no full test run done yet +Running tests right +So the +Also fixes that debug statement +JBT is fine but the installation took hours +thanks! +Hold on DonThis has been already addressed +Ben can you please add to the patch and update the relevant code? Ill add a system test myselfcheers BenI think it would work but the existing TL support wont help indeed and as you said it is actually not needed if it is injected only onceYou are right injecting it into the singleton resource only once is exactly what is needed +Attached is a patch shell script and zip files in order to update the consoles plugins and launchersI have put all the added binary files in zips in the correct hierarchy to be added directly to the trunk folder +Closing after committing the test cases to trunk +Sorry +Removed the java prefix and will add automatically if not entered by the user. +Heres a patch for this +Can someone check whether this works on where it is currently failing +Change based on previous for I got it! Already update the patch +What you think James? Rather than return raw Interface up out of read well return an +You might also want to try the newer version or a build to see if the problem has already been addressed +fix findbugs warning in unit test be more careful about waiting for munlocks to happen +To keep it container independent we should not touch the idSpec at this point +Thanks for the reviews lads and DD for the patch +Is this the same as +Another option is to spawn a process to do major compaction +Also for jobs which need files only for reduces the maps which may be much larger in number than reduces will run faster and the overall time will improve +It may be a good idea at this point to just write our own parsing routines. +My patch is to abolish that restriction +for you +attaching fix +Fixed in r. +Is there anything in the logs that is out of the ordinary +The attached files are the schema and instance files used to show the validation failure +Not for +But for second namenode it will not try to initialize block pool +As this is a new wiki and sadly our moin wiki platform do not share login user creds you will have to create a user account +failed in trunk build Will look into the failure +It is caseId in the Redhat customer support database +all you refresh +BrunoIm on the approach too but thatll be some work +Patches with declared above changes trunk +Committed to sqoop +Ive deployed a wrong WAR file +On a German Keyboard its the POS keyThanks! +reproduces with ie and jsf with under and +Sorry didnt see +Please just go ahead and make whatever changes you deem necessary to close the issue and proceed with the release +Works well! But the solution is a little bit dirty +Ill keep option so that JF is happy +Removed one of the getNodes call in rev +Looks great! Took me awhile to review given how large it wasI made a few minor tweaks to get it compiling with JDK +Only a few new keys this time +I can confirm that the first works +Still working on the update for Roberts review. +Christian can you take a look at this? There may be other options set having similar issue +Works well Fabrice could you just remove the new metrics from the Add metric dropdown list and replace Select a resource by Add a project +I must have had some dependency issues here because now Im able to access the REST endpoint with no problem +Would you prefer that I just change that message to INFO rather than combining them at the same time? Given that the abandonment is a direct result of the bad node I think it makes more sense to have one message +Ok there is definitely some work to do here +Committed to branch and TRUNK +I realized that my true problem was that I was wanting to use rpcliteral which did not work properly in but appears to be fixed in +Was waiting for the release of CC +Thanks Konstantin! +Unfortunatelly not the test changed a lot since Dec rd +Ive fixed this thanks for pointing it outRegarding the creation of threads We are aware of the JEE specs restriction in terms of thread starting of course +did switch logging the to debug and even ALL level but couldnt find related information from the logs +I played around with the attributes for the column but I cant seem to get it to display correctly in PDF format +Isnt this the same problem we have received multiple reports about over time? These were fixed right? so what has changed +Lets work together to get this committed so that we can use it as the basis for the upstream work +Please give a concrete usecase +Please close this one as invalid +Are you using optimistic locking? You are going to need to post more details as you have not given us any real info on your problem +In the Token field paste the token as it is listed on the profile page on +Consider there maybe optionsa Do a simple if statement on the version of the default teiid server as to whether include the extra parameterb More elegant but more work would be to add a method to the to request the statement entirely +Before we commit to doing this in we should test how much it slows down reading data +Update patch to recently code base +Will be available in release +For example we could have a setting that forced the planner to do at most one pass over an input data source but where is the right place to expose it +Thanks for explainingOn the same node Solr is maintaining a heavy weight Searcher instance but the search related information is stored with reader +Fixed +When start is false the installed bundles should not be startedPs the has a simple junit test +I cant reproduce that with the Trunk on Windows havent tried neither Linux nor yet +Thanks Victor +The first WAR I uploaded had a couple of errors thanks to Matthias for pointing them outThis is the fixed WAR file which apparently works as expected on plain Tomcat +I already oprened an issue about that! +Ill see if I can confirm the results here locally +same problem on mac maybe accidentally filtering the binaries +Can we please stop bringing the backwards policy in this issue? I already said I will open another issue +We decided to just roll threadsafe autoloads directly into. +This is a frontend generics bug + wgibson resolved +Attached two that represent Eclipse workspace projects +pushed both in JBDS Build id to fix fix version bcs +So yeah I think we should maybe change it to set your contact info or something. +This stuff scares me +Thats exactly what I said in my comments above the namespace in the datastore and the workspace uri got out of synch +Hi Paulex Both fails +I thought there was a JIRA to address that issue but could not find it +Bulk closing stale resolved issues +I have reproduced the issue +Go to ProjectGlobal configuration and configure build triggers under Sonar section +Thanks JuhaniI filed to do the same for the binary tarball and also exclude that directory from the filesets in source tarball +That version of the plugin I had deployed to the jboss repository +btw this test passed on my to trunk +Im not especially familiar with Maven or the Java OS X ecosystem so I cant say that it works for all cases but its been working well enough for mine +Chris the issue you are describing can be fixed in Struts but it is a completely different issue to the one Anton is reporting which cannot be fixed without requiring Servlets or laterAs I explained above the resolution of Antons problem requires passing a wrapped request to the request dispatcher which is specifically disallowed in Servlets +sounds like a good measure +Esteban Reassigning this to you +i really dont see what value these changes bring +Ideally what we need is that if is installed every table operation depends on acl table being available +The log is attached +It seems to me that you are first storing some data using and then loading it using +This was done for the enhancement +Just imagine situation when client has filter which is applied to the different types of messages and some messages do not have property which is used in filter expressionThis would be an users error +It seems to me there may be two separate issues here the interoperability of different derby versions across a network and the coherence of derby code running in a single VM +Thanks Rick and Knut! Knut thanks for catching that typo I did mean to add method but got distracted +Maybe this behavior could be the fallback but for cases where we can calculate only send events when needed +Now that is fixed we can uncomment the assert in the unit to add the right Fix Version +I fixed it in Rev +Rather than trying to perform error handling in the client thread Id prefer to modifying the local read optimization to continue to skip the serialization but to still perform the read in the read stage +IT it is an issue please create new Jira +I could not find any help on making a Indexed cache block when purgatory is full +Patch from N +The original resolution may not be wont fix but as the bulk change required it that is what it is. +It looks like the first bug Hiro mentions has been resolved but no tests added +patch testing now +It just should say that the assertion failed if it is the way you suggest +and thanks for fixing that bug! +Repackaged and updated existing connectors +Looks like the fix for is still in head so resolving this as fixed. +In this context having file with everything in it is very convenient indeed +Has there been a release of the artifact plugin +Are there plans for new release +Youve referenced a couple things so Im a bit confused now on which one you are looking at +Redo after changes +Why is there a problem with using the loopback address? localhost should normally resolve to this in any case + at +Is there any hope to get it into the next release? Do you already have an idea when you will release? We already use a junit in our build and I do not want to make that a regular habit I dont get the behaviour with the because the pom needs to be interpreted anyway +failure is unrelated +I read somewhere in one of these issue referenced as a duplicate that I need to not put the version name in the plugin declaration and run the update command U like so mvn package U + +Patch looks good +js files is also important + +AVRO shouldnt really get involved at that layerPardon my ignorance but why not? Most programming languages provide a perfectly good way of producing an ordered list of paths to search for the files they need why shouldnt Avro do the same +Right I think the default serialization encoding missing is what had confused me on the first place +Sorry for delay Gavin! +Please findthe patch as an attachment with this bug +Was planning on fixing this +For my needs knowing the line is currently sufficient the column information is off by one +Probably on monday I will build a release candidate and invite you to test it to make sure it will fix your problems +I want all scopes but test and non of the given choices do what I want +I made the change an added it to the larger patch in +Fixed on trunk +Could you take a look? + +Merged to on EAP ER +Thanks for the contribution +Moving to Wish List as this is an Uber Item as Aaron has indicated. +The mailing lists are much better places to ask for general help about lognetWhat is your Date pattern for the rolling? is very slow +As for the latest stack traces they dont even get into the jre itself except for +Thanks! +Good simplification! +But I think we can fix this in the coming implementation +added license header and fixed few problems +No reply closing +However my understanding was that the k restriction is handled by rank within the DLSMeaning that a would look like Ak AVkVk +I tried in different environments but still no luck +even something that just outputs some static text would work just put it in srctest so people know its not a production writermost of the helper methods in try to validate that the responses are wellformed XML but if you want to test a writer that isnt XML it can still be done using the directly +The same patch in the form of pull request on +Preliminary patch +Same occurred on Wink with tomcat and Cocoon Cocoon is a problem specific to tomcat +Thanks +Thanks Oli the patch was applied as expected +The patch still applies cleanly so I committed it +Ill do the corrections +Verified with JBDS. +msmsmsms xbeanmsmsNumbers are pretty stable xbean looks faster than Scannotation +I think it is a convenience but more than just convenient +It is exact but I do not know how to make +I removed the logging of bytes read because its in an exception handler that wraps its assignment so it will always be zero +Let me clarifytake that back +This is a screenshot of the heap dump summary when the max heap size is set to m +Actually I was thinking about selecting a leaf by clicking on it +The following files need to be updated for the new GAVparenttestmixinshornetqdemos + tested! +Would be good with unit tests though as we have some unit tests for the existing type converters. +When you are over the api limit and window is not expired if you are issuing extra API you will get blocked and get error message +Deletions and changes does not work well with Akamai +Added as part of master rewrite +can be quite a large number making it impossible to fit into a long or a double +Each loop needs to wait for the next to be be stopped before they can begin offlining their tablets +cleaned up +Bulk close for Solr +In the future if we really need to support arbitrary throttling we could make the metric type a config value that is used when the Picker is a +I have been unable to reproduce this issue on a recent build of Is anyone else see the the No managed endpoint +I believe the responses all have the same messageid the for loop was a clunkier way of expressing remove the handler but only if we actually got a response with which to grab the id +Closed for release +this a pdf generate by adobe acrobat include bolditalicbolditalic chinese characters +Rob please verify +while saving the fsimage the NN looks at its own configuration to see if a config parameter named is defined in the config +Logged In YES useridThis is really a bug +patch coming momentarily +I reopen the issue because I made a self injection instead of making injection available from the outside +Thank you +They dont read the javadoc they dont develop the applicationBut when a statistician tries to set up a gamma distribution and get a message about alpha being negative he will better understand what his input error was than if a generic negative parameter message was displayedPerhaps this example about alpha is a little extreme and I wont argue long for this specific exception if you insist on it being replaced by a more general one +Fixed by rewrapping the bundles if the original header item implements the interface +seems like the primary problem is the readable mask remaining set afterthe eof is detected +Honestly +Patch is missing images please attach them in issue by +I was allowing blocked flushed to continue before I reset the the var that ensures that the blocked flushes continue before a full flush finished +In environments where multiple multicast configured servers are running within the same network multicast traffic can also become extremely noisyCurrently a server is configured with a regular heartbeat that clients must listen for +there can be many webapps in the container etc etcBut I think its a bit dated and its better to just think of running the solr server in which case all parameters are implicitly meant for Solr except a couple exceptions pertaining to the container like +Id like to add that Ive experienced this myself +Verified at r +Feedback is much appreciated +class +At the moment I see that only the first tomcat service in the gets registered +Critical because it affects all jaxws ejb test dont see a any more +Refactore the launchpadapp to have the bundles to be included as real project dependencies of scope provided in Rev +Donald This is because we have turned off HTTPS access to prevent commits +I think make it a configuration setting will be more flexible +This is now fixed for Axiom SAAJ already set the Accept header itself +If we are making the variables private we can remove mbytes variable altogether and only retain bytes +I must have had a mixture of based installations and manual installations that conflicted in some way +Ill have to do some poking around here locally with a proxy appliance but the problem is complicated by the fact that many users configure proxies for use with Maven and there seem to be no other problem reports like this knowing what type of proxy is causing this problem its going to be very difficult to fix +Oh ok I guess I could add the profiles back into the pom +I agree with Stefan +For SerializerDeserializer we should always open but never close + +Duplicate of +Thank you for the review +I just committed this +Patch to address this issue +attribute is already use by element problem is solved but code guard still shows errors +The conclusion is that its not a very natual style for Groovy to have that conditional after the statementSo were not really fond of implementing this featureWere keeping old feature requests open in our bug tracker though hence why this issue is still open. +I see youre right the expectations are different +Fixed in SVN in SVN +Add simulation to Metamer httplocalhostmetamerfacescomponentsrichTooltip +Updated on review board as well +What do you thing about this one +closed +I would really like a distributioncreate command but this will be a lot of work Im afraiddistributionbrand I dont think so IMHO the branding should simply be another option for the distributioncreate command +Done +Hi ScottIm happy to work on this JIRA and provide a patch with unit test for trunk +Ill review this one as part of the clear the patch available backlog next week +So for first stab at it it would make my life easier to save Y with low degree of replication and then use the rest of pipeline the same way regardless of i +Please use the users list if you have further problems with this. +Its more that trivial +See the referenced +Try this one instead +Thanks for the quick commit JohanCan you close the issue and put the new line in in the Release note section +Reworked accordingly +from the data file compute index BF files on the destination nodeThis seems like the best approach +Perhaps it is an ordering issue +Thanks +Hi Andrew Thank you for these new tests +I see a remove button for seam runtimes in the global preferences +Ill put this in over the weekend +This will automatically be picked up and bundled into oncethe master service is in place +Subbu please base your next patch on current trunk codeAlso run all unit testsI am running unit tests on my laptop +But this example contains both api and impl libraries from both projects +See. +It looks like polledData should be replaced by role in the new if statement +committed to TRUNK and BRANCH +There should be a that contains jar files for each of the mahout modules and all of the dependenciesId like to take care of source and javadoc jars as a part of this tooIt may be worth holding off on this until the collections module structure is sorted out but I there are some things I can do before thats done +Fixed in subversion repository as of rThanks for the report and for the patch +This variables value is overridden is the property is presentAll the tests for passedPlease let me know what else I need to do +Marking all Rejected Wont Fix and Cannot Reproduce Bug issues in ER to closed state. +bulk defer of unresolved bugs to +The behaviour is now as expected +Equivalent script for HDFS. +Great question it is only because there is no in the JDK +Duplicates this really the same BugI understood that handles the class wrongly but does not cause the analysis to fail +Yes this is a well known problem In some browsers collection does not updated on DOM ajform is a workaround that has been invented the most for such case +I will probably also create separate JIRA for port +I just checked all the invocation of namesystemwritelock writeunlock and did not find similar problems +Ok so you are moving the search button to left edge not to the center as you are saying the issue descriptionPersonally it looks finePlease change the issue description if you are going to commit this +Correct solution would be to make the sourceEncoding an attribute of the model so that a bundle created with for example can be used in a project using a different encoding +How do you think about this? Thanks +Also the screenshots are in fact there used to be two examples the TODO application implemented using JSF and then using SEAM why is only the JSF version described in the EAP getting started guide but there used to be both in the version for EAP? If were gonna only include the JSF version then we can as well remove the files for the SEAM example docexamplesgettingstarted bug will be closed when the information in the book is updated correctly and will be closed with the example build error is fixed +Mark as fixed as has been suggested. +Ive also added support for and so I think weve got all the JMS properties covered now +It would be perfect if the path to the which got stored was relative to the root of the Server +Ill try the nightlies to check if the fix works +I believe you guys are going to fix it in +Thanks ChrisI this something I missed in the process or the expected flowAlso I assume that nothing further is required because of you change new Jira emails will automatically be accepted from JiraKevin. +Closing old bugs. +Verified present in +The mongrel tests shouldnt be running in Debug mode anyway +The patch made to apply to the top level of +Thanks to you for implementing this and to Julien for filing the issueCheersChris +Reopening it to be verified on WTP related issues are fixed in on +For legal reasons we are not allowed to do a with dependencies download any more unfortunately +I tried with differents installation all are working +Committing this +I tried to build a simpler test case from wicket example but I failed +Ive fixed that +Taking off of the In version to remove it from the JIRA open issues report +If there is a sufficient desire and need to include them please open up a new issue and attach a patch +Attached patch +nocommit wonder if simple LRU cache would be better Yeah havent considered anything about the cache being used really just took the same cache that was being used to cache terminfos +and there is only one refinement thing in grammar and it has nothing to do with subtreeSpecifications +unless the spec is finalized there may be any kind of changes that break backwards compabilitylisting in in the release notes may still be a good idea. + +patch applied anyone building this might need to build the release manager in until the apache snapshot repo comes back on linethanks jesse +attached PR lifts upper version to to in master for +patch looks good but all the test cases are failing +Moving back on M QE says its trivial to bump into and the freeze is instant and takes the whole UI down and the VPE is often used thus would feel very badawaiting response from Victor about patch +A couple of us have tried to reproduce this problem but cant +No other types of XHTML trees had this problem and it was only seen in the +Hi GailI make a test but this will take some days +I think the best way to solve this might be to put a setting in with default CHANNEL and optional CONNECTION +I agree +Could the JT simply maintain a smoothed indicator of task completion percentage for each task? So the jump from to would get tempered out over a minute or soIf we have a way of solving this with better heuristics it seems preferable to adding more configs. +Mikhail what you think should the Alexeys fix be added to your patchI just need a final patch to check +It does seem like the got dropped from the diff for some reason howeverThanks for the workThomas +Needs some more work for the recent ssl changes +Heres the pom I was using to test +How does that work? If it made the right call it would force the prefix onto the schema elements list of namespace declarations +The timers table creation is in the +However downloading of files is not yet supported +I think an option is the best way to goPushing out of next release while consensus evolves +marking it as Cannot repro +Please apply side of things doesnt do this does it? This needs to match what is on the Simple frontend +A missing dot added +I have committed most of this but the issue is not closed since were still pending testing +wo tests +Clebert the methods are indeed inconsistent +lgtm. +Moved creation of Timer before scheduling on applied to and test +Please apply this patch +As I stated before a portlet should be a unit +on a headless node the only configuration that flume can rely on would be a static set of configs under etcflumeconf +The problem was caused by not having in the statement +Shouldnt be hard +is certainly an improvement over what we have +Now I have to change the MREPO classpath variable every time when I started working with project which uses repository or modify the +Client side mount tables provide unified view of all the namespaces +Made the changes and posted a new Sandy +But not for Twill JIRA +There is a related issue which talks about code formatter for Eclipse +The attached patch renames all node terminology back to cell in the API and code comments etc +So yes you shouldnt need to introduce D load +Thanks for cleaning this up! +Their use looks fine as you are allow one and only one classifier right now but what they are doing is legal +So nunit fails +The timestamp in that column is stored in the local timezone +You can safely disregard them +I have committed this to trunk and +Youre right about rbx +Please be advised that this issue can be closed now that is fixed +Added latency histogramPlease unzip into srcsiteresourcesimages +Does the latest one work? Thanks will not work if you change the format of the facet results +Hi Dennis working well withThanks a lot +Currently we dont care about the charset of the other parts +Im going to throw a useful exception in next EAP version see . +a little confusing i guessthe source distributions docs only has a changes folder +is system level config for hive +Hi Mahadev Ive received corporate approval to contribute this patch +contribution in JBTJBDSSo this project has been fixed and is via the new contributionLastly the project was imported using the Archive File import option which created a nested model project which caused a number of problemsRejecting +I guess you are still working on its documentation then +Gotta be in +Committed to trunk +Do you not believe I found a bug dont care that I found a bugor Im misinterpreting this email and this bug will be Bill and moved to the correct project for the fix to MEJB +cool! +Thus we launch based on authentication requirements +Julien would you say for which of them are optional and what not +Thanks for fixing this Aaron +Ill keep digging +It has been always been NOT safe the fact is that they documentedit starting from after a bug opened atPlease dont use synchronization but instantiate a new in themethod that use it +This does not have changes to the tests and other examples in this patch +I could confirm there is a problem in the formatting method +Thats why Pig break and I suspect many other walkers would as well +My last sentence should end with a question mark ? not a colon. +This issue can be closed now. +jars in my geoserver to test this outAndrew +committed +Merged changes to +Added file Patch updated +Note that it also fixes that elusive bug in which was the original reason we delayed checking this patch in +Is this in regards to the database introspect or database reverse engineer command? schema and package are only required once for the database reverse engineer command +actually the rest services are deployed to tomcat the client uses jboss AS +The fix should be fairly straightforward neverthelessI cant recall whether seqdirectory will take a file url in order to pull data from a local disk +Code was not checked in due to code freeze +let me try to make example be nice also to be able to reproduce this issue to test a potential fix +Thanks! +right? understanding was that till a existsBecause Solr is also an API there are a lot of requesthandlers which people write may find some info useful +I used the updatesite earlier on to get all the updates and did not bump into any idea? Some changes in code or is it a build issue +The above QA run failed cause this has to be on theOpened with this patch insteadResolving as duplicate. +I like this but I would have the default not to use it since some applications dont use aggregators at all and definitely dont have the extra storage to spare +The change is not related to this issue +Need to specify exact version of Forge that will work with this one +conf files to specify custom timeout +I dont think its going to be significant too but itll surely increase performance for textprofilesignature and in domcontentutils +Thanks! +If you would like to compile the SDK with ANT I can help you figure out it out +Please have a look at the patchInstead of inserting only when it doesnt exist we should rather follow the same way we handle dead datanodes + +I understand that it can be a bit cumbersome to download install them manually +There is a linked issue here that may still be an outstanding problem +If you could provide a release number and an estimated release date I would appreciate it +The code included can do basic schema identification +Closing issues +Thanks will look into +Thanks attached +Hi Andrew I just took a peek and it looks like the console works now! Just a note I still see the Run as +if the new Exception is moved under the client subpackage +I expect to add a patch to this issue today +The Source Generator documentation supplied with earlier versions makes no mention of any logging dependency and it works without configuring any logging facility +IMPORTANT NOTEBefore deployment the primary key constraint needs to be added manually on the ID column of PARTITIONSTATTBL if the table already existsOtherwise the statistics might be duplicated for some entries and the aggregated statistics will be silently incorrectIf the table does not exist it will be created in the proper format. +Good point +This is the new Permission edit dialog +im it to the list +Im seeing that if I run with the patch soon as an optional +I agree that this should be at TRACE level +Thanks again for the excellent patches +Take a look at that +patch +I think this is a regression which has been introduced with the class of axis +Looking at the stack trace from create or assignContainer would probably be better targets for a fix +Thank you. +Compare the specificity of the media type in the +Sorry for the inconvenience +Thanks Jonathan +The entry already defines an empty string as a default +Thanks SaschaI will wait so have a good GuysThanx Atul I tested your patch its working fine at my end tooGuys i found something that when we use if it is windowed than it works fine but if it is layered and if we click on Find it disappears i dont know the reason behind it this problem arises only in below are the links where i tested ithttpslocalhostcontentcontrolhttpslocalhostaccountingcontrolagreementIdhttpslocalhostordermgrcontrol guysI found the problem resolved it its now working everywhereI attached the patch for itThanks +Attaching new patch by changing the file path of as it is moved to a new folder +Check out the Environment field above Mac OS X +h files to rampartinclude and +Committed to master. +Issue type changed to bug +I wouldnt be surprised if the schedule changed during the last years +I leave rest to community +Thanks! +I tried pulling out just the RIE changes into a new branch and that fails the test too +Incorporating the comments +Everything appears to be functioning correctly +I kinda thought of that too this morningSo if the will enrich the stream with a then the Camel component can use the stream to send back the reply and when it closes it the file will be deleted as wellThe only concern would be what if that stream is not closed or used after the on completion? Then the file is not deleted and kept on file system to take up space +is still unresolved for this patch +Thanks Ted +Done +This looks greatTwo more comments about the logging GUI It should state on top that these are the JDK logging levels +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +I propose we do the following conversionsMain protocols Misc In addition we should similarly rename all the java API classes backing the above protocols add comments to all the proto files description as to what they do and can containThoughts +Jason can you replace this jar into your gs install? It should fix both the plates and fonts along with +The meaning of these clauses can be found in the Reference Guide sections on the CREATE FUNCTION and CREATE PROCEDURE statementsThanks error happens when trying to perform a DELETE for which the SQL session does not hold sufficient privilegespresumably of a trigger since the delete is called from a method called deleteIt is not obvious that this is a Derby error +Done by pulling in Matthieus patch. +I do not have that part working yet +Eclipse Java EE IDE for Web DevelopersVersion Indigo Service Release Build id have also used EclipseNickAssigning back to you as I cant reproduce the issue on the I have access to +Len I think that would be something to be documented wherever these examples are referred to within SOA documentation so perhaps you should open up a JIRA for the SOA docs team concerning that +After looking at many other file systems where +Attaching a patch for this issue +So removing that entrymanifest filejar file will allow one to run did not try the javant service runner but I looked at the configuration and it appears that it runs fine because there isno in the configuration classpath +bug of a bulk update all resolutions changed to done please review history to original resolution type +Id think the right direction is to make them so only users who care for higherlower values get them and let it be usual defaults for others who dont care +With clean installation of JBT to clean Eclipse SDK + for the patch +I dont especially even like the idea of having to make a whole new class for one little static method +However the real problem lies in the AS which is reading the model wrongsigh Ill close the irrelevant jiras and reopen a new one +Addresses and locally try should take an option that renders the table being checked while it runshbasck is probably also one place where wed locate the and then of all stores featureSee +I use a script that processes each far as I can see there is no download location for the separate Bundles +The IDE successfully autodetected as a platform on my Windows XP box +or it is out +batch transition to closed remaining resolved bugs +Ive never even heard of Beetlejuice +You are defining a local target and a promoted target for a reference that has a maximum multiplicity of +Thanks Amar +Thanks Hitesh! +I have some fixes going in that will allow it to work with either ASM or +If someone else wants to pick it up and run with it thatd be great +Also the unversioned files on the website are files rather than the ones +Then you will get more agent error logs +When the newstyle option was originally added the first patch submitted just made all of the generated classes +to me these both suggest I dont have a strong opinion this early +While I can manually removethe line breaks I would prefer to find a way in dita tagging to forceeither the size of the window to be larger or some other solution sothat we dont have to worry about the authoring tool adding the linefeedsIncidently when I use IE to view the text the text breaks more oftenthan the where the line feed characters are +Bulk close for release +In file has some syntax is not strict mode +However Im not calling it resolved yet because there is still a possible usage scenario where this fix wont worknamely in cases where the buffer that were trying to clear has already been transmitted to the client thus making it impossible to undo the write +We need this for upgrade validationthanks fbcbaebeeeaddff in master +moo +So closing this issue. +mpiexec is used by and supports thatMy point was that is the current standard and that if we stick to the standard the setup can be used with other MPI implementations such as the popular MPICH +It caused OOM errors when running Solr tests +I have created fix for master branchI am just not sure if it should be part of release as it changes xsd schema +Actually now that I look at the stack trace the NPE is not happening in code it appears to be in hadoop code so I dont think that upgrading will help +Since is the singularity should we make client side module a blocker? It seems we might have to break BC for this as well +Ill attach a testcase +Basically the problem is in the handling of elementlevel inCould not figure out what the exact reason for this but when reading the SOAP header the element level goes below which is invalid basically the events generated in case of space events and empty tag events cause this to happen +If it doesnt error on publish for you then let me know +The user can enabledisable showing this dialog +Triaged July assigned normal urgency +Attaching a patch that can be used to resolve this feature change request +but it is possible that it might take care of the variables +Im working with Eclipse Indigo SP for Windows x +caught by distributed testing +Nice work Jarcec looks like its just import changes and my ack came up empty for HDFS after this +And trying to figure out what the problem is quite hard since there are so many circular dependencies in you cant compile one file a time +I am just discussing I dont have a recommendation or not sure if this is an issue +of course you would still want to be able to send to any node I think +Applied the fix to the branch in revision +I belive JCE is only usedby NTLM +I would use the BEANINFOCACHE but it is internal +jks +Hey Jaydoes this apply to EAP as well +Move to +Thanks again +you should try latest snapshot of branch which will eventually become +Plz find the attachement and let me know +We cant fix the jboss runtime streams bad JMX implementation as it is no longer maintained to my knowledge +Is this project on yet +I am attaching the patch for review +bug in handling along with the linked bug + and are different +Thank you kindly Stephen +I included all the changes you suggested +Ive modded the webapp i duplicating effort here or should i submit a patch +Merged the fix to the branch and committed revision The patch didnt merge cleanly to because of conflicts in +Andrew does this patch work for yourather then specify the contentType for all POST request it only adds it for ones that dont specify it within a t had a chance to test that but I believe that would work also since we are only sending anyways +There are some other ways of doing this like adding a list of dead servers and a timeout but it does not pass the unit tests with numerous failures +For this international standard consistent is defined to be a difference of no more than unit +Committed revision Will be released versionThank you for your contribution. +As far as I know its a pain in the arse to support sites from both so Ive been making the for compilingtests so far and figured the move to an m site would involve more changing +With the of contrib this has been made a bit murky and does need to beHow is this any different than the discussion to eliminate Version altogether? I think that should be resolved first and this follow the lead of thatHow can one have a useful index across releases without a stable token stream? From the thread it is clear that few understand the impact of an analyzer on the usefulness of an indexIf this succeeds there is little reason to maintain Version at all +assumed fixed +I suspect this was another issue +DavidI have moved my last comment to Vincents Wiki page +RebasedEDIT FYI the freeze is on Monday morning so that might be too late +The default is ms. +I have the code done for this please see the develop branch +fails with patches spanning across projectsThis becomes a problem patches spanning across projects usually are more involved but we run less tests on it +Ill look into it further +We should do whatever the row cache does +The approach I will take is to essentially have a that will depend on whatever Groovys supported and provided package out there +At the beginning you said that if the is set to use READ COMMITTED we dont obtain the lock +bulk fixing the version info for and all affected issues have in is in. +bulk move Version got lost I guess in the bulk pruning can we push this out to +by setting printer mode service registration property for the bundles configuration printer to text zip thus preventing the bundles tab from being after release +Ive reopened as well tests for these bugs are as important as the fixes +As for advisories they are not going to be fired for every addDestination call but just when destinations are created for the first time +The problem is that untar can create a dir named +The EPR should not be changing so we need to determine what is altering it +to this massive API change without much more dialog +Updated the isvalid check to look for abstractmethoderrors and sqlfeaturenotsupportedexceptions +The issue is expected to be fixed for. +Ive attached patches for both trunk +is a new patch +And not in a hudson context Thanks +What Kevin says is right there was a bit of confusion in commmunication here +Can you run the whole test suite and post the result +In the archive are template the file which uses the template +A dynamic parameter is not allowed as the operand of the unary operators or Are there any other doc hits +It is up to a pointSo I would like comments on whether the patch does what it claims and on the separate point about clarifying the meaning of the patch by using UUID in place of Host and Port in the replicationid calculationsI will also note that at your step replication could not resume since the node could not be contacted on the old port +And also in +Hi RickFirst of all thank you very much for posting this detailed functional specificationFor the Computer Science degree course I am following in my university were supposed to complete a research or implementation based project +Committed +HiWonder if you can try the newly generated xsd if its valid against xml spy +and classes are the recommended dispatch actions. +The field order is the declaration order now +I was able to replicate this with svn using neon using neon and svn trunk using serf +Where is derbynetclientmats instructed to run all of jdbcapiframeworksuites derbynetclientmats derbynetmatsjdktesttruerunwithjfalseExclude for JMEFoundation Network Server not supportedrunwithfoundationfalsetimeoutjunitTestsderbyNetjdbcapijdbcapijdbcapijdbcapijdbcapiderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetderbynetjdbcapijdbcapijdbcapilanglanglanglanglanglanglanglanglanglanglanglanglanglanglangstorestoretoolstoolsand in +In patch vb I simply removed the interrupt +In we do that with a +We need to develop this plugin to solve this issueRather than handling this in our indexer we want to do this beforehand so all indexers can benefit this at once +bq +Makes senseI will make sure to not mark an Affects Version unless Im submitting a bug fix set the Fix Version to the next major release will update the Fix Version to include minor releases only if we decide to backport a fix +Never mind about verifying types +You can run the test using java Test +Finally there are still a lot of generated symlinks and files left in srccontrib after a make clean +I would go with your suggestion and change the snippet to setHeader +Thanks Christian I have applied the patch. +But to dismiss it as is a bit off puttingDiscarding the point about reducing the redundancy of code as is equally off putting +i dont see an easy way to port that well focus on getting released asap. +It would be great then to include another RDBMS example in thereWhat site are you using this on by the way? Oh AOL? It looks like AOL is jumping on Solr Hadoop and friends and contributing bravo +I stillthink the problem is in Xalan I just need to take my actual exampleand trim out pirces until it stops failing +go thru the issues fixed in +push to +Thanks DIGY +In addition to the headers mentioned above this module is responsible for generating some Java source filesIn this case I am going to view the original grammar as the source code and the not expect the generated files to have GPL headers +I attached another version with Arq +Put it on a new line is good idea +you only need to add the new field in the enum + the patch looks good to me +not sure what we can do here the run button works as expected it shows what launches are possible for the current selection +I was attempting to test multiple different configurations of bootstrapUrl and bootstrapHome in a single test which does not seem to work even with the shutdowServer lifecycle called +The scenario orginally reported should never happen because of refactorings in particular the addition of the MSGREPORTPROCESSOPEN mechanism +Cleaned up patch +The issues with maintaining state within the consumer had already been resolved within the revised patches +And we would like some unit tests if possible. +The patch also includes and +This seems like a reasonable request +I typed mvn clean install for my downloads show the +Committed wcleanup patch. +closed all issues which got released with +Can you redo the diff from the latest trunk and also post to? Thanks again +OK but this means that perfectly valid schemas in do NOT work in trunk +Not having elements of depth would basically eliminate all of our schemas from using thisAre there enough people out there who would be able to weather this constraint for the simpler version to be effectively tested before complications are added in? Seems like since in this case we know what the expected complications will be its better to design with them in mind from the start and avoid surprises later +This patch will solve the Unconditional Session ExpirationTimeout issue +This is not a problem with encoding +changed to match the c behavior of not setting the reference to null +Unsetting Fix Version on unassigned issues +I have reviewed Brians patch and think it is correct +Are you guys contributing Hue to ASFNobq +There was a comment on about this patch +Unfortunately this is too big to review in detail right now but I trust Robert to do the right thingbq +Even with the commitlog disk dead the node would be able to serve read requests w besteffort which is the intended behavior +I agree with Nicks point but for a different reason +Reopening awaiting further this with and it seems the default has been true all the way +this class is deprecated +Rails itself starts a Mongrel to serve the underlying application +Committed to trunk with revision +Committed to. +Hi BabakYeah it makes sense +Im running and just encountered this myself +So is it really a war plugin issueDoes it works when invoking maven with cli +Ok I would have to put a container in the path of the default container so that custom components can be loaded +Patch was created at the top level against rev applied Committed revision Thanks Mike! +Unzip under modulesjndi directory and execute +I will apply the fix suggested run derbyall test the fix manually in the scenario I found the other one described and create a patch forThen I will add another JIRA issue to indicate we need a regression test later +Maybe we should reset the error message to and regenerate the old errormessage with the delays in the +See get for details +cannot reproduce +Move forward to out work no longer being undertaken. +Maybe we should even rename the package to something else below qp +Thanks Chunhui +Thanks for the bug submission +Patch looks good +in summary I would say its a producer error to send a value for both markupString and markupBinary + +HiIs any work on this issue in progress? I would also like to help solveve tested this with my local version of Glassfish and the latest SVN head without issuesHow do you find out the version that is installed? Can you attach an example app +As already described I was not able to reproduce the problem with JNDI +May or may not be in this patch though +NO I am using for JSON media type +I have document vectors created from some internal text +each memstore change would be an upsert that would add the new and then remove all that not be flushed to disk anyway +The patch for case is in +Currently hasMany and belongsTo are not supported that is now on the todo list as next item +bulk move of all unresolved issues from to +Meaning no matter what I supply the credentials I can connect to the stomplets +Humm +I was gonna apologize for creating extra work for you but it would seem that it was so easy to fix my pull request was still a net plus +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Sorry +However the implementation uses the Jetty API which will likely be updated as the web socket protocol is updated +Thanks for your great encourageI have finished the other test cases on importing cleaned the code added anotation +This fix will go out in the next maintenance release +Manik theres only wan suggested solution composed of two steps Ive implemented it see the pull request +ive added logging for try and success of builtin libraries tooisolating the logging as ola suggests will be a bunch more work and considerably more change to and i dont have any time to spend on it ve added your logging code in dae thanks very much for adding it! Its a little terrifying to see the number of searches it does loading Rails libraries for example. +It uses without specifying a character set +get will not return null +Please check with Vladmir +just makes this behavior happen even faster +I have a few very quick comments Instead of returning a pair of longs in startForwarding we could simply return maxZxid and read lastProposed directly from the leader object +Our SDE on top of Oracle has NVARCHAR fields and is OK with those +However as this is a pretty elusive bug Id like to try our app at customers site before closing it +Patch committed to both trunk and branch. +If you see others please correct and I will go back and do thesame for each oneCristian +And both JBT and JBDS aggregate builds in trunk are blue! +Seems the include ordering and missing CProgram Filesinclude includes fixes from my patch didnt make it +How this affects and which they dont produce a JAR +Ill get more indepth as I step through the code on tests +they are related but as I said get is optionalFrankly if you have to change your user type in a system doing a smart find in your favorite IDE is the least of your problem +Committed the changes and will wait until the Hudson runs behave accordingly before marking as resolved +ChrisThank you for your comments +This change makes ZK work better +Windows doesnt work well withThere are a lot of discussions in internetOne of the possible solutions +It involved making changes to the deserialization code in the core moduleCommitted code fixes to trunk as SVN revision Added extra tests to serialization itests as SVN revision Marking this JIRA as fixed. +Thanks a lot Jitendra +Ill have to try to debug the setup +Didnt check spidermonkey trunk but I trust that patch is fine +Thanks. +This wasnt a critical fix in hours our disaster patch so Id have liked to have been given an opportunity to review which I would have done by getting the junit part of the diff moved to a separate patch that also updated the junit version to Now was the junit patch an accident? If so could we roll it back and have a separate patch with the junit update and a separate note in? +In this case however the profile would also be active when building only the parent itself an unwanted effect +As long as its done its good for all of us. +Can we go ahead with implementation? Please explain a bit on point I understand the intent but can you please explain where do we pass in a javascript function +madair working on this +on patch +resolve this issue since we can not reproduce the issue +Committed as proposed +One more comment to the patch the constructor does not need to throw +Now we may still want to convert everything to upper case +describe table column +BTW there is no anonymous access yet +It was not part of the patch as it is a binary +Hi KarlThank you for the information +The patch should be applied on clean JSSE sources +out of hunks FAILED saving rejects to file testgadgetspatching file testgadgetsHunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at +it looks like this collided with +Closing JIRA as fixed in fddcdafaafcadfTests in to deltacloudtestsdeltacloud +Everything looks good tests all pass +A new class that extends with the ability to register custom data type converters +I know how to fix it but it is more difficult then it looks mainly because the are a mine field already and I have to be very careful not to make it break in other ways +i expanded every level just cause i dont know +classpath file does not show any exported classpath entries that would be expected to be exported when published +We can clean that up so that it doesnt hit that usage patternre product In RHEL we differentiate at the build system level both for composition of packaging and the default configuration values +Classes are +untouched +As far as I can tell a bunch of the persistence stuff looks broke too looks like it saves correctly for cores level stuff but not core level stuff +bq +I end up doing searches on Google and looking at their cached version of the page +patch with fix and test as suggested by says patched was suggested by Igor +Some cases still remain where characters print +olddoc may not been exists after compaction so youre returning to the source of issueInstead of this have you thought about adding very trivial logic to your filter functions to accept or discard documents with deletedtrue if some special query argument or header exists +file as well as Hibernate testsuite needs to be patched for correct run of this test seeFor the patched version is this issue verified +it took a while before this failure was laid to rest +So taking the documentation at face value it would seem reasonable to get this exception even for transactions that arent affecting the same rows +Attached please read my revised proposalWhen we should try to place replicas to different racks whenever it is feasibleTo minimize write latency when we write a block that needs more than replicas does it make sense to let the client return after writing the first copies onto dofferent racks and let the background process take care the other replicas +Could you avoid the redirect if the NN is not in? As for SPNEGO I think the client does SPNEGO auth to the DN but the token is for the DN to communicate with the NNYou should be able to use the current users ugi to get the token because that will reflect the SPNEGO authed user +Got your point about indexing to multiple clouds thanks! +Instead it obtains samples per block +grr oh well +scans in there +This patch merges with which had to wait for wicket +These tests can now run in modes depending on whether the export policy jars have been installed on the machine +i committed the fix to memorycodec synced the patch up to trunk and renamed the confusing sumDF variable in termsconsumer that actually is no sumDF at allI think this is ready to at the moment we would have to write this separately +I wish there was a way to exclude all transitive dependencies of a given dependency but if there is I dunno it +FF is happily saving the in the respective cookiesdirk +Check on the izpack dev list if they are ok with doing this by think what is broken currently is ability to pass to the installer so that it silently uses it +Selector was updated to ignore version control dirs by Vladimir the patch was applied to BTI branch at rPlease check that the patch was applied as you expected +Lets check Attilas steps +I just removed the guava dependency thanks for the our http module covers this requirement? +Hi Mamta I have updated the patch and included the html files for you to look atThe mixing of names harry and Harry was unfortunate +Fixed in the current java svn +Patch implementing second proposal described in the problem description +I had a look at and if I checked it right and this change to run times create was introduced in +I think the way we handle this in the other implementations is to align the with the skip policy in the step factory bean +Why was the stuff inside +Thats like eternity in project development time +PaulexThanks for the patches +This issue has been fixed as a part of. +The unit test failures are unrelated to the fix +Perhaps this should be combined with high freq terms tool +Provides resolution about whether a property should be or not. +This was decided as not a good approach +Imo it is +okay its with docstrings for the extension methods now and tests are passingRan tests in patch merged into trunk with some minor fixes +Move out of +Sure +use only english names for entities and attributes +Part of Weld master now it will be in next release +here is another patch remove unnessary s configuration configure core surefire to write all its output files inside target directory set svn ignore properties for eclipse adn target directoriesthis makes very happy to see no unknown files in its directoriesexcept the ignored as requested in looks to me like all of the attached patches have already been applied including the one attached after Marios last commentSo unless someone sees something that has been missed Ill close this as FIXED +If you can try to confirm against master or we would appreciate it +And we redirect the query to use the second column family +move to since Alpha is already built this issue wont block it +as well +Its these linux boxes that they threw us onP +Proposed in svn various wrappers dont support this yetAlso it seems there is not test case yet +Oops +It works great +Not in terms of wars +This information is currently not available +But other than that this looks good to me +Hemanth if we are doing the above current directory structure is still fine right? We need not introduce in the directory structure +Hello Alan and Andrewis not exactly the same thing +At some point the master will be a version that expects the locks to be honored +Rev of release. +Ive patched sharedtarget version of as I wasnt entirely sure which one to patch Im pretty sure they are all the same with regard to the bugpatch +Test case which reproduces this issue +Idle wondering +Should be fine to also do this for +i dont think we should be focussing on creating an installer +The problem is that the payment methods are never set in the cart so the system keeps thinking that payment is the next step to be completed +This is the patch to add FOR UPDATE OF to +Thanks for noticing +JET OPT and interpreter configurations fail on windowsxp +However has this change been tested with securitykrb enabled +This is due to a bug in the create method in not using the context classloader to resolve classes +Maybe there should be a little more leniency +Of the two is actually generally useful. +The profiler API lets you control if a method is inlined via the method andor disable it all together but in reality this may be too heavy handed an will add this information to the AOP docs +Prevents the is the same issue +Patches welcome +Thanks Jing +Ive uploaded a new patch to JIRA +Chris to your point I figured Id go ahead and check out what does happen when theres no enabled +upstream ISPN issue +Test comment with image +FYI +I am attaching a new version of the functional spec that containsthe following changes Modified the new commands added to to match the existing pattern Added to each of the commandsI have incorporated here feedback I received from Dag and OysteinI have further received feedback on the handling of Upgrade whichI shall update in the functional specification and submit in a subsequentrevision +Ive changed the implementation in revision its not tested yet therefore Ill leave the issue for fixing this +with entry +I was thinking it was the case also but some users reported that the configuration file are not created from the features descriptor +Patch attached +Ok I moved the Pingable interface to be an inner interface to +Updated version with the actual error +Bulk close for +I removed the section about the dateFormat being reflected in the reports so it wont confuse users. +Also when the master gets a new regionserver on an already running cluster it should automatically trigger a balance +Ill schedule this for +The is now with +If there is some version value that this is needed for I can rework the patch +characters correctly? operator does not work correctlyperhaps these just need some javadocs scoring is strange because its based upon surrogates leave alone and javadoc it +a minor correcton and cleanup to the spring beans xml and the test programM trunkcomponentssrctestjavaorgapachecamelcomponentcxfspringM trunkcomponentssrctestresourcesorgapachecamelcomponentcxfspring +sounds like a plan +Currently the components of the service are initialized from and some of the initialize asynchronously in separate threads leading to race conditions and complicated code to avoid them +Oops excuse me error in my R script for matching up colors to datasets +Just to be sure though I tried yours +Specifically making use SessionBuffer instead of Session +Oh if thats relatively standalone then yes that would be awesome +Im on windows +Add the following into the source files from the API all other patches before are included +fixed +Brian sure +for LUCYVISIBLE +Pat can you check if the patch is all fine and I didnt miss anything +The ear must be packaged to reflect that +patch was. +Moving to JBAS since this appears to be an app serverjfs related I be able to reproduce this wo Jopr +And currently it converts LDAP requests from the Twix format to the common format and converts the LDAP responses from the common format to Twix format so that the server can decode the incoming requests and encode the responses before sending them back to the clientsIn addition the at the moment doesnt transform a common request to Twix request nor does it transform a Twix response to a common response +The row number now really just tells what row it is starting from the top and adding one on each row +Later we likely want to these with getters and setters and these keywords should be reserved for that purpose +Nice David. +This is the browser openshift website having issuesI noticed it too worked most of the time but out of our controlIf keep failing use an external browser instead and see if it repeats there too +This is a good one! This might be one of the reasons why we sometimes see blocks in a cluster +Also the fact that the current implementation is fundamentally broken is an issue +Updated description text from is a component that does iteration +Tomislav this bug was fixed yesterday try to use latest build +I think youre right replacing with underscores would be acceptable +See also +All right! Thank you very much for your time and patience +attach a patch fixing the handling of loss connection exception when creating ephemeral Gavin It sounds right that the connection loss is not being handled properly +Updated this bug for only physical devices +I assumed what Norman meant was not scoped specifically to a what I dont understand here is what it means for something not to be scoped to a context +Test cases that demonstrate the wrong behaviour for empty results and the correct behaviour for results +Raising up priority to blocker as its a raised the concern that asking WFS valid names for feature types might be too strict +Have you tested this? Prepared statements are not based on string interpolation +Java allows this so I think we should do to be fixed +Does it actually work to makecc a recursive task? I would have expected this would cause issues with monitoring and with compilation but if it works fine on your local system go ahead and commit +these tags are added in revision +bq +I committed the patch to and trunk +Thanks + +It should link to the applib module obviouslyIf you could make this change too thatd be great +old related +Thank you Catalina +I saw this with OS X and Eclipse +easy to sort out though +If you dont believe me try it yourself +trunk Committed revision branchx Committed revision Thanks Drew +Description Alter output where exception catching message were printed from to +Additional unrelated cleanup change remove classes +im unsderstand set useful +and continue to be met with push back +The issue is fixed after the issue is fixed. +WDYT +Bulk close for release +This lets you have three panes and you always see the pane in the middle. +But if the name argument is mandatory for constructor arguments I think the error message should state that clearly +tried it out Doron +Also I have made a behaviour class to test it +Im just running all the steps suggested in the wikiThanks against revision +Documented this in the appropriate section of the Spring reference manual +Was planning to run a Pig job anyway + or while redundant since they are required for all fieldsis a regular expression +Also I have following problem still +Never created a patch but Ill give it a shot +Shows both upper and lower bound problems on the same picture +The title is misleading the issue here is that the preview page is not properly url encoding the type name in the generated link this has nothing to do with the locale +great +Couldnt reproduce this bug +I decided to add this code to the Trinidad specific API as opposed to using commons so that the upgrade will be easier +Great thanks! +Junping there is no easy way to review such a big patch in +This appears to work correctly in your case +If you remove spaces from the test it works nicely just to assert that the issue doesnt come from the additional role added to the permissions. +Attached a and applied patch +create can be used for nodes and relationships +All committed! Thank you! +In order to disable the Atlassian connector it is neccesary to set the following preferencesfalsefalseFixed in the trunk using preferencesThe advantage of Equinox Transforms would be that these two plugins wouldnt be started if they arent required +This will show any manifestclasspaths +Thanks +Is the fix included in the latest nightly? one +The existing Flow existing page where you select the file that combobox list simply the list of runtimes listed in drools preferences or a static list with Drools and Drools +Not yet but there are plans to work on it in near is some code related to zookeeper done in yahoo long time back +fixes this for and +None of the links given by Jenkins reveal the cause for the findBugs warnings +Passes tests locally +Workaround Define the element within the file Your project must also explicitly import any transitive Maven dependencies as the Forge structure prevents transitive dependencies to be seen by your plugin +Just looked at the attachment again it does appear that attachment shows a call that used to work from the outside but no attachment to reproduce the error does anyone have one they can attach +This is a duplicate of now fixed +Verified in +Fixed in CVS +No problemFiled for the packaging duplicate jars issue +Log files with stack traces from a test run +Here is a patch to do the others +In a mean time you can probably use launch configuration to run archetype goals +classpath file? I dont use Eclipse so no point in me making updates +A minor nit was the in the deprecation was suppressed +I am having to Cordova for now in order to add the attribute to the backup file +Freddy what is the reason for this? Why not use information from pom +Fixed under revision r. +svn dirs removed now +I just wonder why new controls depends on librarys which will removed in future tobago versionsAs long as they are in sandbox this is no problem but they cant promote into tobago with this dependency +The URI returned by getUri is no longer based off of the for the +jsp or +Keeping the printing in would work but we need to have a consistent way of counting the messages that were successfully sent by the async producer during system testing +Result It will display FindList form for Billing AccountThanksSurya Kusumakar +I fixed the section as described in the description and added a new section about the new +Removing the shared dir +Im still trying to reproduce the exact issue in a smaller process +I dont know I only have the wish for the values +I have added a fix as well as more unit tests to cover this +Given that all data retrieval from Accumulo is returned to the client via a Java Iterator Im a little confused about the application of such a featureIf you have many columns in one row youll get a pair for each at which point its an exercise for the client not to consume more than can fit in its memoryOther applications seem to be solvable through a custom Accumulo Iterator +completed. +That seems to make sense +Since yesterday its part of the generated site so jan will probably deploy the site with the reference manual today +but turns out the interaction w clear requires that it be a data memberneeds seems fine to me +I fixed all the remaining cases in ab +The other option is to drastically change our domain model to match the limitations of Annotations but again this is too drastic for our requirements at this pointI thank you and others like Emmanuel for your hard work on this and look forward to future releases +The Properties Editor does not refresh +If you dont want to run clover well +Sorry I was wrong +Thanks Mike +Please do code review on the existing patch +I believe you were working on this one +Yes that was my idea but I can see why you feel it violates the specIts fine with me if we take a more cautious approach and use a specialentrypoint to the engine as you suggested earlier +Ive just committed this +Latest patch from RB +May by we must write Tool in and call this tool from binnutch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +The classpath for the engine plugin works now correctly +This was fixed the same day as as +Thanks for the patch in interfaces is public by default isnt it +Any update on the test project? I somehow feel that this is an issue with AJDT tools that dont allow to use jars from classpath container +Hoping to clear the clutter under migration work is for the most part finished +Your patch would copy all those resources into classes as well eg classes if I understand for putting the effort into a patch though is there some other way we can solve this? In the past Ive done these as integration tests that would take the final WAR file and use that instead +To be able to successfully buildtestdeploy you have to do it using either French or English locale or one that uses the same date format as English +Committed to trunk +Closing all resolved tickets from or older +I cant think of what a good option would be +I understand the limitation and specifying requiredfalse is just fine + fix a bug on the test + +By adding the Trevni builtin for Apache Hive Apache Hive Shark Apache and Apache Pig will all get Trevni support +Im not sure what reasons you might want to override the encoding other than to provide a default but it obviously was important enough to someone that it was added to the servlet my particular case I am not that bothered +My main concern with and was decreased visibility to someone modifying the file which carries an increased risk of accidentally adding records under an inappropriate operation +Michael thanks for the test case +I dont see any drawback with the patch so that I suggest to apply the patch to resolve this problem entirely +All looks good apart from your imageAlthough you mention x in your image name it is in fact x pixelsOnce you have resized your image to x then I will add your mirror +The plot thickens because the spam check is buried within the processing of referers +Done. +The problem is that Pig doesnt check the return code while calling method +SVN Revision has broken this attached +If people dont like that profile we need to find another way of satisfying optional dependencies without polluting Maven Central +For me it seems to be goodSorry for the delay in replying +Moving out of +Can you please review and commit this patch before others +Seen more than one posting about it previouslylately +This is only intended to happen the FIRST time a Roller installation is started +Adding patch to include new unitsystem tests as well as supporting build file modifications +The rename has been reflected in the code but still need to update the User Guide +Thanks for the patches. +It might break other usages but Id say commit and lets deal with them if we hit them. +I am creating a UMLS Virtual Machine for AA +Geronimo is a bad example because they became an independent projectPersonally I think it is a bad idea for to have their own private lists even during incubation +The outrigger implementation is based on a class that shows a similar failure under similar forms of stress high rate simultaneous add and remove activity +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +Tom could you please take care of this +We are using jdk to run maven and java source code is also for jdk +The fixes work also forI just added another example in the showcase to show how a layout page made with leads to a correct . +New patch rewrites counter handling to make pass +Doing git bisect it seems that both and stop working atab +What you call the white list is in fact the actual fileIf you remove the default mapping then unknown file types will not be parsedMoreover in a previous discussion Andrzej spoke about committing a strings command line like parser that is able to extracts textual content from any file type It is really the good candidate for the default parser instead of the text parser +This issue is closed now +updated tests to check for non existence for version file when format command did not succeed +thoughts when looking for a statement that Roo is Tomcat based I could not find it so it would have helped me if this was stated in a more prominent place +Note that in samples we have specifically asked you to check paths for modules and policy files which you might have ignored +Resolving as Not a Problem given that Benson merged the initial archetypeI checked in the archetype because it worked with not because i found a solution to this problem on trunk +Thanks Nigel +I think that is fine +this patch adds jars to the classpath +where its hopefully git we will be pointing to +Also elimitating equality condition on floats and improving logging +Moving these to for now +Hi Stefan I will reply you via because I think that your question is not related with this bug +It appears the problem is still the same +Modifying the list in different threads will definitely lead to an exception but I wouldnt expect replacing the contexts list with a new object would +Please find the attached screenshots +I will upload another patch file after some better a lot for your understanding +We dont want to delay release because of this +So all issues not directly assigned to an AS release are being closed +Thanks for also posting this to the Gora dev list it is nice work and worthy of review for sure +But I am also thinking about options that might be needed +Gabriel do you have some time to apply the patch and run some tests of your own +I guess the example is enough for you reproduce the cases +Im not sure if there are any good tools to help you but you could try to deploy each of the requiredoptional resources one at a time in a clean framework to see which one pulls in +After editing to remove the comments from the user with the manager role I was able to stop start and reload webapps +I have a fix with test +yes you can +I have tried to reproduce the issue +new +I faced the same issueI noticed that on one of my nodes i am unable to upgarde the agent toI removed that node and now i am able to see the Nest +Patch that adds both the scope on HCD and the clusterId on +Patch looks fine to me +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Quickstart demonstrating the issue +Could you attach your source data I run this script and it works. +Ah by the way in you can pass pairs to your xml parser changing functionality +This is really a reasonable requirement +Commented out the print statement +I stuck to create my own custom enum type which store value instead of ordinal in db +need to be catch too +then upgraded to partway through apparently in an attempt to make the horror stop +As for retries etc Id suggest adding a seperate issue for GA. +wrong patch file +Patch applied to svn revision +Related to +QE could you please try if it still doesnt work with current tested versions of Mojarra +Rev things work fine for me too +It seems that in at least one location in the pread code doclen is truncated to an int +Sorry it took me so long to get back on this +The destination is entirely customizable through the use of an injected strategy as explained in the section Application Flow on Authentication Success and Failure in the reference manual +Could I ask you to try it out to see whether the problem you encountered has been fixed as well? If not please provide a sample stylesheet and input document that can be used to demonstrate the problem +it was only fixing stuff for windows +I just happened to deploy +Adrian +Should it be more explicit that this? Can a real client set its name to empty alsoThis is what we do in the trunk +OK this is really confusing +Patch committed to the branch. +I just tested this with roo and it still does not work +ve just committed this +The tree I tested this with is on commit VijayThis issue was consistently reproducible with the builds that i used a week back + +Verified +I dont have a good way to reproduce your issue but this should fix it +Less than week to pushing off more items to +unless theres justification that the implementation in trunker +I will try with my Ubuntu box +I cant think of a good way to make this work in a way that allows replication across servers +Sorry i have the same problem with the scrollbars as shown in the screenshot +Ill try again but it seems like a false alarm +The null codec is always present so we can fall back on that if we didnt build the deflate codec +Confirmed that it works as it should with bundler and rvm as well. +But what about if the user really wants to set it to true? Live with the WARN messageMy understanding is that the best practice is to not run with checksums off hence the warning +fix appled for as wellMasaya is the problem still there with current build? I could reproduce the problem you had in my local workspace and got it solved too +However the job would have failed if it executed past the token expiration since the JT was unable to renew the tokenIf the JT has to acquire tokens for a job and acquisition fails the job will fail +we should keep backwards compatibility +Do you have an example of a and a plain JSR portlet that does not exhibit this behaviour +This capability allows you to set up an additional physical infrastructure on a different physical NIC and use the same set of if you run out of +Prasad Is this ready for review? Is the most recent version of this patch? Thanks +The next planned release is so if this patch is applied before that then it will be available with that release +Ok +Fixed +if the ledger is closed we use the close value in ledger metadata as last confirmed directly we dont need read it againI think this is also a bug related to read +Tycho allows us to make of ours with more comfortable ways that using TP +attached patch for problem at all +Thats OK its a connection of the bridge to the local broker +It matchesthe version used in the Spring Framework trunk. + is a wrapped exception +Reopening so I can update the fix version field. +Works fine. +I now turn your attention to +Pull request submitted +Verified using +It may take me a little while to get back to this +Can you please elaborate on but this method does not seem to be the cleanest of approaches +Alexander I think there is a mismatch between the raw data and the timemachine as I cannot find anywhere +Well it works +I have improved the scriptI am not sure usrucbps is installed by default on Solaris so I keep using the standard solaris ps syntax +it limits it to new builds only where import scope is used +Closing issue fixed in released version. +I figure this should be an improvement not a bug +waiters is only done during connection startup and clean shutdown +This ensures that a block report will not report these blocks +Attaching v that fix both +ok check the archiveyou should be able to use itDont pay attention to the content it is something used to debug an issue in forgeDont pay attention to the test itself it was just to provide you something +Hi allI added a screenshot of a first draft implementation for this new dialog just to get your feedback +Original SVG source which has the correct alignment and was used to generate the already attached PNG image +a lot for your comment and help +you have to manuly copy from config or configure the existing appropriately +Hmm you should typically not run the SCR plugin in parent projects I assume you do this for inheritance right? We generally do not declare the plugin in the parent pom in the plugins element but inside the pluginManagement element and then redeclare it in each bundle that actually requires itBut we could probably add some code to prevent the plugin from running in pom I think we should check if the dir exists and skip execution if not +Do you have a clone of Wickets sourcegit checkout git show befebeebbf +The patch had been committed at rPlease verify it if it works as you expectedThanks very much for your contribution +Adding to my JBDS Spring IDE from your update site made the aspectj error go away on the spring travel is weird as AJDT is supposed to be optional +Fixed all the configurations and the samples to cater this on the SVN branch +I moved the subtasks to since the fully automated hudson jobs wont be ready in +Attached patch for trunk and +I changed to show the consequences +They will be made public with the next release of the book on EAP s release +In and before the difference was probably near zero but for indy for example I know that not knowing the return type of an operation on two primitives means boxing and unboxing actions that not all are removed by the JIT thus are a serious slow down. +tdataList works as expected +FWIW the artifact filters in question have been factored out into their own project to facilitate reuse in other it a spin and see what you think +There are at least two issues that are evident at this point concurrent access conflicts between multiple nodes in the cluster and problems writing preferences for a portlet application more than once during the lifetime of a single server +Thanks for the patch Aaron + has been marked as a duplicate of this bug +It should go to the outer interation construct +It would be incorrect for these requests to failover since it may result in duplicated the case of a JVM crash the best we can do is to provide failover for subsequent requests +closed +Scrve +This can be counted as final +Implementation of Content Upload in java using multipart for REST +it will be also ok to add the date +So I agree that defaulting to UUID covers more cases +A little more digging and Im a bit more confident that this is a flow control problem in the IPC layer +But its just a wild guess at this point +Fair enough +Patch attached +Turns out that Ubuntu dropped bit support for whatever in and now you need to run sudo install in order to get older executables to run +I know that one approach to solve the problem is to use subclass +Default behaviour of is to close dropdown on resize eventExpected mobile behaviour is for dropdown to stay open so i have exposed property via +I thank that could be written in the hfile now I thank as meta dataJust some suggestions andor other thoughts +Reproduced +did you even try it? the generated class gets changed but references etc +Let me try to put it differentlyFor me we have separate issues here annotations on constructors arguments are not correctly handled produced metadata at the end of the manipulation is somehow incorrect when resource is manipulated a second timeright +Workaround would be to refactor model structure but that is not as practical as making the leaf models detachable +Variant also looks ok for me +In this bug LB offering is not available in the Updated offering +Just post content of Plexus Apache JSF components JSF components dont why but now everytings work +Can you provide detail on what you mean by Yes but mvn is not available by defaultFile paths that the deployer already knows what files to plop in that path +Thank you for the patch Guilherme. +We could also always define the generic accesors but have different code paths internally depending on beans and mode +Closing jira +Thanks Yonik! On the topic of issues which have been around for a long time +Issue revolves around time sync problem between client and broker machines +any news I will close this issue soon. +Patch applied in rev +trunk +Patch applied +This patch look goodIn the longer term it would be nice if we can make the namenode ask for block reports rather than the datanode sending block reports voluntarily +On closer examination Clusterinitialize doesnt need any special handling should be set to local +It could be that we need to pass the to the cipher init method +This issue has been fixed in java M but still persists in the java branch +Anyways patch submitted +I faxed the agreement to apache yesterday +Patch to implement improvements described in issue +I went with a sleep because it was the simplest fix +in fact I get that solution after weeks reading about Grails +he problem is that there is an inner tag and the for the opening comment has no knowledge about the closing tag +Hi AndreaCan I know if are you working about problem that I told you? Otherwise Ill change way for my solution +Looks like we cant solve this problem in code but we may provide some information about this issue in the FAQ section of the webpage +Had the same issue +Unfortunately the new behaviour of maxlength OVERRIDES the global configuration +Hence Ill close this issue to be reopened if there are any remaining issuesJuergen +thanks atm for the latest fixed + +Have not debugged what that is but there is something which is not cleaned by app pool restart +According to the comments of which said the long value must be newed not valueOf +Also the group ownership of files on private cache have to be set to TTSPECIALGROUPRamya to address your other point in the description all the files are deleted as the user now so we dont rely on the group ownership to match TTSPECIALGROUP +Attaching +the problem with r is it does not address you can still have Getter ops on the the service added after the Flusher so they will execute during after the flush +No need to post the heap dump just yet but do hold on to it just in case +You cant have as the first character of a classname either +Someone who uses just records should be on no less stable ground than someone who uses just HDFSMy recommendation would be to simply include hadoops jar in your project and to submit patches to hadoop if there are record features that youd like to add +This is the original WSDL file before it was transformed by Spring WS +Anybody working on this issue? +block writes at compaction completionA better version of the same idea +Attached updated simple IT mistake I had the wrong plugin name in the IT pom attached the fixed IT +This would make the task of adding new metrics much simpler + +The patch looks good on our end running with the Transacton Synchronization Registry in +Rename the class and fix the yet +Verified with +I have committed this +for ivyant. +Looking at the code they want to have precision for the queue capacity +Update patch to include more missing javadocs +Kind of a hack I realize +gig or something like that +In this case we definitely dont have toRe I like this in principle but Im nervous about making public since its such a new API +Updated description to reflect the other problem as s a patch containing a fix and also two new unit tests that verify the behavior +The version is +Since protocol versions are by definition a closed set we only care if there is a collision +To be clear as of this writing resources deployed in a Java EE environment with as the provider will NOT have their fields inside them properly assigned through dependency injectionIs it true that the only way for example to use an EJB inside a resource when the implementation is the only way to do this is by using manual JNDI lookups? As of this writing +committed in revision Mike thanks for review +Well actually we could use nested elements at least for the parser class and the loader reopsitory class +Please keep the as this is the new version +Thanks for the information we definitely want to support this and will look at getting your patches integrated +Verified in ER +Thanks for pointing out the JIRA as it saves me from filing the duplicate +you beat me to it Guess Ill skip the current building to fix Should we refer the service using a string constant? Its sprinkled in various places in +that was x with jdk looks like all the failures on jenkins are due to connection loss issues due to ZK in the setup +thoughts +In this case new extension points must be developed +Simple patch attached +Odd +so do we just change the fall backdefault parser that is supposed to use to +demonstrates the issue by comparing cfhttp and posts to a php script +bq +this patch is for both and im sorry my bad +Reverted the commit that caused the problem plus the one that Ignored the resolved incorrectly reopening and correcting the state to PR sent +And Java is not hostage +This is preventing us from moving to +We are also testing a sample web app with all three of those providers before every Spring that helps for your confirmation! I finally found out what the problem is its definitely an user error on my partIt all worked now +And follow the configuration advice in the online manual +This problem is similar to ticket +Wow I totally missed that you narrowed it to a specific commit +Which part? The s and constants are used in branch +Return an instance of instead of +Regenerate patch with updated indentation +The reason the redirect doesnt work when handling the view rendering exception is an Ajax redirect is achieved by setting a response header named +Made a few minor improvements to the overlap handling code and comments +who will be surprised when they upgrade if we change it? Thats the most responsible way to signal were taking stability seriouslyIf youre suggesting that we fix regressions as soon as they are identified as such instead of propagating them across more releases because we prefer the new behavior then I emphatically agree +will do everything through the which should hopefully make all this a lot easierWill run tests again but I think this is ready to commit +I do agree that in general the fix is the right one in that you need to update the whole row rather than part in the collation caseEither as part of this fix or log another jira I do believe the other block that handles all fields being the same also should be fixed +This would be indeed a blocker for the release +Verified and closing issue +Can you create a patch with svn from the folder? I cant seem to get this diff to apply using svn +Yes but the global binding is a jboss specific binding + +I wonder which instance this is coming from? Are you bootstrapping such a factory on your own somewhere? If this is the s there should be a serialization id assigned to it +I might not have the fix for in before but Ill makesure it gets in before +I hope we could expect a quicker release of a version with that bug fixed +Here are some tests for this but I dont know how to make a configuration option for this as none of the other cache behaviours appears to be configurableJon Is this sufficient for you +Do you see any exceptions in the error log? If so can you attach them here +Committed in trunk at r in at r. +The tss tags are not required ejbs still work fine without corbaI agree the current solution is not optimalI think there is a problem with having the tss bean reference the ejb container namely that if several tss beans reference the same ejb container corba wont know what IOR to serve up when you look up that ejb +Here are the logj dependencies described +Sorry for the cockups +Possible easyfix +Problem not solved after all +Sorry I forget to post it +Fixed +CXF should definitely be included in the standard distribution +Verified +Hi about that +I agree with them tooWe need to upgrade to +Merged to release branch. +The reason is that these apis are async apis and their responses do not give the sensitive information +In class I used classs create method to create cache and object as well +Perhaps our adaptation of DBCP has left out some crucial processing +All the regression tests ran cleanlyCommitted to trunk with revision Committed to with revision . +were the client connecting and disconnecting? How many session is large number of sessions +We will emit the events +I upgraded using the provided ant script to and now my app no longer uses the proxy set by a or set using etc +Ill verify it soon and update once done +If I read that right were talking about a change that at the th percentile saves at most kb? Im skeptical that those modest savings warrant this changeAlso how exactly were these gains measured? In what unit can we expect these memory savings +I think were done with the component then +Updated Y! patch per Devarajs comments +test failures are unrelated +I put the initial version of into extrasNext Ill start to make example and improve details + relates to JBDS Is this issue specific to JBDS or JBDS or both +The problem was that when doing upgrade for we were hard linking between incorrect directories +question is what is a better solution for an example +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Or you do a full build of course +I do like your new testLooks like you havent signed an ICLA with Apache +No matter what options it compiles with should if there is a build failure. +The application works as it was with the fix forThanks! +Please verify the fix in the latest. +A agree that this is a more general problem but I think it should at least be mentioned in the documentation for groovy JMX because it took me almost a day to track down this problem and I would like others to have a smoother journey +This is a the main issue is the fact that ASM used by Karaf doesnt work with Java +Will attache the verbose log run log with problem is due to assumption about filename +Patch committed to both branch and trunk +Eli Im good +on patch version +I fetched the latest snapshort of branch built and use it with my maven mine! it works +Do you know which version of Oozie will have this HA feature +btw that example uses a return value instead of a predicate +Patch committed +I dont think the EC argument has ever been used at least not in our codebase + +Or make the title generic so it fit both situations +The patch was created with Eclipse and includes the methods described in the description as well as updates to the test cases +Do you have any more information? Eg a stack trace +Will changebq +This would involve the following class name changes to to to If there are no objections I will make this change +I left it unchanged so force declared and ignore native +FYI the fix was to make sure S processing normalizes paths to just like everything else in +Manually tested +What exactly you suggesting here? Can you please be more specific +Ill kill the FAQ entry when is released. +Works with While compiling hive fails to untar so I have to manually install it to make it go through +The error log gets a message indicationg automatic connection to jmx has failed +If I have a another problem Ill let you knowAgain and again thank you I can continue my project now. +thanks for cleaning up +One hack would be to see the created time of the pid directory in proc and then subtract it from timeofday each time +These commands have been updated in the code +I cant reproduce this problem umlauts in titles work fine and I believe you have a configuration problem +I am having the same behaviour as Prikka +We can then resolve as fixed +Also only the minor builds +Thanks! Checked in as r. +bq +Snjeza after checking with Tom Fennelly it looks like we can verify that ESB has been installed by checking to see if the AS deploy directory contains the and directories +We will address with a redesign of the renderer after. +You are rigth I meant storing CL somewhere in thread local during loading +drop it in srctestgroovytigerAs Kennard pointed out its something important to achieve a friendly integration with and their frameworks esp commented in a bigger refactoring is needed here so I will have it postponed to +Sorry I just forgot to do it +I already attached a project which reproduces this way back in Sept This and have just been endlessly deferred back since then + contains test source code and class the problem in classloader +I made a change in Git ID ccaadcfdfeecfcf to not throw the exception in but to simply return null and modify callers not to add to the collection in these cases +like it +As both these two commands do unlock for all locks closed within the transaction when the lock is acquired by another transaction prior to the second unlock the lock can be removed from the mapWhen the lock is removed from the map the transaction cannot get the lock to unlock it after it finishes therefore the lock stays closed and all transactions that wait for it fail due to timeout +Lets see the test results +Overall the patch looks good +it is slightly hard to follow but if there is a network failure wont the connection reset and a new be created anyway? it would be great to add a test that simply drops responses to clients without closing connections +Yes this probably a duplicate marking as a duplicate. +Verified in +Users can directly access those from HDFS done folder location +svn now ro except sitewill look at mirrors next +Thanks for your reply +looks Suhas I just committed this patch to and branch +Im definitely for a constructor for all the properties +Bulk close resolved issues +I also prefer to NOT return partial results +Please reopen if more work needs doing on should read Please create a new issue if more work needs doing on this +Id be fine committing this patch for now and opening another to address findbugs issues +We will take a look +When focus is assigned to the component the momentarily gets focus +Absolutely +Done +Im not planning on targeting it for +When running cold I get about s to s which probably would improve with more warmup time +Fixed in next nightly build in the branch +Sorry should have created the issue with a Minor priority but cannot find a way to edit it now +Should be good now please verify + is a proposal of how to make return the correct statement when using logical statements to create the result set +Fixed in CVS +Can you please add a comment which includes the code you used to cause this error? I would like to see what you are doing in order to determine the proper fix +Adding dependency on because otherwise tests are very hit or miss +libsecurity +Thats already possible +Applied PAM patchThanks PAM +I dont know if my workaround works or not +Now when the option is set the task depends on the dependencies of the move tasks for files but the move tasks for files do not depend on the task as there are no locks on these files so there would not be any advantageAdded a new test case for this additional functionality +Resetting until Wayne can weigh in +lets do this before committing the code +youre to fast for me +How long did it take me less than would be nice to see someone takes initiatives rather than just pulls out all kinds of stops +Added version check before synchronizing +we can open new issues for the other data structures that consume RAM +I modified to have enough data to cause a problem importing and I noticed that the IMPORT error message now does have the line number in it +Sometime in the next week or so Ill create an initial patch that has it as a separate component for so that its existence would override the order implementation and if there is support for it to replace the current java implementation it would be a simple effort to add those files to the order component and remove the current implementation +Marking for will have to make due with my customized version until Ideas of the timeframe for +Reopening to change sFixedDuplicate +Fixed in after release +The current implementation of Checkout process is way too complex and rigid +I remember now is the one raised. +If a test case can be provided the reproduces the issue please attach and reopen. +Yes +Its impossible for the cache to be present +relies on the application to store the lowest ranking doc that was returned previously which means storing context between users searchesI think assuming the math plays out that once you show the gains to be had here esp +So I was seeing this when using jrubyc +have applied the patch and merged master into the branch +pushed to and with JBT. +Thanks for your hard work. +Hi NadyaThanks very much for your review and useful comments +Can you attach your xml configuration section relating to this issue? Is the javascript or server side validation failing +Giri can you please confirmAs I commented earlier All tests passed on my machine +Well we could change the log ERROR to a WARNING but the whole point is we let start level change process be a little loose so we dont run into any deadlocks +So I think overhead is fairly minimalbq +David is doing experiments against it + +I think these would tend to be tests since they are unit tests nothigher level tests like we usually have +This was time consuming and rigorous and its actually the reason why the from doesnt diff well essentially I was forced to rewrite it +If you choose to do so Ill open another ticket for more changes to Maven +Updated spec +The complete stack trace would give us valuable information about the context in which the error occurs. +Test Case is are changes included in r which could be ported as I think a fix for this issue is provided there +Will be fixed in have a problem to resolve this properties proxyuser and proxypassword are available only in ant is it annoying if the generated scripts are usable only with ant +Committed to trunk +The PDF opened fine without crashing for me +Other changes in the code are to enable interacting with the implementation in the loader +is only testing whether a function is called but not verifying the contents printed outThe object used in print is different from that used in +Transferring this issue to Jesper since he is the JCA specialist +Also of coarse if there are patches or improvements you would like to help with please take a look at our github information in the wiki +This seems like an incompatible change +The logs juststop after the message creating resource orgapachecocoonwwwsitemapxmap in and beginning sitemap regeneration in respectively +If you have a string preference pick one +Btw wizard is complaining about WSDL even when creating axis webservice +Lots of iops to create all those +Ive tried to roughly replicate this scenario in the unit testThanks for your help +If you take a look at this and its source you will get an ideaImho we should add an option so people can easily specify the ehcache configuration file name to use as a String optionWe should use the API from to load the resource as it handles loading a file classpath etcThen we can create a cache manager factory using this with the that is the configuration file to use +Im getting all panes CLOSED by default +This needs to be fixed +The AM would then as Carlo said continually be the one that deserves a containerIt should definitely be possible to make this problem show up under the right circumstances +Its well known they are buggy as hell +Therefore it does not perform runtime enhancement +Committed for in rev +Please review it +Ill try to fix it this for the test. +Attaching a second patch +Logged In YES useridThis was working fine in RC +I dont see any dependency between invocation of these functions so we may change the order +Hence this patch is for non partitioned tables +I think CP for PTF is broken +Could you please try to run all SWING tests and see the reason of these failures +Use git to export patch from Git +Attaching patch by changing the access modifier to public +Before removing the kres profile we need to make sure that all the tests are fast and do not require an internet connection as discussed on the mailing list +Could you let me know if this fixes your problem +for removing the legacy tar layout +Fix for this +NBD just wanted to explain my logic there +This issue appears to relate specifically to so I have it as such +Can you please add brief description of changes you are making in the patch for the benefit of folks not going though the code +QA Closing this seems to be fixed now. +Thanks +When I then runbuildjava The output is the same as yours +Patch is applied in revision +Theo thinks we should avoid dnscache but I havent yet figured out how to get ecelerity to use the rbldnsd instanceTheo hooked clamav up to ecelerity spamassassin is not hooked up yetThere are two problems with igron the disks arent raided the box randomly reboots with the following note in syslog on restartPMU forced shutdown cause The first issue is minor but the second currently prevents its use as a mail serverThat leaves idunn as a single MX which is hosted at UL +I cloned todays snapshot version and the problem no longer appears. +So the resource implementations are assumed to reflect collections in that they have GETPOSTetc +Indeed Ivy itself doesnt do such a check when you call resolve +Ive regenerated the patch without the test to be added in +move is and is into making the basic IEPS interface sorting by proximity and the rest an implementation detail that only the corresponding kind of should care about +Initial support for +They are in hcat implementation +Fixed with the patch applied for +See my previous comment +It looks like all of the above are there now so things will work as expected. +adding the new after changes in a user fails to submit a job will the resulting error message give them enough info to determine which queues they can run againstAs least an error message like Please run hadoop queue showAcls to find the queues you have access to +The parent pom is not a jar so maybe the pmd and some other plugins shouldnt by applied on such a project +bq +Test case attached +bq +I just grabbed from the tag and other than the known issue the script is working +A notewarning has been added to the Programmers Guide +Thx +The problem actually only occurs if I have and no +Addressed the comments +Thanks for the unit test that helped a lot +How can I use the +I will keep you posted. +install below repo before installing mgmt server cd etc +will post it in couple of days +Raising priority +Attaching and with changes to the following filesM srctuningM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideThe patch provides an introduction for each figure that lacked one and gives each figure a title if needed +Ill provide input but assigning it over to Haleh since she volunteered for this +versionset +Simon we can imagine a very simple solution when message is longer than characters Lilians could you attach the findbugs xml file please +Hey Ivan looks like a duplicate of no? Regards it duplicates +So is it assumes that if you have that you also have? That would satisfy the dependency of to which doesnt exist in sorry for the confusion +Bulk close for Solr +Will work with it a while and see how incremental builds do +Please apply the attached patch to fix this issue +Looks like writing the dummy entry takes about ms +Once its in spring mvc more people will ask of its equivalence in webflow +This fix will be available in the next patch release. +IIRC the reason was to not have them step over the shipped library jars unintentionally with a simple HADOOPCLASSPATH being set +Ill have to check the others to be sureOk a work around that has been in the Validatorexample for a bug in Netscape was reallya bug in the +I have zero time for this at the moment sorry +This looks like it was done? Can we close +Patch would be great +I will start implementation though +Cast that way it would be as though your Contained struct had a required field with an unset value for any reason +We have two nodes A and B +oh I guess that stuff wont make it into the actual release though will it? That means that only the pluginsconfiguration in the POM itself will actually be released which makes your statement for the confusion +Q It seems there is no way a cascaded layer +Ok +So this jira can be closed as invaiid +Please provide the steps to reproduce otherwise we wont be able to work on it as we never face such issue +Im going to try running your test since I have the same software installed on my laptop +I did Object should have been String +I opened the bundle jar with winzip and I see the following files source jar seems to be in there +Presumably the data will compress better before encryption +As suggested this work can be considered closedImprovements are in +Ill have a look at it that autowiring by constructor still works in the old way though For example if a constructor takes two strategy interfaces a single matching bean which implements both interfaces can be used for both constructor arguments +Thanks Chris waiting for. +Making it protected seems fine to me if that does work for you. +Mysterious CPI hangups on the server seems dangerous +Dupes +I know of at least one appserver that is not doing that +You can have a look at the result to see what you think if it works +bq +Discard previous patches please consider only this one +Will need to think about itIn any case the safe bet is to not use the same backing bean for GET pages +Improved patch +Well the main reason I wanted to expose the was not to make it possible to register more but to make it possible to manage an application and thus the facade is not sufficientI was thinking that we would simplify exporting of child elements by something like outlined in +run any tests +It would be better if the server stopped reading the incoming queue but I believe it is not critical for a first pass +bq +Fixed in SVN +In fact the only removeBlock calls in are both directly followed by persistBlocksremoveBlock calls a legacy function which issues an to the editsLog writing the path repl mtime atime blocksize blocks perms clientName and clientMachine + This looks good to meWe talked about to use as an optimization +I see it now. +proposed patchIve abstracted away the store fetch parts of the index implThis will easily allow us to add new storage strategies optimized for different scenariosOpen question How would we define what strategy should be used? I propose adding a strategy property on the index definition nodethoughts +Its awesome to get contributions! +Now that work is well underway with AS all previous community releases are +Removed hive conf from the additional jars list +libvirtd not killedrootubuntu service status startrunning process rootubuntu ps aux grep libvirt ? S May usrsbindnsmasq varliblibvirtdnsmasqroot ? Sl usrsbinlibvirtd d lroot pts S grep colorauto libvirtIt seems both are needed for agent to run +One of solutions to prevent this problem is to set the permission of the dir correctlyHDFS hardlink should follow the same permission requirements as genreal Linux FS and only allow the trusted users or groups have right permission to create hardlinks +I can look into that as Infinispan wants that too +The other options is as previous mentioned not possible to validate as some are lazy used later on. +marking as WONTFIX sinceits working for Gilles. +Well thats Hennings concern +If saveNamespace fails why will checkpoint run into a problem? savenamespace fails if all directories listed in are bad +As one of the aforementioned users who will breathe a sigh of relief when this is solved may I ask if there has been any progress on this issue +Resetting assignee and closing again. +Will commit that +bulk move of all unresolved issues from to +Attached is an updated patch +Hi Tina +Already exercised by existing code +But Ill like to continue further development and keep posting fixes and improvements to this component +Corrected both embedded scenarios. +Is there a Jenkins build with the failure +There arent a ton of them but its better than individuals again starting to say theyll host what are actually community files +This patch contains a fixed exmaple that will now compile +This means that it will not work on Fedora which only has +Not perfect but the message points in the direction of the problem which should be good enough for now. +trunkCommitted revision +In the data model resource book i have hotel offerings are on page not +The CSS file we use for our web console +The geoserver patch applies to both and master +However there are still loose ends in node type validation +What are the immediate requirements for this task? Is it correct to consider altering the buildpackaging process to includethe testsuite and its dependencies into the distribution as the primary objective? Which of the enhancements above are required and which should be added to the wish list? Do we have to include some kind of documentation about the details on running the different tests? Those questions might seem stupid to you but please bear with me since I am a newcomer and it is not obvious for me +We will of course ask Toms advice on this issue +It will probably be a couple more days we will post an updateThanksMikhail +for trunk and for branch. +Username chrholm +Heres a version of the patch that uses a unique exception to identify this situation +bq +verified on into new JIRA for in before on H bulid of extras site +The default implementation does work the way you say thoughIf you wanted it to work as a logger until a appears we would have to modify the Logger base class to use the passed in Logger subclass as the default logger if no was available +Would you care to submit a test case as well +In any case commenting this out might be a good idea with a version test +That would be very appreciated +Attached patch as described above +Investigating further +I have made a few changes to the code and tried most of the obvious ones +All four existing examples have been updated as well +Sorry for that please adjust the priority to something more apropriate +We are in the process of releasing +Then style the different elements that contain them not the buttons themselves to show the relationship +This issue would also address documentation changes required for patch attached would incorporate the changes necessary for the and current feature set of Capacity Scheduler and memory based scheduling +tmp extension then we will discard it +Thanks Konstantin. +The patch falls back to the old behavior if it has a problem parsing a hostname from the address +The latest update addresses all the gaps and generates mapping for pig with input and output feeds as parameters adds endend tests +rptdesign file through ofbiz +fixed with rev rIt seems that extracting the get part in a dedicated class fixes the issueId like to see what others think +Recently for a different project for test result comparison where the order was required we moved from to because the order of insertion into was different in jdk and did not work because of that to compare golden files +Created layout and deployed on snapshot repo so we can remove dependency. +as described in the brooklyn project issue link removing the private from the java method declaration fixes the problem for me +proposal for the short termImplement this entirely in using the hdfsStat structure which has the owner group and permissions of a file +VERIFIED +Looks good +I just looked at the protocol documentation and I dont think currently the btree scan will do what you want with respect to OPENMODELOCKNOWAIT +This issue is a duplicate of s related to configuring servers during installation now it just take path for SOP or EAP servers and use it as is but root fo SOAEAP installation is not the root of AS which runtime is created +The previous try ran afoul of some of the wiki syntax it seems +Adding link to that requirement +Tested this issue in Designer Eclipse Juno using the Designer download +Carlo please make sure the EAP branch of EJB does not refer to AS artifacts in the +None + +Thanks so much for the diagnosis! +Um the example code attachment appears to be the same file as the source code jar +I will use your patch for my current project that is planned for golive in weeks +Resolved with revision . +How is it different specifyingoffset length rather thanleft right Am I missing something +Also used atlassians xmlrpc +We solved this problem in ECQL +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Also attaching this to get back into was released earlier closing. +boot log when webapp is included in my deploy with attempting to earlier application that had already loaded to those JNDI names was you can see the bindings at lines of the The problems with the new begin at line +How does that clearly explain that if you are using vhosts those handlers are still available? A simple intro for couchapp rewriter vhosts is definitely a good thing imho +May I know when this class will be updated to usingSpring version M +Thanks Mikhail +The patch is attached +Attaching which adds tests for bulk import of generated columns +proxy all methods not only read methods +r +Thanks for the followup +Im awaiting for the actual full +Cloning a project is not different from giving a project a new name anymore +I agree with Steve that exposing this config can cause more issues than it can solve. +Lets add it as soon as possibly after +Any update on this is greatly appreciated +Updated in ESB codebase +Thanks very much Sid +Was the removal of this functionallity intended? +This is file of +Committed +Is that OK by you Suresh +Note this is a problem in the language statement cache +It is even easier with GuiceYeah those beans seems to help +comma and space are both legal filename characters +Testcase attached +I have tried to work around this issue but consistently come across stumbling blocks +This bug is also apparent if you build your Eclipse workspace with Pweb Poracle but have a launcher +Youll then get commit access too +There is a precedent when we added JVM locking support to prevent dual booting the behavior was different depending on JVM version so documentationexplanation is key to let usersknow what to expectIt seems more straight forward to only support the change with jdk and up JVMs +This is bug report +In the case of maven projects its not important since the hardcoded path is stored at the workspace level not in +Instead it si better to change the action namespace defined in and +John as a workaround you can roll back Jetty to by replacing the jar in the lib directory +Patch that modifies +Done +how is your opinion Flavio +Ryan held my hand and explained that while not at memstore flush time it could happen on a compaction makes sense +Thanks! I will look into that if you have any additional infos let me know +maybe im wrong for some dbs but if you insert a null the value will be set to null not the default valueits only in the case of the value not being there and hence you need to use before this becomes relevantplease verify +I changed my mind +Patch updated as header file name have changed from to +Hi Justin Id like to talk with you about this I need you help to understand this issue better +This passes all of the tests that trunk does +Im planning to focus more on the scenario where we execute map tasks on the cloud and reduce tasks on the local cluster like scenario assuming that the computing resource requirements are low for reduce task when compared with the map task +A +Two things +hmmmm ok this is weird then +Ive tracked down the problem +While should try to be reasonably compatible with the RI for user convenience I dont think its necessary to slavishly follow bad decisions in the RIIn this case Id be in favour of the patch that throws an exception two JSP tags with the same id is just wrong and for me an exception seems the correct action to take +Moreover users deploying on Tomcat series would have much fewer problems when reloadingJanne if I could I would discourage you from writing a facade for SLFJ which is a facade in itself + to trunk revision +Converters must use thisLocale when performing their conversion +This will work as there is caching per configuration per classloader +I am too in a lot of ways +It has no impact on the priority I assign to this which is that its on my list after a that will be done by mid next week +Please check that it was applied as expected +proxy connectors and repo groups also has alternating shades +I have committed this +I have the same problem with Grails and takes a single int as parameter and returns a this is a blocker +Change status to close. +Fixing this bug will affect the performance of the other of cases for which namespaces are irrelevant +Dejan thanks for reporting +fixed by using style attribute instead of at +bq +BrettThe code attached contains only domain classes and a simplest test case to reproduce the bug +Perhaps you can take it back +Patch to fix issue +all code is tested youre safe to introduce current it must be reviewed by a committer and that takes time +It is way too small using any tools I have here to make anything outSecond Hibernate has been perf tested against both and quite a few different times and each time has shown faster than +As a minor nit you might rename the communicationip to something like preferredipAs an aside Ive never been thrilled with the default that can get returned when we dont know the DC +pushed upstream +After some thought actually I dont think this would be too hard to fix in the a consumer is closed currently we dont actually get rid of it since it might have unacked when doing the close we can check whether it has unacked deliveries and if so remove it there and then +Committed to trunk +Can you tell me what command line you used to run the tests? It looks the slaves process isolator is unable to kill the executor due to permission issues! Does the problem go away if you run as sudo? +Core concepts from Spring HATEOAS are generally useful and should be considered for inclusion in the core framework +Link to corresponding CR upgrade +Attaching patch a which disables automatic index statistics updateI removed the second TODO because the relevant code will still be run even though the automatic update is disabledPatch ready for review +Empty contents are semantically different from null content +My only change to the BPEL was to set the process level default expressionLanguage to +So the alert doesnt pop up +Patch against trunk to cleanup code +It looks like RI checks for the specific case length +Would that still be a problem with a ms resolution +ryan rawson added a comment Mar PM here is a proposed patch the other option is to add a parallel map in addition to servers +HelloI would like to start contributing to HamaIs this a good feature to start with? issue info +Thanks for the since is now released +Size is reserved to mean there is a varint size following for lists items long +cap and decode in wireshark +The intermittent test failure is due to the partition reassignment failing to complete due to +Committed revision dev branch +Hi Steven looks like a lot of good workI made a trivial change shingle is implemented with assert for better checking +Thanks for the quick response Ben +I dont think we need it +Do we already have a model for? I was not aware of that +was missing a change in the moved file +The change looks fine to me +Could catch the exception and do something reasonable like to the same URL or at least show a custom error page? + excludes from the test should be included again when is at revision +A lot of the logs seems redundant +ThanksShahbaz + +Anyone using jslint in a Java project? I cant understand how to allow both static analysis to occur on same project +doc image information than to POI beta in r +We already have a rollback system in place +Fine with closing this +It is that or I am misinterpreting the issue entirely +Lets avoid straw men here pleasethe CF is never optional is a good point +Committed branch and trunk +Please test +Reopened to fix the version tags. +BuildCan you try it again? if no problem pls close it +FOP cant deal graciously with the second last cell whichoverflows the available columns +Lucas you need to comment static tab out to see this bug +The sheer amount of complexity in managing the replicas and partition just bothers me +Rikkola told me on IRC that he has a fix for it and he will push into for an update on wanted to note that it is fixed by developer +bulk defer minortrivial issues to +Merged into the master branch after verifying default and that all unit and integration tests pass with these changes. +I updated the following files based on comments from Army Dan and SusanThe patch for these updates is derbyXMLderbyXMLThe files that were updated areFilename TitlerrefjdbcJDBC escape syntax for fn keywordrreffuncxmlqueryXMLQUERY operatorrrefsqljtypexmlXML data typerreffuncxmlexistsXMLEXISTS operatorrreffuncxmlserializeXMLSERIALIZE operatorrrefsqljCAST +just stick with the last defined relation +Is this not a protocol issue +Thanks Surendra for verifications +If were just going to have CF TTL being sugar for clients too lazy to apply what they want then Im not interestedAlso if that client happens to be Pig or Hive theres not currently a way to set +Slave please fix it for too if the fix is simple and safe +used to just run embedded now I think it will run for embedded client and encryption +and checked into perforceso I assume this is more likely a issue than a general one? +Im sceptical about work day but havent checked a corpus +Please your application with the latest snapshot +After resolve we can enable the schema validation by adding the module dependency +Im not sure what this has to do with localization +Although there is a single writer at any time the complexity of these protocols arises when that writer might fail +Thank you Aung +Closing old bugs. +Apparently an Xss purge was already committed. +Problem due to incorrect serialization in system VM being used +will release to with +added missing apache header to test and removed author the assert to assertSame instead of in commit bbc +what kind of errors +The missing have been reinserted into the Server Configuration Guide +Tim can you please close the bug. +Thanks Lars +That change has been incorporated but the deployment states need further cleanup +Done. +traced the start of the problem to revision for Derby . +Thanks Bruno fixed in you ScottI also backported to at revision . +but as said on mailing list this one is also not complete. +Thus we should not stick private keys or passwords that give access to outsides machines on any Hudson server +Is method shadowing the only way to describe this +This change affects pipes and streaming also since the environment of pipes and streaming tasks comes from the parent java tasks +Fixed with SVN +Version uses eclipse jdt compiler and jdt. +If no one is currently working on this I plan on making a stab at it with the hopes that someone will pipe in and change my plan of direction if Im going the wrong way +Thanks for the help Vinod +introduces a factory to create jndi context and do bind operations +poll will sleep for up to msCorrect but for the purposes of this discussion I think the poll for up to ms vs the thread sleep amount to about the same that is dont block forever but have some wait time +so to svnignore for moduleslunisrcmainnativeport moduleslunisrcmainnativesig and moduleslunisrcmainnativethreadAdd fdlibmdist to svnignore for moduleslunisrcmain OliverYour patch was applied at repo revision rPlease check that the patch was applied as you expected +Unit tests are passed +Thanks for the patch Hari. +How about this patch +Im hoping for a unit test from Kaleb before I commit this +Thanks Jacques and All. +Dont you need a target to build the jar fileThe overall directory structure looks good +Im fine with a single class with two different runrunSingle method +is about mb larger than +Thanks for the patch! +Pulled Justins updates to enAU and enGB localesnlNL is still out of sync with rest of the locales +Where is configured in the child project or the parent or perhaps both +looks. +Hi RaphaelI agree with you is a very useful abstract component for us we can add many useful operations on it matrix multiplication and matrix transpose jobs are good examplesIm now working on the matrix norm the need svd operation its really expensive is there any light weighted method can let us get the biggest singular valueThanks are the beginnings of single machine SVD operations in TedThanks for your reply Ill take a look at Dmitry +Feel free to reopen it if you can provide more details +Closed issues related to Roller release. +Why are you inclined to close it wontfix? From your previous comment it sounds like there are many things we can do without going to a full two pass parser +Looks great and a lot of work tooSo the deal with this patch is to separate out the aggregator handling code into different modules but still operating at this stage on zookeeperThere is definitely code duplication in the masterworker handlers is this needed or will all this be changing in the handler modules as we move to network connections and away from zk? What exactly is the difference for the master and worker handling code? Could there be a common base class handler that implements the common functions worker and master handlers need? Or is the difference hard to factor outAnyone else have any particular problems with this code or can we commit this +Unfortunately test fails with my implementation of the validatorCould you make best of two implementationI havent found how to the issueCheersVlad +Patch backported to branch will appear in release. +As I remember Carsten Ziegeler fixed the component manager thenSo I was surprised that the same issue was introduced again into +The header graphic for the design +Should a separate jira be filed for fixing headroomresources available information which is being sent from RM to AM? At this point I believe it does not account for whitelisted or blacklisted nodes +Not sure I can let that one go by Id suggest that either be accepted +This is not a valid issue anymore +It would be fixed in +Mohan is correct an instance of a tag handler may only be reused for tags that specify the same set of attributes +Annoying that uploads are so hard to test even with Selenium +ok! So the version field must be nullable? Fine by me +vdb file is deployed then you are seeing expected behavior +This is a problem and I want to get a release out with a raft of smaller easier problems fixed before I tackle itNow if you were to post a patch Id pull it into the release if at all possibleSome of Dans recent patches publicized an API that allows control on reading referenced imports +VM creation in created with Ubuntu LTSNext comes puppetising will do that tomorrow now +Fix for +No IT +Skipping tests should reveal major issuesFor the second failure is the issue with Commons? Is that not guaranteed to work on Windows? Why are we using that instead of regular File API deletes though? I dont see permissions being messed withetc +In that case are we Ok to go ahead and I will commit the fix + +Thats probably too much +Done included in the master trees +committed must have not recompiled the update properly +This was done as part of the request to migrate m afraid I missed this prior to my comment on +jgroups is now a complete OSGI bundle since the have not still news on rhq news are you looking for +the attached zip doesnt contain the jsp or gsp grails doesnt zip this up +I can run the full suite does Hudson QA bot no longer do that +Originally I was unable to get any security page changes to show up in the +sorry about that +I didnt notice +I am sorry I do not understand the reason for the question +Basically somehow make the order of insertsgets consistent across by using appropriate datastructures +Is this a good idea +In production we transfer binary structures which will benefit less from gzip compressionI do appreciate the suggestion of using useCompression as a workaround but I think theres a structural issue in the broker core hereThe broker should not hit an OOM while attempting to service a consumer queue read request +Ill give it a go latest tomorrow +My preference is to own the least amount of code possible +It is not easy to add a testcase for it +package +Its done +That potential problem is what the method solves it insures that only one copy of the object is returned +That is nothing we do they do +Thanks +I didnt test it but I think it will work +It also leverages support on Windows for gathering such information +Committed the patch in trunk rev. +added in scope should also include a second profile showing how to install using using usePut jcr install bundle in r +Closing +NOTICE exists at top level +Hi JacquesMy apologies for using a more terminology than specific wording +This means that itll be hard to move to maven if we want in the future +Kim Would it be accurate to say If no permission is set for the user the setting for the default connection mode is returned +Ive hit another issue in the dependencies depends on which uses Mina +In general Id prefer it to be configurable +I think the font cache should have some sane upper limit and stop caching the fonts if the cache already contains the max number of fonts +Committed patch to branch as well. +This is not a bug +For example if you want to flush data after every field has been added today you need to do it both in the finish and startField methods and in both cases you need to check whether startField had already been called earlier on +Attaching patch containing icons and classpath fixes so that ZK and HDFS functionality is usable again +It seems that compiler should instantiate the vtable for stdiostream in testcase object file but dllimport attribute on B mistakenly brings it to think that vtable for base class is also should be imported from external shared library +Closing for now +Yes but that could be a cohesive that would not be worried about serializing youre right +URI is posed to move to its own subpackage where thischange can be made after is released +bulk defer to defer to +Thanks Minto and Greg for your help +I have also sent you an regarding the sameThanks and regards allI am pleased by the interest shown by the community on my proposal +Thanks for review NicolasI just committed this. +Can you put more details on this issue so that we can reproduce it +Running TMU with this patch up on hudson seems to fix the stuck condition where cant find ROOT on startup of cluster around +Try this out in the Pico codeline Mike +G +Based on my testing for in my testing cluster I can easily open k regions on nodes in around minutes +this is expected +that is expected behaviorThe Ids have to be unique within the class hiearchy otherwise it would not be consistentThis would also mean that you can not map a relationship as both ignoreAs a lazy fetch strategy on any of this object graph will proxy the one side if this doesnt exist then you later get the problems described +Closing all resolvedfixed issues of already released versions of Roller. +So I guess Im on this index splitting method because I think we can do it betterImprovements welcome No reason not to start somewhere though +is there any progress on this issue? No Im sorry to say and I dont know of a work around +Then in the unregister method remove the mbean from the server after checking if the mbeans hashCode attribute has the same value +Ive noticed several backwards regions in a meta that Im trying to restore that had the script applied to it +run main method to get detach process to never detach LRS fields +when the archiving have run you should refresh the exploded dir so the changes come visible +Ill commit this shortly +how much time does it take to create splits for million blocks on machines with blocks per split for example +Fixed on trunk but adding optional property on the connection factory and in the connection +Added to Developer Guide rather than Migration Guide +For what its worth heres a patch of my changes against the tag +Im not sure it even is an issue +A second part is that there looks to be some insane fieldcache usage that needs to be fixed in these tests again because the code wasnt hit before we didnt see it beforeI just keep randomly taking peeks +That should be fixed +There were actually two lists on that page and the patch removes it from both +We have this problem of using different loggers at various places in codebase +It would be great to work towards getting some code running inside the Tuscany code base it would also be helpful if you could put some more words around any key design concepts and the issues you have encountered and solved or have yet to solve +Fixed +In the issue it saysThis is unnecessarily expensive because the binary is fetched +The problem in your case seems to be the before that you have in your version fails but using works. +This is because testing of this code requires a KFS deployment and that is done elsewhere +I will drop the priority of this issue if I can +Attaching +This has to get done +is a new feature we do not need to propagate it back to +Czech again Czech added to svn +ZachYes I suspected there was still a problem for you +Taking service registration out of the synchronization on the component manager instance in the method should fix this issue in the Felix implementationWhile this fix targets the Felix framework implementation I assume this does not harm any other framework implementation but may prevent similar issues in themFixed in Rev +Please follow up on the project. +I have implemented a solution that exposes a skin pregeneration service via an internal view +Have I missed something here +I updated those pull requests to use naming suggested in its not possible to testreviewmerge it until is fixed and new builds triggered once changes are applied +Ara el riure se ten va i es converteix en plorera +donefirst action mapping will be the soapaction for that operation +any further feedback to this +Thanks Bertrand +And if its just written for humans +Looking into it now +Released as a part of release. +Thanks Jerome +No I just see eclipse blink when ctrlclicking. +I am not sure bringing version from to is a good thing +Verified +This refactoring has been performed as part of the move to svnpubsub +I still cant reproduce the problem but well keep trying +A general question I have is whether it is ok for certain name node hosts to be unresolved during startup +Edward do you have some time to update this patch for branch +That was the reasonCan you put this also in if you remove the generics +Gabriel can you look in to this? Higher priority than WMS tests though lower than first round of ND work +So Im for RC for the present I have no idea thoughAlso Simple History showed SOLR REJECT even when I placed on Tomcat and posted to the Tomcat +Seam EAR project created inside with as rule that means if I got it right Auto Build take place only after finish its work +It uses a tcpfaulty +Attachment has been added with description Area tree generated from attachment has been added with description PDF generated from attachment need to have OFFO in the classpath for the problem to show +Either way if the set is lowercased or not the lowercaseing is also applied to the values checked against the setIf the is applied before the stopwords there is no need for doing +However the patch will have performanceresource issues in situations similar to those covered by +Reporters of these issues please verify the fixes and help close these issues +Indeed thats what I ended up doing in that page +what happens when you remove the spaces in the list +At last we can close this one +This has been fixed by thanks for the patch I added a test case for this to confirm that the changes have resolved this +PS just removing Dworkspace etc from the patch file worked as a charm to get it appliedKeep the patches coming +Dots were reserved for namespaces +Please do a review +thanx for ur support and waiting for ur input +Attachment has been added with description fo file demonstrating issue still persists in at least on Windows XP platform +Basic support for this is now committed +In addition we also need to put deprecated operation last +Ive used Groovy m using published snapshot but nevertheless Im from the solution didnt cover that case previously but has now been enhanced to cover that case as well +Ideally I would like to remove the redundancy and make it self contained by pointing to a schema file in the similar way as done for schema references in xml files +Fix checked in r. +If a custom directory is used it can be specified by pathto +Cool yeah I was not implying that we update the apis +This is still an open question +VamsiI enabled the tests that you had committed +Its on minicluster so same machine +Forced to treat literals characters as literals rather than using params. +So the giant routing method appears to be mostly a for the error I suspect it does not affect production if youre running on a reasonably recent all errors from jitting should be caught and as a result the method just stays interpreted +Please ask to reopen and provide sample code showing issue if you run into this issue again +This could trigger the deletion of blocks at the datanodes Then add editlog entry outside the lockAs this jira discussion demonstrates between the above steps if NN crashes there is possibility of block deletion on +Would be great to get on and trunk +The attached patch implements what described in the issueHowever we also add simple threshold to determine if it is better to send txnlog or entire snapshot +Documented as a known issue in the EAP release notes +Thrift has its own type system which includes byteii +Patch to add config options to conf i believe these must be added to the default config +Patch for the new in +Thanks for the patch Ryan +A big advantage of this approach is easier upgrades to the state machine +So Edit means edit the object rather than suggests an alternative to the text field +Updated test based on feedback +Thanks +One solution would be to not have a at all +a safe mode exception or other prior to the permission check will divulge the full path resolved from an inode +Hi IgorYou are right wicket already supports this a least in +Please wait or press to interrupt it +Rather then make this issue huge +Well done Julien With Sonar response compression With response compression +Timo I would recommend to open new issue attach your eclipse configuration there and steps to reproduce +Resolved as PR was integrated. +Updated to hsqldb +Ready for testing +Ive committed a number of changes so the test can now be run as follows java false Embedded small Ive described further runs in a comment for and attached a number of run results to which seems another manifestation of some underlying issue as this one. +OK Ill check if this is possible and try to get this to +test VM successfully deploy and runAny idea how to fix it +Derbyall ran cleanly on linux except for message encoding diffs in urlLocale messageLocale and iepnegativetestsES +selckin that problem is more than maven actually +Closing issue +trunk +Attaching new patch fixing a minor bug +The fix linked to by Jochen seems to fix that issue +I will try to commit a fix +The messages can be ignoredLazy mode true means that Cocoon will only load each sitemap component when it is actually required +Do you have sample solution code to submit? for handling in have no idea what King yy format is and as has submitted any code in response to Vics comment Im closing as WONTFIX feel free to if anyone wants to submit codetests. +Changes made and commited +Build with ant jar or ant clean jar to completely rebuildHopefully over the weekend I can hack out some articles on how to contribute to. +So an underlying problem is that the classpath is different for ant clean dist than ant clean test +Thanks very much for the patch. +thanks a lotYes filedthanks a +RPCnetwork exceptions should be thrown as I also considered as a container for passing exceptions through RPC +pom file but the release wont be until after the holidays +This behaviour is the same for all projects using the default template so users only have to learn this the project teams want to put other links on the page to other websites or maybe a couple of additional links to their project pages then they can use this General Link List +At first glance seems like a fine utility method to save developer time +Thomas didnt realize that what the query was doing +This seems like a duplicate of to me +Its not as if there will be multiple users using the same bookkeeper clusters and we want to block them from interfering +There is no planned update for AS Final release +Thanks +In this case it makes sense that it should be copied but does not make sense that it has not yet been set +Once has been correctly patched the test succeeds +Andr the issue should be fixed can you verify it please +This means well only accept which AFAIK is not the Axis goalNote that only the BP requires the soapfault element to have a name attribute +This problem did no occured for s difficult to determinate the reason of performance decreasing +New version that passes more tests +However the code in the that follows symlinks can still return a relative path after following the symlinks +All tests pass but Im going to leave this open until I add another test or two to check that the definitions behave as expected +Reopen +I dont think it will be needed anymore +FYI the next will use last prerequisite +Resolved fixed +Thanks a lot for this! fixed in master so now it should be solved for good. + +How far can you go integrating the twoI think that the result of the process on a document should be the samewhether you do the with Cocoon or with another processorSo adding a few attributes to the xixinclude element to control caching optionsis pretty harmless since that doesnt change theOTOH adding extra elements in the xiinclude element to specify URL parametersetc +ImeshPlease test other scenarios with latest jclouds and let me knowThanks +Why is this no longer relevant +I have not looked at the patch +Thanks a lot Bryan! from the doc end too +In order fix the problem the fetcher should either preserve the original type of the cookie or use cookie representation compatible with the cookie policy used by the fetcherHope this helpsOleg +This stuff use to be all nice working but I think refactor as executors seems to have lost these nice little errorsIve added a test that has a good few of these conditions listed outThere are some issues in that Im looking at but will run patch build to see if Ive broken anything else +So I hope Ive added it correctly +Thanks! +This issue is basically a duplicate ofSince the version of surefire is not locked down in the project POM Upgrading to maven upgrades surefire to which means youre in all likelyhood hitting which was fixed inPlease set the surefire version in your to +All arrays and lists in the should be replaced by Iterators +I will post the artifacts here and commit them to the repo +Thanks Chris +Fix for this for committing this Dan. +We use warinplace such that the front end designers can edit templates and javascript inplace and see changes without having to rebuild +But I do know I was able to replicate it yesterday and not today +I just mean I didnt think it was actually possible to do +Assume closed as resolved and released +I changed the affects version +PrasadI would like to give a try at your suggestionHive will be doing string replace on all string literals returned from parser +The attached patch resolves this issue and is contributed to the ASF with the permission of my employer +The unit tests required one minor change to compile but pass +This patch adds a property to When set to true it will dump the classpath of each test beforerunning it +The server restart is now noticed properlyand I dont have to worry about possibly lost packets. +Because in this testcase the schema is bound by parsing the XSD +Hi JacopoHere comes the biggest patch of applicationsproduct component of Ofbiz projectPlease give me your feedback on thisRegardsAshish JacopoHere comes another big one patch for applicationsecommerce componentPlz give me feedback for the sameThanksAshish JacopoHere comes small patch of frameworkappserverPlease give me your feedback on thisThanks Ashish Vijaywargiya +Easyb also needs to specify a runner class which might require very similar but slightly different setup well see how far I get +I wonder whether this belongs under a different component? It is more related to EMC than +I notice that in the installer we have THREE new plugins but via update only twoRob does the patchfix require or just the other two I reported above +bq +I have no idea what happens to the gossip output inside the ide +I tested this patch successfully on Mac and Windows +Hi GarethI saw your +Average Latency between costs around ms +Great catch thanks for the report +Verified in +Example would be using exec source with a tail F command on a file that is encoded +Regarding API we can add a FS specific api +Applied the patch +If there are any comments they can be directed at the dev list as it is essentially a different topic +James I think its pretty clear Willem is referring to CXF JAXRS endpoints +Could we get this in for the releaseIn addition to the positive result of the Hadoop QA bot Ive been using this in production for weeks without problems +Some things around failure handling make sure we can still disable the automatic storage restore feature if necessary? make sure the NN still shuts itself down if all edits directories become I say testplan? With comes to mind +could you fix this +So for this one we have a pull request that has been already merged on and one waiting in line for trunk +AddedBasic connection testingaddgetdelete participantaddgetdelete propertiesLots of other code workSee Raido great work +Thanks Jeremy. +Tested on +Changed list topics tool to also print whether or not the partition is under replicated +I believe weve also hit this re also using ajsupport on a field and when typing fast in there we get kicked out of the look at the HTTP header log shows that at one point we get two successive responses from the server the first one is a legit AJAX reply the second is a redirect because of a missing conversation +I believe this is a duplicate of +All seems to work for me +Thank you Roland +we will retry the patch once everything is stabilized +As long as its simple clear and document I dont think it matters what it actually is +Please provide a patch for +Thanks for your help +I will change this +So the issue isnt that big and Id like to agree with JukkaIll integrate Nialls patches to both subprojects so that we wont have any naming issues with new releases of +I made a little research and found that this problem existed if try to open the same file using WTP JSP Editor +looks good +Committed to trunk. +Please check that it was applied as expected +closing this one +Tests that was marked needs to be tested manually were also marked as false in +Made the entire column header listen for the click event when builtin sorting is enabled +But this may be a good way to assess how important such efforts will be to people in the long run +Input is highly apreciated +I ran locally on a patched tree and it passed +Compiler is next +thanks +SF was giving me some trouble. +Something else happened after that +t seems to me that determining that a server is an observer based on its weight and switching between observer and follower during ensemble changes are separate issues +Ive updated my Galaxy Nexus to and I dont notice this on the Galaxy Nexus which makes me think that this may be and confined to the Xoom +Closing issues which have been released. +We have tests for Alert in our and I dont have any issues calling them more than once +passed on as the bug fix to branch as well. +Added a patch and a test case +Andrea merged the pull request +As well a link to your blog about using APNS with Camel as well so people can more easily spot it +Could someone please help me deleting with timestamp Jul to svgVthe following visio imagesMade these images a little smallerGenerate png filesUpdated page files to point to new png filesAdded place holder for so that link has somewhere to das contents in +So postponing until we find someone who wants to do that work +Ive committed this to trunk and +Updated the patch against the latest s a trivial javadoc fixes patch +This can help identify if namenode startup is blocked waiting on a particularly slow disk while saving the checkpoint +Perhaps a much simpler approach would be to allow you to feed the primary keys to be indexed to the +I just committed this +commented on tests doesnt apply cleanly anymore + +Heres a patch that extends the test case with testing more special characters and by adding a operator as mentioned in the original bug descriptionThe test fails because the special character wasnt handled by the patch +This patch updates the guide a bit +Please check if you can apply the patch +The jars that we ship in the Spring distribution are only for convenience were not suggesting that you need to use those exact jars +Verified with EAP +On windows I just used ant installed in Program FilesPlease check that the bug is fixed now +Verified. +Here is the updated test that has threads hitting this route +Simonewhile I could provide a patch for for that specific issue but we should seriously take into account that there are highly likely other visuals broken which we havent spotted yet +I am unassigning the issue from myself to leave it for others who may want to work on that +will reconnect if not connected +To detect this seems like it will be helpful to expose the number of queued requests in the? This could either be a separate JIRA or part of this one your call +I looked at the diff between and trunk gridmix directories +This patch applies to version Brian also see for more history and a bunch of options +Thanks Harsh J +Do people think we should do it the way or follow the ANSI standard +Hi Timcould you do a jstack on the pid of the server so we say what it is actual doingFirst do a jps to see the pid and then jstack PIDThe output while be very long so better redirect it to a fileThx I attached two jstacks +Committed to trunk with revision +Therefore theres no actual way to improve performance with gathered write for now unfortunatelyI have implemented gathered write in MINA twice in my working copy and reverted back because it just didnt perform as expected +Yes Tomasz you must use Java +Unityouve introduced new japitools failure with the latest patchfield missing in All other japitools issues seems to be fixed +Do you have an example grammar or POM that can demonstrate this problem +bulk defer to +fixed in If receiving a ME it will be answered as done beforeIf receiving a ME it will generate another and send itto the defined target defined in xbean +Im not convinced that a configurable data structure is necessary +Because of this its fixe automatically since people dont expect to see a p outside the if block if there is a def p inside that scripts rules are a little bit different not requiering the def + +What do mean by the net effectThe motivation behind this feature was that if the operator notices something funky going on in the RM this would allow himher to call a and probe the system in order to find out whats going on and possibly fix it before resuming the scheduling of jobs +NMR cleanup is almost done revision +Benchmark that serializes a structure many times to measure the effect on performance of changes to the library andor the generated structures +I forgot to reset the CCSID manager back to EBCDIC for new sessions which meant that if a client was to negotiate all the subsequent connections would be using +Looks good +we need to get these weird links in our docs alignedcleaned up shouldnt be any IMO +Sometimes this results in the EOF error in system log and in the client other times it doesnt +Close this if its done +I have chased up making some arrangement for the QA testing to use this with an allowable license +closing this item has been resolved for two weeks without problems +why arent system properties viable for this? Setting system properties is error prone +Patch attached +I think this is ready to go +Closing as per previous comment. +Thanks +You could try turning on debug logging on the broker but you might want to also look into the perl clients code for debugging options +I will close this issue invalid as soon as I add the test +I understand there are many priorities in the draft but could someone tell me if thats at least a possible Bug Im not asking for an urgent solution but at least answer simply because it seems quite serious +Added an additional dependency registration +Geotools trunk committed revision Geoserver trunk Committed revision + +bq +Yeah recycling the is probably too complicated +Use the latest temp build + Applications that are running in a JDBC or JDBC environmentHow do I know what environment my application is running in? It seems likemaybe we need a paragraph in the docs somewhere which describes whatit means for an application to be running in a certain JDBC environment +Andrew can you drop us the email address you want us to use please +Moving to +I only went back as far is but it certainly would be appropriate to backport further if neededThanks Brett for tthe fix! +thanksCommitted revision +With the current set up everything is one is to set up a proper property system in +Hi PranvBrianThis issue is not observed in the latest masterPlease check this and update the bugI think this is duplicated by which has been fixed +I have commited last changes in revision +I too accept it +A quick test shows it not breaking anything in which it definitely shouldnt +My client code is deployed in a portal where the classpath is common to several projects +Dev I am planning to commit this change in coming days +thanks erwin! +I guess it looks like its recovering from a replay log? And it skips commits while it does thisSo the problem may just be that we accept updates before we are readyIve got to head out for a while but can look into it more later if no one else jumps on it +Ill wait til after tonights trunk build to commit to leaving this open so I dont forget + +This is the same one that is in the area at the bottom of the new permission type is introduced indicating the user has no permissions in the pool +Were upgrading Bamboo today to see if that resolves the issues +It is consensus to delay this contribution until a later release + +Obviously this patch has been applied no clue who this wasCan you test againregards fixed in the SVN +Ok now I get it +Fixed patch attachedTests run fine nowReady to ship +I just committed this + +Adds get +Fixed at revision should avoid breaking backwards compatibility in +thanks rahul just ping me with the fix for the cascading error messages you might need to be clearing them depending on how you have the flow for this working +Do you have any suggestion? Thank you +Please track documentation fix in. +Mimej handles empty headers just fine +However in this case it is a small fix to retain compatibility with the RI so I have no problem committing itPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +We should support this +Committed to trunk +Resolve problem that after deadline all elements from peopleAssigments was remove. +I am a bit confused about what our positioning of documentation is as I believe we are no longer publishing the Forrest style docs +One running windows and the other running windows +Are you against this feature? Can you please explain whyYes I am against this feature +It should be reverted m on on revision your full rebuild successfullRegards sorry for the delay I am in UTC so I kicked it off before going am pleased to confirm that the full rebuild was successful both on my desktop and on our continuous integration regardsBen +Set it to true to enable it as it works in Camel or older +Update fix versions +The patterns are actually ant path matching patterns so to match all child packages you need to use grails com +If you or another community contributor would like to contribute this support it can be provided as an subsystem that is maintained outside the main AS repo +Thanks for the quick review Kim +Notice when you click from the Giant Widget product to the Big Gizmo cross sell that the link to the category goes away as well as the Prev and Next buttons +Closing issues in RC. +we have got ready too so I will resolve this issue +uses value of to figure out if it has to start JTTT or RMNM +Thats what we do also in for a similar purpose +The older splits are deleted to limit the size of the task logs +Yeah I would think that there is nothing really to do about this issue since each subproject is independent and must define its own dependencies on versions as needed + +move to examples package +Im OK with extracting changes in too +To apply in cocoon thank you +i agree that writing a test would be a pain +Will need to figure out a good way to handle deploying multiple projects that contain no geronimo deployment plan to avoid conflicts such as this +By not propagating this HTTP header element the above exception can be avoided. +Editorial pass over all release notes prior to publication of. +also added logic to then on stale resolved issues +the default core in the example is named collection although it is not in fact a collectionCould you clarify that comment? I mean how is it not a collection? Just because a collection exists only for a single shard on a single node doesnt make it not a collection +Im fully daccord with your explanationRegards +Thanks AndyI removed the spaces ahead of String address +Applied workaround +Tested the fix and able to deploy the ear file +Thanks Shawn. +It works great now Thanks +All tests pass +Why deletes need to catch up with writesHDFS already has way too many config parameters +Wont the above approach trigger too many unnecessary repairs +Spotted that it needed to include and also MSGNOSIGPIPE should be SONOSIGPIPE +This issue has been resolved by the patch to +Using the JDK compiler the build took minutes and seconds +It is not necessary to remove or alter the parameters in any way they simply would be ignored as far as OGNL is concerned +closing stale resolved issues +Closing as Rejected incorrect deployment was due to the ejb project being handled as a regular jar due to bad packaging setting +Bumping priority on Stacks recommendation +Merged into the master branch. +I just went ahead and committed this to trunk since the difference between the reviewed patch and the latest one is just a small comment changeThanks a lot for the review Todd +The module couldshould be marked as Changed but the application should be marked as Republish since GEP currently redeploys the entire application +I am very suspicious my request wiring is not quite right +Also it will help me generate proper patch for +Im not clear on the exact steps to reproduce though +Assigned this to as its related to some other Producer Flow Control issues which really need to get fixed up before we release +Id love feedback and you have my blessing to do as you please with it +My tests have indicated that this is no longer occurring +verified +Ive also ran with no failures +So if the data is not received within the timeout an get thrown +Theres probably an argument to be made in having three simple choosers instead of one complex one that does bootstrapping batching and prioritizingRemaining +What are you looking for in particular +Test to be fix in the latest trunk already +New targets show up automatically +I am new for this +Fixed by csrcCan you please verify? Thanks +all issues not touched in a month to no seen any reports of problems here in a while and all this had hardened a fair amount by now. +Ive configured schema and solrindex and now its working without error. +Just committed on trunk all issues that have been in resolved state for more than one month without further comments to closed status +Derby receives LOB information sent via Layer B streaming +Those of us who want to run on Hadoop can carry around a private patch if you want to be safe Lars +All further commentary will be on that bug +Thanks! +This has also been reported to the Novell bugzilla as bug turns out that I had messed up +What are you trying to do? Why or when or where you fix image links +Possibly related to recent fixes in sounds like this may be a genuine bug possibly spurred on by the recent changes in +Calling it resolved for. +You need to run soapui load test for secs +Committed revision shutdown during Spring bean lifecycle +I found the server ran out of space because of logs now its cleaned up please try +Sorry for making work for you Lars +At a high level +I want to see the new engine working as soon as possible +need to have a runtime exception that is fatal that does that handling and the rest are similarly in m +That includes no call site caching specialties and no constants that dont need to be there +Attaching patch after pulling latest changes and improved test for AM recovery +Cannot reproduce any more +Second attached patch backports to to and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Whats the status? is anything checked in yet +I adjusted the priority of this issue to critical because having clause is very basic feature in SQL +I have the same problem +so we should get this fixed in the branch too +This will be an interesting test +See commit message for more detailed description of changesCommitted b to trunk with revision +We should fix the docs mentioned above +Thanks for the additional pointers +Ive been able to pretty easily fix the problem in the Memcache + by is strange that I cant reproduce some test fail on my machinesomeone have any idea +Tested the changes to plugin by deleting all of my j jars from the maven repo and deleting the jetspeed plugin also +This isnt quite right I misunderstood how the logj stuff works +There isnt currently one and I dont think it works in anywayI have however removed the downloads from the mirror sync directory +This solution while being a little bit more involved will eliminate the need to expose the system session +Added missing consistency levels to CQL grammar. +push to +TJ thank you for the issue report +Ill bug the guy who did it and tell him to post it +I have been working on it ever since this version was released but I havent found an adequate fix yet +In my demo I for sure do not set the dialect +how does this affect legacy code + +So I guess that I will have to spent some more time with it +Oh it doesnt matter +bug at +Reviewed patchRun test suite with patch and didnt find reproducible test failureWill integrate later today +srctestjavaorgapachehadoopmapreducevapplauncher +Splendid work my friend! The fetcher runs smoothly again! Ill check out your patch for this weekBut what about segments fetched with and without this new feature and propertyI assume id have to update the segments before this change with the property enabled and update the segments fetched with this feature without the property +See the subsystem note the use of the this sufficient +I want to evaluate whether the problem is in our codebase +It will be good to check for user group first and then permissions +This has now been fixed +I fixed a few things to make it work like resolution of types from different namespaces writing back from schema objects to JSON text etc +I do mean that the return value was different between original and Apfloat +fixed +It uses a new web and corresponding productAUgoto httylocalhostvatecommerceTestingI have primarily used the following productsGSTGSTPROMOLimitationProblemsRounding problem +I am behind corporate firewall but setup my proxy settings for git and node +These nodes should also get replicated +IBM IO libs correctly rejected toread them using default encoding +Sorry for all the unexpected inconvenienceCommitted to trunk. +Attaching a new rev of the approach +JamesI must admit I thought the combination of icons would scale better +the same patch to +I wonder whether we are facing a concurrency issue here How can use ever receive an reference that hasnt been fully refreshed beforeYou could try to debug and see what state of youre getting there +Why would both DUCCSUPPRESSCONSOLE and DUCCCONSOLELISTENER both be set? Since theyre mutually contradictory it seems the CLI should insure they are not both present. +closing as issues are fixed in +Fixed in in the release of the iPOJO manipulator +Verified this is still a am fixing it for all the commands but the deprecated ones likeAn equivalent which is supported will be help +For blacklisting scenarios first needs to be fixed +Sorry didnt have time to work on last weeks +Tim Im confused as to where the events are +Should I send another patch against trunk in case this one doesnt apply +Based on its state we should prob deprecate in favor of the precedence thats about to land in the new impl +to everything +Thanks AswathYour patch is in trunk at r at r +So all issues not directly assigned to an AS release are being closed +first case should be fail +was the one I committed yesterday +trunk Need documentation at wiki all resolved tickets from or older +The class field should actually be job of course +In CXF MESSAGE mode means something completely different and since this is a CXF component it should definitely match what the CXF expectations are +There are some problems with the zip as theres no license for each jar and the poms are not in the poms subdir +Thanks Tim the patch was applied as expected +Applied in revision ThanksChris +Isaac I think this issue should remain open until the fix is actually in a release until it is fixed it should be documented as known issue +Given the performance results I noticed in I think that the next step is to preparea patch which uses the technique for queries but which uses the new technique for ROLLUP queriesAt the same time Ill also need to resolve the merge conflicts and attend to the other review comments +Pages should be +Hi Julien letters A B +Thanks a lot for your helpWe finally released and everything works as expected! +So it is an upper bound +Is there any method by which the files are consolidated and not held up by a single unconsumed message? Expiring the messages doesnt really work since there needs to be a consumer for messages to expire +This is a really cool idea but I dont think this is going to make it into TRUNK before the branch gets cut +I suppose that the issue is not actual. +Mohit Please open a new ticket for this issue +This is better +sendFile works by putting a in the sessions write +Works on my machine +Thanks Rick I agree it is a bit hard to read +Linking was added +Works in IR fails in JIT and AST on current master +As of today Spring . +there may be some methods recently added to or which warrant addition to but they should be considered individuallyRegarding the specific example of maxDoc +I havent looked at the patch yet +But this test class existed in the mapreduce code +Ideally should store the last in date format +The change I made was to allow a Throwable converter pattern to be specified +Your little fix in to add a check for null just made me wonder about the validity of the solution when somebody adds multiple documentation elements without specifying a source attribute +The attached patch replaces with in all the message files and in +This patch fixes the of the +You should apply your patches on a local instance and create them again +current implementation uses bytes array within the same class now this is moved to +The code pasted in the description was mangled +I think that is a better place for this and its implementation than packageI agree +If modification to the debug info is required then this can be covered by a new JIRA. +Implemented and tested waiting for rollout in coordination with Bugzilla +For instance we dont try to delete data files installed when the plugin was unpacked remove entries or artifact aliases +Fixed in SVN +HiIs there any news with this issueI am using and it seems like this problem is still present +its not intuitive and performant +This should be no burden to anyone compiling velocity themselves and it may be a great benefit to many users who upgrade +But didnt find a guided pathCan you guys please help me how I can integrate the bothThanks in advancd +on patch but please make new patch fixing the empty class comments and the unorthodox test naming convention so we can commit +Please review +Has there been a resolution to this? It definitely appears to be a class loader problem if it works identically in other application servers +I dont like either one of my ideas much to tell you the truth and stick by original assessment of hard and weird respecively +There really isnt a default just one ctor that has a default and actually I think we should nuke thatOlder lucene versions have to specify the Store and users are used to this already +Moving out of the queue as these could not make the feature freeze date +They caneasily be used to manually configure a database connection or a mapping +Account Name Domain Id INFO NULL Completed Successfully completed disabling account +but with the current implementation in the read status is calculated also for traversal isnt it +A milestone release itself is already bleeding edge +I think this works for me as a first have activated organize imports on save for a smaller test project there it did not cause any you need more information I will provide it to it is not possible for me to send you the project but I can easily test some patches for you on this +Not complete in time for +Used for request to fix the issue +batch transition resolvedfixed bugs to closedfixed +Committed at subversion revision +and built OK thank you. +I am running on JDK and I havent seen any in my logs but Im not sure I hit this conditional block +Patch that implements conditional loading if is in the is ready for screenshot a few more Ran! + +Hi FelixThanks for raising this it should now be fixed in the latestPhil. +Added test output that highlights the being sent and the protocolSesison being flushed but the message never arrives on the broker +On second rebuild the got included +Im also There is no licensing reason why such programs cannot accompany Apache software as long as the NOTICE file identifies them +Attached is my suggested fix for the issue +Fix affected to move initial configuration of the console to after loading the initial plugins in Rev +Is there any way to consolidate the versions in the pom in the resource bundle so that we dont repeat ourselves? and need to be manually kept the same as the rest of the project +Yes this and other related settings is what needs to remedy +since the patch stops the copying to examplelib i fixed the topy so it finds them in libm back from vacation and I realize that we have gone a long way in this issueThere is an issue which enables of object +Or we could somehow try and encourage people to upgrade to IE +Was this fixed by +Fixed in r +This can be easily verified The global static is not initialized in its declaration The main method has a local writer declaration the static one is untouched and stays nullSo this is not a bug of Lucene but your code +Had the chance to review patch looks good and integration tests all passing for me on AWS +So i think it would be right to fix this in framework +I feel ridiculous nagging people to review this so applied. +Could you provide a patch to solve this issue? +Hi Jing SureshIm currently testing the patch with trunk +Can you confirm that with an update to trunk you no longer see the issue and resolve this JIRA if not? Assign back to me if there is still a. +Likely can not deploy our updated quickstart to openshift until later because of our deployable DSs +If you see any more breakpoint issues please raise a new bug for it and attach a code snippet to recreate the problem. +fixed in branch and the very first page of demo in IE and refresh the page +My guess is that this is failing to launch correctly but either way this is a very problematic area for us +creates the table fine +Linked it to a bigger issue about using marshallers consistently. +worked through the last two weekends +Behaviour for flush and should flush the data +Regarding the use case where you have a server running and try to launch a nd one on the same ports any poll would be answered by the first one and we have to way to find out about that +And that is then causing the problem + +AndyIn we wanted to mock a that comes up after getting restarted +There are mature clients and software tools available today that support both WFS and GML +I will take a look at next +renamed patch deleted old version +I post the webstart folder as is to +Yes indeed this box was unchecked however it didnt start working until I also checked other java proposals +One solution could be to download the JAR in a temp folder and move it to extensionsdownloads only when the download is completed +Verified on ER. +In most cases it will work if ASM is not available at all but if an older ASM is found it doesnt work properly +Thanks Nicholas for update +Sending a test case preventing it +Thanks a million for the great work! Much appreciated +There is a weird memory issue here when we get the image +Seems like tests would be easy enough to do since these take genericsI see the Pool Interface +I committed this +comment from Mike Matrigali on The current implementations of indexes +How about in case of using local FS or this shared NFS? Is there any mechanism to help? Sounds like were relying on the implementation of file system +Vijay emphasizes that there should be no finish +Changed and committed to CVS HEADIndex RCS filehomecvslangsrcjavaorgapachecommonslangexceptionvretrieving revision retrieving revision diff u u Aug Aug getCause get get getException getException get +It actually works for JSP with configuring enable debug module and trimSpaces Option +Olga thank for review +Alright so can we move towards a consensus on a solutionSo far I see people are okay with Moving over to an interface which exposes get only methods Creating the core implementation which uses a ctor with final fields Builder API can be created and placed in a yet to be determined placeSweet +Applied patchSending implsrcmainjavaorgapachetuscanysdohelperSending implsrctestjavaorgapachetuscanysdotestTransmitting file dataCommitted revision +In general in requires as well as Locale +Flavio can you address these concerns? Thanks +So until more bug reports related to this type of issues come in Im marking this as resolvedfixedOf course new bug reports are more than welcome and the issue can be reopened when this happens +fixed your edit problem with the strikethrough +Actually theres another path that needs to be changed +OK committed +manual is most likely not the best term to describe what a install so I am open to suggestions +nothing +If skipTrash does not work as expected this is a more serious problem +Attached is the patch with fixed test case for blocks and modification to handle a file +For the time being I will put the CMS Adapter bundle list to the full launcher +Mark as fix in. +So in this scheme there would be separate namenode datanode tasktracker and jobtracker war file since each exposes a distinct web interface right? Thats a larger change to Hadoop than simply disabling the jetty daemons since Hadoop sometimes stores application state in the servlet context etcShort of that you could simply package the datanode as a war file but permit it to start its own daemon for its web interface on a different port +will prepend the bundle id in the JMX name so they are nicely sorted by bundle id in the JMX consoles +Patch looks good +trying to run hudson again +From now on any new data types that are very specific to database vendor and not yet supported can be handled properlyPlease feel free to provide feedbackThanksLeonid +changes to to read the help from the yaml file changes to to copy the srcresource directory into the output directory after as part of of the compile target so the yaml is available +Logged In YES useridIt should be fine +Oh patch fixes +The only thing this requires is implementing the NAPTR record type +Are you sure it is commited +its just as easy to diff it ourselves +Each thread now has its own Map +Youre right +This only occurs when assertions are enabled java add node with empty name bug and test case in svn rWRT vs exception isnt that a regression of? introduced assertions in without updating the contract +I believe it is +Im confused +Before deprecating I removed the locks to verify hadoop passes build testsI will submit another patch that deprecates the inteface +Thanks stackIm just taking some test to verify my modification +I have attached the file that contains changes +AndrewCan you provide a PDF that demonstrates the error so we can verify the correctionThanks +Were we violating Apache process back thenI think the issue may be Are the release notes source or are they a build artifact? Im viewing them as a build artifact +Marcus is this something you are planning for code freeze Nov +In addition this issue is about getting the document submitted to OGC + Username clli +Its a regression Ive disabled the test and well fix it in the s probably just an issue that flags arent getting passed along correctly +with Romans +But what if Im in a heterogeneous network such that some machines have eight cores and others have two cores? The config will play a part there correct +bq +For now Ill put a note on the documentation saying that widgets with ids or no ids will load onwookie but will not be updatible at this stage +Attached a possible patch +Hi XutingCould you please take a look squareSchema is missing from +Thanks for Review! just commited that patchfully agree with Christian this is not enough to resolve that issue but at least a little part is done +Closing this since has been released +weird +Thank you all for your attention to this issue +Dup of + +will have pack names with +I even wrote a test trying the demonstrate the class wasnt thread safe but I didnt succeedIs it sufficient to make the fields volatile and leave the logic as is +Ive used jdk but its the strange because the classpath stuff works for mevenide projects themselves quite and where do you define the dependencies? I mean in pom itself its parent project file or maybe some external entity maybe? what is the path to the entityany value resolving done in the dependency definition? can you post one dependency def thats not workingthanks a lot +It turned out the bug had effect only in the combinationof preemptive proxy authentication and a redirect +Closing all resolved tickets from or older +I was expecting this problem when I submitted the patch for but didnt have a good test case to work against and verify the fix +I think it makes sense since the code is effectively duplicated in and it makes the tag implementation much simpler +Please dont hesitate to open a ticket +I think the patch is included in this war +IT added +If lock node is expired one of the Rs will take over the transferqueue process and copy the rest of hlog znode +Patch adds Occur to and matching test for +cd trunk mvn clean install f build So +short of that Im not sure +Looks good +Im attaching a patch to +When looking at methods that take a primitive was not considering that and float are equivalent even though all Java calls end up using for reflection purposes +Is it legit only in some casesAnyway if the override describe in this ticket is not legit a fix like the one from Jose Luis Martin would be much helpful +Eventually we need both +Hi +Even then I would like the behavior of that method javadoced explicitly stating what the expectation is if a cannot provide guarantees +In relation at the file this included in the last release OFBIZ TRUNK +Ported to Branchx. +Owen exposes two pieces of information Number of operations and average time taken for the operation +This Jira handles cast bag to bag tuple to tuple +Found that had references to files and +This just starting point for possible maven files +ThanksI was confused and thought I used submit patch after people had approved it +While running the portlet as web application the url is generated with the correct namespace and action +Weve handled a problem regarding in final +I believe the attached patch resolves this issue +good point may be not +Thanks +Might be related to +I have checked code and found that either its not implemented or code is removedI am not sure its a bug or an improvement I took reference from Split Ship functionality and implemented it for backend quick checkout +Close out issues. +ThanksI included ASF License in example files and added it into distrosrcmaven +I finally manage to make it works by runningmvn clean installthen running sonarsonar command again +This patch will be checked in as a java client implementation +Not seen reoccurence of this bug since related fix went in +At first the root token is ended and then subprocess is ended +I am committer on Subclipse project but not on Subversive +As far as I can see its working fine now +Final recheck +I am going to do the same with the pmd and cpd targets because we actually do not need the XSLT processing on Hudson and not running it will speed up the build +will be documented as a known issue in the SOA release note created. +Perhaps this could be a good reason to change this into a DOM parser +Looking forward to seeing how it evolves +Standards are a good thing +Ill be putting the plugin up for vote shortly. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Latest patch looks goodApplied it in revision I also added a suite to include it the test run. +Not that I know of +If so Ill assign to myself and drop the noob tag. +Yes all of this happens in the comm layer +here is a new patch adding documentation to a reference to all DV Fields +Theres little else I can do at this point +Fixed +Ah +Thanks also for testing on the branch +Fixed with +why are they in there at all? isnt your ideplugin configuration a developer choice? why would mandate that everyone use qe or the or whatever to develop against +Probably the simplest way to inject classpaths is to mod the plugin tools that create the plugin descriptors for injecting these as parameters +Its definitely a gotchastumbling point +Ive been using the formatting rules for eclipse from the project are these not correct? Is there a beautifier I could run before submission to make sure the code was consistent with the projects standards? Ive also got a small amount of documentation to add to the confluence wiki if I can get access +I consider to add this patch before Stefano has complete his work cause its a bug + deletes the test the master output file and removes the test from the suite of old tests being run in the new framework +Reopen +Is there some reason why you cant do this? IMO committers should commit their own workSure I can but you started the issue and youre still the assignee I just chimed in at some pointIll commit this in +separate eclipse tests and contrib reviewRunning Eclipse tests +Do you mean a wiki or more comments + +I changed the code to create and pass the to the generator but at this time the is not consuming the parameter +Ok Ill rework the patch with your suggestions +I suspect the complete message is coming very late giving the coordinator chance to abort the transaction without hitting the bug in +Since they are all acquiring read lock if there are two token operations going on against a same token the order of operations and the order of edit logging can be different. +Tests committed revision . +Are we supposed to guess the details about thisWhat is the name of your methodDoes this class also follow the method naming requirement to satisfy the specification +I thought it was important to execute the cur incrementdecrement after the logic was executed + +I will submit a patch to fix it +Done +I still thought we were talking about doing my original idea on this issue +Thank you +Resubmitting the patch to hudsontnx +Alejandro Is there a list of jiras regarding distributionbuild that we need to keep track to backport to once everything stabilizes on trunk? It might make be easier to create a uber task and mark these jiras as sub tasks to keep track of what jiras are needed for mavenizationdist fixes in +just drop it I will replace that stuff with on disk sort tomorrow +Batik automatically determines how to execute the scripts but I guess there is no simple way to do that +Removed an unwarranted assert in with revision This is a manual backport of revision . +Use of a was introduced with at revision on Feb and in for the pdfbox and releases hence for tika in revision on Mar and revision on Sep +While I agree we should prevent unnecessary VM exits we cant prevent them all and fortunately the VM is designed to terminate and restart quickly +At revision Ive fixed up just jmsBytes to see if this is what we want +Its not possible to instantiate abstract classes currently with Aegis +The file is uploaded at homekamrulPlease let me know if you need anything else +No time right now but maybe you could add tests to +Verified +If you have local modifications Id suggest you to check in more often +Juca can you investigate on hibernates dependency on beanshellIs it just from an example standpoint +Expect documentation update in the next few days +This is by design returning false signfies that he operation should be but then false should be interpreted the same way in beforInsert and beforeUpdate +I have removed the code mentioned above while removing the disabled national character code +Verified on EAP for all settings +I know its confusing well try to fix that +Updated the patch to have the container and temp directory use the same permissions as the appId directory +before we emitI could see it going away if we emitted Key and comparator was s Key comparator +mysql version notI couldnt find the saveUpdate you mentioned to which class does it belong toI dearly hope that this will be implemented as I do not see insert performance as unrelevant +This is occurring in IE and +close it since its fixed now +Attaching a test case which should allow you to reproduce the issue +properties prefix never got removed +Eg if youve got a cluster with all hosts running and hosts have failed local dirs if you restart the cluster these hosts will not come up unless you first unconfigure the failed directory from these hosts +I have the same issue from times to time +On the other hand this could be considered a bug fix and Groovy does have a history of slipping in some new bits as bug fixes +It doesnt appear that using the SNAPSHOT as stated above fixes the issue +Heres a patch that fixes both issues contributed by Mike Rheinheimer and myself +No response assuming ok. +And karma must be granted for helix dev folks +I have started adding TODO Fix in the code each time there is a potential pb related to this issue +Typo here In running examples we should write that the JMS examples are starting a small standalone server that is suitable to run the examples only and not supported for production +snapshots present and abort startup if the proper cleaner is not installed with a log message saying that this must be doneWe do not want to even allow an admin accidentally change something that could result in data loss +This I still see should not be done via NULL check +Btw Im not sure how you generated the patch but was a new file but the patch shows as it was applied over existing file +Patch for util lib +You can still provide custom type detection functionality by injecting an implementation of the interface +As for replacing the template you did the right thing +Closing this. +Can I suggest to move the discussion about whether connectiojns shoul dbe closedautomatically to a new bug? This way we could release the improvement as itstands now and potentially help users find problems with their code +Wow +failure is not caused by this patch +attach here seems fine +Its better than nothing and we can iterate on it as people try it out +takes a larger view for archives +We should check this doesnt break any tck rev branches rev Leaving open until we check tck results +The suggested wording looks clear and fine to me +Thanks a lotNow I optain the Locale but I must find a way to optain a reference to my struts resource messages +Heres the diff for I didnt provide an ant task to run the test +It appears randomly while browsing the sonar site +DimitrisYou are right and I believe the same as you about this issue but the person from Sun Microsystems replied to my bug report with +for the patch +No TouchyMaybe someone will eventually fulfill my dreams and add me to the list of developers so I can officially own an issue +It seems to me like there might be two alternate solutions Kurts technique of compressing only the first rows onthe page in a single run requiring potentialy multiple runs tocompress all the rows Adjusting the array size dynamically so that it doesnt havea hard limit of but rather can be sized appropriately forhowever many rows are on the pageIs there any reason to prefer one solution over the other +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Checking with Bryan to see if he removed it for a specific reason or if this was just an oversight +Thus remote access web console will not have the option to open the GUI at all +Update The workaround to call nodetool compact only works when there is at least some activity causing there to be more than sstable +The attached file includes all of the book files for the specification on format including an empty master file and a template file +wtf +Applied patch into master and branches. +Theres no reason we couldnt enforce compatibility on the query parser so that it behaves the same on both platformsIn my opinion this is a bug that needs to be fixed possibly in both Java Lucene and +Fixed in rev Thanks! +I need to further look in to the simple type handling code as I tested only two scenarios +When the configuration is written back the directivesget lost at least with the current implementationMaybe this concept would provide useful when applied to andits definition files? Here it could be possible to include completeconfiguration sources based on conditional expressions +This updates the lists of votes releases and committers +Id like to see us discard anything beyond the first options in the JT just to keep bugs from the central server +IDEA might have screwed up the svn update +JDBC has fixed the problem +Added another patch in which I tried to utilizes all the common code into one base class +Needs to be fixed +Just checked and it works fine in +Can you add a test case? Take a look at the +But whats the advantage of that? Wont that be the same as using the zip file installYou already know the problem is the bit native launcher +If it had one we could check that into petstoreWebweb and it would just be another of the apps pages + +Please please approve my patch +Look for your new metrics messings +Thanks! +We do accept contributions +looks fine by Boris. +I would suggest logging it to debug by default +this can be annoying +Yeah sounds like a better place for the startcode +Hiranya you should be able to commit now with your apache id +Great! Glad we could help +If it is hardcoded in the sources a fragment or something else is needed +I will issue another JIRA to address the documentation issue relate to this JIRA +Thanks +I made a few minor changes but in general everything was quite good +Duplicate of it seems that created only one hour before +Its not really something that has to be in Cayenne itself though Im just looking to create options +Yeah I think so +I use under Mac OS X and dont see anything that ugly strange +Not sure whatever is going on some instructions on how to configure it with AS would certainly be helpful +attached +Needs some performance tests? My little index couldnt tell me the difference between the three types +User name and a list group names are obtained from RPC +Committed this to trunk and +Status as before +This is basically the same asJVM doesnt provide finer temporal granularity than milliseconds so we cant get a reasonable value +is there a potential risk with this patch that it conflicts with the new feature in to put in the geoserver +Ran Test Suite against Futon on Apache with patch applied +Om I think I have also fixed the issue with Android that you reportedCan you please run the test once again +Apologies it was indeed fixed +Do I need to open another JIRA and submit the same patch to that one +As I said at the start I know little about JNDI and rely on your expertise +By having a layer of indirection you get the following it is harder to figure out which configuration is pointed to by the variable BUT it is easier to change the configuration string later and have all uses updated +This has been true since? You have to be kidding! I am using and it is still broken! +The annotation painting code maps annotations in the CAS based on the type to eclipse annotations the eclipse annotations are then associated with a drawing strategy +What would be phi in the FD is just the score here and we dont care what that is just that we sort by it +So each of the NMS Providers will undergo voting starting next week with and +Cool! Please write down the github url here +to Gary since the initial exploration of Disruptor use within Spring Integration is most likely to initiate within the TCPUDP support. +So the diagram looks some thing like thisplugin MachineOrigin Serveras you can see above the version between plugin and API does not really matter and API always uses request and creates a new state machineAlso you can see that persistent connections to client or Origin Server is not based on the request from plugin to APII have the following questionsHave you tested the API with requests? How are you using the APIThanks I understand the request works for requests and underneath it transforms to request in the state machine +Clarification te fix will already be included in the EAPEWP so the only concern is with EPP based on EAP believe templates using resource is one part of the problem that Thomas have seen but that should only result in a stylesheets beingThe hard error that have been reported by QA and Thomas H +It turns out one of the implementations had faulty implementation that returned false and hence cause the unnecessary for this and keep up awesome workThanks +Yes this is the Jetty that is bundled with Solr Jetty according to the Server HTTP headerI havent investigated on the Jetty side yet it might be a known bug its time to upgrade to the latest Jetty or at least start evaluating itThat would also remove the requirement for a JDK over a JRE and speed up JSP page compilation too +Marking as fixed by +Because I was trying to solve the memory issue at the same time as the caching issue I changed the XML parsing to a streaming parser at the same time I changed the cache +looks like a good idea to idea too here! +Added option splitRows that can be turned off to return the entire content that you can convert to a List and then iterate each rows as a Map objectUpdated wiki as wellCommitted revision +Duh wrong issue! I only wish +request in railo now looks like thismultipart boundary nametestfile filename textplain charset binaryTest file nametest textplain charset bit nametest textplain charset bit +My guess is youve hit a scenario where it just doesnt work EL or regular +Detailed description of the problem Node A runs a prepare sent to nodes B and C A B and C locked +So this test is duplicate and slows down right +JonathanThanks for taking Apache build seriouslyI created with reference to a failure on JenkinsI am on patch v with the minor modification as mentioned Sep Ted +How do you do the insertions? Is that plain Hector or is there some CQL involved? do you do calls to the directly? do you use jsonsstable? do you use the to bulkload data +If the server has loaded a deployment and I copy a jar into that exploded deployment the runtime does not pick it up +This was done a while back see. +I could do this over the weekend and release +I dont see why you need to install in a? +add lowercasefilter and replace TODO more tests with some tests +Also rename the and in the package to drop the +You are right we probably shouldbe on par with what XML haveNice to have. +Reformatting a little bit +Kathy thank you for explaining this issue so clearly +Was there a patch even proposed for this issue yet +Hi SimonI am sorry it is an old format +I guess I originally thought it would be a single toggle combination rather than different combinations depending on the direction +Ok got why its not working +File is attached +Hudson core tests appear still unstable +It is better not to wait for the next half hour to replace the but better to fix it right away +Another thing that needs to be handled is database cleanup of expired sessions +yes it looks like its the same issue +SO shall we close this bug and open a new one related to this which will be to fix the sync +LDAP and Zookeeper could allow configs across clusters +lgtm as r +Now there are several modifications related to debugging messages which dont seem to be always related to the problem is this jira +But we are able to test under two different configurations only per aech CR +oops +patch looks the patch to make hudson pickup the right patch +Thats very different between Chinese and French I dont konw any French too but I think they are both the elegant language +Warning yes but dont disable the finish button! If users want to use ESB then they should be allowed as long as they get warned +Thanks +Is it possible that there is code in your app which does this? Perhaps it might be prudent to put in some logs in this case +We already have a lot of analysis performed on our system and they all misses the version in the components +However for this maybe we have to revisit table ownership rights +By the way how do you do insert the text results in the comment? +I guess this is a classpath problem +Should an abstract class be publicstable +Another important design aspect is that do we want to have separate zookeeper instances for each core in a multicore environmentNo +Heres a patch fixing the basic issue + patch looks good +This was a consequence of the changes presented by +has module refactoring and should fix any issue you might think exists here +I will port these changes to branch if no one has any real objections and will close this issue as well +Without the distance all member vectors are treated equally and this is often not a good thing to doFor some clustering methods the pdf may be used to recalculate the distance to centroid but this method is not general as I recall +Thanks +Jitendra is there no way to do something with ant akin to what is done on trunk with maven? That is dont run some specific tests unless +Manually tested +Please let me know if you come across any more anyone tell me how to build the with Ntlm support? Command line or kindly send me the jar at xiaobof? Ive checked out the project dated at Jan +This introduced a new interface that allows the pluggable API to retrieve the name of a file +The problem is more that we need to change the MOP to support that +He wanted to reuse log files +Ill see what we can do to fix this problem +I agree +Please attach a minimal sample input document minimal stylesheet and an expected output documentThere is no way for anyone to understand what you are expecting much less debug any problems if you dont supply enough information +Thanks +In any case Id like to integrate Rahuls changes so that we can commit this break the stuff into its own subcomponent finish and release +pathtoapachecmsbuild pathtomarmottasvnsitetrunk pathtomarmottasvnsitetrunktargetCorrectly generates the site in local +So since there were no objections from folks in charge of I am commiting this to and branches +Resolving issue since the immediately relevant documentation topic has been fixedWhen is fixed another doc issue can be filed to document the new size limit +Small to and trunk +Youre correct Tom +I solved it by adding the following code to each line that reported the errorinsertablefalse updatablefalsehope this helps closing stale resolved issues +Please find the file attached to this JIRA issue +Added a new test case with revision which shows that updating the statistics will make a query pickup better index compare to prior to statistics availability +My concern however is that this is changing the behavior of an interface that has been around a bit longer within Derby and I would like to understand if anyone feels this will have a significant impact on users +Unless someone objects I will commit this patch later today to fix the build of the eclipse plugin and perhaps even fix the plugin +I added it so that I could call it for the libcloud driver but I havent updated the docs yetIll let you know when I get everything ready +My primary point is that this issue got really out of scope and that I was trying to find the subset of the patch that would address the issue of shuffle should use serialization to get comparator and the follow on jira that discusses the much wider ranging discussion on changing the type and serialization system +updated key features +some functions are missing at all while others have wrong names etc +Here is a patch. +On first thought I think we should be able to handle it +I dont think thats a good ideaHow about moving the stuff to itself as static methodscreate creates an object representing a client for the given WSDLcreate creates an object for the service described by the given WSDLIf we do this should go away too I believe +Nothing to be done +Prepare for Jenkins patch testing +I think terms dict accomplished the same thing. +Sorry for the inconvenienceI found that it happens when is not empty +I just checked and still see that commit notifications are being sent to +Fixed in Rev +I think this is by far the more important work to be done and I would love to see it in if posible +Seems like an ASM version mismatch to me? Chris have you ever run into this? Otherwise Id suspect that there is a different ASM version on the classpath of that specific Ant installation getting picked up instead of our own +No idea why I debugged the code a bit but without much luck +is no more part of Xerces code +Looking at compress package I realize that hadoop compressors are rather complicated beasts +The only way you can achieve this is to leave it as now and for the programmer to code the commit explicitly in the required place +This is client issue only +Maybe +Or is it the case that we will be directly using the vector class from MTJ? I checked some discussion on Commons Math list about Colt +Thank you JarekI will provide a patch to create the bundle +means that join position is not being used +Address potential class loading issues +Integrated into trunk and +This patch moves the scratch buffer into the avrofilewriter instance itself were already allocating that from the heap and its easy enough to add the K buffer to the end of the type +thats probably gonna collide + +is more advanced feature that is more or less coupled with the current implementation +In previous release of HADOOP we dont have this problem because we are always reconstructing HADOOPOPTS from scratch in the invoking process +The joda patch again this time with less cruft and more passing testsFound another bug in it was using timezones like instead of GMT. +pulls the task start up code into a separate procedure so that I can make sure it doesnt throw any exceptions +Later sybase drivers should allow for text retrieval via getClob +This plugin extracts hN tag values from the document fragment in the parse filter i dont see extracting values from the DOM +Any comment is welcome +this patch reflects the new proposed structure +Im assuming youve tried it? All the code is the same as for mapping beans +older issues for Apache since were no longer actively working on these at the moment. +My wsdl contains operation with input contains header body output contains header bodyHowever in the method signature no paramter to access headerIs there any example found in Axis samples on how to handle this +Rebased and fixed a minor JMX bean after s checkin +this fix should handle the circumstance +Seems liek a timing issue to me +Net this raises an So the solution is to add an identical catch block for theThis will cause the lock obtain to simply fail and the calling methods to retry +I am good with thatI assume the matches would work for the GROUPED or UNGROUPED +No need for IT. +The sad part the sources of are not available at Maven Central +It seems to be fixed +Committed with tests. +extends and only have setting for a fixed delay +Javadoc for the test needs modification +As a user you receive a null pointer exception because of something missing in the the very least you should receive an error message indicating the problem +I have changes the file +Yup the in most certainly takes care of this +Stack Todd GaryCan you take a look at the combined patchesThanks +As far as can I see colors are pretty nice in blueSky skin +This clearly had an impact on the level of confusion in this patch +Lets investigate whether or not theres a way in Maven to prevent it from copying srcmainresources to targetclasses or alternatively investigate a way to do a move rather than a copy of targetclasses and targetclassesmime to targetclassesorgapachetika and targetclassesorgapachetikamime respectively +Id need to see a patch to verify if the change is acceptable +is boolean false means buffering if the commits +Attached a test case that has an autowire error because of this issue + agent command portnetstat anp grep collector portnetstat anp grep +So if its No then can somebody please just say so and we can close this JiraThanksWe did not vote +Awaiting your pull requests for master and the the pull requests are issued but I cant find how to change the status to Pull Request Sent +However Im not sure if the current design of FOP allows such flexibility in the choice of the algorithm since the layout engine seems to be mainly tied around the Knuth Plass implementation +Nevertheless reducing priority to minor +I will upload a patch for which actually got committed and is correct +wonder if we can get the to also work as a +Again I suspect that you might end up simply in order to determine a BXML documents validity +HiThe indenting style is not correct and it lacks a patch +In fact both RLE and partitions can +No problem +No need to reopen. +It is just a fix to some tests + As Neil points out nothing can be done at this point Struts is known for its backward compatibilityIf the constructors do not default redirect to true than in this case Struts is not backwards compatibleI have existing code that no longer works unless I manually set redirect to true +Of course that makes a lot more sense +Only the wise ones as far as I am aware +I think this would be a good service to those users who trip the hard error on upgrade it means Solr is not doing what they thought they asked it to do +But from what I remember thats how I thought I implemented it +The root cause is an issue with ZK +I think you can mark it as one +Ill invest some time on this and update this thread with the findings +Ive uploaded the second patch +Tested +Its a pretty serious regression +Duplicate closed. +JS is done also not sure about the others +Resolved in r awaiting clean regression results before closing +Using the values available in still isnt getting the client to use SOAP +Leave this task open until it is completed for all panels +Attached the jstack of the NM which failed to the value for is set to yes this is most likely a duplicate +Serge could you test if this patch makes a difference for you +I have committed the patch +Fixing bug title +the and has same the configure in the cluster +I just finished uploading Hibernate to the as also been uploaded +Again we face logically distinct use cases +It does seem like calculating the average field length at index time should be relatively cheap +If the problem persists please reopen +Sorry for the noiseMark +Patch to enable latest cvs source to be built with Also bumped up minor version number to +Patch applied with thanks +Is there also a situation when xa recovery resource is unregistered from transaction manager +trunkNow its easier as it uses the expression to evaluate a delay in millis it will waitNo more strange stuffAnd the Spring DSL does not have the delayTime attribute +List is the data structure needed for the construct of +This bug is fixed +MichaelYour patch is in trunk rev +We need to be careful how we bridge scan of +If you take a careful look at your version identifiers you will see that you used Beta +it works for me in beta. +Other Java features will require alteration to the shared project +I think the login form implemented by implements this request +The v patch is for +Fix is in s source rep +Im thinking of providing three methods one that runs all the time and takes seconds? a minute? Label it as slow either wayA nightly that runs for minutesA weekly that runs for an hourThe background here is that I see issues after hours running on my machine occasionally as little as minutes +It must be ordered in order to be deterministic +OK thanks TimIm not an expert with Illustrator but fortunatly my wife is will ask some of her time o +Could it be as simple as the two servers are participants in the same replicant and the sstables contain the same keyrowcolumn references the process of streaming repair is sending a set of keyrowcolumn references to the requestor in the same sstable as the data its already aware of via repair Compaction finds the duplicate references in the recently received sstables theyre basically the other replicants copies of that dataIt seems that the act of sending the sstables from one server to the other when both are replicants of the same key range would be expected to result in duplicate references +Patch applied in r with thanks to Jeff +In fact can you please zip up and attach the modified QS as is +So even if the server is manually configured for control a decent error message will be is also in +What the motivation is behind the request for multiple dispatchers isnt stated and so it cant be determined if the current could be modified to handle the need. +I just tried it with new workspace and it is correct +Ill post another patch soon + released issues. +Im attaching a simpler repro +is too much matter to problems +No need to depend on snapshots any think this issue could be closed now +Youre right Uwe I can do that +Thats just how Ive tested constructors in the past the thought being being that the constructor is tested in the constructor test method and the assert is just there to assert something +that copy is to protect access to the message content which is unmarshalled on demand without sync eg the message propertiesAlso transformers are per consumerIf each consumer wants to see a single readonly instance of the message doing a single unmarshal of a copy would work but would require a sync on connection dispatch for the duration +This is a regression Hadoop which does let you create in the same +IMO it is important to take the time to analyze the problems but fixing them is wasteful +This is the configuration file that ships with the ESB standalone +If there is only a single node the for loop will be executed once +Through this binding it is not possible to retrieve the Token from the +Creation date is not even a maven concept +So this becomes a backport of? That patch is nice and simple +Any chance of this getting in before so the few of us on windows dont have to think about it anymore +Applied against trunk +Fixed in in r on trunk and r on branch +Is a better one +Do you have a simple sample app that can reproduce this +If my form is sending parameters shouldnt they be in the parameter map regardless of whether or not I have a and regardless of where I have the in the parameter list +I fixed up one javadoc error and added a new comment in one bit of code that needed some clarification +you Stefan +Copied in serverdefault there is a better solution for that Ive created a forum thread +Patch to remove the Ajax table loading per previous comment and fix the manual tests +next +Im using Jencks InboundOutbound JMS configuration for consuming producing messages and quite often theres a minute delay in message delivery after the producer has replied +defer all issues to +Empty files deleted now. +The DNS issue has been resolved this issue can be closed +Is it possible to use Struts tags with that resultRegardsm guessing youd have a hard time using Struts tags in a GXP page because the GXP compiler would probably barf on them +yes jmx needs to be available for deployment outside EAPI believe its a mistake that the defaults for EAP was changed to use external locationBut you can fix it by using internal deployment or enable jmx on the server + +Ivan Andrew thanks! The patch was applied at the revision verify it was applied as expected +dont have vmware setup please pickup whoever has this one up +I guess Scala builder doesnt provide such a model for Scala source files +I just committed this +The simple idea is to add +JB Will the plugin be released before +However Amar Kamat explained that will help solve this issue +Whoops. +The constant is a percentage while estimate should be for size +I fixed the title +it may be a good idea to change the schema so that user can mark input parameters as optional +Martin Ive created. +thanks +to commit +Ive reproduced this with client mode +Committed in revision . +Then Ive got some time to clean this up and commit it +I just committed this +sql ways in junit but it is currently the best way to execute the language constructs that are specific to ijYes we have in the charter to use only standard SQL but there are some extensions that are specific to Derby and they have to be tested alsoLanguage clauses as get cursor with hold next update where current ofSo its important to keep a test that verifies appropriate behavior when such things are used in ij +I have added a cancelable progress monitor + +After copying the ejb jboss and web modules from plugin dir and refreshing Xdoclet data in preferences everything worked +I can now build a sandesha patch that checks for this namespace when deciding on whether or not the axis operation is user defined +Not yet decided to go with singleton but on close watch +Attached patch includes tweaked for +Found an additional case of leakage fixed now on trunk +I think thats probably a promising path if we can work out why it fails some of our execution will not be in so Im marking it for +Will commit if tests pass +Fixed in izpack with commit edcdcafdecedfdbee +That copyright stuff wasnt supposed to be in there +Sorry didnt notice earlier question +Which should we dealing with this problem by logj setting or source code level fix? We can add a configuration to limit msg of is no longer in the dataset if we fix it at source code level +Was trying to wait for a good one +Bug seems to have disappeared in version of the Feeds plugin +If the user understands Pig data model he should be able to strip the group field off himself +Do you know if supports issue is still open? Would it help if we use our support account to report the issue? Im looking forward to using Spring on WLS but managed threads are in advanceMartin Paoletta +Currently hsync and hflush are the same +I am ok with this approach +Suggested patch that subclasses Springs class +Patch to remove the creation of the symlink and renamed the directory proxyapiinclude to proxyapi the patch worked after renaming the directory proxyapiinclude to proxyapits +attaching proposed solutionThis is not a real patch since it currently breaks some testsThis patch would also address the needing on in the close for the release +regionOnline is already used in many places even in the access control coprocessor +but we need to rephrase such a wishDoes anyone have pointer to Xerces support for something like catalogsSorry to have been a bit rudepaul +Also I cleaned up my imports as per the cassandra code style wiki Committed on trunk for onlyThanks +Still todo Testing +Brian has started looking at it but Ill also take a look +I think that recommended configuration shouldnt cite LZO specifically but rather make a general recommendation to use compression with a link to the appendixIt used to be that LZO was the only game in town but with SNAPPY thats not the caseThe LZO info that is in recommended configuration should be moved to the LZO section in the AppendixIll rename this ticket with this change if youre cool with it +Updated the patch for review +crt go into etchttpd which wont exist for a zip installation +Committed into trunk with revision +Im not convinced its a problem + +Tried it and it works! Thx ll look into s security issue and clarify whats going on +Build configuration includes Main class name Parcel Additional C sources include dirs and flags Additional libraries and linker flags Clownfish include directories Other files to install in ClownfishincludeOpen questions How to name the build module? ClownfishCFCBuild? Do we want provide an equivalent to LucyBuild? Probably yes +Source code and of the first submission +IvanWhy? and how about the absolute path case +Looks like I cant modify the priority howeverAnyway has a build dep on in its pom but doesnt actually use +The lack of support of osgi in the indy jar is related to the bnd tool which isnt out yet in a JDK compatible version +I could but its users is kotik andJose Peleteiro +The conversion to transport is not occurring properly +Not sure why it is called initialAccount +Update to fixes this bug +Apply the patch from to have the functionality working in editor windows using the keyboard shortcut for the Maven Run command +The patch does not fix +patch applied +Is it to intended to be committed or experimental? I can review changes in trunk from the last time that I rebased +Let us know and we can reopen if required +Attached file is having snapshot of jdbc connection parameters +Looks like you have a broken Icy zip file +Thank you Ilya! I applied your patch in trunkResolved +Using binhdfs namenode is starting but i found some exception related to class resolution +I will copy this to the branch soon +AFAIK as any other jackrabbit component are unaware of workspace deletion yet +bq + +I should be clearer the bug was introduced in the release preserved through and finally squashed by the new replicator on trunk +Your analysis of the problem was correct nicely found +Reject PUT if the incoming data is older than the local return code Conflict should be sent back in this api needs to be changed to throw version conflict node should send back version conflict exception if conflict is node should generate status code when receive any rejection within the cleaning after months +Trying to reproduce this problem and could not +From the more complex mockup it seems like this definition of vertical spacing would be nearly sufficient +Each of these would be constructed with a String do you think +Just +I had read the template documentPlease notic only error message in exception in page can be shown correctlyI think this is a minor bug +Made you a contributor Liyin and assigned you the issue +What model is this dialog reading from? It definitly does not look like it is using the JDTs the first tab does not allow me just to write Company i have to write second tab doesnt look like it is finding any of the projects tab doesnt list any recent about just using the default class type selection dialog for now until the above issues are fixed +No it doesnt need to be rebasedCommitted to trunk +I took the liberty of applying the patch as suggested by the reporter +Ill continue the search +Also attached a full eclipseProject which contains all jars code to reproduce the problem +Also fixed a few format consistency problems in these topicsOnce I commit this patch the Reference Manual will have some sections where all topics have shortdesc elements and some sections where no topics have them +Is there any progress yet on this issue? This is still not fixed in version +Thank you for the help but we went with the Java annotation and APT solution for the type safety and being the farthest along +Compressed error log +The attached adds the evolving annotation to the new interfaces and +You need at least two region server failures for this to happen +Thanks for the reminder +Len yes workaround is to have doctype in the file or simply just not use the palette +Does this error only show when using a to declare the security domain? Ive tried starting the server with a deployed application that declares the security domain using the annotation and everything ran just fine +The redundancy is a good convention since it makes it obvious which directory builds which jar file +And the only place where this eror is raised is exactly that place +I will try this running on Ubuntu at home and see what the result is tonight as well +Ill upload a separate patch once all the comments are resolved for this current patch +Please ask questions on the user mailing list +no youre right +Then we can sort out what I did wrong here. +o Rollback changeso you can set encoding and doctype in constructor +Patch attached +Password is s only a guess but it seems that adobe somehow changes the encryption algo +bq +Thanks Tim. +Please review +Attaching +Yes that IP is blocked for attempting to SSH to servers you have no access to +If something isnt clear let me know and Ill be happy toassistThanks for your response on thisBob +See +But by configuring a time between which sstables are compacted you dont control very well how much sstable may accumulate during that time +Suggested fix increases the by +Closing due to release of. +attached a proposed patch which registers the original resource filter into actual resource dependency +Paul can you cherry pick the fixes in the related into ER build +I see the logs being split and then ROOT assigned +Otherwise great work thank you +I am happy to take on that responsibility for now +Fixed in subversion repository with the series of revisions r r r r r r r +This could be done more efficiently +However I dont care too much about the exact packages for the generated protocol buffer code and glue code so comments are welcomeIf desired I should be able to split the patch into two pieces and +The issue is gone since +Marking as since this patch has been applied to branch and is not yet released +This patch is ready for review +Applied the test and patch at r +Cannot reproduce in +This patch fixes the issue with gaining new generics information +The obvious class choices would beThat is where I would begin +Ive addressed this for and also for and all of which initialized in the init callback while actually being dependent on the as well +This issue is considered RESOLVED with the production config +Is there anyway you can try this using the versions of APR +initial version of just committed this +I didnt see the ability to attach source files to it so didnt include any yet +Reassigning to Rick. +I added a little code to clean up the compiled files in that case +Is the part about specifying usersgroups and allowing access only to them similar to which was fixed for Hadoop +We could choose an arbitrary static amount and that would work as long as its the same for all hosts +Patch in last comment is pushed. +Ive got a couple of little changes to make to the docs so Ill handle this along with them. +Best regards +I agree that ignoring blank lines is probably better than returning empty recordsHowever RFC does not allow blank linesSo it is wrong to equate DEFAULT with RFC +I have fixed a few bugs in the original where we were not registering with the RM correctly +Compile error now fix +I had a similar problem when running Eclipse for the first couple of times +When it comes to formatting Teiid Designer still relies on Teiids Language Object framework which includes specific formatting of SQLDesigner did add hooks to preserve formatting of String values within string constants defined in SQL +patch for both and trunk +Will have to do some more discovery +And when I think of it perhaps the title of this bug report is somewhat misleading +close it. +The patch didnt apply cleanly to trunk could you upload a new patch please +Ah yep sorry Im missing something +Working on other bits for now code to if you need it sooner let us know. +Applied the patch +Ideally also with recent versions of Groovy and Eclipse. +Regarding Jeromes comments If the stylesheet is in another projects jar file then you could specify this jar file as an extension for the the Ill consider that +Added missing files +Tests passed cleanly for me on +Nikita not sure if you ended up applying this code change somewhere in your environment but your suggested fix is incomplete because you need to deal with the default region and get the appropiate default cache for it +Committed with changes for +META +The rest is just internal details +Next time could you do a jstack before killing pig process and attach it here for the recordA bit confused +Yes this is available within trunk and will be in the release +So do you think in addition to increasing the accept queue size it makes sense to have a heartbeat here? That is if a server doesnt receive a heartbeat from a client and the is about to expire it schedules a heartbeat to the client and probably invokes a GETSTATUS or some such method on the client and if that method returns a valid response it keeps the client alive for another interval and this goes on +Client is sending a multipartsigned entity which containts an encrypted piece +properties for that reason require a default valueadded check in validatemethodfixed in resolved issue +Patch to lucene trunk to apply field level boosts before query terms are selected in +How does uncache work if the path is the only identifier? Needs javadoc Again here is where we may need to add an identifier for cache ID +This patch contains two changes +if youre using standard text fields it makes absolute sense to use the maxlength attribute +You only see the issue on clients and hbck in those versions cant fix it anyway +Global EL Preferences Page screenshot is attached +but it still seems to have the same problem in uDig +why cant you use for this +If you want wrapping text in a tooltip you can simply use a as the tooltips content +I will post a new patch on Tuesday +There is already a method used to skip white space and this method should be updated to skip comments as well +The purpose of is to say that if an alert is attached to a metric this metric should be shown as green orange or red as threshold have been defined +mailing list change done +Yes all charts with generated labels +synchronized all access of out and ignore lines in the log file with parsing errors +Please close if satisfied +It runs fine on my windows But I agree it should be fixed since there will undoubtedly be users trying this and a NPE at this early stage will send the wrong signal +we can retrieve all available suffix names we can retrieve all configuration parameters for a suffix we can have a default fallback strategy if a parameter with a specific suffix does not exist +Test case and patch to resolve this issue +Unfortunately this patch doesnt apply cleanly to trunk +So now we have only the layer from layer issue and we are doneThanks! +Doing the memstore current patch the max size of the memstore slice is specified in terms of heap size +Each Query would consist of filter AND rule filterEven if we can just apply this for the case where their is only one rule life would be much much better +Taking a look +Implemented this change in Rev +Selection functionality was restored. +Wow a float bands tiff file? First time I see oneWhat about the SLD you are using +the patch looks good to me +projectDisabling Seam support fixes that issueIll verify JBDS from +I cannot fix this unless I have access to a Win box +Therefore retargeting to the release +Could you please add a test to and attach the patch to this issueThanks +Andrei Any progress on this +In ganymede it seems that all the adapters downloaded and installed with Download additional server adapters link cant be uninstalled with P +I would but I am so overloaded with other work right now +Why cant v simply be a new serviceI think we need a deeper for this +Assign for for for trunk +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Please review and apply supplied patchMany thanks great to me patch applied with thanks! +Yes exactly this is for situations where one side of the join is too big to fit in memory +Manik I tried some ways this weekend but it looks like following the same strategy of is quite hard and errorprone to reuse the parents code it is a requirement that the reentrancy key is of type Thread which is not true in Infinispans case so much code needs to be written to reimplement the queue of waiting objects +Resubmitting for review. +Im going to try to squeeze this into +Thanks Andrei! +This is the patch Im applying +This patch will conflict with the last patch in +i cancel my patch and provide a volatile declaration for has no negative impact in field declaration AFAIK +t to +The following components documentation has been added so far CM Deploy Tool Prefs RDBMS SecurityNext Search Statistics +Not sure +If you want to proceed please just the issue and copy in the data as per Gavins request. +Eek +It seems to suggest there is no reverse DNS record at all for the you are sending mail from which is not true alt least not for +Let me know if it works for you. +New patch for specific to +For example if you are replicating to a database you might send a single update for a document representing several revisions worth of change +Please let me know if there is a general interest in a component capable of this +Any expert can help on initial feedback +in new xhtml wizard listed only templates with new xhtml context +heres the TRACE log anyway can also confirm that this is fixed with the RSVP config change. +I committed it to trunk +entry sets the tablename of the to the new Meta table +Dumb question time from meI dont understand the importance of the user +Layer style colors vs +bulk close of all resolved issues. +pt in the surrounding foblock seems to be neededto trigger the infinite has been added with description a short of infinite page has been added with description a tall image +I still dont understand the naming +The scheduler can keep track of required resources to avoid recomputation of resources when the AM is restarted +Obviously this makes it hard to return the value in a queryIve edited the description to remove the false error +Thanks Chris! +I actually fixed this last night +I can add some tests that fail suitably +Thanks for the fantastic base release note textD +NN reports both as duplicates one from the client and one from the DN report above +It does not seem to be a highly sought feature at the moment +hmm +For Java backend fixed by problem still remains in C backend so I changed the component to C backend +If we move version outside the seam runtime it is previous version with path to seam and version number +is the ultimate cause but youd better watch as it will be resolved only when Ill have tests for this specifically or find an alternative solutionOtherwise its fine to use a just not enable async +with Lars suggestions +Thank you +Ok Ill buy the argument that it creates an inheritance conundrum +Verified on ER. +Closing thanks for looking at this task and fixing the code +to commit the third patch as is and follow up later with a proper test case +Can you repro with vnodes and hsha +closing +Thanks +Hi GeorgeI see changes made to at Dec +Could you please try? I could not reproduce the problem with that tried with and the packages were added to the knowledge base with no exceptions +I hope this time everything is ok and the patch can be appliedThanks +If is run from trunk it would use compile and use all bits from the build +Created by mistake +Do the Hyperic folks see a problem with this? This would mean a data processor node would be executing control actions Im not sure if data processing nodes have the ability to execute control actions if not then we cannot use the Quartz clustering feature +I agree document it as it is +Best practice these days is to take contributions without modifications and make the changes in the incubator to allow full transparency +are we expecting users to JAR up their own copy of DIH with its config inside of its jarI hope not +Templating preferences is a feature of the container +Committed to trunk +Same behavior still were not parsing identifiers properly +This will require a new deploy of the site. +Alejandro we have already iterated through the create new container and merge with old container approach and saw no advantage over a cleaner api that asks for increase in an existing containers request +Defer all unresolved issues from to +since changing ledger directories and journal directory might cause data loss and data inconsistency I dont think we should handle in bookkeeper level +Here a patch with the allocation problems resolved +Certainly the current integration exercises for which were hoping to use SI involve dozens of use cases like this ALL of which will poll for files in a given directory based on filename underlying reason for this experience I think is that the application generating the files is inevitably some legacy app whose only way of communicating with the outside world is by writing files to a directory +Hopefully this is fixed +adds some helper methods and support for schema set and fixes the test +User can specify the time duration to be skipped by specifying commandline argument +Will the check command return true for usersabc because usersabc is already created? If so we need to add an END marker +The approach here will be determined after making some changes to the role of vs +Close this. +Initial patch attached for review Im no autoconf expert so this might not be that elegant +Ignore the +I was mainly asked to return stats on nodes in the form of map data structureI dont understand how this is different than the JMX data +Closing this for now +A couple of comments there was a recent thread on the list discussing a motion to reduce dependencies on external libraries and instead to rely on the XML parser +Javadoc warning was because of +Failure looks unrelated +The fix is available in the release and will also be present in the version that we havent released yetIf its still happening when we release please reopen the issue +Thanks Adrian +It also makes the point that it is not the same thing to allocate an Object outside the loop versus inside +I moved to andcreated and +Thanks +Thanks to Shakeh Khudikyan for the contribution! +Single biggest issue seemed to be creating a VM in a constructor +Type changed +You can try the next snapshot or apply the aforementioned workaround on the current one +Trivial there is also in utils which is a thin wrapper that simply provides a constructor for uncertain which approach is better. +It is an error from our side +Should be fixed in trunk added a similar test to the test suite +Unfortunately Ive been swamped at work +attaching the jvms heapdump snapshot and javacore files for one process +I updated my example it is now a full example for the client invocation to reproduce the problem +Doing three splits per server in could impinge on a serving RS but would cut down the WAL split component of MTTR the longest pole +In terms of the book structure I will be reverting this to what is was as Svetlana Max and I have an agreement that in moving forward all Tools and Developer Studio books must be able to be built with both Mavn and Publican +I will continue my digging but I think it just gets deployed first and changes state from UNKNOWN to INCREMENTAL and then first EAR deployment happens +Ill try to add a special check to working around that JDK issue and giving a better error +Patch for this bug bases on version of class while addressing +What we have found is that its true that a java client using the Java api will not get the duplicated messages because the api will remove themWhat happens is if you start the attached BEFORE the message broker is started and then start the broker then the client will receive messages correctlyIf you then restart the the client it self will not receive duplicates because of the Java api handling this but if you look under connections in the admin web interface youll see all messages is sent twiceIf you now restart the again youll see that all messages is dispatched times and if you restart again all messages wil be dispatched times and so on +I just committed this +Opened to track changes for +So lets disable thisTravis the way we disable the test is via EXCLUDEMAJORVERSION macro +Chris can you attach some sample projects to help us troubleshoot the problem +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Committed to SVN trunkOleg +Tiny comments You could inline those b b etc variables for a bit more performance +Good idea I will test on xp. +Can you please let me know either way +Why do we have at all? Would anything break if we just made the response processing asynchronous +After checking on the dev list it appears that there was some inconsistant documentation which implied that this should work +JuergenThanks for the expose workaround works like a charm one thing s changelogs do state that the default was changed to true but the code doesnt seem to have changed +Yep seems to work +Leothe it to be backwards compatibleI mean it should just work if there are NO java dependencieson old facelets +Trunk updated as part of. +Ive forgotten that Spring still supports JDK and this doesnt work in +Other thing to do to ameliorate memory usage when small cells is to up the indexing interval from default to or etc +Bulk defer all frontend issues to +FYI David to test this ticket Ive directly updated some rows in the projectmeasures table. + you Martin. +However Im never keen to just take the option which provides a fix for one use case especially if it involves exposing framework internals which have not previously been public +It keeps throwing a instead of draining the result provide a patch and test cases +Problem mentioned above seems to be a problem in the embedded driver and not just a client problem +backported fix to branch +invalidate comes +If so then might make sense reintroducing a closing flag on regions and then do not return region if its already passed into closing stateWill check these regions out on regionservers to verify this is indeed the case +There can you check if The changes have being made properly +Refreshed the patch for trunk +If the default is Closed then my change is except +If you a redeploying then the is already getting kicked and there is nothing that needs to be done +They may not be in completed state rightRegards Uma thanks for reviewing the patch +Rob can you review this pleaseThanks looks good +And the question is Is it expected behavior +I will consider this patch but I wonder what motivates such change +Isuppose this would have to be a servlet parameter +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +The issue is not in the console but related to the change to undertow +Patch all tests pass + push out to +Ive seen the last update and it is ok +But where is this decision being taken and what can be done to speed things up considerably? If theres a need Id hop onto IRC +Hi NicolasWhen I added this patch I did not notice the and files were missing +Maybe it is not related to this patch +Ok the case can be closed just leave it in. +Updating Fix Versions and resolving +Modified to allow the similar script names as for GET requests with the difference that the request method must be part of the name +I will try to hunt down a similar build machine and figure out whats going on +But as data is None in this case the fix for does not apply here +Added a separate DISCLAIMER file for disclaimer. +There is a certain amount of inconvenience in maintaining multiple IDE build projects which are themselves redundant to the scripts +This may conflict with +Update +If you do a full build this should be picked up by Maven and the dist build would add it to the lib folder +Logs from server A and B +I would like to see that as well +Oops I labeled the first attachment it is actually the diff +Thanks for answering the unasked question. +I didnt put those in before because I thought they were optional +If any of these updates are done in EAP and merged into please update the fix version +Accommodate spaces in paths +Its not something just. +I decided to bundle Sigar with as part of our library source code instead of waiting for Sigar team to publish binary to maven +I attached an example project with the described behaviour When executing mvn siterun and clicking on the goals menu link the error creating the project I found out that the problem only occurs when the is added in the reporting section +Done in the parser the server entities. +I will have look at this issue and come up with a patch +Setting fix version to look at these in Beta if resource leaks even have an effect +for the patch +Archetype related work is now tracked by. +All the tasks pass again but any review is welcome +Andrew can you review please? Thanks +Added compareTo so JPA only does a distinct on the PK and not the entire object +Perhaps the Concurrency sections added to and CREATE SEQUENCE statement should be called Performance instead since preallocating a range also reduces the overhead significantly in the case +Sorry I was busy with other things lately let me rebaseresume everything +Some enhancements have been made to runConfigure in more resent releases +the changes are orthogonal its just that I want to get the service stuff done before I spend any of my time worrying about what we can do to make configuration easier +Maybe +This is starting to sound like an issue I reported some time back namely which hasnt received much attention in recent monthsThere is nothing in the JMS specification to indicate that one should use separate connections for producers and consumers and doing so can have various implications for the design of an application so I would agree with Rainer Klute that this is something that should probably be looked at carefully +Yes thats what that would mean +smile +Committed in rev +gz but just incase I have posted the file which should be unjared +Doug This one seems to have dropped off your radar +Solr handler statistics were already published in JMX this just added some entries + Doesnt this one depend on jQuery? Yes its using jQuery but I think Click should use jQuery instead of Prototype where possible +I am using some version +thanks Richard +Sumit please review the implementation plan and close if issue is ready for it +Great work everybody +Your changes have been made + +and now inherit from the abstract which has a as a property +If you have a working solution for I am willing to remove the file right now +Possible fix +Of course guys I meant that this would be an optional feature +I fixed up this test a while ago and made it use escape sequences as suggested +How is this different from the Separation of concerns +Not marking since it wont apply on trunk and Hudson would fail it +svn ciAdding downloadTransmitting file dataCommitted revision +Ah I see what the issue is will try a workaround and let you know +bulk defer of featurestaskswishes from to +Looks good +We got to test with both Linux and Windows before applying the patch +Included source patch apply against of kernel module +If you update to svn this patch should compile +excellent +This problem is not fixed by adding the missing as mentioned by EnricoAs Andreas noted there is somehow a difference between calling Maven build in the kres directory or using the top level POM +We could turn this into an itest with some work I guess but for now this is just what Ive been playing with +This seems completely unnecessary +The tests are disabled at this point because they run out of memory when run with limited heap because of streaming of the LOB columns whether they are needed or not +Caching of classes is the job of the class loader +Committed to +Fixed with instance is shutdown the global config should be ignored and snapshot can be taken regardless of the hypervisor typeWhen instance is running and false you shouldnt be able to take KVM snapshot +Looks like the easiest fix would be to move the into CLR this is the approach taken already inv attached w that approach +Closing +I would like to take a look at that spec Spring version of the specification to this report +Fix for the maven site goal is already in svn +I hope these queries arent at the limits of what CQL has to offer. +I misunderstood your previous comment and I thought that there are still problems with exceptions left +I plan to address them very soon +This is fixed with revision +Start with empty CF and then write into it using batches of Keys in threads +See comments in issue log above. +So lets fix the issue for you tell me an idea for the date of the release of this version +AllI will be checking in a first cut at the LDAP implementation based on the code provided by Mike Long +make a Groovy class that extends from Throwable or Exception and try to run it on JDK and it will get the constructorsruntime error +All done +Fixed in does not suffer this bug. +Canceling patch +Any volunteers +Please review the code and provide some feedback if you so desire +Stuart seems to be handling this one +Please confirm it is applied as expected +its been noted that this isnt always the case depending on environment +Thanks Ted for the patch +Otherwise for every superstep we would be building and tearing down objects +Im resolving this now because there has been no further feedback +Since removeObject call is expensive I expect mass removal to be slow as well +its just as wrong to me to agree to a LGPL license bundle which it isntBut since Red Hat EULA is the only place GPL is listed then Im fine changing it +Can you please kill this list but retain the archivesThanks a bunch +I also would like to do the test but had no luck getting access to the tck +That is the patch Kasun providedApplied the patch and thanks Kasun for the contribution +Thanks to Jacopo Anil and Sumit Pandit to discuss prior to implementation +attaching a patch which I believe adds these features +So the problem is really with wagon the x bit gets lost when zipping up the site +The patch dated Jan is the solution implemented as discussed with mfojtikNoteTwo tests still failcimi and cimiI had talked with about this and it seems that thispatch is good and she is going to look into the tests +SPRINGCLEANING We can reopen if necessary +Mark is this one done +I think its better if the solution is discussed in the dev forum where people can comment more easily +Or youll have to go after the MR job author for trying to access stuff that hes not supposed to +Basically its working just like except that it checks timestamp not value +Svn Revision fix to for this issue. +Ok this is probably the last time Ill post on this issue +No unit tests should be affected and any tests for this functionality will be far more brittle than the patch itself +This is great +Integrated. +The parser doesnt know the semantics of the mydepend URI scheme +Fixed the functionality that I was looking for +Ive updated the script hopefully it is now +Thanks a lot for the contribution Vinay. +for the patch +s inv +Currently trunk is targeted toI wonder if we could or should fix this in and put new artifacts on our eclipse plugin update site +Look at screenshot +We will take a look at it and commit +Is only trunk affected? I checked back to the code and this problem is there +I cannot reproduce this on my Solaris machine +This definitely wont make the series were primarily fixing bugs there nowWrt the JPA case apps dont have access to the so doubt that even such API would be useful +last +I love! We do that in a number of places and should simply call it instead +The snippet I quoted was from the patch +Fixed a while ago +Instead of fighting with a libtool hack I just fetched libexpat +Ill beast it for a while +If a timeout has occurred the code should and throw an exception +new test case shows an issue with multiple concurrent durable subs +I observed Container does not exist +Thanks Bill! +No assumptions about the type of data in this field can be made without first understanding the technical requirements associated with the Web serviceThe hostingRedirector is a deprecated element since its functionality is now covered by the accessPoint +So there shouldnt be other factors in play that I can think of +Hi AllIs there any solution for this issueThanks +However the problem I have is the provider not being available when an exchange handler is processing an incoming need to be able to determine whether the target of the exchange is a binding +Rob what is this bug about +Its just for generating markupBut now in this hyrbid world other things could happen yesSo if a form is made for a get +Needs to filter by alias deviceType to be supported in java sender s add support for variantID querying as well. +A zip file containing the tests cases +Jody the KML output subsystem is not using the streaming renderer and the label cache at all +not necessary for standalone but I certainly had to install them on my AS in order to get this QS running +add tests for CDI International bundles too +ship the attachedOnly consideration is that we might want to ping the dev list since its a change in functionality pretty late into the series +Till now I had no Idea why this occurs thanks for the hint with the long field +Any chance you can prepare a patch for trunk +Will work on merging them into codeline next +Feel free to close this one +Hi allIf Id like to fail with an exception instead of logging No resources to read can it be done todayThank you! +there was a small bug in addPartition that i realized while testing that has also been fixed +Thanks for testing that +We still need end methods +I think the test just needs to be fixed separately i think it extends because it has one of those things +Usually the directory structure needs to be changed +See spreadsheet attached to parent task +size means precision and scale was wrong and this fix was not necesary +Now to back port the fix +Ive added and a disclaimer intoThough Sobol sequence generator has been adapted from j and thus it is under +Unicode characters are parsed before getting to the different String representations +Assigning to the maintenance lead +If you all are interested Id be more than happy to package this and submit it to fedora and EPEL +my service bean is called eisService and the class name is +Bump to +Diagram the diagram and updated the docs +Looping no longer occasionally fails for me +If that practice is followed then after a while the code becomes unreadable +Luis thanks for the review +Fixed with upgrade in +Verified that all operations on synchronized collections are synchronized on the returned objectUpdated the patch to reflect that +My guess this is because of formating issues I see a lot of Ms in there +If theres any improvement needed it should be made in IMORegardsFreeman +I just committed this. +Tested the latest showcase snapshot on IE and FF and I am unable to reproduce this behavior +I suggest we postpone this until Hadoop as it fixes some new API issues that would prevent us from moving entirely to new +Reverting this again +So I went through it and found the issue +For examplecd mojo diff will get both files +lists table as an option but does cannot decipher it. +Thanks to Todd +Ill have no time for this until Friday because I need to finish some tasksAttached new logs with trace logging for netty +So all issues not directly assigned to an AS release are being closed +Im convinced theres a clock problem here. +Are you testing the latest code Min fixed the issue in commit fcbbcdbbfcdbde which is checked into at Date Tue Aug try with latest code Edison and I recently checked in two commits to address this null issue +I think that the solution is just create a and a with your logic and register in the converters mapHow I assign this bug to me +Rather all we need to do to implement this is to count bytes and files as files are listed in the client before the job is created +Ive committed something that eliminates this requirement rev Im happy to vote against any proposed releases that do require defining JAVAHOMEI still think most of the script crud is not appropriate since it is for starting gshell NOT geronimo +I havent looked at the code yet and I dont know javascript very well but the user suspects that when the file descriptor info is not found it causes the code that puts the JVM memory info together to be skipped +Updated the names of the testcases according to Mauros comment +The current object is not set if it is nullFixed in OGNL +What would making it larger help +or use a repeater where you have total control Great! We finally got to the point is subclass of whose on is producing warning +No news on this issue so I mark it as Cannot Reproduce +So it might be nice if the startup throws an error indicating whats wrong instead of the +Yeah resolving as a duplicate Ill solve this in +For some reason the slice unit tests do not pick up the parameter when its passed in as a system arg +The attached patch should address the NPE but further testing should be done against this for issue out as part of JIRA cleanup +bulk move from M to on this in local workspace against remote sources briefly broke the uberbuild but I think were back in biz to remove a couple tests which arent currently in the uberbuild either +Committed patch a to trunk with revision Ill await at least one successful test run in the nightly regression test before considering committing patch a +Issue is fixed withSending modulessecuritysrcjavaorgapachegeronimosecurityrealmprovidersSending modulestomcatsrcjavaorgapachegeronimotomcatAdding modulestomcatsrcjavaorgapachegeronimotomcatrealmSending modulestomcatsrcjavaorgapachegeronimotomcatrealmSending modulestomcatsrctestorgapachegeronimotomcatTransmitting file dataCommitted revision This disables a test in +However with the change in block reports are always sent from volumeMap +build a tree of projects +Jimmy has already been looking at this issue +I think we should use the wiki as a scratchboard for documentation +Updated the test so that the async case is more stable on fast hardware was a bit flaky on my fast Linux box +I use a stock distribution with default settings and bundled data +I often observed the memory consumption in the reduce phase of Reducers go up to heap limit and fall down repeatly +I can use it in Hudsons build without committing it to the repository +Apache Onami is suitable +Actually I removed the DOCTYPE declaration alltogether because each time the descriptors where loaded the DTD is accessed which slows down the test cases considerablyFixed in Rev +Probably we should make this a in some future release +Its probably related though if I exchange the for a constant it works fine +therefore i dont care too much + +Thanks RegisPatch applied to LUNI module at repo revision rPlease check they were applied as you expected +The pull request was accepted +Yup thanks +Both have acquired more testing as part of this but more tests would not hurtThere is not a lot of common code that can be abstracted out of them as the extensions tend to be very specific to the IDE +This patch should take care of the problem +so you really dont need a can simply have a transformer checking with Repository and if it return null have transformer return the original message which is a Claim Check +Resolved by allowing the usage of PROPERTYNAME just like in a WFS all issues that have been in resolved state for more than one month without further comments to closed status +Doesnt always using SASL result in an extra roundtrip at connection startup? Seems way too expensive to be worth the elegance +We grouped them under the Web UI component so you can filter on that to see the related tasks +Hey Jukka AWESOME! Yep totally makes senseThanksCheers Chris +are you gonna add random data generator +So providing another patch which fixes both the issuesShould I open another issue or this issue can be opened and the patch applied +My patch should apply cleanly to trunk +Seem Ive missed somethingTake it from the source +Committed as r. +Nearly always a JIRA issue so taking Subversion off the component list +you are using a new feature in H not something that was available in HRead the hibernate docs which refer to the exact semantics + my first patch +This also makes it easier to run Continuum without unpacking the war +The JIRA didnt show any for the changes so I couldnt tell what was changedThanks +to slash and burn +Commited to branch +I thought I only need to attach a new patch file to trigger the patch might need to cancel the patch and submit the patch again to trigger the build +Same as +Are u using it on the broker side +I confirm it worked fine with Hibernate closing stale resolved problem didnt resolve? +or something like retry with the current +Ill probably go back after +Attachment Picture has been added with description screenshot showing rendered showing rendered image beside the original +I dont think thats possibleprocrun can run both and detached services and this would limit the usage because it would require users applications to report OK to the daemon +Thanks for that Matthew +I have a flight coming up Ill see if I can give it a +Patch for trunk +Test was using get from HRS to get list of online regions but this is unreliable since it is keyed by regionsize +Unless you are planning to add this to the karaf build can you try setting up your project using the kar packaging +hrm well were not seeing the map file already exists message but we are seeing too many open files which kills the tablet server +namits join results is in the resultsvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalvalI will add a automatic check query in the test and upload a new one +Damjans suggestion seems to make more sense here +Running full tests nowPlease review +Ive committed this patch to trunk and +user interface some renaming and closed +Heres a script of queries that use the alltypesorc table that is checked in to the hive tests +Menu widget refactor +checkout any of the Apache Directory projects +If you have examples where this is not the case can you attach the choreography and let me know which roles is producing the incorrect we should be able to close this one +pending finalization of the release note text that should go with this change +Closing JIRA. +The has been edited also to download the into the piggybankjavalib directory it only does so if the is not present so the download is done once +Closing. +Id say its pretty useless if it cant be set by the user +I too would be in favor of adding the total excluding tax just before the grand totalHowever before we commit it I think that we should wait for Sis feedback even if this could delay the resolution of this patch theres no hurry because in the meantime the persons that need this patch can easily apply it to their local environments +I havent seen this before +RightMy above comment means that supporting more precise types is required to get the exact resultsThus this is just a temporal solution to get more precise results +bq +The link in the description points to the EAP build +Almost all of the tests are now run twice once with the standard and once with the newThanks to all for the bug fixes and contributions + +Just use directly or as a factory inside the Web Service defined in +Ill post the patch here and it will be against trunk at the time +Integrated to trunkThanks for the patch Xing +Please close this issue +Clearly the wait job cannot start until the workspace job in finishes but the workspace job in is now waiting for wait to begin +This is a manifestation of the same bug as that one is fixed this will be fixed as well +Every RPC involves io so always throwing some might be reasonableDoes that make any sense? What is the bug were trying to fix here? +Thanks for your help +That is the reason I want tobe able to use the same service name between the upgradesAbout the benefit of having an up versioning strategy for the servicesdeveloped using Axis well I would think it is an essential requirementfor any commercial product developed with Axis +Is the grant statement the same in as it is in? I dont want to go changing the docs with information only relevant to +Ill create a separate issue once I understand whats causing the weird effects I misinterpreted in this fix version should be changed to but it doesnt exist yet in jira +Is everything up? +Martin did this +Committed a patch +eclipse error log after failed attempt to install the groovy plugin +Attaching patch HarmeetYour patch is in trunk at r +See also +Actually just a second I do not have your Feature Selection tool on udig where is that from? Is it a tutorial or something? +Can confirm that this problem still exists using the latest CVS from +Please try again +this was already done a long time ago so close this jira. +Modified Quick Search and added Advanced Search updated screen shot of search a new screen shot for Advanced Search +META +Ill have an updated version later today +The comment from Noel make me thinks this was invalid or too latePlease commentreopen if still present. +Looks like Nikhil took care of this onethanks +Committed to and +I already started to work on int up to unit tests + Username rchristy +Patch added +I stole the current code for it from where Im optionally dropping the update log on index merge +The unit test fails after I apply this patch +Hi Fay your patch fixed the bug +Closing the bug based on Kishans comment +Works as expected +I believe that the patch for that you applied fixes +Ive changed the target package to depend on rather than so that the package will contain the test jar and the build process will not fail +It should come to them automatically +This race condition gets exposed when hivehistory is disabled because when hive history is enabled it attempts to create the same dir this way but on failure it just logs a warning +I will programing a solution by using derby for my Internet page +Not a deadlock +Its just a configure issue not a defect of jms transport +This patch hopes to fix thisby extracting the retry logic to a new interface +That looks right +Assuming lazy consensus will commit tomorrow. +I guess nobody ever tries those so I never HPUX installers force the user to install on Oracle +The following command line works finejruby rjava ropenssl e class Foo include Java endCan you come up with a test case we can use to reproduce this? No guarantees that its something well fix but it would be good to know what exactly triggers it +is fully modularized. +classpath +Anyway Ive now applied your last patch thanks again for your contribution! +I also ran derbylang suite and it passed without any problem +Debugging will be much simplified +foo resolve to foo when on and barfoo when on bar in all browsers +If theres anyone up let me know if this looks ok +Code checked into so marking it resolved for the release notesIf more work is required please open a new issue and link it to this one. +Stack sorry for the gratuitous changes Ive seen them but I didnt know how to get rid of them +Ubuntu Win xChromeFF think I corrected the problem +has similar goal with this JIRA +Note that DN is the dead datanode hereCurrently I think it will take seconds for the cluster to recover if it accepts write traffic across all region serversVarun + +Thanks for Kevan and David J for helping to commit this. +The user can simply refactor and delete what is not required +I will be careful with the tab next time +This should be fixed +And there is not a great way to sync them what happens if you recreate the datastore in geoserver after the namespace has been changed +There have been a lot of MVEL fixes can you please check this against. + +The config string is still being used in the test +Looks good! Ill commit shortly +I dowloaded the latest build and the problem seems to be fixed +Changing a datastructure from to means that applications can less safely cache values from it +Latest rev looks great +com bits from current customers +I didnt take the patch verbatim +will occur since it will put a null value into Hashtable +However theres some cleanup wed still like to do on the before we use it as a template for a +Fixed on trunk revision +Work done +In this way it can not be displayed if ADS is not on line but this feature can be evolved on a next release + +adc has vetoed applying this to the m branch so thats all foks. +Thanks Robert for your comments I replaced documents with individual documents and added a link to the protobuf docsCommitted trunk r branch r +If we use the connection from the then we dont have to worry about extracting the native connection since that is always an Oracle connection +I dont think that this fix is absolutely correcti absolutely agree +The last commit solved this issue +This is not a valid path on Windows so I dont think I can adapt the integration test to suit +thanks +Replaced mailing list with an alias to dev +Is it worth it +hmm +I believe the plan is updating to pig which also supports booleans +I was wondering if we are planning to implement javabin in any other +I didnt see them locally +Provide an example please +This really should be applied at a global level not just at the eclipse level +I just committed this +update patch for latest do you think about creating Hama RC after commit this patch so that user can easily download and useThen we can collect more feedback from the community +This didnt show before since Doxia didnt created the externalLink class +Add some more assertion about internal state of after unassign while SPLITTING in place +The user database would be serialized at the front of the imageRequiring a file with a username id mapping that admins need to manage is brittle +bulk move of all unresolved issues from to +There is no way at this time to split by package in most placesHaving it be easier in a few places would be nice however +The relaton is there because according to Tom the foreach can be implemented under the hood in the same manner as the group activity +The jdbc tests run cleanly modulo the timestamp formatting cruft +eclipse mem usage diagram during VPE tests without patch +Anyway my question is wouldnt this be better handled by a decorated factory instead of pluggable strategies which will only make things more complicated? After all this behaviour is very much a configuration thing +for me too +Adding compiler option true solves the problem +A change along these lines would need to be more akin with the way filters invoke the next filter in a chain and that would have to be a new mechanism not the existing contract which we cannot change anyway + +The sources have been fixed on branch so Im done here +I personally dont think the uses thing should matter as long as the rest of the manifest is fine +Obviously there are a number of ways to implement this +Im just saying the place to fix it is inside +Will provide a patch later +And a jms namespace and a theme namespace +Thanks for asking Im planning to do it +If I didnt find this page on Google I would have given up +Commited change to ensure a check for the binding is done before creating a new one +Raghu did you verify that this bug still exists on the trunk of piggybank? I thought this was fixed +These semi working test cases will also need to be refactored to make use of the feature andor osgi life cycle +It has been on my TODO list for a long time but I couldnt find time to look into it deeply +Sorry for putting it in the same patch +I have enclosed a patch and new object let me know if it fits what your afterThanks! That was easier than I thoughtI wonder if there would be any speed benefits to communicating to that no one needs stack information so dont bother collecting any +in unix jvm options are in +Thanks. +New snapshots available with a fix for this a bunch for letting me know +all performance optimizations will be resposibility of the Identity Module Implementation itself +In this case httplocalhostaxisservices will work properly +It keeps track of durable subs in demand suscription and unregister local durable when theres no remote durable subs any more +unfortunately including some stages +The code that seems to be failing is new for so there would seem to be some kind of unforseen race condition or similar +Guys you need to enable Include dependencies and source folders from modules option on the Maven Dependencies container either like Aaron described in his comment above or using popup menu on a project Maven Enable Nested Modules +defer all issues to +I wont be able to verify this anytime soon. +please point to specific files in svn that you have concerns aboutI recreated this file myself from clearly attributed sources from scratch +Bulk close for Solr +I would not recommend switching back in the version numbers +Broken how? If you could send in a that shows this problem that would be your other points to have a POM without group makes sense for Continuum but I think even Maven requires a groupId +My understanding is that the statement pool will use this even to drop the statement from the pool and there wont be any question of the invalid statement being used again +Simply ant deploy it to a running ESB server go to httplocalhostcontract then click on the contract JSP wsdl link and also the wsdl link +Thanks for patience Anoop +Solved thanksAwesome! +If a test case cant verify a fix in a short time its IMHO better to just skip it +Extra check was needed to prevent Martijn Ill have a look at your changes asap. +I also filed for increasing the default timeout valueYou may not need to file since both changes of this and are quite minor +This encapsulation also has other advantages you can detect duplicate use of the same id and test code can ensure that all message ids declared are actually testedI also am defining the error code from network client exceptions to be more in line with the error codes coming from the server +this only happens at their environment +Unescape would look the same +Since Rick was concerned about missing closed tags I just generated a new dita file with the code he checked in to see if it still works +to see a couple of lines of my code in a project like this would be awesome for me +As already discussed elsewhere looks good other than not upgrading the store version which has now been handled viaThis needs to be requested for inclusion in so that people upgrading dont find their broker starts behaving quite differently +Ill fix it +But happened because there was a bug in the code +Here are files to harness patch is split into two for reviewing purposes +Where in the test do you find that weve lost the third editSo we name the file when we write it for its first edit then when we move it into place we rename it to be by last edit in the file? Add a comment to that effect Id say else could be confusing +good point gert but i think the opposite is the case this disable parallel then eachparallelloopParallel does not the execute each iteration in a parallel thread +Addition to for property coterminalNew corresponding test in +Thanks for reporting this! Fixed in revision +Will do +Added handling for server to client streaming the caveat is that this creates temp files on the client side +The latter will then receive a wrapped which should behave properly in terms of parameter needs a reference to the bean in the root but at least thats concealed within that extra filter then outside of itself +Is there a way we could divorce names in Cassandra from names on the filesystem? It just seems like an odd implementation specific thingMaybe marking up special characters in a standard wayIm not sure why they are ties together in the first place thoughNo matter what though validation and error messages are needed I was just curious +Thats strange +I request the patch I attached yesterday to be considered for review and comments and if everything is found alright a commit alsothanxNarayanan +Looks good and works for me +Or is there a large overhead on even just creating the scanner? If the hdfs level chown enhancement is implemented wouldnt you need to change the method signature which would make hbase dependent on hadoop deploymentsThe bulk enhancement I am proposing is used for more than just chown +Verified in JBDS h linux +the optional clauses +Yes I do have the credentials to call the webservice with the authentication serviceI am trying to understand exactly how to use the nightly build +Right thats what I meant +Thanks +Were just holding the flows in our heads instead of visualizing them +bq +To answer Albertos question At this point I would replace the usage of this function in favour of the Win API +Please dont move slider to Develop client or above for Client if you only generate web service from a java class you should get a right result +The aspect in which my differed is the to field of the lines +Im attaching the trivial patch to do thatbq +Question about the class and having GCO etc +Looks good +It works here too on Safari +I agree the issue is duplicate +blah is generic and disabling virtual binding for this close of old resolved issues. +Do you have any screenshot and LDIF export of that entryAn LDIF export of the Root DSE might also helpThanks +Thanks fixed in CVS please test +You may want to notify the author +update design document +wouldnt it be better not to need thread pools? Right now we are using them to merely wait on sync that underneath asyncmultiplexing +Hard for something not yet implemented to be a bug In fact as I have said in every Beta release announcement support for entity graphs is the last piece of JPA still being implemented and the reason is Beta and not CR +I was hoping to export to pdf format from there +My concern with stuff like this is that it becomes too easy to destabilize the core components and pages which can lead to an application that not only fails but is unable to usefully report its own errors +The trash routine is only called by so changing the output to a log message makes the output very noisy. +Im for switching to and to logj by default +ve verified that the new windows build runs from a path with a space in it +Ill test it with the new +patch looks good +Is there no possibility to switch to this optimization automatically +Attaching latest patch after incorporating Toms commentsThanksMayank +Now fixed and attached again +Built in brew as therell be +bq +It was caused by the fact that rollback is called by the TM on a stale a resource that it retrieved from a cache that has been stopped in between +Maybe add an are you sure prompt or something? Probably not a big dealLets fix the help for enabledisable and then well get this committed +Bulk close resolved issues +I just tried this locally and its populating right data for me +bq +I just committed this to branch +Already in. +Actually AxisC finds the operation from one of the following dispatching +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +I think this is a new bugApart from this the described bug is still there +Join L with S +HiIve upped the VM to MB for now see how you get on with that the way this is to do with the creation of theThese are now doneYou have Do you need the Self Certs creating or are you ok with thoseDo you need the Virtual Hosts configuring or are you ok with thoseCheers +Correction reproducible under Safari FF too +We should have used in when accessing +The patch does not affect the concurrency of sequence generatorsSo I say to the patch +So wondering how to reproduce your issue +We probably dont have +Please make the above changes and see if you can dig any further +This seems to be related to the different config sets +I have verified that this issue is fixed by the changes submitted to fix +Closed +How does that sound +Wont fix +It is better not to reuse for logging in the test +stacksplit metrics are now persistent across metrics reporting periods and also the minmaxavg time of the split operation are stored +I think the premise here was to provide a simple way without making too many changes to allow users to have there libraries in the classpath +The codes used by EPSG codes used by data generated by ESRI compatible ORACLE datum need to format the three attached files in a tabular format into a single file +Passed all unit tests locallyI have committed this +This is fixed now +Ok but Ive never seen this exception before +Well +That is not a good idea IMOI understand youre probably having issues with this since you went so far as to file a bug and write a patch but I dont think this is a good idea to apply to Ivy +refactored test data set up adjusted test for reversed edge direction added new graph resembling the commit dependencies in test to test cycle detection reintroduced non optional property removed in last commit added two optional relationships to and +FYI JPA final is available for these pointers Shelly. +The following unit test shows the problem although as indicated the issue has broader consequences +Misty Ill take this one +I actually tried to not use some of the parts you use like not setting the interfaces but didnt work +With r r r r and r all files underperllib and perlbuildlib were adapted for ASL as well +Supporting this slow down the application then the regular method returns null when the key exists or the value is null +These are serialized and deserialized just like named records so its nice to be able to share that codeWe might ban unnamed but continue to use them to implement parameters or define an unnamed superclass of thats used for parametersThe latter would probably be substantially more complicated than the former and force us to expand the public interface so my preference is for the former +The patch seems good +A noloop may cause an activation to never be created so terminal node must check for this and commited to branch in revisions and will be included in release you for reporting and let us know in case of any other problem +had to leave on weekend +Checking this in +PaulMy point exactly +I cannot reproduce this at all I will need an export of your repository I think +LGTM +Copied the method from to +currently provides no mechanism for deleting these files so you potentially need a set of OS dependant executables to clean up +targetcxxcompileinclude +And the rd one is a minor issue of loading Seam Persistence extensions +OuchMay be the special tests are wrong We have to figure out what happens exactlyHowever its not a big deal +Thanks Vikram! +This is not a problemwhen running the Flume processes behind a firewall but if some of the agents are forwarddeployed outside of the firewall surrounding a cluster having the content streamed withoutencryption will be a problem +patch is ready for reviewcommit +Ok +Ok the problem that was exposed was related to an experimental feature in xerces that I was using +should import classthe one above should clearly said should NOT import +Merged to the branch in revision +ntm the existing cassandra data model would not be modified +So this is really just a bad error message masking the true error +Unfortunately I think that multiple mapping files is a bettersolution that one monolithic one +Tomaz Great! do let me know if I can be of any other help +For this one were you thinking about want to build Tajo specific storage API or leverage existing library like Apache Gora +Can someone delete this attachment and I will +Great thank you for the commit HarshMitch thank you for providing the detailed bug report + for patch +So since tests pass should this be committed? Or should something be updated +update numSnapshots after deletion +Attached mgmt server log file and cloud DB +I hadnt noticed the incorrect override even after dozens of checks +Could you provide a link +By the way my comment above assumes that the Jencks solution provides pooling of connection session producerconsumer etc +Done closing. +Marking this bug as Wont fix since there are no plans to make any more releases in the. +Rolled over to to get a stable release issues that arent going to make it into to was already done +mrepositoryorggrailsgrails sey mkpasswd Nov Documents and Settingssey +JamesI agree that DB is better than Mock +Comitted to trunk poses no risk for if there is still time to squeeze it in +Fixed for upcoming release +Not sure if it ever was there were LONG types earlier now obsoleteUse CLOB for text with required capacity above the byte capacity limit of VARCHARUsers may need to configure the db character set to at the outset to avoid specific definitions on columns to support international text +I would vote yes +Thanks for looking into this +So is there consensus on letting coproc to subclass responses on r Ill get on it +Ben it looks like select rights are missing on the semaphore table +Yeah but APR uses branches branches etc +fix unit test failure of committed to trunk. +The number of maps is still bounded by a limit imposed on the size of the HTTP request that the servlet can handle +The new implementation does not suffer from this problem so I think we can close this issue. +Now I have added a rule that if the root element has attribute xmlns the file will not be recognized as an ant build file +What others feel +Patch removes javadoc warnings +I find that the problem is introduced by the lazyseek +bulk defer of unresolved bugs to +Rebooted Xen without first removing from management server or disabling it +trivial patch which checks for first being not Tommasofixed with revision . +the patch looks good to me +Actually the MIDP profile is of no help for implementing a JME version of it only defines classes for building simple on a mobiledeviceWhat you are looking for is an implementation working with the CLDC coreenvironnement thats a trimmed down version of the JDKI tried to compile with the CLCD API Ill attach the result +Also there the fact that we create a large number of topics and queues on each run that are not deleted yet since we dont yet have that feature +I would actually favor a solution which would provide a Mime Resolver +Replacing the existing list based fact entry with tree view +Let us know if you face any other applied to trunk. +Have to try another way +This bug was nasty unfortunately we didnt have a test that could tickle itSo in this test after the reload if you add any further documents they wont show up in any searches even after commitSo the test adds docs ensures they are in the index then adds more docs and ensures docs are in the index without the fix it would find rather than until you restart the server +Resolved in r . +The LDAP service usesthe root context for all operations so and not relative names can be used for searchrequestsNow Luke is using the embedded provider and his issue is with this behavior +Then rebuilt the search indexThen when I looked at the RSS I could see the weblog path OK but on the nd hit it went back to the bad URLI wonder if this has to do with caching +Marking this as Incomplete due to lack of aColm +Sounds reasonable to me Adalberto +fixed a bug that the launch time and finish time were getting confused with each made a few cosmetic changes based on a patch file reflects the small changes suggestedNone of them rises to the level of a major change +Closing all resolved tickets from or older +It will require some investigations + to trunk +This issue has been resolved through revision +Im ccing Simone the developer that created png color quantization +I think we should leave it in as some people prefer to change this via JMX +The spec only wants defaults used on read not write. +If its maintained by Yonik at ASF and sucessfully used in Solr why not release the version we currently have in maven and use it? If Yonik thinks its not ready for a release we should not use itSimilar to what Dawid did it took him a few hours to make the Carrot stuff available via Maven repo +The userpass used is purposefully bogus and is designed to pop up the auth box +An minimal example of the bug +Maybe its encoded? Thanks for the help +Reproduced on a machine from beakerSince I cant test on any other machine than those from beaker I cant confirm whether its specific to beaker environment or to RHEL +I hope I have not missed anything +All my other comments regarding and withstanding I am fine with an enum for the +It worked well for a few weeks so not sure that it is related +Our apologies you should now see the attachment +Im guessing Noble doesnt have a lot of free time for Solr these days based on how much he has popped up recentlyIm headed to Germany for a while but Id be happy to look at this issue as soon as I get a chance +There is a new module called +If you want to have minimal effect then add just byte to live bytes count for every delete +Your remark about the second click being ignored is not true +I am wondering if I should keep the lookup for a Executor from the registry in a spearate method or include it in the normal get methods + +If no one objects Ill commit tomorrow +v is v wo the change on patch passed all tests when run locally for me +Largely these cases are rare and the impact of using weak references in the cache is extremely high +This is no longer used its only referenced from the test is a class that should help Tools to do some sort of Deployment view during testing but it was never really followed up on further +Perhaps we should put a comment under related to the sortinginheritance issue and close this issue +Thats too badWell consider this issue as a request for increased documentation of what each Query considers to be its children then +patch looks good +For each DRER release we label a specific run as the reference build so this one will be built using the same commitId as the actual build in Mead +This patch creates a new additional constructor with nothing but the required fields if and only if there are both optional and fields in the struct +Please check if the following labels are still translated correctly after they are merged intoThanks and merged some Common labels from toIf some framework committer could review it I can made some more cleaning on labelsThanks patch for moving some Common labels from to commited in all Facility labels from to commited into rev +Committed to trunk. +Verified all fixes on live docs +I added a specwbuildrsvn ci CHANGELOG spec m fix for merge not according to docSending CHANGELOGSending specpackagingTransmitting file dataCommitted revision . +IT added +Reopening to change affects +Committers it would be great if we could commit this patch. +Fixed the warning reported by. +Had an offline discussion with Alan webhcat is a forcing function as import and export are only available in webhcat and not in +The trunk patch depends on HimanshuCan you run the tests listed Jun +Thanks Stefan! +Would it be possible to get this patch committed to the x branch +Please David do not create any JIRA ticket before having first discussed the issue on the User Sonar maling listI cant manage to reproduce this error locally +This issue is fixed by the fix for RegardsSven as described by Sven in. +although I can understand that those running secure Hadoop might want this as a default they still have the choice and this can be set by default in the configuration if its important +Works with +Added the unit test and disabled parts of it which fails +cancelling patch to use svn mv for history on the files +Go ahead and do your moves +mvel versions are in ER +Currently there is some duplication unncessary classes in the hierarchy that needs cleanup +I would vote for tabling putting it on the site until we can call get together and come up with an initial redesign +Please advise +Heres the patch +It is +and point the the existent zip +This could be solved by replacing synonym sets with a unique token but the number of such sets can grow exponentially so its trading one beast for another not worth the effort I think. +Waiting on Marius input for the package structure section +The question to ask is does the code work if yes is the answer then making all these trivial changes is just pulling time away from implementing the current enhancement under development +Simple testcase that show the problemDeploy it into a then open httplocalhoststruts +Heres the files I have under dataTotal size is you try running nodetool h localhost scrub on that node? Im not sure if youre running into data corruption but I remember having a similar problem before running scrub maybe its something else tho but worth a try I suppose +I today gave more finer solution to this issue + +Trying again on +Also as a measure we store the message count received along with superstep progress +It seems that an event model is missing for notification of font substitutionCan you provide a small sample file that displays the problem +Hi AndreaI enclosed the SLD in the original report +since this involves changing the API im setting it to to next milestone release +Didnt see this exception in Hudson server any more +Thanks for your help with all of this Fitz +Oops sorry about the println +Updates a number of license headers to the current committed in rev +Thanks for the patchHowever I think by definition the enhancement flagging of the bug means thatit will be put back until after RC +Jukka thanks for having a look +Minor It is an empty string for datanode could be changed to It is empty if the datanode is the client +Committed to trunk with revision and to with revision +But the step will so there isnt a big deal specifying it unnecessarily +Can we +This issue was about a specific thing that was fixed some time back and well lose track of comments here since it is closed +Its unfortunate if you dont know what data types youve inserted but inflicting the kind of locking scanning validation that other databases do for schema changes on everyone because of this possibility is optimizing for the wrong scenario. +marking it resolved. +The typical response expected by Synapse in such a scenario is a SOAP fault +The code has not been removed for compatibility reasons +I am not passing in a null +Patch Applied +Im assuming that it has been fixed in either tapestry or ognl since you originally reported it or your versions are different than mine somehow +If you think this is ok Im fine with reusing the Exec andSounds greatbq +bq +I think this can be implemented as separate collector and then together with the executed in the second search +I dont understand what makes you think this is a bug in FOP? The problem appears to be in your servlet environment +This involves undesirable manual parsing of the URL +Stores now in + +it works for me. +Patch applied in r +It doesnt seem like any of the conditional logic gets processed in the helper +NTP into the requirements section of the Gettting Started documentation +NicolasSee for the cause of +To elaborate is part of application logic not data model +Pool doesnt provide this functionality directlyThe detection of leaked objects should be simpler with Pool since all objects are tracked and contains all the necessary timing and state information +I love the debug trace above +Resolving this issue + work! It is an amazing widget! However I found a small issue the suggestion list shows up Donald Ball only if the user press the uppper case D and not if the user press a lower case d +Nits line seems like a mod line has changed the formatting other than that +fixed stale resolved issues +has been marked as a duplicate of this bug +Kindly add Information to explain why this bug is not a problem +verified by am about which is actually not a patch but a test case however it also submitted under the terms of the Apache License v and the terms of the ICLA. +Still not ready for review no tests included +Id suggest we try to tackle this shortly before these new keys are widely to IPTC to alias properties and removed IPTC from Metadata +If I have a Gant binding how can I set the set the verbosity level? It relates to your comment abovebq +There are no such command lines requests for those jobs +attach a patch to delay ledger directories creation +Try hiding and as I understand you package both in your close this issue +I cannot reproduce this anymore with r +I think I will commit the patch with the long running time for now +you mean youre considering to send the queue before the transaction commitsWhatever you mean I agree it should be am actually interested in querying the not yet committed data inside a given transaction. +Document for release and will be a straight move from existing website they are not viewable until the move +Fixed please verify +Should be fixed on Master +I think this is a good pattern +Im not entirely certain on how either of those might be accomplished +Committed awaiting wiki changes +I have factored out a doUnregister method however with an argument only +bq +Increasing the junit java heap size from m to m +This is kind of a PITA if we have to do this ourselves. +patch looks goodTest failures are expected and are addressed by +The failure I get on reads as followsFAIL got unexpected exception Heuristic commitrollback not supported errorCode calling forget on a prepared transactionIt is throwing an error code +Should be a minor edit +I have removed the creation of the module definition in the project file in my patch for +Please consider committing the fix to also by all the same reasons + +Ive attached storatge files +The only use case would be that the file had been overwritten with exactly the same amount of data +I think this patch may be conflicting with the generation +Pass the specified list of action request parameters to the render phase by putting them into the action response object +not worth a project of its think there is a project at Codehaus with only class +patch for schema directories update jonas Patchthanksdims +await love from apache infra folks to revive our git +If you are looking for users to ask for something that does not exist like this it will not happen +Committed to trunk and. +Gnu sed Possibly +HTML file included for review +I am very strongly in that camp having been converted some time ago and I for one am not interested in making changes to support chaining actions +but what patch? There are no files attached nor Fix Version +Dear Markthis is a very annoying issue when you think this issue will be fixed? Theres some workaround to apply until I wait the issue will be fixed +I am also in favour of making maxScore private with public accessors +It seems that the current one is most efficientIn addition I made a change to +ramYes I have also considered that all the entries in the store file is deleted and we dont write any new store fileBut could we generate one empty store file with its meta data alone? Let me do a try first +There is no reason whatsoever for the test layout to be this bug is fixed is there anyway of extending the archetype to include the missing directories or are we better off writing our own one from there any work around for this issue? +use profile +Oops +Thanks Prasanth +Thanks Mamta +Why did you have to change maven plugin and why there is something specific to meclipse is required +Also smart patch is getting a bit complicated with its option parsing +for patch +I see you rebased the patch thanks a lot +What version of the broker are you using +On iPhone you can scroll this list with finger scroll +It adds complexity though +So all issues not directly assigned to an AS release are being closed +it provides an attacker a means to try and turn on credential logging +DERBYHOMEbinsethomechaaseglassfishbin +Looks good to me +Started work on this issue +Rename to reflect the library upgrade +I would add that if I have the time or maybe someone else can. +Work in progress patch +To implement this correctly renaming is neededI have discussed this problem with Vijay when I visited Watson last year and agreed that the Java backend adopts the same approach as Java +batch transition to closed remaining resolved bugs +Ive been using with and a switch on sourceType between +i would suspect that this is the reason +I have posted an updated patch for this issue +Hi RobIll get a test case and some more details to you shortly +Yes this is duplicate of +if we keep punting it off to later never happening +Any application which reproduces this? I dont get that issue on wildfly as or eap +Thank you for the commentI changed the source code so now unit tests should pass even if native code is not available +My suspicion is now that we are somehow trying to use the same http connection for all of the Solr connections and that is leading to this issue +I havent updated the in the docs section +I think this patch can be commited and the task can be closedFor the other things we can open a new task because the describes function are not related to this bug +I just committed this +Sure needs to be removed from scanners +Thanks! +Thx +Imported into CVS +Patch applied thanks Kristen! +The Javascript should the coloring as it sorts +Deletion information could be elided since anything we show will be undeletedI no longer think we want raw data for cluster dumpload if youre through json for that Youre Doing It WrongSo for this ticket I think the minimum useful feature set is to support dumpload of raw sstables +Thanks Aaron for reviewing the patchAnswers to your comments +Waiting on migration of fixed docs from staging to prod +You can support your indexed properties by simply returning an Enumeration to Castor during marshalling and providing an add method during unmarshalling +Just curious why the labels were removedWe are trying to push jBPM web designer as a tool of choice for a big project +There is no way to process parallel requests for a single page without locking +Uploading patch with minor edits to add comments about dependency on +Now that work is underway with AS all previous community releases are +Applied to both trunk and trunk +Sorry I thought it would be easier to just review the proposed changes but now I see that this section needs some reworkingtgssetupjavanvir This page needs a complete overhaul and should probably be merged with the content from tgs and the number item from cgsusingtoolsutils +I removed the gem update system line from the Windows instructions +is it keeping up with the core Cordova project. +T Since we deviate from the spec no one knows the status and when the feature is done + bit bit on trunk thanks to Simone the suggestion worked out beautifully +Ive committed this +So I am assuming it will work even for the mapred system directories childrenSorry I take that back +Hi yong I have checked with the latest AxisC svn and Axis and for me both SOAP and faults are working fine +Could you please leave a comment here with a short piece of code that can be used to consistently trigger the exception +I wonder if there is a way to easily tell whats invalid about the configspec +This output is on the Reuters collection again or on the dense data in the code? The latter is artificially favoring dense data +isnt this massive work and a new feature that wont make it inIve moved fix to later let me know if that is wrong +Ill leave this open to dig into the details a bit more +Any idea how can we get this to the Hibernate Teams highest priority radar +javadoc warnings are in and and are unrelated to this patchThe patch was manually tested on Y branch +I dont think so +Havent discussed the release plan yet need to have a on the dev list +Moving these public constants to a separate class would be an incompatible API change so that would be an unfortunate solutionI wonder if instead we might instead explicitly initialize all of the Symbol subclasses? Heres a patch that attempts this +Verified +Should we make that case insensitive +It is not an issue I thinkAgent will create a storage pool if not exists +lgtm +Unless of course you mark the dependency as optional +Im not sure this is a blocker anymoreIt would be nice if someone could either set a new appropriate priority or possibly close this one as a blocker and create a related issue for the remaining work +The replicas all have bytes of the file +Resolved by new SSO solution can be tested now on our staging instancehttp +Verified. +I spotted mentions of GPL but its not clear to me whats +Same issue for metawidget booking example which apart from widget definitions takes source from booking exampleon AS seam branch rev +No problem Ill add unit test for it also +Yes I need to get back to this +Will give a patch tomorrow +Fixed in trunk in rev +Set by default the console property to false for the hbmddl goal +OS Mac OS X Browser Firefox need more feedback please +I can assign myself but can not assign myself this task + on patch +If you are using have you added the ConnectorJ jar to the WAS classpath +I checked in a shim that will allow existing plugins to work unmodified +Ok i ran as you said and I am attaching the am runing Java on XP Maven output from mvn clean install the sub directory srctestjavajunit necessary? Eclipse complains that you cannot nest a source directory under a source directory +Im also a little concerned about backwards compatibility with what is already in placeCan you please what you would like to see in terms of functionalityThanks for your enthusiasm and support of Webminifier +Not tested on websphere +That only works for server that use this URL pattern and is not a general solution +RN marked Not required as a specific release has not yet been set as a target for this issue and it wont be resolved by +This attribute still presents in revision +I wish to have a debate around it and finally reach a consensus on the designThat page looks a little low level seems like it describes the design of a specific implementationI also want to brainstorm about requirements and zookeeper layoutschemasIm not sure if everyone is on the same page enough to attempt editing the same wiki page +Patch for for Struts were committed at r and r. +Thanks Owen! +I tested this in Windows +Once the items above are addressed I will test again and commit +Yes its not good design solution but it will be compatible with previous versions of the library +Changed the repository configuration JSON Schema to have different structures for the transient file database and cache binary storage nested document +LGTM +Meanwhile the AM times out the call after minute and sends a release container to the RM +On this topic Ive noticed that if an Error occurs Click will simply display a blank screen instead of printing the stackTrace +Updated the patch to also add the attribute to the moduleThis is necessary to get the dynamic enhancer to run with the unit tests when running the package target +Can anyone please help me with this? +Ive applied the patch to both server and client sideThx a ton for the patch. +For the code with the old license yes +But in this case I would rather change s toString method and call get there as defaultPlease comment if we decide to take the friendly URI approach we will reopen this issue +Ivy solves your problem because you assumeneeddepend on a complex dependency management system +I also have and the version installed +Sorry to come in late the patch has gone stale +Geez +ping +bulk close for the release +You are right + +Sorry for disturbing you +to trunk +Original pdf and image of the pdf generated with pdfbox +Adrianmaybe the patch attached to can be usefull for the job you are doing here +Sorry and thanks Michael! +Thanks for the patch! +version committed rev +To Maxcorrect the jboss server and jbos esb runtime in the opened page meansThe default jboss server and jboss esb runtime names in the project classpath maybe different from your workspaces you should change the names according to your workspaces +Hi KimDid you intend to attach a different zip file? The attached seems to hold doc pages for a different issue +Thanks for your help problem is reported when new payment method is created +Bulk close all issues. +Im not sure if this fits the needs of QE or not but it would allow QE to run tests while overriding zero one or both of the urls and test the results +And no it has absolutely nothing to do with order of a set or any other collection +Sande please document this in the Release Notes as a fixed issue +I wouldnt sweat targeting it for anything other than trunk at the moment +Yup all tests were passed in +I put this into trunk yet again +can we have an option to output to local FS still? handy for debuggingSure Ill create the changes and patch bq +Already fixed in. +Also the patch is against trunk not +on the slop idea i like it makes it simple to adjust to the needs of your regions +That is I thinkyou can remove the get get etc +Tom C to +In addition to setting the content this only adjusts the transfer encoding if the current encoding is bit +you get it wrong i m not blaiming asking for a help +Thanks for looking into it even if it turned out that it can not be solved with reasonable effort. +It is fixed on the current trunk. +bq +Well it doesnt help that muchWe will need to duplicate that code for every form creationEven if we create a object for the application we will still need to modify so much filesIts also something we try to not doOh well thanks anyway +Patch looks good +Moving to +I think we should replace this injection with Java EE component injection +Many of the items on the above list could have been taken from +Merged to trunkSending trunkSending trunksrcorgapachepivotwtkskinterraTransmitting file dataCommitted revision +There are a lot of huge ones + +Travis I was already working on this +Provided test cases for custom logging service no logging service +John with your above comment then is this issue pretty much closed? +Fixed by. +thanksdims + +We need the fine grained artifacts for those who want to selectively include modules +Have at it +Merged into the master branch +Thanks Ted and Stack +We can not separate out repo team base we have common repo in entire organization one for SNAPSHOT and one for RELEASE and many company they have only one repo which contain release and SNAPSHOT +I have a potential fix for it so I need to attach the patch files to this defect +Currently has a hardcoded location that will need to be extracted +More problematic is that the temporary replicas under detached may be corrupt if dies when copy is in progress +Hmm +I should probably fix that in XOMs build file anyway +Better looking at the file we have now I noted this the target should read because this is the common pattern living into when the run target is used to start this triggers the build target also and we have made the build target to clear the version info in the file +The Dynamic Web Project wizard is a WTP wizard that we cant change +Reviews most welcome +Patchfile for the patch applied +Trunk was already referencing in so the problem wont happen +I wouldnt use the Bug label on issues like this one since its only really an issue with the test suite and even there just for specific modes of running +I agree argument holder classes are dangerous +Yes I agree the whole concept of loginUser fallback feels wrong +initial implementation rev +So all issues not directly assigned to an AS release are being closed + has the new test which will replaceCurrently the test checks only whether the casts succeed and that results can be drained or fails +Several years ago we went through the main style sheet and eliminated many redundant and unused classes +has been deployed and issues like this one have me thinking itd be hard to do +Bruce pointed out that the information was outdated +PS Im not running Solr just POJO with lucene and Tika +Changes made and committed to SVN on QA for later merge +git svn patch into trunk with thanks to scott. +Stepan thank you for applying patch and adding tests to exclude listHowever I reopen this issues because the problem still existIm working on deeper investigation +Triaged July Assigned normal urgency +While I was preparing the example I saw the exception is not present anymore +Oops my modifications were actually incorrect +Set fixed version to +Will commit today to s more of a test setup failure rather than test yeah its would say yea a test clean up seems like a good backport to me +This shows the dependency entry we use to retrieve a +The Sun v JDK does not seem to exhibit this behavior +A similar change for JDK was made in trunk in +Fix attached +Look for some like Clear Authentication Cache in firefox. +The intended functionality is fulfilled using +Now debugging. +trivial patch +Makes sense +This was my mistake +Shoudlnt we activate failover per default via in the clustering module? So the user only needs to register theDont know if we should also activate passivation failover is enough for my projects +See +Believe it or not it is possible to code a layout that works in IEIEIE Firefox and Safari without any sort of crazy CSS hacks or Javascript +See previous comments. +If its only a Designer fix then maybe it could go into as well for the Jan patch +For example if we took snapshot s on root before the rename operations foo and bars nodes are included in s while the new foos node is not included in s +Therere two concerns +Thanks! +I cannot reproduce this neither on Wink nor LinuxIve attached the exact test file I used +Ill upload a new patch to fix it +Since I screwed up the commit we voted down and are working to get out pretty quickly +A Big Thanks to Chris Bredesen +Im going to create a Jira in apache smx and ask someone to commit the new great thanks Eamonn +SudharsanPls check if the ROOT and META is assigned completely and the znode corresponding to META is deletedOtherwise the scenario could be as in +Trunk version of Solr has this same problem too I just timed comparable difference in reqsec when caching the Version vs the current implementation +The failure seems to be some sort of javah issue that Ive seen in other builds as well +It is possible to change the log message after the fact but JIRA wont be able to pick up the change and show it on the ticket +Documentation updated +already fixed with the current trunk +If no one objects to the latest patch Im going to commit it +If an Action implements and creates the associated getRequest method then one can tamper not only with the request attributes but also the session attributes and public setters on objects stored in the session +Verified by Gregory. +Did you declare in mapping file all fields of the base class in the derived class too +Manuel are you sure you are using the jar that came with Struts? Both upgraded jars have to be in place because actually most of the fix was to +the process stops after about minutes with a is there another way to get a dump file? all in my case the most file types in my repository are word documentsif i remove the class the process works finebut if i enable the filter the process ends with a outofmemory errori think we must look for a memory leak +Also test was unexcluded +New patch adds the unit in the description +I just ran into this on the branch also so committing to both. +Forgot to add this incorporates +It was turned off by default in so as not to break things for users upgrading from +to trunk branches +Id like to open up a new jira to track the java issue separately +In WSDLJ there is a class which is used with the which has a reference to it +fixed and too to be included in forthcoming pull +Shot of example project showing bug results +Looks like this patch has never make to despite the fact that that branch has been cut later then the commit and Fix version is set to +So we are closing this out here. +youre absolute correct +This would work for primitives as well as reference types. +Fixed the issue +Turned off wifi and it changed to none turned on network and it changed to g +Do you have any problem with moving to next stepI run a Korean developers community for the Korean AnalyzerI announced that Arirang analyzer will be incorporated into lucene and solr soon +JBPAPP branch switched to revision for CI test results +A lot +I cannot reproduce this I tried to clean and rebuild and had no trouble +xpdl support is on the roadmap of our new pvm approach +The mentors will look after this as needed +This is a simple problem and a simple fix +Yep Im still looking at having some resolution to this beforeThe more I look at this the more I think that the interface just doesnt make much sense for +Strange I just a couple of times on mySQL and it does not produceany errors +you name +Lets make is permanent for now +It is a little bit awkward to have the value in an attribute but it does have a few key benefits like documenting the availability of the feature +The first statement is more general as chain might represent a configuration where there will be no output message +Brian the patch looks fine to meRegarding the there was a problem with the ant task that generates the jar +It seems we still havent standardized on a file header so I went with this simple text lifted from the ruby README Thrift is distributed under the Apache Software License +Cool +Only set up for needs to be set up for other servers and also tested on windowsScanning local repos not implemented yet + timed out +WDYT +patch applied +core tests +RE My Above through Thomass wordy comments cleared up my question with this +Rather I was hoping that those files would be added to the svnignore property so that subversion would explicitly ignore them +Thanks for your time Frank I will check the system by removing the jar file mentioned and will update the statusThanks +Would it be possible to have a cleaner output such as the one from the osgi list command for example? +cheers and good there anything need to be done about this in meclipse or we can resolve this issue? Thanks +Ill come back to this first thing in tomorrow and create a private fields of to protected for more subtle filtering of column a patch +The class generated from your jsp is holding a reference to the class from the previousdeployment of the ejb must flush this reference by redeploying the jsp put the ejb interfaces in a separate deploymentthat doesnt get redeployed when you redeploy yourbean use semantics if thisis a remote interace +thanks for the applied in r +First part of application +If the content is stillvalid the portlet should not render any output but either set the propertyUSECACHEDCONTENT or and a new expiry timeor set on the of the or and set a new expiry time +ant Dtestjsr resolved issues for already released versions +Perhaps things have changed +Still no answer or comment on our request +Workaround found run ant undeploy before redeployment +Samuel do you have test case for original issue which you tried to fix +We can store those on our implementation of +This would likely need to be completely optional and turned OFF by default +I am convinced that this issue is not important enough to block +This particular feature of cfdbinfo is one that is used by the framework in order to provide database introspection +sure thing ill try to bring that in with which is still on the list +Thanks Mathieu for spotting that yes TTL has been introduced in Mongo +css page +Andrea AimeYes i set the Xmx switch but the exception is still thrown +We just need to add smarts for and +committed to Im aware of that break +I am unable to reproduce this on my Windows VM with the latest +applied thanks rahul +Committed to branch. +Fixed on trunk thanks for the great test +It is NOT a correct action to remove the carrarwarejb in this portlet so we should not provide such action in the page +Now that the webstart plugin doesnt have this special groupId users are forced to either specify the full groupId on the command line or change their +We need to remember to remove that once this failure is resolved +I think Im about ready to commit what I have +The patches I uploaded were based on the strategy employed by +It would be good to pull any misleading documentation that is out thereAs an aside even in areas where we are DRDA compliant there are things very particular to Derby that are implemented as stored procedures for example the stored procedure to retrieve message text or the stored procedures to retrieve partial LOB data so very specific efforts would need to be made to get any client other than Derby client working with Derby +Since nothing has been committed yet Im attaching a new patch that incorporates these changes +host A dont get anything and sends ping againIm going to take a thread dump in next time +Ryan do you care to create a provide a test POM? Or just show me the relevant POM extract that references the system dependency +Submitting the patch +Archive with project demonstrating libs issue +did u use the noWrapped flag in wsdljava? about this issue in the coming new releaseIt should be nice to confirm it has been fixedThanks +Here is unfinished work just for reference. +I am experimenting with registering a custom as a cxf extension +This is my JSF page which fails to render +bq +internalGet and internalSet might also be reasonable names for this +fix committed. +pending to trunk and +For the time being closing the issue +works fine with spring +btw cannot reproduce easily it does not hapen every time +Problem is this will make dev environment fail to compile +along with after release. +A pet phrase is an expression that a particular individual favours and uses frequently enough for it to remarkable from pet in the sense of favourite +Mikes request Do you ever get a bad query plan if you use ? parameters rather than exact values? I would expect not as it should travela whole different path through the optimizerI tried this and what happens is that is called with null values for start and stop +Too large for Lang. +I have a test unit and a code fix +I havent looked at your patch but in general I think subclassing is a very fragile mechanism for providing new filesystems so not sure if we should be making changes specifically for subclassers +I am going to assume my first results where invalid +The signalling was working fine when it was a custom activity +So backpressure would mean having to deal with both and potentially unbounded wait times +Sorry +That would eliminate a number of fragile variables inThanks +Any other build after that gave the documented error of Cause Class cannot be instantiated +bq +Dupe discussion should continue in that issue. +The only seen problem is that test is failing +Would it still use the on the bean model still +The context param handled the case for Document input for endpoints +Jimmy You have a from Jon +that first log message clearly indicates that that is a list of hidden files so its not suppose to match whats in gettableFiles +To have backward compatible behaviour the default value of the emptier interval should be then the trash value interval is used for checkpointing +Committed to trunk in revision +Default font provided by seems to be much better +on patch new test fails if change in is not applied +Im using on the client and it automatically generates an AJAX POST request with the model data as JSON when a new instance is added to the collection object on the client +The base showXXX methods are but the class is constructing Swing components off the EDT which is evil +JIRA cleanup +When the replication launched the will soon reach a commit of G memory as files in the attachment +fix for for the patch David! Do you mind providing a patch for branch as well +Patch has been verified +Removed printfs and fixed memory leaks +Personally I would have put the Build Environment information at the end since I would consider the list of new features and bugs fixed more important +Canceling patch +Have you been using your own patch in any kind of production environment +Paul has previously argued for a very limited set of options for module developers only and I must say that I favour that approach +Jonathan FYI you didnt attach a patch +Id like to see this done for as well +full eclipse had the same error +on this patch +Miroslav thanks for the feedback +fixed to be consistent with commits. +Introducing locks requires careful consideration of various scenarios +on patch and TRUNK +Lets see if it helps Alan +Patch has been committed +We found consensus that Stanbol is a fine name for a TLP. +So uploading the patch with new name +The latest diff includes an integration test I started work on +reproducible at GA +I am going to have to stop working on this jira for a while until I look at +Decided that would be the side NOT importing. +And that to get TPS it would be expected that Netty would be able to handle this without the need of a thread pool not until you get to higher it would be good to get a view of what you see in terms of TPS without the thread pool as that makes a big difference regarding what the customer is trying? That way if we can agree on a flow that works for everything then we can focus on that as being the main test case and ensuring that the TPS seen by the customer and RH is the same and then figuring out how to resolve it if it comes out to be thanks +poms are already rejected if they contain dependencies not in central repo but theres no automatic check yet so some get in the repo unnoticed +Just found it in config +There are classes from the jetty implementation directly included in the console +I have committed a fix that avoids the problem and fills in with Object parameter types if the correct ones cannot be found +is it a configuration parameterYes and it probably needs a better name +ATM I dont have enough time to work with Cocoon until this weekend so if you have you could take a look and improve it +I dont think so +If you think they are critical for CR please them. +Your best bet is to look through the change log for each subsequent release and see if the issues addressed relate to your specific use case +What is the current timeout? Is it minutes +I think this is pretty generally useful as not too many companies make data optional for their users +Committed to trunk. +Binding code generated using from AXISHOMEbin uri resources p d xmlbeans s ss sd ssi o build has stopped our development for the moment so Ive changed this to blocker +is not really useful right now all projects should depend on and not on has been committed to CVS +will be soon after so not a big deal if it gets into the next point release +Hi ArjenThat is great news If I can be of any help at all please do askdavid +A getter would make sense too so that one can see what the threshold isAlso the explicitly states that it is not thread safe +I just committed this. +As Jonathan mentioned in IRC it seems like mapreducing over indexes will be much simpler if getindexedslices and getrangeslices were the same callAlso there have been several bugs in the getrangeslices code over the last months + is Java specific but I guess the feature can be exposed to other languages as well +I looked for an option on or in the javadoc tools arguments to fail if it discovers a malformed javadoc tag but I couldnt find one +I just committed this. +The reason that I use this method is to avoid so much reflection when do the automatically loading of classes but at the moment Ill let it in this again for your help I think that you can close the issue but Ill take a look to see if I can introduce this bug in the JDK issue M +There is a licensing question outstanding on the +Id rather have these messages ON by default Hadoop should be easy to debug and troubleshoot out of the box +Punting any remaining analysis to post +The parallel manner does not translate into an overall throughput improvement seems i have some misunderstanding on this code snippet emm i have to debug it per your comments +Chalk it up to jenkins randomness. +it seems clearer +This is kind of similar to the current approach of dropping table data in partitions wont be deleted if their locations are not under tables locations +RickCommitted revision +Marcus This is as good a place as any +As was stated there pushing the publish icon is an INCREMENTAL publish and will not execute a full publish +jdo files +Could someone please review this patch +Attaching the patchThe issue is related to localhost resolving to on Windows +The wire info seems to be some kind of debug logging +I have a patch for this but I want to make sure there are no legal issues with including the testing directory files in the distribution +I am uploading a tiny patch to fix the problem in the the the same patch but this one is named with the JIRA issue number +patch looks goodI have committed this +David was the description correct? Please closereject it if needed it seems to work fine now indeed. +Given this jira has lot of discussion related to what the current title suggests lets close this as Invalid and open a new jira with the current patch +In documentation many occurences of HADOOPHOME is seen +Well keep the JMS message store for a later release +bq +This is more of an aside than anything else but clients do have some support for delete after close right? The whole +The lack of such tests is not a blocker so I am reducing the priority +For the +I think you may have more problems blocking the clients on the FSN lock during standby transition instead of throwing an exception immediatelyFor one if all reader threads are blocked on FSN lock no more RPC calls will get through even HA admin commands such as HA state transition +Great! As for the commit mailing list wed like to use the Struts dev list dev to increase the projects visibility and encourage other Struts involvement +Will have a look at the pull request later +No its been renamed to and +rebased the documentation patch for the edge output format +Thanks Carl +No verification error occurred +But sure fixing that is fine too +It tries to match the pattern against the methods of which is included in the application context by +Ed Do you have time to work on this again? If not Id like to pick it up +shouldnt be our goal to fix all test issues on windows hereyeah thats true and I agree on this +One is the result of filter the other is just to reset the filters +I was not aware that you want the static variable to be created after a call to inirializeLibrary +Now we can close this issueRegardsSumit per plan is complete. +That is use mvnxml sorry +Backport of the forJianing would you mind testing the appropriate patch for your version +Thanks Arun. +Suit yourself but a lot of folks really like this feature for a build system +not sure +I think we should do the changes in a different JIRA thus this one would be just a backport of the +I think the first sentence is not required + for FLOOD there was inet problem and I thought duplicate submission was is +Normally all decorators see also other packages have a static factory methodBut I have to admit that some things are not consistent some decorators have a private constructor mainly when they are supposed to be final +Rendered by adding three lines to indicating where to find TT fonts +stack Thanks! I didnt see +This applies to Axis as of todays svn +Output from helgrind that shows call stacks +Any will do +They encapsulate logic overriding is only required for the data on which the logic operatescheers Roland +Therefore the attached test case wont have a stack trace the parameter is simply not bound. +I have looked at s +I cant imagine why this would happen other than it must be a bug in the JSF impl +I think Enums are great for this and have wished many a time that the types were Enums while working with PigI do want to point out though that this will affect a lot of user code any that specifies a schema any loadfunc that implements the metadata options etc +I believe there is room here for much improvement +There was added a chapter in part of reference documentation +Already updated the code but heres the explanation +This one actually breaks the loop +Emil thanks for confirming I am not crazy +So not only is this not fixed but its not in +The point is that the Maven CLI provides that functionality so developers should be able to do the same thing within Eclipse +Closing jBPM issues resolved in a released version and reported by me. +Yes seems to be a more common names of inner beans seem also not to be checked +Mass closing all issues that have been in resolved state for months or more without any feedback or update +This page is part of the GWC own UI has no control over it +Junit should be reserved for code associated with the junit harness itself not derbys tests +Removed changes +Your changes look like improvements to me +But I see there is a need for this +Ill commit it later todayregardsMartin +When the user switches server mode the poller section will update itself and refresh the available options in the combo drops +can you do mvn you are right a mvn clean install is working as expected +Need to add unit good +Duplicate of +Patch for for this necessary? Does snapshot snapshot scanner cover the copy table side and adding a snapshot mode to verifyReplication cover the need? +Ping +why should we have to click buttons to type in credentialsAnd again the main usecase is one user and here it will always be remembered those cases where you want to switch you just type the first characters or press arrow down to get the userWe could add a browse button to select users too but I would rather keep the UI simple for the primary usecases +Application Server issue for the changes in submitted files are patches for branch +I use it inside a form and it works fine +Thanks for taking the time to commit this! I really appreciate itJake +pending results from Hadoop QA +I do not think the name neccessarily has to be that way +May be I spoke too soon +All tests in pass for me +This bug is not import maybe wed better commit it when the next improvement in cache +Code submitted by BenjaminVerified and ready for distribution. +I prefer to make them wait +In any case may be it is the whom should complain if theres no symbolizer +One should never rely on the ordinals +not really sure what we can do about this other then somehow cache which classloaders are used but this will be broken in a clustered deployment anyways +I only have tested it with the source from the but as far as i can see there are no relevant differences to the trunk +Attached is the log requested +Should I try with the nightly build on ? +can be used for testingOnce this is good Ill deprecate solrcoresrcjavaorgapachesolranalysis and fix the test cases +Thanks Sid! +Done +ryan this groups by regionserver and sends out to each in threads +Carlo can you please have a look? Thanks +You need anything else +Hibernate is no longer maintained. +Is this really a topic to be included in the usage documentation? I would rather say it is a development topic +Ill check to make sure. +This is a serious bug +This should also automatically enable the to work. +So what is the problem using them? after the server is compiled if we want to change the gbean settings we need to edit the file I guess that the initial request is that we could directly edit the gbean settings in each own folder +Yep I think were in agreement then +Only a few site admin and project admins screens are still using it +I vote for this patchThanks JacquesI noticed if you keep the name of the attachment the same then the older ones are grayed out which makes the newest patch obvious +Moved to and correspondingly its PB to does this seem a right place +But estegosaures is plural while stegosaurus is singular +it explicitly mentions Throws if the argument is not a Timestamp object whilein Java spec not +Add class to to patch as well as some mods to and subclasses +hint text is actually supposed to be provided by a different HTMLattribute I dont remember which offhand check the HTML specs +Patches the problem by appending the branch parameter to the git command in case the version is a branch +Weve been making progress on removing the tapestry testing stuff in favor of Geb. +Move this to since modifying for metadata table is an easy workaround if this actually occurs on a system not running random walk +As discussed I am assigning this task to you with the +Demos dont have to be static once they are in the source tree. +Why was it decided that should be but not the? Why should I endure the overhead of paging when all I want to do is use a table as a queue and have different consuming threads? Maybe I am not understanding the difference but I just read the documentation again and it doesnt make it very clear why I should choose to use one over the other except to say that I found one forum post that mentions Cursors are the responsibility of the RDBMS vendor so you might expect that they would be fairly robust +The additional parameters concurencyLevel softKeys softValues may be passed as config parameters +What purpose would that serve beyond the existing cache saving period +I think adding getContainer to the security token is the right choiceOne minor problem with adams patch is that it has hard coded +Feel free to provide your comments +The library specific PR has been mergedThe integration test is still test merged. +BTW Recently was closed wich is about ehcache shutdown +I just finished upgrading jboss to within torquebox +I think this moots considerations for restricting themWe did discuss code weaving in security policy earlier but Im not sure how fruitful pursuing that would be given the complexity involved and the murky assurance that would be the resultWe can distinguish between external and internal coprocessors with +It this the same problem as when snappy jar is not included in the classpath +Thanks for doing this GavinRegarding the website Im in the process of updating the content so if you could create www in minotaur then I can take it from there +makes sense to be inline with the admin objects wizard +Verified in JBT to add a link to to set release note status and issue now that release note text has been written. +The same goes for the opposite when setting a formobject a property updates some value which needs to be confirmed by the user +Looks good +But i am strucked with a lock +Rossen Would it be please possible to link this resolution to example of use andor appropriate documentation? +I would suggest creating another issue to remove it +Sent +Cluster ID is displayed on the dfshealth web page +Needs fixes in EJB to test +Attachment has been added with description XSLFO file that reproduces the problem in +I tested it with some test cases which I have for my application +Thanks. +Seems like there is come concurrent access to the +I planed the GA release early this week but a problem was found in released date will be late this week or early next the current snapshot does not contain the code for eager initialization +Would you be willing to continue working on this patch? patch made life much easier for streaming requests +Just let me know when you need a release. +New unique index on +Attched unit test for changes made in implementation +Caching just the XML string doesnt buy us much since the httpResponses cache is doing pretty much the same thing +clicked wrong license checkbox for patch +Sample projects demonstrating the issue +Ill create a patch in cvs provider to delete the directory if it we can checkout in a dir even if it exist and its empty i dont know why it doesnt work on your machine +Thanks a lot Jan nice work +When dependency resolved clear this issue +Havent done full ant test or tested with trunk but should be low risk +Suppose you use commit option A +I work on a product that has a class that edits and +The problem was that we filter the directory after we get the list reported from the namenode and the list is limited +If someone can advise me on a better way to create a patch for modifications and new adds Id appreciate itThx checked it out +Looks like the unit test is the testcase and updated the patch +Committed to +lgtm Sid for the review! Karam you too for testing the patchI committed this to trunk and branch. +Block level and column index level are actually the same right? kbThe reason block isnt ideal to me is it makes it much harder to recoversupport partial reads since the block has no context in the file format +Plugin has been working for me but now all of a sudden I have no Hadoop options as if the plugin never existed and my HDFS connections are gonePlugin is still sitting in my eclipsehomeWhen I look in the Usage Data Collector Preview I see a bunch of these errors Could not acquire children from extension Any ideas how to fix this +runs out of memory and dies the broker closes the connection +Im on this approach +The patchmodified files are and for tiger +We should fix it but not when the fix is on the critical path +Good work Sijie. +Patch to for python minimum versionAlso fixes small typo with java we should treat this as a bug instead and fix it back to compatibility + +However the docs are still way out of date +This looks great Uwe thanks! changes to ctor docs +That will be stable if you pick a date prior to today +I think there is no provision of submitting jars for specific query +The test code was satisfactory +s have s in the schemalocation +Id suggest to keep it open +If people want to use different versions of tasks they should simply use a different version of the whole plugin +If one used the JAXP to get a transformer factory and a SAX transformer factory one would have stayed consistently in one implementation +I have committed the backport patch to +I wanted to move it to percentile for the new actions where I will be using this framework +Is there progress here? We could use the feature too so I wonder if it makes sense for me to pick up the patch +do you get same problems if you dont use encryptiondo you get same problem with other types of encryptionCan you verify that HOVantagegmiwebclientdb directory does not exist before running the create script +If you find further issues please create a new jira and link it to this one. +If you have a real application that breaks because of this I would be interested to hear about it since I dont know anyone that uses those level +seam depends on richfaces and richfaces depends on seam +I have been trying to recreate the problem with some variant of the code that you have here +I think that means the new api is working +Thanks! +Really strange because when I dont use the sortable thing that error doesnt appear +This will cause test to failFile to be updated is modulessrctestresources by properly retrieving binding context within the response handlers +should be fixed now +My bad the comment below the patch made admins can always read all doc fields regardless of publicfields made it appear like you were talking about a workaround not the content of the patch +Verified and thank you for your patchGood luckLeo +Im on doing a blanket replace in the name of efficiency +The expert group rejected the notion that a portable mapping for arrays collections and maps will include a primary key on the join table +Hint somebody already implemented transaction support for Lucene a while back +In branch +Can you please confirmIn general its best to post a CVS diff since that makes integrating patches much easier +So groupId will become +Adding Martin Davis to the watchers see if he has a perspective on this +In your case you appear to be using the right type of field +Will not work for databases with no associated +Committed branch and trunk. +It would be difficult to integrate since the user may choose to install the NDK anywhereIm OK with requiring that the tools be on the path +reopen to set fixedCan you explain in more detail why the default value should be artifactIdIf you take the WAR that is produced and deploy it to Tomcat manually +So nothing more to be done. +However for high latency sites the output buffering may indeed result in massive performance improvement +PNG of the is a very large patch I will be trying to split it up into a few smaller patches +We might when the file is present but with a newer date check its size +Is this reproducible? Can you get a packet trace or verbose log +Attached a suite of test cases that reproduces the problem and a patch that fixes it +If a binding is set and the view root changes then the old component which has a binding to a backing bean and doesnt belong to the new viewRoot is still addedThis components have already generated ids out of the previous rendering as they are not considered in the new rendering phase +Committed +Updated patch file +this patch seems to take care of the situation when BOM is present +I realize broken is painfully useless but the release and component was about as detailed as I could get at the time. +Would be good to add a simple test that +Ill keep you posted +Updating Activiti to version seems to unresolve the problem +Ditto for hadoop profile for use by above script so it can find where to set profiles +exactly our sources are not in the standard maven structure of sourceswe are using sonar v is there any work around +A likely much larger issue resulting from the same code error is that an expiration check for every remote session will be performed once per check of a single locally active session +It seems that add new code +I was able to repro the problem and validate that it is fixed with the patch +actually the nd two are likely fallout from the firstif it died and didnt cleanup the locks this could happenas i noted im a bit short on time tonight so Ill get to this tomorrow evening +Thanks +The implementation of this method creates an to record the output stream and to make it available as a String return value +Its because only one process can act as Auditor and others are just workers and participants for next electionsAnyway that can be updated once name decided +This will either be or Disconnected + +We can work it out in hbase and then contrib +Fix already applied on Karaf trunk revision +Hopefully youll be able to confirm that and get back to us +Are you logging in with the right username +Windows Cygwingit +LarsThanks for taking care of committing to the other branches also. +Typically these are data updates that need to happen in the jBPM DB +This is a blocker for +Okay this patch should cater to both keeps it simple and doesnt break compatibility in the existing build process +clients issue cannot be applied as is anyway! +I got this issue while using Rackspace node driver +Patch applies cleanly and the tests passIdeally there would be standalone tests for each of the distance measures that test them outside the context of spell checkingI think the threshold should be configurable via a setterconstructor +Assigning this to Test component. +The latter exception is caused by no request being present in the startup thread where the call forces initialization of a target instance in the request scope +Please test +as r thanks for reviewing Sijie. +v +Attach a patch +Kevin is already working on it +Thanks for the hint we are already using this method but this was done in another component using SolrLucene +This should be resolved by the upgrade in. +Patch that implements this in the +Does Cocoon have a datetimescale in mind +Ill resolve it for now +Thanks Tim +I again need to mention that I never tried that myself but would be very interested to know if my understanding of things works as I expect +Message is not displayed in JBDS +Yes this issue goes into the same errors set with specified known that work is well underway with AS all previous community releases are +Still both require full access to GSE and it must be hidden somewhere internally +Seriously though I do want to help +Jetty scenario fails very seldom now +Just need to check whether there is still an extra in front of the midkeys +Well +Please apply the patch at dir vtestassemblycomposite and the related testcase is method ASMn in file +If we think could fail in any sense I belive this should be done under integration tests +I am testing this on and not seeing this effectJust to check something do you have the version and the version installed in different directories? When you run and switch between them do you run them from those different directories +Hi XiaoFor your test case I think it would be much simpler to just use a Writable which throws IOE rather than mocking out all of Socket etc +Applied the patch to branch and trunk +DidDoes this problem also exist in? I have a customer encountering what appears to be this same problem and we are currently supporting them on +Im seeing the same problemerror using Grails deploying to seems to have been fixed. +Its possible this is a windows only issue +this outweighs the application level uncertainty that would happen if a request fails because of a region relocationI think statement already achieves what persistence to WAL would achievebq +What is hiding behind the better? Any particular cases that can be used in New and Noteworthy? Scrrenshots of how it looked like and how it looks now +I cant believe I missed +Lack of support in Roo inhibits it from using in real projects unfortunately +Using bash or a different solution + transition all resolved issue that did not see any further comment in the last month to closed status +Without these changes the Delegation Keys in the binary image cannot be correctly composed +Yes I use ant clean package +Thanks +I didnt do that for this because the changes could have a wide impact and I wanted other eyes to review it +Trivial patch attached +Once you publish it will copy to JIRA and email +Moved to GA set the fixed version back to beta if you get it in in time +You should use JDK if you plan to use this feature +I dont think this is harmless +It should have all the needed wsdl and xsd files +Thank you for at least considering my request and for all your hard work developing the Groovy look forward to using Groovy in my regardsRichard. +Ulrik I would like to know whether you have completed working on this enhancement +As a workaround if you change the database to HINMEMORY all the tests pass OK +I will commit it to +This patch adds a snapshot of nodes that failed tasks and failed tasks on jobdetails page +Patch to and change to existing or functionality +META +Looks good to me thank you in new version that compiles back to FP further feedback on checked in code so marking this as resolved +Actually +Based on the description this doesnt sound like a database problem but rather a problem where rounding is missing in the code +I appreciate your pointing out setException a handy method I wasnt aware of +That seemed like a good decision at the timeThe truth however is that theres no perfect encoding to assume in mode +Session bind doesnt change something +Hi Lasantha I am having trouble to apply this patch to my local machine +Thanks for looking at this Eli +Attachment has been added with description Sample FO to reproduce bugSome initial observationsThe issue is again with the last +Lets think through whats going on here some more +I just tried compiling with a forked JDK compiler and it seems that the source parameter is still being passed to the compiler +Ready to go into the next release attempt for +but thanks for the very detailed analysis and your findings in helping us fix this issue +In other terms if we look at which is going to be released in September as of today I can maybe say it will have a WCS service in core but thats all Im sure it will likely have another dozen new features but have no clue what they will be. +Turning null values into empty strings makes sense +Are you looking for something different from thatThanksAny news on this? Again the sample attached has a standalone CXF server deployment that was attached toGio +exception which is caught and logged without any other negative consequencesSo please remove +I am not sure how we can effectively mark and inform users of the stability of command line interfaces +There was also an error in which I thought was an instability in that test it didnt seem related to your work and the error did not recur when I ran the test standalone +New is now in placeTo reproduce the bug remove the annotation in +test case attached in and branches +IMO the right fix to the lock contention is to simply not synchronize when there are no indexes present +Thanks Tim. +The bad part is the container launcher didnt have all the information required to actually kill the container so I had to add it which Im not completely happy with but seemed the best fit +Wont make +only got written down in you are implying that uses Ant +Hadoop needs to fail in a way that lets them know its their problem not the Hadoop codebase +I see youre right +i havent been able to find a way to include the neccessary jars into the EJB jar but using a manifest file i have been able to include them into the WAR file in which the EJB is packaged +As this instance is still visible from subclasses I made the change in both places to ensure any lingering usages got the update as wellMoving the method to static as has been done in does have the disadvantage that anyone wishing to override the default behavior cannot simply override the method as needed +The overrides the type personName by adding an additional attribute to it +Ideally I was thinking of being able to do a head to head test +This is my fault +Im going to move this to the release jira +I have adjusted similar template to process this tags so there no big changes and they safe +I have defined the user role relationship as a many to many with an intersection table and have added usage asGordon Dickens +Added as a prerequisite in POM for version +Did you cvs init it yourself? Its missing the passwd file and several symlinks that are standard +we allow acquire and release locks fromIs this in lieu with having some cross region locks using? May be we need to make the try and release to be public in case we are giving the option to acquire and release the locks +As far as I remember execute batch returns the number of documents that it is able to send +Can you try upgrading to to see if it still happens +sound good +Thanks JonathanMarking resolved since already good in trunk +Maybe you could check if this hang happens any longerwith your current ofbiz by simply throwing a Nullpointerat this code line +it was such an easy fix +The script does the following Runs svndumpfilter on the source repo dump with all the proper includes and excludes dropping blank revisions and renumbering Resets andor creates a new repository in the specified Pre populates the new repository with the proper Loads up the new repository using the filtered dump Removes extra dirs created in the new repository during the Restructures the repository to the way it should conversion should take about minutes to attached +like can you past output of java version here +I confirm this change in IE setting fix this reopen this issue because I think we shluld better pass a size when using +starting with Apache Openmeetings the param to run the Debug player isswforswfIf you will use the SWF files directly AudioVideo will not work at all +Theyre now gone from the mirror so resolving as Fixed +To be clear I have no objections to moving forward with the proposed patchWe could continue to try to add other ROWNUMBER tests and window testsas part of separate efforts in separate JIRA issues +I would appreciate it if this feature could be added and the user could optionally take advantage of it +All other instances of calling +Unresolved as of this one I beleive Cory is working on a themer for udig? +Or in the case of STAT without a parameterftp quote STAT Connected to Connected from Logged in as anonymous End of statuswhich agrees with the RFC +Let me address these in the patch for which modifies the same test +site updated +Verified in jbt build. +As far as Im concerned the output of the pamphlet generator can continue to live where it does now +Does this not work or does it take too long? From your comment above I suspect the later while the bug title says different +Thoughts +We have compensated by having each CI task have its own private local repo but this massively expands the disk requirements and has the affect of slowing everything down due to all the duplicate file downloading going on +Three sections and dont removedelete last one well DavidIm reopening this ticket just to suggest a possible improvement When a user click on the Manage Dashboards link this is always because this user comes from either a project or a global dashboard But for instance coming from a global dashboard its then possible to unintentionally create a project dashboard and its become really hard to understand why the content of the widgets are not displayed +It is recommended to use the default value +r for plugin + provides a method to decode text according to RFC It is already used in to encode header valuesIt is also part of the tomcat distribution so we may copy it to fileupload +Schema jbossesb requires that provider element should contain at least one bus element +Sounds good Gabriel did you test the patch by any chanceMy original worry when I wrote that code was that you wouldnt want to put unnecessary load on the Namenode if nothing was needing replication but a second pause is probably good enough +Missing docs I think you should add a section to the Admin guide for this +the time when an identifier is assigned to a referenceable nodeWe can still revisit this decision later on should it become necessary +allows chooseTarget to ignore the datanode load factor when placing had ran using the trunk plus this patch and the patch for iterations without any failure under both redhat cygwin +Committing +Quite right +Patch fixes the followingcoreLucyUtilcoreLucyUtilcoreSearchcoreLucyPlan +Code and doc look ok to me +I reopen this because I have reproduced the attached the try mvn P intdotw on windowscygwincmd and windows my settings doesnt contains the solaris my settings contains the profile but not active and not in activeProfiles the pom to reproduce it +Probably its an issue +I also just stumbled over this +listtointeger is the thing that fails on this input fwiw +Applied to x and trunk +If you call content assistant its popup window appears near Text part of be it would be better if we prohibit content assisant in Visual part of Editor +In my case where Im building an application which dynamically create criteria queries by introspecting entities at creating aliases is far from ideal +Therefore the implementation should be updated to match the documentation +Oops +Committed to trunk! +Sorry to give you a hard time on this but I really think we are closeI can be up for another hour or so if you can fix the last nit and Ill commit +Can we measure if there is any noticeable performance downgrade due to two additional and operation +Patch is attached for branch +Created the issue for not losing the information gathered and solutions proposed +Resolved via +applied the patch +But before javabin we used to use xml which has same Semantics for namedList and it is right +So it is better create a new JIRA in order to fix it since this was already resolved in May +Ive committed the fix to trunk with revision +Manually tested patch to verify it does not inject null queuename pools into +look into the code +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +The current patch will fail in Windows as the commons daemon download will failOne solution is to move the commons daemon download into the native profile +pushed to +No further information on which component this issue relates to +This patch fixes the problem +Patch for branch in the patch to +UhI thought we patched this ages agoSee +Ive also tested the patch on EC and cloudservers +Too much trust into the auto install file generator + +verified fix in ER. +Support for obtaining a SAO proxy on the client side via a and for publishing an specific object instance are completed +To avoid this in the future we should create an integration test that has a few half way in a from the prev Oozie release and the test runs them to completion with the new Oozie +Anyone who disagrees is welcome to reopen it. +Can you attach a code sample that demonstrates the use of this new method? Ideally the example will highlight why this approach is better than what you can accomplish using existing framework functionality. +Select Properties in the popup Find and select Java EE Module Dependencies in the list on the left of the Properties Look for and check the box beside it +You dont have any reason to run the command from MAVENHOMEbin you are seeing this issue in it is due to having the install in your PATH ahead of. +Hey I just found the +Ok I will apply your fix soon. +Can I just drop in newer redback jar files and have the problem magically fixed +Reopening so that I can mark as WONTFIX + +Have directly implement both and Principal +Ive attached file with discussed fixes +I had no problems turning objectspace dynamically just when I ran Nokogiri +Tests passedCommittedThanks John +The patch looks good to me +Have been fixed +Is there an advantage to that over opening up libvirt and querying it remotely +The latest patch looks good to me +Reporters of these issues please verify the fixes and help close these issues +Created a patch +We could please all the tests in one file then we could determine the order +Bulk move issues to and +A side effect here was to remove the logic that allowed for a sort string literal to specify a positional ordering +wrapping the summary around the whole reactor is part of another bug Ive stopped displaying seconds though. +Itll be introduced with the next version let me allow finish the process of releasing first +I removed the and added +looks like the patch assumes a particular directory structure will be used when creating the assembly if so then the required structure needs to be documented +Trivial close resolved issues +The quickstart readme file has also been updated to be in synch. +If you have already reviewed the one on RB then I can attach that here +fyi I asked the gods of jit on mailing list whats the cause of this +This is the that was missing in the last patch +I still need to actually test it against an ES instance but theres probably a working patch next weekPerhaps it can still be released with +The from system test framework was a false negativeThe aop system test was indeed broken but it was not related to this see. +If this is the case please list them explicitly on this JIRAThx Eric +When used for edits it could be incorporated into the Output classesThis should help the development of +Final patch +Patch added +the split into separate patch file so jenkins doesnt to include changes from so that should go in first +fixed in latest svnthanksdims + +Waiting on +I recommend we just make a command line tool rather than a formal ddl +Did you select meclipse features and then expanded Callisto site and kicked Select RequiredAnyways I am glad to hear that you been able to install those dependencies +Can you confirm that you unzipped sonar in a fresh directory not over the previous version +fixed by changing the batch files the old hibernate jars within seam were picked up before the hibernate jars inside seamgen +Also we should modify the to include javadoc from into Lucenes javadocPlease feel free to update this patch if you make progress in the next day or so before I manage to commit itThanks +Could you provide more details hereThat method returns a long so it cant return null +In section the spec also states The Supports transaction attribute must be used with caution +The attached patch makes the build work again on the branch with no file +The volumeMap depends on block scanner to scan the disk and update the volumeMap if there is a difference +The missing error messages we filled in in and the sample now worksThis is still crashing with exception +However shouldnt we just remove requires from completely +I opened for that but I think we are good to move forward here +Implementation in will be taken up in a separate jira +Is currently worked by +And by the way marking it as bug or feature does not really influence when or if someone on our side will write the code for you +Tuscany with out this had already more than most Apache projects instructions on how to get started this seems to be just a little heavy weight Im not closing this as wont fix to give an opportunity to discuss on the mailing list or IRC first +Patch root is set to the project root +I should have mentioned how to use itoutfile SYSLOGerrfile closed the Bugzilla variant as WONTFIX +One problem is that the space reserved for this purpose seems too limited +okay +Ive got no objection its great idea to comment the fake writeAnd it may be a good idea to make frameSize a local variable +So kicking off another build will not be easy +This patch will alleviate the problem +Hi OlegThank you for your prompt responseLooking at my application code it would technically be possible for a client to be started but not shutdown given it was created outside the tryfinally block where the finally block shuts the client down +Rebuilding surefire with this version gives me a different type of problem which may be related to my local environment not satisfying some other criteria for building the project +enabled with r +yes we dont provide ide specific files that is correct +Bulk Close for all issues before incubation +The patch prevents the appearing of Missing Natures Dialog +Unfortunately for various reasons the cluetip plugin doesnt use event delegation +Now I wonder if Rashkos patch is not good for other cases which make more sense +Trivial patch +Then this problem is more serious than I thought +some timing data indicating that this change really is faster particularly since it replaces the internals of so it will affect upgrading users even if they dont change their configs +Moving to RN Not Required per will not be released again +Fixed in revision +And +works great +Agreed we can still get into a bad situation if the AM cannot communicate with the RM after informing the client +issues +In PIG in UDF parameter list does expend as flattened list of columns +Thanks leoVerify by Sean at r. +That is the most likely explanation +Yeah Ill try now +Targeting version +Resolved in rev +Thats the core cleanup to exceptions eliminating redundant code where possible and making use of the hierarchy with as little code replication as possible +Could you add a wiki page about it too +Changed fix version +Patch to turn into a utility in contrib +I think this should already be fixed in master as internal Spring Security attributes are now migrated by default +to make rest and thrift do same as master and regionserver so both can be startedstopped via and logs go into files named for daemon in HADOOPLOGDIR +svn Verified +Now that you mention a plain Servlet doesnt work neither +Please commit +Pushed to master in c and in with a spec. +good no exception occured verified in JBT Linux +minor update in anybody had a chance to look at the latest patch? Its already approved by Hudson and its previous version has two s so I hope itll get into trunk soon +Many thanks Tom for the review! Here is the updated version addressing your comments +Changing it to minor +Then we can use this url to pull the logCurrently this url is pointing to a webpage +The default value of isContaintment is true! +Thanks for your efforts on this featurethanks should be exposed as an us create a new issue for the script cleanup as it introduces new API methods in Context +How do we know if this license is acceptable +I didnt get it working with my Safari either but Ill check again. +on the output +I apply the test suite all is ok +In an effort to keep changes to a minimum for now we can remove the synchronized from the unregisterAttempt and in allocate just check to see if it was null when putting it +Here is a fix for this issue +I like it +Verified in ER +Heres a patch which fixes the deprecated configuration keys in the forrest docs +This is an attempt at making individual navigation action sources for the jscookmenu +Nice one Mike +defer to +It is also something I think needed to get done for as it gets rid of deprecations and increases standardization so a bonus point for that +I attached a new patch and will launch a large scale test tomorrow +there is something wrong in the logic checking +Simple text file wouldnt be opened with it. +I havent run this one through the full suite of tests yet +Can Jesper or Jeff add to example datasource descriptors an explanatory comment about usage of local datasources in transactions +We post pone this task after more Solder in the project +Let me know if you feel strong about this +fixed verify on JBT. +Apache Directory Studio version has been released. +And no requiring Vista is not an option +Forge now stands alone +Sounds like at least different issues +Implemented in r +changed to figure am and pm manually and use instead of due to what appears to be a bug in +Roo itself does not handle the ESC key +Bulk close resolved issues +From what I can tell on Windows the URI fileCtempfile URI may not be considered absolute +Ya Im going to blame TB for this one as wellI cant see a path in the code where this could happen if everything is running as it should behind the scenes +should contain the fileid typical aplications like hadoop dfs ls and fsck should be able to display the fileid +Sample to none still shows the same error. +Same exception is also seen when trying to delete a shared network that it is only IP + +Heres a draft patch +If we waited until parsing time we could display a nicer message but I think the eager validation is more important in this someone knows how to bend XSD into submission here that would be great +Maybe this issue is out of scope +If you dont mind Id love to apply the changes once done +We do not take these patches toIm not entirely sure what you mean by this +I looked at examples from seam one more time looks like we cannot change because it doesnt work for booking app if I remove all non module jars from deployment descriptor +sqlrowset is not suppose to be serialized +Ive tried to commit this but I no longer have write permissions on the new repo location +Seems a rather serious bug to me +and i think other multibyte language fonts cant support those tooas you can knowcommon chinese characters GBK include lettersbut english only include +Should we do that in the future +Thanks Andrea +A better workaround which solves some issues is to return +Works well the only non blocker limitation is that the code viewer is always opened on the same tab whatever quality metric is selected in the Treemap +Ill commit this in a moment +It should not hold up a release in anyway +Unfortunatelly I have the same result with brokerId and with either +maye this could also be relevant for other types of generation +Im usually working on windows OS and Ive had failing commits because of characters +Is this still an issue? Did placing the xwork JAR file in lib fix the issue +Good stuff +That only fixes it for a few days though +I already have dozen repositories registered only for Codehaus projects +Just mark the removal of the wrapper as a TODO item after the next lucene update +Thank you a lot for this contributionI have a question what will happen when we publishdeliver such an file? Will we publishdeliver it with the extends element in it or will we publishdeliver the merged result +In which case maybe we should consider moving the metrics to the request rather than the connection? Or have metrics objects for both? needs metrics for the individual method invocations but theres probably a use case for metrics too andor httpclient metricsIn the case of httpclient metrics might be an alternative as one could just look at the differences +I think its clear that weve still got issues to resolve here +png files and attached +I have got a similar problem +cpp files are in your vcxproj +Nothing from me. +In a non scientific mini benchmark the sorting phase takes an order of magnitude shorter time with the patch applied +Hi SebastianDoes this mean the dependency is to be removed from the maven and move with this addition inIf that is right it is not right as for a mavenized project all dependencies are to be in file +All links on referred page lead somewhere. +Indirectly fixed through the selector has been changed back to +Updating introducing for build and to generate the statements. +Ive uploaded the updated patch +Wont be able to do so for a day or two though +New test casess been added but I think that should be rewritten +Thanks Jan. +I resolved the conflicts and backported the fixes to andMarking the issue as resolved. +However having performed several successful releases on other projects in the past I was intrigued why the issue has occurred + +doesnt support polymorphism yet although Im surprised that its failing this way +Explain how it was missed and how it was picked t forget to link to in this issue and the JBPAPP issue when it is confirmed it was related to that +we cant call super in on +In the case of point layers when opening the style editor the color is not the same as the layer points +pushed Released Closing +Want it? +Our HDFS cluster scans disks in parallel when generating block reports +moving off since not critcial and not enough info to reproduce +Fixed to use instead of STDOUT +Sounds fine to me +This has never been about breaking this for people only about cleaning the internals of Cassandra +The NPE appears to be the cause of +Ben how did you end up getting around this? Im having the same problemI tried to target a local page which would just redirects to my data url but it looks like you cant use system on local links either +Yeah its bogus in Groovy Im flipping back and forth between Groovy and Ruby and get re right silly mistake on my part +Please mark as resolved +There is now around attribute value +In the event no environment was set there was a default exec environment +Im waiting for your feedback +New corrected POM attached +Developer studios name is changed to Developer Studio +Manually by integration tests. +All specs pass +I use Maven with the plugin +thanks +Ok Ive been able to reproduce the issue and I narrowed the cause down to the fact that the fragments are being updated on restart +for thatOne nit I prefer the name to +bulk defer of issues to +Attached a patch fixing theDo you mind to review and apply the patch attached +Fixed by released +Viktor please verify this issue +I suggest that you target the fix to the stable branches +Thanks for the update Hendy and sorry for the delay to review your patchesI gonna make a try on by the way +It can only appear when sharing an old project and causes problems if it had absolute paths which do not match to new location of the project +fixed apply patch Harsh Sorry for too many patches +If not then the resultant DTD from this method will have collation derivation of NONE +servlet source +I will have to change other places where the former message was being used +Needs tbd. +Screenshot of the new patch +Uploaded a new patch to fix the findbugs warnings +resubmitting +Those constructors are deprecated with a warning about removal +Writable makes more sense probably +While it is true the increase in the number of commit flows caused by the redundant commits +Reopening this bug as theres obvious dissatisfaction with the outcome +Done in r. +That might negate the overflow worry completely +I think the overhead is not that big and that we should enable the test. + +Thanks AdrianYour patch is in trunk revision Thanks BJ for review andThinking about it more maybe should be called +Apply the patch to vanilla tree +we decided to implement a specific default timeout that a user may override via system properties was corrected with the fix in which allows users to set read timeouts via Eclipse preferences +We have been patching Geoserver for many months now for our deployments and local builds and we would love to get back on a mainline if it still needs work please tell us what needs to be done +Watchdog removed Database table management removed system metrics moved to +Reuploading patch with one corrected +Attaching which I committed at subversion revision +There are number of fields defined with the same value in many of the methods +Thanks for the fix Jeremiah +With a little effort we could add an api to configure that number +Closing old issues +I was not fully updated from trunk. +OK +because mongoSyncs headRevision is incremented for branch commits as wellYou probably mean headRevision is set rightbq +Attached a patch for enabling java asserts by default for tests in common +Debug Enable this boolean property for connection close processing +Verified by Leo. +Hold up a minute +Had to eventually remove the deployment archive and put it again +This should solve your issue +bulk defer of open issues to +Proposed patch attached +It remains that way to ensure users do not need to update Spring +Diffing producttoolsconsolemanagement shows no I miss something here +Lets create a separate issue for Configure Problem Severity label at preference dialogAnd Im resolving this issue +Im broadly in favor of doing this +This issue resulted in wasted time of investigation +Add note to hadoop versions section in refguide +I was trying to come up with realistic examples but the only useful ones I could think of involve tokenizationmaybe but with things like the tokenization sometimes happens after the Tokenizer is done +Supporting both methods is probably the best idea +Resolved along with changes made for closed as resolved and released +Committed in revision +Im rerunning +Thanks a lot Chris +i have +Fixing the failed test +It looks like the composite key part of this issue is also logged as. +I have merged this to. +I will run tests and commit later today +How about batching into entry chunks or something +Added use case of import to HDFS +here is a clip from the file viewed on Red Hat is a clip from viewed on Red Hat I think there may need to be some action here for v +And frankly Im afraid of stress testing on my laptop which would also mean stress testing my years old hard disc +Probably +am closing the issue +If you wish to ensure your code works wirh JDK simply set the property to. +Theyre good enough for release. +You need to address the concerns I noted above +Patch committed rev bdbacecdfeecfb +I am not sure I would include this since this is dependent on knowledge that a constant exists +Checked into and trunk +committed +Dont want you to be labouring under the misapprehension that my workaround above is solving the inevitable question and ETA for a patch for this +In particular I think RF will be Good Enough for of use cases remember that our goal here is to handle the case when the coordinator has sent part of a batch out but not the entire thing +Icons switched to smoother versions thanks +This was intentionally changed right before release for +I think what you have done is a very good starting point +please reopen this issue with your wsdl or a wsdl to show your problem if it still exists. +Please assign this to me +I have not yet implemented the other comments in this ticket and theres still some other clean up to doAlso I havent looked in detail at but need to think about how it impacts this issue +The first rev of this patch passed derbyall +Ill dig +Did you issue a build command first? Note I havent yet fully tested in the Solr container have been more focused on individual unit testsAlso whats the NPE you are getting +pread opens a socket per access +Ive made an attempt to put this in a separate class such that it can be reused +Ping +Yes a hook like postReplay could work +bq +I just committed this +I think that displaying it would be a good idea +So all issues not directly assigned to an AS release are being closed +We really need to get some tests running all the time +We can edit all kinds of files with and you can even temporarily open external files for editing so im sure we can get them edited if we want toThe challenge is where to find a good way to do thisWhere should the quickfix be? On missing dependencies as a Locate Maven repositories on disk +I think I still dont follow your set up +dblook needs to be enhanced to emit DDL for grant statements +newer version that has better javadoc of bound am uploading a file to replace that section of +Perfect Julien +You should try rebuilding the library from the source code +I think that requesting not to create idle consumers is not valid +The failing test is unrelated +I can see abovementioned attributes in taglib but its not in sync with +Patch applied on trunk +Charlie pushed this patch in d +There are some changes to the patch +open so Id rather leave that part alone +Can you attach the application which is failing? Also are you packaging any jar containing +Fix merged to trunk. +I can contribute a patch to hook actively updated databases for example +I have run junit tests locally and all looks fine +Ideally this should be handled by the shared filtering component +looks good + +Ill commit shortly +The handling of the sequenceflow is very generic and in the base class +We did change equals for Groovy so that if you use it from Groovy you will get the desired result +Oy +applied all patches seems fine now. +Attachment has been added with description simple docbook file that demonstrates the P open bugs to P pending further status from ASSIGNED to NEW for consistency +I just fixed the page in the live Wiki +For a first implementation a to get rid of the memory leak we should implement a for documentsWDOT +The other test failure is obvious +Verified with JBT. +Max those parameters are still in osx profile just not in systemProperties property but in platform +Its like those Com Interfaces with at the end or Pont Neuf in Paris +Does your test case work in that combination with no JOTM involvedJuergen +Jason if you keep the same name for the patch JIRA will automatically gray out the old ones +Updated on latest trunk +Other checked exceptions are transformed to I wasnt able to remove instanceof checks completely I have to use them when examining the wrapped by the +thanks. +Im fairly sure were not dependent on any features that might use drools but I cant remember for sure +in +Done +Now resolved by linked issues +By when do wish to have this? I can start with this and apply the changes +Some people suggested YAJSW as an alternative with friendlier license upgrade beyong is not commercial friendly +On the grounds that it looks good looks like it refactors and imitates similar code has tests has had a look from Ted going to commit it with some style tweaks. +Committed to with some minor changes. +Nice docIll commit this tomorrow unless objection +Just for clarification +I agree that the client should continue to permit both interpretations for maximum compatibility +what should the error message then be? +The NPE is caused by the extended class path defaulting to null instead of an empty array +Dont know why they did not show up in the patch diffsrcmainjavaorgapachestanbolenhancernlpjsonvaluetypeimplnlpsrcmainjavaorgapachestanbolenhancernlpdependency +Logged In YES useridThe class is being loaded by the UCL associated with the conf deployment and the deploy due to the classpath refs +Why would you not have the build version or at least the version specified in the file? For each version of hibernate you need to update so why not? +Note to self remember to issue a pull when github is fixed +Code that pasted in issue but seems to have lost +Agreed + +Extracted out all messages to properties file +I am just collecting ideas +Im not sure if this issue is still scheduled to be fixed but I have a mock jar ready for testing +Sounds good to me +Logged In YES useridIs there any workaround for this? Is there a configurationfor ignoring the indexing mechanism +Does the tests still pass with this changeWill this only apply for webbeans stuff or any EL related compleiton +forCould you check with a current trunk build +would appreciate if you could validate the efficacy of this lease based approach in your environment using the latest +Working on the rest +Updated to use nebula instead of +committed trunk and x +The bug Nick linked to has long been fixed so I believe this must also be solved +Please try with that version and reopen if it still occurs. +on the new patch +will commit if the tests +Click on the Web service runtime link and change it to Click on the Service project link if you want to change the default project name Click on FinishIt should create a new couple of new projects a dynamic web project and an EAR project +But shouldnt there be a way to change the default block size outside HFOF context? Or you have to override each family separately? +So the script should not be forced to be running as root +this contains changes to tests as well +This means that all of the cross site scripting attacks based on bad urls should be fixedIll file a follow up jira to fix the vector where the values in the job need to be quoted +Sorry I cut off the Headline +Fixed on and trunk +Looks like the patch was not applied to either +Patrick Im going to check this in minus the printlns in the new test +oops +I believe I can merge it during one of the next weekends after the release is out and give it a spin before actually committing it +CLASSPATH true Main Will run with rows +I have no objection +Please let me know if you get a chance to try it and if it works for you +Good catch +As php handles the multipartform logic the container can easily access the uploaded files via the FILES global variable +Jared thank you for volunteering to fix this +Michajlo integrated in trunk and +this was fixed in on Developer StudioVersion Build id Build date +Attachment has been added with description A unified diff against transition resolvedfixed bugs to closedfixed +Hopefully this one wont be rejected. +For a native transaction the provider will have to hold a Connection per transaction eventually but it may fetch it lazily on first access and it may release it right after transaction commit +This allows JSP pages which can be scripted as you desire to include Velocity fragments +Problem reproduced investigating +Whats wrong with defining different actions for different methods +Also added a test for this condition in the junit test +Please note that it is mandatory to have this done through a file containing the bytecode setup as setting the property directly on the bean in the Spring configuration does not seem to work +If we add resources like disk and network IO this will become much less necessary but I still think a high number of containers will put load on a system in ways that we wont account for +Done. +But this does not address moving mapred classes from the javadoc does itI think Id prefer issues that more directly address these goals or to more precisely state the goals and benefits of this issue +The issue may be seen in the bugtest project +Next build will contain the fix +Some of the helper methods need to be fine tuned +We can have separate jira issues to deal with CF deletes column deletes version deletes etc +See the user guide for info +Committed revision to trunk +Fixed in CXF which M will be based uponcxf is still a default alias but can be overridden by properties into etc directoryand property to some custom value this is not available then service can be used to configure cxf with +Here up side down and is faster but order of duration for clone operation remains the same +Well be glad to accept patch for this +Perhaps this could be done in cxf +Also this is easily implementable for containers alreadyActually its more a clarification because Extensions are defined as already +I think I will start from scratch +Yeah one thought on this Isnt it more likely that someone doing requests over a CONNECT tunnel is doing HTTPS? If so maybe reverse the logic such that if port is you guess its HTTP for all other ports assume HTTPS +models are already a fact of life +Updated description with results from EAP ER +For CM action and policy classes I think we can make them public +We need to add a third choice ignore +Bulk closing for +This is the latest patch that has fixes for both JIRA and JIRA to integrate struts and JSF tests into the test distribution +Adjusting affectstarget versions appropriately +Defer to C hosted +Version +Yes we agree that the test should display values when the assertion fails +The could you please attach java file as wellThanks and the found that it will raise that exception when a a bean includes a Collection propertyEric +The idea of is to hide the detailsWhat do you think of introcucing an and extending? could extend and then anything we need would be put in there +Not every protocol pass your so I cant really commit itYoud be keen to have a look on it again? Ill be committing your PHP clean up and reissuing the patch shortlyCheersHenrique +No currently it does not signal if suggestions are not found it just returns the query terms themselves +I believe anyone has privileges sufficient to create a new page for the connectors wiki project +is not actual since we fixed +Let me know if incorrect +Removing a user from m wondering about strenght of the test +Findbugs warnings are +Slightly tweaked patch this one wises up on the fact that advanced networks can actually support security groups +agree with Adam here lets do minx maxx miny maxy. +This project is a pure Spring project +Where are we with this? Should I remove this from +Thanks very much Ivan +Files are gone from ejb unneeded stuff is gone from. +For completeness to expand on the parent transaction comment The reason participant records are currently separate from the transaction record is to support cases where the participant does not have a local parent transaction which can happen when running the TM outside an app server and we are not using interposition +Why does async play a part if async marshalling is set to false? Investigate. +Hmmm is correct though it has milliseconds +Reviewed by Robbie +This only happens if Equinox is set as the in the etc file +Their support should be explicitly enabled in the kernel and also supported by the file system type right? If yes doesnt this approach make the utility of this feature much more restrictiveIt seems like option is a slightly better option in that case +BTW as I mentioned before if there is another way to intelligently figure out what hostname to present to the user without it Id truly love to see it +Thanks a lotSure I will add some doc but I would like to ensure first that I am on the right trackI would like to avoid having to update code AND doc in case there are a lot of things to changeIn the mean time I can provide an example configuration if it helps +I have the same environment Mahout and IBM JDKBut I do not know about HEAD +I wasnt sure where I missed the private comments but I think I did everything else +There could be a jaxrs system test verifying it but mentioning it just in case +Deferring to at least +I am this issue because it is not a complete fix +Please verify the patch thanks +Issue has been resolved for a long time waiting for potential confirmation from the reporter +This is the standard way we host the banners +Bulk move issues to and +Or plan B we commit these changes only on trunk and we say that support only starting from and +I create a vm +Good on you Jon +Hey HariI fixed the typos thanks for pointing them outFor the logging stuff what is the simplest possible file that will correctly output to console in this case? I want to make this example as straightforward as possible so even people who dont know logj well yet can use it +patch for trunk that applies on top of the one for +WonderfulYour results confirm all my expectations about this patchIt looks great to meI would keep your cache patch in mind and be sure that it will be easy enough to share one cache even with two but otherwise I have zero concerns +This is currently a copy of the bundle and feature +More complex operations cannot be done efficiently and should be done client side if they are done at all +Re running all tutorials will be necessary anyway so its better to avoid wasting time updating them at each modificationFor the english grammar I guess I dont have to say I wont be able to review itFor priorities about propertiesvariables I second your decision its how it was working beforeNow to apply the patch Maarten youre the assignee of this issue +This is very strange it indicates either that streaming didnt work at all or that the task wasnt run in the right working directory +Will commit in a day or reproducing on my Mac with Java and ant but no successAny clue on what environment I need to test at or what in the test case is the problem +Daniel if we add the defaultMessage to the getText call as you have it then the getText call will return it when key is not found +If we do not care about which generator is being used actually increment should be the preferred generator to use in the tests +Fixed in as well. +This could happen if the user reverts to an older version of the metadata table and newer tablets that were created are not referenced +DanI agree and as its not a big deal I put it in the main task of the main +Some comments requeueCount in FQE could be a boolean thats how its used +And for our project examples the latest are the older examples are I will fix this problem +Theres simplistic support going for this so the sample works nowThis only works for the helloworld sample right now as the service namespace is hardcoded in +From what I understand it looks pretty good +Could you take care of this as well? Thanks +The configuration can also be finished through a installation guide like some commercial software +Changed binary file in srctest applied the patch +For now we are reverting back to distributing Xerces +Committed in rHowever due to on Mac I cant verify +No new tests are added +was added for performanceconcurrent reasons but breaks JDK support +The logic would probably be simpler +Seam Validation seems to work out of the box if you do not use Seam catch and implement your own +The jira was for stale output left behind as the state machines did not handle the necessary state changes to allow for all verticestasks to finishcleanupabort properly +Addressed Teds comments +Works fine +It looks good and I belive it resolves the has a regression for test build target for the flavorPatch needs to be reworked +The document details the proposal to modify the Stonehenge stocktrader application to use claims based security +Old names is the tradeoff we have with strong m to the original improvements in this ticket +And Id like to do the initial of the base directory as outlined here first +Attached test case logs +s commit this patch as is and work on further improvements in a new jira +why can you reconfigure skip configuration per execution +bulk close of resolved issues. +I changed my paths slightly for now to work around the issue. +Found and fixed the issue in +However from the framework standpoint either read returns successfully or it does not +You are right the problem is only with +Thanks Al totally makes sense +Im uploading an archive with straight mappings now +Please let me know what you want to do about the testcode +Ran this from the command line and the problem seemed to be a null miterators +Made changes to to introduce a new constructor with credentials as parameters +Deployment properly added +If a user has to go to the logs in an expected scenario weve failed +Off topic but why would this give a NPE? Is that some other rare bug thatyouve stumbled upon +Had a quick look and this seems nice and straight forward to me +Patch applied. +Thanks +Patch applied with very minor mods +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +the patch looks good to me +Its nice storing the info in the metadata table because you can atomically update the timestamp on minor compaction and bulkimport by putting it in the same mutation +Ill close this now as it seams to be working now +forgot to to projects as compressed archives is the new in +Another option Id like to put forth would be to separate the reading of edit log ops off disk from the actual application of those ops to the FS state +For some reason this fails unit tests both on zones and on my workstation +Committed this to trunk and +We might want to change this later on when we do runtime concurrency calculations +A tiny project with a runnable class to demonstrate the issue +thanks Rakesh for fixing the tests +Hey Aaron sorry for taking this long before responding +Commit made +If you say it is unnecessary i am willing to go with will also point out that we could kill the custom look up all together and just use the property resolver +I have not been able to reproduce it so any information that could help reproduce it would be helpful +looks Eric I just committed this patch to trunk and upon release of +Ill take a look at this today against branch and upstream and update this JIRA. +Same is true for Obtains Instance +I dont know what would set the response encoding and the classic convention is to reply with the encoding of the request +DO NOT foget to set correct fix version when you resolve issues + is the output of gdb where please assign this back to me when the fix has been applied to the branch for the EAP well build it for + +Harry I think the current sandbox instance has just run into the error Ive described on the list under weird access rights issue +Because app client facet is a WTP facet which is declared in I think we can do nothing to add a event handler for this +If I understand it correctly the reason for this custom URL class is that the user cant be bothered escaping his paths +Applied thanks to be OK closing the bug report +Having experienced the pain of the migration Im not thrilled about the idea of figuring out how to get all my data into a new format +Should we this or open a new issue? +etc were committed in revision ecfdbdeeacdefacfebad +Uploading a patch that works for me +Is this a regression from +So in the service we should throw a dedicated exception to indicate something like the ressource does not exist or the resource is missing +I think the test should be modified to add data in the same order in each case which will ensure equals semantics works +After resolution of it will be possible to call custom code before panel activation before panel validation and after panel validation +Upgraded resources plugin in rev +I will file seperate Jira for it +Closing all tickets +Issue still applies to XercesSame changes same files only line numbers changed +This has been properly fixed on trunk and branch +Here is a patch which fixes the remaining master files in the old harness +Committed at subversion revision Touches the following fileM javatestingorgapachederbyTestingfunctionTeststestslang +abcd +Even if the class in question has +closing this is mostly done +I committed the patch to +Replace your and with the ones that come with M and are contained in GRAILSHOMEsrcgrailsant +Resolving as invalid +As you might be aware threads is removed from mostly because of bugs like this one +Deleted references to Unitils since this issue focuses on for the requested hooks to allow Unitils to integrate directly in the Framework. +push out to was resolved at jpl and i never closed issue +Slava see the screenshot +Verified in JBT Linux +test and patch for timeops +Added security to component has been fixed by Dan by taking out the ws security token as the example is not intended to illustrate that functionality +For example line of the patch looks too indentedThere is still trailing whitespaceThe this should be done for all fields probably +It was caused by older libs that were loaded by Glassfish and shadowed the new ones +I tested my patch against clusters at openplaces one with dns problems and one with no problem at all +Committed change. +I have the same issue that prepared statements never gets cached +Rebased on to trunkEven though this bug affects earlier versions its not a major issue and the concat feature is rarely used +bq +Thanks for committing the fix +Updated Wiki text +Rather than using lastIndex you use indexOf repeatedly until youve reached the end +Im getting this error also on and Kivialho is not supported any longer +The Session class has one thread for the connection and blocks inside the run method trying to read bytes from the server +status changes or b +Heres a patch for the binhbase and conf scripts and conf adding placeholder configurations to ease the process of enabling JMX monitoring +We prefer to code in groovy than java but we miss checks when we are not using dynamic language features such as method missing +HelloWould you please try the patch? Thanks a lotBest regards +patch applied +Am I correct +Please add a test +You cant do the binding to javamodule via the CLI it needs to be done via or +Will have to think about it +As far as my research shows there is no way yet to upload multiple files in one go without resorting to browser plugins + +There are still some dependencies left in files +Closing issues before next release. +Is there perhaps anything i could contribute in this matter? I really need this resolved asapthxMaurice MullensVerstuurd vanaf mijn iPhone +Dynamically install Filter implementations provided by a Coprocessor via so Get and Scan deserialization will work afterward +This cannot be right is it? Ill have a look at the test so Ill be able to produce an error state +sorry just saw that you already uploaded a new patchwill start running tests after +Heres a patch that will include the console configs into the build and build them successfully +The hard part is to detectionHow do you identify a slow nodeThe dfs may be able to detect slow disks by collecting the disk readswrites performance data or by using tools like iostatIt may use some similar tools to detect slow network linksThis kind of data canshould also be used for blacklisting task trackers +Also I can find no config test or other conditions for which RWSTDNOTEMPLATEONRETURNTYPE is defined. +Im closing this since it was fixed in the CVS tree. +Thanks a lot. +I think that as of r this should now work but it should probably have a code review before being deployed just to be sure +Ive applied this to the trunk under revision r with a small change to use a finally block instead of copying the policy handler invocation code +If you are still seeing the problem please submit a runnable example that demonstrates the issue +for example what if someone supplies a term of length and asks for distance ofwe should not use this algorithm for nonsense like that in that case I think they should just use the garbage algorithmHere is a quote from that moman pageIt means that in theory you could ask for a Levenshtein distance of ! Well if you have a week ahead of you +Can you rename and to be consistent with other JHS configs and move it to +jars over the HTTPS protocol as well? If so I think we should also test for https as well +The administrator can change the entries in the hadoop config file +Moving to +were referenced by deprecated methods only +I downloaded uDig and the port field in the wizard pagestill accepts only four characters +I will submit this soon I will check it in with wording similar to yoursStatement too complex +I wonder if this would address the problem in +Added new test that pulls in generated configs from the quickstarts and validates the EMF model against them +is in not ivy documentation +Once CLI was released the importance of the issue was lowered. +Uplodaing log file +I cannot think of a way to make the title more clear +Well its not so clear +Thanks for the image now I know Id understand the issue but I couldnt believe itCan you test you page with the Speyside theme please? There are many changes in the css classes for. +The number of maps requested is a hint to the nothing else +yet why does this block did you mean to link insteadDuh wrong issue I fixed it +But how? What did you do or change +It would be great to be able to select based on groups to limit the amount of Value Objects in the build +I fixed all the checkstyle errors except for one that I dont care about and will probably make the code more confusing if I fix it + +Please open a new issue if this is still a problem with a sample have the same problem with and +Are you sure its fixed in +after release +fixed +We move the packages transport and to the new component +After the changes in I had to do minor tweaks to allow thread manager functions to work with dead threads +I am planing to solve it this week +Now that work is well underway with AS all previous community releases are +There might be some updates I need to do for the docs though. +Change to reported javadoc errors. +Changed a comment to to say that the libhdfs unit test is allowing hdfs appends +Comments on reviewboard +KeithIm using the EL the SWF docs imply that only the and OGNL EL variants are available and that the Spring EL wont be along until +I will search for help in the forum. +Committed patch to documentation trunk at revision +To put it clearly the problem I was facing wasnt the same issue Tony is facing +I restarted the machine and problem went away +A user could submit a job to any queue but if are enabled and the user is not allowed to submit jobs to the queue their job is rejected +Fixing the warnings +And seems the issue is in the logic of detecting the wrapper class +On Spring youd need to avoid the namespace and fall back to traditional Spring bean definitions Define the as a regular bean using its sql property and set that populator bean on the databasePopulator property of your +If everybody agrees then we can close it as. +HiYou misspelled it getcapabilties +Id like to take a hack at this +delete this post Please remove our site from mirrors +didnt mean the exact File API as that would definitely be very inconvenient +Looks like Tracy will come up with something what will make everybody happy +I think the NPE should be fixed upstream +bulk move of all unresolved issues from to +So I was looking at this over the weekend and just want to confirm the specific files that need their javadoc comments updatingtrunkapitestjavajavaxajdotrunkapitestjavajavaxjdoWas that all +We are only worried about snapshot views +Moving to page contents to make this message to Edit Configuration Properties page to indicate hitting the OK button will start of services +Patch required for HAR stuff to compilework to be included in the lib directory +I committed this +We have a working now with WS and without EJB +I am actively working on this one +already have such machanism by keeping states +Werner Thanks for your the newest codebase will the code generator produce a set of classes that compile +NN closes the file and frees the lease +Plan to commit this tomorrow +so hopefully you can look this over to fix it +We had some discussions on this in the mailing list beforeI agree it would be nice if we can make the release backward compatible +Committed revision +Unfortunately I have not been able to figure out a way for a Python script to capture what is writted to stdout when it is started as a background process +Id suggest adding JSR unit tests for JSRPFD branchPlease see also +Syncing patch with trunk should I perhaps file this as a separate issue? +see screenshot notice it is on the same project +Khai DoPlease open a new bug for the regression. +Im just not sure if the is so common thatwe need to support it directly +This new patch adds constants for the strings and the unit tests now actually verify that the output contains actual an actual keyval +It was applied in branch and merged into trunkPlease cross check and close the bug +Thanks Rick! The change looks for good for +In looking that the data type conversion routines the other day it seemsthat there are some strange conversions tofrom BOOLEAN allowed +It will cause a JVM crash +Patch created against the trunk +The queue isnt always set in MR +The pipeline calls flowscript which calls another pipeline at the endI am reopening the issue although I cannot currently test with +I must have clobbered something in my credentials file +That would be simple to add and would enable the feature desired here and a lot of other things +Ben Dave I wonder if we could improve situation with Eclipse project configuration in Maven tools +need to research it a bit +This change does absolutely no good for future extensibility or flexibility as none of the functions actually implement the logic to support this functionality. +Uploaded a for this issue please have a look +It is a matter of changing few lines in my bot tests +What do I do then to fix this +Thanks Ill try that +might have to do with private registration +Fix at Revision Dependency callbacks support two new signatures service object dictionary service object map Those signatures allow injecting service properties without injecting the service reference +Its not clear that this is necessary any more +I would just perhaps note in a comment that they are not intended for user extension because we do not want to be locked into any internal back compat for those classes at this point I think +Fixed analog guess and directly depend on this issue to be fixed +took like weeks to knock that out +A few problems public and protected methods on public classes need javadoc needs javadoc inReader and errReader should be closed in a finally clause theres an empty else clause that should be deleted +Have you spoken with your manager about the expense? We dont have a budget to supply shirts for projects +New proper patch +Any updates here +Anyone can post an associated Oracle BUGDB Bug for Weblogic? Is the resolution merely listing of the workarounds or something was actually done? +Also forget to note that I think it could be convenient to have only one patch for layer with both HDFS and MAPREDUCE +Fixed in revision +fixed ve just committed this +Im open to the contribution as long as it doesnt break other code +with a preference for the formerMy main concern with the latter is indexing +I got a second ideal on this to help with hot spotsIf we could add a way to set a priority for compactions this would help with the hot spots regions building up to many map files flushesExample if we have a region with map files and one with Region with map files would have a priority of and the one with map files have a priority of we would compact the region with before If we could addupdate the priority when we do a flush then the compactor could work on region that need it the most in order +What exactly did you remove in the files? Because activiti can cope with processes without any BPMNDI information +Currently I am out of the office and will be back on Monday th of January +should also be fixed in I also fixed this issue in +fixed the issue I think you get this problem with using with the uw option +Related to the change to the in +I never thought of trying that thanks +I opened. +There is supposed to be queuing in the DNS processor but that may not be working as intended +This appears to me as coupling and coupling is bad +updated in Configuring the Registry and Registry Configuration Examples +Actually Im not sure there are any confirmation messages in Continuum +thank you very much +Thanks Arun! +Can we move this out to? If so please update the bug +At this time Macintosh did not even know unicode +Can you please investigate this? +RHEL +GuillaumeIt seems you have a good grasp on the issue unit test +It works for on Wink +For example if the framework is shutting down and its start level is at when the shutdown begins +So Id like to close this issue. +coolAshish Ashish for sharing the link from there the Order Status can be specified for product store +Thanks Jane +Theyre both suggestions that I wont insist on +Also stream lined the loading of VDB during the initialization of the embedded configuration service similar to that of adding VDB through admin api so that there is only one way to load a VDB. +It might actually help in high frequency since names dont occur over and over again +Overlay is failing in windows +IMO it makes sense to move this tracking into the scheduler and in particular looks like a good place to have this logic +Otherwise I dont know and Im afraid I dont have time to look at it in the near future sorry +Then it wont be regarded as a miss and Camel will try this converter combination later on +what functionallity is missing? +but I was trying mvn stagecopy and ziping the artifacts to try a manual upload and I didnt succeed +This probably isnt an issue anymore +Thanks! +Hi SergeyThanks a lot it sounds reasonable +dodeploy is removed first then a +Closing the task as the dependent tasks that worth pursue have been committed +BrentI am not entirely sure what to do about this one +Ah I only saw a spez +Unusable means jar corrupt or about problem? the central repo has to be self contained the plugin cant be releasing depending on stuff in other that this issue should be WONTFIX request release +I checked etcsshdconfig and TCP forwarding is allowed +Patch to fix the issue the earlier code was not catching the exception which hadoop throws when it tries to access a unauthorized directory +Otherwise all great +Fixe +Unfortunately created screencast doesnt show this issue because its just a looks like Edit dialog is displayed with wrong size and position for a moment and redisplayed in the middle of display with correct size +for ptPTinvalidO cdigo postal contm caracteres invlidoswrongO cdigo postal tem um tamanho invlidowrongO cdigo postal tem um formato invlidoincorrectA formatao do cdigo postal invlido +Patch committed thanks Ankit! +Checking documentation is an implcit part of the release process and so this task will be covered by that. +I was thinking more of detecting bad configurations and bailing out early with an error message that could be searched on + +I think the fix for may well have fixed it so that the code works OK your broken serverPlease try and let me know +Looks like Im trying copy to itself +Applies cleanly to tag +This issue has been fixed and released as part of release +Thanks Stack Matteo for review and commiting. +Can you please check again and give some feedback +Trove is GPL and cannot be used or copied +the for default values has been added in Spring with introduction of so I guess this issue can be closed. +test passed +One request can you attach the test schema you are using? I can write a simple unit test to check out the problem then +Im wondering if this issue is somehow related to. +A set of will be delivered for in the CR cycle. +You are right that the ant dependency is a show stopperIm parking this one as LATER +specfile now exists for the project +Applied above comments in some of the exceptions mentioned above are intended to be swallowed +Sorry +So mesages are left in the queue +Basic framework for http authentication and Basic auth implementation is now in place +For howto you can refer to sampleThanks +Im not really sure what the problem is here +Assume closed as resolved and released +Thanks Eric. +The problem was that at regisering xml declarations all java declaration with class name not matching that referenced in xml were removed from component +ChunhuiPatch makes sense +If we want to do this for Beta its easy enough +If the NN is able to complete the replication of all the blocks of the decommissioning DN to other participating then that should satisfy the decommission operation +organize all image in folders corresponding to chapters names +This causes the socket to remain which causes it to return on eof rather than block waiting for attached patch fixes all places I could find where we configure blocking and dont set it back to what it was previously +Patch commited with revision +If someone else wants to work on this please go ahead +The same goes for the base directory in the zipAccording to the guidelines linked in the summary the community release should beDimitris is there anything remaining to do on this issue +tests that HDFS works over +Holding back to TBD EAP since Infinispan TP is not being included until GA and has been closed due to inactivity +Will raise a C Broker JIRA the broker should not accept connections while playing catch up. +Thanks Vikas for valuable suggestions in my patch +This means that the include and exclude tabs will both have two fields eachMaybe the fields should support more advanced filtering rules for instance filtering based on mime types? This will make it easier to filter out video files without having to define all kinds of video files extensions +Username jappe +lets decide on a serialization format to use within ZK that is NOT hadoop writables +We need to switch the paging reader to use a shared transactional em and make sure the transaction commits dont clear the session +Thanks for the comprehensive think parallel startup of Spring is very important in my application spring is taking up of my startup time reducing that would be very helpful during development a major time saver +Its possible to share the same for all called Actions +That is new caches added programatically will use its settings +Colm +Perhaps with an extra room like ? WDYT +Ive checked that it now runs correctly +For now I will log those unassigned znode with target rs being killed +Hi Here i have attached patch for this please review and commit +Ill create a separate issue for parse to deal negative integer and real values +If you want I can describe it in detail the next candidate to look at is simplecache +just sayn +In fact the solution is easier than to pass through +Paul please try the sample and close the issue +What do you think about this idea +I have started to add this into my local working copy and will getit into CVS soonI was awaiting the result of the vote to enable validation by default + +You guys should either bring the project up to speed so it works somewhat or remove it from the internet +This is a really useful feature and would be great to have this out for users sooner rather than later +Error looks similar to the one on +If you add a field you read in a thrift object generated from before you added the field you need some way of knowing that it wasnt actually read in. +thanks +Is there a roadmap for this? any update or fix release set? thanks +DevarajI have upmerged the patch to and on a secure clusterShould I go ahead and upload the umperged patchThanks to latest branch +If you dont have comments this is my final patch +Oh minutes is a lot more than needed they run locally in secondsSorry I dont see timestamps on Hudson +But with or without the patch the numbers were all over the place with large std devs + patch to implement referencing of XSD by can still validate it if they have the XSD +Right now Im adding it to both WAR and EAR +In samples we are using inherit from on is declared the cant integrate Flex Mojo if meclipse does not support extension plugins +Close all resolved issues for Engine release. +However these examples will be removed for security reasons +I just committed this +Any workaround for the short term would be useful if anyone knows what the issue is +I did not suggest that since Im not that knowledgeable regarding the purpose behind ignorethread +Committed in both master and +Resubmitting v of patch +Did the maven wsdlcode plugin work for youAs that is where I was having the issue and raised the bug aboutRegarding the command line which version where you using and also which version of ant and the ant tool did you use +Thanks for the explanation. +Thanks. +One of the params would of course need to be the name of the filter class that it would contribute to any chain +all major issues resolved apart from starting pages which are reflected in all issues except one that is tracked in a separate JIRA are resolved lets close this issue. +Will do more tests before committing but the code is reviewable now +I have attached a new patch +Commit applied as of rThanks for submitting as a Jira Lee +It is now possible to set a flag on a so that it will produce named arguments in the need some tests for this would also be nice to determine situations where the user would prefer seeing named arguments instead of regular arguments +ear and related jars to the +I just tried this patch and the patch for does not work when applied to the from Solr +Should we change these ids? We will finally use eclipse bpel so I think we dont need to update these ids for now +After talking with Julian we moved this issue to CP +Jim is trying to fix the breakage +I posted it to get initial feedback +bulk close of issues as part of making release notes. +If you are still without a functioning Apache account please attach a second patch and Ill commit that too +Without that debugging pig is difficult +Screenshot of jconsole monitoring Perm Gen pool until it fills up to +This test also NTS and the different IWRH for that so its pretty fragileWent ahead and deleted it per your suggestion. +Attached as this to trunk and +Not sure what the problem is either right now +Content is not right yet of course +My HTML is MB big and loading and scrolling is a pain +Thanks for the bug report Olivier. +Ill try it +They appeared in +with mysql +Since we dont use Relax NG I excluded package from relevant POM files +issues resolved in released versions +Could not reproduce verified on Developer StudioVersion Build id Build date still occurs when selecting an existing to since I could not reproduce it while testing for did not happen for a long time +And adds and +Updating and letting Lisa update actually think it was Notabug +In both trunk and +Strange things now happen in console in this case +I am trying to reproduce in a runtime workbench but there are operations that causes Eclipse to crash and close itself +Hi Im Pulasthi a final year undergraduate at University of Moratuwa +Bulk move from to to to +Not sure why it complained here +Ive just started reading your detailed writeups they are very helpful thanksI think it would be nice to spell out the word Window in the code so wed have not etc +Ill remove my but stick by a +It does not require a jira or another patch +No feedback from user on results of using the workaround. +has steps to execute the testcaseThanksRavi +That was irritating to no end because for every bug fix in the service we had to update the clientsWhat actually happening though is that the old client starts working with the newly changed service after +if we start having versions then they have to read the correct version from Hive to interact with the data +changed the default permission back to false for backwards like this was fixed by Brian Fox and Robert Scholte +Committed to and trunk. +We need to figure out a way to mock out Kerberos +What about you Ram +These blockers and critical issues are resolved but not verified +Thanks Todd +Thanks for reporting this +Fixed on and trunk +bq +Sample project +Removing for unassigned have seen this with on vmwarelinux platform with ibm and ibm +I will take a look at this one +Moving this up to since it is getting some attention right now Noel please feel free to adjust the target release if this isnt realistic +If I open form builder as a standalone app in browser then its working fine and even reading processes created in guvnors do let me know if this bug also existed in. +All tests pass except these two due toant Dtestcase Dqfile testant Dtestcase Dqfile test +Thanks for the patch Ning +the first patch is really tiny and I am hoping to update more info soon +for addendum +Then each language reads the files written by all of the others +Any tips to which project I can add this functionality best +The excluded jars should not get pulled in during the tests +Closed after release. +Mauro if you can test against master and confirm this is still broken please this. +This issue is not about trying to build community using platform artefacts it is about the platform artefacts getting in the way of community builds due to invalid artefacts under conflicting with other buildsAs part of the productisation process you are required to deliver supportable platforms this issue is directly affecting the supportability of the platformsDo not close this issue without product management and TDM approval +The unit test failure associated with the Generics removal patch does not appear to be related by the way +If patch works and patch does not it also helps us narrow down the problem +Variables in VTL code cannot be set to null +We found our problem though +Over the course of the years we had numerous requests that we should try and use the most specific class name we could that wouldnt cause compilation problems +i agree with Eds point about how difficult it is to figure out hadoop config variables now and hadoop is worse off for it +Fix description received thanks Brian Strong! +file +All checkbox labels are now in spanish. +Remove the FIXME from the testsuites to enable the associated test + +The parser should probably handle class as well +Also Im not sure of a couple of test cases so I will be looking more even if they pass this time +I will write corresponding on +Patch committed in r on trunk and r onThanks Andrea +I think ive just fixed the worst one +We may need to replace this test data if details are not forthcoming +This issue has multiple streams in its fix version +We could be sanity check the priority on the server side +The one you took which is closer to Unix seems better +Created to track the integration with . +Its simple operation add record to the movingNodes in onChange handler for moveAbout +wontfix +That signals that the connection could not be established which is not true +I changed Olivers approach a bit so we dont need new methods inIm not quite sure we really need all the variants of register it seems like we always have things like the core name whether the core is transient all that stuff already in the we pass around except maybe in the case of renaming the core +Successfully passes tests and unpacking scenario +I just committed this +Meanwhile can you confirm there are no new test cases that can clearly reproduces the problem +I think most people are not searching by the Later status +Also all connection retries should be logged +Why are the client and server from the test dir being deleted? Are they unused there +Nice catch Martin a fix is on its tested EDG with Infinispan the problem I had not tested with the async method version +uh sorry this is about constraint lubs + Split into and classes +Ive tested with and the popup appeared +It does exist in clusters in testing but hasnt shown up because nobody other than me has tried to shut down an FS instance while the JT is still liveProposed actions this issue to be more explicit JT must ask for a new FS instance and close it when terminated +I agree that its best to remove the description of the exact error message from the paragraph +It depends on your environment deployment might be able to replace the with a simple and a +Ill check what else could have caused the problem. +Another patch putting text in +bq +I think this must be thought after is currently no pressure anymore since we already fixed javassist and now the mem leaks are gone +Are both of your setups vmWare? If so can you please pass your setup information? I will get some vmWare engineer to look at it +So we should manual refresh portal after change skin in to affect +The is defined in the release profileIt should be global +Ive committed this with the image displaying to the right of the title +I am not going to slate this for because I think its too risky as a patch. +I have a start on this issue for Linux +Its already implemented See fix the version to one without x and add reference to issue it duplicates +For now we can pack so Pig can run +Whereas in existing serialization it failed with exit code If I change the exit code in the test case passesDo you see any issue with thisThe diff of my proposed changesdiff git aqlsrctestresultsclientnegative bqlsrctestresultsclientnegativeExecution failed with exit status Execution failed with exit status FAILED Execution Error return code from FAILED Execution Error return code from IIRC that will cause some failures in the cli tests but its worth a shot +I will fixed it during implementing other norm methods +I make this blocker only because I want to ensure it gets in the next release not because the issue is applied thanks for providing it +If so Ill commit them +Make default to and well check for that. +Thats causing the Exception +I see what you mean about the RFC now pity it does not have any examples for PWD responsesLooks like the code change was added in the fix forThe fix should be simple for conforming serversStrip enclosing quotes and any embedded quotesIf the reply is not enclosed in quotes then it should probably be returned as is +Mail me or dropbox it +fixed with the new UI +Bulk resolved close state issue transition +This seems possible +Ill at least move it to +Can you test this for me? +There doesnt appear to be a swaptotal graph showing up +Attached patch +The temp file should include the name of the class the method instead of the JIRA +Its a pleasure to be able to help Ask Jayway. +What should be in that file? And how is it created? Does sourcegen create itI dont see the or the in the Castor Javadocs +Took awhile for me to get back to this but yeslatest fixes my problem +The previous one didnt get the namespace right for exceptions that where included from other files + its a great contribution +is a work in progress posting what I have done so far for review +Move comment in javadrdaorgapachederbyimpldrda about client driver implementation to javaclientorgapachederbyclientnet +Fixed +Our integration tests cover this feature as we able to parse same set of files. +Sorry for the trouble and thanks for the help. +Would be cool to enable this for the next alpha that is included with Wildfly +OK cool +No time left today for me to verify +libclasses elements are working now +You need to Apply the changes close the Preferences and open it again to see the added runtime in the table +The version for and +Its not as broken as the current code butit still doesnt fix the word ordering problems so should be left until a fullfix is available +The trigger for this is my having the following line in srctestDEBUGDoes anyone else see the same problem if they add this to their log properties file +Ive just committed itThanks. +It works the same as with Tomcat +sh script and later as hudson job +This was added for fix further feedback assuming fixed. +Check that the application server and JIRA have permissions towrite to usrlocaltomcatattachmentsDERBY now also happens to +Any news on a possible fix? Its been weeks now without an update +trunk +This should be fixed in Spring already +Most likely your select request handler has df pointing to text. +Performance is important and ideally folks shouldnt have to change their applications to see performance improvements +Or like move the zoom inout to the top +Will investigate further if it crops up again. +The two paths I see are either to leave the duplicate code as it is write duplicate test cases based off of the current receivers and watch the current receivers for changes when they occur or alternatively create a set method or similair solution +This would give more insights when comparing various schemes on batching sync callsminor nit long numIterations long numIterations I dont think we need this +Canceling patch to discuss comments. +But it will have to wait some days as I am currently ill and I surely will not do any bytecode things then +can you try to code gen with a nightly build and see +It would be good if you could give it a try +good catch +This is weird +Manually tested! +Thanks for confirming! +Yes +See how it already makes reference to making hbase work on windows? Can you edit that so it points to your article instead and roll a new patch with and these changes includedThanks for the good stuff +Trunk addendum to all branches. +Utility class to process generic class for +For authority connections there is currently an implicit all mapping connections must run first constraint in place but that could be replaced by an explicit list which could help throughput and parallelism +No need to abort if a new pipeline can be established +I dont particularly care about the Windows compatibility one way or another but it seems to me like if a +Apparently theres a workaround so its annoying but not critical +Just woundering if the mirror going to be added +Closed for +Add flags and includes to and files +The shared distinction doesnt make it down into so went with the required distinction which I think makes sense anyway +Since setx wont work without it + +for the patch +Includes Test class and corresponding process diagram +Use Quick Fix to enable KBREQUIREDFORFACET Content Assistant and +Where did you see the warnings? My local javadoc build does not show any warnings +Documentation +Its now possible to configure a other retrylimit for dnserrors +TODO documentation of the new feature a lot for this contribution I plan the integration for. +we could assign the timestamp when instantiating the logId +Maven project with broken tests showing the issue +Martinare you able to see the demo source codeThis is where I fail + +One more thing is the location of +He indicated I should look at src to create the necessary auto generated file + +Besides we already require Java to use Jackson and JAXB ships with Java +verified. +Great! And thanks so much for its been an invaluable tool +Just a matter of local env setup. +is the layering violation too ugly? reaching out to IPC frm within is not exactly clean +Hi Marcel were you able to look at the example I provided to debugreproduce the issue wit the large number of SQL selects +Patch as per above +Trying one more time wont change things we need a fully reproducable set of steps to follow starting from a clean install +Google also stores web link graph to using an adjacency matrix +This is a replacement that includes that file and the other Java files from Gonzalos original post +Will submit when Hadoop QA will be done +We can also add an introduction at the start of the page to explain that the links from the page only work when you have built the Doxygen reference +Addendum is good tooWhy do you prefer sorting on the client +not https use httpI am in spain and connect without probs. +on trunk and on branch +jspx this patch still neededThere are some issues like adffaces as packages etc. +Fixed in branch before release branch was created +lib are it work with previous versions +Updated patch handles more than one of patch solves the problem where the sub project ignores the parents properties but introduce a new problem where sub project can not override parents didnt apply Edwins build from current CVS and see if it now works in both cases +Committed branch and trunk +Can you be more specific about your issuedo you get a different wsdl with axis +The problem is that to be really useful this requires that you expose the instance and allow the to be configured +I dont have any update because I havebeen assigned to do another taskOur evaluation on logj is finished for now +Committed to branches +I unfortunately dont quite remember whyMight be related to the mutable +Thanks Atsushi for your inputRegardsSenaka +The check burns cycles and impacts performance +Committed revision . +Deleted by destructor +Sounds sensible +startsWith starts endsWith and ends also use regionMatches +Added new patch with corba spec dependency removed from configs +I suspect that the feature definitions are to blame for the copyright issues in both cases +The previous patch wouldnt have fixed the problem in all cases a race condition might have occurred between checking the status and actually shutting down theThis patch synchronizes on the Cache instance and prevents this from happening +Comitted support for namspaces in r +Can you update the diff on RB and Ill take another look? Thanks +I keep the patch updated because we use it here on our local copyThe recent Generics work threw off the line numbers from my original comment +Where is this +As such I favour deprecating in x and in trunk and sandboxing the code wit big warnings that itll go when we have a better solutionAny opinions? Robert? David +The custom encoding feature added in might be a good way to implement this +Is that enough +So feel free to move fix version forward +This is not possible since is part of the packageIf needed well have to introduce an we need it because it is rule of our customer that all objects in db must have explicit name to be able better administer database +Perhaps in fact a issue is related to and +Xavier nothing from my side +Thank you Aaron +Since a view like this does not cause the error to be logged in the if the index is already built in order to really change this we would have to cause the whole index to be invalid whenever any errors are throw while mapping any documents +I dont think there is any need to backport this fix to +In the solution relies on a passwordless sudo +Here is the patch to add the jar +That is still respected by this patch +Please use this patch for Tapestry The bug is still alive in +Not able to reproduce +Jonathan I did not see any inttype and I just checked again and there is no such type in the package +Please try to build from brahcnes +Attaching simple fix which unwraps the and cheks if it contains an instanceof +Have you run the ant tests with this +Thanks so much for your efforts! But I cant find a way to close this issue as I only see the following actions Attach file to this issue Attach screenshot to this issue Clone this issue Comment on this issue Create VotingI have the same problem with +I found out that this is a different attribute +If and when we support those we can add support then there will be nothing stopping us on the RPC level +Closing as Wont Fix based on Alessios comment +This also happens with the Guys +Since Grails ships with groovy does it contain this fix +Is adding them to an option +The tests have been passing for some time now and Solr tests pass too +what +Committed Susans patch revision that updates info about SQL +Given earlier discussions weve had that virtual memory is what really matters I am guessing we dont need itRegarding the config variable names a few concernssuggestions This seems like specifying the amount of memory reserved for Hadoop whereas it means the opposite +I think this is going to be obsoleted by +Without reproducing the problem and showing itfixed upgrading will be difficult to justify +This should probably be scheduled for though unless its clear that its just a test problem and there is no actual regression. +Forgot to mention I consulted Hibernate guy for their solution +Ah I see +Thanks +The server will fail to start if a fragment bundle is added to file +Just change of strings in an exception message + +Added the ability to perform this reset from the view output connection screen +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Yes Im here +Also is the session still open when initialize is calledYes they confirmed that the session is still open +If scale only sign should be added to precision to compute max width +It is based on the new behavior to change that exception +Both in Error View and console log nothing wrong + seems instead +This should be fixed in instead to support this feature in other subclasses +Renaming the contents of and extending their namesakes from Axis Java resolves the issue of requiring on the classpath before Axis Java +without transaction support and with no local EJB references to the outbound JMS RA deployment descriptor for the tests with no transaction support and no EJB local references to the JMS outbound elements +I added an updated that combines the ppc i x jni stubs into a single binary in commit ffedeccfbddafbeedc +Upgrade tests still remain to be built thoughFor foreign key and check constraints as well as for not enforced the property will still required till the implementation for those is completed. +Im looking at and it looks like its trying to load +Patch to validate shipcache specs andPi Ive had to split the validation between and and cant reuse any existing class since was the only option but shipcache specs arent really +I checked in a fix for this in the code base +created www added a redirect in s +Reviewed and committed + +patch +Im confused by your implication that return Java +The result of for the attached pdf +Kanna +ThanksWhen I saw the error message I thought the bundle might have been removedWe are still relying on the old update site mechanism and we need to move to Equinox p +These changes in addition to those mentioned in my previous file attachment include I added a test to the base compute to check for theproper value of +Attached patch should apply cleanly to the current trunk +Tested with IE and FF. +its up to you. +Kris we currently use and the issue seems to be solved +This issue has been fixed and released as part of release +Assigning to myself for closure. +Looks like code work is done here and an upgrade to JBCL is now being tracked for consumption of this fix in EAP so resolving this guess youre saying we need to release + for patch +HiIssue comes not from location of but from incorrect template for this pom because as you can see there is a parser exceptionAnd FYI location was changed because of +David Ill start working in this again as after some discussion need to change the implementation slightly to be prepared for extensions +Sahoo I have applied your patch but modified it to take into account security +Fixing the affected version to be TBD +This could be the final resource for all configuration constructions +Thanks for testing +Hello Rob +Please make sure hudson runs nightly +So Ill add spec as a possible component here +bq +I think that dfs client should unwrap specific exceptions that are thrown back to the client +Verified on. +This issue should be no longer reproducable because wtp has removed the launch configuration type +Branch done in context of AWS regions +Just a minor comment +New patch +Ok cool +Patch works good and applied at revision +I should of tried cleaning it before I logged the issue +Its connection timeout +Hi Owen +Im planning to run the site remotely by redirecting to my internet site in js event using +This patch fixes the problemI also changed the test a bit to try both negativepositive floatdouble +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Tomcat standalone +People do + +If the client wants to do VM wide association they canBut on the server side vmwide association is a definite NO +Let me know if this issue can be closedThanks +But Ill be sure when I figure out why my test fails regardless of that +Thanks for the info on the work the example pom does work with mvn +does ant test pass +Hi DaisyCould one of your team quickly review this tiny patch and push it to the masterPS No test possible as a parser that causes an exception is required to create a regression test for this +though the test is more evidence in addition to Mikes experimentsthat something is seriously up with spanishgalician and tikas detector +Ben and I discussed this and we dont think this is the best design +yep greatjust having something there makes it much easier for the user to tune it to their thanks. +If no more bytes are available but needed unread all read bytes and use something like read to parse until endstream instead of using the given length +Fixed with regression tests +patch fixed a bug that write full path of an empty directory to image +The exception being throw is because there are some unresolved artifacts when the composite is built +They are committing as I type this +for the to trunk. +Still works fine now only the native snappy libraries need to be available somewhereSo +So we should be able to build a from the latest upstream Weld to include in EAP +I will incorporate suggestions into a new patchThe rationale behind using a new class for the public API was that we didnt necessarily want to expose the because its part of our server protocol +I just committed this along with the patch from that it requires +Do we have source for this quickstart? Where is it from +May take up to hours for the fix to reach central repo. +Rickard added you by hand but Im going to forward the response to SF +That would be very usefulIve crossed a dozens times with that needI also liked the Propagate proposal +I think this is wrong but Ill make sure can handle such requestsOleg +Run as Groovy works with the latest version of the plugin. +You need to close the directory dest explicitly not only the writer +bq + for the to trunk and +Thanks Bryan looking forward to getting the patch +Going through Array to get the elements will be slower than simply doing array indexing +And a new account will return nil because it doesnt have any friends +Thank you the patch and workaround both work great. +I just committed this +Wow +The improved scheme should work for more casesI am attaching rev of this patch +Ok Bruno changed Patch is updatedThanks Jacques for review and testingCheers of my own patch +I cant reproduce this so I will assume it was fixed by some code I did since this last failed +after latest commits trunk compilation is broken +The reference to is only in IDE project files +Sorry for that I thought I had fixed thoseThis one passes checkstyle +Why does this not result in NPEZachary could you attach the relevant sources for the deployable leading to this? The fix is easy enough but should be verified with a failing test that gets corrected +Ok so we have a split parent with the SPLIT marker in the hri in meta +And the user has to configure the plugin by setting plugins user and password properties and then enable the plugin +If a column is function with single param and the param is a column use the column name as alias +We should also have this for interceptor bindings as well +Could we put this only in? It seems to go beyond a simple optimization + of a bulk update all resolutions changed to done please review history to original resolution type +With multiple places I get the same Resolving absolute path of executable error as previously reported +Remove them or add logging to see the actual failures +The main idea behind all of this is that you will come across problems if the same class is loaded in two different classloaders one is the parent classloader and the other is the reloading classloaderIn your includeexclude patterns I dont see the point of excluding +He he sorry about that +I have merged this to. +I guess the changes are in trunk +I have attached the patch with modified test case +ok i think this one is betterFor convenience I list which JRE versions correspond to which unicode versionsBecause I think that for example if someone goes from Java to Java they should not worry about this +There is also which for example could be used with a to allow apps to add static properties added to every logging event +The section of the JSP spec Implicit Map Entries from suggestsThe taglib map described in is extended with new entries extractedfrom TLD files in the Web Application +Sorry +for the patch I will commit it shortly +Checked PBMAGIC is included serializing and managed deserializing + +Thanks a lot KevinSent from my iPhone +Im not sure what the problem is and have not had the chance to look into it +Merged up to trunk rev +I could see where concatenation sort of marginally makes sense there +AR deployment is already working in +I just committed this +Ok Changes committed +we really need to do this +Like Craig Im very curious about the git ref and whether there are any differences between and a remaining. +I didnt find the latest +warJens +When I read the documentation I also thought that a new appender should dynamically be created for each unkown keyInstead of changing the I think the node should be cloned before the appender is created +I thought we already had an issue which covered this will try to dig it out +Okay Matt +I will close this issue and investigate the new branch. +I think this is a situation where the srsDimension is greater than the the dimension from the coordinate system +It should resolve the problem +FYI It works with but fails with and +AndreasI turned off Hot Deployment and removed the scripting module and my problem is GONE! While this is still a problem your workarounds have resolved a long standing problem for usTHANKS +Tom +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Ive just committed this +Hmm the author didnt follow up +Ill try to get a sample project attached here till weekend +Its a hard choice +Can you please provide me the trace log for the entire JACC runThe security error is because the security checks have not been able to derive the policy context from the policy +This will make phpunit look for test cases inside this directory +always cause to update fix status +All the regression tests passed +duplicate of dont think it is thats a large catch all issue and this one is much more specific the only thing thats duped is the title +Changes Modified evaluate to catch Is the patch ready or is this a test infrastructure problem? Either way if you put up a new version of the patch that should kick off the tests again +Im behind removing two but I dont want to break compatibility +Part of a bulk update all resolutions changed to done please review history to original resolution type +Currently only used in the test but nice for consistencyI will commit this later to trunk in rev +yes should not be able to unarchive if the package is archived +Thanks for the new patch Narayanan and for fixing all the minor nits Ipointed out +The patch has been appliedPlease verify +Normally the default is the bean name pool +Therefore I am the exact same patch named as +I see Ive missed different exceptions +rev of Jopr updates the to reproduce see the bug in action invoke the client like thisUSEPROFILESERVICECLIENTtrue +IMHO either we implement a workaround for the issue or we document with Glassfish in the Seam Documentation +Reopened the issue as the team still discussing the complete project structure proposal and associated names and there are some more things that needs to be done before completing the maven setup +The failure is caused by Jenkins memory issues. +bulk defer of items from to. +Yea Ying I came to the same conclusion +I believe this patch does the right thing I looked around the jars and tars and everything looks right +Adding handling for leading +This server is dedicated to serving WMS and WCS. +Committed to and trunk +New patchPrints true name of parser even if wrapped inUses recursion to support of composite parsers +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Sorry for the inconvenience your first comments unfortunately remained ve fixed this for now +changed priority to wanted to reject the issue +Reviewed by Gordon +This should work with via +This has been open a couple of weeks and so far no issues +fixed in the trunk some days ago +I had to explicitly make modifications to do buffered reads with proper alignment +My intuition is it makes more sense to do this the other way around and have records aligned to blocks +I just committed this after running all tests locally. +Regarding the latest test plan documents I didnt see automatic getting covered by any case +for having a base class like Mutation +Now that the start row key is Base encoded the problem with splitting regions has been fixed +Max please review the patch and commit or let me to commit +youre right I didnt understand this consequenceno problem to let this issue open and wait for a later of and the fix for in r we need to depend on or newer. +Robert thanks for taking a look and letting me know about the validation setting +You may be right and we certainly dont have enough knowledge of client side GUI design and toolings So if you come up with some nice patch to do this you are very welcomeIn fact I assume that the configuration page certainly would benefit from using UI instead of the homegrown stuff we employ now +Integrated in master +I think the best way to handle this is to get the error all the way out to the browser + +I also tested the simple Order Customer example with Hibernate and +so I noticed the problem is happen only when first character in project name is capital +Wow a lot of votes for this issue +There isnt an isAlive but there is an isCancelled +HiI think we should try to create a new material about the project +And that is why I asked how you are doing it +For the second case how do we manage this in production +The assembly plugin is still the culprit. +Should we do it now +Its basically an indexing time optimization that precalculates the bounding boxes ahead of time +delete the old attachment and I would like to attach instead thanks +I introduced the in order to remove null checking and make code more readable however it is not strictly neededTalking about the processor classes do you want me to move them to another specific package and submit another diff fileIm happy you appreciated my work let me know how to proceedIf youll commit my changes Ill start working at improving support by using request processorsThanks +Error in the implementation of the actionProces m not sure to get the situation of this issue? how is it fixed? +Typo definitely doesnt look good in public methods +You disagree I presume +Resolving. +issue is an optimization and is hard to write a unit test +Generally speaking the GS UI does not have action buttons besides the layer group interactive editor and I seem to remember it was a GUI design choice with the intention of keeping the UI visually light less cluttered +Fixed in +RTE has been replaced with Xinha +yes we fixed a big bug in the way we managed the entry fetched from the backend in some case we werent buildling the DN correctlyI dont think we have created a JIRA for that sadly +The fix I proposed Im currently testing it in a loop + +I am thinking about changing our module abstraction layer to make it less abstract in the future but this will be a major refactoring so it has to wait +Ive added debugging support +Fixed in trunk +This is an old and long obsolete element from the days that should be removed as it serves no purpose at all +Isnt in pretty close to that? Admittedly its not documented yet but basically serving the same only difference is that we used to perform validation in postProcessAfterInitialization up to now whereas you are doing it in postProcessBeforeInitialization +test is fixed in RC Ill mark this issue as fixed in RC +That is it +now we have mysql msql and derby +a repair call. +not sure why I thought that JDBC was involved +Ugh we should actually continue to allow using KEY in +If we do nothing either NN will work and there is no way to recovery even in manaul +Awesome +Either setup JOTM to bind to JNDI or provide an implementation of that uses the purely standalone JOTM +try edit simple text element in visual part of editorASSERT text should be changed in both parts of editor press visual refresh button try edit text in visual part of editorASSERT text should be changed in both parts of editor open any jsphtml or xhtml file +This isnt a parsing error +If you do capture netstat capture it a few times with gaps between +Fixed lets see if it breaks anything on hadoop qa +Committed the patch to trunk and +Elimination of Throws conveniently fixes close of resolved issues. +To date weve defined that as just being things that are marked as published +I think they are as good as they are going to get +The annotation was missing the tokenizers for core places +This PHP writer is inconsistent with the JSON if you use php s decodejson maps come back as objects +Velocity does not do floating point math in the template +They are similar in that they both end up causing large amounts of data from being garbage collected however +My patch was only a quick solution +in +Have fun +Remoting will be mavenized +CommittedTrunk r Branch rAlso put a note on the wiki that users will need this patch for this to work with backupAfterThanks Tomas +It appears that your original patch was committed silently in rev +Version of patch that works for +I added two batchUpdate methods one takes an array of Maps containing the batch parameter values the other one takes an array of objects for the batch +It will take a while was just released and I will be on vacation next week +Resolved +older issues for Apache since were no longer actively working on these at the moment. +I had to polish it fix checkstyle add license headers remove annotations from the classes improved mbean info assembler to detect the annotations from the interfaces fix a few missing attributes in the interfaces add unit test that uses the client api add for the new for applying the patch and the polishing +This is a version of that sotores object in the Map only if you give explicitly a otherwise it stores only in the Vector +Fred +feel free to propose a patch in the mean time closing due to lack ofPatch is provided in first commentSergey +interestingI am debating whether applying the patch in or wait until a for backward compatibilityI imagine that todays code is not backward compatible between runs though +The only option is than to delete the task +seem to be fixed already on trunk. +OK Ill talk to Jeff and closing thisThanksHoward +But the server seems to be not working at this moment +I now consider the issue as resolvedIf you guys have further fixes new issues should be created +Kihwal thank you for all the great works +a patch against applied at with the release of. +that probably the input stream should be closed after the data is transfered to the output stream +However thanks for your analysis +no doubt its a very useful tool for the community +I should mention that make check run on indicates a passing test suite though I cant see a lot of error messages in the test cases +I see your point though will think about the phrasing +This looks pretty nice overallThanks + +You can go on with the migration +It still seems a little awkward to me but I guess its a known common practiceClosing the issue. +Thanks Anilyour patch is really very good and Ive committed it in rev +LOG be used aboveNope +Also reducing the of packages to only hue this point the only thing that is left is smoke testing + +That would work too +For example +Whats happening here is that the control flow analysis is slightly incorrect +Thanks Yongqiang! +Question If the AS deploy directory contains the and directory does that mean that the AS contains the ESB componentI spoke with Tom Fennelly about this and he says Yes that usually means its been installed +The messaging client can control the message ordering through creating ordering groups via the producer +once this patch is accepted is there any reason to NOT store the data indedicated filesThis only seems to be an advantage +looking into it +This is a general covariant upcast issue not +This new patch was made against the current SVN repo +Please explain me what you are doing so I can at least create a test case +I took their Fedora meta appliance everything looks normal now +Im not sure we really need to get into SCA at a detail level and the overall architecture view can be a mixed bag depending on what information the user is after +Oh almost forgot Ill also add in support to so it becomes easier to provide custom decorator action titlehints and alt textsSsee for further decoration enhancements above +Ditto for +This is working for me now if I use the build +I think we may need a result returned by the WeightScorer compactly representing the space of valid fragdocs for this one doc +However I think the part of the patch that simply checks whether an existing has been provided and uses it if present is still useful and makes a huge difference +version of the patcha Has a test class which uses a mock FS to mock failures for some operations +I am addressing the issue of how to deal with system properties elsewhere in. +I will document all the cases that are value method call names and outside that youll have to send +Yes an update to JDK resolved this issue for me so if this is seen again would suggest a JDK upgrade. +looks like either recent rm restart changes fixed it or we are missing something +is still as seems theres a visibility issue with +Included explain into random test +But then why dont you just do that in +I wrote it yesterday and pushed the Add button only today +There was no +Committed to trunk +Hi Arpit as discussed in the left part of v patch already include a fix for it so I split it as a demo patch and put it here +Let us carry it forward there for MRClosing out as dupe. +could you please attach the report file from the target directory if you still have it +All usages of deprecated class are replaced using +Fixed that +Seems like a drop and recreate of unique indexes on nullable columns is only guaranteed way to fix existing issues +No more like cp and one ejb in the archive +I noticed that this patch removed some tests for example test at and they are seems validate test why remove themAnd there are a lots of changes which converting tab to whitespace its better put changes in a separate patch +RickIll go on and file the patch then +Also currently no way to specify a trust store on the clientI had the same problem with an application which use two servers +The above list of files would of course include all files json txt etc +This would require to shove Maven SCM into Maven core which doesnt look appealing +Assigning to Morris K +Thanks for the examples that you included +Thanks for the comments DagIll get to work on these tomorrow +Im wrong with Seam +There are custom javadoc tags but they are separate things +Its possible to add this it would have to be added to the interface since thats how the original NMS API defines that the body of a is accessed +This means the wars will have different TOC because the jars within it have different sizes +Maybe a deadline for making comments would motivate more people to present their views +It is a small but safe modification to make the unit test pass +For jfoundation there is no +Verified fixed with RC +Zheng could you please explain what your patch does and what it fixesDid you mean to reassign this to yourself +I think the addition you made to the docs makes this very clear +I incorporated all comments added some formatting to and clarified rename and delete documentations +I will commit it to trunk +From testing it appears that redistribution is mostly event driven +Pushed changes to and +Upul can you fix this one +Thanks Amila for the patchAuthor lahiruDate Wed Dec New Revision the resolved issue. +Applied patch with havent found any other processor thats having the similar issue +I can grant you permissions as needed +Create clients +After adding the application +Cleaned up some long +If I am looking at the first screen of pets and the last screen of owners Id like to swap between the two and carry on seeing the same records in each table +EAR file andCan you send the full stack traceAlso can you try adding an annotation with a name attribute in addition to the? That may allow it to work OK as their will be a type in addition to the global element +Not able to reproduce anymore on latest JBT or JBDS +Compiling one large executable from a single module is not all that interesting +The thread will have to be collected later +fixed in revision +This message has been fixed +Username bruceb +Ill review and apply it +Use the new to generate the doap fileLet know of the new doap file location it was moved under frameworks. +They now could have multiple sources that use +So the actual DN iscnDomagoj MaduniouUserscnCROZdcurieldchrIll also give it a try after i finish unpacking +Will be available in tonights snapshot solved in nightly build +Reviewed the patch +BXBOXINGIMMEDIATELYUNBOXED warning and undeleted class were problemAttach fixed error occured +This issue is generated by an incorrect fix for that I am about to revert. +So unless you find something else Im committing it. +contains per request info so the framwork would be forced to set the field if it was to reuserecycle the beanChanging the sample to take the as parameter to the Resource methods instead made it deploy +I will work on the fixes to that and the other tests and submit YAP yet another patch +Also when strings are created they must decode while Utf only decodes on demandAlternately we could include implementations of and that represent strings with so that folks can use whichever they prefer +Thanks Michael! It works ok nowRegardsHow can be used with? Can you provide some sample code pleaseBest Regards! +Ashwin +Will commit this in a bit now that has been committed +Fixed +Juraj Im not sure how what Im seeing in this screen shot is wrong can you please clarify +Thoughts +It seems likely well want plugins elsewhere so the manager might be more abstract able to return instances of other interfaces determined by different configuration properties +I renamed to omit and added a NOTE to the javadoc about positional queries silently not working when you use this option +Tedbq +the problem was that unaryPlus didnt catch the exception in SBCA now it does +Also updated the of the redirects worked correctly on +Namit Ive opened up +We need to catch since it is thrown inside the block and we also want to catch so that we can warn an operator about the responder thread dying +Sure the user has to work out the right match to put into place but its not the job of the FAQ to give all the detailsIve added a link to the sitemap docs to help new users. +Should we be at all concerned about that? It doesnt seem like a big deal to me but Id like to hear others opinions +This feature seems like a refinement to the namespace quota but is being implemented completely separately from it +Closed. +BTW thanks for taking a look +Patch that should fix the CyrillHow do you have the functional tests setup? Are you actually calling MXMLC can you give me your setup? IE yeah I just got it working minutes ago and got the Exception now +Looking into a potentially more serious security issue with the handling of the path thereFixed by a combination of my change and Allen Gillilands change SVN change s and respectivelyI verified proper behavior for the case of spaces encoded in the url as or Fix should appear in. +Adding patch for trunk for only difference from last patch is actually adding some testing around the changed logic +Patch for trunk +Moved to next milestone release +Fixes appeared in documentation so closing. +Missed in the description that it was the second time around that was causing the issue think the extra second discussion confused me +Since thirdparty contents are different for EAP and one needs to rm fr thirdparty prior to switch building between and in the same build once thirdparty is built the build does not attempt to build it again +Hello Jacques +Applied with modifications to the MAVEN branch +Fixed by removing unnecessary locking introduced by previous change for +Resolved in in jbt build. +ve just committed this to trunk andThanks a lot for the contribution Colin. +Thanks for reporting this its now fixed in CVS +This is nothing close the a issueif you get a random error when fetching this entry that means another thread is processing another request for the same session at the same timeI have to check this point but I think it has been solved in trunk +Thanks Shameera the patch looks good +instead of +I just dropped this jar into a copy of Branch and ran the ejb clusteredentity tests +The task is out of date because the Smooks editor was completely redesigned since the time it was created. +its set to false by ve also added Nick Dimiduk to the watchers list +Make new datastructure that manages the servers to regions and regions to servers MapsNot done yet +The change will be in the next release of Roo which is due very soon +Thanks +dbo +Another temporary updated patch with all variants now in the one class +It is attempting to close any outstanding EMs that had not been closed +bulk move of all unresolved issues from to +If you have anything in addition to it please share with us +Moving this out after IRC discussion. +verified onsubversion revision +It extends and changes INCDEC rates depending on current +You have no idea what currently is in the buffer but you would want to flush what about to be inserted you would want to know youre left with an empty buffer +Added the Repository name information to the Features webconsole plugin +the documentation should explain the difference +actually I dont think theres a default place where to put actual data in the new layout? +Latch patch looks good +I will address the test cases too +I second the motion +I am eager to see a implementation in pure managed code without using classUsing is not the issue +We should consider making this change for when we decide to put out the next major release. +Can anyone pls take a look? TxI am giving it a final look for commit +The tests need to be updated for sbyte +This will fix the problem by maintaining the counts correctly +The location is fine I am just not sure I understood what it was for +Verified with +We should be able to spot the difference from existing code more easily +Ah after the RM restarts the NM can notify the RM that the AM container exited then that would pretty much fix it +This just feels like more backward work to meI actually do not think so +The is duplicate to JIRA . + assign GSOC I am University Of Moratuwa student who is currently involved in research of creating Geo Science gateway for the Apache AIRAVATA +The error seems to be that a JSP cant be compiled due to a missing +gitIgnore file and it still commits all of them +Hi MamtaI took a quick look at the diff and noticed that the normal timestamps that print for network server commands are commented out which I guess you did for your testing +Also fixed the link below the screenshot which was also labelled M +Ive attached the patch for +Then I believe this can be implemented as a Hadoop specific backend configuration +Thanks for clarifying +If not just reopen closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Would be nice if there was a hibernate configuration property for choosing the preferred uniqueness strategy +Can you elaborate what youre trying to test? I dont quite see how an interface helps +Yes agree +You need entries on the public methods of Token +Tried installing patch to see what for an improvement Id see at +Once a day a process was run to download data from dispensers and uploaded data in OfbizI did not see problems with current data model for supporting so often meter reading recordings +Strategy in the future when we know how to do that +The code differs in this respect as it will add the library jar to the regardless of the type of deployer found +I guess the rest are automatically added by Pages +applied patch also made some refactoring to eliminate duplicate code and unneeded method calls added integration testThanks Andreas for your contribution. +Here is the svn diff you requested +I just dont know enough to realize why Thanks undocumented environment variable is totally unnecessary in a batch script +If a fix can be made for then please go for it +grooms status etc +I rewrote fix +Attaching patch +I think it would be appropriate to let Johan do the honors but its up to you +I see the duplicate issues I submitted were removed and assume this remaining issue that was not removed is the one to work on +Not a fan of relying on assert when we really want +If you checkout struts from svn you will have to also checkout xwork and build it as Struts depends on xwork snapshots for development before a release is made this reference is updated pointing to the latest xwork release +Made critical issue +I really appreciate you taking a look at the stage server and trying this out +possibly can let them enter a comma separated list and let them all be connective literals +Does that look like what you were thinking of + +Committed at subversion revision +Which system are you building on +Ill give qdox a try is it backwards compatibleSome more testing showed some issues with my above sample code too but I dont think it matters anymore as youre targetting qdox only probably +So is this fixed or should I look into it +So if this goes out in and a user doesnt regenerate the specific records everything should still work +Also Im not going to bother with hash change based communications +is set to a negative value at the first callI think you are right it will be negative +Install gems etc +I think its only error message diff needs to be sed out +looks good carry on and try access page +Build is running clean give or take the usual crop of failing tests +Hi Nadia Im sorry as Im pretty sure an undesired XML parser is attached to the classloader but I dont know how to help you investigate this issue +Is the proposal to add another log statement in with the lock on the? If so that is a slippery slope +So the fix has to go into two places The Activator to handle startup scanning and the to handle import and other types of I am using build from SVN +I do like it better thanks for listening +This will be in the next Maven build of validator. +Yeah think so too +The title will be replaced with the same value so the queries will be the same trying to understand the effect of having updates to segments on the readersearcher. +Thanks a bunch Willem +In that case why dont you post the code here to be reviewed further longer term for AS the plan is to move on from the current approach of Negotiation to an approach that is less dependent on the multiple domain definitions that would be the better point to review how the fallback scenarios are implemented. +Honestly if the schemaconfiguration set is small +Attaching a new updated AndrewI believe you are right about NPE it was a miss from my side i have attached a new updated patch here please have a look at AndrewDid you got a chance to look at the patch? It would be great if we could close this as soon as yet +Attachment has been added with description thread dump while doing getURL has been added with description Yet another update to the has been added with description background problem background now rendered has been added with description screenshot rendering repeating has been added with description Another update that fixes black problem and inset problem +Updated repro steps +Thanks Jim +I dont think theres a way for patches to work when adding new binary files to the source +Hi Jacopo This problem can not be recreated through the UI +I am still having several style issues +Yes you are right +No regression I vote commit + +Works now verified by some unit tests +I can go either way +I have corrected it in the draft +The problem was two fold was failing but it was unrelated to this patch +As a result interpreted calls still have room to grow when it comes to being more efficient and faster +Perhaps the issue should be marked wont fix if this is the case +Because of this any client code reliance on this behavior is probably a bug +I will get to it today +I do this manually anyway in most of my code +Cheers Matt +Any comments? Thanks! +Committed version the one that separates the calls to setContext and initialize +Thanks Calus +Any strong preference? I can inline if so +Can you try this with the? If its broken well schedule for or +Patch applied at +Patch replacing reference with +to as to handle clearLock +Bug has been marked as a duplicate of this bug +on patch +Verify and if something is missing +This is really very odd +Unfortunately Ive been working on other stuff and that is going to get committed as well so I am thinking I will create a branch for this so it can be inspected before it gets merged to trunk +used to work before his latest round +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Need to be translated to the other supported languages as of today +The old class has been marked as deprecated in and not used +All the tests use the listener +What does this do exactlyif it is about the new maven repo then please call it public Maven repository to not confuse it with any possible official repo +Updated in revision . +In Lucene the fix would only be needed for highlighting in will have no anymoreYou can have custom queryparsers in Solr so you cant say definitely its not usedWhether Im concerned about it depends on if we a if we do Ill fix it +more diff I was green when I opened this +I am getting a few tests failing but have only just started looking into thatNote that with the first one Im looking at the lack of information in trying to solve the failure is not due to a test runner issue but Derby not providing any information +Confirmed by checking Travis result the error is gone. +Attachment please find another patch for reviewThanks Ramin that explanation makes complete sense to meI dont have a strong opinion about this deleteFile issue +reformated and +Seems to be working as expected + +resolved in should be added to every jBPMjBPM blugin to make Legal Info button any news +Jopr r. +this +Is there some sort of a call we could make? Not only would it reduce the compilaitons it would also reduce the impact of a classloader miss +FWIW this seems consistently correct on OS X +Added Note below the Data Source Name fieldAlso tweaked coloring to improve resolved Release +This is the implementation option which uses a shadow column family for cell ACL metadata storageEdit I had to context switch to something else and am coming back to this now +Alexey has already completed this work so Im resolving this JIRA +forgot to attach the context not sure how to edit the environment field so i am adding more here mongod +tried to launch local deltacloud as external process +Unless Im missing something it doesnt make sense to exclude the lower bound +Same db replicated to my cloudant account does not show this issue +Patch +It was fixed in +For some strange reason is does not complain about the names file anymore and i dont have a config folder in my runtimelocalCan someone test and confirm this patch +This was also seen in test batched updates we were always still passing in an empty context array added a check for that. +Manually tested! +Attach patch +I had to hand do these changes in codeline because the changes in trunk depended on +PaulCan you verify what you browser is sending +Our current implementation that supports embeddable class is not OO style + +Here is a patch +requested in So this problem is not a problem of not finding a but a problem of not yet supported functionality +We get this NPE only at sonar the standard jenkinsci builds do well +Reviewed by Alex +I try to find in already existing if the issue has been already discussed but didnt find one matching my problem +Adam pointed out on IRC that Im starting to go down a bit of a rabbit hole to get the driver reloading to actually work +tweaked the patch slightly to move the constant into the constants classFixed in rThanks for the patch! +Logged In YES useridOr we could just get out of the schedule method on findingout that is null +HelloWould you please try my patchThanksBest regards Andrew Patch committed in revision + +Closing JIRAs that are productization tasks that QE does not have anything to verify on the actual builds so these does not show up in our filter of issues to verify and close. +Please attach a separate patch to for scanningUnless I missed something I committed all necessary changes to support managed bean injection via ResourceThanks Gurkan for the patch +fixed in r not closed yet though cause i promised chris i would write a converter for old System properties to new System properties for file +Not the warnings I mentioned in my first comment +Im notintersted in that +This patch turns off schema size checking for complex features + and derivatives have been updated in revision +Attaching a patch against the Synapse trunk +I would expect to see the results you are Kevins comment from January. +Review how this method is called +Either directly or by some root resources +This should be documented as a known issue for EAP CP CP +Patch that adds in a configurable time limit with default being sIm unsure how to add in a test specifically for this but this does just extend on the API which should have test cases of its own in the Java sources +per the JIRA meeting on Nov +initializing axishttpouttransportinfoencoding to NULL solved the problem +The last patch was incorrect because of a know issue with subversion and on XIm attaching a new patch generated by git +For now resolving the issue with the advice to override the protected endDocument method when used with Woodstox +Feel free to add that if you want to apply the patch +Patch that puts back the tests for windows full command line quoting. +i will give it another test without index +process instance ids are taken from sequence so they are always ascendingSo the the result will be the same no matter sort is done based on instance id or start time +Feature Incomplete QA cannot proceed with testPlease resolve Bugs and Implement following Enhancements as per original requirementsEnhancementsAbility to delete Events and Alerts Select multiple alerts and eventsAbility to delete Events and Alerts Search by Category is requiredAbility to delete Events and Alerts Delete by a time period is requiredBugsAPIMS deleteAlerts API responds with error even though required parameters are specifiedAPI UI Admin can not delete EventsArchive from other accounts +yes its trivial to use the marshallers in the client code but this makes the client responsible for marshalling the payload and that was removed from the client so old issues +The correct solution may be last in a service same as first +In general its not a good idea for software to use common names like Assert for macros +At the moment I cant schedule it more or less accurately since it depends on changes in other modules +I apologize for my absence on this issue +If it is still relevant please reopen +computation of timebq +Fixed in svn r +sorry cant convince the cms to publish in another directory +I did find a couple of bugs in the partner link wizard changes fixed contributed back to eclipse +As a the calling shell could test if DERBYHOME has been set before sourcing theset +the only problem i currently see the logger requires java +from laptop Able to ssh to admins public IP +We might have both interceptors and the invoker options supported +Patch committed +Yahor please also keep description updated +I prefer it that way because then it will be consistent with the other views because some views have multiple controls which need to reflect the input changeIf you dont mind I will just implement the change so we can get this issue closed +Added a factory method for warningsM javaengineorgapachederbycatalogChecks for whether user existsM javatestingorgapachederbyTestingfunctionTeststestslangTest cases. +I can see the attached content is actually in the output pdf file with +Applied patches to all major active branches +All tests passes but it may means as well that we miss some testsI will be looking into the various options +Source +We could try to automate switching but I dont think it is worth the effort because it is too likely that the user will be left with no way at all to log on as the old realm has stopped and the new one not started in case of any errorForrest is designed to contain everything you will want to change if you want to customize the admin security for geronimo +java files without doing harm as I say it looks as though the checksums are mostly wrong irrelevant besides I could regenerate new correct checksums now I know howDjunit testing seeming a mildly buggyfragile Ive had one sequence of virtually completing then giving me memory depletion now a bizarre port errorIve sent in an ICLA now so it should figure in a few daysBest regardsthis updates the patch to include standard +has been marked as a duplicate of this bug +This is a necessary fix +I committed it to trunk andThank you Binglin. +It is generally easier to keep track of what is openedclosed at any given time if you create new issues for different problems even if the version is not yet released +Sorry I didnt realize that I had left the priority as major +I was holding off to avoid the build before release but this can be committed in isolation +I mark these as resolved once they are pushed to upstream master +See comments and patch in about? This does not work at all with Quartz Spring since it attempts to instantiate which is incompatible with Quartz I do not agree that should not be supported for Quartz we use declarative XML configuration and would have to overhaul everything to use job namesgroups instead of bean references in order to wire everything together +Here is a simple patch to just return if persistsBlocks already enabled +In the second and third cases one could search but for case zero there is no option +for patch committedClosing this jira as resolved since all changes to accommodate the new interfaces have now been checked in +The small lines of white are still there but because its in a cell I then set the bkgrd color similar to image and its practically invisible +At this point these messages are harmless +Jeff check this one out if you cant reproduce just close the have been able to reproduce the problem by starting with a app and then upgrading the app to +Robert would you like to take a stab at reviewing this +will commit after tests thanks Ramkumar! +Applied patch in CVS +rawresourcesb tests took minutes +When this error occurred could you open the latest version of the file orgcodehaussonarruntime in your maven local repository and check if it is the mysql or derby jar file +Testfailure is not related to the patch +The attachments were uploaded in the opposite order +Moving all of them into context seems as overkillSo we have three scopes of information request scope including portlet state viewId redirect parameters and request type portlet scoped view portlet name config options and parametersI suggest to have different objects for every scope for current request state thats really contains current portlet configuration and some for options Ther are not a lot of options at application scope so they could be put into too +Im also wondering about a spec to show the use case +The JT will not process any data sent by the during a paused state +But this should work for now as Vertex and all extend. +This would be helpful to folks exploring the new and would be appropriate for +Ill try to lift the patch +Some test cases require modification to reflect current read me instructions Im updating the test harness as I find differences +Maybe a different Source would be better +Fixed here +The least count of the memory needs of TT and tasks will be kB or perhaps even mB +Closing it. +In the current gradle build we compile against but run the tests against slfjlogback +Thanks Justinas Ive found and fixed the problem a new patch will come shortly +Please create the same project in a standalone Roo shell and test and then provide the script here +It looks like the compiles and runs fine against of the +Hi BrunoWhich plugin do you mean +swap file to contain the names of the files it has cleaned but there might be nuances +patch looks good +During todays release well restart SBS +This deprecation warning is unusual because the accompanying comment suggests an alternative that does not seem to exist yet +been running for several months now + is the documented and standardized behavior of the bulk operations +In that case the values are written into a bundle private data file which has precedence over the framework properties +Pushing more writes down to the OS more quickly means our disk heads are hovering at the end of the file more perhaps +Thanks Arjen +also requires that Facelets applications use version or higher +On Windows it works fine +Nicolas GilNicolas your patch is in trunk at revision I simply put returns in case of no dispatcher or delegator +We can fix the agent to ensure it adds the backslash but I think it would be more robust to change create to accept either +Why do you say its incorrect? I dont see any difference from RI +The only issue is that this changesbehaviour pretty late in the release cycle +One last question it might be reasonable to move and friends to +I removed the verbose stack trace from the test output +linking to previous issue. +Weblogic as many other JEE servers has its special syntax for classloading and hence why it does not support Apache Camel out of the box +Maybe in that case the message will be converted into before delivering to the is this solved already or youre still working on it +otherwise this task is too vague to complete +Keeping hadoops classpath to absolute minimum will avoid headaches for all the downstream projects +Matt is it safe to assume that your comment was meant to go elsewhere +Depending on Spring is a for core it could be added to the spring plugin but a freemarker class in a spring plugin smells +Seems to be fixed +Presently IO reactors process new connections before purging closed sessions +If it persists then Id recommend attempting to create a short but complete sample app that demonstrates the issue and does not require COM knowledge +Can anybody help +All features need to go through the community as requested in previous comments +At this time i cant make this release because i have a snapshot dependency of and im waiting a release first i dont want to make this release because i currently work to historic support of all metrics in dashboard with a database and i think it will be great to make a release only when the historic support is ok for everyoneNo +Please Verify and Close. +The for is the actual solution for the problem described in the Mailing list +I wrongly assumed that property would always be set +version still says have to remove repositories sectionare all dependencies already in central +back port this issue the fix +Patch that converts the LEE into FNFE and handles it correctly +wouldnt the configuration rework be able to take care of thisIm not sure why we need to guess the type in the way it is done now cant we match the field name to the tag and then get the type from the component using reflection +Both cases were succesful + unscheduled issues to the next release +Fixed on +This seems fixed now +Is there a good place to place the javadocs on the Apache website once they are more completeAdded javadocs +The implicit this is set to true for that must not happen +Adding dependency on as Maven does not support Maven repositories hence cannot build Maven plugins +You could override the default null string value N with ROW FORMAT +Indentation in the class is messed up +This is what I get for staying up lateI have attached the patch as and hopefully followed through all my claims +Targeting this for +Ill test it tomorrow +Then scanned it with Phoenix took s +resolved per Rajesh +The code above uses and does so with minimal extra effort +Closed as of a bulk update all resolutions changed to done please review history to original resolution type +We shall not since RPC is annotated as +When we load the certificates we will extract the subject CN and any subject and search the full set for the requested SNI nameThe initial patch will only support full DNS names +Committed +Trying again +Added following interfaces one for each package added implementation useful if user wants to avoid logging altogether. +Committed +This is ready for review +Tom Isnt the connect timeout is mostly for establishing the connection? I am not sure if the process is down we should be getting connection refused immidiately +Any hintMauro. +From my tries we dont need to do the rd change the remote deploy on jetty just works fine. +This is likely to be a duplicate of and hence already fixed in ARJDBC Trunk +for upgrade scripts +After that issue was resolved all custom functions and date parser functionality is handled by doesnt hold these as private data anymore +I committed a draft of such a Scalar interface in revision +Test for this working run ant and look in buildivylibHadoop for the new configuration and the you wantHow do you plan to test this incidentally? Id recommend having some test +just committed this +Submitted patch +Robert youve removed copyBytes from entirely +If you do please mark this as a dup. +Unit tests pass its a great start +So stupid. +If you receive this message in error please notify the sender immediately and delete all copies of this message +Hauke Im cleaning this up +Marc Parisi saidI think we should have a scan time class loader context that enables fullyhot swappable analysis and iteration +If needed create new issue. + has suggested +The assembler subsystem relies on vocabularies +Looks like updated made its way into main Maven repo but more importantly is there too +Last week I tried the patch called but it didnt seem to help this problem I am using what seems to be the recommended hdfs for hbase +It is unnecessary to call get firstThe problem is that most applications work with +bq +patch to trigger Hudson QA tests +Delete all +I grew up in typewriter world +From what I hear the ZK branch has been cut and theyve only blockers left so it should be out pretty soon +moving off beta if we can create patch to control this text then that would good but not a priority +What browser do you use? Can you try with +i fixed this in properlysimply rewrote whatever was needed it is now in no more need for ICLA or any kind of +This should work without problemBut in the trunk is the same bug +I still not sure how useful it would be considering that it doesnt really make much sense to keep projects or models in memory since there are a good chance that they would be changed externally so you would need to scratch only case it could be useful if you are implementing your own model editor but there you may use your own model or something instead +The plugin seems to be javascript only hence it does not update the The null message on the uninstall dialog is a separate issue can you file one for it +Ive created as suggestedRegarding the incorrect metrics the count does not go to after draining +Furthermore we setup Eclipse and build a simple Web Service containig nothing more then one class without any content +This should be available in the next Spring is finally committed to CVS now +Since a user must be logged in to set their preferences the expandcollapse controls should only appear if the user is logged in +A closer examination of the latest test results reveals that its just wide builds that are timing out +Please review the patch. +it was passing on my machine before. +The only difference between v and v is that the client master file for lang has been updated to reflect the latest checkin for +Assign to during tidy up prior to +So closing the issue. +Im assuming you have no such trouble? Ill clear it from my repo and see if that doesnt remedy it +Right +Resolving it as not able to repro +Ive also merged this to. +When working one starts wondering if what he need is in the catalog or in the factory or in the builder or in the utility +Havent been able to find any cases where this fails but of course its still possible Ive missed something +This patch addresses embedding a pig script in a process definition + +My bad! Checked! +Still keeping this open for further patches +stringchar longintfloat etc +Yeah I suggest you attach a patch based on the trunk code and only highlight a single change +Its much easier to decide if this happens with method calls on the same object +Following test classes are missing in your patch please attach them +Now that work is underway with AS all previous community releases are + we target this for too +Patch to the configure ac and Alex +Will wrap in a plugin for +The script cannot handle the parens in the install path +Ooops I meant the constructor of not +Patch that doesnt clear the outbound messages array until after weve delivered them to the master +bq +Thanks Kihwal and Todd +Is this really necessary? +I think its ok but needs more tests just to be sure +After you pull them out the main project becomes simplerEventually I think the is a good candidate for becoming a module +Currently it happens that whats on disk uses the same encoding as whats in the memory +So all issues not directly assigned to an AS release are being closed +Attached the updated patch +Thanks +Ive got the Javascript changes for part complete but Im sitting on them until goes out +So I removed the Ivy editor one and everything work fine. +Notes specific to applied on trunk r +I did some further debugging and it looks like the wrong URL is getting submitted by the function +Will fix this in +The patch contains all the public synth classes +At first it succeeds later on it finds the one under the build directoryHope this helps. +Yes please thank you i finally dare now! I have send a signed Contributor License Agreement to Tyler Mitchel last autumn +Thanks very much +By fixed I mean will users by able able to execute the following without being subscribed to the RHEL Extras channel and will automatically be selected to satisfy the Java requirementyum install can you confirm that this is a requirement? We need to add to our production plan and to the QE test plans modify the installation guide etc +I only know bits and pieces of the internals where Ive had to find out due to a Wicket bug or a misunderstanding on my part so I dont know if your changes are likely to correct other bugs +Have you considered implementing this functionality a level down in Hives JDBC driver instead of placing it in Beeline? One major advantage of this alternative approach is that it makes the feature available to all JDBC clients instead of limiting it to Beeline +Steps to verify mvn clean package the build should PASS mvn test all unit tests must PASS mvn verify findbugs warnings must be and the maven build must an offline conversation with Srikanth and he suggested that I disable encoding as this might give an impression that Falcon code base is Internationalization ready but we havent done anything conciously to guarantee thatI completely agree and will redo this one with out encoding changes +BPEL patch looks good to me +Care to take that on? And bonus points if you can drop dependency on agree that using are needed as otherwise the items are held strongly and thus indefinitely +I wasnt sure how the age value should be computed so I commented that out for now +What if the SASL mech is PLAINIn that case if the user name and password is empty we should probably throw an exception rather than silently setting it to anonymousAlternatively we could set it to anonymous but check them at the SASL layer and throw an appropriate exception that would point to the fact that the URL is missing the credentialsWe should make sure we provide the a proper notification as to what the real issue is +Aaron if you can add those test today or tomorrow that would be great otherwise well defer this task for the next release +The problem is knowing what goes on inside certain native methods of +Testers welcome. +there is nothing to do for this on the AS side +bulk defer minor issues to +I updated the SDK an ran the all test suite before to apply your patch and unfortunatly many tests failed and some relative to CSS and style so until its not fixed I cant at the moment integrate your modification +If somebody wants to go first dont wait for me +Ill now check the length limits +Will provide detailed comments about this patch shortlyThanks Heath did you apply the patch already? Im now finally back on working on this stuff again +May still need some improvements as it is slightlyAnyway here is the patch for review +Not with rowkeycfqualifier wayPls provide your suggestionOnce the solution is ok I can backport for other versions +Perhaps this is not a serious problem if very few people are using as old a release as me +It seems to be OK +The patch is on trunk rRequestion permission to merge to +target grep r rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopyarnsourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoopmapreducesourcesr rparker staff Apr sharehadoophdfssourcesr rparker staff Apr sharehadoophdfssourcesr rparker staff Apr sharehadoopcommonsourcesr rparker staff Apr sharehadoopcommonsources +Unfortunately I do already have that line in my solrconfig and my index is optimized as well +Finished +Even without the latest streaming changes encrypting the file on the server prior to writing it to disk could be accomplished by providing your own implementation along with a factory for it +Try IT +for +Please provide the username +A constructor argument isnt used as its optional and defaults to an +This will not be included +List of features installedfeaturesRepositoriesmvnxmlfeaturesmvnfeaturesxmlfeaturesjarmvn! Comma separated list of features to install at startupfeaturesBootsshmanagementspringcamelconnectoractivemq CharlesI saw you are using the cxf osgi transport in your and the s default port are Can you change the url of the jetty endpoint which use the port to avoid the port conflictThe port is only used by jetty on the osgi server where CXF does not run +The proposed fix +It doesnt apply to the framework. +Not done yet +Potential patch for the patch! +Updated patch because there were updates to maven release plugin when old patch was submitted +Please let me know if you have issues. +I actually dont want to add yet another dependecy to castor +Added getYNinputmock updated prompt in unit tests +On the other hand given that the bookmarks contain a link to the Table of Figures and Table of Tables in which each figure and table is listed users should easily be able to find figures and tables +stores and attempt is managed by +done +Also make sure to use Guvnor because Guvnor has also migrated to Seam starting with +And which may be used by other routes and whatnot +To me this seems both more complicated and gives up functionality unnecessarily +This is actually done and just needs to be confirmed with QE +Same patches but has the different directory structure +Patch tested on still working on verifying on but that might me take a little while +The issue is not reproducible anymore +Cdrik is it possible to provide a patch without the formatting changesI am having trouble seeing the changes you made tobtw I think only stays around for back compat to support old indexes in my opinion we should not modify it for this reason +Combined fix and junit together +Yes please open a new issue and specify the requirements what we have discussed together with a link to this issue. +Since it has gotten very silent hereIs it still planned to integrate this feature in Apache Cassandra +It works + but we should move the alloc outside of the file loop w suggested improvement +i found that a specific for this is disabled in file findbugs is right the code is you fix the two typos of hyphneation in the same file id completly happyproviding a testcase for this might be tricky the bug is deep inside and i have no idea how to influence hyphenation in fop. +I fixed the creation wizards warning but the other issue was fixed previously +But in my case it reads the stream and Axiom complains about some unexpected characters +Good luck +Im opening another bug for this so we can resolve this one +Switched back to a implementationAlso a caveat the current implementation wont be able to support the bootstrap from offset all the way to the head of the stream use case because only has the offset of the message and not how far the message is from the head of the stream +We would like to have no external dependencies for NIO other than proper and JRE so is not an option +Ill dig into this +Its pretty bad that the first command we ask folks to run after setting up their hbase crashes +fixed for beta +The patch looks good to you for the review +From JAXRPC SectionFor complex types defined using xsdsimpleContent and extending a simple type Tthe corresponding class will contain an additional property named valueand whose type is mapped from the simple type T according to the rules in thisspecification +Alex I applied your patch and testedThe tests should be ok now. +Interestingly I find no matching logs on the or DatanodesI feel like I have seen a jira about this before but I cannot find it +Well create any future examples in aerogear not as additional quickstarts. +srctestjavaorgapachehadoopmapred +Reproducible at at +Ive committed this resolving. +I just committed this +Incorrect MANUAL INSTALL INSTRUCTIONS the path should be without server part just JBOSSHOMEcommonlib and JBOSSHOME you please fix these issuesThank you looks good now thanks +I guess we want to include this on the branch as well right? The patch is probably created off of trunk so it doesnt apply on the branch +The attached file contains the changes I made to make it work with Tomcat and JRE +After setting the date format to iso I received actual timestamp values +Ive evaluated the failure in GC and failed the separate issue on it + is not yet released and still in QA +Can you attach a sample file we can test with +From Peter LedbrookRobert La Ferla wrote Let me refine my question +Why do we need the Trainer interface? As far as I can see all the work is done by Sequence and Event Trainer implementationsA Trainer Factory needs to be able to produce both of these trainer types but we can probably do this with two different factory methods +It seems much less work to me and much less disruptive +I see you are using toolrunner and you still need expexclude +After more work will be needed to get all conjunctions in the same place but it is a starting pointOnce all conjunctions are in the same place it would make sense to add a timeout there +Excellent! Geo is a complex topic too big for a projectLance +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +You have to implement in this situation. +txt +We are using and are experiencing the same truncation issue and would like to use the fix. +Please modify the test to actually have apps in multiple states +still actual +Thanks Sergei +This issue is marked fixed in release of the war plugin please open a new ticket for anything that needs to be done still. +Seems like the locations in the metadata table should exactly match what is advertised in zookeeper +NFS +The patch contains all modifications required to solve this issue accompanied by a module patch +for the property +ok let it work for any input +As a result a bit of refactoring were needed in order to avoid duplicated code between and as gif legend graphics got legend writing in gif to follow the same strategy than for and forced jpeg to run in non transparent mode or it gets a black background +But there are situations where I can see using both +By using this entity provider we have gain some proccesing time for getMap request we dont have to write the image in a file before return it to aleksander for having proposed this improvement! +The argument must be a date timestamp or a valid character string representation of a date or timestamp that is not a CLOB LONG VARCHAR or XML value +Making it according to feedback from Patrick Hunt +Is there plan to commit this to +Some compromise should be found +Its not obvious to the user why this happens and we need to implement the very same approach as we already have for mysqlphpmyadmin +Lets use the new extension point for both facets +type is perform full parsing and look to which objects are listed under root second version of the patch which doesnt break the build +That way the signatures and the checksums are automatically generated and installed into the local +I have attached a Rails project that allows one to reproduce this issue +No it doesnt do the same thing +If thats OK for you guys I would like to keep this issue open while I work on this +Reverted at r +for this +Havent really checked for regressions though +Attaching +And I used the word enter rather than type as it indicates that the text in the CLI should be followed by enter key +Patch just committed to Classpath +It demonstrates how to obtain a and deployes a file through the is a critical issue that must get addressed inInstead of just changing the issue from major to critical could you also comment on the statements being made? Since most of the discussion is not about the deployer and therefore not something that should be fixed in the can be derived from the comments this issue has nothing to do with the GPD +done +Patch to committed in r +It is not a question about what Seam depends on the framework can run on both Branch and is about simplifying the deployment on for the Seam users so the features from trunk should be ported to Branch so the resulting match. +gz so clean package should adequately ONLY build this if this works on the branch and if so I assume you have some web interface to throw away the upload and then I can port this patch to the tagged releases +Agreed +Attached patch implements execution logs availability on CC web interfaceWith this patch execution logs obtained during test run can be seen under Build Artifacts link +Sorry guys one more question +ant test is not required because only javadoc is added in this patch +I just have a clarification on this a remote repository will have its own metadata repository? +This algorithm was based on how the jobtracker does it now +This will issue all database operations left in the session after a roll back +Weird the patch did not change at all +It kind of reverts the changes done to exnathrowregs on Linux in commit +Fixed in r. +This raised another question that how we can guarantee that the algorithm converges if rebalancing is not done in the safe mode +no its not i checked +Yes +password is a reserved word in some databases so the script escapes it by default since if its not a reserved word it has no effect +Ill try again today +Closing again as it looks like a deploymentconfiguration issue thats outside the control of Struts. +It used a validate sql sending any query you tell it to validate wether a connection from the pool is still valid or not +Remember that we have SSL in the picture now +Ill the patch +it is MUCH more error prone than the form widget stuff as were dealing with plain HTML there +done +I just committed this +The bug is not observed in kdeA workaround would be to delete the entry if we are using the DE +I will be more proactive asking about things on the listbefore making changesThanks for looking at it +Any chance you could try the latest trunk code i could put up a snapshot distro if you cant build the complete trunk yourselfThe fix is still not spec complient though and this is an area thats being actively discussed on the dev list in the threads about databindings and wireformat +Exec source should be one of the simplest sources we have so please tell me what this added complexity buys us in terms of functionality +Jim is there something to be done for this issue or was it only to keeping track of the dependency on? In the latter case this can now be solved +As a result the was populated with instances and not the objects youd change modifies the logic to always wrap it with the version and it causes your script to run correctly +I believe this is complete please reopen if necessary. +Brian can you check whether my fixes described in the forum thread resolves your problem +commited +I forgot to say that the patch includes and from Adriano Crestani to enable the precedence code +Is there any workaround for this bug besides removing meclipse? Its messing up with Eclipse project explorer and several other functionalities +Also can you add code for as well +Hello TimCould you please this bug? I am running app server with JBM and the problem still exist +It does not define if the node has outputs and whatnot +If this is deemed acceptable we should probablyport this test back to the branch as well +Closing resolved issues. +ClausI think the plain STACKTRACE should be OK +issue for the +One small comment +Would be great to have a complete German translation but this is an improvement not a bug. +The query is different so it should have a different syntax +Any other comments? Would like to get this into x soon +how does this make testing easier? the updatesite needs the same level as testing wether it is there by default or not +Thanks +looks good now. +Is there any hope of getting this patch into a release? Its still a problem in +So your case would hold true only if there were multiple observers that would attempt to modify the nodes +Works for me Im able to build code with nonoss components mvn clean install Dnonoss P developersystemvm now +It updates to and make the test work +I have added a patch for the stuff +This is really the only thing missing for basic repository configuration to be fully in java +There are at least half a dozen or more committers that signed up for this project but the bar is so high to even do anything that I suspect thats why most of them have yet to contribute even a comment they dont have the time or motivation to go through that huge running LCF doc its a bear just to skim say nothing about go through the steps +So we would need to come up with a mechanism for updating without data loss +This is an interestingI strongly agree that it is an ideally I think this should be something that is done as trycatchfinally but since the case can occurfrom my looking at the test I agree that some sort of fatalError is at least a good start +I would like to avoid a design that incurs an overhead of an additional RPC everytime a link is traversed +But from Camel its all good again due some refactorings done in JMX. +Id much rather move the moving average computation into something like than to have it in hbase +were backported to as well adding the latter fix version. +For example lots of the commons live under these two directories netuiexternalstruts Struts netuiexternalstrutslegacy Struts So given that these versions can be different its probably best to leave them in placeThere are also multiple xbean in the tree and I think that at least one of those is and used by the controls test infrastructure +thats an important one + +Thanks Stepan for your patch +mysql select from patches g systemname patchlevel patchdate patchinprogress xplanner F +Thanks LeonYour patch is in trunk r r r +Testing on the cluster worked on a freshly loaded regions table that I disabledenabled a couple of timesAll tests are green +I make it final and lets just remember to add these breaks to the CHANGES BW break section +Commenting out only that one also produces the expected result +Fixed in svn revision . +Sorry no bug but I had a typo in the locale it would be great I there is a configuration option to set a default locale +this doesnt seem to have fixed the issue still getting the same failure +Ill reassign this to myself and start to work on it on Friday Aug +Were you referring to offline snapshot in the first sentence of the description +Jon any updates on the fixI plan to commit it shortly +Closing as a duplicate +txxFrom the looks of src however it looks like there were more changes than just that file +Thanks +The test and the thank you very much for the quick fix it seems to resolve the problem withI just wonder it it was right to remove the null check that previously existed in commented code +Sounds great to me +But this information has not got to the master yet +Good point +does that for you +table and it has startrow arg? Ugly but user wouldnt see it +committed to release note required as this is a in Developer StudioVersion Build id Build date to provide release notes +Committed revision . +I had set the time out to you confirmed that progress is reported after every second this job should not timed out for time out of it is still getting timed out for all the versions mentioned abovePlease test and let me know if I am missing something +As mentioned constraints arent quite done yet and these make for very good testcases as to the parts that still dont work +Patches the serializer to produce double quoted namespace attributes +just committed this +I have applied it in Rev +included in update site by including into richfaces feature +The issue will be closed +test added in svn . +Hi Jesse thanks for your review comments +properties files worksPerhaps you should add this migration problem to a migration FAQ or something like that + +Ill make the outputFormat a config option +Maybe the problem occurs only with MTOM messages? Simple messages without a MTOM Attachment are working well +for the proposalbq +Patch looks good +Mauros latest changes allow my app at least to shutdown +This has been fixed in the Portlet Bridge codeline but hasnt yet been fixed in the codeline hence the bug is left open as a reminder to merge the work +This is almost a blocking issue for me because I have to restart my server every night +The fix is to make the class public. +Tried abbreviated sample tried +I think this is fixed is that correct +Pushing to iteration of the patch that works with the new namespace code fixes the circular reference between ntunstructured and the default primary type of its child nodes and adds a can patch passes all of the Jackrabbit TCK nodetype tests save one failure caused by name translation +Whats even the objective making the compiler build fasterAnyone else want to weigh in +thinking more about this we dont need to normalize urls to contextfilter just to filter since contexts own their url space +This includes Daniels test and it works on my computer +I used the MB zipfile not sure about +Met the same problem with during profiling my WS perf testsShouldnt we assign it to someone from jbossweb +Seems to have index scans on the left side using key which is sorted on the value that the query in the end wants to be sorted on +verified by Vladimir. +clearCache is not really intended to force recompilation +I have simplified a bit the discussion on closing out ledgers + +Instead of keeping bad records ids in s TIP object I would prefer keeping it in along with other job files +fieldName tatGewicht type field number class +This simple patch adds a log warning complaining about not set jar file +Thanks! Fixed as suggested in revision +Also required is a zip file for forward to having this as a supported feature rather than being out on a branch o Olivers war plugin unzip resource dependencies changes to can you push these other patches to the respective plugins +Unit tests run fine the sort application is currently being tested with this +The fix looks fine. +Preexisting tests check that the change doesnt break existing functionality +Willem FYI I couldnt reproduce the issue with the unit tests provided by Sergey +AndersThank you for your explanations +Thats negotiable +But maybe thats fine for the first go +Cant remember currently but it might be that was stilkl released with M and the POM youre facing has been by Maven folks repository tools +The patch has been appliedPlease verifySian could you please create patches against Harmony class library or module root directory as it required by Good issue resolution guidelineThanks in advcance +Thanks +See for update support +If you have any suggestions on versions maven integration let me know +If you are cloning from git repo you might be able to find the revision number and hard code it in the script +Ive encountered the same Service Utility DAO utilityI need the DAO Utility module on the path for runtime but want to keep web programmers from referencing it directly so I set it as optional compile scope in the Service Utility and runtime scope for the build puts everything in the right place but when I try to publish using WTP it refuses to put my runtime scope jar in the path +How do I make the labels show up? Hmm +This is specific issue or we also need to make it for trunk +Hmm +Fixed the javac warning +It looks cleaner +My suggestion was for the post routing once the exceptions reach Seam catch +in +bulk move from M to M +Testing shows that the FEAT response can contain several entries with the same key for example SITE MKDIR SITE RMDIR SITE UTIME SITE SYMLINK REST STREAM SIZENot yet clear how best to handle that without complicating the user code +Resolved by reverting patch on trunk +Moderators can block posts from these users by subscribing themto the associated deny lists +It embarrasses me to no end to have to tell you about the workaround for this bug the name of the directory containing your app should be longer than characters +These blocks ensure that the thread locals are cleaned up properly +Now I have fixed them all and discovered there was also the file that could be removedPlease find attached an updated patch +Report Execution Service WSDL file is attached with this +applicationidentity is overriden in the file orgapachejdotckpc to specify schema +One thing Ive been thinking is how wed do the notion of LATESTTIMESTAMP if timestamps are set on the client? Client will always swap out LATESTTIMESTAMP for an explicit time +Patches with tests are always welcome +This is an initial patch for quick review +Fixed the comment I butchered and serendipitously noticed that I wasnt checking for null +java source +I think much of the changed code can be moved to so it isnt duplicated +No work in a year +See you there +Although length normalization on fields could be useful we should think about the best default +Bulk close for Solr +No its not relevant anymore given that we ended up doing it as part of +Hi James will encode URL Reserved characters such as etc +File contains the sources which are tagged as asfmigration over at +Similarly it is perfectly fine to have a channel and sink without a source for an Agent in order to drain the channel +bulk close all resolved issues in preparation for and releases. +This project supplies an AST transform that changes the package name +Currently the bottleneck is not in replication layer +Reopening this needs fixed in too +Case b is invalid since its the clients responsibility to synchronize access to sessions +Actually RBE inserts a new entry upon clicking in the edit area without typing any characterI removed all blank entries so that getting default entries instead for the time being +Do you think it could be related +Applied the provided patch +IIRC a big driver of that was when still had crosstalk issues on some and u wasnt adequate to isolate clusters +support has been added in r +They were deleted +Well the Right Way is to add a list of Bounds tombstones to CF SC classes. +Ill work there for now on your behalf +The virtual environment that I used for Linux times out and Windows fails with and without the patch +really? It looked like a to me +I just committed this +Stuff on has been integrated you can add CQL filters at your is more or less a duplicate +QA verified +patch has nothing to do with the test failrue +Should be fixed by r Farah please verify +A simpler test case with what looks like the same bug is also in Distribution +Thanks Jarcec. +Note this should not be committed until has been committed to trunk +This only adds to the precision of the timeout +This is the screenshot with the implemented feature +Weve worked with Contegix and Atlassian to get this have been reasonably stable for the last few weeks. +We cant use on by default because when you activate this env var with on mvn exit with B so the command line is closed and if you use it in a dos environment your window would be closed can try it like thisSET MAVENTERMINATECMD I said +Please if you think there are remaining problems +Those changes can then be made available as next point release of and since at the time of upgrade the trigger get marked invalid any changes made for will get automatically picked up +is now tagged and in the maven repository +Marking closed. +Merged and committed +The patch also looks good to me +That is very weird +That inline abstract feature type definition was intended to be returned applied fix to is not needed for trunk +Looks goodLets wait for a few days +So the tests pass in all cases independent how the compareTo method looks like so the ordering is not important because the queue never gets full +The commit message perhaps is misleading +After upgrading to cdhu it worked. +While it may be desirable not to have gaps in identity values the current implementation does match required uniqueness +Thats the only change in the patchWill commit by the end of the day if nobody objects +Will look at updating support to latest driver in or +The initial bookie is now running without a availability znode so noone can contact it +Lets try to make a patch +Marked the issue the very least we should not assign a region to a region server if it is detected as no good +Includes the basic tests for SQL construction class hierarchy +to why are deleting blocks that it currently thinks are part of the fileabandonBlock happens if the pipeline fails to be established at all so the never even get the blocks +Quite a bit of changes +Patch committed +Very sorry +But an unbounded cache is also not a good POSA practice +Do you have time tomorrow to discuss this via skype or IRCYup ping me on Skype when youre ready to chat my ID is maniksurtani +bulk close for +These items are still not resolved for Please review your items if not ready for please move them out to future it +Descoping items not being worked on for M into Unknown Fix Version for not is as old as the hills in that Qpid has always worked like this +svn revision +The same fieldeditors are used in new wizards prefernece pages and other dialogs +I used a variation on the patch +bq +Am i doing something wrong that these failures are happening +The specification of any reply is part of the business process defined and executed through jBPM +May be related to +Regenerate patch and bump version number of rpmdeb package +Ive submitted which is missing unit tests but is what Im looking for so that allowdeletes can be set without using environment variables +It somehow helps to identify any potential bugs in jsp code +This seems like a reasonable change to me but Id like someone more familiar with the job submission code path to review it before commit just in case Im missing a subtlety +Noticed that this has been a pain point in the past +Since we should be using localStorage well keep the priority minor +New patch incorporating Pauls review comments attached +Works locally see if Hudson agrees +How goes it +Sorry forgot to apply the patch for the tests as well applied that in Trunk as well +connect for remote servers implemented in SVN trunkOleg +Could you do +Adds checks for the SOAP service and reference binding +One other thought I have about paging is that it might introduce unnecessary complexity +Reclassifying from Bug to Improvement. +I suggest QA verify +Ive tried on the build as well as in my TRUNK workspace +Havent figured out how Mult is computed +Probably a sim number as those that already have to lower Im not sure there is anything special about if your already fighting for every megabyte +I ended up using the standard Lucene paging codeHopefully this will get into Lucene soon +wasa issue in the driver used +Patch attached showing exception throwing optionArbitrarily moving to concerns me +issue is now resolved by other commits targeted to allowing the use of workspace variables +What exactly is putting it on the lib folder? Do you have a POM or debug log to look at +Spring Security explicitly depends on Spring Security which I think is the latest +But I think I could look at practically any Java API that throws an exception and ask do I expect an error may occur? and get the answer yes +Updated the naming thats all +all done woot +Thanks +I understand +Perhaps since this issue was created a month back and I did some checkins with some CSS changes it seems to be working fine for me on Chrome +localizeCache seems big enough to split I also think a test case that verifies permissions are only set for newly localized public files is required for completeness of testing +Currently this block my company from using this plugin which would make life a lot easier if we could use the execution of my test I get a so its definitely the same problem and I think it has to do with jars with classifiers +was refactored into its own file and perhaps we can do the same for but I think both should atleast beLet me know what you think so I can make the changes you prefer +Will do better next time +Not even sure if anybody uses this at all +Remove the from the catch block +Any chance this could be ported to? Id like to see it in EAP are no plans to this +Of course we may have not been testing tracing at all or maybe not under the security manager +event not appears with JSP underEverything is OK except Tooltip in modal panel under the Opera. +I will also check if java still alows the permgen parameter or not +This may make sense to add to the load func as well but I agree we can do that separately if we decide it would be useful +Some sort of source formatting has been applied to the +Its shame its been put back +At first I tagged it suggestions but then I thought it might be confusing having two such tags on different other response writers like Serialized PHP for example always output an object or keyed array for named lists +Odd +Regarding jaxenjaxen Sebastian added because it was a cyclic through jdom to an earlier jaxen version or so but Jakob has verified that in jaxen the dependencies are marked optional so this issue seems to be fixed +understand George +Continuation of the work begun in this earlier issue +explained jbosssvc has an command line option to allow either automatic or manual service startup modeThere is no need for patching the source code. +Thanks guys. +Please reopen if not +Updated to be inside webLogic a default path for packaged libs and using a separate option to set the empty will work +Attaching patch I made to make the improvement +HelloI am attaching the patch to this issueI appreciate your reviewcommentsThanks KatheyThanks for the comment +Let me know if you have policy to do such things +Thanks for the report! +This patch is a replacement for jbellis originalThis patch applies cleanly against trunk +If a broken driver is not fixed it is easy to remove +subject to all tests patch for precommit tests +so are you saying youre not getting the standard foo is not a validreference warning in the logs when foo is a inner classif so yeah thats bug +JettroIf you are using maven to start jetty directly it will not work +Updated bugzilla with new stacktrace and test scenario to issue is fixed in webtools source editing master branch +the field have a Party look up that shows only the Person +When storingconfiguration files it might be useful if comments can be set for certainproperties +patch to branch +indicate when using LIKE syntax to search special characters +Regions cant be closed properly by RS for some reason +Need to get Cell objects in place before this will work +Hi BryanAttached javadoc fix +Can we close this report +This has been changed back to serialization +While this is fine for small bookies I think this could fall down for ledgers with many millions of entries +That said the default does appear to be true on my machine so Im now not clear exactly what is going on +Slava please take a is not a bean since its a class +Nice obscure one! Fixed in aa rubyspec update. +it possible to get this intoI cannot find as part of +that is not in the scope of +Same issue than concerns will fix this also +checked into trunk and branch. +My guess is that this is a classloading issue triggering due to the packaging of your application +This is a duplicate of. +sha bffefdfcfacabed +However sometimes they not fit very well so you need a redesign +The second for the address upto where the last skip list entry that was written left off +why should this be in central? central is not entrypoint eclipse wisesounds more like you said as a common thing +What made you pick version when version is available? Ive tested my assemblies using with the versions of slfj endstates and did not receive any errors +Inconsistent Override usage in +Closing old issues +Marvin does KSLucy have something like? If so what API do you use? Is it +TomCan you review Matts patch from Dec ? I confirmed that it works as advertised +Removing git paths in my gemfile helped +Andrea I finally got the time to test this and it works fine +Final patch for Navis +Ciao Fabio +The patch does not introduce any new Findbugs warnings +The same change needs to be made to the ant run sample output in the README for +there is a lot of tweakiing to it but once you get the whole thing configured to your liking there is not much to change afterIn terms of plugin development I do not think Java is that far from PHP so I do not think you would have hard time there +Marking these as fixed against rather than against. +jbdevstudiostudioeclipseplugins there is no I have found it only in jbdevstudiostudioeclipsepluginstemplateslibcomsunfaces +I do get the same problem as indicated in the original post +In inject and parse phases they are rejected by +PS is a sample project file +Patch created viasvn diff the first patch with the following changes Removed the rendering of the enabled versions of the controls if the entire component is disabled +I added a number of additional test cases to try to exploremore areas of what is and isnt supported including severaltest cases suggested by Army in previous comments andseveral more related ones that occurred to me as I was thinkingMy current feeling is that there is a definite implementationrestriction here there are syntactically valid expressions thatwhich Derby does not currently support due to the way thatthe result set of the UNION node is managedBut I also feel that issuing a clear error message is preferableto failing with an NPE so I think that this patch is an improvementover the current behavior +Feel free to pull the code if we arent using it or onto your own dev branch if you are still playing around +I was not able to reproduce your problem +No consumers of this api yet so still safe +Attachment has been added with description contains simplified input with word +By default userDetail object is now put to the principal field of the Authentication object +Would love to see some performance data with this approach +Perhaps its best to place the RAM tracking into where the RAMconsumed by deleted query terms and added documents can be recorded so thatthe proper flush decision may be made in it a central global object +I recommend checking your CSS +Can you supply a simple project that replicates this +The exception was a result of a bug in the jUDDI server code +Just try mysuggestions concerning the XML libraries and if it still doesnt work come backhere +It is likely to put a patch to resolve it +Im not sure why someone would need a job without steps as it seems equivalent to a job with a single Tasklet step +Fix is in CVS +would probably be less vulnerable to breaking if the migration of namenode data is managed outside of Accumulo +reg files in bin directory also they will turn on logging for and the CatcherAlso and should be registered +Stay tuned. +patch added +in +IGNORE THIS IS A COMMENT +I have removed all my other patches +For this ticket bar symbol is fixed + +If nobody is against it I will commit the patch later this evening +This issue still exists and shows this issue +Otherwise I will submit a patch for trunk and make another jira for +are you using the +I have some vague recollection of a DRDA maintenance version that could be bumped for such a fix + +I attached the output +As the does not have Service annotation the serviceName should be the not the interface +Should apply cleanly to trunk rev new rebased patch that includes the recent changes +Then visit httplocalhostcoreWebtestInheritanceoverrideactionAttrand compare with httplocalhostcoreWebtestInheritanceparentactionAttryou will see it is still displaying the same settings as the parentor you may check the generated struts file under netuitestwebappsdrtbuildclassespageflow +This is a problem in the test driver not the tests +Same patch as before but YARN only changes +fixed in rev think that there are more chars which are special to the shell and which need to be covered +Please verify +Works fine for me now +These blockers and critical issues are resolved but not verified +This means your whitelisting doesnt work properly like it does on Android as per +Continued on this I would love to see JMX stats of the +But I agree that if we specify a header we can specify the elements too and would like to see the patch +The pull request has been merged to EAP +Will commit this shortly +Done under the name of. +Committed to trunk at +However I believe that the way it is currently implemented will cause a problem +Patch added +First off there is no patch attached to this issue only a Java source file +Please close if you are satisfied thanks! +This issue does relate to which is still open +Ning Yup I ran the tests and everything passed +Works apart from the Getting Started context +This issue has been fixed and released as part of release +Im changing it to say that the hang is either in the JCC client or else in the Network Server since that seems like a more accurate statement +attached was superceded by +We might live with ids for the original estimate to reflect the time needed to setup the test and verify the results +Verified in +binhbase shell work? Shouldnt that suffice +Thanks for the patch Jesse +I literally would like to start cutting release packages this weekend +The bug is in ant +Please ignore the leading path name segments from the patch +This appears to be a type inference quirk that can arise in some javacs and eclipse it isnt in openjdk or java or clearly whatever jenkins is using +patch is relevant only to the kerberos setup where i manually tested it +Can you please verify and change that too as deemed +Finally fixed please test +This is available on already closing +This is good news +Got it +OndrejIm not seeing this using JBAS r and the latest RHQJopr bits +Note the html file will have to be scrubbed in order to play nicely with the release note tool. +Server is still cleaning the resources away on delete as it should. +This is the issue about classloaders we talked about the other day +Not everybody will be modifying the CQL grammar +is the time that the page is changed at last +In case of an AAE you instantiate for each primitive an own logger +Yea it passes for me locally given that the code paths dont change hscyncflush and append are already enabled by default shouldnt change the execution of this test +Not a bug +To simplify things people interested in the Bookkeeper functionality can simply download the patch marked with complete in the which contains all the dependencies and applies against trunk + +I will respond to the comments and upload a new patchThanks for taking a look +Patch applied +Ive just committed this. +I can only tell you that an Alpha should be ready soon +Inherit from mojosanbox +reading MavenPlexus documentation just to do simple things keeps me busy +Im fine with a but I dont see any reason to persist any flags in the system CF thats what the yaml is for +Maybe we should try designs +Update as per interim review comments from Kathy +Logged In YES useridLook at patch or patch Username hansd +I propose we make building of the jars optional with a switch do it for ant without the switch is too hard during development +Ill see into the complexity of upgrading to first then +We can not accept patch files that hasnt this license +This happens for package names which does not correspond with java naming convension +I hope with release this problem must have been solvedPlease have a look at the echo sample to see how service client instance could be +ASM testcase now resolved issues +So it turns out that the weird issues that we have seen for random reads only applieswhen the client is on the same machine as the master and the HRS +It should be Exceptionhandler Improvements instead of Expectionhandler Improvements +I think weve got a problem with our release standard maven doc files arent transfered when we generate the s doc isnt uptodate as well as the Javadoc have to figure out whats happening there +Must implement interface also becomes an operation that defines a refined receiver object for the subsequent call +Not in jbdevstudio s linux gtk installer +The problem with a centralized gitignore here is there are multiple with various files to be ignored your patch for instance ignores IDEA and favors eclipse +The cleanup command is no longer an issue moved this functionality internally to +Attaching a patch that incorporates Devarajs comments +Ted This patch is only for branch +Patrick the honest answer is none +There is one small problem in Seam Workflow workflow Pull Request Closed step leaves Resolution on bad value Unresolved +already provides ordering capability based on process instance idSo I think there is no need to add another sort methodIs it so +Thanks Girish! +released +This has basically stabilized the build for me. +Will update the patch soon +IIRC the plan was for me to give you a zip of attachments that have been renamed to real names and you would post it to some subdomain of +I missed something before +Perhaps at this point it would be better to internally just set the family to a known value like default or something like that +I would recommend that you post the real section that reproduces the problemRegardsGabriel +that would be wonderfulLalit NagpalCSA issue has already been fixed since +Changed it to use a private boolean along with getters and setters plus added some new constructors +Committed with revision +Jzsefcould you please attach a small but complete fo file demonstrating the problem +It will take up to an hour to propagate I believe. +Waiting to see if there is enough demand. +The important part is that people be able to do facettrue +This one works for me +Look at in +in ECS style is all titles use title case so all significant words must start with uppercase didnt write the docs and they are filled with inconsistencies of this type +has been marked as a duplicate of this bug +LG can you please look into why nsapi is missing from CR? Thanks +Sort out the basic set of navigation items which includes +Namenode should never throw with HDFS as the lease holder +It should become a child of c it does +attributes +After all the control is there which enables anyone to write their own implementation +Updated the fix version to be and updated the change log of the jira to be in the trunk section +I have submitted this patch +Even developers coming over from Java to +The result is similar to those produced by the acrobat reader +Konstantin I think its pretty much the same race +If theres interest in using Geronimos shared version of Dojo then FYI it was recently upgraded to version via so it should now be compatible with your application +And I doubt that these downloads take up much server resources other than the memory used by the httpd slotsOn the other hand Ill show my limitted knowledge of maven by asking is it really a good idea for many people to be downloading unvalidated snapshots with an automated tool +The line numbersdont match the source files of the versionCould you please try again with the release or ensure that there is noolder version somewhere in your classpath? I remember there was an issue with when the file to load does not exist but this was fixedinThanks +Committed to the trunkAttached is a patch + +The bug is that the job is marked complete when speculative tasks are still runningAlso we should discuss the issue in more detail and fix related documentation +BULK EDIT These issues are open to be picked up +indicates a error +Now I can proceed with my work on Spring to hear that it works for you now! Ill close the issue then as theres nothing we can do about this in Spring +Latest patch looks goodTrivial update to add the missing configure goal +siying the patch is not for this jira +to your proposed fix +While this sort of points out that RTC isnt going to work we might be able to sidestep this technical problem by creating branches instead of patches +But compounded with it means that cluster recovery is interrupted and some caches may end up in a corrupted state where its impossible for new members to join +May also consider killing this wizard in its current form? Look at some of the Workflow recomendatinos for guide as what is has been replaced by a system of import pages that we include in the middle of the Add Layer of these concerns have been addressed. +Do we just want to email a private expiring URL that allows them to change their password? Standard fare +Why not force this change on another set of people where it will actually do more good? Just because the other group is larger? I think for the sake of the future health of your framework its important that you can make unpopular decisions that benefit a clean and consistent API + is using the new mapred interface and produces the correct output +Everything works great +Right +Moving to +This is actually the behaviour that was in place beforeIve also updated the javadoc for to clarify that it should only be called for classes +for example say torque was using and resource manager node was set to the wrapper script would allow you to change it up so that torque would say only had and the resource manager would have +You are seeing end of line corrections +Committed +Why load all the databases if your not going touse them or even not sure if you are going to use them? I mostly use multipledatabase for configuration purposes and such so I dont use all the availabledatabases in the file +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +It changes a very longstanding Velocity behavior +Applied the patch in as has been released. +Freeman here is the test for the branch +should we remove spans? Or is there functionality missing in intervals +I was hoping to contribute to a C version but Ive been too busy so Im glad to see its underway +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Using JDK and MavenShouldnt be a charset problem either because you get the Chinese symbols in the error message +Yes i expect +sorry wrong button +If you feel so inclined open another bug with the additional fixes +I think it would just get lost +There are other projects hosting CDI portable extensions such as Apache +This file passes the Woden validator +Fixed by restoring behavior create VOLUMECREATE event at not in +So yes this issue is unrelated to and its patch would not address it +This dump shows that many threads are trying to get lock from and that none is holding that lock +The other services do currently not support this issue was fixed for the Ontology Manager +bq +Now I hit a different failure in the environment an attempt to load the XML into XML Schema twiceSomeone is going to have to find a way to convert your stuff into test cases +Ahha I got it +FWIW Hibernate Annotations does work fine in Springs JPA test +This issue may provide a higher level way to define these example transactions using business process descriptions but is not urgent as the current mechanism is adequate. +Nick mirror search should not affect when the download is actually running should it? That is only used at of dependencies aint it +check your browser +Heres a new patch listing all HTML elements that are missing in the file +You reopened without a comment! Please provide more information +Fixed to avoid during core shutdown when JMX is not enabled +At this point in the chain there should be no difference between a GET and a POSTOne more comment this may already be fixed +It might be cleaner than doing it in Ill add this to smokeTester +Id say its fine as a short term fix if you need to get this issue going +For now Im just going to stick with the dependency structure of the existing build but without class file duplication as Dennis has done so far +Im afraid this is not trivial to consistently support since in order to determine a definitive element type of such a nested List wed have to traverse all nested Lists and check that all of them only contain String values +just look at +The error is really harmless but its still good to fix and avoid littering the error log with these +And we want to let jbpm be a bug fixed version so there should no database changes +Yeah I mean basically that constructor shouldnt exist at all since it isnt setting client id either +Here is the test case there is a detailed readme for steps on how to MartinThe cxf se is not designed to do the schema validation since the cxf se is inside JBI container so generallly handle the Normalizd Mesage which is not the soap message so schema in wsdl doesnt work for it +I get this message when startup all jbds plugins from eclipse with it does not reproducable +As for the builtins subproject at this point we havent really started using that mechanism for new and Im a little worried that we may have remove some code rust before we can get it to work +project +usrbin +Thanks guys +Alex I still have to add at least one new test to the CTF test suite before I am able to commit this issue +I committed this patch to trunk and + reads from disk writes to disk disk space used Amount in soft cache that can be reclaimed in gc maxI do not think I understand what it means by concurrent bits set X block size +You could probably set them to javac but it would be enabling JIT optimizations when running javacTruly there is no optimization magic in javac besides this O flag which we should set anyway +But I just want to prove that its a bug in version as I am not in a position to upgrade to as its a beta +I just saw that trunk has changed since my patch doing the same upgrades to instead of that I have since done independently +Thanks for your understandingFreddy +test pass without problem no feedback no feedback +htrace is available in maven now +Closing this issue +Quick question In your application is the invocation on the EJB happening from an EJB +I will leave the defect open until that need can be addressed as much for the reviews and suggestions +Which revision? r may have fixed this I believe +If someone has a suggestion I can give that a try +I have removed the p configuration allowing to update from as it is too buggy +Move blocking from to +If you like you can also do it yourself +Because initially I thought they are not common like the other attributes are +Thanks for reporting. +Thought about that but theres two problems +It respects the new default buffer size +Thanks Sandy. +I think that this closes the security hole and eliminates any need to expose these properties in the user guides + +Not sure though so Ill write a test or two this weekend +Patch one was cute +Note java has not been filtered outAdditionally and are present in andThese containsjavaxcryptojavaxcryptospecjavaxnetjavaxnetssljavaxsecuritycertas well +Should be ready to go +works with so im closing this +My understanding of having it evolving means that its interface may change +has been removed from source control and is now considered a platform dependency. +I am still working on the ant build script that will enable the plugin builds from the source code +Any update on this? This would be a much welcomed feature indeed. +As the Simple model now passes all of the same TCK tests as the other writable connectors the fix for this defect is to switch from the simple model to the basic model +Gotcha +API extended accordingly to the d like to remove from the short version of the as they are usually unneeded and always available in the long version +I just reviewed the patch and was going to commit it now Uwe +Integrated to and trunkThanks for the reviews Lars Stack Anoop and Ram +Ive modified to check if addressing is turned on in the WSDL when its setting the default +looks good to be applied to to the branch too +Assigned for more information +Will take care of it on commit. +In true it prints all stack traces so simply try with Java and Java +We can only develop a fix if we are able to identify the root cause of the issue +We just need to write a implementation which can handle the Literal +Verified on QA implementation via warning seems meaningful so closing this issue as resolved. +Reschedule if these happen to get fixed sooner +Also would this need to be packaged up into the distribution or deployed to the web somewhere +Ditto for xDEADBEEF et al need to be rebased on trunk seems to be in the right spot in saveNamespace to me +That class doesnt use the remediation have fixed that issue in PR uses the same classes in central and project examples +In version does not work for listsNicolas can you give an example of what you were trying to do and what is missing? It is possible that I hid something when I refactored the codes +The part about the SOAP action was absent in BP +this has been reopened so I am moving it to TBD as it makes no sense verifying the fix if it is not considered fixed +Anyone see anything wrong? If not Ill commit in a few days +You will see an error message like thisHtestscripts x is declared in a dynamic context but you tried to access it from a static context +Sending assemblysrcmaindescriptorsSending assemblysrcmaindescriptorsAdding examplesinterceptorsAdding examplesinterceptorsexchangeAdding examplesinterceptorsexchangeAdding examplesinterceptorsexchangeAdding examplesinterceptorsexchangesrcAdding examplesinterceptorsexchangesrcmainAdding examplesinterceptorsexchangesrcmainjavaAdding examplesinterceptorsexchangesrcmainjavaorgAdding examplesinterceptorsexchangesrcmainjavaorgapacheAdding examplesinterceptorsexchangesrcmainjavaorgapacheservicemixAdding examplesinterceptorsexchangesrcmainjavaorgapacheservicemixnmrAdding examplesinterceptorsexchangesrcmainjavaorgapacheservicemixnmrexamplesAdding examplesinterceptorsexchangesrcmainjavaorgapacheservicemixnmrexamplesinterceptorsAdding examplesinterceptorsexchangesrcmainjavaorgapacheservicemixnmrexamplesinterceptorsexchangeAdding examplesinterceptorsexchangesrcmainjavaorgapacheservicemixnmrexamplesinterceptorsexchangeAdding examplesinterceptorsexchangesrcmainresourcesAdding examplesinterceptorsexchangesrcmainresourcesAdding examplesinterceptorsexchangesrcmainresourcesspringAdding examplesinterceptorsexchangesrcmainresourcesspringAdding examplesinterceptorsSending examplesTransmitting file dataCommitted revision . +The logic applied here could potentially be extended to other image formats but right now only applies to TIFF +Some of these options may not be supportable without a lot of work since we cant fork exec like MRI +Yes checking the actual code it seems that does not detect if two components that are instance of are duplicated on the tree +I checked out a clean copy of TRUNK applied the patch and uncommented the ACCUMULOHOME system env setting in the file +The problem is related to the declared variable in table TEI which cant be a runtime expression +How disappointing +we fix thisMy preference goes to reverting because I dont really like that we add something were only mildly getting consensus on to a releaseAs a side note I would have wished we didnt end up in that state of not being able to release because of some debian problem +includes all changes from the other attached patches as requested +Revised backport for yhadoop patch in sync with latest trunk patch for yhadoops in sync with trunk patch +Thanks Kevin +sorry but we are not going to fix thisbasically this happens because you end up having multiple classes with the same name loaded through different classloaders which is an artifact of using reloading class filter. +Plus the jar is not working + +Sorry for coming in very late +in clientimpl it is not necessary to parse address from string each time creating channel +Patch with test +Commit cefaadfabdbbfece +Added the info on thanks. +So for RampartC is a must in the receiving side +Andrew just warned me about it when I started +I think it would definitely be better if thing inside the reactor worked as expected +That bug isnt reproduced on the. +When prompted for a workspace I chose a new empty directory Go to the workbench and Open Perspective +There are installations where hsynchflush is disabled using +Project interchange file for remote CXF webservices +Thus they never be released +Thanks DhrubaNote your patch included some improvements to the descriptions in conf but these seemed unrelated to this issue and I assume were intended for some other issue and included only accidentally so I ignored them. +I agree that FS action should also move to LMIn the first step we could move the prepare and then in another stage we could move the FS actionI would suggest Kiran to create a new JIRA for FS action +Resolved +Hmm +Does the blank application refer to the blank webapp under showcase? Is this issue to make the blank webapp under showcase a standard maven archetype? I think we have a +I should actually further comment that for this use case it would be awkward to use Avro precisely because we dont have a metadata channel +You could try to wrap content of facet in hpanelGroup and manage its rendered attribute maybe it will fix the problem. +I think around Christmas time I should be able to plug in a nice testable setup in there +Does the problem look fixed from your end? +We will be moving to for community and builds soon and the referenced WS fix appears to be included there +The findbugs warnings are about but not related to thisHowever the failure of is related the cp command usage message is changed we need to update the test +Thanks youre right I did not notice the controls in xircles. +And with this feature enabled hibernate will only be able to support more datababase schemas +Unfortunate how Lucene has changed so much lately that we cant backport this +Go for it +was able to download a artifact on could reproduce in so I think this is fixed by upgrading Jackrabbit +Now we can improve that stuff with source codeThank you for reporting +Indexes +I guess this must have changed since I last looked. + +So if you are testing a BPM process which is not using human tasks do you still need a human task server? Of course all the stuff in the is related to human task processing so maybe its not a huge deal nowMy comment on the port is that the human task server binds to a socket +We need help on this one +are typically only used when developingIs that consistent with the way others use these +Marking to fix for and but it appears to already be fixed in trunk +Only two bugs are actually fixed between RC and +prior to generating release notes +Use Integer for context identifiers instead of Strings +Thanks Liang! +NET. +running hadoop mradmin refreshQueues + + +I guess my question is then did you have the setup properly or were you running OOTB? I havent dug through in depth enough to confirm if this is accurate but it appears that the store is created when the poller is run and when the store is closed is should disappear forever + but needs a WELDINT issue +Its still on the TODO list but keeps getting pushed down when other priorities come up +for patch +This patch attempts to separate out the block manager code from the namespace manager code whereas makes the pluggable +In my opinion the scripts shouldnt change the current working directory otherwise it would not be possible to use them like usual commands on command line +Fixed which is targeted for JBDS Beta and it works fine! +Resolving issue! +Encountered when deleting a templated role in using memory rbac +Weve move the good queries Ill deal with the remainder in a separate issue +When I emptied the field the creation worked +At a suggestion from groves on IRC I tried making my class extend instead of +Reinhard as you are busy with preparing the release I decided to give this one a shoot +This one is working now! Thanks Uwe +It looks like the culprit is that the gc deletes the parent directory and then cant find the files that were supposed to be in it +committed fix to the trunk +Please check all is fine so I can close +Scheduled deleting domain INFO NULL Started Starting job for deleting Domain +Data stored during async service calls +Ive tested it locally with an extra type registered with the uncommitted type registration code +Newer patch +Xslt component doesnt use the as the resource resolverI just checked the code of Spring it will use the directly if the import resource is not start with classpath +We dont have resources or time to do this on the upgrade so I am moving it to the will also inquire about supporting the in a EAP regards and thanks again for all the improvements +Im going to keep it open until Fil gets back then close it once I know whats going on with CI and this task +Where did you get that version of the installer to test on? It was recently respun and those issues were addressed +Rebased on committed +I think in your automated conversion some of these have crossed the columns +Since this is a little bit confusing Ive created a new Jira +In cases like this it makes perfect sense to mock out parts of the system that you arent actually interested in +Right created the and the repository dirsCould you give the script a whirl pleaseThanks +Id like to see a clean test run on Open JDK before we produce a release +Just to keep the discussion on the same page +just run ant run check downloading a file from +Updated patch +looks good to me +The only place where I saw this in our code base is in the script where it looks for HQ in some database row +The view server protocol is designed specifically to build and generate views not to allow streams of information to be arbitrarily processed for concernsIm perfectly on board with the desire and effort to couple couch to better and bigger things but we should not be making an explicit statement that abusing the viewserver protocol is the right way of going about that +Trivial patch attached +Thanks Sam! I was under the impression that After would be the last method being executed +from me as well +Please could you attachupload the patch to this JIRA part of that step is giving Apache the right to relicense the code in Apache projects +This time with license for inclusion +Are the route definition and the message containing the fault andor a trace log available? Thanks + checked the code both in x and and it in fact uses the actual number of optional of terms generated in the not the number of terms found by edismax from the original query as stated here in the issue descriptionI suspect that this is a variation of Edismax parser does not honor mm parameter if analyzer splits a token where the Lucene query parser code passes a single term to the field analyzer and gets multiple terms back and auto is false so the terms are because that is the default operator +I suppose it never really occurred to me that this was unexpected behavior since I am using namespaces to separate thrift code from my code and only have one namespace to worry about as far as thrift is concernedCould you describe the expected behavior for me so I can try and fix this +Stack dump from the broker which is up but refusing to start new protocol sessions +A lot of functionality in depends on opened projects and to open the project means getting the project in the top level +after some discussion and research here are my results plugin mapping does not have any meaning for an archetype LATEST update isnt used since for a long time archetypes were packages as jar packaging which does not update LATEST without any problem packaging is here only for and definitely removing this mojo wont remove any functionality only unneeded copypasteIm going to remove the mojo +I am working on a more efficient implementationof Servlet and will create new issues + for the patch +Jimmy is already aware of this as he got a CC of the EAP weekly report which lists NEEDSINFO JIRA issues and to whom theyre assigned +Yeah thanks +Imagine if Oracle or SQL Server had a flaw that didnt store data correctly +Waiting for EWP build to close the issue. +Thats why this ugly stack trace slipped through +Can you add this to your patch patch to include php and java and added following output to configure Building C Library +Reopened +Btw the purpose of this is to allow me to move state from one context to another +I wasnt thinking along Stacks lines +Most of the projects I have dealt with either uses commons or slfj honestly I have never used the logj api directly +Sorry +Seems we need one more step to unjar script files before invokingId like to commit this patch before we branch +we are planning on removing automatic cookie support from formcomponents +Added a test and cleaned up xml test files +The attached test case makes sure however that adjacent CDATA sections that ends in does not get merged with one whose data starts with and the same for one whose data ends with and the next one starts with +In we had a bug when the same query runs many times +but isnt this enough +Closing issue. +However the code still compiled and I was expecting a compilation error +Added without me +Some improvement was made but I imagine more could be madeAs a side opinion Checkstyle patches tend to touch a lot of the source and quickly go out of date +The last jpa depolyment doesnt work properly +A new patch addressing Myrnas and Dans comments +This bug has been marked as a duplicate of +I just committed this +ASF ID for the late response was traveling +IMHO this is about ready to be included with please. +You may consider correcting the commit comment in revision it says its fixing the problem of but that bug actually is the task of running the tests against a collated database +Pretty hard issue to know exactly what to do here +I left it for future jiras but it could be added later onGood to see no known issue replication blog for some time +Your patch for included this fix and I applied that part of it without noticingSo this is resolved as a part of r. +I love this +Ive committed this +components arent synchronized by default which is a problem if you are using state saving +Radim found out that the M version is actually from WFK maven repo +Please go ahead and commit the fix +for now I suggest you use meclipse or similar to get automatic download of dependent libraries +a patch for branch +Can you reuse for test instead adding a new test suite? +Theres also some pure changesAlso you should add a note in the file to indicate that the new algorithm exists as well as the fact that it requires the couchdbmachineid setting to do what its supposed to +push to +No longer relevant +We can certainly look into this once we know whats going on with the ESB runtime for the next release +and are incompatible in other areas +But that is one of the assumptions that the sink makes the table and cf are present when the sink starts +Agree lets shoot for to release. +If you have any egregious examples please create a new bug for them. +Im going to commit this tomorrow +Seems reasonable changes were made apparently in the timeframe another JIRA should be raised for any other work. +Should these just be listed as provided then if were assuming they will come in with the hadoop installation? Then we can just specify a range of acceptable versions for our code and let maven do the dependency resolution based on the hadoop dependencies +I have applied the simple workaround I suggested above to the plugins but not yet released a version of the plugin with it included +i will change it then and keep in mind in case we run into odd issues permission wise +Refreshing tags table on path input and path validation were should be added +A simple initial scheme could be to have a system attribute for each registered by Derby +thanks hong. +Based on feedback from John Graham the name will be Eclipse BPEL EditorIve updated the BPEL Whats New +Lars has found some clue in m going to rerun it on trunk and on a weeks old trunk as well +Thanks Perttu +but I will close the bug anyway. +Updated patch with javadocs inconsistencies corrected +Sounds like were all on board +Probably code that is getting messy with the internals and is not expecting compression +Bumping to since its not very critical +Additional information Im using JDBC Driver +yesconfigure WARNING present but cannot be compiledconfigure WARNING check for missing prerequisite headersconfigure WARNING see the Autoconf documentationconfigure WARNING section Present But Cannot Be Compiledconfigure WARNING proceeding with the compilers resultchecking for +Patch for trunk +Merging with. +Patch for Query API changes and associated changes for PMF property +I dont think its practical to do this. +xml the jobs pretty much done already +pushed +Clebert Im assigning this to you until we get more details +this is a bug in version please go back to version or wait for upcoming version +Committed to trunk and patch needs to be to trunk. +I am sure there are new and old classloading problems doesnt seem we have problems along those lines anymore. +I think it is a good one to start with +German exhibits problem for a good bug report! Fixed with integration test in r +What I meant to say was Scala is fully supported in trunk +Huh +fixend on too +Assuming admins can add new committers can we go ahead with us as admins and well encourage others to identify themselves or create accounts? +Passed test. +I only changed Lucene classes if that helps +I tried to expand the as far as I understood the workings of Forrest +Still do not see what is wrong +However Ive applied the rest of the patch at the revision +I am fine with removing it +New patch with stored fields +Hey guysJust wondering if this will make it into +It doesnt apply when you shutdown the entire +Im getting a clean build as well. +Sorry. +Please provide a simple test case and compiler arguments used to compile it +that means between beta and +The war is generated by the maven build and looking at the dependencies there seem to more dependencies on as there are on in the project +If i get some time later this weekend ill grep around the docs for anything about the old interface +Has the circular dependency between the server and security modules been broken to do this? If not it should be +I have resolved this issue because the intended outcome has been reached. +In all our use cases the source file may be deleted over time but its content can be still accessed through hardlinks +bq +generate files with sizes k and k for k +Bear in mind the jars were built and my tests run under Sun Java good in the meantime resolving as NAP. +Will will get this into the release which will hopefully be out in the next few weeks +But now at least it can be built +I would very much like to see the suggested improvements +tweaking issue Summary to describe the more general problem with Fixed in +committed to on +DaveYes this is all good +In fact I was able to create a new user +The seems to be more that the timeout is not long enough +Contributing patch for the mentioned workaround in description +These fixes were ported to the branch in the last week +Funambol uses an incompatible license so the integration code can not become a part of the projectMy personal preference would be to make a client +marking as fix version so it shows up in the release notes. +Have not looked at code in detail yet but it looks like you are missing a module dep +Imports are handled differently in +There is a get method where if dataSource is null Nullpointer is including Jaceks revision Thanks Manu! +Ive committed the regenerated design doc to trunk and +Attached a patch for testing this fix. +ClausI dont think this is an issue +By the way I dont think the patch is quite correct +I see no real way to solve this +yup fixed thanks. +Worked with on Windows +resolved by downgrading caln we should only upgrade the library once they fixed the bug +The patch to exclude these new test cases from test run +also look for any typos extra jars archive tag etc and make sure you are using only the jars that are signed and nothing else +In that case the client will call openScanner again so as long as the first seek of any newly created does a real seek we should be OK +I agree +It will also match random paths like abclib +whirr currently has jclouds not +I dont fully understand this yet and it doenst have a test but i see a patch and recent comments from david that the bug is really and the patch fixes the bug so im going to look into it +See examplesfo for a test case that works now +We do this in direct mode an this used to work fine in +Closing this Jira +This patch does not apply to later branch +Just some var name typos as issue title indicates +resolved at revision by committing Brians patch whereby circularities of containment are spotted when building during serialization by identifying a root object and detecting revisiting that because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Feel free to assign back to me if there is something I need to do. +It would actually be nice to not add tests that cannot run to the testing list at all +Posting my comments fromThis jira is essentially makes comments on the changes of +This is the one you committed! Marking resolved. +Its xulruner specific behavior show context meny event doesnt occur under select elements +I would rename CIRCULARREDIRECTS to ALLOWCIRCULARREDIRECTS just so that its more readable on its own +When there are multiple dependent sets such as with a join or a composite key then we will do not strictly enforce the max +Ill address your comments today or tomorrowdaryn issues it own delegation tokens and they can be used only against the service +When the Code will be comitted to repo +This was initially only fixed in the Condition statement +I think the value of having a real resolver outweighs this use case +The Fabric agent can still install as part of a profile installation but you can no longer install those using a plain osgiinstall +There was extensive checking of this issue earlier in the week as well +So its really all clear now +Alejandro I know you have been looking at and thinking about the classpath issue with YARNMR and how we should package things a lot already +Marking as closed. +No workaround sorry +Cancelling this patch since it hasnt moved in a long time +If we did Id want it to have regexp functionality +builds have fixed the hostname problem and this has resolved this issue +bq +Jetty done rev +The transaction metadata object is expecting the session to be a long +per comments appears to be an incompatibility between URL encoding strategy and servlet container configuration +Heres a new version that fixes things so that we can catch rather than when attempting to read a past EOF +For the moment I will just move and see if that helps +proposed solution from Ivan Burskyy +This patch provides some ASF requirements to release artifacts and ASF repoYou could also need CIShould be done in incubatorshindigtrunk added to repo also added default module for javagadgets and myself as a developer. +From your post I was able to make it work +Commit will follow shortly +fix in with. +go for it +Found another that needs to be shutdown +Verified +Theres a lot of caveats to this solution as a general solution to this problemWe would somehow have to assume that is on the same device as your build if not things will probably get worse +ThoughtsI will put it to branch tomorrowRegardsLars +Changing fix version back to AS and increasing priority +We will upload to when we branch updated to use this revision +The Jenkins instability is covered by. +The page would still render correctly for me its just that the action for would be called instead of the action for when I navigated to +I have modified the way Writable is getting created +My resolution for this problem is in attach +Also please note that I worked off the release tag not the branch nor do I know if this is an issue with the trunk +I dont see the meaning of these two concepts in that scenario +Im gonna try to reproduce +That will save the work of creating lots of small objects just to throw them all away +Sorry i must have misunderstood youPatch revoked at r +Window check moved to. +Agree Gert I think we can close this issue and work on the other projects +xsd +anyone can help me? +We run with runit so yes it is restarted when it dies +Please dont hold this release because of the XML Importer +its +Suns JVM also has the property but not +I got puzzled trying to reproduce in linux moving to windows to see if I can reproduce +The list will emerge as a part of work onCurrently few issues block this as the Javadoc plugin doesnt advance to generating the javadocBut I will at least check the repo if it has sources for all jars +So disregard I +My test suite are now all believe the problem has been resolved. +Any objections to this versionThanks for tackling this TODO item +check in and the qpid +In the previous build a different test failed in each pass +Actually I think s does this well? +Im not sure +We should keep the JIRA open until the fix is in the BRMS product code and can be verified +If ftl is ok then Ill do itAnil s ok with me too but maybe somebody knows how to do it with form widget so you dont have to a whole form +Thanks! +Neither Maven nor are passing the corresponding IT +My original intent was to keep things similar to but Jonathans comment made me realize there was little pointlets just use raw keys throughout +Revision move definition and effective node type related code to nodetype plugin package remove duplicate code in add missing test for protection in can add TODOFIXME comments in for code that seems insufficient to methe test for items being protected is still missing with Sessionmove Workspacemove Workspacecopy Workspaceclone +Thank you Ivan +ah tricky + Completely sensible +In the sample application the field myId is of type Integer but validator used is requiredstringI guess thats bound to throw validation errorUse typerequired +Added Spotlight to the homepage +Update the priority and fix version +I created to use as key in the cache instead of +I should have run this test on my other machine also +Solution decided is to rebase to. +At this time no thread is enqueuing more data +It produces behavior +Ning Can you please delete metastoresrctestorgapachehadoophivemetastoreIn the backport patch this file is deleted and replaced with but it looks like for some reason this file was not actually deleted when the patch was applied +The install guide contains instructions on building your own war although it would be nice to be able to move themes outside the war file for deployments on environments like clusters +Therefore we need to write custom logic to handle these kinds of scenarios +Committed to revision +Testing on a real cluster pending +Currently these must be set as system properties for the URL Handlers implementation to find them not in the file since this is not really something that is configured on a framework by framework basisIt set the property using D when launching the framework or the launcher still supports a file for automatically setting system properties +JayThe issue looks please take a look and remove tran label if you agree. +The current implementation supports a simple naive config +This means that when someone invokes mkisofs on a Ubuntu machine with a correctly installed genisoimage package it is not actually mkisofs as in the software maintained as part of the cdrtools package that gets invoked but genisoimage A side effect of the ubuntu symlink above is that the experiment of replacing mkisofs with genisoimage is actually tested by anyone building the cloudstack project on a machine running Ubuntu +Other than IMHO making it harder to locate the test that relate to a particular classmethodI am happy to hear if this is wrong if so then can we please get this finished and applied +I could take a crack at it +Revision +As far as I have been able to determine there are no problems in this area with the Struts release available for download on the Jakarta web site. +for the patch +Thanks that would indeed be a good solution to issue Ill also have a look at those classes you what is your take on issue +This example is fixed in +If you are still interested in this issue please attach a file that allows to reproduce it +A fix has been applied in cvs. +After committing the issue is and could be closed +It would be fairly easy to update the faunacassandra gem to use an external thrift binding source +Ok thanks Robert let me know if you find a specific scenario where this happens more frequently so that I can try it out as well +If we arent supposed to be using then there should probably be documentation that effect +bq +Unfortunately there is currently no good way to determine the JVMs float ABIYeah I was afraid of that +Now that work is underway with AS all previous community releases are +Mikes patch is now in CVS. +Ill commit shortly +It dont have to break this more general issue is any batch or any composite operation that deploys a driver and in the subsequent step adds a datasource using the driver will be affected by the issue. +Hi Robbie can you review this commit please +BravoIts great to see someone else learning the DIH code +Geirthis still a problem lots of awt and swing tests fails because of this problem +Still prefer simplicity of design to raw loc as the right metric to apply hereOk I yield +Is this likely to be fixed at some point? +My project requires distributed facet pivot so I have written a patch to do this and attached it to the issueThe patch affects only and I have tested it against Solr trunk + or would seem to be a more correct and less hacky solutionTempted to close in favour of one of agree with Stephen should be the way to go. +The feature added in master with commit id cecfcaebdceacedeaClose as fixed +The feature has been added to Synapse branch and the trunk +After discussing it in list Id prefer this issue not fix in Branch +The are not included in src assembly as they are not generatedI dont think javadocs should be included in the src assembly anyway as I said above it should match a checkout of the source tree exactly +Deprecation isnt working for MR when client and cluster are on vs still debugging + +And what does resever means? An classlaoder is already setup normally +add Junit committed to trunk and +The proposed change to may solve the same need +Thanks for the changes +For the Cascading use case I really dont think it will be that much of an issue since most data processed is not taken from those tables but it might be confusing for users that might start using these properties in some other context where it will cause more troublethinking about it some more I dont think that it is a really bad option to add these to classes to the mapreduce package and not pollute with more stuffWhat do you think +The output for me is parent pid child pid which shows that they are running in the same process +The page seems to cause a lot of confusion +Folks Do we have more updates on this issue or is it safe to mark this as resolved? +ArjenCan this one be closed +Applied in r. +I misunderstood +for the patch +Yes I agree +Manually tested +Two of the files were new files and thus there are no patches +That seems good to me +link parsing has been improved somewhat +FWIW if kill rather than freeze then I dont think that is efficient from a system throughput perspectiveYes Vinod +lgtm +It was fixed by. +my bad +Could that work for you +The right class should be now generated but the anyAttribute extensibility is still not supported. +Yes it should be documented in the java docs +There are known failures +Wrapper class to solve IBM Java build error +is plentyBut are you certain thats the actual limit? The shell also has its own limits +Has the code already moved to Apache Felix +Closing as invalid as per last comment +Click the button that says Selected false Click the radioButton again +I think its generic and powerful enough to deserve being part of core. +Can you reproduce it in unit test +Fixed under svn revision by properly calculating path when in a web container environment +I think the fair schedulers page strikes a decent balance between scheduler details and readability but it does this by listing less info about each job than +What I mean is that if you look at the file there are two elementsnamed Entity +This is the bootstrapper to get the new variable resolver being used instead of the default +The alias is persisted locally and restored when instances are listed from the server +What JDBC API call does DTP call to get this String? I think a XML string is ideal where you know the schema of it parse and display +So I was still getting compile errors in both and +It is expected that utility projects get zipped up as opposed to web ejb ear projects which may or may not get zipped up +Steps to ReproduceI uploaded as standalone testcaseIt works on Oracle Database +Hey Andre you may have to start Eclipse with clean after modifying the plugins +Hi Damitha Sorry not attaching two filesRegards Issue +Edward your code reviews and contributions have been very valuable I added you as a contributor and am assigning this JIRA to youDo you want to move forward with this patch in light of the progress you and Tim have made over onThanksMike +Trunk patch uploaded +I will +Im afraid I dont see the need for maintaining two independent ways to manage java routine security +The symptoms are exactly as before +But that wont create any problem for further pipelinesPlease suggest if you have better solution +resolved since this were done opened and for what remains +Im not fully sure about that the behavior should be changed to not disable at all by defaultI like what you suggest but Im not sure whether it wont lead to confused users +May I have that unit test? Would youplease also explain the bug in more detailThank youMichael +Not at all +Thanks AlekseyI put this into trunk and +Thats only true if the user provides an explicit binding +Thanks Alexey +ESB issue see not JBMANCON embedded console issue +I agree with the deprecationremoval cycle but i do no see any against listing both changes in incompatible changes +It doesnt make sense to have a user go uh Ill change this from the unknown default to +This patch is incomplete +This patch is to handle such situation +Committed to trunk and. +But you can exclude jcrsystem from the search query isnt it? In my application my queries start from my root node not from the repository root +Dont you mean to make the followingbinary BLOB BINARY LARGE OBJECT +none of the perform commands are my apologies for spamming the comments of this for me in master +Im not sure if well be able to get to this for security is a real concern even though we are pushing just to flesh out core features at the moment +Done +Is it planed for Axis or? That bug doesnt let me upgrade to something newer than Axis +Ill check in the fixes on a different issue +I have made a workaroundCould you please check if it works +the Solr file loading code has some features like trying to guess the size of the setmap up front for faster loading +This kind of worked +This is particular hard since we need to migrate a whole kafka cluster online +There is a note in the Platform Getting Started Guide explaining that this is information is superseded and most of the important information has been replicated in the Platform GSG +No feedback from submitter on whether he has adjusted sync interval +Patch looks good Ill commit it to the branch this eveningIm still waiting on confirmation that the substitution of the newer doesnt break our current version of Axis before merging into trunk +This was the one +okay now this is fixed andclassmate wont be an compile dependency in DR +wait for Velocity to stop using commons root problem is Velocity using and the end user wants to use xstream to output the body as XML and that xstream will output the class that was used to store the it as wont fix. +I have still the same text +I mean did you suspect I misinterpreted the original bug report? Or did I missed something elseNo you did not miss anything +Test for job success not just job completion +on having pre and post for everything +I applied and tested the patch to both and trunk looks good to me +its just that not being able to integration test with configured resources just doesnt cut it for most applications +The gain in checking which message was smallest was minimalA drawback was that memory usage increased quite a lot since we need to serialize the message compress and compare sizes instead of just serializing the message to the So instead i just compressed all messages with good resultsI attach both patches they add a configuration option like internodecompression controls whether traffic between nodes is compressed +Applied updates at r +Thanks Awdesh Sumit for the contributionYour changes are in trunk at rAshish Vijaywargiya +I usually wait for a while to collect couple of bug fixes depending on severity etc and so on to balance number of fix releases versus wait time +Marcin I have committed an improvement in rev Do you mind giving that a test in your system +Out of Date +Verified dir permissions on a secure cluster +Also ran the test and it passed +I also want realtime updates to an in memory term index +In some cases they even cause negative value when they frustrate our users Code is convoluted and hard to maintainThis wont just reduce frustration for our users but it will allow us to focus our efforts on other things instead of wasting tons of time on keeping some of the really bad drivers alive +Matthias this issue is already resolved in +Its small changes to logging so I will commit within a few days barring objection +This patch insures that error is displayed in case +Another solution would be to account for the shading in the configuration but I think this is unnecessary +But before get it applied I should need to test it to know if it also works with different alias at an embedded server +Version patch is attached +apply at root to test launch console Choose JMS Providers For Go through all the steps and deployshould deploy without error in the log +So leaving thisweird side issue for now and focusing on other issues +Thanks! Ok that should be yes its unfortunate that the had wrong artifact id worse due to way Maven repository sync works it is rather hard to eliminate broken files once they get deployed +This comments out from the suite as it hangs on slower boxes +Exactly but I need the json of the execution context +Rebased the patch and updated JIRA as Patch available +Moving based on Justins comment +Ive attached the model project set +Using command line switch is fineI think we can make this feature more versatile by naming the switch nocompressionexport +In other words I think including actual input that causes the bug is a bad idea +But we are definitely doing more tests on this +Please read what I said Take your first example replace tt with t and t respectively +Either make it nonportable to look up beans in extensions or mandate its not possibleAdvantage of nonportable is that it allows containers to innovate here +Changed from Feature Request to Task. +Do you have some sample configuration for this +NET Plugin +So in my opinion we can apply the patch and update the lib when the final version is released +attached mod to lucene catalog to prove that bug is in lucenes get +If the project is moved to a new system chances are it wont be using the same devices anyways and this doesnt create a dependency +Drools is using jxl +Sorry about this +Does this fix mean that we never wait for an existing query plan? Does anyone understand what this means about performance for queries that take extremely long to compile +We can just describe the API changes +Uploading a patch which moves the setting of the interrupted flag to just before CONNINTERRUPT exception is thrown as an back to the user application +is related to and fixed there +Thank you for the earlier comments Oystein +Assign to Mladen to fix +No comments. +Patch +windows booter seems to generate a +I still havent the faintest idea what might have caused the to be garbled nor what or who brought it back to normal +Can you retest with a newer version of CXF? is ancient and unsupported at this pointResolving for a few reasons is ancient +packaging causes linux to see these as hidden files when you copy them back to Windows the file attributes are tagged as hidden and hence the property editor cant save to the file +That will perform very poorlyLooks like what we need is something like a join +imo internal functionality should not depend on it +My second random guess is you could be seeing the same bug from a different code path +The tag element which is inlined in the body is collected twice as its bound to appear in the tags element as well +I think the user language preference should be associated to the context node +I did not ask for help because I usually will read through all the documentation I can for the libraries I am using and even resort to reading through the source code for those libraries before I ask for help +I may not have time to go into these soon so Id advise people not to use multi threaded mode for the time being +It is fairly minimalistic no helper functions are provided just a raw erlang interface +added commonlib to the list +Closing that. +This may have also been present in the previous release +If we dont make that can bump back down to minor +Thanks much better. +Minutes from meeting on November +Bikas does Sandys responses address your concerns? Id like to get this in so we can move to the next step which is getting this exposed in the client API +Keith added you to CC as this patch would require to new libraries from theJakarta Commons project to be added +Verified in GA +This was implied even so in my comment obviously +committed set jsx to +I forgot to set the formatter in my new workspaceAnd the patch has been verifiedThank you. +you can still capture the communication in the client +Verified by Richard. +batch transition resolvedfixed bugs to closedfixed +Whoops +The thing I like about having a component name associated with the limited private interface is it tells developers who depends on that interface +na already using and hopefully soon +Thanks NicolasYour patch is int runk revision +Heres a version of the patch with a bunch of the suggestions from the comments implemented +See p and several following pages +well possibly but was not released yet anyway. +Its for debugging mapbuilder events and should be kept in my opinion +Setting fix version to per my comment on the ML +Hopefully well have a fix soon unless someone else wants to drop in and help us out +Are you using Java +As for publicly accessible runs of these tests being done Im working on it and hope to have something set up soon so that these can be run nightly on Apache machines +The corrupt stack causes the random behaviour noted in this introducing a new stack variable in this fix I believe the corruption has shifted around and its not producing the same problem but the underlying corruption is still there. +Hum this wallpaper doesnt seem so visually nice +Will the fixes a couple of bugs in and +It has an old style and it consists solely of xhtml templates +I reworked the fix for my initial patch if the content of the unselected tabs are not rendered in server side mode there should also be no decode validation and update phase for this components +Pig can also delegate the padding work to the loaders +If that effort is to be abandoned then Avro RPC should be removed from Common too +This task will provide Eclipse based integration to trigger the generic Validation Framework +Single node yeah +Das Encoding der XML Datei auf setzen +could not reproduce with latest branch and trunk +Invitation made as done. +Just added Bruce and Patrick to CC +TTL would be an option on the create table so if you are using TTL you would likely know to not write old stale data as it will be removed +Do I need to provide any more info? It will be great if someone can provide an estimate if and when this will get done +Aidan excellent thanks for this improvementI have a Synapse based exchange sitting on a branch that I wanted to commit but waited until Roberts finished his work on making the exchanges pluggableI will clean up and commit this in the new yearI like having a separate pluggin module which can be downloaded separately than bundling with the main releaseIt would be even better if we can have separate bundles as release artifacts for each pluggin instead of one archive for all plugginsNot sure how to do it with maven but I think there is a way +So for this issue I see some kind of +By leaving it in the file albeit commented it should reduce confusion +The only thing that I see as a problem is that you have to change the container port from if you want to run multiple instances on the same machine +Checked in dev mode and deployed to Tomcat +if not then were going to need an instruction page for each version of derby we can run against +As well documented sessions need to be discarded after exception +Code looks good +We decided to migrate from to +I want to know what kind of class will cause this kind of error on BEAThanks +Already fixed for although we should consider throwing in such cases +There can only be one of these active at a time also +Javadoc warnings are present in trunkNo tests are needed for this documentation change +itemis final position is tmpPathsrcitemiwhat am i missing +The antlr exception should be converted to a anyway +Assigned to Mark to add to the documentation +It might have to do with your Checkstyle configuration so please include that to +Committed +No one seems to be able to get this logic right and well just do it properly in Mercury +If its like most scripts its looking for to decide what flags to use and it looks for that in the path not based on prefix +I am honestly not sure what to tell you itll require someone scouring the code and finding all pertinent uses. +There may be one out there +Thats the clarification I needed +We are out of time to release +Totally missed that class +mrepositoryUserstucu +Thanks for the patch Lars. +The log file shows that jclouds is hammering the remote machine with status update requests every seconds +attach a new patch to add missing line which releasing permit when submitting callback in similar what did +patch from Simon looks like a bug in the segment case +Would it be possible to backport to branch? Cheers +committed to and trunk +the full error message is +perhaps there was a bug in older versions that caused the session to be created without reason. +Ive been thinking +Cannot reproduce the bug in ER. +Github is once a day I think I see a recent commit does it look better now +We probably need a check round there to look for that function before calling it +However the current functionality is strongly related to and and provides four methods for each supported type +This would enable us to have a single servger impl for a SWT app and our external customer facing GWT based this still allow push communications from the server? If so +I can work on a patch if you feel like its not up to you to fix it or just dont have these methods are inconsistent perhaps this is a bug +Status of should be updated to Closed when this testcase is implemented +verified at r +Jake I am seeing this a lot with cassandra version and hector +Notes The trap opened a bit wider in with the removal of +test passes in +was supposed to be disabled in its current state +I believe the attached patch will resolve this problem but I havent had an opportunity to test on a Unix platform +Had to close the first pull request because to test I needed a version rebased on master which already contains some into the master branch +Another things we should improve about the Scorer APINot in this issue though rightI like the idea of having Scorer be able to tell why a doc was matched +I will work on this +You may put an instance of this class in the application scope as like in the might be nice to have also a svn revision number for SNAPSHOT versions +jobhistory from None to The change should be fine if this does not affect anythingbq + that preserves better deserialize for countfilter test in w cql fixes folded into patch +Reviewed +Someday it probably would be good to push the CXF changes into Neethi to simplify thingsId like to follow up with this on instead of here +For any or thrown if it is a declared exception simply rethrow the exception instead of wrapping it in an . +Done but the title is not corrected I only push Dan +Ive just sent the email to as discussedThanks again Andrey +patch for + +line +and derbyall tests are running fine withderbyall has some failures also visible in tinderbox +Attached files and implement the two stratgies I described above in +Thanks for the link! One question though what about Pull Requests that were done before this guide was created? I have one +i am using version and also the xsltc compiler compiles the style sheet so i didnt check it with other Saxon or other tools +Shanyu will you be able to kick off a full test run per Chris recommendation? If not I can help with this part so that we can get this change in +I actually think this issue is a duplicate of +My ignorance tells me it is for code styling +for the patch +QA Verified I was able to deploy a jar via the AS plugin. +Closing. +I played with that setting and found that increasing it had a good effect on the scalability +The times are averages of running the test many times +removed from from affected releases it is fixed in +In other words if the first if is true theres no need to do the second one +Wow very weird +Thanks Alexander for the patch +On the +this need not block +Heres a patch that implements the design as detailed above +Theother test is the same but checks error conditions inadditionKrista I did not include your changes to handle labelstylelablestyleclass and styleclassI dont think we want to get into the habit of morphing names ofattributes using the attribute tag having to then check for caseetcGranted we do have a couple of special conditions we do handlelike class and style I feel we should stay close to the intentof the attribute tag and write out the given name value pairon the tag it is within +Are there any plans to build this? +It would appear that FF and the curl bindings for the command line test runner dont deal so well with trying to hit up to a K docid +Shall we set the default value to GB +Next up bundle creation +Hmm + Should we maintain both symbols for a time in this plugin or is it better to just change the name and get people to fix their code if and when they upgrade their plugins +Number of seconds to allow between the Wrapper pinging the JVM and the response from the JVMShould I increase these values? +Seems like most operations in Admin are facing this issue +Brandon thank you for the review +The revision numbers are and +Sorry the bug title is not correct +Im pretty sure Im using. +Yes this bug is +But I agree it was not correct to invoke process internally without INVOKECHECK +Please review and let me know whether this is fine +on which build? you dont have a new build since saturday do you +I have added a new property on the Cluster Connection to the maximum number of retries and the node is now removed from the cluster after that condition was achieved +I only have basic knowledge on the Protocol side of things though +Tests added at cfc +fixed the review board suggestions added a testcase for the default priority queue behaviourAny objections? Otherwise I would like to commit and close this issue +we should not do that in our client branch +slip all minor and jbpm issues that have been slipping last few versions targeting John please correct if wrong +Michael I appreciate the contribution but works just fine in Castor it already has a special class and a handler for dealing with +It will not be enhancedfixed any further since we dropped JCR in. +Alexey please verify +Very small patch for this issue +Well Im mostly confused by the suggestion to map to true +Then add your libraries as per the older instructions +Instead you are supposed to use a new for each call or at least from time to time +A couple of questions +Let me upload a new patch for +It should be ok to not output anything from reducer with no input +Andrea moving back your issues to +Ive been working on this idea for the past week +Hey Grant bq +dependencies missing in the pom add them +The problem seems to be the JMS component that generates the id that is not file friendly +In our application we set up a context in on and tear it down in on +I will commit this patch after running tests +Should have this in too +Same window but with normal the LF of Forums in version no longer has this problem. +Hi Nikolathe second half of the started now +Our build system is a bit different than AMQ +Adding some missing files that I forgot to svn add +I guess I found the root of the problem If an extended vendor interface is not visible in your bundle well exclude that entire interface even if it derives from which were then missing unless it is being implemented separately as well +Works fine now +The key point is that the secret key must be stored in a secured file that is readable only by the java process ownerThat means that two different sonar launched with the same system account can access to the same database +I cant think of a way to unit test this fix since the problem seems to be triggered by things happening at the network layer +One master goes down it may lose some acks which should be gossiped to other masters in next gossip activity +User ID of ruppert would be great +I committedSending srcrefTransmitting file dataCommitted revision +One practical challenge is that of mapping names as JDBC is generally and Jackson expects fairly rigid name mappingBut other than that it seems like a simpleish thing to do so maybe I should just go ahead and create a github project for thisJackson and jDBI are indeed commonly used together and thanks to maybe even more so in future +These statements have been replaced with calls to logj methods and the messages are now displayed according to the the logj configuration. +attaching the same patch again with license grant +So IMHO I think its better to include here rather than inside Rampart as it concerns the security aspects specificallyThanks it be best as a separate build module +So can you provide another patch for BCB project files +Both live in common but are just used for HDFS tests which is why they are not package visible +Hmm maybe the latter step is not even needed as you said that Kryo only flips a bit and later flips it back but doesnt alter the buffer when done +I created so that the key deserializer bug can be worked separately +bat +After all we might be running speculative execution exactly because a task is slow due to networking issues +Probably we should just delete the check +Yeah +This is the logging that I added +Changed this to an enhancement request +as agreed on mailing list specific dependencies are you referring to? The Lucene modules seem pretty +Why is there a ticket for thisWe are not going to fix Checkstyle rules arent we +Submitted a patch for the old failure at +We need to go through those warnings for the generated project and see if they are real issues with the generated project or that our validation model is not knowledgeable enough about the provided components +Hi CraigI was using JSF before but I upgraded to before I updated my +Ashwin we will take a look after the dust of the Camel release has settled +The first query would fail as is mapjoin followed by tests with the parameter set to true +implemented aaime idea +related by +Sorry for the noise. +Attach an extra patch that fixes some more logging issues that I saw +tried again with revision here is some additional infoguvnor is deployed inside a ear the classes are located in the ear is deployed in jboss version this revision however my package does not build anymore due to argument type mismatchlooks like some other bug mean I still get the error with revision is also that I need +Version Released +doesnt necessarily need to be blocked on this if thats something youre waiting to work on +Probably good enoughCheerss sounds right Chris +If we think changes we need to make are not going to need incompatible changes I suggest making this Major priority instead of Blocker +Please verify +lgtm +Ill see if I can dig up the logs +Can you provide a working example to demonstrate the problem? The example in your original note did not work +so the port number inside cargo can be used to ping but not to defined which port the container should startI will send a more refined version of the patch tomorrow EduardoOK I see your point now +Simplied version of from reopen this bug report with either a modified testthat reproduces the problem and I can runor some TRACE logging that shows the cause of the problem. +Changed the component to Qpid Examples +style nit comment wraps beyond characters not sure comment itself is Rob +is specific use case is unlikely to be addressed however are expected for the POM in a future might also look at Don Browns itblast plugin that helps with this use case. +We should also plumb this through to the already implemented S RPC +Any comments? Would like to commit and do a +Dont be using valueOf in Groovy! Theres no retro +I finally got the compile working +but maybe i am too cautious +Im closing this issue as there is no real bug we can fix hereIve raised to ease tuning there is no way to solve this problem as the second sendSync will only be unlocked when the consumer sends back the done status which will not happen because the first sendSync is locked until the second sendSync is processed +So now Im into fixing the other thing where I want to actually cancel the task and also let that show up in the history as cancelled and NOT completed +fixes that arose with this change because an incorrect Woden programming model was used in these testcases +Bottom line Wed need to engineer a solution that does not issuetrigger restarts on as this makes things prohibitively slow +No response assuming ok. +SourceAll source code is within the zip fileMain source is under src folderExample webapp source is under example folder Why Global bridgeTo avoid registering the services with the session bridge for each and every new sessions +It is important to include all relevant info +Im behind a corporate firewall here getting updates from github to my local repo is pretty convoluted! So github wouldnt be able to pull anything directly from my repo +Feel free to reopen if it still fails for you. +I dont like the performance ramifications but Im on this because I dont see how this can work with chroot and viewfs +I suggest to use for parsing the string +znode lookup is done in backgroundfix some for your feedback i will update the patch according to your m sorry Ive been very busy the last few days I will review and update the patch asap +No NPE when runtime cannot be foundAdded error message when building project with unknown runtimeFixed in Drools svn revision +Merge with trunk and pass along exception as second argument of +AJAX term was changed to Ajax termAJAXJSF into for chapters +Committed to trunk +This should not block the release +The trunk now depends on RC +Stefan closing this issue as I have provided a possible solution +Im going to mark this issue as fixed +Does this means that MR on trunk will be broken untilNo NM exports HADOOPMAPREDHOME to all containers by default today as part of the environment +Patch that adds logging to help w future debugging makes parent check even stricter to ensure its actually the parent we are looking at and adds retry around get of table region if on first try we are returned a null +Closing all issues. +Sohilcould you close this case as Wont fix and open a new one with a correct name and description for it with the planned target release please +A fix has been committed for +Fixed under in d like to get this into +bq +If the AST transforms that implement the groovy behaviour are packaged separately to groovy and put on a projects classpath in it should work +Oleg +jar or copying source or whatever is up to usOthers hows my interpretationTo amplify on Grants replySean is correct +They need a bunch of work seemingly and need to be which hasnt happened up to this +Created PDF in revision add issue list to download page. +Patch version includes changes for review comments from John +clients are responsible to resolve links before using it That might be a reasonable contract but correct use of the API would then involve extra to check links before trying to open them no? But if it is expected that folks will regularly pass unresolved links then we shouldnt use an exception since callers of this method would be expected to immediately handle the case of unresolved links +It would be good to have a file format that only needs with the design goals of the new formatthanks pete +My patch for has been applied in commit rWould you please build it publish on Fuse repo and update +It is responsible for creating cache with specified parameters and options +Please check that the bug is fixed for you +The fact is that all Flex apps do run on a timeline +Assigning to Clebert as this looks like a hornetq issue +This change requires a dependency change on as is used now +No longer happens +I tried a few different configurations both java and and it all works +Well I know this is not a public I want is to use it at my own risk without only solution to do the use case mentioned in the ticket description with spring current implementation is to have another bean that deals directly with the cache to evict entry from it. +Invalid length x +Another factory is overkill here +Thank you Haohui! +Please ignore it and I am sorry +added progress monitors +Very simple patch attached along with comment? Can I commit +Build +When that issue is resolved we can bring the changes over to Struts Tiles +Should we make generation of Callback interfaces optional? in Transceiver can we implement the Callback API synchronously in terms of the existing API rather than throwing an exception? in Requestor can we implement the synchronous version in terms of the API so that less logic is replicated +Thanks Daryn +This is a duplicate of +The major change from the existing code is that I have taken off the part of code which slices and converts the search string to char array +I think Ill have to add a warning of some kind in runtime though as it could slow down the index a bit if the reader is way fragmented +Ill update the patch accordinglyAttached as m probably not enough into the CQL code but in the light of CQL becoming default in the future wouldnt it be more reasonable to add some internally that is being set to true by default and just set it to false if the cql flag is provided? That way we could just drop a few lines of code when making CQL the default instead of altering stuffAnyways Im happy to see this being available in cqlsh soon +Attachment has been added with description A test file throwing an exception when the Locale is has been added with description replace calls to with a big applied the change Mert proposed +Assigning to Sarah M +Marking this as a duplicate of +Thanks as always for the careful review and subsequent commit Bryan +Verified +It would also be great to somehow show the servers that are available but hosting no yet. +Version of that fixes a bug where the response header is larger than lines long +Parth +Anyone who wants to have another filter data structure cached can override this method +Database have also experienced this problem and discovered some other details that might be useful After getting a stack trace like the above once if you try to start again it works +In case of the is obtained from if it new API and from configuration if it is old style job APII am guessing you are using the new Job API +Thanks in any case +So it maybe that the replacements are right and that we have dependency of old behaviour +but why on earth is content assist in java influenced by our VPE +Reopening + due to user test with last snapshotreopen if you still have the issue +Jacques +By doing at AM start things recover automatically without restarts in case of transient issues with the log dir +At that time we had proposed the following Any directory can have an ACL An ACL specifies the list of users and groups that can access that directorys subtree +It is a very rare case when we have to remove all the messagesand restart all applications to let AMQ keep working but if it happens it is very serious outageIn some our testing after upgrading to AMQ we have experienced worse scenario yet when even restarting of AMQ and consumers did not help to keep processing +Removed check for Root annotation when reading an XML source. +Threads of the same process share the same open files of course +Made variables volatile + resolved issues for already released versions +Saphen I cannot apply your patch with CRLF issue +If a path contains Linux it works +It can of course be local as well +Thank you Benoy +Thanks for the patch! +Make sure to update all of the messages to something reasonable for convert local to field +Build and deploy jar file as ejb standalone ejb module +The same binaries built from trunk work on but not on Ubuntu +Thanks Paul +If you are experiencing this problem and have any hint about whatwhen it may happen please let use know as soon as a second +update some javadoc +Im not sure why new test cases would be required +Or something better around these lines +Michael the javadocs for do mention its experimental and There are still some open issues with this parser +We would much rather lose messages than have the potential for our producers to get stuck in any way +Thanks to Brandon Li. +I don t know where it comes from +If they are not matched or if one of the types are not assignable to the other X is thrownIn this particular example when debug flag is set for return type of both else and then clause are checked and found to be not equal +Some points are listed at the end of the document +The docs just need to refer to the these examples +takes care of the upgradation +The exception is triggered by xmllang elements in thesource xml document +But I ran the test files I updated +Hi Wallace +Correcting these test cases proved that wicket does what it is supposed to do +Please let the team know and well spot test to verify +Does this mean that the patch attached to this issue should be discarded +Do you have a specific reason why you preferred an externally defined over sessionTransacted there? And is there a specific reason why youre not using in combination with if the latter needs to be? alternate POC +Please feel free to it into +With this solution it would be easy to remove whatever default behavior comes with seam and put in yours or none at all +Works for me on the latest inattention will kill me soonSorry +Now when using the builder on every cache startup the database type is lazy determined. +Peter cool +I thought so +I just tested this doesnt apply to Solr thus I think this issue can be marked as resolved with Fix VersionSebastian can you also check if your scenario works with Solr +The two separate maps within Memtable are convenient for incrementally maintaining columns in timestamp order as well as doing efficient updates by column names +Do the versions of your NN and NN match +That would be great! +If everyone is fine we can move to r +Here is an excel sheet of the performance improvement +I reproduced the problem by putting an infinite loop in map method and killed Java map tasks when it is running +We should tweak Java Builder preferencesoptions to ignore all compiler warnings +Thank youBest regards by Richard +I will review the others shortly +Ive committed this to trunk and. +Suggestions forimprovements to the tests welcome +Since I havent checked what other work went into this issue before and the changes plugin just had some release I leave it here for review +ah ok thanks for the clarification Actually I thought about it but discarded the possibility without checking since I didnt think they could be that similar +This should be in TBD not V +Checked in the changes to that should be ER +Building the whole project results in above build failure +Oh Thanks michaelIll try to do it +Your explanation of the schemahandling seems reasonable to me + +Site not accessible +This isnt the right path but lets see if we can stumble upon onebq +We do it after diable the table +Sorry for the delay +sbitbig +the implementation doesnt have any at top level its a pretty big refactoring of what was thereill submit a patch soon after making some changes applies cleanly to trunk rename to gen cglib makes cglib optional +can u please post d solution of this problem? +We now store the locator in the lob so network server can retrieve it instead of doing a second store of the lob +line will initialize it for structunion declarations but Im guessing that codepath is skipped for service declarations +Am I missing something +I have also changed the code so that you can reference the same policy file twice but it will be only be added once +To fix this I need some way of reproducing the bug +Test case demonstrating the ve just submitted a patch to CVS +I think this is a great area to explore +Have to grab the revision and to debug it to see how it behaves +The attached patch seems to resolve the issue +The macro solution suggested did not work for me +This approach also works for for suggesting a workaround +Not that it matters for this patch just wondering +Sounds like we have an effective workaround +I know they dont make full use of EMF generics across the code base but they are a step in that direction and a precondition for later improvementsKind regardsBen +Fixed as the earlier comment submitted +Sorry for the delay Konstantin +Thanks for addressing this Knut Anders. +Fix and test it locally +Sowmya could you verify it +I think Bruno Roman Eric have been making good progress towards consensus on this patch +Thanks a lot for the review Cos +That change was primarily motivated so I could locate the methods in which we are getting gcmap problems in tomorrows test run +The patch has been appliedPlease verify +Priority is Major by default if you dont say Simone +This goes to patch fixes Tomcat tests +Not a problem behavior is as think what Bill is saying is that order by DOES NOT use reducer in his tests +Committed revision +Also this is the not the global lock but rather a lock around the object +I like you already! On a project of this scale avoiding cruft where possible is a GOOD thing. +Ive committed this +submitting patch +When GC would want to use rwlocks we can export them right now we cant after feature freeze extend the interface of TM +since the switch to Guice happened is there anything still left to do here or was it taken care of as suggested in the last comment +Line Column for reporting this problem! It should be fixed in CVS now. +I had been following this thread +My wiki username is dcheckowayIll work on some additional unit tests +So far so good with Thanks +propagation of null constants is still broken in polyglot due to an inherent design flaw in the constant folder +This issue is fixed and released as part of release +We now have a datepicker in and the old picker available in Wicket Stuff as complete with migration docs noting these in the Migrate guide on the wiki +There is Vinicius but we dont have yet enough time to work on this + +I took a look at your code and they look somewhat similar to mine +Thanks for all the support +Linked related udig you have one hack in the above list before this one can be closed? Do you want to just update the description of and close this one up? +My understanding is that the classic package are just interface implementations that wrap the new interfaces making it easier to user version with programs that were originally written to use +The proxy will automatically release connections +Removed +Have a look at its javadoc and the in +I am going to fix them sigh +Hrm +I have just yesterday informed the Mojarra team of the very same issue in the RI +I finally got around to checking this out +Thank you AdrianI removed the zh line under +If someone sets then they want that to be the default FS theyre just using a deprecated name to do so +It works in English but not in French +i already explained why I committed the patch without a test No test is included because it is very hard to write tests for race conditions and the fix is simple too +The question here really is whether JPA is mandating behavior that conflicts with the JTA spec +Committed revision +All JS snapshots were updated except for and +You have no rule that it should be exactly the Part from the server. +Found another issue when the management server tries to add a host updated the patch +Thanks +Application code can andusually does lowercase tokens by using an analyzer that includes a lowercasefilter +around the returned resultIt would also be good if clobblobtype could have its location set so that errors can be tracked down more easily +Fixed in +Maybe we could add to documentation warning users that best practice would be to execute each in auto commit or commit immediatelybefore going live with applicationFor me it is acceptable to escalate and go single user for following casescreate table in transactiondrop table in transactionalter table that changes seqences discussion of the issues raised by Mike having orphaned rows in a hidden catalog by design is a problemI agree that we need to have a way to the orphaned rows +I guess that its not clear to me that there is a problem here to solve +basically any reactorized build where profiles are injected from the +verified on +This is also a good addition to go to +Hmmm +Ive applied the code change I proposedsvn ci m Rearranging the test test per shouldnt make assumptions about nowSending srctestjavaorgapachecommonslangbuilderTransmitting file dataCommitted revision +This code has gotten pretty hairy but this patch goes through the list of jobs to find ones that need to be retired +Now the message is marked as a with JBTIS +Has there been any new work done on this +Unhappily I have no ADF around could you please checkThanks +Add tab named Preview +Since theres no SJSAS for development we use and since theres no enterprise we cant have the same re also waiting for an upgrade to to test whether the newest version works +I will commit it in some time +Patch rebased with latest trunk +kenney is this addressed by your build queue work? or is it not relevant anymore given the webwork rewrite of the ve never seen this in doubt this is still a problem since the code has changed so much +I hit this bug today +I can deploy stuff from the console and the DF seems to work from the offline deployer although Im gettting a separate openejb problem +Integrated to TRUNK +Ill see if I can come up with a better solution +Im doubting that this change is worth breaking backwards compatibility +This is a great place for this new test +eg http client will default and use http scheme if you dont specify that +You should not be having a problem with clients trying to connect to the server +Sorry Arne and thank you very much for you have a large system with Geoserver and GWC very powerfull tools! +Thanks +Maybe its relevant +Im probably using the priority flag in the wrong context +One small twist to consider is to maybe continue pipelining hosts but parallelize others +Thanks JacquesYour modification makes it even better +Fixing Affects and Fix in Version +re sounds goodre Why is a separate thread needed for stdout? Can you please elaborate on how this causes problemsre Id expect the io pipes to get EOF when the process is killed +Can someone please review and commit it +One of the first things it does is use this to upload the example config files +Test added in r +sorry for missing a few of those +Good point +I will commit this patch shortly +im in the process of completing this and will add a patch in the next few days +I just committed this +continue discussing at +Updated version of applied thanks. +Removed +We want to record the deletes between getReader calls however theres no way to know ahead of time if a term or query is going to hit many documents or not meaning we cant always save del docids because wed have too many ints in RAM +Test cases in working state added in revision to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It doesnt use site at all except to generate its own site but wouldnt be used in your build time +Thank you Lyudmill +The patch was created and submited +zip file were actually committed +Please verify and reopen if it fails +Closing resolved issues for already released versions +It looks like the testsuite build is failing so maybe its running older code? If its running the latest code Im at a loss as to why its failing since I cant replicate the problem +But we dont +q tests are ran with verifying object store which runs SQL and JDO and compares +TedThis looks excellent +Verified. +I will keep in investigating. +Thanks for the patch + +Patch a adds a new test to +Right +Ive attached the WSDLs that I used +Since Ive been in the habit of mucking about in the build system lately heres a patch +Id like to fix this would you mind contributing a patch against trunk? Thanks +Those symbols are in libqpidmessaging +There seems to be a lot more for httpunit which I presume is a predecessor +Scripts folder is now available as of release. +I described the issue in +we have code like that all over the place where we keep lists except for there its a list +patch file for the patch and committed changesThanksNeeraj +Looks like the constructor used by this no longer works correctly +Knut and I have looked at this and Knut has patch for much of this that will be ready soon +The real problem here is that after unwrapping the the result has null in detailedMassage field which causes the printout observed +for in and probably earlier versions +For this to work it should be integrated with our AS adapter then it would also be able to provide the runtime location so its not an extra burden for the userCouldnt you simply point to a server runtime and get the info from there +I think should not be responsible for pushing down filters +Wendy Edwards Can you please provide us the URL you were using or any additional details +I have followed the chown trick meanwhile and it works now. +Documentation for how to launch a job has been added to Chapter +I dont see a problem with it living in the matrix library +Indeed I used the above attached java code to see whether the generated input fails the same way + +Thanks! +The class wants to load the by the thread class loader +Obviously this is something to consider +Yaml file at seed is but your listenaddress is something else +i have absolutely no idea why i get this exception if i use column names instead for does not change at all +So we can only use +ill wait to see if jenkins takes before backporting +There is also an update to the test case to ensure the fix this change be considered for the releaseCheersPhil +We only support +Thanks for the review Hyunsik! Ive committed this +fix for DougThis fix diverges from the Java code +Just to mention one more thing theres a few Jira issues where theres code proposal that potentially but I am not this will work +Sorry for the inconvenience +I know it is weird but it was kind of intermittent last time that I tested +There are no testcase failures in the testReport attached +Both +I have tested my example with your snapshot of and the servlet configuration and it works fine no open file handles any moreThank you very very much for your passionate support +This example demonstrates the problem +OGC has the SLD WMS request to deal with this binding +So the readFiied method does not know how to read the data node name whether to use or +I opened +Why has this been relegated to an improvement and held toThis is a major showstopper bug for me unless I am understanding something incorrectly +The patch attachedThe patch also increases the timeout and adds volatile modifier to +Paul can you please provide an example or a patch detailing what you mean by a bit more clearly +If you andor others feel v is better please go ahead I would not it +Looking at those logs it appears that a thread tries to operate on an already closed channel +Should we have a version of stopping the namenode gracefully which automatically does the manual steps proposed in this change +that makes sense +Basically taking the following sections and merging them into a single page Building with Eclipse mvn as an external program doenst seem to work very well from eclipse +Added WSMQ quickstart as resourcequickstartscommonwsmqrouterin revision +What do you think +Im still not sure if I fixed the symptomor the the problem there +Of course we can add a manually +The plugin uses metrics from various sources plus some cost estimates to come up with its existing numbers +Im also concerned about the readability of the documentation in the +A trivial patch to reduce noise of Tomcat shutting down +ok +thanks for your help + +Closing resolved issues. +He cannot reproduce the issue with the data he has so youll have to provide your own data and sld for him to check where the problem is +JasonThanks for chipping in on someone or some people out there with Windows and a bit of time can take hold of this problem try out all the possibilities and report back with if appropriate a patch to commit +after the leader becomes active it will sleep for tickTime and then check to see if the followers are synced which uses tick +Heres the patch to make the index from s testbackwards and a simple test for trunk that fails +Can we close it here +I lost track of which source produced which outputs as I was debugging +There is a race during which the new data is there but not the checksum info but during recovery operations I believe we deal with this situation by ignoring checksum errors on the last if there are no replicas with a valid +Thanks updated +An additional patch which merges and together to create a +Any other situation in which invalidation caches might be needed that you can think of +can we name it so that its clear its running w the test conf so its less likely to confuse the unwary? may be too verbose +Verified and added test for disabled +in trunkThank you v much David for the patch. +And once we start using only features maybe itll provide a more meaningful error message +Fixed by Shane +Matthiew in your application start event was what I was thinking about and using the full path wont hurtIlpo this is strange as it indicates the inside the application start event didnt workI assume turning on internal debugging of lognet still doesnt provide anything useful +Testing in progress +FYI and in case someone else is I am looking at integrating and testing thiscurrent batch set of patches +But without durable sync all the blocks being written would have been lost anyway +Ryan added the signed and mdd jars to his personal repo +ok it works now I updated to latest nightly of browsersim +Correction I am proposing four levels for unauthorized scan SILENT WARN ERROR LOCKThat interesting and very different from what I was thinking +What is the current status of this issue? Allen did you have a chance looking into creating tests with a mocked +Al thanks for the correction +Z files so its hard to tell which precise compression technique they use +Committed to trunk with revision . +Thanks for the work +This patch alters the test to wait for the cache to stabilize before asserting on the number of evictions run +Ive just checked in an attempt to fix this issue +Log from error log attached +HiWe are using the very latest version of derby so it is clear that the issue is on latest version +People are so lazy nowadays +Sure for the Navigation page but I also changed the Index page thats not updated +The line number maps are currently generated but they are somewhat inaccurate in the presence of multiple streams +in SVN trunk +Can the fix for this issue be back ported case attached. +for reviewing the changes Kathey +This fixes the problem but slows down the cache +Thank you for the hintMy question is how can I make my mojo compatible with both and and why cant find my mojo which is in theIt looks like it uses the from maven whithout looking in myThank you very much in advance! you for the reply but i think this is not a satisfying have looked at the but i dont understand why the plugin is not using the of theIt is searching the of the maven installation instead +The patch and new documentation file to be applied in the xdocsguides directory +Ok +Committed at subversion revision +Japanese Character cant displayI attach the test program to Attchments +Forgot to point out one obvious thing feature to set is . +Thanks Chris nice catch! The first one is definitely a bug +Am looking at the recovery bits now so that we use the tm if we find one +And one more thing is variables configured in should not append to its own initial value +Test cases for have been checked in +I meant of course + +And mine was a misreport sorry a wine program was eating one of my cores and I didnt notice its testing at about s now so just about the same as the first patches +It must have been the helper file that gentoo packages with +Refactored issue to make it more general in focus +The method in question was not declared static and that will prevent Derby from finding it +Having problems logging into Phabricator today +It makes sense and it is needed +Is that the caseIn any case Id suggest you attach a sample set of bundles illustrating this problem +do you have lot of files changed between each buildAfter a cvs update continuum run a cvs log to get authors modification dates and commit messages +Okay will add this as an for reporting this +I believe this is a dup of +Updating the userguide to describe the new native threading implementation is a release blocking issue +pushed to this record since is now released. +Yes views and templates provided by a plugin need to specify the context path to use so the above is correct and we should document it as such +Now as same as result of embed driver dead lock will be happen in clobTest +Yes as this contrib package is called ICU +Also as this is the marketing video no talking head intro is not vital we are just trying to give the viewer a taste so they dive deeper into JBDS and the aerogear tutorials. +Patch for older version not to be committed +zip and +. +ok +You should then be able to compile this schema into a jar +this is not a bug it was always an explicit limitation of so the fix will not be included into our CP will be in our next major release for now id say it is EAP +Lets see +Removed any speculationinterpretation from for the thorough searching and the factual and direct comments in the Jira thread + for the patch +the change looks good +undoredo patch +Updated patch +the issue occurs when you run mvn releaseperform scmsvn +I too have a big issue with this for I only have access to the not the actual client code that is making the call but the service is taking to long to return the informationWithout the ability to change the timeout in the file there is nothing I can do +Class Logit implemented in revision . +I have updated the two example topics to remove the code that retrieves the default since this property was not set programmatically and its displayed value is therefore misleading +Sebb how did you do the initialization timing tests? Gilles do your single call tests unload and reload the class each time +is there much more to come after Joins +Hey Andrew +There is also streamlined which uses thoseI do not have a particular affection for class but I do find the new API horribly cumbersome and memory inefficient for simple cases and would like to have an option of a less cumbersome APIFeel free to submit a patch +Life would be easier if protobuf worked with old code but unlessuntil they fix that we all have to suffer +binwhirr has the execute permission set in svn +Definitely a tough problem to solve +Fixed +Patch is now obsolete since the mentioned image conversion has been implemented by Jeremias +bq What is the problem exactly with the facet filters? they are not escaped or are they escaped incorrectlyThe query names with leftbracket followed by stuffbecame names followed by stuffThe leftbracket got changed to and got escapes +I assume the heap size of the KV is used for memstorebuffer cache accounting +While you can add a boolean to skip the if distributed is turn offYes thats my second choice in my mindbq +It seems that some refactoring could be done there also to streamline code +Well if youre okay with wicket I am +All help is greatly appreciated +test passes after applying the fix for +Marking as resolved +proper sections in Persistence with Hibernate book and Hibernate Annotations seminar on Hibernate AnnotationsJPA has taken place +distcptmpdvml and user +Ah I forgot about that +Stack this patch is to address Jons comments in the reviewboardIll move UNKNOWNSERVERNAME up into and resubmit shortly +And a change that disruptive is not good for something which we have been trying to freeze for a little while now +There is no DTD for. +slip to next version + +The cleanup task will take the attempt to FAILED or KILLED +I dont know if the new way is remotely m compatible but they dont look like entities +Simples good +Im not sure its fixable by definition it corrupts the structure because you lose all posincs +ill let you know if something else happens +ToddThe problem was seen in not sure about the code line in the description is from trunk +I changed the JIRA summary to reflect the same +Our projects use the same setup and thus by importing them the rules are known +BULK EDIT These issues are open to be picked up +NopePlease vote for this issueThe only workaround I have found was eithera modify XML converter source to strip all javassist postfixesb use custom marshallers second option is just a waste of time +heh heh then coming in late to the discussion Im probably just creating confusion +Could we point discovery to Tools update site +Thanks +If you think this is not the correct behavior please elaborate +windows has symlinkshardlink +Enter B in the test form and you get HelloWorld for input +This is another case where we are not storing the information on the level which we will need to do first +ThomasCould you please then suggest any public API to include one portlet into another? +In that case a standard Sonar backup doesnt do the job + +Event type entity on which action was performed and UUID of the entity should be sufficient do not think entire payload should be in event details +patch without directory isnt failing for me +done news section style +Ideas of android application icon +I think I found it +If we havent already written out a title attribute and showSummary is true output the summary +please comment and assign re not developing Deltacloud any longer + +The fix is now in +I am highly motivated to get these changes into the code base because they permit me to access large TIFF files that were previously unreadable for my application due to memory use +Im using just fine with updating querying cached queries etcCatalin Sanda I need a runable test case. +Great point will do +I ran the program more than a dozen times without error with the fix after getting corruption times without the fix +Implementing an abstract method is not Override so I think the amount of Override is sufficient +This is one of those issues that cannot be reproduced because the conditions where this happens are very rare +For sometime I have been promoting ofbiz to a larger Companyit is National with headquarters in One City and Multiple manufacturingand Assembly Plants in multiple StatesThey use VPN to tie all togetherMy contact was impressed in ofbiz I showed him on the Demo site whilehe was at home +Every feedback or suggestions howto make james more useable or extensible is really welcomeHope the push works out CheersIm still working on this however Im not so sure how it will work together with camelAnd anyway +Smoke tested on Tomcat with Chrome FF and IE +r adds the setup in the parent pomTo use runmvn findbugscheckand thenmvn findbugsguiand select the generated target to review the output +Thank you +When merging releases you have trouble since there are loads of pom changes all over the place but it is hard to tell where something actually changed +Integration tests include a test that deploys two applications both to the same context under unique hostname is known by the AS initially but is deployed with the ease in testing of vhost stuff weve added the following to the real DNS resolving to +this may be related to the usage of the password arg in the SVN provider when password is not provided esp +Can you please commit this at the earlier +I have modified plan to include common jar file and tested it to be have not heard of anyone else having this problem +Jenkins is complaining about port issues +Nothing was released on it. +And the default template that they are using basically looks like the projects is a or mapping +Would restarting the console app work around this too +When I copy and paste into Pivot the image distortion happens +This is not a bug in Tapestry +This patch fixes the problem +When user change part of the application say a given servlet or pojo +Its just that the numbers at the beginning throw me offI dont think a definition list is a better choice +Could we instead just add a new interface which would be implemented by the existing class +To compare major compactions or similar processes between where we dont have the stripe compaction and where we have it +An alternative solution has been proposed +This is definitively not a bug +Just posted another version of the performance patch for +Eventually this can be for new containers as well and possibly all allocation runs +The revision number only took care of the +Excelent work looking forward to test the fix in the upcoming release! +do not need this in hadoop core +Can you please attach surefire logs? Im still missing a possible reason for the failures you see +The whole point of this issue is that it breaks back compatibility +I expect the environment section of our plans to be totally different in particular not including any dependencies just importexport package instructions +This should be fixed on the SVN and available in the next nighlty +Still working on this issue but was really the critical patch +entity java files and derby ddl script to use to recreate this of +is a typo should have been +Let the pool release the concern is over loading mapping files that may not parse Best to know now than the concern is over classloading classes into mapping files before theyre used inline Its a small price to pay to know that those classes exist and are referenced this stuff means not knowing whether or not anything is going to actually work at the time we return a valid object +Thanks Mark the patch was applied as. +What about TAB FF and BACKSPACE? Do we want to change parsing behavior or can we just open a new issue for to think about it again +Could you please remove the TODO comment +was typo in patch name +Does support chunked encoding for us to use +Also for now the code warns under all and fails under none +Hmm +Instead I am thinking about just doing some custom ant scripting to figure out the git revision + diff was created relative to the core folder +I agree with you on the three cases Mark and in a and b the latch will have already been removed +Patch applied many thanks +released! +It is deliberate to avoid select all to install these version restricting features +So far they look good +To Giovannis point I agree that there are a number of issues related to restartetc as well as its real usefulness +That way we can better track whats going on on failures +Thanks +I see the value in setting in a properties file and adding it to the classpath in and this patch does that +Have attached a classStruts needs to move its dependancy on to version before this can be included +If your respect what I do for other screens add and edit screens must be the same jsp with two execution method in the action +done + +Committed revision +With the commit of fsck now works with HA enabledI think I agree with Alejandro regarding the HDFS web and file systems +this then throws a because the current page is not of the same type as the cached pageanyway it works and obviously no one else is bumping up against it! so im happyp. +If you are OK then I can do that +Will commit in a day if tests pass and no one requests more time for review +Applied to trunk in r thanks TinIll close this but link it to for documentation on the entire set of tests +To hide the accessory bar the value for that preference should be true +Tell me if youll have any problem with Lex Im just reopening this one to finish of the outstanding questionsrequests from Nick in the above you provide the fontsize info the xcf file etc +Ok I see what you mean Randall! You basically want to execute certain code on all cache entries regardless of cache mode +I am currently trying to get everything working in with +Bug has been marked as a duplicate of this bug +quickstart showing that it works +Our plans to enhance the coprocessor framework slightly so that this can be done via coprocessor hooks +There is a big difference between someone voluntarily adding their name to an author tag in code and the id that SVN automatically adds to commit logsAlso the proposal here is about removing the author tags from the source codePresumably the author wants to be credited therefore if we remove the tag we should ensure that the contribution is credited elsewhere +Thanks for the help +It just alleviates the need to manage Boost versions when moving Qpid out of the development environmentWould that be a problem for CLR or MSDTC +Fixes the additional filesystem close and a bunch of other problems with the last one +the should take care of this. +Hi FelipeThanks for the patch +Fixing findbugs and javac more comment +GirishIam observing this issue and added few notes +The scope of this bug is limited to the as fixed in on +Closing the issue. +Provider side still needs work +notifications are properly publishing now. + +Closed upon release +Also thank you for the excellent bug analysis! I wish everybody would take the time to be as precise. +Hmm good point +I need to make sure the change doesnt break any of the other scripts +Those are all good points Kathey +on new BU constructor that takes a patch with a new constructor for +And +Bulk closing stale resolved issues +I asked someone else in the UK to run the same test and they get the same problem + is open for pending patch for reviewcommit +properties file for the update I pulled the contribution in today +UPS NM my mistake +Thanks Ivan for the patch update +Patch updates link for Format applied thank you. +Attaching already in the description but now as a real file attached +This allows some of the code to just check the namespace to know whether properties should be hidden from the JCR view +Thanks again! Juergen +I tested one of the tests that was previously failing on HDFS and it passes now with this fixed tests jar +The jpa itest example now uses this builderI added schema to the secret schemas in tck and the jars from that project cant be published until comes back up +I will follow your suggestion while I wait for the newThanks again +Closing all issues for release of the Documentation Guide +If you want to know more details you can email me eu at javatx dot org +Can you post output +I dont understand jacc about this. +please confirm during that issue verification +should not be referred +MRI would not see the data and would get stuck here too +Maybe you have some insight +Sam could you either investigate the underlying problem or improve the test so that it wont fail on hudson +But doesnt work after we Migrate to RAD and Webshpere +Yes I cannot find a single run that was not caused by +Please check that the bug is fixed now +All tests passed +Apache Flex do not come with AIR or files mixed in due to licensing restrictions +We have it applied to our internal branch but we dont have HBASE applied +You need to set the traffic labels +As I mentioned on the mailing list this appears to be working for me +We can certainly achieve a better usability with forge than with our individual maven archetypesThe only reason we havent implemented it yet is the lack of time +? Thats certainly a lot simpler and the only conflict is +that is in general for coprocessor +I agree that the ideal change might prove to be too large for and will require amount of time stabilizing it since it is fairly tricky +I think this should go into +This feature was added primarily for the partitioning use case where the size of the input data set is an input parameter from the context +State change log entries would be informational not warningsHaving a hook where outside events could monitortrigger state changes could be useful +Also lets keep the directory name as main +Not applicable to current codebase +Should be fixed by now Tony changed directory to be writable to directory group membersYou owe Tony a beer now +We shouldnt be referencing dependency versions that do not exist in maven central +This is working now as of the latest trunk +I updated the patches to revert the changes to themeroller and remove the extra directories +Users can change themI believe we cant copy RI colors and look however we must adhere to their feel +please check again + +So the test is correct as per behavior but not behavior +If you can show us more details we can it +Code has been committed in order to allow several test patternsDoc needs to be patterns may be specified using as a in version +created test for this bug +Alena would be able to provide this info +This will help avoid maintenance in future +Andrejus thanks foryour patch + +Fixed the path for the UPICONURL and DOWNICONURL of the Sub Table now works properly in both cases as a component itself and as a template for other is therefore verified and can be closed. +See +I still think getFields on multitermquery is a better option than a Fieldable interface +Yes this would be a blocker +I seem to recall having seen this before tooI ran the test in a loop for an hour didnt fail so that could just be a weird issue on the jenkins machines +Possibly make it synchronized on but use the on trunk +This issue will be addressed by the component +so we are not better qualified then anyone else to fix this +Thanks Johan! +This avoids essentially all of the object overheadMore to the point why not use a search based recommender and avoid all this heartache +thanks +in the future that function could live in a ddoc and be fed by a changesconflictstrue feed and run inside instead of the app +Also the test test in and should be updated to check that the error message contains the original error message +Fixed the issue as described +three fixes from xpg version of tr works with the parameters as got around the tr issue by placing this line at the top of the continuum filePATHusrxpgbinPATH +Thanks +When can we have it appliedNo idea +both changes at BZ and IT side are reviewed and signed note with detailed changes are sent to +This appears to be fixed but is defiantly still broken +Please +I fixed the original problem reported +Thanks for the good repro steps +I have no idea what happened so far +See the patch into and branch. +I had changed the database from oracleg to previous version +Publish the jars etc +Just needs to look cleaner +As it is shown on the following screenshotthumbnail +Fixed in +This change will result in a break in bugward compatibility if this problemexisted in the previous release of Velocity Tools +can just get a fresh copy and apply only your you for contributing +So creating this issue to keep record and effort to fix it. +I am closing this issue for the reasons stated above +Committed code as it was causing problems with Additional fixes are needed and under review but a new beta needs to be pushed to fix the issue and patches need to be reviewed more comprehensively +is interested in reusing as much of this as possible committed this +This patch should in no way shape or form be seen as useful at this point +That page is the second hit when searching for hadoop on major search engines +You also have some minor spacing things in your diff that make tabs be space instead of +Lets keep this issue opened until the next version of bnd is released and I will then align the dm plugin to itFor the bnd plugin please open a jira issue on it once the next bnd version is released +Cant reproduce in and trunk too this bug is really out of date +for fixing error reportingBut why it occurs is even more interesting since null shouldnt occur IMO +I think this should be an option +I will let you know if we have anyproblems with this fix +Would love a patch for this one! tracker sample modified to use GAE persistence with String type allowing definition of entity group hierarchies +There is no xwindows functions on the Zos machine +Verified on JBT Alphaa. +Is it possible for exceptions to be logged at the ERROR level instead of the DEBUG level? Id prefer not to have debug logging always enabled in production environments just to make sure every error is logged +arnthis bug is fixed and is now in the release +I did download the source distribution but Im not using it for this test +conceptually this works for me +Moving to next release for now but will try to add more continually +Option would probably be tidier with newer versions of java but would require larger changes to Woden to have a consistent API +Disableenable of a table is a flag in zk not state of region onlineoffline flag +I agree that should work as you suggest +just committed this +Line which is chars is the new JPA query +reopen for clarification What does first project mean? Why not just iterate until one is found? could the search please be +pushed in ebbf. +This has been fixed in svn repo is awfully old it was abandoned in early in favor of git. +Hmm +no our patch includes the initial patch from julian so only our patch has to be appliedbut the major part of our patch is the work initially contributed by julian updated and enhanced by us + Added comment about type conversion +I ran several tests one of which sent through messages and all were received without errorIf the issue is linked to having a JDBC master slave configuration in the broker than it would be highly unlikely that the problem is NMS related which is why Id like to see if its possible to recreate without the JDBC master slave config +okay i did a cancel patch +Patch is for branch will not work on trunk +triaged for Without a repro probably not much can be done +Just set up your own classloader and assert that its used somehow +Reviewd the code +Program will exitI can provide a patch andor open a new issue if this sounds reasonable +but again not a strong preference either kmscore and overview i agree there is overlap +In its current form this ends up getting interpreted as a valid initialization +I also thought of another possible implementation it would allow for varying latency between writes and visibility +Hi Stack can you try again +On second thoughts should we use before doing anything else? That might fix +Could you please indicate whether you still have intention to fix it +The whole attributetypeceefourannafivarlibdefault git p fdcdbfcdaefebcdfedaebdn ouattributeTypes cn socialperson ouschema birthDate generalizedentry generalized TRUEcreateTimestamp Z objectclass metaTopobjectclass metaobjectclass topentryUUID creatorsName adminsystementryCSN +Its already the case for the xml and atom typeI did not see your todo but Im planning to create a feature request to implement serialization using JAXB as its done in Jackson thats very handy. +Attached is the first patch +I looked at the javadoc for Combo and all I got out of it was that is not an approved value Since SWT has different implementations based on platform the javadoc states only those three values as possible values and I wasnt using one +Sebb Do you have a patch? If so send it over and Ill see about patching it +See +As you can see in the announcement for next major version will depend directly on Java +In +After reading more source code looks like the custom uri only supports library path on local file system +On my system it takes around seconds where around half of the time is spent inserting test data + +Or Elasticsearch for that purpose +uploading updated patch with changes from stack and patch with changes from stack and todd +Done +As far as I remember an update to the sld schema back in geotools module is +Currently it waits for the next block report to send the confirmation back +It may make more sense being maintained in MFC land rather than Solr landI dont think so since then would have a build dependency on Solr which would mean either various parts get built independently or each release of would be certified against a specific release of SolrIf this is not going to go into Solr where I believe it naturally belongs then I think we might as well just to keep it checked in under googlecode as part of the in Action book example +Will add a check to the underlying dom then will resubmit patch once it is for with better test another change +Unfortunately Axioms attachment support seems broken +How to get a readonly tab +I would rate for the +Thanks! +Merged onto the master and branches. +I took the content from and added it a new document +Steve not sure I follow what you mean about making use of get which assert are you suggesting this be used for +I thought the latest patch looked good and as there were no further comments I committed it to trunk with revision and removed the old test and masters etc +Metadata version in the release notes +Ive ignored a couple tests for the very short term while I try and replicate the first fails on my mac linux box or windows VM +Some of the worst testcases have already been adopted to it by me but I did not have so much time to change all the clustering tests +Thank you Patrick +Hi HenryI have tried this before and it caused strange object instantiation errors with Spring +Updated patch for current trunk r +Only problem would be that it would be tedious to do in a fashion and defeat the RAD principle of Roo +Beginning with version Spring Framework JAR files such as and no longer contain files with metadata +One request can we get the master into a state where it actually doesnt segfault before we commit this? Right now its impossible to run regressions or performance tests and on stuff that is this involved Id like to do bothCheer Leif +This is a classpath problem +Closing all resolvedfixed issues of already released versions of Roller. +Now I am sending it in six pieces +It was not resetting state as it visited each operator causing it to get confused on the key to set in the combiner in cases where there were multiple ops that could use the combiner +not sure how to resolve +Changes to mtom sample code to reproduce issue +Tenzo fixed this alignment issue +As I found Add JSF capabilities context menu action uses common so its tied to model and as Denis wrote should be better assigned to Slava +It turned out to be a coincidence that was unrelated to the use of the maven tasks +All tests are now passing for meJust noticed that srccouchreplicatorebin should be added to +patch applied. +Has been fixed with new type mapping +It should be resolved with +In the above stated example the property bar will be set to the string Test and no obvious problem arises +See attached document for the summary of invalid XML characters +Sorry for the confusionIs there a way to correct my commentssreladConfigurationreloadConfigurationsdoesnt fixdoes fix +Denis the link to the discussion on is broken +Just tried this out in Geoserver branch works just fine +One possible and to treat this could be taking the ctimemtime of the file create your new file and only overwrite if the mtime has not changed otherwise recreate +Will commit as trivial patch unless objection +Linking the forum reference +should we allow access to default to everyone irrespective of privileges? ThoughtsI think we should either allow access to default to the world or guard it with privileges +Odd + +Please ask questions on the mailing lists +They are for better or worse stuck with JBOSS this stage please povide a backport to SP +I see a flaw in my logic now +Can you apply the fix or do you want me to create a patch +completed +nope sorry +There is no way we can build the jsr functionality directly on top of the +And it worked for printf statement +This is a consequence of +Verified fixed in IR. +Great I think we are all on the same page then! Well move forward down this route +It turns out that it was caused by s within the error message in the exception thrown by our password callback handler +I tried this and it all worked +Appliedthanksdims +Ive added dsmiley to and +Code changes have been applied to work with new client classes +I must admit that when I tested the patch I could not get the majority of the tests running with or without the patch + case is updated due to the index size is smaller after applied the more aggressive change +Doron thats fine to open a new issue and close this one but it was this issues fix that introduced the bug +Are you looking for the generic way to make an outgoing request from Synapse pass through a proxy? The JIRA is about supporting the PATH only in a POST which sometimes IIS likes and looks for +It would be much better to link a hadoop introduction video from the apache conference +Its better with no stats and higher performance +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Verify that these are on or that there are some corresponding ones on the Wiki site +It would also make sense to allow testing the binary protocol but the current code has thrift hardcoded in quite a few places +Heres an image of the new New Rules File wizard without the extra details page and with the package name field moved to the first page +I committed this +issues reported by Bertrand above in Rev +I assume that the JVM has enough heap and that the issue only occurs after some time +Logged In YES useridThis is a known issue with the JSP compiler javac requiresall jars needed for compilation to be located on the localmachine +But it seems that the problem stands in correlation with the size of the xml that should transformed in addition to the size of the XSLTSorry +Could you please try again with the latest version from SVN? This issue shouldhave been fixed in the meanwhileThanks +the letter appear to be fixed by +Due to the recent spike in critical bugs this new feature is postponed to a later release +Looks good thanks Mahadev. +Hey thanks for the feedback +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Added comments for remoting service config +Tentative patch +Why was thatIn any case the correct way to set up our default permissions for a newproject seems to be right now create a group for the projects developers if it doesnt exist yet add developers to that group create a permission scheme by cloning Old Default Permissions +This seems to be a general problem with closure handling in the C backend as other uses will generate the same ICE +This patch accurately reflects the checkin +So we cannot enforce SSL or anything else for that matter onto the provider +workaround map the columns explicit with addScalar +Thanks Cheng +two versions of chooseTarget +My hope was that some would be possible within the description field but this seems not be the case +Waiting for Hudsons blessings +Committed revision I will file a separate issue for the assert failure in the right outer join and link it to this issue. +cleaned a with fix for avoiding reading from data source continuously +Thanks +does have access to and hence the methods in have to go throughI have run the tests and the patch didnt cause any failures +So I fixed it +Also to reuse the macros we can use the parse of VelocityOr we will just macros in each Velocity templateSo I am creating a subtask for migrating to Velocity can close this issue now because all steps required to use macros were done. +Simple patch based on logic in script +How about publishing these plans on the GRECLIPSE project web siteI guess other GRECLIPSE users are puzzling for a while what happened to the release +Also Im assuming youre using the connector +This allows someone to quickly see the raw data wo having to reconfigure the tableIt seems like the hasArg argument for the formatter Option constructor in the create table command should be trueCould use a functional test + +It is just what makes a difference because on x lazy resolution is enabled +I am on committing +The one exception is building the patched xerces and jars which check these projects out from tags in order to buildIf tests pass we may be ready for the next RC +This issue has been resolved for over a year with no further movement +Just as a comment for posterity +Revision Revision tests work with this thanks again for applying +A helper class to get the Seam version from the should get the version number from the manifest instead of hardcoding it in java idea Ive never liked how version numbers are hard coded like that +Have you tried running this against Spring and Commons? Do you see any difference with those latest versions? for the time being assuming that this works fine with Commons and Spring Framework +Thank you Hyunsik +move to doc version for less overlap with engineering tasks +The schema and validation tests do need updating +Sorry I couldnt try it until now I can confirm that it fixed my original problem! I can now map +Kevin to help out +patch applied with thanks +Updated patch to correct a small bug I forgot to change the accessors along with the type of patch was applied with modifications +Fixed the commentTrying to upload patch via email since the website appears to be down +Can you give me a pointer +I was doing some Issue cleaning and didnt know the status +This is not a but I may be able to get a fix in before Final +Here is the trivial fix +I would reduce it by pixels which miens also reducing the Stanbol logoWDYT +I dont think we should introduce yet another protocol for admin write operations +The forrest doc should be updated +Removed offending definition from with commit aafaefacfaaacaabeaebe +Yes but still Why? What is the advantage over not having them in their respective source packages +committed as r +This issue is fixed. +There was always a lot of talk in the past of deprecating removing the oldclassic lucene query parserI always objected to that in the past due to Solrs use of the parser but now that solr has its own it may make sense to revisit that +Does this need to be fixed inIf not Ill unset the +Thanks for the update +updated with much as possible we should avoid getting from the +I think thats the best news I have heard on this issue in a long time +The version of needs to be updated to once released +Trunk build has hung times in a row since this went in +fixed in CVS HEAD +Fills in the section on Documentation changes +See also related +Ive mistakenly selected wrong file +The work here is actually not HDFS specific +patch to add support for RSB for pointing this out +In my case join keys do fit in bits +Mule would be satisfactory +Thanks a lot for your help +Verified in in +So if after using the new snapshot uploaded last Thursday you see this issue again please make sure you upload both the client and server TRACE logs +The resources in come from the test projects in the plugin directory +I am interested in this project +hi allI have merged the patch from and then all workI have run the test with and all of them passedsome test have to be modified because in +However theres also a part of me that rather likes injecting the fault there so that you can actually tell whats going on by tracing the messageWe could go so far as to have some kind of flag element that indicates to Axiom that the document currently being parsed should be thrown away and the remainder of the content is a fault +Thanks for reporting. +Not sure I understand correctly but +I guess the class should be renamed but not sure what the new name should be +Sorry type no should say is now descending +But yea we should add AS as well good catch +getting the leo of a replica can be done as +Anyone happen to know what the maven equivalent of maven dist is +see comment before +Will change that in upcoming patch +I have opened discussion on the mailing list and on the mailing list today to try to find best practice to overcome the issues presented there and will defer addition of these tests until that discussion has resolved +I see a lot of commits going into but I suspect it is just a matter of coming up with a feasible patch +I do not sure what is the best behavior +Enjoy +Core library should not have any need to call reflection methods or access level changes so you should be safe there +Well the problem with the code had nothing to do with exceptions at all and has actually been fixed and is in the code base right now +Sorry bout not posting the quickstart before and not really helping you with this I just had LOTS of work to do +If you want to stick to cordova then stick with the cordova cli tools from +Attached patch fixes the issue +I created another task for that +Im not an MR committer but this patch looks good to me +is it ok to integrate this with Axiom DOOM and there are no other processors which has full functional HE has only core functionality +Thanks very much for pointing out these errors +Applied the patch which improves by allowing the creation of a context using the username and password for JAAS authentication and which changes to use this approach when authenticating with +Patch for the tests I created to show the bugs +done when it does the delete operationThe problem is likely that the SFTP server you use eg doesnt jail the user so using will escape and give access to the root of your file systemSo I fix this for the FTP operations to remove any leading for file names +This problem is developer studio specific +Not sure about ifwhen it will be private +If doesnt impose a big burden I think we should preserve the ability to run Hadoop from distinct build dirs +No problem at all +Hi I would like to provide some documentation about how this can be used +All files on the console proxy under usrlocalcloudsystemvmcloud files are having version string +I changed to but then didnt update the references +It would be great if you could give me a patch for the branch though so we can get it there as well +Adding initial patch for reviewTest cases will follow soonThanks Test CaseThanks patchThanks findbugs warningsThanksMayank +as originally written was rigged to fail out if anything went wrong during startup +Kathey pointed out this issue might be relate to +sounds like write should only add ENDOFSEGMENTMARKER if its not already at thats not +I would add a test case but I am just learning my way around the code now +Doing a simple POST PUT or GET gets complicated if you need to extends the executor for each type of request +Agree I dont need synchronization +There was a problem with JVM started again the build. +Can you just instantiate a singleton in your rule +Im intending to spin a new RC for the release that includes some of these documentation fixesThen entry should be listed under Release +Logged In YES useridIt also works when im redeploying the sar + +Currently moves the checksum down to fill the gap when a partial packet needs to be sent +I would like to use another issue since it is another release +the session state however is not affected +There should also be some way to view the native retrieval operation of the lucene index as matrix multiplication +I think so too because I tried to use childDelegation but it resulted in the same problem +Patch applied +Dominik something QA could exploreDo you have access to a Mac yet +Fixed the filename of the patch to apply to +As a matter of practice there are very few places where a exception logged at the INFO or DETAIL levels +Matt Bruce could one of you please review this? Thanks +It should be branchesm not sure whether I wanna be committing any extra logging related to this JIRA to any branch yet +I hope to get something like this in Spring along with a implementation of a with dojo code to maven changes proposed by Julian Dreiig and +Thanks for verifying the fix! +I need to stop having so many problems +I like the heartbeat solution especially because it means we dont need to fix anything in disagree with Randals gut +Closed +Are you sure you had version in the classpath for the tests? The issues you describe pretty much matches the behaviour you face when having an older version of Derby in the classpath +Doesnt seem to be occurred once and could have been environmental +Scan with filter may read large data using pread seems not goodIt would be better if thing is automatical however one way seems not able to fit all casesJust like other attributes cachingcacheBlocks +This would avoid having to translate loads of time strings at merge timeI was thinking the same thing +ThanksColm +Yawn +The assumption is that the admin would edit the configuration and then trigger reconfiguration using the servlet provided by that JIRA +Resubmitting Vinods dummy patch to force tests to run +yes Ive set aside some time early next week to get it into patch form +Closing just not to have it filters for jira verification. +Thanks Steven! +Been busy getting a new release of out the doorOverall great stuff +I will comment the test and mark it as deprecated +Thanks for that great advice AndreaI should really learn to pay more attention to these details +resolving as invalid per reporter comments. +Sure let me know when you post that patch and Ill review and commit it. +Triaged forAssigned low urgency checked new comer and repro attached. +I have attached the files and will email you the deployment I was using +Fixes a bug caused by threads sharing a Map when using a +binary distribution on windows and it works fine +The patch cannot be completed until consensus is reached on how to modifyupdate where the queries are stored +Hi Stepan I have encountered a problem I need to add a new security property but then how to remove it after testing in order not to influence the later ones? If it is hard to achieve I would like to use to test it in another java program +This actually opened up a bug in cloudstack related to response type of revoke ingressegree firewall rules +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Fixed the check in Rev +assigned fix. +Also added that caches the indicator so no string concatenation is needed +Please try with revision Alan does babed incorporate fixes that Andrew made in commit dbaacccced? Thats what Im going after at the moment +We ran the tests manually and they passed so please commit the patch +This looks good to me +It looks weird +I remove my veto +Now that work is well underway with AS all previous community releases are +Attachment has been added with description Remove use of enum as variable name +deploying tonight. +Neither patch will apply cleanly for some reason +latest version doesnt seem to need it version +Using generate a new patch from top level of trunk The path name has to be somewhat like srccontribhive +All the findbugs and javadoc warnings are existing problems +Just updated the Sijies patch to update the logId with +A fix needs to be made for this behavior tooThis code path in has been refactored and so the change may not be a simple backport +epf content and if we had what would you save it based on? not sure I follow you could you use More Actions AddEdit Mockup and create another copy with your suggested changes? yes these things should be extendable +What exactly is it about the installation that makes things work for the code that doesnt work when is the starting point for the upgrade +Hmm interesting idea I think the instrumentation magic could do something like that +I tested trunk and the mode seems to be having some problems indeed +This is fixed in CR Trevor you can close this issue I think. +at revision +Ivan for opening sub jiras +trying hudson again +Fixed in due to incorrect handling of configuration +Committed revision This commit is slightly different from the last patch +Test case to reproduce the issue +Thanks for the testcase I reworked it to avoid the dependency and have committed it +From our perspective its the same as Bug as we dont fix this bug +Maybe we just need to remove the log message +Should I provide a patch for with this +I owe you a beer so glad to see this finally addressed! Its been one of our biggest painpoints with for a long time +on inclusion of the in the wiki page +Thank you Stuart +Fixed in rev +If possible it would be great to include details about the class attempting to be loaded in the error message +I will verify that the test still works and post an addendum patch +trunk revision revision I will now reconfigure ASF Jenkins +Create a return for this order by clicking on quick ship entire order and then Create Return +If you have the time then feel free to commit directly to SVN +Retrying +Go for it +I agree that having the abstract class doesnt bring too much gain but IMHO would be nice at least to change to be abstract with get being also abstract have an explicit that will contain the two constructors with the System subject currently in patch that covers only have an explicit that will contain the two constructors with the System subject currently in +Chris thanks for verifying the patch works in cgroup CPU controller environment +Thanks for the mailing list remainder +Fix in resolved issues. +The SET statement would then need to send the package name and section number for the Statement +Version Released +Thanks the duplicate issue as its tracked +So WTP version same version that used in builds +Hmm +Thanks Bill I totally neglected to check the docs on getText assuming it was a straight getter thanks for setting me straight +This has been done in the main code but not in the test +Colin thanks for taking a look +Patch attached +But Im sure there must be a reason why it isnt apart from dictionaries being mutable +Uwe is now also working on this so I wont make further changes +apologies i was not aware of the separate source feel free to close the issue and thanks for the exceptionally fast reply +patch looks good +This issue has been fixed and released as part of release +I think the idea is great +please ignore my previous comment +Hi SaschaDid you try with CARGOThank on the trials done on my end this is workingIn the absence of feedback within one week this isssue will be closed resolvedThank for delay +If anyone else also likes to prepare another Service Interface please make a post to this Ticket to avoid duplicated work +This only takes care of English +Here we are changing only the host name and the port of the new URI value of the Location headerIf the new URI is again referring to a resource in the server to which the original request is sent then replace the hostname and port of the URI with the hostname and port of synapse +Patch from Yang has been applied +NET am sorry but I will close this issue also as duplicateDIGY +Thanks Ratha for the contribution +by newer which is now fixed + for the to trunk. +If this dialog appears during the test session the Suite may never end its execution +This would not be a problem in normal style forking where every thread originated from and System out is isolated by the processIn my case the runs in same process and therefore uses the same err which are currently captured by surefire provider +need to access GML attributes to make this work +It should be simple to replicate if is committed to trunk +generated after svn mv +Review is welcomeappreciated if anyone has opinions on this +I simply use with the attached scenario so both producers and consumer are indeed on the same box +the testresults run from inside my got it deployed to tomcat but apparently some classloading issues since the filter cannot see the hibernate jar +Integrate Zhengs comments +resolved in bbc +Patch committed in rev +If someone would like me to correct let me know +Ah +dont see any reason why wed want to deny the DBOs granting the execute permission on this procedure so to keeping it in SYSCSUTILI think youre right that only queries and triggers are stored in SYSSTATEMENTS +What you need for the override versus add behavior is to use the +I will open a new ticket to track the source reorganization +Thanks for the feedback David +Shouldnt JCR rather say that a is thrown in that case +It was seen in not. +one option for you may be to reduce the size of the data files there will be more of them but the chances of them having unconsumed messages will be reduced +As with disabling most roo features you can disable this json functionality by using the annotation and setting all the method names to the empty string roo then removes the aspect file and the flex json library can be many different ways can you do json stuff! I would like to see some examples of how you guys think the roo flex json stuff should be used and how it should be used in conjunction with the Jackson library or if indeed we should consider a choice between the two. +Updated the release notes and the reference as well. +Problem was with datahypersonic not removed from Hipersoniq database when application undeployed so when application deployed again more timers added to database causing multiple onTimer events +Had to do with the refactoring of AS under the covers with how the modules were structured +verify in ER +Any chance we can push it inThank you +A connected component is essentially anything that that entity can traverse to by following its valid mappings +Will be there new tag for EAPI guess was created mainly for JDK support on EAP +Or simply return +This is now fixed with a temporary solution of setting a variable with the target jar file before calling the Java VM with the jar file. +Thanks Philippe! +But I fear that it gives a wrong image to people that are not so familiar with LD concepts +anyway sorry about that. +verified on +I use similar to using their set and get methods +Attached patch makes proposed changes +explicit is better than implicit convention I was inclined to make it raise an instead of failing with NPE +serialization depends on FREEZE for each element +We allowed this so its like the envars where we didnt do this up front so now we have to live with +This will require and to notify the superclass when they detect a topology change +Moving to +I will commit to the svn soon +If you have a query that expands multiple entities in which query cache do you store it and based on which rules? If you store in both query caches for both entity types thats waste of space +You will need to add the CSS changes to the less files +Thanks to Piergiorgio all downloads are now taken care of so no Apache resource other than Maven is accessed when grabbing the dependent jars + Changed RPM packaging to build daemon tools scripts +Jiras Wiki markup doesnt work +Patched from trunkUpdated dependency to rely on version +didnt open the cache latch if it failed to inject all dependencies blocking all other threads +a debug could help debugin and define HAVECLUSTEREXDEBUG in was closing the wrong issue +Sorry for the noise with the code +CMSstaging all set +Closed upon release of Hadoop. +This is fixed +Anil how does relate to any Seam security api that was in the works? +lucene will open the previous index version automatically +Werner its fine if you just commit addition code. +Stack traces are from dbcp pool +Since is closed and since an agent was mentioned +for doc fixing up the getting started guide to rever to the actual file names in the demoworkingwithderby directory for src fixing up references to pdf manuals and removing links to no longer existing workingwithderby manual +Applied the patch from Norman and a new test for the scenario +I see that these files have both version ids and copyright notices +On top of that I added the handling of filename clashes between source and include dirs +Fixed by Sergey in rev +SCANPERIOD was set to negative value by mistake +Fixed in revision . +Patch adds which takes a and creates a which aggregates by the values returned from for documents +is now supported. +Hi I just did +Reassign if needed +I am working on a fix for and should have one for upload soon +Thanks Karam for verifying the issueI just committed this trunk and. +Depending on the implementation of the VM it may or may not be statically linked against hythrIt would also be possible to have the port library not statically link against hythr if we felt there was value in this arrangement +Indeed it looks like a straight backport of +It probably means no code change at JBM side to make servlet transport work +Thanks again +Assuming fixed in since there are no more reports of a similar problem +Yeah this is what I expect for WSDL +HiAny comments on the patch youd like me to address +I think I have done all I possibly can to help solve this issue +Bulk close of resolved wont fix cant reproduce etc +but once it is resolved we can consider this to be done. +sorry i was to problem +modules +BTW I suspect the older IDEA versions work because they did not support the SVN options and do not request the option +Resolved because stack closed in the belief that was cause of this issue +Much appreciated +StefanHave you had a chance to look at this issue +Test App that demonstrates the Mb for a test application? I am afraid to even look at it +Closing released issues. +thanks Ivan +The is implemented by both and +Use to make sure the result of the hashCode function would be always +The bug got solved from commit fix was available since commit abd +Here is the patch for +So I removed the configuration that was added in the original patch from those two poms and just put the ibm jdk configuration in one location the +tuned logj should add that the file to patch is not conf but srctest think the line numbers are really useful and would hate to turn them off by default +The other tests still fail on me +An Eclipse project that reproduces this bug under Tomcat +The single backslash should be used not a double backslash or an escaped newline character. +Do you know if mvn artifacts for hadoop going to be published by hbase? b I see you changed some of the tests with pig schemas with parenthesis to not have the parenthesis any more along with changing the pig dependency version +for Curtis becoming a committer on this project mostly because I too want to try WPF tricks with Boo +It was my mistake +Hey Edson see +I have not been able to reproduce it +heres a list of of example files I dont think we need to add the license blurb tooexampleexampletestexampletestexampletestexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestsexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifilesexampletestTestspsifiles +Third level collections are probably not loaded +Thanks Tim Im looking at the problem now +I forgot to say that if I force to rebuild the project one or two times then all works finethis happens on all of our projects you can imagine the chaos this problem can to scm cvs ext with native implementation all works for reporter and cause likely to be addressed by linked issue +All implemented Freddy +you convinced meBTW if hooks for are added to it shouldnt use this class just because a Locale is specified that would cause some unexpected results for people who have been specifying a Locale for date reasons +Rich let me know if you want me to do this +Thanks for the backport +will retry it +this only tends to affect graph drawing tools btw +Thanks +See for details +Useful for dev and admin other nodes +To fix the there will be parts +The tag hasnt been done yet +Remove static qualifiers +Done thanks +Are there any further objectives to this issue? Otherwise I will going to fix this next couple of days + applied please verify it +Limiting upstream will require at the very least a contended CAS which will add some overhead as thread pool handoff is needed no matter what it has no additional overhead +But test fails under because uses which doesnt exist in +Thank is listed as a Seam bug because a user reported it and because I want to see if there is anything I can do to improve this situation +Now removed all explicit dependencies on any of the control classes and defaulting to the Java classes +Any chance of including a test for this patch +Triaged for +Committed to trunk and +You may even manage more than one connection with one single connectorWhat you ant to close is the connection not the connector +Target tests failing on server instances action with docs needed changes only in testsuite. +Added foreign key tags to the orm for mylib and also to other packages in which I had neglected to use the attribute or element where needed +Sure I will do that to make sure existing function is not broken and there is no substantial performance dropAnother thing Id like to address in this jira is that bulk assigning currently doesnt pass the offlined ZK node version to region server as regular assignment does +We dont understand exactly why this didnt work +when I run ant +The proposed changes seem ok +And the principle is about data hiding and the likedislike of using methods rather than data members in generalActually this problem was raised by a customer of mine after I mentioned he could use Axis C +I havent looked into the relevant Spring code in detail but am just reporting empiric observations +Closing because this one has been merged. +depends on the class introduced by the patch attached to +DennisNeither of those issues is related as far as I can tell +string is a semantic label in our case distinct from binary in that it is assumed to be characters but doesnt define encoding + is working fine as per the latest changes made to the feature hence closing the defect. +Fixed the defectI revised your patch which can pass our checkstyleWhen you attach the testcase please ensure you also run mvn clean installThanks the test cases based on the patch reviesed by Freeman +The main reason is that they are the only way to force people to think if the method should be not final by intention or if they simply forgot to declare it final +BTW please dont mark a JIRA resolved until the patch is committed +I will add missing pieces to sandbox with the help of LahiruThank youRegards applied +Which platform are you working onMods Looking at the amount of queries in this thread alone i hope this feature gets included in future releases. +One did not seem to perform any actions the other actually did perform validation +Ive also uncommented the equivalent tests for sets and could also do with array and map tests +This work is not intended to move the ESB codebase over to rather it is to integrate jUDDI into the will make specific changes at a later time +Attached an updated patch that includes a test +I still saw errors described in previous my comment andHowever Im not sure that is as same as your cases +Updated my patch against the latest from the previous +Will be fixed when the site for becomes online. +Sounds like this also works as a solution for. +Also simon i just wanted to say you need to be root to change the mount option etcI think this is totally appropriate for us to do again quoting from the pageThis flag is intended for use by indexing or backup programs where its use can significantly reduce the amount of disk activity +Fixing issue description to reflect reality as reported by +only in function has been resolved so the issue is fixed +I aim to get active on it again so any interest would be very welcome +Depending on your use case you have different options. +stated in libperl ok +I think it says to execute the script whatever this means +See we can now update the examples and have one that works on both versions of EAP? +Probably but I think the true problem here is that gossip is independent of any other operation +The only secure way out of this is the solution I already posted on the mailing on Jochens latest comment should this issue be closed or is there something else to be discussed +Its much easier than you just have to wrap the in a delegate that doesnt expose teh interface +Beyond that please discuss things like this in the forums as chances are a lot bigger you find persons that have the time to help +This was filed originally against EAP +I only tested on HSQL I think this is ANSI so widely next thing youll want will be separate constraints in different Oracle tablespaces +With the added information about the user and password attributes perhaps the sentence This attribute can be combined with other attributes +bq +Note that content size can change after AJAX request so it should be in the +Im obviously seeing different garbage collection the result I get is should be irrelvent once unregisterConnection tidies up the map +So there is no registered within the scope and thus actuality the dependency mechanism shouldnt rely on the existence of the container itself but only on the meta data +Moving to CR for additional discussion +has been marked as a duplicate of this bug +bq +Once Ive done that Ill create another JIRA to include methods that can dump thread dumps in tool specific formats and subtasks for each type of tool. +thrTPSwithout any pathces +That looks more consistent with other configs +Attchment fixes issue +We can do the initial code drop to the repo if it is not mandated otherwise +from me too for the patch +My change doesnt stop regular expressions with the and characters from working they still work fine but if this change is implemented then for server side the mask validations will also work if you dont put them in +is the same as but generated with checked in +If can then still be reopened if necessary +Looks good to me +Changed compeonet from DAO to SQL to add that affects iBATIS and this patch is for the latest SVN version of the +I think you should benchmark the cost +Children ask for the parent component check their type decide if they are Fieldset or Form and then ask their state +but if we eant to support it as a connection pool we would need to +For my serviceUI desktop application this allows me to say that many Entry types and core Jini classes are never preferred so that they are always resolved locally and will not cause code downloads to start +PATCH method patch +Yeah good question +Right +Ill take a look at the test case you had which reproduced the problem and revert some of the optimisations weve applied over the years to to reduce contention and which have probably introduced this vulnerability +Yes it turns out the ant build file did not have a clean task dependency +Im sorry +Moving to since that will be our first release after Spring +Kurt +So I added a null in the code committed +will require investigation +However when making it bidirectional it uses load field orders for each customer +This change broke the build +I committed this +see branch +Running a couple of jobs just to make sure its fine +Additional comments on this I dont think the thought is fully baked yetThe ideal output of a build would be a single features file that aggregates the features generated by child builds +You will need to make the following changes in your new patch +Martin any update +Keep an eye out for the is paths +This way we make sure that jobs in waiting queue are served in the order of their priorityI already have a patch for this working on +Could you please upload it as an attachment + +The groovy nature id needs to be first in the list of nature ids in the +Committed to trunk +Is there really a reason the key needs to be an object? Why not just a string like scaleInfoTwo different scale functions in a single function query would then clash +Hard to tell if its since it doesnt have full stack trace +Gunnar feel free to assign to Oleg if that makes more sense +Thanks Vinod +agreed although it is not common to have whitespaces in the directory names in linux +Committed patch as svn +I wrote more extensive testing +This timeout value increases over time +SorryIll check it out this weekend +Thank you to Vinay Haohui and Nicholas for the code reviews and some great suggestions +Rather funny when I change width of browser window group names are not shortened thenWhich browser was used on your screenshot +The regression was introduced by usage of to cleanup the HTML fragments extracted from the Javadoc starting with version of the +Most of the users may not change the default size per reducer +For we could start nuking this relatively easily as its only used in places. +Hi Jing the patch is +As far as I understand the bug was never fixed hence the issue should not be closed +OK thanks Shai! Ill stress test that and Ill also add the seek in +Just replace an empty edits file and startup the +to branch +If there is a simple workaround it should not be a blocker +Tested with the latest build +reproduced with the attached +Status? I think that rollbacked some changes we have donePostponed +Yes I will go ahead and resolve this +Committed to branch and trunk +Committed to trunk +patch to fix OSR trigger of believe that the patch fixes the issue +Probably is missing mclassifier bit +The above works and I believe it should only affect tests +I just committed this to trunk and +Also I only see one java process running in my task list at that Tomcat server entries in JBN EM for those two servers do reflect the status properly showing node up and node down +Existing configuration can remain as it is for deploying without +Need to investigate if I can do something about it after appliance was created or consider a patch forScheduling for the future +Now we can resolve such conflicts by pointing the different files for the different bundles +But this will have to go to +Thanks for the patch Ed +You have to copy some more values after importing the page +inactive is a correct in my opinion its not a correct name +Thats actually a bug in Camel due toString access units and units is null +Any updateThe mextensionscache is still being written to with an blank cache during the remove operation +Now running the tests to see if any more break on my machine than trunk does +Could this be caused by the difference of environment +I am in Australia so there may be some time lag between queriesresponses +worked for me in Ive emailed you the file +I have a question regarding the standalone which might not be a real world scenario but fixing this issue I need to know what the behaviour should be after the fix +The remaining rules have been tested +Patch for orange color included in patch for +Fixed +updated to reflect the change to +As mentioned one solution is to diable for now if you run into this +The text The motivation for is that the specification does allow a +I implemented this in the performance patches i implemented some time ago but have not committed yet +waiting +Added to the list of Affected Versions and scheduled for. +Sorry for the confusion Okay Im removing the references to it now. +Bulk close for +Fixed in +all files in entitylinker packageI left the example class there in case you want to debug with it +bq +The intention of capping is to not let the user kill himself by having too many reducers in case of large mapinput jobs +Done thanks for spotting this +Would you also be willing to contribute a test case and put all changes together into a patch in udiff formatOleg +Retrying +bq +Patch changes only documentation +This sequencer has been written and is located in SVN structure under trunksequencers +Were going to have to figure out an alternative way to fix this problem + +Kindly review +Synchronizations are able to use transactional resources +If the client was just using the wrong value for these use cases and the server ignored those values then the old values were broken and should just be I think I am a little confused on what you are saying here when you say client are you talking about the gadget or the containerWith the current patch if my gadget used the old shindig defined constants then my gadgets functionality could break +A couple ideas +Had there not been any workarounds available I would have been more inclined to apply to as well +We are absolutely positively sure that this is the cause of the problem + + John patch and fix this issue. + +Youre starting using start script in Scripts right +See if gets fixed are we good to go +Its properly interpreting Solr errors as needing retries +I have been using this feature for almost a year +If the database is completely empty and has no tables why not just recreate the whole thing from scratchYou would only need a little flag somewhere for this +Marios thanks for all your work and investigations If you want to assign the JIRA back to me I can track whether this is something that doesnt go away shows up on other platforms than FIf and arent met we can close it out +Latest version of the sample schemas added to reflect changes in the. +Yes you can make svn now +I like this approach! havent run the full suite of tests but I want to get the patch up for comments this evening +The original resolution may not be wont fix but as the bulk change required it that is what it is. +Closing resolved issues for already released versions +JBOSSWSHOME think the system has found the AS +I pretty sure I have a test case for it +Touches the following filesM javaengineorgapachederbyimplsqlcompileReject table function creation if we have onlyM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsVerify correct upgrade behaviorCommitted at subversion revision +Hi Michael what information do you need? thats the same issue. +HiI have a patch for this but have a question +Hi MaciejIm aware of your pull request I just hadnt the time to integrate itBut dont worry Im planning to do it next week and it will definitely be in the release that is coming soonCheers +Patch with the and all core and contrib unit tests just committed this +Thanks Edward +No response assuming ok. +Can you provide a ZIPTARwhatever archive with a demo project +Actually just found that the recent hadoops combine support not splittable files as input +and we will migrate to sooni think this issue can be closed. +Are you using an old compilerAnd the only change to None DTC that some code was extracted from Dispatch into a new method +Proposed patch +Yes I think because I dont think it is a good practice to have repositories storing releases and snapshots at the same time +Done +Noel +And throw in some blank lines where natural too maybe +Let me see if I can reproduce the problem with a smaller stylesheet +Hi a test I could check that tableState property is called on each Ajax really a good I believe there is a regression When you change the state column width for example setState doesnt reflect the changes anymore! NicolasIm not sure if I understand tableState attribute is bound with backing bean property get calls getter for it using +So I think it would be helpful to just include a blurb to the effect of the above in the section in the hibernate reference on HQL implicit joins +Commited initial prototype pending test with composite components +bulk move of all unresolved issues from to +Is there any update on this +The produced relative URL is different than but it resolves to the same absolute URLSee the modified attachement +Attaching the think the existing behavior is correct +Not necessary +Normally the file is rolled by renaming it and then compressing it if requested +Thanks a lot +This has all been applied to trunk +The valid region should always be the value from the sequence plus or minus the increment never a region between two values from a sequence +Patch to silentPrint method to use the passed object rather than creating a new default instance +I compiled patch against and passed. +Note that Im not familiar with eclipse UI development so it might take me some time +flushclear to before page read to allow for of properties of page result items +agree this is clearly a hack that should be addressed +This isnt really necessary you could just parse the pom in the first place +Can we default to yes or no if it is not provided +Thank you +Ok not asking the user to provide version in both the tools +Thanks Martijn! +Perhaps we can create a new wizard for this +Attached patch adds a test and changes to only prepare for a request entity if it is +my thoughts was that these would be good tests in the branch + +This is indeed not an issue +Looks like youre working on the other items so assigning to you for now +Looks like ANY +The test demonstrationg the deadlock stacktrace before deadlock indicating an index with as testcase with only one column +There are a couple of technical problems with the zip +so no changes close Solr issues +I currently dont have good idea +just fails because the JAR was never built +My assumption is that when you call close you already know that youve hit an exception and just want to close the stream wo getting more exceptions +Fixed in revision . +When I wrote the inside Tomcat configuration the authentication process works solution provided +changes verified by reporter +any body able to find the solution for this even I am looking for the solution but not able to get the is a duplicate of another jira we will be look to add sugar to allow this syntax but itll be internally converted to an eval. +Not yet the has to be replaced with a +The shutdown took around on my box so I dont think its worth spending time optimizing it right now +Theres almost no chance this is a bug in xmlsec +Thanks +I think I will also add a line to set the threadLocal to null at the end of each request to be safe +Attached files are the result of updating abstract datastore tutorialThe downside is I have one test failure on transaction independence run the file in the test directory for an easy to step through example +Thanks Tim! +Must have been fixed at some point +Thanks for the bug report +Should this item be Resolved with the change made to throw the exception +Fixed the test failure due to class sizing +Since this is supposed to return only one value per pig query the singleton is a better placebq +The issue can almost be closed now that we dont use anymore Attributes but +I guess this could be fixed either in the AS adapter or make it failwarn in Arquillian which could detect this kind of error +Since the patch went in we seem to be holding steady +Our integration tests cover this feature as we able to parse same set of files. +Hi JustinI agree with you that using XA will solve this problem + This bug has been marked as a duplicate of ! Wrong duplicate number Needs to be +Unfortunately we cannot yet upgrade to because still uses and there are significant changes between the clientsBefore I put a lot of work into figuring out a workaround for can you please try to show whether worked properly with Solr as far as the file name is concerned? If it did not I think we can postpone this ticket +now uses canonical time zone suffix Z +We are still getting unexpected results +Shane I tried the fix as recommended by you and it seems to work in firefox as always +cleanup the log files before and after the testing we should not mess up the logging dirit is dirty +Replaced content of jBPM Message and Scheduler Service section withThe supported jBPMESB integration in the Enterprise SOA Platform has the following limitations JMSJCA inflow is the only supported configuration of jBPMESB integration for server configurations with ESB only supports the Messaging JMS The jBPM is supported only for configurations that do not have the ESB deployed +I would like to use as many Ivy as possible so it could do the work for me but I just dont know everything it can do to help with this like generating a matching revision based on latest strategy or file pattern etc +Agree on the change +Im still investigating to why the memory leak occurs though +Committed to +Now there is only one attempt to connect to the results database and create a schema +Earwin ill try to help with the Solr stuff after i review this monster +I tend to disagree here +pushed +Fixed in r +When deployed the portlet displays a form +sys properties on any systems i dont have personal access toassuming all goes well ill refactor the SSL randomization up to and merge back to x +if theres anything I can tell you to help you ahead just say so +Roller administrators now have a single option to choose if they want their comments to allow html or not and if not then comments are html escaped +It doesnt really affect any version this is just a task that needs to be completed before the next version +In short it will result in the tasks of the job being scheduled on with sufficient free memory and tasks which exceed the memory limits will be killed +Thanks for your contribution Greg! +So using lots of memory in the opt compiler pushes up the time spent in GC threads but has limited effect on measured compilation or measuring compilation phases +Both the patches are not required +Thanks again Ben! +slip all minor and I assume this is for you +As we have multiple parsers we should separate their syntaxes +Verified by Natalya +An option if you dont want to have one agent per machine will be to use Zookeeper to coordinate everything but it will be more complicated to deploy not s the item here? Just documentation or is there code to be written + +Cache the modification time in the +Added flags for AMclient side split gen +Bulk move issues to and +I realize now that my original search was on which is why I didnt see it beforeIve removed it +The patch looks like the previous code +With the append design we on purpose define the get semantics to be there is no guarantee on the length of an file under construction +Sun Java and for other more apps like maven ant jboss AS +Test is not failing because is not run +One configuration verifies the production configuration +Hi Roman append and create are different in the sense that there is existing data in before append + attached +The patch has been appliedPlease verify +Are you refering to the task count numbers below the x axis +Thanks for the patch. +Does the mailing list exist yet Matthias +Merged onto master +Should it be that the callback is marked conversational separate from the service? Also is it up to the service to mark operations on the callback with an intent +introduced with +The patch is ready for reviewMatei Would you mind help me review this one +For those still waiting on this I recommend adopting the approach separate modulesprojects +In general it isnt so important what exactly causes concurrent run of flush method I just dont understand why is used in if it isnt made this class! One more example concurrent run of write and mergeAll may lead to +There are many occurrences but they seems not relevant to the issueCan you checkThanksHoward +Any idea when the ml creation can happenWe are eager to start with codingThanks! +Heres a version of that I think should be compatibleThough one question as I understand the old code the order of the replies in the ack packet was actually in order +Cleaning up +Thanks Oleg +Sorry about it +The changes look fine and the tests passed so I am going to check this into trunk and +Though theres a caveat and a note on purge +We switched to is not None check +This was already part of the initial SVN import from the branch from +spelling fixes for xdocs dir + Username sshparker +How do you think the inconsistency can be resolved when using the JCR API as well +In the meanwhile maybe someone could review the patchKnut youre right +Incorporated most of Teds comments and all of Lars commentsI did not change the order of advancing the rwcc first before releasing the rowlock in the because this will occur only in some error case and in that case it might be better to do things in the normal order +is trying to apply this to hdfs trunk which does not have this code after +Thanks Laura +If were rolling our own Memory why do we need subclassing it +updating to trunkBut if Uwe thinks this is a bad idea +Committed to JBESBGACP with revision . +This is fine with me but the test is broken then +Can it be because the plugins got downloaded even we have them on the hard drive? +That way the scan can throw an Exception with an invalid state and kill the scan immediately rather than scan the entire Range with no hope of returning results +Patch to automatically set the young generation to the heap which yields a increase on both reads and writes in my testing +Minor fix to the patch fixing the spelling mistake +Just remove the workaround second pause before committing +With persistent messages and as the only messaging provider in AS dont we always have to use java? With the old Messaging if you placed the schema in the same database with your application schema you would always have a phase commit for persistent messages +I plan on looking at the merge to and will look at merge to if that is not done by a rick +This information is also stored in the summary file +I dont mind adding all that synchronization but first Id like to understand why it is needed +FYI internally Railo is caching regex patterns in a soft cache +classes are present in would say thats fine +You should grab the scripts from the final release or development head +This requires creating a new consumer and producer +Or suppose a job is restarted but all records processed after the restart are skipped for some have some doubts about introducing new execution context value vs just restoring the existing from the execution context upon restart As far as I can tell and are the only writers that currently keep track of the item count in the execution context In the execution context variable refers to the total number of records read not just the current execution The number of writes per job execution is already exposed to the user in a much cleaner way via the step execution detailsThoughts +But if is not supported then we will have to count the reference by outselves +Find out name of new console +from me Im fine with it +Yongqing Thanks +bq +This will greatly improve fsck performanceYes is a must +Some additional information that may help here +To be reopened if further details appear. +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +In this case since the persistence context was cleared the entity to merge to is not a removed entity yet the toAttach is marked as deleteSo I think the about delete test case is still expected +Fixed in r +A minor point in m an email attribute is required for the author tag I think this should be relaxed and be made optional +You could zoom this attached file if you load in new browser tab +Reschedule if completed sooner. +Thanks for outlining this issue +Heres a patch which addresses the issue +Mainstream text retrieval research applied tech is very strongly biased towards bodies of text +This has been done in the projectsee +Can you make that change and submit a final patch? The additional change you made to pass the name in looks great +looks good +BrunoI am going to assign this JIRA to myself as I am almost ready to post a patch +Subsumed by +I needed this several times in cases of ajax operation +I ran into this problem as well and the patch worked just fine +Moved to making sure is a bug fix and docs polish release not a new feature release +Existing tests were updatedAlso done in this patch reintroduced the which was commented out before because Equinox did not yet contain this class +Thanks Kanna +Dont use transients in Namit! integrates the comments +committed rThanks UriJust tested this feature with the nightly build it works great close for release +Letting this test run completely but Ill check in a fix momentarily +This is the svn diff for the patchLast time I renamed files in a patch in the repository it left size files instead of removing themThis patch has also renamed files so after applying the patch if any size files remain instead of getting deleted please remove them +The patch leave the dependency of on but works towards removing the dependency of on +The file in this patch is much better than the original in the sense that it doesnt attempt to style each individual HTML element +I have completely removed vertical alignment +Fix is now in CVS +Thanks Mikhail. +Issue appears to have been due to an environmental issue in my devcloud instance +I dont think it should +properties files in the serverPROFILEconfprops directory gets updated with the correct user name and password +Owen told me verbally in his cube +I was wrong true does fix the problem on Tiger with Java fixin to as Andrew indicated he would be merging it to +Committed revision Jeff Levitts patch that documents the BY DEFAULT option +I take it that it doesnt but this would helpSo thats another potential motivation +closing issues for released versions +Maybe we dothis as a follow on issue +is unlikely to ever get a new feature that wouldnt so thats fairly safebq +I had assumed that the would just allow open connectivity but Im not sure +Please release the previous version before changing this +The fix dont need to get in if gets in first +Ill add that feature +Please verify it +the old patch overrides acls after deserializing the acls from the datatrree +Can you perhaps elaborate on what behavior you want to see in your UI when a view is rendered by an? What would you want to happen when the view refreshes or a button is clicked on a form inside the view obviously the flow has ended so a refreshresume is not possible +What should we do if the submitted are invalidLooks like this will be blocked by +Reviewed patchThere is a one line change in which only removes a blank line +Committing at subversion revision +Just needed one null check +looks good can you put it on rb please? Thanks +Any updates on this? +Actual patch itself is fine if someone wants to review +I dont want to do the latter without full agreement of what is required +lgtm +May be we should move the xsds and the +This is causing Archiva to hang and I had to it a couple of time already +fixed some time after +i suggest n for Minutes like Java does +Tried with a fresh workspace and just the files above +Thanks for the patch. +Here is my proposal +Best bet is to break into the debugger and attempt to examine the of the that is being shutdown to see who owns the lock since the shutdown method is waiting for the mainlock to become unlocked so it can have a go +Can you added v at the end of the image name to the openstackimage field in openstack? For example to +yes I will file jiras on the issues mentioned to follow up +Thanks +Dialog to warn user of overriding existing SQL when Use Default is checked +Andrew I cannot get your example to run +Possibly maven RPM plugin +Is this against a build of the latest master? Affects Version cant be because no code for that version exists not even in the master branch +there is no is no patch moving to move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Instead of checking all blocks in all decommissioning nodes every minutes we should check a limited number of blocks in a shorter period +Yes this makes it clearer +Docs look great Jimmy thanksFailed test passed fo rmeCommitted to trunk +Alejandros on vacation so I uploaded his patch he made against an internal git repo here +Bulk close for +Stack the proposal looks solid. +which frees the VM shutdown process from having to wait for a long time for agent shutdown +I will continue thinking about a solution +Simple example application +Is it possible to avoid spawning processes in your case? I removed all my little helper scripts and start the processes directly which also helps on it looks like it +also an example passed on jenkins but failed locally for me so it is likely not just a jdk thingWill wait before applying to see if these fails happen again up on a jenkins run +Modified cache resolver to always return a canonical path +Just to clarify here are the differences between the datasource types and datasources contain resource config properties that XA datasources do not o o o XA datasources contain resource config properties that and datasources do not o o o o o interleavingThe other or so resource config properties are shared by all datasource types and the metrics and operations are the same for all this Id be in favor of merging the and types into a single type that would have a Transactions Enabled boolean conn prop and trait that would differentiate vs +Moved to TEIIDDES +This patch increases japitools succes rate fot this package from to +Validated with +Possible to write some tests +Hi Andy sorry that overlapped but the tests were indeed broken +I fixed both in Rev +Windows x x +Message ids XN revision +And just for the record I have not been able to verify the degradation seen in the nightly regression tests +openmetingsconfldapservertypeldapconnurlldapldapserverldapadmindncndummyouexternal usersdcmedintdclocalldappasswdPasswordldapsearchbaseoucompanydcmedintdclocalfielduserprincipaluidldapauthtypeSIMPLEldapsyncpasswordtoomyesldapuserattrlastnamesnldapuserattrfirstnamegivenNameldapuserattrmailmailldapuserattrstreetadresseldapuserattradditionalnamedescriptionldapuserattrfaxfacsimileldapuserattrzipCodepostalldapuserattrcountrycoldapuserattrtownlldapuserattrphoneteltravailldapusertimezonetimezoneldapuselowercasefalse +Currently the patch is targeted to trunk branch only +I will try it ASAP +Workaround is to formulate as a join instead not using the syntax with onselect from where and and to minor workaround fix in bugfix to the query planner would process this query as it is formulated as an outer join +should be removed as it will no longer be used +place and in classlib root +bq +Right! I was wondering why I got this error message when launching from Eclipse these last daysThanks Erwan your patch is in trunk at r +I will remeber this revision and correct my other patches thank you! +Who cares what the existing signature is? The question is whether it makes any sense is what Im asking you +This is the same as. +the patch looks good to me +for turning off the test not relevant to windows +As Mike said there were incompatible store changes that should have caused trouble with +initial fix for rev +Ah the culprit is the JMX stuff that assumes that there is an output +Lets experiment inNo discussion of memory or compaction other than what is in +At the same time I dont see why needs to deviate from how openssl works and not providing a default iv mechanism +The client will add bytes prefix padding for the first chunk +to doing this in as we can +Im willing to reopen this ticket if you find some useless data transmissions or provide a solution +Punting to invalid because of Kannans above rationale. +RohitCan you check on chandans setup to see if anything is different +Lets see +I may not get to this for a week or two as Im overwhelmed at work +The top level was added for Developer studio if they dont use the examples then are they needed +Ive committed a fix to master in ea +Fixed by SVN revision +Have downloaded demo project and checked work of VPE on it +Ive just committed this. +You have specified a name of via configName +The fix for bug is now available in CVS +It seems that you are right identifyHashcode is based on referencesDo you mind if we simply use rpcTime as the hashcode +Apologies I have caused confusion +Migration to Jira has been moved to a separate bug +I assume the XML isntprefixed with the right byte order mark which is why the parser cannot find theXML declaration +as the system didnt update javadocs appropriately in the previous patch submit so Im attaching the cleaned up version +all we have to do is wait for the release Then the easiest way to do this would be to use svnsync to mirror just the BH trunk most likely triggering the sync with svnpubsub on the target VM +Doesnt seem like there would be too much of a downside to this approach +tested on a cluster at Berkeley not sure how to unit test this +applied patchThe same changes need to be applied for the custom build archetypeworking on that applied wfm +Thanks Gaz +I cant even run it on newly created projects +Jiras are for reporting problem +Per phone triage with Burr and Max on +Fix and now working test case +Syed This is a great contribution and it would be nice to include this in the release +Thats what I was trying to sayEven though cacheKey is null the markup is still cached under locationStringSince the resulting Markup is always returned from put the originally updated markup in the parameter to put is thrown away and replaced by the old cached oneRegards Jan +I can confirm that this issue exists on my Windows machine +If I were to have reviewed it beforeyou pushed then I probably would have asked that it be fix changed +I forgotten that the std out of the external brokers is fed in to the test outputI was thinking of saving the Java brokers and files that it generates +Applied to branch +An error is printed to the console if you try to override these application is one of the reserved words +While you have your infra hat on review board isnt posting notifications back to JIRA +I will fix that now +We are closing the connections when element gets evicted from the cache +When the iframe loads it triggers a navigate event which resets all Cordova plugins +Graham your failure is +in +warn on set statements + +Good pointIf defined in the configuration definition file inherited the properties from the current builder this would also solve +The output directory creation issue is now tracked as +So all issues not directly assigned to an AS release are being closed +If you still see a problem then itd be helpful for you to provide a fully functioning test app that demonstrates the problem +And I also add the checks in and +I think overlapping positions makes sense +Yes this is a flaw should accept null message I assume +I think that logging with Debug level is enough +This is really a STS issue if you want to call it that +I see you just dropped the test code but it isnt really appropriate to close the issue until it can be committed +I tested and saw XJ Exception from Cryptography provider + +These failures did indeed disappear in RC probably due to fixing + +Grant thanks for reporting backI will remove the memex connector shortly unless there are objections +Until this is fixed here is a workaround use the vendor parameter formatoptions do define a different output encoding +Thanks for the heads up +Merged +I realized that I did the check for parameter for return but did not account in the execute callQ What I meant is the DDL is defined per model and security information need to be defined per VDB for that global scope Teiid does not have means to define this extension metadata other than properties +It was released with the fix in +Patch is up on hopefully we can move quickly on that since itll have to be redone if the code moves under it. +to proceed with this patch +Also was thinking of changing the async thread pool to a cached thread pool which will not keep all threads alive all the time but will create new threads when required keeping the unused ones alive for a certain time before purging them +Hi SteffenA interface sounds like a good idea to me good +Thanks for committing it +May be fixed by now should at least figure out if its a problem and then decide if we want to fix it +Theres at least a one second delay when switching to the tab thats displaying the page +I also get this problem w Windows Server ve also had this problem w Windows XP and Maven +Attaching a sample pdf to demonstrate the issue +A list of duplicated class files between whats already on the classpath and what is in the client jar +OK reverted the changes from in revision +Having said that I would also like to know your take on HA NN and suggestions how can make use of it +compiled with and run with works with and run with does not work but run with does also you could look at the bytecode of the class containing the super method call +this test case is passing as of the most recent passing. +Can i close +does that seem like a reasonable middle ground +New deployed. +Currently there are only and in the +Since they all specific a correct version +Assuming it is okay. +Also moved the emitter to package +We dont want arbritrary clients of our business objects to have access to the setters. +sorry for the typo its Reader everywhere in the above comment not Writer +the patch looks good to me +Our app does not need multithreaded support for http connection +After IRC chat with Jesse we came to the decision that the is not required anymore and the needs an update +ok we need inheritence and interpolation of the parent here +Thanks for reminder Martin I will have a look. +The CQL parsing is here not the problem +Not sure if have something to do with this issue but the workaround for this seems to be using failover transport +This has been fixed by +is easier to understand this problem +Committed just to trunk even though this affects all past versions +It is possible to turn the generation of processing instructions to SOAP messages since now +Thanks for the patch! +it has to be migrated anyway +Yup sorry forgot to close the issue +I know that load works with that format +Changes Everything happens now in two passes +Rebased and merged into the master branch. + LGTM +Hornetq is not part of EWP I will remove those before generating the SHASUM for EWP +My initial analysis is that there are going to be quite numerous so for now Ill keep this JIRA and open +Able to test this by slightly modifying the testId appreciate if someone could review itThanks good linthanks +The behaviour is still the same on my computer +committed to the issue since the fix breaks the non mysqloracle install +If there are no objections Ill commit it shortly +of a bulk update all resolutions changed to done please review history to original resolution type +It may be a slight naming issue +is present in libendorsed dirVerified on SOA ER +Note jirabrowse related to the same issue and similar changes +I have provided a patch for which provides the same fix along with two new testcases that verify the behavior of +I dont think there is any implementation yet though +Fixed spacing +Awesome thanks again Ton +The best way would be implemente a standar with user and roles +unit test demonstrating marshalling patch that worked for me basically just recognizes if is event and marshalls relevant info if added the failing unit test to master HEAD. +I would love a new fix which generates fewer extraneous parens and lets us keep for a bit longer +Thank you very much +Were upgrading to httpd in EWS so this will be in +pom +I think this commit broke the way in which relative xincludes are handled in Configuration +I agree +Move the fix to +Anecdotally many people on cassandra have been bitten by the current mb size +I did change one instance where it was already specified it had the outputencoding parameter which isnt really necessary +Are you sure there are no issues with the project and its properties being loaded? I have multiple excludesincludes in the multiproject properties all the time and they work fine +Sorry but I dont understand your last comment +sample code +In a real world setting the value of sld we read from the servlet has already been rightSo we should not get any escape codes in it to start with +I just committed this + description contains step by step what is done on the workstation and on the maven repository my modules are minimal they contain empty classes to reproduce the bug on your host you need to change the repository for your own repo you may want to disable the plugins checkspace javadoc +Still need to figure out incorrect offsets but the gigabytes a synonym of Best buy part was a display error in +The guide looks good now. +lgtm a lot Sijie for the review +Basically this issue is already fixed in trunk due to next milestone release will contain the fix +not valid issue +This was a simple bug +I agree with you that it will be difficult to look this information up on a web UI +Im going to make this a duplicate of all macros in parsed files are essentially since parse doesnt happen until runtime. +In patch v active is no longer usedTo verify classloaders being collected we may need to trigger GC after table is disabledIs that what you were thinking +Closing issues for beta . +Schemas have been created and all relevant files now reference schemas and validate +Great that you managed to write test +Wayne seems to have generated some syslog constant code for windows? Is the bug that we dont throw a or that our impl is binding to msvcrt and not working +I suspect this is an error message that should not be triggered and its possible to carry on working with forge +Section updated and Section now moved to in the docs +I did not add the whole statement +In were only recording the addition of more bytes when a new block is created however because previous blocks may be recycled it is the recycled blocks that are not being recorded as bytes used +Thanks Ryan! +I have a fix now and will check in soon +Not for commit here +I looked at the patch too +Had not generated patch correctly was missing file +all +Can you explain a bit more about what you want Archiva to do? It sounds like at least two different +Shai I actually think Earwin is right with his claim that this is unneeded pointless really +Changes corresponding to the latest patch + + +The next upload will be accurate +Moving this issue to the command is working properly the problem lies in the components not setting the description on the service +Quick clarification you are using the JSP tag library not the macros? So this is a bug in the and nothing to do with Freemarker? +The check on active is changed to a check if the is a part of a since it will already be added to the bpmn diagram from the get method +Please reopen if there are reproducible performance issues. +pushed in is the release status for this? is there any minor release in that include this fix +Ive just committed this. +Though the patch is a way to get rid of the dirty files but I prefer to find a way to put those files in the target directory +believed fixed +There is no way to verify the login credentials when Account related components are not fully loaded +No GIL. +Will require investigation +revision fixes SessionitemExists SessionnodeExists SessiongetItem SessiongetNodeand adds a FIXME comment for the identifier path handling + st patch +The bulk of this patch is improving to deal with having two edit log directories +I agree its a difference in design perspective +Hi Aaroncould you attach a small sample project that shows this issue? That would be helpful to can replicate the issue with the attached Eclipse project +Upgrades hadoop to Lets see if similar error comes up +I now see that I didnt grant rights to include the jms and the http test case in ASF work +And the comment is not added to handle since there is no failed offset request rate that we track +Instead of sorting using the in +I tried without the dates and there are no errors being thrown now +Hi will run some test to increase the pressure on the route +So once is resolved this could be resolved as well +And then youll have to add the generated directories to your IDEs source lists +Ill commit the changes shortly +I will configure it as you request then go from there +It should have gone to + +test with esxi on build +I have verified the fix with a set of security scenarios which involve and Conversation +Unscheduling due to inactivity +In other words the repository is real but you may or may not wish to display it in a JOPR plugin +fixing in favor of +Thanks ThomasZhijie to be safe I think we should go with Thomass fix rather than mine since its fairly late in the game for +Add Still polishing and refactory the code +I was never comfortable with memcpymemset of objects because it is not good coding practice but it did work on OSBut what is currently being done is equally badI am not that familiar with the code to understand what exactly needs to change to make this a cleaner implementation +This is part of the logic to simulate failover +Works OK with Maven removing +That sounds good to me as well +Im looking forwared to seeing the next reportSo youre going to debug GGC itself? Wow youre the memory guy +just had a look at the velocity svn trunk the code that throws the exception for control flow is still there at +Fixed a small typo +is not the cache location +I close if needed please reopen. +Maybe by sampling the world every degrees in lat and lon and taking the bbox of the points that managed to get reprojected or something like that +Please verify +However none of the files in the workspace show any inclination to add this jar into the result jar war ear client jar etc +The linked issue has been fixed the JBIDE JIRA is fixedThe deployment scanner is now set to interval by default which breaks deploying from JBDS at the moment +I was thinking of having the heavy weight heartbeats have a low frequency like every few minutesThats too large especially for jobs where latency is important +I must confess that the errror may not occur in that ve upload two images which indicates the difference between QGIS and uDig +When text is selected only it will be externalizedthumbnail +Unless we set to false but then we would not use but a plain +Invoking the stop method on the object should stop the http server +Is that acceptable +Bring it in +I think this would be a great addition +Committed latest improvements of Udai almost as is +Heres a version of the interface with more javadoc descriptions of what the methods should do +Lets apply the patch and then close this out as duplicateregards for this patch to Alexandr Smirnov +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +If the fix isnt too involved I wouldnt mind seeing this in +The call to get is outside the try block to make sure is thrown per the interface no +If the files are three levels down in a partitioning scheme +So to myself for now +Contains source files compatiable with to add gemfireHave you had a look at this? Any thoughts? +I just checked this on a fresh install of build M candidate +Attached patch + +Patch attached +Discussion and testing moved to this duplicate of +We cannot do this in because all patch releases have to be binary backwards compatible and adding a cluster name to discovery requestsresponses will change binary cluster name is now sent with discovery request +Great news +The patch does not apply cleanly for me +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I used this commandfind +Ivan can you please review this and commit it to trunk and when you get a chance +Since there seems to be a difference of opinion can the visibility label syntax be pluggable +Now to figure out how to get it to do multiple builds +Attach a draft patch file not the final one +multiple threads +The issue will be closed +duplicate of +Please mark it as resolved +Forgot Username is dandiep +Other ugly workaround is to put full path in + + would have to either stay the same or be modified to take into account +Sorry I didnt realize it was that badly broken as indicated in the thread you pointed out +Is that value to lowCould a js suggest which uses no special search component also lead to this behaviour +Verified by Ruth. +can change the version affected like switch to to AS if you want to but this is something that should be resolved eventually +I could use some more explanation also +Any hiccup in processing will cause FD false positivesIve pulled a fix for this out to +I think this is really a powerful idea and Im definitely going to begin researching its inclusion into +I think having get as it is in the patch is fine +roadmap meeting. +I think Id rather use a different class so that Views purpose remains clear +Moving issue to. +NOT TESTABLE WITH CURRENT TESTSUITESPEC Testable true Section a portal encodes additional information as parameters it must encode them properly to avoid collisions with the parameters set and used by the portlet +for the patch +They can be included in the code base disabled by default and marked as experimental if not tested well +I will change this to Hadoop Chukwa Team +Reopening as there are some more possible improvements to +Cancel +Preliminary patch which fixes the test queries I havent run through regression tests yet +I think this was not an issue on the broker side +Only tag elements with a tags parent are selected +Fixed in trunk and Branch +The in the examples above have a typo +has been marked as a duplicate of this bug +SF from directory remove digest records from zip back to jar file and replace the original one compute hashes and replace +Thanks a lot for helping me out with this and for your valuable feedback +We return what is in the +Do you want to make a new patch and I will get things up to dateMaybe we should close this ticket and open a few smaller ones for remaining TODO items +it seems to be reproducible only when the jar src jar javadoc jar are located within the maven project structure +But I would be fine with the approach of adding it and then if we see it breaks something we can remove it +What a colossal waste of a day tracking down that one +RobbieCould you please review the change +So no substitution is occuring +As long as you keep this logic in the dialect you should decide +Without trying it I think it would be +Just wanted to add the response times as wellWithout Patch secondsWith Patch secondsThis testing was carried out with with Hive I understand correctly the issue is that even when the blocks referred to by a split refer reside on multiple nodes only creates the split with its locations referring to a single node +Thanks +Thanks the mirror seems to be working OK now. +I have a similar issue with the from +Added Apache License +Sonar IDEA plugin is no more supported so I prefer closing those tickets. +should also take into account fieldresult. +Im sure that this issue is not a bug and just exceptional behavior of RI +issue +They both have a warning about float and double being not supported so we should be fineAbout the error message see my email on +Indeed +The root directory wasnt listed as undeletable +Added a jar of youngms seam spring integration with src +to trunk +This may appear confusing but it is necessary for the packing mechanism of python +The mapper tasks from one node are all in the penalty box +It is now in Thanks for reviewing so quickly! + can this be resolved +As you point out it is duplicating the behavior of cat +Attaching the release notes for the third release candidate +Am I correct +Thanks Ondrej +Need a builder pattern i guess rather a factory +check at live demo pages +Seems that browsers render the boxes differently to other browser architecturesFrom memory the issue relates to the frame size set in the Publican brand and the type of logo selected for the published will have the Publican team look into this and rectify the frame issues in the brand files +Cool! I was observing this too but never got time to investigate. +closing self reported issues resolved on versions already released +If you can get a chance may be you can also do the same for api? Actually Im wondering would it make sense just to rename trunkintotrunkand specs bundle given that it only contains a rewritten should be usable in cases when cxf which currently depends on api is used +I couldnt find any despot docs on svnserve so I tried a quick bounce of apache with sudo usrlocalapachebinapachectl graceful +All the jsp pages do have a baseUrl but then all links are using something pathtourl the initial makes the browser link to the geoserver root anyways forgetting about the extra items in the path +The same problem +The code was corrected to eliminate this errorSeveral tests were first added to replicate the problem and to then verify the change does fix the problem +All tests pass for me with this +Shall we do it in a different jira +Im marking this as fixed in per Permaines comment +So all issues not directly assigned to an AS release are being closed +Looking at our application this has been fixed. +defer all X issues to +All my coding is done +I actually see it with the Sun JDK run as well +The usual problem in this situation is with AIXs support of dlsym as needed in +defining the repositorys URL in the module configuration +I am a bit confused +Ron I replaced this jar with jar in commonlib however the error still exists +Please verify +But it still does not affect the actual code being executedWithout turning the compiler optimization on the input and output variables are copied in and out inside the loop which starves the pipeline +Are you sure you were using with those testsBecause if you do thats likely a phpcassa bug that doesnt respect the consistency level youre asking for because there is no trace of any query at in those logs only queries at +I wrote it with the intentionof just getting people in my team on track with the installation but putting itup publicly as well as well +properties +Later modules will also be placed inside this directory +As an addendum I have attempted to set the formatter within the client to define an appropriate for the server and have discovered that the method get references static method which assumes an Envelope and Body even in a GET method +Are you willing to act as a student or a mentor for the India ICFOSSMentoring requires you to have a good knowledge of James internals +is this not because there is a + + +Committed to HA branch thanks for the reviews all. +The problem is that the test asserts about app state not matching and shuts down the minicluster +Committed to trunk +buttons on the server view relate to the server view not to something other random selection +LinAre you caching the object or the object? Is this done in your own version of a? Do you have a code snippet or can you point me at the source +English messages was translated by Google who knows German could translate them properly +Also setting might obscure the real localaddr Im changing it to volatile thx. +Thank ll try it out +Hope this does the job +usrbin +Nils +Jiri if needed create new tests for these two attributes +However it will be IOdisk bound +I cannot reproduce this issue +If under some circumstance someone had intentional interrupts where this was expected they could set property higher to avoid the javacores so to the change +Here are a couple of newer screens showing and service invocations +Committed revision . +The admin shuts down the cluster turns on the permission checking and brings up the cluster again +See for an approach of adding additional faceting components instead of overloading the now +If we cant find a license that is suitable we might be able to put the component somewhere else +and not attempting to stop the container this is a useful feature we have in Cactus and that has not been ported to Cargo +I messed around with that stuff more than days to get it work +to the changes though dont we want to expose a way for them to set those variables with standard hadoop config possibly namespaced with pig +aar file is not deploying properly +Many thanks for the patch Felix +Thats correct about isX not being recognized +The problem is not with XSLTC in particular +You should run ant checkstyleLocating checkstyle errors is tricky +This looks right +For exampledouble a bca b c a bc always has a value of Infinity +Is this also an issue in upstream tomcat download +Hi Dmitriy do you want to take a look again +Fuhweis patch looked good so I applied and tested it +Rick I marked as fixed as the source trunk contains the fix +I added a backwards compatibility section the the Jackrabbit release notes and included the following Repositories that have used the new JSR security features included as a development preview in Jackrabbit and may face problems when upgrading to Jackrabbit +Can a committer take a look at this please? +The only way I am able to get this node to rejoin the ensemble is to stop zookeeper removing its data directory entirely then copy the entire directory from another peer by hand +Ill look into it. +Please apply this patch over Trinidad trunkJeanne Waldman has reviewed this patch already +you can have the take symbols in its constructor to let people define what the prefix should be I can submit a patch for that version if people want it +It results in dates saved as instead of +on commit +In QE we have maxed the pool sizes however with this patch applied less sessions would time out due to fairness of processing requests +We switched to dbcp as you said but the error still occurs at midnight every night +Ken Cai suggested a solution that we dont need an extra file +Problem could be related to the sample web service I was working with +That is what I was trying to say you can have CMT bean and yet use Resource interface and execute method on it successfuly +Thank you. +As far as I can tell the sockets are still connected they are just waiting for the server to send them some data +yes the initial intent was that the ui shuold be as simple as possible to avoid we look for id properties and if you dont set one we dont add one and you will get an error by the validator that you can easily fix and then run the generation said I agree for missing Id it makes sense asking or at least pointing out Id needs to be on though since its a bigger changefix +Is it such an issue having two namespaces +Attached new version of patch +Thanks Amareshwari +The lookup of a codec by is exactly like looking up a charset by name +Thanks for removing the warnings introduced by raid +Thanks Greg +done for the new tests still need to go through the legacy this really only applies to the changes I have already made +for your helpThe problem was in embedded modeJust fixed it. +It generates associations +So theres some translation from CSTCDT to USCentral not happening either in Joda or J +If some replication has two digits the format wont look good since is a fixed String +That almost looks like it could be going away for project main website stuff +Issue is resolved +Talked to Aaron Kimball offline and he pointed out that HSQLDB can be run in an mode +The last one Ive not done +Fixed by issue is fixed by +I will commit this within a few days barring objection +Adding it should be noted that the noSession filter only prevents new sessions from being created +Closing issues resolved in released versions +of course +Should be an easy to fix will do that soon +I commited a fix for this +Seems simple enough +Thanks for the unit tests +We would soon have to split the for different core hdfs and mapred +Reducing from blocker to major per Max Andersen on irc. +Knut you are right about upgrade test demonstrating what does not work in releases with defect and newly created triggers on or triggers working correctly +oraclesun Java? possibly Java +HiCould you please indicate which goal youre refering to and maybe provide a test project to be able to both reproduce your issue and better understand it? this is from way back +Resolved in nightly build via minor docco tweaks to the supplied patch +Anything we can do for getting it fixed for the next version? More tests? A patchThanks +either the race condition has changed or this is fixedThanks Dain +if this works for hypersonic that is might consider to add a note in the installer warning for this absence of and processes if another db then hsqldb is selected +All Apache Maven plugins have a minimum and maximum Java requirement of +Im sorry i dont have good newsIam using the snapshot version of hibernate which integrates your patch inI have enabled the query cache et added the specific fileBut my tests are still failing in sometimes at the start sometimes later during inserts or updatesI will again verify all my configurationIf this persists can i join some traces in this issue or should i open a new issue +Revision +HardySteve mentioned that you were looking at making this change +you could base it on latency instead of pending countIm not sure I understand do you mean the latency the pending reads would have if they came back at the time of calculating the scores? I was thinking about doing that since we have the timestamps in EM +This will be fixed before is released +Jason could you please provide a patch as well +Ive linked the JIRA for it to this issue +on patch +Take that file listing tar it up and run md on it +Just a note about how I tested this +Should be resolved in the release branch and in trunk +I should note that the patch does not add a format command to the NN yet + patch for review +Logged In YES useridReopen with a definitionexample of a self deployed. +Ive committed this to and trunk. +That might mean many things but usually it means youre pointing to a data dir thats existing but does not have a you provide the full stack trace instead of just that line +However it doesnt apply on current trunk any more +Atomics are not free Im willing to bet there will be a measurable performance impact +Moving open issues to the M release +fixed couple of bugs in dtd +Thanks LeoPatch applied to INSTRUMENT module at repo revision rPlease check that it was applied as you expected +to what Dave saidWould it be possible for blog to cross post to or other places? That wouldnt be a bad idea. +I couldnt quite figure out and how to reply so Im posting as a comment here +from the Tools build and update site +Thanks! +Igor how do you feel about setting up a build bot +If you find further issues please create a new jira and link it to this one. +your commentssuggectionscritics are highly appreciatedThank indicates that there is a patch ready for commit for this issue but the patch available checkbox wasnt marked +However the actual exceptions to be thrown are defined in which doesnt have listedSo I guess we will have to leave the as is +The previous check wouldnt have caught any corrupt data other than the test case +All of this occurs as part of Derby compilation hence the error code But if thats too much of a technicality I wont complain if the error message is changed +Committed to be checked in Hudson. +Carlo with the dependent issue fix this test and close the there are regressions please inform with jira issues +Screenshot is an issue with which relies on xmlbeans. +I have attached the simple project that I used to recreate this issue +Here I am attaching the patch in which I remove the return part +I have fixed the release notesThe release notes was already correct +could it be done using Spring Data +Ive added a test case that compiles but who knows if it actually tests anythingAlso added hashCode and equals with javadoc for the requirements of implementation classesMight be useful if you like test first strategy +If you want table name to not be checked for reserved words then use quotes in +push out to +Ive uploaded my patch +Checked in Label Manager missing +Was stuck reading so bug in the class read in javac? to trunk and. +Fixed in +reduce the icon size to the same height of the darker green band and move it next to the user name on the rightmost but you can do moreI dont like size position and brightness but doesnt matter youll never find one +See JIRA commits on + +This revision does not correctly handle the case of a path constraint consisting of segments with namespaces when the namespaces are remapped +You dont need the Logj jar since it comes with Grails +I suspect its not right +Indeed does not use a delimiter between two Strings that are appended +Patch against latest trunk +sha cbefecddffcccae +The cloned issue is a little different and is ONLY for Firefox +I think so +I have several components that can run with a normal or special profile and I want to be able to switch between modes on each component individually +any log on what has failed? buildbot logThanks +New images pushed toNot sure what else needs to be done here so resolving. +Server View Structure +Ive begun a very simple implementation with a target and a fixed quantum without any attempts to fairness between concurrent IO nor any attempts to achieve an average rate equal to target maximum rate +In that case its probably better to also not deadlock and force closing the sockets too +OK I fixed the findbugs issues and the test failures +On the other hand a can just be seen as a seperated name +Did you try with felix trunk +theres probably nothing wrong with your ive checked in the project and several rounds of changes +How does this work with search domains in +So definitely the patch in revision +We are hitting this problem often +Yes Cedar you are right. +I think the tests were timing out because they were trying to recover the index during startup +Work has been completed on has been submitted to Carlos Sanchez +I will take care of reapplying the changes associated with this issue at a later point +I will submit a new patch as soon as the bundles have been released +Alex yes that is STS +we dont need to unit test +These files can be kept locally and copy them in to +JDT will handle output locations and this version will address this bug +bq +bump +Problem still exists in but it works +Ill clean up the patch and post it here some time this week +Whatever process reads the output you imagine can read the current output +It looks like both jersey and jmx would be similar amounts of work +Can we close this? Or is there something missing from +Thanks for the patch +this was implementedThus setting this issue to in Developer Studio +If I can crank up the buffer size enough that the error happens only occasionally instead of frequently that would be a big improvement over the present situation +defer all issues to +Well in the project that I have attached now running rake testfunctionals here the problem is happening +I can try with another encoding eventually +I now understand that there are two different launching mechanisms +It looks like we are using with JAR naming conventions from Jetty +There may not be much of a user impact caused by this but it is surprising to see this change caused by the patch and it will affect automated UI tests +Timothy Thanks for that +itd be a petty to have to wait with releasing it another months just because of technicalities +I dont think we can assume custom dictionaries are going to be small theres nothing in the which force that +Thanks for the reviews +I dont know why you get the axis as am not sure what happened to the PDFIn the attached JPG I saw the default as Enterprise Application Platform when I was using SOA Platform I think it is using the Type of the Runtime and not the name I gave it +I would even argue two airavata instances should also not check for uniqueness across them +On test setup the code checks if there is such a file and if it is possible to open it +Updated The changes can be tracked by searching the patch was correct but most likely the code has changed in the mean time. +used nd file +Unassigning since its probably better to revisit the issue after and have been addressed +srcexamplesthrift needs license +Patch is ready for revisionAll tests are passedNeed testing on a real cluster +I will create vectorized versions of +All the methods in these interfaces are allowed to throw but Im wondering if the spec allows only some of them to be not implemented +Hi CamilleThanks +So this PR should be pushed to branch too +Thanks yes i forgot to close it +The new api is changed to take a verbose switch +Ok then you can close the prefer to leave this ticket open in order to check situation when provided URL is not correct +This is an improvement of the documentation created by +Which risks would you like to have documented? s javadocs already state that it doesnt work on yet just on you send me a sample EJB stub to my address? I might be able to track down the issue a bit further +Julien Yannick could you try and report if implementing works or not +This patch applies cleanly on master and the is used elsewhere well keep things consistent +As far as miniclusters having just seen this on let me pipe in again pick dir names and ports that will not collide with various Hadoop mini cluster impls OR as they all run tests in parallel and can conflict in confusing ways when their dirs or ports collide +When you say merges some +It seems you either have to import or provide a full classnameIt also means that you have not testing thisLooks good to go to me when that is fixed +Will release a snapshot in an hour or so +Oh i got it so it means we can not use using or +patch contains proposal for the copymove use contains proposal for the copymove use for patch +I dont see that causing too much overhead compared to the convince of having a single file to look at for the har tests +Merged this to. +It doesnt seem like it will be easy to maintain compatibility so as long as we are breaking it we might as well go wild +If anyone wants to take a crack at this let me know. +ON a Linux machine the server is listed as java no mention of not find anything on the Internet stating how to change the service name for Hyperic HQ +Verified in GA +Part of a bulk update all resolutions changed to done please review history to original resolution type +andI havent figured out how to solve the problem when using layoout pageDirection and pageDirection +Thanks EduardoI have committed your language pack to the SVN. +Only if it is declared as Object then it is turned into an array with a single element of type Object array +Thats what this JIRA focuses on abstracting common token from all kinds of authentication mechanisms based on identity products and backends and using the token in Hadoop + for the change +Reopening due to DB test failures due to switched off for elements +Are you sure about the expand? I dont see the expand element beiing used in +Andreas Thanks for taking the time for reviewing the patch +Incorporated issues constrained to next release. +Plus this is something we can discuss in the dev forum later +We need to more careful about modifying serializable classes +If there is immediately following or in reducer and some other conditions are met we can apply Secondary key sorting optimization +Looks good I just committed thisThanks again Ryanps if patches start in the trunk its easier for someone to commit it. + makes sense +Thanks Jon! +I couldnt reproduce this. +Wont fix + +Marking this open as duplicate of to +the short term options are to use lang for callouts no highlighting OR language for no callouts with highlighting +Final patch with tests changed to accommodate a test option. +Fixed in r together with +In contrast to the previous version this one seems to be fine +is wrong +Did you use the closed o instead of open u for a reason in this versionthe white diagnoal lines gives a nice break but I had envisioned it was layered opposite +I ran the smoke tests on dozens of times +If yes then the test program should whole discussion is now out of date and irrelevant of course +Example JPEG file that causes when parsed by Apache JPEG file that causes when parsed by Apache is fixed in SVN HEAD by commit +Thank you for the additional fix +Tom can you create a development forum posting highlighting this issue and pointing out the wise code which is in errorOnce we see the code you are concerned with it should be easier to discuss it and decide how this can be moved forward +Committed the code +Also considering the fact that many people want to move away from ftl driven tags that only reinforces the decision +Chatting with Greg it should be done this week +Overloading multiple meanings on a property could lead to issues especially for tests that test the operation of itself +moving off of this milestone havent had the chance to work on this issue yet +committed to for clarificationLucia home come you set version to BetafAnd is there still an NPE +Go to blank page memory is not released +Its still failing +Andrzej what we propose is actually a very challenging task but it is better that we do not rule it out before careful analysis +involves those poms which do not need to inherit a parent pom +is creating a new instance of for every call tocreate but should we instead have store the instanceAnd should the test if the is closedReason for doing this is of course when using the that depends onthe other words how do we want the to be used +Looks like this might be relatively easy to fix there are direct class references from which should be changed to use class name +Hey StefanI have been out for a while +I still had remnants of +Patch Thanks Loc! +In this particular case its probably not unreasonable if the exception is triggered from Java and not via a Ruby Threadraise or Threadkill +It wont run on BG where we run with one worker thread per place and STATICTHREADStrueThe same hang can be observed on none BGP platforms by setting STATICTHREADS and XNTHREADS to true and respectively +Tests pass except filterjoinbreaktask which failed once during the full test run but Ive not been able to it again +Added code to drop permission descriptors from SYSTABLEPERMS SYSCOLPERMS and SYSROUTINEPERMS when the object they provide permission for is dropped +contenttsson my Windows machine all these steps work +May need modifications +Regarding to Stefans comment I think that there could be a new directory in svn called examples or something like that where we can put different configuration examples +HildaIm not sure if this is really an issue since as you said it is the intended behavior +Alexandre after further analysis I am a little bit stuck +class +and fixing. +Jacoco was in component builds +verified that passes in ER without workaround property in code in Affects and Fix in Version +Yes just for truefalse its not documented yet +it should be at least backwards compatible and have some decent error checking if the configure class does not exist +created +So it you require this support then raise a new JIRA I guess outlining yourColm. +Agreed +See my comment in for the link well keep an eye on that one +For the problem of all types being string that was definitely the case in but it was fixed for +Now Im wiser +I tested it on my windows machine before posting it and tested it on a linux machine via ssh +I imagine I will have to setting for the former initially but def thinking about how to deal with that case +Just copy the new in the build directory of the FOP zip file over the old one in lib +on a trunk standalone instance looks fine +Would it make sense to rethink how were going to deal with this? Perhaps it might be a better solution to unify plugins and variablesAt any rate I dont like the syntax +I dont use test in my project but I think it should work too. +This thing seems to be really trivial to do +I suppose my one guess would be that the compatible image might be doing some gamma correction to the specific gamma or even icc profile of that device your monitor +However after repeated clicks it sometimes works +Havent had a chance +committed +it looks good to me +Done +Thanks ysteinI ran the regression tests over again cleanly withCommitted at svn . +Tried to do testing with a long running sleep job and copying over derby DB contents from a distro to trunk distro but DB got locked +Move to +BTW it would be better to put the patch in jira when you uploading to review board +Reopen if I got this wrong. +Good catch! Ive updated and syncd everything to milestone a +This patch fixes the problem by setting the HADOOPCONFDIR earlier in the script +Thanks +Is there going to be a? Is there anything I need to do to the above pull request to get it considered +The same block is impossible to be on the same node +That and also whether negated wildcards actually work reasonably or whether I should be introducing a nothing here special token that would make the wildcard queries unnecessarybq +The secondary map needs to employ weak references for the values +These need to be dropped into srctestresources so the unit tests can find themAny feedback is appreciated +If RS dies before even startingto work on the region is there some timeout mechanismthat kicks in and master realized it needs to give thisregion to someone elseFor now whats the manual steps to be taken to get master to reassign the region? Restarting the master would workI suppose for now +Ill look into it +We plan to address scoping in the you try to use plugin? project already supports JSR s an alternative but Struts should have it +Hi JohnI think your assessment is correct in that I think my usage model of Maven is a convenience wrapper around ANT and using the event driven model around compilation etc +Fixed with r and close of all resolved issues as part of closing items. +Only the Eclipse VJET features need to be exposed to Central the rest are children of these twoDo we want to also provide the Dom Kit feature? I can see if theres a way to include +but only type is run atm so thats a prob for of date now that guvnor uses the git based VFS system. +Unfortunately I do not have any more relevant logs to contribute they all look like what is already posted. +Bump to well move back to if people are hitting this +once you get a clean ran the tests locally only tests failures are those from +Please be advisedWith this change the qpidd broker will not run unless the directory varlibqpidd exists and is writable by the user running qpidd +I dont understand very class suggested here its resolved problem with error posted by Franois Le Droff? I refer to last lines about license +Hmm apparently you did not update the sandbox examples could you test your changes on the provided examples and then also provide a patch for the examples +It is of no importance whether or not the view with and without docs included have the same as they are different +Ill go look at getting a release out that includes it +Drat +The speed improvement is clearly visible +Im quite happy with this perf improvement +Still waiting for results of following jobs others are OK +The weblogic SAAJ implementation seems to be pretty fundamentally broken +on the other hand the problem has been resolved as thomas indicates so there is no need to see if the cleaning code works better +Bob Drools +At a later point after gaining some experience with it and hopefully some wider testing we can start making use of it when it is available on the classpath via ClassforName or decide to simply fold it into properDoes this plan work for youOleg +Added a rawinfo method to Message as you suggested +I just committed this +I committed this and removed from. +Thats what this proposal was aiming at correcting +Attached patch fixes the failing unit test +It is noticeable after several shut downing +The patch needs to be updated because its conflicting after commit +Hello TeamActually were looking for the bug by producing the smaller application we can +I can get a repro on Ubuntu Lucid as well +not committing your patch can you please elaborate as to why this was necessary +Well if the code follows the Project style conventions has the Apache header and works for Trunk then the only Procons I see to commiting this to the code base are An additional renderer which may be useful for some users although I think its a very tiny set of users +Will commit to tonight if there are no objections +Using in the common header at the top +Originally the issue only covered the SDO samples +i forgot to add the new files +Ill create another issue for that change +Id like to stay as current with the trunk as possible +Ill send you in few days a complete tested solution for encoding xpathquery my last attachment is just an exampleYou can then decide to embedd it inside or to expose it asutility classIll post it as soon as possible +Or the Abdera JIRA issue number +Fixed on revision . +You onboard with that jvs +This better represents what the Teiid sequencer is expecting and is compatible with the JCR behavior +ro mirror now points toPlease allow time for github copy to sync +Dunno yet why +It would look better to me if there were an additional line break before the warning +I added the fix in revision based on Erics proposal +We should add support for virtual hosts as well as being able to filter on ports +Unreferenced +CLI tests for this subsystem should also run in domain mode this would do the verification +Thats why i added a custom validator which i thought could be improved later to test diferent issues with table structure +Please let me know if my patch fixes your problem +Also changed the text to be Metrics Dump instead of JMX Json DumpThis OK to commit +Thanksdhruba +Good point! Ill give it a shot +Many others to go +Already had them test on problem persists +lgtm +We havent had running out of file descriptors for a few days nowI did see the systemload spike to a few times though so the original issue which started the whole thingy still seem to exist +This makes the patch slightly larger than it is required to be +That property can be read inside the type class by implementing interfaceAfter that there is no more fixed properties in methodI sent a pull request to branch +Ill fix the JIRAs targety is the Yahoo! internal release of Hadoop where the initial work on this framework has been performed +This is specified in the parser syntax its not easily made configurable +No tests manually tested by changing to trigger failures +Application server classloaders make the ejb interfaces visible to other modules by default within an ear behaviour should always be spec conformant +We made the decision long ago to add just a few methods for common Java packages and I think the value of having those packages outweighs the fact that weve noised up Kernel a bit +In the new environment I will be using the SNAPSHOT to send live messages to test real world performance +Im using the Sep build on Howards site +Cheers in site project +Calls to trigger message sending and calling of SLSB and SFSB methods +Just like in BPMN +Now you use both the name and the ordinal in the parameter +Any chance of this getting into the version? The current structure of a project with plenty of jar dependencies make it very hard to work with eclipse +Thanks for the patch Izaak. +The wizard comes from WTP and were just integrating a few pieces with it the wizard itself is a mess +Some native process on windows updating file + +Closed +We need to evaluate community desire for this patch. +This is serious +For some reason doesnt need to be shutdown any longer +Can you please take a look +The implemented can also be indicated in the documentation of +Duplicate of +However it could be frustrating for others who have already implemented s if you do change the interface though I certainly would not mind +Im not sure if this patch should go here +pretty much the beginning of the discussion in this ticket +Hi JoeThis is only in test context right +v allows comma separated paths to be given and added some more I meant space after comma and not in the file paths itself path path +grep jar +The tree is available in theThe sandbox will be included in the next release. +We should also write logging guidelines with this +Thanks for your understanding +Since also makes sure that coreWeb is running on the server the check is redundant +You can reach me by email or mobile + +Patches attached for all repos +I am still waiting for legal approval before I can commit +And also Vertex class reflection seems problematic +Tom can you go ahead and upload the jar which you wish to work with into the repository? Ive made the necessary changes locally but would like to go ahead include and your jar before into the build script before i commit the changes +I think it makes more sense to track from there. +I suspect that this is an issue with passing a large amount of data across an intent and manipulating it which in my experience has always led to disaster +This will affect configuration and documentation so those will need to be handled before I close this issue + +didnt handle null +Andrig is right this JIRA should capture this and we can look to find a solution that is sensible for both and Openshift +The subpackage variable shouldnt be too hard even though I dont really see the value. +This makes it difficult to put the project under source control +All tests pass with this patch in the current HEAD branch. +I will apply your patch soon +This patch is patches in patch +It can pass both my test case andBasically the comparator doesnt have problem +bq +Do we really expect it to be used in other contexts? Maybe we should add a Xnobootclasspath option to be more explicit +Investigating! +Matus we use several custom filter implementations that commonly deserialize some of the data in the region server and filters out rows based on data in the row column or cell +Do you agree +You want configurable shell? I think it is overkill +Marked as minor because it currently works and is not blocking +What is the restriction for keys? Do they need to be globally unique in Apache across projects? Or is it sufficient that they are unique within our Apache project +There just needs to be a consistent way to query that jvmRoute via CLI regardless of how the jvmRoute details are specified +there is no sense to discuss the reason cause here is better to provide a concrete arguments may be you are right with patch characteristics like improvements +I have edited and where needed +Sorry my bad not fixing this +Ideally it would do both in one shot which is what i did +I am not sure what you mean by but a large portion of this patch deals with indexing to Solr +So I can declare a managed bean in and still use action params +Ive pushed this feature into the official repo from should it should be fixed now +Interesting changes +Attached is a patch proposalThe patch modifies to enclose the column name in double quotes when generatingSQLThe patch also adds a variety of test cases to basedon the repro script for this issue using a variety of table and columnnames in both mixed case and all upper caseIm still running the full regression suite Ill report back on the results +the test +Committed revision . +Let me clarify +Moving to CR +Added and properties that can be used to control how the endpoints are displayed as well as allow the user to resolve duplicates +When you add a servlet to you need to check if the whole of the contains the new one +Reopened for trunk fix +If they say yes how can we doTwo choices add the servlet into while remain the old servlet that has a same name to our added servlets add the servlet into override the old servletWhat should we choose +svn problem. +Thanks Jerome. +Thanks Scott! +But knowing that the groovy compiler does no form of inlining atm and this issue means we have to add that I would say it should not go into but might be is a bug not an improvementI havent double checked how its working in Java but since the field is a String constant it should be allowed in hence in wish it could be fixed in but if its not it should really be fixed in +Will you propose a patch? If I have enough experience in using PMF I will do +Changed the title so Im not occasionally tempted to close this since we support SSL +Why you refer to group +Ok resolving fixed then +And Ill trap the usage and create warnings or possibly errorsThe errors youre seeing are general SAX parser errors +Works with current under IEFF of a bulk update all resolutions changed to done please review history to original resolution type +Go for it I dont think my opinion counts here obviously +Can you please try this against current TRUNK and report back your findings +Modified +Olivier I tested it locally and now it works fine +FrancisCan you upload the patch on jira granting licensing rights +Ill commit code that does that in a moment and you can see if that makes a difference +Will commit to and next week +It ended up that aix has always required which I added and got through the run +But that was fixed in the next revisions as of late Friday +It is so nicely historical Mark the newer one as duplicate +NET website open the in Visual Studio or Visual Studio Express and Build from the IDE +Patch is workingThanks Prasad +Hi JasonOkay Ill get the fix inBTW there is no license grant on the demorepro +Luke how do you prevent map tasks opening sockets receiving connections and communicating with each other in Hadoop? Isnt that the same case here +Is there logic somewhere in Maven to decide whether a version number is valid? Why would Archiva be more strict about thisIve never had a build fail due to an invalid version number +committed in revision Alexios Thanks for you help and please feel free to post any ideas or code for better handling of temp data +Thx. +Chapter should be removed it was new content in the upstream but mostly related to downloading and setting up the might need to be added as a known issue should go into a migrating applications section might belong towards the start of the programmers to Darrin +Daniel please verify +Rebased applies atop with r reverted +It should be easy to add wsdlURL feature in a similar fashion +Close a project means +Looking great Ross! Feel free with things like this to commit and close out the issue +When can we have this fix? Im using nowI would like to know eap version and its release date approximately? +Unless these hooks are removed when each instance is unloaded all the classloaders will remain in memory and the app container will eventually run out of spaceAdding this optional cleanup method will enable apps that run in an app container to remove their shutdown hook on their own so their classloaders can be garbage collected and the app container can continue on +Reflective Resource some new classes which work via reflection +Ill send an to you offline when I have this fixed up +IMO unless this is a proven problem with data from aprofiler I do not do this type of optimization + +We are working on some changes for which cause this disruption +Nice find wrt the mutating configuration issue +All is good now +It definitely needs some refactoring in the end +The median response time looks good compared to the th th and th percentiles +To fold into the general migration of material into the new CMS website +Rebased patch +This is for trunk +The Loading node seems to appear but only for very very short time +We have not tried in yet +can you see what depends on it? afaik we dont use it +Bulk closing stale resolved issues +I ll attach the screenshots for better overview of this feature +I have a glyph not found message although I havent found out whether it is a real one that was missing and not just a space +Now its officially part of Wicket +You can use the Required annotation to mark properties as mandatory and the to customize the name of the argument used to initialize it +To be specific the earlier patch the differences in this the testcases are in where i setup the staging area root directory with proper permissions so that job clients can create the +I just attached the log as have attached the code changes to have some debugging into +Looks good to me! Thanks Kim +It seems to switch very very quickly for me +I spent a little time seeing if I could fix the rendering of code blocks then gave up +There are some in need of completion and I dont see how the execution mode can ever be set in the driver +The error does look protobuf related +Ok I will do renaming and inline and commit this thanks +Should we make that case insensitivefinalStatus user queue and applicationTypes are better to be case insensitive as well +I agree we need a mbean for but it should not rely on Spring +no threading separate processes +The point of is to squeeze as much performance out of what youve got as possible +visible only on that nodeIf we get more detail on the requirements we may be able to come up with a temporary solution through configuration while we come up with the solution that makes this much easier and smoother +Yeah thats never happened and probably wont happen with our current ui framework +Updated patch fixes problems with reporter +What worries me more is that the generated code fails without indication of the cause and only on specific input xmlIve had a fair amount of trouble with nodeset variables and have got the feeling that there is more to this that just my simple case +In dept we have emp emp emp emp +Also added the from the patch for testing Nadav. +Thanks David youre right not sure how I missed that Maven complained about it at first +Integrated to and trunkThanks for the reviews. +Marking so we can get this reviewed and committed +I commented on +EdIt looks good and ready to get in +Inside this method cookies are already copied from those on session +The problem with on serialization of was fixed. +Removal of the class which was useless +If there continues to be a problem please reopen. +Needs furtherI am having a look at now as well as other open ticketsIf I come up with anything and have time to implement fixes I will assign them to myself +without raising an error +The testcase mentioned in the previous at r please verify +Is that dtd not listed in the xml catalog under preferences? Eclipse caches these and we add a good bunch in there even if we did it would never generate excessive traffic unless you are opening s of documents everyminuteunless of course we somehow have a bad piece of code that reads the xhtml without using that catalog +VJ is there any chance youll be able to work on this in the near future for the release? If not unless someone else wants to dig into the SSL tar pit we should move this to +So if its just wrapping the List api delegating read methods and blocking write methods it usually will be fine other than the extra object allocation for the wrapper which is less important +Added test in rWorksNice +Path committed in revision Thanks Katherine! +Updated to address review comments +OK then we can also remove the modules in Lucene completely! Lets just create a MB fileWe have modules to make this possible and let users start with a small installation without useless stuff they will never need +Patch solving the issueThe tar file needs to be located inside the Archivetestsdata directory +also change the XSLT to recognize that when the change user support is in the ds to not used the cached but to use the pooled +in +This will at least work for xml files since they have the ability for the author to specify the encoding being used +An further recovery attempts should line up and procede one at a time +Regarding the ensuring the node types in the underlying JCR are available in the simplest approach is to just require that the administrators load the node types in +And currently the selection is collapsed to the startOffset and expanded to the endOffset of sourceSelection this is right for selection +Simple fix by now calling the extra method for adding ssh key which before was done during creationApproved to go on beta since otherwise new users will be screwed +Committed fix to trunk +ah I found +When I tried Java it was complaing about a missing ordinal for +If you report this at Spring you are welcome to post a comment with a link to the reported me add a workaround in Camel to cater for that thrown exception + +I am positive I tested this before was released anyway it should now be working in both browsers +We should always use the numbers for compressed dataEstimate on the compression ratio is very usefulIt can be approximated based on the numbers for the data +Big support for Adams comment +Though maybe one of the others guys willIt has been something thats crossed my mind recently +you can iterate over that to check the validity of an API or do autocompletion like you did +Attach include change of the following area +Ive attached a new version +And inherits fromcxf +Verified on. +Workaround is to manual encode the Ill add this to be fixed +I think this is ready to land on trunk! just merged to trunk +I owe you a beerBest regards Pontus + +Test JSON data +Thanks +Sample to reproduce the should check that we never inject null within bind methods +I can always remove the usage in wsproxybasic +Closing this for now +Its the AS that has illegal duplicatesI actually believe it was you who introduced them to make some quick hack into ASThis is already fixed in the next version of Deployers +Duplicates. +Marking LATER pending course of action +jsvc binary seems fine now so closing this issue. +Im really not sure this is still a problem +does exactly what is required by the JSR specificationi am not sure whether accepting other ISO date formats in system view import would be legal according to the spec. +I dont think we want to do any of these things for +I may need a set of valid credentials for the sitein order to proceed +fixed +Man this is ugly now +Its my understanding that exporting from the default package is unsupported Im not sure if that is based on a spec or just Felix +Heads up were rethinking the name decision in +With patches to VM and test works fine for me +To test against a running AS instance start the AS uncomment the mentioned section in and run mvn install +Testcase attached +Use revision +Now I know the test cases exist +links to follow it manually either way +We appreciate and need your help on this +simple eclipse move and preserve history an svn mv followed by the class name change might be better +Added iffusedfs so this wont be triggered every compile +Patch to add the described may also help the user with the problem described in +it should list all the operations available in the wsdlportType +I attached an example application and posted to the mailing list +You know I am a novicewhat I need to do +Aha very clever! I had not thought of thatYes this looks as though it should work +If you do not want that behavour leave off the +Ivo I think this should readWhen nodes unexpectedly die during active transaction processing its possible for zero length transaction journal files to be created +However there are some exceptions being thrown which I need to look into +The attach patch should fix the issue by tweaking SSL context initialization in the classPlease reviewOleg +Yes +Thanks! +If you build the server with JDK therell be no problem +by Heath Borders +I just committed a patch to trunk that now correctly assigns BSNBV to extension bundle exports when attached to the system bundle +cu later +Please comment +do you have at least http hit for the deployment request +Committing shortly +thats when errors happen most oftenCurrently the information seems persistent +Perhaps someone else can take a look +I do not see a way how I can unlink these issues in JIRA +After more investigation Ive found that the file is the cause of the problemI think the file is still open somewhere in the code which prevents Windows to delete it +Work on this module just hasnt been a priority yet +Thanks a lot Suresh for taking a look on this +Hi all Ive noticed that the return section is hardcoded +move to wsm is mostly fixed with the work done to the trunkwsmsrc directory to refactor into the core and the axis. +If you increase your this wont happenA blocker would be something holding a release +Thanks a lot Dave! Yes this is ANSI SQL +Please get in touch with the repository vendor +Hi JBI believe already fixed itBest Regards correct Freeman +is now being registered as a Spring bean. +Thanks Thomas +Most users probably more familiar with usec too +merged some changes from trunk +Configuration archive and producer app +Was there a reason to have it in bothbq +You can close the pull requestI suggest that docs be submitted under this same Jira issue +the patch +I committed this +I have also added couple tests to langPlease find this change in the attached +Brilliant! Thanks for the explanation +about cant reproduce or I was and images indeed can be localized I was using an incorrect resource id the first time +I expect that we want to have option to add new sponsor to the default list as well rightShould the user be able to edit each description of sponsor even those which are available by default? I guess they will often prefer just to have logo without the text +Attaching a revised patch and an updated +Attached proposal specifications please post comments in the JIRA itself +Ok I confirmed that Batik Jar is the problem + for patch +Were relying on record equality +On the whole the performance of the compact accelerated protocol is on par with the binary accelerated protocol which I think is pretty awesome +I have two examplesxyz xyzconversation johndoexyzconversation johnxyz JohnDoexyzconversation johndoexyzconversation JohnDoeI have a workaround for this problem +Then do not define private methods +I apologize for opening a bug unnecessarily +It currently runs and passes on which is incorrect +I dont see socklent anywhere in +Assembly plugin will not be started because of Project child is duplicated in the reactorSuperparent points in modules section to subproject and +Again I dont think it is a scanner problem because in this case it doesnt scan at all the deploy dir +This works for me +Obviously this is due to serviceswhich was missing in previous versions of Hibernate Search +Thanks for your contributionI have added some checks to make the tests a little more effective at detecting regressions +truer words were never spoken +fails with or without this patch +For completeness attached reports the hung state after the blind fix was attempted +it seems dangerous read whole thing cause potentially it could be too big or rather big +It seems Sonar doesnt know how to manage repetitive sentence produced by Cucumber Scenarios +Thanks JacopoI asked this question because for the moment is only used in BSH with harcoded pathes +Committed revision +Note that this only affects interceptors bound using CDI bindings +Will write up about this in detail sometime +DarranWhats ETA when all of these WS fixes will be ported to the CP branchSome of the related JBWS issues have to reference to salesforce cases +I will cover the crash of the AM and others if counter size increases to this JIRA +Jira category and project url updated also +This is regression caused by +dims +Probably it should be the Source and not the as the in case of FTP doesnt know about the file on the FTP server +It now uses explicit checking over character set ranges to make sure it will now handle all radixes and an optional in front but will complain on any other character. +Best regards George +The weird thing is to get the OOME on a particular platform +The schema binding is ok +This was fixed a long time ago +is kosher for daemonizing +I think this is as designed +If needed Ill split the two partsThis patch contains code for adding logging on Bookkeeper +That is the best you see a label called Postnummer and an which has a attached to it +Marking closed. +are all from tomcat as well +to the latest fix +Multiply the score of a page during distribute with or +It seems to me that we have edits with the same timestamps in different WAL files this can only happen when the client explicitly set the timestamps +I plan to commit in a day or two and to branchI updated the back compat test to show the failure and alsoseparately added cases to the test +I had also run the tests as indicated while submitting this patch +Missing file has been committed. +Patch modifying the iPOJO plugin to be consistent with the maven bundle pluginThis patch replace the older +The previously attached patch was made against I just realized that I should have supplied a patch for the trunk +From Steve EbersoleI doubt it though I cannot say without digging into the code +The directories are created by mongodump I would prefer to leave them as they are +I committed the trunk patch to trunk and +Committed fix for javadoc svn commitSending Transmitting file dataCommitted revision +Thank you for the contribution Shreepadma! I have committed this to trunk +on both patches +Thanks Richard! +nicholas what is happening between and +which should be released soon +Attached patch fixes that +I thought that the test case didnt make much sense since it was passing both with and without the change +how to reproduce the error +I also cannot reproduce this decommissioning normally and during app execution both shut down the NM cleanly for meRamya please take a thread dump of the faulty NM when you run into this again +These must exist +Does it make sense for us to just change the name of the exception to then +Remove unwanted rpms in the configs +Please verify the fix I did +File +Thiswill be done for and do not have execute permissions in HEAD +this completely broke writing to CQL tables from Thrift for meThe name check is done against which seems to be populated with clustering key columns only +Small update the still fail on windows for some reason. +the issue is fixed. +Any news on this +When you change the kind generally the discovered renewer will be different +The code that worked fine in gets failed bein converted to M +I tested it against axisc samples and I too found the google sample is getting seg faults randomly all the other samples are working +I think if we can get Python going and make it easy to add Ruby well have satisfied of the potential users +I will replace with and post the results +Works for me +Mark can you check this out and see if it completely fixes the bug for you +What does java version say +If people come up with better suggestions then you consider them each case on its merit +Mumble +Cheers +Attaching the trivial patchThe only issue with this patch is that it is a change +Candidate for +But I also feel like the code will be harder to maintain because this check for split boundary has to happen in both slow and fast match casesOn the issue of test cases I managed to create a test case that demonstrates the bug that addresses but Im having a problem combining the test cases into a single file +Attaching my TiagoThis behavior is decidely weird but it is expected +Ok jacques I restart all process svn checkout apply pacth and load data Its works for meJacques can you see screenshot? I will improve message on the portletNicolas +Preference page and preference added and implemented so the preference is used when adding sequence screen is now out of sync because the sequence flow still gets the name in the create feature +Whats still todo for this jira? I guess it can be closed and well open some new tickets for future improvements +Fromall the information presented so far it sounds like Derbys mechanism is working correctlyProbably one of your two applications is accidentally accessing the wrong databaseunder some circumstancesCan you offer a reproduction of this behavior in a standalone application? Attach it tothis JIRA and perhaps somebody can help identify the problem +logs folder is generated at the root and all the logs are stored inside once you run the program +Otherwise we could have edits going in out of order which could be a prob +Please see my response to the same question by Aaron above +At most? But we need to decide of the behavior in Java EE we cant leave it hanging undefined. +thanks for the test Andrew +Also you shouldnt need to use exclusions if you add the Spring dependencies with the required Spring version to your own pom +I must definitely include info about this in the docsAdam +I will submit a new patch with the suggested changes +Thats no problem at all +Committed. +Heres i compiled under Added your jar to master and +Just to be clear I think youre absolutely right I would add that most of the people would want to ignore it +Cant easily test it directly so will wait for regression tests before closing this +Updated affectsfix versions as this wont get fixed in time for +Fixed in trunkAttached patch with null check to apply to branch +Thanks Thilina! Ill give tomorrows nightly build a try and will update the comments with the resultsI tried this again with the latest nightly build and it works for my test case +yes the problem is related to missing addremove files operation on the new addedremoved cf +Cleaning up old issues +BULK EDIT These issues are open to be picked up +Hudson is completely AWOL and this is a minor documentation patch so Ive committed it +They dont like to see red all over the radiator +closing the issue as duplicate of the newly created one in issuezilla +Completed it for Limbo +That way we can ensure that the code is only at one place and we dont have to worry about fixing the same code in several classesAlso the same thing could be apply to the way we are writing a logj event +Patch posted to has a test for this +Please see the archives for numerous discussions on the subject +So who needs to provide these artifactsAnd yes I agree about the comment on the Sonatype service +For example someone might want to turnoff ha and not provide failover proxy configurations but the logical should continue to workThis seems similar to your JIRA right? I dont disagree with you that it would be nice to generalize the concept of logical further but as it stands today it is not general this patch just works with the existing design rather than seeking to change itbq +There was a commit against this issue but it only updated snappy +Patch applied Thanks +Hello ErwanPatch attached and license accepted +This is the best approach as it lets the logic for executing the method stay in. +The code changes look OKI ran all the tests and got the following failuresFix logs for tests failed did you update the logs correctly failed testsMost of the above fixes may be log updatesPlease submit the new patch with the correct logsI will continue taking a look at the code changes though they look OK to me +Selecting Enable Nested Modules will always set Resolve Dependencies from Workspace project to true and Include dependencies and source folders from modules to this issue is better characterized as a feature request +We use the as the primary solution for allowing JBDS to deploy a datasource +But keep the Row terminology do you think? Or for consistency with and the XML stuff maybe use Position +Should we remove it +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +Its just a matter of refactoring the existing code to pull out the pieces that we need + +This test will show an Alert when you click the unless preview is called first +Would like to take another closer look at the follow up patches +AaronI made the changes to the set method such that it will set the default time zone as UTC in case JVM specific time zone is not validCan you review a new patch? Thanks +That is easy to do +Closing resolved issues for already released versions +Should the Pico continue for the lasting components in case of an exception? IMHO yes +Ill look at checking out the latest source as well and providing a true +I already have this working locally so I can submit a patch if desired +It adds a new property page for validators just like with transformers +Correction The C problem is to do with DEBUG +pls ignore my previous one +Thanks +It seems that patch contains another patch +Just realized you reviewed this Army without my having marked it Patch Available that was above and beyond the call of duty +Oops I just realized this was marked resolved +Any update on this Mark +I just committed this +Ahh I see now +It is not in the stream +I have a custom build of Cordova that fixes the issue but Id rather submit with if its done on time +Sure. +The will then be able to be used as a Filter with the new QueryThis also addresses the current problems with caching calculated distances and means that Spatial will work with per segment +Ill let you know how it works out +Also i would like toconf confNot sure how a rename patch is generated +First cut at a fix +log files in the eclipseconfiguration directory after running eclipse with an apparently installed groovy plugin +Thanks! +piggybank unit testsBUILD SUCCESSFULTotal time minutes seconds +So if I understand it well you want to move the old artifacts right +It is with Maven changed the patch to mavenget and also fixed some other unit test errors that had been introduced +batch transition resolvedfixed bugs to closedfixed +For the in r I have made the schema use a string value for the port attributes such that they can use property values even with validationValidation should still be optional though +In either case getting these stack traces is misleading IMO +It would be great to see explicit proximity operators using ordered +Ive got a further improvement to this improvement allows the override level to be set both for the servlet context init parameters and the context attributes +The reason this appears to be happening is that is a delayed component that has already been loaded so has been resgistered +This includes all life cycle methods and tree visiting not just decoding +throws a NPE +To the best of knowledge everything works fine there +Thats also the reason behind not cleaning all collected resolution information between each root module configuration resolution +Patch looks good Rini +update comments in patch file +This enables to process comments as wellAlso a test case in saaj added to check comments in soap message. +to branch +Perhaps we really should be synchronizing all relevant JCR methods against the current session to avoid breakage like this. +Ill create a patch later this day which fixes this for the trunk version +I dont think we can afford a common regression test for this issue please verify and close. +I described above +Id put it that way it should be possible to define the plugins classpath using the same concepts as the dependencies classpath +Committed revision Jeff Levitts patch that documents SYNONYM support +Verified on the branch that it fixes my problemIm not really familiar with this api hence my notes about +I just committed this +Closing issues that I resolved in already released versions. +I tried this patch and it looks good to me +HoriaI mailed the files to you via email as those are from my live project +Attached ssvm logs +If you preserve the error youll find that it is propagated to the original calling code +For a first goCertified description Features certified with Developer Studio have been tested by Red Hat and they will provide assistance on using these pluginsSelected description Selected features that is known to work with Developer Studio but Red Hat will NOT provide any support on these plugins +Im worried that depending upon a beta is too risky +bug of a bulk update all resolutions changed to done please review history to original resolution type +Well might be the version where a little bit of refactoring would not hurt +Part of the appeal of Juliens sketch was that it was hand drawn rather than a type font +pid +GavinWe are quite attached to the messages and tab functions so we would like to keep that functionality if possible? It just shouldnt mention jira messages +Hey BobI looked at the Maven RPM to figure out some I think I have a shorter explanation for why I built one myself first +RightHowever i also wonder if exposing someall of these properties via Message Headers would also provide for an extra convenience especially for the situations where one might not want to deal with Smack API this delegating it to a framework +So I thought you had done it +No the ticket should remain open +Im still thinking about just making the whole module depending on a SNAPSHOT and develop further +Fixed by. +Who knows what platform ui guys will fix next I just need to be sure that MY listener is the LAST one in that list +If moving the code out from the is not easy you may make things public + still contains the bugRather easy to fix Just insert one if condition doing an instanceof checkWould be fine when someone fixes the issue would be fixed within minutes +If case is New CSP should not change status period +This seems to have two fix versions +Addendum for trunk +Resolved in trunkUnfortunately the fix breaks binary compatibility so it wont make it into a release but will have to wait for a release +Maybe Safari hides combos that have been in a hidden div? +Covered in regressionArea +It sounds weird since they use almost the same flow +Fixed some ruby displayparsing problems found during manual testing +Committed to trunk and branchx +Line height in CSS JAVA roles for IE version of the CDK Development Guide was styles were added to Migration task will be encompassed by +So you get ngroups automaticallyAs for faceting +view would just expose the same configurations and group them along projects and poms +push to +For some reason this bug seems to be gone +So the server cant find the corresponding private key to decrypt the requestColmMy comment still applies +Committed to and trunk +added roberts improvement to I think this is ready I will commit with more refactoring of ForPfor decompression The decompressors take but still use the technique for now +Committed revision M assembliesM assembliesM assembliesM configsjavamailM configsM M modulesM modulesmailM modulesmailM modulesupgradesrcM modulesupgradesrcM modulesactivationM modulesaxisD modulesM etc +Yes +The problem is that println is a method on Object and so the println method from Closure is called then +Sorry about that it compiled successfully on my machine +Though I cant give you any estimate on this because most of the work is actually in the Maven ebedder component +Hi allThe latest Camel unit ESB integration testing has passed +was added to the manifest by so perhaps we can mark this issue as fixed +EXPUNGE seems to be behaving as per SPEC and as per Dovecot +Interesting we havent thought about that +type safety is there for the users of API even now +The updates were fairly minor +Ok thanks guys +I have made two suggestions for fixing this +would backport to We should get consistent behavior from after goes in +interface that could be used for conversion as well as to define partitioning +Default buttons arent implemented and the optional component menu would need some work so that you could navigate thru it with the keyboard +I thought we can use sorted import files +Its probably explain the behavior that you seeI think it makes sense to do it on as the configuration could be different and the features configuration should be the master +The patch replaces the older pathc and adds a JSP sample that uses an SCA reference +From my testing yes it pulls the exact version value set in the Jar manifest forWhat do you think of my +I think we need to add a changes entry too since MP now needs to impl clone in some cases +push out to back at JPL and just never closed issue +This bug has been marked critical since February any chance it might get looked at soon +I removed info +packagegenerated code doesnt pass rtt to its arguments +If I found anything interesting Ill comment backcheersjuan pablo +Thanks +Coming in another JIRA soon +Uses shale static remoting instead of the resource filter +Not sure if voting still matters +Sorry +Looks like the initial fix for caused this +Originally we noticed issues deploying snapshots from the Mojo project +In other words all the mentioned in the manifest only JAR must be in the same directory or in a subdirectory from where the manifest only JAR is +I have not been able to trck this bug down yet +all http related headersThe above mentioned headers are not required to be used by the was done in +are there any installed JIRA plugins for displaying code +Allowing the parser to treat a compound statement as a valid user command +We should look into this +So you would have threads using different versions of the object +Actually perhaps we dont need to encode the workspace part of the style reference just be able to accept it when parsing a representation being posted or put +Ill run some tests when were ready +This jira is for tracking the above community change in productStrong plesae feel free to change the fix version as desired as this is only missing feature +I see some later patches from Lily but I believe that she abandoned them +Fixed on trunk in revision and on in revision by the way I used a wrong rev number in the comment for the commit +The programming model and configuration should be entirely Java driven there should be no XML configuration +patch committed thanks santhosh +Hopefully others raise their voices on this too I really love the toString implementation +It seems like we can get rid of that check if stale connections are stopping the starting of jboss and let jboss handle that itselfRegarding checking of the process running there are times that a process is running but pid file went missing so we also check if theres a process running using the script that start these servicesThe stopping of the service is doing what you describedThe same check as in EAP was done when starting jboss until there is better way to determine if service is started thats what we can do for now +Committed to trunk +Some explanation on output of first n sample +Codes look good + is already ready to go and should be released this morning +Andre what does More Work involve in this case to resolve it using the EMFEclipse XSD way of doing things? I agree that if theres a better hack we can put in to resolve this we should +This patch is just for checking the approachI will include the tests in next version of patch +Not everyone does +Send a message +Making use of CQL now and it is working very well. +push out to +FYI Stuart is adding joda for rest easy +del files +PR accepted and issue closed committed to maintenance +I can see major problems with sites that set quotasIt seems to me that quotas should include temporary space used during job submission +Actually no +the patch looks good to the patch works and avoids this issue I think the correct fix is to make to NOT send this event multiple times +Is there any progress on this Vijay? To fix this we need a language design decision and changes to accept Throws annotations at async and at body +I have a few minor comments on the patch that I intend to incorporate in the next version Move the sections around a little bit +In +I still get this deadlock maybe once per weeks or slightly less +Hi Nicholas Can you please take a lookThanks looks good +Thankscommitteda test case and binaryprotocol fix would be great +So here it is +Merged Petrs onto the master and branches +Its there since the very beginning of +This is a dupe and Im not going to redesign this in Seam +field date jpa class +Import Teiid Designer Metamatrix Projects +Fixed in trunk the Stomp Messages now throw a for any access to a property that isnt set +Tried again and this time I get this error image +Is there any plan on backporting to andor +This issue perhaps throws up some points for discussion more then providing a patch of interest for al AD users +You would need a better cache API so that you could compute a key from the client request which would have to be custom logic in a plugin +Moving to CR +same problem same fix +To make it clear A datastore supporting the hint GENERALIZATIONDISTANCE will try to return pre generalized geometries but does not generalize geometries itself +From what I remember krb vs krbi was like perf degradation +We have improved test coverage quite a lot over the last release and start fixing each bug with a reproducing test case to prevent regressions +You might need to install have made some progress with this but am now hitting another bug +No backwards compat issues so moving to +Yes and thats the whole idea +Binding null values isnt the issue problem is that like return from their Text property +bq +I cant replicate this problem even with the mailreader application to which Ted is referring +The main reason for this is without this restart adb looses the emulators often +Thanks Hemanth +Maybe leave a comment over thereThis issue is a straight up bug +I suggest renaming the option to deleteGone to make it more obvious what its supposed to do +present on all parameters a mapping is created for the constructormethod on some parameters an is thrown for the constructormethod on none of the parameters no mapping is created +Now that work is well underway with AS all previous community releases are +If its any comfort Geronimo has way more embedded jar files than Kafka +Help is updated though the example section was pretty clear already. +I fixed the problem and created an additional test +I think one of the Apache mirrors has a bad file +This issue has been resolved for the next release +Thrift Avro and protobuf all have similar issues filed. +While I think that help links are a good idea it looks to me like each and every element gets a help link even if there is no help link for this element +Another factor is that we dont have much expertise in the team so the quickest way to get this fixed is to provide a patch yourself +Assigned to Remy to take a look at documenting Web for AS +are you sure youre using the same settings in both cases +gitignore file explaining its purpose +Any news guys? Im looking forward to some feedback if possibleThanks again +I dont think this is a bug +lacks synchronized keyword on a few +DIGY +part is fixed part still done now too +the crux of the issue is that as the excludes detection in the doesnt map out to an xml configuration the way to get around it was to use and include of and an exclude of which is interesting in that the seem to mean to different things in both includes and should be against API but I dont rights to tweak. +Reported as +bah +Then we can apply this patch +This will make it visible to +for making any configuration tags case insensitive +Another thing that is really aggravating is that I am unable to set a custom path for eclipse I have set a classpath variable MREPO to Cwebappsbut it has no meclipse plugin does not change if I specify a user settings file CDocuments and SettingsOwner +PR is pushed into master +Kay +Mind this the next time youreport a bug to other projects +Can see two instances waiting forever to get a resource from the is a duplicate of. +Running some initial tests on this patch as well will post it up for review soon +Added set event changed lists to follow pattern recommended by David Jones +Thanks! +We do not use Weld +I general the parameterless constructor should not redirect to the properties constructor to keep compatibility +I meant to say a servlet within a jar +Im going to reopen this issue because theres an extremely similar problem when new rows are addedTry clicking Add a Row in the table component testAdded on callback to call it from when a row is inserted so far just using it to resize sizer bars +It says tests slicing against data from one row with expiring column in a memtable and then flushed to an sstable but it is forcing flush before slicing +Pig will not be able to check that but it is no different from being able to check if the data is sorted or not +While I certainly see the value of this it is unfortunately going to be pretty complex to the plugin is really only using rpmbuild to assemble the rpm +I switched to a in rev and following the docs on the filter to aleaving this issue open for feedback +Seen from jenkins is accessed as fast as a mounted folder since is probably physically in the same room than jenkins +I attach a patch file which I revised. +Volker +Refactor the jbpm requirement builder to include drools and teiid so that it can be easily merged into the installer using reporunnable. +defer all issues to. +I am attaching a patch which moves as part of core +Here are three variations on the JCA UI a generic resource adapter variant one for queue references and one for topic referencesWhat do you think +Merged Ondrej and my changes and sent pull request +Tabbed pane with too many tabs in a with two tab rows +The version is already better looking and if people have further improvements they should submit patches against smaller issue reports +poke +It can be accessed via httplocalhostguessvAs you can see context root for root WAR is guessv second is full file name as subdeploymentRejecting as this is not a valid issue. +Open this jira to track my patch +ad nauseum +See the ticket description for more details regarding Fix Versions +That needs to be fixed +You can also set VERSIONS in the column family descriptor +The final patch will have to update those targets as well so that they pick up all the ec files and not just +Unless there are people running security managers in production for whom this is a problemYou can move the JIRA by looking under more actions +But +Nice jobThe patch looks goodIve verified mvn clean installShip it now +Changed the priority to Minor +Added in revision Thank you for providing the extractor class. +Updated both the caching guide and the reference lookup function docs to use the same example. +let me know if that works +Martin I think it is planned even after git HEAD refsheadsmaster some users are still seeing trunk as the default branch after a clone +Im trying to implement this change only of for now because there are some which extends +The release note looks good and it passes the s checks +migrations where you need to drain all the queues in one system and move it into the new oneMy concern with would be how to handle latency from looping over s of endpoints +willem +That actually amusingly shows all pools in which you arent a member unless Im mistaken +Note its the client doing the connections it can use reliable to reconnect N times etc we dont need reliable on the broker side +My guess is that this is something to do with how we manage the +AllenAny chance of bringing this up to date +except has copyright year and company name is by Redhat +We already try to connect straight away in connect and sleep only if the initial attempt is unsuccessful +Pull request has been merged into master and the branch +Adding to affects versions +Thats a good thing +I suspect a message might be getting stuck in OTC and causing thisAs far as the FD goes we definitely remove it there in so something must be adding it back +Looks good. +Uh make that not +Before this we had snaccj based code +Updations to References sectionThe wizard that comes up when Add or Edit button in References section is clicked is updated to support multiple elements and utmost one pattern element +Something very strange is going on there +Here is a new patch +Ill update the wiki +HelloPlease try the patch +Martin I could not agree more +It makes no difference to me whether you want history or you want history + +This is incredibly helpful +not a third step but part of the actual release process +Awesome thank you! +Fixed by Michael Bouschen +version of slfj that caused the problem as mentioned in the logbecause of the fact that the are not needed by the and the mentioned dependencies add of dependencies my decision was to simple remove itI will mark this as fixed + +I committed this to trunk and. +The original implementation is +JAMES use avalon +Patch for branch +Along with tests please updated the ls documentation in srcdocumentationcontentxdocs +Charlie executed this on the master branch + +Otherwise the same +This makes one change in lucene that I could break out into its own issue but its rather small so I will collect any opinions here for now I take the ensureOpen check out of on getDirectory +Dont know of a good workaround at the moment +It does require a serial reduce phase but it is quite small since it involves the bitmaps from all of the Bloom filters built in each map phase +I am not sure what was causing the problems with this before but I hope it is fixed nowSorry for any inconvenence caused +I was referring to in the above comment +But I do not think the solution is to deliver MIME only when the request is MIME +I am quite sure that I fixed this in +Would it be more appropriate to say +new patchaccount for both stdout and stderr +Is this for M or M with the controller? We should set controller component versionIs this for M or M with the controllerIts for M but I was uncertain at the time of creating this +new images and images it is a really big number of resources to allocate for and store for a long Eclipse run session to get No more handlers exception +Tested and confirmed to work +As these are test only fixes we should still try to get them committed if we get a chance +The idea behind was to provide documentation that could be simply downloaded and zipped together with the release distribution +Updated patch to reflect latest changes from edismax true means that we will attempt to treat something like and as the operator ANDIf lowercaseOperatorsfalse and is treated like any other wordPerhaps the difference between and is text analysis of the title field? and is an English stopword. +My patch uses the underlying methods in all of which were already public +I can confirm that it works when installing smx to root c +I will have to customize the build for RHX team because they are only interested in Issue Tracker beta code +Will retry with +Now that work is underway with AS all previous community releases are +Holger have you tried using or +Cant reproduce this with JBT +I will fix it thanks! +Applied patch to resolve defect. +I can see how this can be annoying +So B is the way +You can already do this +Will attach patch here for as used in SMX +Also included are two log files showing successful behaviour under Hibernate and failure under Hibernate interface changed. + +I have verified this issue yesterday strace shows that sockets are being reused nowYou may close the issue for sure +Included in Extras siteIf theres missing functionality let me know +Ah good catch thanks +I have addressed your comments +I tried selecting build queues in IE and Safari then Save and both seems to work fine as well +What do we pass in? A +Adding release note +this should be in as with fresh builds is in. +Backing devices with QCOW will also break on the longer runPlease do not do thatWido +so dont be surprised if this is already done +that way we ensure it is after all other key for that table +Comments? I didnt delve too much into this code +Moved from to just after merge of old branch into TRUNK +Ill attach a fix shortly +Patch committed to branch + +Once some of the classes has public restrictions I am sending them directly to Michael +I am posting it for comments +bulk defer of items from to +I believe the tutorial is already shipped in the solr download +If VR and VPC are displayed on the diff pages in the UI then the flag should remain +If we want to create a single task for to then we need to rename it to something else and recreate to and then move the old one to be a sub directory inside the new to +I removed native from hbase for now +Another possible issue could be that you were running multiple sources pointing to the same spool directory +Please check that it was applied as expected +So a daemon thread along the lines of replication factor may be a simpler solution +Ive looked through this example and I am not sure about the following + +Would this be like a high performance option for a very limited set of RPC methodsNot sure about high but higher than what we have +How about adding this to the error message The cluster is most likely running out of space +cannot repro on anymore +I will commit it to. +This happens only for Oracle priorities for jiras were looking forward to actually fix in in gt +has just this fix +Would be great if could verify that it is working +Have filed a bug with GWT on this cant change the behaviour at this time. +Please review the patch +The new description can be Clone of XXXX description +I havent run all of the unit tests yet but I did run some of the tests as well as some of the tests and they pass +Ill pass this to all places where we invoke Python +Also changes are made to to expose received tokens on allocate call +Code looks goodThanks DhrubaThe latest patch changes the status reply for OPWRITEBLOCK to be a short instead of byte to be consistent with OPREADBLOCK +I think this happens automatically when the streams associated with the process are closed +Page release of the Architecture Platform includes Smooks +Please let me know if you see some issue after this improvementThanksMarco +patch I will commit shortly added name to description +Forgot to say that I like the idea +Do we need to add a function variance? or we need to move COV and COR? thoughts + +I guess its easy enough to catch that case in and set the defaults explicitly in this case +But the figure will be recalculated next time the diagram is opened +Let me know if you want me to do that +Done for the documentation contacting the repo to much problem see +Especially since I wont be addingremoving nature often +The only option if the exception thrown in case of remote execution has to be a sub type of is to use itselfChange to to capture the causeConvert to in case of remote invocation +The wiki will be nice when its ready but now we have all the essentials +This issue is related to problems we had recently with still open port on vmbuild machine +The shaded jar is only provided for convenience. +closed with the release of at for +Directives with no text following swallow the newline +Additionally in all cases we can move the corruption check to where we read the metadata +RobbieAlex what do you both think +Ive committed this to trunk +The changes arent committed to this tag +You can apply the attached patch to see how it looks + does implements possibly a class loading has been dropped in Id be curious to see if this bug reproduces +They may need to exclude other Resteasy artifacts when including as a WAR dependency +workaround for this issue is to define the relationships on the inheriting class level not the base entity +The first was that allowing collectors to be specified as a query parameter was picking low level components for Solr to use +Part solved in but the streams aspect needs further work so bumping to +Ill get it done asap +I have attached a patch here that adds this new functionality +AgreedWe can integrate once patch for is ready +I just committed this +More info pleaseWhere are the updated files that pear will use to mark the version +seems to be caused by the default argument for db +The attached test passes with. +Please forward all requirements and such through Mike Amburn hes putting the requirements together for us and getting us to the point of actual jira tasking for us to develop off! is the duplicate issueThe problem I am having with email is that I am then unable to track its progress like with a jira issue. +Yes I think the comparator is in fact correct +Patch cancelled +Ill create a new issue for this piece +Nick has taken care of this. +The only exports from commons log should be logging +When installing the security manager it complained of a Malformed URL parsing the policy file if the property was set to an empty String +ul li and +Its now failing because resource is missing +Otis I think this was actually comitted to Lucene but the solution isnt quite appropriate for Nutch which does not use Lucenes but instead has its own leaner RPC mechanism +or eg the common suffixes freeze early on and then lots of future nodes refer to themNot really that surprising you encode common suffixes after all so most of them will appear in a properly sized sample +So unless someone else steps forward to fix this Im going to defer this to when hopefully we can move to Maven +Please add the patch as an attachment to this bug report +It is supposed to be slow and not very general but for our limited purposes that is probably fine +Committed thanks +The priority is raised +I will submit a patch which throws an exception saying wild card not allowed if a wild card character is passed +Committed as part of. + +The jsr name changes have been ported back along other fixes management codebases are more or less alike +Thanks so much +Im gonna take a look at this the earlier possible. +I would need for you to apply an LGPL license to your contribution + will address incorporating into written by JT +Thanks Swati for contributing! +All patches are attached +Wdyt +Of course I can expect that +My idea was that would implement this new session context interface +Updated the patchTheres a slight complication with the handling of the warnings in that val passed to those functions in is the raw string which might not be the same as the val put into paramval +this bugfix released in version +Permissions are a security feature +If someone have the code please feel free to work on it +added in rFor the verbose field i dont think it needs to trigger a reload when you click it +Closing all resolved tickets from or older +The patch can compile using +See the class where you will find the method I would like to propose +Btw the libhdfs test isnt new same one thats been checked in for a while +I suggest to implement Algorithm because it is closer to the current implementation + +Thanks Ivo pull requests have been appliedHavent resolved this issue as not sure if you wanted it left open until the quickstart has been done? Or would it be better to close this one and create a new jira +I will commit a quick fix for that shortly +Actually I just came to realize that the problem is not related to how you wrote the test case but related to the fact that marshalling the object we just unmarshalled in the test case causes an NPE +Duplicate of +Although I can probably get the state of the context and handle that conditionI need Remys opinion though on whether a failed servlet should error deployment +Hi ChristophDo you think there is still some points to address in this JIRAThanksEmmanuel +Since we cant easily detect suspension initialize after every readline +What version of Woodstox are you using +Since TRACE is disabled by default this is now a non issue. +Closed all tickets +Ok The way this currently works is that you would have one pom that would list dependencies on etc so this will generate your file Also if you wish it will indicate what other bundles it requires for classes that you may have not depended on and you can supply this list in a properties file so it will include dependent bundles for you as well +Seen this again upgrading priority +It reproducible for me too +Attached a file to that will fix this issue aswell +Please have a look and would like to know the approach is correctHi everyoneStill the package structure and bookie process dependency is open point +I have checked it in at subversion revision +removed all the remaining nocommits as I think all of them had been addressed to + as Im not satisfied with the implemented solution +So if you have additional comments please comment there +Done +Please use the various mailing lists for that stuff. +I think it is enough to add it for internal functions +I cannot reproduce the bug on and trunk +Are there other Java updates we should do +Maybe try to whittle down the Keynote example? Ideally itd be a minimal test case showing the issue +libs +Juca Can you write up a couple of lines explaining that this is not supported for can then request the docs team to include it in the Database Compatibility section +Sorry for the false alarm +Although I suppose it would work for that too +There was insufficient locking when directories were being created or renamed +See file or for details + was specific issues for my test environment when I fix it I find what currently there are some performance problems and is just trial of this + for the patch +The behavior of the convention plugin changes slightly with this fix applied +I just tested it again with no profile specified and it seems to be ok +The native splash screen is displayed at the correct location but the imageView in which the plugin uses is displayed at because its parent view is displayed thereTo reproduce just use bincreate and run the project on any phone simulator +I am resolving this as out of date +Ive attached a standalone SSCCE using Hibernate HSQLDB and Ant +Rajesh could you reassign this issue to some EAP QE? Or just close it as wont fix if you dont mind EAP logging this annoying INFO +Itll be better if we can have a section on ADB MTOM data binding too +like it is inside think the request locator is ours is it possible it is in there +Initial tests were OK +Axis Standard wsdl located in modulesxmlbeans are correct +This approach assures that file names will be unique and still be short enough for modern file systems +on +Thoughts +Needs some of the comments to be fixed +Should UNKNOWNSERVERNAME be a define up in +You could add tag to of the corresponding package to exclude the package from the +I have modified copy of in plugin which is registered in WTP +Can you point me to some info on how to submit a patch +Hi Bartek that rather suggests that I have made a mistake Thanks for spotting this Ill reopen this issue and look at it first thing tomorrow morning +Processing the processlog means that the information will be integrated in the history relation to specific activity information still needs to be worked out +Also the property in the filemgr has been changed to and checked mistake sorry about that +Undeploying did not remove the files from the deploy directory +Alex check this issue +Committed revision +Francois I mean you had deleted the VM created by autoscale policy from cloudstack UI +I have no other ideas to try +thanks! patch applied to branch this issue to rollback also in the branch the synchronization problem +Thanks Sanjay Luke! +Gary I need some additional reproduction steps here +Thanks Ed and Guy! +Due to +You should fix your tests though. +The patch doesnt update any counters for the cleanup task +Igor thanks for the ve corrected this in the templates and it should work now +implementing the main service in Java is interesting but the code would be simpler and more concise written in Groovy and you could probably even do this in aBTW for the services try the thingy +Ah sorry I forgot to included your test case in +for whatever reason the automated emails are just now being delivered +The doap descriptor is only available on the we have one on branch as well +It wont be included in the dependencies list of theAnyway commited a fix for tunckr to remove the included module before writing to +The rationale for targeting JDK is simply that we havent run into anything that prevents us from doing that +Ive verified this in but I dont really know what the solution is going to look like +Patch to create a new context for action and render requests according to +resubmitting the patch +Hadoop is about sharing a cluster with other usersPlease dont assume everyone uses Hadoop the same way +OK +What about CLIENT and INCONTAINERClient as in your just like a client will be +Will see if I can reproduce with that if not will ask for more info +This test is not intended as a junit test +yes we should get this into the branch +can we just make the executorservice arg mandatory for parallelin my opinion whoever creates it should be responsible for shutting it down no one else +Also updated test cases +Thanks Devaraj! +Id still like to revisit how to clean this up at some point but I think what is in CVS is sufficient for now +Ill look at that test suite +Use with template Recursive template to be used with test to reproduce this bug +Work as expected in trunk +Adding a list of if +Committed a first version in revision Apart from the mentioned changes I rewrote the jcr classloader and removed all unused stuff as there seem to be some problems with marking the class loader as dirty in some casesThe new version is clearer and should do the seems to break one of our test cases Ill have to rewrote the classloader to use a session on demand instead of keeping a session open for the whole time +I locally and all is well +I want to look at the current difference in behavior as there was already a change made for this audit testcase as mentioned in the forum thread +And by the way the indentation in the whole should be fixed but I opted for not doing this in this patch to keep it simple and to a minimum +I have requested to get it changed on the server side and will let you know as soon as Ive had a chance to try it +Latest build just ran into test to I think +It appears that such a feature should provide support more at rather than at runtime +With that behavior we may have two writers writing the same file namepath +Im going to take a look this is the updated version of my Open Search implementationThanks +batch transition resolvedfixed bugs to closedfixed +Ok thanks for the quick are going to be able to go to Spring for one part of our codebase +We already have unit tests in existence that verify accurate stacks will be returned even when reading arbitrary urls via a subclass to + assuming youve tested it Nicolas +Test results will be recorded once the test runs complete +Thisis definitely a bug and should be addressed +Created to switch to a released version of Hadoop . +This would work at the cost of startup time which is already too slow so it is unlikely that we implement that solution in the near future +But it did not work as I have said earlier +Good point +Thanks for bringing this up +Ahhh +Thanks Elliott +What is the known kludge to get the browser to save the validator javascript +Hm ok if we set the default to true then we might also just use the currently available propertiesName attribute which can be used to give the variable a differentname than properties +Provide a reasonable dont care or dont know defaultSome small changes to binhbase are necessary to support separate Xmx options for various daemon typesShould we do more +Why was PTG added to the issue summary? I am not familiar with this abbreviation and I am linking to this Jira in a report and this is adding unnecessary confusion +The we use in the collectors is not necessary because we can lookup the norm from the +Please report other plugin related issues to the appropriate issue trackers +Committed thanks +Isnt this already logged in +On another faster machine it reproduced each time + xsl files were borrowed from the HUTiterative project So it will OK to use common place for this sort of files +bq +i dont like this idea but you might be able to handle this is a Filter or by subclassing will not support case insensitivity out of the box +Anyway I am uploading a new version with the changes you mentionedI dont have an account yet +RSC goes down +Russell when attaching a patch can you click on the option box with the labelGrant license to ASF for inclusion in ASF works Assuming that you want to include your bug fixes to Solr +Its impossible to guess whats going wrong from the limited description youve given +Frankly I have not tested war in a within pluginManagement only in this build section. +Im afraid Im it for as far as resources for working on GWC go +The changes are not yet verified +Some queues work as expected but for some of the other queues nothing gets into them +Good point adding the timeouts +After compaction this reduced to a more expected GB +Pradeep did you get a chance to look in to this +Reading from the Chapter +Hmmm I wanted to do the minimum we could to get this working but I take your point +How do we handle multi arg aggregation functions? We dont be fair with the current API there is no way to express a call anyways we have a fixed set of methods so maybe were thinking too much +thanks! forgot about this one was about time to close it indeed. +The behavior is actually non deterministic. +Same root problem as +Given these two observations it makes more sense to me to keep it separated +META +Done added new screenshots retaken old ones updated text context on the base of RC. +Whats going on? Can we do something to push this along? This is really frustrating +By the way also Apache TIKA uses this mechanism for plugging in the list of document parsers +This has been done now as part of my cleanup. +A good number of configurations in have been copied from and hence not necessary +Fix looks great Colin thank you! And if you want to achieve a slightly higher measure of fame you might submit a patch to +I ran into this while fixing Pig unit tests. +We could extend with our own implementation +Here is the Patch for Payment Portlet +Workaround available +See trunksite for the aforementioned site builders +It deploys jars with JDI implementation to jdklib directory and provides simple JDT stub implementation to resolve dependency on the missed Eclipse stuff +If running unattended and no one approves it might eventually time out and fail +fieldName status type +You can avoid the escaping issues by using and know that this form works +Roller should support syndicationoutput of weblogs in RSS and possibly a couple of other formats without requiring each user to add any templates +There is also +This patch simply turns the condition where the admin user does not exist into a fatal error on startup exiting trafficcop and logging to COPFATAL +It got rejected for not being up to snuff +Moving items not being worked on afaik out of M Fix are now using commons configuration + patch looks good +I just committed this +Thats totally wrong +I agree with the implementation for the get issue that Steve proposed +I just fired up the example +Should I fix it alsoIf not to set will be should I do some PR about my work and we continue the discussion on source code lines +Thanks +I assume this is fixed now? Andreas can we resolve this? +Closing issue Oozie is released +Pleae create a new jira for unit tests +Verified using msvc release build. +Marking as resolved based on the OK from the IPTC for our planned use +Therefore no more threadlocals and fussing about +tcunninglocalhostrosetta svn commitSending rosettasrcorgjbosssoaesblistenersgatewayhttp file data +patch looks good +I see that it just did one round and created Gb sstable from Mb onesThanks a lot for the patch! I will revert the compaction settings to give it enough resources and let it complete its job to see the end results before I restart the test traffic. +applied thanks Rob +Over time it got worse until a restart and then it was fine +We already have perfectly good conflict resolution without itbq +Created for this I dont fully understand your understand client and server connection detection mechanisms are independent I never thought otherwise +thanks Eric for the patch +It seems to solve the problem on my end. +Oh gosh in my case the distant repos disk was full +Ill plan to start another large cluster next week and can test then +We have the same issues described above with version +Retain change in code instead of interface +Closing issues +already has get and get methods +Then forgotten and got out of sync with master +you still have the problem search your local maven repository and nuke the plugins snapshot folder +fixed the hashCode in note that does not have this problem +Checkstyle reports inthe master pom see to see how to generate +Thanks +Removed from the fixed versions as the patch for the trunk is still notcommitted +Either we fix the wrappers or we remove them +Are there any +The patch seems to have solved the problem +uses by default now since that is all it needs and has a set method to specify any with a corresponding get accessor +I would not like to add any features to the but it would be ok to include this into +You should be able to verify in though +attempts to use a heap in C +Thanks Ben +I am hitting the same issue with +accepted into at s request +You rock for making my life easier by wrapping this all up +Please check it applied as expected. +Let me think of other options +So far it is not commited but Ive introduced following rolesetc +Mine just needs to buffer up everything before the parse occursI was just wondering if you know how not to buffer everything and filter the stream in my code I appreciate it very muchIll try using yours +Further a loader may also provide a guarantee that all instances of a key appear together in one split +Resolving this as Fixed +See section have reflected namespace support +That said here is my opinion which I strongly holdThese files must not be put into specs +So that shouldnt have anything to do with this +Its not a SWING bu rather AWT attached is a patch that fixes the attached is a patch to the I apply the test patch without the fix it passes +wsmqrouter +Both are supported on Mac but the bit version lacks support for VPE EditorSo it would be nice if in such cases the user could choose which of those two to install +in filesDont really object is you want to do this and feel it is really important but trying to close out. +This feature will simplify repository usage with custom automation scripts or tools like Chef +I wondered about that +Ming I did not see your latest comment +Yes Ian should certainly get an account +we are hijacking the original issue here +On the same boat +This zip does not have a dependency on another project. +I wouldnt really say this is a duplicateEnding tags with instead of is something many standard HTML generators do and perhaps should be done regardless of any XHTML option. +Seems incomplete to me +Are there other changes to the default Similarity that may be of general utility? Or do folks thinks its better to leave this alone? +Incorporated review comments +without eny character encoding set +My apologies. +Committed fix to trunk with revision with +You didnt swear that works OK +David can you take a look at the issue that Shawn opened? Thanks. +If you do not have any specific reasons to use du command wo path argument then we can close this issue as invalid +addresses the test nitCommitted to trunk with revision +Vincent the fix is not on trunk +its just a bit slow and will move to OPENED in a little while +Fixed in trunk +Uploaded patch with comments from Roman +Bulk move +bq +Updating patch to reflect API changes at the level see for good so far +revision for trunk revision for branch +Though you think the programfailed just after processing entries behind the scenes the earlier opened cursors are still receiving entries note that your program has opened cursors and all are left open +The reason is that it still has some unprocessed bits left in the byte that contained EOI code if code size is less than bits +Undo the changes on the build file to skip queryparser module if jdk was found +I simply ran servicemix with default configuration and manually deployed into hotdeploy folder +Ill just remove all the files from the +Also throwing on set is a regression from and may cause applications to break since previously this method did not throw the exception +Please see discussion on xcore mailing list +thanks +Hey JasonI saw the code you sent me but I dont remember exactly how you implemented it +LGTM +I would recommend you to investigate the issue more deeply and find out why your container doesnt want to grab JSF annotation this might be serious issue in implementation +It was fixed in but I suggest you try the latest release +Defer all unresolved issues from to +FYI I have installed +removed javadoc plugin in main build profileupdated to javadoc exclude generating javadoc on internal api packagesmove javadoc generate to deploy profile so just run mvn clean site deploy Pdeploycommitted in old documentation in docsmoved javadoc generate to regular package build profile doesnt take that longcommitted in +It causes the exchange to fail +Thanks Nicholas! +I havent replicated this error though +bq +but then I dont think you would need the Maven Ant tasks +in +In the meanwhile +Corrected a typo in main stylesheet file +OK help me understand then pathto changeitgives a working server but pathto userPassword changeitgives Cannot start the server ERR Failed to initialize What am I doing wrong +It fails on Ruby since the Ruby implementation doesnt yet support deflate Ive created for that +Since I do not know what kind of summary data you had in mind I am sending you a of the classes overview +Can you make the such that it fails without the code changes and passes with? I dont see that happening +Yes wait is a believe that wtps validators have a problem when editing or opening a also think that the same problem is in +May be only doc change itself should be okAnyway on the patch +I like it. +I went through weakreadRemote and strongRead they are like this as well +Ill robert +There are two main differences from Pavels latest patch +in for a penny +It depends on how the paths are used in the code +Ive checked in a patch that improves my previous patch in please give this a for everyones feedback and patience +But now they reverted their change in so theres no need to update it in sqoopI should have thought about in the first place but I was under impression that Hadoop is moving to all together +bug brought forward to visibility +I will create a document containing the steps to be followed +looks good +Also possibly add error codes to these exceptions as well as to what is sent over the wire +Fixed in. +For one it requires ruby being installed on systems +We will keep running this with new builds +Resolved for a while +Theres a much better chance of getting discussion on this issue if you post a message to the dev mailing list rather than adding comments to the issue +How did you configure the mapper +Updated with Dougs comments +waits for all responses even if the first one is all we needThis may be useful in the reverse too +So now the potential is wrapped by an consolidating the kinds of exceptions that can happen from an execute down to an and an +can you explainFinally Im not clear on the legal status of those Stream classes so Im not clear if we can accept them +Ill commit this once and are in +Its quite natural that a Servlet integration has a slightly different hook than a Portal integration +This is the right patch +Ill commit your patch shortly +Later we can make it part of new whatever the new live config will look likeGood on you Alex. +verfied at r thanks a lot +lets fix that for +Is there anything special to do to see that these containers are not emptied? +for A C could be an alternative but A should be favored +We should at least fix the NPE and provide a more useful error message +Again? I never realized it was locked previously +Note The attached are out of date as they are for an old build. +Please verify and close. +Needs testing against the TCK for confirmation +try attached quickstart to reproduce the final reference to the session? that isnt the best thing to do. +Hmm the existing searcher can be used between resetCache and commit in this handler +Patch line Thanks for reviewing the patch and catching this +I will try the solution as soon as i have access for osx cocoa +Hi It seems like this sync process does not work +On the contrary the problem with the framework assuming that the user is in the servers locale and timezone has been discussed at length and I believe the consensus is that the problem does indeed exist +The third implementation explicitly check argument type while Mojarra doesn added to the script that converts tring value to DOM element reference. +only whitespace differences fix indentation in completeCompactionOtherwise +This matrix does not cover interface +Arjan Steve is not asking for anything extraordinary +This issue is the result of another bug +So this applies to and above only right +I will put another comment following my uploading of the new checkstyle friendly patch +Yes if the user upgraded from that will fix the trigger problem because they will get invalidated during the upgrade +This worries me this tells me that has a complete inconsistent set of jars +the patch looks good to me +In effect these are filesystem checkpoints +Fixed in trunkParameter Annotations were not moved to the interceptor method during the annotation processing +It is working fine for me using the latest svn trunk version of the +didnt get adrians comment until after this patch +Applied the patch with thanks to James I also fix the test errors of caused by this change +As promised Ive attached theLots of credits to Thijs Vonk and Antony Stubbs of course who did most of the ground work for this and the patch actually contains many of changes they proposed and provided through their patches attached to +Im not sure what happened to FYI is almost done Im just dealing now with how to best handle the stuff Ill checkin sometime tomorrow morning +this patch is ready to get in +So test packages names must be refactored +Committed the patch to trunk andThank you Andrey +Thank you for your patience +Commit at your leisure +Marking JIRA as done as all necessary info received. +yes of +can this be simplified using +If it does not know it has a permuted matrix it will get outputs +I checked it multiple times +We want to implement the same behavior for all but have not had resources to do so +IMO just address more admin labels for each node but failure group info is already expressed in network topology +but i am not sure about what the is about +The patch I submitted would fail if there are elements +Oh also should have mentioned in addition to refactoring I need to add tests +Attached +tasty version of the was committed in r +in +that still allows someone to try to reply a commitlog against a cluster when downgrading thoughv attached of to address this + resolved issues for already released versions +Some thoughts here +I am pretty sure this is intentional so I dont think it should be fixedTo set it to use ups will cause the shipping estimates code to try to get a ups rate estimate rather than using the table values in the demo data which means that you wont get any ups rate estimates unless you had set up the ups key etc +I dont know whether this is related only to the experimental changes in the branch or may have roots in main stream code therefore I reported it +Default now is and documentation was added to explain proscons for setting it to . +The same test case pasted in the Steps to Reproduce field +I just noticed that it is actually contained in the spring sandbox I guess I should actually post spring sandbox bugs on the SPR section rather than RCP? I suppose that the only problem with that is that because Spring Richclient is dependent on classes in the Spring sandbox bugs in the sandbox become more important to users of Spring Richclient that to users of the core project +I think that tighter rules render this case unavailable so I guess its OK +Please file a JIRA which will hopefully get fixed in +Yes I had to change as well +See. +Please take a look +We plan to add other repos then + with me +NickWill this columns tag accept any faces data model or this will be supplied through data table tag should accept its own re sorry to say that the component has been rescheduled to +Assigning to Rob for technical review. +GrantThanks! on the patch I think its pretty much ready to goCheersChris +Looks like we have to change the way we deal with default values in special cases like this +C escapes me +this contentbadgoodwhich will be related to the component and not depend on where the component is added +Privacy is another important consideration. +Yes had to do this to get consistent layout with text areas will revisit in the future version +Also since this is enabling an additional REST API please create a jira for that and upload the patch +This is still not fixed after this bug was opened. +the generated code is much largerIn this tradeoff between larger code size and smaller source code I choose smaller source code +The issue has become obsolete due to repository and domain model changes +I think its much simpler to put bidirectional data member support into the interface and have the underlying concrete classes deal with it however they see fit +MarkI think we are one the same line its just that I dont want to go that far nowHave another look at the title of this issue it may be in your title bar but otherwise its quite a bit of scrolling so Ill repeat it here Decouple Filter from +This looks good to me +Can you please specify what the form input looks like so I can reproduce the problem +First Patch to remove Aries dependency from that creates a new module called and moves all aries specific stuff into that +Apparently this is indeed an issue on Mac OS X where only one instance of an app can be running. +Thanks for pointing out the need to switch more things over to using instead of +Juergen +Attaching a simple test to verify this +pdf to theURL +When I enable the server module the error with the beeing installed as nbm some other plugin used among our server modules affect the this way +Jucawould you like do some pref test on this case to see if this change is worthtoapply +I didnt use Juliens patch but just opted to reuse pmd +to +If it cant be done in CLI its not a CLI issue but the domain management one +amended at git true thx +Thanks +Hijust tried with Trinidad trunk and myfaces trunkThe demo for trcommandButton uses the action outcome to navigate to another demo pageThat worked fine hereCan you give me a for this problemOr can you try with latest greatest Trinidad on your side +Greg schemaxsl stuff looks fancy but it doesnt really belong to this issue +Elmer can you supply a patch +Some Security enums had to be split out +btw I added you to the project so that you can now assign to for the comments please look at +We need to enable HA in vcenter for this cluster to create SSVM on another hostSateesh Please let me konw if this is not intended behaviour +HiIn this patch i added a field description in entityAnd also added support for this field in the create update and view payment method services +Olga Chikvina will document how to do an installupgrade from update site or site zip will add a subsection to be aware there is not any updatesite zip implemented for JBDS yet this is for info about updating Tools from update site and site zip as two subsections to Tools Installation section and updated info in Upgrading section of GSGthe information about updatesite zip implemented for JBDS will be added later after the feature will be added to JBDS +I have committed this +I think this was actually fixed by which will be released inPlease let me know if you are still having problems with +The is modified to start up the moduleand to call a flush on the buffer when the is thrown +This seems like a useful patch to have but sadly it is and does not merge with trunk +The changes look good to me +Yeah and removing it in an release means that all the people who dont have any problems right now and use will now have their apps break until the get the extra jar and add it to the classpathI think that the negatives far ouweigh the positives in this case and I would vote for not making this changeWhen everybody moves to and later the class will be loaded by the bootstrap classloader and this will cease being a problem +This is related to see also +Ill proceed tests for this issue in this branch as mentioned aboveAlso a new version of the same name is created in JiraAll the issues about this branch should include in its fix versions. +Future additions to the Portlet Filter tests should be attached to this issue +Done for resolution details see +FYI one additional tweak thatd be needed would be to exclude from the job assemblies since theyre just excluding now +moving to M +It is seem as bug is stil remain + Username djencks +Let me try Avis approach once more +I plan to add more substantial expository doc about in and will note it there too +Maybe lines would be already enoughHope this helps you for your detailed feedback Jrn +Please verify this and then close it +the version of Groovy Im using is +Assume closed as resolved and released +with a few additional text correction +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Reopening this issue +Given repeated feedback about optimizations we are working on updating the implementation and design +As a work around to get working on please add the context init param with a value of true to serverdefaultdeployconf +not dst +Overrides will then first call +This issue was just a regression that I introduced on m sending a PR tonight that will fix this +Can you provide some sample code for this issue +Uploading to address the comments +Committed revision +So there is a disconnect between what can be done in and +I fixed it now +The answer to this question will need to be added to the release notes field of this JIRAThey need to copy all hadoop jars and dependencies into lib +patch applied in rev snapshot pushed +Im going to reassign this ticket please if I am mistaken +Committed to trunk +This only seems to impact jboss could you describe this issue more detailly? How the test case should look like +I cant repeat the issue after the documentation restructure +I was reading security code and the attached patch +Its been awhile since I visited so maybe Ill take a look for +Why didnt we consider this before its after all the spring DMLC that is used to consume the reply messages from the temporary queue. +Added the ehcache bundle to the bundle pom it should be released soon. +It seems that the problem is in a new AS plugin in JBDS H +Im trying to apply the patch +The fix would be write app that starts from eclipse bundle as UI and then install jbds using p API +If the dependency between trigger action sps and trigger table wasnt dropped I think then update sql would fail because of trigger tr finding out during recompile that the it is not valid anymoreAs for ALTER TABLE drop column catching the dependency there are couple jiras open for catching cases like this I thinkI am looking at how the dependency between trigger action sps and trigger table can be maintained during the recompile of trigger action +Would prefer to have the callback final in constructorI initially feared pushing the callback to the constructor would artifically require too much code changes but looking closer it doesnt really so committed with that changed +Moving to as the patches for trunk have been updated and just awaiting some review +fixed in svn that are added to the proxy are first checked if they are bridgemethods +This looks good to me +Thanks Paul +go +is obsolete. +These two functions are inherently parallel +Thanks again Byron! +Id say Jettyx is still experimental until this is all fleshed for not being clear enough this issue fixed +Ok for some reason the description wiped out my whitespaces for the cursor line the comment line should have started just after the first committed a fix please build the plugin from the head and try it out +Note that the Patch should be able to be added a open not just after in progress Id sayLet me know if that works for you and Ill do the same +Patch updated tested +Plus slightly modified test that fails without the fix +I see two versions in the Servers View but the jar in the serverdefaultdeploy directory does not have a version number +Cant replicate for DB so it is not a general issue +Dave pls review +Thats the only wayto find the problemxmlsec can log that information on DEBUG or you can debug into the jar and lookat the digest input that way +What type does the id end up being after this? String +localConf is never serialized to diskWe have tested this change and saw the serialized size on disk drops significantly from around K to proving the correctness of the current fix +The patch includes Thierrys patch profiles to build idp and fixes thus systests run the workaround +wsdl file with the Types tried to work around the problem by removing the namespace attribute from the Response XML and then generated theNow the beans are generated with package nameNow when I retrieve the Status attribute from Response I am able to see the value of SUCCESSThe question here is How do I set generated class to ignore the namespace in the Response document without having to manually remove the namespace from the Response message everytimeCan I do that using? if so howThanksSree +Patch applied thanks! +I felt that I had to investigate a bit further on this one since I was sure that the error couldnt be as simple as described first +Second version of the patch +add parameter config and fetch server from and add config parameter to createServer call +can you try mvn please +Its mostly cutpaste to see how it works and has only WAR support no EAR and such +just got test failure also should be fixed +This is building successfully for me locally however obviously tests are still an issueI have a feeling that this also upgrades our support for Cassandra with both AntIvy and Maven +I should also mention that the above error happens all of the time and is not related to commentinguncommenting the event handler described in the original bug +True but if it is like a big difference other options might be worth looking at +Whereas given that Mahout has the DLS thats the way Im going for now +Is that right +but like i said we are not able to confirm or cancel +WanyikThanks for attending the issue our application framework is depending on this thus it becomes show stopper to run our applicaiton on Apache Derby it will be great if it is going to be fixed in versionThanksBest RegardsChee m attaching a patch with the code changes I believe are necessary to fix this bug +The blank archetype is not needed so we dont need to discuss or add that to the release process for the archetypesWe can also remove the blank generate script and any other mentions of it in the release scriptdocs +bulk close of resolved issues. +Good catch that should be updated +We wrote some JNI code but the user had up the ulimit for it to work +nice catch thanks for the report! I am gonna take care of this one +Ill be able to focus on Shale and get a lot of stuff done +Nope +Ive deleted the test scenario +correct patch +Id like reload working prior to the usability testing next TueWedThur +This issue replaces +So yes they will still see the issue then. +exactly AS never included ajaxjsf nor seam jars +If the plan is to provide only programmatic API the cacheuncache API is not needed in +Actually I think you can actually load a JDBC driver into JDK so long as you do not try to use JDBC features +The resulting +Attaching a patch that adds a new counter COMMITTEDHEAPBYTES to task counters +Please merge them to the trunk +The binary format should have no problem in handling this usecase the same way xml format handles ll take this one on forI incorporate hoss suggestion and then we can see how we like it +featuresmaven +When committing could you dochmod x binbookkeeperchmod x binhedwig +Is suspect we need more Disallows in +Attached patch moves BTI on CC and decreases BTI log messages to debug priority level +Closing issue as version has been released on. +This should be fixed now in both the branch and headtrunk +But maybe it is a good start for somebody else +Testsuite ran fine and did some GUI testing as well and I have not found any problem yet +The portlet handles the request and if the portlets session has timed out our portlet may not even get the request +Ill jump in and help out with this effort +Version id date +Joe the site itself does not have even conditional dependencies +Hi GunnerThis PR working fine for meI dont have the infinite loop with this +I dont know why users are doing this +Addressed build test case failure +Sorry to cause trouble +See +For example checkbox status on those menus in not being updated when settings are changed trough properties +If you upload a new patch to address the feedback from Bikas I will retest too +Since this is I will leave the decision for on reproducing this with JBAS but we had noticed the same problems there too +heres a in after release +Same problem on win nt loan broker and soap binding uninstall is unable to remove component directories +The +To implement distributed search with sort well need to serialize the field values across the RPC channel +How do you know that the engine shut down exactly in this time window +I cant reproduce that and this is the first time you mentioned login and password +Thanks applied to SVN. +How about the implementation? Do you have plan to implement itA minor comment the usual practice for referring other documentation is to include links or a list of references but not copy pasteGood comments +test case updates eliminate a few remaining instances where had been used instead of Brian could you please change the headers in the static files for Apache License headers thanks +Yeah sorry +Is there good to define varitem twice +The build dependency is maven Change the properties in under cfgdev folder +The thread dump shown in JBAS is exactly the issue that led to the creation of +Thats true +A a +Still not a problem seen on though +Initial patch for review +Can this issue be closed now seems like the problem went away when the user put the correct jars on the classpath as per user manual +See the Subversion Commits tab. +Put the stack trace here so I can see +Still has the original SVN ID line in it and some of the whitespace may not be to your taste +Galder please go ahead with the change in AS trunk +Documented as a known issue in the EAP release notesWhen gets a request for a class it returns an instance of +If there is a general consensus then we will go with that and if not we can always vote on it +However this came about the NPE should not occur anymore +We should discuss in the ML if we should rollback. +My current work in progress +Nicolas patch works fine for me with Java +If you do not want to commit this then close it so i do not have to keep this in mind. +Cant think of why youd need this but no objection to including it +To my mind adding these either to the Map interface or a helper class should be a feature +The implementation was buggy +I committed the patch to and +that is not expected unless all of the brokers have the same name +Some points to be resolved +Marcus Then alternative perhaps we can consider pushing this into andmaster +Ah I missed your comment as I had this queued to go once I saw one response +Only the foreground popup colors +Just go ahead and commit your proposal +a patch that can be applied to or later. +Tested on Derby DB +Agreed with Viktor +Thanks for looking at this Knut +Filter out values that is greater than e +Ill ask my friend for a maintenance pack versionreleasehowever WL does it +The and solr core javadocs are combined together currently + Username lqd +I really dont like messing with the name +Attached patch fixes the file so that it will cascade correctly with +This bug is still the compiler complains in this case. +I can see that in the trunk version clicking on the configuration link in the UI shows the list of configuration +ThoughtsBest RegardsGeorge Gastaldi +Closing all resolved tickets from or older +One option might be to add this disabled by default if there is interest +Oops I must have missed a few spots in the docs +I get no errors in Chrome but I do in Firefox +Michael thats fine +The deployer should now be used to deploy your web applications +The split object is created when you submit the job and then only used by that one job right? its never persisted or transferred over the wire to some other process is it +bq +StevenIts quite a work to find out exactly all the spring related dependencies in all the camel componentsThere are pom files using spring +Reviewed by Keith +Sure Ill provide a cumulative merge commit on my branchThen you could cherry pick it back to the in one shotRight +Wed have to think how to handle the fact this flag would cause issues in a JSF environment +In other words the Postgres JDBC driver expects a Double for SQL type FLOAT +This issue is a follow up of sounds fine +preferably before the freeze so that we can test it. +I did not implement tested the auth object +For now I think well ignore this and assume that there was some behavior in the gem +Denis please verify +Back when I had a m document index I did queries that took minutes +Great to see you dealt with the issue! +Cant it be some issues +This issue should be assigned again by +Felix OBR referral my last patch +Fixed this in branchSending srcorgapachepivotwtkskinterraTransmitting file dataCommitted revision Will cross to trunk later. +Any advantage would be minimal at best +And this is fine for us + I think after adding some additional libraries to deploy to webcontentlib +Disable the thread in the standby NN first +committed +And yes they need to be balanced across the nodes of a cluster as messages are delivered in a load balanced way +Perhaps we should be using ? And for anonymous modulesclasses +Null or exception are equivalent +Even though your wasStarted check does eliminate some possibilities I would have to defer to Paul on whether it is too much of a catchall solution +has been resolved on other than done should not have a fix version set. +Opened +Scott Any word from Gary Van Matre on this one? JasonSorry stuffs been a bit crazy here as of late +I know exactly what those changes are +They should share a common prettify +Now you try to add the parent server name to the daughter +Patch integrated to trunkThanks for the patch Harsh +works thanks +SGA has been has been recorded so we can start importing the import ve filed the IP clearance record and requested lazy consensus on generalincubator + +See anything in the logs about zk expirations +There are some pros and cons for such modification PROS If a source file is fully reformatted all lines are suddently associated to the guy doing this change CONS If a guy start using some tabulations instead of white spaces some violations might be raised and those violations wont be assigned to this guy if no other changes are done on those this option could be configured it would be great +Fix UPDATE OF clause for DB UDB Version or earlier and DB Iseries VR or earlier +I assume that when the schema allows it that it is possible to do you say New Wizard family you just mean within the structured editor? then yes I guess all is needed +Thanks Kristian. +I dont have time to dig into this one right now but i think the first thing to check is if the problem is when filesize buffersize If this is the case its the same we had in +If it times out it can check how many Jetty threads are actually actively serving requests +You read in my mind I cant wait to see dropped messages graph in zabbixIt would also be nice to have request time stats as we dont know how long requests live in queues so it is hard to say how long clients wait to perform an operation +Added a debug for no type converter found +are resolved as of +That is the only change to the RSS connector I made yesterday +Confirmed +I want the HDFS to distribute its blocks across the large mounts but without this functionality its hard to keep HDFS from certain volumes and disrupting the other work which receives priorityHas development on this ticket been abandoned? If so Im interested in taking up the mantel +Now the test webapp is compatible with Tomcat and it can be compatible with Tomcat by using tomcat profile +Is this still true +requires testing on device +Patch to uplift the classes in the module +Password and Confirm Password are validated if they match each other but empty password is accepted. +Error problem +Not yet any configuration +Closing per Bryants approval to do so. +Can you include that in the patch +ok. +workaround available by copying the collection in the get method +at the trigger execution time is there no way to determine the column position in the table vs the column position in the row of the update result set? If I understand the problem the current code is correctly indicating to the update which columns are necessary we are just accessing the wrong ones +Yeah but I keep uploading the file after +Thanks for the tip +Verified works correctly can be closed +Hello DeepeshHere is the patch for Return History Function for a specific return so we can see who Created Edited Completed returnFollowing are the steps to test this patch +So Im hoping the numbers stay good for wildcards +Closed with the release of. +Can you do some research in the bug repo and see if theres been a fix proposed for it +Committing to and trunk. +Bulk close for release +This patch at least partially reverts the patch from +Patch applied +This doesnt block +Will commit soon +And perhaps there is more smart naming strategy than just c c. +I have not tried it on Android +web apps to reproduce the should read not indexed by default +Any time we nest a schema in a schema or field this is a problemA truly immutable schema data structure fixes thisIve been thinking about several ways to change how Schema works to make it truly immutable including props and aliases other than setting fields to handle recursion +Details svn mount at using +can you attach the patch to the jira for review +I am waiting on a couple tickets to get finished which are currently being worked on before making the should be available soon +Lets fix this tests assumptions later on +reduce the overhead by a factor of But if the overhead is already considerably less than is this really important? We want to keep it substantially smaller than of what we ever expect the block size to be +Yes Francesco you are right +Probably a problem with logj deploymentDowngraded to minor +Ive attached a fix which should get rid of the +I dont really see why this is necessary but if it is a quick shortcut would be to copy code from Apache Harmony which already has implementations of all the java collection classes +Closing this issue after confirmation +issue can be closed +Because then would know that B extends A and would act appropriately +this is basically done we just need to get a product build of the ESB server +The later has a version thing but in case multi fails it will be called sequentially with the same version number +Im not sure what side effects that may have but it sounds worthy of further discussion +Attached updated patch based on your previous comments +I did include it but maven is unhappy with the fact that the version is a property +can probably be closed scheduling for review +Bah thats a on my part +Guess we need to update that code to roll back the workspace creation if the namespace creation fails +Funny how you relate the two +resolve to and make sure it is still under the root +Thanks for checking that +Current work to do Namespace prefix needs to be added to port types on all interaction constructs Add createInstanceyes to the initial receive or onMessage construct Prevent fault message variable being generated in variables section Fault name needs namespace prefix Fault catch needs fault Need to create partner link types in the WSDL +If you have further improvements please either open a new bug report or this one and attach a patch against the version now in SVN +OK as the Sonar project information is based on maven groupId artifactId and version the patterns should be based on those propertiesappappappapptrunkPatterns groupId com +I am happy to do it +I checked for some queries and successful in VARCHAR but not for INTEGER DOUBLE etc +Actually the comment value is always saved in boxed form so no need to have a special load routine here is also wrong +DN connects to cluster and joins the cluster +But either way I wouldcall it an enhancement rather than a normal severity bug +commit to trunk at revcommit to at rev changes are +It is possible for two reasons a the standby hasnt seen the edit log for the allocate block +I just committed this +I generally dont expect that most people will end up with really skewed selectors at least for reasonably large numbers of clientsI had an alternative that is definitely not better so lets not even talk about it +Replacing template files with code fragments as String literal does not help whole program analysis +Ill resubmit another patch after testing it +i agree that the Filter class is doing the right thing +I merged this patch to. +Switching all issues that have been in resolved state for more than one month without further comments to closed status +JacopoI was a bit hasty with the mainstyle sheet change +So it would be good to at least clarify the current situation. +So marking a block to be corrupt using NN recorded length is too dangerous +This is not compliant with CDI nor CDI I ask how was this tested then +Compatbility with the series seems more worth while than making all incompatible with to preserve compatiblity with +It will be fixed in when it is released in about twoweeks +set the logWriter to an instance of c +Did you try quoting the KS name +I will definitely take a look at the branch +didnt register renderers for and +Please note that the warning only appears when the database has already been initialized it seems that at first start only a part of all the are registered instead of more than about the case when we have properties and both of them could not be converted? I guess we will not see any warnings. +is still broken on trunk due to uniquifying Message objects +Can u tell the exact use case in your mind? Can a Scanner wrapper do the work? We have used wrapper to do some extra work during compaction +My patch includes some cleanup of job history +The speed increase will be visible only on true thought +I flagged that out of an abundance of caution +Changing priority to ve made test runs but cant reproduce it either +This is a test comment +Hey Mingjiehow do I do that? Is there some documentation where I can read about the processThanks +Moving all issues to the CR +Also ifwhen you post a new patch please do try to remember to include the changelog entry with it. +Contains the EJB +We have another Jira related to that +Hi ClausLooks like Willem got to this before I did +Should have been fixed on all components +SimonAttached is an updated fix which does not slow down the detectionIt simply moves the hint checking into getSimon +Patch seemed to be still valid +bq +I think thats just a limitation of the parallel indexing task? Ive seen it not hit the target number exactly due to how it divides the docs between threadsActually thats my bad I divide the number by +For restartability we really should separate the start event from the finish eventI dont think it really matters +Thanks Simon and Robert. +The file path added by should be qualified by correct filesystem +It breaks release notes. +Anyway if this is all it takes to get it working I am happy to make the necessary you planning to make this publicly available? If so you should consider publishing it through +The genie in bottle living right in today. +This is the same as what was proposed over in Yes lets solve this issue here once and for all +I found not only and property keys but also following property keys are hard corded although has fields for thoseSo the patch I attached includes modification for those +I think this can be resolved by using relative paths +Tested this behaviour +Without knowing details about simple actions I think that the method actions are simple for people to grasp and best to use in the overview documentation +Interesting patch though +Hi folks I wrote an proposal for this issue but I am missing a mentor for this issue +I feel the post also can be there +What is this behaviour? Should I log a JIRA for that +This should work with JVM reuse set to ON too +Is fuse under the covers just doing the block device so we can do better ioctl? I mean theres no way to implement a real block device since wed have to make it look like a real filesystem +If it is possible to do w pbs alone serversideLooks good so that we can avoid the tofro conversion +I had a quick look at the patch and I think the approach you have chosen looks reasonable +resetting P open bugs to P pending further status from ASSIGNED to NEW for consistency +My configuration code is working as designed the first transport in the config file is reported to the client upon connection and that value updates the +Introduction screen for Reverseengineeringandcodegeneration demo is required +This needs to be turned into a Chris Thanks your update +Applied in r thanks for the patchIm leaving this one open until we determine whether this file can be generatedduring the build +Do you have a test that illustrates this Maurice +Brian you are never wrong +On the other hand developing a plugin is incredibly painful without thisMaybe if you are building normally it should put in the local repo location but if you are using a release profile it should put in a configured remote repo +Thanks Daryn I just commited this to trunk and DarynShould we restore the old behavior for ? This is going to break people eg scripts that do hadoop fs put file dirfile wo dir already created used to work and wont anymoreThanksEli +It is also certified for Sybase ASE when used with driver version +But I am busy at the moment with other stuff +I think it would actually make more sense to have this on by default +Currently it does support multiple payload types unfortunately a protocolKind field was not added when multiple payload types were added +Thanks for your response Dhruba +Patch looks fine to me +Different story on Mac where there is no proper version. +Good enough for applied at cdd +The same thing happens on Deploy Process Archive as well though the deployment does succeed +Please could you have a look on this +Modifying to only try to unload if the jvm is not which is currently the only jME supported by the test harnessI chose this method rather than checking if the version was higher than I thought it would be easier to check the one exception +Any other idea? +Another hadoop spec file for remove any ambiguity if the license isnt clearly specified it was intended to be Apache we are happy to see this patch +Hm I have successfully installed it on Ubuntu using pipI assume you are probably just missing some dependencyCan you please copy the error message here I can try to find a a root causeThanks +Sorry to come back on this issue but this bug is becoming very critical for Sonar +The patch queue created by that script is administered by the build which runs this script every few minutes +Thanks Rafa! I confirm the new url is now used in JBDS +Please let me know if there are strange unusual things +Thanks Clint +Were going to disable it on AS executions +I have finished splitting the tests into generic and sets and will upload the spreadsheet after a review with Ryan +Just curious +ArchieIt seems theres some missing files in your patch the test requires a websitemrepo directory which I havent found to pass successfully +Assuming we somehow know the median just adjust the input as we go +This issue has been fixed and released as part of release +After the append the log rolled since the max segment size is reached +Were just trying to assist tracking this down. +I think this has been fixed in? +As we manage the code base collaboratively I think it is preferable to have such changes in explicit commits instead of hiding them alongside other stuff as I often observe +So its probably best to keep this since other MR jobs can already control these classes that way +Adding the patch that works with hive trunk and branch +Updated to make sure the consumers are started before the producer starts and give them a little more time to finish +Committed revision +change the feature? I assume you mean add or extend the feature to generateThis makes more sense to add in the New wizard I would say wdytThe code should just generate a Configuration object and put it through the normal hibernate tools generation and it would be in sync with the existing stuff +verified on nightly in d GA +Make sure is part of the matrix +Thanks for pointing this out Dan +We will not change anything in the work flow prior to moving to Bugzilla +Hi Mark Im unable to check the build for the next weeks but would be happy to check the week of November th when Im back in the office working on this if thats not too long to wait and keep this open +Thanks John and Daryn +Hope this also contains docs +Thank you +I believe this behavior changed when I needed to add the ability to add a test for the advertiser mechanism which added a new package and the new advertiser wasnt discovered until this logic was changed +Applied changes in r and created +I think you tables? few with regions few with regions and few with region +Thanks +One other comment not sure if it is the case already but Jaime wants that a non Progress employee can only request to create a project if they are a member of the group also +The artifact doesnt have a pom so Im getting the when it tries to retrieve the pom artifact from the has been fixed by. +works now +Yes make sense +See PR for more details +Guys this exception happens at the stage of WSDL publishing + just committed! Thanks Namit! +Fixed in in JBT Alphac +Windows +As sample for message timeout exception after Remote Timeout on Cached you can look to thread name +svn commit m updated query checking to include query filterSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryapiSending queryjdoqlSending queryjdoqlSending queryjdoqlSending queryjdoqlSending queryjdoqlSending queryjdoqlSending queryjdoqlkeywordsSending queryjdoqlparametersTransmitting file dataCommitted revision . +I just filed about refactoring or uniforming the invalid token handling inI didnt assign it to me right now due to therere other blocking stuff need be done asap will work on it later if nobody pick it up +Patch attached to make Template Libraries work from client JS api +I think that what we have here is in some cases some of the metadata for embedded fields either doesnt properly initialized or gets overwritten +Change the duplication writing verification for based on whether is written or not +Wont fix +Certainly if there is a huge amount of sstables the latency will cause lower average throughput but was it established that this is whats happening? With gigs even with sstables the average sstable should be large enough that it seems unlikely that the major contributor to poor performance is the waiting period between streams +Rebased +Works locally +Committed to branch and trunk. +from me as well +Apparently I found that if you have a local proxy server and it proxies central you may be able to work around some of these problem here by explicitly mirroring central in your +When I updated I went to make sure the maven pieces werent killed +Directly below the project name is spaceAnd then as you say a longer description via hover +And I do see some message loss but when the broker is taken down but that can be expected since the messages are not persistent +This problem occurs sometimes when migrating from Sonar to Sonar +Ive fixed the scripts in too +lib containing the lines described doesnt seem to work +Great +The real problem is that were not pushing snapshots often enough +Fixed at revision +Instead of making a check on the network state which could have been one solution I have compared the old CIDR with the new CIDR and hence activated IP reservation +have committed this to trunk + svn ci m Support for Tomcat Deleting assemblyAdding assemblyAdding assemblyAdding assemblysrcAdding assemblysrcmainAdding assemblysrcmainjavaAdding assemblysrcmainjavaorgAdding assemblysrcmainjavaorgapacheAdding assemblysrcmainjavaorgapacheopenejbAdding assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainjavaorgapacheopenejbtomcatnamingAdding assemblysrcmainresourcesAdding assemblysrctestAdding assemblysrctestjavaAdding assemblysrctestresourcesAdding assemblyAdding assemblyAdding assemblyAdding assemblyAdding assemblysrcAdding assemblysrcmainAdding assemblysrcmainassemblyAdding assemblysrcmainjavaAdding assemblysrcmainjavaorgAdding assemblysrcmainjavaorgapacheAdding assemblysrcmainjavaorgapacheopenejbAdding assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatSending assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatAdding assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatDeleting assemblysrcmainjavaorgapacheopenejbtomcatSending assemblysrcmainjavaorgapacheopenejbtomcatinstallerAdding assemblysrcmainresourcesAdding assemblysrcmainresourcesAdding assemblysrcmainresourcesSending assemblysrcmainresourcesSending assemblysrcmainresourcesSending assemblysrcmainresourcesSending assemblysrcmainresourcesAdding assemblysrctestAdding assemblysrctestconfAdding assemblysrctestjavaAdding assemblysrctestresourcesSending assemblyDeleting assemblyDeleting assemblysrcSending containersrcmainjavaorgapacheopenejbconfigAdding containersrcmainjavaorgapacheopenejbloaderSending containersrcmainjavaorgapacheopenejbloaderSending containersrcmainjavaorgapacheopenejbloaderSending containersrcmainjavaorgapacheopenejbloaderTransmitting file dataCommitted revision +Cool +should be a good place +The filters which are operating in can have a problem in this case and +Thanks +Im running locally vs security profile +Slightly made changes in the order of calling of targets also +Hi Gregor caching the content produced by is another issue because it doesnt send back HTTP cache headers correctly +The patch for also fixes this issue +CurtThree things that come to mind +If you update TRUNK youll now need to run +The part of this bug should no longer happen since SAFE is going to be disabled in +Committed to the trunk as revision +Just so that I understand in your comment in you say that youre observing this problem only on trunk so why is it a blocker for +Unscheduling +The patch that fixes spaces in suites dependencies list was applied at check that the patch was applied as you expected +Maybe Ill get some time later today to take a look +Attaching the three config files needed to reproduce the problem +Thanks again Dyre! Committed revision +Hi Joergyes I know its breaking compatibility +Thanks Corrine +Add ability to disable logging so can see what its like running fast Being clear that commit logs cant be compressedAfter doing above Ill close this issue +Applied the proposed fix to SVN trunk +It seems if the file doesnt exist the f variable in the loop is set to the filename including the glob testhadoopdirhadoop so the continue is needed +Committed r +Update sources action now set compiler source and target versions and chooses compliant JRE container based on compiler settings from pom. +Good catch +I like the simplicity of the monochrome icons but they should be coloured IMHO +After we make this change itll mean a client restart is required for and clients only +Is anyone at Facebook looking into fixing this? Part of the agreement when we started using Phabricator was that the tool would automatically copy review comments back to JIRA +Try to refactor the testable part to a method and test the method then + +A hybrid solution is to have the small medium large configs and the start script just chooses and symlinks one on startup +r brings in of +will do that now +Added fix into +I ran this particular test on Z series and it passed +That was the part I was missing +Carlo leads that team so will be able to help you on this one +In addition the real vote count on this bug is much higher than stated +Bulk close for Solr +Digging looks like is actually close of the value stream +Ive just committed this +Applied in r +I noticed that setting allowFlashfalse resolves the issue +What is the rest of the logic for +Yes the test still fails +Initially doesnt seem like a problem to me + of initial Wizard based on tabbed view concept +We did this to test against two versions at once which had slightly different client but thats no longer an issue so we should clean this up +Looking further into this it looks like itll be running the integration tests on a mvn clean install every time +I think this patch will break recovery +Still hoping to be able to further research this issue. +Good initiative thanks Sascha +updated the check for what procedure to forward merge never happened from +Was an issue with the new auto mapping which caused it to attempt to automap despite being mapped to a complex type. +Yes thats would suit until we need to add second custom to the +Did you by chance run the regression test suite as well? regression should do the trick +We have an application that makes extensive use of timestamp at Powerset +Sorry +A good system is to use if its defined and fallback to cookies if it isnt +This new one also has some comments +Fixed in svn +Weve been using it in production apps without problems for several years now +This is a welldocumented deficiency +This way if a port scannermonitor connected to an avro service and sent it bad data we could identify this and throw and error like Client is not speaking avroThis would be an incompatible change +My tests with the current patch did not work on when running a program +Looks great Uwe +Running tests now +Anyone still using an outdated JRE can do initialization the old way +Im giving up on this because it is simply too difficult to get changes to approved at this point because it is part of maven core and is locked down pretty tight at this point +In the sandbox project for we have already added wrapper for the Exceptions thrown by reflection and introspection. +Bulk close of resolved wont fix cant reproduce etc +yip I need this to get fixed in will a simple merge do it +With this approach it is going to be huge soon since we are going to add full sources for every new version +Fixed +Thanks! +Please put these into core tests for now then we can port them to the TCK at a later date +Apparently this information is not identical across platforms +This means that this issue is still validA am attaching a new patch that fixes the issue and also passes the license header test +The same thing holds goodfor test +line Why is it important to disable HADOOPSECURITYLOGGER? line Should we add a commented out set JAVAHOME line here +Do you use the setOptions method in your code +I will create a patch for this soon +I have concerns on adding failover retry etc +We cannot make this kind of backwards incompatible change for users +Any ideas David +All of them can be considered META issues wo actual data loss +If works please report this problem to g itself +Thank you for the patch! Weve just tested it and it has solved our problem +Zhihong I found that the has solved this issue byAnd can I add this to Version? The is just on trunk +make additional relative links to the AS about WAR deployment and EAR deployment in JBDS has been added to the Seam guide +Does this issue belong in the FP? Is it new functionality +Attaching the patch +I guess it was really time to go on weekend. +It will be great if this issue could be addressed in the near futureThanksPeter +This problem is rather tricky to reproduce code above does trigger it as is but not when written as a unit test +I see what you mean +By a standard links you mean anchor like for line number Lxxx? If so then consider the fact that review can be in another file +Applied patch to trunk and branches +The ZK client code is still looping around trying to reconnect just as before +I have not looked much at logj yet +Tested on Cygwin +you can attach text you have used to this issue as text fileSo I can try it myself +Since there is no way to vote here Ill enter this comment as a way of voting +thanks Sudharsan you are doing great job here +A separate issue with UI support request is created in Fedora to add release note note documentation completed +I moved the initialization around a bit in the and to ensure that they will always be called togetherThanks again for the bug report Anuj +May be it was a server side problem. +Can you please clarify whether this is the same project as reported in? If not can you please provide similar file metrics as I asked for in so I can attempt to reproduce and resolve for you? Thanks for your assistance +thank you for reporting this +In then end youll need folks to maintain this stuff and its much easier to find people that can do it on WindowsLinux vs +The second patch didnt make it into X I added it after the issue had already been for X +Submitting the patch with the proposed approach +The latter is used during major compactions figuring cell TTL and its used distingushing two edits of same rct so we dont overwrite older vintage edits +And I change the max memory from to also generates javadoc +I am busy with some other concerns right now +Im not sure how it handles leading slashes but it should behave similarlyGreat patch by the way +If there are no objections Id like to resolve this issue out of the AMQNET database and have the issue tracked via the related issue + +Well done +use this to check it outsvnsvnbranchesengsvnbrancheseng rapidsvn because eclipse plugin doesnt work right +Committed to trunk +forge +Ah shared sessions you are livin the dream! If Facebook ever fixes the inexplicably case sensitive expires attribute parsing on their application cookies we might actually be able to at least get back to sticky servers but that is a report for a different bug tracker +Thanks for the workaround I confirm it works on my real project too +Forgot to mention in this case is creating the threads and it has the that the DN is checking the count against which is why we dont want threads created with this and logically the thread pool doesnt need to share a group with the data threads +These have been partially implemented for but will have docs and full testing in support for deployed services and connectors +This is awesome! Thank you Roland +Please remember to close issues that you resolve yourselfsvn commit r in releasegiraph +Could you send us the fix as soon as possible? This is in the critical path for delivery of attached. +The value is defined in and needs to be changed to +Thanks Teddy +As long as the record content are exactly same I think it is not necessary for us to keep same format on this with RI? Of course just like exception message it is not harmful to use same format with RI but at least I think it is not a high priority issue except some application is broken on this + +Fixed in the branch under revisions r r and r and in the trunk under revision r. +Im happy to look again into whether providing an is technically feasible but as we still wont attempt to pass the current JPA service specification I dont know if providing it would actually help youTim +Great patch! Thanks +unget is part of the service reference lifecycle +How reproducible is this for you? I cant create a situation in which this happens regardless of the rows returned + wraps a so you might as well just use a whose keys and values are identical or somesuch +Had to manually edit this patch may not apply cleanly +Sooner Ill upload the design document to new jira issue +Im actually not sure whether the requirement is right +Thank I forgot the document the inclusiveness of the timestamp ranges +Anyway it is not a remote exception or any kind of io exceptionWhat about the wrapped style that generates documentliteral wsdl +This attached patch contains the fix for the applied +mirror created for Apache Click on was an error i missed when creating the repo working on getting it fixed +could you enlighten me on how its broken? You say the types are different but I use inheritance +theres no need to update internally +Murty I found the Geronimo version you version is could you please try the new version we did some classloader optimization +Resolving. +I tried to move things over so itd all live in but it ended up having to modify way too many things in the jdbc package making things public and extending all over the place +Since shims needs I added as a separate classpath entry +Fixed a minor nit changed dw dir to targetdownloads since We dont want temporal files in main director svn needs to ignore downloads along with just committed this +Thanks Andreas! The v patch fully addresses my concern about the exception signatures in +Could you try it out and see if it fixes your problem +Im marking this fixed as of the build although I suspect it was actually fixed as of the build +Sure is an example how to use implementation +Files batch scripts to launch VS VC projects for VS +as defined by the Containers +Id really rather outside observers not think we all hate each other in this communityIve been super busy lately or Id offer a complete review myself +Will this bug be fixed anytime soon? Can I safely download the jars attached to this bug and use them instead +I see your point +Committed +One bad record out of a billion killing a job is almost always absurd +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +IzaakPlease ensure that the shell works with this change +I shall also update the functional specification +Setting to Patch available so that it will show up in the review queue +Feel free chime in if youre still wanting this +updated desc to reflect real including taking care of the other special primitives +Jacek did a great job bringing lots of source files in line +HiAs I said I was able to replicate the npm registry with only GB of memory onEven if the replication works on it should work with a similar amount of memoryI dont think that its an Erlang related issue because I use the same version of Erlang for my testsI use the default configuration of so I think its strange to have to change something to make the replication work on if it works without any change onI can provide access to my two test VM and I can provide new VM if needed +Then Ill try to fix most compilation errors tonight if I can +Mine is fixed too +All CTC vdbs are using push down capabilities. +click ok button to get to the tabs page +Style classes are now handled by specific objects not plain strings anymore +adds a classname to codec map to the +We would love to collaborate +I also added an assert inside IFD that we hold IWs lockI plan to commit shortly and port this to. +incorporating Brocks feedback +Hi MikeDo you know how to get the virtual machine file? The link to download from the rcmes web site is brokenThanksJinwon KimDept +Bulk close for Solr +I think we can close this out +I thought certainly type erasure would have made these two methods identical and thus illegal but yes it does seem to work here +Thank you Jakob +Can you open up a new issue and provide some example files that it didnt extract created and modified date from? Cheers all Here i attached the Test class and the docx and xlsx file that i have +the filename comes from the CDI TCK repo need Alexey back to work on if he can just rename these classes or not and still keep in sync with CDI this off the fix version for this release though +Hi SamindaThanks for providing the patch +Jeff I assigned it to you maybe you can check up on what I said +Changing to coding with trunk that nobody is any longer interested +Made documentation changes attaching patch +So you probably shouldnt worry about it +I think the theme templates duplicate the functionality that you are asking for in this defect take a look at the same template for each of the themes +However this can be postponed +Manually tested +These have all been applies +Does this look ok to you +I meant moving the call out of +I have investigated more and I will update the description better now +If somebody gets chance to try it out please do +Seems implied here but Ill state it anyway just in case it should be a separate issueReads should also be +i have committed the test file and updated analyzer to the sandbox though +It compiles and the tests pass +I have attached a patch Im using here in my case I wanted the class attribute passed because Im styling my links with CSS +The latest patch implements requested improvements +I think there is yet a third option +Thanks Rob! +Is now configurable through the admin interface +Moved back to this issue as the related issue was already fixed. +About find references then is there a way to filter that list to only show the CDI related matches? Do you mean if we can add our own menu item only for CDI related matches and not to use JDT action to add CDI result to JDT reference list +Ping it back to me when theres at least a description of the code the intended API and some sort of map of how it hangs togetherThanks +And that is obviously not working +Yes that exception is expected +Developers can extend these interfaces to do customization +Especially nice are the detailed exception messages +Attached the sample schemaThanks a lotSridhar +My original comment was that I merely refactored the and to not conflict with the original and whether I should refactor the actual code wizardAfter our chat yesterday I wasnt sure what exactly you wanted out of this beyond refactoring the so thats all I did +Our change is the sameFixed in SVN revision + Released +Right now all the modules jmx source code generators main stuff all get dumped into one classes directory. +resolved with r and the unit test +Fixed in latest SVN +Its background information and not that relevant for our users in my opinion +Hi Ondraalthough I can imagine that we can get content for a jar file from the class path and if the archive contains filewe can get GAV but I would like to spawn a discussion with proscons +rvm cp scratchjarscratchjar ocO ocprintallirtrue method mergeMiddle can reproduce the problem with dacapo +So it is hard to review +Should we do this in the next release or wait one more +I guess that once this reported error is fixed the hibernate repository will start complaining about the version attribute +It should be moved into our xdocs and elaborated +We should put block placement decisions at file creation time in the queue as well +Hi DanielCould you give more details about commands you would get? Would you get similar functionality like SpringBlueprintBest regards dont have any commands for blueprint nor spring btw +not sure what this means? Made obsolete by changes in struct design +this seems to have been committed without the empty pylibcqlshlib file +The fix for is probably the cause for this issue the correct jars need to be added to the example from the seam distribution + +bq +It this still an issue +committed +The two versions are and +The attached patch fixes the Readline and ReadlineHistory spec failures +Agree we will need changes in both YARN and MR +Bug fixed +what do you meant by link? try mvn install +I will prepare the patches for bothThanksUma +Now i am going to submit this patch within min +The temporary queue will be created on only one of the nodes I agree +from JDOM is working +Do you know if this issue only occurs in or earlier releases too +Want to also include the other change you add where the addr string is included in the WARN message in the catch clause below? Seemed useful +againBut anyhow this patch is very welcome. +Hiseems that there is a little bit more work todo +Didnt trace it +Eli We support having multiple options for writing and recovering edit logs +Thanks +Thanks Viknes. +We talked about making center the default but I think we should keep the position used previously but the lookups topleft +Joachim how would I be able to replay the problem +Hi Lance in this case thats represented by the th percentile lower numbers mean faster query timesIll commit this later this evening if objects +bq +Unsetting Fix Version for unassigned issues. +Do I need to do anything on my part for this +As discussed over IRC +Try cordovarun emulator instead +The failure of is not related +In JBDS though wouldnt a user mostly use the Tools Palette rather than the generic Eclipse Palette? My thinking is they would but you would know that more than me + +While it is true that CI systems could provide tools to clean out Maven local repos this doesnt do much for the individual developer who finds that all of a sudden his dev machine has run out of disk space +Today I installed an ubuntu in my vmWare which hosts my Oracle XE DB +Together well lead this to more robustnessI ran the IT you created for and it passed +I am closing this as each jira ticket must deal with one issue only. +Show the code after is finished in caption YesNo +When using the har filesystem you create a connection for each harpath +Len so doesnt this mean it now works Len +updated the so it contains everything and without the extra old datavisualization folder +Markus this looks fine to me however I am curious about your requirement to add this? Do you regularly need to query on this facet? for patch +Set logwriter errWriter to null and setSilent to and apply this patch instead +Beside opening the mentioned files it is also necessary to open +! Nice work guys +I want to make it more in line with the other Bigtop projects and have it install to the proper directories +sounds like we shuold open a bug against deltacloud for not exposing bogus images +Hi PatrickThis is a bug from when wasnt supported + +I ended up needing for so thats what I checked in +Can you please help in resolving the issue +Thank youThis resolved the compile error +Also added some javadocs +regarding the afaik since it uses a a thread is created for every configuration found so basically if you add another configuration for another directory you end up with another thread +I agree with Camille +Update to previous that fixes schema check when a schema is specified via tablePrefix +Ive just committed this. +Ok on a second thought a clear cache callback interface would be better since it would be downward compatible way to jsf and I have planned an eventing system for post where we can hook it in in a neutral way +Moving this tracker out to TBD EAP +Thanks Christiaan! +Typically I dont expect people to use lots of putObject and removeObject since thatd defeat the purpose of aop +bq Btw You patch contaisn some unrelated changes in +This should be useful to actually review the code +Removing from Fix Versions +Verified with latest build +This is not necessary now as we can generate several segments without writing a new crawldbAm I missing something +It is possibly in JIRA +Ive just committed this. +supplied top project team to implement. +As soon as RF is available Ill come back and will test it with and +this is not what classpath resources are for +bulk close of resolved issues. +AXISPRINTFUINTFORMATSPECIFIERWe need not test this specifically on windows and therefore after you correct this on the Java svn we can close this issueRegards Issue as it is fixed +Cookies are disassembled into a object that is used by the to hold cookie data +I see a hang for about seconds and then cleanup continues +Patch looks good to me +Thanks Mithun +I will mark this JIRA resolved +sshconfig so this problem affects old openshift users only who already had key configuration in their local SSH config file +Unconfirmed testing suggests that the overall slowdown using this method is not as extreme as it it is when my entire dataimport happens without commitsIts still my opinion that a fix for this issue would be a good idea but I do not think it should hold up the release +We integrate several products together and frequently have problems with duplicate jars or multiple jar versions a report that highlighted such occurrences would be most useful +Username scoy +Ive created a branch branches to work on this issue and committed the patch to that branch + and PHPMD successfully tested +Its seems to me you just break on the first and there is no reason to assume the one we want is not latter in the scannerThe way I read it this code only works if the first you get is for +trivial patch currently testing will commit if comes back clean +Thats the intended is spelt subtraction not substraction +Note that the Data Bindable field will be removed from the table +I think you should have similar problems with the analysis module +Patch mentioned in the problem description +If you need this earlier in please let me know. +Please make all changes necessary so that we have working unit tests for these features and please change the wiki to match all syntax changes +compress ratiocompress blocktotal block compress save space T +You probably didnt run into it before the backport because the order in which tests get picked from the junit test are unpredictable. +Created a Task so everyone can see that someone is working on it and provide inputhelp like you did +Marking as fixed for now if anyone experiences this issue again they can reopen +On the other hand now its more important to focus on the integration in particular + +Committed fix on both Branch and trunk +Hmm I guess no patch is too trivial to have a bug +only applies to Anakia and the included convenience class it had no effect on the core Velocity engine which expects a Writer for rendering +BTW what concrete need is there actually for obtaining each individual interface here rather than just the leaves? Im curious to learn about potential user cases +here is the firt patch for the translation +I mean making it patch available again to kick off precommit tests again and check if the failure is related to this patch +LuisCan you attach the schema youre using the code snippet you get the exception and the classpath +Inside the update one will find several topology script examples a link to the file in Apaches subversion tree and a expanded explanation of how rack awareness works +Input welcome +bulk defer of unresolved bugs to +backported to with rev . +Lets open additional if new featuresbugs come up +If you find further issues please create a new jira and link it to this one. +if hadoopqa passes +Have you taken a look at? It allows aggregating resources so perhaps it could solve your problem +The reason being that it points usually to an improvement that should be made to the class under test +patch with build of the problems the original patch solved were addressed in other +Well reopen if its not fixed. +The read and write methods exposed by are used by andwhen the total data size is less than k its stored in array once either the setBytes on blob or write on Stream incheses the size more than k the data is written into a temporrary file and subsequent operations are performed on the fileIf call to truncate reduces the size of the file below k the data is again moved back to memoryChanes made in Changed to pass all the calls which were priviously acessing the bytes array to +That should not be a problem should it? Do you know how that attribution clause should be wordedve been thinking more about this seems to me that if we use the annotations then they should be used consistently +I failed with the following attempts specify in in using configurationsystemProperties specify on mvn commandline using +Marking closed. +Marking this one done +Verified on EAR ER +Your suggestion may require a much bigger change +This fixes the failing streaming tests on trunk for me +I also verified that my test script runs fine with the patch +Alexey please verify +Namit Thanks for creating the phabricator entry +Im not certain however this issue may have already been fixed as a side effect of fixed +Seeing no progress yet and that this is an improvement I dont think wed hold up an imminent over this +Pages like this should probably be distributed and stored inside the content respository the same way our errorpages are stored so people can modify them using the wysiwyg editor and not have to search for the file on the disk and use notepad to changeformat it. +adding and to the ear level of the project and exposing and in the runtime classpath solved the issues +Iliev I am closing this issueDIGY +Not sure Id say its it again yet numdocs and facet counts are quite differentInteresting though does that persist after a hard commit open new searcher +Hum feature type editing is not supposed to work at all unfortunately +Link to related issue. +Corrected the reaper thread main loop so that the shutdown test is performed while synchronized and at each point where the shared state being tested may change +Issue only occurs when a Closure is specified as a parameter to a is a powerfull API but it is not currently working on Groovy Could not find the for Corrado I believe this is the error message you get if your src file isnt on your CLASSPATH +I did not test it but from what I see it does the right thing also it adds a test case +It looks like another solution whichdoesnt introduce concurrency issues is to makemake reserved a private instance variable +There is no text cursor in Visual part of Editor +we should definitely do this +bulk close of all resolved issues. +This is a container specific deployment issue and most likely has nothing to do with Struts +Patch touches only jdbc tests +This is a regression from s ETA for the fix? final CP code freeze is +So lets do this systematically +Dion you can continue your tests now +Hi Francisco I just looked at the get method in and it indeed is implemented wrongYour comments are spot on and the suggested fix is appropriateI will commit this shortly as well as fixing the exact same bug inThanks for catching this and reporting the issue +have just committed this to trunk and +The Application part is probably related to AS integration since the problem should have been fixed in for standalone +Dave mind investigating? Ive run into code that had overly strong assumptions about comparatorsequality or weak comparators in the past and this might be related to those changes +agree Id like improve this for trunk +Im already on it +op Trunk Avg Trunk stddev Avg stddev Gain Gain create open delete fileStatus rename blockReport replication clean +I have tested these fixes in my own application +OK Ive added a patch that already supports ivy and fixes removed done privateRemoved override from delete now it is compatible with Open issuesWorking directory would be great to add itTest Im working in the junit test but it is difficult to deal with authentication with remote sites +Hi allSeems there is NO activity on resolving this issue +I will open a new track for this. +From a quick debugging session I found out the followingThe problem lies in the sequence of Knuth elements that is generated for the block children of the table cell there is a penalty with infinite value for the forced break immediately followed by a infinite penalty for the keep +Before I do that I want to make sure that the content on the pages is what we want +I will test once the object model gets the nod +Can you point me to those +If you find further issues please create a new jira and link it to this one +I can possibly work on this just go through and reedit the portions of the code +Ive just committed this. +On +Ultimately there will be a password of some kind somewhere because the system needs to decrypt the password to send it to the database +Patch looks good. +Are you defining agrammar file and using ANTLR though? The use of ANTLR will make the whole job alot easier IMO +We are planning to do this in multiple patches +Updated title and description because another problem with the same method was found for Derby DB +Makes sense +This has been implemented +OK sounds like a good patch +Eduard I just tested using ACF and I get the same output as Railo +If derby has a place in the documentation it would be reasonable to note the current behavior +Hi Kim thanks for having a look at the patch +We shouldnt duplicate that code unless weve got a good reason +implemented to load from bundles and other archives. +yes the red button is the remove the selected styles link +hmm +Anticipating a different approach post integration +The real bug is somewhere with sessions beans +I think this is ready and I would like to reintegrate rather sooner than laterreviews reintegrated the branch and committed to trunk in revision +Thank you Scott for the info +It will be in the builder after eg number or higher +Tested with both tomcat and jetty Generated html files are created the svn src directory and therefore litter the src tree +At the beginning we had and +I suppose that but I cant figure out why version of plugin is released when it cant deploy ejb ear application to Geronimo by default +Actually I take it back +Ivan FlavioPlease go through the latest patch in for the detecting bookie failure and would like to know suggestions +Why was this not done beforeIn order to this the will need to have a reference to the xfire Service object +I wrote a and a for my app and simply switch between thoseDriver Version Server and it works fine +I tend to agree with Boris +This is a current code of hamaI would appreciate any advice you could give me +It should just be since all you can do is blow up your own AM it wasnt quite the case with MR +need to remove them tomcat source file seems LGPLASF so I dont see a problem there +Cancelling for now + +you probably have paths relative to your own web siteso you dont know the exact path ahead of AMD if module paths are the same and they are used twice that is actually supported by requireJSand will be handled correctlyAlso as you said being loaded prior to wouldnt happen in this case +yes thanks +Oleg Reviewed the code +If the usage dialog is created before the runtime dialog will use the shell of the usage dialog as a parent and wont be blockedIf the usage dialog is created after creating the runtime dialog the usage dialog will use the shell of the runtime dialog as a parent and wont be blockedI have added the flag to the runtime dialog to suppress closing the runtime dialog when closing the usage dialogThis fix is similar to the fix forI have tested the issue on Windows and Linux +I think going to will be too much effort for now +I first started to delegate the whole functionality to the controller but realized that statemanagement ist in +Closing issues for +Do we know when a derby binary release will include this fix + +Should be simple to implement +Thanks for the patch Santiago Ive applied a slightly modified patch was a for getting the error I show in the description +Both the and timeouts will help to improve the read performance +I committed this +So I have just rolled back the change in dfc on Cos meant to weigh in yesterday but the times I remembered and the times I had working Internet didnt overlap much to rolling it back as you did it was indeed meant to fix a stale ivy cache which was breaking Jenkins builds +Im also experiencing the same exact issue +But in my real world application some new problems occured +I will try to fix it asapFor the moment you may try a workaround +Test file inclusion pattern Pattern used to include some test files and only those ones in the Sonar analysisMoreover the sentence Changes will be applied during next code analysis +This bug has been marked as a duplicate of +Im guessing youve built it locally and its in your local repo +I cannot see why the DOCTYPE declaration is a problem +Ill take a look at it and see whats going on with it +HOD is now part of hadoop so supporting it is not Yahoo specific +I also fixed up exception handling a bit and alphateized the aliasesThis could use some documentation since if you edit the LHS of an alias it creates a new entry rather than replacing the old one. +Is there an issue about upgrading Jetty to? Why Jetty does not have this behavior? I saw that Clouderas cdhu use Jetty +This is crucial functionaility if Solr is to be accepted as a solution in any organisation +The Project was successfully created and the ear descriptor files were correctly updated +Fixed in commit eae +As explained there the intention is that method should mimic the servlet API which returns null for an unauthenticated user +Im one of those people but I figured Parquet support was worth it + Username aok +We can always if it turns out to be a separate problem +committed to the trunk and +Also if they are using the attribute it wouldnt be doing anything useful as browsers dont understand itI think the tld attribute should be removed same as the img tag +Just EJB and CMP subsystem +Is there something wrongmissing in my configuration? BTW when checking the deployer I think there is a typo in the name of a bean +Adds factory functions to Iterators that perform the filter and index checks and return null for a miss +Tested latest version of patch +Whoops we didnt this to x! That was a major over site and we just missed a release +The problem does not only occur with bad datanodes but also with full datanodesDuring a long running job the distribution of blocks written to dfs is not uniform +Tests passed cleanly for me on the previous rev of the patch +Jopr r Use xpath to pull out the chunk of xml we do some further trimming of notes +Im marking this as done. +Apart from the XMIModel extension properties there seems to be a few places where translator and VDB properties need to be managed at a translator perspective there is also the issue of versus custom +As of Castor Castor is uploaded to the common Maven repositories incl +Could you please explain +This patch ensures the setup scripts are aligned properly +have moved to this JIRA tracker. +Works well Fabrice and in the documentation I guess we should advice to use by convention the binsonar directory +Added one more testcase to the this is possible to fix in C and Ill add that fix but it wont matter until this works in Java as well +I just exchanged the new with the old and it got back working! Even with Xerces +No PR +Any possible way to mitigate risks of getting the wrong object? +If it occurs again we need to get more data. +Where it is written that must be built with JDK? Officially Hibernate requires JRE + +has nothing to since is now released +Since the class used as key is again part of the cached data as field member of a Field instance the entries are never removed +Added a new Controller object which takes the string and parse the int value +is used in serialization layer wheres no notion about endpoint or the current fix seems ok to me +The default Editor tab this editor tab which will be shows for pages which we hasnt open early in vpenot only for new files +Go David Go +Committed revision . +I committed the main change with revision then noticed I missed actually passing in the null and fixed that with revision +Shindig still doesnt use any of this data + thanks Colin +Its a shame for me +Did a bit of log splunking +only the one for the current locale +Using the dependency plugin the BSF API is copied to a local folder and referred to in the compiler configuration to place the BSF API library in front of the standard JDK runtime libraryThis seems to do the change does not work on Mac OS X as is not availableI reverted the change for now +Thanks Sid +We can file issues with it as we find them +It might be helpful to you +INTERVALMS instead of +Thoughts +I see a few references to others having the same problem +I added the same modifications for the xobject mapThanks for the report and the pointer +That was quickI would love to verify this fix but having to manually apply this patch will take some brain twisting time +Heck we could provide that simple filter +Max put down the crackpipe +My suspicion is with the regression machine +Checked in modified patch at revision +This may be invalid +Hi Ive uploaded a small Wicket app to illustrate my pointIt runs a simple search for items with paged resultsTo avoid db or file config hassles the result items are storedOf interest are classes You should be able to drop the WAR straight into or TomcatOnly these are needed on the classpath Context is athttplocalhost +Sorry forgot to resolve the issue +strange +Verified in +init is not a java standard in any way +This patch needs to be reverted +Committed as svn closing. +There were a number of dependency issues that I had to work through +This patch simply adds the JSTL taglib directive +Manually running the test passed +The one bit of configuration Tapestry absolutely requires is that your get the application package correct +Looks like this is livereload issue which doesnt connected with browserSim +As you posted your comment I was writing mineSo you tell me you cant know if youre a leaf +Place policy in PHOENIXHOMEbin and change to use the new policy jarfileoptjamesbin! filePHOENIXHOMEbin +For our purposes one set of bytes in the MAC positions is as good as another +starts another VM this may be a lead +Did you ever try a more recent version of like? Does this happen when processing any pdf or only if working with a specific doc? Which imagetyperesolution are you using? Can you provide us with a sample pdf +Updated to latest version r in subversionBut I cant get Eclipse to generate a decent patchIs there a sandbox area I can create a svn branch and commit intoThere is a lot of change hereI dont think its more than necessaryDue to the large amount of changes the patch diffs are not as good as they should beFor example to ensure immutability the member variables of have been pushed into an inner class but the diff doesnt show that these lines have been moved elsewhere they just get confused with the context that is left behindSerialization numbers have been bumped where it is no longer backwards compatibleI cant see why anyone would be wanting to serialize between so it shouldnt be an issueSomeone with intimate knowledge of design philosophy should be looking at this patch carefully to make sure its not doing the wrong thing. +Hi David I tried the attached patch and it work pretty well +bulk close of all resolved issues prior to release. +There is no need for an API to map a fileid to a pathname +I began to add enhanced is Bundles revision I keep on running +Ah found it producttools I saw the forum thread and just took a quick look at the page there +creating a patch should be the simplest and assign to meant to say +Write operations at ANY are a problem but not many use cases are writing at ANY +Looks like someone already fixed this actually I see new logic to only close if print +Maybe you have some D data and the issue happens only in that caseThat said we also have automated tests for d data and its passing +Ive merged a fix by just backporting the fix that was made for +I could verify the same failure with HDFS to HDFS copy also +only push consants and thinks that are direct column references +At revision o Merged rev from trunk +We need an FO file to see whatyouve tried to do +ok on this patchThanks N. +The need for selectedCertified in the titles was an the extras updatesite we could simply just have those shown by putting them in selectedcertified categories +Adding support for configuring SOCKS proxy +Patch branch +Updated patch with duplicated common type data pointed out by Nicolas Jacques My apologies for my silence +the diff is because it is unsure what the spec wantsn vs plain +Okay I will separate the into a separate cleanup +Could someone please reopen this as it is an blocker for development with mavenIve tried to figure out what the problem is but couldnt find out whats causing the dependency to the missing artifact +Patch attached for review +Looks like we have problems in such instances see +Alright +applied thanks +For backwards compatibility it may need an attribute that tells it to propigate its exceptions similar to the attribute that it has that tells it to ignore certain exceptions +If you finish it before we release beta please correct the version +Yes the tip has worked but the point is that cpptasks is not yet released as file on sourceforge +Looking into it +is a core property and must be set in +There is a bunch more trace level logging in if you can reproduce with TRACE level logging with and attach the log that may help +I havent been given the ok by Microsoft to start contributing my own code back yet so Im stuck asking other people to help in the meantime +Thank you very much for pointing it out to us +Thanks Ceri +updated with trunk +Closed. +auto no longer called twice +Philip I will apply test the code tonight +Is this issue happening for new templateCan you plz verify on other hypervisorfrom the log Issue is with the services running side the VR +I replaced it with and everything was working fine +I dont have Visual Studio so I will not be able to test it thereIdeally we could get it to work with Mono too but I dont have enough experience with either C or MonoHas anyone else looked at this? Thiru +Ill tweak it a bit more for that case +attach any test you might have that reproduces this +I have ensured that both s server adapters in eclipse now use the same service the service which holds these jars +I think throwing an exception when the connection to the datanode fails would be fine +the release notes are updated but the field doesnt support formatting +Hi Chip can you double check the some jars under deps? The log shows its not valid its possible that maven downloads an invalid file into deps folder +Did you want me to attach an updated or addition patch +Correct +Yes I havent been able to reproduce it either +We should also try that binary installation from JAR does not pollute JVM memory +Janplease look at my last comment +added test and committed the all issues that have been in resolved state for more than one month without further comments to closed status +Hi VinceDid you get a chance to test +If it is too hard I guess its late for the platform +now can get configprefix from properties +Srinath +just committed an initial stab at pulsing cache support could prob use your love again Oddly the reopen test passed no problem and this adds more to the cache perhaps I was seeing a ghost last nightIll know before too long +Ive noticed some similar flaky test failures on my side and will do some runs to try to duplicate failures and that keeps the full stdoutlogs so we can debug +Updated patch based on Nicholas review +Mostly trunk +From the stack trace it is just delivering a service event which causes a class load that fails because the associated bundle JAR file has already been closed +No plans to add this maybe we reopen in the future +I have attached a modified version of your gml example which uses a Polygon and a Linear Ring +Thanks A LOT Ryan you fixed it +Yet there is still memory used to remap ords etc +Are you just proposing to fix the +has zero impact on tools users I assume you mean that a Tools user willcan still use Java EE or Web Perspectives +Pls reopen if it is not +Its better to have different names for each hookother wise same implementation of a coprocessor hook will be called multiple times and can cause inconsistencies +So now coupled with we can create a maven module that doesnt by default depend on spring +This is not a fix though +We are also wrapping the request here in order to provide the request parameter for Pseudo identity assertion +Is it possible to get at least an estimate fix date for an issue marked as major priorityIs there any known workaround? Any help will be highly appreciated +additional broker improvements in r have discovery agent and network connector delegate to the common pool that can be explicitly shutdown. +Candidate for last comment +I dont want to keep breaking build for each major version +One more question I had is how you see this being easily configurable in the? For example currently by dropping a workitem handler impl jar in installerdependencies places the jar in the right places for users +I dont fully understand your issue but this might be related. +Personally I am not rush to push this patch to as this patch is a bit more related to fault tolerance and it may need some improvement when usage cases increase +Lets get right in Gump as part of +Fixed test case in Rev +The javadoc is fixed in +Anyway itll be easy enough add support for this to be controlled through configuration +I think the Clay changes has us pointed in the right directionGary +I will try with a new fresh build from current latest revision and tell you if I still have the problem +The initial chunk can be configured with the parameter result +I will commit by the end of the day +bq +Robert patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I havent isolated the differences +Released in. +However thats not for all options for instance java host localhost does return with an error messageWell this is because there is no host optionHowever thats not for all options for instance java h localhostalso does NOT print an error message +Closing +One thing that may have helped your original configuration was to ensure that the spock jar was exported from the project +Which channel a package came from should not determine uniqueness of that package +then correct fix should go to +To get the behavior that you want you can set to a large value +looks good +Part of the fix was +The test failure results file and a patch to fix the problem +owen runping I disagree completely with the assumptions from which you reach your conclusionsI think you are not looking at the total system costs +and jars is working fine +I dont mind using just ls on Windows +The working on this problem Ive found that vmmagic support in has potential flow that can cause GCGEN failing when compressed mode is enabled on EMT platform +here is a simple test case +The remaining failures reported by Prasenjit are now being tracked as so Im closing this issue +example code is attached as MichaelIll take some time to review that work around later if it works then Ill look at how I can refactor Transfer to use itThanks so muchRobert +ve just committed this +That makes sense and is also more relaxed than the current matcher I think +It is being published now and will be available in a few missing footnotes is however an outstanding issue +New patch fixing findbugs problem +Attaching a patch for branch +Applied to trunk +Yeah +This is the first version of patch +Even user had a propertyDefault of APP prior to the patch it was simply not set +Putting back in +If we raised only once there was finally exactly zero bytes available it wouldnt change anything +The patch does not apply anymore +Hitested +Thanks Colin! +By doing this in an agile fashion it would also allow us to work out a number of potential issues I see when moving this from Oozie which uses a DB to store its state to a workflow AM where that is not possible +I think the name should be different since it only works with fields unlike other and +Still waiting to see if it passes on either platform +bq +But on the other hand just having that few things in core also feels wrong +Yes its OK +Huge thanks to Chris Mattmann for helping me track down the code to change and thanks to Brian Foster for finding and fixing this issue in his wengine branched filemgr code +Here you go this issue is still going strong +This might not be true at all for JME and I have currently no setup with a test environment +No problem +Closing resolved issues. +change applied +for the patch +With the completion of this issue has been resolved +Hi DonaldFinally I found another solution which allowed me not to have to go to through the patching work +I completely agree with the fid requirements and yes we should then clearly document what we support and do not support +I also missed the first entry in which explains itIt is the same like with the Fieldable interface in the past it is seldom implementedoverwritten and so the normal user will not be affected +Are you saying there is something else still to be committed +Hi LewisThe dependency version of in is still +Please have a look at what I have so far and share your thoughtsideasscorn etc +characters are converted to xWhy and wherebq +Its just occurred to me that it ought to be possible to add a temporary shell wrapper around xsltproc which takes care of the core dump and the exit code for the apsite buildsIll give that a try +Hi Im not sure I understand this issue +It would make it more info to clients so that users can debug better +Jasper is Not a PDFA file by any means +Editorial pass over content. + Ill commit this tomorrow barring objections +Closing. +I found a reason for runtime error there was old in my PATH variable +I would expect the JDT to handle this Id be curious how this is handled by the eclipse plugin classpath the second point I agree A should update as soon as Bs pom is saved +Weve found that EXT that limit is around k +This doesnt solve the problemIt still sees an entity callback without the closing +On that issue you said you would research using Jenkins credentials to upload to the Nexus snapshot repo but you never didOr rather you never reported on that issue that you did +Should we close this issue? No one seems really interested +Added tests to my patch +But interpreter run hangs with attached outputCmd I usedharmonyjava Xint jar application dev bin formatter testpluginname classname this issue toAccording the comments from George this issue can be resolved only after class unloading feature is supported by DRLVM +Also I think that we still need versions to be of type because its useful for example in were we need to know if and what migrations do we want to send to others +Fantastic Andrew +Current shutdownNow method changes default behavior and does not interrpurt running tasks +No longer valid. +Each User who is logged in identified by their has a list of that are allowed to use the that user createsI had implemented this as part of when I did it myself but the patch as I submitted is more generally useful to others as it allows one to set up any filters they want. +Freddy Ill send you my database directly +So lets hack it for now and I will file a follow up ticket to refactor this +Thanks Ravi Prakash! +We arent creating a true route +Would recommend keeping it as it is unless there is a definite bug +This is because under at least IF will only check the existence of files not directories +Also take a look at what can be closed here please +Yes +Something broke in the tomcat builder the context path should never be null +Reviewed and pushed. +So yea disable it +OK that works for me +JulianIn the future please submit all patches as a unified diff file attachment and not as inline text +Resolved by updating freemarker runtime jar included into the bundle to +I am not able to reproduce this problem +Hmm good point +The default behavior could be to generate a Map for the row but userscould override that to generate a bean arrray etc + +Hi TomIve now run and on both and +The design intent of the code was to not require all rows to be updated when a null default column was added +One uses the Cathedral as a whole while the second uses a stylized version of the Cathedrals main rose window +Thanks! I do not foresee using different catalogs for source and test compile +forgot to remove some duplicate properties +putting on Final since can be done without respin +The attached file contains the OBR description of the Shell using this file the will generate an improved description for this bundle +Just looked at svn diff for every changes in the referencing module since and found nothing significant +bulk close of all resolved issues as part of closing items. +Thanks Sambavi +Committed to trunk and +One way of verifying this would be to run IBMsJVM with the JIT option turned off +bq +Pls follow the step shown in file I attached +Committed a working package targetCan you put up the output of the test that is failing for you? They all pass for me +Based on Tims comments and test case failures it doesnt look like this JIRA is completely resolved yet +This is trickier than it should be because it is caused by the Spring MVC dispatcher removing what it thinks is a file extension from the URL before dispatching +Added the relevant just committed this +Good point the best we have at present is cachefalse which will reobtain the target JNDI resource for every invocation +I dont know if it works +Once RFC is superseded by a newer specification will be adjusted accordingly but not beforeOleg +Hi KonstantinThanks for your comments and for reviewing the code +batch transition to closed remaining resolved bugs +It looks mostly like a simple refactor to make the connection context generated later +slap coffeerefill +We cant find the cause currently +committed to trunk +I have then made this bug to remove the tracking url printing in the job leaving that responsibility to the client service delegate + Design a new icon to distinguish remote portlets from local application types portlet gadget be verified in gatein master and reopened if necessary +Trails custom descriptor annotation handlers do get invoked at bootstrap as designed. +It makes failure analysis a way easier +Closing this as Dennis has fixed the docs. +Just noticed my patch depends on the additions to the tests in +I am happy for the issue to be closed +This would require adding an entry to the configuration per embedded scheme rather than a single entry for all schemesnot a big deal +In this case some maps are still running +because callee is not accessible in function that uses strict mode +There arent two alive at the same point in time on the same thread +CorinneLooks good +Sergey have you looked atBy using the new stem option I think you can get the same behavior with if you use preserveOriginal along with catenateWords +Anyways the end result is this issue is fixed. +Change follows Ivans comment logging the error instead of throw a runtime added a new comment to the review board +Thanks for the patch Edward +Its not clear to me that supporting a static library version of libqpidbroker is really very important so Ive decreased the priority +I think this patch is enough +Is it possible to post the complete document rather than the snippet? I assume tns is the wsdlsoap namespace? Is the xsi namespace defined in the document +Which is why it fails to compile +Sounds good to me +I noticed that if you fully qualify the name of its base class +Hope this feature gets added. +I think were done here. +Ive observed this with MSIE for Mac +I think that giving a user the power to set all database properties is pretty much a statement that they are a DBAThats an interesting comment is it the case because it should be that way or it just happens to be so because there are security holes that have not been closed yetI tend to think its the latter +Im pretty uncomfortable committing changes to compaction at this pointOh yea +Oops I think this new test is going to break the build momentarily since guava isnt on the classpath until is committed +This most likely cannot be completely eliminated because subsystems depend on some stuff in server and the subsystem handlers need to be usable in the DC +Hi RogerIs it a issue? Its a different project than MessagingHoward +According to the spec of OSGI the does not support all the obrs functions such as no repository admin to communicate with other repositories no resolver to deploy a new bundle into local repositorySo I think it need to reconstruct the code and add the functions mentioned above +As the version under development seems to resolve the issue I had I would have closed the issue I created but I appear to lack the capability to do so +The in seems to be in between and trunk +I think we have fix this in current code base if it is possible can you please check against current code base +Assigning to Alejandro this is a jBPM specific unit test +The null appeared in the path is due to the null in working directory +We can either throw at the server or fall back to checking the value of write +It seems to be a problem +Thanks a lot Francesco for your suggestion Im starting working on it thiscan you provide also a junit class please? Thanks in advance +Correction the proper workaround is to increase the lock of the default cache not of the cache being started +This bug has been marked as a duplicate of +disabled storage pool refresh lets see this issue is fixed or not +The patch for the static error message buffer was committed at revision +Has this ever been fixed? Im certain that this is causing problems. +Our test team developed a testing tool by hacking ab to support SPDYWe use this tool to do stress testing on SPDY pluginYou can also use spdycat client to do simple testIn our testing the in ATS plugin framework consumes too much memory +This one looks scary +Ludovic I like your proposed change +Please to with JBT. +The current testcases for the jca project are pretty simple and it would be great if they could be extended +If its possible ill prefer you to build it and ill download the artifacts +That already encoded stuff should be undone properly should protect us against pre data looking different in +Even before someone was crazy enough to make me a committer I developed a lot of respect for Robert watching the dev list in general and Jira in particular +Thanks for looking at the implementation of Knut +Any volunteers to review commit this? If not assign it to me +The patch looks good +As Ive fixed now maybe you have some time experimenting with this featureI dont have that myself right now nor do I have time to update the documentation soonBut feedback code andor docs are always welcome +Not sure if this is relevant to this particular problem or not but wanted to point it out since its in the general area +Though national character types could be added in the future more or less as a synonym for the regular types. +This patch enables parts of the test +Fixed in Apache Karaf. +Latest patch looks good +bulk defer to defer to +I also improved the console monitor so it now properly handles multiple files being retrieved simultaneously +Note that the webconsole should be available using the featuresinstall command even when using snapshotsIve recently fixed a bug in the maven url handler which was handling snapshots badly and was not able to resolve those for some reason so hopefully well have a new mvn url handler released soon and the deployment of the webconsole should work even from a clean local repository. +I just committed this as r +Test only so far but Im working on a fix +Everyone agreed that is not a bug so closing it. +Thats a major hit on the ops side +Bulk close after release of +The problem with classpath discovery is that there is no way to detect existing resources +That is we block thus removing any possible starvation of this job +Right now were logging control path operations +is targeted for CP now +Stuff like this can only improve the quality of the product make it lean and easier to maintain +yeah +No idea why this was both Fixed Resolution but still OpenNeed to review our workflow. +The Component being created by the various createXXX methods is already being returned +Added the header +I cant reproduce this with another problem is that in a multi module project the coverage report does not include tested classes from given modules dependencies that are in the same project probably because during test phase there are classes included on the class path for dependencies +As it might be used by existing customers already and removing it will break their setups +The semantic of the method is return the sum of messages for the durable subscriptions +Why does the policy need to limit the number of racks to choose from? It seems to me that the more racks the betterHow do you plan to handle the balancer +He can assign the issue to himself whenever he intend to +Andreas I think you have to reconsider your design because xstream with reflection converter does not provide cross version compatibility +Or if you can create an example to reproduce it I will look into it +Hue could be useful for us and Ill give it a look to see if it would work well with our use casesCurrently Ambrose only runs embedded in the Pig client VM so it would only work with an embedded Hue instance somehow +It is greyed +I did NOT see any assert failures when I ran on the latest trunkThis failure is reproducible with +New dynamic web project and new seam web project wizard page Web Module cofingure web module settings +Just add +Same is true on HDFS empty patches still claim lots of new release audit warningsit looks like the script and the properties file are both externaled in via SVN +scott can you check why the enabled attribute is set anymore +Could you please give me some insight on what could be wrong +We are going to close this issue as wont fix +Committed revision Im now going to work on making things more extensible +See the linked issue for further explanations +yes Im delighted +If its been like that for more than two GA releases then we could remove one of the copies +Nice +Jason mentioned that he has some ideas to mitigate this +should be fixed +This was if I match my environment it is likely the tool will work +allowing unicode escape sequences in identifiers +Tested +Looks like something in the calendar module so mail example is ok takeaway example wont cant create mailboxes on atm +I cant create a reproducible test case right now +Here is also the updated patch for +Kabir do you want to take a look at this one? From what I can tell it is the integration of the domain model in JMX that is causing additional entries to be returned +Closing this issue as we no longer see segfaults there are errors but these are unrelated. +Things got a little busy this weekend but I now have the sandbox figured out and can push the code there for you to download +Fixed on trunk and the branch +It will always instantly pop back to the page it forwards to +Attachment has been added with description Sample FO to recreate bugPart of the issue already fixed locally by working with a deferred block list to which the elements of the part after the last break are added before continuing with the next iterationThis already keeps the content from disappearing but as I expected is then no longer automatically triggered for the flow table part in the sample +Ill have to have another look tonight but yes I think that explains what I was looking at +Each DB vendor also has various levels of support for inlining so what you see could be DB specific as well + +Ive changed that appropriately in trunk +Must I use hadoop? I have not use this version before +I think that some of these failures will go away with +added more Wssj does not remove headers +the behaviors in should be defineddone tests PRL +I also tested on a cluster using hbase master stop and the backup masters properly shutdownIf this is acceptable Ill provide patches for previous versions + +Thanks Sture that patch looks goodDo you have an example +We will discuss more in the team meeting +include rather than fix whitespace issue put visible for testing functions in a separate header create hdfsinternal structure type and use that for hdfsFS +Pat do you have a patch for this that demonstrates what you are suggesting so that we can compare +There was a build issue that prevented the first one I updated from appearing which is what I thought was being referred fixed now +thanks suresh! +Since is out I just update the version in the camel trunk I can close this issue now +Ive added a global flag to with default true +You should probably update to one of the and prepare to move to final soon +Didnt completely match the configuration but did notice that if you crank up the number of messages the test sends the test will fail for NIOSSL and the CPU will be running wild +Quick patch to add mockitos license text to LICENSE +new attach +I see no reason why that should not be the case here +Have you seen that one? Would appreciate your thoughts there +Works perfect on runs in sec. +Sorry for the inconvenience Karthik +I kept the old method parse signatures so this is now so when folks upgrade from they wont need to recompile their generated Java +I just verified this patch fixes the issue with those hcatalog tests are flaky +edit some code in the grails app to import a class from that jar file +I am reopening issue because sometimes I am still seeing the problem and exceptions in log +Thanks +It is working fineThanks will be revisited +as r +I believe is +Committed with last proposed simplification. +Another SPAM file by this Charles L Fields guyI found two users with the plain text name Charles L Fields one is fieldscharlesl whose address has been marked disabled before +I assume that the outlined solution works +Assuming it works therefore closing. +One is +oops added whitespaces cleaned up and minor code cleanup to match code style +The performance penalty for a pair of deserializations for each compare in both the sort and merge is difficult to overstate but for prototyping in particular this can be useful +Id be for just removing it +I tried both full and I need to add as a Utility Jar? If so I cannot figure out how to do that +As stated in Adding support for pluggable components would allow exporting DFS functionallity using arbitrary protocols to start custom services +ill file a patch shortly with a testcase +This mustve accidentally gotten assigned to me while I was reading through it with Steve assigning back to its rightful owner +Sounds good to me +I will be out of the office November th th on vacation and holiday +Yeah this is something we really badly need +Um +The manages the in memory and on disk data and has a thread to do mergesWell post a patch with the api soon +is t fix +The snapshot version has been looking good with the TCK +Duplicates +Resolving +for the patch pending Jenkins run +However in an application server environment the Derby driver cannot be unloaded when a single application is stopped since there can be many applications running that all depend on Derby +It is working fine on Android +Work in fixed this. +Will update . +Done +for the patch +bulk defer of feature items from to +Jeff is correct +However looking at the requirement from a logical perspective if Rule R has Conditions C C and C resulting in Action A and Rule R also has Conditions C C and C but Action A the rules are in conflict +Other parts of the CAS definition may still be different such as the type hierarchy +Uma thanks for listing them out +Offset seems to be the incorrect property to use and index seems to be the right one +Patch with fixed javadoc warnings and findbugs warnings +ms sounds more typical cost +You could make this patch much simpler by using the functionality that already exists in the frameworkAdrian +Hi Oleg +Seems working just my SVN build doesnt support SSL +Well revisit this during the +Im reverting the changes for the time being. +Fixed an issue in commit accbfdbaeddd where a wildcard is incorrectly used to resolve the GWT modules config +I would only add that we probably dont want to add this to the actual response document returned but rather attach them to an additional section like highlighting works +At the risk of spamming comments here I am additionally seeing the following symptoms Old are fine +Sorry that should have been r +Do you have any other proposals for names +Could you also post your file +We should also disable event logging for the event log table itself. +updated the test with trunkThe file should be +I think we should make it consistent +Verified +Verified +Not yet clear on how the integration tests are supposed to be run so not submitting test +But it apparently needs further work +havent tested it but it seems sufficient +Note that the root dir of this patch is not project rootThis should not be committed to trunk it is only for those who cant upgrade to amq and need a remote web console +This fixed the leak for me +That is the intended behaviour no skinning +wrong project +I will go on for the undeployed web service s fix this for trunk first and well see if it would make sense to port it to LukasI have no idea for this now it sometimes failed and sometimes successed +All set. +This implementation does not provide relationship support +Hmm +Ill check it in Chrome in ten minutes +When you say we would check the state in the execution context you mean that this value is the total amount of bytes that FFIW have already written in the output file right? So if nothing is written the FFIW creates a new file if it has already written something and doesnt find the file it blows up +Thanks a lot Shawn for your continuous suggestions on code style apis used and the coding standards +A patch for is attached to that issue which includes the rev +Patch applied in txmanager trunk rev and branch rev +Ive attached two different testcases which doesnt work with the patch +not when you add a repo mirror in front to avoid constant downloading nor if you run mvn o and have built multiple projects against different sites thus tycho will be resolving blindly against the full combination of cached repositories +Some minor changes to get several tests on the branch to pass when run with the ibm run down of the changes included in the patch The order of entries in a message is reversed when using the ibm jdk account for that in the test +Hello EdwinHeres the new patch sorry the old one I didnt create it from the root pom +I have deployed a SNAPSHOT which will display a more detail on the you try it? if still an issue. +ScottI just did builds of petstoreWeb in SVN and in the distribution and in both cases the file was available in petstoreWebbuildThis is where the file should be because it comes out of the runtime currently and petstoreWeb doesnt provide its own custom error JSP +Paul can you attach a patch with the change? Thanks +To bequite honest Ill be glad if anybody else more familiar with the XML Schematypes could pick this up and expand +let me know if theres more info to be added to the doc or if any of the current text is incorrect +NET versions and Mono or not +However if the UI download link wants to use a runtime ui widget that invovles larger questions about whether we need an integration plugin or not +Im looking at this patch +Yes I tried it and it didnt break the compatibility +fixed in just upgraded from axis to +What do you think about this version Konstantin? I checked out hadoop and miraculously it compiled and even formatted a NN image for me New unit test reproduces that image and checks that upgrade continues to fail with the correct message +That may cause future support issues +Most IE users should now be using IE but we can still leave it visible +for prev version not for prev +bq +Sorry I have made a mistake in my environnement description +jar with the new changed wiki pages as You very much +Issue with excessive logging in DomainNode carried over into solution isnt to turn off some logging +I found the root cause here its a rare case due to shared doc stores +If we cant implement this without major changes then I dont think we should do itI wanted to know your opinion first but I came to a similar conclusion +Although this should work in AS you should use get to obtain a +It seems that this will happen whether you roundup or rounddown +Sorry typo +But if is only meant to be used during transition this may not be a serious issueIm beginning to think that should be true by default and that the default permission on upgrade should be +Patch was generated against +forgot to say that if you take a snapshot during the RS upgrades nothing bad happen +bulk move of all unresolved issues from to +as a workaround during this is fixed is there any possibility to deactivate automated tagging when doing a release +So it looks like and Mojarra really did have the same bug in themPerhaps you could confirm and notify the Apache guys +hedwig console needs to use bookkeeperopen to read topic messages +I left them public just in case a client wanted to introspect them or use them further. +Ill have to stop at some point but if I run into the bug Ill post it here +Remains resove the conflict with the chromattic APT +Those new patches perform better and they also solve issue . +Why is byte less than block size? Do you mean chunk size per CRC or something else? I cannot reproduce the failure for file size or +In your case you should see a log statement like Hibernate +Also Solomon volunteered to apply this over lunch +I suspected and now think more strongly that there is a language confusion issue between us and the last few comments +sorry I uploaded the wrong am ok with relaxing the constraints +Let me know if Ive misunderstood +Cookies can easily be reset binding it to accounts makes it harder to manipulate scoresAnyway I think this fixes the original problem as users can no longer give a comment a rating all by themselvesIf you want to discuss the intentions and purpose of the ratings system I suggest we take it up on infra instead +Thanks Isuru for the contribution. +Id suggest using yourown abstract Action class if this is a reoccuring problem +bulk fixing the version info for and all affected issues have in all issues not touched in a month to after release. +If you think this is not good we can change it back to your proposal +Pushed the txnId support to master. +This patch sets up build environment correctly +Ill report back +If so we will not do autoConnect for STOMP as I dont want to encourage our users to put user and pass in their JS +The only other way of ensuring that things are consistent would be to run the whole query without offsets and limits or do all this in a long transaction +Attached the next patch and password protected word file that is used for testI added test cases for ignoretruefalse casesI think this is ready to commit +Add the local flag to all the map red tasks if the query is running locally +Thank you for your advice HarshI keep in mind +Theres no real reason you need to upgrade if using STOMP +Ill force a cleanup of those artifacts there + +Same activation via System Property defined in the parent and in the child +I will wait to commit changes so that I can commit changes for all subtasks together minimizing impact to others environments +I want rollbacks when one operation triggers another and on failure both roll back +I dont want to change the id generator of the mapped superclass there might be other childs to the mapped superclass I dont want to change the id generator for! I added the corrected test version with more childs which better describes the works but not in my scenario +Fixed in +fixed patch +Is there any reason why we shouldnt be a bit more aggressive with the default value especially because we can turn off speculative reads altogether? Please fix typos in the comments +There are test cases +Cheers Marek +I did add it to the interface though in case we need to regenerate now or in the future +Thanks for finding this problem Myrna +Otherwise Ill aim for mirroring current functionality as is +but what do I know +The default configuration is to relay mail from localhost +Please use the to removemove files and apply and to modulesluni and respectively +Committed to and +Here is the actual patch +Thanks Luke +I posted a fix for which is related +Would be super if you could +Revise patch to allow deb package to control path structure during build time +Great investigation Maxim sounds like you are on to somethingIn an ideal world wed have an automated system test which did a kill on a running broker and tested it recovers fine and so forthIn the meantime I wonder if you could give us a tarball of the journal files created after a kill and we can use recovery of those files as an automated test case +Patch to not escape amps a second time +patch from from from committed rev cdcddbddbabcceecb +Simple patch that adds get formatting is for users so it would be better to add this method elsewhere if possible +My env Linux Centos OS Groovy on Eclipse have the same issue as well +Is this in svnI svn upmvn clean installbut I still get the same result as originally +Thanks Eric for contributing. +The solution you proposed offline +As discussed in IRC patch to read Counter Columns through CFRR +the tests also check that if you change the parsed file the macro definitions changeIm going to leave this issue open theres a few changes to the docs that are needed +Just saying +Since the test is stable in the past runs I also run the test case times locally in a row and no failure +is for branch +Applied thank you! +Will be installed to Hudson with the next update +Please confirm it has been applied as expectedI also removed this test from the exclude list +nm was looking at the wrong source dir +I hope I have not forgotten any project +Attached patch for this issue + +you should try war +Including the log from a run on where I just added a message in failing asserts which prints the line it is seeingCould be some timing issueI did clean everything + r also caused error forIt looks like a nest class that is translated from an inner class still refers to the type parameter of its super class which is also an inner class +I am very unhappy that we have to write a C client test to reveal a bug in our Java core +bulk defer of unresolved bugs to +Only the default max is respected +bulk close for release +The Wiki comment addresses my concern and was illuminating +Then i removed the stopwords and finally i filtered themI have worked with this analyzer in some projects +Bulk close for Solr +Sorry for beeing persistentI finally managed to get the latest g driver from the link mentioned in the last post +ram buffer size is set to you capture output leading up to it +Also in the CLI If both appStates and appTypes are empty we show running apps for all appTypes If appStates is empty but appTypes is not we show running apps for all the specified appTypes If both are we show apps based on those filters The code to validate returning of all apps is duplicated and is present two times +to make it easier to track the various bits and pieces Im gonna create a few work Jonas its looking good +Only on secure deployments an error will be thrown for restarting namenode +Attaching the Patch for Nicholas I am working on it +I will take a look +looks good to meOne nit Id recommend either removing the deprecated name is never used from the comment or changing the name of the parameter from n to name +Hiyes +txt +Since hg promotes usage of lots of local development branches I dont think it would be a good idea to push them everytime we mvn releasebranch. +bulk close of all resolved issues prior to release. +We no longer have the separate jboss sample files +Attached logs for the deployVM job which took almost mins to complete +Many thanks for your patch Mirko IVe applied it to trunk +fwiw I also ran the test on my windows machine with ibmibmsunsun and this test passed in one run all ok +To recreate it I have started servers killed one follower after a leader had been elected and use top to observe cpu m not sure how we can pinpoint the problem from the logs +bug for at on versionbrowsers fire fox ie safari google chrome List Shuttle Ordering List Pick reproduce +I then took a screenshot of the existing theme for the site and changed some basic colors and created a header using the new logo +then only volume attach can be executed +Marking patch available but lets be careful about commit since well need a version bump on Interface before we can add this +Committed revision could you please test and let me know if you find any problemsThank you for reporting and providing a is confirmation that the committed fix has resolved the problem +Hi SaschaIm very busy at the moment and did not get a chance to test nor have a deeper look +Didnt think that G may affect nonTyped files +set false to disable the confirmation dialog +If the config specifics thecharset all output on all systems is OKIf no charset is set in the config I claim that if is used Ithe right places all should be well +Thanks Brahma. + +Nick no update were performed between the two launches +Thanks Paul +Waiting for final releases of the participating extensions extension are available as final releases +Kihwal and I explored this approach about a month ago but abandoned it due to too many problems +Committing to trunk +I currently dont think touching Query to accomodate a string parsable representation is the right approach. +The patch +The Spring XD option now appears on the ICLA website +If it is a node is being errantly marked down in or later you can try increasing the configuration directive and see if that helps +Stuff like automatic dispatch for certain query types might need some cglib magic or at least req +At first glance it looks good +If you happen to have other on the box maybe you can try with an older JDK release +logs say the session expiredPardon when I spoke with Camille at Hadoop World NYC the verdict was the patch on was not correct and this was the basis for does not propagate session expiration indication +Submitted the updated patch +I updated minOccurs in the schema and updated lib +I got my resultsPlease make all timeout and policies configurable +Fixed the use of the result to ensure only the absolute path is usedI also adapted the test case to have tests for both situations of the return absolute and relative FilesFixed in Rev +frjayasoftivyutil matchesBinary file +Thanks for your fix +Then again since Im fond of recursion maybe I would have stumbled onto that myself +Hope this is OK with youIve tested with localhdfss filesystem +Did this happen yet +This is patch available and its blocker +I changed the default phase for to compile because it doesnt actually generate any sources it directly generates bytecode +That works great manually +Made this class implement +Is the comment about the sort memory being used in an observation about the original test program? Yes sorts currently require one array entry per sort per transaction +batch transition resolvedfixed bugs to closedfixed +think another advantage of putting it in is that its more flexible than putting it in language +Patch looks good to me +Removing from +we decided to change the default log level for the application and exception log to info this change will not affect existing installs where the already exists only new installations +rowkey caches are made backwards compatible and save only plus tweaksfixes from Sylvains second comment +and it misses archivebinhadoop archive sorry to be asking this but is this document supposed to serve as the main doucment and points to the others in case of specific information? The problem I see with this is that it is not a complete documentation +It would be easier to review the impact of this change if you included at least a few example subclasses but in general I think this is a good idea +I did not look at the patch and was thinking it was only to relate a fixed asset to a party +This checkin is in closing as such. +ill commit the fixes soon +bulk fixing the version info for and all affected issues have in +I agree that it makes sense to have a nameHint but why not a nodeName where I can exactly specify the node name without any filtering +When running under the debugger I see that the protocol returned on the URL is bundleresource +Sales people dont have to go in manufacturing componentThis is the very reason its better to have a quick overview on production status +Ive changed name as a groom server +Deployed. +The patch for the dependent issue fixed this issue as well +Actually I forgot to say that Ihave this issue only in vi +Yeah I might have marked this as WONTFIX since it seems completely wrong for them to report what I assume is a version number rather than a Java version number +It has created a regression against so I think its probably fair to change its API in +Im unassigning this issue from me until I can replicate Steves other item if the resolution is not all happening behind the scenes +Oops the new shuffle method does not use the RNG supplied as argument +IT yet to be added +Some like do not have this property set and work fineSo we need to find all binary files with svn set and delete the property +Revision in trunk fixes the issue +Sorry for the double post +Move option earlier +Its a very specialized job that would benefit from relaxing this constraint. +I was under the impression the Sitemesh filters only executed jspvelocityfreemarker templates after the request had been processed by subsequent filters +Tag needs to be moved in order to include the last updates +If we change Stop to also mean loadfalse then we need an Info dialog telling the Admin that the stop operation will be preserved across server restarts +. +If it doesnt the framework has no way to know what the item was nor anyway to determine what boundary the partial item would follow since it would have to be entirely an application concern +Thanks again for the patch +thanks jeff works like charm +Very quick addition! Does it make sense to add the parameter resultType to the constructorThe returned value is a list of values within square brackets even if there is only one match it is not very pretty in if a simple string is expected +I have tested and the code and believe it to be ready for commit + +I believe its important to address this as a top level project +Verified. +Ok +Im going to go ahead and mark the bug in the debian bug tracking system resolved +Id like to commit this issue this friday unless there are objections or other comments +With this assumption all interleavings should result in an error writing to the walog or the GC seeing the walog inMETADATA +As you can see from my previous comment that trunk workspace dir was not present so I was unable to upgrade anything there +A release candidate should be available shortly +if the example would produce a message that list has no for method +Thanks DimitriI understand the existing methods in that class return an empty string if any of the parameters are null but like I said earlier I believe that is a bad design pattern +You could set up sites for years then suddenly have one that doesnt work the same and nobody knows why until the source code is analysed then it is going to be WTF this thing counts the number of letters in the TLD and makes a distinction between UTFASCII +Committed to trunk +AlexeyThanks a lot for fixing my patches! Its a real pleasure to work wich such skilled professionals as you ans Nadya! a bunch of thanks guys! Nadya if there are no objections you can apply patches and +Reopened +Also moved the out into to add spring into its own too +needs annotation +we should minimize the last minute changes +vsl file that enables links after each section on a page +I know that I can turn on compression in Tomcat but only for the whole Connector +Id say its not your patch +Im not sure if its nice to spread this logic over the client and server though +project +Just an FYI that Ive added a few more questions related to the side in the wire format wiki +Miinimal stuff based on standard Spring practices that just ensxures that spring will work +I know this a required feature but at this point we should add none of new features +patch looks good +Looks very nice package restructuring +Are we sure this is the right thing to doI have no idea at all +I have some time tomorrow afternoon to work on the test +The javadoc tool appears to have generated warning messages +Closing issue +Charles Moulliard put me in this direction today as he is planning on a Devoxx presentation on Camel with websocket and he notice this JIRAKnowing how big websocket is going to be I would loved if the core team reacted faster +Can you attach a test case that reproduces this issue +project file and if project is faceted +Denny added a nnAdded trunkdocumentationwhatsnewdrools Drools Component N in the drools whats new folder +Thanks Claudio +Let me know if you need anything more from me making hbase tlp? Thanks +fixed rev +Anyway I had a workaroundOlegthanks for the help +Committed to trunk +Thanks +Part of a bulk update all resolutions changed to done please review history to original resolution type +Evil +These changes allow one to group by complex types +If you provide steps to reproduce we can have a look +So stay tuned +Patch applied no review comments +Maybe Im not following completely is the idea that webhdfs would run on the same port as the NN web UIIt seems crazy to me that httpnnjmx would be the JMX servlet but httpnnjmxopGetread would be a file at path jmx +The have a couple of methods to close resultsets and statementswithout throwing any other useful methods we can add +This has been introduced for backwards compatbility The old signature had in it the new one has +is not a valid response code for the state that the server is in after a successful db DELETE The request has been accepted for processing but the processing has not been completed +we could do that but what if the contents change? What if freemarker or GWT get or a project ends and no longer needs to build? Are you OK with these flow jobs being renamed to reflect the fact that new children are being addedremoved +Thanks Michael +How about instead of +ve just committed this +next time you talk to Peter Kriens please convince him of some more generic generic way of doing this +But youre describing something similar +Indeed the instance must be notified when a bundle is leaving regardless its state to release all internal references on the leaving bundle +Note that the changes are not backward compatible once applied the trunk will no longer work with earlier versions of hadoopAll unit tests and about tests ran successfully +We perhaps dont need to annoint one +doh! Reverted my buggy fix +But who knows what the future may bringSo I think we have to use a of Harmonys BAIS to ensure JRE compat +For we may just have to allow Bag of tuple of arity how would be specified +manysort patch committed. +Committed to trunk +NET specific feature that is being used +Thanks +Here is the patch which fixes the issueHTH +Remove a bunch of out of date documentation that no longer correctly describes Hadoop +has been updated with commented out methods highlighting the missing manager closes still to be implemented as JIRAd in Rob can you review these changes please +All the data is sorted in one direction only and all Scan objects are written to go in the datas natural direction +The functionality should be the same but I may have goofed +I tend to dislike group accounts really ever knows what is being used for +We use a new session in the listener and still encounter this bug. +but the harness is NOT recieving this and is saying it failed the test +It cannot be the cause of the issue but it is a contributing factor to make things worse +Attaching patch with review comments incorporated +This looks pretty good +The word stuffs isnt a correct English word +I also have Performance Evaluation number from running on a real cluster which meet my expectation for improving the write throughput +Hi Julien please have a look on +If you dont install the extension you have to add an ignore action for the and your entity classes wont get enhanced automatically when you edit a source file in Eclipse +Attaching which must be applied on top of the a patch +Thanks Claus and loannis! I will try out changes. +I applied through locally and Ill commit the changes as soon as Bamboo clears the HEAD +This patch changes the decorated keys to be stored as an object with a and a String as member variables instead of both values in a String +could we use ivy for resolvingretrieving +This time unit test actually passes +I dont think we can do anything about this one +It was necessary only a brief explanation since the main docs has been produced in +had an amendment for trunk to have parity since though it was a backport there were some refinements made during the port +If you find further issues please create a new jira and link it to this one. +Fixing findbugs warnings +Update versions for release +Yes Brian I think we should +Add a depth variable to keep track of recursion level +KoushikPolite nudge to see if you have had a chance to triage this yet + implementation of +set proper resolution was rejected not resolved. +This is caused by the general JPA assumption that JDBC Connections will always be retrieved on demand not necessarily held for the lifetime of a transaction +Im not a committer unfortunatelySomeone among the committers should review and apply this fix. +Repository Revision Date User MessageASF Tue Sep UTC genspring Support Java Runtime use Ivans proposal to add a profile for jreFiles ChangedMODIFY geronimoservertrunkframeworkconfigssrcmainetc +Changed location to look for HDFS jar files +for the change havent reviewed the patch though +I dont know if it is a concern or not +Ill commit it after Hudson has checked it +Is there any way to get the earlier behavior enumerations and value naming without reverting back to the earlier Castor? Perhaps some sort of compatibility switchA quick response would be much appreciated as all of my software builds have suddenly broken as a result of this problemThanksBob +Ill close this one when I see the other issues. +Is it possible that it is a product of doing phonegap version upgrades? +I like the mvn jetty a lot +The installers for JSDK and JSDK dont appear to work on amd linux. +How far could you get by using a list function that filters the rows it emits +It is simple to declare the explicitly and avoids the obfuscation which would be caused by further namespace additions +With two ingesters and the native map Im seeing K sec +Any comments? Is the test ok +Existing tests are sufficient runs fine on my desktop ran it several timesRunning Tests run Failures Errors Skipped Time elapsed secResultsTests run Failures Errors Skipped believe the failure of is not related to the patch +Edit release note for publication. +Happy to do that in a later JIRA if this goes in +Whats the status? +With the latest build none of the JARS isshould be copied over to the AS directories they should be bundled in the application WAREAR +Indeed looks like youre right +If youre not sure it may be best to create a new workspace at this pointOnce you do that you should notice that theres no jaJPindex +The reporter confirmed this in the previous comment. +The WSDLS and XSDS are from the OASIS spec so we cant change them if we want to be spec compliant +Enabling by default is fine +Ok then we need to look into this this for Denis or you Victor +Really this is trivial +tools and wiki +Per review meeting +I can confirm that the suggested workaround indeed solves the provided test case as well as two more I have in my setup +My fix for recursive calls at does not work now +Thanks Jitendra! +Yes I try it its available +the only specialization is in using as in using and using using probably could it though +bq +I submitted this pull request three weeks ago now +Attaching an example project to demonstrate the problem +Not for commit but shows the changes that will be needed for +Yes were saying the sameIf a company has a version like tagged as certified I would have other concerns the latest security fixes were not applied afaik basically it has reached EOL +A is now used to ensure that does not proceed while the resume method is in progress +Im attaching two project directories +list item class is deprecated it has been replaced by so the section should be updated accordingly +Duplicated with +Proposed patch catching rescheduling the start level change catching other exceptions just continuingDepending on the cause of the this may in fact cause the startlevel switch to be endlessly done +Things intended to be reusable that are not in fact reused tend not to actually be reusableWhether this more primitive layer should be a library that we use to build mapreduce daemons or a service is an interesting question +I looked through the existing documentation layout quickly and I am wondering if we should include a section on how Derby handles collation in Developers Guide +any progress on this +so its almost as pure +If the database is shared between instances then everything seems OK +However Ive made many silly mistakesmisunderstandings in pig before +So Im a bit confused of what your issue is and in what branch you need to address it +There is nothing else of value in the test so I propose to drop the test from the TCK +Duh nevermind on case Boolean AND query would work for that example +If you want to run Zookeeper youll need to change the binding in. +Please try as it is the most recent and most complete patch proposal +According to several tutorials I read is defaultI added it does not solve the issue +Merge the patch into and branch +If we cant fix it well get it in weeks later +BQ just checked the code the error is already fixed. +I created a test case that reproduces the problem +james is gone +I would expect heavier methods like join to slow down more than map under profilingNot also that while Arrayjoin does show up at the top of the profiling results I would not expect its self time to be substantially higher under profilingIn any case Ill poke around Arrayjoin a bit to see if theres an explanation for the slowdown +Please close this bug thanksThanks for the patch Filippo + Username sigbur +I have another document that I will upload once its been checked for confidentiality +if thats the case how exactly does one go about getting the complete doctype declaration from the cursor API +What is your error recovery policy? All replicas would be of different size can you pl point me to the code that handles this inconsistency? I cannot find this code in your patch +Is this something that is being looked into? I am also very interested in having this feature rolled into the GuysJust reviewing old jiras +Going to close this again reopen if more network issues or missing jars occur. +There must a different way how this should be done +There are two comments that need to be addressed so cancelling for now + +Unscheduling due to inactivity +antlr shouldnt be since its a client side dependency +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Which version of Cordova does this happen on +If you guys have think I should change anything Ill happily rework my structure +See here you test eclipsemeclipse with the or +Mr +This is now not within the scope of the Spring core anymore but rather the problem that the Spring Web Services project +The issue is that the native Java types and their counterpart use the same names +Updated to +Code that renders glyphs even when they start with rendered with modified code note the hollow bullet at the lower right +and the set to Resolved Fixed again to correct resolution which was lost in Bugzilla JIRA conversion +If you generate an ivyreport after your ivyresolve you might see where these logj dependencies are coming from +If the app isnt using it then I dont believe this even system is used at all +SVN update and build seems to be working in . +Changes to both issues cannot influence each other locally I do not reproduce this failure +When we get an exception like this we should error out to the top level and basically restart the recovery worker as these have very little state which isnt dependent on zookeeper +committed in r +Im definitely not interested in that +Thanks BruceIve adapted the config and you now have it building on Ubuntu x Python cmake java jdkTheres an in your irc channel and notifications go to your dev list +Closing all resolved issues for. +First I would like to correct a mistake about conffiles Debian Policy requires that they are specified in a special file inside the called conffile +EdwardIf you can figure out why it is happening +Please let me know if there are any issues +If we remove the value of the RESTURLPOSTFIX property from the axis this works as expected +Absolutely Not having a calendar component to iterate within a day renders this calendar component useless from a scheduling perspective +In one more place javadoc is used without a separate target +with modifications. +Weve already addressed this for the HTTP invoker client side but not for its server side yet +But none of the other attributeTypes have special cases for transforming to xml +To avoid confusion +Fixed with regression tests. +million is too scary +forge it takes nearly seconds +But even though the hash table is randomly accessed Im not sure if MRU will help with performance here +include Online HTML Single page HTML PDF print version! Cool to check status in CVS before closing this +Could some one review the patch? thank you very much +Found some issue with stale left over files in the shared directory +Should it have +Should this be marked resolved? The linked pull request doesnt seem related +Will be a bit difficult because of the syncing but Ill try +Can you give me a hint or point me to a link which is telling me what I have to do +Issue closed +Raises the question of whether should be wrapped around any String +We did some enhancements on all shared branches so I deploy them +Some processors allow this use ofa variable +committed to trunk and. +I dont think this is setting a attribute in the manifest section in the config seems to have no effect +Added thread pooling to the thrift rpc client +Now I am trying to understand the code better +Can you take a look at the failing testthanks +Would it work for youYou may also want to look into +bq +This needs to be ported to branch as well +Upgrade to new network name schema need to do the following steps a +HansonThank you for the patchCan you provide unit tests with the patchThank you GaryPlease find attached for the unit tests +Then I create a little test program and tried to rename some groovy parameters fields etc +Symlinks are used for distributed cache and this will be inefficient +Reopening based on feedback +Attaching small patch for this +Or we need to document the expecting format is s not s nor secThe format is documented and I prefer strict units but feel free to make the matching fuzzierbq +I went with the DilloIve attached a zip of the logos +bulk defer to defer to +set release version +svn mv srcjavaorgapachehcatalogpigdrivers srcjavaorgapachehcatalogpigdriverssvn mv srcjavaorgapachehcatalogpigdrivers srcjavaorgapachehcatalogpigdriverssvn mv srcjavaorgapachehcatalogpigdrivers srcjavaorgapachehcatalogpigdriverssvn mv srctestorgapachehcatalogrcfile srctestorgapachehcatalogrcfilesvn mv srctestorgapachehcatalogrcfile srctestorgapachehcatalogrcfile checked into branch and trunk +yup +We test the create and on all DBs so Im quite confident this wont be the case though +sighAs to who should do this if you see the problem and know how to fix it please do so +You can look into the Artifacts build history +Still trying to figure out whats going on +Those peaks contribute to what causes the gets to have a high average timegetClosest is the biggest contributor to the time spent in getI have data for methods called by getClosest but have not graphed them yet +This patch tries to sketch a solution +So instead we just check so no closing happens when its the same t come up with a good way of testing it though since no external properties change based on this except for the output to STDOUT +With verify I meant that I was wondering whether any subclass will have a different implementation than the one in the default policy +This issue cannot be replicated on either the branch or trunk +I think I can make a couple changes to the patch to achieve the desired results +also issue is the same as this +I can confirm this passes my internal tests and my production app now deploys +How about waiting for and committing this and that consecutively +Removing some artifacts from attached files. +It seems to be down indeed they were just talking about it being down on the osgeo discussion list + +For multiplexing to be feasible and remain async the server must be able to determine what each received packet is rpc call or sasl auth call +If you see a significant improvement feel free +If this is normal it seems that it shouldnt be a warning or if it is fact potentially problematic it would be nice to know how to resolve is strange is that implements so Im not sure why the beanFactory isnt getting set first +Are you planning on implementing other languages? I think this looks fantastic +But Id still like to see my previous question addressed before starting to analyze things +Wow it is a real battle to report the duplicate today +This should be safe to add to +There is a cleaner way of doing this wo copying code from Patch to to remove duplicated code and use the impl +It would be better to point it to trunktestconf +Looks like an attempt was made to integrate a fix but it was done incorrectly +So if one reloads a spellcheck index it seems that theres always a chance of messing up a previous solr request thats still in flight +Well probably revisit this once is out +You can scan anything that is in your classpath +All looks green now +Triaged by EAP PM team on Sept +Ive added Justin as a watcher +Unassign issues assigned to Nate +When a cascade of choice constructs is detected then determine which paths are duplicated and which ones arentFor the duplicated paths process them using the merge function to determine the contents of the repeat construct +The test failure is unrelated to the patch and happens on trunk +I believe the current behavior is correct with respect to held cursors +Then I can get the method in the subclass +Im cleaning up the Accumulo formats to account for the updated checkstyle requirements +The configuration can be made persistent either using spring configuration or store a database +Marking closed. +stack can you help give it a try to publish repository to your account +Patch applied to trunk +Committed at r +Thank you for your improvement +This has been implemented and is tested in the multi tier integration test. +Then cut it out remove the ui and the dead code +BTW the default ctor instances could be cached as well but that is definitely outside the scope of this issue +bulk close of all resolved issues in preparation for release. +Heres the stacktrace Imseeingorgjbossvfsin getChildorgjbossvfsin getChildcjrubycurrentlibrubygemsgemslibvfsextinchildcjrubycurrentlibrubygemsgemslibvfsextinvirtualfilecjrubycurrentlibrubygemsgemslibvfsextinfilecjrubycurrentlibrubygemsgemslibactionviewinfindtemplatecjrubycurrentlibrubygemsgemslibactioncontrollerinfindlayoutcjrubycurrentlibrubygemsgemslibactioncontrollerinactivelayoutcjrubycurrentlibrubygemsgemslibactioncontrollerinpicklayoutcjrubycurrentlibrubygemsgemslibactioncontrollerinrender +geotools file which I duly created but the simple jdni code in the end keeps on using the one in the classpath instead +Please review and verify fixed +You can close the issue +Ive successfully tested tehpatch with Patricks tests +This will also give us some time to release the fixes in a community version first +Committing now since its a pretty simple change. +I missed this thing +Add a unit test for send JMS as by it timeout +Thanks guess we could consider adding a Comparator strategy +Hope this testcase helps +batch transition resolvedfixed bugs to closedfixed +Hence prebuilt OSX alias file +Hi MaxThank you very much for the fast response +Hi Greg is this any different than the one you filed +in +Some spaces are not always underlinedIll leave this open +Do I read this right there is no way to specify a custom partitioner for use with ORDER BYIf so is there any other way to perform a total ordering within PigI will be doing a STORE immediately after the ORDER the relation will not be used again +These are available on the Run as +Lets get those comments editing back +Absolutely +Is there any maintainer who can commit this patch? Or the project is already dead +Thanks for the patch +Hello All Here is patch for this issueI test it by creating some orders with number of itemsI verified some entities like the +However it generates the same given id attribute for both the generated img element and its surrounding div element and duplicate ids are not allowedIm going to change this behaviour to add figure to the end of the figure id if one is given to keep the ids unique +JacquesI will try it with SVN sometime next week and will let you knowThank you +Patch applied svn revision . +Note that the descriptive text can theoretically be localized +Colin if a is used by multiple threads in an application the application is responsible to synchronize their calls since is not thread safe +Thanks +To avoid a lot of full scan collect the data to new structure +Yes this should be a feature request for Allow arbitrary bbox not limited to circle +I am closing this issue again +James can you attach a reproducible test case to this issue? The description does not seem to have the function definition and schema +The on core tests is unrelated +thanks and regardsRoland +Yeah Im not deploying on JDK but rahter on a JDK with hidden packages +The only one that I didnt know what to do for is I think the code for that is correct as is +Note the issue is that existed did existed not exist continues but does think this definitly needs improvement if we are doing anything on the ant side +I failed to compile the attached test there is a missing dependency on classMatthias can you reproduce it against the release? I havent seen any public discussion of plans to make a bugfix release at this stage so it is likely that youll be urged to upgrade anyway either to or to when it is released + +Needs mysql user and password to run properly +May be a and problem as well +I think so +Test cases are passing now +Added patch. +Need to investigate +Thanks. +We should really patch this in given the number of universities etc that still use hod +A region is a DC a zone is a RackThe metadata is so itll have to go somewhere in Config +This is a duplicate of +I hope I got the problem right +Reopen if any problems appear. +Number is the wrong color and the color of Annotation is to dark +One question +Committed revision +Many thanks would see other SF projects that are already being be fixed now +That should be fine in most casesIf you return a and then change this to return a you will break code +This is for Jonathan Hsiehs questionThere are differences compared with + +It would be good to have one to prevent any regressions in the futureThanks! +Updated programmers guide James I now realise that this has not actually been properly fixedThe link to the downloads page still actually redirects to the main project pageThe maven download link is still referenced as the repo even though Byteman is actually available from Maven CentralUnlike the other two tutorials there is no direct link to the maven plugin tutorial instead the reader is referred to the docs page which is less useful andThe opening sentence under Obtaining the source build tree has a redundant extra verb +It would be so much better if Maven and the ant task could both use an environment ve written a test case consisting of six files +Work for this issue has been completed and in production. +Hi Charlie Sorry for coming up so late stuffed with work here +might also have a more advanced tool similar to DtraceI doubt that the file leak is in FOP code +I was able to reset my password on cspserver without any exceptions showing up in the log +patch verified +See also bug for this idea +Enjoy +just comment when its committed +to apply +need resolution on this before the first release +Move issues to next release +Thanks +especially since we now have a datanode directory +Had to extend some of the regular expressions and stop when reaching annex in end of message definition files +Ill mark this as resolved +No the enviromental mix was causing jruby from svn using jruby installed locally +Patch attached to also fixes this issue come back again It was solved in issue I just applied the patch in revision rThanks for want to mark this as Fixed? +open separate issue for created SNAPSHOT version of the Spring archetype is using now +bq +Note that this patch must be committed before and +Updated patch to reflect latest state per Francis review comments +patch was the latest so I didnt see the test +This needs some more improvement +As of version also general asymmetric matrices are supported thus there may be complex eigen valuesSo the variable imagEigenvalues has some use now it is used to distinguish between real and complex eigenvalues +it took to long to fix shortcut creating problems so Im moving this one to m this works great! Thank you +Remove from fix version +Currently its eating it up making it impossible to see what caused the. +Also youll want to read the related jiras to become better familiar with the changes. +Added a one line change to to log instead of throwing an exception if the preferred leader is already the current leader +I am stuck with how to write a test for this thing +I just tried it now and following the instructions in the readme file I was able to get the web app to work +At the core this issue is asking for a new persistence strategy that retains the values until the user navigates away from the page +Change scripts to construct class path properly +In theevent this does happen however the master will tell the regionserver to restart that is close all open regions and flush its logHowever the master should defer processing the servers log andreassigning its regions as the server may still be in the process ofshutting down +We got this error too in production yesterdayCyrille i checked your did you forget to syncronize clear or it is by purposeI also cannot find the Thanks +Any suggestions about API design and realization will be welcomed +The API of the throttler is pretty weird +What am I missing? Ill add the answer to the wiki as well +Project despots are supposed to be able to subscribe new addresses with no confirmation +The issue which appears to have caused this was committed to SVN on got released on +bq +Nice work Laura +committed to trunk. +the whole replication flow will be affected are used everywherein what sense? I didnt quite follow that +To be clear it is only a full build that triggers this behavior +from me +Any reason we couldnot have in couchdb upstream +II have the SAML callback handler factored properly now but I am still working on adding support for the Authorization Decision Statement type +Close all resolved issues for. +Thanks +I dont have a huge problem with this its just that using requires the delegate to be registered separately +We dont do anything that could cause this +It seems finbugs and release audit have also been failing on hdfs before this patch +Whatever mechanism we choose to get these details the source of the data should be ultimately something which the schedulers generally use for making scheduling decisions +Did fixing help this any +I tried running and from outside my home directory and they failed without the patch and passed with it +Right sorry about that im using M at the moment +I have no idea +bq +It actually means that The illegal characters in and need to be fixed The illegal characters in are there on purpose for testing reasons so theres nothing to fix +For some reason the osapi javascript object is undefined when the initContainer function in is called +I can certainly see a MDB on the Stack Trace through the inflow +Please review again +is there anyway to fix both scenarios +However I am not sure I am comfortable making this change for w all of the testing that has already been done +like i thought this method is breaking the input filter making the enable functions useless +Will check what MRI does +Release note setting +Even acrobat cannot extract the text from its a custom font encoding +it added to s just a pointer saying that probably the two can be fixed at the same works now and has been enabled for the full IT run +Assigning to Alejandro +Attachment has been added with description Paint corners of PDF borders with a diagonal border if colors has been added with description Acrobat Reader pixel error with naive implementation of border as separate rectangle patch may be applied together with my patch submitted toIn fact they compliment each other well +Expected result Simple string both in the browser and the s logActual result Simple string in the browser and nothing in the log +This change was done as of +I suppose we take Ingres version Linux as our test DB +Well I do plan on working on this but I dont have much free time to put towards it and Im not familiar with the codebase yet +This fix has broken the scenario in which we provide only the local name +needs to be backported to +Kishan can you comment on this issue? it was changed by you +Committed to trunk and +out +StefanIts working now +One might say the test should actually not compare doubles for exact equality but for equality to the last decimal place or something +However there are conflicts when applying to the latest trunk +Looking at the constructor of it looks like advertises itself before it has completed initialization +Here is a modified patch that handles disjoint heaps as discussed aboveThanks patch also fixes sporadic regression tests failures on shutdown on interpreter +Moving out of +took a look at the javdoc findbugs and RAT warnings and am confident that none of them are related to this patch +in the future please attach your proposed fixes as patches +Still true on Glassfish +do reference in the put tag +Tests are obviously reflected by the automated build +This patch works for me +I just committed this +Sorry. +Broken site links are just one of these +I will check that in a second stepChristian +The stacktrace I attached refers to a OOME using the feature +Sorry Ive searched a lot of issue before writing this one and now I dont look at it I found a similar one +Thanks Christian! +Since AMQ doesnt seem inclined to distribute one I think were stuck maintaining it +Project IDReason POM not found in repository Unable to download the artifact from any repositorypomwhich repository is alpha deployed +Hi ArturoThanks for reporting this issue and it sounds awesome! Im definitely interested in this topic and will be sure to help however I canCheersChris +Added code to to register the authenticators specified in the configuration fixing Added methods to to make the generic header configurations available for the fixing Fixed the ear structure of some tests to include the security configuration as a separate tests are passing now. +Committed at subversion revision +Including your screen resolution will also be helpfulSimilarly what version of IE were you using that failed to parse the column widths correctlyThe tcolums table based layout is under major overhaul to fix many of the same problems you have attempted to address +So itll stay marked as deprecate through all of and be eligible to be removed in +Attached graph showing benchmark results of the latest stdcxx against libstdcgcc libstdcIntel C and gcc. +Id just like to note that VFS does have proper UNC handling Jason Greene spent a lot of time researching it and getting it right +Im curious about the status of the Pax Logging incubation is it donated yet +ukasz do you agree with my idea and did you end up with closing the connection? What else would you suggest +Test changes seem to cover the failing scenarios well +ill round this out tonight +I totally missed the point that it will always be with the installerYes in this case its a better idea to move the pages out of the web app and during the installatin have a default with the option to change the location. +I forgot in the last patch file +kidding me +Attaching permissions test plan. +Did I miss discussion on this +I can commit it upon test passing +Perhaps we can discuss a better way of handling them on the ml. + +committed this +So the patch should be ready for integration +Tagging for +Bulk close for Solr +Youd just need a directory for each conf and to wrap the call to cassandra with something that sets the conf variable +Hi Sven try to use instead ofI think should implement convertInput method setting null for empty strings + Moved setting up a connection into a class +Affects CQL and +Error Server server encountered an error and could not complete your the problem persists please report your problem and mention this error message and the query that caused application works fine locally +Note were about to start our th jira sitemap file +defer issues to +for the patch +MRI IRB with such +Sorry +Thanks Daniel for review. +In my previous company they opted out to purchase after deploying server to production and having to restart almost every week with a noticeable day after day +The next radical step is to remove all checks in the code whether append is supported +Something to eventually consider when creating the new UI +I feel this gets the job done though +Yo can control it the way you like with shell script +Anyone capable of manually entering bugzillaproduct would still be able submit new bugs +Also the member variables need not be public +The bean requires test case fails with the same stack overflow I originally reported +Ive also added additional test case for method +Ok sounds reasonable +I forget to register in my context so proxying did not happened +Thanks Suresh! Ive created to address the TODOIve run some simple manual HA tests in a cluster in which I forced NN failover periodically and kept sending client requests by using +For what its worth the following issues for the same functionality on other have been committed so far +Please verify +With all data in the blockcache as well +Can a committer see if the patch looks good to them and commit if so? Thanks +did you try to test it against that is the lates released? tested also with ack can you provide a testcase with a fix proposal please? TIA! +Do we have a regression test for the original problem? It would be nice to add one before this issue is closed +The idea is that we dont want to commit thing that break contrib projects and thats why integrating it in test and not another target makes senseI am fine this issue with Giri and just adding it to process though it seems that the end result is exactly the same you cant commit patches that break contrib projects +It is better to put new comments on old JIRAs than to put in entirely new JIRAs on the same topic +I hope it was not a problem +You can see from the output that for version correct version is being reportedSince you are getting Xalan Java it means that version of is located somewhere in the classpath +Closed due to the release of Tiles +However the chrome issues are resolved +Hes down to the last failing test so I think were in good shapeHow do I go about contributing to the tutorial +Try this one from myfaces core httplocalhostIm using firefox and I note this effect using firebug +Merging adjacent CDATA sections is something outside of the recommendationThis patch will do that +Ill commit this soon unless there are objections +Hi KevinThanks for the feedbackI hear ya on the annoying part +New patch you forgot solrcontribdataimporthandlersrcjavaorgapachesolrhandlerdataimportAlso I would rather change to debug and leave it +I am working on making a more clear repro +Do you think that I should add a paragraph emphasizing this +Could you take a look at the design doc and see if it jives with what youre thinking +Closing Jira. +maybe we need to start pushing the changes for improved gc algorithm into trunk after released +Old should be changed to something already approved +Well actually we do need to modify a bit in order to spread primary lock ownership within the cluster but we do not need to keep lists of owners as in +If you have a low memory hit ration you end up thrashingThe UPDATE usage pattern allows items to go to disk on an update +finally fixed +I thought only Windows lock filesPerhaps we should introduce the ability to switch off in keeping in mind that provides alternative solutions to the deployment problem +I see what you mean +As jbellis and I discussed on IRC we will instantiate the actual objects for Partitioner and and just the class for in +Closing since hudson has commented success. +I confirmed that the project still compiles using mvn with this patch applied +Im attaching and for you to check +No response assuming ok. +In a corner case when we are retrying qsub operations and the user hits a we dont break out of the loop immediately +Thanks for pointing this out! Fixed in be great if you could give it an actual try in this also be fixed in the branch +Anyway we can contribute efforts in making webdav into the hdfs as contrib based on existing patches please let me know +its race on statistics really important? because if we lock all those vars I suspect itll have an impact on the perfs +Will file separate JIRA for doc update if you agree +HugoCan you review this issueAnimesh +all pass +Thanks for fixing this so fast. +adding new patch +My mistake +Integrated on master and on. +Thanks! Verified. +I propose we remove the parenthesis from group and join to match order as that matches SQL and is less verbose +Can you test with the latest release +Pat Im not sure what youre trying to get out of this discussion at this point +Confirmed + +Conserning Show EL +Do you still see this issueIt should been fixed a while ago stack traces on the log are from an old version +Your approach would keep all of thisin memory for all templates loaded on a per thread basis +I meant it in a dispassionate way +Thanks Jie! +Patch to change the tracking URL to point to the RM app page when the AM fails +We should do a different JIRA for it right +Patch in modified form committed with other improvements and additions +The example that I started with did thatAFAIK thats outside the scope of this issue +One other factor is that strings are not trimmed in so there is a small consistency argument to make +Not sure if this is helpful Sean should be picking up the Dpropertyvalue arugments iirc this handled somewhere in the +Unassign for explanation from Jiri shows how to create a project that will always work when we dependent on Eclipse not supporting EAR libs we could adjust Seam wizard creation to do this then we would solving alot of EAP users problems +Fixed as of IT online. +That makes sense +The concerned merge commits seen in the repo are not present +The menu color will be correct when the skin for is committed. +I havent looked at the wiki recently if we have sections or docs for release Ill add this with a note. +It is our plan to move content extraction out of band +java Xmxm true ij version ij connect jdbcderbytestDBcreatetrueWARNING J Database testDB not created connection made to existing database insteadij update t set status where id row insertedupdateddeletedij update t set status where id row insertedupdateddeletedij exit +i want to think about the code a bit more but ive been busy +Bump +Since these inputs are generated in rd party upsteam libraries and since the upstream issue has been resolved lets forgo resolving this issue untilunless it manifests itself as a real user issue. +Its a only one plus in my opinion. +We encountered this same bug as part of upgrading to Struts nightly In short if the browser and server locales do not match the struts tag fails to include the dynamic part of the javascriptand leaves the static part renderedOur has a single formset with no specification of languagecountry or variant +Reviewing my solution for again I think that it would be better just to disable the when it is false rather than disabling any custom because the developers may use a custom but still dont want an error page to be createdFurthermore the default rethrows almost every exception anyway +If you would look into that it would be good +If someone can commit my earlier patch for bug we can avoid a messy merge later as this impacts some of the same code +This is certainly a configuration problem +Related to security framework wont be added to +Lets keep it as it for now +By the way I have fixed the in JTS to avoid this line problem and Ill be releasing version this week hopefully +This is fixed upstream by but there are many fixes in it that cant be easily backported +The failure is not related to the patch +I noticed this is an older jira but I can not recreate this in orI even tried it using the followingdont start directly with dont stop withwithout error in all should probably close this issue as resolvedThank you for reviewing this jira Heath +Michael have you report this issue to Richfaces team? So it will be very useful if they add these descriptions into future. + reviewed and ran the test code +Still have issues and I have to go +Fixed in. +Previous patch did not actually have the test class +This is also happening with upstream download +I think Im going to leave the sub issues in here for the time being as we knock these off please promote them to be full issues and make sure the jira name mentions table snapshots +Thanks for the test case +It refered to edge case testing the algorithm for getColumns in that sets max integer for octetlength when the datatype is char or varchar with length max integer However I think since then theres been a change in the max size of the charvarchar to size and the test case was adjusted accordinglyI am wondering if that code in is still necessary. +Not sure thats a good idea seems like the refresh policy should be an explict operation not called on every security operation just because it might have changed +Thanks! +Okay yea I see +i am running results Thanks Namit. +On the other hand it usually does work and the failure seems to be a failure not a CXF failure +Its a typical streaming decoder +interesting feature +Done +Problem is resolved in CP CR see attached s another problem the icon is missing will file a separate JIRA for that +Marking bugs with invalid fixed for release as fixed in. +Im good with closing this one +Automatically and silently generating length chunks doesnt seem like a big favor to clients +Bug has been marked as a duplicate of this bug +This is the apt way to knowRegarding Jerrys comment His use case seems different than ours +We should use the latest release to perform this action nowMany patches provided in Jira are already outdated and can not be applied without changing the patch. +pushed into master +Patch to fix this bug +Attached is the current zone file +You can switch to the installed one instead of the missingdefault one via Help Preferences Team SVN SVN InterfaceOnce configured I was then able to browse checkout from SVN +Patch applied +Gregory I dont want anymore +bulk defer of featurestaskswishes from to +Actually I dont think the file is to blame +So this Quick Fix may be useful in case of CDI project +I had to comment out the checking style code from the build as it was producing this error on the build serveroptgrailscruisecontrolcheckoutcruise The following error occurred while executing this line optgrailscruisecontrolcheckoutantbuild of build Feel free to the understand what youre saying but seriously software development was never meant to be this rigid +stack Prakash I will attach a patch in. +Regarding the use of Hamlet is it to be used by the proxy or by the AM? If the later it is expected that anybody doing an AM will have to use Hamlet to get things easily working with the proxy +Wont be done in core. +The real problem here is the tool selection is being given to the but the used by the Map is not correctly delegating to the +Alright I checked it in +For onlookers please vote for this issue if you feel it should be revisited. +updated. +The memory in the namenode is a premium resource and fewer the objects larger the number of filesdirectories it can support +A new patch is attached for testing +Thereis always the option and the spec explicitly allows that for certain rules tomake this somehow configurable +I test it manually on a real cluster +Ill crop it +moving the fix version to with revision +The upgrade was successfull after I found the SNAPSHOT repositories for all my problem is kind of a subset of the original mentioned by m pretty sure that some of his problems are solved now as well +Patch for this has been committedsvn svn Ive verified the repro and the new test cases in both the and branches and all looks to be correct +I implemented this for a customer a long while back but it got fscked up in the business details +This patch has been applied but given the recent conversation about supporting app servers in SVN and in the distribution its a bit dated nowI applied the path but with the ongoing work being done in Beehive it may become moot and need to be reworkedSo Im resolving as fixed but the issue definitely needs to be revisited holistically for SVN and the distribution +Support for in Damien thank you for this patch its really starting to look like the business Can you please generate a clean patch as followsgit diff trunk This means that we can apply it against SVN where it will hopefully eventually be committedThank you v much for the effort We are nearly there with this oneI also noted that you have a whole bunch of additional tests which is great hopefully we can retain many of them and possibly test them against other datastores +target +The more tags we have the more we have to support and the less time we have for core features and lets be honest we dont have a lot of dev time being put into Struts right now +hrm interesting +No +I have seen this tooShould we be throwing out something like until the acl table is in place +Which version of Xen +NotesThis patch does not handle the case of changing andor after the init method has been called by the servlet containerIll leave it up to you to decide if it should be handled +doesnt pass now in which seems to be unrelated +regenerated the patch as per andrews comment +please move back if they will be worked on asap for +You should use SNAPSHOT version or wait for RC version. +Yin Please see my comments on reviewboard +Thanks Navis for the update +Verified by Ilya. +The fix looks good. +Do you have any logs or anything for your case b? That one is more of a concern to me +I guess that is just fine +Chris was this the richfaces demo app? DO you know if this problem still exists +I herbey kindly ask to fix this bug +work around from matt prestrittoInsert overwrite tbl select from tbl where col fake value +Previously in constructing the list of links to detailed notes only the first text node from the summary was put into the link text +rev openejb rev +Patch is provided for version +If not please let someone know or file a new issueThanks +Thanks +Yes I cannot reporoduce it either +Merged thanks. +Seems to work for me try it out if you can and let me know if it resolves the issue +Just to make sure I understand you are saying thatWhen code dynamically instantiates an object via new or mallocs storage via malloc it needs to ensure that it was allocated and return some sort of error if it has not +Please close +This issue depends on LUCENE which is set to wontfix +Fixed with IT in adffabaecfb thanks for producing the test case! +In this case it was most probably the fact of running a script on a Windows host that pushed that machine to behave badly reducing the TPS to the extent that looked like Synapse was locked when infact it was fineIt has now been confirmed that with proper tuning and better OS environments this test passes without any Can you open a new JIRA and attach the sample request and any other resources youve used along with the above description +They are particularly challenged by node failures and graceful handling of those is a key requirement to scaling up +fixed in trunk +I saw it on Ubuntu LTS x with sunjdk +What is missing on this issue to close it +I updated codings style +Attaching the userguide from +is this on trunk or branch +The check does first open the file in rw mode to ensure that the file exists and then it in rws mode +I also cannot see in the jenkins logs which line is causing the problem +backported to and +I think the important point here is that we really should not be displaying Java stacktraces to end users +The is the original interface class +As I mentioned to you in this is a work in progress and so I have changed this to an improvement +Manually tested! +Ill just start working on this +Hi Cheng for running time could you exclude the system resouce factor +Karthick you should not delete old patches from the issue as it makes it hard to follow the issue +Did you forget to included it in the patch? Also all tests of failed +Committed addendum +The proposed plugin composition is a good idea +Fixed in +If messages get dropped for other operations as well you probably ran into wrong timestamp calculation from +I think in the short term I like this patch as is because I think developers will be able to port it to the new API and users will be able to understand what it doesLets get it in it is a huge improvement to the Chinese documentation and needed toobq +its already fixedthanksdims +OK Robert might be right we can only suppose how people used alldocs but we cant know it for sure +Just nitpicking and resetting to more appropriate WORKSFORME. +this was done anything else for should be in new issues +Spring was very nice for this intergration with maven was easy and smooth +I also wanted to comment that Id like it if we could include the visitor option in the jjt once we fix this +This is considered an incompatible API change by the clirr maven plugin which now fails when comparing with versionShould the next version been bumped to? Previous discussions on the version numbering missed the point with this issue +We will move it to a branch soon +Screenshot +appears to work fine +Duplicate of +Two should be provided one for value column and one for weight column +Hadoop binaries dont ship with Mac native codeAs Hadoop with Security becomes more popular this will become a bigger issueRegarding the name of the methodproperty we could change it to or something like that +Since the description of the command says List all the metrics currently configured for the resource I suppose this is acceptable behavior +These would still be separate quickstarts but the services should be in syncThe trick is that quickstarts are supposed to be self contained and easy to deploy and understand +a test case would be useful here this is a subtle but important behavior itd be useful to check it +Can you please clarify how did you launch Maven when this error happen +fixed errorconfirms it does compile locally not sure why not on jenkins +Committed this to trunk +Thanks for review Sergey +Adjusted the patch for the new trunk +It specifies what I described above +Andy put some ajax stuff into the framework and wanted a general standard pattern and I think thats what we should try to doI did this as an example or POC +If we dont handle this it leads to invalid memory reads +Just to be sure I followed it all if I do ant clean eclipse test after svn co all of the jar processing should happen as desired +new queuing structures smell like new features to me and we are in beta forbut thats just my opinion +The problem is that when doing it on a more global level you may introduce unnecessary project dependencies for other. +It seems you stumbled over which should be fixed in master already +The spacing seems to be off are you using tabs instead of spaces +Patch applied +Sorry got will post something this week +adding a seems that the workspace job in the marker dialog schedules another workspace job and joins on it +More tests more better +I just committed this patch +For your usage model simply use with the attributes property to acheive what your subclass acheives with a very similar usage +Merged into the master and branches. +I run some experiments with combining less often but there was no speed change +Thanks Jon +I am closing this issue since implement C for services handlers and RPC will not be done +Fixes a couple of minor bugs +I was mostly parsing it in as an alternative to grails configuration merging +would like to get RC out soon +That would allow Camel to work easier with compressed files and what else you guys have in your project + +Closing this as wont fix as at the moment they are different enough to warrant separate parameters. +Closing. +But the region is actually using the small flush sizeI think the test intends to use the small flush size and doesnt want split +This should pass the acceptDocs down so both filters are togetherAlso made also use the autodetection +Im sure it will be ready for tomorrow for your help +I had to change the pom to enable the tests +I see that I missed to do it in the new provider wizard +Added release notes Lars Thank you very much +i am too working on a patch already +Thanks Aaron! +Or does this mean if you use applicationyou cannot use reqgetParameter and reqget at the same time +Well see if any weird exception will occur in production. +Verified the patch fixes the noted problem on the Tomcat assembly when using Axis +Thanks! Glad the patch got accepted +Sorry I meant RC +Im going to look more into whether bind updated and deactivate methods can plausibly return a Map to affect service registration properties I think there are some valid use cases +This seems to have resulted in about a performance increase already +Thanks to trunk At revision +Closing issue. +Please check that the patch applied as expected +Fixed in SVN +was checked in long ago and on Feb +Unscheduling prior to moving to JBAS +This attachment is a fix for this issue +Im able to reproduce the OOME on trunk +Not very well Im afraid +Patch for trunk +I am not sure if I understand the description correctly? Does this mean there is already an implementation to declare process variables but it needs to be extended so that you can declare within a group activity definition? Thanks for any explainations +If the timezone column is NULL reads it as the default AmericaNewYork +patch applied against master +Youre lucky I did not move to Awaiting Test Case +Additional defects will be opened to describe other regressions +Is that the direction you were going in +Thanks Sergey +Maybe you noticed it already it is ? super V +Probably this is a very edge use case for you guys. +Its supposed to go to +So far I havecreates a zip file with a windows and unix folder that contains the zip and versions of the Apache release of at present +Ive reached out to the OODT folks Jordan you beat me to it will reply there +See and +See also +is added to directory +Here is the patch for the same +Sure but we could work here on that further and stabilize this classCommitting to the Camel codebase could be done later +There is also several renamings svn add trunksrctestsapiorgpunitwatcher trunksrctestsapiorgpunittestclasses trunksrctestsapiorgpunittestclasses trunksrctestsapiorgpunittestclasses trunksrctestsapiorgpunitrunner trunkpunitsrcorgpunitexception trunkpunitsrcorgpunitexception trunkpunitsrcorgpunitwatcher trunkpunitsrcorgpunitwatcher trunkpunitsrcorgpunitwatcher trunkpunitsrcorgpunitwatcher trunkpunitsrcorgpunitevents trunkpunitsrcorgpunitevents trunkpunitsrcorgpunittype trunkpunitsrcorgpunittype trunkpunitsrcorgpunittype svn remove trunkpunitsrcorgpunittype trunkpunitsrcorgpunittype trunkpunitsrcorgpunittype trunkpunitsrcorgpunitevents trunkpunitsrcorgpunitevents trunkpunitsrcorgpunitexception trunksrctestsapiorgpunitrunner trunksrctestsapiorgpunittestclasses trunksrctestsapiorgpunittestclasses a cleanup issue with and added two new testcases to the whole test run +Vlado all pass mean that there should be no regressions when running the maven tests locally against a new version of +Thanks +With these asserts Im not able to create a new databaseij version ij connect jdbcderbydbcreatetrueERROR XJ Failed to create database db see the next exception for detailsERROR XBM Startup failed due to an exception +committed the patch from Vanessa to add equivalent for karaf commandsTODO testing doc update +Good point will do +One of the test cases invokes an MDB listener which starts a JTA transaction and propagates it back to the caller who then completes it +Let me know if you have any questions +I dont see those values updated by the jmx console mbean views either so either the conditions for them being updated are not met or they are not working as expected +Please pass thanks on to Rod too +See for further details. +In this very case please provide them yourself if possible +Vadim +What is and not relevant +for Martin patch it fix the attached test case asserting that when responding an AJAX request an AJAX response will to be with rPedros tests are also in. +Resolved for a while +Implemented in revision test case for order by with relative path svn revision +Scan must take in the name of the table and may optionally take an array of columns or an array of columns and scanner specifications +interesting that it should be necessary though what was your previous version of Maven +It does be a defect that the default protocol is not setBut username and password are not required for JMX protocol so they are set to null by default +why does the editor need to modify? what kind of changes is that for? profile names or +Marking as a blocker for +Committed with the Nits fixed Thanks +Can you change the JIRAs status back to newThanks per the above comment. +I have reviewed and tested the patch +Thanks Ning +This change restores the old processing of forc and supplies a different fix for +we should not target it in for the moment +Guys enable cookie from internet explorer settings of WP +Modified files for fixing this issue +Done +Its just comments no functional changes +Committed patches to resolve issue +I couldnt find the patches that provide the bridge between HDFS and the referenced patch and JIRA issue contains only patches ot the external libraryIs there any progress on this issue? Could you perhaps upload those preliminary support for HDFS +Patch for the zebra smoke test +Verified on. +The parameter is no longer used +In particular all of HDFS knows that a link can be to an external since this is a generic property of all +It seems to be working ok for me in Im on to another project so I cant test this tried looking at the source rep to find where the was to look at the diffs but I coulndt find it +Please see +It may make sense to write the props to a buffer and then include a header make it easier for things to detect breakage proxy servers playing up etc etc +Yes absolutely +Any chance you would be willing to take a run at improving the docs +This was recently fixed +In any event I have spent a lot of time implementing JDBC and I do not appreciate the problem being kicked over to me +Thanks +What is needed is our keeping an id rather than the table name everywhere so a rename requires our changing the tablename once in a table to id map rather than in a few places in the filesystem as well as in the metadata kept by each table region. +In APT format the HTML page title is generated by the plugin by using the main headline specified in APT markup +EJB is not included in this list as is covering propagating the JAAS Subject to be handled by the existing integration based authorization checks +A version of that closes the source and destination after opening them +Commented out rules no longer appear in outline other issue was fixed a long time ago. +Would any committer take a look +Committed! +DM Thanks for taking care of this large changeBy Mikes comments on seems will be committed and this one will be canceled so I feel kinda bad for the time you put in the last patch here +HiTaking a look at this old bug +Ubuntu LTS ATS version no compiler optimizations +My current theory is that this was caused by which is not in + +This issue is now fixed with the latest plexus and maven +The issue still happened on jdk using the newest trunk and while swiching into jdk the issue also happened +I forgot to mention that I wrote the keytool keystore generation Mojo yesterday +in release +I think we need to do blocking mostly for async jobs in the test case +This distinction may be necessary as other may not visualize a drop down list and a multiple select box as closely as the WC does for html +Is it possible to send sample data through sample triggerCan you please help meThanks allPlease help me in this issue +Thanks Vinod! +this was an old issue +I do not have the permissions to commit the change +All committed now. +ADD INFO I just want to be sure that for instance none are called from eCommerce where an user can be anonymous +moving this to since we ran out of time for this +The error is coming from the updateartifactstatus function and thus is only thrown if youve uploaded artifacts +Seems to be Is this issue a duplicate of +I dont really see why magically removing and the and adding all the again is better than simply changing the socket bindings you want with +running tests but should pass fine +Chris do you want to talk to Lars or should I +Please fix it +Well I have reproduced the issue at bundle startup +Itll be great if you can submit a patch for common only +This issue is a duplicate of +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian ve just verified that its fixed in releaseThe issue can be closedThanksFilippo Munaf +I like to see these issues solvedThanks in advance +The attached patch includes this +We certainly dont want to copy just a subset of the data as that would give very misleading resultsbq +I created a new one my branch is up to date +This time Im closing this as invalid its a setup issue with this specific jvm. +Yes closing this was fixed. +Paste the link using CtrlV +JPOX only queries the database for classes that have been registered in the runtime environment instead of classes that are described in the metadata +This JIRA is not relevant any longer. +Rajesh Again if we go to check the ROOTMETA while startup we have to again see how long we should wait if ROOTMETA goes down at that time +For this reason the dependencies page iswhere it is +This is fixed in the but not in the once is used in the later this issue can be closed +elemWell I think its a question of taste + is missing +Thank you very much for the help! shouldnt need the keepAlive enable as the only thing that does is check for TCP liveness every couple of hours the inactivity monitor would be a better choice if you really wanted to know when the TCP connection has dropped otherwise just add false to turn off the clients inactivity check then the send operation will be the only thing that will trigger an when the TCP connection is down +Committed patch to documentation trunk at revision +The new snapshot of meclipse support is now available from the dev update site +and should all be attached together into there appropriate project +btw unsure how this relates to might be for the delay reviewing +Not FUTURE releases but NEXT release +We recently started working on sqoop where we will rewrite most sqoop parts and where we will be able to do backward incompatible changes +I have try struts tiles in jdkbut the same exception occured. +And the title is misleading will fix that +works fine locally via tycho +There is a known issue in jffi where it selects the wrong width binary on Solaris +Preview shows what SHOULD be shown in the view but clicking OK does not cause the view to refresh w the new query +So I cant provide any patch +As regards to contrib components lets have a separate discussion on how BK related code should be made available if not contrib +ThanksIve committed a fix for use but leaving this issue open so some tests can be added. +Oops there was a silly error which led to NPE fixed now +Hi Patrick yes +Resolving Im not able to reproduce that anymore +So these segments remain alive when program exits +The locking patterns on issues and look very similar to +we do not need to clone all the nodes from a node to its root we could clone only the node that is being modifiedI wonder whether this optimization is worthwhile +Thanks RegisPatch applied to JNDI module at repo revision rPlease check it was applied as expected +Ive fixed the code style violations +It is just that resources from a previous usage of are not recycled during the whole run of the test suiteI see you reuse the application instance for the different instances of +D +Hence closing the bug +Do we really need it +Patch for possible way to fix attached +I will try the change you made and will let you know how it goes +Yeah +Scary to think that this has gone unnoticed all this time +The wrapper element should not have an xsiniltrueThis seems to be a client problemCan you please provide the exception stackAlso please provide Name +in maven junit it probably didnt fails because JVM is reused and its already initialized +Fixed in r with a test added to test +account tested passwd changed. +Backported to branch as svn closing. +Nets requirement +Ive just committed this. +Could you provide an update on this? What is the schedule for addressing this issue? Our request to transition preceded your move of JIRA to a new host +Max Denis Mickael Please review for inclusion in xx branch for CR +Thanks! I updated and didnt see the test I added +Hi DavidI cant find attribute lang and dir on struts form tag +Yes it is possible to do so via the configuration variableInfact there is another reader which we supportThe new paramter that you are adding should not be specific to but should be general to any recordreaderMaybe it should be namedset and all recordreaders should be modified to take care of that +Attach patch for trunk +Bulk move of M to the issue here is with having to scan the entire xhtml document to register uiparams variables then maybe initially limiting to ones defined in a fmetadata would be a good compromise first please verify +Thanks working fine for me as well now. +check Do we expect at least one of the directories to have the file +Integration with the jboss security layer to allow authentication and authorization of jconsole connections as an example is something we need to support with this +Bulk move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Attempted a fix in revision +I think weve reached consensus on this so Im marking it fixed. +And I would have no objection adding it as a hard dependency for Deb had forgotten that this discussion had been held on JIRA and filed a separate ticket for this and some other standardization issues +I take it back we can expose uberness in I confused this withRobert if you have a patch Ill reviewcommit +In this case the boolean query would be built and the terms cached +Putting obj files at target directory also work too dont you think? If the user want it to be at somewhere else heshe can set it +seems to have a few nice properties but do we actually require themNot we can get around s short comings by having the create factories that take in arguments and pass them to a constructor however it would be cleaner +superseded by +No integration test +Shouldnt I be able to retrieve data from Foreign Tables? Or do I have to manually create a table inside the VDB? Is that what I am missingThank you check with if you can see your tablesviews +I reran build test failed interpreter passed +Screenshots are for embedded jopr but i expect this to be a problem for Jopr too +Your patch has been applied in rvI will close this issue now even when is not complete +Remaining samples converted to version of. +Sorry for the miss placed packages +We can simply point to a standard +I dont think a parameter is necessary or a good idea +Hi Dave +Fixed on master branch in that be backported to branch? It probably should +Test added +java native date +here is a patch +Daniel please leave until we have a chance to talk about it +I have a fix for the tests so they check valid expectations and the tests now pass on both the RI and our java jdwp Ill rerun the full testsuite again to make sure nothing else is affected +Thanks Gera! +but this should really point to the correct branch nightly or nightly +Thanks Jason! Feel free to commit it +Would that mean it should generate a FOR UPDATE OF clause including every tables alias except cat? Surely thatd be a lot to ask +Moving it out of release. +Softera LDAP Administrator is not doing this +I dont want to be in your way really +Then leave it for Mark to explain when he is back next description seems to suggest that we are enabling the ability to inject XML content as well as element values perhaps the attachments back it ll go through them later and see what I can make of it +Hey Todd No problem on that I think in terms of running a MR job theres no real effect unless the framework is the same Counters reference for different jobs which could happen if JVM is enabled +that is because it was not like that in latest JBDS release that Galder is must have been fixed after GA +Thanks for review +The workaround for now is to just have the customer add the entry to their startup command manually +I had a closer look at the file and discovered that the HTTP connector is configured as well as the HTTPS one when launching an app with grails https +Attached patch aboveThis should be OK with both sandbox and sandbox +Hi MattOne question +Updated patch for bug still exists in Eclipse is there a workaround +reopen to move from LATER to actual fix again. +Hi Jinghai I closed this issue please create another one it will be easier for me to pick the new patches +This is an API that exists already + just committed this to trunk and +Added preconditions checks in revision . +Due to a refactor that was not completely finished +Sorry begs the question who would it confuse +The way are initialized is not correct currently +No worry I just wanted to give a ping in case you wanted to help out with the last hurdleAs I am closing down issues to keep within sight for a Q release +Okay I originally missed some new files in the commit +jar is missing +So far our upgrades of the counting clusters have involved recreating the data from our source of truth so once more would not be the end of the world +Specifically it is somewhat visually distinct from the project logo while retaining obvious association and incorporates the Apache featherI also think it works well with the tradition of wearing different hats at Apache +Initial publishing added for M +could be important +Thanks Ted Dunning for his hint +pushed to arqmaster +The correct way to build Axis is using mvn clean install +The catalog has been added into the juddi webservices war and tested on both CXF and native +Looks good and the proxy part works for me nowSo really it is just a list of configurers which we just iterate through passing the same httpclient instance to setup the various configurationsI havent tested it with both basic authentication and proxy configurationThanks for updating the wiki that was on my list also +heh this was the very first thing I did with the code base and this particular patch isnt exactly correct in all cases +Verified with latest build +Then create a snapshot of dr with the same name +This link does migration make sure it works in line w all going on in here +the classname you have is not identical to the one in the process I assume the one you put in the report is to be used in ther process +Are we not talking about objects that are local to a specific database? I miss a rationale for how you ended up with three features +Please check that the bug is fixed for you. +Meant to close this as Wont Fix +bulk close of resolved issues whose fix target is either unspecified or from a prior release. +FYI +And the Run as +Verified in Tools +Even when you get your hands on the device you are not going to be able to fix this +It does not protect against some crazy initial value from the old log file format which might cause it to read the entire file as a single UTF string +Attached you can find a test application mxml +I did +Bulk changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Patch looks for just committed this. +Hey Gauthier I updated your patch for the new project directory layout +We wont use the Flint Editor so I close the released! +See there is no necessarity to include the fix in. +Thanks! +Create a Hybrid Mobile Application +The patch expands the package imports to the used classes +I think this would be useful feature to have in +Issue is fixed for me +This is a duplicate of +Please check my patch +Logged In YES useridFixed in and +When declaring variables already in the process definition it is possible to specify whether or not they should be taken into account for the history +Thank you Claus +That is mentioned in the description but since the example of the current workaround is on I want to make sure that point is not missed +attach a patch to let hub server could configure write quorum and ack quorum +Yes I close the stream after read or write but it could be possible that another thread are also connect to the same server in the same time and create this issueis it a hard limitation or a parameter? do you kown if is it possible to increase this +BUT when a new asset is created it is saved into the repo but not checked in +Could you please suggest some Hibernate material that I could read +Since the queue will be nearly ordered that is not a terrible assumption and in the case we are wrong will only mean a few events are delivered twice in the case of full replay +Attaching a patch which extends with test cases for this bug +But then again whats the point of deprecating coretests and soatests and creating integrationtests and soaintegrationtests +Denny what branch is this a patch againstWhat does enable the modify button have to do with setting the to +GDB attaching was moved into OS signal handler because GDB shows bad stack trace when attaching from chandler with transfer stub on the stack +Committed rev to x +since we have classes from Avro and Thrift already maybe we should just let those deal w the serialization and wrap them in a class containing our internal logic +Glad it works out for you now! +What happens if I call manually +Screenshot for installer when select chn for install a Installer bug but JRE incompatibility with the system. +Based on Steves suggestion I made single custom and simplified everything else +Ok +Seems to work +Ill have to get back to you on thatIn the meantime Ill start making the changes you requestKarl +This issue is fixed. +The webapp gets it directly from the jobProfile +it make more sense for DIH to farm out its content acquisition to a library like DroidsIt would be great +I will fix it and add a unit test to cover it +The subject and the description of this JIRA was updated to provide a correct explanation of what causes the NPE +Heavily commented as this will serve as the base to other structuresStill partially untested but would like a general code review before moving forward +Yesover the iteration on parent entity child entitys parametrized stored procedure params are changing but returns same over tracing DB calls it revels the fact that those are called only once during the DIH run +Can be reproduced without sharing the EM between threads? The added locking in this fix will prevent two from getting the same sequences Im not sure about a single thread using the same EM +MVCC in the cache allows for reading as well as concurrent +Sorry it took me so long to get this one in +Ill try that tooThanks for testing and for the useful error with and +I neither insist on a test as the performance improvement is smaller than the variation of testing time on my machineWe agreed on our Weekly Development call to take this fix into the upcoming ODFDOM versionThanks again for pointing this out PallartBest regards the changes +The patch contains changes to relax the collocation constraint in case of mappingThe document contains information regarding what works and what are the limitations. +I can fix this problem +I think you have broker and broker reversed based on Lowrys description +was renamed to in my Git fork at some point but when my fork was merged to SVN I realized its still +Have you had any issues w this +I think data locality is broken in in general not just the counters +This class is a one it would be good to keep it that way +Thanks Kristian +Its really a general purpose encoding layerJaceks codecs were all delta based but Ive written a TRIE format where keys are not based on deltas between each other +this code can never generate an angle greater than pi and yet it uses code that falls through to standard cos for anything smaller than MAXINT e +Maybe it can help +I would prefer us to try and define a common interface layer from this API and the implementations that are built on top of the forthcoming serialized metadata representation +I have been unable to copy text from the Clearspace editor and paste it into a text editor +Lets please move on +I have attached a zip folder with a +nbrow rownumber +Issue closed with release. +Fixed support doesnt need to support forward as the type is fixedJust just have to cater for the const support +Later when aliases are expanded by function objects are instantiated from those instances resulting that args specified in DEFINE statements are used to instantiate function objects instead of ones specified in LOAD statementsMy question is whether this jira is to solve the same problem or not +Rob Can you also point me to you presentation slides in Hadoop summit or any materials related to it +HiThere are a couple of issues Its not possible to embedd the package all has to be prefixed with All files have to have the Apache License plate youll find in all the other files +What comments are you referring toThe comments just before int float long double say For faster range queries consider the tinttfloattlongtdouble types +Nice set of tests +If I take this patch an index created with will not be cross compatible with Lucene Java? +Patch for initial review +And I would still need to add the executeWorkflow function in as it is required to execute dynamically created workflow objects composed of wings tasksRegardsVarun +Closing resolved issues. +Flushes are scheduled just before we put up a block +Looks ok to me +There is a solution for this in the current snapshot were going to release an RC this week +I believe INFRA issues are not right place to discuss Openmeetings buildsPlease use mailing list instead. +OK thanks for the explanation and patch CarlosIve just applied the change +Now when edit this portlet show preference tab may be change URL but nothing happen when change + comment reported as user from commenting this closed case new opening is more appropriate +nothing to do anymore here +I believe the release was already built so I dont believe I can mark this issue as resolved in that release +I am not sure anymore if that was with the older client version or if it was with this weeks update +This issue is a duplicate of +Committed the patch +The attached patch should allow for server global template files +Thanks for your support +Thoughts +let me know when its ready for applied +Can be when that one is fixed. +I will create a quick branch because Robert and I want to fix this +Solved after solving after releasing +Query did not succeed with million row table with existing plan +jberet was just released on +I think this will be very useful to have +message +The have no idea what you are trying to say here. +I think the findbugs and test failures are unrelated to this patchEric Todd does this patch and the one on look reasonable to you guys +Fixed in CVS +There is no implicit version assumptionThe assumption is that we can get a cluster UUID from and if no UUID was set we get a reference to DEFAULTCLUSTERID which is true for any version of the WAL file format precisely because we kept all the versioning inside ofThe new code can read the old file +Sanjay sorry that I misunderstood the advantage over +Can you please explain in more detail what youre doing +Transferring back to Shelly +Thanks Alexander the patch was applied to BTI branch at rPlease check that the patch was applied as you expectedJust for the recordI left the test in the exclude list because ofIve checked that if I hack Harmony as requires the test passes. +Checked in fixes to trunk +The diff is unrelated +thats why Ive filed the issue +As a first cut at testing a fix for this issue try running and the compatibility tests +This is incredibly useful patch thanks! And wee need more things like thatResolved in rev +More comments in code per the reviewers comment + we should refer to mention the newFor VFS Annotation scanning tuning I suppose you need to talk to Ales +mirror is back +Ok thanks Alex for your second patchPlease close the bug if everything is fine now +Thanks +Syncing patch with trunk +implemented in HEAD +i added your changes but against a new issue +I am also eager to hear about solutions to this problem or possible workarounds +Nevertheless does not cause problem on maven Though this is probably a checkstyle issueSorry for disturbance. +Is that a philosophical statementAs Cassandra at least for now is a pp network with fairly clearly defined boundaries I will continue calling it a system for nowHowever looking at it from the pp viewpoint the user potentially have no clue about where replicas are stored and given this it may be impossible for the user to issue repair manually on more than one node at a time without getting in trouble +Alex isnt the right approach to fix the public id in the dtd instead +Hitting submit patch so the QA bot will run on it +Hope there is more interest in the community to join us +Uploading the first attempt at a release note +So I still dont see why in this particular instance direct access to the registry is needed +Ive applied patches to the branch and the trunk +I tested it with the current code for which uses Xalan +Seems to have bugs in the generated javadocFor eg api only show hwiapi only shows qlapi only shows hwi +Closing this one +btw +A work around is to use the Trinidad to submit the request +Any chance of a clue as to what data has changed? Something to do with medical and dental insurance? Wading through this object model and trying to debug some clue out of it is too time xml files company and company define the same pc instances but company does not set any relationship fields +But I can say that the results are very good + +patches applied +after running a maven build one can with different options. +Juergen can you backport the changes I made to to +if we have collisions in node type definitions i feel that this will cause troubles +So if we can come up with something soon that would be good +so +it still necessary to increase the sleep timeout to Ive increased the timeout to give some more window period for the async operation +Thanks for the patch UdyWe should probably make this a general feature of MR jobs that you can specify job name from command line +Milos is there anything left to do on this or should be close it for +I actually get the server stopping in DEBUG still no idea why +Thank you in advance +The hardest part will probably be updating the tests +pending Bowen +I would suggest looking for another alternative. +Version Released +The composite pool implementation in the trunk for pool has the ability tocheck that only borrowed objects are returned if you use the configuration option when creating the pool. +Its a modified demo from camel so youll want to put this into the examples directory +It only needs to be integrated in the Solr facet code +Fixes the final step logic in the Aggregate to release if multiple Cas Multipliers are used +Dont agree +Im checking to see if its not possible to have a the packageName attribute I fixed the xsd today on the web can download it +Will you be available to log on asfinfra at some point to be available while I migrate? Mostly be looking for some basic things test cvs is once the migration is complete confirm that you can check things out commit them that the email goes to the list okay etc +If an entity was not found by sourceURI there is a lookup for namespaceSo i think Axis is not able to handle multiple xsds or wsdls with the same namespace at all +One point dont have to be commiter to review a patch +Looks like as a duplicate of +The code only issues this warning when the expression is actually hit so Im pretty sure this works +Also I agree re commit though I think we should probably add a couple more tests in this area +Attached patchCheck if +So moving to. +From your log Caused by Address already in use JVMBindmeans port is already taken by another process +If the method was in you could only call it once the workspace in question has been openedHow about making the method protected then youd need to subclass to access that functionality +Great point! Looks like weve got a couple of those references to varlock going on in Debian scripts +The service adds FS to that view but the delta CAS output assigns all the new content to +The connection seems to be failing because I changed the jar files under the running code +I have added scprepositorymaven as repo info +After you run that you will have a maven repository installed at + +I tried reproducing that on IE SP running on XP and did not see it +Moved the TPS issue to has been documented as a known issue in the EAP GA release notes +committed w above modifications +In my writeup on the diffs ignore the comments about there being a possible bug + +The change I made is in and it has the better solution and avoids the unwanted changeThe changes are in SVN rev +Actually we need not compare that content +I can then mark this ticket as officially resolved +How much more difficult would it be to just have it do a straight HTTP POST or PUT of the new image instead of the Ill ask you to ask me for this image danceI investigated what it would take to do this a little bit and concluded that to do it right would take a fair bit of refactoring thats well outside the modest scope of this JIRA +I read back and saw that literals count as well +wouldnt that suffice +Backtracking on this because JPA states mappedBy +Patch with typo fixed +fails to build when applied to due to a missing file +However I think it makes more sense to make serializeable than to make a whole alternate servlet +well you have two of them in your pom and you dont seem to complain about it right? There are tests that validate this scenario +bulk defer of featurestaskswishes from to +Thanks! +Ill use it carefully +This change needs to be noted in the release notes for CP the quickstarts files should be modified in the next CP +worked just fine +The zip is not longer available +Ok thanks for the quick response! Ill create an issue in the Hector bugtracking system +Anirudh +Committed as r +Optional jars for would be a great Issue to solve soon before a lot of people jump on to Maven +I thought about that but since streaming is primarily used by developers who dont want to write Java I think the usability is better if we just have a set of enumsstrings that we map into classes in streaming +Also there seems no coverage for wrapping guava function in better to have some +I updated CI Trunk jobs of component to add the pack profile +Make sure to resolve this if you find the latest SNAPSHOT fixes the leakages you were seeing Need to know when to start thinking about a new release +Youre right +I was referring to your comments onI also took a look at the predicate push down mentioned there is actually a bit different from what we plan to do +Fixed the calculation +Steve I gave this a go in between meetings this afternoon and it seems to work great + Support for multiple idl files to be generated for multiple bindings +The adminhandler attribute must be persisted back to close for Solr +This version of the patch renumbers this item to like all previous versions but it no longer includes a at the beginning of the items text as all previous versions didShould hierarchical change item formatting be supported? AFAICT the item is the only occurrence of this phenomenon currently in +Should the failure count be reset per block but not per readAlso I tried the patch +Write an which initially creates an index of size N from one content source and then calls the several times +Why not rely on that list if no explicit list is given rather than coming up with some other hardcoded default +Source artifacts are useful and should be separate artifacts and installed into the repository as a separate jar +Committed revision +I think I can come up with something that will work ok for now. +what do you use as the group on the discovery multicast addressEnsure that it is something unique discoveryUrimulticastdefaultgroup +Committed on trunk and branch. +we should at least have one basic template for each of the file formats we provide a template mechanism for +if the problem still there please attach your wsdl and reopen the jira. +The change to file input split javadoc is still present in the latest patch which causes the javadoc warning +Not usre I like but there should be some way to define a targetEntity on the Maps key other than defining it on the Maps generics +Final patch was attached +Applied thanks +pick arbitrarily +Ill try on Lucene soonI had assumed that the second exception had the same root cause as the firstI also switched to fsdirectory and let it run overnight +fix var initialization unit test styling +If you find further issues please create a new jira and link it to this one. +Good catch Doug fixed now +Committed patch a to trunk with revision +Just modify in run ant dist then ant deploy +BTW this wont be fixed as its better to return Map rather than the implementationBut I will add your project as an it test to ensure we wont break something in the future +are the bits from the wild which are included in our featureI could omit the features from the update site if youd like but the last time I asked the decision was to include BOTH the raw features and our wrappers w the updatedcorrected license text certified identifier +war pathtotomcat think that Glassfish v is affected as well +Thanks for the patch FrancescoI applied it modifying the code a little since there was an easier way to implement it using the already implemented that helps us respecting the DRY principleVery appreciated thanks a lot! +It seems to me that we need a set of tests which exercise Kerberos authentication in all the supported cryptographic modes perhaps by using the Java Kerberos login +It seems better to add an alternate check for another driver. +Thanks a lot for addressing my feedback Todd +is confusing to me because it looks like new file wizard is actually files generation in for any source class in workspace +Id like to ask you guys to do a config file property as well +Page is not properly renedered +I will look at this +In many cases you can happily add a new consumer and even set a new message listener to a running consumer without problems only in the tangled cases of synchronized methods cases youve created does this run into problems and thats why the spec makes these notes about the use of sessions and consumes by threads other than the thread that is delivering to the session with a running consumer +Yep Ive already taken care of this +It seems now that qdox has been stabilitzedJust changing the dependency from to fix annotations parsing i m still having parsing errors with qdoxSo no need to upgrade until the error is fixed in qdox +Switch over complete rev +Hi James I think this is not the right place to discuss issues related to my fork please use Tayler github tracker for thatThanksSergio B +no new tests because this is only when something is drastically wrong and in the current libhdfs environment that is almost impossible to test +I believe we need the jaas for JMXSSHWebconsole aslo +I ran a lot of successful tests and I hope I didnt introduce any side effects +Deferred +Do not lookup in the constructor +can always override the defaults using or whatnot. +It is literally there as an escape hatch for people who really dont want a CGLIB dependency +We can move this to HDFS and convert it the subtask of +I just committed this to and up. +I think would be a good way to encode transports and their options +Also embedding a slash like to +No activity on this report in over months +Please provide a svn diff format +Tests pass locally +I have been trying to build this for two days +You need to pick the right one +Note that you will need to upgrade your cluster to hadoop once you update your tree since and are not backward compatible. +Just committed this +Im not to unhappy with the new structure +only thing i didnt understand is how we are conveying to the combinefileinputformat that combined splits cannot span tables? there should be some data structure that captures this information that we need to push from hive to combineinputformatalso how is the required number of splits configured +Arrr cloning an issue didnt work as I expected +Please contact me if you need any clarifications +It has nothing to do with plugin classes +Thanks for taking a look Vinod and Bobby! +Got caught up with some build issues today will post a review for Edison tomorrow +on the patch +Ran again without the USING clause and got the incorrect result +I finally got the junit test to work on windowsIll assign this issue to me and Ill try to commit my changes within a few days +Thanks Christian +The test migrated in different iterations and I no longer need the trycatch +Mladen no I dont +closing it. +This needs investigation but Id qualify that as a vaild case and hence a bug +See upgrade some supporting product dependencieswhich lists all jars and has some upgrade notes especially for each jar that Forrest also uses +Manually ran that particular test case and it passed +Havent seen any user reports of regressions in behaviour or performance as a result of these changes marking as resolved +Let me know what you think and Ill start working on a new patch hopefully later today or tomorrow +patch for which excludes facilityId +I think this issue has been solved byCould somebody confirm it? that resolved the same issue. +lgtm! Thanks for the patch this is very cool +Hi thereJust out of curiousity since this issue was already closed for months why not merge this patch to code tree and add VS to supported platform on Build pageIn addition looks like I have to use xerces if I want to compile xalan with VS rightThanks in advance dont modify a release once its out the maintenance overhead and the potential for confusion is too great +Committed to trunk +I forgot to commentIf there is something Im missing though and there is a good reason for this then the object also has a GLOBALMESSAGE key which has a different value to the one +Assigning back to you so that you can apply the patch +is the status of this request + +Applied patch to source repository +Could an xml file be passed to jobs on the command line +Ah that would explain it I hadnt realized the single instance was shared +I cant reproduce it +me too +Cant you use delegation instead of inheritance here +Ahmed looks promisingCan you generate a +bq +Closing this out as incomplete until a bug can be verified +Please fix this for +I believe the schema is in error the schema spec requires that the canonical representations of defaultfixed values have to be valid with respect to the type +Eric +just committed this Thanks Suraj +The current mechanism is not great it pulls ALL the documents from SOLR finds the duplicates then issues a delete command +Acegi is now Spring Security +Im a bit confused +Any idea is welcome +offers a way that a single field type could write to multiple fields +Arun +But Id rather haveall of these tests fixed in a single patch since we seem to understand thecauses of all of them +Contains modifications to camel to add ibm jdk profile that will force surefire to use properties file with IBM JDKTested on AIX and +its checks tokens then services and it should be token service and then tokenservice +will ship with other fixed issues through the update center +for the current patch +Committed to trunk +The use case I was thinking of was dedicated server with gobs of RAMThe index could either be updated often or not updated at all +Ok went ahead and committed this +This patch implements the new watcher object API +When does each thread do its commit +Getting into the code heres something to note The javadoc on setParser indicates that it is not thread safe so I think that throwing the exception is the right thing to do +Fixed +Im closing this for now +Changed priority to critical we cannot really have a blocker bug open this much time +I had trouble starting the examples given in examples +I run the all test within itest directory inclusive itestdatabindings and all is ok +There are two attachments for this issue +ssh because that is not guaranteed to work. +So the apparent river end that I tried to extend was actually not a river end at all but just a vertex within the river therefore extending wouldnt for us being +Thanks for that hint +Verified in ER +Superceded. +Please review the fixYour patch looks correct to me +Earlier today I made a change to that optimizes its paint behavior +Updated documentation as suggested +keep the solr package names and make it clear that its primary purpose is supporting higher level features in solrAre you serious? If you need to change something or make additions to such an interface abstract class or whatever you should be able to do this +However I retested against and that is not fixed +But the directory did exist just the files that dont +However on a purely API basis Id still like to entertain the idea of not forcing the client to catch a checked exception when borrowing an object +This patch resolved conflicts with the newly updated append branch +This will be included in the upcoming release of Avro which will support mapreduce over Avro data files using Hadoop or greater +I cant create a solution for only javadocing public API sources without knowing HOW theyre were going to be marking at a level lower than level the solution for this becomes substantially more difficult +does not inform the listeners when the job is submitted and it waits for the setup completion +I have checked the code in coprocessor it cant support the dynamic class during running though I can refer to the code of how to load jars from HDFSCan you rephrase thisIdeally we should have one classloader that can load from HDFS not more than one only slightly different from each other +Removing the framework adapter JAR in the plugins directory and cleaning the again did work +The statement to set the base URL happened after the inclusion of and when cached requirejs didnt know where to go get scriptsI moved the base url statement up in trunk but I am not sure it is fully resolved. +added an entry for running on OPT in H regression test since they are identical. +but you see the problem with trunk still +It works for me +Sorry guys I wasnt ignoring this! I tried to apply these patches last night but it looked like they required some other patches first so I just waited a day +if you prefer a list here are the filesMy question would be if the build would fail if one of this is not available as most of them would be used only for release distributions +All setup on Nexus is complete +It is tiny and I dont want to create additional Jira for that +Hi AlekseyI dont agree with the changes to the classes in the +r. +hm +The JON integration for message alerts collects the alerts and displays them as JON events or if the administrator wishes to do so as JON alerts that are fired triggered by the message alert JON event +Then the message wont be delivered forever even there is new consumers on that queue +dfsshell uses the dfs client which in turn verifies that the crcs match the data of the file that is being fetched +Does auth require? Normally HTTP requests expect. +So you can see the old code in eg +Patch committed on and trunk waiting to close for some what I meant above is that if I set the max LOD to Google Earth starts making a lot of requests +Had to mess in transactional hbase to make it use new rather than as value +I reopen it as its not really fixed +Marking this as an incompatible change +I dont think this is related to the changes made for this issue +Are and still needed +Will avoid surprises like that reported on this issue +Edit release note for publication. +Awaiting good Hudson spin to verify config there is good too. +Moved class to same package in docs module on trunk That did not work for me and now uDig is broken +Until we develop a protocol that addresses these we should only describe HTTP in the spec and we should use HTTP for interoperability testing lest we end up with a of incompatible legacy protocols that folks might expect us to support forever +I will update my working copy to get the more current version of the test classes +Feel free to start one +Please use the user obviously could go to user forum now but after reading specs it seems to me to be a bug +Committed to trunk +Ill have something that fixes those shortly +removes unnecessary import statement in +Concurrency in schema sharing is dealt with by using a loading cache from GuavaThis ends up tidying up some of the zookeepernot zookeeper logic in as well which is nice +Bug has been marked as a duplicate of this bug +Brock do you have time to review commit this? Unit tests pass and it looks OK to me tooChristopher Israel apologies for just jumping in here but I want to roll an RC of today if possible and this is something that should go in if we can make it happen + can you apply to WRH and too +Test for the report! Fortunately this has already been fixed in CVS HEAD and will be released in +Since this problem relates to DRLVM kernel classes Im changing subject and linking this issue to DRLVM componentTo reproduce it onehas to run JDWP test several times until it hangs and then attach gdb to a hung debuggee JVM +bq +In this case however the profile would also be active when building only the parent itself an unwanted effect +Maybe if you describe your needs for the application I could tell you a few things +When agent turns off Exception event inside of Exception event callback interpreter went into an infinite loop +I dont think assert really has a bug its written more to match lucene queries and indexes not to exactly compare one token stream with anotherI see your point about assert but it was being used to test tokenstreams +If I would test this I would have to do it in my production environment which I am not so keen on +Changing title to be more descriptive and reflect that the problem is not zOS specific. +Part of a bulk update all resolutions changed to done please review history to original resolution type +Select the correct project please +It also uses this +again! +Ok looks good! yay +Correction test for the unit test for +superceded +DFS and Hftp override get to call build +It shouldnt take nearly a minute thoughSuspect swapping GC or other external factor +Tested with SNAPSHOT on and Oracle XE all produce the error +Currently this works in Branchx Ill have a look about trunk when alls merged in there and brought current +I guess this isnt really doable unless this plugin gets bundled in with Fedora s Eclipse spin +Im wondering if we covered this with later commits +I just downloaded via subversion and see that this was fixed +Closing this no bugs were found in certifying CP with on RHEL +Thanks Remus! +Im not as invested in this issue as I once was +deleteDocumentIn the internal reader is reopened with the newsegmentinfosIncluded a very basic test case +Hi KatheyDo you still plan to backport this one +Any objections to having this in as well? If not Ill make a patch +Now the pathStamp works wellAlso a similar problem related with the paginator on the webkit browsers was solved +The tests seemed to behanging and from the output of I was not able to affirm which test was hanging and alsoif it was due to my changeI will run tests again and revert back with the results today +committed +Assigning to apologies for late responseWill try to get by with using more mappers producing less output +Planning to cut a release +Fixed the other test to refer to directories in the test build root so it doesnt pass or fail based on whether test exists +OKWell no dates set yet but we see the end for many tasks so Im checkingPerhaps in a few againAny updates possible preferably before the end of the monthCheers +no +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +I dont see the models in your patch +renames to and updates the javadoc +However the file was still looking for a folder called todo +The welcome behavior is how eclipse works afaik assigning for GA to look at that automatic deployment etc +This patch allows similarity to write all numeric values plus fixed bytes +It works fine just using the brokersTheres still an open question about collaboration in +Depends on having pdbs and xmls attached along with the main artifactThen wed like to also get them down and copy them alongCant remember to have done something on that +almost done +Maybe not a convention but at least its not unprecedented +no tests. +Sorry forgot to say that for me this a bug no an improvement +Setting a flag on metadata changes and testing for it on reads and writes would save unnecessarily copies and synchronizationI will open a new issue describing the improvement on and to use with +Please in EAP CP CR using the testing war from. +And it works on trunk +If you can contribute the documentation by tomorrow Ill spin an RC after its merged. +I believe this catches them all +JacquesI have made the change in the bullet styles +If you still see the issue you should rename your project to not a have a first capital letter +Its confusing +Two small safe fixes +While it certainly seems inappropriate it ought to be addressed by a separate JIRA such your +Based on +Attaching schema +Attilas patches are applied toThe first one is small so I dont mind to port it to but the second one is rather big and I dont want to compromise stability at this pointThere are several other improvements in which are only in anyway +a much more sensible patch that adds the configuration +ran unit tests in local and sense +However I am wondering if we should support a more flexible rename strategy so you can move to a folder if OK delete if OK move to another folder if ERROR delete if ERRORWhere you can mix and match what you needWell something to consider for Camel Well spotted +Thank you +Let me know if you would like for me to open a separate JIRA on that or change this one +Its the same as in the bug description but unfortunately the newlines were removed in the description text +Unless all these new method are placed into their very own separate classes and existing classes are split up to reduce coupling then the permgen issue will get worse and worseSo having a declarative syntax available so that a custom permission service doesnt need to exist it a feature worth having +The licence for the font must be clarified +Potsponed +The lock contention issue here should be resolved by the changes in pool +Will apply to branch as well +Ah I see +Can you provide this in the form of a unified diff uploaded as a file attachment to this issue? Thanks +Ah good thanks for testing and the report +The patch as it stands does not incur any token related overhead because existing clients wont request and send tokens +Then you can switch language to French and then write tha label for french and so on +Moving to +to trunk and merged to the branch. +Additonal method in is needed +Move this to a future release +Reschedule for the next release +Shouldnt the holidays list come from the dispatcher itself? I dont see how this is something that should be scoped to the invocation +HiAny chance you append your working testcase? You can remove your business logic and as simple as just reproduce the behavior you describedIt would be a great help to reproduce and fix this issueThanksve attached a small Service Assembly zip to illustrate this issue +BenoitWhat do you think of the proposed exception hierarchy in the patch? addresses teds comments +should it be release noted +Remove added and update links to point to the correct location for this +Why didnt you apply this to +set is not in so I guess you compiled using the wrong compiler again +This umbrella jira isnt a blocker +Closing as requested by reporter. +See for alternatives +I have received a response from Refractions forwarding me my reset password +Im not going to the patch but Ill fix it in another patch Ill put up shortly +ThanksJust tested also now works on Chrome XP. +Should be built as part of preparing distribution +Hi IsraelWell the code is in the patch ready for when the maintainers are ready for it +Yes it is a side effect of +Patch for distributing Instrument for building against Instrument component +Fixed in tested with both store and and things seems to be working now independent of the host environment your app is +If you cannot patch core you could patch the client side and use which is what we are doing. +fixing the path to be windows like doesnt work +This is the ORDER BY case +Thanks for your help issue with CPD has been reproduced and fixed everything should work fine for our rookie developper seems to have commited some code! It should be good now +oh sorry that makes sense +I checked the patch seems nothing will break existing code +I see no reason why we cant change the API +And btw are you going to use MS Vista for any production environment with Camel? NoOr is it just Vista is your developer environment? Vista is just one of many test environments that I am using +verified by Dmitry +Can we make this an option say configurable in? Im not sure I want my downloadable pdf files to be flagged as errors to my sites users +Going to commit +Strangely the parser produced errors in all cases except that of a field having two consecutive attributes this I have fixed +Oh! Thanks a lot for saving me some time +Restoring and deprecating the old methods is a blocker. +do you have anything against adding the staging repo to the parent pom +Brian contributed the patch +Check it out and let me know any suggestions that you might have regarding enhancements +I think the overall approach is fine and I dont have anything against depending onTentatively scheduling for since Id like to keep a fixed scope for the release but if this turns out to be quick and it should find its way into the first release +The project structure of Apache Directory Server is complex with quite a few files so this sort of version confusion will occur +Did you run into this with the latest code from the branch +Closed as Wont Fix. +This has been fixed already. +Hi Richard Thanks for your suggestion it works +Probably the best place for this is in SWF +file location orgjbosswelds fixed in the latest build however the older builds are still present in the repository too +I will fix the test case +Work specific to this proposed solution has been abandonedDiscussion should move to in order to properly arrive at a solution. +I dont know if it is relevant to the idle check mechanism but it might be worth trying out the fix +Cool +Initial support will be available in Launchpad Base +It seems that Viacheslav has fixed it +version doesnt seem to be quite ready +Ive made the change. +I tested importing from our test BQT Oracle instance and was able to import procedure with a parameter delete the parameter or procure replaced the missing parameter or procedure delete the parameter perform Metadata Update From Source action parameter was tried to Imported BQT model from SQL Server created View Model containing Procedures Created and deployed BQT VDB containing source and view model Imported this VDB and was able to seecreate VDB Source model via JDBC import parameters for all procedures were not able to t reproduceAdding additional artifacts for a failing case might help to reproduce +Also I want to check the links when everything is finished +Any issues with it would now be considered bugs. +The problem has been fixed as of build. +Applied +By the way I have the bruises to show in this area +Finally xsd based autocompletion works in all the above combinations +I could not reproduce this using Groovy on Windows XP SPEven if I keep pressed the console responds after all executions are finished +I understand the areas of concern now +Please close this issue if it fixes your problem +DohNote to self dont just run the tests remember to look at the results as wellThe failures make sense they use a schema with uniqueKey defined but add docs wo that field to test other behaviors of toDocument +Can you add a testcase for this also +Just needed to delete the file and uncomment the bean in the regular +For copyright text was used Arial size on +Its not clear to me how well this satisfies the Open Source Definition +I dont have the command at the moment but my version of boost is rpm qa grep boostand attached is a patch which will cause these tests to work with that version of the lib +Im not up on whats involved +I am happy that we have the rat plugin to automate the tedious process of checking licenses +Great catch +The new patch file against the truck failed contrib test is not related to this patch +Committed revision +Master moves the region to failedtoclose state once tries run out +I have applications that work with httpclient. +Another option would be to not cache the methodclass if the class is a singleton type +You are correct +Perhaps theres a hashing issue or something +The patch causes the Eclipse classpath to differ from the contents of the lib directory +id for non domains +Choose Files modal dialog appers butStep +We have to do the inner work in seperate transactions and keep them small and short +i have changed back subject and examples +Hi RobbieCould you review these changes pleasecheers Keith +please reject or requalify to RFC if think that its important +The lib directory is full of dependency jars some which exist solely for the cli +No similar reports for quite some time +Martin says it works with a full distribution. +Seems to work fine with the latest +All the tests passed and I integrated Bryans suggested comment changes +New patch that deprecates get and updates documentation for the same +Thanks +We have looked into the lib directory but there is no Javassit library found +Is this plugin deadShould I apply the patch for you +Can we make sure that appropriate info is documented in the users guide and Java docs for creating String features before closing this issue +is probably a good target for the implementation helper classes as it seems to be a library already +Committed +aj is not correctly interpreted +Linked to +Not needed If osgi is not available the dialogs will be be suppose the idea of the issue was to also remove it from the codebase not only have it hiddin in the UI +Luca to both of your changes +I cannot reproduce +The check should check that the txid is greater than last successful txid and less than the failed +However the current portlet doesnt return any log records after subsequent changes were made to support multiple containers +Any chance to try the GWT addon before the official milestone release +Thanks for the reminderThis should go to first +One minor comment change the order of antcalls do first +Fixed in subversion tree as of rThanks for the report and the patchBeware that the Adams integrators are still experimental one +It is fixed in upstream and will be part of the release +Really we just react to change events and start writing to the corresponding session +Err ignore the random Jmx file in the patch +Plz set the fix version of this issue to verify it +Diff with the above fixesimprovements +No noticeable performance impact with defaults or with one column per row in periodic mode +Bumping since wo this secure distcp using webhdfs wont work for some people +Removing the services from the subsystem fixes this issue +Looks good to me +Id really need to see a test case +Hm I didnt think of IAE thats a good idea +raised up to warn +So essentially the same as patch +So for instance the Standard has nothing to say about whether an application can create a function called ABS or SIN +completely off base +I think the proper fix for this solution is to switch to commons code for zip. +Currently we install cars dependent bundles in cars activator which is not that +Wonder if the Deploy To button weve now got on the toolbar should include the Run as Local Context option then you can run locally or in any configured container from the same button +Output messages that look like they are informational and not errorsAnt environmentBase Directory homedjdcruiseprojectsderbybuildtrunkBuild output homedjdcruiseprojectsderbybuildtrunkclassesCompiler modernSane falseProceed noGeneratingGenerating ij parserGenerating mtGrammar parserGenerating SQL parserBegin building toursdbFinished building toursdbRevision number set to Beginning insane buildcreating class listcreating new fileBeginning insane buildcreating creating new fileBeginning insane buildcreating creating new fileBeginning insane buildcreating new fileBeginning buildBeginning buildBuilding Building Building Building Building Building Building Building Building Building Building Building Building Revision number set to Revision number set to +It looks like all of the changes for the rest of thisissue were already backported to as part of the backport of + +We seem to be sending the correct amount of data out for the target message so we probably need to dig into the AMQP JMS client code to see whats going on during the message unmarshal in order to get a better idea of whats going on +Now picks the correct eventCount to fetch based on its input index +Temporarily reopening to update release note info. +i think you interchanged the fix and affects versions + to +Ive created the version +Feel free to provide a patch +We may have to push it toAny interest Uwe +I thought it was but after futher investigation it is between a datanodeand a after I started digging a bit more and reran a job toverify +Patch containing the modified regular patch +The hbase table size is M urls and my batch have about k investigated Enis change and prefer not to mergewe should keep filter in query after convert successfully at least for debugging or maybe submit query againsome time we can filter again for example the in Hbase it return first N records from each region and may return to client more than N records and we can filter again in Gora to return exactly N docsAnd to not confuse with Filter in dont know why eclipse dont include new file in patch to for and other Folks is there any motivation to put some tests into this patch? +Thanks Peter +Thanks for the effort archiva is a great tool and im happy to have it at my disposal +While Java does not allow byte literals directly making your AST support them increases the expressive power of your library without exposing any fact this ability greatly simplifies targeting Janinos AST from a different compiler which is exactly what we are doing +Second build passed +contains one more test In the test is possible reproduce the error that I mentioned +This is done in delete service because record should be deleted when order adjustment of type promotion and have promotion code gets deletedSpecial Thanks to Ratnesh UpadhyayThanksChirag got a problem that for promotion in which discount on shipping and handling charges is not applying on clicking update itemsFor this I just called doPromotion method after save method in RegardsChirag fix in previous patchRegardsChirag this be classified as a bug and applied to releaseAlso are you promo rules or just saving the previous promotions applied? Seems to me that it should the rules since items changed +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +you apparently have more than one foreign key for the same column or howcould you show me the sql DDL for the database you are trying to reverse engineer +Closed after release. +There is no reason to force this assertion in the code +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +It can keep the DB in memory which suits the purpose perfectly as the test wont need to clean up after itselfHeres even better part H can pretend to be anything between mysql to Postgress with stuff like mssql oracle and else in between +Hi Stepan Thanks for the patch +Yes it seems to work +Patch is applied at +Just wondering when it might actually get looked at +Agreed. +A parallel set of implementations is certainly an option +According to logj The WARN level designates potentially harmful situationsI agree that this is expected behavior by design +I have not yet tried the new to be fixed now +Ill post a proposal patch with the implementation and some tests on Friday +Pat Hunt is back today and I hope he can help me out +OK here comes my first approach which doesnt seem to work +Look at the red cycle +No +Thanks Pete. +These already existed in couchdb +We changed the root artifactId to since the root artifactId controls the name of the release packages the sources jar and the jar +looks updated patch +This is now fixed with the changes from at revision +Add licenses on commit +Where does this code snippet come from +coolio +I havent had time to investigate further and wont have time this weekend +When I checkout the source code using branch and build from it everything works fineSo are we going to redo the release process for or are we going to release + it was in the bag we got at A small we use the flexible project API it is probably supported in some sense There is though one drawback from this namely that if you change either packaging OR the WTP project you will have to maintain the other +Sorry been swamped + for the to trunk. +Ill keep that in mindbq +apachectl startClosing issue. +Thanks for the patch btw +Bulk close after release of +A redirecting node will ask the node it redirects to for the the interfaces +The bug was fixed in July +It seems reasonable to me that we should ensure weve contacted a seed during bootstrap when one has been set +These issues seem to report similar condition for older releases +Also included in the patch is fixes for some test cases +I just committed this. +The problem is actually with NetSSH and it will consistently fail whichever version of you use because it incorrectly detects the platformAny ideas on how we can get it tested so it doesnt break again +Ivan thanks for your patch! It was applied at r +I dont want to set the color to black as then we will lose a visual cue that the input is disabled +As we add more operators from it doesnt seem to make sense to have a separate symbol for each one +Closing for release notes. +When I answered that dialog Igot CTemp + has changed since the first patchIm attatching a new one +Revision . +All my slaves are syncd using NTP so this would work better +Run tests first on dev machines and then asks QA to validate +I need to ensure it doesnt conflict with tokens +Ok looks like I was able to solve the problem +Thanks Gert for the patch +Ive tested this patch against SMX unit tests and its included examples +I meant how many wed finished the flush process on +However I dont know if this is a workable solution for you also the argfilesOnly property does not seem to be documented on the properties page of the aspectj plugin I will add that shortly +This issue is around enforcing the conventions of mapping and populating Ids +Done the first round of this +If there is a wiki entry on how to convert existing tests to or if there is some work underway let me know +I could commit this but I have no idea what I am doing can someone look at it +You would need to show a video in the monitor while recording then I guess you could see sth +Its essential if a eclipse project was set up using an target path outside eclipses project environment +Lists moved +nkeywal nice one +Note that is fixed in CP too. +Workaround to SWT oddness added to +I modified the TTL test case to use timestamps in the past to test memcache expiration without requiring any sleep + to Martins suggestion +to reproduce +I assume the issue is fixed +Im happy to backport this fix +Leave the issue open in case we find issues with it once its delivered +Applied patch at r. +This should be trivially easy just make it call counterclockwise from the oracle filter sql encoder +This might be an FAQ or How To item if this is not an actual issue in I applaud your effort details and drive to get this resolved! +The root cause of this issue is the fact that a broker that is not really ready to serve requests ends up serving requests and misleading clients +Thoughts +So I am closing this for now. +Ive added the tweak to the rules Actions tab that the InputsOutputsGlobals tables are only enabled if you have an Action selected +er +Ideally is for GFAC service +Tests look good Nicholas! Do you intend to add unit tests to your patch +Then I ed B +Probably this was the consiquence of but not shure +Updated the patch to add comments and assert in to justify the number of allocated containers is not zero +Verified on. +If you beleive that should be please and contact the program team +Yes +Patch attached with an implementation of this +Hi Sudarshan sorry that was my bad I had svn mvd but when I created the patch I failed to pass to svn +ah yes fair enough +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I think it is okay to keep the new message +now there is a for the component that does the rendering of the children automatically +Bulk closing stale resolved issues +FYI the search box is larger than expected on IE +A determined user could probably work around them but Im fine with thatI also ran a script on the DB to reset all bugs to the default assignedto +Ill spin a reworked version and commit that +I had a very quick look at it and I think I like itSome things I would change arei I prefer the original definition of what is Apache Karaf +Patch addressing comments from Bikas +I created and to track the +change in bugfix as +Couldnt we just check for the type of merge scheduler and on the right types add the right mbeans and avoid introducing this new solr specific impl +Thanks for the review Kihwal +build updates backported from trunk to everything it setup nowWill update the release process on the wiki as I perform the release and any required updates against this JIRA if needed +gtr invalid backslash escape at end of stringecho i tr sed s +So voting on the downloaded package not the SVN tag +The ord in one segment isnt equivalent to the ord in another segment +The fix with refresh is not so difficult and critical but should be tested carefully so it can be moved to +Am I doing it wrong +But if it is relative to the number of modules andor classes it would be a big hurt for big projects +Java output of test script with full stacktrace +Its creating a copy of the model now +Looks like this issue is the same as richdatascroller does not reset when data model is changed +I hope this can be improved +Hi Sujit +Any chance of this getting merged and resolved in the near futureIf I can do anything to help let me know +I dont think this more problematic than the rest of this horrible code +Attaching stylesheet needed to reproduce the problem along with a sample input document and expected a patch that resolves the problem +Also Vivek from next time please keep the name of the patch same as you updated last timeIt helps in identifying the updated patch easily +Any objections against bug No go ahead +But and will never implement our interfaces so you would be forced to create whole wrapper classes around every class +r +I am still not entirely convinced if we dont have a problem here +There is a property called but I cant find it in the documentation +By the way it looks I will be able to do so tonight +Is there an or something which handles the magic +We should leave both just mark the older one as deprecated so no new customers use it anymoreThe customers that use it should have a chance to change their systems in to use the new way of setting up the param +Can somebody help meThanks in still need to do some research to understand more fully the interaction of locales and input streams and parsers but I am still looking into this +But there is another problem +It is clear that overlapping copying and merge will pay off a lotParalizing merging will pay off too +Could you try this project +helloworldbridge +I have added test cases and have uploaded a new patch +In Java the situation is quite a bit better but still not idealA failsafe workaround to these kinds of issues is to run the PDF processing task in a separate process like Ive since done for example in. +maybe even with some text explaining that you need to change the name because it conflicts. +David Kerber has started work on this issue +Therefore it doesnt find all the properties adding all the missing properties to the local +I am on vacation from to If you have any question on deployment and JEE bugs please contact Saurabh Arora or my manager Maruthi NuthikattuFor emergency contact me at. +Currently collecting more information +Canceling this patch as it is over a year old and no longer applies to trunk. +So it doesnt seem to me that the limit applies here +Unfortunately its still a dependence on a SNAPSHOT plugin but when it turns we might use in geotools as well and throw away our version of the plugin +Thanks Marius +Working with the Infinispan team to get a more concrete test defined so we can stop discussing on theories +Thats the goal +sh versions +Attaching stack traces +Added simple weighting +I checked out activemq to double check and the wildcard is present in the nativecode sectionIts probably worth noting that the karaf instance is created by the karaf tooling plugin so perhaps something something is missing there +Users have to copypaste the commands into a SQL client to upgrade? has this set to resolved in +Could you pls tell me where I can download this patchThanksAshwin +bq +Thanks for the comments +Im with David on this one +buildivylibcommongriddevbin +The idea behind your code is ok but it should work also with raw visitors +Ir converts the first livecolumns n columns to tombstones +patch +Thanks Sijie. +you can use it inside any of the spel aware components so filter router etc +If the provided context to your converter is of this type you can cast it and call the getPath method +Ill look into it though +The patch does not change the Input Output Processor +OK now Im done until further review +Ready in a month +Bulk closing stale resolved issues +NET Framework IS required for the already got +To make it more usable can we get this +Ill try to get this one fixed for as well +contenttype +That means I no longer need a rabbitmq or something like that to put stuff into Elastic Search +Tim also added code to remove before calling is a bug in remoting +Applied at rCould you please verity itThank you very much +An alternative would be to mark these indexes as something else than unique when constraints are deferredIn the deferable case I think these should be just standard duplicate allowing index and do theallow duplicate checking in language +Proposed patch Second Password also needed when creating accounts move ASF query check a bit later where we know if it is a new user or a shot to commit +Works great now +If you want to take it instead feel free +let us choose a name and close this soon +bulk close of all resolved issues. +jar from Tapestry application to GLASSFISHHOMEmodules directory restart Glassfish. +It runs in a loop sending a message consuming a bunch of memory then sending another message +Fortunately the vast majority of the working group members and WC members voting there the reviews by the PSIG and lawyers on that committee and the stated opinions WC staff overwhelmingly supported the use of as a license for despite that vocal but minimal opposition +NSA is looking for mutually acceptable terms via which its employees can be allowed to become committers to Apache projects as part of their official duties +attaching the fix +RobertI understand +Username sporte +Both File and HDFS sink support text output +I agree +all the tests pass but it could use its own tests tootheres a nocommit mentioning what i dont like +hmm it would be nice if there is a way to make both filter caches request context work simultaneously +This behavior should be fine if document that +But if its too hard thats understandable. +I think the execution time will scale with the square of the size of the comment +I think Oleg took care of this so Im marking it as resolved +MarkWhat is the status of the? Thanks +As the internal list should not be available outside we have no problem anybody else changing this uncontrolledSee also issue +I suppose if I reboot and clean all my caches it may draw correctly but who knows +I will definately give a try with new release +And there are probably on ly or handlers per service +Fixing Affects and Fix in Version +right +confirmed with FOP and TRUNK rev unsupported format +The testcases for the current fix seem thorough but if it helps I can do the testing with my own data as soon as a new patch is available +When I test I use containingtruetrueWorks for me +Works for me for now +As side note the surefire fork mode is set pertest because Ive meet unexpected results running CDI containers multiple times in same JVM instance +Now bug is fully resolved +We can change to make find stateless +Can you give it a shot? And let us know what you find please +Is this a duplicate of +Daniele Simone opinions +is only used to store entities etc +My guess is that the exampletest does not display the weekdays correctly if your locale says that the first day of the week is some other day than Monday +For DN having couple of million blocks it reduced block reporting time from sec secIll resolve this issue as a +The shows a failure for deployment of the helloworld quickstart starting at line +committed fix to trunk as change sibmsvn commitSending javaengineorgapachederbyimplsqlcompileAdding javatestingorgapachederbyTestingfunctionTeststestslangSending javatestingorgapachederbyTestingfunctionTeststestslangSending javatestingorgapachederbyTestingjunitTransmitting file dataCommitted revision +Now on staging and soon also on production +Do I have to update some maven files if I change the jar file? Im not a maven person so please help outsimon +I think those are the least likely to cause issues +In the old situation it would be overwritten with the platform version while running prepare +Patch that moves non generated java source to srcjava +Thanks +Somehow my thoughts got truncated I meant to say aboveI can think of an approach for trying to deal with this where we could determine +So are you trying to say patch my code? please provide me the trunk url +Whoa! Big issue description! Perhaps you could have gone with a link to the mail archive? Descriptions are included in every message about the issueIn any case I think perhaps each recipe deserves its own and should hence be a separate contrib module +Backported in r r +It worked +Assigned to me +These are test program for reproduction of the phenomena what I found and result of the program +Attaching a patch for testing the readme files when logDevice attribute is used +New patch addressing Joeys nit +svn statM javaengineorgapachederbyimpldbM javaengineorgapachederbyimplstoreaccessM javaengineorgapachederbyimplstorerawM javaengineorgapachederbyiapistoreaccessM javaengineorgapachederbyiapistorerawM javaengineorgapachederbydatabase +Anh Vu does the issue fixed for your caseYou have reopened it at April +Any reason not to use an OUT parameter as the returned state +Fixed on and closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +I think for the safemode part as Vinay mentioned the key issue is still the current code fails to recognize if the file is in snapshot and the deletion is on its parentancestral directory while loading the fsimage +Well need to fix most or all of these before upgrading +Added the new configuration option in revision issue after bundle release. +Scenario to reproduce the issue using +Sending componentssrcmainjavaorgapachecameldataformattagsoupSending toolingmavenSending toolingmavensrcmainjavaorgapachecamelmavenTransmitting file dataCommitted revision After much pain and suffering I found a solution using tag soup instead of jtidy +My and Juergens comments were about points and from my description +Id rather not block exit on writing out a file +I fixed it so I removed the quickstart I previously attachedHowever I verified that my patch does fix the problem in my real codebase +As the sort should happen before the Row is actually constructedOk there may be better ways of extracting this info Ill digg into removed the get method in revision +This was actually fixed in and the additional backref issues are filed separately. +keeping a count of versions on a basis instead of +tried on buildusing geronimo runserver startup successfully +Id change the default to but its a back compat issue Id rather not deal with +Oh first gotta wait for jenkins again +SanjayCan you review the patch I have attachedThanksgirish +jakob you can always implement reader of customized data in a rd party lib and let hive load it from there +The bindings can be different +What work remains to be done on this issue +Applied the part of patch with thanks to CedricI changed the part of how to get the address from the and also add the spring configuration file of testing which is missed from the patch +Also I dont think it is feasible to limit this to only the primary data owner +Hi Julien Im assuming this is required for proper relative link resolution when using something like the right +I believe that this particular issue is resolved +All versions are wo sources and javadocs +The fix was already included in as part of the mods to accommodate the Translator Overrides panel in the VDB appropriate UILayout changes were incorporated into Teiid Designer and checked into +Think that we need to consider more for the changes the original logic is that it firstly checks spec deployment plan file to determine the application type then read the corresponding Geronimo specific file the real problem here is that with the new specification those plans are not must +We can still change the mixin node types +so its really a step backwardsCant we just use the same UI and update the project settings with it +this is not done we currently overwrite the file we can take the approach of write new file delete backup file if present rename current file to backup file rename new file to current this a couple of weeks ago but didnt know there was a JIRA on it. +Assigning this ticket to Charlie for review +There is no need to create requirement. +Why does the example schema define as characters and the Dao truncates it to characters? are using and there the length is cutten at in and it looks like it wasnt configurable +You are saying that datanucleus and jar causing dependencies and we do not need these persistence jar files +Note that this wont work for FDSOCK as multiple FDSOCK instances may use the same config needs freedom to find an available port +However the problem has gone when I tried a clean svn checkout +but this JIRA is not about cleaning up imports just javadocs +We might be able to use the same handler but itd need one of our XML experts to weigh in and say for sureIve fixed a bug in r where startend document was being called for each htmlxhtml file which solves the use of metadata from but doesnt affect the text extractionFor the mimetype subtype I thought it made sense as both use the same structure and a very similar format +svn commit m Added check for methods that require the to be open but where the Broker would throw a different exception srcmainjavaorgapacheopenjpapersistenceSending srcmainjavaorgapacheopenjpapersistenceTransmitting file dataCommitted revision svn commit m Added test for throwing after srctestjavaorgapacheopenjpapersistencesimpleAdding srctestjavaorgapacheopenjpapersistencesimpleTransmitting file dataCommitted revision . +Heading home today and once I fix the build on Macs will check it out +With this patch I get failed JVMTI tests Breakpoint and +Use one of the existing source files or insert from an existing table like src +I just committed this to trunk and. +This will avoid having to carry the privilege list from compile time to execution time and will make the grant statement a bit faster +Andrew are you still opposed after seeing the patch +Thanks JimPatch applied at r please verify +Bumping issues to except those that involve with Struts or +If you can help us with creating a smoke test for Giraph itll be all we need +This patch makes the client and server exchange full timestamps as happens in the embedded situationAlthough the regression tests pass this patch is not ready for commit +jaxrpc saaj webservice webservice passed without failures +Should have a field type patch today +Hi Jene its always possible to exclude sources or packages from quality control as mentioned in nevertheless this feature of reusing existing build results is still important +In a controlledproduction environment we cant give access the best is to fix the above code to properly fall back on the folder! Two things +Since I am pretty confident that this is an EMF bug I went ahead a submitted EMF bug +Confirmed +Patch with fix to increase default to from for Memory Channel along with corresponding update to +The problem is that the xpath is evaluated as an expression and not as a predicate in the filter EIP in Scala DSL. +on patch +Version update can be done using a linux grep commandCurrently both Eclipse plugins contains a common sub set of jars embedded in them + +There is no development on and anymore so this is not relevant. +Any insightsexplanations +Thanks Krishna! +OK I cant see me to get any changes to take effect so perhaps this is part of the overall reason Ben left open +reverted +That makes sense about the UGI +The issue is gone after using CP +The clients also use this key to figure out server principal +The patch does not introduce any new Findbugs warnings +let the configured handler take care of it +This is fixed after removed the information from job list +From its name you would not know what it is forbq +Yes I had gotten used to seeing this so in the end I forgot to fix it +Updated the Derby website Added the release of to the news section on the front page Updated the pointer there to refer to the upcoming USA rather than last years event Fixed a broken link on the release pageCommitted at subversion revision +The point is to minimize the extra serialized size on disk +fixed in for contribution +Removed as a fix version +Changes look good +This may be presumptuous so reverse it if desired but I am quite sure we not only shoudnt but cant use Java for the project in general. +Lance can we get any more info on this? The exception in question to start +I would definately be able to follow this if I was seeing it tutorial pages themselves are excellent +Patch with testcase +nit I dont like the name copyStream a tad too generic for the specific function its doing +This patch uses a new filename on recovery +I was not able to reproduce it Brian +Fixed in acaa. +Closed. +Im sorry to hear that but I fully understand you have to set priorities +thats good to know +This is actually what needs to be invertigated what is in the tar ball produced by one more from +bq +Thats the only difference so I will commit if the unit tests pass +My current idea is to have each btree use a pid between btree actions and the disk +I just committed this +Sorry I think this is because of a change I have recently made to the configuration file +mistake I suppose there isnt option move issue +The Inode object is a DFS object and has no place in the fs package +Since MR is ignoring exceptions will see SUCCEEDED but on RM side since unregister event is not received stays on running and eventually failed +would be cool to provide some easy way to use them as is done for example for the Solr additional analyzers especially snowball which should be part of some i integration +test compiled successfully and ran correctly on my local machine +Can you see the problem here? Maven will amount to nothing if you dont get out of the way and let us upload libraries to the official repository. +This just shows an example what needs to be added to a projects to run pmd +was released +Hi here is a patch that +Attaching a second rev of the functional spec +the list of parties already show the clients and provider parties so is there really a need to split that upRegardsHans +Patch and tests to enable resolution of default virtualhost name when defined in an external virtualhosts file. +This should not happen until there are no other nodes in cluster to run the task +Nice find +Good ideaAnd finally we can send to all peers the exit message in the same way if when the sum of previous updates counts from peers is equal +to elaborate on the DB being removed the JNDI connector also documentation added +Never mind the patch actually broke the build because some classes like for example are not available in java the only way to build is to use java +There might be no cheese down this avenue +Patch implementing this eclipse filterThe id values seemed to be incremented by so I just followed that convention +Shouldnt RO support only be enabled via an explicit configuration option +do you have another link or list of tests that failed +jstack any chance you guys arent closing resourcesremoving listeners properly +Checking the format only works if theres a definition to compare it with +I also had to change parts of but just fixing some compilation errors that occurred because I no longer needed parameters everywhere +Vladimir Ive committed the fix at the revision +I did add a test for this specific case which ideally would catch the same issue in the future +but the winner wins by some very arbitrary rulesyou cant reason about the final state +This is fixed in revision +An aside if youre cleaning up the C code is were dumping the wrong floating point registers +If we kick the connection how should we handle this with HA? still disconnect disconnect without recreating the consumer on failover or should we maybe just close the consumer rather than disconnecting the whole connection +Weve already deprecated and are about to remove our own implementation in favor of using one of the other implementations nowavailable +Thanks for fixing this issue Swapnil +While some issues in here are easy fixes some are also issues with Publican +We already have a pretty error message page with a place for a message and the stack trace presented in a nice window with a scrollbar +Will commit if tests pass. +Hi Freddy the Sonar upgrade to version solved this problem +Adding this comment to make this easier to find the error message is below +I think this is not just with also have the same issue +Still no support on the searchresponse side but groundwork for adding multiple fields per is now laid +git is already present in the EAP buildroots +Updated summary to indicate patch is patch applied +Attached patch for patch for trunk +Ive added a page to the Struts Wiki mentioning this issue +I dont think entries need to be or do theyI thought it would also be nice to print a stack trace on shutdown so added that too but left the property name atI have not run tests and will be out for a week so if someone wants to pick this up run tests and get it committed that would be greatKathey +Looks good to meWhile we are at this maybe we should just get rid of the old calls and pass default from ql to the metastore calls +See my steps above. +This would be more efficient too +I only get it when I edit the file in eclipse +Looks like youre right Len The solution is to simply remove the definition from the plugin this facet is already defined inFixed in trunk patch for branch attached here for review +Change merged to the branch +I will take a look at it +My major is Computer Software and Theory +see the attachment to for code that I have in that does this +The thread local still scares me +Rob could you review this please +weeds out any items that arent jar files +Go ahead to commit Shawns patch to trunk at +I can see both maps in my geoserver and each one in the declared srs +Looks like this is likely to happen as part of the SGD frameworkWould you like to help with that +Tests on dev and dev showed this works +we incorrectly ignore adding whos package starts with where did you see that? I thought it was +The query value was changed to instead of Integer to save a little on object pointer allocationAs a side note there are a number of unrelated generics warnings when compiling +Closing this issue. +This is corrected with this patch +Are the total resources of the cluster adjusted downwards as containers on the node complete or does the cluster run above capacity while were waiting for the remaining containers? In that sense this seems closely related to +Patch to return a code when an exception occurs +And please feel free to backport the other patch as well +So when Im going toquery JCR my expectation is to deal with paths Im using withhierarchy managerReal caseWhen Im going to query along path of the form AlevelBlevelLevel thispaths breaks XML Specifications an parser cant build the treeJCR Specifications says query must be run against xml document representation of jcrstructure +Hi Nicolas Great I was studying the to try rewrite the test +Because of this external maven repository it might last a long while till the updated version can be released +may we do this before the simpler patch you refer to? Also it is consistent to other windows related patchesThanks a lotSent from my iPhone +Maven project to reproduce the issue + +Thanks d like to make sure your team and ours liaise on this one so we can ensure correct versions are being used and also give us an opportunity to review your teams work before going live in EAP +We can definitely use more documentation help +Change should be available in ER +If the problem is not solved can you please post a test case that I can use to resolve the issue +This is a big patch can you post it on reviewboard +Ive coded what I told before in the last post +Have you ever tried activemq itself? I just want to make sure its not a environment problem +Patch applied to Apache CVS main branch and branchNicholas please verify this fixRegardsBrian bug has been fixed in Xalan Java +We provide the auto as a convenience and if theres ever a problem with this then the only valid procedure is to manually set up the database and set auto to validate. +committed as svn on trunk +I do it all the time and it drives me bloody nuts to have to go around changing method names just because Castor wont match types on signaturesHowever I have a distinct gut feeling that this is going to create some very interesting behavior if it gets fixed +Dimitris I assume that EAP has already been taged somewhere and this should only go to build +Committed revision reopening to update release note info. +Merged to the branch in revision . +Hm I dont recall a JIRA ticket related to that change in July did it just get changed without a ticket or discussion or am just forgetful +Not sure why different jvms will instantiate session attributes in different order but this is something that should be fixed +Besides Properties view also throws exceptions in that case +With coming in we can use binary format for both searching and indexing +we call which internally may call its writeBytes but not +I meant especially and +Waiting for confirmation that our Corporate CLA has also been successfully submittedMy changes to the message serialization can be merged at will as they should be covered by the ICLAThe changes I made to the driver and stress application I have donated to the driver project so you can use them as you wishI believe the write batcher could be incorporated in and then used both in the driver and server without any legal difficulties +This factory method would care to have the thread running and the flags cleared as well as having the use counter managed +This works when the AMQP JMS client is updated to the. +I havent tried to instantiate on pushed the fix abfd yesterday and probably the reported exception wont be raised +Attached patch +Thank you! +The following scenario does not work Have disconnected application Try to create new app First credentials dialog appears Cancel Second credentials dialog appears Fill the credentialsRESULT Wizard with credentiala page appearsEXPECTED Wizard with Select Application page appears without the need of additional credentials need for credentials wizard since the application creation wizard provides its credentials page too +For example use the dash when separating parameters the slash when separating methods or something similar +retesting +Thanks for the patch Jonathan! +Wont it lead to a lot of small files MB assuming the reducer maintains the same size for the the option is already there +Defer all unresolved issues from to +Discussions on LazySelective field loading indicated that people wanted more control over the process +I should have mentioned that my home directory was set the the xml directory and the command I used waswsdljava wsdl wsdlt this a duplicate of ? Let me know if theres extra info to consider here +Just to clarify and save others the time this is fixed in nothing prior on the branch. +Page Design Options for external files was disabled toolbar items creation was updated disabled icons were added +Is it possible that hudson does not do ant clean? This test has not changed recently but turn it off while I figure out whats any update on this? AFAICT there is no way to turn it off in the precommit testing without committing a change to the script and of the project +The file should always be there in the jars and you shouldnt be hitting this at all +Hey Ashutosh I have been busy with some other project priorities +it could use another pass + has the same issue +Hello PeterIve attached a sample project just run grails to see the remove the need to have the version attribute in the class if possible +Reopening to resolve well close when the release is out. +RightI mean the end result is the same but IMHO it is more concise +Im forked pretty badly at this point with all of my recent patches so its hard to keep track of what needs to go into each oneThanks! +Yoko could you give any details on why this takes longer on +I would assume this is fixed so I resolve it +Input data for data sets given here are too small for Pig to split keys into multiple reducers +JCA was not final +Also this is w +Reopen to mark affects version + +lets expose in nodetool and call this goodAlso naming it something that implies an upgrade would be helpful would like to hook into this upgrade process and it isnt really Yaml related +Hi GertDo you have an update on the state of the tests? Are somemostall tests executing successfully on the trunkWith the sole exception I mentioned in my first comment all the tests executed successfully on trunk +dialog and select yes or no on the Publish mytestapp? dialog as shown on the attached screenshot +implemented +I will commit as soon as I get a +If these were threads that are fighting for shared spindles we might consider limiting the number as we do for +Great catch and my own fault +This was exactly what I needed +So all issues not directly assigned to an AS release are being closed +Robert lets enable and by defaultThe stopwords list in the Lucene analyzer looks too small unless its always used in combination with a stoptags filter +Patch appliedGood catch thanks Aysegul! +Im copying your latest comment back to the original issue and well pursue from there +This provides a clear visual separation for people when they are using the official API classes and when they are using the classes +I am sorry I was using the version +css file I could see in Open With only CSS Editor and Text Editor +Indeed +Tracking references across projects would be painful and mostly unnecessary +Verified that native code has listeners in place and fires appropriate callbacks to update properties on the object with the right data coverage type. +bq +Initial version of this is your test plan for this? Id suggest something involving html unit while jobs are pushed isnt the right choice here since Im not generating HTML +changes based on Doug think this should get from the Hadoop Configuration object rather than from system properties +Dtestjaxwsjbws s been fixed in the container integration and will be shipped with AS +Ive edited the master and files to implement on my local machine +bp file theirautogenerated C declarations all end up in the same file + in JBDS h Linux +I am considering using and Push to push data to web clients +This ended up including a first pass at the atom table stuff as well +needs to be open until change committed to trunk +That was a bit harder than I thought but the logic Ive described applies +I think Elis concern is about some releases which did neither +Thus I think this can be resolved. +nt blocking were you ever able to test this +Fix by Heath Borders +I killed the JVM approx times the last weeks +But if it are pages then it is not that easy because then we also have that in the resulting page as far as i know +ben Is this just a release issue or we need changes to our ant builds +Closing this as duplicate. +The upgrade to Hibernate fixed the problem +The old one for back compatibility the new one for newly created pages and replacement of the old one +The JAXP Expert Group confirmed this interpretation meaning that XSLT processing should begin with the node that is the argument to the but the input tree is comprised of the entire DOM tree +Hi DhavalAs mentioned in the README of this sample looks like maven should be used to build the sampleAlso looks like ant will not be used for running the samples tooProbably we need to exclude the from the distribution +Deferring could solve most common problems but might conflict with when for the first null values in a column a different message is pulled from the resources +Trivially adapted version of Harmony implementation committed in r. + +The primary benefit of consistent hashing is that it reduces the amount of data you have to physically move around +Patch committed to branch and trunkBrock Thanks for the contribution +what does it do? +submitting +Andrew please check +Were going to have to push this simply because that opens a BIG can of worms and wont be able to get resolved in this time frame +Keeping time zone issues in the users codeNot sure what the value is in dealing withcasecasecasecasecasegiven that this is a private method +SVN +Chances are high the shutdown process may not take into account parallel forks ATM +Pacth Robbie can you review this change please +nothing we can do about this before GA leaving here for triage +If MANY schematargetnamespacepackage name mappings are specified then push types to specified namespaces +Im new to both open source and nutch so Im not sure of the right way to submit corrections +Latest revision of patch +Hi RonAny chance you append a testcase? So that I can see whats the fault message looks like in the flowThanks test case is attachedExercise the test by opening in a browser changing the personId value to RE then clicking the Send buttonLet me know if you have questions. +How come that is not in Tikas trunk? Nor it is in Nutch +Behavior would depend on defaultexternal +The macro definition doesnt seem to be there yet and the themes dont appear to have been modified yet to call the macro +the fix for unresolved count is not accurate any more since i wasnt aware of the Closed status used by count for unresolved apparently includes all tickets marked as Closed +Let me try to meditate after two cups of coffee +Configuration parameter also in trunk for FLEPlease review +For this one I can use attachment +There hasnt been a change in yet + am not sure what is the point of removing the keys from the iteratorIf the noof items in that list is huge we want to leave the objects to revisionthanks Ali close for Solr +java maybe +mark can you please sanity check whether this was an intentional change or a fluke? is the original risk in now eliminated by some other change to how the is used +trying to pass the security context already on thread to propagate to new session +Please check that it actually helps with performance +If you want a more permanent solution I suggest creating your own copy of Wicket with the markup files modified to your liking +bulk close of all resolved issues. +Comments from JIRA bug Triage on Febrary Patch needs to be more robust to be taken into Other parer options need to be considered such as validating This is the Process command +In other words each form would have its own cache entry instead of each form file being a cache entry +Do you have any Closed issues for the current version? +Please see the attachment of das web client for new files these need to beadded to sandbox svn +a diff of kb is pretty big for a trivial change changes seem to be ok except the additional constructorfactory method ini will take a look at it + +Nicholas thanks for catching that +Can someone please let me know how I update my Cassandra with this patch +Confirmed that the user is indeed doing Proxy authentication incorrectly since unit tests confirm that this is fixed closing the issue. +Root cause may be? I think is just symptom though we should guard against it +HiAttempting to ping fails with Destination port just tried editing the but only user servicestudio has write access +shared libs dont work for me +Cool +compilation error while installation spring security core dont set time estimates as the instructions state Only to be filled in by the assignee before starting work +For example above I tried to reference but it was not +Is this report still validIm looking at the code and anytime adds anything to the security registry it unregisters it after the the original report what was it that was retained that caused the memory leak +Yeah including those spec jars in the application packaging will lead to classloading issue and can cause problems like the one you sawClosing this JIRA. +Started some of the t get the validation messages to stay through a +Hey devs are there any further comments youd like me to address on this patch? Do let me know +Need a good text for warning message +I also ran into the corrupted source paths described above +I have attached it +In case of a particular failing testing returns an error code +this issue is outdated as modssl is now included in the installation command +However for that to work we would need to keep references on the sstables after the initial flush which adds its sets of complication if for some reason a node never receive its you can start computing your tree message it will keep some sstables around forever +This created a new file named trunkcoretarget +It would be very useful if this could be configured in a way like +Apparently there is a bug in the jvm under windows so that the files are only deleted if they are closed +Its also in rpm package +The class keeps the entire thing in memory before flushing to disk +No response in the last months as I guess that this is fixed in the latest Cocoon I set this bug to fixed as wellPlease reopen the bug if Im wrong. +Isnt the status field in? Otherwise looks great +Maybe Mike can help what the messages mean +maven doesnt set it correctly in exit B errorlevel at the end of ll add a faq for this pb +The specification now statesPortable JDO applications must not depend on SCO or FCO uniquing behavior nor on the storage mechanism in the datastore +If I try to rename a folder to con the attempt is totally ignored +should no longer be using properties in the namespace +Patch for attached +All the methods were added. +Sending some emails on the dev list to clarify things +ubuntu latest only known workaround +Issue is resolved by commit r + also removes the use of junit in +Please verify and close this Jira if its. +do we need to make changes to accomodate has no impact on current labels it just fixes the way the response files are saved in the filesystemWhat would be nice to have is to show the high value report if it exists but we can consider this an enhancement issue separate from this one +Modified startstop script with permission set for StartStop script +Also I found that contains strange characters for exampletext If true sets state of the item to disabled state +Yeah Id agree a year is a bit much +Add dependency on was done by Bill. +svn ciSending demosbrandingsrcmainresourcesresourcesdemosbrandingshellSending demosdeployersrcmainresourcesresourcesdemosdeployerkarTransmitting file dataCommitted revision . +Thanks to Julien and Gabriele +The examplecode for using was wrong wrong constructor for wrong constructor for java code was not valid for creating anonymous arraysSee the attached patch + to trunk +I agree with Max +In your version only first declared pattern would be taken into account +We dont have Ant task. +I can get xml mapreduce plan with the patch +Sorry the idea is simply to make positioned reads I understand that +Tom thanks for reviewing +This happens because the tests before the failed test does some regionserver abort +This should probably be closed since we have the plugin mechanism plus the and Remoting plugins +Effects the Catalog Product Screens and Product SaschaYour patch is in trunk at r +Looks reasonable to do for the next s excellent and thank you very muchWorks for me now +Attaching patch in progress that hooks up RC to Bigtop renames into cleans various bits and pieces upStill TODO updates to the puppet deployment code updates to the integration testing code updates to package testing codeStill this should be enough for anybody to try the new flume in the Bigtop context +I committed this +Note that Shinobus test didnt work since the method returned a null +To be clear does your fix mean that will be the new JPA implementation shipped as part of? Or does it mean that if an application supplies those libraries they can be used +I will follow up with them +Is caused by my implementation for defaultSuch a simple example and yetWill get back on this tomorrow +this patch fixes the bug in trunkunfortunately the riskbenefit of backporting this to the branch is past my threshold of comfort +committed thanks namit! +The s transform goal has this feature +issue has been resolved +This is just one of several place I will have to look into to a make sure we do the required checks and b as in this case make sure we do have the right doPriv blocks in place +See last comment + +My apologies for raising it as a bug +Thanks +patch to will work tomorrow on it +Fixed according to Fred +Verified in +This is caused by the fact that we use a callback server on the local device for the bridge and we just killed the routing table by doing this + New test Add getter that returns the object Add check for scanner null before trying to close it Enclose test body in try catch block so that exceptions can be dumped to the console at the point in the test where they occur +Actually apt does specify the format of an anchor The name of an anchorlink is its text with all non alphanumeric characters stripped +Should it be patch available or there are still things missing +bq +I would prefer going forward to use the new antlr library and not continue with the meantime I tried to merge the pull request above however Im not able to now as my recent changes to the code base to improve the optionContext strings in s seems to now prevent automatic merge +The JBI component should not use getExternal endpoints are usually activated by BC for consumer endpoints for example a soap BC which will accept incoming http request can activate an external endpointTo resolve one should use +Selection should be revised additionally. +It should be noted that does support body content so it itself can be used for in purposes right now +Thanks Konstantin Ive the issue +Added the bpmn AjithMy initial guess is the process variable that doesnt have the type set look at the screenshot +Ive seen this problem happen before +This issue has been resolved and a patch posted to. +bq +Ivan Sijie do you guys have any other insight to add here +Triaged for +If this problem goes away with clover Im inclined to just close this out rather than spend more time on the possibility that it is a Sonar rather than a Cobertura problem. +This version fixes that +Could you please download this snapshot and see if the manifests are correct now +The current patch will fall back to CHUKWAHOMEhadoopjar for the hadoop jar file +Attach a patch +Thanks for pointing this out! Fixed through changing s interfaces Set into a List with additional contains check to avoid duplicate give one of the next nightly snapshots a try if you have the chance to! It would be great to test this with an actual use case before the that thanks +I committed your patch anyway +This is Exceedingly Bad +I just reenabled minutes builds on hudson +Committed thanks Robert +I owe you a beer +Hi BryanThanks for identifying that typo I will correct itBryan You can avoid setting up the CLASSPATH +Awesome thanks +Im changing the title to reflect the origin of the problem as it seems it could happen not just during unit tests but also when doing a normal shutdown of a master +I contacted them about it before they said it shouldnt happen again +What is the use case where you would need such a behaviourThe idea behind dropping the directory is to be certain that you start on a defined base line +Tapestry allows you to use templates with unquoted attributes +adds timeouts if the tests take too long +automatic git apply +a related point is that we currently generate explicit initialization for blank fields that correspond to Java primitives and for objects but dont do it for other X struct types +Hold that last patch +Thanks a lotIt works like a charmRegards! Tony +You did an excellent job +There may be errors in the comments +here is a patch that adds estimate to DISI +A few questions Perhaps we should make the message more verbose in the event we fall back to kill ? Im thinking something along the lines of Daemon did not stop gracefully after signaling it X seconds ago +batch transition to closed remaining resolved bugs +Proposed patch +So the Thai problem was also a whitespace problem +STATUSOBSTRUCTED is funny +And so I am happy and close this issue. +If people complain lets just write another renderer +I am not sure if this requires an OK from a project despot but i am giving an OK anyways +Provided patch for version +Alternately we could just try to launch Maven from the highest common directory? There may or may not be any POM in that location but it should still work I think +git ffadaebdcdeeebErm +First Ive updated all of the project example zips to remove the SOA or or whatever line +Thanks for the patch Ill take a look +Like ordering the callsThe mock is only a proxy to get to the control +Hi Keith could you review these changes pleaseI have made the connection checking operate the way the ones do as using the connecion instead will need further work to take the federation linkbridge functionality into consideration +Heres a patch that implements HADOOPCLASSPATHMichael is this acceptable to you +this is because the bundle is not startedThe system bundle context can be injected like thisInjectpublic! I should have known that +So Ill either have to debug from source or see if something like ethereal can give me a quicker answer +do we want to semd a message w the version or just infer it in +first half of review really only questions and requests for more comments code so far looks fine +So this warning should not be a problem +Fix findbugs warning +I would so love to see this happen! +If you dont have or instance available then we can just resolve based upon what you have in +BTW Im not sure this is exactly the same problem as as that is talking about user defined classes whilst I am talking about serializing the handler instance along with the proxy and reattaching it when deserialization occurs +the doc is for the wengine branched version which is pretty much same as the trunk except for reserved metadata fields which are named slightly different and file staging does not exist in the trunk version +IMO we were missing a core component for frames so Im trying to resolve this ticket introducing the Frame component +on latest patch after some testing +As Hyunsik said above you need to understand how Tajo works +Theres probably some duplication and Im not sure quite how to distinguish between what to do with and without +Attention EvaNo problem about the delay +Fixed in commit c +It is closed now anyway +Patch v move all logs to a logs directory to avoid polluting the main directory with a half dozen rolling logs +I tried to reproduce this problem by using the lightweight HTTP server example in the sandbox and the http codec in the trunk but I failed to reproduce your problem +tested +And what do you think about the following messageNo projects have yet been analyzed +The folowing tests dont do a proper resource cleaning clientsrctestjavaorgapacheqpidtestunitbasiccloseclientsrctestjavaorgapacheqpidtestunitbasicclientsrctestjavaorgapacheqpidtestunitbasicclientsrctestjavaorgapacheqpidtestunitbasicclientsrctestjavaorgapacheqpidtestunitclientclientsrctestjavaorgapacheqpidtestunitclientconnectionclientsrctestjavaorgapacheqpidtestunitclientmessageclientsrctestjavaorgapacheqpidtestunitcloseclientsrctestjavaorgapacheqpidtestunitmessageclientsrctestjavaorgapacheqpidtestunittopicWould be good if somebody can double check whether this list is exhaustive +Thanks +Basically i currently work on and is not yet started +IIUC Web Connector configuration is not a very frequent task from my understandingFor the two reasons listed above I highly doubt if its worthy to add many wrap connector for connectors loaded from just to meet a function that is not very useful anymore +Class can only be configured with an not a +whoops +Last checked is with rev. +Committed revision +for the patchBTW we also need to add Timeout annotation to test cases in as has done for other snapshot tests +Suppose you have two projects depends on where has Java builder and have no builders at all +Hi Kathey! Yes I believe this is a regression +Adrian This is good to know that we can easily extend menu widgets to deliver desktop type menu +This patch should fix thethanksdims +It has been so long that I dont remember the change but I believe we play with estimates to makesure that in this case the optmizer picks the index case +Ive deployed a new that you can use. +Thank you very much for the quick fix +Marking closed. +Well if you look at that stack trace I pasted youll see the lines repeat themselves +See the discussion in +In this case the JVM uses a random port that is unfortunately different at each server start +yes we should move to miniMR + +i package dest +I may try other values to see if I can find the upper limit but if it is near an hour starts to seem like timezone jumps +Patches welcome +Old JIRA am on vacation until Dec ndI will respond to your email as soon as I get back. + the v patch into and trunk branch +unit Vladimir the bug was fixed in CRYPTO module at rPlease check that the fix fully resolves your by Vladimir. +So what we can do is that rpc specified in server can be a max +I have not yet plugged this into the phase but comments are welcome in order to catch any early omissions +Actually I suppose session scoped components components would be affected too. +I guess the question is whether JDBC exposes it +Tested on on a testcase for this to trunk and +My appologies i created the zip but forgot to include the files with files within +Resolved as duplicate of because we get the same error on Hudson with. +We could do more harm that good here by evicting wo the applications consent +It assumes not only that your subjective decisions will be more intuitive than documented objective criteria but that all your successors decisions will be as wellAt any rate I think we understand each other so its probably better to agree to disagree than to continue this discussionI just dont see it but I wont get in your way if this is what you want. +Hi ChristianHave you had a chance to look at the patch? Let me know if anything needs to be changedTake care you Daniel for the patch +Try it +You find more links and the parsing is the maven core doc I obtain minutes seconds without your patch minutes seconds with your patchFurthermore it fixes +the patch looks good to me +I was looking at the read side apparently we write a copy of broadcast address as the value but it is never read +AlexeiI have fixed some synchronization differences with RI in r so Im marking this issue as resolved +Other committers please reopen if you think this should be applied +It should be ready within a week or so +There is a Java classs in it that is stored in the session which is responsible for tracking the state of the list +Committed to trunk and. +Questions should be asked on the mailing list as you just did +Second is that in order for Nexus to generate proper metadata Id have to pull the old releases into the repo from +The Eclipse project used to build both test cases +Im still chugging on the consolelogging should probably figure out if we can do this sort of thing across all the platforms +The approach seems reasonable +We could still do this at the CFRR level +Alex? Can you try with? +I dont appear to have the ability to move the issue to another project +reopen the issue to attach the migration doc added to and trunk. +Sample WAR demonstrating the issue +not a Core bug +closing stale resolved issues +Closing for now since theres no feedback on this ticket +I was doing a lot of of block sizes cache sizes etc +Resending to Hudson as it didnt seem to get built +The UIMA As aggregate should not call delegates marked as ignored +has been approved for incubation as far as I know creating the mailing list is the only remaining task before I can open it for public access + +I can have it within a week +Patch is applied at +Thanks! +I think in your case this bug might be related to a bug in the German keyboard mapping +Please let me know by then ifanyone needs more time to review the patches +It looks like the issue in is not in the rendering but in the way DN is formed therefore not exactly a duplicate but both can be resolved simultaneously +I will put documentation how to use this feature on the website shortlyBest RegardsChristian + +The main benefit Im seeing right now is it lets you compile and link against methods that arent really implemented in a useful way on NIX +Do all tests pass w version of this patch in place +a unit test for verification? Im trying to get the RC together and this is one of the only remaining open issues +bq +This is the patch fixing all the possible issues with the test I found + +Resolving this since Ted committed the patches +Please feel free to ask questions on the lists +Improved with rPlease check if more generification is needed. +If we want to validate the monitoring specific tests should do that. +Applied the patch into cxf branch and cxf branckTODO we need merge this patch into cxf trunk +I am using Mac OS X Lion +Pavol what is in the error logs +well I would like to try it first to see how it is to work with it +It was trivial change not sure why I was hesitant before + for the patch +This is the common behavior for rpm installed by packages +Thanks Diluka for pointing this outRegardsSenaka +galderz yes the implementation followed your suggestion +Interesting point it sounds like such a design might be achieved by moving the serialization stuff into its own layer above core Map ReduceEXACTLYIve created to discuss this point +Moving all issues to the CR +Committed to trunk and. +It would be helpful if you could write a Byteman based unit test which shows the problem +akazakovThanks Alexey your PR works like a charm +Perhaps we could get it uploaded to the central Maven repository for easier integration with the Tika build +It appeared to be the result of the current request thread local not working properly +bq +trunk Need to update wiki pages when they are working wiki pageNotice that in onwards the wire tap will by default use a copy of the original tapped Exchange +However if these methods are merged into the Filters class and we drop the addFilter method in favor of the existing and and or methods then we can sidestep the issue because none of the other methods need a factoryOther responses coming +Thank you for reviewing the v patch and giving it a commit Rick +That fixed the problem with Eclipse but I was able to get our technical staff to grab Helios and the archive file e version +If not should probably go in a release or LATER +My vote goes for or perhaps if it must be +Many thanks. +Sijie! Committed revision . +I should read email first +I can accept it as a request to add a command that will describe things in a nicer format +yeah this is an invalid bug +I havent taken a look at the other command scripts but this same change might need to be made in them as well to allow the same functionality in the scripts if we wanted that +Made the changes leaving a deprecated stub as suggested by Adam +The previous patch missed a copule of includes that gave error in Ununtu +It will either be resolved but not as originally stated or go back to WONTFIX +Okay once we have finalized the spec we can go back and revisit the implementation of scoping +It sounds great + +Just notice the field of targets in writereplace is required in +Is this the expected behavior +I can just file a new issue if preferrable +I was going to build from trunk today and test this again +I wouldnt like to spend time on documenting this legacy method since Ive already moved forward in my devel branch and moved all tree manipulation methods in separate objects +Added a way to run the system unit test more easily and fixed a couple of bugs +Dimitris what do you think +Seems to have gone away magically by itself +New name doesnt seem to follow convention based on other names in that file +Bug has been marked as a duplicate of this bug +Also this updated archive doesnt contain Subversion metadata +Run the main target in downloadswcresults to generate the reports +As the case with math functions we should only through +And not just any number but precisely the number Y prefetchSizeIf I restart my entire application the messages then get to the fresh consumers and flushed out of my +Please have a look at and +So adding the name to the section added by each component would help identify where it came from and you wouldnt have to rely on order and meta knowledge of the order defined in +The javadoc tool appears to have generated warning messagesNo javadoc warning is introduced in this patch +Hey VinodIn this case Ive killed ALL but the are still running +RN marked Needs More Info as little information is contained in the issue at the moment +Olegs descriptions are especially good +Changed Fix version to +What is the status of this bug? Has it been acknowledged? I can see it has been assigned +We fixed this and it works like a charm +I dont think the tests are in the correct location as they will only be executed for the embedded driver with the current patch +I just committed this. +Removed old and replaced with new one completely +heres a patch +committed to and trunk +Ok please update this task when you take the decision because itll influence the way I do it. +Worked for me on the command line then used mvn eclipseeclipse to set up the project +Avros POM does not require Hadoop +Packaging and Deployment are now using the Job API from Eclipse +pushed to +How about a refresh setting defaulting to s but which can be changed with a dropdownAlso is it possible to dynamically adapt the size of the graph? When you have a small node cluster it is displayed on a fraction of the available space and for my conference audience they could not read the text on the nodes +This might mean a few trips back and forth on machines with some locales missing but in most cases it would avoid the overhead of going to the OS at all +Thanks +Ill get a version of Grails as soon as possible and check if I can reproduce this behaviour +Unsetting Fix Version for unassigned issues +Complete patch for jenkins +doesnt have is does have is does have is but not means it must configured to use itIf we need work with a hadoop I think we need find another way to identify it +Cancelling as this is going to not compile now since Dhrubha has already checked in my txn of explain plan +Thats fine +Perhaps giving us the path the the git repo would help too +Its semantics are a little odd it assumes that the src tree matches the destination rather than the usual cp semantics but it overwrites the destination file iff its size is different than the source file +This is a result of the architecture of Cassandra which fundamentally requires a thread for an active thrift requestIsnt the whole point of HSHA to fix this +Thank youYevgeniy VostokovSoftware DevelopmentWolters KluwerToll Free x Direct Cell +All other JSF input components do behave the right way at the moment +Ill never pretend to understand Eclipses Update Manager but Ive never been able to update from a milestone release to a major release through the Update ve installed the Subversive updates and the meclipse plugin and it seems to like them +I cannot see the use of any algorithm that kicks the facelets support away though the developer hasnt configured the system to do so +bin and conf directories are moved to project root +Same as previous patch with usage addedThe majority of the usage in Outrigger only used the keyset +Still getting same message for embedded Tomcat on embedded Tomcat and get the same error message as for Tomcat +Im a bit concerned that myproposed change turned out to be more complicated than was necessary and perhapsyou will see a simpler way to fix the problem if you have a look at that issue +Patch integrated to trunk and +Attaching SQL authorization examples +Which isnt the case of spidermonkey +All of the RHQ based classes include test coverage +Latest phoenix trunk also correnctly handle classloaders for contained applications +Attaching a combined YARN and patch for Jenkins +Update has been committed to trunk +Password is now replaced with before logging. +Thiru thanks for catching that +Feel free to reopen if required. +I have moved it up to but lets discuss this over +Subsequently the management agent can contact the new server again and get the latest version of the software configured for itIn cases where the current version somehow has stopped working completely this might also be something you want to do even if youre not switching servers or anything +This will enable us to implement serial communication transport type in the future but not right now +HiCan you confirm that the fix in solves the problem for you +PhilThanks! The patch looks fairly clean but can you perhaps explain in more detail exactly what you need to do to the messages that this patch allows you to do? The Oracle link is not working for me +Latest patch looks good to meI ran twice with the patch and couldnt reproduce the test failure +If additional tests with reference results could be added this would be even better +Ill give it a try tomorrow +That means it uses a new timestamp for each map taskAfter patch we are setting ts to which is same in all map tasksHope I understood your question correctly +That was my suspicion that this is a regression due to the changes We had to make a change to one of the compilation phases and it looks like it has this side effect +I was trying to adjust size and position of dialogs and it wasnt working correctly when monitors are present +You probablyneed to make adjustments to the pom file +Note examples and docs need to also be updated. +I backported the change to of to with revision +It seems like it could be confusing if someone configured the ttl manually on the trace table then restarted the trace server and found that the ttl was resetchanged +Derby is behaving as expected when JDBC url has no territory attribute but it has collation attibute with value as terriotry based +And in it there are also comparison code of and does not perform better in writing and compression ratio but much better in readingThe test results in previous post is generated by class +The next version planned for July will include this fix +I noted the failure yesterday but I didnt ignore the test because the failure doesnt happen always +That was my understanding Florian +Maybe this issue can be closed? It looked like a duplicate of a bug that has been solved +testcase for me on a fresh buildThanks for the fix! +superceded by local repository proposal +This issue also affects which cannot be exposed as an array +It will end up to be a minor but breaking change which we would have to document. +If it is forcing too many cache flushes adjusting this parameter may solve the problemThe config parameters name is the default interval is seconds +on second thought you could probably turn this around and have the server learn the polling interval just by assuming it is the average of the last couple of polls +Ill wait to hear from Kristian but if hes OK I can make the change you suggest and commit +We arent going to maintain legacy code as part of this project. +Assiged normal urgency +The latest patch is tested fully and painfully in trunk +This was fragile but if this does all that can then that is fine with me. +In theory your query should not have worked before version +Jason Please update this bug +Should we also limit the number of filters allowed in an analyzer? Should we watch how many docs you are putting into the index? Should we make sure you dont add a query filter thats too long and crazy? Its a silly direction to go down +Yeah definitely had the wrong rebase +Also there is not a specific any messaging system creating a consumer and session will involve a network round trip +HiAs per subject +After running a time or two we suspend the and wait as we observe the timer continues to fire and the route executes despite being suspended +Attaching the patch. +I would suggest a fix day for me adding a new patch which gives you a week to commit it +Is it going to be fixed soon? If anyone have a workaround other than downgrading Hibernate please do share +I like the new tests in. +fbricon they are all working hard on having archetypes aligned by Monday ready for your return to update the archetype list +The behaviour changed in adccdcfaebfb February +While I admit to still being a novice at building Qpid I have to wonder if bit is even coming into play with the errors for and Qpid proton not found required for amqp support errors +I will try to think of a case where old dbs may still have the old format for the empty string and thus would not be fixed by your writeExternal change +Next Ill be doing profiling to figure out what are the hotspots and see how we can improve things on the JBMAR note that when using Marshaller we havent yet added VAM information to it so home grown marshaller VAM is faster than Marshaller ve also created a test that compares size of payload with both marshallers and the result so far is that they payload is smaller with the home grown option compared to Marshaller +Looks to me like this was fixed by the patch for can you verify that it works for you in the branch +Approved for. +Look at the digits in accuracy per evaluation for LGQ vs improvement in accuracy as suggested by Konstantin +Reopening to set release note fields +Changes seem reasonable to me resolving. +can you please try with the latest SNAPSHOT? The behaviour is not yet final as we need to specify the correct behaviour in the EG first +I am all right +I just committed this +What was the reason for removing initializaCreation and finalizeCreation methods of theChamikara +When resolving type parameters in the editor and we come across a type parameter that references itself we simply convert the to an Object reference +Ive created some Jiras on the ESB team to document some of these options but thats going to be some time before the documentation is updated and still if the options are not in the tooling I think well have users stumbling +Good stuff this effectively covers the scope of this jira +Previously we only ran it at debug level from time to time which can let nasty bugs like the present one slip came across the same bug in RC +bulk defer to to +Ryans graphs and tooling looked really good so I was hopping he could quickly catch something I missed +We often run into this problem when the local named caching daemon fires up first +Check the class and look at the end how I add the items to a user +Screen shot showing schema browser shows the problem +Ill see if I can add a sed script which filters out the differences +If the runs some warnings get lost +However Im currently working on some otherissues +I just committed this patch +Check out the set on and the great to me! Im glad we were able to work out a solution +At simple test case would help a lot +I dont really get the point of this would you like us to change precisely +Please provide a test case that illustrates the problem +I guess +the onMessage method will be the entry point +I added a file missed by the previous commitCommitted revision . +I also do not see anything obvious in the patch that would cause new problems in the tests +Is there any hope that it will be fixed soonBest not before late this year unless someone from the community steps up to address it +Yeah I noticed today +Merge and appropriate usage of has was required to get around this although the design of this area needs to ve revisited. +Hi SianThanks for your kind help +Robert We need this code because the filter language only recognizes some base set of filters? This code extends the set of filters known to the filter language +The on that line should be acceptable so it would suggest that there is a in one of the attribute values and that the values should therefore be escaped +doesnt have some mutually exclusive options +Could you please update me on the sameVivek JaiswalYou can easily judge the character of others by how they treat those whocan do nothing for them or to them +Added page to menu Changed path to admingraphslocks Added to parent +The main reason why it was converted is in prevision of JSF which is based on Java +For example here in Jira when we lookto the left hand side +has been marked as a duplicate of this bug +Are you able to manually access the DB using the cloud user +I will take care of porting it back to branch +Sijie going by your analysis I think fixes this one and +superseeds +With speculative execution true +Also I found a test scaffolding to be tough to use so instead of testing Ive added interceptors part onlyThanksAndriy +Fixed in CVS. +All projects are hosted on labs +created for the part of this. +Ill add the additional test case in a patch +The references to the package is problematic for similar reasons +If you feel it is the same as please watch or comment on that bug. +That functionality is that makes no attempt inany way to apply a default configuration to logj +convention preferred by QE devs +Mockito comes with a NOTICE file BTW but since its MIT licensed it has no effect on usThis whole episode is another excellent reason to keep your set of dependencies small +Also we probably should provide a option to allow user giving a raw URIWe only support a subset of so we need to have our own constructors to validate and normalize +I almost had forgotten it existed +We can consider it if you propose a compelling use case +info that can feed into the is about not waiting for a reply to a send so that there can be pending sends but it still blocks to do the send which may be dependent on the failover is a option that you can use that will ensue that a send will not block for ever you get an exception if the time expires +I was just making sure we were ok with the against +Rejecting for now +I am not sure this is the correct solution +also since we are working on an upgrade feature on zookeeper we can upgrade the version to be long sometime later +the test is done with a single server wherein a client authenticates with the auths the first one being the required one and after a restart of the sevrer the client is able to authenticate itself +We need a generic framework where data from Databases could be iterated upon as a vector and algorithms can use it seamlessly +What do I need to submit to help move this alongThanksilya +I seriously think that entry is the better +Thanks for pointer N +if there are some localized files under the task cache directory that need to be loaded at launch time well need permissions for these also +Ive found some time to look into this issue and it appears that in the delightful method the chunk of code that handles external wiki linking has been moved after the chunk of code that handles the result is that the wiki formatter creates external wiki links which the URL reformatting code then merrily processes again! For the time being I have moved the URL handling bit to immediately before the wiki linking bit and this fixes the problem for me +Patch has been applied to branch in SVN rev +Null pointer check +so in short jbosseap doesnt change just something that allows us to have seamgen in a working version to use for RHDS in eclipse format for exadel Fix Version since is not part of verification on EAP platform required closing. +suresh +Can others please comment on this +As a workaround Solr checks are currently not so thourough because the full classpath is not available on +I added more tests too +Im able to have it fail in a bit other way +Verified in CR but as the occurence is rare will reopen if the problem will reappear +strange Ill take a revision Thanks Drew! +Sorry +After goes in Im going to try and get to some of the metric that are needed +Anyway thanks for the patch and the effort to improve the codebase. +Please RajYour patch is in trunk rev +I have merged this to +Problem is that I do NOT get any exceptions but after some debuggingand reading it turns out that mySQL does NOT store any fractions of a second inits date fields +But we need to mark this JIRA as an incompatible change or put in a deprecated config mapping +Just what needs to be done before we can tag this for incorporation into the branch so that the tests can be incorporated and testing can start +With this patch these tests always pass on my box +The occurring is a risk of the in the design +I wasnt able to get the existing ibatis trunk tests running in my eclipse so could a DEV take the two testcases from the and put them in the trunk? Or is one of the already using eclipse and could commit his + test case is flaky one unrelated to this patchcommitted patch to trunk and +Thanks Jakob for the review +One of the new files was not added to the patch +Radim Nutch is based on the Apache distribution of Hadoop and already works with it +I guess Good point thanks Will do Will do +patch for at rev +Cant simply change attributes to auto because then bug with iframe appears in IE +Updated patch attached at +Im attaching the patch with the fixFix was tested for invocation of both proxy service and sequences with Default EndpointsThanks +yes +So it looks like youre seeing similar results that for larger values native is faster +Reading snapshots support finding the table descriptor in the subdir or the orig directory so no migration of snapshots are needed +Still reviewing this one +Ive taken approach for now and will loop back when I upgrade +Please note that any Interceptor instance is per spec definition a Dependent scoped instance and thus gets passivated along with the passivatable bean +This bug appears to be resolved in version +Can you please explain +No problem +Waiting for the user to test an updated copy of the connector before I post the patch +However this approach is not very correct because the server could have restarted with a different version of the software +Marking for move is attached as requestedHaving without the deps will be great +Should this JIRA be moved to as it seems most useful there? +batch transition resolvedfixed bugs to closedfixed +out from under hadoop wiki also move to confluence rather than the legacy wiki technology +We need to find some solution for this in +batch transition to closed remaining resolved bugs +Can you give an example of how you imagine using the FEATxxxxx +How did you build locally? What profiles update sites did you useCan you reproduce if you move your +Done that for all existing skins +Im thinking more along the lines of embedding the functionality inside the CGLIB enhanced method +Tharindu May be I am wrong here but to my little knowledge on hessian we cannot reuse the MTOM code here but just the approach +Grant I was trying to verify the patch but for some reason my hadoop setup refuses to run with Error reading task output message +I updated the bash script to show how simple topology scripts could be +I think we should have a way of defining a common list of scripts that need to be executed for all services before and after bootstrap install. +This included all but publishing the DTDs on the web side +hoping someone codes it before I have to +fix single line are seeing faceting on fields as a significant performance problem so wed very much like to see something of this sort +Fixed in +OK I didnt find the jar inside the file but I did find the source under javam having trouble explaining why xerces dependency is failing after it hits this relocation +As discussed this seems to be a core jBPM problem an not a GPD issue +Jason please note that the latest katta code is actually in sourceforges git repo not in svn +Sorry but I dont see a difference +This issue allows client connections to proceed even if permission is not granted for the new system properties. +I have seen it happen on a page which will only ever display up to use display tag version along with OCJ problem is hard to reproduce but it is affecting our live systems +Patched quickly tried the suggested changes and it seems to work at least for my case +This indexing session did not produce so many inconsistencies as the previous one there is only shard of which one replica has more documents +first upload waiting for opinion hints instructions +so you didnt do anything special in RHDS like enabling JSF on it or something +Declare a decorator and thats it +What do you mean by serialized to a Java object? Do you mean to a ? What info is this referring to the Java object will contain info that was not serialized? Can you provide an example and show exactly whats wrong? I would prefer a solution where we avoid producing the info that is causing the problem without temporarily detaching the object +in Maven there is only one POM for a given coordinate classifiers are for derivative artifacts and so the project information should apply to all +One thing that was noticed was that the getCounters call in is synchronized +Please see the in the toplevel I was expecting entities and a test case running HEM not a full projectI dont have time right now to dig into this project I did add a unit test involving a bidir one to one and a merge operation and it works fine for me +Thank you Tanping +I misunderstood some things +Jakub is this issue something you would like to dive into and get it implemented +This patch also contains general simple methods to create Person Role and in partymgr +text file go with the text file +This issue does not only affect commandLinks but also commandButtons which are in forms inside a panel +For future reference When changing the sources use mvn clean install Pjavacc in and commit the updated files +Still weird that only a few people seem to have trouble with it +I updated RSPEC to transform this rule in Useless imports should be removed with different meaningful messages + +The account is though removable from the API +Patch and unit test look good +It sounds like you have a reasonable solution here is there any reason not to implement it forJust that its quite a bit more complex than simply disallowing overlapping ring movements and the extra problems that come with higher complexity +Try running the test with Jenkins +Thomas the issue mentioned above is not fixed in Haritha Reddy can you please provide a patch for your changeIs there a way to get this fast tracked into SVN? We are also finding that this is causing headaches with our clients +As Alejandro also pointed out and as your environment suggests having a running job limit you were running into to close with proper resolution. +In factthis is the proverbial feature not a bug in that this approach consolidates all JS serving in a single file rather than requiring several HTTP requests one per feature bundleMy understanding was that you found typeurl gadget functionality to simply not be working +So without the viewBox attribute you are changing the effective width and height of the SVG in userspace butyour document uses the same portion of the userspace. +Thank you for consolidating the Login Module setions +Odd +Can you please attach an eclipse project that shows this error +It is because MR jobs are submitted through and managed by while checking AM size against the maximum resource capability is only implemented in Client and of the distributed shellIn addition since the requested AM size will be rounded up it is at the risk of the roundup value goes beyondAnyway the aforementioned situation is a separate issue which Ive filed in +How is this different from the implementation? it looks identical unless Im missing something +Didnt seem worth +The entire contents of the file referenced with xiinclude should be added no more no lessThe header described above is not in the included file but seems to be generated automatically as the today entity updates every time it is builtThe inclusion works correctly in version so the issue must have been introduced inFor the example described in the steps to reproduce the docbook fileComponentReferencesrcmaindocbookuses xiinclude to add in the xml example written inComponentReferencesrcmaindocbookextras I seeThe issue might be that jdocbook is not honoring the parsetext attribute for xiincludes +Heres an example of what is exported now true false +the patch is for like a worthwhile and simple enhancement. +In general adding an to a type used as a parameter in a wrapped message seems to have well spectacularly bad results for the resulting schemaThis is actually the same issue as and was fixed as part of that +The issue is discussed in detail in the thread but if we stick to the direction it was taking wed just close this bug as wont fixMy +You are right cant change this in at this point since we have had branch releases on +The will also perform this task when a node goes down and any subsequent request for a connection will then only use the available nodes +fixed in revision Mark please verify it fully resolves the good +Ported to branch at subversion revision . +This doesnt need to be blocker for code is refactored by +I validated this using Carlins example and verified against SVN +Wow I must have missed when the queryName property was added to that reader because it definitely wasnt in the original version +BTW it has been discussed a few times theres no need to add numerical modifier to a patch file JIRA takes care about this automatically +Im tempted to spend some time looking for a more way to code the merge operation not because I disagree that this is a valid Xalan issue but because I really think folks need to learn how to write stylesheets that dont rely so heavily on variables +is there any need to target this for +We should check if this still does give problems +I guess we need to merge this improvement there as wellThanks +Doco written with update needed on Apache site. +Update some files +Yep this is no longer necessary + marked upstream issue as invalid +Thanks Bryan +There is another issue to track the issue where IPV External DNS entries and IPV Router details are missing. +Integrated recent commitsAlso waiting for +java and Groovy have lexical scopes so you cannot define two foo variables +Is this a bad idea +doesnt work anymore +It looks as a complete failure of loading Seam Components View +If the only dependency is on a standard java library class then is a good package for the renamed classI think that the whole business of and needs to be looked at +on to snapshots branchThanks for the patch AleksThanks for the reviews Jesse and Matteo +I just thought I would fix this while it was in front of me +First cut of the sync code +Pull request sent +It will not work for me +the elementscomplex types referenced in the binding file do not exxist in the schema +I am not sure if the configuration fits in the DSL as its only applicable in the scenario when you use the which you would need to configure as well +added info abaut class +I was afraid that may be the case thats why I made the other attached a new self executing stack overflow example +removed the tags +Quick update starting fresh i was able to get the issue resolved once ant properly rebuilt the file +Resolving as duplicate + +committed by that there is no way to set value to null +More comments and explanations can be found inside the code +Removed Transactions Web Services Programmers Guide from the and repos +The patch looks mostly good to me but it misses a test case +Verified. +The scripts should be OK now +Great that finally theres a bug for this issue and Im sure many users have to face this problemEven worse is the case when a Dispatch Action is used after form submission and the file is too large Even the do parameter that determines which method to use in the gets lost! Very annoying problemIt would be great if the struts developers may recognize the fatality of this problem thanks +They would be nice enhancements in themselves +Is this going to be included in EAP? A patch was made available for +Hopefully all issues are now addressed +Unfortunately I didnt pay attention to the fact that text is indexed but not stored +Hi Johan could you give a try with java as it sounds like a bug in the JVM +But if the namenode slowly the admin can always kill restart it manually so I dont see the harm in it attempting to +Taking this off of as nobody responded that they were signed up to fix it +Looks good except for CVD still has old screenshots with Gladis and wrong number of menu items +Hive packages build and install fine on Ubuntu x +That is a fundamental constraint of mapreduce programming +It would be cool to test this but I cannot imagine how to do it if not in the itself +Jorg this is Cocoon trunk I take it +This is a request of Streamable interface within this project would have the same effect as should be a Marshallerss feature The advantage of such Streamable interface would be to avoid using any serialization at al and do a direct byte buffer. +Marking this as fixed because the patch was committed on September +Just make sure all the tests continue to pass +It might make a little sense when all the resources are bundles but if they are configurations or whatnot it will not make too much sense +that should run all tests including the storage drivers with the necessary dependencies in place + +It can be run manually +xml if it doesnt exist +Patch committed to trunk +Submitted patch for review thanks +This issue seems to have been caused by the ASs inability to process address literals in correctly which was covered in +for changing all properties t it seems reasonable that any thing other then true is false? Cant we add some notes about it in Practice Application or any better place or even in the property file as my vote would be to keep it intact +UmaThis is what I was suggesting +Submitted for Benson +Thanks in advance +It depends on core and the reflection layer used in hibernate annotationsNow the Entity dependency comes from the reflection layer but this probably can be easily solvedWhether or not having one or several jars is a deployment issue that can be solved only after the previous ones are +Thanks to Brandon! +Removed and my initials +There are several ways to work around this particular issue and from what Ive seen it occurs only in a very specific case you are excluding a dependency from being pulled in transitively but then need the same dependency in the test scope +Hi can you make a small example of the rules that you used? Im going to take a look at this +Thanks to both! +The Process BuySell +Reliance on hadoop script to launch MR jobs is not cool +When you diff the with this two versions you can see the let me know how I can help +Reading directly from directory is a permissionsWe should investigate standard apis +Will try to submit patch that respects WFS option feature bounding for JSON in order to not calculate the bbox member for response +callId service methodName Mutate size connection param region t +It is possible that another error raises +Seems to be OK will that patch be applied to the trunk +Not sure what other info I can provide without specifying server and port +Sub Tasks seems I need INFRA permissions to change some Jira settings +I think somebody already added it to Solr too dont know the issue but have seen it. +Wssj depends on which is supposed to fix this issue +If you define different dependencies for a plugin in a the first definition will found win +Again when is it not creating the constraints? Schema create? Schema update +Ill look into adding your repro as a test case in something likederbynet +Pinging this issue still liveHasnt this been subsumed by other work + components have been tagged and are ready for pulling into for EAP release +How can i turn off extension filter? if i dont configure that listener i receive an error not correctly configured +Got a day off for celebrating Independence Day +Hi GeorgeYes I agree all class files can be deleted +Fix from Daniel committed in r +This is exactly the same as with the Tuscany module which also uses provided scope +Great! Thanks for the quick update Colin +Base version of the Externalize All Strings dialog was implementedPage User selects the bundle file +I can see the changes verified. +Very effective thanks! +We should also have a look at handling any other errorCodes that are returned from so we can get a better exception in the logs +Please test and close bugUpayavira +What do you think of this patch Raphael +Wontfixing for now although still open to a pluggable solution as above +We could fix that by picking a random node to start our scan of potential hint destinations theres nothing magical about starting w endpoint +Whats causing the native exception? I cant reproduce the problem you are describing with the information at hand +I guess hiding is the right word to use The launchPig method in localLauncher or map will hide the launchers static method +Reopening due to test failures +Don your suggestion still looks like magical convention to me +The wasnt being given a parent classloader from which to load class data +I dont believe many users will even notice it not being there since it is for compiling you agree? We would need to document in the release notes that users of the jar would need to add it manually if they need to +The dollar slashy in particular is good for the embedded regex cases you mention +Fixed in r. +Yes the test is unrelated +deployers +Finally done! It could use some better documentation though. + of Aug displayFormat ddMMyyyypick a date okerase the inputpick a date all dates are me know if theres anything I can do to help move along the process of the patch getting committed +The version comes Team to Burr to figure out the version after speaking from Team Errai +However Im not sure whether the additional complexity out weights the initial goal +Those were never providedplease dont open bugs if you cant provide information that lets me understand what the bug isasking for help in the bug list looks odd +Marking closed. +bq +This was fixed some time ago in +Can the issue be closed then? +This seems dangerous to me as Injecting a into an or bean would mean that the injected would quickly become stale +It will create a directoryCd to the directory and run mvn packageIt should create a target directory with a start of a directory structureLatest changes for distribution +Sounds good to me +Well Ive investigated the problem and it is obviously generated by the Karaf branding pluginIt has its own copy of the +Thanks Tom! +Id resolve this if I had the right perms but the new version of surefire is compatible up to version of +I tried to be as little intrusive into s stuff as possible +Works great +I changed because I was under the impression that a stream request to an endpoint that doesnt contain any of the ranges requested would create a header with null for pendingFiles +Make sure webapps directory is bundled and class path reference is corrected +Heres a patch that fixes the problem +Please fix these issues because they really destroy clustering +Can you clarify what you mean? With my patch to tez local mode Ive had no issue running Pig on Tez +Not for commit +Bulk move issues to and +Im not sure if this could lead to any compatibility issues? Is the old mechanism to close the JCR sessions still needed? implements the interface and collects all JCR sessions after getting EVENTDESTROY +Patch added +The distinction between properties which are defined at an AMQP level properties which are common to the Qpid brokers and properties which are completely bespoke does not seem to be terribly interesting from a generic management perspective and only begins to become interesting if you build tooling which depends upon the existence of certain properties +I added links to the XML schema and the XML Schema Reference in r. +If auth no is specified then the broker does not send anything in the SASL mechanism array to the client +DarrenIve just looked at this and tend to think that its not worth it +Attach the patch files +BTW I am using ant on Mac OSX +The latest patch looks pretty good to me and I agree that the test failures seem unrelated +My suggestion isnt as simple as offering a choice between preserving IDF for all terms or not +the plugins classpath +If you can do a failure then revert it thatd be great +Just the toolbar icon needed to be adjusted to allow for the new Central iconThank you for the quick turnaround +Your call though +Is the intention to continue to use the same port but with a different protocol and if so how is protocol detection going to happen or is the intention to switch to a new port number? The design also does not explicitly state anything about sunsetting support for older adapters +How many records are you parsing whats your mappers heap size +However I can understand the point of view that something unexpected happened and we should bomb +Seriously nothing goes to nexus release before we got these things cleaned upWhat is to clean up? We cannot work on improving the way parent is published and more important consume by components if we dont publish it to Nexus +I suppose let me keep the official version of the patch for now since I need to make some fixes +The earlier patch I submitted can be ignored +The stop logic is not in line what we do in other componentsThe class loader comparison is special and we should avoid such things. +see above +Konstantin the new patch contains what I originally intended to do incorporates your change to +The latter is provided for ease of migration from Jackrabbit toThe Reference Guide was updated to document this new approach +Patch for will be attached after Hadoop QA +DanThis is great news +It copies the v to a v HTD +Removed listeners in Git ID bbcfeaddaeaeba +Add more descriptions in the config file Set the memory usage of hashtable sink op and group by op into their desc +The only fault I really had with it was that all the configurations have a and with parent beans but it never really explains whats in the parent beans I guess that could be a talking point but a brief explanation of the fields in the various configurations might not be it looks like the getting started ppt has a few extra materials required for it such as the archtype and some kind of +try this patch +Applied patch with thanks to Scott I also fix some CS issues and clean up the codes +If the building system would have installed all versions of supported distribution it shouldnt cause problems +TCK were fixed a while back +can you comment on this please can you also comment on the case where have say a mix of jboss and keys in the same appserver ie a system gets patched with a different Coxs team who own the keys had been updated about the keys expiring few months back +Most of copies the change made in into +I would like people to review this patch to find any errors I didnt + +In this patch the outputschema and the mapping table have been added into STRSPLIT functionAn ee test case has been added into theThe has been successfully run with this changesThis patch should work on both two versions +Lowering the priority because less than of Android devices are running so far and has this fixed +Committed to master. +Let me fix that +This is related to have double checked this with the current code and we are using the correct name as defined by juddi +Ah ok +The javadoc tool appears to have generated warning messagesThis is not this patch +Please disregard was not able to include the new classes +There are numerous cases of this and we are fixing them gradually as we come across them +all resolved issues due to a successful release +Committed to trunk and x +Olivier can you give a way to have both main and test files in the same drilldown page? discussed with Olivier we wont fix this issue. +And based on my comment I think this one is a wont fix +These issues was set to GA which does not make sense since CR is just supposed to be renamed to Final +It works OK now +Test those? This is possible in ecommerce application +The core of this from the patch has been tried and tested in production +also fixed a problem that should have said tests should end in Test not start +Patch looks good ChrisI ran the test with and without fix +They can be used to find the filtering error if there are special characters like in a column +Touches the following filesD toolsjdbcD toolsjdbcD toolsjdbcD toolsjdbcD toolsjdbcD toolsjdbcCommitted to trunk at subversion revision +Where should I upload itAnyway let me know how I can help +Im neutral on the warning v +Now that work is well underway with AS all previous community releases are +But that is just an idea for the future +It inherits from and doesnt use anyPerhaps in future we will come back to this point +I created for storing last flushed sequence id for each store of region in zookeeper +Thanks Brock for review +Just fyi what they do is not magic we can do the same thing with our own style sheet and a little javascriptOne disadvantage to using Dojo the visual elements are not scalable +I think separating out all properties into two or more files will just confuse the user at this stage +I made a test on intersection +fixed in bfd +No now it works +Comprehensive upgrade patch from candidate to candidateAffects the following filesIndex srcAnalysisStandardIndex srcAnalysisStandardIndex srcAnalysisStandardIndex srcIndexIndex srcIndexIndex srcIndexIndex srcIndexIndex srcIndexIndex srcIndexIndex srcStoreIndex srcStoreIndex srcTestAnalysisIndex srcTestIndexIndex srcTestIndexIndex srcTestIndexIndex srcTestUtil +New structure +This issue is not Mac specific +supported vs certified verified on PM list description mentions the verified Spring version as an alternative Snowdrop guides now include the and its description in the package structure sectionResolving this! Thanks for your hard work Ondrej +For instance if a region is offline you wont know by reading the +The patch looks good to me thanks Bill +New patch with unit test +Please have a look and let me know if you think it would be useful +It works fine if I change it to create the after the web applications deploy with the EAR +Ill leave this open as we do implement just one type in concurrent and I guess we could internationalize that one type but it seems pointless to do just one +Checked the code and added an example in the userBasic directory so closing this. +PLEASE do not create any issue before having first discussed this issue on the user mailing listCould you please attach to this ticket the java file comcis to help us investigate this issue +as long as we both agree the implementation should change +Im currently testing it +will file a follow up jira +bq +It should not affect complex features +done +Had if verify email address functionality be implemented setting partyId NA then i would call it as a bug +SaikatYour outline is just fine +What else is required to get the patch committed? +Focus on file behaviourWhat can be done with configurations that are not exposed as eclipse classpath containers? Can we get them visualized also +Please explain +There is no more problem with this test in EAP CR bulid of + +fix will be in the drop +Something with sane design preferably as opposed to the design ofBut hey obviously people dont want or any promises of stability +Therefore in the case of webservices support not being able to set the on http responses effectively means we cant guarantee interoperability since many clients will refuse to process responses that dont have indicate an appropriate +Id like to keep this issue open for a while Im going to test how it works with the new infra +So Im surprised you dont see any real difference +Please close this issue +Bulk closing stale resolved issues +Hey RobertA revert is historically considered a big deal +Custom event can be handled by your own protocol below FLUSH +I dont want to pick up a dependency to the apache package just for this and the JDK regex stuff is only +Reviewed patch Nit Calling an abstract class seems a bit odd even if it is technically correct for this +Bjorn your results sound a bit odd compared to my experience +attaching possible fix +New patches added +See for more info +I think in should use urnswitchyard cause its not part of the SCA schema. +It runs in second on my computer so the bottleneck due to the synchronized is limited +Why do you use the same action with Portlet result in environment? The NPE is thrown because of missing should I apply this changeMy project uses Maven to get the Struts portlet plugin versionIs there a Struts snapshot repository I can use to get a version of the portlet plugin with this changeOr do I need to checkout the source code for the portlet plugin make the change and build the jar? If thats the case can you give me the URL where I can checkout the code +Sure no problem at all +It seems no issue about this +hibernate bundle not yet uploaded +This is the correct addendum +Michael which OS are you seeing the exception on? Im trying to repro on Linux w IBMs JVM with no success +clarification I am not saying support compatibility as I am saying provide a migration pathIt does not make user comfortable to know they investment will cost them considerable to upgrade +Stack said to go ahead and commit this +Eventually you can reproduce this problem. +done for reports these projects are unchanged so must reopen to mark REJECTED instead of RESOLVED. +Sorry its a core thing +But Im not against copying the dtds to the src directory so Ive put the copy of the dtds back with revision +Are u sure this change is enough? i see add is private as well +I just need to finish something maybe down to the way the design of getting a in CXF and what we want to do with it is out of. +fixed through resolution of +If I got it correctly this Web Module facet install dialog is available only within Seam project create wizard +Im working on thisIve already solved the currently working on the other didnt see your comment so also started to work on this so I also made patches for the issues +v has been been voted in the last few days by the committers on the dev mailing list +Only affects as this issue was fixed by upgrade of Cocoon in Wont fix for +Thanks Charles +Patch applied +The problem scenario still exists if I use the current latest bibernate stack based on row in compatibility matrix for. +If folks think it should be put in happy to do so +to keep shutdown from waiting for the compaction kickoff +I will not be in position to commit myself to SVN for the next few days +Thanks Anton for the improvement patch was applied to TOOLS module at r +looking closer it is a session expiration issue +Once that happens Ill commit this +I am testing the newer MS version with the object store in the path problem we saw may have different behavior +Increased is how large an increase in absolute terms +Fixed in a previous fixed in iPOJO +from me +The patch seems to have a very large amount of changes for such a small thing +Bagsie not issue out as part of JIRA cleanup +Just fixing it. +Benot just tried out your github branch +This has been fixed forDear DupuisThanks you and Spring Source quickly ResolvingI have differed the source from trunk and fisheye and I have a small question as following +Since will contain the RI we will do the opposite logic +thanks cant wait for AMP constants refer to amp instead of and uses the AMP constant instead of ampadded additional test case in +Attaching a patch for this issue +Added sanity checks +The datanode first forwards the data to the next datanode in the pipeline before it writes the data to its local block file +any thoughts based on my last comment? Im trying to clear out issues for the release. +Uploading a patch to check if the fix is in the right direction +Yes I did and it looks good +Retries shouldnt do that code path +Works with Mojarra doesnt work with Mojarra +Tailer would read from start of file on frequent occasions +And I am afraid of overlook something so I use wizards method +Here YARNHOME is checked first +Should we also add this srsProfile param +Moved the creation of Web artifacts from to +Committed in revision . +For holding unified cardinal number for all Sinks +Then it is of no use for user +Minor comments I know we have not been following the java coding convention in terms of defining the static variables at the beginning of the class +I hope itll work for you but probably youll need to change it +Implementation code is attached +bq +v improves the explanation in and changes thresholds to kk +Also thanks to Arpit Singh Pandya for Researching and working with me on thisChirag Manocha +bulk defer of features from to +They look correct now. +The flow of HTTP communications is alwaysREQUEST RESPONSE theres no flow like thisRESPONSE REQUESTTo be more clear the Response object where you are adding cookies in theoretically came from a different request than the one you are trying to get that cookie back from +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +This patch uses create table as instead of insert statement to build the index table +If logging for webcontainer is set to debug this happens on EAP as well +Good to hear that this was fixedWere still on izPack but will check later if we can upgrade to to have the fix work for us +Create an International sales order +If you dont the tasks can be lostYou should also supply patches against trunk +Im not sure I understand the problem here +Looks good to me for . +Fixed in trunk for patch added for +which also would somehow coordinate with these to avoid duplicationsoverlap +Stephen I have just commited as is +Attached patch +It is one of those things that just bites folks over and over and over again. +Right there with you Im thinking in this direction too. +for the deadlock trace Saurabh +Patch is attached for specified functionality +Unfortunately I think this is a side effect of readline +I suspect that some threading behavior caused that effect since with entry into the filter the second check shouldnt be necessary at allJuergen +plenty of multimodule seam projects out there +Unit tests are also included +Wrong state the new found problems are resolved fixed and can be reviewed and pushed if you like to +Added +See for discussion +Correct behaviour is for send to be checked every time otherwise if write permission is revoked after producer is created then client can continue to send messages +Its a shame the API of the do not have a nd optional runnable parameter for rejected execution + +resolved based on the previous comment +Claus what are the platforms where build fail +I am trying to checking the other problems when I match the coming project milestone I will arrange the modifications maybe will add one or two more reports and do upload the patch +bq +IW would only need to know about attr source +I am trying to fetch the id field from the secondCore using the join between parentIdsecondCore and idfirstCoreI want to give it a try for the long support +to inform server listener about switching to another month +Ok but this is a new feature so probably will be included in the release while the fix for probably will go in +You already have a base rewriter that can handle many types of rewriting operations in the future so you should just put it there +If you have a starting point that would help me greatly +This was committed to branch and is still in RC so I have to include it in the new one Im rolling out. +Thanks a lot AndreasDo you have a date in mind for the? +Pull request closing. + resolved issues for already released versions +I am for the last patch +Thanks for the catch Doug +A zero value means a transient cookie no? Does httpcomponents not support transient cookies? If so we cannot use this library for session login since most sites use this +HelloWould you please try this patch? Thanks a lotBest regards I cannot apply this patch would you please look at this +applied thank you. +The patch did not correctly handle paths with spaces on Win +the patch +No no plan to drop there are just too many users still on it +But we could keep the jira around for such discussions. +Patch for the rakefile to set the test environment variable +MartinCan you verify if it at least completes on time? Even though the progress monitor is incorrect does the window close properly once the download is complete +Thats a good theoryI guess the worst case here is the uncontended case with lots of cache missesFor example a single threaded large scan over data not in the block cache in that case well always needlessly go around twice in that loop +Well just have to wait for an integration build for a fix +Updated the patch to get size right and account for new layout in trunk +I see nothing in the Tomcat log +I am consuming my portlet from to Jboss +xml file in my conf directory +Verified on EAP CR +Once you get your new patch ready it would be good to do some testing wtih and mixed versions to make sure we didnt make any wrong assumptions there +The fix doesnt need to no +Thanks! While I trying to write some codes for it I find the current issue is caused by another reason it seems that pluto will redirect the client to a new url while it finishes the action process and the redirect url is usually very long it will caused IE corrupted +fixed the problem was that the property was defined as system environment property and that was not taken into account in the UI. +This adds to toolsjava so that osgi support will always be built +Hey Joe sorry about all that delay got busy +However if the installer promotes these habits they tend to agree with David +Andy you tested the attached patch succesfully right? If so Ill commit that +I found that this bug does not exist in ecommerceI will atach is an old issue +needs to be updated when this issue is resolved +support is there leveraging a controls property +elementBefore and elementAfter methods have been added as a part ofNo positionOf type method as yet +There is another more immediate issue with the patch +The logout functionality doesnt work +Attaching patch for +CommitMahout uses java where there is an advantage to do so +Patch should not be applied until other two are applied +uploading zip containing test case mentioned in theThanks for the prompt responseIm glad to report that fixing the iterator does eliminate the NPE both in my example code as well as in my m especially glad to hear that custom collection types will not be specially processed going forward +So Im fine with if its that close anywayThanks +Patch provided but requires updating release +Enable the use of locators +Just because Perl and Ruby do it doesnt make it good and desirable Python doesnt have the structure as was originally stated +I am going to manually verify the issue but I wanted to kick jenkins in parallel +Need to review this but BB is now +The error was produced because of incorrect version of spock +Appears fixed as filtering Draft Entries by date now appear in Roller +Sent from my phone +Can anyone help review and commit this patch? Thanks +Ok let me try it again +I keep staring at it for a little more time to see if theres an obvious error and a fix +I have a solution to this problem Ill roll it out in the next day or soI pulled the new plugin this AM +havent seen this issue occur in the wild but its something weve reasoned is possible +Please delete after usage +Verified on. +NicklasThis should now be fixedI did try using your examples but both of them had issues due to missing properties files +FYI this is the only Type that does this every other type appears to function properly including Float and you include a unit test for the problem +Hi BryanSorry +Ill put this on hold and commit in a few days if nobody has anything against +Changing affects from to so version can be removed +It didnt make a lot ofsense when that approach worked to start with but it worked right away soI went with it +But this cannot be implementedAs per the Configuration Admin spec the actual PID of a factory configuration is unique and implementation dependent +To get completely seamless support for multiple encodings which is what we need well probably have to look toward Ruby string file another bug for the new issue. +As mentioned it would be nice not to include the but some meaningful value like expectedSize expected or whatnot that would work both for merges and flushes and also wont introduce needless dependency on +We dont use anymore the thus the syncMillis parameter is not used anymore. +I guess I missed the submodule rats on trunk +I just committed this. +Hi KarimIf you still listen would mind have a look at TIA +. +Thanks Jonathan. +It could be that the message should be printed somewhere else than it is today but I dont think it is good to remove the message as a whole +Im not convinced that its worth the trouble of caching the changesummaryobject given the problems it causes + +Great patch Enis I think we can use this +Reconnect logic was added to +If you finish it before we release beta please correct the version +I amtrying to think of the recovery test case but need to be able to createthe db using a territory based sorting and then the next test case needsto boot the db and find that collator not availableI assume you have the simple test case for creating the db with justpicking a territory that does not exist anywhere +to failI dont see how its possible to use without either breaking compatibility and some apps OR duplicating the the Appendable interface throws which does not help +Fixed in new build version. +talked on irc and pls proceed accordingly +Sorry for the links but the aragorn machine is not visible on the net +I was only able to recreate the problem in Tibco EMS +Only master is compatible with Immutant overlay +OK starting this tonight +I think this is an instance that I discovered while working with a local VM + +We also need +converted port numbers to constants +Unit tests have been improved +Provided a patch to fix the placing of query element +Implemented and now in Roller issues related to Roller release. +with geronimo works believe that what Donald meant when he said to copy the newer files over top the old ones was to replace them change the new file names so that they overwrite the old onesNot to remove the version and install a versionThere are other areas within Geronimo that will still be looking for the old version +Couldnt be simpler +Ive updated the pull request. +Are these intentional? If so can you describe them? Thanks +It is possible that only the xdocs from this patch is appropriate now +Didnt we just do this a month ago? We had to back the changes out because it caused a circular dependency on Windows builds +Also adds unit tests for it +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Contentsee my previous comment or you need more info +There is probably potential for a null pointer in there +As Leonardo said components will always register resources that existsNormally myfaces produce warning like Resource referenced by resourceName +But like I said if the data being pushed is more it will be interesting to see what overhead the master would need to incure to aggregate all the metrics and keep them in sync on a regular interval +Thanks. +Im certainly not using this I just noticed it when I was trying to track down a +Maybe it could be more along the lines of Implicit schema creation is now performed in its own transaction to avoid deadlocks with other connections accessing the same schemaIs this change in behaviour for all implicit schema creations or just for the default schema it was hard to tell from the comments in this entry? That might change the release note as well +patch added +Tx +Sorry +Are there an open source data model and UML tools you prefer to use? we can start to work do not have any preference for data model and uml me the tools name and lets go +Is this patch only for branch? Or it automatically will build the snapshot for cocoon +Remove this issue as a blocker for release +Thanks Jitendra for Review +It would be difficult to write one that checks things from the client side but it should be possible to write one that checks that the application ends up in right FAILED state +Portlet support was extracted to a plugin in +There is no way currently not without creating a role account with the dev list as its email address and we dont allow that as password resets etc would also go to the dev listThere are many other ways +But thenagain who uses JSP in cocoon anyway Feel free to close this +I confirm that in the past analogous issue did happen +newest patch includes updates to all contrib classes that ensure they work with the new package Makes it a pain to have to use find to code and response seems like the right place for the writers +Thanks Jiaqi! +Thanks! +Resubmitted with AndreiThe problem with the patch you submitted is that it would override the work done in +Lets see if Hadoop QA agrees +I confirm this see my email +Gee put something down for a moment and someone else goes and does all the hardwork +This updates the News section of the Derby top level web page +Issue closed with release. +You create a +Right now part of the parsing still needs to be done outside of the new interfaces +Limited the scope of this issue to cover only namespace handling in +My pleasure +bq +Thanks Alex! +Also a disposer method may be bound into multiple producer fieldsmethods +It should work but is missing the new configuration parameter +no reason we have to fix it all at in trunk backported to close after release of +Committed new tests for nested support in wsdltoidl toolEdell +I think we can close the issue and open follow ups if and when we spot issues +Subroto Can you please take a look at the failing tests +merged with latest TM update resolved conflict in The patch is complete replacement of in r +Thanks for in rev +Fix applied in master in JBDS with HTML archetype. +is a better name +Id be fairly sure the correct one is projectsjaxbintrostagssrcmainresources but perhaps we should let the guys make the final decision +test was added +They are instead parsed into a +I dont recall we removed it from the feature set +ve just committed this +Was it another operator +Will be merged to the trunk when we will have the okay from the PMC +I believe as long as the existing tests pass it is good so no test is addedLet me know if Im mistaken +Enabled some of the tests which were not running under security manager earlier because of this bug to run by default with security manager +Adds a Map coercion +Seems like the expected behavior would be to look for a base constructor that has the same parameters and types first then a parameterless constructor if there is one + +It might make sense to introduce a hook to have multiple sinks for this sort of log data +Fixed up tests +Patch for yahoo dist +Ill fix that in a moment +I am meeting with them on Monday to see what we can doAlso the hod documentation is not up to date so the interface there is not correct they have changed it last weekI will update the patch today ignoring ssh part and then follow up with ssh changes once I figure out what needs to be done +The WSDL which wotks for but not for the nightly build of wiht latest svn +Committed to trunk and revision +WillAre you OK with splitting the guava update into a separate JIRA for now? If so Ill commit Prasads patch +I thought that session usage for target url functionality is just implementation solution and didnt have to influence on it +So the problem is that is broken after this change its at least broken for Strings from the bootstrap compared to those created in the running VM +The above wrapper is a good workaround combining criterions for one association path but doesnt help for +bulk move of all unresolved issues from to +The mapping from to xsddouble was missing from the default set as a result of an editing error in the recent remodularization. +That person will review and decide on your patch +The from the tarball on does work with +Thanks committed. +Also I tried Tika Jaxrs network app server which throws Server Error +Yes these are definitely linked +I kept the general gcd and gcdPositive method separated to allow gcdPositive to be used directly within CM classes when we now for sure that the numbers are positive +Your patch is fine +What do you mean by this class? Configuration +I already pushed your Service Interface to trunk as well as a new Task Service +Can we also find a way for Windows +The war is not getting deployed saying Substituted for the exception which lacks a String contructor original message Attribute cron is not allowed to appear in element pollerCan you provide pointers on how to implement spring batch admin for version in Y +New version of the patch the comparison checking made more generic +This is resolved in rev +patch +We are supporting WeblogicWebsphere and Appserver +Well in that case uploading new should be easy +Hi Jeremyi attached a simple JSE Test +The docs say the deprecated one simply forwards calls so there should be no of this change +I am defining an and using it in FOREACH +an empty string +I then change the limit of the number of open files for the process from to +I dont think this is true anymore post at least when the keytab isnt set +For synchronous routing the event is emitted after the producer has completed +Do you think the name is good? or simplier? Its not really one of both its a mix between wrapper and the real filter It wraps the query but does the get and himself. +Committed as r +Thanks Sergey +ASF KiranYour patch is in trunk r r r +com +patch containing a suggested for reportingWe are currently investigating a major modification of the API which would involve dropping the Component interface +bugs is already subscribed to + +Closing as a duplicate of which has more work on the same problem +I think is a good train in which to do this rename when people will anyway take a thorough new look at all the changes and most will hopefully discover that they do not need forceMerge even if they used optimize before +I will create an other JIRA to trace this +Almenys en el llibre que jo tinc! +Willem did you fix this for the release which is in vote now +Making patch available +For earlier versions number of workarounds already mentioned hence its OK to mark this as fixed for release +There seem to be some discussion on that issue whether this is a problem or not +Would you please help to try it +I think new patch is fine +back port this issue fix +This bug was identical to except for being in the gateway server iterator instead of the MX hosts iterator in the code +From the Subversion reference card +I have removed the check for listener null and set the thread timeout to s to tally with +Resolved with the fix for in CP via +This is really cool to see it had always bothered me that the called run so its definitely good to be able to have multiple running in parallel in one job +which contains a test case that shows how triggers fire out of order if the time zone is changed between two CREATE TRIGGER statements. +In our code base that will never be autodiscovered anyway because we did not license the postgres plugin from Hyperic +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +Ive just committed this +Btw Im going to hold off on committing this patch for now because like it needs to support KSSL for getting the delegation token +I think that should work +Closing issues before next release. +Patch looks good +Thank Giri! +See for the diagnostics plugin. +Committed to trunk thanks for the test cases you can provide in the testcsharp folder would be great +Ive now done so committed in r along with a unit test. +can this really cause the bug described +Fixed +Fixed by an update to in revision resolved issues +I chose to resolve as wont fix to reflect that were not going to pursue changes for the JCC Driver. +Agree with Kevin that this isnt meant to be a complete catalog of quickstarts but simply a starting point that mentions a few of the easier ones to start with +This will happen as part of a more general refactoring of activators in +Why are they abbreviated like this? It was much more when the classes were named with full words instead of would like to discuss further at the docs meeting if possible +Fixee in revision +Arraycopy code generation relied on the fact that jithelper call is the last in basic block +I have checked the source code and it used the default excludes from which excludes +See +Looks like is the culprit +Duplicate of +Reassign if wrong +Thank you for the patch Hartmut +Updated the SQL State messages file based on Jeans commentsThe files are patch to the trunk revision Rick Ill let you decide how you want to incorporate this into the branchThanks Laura +This one then holds the unmodifiable instance stateI already explained The code duplication comes from the two ways to do itOf course for lovers of telescopic unreadbale methods we can still add some conventional factory methods taking tons of parameters but internally use the builder +Still awaiting PMS ve applied the PMS colors to the design +The test failure is unrelated to trunk +Remember to update the cache so it caches on param types too +Thanks for the help with this! +The Guvnor app is using Seam which is intended to be independent of app server and so Seam doesnt really provide the Subject so the can access it +Great! Ill try it out on and trunkSpeaking of Tika have you ever seen a tikaconfig file? I cant find on anywhere on the web or the Tika source +Simple patch to make fileinstall wait for configuration to arrive instead of creating +Is there a hurry? Ill need a couple more days I have several comments already +The problem is caused by a bug in which erroneously interprets character in the binary stream as EOF +REOPENED +I think the problem arises when the reducer the data +That should by the way be part of the Spec +That sound rightPerhaps another solution is to not override the get but instead just provide another method like get or getCompatible +I think can be used for this. +bulk defer to defer to +Tapestry logs at INFO level all the module classes it loads at startup +Thanks Chris +I was wondering if anyone has run and derbyall on the JDK on Mac +Thank you for reporting sorry that it took so long. +Ignore last comment attached to wrong issue +The submitNext logic didnt correctly handle having the final row number proactively sent +Ok I got the same error as with other bugsErrors read past EOFThis JIRA however doesnt appear in the assigned to me list. +What is it goodfor when I say shutdown server on localhost and thesoftware does Oh I doesnt find a local server try toshutdown any other server on the net? In my opinion thisfeature should be switched off by with this property all works as again and sorry for calling it an +Im following up through the support portal +There are assumptions we are making about the setup that maymay not be true in practise +The developer only sees the methods and its Javadoc and of cause assumes that the method really does what it reclames +Please reviewtry it +Tests passed cleanly on +Attach an example that reproduces the priority until an example with steps to reproduce is reproduce without an is currently happening on with Grails. +We want one service that connect to every node in the cluster +Its intended as an optional setting that will default to off +Mostly changes to the configuration objects left I think but it didnt pass the tests when I applied it against the origional point so Im not overly optimistic about this one right now +er +looks good +In we clone the delegation token associated with logical service name to two physical addresses corresponding to the active and standby namenode addresse +do not reproducible in at +As a confluence administrator select Auto Export in the admin dashboard locate your space in the list of templates and click on edit template +verified in revision +Closing all resolved tickets from or older +Assume closed as resolved and released +It seems like we should highlight that somewhere in addition to your comments on the command line +Tentatively set for Engineering release actually fixed in Engineering Release Will be part of release. +This is related to have JBDS installation visible on windows in control paneInstalled Programs +Love this community +Maybe it is possible to catch and as for return hostIpSounds reasonable to me +I dont think we should allow worker threads to be configured +Changed default table and +Orchestra might have fixed it but what happens with other libs which do the sameMy guess is Trinidad itself should look for javascripts in hrefs and then omit the encode since technically spokenin a javascript trigger no other url info should be appended nor the code has to be altered by other layers +CorrectionThis ticket should be marked as critical the workaround allows the dataabases to be created but Hibernate doesnt know to quote the column name and chokes on startup +Also take a look at what can be closed here please +MartinI think this will pass if you shutdown jboss delete the data directory start jboss and run the test +Doesnt look like theres been any progress on this? Is there a known workaround? +require release closing stale resolved issues +Verified in +and it should probably move to common +I committed this +The jar included with servicemix is +Heres a possible fix appropriate for binjruby at r +Fancy taking a stab at this? Maybe we need a ping protocol +However I do not think this is either the cause of the OOME or is likely to solve that issue +Updated patch the version actually needs to be updated in multiple places +I shouldve mentioned this patch is for trunk +Hi CarstenI cant see the nbsp on first glance could you provide some steps to reproduce and be more specific about where the text is located +The root cause of this is the fact tha the outputformat in the escaped gets reused per file it attempts to write and that the seqfile output format assumes that it is only being used by a single file handle +I have no intention of what so ever to provide a workaround for each and every broken CGI script out there +But unfortunately it does not suit us because maven convention is best suited for single artifact builds it makes it difficult to support language specific builds +Closed as duplicateI searched for duplicates before reporting this but JIRA is kind of hard to useAppears I was only looking at the reports for did not even know existed +Committed to trunk +Sorry for my spam. +Do I have to do that for every infinispan component? I would have thought that the internal artifact structure is not part of the public face of a maven artifact +I have found a similar error but Im asking just for in the Java Side but +VDB and both datasources from portfolio example I only modified tried with latest version and I can not seem to +Hi allWe have integrated the most recent patch into our install and the Out of memory fix suggested by Peter +Use to send the and to send the or in the second example +NPE is not seen for a while so closing this issue for now If NPE pops again we can reopen this +New jacorb upstream branch with all patches created +legacy version refers to tickets opened against the Embedded project in the prior implementation as a module of the AS +bq +At least now I have a test independent of +And we probably want to fix the verbageexamples under example to make it obvious it goes to logging and remove the deprecated file example +This is marked as a minor issue but for modern sites its actually quite a blocker if you want to use webflow +I like v fusion I guess its just a matter of making it into the shunky style as per v the existing v is a bit too basic and looks like a knot rather than a particile collision and that make any sense Cheyenne thanks for all the good work they just keep getting better resolved with the close of the Drools. +FYI Tika does provide the Fork Parser for cases when you want to ensure the parsing cant affect the parent +Closing as WONTFIX +No I dont really need this protocolOk lets focus on other stuff . +Instead of adding new parameters breaking all implementations of the interface we can instead add a new method to the context and deprecate old context methods +today i just try my luckto my suprise problem is done Christian BauerI love you. +Agree +Closing this record since is now released. +In a test this happens during a Before method +it works +as described in the jira pool size is configurable. +Ill also update BZ as requested. +Running should be good enough mvn clean install ant veryclean Dresolversinternalant test is a little unstable with some failing tests +I left out the tests. +Fixed in the trunk +In the case of the JSF that would mean adding them to the deployment after the JSF SCI has run +So apparently this new is not the same as the old +Freezing gets closer and we need to sort out the most urgent tasks +this issue needs to be retested since resolution +Looks good +Its time to move on +I the timeouts are already there +believe line should be not cat +Added to templates close after release we shouldnt add Google Analytics to old Labs pages as its a deprecated system that users will be moving off +Tomcat only accepts connections in our config. +The original files will no longer be there to use as replacementchange templates +All issues it depends on has been fixed branch also merged to trunk in r +What do we think about using Flash though to allow this to happen? Requiring access to the filesystem that Solr is running on is can be pain +Especially since theres absolutely no reason why one would occur on that line +Hi ClausI just saw that the release is now GA +Patch applied +I also have it appear with at the first deployment +Fixed both on and trunk +Sending trunkassemblysrcmaindistributiontextetcSending trunkgshellsrcmainresourcesorgapacheservicemixkernelgshelladminetcSending trunktestingsupportsrcmainresourcesorgapacheservicemixkerneltestingsupportTransmitting file dataCommitted revision +But I agree with Howard on this one +Adds a utility class called which allows a generic framework for stack based walking of Node trees +Andreas recently checked in a fix for which might have fixed this issue as well +close off release +So I will wait a little bit to see if this JIRA is needed for out task that needs finishing +VincentI tested your sample with Hibernate and I worked fine that why I believed it wasNow I can reproduce this bug with Hibernate too and you are right its strongly related to the log levelBut the bug is always there. +Something is off thoughFalse alarm there was some sort of build issue +Quick questionI dont remember if weve gone over this before why is the sortedness information considered part of the schema? Shouldnt it be part of the statistics +I should try to keep up again Your code and the comments on are an interesting reading +path relative cool how can Tomcat WTP adapter have multiple pages that works and we dont? if the page cant be persisted how are users supposed to use this +The user probably needs the ability to configure a timeout setting that will apply to all client operations +Minor comment that itll want to be logginglogj instead of +I feel like scrub is only for attempting to repair damagedproblematic sstables +Is there a jira yet? I can help with this +allow specifying a subset of columnfamilies to check for version didnt actually limit its main loop to the requested only the locking +The workaround is to use the next version which isnt release yetApache motto is release fast release often a motto we dont follow sadly +Reopening this issue to keep track of the fix not being applied. +ear was broken for exploded deployments +implements notAn administered object instance that implements interface must ensure that its Java class and the interface type implements and interfaces +If we run into this kind of problems in the future we might develop our own or maybe use the shims layer +its actually a legit conflict not just a merge issue +How do I achieve thatCurrently upon recreating a page I get an error +Also added implementation for updateCountersStill need to figure out how to represent conditional tasks +so I think this looks good though +main issu has been fixed but this one is still +I will close this issue when this is done +or we have to provide some way to update RPC tickets +I am for checking to see if there is a builder subclass when the user specifies the class itself without the builder +for the patch +Refactored to use class instead of interface +revision in JBDS h Fedora +I wouldnt call this a blocker scheduled for +This issue is fixed and released as part of release +looks good to me nice catch erel! +One question have you considered perhaps moving it somewhere even more generic? Perhaps somewhere in Common? This functionality doesnt seem to me either +listing both will check both every time looking for the newest +Pulling back into so that we at least have a look before goes out +Thanks Willem +The + Brett +Philippe I wanted to tackle that one but your patch introduces a dependency on +Suggested fixIf we find presume this is standalone and just break out of the loop looking for our host name +FYI the branch has been deleted since all modifications youve made were merged into the trunk +Not sure whether that comes under the purview of the question you raised +The extension points are not very well defined so its hard to figure out what to override and what not to +Thanks Harsh! Ill take a look at yarn and see if a similar situation is present there +A Weblogic Jar is now generated +Scratch that last comment the error was on my side +I am sorry Sebastien +Way to read the description carefully +reopen for the suggested close to M to risk it now that it was clearly a memory issue +No run no support from my side sorry +Containers must contain values all encoded the same way so the protocol will have to choose the most effective encoding for all the values +set href to when its leaf r +no compression is done in exploded archiving +Just returning a constant is maybe not a case in which we should do this kind of thingPart of the optimizations I added is to gradually change values from being always boxed to be used native as they are +I attached a which contains exactly the most famous case of failing node while starting the cluster and it breaks the entire cluster at the end +patch? +I doubt it is due to real network issue the proxy is a little slow and Harmony does not wait for its return +Ouch +Im pretty sure there is no way to tell the difference at runtime from the framework point of view between a dangerous error and a simple inconvenience +We havent thought about using a library to generate the JSON +Some fiddling with is still needed +Sebbalso almost all +Probably more useful than getting the release posted for us would be getting instructions on how to do it ourselves +Use precalculated expiration to invalidate cachePlease apply + of a bulk update all resolutions changed to done please review history to original resolution type +Do you think I should ask a solution in spring forum? At the moment the workaround is not to use generic types +Made change +It does support it but its called bindAddress +It also resolves the problem where I was using the actual column position in the trigger table rather than the relative column position in the resultset for the trigger table which will be created at runtime +A stressful test on a slave cluster with is still stable after hours +I dont seem capable of this ticket to Juhani if anyone can please do so +Thought it was good practice to delete old versions to avoid confusionWill follow in the future +This looks like a dupe of +Luke Daryn can both of you please summarize the packet exchange for simple kerberos and for each of your approaches + +This is fixed in upstream master +Does the spec state explicitely that instances can hold DIRECT references to the metadata +Sorry Ted I should have realized that my patch was only againstCurrent state applies to branch and was ed by Ted +Fixed in + done githuib will catch up from that shortly +rUbuntu smoke kernel +If I face the same issue I will reopen itThanks for your swift response +Patch is applied at +Mouse selection would still work in that case +DhrubaSuresh I think the suggestion has been a little mis understood +Thanks Mario did you remove the Ignore on the tests showing the problem? +Thanks +found some other places that has the same is a weird issue +The fix looks fine. +Implemented pairing with Henry StorySocial Distance based Authorization is left to be part of other issues. +Example application that uses a image in a lazy loaded panel +Bulk closing stale resolved issues +There have been a few small changes in the english resourcefiles for struts since the japanese patch was supplied in FeburaryAre there any patches you or others would like to provide +What would be the advantage of this changeI can see a disadvantage The proposed error message would not point out the actual source of the problem anymore +Im in the midst of a big JI refactoring that might make it easier to sort these sorts of things out +Sorry I was in the wrong log +editor is implemented +Ive explained why several times above and will try again now +Added Alexander as a contributor and assigning to him +I checked in a failing test to the branch +There is probably a more correct place to add these +why does this happenIt is because there are excluded datanodes +Plus I absolutely hate it when plugins get too clever and start forking off lifecycles as this atomic build behaviour is hardly ever what you want in a complex build +What am I missing +This patch is based on axis revision + +Closing all issues +We need to note that your require statements are going to break in the changelog though +Fixtest patch +Everything will be BASE encoded for that field so expect a significant decrease in performance if its stored which it most likely will beAlso the parser has to store content and it has to be loaded in the indexer +Please reopen if still a problem. +Committed to and trunk +If this is an issue we could store the value as a +Alternatively maybe a random name could be generated but each time a host is started it would be assigned a new name so hard to track which host is which when managing +Predictability between clusters seems to matter a lot less to me +Not really is about improving the workflow to build and publish one is about relying on multiple site in rather than using a issues are independent improvements +Forrest seems to incorrectly use the system locale in a number of places and thats how it is broken +Patches have been recreated using git instead of intellij idea +Thank you very much +Is there a fix for this version? There are currently no plans to upgrade to version +One could argue that this can remain open and we would have support for JDBC metadata but I think its a real issue in terms of JDBC compatibility +l is defined outside of a try block and therefore live at the resolved new array allocation that is within the try block as an exception would mean it could be read with the value of null + + +Unfortunately Dougs C +I would be willing to help test this out if anyone is working on itSticking to ant in the mean time since netbeans uses it +lemme know whats easiest +the panel is integrated and one can choose the application server to deploy to +Please update and check again +Best regards GeorgeThe fix looks good +This is a typical example for that in my opinionbq +Pls do mention if this is not the right thingto do +Regarding Johns comment on Dec AM +Committed +And I dont really understand your remark anyway since the patches you attached to this issue targets both and +Im sure it breaks security somehow so its definitely not finished +There were some that could be changed to but some that couldnt that applied a default value +Please close +Here is the patch to +So this shouldnt require new properties unless Im confused. +JieshanDo you want to attach patch for trunk so that Hadoop QA can run tests +Feedback is encouraged! +Its a bit ugly but it works. +patch that mark race by special API for prevention of further alarms should be patch verified at Thank you for applying it +Updated the and the plans to support the JSR geronimo naming. +This is strange +I assume this does not sink the RC so I am targetting this for +Im running the tests now and if it passes Ill resubmit the patch +Version has been releaseClosing all related resolved issues. +lowered priority +I wanted to add that the part of the query exhibiting the error in the second case is a boolean query with three clauses all +I have done this accidentally I need to finish this task first +This patch is in the svn repository + Use advanced config to set the right parameter as a remedy +Thanks for the patch +Even if you dont get those patches you can read the source any way to see what it does +Makes sense +Thanks. +I see no down sides +Sean I know +yes i think it is too dangerous and would create a lot why doesnt it work questions +The only time I think this is an issue though is when selecting which dependencies to export along with it +Well as I see it +to define the method as protected and create a subclass in the test +Yes I dont think there is a Jira to consider for Groovy +applied patch to HEAD and leave it to Jeff whether the issue needs further work or is resolved +Gonna see what I can do +It will be good to run the nstest again hoping that we will run but this time around we will know the root cause of the problem in clearIdentityThere might be other ways of handling this problem +Do not know of such feature now or before +On debian its named +Ill open a separate issue to address thisI think this is ready +As far as I understand your particular fix doesnt work for case +We are still maintaining piggybank and we will review and commit code once theres a patch +Alright you win Your setup looks good +Attached testcase to disable a be covered as part of new Admin UI work + +Addressing Teds comment +Perhaps this is a larger issue +Here is a patch that should fix took the liberty to refactor the creation into a private method inside the +Ok I have checked and I think it is a good idea to copy that code from deltaspike to myfaces and use it +Ran the tests locally and this just adds one option to create and a test so Ive committed this and merged to +Committed to trunk +I would appreciate your help in verifying that the fix works for you +yes please create this demo project either this will prove there is no problem any more either it will be useful for me to change it into an integration test to work on a fixand well even be able to test with every version then determine precisely when it didnt work +please make sure other jiras contain info that allows QEdocs and users to understand the changes made when it involves changes to UI. +the last one is another unrelated issue that has already been there any workaround for this? Is there some way I can skip the first test run and just let run the tests for me or something +You can get a handle to the service prior to configuration of the caches +Do you want to modify settings on Sonar server or modify property in +The example works fine for me +Linked issue was JTS recovery issue +Ich werde am wieder im Bro seinFreundliche GrsseOliver StrerThank you for your message +Maybe we can create subtasks for this JIRA beginning with RPC and changesIMO those changes dont make sense independent of each other +bq +Committed. +Youre right +There is no security issueMy final point is this it fails SILENTLY +Error disapears if uid starting number is increased +But because there are two types of tokens application level and level we should have two credential fields +I just tried bb and I see the same behaviorcd checkout bbcd wrapping mvn clean install cd +deleteTable causes appropriate deletions +patch looks good +The problem here is that theres no way in Java for us to match recvfrom behavior for normal TCP sockets +On second this this should probably be raising a normal Ruby error like rather than just failing over to the Java error +If no digest is specified then the passwords will not be hashed +I had some related problems with concurrency besides the optimisticlockexception in communication layer +No because leveled doesnt overlap ranges within a level so it doesnt matter what state each range is in as long as its consistent with the replica for that range +Yes If the timezone is specified in the string pig should use that value +asap Ill upload the patchThank you again +Added two new endpoints in the postRetweet and code is pushed +The buffer size can be used as the read size in random reads +The resulting patch is way simpler now drop the call to forbid from +Might be better to drop out instead of running a broken nova command if the instanceid is empty thoughSo at this point I can capture a linux image from vcld setup on Openstack Essex with your nova api moduleGoing to continue on doing more testing +The DroolsjBPM services it wraps were designed to be global vm services +I will commit shortly if this looks good +So +Changing status to FIXED because we have at least one confirmation that the issue appears to have been resolved +Compilation is OKThen I tried hosted mode with firefox +This probably has two Findbugs warnings still to be fixed +Pax Exam artifacts are now on Central Im updating in Karaf. +Does this mean that it wont be fixed then? The way it works now I can hardly see any use at all for the JDBC message store! +Updated after review with balfanz +Lets see what happens when the results for r come through +Pushed fix to latest build fromThere is no regionid in domainaccountuser tablesThere is no apikey and secret key being stored for region tablesWith above changes there is no concept of owner for domainaccountuser entries. +Installer probably updates text first then starts p meta info generator at the end and then update progress bar after meta generation is finished +bq +For example has a link to which in turns has a link to the +Changes must even resist crashes +Patch for branch +With those versions parameters of the mentioned queries get lost somewhere +Changing has implications on the logs +Nicebq +closed after releasing version +To answer Dougs initial question yes we are using this in a desktop application inside a Swing. +Attached an initial patch need to add some more tests and recheck some of the in itReady for approach review +Hi PeterYou bus cfg file with name will be loaded by all bus initalization by default which means all tests will use this bus configuration +Actually it might be possible to implement it in pure ruby +What I am saying is when the connection breaks we clean it up this is not like http session where you can session with ever request +try the diff and shell +Attaching a patch that removes the JDBC topics and modifies three more topics by moving material into them +So all issues not directly assigned to an AS release are being closed +However and newer generate the code fine +The offset row id table should never be too big for ram +Please tell me if I make any you +Looks good! With these latest changes I am not seeing the tiling artifacts anymore +Change column header can be done by click Alter View on List Tickets view page then change the label and save the changes +No sign of an exception yetResponse rate measured via bounces all over the place ms up to ms +Resolved with the resolution of all resolved issues due to a successful release +It uses applicationsecommerce files but now all is in the specialpurposeecommerceThe approach adopted to do as per description was looking fine to me but you say that createcustomer executed twiceThe work done here is just checking the user available or not before creating new user was not able to understand your last comment on FebPlease let me know if works looks fine to you as well then I will engage somebody to upload updated patch +Its not forgotten though Ill be back +Attached a patch for learn under +Yonik Hoss can anybody explain how can an contain an empty string as value? What should be the fix? in or in +Committed revision toCommitted revision to trunk +Thanks Jarcec! +Im all in favor of using REST instead of SOAP +I think its better to file or add docs to more specific bugs as appropriateFeel free to reopen if you disagree. +Perhaps we can even make it so schema checking parameters are kept under ousystemouinterceptorscnschemaService where attributes for that service are stored +can you attach a test case to the issue please +anyone got any ideas +Will not modify current behaviour until the SCR plugin has been the plugin has been updated to take care of the correct service replacement semantics the SCR is also fixedFixed in Rev +at the end of the link was meant to signify the rest of the url parameters and form a complete request +Each Reduce task needs to know when its map task predecessors have completedIn the old days this was easy +Ok that makes sense +clienttrace format into +mingjian Do you have a log? Do you want to open new issue +thanks Aditya +Thats what happens when you dont ensure a test fails before it passes +LICENSE NOTICE and DEPENDENCIES files are generated by Maven and added to all build artefacts +Since we now use native threads the discussion about subsampling for thread switching is moot +As a workaround the configuration in the parent must be manually updated to +Acquire and Acquire is only issued when Acquire +There doesnt seem to be a clear logical line between the methods used by vs +I believe this bug has been around since and it seems to be an important feature +the read and write are done by the same thread +Attached is ACR a proposed change to the DRDA spec for UNICODEMGR which will allow us to implement a fix for this issue +Would it be possible for you to correct it and commit? Otherwise cant interoperate with the Identity Server +The autoload branch might be measuring the wrong thing +Updated for current for current version of the patch put in sync with current like some tests failed due to timeout +Closing all resolved tickets from or older +Im really at a loss why these tests continue to cause us trouble release after release +Thank you Jing +OK I have merged in the latest trunk and it looks OK +So we try to introduce delays between each batch of same progress display which increase exponentially with each batch completing +Mean the performance is important than the correctness? I think Correctness should always be first +bq +We can weed out any issues we want to push to a later release from this set once weve done the consolidation. +and gridmix junit tests passed in my box +Updated with a few changes handle should call handleRollback if streams is null just like it calls handleCommit handle should call handleRollback if streams is null just like it calls handleCommitLet us think about changes if users ask for it +This is because after accepting a connection request on the distinguished port the server then opens up a connection with the client on another port whose number is unpredictable +Addressed Alejandros comments of configuring logger via logj via a newly written appender +Hi Lars can you share the patch? just do an svn diff and attach the file here please +Request the community review this design rev and provide their comments on the JIRA +Can you show us the template youre using +I can run more +Tests could just explicitly specify the update handler on updatesNow that I look to fix this I am not understanding I dont need to change the update handler I need to change the update chain +I have not tried adding new measurements after that as have running sonar was of higher is used to manage the version of your Maven project +Robbie could you please review it +This bug broke my list of values links throughout my application +This issue is related to redirecttrue and output at the end of the last test in the fork +I switched to elephant birds loader which handles this case and more complex json objectsDavidSent from my iPad +Great leak is gone! Just tested master and the size of the thread ist stable and so is the overall heap +Why dont we deprecate the default one and make this one the new default +am i missing some thing +and ant test passed on Y! patch except +Ok I found what was the actual problem +As a note old servers will still retain this information in their server clogging them up in case the user switches back and forth between versions +While one could always define a Xerces feature to allow SAX applications to get such events it must not do so in the absence of such a feature being set by the application +In there is an exception +AnimeshThis is an API bugPlease assign it to API developer +Committed to trunk +In the mean time I sent a pull request so that you can proceed when Andrew is fine with that approach +There are also some other examples on of java container +Per review meeting fits under Usability +It might be helpful to include some more information about the cause of the error say like the JSP that produces the problem configuration etcThis kind of question is best asked on the mailing list first before filing a bug +Ill ask them if they can help. +Theyre also filtered during fetching when following redirects. +Ive attached the mentioned file so that it can not get lost +Looks good +Checked into tools folder +Uploading a new version of this patch whichintroduces new pushpop methods to the compiler context to save thedefault compilation schema when it needs to be temporarily changed +bq +I fixed the bundles on the update site +Thanks Steven. +So you are welcome to work on a patch +That will happen automatically once the svn move is made +yup +Not sure if I am supposed to close this issue or the reporter should but Ill close it since its been committed +The patched table do not exist for the other databases nor will the test be you please create a new bug report making mote of this? TIA + +Ill submit another patch this afternoon or far as code conventions go I now imported the maven codestyle and let eclipse format the code +The two objectives are generally incompatible for example is using a custom shapefile reader instead of relying on the OGR for this very reason +Just providing patch I had quite tough weekend and could not run all tests +All examples have now been updated to use the security API except the todo example which doesnt need it and icefaces which seems to be too buggy to get working properly +Defer all unresolved issues from to +Because I am using Java I need to remove the xalan and xercesImpl jars from s libendorsed issue is apparently due to a xercesjava incompatability +Thats especially important for ajax updates which supposed to be really fast but in size of responses were really huge +Does the bean validation implementation youre using return constraint violations in a fixed order +So why do we use connection profile at all in this dialog +I just had a classpath issue +It will simply help more people to use Seam and help its wider that would create an easy clear migration path towards JTA for applications that at the begining do not need it +What does this mean? AFAIK WTP always used for this cause breakage of the example on previous WTP versions +Issues like this are a result of storing data version in the node itself and as such OL is considered broken +bq +Committed to alsoHence resolving it. +Hopefully it wont make any difference with +Ive converted all the header files +Nice review Thomas +Fixed on both and closing all issues that have been in resolved state for more than one month without further comments +HTTP Header has not been returned when you make the requestThis java script will not work unless returns a value +Please reopen if it didnt. +The Accumulo web page can be updated to point to this example anytime after is released +I think the value added by these assertions is negated by the fact that it is so easy to forget to update them +Actually graphbase is using an hbase table as an adjacency matrix +See for more details +I agree this is a matter for the plugin in the ide +After trying this over and over again to try and isolate an error message or repeatable pattern I think Ive narrowed down the circumstances that trigger the things only go wrong if I delete a subproject while things are still being checked out of subversion +you want to close this issue as wont fix and edit to make it more generic or keep both issues +Help text is hardcoded into each command class +Will submit a patch sooner +Just work around this issue by doing the input stream to document transformation without using +Verified fixed +It thereforedepends on the log shipper integration patch that has been submitted in +Mladen can you please investigate? Thanks +I will get to it tomorrow +Patch applied at r. +This bug has been marked as a duplicate of +Once the clustering is added back we need to bring back the needed administrative calls +Verified by Sean at rThanks +The attached patch fixes this +OK tomorrow morning I will crop the two of them in order to save more. +Im not sure this is a valid issue but I think the current behavior is annoyingI read through the specification and didnt see anything specifying that rendered attribute must be checked during Apply Request Value phase which also queues botton actions for later processingThe condition of the problem is that a commandBotton has the attribute rendered value bound to a backing beans property +Backported to. +Its a little late in this cycle to do a change that have a ripple effect on documentation etc +FixedAlso files from excluded assemblies do not have any coverage data. +Btw Jake Ill let you review this one +and committed + +I finally got it +This will support the above usecase where the update has been made through the Admin API however if the modification is made through the then the is updated however it will not switch the sources until the next of the server +I think the same change needs to be made in control +Attaching the original implementation I need to update trunk and see if I can remove the code from reordered to print summary first +And as I seem to understand a bunch of error log I see are fixed in the recent recent version +Im wondering if the environment and its properties would also be available through applicationScope? +This patch improves the handling of correlation names for the source tables of MERGE statements +thanks for raising this +Well have to fix this issue in a later version of this plugin +Also to better match command line behaviour only dependencies are added to jee module dependencies now. +Please close one of them +Ok JacopoIll take a look into it and hopefully have the another patch to allow to create the RFQ like a PO in a couple of daysGreetings Rodrigocan I close this issue or are you planning to work on this +Upstream issue is fixed in AS +Hi JuvenThank you for these detailsHow can we have Bamboo to deploy to the Nexus snapshots repository automatically? BenAny update from your side on setting up nexus snapshot deployer account on CI server +archive +This involves using the from Hadoop contrib writing a custom and which splits the document result across the mappers +it creates an element which is not there in the wsdl +on patch for trunk Enis +Consider splitting things up into more granular change sets and submit least complicated intrusive first +Committed at subversion revision +Tested with a patched Weld core jar and this works so sending pull request +Close after release +Please review +I think you missed removing the lines that stores the infoPort after is created in got and registerDatanode from +I have gotten to a point where everything will at least build but I havent tried testing beyond getting timestamp validation to work +An append to the end of the array is performed then the linked list pointers are changedDynamic resizing needs to be implemented +We can narrow the scope of test failure and make this test deterministic +Lets tackle more urgent issues for now +I think if you want to record the inner codec nameyou should just readwrite a attribute +This patch adds a change to the previous patch +It looks like the generated classes and do not derive from +Hi Patrick Id like to push for this to be included in +Is this a problem + +Marking closed. +Looks related +Im not looking to actually store the access token I just need to know that its valid +Later comments use the term User Defined Types +the was for sure some additional potential for improved ac readperformance but those most probably require changes +Hi AndreaI have uploaded a smaller file which exhibits the same problem +Bowen Dave can you check status please and appropriately dispose this off? Should be postponed if its not fixed yet +And attach an example generating the error if you have this issue again because as you said original problem was fixed waiting for the new in advance. +Ive reopened as a reminder of this. +Thanks +BTW +fixed in rev +Thanks David I will look at your patch ASAP maybe I will be able to commit something during the nextIve noticed youve comitted the patch to CVS +no I ran it on a sample project +I would not mind closing this issue +Moved to CP for triage after ships. +Ive rebased and updated your patch +I have applied it in Rev +Thanks Gary thats exactly the info I was re targeting the end of the week to have this completed +has been integrated with trunk so proceed with submitting this second part of the fix +Im using jdk and my OS is Windows XP +Closing stale resolved issues +Verified by Vladimir. +So the regression is real in your micro benchmark but is caused by the more compliant EL code that I pulled from Tomcat +Bulk close all issues. +Do you mean because NNAs ZKFC does not have Cs address or there is a fixed mapping of the ith NN to a server and that mapping is staleRight in this case even the namenode ID changed so NNAs ZKFC didnt know what to do with the information in ZK which said it should go and fence the old namenode IDbq +I hope it doesnt add much runtime to the test suite. +And also it is very obvious that LZO provides good compression speed and good enough compression quality +May be we need to come up witha a solution that works for both if not then we can make it invalid +Changing the VVS behavior couldpotentially affect how other resource loaders work with it +Please review these changesIn addition to the changes mentioned a bug was fixed in where the tearDown classes are obtained always from the default reader not the reader used for persistent instances +Update from Aaron it will be a addition since its not a bug +Updated something strange went on with this build +Thanks LeonYour patch is in trunk at r +Gary can you create generate files for your plugins? Is it safe to refactor the features so they follow the convention instead +Uwe good idea Ill do that +I agree there should only be one way of doing this +The third option is certainly the most user friendly oneHowever in this case something is changing and I dont know what +I hadnt considered this environment makes sense +Building a custom jar file for customizing the images of the input calender is somewhat of a big hassle for some most users to do +If I see it correctly the current implementation creates a resource provider for each symlink I fear what happens if there are hundreds or thousands of curious why do you think we should add both patches to the extensions +Thanks for the quick response Eugene +Rob are you saying that none of the other tabs are applicable when using RSE? Because you only execute the server using one ssh command so things like PATH and everything else dont apply here +Im uploading the patch +Done and done. +Closed upon release of. +Assume closed as resolved and released +So it looks like your assumption about a bug in the progress dialog looks accurate to me +wherehas it been documented for other developershas the site plugin for maven been configured to use it +buildr seems to share a classpath for java calls so after creating the databases we have Derby in the classpath +Updated documentation +Feel free to set it resolved +Handled quote characters in text selenium tests added. +thanks Lars +To get network for the private gateway we searching with vlan and cidr +Almost every review I did with projects using Spring Batch had provided some kind of hacked way to implement this +Yep just checked and it is definitely there now. +I couldnt reproduce this bug with trunk +Yes entering in the variables is much better than typing up new messages I look forward to Andrews solution postFor the most part I did use what was there for +Very often you have to guess reformat redo guess again and so on +Forgot to mention please remove author tags from javadoc +trunk +Sorry +So currently does any spring taglib work with portlets +However this may be hard when the patch has got in for a long time +I saw +Seems fine to me +I will look at the code to see what is involved in implementing this +But Seam should not look away and tell us what went wrong +Looks like you have to be a developer to close an issue and it doesnt look like I can change that +Hope that helps +Thanks Brock! +This problem seems not represent in and trunk version +Patch looks good to me only change I had to make was to add the new containerId param to the base class tooThanks for the patch +To isolate the issue time values were collected for each module in each request +Also can you run test scenarios in guvnor that use that global? +No sign of automated builds so deleting that item. +Also I dont think Im crazy about stepping two years +If you have the root category of your site set to root you should not see this problem +If it still seems worth doing I can add a Reference Manual topic on the property +Data will be stored in a extra file +The error above is what I get with compact python support is very important to us +I cant see any change in +This patch changes the way that the NATIVE authentication service opens a connection to Credentials DB when privileges need to be authenticated +Running tests + +However I am updating the restartstore tests to verify that the master key is being persisted and restored properly so were prepared when restart is implemented +Hi matt is there any chance to receive file from you +I am thinking that it can be made better and more complete +For search engines this is not nice either to have changing links for same content all the time +replaced Model with in to incorporate all and inheritance calculations in a model before using it to transitively retrieve ve tested this in marmalade +For the db connection exception thats environment setup issue and you can be better helped by someone local to you +since the apache servicemix and servicemix share the same component maybe we need to fork a branch for the maintenance of camel version +Thanks Devaraj K! +That issue is about doing each mutually exclusively +Ok phew Ill commit the improvement to test. +Ill send you a schema +If Im correct we can close this issue +My had assumed that we only had one message counter and hadnt looked closely enough at the ll get it fixed and do a full search for other instances of this. +These issues have all been resolved and just need to be closed. +This is working as expected +The second one happens after the buildfile loads and you can extend it as you see fit and also use more specific tasks like helpprojects +typo in the section getting the property value from o +Should be fixed now +Ive just hit the same issue +I think we should create a module that contains stuff that requires dependencies +And all is ok +Perhaps we dont even have this header comment since the properties are defined in the documentation for agent configuration + +Ronit is not really done like in this IRC discussionDoing it like this would affect and alsoI will evaluate and fixAnyway +is the real fixPer I made it work both with and without but only by making punt in the case where it sees a table +Checked in documentation about the process for IP and licensing +If the continue breaks instead of continue next loop this should make no difference should not be present in the counter is in the switch as well +This will receive the messagesand pass them on to the other classes that will perform appropriateaction with themA javaengineorgapachederbyimplservicesreplicationnetThis class has been designed along the same lines as the only difference beingthat it does not have ssl enabled socket creation ability for nowIf this is required this can be added in the same lines of the originalclass at a later stageA javaengineorgapachederbyimplservicesreplicationnetThe generic message unit that is sent between the master and the slaveM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceThe for the exceptions that are thrown when a exceptionoccurs during Replication and their corresponding messages +Thanks for that analysis Knut +Attaching patch for with fix from addedAlso attaching a th patch with the thrift generated file so that those that may want to use the patch dont need the thrift compiler. +jar file upon redeploy the home interface isstill being loaded from the old jar while the remoteinterface and bean class are taken from the newly created +should be close to there +Previously I was missing a filePlease scratch this and thank youSinclair +User config error +. +Brett I am fine with the current patchs behavior unless someone else in the community feels otherwise +this was previously required because the interfaces members from all our RPC interfaces were sorted then sequentially assigned which were used in place of strings +Thanks for the patch! +The Jboss can be used with API jar and vice versa +If you look at the file in eg or you will notice some maven stuff that includes the source from into the artifactIf you look in the and +By the way this patch still contains some casts in getAttributeaddAttribute and is deprecated should be replaced by UweI updated the patch now it uses and does not cast anymore have done this during removal of old API. +My bad sorry about the pep issues I did start to clean it up but missed a thing or two before I patched it weird my vim plugin didnt catch itI started looking at the tests and Im working on it right now +Koen examples would be a great addition for but if not we can push to of since no answersreplies and it seems related issue is not solved either +This is happening because the itest is filtered from the source distro but still present on the itest +Instead of modification dates it makes more sense to keep track of creation time +Issue pertains to former Component API and has no correspondence with the new Sling API +Updating test case to reflect the fact that offset info still gets screwed up in the XML processing instruction case even if there are no XML elements in the source am getting a ton of these errors in real documents with the nightly build +were backported to as well adding the latter fix version. +That might need to be revisited for +I updated release note +Now just need to add private ssh key to ssh preferences during cloning processVerified in Developer StudioVersion Build id Build date +How did you manage it +it shouldnt be in there so that is there any other s in the server? would be good if you could search for that file in and come back with info on were the files are and their screenshot with search results +But hopefully we can avoid large patches like +Generally Id expect things like that to be careful about maintaining back compat or a mode that can run older version algsThat said Im not saying this change isnt worth a little algorithm disruption +Why is used in some places and in others +Complete with rev dbeadbddfef +such an issue cant be a blocker because there is no official osgi support in this I just tested the injection on the servletConetxt and its Ok Thank you so much! +saveNamespaces javadoc should update the sections Usage and see + +Do you have any idea of the problemNo offense but seriously how are we supposed to tell without some demo project to reproduce this +Im fine with going a bit ll chat with Jim and see what we like best +For below it appears that both of my emails are subscribed as list moderators so thats great +Ill put up the results here +Attaching patch against Solr +I ran into an issue withTimestamp but that was resolved in +to and +fieldName cause notNull sizeMax add field string class +patch ignore the file doesnt work. +attached is a patch that enabled for compilation and bytecode generation updated building doc page as wellWhat is missing is to add the missing JDBC methods in the fake connection when doing so please dont use Override annotation so the compilation does not fail with jdk +Until then I dont think committing this is right yetThis test exposed in the first mins +Votes +I find it a bit troubling that must be specified explicit as the map and job output types instead of just Agreed but I was following your leadSounds good on MS and Hadoop needs to know the concrete class so it does not have to marshall the class name with every instance +fieldName storniertadd field string class +Do you receive an empty Ajax response or the page is reloadedWhat is the expected behavior +on +Ive noticed that in the moment ive clicked on commit tooIf i rebuild that component with and replace it in my deployment then i get the same error +Thanks Fay it does resolve the issue after setting to false +hope nobody objected +Therefore neglected it on purpose to avoid the overhead in downloading one for all tests for this tiny feature +Thanks Sushanth! +Verified +I dont understand what you already can use such keywords by backticking them User +If someone wants to fix this by simply changing the space to a plain space I have no objection +Great! +I must have missed that in the release docs +If Groovy works then Gant should work since they use fundamentally the same batch find I have to set GROOVYHOME ANTHOME and JAVAHOME for the batch file to get anywhere which is reasonable as these are fundamental dependencies +This is a key feature for without it the page customization wouldnt workIf you put the before the your tapestry pages wont take precedence over the routes +Ive read through the spec and couldnt find a naming rule that applies to Hibernates behaviour +Id have to look for the specific reason +You can even search for yourself by just typing your username into the global search field and pressing enter +You need to modify linking to DB server in and the class path of the Jackrabbit components +If nothing is passed by default the history code selects filesystems default values for block size and bufferbq +This feature was supported in Castor JDO as noted by Nick Stuart but has been lost with the release of Castor JDO noticed there was a change to to support multiple database elements per element in the but this does not appear to be reflected in the Castor DTD +The stuff is fine but according to Jaeho doesnt seem to be working +I do not understand the design +We have not found out where is wrong +Because its a feature request +Please feel free to make changes to the test or change the parameters to the test +Im sorry to be so harsh here but IBM is wrong and they should fix their IDE +Also I dont see why this is actually a problem apart from copying unused resources +idea reports these as still unusedinvalidateHelper prightcompactHelper pleft pright +Ill improve the function to add the charset as a argument to the function as well and use the default asThanksSupun +Ive never done it before with Git so please let me know how it went +There is an example configuration file in the contacts sample source directory which you can also use for reference +And harmony uses similar or even less time than RI but on windows XP RI costs less CPU which deserves us for further studyGood luckLeo +However we do print some stuff out to stdout and stderr in the server on startup +path segments +Added list to +I double click one of the containers in the properties view down below +Give it a shot Horia +This request will go into the NEEDINFO report for Release Notes. +You could debug into and see why your isnt picked up there +This looks like the right fix although we should combine these into a single block in the switch +Any dual boot opportunity is very very bad and hard to track down in a production environment +However the performance impact is caused by synchronization +This way if there are not very many events we do more writes and the events are output quickly but if we start to fall behind in the writes then we start batching them up into bigger chunks which are more efficientYep +Thanks for the patches! +this issue can be closed from my point of view +On a second thought let me post an amendment here +Workaround build first as it will download the mail jarActually wonder why need a mail jar for testing +Test failure is unrelated +PR applied on master and with JBDS L +patch from RB +It is not really a Grails bug but a Hibernate one so there isnt much we can do about it until the issue is resolved in Hibernate + patch +Thank you +So myassumption now is that you are just using outdated version of Struts +If so thats a you are right i can see from the source code that there are several throw exception +That was always a short term solution the long term one being +to reporting an error +is the file that I have modified by hand to be what I think that the results should be +My problem is that when there is to somewhere in classname or path the no modified path and classname is not checked +To me this is exactly +Found itthanks. +When I started the second it failed with a lock error but didnt delete the lock file +but if u repeat the recipe it will fail because a points to atmp next time +Could you give more details about what you want to do? I would question why anyone would want to use in a web container environment AIUI is a way to get the Tuscany standalone runtime able to run web style applications but the approach taken doesnt really fit in with the JEE web applications or with the SCA +Notice the prints surrounding the lucene API calls for adding deleting and reopening +Brill I put a fix for this issue but dont have OSX to test it +You need to specify the list of parameters to pass to the next phase by setting the renderParameters property +OTOH if you do want to get a support contract why not just email salesCheers like to point out that we are using in our development of a distributed databaseThe UUID is absolutely essential in this environment as we cannot rely on an integer identity or a sequenceIf you can suggest another way wed be greatful +See +hm +Raise instead of dont issue the backtrace andupdate the test to handle the new Conflict Exception +We can give coverage layers a different name closing all resolved issue that have not gotten any more feedback or comment in the last month +Will commit once Hudson gives its report +Will continue looking at this early next week + IT test as there is no default global we could use a fake plugin to test? +But David I allready promised you that you ll get everything as you want it +Thanks Arun Divesh Done at rAshish will be making some more improvements in this work +The fix was simple I modified resolveScript to return null when we plan to use an internal command removed the get from the relative path logic and all is well +Thanks ChristianYour patch is intrunk r r r r +Still I was wondering about your exact because as described above the you mention should be covered because it would trigger a which will make it to the caller +Seems nothing else needed here. +Good catch +By default we want to turn ON mslabs? If the config param is missing from the xml previously it was taking the value as false +It should be a deployment error +Being moved to release so can evaluate how this functionality can be aspectized so that can be external to remoting core code and in a seperate area. +In my opinion the best we can do is find out what the officially supported way of undeploying an application is and recommend people use that method if they are concerned about this +Unit test illustrating the issue +I really appreciate it +Hi Mike I am fine with that +please prove me wrong +Same goes for addIndexesIf you want to add fake data to your index thats your decision again wrap your reader with a that adds fake data. +Im not sure the impl makes sense in Groovy since is kinda a messed up concept +Accident or wishful thinking +I have tested the version sample and it does not send empty headers. +So the culprit might be a buggy plugin that generates the file with that problematic name in target +HiIn this page there are some improvement in file of party and ecommerce componentRegardsAshish +This functionality should already be provided by pread andor seek read +See Marks comment above +Closing all ResolvedFixed issues lacking a Fixed Version +Committed to trunk. +Refine the patch and add a test applied at r Thanks +We can open new issue and cite this one if we can reproduce. +The patch incorporates Konstantins comments +Corrupt replicas correspond to blocks that exist on the but the state of the replica is different from what the expects +I havent forgotten about but have been swamped lately +So Im effectively proposing a depth of +is implementing online schema changes is adding persistence to online schema changes during master failover +Could I add it to so we really could get rid of the requirement for +The libs can stay in the SVN but I did not evolve it to handle contrib and parent and build to same place as Maven builds or nothingI also noticed that the trunksrc is still in there so svn rm trunksrc once patched +With that change I can modify the existing unit tests to implement a run method where the prickly scenarios can be run against an external cluster as configured by Having being able to run from command line or as an integration test or a unit test gives us the most flexible approach +Touches the following filesM javadrdaorgapachederbydrdaM javadrdaorgapachederbylocdrdaM javatestingorgapachederbyTestingfunctionTeststestsderbynetCommitted at subversion revision +This issue was fixed by Mamtas fixes for embedded holdability for andThe following checkin adds the test to the client suite and removes netxaPositve which was a subset of xaTransmitting file dataCommitted revision +any possibility you also added a test case for it? +Hmm ok took a look at your code +I agree +Wont update JBDS CentralUsers who need this can simply uninstall the feature from JBDS then update their JBDS from the site itself. +Ive never actually failed to replicate something due to the timeout if I retry enough times but sealing of the Ref leak or catching unexpected messages seems like a Really Import Thing to do +tobi you broke the SPI contrib which makes intensive use of the commons projecti deliberately did not copy the commons classes over and i dont want to do so +Hi Bill thank you for notifying me about it +Attaching +PR coming once is out. +But actually the XML discrepany is just one symptom of the bigger problem +If someone could point me to an examplesample junit test that uses the HTTP protocol when communicating with the broker that will help me get aMy test program written in SAS just a bunch of messages using HTTP then them +Here is a is a patch +Works for me tested on master and closing as resolved +Sorry for the delay I was on your suggested line in allows Tomcat to start successfully but my transaction interceptor is not executed +Thanks Zhijie for the reviewIn the code we are throwing and if in the previous world somebody is depending on the that will breakSo probaly we need to do other way aroundFor other comment I updated the patchThanksMayank +contains the complete just commited the fix and it now builds for me. +Attachment has been added with description ZIP of test stuff fixed already and the rest is a dublicate of the +The regression test is configured to run with Xemopt optionx platforms dont support this mode and test crashesThe test should be added to exclude lists for x platforms +I think this is a really bad plan +Im surprised this issue is not of higher priority +Closing this issue because it is a JVM issue +CR has already been released. +I consider this to be closed since associated ISPN jira is fixed +Additionally the Override method appears to have been commented out you attach the patch to the JIRA +Thanks Paul! +And regarding the question from Lorna we need more information to diagnose the problem +I hate instanceof checks and the listener registration with the step factory bean was implemented precisely for this reason to prevent anyone having to put an instanceof check in a readerwriter +Mark any comments? Id like to get these forums up by Friday so we can link to them from the product pages well be pushing live +Ive added this so you can check thatt he topology is passed around correctly +Thanks for the good report lets fix this right awayDo you have a bundle deployment that works on Linux +i removed it as part of close Solr issues +ResolvedA filter which excludes platform is applied to the plugin +Ill set up again and have and doing tests only no snapshots or rat sound okIll be happy to turn off when you say it is ok to do so +tcksrcconf +new patch similar as to remove zookeeper integer version dependency +Using an enum here would prevent folks from accessing digest algos in the underlying library +In my opinion all are wrong +Triaged for repro attached and setting urgent urgency +Fixed for +If Mark is OK with these changes it could be published as a contrib +Username ruoso +They get their main jar syncd from the apache repository +Ning did you mean to click Resolve on this one +Philipp whats the latest with regards to the regression issues introduced? +I did not run the test class against it +Oh I see now +This is now a separate issue but Im thinking the local repository needs to be configured on the slave side and master selects an available local repository there +Im just saying I dont think our filters need options like preserve or inject because I see generally no value versus just using another field its typically just users who dont understand that the underlying cost in an inverted index is the same. +This issue also affects version dont know when this was fixed but the example given now works Grants tests are already committed and I think any stop word issues that remain should get a new ticket +latest patch addressed Rohinis comments in RB +If you want to remove data you explicitly need to delete itregards hint I dont know if you really want to do this but if you want to reset your data in the you can use the attribute type and set it to reset in your commandButtonregards for taking so long a final comment you need to determine what will be shown in your detail page by determining the state of your managed beans this is not a problem withYou need to take some action on the cancel button to clean out the detail data so that upon entering again into the detail page the detail data will be read again now with the correct idregardsMartin +when I move the mouse over the textarea the message displayed as expected but the disappear when I move the mouse again away from the textarea the appears again +The seem to work only when they are part of the primary key + select ftl the parameter table will be hidden and all param element under will be removed +in +confirmed +Assign access to the mapbuildertrunk directory to a group of trusted +Committed to trunk thanks Colin + Testcase for using namespaces in why not retaining namespace prefixes for all elements? This shouldnt be too much overhead +hiTamasthank you very muchyou are rightalso i think are not very efficientthe readed column ids are transfer to +Thanks +Since all Community projects are led by Red Hat developers and contributions are welcome from anyone at any time there is no need for a keywordflag +I never documented multiple docs per add because I hadnt figured out how I wanted to handle errors etc +The Annotations mode is the default one that we are most familiar with +I will commit this unless there are objections +Forgot to mention Im using rev have verified that this problem still exists on Drools Final +This will happen much quicker if you submit a patch for code and unit tests with your request +Luke any progress with this one? Should we leave it as targeting +If the formatting changes are too extensive let me know and Ill make a cleaner patch +Patch for the modified +No problem Martin as you can add independent test cases to srctests +Problem seems to have been resolved in the meantime +Ill take a look +We have found this case before and are sure it will only happens in IPC handler rather than core thread of regionserver thus mark it acceptableNice addendum +The patch is not a proper solution and is not thread safe however it proofs that the same JAXB context can be reused +T immediately do a removeObject to remove it from the cache managementAs a result T is not aware of it and any pojo operation T does will not be propogated through the cache because it has been detached +The reason you see that is that the Java client has am class that sits on top of the TCP transport sending periodic messages to the broker and also checking for receive inactivity +Ill keep this bug open until then +As far as the disallowing of the capacity resizing that could be a valid concern though it would make the simple code more finicky +Closed after release. +I forgot to make the methods public which Ill fix when it gets committed +For documentation it is better if you can provide the patch for the content since we are not familiar with cygwin at all +A bigger one will be provided with issue +It does not contain the librariesI tried it with the latest snapshot from th July How to reproduce +Yes is still open though we can separate from this +not deprecated +i found the task in jbpm +Sorry for the trouble +If I use Firefox or IE I dont get the message +If we have that we can tune is not a big problem +The new tests introduced in revision look good for me +Nice! Thanks Varun this looks greatOne thing that might be nice is a bit of testing that batches a few hundred or a thousand documents something that might catchIn any case this looks good Ill commit it soon +I dont like UGI being exposed further +You mean that we need some benchmark program to test the performance and scalability of message passing methods +Do you also need shell access to the Hudson build machines? Or web GUI only +Thanks SamuelCommitted +Disabling Authentication examples refer to HTTP GET and POST methods in +I have starting to work on this issue but I dont have write access to CVS +Applied with changes thanksIve opened for tomcat +Has it happened just once or it is reproducible? Does it also happen with +please open if this is still and issue but doesnt seem to be an issue with the am not working on this from a long time so could not test it. +Oops +I committed this +Im reopening this issue since the compatibility tests have not been able to complete in the tinderbox or in the nightly tests after the patch was checked in +Targeting revision . +I am currently working on deploying an app using JSF to AS +This sounds not really sane +Logger is initialized even before attributemanager can update the changed Logger attributes from title to really describe the reported problem which was you cant change the default configurationFile value from to another filename. +Finished patch +Unless all of these classes are made such patterns cannot be universally employedI suggest that you generate a list of static final classes and attach it to this issue for review +This feature was mostly done for +However when I included it in the ear file I get the familiar for the missing templates +Thanks + +Looks really nice! Just as wanted +This is patch file of generate from eclipse file encoding is +One of the scenarios was resolved as a user error +Further dev should follow on its own issue +I am currently not planning on doing a release +Enjoy! +ChangesRefactor some code in and Add test for failed test is unrelated +It was removed and not replaced with anythingRight that was the entire change of at the time that second call to init was causing file handle leaks if I remember right +Do I need to make some settings for it to websockets are able to receive the large content as property in message header why its giving issue with +bq +svnentries The file or directory is corrupted and unreadable +passes test suite with Dhadoopnonsecure does not pass checkstyle yet +Duplicate of +defer all issues to + is different from Entity + +weird +I checked in the test case jdbcapi to +Logged In YES useridIm attempting to test this and one thing I am encountering is that the Ajp connector is no longer part of the jbosstomcat bundle +Heres an initial patch that removes children array from nodes +javacore with the threads at the time of the hang +Sure I can add it but is that only used in the examples +Out of curiosity why dont you use custom Type mapping to expose the Joda instance rather than DateCalendar +As a result of the work performed in introducing useJava properties to and has been a lot easier than with the original. +So if you have RMI connections to other long running services these get the JCS timeouts +I think that will take care of your issue as well +It might have been making sure things run in the default heap or maybe it had something to do with permgen space + +from me +The patch mostly looks good +I assume this is critical for the build to run +Adds a second test to force split that has an empty family first and the full family second +Dont worry about Tiles +version that pushes into objects +Is the old name still supported +I have added a check to handle terminated is fixed in revision . +The autoconf directives ACPROGCC and ACPROGCXX set these variablesautomaticallyThis is whats happening when configure outputs checking if we are using GNU Cand checking if we are using GNU C +The ddls are stored in a dbscripts of the war file and a link was added to the configuration page to trigger the creation of the database tables and indexes +Attachment has been added with description files needed to reproduce the bug +jar +this one is looking pretty good but needs a few fixes +For example if outcome message is codified in must be charset +yes sorry I started my project more than year agoeditI dont find how I can delete or close this issue any help please +Comitters please reviewThanks +Id like to move this to +CP updated in revision . +Moving to TBD as this was not done for for the next release cycle. +Steve I have attached an updated project complete with a unit test +It is not possible to set Java Execution Environment for the server runtime +classpath should be added to the ignore list as well +foo files when it cleans +Weve done something for our build to support different in several combinations Ill make sure it applies on trunk and upload it +Thanks Oleg and Shameera for your patches IMO this is a very significant improvement +Ive asked for a new moderator multiple times with no response and am today directing all email from the Jena mailing lists to the trashPlease remove me as moderator +Koen I recalled you talking about this at eclipsecon +The underlying JBAS issue is now fixed so Ive removed the workaround code +WSDL file that triggers the described bug +I gather that the solution is to keep this simple and just add the new locale jars to the list without adding any caveats +studioeclipseplugins +and then start editingI could also entertain the notion of users being auto added as contributors or some other user Codehaus doesnt control the security of Confluence spaces we set them on creation and presume if you change them you own the to discuss alternatives +Confirming that clean package does the trick +Released. +I think all the version changes that can be applied have been + +Ill open another jira for publising bookkeeper maven artifacts for +This is not any different +Done ready for release +So this JIRA can be closedThank you for the clarification +Got it +Thanks + +Hmm Im using the very same feed you are with and seeing perfect resultsCan you attach a screen shot of the View Job page of the job in question? Also the View Connection page for both the output connection and the repository connection +I wonder about the exception +Then the bookies garbage collector thread will read his own ledgersdeletedBi node to get the deleted ledgers list finally bookie will delete the ledgers and remove ledger metadata in meta storage according to the list +We will no longer support interoperation with the +If there is a problem with this please contact me +Added test casesm sorry My mistake +is the right thing to do here I think all the tests pass when I did this manuallyAnd as Yonik says theres nothing similar in trunkSo if I remember the patterns from lex and yacc theres really nothing for me to do except close this JIRA right? Mike can freely check in the newly generated and I can remember to actually read the header that says things about this being a generated file that shouldnt be changed +It smells like a line that should have been removed +fixes the unhandled timeout exception as well as adding some cleanup code when a client is disconnected +Lastest group of changes +Please also consider the case where a project needs to deploy both Debug and Release artifacts where the Debug +This poses undue deployment constraints on production to blocker +In trunk we already have a new version of that tests the clientserver interactions for token authentication +As seen in the various stack dumps here crashes have been in a couple of different places +Should add the test to trunk as well +It would never start with anything else in the current implementation +Committed to branchx in r +I will look into the differences in configuration and see if I can determine what is causing the issue +Harsh can you please put up the patch here on jira and grant license to ASF for inclusion? Thanks +Reverted the BOF change in dffccddfdbbaf + seems client all classpath container is a large part of this +Please verify +Included in M is RSE support for regular jboss server types +Find the issues with and fix them I wonder if we can take any other approach but maintain message queues for failed deliveries in the presence of network partitions +If an equals method was implemented it would inevitably use the same values as are in the cache key which would result in the same behaviorDo you have a test case to show there is a problem +please provide a full stack apologies +Ok Peter says hes finally figured Maven out so I will see if I can update to a newer version +Will be in next code drop +in +Nice fix +Patch applied under r +all tests pass with it removed +The close will be successful because there are still valid replicas of the block in HDFS +The outputs are not emptyThis was a bug I fixed but I think I forgot to update some of the log filesWhen you run the tests can you update them +Can we make this a blocker +greclipse is still very rough but will be out in the fall +We must make that decision when we construct the object that actually opens up an fdThe question for me has always been this why cant a user decide that its a when theyre opening itIll poke around and see if anything has changed in the JDK we use but thats the short story +Committed to trunk and +To be completed closer to release + +MR job would take optionally table name so could per table rather than all +attached screen bat file throws no visible errors +Patch is included tests seem to pass +After that you decide to add some field into java class bean +I just need to generate +ExampleHistorique pour Initial Check parEtat ManquantRvision Vue PIM Rvision de la branche Chauveau Date CESTDonc cest pas gagn +Sure if we can get a patch for it +Actually the proper url for wars is webbundlewebappwarcould you retry with that +In embedded we have the actual string so thereis no stream so nothing to get closed after draining so no problem +emailed the PMC directly asking for commentshelp +Doh +for the patch +I just took a closer look at the resultBut do you know how hcommandLink should behave outside a form? just not render any javascript for submitting in onclick? mojarra prints a warning to the that the related hcommandLink is not embedded in a form +Patches applied +CP fixed in revision fixed in revision +They are at the same level +Good catch +The code below the assert did this so I just removed the assert +Dan suggested JMX controls +Ok Ive got it now +Verified with latest run +Resolving as invalid per previous comment and no further feedback +However I installed zookeeper and then reinstall Giraph again +ant test and confirm that all tests pass successfully +I think your ordering mechanism is also far more intuitive than my idea of order sensitive annotations +This issue needs more could you take a look I just noticied that the command was actually from the hypervisors are using currently +Note The includes changes to the unit tests which all pass now +The corresponding has also been moved. +eclipsewithivyde a directory +Please attach Eclipse Eclipse project and exact steps to reproduce this issue +refine notifyObservers try this patchThanks a lot +feel that we should check for the data part whether it was set correctly +maybe this needs to be a feature of the deploy plugin the ability to deploy a reproducible snapshot where the versions get locked in but its not the default +Fixed in revision I was able to solve the issue with the blank node and list shortcuts by making just those parts +But is the Path object actually in the right state +If ppl dont have problem with incorrect batched sync metrics I will take it out +Updated patch handles combiner exceptions as in the +After much deliberation and discussion Matz has agreed that this behavior of retry is not worth keeping in Ruby +Matt sorry my mistake +This is a job for the new MOP so John gets the give this issue a higher priority +Seeing partial files could be considered a feature +I have applied it to trunkI also did a bit of refactor to better handle lifecycle of the shared and have it shutdown when Camel is shutting downtrunk +It seems to me like these problems also arise due to overuse of wildcard scanning of the classpath +Please provide details when you open bug reports the requests the responses +Unit tests were added to show results of expressions +What should we be doing here? I feel this at minimum deserves a commentCheers Kelvin +I dont know how to deal with this situation in other Apache project like SolrHadoop or etc +this sounds like a good change but we should check into the velocityfreemarker aspects +That makes sense +Move creation of the new SYSCSINVALIDATESTOREDSTATEMENTS procedure from the upgrade logic to upgrade logicTouches the following filesM javaengineorgapachederbyimplsqlcatalogM javaengineorgapachederbyimplsqlcatalogM javatestingorgapachederbyTestingfunctionTeststestsupgradeTests +It affects all constraints of are disabled when the widget is in mode +Updated issue title to reflect the action item. +just committed this +In the above example tx is a got from JNDI +As a note I have also hit the wrong offsets mentioned by Sebastien using the current code +Jody +I do not think this should be done in a bug fix release +Projects only in solution files +Rat test passes with failures +Since the native libraries work regardless of JRE signature the patches checks for the IBM bit JRE signature and if found sets the JAVAPLATFORM environment variable to the same signature of the OracleSun than the reported version numbers isnt this identical to +Just a file an file to invoke logging and +Im not sure that will workLets say I have an automated process that adds a splits for the next day at midnight +updated patch mike conflicted me in the way +I test it with Xmxm Xmxm and always obtain an error in dvsl task +But you say autoReconnect is set to true so there must be something else wrongMaybe your JDBC driver is so old that it doesnt support the autoReconect flag? Maybe there is a type in your database connection string that is preventing autoReconnect from being recognizedTurning on debugging probably wont give you any additional information +The NPE can be avoided by testing accctx! null in retract and avoid calling remove and executing the if statement +Thinking about it it might be easier to enhance the bundle to optionally acquire a Repository via JNDI or RMI based on an environment property +No longer seeing this failure +Explanation at forum +that is just exotic and that its ok if the only way you figure a behavior is configurable is by digging in code +Support for Spring Security is still available as well in the original package +Fixed in subversion repository as of rThanks for the report and for the fix +Not sure why the patch application failed it works for me against trunk locally +After applying this patch the overhead of the is recuded from to +Shall I just remove the test case for the time being and commit the patchOleg +Bulk close for +Move to ll give it a try with if suitable +It would be better you narrow the issue down a bit so we can have a good start +The Roller feed is supposed to be RSS +If it has an XSD clearly the user wants validation if not clearly the user does not want to specify one +You can fix the trailing spaces as well when you commit +batch transition to closed remaining resolved bugs +Removed a comment and unused import +solved this. +The patch failed to apply cleanly afterCan you regenerate the patch +OK I think all of the migrated repos have been moved and the new ones are read only + test case relative to srcbugsxml +Fixed at r please verify +We can theoretically introduce a package for snappy but that will be somewhat orthogonal to Bigtops charter of packaging only Apache stuff +Whats odd here is that we are trying to transition the state with a version +David you were very close on this security did not have the filtering option set in the copy ant task +The extra check is better than before! I think its ready for commitThanks for the discussions +sounds reasonabledo you already have something at hand that you could contribute? that would be cool +How do we show the dependency of f when the selected time interval for health monitoring covers both the casesSolution For health monitoring the motif is to show the current state of health so we will always rely on the dependency fetched from objectDetailed instance history of the feedprocess is persisted in DB +Our url is tcp +Ondrej can you pass this on for testing +Increasing the memory in the +Attaching a new patch incorporating Devarajs comments +I just committed this +Ive just committed this +Thanks for pointing me to the right part of the standard Mamta +Sites verified +The two rows of buttons should look different because they use different CSS classes +Ok +FF into of a bulk update all resolutions changed to done please review history to original resolution type +this also introduced a bunch of javadoc warnings see if you dont have findbugs to exclude remove that from the pom file +running mvn assemblyassemblyEt voila tested it on a simple case and it worked ll do more intensive tests on monday wih Ant buildfiles that fail with Maven Ant Tasks +Watch out for white space changes in your patches check the patch manually before submitting. +Struts will absolutely not nor will it ever support HTML out of the box +Lock and Unlock CA +From my point of view looks good +You should build your components so they are resilient for doStop callsFor example when Camel is shutting down it will stop and shutdown all its services routes components and whatnot +BrettYou help is very much appreciated! Can I just ask you to update your email address in your profile so I could contact you privately if I ever needed toThanks again +Fix included in +The problem is CDI is optional so we are forced to deal with the case when CDI is not available or the case where we are not in a compliant server + in branch and resolve to both versions please +No further work expected and no problems reported. +If I install then I get the sun JDK and hadoop if I install then I get the openjdk one +upon the size of even the initial hintingcapabilities changes I will initially target the changes at +Working patch with unit test +You dont need to fully understand something to point out an error +Even if they can physically have sources and tests the lifecycle for POM packaging will not compile or execute them so what would they be used forDo you have a real world example of a project using such java sources in a POM project +It would be unacceptable when seeking having to deserialize the values which can be quite slow +However it still seems strange that we need to add escapeAmpfalse whenever multiple parameters are included within an surl tag +Attached config DB dump +taking care of javac we add in the initializeContainer doc exactly when this will be called +Ill put it in. +Committed this rd patch after verifying that all unit and integration tests pass +yeah +Sorry I dont know what you mean about the problem with the session +I am closing this for now with the fix I applied earlier +r for pulling loadtests out of their own hierarchy into the integration test hierarchy +Ill try to remember to include this in the release notes of the next versionSorry for the inconvenience +RamThat is a good suggestion But timeout monitor needs minutes +Added a patch and a the patch localizing wide opcode parsing +Do we want to try to fix this forThis could break existing mailets so Im not sure this worth it unless we touch the recipients collection for some other reason. +In any case those arethe results I get now +fixed in fffeacbabcdfbbbceeb on master +I have now posted the working patch +Could you also post some behind this? What would drive one to use different placement policies for different files in general +New readers will see the updated generation stamp and thus wont be able to read from the node with the shorter length right +Its probably going to be at least another few weeks before I can look at it +Default value is false +You just need to find an bundle of and its dependencies +HiYou really should post this kind question on CXF users mailling list first but not create a jira directlyFreemanMarking this invalid as its a request that needs to be on the users list to determine what the issue is +has been closed as fixed +The following links should give you the code you seekgit s ok for me +Nice! Resulting jar is still just MB +life in a world +This is fixed by the following revisions Fix cluster elder calculation to ensure unique elder +Please see attached mini project +I was able to reproduce the problem by running the unit test times +This is what I was afraid of were going to be chasing bad dirs forever +Assigned to Devin to review +Yes there should be comments on the closeSocket method +Also the fix with respect to start time seems fine +Thanks Dhaval +See have updated Springs to be compatible with Quartz s variant of the interface as well +I really dont see a need for a second plugin +It is not complete solution but it is usablehtdlogical is not used every argument is passed as string +Since the package actually gets included with the standalone I didnt need the extra copies of and the classes +Patch from applied to and M r and r +Ok thanks. +And in step we can see the clients side retry +Replacing get would have implications on how to correctly handle temporary holes created during splitting +WALVERSION could imply the compression type +It also already contains a method in for +When we update the versions of these projects what bug are we actually fixing for this bug release? Should not these version changes be associated with the bug that they are intended to fix +Thanks for bringing this to my attention Enis! +Committed at subversion revision +It can be applied to the version of the archetype would be great if more people test the patch and that someone will intergrate it into maven +OOM exceptions should never be caught since the JVM is in an unusable state once it is thrown so we are toast +I dont like previous versions poms +again this patch is for yahoo +Interestingly while there are extensive xml conformance test suites none seems to test for this case +Also take a look for two remaining problems with Generator Generator created from empty Enumarable just hangs during cant handle blocks and its yield method works latest version of the patch now includes changes in +displays targets implemented at this time +I can send you the logs if youd like +I grouped them because the code to implement them is all in the same my code creates an Entity that needs to be saved or updated in the database +Ill gladly review and test a patch if I cannot get to this before my own opinion on this has changed somewhat and Ive been meaning to circle back and resolve this as Wont Fix +I fixed this in Plexus +The policy should check that +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +anyone have any ideas +These files and are now included in the directory JBOSSHOMEtoolssystinet +no sure what this means besides being able to checkout the src from svn today +So I would be very happy if somebody who encounters that issue could determine the implementation so that we can either rule out that explanation or try to find a workaroundsolution +Is there a test script you could provide that reproduces this for you +resolved in M branch and was removed +I think the solution to this is documentation of the current behaviour and how to make it possible to include it in CSS +Hopefully someone else will be able to find a solution which does not expand the synchronization so much +This is making your patch noisier than necessary I believe +It doesnt work and trunk +Unfortunately I need to focus on other things today +I added the component type so that the failure information would be durable but if thats not as important I can keep the information +SamYour patch adds a new public method to a public class and deprecates an existing one +I was hesitant to second guess the original authors choice of picking an identical number of for what appeared a random size allocation as the yarn nodemanager localization port in those unit tests +If the index files are deleted those requests may fail +I did try to report this as well a long while back but go nowhere probably because there was nowhere to report it. +Usename is in the log entries right +Canceling patch +Thats the reason I created this patch +This problem is usually related to a missing Spring IDE dependency like AJDT or WTP +Verified with JBT +Would you suggest that this value default to false? This would agree somewhat with your comment that often one does not careI am for the second patch +If applied over trunk there may be some import conflicts in on the sync of files case if we ever did want to support I think wed add sync toIe it makes sense that this sync only works after the file is closed +Some comments Follow our codestyle I changed a case where a exception was thrown in one method and caught in another to just throw a The usage wasnt required and replaced with a simple expression +Im experiencing this problem to for regular REQUEST and SESSION scoped beans too +This is a that will help users to switch over to the new query parser +The account returned is correct but embedded user response in each account is not right should be +The second table has rows the last row has cells two of which span and columns respectively +Ive delivered two lectures on technical writing and covered all topics I planned to talk about +leaving in the old patches so the changes in implementation can be compared +Committed into branch revision +The new method should be public of course +I can try running nutch inside jdb tomorrow if I hit it what do you want to know +note that this is now in the by default so review that when creating the doco +Its worth looking at whether we need to guide the inliner further but we current only have two r performance runs +Also in this case vnet column in physical networks table is not updated properly +Closing as Cannot Reproduce as per the previous comments the attached test is missing files and out of date +Updated patch +This is remoting remoting HTTP client creates that following path fixes this have been created againstremoting that is in EAP branch +When applying it the bug should no longer reproduce with a clean database +I always wait until then otherwise its just asking for conflicts if you ever merge up the patch +During debugging a test I see the number of Cookies is doubled a lot of times +connectand execute commands deploy deploy the serverGo to AMQ web console and send a message to the server log file the message should be redelivered times seconds apart but that is not what me know if you need more help +Im a little torn as to what to do the simplest approach would be just to cut off the output when the error happens and let the client deal with malformed XML +Im not that into the grails source yet but it looks like is the place to change the artifacIds +yup we made it bigger trying to fix a problem with security panel not rendering properly on Fedora. +Somehow we need to know the exact diff so that we can apply a patchAnyway in this case i will your few changes +But since we dont know the number of parallel maps that can run setting replication to a fixed value like may not be enough for some cases and also can become overhead on namenode +For example C would interpret n as a newline character +jarExtracting systemorgapachekaraffeatures The system cannot find the path create CUsersAdministratorSourcefusecdctargetdistsystemorgapachekaraffeatures +Could be kind of interesting to have outputformatajaxmap +Some additional modules were found to be missing the dependency since the patch was first posted + +bulk defer of open issues to +Same story with Karaf on trunk was able to get the start level of a bundle using the symbolic name or the ID via JMX +Thanks for the patch because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +didnt want the full class name +attached is a patch where exception is thrown when the configured keytore andor truststore after configurable time +At the very least main and wfs build properlyTake it as a suggestion for an alternative solution if it does not work for you just go ahead with your nice your patch is much cleaner Andrea +batch transition resolvedfixed bugs to closedfixed +Ben think of the following scenario +I agree option seems to me like it will result in clearer code +Note the test still have a problem because of the wrong asssumption for the mapping in one of the to Max for why do we need to remove the renit callIsnt that there for a reason +Instead of a change request it is just a document tracking changes +doesnt contain any history +Uploading the correct patch +I think a LocalGMT button in the UI would be perfect great idea. +Here are the contribution jars used +Fixed in trunk +That wont quite solve this particular issue though because we also need to pass the feature params for the gadget +So we need another release or just update the dependency in CR + +for the patch pending Jenkins run +I vote for removing access modifiers as well +I especially like that its configured via the configuration file + +Completed and verified for BRMS and SOA +Noting that the checkstyle feature was resolved as too difficult for the value. +Just gets the code synced with trunk and rebuilds Hive +The time performance difference is unnoticeableIm providing a patch that allow xmlbeans to select the platforms default Sax Parser instead of relying in a another configuration option. +Patch is attached that removes Charsetencoding from Authorizations and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Its hard to envision why youd want the current behavior +Otherwise when unit test are the part of the project itself PDB file and associated DLLs are in fact project generated artifacts +This will be part of the upcoming release +Lets release FT version as a Hama +latest patch available for review only +Guys +Attachment has been added with description Output file created by my Fop table is missing because fo is notyet implementedA workaround is to remove the element + +Planning to look further into what we can do with this specialization for +I This is done to be the same with HQL editor do not load until user asks about it +Currently awaiting feedback from OWB +Thanks +Pavol what browser are you using? Ive tried IE chrome and firefox but I am not seeing this +At the level there should be a so that users can build the plugin for themselves +Need to create new plugins since the old one has the word beta on plugins for the this issue as fixed +Please let me know if you still see the problem in the next build though + +I shouldve mentioned this issue was discovered by Stephen Chu +The problem is that Enunciate uses Suns Annotation Processing Tool to read in the source files and get the javadocs +Im totally with that too +jar if we guarantee that the base classes dont have compile time access to lucene classesI also agree with this and we should bare it in mind as we progress +Will replace with the zk quorum info if that seems appropriateYeah that might be more appropriateGood stuff +Yeah I know about common config for java and php but Im going to use only php in my project and wanted to decrease of integration code Thanks for tip about containerfoo +All the other languages we are evaluating as candidates for THE applicatins scripting languagedo support preserving the state between what I tried so far using code similar to what Groovysh uses only imports failed to work +Mustve created too in a bought of sleep deprived typing then forgot about it +Attached is a proposed change to address bugs and together with a detailed document discussing the thinking behind the changes and explaining how they work +The fix had unforseen side effects +But perhaps the lesser of two evils +Removed the patch attachment because I am updating it and will then post a review request as noted above +Thanks Jacques +Sample test program +not every component has a as least up to now there is no requirement for that +These transient fields are not serialized +By the way the provides access to by the meantime we got it all working and we are now also aware of all the methods which provide the in one or the other way so from my point of view no further investigationadjustment is necessary +Im a bit surprised at the debug logs in the test cases +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I havent tested this changes on any other but I tried to do the changes multi platform aware +Isnt this an issue with CVSparselog using a joined stream and hence possibly parsing error messages as well as CVS output rather than CVScheckoutAlso if we split the streams well have to read from stderr as well as stdout otherwise its buffer could fill causing the subprocess to block on a write +Looks like the proxy requests are involved in this somehow a core is not being found when the leader sends some updates to a replica and so the request is proxied back to the leaderWhat is odd is why the leader would be sending updates to a node and not be able to find the solrcore +Bulk closing all resolved issue that have not gotten any more feedback or comment in the last month +What about support for? It doesnt look like that is currently supported +Thanks I applied the patch and it looks better now +Update the Fix version of this Jira as it will not be included in +And creating something like this for an extended form like sform wont be very applicable since a form can contain many components that are not displayvalueonlycapable +The bug was raised against so the fix version should include no +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +v fixes an NPE if you try to removetoken force before calling removetoken +If your HADOOPCLIENTOPTS already contained the JDWP options then they would be duplicated after this step executes +close issues for. +Seems like a lot of variation +I just committed this +Thank you for your quick responsesIf you could implement such a thing this would be a major step in the right direction +Closing issues for beta . +But Im marking this incomplete since we dont really have a good understanding of what is needed and dont want to add something based on our guesses. +See +Attaching patch which expands the scope to include all processes +patch looks good +But our handling of an empty qualifier is inconsistent we need to break someone to fix that +Whats missing is that Token Introspection IS a remote service full stop +Thank you Ravi +Not a problem anymore at least for me +If we have to rewrite code to migrate to then we shouldnt break any user apps that used the level we supplied in +Thanks for the nice work +now I need to check why the test works and my real world app crashesBut it definitly crashes currently +leads me to think its some wierdness in the underlying impl +a TODO item +Add useSSLssl attribute for cfpop and patch suggested will not work for me +Im terribly sorry about this +Thanks for the patch Roger! +Unknown release vehicle +For M this does not matter the plugin is loaded once and once only. + +A critical bug fix is one affecting a large amount of users +Can you provide a more description of how to reproduce this issue with a clean workspace and? If I can repro I should be able to figure out why its barfing on the NPE +I updated JBT to latest version and problem is not present anymore. + Patch +Should take abot hr in PROD +to commit +Now that work is well underway with AS all previous community releases are + +For example all the methods related to lockingmay be as well because they are independent of the catalog implementation but will always be s also the mess with ids and namespace prefixes that should be m now working outside of but when this will be finished and most of all if it will be of interestit should be added to and after that should be modified to use it but this decision is up to we need these functionalities so well have to go on with them anyway but wed like to see them integratedotherwise well end up using a forked version and that would be sad +It seems it does not work properly if the property does not exist in the original class +Thanks Chris +once we want to persist the dataI dont think wed need that for unit tests +We need to choose an algorithm to provide a balance between compression quality and speed +Fantastic news! Is there any way to setup the Apache repository now or should we continue to develop in and move it over later +Thanks Michi +The Javadoc says that transferTo can be more efficient because the OS can transfer bytes directly from the filesystem cache to the target channel without actually copying them +Rajesh can you this test with the GA bits +Progressable is about whether or no progress is being made +again +Nothing to worry once this is committedJust saw that changes the conditions under which the still has rows while returning false +Consider a sorted HTMLdocument with a sorted table +war file with the files to duplicate the bug +Note that this section in the article Configuring the Enterprise Service BusRefers explicitly to a server configuration we should add instructions for servers too. +I have no idea whether there is a way to download dynamically +Most of the xmbean descriptors were ok as they can easily be edited and did not need to be customized by the installer +We could add in a config to turn off the proxycontent rewriting all together +Thanks for now. +Works on todays version of built from master branch. +Making it as patch available for precommit tests +Is this patch accepted or rejected +I am running regression tests nowTouches the following filesM javaengineorgapachederbyimpljdbcM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapireferenceM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javaclientorgapachederbyclientam +Thanks for the pointer somehow overlooked +Patch looks good +Also tested formultiple nodes blacklisting and coming out of blacklisting +In the older days we preferred using HSQLDB +This doesnt actually require a release of camel to fixRather than have people wait for Camel this can just be patched in guiceyfruitBecause of the nature of mavens dependency system this will allow us to exclude the offending version of guiceyfruit and use the guice compatible version of itUnless the plan is to get rid of guiceyfruitI guess that makes sense as the project seems pretty dead +Option sounds ok as long as the declaration is at the portal metadata level so that each portal can have its own configuration for error page +I think Prakash has a little more work to do here for the case but this is an existing bug +cancelling patch hudson failing +Looks like no benefit on Mac OS X with the Apple anyway +Based on the attached proposal I am going to create of this jira +Hence NNs isnt always needed and shouldnt always be startedbq +linking issues discovered during this testing +Attached the report of running Benchmark on Hive together with the package that is necessary to reproduce the benchmark +Im going to revert the listUrl command back to its proper form and include your listRepository command +I had a conversation with Cos and learned that I completely misapprehended Herriots scope +Seems to me that they belong to one place +I will probably need some support from some of you core developers and will connect to the irc ASAP +I think you mean that would extend not implement it right? That seems fine to me since s public methods already all come from the Encoder interface and we dont lose any abstraction +Moving to will not be a moving to for m not sure if this is still an issue +It seems primary key index for snapshots table is very fragmented so a regular reconstruction may help to have better performance +First draftAll tests pass with a few caveats Replication tests that replicate config files are currently disabled and its not yet supported +I used Quantum GIS to make this shapefile but it seems invalid file format and geoserver couldnt insert new data into that +I already did the merge to +which indicates that jetty wants the connection closed + would be nice though I imagine a fair bit trickier than a unit test +It is targeting the old version of the groovy eclipse short term fix is to right click on the project and in the configure menu select Convert Legacy Groovy Project +I havent had a chance to go through all of the changes in Erics fork yet so Im not sure how the two attempts compare in terms of functional completeness +Yeah +I cant reproduce the issue and due to lack of further information Im closing this down. +Thank you for your patch! +I think this is valid +Same OS running Erlang RB and similar hardware E with GB RAM K SCSIRaid it may not be hanging just taking a really really long time +i initially thought this was coming from a namespace corruption bug and redeployed trying to reproduce the issue before narrowing it down to the edit logs +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Villu name retained as author +So what should we do about this issue? Close it? Keep it +Therefore it doesnt need to be of size maxDoc but rather close to how many docs were visitedMakes sense I guess for queries that hit a low percentage of the index +Committed +Fixed in revision Note it is not a good idea to attempt compaction on a database is near full capacity for its write load +Dont forget to help us with documenting it Vasily Thanks! +Fixed in r r +Note also issue which affects fault handling +Removing extented ACL control should solve the problemsetfacl b the tests were failing due to build env issues would that mean this patch is going into the tree sometime soon +Snjezana could you look at this? We need to get the runtimes configured for the servers installed for EAP and and in a way so the names arent colliding and not too long either +Geoff could you re attach the patch file granting license to ASF +Even after applying the patch from this JIRA the problem remains but in CR it worked with this patch was applied +was not expecting it and had to modify my addons but made sense +Point to localhost +This bug will serve as the marker that we need to propagate fixescleanup to do the right thing when calling Java +This is not being scheduled as it overlaps with existing EJB functionality +nothing obvious from the logs except that one problem is common in all test failures +Can you please submit a test case or at least some excerpts from your configuration? All our existing tests for such combinations seem to work fine +Fixed in rev to change issue type. +Committed revision now shall support binding and using a variable in the same pattern +The question then is how to handledefault args +Ill try to create a patch once I get the CVS connectivity straight +Solved by Gilles as of r + +So there is no exception neither unworking code that magically became working after applying the patch +Hmm works fine for me but Im on Leopard +Makes sense +Ok here is a test that reproduces it +Close all issues. +Talking with JD we figured might be a solution as well +it will be useful to user if we provide the error message instead of keep quite +Avro and Jersey were bringing in different incompatible versions and causing problems +works fine using a singleton with read lock mode instead of a replication project +Workaround provided +Thanks Ramesh +Looks like you have lots of tombstonesRead live cells and tombstoned Do you have columns on that key that have been inserted and deleted several times? It looks like you have of them +So the requirement for this task is pick at least two components whereby Google Maps and the multi select together count as one +Assigned back to David to make the edits +oops wrong browser window disregard previous comment +Users should not get any kind of nasty surprise when they write a Transformer and moreover it can break back compat +Add a page flow that demonstrates the verify +Fixed in patch for committed the fix on. +Updated patch +They should live on we probably should at least link to the mapbuilder stuff +If I simply changed the method parameter it would be incompatible but if I remove the method then any code that implements the method will still work +Actually this is or the column name and not the field name is it not? The attribute should then be versionColumn +I would only use it if there is NO other way at allsimon +Too closing of issues that are old and have been released or marked invalid. +The patch looks greatHeres a very slightly cosmetically updated patch which contains no functional difference vs +Also when a version bump is done then build goes crazy because there are no jars of that version in the +The same file generator can now be used +Im closing this as weve reinstated FLUSH on +Distributed caching is extremely difficult to get right +Can you reproduce on current trunk? I cant and I suspect this may have been due to +It seems something else fixed this therefore I am marking resolved +Perhaps we can configure Solr to disable requests to any URL in a given list of? Or did you mean outside Solr itself +Ive changed that import to an explicit declaration for which should hopefully solve this issue for +assumed fix since bug is fixed in WTP which we now are ok by using +Does not impact existing code outside scope of this filter patch +The total plugin count is down to from the that it was previously +So I dont see why the new implementation wouldnt have similar gains +This functionality has been incorporated into because of dependencies between to issue that contains this code now +Too big for scheduled for +Cant tell yet if its your fault or not +I mention this above but it is worth repeating +Fixing it would help people andimprove the code +Im going to play around with some other changes and test casesHave you tried sending a with a string that has ASCII values greater than Im betting that will fail to unmarshal correctly as well +Otherwise we need to fix the underlying problem +for me! Thanks +bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds bburruss bburruss Mar datadatauds +bulk defer to defer to +My personal opinion is that it should just depend on it normally but I would like to know if I missed something before submitting a patch for that +Please look at the +slave receives the failover command and performs failoverThe flush buffer implementation can be shared with the stop command implementation +make and package private fix comments in and dont have other comments +Will commit this shortly +Added in commit d on branch and already committed on master +The pull fixes the issue and it now works whether on the wizard page or on the property page +Ok now I committed it for real this time +we set it to wont fix and I close the PR +verified in CR +We can update the file once a job is done or serialize multiple versions so that only those unfinished jobs need to be rerun +We copy the coprocessor jar locally so it can be loaded by the JVM schedule it for deletion at exit and then ignore it +Need to implement this bounded mechanism for constraint entailment +Oh i havent tried a full build +Thanks Jarcec! +If you do not have any objections against this Id like to take your sample schema and use it for an additional CTF test to be added to the Castor test suite +describes the actual cause of this issue very nicely +This is in. +bq +py script is the eventual goal then rather than having divergent shell scripts that will have to be unified in the +I dont think is supported on Windows so this workaround wont affect any other work going on +Issue lies in base POI and is now Jeremy! +I have to the unit tests though +This fixes all issues Im aware of +First edit the file and change the properties at the top to your systems local properties +Perhaps the in the test classpath should be taken into account for test classes instead of in the main classpath +bq +Should get perhaps be marked Eh I dont see any reason it shouldnt be used elsewhere in the code either +Bulk closing all of these resolved issues. +Aminewhat you wrote is probably a prerequisite to get it working at all +Ill wait for a day to see if anyone has other use cases for me to consider then propose a design or start working on a patch on Monday +I was hoping that someone might respond to that question on the list +Please review my changes before applying the patch +Seems like a reasonable use case to me +Thanks for the information AndyIntegrated to and as wellThanks for the patch Matteo. +I thought that the xsd had been changed according to the previous comments in that issue +Seam tooling works in the current GA version of why not +Restoring the old constructor and marking in deprecated is probably a good idea +The new patch also fixes warning aggregation in to use the counter support now available in hadoop unit test failures are because hadoop QA process is not using the attached in this patch I ran tests locally on mymachine with the new jar and they all passed +I left as is since I think it is more consistent with how CATALINABASE is set in Please let me know if you have any other concerns +The wikipedia page tells you ways of doing it +When a given server needed to be restarted the load balancer changed the servers state to something like restarting and did not send any sessions to it +This is indeed cleaner +Ported fix to and branches +Could you also upload your settings file Cdevivy? +Im wondering if it would be better to tie this into the functionality in grunt so that it could aliasIf that isnt feasible Im fine with having an aliases keyword to show current aliases +Format each of the source file with the hama formatterNoteThis version of MLP is the basic version I will add more features if this one passes the reviewBTW I have checked out hama from Apache svn repository +This issue is fixed in fix version to sorry Ive been too busy to be involved here but could you briefly sum up what the change was? I banged my head on this one for quite awhile when it was first logged +Just tested with the latest the problem is no longer thereBut it is reproducable with the snapshot I picked up on it not reproducable +I also want to get sure that this feature will only be used on spool repositories +OK then I am a on the patch +Updated the patch slightly to add license headers to the EC scripts +whoops +I looked into this yesterday while testing candidate for inclusion in +just have short test show that cluster mode is broken deeply +How are you creating the BATCHRESULTINPUT class? Is there a JAXB customization file being passed in? If so you would need that at runtime into the as wellResolving as request for more information unanswered for a month. +Which one +will commit committed +This is actually a relatively harmless scenario in that the algorithm will still function itll just collide a lot +Looks like this one is done now that Lance has implemented env cloning +What happens if the condition is met at some but not all replicas? If we want to fail the insert in this case we need to be able to roll back the insert at some replicas +Bounce back to Rob after reviewed so he can commit stuff +I am doing in jboss application server +Ive fixed this with a simple cache on a method that gets called every time just before the launch config opens +OK I think I know what the problem is +Need to see if its still happening with current amq before spending any time on this +Specifics of the root cause and solution are on forum. +If you disagree with the resolution I propose that the SSL initialization logic is moved from to a special factory class that takes into account all system properties not just SSL related +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +Im talking about the parsing that happens in a chain + resolved issues for already released versions +There are only so many places you can look +it is not just a warning being printed it is an Error and it says invalid project example where as far as I can read from Snjezana it is a Project example Not found +Current patch is still not publishing hcatalog artifacts in hive namespace +Dupe of? Can you try the patch attached there and if it works can you close this one +So I think we should just include the common serdeser code in both the client and server for now? and have jars but with some common files +Hemanth if you submit another job that is also it may interfere with the first +Verified by Mark +Colin nice additions! Patch applied. +addClass isAccepted methodsthanks +just nto sure if we already have tests for binpig that I can lump this in with +The wizard window should be on the top of the dialog +Also why do we need a call to get jobs that are still running? I thought the call was made only on restartreinit? We should be able to clean old user logs without calling? I think we should avoid adding a dependency on calling jobtracker client methods in the tasktracker itself +The patch is fine to review though + +We have a test called in that test a bunch of writer thread keep flushing data and send the info about what is flushed to the reader threads which in turn query +Please provide more detail example as to what was done should be done instead +Committed to trunk and +I just committed this +I looked into this a bit unfortunatley it seems its going to require extensive changes +Ive never done it though and I kind of get a headache thinking about managing that +update as of these qss have tests writtentrunkproductsamplesquickstartstestssrcorgjbosssoaesbquickstart need to create new tests none exist for these qss far these are done as of Aug blocked by transform quickstarts are finished +From my understanding on most architectures the JVM will turn get into just one cpu instruction rather than using compare and swap +ve turned the messaged to debug message so it only appears when the X switch is on in maven +I dont feel that strongly about the case as its an unlikely where someone would care +Additional change to Content cuts down time needed in effective fetching +some helpful info could be found in the linked s also necessary to check whether it works as written to add screenshots to highlight meaningful words introductory sentences definitions to firstly mentioned elements of JMX perspective +The patch looks good +If the pattern only expands info such as username or the thread ID of the initial thread the pattern could be expanded at the time the appender is initializedThat way we have no problem with openingclosing log files and still can satisfy the need to have or log files +I just committed this to trunk and +Moving this to as some of the solutions discussed may involve API changes +Thank you Eric! +Incorporated Adrians styling and bug fixes from latest patch dialed back styling from my latest patchAdrianI see where you were going with the whole resizing header idea but IMHO I just dont think it looks good +this is a known PTP all issues to +This looks like the same issue as If so please link this issue and close it as a duplicateThanks this is a duplicate of fixed in trunk. +So I believe here we should exit when a timeout occurs +Closing +Any suggestion is your changes are in SVN rev +not being fixed in m if actually m problem +Yes we can add conditions to run specific tests based on the framework +Looks good +Great input +Do you have an alternate proposal? Do you see issues with the proposal I described above? Part of the reason for putting up a proposal was to trigger this discussion so your inputs are really appreciated +Sorry +Maybe we should do something explicit to make sure that classes enhanced with the old enhancer cant load anymore + +intellij didnt complain +You want to loose the intermediate page +verified in ER and ER +This patch includes a few bug fixes and an integration test +I thought it would be something easy to fix +I have put the latest snapshot of in our production environment and will replace that as soon as becomes availableThank you very much againGreetingsRoland +BTW why are we not doing this directly into? Its kind of duplicate no +Fixing findbug and release warnings +Per Teds request removed the redundant null check +Bug has been marked as a duplicate of this bug +Jar file containing patches +bq +This patch also fixes failing compaction test + +for the method +I hope I will be selectedRegarding your question now I havent decided to use what kinds of database because I know that there are different kinds of database and the of them are different +Looks great for me +Closing this issue to avoid any confusion of patches +I will make a xdoc the CLI need this wise logo to fix the align of the generated the image in the vsl to fix the most of the language this work until +Absolutely +patch mostly works +This was an extremely recent regression +And why is this an issue? We have a logj context loader listener that loads it from makes more sense to place it in classes because we need to add the file to the class path for the Eclipse launch configuration +There is a patch from them though so I could patch it and then release it onto maven central +Committed to the svn repository . +See file could be completely removed only after this dialog is remade +That would be great but me cant find sources if an archive isnt in the Maven classpath container +Martin are you using CVS code or a vanilla release? Looking at your samplefiles provided above it looks liek you are not using CVS code as youd berequired to rewrite your file to meet the new you are running against +Is there any chance that you will release a new binary soon? I dont mind building myself I just figured Id ask because it makes it easier on the other developers +Maven version is something I would expect see +Why dont we open an issue and track this down +Setting execution environment for such projects basically guarantee that and it has to be done +Thoughts +A restart was needed to activate an updated HMAC key as without that the notifications were just being dropped +one minor nit for future patches it becomes hard to review the patches if it has a lot of indentation changes +Sorry for the delay Ted +Do we have a plan how to create proper patch +We have a lot of users who operate disconnected clusters +r ldimaggi ldimaggi Mar +from my side +If feels too restrictive to me but I dont know +movedusermina and privatemina created +In other words instances are and can be shared in which case the getLog call will only happen since this is easy enough to fix Ive changed this into a static Log instance for Spring M and for pointing this outJuergen +Forgot to resolve +Updating patch +h file to +if you have run deployconnect command and successfully connect to a remote server then the credential information for that server will be kept by deploylogin command +What extra functionality will this provide +to replicate in JBDS Version Build id vN Build date . +The portion of the test has not changed +thanks VinayTrunk Committed revision +We are going to work around this bug in in revision +I havent yet looked at our anteatertest framework yet +Ill give this a try today +I think Ive changed my mind about having it on by default though it breaks some tools which cant deal with the UPA stuff +resolved in + removes five tests from the exclude Andrey the patch was applied to BTI branch at r +Tested with EAP ER no more occurence of this exception closing +the style attribute is sometimes now a symbol and not always a string +Can you attach the entire XML document? Given the fragment you posted we cannot see why that error should occur +Assigning to Arnaud +This causes problems where sessions are continually opened due to occurring even when there is only one active session at any time +I think we should commit it +Cool +Incorporated review comments + +No after several months and a request for a reproduction. +I will optimistically call this resovled if it turns out Im wrong +Calling this method causes the crash +Please keep operational questions on the user mailing list +Ive committed this. +In return the getMeta reply would contain a binary version of the type system as arrays of type codes feature codes and feature range types +Refcatored the so that the work on wrapping the is now done in a separate class which has been moved to the svn rev +The connection is not of the right type +I looked through the code and it all looks fine +Based upon feedback from Graeme marking as partially completed +I should be able to find what it was hanging on but a lot of logs to look back through +I didnt bother with the configuration since were just restoring what has always been there +I am simultaneously running went through the patch and it looks good to me +As discussed in this is disallowed but we can revisit to provide it in a usable form later +It should add them to the classpath when running tests if no other time +The issue is solved +I would prefer we this based on the successful IT and open a new issue if there is a variant use case that fails +Im getting started on this one now +calls parser with empty string and parser prints this line to the standard output and throws exception which is ignored +However its unclear to me how that can ever arise Those exception should always be thrown with a proper root cause by Springs method never causing that NPE in the first place +I would prefer too +Im not +Can you take another look and make sure we can get a clean m sorry Ive updated the patch and now it should +Eventually we might need to separate out some of the file format parts from the tika parser package +Right now sends globals that are required by it to categorize and organize data +Therefore if a solution cant be found the previous behavior should be restored until someone can actually come up with an approach that can gain consensus +The DTD is included in the ocm jar under +A few other details +Closing this as reported is providing details and the problem is the flatpack library and not Camel +Yes or use the zookeepers passed in on the command lineIt would be handy for an administrator to not have to go looking for and then remember which nodes are the zookeepers. +I agree that checking for nullability doesnt need doing in this JIRA +to commit +Sure thanks for the clarification Ryan +Heres the patch +Deferring all remaining issues so I can release in JIRA +extends instead of extends Just need to fill in the test and it should be good to go +I have cleared your initial attempts from ortpasswd +Then as the projects are processed their files will be modified so that there are no variables for the artifactId groupId and version and they will be placed into their respective target directories +Justin this one is a matter of discussion as well +That was a good catch with the floats +Ive just committed this to the HA branch. +Thanks for submitting the patchI have quickly checked it MD checksum matches source files have license headersnotes pom files are missing the license headers the runner folder contains a jar file and some batch fileswe can probably fix the pom license header issue on import and decide upon handling the runner folder while importing into forgot the license in the pom +forge folder so that the newest version of the plugin is installed +Hi WernerThe behaviour described in your last comment is what I also would expect +Kevin you need to the patch making sure to tick the grant license to Apache box on the upload dialog +It would be rendered to a mail and be sent +The method is to avoid passing them thru to policy just to use in request creation +Maybe polish it a bit +smallest be empty in this caseProbably not assuming that there is some set of files that can satisfy the other conditions on a compaction +mod for +Doesnt seem to be anyone taking this up so marking it as +Day end will be run multiple times a day each for a specific market +In my opinion We are building an open platform so all the relevant features should be exposed and let the user decide whether they want those to use or not based on there use cases. +Keep em comingtrunk +further bug fixe and improvements can go in other m curious whether the change in the default value for the parameter from in to in is an intentional change or simply a bug that needs to be fixed +Looks like this may be a mojarra bug +It seems the first few tests dont shutdown the java processes and then the remaining tests fail because of that on these platforms +I figure we can move it to zk later if we want +For now there will be no option for specifying the item label via an additional command option for several reasons +Resolving as incomplete description for now +Well +Hi ChristianI just went through your patch and did some changes on your patch before I committed the patch +Any other thoughts from other committers before I commit this +The problem is that if a broker is down returns an empty replica list +Fix applied to trunk and branch. +They still have the +I believe it takes the default last used if nothing else specified in the archetypein any case nothing serious enough to block beta since users can trivially fix this if they dont like same package name in two different projects +I fixed this in d so would you give it a try again and report whether test cases passed or not +Mladen confirmed that and are backward compatible so it shouldnt be a problem upgrading +for the reroll +Is this issue fixed +bq +The ability to fetch the would be an alternative d prefer a lot to be able to get some sort of list of the errors that occurred while starting +The proposal is as follows +Thanks StefanI will update the file you pointed outI agree on the Powered By +Hey namitI did not realize that you have not given a on this yet +Autocorrect and me being sleepyWhat I meant is it seems like there exist cases where log replay time and Hlog size on disk didnt correlate well before and there will still be some now has it really ever been much of an issue +Once we get the database connection pooling code into trunk could we take advantage of that to avoid having the temporary files? Currently we need the temp files because streaming binaries directly from the database would block the single connection we use +JAXB was doing its job just fine +Yes the database has views and it happens always +Gary in general opinions from other team +If anyone sees it feel free to reopen the issue. +I agree with Adrian that this must be a per product setup I am not sure if Product is the right entity I think that different stores can have different maximum quantities for the same productI also think that some rules are missing is a customer allowed to place two orders for the same product? this would be an easy way to get more products +Leaving any other commitsuntil after the release. +web service and ESB queue +Thanks for the update Gary +maybe because the error occurs in a separate thread there is no reference to CGI variables? they exist in ACFs not there +Paul I have created for the optimizations to getpartitionsps +Svn diff couldnt show me moving the png files to new images directory insrcmainwebappadminthe complete fix should do thispushd srcmainwebappadminsvn mkdir imagessvn mv +If you want to I can provide the fix as well +push to +see latest graph +where is the documentation +Thats my rationale at least +Well actually I think JIRA does allow this. +Its not a problem but its preferable to have patchs apply with p +This is a minor issue because the relevant text is bolded but it is strange to have the name appear at the end of the sample without explanation +Dealing with console evaluations seems like another obvious place that could have hits here +What kind of carbage on hypervisor +Committed at Apache and backported to Fuse Camel branch +Ill go for holidays. +I was blocked on the patch before since we were reviewing a partial patch in time for +that could cause an endless loop when moving the mouse over expressions near the end of the file when a was missing +their input format is using mapreduce +Id appreciate if someone can review this so it can be committed +Moved the submission to the thread pool outside the synchronized block Removed the object poolingThere is also an experimental version that divides the queue into subqueuesbased on a internal the issue now revolves around finding the most performant +Hi CraigDo you plane to include this fix in your next release of +I agree this is a serious problem for any production use of nutch a blocker since you end up refetching again and again the same pages +While looking at this patch I noticed the need for +I hope I havent forgotten or something or that recent changes in Ivy havent introduced any problem feel free to open a new issue if you run into a problemAnd many thanks to Jing for your contribution to this task. +Reopen the issue with more inputs if this not the case. +I also tested this patch with a job that hit the OOM regularly during the shuffle and it completes successfully with this patch +Jeremy Good thoughts about testI am fine with the other proposed solution +Ive committed and deleted and +Laura is documenting the behavior of this jira entry as part of +Will this patch be added in soon or is it available somewhere now +For for apply Patches reviewed +I think we shoulda make sure ant is mentioned in the docs as a requirement for using add a similar warning as seams build file has to the s build a I dont want to confuse b it is not possible because is called by two ways in seam dist like seam COMMAND and the COMMAND is created as call other ant file without inherited is tech preview so I removed unsupported attribute in available ant the Release Notes flag since this affects and was fixed in +Given the age of this issue and the absence of attachments and the absence of a description of the problem other than the presence of a stack dump Im closing this as incomplete +It seems like this issue would not be appropriate for backporting because of the behavior change +Indeed it is tied to +Committed to trunk +Im marking as relates rather than duplicates it because that JIRA notes how performance of the servlet collapses if the file is big almost a minute to get that trailing few lines +Is this a duplicate of +If an alias is dereferenced during the search the search isnt continued under that alias +coresamplesjavatargetweblogicxinstalledtestcontainersecurity +If for any reason this site is unavailable drlvm build will fail +Today is a bad day +Thanks Xuan. +Ill have a play with what you suggest +K would be good but M or G might not be as good +bulk move of all unresolved issues from to +Jun +So all issues not directly assigned to an AS release are being closed +on this bugWorkaround of lazyfalse works for me on Railo Seems to happen with all my relationships unless I use the workaround +Jboss Maven integration facet is the one that should fix it isnt it? +i think too the flush should be done before the redirect cause i cant handle it +outSending javatestingorgapachederbyTestingfunctionTeststestslangsynTransmitting file dataCommitted revision +When I add a concept previously created as free concept its missing Preflabel and the uri field shows the preflabelI cant reproduce this behavior or I do not understand your explanations tight +Can you revert the parameters for the test back and set the timeout scan to ms and see if it passes? Thanks +Set to resolved. +I dont understand how the gets the resolver and resolverCtx at execution timeAre they serialized together with theIf so we need to mark those classes as serializable and move them to rightOtherwise it kind of breaks the implicit contract that everything that needs to be serialized is in the Work instead of the Task +I did not perform a complete minimization but I dont think youll have you much trouble using this to debugThe test that failes isIn at line the working code is commented out and the failing code is commented in +postponing to the +I think its okay to committ this patch +I just committed this. +pom file +It is in the trunk I have a more permanent fix up on the JIRA +No unit tests either yet +Isnt this a dupe of? I dont see how the bin scripts hit this since they dont appear to use relative paths +correctly handles new line characters now should fix the ant problem here. +This would be a major change +Errai minicard for +LGTM. +Great Peter +So I think what is needed is to update the encryptionProvider and encryptionAlgorithm topics to mention that they work with either bootPassword or encryptionKey and to provide an example for each that uses encryptionKeyDo you have some tests that use a encryptionAlgorithm? DESCBC is the default so there is really no need to specify it +I see the filter doesnt return null but the page is left blank using the +Redefining title back to reflect the original intent of this issue +Committed to branch +Copy or to with package substitution to use as thread pool +I think it makes sense as a default skin for widgets +That looks like a browser with Toms comments incorporated +In my opinion the client SDK has more purposes than just to facilitate HA +The reason for getting the connection is to find out the default schema default catalog and dialect +shp is only Kb but the index is MB +Added details http headers so someone will be able to reproduce it +Well +The config builder does have write access to whatever its buildingI havent looked at the code today +I would either make no check at all let the backend fail or eventually load all apps from the backend to check +Since you hadnt mentioned that sounded like you had actually built something that did AMF deserializationGiven that your implementation is tied to perhaps supportblazeds or supportflexblazeds would be a better module than straight supportflex given that there is a specific technology +patch looks good to me we have already other conf property name hardcoded in the Admin should we start to move them somewhere like or similardo you want also to include the rename of the rollbackSnapshot or is abandoned as idea +Hmmm Stefano I think you have attached the wrong txt file +Hi MirekThe Exception can be ignored +Yes this is close to what I need +patch applied +Maybe we will fix this by removing trafficshell in. +It may be related to more complicated JS libraries that themselves load additional JS files +OK I think I see what youre saying the main classes used for standalone deployment arent activated when you deploy as an bundle so indeed the three configuration files are being used for deployment +Taking a second look with checkindex and all of the docs terms are piled into the first couple segments the rest are a long tail with few terms +I am using oracle g +Because RFC is the only internet standard related to CSV? Excel is not a well defined format the delimiter is locale specific +ditto craps out on FF and Leopard +Below is the correct one +Benoit +I think I have identified source of my observed error +Reopened to set the Fix Versions field to JDO beta. +I just committed this to trunk and +Defer all Harmony issues to +Set the key in generate and add the key value to the payload +Which is it +Thank youBest regardsSpark by Spark. +It breaks compatibility but I dont know that we have many folks yet using negativeI dont think that anyone can be relying on them since they didnt work until yesterday +I thought it best to at least report the problem instead of do nothing +META +rev fix a bug where every proxy to was registering a new thus preventing our shiny new framework from actually creating any managed beans for. +Thanks for factoring this out Tommaso +This version fills in the Documentation section +This patch updates the news section of the front page to include the announcement of +Was fixed in +batch transition to closed remaining resolved bugs +Resolving as duplicateWill close when has been closed. +v +That suggests to me that isnt smart enough to find your IBM +Same error without my custom +One thing to review is the calculating of offsets + minutes after the end of Apollo is still dead +One suggestion I had was to these offenders +Whats wrong with using the one from the EAP distributionworkmodulesorgjbosswsmain I see that the jar is already in the maven repo zip and eap zip +Addendum looks good +Verified +Resolved issues being marked closed in conjunction with the release of. +I have attached patch and committed changes to trunk for further testing +Tests and fixes for the bug +again +Reopening to set release note note not required for new feature of a new issue now release note status has been set +So it looks like that now Im able to reproduce the first problem +Are you sure you want to use LUCENECURRENT in some ctorsThe lucene core subclasses used by some analyzers require this so another alternative is to create a static CONTRIBANALYZERSVERSION for this purpose and bump it every release +I would suggest removing JCC from the testing as it has not been supported by IBM with Derby since +That doesnt make much sense to me +Task is of a bulk update all resolutions changed to done please review history to original resolution type +Please open another Jira describing what you would like added in more detail +Using the class +Gorgeous +However Im not a maven expert +Two more points which are tangentially related The patch is not attached to this ticket and it looks like Phabricator stopped automatically attaching patches some time ago +But I think is just fine for this issue +Im nidaley on Jira and nigel at ASF username is mattmann and my JIRA username is chrismattmann +Attached patch for your reviewI agree with your idea the additional Context not bring any more information and flatten such objects are easier for us to add more information +teiidexamplesclient +Patch looks good to me applied in SVN +I will get back to you guys Monday morningMy guess is that theSpeculative NONE? is hit by the initial request reading path which is successfully resolved by the Speculative Retry +I was wondering how serious is the issue that empty groups are cached? The groups will be recovered after cache expires +Ping +Do you have some insight about my problem with http session +But it shouldnt be too difficult to work around this one issue. +Issue has been resolved for a long time waiting for potential confirmation from the reporter +Please review +The only problem I saw now is the scope especially in the builder the scanning scopes depend on some other scanning result +I have had the same idea about a much better sample and the one from the integration bible must be the best example +I fully agree that test code should be as lean and as reusable as possible but lets be honest here +I installed for All Users on Port as a Service Recommended +Manual unit test pass. +Thanks Xuan. +The login form is provided in the bundle as an HTML template +Mayuresh attached a patch for this issue to +I fixed this as part of +May be related to previously fixed bug + +I will move my test cases to systest having good coverage here we could experiment with invoker as well +The question I wouldhave to ask is where do you expect the Parser to find your document +It is not a straight revert of since in the intervening time since this was committed the codebase has been split into three projects and has been introduced which uses some of the same underlying codeThe HDFS part of this patch is in +Sorry that should have been +It will be a different development branch +Now AFAIK the Forge runtime also contains an embedded Maven +This disk cache is like any other +Thank you for informing us about the run with. +Adds a fixture to test that prepared statement is invalidated when atable its depends on undergoes an ALTER TABLE ALTER CONSTRAINTstatement +There are too many possible combinations to define refferences between component tag and renderer so I going to limit them to next cases explicitly set renderer type for component by annotation or element using type or template attributes +We can have a kind of modular approach here module will translate plain set of filter queries into the BJQ tree expressed in crazy expert syntax +Mass transition all resolved issue that did not see any further comment in the last month to closed status +libPIGHOMEbin +I can take an update deploy live and give you back what is going on +At the end of hrs initialmap tasks were finished out of tasks +How ever just this check may not be sufficient additional check is necessary to see if the of the Node is same as the sessionId +The first version of the plugin is ready for release +Cleaned this up while attacking another issue +Patch with fix Spring version range on the plexus bridge +bq + +The codegen plugin does not depend on any version of xerces +Latest patch includes a just committed this +patch submitted +I think that while polishing my code you removed an else return from around line +Reporters of these issues please verify the fixes and help close these issues +Therefore Id recommend you pick up +Here is an eclipse patch +Most of the changes made it through in the last commit +Updated to trunk +But committed something that should at least fix this blatant problem. +Removed first patch file I attached here because it is harmful. +This is part of SDLC +. +Closing this record since is now released. +Not related to this still an issue +proguard is GPL we cant distribute +Ill email the group and see what the timelines look like +I think we dont need the new conf key DFSNAMENODEAVOIDSTALEDATANODEFORRECOVERY since the stale datanodes should be avoided in any caseI also think that we could simplify the patch currently chooses the first live datanode as primary +Very well Carlo +Considering how repetitive this issue is you may want to consider because there does seem to be a refactoring opportunity here +Ron any chance of getting this fix in eap cp if its such a simple fix +Please raise an improvement request for secuity JSF check for JSF and GWT in current module to prevent security setup command from being available +Verified by Sean at rThanks. +To get started with it quickly you may want to download with Java EE support which includes +We should be able to take the latest jar and drop it into GF modules me know when the patched version of Weld is ready +Ajith is this require for ve generated code for the service implemented the skeleton and seems to work for me! Please have a go with the latest code and let us know +especially since this is a trivial fix to verifyfix and having this dependency even on Indigo seems wrong + +here the element you mentioned in the second schema and the generated code is correct. +I have fixed the issues you have addressed +I think Im more comfortable putting this in myself +that this patch does not change any checksum guarantees provided by current protocol +I cant accept your Job Manager patch +Its pretty nifty but there were some issues Id like to see resolved The Copy button is confusing +current for response it return false I think for response it should return false too +can remain resolved due to the patch +If the jar only contains packages it has to be in that group same for +Hmm I dont seem to be able to get this test to fail for me any more +Make sure when you test remove from the default project template +As already c ommented I could not reproduce and need to try this with oracle. + Is resolved +HiPlease note that I can reproduce this bug with a grails what Ive understood this issue comes from the typed service injection wich doesnt Ive replaced all authenticateServicebydef authenticateServicein the following files of my project + doesnt have any childrenIts not a bug its just that those valuesources are not scorers +Am I missing something +No releases older that have been added. +Problem is in the iteration containing find it runs x slow on my machine approx secs for RC and secs for investigating for me the difference is not so big for your simplified test case +The patch looks good +patch committed +It would just be nice to be able to remove that when the next ESB version comes you google the issue a bit you will see that this is a very common problem not only for ESB users but for anyapplication that connects via SSH +Adding the classes and change looks good + is the defining case but this is better handled by futher up the chain as a validation gives the necessary to perform this theres no so we cant yet check until thats check for local check was in place Unit Tests passing +For example create does not check if append feature is supported +one minor nit can you add some javadocdocumentation to +Attached nd patch file for changing the test case in as discussed and decided by Charles. +I also had this experience with the new drop downs but Im wondering how much its related to the thing that we were use to the old UI and because of that we forget to set those fields +The calendar worked fine in FirefoxToday I downloaded and can confirm that this problem no longer seems to exist for me +I also dont think the limitation comes from +This is only reproducible in the hight load scenario +Jochen the nd patch also wont do +please test with latest snapshot if still broken please attach a quickstart +Ensure Hudson +I am a bit nervousWe should discuss it a little bit and commit it +If its acceptable to tolerate exceptions in some indexing filters such exceptions should be caught there +I have made changes according to your suggestions +Break it in subtasks. +Additional patch features include Custom SFTP server description string Optional open file handle cap +Here is the Weblogic error message received to the JMS Spring config file used +Waiting for Nithya to test before adding to on Tuscany and version +Event Listener implementation is ugly and still needs work to reflect on new APIPlease review the new API changes and the new design and the changes made in unit tests +that file was encrypted with empty password. +Please try latest SVN +Heres a patch which removes keep and any references to to grant license to all committed in rev +Hi HenryThe difficulty in applying this change is that I am not confident that it will not reintroduce the Spring errors we observed earlier +shazron Really appreciate your work +I believe most of the issues reported have since been resolved and besides this issue is now a bit too old and messy for it to be of use in any future discussions. +passed + +In all cases the slaves response to the fetchindex command is OK and any failure must be determined through examining logsI added logging calls to the top of do in to display the contents of the instance it is given +java +it works +It is version from SVN +Grails uses the collection name to define the column names +Heres a patch for thisNote that this has the of masking in that interop tests will now pass again with this patch +i will take a look at it +In the attached image its the th icon in the route +This particular change available only in versionsAnyways marking this JIRA as dupe with I did not know this issue was already filed +It could be as simple as a field value although the ability to do more complex queries would be nice also +Will fix +Any idea here? There are and that we can revisitresolve I think +Do you mind testing the code on trunk to see if that works on your system +As some of them are not happening in trunk we may directly fix them in +If a task hangs after completing its output but before exiting then the tasktracker should still kill it but the jobtracker should not reschedule it right +Persuading the launchpad to create slingconfig would be better although I havent been able to do that yet. +From what I have read on the web it looks like a classloader issue when you have a different trasnformer impl +I dont need the workaround right now but I would need a proper fix in order to upgrade to +looks good +pushed +However if test scope is to ensure correct behavior of business logic it is not convenient to add Test Enrichment to an Arquillian test +if threads create components that need to have different factories injected one factory referencing the other +HiWould you please try my patchThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +k sounds good +We have the same situation where we need to upgrade exisiting databases and adddrop some columns +This patch changes to using relative time everywhere that was previously used +Removed some unused imports +small updateIt seems that internally forge figured out which command I meant but just doesnt display it +Bulk close for +Bug has been marked as a duplicate of this bug +While testing my change to make advertiser a plugin I had to define the configuration with packages as the is under the test source tree under the core package hierarchyThis wasnt sufficient to have the configurations pluginmanager resolve the plugin as the plugintype cache had previously been built for core +I am fine with your comments +Great addition +committed fix at revision committed to trunk days ago +Fixed by bug have put the bug into the Resolved not closed state. +Okay we got a bit of starter documentation that can get end users on the right track +JON JOPR doesnt support this notion and this will only lead to confusion in explaining differences +ENTERED TOO EARLY BY ACCIDENT Hi TomOK this looks good but I have then created a new wish issue Proposal for a new help system and moved your comment there please do the same for the imagesAfter before doing anything else let know the community about your proposition +Using now +For doing step do we need to use the information from and add it to Meta +Close after release +Which popups should be singletons? The slideout and modal? I believe a wrote the dialog as a singleton because the bootstrap dialogs seem awkward with multiple instances they just overlay +Could you check please? Jacquesdid you mean that after save the xml file from Label Manager you have seen some characters not encoded correctly inCan you give an example to understand the issueAnyway it is still not completed because the apache license header it is not savedThanks tested without the labels very useful +Please attach stack of a bulk update all resolutions changed to done please review history to original resolution type +Id actually like to rewrite bulk of unit tests so consistent redundancies have been removed and they are pertinent to current hbase +Really say +But given the new opportunity for spam I wonder whether its really worth it +Hi Willemthanks for the info +New default implemented for trunk and pushed to Nexus +Looked into this +Client file to be posted over http +To be checked before the RC +Set safe to true in the only when is true as well because unregister only happens when is true +fix adds a backword incompatibility so should not be backported. +to go to trunk +Pardon me for any inconvenience caused. +required +If its bound to zero interfaces then theres not much we can do about itAgain we can more easily assist with this support question on a support channel rather than our bug tracker. +Mark I think youre missing the point +The JSON Object had an error +bq +This is a fairly harmless workaround but Im now wondering if others might be seeing this +Minor changes to to reflect the changes in and +The current framework is still just first generation and really needs go get a +Good idea +Thank you for accepting my revision +I promise to be careful since I now can cause rather a lot of damage. +Ill fix this after after release. +symbol exports to allow C programs to avoid double memory copies and process application headers +markus +I have committed this +then we dont need to up the global versionTrue +Formatted documentation for the XML sample at the beginning of the document +This exists for tracking through the SOA platform +I tend to agree in the sense that if there is no strong reasons we should keep the groupid BUT there are two rather good reasons will definitely impact the dependency declarations of existing projects we should be consistent with the other project which already moved away from the group spaceI think we should do it ASAP as until were in Alpha stages its unlikely that anybody is using it other than the hardcore people reading our mailing lists and blog posts on a daily basisIt is getting more critical now as I dont think we will be able to finish before the beta milestone +I can confirm that something is wrong with request +add a patch for the class +Here is the patch relative to trunkwicketwicket that changes Radio and +For junit tests enabling the prefer to true +Changes across quite a few classes in the Pos area +The start level problem is related to The small bundles are broken for now as there has been a problem during the release process and one package is not exportedThe jaasjasypt module needs to be fixed when the blueprint utils package will be exported again +But I checked the management server log which seems not complete +Specifically I had forgotten that this regression test only shows a concrete symptom in the JCC configuration not in the configurationCommitted the change to the branch as revision viewcvsrevview has been fixed tested and ported back to so Im now marking it as closed +Okay lets commit this with appropriate comments +Darren merged support for recursive types and I merged support for the forward declarationsEdell +Thanks! +I change for client and embedded to throw instead of for exceptions +btw +I thought these pages were simply editing preferences +Attached please also find some changes to which improve resiliency especially if there happen to be multiple concurrent operations on the same table +Changed name of the to for readability and also added a POM section to ensure that the plugin is at minimum level +Default is which applies no validation +shows how we can easily deal with hadoop dependencies +Its better we cleanup using which cleans up any remaining +The latter one which is the maven convention is reasonably well understood and widely used +so I am going to say that mxj probably depends on mxjmxj +has been marked as a duplicate of this bug +the problem here is that the browser will act on +Mea culpa +added juergen to watch variable is now persistenceContext for both hibernate and JPA +Test passes with the attached file +Marking closed. +Logged In YES useridFixed. +It appears that if is resolved this problem may go away +Ithink this is a small sync problem because of the manual mergebetween the two writers +I tried this example and a related bug reported on the mailing list and cant replicate the bugThis might be related to bad artifact deployment for which we need a workaround or a genuine bug in Buildr +Hudson has the same problem when authentication is done aginst AD and it took some days for a collegue and me to figure out why the login was successful but the user had no rights +shared memory issues? try to change the loglevel to debug and check the errorlog file +If the checkstyle artifacts exist in the Ivy cache it operates in offline mode by default +And is fine for both the trunk and stable branch since it is is simply a null check and no change of any major behaviour +Please check your mailDevin Devin Can you please send me a private so I can verify that this is indeed fixed? Thanks +Part of a bulk update all resolutions changed to done please review history to original resolution type +If works out well have an automated schema deployer and database upgrader +comments +See this issue is related +I modified your test Cedric to test for private directly instead of excluding others +Setting fix priority fp assigning to Joaane T +Probably this issue can be closed. +Leif you use Fedora +No name given to the +But the overall total memory usage of the namenode will not reduce by +Thanks Adrian I guess that makes sense +We need to fill these out +Yeah confirmed +Closed upon release of Hadoop. +Baar thank you for your bug reportThis particular issue was already addressed in and you can verify the fix using the x or trunk branchesIf you find anymore locale related problems however please dont hesitate to report close for release +Maybe the Jackrabbit versions are important +The most we can do is provide more informative text about how the admin console is running on the domain controller. +Im sorry to jump in late I was reviewing this patch with the intention of committing it and noticed some related changes +One other nit Apache Licenses missing +OK try now +Are the service implementation objects inherently unpoolable even if you implement the interface? If you REALLY need this sort of functionality you can implement it yourself without having to create some special service model +The revised patch looks great +Please commit it +My bad Ill fix it in the next patch for now this should workmvn clean install PWDsrctest Dtest will need in srctestdata to run unit patch +The Properties javadoc says of loadthe inputoutput stream is encoded in ISO character encoding +Heres an SVN patch that will generate unique but consistent +lisStatus from each map etc +Seems related to +As to the minimum versions for heres the list of plugins installed from the maintenance build from Jan find +Unzip this into the sanboxwebwork directory +I just committed this +It looks good I just wonder why the retired flag overrides the status filter wouldnt make sense to still respect the status filter and if you want all jobs just use an all status filter +has been marked as a duplicate of this bug +Patch integrated to trunk. +doesnt help make things easier in this area there is not a custom action for installing help into +I disabled console logging and started to log to file only and it seems to disappear +Sorry Stack +Its OK if the feature has a different implementation but Im not interested in reviewing a patch against before the issue is addressed in trunk +Filed about the issue +The former would provide the user data while the latter would not +zhihong its strange that the tests that fail without patch are not a subset of the tests that fail with htrace +Great stuff Minoru! +Seems to work fine now after building from the latest appears to be resolved. +Uploading a patch which replaces the thread local storage of the referenced in favor of storing it in the lcc +Hi Sagar this looks to me to be most likely a configuration error in which case you should try emailing a user mailing list +We can close the issue +I didt include the part for because there was already a similar change in the codeThanks to Sean for the contribution. +I let it convert the project to subversion when prompted +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +i attached my ttree rendered file +Mailing lists are open for business +Thanks for taking a look Otis +If that is the case does anything need to be documented +BTW RI also prints the exceptions occurred in EDT to stderr +Brilliant workI need to find some time to let the tests all run their course Ive left them running for about minutes and they havent finished yet +bulk close of resolved issues. +OK I added a unit test that hits the NPE and corrected it so the testnow passesSorry about this and thanks for catching it Aaron +The blog post is quite vague +HeyThanks for the bug report +Heres a Groovy test case that proves the issue +Closing all resolved tickets from or older +Thanks me know if I can be of further help getting this contribution included in +trunk +PDF created with and PNG created with code in bug JDK cant read the embedded ICCprofile which leads to inverted colorsI fixed that in revision +Added test cases for testing transformation to legacy EAP versions. +Seam GA Tools GA +Yes its work fine and its not a bad concept to have a build pom and a parent our case the root pom is a child of the parent in order to benefit the release change version system and retrieving SCM only bad news is that the generated site mention the root module as a real separated module +OK +eg when reporting issues be more detailed +bq +Maybe after we switch to Teiids query parser this will be easy to implement +away + +Verified in Developer StudioVersion Build id Build date +Its an FXP file +For me it does not need more sophistication but I could be missing something +The issue doesnt seem to be present in issues I encountered with Hibernate and Aries JPA were mostly ordering issues that prevented Hibernate to work if the for Hibernate was not present when the aries jpa container starts. +Hey JacquesThis issue is waiting for your action please have a look at the functionality and patch attached +This should apply to runtimes and above only +Yash is removing testcases and and renaming testcases and to and respectively +Slava please add java as a key word for this wizard +Any update Thibaut +I think the best thing to do unfortunately is make an for each snitch allowing them to expose their specifics if they choose to +Attached is a patch to implement probably get a test for this too but I did not find anything for testing sourceSchemas parsing out +We have the module prototyping this now so we can close this jira +Uwe I just ran all Lucene tests with your patch and they passed so +This patch solves the bug +PMC to create any sort of dependency upon code that is more restrictive than the Apache License +I got rid of the auto feature long ago +So one issue please +That handles toosvn ci m Fixing +Previously seemed to be an option but was not honored +Thanks for reporting this though. +Here is the correct fix for it +The size of the large images has been reduced +They are outliers certainly but with the current scheme where they are copied multiple times on the heap it is not only inefficient but also occasionally fatal via OOME +I am reviewing this with modified comments as per Olgas recommendation committed to trunk. +That is a big advantage +Kindly verify. +This simple patch should fix itInstead of just looking at content calculates hash of content parse textI think this keeps intended behavior of as well as a tiny change in pagewill create a different hash as before. +Ive applied the pathc in revision as proposedThanks for the contribution! +Rebased the patch against trunk +I need a larger patch that can deal with all the changes +DebuggerSince you correctly identified the root cause of the issue do you think that you could do the required modifications to make it work on Weblogic and submit a patch? +Thank you. +We should rename type to recordtype +This would allow us to continue to do speedy checks while still expanding the user base outside of English without concerns of encodings breaking their expressions +Workaround for Linux was applied. +Also if you want to continue working with the github clone I think git svn rebase may pull in the karaf updates needed to get to svn HEAD. +After I restarted the server and agent this now only happens for some resources listed in the Browse Resources these are resources that arent currently running +Bulk closing stale resolved issues +If you dont have any success with callouts let me know +Ill keep an eye out for it +example does take the messageExchange element into account +Heres what I did I downloaded the wsdl I need to generate stubs for modified it then generated the stubs +This has been mapped in branch and head in products have been added in prod and Communication Platform downloadsPID need a test plan on this keeping it open +Maybe you can think of an even better approach +Give it a. +Would you consider documenting it in the where you make this setting for WindowsGood idea! Here is a new patch +Forgot to say The word CJK means Chinese Japanese and Korean +Additionally i cleaned up the core +Nick thanks a lot for your explanation +Issues with the following files from deploy bad JUDDI datawsdl possible bad soap URLI seem to remember these errors from EAP +In that case lets change NULL by adding contextual prefix +The bug was that a root xml element was expected under the xhtml div element +Yongqing can you add the change to wiki after this is committed? when you do that please mention that the underlying data is untouched and it is the responsibility of the user to change the data to fit the new schema +please try havent heard complaints about arcsde delete not working so Im going to assume this is done. +Spring provides for tiles We were not able to find established and supported practice of using tiles with spring online +MistyThis change is pretty much under the covers +Fixed the min variable and refactored some parts of the binary storage subsystem +Shouldnt be a lot of work i think +Its strange that it idndt happen until I had toggled between connecting the same user in short order +Backport of if anyone wants to review +Sid from your comment on +require is to load a directory or jar file so that you can import classes from it +It has been introduced due to +New version added release +Howeverthis doesnt appear to be a Xalan or Apache bug and I doubt theres anythingthe Xalan team can really do about it +Version has been releasedMoving all related issues to the next version +Thanks very much for that explanation Kristian +The fix for embedded camel was the blocker portion +Similar to how eventual consistency may prevent some applications but are good enough for plenty of othersAre there any specific code changes we can make to the patch in order to make it commit worthy +Was this on intention +Do you want me to create another jira for this +Time taken for adding hosts is now more comparable to baseline numbers +Patch applied! Thanks Bill and Senaka for the Input +Did you try with the official releases that I flagged or with a more recent build +Ive applied the patch and all seems to be well +When a task fails it sets the following task to be the first to be checked for assignment to a +Which obviously raises it head in as the is an internal implementation detail not something you want to export as one of the +Patch commited and updated +No need to go any further anyway +curly brace patch committed +If we pursue spawning a request or delaying statement closure there is still an issue at session closure but presumably it would happen with less frequency +Username javajedi +Alfie an issue for your problem exists in JIRA for as +Bulk close for release +This patch adds the missing newlines +ChunhuiYou have patch for? If so can you upload it here? There is a patch for but not for +Will commit within a few days barring objection +and also +I was thinking about the approaches +Why not leave it always as a binaryThat is avoid at all trying to convert to an atom +After went through the patch of I found it does the exactly what I want do is a FIFO queue in the scheduled thread pool and it will reject the request when the work queue if full +The class I posted above has served me well in several applications +It always reported coverage on classes hit by ANY tests and didnt show classes that had coverage +Not a bug +Didnt feel right to add this to a performance releaseYou think this should go into? Might be good to have this correctness fix early +This would keep us from waiting around for scheduling opportunities that will never help us meet locality for an ask thats + +It also removes checks for these constants +Changing old diff new patch with changes in +This works properly now +Thanks RobI checked this into trunk and +I have restored the completed description. +Ok later +Indeed +The juel and juel artifacts are already in the repo but the version deployed is a release candidate +The regex capability is actually in the usageAdding the doc +I shall upload the design for the same soon +to +the way ErikDojo just came out Sunday So is already out of dateIf it would be easier for you you could just nuke what I had out thereJay +It feels a bit like a race condition the viewport id is not ready when ripple tries to call for it +Verified with JBDS +Thanks +Will look into this over the weekend +So each switchyard app would be treated in isolation will that be a problem in the same vmRegarding the third point if the mocks currently need to declare the binding they represent then the only problem will be if a switchyard app uses a new binding that hasnt been catered for in the simulator +Even though this is release noted for I would like to move this to and start a discussion on this +My manifest still has all the stuff +Built and worked fine +We fixed so the text installer could be used as a have not had the time to setup an environment so that I can reproduce this + +bulk close of all resolved issues in preparation for release. +However since I was interested in learning a bit more about Derby and because I didnt know the state of the existing patch I started from scratchIt is my impression that the existing patch isnt production readyI think it would be cool to compare our approaches +The stree classes have been removed and replaced by a new scheme called DTMThanks! I am having the same problem however its not just while calling childNodes this happens while doing anything with the contextNode that you derive from the context +We probably need a test that delays the flush process otherwise flushing to RAM occurs too fast to proceed to the next tier +Fixed in trunk your feedback would be appreciated. +trunk +Then end user can chose this if they want the JMX stuff goneAnd then can just unregister the at the end which is does by default +Can we live with Roberts solution? Please comment +Ive attached the updated patch +Hopefully final patch with added Javascript tests for continuous repl cancellation success and failure cases +My dear svn managed to merge this commit with my changes of andThanks +Adding XML support could easily be done +Is there any special considerations with editors an such to not corrupting them? I assume I just need to change all the XL references to message files only contain plain ascii characters so any text editor that handles ascii should be fine +we do run for x just in a slightly different place +If i move the file to srctestresources dist build is working fine +Yes lets not confuse things by treating them as one +Check in to xbean trunk and prepared a release +Thanks +Thanks +The constructor for now has protected visibility allowing for extension in combination with extendingIt would also be possible to make static or indeed a class but that seems like a lower priority change to me and we currently dont have a strong exampleuse casejustification driving that change +With revision adding a test case in main codeline +I was closing streams in a wrong orderThanks to Apache Jira Issues Team +Glad you were able to figure out how to build etc +Also to Julie for verification +Unscheduling due to inactivity +Jarcec that is a good point +That sounds like a good idea to me +Im not sure whether there will be a lot of demand for such a degree of sophistication in the strategies used to obtain user information from requests +on patch +Hi therefor the mailbox api part I suggest you to have a look at the jpa implementation +Please update to the latest version and try highlighting is incorrect here but there are no compiler are you trying to launch this application? I can run as Java App Groovy App and Groovy you continue to experience problems please attacha failing project +Added support for this issue with committed revision +Committed to and trunk. +Hi SergeyI would like more detailed information a diffpatch for the code you added in your comment of MayAlso pardon my ignorance but what do you mean by you may include some popular fonts into your program distribution and load them into some static mapDoes this mean we need a special build of that includes the fonts we wish to use? Or is there some way of specifying the fonts dynamically even though they are not built together withIm a newbie to this area so Im sorry if my questions seem elementaryThanksJonathan Levinson +Attached the v patch that Im going to merge later today if no other commentscontains the written by Jerry and removes the Override comment pointed out by Ted +Hi gefreyDownloading community look for examplessourcestarget +This is not in master yet hence still being unresolved +Ive just committed this +Oh my the script generating the file was broken +Rather than changing the documentation I think the behavior would make more sense if a dryRun could be followed by a prepare without deleting +I know that had some major performance enhancements especially on things like threadsScott +From the aforelinked PDFSince the buffers are allocated when the SQL is parsed the size of the buffers depends not on the actual size of the row data returned by the query but on the maximum size possible for the row data +Id like to get this done this week and to my horror it is now Thursday +See. +Gaojinchao will address your comments in the new JIRA +Also please confirm which Apache rsync repository you are using you stated samba which is incorrect in this case +Ill take this on as one of the things to do before the RC +I upload +Not sure how to handle this its not precisely a duplicate +I see that this issue was moved to +The Support Team will review these documents and make the appropriate changes this is fantastic news +Hello PranayI am uploading the patch for showing billing account and return information in party componentSteps for testing +What file are you using to train the classifier? +In fact the issue is also happening when we build deployment artifacts the multi is removed from the generated +RyanThis looks great it does what Id hopedIve never been all that happy with how the was organized so Ive attached a patch that builds on yours and refactors a bit +tng tng Nov +Reverting back to track the community change +The issue is not seen if the below patches are not applied on the Xenserver but if the patches are not applied the PV drivers are not getting installed for the VMs deployed on the Xenserver host +Is it possible to have both the namespace parser adds a tasklet but retains the properties of the parent if it is a FTSFB +Hi ScottHas there been any movement on in the last two years? Do you know of anyone using it in production? Or is it a now defunct project? Any information would be helpfulThanks like the project is dead no update for years. +The difficulty with larger puffs is keeping the shape of how the blocks interact with eachother + +The change of will be included in +For M Ill just remove the name mangling from the runtime +I can wait until you release the next +its a bit icky that changes to the list passed in effect the writingI agree but thats a more systemic problem +Otherwise the patch looks great +Identified the problem the skipping of setting name rules is done in the wrong place + +Thanks Hairong +Cool happy to do it +DaveI think if you look carefully youll see my logic is sound +Thanks Chris +Will also reopen for a solution to the specific problem of. +That is what N has also suggested in the above comments +Testing testing let me know if this works Bart +think the first are the most visually appealing and the balloonsv and church and state convey the message better +I can access now +lets retest this now that some tests have been fixed in did you want to review this before commit? from me Ill commit it on Monday if there are no further comments +Unpack and read the you maybe have an old version of your Component class in the classpath which does not have the setValue method yet? The root of your problem seems to be something like this at least to me +Tested with its adding selected features but displaying annoying message box +Let me know if I am doing st wrong +here is the file for the Maven project im having trouble a lot for the pom +sorry for the with the right resolution +This is what startAppstartGroovy are doing For Darwin add GROOVYAPPNAME to the JAVAOPTS as Xdocknameif darwin then JAVAOPTSJAVAOPTS XdocknameGROOVYAPPNAME we may also want to set Xdockimagefia dock image would be a nice touch as on only +Having with distribution would help users use vanila on non HDFS without having them include file +Calling finalizeBlock in the last may not work because the replicaInfo in the replicasMap is already a finalized one +The server wont do anything +bq +My guess is this functionality will be used less often but it certainly is stay tuned +Merged into the master branch. +The change looks good +Hi SewerynI will commit this only in trunk as the release branch only gets bug corrections and no new functionalitiesCheers +Just to add +Have you installed Update Pack for BCBYes I haverlc +Patch created with git not sure if I did it correctly please let me know if you have trouble reviewing +Added the methods and the variables annotated them deprecated and added a test +Thanks +There are few binary files on RB + +Yes you are right +Thank you Eric +Patch to point to correct locations +I will rename it from to and we will see if jenkins is base class so hopefully we can keep it abstract +I can reproduce +So Ill run tests and get this checked into the trunk +This relies on specific code and we would have to reimplement at least regionMatches on top of the APIOf the other ordinal last and countMatches are easy +Note that this can currently be done by writing aSo its trivial to implement just create a that transforms to javabeans +Your patch has a bit of a problem if someone renames now one of the headings or rearranges paragraphs all heading anchors will be invalid and all direct links to them will be broken +Bump actually this prevents compiling from git on OS X. +This is a bad habit to get into +Im working on this dont waste your time Uwe +Replication workers start replicating to local bookieSo here are my thoughts about current Jira Create a class do start Auditor service and replication worker +I cant get why you cant just use current selection attribute +We need to get some clarification from the spec people about whether a property is valid +Actually it is not clear to me why this patch will help improve perf +fixed since r +is this causing any problemitd be good to make the release plugin expand the vars but it shouldnt cause any issue is very annoying because right now Maven tries to findjar and therefore fails. +I just committed this trunk and +So in other words the Resolved state is as far as a fragment will ever get and that is per the spec +This could be because the postgresql db i was using had some data from the earlier tried with mysql on my local box and it works fine +Looks like memcache has the same issue + +Joe I have run mvn eclipseeclipse after your last comment on the JIRA +Looks good +Ahhh right got you +Help in resolution is appreciated. +Assigning to dev for future EWS release. +Alternatively you can provide a URL with a root DN ldapDCcorpDCdomainDCcom and Suns implemenation will automatically look it up in DNS. +Hi MaciejThis might be something you know a bit about +If everything looks fine please second set of files with +Going to apply this patch +Fixed in +Committed revision . +You also might want to take a look at the remaining issues as Im not sure whether they should be fixed or how they should be fixed. +ArjanDo you have a version of Groovy andor a version of on your CLASSPATH when you run Grails? If yes try to remove them and see if that has an impact. +I see that there is lot of testing based code in and corresponding impls +push out to +We could have POST to a graph be streaming triples and POST to the dataset itself for streaming quads +not yet sure why +Yeah looks like a duplicate thanks Flavio +And that contains bad corba api classes +I understand what you meanDo I have to close this issue +Hello Klaus can you have a look in targettaglist of the root project and the modules to see if there is something obviously wrong? Just trying to find out if the issues comes from maven taglist plugin or from sonar taglist plugin +Applied the patch +My thinking Since the takes are always happening from the head then there should not be a need to shift n items during the replay of every take +Hei! The duplicate issue is THIS ONE Mine is and is previous to thisJust joking I only wanted to point out that the comment of is more complete than this one and should be considered while fixing this issue +Wiki pages updated. + +Per phone triage with Burr and Max on +Attempting to persist a property that is set to u still results in an exception. +yes no worries +Im not sure why ignoring +Upgrading to to see if this fixes things +I am new to this Apache license stuff but you might need to reload upload the patch with the license checked before we are allowed to look at it +Patch applied +applicationsproductsrcorgofbizproductspreadsheetimport +I started drafting an improved Jackrabbit repository bundle in the sandbox area of our svn +So even if it takes kb of address space it does not take kb of memory +Andrew and I spent more time on this issue today and realized that were far beyond the scope of the project this is a general web services and problem not something specific to WSRFWSDMetc +On thing I just thought about enabling this feature doesnt change that we currently still roll on the total size of the file +trunk Will now ignore this exception and log a message at TRACE level in case you want to see the thrown exception +Couple more tests added +The current implementation doesnt support the whole command set of a typeC font which leads to the decribed issues +The problem arises in which only understands two types of modifications that can be made to a columns default value changing the INCREMENT BY value for a generated IDENTITY column changing the RESTART WITH value for a generated IDENTITY columnBut there is a third type of modification that can be made to a columns defaultvalue which is simply to change the default value that is to be used for anordinary column if no value is provided when inserting a new row into that tableThe fix is to add a new modification type code MODIFYCOLUMNDEFAULTVALUEand to teach and what to dofor modifications of type MODIFYCOLUMNDEFAULTVALUEIve got a diff ready and will post it after running the regression tests +Fix confirmed in a manual build +Now wfsdispatch files gets deleted as soon as the request they are still created whilst in fact it would be better to have them not created at all +It will be removed later +Hows this look + +Please refer to the spec on how to configure the guest for get automatically. +Fixed in r +I remember there was a change for the header handling or something without it messages didnt go through +The Webservice annotation can be put on an interface but only without parameters like the serviceName parameter that identifies a service +For a start we can have a Combiner extending the current Combiner for Message extends +This problem cannot be fixed per se but it can be avoided by compiling the rule code to bytecode. +Thanks Zheng! +From all I can tell it is a useless property that could be removed withoutconsequence +Sending information about bad nodes at bytes per node for AMs every second is about MBs of trafficSending deltas means tracking last and current states on the RM on a per AM attempt basis +I tested deploying a WAR file using the it uses the same path so you see same issue if you deploy enough WAR files +bulk close of resolved issues for release. +Obsoletes. +Barry can you look into this? Minor tweak +I have verified that does not exist in the branch +coreWebminiTestsexceptionMessages cause the the same exception +Look at the +Igor the current behavior actually cause problem to existing users as it been reported in the mailing list and several jira issues including this one +But lets talk about it when we need thisbq +The patch assume state saving is using a map to store state information +I should note that the merge code was much easier after Randall provided an insight about using a second session for the same user that pointed to a different workspace +The problem with freehand editing specifically with nix boxes is that I wont know whether the path they enter is relative or workspace relative +for the patch +Donald Thanks for breaking out the database versiondriver support into a separate patch +fixing this now +Thanks for the poking around Colin +much better IMHO + + AS has a nice JRE comboWhereas its missing for AS runtime +Hm i think i have a patch for this issue +Assigning to Benn to help integrate the latest plugman implementation into +You dont need to create the proxy per thread +Gregory +I saw you have fixed this issue in a different way not using the adapter +To this point it is unclear what impact such a directive would have +Great idea thanks Christopher! +Now all attributes are editable +I was about to resolve it for you but then I noticed that you had the affects version set to but with no target version setDo you intend to commit this to as well? Seems to me like we should. +I do confirm that has been correctly fixed but this fix works if only if your result directory doesnt contain any files starting with the TEST prefix otherwise only those reports will be analysed +Attaching the patchJust for reference +Oops actually I misinterpreted an unrelated error as being related to this issue your fix does in fact solve my problem +From Jason of the JCA spec +HADOOPHOME and HADOOPCONFDIR should be the standard way to locate HADOOP +Could you either or provide comments shortly please +Exceptions reading the edit log as opposed to exceptions applying the edits +All plugins expect coverage plugin are using this path +It reads it fine but writes it literally instead of outputting the ampersand character +I think the fix version has been wrongly set to EWP +Ajax and bigger thread pool is not the right solution +There doesnt seem to be anyway to manage versions in a centralised manner beyond this approach +Expose this in interface first +This has not been done +This might be a Facebook problem however they dont say whether data element will always be there or not Structure containing a data object and the number of total likes with data containing an array of objects each with the name and Facebook id of the user who liked the post +I have already made a similar change to to allow it to be raisekilled safely +Please update the class javadoc +There could be a transit container statusHave run the test for tens of times and no failure occurs +I created create as a replacement to createWeightcreateWeight was protected when you made create you made it public so that was the change I think +This problem got in our way when we tried to implement a repository system where each development branch would have its own configuration that would be linked together using include files +Id appreciate it if you could try testing with tonights snapshots +Sounds like we can close this issue as Not a Bug then no? + looks ve committed this and merged to +A already exists +Is there any progress on this issue? Currently the templates cannot access the out variable making it impossible to create actual output for a script +That alone did not work however when I deleted and then it worked perfectlyThanks again to fit JIRA size limit +Ran through the petstore sample created an account added an address bought some animals checked out saw the updates in inventory signed out and browsed +I checked it with the latest SNAPSHOT build IMO it works as expected +So all issues not directly assigned to an AS release are being closed +Yes that would be nice +And here it it be better to just remove upgradeProgress +Resolving as invalid +I am closing this issue as the NPE is definitely a different issue and is already covered by +Committed patch to documentation trunk at revision +Updated resource file from apache svn before creating this patch + +Also in my previous comment there is no way for a change to the provided dependencies to effect a difference in runtime behavior +Thats fine with me +We should figure out how to use a keystore gbean instead +Oliver it would break nothing default behavior would be as it is now. +So Im not sure that the gain we will have by avoiding the exception which should occur to times on avarage will make it up the extra cost. +I believe has been fixed since that first change I made mentioned by Steven to account for positions returned from an Analyzer +This tests show the problem most of them fail on Harmony +is the combined patch including the mv then the patch changes +why is this issue ResolvedFixed +When you join tables it is most probably because you have a relationship +And wed need it to get into if its going in soon +Patch with test and fix +Another improvement in Rev +Canceling the patch while i address the feedback +specific names +Well OK but I guess of interested people are interested in the source code of not of its site. +Thanks Paul this is a good writeup +It was a classpath issue +on patch +Martin looks great +if anything urgent please contact +Reopen if still an issue with latest to set fix version to again with correct fix version +I had dependency issues so I was running an older T than I thought +Schema utils modified to move taglib tags into a new tag +Because when the follower writes a new log file without writing a snapshot with the old transactions on restart the ZK thinks it has the transactions up to the zxid in the log file +Yongqiang can you please create a review request on? Thanks +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +To clarify file generation will fail means that the ant task will abort with an error message and refuse to create the true eclipse files from their templates +The FTP outbound Gateway should serve as a good +It requires the request entity to essentially be parsed twice +ClausThanks for digging into this issue and resolving itMange tak +I noticed that there is no drive letter C in the path quoted in the exception messages in both cases +Changing the fix versions as we can only make fixes to future versions and has already been released. +Hi Chad can you attach the trunk patches to insteadI have marked this ticket as a dup since its been superceded by. +What if the namespace was truly wrong? +The open child too. +to trunk +Fixed thanks a ton for the patch! +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Already removed above +Prior doing that you should the maxent moduleThe attached CONLL files can only be used for research and therefore incompatible with the AL license +I just dont see how or where such a structure would make sense +I will investigate extending the CLI tool to getset properties to do this +that was a bug in my script +Is this an issue with +This has already been implemented in Camel. +Fully agree we need an improved Exception handling and I like to get some more contributions from your side +Thank you +Can you go a little more into what the issue is with this definitionAlso whatever we decide is there somewhere where we can document exactly what these mean +More importantly are we actually checking that a DCM can be started withIt appears that were attempting to parse it in but trace of being able to start a should be doing that +final classes is not possible to use when using proxiesOh and please reopen the issue until youve had a chance to inspect the example. +Please clarify if your understanding is different +Addressed Matteo and Nicolas comments +what do you want to do with this one juergen? you added those xml headers i guess explorer chokes on them +Ignoring garbage Input seems like trouble to me too that can only make things worse +Fix is in SVN. +At the time this was happening we had C +This has the disadvantage that if numRows is really large but most rows are empty this will waste memory but has the advantage that it doesnt care what impl of Vector is givenDoes this work for you Gokhan +In standalone Web app mode this doesnt occur because I believe a true new request is being madeI believe this might be able to be resolved by creating a new action instance in the instead of using the original one +From the trademark point of view what informed consumers are looking for is a software product that does something useful for them +It produces invalid grapth which confuses loop peeling +See for the status of this configurable logging change in EAP +For me it still doesnt work in the trunkregards ronald +Can you attach a diff file with all your changes? +From here we could easily try and exchange and make your test run +Resolving as invalid since well be using a different bloom filter implementation. +Hi CodyCan you please explain me in detail what I need to do to reproduce the problem? I have created the above class with Tools and I dont see any error messages +bq +I dont see any more WARN messages +This appears to have gotten resolved somewhere along the way I no longer get the NPE upon attempting to install was recently seen again when upgrading a base JBAS install to Verified. +Can you please verify if you can reproduce it in environment without proxy? You can also create test remote repository using file protocol +Yeah seems fixed in trunk but it is still broken in so maybe just close this ticket against instead +Will aim to do that soon +scrutinize this one a bit moreTake your timeDepending on how long the review takes we can decide which one to commit first +Hi RickWhy did you unassign yourself from this issue instead of just closing it? It appears your changes complete what is needed for the problem described in this issue to be closed +The failed and are not related to this +attach a patch to expose number of worker threads as a configuration setting +Also can you write yourself down as module maintainer of this module? You are the only active developer I know of on the list. +! +Thanks Emmanuel looks much better now +just keeps getting put off due to other stuff coming up + +I have merged the changes to the branchIn we are using the from codehaus +This looks good to me +Its a secret feature for when someone asks for something wierd +Broker not knowing it is not the leader anymore continues with the produce request and appends some data to the log +Fixed no updates for symbol table because currently we dont deal with generics in it +Ive got a hack patches that clears the state in unit tests +I think Amit checked out the branch at a time where the eclipse config was temporarily brokenI dont know anything about ruby to look at this script though so it would be great if someone that does could evaluate this +Revision has added new version field to persistent entities for detach tests +If you wantI would be interested in this +Fixing the typo in the title +Thanks Navis for taking a look +How do you plan to get all tags? Is it user defined tags or we could get from Plugin Category Piority listsPlease share you ideas about location of this functionality on Violation Drilldown page +Resolving based on the you help creating a doc bug for this? I think the libvirt version causing the usage to go wrong needs some documentation if not available already. +Thanks Devaraj +Thanks B Anil Kumar! Ive committed this to trunk and branch. +If you realize that the container needs to be even larger then you can increase the size of the container +If a transform is added ot the same phase then we have to process the remaining ones from the stack +It cant go wider because the preview off the edge to the right of the screen capture expands too +I was on a other way start working on improve your to have a better maven build which is more easy to integrate in SonarLet me know if makes sense for you +More patches are attached +Im not completely sure what you mean by thatWhat I mean is if you lose or corrupt the +Both sets of tests can be controlled by the same property +Id prefer to change the behavior of the constructor to omit the reason phrase altogethercheers Roland +I think it does make it worse by adding a new global collection that is being updated by the lower levels of the call stack +Agreed bumping to once its out instead +Working +Fixed in for. +Minor nit should we set the default value of to or some such rather than? Also we should call it the current one might confuse users +Looks good now no warnings in the catalina logs +Verified in the release notes. +Dupe of +should be +For Hudson to apply the patch it should be generated with +Would a different UI element work better hereIt would also be good to add a column on the Schedules page and show the list of queues associated with that schedule +Groovy cant solve this atm the full implementation of the new MOP is required to resolve now +Another use case that comes up in Grails is that you have a with a nice environments block in it already +Using spaces instead of tabs +Once I updated themanifest to include the everything deployed fine +Thanks for the contribution! +I needed to modify the cached config so that importDirectory did not immediately implodeI was not able to fully validate that the importDirectory additions from still work since Ive never used that command and wasnt sure what input format is required +Btw the attached jira contains a way to reproduce the problem in mind you have to use the WMS request provided by the user which has the bbox entries swapped compared to the wms since the official EPSG definition of that layer says NORTHEAST axis is the same patch from but as it is a problem it has been attached to this ticket +If Solr crashes while downloading that will leave unnecessaryincomplete files in the index directoryIf we dont want to try and pick up from where we left off it seems like Lucenes deletion policy can clean up old index files that are unreferenced +I didnt encounter memory leaks after more profiling so Im closing this bug +and it works in my testsIll need to package it up as a patch and probably write a test so I need to figure out how those work first +I will run through a soak and try to hit the race condition then make the fix and soak again for a few days to make sure it is fixed +I think this is already isnt fixed at your aspects are not in srcmainaspect they dont get s patch permits multiple source directories or completely different source directories if thats what it consider adding it +I am not sure if i would have the need for it all circustances I think it should be optional whether or not to include blob and clob fields +I agree the path manipulation looks wrong +Ran test to make sure that this patch unblocks +bq +Id probably subclass if this is the version thrift goes with + If an extension is provided in a shared library then it would currently see the BM which is applicable for the bean archive for that shared libraryhumm in OWB even if the Extension is in the shared lib of an EAR you will get instance of this very Extension per warThats what I meant with per WAR in OWB in my explanation above +We cant resolve this issue until we dont add Seam Solder to tycho build +Looks like that last patch should actually have been version +The idea is that as soon as the packaging was defined aswar? there is no reason not to set srcmainwebapp as Web Content for the course this can be accomplished later by runningUpdate Project Configuration. +If cp fails to load will all the data then become visible Andrew? This is straightforward to address +Thanks Ning +Remarking this resolved the follow up is in instead +This also include the change where the will use the as the underlyingThanksKishanthan +This is prolly not that related but I thought something thats a bit counter intuitive and cumbersome for developers +I dont know if you noticed this comment but when I recreated the project following your steps the list of jars was a little different from the jars listed in the Simple Project page could that be part of the problem +All what is required now is to have those two extra in the classpath +the dif to the new base clazz for the recent recent class support added to +I would say World RecordD +we aim to please! smile +My comment was related to if it was a archive +Hi Marcel +I take it back Actually its working now +bq +Please verify +into with . +In that case its most probably intentional and its their authors responsibility to document the proper order in which they are to be applied +Merged into develop +resubmitting because the build was broken at the time jenkins patch on just got committed with basically the same change so I think this is a duplicate now +I changed the features as well +I attached a file hope its in needed formatNo test case could be provided since it is a simple close operation surrounded by try +this was a development bug in the version of the code but has been fixed +And right clicking the and click Make Deployable is not working +in light of comments +Works like a charm +Moved to the release since this is needed by other projects +Correct i thought about that too but changing the namespace to solr would break existing configurations that rely on solrindex +Daniel the patch attached is basicaly fine but please try to add more unit tests much in the same way as Andras has done for his last patch +Ill give your patch a try +Addressed review comments +There would be a row of circles for each shard and a column for each replica +Is this still a problem? Please reopen if it is +ill ask rich how he wants to proceed perhaps we should close this +I couldnt figure out how to do it otherwise in a standalone patchI hope you can review these patches soon +VladimirThis bug cannot be reproduced on r and I see that we now have an explicit null checkPlease confirm that this issue can now be closed +Using the Web Developer toolbar in FF my page does not show a form element for the date pickers but yours does +Hi AndreasDo you confirm that directory exportbuildhudsonjobsworkspaceprjbuildsonar is not under version control +Deployed initial snapshot version +That should at least provide a workaround for QA to continue testingHowever Alessio is correct that we shouldnt modify the dependency scopes during productization so we will try to restore the original scope configuration and then deal with any remaining missing dependencies individually +Still not sure about need for embedded gradle wrapper in project source may remove later +No need for integration test. +Of cause it is possible to add these jars manually into the classpath +will commit after tests +file sizes inMETADATA table are estimates when tablets split. +Patch for discussionIn read only read for the first edit may throw EOF +Fixed under revision r. +Still appears to be present on minotaurwwwcontentpwd content ls l bigtoptotal rvs incubator Aug css rvs incubator Aug imagesr rvs incubator Aug r rvs incubator Aug r rvs incubator Aug r rvs incubator Aug r rvs incubator Aug r rvs incubator Aug +Thanks Grant sounds cool +Will go over the patch in more detail +has this been tested with? If so to get this in asap +Give me a little more time here to figure out the problem and clean some small stuff up on Davids original patch +Running test patch locally showed several license issues +think this is not an issue thinking about the way messages are attached to the choreo task boundary got me thinking that possibly they just mean through sequence links to a catch or throw +In Cpp defines are not overridden by values supplied at compile time with D +Attaching a patch to address ysteins comments +The rules have been updated now. +Is there a best practice when it comes to using special symbols to separate entity fields for unique selection? +patch looks good +Modified moved to in clientpositiveModified moved to in clientpositiveAdded deleted respective output files +Fixed in cant find the source of this dependency issue +Do we have a recent plan or update on this effort +We replicate all up to the exception and then next time in we should pick up the IOE again +Maybe there is an issue with message uid ? You could move message by hand with a SQL tool to another mailbox and see if the problem is still thereWe need to find out where exactly the problem occurs +too +HiOk +Assume server started checkbox +That wont fix it +HiI understand what youre doingMy thinking is that Spring is not necessary to expose the as an OSGI service +Bulk closing stale resolved issues +Logged In YES useridIt seems that there are several problems defines four that are onlysupporting the debug mode +Verified on EAP Weld +If theres a better one let me know +Do we also need an implementation of lookupFallback in +Updated patch file per previous comments +We should integrate that with NEWS +they really only need to build similar Dependencies should very rarely be in parents if at all you should use standard OO principles to encasulate them in composites that are reused by muliple projects as would take a very different approach and completely disallow modules projects from being parents and vice versa +However I being new to Java and it sure would be nice to be able to know this issue isnt about the tutorial but could someone please point me somewhere that will help with debugging with with Eclipse IDE? Thank you so much in advance +Fetcher remove shutdown and keep only stop + passed unit tests and + +Committed. +Contains the along with sample Order and the XSL used to transform the xmls +Logged In YES useridA sar does not support an expanded war inside the sar +bq +Thanks Russ. +If they are the problem would be very easy for them to fix by renaming their own references +Thanks Dave i have added your suggestion to David +Its probably time for a file to track these configuration possibilities +It appears as though during integration the file is being manipulated the file fresh from SVN is fine while the one used in the unit tests is substantially larger +Im not sure how to test your change easily but it would be good to know that the fix works +This issue was actually addressed in CP +Thats what I thought +removed description that was a separate problem. +Ill throw an eye at it +As the bug noted the is acquired outside a transaction and there is no guarantee that the is even valid after beginning a transaction +All the sample files still include the incubator maven repository +Normalizing this might be helpful for data warehousing statistics. +Fixed on and commented on a saturday because during working days I have to follow paying customers +Thanks guys. +Thought about using then punted +Or we create a second AOO only Pootle instance where we can get more admin control to create and manage projects and languages +I didnt realize this was intended to be the goal of copyunpack +You intentionally left it as it is +it isnt perfect but might serve as a starting point for a better solution +Here is the Cheersve improved the batch exception handling +I would suggest you rename into because is used somewhere else in a different context and the new naming would match the naming of the BSF part +The Layout Demo page was the result of that discussion +If I dont commit anything from toolsstress or toolslib is that sufficient to keep the stuff separate +Seam is ok as a default for EAP with FP installed but not for not sure we are able to differentiate without having a sever type for SOA +to and trunk +Is this going to actually get done in the next two weeks or should it move to? +Maybe it is because I use Maven? I will have a further look at it +looks fine to reviewing the patch I recognized that you prepared a new one in the mean time +gitignore rule + +I dont understand this you elaborate on the use case and describe more in detail what happensRegards mean the array should contain only right +Maxim I saw the Mac issue +Let me explain my problem +Thanks +It would help the user to avoid having to write load statements which is +It uses the same environment and techniques as +Steps to reproduce added please try +Removing component and fix version as native stack reached EOL +Im currently working on getting this new version of javassist to so we can use it for releasing CR +Flex calls it a +Updating summary to reflect that this patch is the patch applied to. +Not sure if I am missing something here or if this is a bug thats crept in since this issue was fixedColm +Could you provide the full stack trace so we can see in more deep what is the problem +Remove Fix Version as the issue was not fixed. +So what to do +do you mean vodafone +Allow for looking up option with either the short or long option +Can we at least make it the configurations value +So it is not necessary to update as the number of rows grows +Committed revision issue with release of +Triaged for +Tapestry asks to determine the properties of a +It comes with following features Injection of a Spring configured Hibernate which is a transactional testcase +It depends on JDK specific classes +I think we should also fix so that if the write lock is in the index dir it doesnt generate the large digest in the file name +Thats why I didnt change other placesI dont think that will see those changes so I tried to cover up all possible cases here +If the item is found there build the item state from it +Set patchAvailable flag to request more feedbackI havent yet run the complete test suites with this patch but by itself runs correctly in my environmentIf people are generally comfortable with this patch Ill moveahead with more complete testing +to the queries moduleCommand coming upEverything compiles passes +But I dont have strong feelings about it if anyone else wants to chime in +might give us the tools to do thisI am halting work on this issue +It is a requirement from the XML spec to translate the character +Okay reviewed and tested patch with sqlite and bdb on trunk +I dont see why the code constructs a key like only to parse out Z again to use as the ID +I assume you tested your patch and it did make a difference right +Any resolutions +Is this still an issue + +The patch doesnt apply anymore +fixed and Renamed pointertypes to lowercastpt +We should find the apps that are deployed but not in ZK and undeploy them +Looks reasonable +Ive applied the conceptually changes by hand and your test now passes +Right +I committed this + +Used instead +We should ensure that our impl for Wookie Widgets is compatible with the Rave OA Hub it as a feature I have no for basic feature support to make it easier to create widgets that talk to other widgets and gadgets in a managed hub environment such as +UGI loads Kerberos tickets into the subject +Good catches put up v addressing your comments +fix merged into branch in +I just committed this +Does it look betterThe Hadoop change sounds like the right approach +Im going to take a look at the helloworld ones next +Thanks Niels for catching my mistake. +Waiting for the release of them now the release process is started by to last remote resources plugin in rev +You are right SaschaIt might have been a browser cache issue +If you were in that boat not sure how you would have voted on this issueIn this kind of voting I think the community suffers +OK thanks +bq + in + for the to trunk. +Merge complete +I deployed the new jar and set true and experienced the hang +This is ok since the calling methods within the library are declared to throw a +In the i have fixed the reported have applied the to trunk at svn revision +Targeting to as has now become a release with mostly fixes and minor improvements +Then we can get it in Camel file that fixes and +I dont see a strong need for a custom entity as long as you dont DOCUMENT how these things are used the are not used +We have to recognize such a dependency and ignore it to prevent the +Fair enough Laird +Ok this does seem to be browserplatform dependent +It seems like Peter implemented to work with any thread pool implementation without compromising the order of events +Failures Errors Mostly its authentication failures and I think expected column Nullability changes +The advice that produced the error +it should be caused by using ant in brew build because there is renaming to but it seems that ant has a bug with the construction of the renaming as is in build for copydependencies macro +Not sure this will make as we are trying to not add new features at this point +changed patch for better error as of r +Bulk move am trying to migrate from Solr to Solr and so I need to patch the branchWhen I use file on the devtags branch WAR file is built successfully but then it fails on loading with Error loading class message as if the collapsing functionality was not implementedShould I try using patch instead on sources? That doesnt feel right but maybe theyre compatible I dont know +Can you attach your test case +The workaround is to demote all atomicly updated fields to package protected and use proper privilege this to work the common jar will then require access permission +The difference in not utilizing the metadata entries would be in the naming of the tablesprocedures +And also when there are only one version of row exist no matter how many hfile you have a sequence scan operation will always need scan all the hfile row by row +r +Put it into commonsjcitrunkcoresrctest it into commonsjcitrunkcoresrctestresources +The download from is the right one +Now I know the file is reopened to make the power assert text and I guess the code to close them is missing +Note that the end state is reached twice and that inside the forkjoin the parentchild token relationship is empty +Tests is not added right +This issue has been caused by a dependency problem in an external module +updated patch with buffering in +Patch v starts in for running the snapshot export job +Your suggestion to wrap the ant script is a better approach +I think they should all be separate classes filters are a public API that clients can use to create queries whereas Iterators are an implementation detail +Looking at history of this file I see Kristian seems to take the approach of always using spaces for new lines even when the surrounding code is tabs +I clearly didnt keep around enough information on this +And lets document the change to the compatibility break wiki page. +Looks good +Showing the section with total can be useful +Like I said this bug is possible but in this case not likely +Apache Hadoop does not have anymore nor I believeNot sure how to fix that without fixing it from within Spark itself +If you finish it before we release beta please correct the version +Oh yes and the other design requirement we had was that messages not be on a fetch request +That is an excellent link Enis +Sorry for comment spammingI just made a test with maxbatchsize I add more messages to the queue after this startup error and reconnect the maxbatchsize is working +I thinkcassandra AMMakram Saleh kBIs the best +uncomment Foobar line saveNow the will be in a broken state +Oh this is already implemented with tests +This is also inline with the jdk whenever there are are methods that take a Collection as input. +I thought this functionality was removed from maybe the configuration is a straggler +And thats where SDN you raise a Zendesk issue with your customer account to get it fixed in Cypher Neoj can build a work around in SDN but that will be more expensive as it will have to include a filter for just allowing the official columns for each row to be visible to SDN. +version in your BND instruction to remove the version +Is it ok to put them on pdfbox homepage? Or is that too official? As an alternative we can put them on someones homepage on cant weAny ideas suggestions objections +Also are the supposed to be in thereIs the patch intended for commitCan you say something about what has to be done to push this issue to completionthanks +For some reason it didnt +Skipping for this feature. +Can we move this to for tracking purposes now + +To reenable replication you would have to go though all steps +Improved with r. +The latest JPOX build has fixed this issue +ManasaThis was failing while upgrading to So any source version is fine as long as you upgrade to You will not be able to reproduce this as this was some faulty dogfood cluster entry +Heres a patch for this +Coming up +Thered be a row per regionserver +Is intended to be assigned to Hui Wang? If not please consider assigning it to Mike Musgrove on the JBTM team who developed that part of JBTM originally +That sounds promising to me +Links and docs verified. +Seems there is a bug in Apache about handling concatenated decompression +Fixed on and trunk +This case should be closed but I cant find a way to do it +I could not get them to fail when run individually +we should crisply define the requirements of a session and design for that +Thanks MarkUpdated patch was applied at repo revision Please check that the patch was applied as you expected +See +This issue was fixed +issue is in around the + +committed +Ok this patch still keeps the cache in +Looking at the sld I cannot see how you get a non boolean out of lessThan +Clearly with the provided quickstart I cannot reproduce the problemIve to add which is missing in your quickstart to disable cookie and Ive played with the appThis is why I ask for more details +Thanks Anthony! Ive committed patches through +It is expected to be included in the binary when the driver is released +Whilst I agree that it is JSF implementation depending whether the header gets cleared or not I still recommend to check the response code before applying the filter in any case +But HDFS does not have hard links +Output of a run of the test showing deletions without potentially more interesting trick would be to only recrawl those documents that have a hopcount that is on the edge and whose hopcounts decline +Thanks ZhengI also looked at the default and capacity schedulers but the default scheduler already seems to have this logic as part of the patch for and the capacity scheduler doesnt try to do this kind of load balancing when there are fewer tasks than slots so I think this should be a separate JIRA. +Applying the margin to the Field and Label still leaves the Label floatingWhere do you expect the Fields labelStyle to be rendered +Patch submitted +Isnt the whole point to make the second thread skip the block completely? We only need one thread per instance to go through period +Note that when running under from jar files outside our test harness the drivers correctly autoloadThe new test is wired into the jdbc suite +Ill update the snapshot were using and see how thing work from there +May be it is possible to do the same for NB? Please +Marking as incomplete. +This ultimately comes down to a limitation in the way were currently using s and XA +One last comment it seems like none of the assert statements have error message +The original comment mentions the WHEN portion of a DSL def +After fixing this fannkuch runs correctly +I dont know if there is a better way to handle it in the XSD but I dont think so +Sounds to me like a bug in MINA we never receive any byte to decodeDamn it +In the CDI implementation it is possible to not specify any alternative in file that forms part of the JAR +Is this a duplicate of +Ive seen a lot of strange things so be assured someone will do this and if we can prevent doing this by picking a different solution we should go this route +I successfully generated the jbosswebDescriptor for +I however thought it would have been better to have had logError as a static method in the class rather than extending the class into the controllers +Attaching a patch for this issue +J attempted to use the OJB cache in the past but the effort failed +I have a patch ready for shared libraries and currently working on engines and bindingsIll attach all patches as soon as they are ready +so it will be difficult at bestI suggest this is closed as a differenceHowever you say this difference cause the junit test case fail in hadoop +There we have some auditing built into the framework to suppress duplicate messages that prevents this from happening as often theres no such support built into the NMS framework currently +Fixed by adding ASM to the main Available in OASIS SVN in revision definitely fixed in OASIS SVN revision + +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +resolving as fixed in trunk +bq +Ill take a look this looks like classloader lookup order issue +Unless there are objections I will commit it in a day or so +Linking this issue to a master issue for overhauling the table compression code +Here is where I am at +for both investigate further to check for options to retrieve the job from history and push for standardized access mechanismsFor now can we document this behavior the only concern I have is around the duration for which JT retains the job in memory +During this process they can all exchange the most recent view that they saw so that E realises the current view +I understand so basically we go back to your first approach where its the node who stores the destination vertex that spills to diskIt sounds like the simplest and most effective way to me +Colin let me know if you are free to work on version of this patch + needs to get redone without +Image manipulation should be moved off the device to somewhere else due to memory issues and the inability of Android to handle these images well with their Java API +the behavior patch has been applied. +Included in. +Could you please check the code in master to confirm that this is no longer an issue +emmanual can you have a look at this oneand Dennis please stop put in front of settings that has nothing to do with annotations it hurt my eyes and confuses my sleepy mind +So I agree with you its better to let the library deal with the default value itself +This exception is send when the property stored in the properties table is not the same one on batch side and on sonar server side +Thanks for the review Ted +But the patch the technique weve use there to wait to reach twice the normal phi convict threshold to hopefully make sure this wont happen +Zip of the project which you can use to reproduce the problemIt has been created in order to get this problem dont blame me for its poor quality +fix verified in thx +Furthermore the fix should be moved to jetty +can we not just output the label as title? seems to be what you are doing +None of the instance data in any of the static singletons is actually used or depended on by the methods of this class +If a developer chose to provide in a monolithic jar his code and used Eclipse to export a runnable jar while choosing to package these jars dependencies on the rsrc path were not found +Sorry forgot line numbersobj srcxercesc +Need to be fixed in too. +I reviewed the patch on and it looks good +Also expand on the description what it means when it is used etc + +SSL is indeed orthogonal to container authentication in the sense that you arent required to have it turned on +So this problem is a basic lack of permissions in the appropriate fileNote that this code has changed in Tuscany and later +For supporting new api in streaming the implementation involves two major tasks Setting job configuration for the streaming job set appropriate mapper and reducer depending on the arguments passed +slip to think this is relevant anymore. +In that way the full transition history is available in the SVN +Patchized version with latest modifications +Java files +It is due to undefined in +Testing it on wont work as the exception occurred only with Java which is not supported as mentioned above. +Yes you are right there was a space +Installed the library +Was this fix included in version +patch now including unit marked this minor by the way as its a relatively trivial fix +Thanks Yip +pdf file about misspellings grammatical errors and usage suggestions +It isnt a problem any more the latest already uses please upgrade Struts to latest version +Metamer was updated +Thanks a lot Robert and Ahmed for reviewing the patchI have updated the patch with the above changes +Thanks +Patch attached +I am not clear about the rationale behind which patch goes to which version +Linking to +Does that make sense +Simple patch that fixes the original patch to use +Looks fixed +John Perhaps you could come up with some basic specs for the methods I modified? It would help us mark this resolved and reduce the load for +SVN +My newest test proves that this problem has been fixed in +Applied the patch +What does this meanYou want to use something other than the simple mailer thing +Thanks Mahadev! +This patch should completely resolve the bug it is relativelystraightforward +Committed. +I wanted to just include this information to try to help the openjdk developers reproduce the bug and so on +The unit tests use these classes as well so I cannot use ANTRUN to delete them after the compile phase +Unused import Property +Reopen for backport +We nullify the split until we ensure that the is notGreat! Can you please upload the patch here so that it can be part apache source? +When this problem ocurs can you pl run it asbinhadoop dfsadmin metasave filenameThe specified filename will be created in the namenodes log directory +In such an environment I think there will be other problems +The +Resolving this one as Wont Fix but this is a good discussion +The assertion should report a warning for any exception that results in the attempt to resolve the namespace +This issue is not so important any more in my eyes because the xml config was split offDo we plainly want to remove the logj settings in out xml or make that pluggable somehow +any status on this +Beyond understanding the design you need a testing plan and test results to review +Or if you define the response element as an anonymous complex type then the problem goes away +Thanks Paul I just reached the same conclusion looking at shimsWill try to get patch ready again soon +changed a lot these two yearsMore should be removed from package +At best this is a minor issue that should be addressed during a filtering code refactor +not have it source other hbase scripts so this feature can be installed apart from an hbase install +log file for error messages? Could you try to close and open the connection does the problem disappear thenI tested against and there were some issues +Resolved at the openshift client library level +Running locally worked fine +I also had to add some extra excludes on the metastore Ashutosh I was waiting for the test result and as you predicted just got a build failed +This test gives another error logged as another JIRA issue. +Integration tests were updated. +I expand each dependent war into the target folder BEFORE the current projects war gets assembled +Overall I like this approach though very easy to use if youre already familiar with stress +Patches are updated +can you please file a jira for Process code +May be if you find a good project partner we can have two projects to go hand in hand +I believe the misunderstanding about rm flags is cleared up +If construction fails the server will be left up +needs to be as a new issue but this bug is fixed. +We should remove reliance on JDK internal classes +Tapestry should provide this component and let the developers choose whether to use it or notWhen all browsers support button then the styling of buttons can finally be varied as much as wanted +I will provide a patch +HiHeres an updated patch that would also check if wsdlLocation property is specified in the service endpoint interface if it is not specified in the service endpoint implementation class +The patch depends on which provide the Pig needs +Comment that JTA will be pure JTA soon + Test for the patch I will have a look at it and commit it into the the patch to revision +from matt Can we get access to the reports server in order to test data migration from the current IT schema to our new schemaLisa already has access +It would also make it easier to spot errors in the translation like the ones Knut has pointed out since you can easily view the English and the translated version next to each other +I experience the same suggest to change the priority to critical +A bug? Well no just a sensitivity +I made this modification to make the sample working +If I set Dgoalsdeploy and forego the site deployment everything works fine for me. + looks good to me +This command does what the issue requested. +Axis fails to validate its signature. +OK that test should cover my situation +I think that those vulnerabilities are just examples of how you could go about attacking Derby once you got your hands on a database +That is not strictly a technical consideration but is an important point +This should be fixed by r with one exception +batch transition resolvedfixed bugs to closedfixed +My upgrade uses ant but there was not currently a build file on the system and in fact the upgrade process build file did not need to exist after the upgrade +Trying patch with debug again +I contains a single Groovy file with static imports +But what happens when i arrived to max size poolProbably it is needed always use async mode to limit this problem +Deferring for now as this may be reverted from the spec +While we want to be able to make Groovy programs we dont want it to become too much like Perl +Really sorry about that! Do you know if there is some equivalent to devnull in cygwin? ie a file which is always guaranteed to exist and be empty? I wonder what happens if we just pass an empty parameter to the ant excludeFile tagIf this is getting in your way feel free to revert and we can commit a new version that works on both platforms +problem is that iw was build w Java a release now +Ive also been interested in this +Perhaps you can have a try +Patch committed as supplied +I havent used custom formatters in the shell before +So if we dont create a symlink user needs to update their script every time we upgrade a hive version which is not ideal +Eli we have not started the discussion on why was this jira closed +The fix is in SVN. +cant think of a test case for this +From the main method it looks like the whole job is frozenAfter youve killed it you have seen that the code was just syncing +Solves as it still needs to be ported to +Server deployer path should be changed to appupload by default or the bundled jbpm console should be adjusted accordingly +This is after clicking on search button +Ive got this working well against the utility that one would need to add a to their file +Patch committed in r. +Partial conversion patch for auto JarayamI think we need to rearrange Jira issues a bitThe issue with the titleMentoring Project for Derby Test and Fix for Jayaram SubramanianShould be filed under the Community Development Project and attach your proposalSo file a new issue under community developmentThere should be a new DERBY issue Convert to and atttach your patch there +some comments on iteration +lets use a new one to track the integrationThoughts? commit this to trunk +Fixed with in with maven repository and maven repository. +Thanks for cleaning this up +Its even used in the +Committed merge at subversion revision +I think it should be significant +Patch checked in to SVN trunk and branch +I also just spotted there was not a single entry related to the in the log +I uploaded a new patch although clearly you already have the fixed patch +Im fine with the patch +traziam eram ANALYZEDI am starting this PROJECT Minhou contribuio half and still speculativeIf ja Issus had been solved I Sorry +Thanks! +or none as at repeat demo +Found issue in GPS Tags are mixed Ill propose a patch soon +In this case the TCP RST does not go out so the DN side of the TCP connection never knows that the SNN rebooted +Do you archive any other files besides log files sayAFAIK no +or could also be reasonable targets +Is that for a specific reason? +Thanks +ok +Thanks for the review Bobby +Attached the META table migration code introduced by is highly suspect +Well add the tomcat to the zip distro for EWS +well I dont know I dont have a problem with adding this job actually +I think immediately turning around and asking for the next MAXEVENTS maps if we just received MAXEVENTS entries would be a straightforward way to eliminate the sleep penalty +Bulk close for +Changes made to address will prevent the file corruption issue noted here and future work to authenticate the based JMX will remove the root cause +Alex wrote Therefore no reason remains to keep the default DB at agree +good +I really hate discussing anything using this stupid tool +How are you counting test coverage +Applied my own patch check the new selectors in file. +so +Ive fixed the issue +I have them also in the lib folder of my +I propose we add the taskss working directory to its LDLIBRARYPATH then you could distribute your native libs with the and also have it symlink them into its cwd +The only trouble with Mimej that it is still API unstable and there is no hope of a release any time soonOleg +Im going to say lets get logj fully integrated and working well logj is common and the infrastructure changes we need to make that work are the same needed for logback or logj +x +Ill look at it once more +JDG is built on Infinispan +The tests will be added with +Try against Jenkins for kicks +We should rely on getting parallelism for large data sets by having multiple splits processed in parallel in different processes +it works anywayon Cayenne DBCP +Please try this patch with a one line change to see if it helps +This issue has already been committed +Brian bug has been fixed in Xalan Java +Attach a patch to add a shutdown hook which can shut down bookie server safely in following case the JVM is terminated +Wed also need to have some concept like Writable so the split can be sent over the wire +But Ill let you know if I find something +Just in the course of developing the unit tests in this patch I accidentally did that and was confused by the NPE +Ive tried to extract the text from the attached document and the exception is gone with version . +I had generally increased some timeouts by a factor of two to three +Mixed checksums in a should not be allowedIf we dont allow a different checksum for append then mixed checksums can occur only due to concat +I think but I cant be sure that I did this before doing the initial import +Youre getting the exception the watcher code just didnt execute fast enough so I think its +this issue as incomplete because nobody was able to provide a test case that shows how the Axis SAAJ implementation differs from the reference implementation. +py svnnative +Still I will try to discuss this with a few other people and see if there are any other options. +zipLet me run it now. +For now we can have the api inThoughtsMoreover If the initTasks is done asynchronously due to we wouldnt need this change +Please also read this issue is really important for many usersMaven has the philosophy convention over configuration so in such cases we should avoid ending up with workaround over usability +For destroy how about handling in? Then we dont need to repeat the code in and +No worries +attaching experimental patch for views +Sets Slfj as the default loggerErases the Loggable class +Committed revision Merged to and committed revision + push to +can be closed fixed with issue t reproduce on the latest build +Thanks Amareshwari +But maybe thats not such a bad thing +Option We can search in the connection pools to see if we can obtain a data source matching the database name +I presume that the now available type support will still be thereb Im not sure if I understood your question in detail +I will upload a fix patch later +That is why I have fq parameters with different qf listsDoes this make sense now? +This patch will make many parts concise +The latter case wont fe fixed by LOOPFOR right? I think AWAITREADY should depend on the system clock time rather than the libprocess Clock +Mikhail Ive checked the attached patch on my side iterations no failure +. +Ive committed this +Copyright year should be year of the last change of plugin as far as I remember +Once trunk is open there should be a good window of opportunity to try out those closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +The was based on a mode string so the simplest thing at the time was just to go with that +The bit of code that is there is there simply to get a handle on the Ant logger +Target framework that simplifies generics processing update spring factory and test classes +Gabriel Justin anyone else +It shows permission as +Patch for +Patch committed to trunk +Do you mind attaching it as a file? Thanks for the help Dragos +metadata +Patch is applied at +The problem we are seeing here is +Simple patch fixing the problem +Same for trunk +I think this happens in all components not just RM for instance jira is for the proxy doing the same thing +Thanks Devaraj and Ted! +This blocks also +The ease of typing is NOT the point it is the identity of the two participants in the operation +This convention works as advertised. +Same applies to a web service with no some preliminary work but is not finished +Looks good +The links to the unzip and copy methods are preserved but I suspect most people just runs away when they see here is a doc patch for your for or also for +Here is the fix +If you dont know exactly what size you want then you can ask for a larger container than needed +Is this still an issue in? +So this problem is more constrained than I originally reported +Moved to upon release of +The pointer of the ticket for Lucene is missing +Whats going on here is this the changes inside the source distributionDoes smokeTester verify only in the binary distribution? or is it just not checking well enough +DavidI definitely like your proposal on making this more scalable and +An alternative would be to just pick this less often in see the hack Another option might be to make the pick a ludicrously small Bitset sizing option for each field so that we can accommodate tests that create silly numbers of fields +Hi FlorentCan you please test the following patchPlease note that this also requires the patch from as well! Otherwise the correct path will not be found +Looking at the attachment from the issue reporter it seems that this happens from our Ant task for Jing +Closing issue as this is fixed by Ralph Goers added a comment Nov +Thanks Kim! This is extremely helpful +Shortened the summary line in the release note in order to make the release notes more readable. +Committed to trunk with revision Merged to branch and committed with revision +Also it sounds like this always invalidates on update +Thanks for the patch Jon. +Should affect both as well as bulkdocs +and also the changes in tests make sense after I discussed it offline with dhruba +This triggers finishLoad +This will be in the next update +Ive fixed it now with adapting the files. +I found this site http which says they are needed otherwise it is doing a contains rather than matches type checkHaving said all that I think this is still a valid change +The getScrolling function will be rendered when the fview tag is closed afaik +So that we are just changing the oreder of read trails in quorum bookies for entries by keeping track of the failed bookies in a faultBookie list +Just made a commit to fix the branding problem in some of the other features that is visible in the for the x branch doesnt include these two features so Im not sure I can simply remove them from the site and have them be foundTesting +Ivan could not reproduce this on the latest build so closing. +bq +See +The parent class method allows Exception so it seems like an oversight that the declares +if were breaking BW why not add the private ctor? Im sure that when will be out that wont be the only BW +Never followed up resolving as incomplete. +Adding file +MarttiThere is no mention of copyright in the code +it was resolved by fixing issue +Can you please let us know what flag you are talking about? Is it javafalse that used to be in +Please attach a Postgres DDL for this sample +I dont see much value in all idiosyncrasies and features of unix tar in Java +Pick the first centroid randomly +What is causing the consumer to closeAre you explicitly closing the consumer or is it from session close or connection failure +The added patch to the current trunk fixes this +For me this small patch makes the sam ple code work as expected +The deadlock appears because is called while holding the listLatch in +now keeps track of the updateModel settings so that it can be swapped out if user later changes their selection +Also need to implement the queryAll operation +Jerome i recently made some significant performance improvements to could you please try the latest nightly build and see if it is faster +Thanks Yaron +Heiko can you comment on this issue +Oops nevermind I botched the mspec run +ClebertWhat you have suggested looks OK to me +looking +Thanks for clarifying. +The second time was from a local cluster +Any id propertyfield is mapped to the id field in the document that is stored in +I tried both listing all the files and using the and didnt see any difference +Add path of Axis CPP folder to PATH variableBut we didnt resolve problem +Patch applied +That is not a valid response according to RFC however it might be worth adding the checkWhat is your server typeAlso if your FTP server supports directory names with spaces in them how does it report PWD for themNote according to RFC embedded quotes need to be this is not done eitherFor example MKD foobar usrdmfoobar directory createdIf we assume that the trailing comment never contains a then the code can handle embedded quotes by looking for the last quote and then the +in the breadcrumbs when adding a resource to the list of when adding a resource to the list for which alerts should be displayed on the missing properties to file will be in next code drop +The question is how much longer and how more often +Thanks Suresh +What does sstablejson sayWhat happens if you switch from mmap io to bufferedWhat happens if you run scrub +First cut of the patch +If you can wait till then I will include this patch in my patchThanksOk! I did not get that Of course add a comment to the issue to tell the projectBest regardsNicolas +I had forgotten about that question +or later +A test case for this has been committed into Qpid trunk at rev The test case has basic tests to check for trailing spaces when defining queue topics and topics with multiple binding keys. +Many thanks! +Fixed in beafedefaffdee +What exactly did you fix? +Would it be possible to move all the servers to java but keep the on java +interesting +great thanks! can you please link related new issues with this one? TIA! +Should we release note it or update the POM +I just tested and here are a couple of pointers +is this also affects eap ? should it be documented in the eap release notes +Normally having Maven configuration outside of the pom is not the Maven way +Anyways there is few method contents that we can outI am currently working on a sample +Were not talking about a single use case +It could be made configurable in Tika anywayPerhaps we should deliver a patch for Tika +If the configured value is empty then the authorization must fail +I agree too +So the solution for avoid this is as suggested surround submit call with a try catch +Thanks for the submission Chris! +Not sure if crawling is fully working I am using Null output and null authority at the moment +committed +Whoa +Just wanting to make sure it isnt waiting on us or something that can be closed +Please install the extension feature if you want to run Java +Please use an Apache httpd server as my colleague has advised +Function metadataneeds to contain the location of the jar +Thanks Dhruba. +ok seems to be a bug in woodstox as the RI doesnt write xd for the r elements +Not all fields have indexed docvaluesbq +I confirmed the fix +TODO given the discussion above and the unresolved SDO spec JIRA targeteed for the SDO spec the assertions in the test case relating to the types mentioned above in the description should be removed. +I dont know if there is any other problem or something is you in advance. +I add a judgement if file give error message then cant go on the provide steps how to verify this on JBT CR +Still ready to commit though +This bug has been marked as a duplicate of +Tim thanks all works fine with with latest sources +Fix should be available in ER +Freddy Im not sure that this will be possible to implement due to the same reasons as some Maven modules may be empty or contain only test files +HiPut the you build yourself in KARAFHOMEdatatmp folder to replace the old one this your will take effect +There is a support release being discussed on the dev list +How did they fail? As far as I know the package name is still valid Java so it should not failChanging the generated package names is covered in the FAQ under General Questions Can I change the default package names for the java classes generated from my schemaWhere would you rather see it documented? +Will commit if tests pass +This is caused by +No problem Im closing this JIRA then +We have to do performance validation of changes like this + of the other are merged in +This will be a known issue for GA and we should recommend users to use zipped deployment for remote server not only because of this remaining issue but also because of the poor performance of remote exploded deployments it is very slow +Its the latter case and waiting the cxf fix be picked up +Many guides have been no longer exists and many new ones have come along +Added some page flow classes for testing this bug +I think its worth calling out the things we know at the time of the release as well +Are there going to be any additional testcases to test the JNDI functionality that is being addedThis would require JEE environment to testIf someone wants to use this code contribution allows users in a JEE environmentto specify the JNDI name for the connection pool and invoke the JNDI methods toconnect to the pool +I started with +The patch has no context nor tests so not sure what it is you want me to do +I guess kill kill QUIT should work too but I dont know where the output goes +Resolved in CR + +Im not sure if this is exactly what youre looking for but we can adjust as necessary +Not but has been running in production +nice quick turnaround davei scanned the code quickly and noticed that you removed the component from thethat simplifies things a bit but at a glance it seems like it might introduce an issue with a provider that serves up multiple resources +HelloWould you please try my patchThanksBest regards Andrew Changes committed in revision +Pushed to +Fixed in trunk +I do remember XML namespaces the different handling of xsdany and the empty namespace across different XML parsers other things that dont bring back fond memories. +For what concerns the EOL Im aware of that and it has postponed +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Im still seeing the NPE in when using this fix +I think being able to specify the schema documents is a good medium term solution at the very least. +Verified that this does NOT occur for so looks like a regression between and probably related to major rewrite of property discovery process +Thats the reason i assigned the transactionboundxaresource the newly created xaresource from the new created session +There was a bunch more code using mARF in mainly from forwards +patch allows to create keyspace without being authorized to one fixes describe +Committed revision . +Andy I think seems reasonable level for it +This fixes the problemI tried to make a unit test but it is really hard to make it deterministic +I just committed this +well no wonder no one did anything with this the code is overly complicated and even the minimal test case provided doesnt work +Thanks! +Still need to add a new CTF test case +issue resolution verified by submitter. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +If you set the war to be an isolated deployment then this will not longer be an issue as the war will only modify its own copy of the bean +is that I applied workaround described in but issue from still existsthe word here referred to +address this requirement the number of records read from each user input and written to each user output in a script will be written to the Pig log at the end of execution. +Heres the patch +Note A temporary workaround for that will match default behavior is to set the upload directory path on The Admin Configuration page to the absolute path of the resources directory within the deployed Roller webapp +With ER and ER this didnt appear +Updated patch that adds the requires to the files as well +JarekCan you post this on with group flume so we can get it reviewed and in +Forgot a bit about this issue +I think thatpages may only when the database is checkpointed or shut downAnd to cause a page to be read you may need to shut down andrestart since the page you wrote remains in the cache + +Patch +It should sort in alphabetical order only regardless of the case +Olga can you make the feedback image smaller? Actually you should know that the display window size could be x px +Bulk move issues to and +Maybe there is a problem with Oracle and commentsRegards looks to me like a oracle driver or db bugI can reproduce it easily and it only fails with sql comments onThe failing sql isload select as category as category as category from category categoryda where and its totally legit and the single parameter is setted by hibernate +Ted In current implementation duplicated filenames will lead to duplicated input data +Cant see the failures that you and Neha are seeing +Im testing deleting the temp dir in finally +Should I make an issue to add seeking? Wouldnt save us much but does look like seek to the correct offset in the CRC file is suboptimal when random reading +Carl I agree the comment on list bucketing +This was probably overlooked +Ill check tomorrow +I have the same issue with glassfish urroller on immediate commentsThere isnt even a hint of a problem in the logs but comment authentication fundamentally does not workHowever it did work once which is even more of a mystery +Funny was just looking at the same thing +Sounds great will do +not sure if this is intended to be fixed in or but dropping in here for a trimmed down POM to reproduce +Unless youre doing a collection all properties should be of type String +of docscurrent +Trivial patch suppose this is for the branch. +Defer all unresolved issues from to +No problem Andrei +Tested the patch with and testsuites +Joachim I should have looked more careful as well and are completely different things ofcourse one is put in front of the otherThe new patch looks fine by me and it doesnt break anything afaics +Moved to so Mark can look at it +Committed revision +Our local test cases always fail I will apply Teds patch and run all testcases +In this case the session closure is infact a serious problem with the connection +If there is an exception or a timeout the UI should show the problem + +I edited the comment after Uwe pointed out to me this was an authorization errorDescription shows the error on my mac this has something to do with parentI dont know why it works on my linux box +In one of these the swimlane is to be deleted but since the not deleted task still references it a constraint violation is solves this is explicitly setting the swimlane to null in the execute method of the +I thought we had decided as a PMC to use etc +Fixed at resolved issues to so we can generate release notes +includerw New header file to enable autolinking feature +I see some tests are failing on the CI job for Trunk +The algorithm which we use to check for the applications to get the scheduling opportunity is based on heart beat from the NM hence if we just use the number of racks it will not be much of the value add for actually to achieve the node locality +Wed like to add the for the ugi and set it as the login user because in addition to the register operation the following operations may require the ugi and its as well such as makeRequestTherefore IMHO the current patch is doing the correct thing +Bruce any comments re the proposed solution offered +Sure let me work on that and resubmit! was able to my fixes against trunk and add a test to the grouping tests +What still needs to be done update and delete of metadata properties in web UI and call backs to action class selenium tests for the metadata for adding additional info to the metadata I think we can start with just allowing addition of new mailing lists dependencies licensesIm currently working on this and Im planning to use a for adding the data when an Add or link is clicked beside the section name for one of the above properties +I would go with +Was a JVM issue fixed by moving to Java. +Sorry I cant reproduce this without more info +Could be backported +Why pass in scannerRow and startRow? When would they ever differSee in +Delivered and validated. +Hence my need for an alias on a list name +Attached a thrift system test which fails until this functionality is complete +have committed this. +in keys is +The fix was not included in beta release it will be in beta +Im putting a Fix Version of on this but that should be regarded as very tentative its mostly just a reminder to me to think about it in FebMar +I think my best was in a row on target +IMO it should be fine having us depend on an LGPL at compile time as long as the dependency was pulled by maven at compile time +Bawany I can provide more information about this +looking into it +I added a control per server the client cannot have more than X request on a same server +Whats going on with this issue +It should record that in the AST so that it doesnt have to be maintained by other parts of the system +javatestingorgapachederbyTestingfunctionTeststestsjdbcapiIf a clob is accessed after commit or after free gets a of XJ +no tests added as only deprecated class and methods were removed +I cant duplicate and submitter of bug cannot duplicate anymore. +Attaching cropped server log +id not recommend upload a war as there isnt one in the official zip filethankst the webappsaxis directory the equivalent of a war file? Deploying it is as such is mentioned in the users guide so it seems like a very standard procedureI embed Axis inside of my ear file and Id imagine that a lot of others do the same and including this war would be a great help +We also need to consider if we need to produce two otherwise identical modules in a build one compatible and one for the users +twitter accounts + +A patch is now available for review +Is there any more work to be done here +Setting these to Not Yet Documented since they required updating from known to resolved issues +I dont think this will make at this point so Ive set the target to backlog but contributions are welcome +Is this what you experience as well +See. +But I thought your idea of use total binary input would be really helpful in real production mode since the SGD algorithm is really blazing fast any performance improvement in the feature parse hashing encoding would improve the overall performance a lot. +Heres a new version of the patch that solves also the issues +is in usrsharehadoopconf and etcdefaults is symlinked to conf +Attached simple test on EAP ERRN contains expected text + to trunk +To prevent accident I added a comment not much else possible to do +Im unable to find magicGball sample? Where did you find it +Hi MartinI dont have rights to attach files to the issueregards should have the rightsEvery user can attach files +Im not too familiar with these changes +Should we blanket disable them now and then reevaluate once the framework code has been updated to use only New API objects? is shared by common hdfs and mapreduce +done +It fails on Mac OS X if run alone If run after or is works correctly +Done. +schema is now set to ignore system rows when version is calculated and when migrations are to be send to the remote node +Also I think we should have separate jiras to track That the memory channel does seem to respect the capacity That the sequence file source does not seem to stop on reconfiguration that disables it +Thanks for the suggestionJuergen +Unable to repro against the branch or even itself with a node cluster +CommittingModifying titleJust as a workaround for people not having this patch they may use Base codecs to get their delimiters in and out from the XML +Then we can discuss how to move ahead +Hm applies for me +People just need to add a reference to whatever implementation they want to use +I am confident it is safe to be used still Im not sure if it is really fasterI would use it for myself as I believe on theoretical basis it should be faster but still didnt manage to proof itIf you think I could get more feedback by enabling it in Beta i will document it but are we still having the option to leave it out in ga after being in beta +Zuku steel is known for its malleability and tensile strengthCheersZubin +Modification applied. +First draft of interface for query objects +moved to target added some tests to show that this got fixed in the meantime +Comment is removed +I think you can resolve the ticket since we have an initial version commited +I did some regression testing with your patch for the B part +I did note the offending was packaged in Eclipses but not in Eclipse +Just got back to this patch +Does not inlcude test +Verified on. +It relies solely on binsh and it works +All required bundles are present into the SMX OBRPlease reopen the bug if new bundles are required +Also if you take a closer look the maven repo you have provided the link to you will notice that it does actually not contain the driver jar file at all so unfortunately I dont think this will be saw the mds and just assumed the jar was in there +Oh and also the client failure is the same with both MRI and clients +Open server editor and change deployment to Workspace metadata save +I committed everything but the step since alone takes K +Tests to launchpadcontent in revision link on httplocalhost gives access to themNote that currently running these tests requires setting Allow Anonymous Access to false on the Sling Authentication Filter configuration page Ill open another issue for that. +Second patch applied in r +Also I understand your point about a BQ is a Q but I dont believe client apps would use BQ in that way +using could help you +All tests pass +Im eager for news about this did you publish some numbers or any thoughts to share +This fixes Owens offline comments about having a finite limit on the split meta info that the reads +Assertion fails +AncaYeah I know +Please submit a correct one +Also my previous comment incorrectly stated that there are two test cases added by the file +Was there a specific reason for going straight to the tablet serverI could have just written the entire test in but it seemed reasonable to test the thrift service separately just in case something small changes in that layer it would be easier to isolate any problems that waybq +patch committed as well +the script is out of dateI just fix it +Saul I hear rumors this may be done for very soon? We should get out very very soon so if youve got stuff then get them in soon +Sorry no Im eager to but the priority is to release beta +When you run a Grep if you apply the same slot numbers on both cluster I guess the processing times are similar +I would vote for option above +Thank you Alexei +Have refactored sufficiently to avoid most code duplication +I think one giant key value store is a good first step but realistically you need different permissions based on if you are making a change intra cluster inter cluster for one box for one user or for one job +As for why the value is called a threshold its because this is the threshold below which the item is retained in memory and above which it is written to disk. +The rd Ref Guide patch looks good and was committedCommitted revision +Thanks Mridul Jacopo Scott for the collaborative contributionYour changes are in trunk at rI am keeping this issue opened for next few days so other can comment on thisMridul or I will take care of closing this issue soonThanksAshish for comment about closing Ashish +made the dry run the default added some more input validations for the tool some renaming to the controller methodsSwapnil +Using works fine and will resolve as expected +Steves comments seem to imply that we need the capability of launch the health checking script in response to external commands +Alright committed thanks +I dont understand youre saying that you have in the parent and then in the child and is selected? Thats the expected is a dupe but Im not sure what the other issue ID is +Thank you Robert for the explanationIn this specific case its hard to understand the differences between hunspell and Lucene since Lucene doesnt even parse the affix fileIve been in contact with the authors of those Ducth dictionaries as well as with the hunspell author +IMHO it is extremly important to get this into maven asap and test and improve +Plugins and trunk changes to allow enums within converter metadata +Galder The feature that I want is to create an evicting cache store but have no defined upper limit on the number of entries +Apache Directory Studio has now been releasedLets clean Jira and close this issue. +I just committed this. +removing hcatalog from thriftif build Travis for review +if there is a Page class then it is reachable at httplocalhostwicketbookmarkableIt has a bug currently that it tries to create a page even for classes which do not extend Page +Surely +Thanks guys! Heres one minor change The last bullet point of the PMC responsibilities should really be a paragraphIll upload this to our staging site so that we have a fixed revision we can vote on +Files appear to load just finePlaying with it a bit more to ensure this is kosher +removed the enforcement of a destination on policy entry not required for default entries +and Cli all use +I had previously understood that what you wanted was a form with a submit button and a back button +Im closing it then. +Not sure when it was fixed but its fixed. +Will the proposed metadata service subsume those attributes as well? Curious to see the proposed designThanks lower level of metadata we dont mean storing information already present in the namenode +Dave On Dec at PM Allen Gilliland wrote ugh i just realized something icky +I just committed this. +Verified by Leo. +Looks great Andrew +The attached patch removes the field +Empty are handled ok on the server side +I was thinking to install the maven and jdk in the cygwin +I guess it should be documented properly +I was looking at your patch and dont understand how the lastconnectindex works +Any news on this issue? its nearly week old and we really need it for start coordinating the project +No the configuration parameters are not different +Now finally I see the solution has been committed in a SNAPSHOT but I cant find it in any repo +I cant state something explicitly till the time I see your project jars in your project classpath +borders should be taken into account on if the property is set to collapse on table +Fixed. +of. +No further reports since last release moving down to. +Thanks Jian! +OK so next question is want to create a patch? Should be a simple matter of changing the Selector class +Thank you for the detailed information +included in release +Reviewed looks good +Yes Michal if theres no httpMethod annotation present will treat the call as locator Thats the design currently and it will the unclear log like this one +It should also be applied to remotingGA and remotingx +The amq ssl Context maintains a thread local for the jsse context +Sorry. +thanks Ben for the patch +Once upon a time I think we used +Uploading a version of the basic patch which addresses Knuts comments +Nice to have but moving to to get in bbebdacfbbaffe on master and in cdbdcacbafcaeaecedcc on +Chris or WillPlease remove synchronized key word from s render method +Ill look into this +How was it dealt with in MR +right its basically the schwartzian transform pattern +Removed noop opcode which I had introduced for testing another thing +Authentication changed +Tammo I didnt wanted to change the method signature by throwing exception throwing exception and catching would have been ideal but that would have required some more changes I took a conservative approach to code change +In the same way to serialize properties we use algorithm implemented in if it would not save as but as uE we should not improve it to generate invalid byte sequencesAll we can do and are doing now is to check that some property has not been touched in UI editor and so may be left as is +Max which run asdebug as do you mean? I can see it on toolbar and in popup menu +In windows I didnt get it to work neither in trunk branch previous versions double quote single quote +This has not been fixed +New patch with and testUsing the did not work because although it is a list it does not keep track of the tail of the list +Its been more than a year and no one has added anything + +So your feedback for different issues should be directed to the proposal you are commenting on +how does code access the types names in the multiset +somehow records are not getting saved in child audit closing stale resolved issues +Updated patch to resolve conflicts after +We should remove this +Sometimes you get and sometimes not +Im not sure we need to make a to fix this but that might be a good idea +Raphael is this actually an error with the tapestry archetype? +Check whether fsimage exists before return +LGTM committed. +Please never reindent code when submitting a contribution +Essentially if we have a portlet request and if the is the one provided by the +Right so you cant add them dynamically once is running +How about improving documentation around gcgraceseconds for when it makes sense to increasedecrease the default +Colt remains within this codebase +I also added two unit tests for the scenario by reusing some existing test methods +This happened because when a bg popup is created from Hosts level the Do not show this dialog +derbyall passed tooInitial perf numbers seems to be comparable with the previous ones when using the back endIll start some longer runs tomorrow with the back end and upload the test code +We really need a test case for us to reproduce the problem +more css fixes to address wrapping div containers +Reopened for trunk +Uploaded a patch +This allows us to use the dist to generate the source code in a Maven standard way but also use this extension to download the code onto the local hope this gets into the plugin soon + +Thanks. +Cannot even get people to maintain their list of Mentors +bulk defer to defer to +No +Thanks Namit +This has the net effect of the described symptoms in this issue +Also I am reading the code and I think it would be cleaner to stop the subprocess in a annotated static method +I admit I only ran the build tests some empirical crawls and other types of jobs would be more then desirable but I dont have the infrastructure to do it +Hi LionelSeveral changes have recently been added which should help keep memory levels down in large connection scenariosr Producers were holding on to large k buffers needlesslyr keep it simple +We will fix bugs in the core classes since theyre all our own code anyway but we dont really want to maintain a whole bunch of patches for stdlib +All the regression tests ran cleanly with the patch so Im setting Patch Available +Patch for branchesBranch closing stale resolved issues +Thanks for this suggestion +Uploading +Volunteers welcomeTo some degree it doesnt matter where its hosted +We definitely want to fix it so if you are able to provide instructions on how to run tomcat as a service it shouldnt be difficult to find the problem +Whatever solution we go with I dont think it should be to grant get permission for all the jars for every test we run +Heres the text from the docuse only all set values in this config will need a lot of properties but allow to not set some valuesDoes it mean thisUse only properties set explicitly in the configuration ignoring default values +I understand the rationale for overloading addData here but the actual parameter is another instance not just data +Actually that would be really nice +Youre right Alphonse +The bug was some internal but basically it was adding all such classpath jars rather than just those specifically designated as going in the ear +I havent got a test environment handy at the moment but I did look over the code and it looks like the caption style was not being written and the data table header and footer styles were not being persisted but it looks like headerClass should be persisted and the two attributes just never rendered? Or are they just skipped after postback +looks good thanx ivan +Simple fix which just fixes nodetool to close the connection which it creates +I also adjusted the reducer and test logic accordingly +This sadly is a big pain since one has to setup the dependecies then eveytime the IDE is restarted +I had a very quick look at the patch and have one initial comment + thats works +Qian can you try it out as soon as possible +looks like a problem in the expression escaping support of the parameter expression evaluator +I just committed this +But it seems to be reasonable to hide such targets away +Committed Fri Jun PST +Jon describe the benchmark methodology +Accidentally added in here +this to trunk and since this is a bug fix +Integrated to branch and TRUNKThanks for the patch Brock +But we know a lot about docids and extra hashing should just lead to an slowdownOk. +Pull back in if the new bugs that master rewrite has introduced require this described mechanism +theres an apparent bug in the treatment of a directory +all newupdated markers should be removed from JBDS docs +If were actually checking that the script exists with success then it seems like the option is misnamed should be more like or something like that +Yes you are right the inheritance mechanism seems to work now +I dont do C development and I dont have time right now to ramp up to make a meaningful contribution +Can you create a patch against the trunk please +There are still pages that are coming from the UI +Modules and are flex and DTP Profiel v is a flex profile +Patch committedOleg +Thats also a possibility with the builder approach as well the developer might pass null values by mistakeBut like I said you can go with whatever you prefer +This reminds me that starting a package to gather these rpm snippets as macros could be quite useful. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Makes Transport and Protocol configurable through properties file +Likewise if there were two more updates followed by a refresh then the current revision directory would be since there would have been three updates and two refreshes total +Thanks! patch is now in +Is there any prolem with this +The attached should implement this for bundles +for the patch I meant +Otherwise this is an excellent start +Rick Please review and apply this fix if you agree with it +The main problem is The universal possinility to use is the fact that you only have to define the swap operation and nothing more +User want edit page in source part of VPE but want see visual part without current visual artifacts +Im not sure why I didnt notice I did after trying to rebuild it this morning +Ref language is not in IDE +I would really appreciate this feature +Two notes I it to as it was a surgically small change and Im hoping we dont need to regenerate all the thrift classes +Sorry for the delay +New version of the spec and updated jersey implementationAs before unpack the archive in the srccontrib directory +Can you comment on this jira Kurt Is this still an issue? If so we will add to the Release in ER +I am certain that it doesnt happen under postgresql +Ive committed a partial fix to the native code in r +I observed it with but not with we create a jira to track this issue? I can do that if you dont mind +Ditto on deserialize +Attaching a first rev of the release notes +closed after releasing version +If I could get some reviewers to make sure I dont need to handle other broken operating systems that would be helpful +jar or you name itso I need only name pattern for installers without EAP and name new section in Build Results page +Mark would you please take a look at this? Thanks Ben +Can you please attach a sample XML document instance that matches the XML schema given? Id like to add a new test case to the CTF suite as a result of this bug + to trunk +Hi thanks for the JIRA report and identifying the issue this has already been fixed by +was committed to +K let me know when its ready for review again +targeting a version if no response or we cant duplicate in due time well have to reject or reopen this jira if you can provide more details. +Patch for german for english for content. +I agree since it appears that all handlers are blocked this could be a deadlock +do you happen to know what value causes a problem maybe the routine in derby could workaround the jvm problem +It is a special type of mapjoinIt will end up overwriting all the functions of but keeping it this way keeps the hierarchy +Hi Venkat I tried this this current trunk but it worked fine +OK Ive moved all the old files over to the new repos +Hi Oliver +Initial version committed on branch +I knew about the escaping but I tried with single quotes +num should be final in Ok +Let me get back to you with something in that direction +It looks like a bug +added missing in revision . +properties of template created form with latest build and still see the same result +a Declare as Spring Bean Action in the context menu of a Class + +cache attribute for tag +The logic would only be invalid if disclosed was already set to true on the being expanded +Any fix must be careful not to regress +the HWI stuff does not compile against and prior because of the change to using new jetty apis +I committed this to and above +click on profile link +Good point +It is hard to replace the dictionary implementation because everywhere is using instead of the interface +Both are used internally so there should be no compatibility concerns +Curious when it decides to do that vs merging w L +Can you add details on which commitjira merged out the trunk change +Updated CHANGELOG with Features Improvements Stories Bug in +Will also need to go into aka +This patch fixes the problemThe path will be cut after the first colon instead after the fix value +Thanks Jarcec for your contribution +Fix for operator precedence fix is in SVN +changes Added a way to connect to other LDAP servers and not just the embedded Apache DS +What exactly is the problem +Im a bit out of my depth here but I wonder if its more reasonable to examine the eigenvalues in sorted order and examine ratio of one to the next +Hmm OK so now we now the bug is upstream its good to guess the next step would be to know if the bug is in Compass lets hope Shay will be reactive and investigate soon +Im not using inconsistent versions of DN jars the reporter of the issue is +only moveMessages with filter was doing it wouldnt consider this a Critical issue but its fixed on nightly build with. +Confirming successful subscription to users +This stuff has been out there long enough that we want to weigh the pros and cons of a hard break +Using JBMs own security manager we can easily combine the two into one +Would have it run if it was under? If so we should move it +Committed in rev +Any progress on getting this patch into a release? I can take a look if nobody else is +Youll need to be more specific about your environment + is probably the same thing as +Wed like to be a little more friendly to the sender +I dont use SAX parser so I dont use +Thanks for patch Elliott. +Committed to trunk and +Werner was working with someone to have them create a test servlet and access the +Its ok when Server is not running +It appears that the implementation creates a new instance to handle every invocation rather than reuse it +I kept the old to support the old API but deprecated it and its test +Good idea Mike +I agree that we should keep this ticket open now that the security vulnerability has been emphasized +I think adding a requirement that the NN upgrade before the is quite inconvenient for operators +Youll have to plumb the prefix through from somehow to get the proper error message +To clarify what this patch does on the first round behavior is unchanged it will fall back to the subsnitch as usual +changes look good to me +Mmhh wouldnt be better to get MAP as default if not set +I agree Marshall +Maybe you can try the three patch files that I put on that see if they work for you +I prefer ping because it is shorter +This fault lies in with tests. +Attachment has been added with description patch +This one is mainly based on +I will have a look with Mint that I have installed as dual bootId be interested to know if anybody is reproducing the same +Still testing evaluating performance and verifying correctnessThis modifies the existing sort as follows +minor title enhancements release +Yes it seems to be duplicate of +I am attaching the patch containing the modifications as per my last comment +I simply changed the help text for now. +They should point to the corresponding implementations in Axis +Marking this resolved +Thanks! +Can you please elaborate on I am getting connection object from a third party + +It is up to the processor to provide the proper object coersion or to simply pass the object on for use in an extensionXSLTC is simply not attempting to perform object coercion in this case +In Infinispan this should be the other way around it should go in the externalizer class cos the end user might not be able to modify the class thats being serialized cos it belongs to a rd party library or its closed sourceSo supporting both method calls from some abstractions to be created which Im working on +Now i am getting your point and it takes sense +This issue still occurs if KC is redeployed to AS many times could be related to issue still happens even with factory being closed properly now that is fixed +Yes it is fine +Marking as resolved. +Installed m plugin via update project enabled m on package goalno problemsPlease provide detailed steps to reproduce problem +We may be able to get a nice boost by moving the watch call outside of the synchronization on regions +Im not able change any value or use menu option since Im getting different or or NPEIt is at least not how I should get according to user manual +Hari It could be an option to shutdown however Id rather not get a call at AM +Committed revision . +Here is a new patch with spaces but some files have been affected by the autoformatting tool in Eclipse +Note that the DN uses a separate thread for every transceiver during its normal operation and the default value for the maximum number of these is +Leif Is this not wrapped up now? ISTR a DNS entry for +in +Applied to trunk Ok with some import order fixes and manual mergesHowever wont apply to branch as there are quite a number of previous patches missing that this depends on in the branchThis patch fixes the compilation errors for that are in trunk atm +The problem is that the short read protection hasnt been updated after the changed made by namely the fact that now group columns by prefix before counting themAttaching a patch that fixes that +As long as code is licensed to the ASF for inclusion in ASF works it can be copied from project A to project B as long as both A and B are licensed under the same ASL right +Jan please update this issue with more information about how to reproduce. +currently recover code path is different and only executed if recovery is supported +It is going to be fixed after we switched to build to assemble JBDS isnatllation based on definition +Transparent image to be placed in srcmainresources +Check if Driver attribute has a value +Simple patch to to fix applied +Thanks Adrian for your review please have a look at modified patch +Verified by Leviev. +Please see v thanks +Youre right +Verified in +I wish to reproduce the problem and do a step by step analysis +If someone else sees it they can with the appropriate information so we can reproduce it. +pat since the test passes now and is fixed do you think we need another test case or should this suffice +Will create new umbrella issue for +I cannot reproduce this Access to request parameters in an EL expression works nicely for me both standalone and as parameter for the springmessage tag +Resolving again after his application a while back. + +and put is highly likely to fail because current retries are no more than sec +Now the sources have changed so much that i am now having difficulties to reproduce the problem +to address this issue in smx we need cxf jms transport continuation support throttle FreemanThanks for your but I will not be able to run these tests before the mid of May as I will be out of the office until then from tomorrow onwards +I will attach couple of simon graphs that show what was going on during this time +I also run sysinfo manually for the Tools guide example which takes next to no timeSetting one or more of the properties is a bit too advanced for Getting Started or the introductory topics in the Developers Guide +Commit the be backing this out +Mircea TCK requires JDK to be run +After this change the class doesnt work for typec fonts +Thus all above changes are not relevant to this issueSorry for the confusionThanks +Great +works fine. +still an issue in numbers only seem to apply when running the client in jirb +Patch for the unknown models row this patch now +This change has been introduced in +while i agree that it is possible to invalidate the cache manually and I am doing it now it just seems more logical to make it the responsibility of the container since the problem relates to syncing the container cache with the db +Attached a wrong log +Though there is a slight inconvenience of creating a separate class i guess it should be fine +This works on the desktop page and on the nd click of the button on the mobile pageLooking at the ajax requestresponse comparing what works to what doesnt work we can see that in the case where the submission does not work the is missing from the submitted form data +Last commit should make it shutdown +There are a couple of bugs to nail down and the UI needs to be modified to properly display the deployments +Patch applied rev thanks joe! +if so I can wait +trunknetuitestwebappsdrttestRecordertestsLet me know if there are any issues +I committed the patch in revision +For example I booked A and get one promotion item B +This should be rectified by an alternate design +Rescheduling to implementation has been checked in +Ops folks will need tools to edit the fsimageedits when things go wrong if they are stored as compressed content +Got it thanks +The test can be improved but is enough to illustrate the issue +Found a great new scala testing library so Ill want to try this out. +Hi Jacques thanks for testing +resources and this was working fine in last weeks SNAPSHOT. +The norwegian translated thank you. +another part is keeping views up to date asynchronously so clients can use staleok all the time +definitely must change changes our jar libs +Ill try it out +this was attached so we wont need ICLA in +Added another bean to help simplify converting a Cache into a +create +So that I dont know if I break some other things +The fix is in the +eclipseplugins +Hi AndreasCould you try to provide a test case for itMight be a good first step to get it fixedThanks +This is a common problem with certain project setups that a bindiagnoseproject will help solve +If used as suggested it should be helpful is what I feel +I still think some of octopus would be great for like circle +I just committed this +Thats correct but for some reason it is not working with GEP +FlaviocheckOpen is introduced in +This is a known issue Incomplete use of nested generics against Spring may break with the fully enforced use of generics in Spring in a few places in particular with signature change is binary compatible actually so things should work fine when just upgrading the Spring framework jars and running existing application jars against them +Coversions between the Number types should be possible +Thanks for the reminder Claus +Closing at reporters request. +I rebuilded +Works as designed. +Hello TakayamaThanks for the example! I made some changes with respect to this and I believe that covers your issue too +Click on Layer m guessing maybe it has something to do with computing the you are having this issue just remove any layer groups that have no layers in them and make sure your layer groups do have bounds calculated for them if they do have layers in them +If everything is OK for you I can rebase and rewrite the history of my branch to obtain a clean commit before you mergeThanks for your feedback! looks like perfect merging. +Hence snips not to be taken as is +I still need to test. +I logged bq +Will take a look +Oh of course +can you please provide some more details on how to reproduce +We use hard spaces in Hadoop +The only references I see are in the assembly descriptors for the binary distributions +I still see the variables in the installed also tried the latest released +I think this is a dupe it requires classloader believe we can try this in Ive done it before and it just requires carefully a documented break in backwards compatibility +He is a possible change to accomplish this +Test code is far from having a structure that resembles the system code +To secure the console in the embedded EAP version Start the server with no commandline parameters or with the parameter c productionProcedure +I dont think we need additional check for that +Patch looks good +Assigning this to +But better docs might help +Looks safe enough but wondering if this wont have to change again when the client gets updated +Thanks Navis! +A quick look at the equivalent community documentation confirms that this was a direct so it cant help confirm just what the formatting should be + Implemented functionality to generate pdf and csv file formate for balance sheet report +No time to do anything else before the releaseWe need to get exception throwing off the return path for performance anyways so I want to leave the warning disabled until we apply replication of finally blocks for returns +Collect the distinctValues can be expensive but in my case is a requirement that Solr cant give me in a easy way +Thanks for the review Suresh +Graeme assigning back to you as you need to make a call. + +Path and currently throw for path problems like this +zip dependency in addition to +patch of correct but Ill commit it in the next day or two +Raghu thanks for your work on +Initial code +Btw why was the corresponding patch from GWC reverted? It seems to be it would have been good for a GWC +Fixed in svn revision +with older versions of Castor? Is this something youd expect to work? I just finished a test case that shows the problem as reported by yourself and the href attributes is missing on the unwindCashflow elements indeed + code generation for stories and scenarios. +You should add entries for the new settings to document them +Alternative patch If resolving a relative url removes a leading dot it should remove a following empty segment to +This is a different issue is working on over months without a release yet there is a fix for one of the show stopping bugs which prevents you from using the plugin +do I need to add something else or is this a dupe of +Thanks a lot Im waiting for one more thing and will start preparing release of +I can reproduce the problem with a simplest dummy app attached below +Back ported to closing after release +I dont think it ever can with absolute certainty but we could set the severity based on how long it has been since we last received a hint and mitigate the problem +There is an additional issue with the feature +Thanks Benjamin +Thanks for filing +Fix now in CVS +But I guess theres no rush until someone specifically needs that +I hope that may help a little bit +In the attached patch the constructor of is mocked to return wrong or correct address +Thanks for the update Claus +Committed fix for that in revision +I think including first lines of the file is quite descriptive and shows the overall structure of the file quite well +Fine +Yep +Thanks for the fixIve applied your patches in r +Chris kafka was released a month ago anythng else holding this migration up or can we resume +With the latest synapse build synapse responses to the client in same SOAP format as it was originally received to the synapsenot in the POX format +Either the has children and the label needs to be defined inside a or the is a leaf and then it is acceptable to use the tags body content +The latest regressions have ran without any failures + +Assume closed as resolved and released +Proposed solution to allow adding arbitrary aliased jars to manifest cp +There still seems to be an issue with this using the axis transport +These can be passed directly on the commandline with Dnamevalue +Can you give it a try? Keep in mind it will take up to a day for the mirrors to be updated +Which part of? Where are the new tests? and extended the symlink tests to +The patch is IMHO ready to be applied maybe needs some works on adding Javadoc and corner cases tests but it is goodI assign the issue to Bill whos the current mastermind behind +It does not make sense to externalize aninternal request + +Thanks again for all the detail and example war +Yeah right forgot svn add this is the patch for it +Assuming we have that it is not clear to me how much we can gain by doing on top of this +the tasktrackers rpc port for thebq +Thrift has i which allow timestamps in ms to be storedtransmitted no need to add datetime complexity +weve got about success builds and no failure +So you could for example create a Java class each time the script is run and change its like above and we would get the same problem again +To me means this shouldnt be committed as is until changes are made with the same force as means commit away as is +Would you please try this patchSpark ShenBest patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +What would be cool is being able to keep the exactly the way it is with a Data Source JNDI name but in the console you can create a and give it a name that matches the JNDI name you gave in the and the console will use it +pom jar war etc +pom wsdl again Daniel this is correct and works fine with the wsdlRootJohnThe default options are now merged into the configured options more or less as documented. +The watch executor thread doesnt shutdown properly +Attachment has been added with description FO file showing the P open bugs to P pending further review +Change state to Later +hadoop tests run fine on and to as a fix for ve just committed this to trunk and the branch +When deserializing a variable length number into a java integer do not ignore problems resulting from truncation throw an instead +Have always return this + +Bulk closing stale resolved issues +they should be added to the factory in geotools otherwise they are not going to be discoverable there +Im planning to fix this in next version once CR and GA testing is over +Our intention is to move development of medic into an apache repo +bare in mind the method taking these arguments is protected as it may affect other injection the patch into master and and branches. +Bc doc also requires looping over the +I looked at the Hadoop checkstyle file and found it to be more orderly +bulk fixing the version info for and all affected issues have in +Thanks for the review RickCommitted to the docs trunk as revision . +Code commited +Here is a simple unit test for the PDF link parsing +The attached test cases no longer crash the compiler +Adding the full path of jps to PATH helped pass the test +fixed in aceeddafcebcfeafea on is not fixed in branch +Were going through old bugs well resolve this before release +That did the trick thanks a lot! +I tested this with Apache Derby Network Server and get the same behavior. +Ah right +After this patch is applied Ill create a new JIRA to track that workRunning full tests now +The Win executable has been updated in the SVN trunk. +To be safe + +Thanks for the explanation and for mining and correcting these upgrade issues. +This may not be the right way to fix it but I think it is. +This is similar to Ill fix the pom in a similar way +Closing issues which have been resolved for some time +you said you created a VDB when you add the source model in the VDB explorer then in the Source Bindings tab you can set the translator name and jndi name of the connection +Whats happening? Do you need unit test +Rob is this still an actually open issue? Should we close it or at least move it to some other Fix Version? for the time being. +We dont use markers so thats fine + to illustrate the for finding this +That would be cool because there isnt any documentation on creating a renderer akaik +Old UI design for +So we should add it in as a example plugin for the PDE Wizard so it is available through that have the editor that works on the +Header check added to checkstyle and all violating files have been amended +I do not really have my head around all the cases and handling for parameter setting and clearing etc so look forward to hearing review comments from others on the patch to understand better +Does this show up along with the message that outputs the number of reducers etcIf so would this just not go away with running the cli in silent mode +SamuelI really hope there will not be any data loss I have tried to make sure all upgrade paths are coveredThis most recent folder location change should not affect anyone since the folder location is already changing from release this is just changing the changed location unless you submitted an app using the git repo version of cordova you will not be affectedAs far as leaving the option to backup data to iCloud that option is not possible for localstoragewebsql so I have not included a user setting for it +Did you install some specific jruby items before trying sonar? Could you attach the error logs pleaseThank no further information was provided I close the ticket +Recommitted +Another consideration is ds quota is based on a multiple of replication factor so who is allowed to change the replication factor since increasing it may impact a different users quotaGenerally when user creates a hardlink in Linux it requires the EXECUTE permission for the source directory and WRITEEXECUTE permission for the destination directory +Not exactly +Will this be fixed in an upcoming version of xerces +If any of these updates are done in EAP and merged into please update the fix due to inactivity +project servicespig +Kathey If you think this incarnation has a different root cause than yours Ill file a new issue for it +Is this acceptable or should an attempt be made at reimplementation of intention using removed interfaces +And it is especially unclear that we want to do so without any discussion +This is now in the branch +One case I can think of in which not retrying might be better is this +For examplejdbcderbyencryptedDBbootPasswordencryptionKeyThis needs to be changed to something like To boot an existing encrypted database that was encrypted using the bootPassword mechanism specify the bootPassword attribute +Because both of them changeThis patch is available to apply on the vectorization branch +For the record this seems fixed in Confluence +The title should be handles headers incorrectly + in Brew and confirmed with Mladen that this is in the Hudson builds marking as resolved +Can you please let me know what you think +Cool +Hmm those were some weird errors looks like jenkins is just in trouble not the build though +nvarcharupdate col colalter table drop column colalter table rename column col to colen als er geen te lange varchars inzitten lukt het +For repositories this patch slightly degrades performance by a factor of about +component created. +Start the producer and let it run until it terminates with the broker full exception +I just committed this. +Is it OK that I disable this feature for now so that unit test could pass? Good that Dhruba already enhanced save in that could be used as an alternative to restore the failed image directories +Example project tried your example project and can confirm the described wrong behavior when used with +Checking now to see if it was fixed in the last release. +Now that is fixed we need to use the EJB resolver provided by that to change release version +recovery doesnt work for Topics eitheron a restart non of the subscriptions get initialized in hence messages get put on the DLQ +on assert removal +An SPI entry in wfs is not needed because the required SPI entry is included in which is on the classpath +It strikes me that one may want a undoing after a scenario but not during the execution of the same scenario +IMHO this issue is questionable if string type is used as value then application developer is expected to do the necessary checks otherwise if any numeric type is used as value then conversion will happen +Last patch should be good to go +Fixed on and trunk +Will work on the rest +All HTTP conduits are configured twice once using the name based on the service name and once based on the URL +As described above using a link file is not supported directly +Patch fixes the broken test +Thanks Johan it was a pickle but i think i nailed italthough you might want to check if i missed anythingPBTW Martin what on earth did you use to create this patch? it simply does not match anything on trunk or branch +I have added tests reports to big TODO is determining which set of tests to run for each project +please reopen if still an issue. +Its the output of a Webservice and makes it difficult to use +The stack trace in the description is coming from the modcluster subsystem not the web subsystem +patch with jsch added ve just stuck an that has an explicit sftp configuration that extends the ftp one and is pulled in for testing +completion of issue i would like to complete this task and also promote and +Yeah matrix multiplication can be one +An interface is still handy for using the override proxy and as John indicates using enums with value maps isnt really any simpler for the connector developer. +running +Not sure whether it is Java or Mac OS that is some how prohibiting read ahead for these sequential readsSo we could enable this on datanode as long as the buffer size is k +A fix for this bug is needed because of a regression caused by the fix in We could just pull that fix out and we wouldnt need to do anything +I am not able to reproduce this on trunk with the latest IBM JVM. +As my understanding is that the current sort in map reduce is participating by both Map and Reduce +The is used to release the dependents of the bean rather than invoking any methods directly on the Beans. +Workflow changes to force hudson to pick up the the patch no longer applies +Before looking into maven profiles for other database engines let me try to execute scripts through junit as maintaining the list of scripts to execute for every supported database in maven pom sounds like a nightmare to me +But it is tough to do that when you have a full blown app around it all +It would be nice if Enis can review this as well because this patch modified as well +I dont see a need for this changecheers socket is a class that can be difficult to subclass to do as you wishI note that the WIRE logging is spread across classes in httpclient in places +Anyway extensions worked for me in first tests I just did +I just committed this +Patch is attached +I improved the wording and the buttons +Thats extraordinarily difficult to get right given the current structure of the codeWe allow for objects to have fairly names and rely on object hierarchy to provide unambiguous paths to objects +Remember that is usually set to a value due to the configuration XML files even if the user didnt set it explicitly. +the issue linked with jbosslabs xwiki integration +Ive tried to build latest source from branch +This patch is for branch +Bulk closing stale resolved issues +is this a valid S issue +have decided after all to reclose this and instead to create corresponding subtask. +Support for the or find qualifier was also implemented. +Hi Jim can you please check thisThis is a reproducer that publishinvoke a web service using Endpoint +The new page to mention also requires that implements rather than not include the pagestore as part of terracotta wicket module? i dont see any reason why this should be part of wicket source +But we accept that since it increases scm compatibility and since it unlikely to happen in standard CI env +So I have reverted jenkins subversion to version the one before before svnkitAnd that looks to be really better +And with xpointer I was a little inaccurate +Im concerned we do not understand the others enough to design the interfaces well enough +I have no idea about NTLMWith preemptive authentication do you believe that the nonce can be? Maybe some servers can be tricked but that doesnt seem like the way the protocol was intended +The jstl jar is missing from +files generated by m build have been checked into the code. +And Willem there is another jira ticket for script unit test is not running you have fixed this now great +Please discard all the other patches +Closing all resolved issues due to a successful release +At least in case of web applications etc will not be available through s classloader +Attaching a diff with a proper fix for this issueAlso a test case is addedThanks AmilaThanks for the patch and it fixes the namespace interpretation issue of expressions +The now does the authentication for both plaintext and password digest types +Why not just add a new method which calls the existing method +May need some debugging and definitely needs unit tests but I though I would post what I have so far +I put them here since they are tests failing in the same +I think this is the samve issue we are tracking in taobao where it will case many http SM hangup in system it never time out and close it is very dangerous in heavy system like us Wei jin and Koutai working on this and have a patch for testing currently will submit it later. +Currently the list of supported sql states is and +cleans up Interval and a bitIll commit this shortly +Yes should be easy once I figure out a good approach to configuration +ok rewritten the unit test to correctly reproduce this +Of course this is just for the file protocol but that is know from the prefix. +Please confirm they have been applied as expected +probably the bat script does not recognize if the input argument is quoted or not +Major +I dont want to risk to break the release now and delay things even more +is there some reformatting happeningIf we cant implement this without major changes then I dont think we should do it +Hi MartinAs it involves a change to the parser the patch requires a bit of extra work to apply namely compilation of with javacc +Whats the verdict? Will there be a vote +We need to clear the import section in The name of the for is Namenode Retry Cache this will make the name used to register the metrics Retry Cache +I just tried to do the same with unit test but all passed successfully +fixed at r on trunk and r on +Need to review if some other changes are required +maybe this set iterator isnt really needed after all after release. +jsfdia files. +Checking for nulls and throwing NPE for patch for the ve created a new patch and applied itAlexey please verify +Ive just applied a patch to the SVN +Committed to trunk +Merged into +Or upgrade to newer release of wrapper yourself as we cannot distribute a newer release because they changed the license terms so ASF does not allow included that in their releases +pending therefore keeping issue is now added so I close this issue +Bulk close for +From analysis of the the stack traces I believe the original user reported problem is likely fixed by the interrupt work done in in the linked issue. +However after recover has run there are no more logs +Please check that it was applied as expected +I tried to rewrite as a but I could not figure out how to use did not find any documentation nor report using I have code to write the new HTML report as a unique page overview section per goalIf I dont find how to break it into multiple pages Ill publish this version here its better than nothing and I hope it willHow was your work here? any progressAnyway the suggestion is to make the page more like the page in mavens project descriptor +Ashutosh We would not hit the wrong token issue mentioned by Daryn because hive metastore proxies as the user requesting the operation +If I forget about my workspace and completely start from scratch its all working +That means that in a Java consequence we cannot use Java Bean Property type Accessors +I will try to resubmit an updated patch this version of the patch +It applies cleanly and the newly added testcase passes +new patch uploaded without CTRLSHIFTF usage +Thanks any release foreseen? +I believe this is for the most part done already +I dont agree with this as a default +If we go the JMX route we can expose via and quickly fix this for both Hadoop +Committed revision . +Except as Brett mentioned if the applications check for XL without also checking for XL +code makes the entire sortshufflemerge pipeline more complicated to maintain +Verified with M respin +but yeah +Didnt know that this was already setup + for the second option +I did the update to Glassfishv last weekend andthe problem disappeared +Stripping the Ext suffix is out of question since the compiler will start complaining due to ambiguous method signatures whenever someone adds a using together with a using lognet +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +reviewing issues for to where we can take another look at this feature set. +whateverit is Dnode Dnodelocalhost unfortunately the failures described in the above still remain there +whoopsie I actually checked this in nearly a year ago and forgot! +Returning Query gives the Strategies freedom to have a meaningful score if they support it +commit to trunk At revision Michael FangStill found a little problem to the Monitoring porlet on admin console the Enable Query link associated with this default server server works well the Disable Query link but found the value of for the default server keeps Stopping +We can change it to use the new metrics typeAlternatively we can have push out two values to the metrics sink one is the current counter value another one is ops sec +I have had to upgrade to using version of Hibernate so my queries will run without throwing an exception +Could you update the link +bq +Similar error with blog but shows up at a different spot +Then we could separate out the toolbox into quickstart and sitegraph plugins +the only difference being the amount if memory on these types +This issued should be resolved by +For some reason parsing symbols make this code ridiculously slow +HelloI can now reproduce the problem! It only occurs when working with running against HEAD is successfulHow far can you switch versions? Is it possible that you run against the latest snapshot +Yup Ive noticed that you checked those changes in +Ok make it fail with a better message + +I just committed this +If you want to update the wiki create an account there let me know the account name and Ill grant you write access +This is an project and all code and documentation is available to download and patch +Let me submit to see if it passes +VFS will be using vfs moving forward + Britt +I feel its a core piece of the Jetspeed architecture and dont want to throw aside in the zealous pursuit of COP +Includes previous fixes plus couple minor changes to namelydecreased snapshotIsolation tests iteration count to so the test runs in a reasonable amount of timesplit testBlobs into test test test so we get more feedback as tests progress +Updating versions as it probably will not get fixed for +Thanks Scott +Closing this as +This is either a candidate for Spring Modules or a candidate for Castor itself actually since such specific integration code would fit into the Castor distribution as +It stuck around for over an hour so I am going to assume the resource is stuck open at least until the threadthat started the server goes away and in worst case until the jvm comes down +Well we gotnothing to stop him with this struts class +Just got a little confused reading the comment +I believe this is because of an error initialising the indexer and it returns null +Lewis We will create a jail for you that will be something like podlings are not permitted to have a top level DNS entry as they are of the incubator +Could we please collect a list here of what information should be made visible out of the ZK data? Also where +Going back to the original issue Ive got a patch that should help some with the Eclipse issueIt simply generates the Java code into a different directory qlsrc so that Eclipse users can include that separately from the preprocessed source code +Filed and submitted patch to can you rebase this patch? We need to remove the YARN changes as is in +Added a test case in trunk that creates a consumer on ran it on linux and windows no errors encountered +please use the attachment as referenceSwitching between the tabs would not render the scrollable table properlyalso the tooltip for the deleteremove button gets cut at the table of a bulk update all resolutions changed to done please review history to original resolution type +I currently dont have a working standalone testcase to demonstrate this but my class bridges all implement +Bundles wouldnt be able to list future so it is up to the framework to indicate backwards compatibility of I would thinkI agree with Christopher that we should set the value to be all SE up until the detected SE EE +The patch has been appliedDenis Alexei please have successfully reproduced the failure at test +Ive deployed a new version of the plugin with your patch applied +Two tests remain commented out because there are some situations involving where Derby wrongly finds a single matching signature instead of raising an exception complaining that more than one candidate matches +OK Ive just tested Sonar with Sonar Runner and it works perfectly +Sure sounds good i am going to reapply your original patch and test and i will look into that as well +If we have class which implements one interfaceWe should do the same as in +If this is needed Im happy to patch this as well +However I dont think its a good idea to try to come up with a unified approach of generating hadoop based recommendations +Looks good for building Composite configuration using Spring resource handling but the creates Properties +Revision . +Yep your idea looks good to me thanks +I may be wrong but I dont think the IOE in log extract makes it out to the regionserver? Its caught down in? +I really need this too but Im too bad in +Anybody got chance to apply this to trunk? This bug is extremely applied to the trunk +But when I implemented this for myfaces I tested this scenario with RI and they really created an for one component +Applied changes to testsuite for and to eliminate dependence on localhost +I have made the changes to SVN after run the test frame work +heres a before vs +bq +I unfortunately cannot +Since we refactor the webservice support it does not occur anymore. +Hi think I got your point +I am not much familiar with WSDL and rest so you get the proper answers from that +renamed zkstub per bens revision +Looks like the jTDS driver doesnt provide any way to customize this datetime format +Patch relative to the current trunk +So the master sends a message +Added logic to account for procedures without parameters when generating preview Released +Committed patch with revision and other patch with +updated fix line after keyvalue pair was removed junit was corrected +ported to +Since we are still pushing filters for joins the assumption that we will always output a row by the filters is true and therefore we dont need a progressCool I will take a look at the patch +Damn it does indeed not fix it and after reading the title of this JIRA I understand why +If its trivial we can fix it if not then Im not sure its really worth the effort considering Groovy is the current official stable branch and that everybody should migrate to this version +Fixed on trunk +Resolving obsolete issues. +The attachment is requested for public reviewThanks +test the vm is pingable or not +Giving implementations for graph algorithms find components in a graph find in a graphFeaturing javadoc toy integration tests with multiple recursion runsI am happy to receive further ideas on improvements +I can change it if we decide we dont like tr +Nobody seems motivated to execute on this. +I would be more than shocked as I create new servers and runtimes every day +I guess we can switch the new api in karaf maybe +Hi Konstantin I an finding that this patch does not merge cleanly with trunk +DIGY +This should IMHO not be the case +NN begins lease recovery and finalizes blocks +I have sent pull request for what I did to migrate the existing Tests are still failing +If theres a concern about having too many apislistnamespacetables cli command needs to be updated to use the new api +bq +Thats unfortunate +If I had an ij connect somenewdbcreatetrue and it was not my first connection would it still get created +Yes I should have commented on the title change +If the problem persists in please the bug +UG +I update the patch with manage on imporve resolving patch when the flexibleString is resolved search fields on root context and parameters map +Btw after creating the database please execute the following queryupdate owner set product where id And dont ask me why +We are going to address all Administrative portlets issues in wont be reworking these portlets until with Ate +Disabled attribute doesnt work the bug was reported some time ago by Aleksej Yanul +Hi Gong looks like youre facing bug directory cant start with numeral +Removed the unused file in +So probably there is none and the small performance improvement isnot worth to be considered further. +patch looks good +The code already exists in Cactus land +Where do you think of placing it? I did not see it anywhere in the patches +The tests created a bunch of bookies and wrote a load of entries and failed the server +Experience of answering questions in forums etc over the years has show that people tend to just copy and paste it without really knowing what it does which isnt idealIve corrected the description in the and RNCXSD files and also labelled it as a legacy attribute and advised against it in favour of using explicit configuration of the features you requireThanks for the report! +Patch committed except that has not been added the test has been added to instead +That thread contained a few topics but was mostly focused on the combined jar which the consensus was to remove it +In the meantime a lot of changes have been done to both the schedule and the calendar components +Start Aplication Server icon I think I have located the problem in the EWP installer code +Committed to trunk +Ill back it out +However I have a counter proposal to changing the typeSince is meant primarily for legacy purposes why dont we change the distributed example to not use it for the float field type + +I wasnt able to reproduce test failure in Addendum for adds debug log for the case of recoverLease timeout +Perhaps if there was a command to fetch the names of the nodes with failed drives or somethingAlso does this data appear in dfsadmin report? Thats pretty unstructured but if it shows up on the UI it should be there too +There two unit test still using HDFS that still take a while to run +Woops I did miss thisShouldnt the if statement be fileOffset numBytes length instead +I dont know if thats a fixedAttached is WSDL I generated after fixing some of the issues by me earlier +This sounds like the approach that will alleviate this problem to a certain extentNot in the current form +Ive cleaned and updated the patch +Seems the same problem of import from the classpath +this is not done in a reliable way +Its not in Gits master. +Added a test at adabfbdbcaeThanks nick for applying the patchSorry for late response. +This is fixed in the branch. +Edited comment for clarity +btw i am working with think actually I must have something else wrong +Another issue is that the always holds a reference to every courier created +But this issue seemed significant enough to be given higher visibility as a major feature of the release +try mvn o install to get the error in tests +The attached patch modifies the internal pageflow processed annotations to remove dependenceThe patch adds a new directory with a few new Java Bean classes to the svn workspace +Im going to mark this as requiring +works with this you are right +committed +Would you or anyone happen to know how to get the bookmarks if you have the object number and generation number +Can you take a look at this +Nigel reviewed this and gave it a + was used for compiling the build +Contributed by Erik Steffl +Curious what does allow them to escape mean +These issues have been resolved but never closed +bq +Thanks to Mel for his contributionI think we still will have an ant build file even if we are switching the build to maven +Initial portlet support committedAs this most certainly will need further enhancements and refinements Im leaving this issue open for now +Please post the full wsdlthanks file illustrates the issue +Integrated to trunk +Each test was run against a node cluster at RF with a separate client machine pointed at it +svn diff format +but there is a BIG problem with the referenced assemblies they are locked by because I was unable to use another for the compiler +Try the modified example I made using the patch Ive attached +tested! +Nixing the string output solves that in a way +Now that work is well underway with AS all previous community releases are +As a quick solution to this problem I guess we could write the following query to find the associations of a particular user by username and delete all the while deleting a particular +is not sending the right part of the file when using encryption and that causes various described herePatch attached for fix +We really need to introduce our own type value class or type class to handle this consistently. + +Changes Support for date types and keep a static instance of class whose parseMath and toObject methods are used +Uploading a new patch +Looks to me like this is tagged now +Thanks +Re Jacquess comment of Apr See for a fix for initial state list. +Found some more issues +Ive just applied patch +I just wanted to stay away from the root but am more than happy to update the poms to be placed weblogic if that makes more sense based on standards and package me know +bq +Linking to which represents the ESB WS Proxy stuff I implemented +I dont think the synchronization code in the class is being tested +In my test case there no datanodes registered with the namenode +Looks good to meWhat Im a bit concerned about is whether it is clear to users that this is SPI only +The jUDDI DB configuration would have to be modified in any case before this would work and this is probably more hastle than it is worth +More specificallyrvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodelheader rvmsrcorgjikesrvmobjectmodelheader rvmsrcorgjikesrvmobjectmodelheaderlocking rvmsrcorgjikesrvmobjectmodelheaderlocking rvmsrcorgjikesrvmobjectmodelheaderlocking rvmsrcorgjikesrvmobjectmodelheaderhashing rvmsrcorgjikesrvmobjectmodelheaderhashing rvmsrcorgjikesrvmobjectmodelheaderhashing rvmsrcorgjikesrvmobjectmodelheaderhashing rvmsrcorgjikesrvmobjectmodelheader rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodeltables rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel rvmsrcorgjikesrvmobjectmodel of what there is cruft that Im in the process of moving to the new model +Please check and if necessary +Few comments I assume enabledisable all just offlines user tables not root and meta? The help for enable should probably be something like Put a table or tables online and disable should be Take a table or tables offline +Thanks for yet another last minute fixDan +First version of a possible changeset design +To be more explicit the solution proposed by Michael works and it could be good if it was committed +Implemented that caches definitions and resolves them +And we reached the conclusion that it isnt after an elaborate analysis of the order of patches that went into trunk and yahoo distribution. +Im not sure if this is as realistic of a use case as the out only producer but I guess its doableI looked at the Cache component which does similar things +So the fix is perfect for will need a tagged snapshot or a minor release for messaging now. +What do you mean by an external media file? Can you provide an example URL +I have to study more the sourceThanks for help me +Fixed and deployed as to npm. +bulk cleanup of Jira versioning +on v +This is probably broken now with the recent changes done in the NHTTP transport +This change only affects function call with constant parameters so it does not cause performance degrade +How about does that somehow make it easier +Any idea on when this will be released? Thanks +JakobHow about the progress of this issue? I have little experience about developing Yarn appIf you share your progress or separate this issue into sub tasks I can help you a bitThank is coming along but there are a lot of issues that need to be done first +Attaching a release note for this issue +Thats exactly what Im doing now I will upload the patch archive in a few minutes +I tried to clean up the patch in order to make it more readablePlease let me know if it is OK +NaveenThis assertion was used to find the situation when we have GC worked at IP that is neither callsite nor NPEWhen it was written I knew now such situation so the assert was goodAfter we added SOE handling any PUSH op can cause GC +Checked in Stefans last patch with minor changes +there is now a new constructor in that takes the factory +Please if reproducable on +For the moment we could turn automatic garbage collection off there is anyways a manual trigger available that can be called via a web service +Ideally the programmatic query building should be able to build a query for null valued fields. +once injected however the scoping is handled by the container via the use of proxies +Thanks for your fast help +I agree with Mike that we shouldnt attempt to force an Eclipse project format on the development community +Since there is no standard xmloutputfactory option for invalid character replacement odataj does not presume the use of woodstox and theres no simple path for user interaction of the xmloutputfactory to set a woodstox specific property its fairly safe to assume that this will not be addressed as an enhancement by odataj +Fixed in commit ecd +Havent looked at the yet +Do we have any docs about CDI plugin +Could you also provide the full stacktrace please +I can reproduce your problem +The patch file +One reason was because has log threshold set to TRACE while to INFO +registers its self with the Hadoop Metrics +The former code misused the default charset so that Chinese characters can not be encoded and decoded normally on windows whose default charset is +Ill commit this soon unless someone objects +Are you able to post a patch via reviewboard with the fix +Simple ideas are often the best ones the new patch has a single loop! Thanks Robert +If this is not supported maybe support could be added +doesnt it need to call super in setThe set method was only added in order to be able to write a test that determines that the connection is available during the execution of the metaData closure +If someone can try with the sample client +The documentation issues are interrelated +Done +But you assigned it to you before meSo wait and seeCheers +Still this will be very confusing for the users +only for s and wicket is not creating that url the user is. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +All tests passPlease run a fresh build from trunk and test against your apps +I was getting an invocation target exception +Working on it +Could you please review the attached patch to include it in the branch +This can cause problems when importing +and toolbar dissappeared as well +I will first address the jira issue validation so holes cant overlapp +Thanks Andre. +We can push parsing to a different thread and kill the thread after a while but IIRC forcefully shutting down threads is a looked down practice in java +Confirmed a working install on my Ubuntu dev laptop +but in one scenario it doesnt work as you said the changes should be enabled only if the server has modules and is fully synchronized +can you try loading not storingIf that works then the problem is that the bit of code I pasted above only gets called during the loader constructor but not the storer +All unit tests passed for me +Thanks for pointing the problem by throwing an exception if the operation is missing. +If the test is flakey lets disable it +After updated with all the failed tests are now passing +I dont know whether you ran the code against the latest SVN +to set proper after having corrected fixVersion +I like that approachCommitted with minor changes to to move a little more logic there +Please feel free to reopen this if you have any objections +Thanks Jaimin. +This was introduced in commit fb +Installing lib or getting to use libncurses will probably solve it for you so perhaps its actually an bug. +Google doesnt seem to know Kolbak for software for example +Thanks for digging in and investigating +new module osgicore introduced that builds the original sources in a java compatible way +verified thanks +Attachin stack trace of the loaded server +Does this have a negative effect beyond the warn log entry? That is does your application run fine in all other respectsWere going to address this in some form anyway +The container launcher already knows what containers it has that arent complete so I chose to have the container launcher kill any containers that havent completed when its stop routine is called +I actually dont know is it good or not +Im just going to commit the current fix and worry about any performance improvements in another issue +Indeed you can explicitly specify the profile with P profilename option while building the docs +I also think so and your patch is better +ah i see +WSDL that causes the AndreasI have attached the WSDL that causes the issueThanks the WSDL file appears to be copyrighted and we therefore cant use it in a test case +Patch for +OK then well leave the issue open for now +Logo would be good +re scheduling for in SVN revision +Besides that this fix looks good to me +At least we need some way to override the configuration for some servicesA different approach would be to define each web service bean in the Configuration class with some nice API to do all necessary configs +Issue has not been ported but there has been no request to do so +It would be good if it tested the three destinations exist and +Is someone working on this issue? Otherwise Id be willing to contribute +Patch for database specific decision on whether to a column on which a unique constraint has been placed. +Its not relevant to AS +Especially when it is disabled although the the task logs are no longer available via logsuserlogs but they can be available from the job history server +Do you need some help to move Velocity on M +Part of a bulk update all resolutions changed to done please review history to original resolution type +But its worth noting that Hadoop already probably uses too many ports and we should probably aim to use fewer and rely instead more on authentication and authorization of requestsFor example we might use HTTP for all interactions including RPC shuffle and datanode access +Nice work Josh I see youve included a sample as wellI have one small comment +on to trunkThanks for the review Matteo and Jerry. +That is now done in the main script and only if CATALINAHOME is definedI think I should put a test for this being defined in this script and fail if it isntIll update this later tonight have to go now +Isnt really enough for me to say for sure on that report looks like all threads are doing what they should +I believe this is now in proper shape +I expect to finish tonight or tomorrow +The fix is available via this jira for community user use +In this query the same load provides input to two cogroups +should be fixed for +Thanks for the fix fbricon +This stuff needs more study but inside in is not the place to be doing it +One of the questions I have is that it seems like the scoring process doesnt take into account the distance from a central point +Hm +Ok its definitely not setting executable under Java +Hopefully the test fix in will also fix this providing it is also applied to the test +In reality any one to many relationship with a back pointer generates two joins when only one is necessary to pull in all the information needed for the queryDaves example in the mailing list should be something you can run pretty quick and see the problem + +If there was a general exception in executing the script ignoreOutput is set to false and we are trying to read the output which may not be valid +The constructor Im referring to is the not the constructors wont be able to take implied closures? Thats no good +Spring webflow has not yet been released + +David Perhaps we should resolve this and I will create a second Jira item for the other issue? Or should we keep this open hereJust to summarize the behaviour is that if is indexed from the root the index itself will only contain those fields that were specified in the root of Item +Im not quite sure that we should test functionality that provided by security provider +Amaresh can you clarify is there an objection to providing a public API call that allows addition of jar files to jobclient? If so whyNot everything is a Tool +in M +But applying the resulting patch file runs into the same problem I faced when trying Matthews first patch +Heres a patch that does what the old scripts used to do scps the private key to all machines so you can ssh between them rsync config around etc +I have now closed as it has been resolved I have also renamed this issue to incorporate the commit recently made to our nutchgora branch which is also in development +Was just fixing testRecovery and the in this patch +No newspaper is also an individual component in Tomahawk +looks like ok not porting to branch for now to have some beta time. +Thanks for that addendum TedLars thanks for the port +I havent actually tried your patch yet but I would expect the to have a similar check +See attached screenshot +One hint is the stack trace has your new package name +I checked the standard and SQL savepoints do allow nesting cf +My impression was that putting it in a separate space was intentionally since it has different access characteristics than normal bootimage objecrs +I have uploaded a slightly refactored version of your patch +Move it back in if you think differently +I will be creating a release candidate issue as this has been fixed +Alexey You comments were addressed in the +Nice idea seems reasonable to me +It should be released on the update dev site soon and then shortly after that be made available on the regular update site +Committed revision +I did my best creating ones for wp wp but guess windows still needs one +At least I would like to know if this is an issue with the user code or Lucene code itself +This still plagues us when using RTE +This patch includes caching so performance should be comparable to what you see with the other region location lookup approach +Bruno has been able to reproduce +Thanks Lukas +Identical logic to main patch +Dhruba mentioned there already exists retry logic implemented in the DFS client code if the is throw it will wait about ms and retry +im not sure there is an easy way to fix the behavior. +It should say something like Please first select a connection instead of Selected logfile already exists +I committed the change to trunk and +you need to provide a jar with the following contentsthe jar you want to uploadthe pom file +Having a patch ready to review would be faster +Thanks for the patch +I wasnt sure what to set the status too its still REOPENED I couldnt set it to closed again +Add the spec jar poms have been updated to extend directly from. +Updated Enlistments are now registered as Durable +The stream cache should be changed to use or the likes for a better cache store and leverage Exchange to cleanup the cache +example model with a new test test patched you for the patch Ill take a look at it right away +Assigning to for investigation and resolution +Could there be a reason why a user would want to reuse an old? Just saying because by introducing we alter past behaviour unless we introduce an option flag to negateenable the correlation ID +A simpler workaround should be to just specify the path to the parent POM directly +Machine has been installed and set name is do Set up Xen dom and two xen domU set up at the moment the CI handover involves a manual step this will be automated once system is proved to be running correctly +Dotnet does not recognize the default way the command line is handling some arguments that has spaces and double qoutes then just wrapped a single qoute arround it the arguemnt but the dotnet tools doesnt recognize is overridden to handle the Dotnet Style switchtext where text can be value or a value with space wrapped with double qoutesEg +i will go ahead and commit this +Actually the was not generated +So far object graphs seem to get saved and loaded correctly +So more information would be helpful +This issue occurs again on the is a Windows cmd script that uses instances of the sample app to create embedded brokers and the with a failover URI to demonstrate the problem +On put bytes are added to current total +Touches the following filesSending srcrefSending srcrefSending srcrefCommitted at subversion revision +I committed the patch to trunk in rev +One use case Ive had for something like this is the followingI have a thrift definition which I want to manage as an rpm +to Henris response on July and to Roberts response on November +there is no other explanation for the transientivity of this error +The patch was applied at the repository revision +Only if we expect continuous changes and the necessity to keep the integration in sync inside one release then it would make sense +All JDBC tests are now part of the Junit All suite. +Configuring the keystore type now support in rev r +This matched behavior of when is not local +Fixed in previous release +Coincidence or part of the problem I cant say +LOTS of ppl use like that approach so it may needs be around until to Shaz setting to iOS +Fixed in trunk +Not all decompilers produce compilable code +I see the problem its similar to Microsofts default settings ones +I tend to think that we should remove all of these dependencies and add them to the POM but I dont know the effect that this will have on the AS +It does not appear that you plan to fix it based on your comments +Do you think its valid to link Requirement Entity to entity so we can track what caused a requirement +A few notes +And fixed CS also +leaving us the chance to do a better solution in down the road that prompts the user for cert acceptance and usernamepassword +Since theres no RECURSIVE merging strategy available for egitjgit wed have to push force to the remote imhoI see the following approach add the openshift repo as remote repo add deployments folder add +So if the user sets it to something absolute it is left alone +Else they will need to create another user and create a backup again +its wrong there as for the reminder and approval +DeepeshAfter doing a quick review I found that existing refund process will break +I could change the patch so theres a couchidgenerator where the algorithm is configurable defaulting to the current one if that would move things along +username password +For mapreduce jobs whose output size is proportional to input size the cost of output would still dominate no +Thanks Joey! +Thanks for reminding +Ryanright thus you patch should include a check for the html tag +Are there more recent versions +protocolhttpapi I noticed reference to a TODO about two thirds of the way down the classWhat does this mean +Ill wait for the build to pass and check them in +I committed our latest release to releaseetchThank you very much +The website is now visible +It would be helpful for hive users +Out of Date +I attached a WAR file to reproduce andexception stack trace and richfaces component structure from Tomcat item is invalidThe backend bean of the custom component has to implement. +I am sorry a slip of fingers +This patches libhdfs as well to create a Mac +Thanks for the clarification Dan +This issue is more open ended +Thanks for the review Daryn +clarify AS the changes that Ryan suggests here to +This may not seem like a big thing in a smaller application but when you need to parse a product catalog those descriptive names will add up +And further to default the pattern to this would satisfy the convention over configuration but also give some naming flexibility to can file another bug but it seems like changing the default is a binary decision rather than an enhancement on top of this feature request +txt contain patches that were used to print the the interceptor chains are the same now dont forget to check the HR client configurations to see whether theres any differences and try to provide a couple of snapshots of the client side with the old set up and the new one +Methods and fields now work still need to test constructorsThe tests do all compinations of Base global sar with classes loadedclasses not loaded Child scoped sar with parent delegationno parent done +Changed to Enable Hibernate support +Im still worried the intracacies of this feature +Database is already set up on vmbuildGav has said hell reimplement the data and database backup measures there +The difference between this and the qpid is the rejected goes back to in progress in stead of to Open +many thanks for your time on this +Looks like the problem is the is based on the propertyNames of the filter +Forgot to say the xml above should be added to +Your original problem occurred because of bad documentation +one groovy file has an Singleton on the class and one file has a bunch of Bindable on a bunch of variables +is the EAP report +OK should I provide full stacktrace for this case? I will be able to do it tomorrowThank you +Ill give the latest stuff a look through on Monday and report back here +for the patc h +But version got merged to released version making the version incorrectSo marking this as found in +To be more clear maybe we should prevent this error IMO having No set of EXAMPLE is not enough +latest patch looks ve committed this and merged to +Partial patch just for the common conrefs file +ill mark it as resolved +In the code its just the other way around which was a bit of a redherring to me as well +In case Project and Archive node under it selected together Del key still works and if confirmed archive and project do get deleted +would be good to get to the root cause of how this happenedYour terms dict which records the field number and character data for each term has somehow recorded a field number of when in fact this segment appears to only have fieldsCan you run on the index and post the result backAny prior exceptions when creating this indexI dont think adding a bounds check to makes sense the best we could do is throw a exception +The code is in trunk for the upcoming release +Up until now the dialect simply provides bits and pieces of the query to the datastore +Should be sufficient for folks familiar with Java and Stonehenge to get it all up and running +Draft complete and committed to SVN revision Ready for tech review +This was fixed on trunk with just FYI +you are right about the string size not being affected but then the structure you create must not be something like a constant +buildivylibcommongriddevbin +Dan sfilt can take a units measurement but internally it uses miles +The class being patched is for Tiles and so it wont make a difference to any Tiles code +If this is not already available yet it might be worth to put up at some DVCS like git or Hg +If you finish it before we release beta please correct the versionThanks +This stuff is now integrated and seems to be working well +If we cannonicalize both sizes of what were comparing shouldnt it work either way +Its an option that component was mentioned in both the spec and the refguide so I stuck with the familiar name +I will upload the correct patch soon +Its probably already fixed +Ah now i remember +The added unit test fails without the patch +Thanks to all of you for your valuable views +Hi JarekThanks for looking into it +issue fixed by other issues please check the new and closing stale resolved issues +Lars and derp for not making that comment in the first place +tccell is deprecated because it has quite the same features as tcpanelOriginally the sense of tccell was only to define spanX and spanY but this is no longer needed because of the tcgrid tag which can set this values to each component directlyRemoving the unneeded is an optimization and will be done in future releases +Minor nits does not need to throw the exceptions It would be good to add an assert for null at the end of Jothis comments +What is the status of? It seems to have been part ofthe public API at one time? This is the only place isever mentioned in the docs as far as I can see +Hi AlfredoThanks for this patch! Ill push it to SVN ASAP +This is a particularly important issue for me in relation to my cloud related projects +Yes its is for backward compatibilityAnd its use is to set default umask. +For your first example we may set snapshottable for all users +Heres a cleaner version of the patch for Seam maybe youll like this one better +Attached trunk patch uses to compare two file systems +Shouldnt a decoration layer a Node instance in the first place when it passes it to the decorated repository +Attaching samples and Timestamp Signature and Timestamp Signature Encryption and TimestampEach sample includes srcwsdlREADMEAdditionally Signature and Encryption samples haveetcprops fileskeystorekeystore filesThere is no assumption in the examples as such but we can put them all into sampleswssec ve added the samples to the distribution +The compiler knows later but too the reason why the class is not seen is because of how the compiler is designed +In the former case it makes sense to completely consume the delimiter and keep the resulting tokens next to eachother +Running tests will commit if tests passIdeally checks like this should be part of plan and we should not look at the conf +So I dont think duplicating the java code would fix this +But live volume migration support has been added from onwards +Patch for that adds user attributes for and avoids +This is not the appropriate place for that discussion +These arent Spring packages but Spring Security packages +The issue can be in +Hi Claus thanks a lot for fixing this issue +At this point Id suggest peppering the docs with references to Resource wherever it discusses autowiring things or about collections +There is no special behavior for +I just found that this contributes to another symptom Im seeing for RF and a ring of nodes if I bring down box the remaining still work fine for Quorum access but the latency is x highI can see from debugging that a lot of time is spent on storing hints into local system table on the coordinator +The jobtrigger groups will bejob group mygroupjob name myJobtrigger group mygroupTriggertriggername my +Thanks Navis + PatchI have tested the problem an the solution seems to be acceptableMaybe it could be other enhancements like focus the selected option on or focus an option based on its index but for the problem described it is fineThanks to Paul Rivera for provide us this patch +We all want you done with this and back on +See the discussion in for justification for returning the date portion of the as the current date +Committing this +Thats not right i tested it it will not start try to delivery it now +I was able to add Mike to the initial developers list +verified +Patch applied in rev +I think the problem we identified with transitive dependencies is that it was expected to be problematic in terms of merging the transitive lists. +Instead of Browse git and git organization we should just have which points to the projects git organization +I guess you are referring to restricting the number of tasks per node per job +Added a little to the test case as well to test more realistic versions of fields with Strings in them. +Fixed in trunk. +BTW the site is a wiki so you can fix it yourself +Thanks for reminding me about itIt will help indeedIll debug what is wrong with REDIRECTTORENDER strategy. +Confirmed fixed in trunkThanks! + +Attached updated job matrix +If the problem was not related to exception handling please this defect and provide all the information we need to reproduce it. +Looks like a good solution! Im sorry I havent had time to look at this earlier Ill see if I can get this included in new future. +Quickstart app that demonstrates the problem +The patch and resolution for JIRA Beehive issue willrequire that the paths to tiles definition files be absolutefrom the webapp root +Thanks Jian! +Hi AnnaIs this a command line application or within a server of some kind such as Tomcat etcWhere have the digester and validator jars been deployed? The errors seem to be suggesting that they are in something like JAVAHOMEjrelibext +Thanks for the patch Jing +Bulk close of resolved issues of +Set version to commented out post here the fixed poms and Ill upload pom vtest scope set servletapi group fixed transitive dependencies jetspeed dependency cant be replaced with anything on ibiblio I wasnt able to find out the real version and an official binary distribution so version has been set to and scope to provided. +db driver ddl something +Committed. +This is the mapping I guess that the tools said Cant find the node right +Committed with revision Committed revision +Changes for hibernate maven AS jst runtime usage and teiid done +check in the fixthanksdims +And we know the importance of having updated stats in production for performance therefore the usability of keeping the stats updated would be very important. +No need to do this anymore +No problems from performance side patch can be committed +I agree. +It would be better to report the failure to the Datanode have the Datanode validate whats on disk and report corruption to the Namenode if the validation fails +Central should be onup by default and when the Welcome screen closes it should be front and center +Moving to since the shapefile mantainer is now answering me and rc is due out any day now +Move it back in if you think differently +a custom format +Are there any performance metrics availableI do not see much concern here + the patch +Has anyone checked to confirm that this works in JettyIf not Ill check in the next few days and close this issue +Once constructed they are not meant to be mutated +So starting up the workbench after installing the plugin will not start the plugins until and action is performed that requires its startup +Thanks +Lemme reopen to keep of this for the addendum to and trunkThanks Nick and Enis. +The is just an empty pom stating module the pom discussed ll find the it for it and augment it for this issue +Yeah my bad +patch looks good +push to +Closed after release. +Hmmm +This patch needs to get checked in with since it modifies the signature of an abstract method that a class in HDFS overrides +Dims will watch continuum results +JIRA cleanup +bq +Give me a few hours to make a cluster to test with and Ill upload the new patch +can try the latest cvs +tests pass will issue is fixed and released as part of release +missing class added +If so then Struts shouldnt make the same effort as a standard like JSF +Does this sound reasonable? Yes perhaps getting this checked in without touching handlers or the side is a good idea +Also Derby will today not allow the explicit creation of an index if such an index already has been implicitly created through the definition of a constraint +With any luck we will quickly find a way to meet everyones goals with a minimum amount of conflict +on patch +Unit tests are passing +This is likely very difficult to debug at the moment givenIm just trying to add large docs tests to all the tokenizers and opening issues where I canIf I can get them all in Ill try to look at and see if I can fix this random seed issueso we can actually work on these things +The new region assignment goes via zk +However I like the encapsulation of putting everything in one file rather than spreading them out +This is enough for a Simple scenario We also have the NIO thingy checked in as well +Ok I will install it then what +I am resolving this issue since the remaining problem being discussed is from +Can please get it in to? Thanks +The passes with no failures and our own code no longer has any problem +Deploy jbosswscd webservicesant will create the deploy the following filesdirectoriesserverdefaultdeploy serverdefaultdeploy has a depends clause on the deployment interceptorthats installed when gets +updated the deletion task to not bother retrying until reschedule is called theres negligible chance that GC will happen to force an unmap otherwise +that looks funny to me isnt that linking the same filename to targets +I thought you wanted to be able to pass objects between threadsAs it stands the implementation is not completelyAnd the Javadoc comment is still misleading +Wont be working on these and theyre old +Not sure if its really doing much but if it aint broke dont fix it +The closing issue here indicates that it was fixed and should work without workaround is it so +For example you ask for a GB container with a min allocation size of MB +So it would be really nice to see how the memory is impacted with and without the change +Package details to followCheersJared +I asked about the duplicated id just to be sure but I believe it is a typoI agree it is better to use div instead of spanSure feel free to send a patch +Ill push in a fix todayKelven +Jody can you check this for me please +We need to track all hadoop dependent jars I am not sure if it is realistic +Please close this issue if satisfied +This patch looks good to meAs Jinho said however the patch is successfully applied but it seems to require to be rebasedAlso please remove an unused variable initTime in +Ive just committed this to trunk and. +Does this also affect +Not sure why this is filed as a perl fix +for trunk patch +Since this is something of a paradigm shift some of the related tests dont even make sense now +Ping +I thought that option is your preference thats why I askedRegarding the oldVersion Your suggestion to use a range is way better then my +Ill investigate and fix this wrong behavior asap +If we simply do distcp then it cannot finish copying everything within a single dayAlso since currently copies files sequentially files in userdata will be copied first +I think this is one of the most FAQ. +Take a look at urlGet for a handy helper to get the HTML of a page and for an example of verifying the contents of one of the web HDFS web +Very coolGonna try to get this in for +Created to ask for addition of Sling to the Nexus that we are registered on the Nexus Repository I committed a slightly modifed patch in Rev +screen shot of the current repo config generation page +GREAT on you stack! Thanks for the hard workTesting now on our full dataset. +Or is this not how Seam catch behaves +That will keep them sorted when they are sent back to the client for the +Can you pl check if you can reproduce this condition +PR ready for that? +Andrey Ive committed the test at the revision +merged +I still plan to revert tomorrow +On my desktop it takes sec to import a project while a test runs in sec +Fixed in branchx and trunk +the jndi names are the samejavacompenvjavacompenvplease check if the jar contains services for a detailed we must also support the standard JSF names +Hm +went to review your commits to do a sanity check on them and make sure I concur but the most recent commit is five hours ago +Push to a later fix version +It should be Create a folder jsfTestresourcestest container for +Closing xml and console issues. +or should it just be rounding down at registrationIt sounds make sense +as there are additional issues discussed in up fix version. +a +The warnings with Cargo plugin also dont appear any longer but I think its okay because these warnings were just Tomcat startup info logsThanks for this quick fix! Well done +But it responds really fast and round trip time is always visible on S especially remotelyIm to this patch its not going to break anyones existing code as the operations would always fail to date +r fixes the script engine +The Java samples tests show this we allocate tens of ports for these containers +potentially compatible change bump to critical +Try setting your contianerId to profile was specified it should have been tomcatx +There are still issues that need to be resolved to get things like security and such working but those can now be logged as separate bugs +Need to determine impact to this take a look +Trunk fixed in revision . +Yeah but since the xml is not very complex most of it just seemed like grunt work and perhaps not too much of it? I have not looked too closely in a while but I would have approached it by just making a java object that I stuck the settings on and then pass it around same as the xml dom was passed around +fix verified for the repro query given here +Attaching patch aThe patch changes the access for static variables from instances to class names in three Derby classes removes an unused method in and removes some unused local variables I spotted when browsing these three classesPatch ready for review +Im meeting the same issuesaveConfiguration function returns a BOOL value and I believe it should be returned as the return value of docmd function +Please also warn that recovery may cause further data lost and tell the user to first backup the image +So sorry about this. +Attached patch solves the problem +Since I am currently working on and I have thought about how to recover a failed pipeline +David. +As you noted the problem is that you used the wrong namespace +So i have registered vpecontroller as selectionListener using selectionProvider +Same addendum for to all branches +I know this cast is not perfect but until we can use Java Generics I dont see a way around thisAgain this is just a proposal I am coming forward with because I needed a pipeline in another project and I liked the one in J except for the dependency +fixed +applied to and of patch +It will be fixed after the database is cleaned up +Its really an issue +Im looking into it further now +The comment in SQM should say set to false not set to true I will fix that +Meanwhile if someone else wants to the patch feel free +A router can only exist as the last component of a chain +What is the usage of? Support new special apps which need a local AM or support all apps? If the later it is far from complete so either disable this test or writeuse a simple AM in test +Were there any errors in the logs? Was this JSF or We need a little more to go on +got itwe had a todo on handling this for pure driven hiearchies that is now fixed in cvs please test +You can create a patch using svn diff and attach it to this JIRA using the attach file link +Steve Todd could you please state if you tested this patch in your cluster as wellIve tested on my laptop but not out in the wide world of varied deployments +Let me look at it and revert it if I can reproduce the build issues +This also happens to DateCalendar objects as input +for today just letting something external be notified when a reader is gone just as is privately notified today is a good baby step +Therefore retargeting to the release +I would like to hear other opinions +The shell with find EXE and CMD files you do not need to specify the file extension +since there are no more comments Im marking the issue as resolved. +Here is the patch +Just paste it in the H console and refresh the database tab in Probe +The exception appears shuting down the all config +Something like adding Status ok or notok +Its the Schema or S icon youll see when you expand to the first level below the modelSo dont think this is a previous note +Hi Kathey thanks for the info that is very helpfulAdjusting the if statement as I surmised in my previous commentdoes seem to make the repro script passIll try working up a more complete patch proposal for more testingOne thing concerns me we need to have a regression test which verifies that the MinMax optimization feature is working correctlyIll have to look to see if this information is available in the queryplan andor in the runtime statistics and whether we can use thoseto fashion a regression test +Moved the scripts to the bottom of the pageRendering of the buttons is now done in instead of inlineMapping widgets to region is now done by looping over DOM elements instead of inline calls +Hi Alan Looking at this attachment the tests appear to be run twice +Dominik I just tried the attachment +I think youre right +Therefore this issue is now closed +to critical to mark it as must fix in is this still major and targeted at? I thought wed only fix important bugs in from now on +The name of the folderzip is the package in the engine project +Hmmm +appears to be the ocation where the AS binaries under testing reside +NN detetected that the block was therefore scheduled to replicated it to +Are you building on Windows? I ask because this looks a lot like an OS or JVM related problem +Ive committed this to trunk and +and print a warning if utimestime +I just committed this +Ive looked at the and it seems to test things properly and the unit test runs without errors +Diffs should be sent to at +after that issue we dont have any fixes on cpp client so it is difficult to find this issue +Will ask Arber to see if it helps him +Another way to find this information is to run the following commandbinhadoop jar Does this satisfies your needs +Looks like Lingo is dead +Bug can be closed after that +Attached patchI changed these defaults Use so merges are run in the background +Patch applied +I added the possibility to add custom validators so that more than the default checks can be done but didnt remove the default checks +waiting for input form Barry +Guys this was integrated after the release was cut so it did not actually make it in +HelloI also experience a similar issue the number of connection rises above any limit set in the config +This would also require a rethink to accommodate any changes you have +I would like to commit these changes as soon as possible +The patch has the update for to make formatting of definition lists look more naturalThe screen shots provided show the difference in formatting +I should add for a default Nimble install like this account details areuseruseRadminadmiN +Ben what was the problem you hit to get the misfireInstruction to work? From looking at it a bit today it looks like it would satisfy your requirement +After closing this ticket Ill open another one for working on the docs +ping is used by the communication thread to ping the TT every sec +if you want to go that far then a java class would be used if I remember rightwill look into expanding itsince is say create should ask of things like graphics to replace in the directories +Martin +The forkMode setting should go into the configuration of the. +Will commit if tests pass +IMHO they unnecessarily bind the persisted items to specific node type settings +Patch looks fine except I dont see a reason to introduce new libraries just for asserts +Closing issues. +We dont want to introduce things like NPE into the code just to print out strings +Actually it looks like a Mutation can hold two identical keys but one of them gets removed when they are inserted into the map +reopen if still problems with h toolset +Uploading update +As it is WTPs issue not reproducible under Ganymede now +Let me know if you have time for doing it +This patch is based upon the branch and thus might have minor differences due to different code base from the final patch to be applied to the trunk +Agreed that would then be a bug for the Felix Http Jetty bundle right +attached patch switches remaining extends to extends for better coverage +Its not a beauty contest but heres the code +I think this looks pretty good MichaelToo bad so much needs to be cloned in the case that closefalse +This new patch just change a intSize boolSize in the serialiazed size of an +Remove ! operators overloading add Iclonable fixed +Phoenix launcher implementing the Daemon interface +Updated fixes in committed revision . +setting target version to as this is enables using cgroups without an important security issue in clusters +Fixed for in r. +M javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapi for methods not present in JSRM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiUser and password are specified in databaseName itself so that they dont get for other connections in the testM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreUse Datasource instead of M javatestingorgapachederbyTestingfunctionTestsharnessM javatestingorgapachederbyTestingfunctionTestssuitesChange jdktest property to jdbctest property +Just for my own understanding Right now you are setting output path string as revision key +Weve had some pretty significant bug fixes on file modes in archiver so you may want to test in a few hours when it hits this issue been fixed in assembly? +Will review the PR later today and if acceptable merge +Andrei please check revert issue that Lukas points out and when you resolve it please resolve it as rejected +Not sure what is being proposed here the incompatibility is an issue regardless of upgradeAn existing application that creates databases will hit the incompatibility as well as one that upgrades +Will commit to later today if there are no objections +Forgot to mention that patch does not contain the required jars and which I will also commit +If anyone creates a patch please post it +The central maven repository can download artifacts from the jboss repo through rsync +cpp are missing from the patch +the scope of this issue seems to be broader than originally expected Ive updated title and description accordingly +new page for remote ip page class for remote ip file to apply to tapestry trunk with integration attached implementation is ineffective if theres a proxy sitting between the client and the application server + +A testcase showing the problemChange the version to in and it will work as expected +Please use the same username as my JIRA account sirenianand the following public key dajrrcc your project despot powers to assign her to a project etc. +Heres a patch against that provides the behavior Im looking for + Updated output location of integration test case to prevent false negative scanning by rat plugin +I dont think Archiva remembers where it proxied something fromI am also not sure that configuration changes should be retroactiveThat is once you blacklist something Archiva will not do further downloads but it wont go back and remove what it has already downloadedIf you determine that there are bad artifacts coming from a certain repo you can use the delete feature to get rid of them +The to verify the fix +The console can do things to detect the web container and act appropriately but the console web app itself should not pull in jetty or tomcat elements directly +This patch will contain the necessary changes to create the axis eclipse codegen plugin project +Can this be closed +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I like what I see +Makes mention of and to trunk and to branch so has the updated grid too. +the patch looks good +I do not remember why these jars were missing but it seems to be ok nowIf so could you resolve this issue? +May be we should close it +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Resolving and leaving any additional cases to be logged on a Released +Hi ChristianIll take a look at your issues this evening +At the time we had to pick an algorithm for turning Xml names into Java package names and this is what we picked +Need more information from the OP to do anything with this +UweHow do you imagine passing the XML handling decorator toShouldnt the be as safe as possible by default +I this case Id judge it worth it +bbaa and +Also an important test that we should add is to make sure that leaving a shared doesnt leave it in a incosistent state especially if other components share the channel +I have filed to capture the fact that there are test failures in code in trunk +He receive order from the master to send package with given shipmentId +I think we should add a comment there about why we are doing it that way +Ok Nick as we in agreement there lets implement itI set M but if not fit please move as appropriate. +The resolution fixed the problem for streaming state transfer but not regular state transfer +Hi Ian Im ok with changing with some function of tickTime +Fixed with a constructor is for a java native class or a class in java +Transaction hbase is no longer part of core. +Is the release failure exception appropriate in the first place? +rbaxter doesnt exist the user is +This isnt really a VM problem as much as it is an issue with your system not allowing you to run the client +Clarification to my last the maven PDF compilation fails +Well need to deprecate the API in before we remove it +bulk defer of issues to +As we made very good progress on the last meetings I think we can just leave it as is +However I doubt this was the original cause of this bug +As Frank suggested before pointing to the comments of rev . +patch +Done added two new builds for windows currently fails tests so youll need to look at that the other passesSee juddi on IRC for comments feel free to tweak the config file yourself if needed. +For example Client will always look for token and try to use it if found even if configuration says otherwiseI dont understand this objection +If no one disagrees I think this issue should probably be moved to +But I dont think we document the new index statistics feature not yet anyway +Ive applied Stuarts patch please try this out and report any problems +I guess either way it would be nice to document the expected behaviour +Lets sIt is a little um funny to have simple regression implement multiple regression +I made the highlighter module depend on the query string parser and analyzer modules for testing + +Justin and RolandI have had to add a few more strings to the locale enUS bundle +ThanksIts obvious nowo +Its behaviour is just as I described in the task +Ah yeah Karaf does so we can have the version range as suggested +I have applied his patch in revision . +This is based on your patch butextended and integrates the concept +Editorial pass over all release notes prior to publication of +TedThanks for the reviewLarsCan you review the patch and provide your commentsIf it is ok i can integrate to and take another RC +Yep I forgot to svn add the new fileHeres a hopefully working patch +But if I provide a dummy location not starting with http then it gives a different error Cannot infer transport information from the URLBut I have rechecked the targetEndpointaddress it starts with httpand its correct +Marking patch available as I am testing patch available testing for to trunk +Ive added the settings. +The findbugs warning is a known issue in trunk tracked by +Is that safe to remove yui library +I tried various rename cases and the patch passed all of them +This generally then quickly goes on different lifecycles schemes logic naming etcSo I guess you can put that to medium or something +Ive attached another rev +I committed for now the impl without interface in the branch revision Comments about the clean new API welcome tests of course do not even compile +Can you attach your job configuration file +The mailing appears to be being used to ask questions about Drools Guvnor not Guvnor Core +Im also wondering if there are any generally accepted practices regarding giving the cassandra group itself write permissions to these directoriesfilesFrom the perspective of someone writing a monitoring application I would like to be able to have our packaging create its own user and add that user to the cassandra group and at that point have readwrite access to configuration filessnapshotsother things +another log message +I think its perfectly acceptable to require a core reloadCore reloads in the presence of updates have always been more scary +I have reviewed tested and committed the submitted patch for this issuemsvn commitSending javaengineorgapachederbyimplstorerawdataSending javatestingorgapachederbyTestingfunctionTeststestslangautSending javatestingorgapachederbyTestingfunctionTeststestslangcomSending javatestingorgapachederbyTestingfunctionTeststestslangsysSending javatestingorgapachederbyTestingfunctionTeststestslangwisSending javatestingorgapachederbyTestingfunctionTeststestsstoreacSending javatestingorgapachederbyTestingfunctionTeststestsstorehoSending javatestingorgapachederbyTestingfunctionTeststestsstoreteSending javatestingorgapachederbyTestingfunctionTestsutilderbytesTransmitting file dataCommitted revision +Thanks fixed in SVN r found two more files that have the same problem +Attached correction if the containers JVM arguments parameter contains host then the CARGO container will not set the hostnameAs a result in order to use +Still the issue here is not performance but correctness +Any operation that expects the object to be known the Session like lock is indeed bound to fail +When you call mvn install the first time it works but when you call mvn install a second time it fails with a stack overflow error +I think this a very good implementation +So we might as well go with what Java gives us +It doesnt look like Cache will ever switch to let alone but Infinispan still has a chance +The whole chain of all nodes from child up to the root should have all the child entries +Doesnt look like were going to sort this for CP so Ive move it up to CP +Bulk close for release +Is this some kind of problem +Shows the error popup when it occured I also highlighted the position where I was editing code when it for looking into thisI had the error a few times more since my last comment and each time I had to restart clean after making some changes to the file that was edited beforehandThe code is fine I can run it ok +In some next moment the can keep only the component path for the UID and not an weak reference to it +Ok the changes have to submitted as an svn diff file not a zip file +Hi Would you please try my patch? ThanksBest Regards patch applied at revision r with minor changes thanks a lot for this enhancement please verify that the problem is fully fixed as you expected. +Good idea It would be nice to have a sample EAR to submit to them +passes on Linus and Windows +Its default balancer relies on there being no per table balancer and in fact I think that it should only be in the old balancer at all +Brian this bug down +Sorry I release this ticket +The header of document and its mimetype is different +Now you are trying to update the parent region server name into the daughter +yQ +as comment on i am not in favor of that move +Note This patch includes changes in the previously posted patch +I think sorting the imports exports by default would be nice too. +This patch only add support for using as query result +The attached patch renames the version to and split the dist module into two modules and so that the two tarballs are deployed under a different name +Martin why do you think we need to do this? +This is done r +swamy is this specific to or happen in any of the HVsthanks issue is not seen on KVM setup and it looks like only happens with vmware +This is a hacky kind of fix to the existing JSP pages +Paul do you think you may have a look at this idea? What do you think +This was completed as part of Ill add a changes entry +svn ci Adding srcmainjavaorgsuperbizattachmentAdding srcmainjavaorgsuperbizattachmentAdding srcmainjavaorgsuperbizattachmentAdding srctestjavaorgsuperbizattachmentAdding srctestjavaorgsuperbizattachmentTransmitting file dataCommitted revision . +AK found the issue. +Committed on trunk +Thanks Amareshwari +the fix looks committed. +Hi ChrisId like to clean POS issues do you have something new about this oneThanks +Seems to be a issue +there is a critical security risk +Need repository or some details about jPDL editor +the heap dump analyze of reducer +The other option would be to use Properties in the api but just serialize it as JSON +Added licensing sectionPlease review +Looks like the adding of attachments was only executed for type all tests pass this patch looks good +We will need to make our master pom available to them to use to get all the versions Cocoon is using +So I would leave the name as it is unless you fell strongly about itthanks again its great to have such a thorough review +Would glady submit one +Ive just committed an amendment to include the new file. +Its been an issue with addressing extensions within that particular WSDL +Ill make that change +The basic mechanics inside the IDE seem to be fine +Project Leads If you want GA for your space send me your google account ID +Thanks for the quick fix! Ill use the custom built JAR for the time beingAny idea when will be available? +and this is not a issue imo +Actually hedwig console supports run commands from the commandline like binhedwig console show topicsAgreed to add some commandline options +Fixed for +I think it would be a good idea through to leave the module directories all in lower case to reflect the artifactId +It may be fixed in later versions I dont know but it was a significant problem +Can you try with Camel as have been improvedI think currently it will update the cron trigger if you got the same trigger nameWe could introduce some option you can enable to let Camel barf if an existing trigger name exists + unscheduled issues to the next release + +Is this critical? Id prefer to bump it to the next release +It is not! has the correct hashbang usrbinenv jruby +Fixed in r +I would prefere this tests to go into CTF Testsuite rather than examplesAs not all combinations of possible setter can be tested the cases tested cover the important and test coverage + on fix +Ive just seen your commits +change gatewayDirectionConverging to gatewayDirectionconverging i think its the problem +I agree it as an improvement rather a bug +Please dont open multiple issues for the same problem +use the archetype and just create a very minimal example that duplicates this +It should have been instead of is what should be used there +I thought the problem was concerning forms of type list only so my solution was addressing these types only +Most likely if both were using ths wouldnt happen +Its hard to tell just from looking at the jar files in the lib directory +I just committed this. +Sorry that I thought was used for checking overflow +The rest are mostly internal classes +It must be planned orchestrated to ensure a smooth transition with as little disruptions as possible +The two queries in the test both fail if they are executed +Neither should be all that bad since there is HTTPS support in both jetty and the apache HTTP clientKey management is likely to be trouble +Robert Fixed in r +Thanks againStill reading about the Listeners +I will see if I can split the patch +Another situation is if I develop a web app and its target run time isthen the app is not able to run on geronimo in eclipseError below will occurThe server does not support version of the JEE Web module specificationIf I export the web project as +Ill fix it and resubmit the change +The will only be set if the instance is a and action listeners can be added for both and +Pushing ppc problems off to dont think were going to be able to nail these in time frame +if you dont have any of those in your table you will have rows and cause huge problemsWe should protect our users from this very nasty edge case +For certain problems they should yield the same answers +Setting release note works as expected +This patch allows deletion of partitions by specifying any subset of the partition namesThis should be good enough for most cases for now +Hiwould it be possible to add a fix for too +So we have the same bug report in bugtrackers and command line tools which do the same thing are in repositories +CI fails due to switch from to Firefox +The should be used in EAP instead of in community project +I read the but didnt get how to fix? Can anybody give me simple stepsThanks +released closing. +I dont think we should respin just make it the default MP forBut I think we should use lucene to conditionalize itIts a biggish change that doc can now be shuffled so in casethere is some Solr app out there that relies on doc staying in orderwe should respect the lucene +bq +Touches the following filesM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrda +Looks similar to +Pushing this to and beyond +Ah shoot my specs got lost in the process of building the patch +OK lets do it +Some recommenders cannot do this because they compute internal numbers based on the prefs and then toss the prefs +Worked as advertised +updated. +And if moved to Descriptors should the same rules for adding classes apply +This is a problem for languages that have for example an anchor embedded in a word +We cannot fix this for this release as it may need too much code that we must need to say in the Release Notes for EAP perhaps that this example should not be used for now or some workaround +I have migrated to the new style the eBay application +Can someone please help me with the steps I need to install the fix for the localstorage issue or is it included in the Cordova install package? I appreciate any helpRegardsFrank +It was a long time ago +This pacth is also targeted for the release so it needs to be on the branch tooIdeally specific referencesimplementations should not be made in generic concepts +no more noise +htm +See +I am working on the pipeline FI tests +but I want to controll where my temp files are created per instanceIt would be nice if I can set +You are right ClausThe second commit is just cleanup. +I will need to figure out some other way to make it work on x +xml configuration +parameters to be neglected +The default it and should work better atm +Attachment has been added with description Patch for the test has been added with description Patch for last everything seems OKThank you again Glenn. +Hey TimI know you are super busy but just wanted you to know that if you have any troubles with the book or need any asistance be sure to let myself or Samson know and we can assistPlease read the above and provide links where we can review each of the bulleted changes as a priorityThank you very much +Let me try to draw a clearer picture here of what I had in mindIm sitting here in front of eclipse +Issue always happen when open ESB Editor in tree panelIll ask to they to clean their PATH env variable and try again +Make as Incomplete +Hi You can find a sample which demonstrate the XML Stream API base approach here +Ollie +If you like to see such changes to apear in an earlier release we are open to review your patch +I know we need to fix this issue but changing the code at this point is not good idea +Can you attach a test war to this JIRA issue and continue discussion at the forum thread with some new concerns I have +Thanks for the contribution and apologies for the long time it took to apply it +The random jitter isnt really an issue because the compactions should be put in the large compaction queue and not affect normal operation +Because many cjk developers are watching the status of developmentThey are high level developers +Hopefully we can get this into +This works in geronimo assembled using blueprint but I dont have a good way to test the spring assembly which may well not work properlyPlease let me know of problems and I will work to fix them +If you give me a few moderators I can add themIll also try to clean up the allow list based on the committers list +Closed upon release of Hadoop. +Yeah I just want to verify first whats the situation maybe Im missing something. +It wasnt my code that was attempting to resolve an address +Heres a backport to as no objections committed to as well. +Karthik can you point out which patch? Anyway this really only matters for when we merge +Attaching the initial data model implementation +I agree about changing the reservations +At least thats what it seems like to me. +Version is attached to add the fix by right lets go with the Map versions then +For this reason JBPAPP can be closed only by QA +Rick Ive just backported this to in the course of some broader backporting +I have made a patch with the after the set and the test passes with both and sun +maybe not? I think the runtime is the same in either case +Test issue only. +Your fix worked +Patch applied +The real variable name is +Im willing to leave things as they are for RC +Verified in CR +I started a PR for this +Note that is really a static file +Attached patch a which fixes the int overflow issuesSince that arent transferred using layer B streaming are represented as this results in a maximum length of GBRunning regression testsPatch ready for review +to add fix comments and to change the revision Changed the references for the faces plugin to +WFM +The change breaks a unit test for another part of the system so I think the test needs to be fixed as well if anyone wants to build fix adds a nondependency element to each maven dependency +My talking +The semantics are the same though as it takes an optional array to fill in with the results of all the ops in the multiop +If you empty a mandatory parameter the build should fail +Ill change that +Needs investigation +Im able to reproduce it with following settings cleared temporary filesApache Tomcat JRE hope I can give you access to machine where you can reproduce it +Hi Lewis Im sorry I generate a duplicate issue +Can you as a p patch +This should be resolved +is now out so closing all resolved issues +Marking Patch Available +Though I understand what youre going for I think wouldnt Include Field Names in Output clarify where theyd be included a bit better +But I saw no issue with the link to the screenshot pageBy the way thanks for the report +You likely need to add the versions of and as dependencies of the plugin +This patch has the following major changes +Thanks DiveshYour patch is in trunk rev +Some risks can safely be ignored others may subject you to patent litigation +keep all auth config in one placebq +I just wanted to make you aware of the document +I think these have finished all the documentation works for +v contains a real fix on but the logs are still redirected to the console to see the other stuff +More interesting would be to know whether you can reproduce it with Java ? Java is not supported for that version + unscheduled issues to the next release +That is fine for me +Resolving the issue since did not heard from the reporter yet Please reopen if error still persist on the reported environment +Thats fine +My approach would be to have a command to add partitions manually I have created a jira ticket for it +Document these in and its good to go +Im still lookingFrom a purely personal motivation it looks like the system that produces the TIFF files that Im required to process is going to be pretty soon +Lets close this issue for now and reopen if that is the case +It should be able to evolve at its own pace independent of the Accumulo release schedule +None of the lines of code in the stack trace seem to come from jaxen +I can also compress build log with zip +Thanks Nicholas. +This one fixes as well and checks for back compat +a +Skipping Hudson since it wont detect thisI have committed this +the patch looks goodRodrigo Just a minor nit can you add some comments that propsplits is a different column if the entry in index is a directory or a file? Its a little confusing when you see the code +So if we just pool max chunk size buffers well probably come out ahead +This issue is related to the fix +Patch welcome. +Patch applied thanks for the fix AntoineTests would be good but as the class has no other testing already it seems ok to not require them for othersI needed to get the release out for a fix for another user and as it was already being voted on it would have delayed it to long to include this +New features should be introduced in the Beta to Ron +By specifying i mean defining cassandra specific properties and loading them implicitly just like having any sql client holds or held by Hibernate. +Hem this is actually may be an indication we forgot to extend the sld schema embedded in that already contains most the extensions +Attaching a early version of the patch +Was fixed while resolving how the user should be able to disable it? Disable Support on this project? Or there is some other way to disable only validation which I didnt notice +See +We have a few other Mac users who can also hit thisOn a related note one of my colleagues also ran into some failures on Linux but its not clear if the root cause was the same +That is way I said side issueWith Railo can you build a web service with GET method that accepts BODY? +We might be able to change this if we have issue for more info +Unassigning jjamrich +thanks Leif the ATS is good! this is not a bug! +Matt you may want to remove the redundant key attribute from as part of your patch +I have committed this +test has a test that check that the cache is empty we could reuse the same concept on the to make sure the value in the map is truly what it should +Actually the test fails in the interpreter mode only so please commit the updated patch +I can see this behavior simply stopping the consumer ctrl c after the producer starts sending +The patch has been appliedPlease verify. +Posting new patch +Then I believe that you could exclude the projects main artifact from shade and ask for the classified one instead +If someone knows the solution please let me know if there is anything that i am missing or is there a patch that i need to apply to avoid this youMansoor. +The now sets the actual Property on the +I have done exact the same as you and youre right both bundles stay in the resolve state +I think is about your greater point of more robust exception handlingLet me link this issue to that one and see if I can come up with a unit test that repros your exception aboveThanks James +Patch v for hadoop which incorporates Nicolas first two comments +solution is to use encoding +Ill commit it soon unless someone objects +Add samples documentation +Thanks very much Bryan for the review and commit of this fix +Fixed along with a test case. +OK I understand now +New patch addresses Alans comments +he need is always didnt find how to fix it +Essentially we want to be able to add an arbitrary filter query to the request based on some notion of the user +issue is fixed now. +Reopening to change the Fix Version to future since tooling patches will not go into the BRMS release +The remaining thing to get working is links for inherited fieldsmethods +The patch should work but it contains some patches toHowever not much changes there +Thanks. +I saw the app token in pi jobsRemoving the appTokens file prevents a leaked app token from squashing the app token in the child AM +I removed the redundant null check Knut Anders commented on and fixed some bad formatting in the string created by +Also casts bytearray objects to TupleMap in and! operations +separate task +Making s a patch that adds how to add a metric to the documentation +The following property names have changed from the previous comment to make parsing easier +this latest patch cleans up the logging a bit more it also adds Nio client Netty server based unit tests these are a subset of the base tests but using the netty server cnxn factory +Yes you are right +If takes some more time can we check this +Please move back to if this still requires code changes +The benchmark terminates early as reported in java Xbootclasspathp jar n antlr antlr starting warmup Running antlr on grammar antlrcppANTLR Parser Generator Version +Therefore action are indeed updated and will pick that coord job up +closing this the and builds are cleaner now for me. +Required copying code from Sling class but I believe it was necessary +According to the test case it fixed +Can you elaborate on extensive testing +If RS went down after setting DISABLING state then we are unnessarily assigning and then unassigning the regions in SSHThis we can solve using table locks then +DohThe issue was caused by the creation of a Display object in an implementation +Patch attached +I worked around this problem by adding from the Groovy lib folder to the Eclipse project build path +Id much rather use the standard because my method also doesnt pick up the dependencies +Bob what OTP version did you test this with? There are a number of RB fixes for various SSL fixes that superficially could be relatedSimon if you can reliably duplicate this you might want to try bumping erlang to and seeing if this still happens +Shorting out the NEGOTIATE for a becomes a bit complicated +The zip should be unpacked as a direct replacement of the workingjdktoolsmodulesjpdasrcmainnative directory +bq +Recreated the archives +thats a good idea +Are you planning to work on thisAgree with the blocker status +The courier is sufficient for and will remove this entirely. +pending separate patch for trunk +Now that LFs are fixed in I resubmitted patches to both sources as an attachment +Sets the mappers to for sort +that will be problemThanks for your help very much +expects the data to receive in order but the seda is async so some exchanges can arrive out of orderI have added the produceDelay to the unit test and fixed the seda consumer to not minimum sleep for millistrunk all tickets +I think the right thing to do is to use the verboseness level to determine whether stack traces are printed or not +You could add the Equinox JAR file to the class path and boot delegate that specific package too and see if that gets you any furtherAnd yes the +During the investigation found out that JDBC Connection Properties View was changed and no longer corresponds to reality +I dont think this belongs here +I can fix the comment on commit +Defer all open issues to +can someone reviewcommit this? i dont think i am going to make more changes to thiswill work on long regression framework separately +example now has the option to use the drl or xls file +Once we have integrated the IO in the peer we can use the but just with another sync serviceThis will reduce our duplicate code and we dont need these Thomas please fix the localrunner tasks in future +introduced a concurrency issue with being accessed outside +Small patch attached to add the Java Builder into the +Tom and me have been removing and adding stuff around this all the time but havent found a good solution to it +no response from submitterfeel free to if turning the cache off does cause a lot of annoying problems for which you have to be prepared for +I had to dig into this because I need the older in my project +Lets figure out what the user interaction surface looks like and carefully create new +Is there a way in the +The second is that this would cause a redundant error message to be logged +Netbeans is really not very convenient +thanks MikeContinuing testing +I think he meant this should be also validemail email have a look at the patch +Actually then I want a clarification here + on patch +add unit s the reason for giving access to the entire KV map? Note that I just added which assumes that unreserved KV types are config values +For now we can imitate HDFS HA +From a user perspective I have a problem on my hand if I cannot accurately determine the number of messages on a single JMS destination +thanks for the tip re joda +This was fixed after was packed +Hence closing the bug +I missed this one. +Can you run the test with latest change and confirm if the hints are avaialble with target Fix please consider for the branch +Some changes are made for better comments inI finished some preliminary testings such as and on Windows and Linux and all of them are successful +As I see it it is the behaviour of the Field isDisabledisReadonly methods that determine this cascading behaviour +So the problem like i said before happens when i stop theserver when it was publishing after this i try diferents ways fix theproblem but the message error is still imutable +Hi Kohsuke thanks for the patch +git might be higher up but that will be found by egit automatically +in the Standard container config +Tools plugin needs to use the latest jar +Rationale count is not updated for WFS in the case where no global max feature has been set or in the case in which there is just one collection to be returned +I am not sure would need to discuss this with the other project members +Even when using that this behavior still occurs +The fix could be manually backported if specifically required on an earlier branch. +FOP takes as input a FOdocument that is a document conforming to and produces output in thedesignated format +The current documented functionality of the ignore attribute relates to the bean identified by the name attribute +I might decide to add overloads which accept an array instead of a collection and want to be able to view them side by side with the other static functionsI have a class which extends the +behvaiorsajax calls is not always +But I will verify very quickly +I didnt know it was so sophisticated +Does it notThe lock file is the right way to determine that the is running not the web server +We already have some benchmarks in the examples source tree +Meng please verify it also +I didnt even think about impact on data caching +bulk move of all unresolved issues from to +and make the failed hash table insert run back into collapsing in the next time it will not solve the cluster problem but will make collapsing much usable in single server env and get cluster debug much clearfrom what I have learn from the buggy codes I really want to move collapsing out of transaction why you put a decide function in DNS? and looks the author is not familiar with most of the httpthreading codes we really need to get it reviewedreimplementedafter my patch I have identified one cluster write bug it will make collapsing evil +You can still run +He needs to sort out an issue with a dep on a hl osgi bundle for component +changes patch to make the changes in the abstraction layer backword compatible with and subclass Exception as initially For you want to use the other constructor and setup the causing exception via initCause +Cmon folksThis issue has been beaten to death on more than one occasion +Also this never appeared with jboss GA +Before the fix it was something like mb +Yes +To be finished as part of comprehensive tutorials in the timeframe + The link in my previous comment points to the description of some initial work for this issue Thanks to Jay to mention this issue as I completely missed the fact that there was already an issue for that +This patch does a bit of clean up and refactorings of +Turns out theres a VDB folder URI causing the problem and ONLY when there are SPACES in path to the imported model like Documents and Settingsworkspacexxxxxxx +resolving as WONTFIX since we cannot support the unofficial at this point. +It looks like a copypaste error from the +Ill attach the fix for that to and will look at this issue later on +Lohit as Jason mentioned should work for you +Some of these issues are awaiting retesting after upgrade of ODE version and merging of the JPA work from the ODE trunk to the branch +I will guess that we need something like mvn execjava pathtomyjarand mvn execjava DjargroupIdartifactIdcan the one watching this issue give some feedback? +But that use case is likely to be rare so i dont think it needs to be considered for memory size limit estimate +Ill look at an alternate check to do instead. +Trivial patch skipping RB +Thanks Binglin! Upload the patch to fix +This issue is no longer relevant for it is fixed inChanging the affected version to. +Include more properties in producer property file +Last comment provides the solution for JDK +Note without the packaging option a src directory is created in the root directory which is to be expected +We will review and discuss it soonRegards MichelI reviewed your patch it makes sense +I am unassigning me this due to time constraints +Patch applied +To commit it separately we must have separate tests +The variable is not assigned when a new object is created so session data will be piggybacked every time not only when they have changedIll upload a new patch +Can you please run your test with the +On the whole though that should be another JIRA issue with a pretty low priority +Eliminates of contended locks on +awesome thanks +Id like to go with the solution proposed by Mike +Rev close of all resolved issues as part of closing items. +So I could add another bit indicating a field +Also just because you dont make changes doesnt mean you cant update the version +update set to Harmony jre home +Also I would not mind at all if the Hudson or something would publish a nightly snapshot of the Lucene projects to that repoYes we could think about doing that in the futureFor now Im planning to commit this patch shortly for unless thereare objections +Already fixed as part of another ticket. +Ill have a look at Python source code and see how HEAD requests are handled in andAs far as only running the code in some versions we use. +If you pop up on IRC Ill be happy to walk you through the git commands you need +Ill wait with applying until its fixed! +Hi Paulex Anton has agreed to open to track the attributes problem +The biggest reason why is theres an awkward hack to the fact that a Solr cant aggregate multiple values into a single +No the latest version is and youve reported this problem against +Thanks Chris +I will attach a repro showing that +I am going to commit soon unless I hear objections +Last patch applied in revision +IMO that felt quite natural +I see that is fixed in the trunk and in the branch +picked up this trend and used it for polygon symbolizers too but as the name say vendor options are vendor specific not portable +The client parses the SQLCARD and if finds null it sets connectionNull variable in to true +Closing resolved issues for already released versions +lgtm +This only happens when running under a security will be fixed in the next classfilewriter release +Seam numberguess ear where the has a manifest with values that say from numberguess jar +Checked into SVN. +Instead I would recommend either of or I could see Solr being a top level project at some point even though its directly related to LuceneDependencies like Lucene are already available in the Maven repository +Anyway this is another case and another issue +Thanks for the on the changes +Thanks Nick for adding the alias +for version all is OK +Agreed great work Eli +it was the browser cache +The attached patch might fix the graphic image so you think your patch will fix both sgraphicImage and the email rendering bugIsnt it an example project for multiple war in an ear? If so have you tested your patch on that? Because there is really nothing special with my again I will try it out when the nightly build is done +I am glad this patch did not make it into the release +Committed +fixed in r and close for the release +I guess I will have to look into it more +You mean if the server crashes between a successful send and an ackThat doesnt matter since the actual persistent transfer is always done atomically in the database +JimmyStack also i have rebased +Damn I got fooled read here should be pattern +This can be confusing +Marking this jira as rejected because it was a problem in test design +This patch improves logging for Cassandra mapping +Modified the service receive in oagis component according to the following todos +Theres a CVE out there somewhere that discusses how to inject control sequences so that if someone tails the log as root itll execute arbitrary commands +Thanks +I know that it is not yet released but Im working against trunk so I hope that is okIve included one test which shows the expected behaviour with a xml based configuration +Ah I see +I have tested with the latest version from CVS +Fixed master for to remove white space diffKathey +Use consistent sorting also in classpath +Ive used the demo in + applied on branch revision pretty having quotes on everything but this patch is better than breaking it approved +year passed Not fixed yet in Seam Seam team agree the proposed fixIt is OK that I fix it my self in my projectthanks a lot +Hopefully theyve seen something like this before. +I concur with Ari +Patch Committed revision +HiI just pushed the changes +SMX bundles have been branded for FUSE and set up in +Good point +All jars are fixed. +We should test as much as we can +just run and see if it works +There hasnt been any additional comments on this item since Ill finish out and move that into the examples folder in svn +Seems like this should go to the branch for too +It seems that this is an issue with class loading +issue is fixed in r in trunkThe correlation id is copied from the in message ctx to the out message ctxThere is also a test case added to verify the functionality is not broken during future modificationsRajith +Sorry my Fault +They did work to dynamically generate new metric names etcI think we could work on this at the same time we start to think about using the info for better load balancing and such +Committed at subversion revision +Thanks Timothy for the code snippet +The task I ran is as described in my Comment. +to be +If there are issues I can remove the fix or modify the fix if there is a better number +Again this works like these properties are used for sys prob variable substitution in +disabled and selected are absent +However unless someone wants to write we probably have to live with itCould I ask what bad things happen if you leave the jstl dependency out of and? Also isnt jstl a jsp feature? If so I would think that the default environment would be an even better place for the dependency and it would only have to go in one place +It will be taken care once the job completes +Most of this is already possible I guess but would need documentation and an example +The proposed patch adds a configuration parameter to the Management Console as shown in the screenshot to allow users change at the directory where storing temporarily uploaded files +thanks for the work on this one + +This would make registration automatic +Carl thanksNote that something might have gone wrong though the changes were either not published or not propagated because the side didnt change it was last published in October . +If you think they are critical for CR please them +Correct somehow I looked at the return value of readline +Updated in ESB branch will be merged into to release notes as resolved The performance of has been improved for concurrent access. +JacquesDo me a favor +Logs when running using m able to reproduce Exception E when I run the benchmark concurrently +Here is a new version without the property changes +But only with a TTF marker with the star its working fine? And with the slash symbol too +Updated affects version +In the test case I suppose we can be a little more strict in verification +file contains the keys that are assumed to be rightly named and will not be changed +See linked issue for more details +If geoserver can overwrite the properties file when an error occurs the exception generated by the error is not reported +submitted in r +Did some more testing but cannot see the error +done for tomcat at +Fixed as part of Bug . +Couple of comments +It does make things a little less convenient Ill agree +Your patch contained quite some garbage just changed the related for the mess eclipse seems to have converted the newlineswhitespace +Open to comments +BTW Id also add some information on how is integrated with AS and how to remove it +All unit tests pass +Thanks so much +Note that modulessrcmainresources should be deleted as part of this change +even if it is only a configuration I agree with your last point your solution has no real penalty +Nothing in the BYOE feature prevents future update +I dont know what the plan is to implement multiple ports +Addressing a very minor nit and uploading it +changes to to get the CTF test suite running for the new test cases TC to TC +Fixed as part of +Ive not used meclipse before so I started with versions will dont have proprietory Eclipse settings for local repository and will use what you have in +Another twist on this +At least it should log at WARN level insteadBut I would prefer to insteadThen that option limit can be used to bypass and then you can have if you really want +Attached latest patch addressing Ivans comment and also refactored tests +I linked to trunk but the same line is in and StevenIts my bad I didnt update the code after I copy past itI just commit a quick fix for it please feel free to check out the code and verify it. +Currently I see two timers for hiding the message One that is automatically started and removes the message in seconds +It works fineThanks Tim. +Let me start with the version in stargate +Minor request please open a thread on the dev list for these kinds of architecturaldesign discussions + is also used in and it seems tricky to remove it from there +Duplicate fixed in m and upcoming +I believe this is already fixed in the latest trunk codeline please in rev +bulk defer of issues to +It sounds like you know more about ARFF than me so what do you think +It looks like Eclipse issue but I need more invstigations +Will close after we QA and doc the feature +spring fine by me to discuss the use of spring in another context +then add folder +master broker full DEBUG broker full DEBUG masterslave removed in upcoming +Is there anyway I can get a patchset for so I can get you the exception that you need +I believe that the upshot of this is that it works as designed +You havent left anything for me then +Added this to the mojo a docbooktransform on the existing docbook directory +They will however be closed when the day is complete +Patch for the encodeNamespace accidentally clicked the Provide patch link and changed the status of the issue now I cant revert it +Please help to review and see if the workflow and logic is correct +Hello WendyDCurrently the build agents are grouped in build agent groups and use installations from their build environments during building the improvement is to make build agents utilize their individual installations configured in +Fixed on and closing all resolved issue that have not gotten any more feedback or comment in the last month +dont think i understand your comment +dims +Need an extra configuration in the to figure out the bundle info +The patches have been appliedPlease verify +I missed Mikes comment after resolving this issue +The envers listeners will still work as is if no audited entities were found +Looking at those two line Inject private resourcesand line Inject private componentResourcesI think the second one could be omitted right? +solved in CVS +James can this issue be closed +One more thing I forgot to mention in the bug report Funny enough openingreading such files works while for example? doesn one more method is missing to make attachmentfu happy +I think it should throw anHmm +As compared with the earlier alteration of the interface this change doesnt leak into unrelated areas of the code +I think it would be helpful to address bugs whether are from vm or classlibknown issues +When immediatetrue the returnListener is executed but you cant see the new value in the browser like the PPR didnt work +Should be fixed in the upcoming IBM SR release. +While were on this topic +bq +So we need a more complicated patch +Thanks +The original issue has been fixed with reusing if this already exists in the new version please create a new issue for this with a test case. +Can you attach a unit test for forward regression on this? Also just wondering how your patch fixes it not because I dont agree with it but mainly bc I dont understand +this modifies the original root pom +user can themselfThe current api is bogus +I got correct HTTP response when I run it under JDK but I didnt got any status code when I run it with the option JimIll also try your suggestion but this is Oneway service so I think HTTP status code and having body is still wrong behavior is that right? It should return the response with HTTP status and +Is there an upper bound on how long collated terms can be +Verified with JBT +Now the table returns expected +switching on patch available for patch for working with myself done for now +These tests have been rewritten. +When I ran mvn clean install it succeeded but had the messages noted above +Im not entirely convinced by the spec compliance argument +A couple of comments but were almost there since its probably the browser makes a reconnect to since its probably the browser reconnecting A test to confirm the header overrides the since value in Add a sentence to +It looks like this xalan issue us our root cause +I will set the following to WARN else that could use trimming down +From the standpoint of a build server the last startstop request should always win +Fixes the NPE in representative points mapper +It would be better to create a different method to read one byte which would then call the read method the returns the unsigned charI can probably work on a patch but I have never work with an open source project before +perhaps a solution to the password visible in the URL would be to use a redirection done by a web page with a form instead of using the HTTP redirection status code +Note this secuirty policy used here does not cover simple actions. +It contains as soon as is committed I will upload a new patch +OddThanks again Mr BishWill try to update gcc though Im running out of ideasDont hesitate to ask for logs tests +If JNDI lookups of resources is allowed they should not be citizens +The change looks good +Ive just committed this. +Thats why you will see build successful even when running ant checkstyle +Already fixed. +The exception is now caught didnt cause any damage though I checked and the code was changed there and we dont interrupt any sender anymore but instead cancel a task +Unassigning this so it can be picked up by a committer for inclusion in applied +sh file to toolsbinNormally c programmers hate working with java so if this script is available what they need to do is download the java snapshot set AXISHOME and then run this scriptThanks added and +Behaviour seems to be quite random to me +Currently clustering requires all blobs to be stored in the database to ensure transaction semantics +Committed in r on trunk and r on. +Suresh Thanks for taking a look +I will try to create one and I am sure it will. +following Java the toString parse methods dont include the suffixes that appear in the source language to distinguish the different sizes of values +what is the requirement behind thisthe aim of the simpledavservlet in jackrabbit is to allow webdav clients to access a jcr repositoryand dav clients are known for being limited in their ability to deal with authenticationon the other hand the sling authenticator adds all kind of fancy configuration options that donthave any meaning in a webdav client server setup and are prone to simply disable thewebdav functionality +Closed upon release of Hadoop. +I just committed this +Can someone else also test it +Github will then update from that +Ive just committed this to trunk and +Hi Antonio +I will try it on my side and let you this is not related to the version number of Castor you are using +This is a rare case and as long as you have long enough pipelinethis is not a problem +Ive submitted pull request +of course acknowledging that container level configuration could exist would imply that not all instances of the container will act identically which starts to head back to your point of not wanting blueprint to act differently based on the state of the framework +Is there a to this? Im having the same problem +I changed the JIRA title the relevant components and removed the references to clustering from the description +I got the development environment set up now so I was able to test the fix and create a patch +Verified by Richard. +Hi SaniCan you explain this ticketB +RAID is in an very advanced stage now +Go ahead and check into main +fixed with revision +Thanks SumitYour patch is in trunk rev +validation aspect required +You could also say The syntax for the for a new column is a subset of the syntax for a column in a CREATE TABLE statement +May be because of same name the QA did not pick it upPlease provide your comments on the latest patch +Ok I think we should actually set the default in to the real value instead of or +With fixed this does not need to be opened any longer all tests passed successfully Jacques for taking care of this +This issue is hereby resolved +Youre totally right Im not sure what happened there since I obviously allocated a variable for the future like I was going to do the right thing and then just +I upload new patch after rebase thank you +here is the workflow +Yeah in this case updating an existing pull request should be fine +There was a bug for this already +HiI am developing my applicationworkflow using Drools +Fixed and deployed to snapshot repositories of and branch. +I will talk with him about this +Will create new issue to update docs +Please see comments +search results of searching clusterName in hard coded names should be replace by following items defined in DEFAULTWADICLUSTERDEFAULTFARMt fix this in an issue according to our instructions to setup wadi cluster. +Let me try to commit that fix for you then to see if that helps solve the GertYes it looks like these tests pass reliably on aix and solaris now with the longer receive timeout +Configuration elements are not components +While nobody objects Im going to prepare update for all modules +You can control the policy via the two system properties and +Pushed to master PR for maintenance is still there +Thanks Adrianrev +Rich please take into account that introducing a change that relies on new code in Axiom at this stage will probably delay the release +Sorry for the confusion and spam +Marking this as resolved against older version since this issue has been fixed long ago. +Hi great work gentlemen +I think its ready to commit +Since this just modifies a test and not code dont think we need to wait for Hudson bot +But once this was sorted there still is a problem trying to locate a port so thought best to pass back to Alejandro to check out +modulesshell? there is already a shell in +Assigning to Frank +Here the patch to fix the pagination error +for pipeline writer +A couple steps svn diff log into JIRA and this issue click More Actions Attach File attach to itThanks +classes will removed from the domj jar as well +has been released closing issues. +Here at Comcast we are happily using and variants using Marks additions and would welcome the solution you propose or a mixture of the two +So since we have a plan to move to maven if maven plugin for nativeascii is provided Ill check it +Fixed about a week close of resolved issues. +So Im closing this bugA polite request please ask questions on the list before raising a bug as you will get an answer to your problemquestion sooner +Thomas Patch applied with a few minor xml edits but documentation itselfApologies for delay in applying. +Thanks! +Please have a look +And the percentage is but I would expect instead +Thanks for the report! Ill take a look as soon as I can hopefully this weekend +Previously I already experienced this problem on my home pc +wait a sec +deploy both new download section and other changes +Bulk defer all frontend issues to +Another solution would be to remove each you added +Hi GlenCan you check this with the latest trunk i think i have fixed thisThanksThe bug is still there on SNAPSHOT as of this morning. +I am not certain that it should be a copy so I am not touching it in this update +patch looks good to me +If you know of any let me know +Any further complaints should open a new issue on the subject and much more specific in its description. +Im not familiar with the code but it seems to me that the code is testing the secure set up +Fixed +Marking for so that it can tracked accordingly +Same patch for +Ill leave this issue open for a little time to give others a chance to does overloading the name with offset and length work for cases where the filename happens to contain +I have doubts about those patchs also but a third option is to change Pagedirty implementation to always give up executing if page is responding an AJAX +I used Robins document schema and in the form of a unit test and it gives an empty spell check response but no exceptions +I will attach the patch tomorrow cause I have to run now +Nik please work with Shane and Clint on this so that we actually have an API that addresses the use cases of other view layers needing conversation management code +Hi Justin is there anyone still working the module? Or can we just assume its dead +The problem I had with logging inside is that it is somewhat problematic to get a logger I should add at least a print out for now +Ill be glad to see some code for this idea +I think the parser is getting stuck in a particular mode because of the exception flow bypassing some resetting +In addition it starts up its data receive server in the beginning of offerservice instead of in the constructor +We can commit it with the understanding that we will overwrite once we commit hadoop support into patch fixes two minor conflicts with the current pig trunk +Logged In YES useridHiexact the same problem here running linux fresh eclipse and doing the same steps t got it Username letiemble +The people themselves will take care of them +The interface and implementation are already separate +jgray recent OOME in +Thanks for the update Amol. +Please provide the full stack trace and your file +I saw this issue on the branch +This patch is for Apache +Hi NIm Aleks over at Cloudera +I see the same thing in Aruns Datanode logsDoes it have stuck clients? DFS is expected to handle connection errors +Why is it a future requestI need this now in production +Added one more test case +Help in resolution is appreciated +I didnt touch the cvs code for a while so Im not setup +The jar is no longer includedIt shouldnt be should it? Avro used to pull it in as a required dependency which was a bug and Common explicitly had to block it +Thanks Alexander for the improvement the modified patch was applied to SOUND module at r +This is not reproduced with FOP Reader but SVG standalone gives badresults with SVG transition to closed remaining resolved bugs +we do not add such information in Apache Commons either +We could perhaps commit this sometime tomorrow +When that fails we now end up throwing an with a message indicating it cannot find for Solaris we should have JNA support in the next week or two +If it doesnt then its unlikely the script is going to work anyway +Extracted +Moving all issues to the CR +No discernible difference +Hi unfortunately I dont have enough time for now to split the patch into several parts +Thanks for reporting the issue and supplying a path +From your description it is a bug in the waythat are called +Did you tried to use datatype for processing the incoming attachments +bq +OK +thanks Ivan for rewording. +Thats for unit tests for the UDF +Theres nothing obvious in the above client log +now returns a list of topicpartition tuples that represents those messages that need to be resent due to an error +Can you mark this as a in plugins? +Isnt this a serializer issue rather than a Tika one +Right that would be the best +Im going to mark this bug as resolved since all the requirements have been addressed. +Warnings fixed in +input and output are now gone in favor of the standard and +rough code not real the query classes for and added tests for both the findbusiness and thefindservice instance and added calls to both within. +What is the status of New Api? or What to pick up now? +Replacing old patches with new ones +mvn install is successfull on my project no difference between JDK and JDK what I have not tried yet is as this script is currently I got the running also +Its everywhere we readwrite models metadata checksums. +Its just that we have to do it the right wayMany thanks +You can see the changes hereThe patch works perfectly fine + +Id like to close this bug downRegardsBrian Minchau +Committed to trunk. +Conversion of apacheds to use for configuration +This is still an issue but it is not a big priority right now +Hi Philip the annotation which is selected in the outline view should be selected in the text editor +These Policy settings do also go into the and into the One example of a is also attached. +To collect all hits implement your own Collector subclass and the score would be unimportant so the whole overhead of score calculation is uneeded +Brian I can take this and knock it out quickly if you dont mind +Unmarking High Value fix +I agree that this definitely seems like a bug in WLS +This is an exception scenario indicating that the producing system had an errorBasically im trying to distinguish between three states Value present and is Value present but is nullempty Value not presentI guess this is analogous to the nillable attribute on xml element definitions that allows a element to have a null value against not being present at all. +Done with FIXME +Brought uptodate w trunck and code simplifiedThe attribute names are pre post +Allow null +Stephen I just looked at the saved output of my past ant runs +All lines will be converted to +The issues here are only related to Tomahawk +The AM should get a list of files to be deleted after the DAG has finished +The timeout value in the tests has been extended. +This is a current web siteIf you have some opinion please tell meThanks +Committed +Actually Im closing this because we already have for adding NRT support to Solr +None of the workarounds work for Final so this is very bad +Blows up which uses SSL to speak with Amazon REST +Please add new constructors of records to also log in racknames and deprecate the current ones so that MRV doesnt break +Can you confirm with your observations over multiple installationssetupsReduced severity to major from critical as this is not seen consistently across setups +we could provide an optional for even needed its much simpler will hack it quickly +Cleaning up old bugs +Thanks Kousuke +bq +And again restarting first node cuts number of missing row by more or less half restarting nd node cures all missing rows +Martin can you review this change pleaseI have modified the management console to only print a notice to stdout upon reciept of a notification instead of upon every registration unregistration and removal of an mbean from the server registry +See patch in move from M to access to svn causes this to to next version +Thanks +We should send Invalid content type or something like that +Its a request for a feature improvment +Fix the issue completely by generating a list of CSS files with a maximum of selectors rather than a single big one when the user agent is MS IEHowever due to the central aspect of skinning within Trinidad and the major impact this patch could have on some other sections of the code I wont commit it before getting a revue by the peers if possible by Jeanne who is as far as I know our most knowledgeable resource on that matter +Seems we need a platform specific wrapper for the cast in callback +No there are standard junit test and plugin unit tests +This has been tested and verified on both platforms to work as you guys have changes of enhancment suggestions please create new issues for this +Moved to future +With the final release we did fix some problems +Pull it back in if you think different +The upstream issue seem to be fixed +Maybe thats part of it +I also ran a full integration build cleanly +Ill make a new noggit version for trunk shortly +push out to +Try it please +do you mean in? Or +Disabled the automatic JNDI registration on both trunk and branch as of revision +From the description the client is expecting to be able to connect to a running AS that is the part where we do not support the rmi connections +Next run I also killed the master while it was splitting the logs and then restarted the whole thing and Nicolas patch did its magic +When the ping times out the is called instead of +I checked out the new stuff in the console and it looks quite nice. +We arent making progress here too quickly so maybe in the interim we should add more logging around this and log something at ERROR instead of dying. +This was added inThe proposal here is to also treat raw comparators this way so that when a container wishes to create a raw comparator it asks the Serialization passing in the metadata that describes the serialized data +Thanks a lot +I am using HOD to create a HDFS clusterI am not making use of the static external HDFS. +can you clean up the browser cache and try again? As it still works for me +Essentially there are the three ways client learn about final status RPC call SUCCESS markers and +Will post a patch soon as it needs to be polished +Instances of this class are recreated whenever the component is activated +Im closing this ticket if we dont get feedback by end of this week +I tried to use changelist +Verified in JBDS h Fedora +Quick follow upSandip could you please tell us what is the current status for this issueAlan suggestedms timeout scan? Yikes that means that slower computers will be doing nothing but scanning +For the rebasing youll need to make your new handlers work with the refactored base classes +Commented out all the code were not using +In this case what would be a fax or TIFF? An image +Yes this refactoring is just internal optimization no changes to API interface and response itself +So if you had another provider of it would be wired to that one +Shouldnt this be marked as resolved? +to have this in +I will resurrect the code +Should we force Java inI think its too late for since thats already been branched and we should no longer make incompatible changes +This difference could be a reason of following questions +Untested impl for jetty in rev +I dont see the value in special casing for this very specific use case +Attaching slightly modified version of my previous patch that ensure that also have non zero arrayOffset +rtc item defect mustve been fixed by the recent compiler robustness commits +sorry for late comment but i dont think this works if some one created a table with timestamp datatype using metastore thrift api +Looks good to me +Adds changingPattern for snapshot resolver and adds changingPatterncheckmodified for fs resolver +jQuery is my preferred library but I dont agree in forcing people to switch from Prototype +Update the patch to address Sureshs comments +The constructor of Meta translates all input Class arguments to their stringfied form +Resubmit the patch because it was not picked up by hudson yesterday +Oh well. +Yes please see the snippet in the stackoverflow thread linked above +Ilya review this bug please +I wasnt able to reproduce +Its going to see a bit of further polishing but functionally it should be completeLooks very good +Hive actually throttles new if the heap memory on the JT exceeds a certain threshold +Correct me if Im wrong but I think the conversion of the dep of this plugin to domj has revived this previous fix was a patch to plexus concerning +Beyond that you would need to refer it to the JCP Executive Committee +Yes +Shows there is no performance regression with the latest changes +I have a test case for this but I cant attach it because this instance of JIRA doesnt want to let me +Any interest for this +After that patch will be applied we will have to rollback to SAAJ Sun impl +Here is the line change needed to keep the plugin from assuming your deploying from a file local to the allows the url of the file to begin with http +Thanks for reminding me I agree Ill do it +DB assigns it to the current schema +Forgot to mention earlier but if we decide to fix this issue then it will have documentation impact and we would need to mention in docs that synonyms cant live in SESSION schema +In Maven projects we would changes SCM in POM +Thanks Martin i clarified those requirements for the project status files +Moreover everytime I click experimental site checkbox another one empty appears under Others category +less bootstrap fragility +For see +We could estimate from the FD except in cases where we have no data at all which is most likely to be the common one +I think thats even more surprising behavior +OK Ill change it to but still not see what is bad because when we parse the string we use the same separator +Besides it would be even more annoying if you need to click additionally on the first tab to load its content +These patches are not useable they were not generated by CVS and Eclipse cant use them +It tested my project well +JinchaoCan you attach a patch for TRUNK +It looks like the latest +Attaching the new patch incorporating comments to make it comply with hadoop coding conventions +I think I would prefer to see updated to handle having both methods called +Hi Alejandro I ran the tests on my box +Hi Ivan +Attached is a patch proposal with a simple test includedThe patch is the same idea noted as a comment to namely toset the tableDescriptor into the column descriptor when building the columndescriptor inPlease have a look and tell me what you think +then drag the message annotation figureThanks Bob +The plan is to maintain compatibility of and performance +Any unexpected exception in will kill the thread and the error will be logged in +Heres the patch forit should expect generic IIOBE according to the javadocs contractI opened for the difference +This says that its a bad request a problem at line column where it encountered mismatched input in the form of the word IN +Ive attached a patch to fix this +horizontal is obsolete. +Deleted through intervention of administration at our request. +Patch attachedbq +Anybody can comment why? I can extract the library info when reading class in jarfileparse but it is not efficient as current the issue of extracting runtime library info from asm module +Ahh you know what I neglected to notice that youre having a Windows encoding issue +As hewrote the absolute pathcontext path is missing +This is why I wrotes parser is all kinds of bad +Would you please provide a sample data +Thanks Alan I found the same problem while trying to generate an entity called User and your suggestion works fine. +I tend to agree with Vladimir that this should be an optional behavior +The first one corresponds to the User Id column heading the second one the Add person me know where each one is picked up from +this would be weird wouldnt it? Ill check +Use the web app classloaders copy of Servlet to test inheritance +What the main rationale orIn my apps albeit mainly based on the rest plugins paradigm Ive never encountered a need to chain +Will update shortly +It is the generic wrapper around the Java Persistence API the one extreme weak spot in the EJB improvements +I can access it fine from welcome screen I assume you mean that you cant writeupdate the mvc archetypeThen yes this is for Marius to updatefix +Map slot seconds reduced by about reduce slot millis reduced by about Of course we need to reproduce this a bunch of times to be sure Im not seeing things but I think were onto something good! Ill upload an updated patch soon +Jene do you really think this is a bug or more of an an optional thing +Closing resolved issues. +IIRC this was deliberately excluded because resetting the time period requires a restart of the route whereas setting the max requests is dynamic +Section names was unificated by Olga Chikvina +Since that is the only place I know we require perhaps we could set +I believe the fix for also fixes this +I believe this was or a trunk build around that time +Changed to extend +It is lacking comments at this point +Consequently the editor doesnt load the workspace anymore NPE but you can see where Im going +If yes can you please add an patch +In the end you can always write your custom constraint eg a or current implementation is intended to cover simple cases like file http ftp which are supported by all +However a previous run with almost same patch worked fine +So we have to work around the JDT limitation and has hardcoded list of allowed archive extensions +Committed to trunk with revision will backport to if anyone feels it is necessary. +Else I will do when I will get a chance +reopening to add testsfixes for other DISI scorers i found no bugs at least in core tests +If were going to make a low default we should make a note in default scanner caching config and scanner cache javadoc method about the interplay with this parameter +Also need to work on find occurrences and search for methods with default values +Btw I can even change this class if I have Java sources +Thanks Dhruba! +My only nit is that two spaces should be used instead of tabs +I believe that Polexis has a directory for utilities +I can see the elegance of using Python scripts because its easy toedit and the pickling is nice +We could fix that if neededI dont know if this works is there a way to test reading on real input +think you were right earlier there must be some kind of bug in the active by default maven stuff +This probably needs major rebasing after recent changes to the state store apis that retain completed applications instead of deleting them +I am on is quite old but it looks like the class has not changed since thenI think there is a problem with always delegating to the super implementation forIts class loader may include classes that are not in the specified class loader so may resolve classes that should failIt looks like the name for primitive types is the same as the keyword +Closed due to the release of Tiles +Ahh will correct the spelling +This is the initialisation of the jBPM database from the scripts provided by the schema tool so is unlikely to be anything to do with ESB +There is no need for a we need a of the native launcher +Willem no worries +You can assign this to me if youd like +Thanks BrockRev dcbdbbcfcefed +When I apply the same change here Im getting a different diff +I have a provider holder that the provider is injected to +This is not going into +if you use the same name for the foreign key it will override it not add it +Here is the updated patch where I removed the orphaned imports +Ill keep you up to is the current release schedule for? or is there a URL that I can view the release schedule +HADOOPNAMENODEOPTS will eliminate the current problem but it also enforce HADOOPNAMENODEOPTS to be by the template system +As you point out there is some loss of history in trunk so not ready to mass do this for the sake of doing it +Current properties are mapped with the underlying Castor properties. +Thanks so much for this patch Jason. +Example of streaming job that succeeds with sudo u templeton hadoop fs rmr tmp sudo u templeton usrbinhadoop jar usrlibhadoopcontribstreaming input tmp output tmp mapper binls reducer usrbinwc +What was actually the reason for that choice +For values I used the method to compute precisionGetting the scale and precision from the data is the only way to make sure the server gets the actual data value set by the user +asc files are required for deployment to centralSomething that the default is nullfalse in which case no +Now assume a region that was split and still not daughters are not added to METANow the RS goes down +Hope to see test cases soon for these classes +Hi PrasannaI tested with latest marvin account not getting appended with test cases problem is that many tests create the account in the setupclass so the account is created only once per testsuite and not per test class +So at the end though we created the correct socket the Host http header remain in localhost +A fix for this was just committed by Dan K +DarynOnce again you are right +But how would you programmatically know that the is not operationalWouldnt it help to get that information directly via an API? Do you know if one exist? Let me open jira to add one if it does not +This appears to have been fixed added a test to to demonstrateThanks as to fix version. +So it is improvement to add this information to all packages +Please stick to stream until we officially release it its API is not yet stablized and may be changed even withn in minutes +Can you please attach your project so we can replicate your steps? Make sure your entities which are supposed to be persisted into contain the annotation +Let me know if you need helpdoing the svn move +This issue has been solved almost one year ago with the dynamic symbolizers proposal +Eclipse can run fine with only a JRE because it has its own compiler +Made the change +If the clients set log value as negative number will use default value instead +However the big difference between this and my patch is that the latter only displays stateful components and their parent +HelloIm new here +cocoon from svn jetty sun on Win XP Proall blocks disabledsome calls to cocoon status page and samplesresults after h handles event handles file handles semaphore you need to clarifyprove that this is actually harmful and drains systemresources +Attached is the with the assertion +I forgot the new class for the encoding +Patch eliminates the exception for me too +Thanks for the suggestion! The response to this on the developer list has been pretty positive so we will aim to provide simplified XML syntax in that style for Spring has this already been tackled in the source code or could I give a try to provide a patch for this? +ny luck +It looks like this was actually just fixed today as a part of +What do you think Jukka +Does this issue still occur after Groovy moved from Ant to Gradle +It is quite difficult to review it here +Ported security permissions for bindrebindunbindcreate +Bulk close for release +This sets all false but no update happens +is no longer applying on trunk because of minor merge conflicts +Triaged July assigned low urgency +While the patch might make the CCE go away I fear that it would be masking something else going on in either or application code +Then we can take a look and maybe spot something we can fix +Rejecting this. +Thanks +Backported to branch as svn Bryan thanks for noticing Ill fix it +For pipes we still need some work I propose it to be done in another JIRA if nobody else have any objection to thisOffline Ravi commented that it can be done in this patch itself +Looking again I see nio is available in +Is this what you meantThey probably shouldnt be local variables? Why do reflection every time this method is called? Why not save off if method is available or not and if youve looked for it or not +Please edit your test cases so that generated sources from current build use some location under target directory not within SVN sources +Never mind the last comment I manually applied this patch but the next one if you could do a svn diff from the top directory it will create the patch +fixed by view factories checking view reusability taking random access into are in release + looks good +Verified on JBDS and BRMS +Nothing in the system actually uses any of these changes +Thanks for reviewing +Assign to Thejas for further investigation +The only things which are safe to use from are some constants and it shouldnt be a big deal if you have those constants in N places where N is the number of threads +yep +You could try to run the JVM in verbose mode than you should see what classes are loaded form what jar +I applied the patch +Please reopen if you provide info. +ran all conflicts with the current trunk +So I think this should be closed as +Do you really think that an overlay in concepts on a specific term makes things easier to understand? Its actually the other way around +If you have the information reopen this issue and attach patches should be context diffs against the latest version +Unless someone wish to work on this in the near future Im moving this out to target +Can I just say? Awesome. +Small demo for bug see failing maven test it works if is was able to reproduce the problem +Hi DhrubaDid you get a chance to look at the new diffZheng +Great! Lets wait up what Jonathan has to say about code +Thx +I wonder why as the spec specifies only a single value +Another JIRA will be submitted to fix this. +I think we should close this and chalk it up to the learning curve +Bulk close of old resolved issues. +true false as in you Jean do you know which release versions this switch will be available in? I mean +looks good +none of this seem to be committed and Dima is on vacation thus assigning to me to look at in release +I would apply this patch +Jini did not have a specific problem it was trying to solve and so was too complex and open ended +Of course you get no caching but youll never get wrong dataSo this worst case is pretty much like not caching it at all but with some luck you get the value cached at least once for each threadLooking at other classes this seems to be a common technique in +A new CF isnt considered to be added until the CLS is ready to go +David Jencks changed the schema conversion util to move the taglib tags into a new tag. +The idea of this patch instead of backporting the other patches is that it fixes the specific problem without doing any refactoring minimizing the risk of introducing new behavior or bugs +In neither cases it is unrelated to binlib changes broke the libhdfs test +Committed this and pushed to branch. +Patch adding equals and hashCode and clone to +Hi AaronThat was just a sample of measurement for a day +Patch applies to the current tag incidentally +At any rate running the test in Eclipse there is no jar in the +Its much simpler to have the preemption be an RM only action +the first thing i noticed was the in +yes pleasethanks is the entire java package that I use when I am generating the WSDL +We want to stay backward compatible of course so this might be a moot point +In most cases this would only be a single entry +What is the status of this bug? I am interested in getting Mesos merged into the Gentoo Portage tree but this is impossible with the current state of the tarball source distribution +Note that these are intentionally not targeted at because they dont belong in the release notes +Thanks! +Im fairly against calling mutation handlers on replication as it seems like it could very easily lead to an inability to reach steady state +Fix me if Im wrongJochen Emacs will support and +Actually the current solution only works in target specifications +Moving to M +Correction to startup and stop scripts +In the did not use the process class loader hence objects of classes deployed in the process archive could not be deserialized +From the phabricator review I think this patch has Gunthers +Need to get this fixed being able to have the projects not require setting up the runtimeShould we simply just have an empty drools runtime +Im opposed to doing this +BTW we already have srccontrib +I get the correct parameter value from the query string you specified +Only works with java and doesnt work if webapp uses logj. +assign to Daisy review +Bulk closing stale resolved issues +If you can provide me with a simple project set up that shows the error then I can also take a look at it may be thisll be faster and less effort for you +Like the last patch however separate READWRITE CONSISTENCY is replaced by a single consistency +Not sure why does not detect that it is already closed +What is left for this task? Is it just a review of current functionality with myfaces +Similar issue occurred in our case as well +Bryan Im not sure if you have followed recent developments +So this patch builds for me +Steps to reproduce Download Apache Gora +Closed after release. +Fair enough +as discussed offline this should be considered a back port of the changes made in the oak projecttherefor it should work the same way such that we dont change the behavior twice creating users with null password should work changing a password back to null should not succeedthe corresponding test cases in OAK are test set removePasswordcan you confirm that this is the case +Im currently working on this one but I dont yet have a complete patch ready +Verified in H build +One of the issues I ran into is that while its seems technically correct to purge a tombstone after if it is not written into an its lost +Jason thanks for the thorough review +Maybe we dont care about wire compatibility but if so we should explicitly state that we dont careWe dont care in this case because this code has never been released +Obtaining Spring and removed the User Guide +This issue will be close when consumer tests will be implemented by someone. +I was missing a space in the sql query in the previous patch +However as you say we have more flexibility on the project name field +A fix for this has been checked into cvs and is on QA +Thanks Buddhima for Contribution +I presumed these base classes would be included inlib +Reposting the patch +will target Spring but there are only milestone releases so far +Just fixed some misspelled words and bad sentences in the comments +Just submit an update to the pluginThanks +In the UI go to Networkspick up NSselect non source nat ipConfigureOpen LB rule page +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Lets focus on the index part in this issue +Thanks Andrew +The annotation name shall be . +Please note that this list could be be a Groovy script with a scenario will be better +In we chose not to show an error dialog when the user was cancelling the job +and rpcencoded changes are still being applied +Reproduced in under is not reproduced if the timeout has default value +Im neither assigned to the issue nor particularly competent in this matter +Havent checked back on this for quite some time because we didnt need an archetype anymoreHowever the last time I created an archetype I couldnt reproduce this anymore. +Resolving as fixed. +This is not enough +In a dev environment this value is set by ant in production this value is set via the jnlp fileThis allows us to compile the application without knowing what the server hostname will be when deployed +This looks good +Hi QipanThis patch only bumps up the hbase version and adds a few dependencies required by hbase version +Wed write this region znode every time we flush +Should this be fully automatic or should Greenhouse ask the user before setting the username +Difference is in the inspect instance page or the normal tasklist +Ill update when I find out more +This led me to think that if I was told that Im including that wont work because I dont have the dependencies needed I would just go and remove +Ready to goNow left to resolve. +Ill make sure to next timeI changed the tests so that a server was available and then disconnects these ones pass for me +Clearly and to avoid confusion doing mvn install will do exactly what it does today no extras will be removed +Built fromembjoprtrunk Revision jbossastrunk Revision +When files are deleted there are other contention points but not during searches as is opened RO +Attaching patch with above mentioned fix +Added beginnings of and +xsl stylesheets that generate the +I think it is +all of that javascript is completely internal once you do you are on your own +To add +Would it make more sense to only support using the standard mechanisms of an OS if that OS provides one? In which case the functionality is to make hadoop a good citizen of that OS and not to make guessesThere is no standard path across Linux distributions for Sun Java these globs match Sun Java on popular Linux distributions thats how the list was generated +The transient keyword having to learn the difference between how they should be applied. +Caused by fix for there something I can do in my code to patch for now without recompiling grails? I think I am running into this issue +Its not unreasonable to need a default formset before a localized one but it probably shouldnt be a requirement +Wellit is in the documentation but not in the attribute list of the task but in the text above it +Need to add that property file to the Install Guide section on issues related to Roller release. +Great to hear +I suspect that the lock isnt released quick enough by the running HSQL while the app was being shutdownNow that Ive tried other webapps I can confirm that all the other ones deployed and undeployed perfectly +I am unable to reproduce the problem from your description +Rob how was I supposed to know you added this new checkbox? I can see it now but I wasnt aware of it +Applied workaround in rev baeeaaaaacbccba +in branchthanks! Would it also be a big deal to svn mv all AxisImpl and classes and to package just to make the code a little more structured? I could submit a patch for that but that would cause the commit history on those files to be lostAm thinking of a separate maven module after release +the patch looks good +Please wait a day for the changes to propagate +Ideally a unit test should run with as little extra code as possible only in integration tests should we have all those extra components enabled +So it seems to be okWill try add unit tests soon but if someone has time to do it now please do not hesitateBTW the patch is failing in hudson because of a failing unit test which seems not to be related to this modification +thanks! +assigning fix version +Sorry for the mess +actually the next possible point of a recompilation is time of last compilation plus plus minimum recompilation interval +I just ran and it worked +Completed At revision on At revision on trunk +Basically synchronized log split tasks before SSH is enabled +Its easy enough for set to change the size +I simplified the original patch +Patch looks good +would though be interesting to hear if maven has challenges in using +Create case is using DWR +The issue was fixed in commit bddaccaecbfedeeacd +This can be overriden by passing HADOOPHOMEbin +Apparently this is known to work on the current Hudson version whereas the notifications appear to have broken in the recent Hudson ll leave these new settings for a few days to see how they work for our builds +Certainly not an EWS issue +Seems not fixed yet replicated with in branch and trunk +Could this jira be merged with +you need to replace and for these issues +For example colors and fonts to be used as background +Is there any plan of integrating this patch in the official lucene libraries in the short term +After removing the EAR from deploy it cannot be deleted from inventory +Changed the sequence of methods for the generated translet +I want to download big file with slow connection and process abort by timeout +Thanks Erick and Robert for committing this +Resolving to test if newer versions of Java have this fixed if not reopen and well have to find alternative or detect the problem and report an errorwarning +Hiya have checked our source code and the file is not present apologies for this +Something to report to the spec group and mark as wontfix +If you have a Red Hat subscription for JBDS or EAP you must go through Red Hat support and not here for community project +I would think though in most cases the is the way to go +Ill gladly try it out and provide feedback or any other help youd like +bulk defer to defer to +please open a new issue for that because this one is now closed and the fix is in the upcoming release +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +so removing the regression check. +I was caught out by this as I usually install the postgresql drivers with the group and was surprised when no jars were presented +Can you confirm that version will be available in WFK +I can help to test and review it if neededThanks a lot I will send you the patch when its ready. +Verified in could you plz summarize what exactly was customized? As from comments its quite difficult to understand what was finely decided to change +Whilst we dont need to support this in WB its maybe something we could support as an option +Can anyone elaborate on this +Reopening to merge to branch +This issue proposes to add many lines and to add a substantial new requirement +Hi Kevin Not a problem +ThanksPlease close this defect +can you change Throwable to Exception +Submitted another patch with similar fix for reconfigure +I dont really know much about Guice though and even less about how Struts does dependency injection which appears to be based off of Guice +I tried with nightly build and the problem seems to be fixed +code that demonstrates what you are trying to accomplish +Ill try to do something along these lines under +Alternatively maybe to better indicate its purposeThe rest of the patch and the unit tests look good to me +This includes proper modularization of Spring Social as well as ensuring that Spring Social avoids using any libraries or classes that do not fall within the Android whitelist +bq +Can you drag and drop a file to the mounted folder +Why dont we just look for another XML encoded decoder and use that instead of the java one +Giri thanks for the reminder +Thank you very much for pointing it out Ill apply your fixSergio B +Thanks Jon for review +My goal here was to implement the minimum needed to work on and +Does not work in the following does not work +Tony patch applied at r thanks a lot please verify +Your additions look good but please help me understand why the old was off +Committed to the trunk and branch +Yes please both are bad bugs +We already know that the bean can only throw valid checked exceptions as this will have been validated on deployment +With the implementation of keeping and checking revision numbers of assigned configurations it can now be prevented to call the same service multiple times with the same configuration version +I have added each of the tasks that can be performed withfor GEP as a FAQ +This looks like a duplicate of to me +If not found the following error will be displayed to the userCould not locate Maven executable please ensure mvn command is in your pathTesting with Mac and Windows machines is required +Remaining problems when changing anything in the new tab the project is not marked as dirty When typing a factory name that is not a standard factory password encoder is not disabled. +Did you forget to add it to the patch +We do already validate the size of the TTL value since so either Vladimir is using something older than and I think we can keep that closed or itd be nice to have more details on how to reproduce +No tracing used so far +We will do a new RC for so this component is now a part of issues. +Hi SagaraYes I think these are really good numbers compared to what we got previously +This does not seem to be an issue any longer in SWF +As far as I understand I need to add documentation in srcdocbkx rightAny special guides format guides etc? Or I will just do that by analogyAlso found filters mentioned in too +The commit comments which describes the fix are as followsThis commit modifies to ensure that expected Connection events are fired to the +Ok source has been committed and updated with package namespace +Yes the framework and its dependencies must be replaced +I found the problem with the latest patch and have hopefully solved that with a new patch +Would you be willing to rebase your patch to trunk +Infinite wait where? Presumably either the collector will eventually reboot or else the agent itself will give up and exit +more tests and docs are wrong stop The example I have mentioned in previous comment reuse files from another solution +I think your description and title differ in meaning a bit but now theres an argumentArray method that you can pass your varargd reference to apply properly +See the attachment and try to obtain the wsdl of the published endpoint +Hi Werner what is the status on this issue +I thought the nested class remover would get rid of all evidence of nested classes but apparently the types remain in their original form and have to be treated specially later on +Sorry I dont see currently a demand for this ticketWe love contributions so if anyone in the community want to step up and create such a component then that would be great. +Can we modify the references to sandbox and hosrtonworks to something generic +I removed the new save method and implemented a saveView method which calls into the private save method +Im curious why you used svn copy instead of svn moveI used svn cp since the patch removes the files and otherwise some systems have problems if the files dont exist anymore +I can provide patch +Regressions ran ok +That was bad. +Im now inclined to this from a defect to an improvement +updated in line with recent did not make it into the release notes for so changing the fix target to +Can you please post a sample that throws this error +yep tried with rev and it looks good. +install it +May there be a randomization issue +Tomazworks as expected and definitely code is better placed. + might indeed be a better underlying data structure for this class +closing since I got word from ben on irc that mirroring will likely come sometime and I can mirror myself in the interim +trunk all tickets +Just checked the SOA issue it is now against and not the CP asked for the SOA issue to be moved into FP +New patch moves the call to get so theres only one +Closing all issues that were resolved for. +But I doubt it will be much of a documentation +IIRC the implementation of is optimized and reuses the char array of the original string while conversion from CAB to String requires copying of the substring characters +this patch includes update to Patch looks good and all tests passI plan to commit this later tomorrow if nobody objects +apologies the editor ate my whitespace note that int he above example the two dated files are meant to be under the archive application context should read enclosing execution context +I committed the change to as well +This is caused by the cadvise bug described in +The other colors have been taken from the Apache logo other than the blueI wasnt sure how I liked the angled T on the icon I think we should keep it horizontal +Nevertheless I think that this is a good start +My guess is the needs to be told about property assigns +How does C Ruby handle this +Feature disabled in. +The parent was the one that seemed to have locked up +its definitely a separate issue +omg facepalm thanks +API breaks for +This patch optimizes getDistance contains and is +patch +Patch attached +sample just run via jetty have attached a patch with the correction +Looks like the patch carries the corrections specified by Rick +If not this should really not be critical for beta but will work to address for beta immediately switches to error log +Fintan +Committed change in revision +I left the pass in +So my statement was by no means intended to be marketing related or spin as you put it but a more elegant solution or showcase where Appcelerator is used at the end of the dayIf you took the liberty of editing my previous comment you can safely remove now both completelyBest of luckAlfredo +This gets rid of the nested try finally blocks as wellAlso if there is ever a need to override handleRequest Context is already bound to the thread +Dave please keep track of its progress upstream and if its not fixed in the release that were shipping for EWS please apply the patch to our build +New version of results with failure issue +Committed to branch and trunk +I wonder whether we should fix this in too +Perhaps if we add it theyll follow +I just asked you if its right to commit toSecondly as your last commentscd jboss buildparent mvn installSuccessfulcd product mvn installfailedcd product mvn installfailedI do only as your comments not do anything elseYou means we need do steps as what we do in jboss tools trunk +htmltries to reference a document like thisIt gets transformed into thisdevelopingWhich is incorrect +I have a patch mostly finished that Ill upload later today but Im also open to any feedback +Its not a big deal as Karaf provides and wc dom is provided by the JVMAnyway I will include this in the next bundles release +I agree with Hans on this +Sorry for the alarm +Commits in passes tests in rev +This command is intended to allow karaf console users to logout of a running shell without using DI have not tackled the issue of untieing the lifecycle of the osgi framework from the lifecycle of the main shell with this patch Ive only implemented a basic logout or exit functionality here +Assigning to QE to verify that the change is in the CP docs +I think this could be caused by nightly snapshots being a bit borked +Teds guess at the change is probably good and it passes tests +Not on all +In trunk get calls analyze before it tries to commitcomplete the block and that will throw if the penultimate block has not completed +Documentation has been added along with an additional page to provide architectural guidance and coverage for the topic of Camel Security +However interceptor support for is explicitly mentioned in Chapter +I see these things when I run perf tests all the time it is unreasonable to expect for all data to be in cache in scenarios imho +Verified in JBDS B +Edward Ill set you up with svnpubsub today keep an eye on this issue for more instructions +good. +Esteve Ive review your patch +So it looks good to me +Replace responder thread while writing +Suggestion from Hairong If user doesnt specify a topology script for rack determination we can ignore the check for enough racks +This is especially critical for the stability of applications durring rolling upgrades to YARN +The should be prevented regardless of whether we support both +Patch added based on change idea from Brian Relph +If a cookie has the discard attribute set it cannot be added to a cookie store correct? Effectively that means I cannot retransmit cookies of this kind to the server unless I change them to not have the discard attribute +This affects things like the source for SCM information in the release plugin as is what do you do when you have multiple roots as in a m r scenario +Lets put this on hold until after the release +At least we should also provide an format forBut this would be a new feature then after the movement we can close this epic issue here +xsb files are only generated during compilation it becomes impossible to do thisThe workaround appears to be to generate the class files into your projects class file directory and then jar from there but it should be much easier to do the source code generation and +Will discuss with Philippe soon +Ive been annoyed by this tooHere is a patch which adds info when there are no results from the shards +I think the first case we should implement is not export +Perhaps it shouldnt have default in its name being internal it kind of doesnt matter though +We need to apply the patch in thrift and then cd libjava ant publish to publish the artifacts +Nope nothing done about this so far +This has my proposed changes to is my proposed readme file. +Guess this means I have to start arguing with our admin for that partifcular machine that still runs ancient +bq +I just committed this. +I will do that from now on. +Well do this soon after but not in is a patch to resolve this JIRA can someone please review and apply this fix +A patch that transform the breadcrumb into links to allow easier refreshOverview dbnameOverview dbname docnameOverview dbname viewnameNo more h strong only h and +generation stamps dont increase on every operation +I appreciate the patch but I wont apply anything that touches every line in afile including the license +why should everyone who wants the capability have to write their own just because we dont want it to make assumptions about a limited syntax +bq +Time to close this JIRA the problem never had any real functionaloperational with CPCR its not reproducible see attached server log file +Correct +Ive tried reproducing but lost patience when it didnt reproduce after K iterations so cant say Ive tested the fix but the JMX documentation makes it pretty clear we should handle those notifications and that its very likely the source of the problem here +Failed in the unrelated currently erratic this time in? Unrelated +This is definitely one of those places +The names are all regex expressions and matched for both of those +In summary this problem is difficult because hooking into the existing session factory can have the effect of forcing a roll back on some databases and some database configurations +Fixed. +I added another patch which also sets to use DELETECOLUMNS by default and added some tests +Looks fine other than it seems like a bit of overkill to have and +rar etc we ship so the inside it are signed +Ouch that definitely wasnt deliberate +I even read that the JVM guys think that pooling can be slower thanAlso Ive mostly seen gc performance problems so far if there were a big number of objects it makes the mark time of the garbage collection very long +Havent tried this yet though +Will backport in couple of days if there are no objections +Search service depends on indexing service can be categorized into two different parts +I have verified that the does indeed fix our issue +Tested on the latest Hadoop trunk and inconsistent output when using hadoop fs dus during globbing seems to have been solved +I have even built a brand new VM and deployed a brand new clean install +Aha! Quite right this is now fixed +the options here are to either show for all or show only when the nature is enabled +Thanks Jing! Ive committed this to trunk +but perhaps we should get feedback from more people? Id be fine with any of these options but lean towards AS and shy away from +The attached patch adds the missing call +This patch removes all of the corrupt replica detection from add and moves the disk space consumption code to commitBlock +If you are interested in owning an extensions project me +MapReduce would also be useful in the +Thanks +Bug fixed at this project +Kurt said no I cannot get the testing working +Ill check how it works for other products and then decide what is better for us +Build failureThats my mistakeSorry +Thanks Amareshwari Ravi and Vinod for checking this +Attached is a patch that resolves the issue along with a test case +In the absence of this feature changes have to be made to the domain model to workaround the problemI think it would be very easy to implement it at least for the table per class hierarchy strategy +Wondering how good that would look printed on a shirt for instance +Estimate does not include making the jars much done but the needs to be moved to the root of the project. +Revisit issues that have bugs older than a year as Out of Date +OK button MUST be on the left +so I answered my own question +Right now there are some tests that fail +actually why shouldnt it be possible to use only Id deprecating +That looks good thanks Paulex +This forces the web application to use wssj instead of weblogic can be closed +The problem is missing validation on feed name +Just a question have you extended the schema to contain the number attributeType? If not then the server behavior is correct as the does not exist it reject the modify operationwdyt + there a binary distribution for this and if not when can we expect oneIs the latest version on source control system stable so we can at least build on our own and patch since this is a major blocking issue for CxfThanks in advance + +Satheesh might now the specifics +Thanks for reviewing the patch Dhruba! I changed numDatanodes to be numIterations +Test failure and findbugs are unrelated to this change +so in the end its a jetty bugConfirmed +Amila I think it is better to make this to a service parameter because that is how we handle the use as wellWill do that change and apply the patch +Does this sound plausible? The only thing I occasionally notice is that a client starts up with a view containing only themselves then a merge quickly occurs to fix +I have tested it locally and it appears to work for me +The patch looks good please commit +The ParsingResolver got adjusted according to tobis feedback. +I second BJs you BJ and Jacques for your comments Now a days I am little bit busy in some other work but I am sure that Ill implement the suggestion given by you As I get time for it +updated dependencies to point to the latest google sdk plugin +I think either would be better than the protocol exception since the protocol exception doesnt give any indication as to what went wrong +Having scanned RubyinitBuiltins this appears to be the only case quite like havent gotten my local environment running and debugging yet but once I do I plan to do some more research on how the loading routine works to figure out the best way to safely trigger their initialization along with the parent +fixed in cbadaeefbfaddcafd +Yes thats what i thought was happening +I also moved stuff into as it simply included some utility class used by the jdbc and file persistence modulesI think we need some more consolidation so I leave this open to find some time to have a better look expecially to the api modules +Patch applied at +This jira is not rearchitecting what is already available but embedded REST api in region server makes a lot of sense for long term maintenance of +requires as a patch looks good +Thanks Luke that has fixed the problem for me +This really look transient to me +Id hate to have only a partially complete fix in and and the implementation we really want only in the latest branch +Committed revision . +is relatively much to inherit just to use these classes +Closing stale resolved issues +Is there something else I should do before committing this change +Note that executing multiple queries in a session is possible with async execution +I would expect to see used in an inner loop not in exception String construction +Trunk Committed revision +Its really a subtask of the major +Verified with and! +Thank you! +the the new SNAPSHOT version works like a charm! +Making the administrator log into every single node to delete a file is not ok +I removed the HDFS dependency +The following test passesant test Dtestcase Dqfile +The fix has been validated + Change aggregation to calculate rate for datanode first then sum up the datanode rate to get the cluster rate +Fred see +testcase for your prompt attention to this. +I feel it is not required +The only type of project you can create a web service on is a dynamic web project which is faceted +I am marking this as a blocker so that it gets into +Closing issue fixed in released version. +Thanks! +As foretold here is a trivial preliminary patch to move computation of inside the cluster something weve craved for a very long time as evinced by the interest in this jira and the number of times it comes up on user listsThis is huge because its a significant step towards various improvements such as job submission etcShameless plug for it took me mins on a Sunday night to get this done +It would be in the classpath local to the invocation but added to the classpath for any of the server processessigh + +As you said there is an inner transaction which is trying to acquire the lock on same entity +Resolved with +patch looks good +What happens if that build agent is busy building some other project when you click ReleaseIt will still continue with the release +Fixed when we moved to the new zone server +Seems to work using the current trunk +I think thats the new firefox update seems this is a problem of Firefox as I can reproduce the same problem with revision before this issue commitAnd it works well in some other browsers like IE Google Chrome Safari +I believe the original patch failed due to +Not working in by code change in +I had verified it in revision . +About the new file added at the end of file this will be done in a second time +People are free to work with them subclass the classes implement the interfaces but it is not the hadoop teams concerns if a new release stops their thing from recompiling +Sales Opportunities just need to be linked to accounts +Ill look into adding the specification code and not worry about the licenseThere is a subtle problem with fixed delay timers right now +The update you proposed looks good from a purely dependency and testing perspective +Say you change the installer dirshould bebq +Please set the requested fields so that the issue can be properly triaged or it will be rejected due to lack of information +Thanks Srikanth! And I agree a container would be interesting +For accounting it would be an internal organization +Patch for this issueSorry about the code formatting +Patch applied to trunk with revision +Any new progress with this issue? +Merged to the branch in revision . +This appears to be normal behavior for rubyruby which checks modes on your home directory +The cleaned for the patch! +This broke bootstrapping the node thinks it has the schema when it does not then of course streams nothing and joins the ring +If we could know the version of Solaris being affected and also if it is possible to know what modifications were done to get make the compilation successful I think this issue can be solvedRegards AllI also am going to port to Solaris starting very soon +Interesting +committed this +Patch on top of that addresses Bikas earlier comments about ACL checking as well +Why does this need to be deprecated? Generalizing it as a way of setting up other limits or ENV config seems more correct +This causes the client to see a DRDA protocol exception and terminate without ever ending an in progress XA transaction +Started a new security module with revision its right that only tokens prevent full CSRF however a referrer check can help in certain situations +Sample application that demonstrates the issue +I dont see how that screenshot is different from what we have today? about quickstarts not the job of project example engine to scan it should just be told JBOSSASHOMEquickstarts + add a style class for table put it on the. +for the and y version of the patch +At least theres no unit test failures +Instead of byteCount we should use fileCount which counts how many files needed to be copiedEdited summary and description to reflect this +Bulk update to fix version +Emil patch has been applied +Getting an object back from when it fails strikes me as odd +Which editors do support this option +is the current level of logging at FINE suffcient? Or are there specific places where sql is not being logged +Attaching the source code and the +I have not tried the patch will do soon +Created for the MR parts of the patch +Committed +Reverted the default back to +The proposed patch fixes the Locale ctor issue with JavaPlease note that the needs to be updated to latest for reporting the issueYes I agree that adding the sun +Isnt the exception telling you that you almost launched a job with the same parameters twice? On a retry wont it just throw or +backend directory is for multiple backends that we support +Fixi is in place in l GET Quickstartwebserviceconsumerwsdl HTTP Jakarta But the text is missing from Release Notes have been submitted for QE see the release note page section +Thanks +easily? Or is this a silly questionRoland thank you for the patch +Thus setting to +fixed the problem with hdfsproxy +default is not something we would want +Passing patch through Hudson +Can we use this cofiguration for publishing our ivy artifacts? We use ivy for dependency resolution and we have files and not pom file + +Are you able to reproduce this behavior with applets in FF? If you can then it would certainly suggest an OSBrowser issue rather than a Pivot one +The the indexer could remove the dupes from the view indexes the next time it runsOf course catching all the specific dupes will be tricky +This issue as well as the solution was also quite educational experience for me +It may be that some of these typos that had already been fixed in other patches applied since I made the patch I suppose +The replicas update their HW only on the next replica fetch response +If you really wanted to go crazy you could make the property follow standard syntax and do something like telnetportmaxconnectconsole +Packaging the Avro C code should depend on instead of I dont follow this +im going to postpone this one since the webapp is mostly there to show the capabilities of the engine +Please use your custom patch until we release +we need this behavior back in order to upgrade to rc you set singleWinow to trueI works for me with from the snapshot version to the released version works ticket can now be +I believe thats why we had iterate and iterate on Vector to be a bit more particular about the behavior +yup +add test address todds comments +Note that I took the liberty of extracting an interface out of the abstract daemon class +looks remove the map entry in the ve just committed this +AwesomeAfter applying the patch mvn clean install is successfully finished +change in enhancements release +There might be a problem with the Windows script +Uploading a repro as a patch to which shows both that the embedded driver behaves incorrectly and that the client driver behaves correctly + fixes the warnings +The interfaces need to be a separate package to avoid sealing issues The proxy class loads the implementation class using and then casts it to the appropriate interface and then invokes methods via that interfaceLet me know if you have any questions +Fixed patch. +Therefore the is best be placed inside the common moduleThe goal is to have a place holder for all configuration properties +We do need to make sure this happens for also noticed that is still using as its default path +Without which AM cant update the askrelease data structures correctly leading toresource table in AM and RM going out of sync +The installer shouldnt contain the pack jars so it shouldnt require unpack +Do you think there could be other ordering bugs in this test? I presume it hasnt been run much with Java yet +Can you also take care of as part of this? We need to make sure no test depends on because of the version +Rename srcjavatest to srctestjavaI seem to have missed the file +thanks for the clarification +If the scm command returns with a resultfailure it dumps the output from the command in the logs +Tentative classes have been coded +Just committed this +Ill go ahead and open the Jira against that myself since I have a simple program that demonstates the problem +At the moment you should be able to achieve this through the help of static methods +Also tried with adding sleep call in it passed +Reopened to review if the fix should be committed in Beta +Forgot the on the splitPoint for second +What is wrong to have possibility to name unique constraint in similar way like it is possible witn Index +I would like to contribute IO Connector Service implementationPlease see atached sources +Related to the previous comment it might be a better idea to state somewhere what properties we require from the backend store + +Is atomic construct an abstraction of transactional memory +Excellent! Thanks a lot! erm Whats with the blanktransparent area in each image +One of the bugs I had to fix was in the leastmost items methods so that I added new items to the before removing old ones because the one it just added might have the same hitcount as entries already present +Because webhdfs uses the HTTP header a URL is not sufficient to specify the desired range +I just checked in a fix for this problem +Thanks +Sounds good +This will be picked up by the configId change +Hithe same piece of code works fine under Windows OS with or IBM jdkIf yes i will correct it and make regression me in and released the July th as +The Solr update XML format is a specific Solr XML format not XML not notanything elseTo index XHTML parse it and convert it to Solr XML update format +Can you give an example file +ok it works now but I have little sugestionthis combobox with mimetypes is not userfriendlythere is always variable that contains mimetypeid and can be rewritten to mime variableso combobox with this shouldnt be required and used only when you want force to another mimeTypelist with mimetypes is very long and it is hard to upload lot of files beacause of choosing mime for all of themBTW +No longer reproducible +if you send to topic then subscribing to topicblah +I would like to keep this open for that perfect solution +Added to the architectural goals for +Improved error handling for produce fetch get offsets request to handle unknown topic error code +The patch has been appliedPlease verify +attached the second version of the bnd scrplugin +committed to trunk and +Patch seems fine except that it needs to check is JMX is enabled in the if statement +Dirk can you commentThe dropdown was removed from the editor UI because while changing from one editor to another all edit actions are lost +in didnt instance after create a new portal for edit inline +Actually many of the functionality has been overwritten +In the patch Im removing the old bookies registration path if exists and continue bookie startup +I have fixed all the culprits and a new release coming now. +Cool +Eclipse itself shows these compiler warnings +Should Woden use a wsdlj dependency compiled from the attached source drop or continue using the more recent versions from? In the latter case the license information still needs to be updatedAnyway thanks for the source drop Ill definitely find use for it elsewhere +IMO not setting agent name should only block fetching in protocol http +Ill commit soon! +I now see that the javadocs can be improved here especially in the reader +Im uploading a new patch taking your comments +Fixed by r +verified by Tony. +Since the spec says they should raise an error when the result set is true these methods cannot be used to check if a row in result set has been changed with a positioned update +Ah second thoughts maybe you meant something different +Hi can you try an attributes property? It is meant for suppressing Sorry for the JIRA it seems it will work i didnt notice that attributes was mapped to the set of config +Now at to move to in order to pick up fix. +Here are the attachments for this bug +This project has been waiting for someone with an itch for making the build easier for a long long timeThanks! +Scheduling for AS as I want us to move from having a realm name to encouraging the admin to choose a realm name on entering the first user after that all subsequent users will use the same realm +I tried to remove the dependency and found need to to look up the which holdsSo I will just commit change of part. +No sweat +Ideally that could be build into the shell +Reopen if there is still a problem. +However as I changed it to some other folder like tmp it worked +Unfortunately that doesnt work an empty profileName translates to the default value +But I assume that fixes that underlying problem so its probably not worth addressing it at this pointShould we commit the attached patch today independent of so that trunks not broken +can you provide a minimal sample? It shouldnt bother about Java classes +Maybe we could even add a shutdown method directly to theAny thoughts +Im assuming that the portion will be handled by +TimThe fix is good +There is a doc request raised for that so I will add this as a comment +Upgraded to the cometd as this one uses Jetty which is now also used for the other projects as well. +Greg? +Thanks Dhruba! +Perhaps Im mistaken but according to a test by another member of my team the conversation id is currently globally incremented meaning that although it is only valid within a single session the number is incremented at the application level so two users in two sessions starting four conversations would have conversation Ids and +Resolving as a duplicate of the mechanisms we added for detectors and parsers and the multiple type database support added in +Nick Ive just tried running locally on trunk +This is not an EAP issue so please do not touch the poms +ok you know better +this patch compares the existing remote user principal name with the new one and does a remove update of the user credentials if there is a mismatch +I even removed the prompt for controller I think it should be configurable or an extra parameter to select the desired style +This can be integrated into the eventing support inI cannot access this feature using latest GA version as it contains instead of neededThanks in advanceRegardss right as the support will was introduced in the timeline and will not be to the one +I did some more testing with the patch that I suggested +Will get back to you as soon as possible +The geolocation code will be a nice addition +Do you think you ever will get a chance to work on it or should I take over? +Ill try to fix the problem +I looked in a recent build from master and is in neither the main geoserver war nor the plugin +Nice find Carl i say we try that and we will also need to add the full Apache license header to the guysthe license header +As I have already used the previous patch and done some other minor changes can you ensure the new patch is off the master version +Sorry that no one looked into this problem more promptly of course performance problems are of great interestThe problem appears to be caused by the parameters to the object that is used to store the character data of the input document +Well address renaming to after beta +But Im not sure if that answers your question +Sorry about that +Well have to supply both eventually and my vote is also for a truly complete jar. +Andrea can do it for me please +So it should be careful to make them behave consistent +This is because the test method does not save the process instance +Is it possible that the issue Im seeing isnt necessarily caused by corrupted logs but some sort of failed leader election? It seems very suspect that if I reboot f the leader node that it cant joint he ensemble on reboot +Attaching a first rev of a functional spec for this issue +Im using a separate flag to track when a rebalance starts and when it ends +OK thanks Scott forYour patch works wellThanks for your Eric fixed in r +Can we please summarize the current state of the discussion hereThere seem to be two main issues as far as I see number parsing Im not quite sure what we can do differently here +Arpit VivekArpit your modified patch is in trunk rev +Previous tests on other dbs need to be executed again +But in EJB is not a must file +Thanks +Heya not too familiar with whirr so I did not exactly saw where the discovery part is configured? Does it use elasticsearch built in ec based discovery module or does it build a unicast list of hosts for it to useAlso the default gateway is local meaning that data is stored on the localebs drive +In the build this error doesnt exist +Client must have specified if you do not want to use the latest protocol +I tried removing the plugin from the plugins on the sona server and running mvn sonarsonar inside the project +Patch to raise when keys are not found overridable by command line option also allowing the ability to ignore insert errors if desired +patch issue branch for the new test to pass no longer need changes to instead add headers to response! thanks for the patch! +Toby have you tried reproducing this against the trunk? should have take care of this with the new java option +Fixed a couple of minor issues and added the ability to exclude the lower bound +i have checked before committing this patch but it seems did not intentionally delete links and contains some info that appears nowhere else +Please use Commons IO to work the problem aroundI am not saying the issue is not real +Do you mean? Its already connected as a referenced issue +We are only missing the UI validation +I think we should document that our current impls wont handle this +classpath files these files will be generate automatically when make a release build by maven if we commit these files it will affect the drools build +Jessica Can you please add listalltrue when listing by zonepodcluster from UI +is it possible to validate a snapshot +Resolving this issue was not appropriate as it was not fixed +I believe this factory will sometimes generateI havent run into any NPE in practice but Im willing add some null guards ifyou can identify where they are necessary +I assume the API allows this and thats what Id expect rewind to do +It seems to work although deployment of exploded apps is extremely slow +OK with me we wont break RCS support with that +To avoid problems be sure to not use a PARAM without arg Remove default hardcoded settings to allow users to remove them using an empty still wonder why I couldnt reproduce it +Suggested fix +The examples project is not included in the distribution +I tried to implement the fix you suggested Ivan +A bit cleaner patch +could store that its an int and expose more stronlytyped from as well +Just track for now +I want the clustered results should come with keyword or phrases not by the title +Why do you run it every second? I think its a little short and will use lot of cpu and thread +Assigned back to Dave to determine approach +i opened a new issue for the oracle problem +Thanks +Its going to be more complicated than a but I still think quite doable +Marking this as a blocker for wastefully sets maxMemory to needs to be fixed along with +It wasnt until I looked a second time that I realized it was is or some such +bulk close of all resolved issues. +Patch to add a copy method into the Bytes to add a copy method into the Bytes version based on Stacks comments +Could not reproduce as well. +It has to be done for everyones sake user and dev +Applied with Does not work on trunk due to missing configurationbranchSending assembliesSending etcSending pluginsSending pluginsTransmitting file dataCommitted revision +If an existing class is to be used the error string has to be made a constant with the test referencing it directly rather than having a copy of the string +I would like all hosts in the includes and excludes file to be displayed in the +Thats actually a design decision +Applied a for now but this issue still needs to be resolved properly at the WS level. +It turns out that the other functions do not have a problem +builds were doing too much which matches exactly with the problem that changes are too significant +Patch to trunk +Merged patch with recent changes to +Closing this issue as we have not hit the exception since fix and no commit related to the issue was found. +applied in commit cacaacedffcccdebThanks! +Now that work is underway with AS all previous community releases are +Here you are the attached files packeds in a zip archive +I think we can fix this by including the jars in the binary distro +Remember the feature freeze date for is and the last day for merge request with hour feedback peiod is . +After some time the Component state changes to active +Yes +I added a legacy test and fixed a small NPE bug that happens via rthanks because the mergeScheduler and mergePolicy specified in the indexDefaults section is not honored +bq +For now we stuck with the approach of providing what Spring upstream provides this being a technical preview +So I dont think that doing the step described in the comment will affect anythingThere is nothing particularly strange about this Eclipse headless build so I dont know where the leak might be coming from. +Ive thought of using which can handle arbitrary precision numbers +Patch +Hi HiranyaThere is a minor typo in the sample documentationPlease apply the attached patch to correct itThanks +POI beta has been released +Opening a pom the first time the pomEditor works can open multiple pom files without issue occurs when I close a pom and then I reopen closing a pom I can open other pom without problems +Notice that these two classes need further processing to put into Droids +sounds fairly specific and its hard to compare across heterogenous nodes furthermore its hard for to understand and specify it +Good catch Daniel +With or wo Internet connection +Thanks Jukka I dont see the attachment attached to the issue though +maybe add a read only keyword or table property +I think this will allow the region server to shutdown if the native session expires +Important note if you guys consider that the interface is used BOTH for a whole document and a single element which will be incorporated in a document the patch should not be I think we should remove the thing which has been applied in because it does not make sense +i cant build on trunk Settings Repository is null +Now Im trying to utilize and insteadThis method seems to work well +I was able to reproduce it on OSX +Its still require and needs to be optional but at least controls work in web services nowAnd theres an automated test that verifies that this works. +Is there any maintainer who can review and commit this patch? Or the project is already dead +Thanks Bharat Rishi for the contribution Initial pass is been committed in trunk at rCould you please provide patch for other email templatesAshish Ashish for committing the patch will soon provide the patch for the other Email templates which follows the same +Let me know if I have it wrong Lohit. +I can always do it after the initial commit +Maybe a plugin to generate the endpoints based on a xsd schema could be useful +Thanks for letting me know +Thanks! Ill try to have a look soon +Verified in EAP the core issue here that the authenticator although enabled by default is not installed by default +Also some like Column etc +Thank you Jakob. +Please let me know if this information is correct +fixed note average fields for stateless will display inccorect value because the numer of opened session is nearly zero all the time. +Im not sure I understand your second question +Can this be tagged? So i can build and upload to the has to be updated to build the zip and the src tag can be updated later when releasing on sourceforge +Fixed with. +I am so sorry! The truth is that Im extremely impressed with and grateful for Sonar and my use of should wasnt meant to demand someone to fix it but rather because Im convinced this fix is the right thing to do +The new website is checked in under mahoutsitenewwebsite Isabel had taken the pains of porting all these information to the wiki +I can think of two questions to answer Whether to wait until the DIH response is more consistent and or to forge ahead with this issue with the existing DIH response +Please proceed +I tried to checkout from c into ca but the filename is still to long. +please review patch for patch definitely looks safe +Like this I cannot apply your file as it would break other things +However since the element has xsanyType we dont really know how to parse it into a real model object so the object ends up just being a hashmap representative of whatever content the server decided to place in the element +Sorry for being unclearIf that is the case the suggestion of cp p to enable the semantics you are looking for makes sense. + +Moving this to since its a new feature and not a blocking bug. +rar file insteadRemember the also +I run it on Win XP +Hi John I can not reproduce the issue on missing uncovered conditions +If we are adding external dependencies on cryptography libraries I would prefer them to have seen more mainstream usage tend to focus too much on issues like which hashingencryption algorithm is best when this is unlikely to be the main vulnerability of a system +Dobbs Journal. +Not able to reproduce on the Galaxy Neuxs Samsung Galaxy S or the HTC One X +v fixes missing note about userpass for connect command when just help is used without connect attribute +Is it not just a matter of appending a list representation of the cycle array to the message +This issue needs to be better organized +see description for +There are some issues along the have preliminary version of the dll built need to clean things up test a bit and then submit +verified by Julie Zhuo closing on her behalf as she does not have Jira rights. +Yeah this can be closed. +It printed a lot of output to stdoutlines likeTester Run Select Mon Jun PDT Tester Run Roll Mon Jun PDT Tester Run Insert Mon Jun PDT Tester Run Roll Mon Jun PDT Tester Run Mon Jun PDT The command I used to run the test wasjava +for commit +I have reviewed the patch and approve it +Please check that it was applied as expected + +Also menu items created on Linux has version +The SA will create a directory called inputdir in whatever working directory you started servicemix in +aggregate hitmiss counts will definitely help +Ive added this +Reverted in revision . +If that output is acceptable I will commit this patch +after release +forgot to mentionadd apache license to the top of all filesadd a to the contrib directory giving some basic information on whatwhyhow type information +This patch cannot cause these test failures +Caused by the way Configuration handles key deoprecation all deprecated keys for a specific key need to be added at the same timePatch removes the tez and keys which should anyway be accessed via +Fixed on trunk +But in case it happens to go into there there must be a serious inconsistent state i suppose two region servers are having this region and very likely masters region info is different from META +I have about projects +I have a debug +Maybe in the next cycle well be able to get to it +Im trying this on JBT +Nothing otherwise magical there +files are updated in the continuum working directory +Reopening +Did you ever find a fix for this? We are currently experiencing this error on and off and havent been able to fix it +Patch that adds DLQ detection to the +By extension is there a way to modify streaming to take advantage of this? We have a problem with streaming littering orphaned processes that we need fixed +Bulk close of issues +Changes Make splitshard retryable by unloading cores part of a and starting from scratch again +With the attached patch I dont get this stacktrace anymore but I dont see why this should be a specific issue +Im a bit confused as to what page is rendering when the redirect happens at stage +I hate when I cant reproduce bugs +The reaping model can take away those problems +As per discussions that have happened on this issue there are two separate issues identified Tasks get initialized without a limit on the count initialized at a given point +Will try to knock it out this week +Hopefully it will move on +Thanks Travis +The problem is a limitation in features +to branch and trunk. +This bug report is indeed a duplicate ofI added a unit test in r to verify the now correct behaviorThanks for the report anyway! +Is there any progress to report on getting this pulled in? Do I need to further revise my commit in any wayThanks I do not need you to do it but thought it would make working on multiple issues at the same time easier for you. +Hi Its me again + correct Fix Version to one without x like depending on where it was fixed +EAP are fully opensourcethe bundled maven repository have the source zips +Aleksander did you try out the current svn version of geotools trunk? As I told you in +added tests +Can you provide an example where no facet results are returned +Should i add one job for solaris sparc to ci runsyes +It may have been though the fix you suggested would take care of it +fixed addingservice phase not to break out in the middle of the process. +libext +Feel free to review changes in SVN and make adjustments you deem necessaryOleg +unmarking for +Hamcrest Mockito matcher for Message Payload Header +Still make public in +However both of them run nicely with my patch enabled +The alternate tactic is to have a mock implementation of the code to determine disk space use and simulate failures when a node comes up +I guess the issue is in SMX http since it returns different objects depending on logging levelIt should return a payload that Camel at least once can read +In our production where QPS is very high during the server start the proxy is ready before cache ahead about we dont want the proxy work until the cache ready +First rough patch to follow shortlyMarkIm very interested in this +I suppose this would have to be some xml configuration. +Itd be good to see this fix in EAP will mark this as blocker and will resolve it as soon as we make a issue is listed as Major or below and as such is not targetted for the EAP release now that we are in Blocker or Critical issue only mode +Carlo is back so. +Attaching a patch for this + is working for me seems to not works in linux adobe reader and not on windows for +Then it should work do you use the latest from the SVN trunk +Add a getter for the errorCode for a unit test which would obviate the need of findbug suppression +Ah thanks didnt catch that +Since plugins are released separately from the actual platform there is absolutely no reason to block this release +Patch of the points bug fixes for rendering TT fonts +Move the fix to +Verified in ER. +So here is a final patch without Spring +I did this a couple of releases ago +Regards in the new sample with some minor changes +Simple patch which adds for total launched mapsreduces and maps these help for +We would like to have the patch +I did what Chris suggested and its been working fairly well +Its in my INBOX on Sep as well +lgtm +Note that no is included in the jar ever +I would vote that all paths are uris and thus must use as the separator on all operating systems and file systemsThat would certainly be nice and we try to do that as much as possible + +It is in that doesnt affect delete action and project finally deleted +I think for simple fixes like this the is implied +The previous patch did not apply cleanly to trunk anymore because has been split into multiple files +Although Markers are part of Im leaving this open in case someone wants to implement one of the other techniques mentioned +Patch with the changes +Committed the patch + +With the patch the string representation are valid json strings +Samples also added for each artifact with sample test cases +was pulled into prod +Committed to trunk +It looks great +fixed in cvs neither of your solutions were enough as they did not cover embedded subclasses +I will update and test this eveningRegards is fine imhoThe duration stuff looks useful to me as wellRegardsLars +See related issue +Ram currently I was thinking about three changes +Completed At revision Unit test activated in Committed revision +Hm +extension of Stephan Bublavas patch +attributes which I am not sure of +educate me here What is xmdesc +Maybe UIDVALIDITY changed and Thunderbird still has the old? I know UIDVALIDITY can be quite funky sometimes +Yeah I think binary does the same thing but we definitely need to add that to the developers doctwiki +It also patches the plugin itself to use the proper date format for those fields +Fixed on trunk +to commit this especially with the tests I like it +this patch warns users if the source code contains references to s extension to EJB annotations +Since it doesnt I suggest bumping it until someone can provide a patch +there are also s +configapplicationContext +Just FYI for whomever takes this this is fixed on trunk so take a look there +Thanks Stepan! The patch was applied as expected +RajithI tested the issue in Jdk Tomcat and it workedThanksDeepal +Thanks +They are part of both HDFS and Mapreduce and are very important +Im going to back out these changes +Ill change it so that if a username is provided for the distribution repository authentication is enabledIve resolved the two linked issues as duplicates but I now see they concern a different albeit related problem +Go ahead and commit whenever youre ready. +The committed changes look good for me +I cannot reproduce the error +I unfortunately wasnt diligent enough to track exactly which step fixed it but the problem is no longer presenting +Im therefore not talking about anyuser but specifically users with WEBTOOLS permissionsGavin +As its blocked our testing so raise up the priority to Critical +An iOS has the same behavior +With that and fact that we only need to update ZK on a flush should be fine to keep this data in ZK though this would be first time we retain info in ZK so may need to start thinking about clusters with s of regions +Especially nice to catch regressions in the future +fixed +But setting up a focus manager or something like that on the pagerequest cycle would be nice +bq +It seems more intuitive to use lib as default +notNulladd field reference jpa class +In this case the file that has stale markers is the very file that changed +Thanks that helped a lot +Test is incorrect +maybe we could copy the continuation url to the new pagemap but i dont think we can ever know that we can remove it from the old +Could not repeat this issue +In the near future this requirement will most likely be lifted but I cannot promise on a time frame for this right now +I agree with your solution my patch is too complex waiting for your new patch +Committed +The main problem is that any of the join queries does not highlight its required field. +Thanks Guido AmarillaYou patch is manually applied in trunk revision +Patch from Amit Sharma +I havent tested this so be kind Im pretty confident it will work though modulo any automatic writing that seems to periodically interrupt my coding +Id have much preferred folks to speak up then instead of now +This guide should satisfy you Ate +The patch worked fine +Applied patch and generated the docs +Thanks Andrew well explore this in after is out +Denny when do you got this planned for? Ive set M as my best guess target +Sorry to be late to make this comment but what is the rationale for having value in the subelement? If someone is going to use a literal value they would include it in the url +I dont think that is a good expression for a number +It is Cordova its Apples fault and it has been resolved in a future version of Cordova +mark when a trasaction ends allow a reset without a mark while reading to accommodate the first reset before a mark is set call mark on the writer when finishing +We need this in trunk and probably too not sure how it got missed somewhere along the chatted with Devaraj and he is OK with me taking over this JIRA to have a patch for trunk +If it doesnt can still be fixed in time +Max shoud we commit it to svn? At least to trunk to test it more carefully +Should I create a patch for too + patch looks good +Try that out and see if its works better for you +Fixed in for your work Kristian! +IMO database should act in same manner +Are you using the embedded Maven in Hudson by any chance +I guess its pretty clear this wont be addressed +switched to embedded fail under Windows looks like the same problem as with other tests cannot clean repository between tests +Attached the patch +It wont affect release but it needs doing soon after +A source does not delete a block by itself +Agree completely with Nick here! Nick I have added a comment to your pull request could you check if that make sense +meant +But really with adding filtering that requires LDAP filter this one can be removedIll make the filter field stay and not being cleared when someone clicks inside itCommit? Which patch? The first or the second with categories +Is there a reason why the default lint settings are different than those used by Hudson +Patch applied in revision +I check what could be done on that Today implementation of spatial search in Hibernate Search is filter based The scoring processus is done at query time not at filtering time We would have to compute distance to center for all documents in the index to compute the score before they are spatially filteredThis is not what we want as it will kill perfSo to do variable score we should move form a filter implementation to a whole new Lucene Query level implementation of the spatial functionnalityI think this should be better achieved with a native Quey that re implemeting on Hibernate Search levelWDYT + investigate this asap so that the seam team can complete their build +I had checked your comments fixed it +Check the ML archives where he posted the link several timesPersonally I use the default for Emacs and I correct warnings manually +Great job +I have forgotten to check the option in my new eclipse filter I will take into account the in the implementation of temporal predicate. +rev actually makes static support work +Maybe we should fix the unit test instead +Not sure what this means +I think the reasoning for return instead of break in the FD block was that if we havent finished sending hints then we probably dont want to force a major compaction that will rewrite a bunch of undelivered hints +Jira is not a help all issues that have been in resolved state for more than one month without further comments to closed status +I guess there is JNDI +Closed. +Sorry about that +But making it optional isnt probably going to be completely straightforward +Oh no the is not for cargo +I committed a fix for it too. +I agree that consistency in this regard would be goodI dont want to start an argument over something so unimportant but I was just wondering is there a consensus in this developer community that two spaces are better than four +Done means it has been completed +Filed and for those remaining to addressPatch updated with fixed indentation and references to the above +Ill run that +Sorry I forgot to add that I tested this against version as well as the latest + +I just committed this patch +fixed in svn HEAD +New file in addition to patch +The mitosis replication subsystem needs to be completely rewritten since we have discovered criticial flaws +I can give it a shot if its ok with everyone +OK no problem +May have to disable the inactivity timeout in code for fixed and temp tcp queues +Just out of curiosity is this a Java convention or a AS convention? I had a look on the Internet and couldnt find any mention of it +Hey DavidThanks a lot for the patchCan you attach it to a new issue though? It helps with tracking in and svn. +New patch that fixes also the two failing tests by setting only to false for themI will now create a new subclass that tries tests with old and new api automatically +The recommended deployment is to not put hbase on the TT classpath and instead ship the dependencies as part of the MR job using the provided utility function in +From this change I modified that contains most of all property keys +Im not entirely sure this is possible using existing Compass but its worth a shot +Thank you Mayank. +Screen shot of the odd tagging performed with the flat project reason this was reopened turned out to be a variation on this issue +When you fix the build and the testsuites dont show problems yes +I didnt find a way to change it to warning and that it why error is still shown but the Reun button is not blocked +I didnt test yet +I am also trying to automate and parallelize the release process hopefully in time for +As Alexander pointed out this seems to work therefore closing this bug +in +QPIDCPPROOTexamples QPIDCPPROOTsrc QPIDCPPROOT QPIDCPPROOTrubygen QPIDCPPROOTsrcqpid QPIDCPPROOTsrc QPIDCPPROOTsrcqpidamqp QPIDCPPROOTsrcqpidclient QPIDCPPROOTsrcqpidframing QPIDCPPROOTsrcqpidmanagement QPIDCPPROOTsrcqpidsys QPIDCPPROOTsrcqpidlogPlease let me know if you find any issueThanks DanushkaThe patched refers to qpidcppsrcqpidbroker and qpidcppsrcqpidconsole but there are no patches with those files or patches to files in broker and console +How do we benchmark for good size? Can we push that to a separate issue? This issue is about removing checksum files from the HDFS namespace +I changed the class name to show the relation to a stackThe method names are still the sameThis patch already depends on +A simple way to get around it is to attach the iframes current path to the end of the main windows url +the findbugs +comments on phabricator +It works perfectly for me with Tomcat on +One good thing is that the DNA code change will work in either situation and the overhead is pretty minimal + +Thanks for catching this +Swedish people will most definitely in that situation write raksmorgas when searching for rksmrgs and most probably blabarsyltetoj when searching for blbrssyltetj while my guess is that an American would write raksmorgas and blabaersyltetoj +Perhaps it might be good to to point to the example of use in from The result offset and fetch first clauses sections examples or make another one there +On second look a couple of very minor things the guarded log statements are maybe If kept it can we ensure they are surrounded with braces? +Thanks for making these improvements +Closing issues for. +It was my mistake +Very ThanksI will try +Patch verified thank you +That seems to cause the problem I was seeing possibly only in Mac OS X? I couldnt reproduce it in Windows Vista out that line fixes the ll explore further to see if that is also the root cause of the issue in the Installations page +Bug has been marked as a duplicate of this bug +This issue has been last updated more than years ago has no assignee affects an old version of Tapestry that is not actively developed anymore and is therefore prone to be in the near futureIf the issue still persists with the most recent development preview of Tapestry please update it as soon as possible +I just selected a wrong version by mistake +And we should be careful to add more to it +means this inode represents a file +I tried the same wMS SQL Server and saw the same problem +Good point Doug +Same thing happens in +Ive seen this before too +Unknown tag goal +Code optimization in +You know very well the price of forking you cannot assume that well keep everything the same to please you +Using to merge in configuration would be the strategy for other tests which dont currently pass based on Should the call be placed in a common code path so that other tests can benefit +could you check it into sandbox so we can start playing with this plugin +If you pasted this directly from your file then the answeris clear Change fowward to forward +if youre curious. +Item are ordered based on extension point id which is formed as plugin id extension point id +Hi Julieni have updated the patch +Correction the JPA configuration provider does use globals +The problem has been fixed. +Thank you for your quick reply! Ill modify the code as you suggest and Ill get back to youIt does make more sense that Fragments evolved more quickly than my understanding of them rather than Wicket being horribly broken +v wasnt fully baked yet +mrepositoryorgjbossaopm continuing investigation +People shouldnt have to install openssl to use libraries that doesnt actually need it +Me thinks that we have to add a bit of documentation for this new feature +Let me know if you need help +Thanks Prasad for the help. +Sorry there has been a problem in one testcase of the last uploaded archive so here is a correct I included an adjusted implementation for based on the proposed +but it will be a couple days work and we dont have time to do that for +Eric I think you only had to ad lib the init script part right? That would be because of +Scream if you think this is wrong +I tried to find it myself by pretending to be an anonymous user no luck +They could probably use the default JDBC profile and the default processor right now +bq +Thanks Prasanth +Will this prevent us from putting the in a different portNo more so than the current implementation does +Valid formatting of dates is important +committed to is an occasional situation +public evolving Cluster Job publicstable Mapper Reducer Counter Counters ID Mapper Partitioner Reducer private public stable +Of course the results dont really speak well of SKM but at least there are some results +After clicking on check box it converts into Viewed Yes and update entity +So some magic is happening in the middle Fixed I will be on vacation till Tuesday +I have the same issue jdk tomcat roller mysql +I believe Eli committed this +You can get this by building and deploying the plugin on your server and then querying the service for the wsdl using either or curl +This patch is for the branch not trunk +Im not sure why the thought persists that we are denying all bots at this point +testcase for +Changing to blocker based on conversation with Sameer +I think the I saw is because my caught an OOM and then when it tried to log it another OOM was thrownI like the idea +bq +There have been bugfixes since the release but not necessarily committed IIRCThe robots parsing is working OK in Nutch and we have loads of other things to work on which are probably more important +All the code have been tested with LuceneThanks Art for the infoURL +I suggest though this patch goes into CVS HEAD only +After poking around a lot looking for a way to bump the reservoir size I finally came across the paper on reservoir sampling by Vitter +Looks great Ill commit in a day or two +need to fix for also +Sure will do +So I think the thinking on contrib is to instead either take it as a part of the project that we maintain indefinitely or let it thrive elsewhere +In the case I had it was a uses constraint that failed +Duh of course the contribs are also ivy modules that depends on core +Ill rebase the latest patch soon and wait for review +Each are distinct exchanges at the transport layer that require no shared state +Thanks but you have not granted the ASF licence to use the file which means we cannot include it in our test suitePlease could you delete and reattach itAlso we will need the equivalent zip and Win archives for comparison +Well proceed with general Maven upgrade discussions here +The buildfile generator should be able to work with the local repository specified there since Windows doesnt have good support for symlinksIn the meanwhile you can set the environment variable MREPO to point to your local repository +I assume this patch doesnt fix issues with and is targeted to and later +Pull request merged +and so we dont need to change the visibility of andRemoved +reverse the list and join successive sublists with +bulk defer to +I believe we start new caches for each app Ale knows the details regarding actually start caches for each need to be eager started so any MR tasks jobs work as expected see +Checkin r reverts the change in trunk. +This patch contains a fix for the issue mentioned in this JIRA +Duplicate of +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Bulk close for release +but not for the +Prepare phase actual writing is launched when all the writing threads created +I assume you have multiple application deployed at the same time +Fixed the findbugs warning in +So i reproduce the scenario +Because Accumulo makes Hadoop hairy and more powerful +A ZIP file containing two schema files and the generated java source files +Doesnt seem to matter what order I put them in one dependency does not get listed in the project dependencies +Looks good do not understand Step ? That is not normal I can run eclipse without the vm attribute on Windows Mac no problemAlso Step and are mutuallyexclusive +This simple comment can keep auser from accidental results. +I changed my benchmark to better show the seek caching effect +Yes reasonable comment to commit +what protocol etc etc +Reverted attrunk r r finally close because I have no time to look at it further +Please keep in mind that the main deevlopment infrastructure should be resided in federated build repository in enhancedtrunk in case of jdktools it should be in workingjdktools +Could you provide more infomation +Thanks everyone +Please lemme know if this covers this one +Or are there any volunteers for thisReducing the priority +The suggested some reason the patch fails on +Yes but the original regexp youre talking about has ALREADY been changed inthe code as part of the fix of defect +Thanks Giri! +wsr suffix +I working on waffle integration and looking at Ryans attachments +David apparently ran into this problem on +Thank you +Thank you that does serve as a workaround for us +It works well +PaulexFix applied to TEXT component at repo revision Please check that the patch was applied as you expected +So perhaps we should simply drop the entire classAnyway I changed this to an improvement issue as the class being in srcmainjava was not a mistake. +With the use of Nexus it does not make sense +Has anyone tried the twoservers quickstart and deployed it to +Verified by Ivanov. +its really a lot of copying around to fix a shell issue +hi erickson +The POM structure in CXF was lovingly crafted by Dan with occasional from othersIt would be better not to reformat in this sort of patch since it makes it hard to read the diffs and I do not think that empty elements are desirable at allPersonally I use xmllint on Linux or when a pom needs a format cleanup +I am not bothered by the warnings really just helpfully suggesting I have a changelist ready to clear this out if there are no objectionsOn Taste erm what should be deprecated? I already took the opportunity to remove anything that was deprecated and remove more stuff that is kind of old and unused +What are use case where the version range is a problem +ping? any progress to get this merged +Sorry yes I believe this has been resolved +Fixed in branch and trunk +theyre all the same +Robfind the small icon attached +I have committed a modified version of the patch just add a protected method to get the right column count that represents each component to be able to modify the code correctly in Tomahawk +have committed this. +only occurs if the server does not exist and I use Run on server to try to bring up the HTML file +So for example rounding gives and rounding give which IMHO are both correct since the hour does not exist +This patch is again off branch +I dont understand what the issue is here or how you are trying to compile itThe java source file is generated during the build process and works fine hereHave you followed the build instructions on the wiki +Maybe the notion of the lock icon in the other ticket is something to explore for locking fields +I just committed this +fix definitely does something startup time of cluster using preloaded caches went from minutes to minutes +Same result +It would be great if you took a look to see if that is acceptable +Please remove the fixVersion and close once you verify this on Alpha release. +Code is still the same +Then we can decide in the using code what level of interface is neededNot sure which one is best +Verified in automated testing +Actually it looks like this can only happen in exotic cases +Note that this does not occur with the pageflow editor +Will take a look at it +Message variables are maps by definition +In trunk with derby it works OKI am now updating postgresql to to see if this happens postgresql works fine also +I will do so tomorrow when I am back at workI have the current dev sources so maybe I can provide a patch then +Couldnt reproduce with subclassing disabled reopen if you see different results +So we will need another idirect mapping to the cache store from our individual implementation +Can you please elaborate? MRUNIT itself is not part of Bigtop so theres nothing to test +I guess I would not be in favor of a special term I would rather see it integrated into the file format somehow +Also I dont like the solution +I just committed this. +but the patch does not applyben can you provide a patch for? pinging again can you provide a patch for +We might also consider to move some modules from the to normal features +Could you regenerate the patch with u option It would be easier to apply +Why is this marked incomplete? Not sure what that status means +Heres a proposed patch to workaround this issue +Otherwise we are saying something likeYou must use an odd number and by the way zero is invalid +Closing. +I see how being able to set the adapter directly could make testing easier but do you have other use cases in mindThanksJeremy +revision . +Looks this is the bug in branch only +Works for me +Then when loading and computing each partition we load the relevant messages for that partition and populate every vertexs message list +Was there any reason it was downgraded from to? We have been on for over two years so it doesnt seem like a merge error +with ip on geronimo but still failedCaused by Could not connect to broker URL tcp +Please post on the forum. +does get pulled in via transitive dependencies when we depend on progeny +The implementation class should then be moved to some contrib module +Done for do we have tests for transitive interceptor bindings now? +The other plugins are not part of the Maven project +i sayed that it works with separate properties thats not true +We are not providing Zones in the sense of a Solaris OS ZoneWe can provide you a Jail but we are awaiting a new array for the extra spaceWill keep you posted +close it and dont tell me youve closed it +Resolving will automatically resolve this issue. +it will be some time until I can take care of s a take on the +The stripped test I used to narrow down the problem initially +When is this going to be released and available in maven +fixed a filename typo +Attached herewith is the patch that implements basic traversal for XML schema type alternatives +But it also exists for us and we dont get the tack trace +it in addition sends the hashcode of protocol methods b +Thats right +From reviewing the code and looking at what some of the sync points are I dont think even with the you can remove all the syncronization as it servers to control orderly shutdown as well +Rajesh are we supposed to enable ipV from EAP onwards + ant clean compile fix and test casethanksdims +Ryan fixed it +What is the better other opinions +Will it be sufficient to demonstrate that the alias is ignored when passed in createCriteriaAshish +Im getting this error relocation against a local symbol can not be used when making a shared object recompile with fPICThis seems to have been a problem for others on amd too but I dont know what the fix is other than adding fPIC to the library file build but that has bad implications on non amd platforms +Hi KarlLooks like this patch is not working on trunk +I hope its acceptableIll submit a new version of the patch soon +Heres a patch that adds a Previous reference to and code to to build the servlet in sequence +I never did get around to the converter script but everyone has survived for over a year without it so im just gonna close out this issue. +closed +Thx +Attaching source code for this use caseSome notes is the class with main method which demonstrates the problem points to my SQL Server Database so to actually run it you may need to replace connection information where I have pointed piece of trace which indicates the problemPlease let me know if you need more for Hibernate trunk to change query cache timestamping so that cache hits occur within a hibernate only the part of the patch is necessary +With speculative execution false +Sounds good thanks Bikas +Talk to Tom Ross BTW about his changes on +Templates wiil be loaded from the specified location each time user starts my mind the default location should be the workspace root folder +And even if the target objects do not implement equals themselves arguably the shared AOP proxy should be returning true at any point since the target objects would be conceptually equal whichever one got picked for an actual longer I think about it the more it seems OK to me to keep comparing proxy identity there due to the inherent stateless nature of the proxy instance itself as well as due to a full equals evaluation leading to the same result for all typical scenarios +Fixed the code and the README +Trying patch submission again +Thanks for pointing that out Edward just fixed and committed it +The specific issue has been resolvedIf we can reproduce different OOM errors we should create new issues with specific information +Paul can you reproduce it and attach a jstack here +Thanks for taking this up +I faced the same problem in the past but prefer to use to add javadoc for my packages as well +I have opened for this issue lets please track the progress from there. +According to Georg it works with the change +In the yarn code we standardized to put the units in the name of the config property +The above was in reference to Dhrubas approach of using the abstract class verses Owens more inline approachWhile Dhrubas is a little more complex it is not complex enough to warrant doing the additional RPC +Since authentication happens once per connection wed want a separate CF? for information +Patch integratedTook a while to migrate gwt code to rails. +Thanks issue as it has been released as part of release +Is this useful? Shall I improve the patch in some way +We can add prerequisites in README file or download them on demand like Harmony did +Here are patches for geronimo trunk and openejb trunk to fix this problem along the lines suggested in the dev list email discussion +Changes have appeared in and alpha docs. +for as always looks great Stefan +Submiited a modified version of the patch thus removing my veto for this and the patch +I have arial on my system do you +Resolved using Approach as outlined above. +Resolved with resolution of +This looks good to me +Cheers +Hadoop is pushing jars to apache repository on successful build +But you only add a empty element user can not know what should he do +A similar problem was reported in will be one set of threads for each database that is replicatedHence the name of the thread should include the name of the database +Moving items not being worked on for M afaik out of not thess issues out as they relate to the old and +Id prefer to have the class in AS +Im not going to be able to get to this before release +Thanks for the patch Stevo +The hashmap implementation is closely tied to capacity being multiple of two and the code added that calculates capacity should be future proofI will make the other counters volatile +Should work with the project split +Thanks Pranay for reviewing and testing the patch +JBT site cleaned up +with thoughts that should handle scenarios when we try to change the state of Container queue the state of will always be stopped +As it is the client which drives the this strategy of locally changing the works +We are planning to support the as suggested by Ciemo to avoid name conflict +That is for sure a better idea +Now in trunk +As such I think we should deprecate these methods write a unit test to confirm that if the key is a type actually fails somewhere with a good error message +Hmm it compiled locally +And since its a bit unusual I think its reasonable to mark this as minor and punt to some unspecified future release +If folks are writing individual records to a database then a best practice is to maintain a registry of schemas used in the database as a separate table and have each instance refer to its schema in the registry via its MD hash +Thanks +Per the previous comments in the thread the location of the WAR file is set in the sh scripts and generally wont change for a release +Solved +Only a bad work arround to handle documentliteralwrapped SOAP upload the soap message that is being deserialized and the wsdlxsd as wellthanksdims +The problem is that it introduce some incompatibilities with previous releases +Is this still important for BRMS? Posted from Bugbox for demand +gemspecsThere may also be a commandline flag to Rubygems to force it to rebuild this cache but Im not sure what it is +What do you think about removing the id altogether +Please reopen if anyone sees a failure in. +except I dont think get can be relied upon to always return a valid reference as discovered in ticket cannot reproduce it in and +My take is that should have never attempted to guess the type of the fieldIt creates errors when you query on something that looks like a date but isnt indexed using orI guess we still need to make a choice on whether breaking the compatability is OK or not for think we should commit this except leave deprecated rather than remove it since that is the promise made +Please verify this patch. +Shall I remove these myself and commit this to trunk +Can you close this issue please +Empty by defaultexport HADOOPOPTStrue HADOOPCLIENTOPTSthis is assigning teh m or any other hadoop heap size to early to HADOOPOPTS and places the Xmx after the HADOOPHEAPSIZE var is set so overriding that +So for each of your cases in turn there are two different objects with two different class hierarchies intended to provide implementations of the same interface +Fixed in buildUse correct JDK +duplicates +Added privileged blocks for test util file functions that are called through SQL functionsprocedures +A particular issue described above is not in fact caused by massive class loading but by some problems in class itself +This is Resolved per the initial description +Oh forget what Ive saidHow do you do represent a null value in the DB for a if your greatest index has a null value +The approach of making them connection level severity seems to workbut I see many red herring from the replication regression test +Will use this for the pending testIs this jira done +and all child tests of +I kicked this off again +The internal Y! release does not matter to the larger Hadoop community except that he may have helped detect a problemMain questions are Did the semantics change and if so accidentally or on purpose + +i think more can be done here +Then sometimes it becomes empty but the system will keppe running fine until the next restart +username +I havent tried yet. +Thanks Alexander +The aim is an easy way for folks to add Axis to their existing WARs +Hm +If something doesnt look right we can reopen it. +Ive committed this. +Changed the code to not use deprecated methods anymoreI am wondering though whether we should deprecate and and introduce ANALYZED and NONANALYZED the same way Lucene did +Ugh that was definitely a break in my badI missed that this change would mean we silently stop calling the methods in subclassesIll send an email to. +Unless Im missing something it should show up in that chain shouldnt it +in the import wizard +Thanks +But what is the correct behavior? Should only resources be checked for existence? Should nothing be checked for existence? srcmainresource doesnt get added in the first place unless it exists but that may have been different when this eclipse exporter was first written +But it is a common fix not HDFS +Stack could you assign this issue to me? Thanks +Probably can get the master side right earlier next week +Seems reasonable +I made a fix at that now treats all KindReg calls as calls directly to the compiled code of a method +Toward that end I am disabling tests that no one is looking at or that folks are looking at but meantime the build is failing because of the test so I can move on to find the next obstacle to a clean build +The unsyncnotification problem was introduced by +Fix in r +all the these resources should be moved out +bq +Shouldnt this JIRA be in common since it isnt HDFS specific +I just checked with IE on Windows and I dont see any diff between what I see here and what I saw with FF on linux. +It will just encode it out as abstrct +same issue of +branding feature is not built during devstudio build and hence not included in problem resolved in H +Not a high prio but certainly good from a marketing what should we do with this issue? Performance results for are clearly superior to INRIA results +Im not convinced we need to add the suggested controllerattributes but Ill look into it if I can at least get a from anotherdeveloper +Marking this as a blocker for so that it will be fixed +Very large generated XSL so not attached but relevant section as aboveWhat source files in CVSD for the fix? Ive built from available source +This merely changes some Eclipse configuration +What do you expect instead +Martin please verify this fix on the CR docs bundle and close the issue +Emanuel is right +Thanks SianApplied OK. +Close after release. +I took the file from the portletbridge directory of EPP and did not see this error within the logsOne difference I noted is that EPP has a newer version of and Seam for that example so not sure if it is an error of those old versions running on EPP +Integrated into trunk and +Actually +I just gave it another go and googled the solution to my problem and confirmed that the spaces after commas are also present in +This would help us verify that nothing is amiss +AlexeyThere is no need to commit my patch nowI agree with solution you suggested in workaround +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +It is very to fix the client to use the same SQL state as the embedded +I have committed this +So I am attaching it now +in addition having that optional dependency to would finally allow us to provide utilities for which up to now was put into or into project specific utilities +Even if its a legacy database +From what I can see there are two tables which reference this table as a foreign key JBPMBYTEBLOCK and JBPMVARIABLEINSTANCE but I cannot find any indexes for their constraints +A relocation POM is wrong in any case +I checked the latest bits but not changes in file to specify the resource bundle name! Is this fix part of the package? Where can I find the +Nice work Thawan! I havent reviewed the patch yet but I have two quick points about it Jackys test case was written on and I produced one for ZabTest which I posted on +I will keep you updated +Does that mean we need to track down who thecopyright holders are and update our NOTICE fileI dont think the fake interfaces increase the potential forintroducing errors as I also suggested that the real interfaces are used if is set just liketoday +So I feel this was false alarm sorry about that +The patch attached to should fix these integration tests as well +Actually logging happens after the prepare and before the execute +The issue was the following when looking for window definitions in the select list the visitor looked to deep it should not move into subqueries in the select list +The idea is to include this resource with some different selector andor extensionWe could prevent endless loops if we would throw in this case where the target resource is this resource and no selectorextension overwrite is taking place +updated patch +bq +Revised binjruby which works with Solaris binsh +reassigning this issue to myself +Hi SeanI see what youre saying about coupling restore +whats the reason? cause by? +is fixed external links now open in new and are fixed refactoring text and is fixed now have a Start link to the left of the other header links on all added a Open a Support Case start page so Admin links arent shown unless your role has the JBN EM Administrator flag checked +This was done by purposeThose are the same executables butbin is dynamically linked to opensll +Previous submission committed at subversion revision +It was probably fixed with the newer versions of Tycho. +There seems to be no Analyzer in Lucene that uses this class +Yep were not looking to make that change right now +Hello GregoryIve seen but Im not sure that is duplication ofit +Added PROPPARENTID to documents fix is incorrect actually doesnt belong on documentsActual fix is a bit more complex +Patch updates and with datatype mappingsThanks to Thejas Nair for input and multiple reviews +Moved to Bugzilla. +Can you it as ZIP file +I just realized that adding the buttons statically means they are shown also when javascript is disabled and then they do nothingI would like to have both buttons that work also in IE and view that is consistent also with javascript disabledIf this is not possible I think it is more important for IE to work properly +Appears fixed now. +Thanks +Thanks Cheolsoo! Its to reply late on this +Is this the right methodWhen i executed this method it didnt clean the selection totally +The existing tests even call it they just never call equalsAdded a check to avoid the NPE and a test that equals on empties works +Post a test caseexample or a patch that makes it more clear if you still think it is problem. +Thanks +Query parser not registered by default and a commented out entry in example solrconfig was addedHopefully ready to thanks for the work on this +So you want to define it once for every definitionIt should be some sort of shortcut rightIn this case I like it +with JBT +Ive just committed this +Any C folks want to take a look at this +No matter how great this works I dont think we should pull it into proper +Ill keep you posted on how it works out +Reader +See. +Marking this as resolved as the complete code to handle property files has been rewritten. +It doesnt come for free with the factory change but I suspect that will make it more convenient +First patch for PAYMENT SET REF +Hi PeterThanks a lot for your contributions +Both code and body should be preserved and passed to caller as such +Included in +You are correct +Ive got this error with commented finalize method and the problem of SOE handling in gcallocfast is not related to this test +Changing assignment to Joanne T +The configuration of jUDDI itself in +It could be backported fairly trivially +It looked better for an artifact id +In the mean time let us know if there is any specific debug infotraces you think would help root cause this issue +Eugene could you take a look at resolving this +Inaddition Ive patched srctests to use one or the other whereverappropriate +Patch was committed before test was fixed +is held by to be used globally +I can fix this issue with the attached patchIll leave it up to you guys to decide if thats the right fix or if there are more exceptions to be considered etcBut from my narrow scope of understanding it just seems that the SHOULD swallow exceptions like instead of throwing them up +Perhaps all of the accessor methods can even share a mechanism +use context in instead of context passed from as r +Upgraded to version for and upgraded the following dependencies from to from to from to from to from to thanks +Im looking forward to seeing this in our first releaseI see you updated the unit tests already Ill see if I can find the right place in the Selenium tests if no one beats me to it. +Logged In YES useridMatt have a look at whichdefines a custom Resolver for loading the DTDs from withina jar itself depending upon the PUBLIC +Good to know +The challenge with the backport that Benoit identified will come if an Ubuntu maintainer wants to update the version of in from to something more modern +ups sorry I meant ShifAltQ L +c default b As I said before setting the LANG to C which I understand as a generic locale solves the problem in my environmentFor it also renders an empty page +I was chatting to Sijie earlier about how this is a more general problem +Verified in dist built from SVN After signing in to the petstore a banner image now shows at the bottom of the pages. +Making blocker on +Fixed in v and vdirk +is not included in the API coverage coverage as of July +Any other feature requests for Session Handling? +See comments at per Lukass comment need to escape the so this is not a bug. +It caches a instead of a +patch looks good +In this regard I have made an attempt to modify the escape handling in the Commons CSV parser +is a sketch for how we may add arbitrary metadata to returned documents +The existing attachment is not readable on my machine and there are too many missing pieces for me to see exactly what youre talking about +Only when we set the permission on this particular directory then only we will hit this issue +Fixed! Im off for this year +Comment moved to +A simple patch moving the interface to in the module +If you could help me out with an example code doing an SSL certificate authenticated persistent queue consumer with a selector I will try it happilyBTW you ask this because you want to know whether works the same or this is a STOMP related issue? +The swapping is something we can do +Please see looks the same as the issue that was reported by Chris on January th is it possible youre using the version of the patch that was reverted? The way to check would be to see whether extends Path. +The patch file containing the change +Frankly throwing a lower level exception will not make it any faster. +this patch makes necessary modifications to enable defining postion for +Thanks for pointing that fixed +In no way its supposed to be more than exampleHowever my first intention was to add the comment you suggested I just forget to mention it in my previous comment +Thanks! +Launch this testserver with mvn jetty the class to access the testserver. +Thanks Peter + +How the tile is implemented underneath the hood is in fact one of the benefits of doing it as a +If there is anything else i can do to debug I am happy to do it with some hand holding +Included all of the files in a new patch diff and also attached a new file with all of the updatesPlease review and comment +I posted a new patch that only makes private +Still fails tests etc +We need to know how where the pics differ to debug the serializer where it happen +This patch to add Support for iiop keep and idl rmic switches +trunk +I would temporary remove all String methods from Term and try to compile core +All ant tests except the following have passed All these failed on trunk too +I have included an additional test that would have failed with the previous patch +So Im not sure toolsstress will work? However if toolsstress has a way to generate keys that would be useful +Can you try +While adding the project files I reorganized a bit +Wiki updated +Ah sneaky +Looks goodTested on and have committed this to the trunk andThis chance requires Clover to be used for test coverage runs. +I created a follow up issue for the questions you raisedChanges for this issue are included in the release and we shouldnt it +Created and closing this one. +Mind parsing this for its goodness Mr Meil +For this test I used the old style database persistence that is still the default configuration +This can be resolved because it has been confirmed working +Updating the fixed version to patch will remove the memory leak created because of not freeing the subscription processor +It seems we still some consistency issue with scanning sometimes Or is this a bug weve introduced recently +Duplicated some data in the database now the queries are a few hundred ms faster. +As I commented is not affected. +It works only for AAR MAR files and does not work with service deployers +Btw you are aware there are edits to code right? Also needlessly turning objects into serializable etc is not making me happy here +were not developing Deltacloud any longer +thats just it there was no commit +Should probably be looked into when doing view template support tags +packages all +Looking forward to it +the current definition ofunique without any defining nodetype information will just prevent any kind of access control content to be stored to the repo +committed these changes as revision on branch work just matches the work done within the platform build process. +Compare your Tomcat config to that found in rollerpersonal it sounds like it may be configured wrong +Should I resolve this issue +localhost jboss working on this also take a look at the wiki documentation and checkrun the advance samples + works on both branch and trunk +Added namespacing of plugin files within the Plugins dir by plugin id in latest masterPublished as on npm. +Appl file with postinst section which modifies user passwords and etc as this wasnt a bug. +But I found that I inadvertently broke e +Closing resolved issues for already released versions +I was using sane build +Update No release yet for +Forgot to grant license last the patch last time trying again +for the delay +Test is passing in to update Remoting Guide +Because flow extends from flow as well the list state gets merged into flow twiceonce from flow directly extending from flow and again from the fact flow extends flow which extends flow +attached fix that unfortunately did not make it to BRANCH by mistakeMax please review +Also the forum discussion contains a patch +I just updated it +This happens because JBOSSHOME is incorrectly setCan I suggest the script error message make this clear +I will need the vmware setup and Esx to debug this further +That stack trace looks really long I wonder if you are running out of stack space in the class loading process and the underlying error is being swallowed +Verified bySUID added to moduleslunisrcmainjavajavalang at repo revision Please check that patch was applied as you expected before I close this issue. +Looks good +As you said the master node isnt so if it goes down the system cant index new documents or perform replication +Just a note Tested on JBDS GA eap package Linux have removed the Add Richfaces Libraries from portletbridge distribution optionCommitted to the trunk and branches +for the patch +Its unclear why they werent already there +We discarded testexternalsrubytest late last month. +While I was at it I replaced the parsing code withI elected not to have it ask for confirmation since this is only formatting the secondary and not a primary +I can submit as either patches or to a branch +There are some exceptions thrown in tests +Bulk close for +directory removed from the SVN trunk and added to the svnignore list review and test I dont have the environment running at the momentcheers Roland +bulk defer of open issues to +AlexTake a look at I agree with Alex the problem I se is that now the code it somewhere in the middle the filter is applied to the first column and selected columns +Resovling +I have committed this +it would be nice if somebody can review this patch + +is not public and therefore the needs identical methods +The two queries demonstrating the importance of parentheses actually returned the same result +Thanks for the review Alejandro +Im building on Linux w standard Clearcase +While I agree this isnt very backwards compatible as the reporter has said Sling is behaving in a way that disagrees with the JCR specs so I think it should be fixed as described +Patch committed to branch and trunk +it would be great if you could validate the fix +r uses JAVATOOLOPTIONS to set UTF for the JVMS that forrest starts +If not then use Default +Confirmed with as well +Go for it working on something else right now +The last commit may or may not help +This is not a bug rather a wish +I enabled these tests in the srcconf files in SVN trunk and passes them +Closed. +Well see if that changes things +Let me play some more +Coincidentally I noticed this same bug yesterday its a pervasive problem +On second thought we may be able to unwrap the proxy in the finder before returning it which would improve on this I guess +Verified at rev +Oops I didnt realize that I have to manually lock the record in question of Castor +If we add another mechanism now well have to support it too going forward + +Running patch through hudson +IBM has indicated that the fix will be available in SR for JDK +New version of patch available +So fixed by asking the compiler to generate classes but linked against the runtime classes +What is the diff between the HQ and theThat can give us a hint +In fact we could just make the doc or any context global +I realize that it is possible build it on my own and utilize it but not updating it on makes it difficult to create anything that depends on Thrift for the public. +If there has to be an op for this maybe would be nicerIt could reset to undefined or the default value if there was any +Could you AlexeyHard to draw conclusions from only one thread dump +I think the specifies that the bundle packagingtype generates artifact files which end with a +Actually the change I mention above stops the errors but also schema propagationD Ive sorted out what the code actually does and made a change that properly fixes the issue +An improper separation of concerns if ever there was see under the Codehaus Confluence pages that there is an Aiding Maven Adoption page Ive got a bunch of suggestions concerning that topic but Ill spare you for now +Added error log after doing steps described in Steps to Reproduce +We can also document it as a Known Issue +NPE Fixed in master +Issue closed for release. +Closing. +This test rarely fails in recent builds. +Maybe will do it +is my first time to upload the patch so I dont know which project is mrunitCould you give me the documents for help please? apologies Jiang I did not realized that we do not have the review board repository set up yet Ive filled asking for it +These packages might contain nodes where both properties and subnodes use the same name +The normalization of the paths has addressed it. +The repository factory referenced in the context configuration +Essentially the jdebjdeb goal will only run once in the lifecycle rather than submodule +bug brought forward to spent way too long trying to figure out why I couldnt edit my layer only to find it was because I hadnt defined my id column as a primary key +Closed. +Please disregard my previous comment I believe that the root cause of my issue is Rails +This is compatible the only thing that might happen is that jobs previously overwriting the queue configurations are now processed by a different queue +Any commentsquestions about this patch? Thanks in advance for reviewing +Too many tools count files with line as the first line +. +Reopened as patch was not applied +Wrong usage of interface +If it werent for some subtle problems with updating partition schemas Id probably have just gone ahead and made readwrite from trevni a table property of tables using the rather than have a separate +Uhh seems didnt work so well in the mapping there +The field preferred is fully implemented +Otherwise managing a local projects build can become very arcane since it will have to deal with profiles that magically appear and start mucking with its behavior +It has downside of some tasks being rerun on recovery but thats completely okAs suggested above additionally doing it on will even be more optimal +Renamed to an extra log statement I had added for debuggingIll commit this shortly +Hi Cheyenne it looks very nice +Since i reckon that the devices are distributed to uncontrollable environment i depend on the fix to avoid the possibility of spoofing +Got it working now thanks! Will raise ER for Eclipse to include generation is WS wizard +Marking as reolved and assigning back to Shane just in case he wants to veto or tweak anything here +note multiple openons now also fails in java source editor +An then we also have an? weirdI want to bring this to core too +This document also needs to be updated to include the change for +I found this JIRA issue because I was copying a script into the shell for testing and was confused why it was wasnt working +It is good enough for us to copy the system properties on session start +Would you please try the patchBest regardsspark patch applied at revision r thanks a lot for this enhancement please verify it is as you expected and Im looking forward your more work on +How about filing a ticket to deal with it +Thanks +Im going to commit this momentarily +Committed to with revision . +Thanks for that Caleb +The database and UI are getting updated with new disk size after the resize operation +Thanks. +Most of use use Linux or sometimes we forget that even Windows exists +Hi Sebastian I looked into the module and it does not look so bad +Anyway just do what you think is good +I cannot compile with the java ant task as described on the wiki using +Refactored and merged BMT to MT codepath +The code is complete and so are the examples and docs +Then adding additional features in another context is no problem as this context would only use someone which wants these and doesnt need the compatible to other processors +There is a conf unused variable tooAlso fix the additional test case failures +AnilYes you are right +This produces and +How I can test? Are there any snapshot builds that I can download +perhaps thats just a problem on the Groovy Web Console +Will fix the test right away +Final patch for commit +It does not modify mapred nor distributedshell code +Are you really waiting or are you just wanting to get off of local forksAim is to get out rather than doing a +Thats why some data from changeslog succesfully stored but must be not. +I just checked that we had this file in as well if the customer just unzip at the same place and we leave this file out from this release the old one will be there and could cause confusion as it indicates an older version of the components +Ran a pi job +Closing issue as version has been released on. +On Windows I am pretty sure the Actional people have successfully used Explorer to upload files +Hi PaulexThe fix looks fine thanksBest regards by Andrew. + +fixed c +Tests were created and put to Metamer. +To avoid overly complicating the patch Id probably just add an option to allow users to enabledisable node exclusion based on numeric ips +Thanks to Antoni for the patch! +lets do it to revert the change! +Ill look into it perhaps theres a hidden bug when a bean to be editted has just a single property +You may review the branch later on and I will address your comments in separated if there is any +In a nutshell we now give back directory entries as well as file entries +The jmx interface also currently has the check for if user has administrator access which I dont think we want here +In a plain Marmotta this should not happen +Can we please rename the method and update references to it in the compiler +You can probably fix this one and send a pull req +The backup namenode continually loads new edits and ends up creating new startup progress steps under the Loading edits phase +If you could write up a jira for that and mark it as high priority so that it gets fixed in I would be much obligedThanks Leonardo and sorry for the mixup I have like pokers in the fire ATM +Seems like the current default user name is not used when opening a connection in the setup +ETA unknown so moving to TBC for clarity +Chinthka can we resolve this issue +Updated the mime type list with the types from Apache httpd in Rev +You can just throw in an XML file with some custom DDL declarations that will be used by +They describe not being about to move the cursor out of the quote block in order to type around it +Having the same problems as Michael I cant open a rule asset that has a date picker in it +contribsounds reasonable to me +Fixed the test and +Therefore resolving the issueThanksIsuru +and what do you about the surplus cheese from this carve out +So for that does not really use fidmapper i guess we would have to implement interface directly + also covers your proposal of adding only RFC MIME support I leave it to your judgement but what is the client support for it? The Classic inline PGP is always available to anyone having PGP or GPG installed so isnt it a pity to miss its support +Can you tell me what the fieldtypes are for your sort field and collapse field +Delivered the fixes to CR and trunk +OKI will commit it if the tests +ThanksChamikara +In the attached patch I added an additional test case for an unknown short option just to be on the save sideSo this issue does not affect CLI +I think that is a very good idea the above script is in fact calling the API just making it more convenient to do so +Thanks BrockRev fdbcabdeadddbacf +The only thing to keep in mind is that such configuration is static for the lifetime of the endpoint but I guess your client isnt going to be expecting it to be any different +So the and deps went to back to the solr contribs +We generate a lot of garbage terms when parsing things like time stamps malformed attachments which parse badly etcAfter the change the big time waste is converting the terms from utf to utf when reading from the +Handled Throwables other than Exceptions too. +Youll see the different sources +James have added the new console that has this feature +Do you have a JIRA number for the bug in destroying? Would that be the cause of this WARN logging accessing the seam application +Tabs should be spaces +maybe I should reopen this since I didnt hear anything back +check that importing of existing projects into workspace is described there should be written something like that Go to Main Import General Existing projects to workspaceSelect root directory test all checkboxes near test All projects and parts must appear in Package Explorer +jax Post Servlet is long gone marking wont fix. +scrubbed rn +Will commit the example shortlyChris. +For your suggestion which place do you think I can put these filesYes I can only invalidate module changed since last time +Could you please take a look at this regression? Thanks +I read the JIRA title and couldnt make sense of it how Flume and and ES are related knowing Flume can write to ES but so canbq +I can test this only in two weeks unfortunately +OK +scm connections are now rewritten by release plugin +So making the fix version for +Thanks for the review Claus +For the use cases do we know how Accumulo is used today +so I didnt refresh my page in hours and missed Sebastiens comment +Has been fixed so closing it. +Updates dist target to include all files necessary for a release +Was this something that used to work on? How was it implemented? Cant we recursively introspect a jar if the contribution processor finds the current artifact being a jar? But I think that we might need to manually add the embedded jar into the classpath to load the resources +This patch adds dist module to generate release tarball +Thanks +If theyre not then thats the essence of the bug and we should fix it +Moving +Once someone fixes this issue we should change the documentation to remove the restriction imposed by this jira entry +After fixing create the at s set method +Perhaps Ill need to hardcode them as well +I hate to be holding up good and important work but dont know what else I could doThanks to Rick for the reviews and for running the tests +This command needs groups of arguments and the number of arguments needs to be a multiple of three +It is not generating the wsdljava packages your application built with will still reject the usage of them as they use the Enum type +I was taking a wild shot in the dark obviously missing +I found it because i was looking for a query equivalent to a nice point +it stays in Beta as it is now +It was too much detail for an overview +Svn diff of changes needed to use json message object +This problem is not present with other entity processors +HiI didnt have much time to look at the patch but a few remarks The USERAGENTAPPLICATION should be changed +cant you mixin GLTC +Sorry everyone +Hi Guillaume sorry for the long delay +deployment error found when Guvnor use jboss seam with seam does not fail on I ll try EAP instead +committed to as eebbdcfbfbdfaca +We need to add some documentation and also to add the proposed configuration as a default value in the server +Configure UI for next release for JSF Portlet Capabilities page so that it gives you a short description of what jars should be there? For a user not familiar with bridge its notIf you have any suggestions to the new library framework please create a new jira. +That looks right to me +After that the messages regarding Not a directory and Is a directory disappeared +Dan had commented on my earlier patch for jdbc clob blob methods +therere patches +sorry about this one guys +This however sounds like a new issue to me +Ill test and update +This was mentionned in as wellIm also interested in a solution to this issue +again Knut +It should be fixed +Dhruba can you review this +Do you have a specific scenario where you are specifying a as well as custom editors with the former overriding the latter? Are you specifying those custom editors yourself or are they possibly being registered by framework components? scenario is trying to provide a converter for StringResource +You may want to warn the dev list about the need to manually rebase any local Git commits. +Wouldnt closing the socket on all exceptions make sense since we are sending a fault back in any case +The sooner this gets in the fewer other patches out there get broken +All features installed +I have a class in sharedclasses that imports +The bad credentials error shows up in coretarget +While itd be nice to have a full showcase and documentation for an alpha I don t think its necessary +I swear I could before +Test zv from XSLTC groups internal testsuite fails of this bug +Will add more documentation in +CDI TCK contains jdk sig file +Perfect! I had no issues to apply the CP +I think Jeff means that Pig does not support without parameters but shouldI agree + MojoLifecycle rewrite for latest Maven have verified operation under maven and these corrections do allow cbuilds to work again with maven +Patch looks fine +I am attaching a version for Hive +reassigning my works for me on linux +If so we can publish that for Beta +Heres the patch with tests and Solrs factory changes +Having faster builds sounds good +BtwI found only one instance Divesh atcomponentexamplewidgetexampleprint +The permissions associated with the user have nothing to do with the session id +You can simply move the Publishing to +Is there any solution or workaround to this issue +Added a little sample on the aggregate eip wiki page +You can turn the security off but thats a option not a buildpackaging time one +not for objected. +and the best thing Math can be used as category +We shouldnt put to much magic down in the guts of anything here +MistakenI will back out the patch and suspend release ofI will back out the patch and suspend it until release of +I set the incorrect Affects Version its +Also the above points mentioned by Hemanth are taken care of in the new patch uploaded +Great Hardy! So is the difference between discrete faceting not supported and supportedIndeed lets fix the docs in this dev cycle both of you +Both are simlar behavior but env is different +Close this issue to make things clear +Closed the nd pull request and added a rd one which contains the commits from the previous one plus a couple of others one which further improves performance by updating indexing only when property or path related information has changed one which fixes the index rebuild options in the AS subsystem they were ignored up until now and the changes brought in by the previous commits exposed this +Beyond that the bug is all yours +I never said anything of the logic is correct and I made a mistake in the forum post we dont assign the last parameterless Remove method to be the default we throw an exception if there are more than +Thats right +Patch for branch so that others can test +Bonaza! This made it work +bq +Unfortunately I dont have a installation to test this out on so Im just going to apply the fix as recommended by Jonathan +Attached is an initial patch for this issue +I will commit it soon +Setting patch availabletrue and target fix version to the fix for this Ill close this JIRA and open another for the test case which will allow inclusion of the test case once the groboutils licensing issues are because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +All works now +Norman will maybe put it in this week + +This will catch cases where tests ignore and therefore ignore a that is turned into a or a testcode path is swallowing +Patch looks good +It can be finished today then I will attach the patch for trunk +The book version makes it clear that the properties section is gone +This issue has been fixed and released as part of release +A translation is done +Oracle Oracle Oracle +This is under testbenchmark +Fixed interop tests in itesttestinterop +I dont think it belongs in the docs hopefully people will find this comment instead +Regardless of the Dialect the length will not exceed +bq +I think this one would be good to have. +A patch which adds this functionality +I cant download the demo files +Seems a reasonable inclusion in forgot doing this will require changes to several tag libs +With ignore I get a which cant be caught and causes login failure +In theory using this scope breaks backwards compatibility with Maven versions older than even if I cant reproduce the issue with maven when building and executing the are you sure that this new scope is mandatoryExcept this detail the patch works great +Hi Thanks for applying please note that the original issue description includes more things like agent integration et al +Even if the click ends in the same +I agree that we could do even more to help new users +Closing this record since is now released. +Patch contains a fix for empty jar path +Whats wrong with defining different actions for different methods +Fixing Affects and Fix in Version +Thanks for the clarification Dag +CreatingClosing a connection is supposed to be expensive +Actually I think I see +Nor does it disappear from the screen once an item is let me know how I can resolve this issue +Spot on the domainTypes is an array regardless of whether you have one or more should have been more clear earlier I added a new validate statement with words to this effect +Committed to subversion as revision +in path specification +Dear Steve Dont feel it like a moan man most developer like me never get time to solve Hibernate or similar big frameworks issue due to too much of routine works +Its correct +This is confirmed as a blocker for BRMS passwords for the admin and mailman users can now be set in the some clarification for the release notescan now be set in the Does this mean that those accounts now pull their passwords from JAAS setup etcOr is it configured with specific elements in +defer all issues to +What is most accurateFile tgssetupjavaenvirDo we really need to issue both java version and sysinfo? The way that the text is currently written it seems that sysinfo provides the version of the JVMFile cgsusingtoolsutilsThe title refers to tools AND startup utilites +Add method encapsulates calls component creation and also it will add the parent to this created component +The documentation on the web site has to be manually copied there because it is for a specific version and because it requires Maven and Sphinx to build it +Bad reviewOne other thing we missed is that currentAttempt can be null at any point of time may be we should document that + +Fixes the issue by executing the first class in the file even in case of inheritance +Attaching patch which seems to fix the problem +I can do that +seems like a very good idea +Id prefer to see both Server bugs fixed so Ive linked themHopefully a dev can look into it +patch so the revision numbers still match with +This is currently a double arrow +So all issues not directly assigned to an AS release are being closed +Well it can always create a new file too rightWhats the tradeoff between appending an existing file and creating a new oneThe future problem a db file growing past the point where it can be anticompacted is currently hard to keep track of and hard to recover fromAnyway it was just a suggestion +I had excluded it from the + Sending assemblysrcmainassemblyTransmitting file dataCommitted revision +Fixed interop tests in whitemesaround +final patch i will move forward to remove this trap +Needs to be to +Closed with the release of. +snapshot directory and its tree if the path is a snapshottable directory +Thanks Paul! +Patch for the the new test case for class fails on RI +There are tests validating if a bogus user would fail as latest run was clean +The infrastructure is already there and the FTP and FTPS producer endpoint already behave like that so I codsidered it as a bug but I have no problems with changing the issue type from bug to improvement +The reason is that the tracking URL is generated in and the parallel tasks are generated in Driver +Updated patch +Was that what you wanted +NET WS client +facilities are outside the scope of the project +TomcatSun JDKWere upgrading the JDK to see if that makes a differenceWe have tried Cocoon yet +The issue has been fixed in currentCVS. +Perhaps the Transactional annotation used by Spring would be a reasonable example to copy. +The use of classpath was required +Added shared consumer test case +I see the in common is in the jar +looks like I could add him +However upon testing this Application Server that I build still has the same problem as before +However CP is very old and should not be used +Actually there are a bunch of obsolete options referenced in there +Hi DavidHow is the migration going on any updates? We are on the standby to test the Git repos +from me +the needed patch is in issue ! +Closed after release. +fixed is just a wrapper around cobertura so you can use it with MavenThis should be fixed at Cobertura +Can we add a option default to the WT select on query tab which forces Solr to use the default one? Also I think we can build a select list directly from the current solr configuration +Committed branch and trunk +We think this problem has gone away due to some other fixes. +Fix the test failure +What I think Ill do here is remove the +What! This is a major bug that is easy to reproduce +Bulk close for +rolled back in +This means that Claim Check only and also represents the Message at the particular state and the same Message could have many states each backed up by its own claim other words we must support two scenarios for and the more I think about it the more I am convinced that it should be the responsibility of the to remove data based on some flag that we need to introduce with default set to . +Hi all Im here +Patch applied +We must have this functionality in production environment to isolate flows for security reasons +Should have one ready in a few days +Mistake happens because of the environment including Jakarta in the distribution of Tested with Jakarta and everything looks like it worksThanks to you +Oh touch +That would mean trying our best not to change the unnecessarily +What Randall said +This one wasnt what i expected as the bugappears to be both in the spec and in ruby itself not in spec was wrong to want to test comparison by eql? when the actual comparison is done by on matrix rows and vector elements +What is preferredCreateEdit screen is fine +So my addition to the issue can be ignored +please take review the tests +Hi Brandon Oooops +In particular that +I know it is kind of anal on my part but I have be pushing to keep Taglist at zero CPD PDM and Checkstyle for your work on this Christoph and I look forward to this feature in. +Fixed a few more issues based on Igors comments +Sorry for the belated updated review +Why are we using reflection in the first place +check Bean responsehttplocalhostaxisrestsingleBeanvarxyzvarvarvarvarvarvartrue +This needs to be included in the distribution and documented in the user developer guide +branch revision trunk revision on. +Lucene trunk which will be Lucene needs Java +We have been upgrading spring to tie back to changed with project and we are in a bit of bind as to the versions of spring that we may now select from +The old client treats it as CHECKSUMCRC and runs into issues +Alternatively we could do a more granular check inspecting the GSON library for a known working version +Whats the status on this? Id like to commit but is there a new C version out that we should go to +The first test case uses static values +So here is the ideaSince the file is empty there is no TIMERANGE stored into the file +A deferred issue cannot have a fix Baeyens Lets fix this in jPDL Thomas Diesler A deferred issue cannot have a fix versionThen why not leave it just open and put a fix version of on it? That way it does not get lost +Jozef could you review the changes Ive made and check you feel it sufficiently specifies your scenario +Committed this to as well +Waiting for PMC in order to decide if and when it should be upgrated as a supported module +fieldName price type notNulladd field date jpa class +Thanks for the review +I believe some work was done on an implementation although I am not sure where that has ended up +Prashant AvinashAre you guys working on this issue already? Thanks +I suspect its easier to figure out why its failing now rather than bisecting +I think this issue should be marked as version upgrade as part of a CP release would be inappropriate and against addition to this there is no GA release of the branch and no TCK testing has been performed so upstream work would be required before we could consider it in an EAP release + +They should be placed into the landctests directory +Closed after release. +since this post has been a year already +Let me your email isnt uptodate anymore +Plz have a look +Check out the output from class C I think this is unexpected behaviour +what can I sayI think it took them a while to understand why its optimizing away +Created to trunk +This should be fixed as well +Patch to track remote tokensendpoints in a new key in the system CF and load them on startup unless is false +Is this regarding Commons +If someone feels that the logging isnt enough and wants more feel free to reopen. +Can you be specific about what version you are using +Sorry +I thought I got them all? I had a tough time reading this particular xsl +BTW Steven if generating the patch with git make sure to use so can pick it up +So they had to break backwards compatibility to move to Eclipse +renaming breaks quickstartsconf in platform branch changed to look for +Cleaned up the formatting and wrapped the example in a proper code is my TCSH flavor of the helper aliases +Not a problem in current HEAD +This just needs an updated Xalan jar in our libs +JAVA we remove it +Tom do you want take a quick look or run a RAT report with this patch +The patch applies to trunk reviewed the patch +Makes sense +I think we forgot a rd goal of the Big Lock make sure that when a memtable is full all the flush threads are busy and the flush queue is full we want to block writes so we dont OOM from shoving more data into the heap before we can finish freeing what is being flushedIdeas +Applied +buildivylibHadoopcommongriddevbin +Disabled as in deleted +Trivial patch to throw when listenaddress is +b The purpose was and now replaced with a rather trivial semantics +ive applied on M xavier do you put in trunk +the last one wins +Submiting patch with Devarajs comments incorporatedAlong with the comments This patch addresses couple of things +Setting assignee to Max as it shouldnt be me and I dont know who it should be +Thanks Liyin +LGTM for the review +Closing issue Oozie is released +Note that this patch will get a from hudson because it doesnt include any new tests + +I think its closely related toWe are really close to find the solution for this problem +This you are talking about is a new issue +Made it so we build in all modules Did pretty print on our poms + +Then our hash map wont have this issue +Even if it requires an additional configuration element or something +I am not specifying any custom editors since I assume that this is supported but not only follow your argument if I assume that are the preferred mechanism for supplying a conversion strategy +Jira project created with standard layout using key peterI dont know if I need to ask specifically for a confluence space but I would definitely like one +committed to trunk as revision +I will look more into thanks +Communication with Alex has resulted in a new tarball for the attachments and a new xml export file are being produced +I see no reason to not be forward compatible with the JEE spec as long as it doesnt break our existing spec compatibility +Ive verified Todds patch fixes on x and committed to branch +Initial patch for review +Both provide a implementation +Thanks Chris and Chuan for the review! +OK is this a Date field or a Contacts issue +Now that is fixed feel free to mark this issue as resolved +Integrated. +Want to start a discussion on the DEV list? Lets not clutter up JIRA with discussions +No no JrockitIBM +This feature is very important to us at Wanted Technologies +AgreeIf IE is OK then close it +There is little chance of this happeningIs there some compelling reasonIf this is because of some internal requirement in your app you could always map the +yeah but the really is ridiculous if you try to use a large dictionary with it even without cutting over to an FST it could probably be improvedfor minority languages without really nice dictionaries it probably doesnt matter much but forthe languages with really nice dictionaries you also tend to have options availablejust another crazy idea I dont know how much of morfologik is dependent upon polish itself but if it already knows how to compile ispellhunspell into an efficient form and work with it maybewe should just be seeing if we can generalize that and work it from that angle +attaching proposed new Tommaso fair enough and thanks for your commentI have attached a new patch with the test aligned +Reopening to set Component to have moved the Release Note from Seam Resolved Issues to Seam Resolved Issues as prescribed +Attaching patch +for the patch +Ive just looked at the test cases and it appears that all of the tests that generated the bugs are disabled in the code which I assume someone did because they wanted to make sure that the bugs didnt prevent test cases from do we want to handle this +The fix is to use the class to compute periods instead of calculating seconds elaspsed in a periodIll create a patch in a few days if someone else doesnt get to it first +Also contains changes based on peer review feedback. +Yes the nd version is working fine + +for has been committed by BorisResolving. +Ive not seen this error for a long time and the reference to jcc is suspect +I actually have an implementation of that knows to pick the right file in different runtime environement +Can this patch be backported to Pig? were wondering if this change can be committed to branch +batch transition resolvedfixed bugs to closedfixed +for the patch +Alternatively could each component run these tests as part of their own job provided that they set the correct settings in the test plugins +Proposed patch for +Also the jetty identifier string in will need to be changed from Jetty to jetty since it reports in lowercase now +You may well say that this should be changed by the project +Well at least its consistentLet me bounce this around here and see what I can find out about this particular set of exceptions +I need to look at the jdk release notes in more detail but at a glance it says Java applications invoking JDK from a legacy JDK must be careful to clean up the LDLIBRARYPATH environment variable before executing JDK which makes me wonder if it applies +New patchset is refactoring to set up adds SESSIONFINISHED reply so bootstrap target has time to do ary index building before source declares it finished are old split ary index build out from bloomrow index build and move into stream session that we set up in +That cant be OK +Ive been maintaining a backport of this to and the attached patch is for all intents and purposes identical +Verified by Paulex +MThanks! Bruno Aranda Mark Sanders There all is ok! I found in commonlib + +Applied to At revision +This is a third attempt to fix the truncated download page +Also the code to count needs to be removed and instead compare against this map +the same problem may apply to foinline +Hi NicolasI found some spelling issues WIP +I just needed to add a dependency of onI took code from Daemon and bent it to my willPassing an object as a parameter is a bit dangerouspromiscuousWith the attached patch I was able to start the server from the eclipse plugin +Being able to assign values to nodes will be useful but I was thinking that until we are in situation may be we should reject it in a more informative way +Not supporting it we can save some time without lowering usability of JBTIn case we have only Central its less work for QA and all update sites used by Central are already tested automatically via automated testing of installation from CentralAdvanced users can use installation from Extras update sites but we would not guarantee it will work for +patch Davide patch works and code looks clean + Bug has been marked as a duplicate of this bug +Holidays and allI applied the patch to as that is the trunk now and likely next version. +It merely improves stacktrace reporting +What is the reason for upgrading antlr +Patch reviewed +If some puts fail it continues with the ones that succeed +Thanks for the bump Andrea +Committed revision Thanks Mark! +The old mapred API is being for and is not going away soon +A new instance with the same field values as a persistent instance wouldnt be considered equal even though they are semantically equivalent +Client for any case i have used same test conditions and procedure for both and +currently assumes that the new configuration is being used because it will call get with a null configuration that loads the new +Claus luckily I havent had a chance then to implement this with the code change you mentioned do you think this improvement is worth keeping? I suppose some folks may like keeping their schema in various different locations so perhaps yes +tweaked diff +do you think we are at point where we can consider starting to write some code? Another thing we need to do is identify which of the functional tests we are going to port +In my understanding there is no work needed for this +very nice! thank seeing this yet in the JAR +Thanks +You need to fill in the description of the issue as accurately as possible +Fixed some some fingbugs warnings + with maven when becomes +Then the source processor can ignore anything under that single directory +Yeah I dont believe we can force our referencing subsytem to use a certain abbreviation for meters so I fear well have to read the prj apply some dummy transformations on the string and write it back +What is the average size of the files youre sending to Solr? How many files are you sending concurrently to SolrI believe that Solr Cell internally saves the parsed content in a String before it adds it to the index +A schema is internally represented as a Map so no problem writing to it but a POJO as an underlying class best adheres to and for merging needs to be Serializable +Thanks again +After fix the Apache configuration we will see if it also fixes this one +Ignore the first one +That releases is tagged and ready to go and Ive tested it in the JBPAPP clustering testsuite +I couldnt figure out how to do the job cleanly after hours of debugging +Dkulp also provided quite a few fixesI think the issues left are plugins like the assembly features +Small patch that merge the smallest CF with the other one instead of always the existing one with the new one +I vote for including this patch into the forthcoming release agree +It does allow it +Thank you very much for the great work Junping +patch implementing longpoll option +Ill poke around. +Thanks + Possibly found by users in the communityhttp +Mail sent to announce general +This caches the lookup once per Hibernate transaction rather than once per +You dont think that is a valuable contribution to the seam documentation? That documentation cannot possibly be used by the Hibernate documentation group as it is completely except for literally one line +I think so but thats Brians call I dont see any reason not to include this in a CP +Thanks Nicholas. +is also very new +I am not sure if that is a fix for this jira since we dont know what the bug is yet +note since usage only was and is proably more trunk to and branch to so that we can distinguish reporting installations +Hi DirkGlad you figured out your issue but just Im a little confused if you are defining the worker in the XML configuration then calling will return a singleton +In order to get it worked tfpt should be in PATH. +Ill commit this shortly + +Patch to solve the is applied in revision +This issue is fixed and released as part of release +To be precise related to the inAnd in relation to groovysh in and +Actually I was right the first time fview has an encoding attribute in facelets as does fform. + tested! +The issue is rescheduled to as it could not make it for the cutoff +I put the test in which is to be used by +The trouble is that this resulted in the standby not having sufficient transactions in the shared edits dir to be able to become active since its fsimage was so out of dateIf my understanding is correct lets please change the title +have no idea how that got missed when I did the global search and replace +Attaching a simple design document that says how tags will be supported by and the advantage of using +The patch looks good +Thanks a lot for the reviews and discussion everyone +Most of the original functionality is moved to +Yes VERBOSE should be good +Thanks AkashYour patch for formatting changes is in trunk rev +Attached the patch +It introduces an incompatible change that made get to return null for a path +It seems that was never applied externally +Would it possible for you to develop that piece of code and grant it to the project +This appears to have been broken when was committed since it changed the way reducers locate map outputs when not running via normal containersUpdating the summary to better reflect the issue +Closing resolved issues for already released versions +Please confirm the and then close this issueThanks for a late replyAdam gave me a clue so that I can fix this issue finally +I get the same problem for Oracle +bq +Please confirm +I am pretty sure this is a bogus error message +Sanjays idea seems like a useful capability in itself but I dont think it solves the whole problem +Andrew looking at the patches I see that all of them are now in Pigs trunk +should be release either today or early next week +how to jive what is returned from a SQL resultset with Solr unique keys +We should make that consistent +Working on a better fix +i would suggest clsoing this patch for now + +And I believe this has been done for a reason +bq +Closing with the release. +no that wouldnt be a good idea +A couple months ago I made a contribution to remove the direct dependency on Spring so that HISE could be used in other frameworks such as and SCA +This will allow users to easily be able to test J authentication with LDAP +I think they just need to be updated to use the new method signatures +Fine by me +See. +Ill commit soon +The upstream bug has been resolved in +We fixed a lot of problems in this area in the last months with introducing the dynamic class loader changing the repository classloaderand some areas in the scripting these changes fixed several concurrency issues during compilationHowever it seems that were still not through with it a new problem is tracked withAs weve changed and fixed a lot I think this is a duplicate of therefore closing this bug +Any further requirements after please reopen! +Anytime JohnIll update the wiki +One use case is embedding the XML files inside other XML files for tooling but my use case is to allow them to potentially be embedded in spring containers +I wanted to integrate Outlook parsing with so we could make the magic autodetection work also for Outlook files +Ill check it tomorrow +toolsjava +Attaching a test case from Carrot that causes AIOOB in a deterministic way +Just make our strategy such that we could later slot in more analysis do crazy AND combined filters in the way the Filter stuff works +However contrib doesnt seem to be set up to run with maven at the moment +Thanks +I dont see this as a deviation from the spec for various reasons +Im not looking that code right now its my weekendevening job but Ill give that a go this evening and see how it goes and report backCheers +fix by Matthias is a combined weblogic log containing messages from both instances +Also I didnt implement a conf option for unique key + +Hello Paulyour commit fixed the issue +mavenrepositoryjarsThis patch changes that to CATALINAHOMEsharelib applied +Will edit the comment +Understood +will change all null string Noit not resolved my problem +The first file is the patch for the projectThe second file is the Jspresso DSL AndrewHere is a new proposal +They should be consistently called +Verified on EAP. +Carl aside from semantics what would be the valuedifference of the temporary topic instead of the queue +Unfortunately I have two comming deadlines so I will not be able to do it before March +Dont think so when resource is in another bundle. +If you add that does itstill fail +the problem is how to resolve the issue then +How are you signing the request? Are you using the support or configuring the WSSJ interceptors directly? If using the policy support can you paste the policy in here? If using the WSSJ interceptors directly can you include the actions you are configuring inIt looks like however you have it configured its putting the key directly in the wsse as an ds element whereas the service is expecting a wsse and a wsseReference childPotential solution given over a month ago +In general weve spoken about sidecar profiles that could be optionally downloaded providing more specific subsets of servicesThese are their storiesdoink doink +Thanks Klaas Ive committed this to too. +Ill do some research. +Okay will do on all points Gary +Committed revision Thanks again Diogo. +Solved NOT a Lucene reported himself commented that this is not a Lucene problem. +Thanks to Jakob for provide this javadoc does not have getter and setters for them so we have to ask EG for this and if is true update our code +The false link should be disappearing on the next export. +The replication test suite completed successfully +Ill take a pass at this one this weekend +Mark are we broken again on Windows +So we will expose the keyword file to the user +Sounds good +Ping +Snjeza I dont have MPC installed I can drop the links in some areas of the workspace but not in the Central view +I have the full logs for runs of which failed +I only asked for LDLIBRARYPATH by the framework for the convenience of the application programmer not to have to write a wrapper when the executable needs some shared libraries not already deployed on the nodesNo big deal +Fixed in headSending modulessrcjavaorgapachegeronimoaxisbuilderSending modulessrcschemaTransmitting file dataCommitted revision tck plans fixed +Thank you Chris and Stevo +Sorry for that surprise I created! By the time you saw my post and commented I tested the code and found it to be a bug +Modernizr is not meant to be used to swap between min JS +And the JRE internal xalan is almost certainly what was causing the changed behaviour in owsAndrea can you explain what is going on? Does this changed behaviour have any implications for deployments if we no longer ship xalan +Ill hopefully be able to take a look in the next few days other work to fixed this +Best regards started to feel that this issue never will be addressed +They simply have to decide on what their column families will be building some information about time ranges into them and assign them to locality groups +The name was changed to better reflect that it is now readwrite +yep +on the rest +Btw +It is not enough to just use at a time to determine the collation info for the final dominantDTS +Shall we resolve this issue as Wont Fix +Other possibility would be to include skip parameter for that goal +However Im not sure that we can amend that behaviour short of a complete parser rewrite +xml files are empty by default they work fine +Clirr no longer errors. +Resolved as part of QE work on this document +Have you had a chance to test your scenario using Geronimo? Was waiting to hear back before closing this Jira out +nothing to be working now with a bit of modsubstitute magic. +Thanks Rohini for the fix +This call is returning DOWN even though the ear is up and running +AFAIK the only changes made are to the makefiles etc to change library names and so on No source changes have been madeIf you have any more queries or want us to try anything out just let me knowRegards reply from the reporter +Initial should work! I was storing something similar to saveCount in threadlocal map which can be saved as an instance variable +Proposed patch for this issue +Ive rebased the system tests so that unit and system tests passes +Validated with a variation on the duplicate suppression testAm going to ensure that set ignores duplicate subs such that it does not get in the way of normal durable sub reconnect where the networked durable sub is offline pending a reconnect +m currently using too and this debugging fault is a big issue for our project +It works with the jdbc port +There are some problems with nexus on +If you are still having issues I suggest raising a bug with the Spring Roo team who will be in a better position to track down the root cause +deploylib +Committed to. + +The fix for appears to also have resolved this issue. +Implemented in trunk rev +It was already verified +but N should definitely be copied to invoked command lineI fully agree +When there is no Jsf Core on the page it will be added automatically +Raghu Yeah I did +Pull request for docs has been merged +Otw the code is unchanged +It should be done in ThreadinterruptedisInterrupted logic +Im currrently trying to figure out why +Two big issues here +The is for branch +Fixed in r. +Patch supplied by Matas closing all issues that have been in resolved state for months or more without any feedback or update +Thanks JeremyToday I ran the single node test with your log above and got even stranger results instead of giving me NODEEXISTS errors I was simply overwriting nodes +The line number in the stack trace looks like pool +Are there some user level instructions as to how to configure and runThanksttylDima +Thanks Greg and Elliott +Yea sorry sorry about that +This issue deals with multiple hierarchies in the same project group not necessarily a flat least we could disable the Release button if its not going to work +Hi DanielIve test the new patch all the systestsjaxws tests are passed +We should request params for URI using only node uri instead of both of node uri navigation id +The Tools manual Tools manualChanged placeholder name from databaseURL to connectionURLTools manualChanged placeholder name from databaseURL to connectionURLThis topic mentions connection URL attributes so thats where Im putting the to the Reference Manual +But we should use the same concept its used at different places +Also the tests needed updating to work correctly when the system running them was Locale. +People can configure JMS in Camel to use if they want. +Attached is a patch against the HIVE trunk adding the code that Carl describesIncluded is also a small change to to exclude inner classes from being treated as test cases themselves +the patch looks good to me +I think this one is still left to do? Demo sources are no longer shipped in a binary +What do you mean by app? Can you describe the deployment please +OK +didnt list ant at all and I couldnt find any Debian or even Ubuntu packages that included anNot sure if I was doing anything wrong but I reverted the commit and was able to go on and build just fine +please change in simplifier looks fine to meRegading changes in EA why shouldnt compilation fail in release mode? We may put jitrinoassert there +In this patch Ive documented some of the limitations Im aware of +Attached original patch +this is deliberate as indicated in the source suggest its closed + version dont support gwt version does and is releasedWhere can I get version from? +Waynes recommendation is a good one to follow and if possible not using native libraries at all is even betterI didnt see that theres an actual leak for a loaded server +fail the build +isnt a database! If you must use a use a instead. +Obviously people should read the whole page but since the resulting bug is very subtle I think its helpful to be a bit more about this behavior +I dont intend for WCFInterop or qpidxarm or qpidclient to be placed in a static library they would all remain +Actually Chip should probably announce this feature himself +We see a few notoriously flaky tests fail spuriously pretty often but not usually these and not usually with timeouts +When running test case I got an unknow exception +Resolving as per Igor. +It no longer reads like a chalkboard with the background being blue +Werner done makes sense + +It seems that the specs directory is no longer built by default +Adding the updated patch +It would be great if you could run with true and post the contents of on that issue +I worked today to address that but if someone instead promotes a patch to remove Avro RPC altogether then Ill stop trying to fix thisSo no I do not intend to turn off the test +This issue has been on my list to report for a while +Bulk close resolved issues +But that seems like reinventing a lot of TCPIf the datanode goes down then currently the entire block is in a temp file so that it can instead be written to a different datanode +Karlthis is great news let us know if you have any questions +checked in trunk +I wasnt successful verifying the above mentioned thing today I couldnt install EAP from the errata +Eike We dont have unit tests for now so we can only base our selves on code reading or implementation usingI dont use so I can only say that your both points sound logical based on code readingI will commit your patch +Bulk close resolved issues +Defer to all open issues to Nate notes the Java backend can be made to deal with this by inserting appropriate is the plan for the C backend to deal with this +Reviewed patch +See also believe this has been do you have a test to confirm +but at the end of the debug stream BUILD SUCCESSFUL Total time seconds No emulators found to deploy to +The test case looks fine +It is possible to not reproduce this problem +cant be released without this because now calls a native method so anyone on ubuntu with installed will run into this issue and as a result well get lots of reports +Could you also post this sample response so that we can check if the content type could be guessed +Updated patch to trunk +Fixed ready for +log to trace and do nothing +Need to work out a good way to test it before applying +If the flag is set the log shipper thread may exit its loop without notifying the waiting thread +No further information on this isue so closing it. +I can attach the Service Class implementation if it will helpAnyhow I think that the is creating this SOAP message so maybe the problem is thereThanksShahar +The only thing where Im not sure is the code which picks Unsafe or reflection based on the availability of Unsafe +The bug in openejb has been fix in its trunk and we still need to wait for the latest openejb build uploaded to maven repo +abstracts the it already provides a window for its customization +Running within Eclipse works so its definitely a problem with loading the wrong configuration file on the classpath +To fix this issue we need remove span in which we jsf inputs and update junit tests +Includes passing cloned subtree with joint children instead of full tree testing prefetched objects in account all your comments in that new patch +Apparently the native cvs provider only handles the first Geert +Of course at runtime it will still be contingent on the wrapped collection object itself being Serializable +Its a small change to JIT code +Presumably I hit some magic key combination to change the method name to lower case while investigating +I think its ok for a commitFrom the code I read we should have the same behavior as before on split +I thought of some things to add to this Ill try and get them in tomorrow and attach another tar or namely support for not running sablecc everytime based on a stored timestamp from the originating grammer support for a default mechanism that grabs all the +Delete is a GetPut is a Get is a From my point of view the only major flaw is the Delete and Get arenot doable in operation which also causes compare and set to befractured into API callsryan +Cant you use fragment bundles instead of adding bootdelegation entries? This doesnt make this issue invalid of course but its usually cleaner +Completed +The patch attached +We should add a link to the build results from the web site probably from the Project Info section +theres no new designs coming in the near future +Could you please write a message to the dev list +We should backport the ability to do so to the tomcat to as this is not a should be documented in the Release Notes as a known issue +Im not sure I got all the real code changes in +Karelalthough tests passed your line fixes indeed look valid so the files in are updated accordingly +I already have some patches which migrate the Velocity engine to JCL and +Good catch Freddy this is now fixed +Wuhoo +Verified in EAP +CDK should add this attribute by default for subclasses of +What if we just made the minimum token for MP instead +Resolved in with r +Thanks for this +The test was written assuming that the default transaction type was datastore +Whats the difference between a lst and an arr? From what I can see in the enclosed sample they both behave the same way +New patch giving credit to Christian +We use content to get highlighted snippets for whatever field is matched by the main query +Not acute right now +I dont there is much justification for this +Hi Donaldthis patch fix some minor defects pls help me to commit it thanks. +Despite the fact its not a bugfix it is a very useful feature +slava we should allow EL for this in the editor AND we should report it to the ESB project with a link to this issue +I would think it would be most common for scripts to check the exit code value from ping which returns a value for a failed ping +That means I have to save copies of all my credentials in my Template subclass and then my in the constructor of my apis Template after calling super with those same credentials and I have to replace the interceptor list with a new list containing my signer instead of the header one +Will fix for +The first and most serious issue is that the JDBC driver is included +Ah I think Ryan must have got itYeah I think its fine to remove in both +that was hours of my life. +really agree! this is most needed +Wasnt able to pull artifact from maven repository +Someone should update the section on Category B doesnt say anything about where a Category B work may be placed +Right but the topic already exists at the time of failure so nothing is going to trigger +I propose that all thats needed to close this issue is a modularization that has one module only in it and that module is +Please post which you used thanks +This is my first crack at the mentioned above +As Im unable to reproduce +I would remove this from the distribution all together +This works fine for me +It will take a few days of testing but for we can try to upgrade the container which will bring people in line with but will require much testing +Eli the serves as documentation of config parameters +was already released with +patch looks goodThe test failure is a known issue see +From the ContributorCommitter Best Practices things that already exist must be understood before they are changed +If you wish to create a new collectionshard wo enough nodes we can just use something like forceCreatetrue so that the collection and shards are created empty inactive +Thanks for adding the componentSorry I was not clear from the branch and yes we are only talking about bug fixes +Ill make those changes and resubmit +WIP patch for this which were already using successfully in UbuntuIssues I know about so far The Unable to destroy invalid instance warning is triggered on shutdown but this seems to be mostly inert +Jeff can this be resolved have confirmed that the paths are all corrected in including my comment +BensonAny idea what needs to be done for this +adding the jboss repos did help +Might be something for the scratchpad samples? Didnt look deep inside though +Only one concern and that is the assertion of segments +Maybe I didnt understand your use case? Or would you propose that we simply move the new method from the to the? +We are now pretty far implementing this +Deferred beyond beyond beyond beyond Release +Can i do this +Just replacing setParameter to addParameter seems to do the made a unit test for this case +The still failing tests is also bound to the fact that ASM doesnt work with Java +Hi SergeyI apologize for attaching a NB project +More CR feedback +All ledgers will be marked for checking by the auditor but theyre still available after the bookie comes back up +and dont forget the jaxb flag so the server can actually start without trickery +When I edit I have no code completion +The only part I dont like is using the old class filters hidden classes +those keen to work with it will come back to the mailing list anyway +Where can one check whether or not the sender ever receives its own messages? +Sample Scala program with included class + +I dont know if I can give you a copy of the project +All tests pass with this patch. +r This is code complete +I see this out of heap space on linux running without with todays trunk codelineI tried these combinations from today from svn that fails with the out of memory error from svn from todaythat passesSo it would seem it something in the setup that has changedI always seem to see the failure in the test fixture of ran on linux on svn and and saw no memory leak or difference in behaviour +In fact the session is still there because the application works almost fine I can use the app successfully but observe occasional HTTP that result in some images or scripts not being retrievedAs I mentioned earlier the request that failed when replayed using wget or the browser itself succeeds +Checked release notes +All commons logging and logj classes are already included in So of are already in the classpathps I dont like too much dependencies either +A full test for streaming to agent to collector to demux and load to database then verify for overflow result +Itll save me time tracking it down to fix the error message +Create a new Roo project +You are free to choose +cleaners should not have much impact since the snapshot caches are empty and therere no links if snapshots are not usedWhat about the thread pool and the zk procedure coordinator created by the for the online +My guess is that you found a bug that exists in but has been fixed since I dont remember which one but there have been many fixed in various ll see if I can find the specific one link to it +Wont fix issue closed. +Ive attached a patch which should fix the size problem +Frankly I didnt check tracing on bits before I applied this patch +Thanks. +It depends to version +review comments from Devaraj +is created for the problem with GC crash +issue watchers may I bring your attention to this issue? Itd be great if you can tell whether the patch is acceptable or can otherwise provide another suggested approach as at the project we need to find a resolution for thisCheers I have no time to dedicate to this one +Yes ill give it a shot this week +However in this case since the cache flusher held the lock and was blocked this in turn blocked the region server clean up code and the region server hungThe solution for this problem is to catch run time exceptions in the scanner constructor and ensure that the read lock is released +again +The problem is likely the doesnt specify that Groovy must be installed readme should not contain that because it should not be is more likely that the location of the groovy jar has changed and we have not modified the to reflect this +This attachment does +I can provide a patch for scan too +Ive just committed this +Applied in rev to trunk +It would be easier if we dont support escapesLohit HDFS does not permit colons in file names while Unix does +Nigel can you the value of ? Thanks +You could compare it with the adding generated to the correct project resources +When running the test case I see the being called upon deployment of the and this generates the policies that apply to the ejb layer +yes the api in the patch is new in +patch ready for review andor commit +As we already have discussed that issue in length as it is on our Macro wishlist there is no need to keep the report open +The changes at this point are only in codeline +Patch adds a test for and fixes both the last doc problem and the thread leak. +Alfonso when you happen to have a chance please retest your application usingthe newest and HEAD shapshots +Thanks Christian +IMO +uff you know the minimal info for poms dont you +Hi Hairong glad its working well for you +The comments in the original test seem to indicate that the intention was to test certain properties of the predicates +Attaching minor patch that will update the test case +Committed to trunk +samplesquickstartshelloworldsqlactionlib hibernateaction and sqlaction QSs have been changed to use hsqldb and the mysql driver has been removed from both +Outputstream is not +This means updating the publish options and docs +out of date. +This will also prevent retries of any errors that have not been retried yet when the exception occurs +Improve string value sorting with new class +this patch is faild on my Windows machine +Patch for changes in and the crawl script +User specifies security configuration with name default which overrides the default security configuration in the runtime +In the file this is done by editing the list value injected into the xtsInitialisations property of the +Patch for trunk +Some of the AS stuff Ive seen doesnt use the interface so I think this probably isnt needed any longer but I may be missing some important detail that Paul knows +From my observation currently All keywords are All names are +and vpn is works perfectly for me to tweaked application is now built as part of the same reactor than the product and was removed from repo +Not a real issue. +Thanks Vikas +the classifiers are parts of the main saxon Ant build +Im not positive of the benefits of this but thats what I could see in the code +Patch for trunk +Ill look into it +Changes to utilitiestrunkinstaller added the source from their Google Code SVN changed the licenses on the +Could you please upgrade both jempbox and fontbox to as done in +Sometimes were to focused on one aspect of a problem and dont see the obvious +Pete thanks for the note of an injected decorator +reopening need to better handle defaulting of decoder configs +The building process is unfortunately RAM intensive but there are settingsknobs in the FST Builder API to tradeoff RAM required during building vs how small the resulting FST is +Patch already purged Cell of any notion of CellsIve made a new Cells class +I guess I could do it earlier though if we think Camel is coming really soon +Committed revision Thanks for reporting this Michael +The attached patch resolves the issue without any changes needed to affects versions +the only way user deletes is manually editing the in that regard they delete much more than that make system render useless and we have no control over it +If we keep them Id advocate for bumping the default to and clarifying the error messages to be very clear on how to resolve the configuration issue +I clean out all my temporary files which allowed it to build last time but no go this time +genmodel files inside targetI just built on windows and noticed that the value of the path folders is correct but myserioulsy the s are missing from the path +Thanks Siying +It appears everything will be driven from the order item level and the order wont be completed until all lines are invoicedInvoicing is driven from Shipments so how will this work on drop ships? Are shipments automatically generated for drop ship items? Would it be possible to wait for shipment confirmation from the supplierWill payment capture remain the same +Please verify if possible. +Other than auto generated Java docs try to add more comments so that API provide more details +Created screen for leave list with name in file +Fixed +Since restarting Eclipse with the editors in the same state as they were is useful to users I hope some other fix can be found +or +New version of patch that fixes another namespace assumption +Still would like to dig into this more +I think I found the problem +But the code does not produce strange exceptions so at least in the Sun VM we can detect compressed Oops and guess the reference size better +However since the module is currently unsupported and the only module requireing these modules I would keep the version declarations inside the specific file itself rather than move it to global now +Without additional details given I have no other choice but to mark this bugreport as invalidOleg +The patch is already finished and ready to commit +Sanjay Konstantin Did I address all your questions? Can we move on with this patch +Hi WillemI think we need to put some more work into that split +to run after properly setup the environment executeant tested with but in test suite it is ok +The Return Item page contains a drop down to select the Return Type from here select Refund Immediately +This makes it impossible to label builds on the fly AND build from any directory without editing all your files +Im talking about log splitting +If you use a queue instead of a topic Im assuming the messages are only processed once +I have solved this problem myself +The RPM build process needs to be cleaned up +fixed refactored RTTI handling by sort of cloningadded a TL stack management for it +I can do it but I dont find it I see no corner test cases for these bugs that could discover something that couldnt be discovered with performed tests +The bridge provides the intermediate that translates between portlet and JSF semantics +I had and I accidentally switched the artifacts +may be the Hudson got busier +First patch +Not a major issue I settled on doing what I was after in another way but wanted to document the compatibility issue. +YARN can launch the zk process from the app manager and populate zkList to signal Giraph not to launch a ZK instance itself +I have commit the adaptation withbut this adaptation requiered to use instead of +Please look in history for what the actual resolution was. +Switched to has slfj to do with jcl? Nothing has changed +I am looking at buildingtesting this patch against the trunk would appreciate at least one other person to review +Now I need to resort to setting the to lazyfalse in order for the proxied pojos to be recognized as being fully initializedNot sure why this bug is not fixed low prioLeeds me to conclude that not all users are experiencing these issues when having query caching enabled lazy associationsCan anybody shed some light on the status of this bug? Have people successfully uesed second level caching query caching lazyIs there a way toRegardsFaizal too have been experiencing this issue when enabling query cache +My first idea would have been a duplicate xwork jar somewhere on the classpath but if thats not the case then it might be something more dark deep inside the OGNLFreemarker integrationI cannot promise anything but Ill try to have a look at another project that recently encountered a similarly strange error in a single Freemarker template after the upgrade to OGNL +Sorry for the delay +As far as I know everyone can listen and write to a dynamic queueCan you tell me how the security in servicemix works +moving to M +David Sorry and thank you for the move +Note ensure from a naming perspective the artifactId is specific to EAP such that users are not confused in choosing an appropriate adaptor +Thrift should have been the same +Deferred until and Portlet implementation unless theres a compelling reason to implement this feature ealier +Attached is a fix for this bug along with a test case +Otherwise issue made no sense +Resolved as suggested in revision +Please dont change the priority +added a test +I recommend using cascade constraints when tables are dropped +After I updated st issue locally run the build I saw error error reading optmrepoorgspringframework error in opening zip file +that shouldve been hadoop fs touchz f hadoop fs stat o f +Then where is the jar of TEST or FW which ivyretrive should copy into templib +There where no changes to Source code and Java is specified in Hudson Job configuration +If there arent associated regressions tonight I will try turning on explicit constructor inlining for the Java tomorrow +How about just having an unbounded queue but having each CF just keep a flag that says whether or not there is a calculation currently pending or executing it would be reset by the task when executed and CASed in the write path +Summary Panel shows empty page when built from code +The linked issue has been resolved verified and closed so does that make this JIRA outdated now? Thanks IsaacNo it does not make this JIRA outdatedLinked issue is about NPE and workspace crash + Bug has been marked as a duplicate of this bug +i will try this out with a profiler and post the results +fixed timeout by correctly closing transport connection which allows Cassandra shutdown hook to proceed without waiting for RPC connections to close +Marking closed. +failed +Ping +Reviewed your code nowThe phrases should be initialized to the sice of the inputsentence that will always be the max number of Span objectwhich the code below can create otherwise the internalarray must be resized if sentences is longer than default valueThe test for this method should contain a case where we have aBOB sequence to test that after an O everything still works correctly +Attach test case +David I tried to count the number of times currency is used through the entites and stopped counting at + +AS aaaah so you actually did not use jboss tools for your server adapters but the generic ganymede ones +Should be verified in gatein master and reopened if necessary +In my approach Ive tried to minimize amount of changes +Revisit issues that have jBPM releases can only contain bug fixes +I dont have the detail of which test this was but it shows the failover caused by the get call when the queue wasnt created +Checking out using tortoise svn or in linux works +C Java and Python serialize them in IDL order +stress on ec would be a reasonable smoke testwould definitely prefer something tunable via JMX even if not exposed to nodetool changes require a node restart to take effect +and committed. +yeah fair enough unfortunately I have been told to use the file system as my manager doesnt believe databases are good for data storage +Script writtenbin to build up the courage to remove the echos and run it +Added Testcase to this scenario as into master +Youre missing a bundle for version is incorrect its set at in the openws pomColmLooks good to meColm +Thanks for the patch Nicolas +proposed changes +Applied and to branch +I have just upgraded to Gradle and in the process regenerated that file +Closing as verified. +It is like only dependency I have is the size of the page and how many inputs is thereany conversation parameter did not help +Bruno could you please do this? Thanks. +Patch to make the instantiate external refrence resolvers using the object fixed in +wonder what happens retrying after an IOE +Also includes an avro test in +For example by default Debian uses dash for binshIf this is the cause of your problem you can fix it by simply running the program directly or by explicitly running it under bash with bash HADOOPINSTALLsbin or by changing your systems binsh back to bash with sudo dash and answer no when asked use dash as system default shell +Im sure Im not the only one confused +the patch looks good to me +Ive checked this feature +Hi JoseI agree with the idea of using can you test from a windows and mac client before committing +Reverting it to modifying the coordinator nodes state +Link to some previous instance of this issuep +The bare minimum would be the blueprint and console bundles as shown in the list above but without security there will be no ssh and no way to access the console +I am trying to externalize it into a configuration file defined in the TOMCATHOMEconf help would be appreciatedThanks +Reviewed merged +Great news! Thanks. +Just FYI looking at my own comments in bergerspec I think you were the victim of bad documentation in Programming Ruby nd Edition +I just committed this +for the update and the tip on how you can use params to drive the my scenario the GET was just an example +There is no silver bullet but verifying through the file should be avoided IMHO +some are the JDBC class when they have additional functionality compared to JSR +Assigning to Max for approval. + has superseded +Id be happy to provide more information +Please review and apply the patch +These metrics are only computed over the itemusers that are common to bothYou can write a test to but after looking at it again Im certain its correctIts probably the fastest implementation possible and because of that not the shortest +those tests are not part of it +I backported this fix to both the and branches +When removing a datanode reset its scheduled invalidate blocks +I confirm +Memory leaks related to handling is a bit complex to solve +Lets please remove this catch block and address the underlying problem +Thanks +Would be easier tracking failures and running reversals +Only the code who created the resource resolver has the factory and can therefore dispose the resource resolver +Angus do you have the same error when you use a hibernate configuration in your projectCould you please provide us the library names and versions in your project build pathAnd how do you configure the logger? With? Could you please provide the logger configuration file tooWe have some libraries inside our plugins which we are using to build the hibernate configutaion but also we load the libraries from your project class path +Let me atach the document here and you guys let me know what you think of this okay +Patch looks good Ill commit and Ill work on the conflicts +They could perhaps go in contribecbin +OK + +I will request this for once it has been reviewed +Sorry I didnt clean out these classes they are unneeded to run the sample client I can enable or disable consumingproducing by modifying the file confconfiguration +Any chance of this happening any time soon +Also cleaned up configs by using dependencies instead of references to assure start order and commented out the sample css and tss beansg +Works for meThank you Alexey. +Thanks Paul! +You just need to broadcast to an ensemble and wait for quorum responses as I outlined above forI think ZAB or a protocol with an equivalent set of guarantees is a necessary and sufficient tool for the job +Heres a new patch for trunk +JS +The console still works with unsigned jars inside it +Theyd simply need to make sure to properly enable caching +The first thing I did was get it in the local repository so I could use it in maven build environment I see others on the mail list do this tooWould it break any licensing for someone to put it in a public repository? Hive could for example put it up in their own and reference that in their pom +Namit are you concerned that inheriting the entire classpath of the parent process is not a good idea? I agree that its unnecessary and can modify the patch so that only the that were added by the user appear on the classpath of the child process +Committed to branch +Could it be that the issue is not resolve in the line? can you please also backport the fix tothanksAmir +You can not have two versions of the war in the deployment the EAP version by default the all profile has the development version of the WAR and the production profile has the production version. +Have you looked at the new requiredif validation? It may handle much of what youre talking about here +Please post to the user forum providing more information I assume youre using the native stack that comes with AS is that right? Are you using Metro on client side? Please on a new user forum thread and link it here +Next step is dcexampledccom +What I do now though is process the arguments twice +Type system and XMI file to reproduce the problem and a screenshot +Maybe instead of using regular expression one could utilize the class +realized it is same stack trace as other NPE +reopening issue to backport it to to fix some nightly diffs showing up against ibm jvms after recent backports +Patch added +Committed to trunk and +I would like to check in soon if all known issues are solved +Ive already begun developing a to see if that could be a solution +Regarding the comma +Done. +tested and can confirm this fix works +Please verify +works properly for me as well +Incorporated the fixes from and +option it is possible that the command line error message and the Run As Maven Build option could have been the same. +Could you please check if my fix is correct? Thanks +I have checked in your changes instead of leaving the +This was documented by Ellen Evans who does not seem to be in the JIRA system. +Available in resolved issues for already released versions +Closing this for now since it is configurable. +This is a bug with Sun JVM +Like the simple form example +This was caused by builds accidentally publishing incrementals +Editorial pass over all release notes prior to publication of. +Moved to SMX as its Container related +Looks good to meAny other comments +I will leave this issue open for a while in case someone wants to port this fix to older branches. +wrong with my patch generating +I anticipated that answer but it was quite helpful to read your clear and detailed answer +I test locally and yes there is no expiration when using +It seems that we have consensus that we can keep things as they are for now +Thanks +Please try again +I noticed while skimming the patch the the author tag is used in the Javadoc in places +Thats why I wanted you to look at the cert produced in the directory so I could modify this if that was happeningUnfortunately all I can do to get a cert out of a keystore is get it in whatever native form its Certificate implementation has +No reproducible with +Vijay it looks like we took a similar approach +I think the at least could benefit from having a method along the lines of is to denote whether or not it is safe to read the returned by getBody more than once +Recent discussion on mailing list didnt come to conclusion of removing offline merge +a blocker for Axis release +This issues did not appear for a while +Patch vector branch +Are there any test service for this +Hi Tedwhy not test for the return value of? perhaps if it couldnt write the entire chunk size we can raise an exception +to not requiring protoc for simple dev use cases +It also updates the to not use NA as a flag but instead go off of the applications state +If the port is missing treat the includeexclude entry as covering any Datanode on the host +Please check the patchAnd I have a question what do Line set isolation serializable mean in? Could it be removed? Thank you +rev ports rev to branches txmanager +patrick It seems you patch is removing the def of clover plugin from the pluginManagement section to add it again a few lines down +History files checked into trunk rev . +Thanks for the patch +In the mean time you can use the attached fixed jar +yes this only affects test on IBM as wont fix not a hibernate issue but the byteman used in the hibernate test +I only think we can add to the DSL if its part ofAlso I would like if it was possible to use a common prefix for the DSL methods so they are grouped togetherSomething like signatureSign signatureVerify +The Using the Context Menu page also need to be similarly updated +What are the exact commands you are runningAre you using or +This has serious performance problems when CLOB columns are involved +This issue is closed now +This is necessary for the upcoming Ambari release +Other formats are just for completeness nohing more +At first glance its not out of the realm of possibility +Therefore without this patch Sqoop will not execute from a release +Err +ThanksAs you said the API inside is a kind of over kill so could be like you mentioned earlier +STS plugin is Im going to close this one then +Meanwhile connection n sat idle until n timed out and then it too would receive the initial handshake message and so onSo it would seem that given this behavior it might be easy to instigate a attack against the broker simply by creating connections to the openwire transport and sending nothing on them + close and indicate where this was done +Patch throws on short read from the entry log in all other cases +I see the latest in the series is but I dont know how easy it is for you to upgrade +Many thanks +HUESHUTDOWNTIMEOUT does this mean the shut down process could take up to seconds +I solved the issue by adding dummy script elements to the head with a src attribute set to the script src via ajaxSuccess handling after jQuery initially loads each script + on the committed +note doc fix was backported to but actual message was only fixed in. +Ooops! it looks like I still had these files in my project since I manually implemented the solution that was proposed in the forumThe files are artifacts that were not generated by will test the relationships after deleting these files and reverting back to Long does seem that Roo has trouble with ids as per the comment above +Applyable inside of eclipse +Its almost years open and still Minor + For cert expiration warning set a day threshold +Fixed in trunk +However the exception was misleading +ARFF isnt supported you need to remove the header from the file first +When the connection to host fails we move to hostAt the same time that we support failover we need to be able to do round robin on both of them +If we dont do that it will cause a lot of IO in mappers and kind of defeats the purpose of MR imo +The readFields for Task should always set the task to either a new value or null +If it is trivial lets do that? If it is Ill look into a rebase +This flag creates the Helper classes which help keep things in the right order +sadly +Thanks! +You know I have absolutely no idea why I used Res when Res there +Can you please submit a review request for this patch on? Thanks +Hi thereDo you people actually have this working? Ive found that once you click on the Browse link all of the left hand menu links get rendered minus their +This test is failing when it is run as part of some other target +Also see duplicate cr +With this setup aggregate can be considered harmful +Changed issue title after some initial digging in the code +This should be fixed in see +This is an issue within Eclipse assigning to Max to fix +Please your application against the latest snapshot of trunkOlegPS Sorry it took a while +Yeah Im fine with a new parameter for this +I applied the patch in rev after fixing up the problems with trying to install plugins +Hi Christian With Compiler issuing Java bytecodeBeans Cross Reference has the same problem +Hi Konstantin I want your opinion on this +I have some prototype here +Thanks. +eh? what is the bug +Ive looked into Adrews patch and verified that it indeed fixes the problem +psr abi to the beginning of the file +Keep the statics around for single AM per JVM case which I believe will cover cases and add new This is what the patch does +Done thanks for cool thanks Lukasz +I just went through an removed all subsystems from the CLI and they all worked as expected +It should appear like the portlet doesnt exist at all the user doesnt know about the portlet on the pageIf we display some message like Access Forbidden then it may be confusing or irritating for the end user point of view +I have an immediate fear of trying to fake this since it opens the door to having multiple threads all appear to be the same objectWere going to have to dig into this more deeply with the help of the Rails folks since the implementation of Enumeratornext is not something we can change and making one thread pretend to be another worries the heck out of me +Ive replied in Wiki +After a quick look I was wondering if there is a way to the SDK already installed locally and configured in Eclipse +I attached the test file because JIRA removed all my entityrefs explanations +I noticed the size of svrs is different from that of svrs + +Maybe we could help users by adding the following factory methods to say andWhile this may help in some cases note that configuring systems using data sources often consists of specifying the data source class as a text stringAs such I think it is important that we document this as good as we can +has to provide a different solution +Ill take it for a test run and comment again with the test results as soon as I get some resources for a VM +Fixed some indentation issues +Sami it would prob be good for you to review that a bit Im not sure if we can do that in a cleaner way or not + +Ive asked mark for tests for +We happened to have the same requirement for debugging purposes +I believe it worksfor the seems to be falling down only for version ranges correct? I hadthought these should also operate as a transformation but that doesntsit with the whole resolution actual version selection should be assigning a single repository toit where the correct version was area of the code needs a lot more tests in the order of what thecollector itself has +Oops sorry about the confusion +Please take a look +These blockers and critical issues are resolved but not verified +Based on the comments in it might be related to your jdk version +Ratneh RatneshYour patch is in trunk rev +does not do this so you simply need to place the shared classes into a common jar and drop it in the lib directory +Blooper after blooper I hate SVN +Verified by Nathan +Review as part of EAP QE +Pending further details Im closing this issue as Cannot Reproduce +Raised to major as we are breaking the specs +Patch to switch the logger to take in the throwable not just the name of it +Ported and committed to branch +Running run Failures Errors Skipped Time elapsed seccannot reproduce with r + removed the from the kit its contents are added to the wgibson resolved +It would work with any query parserSorry I used the wrong term +patch looks good to me +Im not mystified +Patch looks very good +No reply in two months reopen if issues are still found with current build. +Does not include the content of the first patch +With this patch i have not attached any new test casePls take a look at it +ran overnight on node cluster and fared much better with the patch +A redirect is definitely a good idea because right now googling commons Lang javadocs results are s +Let me know if I should create a new issue about the aliasingThanks guess we need a separate Jira to deal with better error messages +I think if there is any we want try again +The subject of this issue s not correct +Thanks for testing. +We need to beef up the testing of the behavior +Thanks for all of the time you spent looking into this +Are you with me? Think about merging comming from test environement to production environmentThe solution ive found to avoid this pb is to hit the identifier generator when inserting new data until the next identifier would never have the same value than ones inserted previously +the naked and unpacked bits can wait +Im probably thinking too much +Not sure whether this still applies in can you comment? +reworked for current trunk. +Hello Julien as the password is a simple maven property we do not have control of whether it gets displayed or not +better to do days +yes the patches should always be relative to trunk +Aspi I believe we fulfilled your request along the way as we worked closely +dup of +Then we could usefully compare server time to data timestamps and create a tree of data inserted before repair beginning +Create a new version like or you have plans to do this for the EAP series please clone this JIRA as the EAP and EAP cycles are different we cannot have both in Fix EAP is history so please assign to the clone either EAP or EAP clone and have one for each of EAP and EAP or we wont be able to triage it into the correct releases +making the documentation task +Seems like its +However I am still not clear as to how these groups are tied to the authentication system +discriminator is not gonna help your case +This means that what I have in mind does not have anything to do with your original intenetion of this issue and does require a completly new approach how and where to implement this +Hi Bobbybq +Updated to latest looks good to me +This issue is the follow up to +Sending karafadmincoresrcmainjavaorgapachefelixkarafadminSending karafadmincoresrcmainjavaorgapachefelixkarafadmininternalSending karafadmincoresrcmainjavaorgapachefelixkarafadmininternalSending karafadminmanagementsrcmainjavaorgapachefelixkarafadminmanagementSending karafadminmanagementsrcmainjavaorgapachefelixkarafadminmanagementcodecSending karafadminmanagementsrctestjavaorgapachefelixkarafadminmanagementcodecSending karafadminmanagementsrctestjavaorgapachefelixkarafadminmanagementinternalSending karafmainsrcmainjavaorgapachefelixkarafmainTransmitting file dataCommitted revision . +Note that the lines for dont match whats in cvs as Imusing the +If the problem recurs I will update this +has a chance of breaking some behavior given our reconfiguration historyOne I can think of and came up recently from Koji on the ML is the MR child opts vs +change looks good to me +Yes try this combinationfalse +Removing the functionality entirely would severely break backward compatibility. +Verified on revision . +Yes the rest of the cluster was free +Done +Making the policy pluggable should be sufficent +if OSWindowsNT SET MHOMEdp +Sounds good +Original opsmin Patched with MAXINLINEGROWTHFACTORPROF opsmin Patched MAXINLINEGROWTHFACTORPROF opsmin +There is no facility to avoid this until jboss . +I am sure all for it +Hey thanks for the comments If we do this how will one fetch reports of all the tasks that are either completed or pending? You require to call it twice and then do the processing +Fixed in revision in trunk +It looks like I committed a patch and then backed it out and then there was another one posted that hasnt been committed +Though this could be done using cucumber lot of shops like us have invested heavily on the Geb and Spock combination and all of the team is used to the tooling and frameworks +I still believe this patch is sufficient for the fix in branch +Therefore it would be really nice to have both operation modes +Lowered priority to Minor since this is unlikely to affect many users +Deleting the configuration helps but then I cannot disable or uninstall any plugins anymore +Whoops that should read. +Im happy to include it as a command +I think as Pig becomes more mature we will start collecting and storing needed metadata such as data sizes column cordinality sortpartition order of the data etcTrying to dynamically compute the information if it is not available sounds like a good idea +OK sounds good will do +This may be fatal when running on CI server + +Hehe Ryan I feel like such an newbie +ThanksSorry for the deplay in closing for some reason I didnt get notification. + + in +I cannot think of a reasonable scenario to enforce case sensitivity of table and column names in the real world +committed with a slight modification to allow the unit test to pass in a different partitioner than the global one +You are right Nicholas +You can find the VS build script in the +Thanks +but then again if its a param it can be specified as an invariant if you want to ensure that doesnt happeni guess hte key thing is just that the has final say over what Processor gets used +I thought this was already done but I checked and it wasnt +You need to attach a sample project to reproduce your setup +AntonioThe same file can be used for cocoon and +The D option is a first option to populate the layer by default +Start of documentation section is now to move to as was issues after move to in +I just dont know how to set the Eclipse classpath +I didnt get any further information yet +From that we will create subtasks to create the new structure move the existing help add new help etcFor now I will definitively close this issue and if we go the new way I will later close and all related tasks and subtasksLooking forward for a continued effort +Backwards compatibility can be achieved by supplying a if one is not provided +However it is not reproducible on RHEL +Fixed on the new Quartz archetype. +The first ID is getting clobbered with a new one when the first ID was intended to remainNo support for multiple IDs for this semantic +Mentioned artifact is not being created anymore. +Not too worried about doing this inFor yes this is the issue I asked you about +Could you provide some easy steps to reproduce the problem? From looking briefly at the Grails JIRA issues mentioned I didnt really see much there to help us reproduce the problemObviously ideally steps that are independant of Grails as otherwise we cant really add a valid unit test in the Groovy test suites to check we dont have regressions moving forwardAlso if the related issues are separate please create two different issues in JIRA not just one +It should do it during first looked in the seam example and there are indeed no gpd files so this is expected behaviour +Thats already done no +Have you thought about how to handle multiqueries? I think this ties back to our conversation about EOP and your comment that in MR EOP means both end of key and real end of processing +Didnt change the parameter name +qmailsomethingorotherxargs perl pi e susrbinspamwatchor change pi to if you want backup perl commandperl pi e susrbinspamwatchseems to be using the character for too many thingsThe line should probably beperl pi e susrbinspamwatchor possiblyperl pi e susrbinspamwatchgif the string can occur more than once in a line. +moved to +The only idea behind putting the condition into the was to keep the registration page as simple as possible because it should be editable by users +Reopening to check attached application +Can we please wrap the map in a to allow garbage collectionThere is an Override in the source code and not inside a comment +I believe this was fixed in could you verify? This fix was applied to and trunk +I fast to fast againYou could also add the patch here as it addresses the same issue +This one have better error messages +it is enough that inside the services the service is declared as in release +launch files works in eclipse +One way this can be done is to new config files with rightimproved parameters and let the update themselves +Thanks for your comments +Committed revision +The method also batches the reads to help optimize access if there is more than one segment in the relative patch catches the that gets thrown if more than one segment in the relative path doesnt exist and attempts to read the relative path one node at a time +Great! Thanks for the fix +System updated the POM +While adding document to index we need to eliminate family if value is nullCould you please help me in creating sub tasks and in finding the places where it could cause a problemRegardsGagan +I will start work on it this afternoon +Hi ClausI am extending the abstract class and add a class in the camel core which I intend to use in the Scheduled codeCheersAshwin +That will choke things sure your classpath is clear from any jpa persistence lib other than the one comes with beta +Wrong project +Do you need the base configuration then configured as a beans config in your IDE +After I have done this change I was not able to reproduce this issue any more +What I meant here is to fill the resource bundle with the messages in OM with support from commons module +Could you let us know what version of SDO you are running please? Perhaps if the issue turns out to be not already fixed you could also attach a schema demonstrating the issue. +I also see this issue on Windows Server +The correpsonding orm test case will be checked in once the orm schema is updated. +Unfortunately I will have to make hte interceptor chain factory accessible because of see that it is related to the microcontainer but what exactly does it do +I just ran mvn site in the project +Hi Matthias As we discussed the filter is not the user if he is already authenticated +Note that I am familiar with these properties and used them in projects in my last company +I will be picked up by when I merge trunk to that branchThank you Chris +Another basic question we are on version is this jar also backward compatible or do we need to upgrade to +What kind of results or terminal output should I expectBill the terminal output should almost be similar to what you see with HadoopPlease create separate tickets or use mailing list +Correct me if Im wrongbq +CLA received ac requested +Can you give some examples of SFA application in market that follow the workflow you are suggestingHave you evaluated Capturing tracking code linked with campaign +However there is no to run +I also get the same error as whats in the description using Struts and had the same errors but I found the solution +This information will be part of Security Guide so no mention in Administration and Configuration Guide. +Maybe we also need to support node as the basic node typeWe are still working on it +Submitted for this issue +However it makes a deep copy of data structures which probably adversely affects performance +This is seen for Streaming applications +The find bugs warning does not come from this patch +I have attached patches for the trunk and branchRob could you please review these patches +Thanks thats sorted itHowever the hang still occurs if one uses explicit mode the new example uses implicit modeWith no I get a no availableI need to look again at the issue +The configuration is purely so at this stage the console can only display itOnce is fixed modification can be enabled +One minor comment is that we need to heavily document the fact that terminate is +It would be good to have this patch in the Hadoop codebase +What project owns this and what are the group ids? see for the info turned out to have been dealt with by someone a long time ago +This is most likely a duplicate of +I have tested the scenario in the server built from branches +Youre welcomeIm happy to help you outClosing the issue +Messages which are buffered for later delivery need to be associated with the correct consumer +See screenshotsIn the Edit menu and in the context menu for the visual editor the shortcuts are correct CommnadX and CommnadV +This change filters them outNihil Obstat +Thats a good question +If thats the case we should probably file a bug +Derived marker is set on Maven project import and when Maven Update project configuration action is with and this didnt seem to work +Whats level of effort to this to? Are there other which this patch is dependent on? +I think this assumes that the only possible exception is going to be NOCURRENTCONNECTION +the netbeans integration doesnt require maven to be installed on your machine its using maven embedder +I just have a trivial question will it be merged to as wellThe initial plan is to merge into trunk will come later +the directmapoutputcollector takes away sorting but it also takes away partitioning +I just committed this. +I dont think thats possible to do without doing a get first as we do not want to override current status on URL +Investigation revealed that the Namenode CPU usage was caused by lack of heap space +RN marked as Not Required because this is a component upgrade +If you can provide a patch well consider it +For some reason I never noticed that there was an entirely separate API doc for components +Would make a good addition both the tests and the associated distributions +PeterYou must run mvn in directory +Thanks Virag for the review +In this case the solution is just move the code that check for collectionType to the outer block +I have written a small piece of Java code to check the LDAP error codes +No +and I dont see many other options right now. +Attach a patch +This patch resolves both and +but still strange +I managed to trigger it with as well +Patch to correctly generate etchbaseconfregionserver file +changed the code to always write in when doing automatic updates to or files +See for discussion on this issue +I found this solution solve the following issuesUnload Derby classes after engine shutdownPreserve behavior and support any future loading of connectionsDisadvantage It does not preserve the autoloading idiomI am proposing Derby allows support any future loading of connectionsI did not testing the impact of performance for this patch +Ive slotted it just afterJAAS Authentication +If I can reproduce an OOM via this mechanism I will try the CLI +In only when I try to run a groovy script by only typing its name a cmd window is opened for the output +I will document the issue in the user guide and hopefully someone will eventually come up with a script +thanks. +Engineering QE Docs all set to the requested directory structure + This bug has been marked as a duplicate of +Applied patch and added Mattmanns test to +yes kill will skip shutdown hook but I think this is ok + looks good to of +Tapestry should probably treat boolean and Boolean slightly differently +punt to peter for jira +Groovy complaints on second JAR in directorygroovy cp homenovojPrototypesgroovyScriptslib cznovojinfrastructuregroovy startup failedhomenovojPrototypesgroovyScriptslib unexpected char x line column +This will be solved with separate refresh threadIn essence this issue is similar to +rev new version of based on Rafis and Alans comments +I will put this info to in the new jira +Any comments regarding my patch +Maybe a better and more elegant solution is needed. +Thanks Gopal +The last patchbuild with cvs diff command and tested with cygwins patch +just uploaded +Committed to the two branches and trunk +One proposed solution is for spellchecker to record the index version when doing a rebuild and use this to determine if index has changedThis can probably be utilized to avoid unnecessary build on empty commit as well +Mobile platforms have support for this script now +The compiler has to emit DCMPG and DCMPL based on what sort of comparison is being performed +In our org we have standard packages of tomcatapache which has standard settings for all these +Customers reply is this enough information? If not what would you like me to ask forThis is a web service +I could not see any indentation issues in +Closing all issues that were resolved for. +This issue doesnt exist in current Axis trunk +My suggestion would be in a but please open a new policy JIRA and ping the legal affairs team +Patch for the test +This patch has been tested +Heres a patch for the branch including aIve introduced a property called which should be set to yarn to use YARN and if unset will fallback to the existing behaviour +btw this was not just with jdk +btw you can create the same behaviour that emacs does by doing the symlink yourself +Can you add tests for redirection of urls? You can look at it already has test for the patch until testcases are added +To add a little more clarity using your browser to issue a GET onhttplocalhostmypathmystuffxyabcmeans the ESB does not see xyabc which should be parsed out as easy to read properties inside the ESB message however it would appear that JBR does not pass those through to the ESB code +Just to correctFirst we have to decide if the pointed variable has its factory method in the same classcomponent not just the whole Factory component can exist in a superclass though +fix committed thanks for the clear test case +it looks fine aside from the things I mentioned +Thanks +Is it to be provided in another patch +Fix is in +I have sources for this +Hi MilindIf a user wrote I guess heshe does not suppose to put entries in the command line right +Changed the visibility of as per your suggestion +This should be committed to as well since its just a test bug and causes test failure on some systems +I put the attachment at the web site because it was slightly too large for bugzilla +I think there is consensus that Hadoop should delete the mapred system directory and that HOD should name directories to avoid clashes +Now if an event comes in which has header value which does not map to any of the configured mappings then the event is routed to default channels +Pushed to +Are you looking at this bug? Hope not since I just accepted this bug and had a fix in mind +Is this something that needs to get added to the doc +When you add a parameter to page verify that url changes +The current spec says to do this I must put into every bundle aThere could be two solutions the first one which just says which configuration in which wins or an alternative way in specifying entities +It should at least provide a better error message to help diagnose where the problem is. +I see clear API in which assigns a new obj to events object but dont see getting called anywhere +Done +But other commands like less more open edit should also be reviewed +If you need this feature for your company the first thing to do would be to test the attached patch add configuration param for enablingdisabling add tests and work step by step towards a mature patch. +I am not sure why Michael noticed features being transparent prior to version +I think changing them to scale and shape would indeed be a good idea +Fixed in revision by adding a front page check for the availability of the interfaceIf the interface is not found then a special error page that describes the problem is displayed +Im not working with Nutch at the moment so I cant provide very precise information about what percentage of time generate and update consume compared to the fetch step but I do recall very very long generate times from my previous work with Nutch and I remember others reporting the sameNot sure what the best thing to do is just wanted to voice that what true provides is nothing like what Rod was working towards so there is no confusionRod are you still working with Nutch? Any chance you can give us a more updated take on this or maybe even a newer version of the patch with your changes +copy to severalllib +If youd be willing to try this out Id be grateful +Hence the problem +Probably a bug +Request to move it as it looks more like an issue to me +JihoonCould you commit this patch +It would allow users writing tools on top of Hive to depend on just the Hive classes freeing them to construct their classpath as appropriate for their site. +Grant the latest commit to trunk is much better but we are still missing Also now that we have upgraded to Lucene there are bunch of places still referring to that now show up as deprecated those would need to be modified too +Added an itest for spring trunk all resolved tickets from or older +Use a secured webdav server already maintained and actively in use by you prefer when I have time Ill make my own project file and sample configuration file pointing to an external webdav server that Ill setup +Ill start by making this patch to apply cleanly to trunk +contains a buffered input stream align operations on chunk boundaries and verify checksumsSince this jira is not a regression I would like to submit a patch without the buffering optimization and then address the optimization in the next release +Hello AlexeiI have made some changes against the test method in the new patach +See for more details +Hi Bernd please send first any support request to the Sonar Community list before creating any JIRA ticket +Ive voted for issue +I see that the script is in place but it didnt help +What about being able to inject a reference to the endpoint class instead of relying on a not so clean cast to? It would have the same effect but I suppose its a bit more generic +Perhaps this should be changed to a feature request to add defaultEncoding settings to the +Ive committed the header file patches +It wouldnt make sense to have multiple executions of this plugin yet but ifwhen it does the would need to be distinguished too +Patch with webgraph commands +Committed branch and trunk. +Merged and. +I modified the and it works now +The uses of in AIOCallbackhandlerhandlediskfailureshould be incriocoreaio has statistics global variables updatedwith normal increment then copied into the global stats +For more complicated information the handler can implement an additional interface which will allow the handler to control what it wants to saverestore when the is savedrestored +I thought goeserver creates a new reader for each request but that is not we have a number greater I think we can think about closing this issue +Committed +Unfortunately we dont really have a good jar to put this in. +Running tests again +I discovered this bug when building a simple search form that is bookmarkable +Out of scope for HB +fix files if they are available +As to differences covers putget methodsFailing scenario uses putget +We could submit a patch to this project to allow configuration through some other mechanism I think he hardcoded the config file +This should fix this issue +the keyinfo object has a object and aOk I am closing this as not a bug +I realize that I am on my own here and provided the details and comments for the benefit of other prospective and current users as I see I will receive no help here +This is how you would create a custom +fixed it +I do not understand why WFS CITE tests passed for +Once I commit the changes I will be closing this issue as it doesnt seem to be an error in the but in Flume itself +It does not look like this issue responsible for the findbugs warning +Hi David for the prompt reply +I ran the application from within IDEA +special character sets must to start with + +None of them use any methods specific to so its a pure signature changeSome of the constructors did not have any callers +The lib isnt backwards compatible wrt to old generated code +Hi GaryIm not sure why they got dropped again +Ive attached a patch for +Another likely candidate is the rmi binding but this does assume that its a java interface so I dont know what happens if you specify a WSDL interface on the reference or serviceOn the face of it the patch looks OK so I say we apply it as a starting point and look at other bindings to see where the same issue might exist +It doesnt need a because its simply a set of extensions +Dedup the locks requested preferring exclusive locks over shared locks + +The test will now hit the STS and get the SAML token +this patch doesnt address encryption issue when the data base is encrypted +Ive only done some initial tests with and it looks good +Andrea thanks for your answerI downloaded the installed and configured it following your recommendation +Is anybody still watching this thread? +I chatted to Simone on IRC and he did something with respect to this bug we will need to verify before marking this as closed +Known and accessible parts of messages invocation such as args return etc +My reason for this opinion is that in this particular use case the RM will have a new version of the configs and you usually want the configs and the configs to matchOther than that I am happy with the patch +Btw Ive got one thing I am not happy with and that is naming of the old andnew JDO classes +for maven projects it gets triggered when a properties file or file is modified +Between Ritikas email on the dev list and Mikes Sybase research it sounds like this feature is supported by enough databases that I think its useful to expose as an API rather than just a hint +Committed +Test classes will be always compiled +I will exclude them from v to reduce the noise in the patch +Patched in Working on the version +opencloseopenclose +The upgrade should also change the s DTD declaration in and projects too +I added some warnings +Note that the SWF code excerpts above are +Did you retest with a later client or on a later version entirely + +Resolving this as not required. +I just committed this. +Reopened because some files still refer to old Tiles configuration DTD +Entenc que travesty t el mats de mala cpia mala imitaci +TomThis issue is scheduled for this means that itll be fixed in that version in your proposed way or another +I checked all these queries do work correctly in Lucene trunkThere were several related query parsing fixes recently that might have fixed this +I guess we have to reopen it firstAnyway Basic idea is to have an entry like thisdn cnTori Amosousystemsn Amoscn Tori Amos objectclass personobjectclass topThen change the RDN to snAmos and keep the old one +Had to revert this because causes problems with +Marking patch available +We do a check for the ACPROGSED symbol in and invoke it if its available +More or less for certain cases + +Interesting fact about integration test is that by default when CM kills RS socket is closed to ZK so ZK session and master +So +I just committed this +QA verified +Your suggestion to remove the descriptive outputconnection label comes under that heading +I wanted to make sure I had the functionality down and wasnt missing something from the RFC +Source. +My changes are showing this up by actually reporting the exception +These files are recognized as facelet libraries by but they are really arent displayed in Web Project view +Actually I think it might have to do with Infinispan +Fixed broken merge post +Now I just have read source code and there exists somethink not clear yetIt seems to be needed to trace execution of program to know the fact more clearlyI will continue to survey and report the result +with in a new CAS format or as part of an extension to XMI +I have the beginnings of a working fix which involves deferring session writes until the end of a chain of forwarded requests +Same for german umlauts with. +I just called the method witrh a block id that does not exist +It required lot of changes because now I declared preparedStatement as a method attribute inside executeStatement and then I had to passretrieve this preparedStatement to all the method that are interacting with the the patch you attached does again contain a bunch of tabs and some method parameters are not declared final +bulk defer of featurestaskswishes from to +we already had a snapshot repo defined so your patch broke +Daniel feel free to fix any inconsistencies in documentation you found. +Cant think of anything better +That is what I thought. +Memcached is something weve been talking about internally +Tested with revision +Committed to trunk +using a script or from a double click +Fix in by +Includes some new metadata keys for this and updates the Mbox parser to generate these keys too +The proposed solution works +Closing all issues created in and before which are marked resolved +Thanks for the review +Verified in. +Updated patch that does solution a above +Files were not in patch format +Setting the protection domain when defining the field extractor class +Once reloaded the fields used by the sort functions can no longer be foundRestarting tomcat fixes the problem but isnt an option on our production servers +We probably should have fixed that for Maven. +Patch looks okay to me +Sure thanks for finding this +It will be most useful as part of. +New patch includes fixing nested maparraystruct +Yeah thats where things get weird where they shouldnt be +There is no morphological processing or any other functionality in this patch +Yes we should because this pops up with every release +Tutorials might best go in the forrest docs no +JonathanFeel free to create and assign them to Jesper for any JCA work you need done as part of this +could you give an example of what fails +Wed need a test case that reproduces the issue to show that its not just a limitation of the NMS client at play +I think we can close this issue as Not A Bug +Yep same issue here in version +But I think you are right that if we arent going to skip we can fail immediately +The regions are excluded from validation as much as it possible +Attach pls the image what you used +In and store references am trying some local changes to see if we can get rid of them +Resolved for JBDS. +If this returned value is disputed Id like to see each party document it withreferences as to where why and what they are basing thier results on +However I see regularly failing in my env not sure whats up with that +Removed monohtml references from subversion revision +I guess I dont get it +JavierThis patch applies only Regardsilya +Ouch that was an accident when refactoring the AOP check into an optional dependency +test and test fails for me +I dont think storing UDF jars in metastore is necessary +Failure repros for me +But the other item will remain within tabs. +within the cocoon pipeline null pointer exceptionNo Cocoon issue +Its unlikely a cleanup issue since the table is created during deployment and dropped during undeployment +Ill improve the code in a future patch +batch transition resolvedfixed bugs to closedfixed +This has been completed. +See and. +The timestamp is of time when task tracker receives a state change update + +Please review it +this should do it +If you find further issues please create a new jira and link it to this one. +I believe the file to be patched is project +Run ant you want to close this issue later I can create a new one when Im ready to submit the patch +Uploading patch with the above mentioned corrections +Thanks the patch Elliot +That might at least give us some additional confidence that this wont introduce any regressions +I like it +Well need to return a struct with both and though since has other pieces of info such as etc +Brian assigning this to you to see if there are any changes required in classes for jboss messaging support +That way I can get out of the business of bundling them into my appm looking at fixing this right now +For example a DFS directory could be used per job with a message per file named with a timestamp +What it the? What has it to do with your converter handling? How do you call the instance? What is the outputPlease provide more information or better a small unit test +Added since to the class and the methods +Hi! Will there be any progress on this issue? At least you can change the state I attached the tests already! Thanks +The entity reference report feature Im trying to add to the Webtools component ran for an hour then I finally gave up and shut it down +Screenshot of the icons in use +I had not considered some of the things Uwe mentioned +My intention is to include this document in the doc folder of Thrift repositoryNote that the protocol described is not yet implemented by the previous patch +Thats a separate issue +Please make sure you are using Spring and supply a test case which replicates the issue +Backup master becomes active successfully +I am not sure but I think the original duplicate row on create problem fixed with +The attachment has a test which passes in wicket but fails in QUICKSTART IT FAILS IN RC TOOMartin +Gregorys patch from to branch +rambq +There might be other causes for this problem but Ive fixed the problem for the cases which I was able to replicate. +Must make sure that results page is updated too need to properly show the zipsjars +Fixed in rThanks Dilshan for the patch +The domino effect of channels reaching capacity is not really a problem +GreetingsI really apologize but how can I configure jexl to use scale during divison of two bigdecimalsThanks a lot +I have most of this implemented ping me if you want to see it before I have a chance to commit it +The advice about best practices for choosing a good name might admittedly be overkill +that would make it more and hopefully easier to maintainIn the meantime should we disable the fault injection tests in HDFS +For example if I use a driver like it wouldnt know how to deserialize a given column +It does not take into account the possibility that the store was not located in the classpath and that the passed store name was not converted into a urlThis patch fixes this issue +IMHO we should let IT know about these problems with the internal instance +Whether you understand it or not the people who requested this feature have a use case +Patch checked inOleg +Did you try that +This is an Eclipse project with test that fails on this issueIn order to run the tests +When i try this PHP succesfully returns but the +Unfortunately we are stuck to M and I would really appreciate another compatible version +Will upload a patch soon +Patch attached +on my machine gave a +Hi Irina Sorry for asking Just curious does assigned to yourself mean this project abandoned from Gsoc +Issues dealing with bundling your own JSF impl are now Open to Community and unassigned +Thank you. +Its complicated to implement but it will free you from relying on the old node naming scheme of etc +It looks like either one or both of these didnt run before the package attempted to install +this column can be referred with the columnName or in the above clauseswhere clause can only contain and is different from above rulesI will upload modified patch with tests for all the above conditions +Attaching test projects +It keeps trying to get the new leader of the partition and add the partition to the corresponding fetcher thread +fixed with revision +Hence we end up with the topic being bound twice +When i save a xml element entity in domjsessinit doesnt save child pojosession works well + +It is possible for users now to add new files to the tablepartition directory without going through Hive to load files +But since the investigation has low priority I wouldnt finish the investigation until I finish analyzing the new test run +With the completion of this should be fixed +only concern would be breaking older compilers if we can document a starting version and above for a supported compiler im good with would say gcc is the starting point older ones should work as well +It works fine +If you dont think implementing this in the timeframe is realistic feel free to move it to +Just adding a wait method that we call w region before proceeding +Thanks MridulYour patch is in trunk rev +you are referring to I think this is useful lets see where it goes +But if community support dies out then drop it from the next release +the changes to trunk At revision +can we mark this as wont fix? Even since Chris comments above things have progressed quire dramatically +That should not affect the CSS at all but know that certain aspects of the default templates will changeIn particular you should assume that templatesdefault will be the template that everything in will ultimately inherit its looks from +If you feel those are good this issue can be closed +prefix on the transport server uriThis replace the initial patch that did not include this fix +Any chance this can be included in a release? Do you need a patch for this +I assigned myself for the patches till Apr where I reolved this issue +I get the thing too the prompt comes back before the result has been processed +Thank you very much for the patch +This looks good now +What JDK are you using what is the command lineBelow is a copy of the batch file that I used +Dupe oops. +That way we minimize the danger of causing the TT to fail +This has the benefit that we at least know which exact version we have without trashing other project or getting trashed by customer dependencies +Resolved as best as I cannot edit the console though so adding a line to event log is the best I can do +I have committed this. +Carl Edward answered it correctly +This is verily a delightful opportunity for meI have started working on the catalogs as you suggested and will submit a proposal for when the accepting of proposals commences +I think it should be part of the Socket constructor +Thanks it works by adding duplicate definition of our Archiva instance with legacy layout I think behavior is quite strange IMHO I would have expected that specifying the remote repository as legacy would be enough as it is the proxied repository that is in legacy mode not the Archiva instance +Heres an idea why dont you do what you suggested above zip up the files and add them to this issueI still dont understand your reaction as writing your comment will cost you the same amount of time to setup the above assuming you are a PMD plugin expertGood luck. +Testcase added to testsfunctionalorgjbosscacheloader and passes. +As the server is in fact a server I do not see anything logged to the console +Should make it so can change settings too at least when going via jmx +This is closer to what I had in mind +In the particular case that caused me to look at this the deprecated API wrappers were having issues with this reporter +I do not understand the whole problem +I think it is ok to change it on trunk and let the other languages catch up +Updated the page to add those two Windows versions +Undid the jenkins change +ok I see that now thanks +This will be addressed implicitly when we the docs with . +It makes sense for me +Some clients are configured to always reply in the same format as the original mail so that might explain why it works when youre responding to emails +We already have broken code in the task that tries to only send counters occasionally that should be removed. +and Im not sure how to deal with the issue comes with capabilities request parameter +If a comparator is not defined forces the class to be initialized by the class loader +Would you mind uploading a version of w patch included Alexey to make it easier for others using the script +Sren +This raises the question should directly inherit from +Thanks +If the case passes against DB with JCC then I suspect a network server issue +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +changed subject of the for input +bq +Greatly improved cheersI can confirm point a you mentionOne suggestion I would still make to improve the intuitiveness of the UIThe preview box shows the webcam feed as it will be in the conference +Its the slowest container of the pack because it doesnt currently support the notion of configuration so Ive had to copy the full runtime too! Also G does not support static deployments. +Is this a Struts issue or a Freemarker issue? At first glance it seems like a Freemarker issue +This will be fixed in subsequent releases of the AS +This issue require Arquillian to expose more data in the Packaging work around is sufficient for now +I have added a check that the can be retrieved before adding it to the navigation tree preventing the mbeans from being shown to users who cant view them anyway +all issues concerning the license review are solved Ive added to the trunk Ive reorganised some of the targets of the Ive added a target for a standalone version of pdfboxAnd now Im preparing the first Apache release of. +Not sure if this is going to be tackled completely in but we need to investigate and have a time frame for sure +I cant think of anything else to add. +Heres a better patch is not a new this is a modification to +At the moment I am pressed for time +So a Block object is initialized we set its block size to be zero that leads to a parameter of zero when get is called +I agree that makes more sense if you explicitly set the instance thats the only one that should be searched for callbacks +There may well be other issues with this approach thoughI need some time to think more on this svn diff for the changes referenced in my comment +Ive attached the log that Ian sent to me +How about a nested SELECT? This is a familiar paradigm and allows you the staged isolation and no new syntax +I cannot reproduce the failure you receive but I reproduced the similar failure +this patch keeps accepting connections until it runs out of file handlesThe number of connections that a server can accept might be far less than the number of file handles +sync +Theme selection screen in componentI wont be able to work on this for a few weeks +See +svn diff for adding xsdconfig switch description to to make a bug but without the patch the xsdconfig option is not documented in wsdljava the patch +I guess deprecating ALIAS can be committed first so that the changes for this issue will be smaller and more patch looks fine +Thanks Piotr +Most of the people who contribute to Hadoop proper have a systems background +A better solution is to simply use the qualified name instead of using imports +It could rely on Jira rest API or Jira Java Client for rest API +Ill take a shot at creating one but I dont know when Ill have it done by + +Replicating on the same machine runs without issue +Configuring and Executing a Job Taskletswhich seems lopsided +Maybe something is wrong with your SAX parser configuration +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This is an ant project that will reproduce the issue +As you can see if you use Token offsets could be incorrect because may convert char to chars or the other way aroundIn the thread you suggest that this API could be aded to lucene java +Added a java api link POM property that is configured in the javadoc plugin +Verified on EAP release note fields +So that my JNI at windows env can run the test a directory that contains the dllie target dir +bq +Since we dont currently have a means to test the XML stuff with a DC Im going to rely on Robert verifying the fix +Onto the versioning xml readers +This is the WSDL I usedtestwebservicewebappwsdlschematypesVerified using svn rev +Done +Im even using OSX +This issue is fixed and released as part of release +Run Lars proposal through Hadoop do not think this an issue to hold up a release +but i could not find any docs or examples +Fortunately whoever with the latest JDK or and jdk wants to add a Word Filter or else can easily do it +For the time being gotten around this issue by increasing the heap from G to G on our regionservers +Verified in EAP ER test cycleAll jars are present in client directory after CXF is installed RN contains expected text. +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +save the log and wait for the next log +Yes as Rob C says we want this feature to implement processing of very large files that result in long tails +Something that would be nice is an integration test for this that usesYep I agree +I have committed it to +Includes fix suggested by Martin whereby we check to see if the phase listener wants to be notified before calling it +In other words that throwing away our tombstones wont make deleted data start showing up again +verified by Alexander +In the case of Linux that is almost always +Could you attach your class file here? We could look into it and possibly provide more help +Ok then the agreement is to prioritize system properties over environment variables +is but one system people might migrate from do we want code for each of them applied in this manner +I have tried to reproduce both test cases ant lt isnt reproducible for me +I just verified it is working with the latest HEAD +Then it would also be possible to copy the pdb along and try to debug +If not we can make uDig sanitize the before it passes them to the other problem can be worked around by giving uDig the full urlie httptaosREQUESTVERSIONSERVICE filling the Service really helped +So this bug is not fixed and should be IMHO. +committed revision rregression test added. +Thanks for the patch +This change works around the Derby bugsvn commit srcjavaorgapachejdotckqueryresultSending srcjavaorgapachejdotckqueryresultTransmitting file dataCommitted revision . +The test actually references String instead of the NPE occurs generating the error message +FAQ was updated. +Thanks +Scott you should now be able to replay Ians steps though you will have to work around +hope this you provide a test case that shows the issueCan you provide more info such as your URI stringAny other additional information that might give us a chance of reproducing the error +I think generating should be just need a small info about examples contents could be even derived from did doInvoked AS to undeploy and only then do deletion +Whats that remerging business you are talking about +To support the Port offset in the may not be very difficult but the question is that the port number also exists in the and it is serialized in the +I tried to recreate this using SNAPSHOTS but is seems impossible to me +Plain and simple a block wont be treated as an argument to a constructor +be you have any specific conditions to make that happend? As example it could be Happened only when I have it installed from archive when that was unpacked to eclipsedropins directory or only after JBT started first time on new workspace +to be closed? +Thanks! +The annotations which are displayed are the ones of the current mode plus the annotations in the shown annotation listAll displayed annotations can be selected +is no longer needed embedded jopr does a scan of whats deployed when the app starts and another one when someone tries to hit the app for the first time then does scans every seconds following on use +settings openshift server wizard make these +Will investigate a bit more with an additional test +Interesting I checked it against the SNAPSHOT from last night and it still failed +The case is broken by the patch since the private Fields are now missing but the getterssetters are still using them +Do you have another approach which you think would be appropriate? Not batch junit tests is a real problem for ptest since we pay an overhead of up to minutes to execute each junit test many of which only have a few seconds of real work + +The last patch went with the trunk +Attaching source and war file for modified we need to restrict the scheduling step to the local context through a source check in on there +Anyway if it is intentional then all right +This is something best fixed in one go over the whole doc set. +Reporters of these issues please verify the fixes and help close these succeeds +Thanks for the review Bikas Ill remove the bits for modifying the windows stuff and update based on your comments +That is for reasons +The first patch attached dont fix the problem +If we use resetValid in the constructor we can get those elements which are not valid which should not be allowed +Im getting this error too +m folder during all my tests +a little test is another bug for failures +So their real password is being shown on the logs and our security department is not happy about that +But please in the future provide patches with no absolute pathes +that was fast +Patch with an example that compiles +Changed the old version to work with the new Scanner Delete and +This is not a bug +Patch looks good I like how now tells you version range of your segments +Karl and I are looking into it right now +If there is a special case where this is not correct then the type annotation will need to be specified in order to handle it +Finally got around to reviewing this again looks good to meThere are some minor things like cutting down the test solrconfig so its more minimalistic but that can always be done after committing +Obviously any such action should be done in accordance with procedures laid out by Apache Legal however +Guess I was a little too much when I updated the tutorial page +The good news is at least it picked up RHEL +Is there a reason why you do not apply this filtering to artifacts of type pom +I finally found the source of the brisk version +Asian character in Text Characters in Text Area +NicolasYou should use to generate your patch so that Hadoop Qa can run it +Thats something you should raise against Seam and Drools runtime wizard creation make sense to me to fix i +Randall +Closing tickets +Cool! I would work on itCreation of +A grep for file references turns up no unmodified reference +Alternate patch I wrote up for this patch I wrote up for this uploaded the wrong patch file +Yes I am +The projects most impacted by this will be and Errai +The question here is whether Id like to add in either as a dependent library or as raw source the project which deals with investigating jboss modules +Verified in +Found missing scm url to update for release preparation. +Will close the issue as the result of this decision +Accepting for review +This was exactly the issue that I pointed out tried to fix it back then which did not work out and then forgot about itrelated P open bugs to P pending further review +Committed the new doc page to subversion as revision +Patch to allow access to the real object +The workaround is triggered by adding true to the command line +And the backport of hibernate patches would fix the jdbc problem +Actually this is against RC on sigma so I suppose theres a chance it got fixed +The NPE occurred with the Beta snapshot I built last week +Hmm after looking at the class a bit more I dont think we should do it +It seems to break some naming conventions and it looks like the datacoverage is actually duplicated now in the release directory +I just came up with a different way to express a pause phased in the startup processThough I think wed have to increment the level and start the bundles manually as the start level does its job asynchrnously so we need to know when it has actually finished its job +Thoughts anyone +Is the patch we need just a or do we need to update Avro as well +Thanks for the update I wonder if you doing it right or not +Resolving this issue as we have a workaroundThanksIsuru +I dont think any logging would help on locating this one +We do not need token to be passed back thats why +Please create new JIRA if you still see this issue with the latest code +The wiki page is my attempt to capture it +has this been raised with atlassian? its their thing +Im all for solving the problems that you are bringing up +Yes that works for me +Part of a bulk update all resolutions changed to done please review history to original resolution type +Whats it to youI understand if Rob Mike etc +Lars HofhanslI prefer to make code short and let the instance thinIf implements there will be more infomation in methodinfo and interfaces and the interface method referenceZhihong YuThank youKill the code which is not usefull +Patch was reviewed by myself its OK +We could also turn it around and only use Objenesis when explicitly reasonable +Next I plan to write upgrade test to make sure that this method will not be available on dbs softupgraded dbs +Tilde works OK in the Roo shell +Thanks Liyin +A similar change as made to the api in is needed +Ravi the ask is for an additional field to display the URI of either AM for running apps or for URI for completed apps +The change didnt belong to this patch as Satheesh guessed +Problem you for reporting and providing test case. +I just updated this patch to merge the changes that have been committed since early August when I originally submitted +If the file has an error that causes it to not have an AST then all imports may be removed +It looks like I made a cutpaste error of some such +What you suggested sounds good +Beyer I saw DRLVM doesnt use with values though there is implementation code +Did you try in local mode or with mysql +I dont know if this can resove the problem +Maybe till then we should not allow +Let us know when FTP is ready and well add it then and please dont send any doublettes of your reports to mirrors were receiving your responses here too so nothing should get lost +This is the log file showing the debug output of the current init script and how it fails with httpd also running on the have been using this machanism since EAP but as you mentioned if the port has changed it will still be checked +I think this is a better approach since incorrect dirs is likely to be an accidental configuration errorThe patch itself needs one fix +This doesnt include any stargate stuff only issues there are the tab issues reordering and whether to expose the pools or notIm not sure we want to remove the ability to work with directly or notThis patch a bunch of stuff cuts down on the API quite a bit and reduces of HBC when getting pools that already exist and will always reuse the original when getting tables +If no one has a better idea Ill close this in a few days +So the patch is fine just that we may have other issues if we set the timeout to something different than +and release +I had a working skeleton and was working on the GUI cleanup but that was several months ago +It replace the issueCan someone close the old oneThanks so if any desagree with the rule FTC Day this is not an issue and it can be closed with the causal Wont be fixedThanksMarco +Uploading will add only missing should be added mostly for the correct file path +Committed revision on trunk and revision on branchx +Thanks Benjamin for the patch. +One for the portlet one for some psml changes in Jetspeed application +maybe not the pattern tokenizer even though I think its duplicated effort +If you want a more specific scope declare the variable appropriately +Also one may choose to add another primary storage pool after adding the volumeBased on the above I see no reason to check for pool capacity at the time of adding volumes. +Why is the issue marked as Awaiting test case? Tony OGrady submitted one hasnt he +Nothing critical +This is coming along well but Im having trouble with the automatic coercion from to +Hmm I changed the method to explicitly decode as UTF +there is a javascript version which will create the inlining on ryans blog i believe +We have pending now sent received +But a patch for trunk is needed +I uploaded another version of the patch that is almost the same but removes a little bit of what I believe is dead code from the php build file +Had this been a good sized feature then it would be another storyWere looking at making a hotfix for this and other issues on Lucene relatively soon +Even with the new changes Im not quite convinced yet +as just some text +We do add a number at the end but never change anything else +This feature is becoming colleagues ask me when they can start to use Jackson +Hi Lari Below is the observation on +hftp API is significantly different from webhdfs even for reading a file +In our case that gives problems because of application state manglingThanks +Cannot reproduce this in I have added specific tests for this scenario +Should we provide fix for this issue in +The fix does not work when migrating from a installation +Thanks for the great find Alex +This is great work! I have some long running tests running again and will continue digging as well +This is a one line patch that fixes the problem +This patch makes the changes to the Tools Guide and touches the following filesM srctoolsM srctools at subversion revision +the issues is about invalid dependency paths but you are also talking about unexpected char think I know why the unexpected char happens can you your build that causes the invalid dependancy path with X and provide me with the full log please +When postgres features working blob support we can revisit +If you clean up the patch in time for JM we can pull it back in +Thanks yes that was clearI take a closer look on wednesday +Right +I added a test and fixed some remaining problems +closed this in place of the QA bug by mistake +Some comments to clarify would be nice +We did not want to make have a dependency on AJDT so we kept the debugging support in only +Still not +So if you are comparing completely disparate objects instead of raising an exception the engine will simply evaluate the constraint to false +This issue is in a fixed state but should an enhancement issue be created to actually implement the functionality in the provider +wiki and doc are pointing to the VMIm not sure but are you waiting on me to do the Apache Configuration? It is still in an untouched default state so just wondering +End user might mistake that Camel uses any of these +Do we ever want datanode to ship corrupted data? If the benchmarks dont show any negative effect of double checksum on real clusters would that be enoughIf source datanode does not check for errors how would you propose we handle CRC errors while transferring to another datanode? I guess destination datanode should report the problem +The patch also fixes the bug by resetting the position in the BLOB before copying the bytes +So for and onward this is not an issue +Line does not have both nowI am not looking at othe demo yet +The patch works in conjunction with the patch in +htm extensions +Fix and test case checked in under revision . +Done. +New options that control howwhen indexes are rebuilt +But then I tried creating a bean definition with a bad class name as well and the same thing happened +This issue is fixed and released as part of release +I am currently using LUCENE Net version +Reproduction scenario includes +minor comments in +this is at the top of my list of things to do during the hackathon tomorrow +When I try to open a pom file the Pom Editor is not working anymore +Thanks +We should be able to remove these copies when a version of Commons VFS is released that includes and +IIRC you said in another issue that you changed something to force it to use that file. +Well open another jira for this Additional Affected Code from trunk frameworkfeaturesclientsrcmainresourcesinstancesclientetc RussellThanks for providing these patches +please no +New patch without tabs or superfluous ve just committed this +Hi I am also getting the same issue +Patch committed +The javadoc tool appears to have generated warning messagesThis is not right +Now also framework restart is supported by calling update on the system bundle +Updated patches has been attached +I dont know how I feel about having scripts to support external utilities like this +This is probably related to where filters may be scoped +err my patch above is broken +I believe my error here was that for I changed the AOP Dispatcher registration key for an SFSB proxy factory from one based on a JNDI name to one based on the SFSBs OID +If you still have an issue please attach your catalina log file to this ticket +So exceptions are still delivered to callers +Using Spring for JSR support +Ive just kicked Jenkins for this patch again to see if we can get a clean run +I guess I cant say that Svenson is unequivocally wrong but given that the RFC explicitly defines control characters and the major parsers dont include that high range Im closing this as not a problem +Tinderbox and daily tests showed no failures +Patch removing runtime upcasing as in my previous commentCommitting shortly +Ive just committed this to trunk and. +Ted Andy and StackSorry to all +I dont remember for what version we verified all of these +I have added two more keyszkopenfiledescriptorcountzkmaxfiledescriptorcountTravis could you post the zkganglia script +Thanks Ronnie! +Interesting +However before applying the patch attached to this issue please apply the patch for +Is it possible to get a SSL trace with Harmony similar to? If so would be great if we could it attached to this issue +We want to have factory in generic form because we are planing to support various transports which should not depend on any external checks for framing etc +It is a mild change that also adds a Logger instance +Im afraid we will get false positives if we do ll add a section about tag counting to an FAQ +There seems to be no online reference for the error Bootstrap failed +I think we could do this check in scriptshudson and it needs export BYTEMANHOME which install byteman on CI +I wonder if the tackled this somehow? +We should probably mark as Public and Unstable or Evolving +Hi DanThere are a couple of new things that could help +Should change to use yarn apis for most aspects until we add a more fleshed out client protocol into the TEZ AM. +To expand on this the problem does not affect just the mentioned revision but any checkout thats supposed to contain the effects of that commit +In hindsight that is probably a much better solution +Its gross but it works. +We should prioritize this fix +When the host becomes overwhelmed with disk IO it reponds by protecting or more sending them into read only mode enabling the rest of the to perform +Thanks Miyuru for the cotribution. +It now recently does some good sanity checks for things like clearAttributes even with saverestore state etc +my excuse is that it was late when I added it so I was half asleep +The code may be overly complex as portions were designed to handle more generic use +These files should be deleted after the uninstall JAR and the of the bundled JRE have completed +Im closing this as WONTFIX as theres not enough consensus on it for my liking thanks for the feedback +since were using Equinox Bernd thanks for the pointers and offering help +Anupam can you pls check to see if it works for you? found another bug in is broken it overrides and does nothing +The problem is not about consuming from multiple consumers the same message it is about consuming different messagesThat is subsequent consumers always return null if theres a previous consumer with a non committed message which doesnt make sense as if you want to preserve ordering with that you should really throw an exception rather than fail silentlyIs it clearer now +Like Bram I can build ACE with Java . +Any reason why you have so any stuck in memory? They should be discarded as soon as the session is terminated +Updated patch +Thanks Sebb +Which places the value in the three values forIf you specify installDir in the pom itself there is no change it never gets set +In that case synchronization of this method may not be enough +IMO the idea of exposing hadoop classpath was to let users learn about all the jars included in the classpath +without keep alive support of stomp we could send some period msg to simulate keep alive. +I think its likely that the bug is fixed in newer versions of as I havent run into it again +The find bugs is caused by another issue not related to my patch +BrianI think that It would be great if Required validator would go to both onchange and onblur and the rest of validators would go to onchange event only. +Should be fixed after the merge of trunk into the branch. +I really think we should not mark this bug as fixed until we address the jar question thats the issue not the package name +The data file got deleted and recreated +revision fix works fine for me +logger configuration used for storing hibernate SQL queries only +CP fix is in revision fix is in revision . +Added a patch for TapestryIt fixes the issue highlighted in the sampleNourredine. +Any ideas Christoph +See the myfaces website for the email list detailsSo please subscribe to the list and post your question there +Output stream for this item +patch with additional debugging so that problem with unit test failing can be debugged +to write release note issue now that release note status has been set +Classes marked as deprecated such as Category should be removed +click on input to focus it +Matthew I am not sure if you are still around but I have attached a patch to try to address this issue +added to release notes as resolved All the byte total metrics in are now measured using the data type of Long +report back if it still fails and with a simple failing comments from the user but I have verified that it does not workA workaround though is to use the +Have you verified that the messages were succesfully received before attempting to read them +this jar contains extra information that will help to trace the actual problem +Attaching a program which shows this problem +Attached patch uses pom and the profile +Currently its not that easy to avoid the cglib dependency of Wicket +Ive manually verified the fix +Ill attach a patch today so you can check it over +Whats at line in your repo +Manually tested as part of a job +In fact the Home class does have logic that deals with the case when the parameterized type is not available +Patch for trunk +Thanks SvanteAs we discussed on IRC I think we must clear two issuesthe path relative to the root document or current document +if we dont need that code in line can we just get rid of it +You should be updating GF with Weld trunk build as long as Weld team thinks the weld changes are backward compatible +Sorry for the delay on this oneI have merged kristians changes to master preserving the updated jffi version we just spun to fix Win issuesIve also merged the pull requests into the JNR projectsWell be looking at a big release party when we finally push wont wePlease look over everything all and let us know if theres anything more we need to do to get maven artifacts in proper order +But my question is what are the use cases for them? You propose two where do you plan to use each of them +jks with other workable jks in systest +Regards! +This also cleans up some javadoc warnings in +Steve Blackburn will review the attached patch +resolving with proper tag +The prototype looks good +I just tried creating a new project using the create command then copied in my plugins and www folder and all seems good now +Attachd AMQ wrapper logs and amq wrapper problem is related to the being activeIf the queues are switched back to the default queue cursor AMQ starts up smoothlyHowever the behaviour is strange as AMQ is nowhere near its memory limits when it crashes +Hey Rajith bq +access done please you Gavin I tested it but it didnt workssh jacopocConnection closed by +But I will redo the processRegards path fixes the licence header and updates the text in help files in locale now against another trunk made a booboo with the latest patch +New patch is done with svn diff! +Upgraded to +I wish reference guide had a section on these things +We now only test that commiWtihin is set on instead of actually indexing a documentThis patch must be applied AFTER same patch for close forAh will look at it again +last patch handled the write part to quickly determine the mappingThis patch handles the readerscoring partThe internally uses a class to open readers for searching +Heres a new version with setJargetJar deprecated +Dailleurs merci pour iPojo! Je naurais probablement jamais utilis dans mon projet pro si javais du vendre lusage brut d +It is possible to specify a interceptor +Sorry it took a bit longer to get this merged but thanks for being patient. +I testes the patch on my machine +I just opened to revert the metrics system to the old behavior +bq +now you dont get the exception? I guess I dont understand +Quickstart to repro patch looks fine but I dont think we have to cover every corner case +This is fixed already on trunk by synchronizing all methods and I dont think its used on branch +Need to remove the old Working with Derby guide files +Niels I have added myself Rini and Adam as watchers +With version Ive added some type checks +That way if several s are using a single common configurations file the defaultconf can be defined in one place as wellI would be happy to submit a documentation patch as well if this is committedThanks +I am not sure what else we could do thereEverywhere else agreed +When set it deletes the locks so normal operation can continue +Created another target called which is nothing but same as package but minus javadocsdocscndocs +Take two excluded the patch changes +Thanks a lot for your work. +Attached is the modified version using Modified version +This is actually a bug in IO not Socket +Please see for information on the dictionariesso those dictionaries are not ASL licensed right? I need to check with legal if we can include them into our distribution at all so we need to figure that out first +Since Foundation Profile Specification is a subset of Personal Basis Profile Specification which is a subset of Personal Profile Specification I suspect we need to include support for the latter as well +Committed to trunk +Ive not been able to test PHP +In of has the same issue and needs fixed +If the selected disk turns about to be bad then fails with a +rev Actually got the yoko spec jar into libendorsedput the yoko jars in the same client config as openejb +Changing JRUBYOPTS works fine too +It seems that the method of counting the messages in the test case was flawed when running against the embedded version and it doesnt duplicate the issue we are seeing in production +this is richfaces nod and will be released at of a bulk update all resolutions changed to done please review history to original resolution type +a classpath container can neither reference further classpath containers or classpath variables +If you wish for more containers to be supported please create sub tasks linked issues. +Thanks N. +Can anyone provide any additional insight into this? As of I see no +Contributions are welcome +If nobody has any objections to this and once I have extracted this information to another page will notify everybody and somebody with enough rights will have to delete this page +Just to be really precise step still mentions Cp instead of second typo in step of Steps to Reproduce +Virag please edit the lines exceeding chars on commit +to the javadoc of what class? +I am running tests now +Lets close old ticket that are not in demand. + +That looks easy enough Thanks for the explanation Arjen +Id rather we just settled on the blue theme and required users to have to override if they want something elseAlso we should include that in simples and then include simples in xhtml +It is calling rootbinIll sort this in a sec +is referenced twice so Im removed one in in branch +Closed upon release of Hadoop. +I will try to make another test on a Samsung device to see if I can reproduce it. +Committed r +Im not sure what the issue is here the two calendars objects are not the same because they have differentThis seems to be more about how calendar objects behave when their timezone is changed +Not critical to move to Beat +I also checked for the EOL only changes you mention +Were going to put this on the CDI wiki and not in the spec. +I am testing a patch to restore these and will have a new rollup out shortly +So this RS is like not in cluster at allWhen Master and RS are in seperate nodes and RS node is having nw interfaces and operator want to bind RS with a specific nw interface then also this issue may come up +Maybe not that strange given the name but I thought Id ask +And i removed all tabs in the js file +Can you try tests with testng and a? I would like confirm that it doesnt work or that there is something screwy on my machine +Please verify and close +The gump complaints about led me here +Interesting idea Knut Anders Ill try it out +Attachment has been added with description The transition resolvedfixed bugs to closedfixed +In standalone mode changes through the management API are transactional all or nothing +After this is in we can uncomment that lines +The difference between my patch and the original only exists on the linux version of the jre +How many of the steps in Starting and running replication would you have to repeat? Would you just have to specify the startSlave attribute on the slave and then the startMaster attribute on the master? After failover you can not restart replication +Right should have run ant clean first +Sure just checking in +Uploading patch to fix this code in under discussion was introduced in +Hello YuntaoIs this project still active? I would be very interested in participatingKamil were you able to resolve your issue with the number reducersThanks guess the Q failed is because hive set the default jvm setting for task launch is M so it failedremove this setting or set jvm memory for a larger value would be nice. +logStrangely some of your coworkers actually managed to make plugin work +Thank youI shall create a new issue for this and attach a patch there +First attempt to checkin website changes for subversion revision +Updated patch verified for both and reduce views +I can swap something you have on the physical side +it was not fixed +Is this seconds timeout for whole test suite or just for one particular test? Could I found complete log of this failed test +Hudson test failure is a known intermittent failure and not related to this patch. +DRLVM has been switched back toSo this is not a blocker any more +This patch is a change from the previous version only different in removing an incorrect javadoc line +not via the proxy does it still error +Move class to package +Done thanks. +Verify that trunk jar patch does not leak connection with pool size set to set to and launching client threads PASSED +Builds build with bytecode +Patch is now on branch +Oliver can you test the latest code and verify that it fixes the problem you are seeing +Please dont close as invalidI agree that to solve this for we would need some way of identifying different windows +recreated patchprepared integration test for testing of inherited namedQueries via mappedsuperclass +Id ask you to put a new plugin version to the unstable build site as it gives people whove been struggling with the issue a way to work with a fixed plugin and dont see a problem anymore. +What version of are you using +Fixed in revision +This problem is also actual for change and node +The custom changes that we had in is not ported on to +hm +Previously there were two sizes x and xNew trend icons are located into imagestrend +We need a patch to bring the old test case up to speed. +I will write the tests after a committer agrees with using encoding. +Its good enough. +I just ran into this exact same error on Windows Vista Sun JDK u SVN +This is separate patch from native resource cleaning +This looks like a good area to start in so any opportunity to jump in would be great +I was going to create an issue for the examples link not working +patch was applied +Now that the big changes in are winding down Im planning on revisiting soon and preparing a patch +Hari +Thank you for your mail please note Swapstream email addresses have changed from +Lets punt to d really like to see happen for +ScottSomething really strange +Ive updated a new patch incorporating the comments +attached to confirm the issue is in this release issue exists in heap profile results base test for the following patched version on same platformenv same test as above with patch appliedlooks to be in the right direction still a bit off the results +Duplicate of have searched better before opening up +Ive made it a minor issue as any other variable name including Date with a captial D works so there are many work arounds +Patch for issue +I tried a JDBC client example and it did not require any additional other than and so it looks like we do not need rework the changes to avoid dependency on serde classes +new patch to nail down some findbugs warnings +The NN just slows itself down under loadDo we already have a jira for this? Otherwise would you file one +If by copying you mean you will keep the old classes around in a defunct manner it may not work for thirdparty sqoop connectors +attached a like the first sentence of the warning message Apache Geronimo is currently only certified on a JVM +this patch removes redundant javadocs +For your concern its ok for get to return empty leader in a transition state +no there is really no magic in this sling symlink feature concerning advanced features like jcr queries or observation because the JCR does not know anything of the symlinkyou have to handle this via code manually +This comes as news to me Ive been building on cygwin almost exclusively for a couple of years now and has served me excellently well +Here are just the tasks and the change to to support sort +Im already doing this with and launchj +sometimes involves doing something yourself and send in a patch for functionality badly needed +But this does not solve the problem here +After a few minutes of thinking about it its probably not any worse than the existing ways that people end up with huge tlogsIt seems to me that honoring both idleTime and maxTime would be the best way to go +Thank you for figuring out the detailsIll check the faces messages from maybe this is even better for my componentsregardsDaniel +Ah btw if youre not familiar applying git diffs its just like an svn generate one but it has an extra element in the path that can be ignored using patch p on the command line or asking Eclipse to do so when applying the the patch on trunk DB tests are fine +I got failures in you give more detail +Yea definitely duplicate +The delay for the first attempt to get a thread gives Weblogic more time to get its setup in JNDI +Commited in pngs and link on from launchpad added in Date field seemed to be editable even with no login commited in r +Test case for I tried change only mapping to your but nothing change +Crash dump might be useful +Running thro have committed this to and above +It was really fun for me to help you with that somehow My first contribution to something important +NN is already a bottleneck of the system +is a new piece of code and the stack trace doesnt show Lucene so Im marking this as Invalid for now. +Was the value modified directly in the DB +Would that work +mrepositoryxdoclet +See the effect that it has on entities in the main test document etc for release +The proposed patch is committed with r. +The new patch contains only the necessary change pertaining to startLocalizer +The default for eachof the two classes is the class itself. +for and trunk as long as hadoopqa passes +not finding a forward except in the case you outline is a showstopper +patch in revision +This looks like a duplicate of which has already fixed in trunkCould you please try it with the latest trunk and report if it also fixes your problemOtherwise please attach the eml file of the mail to be parsed +Thanks you ChrisRegardsEric +Added to cover the issue this patch fixes +Have you disabled the inactivity monitor? Its possible that if the inactivity monitor is disabled and the transport doesnt detect the socket drop that you could get stuck in the same scenario +INFO Running for rows took msjmspaggihbasetesthbase cat grep took grep Write INFO Running for rows took ms +I can also uploade the entire spreadsheet with all the formulas if required +For doing a full release binary tarball ant binary is still not sufficient since it makes a tarball the native code is only compiled for the architecture on which youre running unless you pass some CXXFLAGS and CFLAGS environmentsPlus ant binary is useful for people who want to distribute a tarball around their cluster its the full dist dir minus the native stuff which some folks may not require +I guess its fine then +I think the attached patch solves the problem for clients but it causes other problems with at least cxf tools that I havent solved yet +no news is good news. +Build failure is not related to this +the wiki also is updated +Oh ok I thought would make things faster good to know +On linux there is a simple workaroundPavol what is the workaround +Please let me know if that look ok to you +What is the heartbeat on this cluster? Is it is say large like min? invalidateSet that contains the blocks to delete for each datanode at namenode is actually a array +I just committed this to trunk. +Cancelling patch as test failures have not been addressed +The Id is specific to the JPA entity and is covered already in the entity command and since there are many field commands this is best left where it is now +Also it might make sense to do a benchmark before and after this patch is applied especially because it is tricky code and can have a profound influence on performance +The collation server doesnt have any plugins enabled and the other servers just have enabled +Besides there will still be cases when Synapse core may need to be embedded into runtime environments that are not enabledAnyways decoupling transports from the core would improve the situation greatly +Sorry for the mistakeNo big deal chief +On the other hand these nodes will be there either way if master restarts like now so shouldnt matter +I found it and fixed it +That may be but i think thats still atypical +IMHO it would be sufficient +Added to flag all as fatal regardless of code +Did you use the Roo shell backup command? Your zip still does not have any directories in it just file names beginning with srcmainjava +has the implementation to fix it +So the real fix is to remove all the caching logics +Given your comments we should definitely finish this up for Wink +I took out for a spin +Description of how to add multiselect boxes to reports configuration +and I think Patrick has answered and I look forward to testing the patch +add code issue sounds like one I fixed for +This issue may be closed +Committed in revision +Im going to assume thats the problem +Please let me know your thoughtsNote This has to be applied along with the patch otherwise this would break the client functionalityRegardsVikrant +Additional analysis should also be done on how this information is actually used to determine how best to design these object graphs for efficient use during bean construction +The WSDLC command used wasjava or d none wv uri the issue is fixed +I wanted to call for a vote this week but now Im waiting for +Im agreed with most changes except oneYouve converted stackbuf to static variable +Excellent progress +Also it would be risky to fix this at this late date +I agree with the Oracle Dialect apprach +Also Please dont change to include YARN stuff just return values from old MR Arun +in it and it failedIt seems then that the key is in fact the +it seems like this occurs with sizable data +Added second cut Now with global scope as well +See for an explanation of what the problem is +Next I will work on backporting this to and +Much better. +In all cases I have changed them to harry +We dont as of yet have a similar extension framework for the Master +Theyre not data movement applications +if I use then it works fine +When closing the issue the fixed version should be set +I used the function to parse and check the help argument for commands ZKFC FSCK Balancer and +I cant commit so im wait your vote +If somebody really desperately needs the old behavior they can use an old version of the tokenizer +Great +Hi SusanI just attached the patched jar into this JIRA +I would have to go for a new parameter than the current patch +i decided to move the token stuff to a separate configuration Committed revision +Hopefully +There was a typing warning from the use of that I fixed +So this is good +When Ill find some time Ill try to test it +This bug cannot be fixed because the issue requires a spec change +so other concurrent clients would be failure in this scenario +Fixed and deployed for branch +Pushing this to future release until there is a or demand for it. +Its ause this at your own risk scenario +Odd +The patch looks good but I have a question +Wed then need Denis help to provide an updated url default for JBDS +Therefore I fail to see a compelling reason for those maps to be synchronized +Fixed upstream +Ill see if I can regenerate it as you noted +I have discovered that by adding endpoint to the address of the pullpoint in a subscription message fixes the problemIf this is necessary a nicer error would be nice or slightly better documentation on the servicemix page on +implemented Shinobu Kawais suggestion. +I noticed the same behavior on all pages +I do not have enough rights +Let us know if there are any issues +I dont know +However if you are not following links there might be situations in which you want a page to be generated multiple times +Still present in version Its possible to select nothingLabel text in suggestionbox component +catch statement +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Then it should not complain if you specify a listener parameter but dont provide a listener as a value +Ive created a branch branches to work on integrating this contribution +Reopen if a some alias method is really required. +Let me take alook at it +For instance the in Solr is in the default schema and is thus used albeit naively by a whole lot of people +I have no access to this platform but trying to check the build with intel c Windows +The issue here is that is missing from the tasks classpath since sets its own jar as the job jar so is used instead of +Thanks for the clarification +The last column is assuming the patch I submmit has been applied +Even if its a small performance benefit thats something +Lets fix one thing at a time +Closing old issues +This trivial patch works for me +Thanks! Im assigning to Ben since he is mostly responsible for collectives these days +Sorry about the duplicate effort there guys +Changes commited in trunk +Hopefully there are no serious systemic problems with these translations +Ill do that +Can you rebase the patch against current? Thanks +At its time this was not uploaded due to license constraints that might prevent its no version of the hibernate dependencies cant be changed. +I have granted all users the Browse Users permission +Thanks +I am also more inclined to keep in bookie process for now +to for taking a look if there are no objections ill commit in a few days and open a followup issue to ensure that all scorers support traversing there are still a lot that need to implement the method in order to support access to any inner scorersFor example in my opinion even shoudl allow you to get access to its inner scorer via some CONSTANTSCORED relationship or similar. +When I change to pass an explicit encoding to the constructor of everything seems to work just fine +Thanks KeheliyaIll take some time to review these patches +Super thanks Rudi +Im working on a patch following the simple logic suggested by Riyaz +Agreed +Hi you indicate precisely which code drop this is source was fixed during development and got into the release +Just as a courtesy to those that follow Im not sure whether attachment was ever tested +Big thanks Steve +Reopening because its required by +Committed the patch to trunk +I hope that others find this issue usefulvrAdam +Do you think these items can be tagged as a +Rebased Mon per dicussion about AS pull request merging into master will happen in few steps starting with Stuarts tests branch. +So instead of changing the existing authentication event here are some alternatives Create a with the previous session ID and the new session object available as event properties OR Create an to provide the same details as the but published after session migration instead of do you think of these alternatives +FYI the classes and have been removed +The explanations of the examples do not reflect their current state +wiki updated +Ill include the class hereI doubt this is worth including in the standard webwork distro but I figured Id share nonetheless +Chris the class is created for just one utility method and several test classes have to extend it +Post the results here +for commit after javac warning is fixed +I like what you did with the shims that was one I just punted on +I assume this versioning issue with the libraries is an ivy problem +Patch file for IIOP loopback transport +Reopened to set the Fix Versions field to JDO beta. +So for users theyll have to upgrade downgrade or use a RF workaround. +I did not check for Mojarra +If you could introduce some recovery for even longer values that would be greatBRHere is a new version of the patch following Timos recommendations +OK +bulk defer of featurestaskswishes from to +Resolve as dup of. +was conflicting with +See comments in for details on how were approaching this in Continuum +So instead of the permissions on the history directory should be +Merged to the branch in revision +Pardon me +Check Duplicate as I have included QA Subtask +Rev the future please do not jiras whos fixed versions have been released because we loose all the history what change went into what release +If you simply rewrite the contents into the output stream and message compression is enabled youd be writing compressed data with an output stream thats doing compression again so thats not good plus you be adding new data onto already compressed content if you use any of the write methods so you are doubly corrupted +The test was actually based on a bad assumption wrt managing of associations. +The actual response has a content type of texthtml that comes from Tomcat +on the filesystem I do find a jobfile like this which contains the correct value in a compressed formatE VQPerhaps hive reads this as a file +push out to +Thanks for the patchI tested this using the attachement to +Should be assigned to someone else +No seen +Added under the ws seems ws staging profile not working properly for Woden +runtest this is a simple script I used to run a single test repeatedly +Patch to fix sending xsiniltrue for parameters that are minOccurs but nillablefalse +That said if a patch lands for Ill gladly accept it Im just not going to block a because of it +Will release changes after you also require rename participants for field renames +IT no solution has been currently found to prevent using the UPPER function a ticket has been created to fix it later +I tried on Linux where the platform default encoding is and on Windows where the platform default encoding is +Anju are you able to move to the JSTL cimport tag +Let us get rid of useless get methodsand deprecate set oness patch committed +The attached patch v adds error handling to +works great +Yes thats right but it is also a service rather than an instance every reducer runs +The description should be as following The oncomplete attribute is used for passing that would be invoked right after the Ajax Response returns back and DOM is updated +To solve the issue we need a reproducable use case +Removing patch available as this is not really a patch rather an attachment to illustrate things +Since Rory has been doing all the work on Net and saw fit to include it and it is a trivial implementation then I dont see any problem +Hmmm +Go for it +Not sure whats best here but I dont like the idea of adding another variable just to know which case the error is +Give it a whirl and let me know if it works +The state stuck at Allocated state It is a dup of +Bypass the usual processing in when negotiations are complete so that it flows through like nothing ever happendBasically I just made sure to do the NTLM dance for IE without altering the normal Acegi behavior when a user is already authenticated +Im on this. +Added the lisence Made the decompress look more like the snappy decompress +This is the certificate used as exported from Firefox +updating target expand and infuse options are now used successfully as part of the main installer and demo application buildsInfusion was apparently fixed in commit by Ate +Paul have you experimented with changing the buffer sizes and how it affects searchperformance +binary version of niochar binary binary of performance tests with new tests results with new patch has been appliedPlease verify +Im undecided I dont disagree with either you of +Here is a patch that adds a new class with file format docs +Added servers in Eclipse +applied patch with huge thanks to +Commit at subversion revision + +Will reopen if anything breaks +Any progress on this one Adrian? +Just noticed that this issue hasnt been addressed yet +batch update of issues with fix instead of the correct +Thank you very much! You are a magician +Patch to make running the test case easier +I get a similar improvement with my app when setting Thanks +Compare LF of andIn contrast I am for different look of Harmony so that anyone could instantly say Hey its Harmony +SlavaBut anyway we should report this issue to Eclipse bugzilla even if we are going to apply this workaround for CR +I would say lets put it into. +I think I got confused here +If you can describe better when this happens I will take another look +Thanks Mahadev +While communication with server in version looks good log shows multiplerepeated error frames +This jacks up my world as my jar signer freaks out when it sees the duplicate files +von betreff nachricht gesendet mittwoch +John Yes this issue could be closed now +Davidthis issue is a duplicate of which Ive attached a patch to that drops the dependancy on +Not sure about the javadoc +When does this issue arise? When we try to getclosest on split key +Patch attached +Ill commit it shortly. +Looks like this was fixed as I am unable to reproduce this in any version after +I have tried installing older versions of solr +I created a standard Groovy build script and the bpelc task worked fine +Embedded class loading is be now further refined under +Cant reprodcue against head +If it does not feel free to reopen the issue +Hi MirekThe sucker password no longer can be set in clear text +I had different types of first actually used the state transfer and implemented the getset state properly +like so many things in whirr +No white space changes in the patch +If you could this issue with this version we can close this issue and get released +I think you have discovered a critical bug and which I believe can be solved in this manner +havent heard anything on this in a while does it look like the new compaction code in fixes it then +its not collisions then it was worth a try +Now its time to decide the package nameappropriate packages for each of these classes +By this issue I have absolutely no idea what is failing for you what error you are seeing or how to reproduce +The same patch applies for also +It was released with the fix in +Could you try to execute this test with EAP ER please? See my comments above +I didnt include since it requires a instance which is not necessary if we just want to do sanity test for Crunch itselfI have run tests on a centos VM +Fixed in svn +You are using right? StanYes I confirm VictorI rediscovered this issue while cleaning out Jira +I think some sort of testing in the ZabTest model would be good +I think Prakash and Dhruba are specifically thinking about rather than Puts +Moreover the actual position changes with respect to scheduling would happen due to user actions like another job submitted with higher priority while user job is in waiting or user changing its priority +I think this issue will be addressed by the changes Pradeep is review the attached patch and give commentsthoughts +Expected cant reproduce the issue on the latest build +looks good to me! Since this is a blocker Id like to give an opportunity to others to have a look at it +I suppose that solves the issue? Please correct me if I am wrong +Turns out components are used more widely by containerkey components +The Apache UIMA Java and C frameworks support and bit array types for arbitrary data content +In the current version the Tree widget is as fast as the dojo oneDHave a good day +Okay +Anyhow this artifact was the only one in my repository using in that manner I searched thru POM files found the logj approach thought it might be worth raising as an issue +Hi AshishEverything works fine for me the ui is already defined in the +Could you try it again? Now it should workCheersBruno +Canceling to add unit tests andor post manual testing +Might be a tight deadline as I believe today is the last day +now archiva calculate what is the last SNAPSHOT and return it. +The trouble was that the test JVM was exiting in this case post +Similary rename them in the correspond java +tmp dirs are there but theyre empty even though compactions have run +This patch fixes the behaviour from my perspective but breaks a bunch of other tests +Simple patch which fixes the problem for me +Fixed + Seems fine +This error is printed when the user does not have the executable in their path +Capturing dev email discussion here +Reverting this due to reverting +I see that the patch was not acceptedAnd stop to spread FUD. + looks good +Use the to post your issues at first +Actually bag inside map is done by not +Any news to this long running open issue? It shouldnt be that hard to fix the regex +If user need to use the original one check Use manual attenuation algorithm in each config from the workflow +Instead I have attached that induces a single shot failure that illustrates this issue +Hey DhrubaTrailer works too but thats a file format change +Thanks Phil +define blow up +This JIRA can now be closed. +And for the invalid xml problem I really hope maven guys will do something about it using something else than an XML parser to parse data like poms is a non sense to me +Do I need to get them to upload it +source generation throws an exception +Addendum addresses Lars Fs finding +So is reported as cp is reported as cp is reported as This looks correct +Created +It does seem to be some odd combination of s spec runner yaml and the date +Closed after release. +looks checked into trunk and branch. +But here it required prefetch size in as well +Dynamic snitch is designed to mitigate this but really it needs to be solved through monitoring and tuning +and draw attention to the different subclasses in the javadoc if neededWhile is a subclass of explicitly calling it out in the API isnt super useful +for the patch +Thanks +I said it is global +It cant be of the time +buildivylibcommongriddevbin +Will follow up on underlying cause. +Patch applied in SVN rev +An exception should be thrown + This problem is causing real issues on running clusters affecting customers +Do you think there is still more information needed than currently given at the beginning of chapter three +Arin can I take this one over +hi Yonik yes you are right! if i replace or escape this erverthing works fine and Solr still rocksi was so focused on that strings and maybe wrong app encodings that it took me until today to see that unescapedi played around with that stats but i cant get this error simulate on other queries i mean my problem case include three items with the same priceI did other request which response with items with the same price but i got a stddev of +Also too much noise on the comments on a few issues if still getting errors like this another JIRA would be appreciated to make it easier to track! +Committed. +Also thinking longer term we would rather work with Hadoop by default rather than earlier versions +More specifically we are trying to make play nicely with hadoophbase conf so that we can interpolate things like as the default for the ZK dataDir +Thanks! +The issue dependency link should checked with the person who created it +To be determined in a later time frame +We encountered some dead lock situation wrt to the advancing the mvcc and the order in which the flush was happening +The version number on the DTD is not incremented as long as the mapping files validated against this DTD are compatible with a major version of Hibernate +Andrej I need more input otherwise I cant work on that issue. +looks good to me and ready to go +seems best served by the enforcer plugin or similar but worth considering in some core way in the is the job of a plugin. +Im sorry I just realized I run the tests under Linux +docs build OK on my machineNote No new test code changes to documentation +Fixed under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Attaching a hotspot snapshot from +Ill pull the acegi library into trunk but since were in RC and release is imminent it probably wont be tested enough to make +It does not have the same meaning as clusterFilter though so would need to be a different flag +But the database is not created +HiWell you have to use a VARCHAR column as collation are used with string values not int +I will commit this soon if nobody objects +bulk move of all unresolved issues from to +Attached is the patch file to the files that required updates +committed revision to newtrunkif you have problems please just revert and I will help debug themfor future speedups we should try to move ant logic to and it for contribsthis way DIH tests etc will run in parallel too +To ease fixing existing users of the generic api would it be fine if I added a constructor to that allowed for Java Enums for the passed in symbol +I have found that Derby may be optimized by INDEXINGCache increment does not helpresults in attached file +the former will preserve the history where as later do not preserver the history of the file +Ill check it in +Testing in secure environment is pending +bq + +Where can I download EPP CP fromYou can choose your own portletbridge or Seam distribution +patch for quick fix of the issue wrong jar filenames in in EAP +I still think that removing the changes is a step backwards +What else do you require +Im confident that the test failure is unrelatedThe javadoc warning is a little curious though +making it run through hudson +Hi JeneThanks for the patch Ill try to get this one into the CVS shortly +Is this how the Alpha is patched now or could anyone point me to any sort of temporary solution to achieve this functionality? Thanks for your great work +Initial support for Selenium test scripts are not managedupdated when a domain class is modified. +ok is a folder in erated applications that just contains temporary deployment files +s usage is a bit involved since it uses an additional anonymous class that extends the setUp and tearDown methods in the containing class +So I posted the question hereMy question is whether this new API should handle recursive traversal and symbolic resolution +has some notes. +superseded by +so back to a very generic name like slingContent? +Rebase combined start reviewing the patch by tomorrow Jeffrey Zhong +Ive run some not all +That would also speed up streaming +Getting the same error in CDH Hive while running query with lot of dynamic partitions +I and Amar had come to a consensus that srctest needs to have that property defined and it should be used +Robert great news Ill try it right now report back my findings +Ive attached a patch to add this functionality +Let me know if there is a wayI will generate a review request once has been committed because currently this will not apply directly on trunkThanksVikram +Im reopening the ticket +After adding with ant and removing saxon and xalan jars from toolslib I am able to build portal trunk and run reports as well +NOT TESTABLE WITH CURRENT resolution state to Partially Compleated +Temporarily reopening to update release note info +Fixes fb compile issue and other pure enum related compile errors +As was discussed with Freddy blank commens still should be excluded from comment lines metric +Are there any other combinations that you want me to try or can you reproduce this +See the proposed system will be available since version +I just committed this. +I definitely cant see them behind a corporate firewall of one of my customers when using IE +Can you try with those versionsThanks +Committed to trunk and. +It would be good to finish with clean style indeed +It sounds more logical there. +Uploading a new patch which keeps things backwards compatible +If somebody can change the contents of a bundle he might as well remove all traces that it had been signed in the first place +Published! +error messages now link to the location in template where the error error message model also has info to be able to link to the original scala source code location but the view is still not rendering that bit +The UI is like pictures attached +ClausI just tried with Camel but unfortunately I get the same error +Patch integrated to and trunk +Gabriel the two gwc patches represent the changes against the stable and master branches +not blocking issues still relevant most not resolved so we can actually benefit. +This is a first release of the code just to get it out and have peoples opinions on it +Hey RossSorry for the late reply from my side +bq +You can see that garbage collection works in the profiler the number of instances is held at a rather constant level that varies with load during the hours of the day +For example I wrote the sort and word count examples that many people use for the start of their first mapreduce program and my name is in the author +Updated patch to trunk very minor cosmetic differences +As evidenced a bad POM must be circulating on the Internet +Right now we just pass the schema string as is to the +Please apply the following patch to trunk +Candidate for SOA plugin provided in er but not yet in plugin pack +And the translation task is been done by a professional translation company +But I agree the error message is not clear besides the there is no indication that this is a circular dependency error +The whole ESB group +May I ask you to try to reproduce the problem with +Well just the number of completed tasks probably isnt particularly useful you are rightI feel like we need some additional tracking around streaming though +As there are no more slots in YARN and it is not a public API would it not make the most sense to remove it? lets remove the method itself here +Fair enough. +We should open a new issue for this +is covering the fact that report is a deprecated format and will be replaced with a format +Keeps it simple in terms of an equality matchFor the signature will essentially be a type with some matching functionality +I will be travelling and will try out the patch ASAP +Now script mediator will always pick up the correct JS engine in all environments +but that is not crucial. +Sure thanks for checking it out! +Fixed since tags +AFAIK there is no Harmony equivalent yet +I agree with you that minimizing external dependencies is by definition a valuable approach +How about a more structured performance report instead of just logging the numbers? Id like to automate the generation of some kind of a trend graphWe can take care of and also later on +No one to blame just saying. +Can you please confirm this works for both the cases where the parameter is specified and when it is omitted? +Ill be careful in what I say +hdfs probably needs to store metadata with the files or blocksInstead of storing replication policy metadata with the blocks could you keep the responsibility for this back with the replica placement? So assuming the balancer is also updated to use the block placement interface if a cluster has to support multiple replication policies it could be the s responsiblity to decide which policy to use based on the file ownerpermissionsfilename for the block + +Merged with trunk and regenerated the patch +just related to m able to install a cumulative patch now tested with AS SPcp installed on top of a instance using ON +It still misses integration of the added Path class in the existing code where it makes sense and doesnt break the APIIm waiting for to finish this +The former change is needed so that we dont have to keep flipping the default in fix branches +Cant you do this with a custom +Ahmet I just checked in what I think is a fix for the above problem +I think in this case FINE is ok +Bulk move of M to get desrciptions from TLDFaceletscustom xmls now and show it in code assist. +Committed revision +If you are looking for PB only serialization why cant you directly use the generated proto records? for example uses them to persist data to +Hi Jacquesmaybe this one will work betteroops forgot the asfopt seems to contain the productId of the variant +Code modified accordingly to the Katheys commentsPlease review Ugo +Can you check if it works for you +The biggest impact of this change is that users of either should work directly with the last execution for status or execution attributes or with the current execution +Fixed in r +Not worth fixing locally +Ill commit shortly +It doesnt feel right to have a config file that is specific for mapreduce but I think there should be some kind of default config file that the user can override in the job but that is editted by the siteThings that probably belong there +Please release a new version of the plugin with castor prefer that Werner should take this plugin over +So this issue is just about annoying exceptions in log +Thanks Luca +If I want to from the milestone site the JDT patch is greyed out as if it had been already installed +I should not have the patch to Jenkins +Thanks Sandy +This may not be fully polished all corner cases covered until a later beta but we should have the basics covered for +Yet another try to improve the documentation this time targetting the javadocs for the Plugin API +We need to recursively delete the entire dead branch depth first +Id prefer to batch any requested changes into a single update if possible +Heimberger +This creates a dependency shared by all distributions but if we are willing to manage it then OK +The good ones are closed source the opensource ones are use for all your pasting needs. +Hopefully switching the default to scratch buffers in doesnt raised a problem +Repeating the questionbq +Should we perhaps do an sNNNMg in the description of this JIRA? Or does this have something to do with the Name Node and Im completely missing it +to Daisy review +attach my test patch +I committed this +You make a great point Erik this sort of thing surely has been considered before +if the patch doesnt involve changes to have been having trouble getting Jenkins to run for patches that I submitted into the directory +Galder whats the status on this one +Is that Ok +Verified fixed in ER build. +is a mb repo which can be slow over a slow linkYep minutes is what it took here +It usually subsides after a few minutes +Just my cents +It worked so I will publish the patch shortly +It doesnt crash on linux with acrobat +For identical queries in my test index bjoin is twice as fast as hjoin for both small and large inner set sizes. +Thanks Tom +Slightly updated version that uses appendEncoded to avoid an extra deserializationserialization step +The eclipse plugin has been updated to discover the available workspacepaths from the server and provide a drop down list of those for the user +Thanks Jaimin. +Please see security documentation for configuration instructions +If we provided a Javascript one we would need to support JSONP Dojos transport or some other hack to get around Javascripts SOP restriction so you could POST PUT and GET from a different host +Looks like any class which implements is considered as a in even with a name declaration +Patch applied to resolve defect in testcase +need to be sure we dont remove things from central that centrals tutorials are dependent onSo a start would be to add these +See attached patch +mvn o N install in GERONIMOSOURCESHOME +Thanks Deepal +Patch for this for the patch +trunk The non osgi has been updated to Claus +Got itHave update the JIP to also use the JTs system dir to get the file system patch while Eniss concerns are addressed +Patch for trunk that also addresses points above +I would wait for Lars decision on inclusion for test suite validation is also neededCan you attach patch for +I thought this was handled by Faban +The endpoint property name was and needed to be Endpoint to map correctly to the deployment template +Assume closed as resolved and released +ok the caller security principal +The reason for this is that their names are not unique +Sad +I even dropped a note to the dev indicating that all the existing tests are passing in the extensions module +Maybe the formatter settings have been the cause why you could not reproduce itCan you try to reproduce it with the below settingsPreferences Java Code Style Formatter Edit +I will change it based on what that code +Tugdual Can you please tell us if the JMX approach is acceptable for you? +I verified by running the test largedata successfully +I will see if I can get help fixing that if not will submit a patch without moving them +Moving to next release due to lack of information to work from +Removed stray prints from the for taking this on Ashish +Actually I have open a new issue for this +In total five date comparison methods have been added to +Marking as a duplicate since Im pretty sure those tickets cover this issue appropriately +Knut Anders thanks for committing the patch and particularly for running a test on a buggy JVM to verify that the the buq was detected and that the workaround was triggeredI will submit a new patch shortly which do the changing of mode for opening of log files +Im changing it to that +could you please attach log files and error messages? just use Luke to see the segments its display good and I cant find any log about this errorNo tomcats and the its looked goodI think some wrong with image is the search chinese word before and after I think you know what I just test the tomcat linux and I seach the chinese wordthe search result was good and when I click the web cached linkThe web cache was good +Can you either attach a test case or point to a test case in the Tuscany itest suiteThanksRaymond +I wonder what versions are affected +mdsum fceeedcacacf. +just callvalidationfaces in summary fixed +The pull request implements generic syntax lexer but detailed and accurate ddl description still required +HiIm not quite sure from this bug what the underlying problem isIf the JPA provider is trying to use the application classloader to load its internal implementation classes then that is a bug in the JPA provider not the JPA container +locally run these test and confirmedJPA change included in patch as previous commentthis patch is ready for commit +I think that the presence of security annotations metadata with no security domain should probably trigger some kind of warning +Question are Can we afford the increased memory consumption? Is KB also the best choice on other systems +Otherwise leaves in jgray fixup +Since Falcon is doing a build of oozie locally by applying patches thats how war file is present in local m cache and build is going fine +after release. +showed that the patch removed the additional listing call to the parent directory if the input path did not contain a glob +Ive added a test case to try reproduce it seeYou can try it out in the assembly modulemaven DcountMaking about or so seem to run out of heap when using a max heap of Mb +Attached corrupt file +Theattribute is stilled used in some examples in the pages +Thanks Harsh! +That will be the causeBy putting the last test to first it is reliably reproducing in my eclipsePlease post the patch It help to get the trunk green soon +I committed a fix that should address all of these issues in roller trunk revision +This patch is now included in the ueber patch in bugenhancement +Any comments about that +Seems to be good not sure what you did but thanks +TomaszAbout library classpath issue was already postedIncluding your posted how about resolving them togetherFor details see the issueThanksKan for your reply +yes +Also killing servers didnt kill regionserver hosting meta because that makes a mess of counts tooBut killing hosting regionserver balance came back promptly +is using a under the covers which forces no read to span a chunk boundaryThe jetty server on the other side is arranging chunks so that the characters start the chunks +If we can save this minutes for each patchdeveloper we may easily save thousands of minutes before the native compilation is fixed +OK +Filed for this +The DROP COLUMNchanges to SYSCOLPERMS will be a tad more involved than the ADD COLUMNchanges to SYSCOLPERMS but we can get most of the heavy lifting outof the way by fixing first +Thanks Kathey +You may add subview to your pages pr use existing +Well this issue is opened on Aug +Dan if you want to continue testing the jobtracker recovery feature for now please revert the patch in your local copy +Artyomthanks again for your fix compilation passes nowYet Tomcat does not start for me and I get series of exceptions in the log +Yes this indeed is annoying +This version is for trunk but will be merged to the branch and edited there to include information about changes to the release +Goerge I agree with your correction +Can you create a unit test to reproduce this? Or at the very least give broker configs more detailed logs and steps to reliably reproduce this? Looks like the index pageFile is null but no way to know how that happened +please provide a quickstart +The barbarian example isnt very real world +which have nearly the same functionality as +If you prefer getVersion instead please change it +The reasoner was enabled +Not sure we have enough resources to dedicate to some JSDT extension component thats maxs call +Will commit +This is clearer thanks +Marking for so that it can tracked accordingly. +Maybe every s? You could check a boolean to see if the timer went off every tuple if so report and reset timer + Username cjohan +Robin check out the integration here is this what you had in mind? I have the structure for unigram passthrough in there just isnt done yet +I hate class think we need to fix this before final +I cant think of any reason why not +So once we have a new build that includes the refactoring on both ends well see if we can reproduce +Attached patch adds this test to exclude list for intermittent failures on Windows +Committed. +Added the String icon field to hold a location for an icon for the +This JIRA intended to track the work to make those only failures passing on Windows +Changes Changed ACTIVE to NAMENODE +AHA my mistake +Actually Ive met occasionally this bug +Let me know if this is what you needThanks. +This was originally how the command worked but somewhere down the line it was changed +This begs the question Is an the appropriate place to launch a subsequent job? My guess is no +I was reporting this issue to make it possible to have an optional descriptive label attribute on graph the label is not specified it should report the same value as the name it would be possible to have more intuitive graph element descriptions taking in consideration that names being identifier should be unique and could follow other I think it could be easily realized a adding an label attribute accessible from the Element Properties Dialog and from the Properties would also propose that the label text will also replace the element name on the graphical canvas. +I am running tests now +Creating a that allows writing? Glab +I committed it to. +Patch looks good Ted +The web application will still be packaged distributed but will not start together with the subsystem +Ive created a discussion thread for this issue +Same problem in as well +The offending line even as the word ugly appended to it as a comment +Any progress on this? I dont mean to nag is there some workaround I can use to search for and install Roo +Lets try it if you can easily create a scenario of from aboveManik no I meant that and other iterating operations we create a clone of map and then iterate over its contents +Have not had the time yet to look at this one but if youre right this is really serious +The methods signatures nevertheless are already forbidden to break those rules +I am not sure whether this is a valid issue or not +Since this is my first podling Im not sure who should have access but for now the Struts group would be fine as long as I can add folks to that list once I get them Apache accountsThanks for the help +Looks like an agreement on the current behaviors correctness is being reached and I can defer the search for the documentation +marked as all old Resolved issues +Should be faster but need to verify +fix test outputs of two new added tests after rebasing +For eclipselink there was no workaround mechanism available so the patch was included into the eclipselink projectHowever since there is a workaround here which does not require directly patching I think this is sufficient. +Thats why I want schema export to do it it cant be that hard to add SET NULL in addition to the CASCADE option we already have +The fix as mentioned in the first blog posting solves the problem in case of +Richard Please post a review request on +The replies just keep comingJeff +What if we removed the tabs? Better to not tease the with things that are broken +Seems the consensus is that even though the names are not perfect we should stick for them for consistency +So this is a bit complicated +Bulk defer all frontend issues to +So the may be in an inconsistent state and that causes the NPE if you rename page A + +Sorry guys I selected wrong version of Xerces it is not +This was a problem in +Hi zhaixuepan +Will close this ticket when is fixed +Oh noticed one other thing +should be removed along with the later references to utiliies as its unnecessary to mention them hereFor the last two comment if the variable is empty or not set would be correct but as for whether the section on chmodding the files is even necessary I actually dont think so since I just checked both the bin and zip distributions on a Linux box and the execute bit was set for all of the files after unpacking them +Now its too late for such impl additions + +also check new whats new coming via part is done +Now has been fixed and and derbyall run cleanly with the patch +Should have the ability to dolt nameproductForm typegt lt nameproductID type initialgt lt nameproductName typegtltgtlt name type extendsproductFormgt lt nameproductName type initialgtltgtSimiliar to what Tiles provides with the definitions +See the or the developers guide for more on this. +Thus providing test case see attached zip +With and without the patch the test runs cleanly for me on both OJEC and phoneMETouches the following fileM javatestingorgapachederbyTestingfunctionTeststestslang +Initial patch for review +Do this problem still exist and where is it located JDO or XML +Where do I get access passwordsto visit the admin UI to attend to moderation requests? What is the URL to the admin UIHow do I subscribe to these new listsAssuming a pattern here is that by sending mail to Pardon my ignorance Im new here a little more info would help in cluing in me inThanksAndi. +Please cancel this issue I have determined that it was in fact caused by the bean that used placeholders was in fact referenced by the bean that was injected in my postprocessor and was therefore early initialized +Patch for extending the with the Annotation Ontology and iCLA have been acknowledged so this can be committed +Sorry I forgot to mention that I updated the file accordingly so that the subclass is properly registered within the web application So it was wise not to close the issue +Here are the actual of JIDE Integration demo is the correct screenshot the old one was only of the internal frame +Ok theyre not trunks got a few issues but not near as many as Im having with my patches +We might need to use a that provides API to close open JAR files and then expose the ability to reset the class loader for third parties like no luck +Closed due to the release of Tiles +It would still be cleaner to pass the cache instance to an than the since the manager is really a detail of the cache implementation and should not be exposed to other components like the +I reverted things and went with a simpler change to accomplish the same goal in r. +test is now failingplease check withmvn clean install PtckI think we must only disable observer methods if they are overridden in a subclass +on trying to put it into I think needs to be fixed before I commit this. + +Okay got a better matching now considering the MEP patternWill now skip void methods for out capable exchanges +Ill provide scripts to move the remaining regex tests when is fixed +I feel like I missed something important +Opened to log the inconsistent shell behavior +nd try of patch to fix the condition that show up after second restart +Also Thanks for Vince and Eva for help in designing tools +The test case includes test If were updated to use the new API it could be useful for the iterators of and +The compile triggers a visit to the enhancer processor which creates the query class source files and then model query classes are compiled together +I tried against H both with a new database and a previously seeded database didnt make any difference. +It was a problem with server side state saving again but now the demo run without problem +Forgot to mention I havent set up a notification scheme yet +Fixed +Thanks GarfieldFixed at r +Bulk close of old resolved issues. +This jira cant be backported to released branchesIf Rick sees a need to cut another release jars we can backport it to but otherwise it will need to go into next major release +Correct migration +KB Time ms +When can we get this committed the current patch onto the trunk for now +Marking patch available for reviewcommit +Update patch after merge and adressing comments from ArunAddressed and +I understand that but its not a complete rewrite as youll see +bq +Oh ok +Im attaching the new one +But once this issue occurs application cant access any asset or process definition from Guvnor using rest +We do not provide rsync access to the repositories +bulk close of all resolved issues as part of closing items. +One more patch attached fixes silly NPE bug when was null +I looked at I thought this use case was simpler +main +disables creation of the second Connectorserver +Getting urgentHibernate Search was just released supporting Lucene as well +Build results are attached above for before and after the change +to updated patch for the current revision of the trunk +For instanceLets say a database is created to use a territory based collation for character types +What do you think of that approach +I accidentally revoked the ASF the grant for this patch and now Jira wont let me set it no matter how I tryWonder if this works I hereby grant the ASF this patch which I have already implicitly granted for trunk. +Im aware of the per issue the aprpoolcreatecore method is a new +Heres the example Ive been playing with +This is required to be fixed in order to use Travis work started on this branch +This looks good but one minor nit instead of cipherSuites in the yaml can we use the more idiomatic ciphersuites +This was never the intention of deployment overlays and we are now looking at addressing the OSGI use case in a different way with deployment specific metadata. +Sorry was getting error because of the LDLIBRARYPATH not being set correctlyI needed to set for both Xalan and Xerces librariesThanksJim +Thanks for the feedback +I thoughtResolved means that its fixed and ready to be tested +That could be the case if two threads try to release the connection at about the same time +All tests now pass +come from Java +This class no longer exists in the codebase +Committed revision +Please let me know as I am stuck and we need to release a certified version of our app asapYour help on this is much appreciated +whats up +here is the issue try to summarize our discussion about problems purposes and solution +This happens because when the thread runs loading of Daos is not yet finished by Spring +Committed trunk revision Committed revision +Ditto for +duplicate +A new binary was created that doesnt include the javadoc and sources +In that case this should be documented in some paper on migrating from to +This also solves problems when running huge builds via the Maven reactor +A courier was like a tour rep who actually accompanied a party of holidaymakers on their journey not just being based at their destination +for qa +Also not that the issue happens at random times +Im pretty sure though that we tried just using the b with the customer and that made things like the admin console work but not the JBRHTTP gateway. +Yet this does not use Version to preserve compatibility +Fix test in +Ovidu was seeing associated with the timeout +The team was very happy to know that this issue is being looked into as there is quite a lot of support to get to a elastic vm solution using ATS +I will put this under extensionsgwt as servlet and sample +Full wont be necessary if you have full hdfs test result manual test result of the fix +bug changed after todays pressing the key DOWN is corrected +From my side the blacklist is NOT propagated from MR AM to RM +Yes Ive actually run into that problem too +One of the jars I added is if it is on the class path the above exception is printed outMaybe this is related to the usage ofHow are usually deployed? Using PEAR + +need a different fix +I need a more complete bit of test code to progress with this +So I dont consider this a minor bug but a real annoying one. +The s for EAP are being reworked in +Also patch looks good +Thanks JoydeepThis isnt a regression so I dont think it belongs in +Im about to write an integration test for this +The screenshots dont tell me much on their own +reverted the revert seems to have fixed it. +Not sure that it will expose ALL of the cool widgets the way the reuters data does but we could also point to the reuters demo or make the download of the reuters data something we document and not have it be automatic +button has button reappears only after changing selected node in page will roll back the last chang to evaluate if this issue remains fixed due to other commits or shoud stay open until a better solution is found +A selection event can maybe cause unexpected things in other views +There was a bug in the so the real bug was hiddenNow Ive fixed both for the next nightly build + of release. +Thanks Amareshwari! +Change made and committed thanks +I also updated my sandbox sohopefully you will be able to apply this version +Update the diff report to work with the file diff change +I am going to try out other samples with empty repository and let you console output for running mvn package in wannabecool sample sample works just fine +To get the send error status returned by httptransportsenderwritemessage out to the client changes were needed not only in axisenginesend but in axisopclientaddmsgctx not to destroy the error status when a NULL incoming parameter was seen +xhtml so when this is the case it fixes this problem and sends a redirect +Its a local problem +The sleeping one second is removed +Id be happy to submit the change if anyone is interested +it requires a combination of metadatatemporary behaviorcallback behavior all used within the same component +Lets stick to that pattern for +Im not talking about caching in the Digester Im simply talking about keeping a reference to the Method instance in the Rule vs +A README should indicate the location of decoupled rd party products and their licensing files should be clearly located with the filesI dont know of licenses that would require attribution in the main NOTICE but I imagine it depends on how you interpret attribution clauses +We have a features file now +You are indeed correct but I do not yet see how it could do anything else +Resubmitting patch +Internal Yahoo! patch for the issue +You can SELECT from schemakeyspaces schemacolumnfamilies and schemacolumns now +Patch committed to trunkThanks Woody! +Updated readme file parameters cartridge file names default configuration values +is in commonlib as of +Whats the relationship between priority w and priority ZWhich one is higher +The new patch seems not work or involve new deadlock +Specifically they are instances of instead ofI remove the folder from the deploy folder and made sure the proper JSR annotations were included in my deployed jars +I realize I began with my patch which is not appropriate for upstream +The previous commit in this series or this one may have caused this breakage +This issue is not seen with EAP its just a warning? Still works right +We just have to make sure we announce it correctly when we release. +Any comments are welcome +Fix version set to +Thanks Vikram! +Im providing a patch named for trunkThanks to as to trunk as r +Ive committed this and merged to. +Let me play around a bit later today or tomorrow +s does NOT work in my environment +committed. +Remove the comment about checking available since it seems to work +so rather than have a mixed approach to data handling I just avoid using any of the events +When using the for instance you dont need WSDLJ +So far no luck +abdacceeebcbaadf +I really wait for some competition for Maven +Feel free to create new issues +Matt combination of lack of parser support at the time in conjunction with the decision not to use it +Ill post a sample mapping as soon as I can +Fix javadoc can you take a look at this latest patch rev? Weve seen users have a big perf impact due to this bug when their DNS infrastructure isnt well set up with nscd etc +If I remember correctly the serializer was fine but I got a message something like cant find parser provider for JSONRDF when it was asking for a Turtle parserThe tests should still be valid testing for particular strings in the Turtle serialization which will be fine unless the Turtle serializer changes radically +Edward Im sorry I dont understand +Removed code that was marked as deprecated in Spring +Can you fix that +Closed as not reproducible by request of Alexei +So I think for this one the first option is ok +There is slight difference between +Perhaps my issue was fixed simply by updating to the newest trunkI was wrong about the copy constructor being the issue its only called in testsBut I still think this little patch has some value so you can see defined indexes in the shell +Could you suggest a workaround +Attached that fixes file swap issue +Here is some falied tests from the last nightly build +Its something weird between css and chrome +Original problem is fixed +its one of the least performant and scalable +Something like that does not happen in a normal environment +Hi David I think that as well + +Please use the latest CVS code. +Thank you Rohini for your feedback and for accepting my patch +Not good +How should it handle prefixes +F B Can you please provide me with a sample XML schema +I dont have websphere or weblogic to try this outthanks to blocker +test test hanging threadsunable to delete test +Ted Will do +We are creating immutable classes without using Immutable and we would like to use these transformations with their caching behavior +Havent we fixed this +I applied the patch. +The change that changes from public to private cannot be backported easily to previous releases becaase it changes a public API isnt it +If you have a description of what components you were using and steps to reproduce feel free to the issue +Sounds very odd +Here you go Ed +Max means that if we insert for example amp into the source pane +Closing issue as cannot reproduce +This problem should be fixed now +It looks like a problem in the Eclipse platform where the job framework NPEs when accessing some progess info not reproducible in Developer StudioVersion Build id Build date +Thanks Chris +Ok +Thanks Tom! + was failing because it was relying on the block pool to be called Test which is no longer true +Since were choosing to focus this on updates done in the OLAP style work loads isnt going to be a great storage mechanism for the data +I just committed this +patch fix indentationNote No new test code changes to documentation only +If it is not you get an error +All relevent fields are qualified with a specific schema +I wonder if we should refactor the admin service as we did for the features service +I generated UUID times using the initial method and the new method + +Have you or he been able to reproduce it +the client does a lot of the work described but the ui should be redesigned. +Sorry about that +But if thats your take on things thenI didnt know that not submitting a solution means that there is no bug? +It tries to modify srcmainjavaorgapacheactivemqqueue but that file does not exist yet +Ill have a few hours over the weekend to try and get a better patch out +When we are not using a grammar pool the enitties are resolved correctly every time but then parsing takes too much time for our requirements +It also includes a new Filter for the AM that will verify that all requests are coming from the proxy and if not redirect the client to go through the proxy +The fair share scheduler page is not there in looks pretty good +Ill look into integrating your openwire examples nextAnd yeah I think splitting out the STOMP docs into a separate file makes lots of sense +Another iteration with Weight as a protected member or Scorer +after hadoop calls close on an outputformatter would it still potentially the output data? +I might switch to the dev branch of Railo soon +Go for it +Created ticket to bring this to the attention of team +bq +Add proposal to load Console Configuration if it is not loaded +This version addresses the final typos and makes it so the file creation throws an error +I believe bigtop is awaiting the hbase release maybe we can do the same here for the profile + too just sField to port the test +Theres no longer a requirement to have the BEEHIVEHOME environment variable +All the files will be extracted to the destination folder +Manually tested +Ship it +Can you update the testcase to use an index on a compound key so that this case gets coverage +Created a separate issue for Windows with JBT. +bq +switching the code over to CFSQLVARCHAR sorts it under +I will review it +So property key was corrected +updated to autoload NetSFTP and cutting down the startup time down to for Ruby and for +When you add a new entry you increment the value of each schema entity that entry depends on +Committed thanks Nik! +Thanks +endif and ifndef MFPENG +But this will not fix the problem for other users +MichaelIt is the same situation as with +BTW the same algorithm has to be implemented for event in +for the patch disabling the test +Engine issues that prevent the examples from working will be reported and tracked separately +The is tricky here since we clone the MP in IW you need to actually pull and cast the MP from IW to change the setting if you want to do this in realtime +Persistence module are not in the moduleDo you mean and a JPA persistence unit? These can be in the module of courseOr do you mean the driver file? This cannot be in the module indeed but the has access to all the class loaders of its parent EAR +Closing all resolved tickets from or older +I just think it is helpful seeing the attached to the +Added unit test to verify problem fix backported fix to be included in as well +I have no problems with being told a JIRA is filed in the wrong project but it would be nice to be able to correct that and not have to start over againFrom my perspective it doesnt work +However when the schema is being searched for in the entries it will being looked up wit the relative path inside the VDB +The amount of testing that the current code has undergone is not IMO worth upsetting unless were serious about refactoring +Thanks. +Have a look in the attached of the things missing is using the to fetch the jms provider as Adrian was using in his MQ recovery +Ill commit it if I hear no objections checked in is a public method +resetting P open bugs to P pending further transition resolvedinvalid to closedinvalid +create the member for the Item class as multivalued in the Choice class and change the cardinality for the Choice class itself to + +It starts to crack open the door to much more advanced scheduling +Is there any other workarounds for this oneThanks seems that there are many duplicates of this issue or that the issue manifests itself in many different ways +More progress +Id recommend using unless you have a compelling reason to use an older Avro version +The same is true of a number of other mbeans most notably which will require two nodes to appear +Incorporated Konstantins code review comments +part of release. +handled merging region at master recovery properly +Unfortunately the GWC class which is in the integration uses and imports classes from the quota jar so removing it will result in class not found closing all issues that have been in resolved state for more than a month without further comments +trunk Applied patch with thanks to Martin Krasser +Any comments +pushed +This is at least in project of your workspace not the case +But the WTP web service framework is a tool to resolve the tasks of creating web service find create publish and so we realize a new wizard by ourselves we need to do more works for these tasks +Or if you are used to using github fork the repo and issue a pull requestAs for making the option disabled by default that is a wider discussion and while its certainly debatable others would probably disagree that having the option unchecked by default is appropriate since for huge layers full extent calculations are very expensive +Good news! Thanks +patch that makes the jsf part work with updated structureNot committing it since the generated code is not kosher enoughWould be really interesting to make a total rewrite w the following changesno more objectclean up sweep of VTLall managed crud beans have been collapsed into one +Bulk closing stale resolved issues +OK Im not sure I understand what you are thinking yet +on bringing over the parameters +I just committed this +should read +Koen im not sure if we can do anything sensible here without risking deadlocksinconsistent configuration but it might be worth exploring that if forge just completed a set of commands that if this error is present that we run the update project configuration once for the with Fred about feasibility of this +Looks very nice +Deploy both applications and run +It looks like its no big deal for you its a matter of modularity +helps to think of modules as things that can be pulled out of the main project and turned into black boxes that just work +Im not sure the additional complexity would provide sufficient benefit this seems like overkill IMO. +They dont actually harm anything but it makes it easier to compare with the version of the dita file +go go common nav +Lets get it in review queue +Sure Ill update the after Christmas there is already a half working example patch if you want to have a look +Please comment +cloudusage ADD COLUMN virtualsize bigint unsignedalready exist on branches +Compiler side of fix on masterfefbe and eafaReopen if things still look busted or open another bug if theres a different failure +Please reopen if you think that more work is required at this point with specific comments describing this work +Ive checked all the s and we no longer seem to have any javax +Uploading a patch which reintroduces the bailout code in plus adds the repro from this issue to +This test can be verified but the automated test is missing +Yes will do this today +fixed in +since hdfs is an invalid uri +Ill upload to svn when you verify its really okthanks good to me +If you believe its resolving to please check your classpath and make sure you dont have +Thanks +I just checked the documentation from Spring and the example seems to be up to date. +MINA component patch and for these patchesI was hoping that the Mina project much sooner have been able to release Apache Mina which they unfortunately haventSo we are still using the older MinaAnyway I think you patches looks according to the recommended documentation at Mina itself so I will get them into the codebase +Could you attach a test case to thisI have just tracked down the cause of the WARN Duplicate message add attempt rejected +Ok thanks to you Gustavo I understand my errorThe issue occurs only with in this case we have twice netty +No changes were made to the mapred API +if this is still an issue with the current Thrift please and Id be glad to have a look. +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content fine with EAP CR closing +Checking this in +Thanks Hairong! +Let me know if this is what you expected or notThanks + +I try to review and commit in the evening +Review complete +Revision +Let me know if youd like me to test your patch in this environment +That was certainly introduced in and probably makes zone support broken +Thanks Milind! +So far youve waited about hours and had responses from helpful interested parties that are not build adminsYou need to wait longer to allow the volunteer build admins time to look into the issueThere are Jenkins devs on that list +The patch looks reasonable to me but fails to apply cleanly +Committed to trunk and branch +fix committed in r +Yes I am using this in a environment so thread safety is vital for me +Committed to branchx and trunk +Thanks Jimmy for the patch. +What do you not like about exposing BTW +This is the for the feature! +Will commit it later +class or as a +This was mainly addressed by which added better transaction detection logic and removed the need for explicitly having the update flag +Yonik youre right therere separate acceptor threads setting accept just affects how connections are handled when they come in too quickly to be accepted by the available acceptor threads +Handles the case when service methods have for arguments and retrun types Provides a fix for the issue raised in Has testcases that demonstrate handling of SDO Types andIMPORTANT The SDOXSD generation is not complete +This one is a bit tough +simple and xpath expressions? +I think that this is more polish than anything else +That chapter was updated with the new names +Kathey I wont have time until later this evening to look into this but how do you run with collation +Final patch for review. +Can you try adding this to your MVNOPTSXXm XX issue is not easily reproducible +I meant to put the test in but accidentally forgot the test part +Other assembly patterns useful to people could easily be added +Since this is not a regression introduced by by policy it should not be committed into +Any editing button like Add is ok to do even though there is no configuration +I dont think we want the last e in there +Oh another thing +Please review and commit it +Expected release date of Sesame has been set for about weeks time +The first solution would be for the message method in +The real affected versions are and grails +Ive tracked this down and I think I was able to resolve the issueFirst off my testbean was not public as for one reason for the test to failThere is a method retrieving the length of a result of a bean getter in but a real check for the was missing so Ive patched thatUnfortunately this does not fix your findings so I guess this issue should stay open +Smooks editor is no longer being developed +ok! great +Hmm +Jan what else has to be done +Committed +But our application has cookies turned off which means that the url has to contain the jsessionid +But I seem to have gotten ahead of myselfI will prepare a unified patch that covers the affected classes when I complete +Committed patch as part of revision patch using revision +x and +So that for a given jar path we just load it once +Adds security forMoves related configuration params to a separate class +Closed after release. +Create the in the and pass to the so that its only created once and not for each evaluation of the for the patchI polished it a bit +thanks +I can see this too +Closing the issue after verifying the fix with the latest builds. +This patch does not fix or adjust any projects othjre than and +I committed this +To make this more useful we could pass in the version number in the call to this script in beforeBootstrap perhaps reading it from the property in s configuration +Well I cant think of any other special cases that would help outside of so closing this one +Probably would be good to point the users to the same place for consistency +I was just wondering why we keep it enabled by default even though its not output anywhere +Given all the tests I run with different HDFS instances etc I find this restriction very inconvenient +Good point that the absence of a should not prevent the application from starting when is used +Ill update the deployment documentation to include this for the upcoming release + just to set release notes in +I just wanted to say that we should first search for simple ways to add multigraph support transparently on top of what we already have +Mattias I believe the fix should also fix but I havent tried it +Kicking Jenkins +The number of javac warnings from stayed at I dont know why reported there is an increase of javac warnings +Looks like a good way +Integrated to Thanks for the reviews Lars and Sergey +With only entities it takes seconds with it takes minutes and if you have relations between your entities it looks to be even did a try with ROO removing all controllers and views on my entities and it takes about seconds to validate it all +Thanks Dirk! +Thanks Andrew! +Proposed changes to javadoc and tests +Comment should have been just the first partThanks Knut for looking at this +so regexp would be +There should be a way the receiver could filter out the duplicate messages +A new feature +to trunk. +Pretty awesomely it actually works without any code changesI also added a nice log message to mention that were continuing from the middle of an edits log file +OK +I cannot reproduce this at all on Linux Windows or MAC so Ill give this a day or two and then consider it closed +Thanks for the patch! +some minor comments +The applied patch does not increase the total number of javac compiler warnings +Applied patches in r +bq +The XSD also uses the value attribute instead of ref +Thanks for details +You are correct Niall +Committed to and trunk +I will look at committing this patch +Makes me nervous +I think it is better to do the cast to prevent a but in theory as you said it is not necessary because default implements +Bounce to Fred to commentreview +Modifications to default config file that still show the problem running the provided sample code +Since it passes through a Python dict Id have expected duplicates to be dropped silently. +Revisit Jira assignments +i havent looked at the patch but i dont think you need two calls to the sanity checker +Remove now works with optimistic locking even if the node does not exist in the local cache. +resoled I overlooked that it cuts off the jcr +I really likethe forrest project so I think it is time to give something back +To summarize I refactored code to parse the warnings text and build the xml into separate strategies one for the standard doclet and one for the standard doclet +This is the fixed up patch +for the patchPatches using svn mv +Unfortunately there will be more to it than just implementing these methods + +Version just doesnt work for us because we have multiple projects and it really does weird things with project references +Actually abortJob does not exist in +Will it be released on branch +Can you take that part out of this one +Also the precommit build that Jenkins is pointing us to failed so it must not be the correct build +Thanks +In srcclient you will find the client for testcase tries to insert an entity in the database but an attached Hibernate is configured for rejecting every and to throw instead an +Attached v which fixed one unit test by restoring the region plan in case a region was opening +It looks good to me +BTW what JVM +It looks good to me and I think it will be helpful in evaluating whether affects plain GROUP BY statements in any wayCurrently the test uses a mix of three different statements with different characteristics +gdalinfo output on a raster +Committed revision +Note the failing test is running on Windows +AgreeI am changing it to only under with TODO comment to remove this condition once gdata javadocs is fixed +Can you describe the query at least roughly in terms of the number of nodes in the result sets +Of course there are html pages not pages attached +The classloader should not be parented so that the right versions are found +I have spent awhile trying to reproduce this error so far I have not had any success +Alexey Im not sure that localizing strings like Content type is a good idea as they reference the API documentation and localizing them would make it harder to find out the cause for the error +no no i know that it has its own tool but those working on annotations support might need more methods on itand let me know if you are up for making a patch that work with the current changes +evolve the the way I would like to see it support parallel and sequential per my comment above +Committed. +Looks good. +However it will not fix problems where the users are in the wrong group too many groups +I dont believe the makes the guarantee that watcher callbacks are called exactly once +Started re basing for +This doesnt work for people who use the IBM JDK or deal with cases where platforms have custom like the Blackberry +Denis please specify which buildjobbranch youre talking about +Thanks +patch for your comments and continued effortsMike +bq +Hmm looks like removing the old bayes code was a bit too aggressive +I have verified this problems with CA on the last available build +No problem thanks for the fix +Thanks Paul! Could I get you to add the groupIds and to the routing rule please +It is more of an improvement than a bugfixBy silently swallowing interrupted flag hornetq library code destroys the opportunity for application code up on the call stack to act on it. +When i integrate some updates then are getting mapped to a simpleType by really sure if there is something else you wanted to do with this +Yes the openssl c libraries have a max length constant for +Thanks for the addendum Sergey +Agent crashed on him on +I see messages coming for vertexes not present +Thats because a struct is a collection of fields all or some of which may be unset +Hmmm +if we do not let it fallthrough then it will not be registered under the locale and style which may lead to some undesired sideeffectsthis is a more general problem then just the resource laoding +We may get to revisit this some day when things change in the Ruby world but its not likely. +Thats about as far as my thought process has gotten on this one +Please let us know if you believe you can work on this for M +I added them before as I assumed we wanted to fix them +On page example of registry configuration should be formatted better +This issue is as part of a move from Wishes build +Part of a bulk update all resolutions changed to done please review history to original resolution type +I will also try to check sth with this maybe somewhere some additional file would help +After a quick review I think this is definately a step in the right direction and it should be considered for inclusion +accepts which is user oriented +Look for deprecationsAlso can you check to see if your use case is better served by SGD models rather than SVMs? That question has come up several times and I can only say something about the SGD side of the house +Deferring it for now and going for the next artefact +Committed revision Thanks Kay Kay +Problem was only in a fresh install where new is constructed before the agenttoken is setup and put into storage +I plan to commit it soonI still think we need to do something about handle possibly just remove it +I dont really want to maintain an API with JT details etc +Other than that it seems to work greatNow about x faster thanStats for test documents +Do you want me to make patch for you +Ah got it +Running mvn validate on either of the modules works +committed thanks! +this is not the leveldb approachCorrect +Attaching and with these changesM srcgetstartA srcrefM srcrefI removed a couple of sentences from the end of the Reference Manual version of the syntax topic because they applied to other manuals but not to this one +I think we did some digging in when we made this change andno Windows code ever returned ERRORSUCCESS for the cases we caredabout +Need this to fix up the bug where wed see splits before region had successfully opened +Hi Leonardo patch works for me without problems +This will not work in other containers that allow the dynamic cloning of portlet instances +Please fix this +Gabriel ever seen that clear wicket caches button on the left side? +Also fixed the version the JAR was shipping as a SNAPSHOT version +Updated to trunk and continued passing in the default fs +do you mean the could make some other readerIm less worried about that +Merged changes into trunk at revision +Properties file loading was updated Unicode codes should be parsed correctly +design phase complete +Thanks Suresh +I was not aware of this +Most of them we already have +Idea at the prompt propose Choose a number or filterif data entered is not a number then assume its a filter and display newly filtered archetypesthen you dont really need to know Dfilter you can interactively test filtersI didnt if it is really hard to implement or not +Committed in r +if I turn off my master in domain mode happens this error! +Marking minor difference in exception text between jvms consolidating with bugs +HI Andy not from what I know unfortunately +Another argument is that in JON this feature is included as well +This is also a problem in you please attach a modified project that resolves this issue? I can then see about changing the ITD generation +Fixed on trunk +With a little help from Hairong I could write a unit test using the mockito framework to test this out +Using looks to be largely better +Ive added the ASL headers +has a broken expat build +resetting it to default after test +This feels like a Rails issue but Ill investigate a bit deeper. +fixed snapshot deployed. +Updated diff to apply on trunk + +Hey Michelle That sounds good +I cant reproduce this on my Galaxy Nexus which is synced to three accounts +We have realized this bug before and tried to fix it but it seems we still missed some cases +In addition its worth mentioning that theres plenty of alternatives available to use with Castor incl +Its not pretty to look at but it at least would be a correct solution by removing the dependency upon locale altogetherright i definitely have no problem with that from a localization standpoint +Merged to doc branch at revision +for the review guys committed to and trunk. +Then I noticed that in the link above says right truncation was added in which is very recent. +I could probably add blocking methods if there was interest in experimenting with it for +the number of cores +Since I wrote it its heavily weighted toward my style of developing but I dont think its too far outside the mainstream +There is still some inconsistency because the change to the branch is more like a new feature rather than a bug fix +We have started working on the content when we are ready for production will ping back +I will submit it in this weekendThanks SureshI will wait for your new patch soIn the mean time I have just committed a couple of utility methods +Got it fixed in +Im assuming it works when started the normal blessed way +A number of fixes have been applied to CXF code dealing with the HTTP host resolution across multiple releases +The problem started happening again after I added JIRA back in so assumed it was related to it +Ah my bad I see that package should not be exported +This is hot +I have resolved this issue as all points raised have been addressed andor corrected as required +thx for the quick response +All compactions are disabled? Or all major compactions are disabled +Question is should I be using the create script to build an example project and including that in distribution or should we be fixing ant dist to include a proper name +Ill try integrate it +part of release. +Besides why do you want to turn off the feature if your input source uses namespace? You may want to reconsider the usageAgain this is an invalid bug therefore I marked it as resolved +committed with unit tests +Similar to +attached ouput and error of ant ouput and error of ant debug +If you Tab through the entire page focus will eventually get to the Upload button and files can be uploaded +Committed to andThanks for the patch Himanshu +otherwise for the patch +Minor updates to the patch to better align with yarn scripts +patch to address hudson failure +Thanks for the commentsone question about the dailyrolling appender +See for details +Patch for for someone familiar with the security code please review this +Fixed in CVS +Patch for +The icons can also be managed using facets +hm we had this also with the maybe guillaume knows how to get rid of those +I am not sure how you actually implemented it using newLink? +It appears to me that this bug has been fixed. +btw other than making the condition for ending cycles of input split reads that every path on the znode list is marked RESERVED rather than checking also for FINISHED marks this patch is identical to the patch +Would you mind making a patch Mubarak? Thank you +Should we just mark this as a duplicate of in that case +If those files are missing then the application will be able to do everything except for send email and store user profile images +I will leave it to you if you want to keep this defect open until it gets fixed +Commit setprotocolspbcastI think there is still an issue in a simultaneous flush am seeing blockunblock blockunblock like I expect but occasionally I see blockunblockunblock and that screws my software up +both and should be reviewed and committed since I split it apart to make jenkins happy +Will bounce it to TBD then Back to V +Why would a new plugin interface be needed for this? Cant this just be implemented with a wrapper transport like +Fixed the new findbugs warning +The widget became a strange thing to have and confused users +the error occurs on windows too +However it now gives a lass def not found instead of s no java test frameworks found +Brads patch has been applied +I dont think whole require fixes at master is suitable for stable branch. +However for a long time we have recommended using the hibernate Batch specific session +mtcompactionwhy not just log this at startup the way we do some other settingsbq +If late October then it should be done by then +In EWS on Windows this behaves as should until it throws Too Large +I have committed this to and +Yes +He needs to find a JAR that contains the method hes looking for +Thanks Sid! +We construct endpoints at runtime to send a message to the XMPP subsystem +Or please suggest ways to patch this bug +Thanks Laura and Susan! Committed revision +Just tested again with upcomming rule does not match here. +Is somebody have an other way to solve this problem +Thanks Ted +Thanks! + +I have multiple Railo instances serving the same sites on a load balanced set up and I have to do everything twice! Please can we fix this +I committed this +I also tried doing this by wiping out the sourcecache which should be a worse case scenario +btw +Improvements can be tracked in a new issue +Disabling the extension is a way to fix the problem as having the exact same class in two separate jar files can lead to very odd bugs later in life if the files ever diverge +Wed like need more information +Wasnt a huge deal to add a line test case +did you resolve it +This is going to be fixed in +bq +It checks in for axis tomcat and jetty modules +Are we blocking the recompile with instrumentation somehow? How does cobertura normally force the recompile? Or does it even require a recompile +Closing all tickets +There are very useful widgets +However they have not fixed the worst case situation and it is still possible some set of bundles could cause the resolver to go on for a very long time trying to find a solution +Theres just too much to change in too little time +Please reopen if you have a definitive answer. +This problem image was produced by my Fuji camera +So some files have to be moves from the jms directory to the deployjms directory +if we could get and commit that would be nice this has been sitting for a while with nothing blocking it +Thanks ShubhamYour patch in in trunk rev +v fixes how we handle deprecation I had it backwards in the v version +Funny you say that I was just thinking the same thing over night +Now my new implementation of and can each be used in Slide +not in svn rthanks for reporting this issue +minor improvement regarding +Thanks Gregory +Thanks for providing the sample code +odd since I used Mac to test it Will check +This one turned out to be simple for once +Outdated non mandatory stuff really +Attached the patch +This may be used in order to allow one component to continue a conversation that had been started by anotherFrom the above if the service is Remote then it must be in a separate thread +and paste the response +IasAm not sure + +Rails app fix to paginate friend cloud needed. +A Sonar plugin can be developed to cover this feature. +Hmm I do not understand Please do NOT add these patches as it will affect other Forrest instances using the Dispatcher +Btw is configured in such a way that the reporter of a message shoud get a short notification on any update of the issue at question +Please verify it +Is this something you are planning to try out in the next little whileIll continue investigating this myself as best I can but Im still pretty new to SAML so it takes a while for me to dig into these specs figure out what is going on compared to what is supposed to be going on +It looks like v is against can you attach a patch against trunk +esb archive with a new version of does not work for me +In the form of code You may want to have a discussion with Anil on Gary Ive sent an along to Anil to kick off this conversation and will try to capture the results in JIRA +Updated Chpater with the new data and screenshots pertaining to the new Embedded JOPR Core console +Any chance you could describe how we can reproduce this issue +The original is then used to check the results. +batch transition resolvedfixed bugs to closedfixed +Here is test only from the patch +what about caching of pathsA similar solution would IMHO nicely cover path handling +Eh normally in the GUI and REST config thats why Im refusing to apply the patch as is its unusable for the normal user +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I did actually run through this with Bill and Linda and Jason Greene and there isnt really any extra integration to do apart from deprecate the platforms annotation. +Instead of generic LDAP servers just come out and say schemas +After the change in this JIRA +Jan can you verify this fix in the ER build +didnt get this in latest code drop +Also is this patch read for review? If so you can set the JIRAs status to Patch Available +I ran and myself +If youre not interested in helping then this will probably have to wait until someone else who needs it comes along and is willing to make it happen +I think you are correct but Im not sure what the Enable action is supposed to do either +Disallows the use of an identifier which ends with a +Ive attached the logs and the metadata files +AFAIK they need to be defined on apache virtual host level +This would still require the code which merges the perdirconfigs to give precedence to a configdir in a parent context when it is an absolute pathname and the current isnt +If dsnitch actually worked then I think this would be fine but fixing dsnitch is probably out of scope here so thanks +I know how to set up as a report at least so that it can generate something like these results on demand but not go so far as to fail the build +If using a Rule Timeout Id suggest additional things Have a annotation as well to be able to set a timeout other than the default for particular test method +in +This patch does a better progress reporting +So even if we move content handling out of band we still need it to work reliably and correctly +resolving as a cant reproduce now that trunk is on Maven +The differencesbetween this patch and the previous patch are as follows Added the new field now includes handling for DROPCOLUMN and DROPCOLUMNRESTRICT if RESTRICT is specified a dependent view causes the DROP COLUMN to fail if RESTRICT is not specified a dependent view is automatically dropped when the column is dropped +Thank you Xi +note there is only one other place these jars could be and that is as well known stable Maven GAV identified artifacts +It is set when a fix is committed to a branch with the corresponding release numberIf you intend it for certain releases add those release numbers to Target Versions field +moving this one to where well have time to get it right +passes +Good question +Steve would you take a look at this and please +This would have the advantage of not forcing incompatible changes to the users +Verified in +Hadrian if you have the credentials then could you keep and the release only +It should actually be showing a rolling average +Replaced a poorly sedd patch. +Turns out setting timestamp for DELETE wasnt working at all anyway because the parser wasnt allowing it +I have corrected the test cases and the patch is based on code +Rob did you verify this with GA or some other branch +CSN arent consecutive you may insert a CSN between two existing onesBy consecutive I mean you can immediately know if a CSN follows the preceding one or not +Wikipedia article about URI and related links are explored +upgraded as this will trip up a lot of in +in +This issue was migrated to +Our editor must respect characters in souces as much as possible +are you sure youre using the latest version +seems like all the extra copies will be on the same node rightso if it dies so does the filesystemperhaps the bug should be cloned with the patch resolving a bug but leaving the current bug open until its more fully addressed +Shipment Plan is one of the tabs of the Shipment screensNo new entities are involved +I was trying to just allow the same way of version delete as from the normal client side +But moving test classes to src package isnt acceptable solution to fix this problem +I also want to fix the edits unit test after we merge since historically checking in the new binary file has been tricky to get right via patch +Iterative method is one way to fix this problem +I did this recently but I dont remember where I wrote it down +never heard back from Hugo. +Another potential Mule contributor asked me about how to gain contrib access as well but I have been unable to find this information who to bug where to register +The content of the jar doesnt matter the name however does if its in a maven repo +Ive committed a whole bunch more javadocs and aTheres still a big nocommit in but other than that were looking good +im working on mojo to do this +change in bugfix as +Tested +When we like to support class based mappings it would not be enough to specify a filename like +Ill keep trying but it seems that fix is working +Move under hdfsproxytestSorry I meant hdfsproxysrctest +Committed as r +HiWhen I run this the job finishes but no documents are indexed +Committed this to trunk and +done in trunk in +when did you do thisthanksdimsNow the Codegen generates the stub with out any issues +It has the following changes +batch transition resolvedfixed bugs to closedfixed +The situation for relation is similar to n relations. +Thanks for looking at this Todd and ATM +What were invisioning is a way to slot Infinispan between and our app +Implemented please in Developer Studio +However the dependencies of the subjects are not added into the mix so those symbols will indeed not be potential approach to this would be walk through the dependencies of the subprojects and add those into the classpath of the report but that probably will break on windows since there is a max length of the commandline and this plugin builds out a plexus commandline object which would have huge potential if this were a larger project using the might be to get this plugin to not be using the commandline? have to look into it I ll attach the patch in a moment pending commit until svn is back up though +This is about being able to force things in without that wait and irrespective of I get it correctly what you want is use Yarn to run a process in a particular host completely outside of the resource pool +Is it ok if we delay it to another JIRA +I see that you enabled it only for AS +The designer kind of locks up for while once I do that +Unfortunately I have to configure it all over again +Would that be okAlso a specific implementation should be a javabean so that its possible to generate an editor for it on the fly +Im putting this here for discussion +It seemed to work when I applied it to a new checkout locally +Without it I cannot take the patch +Now that work is well underway with AS all previous community releases are +I am not seeing this in +Committed to trunk thanks Marek +Fixed works perfectlyThanks a lot Vladimir. + +revised patch adds c docs to +Hi of interest what did you do to fix this +Attached patch fixes the problem +worth mentioning here is that currently it seems there are case sensitivity issues with vlt jcr is case sensitive while a file system typically is not +Is this something that could also apply to +But if the tests run fine on your backport to codeline please feel free to commit them +You re right all our use cases are administrative tasksBy introducing this feature most of these troublesomes tasks can be simplified or avoided at allCreation of workspace is also admistrative task and can be done manually +Inside Yahoo! we have recommended the use of BBWC raid controllers +ThoughtsSounds good Ive made a first pass at this +So I will update the patch +not minor it greatly reduces number of amount of clicks when wizard does this properly +Are you sure its because of nekohtml and xerces? Can you provide a stracktrace +This is because the size tiered approach could combine an old sstable with a new sstable which renders the sstable to look like it is quite new +Thank you +Reducing priority since there is a way to get this info now +I should back out regionserver reopening bottom half or I should set the +In this case it looks like we have been a bit to ambitious when it comes to error handling +It also appears that the behavior is such that when you do the mvn clean verify the first time the plugin compiles and the integration test does not +We need this bad but not being worked on currently that I know of +Bug I commented isnt related to this issue creating new one closing this issue. +Hi I fixed class to fix the problem reported by Elder +Can someone whos seen this bug please verify whether this fixes things +I trust the lib to sanitize correctly the parameters but I am afraid that my function will be used in other context +Since we are no longer need to push this to +Maybe let Sailaja retest upgrade on a clean setup to get a clear picture of what the upgrade issues are +Will double check and close this issueThanks. +Hi Kaustubhif you read on for a few sentences you will find thisRedelivered messages do not have to be delivered in exactly their original delivery order + looks good +Doesnt make sense +Any objections +Hi JackIll look in this issueThanks and Regards JackCan you please let us know how to reproduce the error which made the wizard to exit in the first place? I tried various things to produce an error including adding an external jar and removing it before the archive generation +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Thanks Mark +If you would prefer to leave the JCS Section as is please leave a comment about it +The attached patch removes XXfalse VM argument settingIt also corrects the properties access and report results name the errors found by Alexey Varlamov +This has nothing to do with the CLI not sure where it gets the template +Closing issue. +Youre right I could do something to make the headers a little easier to access +Next time you have to provide all the contents of the directory in the easiest thing for us is two bundles each with one of the jars each with the same pom +However these same fields would be useful when specifying input and output directories in command lines etc +Have you perhaps have an idea how this came to beRegards PierreActually its not strange +reopen with more info if it is +Shouldnt be too hard though +Thats the version I need +Moving of CR since seen work by others Snjezana some other things need to be installedI guess meclipse should be installed and activated? meclipse Extensions too +This could be related to that +Verified in CR +Fixed default selected plan is the closest one option Unplanned added button Unplan addedThe shortcut on the closest plan is not available yet +Wayne is this correct +When another client is holding the lease I do not think we should allow create with overwritetrue to delete the fileWhy not? Doesnt overwrite imply that you wish to delete prior to creating a new file? In the same way that we allow deleting an open file I think we should allow the atomic that overwrite implies +I just got an update to the latest plugins and the problem came back now have that version of has the same bug that was already fixed generated output goes to targetsitecobertura while in the stagingDirectory specified with sitestage only an empty is am struggeling with this bug in at least different for over years now +Thanks for the contribution +Sure the question is if we can somehow consolidate and ICGC into the same class +Not sure if this changed in this jira but YARNAPPLICATIONCLASSPATH shouldnt need to go into the classpath when running against a Mini cluster +Reassigning to myself per update from Namit. +I confirm that works fine with when not used +In this case this attribute must be specified with a unique name for the reference to avoid an error +File is still needed to create global EL variable dialog +I have some local code that actually fits a bit better with the interrupt model and it might be useful +This should fix the issue for any Maven based projects that we rebuild +Sorry what change did you do +Closing all resolved issues for. +Oops sorry +And congratulations on your first Hive commit +Are you going to submit this patch +component fileClosing and reopening the workspace showed the same content in both pages andFred D I can work on this now +It has been reported in FUSE MRWillem is looking into it +Tests pass but Im afraid there could be a number of possible problems in complex scenarios. +OK +I dont manage to reproduce the issue with IE and default dashboard +for putting this file into the componentBut I think the user should write the himself to load the camel context and setup the servlet +If we do ref counting cant we still finish the compaction but not archive the files as long as there are scanners against them +Compared to version the last pannel is very slow also on Windows +Unassigned issues already received Chriss privately so I will commit shortly +I think you may have put your finger on the problem +OK I think I understand it now +I hope this will clear the confusion. +Attachment has been added with description Fo P open bugs to P pending further priority for bugs with a status from ASSIGNED to NEW for consistency +That is one of the main purposes of a cluster right? +Heres my proposalThe current x patch goes in largely as it is now +Thanks +caused conflicts with this patch during development +If the user enters text into the field then clicks the popup button we should then parse the text field and raise the date popupIm trying to find a library with a compatible license that does all that +It looks like I am in the minority here but still +I think it is getting stdin +Pull request has been merged in time for AS +Deferred to as usual makes good sense +is ti from groovy or Java? I get now the impression that it is java telling this +Closing old resolved issues +And because the timeout is hard coded to s the WS call fails +This requires more tests so move to +I agree that there could be better coordination with shared files between the ruby and jruby FFI projects but in the meantime heres a patch that applies cleanly to the jruby repository +Alan Thank you for your help +Weird +Can we do that? Thanks a lot +I have put together an enhancement which resolves this issue +Martin So specifically the issue is that on Mac using the browse window functions properly but it appears behind the workbench? is that accurate +The nice thing is that you can do all of the parameter mapping and extension you want now and in the future without having to anything and making the switch is also fairly trivial +build the latest Harmony jre +Verified in Tools. +I add this interaction to avoid the timeout effects +Were not maintaining the branch +Could you attach the spreadsheet that cause this? spreadsheet has been attached. +Have a look at and see if clearing your session cookie after reloading changes helps +Updated patch with correct visibility on get method +It seems that the problem is with libtool +And the mailet could be called for any mail being in the spool +The change is in the logic that iterates over the nodes and the query +If you allow it the page renders correctlyIs this something we can fix? Or is it expected +Bob agrees to this change and apologizes profusely for the mess of licensing ERH is doing a great job of fixing +It may even be a case of having to resort to VM magic to keep the type system would be nice to JAPI to see the result with the low hanging fruit gone +Thanks Tim +bulk defer issues to +I guess we could document somewhere that you shouldnt perform ALTER TABLE DROP if you use timestamps in a way in that CF +Dirk sounds like your domain +It also updates the order total inline as a result of this +It would be wrong to force them to deal with it +Further when sending the response we depend on the requestObj at multiple places in update one of them being the requirement to log the complete request +But if only one single exists on the server with the blob field and using Oracle the server throws the same exception for the other too +JIRA for the filesystem issue is +Patch introduce a new problem See +and to constructors in and +So hitting ENTER wont carry out the action by default +Updated patch addressing the review comments +UI bug +Anyone else has some opinions here +bq +I added +also addresses part of namely the ability to collapse similar or identical warnings into a few warnings with a min boundary on repetition of the same warning +Can you please attach itThanks I will take care of it thanks +Feedback welcome + +Does this effect? Or is it a problem with the new faster indexes? +Thanks Corinne. +Now at at +Committed change so that setUp also used a connection managerHopefully that will solve the issue +Actually that is what we do for nowThanks +For macro Lucene has a nice benchmark toolkit that is pretty flexiblepluggable that we maybe could leverage +At revision I then successfully applied the patch with and built successfully +My apologies. +Now the text isnt rendered as intendedThanks for the given example contains an inverted imagemap which has to be taken into amount when calculating the colormodel +I can read but not write +I am unable to replicate this it seems to be working fine from my machine with Whirr trunk +fixed in revision not apply to trunk +Thanks for the review Andrew and Anoop. +For many of the defined it seems to make sense to include some more information like which job and which task are affectedbq +I like your idea +fixed in current SVN +Committed to trunk. +Well I dont know + new constructorsinitialCapacity and actual List to use +JBDS worked this way from the beginningIf you think we should change this behavior please create a new jira + +However for symlinks to directories get isnt applicable since directory walkers are going to rely on the status returned from listStatus rather than doing another get on each of the results from listStatus +I committed some ground work in rev and all props except tcpnodelay could now be configured at a broker level +Please review this +Run both these tests on both targets and the failures are diagnostics for this junit test so Ill mark this issue as resolved. +The userSelect and groupSelect SQL statements can have a predicate with the username parameter to be set in at runtime +I am giving my because this makes code much cleanerI havent done any performance benchmarks because I agree with Juliens assessment +Could you please fix this withDer Wert fr das Feld s muss mindestens d Zeichen lang seinDer Wert fr das Feld s darf maximal d Zeichen lang sein +Did anyone figure out work around for logging out issue +Moving to resolved state so that issue can still be edited. +A few observations You really really dont want the user configuring the number of tasks to combine +reassign to maz for verification and closing +And length is nice but not the most important +incorrect test of a bulk update all resolutions changed to done please review history to original resolution type +That handle will be logically closed first then the underlying physical Connection will be closed +Im removing the JPA component but this task looks like it still might be useful + Fix C union issueDefault field in an union should be treated as if its optional +patch for contribpig +Fixed in r. +Wheredid you find it refered to in the documentation +Will show up next time we push the site. +Thank you a lot +Yes this patch has been applied +Closing stale resolved issues +What about Ryans query about how you are testing your patch and what you are seeing? Thanks +This information can also be provided through Thrift +Added connection profile and changed the Teiid importer to allow for sources +This issue has not been commented on in a while +the offset screws up or something in the readIll make unit tests though +Its not the defaultso it needs to be configured explicitly withIn order to ignore the order of nodes in the members list I made Addressimplement Comparable the Address was already Comparable +Existing jira is +MartinGood idea +No I only use Windows when I have to so no DLL hell here +Update and patch for several days now to branch and trunk +So we can store the shuffle address also there thats the part of the code that you may want to look atWe can definitely separate out the fixes into another JIRA issue if that helps +The call to sh could now have a spec but Im not sure what the best way to do it is because its called with command and command is a protected function that the specs dont have access to +looks good will commit if the tests +Patch reviewed and tested under load +You can easily verify in the current version of that and will not be included in the installer and uninstaller because the patch of the zipfilesetincludes for. +In that sense this JIRA simply made sure that the feature in code is documented in Forrest +Any references will get evaluated as expected by the access node. +Do you care to work on that too +So Im using that instead. +Since their most likely pooled it makes it a bit more tricky +We can start with s for small tests s for big tests if you agree +Speaking of CQL versioning created + Username pilhuhn +As far as I can tell all initial committers are existing Apache committers +The code implements the specification and Harmonys results arent surprising +Any more specific analysis? In particular if you ran into something thats trivially fixed I can apply it but otherwise Im inclined to close this as Wont Fix and would recommend moving to +Setting component to the appropriate plugin Android as this issue seems to be specific to Android +And just for ref MR details can be drilled down from via. +because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +If you post it to the search or branch it returns not found +srcmainresourcesorgapachearchivainnote fr is empty +Perhaps we should still bump the version to and include a note about this change in the next release though +Here is a simple patch that removes the caching +Denny is this fixed +fieldName description notNull sizeMax add field reference jpa class +This diff contains the added feature mentioned in this issue as well as updated documentation +About the plugin discussion I couldnt get a sense if you folks are more inclined towards having the auditor as a separate process or not +bulk defer of issues to +Can you reproduce on trunk? This sounds like a duplicate of +I had to move back to the editor I have confirmed it works +ZhengIt would involved some logic out of the +Heres the initial drop +Great work TomAnd very little impact on the code in the parser +Could you please apply your changes to the trunk instead +Run the testcase times and all of them passed +Aaron there are number of places when Exception is caught and logged into the Eclipse log and it is unclear to me what path you are concerned about +In reponse to your general comment aboveI dont share your lack of faith in Abdera +Thanks s silence interpreted as a verify +But than when I uploaded the diff +Am I right about that +Basically associating the Remote Application Launch with the current project your in and launch it +This appears to have been fixed now. +Please with the latest version of Traffic Server + diff to change the fully qualified name of the class +Trivial patch to fix the m going to commit this without a review since its trivial and fixes the broken compilation +Duplicate of this may be the case +The purpose is actually only partially the maybe higher speed you can directly see in a microbenchmark +That sucks +Is there another place to define it +Visions of Schrodingers cat come to mindApologies that it took so long for me to follow up +Oops +AaronEvgeny Yeah I revise what I said +Thanks. +eclipsekeyring showlocationENTRY MESSAGE System property has been set to local +Can have a detailed look for all scenarios +Most of the namespace code has been reworked to remove bean decorators and the use of for wiring up things like the filter chain so a realistic portrayal of the final bean configuration can be achieved by the tooling environment. +ok ive changed the rendering of the result to put? in this specific case. +Thanks JeffAs I see in our environment is being used not nextListener +I dont have a need for it on the client side +That may be just a safety first commit adds an improvement to absolute positioning using Teiid JDBC +All came up balanced after a little while +tests included +I just got the guessnumber demo from Suns portlet bridge to work with eXo +We were using in PHP and ran into the bytes error +That material only makes sense to me if it is preceded by the explanation which appears under Equivalent to +I tested the mailbox with and everything is working fine for POPHowever exactly the same database data and client conf gives timeout with M +Watch out for todays regression runs +I agree on MALLOCARENAMAX too +Thanks for reviews and Ted and StackCommitted to trunk +I will take a look how easy would it be to run both stacks. +In English we express this concept verbally to want to cryweep +not much we can do about issues. +bq +So it could be wrongand I wouldnt know it without doing more than I have time for + +Please? +I just checked and no more NPE we can do jaas realm karaf without problem. +It would not cancel +Cleaned up the patch to remove all nocommits and improve the javadocs a bit +FWIW I have a question in to our guy since we had and resolved the same problem I just dont know howId mark it as not a problem since its a container issue but if theres a way it can be avoided in S we should look in to it +Screenshot that illustrates the at other features in eclipse none of the source features were nested inside its functional feature +Resolving this issue and raising for the remaining work. +and I also provided a test case which currently fails +Here is a patch to address the issue +I think may be related as well +ones that are not in core Lucene it is undefined what the filter will do with those Attributes +Sorry if its ugly I dont blame you for a few dry heaves +It creates cores datadirs threadsIf this seems worthwhile I can port it to +In fact since any replacement is likely to be a BIG chunk of untested code Id actually really prefer to keep for + +with JBT. +I would like to have as much as possible done by ant so that Release Manager do not have to deal with everything manuallyI know that were going to abandon ant in sqoop and thats why Im saying easily modifiedJarcec +Derbys charter is to follow standards to allow portability of database applications to migrate to other databases +Do we need to update the Release Notes? Or? Where is this +Thanks Hairong +try and use this on a seam generated project where you have srcaction and src action shows up as a package +Once when going from no setting to a or and then again when the user clicks on or +The power and responsibilities are in the hand of the Hadoop PMC +Yonik strongly suggest you open a new issue to address this since is already listed as a feature added inif you fix this mid string slash issue youre going to want a unique jira id to refer to when citing the bug fix as a for or no one will have any clear idea what works where +Patches v now works with Mono on looks goodA couple of comments In there is a todo comment in Process +Unit test added to testsuite at +scheme check was removed +With the configured render strategy ONEPASSRENDER the page is reported as stateless in the testcase and when rendered in the browser +Adding new security realms would be managed by the admin console or CLI +Thanks any reason why we need final application status and the tracker url in the report? arent these available in the overall application report? what is meant to be retrieved from the attempt report as compared to the app report +Stack trace is not wholly preserved though +Changed element to type as well as updated sample and reference documentation in Git ID good but would it make sense to leave entity in as well in case theres people who depend on that in other scripts? Is there a way of having an option enabled but not showing in the list +Theres a minor conflict applying it to +My contact info is somewhere in the forum thread +I will be investigating this tomorrow I have seen a similar deadlock when using confirmation and failures +Enunciates GWT compilation mechanism is deprecated in favor of the +In the modified jdom user will need to run ant ant jar to create the instruemented jdom jar +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +It typically happens indirectlyA imports B imports CA imports D imports s an old problem in C the ifndef directive was used in similar have opened a related issue which I think would make the log messages a lot easierto interpret in these situations. +The test failure seems to be due to Hudson being flaky because the test passes on my laptop too +for the latest patch +This was an test +To avoid any potential problems with the release leave changes until after this release ships +I will do that asap but leave thetest under srctestsjdotest and integrate it at bug +For certain use cases being able to fail a region over to a host that already has data locality for the store files is going to be criticalOnline serving cases in general will notice the increased latency of remote reads +bulk close for +Thanks Rob! I committed this to trunk and. +It seems a little odd as the expected default output +That was my main problem for trying to making it work +Therefore this patch would go to their project. +bq +Compatability with such a packaging should be supported +I agree that expanding the notion of interceptors in a more general purpose sense for services may be a good feature +fixed in resolved issues. +dont we cover JDG here +And it cant work actually +I was asserting the object returned from a method was the same exact object everytime +Fixed in the latest versions +Its now MBThanks Devaraj! +You can just grab Forge generate an Java EE app and you have an example right there. +I will give it another go in a day or two perhapsThanksRaj +Ah? Can you recompile XRXYou will have two versions of the runtime +Jeff Im still trying to understand the use case you have in mindMost folks writing data to files should use an Avro data file which includes the schema +You are not using Apache Xerces at all +I submit that this should be moved from major to critical +It is normally much faster since heaps of people have authority to do this +When node checks in saying that something can now be scheduled on it the scheduler expects to find a reservation with priority Y on node but doesnt so NPEs +I just committed the testcase. +Fixed as part of release +Thanks TimNamit thank you so much Tim +Could you please relate as to how did this NPE occur so that I can reproduce it on my system +I assume the POM changes were necessary for getting the user guide to build? If not perhaps those changes are better separated out +Should bitThanks for looking at the patch +Fix coming up +Updated also the error message about missing jflex when calling ant jflex to regenerate the lexers +Also added an assume when the locale breaks the test until can be fixesTrunk rbranchx r +Need new file format to be able to do stuff like limit indices or to store them in soft references +question whats the compatibility path for all applications written with Writable +I wonder where why there are test errors +Patch looks pretty good to me Todd +why on file? why not on +Now removes some of the limitations mentioned earlier at the expense of keeping some String sets in memory +If we cant fix this soon I would like to move this out to +and by the way you wont be able to commit to the code anymore +Lucas uses now the suggested API of Lucene. +Updated patch with Nehas suggestions +Iterators would then use this to decide when a column has a value +I believe we need to close this bug as as duplicate +Happy to leave this specific one closed and leave it as an architectural goal for +If it was to throw an error it should be meaningful like uom is not set +bq +added handling for system property +Click Subscribe +I believe this will be a configuration issue on the Window box +Thanks for the quick feedback! I have added range checking in the setSlice method to throw an if it is guaranteed not to return any values +Update to quickstart to fix in ESB codebase will be in next merge +The default cursor name still prints +Applied +brett I dont see any + +May be you could add me to developer list in JIRA to be able assign tasks to me? I think it wont be practical until you can commit code +perhaps putting the daughters into the same row adds some transactional benefits that we didnt previously haveIndeed +In theory yes +We need to provide guidance for user to use both functions +The patch resolves all compiler warnings +I believe the plan is to merge this to after and then have been merged there +The example uses uxxxxx unicode format which does not work +BTW if the openjpa or the jdbcdriver do not timeout the query and the command does not complete the lock should be held +I didnt realize it was possible to do it that way but it would probably make the most sensePerhaps you could change this ticket to be an open ticket with the subject Create a custom module to emit Apigee metadata into WADL files? +Attached a patch on top of branch +patch does not include and classes +To KelvinYeah +Reopen this issue for the potential update caused by patch by Devin at +Please attach a runnable quickstart thanks +Current code is implemented such that authorizationID of the owner is recorded in system catalog for system schemas regardless of the value of property +Most of the tables have multiple columns and it makes sense to provide information about them +mvn test P localTests DtestTestMeta should fix this right +Pig has schema +I created for the issue where the master didnt do anything about the region being closed at the end +The record in in trunk should belong to section as it does inAlso this ticket should be market as fixed in release not only Cos for the reminder +I was just wishing for this in the NEWS and CHANGES files myself +Still needs for security but this patch passes mvn clean verify +committed to branch at revision +Committed this to trunk and +Closing this as per dev Comments + +context simulator of a device reports an error +Andrew that would make sense if there were several other breaking changes at the same time in which case this one would be trivialnoiseFor my own understanding are there other breaking changes other than the in each plugin repoFor an app developer with a proprietary plugin that isnt at a hosted url is there anything they need to do besides update the import path? This is the scenario Im looking at +bq +Ignore my last comment. +Ive always thought this was a cool idea because someone is not always around to demo this functionality on the backend +Thanks +It would be great if you could verify with tomorrows snapshots +Ill leave this issue open for the backport to! How join would work in distributed mode that would be very useful for projects +Patch against windows jvm equivalent is and has the process handle in the private long variable handleI suspect having an equivalent for this would be straight forward but I dont currently develop under windows so I didnt try to implement an equivalentThe code uses reflection right now to get at the private variables definitly fragile +Applied patch +I also deprecated and replaced it by. +Monitoring the state machine becomes much simplerIn general both approaches are doable but we need to weigh the cost of patching the code to make it work VS choosing a simple design that will be easy to maintain and monitorI would like to see a discussion on this fundamental design choice before jumping to code and patches on and +Attachment has been added with description patch against svn of has been added with description image results from cubes green area in is has been added with description image results from normalized cubes green area in is bright +Would you mind posting the code where you have this issue +Admittedly its better than a totalA combination of host ugi would be a better identifier for throttling but thats infeasible because a client can by simply spamming socket opens or even just socket opencloses in which case the ugi isnt even known yet +Im looking for a repeatable data management design that doesnt bring all the fun of Admining either a high availablity RDMS or distributed store into the mixOther approaches might be to hack up derbe or or postgres but all of these bring more bagage since thet are not already HDFS native +Duplicate with +The patch to apply is +to the patch +Also thread leaks were removed and a log to dump the current system threads was added in case an OOM happens. +resolved the issue +Now the identification and loading of Tiles definitions files are done in +Igansithanks for the responseso i have sent the Individual CLA by email and i want to start working on the fixi have a fork of the jclouds repo in my account but i am wondering which branch should step out from? master? usually open the pull requests against master and then to the branches for the released versions +As Karl statet the current implementation is according to the spec however it is very slow in delivering synchronous events as they are put in in a single queue +For the time being well follow Carlos road and give each module its own client assembly to be brought in as dependencies of plugin. +the fix looks good to me +Removing references to shouldnt impact D coordinates +retriggering +I believe this was supposed to be addressed by Teiid designer +I presume the other spurious repos dont need build jobs and we arent planning to release anything older than. +Usually we name patches w a version going forward +Adresses Joeys nit +orgmortbayutil matchesBinary file +use padding and set margin to Works for me see illustration +Closing issue since there have been no complaints since last fix. +Now we update model only when we switch to preview tab click apply or ok button +So all issues not directly assigned to an AS release are being closed +I plan to see if there is a way to share logic for the cases where we temporarily set the current schemaback at a minimum to make naming uniform +I would appreciate if a committer reviewed them and gave me some feedback before checking in +Your patch was applied in revision +consider nodes ab and abb +thanks pat. +this is a debug output from a node with this NPE happening around the same time +Here is a new patch I have run GB file test included in the previous patch +Can dev team take this JIRA on high priority? Thanks +Now that work is well underway with AS all previous community releases are +Should we also remove organization column from committers table? This was discussed some time agoIf too radical could you please clear the field for me +committedOleg +if its not please reopen with a correct quickstart. +Got a patch +The Derby connector can access Geronimo thread pools via the +Need to change the fixed version value. + Mailing lists done as per SVN folder already created added the podling to the auth templateNext jiraAlan can you please clarify the comments about CI and hardware +tested +In lieu of that this patch looks fine to me and will cause the least impact +Containers can pool threads +On a test grid with map slots the first user in running in the default queue only got map slots +Yes the TT will be blacklisted and it is sometimes unlikely that the TT will continue its normal computation +If it is not resolved by Due Date its out +The posted patch makes the test pass +This is the original fix that introduced looking up the in the +cf cfcf +Committed revision +This issue has a working hack patch now +If it is set to false then the header is populated with the action +I made some changes to and files +Please close if you are satisfied +to Eddie for review with SVN +reduce task settings we have to not run the job as uber or have some hybrid where we run the maps within the AM and the reduces as separate tasks +patch along with checked in at revision +Especially exceptions arent easy to fix if there isnt a recursion issue or something like that +This isnt a showstopper since Maven is recognizing the profile +The primary motivation for making the change to the schema in was to avoid storing column statistics fields as a BLOB +fix the error message + pending you did not touch the relevant code my guess is that trunk fixed some findbugs warnings and the change didnt get ported to our branch +I have made changes to make pick up all xml files in the confspring directory and carry out configuration and to to watch for all xml files in the same directory for changes +This patch modifies the class to makesure that the paths for the tiles def files are absolute from thewebapp rootThe patch also contains tests for inheritance and nested page flowsthat use tiles +I should add one more note following up on Gails hint regardingDisabling null checking just allows the save operation to get all the way to the insert operation before a is thrown which still occurs since the required attribute is being nulled out by +update the server trunkplugins to use the external war changes rev geronimo patched rev use the new war remove unused jars from svn repo remove unused war projectThe dojo trees in the console viewer portlets look like they basically work but like some images are missing the trees dont seem to have the symbols at the start of each line +If you want a thats not unique it sounds like youre describing +Resolved in both master and and tested on my local system. +Is this done? If so can we close this +Tested the patch locally +Moving all unassigned bugs out to +I side with Greg here +for Hibernate Tools beta which uses all plugins directly from WTP rather then from the Hibernate Tools archive +Im satisfied with my fix for this and will make a mental note to add better specs soon. +Just a couple of typos to fix then looks good to me thanks for updating +Changes with site +Just cd into that dir and do a forrest siteNote that Forrest Run also now has problems which Ill bring up on list +The code must be enabled by setting the property to DEBUG +KeithI removed and from excludes for profile and uploaded a new patchThanks for spotting thisKind Regards applied. +Patch applied. +Thanks for the patch applied at rWould you mind to verify it +So far everything looks great +If you could take a look at that issue that would great +Sorry the testing showed my error in the exclude file +add prev to both methods when crawl status is equal to retry and gone in class +Wont fix this EIP issue +i see +Or would that be overkill? If there isnt any interest in the codec idea Ill add take the fork option and add the jfastlz code to the patch. +Thanks for the bug report +generating WSDL dynamically at runtime +done +Uploaded a patch that adds some live data to the network server Active sessions Waiting sessions No of connection threads No of accepted connections Bytes sent Bytes received + +Patch to allow the full lexical space of all numerics +here is the new using my patch for a while I have found that using the setting the to is what fixed the issue and not changing the to consume +Veryfied are +I see the point of the issue but do we want to make such changes between RC and final release? I mean it is quite notable change for user API +Yes thats what I meant +What are we going to do for this JIRAWe wont be able to release until we resolve thisThanks as critical we will definitely fix this for next release +Max John are there other magical property valuename pairs that should trigger an +Client Side Validation supposed similar functionality perform whole form validation before submit +Ill let you know what we find +Ive just build the standad binary distribution snspshot from the current code baseHovever the library is missingthe class fileIn containted that class file so it seems to have been removedI checked the current SVN codebase and indeed the file is removed there +Fixed in r +Sounds good +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Resubmitting the same patch because I could not connect it with the failures in the tests +KB Time ms +Thanks again for the patch! +Ok great! Thanks for the clarification +Why not leave alone and manage execution plans with another job typePrior to I was using getSteps to present real time and last run status to users by iterating over the steps retrieved via getSteps +Sorry about the unit test failure I had it all working yesterday Ill findfix the problemIn terms of fixing the Script API +This change make sense +The fix should have been pulled in the last rebase +apologies for email storm created by multiple edits to my prev comment + +bq +Changes were made to +Im working on comitting this javaclientorgapachederbyclientamSending javaclientorgapachederbyclientamSending javaengineorgapachederbyimpljdbcSending javaengineorgapachederbyimpljdbcSending javaengineorgapachederbylocSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiTransmitting file dataCommitted revision . +No we wont apply the patch without unit tests +Thanks for the JIRA . +This is expected +this is currently handled in which is nearly done now. +not even any need to change it to use the System property +The only thing a needs to function is a pair and those were days ago I cant think of what else would cause this to fail at the moment +Perhaps soon we confirm its not JBDS Ill get in touch with engineers +on closing reproducable IE not supported anymore +However they are now displayed fine in firefox but maybe some caching in firefox is going on so this jira still needs to be fixed +Is this something for Groovy ? +The error comes from the date data which is sensitive to the time zone +The existing functionality is a whitelist with no wildcards +Same issue for me +Thanks Bryan! +code +HiIs this still an issue? Id like to work on it if so but it looks like all the info can be gained from the changes feed? it hasnt been resolved +If a hive gets compiled with and then gets deployed to an earlier hadoop it probably will not work correctly +Hi Harshyes but is merged into and I want this patch go into Friends I have done this changes in srcmapredorgapachehadoopmapredBut can you tell me how will i build this code +It is difficult enough with so many class loaders +Bisocket unit tests pass on cruisecontrol. +Somehow overlooked that Jukka was assignee +Hi is the tag for available +And the same setup works with webwork As soon as I update to webwork the displaytag components stops working without any errors and the logs are clean +If I uncomment all tests in jdbcapiSuite before except I can see the errorSomehow it seems the problem is a of which does shut down the wombat database and bring back up again +Correction by providing a list of projects using Maven CLI option pl only the first one is evaluated by sonarsonar +srcjavaorgapachesolrhandleradmin +Something simple like this will work to update the iptables service command for FedoraAlternatively something more sophisticated could be done to search a list of supported services to figure out which command to use +Unless the TeiidModeshape jars have to be respun and this needs to be reopened I think we can call this one resolved. +ssh account created privs bzotter access access NOT granted since apparently hes not signed up yet +Okay so I guess the question is who objects to pulling this from? I dont think we should release a class that extends a deprecated class and I dont think we want to hold up waiting for an adequate non deprecated replacement +revision adds the gbeandata registry and queries against it for resource refs and gbean links +Thanks for the reviewcommit Cheolsoo +Since the Tools BIRT integration depends on BIRT I have fixed this using the has been a similar issue with the TIME type +Since I already had a patch and I thought of uploading it +Assigning to me to track the test results and will Resolve accordingly +So bundle a default version of hadoop in just in case +Resubmitted to include the ASF license +off release +However there is lack of unit tests +All the other cases are reported here +Chris are you still working on this +However I would love to see an example Java code and a cursor position that triggers this error +Commented on +If theres any failover to another one the whole thing is toast +That is correct +changed Issue Type from Bug to Feature Request based on discussion with Gary +Re I think we can first add the terms and frequency in separate using the iterator and then start from the middle +Im testing out a more recent version to see if I can isolate the crash +Thanks Charles +However if I am mistaken please reopen +One can argue that logjs approach isnt the best but itis the standard for practical purposes +I see I didnt have fields in the object +Well how about a patch with an integration test? +Thanks Alexei integrated as expected +is a fuller version this issue +In the new test I think you used the testDelete method instead of the new checkDeleted method +When the script ends the is flushed into the request writer +handling maven and genereal project correctly for now +But atleast you should be able to use it from the trunk +Okay I will move all issues to a new Jira component. +I wonder how you would do that +AgreeJust recognized that using CODI bean reference was already on my list for today +done +What utility would major version numbers then have over a list of methods? If the issue is simply performance then perhaps a hash of the method names could be sent instead of the full list +Sorry for reopening +also for any future issue even if an optimized copyBytes is put in we should be careful if it only works depending on s internalsat the least instead of setting a trap we should make NIOFSIO and maybe NIOFS itself finali think in general there is a lot involved and we should just open an issue if we want to go this route but there is a lot of complexity involved personally I would like to see what the gain is. +Can you rebase to trunk when you add those +Looking forward to the repro issue +That was pass pass coming in a moment +Added close attribute to tcp gateways and inbound adapter. +trunk and branch +Marking this as fixed so Hadrian wont stop cutting a release because its listed as unresolved +In r. +to where we can provide a better fix +So much for back compatibilityThis is entirelycase adding methods to could technically be viewed as breaking back compat but in this specific case its OK since no one implementscase removing breaks all applications that refer to +is directly related to this issue +will fix it for same branches as initial fix +Will try this immediately +Cleaned up with some macros form apr and httpd that allowto subst expanded vars which could be used instead current DPKGLOCALSTATEDIRdefines in a file +As for the generated pages they would be patchable in the same way if the sources of the site were also in asfnutchsite +And BTW if would be easy to access services from the test bundle using an Activator that stores the and makes it available via a static variable +Patch replaced with the one based on the latest svn revision +Our end is complete with the sections in our httpd configured +Should be fixed in nightly builds starting Jun Thanks for reporting this. +Please consider committing +Looks like this may just end up being a resolve +bug from do not exists in and +Looked into it +I tested the fix and now the exception is send back to my client as expected thanks. +A small patch fixing the issue +tests were updated +Users might have different correlation strategies and having an example of how they can override the might help them +This also means that our local NOTICE files only include the correct attributions specific for that sub module +the Recent Changes plugin and checks if there two pages or only oneBy the way which class replaces the Reference Manager? +Committed revision to trunk +which patches are relevant? All three +to special directory along old standard examples watch also where is current progress attachedtracked +I think your solution is bad +Changed the title to only fix logs specific to YARNmetrics link is not working in common opened for this +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The changes in trunk over the previous weeks have caused different performance results during my tests +Although I have not testing this client vs +Thanks ATMI updated a new patch with fixes from both your suggestions +Update patch for changes to trunk +I know I know +I have developed more complex test and committed it with actual fix to the master branch +The auth scheme mechanism should now work as expected +Moved out the code that deletes the user to its own block +and alexoreeThx. +works perfect. +Thanks for clarifying that Jrg +And I found some codes in ws are similar with esb +I just committed a fix for this +I see +So marking this as fixed for. +Maybe rewriting to a more simple syntax would help +Get the latest code from Git +I can definitely see this beeing useful +Thanks for reporting this +Interesting +Thank you +Hi Karthik Kambatlathanks for your you plan to test this patch under secure cluster? May be there is something different between secure and unsecure clusters +May need to come up with some stress tests but it isnt going to be easy +The only improvement I can think of for the future is to well define the content of returned unknown properties in SDO terms instead of EMF +Updated to as well as corresponding fixes in due to changes made in Derby. +I believe the correct attribute to use is nameValue +on +Actually its not possible to have an annotation on the Externalizer itself cos the framework will receive the object to be externalize so hows it gonna figure out which if any externalizers to use dynamically without any other hint? The annotation needs to be part of the class being externalized +Added NPE check and unknown feature reference warning +Logged In YES useridChanged identified locations to use the transactionmanager to disassociate the threads transaction forother +Maybe we need some kind of hook out to contribs for migration +I was able to compile the xsds listed and the case radu mentioned +This really should be corrected +Ive created to educate users on why it happens. +Reverting to fixed that problem as well +Suppressed unchecked warnings in a test to get Hudson to notice this +Doug wanted the split file to have a fixed ascii header +Adding to branch as committed to to branch and trunk. +Confirmed that it works fine against upstream +Hi Jitendra I think that long term it may make sense to extensively change the protocol authorization system to support multiple expected client principal names andor groups but a change like that seems like overkill to address this particular issueI think that being able to configure multiple expected client principal names in the annotation still doesnt address the issue of the balancer which could conceivably be run from any node in the cluster +I ran across this problem too +patch creating a clone of examplesolrcollection at examplesolrcollection with changes required for schemaless mode operation details in the issue description +Patch committed to trunk and branch +Alan Any issues with this patch? The extra ntohs in should be removed in any case +what replay buffer +And additional tests for Quick Search were added too. +. +I increased the serverity on this bug accordingly +Ive always viewed as purely an encodingdecoding utility that may not even know it is talking over a socket so I would probably expand the exception into a small exception hierarchy that has the necessary granularity to indicate the various overflow conditions you have addedAdditionally Im not sure I would include all the exceptions you have in +I cant reproduce at all +I didnt include negative test for Hive as David suggest +did you check if there is a wiki page? if not this feature hasnt been documented +I plan to make a functionalspecification based on a subset of the SQL standard +Consider adding a comment at the top of the if block mentioning why it exists +BTW what is the origin of the class and is it originally licensed under Apache copyright? Is the copyright notice on top of the file correct and also in the original +If it simplifies things I can manage changes necessary to the authz file since I have all the necessary karma +According to the log the client has failed then tried to fail over onto the backup server +Fixed in subversion in revision +Also it might make sense to take advantage of to handle cleaning up old hints +bulk defer to +Thanks to next release +Sorry back to the beginning +Fixed in V of the plugin. +That was done intentionally because doesnt render a twitter page properly +Committed revision . +Oh and we should probably work towards an initial release and mark these more basic issues as affecting the next release +In lack of a better size I still use k but now defined in a constant in +code extracted from and added a small interface to reuse some good on a quick scan go for it! +We would have to solve that on the java part +will take a closer look at this +Yes in the raid contrib see suresh few days ago i got this same idea of distributed in hadoop i got impressed with your idea +Is there any sort of timeline when this old bug may be fixed? Or at the very least the submitted patch committed to trunk? Only after a couple of hours of frustration did I happen upon this bug +Another thing to consider is our backoff policy on client retries +I have this problem too and I guess that it is not due to the new palette because there is the same problem in the +Release Notes updated. +Ive added an usage example to the examples application +Part of release +There is a transient counter and the number of retries that both get incremented when this happens that shouldnt be incremented when the suspend command is used but Im not sure if these are exposed to the user on the web UI or CLI +Untested patch that revises the original patch +The data is never copied from the local disk to hdfs +Will upload a patch that incorporates the comments related to the code soon +This indexing does a deep traversal of the entity children to collect any interesting text to index however this causes an as described aboveIf I only do a shallow indexing then there is no problem but then I have an incomplete lucene index which is not idealIf this is not going to be fixed then what could be an appropriate workaround? Im thinking of having an indexing thread which waits for notification from the post update event listener +so I hope that these changes dont cause any problems in other scenarios +The only concern I have is if there is one message pub failed it will cause subsequent messages been blocked until the message are +No activity cannot reproduce. +I have committed this to the branch. +condition on its machine regardless of other consumers in the same job +Sn are balanced by one LBThere is a custom LB which will balance R +Just follow the commit links in my previous comment +Canceling patch available status due to a lack of ICLA CCLA and removing myself from the assignee +Hi NirmalThanks for taking a look at this bug +Ive also seen this happen frequently in my test runs on phoneME running inside a instance so it seems that the fix was incomplete +The above is from AM? If so Im not sure it a bug +Upstream issues in are still unresolved despite their releasing a build so bumping back to CR +Will take a closer look at this one +Let me know if the Maven build ismissing something + +Defer all open issues to +Changed the debug GeorgeThanks for the patch +Does moving the discussion to a jira make sense +Committed patch in revision to the branch in revision . +Code is still there but stopped +I am able to get the application loaded with it +Update patch as last comments from Nicholas +Thanks Yonik for the very prompt reply +Attached patch includes only the unit test and is generated on trunkThe other attachments shows that the added unit test passes after the proposed fix and no other diff is introduced +New version of the patch theres no need to startstop listeners because it is done by the server startstop methods +Committed to trunk. +I commented the patch to. +this resolution violates the SCA specYou are not supposed to go adding stuff to the SCA namespace that is notapproved by the SCA spec process +Anyway is in the Teiid plan the fix for this issue +Are we happy with the current level of tracing available with Axis CDo we have any plans to implement the said level of tracing? Do we have this kind of feature with Axis JavaPersonally I do feel that this level of tracing would help however would make the code too clumsy unless otherwise we use AOP which is far too much +Generalizing it a bit the place where it should be more easily relaxable are the cases where we know very few people make customizations as in implementing Fieldable orAs for this specific case the original change was the thing that broke back compat +Because I added a check in instead of The check in will pass in the normal case because the app has unregistered and this is the first call to remove app +to the solution Kathey proposes in KatheyI was searching for that API thanks for the link! Ill attach a patch soonThanks for addressing the issue with weme too +I am on this one I really like the idea of having the css equivalents of all the built in styles but dont really like the idea of shipping them when the css extension might not be installed +I tested my new patch exclusively with times in a row and it didnt failIt passes all the unit testsI tried it on a cluster during an import job that created regions on region servers and then played with disables and enables without any failure +Done with unix project creation +Throwing an error there might shutdown the db i am not sure what itdoes at top level with an error +Using plain kit works without any problemI suggest to provide a classloader in the so the Foo class can be loaded +lgtm +Assign to Manik to integrate into release processManik I added a profile called configdoc that will generate a file in targetxmlxslt +So theres something not quite sure you run both ant and ant test to confirm your fixes dont introduce additional failuresregressions +I can see that you spent lot of your time to analyze this issue +Fix identifier capitalization issues in schema scripts +Doh I forgot about the configuration in EPP of proxied gadgetI still dont understand why its proxied in maximize mode and not in normal mode thoughI dont know if the patch is needed +If a non owner updates a value for which it has an L value it removes the value from its own L cache +Workaround If the german file is encoded in the the special chars are shown correctly +Ill check this in later this afternoon +Added a patch for which keep the old functions and deprecate them +Thanks for the timely report! tested my internal project against both and and everything works as expected now +You are right +All I can see is +svn files when I added it to svn +jbds doesnt include the maven plugins so no this aint for JBDS +Thats odd we have observed a change in behavior in the UI recently +No copyright notice or license doesnt mean public domain though +Thanks Tim +After clicking enter the value is saved and propagated to xml +So Ive just wrapped the offending line in a trycatch and default it to an empty set if an exception is thrownIf the task is truly dependent on some generated folders being present I guess we should make sure that the build task is marked as a dependency of the idea task +Console output +Project checked into httpshuntcmojotrunksandbox in the sandbox +I have left the fix version unselected +xsp svnnative +And please try to narrow timeout block region of your code +To change the in a cluster this will need to be set +Something as simple as setting a high water mark on the number of pending events and throttling events from incoming connections until the congestion subsidesWe have see a similar issue in the IPC layer on the AM when too many reducers were trying to download the mapper locations +Closing +difficult to assert the stream closes alone +Somehow the unit tests do not catch this error as you mentioned +The issue is resolved with the current JFFI head +Will be done when we will use Sonar Runner to launch local analysis +Im with that change assuming tests pass + +I dont think we should touch at all +they are aware that they are leaving the sandbox and that they need to play nicely +Committed by Paul +Hope you understand the thinking here +It should apply to the trunk OK +I could make the packaging system do the reverse if hadoop community prefers this +I made some minor edits and uploaded a new patch removed unused imports in the file +Hi CarlosThanks for the answer Ill do this during the +Recent is managed by the server onlyIve updated the newer IMAP implementation so that the parser ignores Recent but does not throw a +HossI finally got by my problem with running this test individually and ran times both on trunk and x +The patch is for debugging stack map recalculation +it was picked up a little bit later +In that way the older RS wont be able to update the table properly +Removed a bunch of debugging echos in the ant build files +Made things much easier +PavolI added the wait in TRUNK +Im frankly a bit stupid sometimes +Fixed in r +Initial tests are good Ill attach a patch soon +Please resubmit with a test that exhibits the problem and the code that fixes it really is difficult to apply a patch generated like this +thanks +This patch shouldnt create any issues we didnt already have +Sorry reopening this again since while works with x JVM but it still raises on Windows x on JDK really need this fixed on x JDK as well since otherwise rubygems wont work on that JVM +oops was in the wrong bug list +Do you want me to break this up into two separate patches +assigning to kurt to take a look at and provide any issue was migrated to +Deprecating old metadata classes have been deprecated in favor of the ones available on coverage +Does this occur in Apache Flex? Can you provide an simple example that shows the issue +i think this should get it +Hi I tried some examples and in those there were no but some relevantexceptions +I atteched my changed file +Added a new patch that contains the pom +In theory different channels could go to different servers so the current negotiation between the two parties ensures that only feedback is exchanged that is wanted by the receiving party +As a result were using this as our base resolution +There is no default and is inappropriate +I think it would help developers to have the Pivot Collections interfaces look more like the so Id like to take that on as a project for +Verified on Hudson +I suppose there could be a list and the parser could add itself to the listBut Im not certain that the benefit exceeds the cost eitherOTOH one of our previous discussions mentioned the possibility of Tika not doing any logging at all +Attaching patchThanks up patch and moving rest of the stuff to corresponding JIRASThankss the relationship between this patch and make web apps to work? It seems to be related to RPC instead +for the second +This attribute can be set as a servlet parameter +We really need to support since the deprecation of is not very friendly +Although based on prior discussions between Me Runping Mac and Milind it appears to have better fit and more potential with Chukwa framework and so was decided to be part of Chukwa projectMilindMacRunping comments +Patch applied in Git ID bfbbbccdbfceb +This is a forward reference issue +This looks very promising +The ability to blockcascade state seems like an edge case not the normThe current impl covers the common case where disabling the parent will disable the child controls +its just a sys property name weve specified in Keeping in mind we didnt really specify it this is a jetty default we just changed the port +contained projects will create a textmarker folder in the sandbox and import everything into it +Thanks Andreas +Thanks! +The fix was checked into the code base so you will need a nightly build or the release candidate when we get one built in order to see if your problem goes away +Thank you for the feedback +Well see what the Jettison devs will say will use what they provide +Fixed in commit bfbfcbddc +Thanks for the quick review +If user selects a class from jars in the project classpath there is a NPE error +getChars insteadThe method still needs to be fixed in the PC component +onChange only ever sees one piece change at a time so we perform a kind of state machine here +gradle and restarting Eclipse gradle files open in the groovy editor with no problems. +Fix the localrunner time the Checkpointer failedSeems that Jenkins is not working properlyWould someone please apply the test on his trunk and run the testcasesFor me they are working +I closed this issue too quickly +Please verify whether the problem is resolved as you expected +If this method can be used to manually create new accounts that is fine by me Im not asking for a fancy UI to create accounts +Commited on trunk and branch as of revision +There is not as well as y in the end of Derectory +If you disagree please comment +trunk +thanks Ivan for explanation +any additional info additional info here? Otherwise we will close this ticket within the next days. +Downgrading the urgency of this issue +I see thats a good argument for leaving it as it is +Committed to trunk +Please remove the unrelated changes adding Override annotations to the other methods ofEven though this patch introduces an interface which as yet has no actual implementation Id like to commit it to trunk anyway so that can be used during reviews on +the whitespace changes were only in the new lines bill had added just spacing them apache stylethe original patch cleaned up on failure but didnt rethrown the exception this version does +Again client can cache the server mechanisms and associated info which almost never change across connectionsbq +It would be really nice if we can split this into two patches one patch that fixes the infinite loop and the other is a related to making private +Thanks for investigating Nicklas if you get the same issue again if you could please capture some network traces and this Jira we can look closer. +Which means this is indeed a bug not an enhancement +shuffles the directories to distribute disk usage when using multiple workers per machine +Checked on win XP and lnxBut I found one more instability issue and it would be great if I could add fix for it to this jira + This bug has been marked as a duplicate of +On the face of it this has some overlap with the recent ofAny comments +It worked well +Retargeting to and +This seems to incorporate which the first time it was floated had some objections +Well if we really need observer to observer responses for reconfig purposes I presume then should we be sending them to observers not in LOOKING state? See the conditions that apply when responding to participants in the lines below my patchBut even still with that being correct it might be too much overhead for large Observers deployments +Too many boolean flags I dont detect a sense of clear minded purpose +Can we fix the cause why did a to this one +I just pushed this to trunk only! Thanks Eugene do let me know if this needs to go into branch. +I think thats likely as wellSome easy ways to integrate would be nice however similar to the way the Spring folks set it up so Spring could handle the requests from GWTIm planing to take a look at it but Im going to have to get a lot more familiar with Wicket first +Christian can you please attach test project that would allow us to reproduce? Please also specify how exactly you created a war project +One of our customers ran into a problem that makes the results page useless +on patchMay be a small release not is needed +Added explicit handling of empty Strings for Calendar and deserializers it will be mapped to null values +Ill follow that up and see if I can resolve it +This is where it should be fixed +Using chukwa this disk is now the bottleneck +from me +Id like to give the ability of moving jiras to hdfs to contributorsof bookkeeper +No problem Patch looks good to me + we fix whats causing the error rather than changing the column to +Ive added licenses to all resource files +Whatever is convenient for you +place in a directory called Also fixed issue with using quoted csv records. +Technically no it is not an bug but having a default configuration that is broken in current versions of Safari seems like bad form to me +I will comment out from all integration tests and create a patch +Fixed +Could you verify if this was already fixed +Do you also want this for +The changes are not effective on ibiblio +The bug you linked to describes my problem completely +so I think it is a blocker +Could you explain how gtest is related to Tajo +Let me know what you think Eli +The bulk of the changes are in the trunk at rI still need to commit the tests after testing changes on LinuxAlso need to add the trigger creation which removes unreferenced messages from the messages table +Closing it. +Can you state which TCK test fails? I can take a look to see if the TCK test is following the spec rules +It works all fine here +Ok will try thanks! Just to save you time problem seems to be with readUrls method in that for some reason returns only first jar from whole war file +I merged the patch to. + to Paul Rivera for provide us with this patch +Thanks Thomas and Jukka I was struggling with finding good names +Let me know if you are happy with the proposed fix and I will merge into code freeze is a few days away +I have finish data model refactor job for this project and sent it to Rahul he will help me to share it in projects sandbox SVN soon +This is not a question but one that should be made to the JSF specification team if you desire +Reporters of these issues please verify the fixes and help close these tried to repro the issue on host XS cluster and things worked as expected +But s javadoc says return the Session object +Thank you Robert +Also can you please post this as a single patch file? Thanks +Committed to trunk. +This makes sense now +Looking good +This issue fixed by revision . +Currently resolved by providing level and modifications +Test case project +For the first one I did some changes in trying to fix the bug +Is that correctLooks like the class level javadocs got overlooked when the whole similarity API contract got revamped ill remove all those references to global settersbq +Hi Andreas the changes fix the issue please help to port it to too +Note that this is a regression from +So Thawte will only issue code signing certificates and they wont let me renew our one that expires in yetIdeally wed be using timestamped jar signatures where supported by the jre that would allow us to continue signing with the old cert right up until the expiry date but the user would never get a notification of an out of date update +Reassign to a release that you feel comfortable resolving this issue in or leave as is. +what Linux does if a remote directory is mounted to a local machine but some accounts used in the remote directory does not exist locallyLinux reports the uid and gid of the files instead of the username and group name +Not approved for EAP +addind an example to test and some work on proxies +I didnt implement the collate thing yet but I think that can be handled as a separate patch +Hey nick sorry to pass this off to you but I at least need your advice on what exactly Im supposed to change +For example disk space usage could be trended via Zenoss much the same way we trend CPU utilization +We could modify to doCompile only when sourceFile is newer than the destination fileThese are just seconds of the seconds needed to build +Cant say for sure since get the test case to reproduce the problem +Disconnecting and reconnecting will allow you to publish however all the threads that are waiting continue to wait +Removed that and with the other fixes I mentioned before everything is humming +OK I found where the bug is coming from +simple fix attached all tests code changed to existing commons lang implementation +What you need is a +It would be better however if it can be configured through the hibernate configuration fileQuery substitution is very close to the idea although not straigtforwardRegardsPeilin Zhang +As is section I imagine +I had the same problem and an upgrade resolved it +Closed after release. +Patch committed to trunk +Make sure the first directory in the path that points to a JRE points to a JRE dont just tack the JDK JRE path reference on the end of the PATH variable and hope it works +I am verified that the update of class fixes the problem + fix unit test failures +Thanks for the patch Niranjan! Since I filed this JIRA a newInstance method was added to Resource itself so I think it would be preferable to use that one over the one in which hopefully will be phased outAlso would it be possible to remove the ResourcescreateResource as well? It will require explicitly putting or in the places its used but I think this is probably advantageous for preventing unexpected behavior as YARN becomes more centric +on patch +Note that the fix is different than the patch so that more general cases could be supported +Then each of the threads would equally be in midst of retrying +Thanks Chenta! +None of them have external dependencies +I dont have code ready but I dont see why this shouldnt try to go in some of? Its not marked a blocker whats the purpose in punting so long as they wont affect stability of the running system +Is the client code logging? If so we should file another bug to make this independent of all of the server logging for sure +updated patch +Did u get any fix or alternate way for this issue? we actually stopped using so basically my problem went away +assigning back to you +Yes just tried this on trunk and its definitely fixed +What about changing the method to return a list of one item per geometry column +In other words whether table layout or tabxxxcenter styles it will be ultimately refactored +Not a bug. +I modified to generate with the ANTLR Xwatchconversion option +If not something is wrong with maven or our pom file is not specifying all of its dependencies correctly +HelloWhat status does this ticket haveCan we expect it to be resolved? +Wouldnt this be just a matter of generating a static initializer block that tries to access all static fields of the class +Updated +Hmm I thought all of the methods which worked the old way were going to be deprecated +Canceling patch +Please check that it was applied as expected +Yeah I think there is some overlap +Doesnt need to implement Writable? Otherwise looks good +Agreed John +Thanks Mark for the pointer! +Not for is being used in the course this week as is +Relatively small patch skipping RB +I did not have a clean check outDon +But everytime i start the datanodetasktracker manually +Whats up with this vector rule? Only vector logos? Weak +Patch fixes problem described +Well I didnt view the configuration as unclean or a mistake necessarily just as containing redundant information +adding the missing instrumental word to Fuhweis comment above +Hi ChrisOne thing we could simply do is to improve the labels for instance here CreateEdit Shipment Method Type +Part of a bulk update all resolutions changed to done please review history to original resolution type +Different block sizes are good because blocksize is minimum split unit +I definitely think that this is a useful fix I dont think that your original patch is correct because there are cases where repeating the vsnprintf is required +Moving to since we have disabled baremetal for +Do you know what I might be doing wrong? I couldnt seem to find any documentation on this + AFAIK a new has been in use for a few days +Needs the plugin issue patch applied. +After quite some time without response and in the light of the implementation in the it seems proper to close this bug. +bulk close of all resolved issues. +Attached is a zip file with new files and an updated +Please use the community forums to ask for assistance Jira is for bug and feature tracking only. +Now menu item Delete must be disabled for JSFCore as well as for other sections distributed with RHDS that is what I called protected +My opinion hasnt changed and I still dont want to be associated with the project +We should change and perhaps the Query class itself to support a better nothing of sortable columns than passing text strings that get appended to the query +The kind folks on the Eclipse Platform Development list helped my find the the file the feature tag has an attribute called plugin +What this JIRA should refer to is not that is failing to work but that the tool doesnt show any indication that its not connected to a broker +Thanks everyone +Can you please confirm that the restored file is the correct one? For me it looks like that it might not be validit was the oldest one now back to CAS API assumed it should look like the one inside our test folder +Here are changes to impetuss patch +for we need a more detailed discussion on how and where to filter out the hidden items +Thanks Tim. +Ill reattach the dumps as another attachment +Charles it is necessary to be running a STOMP server on localhost +You want to do those now Ram +Canceling patch until we converge +The attached file contains the OBR description of the preferences service +Perform deep surgery on DBCP so it works with existing pool implementations but as you said this probably wont be backward compatible and I think its a bit of the tail wagging the dogI personally think is the best choice but Ill work with you to make happen +Hudson uses to send mail to +I would be happier with an or supporting about it some more it should be possible to reset the counter in Tables back to if we happen to overflow the long +They may +prefix +bq +Geir please use only the patch +Here is a patch that does both sides +If is properly implemented then I dont see why it should break the behavior of the existing methods since it defines a policy on top of a mechanism and these classes implement the mechanismIn any case there is no harm done in splitting them and its trivial to do so +I have committed a patch to fix this +I made the patch in a way to be compatible +Avro is not with Jute +Thanks. +The property is true turns it on +Regressions passed with no failures I think its safe to apply at least patch psanity to fix the sanity issuesTesting older clients today +This issue is similar to fixes we are making in and. + +Im attaching agents sources with permission for ASF to use them +Attaching a first attempt at a release noteIt doesnt say anything about the fact that you can invoke getBytesgetString multiple times +srctestjavaorgapachehadoopyarn +Im resolving the issue. +applies cleanly and solves the problem +I updated your patch to apply and compile in latest trunkI am not committing this patch since I dont want to mess with ToddsFetcher work +great work Puru +RB comments posted +Reporters of these issues please verify the fixes and help close these state is being correctly shown in the UI after the specified +installed Geronimo +The amount of time it takes to do a DU could be significant +seems to work again. +We dont need that any more +Please try again +Ilya I agree with you +ve committed this to trunk and merged to +turned out the javadocs were fairly accurate +It goes without saying that if you strip the accent you change the meaning +Should I resubmit this patch wo button group? I filed this seperately from for protocol and testability resons only + I agree seems like a category A licenseUpdate Upon reading more about the license the restrictions that are imposed exclude it from consideration as a Category A license +Will take a bit longer to get out there +Fixed before was even completed so reopen reclose after removing the label +I just committed this. +review comments addressed +Log and swallow. +Fixed in which version +However I am not not sure the proper place to perform the clear +THANK YOU! +Committed you gonna try out the font support +Issue still present +Unfortuntly I cant force users to write HQL in the correct form hence it is still a blockerAttached ZIP contains a sample scenario which reproduces itString hqlWorks select from Component as library left join as book with parent left join as height with height A left join as width with width B where librarylibraryString hqlBroken select from Component as library left join as book with parent left join as height with height B left join as width with width A where librarylibrary +In the init script Use HTTPFSUSER instead of hardcoding httpfs I understand its copypaste from hadoop init script but I would rather use one of the other init scripts +I just reopened As said This is a tracker bug to track work in JBIDE +The login jsp needs to be rewritten to work properly on mobile changes which need to be made no giant complex background image +It will check yoru database for inconsistenciesBefore you can run it you have to edit a couple of files +Rebased + +XCAS service communication between Apache UIMA and IBM UIMA already works +For remaining devstudios to work the bugzilla needs to be fixed +Writing that into the relevant documentation would definitely be a welcome you for the job done anyway +Just add a in front of the path when adding the servlet +Includes fix for +Thanks a lot +In falcon we want some jars to be added to all actions of specific entity lifecycle which cant be done using oozie sharelib +A massive brain cramp today caused me to write that last comment +And where exactly in the JSF specification does it say that this behaviour is not correctThe fact that you personally might want a different behaviour does not make this a bug +It serves as a backup imageI want to restore snapshotname back to A from the second cluster DFS +Will be included in the next release +I just checked the tag doesnt have the patch I merged this morning we need to wait for the . +The patch removes references to identity columns so that the javadoc now only talks about sequences +Ill second that +This is how does it for seam do you know if this is a bug or intended for some reason? Maybe for +I like your idea of having removeAll and removeArgs mention each other in the javadoc +If you experienced this phenomenon only in AS then sometimes Jboss AS is doing something with these jars dont know +Please check that the patch was applied as you expected +closing per reporter. +Attaching the patch merged to the latest was resolved in. +Even your sample pdf doesnt throw any exception or anything similar +I have fixed the problem with empty attachments +Fixed in branch in adf +This explains why when I manually set the version in my first submission that the problem was solved +Okay I rewrote things to use a interface as you suggested +I am still baffled +Shouldnt be too hard to fix though I just dont have time right ve fixed the bug I found yesterday and added some documentation +one line change fot this ticket +Whats the reasoning behind putting the full pathnames? The hadoop metrics output directory is configurable right? Wouldnt changing it now require changing all the filename properties as well +This properties specifies the number of retires before the producer give up and drop this message +No need for integration tests +introduce this feature as an option initially +It was applied with minor changes to and +marking fix in and found in versions to to be merged with later +Lucas can you create a test for this problem? Ill try to investigate it too +We should probably mention this in the docs +The rss test failure had a somewhat different signature in that it looked like the document delete stuffer thread had just given up and frozen or exited leaving lots of documents in the ELIGIBLEFORDELETE state +Form should be inside modalPanel and not outside +YogiWe need your input on this +We will soon have a release to accompany these as there is some issue with the original groupId it would be good if you could explain it to me so that we can clear up any +Checking this in to unblock nightly builds + Bulk move to on JBDS +I havent used JDK in a while but Im guessing the problem is still there +agreed +Triaged for +The other fix is to make sure the iterator supports failover as well +Thats a valid point +AntonI cant replicate the exception youre coming up with +This patch will ensure that +I may even wait for it before I start using STC. +Which may be by design since I cannot think of a case where I would want to add a message to a control and then navigate to a different pageSo this issue might be already fixed in +So it is more forgiving than the status quo and the warning is an added bonus +When software become more complex it is possible that self contained binary tarball directory would not scale up for large projects +Ivo I can help you with getting set up on +Ive tried to use a jquerycontext menu plugin to do but didnt get far +should probably really only have bug fixes and minor improvements at this point keeping it stable and we should make the next bastion of the high tech pigs +Another way that I can see this behavior is to just hit httplocalhost with a web browser and watching the open files for the synapse processOur server is running out of open files I think mostly due to the first scenario above +On the logger that I need to resolve it is the base class that is putting out the log always whatever the subclass but need cleanup +As I recall the test case bypass for Javas not supporting XML is testing the wrong part of the support +Tomahawk examples are now all working using the new +Thanks. + doesnt detect changes to on the filesystem +This could be due to some code between the successful query and the call to next in the current thread or some other thread got hold of the result set +Assigning to to Michelle as new gatekeeper for doc +should be I can confirm that your fix worksThanks! +a conversation name was specified as a parameter but not the conversationId itself then the EL of the is evaluated to find the curernt conversation id + +The next xalan versionwill fix this bug +Chris feel free to resolve Ive pushed your commits upstream. +Sandy if you havent already can you do a unit test run to check that there are no regressions please +For or just +It just allows users to pass in additional libs in the path +Cant understand why all jobs in Job Executor will be processed in one transaction +Updated as recommended by Army Added code comments to describe the reason for adding a var entry Added constant to reference ECLIPSEHOME for your comments ArmyI made the source changes that you recommended +Closing resolvedwont fix issues. +Yeah the compile step probably works for me because Ive run mvn packge install on trunk +According Legolas Greenleaf in this issue could be closed +Getting a fix for these will definitely help my SEAMEJB productivity +Removed busy loop from this change then requires AMQ at a later than do we say that in our documentation? +Need test results +to define the HORNETQ home +r too +If it gets handled by catch it will likely get logged or not if the developer doesnt want it to be +oh even if the number of tasks is greater than it should not attempt to calculate the avgLocal test passed +Changes look fine but could you fix the indentation +This patch removes the try catch block to allow the tests to fail if an unexpected exception is thrown +Hey you still around? Im not sure what to do about this issue really +In this case we should probably also adjust the current Jackrabbit implementation +No response assuming ok. +Please look in history for what the actual resolution was. +We could add a specific method that just returns count as it may make it easier for JMX clientsDo we want to return the total size used in the data directory for each CF or does do we want it globally for level? I was thinking of adding one at the level and it would just return the free space for each data directory +would it make more sense for this to go into mahout +The only text in view page source is also NULL +Btw I think I saw a thread related to this topic on the geotools mailing list as fixed added tests to double check the thing stays in that state +Fixed +Resolving as fixed +When you create plan from the if any error it just shows the root causes detail msg nowCan use the +I think I have located and solved the problemIn the jsr component I added a line to set the operation in the exchangeIn the jms component I use the endpoint name and the operation to set the soap action +The only time it would be used after that is prior to committing a modification to the code +ve committed this and merged to +Which might explain why without the seam upgrade it doesnt appearSo how do we solve this? Reverting jerviss commit is to big +i was using Velocity and Velocity Tools View Netbeans and Apache Tomcat and SQL server for build the my web applications +Committed revision +First merge patch fixes almost every lucenesrctest +Oh boy taking a look at the patch its pretty massive +I have recently had a discussion with one of my colleagues turns out this issue might be something on our end +Can someone please commit this before the next release +So is there any chance how to suppress the NULL value when using sum? +Nice job +Latest patch is good +The artifacts will be updated manually +try deploying to jboss see if you get the same problem +Fix bug introduced by catch Renaud! The patch looks good can you add a test case to to make sure abort always removes files correctly + looks good to me +Or even better rewrite the sentence to one it has the advantage of yada yada +It looks like a simple change but this late in the development cycle it probably wont be accepted unless we can show a significant performance do you want to file this as a bug in the iBatis JIRA? After you do I can work on some performance benchmarks to submit as part of that issue +It seems that this would be the final step in fixing as well +There is a problem if the query contains a pk condition +Shazron what would be the impact of using such a fix? Would it be similar to setup? Could it affect localStorage backup in some way +I am fine with going that rout so I will just update the test to expect the new behavior and then document that behavior on the other JIRA +Thank you Tanping. +Greg No its not related. +It will most likely be a typo on one side or the other +OK understood even if I find this disturbing as a change in behavior between andBut how is one supposed to configure the start of bundles after deployment? Not using the CLIConsole of course but to configure it to call the activator on each +Good call Simone +Added two interfaces and invocations at appropriate you envision it to be used only for intermediate data between map and reduce or is it for end user data also +We should remove filter altogether once its been pushed +Comments werent present when I broughtup the issue page. +Filipes work is a feature people care about and any objections of correctness have been addressed +Agree +Hi Rob can you provide feedback on the full solution please +Updated patch modifies a test to cover the behavior and handles the +There are still some tasks related to this issue but since they are not needed for the fix I think Ill resolve this issue and open a new issue for the tasks +If it is not the case please feel free to reopen and provide more information and sample test project that would allow us to reproduce this. +senoughnot enoughg +Hey Eric This issues been resolved with Andres patch +attached patch a clean checkout typing the following worksgrails upgradegrails +castor at +I was thinking about it +for this patch +Im going to mark this as fixed +jar so that upgrades dont require changing the +DavidAs I already explained on private list I will be happy to help too +That aside as this incurs api changes I am not sure when would be the right time for it +This is of course not a blocker +The fix changes the plugin +HelloWould you please try this patch? Thanks a lotBest regards Richard patch was applied to SQL module at rPlease check that the patch was applied as you expected +It just occurred to me that this is wrong EVEN for SVN because the parent is NOT an aggregator its JUST a parent +Looking into this now +Reopen to move to right a temporary fix we can just add the file to each projects srcmainresources dir then Maven will add it to the jar +I am not sure that it is appropriate case +Please try the new patch +Added support for includeexcludes +Not the count of files possible to add to for this component reads every time component rendered +this code logic can be simplified to say if exceeds timeout period return FAILED or Timeout otherwise return InProgress +That has now been fixed in Axiom +You literally read my mind rats +Server side we have the issue +Ready for review +Thank you +bq +Ive just tested scenario described in question and now it works as expected so no need to create new jira. +After running this through a debugger I have determined that the source of the problem lies in the Profile Service +The fix is to publish the project layout in our maven repo and allow it to be pulled down by the release build +Temporarily reopening to update the release note info. +BTW the one thing that I did not do yet was change the groupId +The behaviour should then be somewhat consistent +Thanks for the patch and your contributions +I found the problem +What if it doesnt? You still end up in the current situation +Ill see if I can get a fix in for the next release +ThanksBest regardsWould you please try my patch? To apply this patch please +Committed revision close for +The pread performance of hdfs is too low to be used in the shuffleDo you think this is an inherent problem or just a problem with the current pread implementation? If in the vast majority of the cases the blocks were local to the reduce node then pread performance should not be terrible should it? Clearly the goal is to remove the task trackers attempts at managing local storeYes +Can you post some exampleIn ODBC PG does not use Portal +Sanne Im closing this one +I will attach a patch which adds deprecates ctors for all analyzers having those setters too +we would have to see what it looks likeMaybe better to just have a thats confusing than a AND thats confusing. +The see elements is broken as it relys on the Citations Plugin +Added useQuotes property to +If that is in face the case Id prefer to use those existing jars and request additional guidance in such an implementationMy understanding was that if the jars were already available the existing code would have found them +modifying patch to include test case +Thanks KarlWhile working on this I have found other issues the files used in hhfacility and myportal were not tenant enabled +optionsread otherwise we get Configurationoptionsread as the logger category The is unsynchronized multiple writers to an unsynchronized hashset can cause runtime exceptions or infinite loops why are you importing? extra indentation in the javadoc as per Todd +It seems to me that I can resolve the issue by detecting when are used in the side of findAll statements and generate an appropriate warning indicating that such expressions arent supported +Dont ask +You can run the application without the Clustered annotation in a setup as I am doing now +So look forward to having a clean solution to extract URI variables from the inbound http gateway or inbound http adapter +This patch adds an always reload attribute to the saxonxslt saxonquery and saxonproxy elements +Thanks Nick! +Making changes to didnt help but in debugging I noticed that there are indeed two instances of around and for some reason the is using the wrong that on my laptop the problem only happens after a mvn clean so its probably related to the repository taking a bit longer to come upAlso when the problem happens Sling works fine after stopping and restarting the sling core bundle +Yes that workaround could be used as well +I initially wanted the aggregation to be recursive so that in structures like this submodule submodule subsubmodule subsubmodule got a in submodule that was the merged +svn commit m fixing broken unit tests srcSending srcmainjavaorgdrools srcmainjavaorgdroolsrule srctestjavaorgdrools srctestjavaorgdroolsreteoo srctestjavaorgdroolsrule file data +I just put in similar code to address the other possible in that class +compact file is fully written and has a valid currentseq +Deferred until. +This issue is a duplicate of is caused by a bug in IIS +Arguments dumped from withing the function that is throwing the error +Now that you have the logic working I can go over it and clean up the appearance. +Thats what I did with the jetty classes anyway. +The subject is The ROWNUMBER function but I agree its confusing that way +Miguel as Joseph Kesselman indicated the default namespace problem isnt a bug in +Im going to collect all similar issues and raise this question onThanksStepan +Nick Could you please close this issue if it is done +This will take quite some time +Can someone please review the attached +I believe that the base interface package names are probably the most appropriate in this case since the package refers to a reference implementation? Let me know if this does not work +Added to track a proper fix later +ogcsrcnetopengiswfsimplmodified +what I was asking was +Where how exactly is it failing? Are you extending an interface or using a constant? Both of these are known to trick the analysis into thinking its not needed because the compiler optimizes it out +Other services deployed by hot deployment are not visible. +I thought it seems unnecerssary +Verified on EAP was not able to reproduce all the issues so I verified them just their code presence in CP tag. +It seems that this is a side effect of which does not resolve the deployable composites thus causing problems when trying to process them +Clebert did you find out? I want to get this release out asap +Patch removing the inclusion of which was not meant to be part of it. +How do I get the Xerces code? Is it available as a tarballI tried to look in the snapshot directory but it could not be found +This is the args that we pass to the mavens child process which is forked to the unit test +I canprovide an updated patch with your test right away +Lahiru may have patch related JIRA IDs +Attaching a v that does thatbq +patch I thought heartbeat and generation were the same thing +close this jira +Have we got any example on how to do it on a XML fileCheers +I also want +Lukas a first problem is caused by renamed to second issue is more interesting as it is relevant to newly introduced programatic Resource you please report that separately with a link to reproducer steps to reproduce +I will fix it and resubmitAs an aside i normally get all kinds of junit test failures so i guess i didnt notice this +That looks like exactly what I was expecting good catch +I cleaned up some directories +The problem is that its then easy for yall to add these comments but a total pain to remove the spam comments as we have to reopen the issue delete the comment and reclose +ok file a follow on issue for the issue and Im +Whats the use case? Is there something that isnt possible with it as isI would say no if we can conclude that there is no significant difference in performance between the implementations +the current patch takes care of themsrcbenchmarksgridmix contains references to +Not very happy with introducing an argument logReplay in unprotectedDelete and unprotectedRename to differentiate an operation while replaying edits log from user operation +I think the only thing that needs to be provided is a way to really hide the command maybe with a Hidden annotation or something alike +Thank you for reply. +Thanks for reporting this Kai +This includes jars pristine c and c tarballs python eggs binaries etc +Also Ill be a moderator +It doesnt address the leak we have because we dont the first we set but that shouldnt be to much of a leak +Itd be great to get this working! Does anyone understand the Maven issue with generating the +The file was just an unintelligent argument for what I really want +Ah my bad +the inmethod grid allows tables of unknown size +Thanks Alex +Would you like to remove that +I see I understand now +Thanks Roy! +Note the combined patch with all related jiras was d under +Also we nailed down the problem to what is shown in the description although our hql request is much more complicated +Replaced single identity objects and Complex class at all areas where identities are handled with a new Identity class +You are right +I can do this if I am the assignee I cant do this if Im not the assignee +Fix the unit test to not catch exceptionsActually wait for openCallback to be called by using a new control under discussion +Went with problem +I wont get to it before going on vacation +So we never end up with a request with as the partition +Mridul your patch is generated using git +Committed changes in r r and r. +Hi Sebastian I expect it will help with that somewhat +For remove its the network corresponding to the nic you want to remove +If I changed the address to use the jetty engine I can access the service without any troubleI will keep digging the issue +dims +Ill have a look at your snapshot +The exception gets swallowed on the way so you only see a can not use mapping file or something +Thanks Tucu for the review! I wanted to incorporate any comments on the original patch before preparing a corresponding one for trunk +Ive been digging through the code but without many answers yet +My recommendation stands and I think this should be implemented as an external plugin hosted on googlecode +I cant find it in the code that was checked into the sandbox but it is in the zip file +So I have added a thorough explanation in the class javadoc without any code change as of rThanks for reporting the issue and discussing about it +batch transition to closed remaining resolved bugs +Should it nor work for you then I will reopen the issue +like explained in encoding support in javancss itself is neededLets see what the guys at javancss will do +probably wont apply wo first +Im running regression tests now +has been marked as a duplicate of this bug +Thanks for the reminder. +Thatll be a separate fix +These are probably all obsolete +Its not the same as but Ive created to track it in the AS project +I used a slightly different algo for get to avoid pulling in guava just yet though we will probably end up pulling it in anyhow +Instead we could compute message digests over the crc checksums +FixedThanks BertrandFunny I reported similar bug to another project few days ago. +Brian this seems to be in the jboss tools build script still +Code commited to trunkmodules +Did anybody test it recently? Maybe also with the refactored in +Ok found the problem I have Sonar behind the Apache reverse proxy +Robbie would you be so good as to review +Transitioning to closed. +I am working on that elsewhere +I still dont understand why you wish to exclude some files from src from being compiled +The best solution is to pretty much make sure that Hadoop comes up first or configure the other service to use a limited set of portsAny port that is picked is going to conflict with something when talking above the k port line +I dont see this tested with a debugger as well as log messages +I agree the test is not correctI think we do not need multiple threads to test the assertion +No at field part of Hibernate Table annotation +At this point I think we should pass on this for since we anyway have it in trunk. +Hopefully this fixes the issue for you +The problem is that the HDFS client changed the property to in HA +Whats the stack trace +Were a long way off migrating to Spring at the moment so well probably just have to wait for +We probably just need to get the following stats in jmx +Patch to ignore gossip messages when the gossiper is disabled +Verified in its ready to be closed. +Please verify and it works +Fixed under svn revision +GB GB is probably betterCommitted with changes to GG. +Guys please can you have a look at this truckload of issues and see if they are really still unresolved +This patch removes the profile settingI also intend to backport it to +The only errorexception that could happen in this clock is OOM since the IO is to memory buffer +In the future please do not reformat the code when submitting a patch +This might be related to have a fix? Why is zk not coming up? How is this related to +fbpurposestyling shouldnt be in my code in the first place +Raising severity as this completely blocks running applications on the Aries noop resolver in a separate bundle as suggested by Alasdair and make blog and aries assembly to use the noop resolver +I still think that the right thing to do is to add LDLIBRARYPATHHADOOPCOMMONHOMElibnative to thmakee env unconditionally AND add to that whatever the property specifying +Yes these have been fixed +Will address when I get back from Thailand +Volume testing will be taken care of in +Verified that without the patch the unit test fails and passes after patch +Fixed in the addin. +yeschecking for SSLgetservername +Waiting for a few patches to be included in the RFC assemblySending assemblysrcmaindescriptorsSending assemblysrcmaindescriptorsSending assemblysrcmainetcSending gshellsrcmainorgapacheservicemixkernelgshelladminetcDeleting managementsrcmainjavaorgapacheservicemixmanagementosgiSending managementsrcmainresourcesspringSending Transmitting file dataCommitted revision . +Moving out of +There is no problem linking all reorganization issues with this parent one but please no subtasks +This is the problem I expect to occur but I am not sure +Procedure +Im gonna move this to to unless someone wants to work on this right now +If I had Id noticed that I had not providedNow with therere no more problems with common compress xzWrt patches for parallel bzip I dont understand enough of the subject matter to be able to contributeGreetsH. +It seems like the only motivation to call this an error is to justify using exceptions to handle links not that it would in fact be a user or program error to call it in this wayIt is nothing wrong to assume parameter f a path +This is a duplicate of +Mark may know more as he moved it into but have you checked out what has been done in the attachments +It took me a lot of time to understand and solve this one fullyIf no objections Ill commit the proposed solution soon +LuckaTodays Beta release includes a fix for this exception +incubatorambarisite is now writable as is your git repo David +Our retention time is at least hour +If the core service does not have a registry method add it +If the fields on the Product entity are populated it wont look at the corresponding records that is part of the pattern +Sorry I accidently this issue in the first place when I thought I was assigning to is now closed so I am this issue +Is there another way to clean up connections? If so perhaps we could provide a method on to expose that +Committed to SVN +TedOne more thing we should abort even without checking the file system +Ugh nevermind +mp JBOSSHOMEmodules logmodule jaxpmodule JBOSSHOME +oh a leftover from the move trunkcontrib sandboxits fixed at revision thanks for reportingangela +Pull requests merged +Thanks Robert! +Closed after release. +Please ignore the patch looks goodSome nits Please use a consistent coding style from Hadoop +Why has this fix been rolled back since Dec ? Now this error happens again for Tapestry and +Committed to master. +Hi KevinCould you please give a try with Oracle JVMAlso seems that this problem doesnt has a connection with detection +Note +The popup window appearing and showing the stack trace when the bug is triggered +wicketsrctestjavaorgapachewicketproperties +Uh +Following the submission of this issue it occured to me that the class loading structure of the Tomcat server in which this issue is being experienced is oddAll of the can be found in the jrelibext directory and are referenced using the in +Here are results on bit Linux with Sun JDKjdkchangetest nametimeM entriessec M bytessec bytescycle ms Charset ms ms Charset ms +revision . +The two major issues discovered have been resolvedNote If mixed jars no longer work in then the test infrastructure will need to be changed to allow isolated class loader for client. +Filed to fail the random failing in. +Pavel you are rightI propose fixing this bug by setting this option to false by default +resolved with revision the fix in nightly build and worked fine +I think in general people consider Maven anathema and use only Maven +This is so we can be sure that no code outside of this package uses the old constructor +Denis I implemented your suggestion by using our extension point +Perhaps this is too simple Ive mailed the dev list +Actually an incoming HTTP request gets upgraded to HTTP by and hence is suitable for being satisfied by the cache +bulk close for release +bq +Hmmm this does not sound good! Ill apply the patch and track it down +Somehow I must have worked with a wrong jar. +rebased on the latest is committed to trunk +Yes I also thought that using only the hash could lead to some false positive duplicate messages as i wrote in the very last line in the issue description +I am using patch resolves the issue for me and the tests in the ec submodule pass +To switch qualified and unqualified +It mirrors also the develop branch so no separate repository should be needed. +Tomaz Muraus Im sorry to disturb you but I need some advice for this patchIf you tell me about some parts to mae up forI will fix it and sent patch again +Changes from bug to improvement major to minor to trunk. +this patch fixes the problem for me Justin Ive applied your patch with revision and deployed a new after release +Anyone would be able to check latest code memory lakes and report would be great +what name that repo would have +I think we can refactor the code a bit so that all the refresh logic is at one place +commit m Refactored parsing and reservation of passive ports into its own class +It has to because it references which is not available on +With the resolution of this has now been fixed +Presumably the underlying XML parser is these correctly for us. +The dist search components need the reference to the +Can you take a look at the log please? A local ewp build from JBPAPP can be helpful as well +This is part of the leaky abstraction concerning the use of a seperate class loader for component classes +I think it would be more safe for keeping scheme and authority to null instead of setting them to empty stringsNicholas do you have an objection to committing this patch +Sorry Guys but fixing a ticket do not just mean integrating a patch on our side +Best is to start a thread on the mailing list and present your ideas about how to proceed +I think a fix should go into that direction to complete the work the compiler does for local variables +Is your caps lock key stuck? +It would be benefitial to easily see what belongs to a deployable archive and how it corresponds to the server by +For now I mark it for two versions and LATER +to vectorization branch +Do we need this for Alpha? I dont think so right +HiWe can keep this issue in that one +Marking closed. +The monolithic has beenrestructured into smaller interface implementing modules anddeprecated +We should have a common mechanism +I understand your argument about the flexibility that the spec gives an implementation but how useful is an implementation with a behavior that is not predictable? Executing a series of methods should always result in the same state +Just a delegate variable +Claus I did not find any other areas that were utilizing +Is this issue still valid? Can we close it +Committed revision +Also since we are arguing that the hard constraints set by us are not going to be challenged in likely usages I have also hardcoded topic name length to have a max value of since its very difficult to exceed this length as well +Rahul can you sanity test the changes on a single node cluster +As a user I think it might be confusing if I have both and I will have to remember about the lookup order +It was a bug in Portal. +To achieve this most of the Interfaces should be extended form +Not sure why I didnt notice that sooner +Integrated the v patch into trunk and branch +This patch still doesnt work for me tests still fail +rvmgemsgemslibactiverecordin sanitizesqlarray from Usersme +Patch for temporary patch to trunk until is resolved +This bug should also exist in Camel +Thanks a lot Nicholas for the review +We could do that as part of or even remove these bat scripts if those new testcases cover the ones in the bat script +I uploaded it into Messaging lab you can find it here mntjbmcommon +I could perhaps see doing it in or later though thats not a definite commitment +Either this works or is not fixable +It would be helpful to know if this happens only when compaction completesThe replicator has retry logic for transient failures but that does not include a response from the source +Thanks David +Ok +So I commented thatThanks it this works +The patch and the regression test are attached +I need to check if its the same for Mahout +Update screen was being caused by the output buffer being flushed on the page include +please see the attach for the revision used is +Regarding the command line tools I wonder whats wrong with my machine +I think k is a drda on byte size of strings passed along +I dont think this will interfere with a releaseif applied before the release nor do I think it should hold up a releaseWe should ask Rory Winston if he would like to become a committer +Looks like it may be a race +I tested this on and onwards upto trunk So the behavior for indexes hasnt changed +Rohini pointed out that last patch was based from and not trunk +Please let me know if you see anything +As of version all above mentioned classes must be removed +actually that would be a good thing to do +It seems that Urgency field was accidently set so now UrgencynoneIt certainly would be a good thing to have so raised the to MajorAny takers to build upon Johannes work +If that does not work lets disable em +Any suggestions on the path forward? Is there a standard maven way to handle the above or is this use case outside what the standard model supports? Or to rephrase does moving forward while sticking with the current model involve splitting the deliverables into two one which would have the required dependency the other would have to remove the class and remove the dependency? Is there something equivalent to ivys profiles? I am not sure about the full capabilities of ivys profiles but I believe they can handle scenarios like this +Forgot to mention aside from the test I also tested a build from a tarball running a couple MR jobs using the LTC +I am jwhite +The interface to be modified isI missed out on adding delegated method in +Oh yes! That will do away with the need for a new entity field like isRehashed +On my old system I had no problems creating Grails projects and I had installed the Eclipse Grails plugin on that system not long ago +If the test suite finds that directory it will process its contentsSending Sending srctestjavaorgapachepdfboxutilTransmitting file dataCommitted revision Now we need an automated way to populate the directory with the files that were removed +If not a subtask of can be filed to fix this one. +only change review should be enough +missed to enable asserts in testcase updated patch to correctly assert offset length +Afraid its still got the same problem see attached build output log +Same comment as before please provide gitified patches. +Hi MirekI think it is a jmx console issue +So I am asking there is need to use clean build when replacing REST web service annotation with or this behaviour will be fixed +Bug verified and I think Ive got the fix +Thanks for addressing all the feedback +Thank you +I think things are good to go as of Andreys third patch. +Why is it initialised differently when looked up locally? I think the only thing that would really be different would be the TCCL +Why do we need to sync inflightPuts and inflightTakes to disk when the channel is stopping? Inflights are only meant to make sure replay is consistent when replaying from a checkpoint +Once the request is sent the number of outstanding requests will be incremented by one +Patch files to add jmeter project files to branch +Can you attach a test case please +Marking as fixedthanksdims + +Testcase attached to reproduce unzip attached in the root of install start servicemix both https endpoints should deploy once deployed delete from the hotdeploy directory +Youve added a new subscription event called NOP which has the effect of not closing the channel in set +I mean users dont care about your automaton at all much less transitions +I tried with patch but assertions still occur +Issue can be closed +This helps avoid mutual recursion problems when loading graphs of nodes with circular references +Sorry for that +If you find further issues please create a new jira and link it to this one. +Moving this to as this is where it will really be needed with MVCC +RE text default that would be yet another default and worse IMO is it would be too hidden of a default +There is also the suggestion to include details about the class attempting to be loaded in the error message. +Such is a base class and +files removed. +Enis thanks for reviewing it. +Heres another go +Thankyou for invesitgating this issue and making suggestions +Are there any other suggestions +Ive tried it again and it seems that it was my mistake. +Both and derbyall pass on both sane and insane mode +Will commit it soon +patch open as the patch is not tested or verified to work nor does it provide completeness of many patch +No response assuming ok. +protectedprivate field but its setter is public there should be no injection too + for the to trunk +Very good idea +Generally what can one do with statics in coprocesorsI think that should be discouragedbq +No other specs on artwork received +They were merged with a beta release but you will still need to manually merge them with a good diff tool +When deserializing the configuration on the grooms does the classloader refer to the path where the jar has been unpacked +Stefan there are plenty of examples where we provide a convenience setup in Arquillian for systemenv property +I was hoping you would create a test case +checked in JSFUNIT +The answer was never return null for the array return types +Oh hrm +There is actually a deployment classloader that is available at the time of the rewrite it just needs to be passed to the endpoint publisher +If there is no foo but there is foo you could map foo as well and it should get redirected. +With my companys asinine IT policies Ant is just easier to work with +But honestly I think we should just update the rendered code and then report any improvement found to the Snowball ml and keep track of it in the package readmebq +Missed in the original ticket was which is pointing to but isnt redirecting to cloudstackDepending on how graduation moves it might make sense to just let this sit for a few days and deal with it as a TLP but I just noticed the problem last night and didnt want to forget hence reopening +As an addendum to this I believe that I have updated to Leopard in the meantime +Minor change increased the timeout for test +What if this second call has some? Would the user table be dropped in that casebq +Committed revision stuff to optimize no longer andHopcroft Policeman is back +Pushing to +In my experience something like of uses of do not require thread safety +Thanks +Here is a testcase that should work once Tuscany support the INOUT pattern +Anyway we will change the Axis components to JCL as you have suggested +I dont think it would be too hard +I opened +One of them needs to be marked as a duplicate so we are tracking this problem in only one issue +management server log file this issue while deploying vm from new CIDR in Advanced zone shared network as well +blocker for one is fine +bq +Updated and transferred to the wiki +I will fix that in next version of patch +Sorry about that +No response from submitter to April question I posed. +The test also includes a table for adding descriptors of methods which we can ignore because the spec says they are allowed to raise this exception +Observed duplication of persistence and hibernate configs +I just committed this +This allows secondary NN to transfer the compressed image to primary NN thus reducing network disk IO overheadJeff Id like to take a look at the Avro file format +We use iptables to forward port requests to where only tomcat is running +this seem to also happen for rhc command line so for now assuming this is a server side issue +push out to +I am going to create a branch off the branch so that we can quickly iterate on the patches on a branch and then see if it all falls into place +but basically we just check the axis order in world and swap when we have a latlon ordering +My patch made the code more difficult to understand especially for the twice access and volitable value +mvn verify takes about hours to run with the unit tests +Not sure who gets to close these but I believe Bryan does. + hours later the list again appears to be perfectly complete +This bug is a duplicate of. +Hi Danielyes I was talking about thatMoving and to ANTLR should be easyWe can use this Jira to keep track of this work +moving to BETA release +is currently IgnoredYeah for some reason it was failing constantly on freebsd but passing on other OSs Ive never gotten back to itbq +This works great +JonYou are right +That should be changed since you have removed property +And the test is successfulSo here is the patch with just the unit test +tentative patch +Yes +It does not for me +The components build off of the Maven repo stuff so the comments wont get picked upCheersEric +fixed in +Go for it. +Late comment this patch doesnt have new unit tests for its correctness is verifiable by the existing test when s fault throwsTest should fail in this case rather than merely hang forever +If there is an exception attach it to this issue but first check that it is not a common problem or something already solved +The problem only occurs if you add the server location during installation +Forgot to mention this patch also covers DanSame suggestion as for could you please add unit testsThanks +I suspect problem may exist in +And I was not testing that kind of use case +In this case a CAS is generally going to be just about the smallest tick on the CPU consumption of what is after all otherwise going to be a blocking handler +Of course there is code behind the configuration but the producer and consumer are none the wiser +bulk close for the release +Proper patches and links to rest of docs to be provided once review comments implemented +fixed misspelling +I committed this +Every little helps +PatchPlease review and advise +updated checkbox label as discussed with max on label for binary deployment checkboxdisable automatic maven build when publishing to +Thanks a lot for the review Tom +committed +Make a test with curl or something else which does NOT wait for before sending credentials +new patch using existing jclouds stub the attached patch on hbase and cassandra +bq +Hi Carlo can you please commit the fix to the JBPAPP branch? Thanks! +Want to write a patch for that +Oh I had a patch +Tests failing due to dont think this comment is correctP The failure to become active is already logged aboveThere is no log above +This patch doesnt work with Orchestra releaseThe problem is that this introduces a custom that is always installed automatically when tomahawk is in the classpath +Sorry I didnt update the bug earlier with status +I have found the problemIn fact the maven repository is not updateI have test with the nightly build of and and thereare no problemSo I close the bug and Im going to change my maven dependencies ruleBut it will be perfect if the ibiblio maven repository is updated +I see even the current release is unusable +Which page is this on +it seems that you use the RMI client on a Jackrabbit repository +sample project which I uploaded on this page works correctly with wicket my project on which I found this issue DID NOT work correctly with wicket +The cool part is that we reduce a bunch of the Maven dependencies with CXF and we are eating our own dog food +similar one marked as to Major because of workaround +Ufff finally I got it +UF has support for this +We still need to add test +bq +but I believe that it also adds a bufferingthanksRich +Helma since we are already happily using Thiens work at our site we can close this issue +Looks good to meThe motivation for the change from the good name value to a bad name i is for consistency with the RI +Hopefully we can get this back in for in the next couple weeks +Committed trunk revision Now backporting +Thanks BenI know DC is not standard we needed som coexistance of CVS and SVN while migrating from CVS to ll check in my pending changes later today +The attachment includes the WSDL the service implemenation class and the test client +Thanks Markus for the initial Patch i committed my modified version inTrunk Rev Rev Rev patch breaks the http https redirect +patch looks good +simple and good for adoption so lets do thisCould you attach a template for what the different test types should be like +I dont know purpose of Resource sure it can be fixed in Beta as you wrote its more feature than bug +Done in revs and . +Not safely there are still issues regarding HTML parsing with Tika even without this nasty boilerpipe hack +Except that messageId is very straighforwardly generated by an increment on an static Right +Its true that an interface would be cleaner but I dont know if such a refactoring worth it at this time +By examining the origin of this issue I concluded it can happen on any TX cache no matter if its pessimistic or optimistic. +bulk fixing the version info for and all affected issues have in all issues not touched in a month to move issues to and +The question is whether we should even try to return that many services I wonder if the spec ever talks about some paging mechanism like there is for subscriptions and the use of chunk tokens +If you want to actually execute on youre going to have to use a retroweaver of some sort to do it +I cant reproduce this with a clean installation of Sonar and Jira Plugin so issue closed +I have attached patch to address this issue +It should definitely be possible to get the facts from the parent +Just to clarify I need to uniquely identify a connection to use as a correlation id for messages generated from that connection for possible downstream resequencing +Please test and resolveBen as fixed against revision . +Otherwise on would continue to execute even if in reality no documents were imported then that is not so useful +Just checked again In listStatus resolves symlinksIn HDFS listStatus does not resolve symlinksI did find this conflict when I was doing and followed HDFS convention and uses get +I was able to reproduce the hanging test test locallyThe stack trace I obtained was short +Attachment has been added with description unified patch cause several of the test in the NIST testsuite which worked before to throw exceptions +The deployer is looking for domaindefault servermyInstance namedefaultAre the domain and the target name correct? Also is myInstance the name of the server or the profile +Sorry about that +thanks +Thanks for pointing out our sloppyness Lars +It works for me +The priority will be held within a thread local context so that it can be propagated through any subsequent invocations of theIf the transport is JMS invoked through the JMS Courier then the priority will also be used to configure the. +I will try with the configuration you are using and compare it with mine +Any Luck +I think we should also consider passing the sure if this is related with this issue but why we need a method +A system call every few hundred bytes though might be tolerable +I beg you +Ok after some offline discussions I am now tending to go with Viveks recommendation due to reasons outlined in his last comment +IMO you should use instead the to initialize and destroyYou can set credentials in the initial context propertiesMoreover if you use the same initial context instance to initialize and to call close the surrounding instance we get the sameFor that be careful to set the property destroyCan you give it a try and let us know +lgtm have just committed this to trunk andThanks a lot Ivan for the reviews! +server side state saving +I think all this patch is trying to do is establish some sensible defaults for script resolution when a request like that occurs or if Sling doesnt become more aware of Accept header +Attached is a patch of the changes I made to the application +Uploaded patch +Addressed by work on the X builder +If you check the patch youll see its actually about both +To implement this randomData would have to be made and the setter would have to recreate it using the supplied generatorIf this is distasteful I would suggest as Gilles mentions above just using directly which provides direct support for sampling with configurable +attach a new patch rebased on latest trunk +Thanks. +Keith Please dont forget that this project is based only on voluntary work so contributors chose what are they working on and what questions will answerPlease check the archives of dev and user mailing lists since this was discused quite a few timesThe bugs are filled to track issues to disscuse problems come up with ideas for solving and to submit patches for fixing themCezar +it runs a java program +The files and do not apply to the trunk version +assuming replicas isnt available? A dont change capacity flag would be very useful or perhaps making the data structure much more conservative about resizing. +This would be interesting to try +calls grow on arrays and other collections except Sets looks like an easy fix. +Can you help to review Michels patchRegards and push this patch +Ill let you know as soon as Im getting any and regards +Attaching +new version of patchrm code +I forgot to add that comment to my pull request +BangCannot see from the code why this happens but it seems to be related to usage of the +Thanks Knut +This is a regression caused by r with ironic log entry Almost all changes and fixes to get arity working correctly +It uses the thread context class loader if set +Applies to trunk and branches +Closing the jira commits already closing all issues in resolved state that have not been reopened nor commented over in the last month +CP reverted in revision +trunk +i have generated this document with openoffice and pdf exporta tabulator is missing +Thanks +The java file should be put in the right place if this patch is accepted +Jan could you attach a simple test case that reproduces the error and which is fixed by your patch? Well commit immediately once we have a test case +What would be a good place to host this project? I couldnt find a way to email the owners to find out how to submit patches etc +The one part Im still trying to think through is if configuring the custom factory on a global scope vs per source is sufficient +The SPI implementation may not have the property definitions at hand when just the node info and the property name is passed +Start and end parameters must be in hex to accommodate binary columns +I have resolved the issue by removing the local variable context who hide the superclass variable +old title was allow to recognize elements that contain minus dot and colon while keeping underscore a valid char +Tests in branchxsolr all passed +A sensible message expiry policy can help in tandem here +May be it should throw them away like would do if running in a safemode +Jan Cwik said on the mailing listHiYou are right I switched ODE to asynchronous mode and I got theexchangeAccepted event +I think that we have figured out that this happens because AST transforms interfere with the AST of the source code so that it no longer matches the AST that JDT sees +My approach is really hacky as it requires a custom spring social webapp +to TRUNK +Marking LATER pending a course of action +This jira is a natural extension of +you already tried with +If it starts happening again Ill see if I can find more specifics as to what in particular is happening at the time it happens but for now I suppose I should close the issue as something I cant repeat. +Yes this test is failing at exactly the same spot as it was before +Assigning to me for investigation +Attaching a patch and html file to fix this issue + Username triathlon +Note that serialization test I added fails for the fastbinary protocol because it the Python checks I put in + Test Test BernardIs this problem different from the one you reported in +IMO this would be more consistent with the way annotations works anyways A better approach however could be to turns things around +Ultimately my problem involved passing an empty string as one of the arguments to exec +Above tests failures are not related to this change +bq +Erick feel free to take this one and iterate as you see move its clear I wont get to this in the time frame so if someone else wants to pick it up feel free +You may try to apply his fixes +Following up on this one because I was looking for it myselfIs this all thats necessary for checking proper headers? Do I need to add them to each artifact or just the parent +Apologies the previous patch is not sufficient I underestimated the breadth of the configuration +This can be closed now right +I dont have anything to attach +verified fixed +When thread pool creates a new thread then the thread inherits the current thread context classloder +How about when we are going to create new ORM mapping by using Cayenne modeler? Should it allow us to open a new Cayenne modeler instant from the Eclipse IDE and do the ORM mapping +There is a new signature for invoking the parser that allows setting the property map enabling the desired properties to be aggregated +Moving to due to release management restrictions +Each Yarn application can specify its individual retry number in +I guess Im confused +Do those work ok +Following is the screen capture for the project structure +Looks good! +First attempt on patch +Wrapped lines to reduce width of generated page +htaccess filesThanks anyway +Regards well its not this in my patch but from what I see needs it so it cannot be protected nor private +When we have more better design it is OK for us to switch to the new api +resolve as out of date because of month no activity +bulk fixing the version info for and all affected issues have in all issues not touched in a month to this issue because its a small piece of what is trying to do +Dont randomly assign the issue +Im closing this as a dupe as the discussion is happening on +Better more relaxed then too strictNot necessarily +I am resolving this right now as all the links have been fixed as part of +I may have been a bit quick to state that my disk was full +has been upgraded to and that contains the fix for the on EAP and +They do however affect each other with respect to timing of the locator release and round trips to the server +We need to implement it I thinkKoji Thanks a lot for giving the related jira id +Implemented prepared for next release new table in DB must be out +this was created in anticipation of work to be done on installation scripts +It makes a huge difference. +attached patch should resolve this testing by Michal Fojtik on dev mailing listpushed to master with commit cfaffceceafc +agreed? If so you might as well change this issue to read Its good to have ADB support all of XSD +Fixed on and trunk +Thanks Jimmy. +Id say we should +There is a bug that is starting a pinger for CUSTOM services +How do you think +In steps above I mentioned lib folder just for example +My commit r was to highlight this issueAs this JIRA has changed to code base it should be reviewed independently of +for the patch +How it is sent to the secondaries is up for discussion +Why does one need to remember any URL in the first place? Couldnt we have a root webapp with a central landing page containing links to all applications and resources +no follow up for M so removing it M it is then +Thanks Suresh +Thats a good question +Hi Simone +Added an appendix referenced it from the first chapter immediately after it refers to the SOA Overview Service Oriented Architectures +Updating of the node consists of deleting the old visual element deleting source node from the dom mapping creating a new visual node based on the changed source node creating the new dom the attached page after updating comment that is in uidecorate tag uidecorate gets updated together with parent uicomposition and uidecorate contains a lot of other tag inside so all of them have to be recreated too during creating of the new visual element for the changed node +Unfortunately we cant account for Hibernate bugs +So file is the right patch to try +And when you look at frameworks with superior URL mapping capabilities such as Grails this is a must have for WicketErik you saved my ass with this one + to trunk. +Glenn is it possible for you to attach a reproducible test case or at least a derby stack trace +could be replaced with +xmlrpc currently you have to give a project ID and its a bit hairy to configure into your svn post commit hook +Do you think flaky unit tests should be pulled out or put into a separate categoryI think the correct approach is to revert the committed patch which breaks the build +Agent needs debug enabled all lines are INFO +No response + code style +I will dig into it +Is this still a possible issue +let us open another issue if the testcase needs to be fails as didnt make it into trunk +The Cluster Summary prints total capacity used disk space etc + noncloud mode warn that implicit is being used +The user can then discard that work and try again as per the users normal error handling will only occur to transactions with persistent acks in them +My mistake +ok. +Also can we have a more descriptive title? +Let me know! Thanks again patch revision +Here is the patch +Try searching the documentation for +seems like a good ideaThanksps Thanks for answering so fast +Mixed mergepersist usage would create unclear semantics for storeOwner callers here +Just use the signature that has in itinstead of +But the premise is that Tom provides the suitable exception and error message +I believe Tyler is correct in his diagnosis and I think that the patches should fix the issues I am having +Hi Simo I Fixed the patch now is ok ciao ciao +Adding EAP to affected versions +This has already been implemented. +Shall we close this one +Just out of curiousity is anyone using the web services app client? Was wondering if we could simply remove it along with theThese two are actually independent +The constructive thing to do at this point is to make a patch without that one item then perhaps pursue that separately +Thanks Jon +I like it +Hi JamesI am currently unable to reproduce that error +I think we have a clear feature request now +Issue may already be resolved may be invalid or may never be fixed +You dont happen to know which classesjobs do youIm working it out just committed a bunch of changes but were not there yet +Both worked fineSince only is changed by the patch and unit tests are not related +Thanks vinodbq +I know this must be frustrating +Until then it will be some sort of a work around to get the latest state changes + +Also working applications may see a regression doing a soft upgrade from or to and releases subsequent to +I dont see any reason to keep it at all +Hi I tried to check RC but Simple History still showed SOLR REJECT many times +Review up +Oh +Unfortunately its at the cost of readability and maintainability +The deviceready event never fired on iOS even with Cordova being a top level script +Im afraid that one will be difficult to fixLets see if I can move this to commons sandbox for information +is fixed so is this. +This looks reasonable to me +Thanks for reviewing Matt +All set thanks. +Applied the part of the patch that did thePlease open a separate issue for the stuff. +Unfortunately my experience is that if a predicate matches for a client and doesnt matches for the others the matching predicate sends the response to all clients +To reproduce the error simply run +Good luck! Leo +In It appears the JIRA link takes you to the page +Resolving as invalid +But if you apply s patch this one becomes patch available +Im not sure about the null either +The patch explicitly setJar so that can find the resource +Was the patch ever acceptedve got the same error with release +Needs to be backported to as well +yes sounds definitly like a good thing +I guess this issue wont be solved until someone gives me a way to reproduce it step by step starting from a clean install and its demo data directory +Thanks I actually ended up doing something very similar but itll nag at me until its fixed I did find out that you can extend the and override the comparator if needed +Resolving +System tables were not getting flushed +Pavel thanks for your attention +Thats a reasonable explanation +Since loadEdits already hanlded stream close +Just in should have noted this earlier +I know this +Tough call +Ellen and pageflows are different in this respect +That seems to be a only problem introduced by +I can also help out with the docs but I currently dont know how patch with thanks to Christian I also fix some issue on the HL data format which could cause the test failure updated the karaf feature and made the integration test passed. +This patch add a configuration option +k changes think I would prefer to get a back as this falls into the general category of that node is pretty busy right now and I would take the same avoidance strategy as with other cases +That was always the behavior even in the failing cases it was always present in dependencytree even if it did not appear in the class path in the appropriate place +ESB is now running CI against and jdks +It certainly could be a datastore bug but I dont believe that it is a TCK bug +Hello Thomasdoes this patch also addresses the issue reported in +Ok Ive applied the change on both buildtestsuite for Branch and HEAD. +Its hard to fix as redirects in flows are statically produced so Im moving to as its not that is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Does anyone have an idea about an official released fix for this +Should test against a new release and see if you still see the issue. +Reopened to try to merge Desktop with original spark and committed to developcaaefbfeffcfcefc +OK +Thanks +Now I remember what is related urlfilter is the first filter then it fixes urlfilter is the last filter then it breaks web +Made the above changes to the patch +The will look up the security login module by the realm name instead of the app name +about that +postponed +Yes it makes more sense to be on the lowest usable version +Started work on +How could I have forgotten that? Thanks Knut +revision +Note All this can be achieved by changing the build files in SVN +alphan Bn in a jobthe will receive tow parameters duringfirst one is summands list which is a summands list composed by the matrix paths seperated by comma +change the getMetadata request from to in +No decision +Includes little test +Release Note now obtains a different sessionproducer from the connection pool to be used within the scope of each request rather than sharing one across all invocations of the router +there are good guesses possible +Fixing last two compilation issues +Hopefully this will all be done in a week or so then we can move forward with patches + have been done +Each generated class file represents the query plan for an SQL statement processed by Derbys SQL interpreter +Patch committed in revision +Actually posts the documents nowThings to do Need an icon for the documents tab SuccessError handling Test other content types File upload Document creation wizard Better layout of the form etc +we are sampling the first max splits +That is not enough to move documentation to the rd line +Suresh thanks for letting me know +Just wondering Why would this be used in the Scala script? Shouldnt it just use SLFJ Logger +Attached patch with extra test +I have no problem with that +Nitin why do we have to drop this table? Is it no longer in use? Ive checked you were the one who introduced this table +Agreed +Note the patch still needs more tests and needs to check headers etc +This functionality is nothing like that in Hibernate +Ah! yeah +so we should probably expose those things in the context directly +Which explains why the classloader was not initializedNow the initializes the first +However it is likely to lose precision when constant values are used as function parameters +Uploading new patch +to correct fixversion +I didnt do any measurement +Changes applied in trunk r. +sld file instead of creating a new +Werner the close fails every time see comment for a description of the culprit +Should I apply it for branch +The issue could be closed +The problem is after seconds we assume gossip has propagated and forget about nodes that have been removed + +Because you have not configured it tries to find a suitable logging implementation to use +Ive updated with correct information +what are the possibilities of bundling spring javascript now bundled as part of webflow +I think the issue I had with data loss after node restart actually had to do with this bug +I have this issue at our place I want implement your solution +No feature of course but this is not stating a particular change to Mahout +It probably does but I shall leave the decision to you simply because you know the library better +I think the pay off of this change wouldnt equal the amount of time for deprecation refactor etc +Also its not clear if you can use classpath to add to the standard maven classpath +adding specification patch removes the member functions from the interface +hmm thanks AlanI am looking in the types branch +Running through patch addressing two defects in the older one Incorporated Rajivs comments on exit code +I dont knowwhat your facade does but you might want to look there +Closing this issue +Wouldnt something like sverify be easier to remember? Personally for me Id probably forget svalidateEquality and have to look it up in the docs +bq +How to fix this bugHF link extension should not change URL for deletion linkExamplecommunitydocsdelete instead of communitywikidelete This ensure that deletion is done on required page without Article navigation that cause exception +Ive moved Dmitrys issue to it is a separate problem +Otherwise it will not specify collation +If I add something I would like to make it as simple possible +in trunk and Beta in +You can setup whitelists allow administrator approved controls or place their internal site in a security zone where this will not be a problem +It make sense to provide the top score somehow +Whats the implication here for people who have code that doesnt set the to auto close the stream and then upgrades to the new jar +patch committed under rev +Looks good Nigel +Stefano thats set up now +Verified this issue with latest build from ACS branch +Bringing back to and retitling +running +Ill change the proxy thrift to have its own data types +The thing that bugs me is that the documentation and the behavior do not match +Can this JIRA be resolved with todays testsuite commit +Closing. +Sure close this issue +I think there is an issue with patch id +and some additional variations +There were also changes I had to make in andHowever based on a comment by Steve Loughran I decided to add the abilityto scale the timeouts at run time instead of just poke in what worked forme +I would be really happy if after the next RC there will be a couple more because I still get other crashes in the area of session expiriy and others +Attachment has been added with description Patch to make FOP ignoring too loose or too tight lines +Patch committed to issue has been fixed and released as part of release +have you also considered Strings that only contain blanks? Like or t etc +Definitely no hurry +Here you go + + +However the knob settings would be lost when you exportedimported the database +You cannot override the initializers from it is the other way around + ESB can be found in and directory +Its broke +Implement the concatenation described by Scott CareyIm assuming that for concatenation the following would be considered reasonable behavior Only the metadata from the first input file is written to the output file +On and trunk the file above seems to work fine for me +Appears to pass basic system tests +Yes +I have just commited a new fix for that scenario with the getProperties on theCan you test again on your system +I hate subcomponents in jira why doesnt it allow updating fixversion based on parent one +actually a better description is initializing lazy objects +Attached failure stack trace found from the server when the test attempts to deploy a +I changed further config files so that an icefaces application works fine as well +It passed all the regular tests for me +Note that this will not solve problems where the diacritic comes much after the text chunk it overlays but we have not observed PDF files like thatSending trunksrcmainjavaorgapachepdfboxutilSending trunksrcmainjavaorgapachepdfboxutilSending trunktestinputSending trunktestinputTransmitting file data +Please review +Fixed in r and of branch and trunk respectively. + does contain the entryand looks to be fine +the parser in its a lot more permissiveOK that explains. +Would it be possible to backport this change to the SMX branchThanks srcmainjavaorgapacheservicemixjmsendpointsTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixjmsendpointsTransmitting file dataCommitted revision Backported to branch +Updated the patch with modifications to +Still need to add to the wiki +Thanks David I just changed it to this version instead +thrift +Do you happen to have a simple testcase handyThanksIm really sorry looks like this bug allready has been solved +This patch has gone a little stale now +Until then as long as the controller doesnt failover it will not be elected as the leader even if it is in the ISR since it is one of the shutting +When building the comments all n will be replaced by n +The current implementation uses Java Ruby RMI Javascript and Axis extensions +I cant right now the system tells me its been edited by you +Was committed as part of. +When servers upgraded they are all stopped then new started one by one +The added testcases passes +I do not have any new plans or ideas about this +Seam is more like extensionintegration of nonEE technologies with Java EE +Closing it now +The fix Jessica provided works only if the zone is not the first zone +This patch covers a much wider portion of the system in accordance with the feedback +Also like Rick mentioned can you share more information on the casting part of the syntax +Moving to will not be a moving to for now +Setting the target version to for now +So all issues not directly assigned to an AS release are being closed +Could you first post your question to user groupuser +Hi AlexAs discussed Ive incorporated your latest comments and recreated the patchThanksThanks for changes +Theres no easy way that I can see to know if the DNs block contents were considered stale due to startup or having been made active as we only store a boolean in the indicating stale or not +Update the patch against the latest branch code and add the code to initstart and stop the writer in case it is a the me +Adding annotation in you responsible for the two new javadoc warnings Chunhui? javadoc +For those users should be and cautioned +The producer audit is one example of metadata that can be shared but last and inprogress transactions need to be separate +for the patch +Ping? Sorry to keep asking for this but it seems like a simple fix which would prevent lots of headaches +If I move my Seam apps to other App server other than will SSO still work +Rob I dont have the same behaviour as you described +There seems to be other issues with how a hub handles a bookie failure +DonaldA mistake was found in the monitoring porlet which is that it cannot successfully enable the registed serverI submit the nd patch to fix this so please review and commit it +It will depend on users configuring an external forum. +Thanks Abhijit +push out to +One might want more examples now but thats a separate point. +Implementation of version of project could not be found +As Edwin knows I have experienced the same problem +tests included +Sounds like an interesting idea +Clicking the OK button resets the password to Am I working with an old versionThanks in New Password mode if the field is empty the button should be disabledLet me verify this on a Windows VMBTW we released a new version yesterday Apache Directory Studio +Ive attached a sample maven project could you please take a look at it and see whats different from your project +Dont worry about codestyle I can take care of that and any other pieces +But a poll should not block for ever + It turns out that only uses it + + fixes this +Issue resolution delayed until following addressed in maintenance release +The problem still exists with the bug is fixed the assembly plugin Im using wont receive accurate information +The log is nearly identical +My patch still does not fix TCK problem +Fix to be confirmed +Ill submit a patch as soon as Ive worked out how to use git +We will use such counter to track every region state change in the system +trunk I have fixed the issue in to type converter registry on restart +Is it the +ve committed this +Ignored the two failing tests in faaffbaaaefcbbbcca master. +Looks like those http test case failures might be related to this patch +Out of five resource types currently i am working on the addition of CPU and RAM as new resource types and adding the methods to set the limits and to get a count of these resourcesAlso adding check and incrementdecrement calls for CPU and RAM in VM life cycle +the problem that was fixed was that if it failed it would get stuck. +A tile is a box containing lots of points and is used as a quick substitution for all of those points in the box +Screenshot attached +Here is the patch against master +Just dont uninstall apps assemble a new server insteadIn particular I think that altering the behavior for mulitple server instances here would require some kind of registry of all the servers running off of a particular geronimo repository +I fully agree on the standardizations and had similar thoughts while creating the patch +We check the problematic relations before refresh and remove all Entities which have no identifier yet +Im not sure what the description mean with reusing the file descriptor +Waiting to here back if there is a planned release of rhq before is release in which the deprecated property has been removed +See for further details. +If that does not work please provide a fully reproducable test case +I have updated the patch according to your comments +This was done intentionally +moved to since this is a +Im not claiming I understand what the issue is anyways it just looked like it would be more sound this way then introducing a magic timeout value +Probably we should have a first mail on to ask what developers thinks about it. +Havent got a chance to write a test case for this yet +There were no new findbugs warnings +Could you please have a look at this old bug submitted by our fried Geert BevinAfter this bug is fixed I think its worth releasing +need to consult with fbricon about that +It looks good to me +just buggy +Are you EOLing EBR? In which case switching to Maven Centrals the right thing to do anyway +I have made a and if we agree that we want to continue with the Bizhan version work can begin +I think it would be nice to have this done in as a minor bug in +the spreadsheet from Jan has red crosses for messaging subsystem attributes profilesubsystemmessaging +Looks like our Andrew saw this before we did over in and are only interfaces in +Should probably note that the patch is valid from the test dir not the top level +Removed guava dependency from the patch for +Im fine you just attach the bloomfilterfpchange addition here as a separate patch and we will just close as Duplicate or vice versa weather you prefer +Please review +Committed revision . +Am I right that you are using the quite old version? Did you ever try a more recent versionIf the pdf is somehow scrambled what kind of did you add to those pdfsIs it possible to get a hand of at least one of these files +The patch passed all of them +Thanks for the answerI cant really close that JIRA since I dont have permissions to do so +So it is about facilitating component reuse I guessI will supply a patch later todayThanks +Make RM provide heartbeat interval to NM +Michael isnt currently being updated +implemented via the new stale resolved issues +It would be nice if anyall incompatibility workarounds were included in a single module for GAE users +I had been using Reuters but Im not convinced that its ok to include it and I was looking around for something better +Suggested patch please review +Looks good. +Will if that is enough +Tika has a number of unit tests for the extraction of embedded resources from Word documents in Are you having this problem for only some files or all? Do you get some all or none of the embedded resources out +I forgot that I have a question +Any chance you could apply the patch for me so I can upload mine +If this is a temporary problem please let us know +This is for judging if clientserver are compatible or notWe could either send minor version or a list of method names +please run a complete build of all of jackrabbit with +Introduced with the fix for reopening related issue +I merged CDI Perspective into JEESo there is no CDI Perspective anymoreShould we do anything with Seam +Well try to have a fix by +Not a problem if we upgrade to +Username gohde +Attachment has been added with description updated sample fo file shows problem with footnote on repeated this bug is not fixed +Also since recreating these objects is that expensive we should explore the option of not combining for every single new message but in small batches +BULK EDIT Moving the unassigned features and improvements out of +Now the application context is created before the Tiles container +tests are due +Btw +Bug has been marked as a duplicate of this bug +As a first note this seems like a good use case of why moving into Hive is a good ideaBut in general could you explain more why isnt a good solution? If it doesnt work with Hadoop wed certainly be open to patches to fix that +The new patch addresses Alejandros suggestions Moved and associated tests to login module which now produces a WAR and a JAR Removed addition to Updated documentation to reflect patch fixes the build issue +Usage question not a bug +A partition is deleted from this list only when it is deletedRest of the comments are addressed +I have fixed the failing tests +There may be system specific ways to clear unix cache but only way I can think that is completely safe is to do setupcreation of a large table shutdown machine do one run on the table +For publishing a snapshot we should set the version to and increase z for every snapshot +Avoiding explicitly to send the whole stack trace +I have added a few but I am not able to find these onesand I am not sure about this. +Now got a full test run on the branch on windows +The form submits with POST by default but it is not possible to copypaste the URL and send it to another person in this caseRead about redirect after post pattern +Ive made changes to depend on Apache Yoko in revision When releases are available we will change in the properties filePlease check if this solution works for thanks +Hi AllIve just started using display tag Libraries +I have made one more attempt to apply this patch after wainting for and a half years and hit the same issue we cant get ddlutils +If you really need that many field names the only option right now is to increase the size of the perm gen +Should be good now +bq +Pushing issues to the next CR +I also added some logging for +Can you provide a simple sample that illustrates this problem? The line you reference is just doing some memory allocation through the allocate call so I wonder if theres something else going on thereA couple of other questions What is count newCnt at point of the crash and can you trace it down into the allocate routine? Is the memory manager valid at the point of failureJames +seems to be the correct issue according to your github link +I only see the lutter acctIf you and the other names havent done so already please create an account on confluence andthen let me know the usernames +Is this related at all to +we did not turn on validation yet +The following file is a manually zipped files + +This can happen if the database recovery needs to be done at the boot time or it can happen after the database has booted and user has issued a SQL which requires Collator support +PradeepDoes Stacks comment not address your needbq +In this case How to get rid of this issue? +Thanks Ill get to processing the pull request very soon +Is it still a problem for you? If so please try it out on and +I will commit this patch tomorrow if there are no other comments +Original documentation shows incorrect code that will change root element instead of adding child nodes to it +it works fine for me too now +Building a +Please add this script to in the posted zip file +A suggested sample webapp and documentation along with a maven project to build the docs and webapp +Sorry for the inconvenience I caused +A quick workaround is to lower the log level for +My first guess is that out single background daemon is stuck doing a checkpoint and cant service the switch request but I have not checked the code +Yes its doable +I am on this Thanks for the quick fix +Johnathan Yeah I edited the steps above with a better benchmark approachPavel Yes JNA is no longer required +An news on this? +thanks for this Andrew +They could be small bumps on the things they are connected to +Will post to RB when finished +No worries +This patch goes on top of +Thanks! +I could follow part of your work through the apache dev listThanksAlvaro +It would be interesting to somehow call that Action via Ajax but get JSON back instead without the Action having to return a different result codeIm taking this off for now since it will require much more work and discussion than possible for a pending release but I hope we can incorporate it in a release soon +hi Sebastianwe will upgrade the installation in several days and take a look on it +The stacktrace of the patch to solve the Problem +It seems the case +May be I am missing something +as described above the Clock approach was a false start and should be removed now that we understand that is a dead end +And I expectthat a private method with only one caller would get inlined by the JVM anyway +Still waiting for Hudson blessings +Sure enough all the unreproducible errors could be tracked to the top two nodes in the rackSo if you put all the copies at the same height on the rack the probability of losing the files at the top of the rack is definitely higher than the probability of losing the bottom of the rack +Thanks! +We should probably just create a modified file from the bundles available from the Oscar Bundle Repository since we wont be officially releasing Felix bundles just yet +from me in principle suggest waiting for Terjes test before committing tho +This patch makes all calls consistent and eliminates the race condition +The second looks more like final codeWhat can I say? Its pretty obvious it is better +changes look good to me and is aligned with wrt +The difference between the compiler and groovyc is that the compiler is file based and uses the eclipse infrastructure to inspect the classes +Thanks +Thanks BrunoYour patch is in trunk at r +Thanks Bryan for your comments +Jeffs comments are correct something must be using that port +Ive created a patch from the changes Adrian suggested +Thats what I ended up using +Thoughts +Is that reasonable +this new patch is showing a modest improvement of about +And i always do drop table and create table stuff and i can switch machines and only need to redo the create table + +Id like to revert this commit for the following reasons for the full smx distribution snapshots are provided in the system folder so it slows down things a lot to force the lookup in the apache snapshot repo for the kernel and users installing snapshots features we can expect them to build the projects locally or ask them to add the trunkassemblysrcmaindistributiontextetcSending trunkgshellsrcmainresourcesorgapachegeronimogshellspringetcTransmitting file dataCommitted revision . +The description is slightly wrong +But can anyone explain what kind of technologies will be needed to implement this +Simple case of producer flow control kicking in when low memory limits are set and no consumer present on the queue. +Chose not to deprecate since its minor is not yet released +I have developed a file to do this automatically as in the Linux case +The interface now conforms to the current rest interface +Ill make these changes and would submit an updated patch +Which release should this tool be packaged withFor users running they would expect this tool in the same release so that proper action can be taken before upgrade +The following incompatible changes were noticed between and caused by +Is that enough? What do you think Mark H? Does this patch still make sense +Updated avro deserializer with literal mode to be compatible with literal mode in the avro serializer +We never got this addressed in Camel timeframeIf really needed we can address or look at it again in Camel or later +I like Hardys idea +Is there someone out there who can verify thisThe structure you have drawn as what it should be is stating that the submodules inherit from myproj which they dont +See comments for the resolution +I also think that reducing the number of restarts on a per module basis is more practical +needs fixing +Isnt a response header? Wouldnt the client be sending a Range header insteadFinally how can I test this from the brower? Can you provide a simple HTML page I can use from the browser +Patch fixing the problemCommitting shortly +Apparently it was I got it working I was able to put back the version without problems so that was not issue can be closed now +This has been reproduced with using +Resolving as WONTFIX. +Rebased on Todd thanks for posting a patch +wasnt fixed +Thanks for creating this JIRA +Hi Carl when we execute select query it is not invokingLet me know if I am missing something in my analysis. +Suggest you include them when you apply the patch +Carlo Ive got a few questions Is this something that worked before +So making the change will need a more thorough investigation and clear design proposal +Passes all tests still needs units tests for with SASL +I think were good to go now. +I work on a small project were two developers we dont have an upstream repository but we still do releases + +Patch to fix DRLVM +Marking this as a critical issue for +robert I agree this could be a problem +This may cause problems in css +And I also see this deadlock with the release candidate +Or is there some sort of technical problem? +Fixed in revision +Id have to test this to be certain +Also another problem is that getReference is binding the class name to class implementation and not +I tried forcing instead of for but got the same result +in our with Max discussion we find reasonable to have properties file be generated +Backwards compatability and casting rule this one out +All yours Sami +Verified seems ok thus closing. +What am I missing? Does each data representation need to loop over the schema itself implementing that logic? Is there an example of how a record is read somewhere +As long as I have access to the datasource object its all good +Requires python or above to work correctly due to Queue module +Not quite ready yet will put up a new patch on RB then +Thanks +Click on test event test +Let me know if there is anything else I should be doingThanks for all the help Dont worry you should have seen the hash I made of my first patchHudson is misbehaving at the moment so Im not convinced that the test failures are as a result of your patch +Otherwise the dfs client has to handle the issues like synchronization and out of sequence packets etc +Please verify +batch transition to closed remaining resolved bugs +Sorry I meant to say same error but different trace +If we eventually need a for some other reason then maybe we could move the method to it +Unassigning for now +Any chance you want to supply a test case? +Fixed at r thanks! +Grape should try to handle the exception and give a meaningful error report +Patch isnt complete +Resolving the issue. +It is intended for HadoopSorry for that +All Thrift transports are for serialization and deserialization and Thrift structs dont just pull in a big blob of data from a memory transport they get exactly the number of bytes they expect to be there +We can probably read theold format write the new format if this is deployed +you have to prove that the project owns that domain to use that group only see yes you need a new bundle +Revision . +Since initialize swallows most and returns a boolean Id update the throws javadoc to indicate an IOE isnt throw for most error cases +Perhaps this is not an issue in Tuscany point of view can we have an update on the matter from one dev +In the previous patch has deadlock bug +Im going to close this issue +Better to create these kind of issues in RFPL +It is not that fails to workwith IBM Web Traffic Express Caching proxy on https rather it is IBM WebTraffic Express Caching proxy that fails to work properly +would have been sweet but it didnt even make the cut inQuite strange for a major issue +This should work when using +the problem was not updating the cache version or clearing the cache after an update to a structure stored on disk +Attaching fix +The patch adds support for reading and writing to +Ill transfer all the text from the github account shortly +I set the default to minutes +It implements the requiredFields feature in all the relational operators +related test case +Corrected the datatypemanager to return the appropriate implicit types +updated to in svn checkout next nightly build +In the previous patch I didnt handle the robust in only case +I agreed with Yang that I would not attempt to apply any nore of the files from attachment +Will adjustbq +I will work on a clean installation to the connector +Versions and appear to be addressing this issue +Its a fine idea +Grrrr +So for Derby versions before on Java the class loader falls through and returns a version of Derbys JDBC data source +HelloI can verify that Andys fix works for me tooRgdsJesper PeterThe only solution in my opinion is to patch the source code yourself and then rebuild the hibernate core +Closing all resolved tickets from or older +Lets hope it survives the CI this time +Validator is not supposed to change the data when validating it +My takeaway was that was somehow screwing everything up by being the only component pointing the way it was pointingIs it ok to leave the fix for now and pending the results of discussion on go change it back if necessary laterTimsounds good to me +mvn install D skipTests +The charset is a valid part of the mime type so Im not sure we should remove it +Agreed and yes I think deprecating makes sense +Modified package description file which reflect interface changes +Making patch available +In future please also attach the MS logs DB and ssvm logs or any series of events you followed which could be helpful especially when filing a blocker +patch applied thanks! +Rebased the change on the trunk version of th Jan. +Nifty +is a potential user of this +I can duplicate the problem with any test run +Thanks have two comments here +retrospectively setting fix version in preparation for SDO issues fixed in SDO Java M in preparation for M release +Scott thanks for noting the JSTL update issue + +Hi SteveFrom my point there are mainly two cases of The admin replace a disk and then trigger the at once or in specific time +Reuploading Jeffs patch relative to svn trunk we generally require unit tests but it doesnt look like the eclipse plugin is really testable +Patch to ensure application resource usage report is always provided when building the application report +I am now able to use xae under emacs again +Patch to check the version number of the appropriate schema type to ensure it doesnt change set for a version that does not support it +I have my mgmt breathing down my neck for this one +and see and +In the testcase produces correct is associated with a and the parse method will only deploy resources that have a or no deployer at all +I HATE the priority choices +Ok thats a good point +The VM has been running for a while and the SSL endpoint has been offloaded to +There is really no content that needs to be preserved +Thanks Sandy +A patch request has been filed to the review board for branchPath is pushed to masterRayees is on PTO Chandan and Pradeep would be able to help if there is any testing requiredthanks to branch now +It should be possible by using just one queue for all the fetchers +As far as I remember I left out some formating changes I didnt like +Here is original debugging patch + +so i declare this issue for further Please add the +But if they wish to be credited then of course we should permit this +Patch consists changes in mahout code that passed unit testsThis isnt fixed until the fix is committedCan you your patch in a standard form such as produced by svn diff +Closing RC issues +Thanks Vinod for the suggestions +There is very little variance in the cell size +I dont think this information is important enough to deserve a place in the header +Thanks Wes +closed for release +But i am sure it need to be adjust to trunk code +Sheesh youd think I could spell +Im setting to incomplete I havent seen any response to Nils last comment. +Committed revision There is a problem with the file which I would like to resolve but I still need to sort out our dist area and ensure that the mirriors are configured properly with minotaur +to trunk +I would be expecting an identical behavior in both init scripts and use DNUSERIDENTUSER for both casesBruno Debian system uses +Attached slightly adjusted Set class comments removed one of the constructors that was causing a bugAdded Test script for Set important thing I forgot to sayPlease use in stead of the other to the repository +Thanks for the input +One less item on my agenda today +Deepa Ill leave it to you to file the followup JIRA +Or something like information I have tried closing and reopening eclipse and it does not work it is really the project closing that make the refresh happen +something like remove timestamping till we do it right because meantime it only gives wrong impressionOn +Committed revision . +That diff was pretty untractable due to the lack of u or c and the reformattingWhat I really need is a testcase of the problem +It just adds a convenience class that can be included or not as needed +See previous comments about increasing the stack size +userdheerajkdatapow +This issue occurred using a snapshot of the trunk which was still in flux at the time. +I merged again today +Hesitated it might be trivial to test formatingAnyway now a Test Case has been attached +The default value should be to use the same name as the input +I have checked locally +core tests +I dont suppose theres any way around using node? cos wed still need to find the binaries for the specific platform were running on which would be a pain in the ass +patch committed +It effectively disables the troubling behaviour + attached +Mark Kevin and I talked this afternoon and I think we can swallow the xbean dependency in going forward +Will the interceptor block the call? if just be notified seems like some sorts of ? looks interesting for the dynamic subclass although potentially problematic regarding classes without default constructors if not cglib then lower using the asm library insteadIn light of the constructor limitation of cglib Ive written a patch which utilizes asm to do the subclass wrapping of bean objects +Any chance of the attached patch being integrated into +I added your stuff in revision +If Ken told you jump off the golden gate bridge +Minor +Now fixed +Sounds fine +Now assume that has eating up ALMOST all of the memory but there is some left +it will be easier to build a spec if we start with posixs and then look for problems +It works against the standard release running pure load I had the impression is a lot slower on trunk +I think we can close this as the test is excluded anyway +I saw the error on IBM test runs with IBM jdk on Windows machine on trunk revision jira was created in July +How do you plan to detect downtime? And how that will be different that regular recovery logic +The code changes look good will commit if the tests look good and they to fix sampling in subqueries +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I am not I think the current change makes things worse not better so I responded with +It seems some code in have been lost when resolving conflicts with recently committed codeOr was it intentional to remove the calls to +Actually bug is the bug I have been talking about +This is a general requirement that does not only apply to +Cleaned up github stuff right ASF license +Will this be an issue? Blocks of an awkward size say a block that happen to have a massive in them and they exceed massively the default block size would need to be treated differently +Another quick pass biggest change is sharing thread executor across requests to reduce the need to spin up new threads +The only provider where it is overridden to be false is in +Perhaps nonce recorder is actually the simplest solution to both +Let me try out the patch if i have some timeSo once we say HBCK fixAssignments then we need to see if the region assignment is done and then forcefully enabledisable the table stateAnother way could be forcefully provide an API to disableenable the table based on region states and if needed correct them too +The multiple repitions of DimageSrvbinnative is not a cut and paste error +The refactoring looks good +That means in other word no one except Apache committers could do the work and releasing it hm +Did someone already try the configuration described in my second comment +Fixed on trunk as of revision +Moving to thought I found a shortcut to moving issues to the next release by the resolved link but it didnt work as intended so im reopening this one. +Patch for TRUNK +EL validator shows only first problems by default +remove the infinity in the statement +I filed to get this issue fixed in general +Fixed +I think that the beep has always blocked the UI +Closing out this issue. +Well spottedll update it +Please revert back if you disagree +The current style is leave that to applications folks shouldnt generally modify data after passing it to a Schema constructor +Patch applied at +We might want to wait for resolution of that issue before we introduce this new API that we know wont work as intendedIt will work for setup +Closing issue as version has been released on. +thanks +apply this patch to modules MiyuruIt is better if you can add some context to the Jira +I propose to have running jobs in a running folder and move the history file for completed jobs to the parent folder +Can someone review this patch pleaseD +There have been several people who have said they had already done a basic of this so if anybody would care to post that code I can pick up where they left off +Thanks Dawid I am looking forward to that +And it is not dangerous +The main issue of recovery mode is silent data lossRecovery Mode prompts the sysadmin before doing anything +Resolved with revision of buildersriftsaw buildersriftsawresource buildersriftsawresource buildersriftsawresource buildersriftsawresource buildersriftsaw message Remove ESB files from install +I think I have found the culprit +My current svn checkout of the contains just enough differences to make the plugin work for me as I recall it +Heres the fix +applied patch with minor changes to obey javadoc htmlfixed in subversion repository as of rthanks for the patch + +I was trying a new CR candidate of Seam FP and it finally works but I had to open Seam setting dialog on my Seam web project +Sorry we can do this whenever you want +Moving to TBD EAP +DO NOT REOPEN REJECTED +java output for every build +Manually tested +The attached could be a solution to the problematic presented above +Modified code to check for word boundariesWhen searching for India code will match Politics of India but not People from Indiana + +This is different than than the pool stats which tell you how many instances are in use +However no way I can work on it until Monday and well probably be ok without it +This should go in Accounting ComponentCreate CRUD UI for in Accounting Agreement to AddEditDelete an for an agreementIdCreate another CRUD UI in to AddEditDelete an for a workCreate Lookups for Agreement and + +What about sasl lib libxml libxslt etc +Maybe a unit test case that tests this functionality would be really niceThis is a very functionality +HiI will appreciate this being included in the next release because makingCXF works with the oracle soa is a mayor compatibility stepSimon P +I didnt originally anticipate writing a design doc inline but you know fingers started typing and a few pages later it was a very long JIRA comment Ill do another rev to address your questions as well as flesh out the BN and Upgrade bits and upload it as an attachment here some time tomorrow +Please remove those line breaks +Its just that I consider the original code as quite a mess Im currently working on this stuff again and Im available on IRC +Fix in hotfix +ah +Less than megabytes +After a quick test it looks like using the new embed API will work for us so Ill take a look at converting +HDFS bytes usage is calculated from dus periodically +No there is no SQL object to back up this comment there is no closing stale resolved issues +I also upgraded to and configured to output in +would it be possible for you to provide a new patchThanks I understand why you are using the prefixThe users locale is already in the session and in the screen rendering context there is no need for locale suffixes +but the rest can be sameSolr shold startup an embedded jetty and manage the lifecycle completely +Here is the updated patch based on the feedback from Paul Linder +Verified. +The timeout manager kicked in times but after that the three threads were taken up and all blocked on META but the server that had been hosting ROOT was stuckIm thinking the best way to deal with this is to divorce the threadpool from the reassignment of regions from the dead server +So from me for branch +Wed have to make it so IF THE EDIT IS FOR THE CATALOG TABLE ignore general flush settings and write out the + +I think I prefer the first solution but I dont care to much as long as the bundle data area is used by default +Making a start +the same patch with only a change of path in patch +Even i cant understand why this bug is rejected? It is valid requirement +TedThanks for review +provision of get delivered. +This can lead to all sorts of unwanted side effects for example with the propagation of session attributes in a cluster +closed wrong issue +Fixed thanks! +fixes this problem on branch also adds some tests +I would look for similar race conditions in all Spillable implementations +The test failure on is irrelevant with this patch but due to +Also addressed in but made the fix optional based upon the vm property setting +Checked in trunkdocsmessenger rev +I see youve updated the issue +Spring Security does this for their own authentication endpoint +what is the status of this patch? +I think with screenshots and changed log level of testng is enough. +Flat structure under trunk has main advantages shared libraries between subprojects in a project can not be stored in duplicated way Subsersive can handle on flat structure under trunk to do tagging has new structure. +Attached is the code submitted by Chris Brown plus some mods +Bulk closing stale resolved issues +Seems Id accidentally based the patch against not trunk +when in fact I hadnt read it carefully enough +I just closed the pull for the contribution! +Note that this still only works if I disable the SIGAR getline native code you do not need to explicitly quit on UNIX when the file is done the shell exits cleanly on be clear heres the workaround until this bug is fixedIf you want to run the CLI shell on a Windows platform and you want to pipe commands to that shell you must do two things edit such that SYSPROPS has false SET SYSPROPSLOG false ensure that you explicitly pass the quit command as the last command piped to the to in SIGAR source still waiting for binary update of will be in the build just breaks out of the getline loop when reading from a pipe after all input has been read. +Implements proper validation in +Thanks Jitendra! +We havent seen this since this one occurrence +This is the mentioned heap walk generated with jprofiler +to commit +Closing all resolved tickets from or older +In there are mentions of only Tiles + description says that the termination handler will not be fired when the process reaches an exit activity +Then its just a matter of iterating the sstables and comparing checksums +But due to the above patchthis situation is handled in a way that the reducers dont hang infinitely +Will add to webapp component and trigger a schedule +When the splash screen needs to be shown its plugin is not loaded yet so you will have a blank screen when you are expecting a splash screen +ArunIll take a look at testCompactionJim +Ill fix the containsKey error well spotted! I think we should view the ability to override install configure functions as an advanced feature not a common mechanism for installing different flavors +I changed that because we had hanging tests +A few commenst about the patchClass The might return as a token +Decided to stop using homebrews wkhtmltopdf I cant seem to fix the problem by reverting to +If you remove that things will work as you expect +Particularly since the issue is counter intuitive the dependency plugin will report your dependencies as fine but the src plugin will fail because of a missing dependency which the dependency plugin would report as unreferenced +Yes +Jake your patch does not apply cleanly apparently because youve not diffed against but against a more recent version which was updated by a patch addressing the trfunctional issue +The new patch with the inner class changed back to static +Just saw you already changed the test to show how it is is a better way to use properties +Email conflict path in the for cachefix for setPreffix for has packages that have to be imported +Lets reject this +Cant reproduce but not sure Im doing exactly the it appears again in last trunk revision +Very much agreed +Such information would be welcome yes +Added new class added test case to validate behavior and updated to use this. +The new patch looks good +This issue is blocking for my companys project to use SONAR for Java code quality assessmentPlease keep us informed when some progress is made in fixing it +It updates the bundle when you update the bundle file in the directory and when the file is deleted it will stop and uninstall the bundle +I have changed the error message based on you review comments given patch which has attached to this ticket has appropriate error messages +We should also take the ESB doc structure for the to minimise changes and resultant errors +Tilman Hausherr I really appreciate your helpIll try the revision thanks again +Ill commit shortly +HadrianIs there some wiki documentation that should be addedupdated with this change +Thanks Tom! +Please +is it possible to add a flatten mapper inside the fileset with the refid or is that an option youd need added to the dependencies definition +The Map types in your service are trying to end up in the default type mapping instead of your type mapping +What should we show +Max suggested that the app creation wizard provides the user with a dialog to create his domain if he did not have one yet +IMO Spring EJB Web Beans Seam or Guice should handle dependency injection +There were a number of issues spawned out of this one the rest of which have been issue is concerned with the exception raised when a flow through the jBPM console code results in a call to set as described in the first comment +Adding a JAR does not add any imports automatically nor are imports handled at the package level. +Its also possible that some behavior changed during the timeline rewrite that either fixed this issue or invalidated this way of showing the issue +Hi MarkI cannot recreate this problem +will commit shortly +Its not in the spec either +the ASF acquired filesGuess we can setup two TCK profiles one that uses the files and another that uses the NDA covered filesThen in publicly discussing problems well just have to make sure were using the sourced filesSounds like a good topic to add to the wiki +release bulk close +As it is not reproducible on latest master +We need better documentation but it should be dealt with in another issue. +Ill upload a patch shortly +Am I supposed to do anything here +Yes trash is enabled in tests on server sideSo looks like its expected behavior in that case +Also this patch is independent of the upgrade patch and will not cause any failures until the code that uses these stored procedures is actually put into the clientAlso I will take your suggestion of moving the stored procedures into SYSIBM +Viktor thanks for reportingAnd Thomas for the hint for the solutionBut there was other problems on the receiver side when using transferExchangetrue and with files +I would just like to chime in and request that there be a way to configure things so that absolutely nothing is on disk including however that might work outThanks to say but if you want to spend time on some of the known weaker spots of the patch search the diff for TODOAlso Im wondering if it would be wise or not to ditch the path handling and rely on instead +fixed in revision old resolved that dont have a fix version and therefore cloud queries for M release +Thanks for looking into it +Raising separate JIRA for those +This is a old bug so please verify if it has the same behavior +Can you point me to the debug log statement that apparently catches and logs the configuration error in your case? Such exceptions should simply propagate up to the test runner not being caught anywhere +Another thing to remember is that counter mutations are not allowed in atomic batches so the two statements wont be functionally same anyway +So remove item +Iteration of the replication filter patch +Native SOAP stack is not supporting JAXWS any more + +I dont believe that the systemvm should be used from +Patch applied in branch +Thanks for closing that gap RC is scheduled for late September +bulk defer to +adding a outlook related issue +Heres a patch that updates the methods to throw exceptions to the user rather than just a boolean +Attaching the right one +The dialog problem is fixed to the degree that you now can embed a form into the dialogyou cannot embed the dialog into a form this is not possible at all due to IE bugs in its rendering engineI would say this bug is resolved now! +Google reader can only read available publicly as it polls the RSS on google servers +Marking as resolved +I am leaving the JIRA as open since I want the fix backported and have updated the Fix Version field to include backported to branch and passed git pull request +In the current patch I only removed the three unused fields from +This sounds like you are missing the m directory check to ensure you have that directory and that it contains the following files +Please comment here or on the dev list +This bug has been fixed in Xalan Java +I checked out your test case and made some mods sent some first directly to the consumer broker made the consumer use failover and restarted the consumer broker +the th attempt I did now reproduced it but the application was not rolled back its still responding +Will open a separate ticket for moving out tools into the client packageCreated +Heres a fix +I think saying +static String PROPERTYTRANSACTIONISOLATIONLEVEL but also five more specific property names +Ive made the whole package optional because it is only required for dynamic wsdl generation +The JPA example worked without a problem +Dont know how we can create a table with the right collation with unit test means that we have but maybe at least create a regression test running queries with alternative syntax without checking the results? Otherwise I think the patch can be applied +linking to provide steps to verify or close this to close this Jira +Merged thanks! +It is only for namenode and datanode metadata where formats may change more rapidly than a general object +While moving it into the context class I added more sanity check to make sure that it has locked the correct locks and only once for each +If there are multiple versions of the SLFJ jar on the classpath it looks bad the shell dumps a warning about multiple bindings after the first command +Its not actually in the test it just manifested itself for me via the test +thats better indeed + LGTM +apologize I uploaded the wrong diffs +Mahout has an extensive set of tests that work this way for both Vectors and Matrices +Patch looks good to me and the change makes perfect sense trying to apply now +Sorry about your problems sending mails to the user list +Have we got a potential fix for making this work as expected +AndreasThanks for bringing this possibility into focus +Thanks Rick! Habe a nice weekendRegardsAndreas +I attached a patch instead of committing because I would like to know if it is ok for request that can be done after support to nested definitions +Vinod I think you are very familiar with this part of the codes +Closing due to inactivity +I vote we release without tackling that issue +Thanks Gav this is done +New patch changing the parameter name and improving the documentation +theres a test failing so Im reopening this issue to investigate Desk ticket indicates that build is not? WhereWe didnt get any input from the ticket saying this is not resolvedCan you close the issue then and only reopen it when you get a clear input saying its not fixed +Wow I just intend to show progress of this job but youve already tested this +Committed to trunk. +Thanks to Gordon Sim for the fix that resolved this issue. +Why not just test on or equivalent? If the flash payer gives different results in different version making the test depend on that make the test clearer I think +update because we now require a Java SDK +So it is inconsistent and really not usable. +But the names dont quite reflect the intention +In revision JCR observation events are now being generated +Closing the issue. +In Maven it is built in and thus we have less to maintain +Verified in IR +but lets see later this +I modified the files in svn +Have you tried AS? From the log it seems an issue with the deployer in that cannot load Guava also as you mention +So the idea is to expand to the greater scope +I dont see any other possible conflict with other databases regarding reserved is a must if we call the module JPA you very re right about the portability +thnx +h +Thanks Kevin +Whats the voting system there for anyway +Ah Ive clearly misunderstood the semantics of and +Hi PhilippeHow did you finally manage this issue? M +christiani havent tried rc yet thus i cannot tell if it is something specific to your configuration +I tried with Meclipse and it picked up the right configuration +This isnt surprising since we generally expect an to fully refresh in a single thread before getting exposed to other threads +I intend to pore over the code to deduce what rules it really uses and to document them +Go for it +I think thats fineFrom the log snippet above though Jimmy it seems like we are updating the znode every second almost +In G heap fielding a withering upload could be what throws us over +Reopening because further testing reveals that the related objects are not put into the model when the form is redisplayed following a validation error. +Thanks for the patch you want to add some specific unit test demonstrating this please feel free to do so +btw this version does not have javadoc at its site +bulk defer of features from to +Hope this helps +and are IDE specific +bq +Done +Description In some network designs bandwidth is never constrained so this doesnt give much of an advantage +Hi LeonardoThank you for your code review and for your suggestions! I attach a patch with the corrections you requested +Is there really a solid use case for this? Underlying zk based lock implementation provides timeout support at a more granular level but Ive opted out in the level +Feedback welcomeOleg +This has been rolled back to the previous implementation and the incorrect unit test case has been removed +Any objections to committing this +Yes separate ticket +makes the whole CI bit less easy as well +Noticed incorrect comment throws declaration in code copied over from decided to fix that too while i was in there +These docs exist attrunkdocsforrestreleasesrcdocumentationcontentxdocsnetui +committed in +applied +Most likely a FAT jar was built which included classes from all dependent libraries and as a result classes got packaged. +close of all resolved issues. +The current behaviour was introduced with +Can you post the infoStream output leading up to the corruption +What fails? The client or the server +Resonable but my services would stop and consume too much time for repair +The patch seems goodTwo things please commit only on trunk there are too many changes in core areas for this to be fit for the branch this close to a release +Did you mean this +Not only is it the memory but also the extra initialization time of creating these empty maps that are not used +I assume this can be closed now? I think everythings already in repo +Updated patch to latest SVN +Thanks Doug for reviewing the patchIncreasing from second previously to seconds in the patch appears too excessive +Instrumented code for failure needs to be removed +versions +How do we implementation this? My preferred method is to FS implementation to do the the buffering +twas a delight +Now we have two serialization format and +This is currently done by Javas +i am looking at revision of the trunk and this bug doesnt exist so it wont show up in a future release +Needless to say we dont miss it at all +I filed for the lack of use of the scale parameter in register for both embedded and clientI added a test to which was only testing whole numbers for decimal valuesI also noticed that setObject with big decimal and specified scale has a similar issue but it is not fixed by this patch +I expect a realistic usage scenario wouldnt stop at querying root so the client would just see those S errors a little later in the sequence +Moving open issues to the M release +The only thing Ive been able to find remotely close to describing the problem is +I just committed a fix for the empty state file + +I will fix some formatting changes and commit this patch to trunk. +i dont have any strong opinions but i agree with your plan about putting some warnings in the javadocsRandom thought it might be cleanereasier in the long run to createregister a hog Handler that records timing differences and add calls to and on the various methodssomeone may have already implemented such a some average stats to +Will be available in the next drop. +WDYT +Maybe its simply the checking not properly detecting JDK on Solaris +I would guess were missing some drive letter logic when splitting up the path components or simply not recognizing the drive letter as a filesystem +I believe this commit fixes the Tab issues for. +bq +Take it one step further what does constructorDelimiter look like with space delimiter +Will fix your comments in test code and submit a patch +I would like the salt to be appended to the password hash after a so we dont need to create a new entity field for the salt +Right that sounds reasonable +Youre right Jonathan maybe the framed transports should advertise their maximum length? Its not perfect either but its a small step in the right direction +configuration error fail hard +Changing from Mx to M as nested composites and wires across composites are starting to get implemented the sample should be adjusted to use this capabilities +to trunk +bq +Thanks Jitendra +Regression due to fix +stackted any idea why its failing these tests +v is what I will commit on trunk +There are a whole lot more problems with links and disabling javascript first thing to note is that maybe you add the here but what about all the other parameters in the form? They wont get submitted and the action works on stale dataThis is something that has to be seen in a larger scope and fixed in that larger scope this patch alone wont help +this file +Thanks Surya for patch +Closed due to the release of Tiles. +Please update and see if it works for you. +It has become apparent to me that having sysinfo in the especially in combination with would improve supportability significantlyIn complex systems it has been difficult to gather the important version details because there may be many different distributions of java and derby on the problematic machine +Seems this problem is related to my database +This caused tasktrackers to go into a bad state and not service any more client requests while still sending heartbeats to the jobtracker +Does David know about thisAlso could you please give us a before you go off creating projects? hausmates typically get their projects approved of course but its still best to run it by the if youre a funded startup that should be running a or suchJust remember this isnt +But since eventing is the only impl +There is seriously some issue with High memory utilization in Solr with large multivalued field when retrieved +Im creating a new JIRA instead. +Creates a default value expression if one is not set +ok it goes stranger +Sorry for the confusionIf you want I can send you our test code. +Second I cant even get the most recent snapshot to work at all +Unassigning I wont work on this right nowIm wondering if the way log files are rotated causes issues with the way Windows caches filesystem info so that tests see stale information about the rotated files. +Personally I think the Java XML Serialization is very cool + +The basic questions for a new section in the guide were defined +Applied on the java level +Thanks +why does the tentative fix contain modifications to the inRefactoring accident +Jukka Id rather move the is method to the streams and since it is really their job to readwrite the data +Not a bad idea to improve and make testing all resolved tickets from or older +I can still reproduce this in +New parameterlambda rank of the rating matrixnumber of features of each useritem vectors number of iterationsepochs result on all evaluation uses RMSEJul PM infoINFO Recommender With time spent Jul PM infoINFO Recommender With time spent This is fast and accurate enough Im advancing to netflix prize dataset +I dont see the point of this patch as in the comments it sounds like it is ready for commitJust an FYI for the general layout stuff we are moving away from layouts +Updated the jQuery ui dnd libraries to +Nick Is the Primary storage local storage +Hi oliver Havent been able to reproduce the issue tried with the openejb project everything worked can i easily reproduce the bug +This is a docu patch file for the site directory in svn +I cced the support email to both +The patch also adds testsAll regression tests ran cleanly +Quite amazing but believe it or not I figured out about minutes ago that would solve the problem then I got your email +Patch which moves Sandesha up to the new spec level +Would love for somebody to confirm that and also of course review the patch +This looks very similar to +ALL D W I like JOTM is sending commit statements there on resume + +Fixed in r +revision in Thank you very much Nathan for the contribution +Thanks Sohan! +Bug is fixedAfter fixing the NPE i noticed a second bug did not work in last row of a table +This is ready to go now +We will modify the code to give qname in those places and let you know +Will do +is not part of the new API in Hadoop and we wont upgrade in the near future. +I have pushed a fix to master and the branch +I confirm that velocity and dots are not very good friends +I just committed this +If you did mean something else than the mere syntax issue then feel free to reopen the issue or leave an as a final comment if you really did need that syntax you could roll your own support with some tricky metaprogramming or AST manipulation though the earlier mentioned supported options will be much easier +Are you using the plugin? If so please show us your configuration file +Do you think we should add an enableStats button to stats page in? It would be nice if you could fix +Ive just committed this to the branch. +This patch implements a bit fileid per file +We have a concrete use case and the proposed solution doesnt add that much overhead and doesnt require API changes +Any update since last comment +The new version of pulls in logj avalon and logkit as transitive dependencies so I have excluded them from assembly in the uberjar +I have some questions +This example does not work in Internet Explorer +Fixed tests that were breaking because of semantic differences between how certain commands are processed using andThis patch fixes the branch + +Looking at case I can think of separate cases which could be useful since I have projects of all A project is already a Maven project but does not use Spring +The only type of POJO implementation that could be specified with a side file is statelesss with no intitialization or destruction callbacks +Sorry if this spams things however its unlikely that Ill work on these. +Tested it on master it works fine +I think this issue is related to no +Patch to turn off ssl compression still need to test with openssl +I have committed this +reopen to add fixed in trunk and branch +Project can be deployed and excecuted by coping srcorg in to srcmainjava and issuing mvn clean install androiddeploy +Will submit a patch for soon +Keeping it open until we find the true fix for it but moving it to since got a workaround AND fix for it +Nice small comments Rather than injecting the Extn class into the FS Cache could you instead set up a new scheme by setting to your test class and then calling a command on testfsfoo? Seems like a cleaner test pattern +It should be right is +Whats missing? They now pass reliably enough even if they still depend on time +I will attach the revised patch and will commit Monday unless there are more comments +shouldnt be setting this parameter any longerDone +above + push to +I dont think so +If this hasnt been fixed any leads into where to insert the detection would helpful +Did have the interface at the point Curator Ensemble was written? Is there a reason why Ensemble isnt implemented in terms of a +Hi Keith could you review this change please? no comments from me. +If you have a test case for it Id love to justadd it to or +applied patch with minor modifications and added think this is very misleadingA CSRF attack requires the server to accept POST requests with content type. +I tried testing the patch but thebug reported in seems to get in the wayYou are correct +For the feature you guys need you guys should be able to add metadata to the Sessions and have the information you need +Was my own fault was creating cachemanagers using the which does scope ports on a basis to isolate different test runs. +Simple fix looks good to me +closed. +Request to fix this at the earliestThanks +For now we should get the required jars in SVN under contriblibAny reason not to go with +This patch makes package fully signature compatible with Java This patch can be applied to core and modules +I guess its the bundle plugin update but for now thats just a integration test failure is indirectly caused by the bundle plugin update in that the JCR API dependency is not imported with a version +Thomas HelderI am unlikely to respond to further bug requestsyou may feel it is simple and easier for me to find evidence in the spec +I am not seeing how that can be done dynamically +Changing owner to Unassigned +Is anyone from working on these docs btw? +So any suggestions on how to proceed? The maven profile for now could simply prevent the tests from being run +The statemachine for need to offer a way how to lockout the instance in STOPPING state +Do we want to invest time fixing the usage if we plan to move to maven? I know it was copied from pig and was a quick way to get it working +Sorry Ive set the scope of my dependencies to runtime instead of compile +fix in modcluster repo +Howard we should get this fixed for the release +Backlock seriously +Rather no please upgrade to the latest available version few security vulnerabilities were solved in mean time +new version of doc with added OPT text uploaded to website. +Added +Updated patch that does not requires UGI initialization in advanceThe UGI initialization will set the name rules only if they have not been already set +This is a partial fix that corrects the names of the new globals for caja the missing setup of valija and uses a more sensible test for whether caja is enabled +With the factory we can load more serializers and finally loading more artifactsTo do that I had to change the constructor moving some of its code to methods that can be called by the at the right timeAll Model constructors had to be modified to call the post constructor methods +The accept thread run loop now catches any and ignores this with a warning message +Attaching patch which implements the sameFollowing is the way the Feature is implemented +I committed this to the branch +Note in the original forum thread I was requesting also to allow Parser and Printers to handle null values +I would like for that +Sending srcmainjavaorgcodehausmojojaxb file data + for the patch +Seems my previous comments was a bit off. +Can be reproduced with Tweetstream sample app +sorry +Hi AndrewThanks for looking at this issue +Something for the +And I cant reproduce it in Tools +Any success in reproducing this issue in a test case? for the time being as we cannot reproduce it. +I wouldnt mind adding an insertLocal path there but optimizing atomic batches for single node clusters isnt a super high priority +My fault Ive used the wrong classloader +However I dont understand how to fix +pulled the query beans into their own file and include via import so that the duplication is removed. +I did see some similar comments on the Findbugs forum +trunk +For me this test failed with message mkdir failed for IBM VMIf I try to create this directory by mkdir utility it reports Invalid argument while directory with symbols create OK +But at this time I think it does not sound right to tweak for some functionalities we dont have a clear consensus +Ill try to create a test case as soon as possible +Attachment has been added with description Test file for right alignment makes perfect sense about not knowing the font metrics for each device and explains why we are seeing different results +maven first looks in your local repository and if it doesnt find the archetype it looks in the official repositoriesi think you can put the wicketstuff repository in + +my tests showed that this is not related to getName in any way +This is just the deletion of a redundant example descriptor file +If I find a way Ill definitely try to distill it +Hi Simon Looked at the latest code on trunk it seems that I can just add my own on line in DWPT +I am inclined to say we document carefully and resolve this issue +will be always compiled with Spring +No we didnt test on the release of it was pretty close to so there was no time to test specifically for jboss +Thanks +Thank you for reporting the problem +As such you should be able to change the constructor to not take in the class and change the fromHash method accordingly +For arrays of primitives this wasnt a problem since the space for the array element would be overwritten by the new value once you called the appropriate set method +The merge had conflicts so it lost its committer log sorry. +Thanks for providing the patch +If you want to test before the snapshot is available +and both recognize the for the simpleBeans property but fails to find the simple readwrite methods +The examples in our code actually all subprojects do depend on the same snapshot version +Since this is the first time I do this I would appreciate some feedback or help from other with more experience here +Im good with that +First round of patches applied in r. +This issue was resolved lets open a new one if we want to do more. +gives more information on a workaroundalso please do Not ask on other tickets to vote for a ticket that you open in exchange for your for seeing progress andor work done unless you managed to hack into my local machine you would not see any as of right now so thats a ticket is really importanturgent for you then you can always sponsor it and we will give it a higher priority +This makes a temp change to the max that can break dispatch if it happens concurrently with there be some admin task that uses the console or jmx to copy some messages by any chance +There are a bunch of uses of log that do need to check if audit logging is enabled before constructing log messages +Replacing USER makes sense also the hint is a clearer message than what I had +By the way Ive run unit tests locally for both and patches and all the tests passed except which should be unrelated and has been reported in +gonna take a spin with ibm j +I ask because it doesnt seem to work for Alpha either +It isnt recommended because username can be ommitted but some users need it +Regarding the Master more work needs to be done to disallow a user from taking administrative actions on other users tables such as enabledisabledrop +The generated file doesnt contain all startup bundles +This is going to be closed as soon as the subtasks are fixed +Not at all as I intend to through out these to methods +Did a little debugging and I think the problem is in theIts creating a new param to handle the not checked case this isnt correct if we have an array of checkboxesI think it needs to check if the parameter checkboxxxx value is an array with size it should not create the new false paramIf I can make this work I will submit the patch +Verified in JBDS B installer + +Assigning to Hardy to tweak as necessary. +Jonathan will remove it no patch necessary +Reopening to clean up the API +Should the same be thought about for dataflow operators +The IT for should be extendable to verify the proper orderAnyone wishing to experiment with this can do the followingIn directorysurefiresrctestresourcesRun commands tests like thismvn methods clean testmvn classes clean testYou should be able to test observe the effects of any patch Deployed updated are unable to do anything about output that is not associated with a test because making assumptions about this would have severe implications for embedded use +I am raising the priority to blocker as IMO we cannot release with such an outstanding bug +Moving to for implementation +IMO its always required let user get version info quickly and conveniently +This issue was introduced in +I agree going with Commons Compress would be the best approach +This issue prevents Hudson runs from automatically verifying the projects install d also say that this is critical +I have a watch on this JIRA and think it is very useful +JIT compilation will inline functions that it sees fit and unless you have a performance test that shows that this implementation is affecting performance that adversely I would advise that you keep it in a method until it is determined that it is a performance issue + +also test the task comment replies +There are no multithreading calls in our Java code +When are you guys releasing? Id like to get it in that release if possible +This patch keeps reload untouched +Unfortunately in order to make all socket operations properly interruptible they all need to be using nonblocking logic plus selectors +Presumably the DN must include in block reports the disk location including the media type so the NN has the necessary information to accomplish that +This works for you Benjamin? LGTM +Anthony whats the status of this issueplease close it if it is not a bug am having the same issue +I also thought I added checks in the code to disable revoking owners own permissions but I will check where I missed this + +Andrew can we move this out of? Do it for +At the very least this gave developers a way to add data source entries in the short term for persistence testing +We just have to coordinate with Jeff to get the branding done +Ill commit the patch for now +JON for SOA +I moved it to positive tests +The system Vms are going into a continuous reboot because of some issues with etcudevThis is expected as a file but in the template this appears as a directory +the patch is a little big main of part is test has been applied at r thanks for your contribution +Fixed by on trunk and r on the branch. +Updated patch file +It didnt actually delete what I typed but rather changed the highlighted area +Fixed with commit dcbdaeaabcafe +Changes will be in the next Brew +Do we have any tests for this FS yet? +This is the the same issue as +Why we need to track cts failures in both JBAS and JBCTS +Patch application failures are normal for this type of change +Thanks for the reminder and sorry for the long delay +As you say the point is toobserve that the Derby bundle can be activated and your test does thatSo whats the next step? What sort of feedback do you need in orderto make progress with this proposal +A mistyped absolute path will be treated as a local better workaround would me imho to have a different url type for local references such as datashapefilesso that absolute paths can be spotted easily from relative ones +Closing! +No objections so patch as fixed in. +Do you like the tone of that better? +Also in our case if we would just silently ignore the empty preferences file that would basically mean a reset +add was avoiding handling over blocks during safemode +Given no response and the exception being identical to Ill go out on a limb and say its a duplicate of as Kathey points out. +Would that help to address your concern +hadoopqa is right +If its a public API then we can only deprecate it +Hm so what exactly is the issue now? Youre now getting the correct fault right +If they wont backport the fix for should we wait for their release and base our new hq on +I like all the removes and the emergence of a simple pattern all follow when they want to do region moving +The attached patches strip view name parsing and the associated testingThis makes the parser output more stable. +Thnx +The patch for this was committed a while ago +I still cant find anything that locale namesIve changed my app to use support a locale of enUS and it seems to work fine +unfortunately Im not so skilled in thanks for your interest +We need to keep custom fields searchable for SBS and JIRA integration +Its a little rough as in the prose isnt fully fleshed out but the major ideas are all presentWe also decided to remove the secondary write path and write flag logic +Wow +this one moves some deps to java If people want to run the social code and gadgets code together it makes absolute sense to keep everything on the same version +Pleasepick up the latest CVS and tell us if it works for you too +Hi IsharaThe xmlid specification says The xmlid processor performs ID type assignment on all xmlid attributes even those that do not satisfy the constraints so even when the declared type is incorrect you should be setting the type to be ID +No errorsYou can leave the issue open until you figure out what is going on +Patch contains the that supports usage of this time is the factory capable to create for Entity classes +Please provide quickstart application with steps to reproduce +probably the incubation can be as simple as creating a techpreview or incubation directory in the download +This change causes that is recreated in and Vehicle instance refreshes from DB so reference to Person instance is lost even it is still available in the conversation in but that is not in SMPC expected due MANUAL FLUSH mode for affects Seam framework special classes and which needs to be updated to can react on that +Looks good +Im wondering why is an xml file? From what I see why cant we just make it a properties file +The current specification does not require the container to support interception of other beans than any managed bean that is not itself an interceptor or decorator or with any EJB session bean +This attached file shows a screen shot of the protoyped package structure under eclipse +Moving to. +A silent attempt in the background probably wont do much harm and is the safest approach with attempting but not annoying the user with the output +Removed from since its just scripts for bugs and jira searchcommit id efdcadaeacfaaeedcfa +Can you try this +Confirmed fixed in not able to reproduce and artifact loads ok now +We should start there and create connection resource managers for the individual containers +This sholud be resolved by implementing the missing methods in the classes and just throwing +In Camel we have the module which can only be compile with spring +No one seems to know anything about this issue so Im closing it. +Fixed the tests I think +For those interested parties I have almost completed this work +Would be great to see this problem fixed in but would hate to see a Jetty release holding up the releaseTrygve many thanks for tracking this issue down +Anywa i sent you all needed files this morning and just did again just in case +This patch looks great +This information has been added to the running the installation section of the release in the release notes. +Ill commit to trunk and x later todayIs this fair to commit to branch as well? Itd be good to get this in there if possible +Thanks for the quick response and for a great framework +Integration test updated +Looks great +Vivek can you give a try +Well take a look and see if we can get this in for +Fixed a couple of minor bugs in and +This means that the while loop in writeFDODTA usually runs exactly one time and it is the responsibility of the while loop in writeQRYDTA to compose the implicit row set for return +We will definitely address this issue early next week at the latest +was already merged by Dan via gitAndrei +Ummm did you forget to annex the patch TobyIf you still have it just go ahead and apply it +That way we have fewer code paths to debug +. +Here is how it will look like +Hi Denis so what should we do instead? AFAICS theres no possibility to pass the classloader to use into the +Hi Tristanwe have tried out the new query sent by you but it gives the same exception described in the jobs logs +Hi Would you please try this patch? Sorry for inconvenience +And another thing if you choose not to install these docs the doc directory is created anyway and remains empty that seems strange +You are right my ant classpath was wrong +Bulk change issues closed by JEXL. +per JIRA meeting June +Thanks TimYou may also close as incorporated. +deploy into Tomcat +Ive attach the scrinshots +Going to create for things found with the new code drop +The alert will also recommend refreshing the page but the page is not automatically refreshed + +Heres a sample project that works +And committed +It would involve including an additional test in the fast path which may or may not impact performanceI hope you will understand we are governed not just by compatibility but also by trying to make a better and better implementation +Heres a patch which addresses the issue +And the way it runs now it DOES NOT work +Patch also updates Hadoop and Java versions +If we have a then I think this should make that release or if we find other more serious problems with then this could be included in the final version of but I dont see it alone as sufficient for holding up +Attaching another puzzling case +I will take a look +done in have a third option submit a patch to Jackrabbit +I do not think a script for one documentcan redefine functions for another document because the Contexts are seperate +Heres a patch for this +On first item yeah could be clearerOK on the second item +These jars were never used by our current integration and have since been had been retrieving these files as part of the deployment process for the EJBJPA deployments resulting in the error messages previously mentioned +I think when such needs arise is your friend +Moreover they rely on annotations which imply a declarative contract the presence of the annotation implies the contract needs to be you want to override that then you can customizeextender the processors so that various types are ignored +Is this what the parameter could be used for in the goal? There is no documentation about it at all which is why I find it difficult to work out what its for and what values might be acceptable +you please let me know the steps to upgrade? Is it just replacing the new jar with the old jar? thanks is a question for the user listJIRA is not a support forum. +Formatted with DamienI cant apply your change into the trunk since we still need to support Spring context inBTW you can try out the for using CXF in Web container without SpringWillem. +Thanks Knut for all your help +The sonar server has all the permissions +Its been deprecated in trunk in favour of but this has been deprecated too in favour of methods on Job +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Im trying to get access to one of the dev servers so I can try this out on linux +Verified on. +His perspective is that it makes a nice convenience mechanism +in you can comment out wsdl added to wsdljava test case +Here it is enough to catch additionally that covers all the problems that arise from dynamical class loading including the ones mentioned explicitly by Michael +bq +I am going to close this as its pretty old +In particular I would be happy to help out with integration testing of the new release candidate against Hadoop via Apache Bigtop if it does get released + +Dprojectcamelcomponentssvn commit message Introduced for file filtering using ant paths +Ive documented the onlyonce property so I can mark this as invalid +We still need scripts to manage daemons but client operation would be simplifiedI dont see your start and stop scripts hereSpecifying a conf directory could be done by manipulating the class loader or classpath rather than adding a lot of special case codeShouldnt we standardize on for processing? Shouldnt this somehow integrate with the Tool and stuff +Thread notifies ping thread which sends ping but when it is to get the receive it has already been acquired by readMessage which reads the pong and calls notify before the is has started to wait +If there is no id it will throw a deployment exception +I successfully applied the patch on the trunk +Logged In YES useridI did a global change from to in all xml files in the serverall directory and still got the same error +Made this a blocker issue as it prevents us from running +Write the wiki cannot find any test for the annotation on a webservice is this already supported +My mistake I think +The tomcat server runs in the Debug modeAfter the assign activity excute completely shut down the tomcatAgain start the tomcat server the process instance can not receive message +What is the impact of having this done? AFAIK DRLVM has an implementation of VMI and there is no any significant limitations showed by this implementationIf there is no something specifal here then lets get this JIRA resolved as WONT FIX +Please modify it to show the problem and reopen the ticket. +If you need this fix for use the attached patchThanks for the bug report it works! Thanks for your fast fix Jeff. +In other words it looks to me like I could create my own PAX implementation but if the client has coded to logjs api then my implementation wouldnt work +Just to make it clear please dont change the existing list method +Hey BrianCan you validate that this works when installed in JBT or JBDS? I installed the latest SY into JBDS and was getting XML errors in the file +Failure analysis in progress reopening +Thanks Steve +Attachment has been added with description SVG to is a littlwe app I tried to see if it was a squiggle specificproblem +After looking into it some more I agree this is a minor bug as it should keep the feedlist in memory to avoid the feed when it wasnt necessarytrunk +One of the reasons this is not there is that it is very expensive +I would vote for it as well as the additions to the php and erlang libraries I attached in other tickets +But to some extend abusing this parameter variable is just as wrong +Thats true +I changed my lookup call to see if the invokation stack has bypass +Please never do that againPlease try to set up a straight cronjob doing anrsync once a day and see if that works wellenoughThanksHenk HenkWe are trying to revive our Apache mirror +its possibly very specific to our setup but the patch should cause no problems for any other setup as well +One issue this highlight is that we had an issue open and patch up for months and nothing happened with it when evidently it was valuableMay I suggest this is for if nobody has otherwise noticed or taken this up in the intervening months +Patch with test +Patch committed to trunk and branch. +It also extends jdbcapi with two testcases which verfies that the correct exceptions is given +Checked on EAP ER +Integrated to snapshots branchThanks for the review Jon +ah it is clear now +Based on those situations it didnt appear that it would impact the intended functionality so I made the change and then tested the referenced code and didnt immediately uncover any issues +Please commit the big comment at the end but please retain the existing comments about how faceting does not yet work +shouldnt be the case anymore since introduced a new handling for node creation +I think how to add a new request is a separate issue +Now I think the patch is working +Deferred to bugs older than a year as Out of Date +Your patch is in svn rev . +if fixed in trunk then its. +If you have code depending on the accessing the some people might run into problems +Can you add more comments for the use of the newAlso since it is specific to Tasks should it be in the lib directory or shouldit be moved to exec +Compiled with assemblyassembly to produce jar with dependencies +How do I change what branch its applied to +numTerms seconds numTermsflprotocol seconds numTermsfltext seconds numTermsfltext seconds showschema seconds after above warmup queries These numbers are consistent run multiple times against the same index load in various ordersGiven the above numbers the commands should be to get a list of fixed fields numTerms to find dynamic fields showschema to find unique terms for a field allow user to choose between numTermsXflfield facet callIt needs a new showschema option that does not scan for dynamic fields +Logging out and in again does get around this issue +Yes this may be related since the component is probably not properly stopped due to the uncaughtCould you try with the patch I posted to whether it works? Thanks +This issue has been fixed by the mysterious Grunwald +Yes it works for me now +It turns into a consistency nightmareI have some code that does persistent storage in a local branch but Im going to punt it until we decide to make this into a Factory method or figure out in more detail the context in which a is going to run +Typo in desc +Patch added +Committed patch as svn resolving +bq +I have added support for setting system properties named disable disable and disable +added port number to the hashed url +If you have a client sending a message to a server and the default value for some parameter is different across the two instances whose default value will you useThe servers default value is only used when the clients schema does not include that field +Following are the reasons for reopening the bug were tracking the snapshot on primary storage using path attribute in snapshots table + trunk. +Added ASF license to the ve committed this +unless you do some tricksWhich tricks? This could be part of it by the sound of things +Ive updated the pull request with a commit that fixes the above problem +I think we should revert this change +Thanks for asking I had already created a patch and can upload very soon you can review patch thanks +Thanks Grant +some reason it appears to have missed the merge +The main idea was to disconnect the synchronization between agent collector and data node while improving data reliability +This should fix it +I still think it might be a little early +hi i believe i have the same issue at least the stack trace is exactly the sameIve tried deleting the project through the UI but it doesnt seem to remove the entries from the database and so doesnt fix the issue +I really considered making it a rather than a simple error log but decided in the end to do the log in the off chance that someones code was working with overloaded setters +One zone for all Lucene would be fine +The performance improvements for prototype bean creation are significant +So it would be exceptionally fastDo you wanna cons up a patch? It should be a refactoring of whats now done inside s init method +But it didnt work fo mme +After submitting this JIRA I continued playing with the testcase and realized that if I comment out all of the Fetch annotations the problem goes away + +resolved issues for already released versions +Looks like a useful library but it seems that its main benefit is being able to create and Word documents in addition to Excel XML docs +The default values have been set by ECF +Patch upload +That should alleviate most of the danger in making this change unless theres something Im missing +Thanks for thing we do not use old test folder anymore +Here is a patch for fix this issueThanks RegardsArun for the contribution Arun I reviewed the patch and looks good to me +Let me think on it over the weeekend +Ours is a range based row deletes +behaviour with current trunk +This is similar to but applies to all readers not just and proposes a different less complex solution +Thanks Sravya +I will investigate a bit further but concur that it is resolved for the time being. +Is there anything else from the logs that will help? Let me know and Ill get them ASAP +Release plugin should be flexible not it isnt +A new patch that reflects the discussion with Doug has been uploadedThanks +I want to be able to read and write files with the same format +See the PDF guid for more +fixed for +Id argue against adding an INITED before service because the child wouldnt share a Configuration instance with all the we have places in the code that could benefit from this +I first it looked like this was a problem in both trunk and. +and now Ive done the port to Webrick +Its supposed to be a dynamic client so it seems like sorting out the shrapnel in the build might make sense +Thanks Magesh +I searched all calls to purge and this other code path seems like the only other problem area +now in beta m closing this JIRA as it is now possible to disconnect and connect again +In clearing of the caches is done with an event listener +Its deprecated but it will still be picked up and used if its there. +So once we move to that we will need to update that +The very first Section in the Switchyard Documentation is Service Implementations followed by Bean Services which begins with The Bean Component +Dale what would it take to have a make target that we could integrate to build docs on +So I would opt to do a simple transitive ivycachepath to build pits classpath +Hi Vadimwould you mind trying the latest code? Ive done some bug fixes in the kaha area today unfortunately one of the bug fixes means its no longer backwards compatiblecheers RobYes looks like you got it +This is an AMQ issue. +savanaTheres a benefit to being able to search your working copy to find out where the SVNA project roots are located +I cannot reproduce the issue any moreFriday th +Its missing test cases so I need to write them first +please with rev +The old version VR wouldnt guarantee compatibility with new version mgmt server +Yeah lists are not even supposed to work inside tables +how they appear in the attributes map and how the values are changed incorrectly in the pojo? An example of the object structure and types would also help +I do not know if it supports or not because documentation suggests to use Handler providerIf you are using on instead of version of it then can you provide more details such as steps that you followed to deploy axis on your test code etcThanksVenkat +Here is a simple example that has two projects to highlight the problem +I have made the changes in v patch init is reverted to its original signature topic is passed into from propsJohn +thx for the hint then please mention that in the section referenced by theabove URL! +Thanks very much Andrew! Yes everything looks fine now +A dashboard has been lacking from Solr +Joe can this defect be resolved? I see the code is checked in +Is there an update to this issue? We also need that bugfix +For now just checking your basic message should be fine but Id suggest we also try to get an email with plain text html images and similar in to check the more complex bitsIn terms of the nested parser Im tempted to say we do something so that plain text comes out without any extra work needed + push to +I went ahead and moved this to since theres no way we could get this into +Because of test scenario which is quite common +Thank you Lewis I looked if I can quickly compile a test case but I didnt understand your test infrastructure quickly enough +Closing. +That still needs to be done but I can do that later if you are happy with this patch +This one needs to be against but it should be not reproducable +Patch for both LUCENE and since they are both related to term vectors and both are relatively straightforward +Duplicates as dup +Ill take a look at this soon +The way Ive documented this in the release notes makes it sound less secure to me +Thanks! It now works in IE wo compatibility mode +Closing +nvm +Does it work on as well as +Is it possible to exclude long lines from files in the following directories +If a happens then it will report the error immediately and will not wait +Werner The files contain path information from my workstation +splitting uses more IO than not splittingSorry +Good catch fixed and deployed to +is not registered +This fixed the issue on my dev machinePlease tell me that this is not necessary on the production machine because i am not allowed to install anything in the GAC there +hmm +routeId has been added to the class i think you also need to add the new field to a class in have added the field to as well +This is a very important fix +Same as +Tests did not fail with the after the change +It is all correct except one thingIt looks as if and bootstrap were inside of instead ofconf contains required port bindings bootstrap +And for the record i dont particularly care anymore whether the output is buffered by default or not +and for this release we are focusing on the postgres and mysql +Seems to fix the problem when using possible fix for this problem +I just upgraded to hive and I had to but these in the auxlib +Dont know where I inverted that +Karthik +A quick inspection of the code finds no obvious issues so Ive checked in some diagnostics +Committed to trunk +Thanks for your contribution i begun to work with itYour contribution finally convinced me to implement a thing likeI do not want to change the filename pattern maybe sometimes comes another oneand would like to have another styleSo i will make this configureable through the newI see some other configureable things and so this is a good chance to bring thisto live +Only very rarely would they want to go back to the first coordinator action which is more than actions away +Please apply this patch only into DAS M BranchLets keep the JIRA open in case I need any more updates + +Since the vhost is depending on the Host header +Thanks Rakesh for your suggestion +Done removed maven build files. +Likely fix for this committed branch rev trunk rev In addition the server branch was updated to use the snapshot tm to pick up this fix + +If you configure your application to use the legacy RequestProcessor instead the Struts behavior is observed +Regardless of the approach we take we could add a scrap of upgrade logic which looks for conflicting function names +META +The code is fairly and avoids creating an until the encoding that will actually be used is determined +closing since is now released +You also may be able to change the layout right before export to something that doesnt have a side bar do the export and then change it back for the online version +Colin apologies that this has taken so longI installed lzo on my Mac via macports and I needed to add I optlocalinclude to to get autoconf to find my lzo headers +Note to committerPlease commit my changes and let me know of any issuesRegardsRavi + tested! +getPosition does the right thing here it checks the index file only on bloom filter positives so a high bloom filter setting will benefit automaticallyThe only improvement I think makes sense would be adding support for compaction strategy tombstone threshold +Again this is only accessable by default via the local loopback address connection attempts from any other machine will be SNMP v will likely just cause admin headaches both the apache snmp module will need to be configured for it as well as each Apache servervhost resource in EM +Unfortunately Im stuck with WLS so if you have any thoughts on why it might not be working in there Id love to hear them. +Textures on the background really gives a bit of work to find patterns for in but its still possiblePlease let us know about the progress! +So in your previous comment you have br as name for the bridge interface +Added a fix to the class to store attribute sets in the orderin which they appear in the stylesheet +but i think you underestimate the QA effort and documentation update effort that would go into this +Thanks for looking into it + +yet to complete include finishing the showcaseAnd finally removing the template +Would it be a good idea to rename the new test in from to something likeOr maybe would be a more general but still appropriate nameDyre if youre still looking to commit the changes for this issue I have no further commentsto make on the c nor the patch +If an exception happens during installation then the state of the features remains unchanged +apply patch +Thanks for the response Patrick +The same exeption appears in all pages in this case not only in in + +But the semantics you made in seems fine to me too +Currently is left inThe wiki needs to be updated with instructions on how to use the war to run jetty with an alternate binding such as logj +Abhinav can you present the steps to reproduce? Also DB dump of your current DB would be helpful please attach it to the bug +Need to test a little more +This problem applies to any case where the scope of the datamodel is longer than that of the action bean outjecting it +creates a small index and then Expand it thatd be greatIve looked at the different implementations in benchmark and Im thinking if we perhaps should do the following Create an which receives one or more Directories as input and calls If one wants he can add an call afterwards +Ill take care of it +Put water in kettle turn on switch for outlet on wall turn on switch on plug turn on switch on kettleit takes a while +Verified by Richard +Ah did not catch it was just a timeout issue +Migrating bugs marked as resolved before and tagged as version as fixed in RC + I think initially we should just add a note to not use when the users will be logging strings +I tried to work around this issue by adding the THHmm to the POTENTIALFORMATS but this did not result in the correct parsing because in that case the happens at a very late stage in parsing and the original value of the date parameter is no longer valid +I think its better to do finishInitialization step fully +To add these methods the class needs to introduce a reference to conf object which Owen did and it is just fineI agree we should have a separate patch for Issue +todd I am using it for the RPC server address change detection test +Log from Vans testing +Tested +There is just one pair of settings for all sources. +Ning I removed the tag +Marking as a regression since these operations worked without error on +patch implements the suggested changes +Shortly the pao folder will be archived and made like the others +moving back in line for sorry to move it out glad to see you are working on it +i see thrift does not support optional function arguments +That bug was fixed a while ago but the problem still exists inmobile browsers as far as i can tell +It contains just the package header imports and globals. +In the previous install I said no +Reopening since was I cannot reopen this one since I did not create it +It uses which should take care of this not send me a sample of your code that is causing the issue should be +If it is true then we throw an exception if the proposed change would exceed the cache pool limitsOne detail we should update the bytesNeeded and filesAffected after addDirectivemodifyDirective so that multiple calls in a row have the expected semantics +Does anybody have any ideas on how to convert our user guide into pdf with little effort +This patch fixes the problem by checking the message of the to see if it is No space left on device +A more elegant solution may be possible but the patch improves the code and resolves the reported problem so it should be applied +That will mostly be some rehashes of your ideas. +The perspectives have been reviewed and backing classes are now in appropriate packages +Attached patch with test case that shows the issue +I sent an email to builds +I am unable to reproduce this +Ill mark the bug as resolved for now +Relationships to self seem to work just fine +Ill change it slightly and add it to the HQ release notes. +Added cleanup channel group +Why do we need these for client +I also noted that you are running your tests without specifying server +Updated warning message to be more useful in a world +I have found the issue in which the same DOM document for multiple requests +IE on Windows m able to reproduce the problem now +Ill post results as they finish +The previous version had some extra REGISTER +Yes you are right +Thanks for test. +Merged to the branch in revision . +I agree it is not an highest priority thing right now but I would recommend keeping the jira open so that it stays in the radar +Will do and committhanksnandana +Docs added +Ready for issue out as part of JIRA cleanup +If you could specify what CA cert to use to verify then the problem would be alleviated +If we use an LRU implementation extends and wrap this in then we will lose the features of we will lose scalability since only one thread can access the Map at once with synchronizedMap +Thanks all for the reviews! I guess we need to sink the RC otherwise Hive cant use it. +Note that the current version of ajaxPortlet doesnt build in trunk its referencing improper versions of portletbridge in its pom dependencies +What do you mean by a debug jsp +To the question does an appdef event get triggered if the agent is down and you add a resource to inventory +Updated PR. +years and no fixWhats the right thing to do hereShould unpack perhaps do a copy when it encounters a directory +file in vc directory does not have win configuration +I think there are two different issues here with jcr api slings jcr api itself is not tied to any JCR spec api if we keep the dependency to JCR it is usable with JCR and JCR the real problem is that there is no bundled version of the JCR apiI think we should keep the dependency in slings jcr api to JCR +As an update ATM KF has some issues which have been reported and on which I wait updates +Ill review it and check it inBy the way does anybody know how to specify the target branch for the buildThanks +The patch is against openjpa version version idMApache svn revision submitted in revision . +Right now all you need to do is to add the correct prj file like the one I attached +We faced the same issue and fixed that by patching Axis code +Patch submitted seq +Apache Karaf is now available for download +I wasnt checking up on you honest +As per request by Damjan file attached with license granted +I think this is largely complete +Can you explain what does a mapper spans a file mean +Fix committed to trunk x. +Pagination has been overwritten by the layout of the x product listing +Simply drop it into your source note that by providing this workaround I am not saying that you should avoid the use of manual flushing +Im not sure how to fix it because I cannot change the test code since Collector is not thereSo the only reasonable solution I see here is to Change to extend document that in it will change to extend Collector and that in the meantime use if you want +Here is the patchNow the field binding can be used +I know the queryNorm was based on some long ago research on it but I dont know that it has proven to be all that useful +I just pushed this +Command line maven doesnt suffer from this issue +That is the boostrap loader +reflectively checks for the presence of and uses it if available +bulk close of all resolved issues. +Eventually this can cause the UNDO table to become full disrupting all locks in the DB instance +Ill generate a wire log +Ill try it today but does this mean that this wont be fixed or changed in the next release of Struts or are you just helping us get a workaround in the meantimeThanks for looking at it +Committed the patch with an unit test into the trunk +I agree not a very common usecase +Happened with +Andrewthanks rev +Ill try to investigate further +when you have a url likeresources +Todd I just commented about reverting the change +I went through the attached diff and html files and have couple of minor comments For the new message for the SQL State ZD please add a new row instead of appending the new message to the existing message text +Initial versionthe current patch is more an example to show that it can be done +Actually it should have a codepoint at that point if its a unicode character +Absolutely agree that we should do the right fix +I cant reproduce it on linux to later by following reasons we cant reproduce this issue eclipse for me very unstable on jreand we work now on jre we dont support now x platforms +I will make the change +I have granter your JIRA account karma to assign tickets as you seem to contribute a fair number of patches to Camel +Until the libraries are bumped from to inside this commit breaks +I have added some initial comments on phabricator +This issue appears to be resolved with recent patch web contexts go green +antoine AT AT djspiewak AT Thanks +Ill start working on a fixMy apologies +Not sure if this matters to any of the tests though I just wanted to raise the issueYes I did consider this but felt it was an very weird way of creating the database and the connection returned would be with another current user than the one requested which seems plain wrong to me +I ran few those tests failed on jenkins on my local machines and all of them passed +need to be updating testcase +Hey Stack any reason this never got committed? Just needs followup update to trunk? Patch looks good and relatively small. +This is essentially the same as the return types of the operators are incorrect +I checked some more documentation on format and this is a valid formCase one and three already throw an exception +pLease include the version of JVM and the version of Xerses as well as where can we get it to try to reproduce +Are we still supporting pre Hadoop versions with +Looks good +You can achieve this by creating two separate endpoint mappings one with the which require and one without it +Since the key is peer id and peer should be registered first I dont see a problem hereIt should be changed too originally it wasnt used by multiple threads because there was a maximum of one peer +Fingers crossed +Fixed threadsafety issue in r +properties and BBB +In the current implementation we do not provide an error window from when the user provides a incorrect credentials +Perhaps the problem is the new build setup +Each log context that gets created for whatever reason could just add the same handler which would output all log messages to the sounds like the first step then might be to write a log parser which I know Jesper has asked for separately anyway +Hello Luk FryThank you very much Ive also voted for the feature request as we use skinning As we use Servlet Ive applied the recommanded instructions +This code works perfectly fine with CordovaYou could use the sample from and use Cordova to reproduce this problemI am not using the define methods +Depends on what you understand with bloating the api +Actually I think I found where the property conglomerate writing happens +In the more general case its just hard to know whether it would always be possible to call pnlinkoffered in timeAnother option that might be worth considering is adding the notion of interesting conditions at the link level along with a query API at the connection level to allow an application efficiently navigate to the set of interesting links +Hmm it also looks like many of your changes are basically changes that are already in trunk code +Apart from fixing the tests I installed the Log Parser Plugin into HudsonJenkins and configured it to promote this log entry to an error and make the build fail so now this is and I hopefully cant miss this again +The idea was to keep Java compatible +After changing access on the control to public the code worked as advertised +This approach would allow the admin portlets to have access to the portlets +asking that he be added to jira and this be assigned to him. +Whats the likelihood of porting this back to the branch +Might it make sense to this issue and change the title to documentation +triage for releasesetting urgency to low as only seen once in a one off run +I am following the same in this patchThanks to More work for you to add something new to best practices +The classes directory is put in front of the directory in created another issue under to fix this +pete +I see no problem matching the last description in these builds +See attached for a proposed simple patch for this issue +Thanks again Rick +The main problem that persist here is that we can only guess what requires what +I could also directly hack all the as properties in the toplevel POM without any profilesThen I would need to be resolvedThe problem is that I do not want to release all parent if a child shall be releasedAll I want is that I can update the version of a child in one single place instead of also spreading this all over my files +Ive traced the issue in my case to the file Im using +Dont you think that a good error handling is not enoughXavier +I have successfully tested the Digest Authentication Scheme againstRegards committed at revision Thanks Senaka +Ive just committed fix today that also allow to map Maven goals to Eclipse build cycles +Corresponding use case is when a property is defined in a parent profile +Will update shortly +Read the reference documentation or check the forum +I am aware that my test currently fails +Might it be better to pass the value of the flag as a parameter +The issue seems to have been resolved +However this patch seems to be missing three classes that has the missing files +Please also do not forget about +See attached snapshot for referenceAnswers for your questions +Looks fine +makes it hard for new Solr contributors like me to know whether their patchs are committable or not +this looks like opentaps distribution +As the is included in would it not be easier to have a well structured easily readable file and simply include reference to it in? Maybe Im not getting it but would that not work out easier +it seems like jackson is split in two in +changed unit tests +Thanks MarkPatch applied to make at repo revision How easy was thatPlease check that the patch was applied as you expected +Proxies will be proactively notified of view changes specifically crashed members allowing the proxy to gracefully reconfigure itself potentially avoiding failover processing +Closing. +I merged the trunk patch back to the branch without issues and my build and test runs were clean so I committed the patch to the branch as revision . +Juergen +is prefered over creating new connections in this case rightDo we perhaps want to solve this on the level since thats the part which doesnt allow the multi threaded calls or perhaps more on the since thats the part which has the requirement of multi thread safetyAny ideas are welcome on this +I also changed the name of the property from to I can not commit trunk can you do once I have the bdb in? I committed the +You would have to search for the text files in the dependencies of the Ruta project +The users of my LDAP are sucessfully authenticated but i cant do nothing in the wiki i cant edit pages create pages create users nothingHow can I solve thisThanks in advance. +Forcing everyone to get rid of their existing taglist config is not a good option as it might be used for the plain HTML suggestion is to emit a warning when obviously duplicate issues are found and skip them + +The patch can go into too +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Fixed in ESB will merge next week +My req +It looks like get and get are only meant to calculate shallow sizes +Should be live in an hour +It should be pretty safe though so please feel free if you have spare cycles! +User can always do their own dependency settings and exclude what needed in the pluginManagement atRevision Author dantranDate AM Thursday October MessageRevert Modified trunkmojo +the and were created but never read again when initializing the server +I was torn and as mentioned ended up just stealing what the jar plugin was doing +This is a great have Roo gvNIX projects without persistence requirements +Needless to say the property is only relevant to fox +Incremental improvement is always nice and if it doesnt work out we can always roll back +Logged In YES useridI attached some of the files I used +This appears fixed based on the comments in. +It now also crashes again in my real world project +Ill go ahead and update some of the other recent issues that I have resolved with this information. +added test cases for the bulk apiswill test on real patch fixed the findbug is reporting is not derived from Exception although named ending with Exception which is not an issue +Fixed in seeing hangs run in a loop runclustertests +Also please if you have time review as well. +Integrated to and trunkThanks for the patch SergeyThanks for the review Lars. +Julian see the test case please +Should this be committed +originally suggested by extempore on scala just committed this to TRUNK. +Im going to mark this resolved. +Ive seen it happen once or twice +This feature already exists as the extends attribute +Patch changes examplesbin from being a standard shell script to a bash shell script +I do think though the actual issue is not far from this +Ad The code comment explains why +Applied patches to the HEAD of all branches listed +Yup I understand Tristan +The unit test failures seem valid +Yoko Can you have a look into this one +The top level representing the Object Store probes for records using the object store API +These are not as accurate as the tables estimated row countbut if you call every now and thenthey should give a pretty good estimate +This patch adds the functionality of users registering with this contains a web page to get user info and codes for storing and retrieving user info form JCR +I think I figured out where is the problem +Sylvain thanks for the feedback +It runs the test times and prints out the average time across the runsIt turns out that the atomic counts are a bigger performance hit than I thought +wil change all that this patch depends on. +It would be great to get some feedback on this before I commit it in the next couple of daysIn terms of performance the rollback adds about to the time taken to perform all tests which on my machine equates to about for the feedback David those are great points and Ill get them addressed as soon as I get a spare moment +The same exception is thrown so this is not a regressionSergey can you provide a test case which works with but not with +We ran out of time for any additional testing beyond upstream tests within the GA time frame +It would take us further away from one day being able to do memcmp on +It should be extended to check test was originally trying to spot a different problem +It should now run on any OS +Will do for if no complaints +Hi LilyI think the patch is fine +perhaps we should not record failure until a job has been tried on two nodes? Then we could collect stats on frequency that a job fails on a given nodeAlso failures seems odd +Could contain an optional parameter to switch on or off showing the inherited operations? Just wondering if there may be scenarios where a client wants to know exactly where the operations were defined +I applied the patch and ran ant clean then ant again to the project. +Have the same issue +Many thanks DanielOleg +While starting the bundle preload the the most reason to do this is for having a chance to mantaine an artifactid bundleContext map if there is another way to do it we would do nothingWhile the bundle is active we could do any configuration lifecycle operations it will not affect the bundles status it only changes Geronimo repositoryWhile stopping the bundle we also stopunload the configurationWhile uninstalling the bundle we will stopunloaduninstall the configurationJust an idea not sure anything importand is missing +Dan take a look at that code +As always it would be helpful if other people could this patch in their own environmentsMac OS X Java compilerMac OS X Java compilerUbuntu IBM compilerUbuntu IBM compilerTouches the following fileM +Closing +Amila I see the registry patch needs to go into trunk and as you guessed probably the JS file within the js directory in gsoc sandboxjs folder? Shameera this is good to see the patches coming in +If you find any problems please let me know by returnThanks +But if well see net gains it doesnt mean anything about how it will perform on small set of integersThats why I think this test has nothing to do w the EncoderDecoderBut I dont mind if this experiment is done here anyway +This works for me on without any change that Im aware of since + file used in +Another point is that for HP Cloud different authentication servers have to be queried to get a list of all regions. +Good catch the under neath untar implementation is from Ant +Patch applied +Bamboo is happy with build +This is not a WSRP admin issue but stems from how deals with sessions. +Ill have a look as to what coul dbe done to identify the misbehaving connection +For instance it would be a mistake to reference JDBC extensions in code which is supposed to run on Java +More efficient fix test to show issue +involved into testing so of a bulk update all resolutions changed to done please review history to original resolution type +Diff for changes to and make OliverThat makes sense applied to make directory at repo revision +I would prefer to see both go in and we deprecate the push +Like which added these environment variables to the case where the hadoop script is run with a classname they should also be used when run with jar as the command +One outstanding issue see +Ok so thats correct but the session id is then not used at all +But hey they must suffer from backwards comp +The fix is about the subject of this jira only +I tried to use hoping that that would give me what I needed but it does not +now the Insert w identity column fails on Sybase with exception I dont know why this happen has someone fixed this bugwell Im working on BranchSPCP somebody can take a look at this +should help reduce updates to the root node +SijieCould you please review the latest patch +resourcesschema +That works +Ive attached the patch +We still havent clarified that you are able to pick up libs yet Please can you set up the file and set AXISCPPDEPLOY environment variable to the directory that contains it +Patch a implements the unlimited growth functionality and contains one additional testPatch ready for review +rev +patch applied in r +You need to checkout your project on your continuum machine for register your password +Please provide more inputs on the requirements and design +Attached is a patch to fix the incorrect state transition assertions +The All content simply shows all the discussions wiki articles and blogs in a particular space. +this issue closed +closing as WIKI is deprecated and EOL +This means that there is no mechanism on the site for users to navigate back to the at Apache JDO welcome pageComments on content or wording +I have the same issue but couldnt proceed with thisI could run the build successfully with out the agent but when I enable the DBM I could see the same issue as above +The other problem Im now trying to fix is the reactor order +This subtask is first step +An empty map seems to be initialized anyway it is replaced only if headers is not null +It just keeps things more organized +Excellent work. +Reattaching same patch for qa +Why not reduce the number of elements instead of adding +I just made those pass for now but they need an additional look +Doesnt look like the new languages have been added to the split targets +Patch applied thanks. +Overall Part I Chapter looks rather poor in content explaining the overall architecture +Ive committed this and merged to +I apologize but I cannot follow the full thread +We have a GB heap for these regionservers which are used for a cluster which is primarily doing random reads +For convenience heres a port that will patch to release of Todds patch +Resolving +See for alternative approach +I had a look at the source but the criteria query is really buried deep not sure I can easily get to it without making a complete mess of things +I have committed this +This might require some slight refactoring of the response writers +also on my hit list having to be super careful to call make in the right places with no type safety since its BB in BB out +The problem with editingdeleting an agent not seeming to have any effect should have been fixed by +Patch was missing a file +I am going to limit this JIRA to Comparable and open another one with lower priority for implementing the menagerie of other interfaces +after reverting the import changes +patch applied in r +I just committed this +For Concurrent see this to our next release cycle. +Unfortunately the random grouping tests fails now +Logging changes I made to Cache to help diagnose this issue +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Ive moved offset to the other side of the inequality +Along the way we built a better understanding of issues faced by frameworks like HivePig and came up with +I will apply after the release +This is just an idea that crossed my mindWhat if we can describe SOAP fault with XML schema and use that in code generation to enable the handling of faultsHere if we create the Java type for the SOAP Fault using whatever the data binding framework then it should be possible to deserialize the SOAP Fault to the particular objectThis way it may be possible to support the handing of the custom faults wsdlfault definition in WSDLWhat do u folks think +I am looking into it now +M +This patch adds two parameters and +First version of the patch for early review comments +This issue can be demonstrated in the quickstart app. +Thanks Ted its another test that was counting on having catalog regions +Thanks IsmailYou patch is in revision . +d favor an exception to be thrown not required +Simon I disagree with your comments here +The fix is similar to what Michael had but in different class when all the properties are loaded in +One of the files I was referring to is frameworkslocaleptPT Justin +It would be great if this could have been fixed earlier +Hi Michael Im new here so please forgive a potentially dumb questionAm I correct in thinking that by filling in Fix Versions above you have managed to reproduce and fix this problemRegards like you can see ticket status is set to resolved and it is fixed for the versions set with fix versions +now it should for fixing the missing quote +Can you provide a test configuration that fails and explain in detail how to reproduce the error you are seeing +Can you comment on these failures +Thanks Mike I almost expected thisIn my opinion the whole overhead of this thread management is making it slower times than the slow +that works for me +Anoop Kristam thanks for finding this and hunting down the root cause of the problem +So it is not possible to fix this issue +I just feel the change is more complicated than necessary so Im working offline with Arpit and Chris to see if theres a simpler solution +Will commit unless someone votes the patch down +New patch for release +Planning to get this one into to implementor the AMF and GWT endpoints currently expect only one endpoint implementation +not like now by looking up the class in the compilationUnit now +Seam Components View in new window remained empty because the window was invoked with null input and view expects +No its a private app installed manually via APKAnd yes thats the default factory version but the carrier sold me the phones with on Phonegap +Patch file containing diffs for java classes that are already in the BRANCH code +changing fix version to +But will keep open for now + +Tests passed cleanly for me except for the known errors in the upgrade tests and the errors described on +I added a bunch of new tests including one where we set readahead just less than the block size +I ran the failed tests from last run locally +You will see exceptions in both the producer and consumer related to problemsBasically instances of the producer are initialized +this second patch seems also to have been applied but I see no record of it +See previous AndrewOne suggestion I have is to strip this xts config file down as much as is reasonable +Fixed issue where this patch didnt work with direct WMS integration +So this issue should be fixed +Moving it to Lucene changes it to more of a supported API +Are there a similar number of references to the feature declarations in the class? If not do you have a reason why the feature should be defined in the class as opposed to the class? Does this follow a convention set by other projects +Sadly Im not really on the integration stuff in +I struggled getting it to work at all +Added formatting +Yes please +Yuck installing CXF overwrites the fixed version +I think I have also added all the features described by except for point K +Not sure where to get this package from now +This is a slightly improved one +for the patch +Please Sumit create your patch from root! Attaching a new file +Later the class can be expanded to parse and return other information as needed for other tests +part of the process outlined in idea of having clients watching on different nodes when waiting for application events is crucial to avoid the herd effect and it has applications other than leader election +The activemq die faster as it only can run LeoIm profiling with right now and I dont see a leak yet +closed for release +The problem is that it makes decisions even when it shouldnt +OK +One is to move forward the holder support from +Set connection timeouts to sec +Point taken on documentation but if you dont think it is a macro then what is it +Can you please specify the problem? What do you mean with does not work correctly inside a pageflow? What happens? +Some possible improvements Its a bit opaque where the jar files come from +Just look at the output format for WFS and adapt it to the getlog call +So this feature is available only for those platforms at this time +Sorry I figured out the problem no need to worry +Our query volume is low enough that the only reason we have two copies of the index is so we can survive a hardware failure +Also added some test cases for createString to +This issue was fixed in +This is going to run in QA so I am assigning this to nathan +Application writers need to think about setting context for containers and not AMs themselves dont theyIm talking about the set call which folks writing YARN apps call in their client +I dont seem to be able to find any trace in the repo +This patch adds extra lines to +But the Valve doesnt change anything the bug is present with or without it so dont pay too much attention to it +during a discussion with Paul I stumbled on the problem that this is maybe caused bye a wrong handling of DGM methods added to interfaces +but then we depend on the handlers +Fix commit reference actually patches some other doc. +But the best hope of making this work is to pass DNS names around and not IP addressesORIs it that the admin needs to take into account the difference between networking where he is and where the broker he configures is? In this case we need to improve the documentation and tools substantially because doesnt really make explicit from which viewpoint the network names are resolved +This was tested on a linux box +also fixed up help need to better enforce that the system keyspace should not be cleaned up +Bug has been marked as a duplicate of this bug +Thanks! +I have attached a patch which generates this pattern of codeThis patch needs to have already been applied before this will apply cleanly +This patch removes the incompatible change in the patch to but keeps the critical code to prevent the Balancer from overusing network bandwidth and avoid the deadlock problem that was described in +Let me first get the thread dump +This wont probably go into the release line +Thanks Victor next time just resolve issues you find to be resolved Resolved because the related is now solved. +Let us know +However ive been looking around for an example of a single part fault without any luck +buildivylibHadoopcommongriddevbin +Updated patches are for review because I changed the visibility of Storecreate and get depending on branch +Thanks for reportingIve committed a fix in SVN trunkCould you give it a try +If you have a test case for this you should log a bug against IBM JDK +Hi ChathuriYou are right +basically keeps track of the versionassigned to the action +Anything that requires the client to decode a token is a because of all the future incompatibilities it will cause +I think that all management calls are synchronized and run in transactional fashion +Typically hthe issue will happen after added a new set of shape files also this is not just limited to OSM data +Clearing proposed Target Release value from to need a link to fix and verification that it is in our product branch +New patch without wicket patch applied +Edited these issues to set the fix version. +Upgrading to critical since this is reproducible and causes complete pipeline failure for writers +The setting has no meaning if maxThreads is not set +The text looks fine to me +Short term constantly being rebuilt snapshot leave it on term snapshot rebuilt and deployed when youre happy with it put it on term its not a snapshot can take care of transfer of items between the server fairly easily a lot harder for end users using raise a request when you want it done at least hours before you actually need it +Hm interesting +Will probably not get to try that until Monday +Same here are anyone working on this oneIf not do someone have a good description of using the WEBDAV protocol instead for those thatcannot use SSH due to infrastructure restrictions +in ended up not being an operator +Im using JSP +Patches upgraded to rev +I just committed this +Closing the ticket +Some existing messages in this module were already marked with tags and were skipped by Eclipse externalization tool +See previous comments for details. +Olivier thanksAnthony do you have time to look at this +Done +It is needed in also +I know control s is a shortcut for saving but the tool needs to be able to detect that theres a model change for that thats triggered by losing the focus of that field +But there may be more empty elements can be removed only if the removing wont result in any deployment problem +Going forward we will continue maintaining the branch +DmitriyThank you for your comments +What does the stacktrace for that NPE look like? for the report! This should be fixed in the next +Some early thoughts +Thanks James +A workaround might be to change our root POM file to specify an older version of that plugin +This patch ensures the pluginManager script variable is always initialized in the loadPlugins target +i am not able to apply the patch to latest trunk +Certainly a neat feature +Thanks Dave! +What about this implementation is so attractive +This is the war that I was deploying when I encountered this problem +Origin of test +Actually the initial patch does not avoid the exception Im seeing since the start of the token is ok but the end is beyond the strings end +I am wondering if we should push that to as sure if it makes sense to remove support for V after +But Im not sure since I have been reinstalling and testing so many different versions lately +Fixed using patch from +commit dabbdfbcccbbbccfbaf +Please look in history for what the actual resolution was. +This is difficult to realize since the http error code is readed from response in the fetcher and setted into the protocol status content analysis can only done during parsing +This patch fixes the problem of removing root privileges and restoring the original user +There have been some errors in the code which havent been displayed through the IDE +You have been very helpful +Please verify that it is working for for the good I checked out the branch from Subversion and verified that the patch works correctly +Manik the results with CR are the same as with the snapshot I reported +Suggested fix +See if we can get the value from logj via the jakarta commons logging api +Existing implementations support configuration through injection. +Hi MarcelWhat else should be done in order to fully implement composition support +The only two ast xforms directly calling this method areHm I think the error above did not occur before introducing Delegate but of course I could be wrong and it is caused by some Grails xform +bulk close for the release +Addind Doug Meil as Ted Yu said he owns to TRUNK +yes we can not guarantee the message ordering at the transport levelBut I think this means RMS client should not send the message until it receives acknowledgement for message +Database owners also cant GRANT or REVOKE privileges to object owners +Offering resources templates page in Code Assist window even there are no +Steven another idea what if we simply added the options to so we could turn off the tokenization of title and date fieldsGood ideabq +Seems valid though although I doubt the usefulness +Thanks Ted +Maybe I will look at the conversion later if it is easy I would like to reduce our dependencies to +closed as per user comment +use is null obviously +It should come from the cluster in which the app is running +ping +Please let me know if you are able to do this +After updating the plugin to the RELEASE version i was able to uncheck the option +Even i had the same problem +Please reopen if found. +Please look in history for what the actual resolution was. +WAR file with simple +I will chat with you on IRC to figure out which approach is best for our current situation +Walter wrote The context required to resolve the ambiguity is a wiki page that I didnt know existed +The root Entity makes its way into the index but not the sub entities +Is there something I can do to help track down this issue such as going through the FLEX source myselfRegards +I think that was a from Jon Im assuming it was a smiley face +needs to be upgradet to the +Issues that affect Tomcat standalone should go the the Tomcat issue however it is not a bug +Thank you for fixing this bug! +Um what did you change +Also you cannot put some jars in solrhomelib and others in a location specified by sharedLib they must all be in the same locationThe same config wont work on and earlier if you are putting jars in solrhomelib +It is nor possible to set any property starting with +Close this for now +So this patch does not degrade +What about then? I guess it sounds more straightforward but its up to you +If we want to implement something like we would need this option +Finally I decided to commit my changesBilgins tip at least the Financial Account Summary screenlet will function correctly for nowCommited in revision +We know what the problem is the Sonar Eclipse plugin embeds version of the Sonar Findbugs plugin and not the one of the Sonar server +Might be a little +Presumed fixed by code review change ofPlease reopen if there is new evidence. +Marek said that the URL redirects to automatically so we should leave it as since thats the official one. +SebastianThe differences in regioninfo that you have mentioned is as expected as we moved the HTD out of HRI +How to compact the index CF? What is its name? Ive tried a variety of things and cant seem to find the magic expression. +Please build the plugin from the repository and test it out +the attachements are no longer functional + this is a reasonable approach +Thanks Mike +no real changesThe final version of this will need to move updating logic out of the processor into the m having second thoughts if this is a good enough approach to really put in core SolrRequiring that all fields be stored is a really large drawback esp for large indicies with really large documents +s con no tant si vols com si no vols de grat o per fora vulgues no vulgues +Thats why is a terrible better example of this sort of dependency is XML Beans +Thanks Yongqiang! +I think to fully support this we have to generate bytecode +. +Closing resolved issues. +Our stack trace from the running node in question +authentication is usually done on assuming you are talking about client +The changes span both HDFS and Common but I attached two separate patches to this ticket for nowAs previously noted this patch addresses the same core issue as +In most cases the pk attribute is not required but when it is mentioned it should take for your usecases the pk attribute is not required +FYI the release is imminent just waiting for tested it and you appear to have resolved the it took so long to get back to YouStephen +I added a small instruction how to invoke Doxygen with this file +You may have discovered a bug in Grails s ability to allow you to provide your own GSP to preempt the GSP provided by the way this is not something I can address in my plugin +in? cBuildetcconfigsrc? cBuildetcnlsgenlist? cBuildetcnlscharmaps? cBuildetcnlsposixgenlist? cBuildetcnlsposixcharmaps? cBuildexamplesmanual? cBuildexamplesmanual? cBuildexamplesmanual? cBuildsrc? cBuildsrc? cBuildutilcodesetaliases? cBuildutilgenlistOf those only genlist and codesetaliases are actually expected to not have the ASL header +for the patch just committed it. +I didnt find the libraries to work without the XML libraries +From the discussion on the mailing list it appears someone needs a PDF editor and to do some hand edits to the PDF exported from the wiki prior to commit +Sure it makes sense +Farhad thinking more you should be able to make it work for mkahadb just have a single kahadb nested instance use a locker and for all others useLockfalse +It turned out that allowed types to coerce to void by just returning null and stricter coercion logic in prevented that in a number of cases including for String +Heres a patch +Posting empty forms works as expected now +The start will also be called from the deployment which is created by the deployer and will be executed in a different context +This causes the resolve issues mentioned above +Zip of updated images folder received from Mike m a little confused +Access Violation Thread thats helpful have an idea of whats happening +As I said I would rather see them combined into one like my example since it is easier to see the relationship +bq +But this should be possible to do for other libs like avro +Attached patch to fix the committed thanks pradeep +We could try to test to see if we see a difference in practiceit is also very wierd to me that the method you are using is the one being used in ICU +I think because it is a VM combine changes +Works in. +It complicates the public interfaces a lot and doesnt have any paybackIn our experience given a file format the code is pretty independent but it is tied to the fragment splitting +Its the responsibility of the serializer to know the schema that its to conform to +Moving this to by Herms request +What if you were using a custom persister that you specified on the class? Shouldnt you be allowed to have access to that information when trying to instantiate an objectAccording to the javadoc the interceptor should be used to create an instance of an object and then return null if the default behavior is to be used +I tried adding a to on seeing if the focus was the problem but it still didnt work +Should I open another issue or shall we just rename this one to something like fix links in the component reference +Should be protected at the package level like +Maybe there is an issue in surefires resolution itll need to be investigated +HDFS doesnt have any of the security issues that MR does with user applications having web interfaces which was the main motivating factor for developing this framework for MR. +I believe all the issues which came from comments are covered specifically thus Im rejecting this issue +This still has to be integrated into the current implementation +As are startedstopped there will be many that are just stopping and this method will create way more running than requested +Is there any way this can be applied? Every time I upgrade Seam I have to this myself +So if this is per spec is there no simpleclean way to enforce that empty ListsCollectionArrays are returned if suppliedDo you have to provide a to control thisChanging the return type to an array doesnt seem to matter eitherWorking per spec +Tested the solution on local PC and seems fine +This helped me a lot to find the problemI attached the patch to fix this problem +Id say leave itIts actually exposed two problems that I see +Increasing file descriptor limit and backlog are administrative operations on the OS +I dont know if there is much we can do here +Could you post an updated example that uses only standard fonts and no images +importing proven code would be okAdding more and more Maven dependencies to small single owner projects is the concern +As the numbers would indicate many files do have a large number of revisions so there is a large amount of information for a cvs log command to sort this case it would be considerably faster to execute a cvs log against just those files that changed +then they should call addScalar at least once right? +Marking as extension as this seems to be ffi behavior and not specifically this library as far as I can eyeball +So +done at rThanks Youssef +Its a minor bug anyone interested in can pick it up +Thanks +Merged into develop branch. +The experimentation mentioned in my last comment was a success +If that is the case then the most efficient way to solve this issue would be to open a PMR for that product + +Mostly my original statement was meant to mean Is going to have lots more releases? If not the would be nicely supported in LangHaving them in is a bit odd but as theyre there theres no point duplicating unless fades away +Im not sure if we should store the password in the secure eclipse store +Delaying resolution as new maven embedder wont be available before release +Reverted the patch confirmed passes +wait for committer commit to the svn +Here is the list of affected filesFunctionFunctionPayloadsPayloadsSpansSpansSpansSpansSpans is ready to goDIGY +The second you are about to share these with someone else you have to have some reliably reproducible build that has undergone some form of release +work for but doesnt work at +Thanks for fixing thisYW! very good feedback I will port it to soon +And as I said the restriction is that the results cant be worse than they are today. +Would be nice as a first commit for you to do. +bq +The typical scenario is when we need to replace a faulty memory module or CPU in our server but have critical tasks running on the node which we need to allow to finish +Mapjoin Select operator is split into one single stage +Now the WS connector doesnt work at all because AS is trying to type cast value to long +Committed trunk revision Committed revision is the missing test with real byte streams and charset inside document and outside document +Will look this weekend into next week +It was result of changes of +Thanks +However it is a very good starting point for collaboration +comment string in etc +Attached Sources of the Clirr plugin for Sonar The binary +Destination maintains the list of files in and takes control from this point +is this still a problem? I thought this had been in xwork rv +Dunno why this hasnt already been closed +Looking at my browsing history I find that I was going to the correct URLsI cant account for the difference and Niklas comment that I didnt have an account on the old system also doesnt jibe with my experienceRegardless the problem seems resolved. +all the files should then be moved to the Attic and preserved that be all set now +Applied patch to tomcat builder in rev thanks! +As the massage context is freed at the end of serving the request this instance of the stax builder will live till the end of the request and get freed when message context frees its propertiesAn example of how to set the property can be found on line number of where the http client instance is set as a property to be freed in a similar fashoin +Back to send an but handling it in instead of a special case in the +A combination of circumstances including some quirky behaviour on the part of Tortoise svn caused the previous problems +Is this really worth it? Set maintenance is more expensive in pretty much any languageEDIT Make that in all languages +So the new API should be a convenience to developers to get those inner beans when we need to know those inner beans +Theres nothing wrong with using NFS HA if you already have a substantial investment in NFS filers +Keuntae thanks for the further explanationYou are right +The changes related to the issues have been rolled back +Jason can you please take a look at if using system dependencies in dependency management is not supported then it should give a clear error but not the NPE +Possibly something like this +This was fixed by the commits listed under all +When gets it merged with the trunkIs there a way to override it in general so it can be used for all queries in +This is not an issue but rather a question to be asked in our mailing lists. +Hey Emmanuel So from what you have said should i understand that the epoll selector bug still exists in jdk or that you are including the workaround for people using previous jdks +Always set the time BEFORE creating the the initial time event +And I strongly recommend to stick with at least one of them instead of coming up with an own proprietary creator scheme which no one knows aboutIts nice to be able to be pragmatic but not using standards for metadata today causes a lot of headache in the futureRegarding Geo data Im ok with using the WC properties for the core properties +Better. +Verifed on SOA ER. +recap findings this far Thomas thanks very much for all the investigation and experimentation it is very helpfulI think you should assign this bug to yourself to reflect that you are studying itGiven the subject of the bug and the fact that the behavior was markedly betterin recent code I wonder if the fix for is part of the behavior improvementsI see that Mark Helkamp was monitoring the handle count while you have been lookingat Java object memory usage +Proposed fix +Nice play on monopoly guy concept it is a rich face Gems have facetsfaces and so does JSF +can we improve the documentation with javadocs in? I am happy to open another jira for it but seems a little redundant +call LIL instruction now provides shadow space in stack +that they can not use to open new window +I think that you missed on some of the files you must have made changes to else the unit tests could never have passed +Remy could you please have a look + +This is quite close +I apologize for the confusion +Hej DennisIm happy if I could are right +Here is a patch that seems to fix the hang +I see the patch that you provided but would you mind telling me exactly how to use it where to put it etc? It would be really helpful if our links in the pdf are underlined +In order to exercise it youll have to do it directly as commons logging will have given up on logj and be using something else by nowJust creating a logj logger with should do it +submitting +As I found out putenv should be used instead +Issue closed +Not heard from Neil as to whether it fixed his problem +I willprobably have to find a better mechanism for thisThank you very much for all the feedback and help +Fixed +My understanding is that its not part of this patch as its not required to get CITE tests work +All priority was broke +Sorry Alexey it was still unassigned as I was looking at it andannoyingly JIRA doesnt tell you when there is a collision like thisTim +This patch provide fetch time saving up to +Cannot replicate that on Ubuntu Open JDK it is rather possible with KDE +We want it to read any directory or file regardless of filename +I did couple of runs this week without success +Seems to work fineWill this patch be included in the release + +Attaching a patch that refactors the test code somewhat +Unit tests pass with ant clean package with you on not breaking backward compatibility +These blockers and critical issues are resolved but not verified +It is a regression that occurred sometime after the Seam release +Looks like uses all raw types and javac warning is expected +Applied patch retested +That actually covers all modules +As Olivier does all the great work here I assigned the issue to him +But if the workload has changed such that it has us changing configs +It does break the new ant target I made +Updated the reference basic transforms and the connector logic. +So your example would throw an error +Context menu seems to be already skinned +A test case demonstrating how this issue occurs under normal usage is needed in order to reproduce and investigate this issue +Continually working on this hoping to contribute for release +We had a lot of tokenizers abusing setReader for stuff they should be doing in resetIt would be really nice to make setReader final but that sounds like a challenge +Theres no root cause there anyway so that is why we might need the test if we cannot get the reliable repro otherwise +So I wonder if we should post pone this? +for the patch +Is it running out of mem? Its not because you lack MAVENOPTSXmxm or something +try the Schema in and confirm that it works for you +This time I went and manually deleted CDocuments and SettingsuserLocal SettingsApplication DataSunJavaDeploymentcacheThat seemed to helpThanks and sorry for the bother +Committed to trunk. +Eric what is the advantage of retaining only one replica? Unless it is temporary data that can be easily regenerated having one replica guarantees data loss in quick order +Thanks Marius +FYI Ive uploaded a patch on that updates to add more complete instructions for building on Windows +Since this issue is not critical and hard to reproduce we are moving this to next releaseRegards to next release +I still think that calling is is better because the message formatting stuff may do unnecessary things like casting autoboxing etc +Revised code available at httpakkumar +I would recommend making a very simple test in an project that just calls navigate to eliminate any potential side effects from Flex codeThe first variation I would try is to not call encode on group but rather call it on each phone number so the comma doesnt get encoded. +For this case we must be able to use dfs to do rms +released +Well have to push new versions +A tailored MVEL jar is about KB in size larger than KBI think we should run some performance tests and see how it comparesYour cents is always appreciatedregards Malcolm Bob for the analysis with regard to the performance numbers what is the difference between caching cumulative and thread caching cumulative +According to your log file youre using an Oracle DB but which version of Sonar are you using? Is it a new installation? my understanding is correct youre fixing bug which is fixed with Sonar so Im closing this ticket. +Unit and integration tests pass +will add that feature as part of this issue +bar + +Should probably not include this in SVN anywayWe need the JSP jars which dont resolve through ivy +could you check into this +I suspect its some kind of weird consequence of Groovys updated metaclass Testing plugin then will just have this as a known limitation in +Im to commit it in master once is out so we can have enough time to test it in different browsers before is out +Chrome works fine +The client just stops receiving messages so isnt much betterThanks for the ignore +Figure Image is page +Nice didnt know about it either +File Contains rebranding of servicemix table name to karaf +The idea of embedding and importing at the same time is flexibility in terms of what is installed in the framework +In patch the idea is that when there are only one scanner left then there is no need for lazy seek +But the Second Level cache does get written when I set thatWhat good is the then? I can only avoid writing newly loaded items to the cache but I CANNOT avoid writing objects to the second level cache in a saveflush operation +Structure is needed to correctly initialize and acceptance tests with this code +Davis has asked me that the st time I did it +Take a look at methods for billing account balances and youll see what I meanIf the code I mentioned is called then this would happen and it would be a bug +XuPlease read againWeve changed the string input semantic from being a normal char to being an escape char +defer all issues to +Fixed committed with tests +must look into alternative ways of doing the weaving +Thanks Gavin +rb files in the project to +Adding link to ticket about going offline +Problem is that maven doesnt find the needed classes +Its fully reproducible in my production environment so I should be able to get a thread dump at a minimumI wanted to supply a standalone example but that might take a lot of work to craft an app with an embedded db table creation data etc +Im not sure how this code would have been released without it +This issue seems to have resolved itself with an update to the latest snapshot +RegardsSumit Pandit +I verified on both Mac and Windows +Glad to hear that +Can you provide us with an example on how to use the max parameters when using the http component in a spring context file. +ISTM that FD processing Gossip updates synchronously is a fundamental problem +yeah creative the same way of actually using IE +Passes locally +Hold I know +Thanks! +Done added numinitialsrvmembers and breakoncoordrsp +This is not due to any code gen error but the build script + +I think it would be better to say in our documentation the following unicode characters are valid in a topic name and just have that be definitive +Dont you think the Priority of this should be set to a higher level? For us this is a showstopper because we cant show validation messages for one or more properties of mulitple instances of a class +Right now not only must there be an existing working copy but the most recent build result must be a success before you are allowed to begin a release +If someone else wants to make an attempt to fix the script feel free +So this patch needs a few updatesHowever the code looks good +Now it is not possible for plugins that are outside component codebase and are not part of the build +I think a slight performance degradation for BLOBs is an exceptable trade off for not having them materialized into memory +I have removed the file from the as its not needed any longer +Thus currently Im working on debugging the code base +The only jars which contain service loader configs are and and the jars both contain the correct impl classes +As requested raw untidied source +My only question is about the default value +so i think it should not be added to +bulk defer of issues to + +Worth investigating though +Will you point me to the code where the additional loop iteration would need to be added so I can understand your performance concerns? +is not pure problem +VERIFIED +On the other hand when I run google provided wizard called Web Application project I am then able to select GWT facet together with required Web Module and Java facet but facets are not enabled by default in such type of project +Its easy to sign it in place though with a different syntaxSystemLibraryFrameworksVersionsHomebinjarsigner verbose keystore UserswoodjLibrarymy storepass keypass VolumesJaguarUserswoodjDocumentsprtprttrunktargetjnlp way to fix this problem is to sign to a different jar file delete the original and rename the signed jar +If someone can share any input that would be great +Done +RHDSupdatesdevelopment update site. +In my case i want to write a CRUD application for administrator but we already have JPA entities +Standalone JPAEJB test case to demonstrate this problem +Just tested with the ftl without the region it works the other hand I called the webscript directly in the browser it works I guess the problem comes when rendering the region tag +Please see parent bug for patch +Patch submitted to trunk +This is now duplicated by +This patch adds a new thread to the that every three minutes checks the list of launched tasks that havent been reported yet +This is exactly the mechanism youre proposing to remove +Tested +I just think disabling all zk INFO log is excessive +Preparation for Gora release candidate +r +I have changed to include component template file +Attached the sample project used to trigger the issue +Ive also prevented it from being added if there are no pointcuts defined so the additional lookup wont be required unnecessarily +Thanks +This effectively means that the NN is free to resolve all symlink targets because the client has no intention of resolving them differentlyThe downside is that wed be leaking the viewfs abstraction down to a specific filesystem implementation +Last patch missed some generics in and looks great Uwe thanks +Steps to reproduce attached and the log +I changed the accessors on Schedule and +has been marked as a duplicate of this bug +An updated patch for earlier version of hadoop +and deployed new snapshotPlease close this issue if this fixes your problem +Thanks for your time and contributions Markus! Looking forward to more improvements and reports from you +Can you provide the steps to reproduce? It is just adding a shapefile? Or any other steps involved +Can you take a lookJesper is this really critical for releasing or can we downgrade it to some other priority +Thanks +Well get to it soon +Even if it were available the bindings dont match +. +running tests +Once upon a time a major factor in changing jobs was over an argument over Vector vs Should the declaration be of type List while instances be? Do you want help? I think I can tackle this one with searchreplace? then tests +Thanks a lotForrest +This patch should result in EAP using JAXB +This is fixed indeed. +this is included in in JBDS Build id to add newandnoteworthy +The same issue on version +I uploaded a patch and I will upload its design specification on today night +To hook into the releasejavadoc mechanism +We have been using ODE in some of the enterprise applications at Tata Consultancy Services +Thanks for the patch Billy +I like Cyriaques idea of plugin specific CSS for this is a new Issue to implement this in general +I no longer see the infinite looping when running my test. +Yes this feature is not really required for materialization +Issue is resolved after issue is resolved in turnk and on. +It should be in initialization phase somewhere when indexedembedded fields are being analyzed +Btw +DWe could add a message dialog +Maybe it would be a beter idea o ad a extra handler for that +Can you explain a little what kind of extension you have in mind +Heres a patch that implements this updating the spec and implementations +As I wrote in my mail try to remove the basic auth from the nginx config +Thanks Sanjay +Please verify and close +You should know that +Have you found the need for weak references to beans in the past? We have an EAR project consist of Modules and s of services We are injecting these all services through spring IOC in these some of them are singleton and some of with other scopes +Or we could just have the public static final and add laziness ifwhen it becomes necessary I am fine with either +Rajith please the issue if you still have the issue. +Server works on RHEL now but it doesnt work on Ubuntu and higher +It involves fixes of those holes add the flag to tell us what statistics are accurate and what are notWe can discuss more on +The same MC version cannot be defined as dependency visible to all target containers +Easier configuration of declarative transactions in XMLSee page springdocsreferencehtml may want to add the main content toSee page springdocsreferencehtml reference the tags in an appendix like this oneSee page springdocsreferencehtml of information on the tags include the Core Spring deck in Day of the Core Spring course +The segfault is happening in the ICU transcoder which means you must have selected ticu when you configured the parser +Until then my version of is flawed +Thanks for fixing that are at least one issue depending on that one +I wonder could you try the current build please? Weve fixed a number of issues in the Stomp support to do with handling disconnected stomp consumers which could well fix your I will try that +creating class for plugin is deployed +In this situation if the outputs of multi map tasks on the same dataset are different for example outputting a random number the outputs of maptask and the maptask will probably are different +Then the is retrieved during the deserialization and the types of the two objects are compared resulting in the error +Addressed Ashutoshs comments +If no one can come up with a reasonable automated test for our existing testing framework I will request for an exemption after more through testing on a real cluster +snapshot of findVisits request after user of findVisits after user of findVisits after use login again. +When I referred to shards in my comments previously really I meant replicas +Closing as requested. +Please review Artem new patch looks goodThanks for the tip about! Ill update the TODO with this information but I dont think we should pull in a dep on commons for this +Im not sure what that means given that the test has only a single GET +The chance to hang in deleteTable could be higher +This cannot be done +Thanks Devaraj! +Thus if consider it is specification and note a bug we have to document this change +patch file for the disk aio api change for upload proxy support +Attached a patch for the file found in standardtools SianPatch applied at repo revision rPlease check that it was applied as you expected +Updated patch to test the new public API of +The change makes sense to me +Is it a problem of build? +Has anyone had a chance to look at this issue further since then +Portlet support is dropped from Wicket because none of the active developers knows that technology and uses it dailyIf you need fix for this youll have to provide a patch + was due to holding Combiner test inside +consolesrcmainwebappviewsecurityrealmmanagersegroups +Test checks that watchers Hadoop QA to it since this is a git used to script to convert git patch to svn patch +Also possible there is a Thrift bugEither way unlikely we can do anything about it +The sample web application uses +Originally Ben is complaining that updateReaders is taking a long time so long a flush timedout and regionserver crashed +says that will caused dead lockso use versionthats okno problem +Now i am getting correct response +This is a duplicate of bugI am sorry for the duplicate + for patch +Should references to ibatis and be removed fromIf so should the iBatis LDAP groups then be deleted +Its actually kinda funny that the existing line isfor f in for f in GIRAPHHOMEtargetgiraph +If this is still broken please this issue. +The unfortunate limitation with patterns that point to the root of the classpath is caused by the JDKs behavior not by Spring which is clearly documented in +That client is not exposed in any of the javascript functions runs but specifically exposed by the test harness +it seems like we should keep returning a NULL +Is there another jUDDI jar there? Could it be picking up the wrong one +Problem solved by other way move processing of a single attribute from client to server side +ronald i dont hate you +Marking this resolved to be verified in EAP ER +Nice will try this out this evening +I would rather that a person gets the opportunity to deploy it at any time and not force any other dependencies +Workaround until I can get a new version of the plugin release drop the attached file into the scripts directory of any grails project that has the plugin installed +So we must postpone this ticket +This fix to this problem has been incorporated in svn revision . +bq +bq +The log information is definitely misleading and it would be nice to change itoCurrently Im defining the endpoint on the target level but if the address is also used to compute the endpoint reference which one will be chosen if both of them are defined +This is the patch file for the branch. +if it isnt I can adjust Jesse I didnt apply these before I added the forking code +If the last comments suggestion does not work then please reopen this isse. +I think the only way to test it is to create a branch apply it and then force a build of the branch +Forgot my previous comment this is solved in closing. +Thanks Dirk very nice additionsimprovements We dont need to collapse the navigation as it is a navigation +Thanks Shinobu! revision +If its not done on time would be acceptable +ThanksIll have a look at the existing pull request +Would be great if this could be fixed for is to put the choice in a group and have maxOccursunbounded on the group rather than the choice +is present since we are able to show SID and ECW files +Hi Galder I have attached a unit test and some config files +In the patch you omit some tests with the comment The following tests fail due to limitation +patch looks good +Thanks for the patch. + has some ideas on how this could be used +No sorry +I removed the sleepycat pom so the pomjar can be +cvs diff of the committed to cvs. +this issue does not affect jcrspi any morethe only usages of in jcrspi got removed with revision +no page outside of pageflow is available or +Anybody has a good idea for a unit test +Indeed Dinesh I agree that we should improve names and descriptions +Ill try to figure it out +in totaly agree with you Doug on API back compatibilty +Verified that the pid generated by LCE matches to that of the running container +New patch removes syncing completely as a this is only called from and b flushLedger and flush both have their own synchronisation +bq +Done Freddy +Use the Roo backup commannd +for the patch +Unit tests the use of multihome configuration with objects is tested in tests are passing in hudson. +Works fine! +for a new patch for branch +Script creating the database +Reapply to QA test +r was checked works fineThanks! +I would love an intro to Backtype because I know theyve created a patch in the past for the Python contrib +Do not commit this patch until I have had a chance to get it reviewed and run tests +This leads to a Stronger codebase which is a good thingThe add methods predates DP and not much we can do about it unless we deprecate them at some stage +Am I missing something +Hurray! Have a nice day! would have an effect on the ordering of dependencies from a plugins POM that are used in the plugins classpath +How is it looking? I really really need this one fixed +bulk move of all unresolved issues from to +Revised backport patch in sync with trunk patch for yhadoops. +HiI found the reason for the circular dependency the POM loading part should only be invoked if the artifact in question is not a task +Thanks Knut Anders for your insights on this problem +Marking this JIRA resolved as we have had a Maven build for some time now +XML EditorIt does not even show me the option of our XML s weird is that it does display the file in pink so it knows there is something special about it +If it is purely a state I guess it should be OK +User will only have tospecify the system properties if he is using custom keystore truststoreThanks a new patch for g after incorporating the comments made name of the file is +Have the same issue +I think one big patch is preferable +Committed revision +Actually I dont think it would +There are places where I have made the ContextOther some of which might be wrong +I solved the problemby prefixing clean to the ant task list as in ant clean test +Patch coming +I applied the trunk patch +Something weird happened to the testsReattaching same patch to trigger Jenkins again +The bit of code that gets an underlying DBCP connection could be split out into a utility class and put to use for the streaming support +Thanks for committing the patch Knut. +Also we should add this as list of jiras along with the maven backport to +ant test and ant passed for the common projects patch +Preferences view after changes names +Would it make more sense to have one package instead of one plus core +I just stumbled over this problem as well +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Lets verify +nota patch file compat Cheers will download your tools +until we get a quickstart of some kind cant proceed +Fixed +Maybe this should be reverted +Place this file insrctestcomopensymphonywebworkviewsjspui +Minor change please review +We have to abandon relying on exact row count and enable UNREFERENCED rows lying around I think I can live with it +This is fine sincehcat will share a lot of that +Now I could move this to issue +If you could take the time to create a small example of the phenomenon that would be helpful otherwise one of us will have to do that +Committed to trunk and +Committed at subversion revision +I have tested with farming Tomcat Native Clustering +Am definitely not keen on having dependencies on SOLR in pretty much every indexing filter +Isnt this a duplicate of +ESB Editor Reference Guide is Reference Guide is Tools Reference Guide is checked +Test Richard Thanks for giving the insight into this issue +Scheduling for +It seems this is no longer an issue +I think we should pull it for +Would be really good to see use cases description and test projects that demonstrates the problem +Windows +Fixed and Committed in revision on Dec +used the javaName generating the variable name +Lesson learnedThanks for joining the IRC and talking this overAs discussed in IRC we migth ask yourself the general questionWhen does an DOC API user needs an exception when is it more comfortable to catch it internally and only logWe will have to discuss our exception handling next week in more detailIssue is indeed resoved +It fails to build because it cant find version of the in the central repo +Only articles which contains only uploaded file are affectedNormal articles are OK +Here is the short summaryFix for in client under Security Manager when trying to connect using EUSRIDPWD mechanismPatch removes the unnecessary addProvider call in that was causingthe +Good idea. +Instructions are included inside the attachment +Thanks +Would you please have a try on the patch? Thanks very much +Patch committed with a minor tweakMike +Another workaround would be to customize yourself if you never usespecify then you could just return null straight away and cache nothing +We still need to fix this sooner or laterI got furthest by replacing it by and. +Committedr +It looks like the resource localization being performed by the is also leaking filesystems +I have attached a version of that has been updated to use the Tiles +This would help in compatibility of with future releases +This second bug is not related with jitting and then neither with the other bug you the exception probably I know the cause and if I am right it has been already fixed on the Id need a reproducer to check it by myself +yep I agree with just from watching the Lucene discussions going on lately I think it is worth letting a few more things be worked out before using a nightly build +I am attaching the wsdl and related xsds hereOzgur with a cursory look would you be able to tell me what might be wrongThanks! +Yes I am fixing it now running tests to ensure all is still okay +I can get a higher quality scalable version make if this logo is accepted +move command already piggybacked on copy command so tweak it to delete the src files after it copies changes look fine to me +No apparent issues after port to as tinderbox and daily test results for have been clean for several days now +Weijin Is this going to land in? +Sorry for the late reply but here problem was related to the use of IIS as web server +javatestingorgapachederbyTestingfunctionTeststestsjdbcapiModified tests to get new value for locators +Maybe it would be better to specify the stored field and the index fields separately? Or maybe to retrieve the index fields from the terms of the query? What do you thinkI dont like retrieving the indexed fields from the query what if you dont want them all? how can you make sure that the ones that you take from the query really do share the same stored copyAs far as calling out the stored field and the indexed field separately I think I like the idea +Sure Ill upload them shortlyThanksRatnesh +Confirmed fixed +In other words Id like to understand why switching from to makes a difference at all +Got a sql script we can use +There are valid reasons to add spaces for example if justifyis specified +Is this easily reproducible? Can I reproduce this with the single rule above or do I need variations? If the later then youll need to provide me with a self contained project that I can run to this +Aint this just one side of the story? Where is the enabling ssh server bit +Attached followup patch to make it +checks for token type gram +sounds fair but dont we have sensible defaults for most of these or +Ok I just saw the images and that the WGS is the CRS so ignore that +bulk close of resolved issues +These counters are sorta like to illustrate that MR is working rightHowever if tools like Hive and Pig rely on counters this is pretty serious for their move to new API if theyve grown to rely such a count behavior in the old API +Patch is attached +Hmm I am not seeing this error at all but I am using Win bit with bit VM +Hadoop was released today. +Philip this looks good to me +I read the description the other way around +reading a record key value read line value read line value read line +fyi i fixed and in my travels. +The version contains an inlined version of the OAEP padding routines that take a custom digestTested with still need to try +can be marked as superceded by if we agree to implement that +Note works only in Project Explorer Rename +By the way I have opened this bug at Carlos Sanchez suggestion on the mailing list +Added multi threaded tests wrote test forTo StackTed Yu now is a good time to review +RTMP plugin would almost certainly not be a streaming implementation and would likely be able to store the the contentSo that seems like the plugin would not be covered by the licence after all +proposed required in +isnt a like ae introduced that test failure +Allow localization of feature params +I think it has been fixed in +Thats the solution thanks to Michal +The classes are in their names are and +Attaching a new patch +pigunit is for user to write their Pig unit test +Steve I see where youre coming from +Thank for clarifying Radim If this doesnt cause any problem right now other than documentation lets keep it as is do you agree +That was fast +I guess we should fail gracefully when the file references a class that does not exist +properties renamed to and respectively. +Ok Its clear now +if the approach accepted in this issue is ok I just need to figure out how to convert to the dct mem format and i suppose it is doable +FYI I ran our own tests on an machine this morning hoping to tease out some more reliable failures but to no avail +Because the lookup method is internal this is a pretty minor issue +So in test execution time they are visible to test so then there is no problem +Doing so resolves the scenario in m not quite sure why the recommended version range of the original takes precedence however +Also this is a question not a jira bug +Checksums are implemented to catch thesekinds of errors and there is not anything derby can do if the page IO does not complete correctly except report this error +But make sure not to do this until after logging something with commons logging +Closing all old Resolved issues +Patching xalan only solves a half of this issue so there must be another point which consumes heap +Was a regression that I introduces when rewrite it in waySorry looked at your code doesnt it overload too much the processor that way +You can always run particular tests from the IDE or on the commandline via mvn +Also per driftx I ran gms trace and grabbed about minute worth of log entries that I have attached here +Ps the classthats not used anywhere is it? I think thats a copy and paste leftover from the reference class we used from the oauth projectWe might as well remove it since it might confuse people into thinking that we do actually use that and use dba functions to store our tokens +Checked again still looks done. +The patch seems to make sense to me and should fix the core issue +But how long is an appropriate time to wait before moving on since EOL seems to be the incorrect timeframeFor Accumulo if you are interested in switching to Java I propose you call a vote on the issue +He is driving the Policy Plugin +At minimum we should produce some exception error or warning not crash +An empty set doesnt jive with not stale so instead I simply return if the directory doesnt exist +Waiting for review to Mar +Could be that this got fixed when I did the new popups and was removed +so I dont see how this would fit in the future at all +The change is quite small and Ive asked on the lucene list if theres interest in merging the change +Thanks Martin + svn ci Sending assemblysrcmainjavaorgapacheopenejbtomcatcatalinaSending assemblysrcmainjavaorgapacheopenejbtomcatcommonSending containersrcmainjavaorgapacheopenejbassemblerclassicSending containersrcmainjavaorgapacheopenejbassemblerclassicSending containersrcmainjavaorgapacheopenejbassemblerclassicSending containersrcmainjavaorgapacheopenejbassemblerclassicSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrctestjavaorgapacheopenejbconfigAdding containersrctestjavaorgapacheopenejbconfigAdding containersrctestjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbspringSending serversrcmainjavaorgapacheopenejbwebadminmainTransmitting file dataCommitted revision +make check in testcpp didnt do anything for me +Will boot db wombat in RO mode +And when users set the heapsize this tmp entry would be overwritten +Example code +A +However they can indeed see each other using VM endpointsIt seems Camel creates separate camel contexts ignoring the fact that they have the same context id +waiting on the hudson results before commit +I should be able to connect to that with a browser and see some simple status message that lets me know all is working and is running +New version of the patch now with added files +Thanks for the investigation Myrna made it easier for me to find the problemThe problem is that the code installing the class loader in assumes there is no context class loader but the IBM JVM seems to be setting one up +Tentative classes have been coded +Thats greate but until its released its not really idea when your going to release the version with the fix +There is no failure but the handler doesnt disappear right away +I agree with the observation that the English text for the messageparameter should not be but should be fetched froma message fileIll have a look at this and see what I can do +Leaving local definitions all over that can get out of sync with other KVM hosts or cause conflicts like this is a bad idea +Over time the one they dont care about will become less and less used and we can slowly phase them out +If they just grab any old resource and cast then it may be the cause of these seems to be fixed +If someone wants to work on it please contact me +Please if you figure out a way to reproduce. +Closing as I still cant reproduce it and there is no user feedback anymore +also describes problems when DISTINCT and ORDER BY are combined in a single query + for the to trunk. +I can indeed confirm this +Does everything else workI assume so thus putting this for replicate that uder Mac OSX for last CR installer +And is not converting them back to nested types +Its working fine +The only case where we can see an issue with the global presence of alias would be if we update a table and then join in a single query +GavinI thought Ross would have let me know if there were something wrong with my latest tarball +All Pivot Maven POM files unpack from the Pivot directory +FYI this broke the auxpath +Patch looks goodThanks for the quick patch! +Thanks Chip! I see logging issues maybe thats why start call end call statements are printed + +Should apply against a clean trunk again +Im having a problem with archiva stop on Mac so Ill open this until I can figure it we need to update the following docsquickstart standalone docsadminguide docsadminguide also seems the and stderr logs duplicate information +patch applied to JPOX CVS +Lowering priority since there are workaround +After you install it it adds a new update center with more mevenide modules +was our choice that we duplicated these properties on the model and used for creation of data source +Having all these issues added to this container ticket will also help with Docs QE and Engineering QEIs this guide going to make Post GA? Should it be realigned with the normal release +I have added the egrep script as toolsrelicense toprevent any problems with copyingpasting from here +Here it is +Is it not a protocol error to get a negative list size +Patch based on current trunk +done by ben +Such documentation seems small thing but it stops me from using and is very frustrating +For what I can see you used your own gzip implementation I didnt thought in very detail about it but I think this is not necessary It seems there are classes within the package you didnt need to fulfil the proposed function +I am currently reading the dev guidelines and will provide pull request later today +This issue and its origin can be closed. +field number class +Committed tests with revision +I began working on this tonight and came upon another pro for the sequential numbering of the logsThere may be occasional times when we want to roll the edit log when weve not made any edits into it +Is there any work around for this short of dumping and recreating the entire database +Okay Ive got all tests passingI rigged up a smarted that works for these tests +I can do that when I check it in +bq +is fixed properly thanks a lot! +Can you please provide the complete test case or a description of how to reproduce the exception? Thank you + +o Merging rev from trunkThis fix will still bomb if the contextPath is or base url ends with +Will check back when I get it +Im going to mark this as a duplicate of and add IGMP in with that issueBeen too busy lately got the code written will test and commit within the next few days +The scope is much larger and theres no timeline its a brainstorming issue +Looks useful +The Annotation Editor should then be fixed to visualize this selection correctlyI am happy to work on this fix I believe it is also a bit annoying when you select multiple annotations in the outline view +seems not to have this issue at all its much much faster +bulk close of resolved issues. +Thanks so much +for the patch other than the correction in the javadoc similar to whats been applied for trunk +I think we could peal back some complexity if we treat rich data types as the common case and the case as the edge case +Please see ongoing updates on linked issue +Regards the question for reviewer yes first key would be not a +Patch was applied at +Ive fixed that in the svn +Verified by Sian. +No objections here +The problem is that doesnt solve the original problem just the test case +Any update on thisDid you try rc? +If I dont hear anything Ill plan to commit the patch at the end of the week +Thanks again for doing this after struggling with I know how much work it is! Knowing this problem well I have learned so much from reading your solution +Any ideas +Ive made one for the trunk as the test +BTW Im not sure if you already know about this but there is an ANTLR grammarfile for OQL available for free with a license on the front page of theODMG website that we can use to start this work +because the actual speedup is achieved because you reuse the same Block objectI did an experiment to not use the same Block object with the new shortcircuiting code the performance gain remains the same +addition to upgrade to there should be some code addition to the nhttp transport so that Synapse uses the new improvements +Could you add some testcase scenario +The lookUp method should take care of this at least it does on other datastores +So all issues not directly assigned to an AS release are being closed +Starting up and shutting down for every testcase will just increase the test time +I want to search the text of the transcript and return the timecode so I can display the lecture video from that spot +Closing this JIRA bug out as per Andres comments above. +because mageia recompress man pages in xz dependencies or at least make it more also ensured everything still builds on centosopenSUSEMageia after this patch +Add timeout back +There may be cases where onemay want to unblock the socket but not necessarily kill the running thread memory serves me correctly you must close the socket that a thread isblocked reading on in order to get that thread to continue processing beyond thecall to read +Ive also had occasion to contemplate the limitations of and thought about something like the mentioned above +Somewhat tempted to revert this patch +Well integrate tests for in test case test of as we can use same entities to replay things +has many other options like limiting the number of connections per server using a proxy using various connection factories etc +One idea that Ive entertained is to automatically rsync the local copy of sharelibs with HDFS location on Oozie startup +Added Thread Safety to the list of components affected by this issue +This is all testing with maven +Moved from JBIDE to Seam on the request of Pete and Jay +This has another deeper implication +this looks like a critical problem +Your understanding is correct +I did changed quite a lot so the retrieve is better supportedExisting retrieve configurations are now considered as standalone retrieves +Done +It would be better to invest into upgrading to a newer widget version hoping that writing tests is more comfortable +The reason of this behaviour was that the script has NOT failed when some of the features were not found on any of the updatesites just displayed a message +they should be entirely unrelated at least as far as itself goes +I tested it using the project +Whats the use case for this functionality? ie who is the new user of the added function call +Logged In YES useridI was working around here so I fixed this +The mapreduce use case does make sense but Nathan brings up a good point as well that I didnt think aboutThe sloppy code I was imagining was people using the assumption of a consistent order to avoid hash lookups on field ids in deserialization code +committed to +Steps to test the new EJB Stateful Session Bean Resource type Copy the attached into the deploy dir of a JBAS instance that is being managed by Run discovery f from the Agent on the same machine as the JBAS In the Jopr GUI go to the JBAS Resources Inventory tab and verify that it now has Stateful EJB Resources as childrentest test test and test Test that both metrics and avail are being collected for the Stateful EJB Resources +I missed this one +I tried upgrading to and it doesnt break anything in the specsso for upgrading ant +saw this with with ibm +Since your indexes are not being held in ram there is some other problem +And I tested it by using both Sun JDK and +Here is a try +Yes that is a bug +Thanks +This patch and issue are obsolete. +I think we should just drop this JIRA as. +please has been committed. +jarExtracting systemorgapacheariesjmx systemorgapachearies +OK is supposed to be released tomorrow +Whats the point in having a with only an id +Reported as fixed +The great thing about ROO is that from that point we can take control and make manual adjustments to the code and queries as needed +However I cannot predict whether such a release will appear and whether it will be upwards compatible. +Marking this JIRA as resolved as the upgrade to jboss web is done decompiled and verified that patch applied verified unfortunately no idea +Verified on ER Affects and Fix in Version +Changed fix version +The javadoc needs link for the deprecated classes +If a can be made we can reopen it. +Thanks! +closing all issues that have been in resolved state for more than a month without further comments +Well certainly you could mock out the call to getSunrise and get it to return the required String but you could do that with EMC too +Ill create a separate issue for that +This was because checkboxes are no longer supported +Thanks Ravi +Actually it seems like we can apply the same patch on trunk +The edits log is meg large it is included in the attachments here +Hmm it would have to be a static cache thats not great +class files on Open Type if you are willing to argue more about this issue please provide example projects to reproduce your issue. +Will try to see if this is the same issue +Are you sure this is not the issue that you are seeing +This is probably not to hardThe advantage is that its much easier for a human to reason about small integer than biggish MD hashes +Im going to cc Gabriel to take a look at it it was odd that I had to do it but I dont think its that harmful +Just rebased the patch to latest trunk +Updated the XSD todayFor the schema will only support Source relational models +Hopefully it is easy to review and maintain +Paulex The fix looks fine +Then you could have a thread local registry and simply call clear later +shall we just close it +One of the solutions for these types of problems is to take a RW snapshot of the production dataset and then have the development app run against the RW snapshot +I definitely know that one of the issues is related to an existing bug report +bq +Updating component +Good progress has been made with shared libraries build on AixPower +I added some convenience constructors for I hope this will help. +Pavel yes the only thing i changed was the timeout settings on the ra from sec to secs as sec is not long enough +If no one beats me to it Ill work on creating an updated patch after I finish working on +Please go ahead and make the repo writable +fixed. +It could result in service availability issue +To reproduce start a network server on localhost then run the attached class +Im trying to adapt it into our test suite. +The query timeout test has failed in the last two test cycles on both Solaris x and Windows +the example file through FOP Trunk and I agree with the interpretation +Can you extract the functionality of creating file as a script so that user can invoke it to create and at the same time + fixes this +Committed the change causes test failures in the xmlbeans and jibx modules see Hudson +I have already created a backport that we will be using internally until the next release +resolved r +Okay was very easy to integrate as I told you the number of messages were used for updating I just replaced it with the number of active verticesTestcases pass so I will directly commit this. +Will commit in a few days after I get a chance to look at it more +I afraid we cannot commit such patch into the main repository +This time its much more flexible than before +Discussed severity with MartinRobRupert and agreed to downgrade and move to M +I think I had this situation when trying to access a protected action with no habilitations but with a SSO token from another s why I was suggesting a redirect with a message like but May be I missed something. +If a kernel developer would like to review please do so +Patch looks good +At that point it does not properly attribute the type to be bag if that was the case +The main argument for me is not aesthetic it is practical +Patch for an and associated test +Is this still a problem? I tested on win and win with no fact the test is running and as far as I can see there is nothing disabling it on windows +Tested this on nodes and the issue to do with stale values after the job is complete is addressed +Out of curiosity is there a reason to use volatile for instead of making it an +use waiting to ensure that the tests pass +Can be set to values greater than and the read check interval will be multiplied by this value +Please feel free to this issue at any time +Bruce please provide a patch there are not enough developers doing and way too many users asking +Ive attached some code which could be used to dynamically generate the for each LDAP Authentication Manager +rebased +Correct +Are you seeing the problem in +The problem is here that the attributes are initialized after construction of the Tokenizer before the consumer starts to consume the tokens +Marking as WONTFIX. +Added to committed +Patch checked in +A proper patch file would be better the next time Do you mind if instead I provide a getter only but an additional constructor that would let you specify the classloaderThe reasoning behind that would be to avoid people from change the classloader in the middle of the usage of the engine that could lead to weird if one sets the load at construction time youve got the same ability to customize it but at least we prefernt any weird behaviour in case someone inadvertantly change the classloader at a later do you think? +If this is still an issue feel free to reopen. +Ive realized that you need to import multiple tables in paralel however I would be interested more how youre actually executing Sqoop +This one got fixed for with +Im kinda surprised that nobody else is bothered by overlapping graphics +Applied to branch +not sure if it is a great think the PDF file is really buggy in the policies the metadata property redirect to the catalog xobject of the PDFThats why there is a class cast exceptionHowever it should not fail like that +r added inline sequences for checkcast of a resolved class +Hi Geoffreythe latest SNAPSHOT version has already a interface +I have several questions on the review board +We will also be updating the document for which should solve many of these issues. +I ran the tests under Dmkuttyscenario and Backing it up at Dmkuttyscenariomailbackup +you have to build it from SVN the release will come in the next couple of weeks +is about making online friendly +IMO all selected cars should be enabled by default +closing issues for released versions +These are real servers spread between two racks +Moving openshift out makes this less important +But why do you all this instaneof and isAssignable stuff? Just register impl or instance and it should be enough. +test to catch this problem and a large patch for a test but the idea is to extend it to test also other transitions of states +So what if one of my rules is that I cant use snapshots +This was removed on purpose +What does your jar say +but as i said earlier a kb model description is not small at all +Can you make one for trunk please? Thank you +This is mainly for improving availability since we will have much lower downtimeReplay txnlog conditionallyI think I was trying to mimic the existing code that initialize database conditionally +Jev can you please find the issue and link it here? When I search for this issue key in my email archives I dont see any docs touched and there are no related issues. +In addition it would be a big compatibility break +All tests are green after patchPlease review and commit if reasonable +After closer inspection it turns out that the problem is not with but that fact that the generated class is using and NOT base +bq +Can you provide us with a failing nunit test case or a simple console application that contains both indexes content and fails to search within the indexed content? Or perhaps a small console application that shows us the errorIt still sounds like youre using the default when indexing +But doing something like this is still likely to be useful for helping move code from Java to X +in +i dont see that behaviour on UNIX +Please note that Maven does not work either so Maven is required to deploy correctly +So theres no need to fix the doc before vMoving to TBD +uploaded +Seems like that version of Drools is not supported on Java hence the error message +Committed branch and trunk +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Yes +Passed +committed to both branch and trunk +I simply theThere is still a commit log thread local which holds a strong reference to a jackrabbit class preventing the GC from collecting the +bq +Most probably this happens because of too old installed on RHEL +Moving discussion to +This can be done however the URL should be reject by the JNDI provider if the host componentof the LDAP URL is not the current host and the port component of the URL is not an LDAP or LDAPS portbound to by the current server instance +Lets think of it this way +Of course can be reopened any time later +Closing issue out as part of JIRA cleanup +From the specification it looks like you are right that the persistence context is in a bad state but I think that you should probably mark it as unusable rather than corrupting dataIf you could please explain the relation situation as well +When sending something to the new lists I always get a deliveryfailureDelivery to the following recipient failed permanently usersTechnical details of permanent failure DNS Error Domain name not found +Im now having troubles with things under cygwin +Sound ok +But I wont do this since Xerces being able to process the files without error when theyre sitting on the file system is a pretty good indication that Xerces isnt the root of the problem +Committed revision +If configuration updates are not supported at this point then there should be no gui for it in the console +Web visual tools guides replacing the image for figure done for Web visual tools guide +all xpaths are correct this was just a small UI error it seems though the UI error did eventually cause the xpath to break +Dmitry do you have an use case where previous doesnt apply? All caches are automatically disabled in myfaces in Development stage to allow development without servers restart +The persist the bimap only if it has changed otherwise references the previously written map +Fixed into rev +All the test cases that the bug description says behave as expected +But I pressed Cancel so the runtimes shouldnt have been created +artifact +I have updated with a description of the problem +just to double check did you remove the previous version of report code from lib? since attached one has different version could be that you have both on classpath +When later a reference to the block is needed we can get it from the block setWe can furthur remove the blockMap and activeBlock list by adding two fields to the Block data structures +If so open new issue and Ill commit thisYes beforebq +This patch adds the dialog and pops it up if there are parms in the URL +Hemanth yes agreedBtw great to see you around +for patch +Im still optimistic that we can get rid of these map lookups +Is there anything else needed on the indexing side +Wouldnt it be better to rename the servlet into something like war or web +This is scheduled for fix in GA +all resolved issues for Engine release. +Regards blocker critcal was created before July please review and resolve we are approaching code freeze in days +Ill open a ticket with a general description of the current and my earlier proposal so we can track any interest there +bulk defer of issues to +That makes it much simpler to +Thanks for testing this +I think its a reasonable assumption that this problem had a similar causeThe failing test was removed inMarking the issue as resolved +Tx Omkar +They concurrently start to read from bookkeeper and in the process they try to recover the ledger +Marcus pl do review thisthanks if we dont want people to be able to be able to use basic zones without security groups in then it isnt a blocker +I added a link to this dozer docu from the data format as its related to those +A suggestion from Blake look into using a soft reference instead of a +I suggest to commit your original patch and dont add this support yet at least not in +So no need for a patch from me +If Im not mistaken the old email thread this issue was being discussed on was initiated by me +Verified by Aleksei +I was trying to figure out the root cause behind these occasional launch delays we encounter them once in a while on a highly loaded cluster +Thank you Kristian +upgrade to last svn java provider version done. +Waiting for pull to arq repo +I havent run any other tests yet +jar and should be fixes the classpath with Prasad offline +Maybe when I work on again I can take a deeper look +I can see some fun slogans coming out of that like try and catch up Get running light speed runs +Hi AmilaI am sorry but I am no longer on the project dealing with Sharepoint and donot have access to a environment to do this +Ill have to regen +I ran my project only in single java project mode no any application server involved +Ill clean up a bit and then how to configure the different host +Some more comments In case of errors can you say need to be manually deleted instead of just need to be deleted? Please add tests to verify fileCache and NMPRIVATEDIR deletion verify delete by rebooting NM multiple times when a previous deletion was in patch for addressing review comments +Ive been meaning to try compiling the project from outside Eclipse using the compiler from the latest Groovy release to confirm that the issue is specific to the plugin but I have to assume that it works fine since I havent heard anyone mention this in the larger Groovy community +New version in morning need to adjust a few mocks in +Ive also relaxed the requirement for an authentication +Latest addition is applied +Hmmm what does do for testing? How hard would it be to have a mock? I havent used it before so any suggestions on what to do would be great +According to you its not failing in your test setup +It would be good to hear others here +The point of making the sorted behavior explicit is that it allows us to fulfill the single use case this issue is actually calling out rather than arbitrarily affecting all situations where we iterate over the fields of a struct +So youd start with the assertion error then grep backwards for the key in the debug lines +this is another topic I have been meaning to address on the mailing list +In the facelets and JSF environment property type check have been the original design even from first Ajaxjsf versions for an event property have been could provide unpredictable errors if attribute value changed between requests +Thanks going through an updating all of these +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Thanks again Francois for the patch +It would be nice to fix this properly and allow support for specifying unicode characters uXXXX and supporting standard escapes r n t +I need to check this in real example if it works and then I will prepare the required changes +The standard way to set is to use hostname but in previous versions setting ip address can also work so I think it is better to keep the old behavior +escalate this issue a lot of folks want this to be fixed thanks +Is this intended for the job summary log or for feedback to users? I guess one could consider the shuffle bytes and these times to help tune their job +I see no issues in the code so long as is set +Just this Problem Occurred window information Ive provided +It was established that this issue was not a bug. +Fixed javadocs snippets and code references for S and +the two known current solutions are now in the samplesTo whom it may concern Please test and enhance these and also find othersolutions +Need to review cannot be made panels can be autosized +The two concurrent tasks are firstly to await completed requests and optionally create new requests if the request takes too long and secondly to send the request and accept the response +eu server if your IP is listed as being in the +Is the current text a bad way to say thatThe current text is good +Thank you Eli. +Why is the fix version! It basically already is fixed +However how do you explain todds statement that the original patch has been in production for over a year +Closed. +Moving to. +Committed revision +We dont have to figure out how to +Closing as this is not a Grails this fixed the problem for me too Thanks guys +Original issue is likely a duplicate ofSee also for issue +Since breaks the compilation of MapReduce +Thanks Jeseung +If someone else wants similar functionality for testing we can add in targeted configs to do things like what Karthik is suggesting +It doesnt look stupid +Thanks a lot Simon +admittedly I havent fully thought out if merging from multiple sources would still be validNow that Ive had my coffee no its not +Closing old resolved issues +there is a JIRA issue for this someplace +looks good +Please mark this as resolved. +I tested with rev for and it doesnt affect at all +Although I can see where it wouldve been nice to pop in different lengths of the bootPassword for experimentation I am closing this issue as wont fix because + where is it? The download page for the only shows versions up to and has the old licence +Do folks agree +The issue is most likely due to missing configuration +An Error was thrown instead of an Exception +Please use the forums for this is a feature request please reword the jira title and description to be more meaningful +Attaching and with changes to the async get cursor and get scroll insensitive cursor command topics in addition to the prepare and execute command topics and the identifier syntax topicI was able to test the cursor commands thanks to Ricks commitPlease let me know what additional changes are needed + +We can say so only when we expect a fault +Tracy Claus Cool and thanks again +Not a issue. +I spoke to Vinod and found thats the same reason why he too wants fixed first +Its just performing the cleanup of stale data +Change CHUKWACONFDIR directory for alert +I dont think we need any action on this +We have to change all the classes that use xsdhexBinary to have of deep copy of above value +Yes I think we should apply similar change to closeRegion too +Heres another option Add a shutdown hook manager to Hadoop common which installs a singular shutdown hook +Sorry for the confusion +Patch for Lokahi mysql +One of the writers would eventually fail +Request to merge this fix onto branch +If the decision is to not support Serialization in Lucene then I recommend removing Serializable from all classes in Lucene so that users do not mistakenly expect the search library to behave the way other Java libraries such as ICUJ JDK class libraries Spring etc do +For annotations are needed to instruct Jackson what to serialize and what to ignore +May fix the should fix the problem +Fix merged into the trunk under revision r. +handyande has now applied this +Blame is a strong word and I didnt use the fact remains that there is some code in which causes a problem for AS end users and there is disagreement between the JBTM team and the Hibernate team as to the correct course of action to resolve you please liase with the JBTM team to see what to do nextRegardsRichard +Based on how many valid status can a financial transaction have while also be part of reconciliation you may decide different strategy for checks and return message +So DM can allocate the right amount of memoryBut I think this leads to a real problem because you often dont know the size of an object +Thats true perfect +Use some tool to determine what code is holding onto a lock +The text for use in AS should be updated regardless +this patch fixes this issue +Once this is copied over the Jetty server of the binary distro starts up some reason this file is not packaged with the binary distro but is there in the distro +Does it have a pool? We need it to budgetreuse the use of off heap buffers +bulk cleanup of Jira versioning +committed changes for which bring the cmake install target fairly closely into line with the automake install target +Removed attachment at attchers request +Missed the CVD launch configuration. +I was however a bit perplexed that the creation of the class file still seemed to happen even without the change + Added SQL batching limit for +I will look into this +Must have been a SVN problem + +I recreated the whole scenarion and found that if we define the bundle inthings works in the expected order +Wouldnt it break backward compat onWe can do a solution where always has reported so all is good with wire compat +Fixed part of this issue with revision +Thanks Nitesh Deepak and Rishi for the contribution +Patch commited in the tested! +Noel +A Solr client library could run the distributed searchmerge and return that to its calling app + is a better first choice as it is much more accessible to average users +This is achievable with existing resource loading mechanisms. +Moving these out to please move back if you plan on working them in the next few weeks +patch for HIVE QA to run tests +Our workaround thus far have been to manually sideline the corrupted hfiles check them using hbase hfile and if they arent read properly we set them aside +that one is defining a transformer this is defining another use case for the service activator +on that +Improved menu widget patchReplaced rn with appendWhitespace method callsKept markup formatting for reason stated above +Im almost ready too +Ill shortly provide a redone patch that allows the to work again but unless we have evidence of actual use by other language analyzers I dont think we should be addressing them right nowOK this sounds like the right approach +Yes works fine +Temporarily reopening to update release note info +Verified in jbt build. +character encoding isnt handled properly yetBecause the correct character encoding is determined in the instantiating the cannot be done yet from within the constructorA new Valve is needed doing that which thus must be invoked after the and at least before theThe dependency of the on also poses problems with the integration in Fusion so this dependency must be removed. +While solving Ive put in place check on the maintain session prop so that if thats not used clients calls to oneway methods just wait for the http message to be sent and do not wait for the connection to be closed by the server +Resolving fixed +If desired we could separate these into tests that are really tools which would go into tools and shouldnt rely on junit or other test framework stuff and tests that are really unit tests that go into something like tests +In any case we should increase the number of threads in the example jetty config and document this issue +Yes before creating my bug report I have disabled loggin as it has been described in other bug report with a similar problem +closing as issues are fixed in +that pool clearly indicates when hibernate starts up that it shouldnt be used in production +Hi Simo sorry for delay +This is a strange one indeed +Can you please update this ticket with a working patch +There is now a background segment merging feature available in Lucene +r +Attached patch removed from throws clauses javadocs added ensureOpen calls inside as well. +Can open new issue if this doesnt squash it + +This is a stupid bug +Attaching committed patch for +Thank you nicolas for this contribution tested and Committed revision for the review HansYour commit leads to create two issues complete the documentation on wiki complete the unit testI will dot it +Some of the directory and jar references are only applied to those apps that operate independent of the Geronimo app server +For the children they will be handled by other options like fix and fix +I will wait until we merge everything to master and release have merged into master and tagged on my own machine and am almost ready to push it all up +I use Suns JDK but I will try as well +The change here simply detects that there is a mix of messages with and without dont care values which must correspond to a rolling upgrade so it ignores the corresponding fields and simply verifies that the epoch is greater +Patch v +Not reproduced with and withmysql +try +Currently it locks to get disk stats +But taking it further since the base requirement of a sqoop ETL flow that a jdbc driver is available for the data source maybe a mock JDBC data source would be an even more appropriate way to implement the tests +piggyback +Committed to as well +There are other request handlers that leverage this as wellPerhaps though a solution is to not allow streams to be resolved unless they are truly needed by the request handler? In your example there is no need for the standard search request handler to access any streams and thus that URL shouldnt be hit +Fixed for upcoming release. +What I meant was that the original autor of the patch has to grant permission to the ASF which is done by uploading it into JIRA +Do we also wish to identify the HostnamesIP addresses for this kind of authorization +And yes the cascading failure problem is the real kicker +Fixed on both and closing all issues that have been in resolved state for months or more without any feedback or update +This problem also shows up in +Please email a copy of the CCLA to akimball at gmail +I have changed it to the standard +Moving back to initial fix has been committed on last week hence Id suggest to lower the priority of this tasks since I would still like to investigate the whole chain and see if we can address this problem once for Justin sugestions +thanks david that explains party of itStill need a way to do what you say if it is set to Y +reopening since need to commit to patch +on asking the job tracker for the size of the cluster and replicating based on that sizeI think you mean namenode not job tracker +This is genuine bug but instead of temporary trying to introduce code all over the place in benchmarker I am going to wait with this one untill is fixed which will fix it directly +Actually it seems the attributes are not handled correctly when using JAXB and JWS service tried to set RPC style instead of DOCUMENT but in the generated wsdl i get document +fix openjpa shindig in JIRA a start Unix group shindig created +I can change it back +for +Justin want to have a look before I commit +The BIRT online help is not up to date +Smooks should be modified to use FQN in service file +Premature commit +Moving all issues to the CR +Would you mind sharing the code with me +Compertitors Partners +Marking it as Invalid. +this is multilined would be good enough as far as im concerning but got this is when i opened this issue +We dont want to introduce complexity into the namespace options where we can help itadding support for a custom handler mapping instance exposes though docs need to be better so this ticket will address that +Looks like an easy thing to grab +We generally use JIRA for confirmed bugs to track for fixing themI just noticed the flow name Question But I never used this beforeRegards some format problem in my previous comment please read aswhether you have under the path HADOOPINSTALLATIONHOMEsharehadoophdfslib +Git ID bbfdefaebebacabaa +This patch should now allow any instances that happen to be left in memory due to belonging to a cached Stylesheet to be garbage collected when the process that contained the stylesheet is +right I update the issue description +Committed for in rev +I added both bare and async flags to our wsdl first sample and with it does generate bare async methods +I think is required because Derby is compiled with bootclasspath set to toolsjava +Attached patch +Reopening for review +Just found this comment in the original discussionbq +In other words for each parser foo create a pom that depends on but excludes the dependencies that are not needed by that parser +Very cool! +This SHOULD work however I have not compiled it since I do not yet have a working build environment +Attached fixed version of the patch +Reported issue is resolved. +to be honest i am not sure which one is more suitable i am not a native speaker and could not distinguish those two exactly +Closing. +Acknowledged thanks Ollie. +No updates yet but I should have time to start working on this again in the next couple of days right after I finish some working on converting hadoop RPC to NIO +verified in JBPAPPGACP branch +Sorry. +Applied thank you! +Sorry +Regards we dont need to flag folks +Move it back in if you think differently. +For some reason I though itest is independent project +Thanks for the patch +Test failures in TRUNK is down to . +Fixed as discussed. +Patch for trunk all files together because branch patches do not apply correctly +sofor the platforms +This patch adds beans at right? and are bad names for mbeans +Heres the first step which makes use initCausegetCause instead of its own setget mechanism +Thanks +it seems like the bugs I mentioned above areall fixed +Closing +Why not make it configurableThanksRuchith +Can you please test latest wicket +The easiest would probably be to have all the factories and the class loader in the same bundle +Is there any intent to try and fix this +Is that true? Or is there some way Im not seeing that I can set up the to work for both faces requests and non faces requests +what goes wonky when the binding configs are in there +May be we could go futher and showhide not directly connected entities tables +Closing. +The patch that Darren provided I have tested more extensively but no longer applied to HEAD +Bulk close for +Could you try to open the file wardeploypluginsfindbugs to check if its corrupted on theCan you also attach the full maven logs with debug mode X please +upgrade in my eyes +I hope someone can find clue in it +that seems to have done the trick +I would be happy to move away from the old logo +My suggestion is to make the first write to the empty file fail if there is not enough datanode in another word make the policy check more strictly +Your example shows loading of a MAPPINFILE +Problem has not been evident with later versions of FF +Needed to change a couple details to allow me to fix them a better way in will continue this patch when that one is squared away +The service ranking value that is set on the resolver does not guarantee that will be used if present +Move open issues to CR +No has a single copy of andI made Ant build script to make zip file to release However it has some constraintBuild script is and please read Source code is included to exported by the plugin exporting task which provided by EclipseCould you review the release zip file contents +Applied +Applied the patch at revision Thanks Thilina for the patchThanks regards +Ive downloaded and Im going through the laborious process of integrating this into the product for testing +Moving back to M pretty much have this finished +We need to ship these classes +For the post hook if an operation succeeds at the client specified consistency level a successtrigger fires +Otherwise we risk breaking the behavior for folks who expect the official behavior +And reviewing the logs it is seen that the test cluster leaves Safe Mode successfully so it is unlikely that this patch has anything to do with the failurePlease consider these test failures to be false positives +My humblest apologiesThanks +This will be easier once comes +Pete +Seems reasonable having a cataloguser space divide +You have to install the Central Community feature +There was slight typo in the old code +Ive modified the test case so that new messages are pumped into the route every second so it keeps running will let this run for awhile and see if anything happens +Done +If not then you need some kind of RHS transformation again +Thanks for the quick response! +there are tests for Otis commited on March but those tests dot currently compilerun without the s themselvesOtis do you have these s in your working directory that you perhaps forgot to svn add before committing +But dont think it blocker on RC +Konstantin are you on this latest patch still +Ah i see makes sense +Are you suggesting that it does not work in standalone or when deployed into the application server +However if you create a document and free the builder then the document refers to a object +But without the workaround I think it could be considered a blocker from a point of view +Thanks Ramya! +what I meant was +Committed to +the window is actually half a second in which the service having gone might still be accessedWhat I do not understand is why the Configuration Admin and Event Admin services are bound and why the and services are only registered when the admin services exist +I took thread dumps from both nodes in a cluster +For now we have synchronized start up of the threads but I dont know if that is an adequate solution +It should resolve this kind regardsF +I am looking at backporting this fix to +Sorry Uwe I was really busy today +A patch that addresses Brandons commentsIll address in +Sure +There are still some issues with the patch +Patch to add new just committed this. +I committed the patch +Currently it looks like a datanode shuts down when it encounters a read error which is probably sufficient +Better patch which removes changing the state of a static variableIll commit this in day or so +What is the scale that you expect this bayesian classifier to handle? How many training documents does it needIm doing some benchmarking in these days therefore I should be able to say something about this shortly +Because trunk needs a +To clarify you should build ruby branch to confirm patchlevel and whether the issue is still there and file against as a regressionbug +GUI installer comes with a doc directory filesdirectories are present only in the GUI version in jars were not in other files buildoutputseambuildoutputserverallconfbuildoutputserverallconfpropsbuildoutputserveralldeploybuildoutputserveralldeploybuildoutputserveralldeploybuildoutputserveralldeploymanagementbuildoutputserveralldeploymanagementbuildoutputserverdefaultconfbuildoutputserverdefaultconfpropsbuildoutputserverdefaultdeploybuildoutputserverdefaultdeploybuildoutputserverdefaultdeploybuildoutputserverdefaultdeploybuildoutputserverdefaultdeploybuildoutputserverdefaultdeploymanagementbuildoutputserverdefaultdeploymanagement diff is s the status of this +RonI left the assignee on automatic when I filled the JIRA +This issue is caused by a patch committed on +This update addresses all of phils review comments +Ill commit to trunk tonight give hudson a weekend at it and backport monday if there are no issuesI dont expect any problems though I think this is just as simple as before and we can move forwards to simplify CFS access +Oops wrong file is the right one +ThanksApplied the fix and the test in rev +Patch applied to branch. +Thanks! +This issue has been solved in version +Basic utility code is moved to with delegating to it +And we would allow STARTING to run through uninterrupted but support aborting the initialization +First crack at some Analysis level javadocs +If the resource is deployed then it doesnt work anymore +The idea for now is to have still a thin disk but without metadata +mylyn +The communication of the system counters information can be made efficient by having a customized serialization for them +Hi ValentinThanks for the comments +Control of the default values is isolated in thereset methodThis doesnt fully clean up the socket options but it makes it easy to add or control them by tweaking just and Connectionopen +This works fine for me on current +Its better if you could do it if you dont mindP If you are too busy i could work on itThanks Rainer +This was never meant to work at the top level +It seemed not a standard SQL and also could be confusional especially with joinunions +Adding patch to partition into discrete ranges per vnode +Attachment has been added with description Replacement that avoids touching raster in as previous attachment but as a single complete file instead ofa patch file +attached is a patch with some suggestions from Uwe use reflection for the compiler instead of a tunnel remove a unused variable in the patch missed file additions +The Reference Manual needs to be updatedfilename rrefdmdfns title functionality not supportedRemove section about methods get and get not being supportedAlso what is get? A method a class a? The section that mentions it needs to be qualified +When we start seeing these messages there is something up w our commit logs or with hdfsFinally prints length of edit log when running split +Changed the fix for which did not cover all possible cases +I see it looks like all the bases are already covered then +Confirm fixed on master +Ashishwhy is virtualcolumn class in the serde packagewill put it to package +Please see attached patch +Hi Joep Thanks for reviewing the patch and providing feedback +Test Case attached +Instead it happens after some time +It is way more flexible and should be really easy to a blank value in the property file so that the message doesnt show should only be a temporary solution while waiting for a real fix +Ill try to give this a look as soon as possible during next week +A mass upgrade like the above patch risks introducing unintended errors to code +Any answer for Sebbs question +. +a +is there any update on this topic +Daryn thanks for the explanation closing as invalid +lgtm +Closing old resolved issues +People reading files with Windows paths should appreciate +Yeah ideally they should be a different data type +FYI tested the same scenario using CFIMAP exact same results +If its not then is there any specific reason to default to false +Hi BenIt was Nick who gave me the patch as I dont have access to his machine right now +Duplicated +Tests passed on my local machine +Ok I get it +This is actually a docs issue that gets bundled in the installer +GWT tried getting around people having to implement Serializable by providing an interface with fewer promises isSerizable +Still present in +I am seeing the same issue as Nathan Boy commented on in May +I tried to make the ODBC queries compliant with the ODBC API as I read it and the current queries seem to satisfy the need +Thatd be fineThere is also the mechanism whereby if count of some configurable amount we start flushing memstore that holds oldest edits +Word of caution If you dont have then you are testing th of what is intendedWhen you have a moment please explain this a little more + issue is invalid +codec format documenation +It was released with the fix in +You indicated in the issue description that you may want to add some WARN logging +Users should not have to update and recompile code to make this happenIts been over three years at this point +Closing. +Will need to check it out +Just venting but could the methods be any more poorly named? For getting get is the user API and get is the internal API +Patch looks good One thing I believe Jody changed the interfaces implemented by Not only on trunk so on we should be good in theory even without the with DB appears to not be an issue on +There is no difference +rar not so much +Added some internal error code mappings not considered previously +This is what Im proposing for the proxy UI +Implemented the review comments +This depends on what else you want to do but any phase after resolve visitor has done its work and before the bytecode is generated should be fine +What browser? what OS? Do you have any sort of browser developer tool extensions installed so you could check for javascript errors? what about on the solr server side? what do the solr logs say when you load this pages? anymore suggests this worked for you at one time what changed? did it work for you in an earlier snapshot of the x branch and then not work in or did it at one point work for you in and then stop working after some time passed +Dhruba I understand its a nice thing to have and that it can be used in many fashionable ways and that many file systems have it +By the way I used the following environmentLinux Fedora Chrome betaJava Hot Spot tried this with chrome on my ubuntu but still could not recreate the problem +Committed to trunk with revision +Uploaded patch fixing typos eol whitespace and long lines +is for trunk is a backport for +Fixed in CVS +There is no need for JDK but as far as I know there is no and is set to JSE +I moved the duplicated class to commonlucene in the core +Its a dependency issue with the staxx library +git apply error patch failederror patch does not applyCan you rebase the patch +Implement support for system view XML import in Rev +Sure but this is where is gets a little more complicated since after receiving the error the init script needs to check that the pid is really gone and if not get more forcefulwas about to implement this when a problem came to mind +Closing resolvedwont fix issues. +This is reasonable but I think the gains will be zeroed out in trunk by since we keepalive a single TCP connection only one thread will be created when the connection is first acceptedYou can run trunk with HDFS trunk using I was able to run a node cluster last week for some benchmarks +The ejb tests build their cluster config off the cluster config created in the main AS testsuite +Thanks Sbastien for this patch +Uma either a new jira or changing it in this one is OK since the changes are quite minor +This tar ball contains the jar files for +Just notice that +We will delete the batch file +It would also be a good idea to update the javadoc of to mention the use case you are enabling +Add files from to configure jUDDI for +The serialization will switch to using Avro binary +But at this point Id be good with either approach I want to get this going +but this sometimes lead to independence problems as it is the case with this issue +We have just one shell script in the Java part of Avro +The tests I wrote show this working perfectly fine +If these look good Ill upload a Trinidad patch and check everything in +Then my client will run antthe recive client will runNow you run server +Changes PRETTYFUNCTION to func which is more portable +LOG allows changing the logj levels and USER grants the ability to adddelete users +All subsequent calls will use the version stored in the variable +We was not compliant before was fixed but the fix for didnt properly handle the caching +i absolutely need a way to tell the difference between a correctly spelled word with suggestions and an incorrectly spelled word that has correctionssuggestions +I agree seems very much like my issue except that I experienced that messages were lost not duplicated +At least it should display why the bundle is in GRACEPERIOD +What Andrea suggests seems like the way to go for me +Its not that simple +Im sure I came up with a workaround at the time just to move forward. +resolution status to Partially Compleated +It might be also a good idea to do the same with OPENNLPHOME +The fact is though that there are a significant number of people running who would like to use Pig and supporting them out of the box seems worth it +How can you create the Return Invoice if when you try to Accept Return you get the big ugly error messageConfused s not a workaround +Youre right my apologies +Does anyone know do you need jar files vs +Hi ChristianYeah If camel team agree to upgrade to then we can use SMX wrapped bundle directlyIf theres no other objection Ill append a patch which upgrade to also +This is used by the failover logic only a user does not need to be aware of them when creating applications +Heres an updated patch that includes some testsIll commit this soon unless there are objections +I designed a very simple overlay that will go on top of groovy projects +I agree with Lewis wrt closing this issue as wont fix. +Yes I get Page Not Found +patch looks good +Oh one other thing removed and support +I need to provide a patch for that version +I understand your point about extending the output source so that you can span rollbacks and commits across files +Verified the removal of duplicate record on Build. +it is not getting recorded with the correct zxid +I applied Seans patch to the code which had optimization fo distance calculationWhen I applied this patch to the trunk it took hr mins to execute +Added the Testcase to the +Only customized configuration for nodes is supported at this point +Linking issues +a scrollbar doesnt solve the problem IMO +I have a fix for the tof behavior which just defaults to Infinity or zero if the exponent is outside the representable float exponents +I havent looked at the code again today but I believe that that is what gets thrown if the String cannot be converted to a number for example if it is any other group of lettersfix for this issue +I have also committed to branch +yeschecking openssl presence +That breaks Bikas intent of trying to make this a generic scheduler though +Im going to now start working on converting over a bunch of our tests to use this stuff as part of another issue about speeding up unit tests +proposal for delegating resource serialization and making more +to the local disk and then editing the file in the Apache Directory Studioconfiguration directory so that it point to the local disk instead of Now its working again. +We need to test at least for possibly for +So I changed the component to be results of boolean and boolean +I can make command execution in the Karaf launcher work however +i think it is just a matter of running mvn deploy with the right flags right? i was thinking we would run it right after do do the release +Or maybe we should check the flag and only specify the workspace if global services are disabled +Itd be nice with a new feature if you could run on one or more of the other databases though +I think the javadoc is correct because items are recognized based on equals checks but Ill look into making this configurable +I dont believe theres any work really being done on AS anymore +Initial triage for release suggesting postponing to move back to if you disagree +This requires changes to the unit tests +BabakIn due time you will be able to make the commit log changes yourselfIts not really that important and just risk something odd up with svngit syncing and whatnotSo I would just leave it as is +Looked through all +Scratch the need for get thats JDK s get as Tomasz points out and I utterly missSo what we need is getget +v shows what the CFSCFM reload approach would probably look like for +I somehow doubt that messing with the classloader will really make things better +Please check that the patch is applied as expected +I will change the patch +Attaching a preliminary patch for review +Thanks +Assigning to ALR as he did the Bootstrap. +I renamed the parameter to fail so it could be and reused amongst the other mojos in future too +The bean is and is storing that application context in a member variable +Hi The fix is fine Ive committed at r please verify +Verified on. +The same assign call is used for user regions as well +We can doc in the wiki when we do the stuff for addingremoving disks in t the cookie related to management of bookies as well? If so it should go to the admin guide page no +Oh crud +is a legitimate bug this is required to support tracking and trending code coverage from the CI build server in the light of issues such as +Take a look at the updated patch against the current trunk that Ive attached +Attached patch fixes the issue reported in the bug +Its not a regression +Is this still a problem? +Jacopo AnilHere is the patch for this issueAshish thanks for patch +LGTM +Aaron all the comments are addressed can u review once Aaron for a thorough review resubmitting the is Apr javadoc +Didnt dig into it may be related to my environment +The problem Im having is that the Velocity templates read from the filesystem are not read in the encoding in which they are written +Ok so our npe check is enough for GA or we need some change to jstpublisher +No such implementation yet exists so getTable essentially is a placeholder +Patch produced from latest from looking forward to your colleagues patch +was just about to submit my own patch +fixed in resolved issues. +Ill take it +dstat command added on trunk +The problem comes from a new module is added to the project +I once looked at the Tomcat code in the stack trace and it looks like they are doing the wrong thing +I am still unable to trigger the condition on windows so moving out to next release +to trunk in rCan someone verify that this makes useful artifacts before I resolve the issue +The maximal value is computed by max heap size divided by huge page size +The import is in the header of your bundle +test failures are not related to the patch +Pat Any thoughts on how it would be to port from Forrest to Maven site generation +bulk fixing the version info for and all affected issues have in is no indication that anyone is actively working on this issue so removing from the fixVersion. +It is a file that makes it easier to build projects written in +and fixes findbugs the patch +Cannot reproduce this +Chris look good to me! It makes sense since this whole unit test is targeting Linux as the name suggests +save lines to create new class files making it what their scope is? bad trade. +Can you do a svn up on this file and retake the patch +Not +Have you made any progress on this +targetprotoorgapachehadoopyarnproto +This seems to to be nothing more than some optimalization in the Maven core but unfortunately bad for using in plugins. +Sure I can solve the issue yesterday I notice that too +cmpilato will be the one to handle that I believe +Also removed metadata from as this will be handled by a new instead +With patch v passes +Is this something that can also be fixed in the tree +I reviewed IO reactor classes for possible memory and was unable to find anything that might lead to memory leaks particularly when opening and closing lots of connections +This code passed on Harmony before and I can not say that this is regression testing +Should the null check in the get method by synchronized? Or does it matter +I believe this is a consequence of +Update site navigation to use maven site +An improvement of the revised documentation of issues are obsolete since the version of the plugin +Patch against adds a test checks for the text and the nested that in theres a hack to get a private client and theres a proper API for this in I am going to propose closing this as a WONTFIX +adapted the tests +message should appear on the least thats what happens with some other programs on Mac fact that was the issue I was working on at the time I created a trivial program to debug my changes to for Mac OS +Read description. +Ran a full unit test and verified that theres no regression +After the scrub I still cannot read the files with sstablejson +This is how it looks like is an example of something similar of what I want this is from Visual Studio +I have the same issue on page with components in IE +Please see +Closing out the issus for the release. + +I cant reproduce any of the failure modes above with these patches applied not even the no viable alternative one so they must helpIn reviewing the patches though I dont think I understand how helps anything +If you want to test it you currently have to build the plugins from their SVN source +For other jar taglibs loaded +Do you think you could fix the issue and provide a patch for it +We do check during initialization itself that the path represents a directory right? I didnt notice the IO error part in the Javadoc +Thanks! +Scott could you please point me to the section that you modified +We could even write a script to do this for themIf we leave now with a +Moving this out to please move back to if this will be worked on soon +when I get positive feedback from a PMC member. +Here is a patch and for the version of SUR +What if I need possible conditions +All MINA modules are now bundles thanks to the bundle plugin +That would be a good initial goal of an expansion +Im surprised this bug is never addressed whilst so many other ones are +It will now match only on method name meaning you can add as many parameters as you like +Add the configurable attribute +I dont but Id probably do the rip work +NET client +The number we set up for the queue at is configured as +That is done by otherwise why then do we need? I dont get it +Hmm I think the order of Tokenizer in response is unconsidered +I tried this again on r and the issue seems to be fixed already +There is still one issue with the openjpa enhancer we need to fix +Adrian Im doing triage on things that break when I try upgrading the tx mgr +I took a look at the spec issue and it is not completely the same problem but they are definitely relatedWell see what JSF brings +A independent of demonstrating Pig bug will be highly appreciated +Thanks issue has fix version and is marked with either Release note needed or Existing application impact but does not have a attached to it +Now that work is well underway with AS all previous community releases are +HiI think maybe there is a misunderstanding of idea behind the +This wayonly one merge can happen at a time even when the application is usingmultiple threads +We will clean out all relations between Ftp and File related sources +All issues has to be filed against from now +I am not happy with that from a workflow point of view but it does accomplish what is needed. +Username marcohu +Is anyone else seeing this? There were changes to in but none that should cause this +Changed fix version to match the version on the branch. +We can add one new method in and add also one new method in to implement dont read the meta file that can keep protocol compatibility in +Port to branch Adding applicationsAdding applicationsAdding applicationsAdding applicationsAdding applicationssrcSending assembliesSending assembliesSending assembliessrcvarconfigSending assembliesSending assembliessrcvarconfigAdding configsAdding configsAdding configsAdding configsAdding configsAdding configsAdding configssrcAdding configsAdding configsAdding configsAdding configsAdding configsAdding configsAdding configssrcAdding configsAdding configsAdding configsAdding configsAdding configsAdding configsAdding configssrcTransmitting file dataCommitted revision . +compaction is optional meaning if it fails no data is lost so youshould probably be fineOlder versions of the code did not write out time tracker data andthat is why your older files were giving you npes +A separate boolean flag would just be redundant hence I prefer to consider null as an explicitly supported value for the exception attribute +Closing resolved issues. +The insertion of the extra should have forced a renumbering of the result set number in the column reference +What columns would you add the unique constraint for? There is already a primary key constraint on the JOBINSTANCEID +This is only a blocking version and doesnt work forThe tutorialjava is also updated to show a working example +Moving it interface changed +Thanks for the quick action +JBESBGAFP +The correct answer is to not visit JSP pages directly which is a best practice anyway. +Hmmm? Maybe this is related to the fact that the domain expired at about that time and people now basically point their XML documents to the at? If thats the case Is suggest some serious user education through the mailing lists etc +rules can now be defined for interfaces +in Error On Page still exists +is not the config for making the job fail +Is there a patch for this +Also still wary of many threads synchronizing on a single important master object afraid that someone in production is going to find a codepath that deadlocks and that it makes the master less live +Reverting the retry number to as set that to on RM side +Its still a pretty small amount of data and I cant imagine why the call stack would get very big either +Fixed with svn revision +Please provide a simple test case +to the latest to branch thx for the reviews Eli. +Looks like we just forgot to update OIVs loadable image versions in +If you do not do it now there is an opportunity to comment again when federation branch gets merged to the trunkAll comments will be considered and new jiras will be created to address the required changes. +Thanks +Suggested patch fixes this issueWith this patch the whole fat lock table is represented as an array of fat lock tables +I strongly believe this is caused by the Maven Facet instance being persisted in memory and not cleaned up between different project knew this was gonna bite us one day +Shouldnt sources be optional for those who want them? If we include them IN the TP sitezip JBDS installer will get bigger too +The js was updated +Judging by the comments in the linked issue this has been fixed already +When the testsuite is configured to use datasources from then it does not display on HP any behavior different to other systems. +thanks shravan +ToddOk +OK I did not know about thanksHonestly this seems like a pretty dangerous option particularly since the default seems to be false +In shell do thishbase disable TABLENAMEThen run the script from the +Also of course usually you will want to secure the server we have similar use cases where there is no target bean name defined for the. +An evenbetter less blocking approach would iteratively merge the changes seen in external updates to the local statesThanks a lot to Rafa Kwiecie for reporting this bug and providing the test classesFixed in revision +Now that I use the new convention it works as expected +Adding trivial patchThanksMayank +Replaced the previous with a new one that has a simpler way to expose the encoding and decoding methods +Please try from SVN and closing stale resolved issues +This leak can be very considerable as the conversation may keep a session that can contain many objects in its cache from the database queries +Andy mentioned this is critical for his app +HiWhat is the of a with TLS? If you are using TLS then why not just use a policyColm +I didnt took the time to test it under Windows. +kafka +The update seems to have fixed the problem on my test Eclipse environment +Path has became staled after mavenization +This is key for making startup faster for big clusters +I actually agree w Mark +functions are static methods and not +I am planning to get back to this next Monday but if you need it before let me know and Ill prioritize accordingly dont want to jeopardize releaseThanks! Charles +Making critical +I checked out trunk and it passed then applied this patch and it failed +One liner. +Should be easy to fix in +Please discuss this issue on the mailing list +isnt gsf more or less a replacement for gs as far as gcc is concerned +This may not be a huge issue but the case Im thinking of us is someone inadvertently messing up a relationship that the modeler created via the sync button +So I guess the release profile makes something differentThe tools JAR is thus not needed at all as a dependency to use Camel so what we want is to get rid of tools JAR in the file +this occurs when the directories dont exist the latest version creates directories that dont exist +So your hbm schema will be different from the actual db schema +There are no visible controls which can contain accessKey attribute according to HTML spec +I am going to clean up the code in before adding storage type to +A small correction The test will fail if one or more files have not been deleted after four attempts +The ABS case arises in Derby because Derby treats ABS as a keyword +Do we want to add this +The fact that you can serve pages via HTTPS says nothing about whether or not you have SSL configured properly for Java useTo see what I mean write a servlet that does what does and debug it + +Could you fix it + +to handle the possibility that the first namenode isnt the currently active oneI think thats what Haohuis comment Putting the work of automatically switching to means here +I get this was flagged as Wont Fix but I feel thats a mistake given Sonars goals and the Enforcer plugins goals +bq +I have committed this. +Fixed in r +Committed as rev Modified previous patch to accept either upper or lower case in thread name patterns and to add x to char buffer before aprsnprintf instead transcoding then appending. +Seems escaping is completely wrong double escaping is applied for content inside CDATA block but there may be no any escaping for itExport is done by some third party plugin and I have no control on it I cant change it +Robert do you happen to have opened issues for the issues with the filter themselves? I could not find any in Lucene nor in Solrs jira +James the overall design of rv is what we will be going for +Ill post further information as I get it +Different OS projects uses different versions of +If so that wouldnt be the first bug in this area sounds like a good candidate culprit +Another Con for is that is a version possibly irritating some users +Couldnt find style and styleclass specified in the html rendered +I think option is better because it would prevent future issues like this +msgpack requires lgpl software see of the libraries distinguished themselves as being a particularly crappy choice for serialization +This feature is available right now by using one of the shade proguard or jarjar maven plugins to organizeobfuscate an interim jar which is then turned into a bundle by the bundle plugin. +I have discussed this with Ashish we didnt find any issue in this approach +is this one archiver dependent or is it something to fix in the plugins +Closing myself since there is no way to assign the bug to Rebecca W. +Lets get this one in too +fine with me +therefore it is not allowed that a role has more than one User User has a collection of Roles +The code in was almost a from the code in except the code to add in the from the current JVM was deletedIf you want me to add in some more tests to verify this I am happy to but I think the patch is valid as is +yeah +I was able to get remoting to work properly with your given. +I just fixed the problem in which should appear on Maven Central within the hourPlease try it and let me know +Although its not perfect it is the best we can do right now. +The patch command could not apply the patchMostafa your patch fails to apply +Ignore the attached patch +This process then becomes a zombie process +Would be a good addition toI can offer to load this on our test cluster and run a large PE workload on it +Committed to trunk +I am not able to find in the current trunk version +If you could test with it Id be much obliged +ThanksBest regards by Andrew. +For instance Hitachi returns RMFAIL when it should be RETRY +marking as. +I have found the underlying issue +Hey Im just the messenger Kay Kay is the nd policeman +So Im rescheduling this for JB and assign the remaining subtask to Elias +Completely right +Need more time to fix this. + +Tom if you think it can get into go ahead +This optimization makes it easy to implement datanode side fix +The minimum flush unit is per region not at individual memstore level +Thanks ArvindMike for the Patch +newer issue has better description so keeping it +Closing bug as no feedback was received +There is a PRETTYHTML option and its only partway there +Attaching a patch with more bug fixes +Patch looks good to me +Thanks Mayank! + +Should we commit it to as well +If so please dont +This is also happenning if I and pick Resolve while multiple Ivy containers are selected +The specificconflict varies from run to run depending on the timing but there arealways conflicts +If and M is really big the amount of data kept for doing LIMIT can be ridiculously too much +I agree but I thought the fix would easily make it to +bq +Then Ill commit the changes +Unit test for cruisecontrol results +Thanks Tim +wrapper can support a more conservative set of restrictions and naming conventions based on what we think names should look like + on the patch +When I changed the Rules and +So for now its just displaying the image names +Attached patch contains the change +The workaround continues to work okay +reopening for possible backport +SebastianI see no patch here +Should be resolved by an upgrade of JCR to or above. +Considering this issue fixed. +What do you think? are both right is there some type of patch we could introduce to toggle for both use cases? Perhaps in Logging do some parsing that would output multi line when some default configuration is overridden? Or better yet we have a tool or known command using awk in it to to use when tailing the log? +You didnt answer the question however +MikhailYes the patch is duplicate +Hi Chris Im reopening it because of the regressions +Use container tag like fsubview or whatever + is a class so needs to be proxied via CGLIB +Hi RogelioIt seems to me that requests in addition to exposing the finders with lists in the that Roo also generates a GWT user interface to access the finders +I will double verify though. +The fact that the test passes of the time says a lot even before if it didnt run at all we would just introduce more bugs +Enabling this is one reason why we put region information into done as part of master rewrite +Could you make the change for me? in plugin version +Colin please read my suggestion and my analysis of the difference before commentingThe simplified API absolutely provides a means to releasing the and yet it is lines long instead of +is a private method +Runtime may result +They both implement correct behavior in a more general fashion +do we want to support both? its making the code a bit ugly with all the ifdefs for a platform we really dont support +Just commited the patchThank you Mike. +batch transition resolvedfixed bugs to closedfixed +It could be difficult to implement pruning at this level for all algorithms +yes thats what i thought id need to extract the war change and redeploy it +I just tried the latest snapshot and it still exhibits the same behaviorFYI I have downloaded all the sources and removed what I thought was the offending line in but it still fails +It has an executeReport +Wait for changes to propagate +Thank you Alexey +Looks okay to me. +rather than? as requested by the original raiser but it does the same thing. +Ive just compiled a cvs HEAD and it work fine! Thanks for the attention! +the Thrift and Protocol Buffer communities are larger than Avros +Change And Or to pass out an unmodifiable list Update implementations to make a copy modify and then return a modified filterQ Did you look at the existing duplication filter visitor +Triaged for +Community involvement and contributions are always appreciated +Patch applied +However I would refactor it a bit to regain readability of the code +Im not sure I understand the problem +Please reopen if I am wrong. +Best regards class that decorates the class +Can you test it? +Nothing seems to be changed in terms of options parsing of fsck +The configure script will need to be regenerated using autoconfDiff was generated using svn diffChanges were made to README and Doc for flex detection has now been committed including documentation +Small patch to fix javadoc warnings introduced into Y branch +sample of sortingpaginating of have this far as I can tell none of the Myfaces examples have ever listed in the file +The error Im seeing might actually be +Thanks for reporting +It also moves two header files from sharedcommon into moduleslunisrcmainnativeincludeshared +the logging redundant RamYes I can remove this +Test cases are still welcome +RameshIn the handling intermediate converts are being removed is there a reason for this? In general they are needed especially if the conversion is lossy or in the case of boolean knows to convert between numeric and string values +Different methods belong to different applications thus their messages should come from different resource bundles +Only one subtask left +the same test may or may not pass on random runsStill looking into d love to be able to help debug code if you have any available to post +Changed wsm so that it prepends beehive to its jars at revision beehive jars are now renamed and the controlhaus controls have been updated accordingly. +Marking patch out of +Fixed in CVS head testsuite passes at ! Fixed in branch testsuite also passed with +Thanks Krista and Carlin +I just changed the impl for MIME type declared on the reader instance MIME type declared for the reader component Ask the Reader for a MIME type +Agreed burying exceptions like this could be annoying +We need a feature for this so you can easily install it in +Hey Juraj could you proceed with the readme updateAll updates seems valid but clickable TOC it will be useful in rendered but it wont be useful in plain text version +A hudson run would show a blue dot despite test due to inactivity +I am OK with these small changes to be fixed upon committing +I think open source is using one memstore flusherMultiple memstore flush threads was added pretty recently +Looks good +Documentation patch has been appliedThanks again for the contributionMaarten +The problem turned out to be not resolved completely another bug with the same symptoms was found. +I agree that making the prefix unique will make it easier for applications that process the PDF to extract or font resources when merging multiple PDF Files +Service Create Manage the services provided to customer by the organization + +patch or +Grab item out of cache map +it can be set defualt true +Anyway I ran your attached test and it ran successfully +May be out of date with all the changes for CR +Hmm so this one still seems to be dragging on +We need to turn on the openejb wrapping +bq +Done for good this time I hope + +Thx again for all the hard work mate +We added an instanceof check if Copy all descriptions option is checked +ec extension in the output folderIve already forked to implement such a feature and its working pretty well +This shouldnt be that hard +Can some one commit this one it has already been discussed at +thanks flavio. +Hints +I cant reproduce this +Mahadev Thanks for the review +We are allowed to redistribute the WC schemas +Please let me know +Reopening to attach QS QS modifications +patch looks good +Thanks +Hibernate supports this functionality using the xml descriptor but not using annotations +I have checked it with beta and it was for verifying +This is a problem for us as we do not allow remote SQL connections +If possible please verify the issue is resolved. +think its worth remembering that if we want to print this logo or make it smaller then having fewer dots and a smaller angle is d looks like the best option at the moment to me +did not know how to patch direct calls in register form so code patching did not work at allthere are three issues returning target address type into Int to cheat constraint resolver +Besides the hanging the adjacency file has too high weights so when calculating the integer have an overflowIll generate a new one +Sorry Im late to the party on this one +Its a task for the cache provider dont think it affects ISPN itself. +Attached screenshot from MAT analysis +fixed for +I will try out with latest source build and once gets release I will use all resolved tickets from or older +Im saying its a separate issue because its a separate issue +The unittests in do not leave empty args and cant quite figure thatout just yet. +That used to work fine for me but now it doesnt anymore the result files have inconsistent line endings +The issue doesnt happen when using RANK BY so another workaround is using g RANK e BY a +batch transition resolvedfixed bugs to closedfixed +is the screen shot of the test application running on RI +A quick overview suggests that extending using the class Multiclass from the jni library might lead to results +I would rather not commit this to unless is also committedI dont understand the rush to put this in +Steve can you take a look at this +But I thought it was worth mentioning it. +The problem only occurs in Eclipse and its probably related to me +This is my incomplete attempts at implementing this +Hi SergioSorry for the delay getting back to you +I believe it has been +I will repost to see what the unit tests think +After that it will be easier for multiple people to work in parallel across the various interfaces and implementation details +I dont believe full compliance with C is required or desirable +Attaching new design document to cover comments from Dhruba and Konstantin +It passed all the tests +Andre that sounds good +Not a priority for +Usually the number of jobs in one batch is very small so checking exception at the end looks acceptableIf you specify stoponfailure in the command line option then Pig will stop after the failed batch +I dug the UT up from an old branch. +done +Yup bug is on trunk as well +Its still happening +These blockers and critical issues are resolved but not verified +If you want you can check out the attached JIRA and consider a backport +it was application bug error in for +Added functionality in that runs whenever a project is opened +Unfortunately they are blocker for building hadoop +Reopening as we need to remove the workaround from the assembly before we release the to Axis issues as this is an Axis because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I dont remember what I read or what I might have tried in the past that might have made me jump to the wrong conclusion but Im glad its straightened out now +They must be different +what are the contents of both your validation files? Just putting a validation file in a package isnt enough to cause this error I find it more likely theres an error in one or both of your files +I committed this +Thanks Dhruba. +pull request for JSON released + +Yes I am with Hiram I think when the broker startup then its services such as the datastores and whatnot should ensure their JMX stats are consistent +Hi Aaron thanks for the comments +If you have any ideas for rules around this feature Id be interested to hear them +batch transition resolvedfixed bugs to closedfixed +Indentation changed +Please do not combine or more patches into one +TedThe existing emphasizes behavior when bulk imported have the wrong region boundaries and require split +Is there any way of observing the work by being able to get your materials and try things out +I think improving the timing for launching reducers is not just for small jobsIn the case of for larger jobs with mappers the mappers needs several batches to be fully scheduledIn this case if we launch the reducer when mapper finished those reducers will just be idlingHere is theIf we launch the reducer too late we lose the parallel execution for the mapper execution and reducer shufflingBut if we launch the reducer too early we waste the reducer slots because they have to wait the mappers to finishThe optimal case for this is that we launch the reducers as late as possible while the reducer shuffling phase finishes right after the last mapper finishedThe goal is to somehow estimate the mapper finish time based on the information we have and launch the reducers at the right momentI think this decision should depend on because different scheduling policy affects the mapper finish timeThoughts? +I will update to it when its available +Made few more fixes while testing for + +As I said in private email yesterday trunk also compiles fine for me +Will continue to but not expecting to see anything +I tried with the most recent nightly buildof +I should be able to proceed with the work now +Im trying to implement this improvement in small incremental steps +Thanks Ahmed +To build with an alternate hadoop use one of these three options mapr cdh hdpmvn install Pmaprmapr will build with version cdh will build with version hdp will build with version by default will build with version +Seems like a good idea to me +Unfortinutaly wedont have enough resources to start the development right now +moving to on. + +Thanks for the patchesCommitted to all major branches +I committed +Ah! I didnt realize this was fixed in already and is being backported to +Ive added a check in to make sure the POA being destroyed is not null +hope it will help for the with JBT H . +A simple git rebase fixed the patch +I should have a new release ready in a few the new release ready +Rebased onto this weeks activities +Are you make any development for this issue +For Rafis consideration +Thanks +stack solves synchronous calling +I prepared a patch against the branch that solves this problem for our environment namely H and SQL Server +apache commons you caught me in time Simon It all seems to work and I just committed it +That mean that this plugin is using the same version of that component as the dependency plugin new have been uploaded to the SNAPSHOT repository +Can you check this again? The error appears to be happening in compiled code which has matured tremendously in the past several weeks +supplies fields to its child entities and that has been enough for its users. +This issue will go after we merge blacktie and narayana repos in for pointing it out +I know the overhead is likely small but would like to avoid if possible +I actually did try that last night and it worked I misunderstood how to use singleton with the +Sorry too fast to close it I have reopened itI have tested on Safari and Firefox with Tomcat and clientAuth to want and the problem still existing. +Linking to the original issue that caused the switch to the http wagon +I dont think we intend to support marshalling Java objects +Closing issues resolved in released versions +I created vertices with edges each it never finishes with two threadsIm debugging right now +Thats why I want to find some other mechanism to manage the reference counting problem +This patch allow you to add handlers to the list of handlers and delay the calculation of the handlerChain to actually use these handlers +Im experiencing the same problem +jse file +Should I have the user enter it in as region zone a or enter availabilityzone and split it myself +But yes the way TCP is supposed to work is that it takes the minimum of the size the server or client can support +has been marked as a duplicate of this bug +And since its an internal change we can always make that change in the future if somebody asks +Make very clear in the description of those bundles that they are custom versions and as much info as should use a classifier to denote that they are custom builds +I was able to successfully build the gem last night on netty ill setup the latest desktop version and attach the packages list and commands used when i have it can confirm that Natty works fine + has been fixed in the SVN +If anyone else face the same issue with they can consider upgrading to. +Unassigning myself so a committer can apply it +I am attaching the NM process thread dump for this issue + +Agree +This solves again the risk of defining it twice and thus being able to define twice which would trigger the Liskov BSODThere is one remaining problem In case of parallel interfaces defining the same method the return value can host constraints which means we could define Valid on both interfaces and thus have parallel rules and contradictions +Good point +Do I change or? Because I have seen that there is for download but repository link is for only +Id like to see this backported to as well +So theres no way to make that work +hm doesnt that refers to metadata like websrcmainresourcesschema +Quick update I havent had time to progress these issues because Ive been working at a customer site but I hope to devote some cycles when Im back in Malaysia next week +We need to fix it and its def a blocker +You can submit a patch and if it goes in we can add your name +Reopening because this seems like an issue in implementation of +Just spent sometime reviewing the patch +Patch for the release that addresses this issue +The forms where a shipping group is specified should be kept +thanks +Assumed this is fixed since is being compiled and included. +Lev please recompile the classes in these with Java +And why only in there and not the other see that there is also occurrences of it in the Windows batch scripts but that will have to be someone elses problem +In a discussion with a couple of the people very familiarwith Xalan code they asserted that Xalan relies on the JDK to handle and be aware of its platform +jstack for Region should fix this in if this still exists +I enhanced this a bit in revision +I am closing this issue +I committed the addendumtest appears to be broken after this patch +I just committed this. +This issue is one of the steps towards introducing append +The last run on ubuntu finished in almost of the time of the previous four runs +It would be very nice to get a fix for nokogiri ffi support for + +JochenGuillaume The regression is not new and I have a workaround so maybe it can be downgraded from blocker +The rd party jars and jars are consolidated in there to allow one to download and fire up the server using a single Jar +Thanks +Im for the more conservative approach +Im not sure how we could do this from the client +Ill either disable this test or ignore it for M as a known failure which depends on whether you and Jason decide to support this in the future +Added to the list of affected versions and rescheduled for +Move original implementation to +Closed with the release of Sling JSP Taglib +I might be too late for the party but I have a question on removing the field name comparison for struct type infoWe have choicesC Compare field names case sensitivelyC Compare field names case insensitivelyC Dont compare field names at allThe old implementation was following C and the new one is following CIs there any reason that we dont do C? C seems to provide some minimal sanity checks that users will need in practice +I am not passing any arguments to above classes +Then at the end of the ST process apply enqueued writes and then throw open the latchThis way the whole sync approach to deal with replays goes away +Thank you for applying it +Adapting summary and environment to include Internet Explorer Rev +I realize that has fixed your error for now but in production that tag will still be required to be there or some features on the server may be missing +more update to this patch sorryDIt doesnt actually make a difference to the test itself because grants are irrelevant to SBAP and because the table and db were already set to rwx before but defining the overrides in the SBAP tests for allow and allow is a correctness fix +WSDL and windows you give any other parametersI generated code by giving only uri option and it worked fine +can you run tests +Off my default +I was using this with though not reverted to test thatI also noticed that this appeared to be changing the schemaversion every second as well +Ive the same code passing with MRI Ruby +Can we meet and talk about this in personCertainly we could do that but is it really necessary? I must be really missing some thing I see this as straight forward and important change +Stefan can u do some preliminary investigation into this +I saw this failure on merged there +patch committed +patch looks good +The component is included in the current version of the EIP Component Reference as released with Fuse ESB. +Hi Jeff do you have any idea when this feature will be available? you Jeffyou are right Ive overlooked the empty is the AS on EAP and on docs stage. +I am reenabling the test as part of +It would probably be very convenient for us at time not to but is more painful at rollout time +Fixed in in included in. +If there is a problem please open a new jira and link this one with that. +There was a logical error in the testsuite test that made this appear to be working +Not sure what it means +It just sets the anchor text param to the empty string when get is called +latest stuff is in the trunkcompiled a new jar +on ported patch +on making it default if there was a phrasequery +It looks like the C implementation of value equality does not take into account the fields of ref type +Isnt causesAllowed always false? In other words it will never throw a style Exception with causes +This is really annoying +With a build including the changes for I do not see ERROR XY anymore in the nstestHow to go on do I close? Or? Or both +Thank you for the patch Benot +Fixed upstream +Committed to tez branch. + +Thanks +I committed this. +Looks like this would do itsvn mv +Hi ClementCould you submit a patch for this issueColm +It only requires modifications to the server and client interfaces +Could you please tell me how to take this discussion to mailing listsThanksSiva +I tested this patch fixes that error as well +But junit simply tells that there is no tests +If an empty value is passed in it will be handled appropriately +Thanks for your efforts Erwin and thanks to Xebia for granting the developer timeI guess this is not gonna be finished in the Spring timeframe then +Worked example for ve committed a patch that turns such an occurrence into a assuming for now that this is the right thing to do +Its good to eliminate the overhead for installationsBest wishes patch in revision Thank you all for testing +I think there are also examples using Freemarker in the ESB docs as Len +Committed thanks +The war file contains an XML catalog and it is the resolution of these entries that are causing this issue +This is caused by EL method expression used in CDI sample this type of expression is not supported by Tomcat Taking this issue and assigning to M +Thanks very much KatheyCommitted patch to documentation trunk at revision +This is an issue only when you use for some reason onCreate isnt run half the time and it resumes without any data +Thanks Dan for the cleanup. +The Java Reflection API is useful we use it throughout Castor already +Thanks Christoph! This worked +This ticket has nothing to do with documents just database creationdeletion. +Kris can you please try the latest to make sure your problem will really be fixed when we ship the next version +Here is an updated patch for this issue that changes to look for the XN error when the stream is reused after commit instead of XJ +I was testing and found it is the problem because of onclick +We already copy xml files and we use them to generate the report. +Thanks Regis Verified r +We just need permission to do the release from anyone at Yahoo! with signature authority +Realized a lot fewer changes are needed and updated a patch with a much simpler approach +I did a minor rearrangement of to start us toward a fix for the problemr +Im setting this to resolved as Ive changedimproved the wrapping lately +Hi Can we backport this to the latest release in Hive? As this is a major bug +bq +This patch does that +oopsForgot to run mvn clean test so there are currently failures +Reviewed the patch and applied it built a version +Attached is a provisional patch that seems to work for thisIt works for a pretty good array of tests that I have here plus seems to catch the recursion nasty +I am not sure closure such as scalas would work on a distributed multi JVM setup such as Hadoop +But making Investigate as an analyzed test status doesnt make much sense +Removed the code so it isnt a problem any more. +This should be fixed in the version of jBPM we pick up for SOA +I guess we just need to clean up the workspace of Camel to remove theClaus Do you have right to access the hudson box +Any news +Looks like this was previously resolved for Velocity +HelloI am also getting the same exception +Thanks Vinod! +ThanksOn second thought my patch may introduce an endless loop without exit The current request is placed at the first position if an is thrown under the assumption this is a transient situation +patch looks I understand this patch correct we inject Oracle specific code into the common code base +call setLang on reasonText if Soap is the patchBranch revision Trunk revision Cannot mark the status as resolved as I cannot assign this issue to methankssumedha +Patch applied at +Easy way to check from parent directorymvn f clean with the patch I get tests failtest me or do you encounter them too? This happens whether run as a normal project or in the reactor or in IDEA +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Expected output After remove some gadget need to re size Dashboard workspace form +Im starting by replacing content of the client package with a single junit that spins up minicluster on startup and then runs all tests +If I query for points in the box Lat and Lon then I find this document when I shouldnt +This would be a great feature we would gladly take a patch +A patch ready for review +Sylvain and are right compaction ignores the cache entirely and always reads directly from diskWere unable to reproduce using Terrys test harness on our hardware or on EC +This has been fixed by recent build changes +However since ICU +See the is visible with typing letters works as expected +Thoughts on what this could be +Oops wrong call +Created reviewboard +Will need some specifics on how to cause this problem +I also upgraded your JIRA account so you can assign tickets now +sorry as noted thats an awful name metadata length which is the amount of data for which we have meta dataie the length of data for which we have valid crc data +Closing this since theres no votes and no volunteers to work on it. +Thanks! +Any preferences? nice thing about is it gives us a namespaced place to suff data and later retrieve it in with a single call +The only place where it makes sense to use validators is in the marshalling endpoints since there an object is created that can be verified +Thanks +As a programmer what would you rather see private ncontactor private last +Still need additional test cases +Ill take a look at changing the build target for the network client +Fixed on branch revision . +This patch looks good and passes test for me +It is by purpose because of possible ambugities that we currently do not search for in all possible properties or even only the pages properties file +Remove the requirement from the runtime Fine the commit and manually remove it +and I think we have some work do to in defining exactly what +Bulk moved to s hide Mobile items in the context menu of the visual part +Tested +I just committed this +Im going to commit this in a minute +Uploading patch for +BTW the pisoa tests wont run on x hudson nodes +Think of all the cleanup jobs etc + +Fixed +srcmaingroovygroovyxjavafxfactory on +Didnt run it +I think this is still something that merits more investigation +Thanks Thiru +The other version has a solution proposed has problems with the can you provide your changes as diff and add a unit test? This will help us to track your changes even if we have made also some in the meantime and will also verify that the functionality still does what it diff the source code with the bug fixed and the the +Could you your patch and check the box which grants a license to ASF +I ran derbylang this evening and everything passed so Im posting the patch +Thanks LangPatch applied to imageio module at repo revision rPlease verify it was applied as you expected +The rest of the integration worked fine without this feature after all. +Merged r from trunk into branch. +This looks reasonable but rather than silently eating the perhaps we could throw a different type of Exception for the version mismatch and only ignore that type +exclude regression test for x lnx up to end of patch to disable temporary so optimization is attached +Someone please check me but I believe this wont do any good now that I have another look at it +Trygve Harderson noticed that the use of a for the subject used by the ejb timer so the timeout can access MEJB means that if you change the systemmanager password you also have to update the credential storeThe obvious way to fix this is to create a new realm solely for the use of this credential store that generates a subject with only the principal +I can see your point about it being a server side tool but in general am concerned about setting a precedent for loading up with tools that might cause it to grow larger especially tools that are not to be part of the public API +Please review the patch +Bump up the number to e and skip the keyvalue pair if it was a overflow +Since hes the one that did the original mapping I would like to hear his opinion too +Its just pointless noise in the code until theyre actually doing something +In general we really should think about packaging everything that has to do with alternatives into the RPM macros +In this scenario I would like to have two test sets +But the thing is that just reading the markers +Jene do you need any further help to get you going +Ill need to regen the pbs +Please see details at the other bug report. +So it seems the best method is document a token for this and +Updated on branch +Upgraded to jBPM +Closing the issue now that it has been resolved. +Once a these test cases are vetted via snapshot kit testing I could then commit these to the SMX +I can put together some tests +patch to make all internal Range and Bounds construction use the constructor +I looked at and that allows generation of C classes without the need of any templates +Thanks a lot! +I used as the server +Committed to Thanks for the review Ted +Thanks! Integrated +No response to my Q from N months ago +Hmm +I havent thought of a way to detect and log cyclic dependencies +Thx Nadia as far as I can see in your graph it is even worse in term of memory than the current code thus so far I dont see any reason to use the trove collectionsDmitry Are you sure that the trove collections are used a lot during the daily tests? If not could you please check and measure how much it is used using a counter for example? If so do you have any idea why there is no benefit in term of perfs and memory +The easiest solution ive found for now is to unpack the contents of the file and add the unpacked org directory into the +I meant to say What is Aerogear? section +is not intended to support connections to SSL proxies? I am sorry my previous comment may have been misleading +These are functional tests they do not check optimization has been applied they check for correct query results +Id fixed this locally but forgot to commit +Changed to make it more tolernant of changes to the source pointing to the configuration repository +The problem was in have fixed this problem +Test project for FOP updateFerdinand feel free to assign this bug to you +patch +QA Verified. +was to remove soon and replace it with one written for the newThanks Mark I do not recall this so good that you mention it again +Hi BillIve implement this patch according to your fourth and fifth doubt on how to implement support escape characters in attribute can simply use the we use with text you told its not im confuse on how to implement it in lower levelAnd the next thing is the performance this patch i did a benchmark and theres no performance hit with this ideasregs looking at the code again I now agree that you were right to replace the character by sliding the token data down +Ill commit this soon unless someone objects +Im concerned that I may have polluted my environment somehow because I have been able to call from a client built in this manner before and in this case anything that returns a response triggers this error +This issue has been resolved for over a year with no further movement +Btw just checked out on a XP virtual machine and I can confirm the error even with Java +Patch fixes compile errors +Committed to trunk with revision +That is normal can restore a perspective using WindowsReset Perspective +It could maintain its cache of when it first saw a container +I assume this is issue should just be closedrejected +When using with there is no way of gracefully handling users without cookies on a server that only supports cookie based sessions +If we whould change the lastupdate to the email whould be permanent tried to deliver +This will be divided into the following +Sadly Im unable to reproduce this on the current Forge master +Build mostly back to normal +With version and the brand new squid C plugin this issue may have dissapearCan you confirm that +Thanks Chase! +Will can you please provide a log output for this problem? That will help me understand the issue better +sounds like an annoying limitation? Is this for CR or later +Specifically the following causes the bug +This appears to be a latent bug in GC thats probably most easily resolved by bumping the heap given to compress +Does it make it easier to use if this can be binhadoop hdfs concat fileA fileB +It is very helpfulI tried nowbackgroundColor works as expectedcontent does not work +Those are the Java types of course from the API +Is this sufficient for you +Pruned the dependency list +Thanks for reporting occasionally it seems there are problems with the failing mode that is it sends an email on every failed build no successesI have changed the config to modechange meaning you should get an email if builds go from pass to fail and from fail to pass subsequent repeated failspasses are ignored to keep the mails downLet me know if this works +It will always be from the except in the test case where the mock was not setup to pass along the configurationHoping to commit this soon to make the deadline +There needs to be an rpc context and a master context +Daryn I agree that we should restrict the primary and secondary using the same port in HA +create a full set of seam projects with associated servers and then store these in cvs somewhere and on a normal filesystem and use Projects from CVS and Import Existing Projects to verify that everything still works +A distribution folder and tarball are created +I am also blocked building caused by the exact same error environment do both know that Java is not an officially supported build enviroment yeah? You should build with Java until officially switches to Java Neither of you provided an important bit of information is checked out in a path with spaces +Looks good to me +Fix is committed with regression tests. +Now that Derby has graduated it really does seem like its time to vote for a logo +applies to. +Had to move decision making code up from codec into the. +Thank you for the fast and good support so far. +Clearing the regression flag +So I went ahead and made our own version of from the sources along with the proposed patch +So it cant be case of waiting for txs to finish and while then someone closing the cache manager and then when transactions are finished try to remove the listener and not find the cache manager notifier +If you have not installed the launcher plugins manually you should not be looking to use them and should instead use hte SERVER api introduced in eclipse from webtools +However I think it might be good to promote the section up one level to be on the same general plane as the + in +Its too late to fix it in +Hi there me again +Thanks man its pushed +for the patch +The deploy still doesnt work after this +I will review depends Jira issuesNicolas +The attached patch seems to have lots of whitespace changes making it unnecessarily hard to pick out whats actually changed +This is working with r. +Reopening so that we get the correct at the code this seemed fixed +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The class is not a public class +Fixed changed and added a new test case. +just fix typo and improve else could be wrong on the side but not the initial parsing ans spec incompatibility report +I hope that this can at least provide some level of assistance in rectifying this situation +Make sure that eclipse is actually running with enough permgen etc +Issue has been tested on CASSANDRA and should be resolved and closed +Yes but its not a simple answerThe update to was done to achieve file attribute correctness +IIRC the streaming parser does some when searching for boundaries etc +The code has been put in attachement of the Remark After osgishutdown and start the endpoint can be resolved and route created +Surprised to see this hcat test fail shouldnt be affected by the changes +It this ok to for +What problems does that solve +I find it makes the code clear and small +This issue can be considered solved by the implementation of a different approach in displaying the configuration factories and configurations +Works as expected now +Im thinking of taking the JAAS section out of the docs for the test app since they dont seem to be needed I think simplicity is key for getting people up and running with a repository to play withAt the same time the JSR spec says very little about how security works inside a JCR repository and Jackrabbits documentation on the subject is also rather sparse +del files when is run +AM Nicolas Spiegelberg should I push to the titan hbase or should we wait for the trunk refresh +It looks like the ec client that we are using is using a deprecated Scheme constructor I will file a bug to them then +Andre I am double checking the installation will report next +Looks like its am running these tests locally to see if can see hangs +but his counts only for programs started from +The DATASOURCEKEY is the default name for the attribute butany name can be used +Workspace resolution of extensions plugins should work now. +fix attached to +One thing that bothers me is that it requires a fork and weve already seen that forks are expensive for our servers +Also changed the scripts so that BDB library version is no longer explicit and the scripts behave in the same manner as etc +Thanks +Give it a run in the next build and see if it helps any +Seems to me that couchindexserver and couchindex should take care of all of this without requiring each index to reimplement this simple code +point +Is it just me +test for initialize i said on the mailing listThis is not a right test for a end user +Unscheduling due to inactivity +Unfortunately similar documentation is not yet available on the Ambari project website but it would be helpful to include such documentation so we will work on that. +Does it mean that a fact update is then NOT equivalent to a retractinsert +As per users per users request. +Verified in JBDS H Linux +This has been fixed by making the code recognise local pairing established on previous calls and discount them. +There was also a sitepublish goal to push from staging to live +Thank you so much it works on Rackspace! Next step Mahout +Rebase after went in +New patch attached last patch had unrelated changes to the build xml files + would leave optional resources and other stuff like tiered resources and data locality vs optional resources for Resource scheduling for +Please elaborate because Im not sure I follow your reasoningThese methods dont need to throw a checked exception +I am just attaching the diff with changes that I made +See for a test similar to the example on pages in PDF +Let me follow the installation for go and see if that helps +If I delete one parent doc doesnt stick orphan children to the next parent but it happens after optimize! This seems fine deleting a parent doc and not the children results in undefined behaviorbq +Please let me know if more work is needed +then well just make sure that its branch stems from rather thanI had not thought of that and have kept bunch of features in limbo +Closing old bugs. +Can we consolidate them +If a user of the resource manager is not given the right flags how can it be expected to behave correctly? The bottom line is that does not support transaction suspendresume +Done +Thanks for the pointer Ted. +This seems to be more behaved now with all the latest changes currently the tests are causing most of the errors I see when I run the unit tests +Ok +This breakpoint should behave exactly like except it will have a different marker type that excludes it from the scope of +The fix essentially disables the external ivar list for objects reverting back to the original logic wherein ivars live directly on the wrapperIm marking this resolved but Id like to take a longer look at it to see if theres any way to solve this without moving ivars back to the wrapper +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +DiveshIn the screen the order status is showing as CreatedAs mentioned it only happens when the product type is of ServiceAswath +This issue duplicates by +The bulk of this issue was fixed in b though in CR ive modified the default for the remove scanners on shutdown setting to true rather than false +r +there was one person reporting a problem but I havent been able to reproduce it +bug of a bulk update all resolutions changed to done please review history to original resolution type +Will need to check with Donald if he thinks we can include this fix +That could work +Added validations Added junit tests for string data Added tests for wrong think there is still a possibility in a pathological case that you could return a value when you should return NULL +thanks Ivan for fixing typos +This issue will need a reproducible testcase to move on +And than floadBundle tag will be added +Fixed now. +But let me know if you want patches for all of them +this is the review board Johnny are you planning to work on this patch +It is not an API I had been aware of +Closing it. +I think in my previous comment I digressed and went a bit tangential into HA +unnassign +This issue is introduced by the patch of +Ill do my best +Basically the logic should be find the best node to stream from +Double check CF is exported as service +Ill commit soon +Aww man its better with the profanity. +A test case showing the behavior with circular dependencies +zip files +It seems we set the state to started waited for the server to respond the module was up but then we didnt set the state after that +The main risk is if it is used as a coordinator when it starts up before it discovers the rest of the ring it will think it owns all writes even though it doesnt +It passed all the ant tests successfully on my machine on branch +An outright reject of the form submission should get the job done. +Investigating +I am tentatively marking this issue for because my belief is that it causes most of the tasktrackers to freeze up during times of heavy load on JT +The conclusion that Doug Owen and I came to during an offline discussion is thata Deprecating code doesnt mean it is an incompatible changeb Removal of deprecated code means that the change is incompatibleWhat do others think +This has been resolved in which is great +Marking as closed. +Updated patch +ok +Not sure this issue is actually stale +Before the first logical insertion we make a copy of the current +Bumping off valid anymore load path can be modified by people using and branch is dead. +I will look into this starting next week +the fix is in WTP helios SR release so I assume we can close this in time for issue has been resolved in the updated WTP version included with Eclipse JEE Helios SR +More investigationIve located that spellchecker collate function is the troublehttplocalhostsolrspellqclefrbrAspellchecktruetrueThe query failed with the same error as abovehttplocalhostsolrspellqclefrbrAspellchecktrueThis query works fine +This issue was fixed ages ago +When I checked the default encoding on Mac it was +Should be ready today eveningThere really lots deferents in branch +Committed as of r. +Closing. +Can you provide a testcase or environment to reproduce +We have no option to store our jars in a maven repository +Everything worked well in OSX once I manually copied the json fileWe should attempt to recreate the file anytime we try to access it and it does not exist +plus each thread has a stack of MB +So it includes etc +According to originator this is occuring on as well although I am unable to reproduce +Ill look at committing this +The failure is expected +Bulk defer all frontend issues to +The original patch for is outdated +There is already a issue for the parse and delete XML was chosen over JSON to remove the need for a extra +Where is set to empty String I would expect the same behavior as if it set to Null +Committed +since you volunteered Im assigning to you +Applied to the subversion repo +Ill check why probably issues with absolute path in the package +I am sorry about teh last file +Ill get the CVS updated once I have a chance to do a final review of the agian +Itll be trivial anyway +I believe there is enough info here for me to look at this and will make every attempt to spend some time on it in the next few days +Moved from to just after merge of old branch into TRUNK +I would like to know how much churn is happening if I have members and constantly changing that means my cache needs tuning +Hi PaulexThe fix looks fine +The blurb alerts users to the fact that we intend to stop building Eclipse plugins after release +Thats why it was not working +Learning to point the Hadoop classpath to the installation was a difficult experience for me +The patch does seem to resolve the issue in my environment thanksAre there any issues with regards to security when spinning up custom class loaders inside a JEE application? Or will this architecture work for all byte code generated classes +This patch fixes theIve committed this patch +Im resuscitating +Process Definition to test the Handler to test the issue may not be relevant for a standalone or servlet environment because they do not rely on the ENC but is definitely relevant for the ejb module +Ive reproduced what you have experienced here on Mac interestingly I can ssh into the instance as per normal just not using the adminconnect method +Can you please check this and apply for the release +This looks better but it still unclear for me how is built in this case +Or the atifactId for the groupId on or after version has the dependencies marked as optional +As as been resolved recently could you provide the patch you told us about +One can argue that things can change after the preemption which may have made the preemption unnecessary +defer all issues to +Maybe get should accept a delTerm and lock on every numDocs DWPT and add the delTerm to them then unlock each locked DWPT? This is analogous to how trunk adds the delTerm in the synced method +What I meant was that being able to grep on the properties being used is fairBut if the cost of it is to have such template copypasted in such form all over the place I am not sure I will go along +If not should probably go in a release or LATER +Just logging the message of the exception is enough +Previous patch had an error +This was ed by Jesse and myself in the review board +Fixed in +Agreed. +marking resolved until being able to can you attach the test case +Bryan looking at the offending code in trunk its already been commented out +You are right on the debug I saw that libext was added to the path for each server is in the domain but on standalone mode I didnt do anything the server looks on the standard jre ext lib without I specifying it +Attachment has been added with description Better example of FO file that uses will also attach generated PDF +new source is the new source code file is a for the patch its applied +Sorry this was lousy work on my part it doesnt work and I didnt check dont know what to do about it yet but there must be some way that can be declared +This is now fixed +Corrected some typos +I will update the patch to address these issues +General DependenciesIts the full list of +Andrea when I attempt to add a layer this are the additional errors that appear in the logs +Thank you. + +The and the real run are two separate runs which means the file system cache is warmed but not the buffer poolYes itd much better if a real query log could be obtained +Done in revision +something +I just had svn update issues yesterday +Committed revision . +library lib was upgraded to stable on EAP +We do run our test suite with DEBUG log level as well testing all sorts of bean definition variations +It also persists with and in Sonar havent attempted to go to Sonar it be possible to create a fix for this or is a script the only solution +Should people on the list be able to successfully send a reply message to user +Updated ve updated the believe that things are much better now but would appreciate it if somebody else can review the change to activemq +So when the SMX has been patched and the bundle has been released to the SMX repo you should remember to adjust the version in the Camel features to match +Attached zip file contains a package with the test that I ran in Java to demo a bug when an Element node context is passed to evaluate an xpath expression starting with +Workaround is to use with as a Element attribute + +Thanks +Vinod please reference the Jira here that will test this +On the other hand on the installer side there is no good reason for using an alternative to JAXB to make the installer as small as possibleBut JAXB seems not to be that bad +That can be done underThe Axis binding does not yet use an that will have to be a task for a later release. +And if someone could actually run throught it that would be greatShankarCan you take a look in particularly at the wording for the +currently you can see all the regions that a region server has open by viewing the region server gui on port Thats true but I have regionservers so it takes a long time to go through them all +NET apps so we have a lot of jQuery code already +Actually Maven is always trying to fetch the artifact from the repository rather than from the original repo +Unless there are specific problems that where not addressed within the original JIRA I see no reason for keeping them in this state when most of them where fixed and resolved almost a year ago. +Needs testing and unit tests +This is not +Nice catch Kim +As I mentioned above this should at least use the we have for this purpose. +WSDL generated by wscompile used in deployed war file +In general the image size should be increased by a factor equal to targetDpi and the target dpi set in the format options +the patch + test to this test sometimes finish with success sometimes fails with and sometimes goes into infinitive with when test goes into infinivie loop +Thanks +then it should through an exception +So I think it will be fine +Round of the x stickers for World +back port this issue fix +Tried with latest and branch +I have worked around the issue so I dont have a pressing need for the fix right now. +Well need a fix on trunk too +The open tables should also not be leaked +I have not seen the spec for it yet but I know that it is coming +first few additions You forgot to set chmod x +Add common code to shut down mini dfs +This was ridiculous amount of work but now there is which will work as expected both for serialization and caveat is that this will not work well with polymorphic types +I couldnt reproduce the issue on JBDS nor on Eclipse IDE for Java Developers with AS Tools installed +Thanks Krishna +Im just assuming that ! is at index because +I havent checked +A case insensitive search for break returns no results +In my implementation I was trying not to create any new entities so I just used entities that are already there +As it is now we can close the issue but Id like some other users of ADSWindows to report their experienceThanks for help and apologies for bothering +Correct terminating end events currently always terminate the process instance which doesnt seem to be as defined in the spec +Try searching all your XML files for class +This was a low priority for meNice workJohn +Hi AlexeiThe fix looks fine +bulk defer to to +However as I understand it the only default value is srcmainresources and this seems like a lot of effort to avoid typing characters +Thanks. +Yes it works fine +Due to delays in release incorporated this in from. +Updated the patch with newest directory structure +Castor JDO tests can be executed from within a Maven build as s the best way to go about this +Thanks +fixed but the algorithm is weak since I check for ? and the following should really consider disallowing mixing ordinal and named parameters in core +Attachment has been added with description patch has been added with description list of has been added with description Test FO showing the problems described has been added with description patch has been added with description list of thanks for the updated patch +Logged In YES useridThe problem was solved by setting the directive true +Let me know if the changes look good +I added a trycatch for the conversion +JulieCan you fill out this bug a bit more? What problem are you seeing? Thanks steveh +Added patch Fixes +Hibernate depends on the Cache release +ReopeningI committed though there was review still going on +Ill give it a shot today +Revised the test to account for the changes +Yep I expected this to be an SDE specific issue thats why I suggested to upload the dataset on Arcy so that I can use it as is +whats the status of this? I see there is some code in mavensandbox +Although Amys patch caused more failures in streaming LOB functionality that functionality does not work without the patch either +I will make the necessary changes you recommend and the componentI will also check out the new async routing engine and see how it worksCheersAshwin +in the local storage pool was registered as varliblibvirtimages and in it is registered as varliblibvirtimages +Updated the patch with some more checks againsts NPE +Yep I was rushing to convert everything to strings +It can be applied to stest directory +Committed to and trunk +This is probably related +ping +Attached a patch to also upgrade dependencies and parent +In this case the offending hostname value is taken directly from the execution context +Could you please confirm thisThanks +I think I do recall seeing the image a little messed up when I installed GEP from the staging site +Time for a judgement call +When I restart jetty and try to click the link i get the page expired error which is to be expected since there is no longer a page with id in sessionin order for to be stateless it cannot contain any components that generate clalback urls such as Link Form etc +oh I didnt know this wouldnt want to run an external program to find it so I think ill add a new property in the conf file to change +Looks like these files got dropped in the V patch for common +Many thanks Sam! +Upping priority and marking against so it gets more consideration +The problem occurred with scripts with a very short execution time +Thanks for the patch but Im afraid I cant apply it +As someone who uses key X with random partitioner all the time to walk through results Im concerned about this ticket +The trick will not work on databases that use upper case column namesThere are many instances where provides more than one path to a solution +Now in the second pass we disable validation and build the config just as beforeThis way we at least print helpful errors on the console without breaking the startup for files +Thanks Timothy! I will try it soon and let you know how it goes +canceling and resubmitting for +I was at the point of the first test always fails without your changes likely because of the +User can add S only through zone creation wizard as of now in UI +Shouldnt we throw an exception of some sort if we fail after the max number of tries +In light of the maven failure I think I need to move some of the reflection calls from the nd trycatch to the nd +I checked that and tried it with hsqldbmem also +It passed all the test cases on my side +BrianThanks for this very thorough review +Can long varchar be used in a MAX +Yes because this will block the NRT implementation +Are they going to be generated on the server side? Or should I have included themAlso should I provide a version too? Its pretty different but might be very usefull to have this on too +Ant tasks now use filesets. +This means there is an inconsistency with the current functionality where desktop mode on different devices is resolved to different device types +We still only have files in distdevincubatorbloodhound the IPMC vote on release finally started about hours ago +this help will help much thanks +bulk close of all resolved issues. +This requires lot of changes as the pages to get a reminder require be verified in gatein master and reopened if necessary +By effectively dead I did not mean the whole server has stopped working +This is expecially true with git +Resolved via revision . +Sorry for the inconvenience +Is that still a concern +We should file another JIRA for that after this one is fixed +That is definitely a method that makes sense to me +Do you want that just for portlet or for any other components too +If token string is present the username from query and the remote user should be matched to the user in token +I will follow the second approach you have mentioned to log t and to throw tBest Regards +Now with unit Test and some minor changesI will now try to generate a diff and attach it +Success! Thanks! +blocker for M bad synchronization bug +Great these properties to the file in Rev +Might be related +Reopened +reopen since this patch has been patch since test structure committed in r. +there is no page for RCF yet visiblecant reopen this issue +OK I seeSo Ill leave it to you to apply this patch to pick up the fix you made +Verified thanks +Im now using +makes sense to will be done as part of +Jesper thanks for simplifying the test case +Changes look good +The same applies to wsrf faults i dont see proper wsrf faults action uris being returned either when wsrf operations fail on Muse +I ran the patch on my databases in Ubuntu and did not have the same problem +It looks like yesterdays patch was incomplete in that it ran against old code with some commits from the last days or so not included +Thanks Rakesh catching this. +Fix committed. +Error checking released closing. +Can you still reproduce this Paul? Im using the latest code and cant seem to make it happen +Updated on revision . +No test case with broker config provided +Im curious to see if there is a difference between the way interfaceabstractimpl and interfaceimpl are treated in the various contexts +Data and ranges are prepared to transfer in the first loop +the Directory maven remains containing an ant build file and some other misc that overlaps the etc dir +sql +You can create Flex applications without any MXML code +please attach the file webrootrequirementscoldspringbeansa repeatable testcase working outside your environment would be helpful +Oops my mistake +Attaching new rev of the functional spec for this feature +Resolving as Not a Bug +Sure just adding itcan we add the build parameter pullNumber both on and? also need to wget and wget ? +Agreed +Hi Alex the patch is not up to date anymore so can you please come up with a new patch +So its up tothe admin to decide if they want to allow these tags now +Rajesh working on this +dig where this option should update the ast +Thus I made the assumption that all change entries were in the timezone of the person creating the site and just used that timezone to correct the date entry +My own build still wont pass even after i restored modified files and completely recompiled +Hi TimI would recommend posting your question to +fieldName status type +Login issue has been resolved +April for desktops and until +Here is a fixed version of the patch +Attaching a second rev of the localization changes +From looking at the consumer my guess is that you are getting duplicate messages +We dont expect to see Derby support for JDBC on that release of Derby +And I guess since it is not loaded through Groovy ant tries to find it by itself and does that wrong +In revision added a system property for controlling the cache size +We should generally wait for that event and not seek permission +Better patch +Fixedupdated test case showing slight fix using +JIRA closed. + module is operationalI plan to migrate the dependencies definitions to parent +PrachiRegular user doesnt have access to list APIRegular user only have access to listZones APIPlease add the properties to listZones API responseisDedicated truefalsededicatedaffinitygroupid After you are done with this API change please assign this bug back to me +Committed revision trunk typo I added docs to the link above +Can not reproduce +I will remove this upgrade step +That is why I needed to make the operations at the stack level +I removed the part that tries to add ORDER BY the nested table PK +Yeah thats also an option +Would be great to get this in before like the homedir configuration isnt documented here this should be covered as wellAny chance we could get an updated patch committed before? Were still at the point where users have to read the code to understand how to use it. +new patch is attached +Thanks Philip! +You could probably do some custom logic to detect if they have no roles too +This patch fixes the issue. +If an application completes after the has asynchronously retrieved the list of finished applications but before the status update event is posted to the then the application will be added to then cleared from the list of finished applications before the had a chance to notify the node of the completing application +Moving out +The is not encoded correctly +Let me attach a screenshot of what to select to Errors View where do I find it +So it must be something else +Ive been puzzled by this myself but thought that it was because a string is per definition one token and thus not a phrase but a word But since this worked with dismax it now looks more like a bug +Small update to +Is that expected? If it wraps in the PDF but not in the HTML it might be better to say something like +Also all the edismax parameters +Patch including all proposed enhancements for the master Ill also provide an adapted version for for commit m Adding property to of standard distroSending srcmainetcTransmitting file dataCommitted revision . +Closed as fixed for +The data flow now is simpler c +Currently we are automatically assigning a free character to it and it is always coming as g but this is not required for +The status for this RFE is set to resolved not yet closed because we may revisit it when future version will be out +Using would be nice +Messages should be ackd from the same connection theyre consumed +Ill do a quick search +simple bundle project that illustrates the issue +Test to demonstrate the problem attached +so what are we going to do regarding this jira? Remove these demos that didt work or add dependencies that required by these failed demos or what +Its better at least to me to see it all wired up in one placeI think Im going to checkout from SVN and take a look at creating a patch to create this behavior +Please read the README in the directory +Odd +system tests framework +I do not see any major issues with such an addition +Applied on closing all issues in resolved state that have not been reopened nor commented over in the last month +I might have to leave those changes and some kind of test to someone else unless I get a good Internet connection in San Fran and get time to look again +Thanks! +Hi BrianKaraf leverage to do this stuff so I dont think Karaf can help here +If these wsdls have conflicting namespace declarations say for instance two different tns or targetnamespaces then when we merge these two we do not resolve or treat these conflicting namespace +Fixed in nightly build +Submitted and tested +Stack Did you commit this to? I dont see it in the SVN log +Deadlock between and +Thomas big thanks for catching this problem so quickly +Attaching the result of running ideamodule goal with the X flag enabled. +Trigger Jenkins to test again +Sample FO file illustrating the file +Checkout +bq +Attachment has been added with description Teach FOP to start from any dir on NTXP like thanks +Improved patch which connects the PMD tasks to the ivy checks we have +The patch is also a slight performance improvement +Min has added the code and Jessica W has been working on the UI in the past couple of days so this looks ready to this record since is now released. +If its ok I will upload patch for +Hi Simeon was this done? Is the version that Yang has with these fixes +Were going to have to revisit how our bridge works +binmahout seqdirectory i +Anyway this is a minor problem just thought it would be a good idea to mention it here in case someone runs into the problem and doesnt know whats going on +nikita first thanks for the patchThe changes to Cascade happened to be the first one in the patch so its what I noticed first +There is a workaround for now +Thus there is nothing left to do with this issue. +If I send the note with several CRLF at the end then it works fineRegards Chemi +Ive just committed this +tmp file and deleting the old aggregated file +You can create a singleton that can process multiple sessions at the same time but its quite hard to implement +I tweaked around and Im sure I not setting properties +The second one tests writing multiple types to one file but adding a second type later when instances of the first type are already written +Closing. +This occurs only if is set to true +Jon This patch doesnt change behavior +A jstack of the hung process would be most useful +The reason we do not is that thrift does not support overloads not because its actually better to have getslicebynames getslicefrom etc +Even though I wiped out some of Daves changes accidentally since SVN doesnt quite handle file renaming that well the issue should be resolved +This should result in a and error sent back +It always marks the type as tuple whenever the field schema is of type bag or tuple +Also see for more on what Kay Kay is on about +It is all easy until you get to arrays maps and recordsSpecific reflect generic or future representations can all be different +Any news on this? Does the option make a difference? How have reacted to the connection pool provider issue? wanted to add that I am waiting to hear back from after I gave them the code last night +Deleted spurious Added svnexecutable from the patch +bq +For me its working fine +to branch +API might need to change +This version patches against the current trunk version +type +Tom I dont understand the distinction between andBoth classic and nextgen MR use the same runtime +One way to avoid it could be that in the Online state transition send a leader request that contains all partitions +Your should work the original one is used instead causing a malformed +erb template +However the productization team is not publishing the tests to us so this is what this task is about +Yes I understand that people run code from branches all the time +I have however fixed it in my zipper plugin +Patch applied with commit revision edccfbceebbbacfThanks +Closing as invalid since Ivy doesnt build jars + +The sample pom also includes this one +In the code above if I comment ILLUSTRATE the code works fine +Which console? Makes my life easier when I the task +Hive will query the location and get the list of paths at the time the query is being executed +Thiruvel please assign it to yourself if you are planning to work on Thiruvel +Committed to trunk +It is fixed in +I am pretty confident this fix works +I have already committed a fix as version +apparently Luke fixed this while I was working on the patch +Could someone review the patch before it gets too stale please +it is never really unknown +I had a question about +Any update on this +Committed to trunk revision +Here we have no componentembeddable referencesHowever oddly this noofficenettestPersoncompany property is defined as a component a component with exactly one property companyId which is the entity property that is the target of the +Question When Ivy is iterating a chain and it finds instances of the same modulerevision in different subresolvers what is the expected behavior? Can Ivy use to decide which is latest? Or is it incapable of evaluating these as distinct revisions +It would indeed be nice in that case to try and pick up one with a single column if there is a choice +Punting issues from RC to final +I can take a look. +Given there is which can specify all these things in we can ask user to configure in instead of annotation +Looks like the consensus is not to upgrade the jetty in +Since its busy writing to at least one file in its lease the lease wont expire so the renamed files remain locked foreverActually come to think of it isnt the issue in much more simply fixed by releasing the leases +An error has been corrected in the signup page please try again +Need to add a unit test to verify that things work as expected +The previous patch had missed a few changes +Currently a lot of manual testing is done for the capacity scheduler where we need to measure details per job +Juan +I applied this +Im going to close this after I created for RARWe may want to open a new one tracking write support of compressed z archives right now we only support the Copy algorithm +master PR requests merged +At first sight everything seems to be ok +Create object inspectors for some of the classes that handle complex data types +I must apologize the previous attachment was a patch against the entire directory instead of just +do you have maven archetypesexampes yet? have you tried to see if they just work +The test case for should drop the table it creates +QA Verified. +Makes sense we can just live with the existing counter for nowAlso lets register the controller stats at startup +I have just tried +Scope detection like these existing filter classb +Will do later today +The implementation assumes any class loaded by the System and bootstrap classloaders are system classes +This leaves only the openejb ones to addrev go back to released xmlbeans plugin +This sub task adds a sample to demonstrate how to use SOAPJMS transport +AndreasSorry for lat confirmation after adding the Woodstox to class path we never get the issue mentionedThanks for your help. +Is this likely to be as a reult of bug fixes only being applied to one of the two branches we build fromPlease retest with this buildconsole and apply fix if appropriateUser infoI see when using the Management console in that it shows the queue depth at one less than there actually is +Yeah because of tests are going to keep failing until the hadoop version is upgraded on Hudson or someone makes that test check the current Hadoop version and bail out if its less than I guess that if the contrib tests pass then it means the patch as far as the hudson build is concerned passed +How about this to leave things as objects until were ready to turn it into a string for tracing +Axis svcclient detected faults both time correctly +ThanksmavensynchomegroupsddbdbfithtdocsmreporeleasesrsyncsshPal Brattbergbrattberg +Modified this for the general case I think the warning needs to be chatty to drive the point home. +Maybe should not be starting but either way the deployer should read the file to resolve the variables correctly +Ill wait and see what happens when we upgrade. +This is a preliminary patch to upgrade the build to SBT and for and +buildivylibcommongriddevbin +Getting this error message is quite strange as it should not been seen under normal working conditions +Cant think of a good fix for this yet so pushing this to the next release +Servicemix seems great to achieve thisThanks older issues for Apache since were no longer actively working on these at the moment. +Fixed +I just committed this +Because of this there is no way for a database function to return a NULL +Geronimo done revision +These tests are live and run against amazon ec and rackspace if the live profile is specified AND if live profile contains credentials for the each cloud provider +Todd Did you look at this issueWant to invite you on this issue for solution discussions +I have added your fix to the mavenized Seam Project +need to look if it can make it for +x Outside of this test measuring inserts gets a bit iffy because GC begins playing a significant role though this should even out with enough runs Even in this test GC is still likely a factor at this scale since reads end up being faster than writes BBs dominance on inserts is undeniable though both it and PS have a significantly higher standard deviation and BB is very bursty reads were very consistent across the board here these are very trustworthy resultsh +Thanks for the explanation Rick! If I understand correctly the fix for removes the hang but it still ignores the property if the directory doesnt exist and the directory is not created +Fixed in r +I use the same technique as facet sorting +This patch makes it avoid that only during startup safemode +Their current descriptor is just wrong I dont see what could break by fixing it. +Please check your class path and extension directory for other versions of the the DOM classes +Our web clients always use Accept textjson while our clients use something else +Attached a test patch +The edit panel does not show the current value of the portlet and the view is showing the long gbean name +We will upgrade as soon as now we solved this problem by setting the listeners autoStartupfalse and then have our own bean to start can close this issue. +The real issue is the failure to remove the connection entry in unregisterConnection +Closing as this has been fixed and verified. +Perhaps the resolution of this issue is to update our How To Release page on the wiki and call out the exact details of how the artifacts should be named +As of Spring GA subclasses will work fine with Quartz as +Need to provide a walkthrough of the API standard practices tuned for a new userExample code is handled under another bug +If you wrap your query in marks before passing to cfsearch the parser seems to leave it +See the XML spec section definition for match +Let me note though that things will chnage with regards to instantiation of such classes once we switch to Spring +Fixed in for the patch. +Minor doesnt need its methods to be synchronized +Thank you Luke for the patch +I marked this issue as fixed even though it may be improved as part of progress. +I uploaded some minor changesThe name of the error label component was added to Form class as a new property and a fixed some minor errorsHas anyone tried the code? any suggestions? +Matt Can you make a patch? I think your patch is off an older version of trunk rather than +The provided solution of is fine with me my proposal was just to keep the orginal behaviourSo it can be closed for my sakeBest regards Trym +hehehe +Merged into the master branch +Whats requested here beyond what Version Watch tool already does +Agree this can be improvedI noticed a far worse condition in the logs today where we would log s of exception per second in some situations when the peer is down +Viraj could you please rename the patch to and cancel and reset the patch state to? Id like to have the automated build pick this up +If you want to help help +Its the fact that if you create new loggers over and over again things are going to be slow +Im now testing the patch if everything will look good to me Ill commit it soon +batch transition resolvedfixed bugs to closedfixed +According to the Enable database class loading with a property section of the Derby Developers Guide should be prefixed to the beginning of the classpath specified by ALTERJAVAPATH not appended to the end +IMO this is not a jquery taskI mean For and we should use to make those settings persistent for each user +May be fixed in M but need to verify +But hopefully will be able to nab it during the coming weekend +Thanks a lot for the contribution Binglin. +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I am not able to reproduce wonder if some compilers interpret this differently +Attached +environment variables +IIRC it doesnt implement filtering on modification items +You are right the main objective is to offer a way to delete the original file +Thats a great question +Resolving this since it has been committed +The consequence of this deviation is that the authority component of a hierarchical URI such as s will parse as a authority +do we need to do anything else on this jira? Does the release process on the wiki need to be edited before we close this +This looks interesting and the gem code base does not seem to be huge +I havent worked with this tag or the other tags so I will defer the work to someone else +In case your time permits an attached unit test could speed up progress considerably +If you would still find this useful would you provide patches against a nightly build +The files are attached in a zip +As you said its something we should consider if nothing else provide a section in the doc discussing a pattern for securing an app based on Home objects +Oh see o +It proves that works +I can attach a patch with the regenerated classes do generated classes need the Apache license header tooIll update the doc too but Ill do that in another patch +My jni directory contains an executable and the following directories build jffi libffi and build directory has an empty subdirectory of +Yes I found the statement to be required otherwise Abdera will not be able to find the class once itis outside the trycatch block +doesnt version is you provide more info about analyzed project? CI Engine build log and so on? Tomorrow Ill try to figure out what can be wrong +you are so nice +There will be no more enhancements to +If the shards file is an XML file we can add parameters per shard +Id like to commit tomorrow unless there are objections +But there were some further problems which are fixed were always wrapped and the algorithm for choosing the correct break position was not correct inside tables +Sorry I didnt see the Larss comment +Please review this patchMost functionality of Enhancer is implemented and tested using a mockStill missing classpath processingSome notesUsage is always printed to stderr not stdoutMultiple persistence unit names are accepted and passed to one at a timeMultiple jar file names are accepted and passed to one at a timeOnly one classpath entry is expected but currently no error checking on the second oneOnly one output directory is expected but currently no error checking on the second oneTesting is done by adding testresources to the test class path but these artifacts arent put into the api jar fileComments +A version of the patch for branch +by delaying the actual binding of until after the component is activated and the is availablePlease close this issue if this fixes your problem +BTW you could also extend and override the sendRedirect method +Verified in CR +Awesome looks great to me +No response for months +If someone later comes up with an idea on how to improve this code that work could be tracked in a separate issue. + +today I will update the patch I am testingplease it was referring to the patch I think you should think about this scenarios in my patchI hope that the hbck tool will be strong thanks +Is that something we should be concerned with +deferring past anyone looking for a workaround when run from the Component Descriptor Editor in Eclipse uses the projects first folder resource as it output folder +Commited a fix for the Ivar. +Try +Maybe its a desynchronization between and +We can move trunk over to ready for JBT +patch applied thanks! +It would be confusing for them to be in two places no +I think theres a missing addition to the CHANGES file here +Do I still have a couple of days to jump on this? I dont know whether I can work on it tomorrow but Im fairly sure I can on Thursday +Thanks for the explanation and fixing it +I updated the due date just to get it off the top of the TO DO list +patch that solves this looks good at first sight but has a serious drawback The Servlet Spec defines the result of as +m Same result +Lets do it +Ok then +Not for GA +This issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +Pulled the change out to +Here is how far I got until now +I dont think we should do this see my comment in +This patch deactivate the modified stated when a merge is made with a deleted state +bin geronimocarwork +thanks. +The wsdljava content in that zip doesnt work correctly +I think its a great startIll apply these later today unless someone else does it firstThanks for the patches +Correct +This would be a page ripped from S +Manually tested +Actually I have another question for you Jonathan +So although the collection of source directories may be manipulated by any one plugin for consumption by another this does not hold true for the output your issue did lead me to some necessary cleanup of pushing the generated classes to another location then copying them into place +Somehow didnt work for me yesterday but now is working just fine +This patch observes this policy +I did not go over all the test plan outputs I made sure that there are no data changeslooks good +I lost the whole morning cause s was dying with no logs no nothing +Can we please just not simply delete this but find another place to live +But indeed this is a quicker fix to the problem +That way wed avoid completely thread scheduling issues +I got fewer changes than yours +Validated the fix to with Teiid build +It says the database is in slave mode even after a failover when it is not actually in slave mode and the real problem is the jar problem +Merge without completing all the +Move the fix to +It had to be some kind of environment problem because I tried many combinations and all started properlyNow I didnt have to increase consumers at it worksSorry for concern +The web UI is broken now +There is a manual work around to get things running again so Im reducing the priority tooIf this happens again please reopen the bug. +systemslingselectorloginDoes that make sense +I just committed this +the HDGF code in POI fails to interpret this file correctlyIt would be great if someone could report this issue upstream to POI and add a reference to that issue here +So thats already a unless were talking about adding some ajax detection to the logicMaybe the fix is just to alter the documentation surrounding Forms method property to say that its a recommendation to the markup engine and doesnt guarantee or even influence how the form is processed +I think the very first change you made where new Comparator is brought up a line violates the col rule +Got this done but Im sitting on it until is out +agree in rel +As soon as we support multiple the solution is effectively becoming more general +The sooner the changes are in the more time we have to do before. +might already be fixed but should be nailed down once and for all in +Probably one of us has something special on the test machineWe really need a continuous integration environment +It will get fixed automatically when release with EAP + +This appears to have been fixed can this issue be marked as resolved +glpk does a scaling of the coefficient matrix by default which would be a nice addition to too imho +cfg files +Migrating bugs marked as resolved before and tagged as version as fixed in RC +Do you have an error in your logs +Lets just cancel and resubmit the patch +Cannot be reproduced in bug of IE but would be good to have this fixed +The order of execution of callbacksannotated methods is not defined +Its now running with threads +The test cases is attached + +Patch contained an insufficiently long wait time to make the test work +Then developers can choose the method of using Maven thats best for them and get consistent results from the or through Eclipse +Ported to all library specific configuration parameters are present respective libraries not +Closing as Rejected as there isnt an issue here. +Alexey please verify +I will review on Friday +not reproducible at for now +The problem is that your is missing files building on our redhat boxbuilding Apache error cannot find input file find name are NO files in your +Going back from to solves the issueFor that bean there is no simple workaround since it must be request scoped +But we need to refactor related Quick Fixes +Which annotation is reported to be illegal? May be you can open a new JIRA with a sample war or test project attachedthanks Thanks for your response +When DFS client commands are used along within other hadoop commands say job submission within a script +I think that may be more trouble than it is worth for now +Also updated the FAQ in the documentation +I have checked out the sourcecode so i can see if I made a misstake or there is a real the sourcecode I can see that the facet for ejb is set by a search over the dependenciesbut i cannot understand why it is not set by the value of the as you can see from my sniplets you can set the ejbVersion there +A more precise statement might be there appears to be code that implements truncate table its state of completion is probably unknown +This makes the accessibility changes identified by KimTouches the following filesM javademosimplemobileM javademotoursdbM javademonserverdemoM javademonserverdemoM javademonserverdemoM javademosimpleM javademoM javademoworkingwithderbyM javademolocalcalsrcM +This seems to be the same as +Are we worried this is +Fixed this needs to be revisited once throws the itself +Applied thought this was applied too but when I did a fresh check out today and didnt see this patch in it +Resources like queues datasourcesetc +Deferring all remaining issues so I can release in JIRA +Anyone know the status on this or how it will impact the need for hbase to implement this +Working on adding unit test +The fact that somebody is using means he is using it as an API +It was a long break from workThanks so much for taking this over +Im using and Im bootstrapping via XMLll try to prepare an example that demonstrates the it any chance that with the latest snapshots these problems will gone +Page for removing +But I sooner found out that pagecontent was hard to work against if you wanted to add some standard buttons on everymodal window +Add label tag for those form accessibility issue in the plugins foldera +Also to document that when you turn RESPONSEtrue it turns an In message directly into an Out message with all its headers. +It was failing when Virgo was running under Eclipse +I am seeing the same that Dave mentions above +Leaving JIRA open for now I am going to add some additional tests +No need to vote on website changes so this can happen wheneverUnless there is some marketing item that we need to enable for the announcement +t this a too big APIbehavior difference to not have in the +This problem has been fixed by Vivek and confirmed again. +You would only need to modify the Party Manager to restrict access to certain dataIn our deployment various party roles have their own data entry screens +One way to add this automatically is to use the +Attached a patch that simply enables it for tests +IMO these buildversion changes should have been done the moment was released before any code changes were made such as to fix +Q What about the +Added a test to catch the problem +Copy pasting code doesnt seem like a good idea +If youre unable to change how your classloaders are structured then Ive got several options for fixing this situation +Sorry +Im not sure who causes the delay the spring ide or the subclipse plugin +Ive also added a license to some files and removed unnecessary javadoc tags at r +should be relatively easy to add helpful logging and would save a lot of time and frustration for folks +Please try again with and see if problem persists +I did not revert the switch statement changes since I donthave an automatic way of doing this nor a canonical list of where I made those changesI also maintain that switch statement braces should be avoided unless you really needto define a new scope in which case you probably shouldnt be using a switch statementin the first placeZheng please take a look +I just committed this +Thanks Knut those suggestions definitely improve the materialAttached is a revised patch diff and revised HTML versions of the two pages +This is because nillabletrue attribute is missing in map parameter element +Support is good enough for now. + +Moving to since nobody is working on issue has the following side effect when a Pig script with a macro is embedded in Python execution will result in an error +I could have sworn with this worked but I redownloaded it and it did not until I added in the host attribute +Closing internal build tracking jira. +Interesting +Adding function to support non anonymous complex type in wsdl schema +Closing xml and console issues. +Renamed +kind of funny how it likes over and like it doesIve also noticed this +Did you also manually test it out to view the log +It should be Nightly Development and discovered that were setting nightlytrunk as the path to the file which works because theres always a nightly trunk site +well directly from the shell without having to run inside a torquebox instanceIf that is not the case then I dont see the use of that example +updated patch for this issue. +Fixed the issue by sending the list of values for the HTTP head as a listThanksThe change is not correct +This patch has been checked into the dev branch +Yes your change solved my for the very quick response. +Candidate for SOA traced back to a cvs can you double check the operation of the brew job and liase with Martin to get him to check the release you sign +WeihuaJust had a quick glimpse through your packageI really like in the of the tests the and the pre postconditions +Included in +Most GNULinux distributions provide a servicelike implementation for backward compatibilityFor instance systemd provides a service implementation on my Fedora +belongs to JBPM module +war distributuon where we could test this? +plz can you do it? Changed +Committed to trunk +Duplicate of this record since is now released. +From the above it looks like has been deployed into the SOA server rather than the standalone tomcat +Fixed in V of the plugin. +If you are not working on this I can port the patch to branch and trunk and post the patches here +I hate to say this but I couldnt reproduce the OOM error +I dont want to get on anyones case here but I submitted this bug more thansix weeks ago with a fix and it still hasnt been committed to make it easier to commit Ill attach a diff for the fixed XSD against thecurrent CVS version +That should negate any confusion for someone looking at the project docs and for future inclusion of those docs +Applied the changes from to branch test for results in hudson +So disabling for a quick test run or two may be an easy test +So hopefully after is fixed Ill be able t close this JIRA too +It doesnt extend because wont be used +This configuration option will be supported on inbound as well as outbound adapters. + has the Flume implementation you are looking for +i have not added tests since the test require creation of large input files +Has the ability to resize images and store them in a local directory +Go for it as in my view this is stable code now +as this funtionality can be achieved by other dont know how to do it automatically manually doesnt seem to be a good idea +Otherwise the old message continues to be used in other locales so for a while those locales dont get the benefit of the updated messageIn past releases the new messages were translated by one of the companies interested in Derby and the new translations were contributed back to Derby at a later date that was after the original release +which is the full source with the patch +Sigh +Resolved by commands in the comments. +Per Romain I am attaching catalina log which has times it took to execute SQL +Some change in is needed +Tested locallyI have committed this. +There is a lot of functionality here that needs to be built out +Reopening this issue so we can keep track of the upgrade +replaces fixes comparisson induced by mild heat stroke +Investigating +Sorting works at +I think this was introduced when compression was added +We need an EJB RC release that is compatible with andor an updated installer to a task for producing the next installer release +But then why not update the status to in progress or partially fixed and set yourself as the assigneeRegards at this stage the workflow allows only Cancel Patch Resolve and Close +Freddy please check that description of this rule is ok +Right +Nope that sounds right to me just a target on its own wont do it I think the preparation of the output will need to be included in a utility +Moving to +Also standby can not read the complete edits and may end up reading only till the gapped edits as max and will not do any tailing furtherIf this continues for long then standby will have big gap of edits and on switch standby will fail to become active as it will not have continues editsXiao is this the case you are seeingI think in HA case we may have to expect shared storage edits mandatorily instead of falling back on to local edits in case of edits missing in shared storage while loading on startup +If the workaround doesnt work for you we can dig deeper +But these are exceptions that occur at most once per document and in most scenarios I would think far less frequently than that +Triaged for +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Next version is ok for me. +I suspect my attempts to repro a smaller version failed because there is some race condition andor other moving part that only gets triggered on this more complex project +Jonathan we want the ability for clients to explicitly not use the column family default TTL +Service engine archetype updated revision . +In future this could also be used to throttle the load at the map reduce framework +attach has been introduced by a similar ticket in I think by PedroUse svn blame to find the other ticket and if we find a solution for both quickstarts itll be perfect +Another possibility is that you are using a fake sender for your tests +The should be maintained on the clientwriter side +With no objections Ill commit it later today +Those changes will probably go into our current patch sets for upgrading to hadoop +Im a bit puzzled with this one +You might get more repsonse though if you ask such question to the mailing list +It is a Cocoon issue as we provide a util class abstracting from the environment +Or will a finalizer introduce other problemsFinalizers have their own share of problems +Can someone validate that this code works for them +It looks likejar is causing the trouble +Working on resolving issue now. +the link is deadany progress on the need to document an unused API +If this is going into then a light touch would be strongly preferred +Im trying to look into such kind of issues and possibly patch them up +Looks good +Instead I would suggest to either add a tooltip or to print a note directly see attached scrrenshots +Please let me know if you find others +whats the plan for handling further additions to authentication mechanisms? For example Kerberos or Windows login keys? Would this require more signatures with the needed arguments +fixedThanks for the report +You need to remove the test from +Closing issue fixed in released version. +Added some comments +Closing as a duplicate ofFred Bricon +test logic has been adjusted anded new package for selection bar on +this is lack of proper documentationafter adding user as admin instead of able to add the RBD cluster so i am keeping this issue open till the documentation for this feature documented with detailed information +straightforward patch +Ah I see +Good point +On the MR cluster is there a static method for getting a hive conf stored in the job conf? That would be much preferable to passing the conf through from all places where is calledThoughts on this patch or suggestions on a better way to add this conversion +Small to branch and trunk. +As far as I can tell these were all resolved in the massive rewriting +Duplicate of +Will not be closed until tested with usingJBDS Version SOA beta and ER. +I would be real hesitant in doing so +Due to lack of time I am unable to resolve it immediately +Patch +Sorry its actually fixed in Grails +Patch files for codebase +So this can be safely resolved +I was just unduly scared by the log messages and default slop settingThanks Zhihong and everyone for your help +I missed the underscore trick +I used SVN whilst logged in to the hudson zone +This is almost certainly because the core uses an old version of of fixes this +You can tell the actual version by unpacking the jar and looking at the manifest which for what you attached is +But even then I first need to know when it does not make sense. +We shouldnt commit new features in branch +not going to x at this point +Made it into a feature request rather than bug updated the summary +This patch is what Im suggesting +Provide example ofthe bug please +Hacked version of as of to include multi start and end key functionalityAlso a diff from +Yep youre right +So it will work in JBDS too +I have no problem with this sort of change except as mentioned belowThe Ui Label and though look same but have minor differences The language translation for nl is different in both the labels +Thanks Gregory! +Done. +Just fixed above strings and committed +For this issue will address the directory structure proposed earlier rather than eloberated structure proposed later +Ill notify the guys and see what they sayt realize it was for sure postgreSQLs fault +thank you for providing the unit test! +Jay I over there as well +This appears to be a duplicate of +Now coming to know that it was half working by assumed that absence of any comments refuting the requirements of this JIRA meant that the requirements were valid +Thank you for pointing it out +Patch v +Previous patch mistakenly included other stuff instead of +Thanks Mike for looking at the +Ok Nicolas dont worry +Your patch looks good to me DIGY just tried it out and it fixes the tests and I dont see any negative impact +The policy is that every GSS issue has CCFR provided in the Release Note Text field +and that are not managed by the entity being edited are displayed as an and hence cannot be used with Primefaces +Can you provide a patch and Ill make sure it gets merged into svn +The discussion of relocating packagesclasses seems germane to which is closed as wontfix +Should we take this up with the EGbest regardsMartin +Doing this also allowed me to remove the Ignore comment from the test of doing a duplicate define +Only if you query the builder again it might return another configuration instance if a reload was performed in the meantime +This is a data dictionary representation of a schema object which has privileges granted on it +Okay I the repository +batch transition resolvedfixed bugs to closedfixed +master pull request merged +fixes compilation problem with the recent trunk merge +Thanks for all the hard work on this one +I wont mind working through the documentation to describe clearly how interacts with the other params and also implement a parameter check which throws an exception if the user supplies incompatible paramsWhat do others think +passes all has a config option to cache index block with priority +This is the version to be committed now +Hi Thomasbut what if the state is longer than the kBs IE allows? This is essentially why the commandLink needs the state can be larger than thisregards for taking so long to respond a lot of work kept meFirst I used server state saving in testing the proposed patch so I did not run into the k limitSecond I was still a jsf greenhorn when proposing this patch +I will review it +This has been fixed upstream assigning back to Dominik to verify in next release +The problem with jetty is the following When I install the utility jar into the local repo +Is there a facet for the web site content? +Marking this critical since wrong data returned is a serious bug +Looks to me like fallout from switching the driver to be fixed now +Why do we need to store additional metadata? ISTM that restore from snapshot should just be make the snapshotted sstables live +ChrisAhh were just classifying differently +Committed +Its only supposed to be for use in developmentThats not totally true its fine to use it in production +Bulk defer all frontend issues to +I will any way update the patch based on the latest code changes and remove that commented code +on commit +Looks good +This will now not happen but the fix needs can confirm that Groovy r does indeed not have this problem +bq +I asked for configuration what did you have in andor? devMode and other news +all through a pluggable connector framework +If this is the case while configuring an Eclipse project the settings for source target compliance for aspectj compiler plugin should simply override corresponding settings for java compiler plugin +Looks good +Attachment has been added with description File that exhibits the has been added with description Fiel that doesnt have the error because the first block in the table contains some has been added with description File that exhibits the has been added with description File that doesnt have the error because the first block in the table contains some SteveThis is a duplicate of which is fixed in the Trunk so the fix willbe available in the next releaseThanks for taking the time to report the bug anywayVincent This bug has been marked as a duplicate of +be careful +patch for +It will probably be similar to support for creating the clients from API with create attributes set on jaxwsclient in SpringBlueprint config and then those clients loaded from the code +Removed ability to use +I implemented as simple background job that restarts the given application and does not state checking +the screenshot shows the tray icon +Also attaching etc so that we back port these broker config changes to as well +I think we should either fix the ZK NPE or upgrade to a version that has it fixed +Almost always almost universally code change +Oof we should really fix to notice this sort of failure +Yes that would be useful to have pending map tasks and reduce tasks separately +The previous resolution was correct. +Changed to use WELDX resource loading +Ran the failing tests locally +Verified. +This redirect problem is fixed in build +Tested it with and it works +Just wondering if theres been any progress on getting this fix released +Partial selection behavior should be considered too +Show please the mixins of the node exo fact the patch is not the problem because before and after patch application the node exoapplications is not hiddenI will investigate to find real problem of this cannot reopen an issue that has been fixed for an already released version so if you still meet issues create another JIRA issue +I feel this issue is now closed with recent commits after length discussion with max +Trace log from the Infinispan test suite +anyone left in the project who understands the intricacies of the compiler +Chrstian since you are using Maven can you attach the original Maven project? It will be very usefulThanks +Sounds sensible to me +the configs are not enabled +However we would like to know why this NPE exception happens +File a query on the Findbugs group to find out the rules for using include and exclude filters simultaneously +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks +I run your sample and I didnt see the error +No sorry for bringing you false hope By undocumented I meant that it is not possible from with IDEA GUI to achieve that result +Cheers +Thanks Owen! +I guess my question in a way is would it be the same problem with ORM? + discussion with Santhosh I get a better patch +Closed as fixed however the fix requires that either streams have direct length fields or is used +How does that patch compare to this? Esp wrt riskDo you think we should choose over this changeFlavio +It also would blow the JSON wire format +The works fine +Changed UPGRADING file Committed revision . +Setting to future pending a patch +Released. +Load tests executed withjar uploaded by executed without errors no performance degradation noticed in load tests +A behavior more inline with our python heritage +would be nice to tell the embedder to construct teh parameter object and resolve any variable etc +I dont know about but AMPATHCPPUNIT isnt available without it maybe its installed by default on some systems but not any of mine at least +I will take a look at this test failure + +Also given that webhdfs is using json defaulting to json may be the right option to go with +If that is the case then would ensure secure transferIll read the design doc in more detail later this week +Thanks Amar for the review. +extension +Yes its a workaround +Changes for this issue appear to led to another regression +However need to confirm the reason why the two lines were commented out in the first placeAssigning the bug to Edison as he had made the change +If that does not fix it we will try a change to the ajp proxy suggested by Mladen +After discussing with Brian we came to the conclusion that this state is a good first implementation which errs on the side of being more strict +The error still occurs after running ant release +Not sure I like this solution in this case +Closing all resolved tickets from or older +With Rhino from the only way to disable it is toeither remove from the jar orremove srcorgmozillajavascriptoptimizer from the repositoryIf you can reliably see that the optimization is only slow thinks down on JDK or later I will disable it by default in Rhino Cocoon usage is a good testfor this +But dependencies with such scope currently still included into war +In other words encryptiondecryption happens on client side when upper layer applications use Hadoop filesystem API +Both runs passed all tests +Thanks Jacques for suggestion +trunkCommitted revision +The JDBC connection port number should be validated the same way the Admin connection port number is +Is there any progress +Any progress on this issue? I understand that there are workarounds but I would be nice if this got resolved sometime in the near future. +Timestamps are correct when the broker is running under linuxI also think that its quite a significant issue and that it should be resolved asap +Well if you are doing V then indeed the javasource attribute is not supported its something that was added in V +In my case it forces initialize to run twice which is causing the build to fail +I dare say to some extent it could be considered as a bug for the following procedure for spring edition update I never found such problem ever before since I update to the latest stable edition +Would you mind trying these out and commenting here? I also agree it is likely a problem so itll be interesting to see if this resolved it +Cleanup +Thanks Thilinaregards for applying the patch. +will see if i can fix that easy +Im not sure how the user will be able to determine if the remove finished correctly or repair should be run +Patch applied in trunk and fixes branch many thanks. +It seems that the reason why this bug occurs it is because the xhtmltrue causes the html tags in the page to render as XHTML instead of XHTML +If nobody is working on this issue I will apply a patch after gets reviewed +Looks good to me +This is a pdf which has some images which is wrongly rendered after extracted from pdf +If this patch is applied first it will be okay +well there is a lot more to convey than the old text is treated as if it went thru keywordanalyzer term frequencies and positions are omitted length normalization and boosts are disabledThe idea of is to describe that the field is really only useful for matchingnot searching +Please double check with the WSA spec +Arguably the mapreduce dependencies should be removed to make it a more general class but that would be a different bugI havent run this bug but it looks like the right direction for streaming +open a reader writer +The refactor in caused the failure I think +not just in the server branches +Should Guvnor be shown on Server Properties wizard? Look at screenshot plz +In what file have you seen it? Context find founds nothing when I tried to search for +You can use before to make sure it runs first +The issue is or is squashed flat into string before it log message gets to the custom HandlerAppender options we have is Persuade AS to provide such extension Provide a different way to intercept the message object in Teiid that can be extended +Jean I assume this ticket can be resolved now +And details here Justin? Right now the bug report is not complete as I dont know what the temporal filter classes are +It is similar to +Again you might consider this to be fixed in the Simple API project first as ODFDOM will stabalize lower layers first + LGTM +Oh wrote comment that was before I decided to change the default +It is limitation enforced by SQL scripts delivered with package +This applies on top of since one of the test cases fails without it +What I meant to say was is possible the Xalan warning on VC could be related to the build error in VS +If more memory is needed users have to increase the total +Glad there are people like you out there doing it +dims +Not convinced we run them as part of the default jackrabbit build as well +If your application does not explicitly modify the default floating point behavior the best choice here would be to determine if a compiler option exists to change this behavior +on your feedback and our own observations from inspecting the code here is a rough idea of what we would like to doIn the first step we make an interface exposing the public methods of the current class and make the current class an implementation of the interface +It looks like this was already added by somebody prior to +But its a possible workaroundAndrewI dont think a static link is the way to go as Josh is creating another version of the ACL model for and I expect additional ones my be created and the user would want to pick which one to loadCarl +Updates have resolved this issue added test cases for redelivery to verify that messages are redelivered properly +Hi Alan thanks for your response +This only happens in not fixed. +Thanks for your persistence Scott + +StackWhere is the NPE coming in the client side? Could you paste the stack trace to chk it out +I have with drawn this change proposal and patch as its time has passed due to lack of review while the work was funded. +Generally looked good to me +Needed to enter a version for because I couldnt build with maven without it. +I have tested with a svn repo and didnt you have again the issue please reopen the issue. +However that navigation bar on top is missing now +The recent journal is not flushed to begin with they cannot possibly be replayed +However after looking at nested conversations and the thinking behind them Im not sure we need this JIRA given that nested conversations werent meant for editing details but to provide multiple paths for editing data in the original conversation +you have the full log +Moving to next release +Its already there from onwards. +The jsp servilet is provided with default mappings for sutt like +Bouncing to Max instead +so it likely wont be fixed on their the moment Im shelving this issue and hope a more elegant solution presents itself +Diff against release in svn +To clarify MRI prints e abc and we dont with my change +At the moment only supports for WSDL but it should support for WSDL ideally without introducing any new interface +This is why i pointed out that there is point in defining more than one agentWe patched the code locally and now the patched version works as expectedI am not sure how this was working for you beforeConfiguration we defined one group with agents win and linux machines +There must be a method to get all the openclosed issues for a release +When can I expect to find the file in the Maven repositories +This actually ran the newly added test case from the patch +Im using it in a few recent projects and it works +Will go ahead and checkin the changes into this to done +Changes have appeared in Latest Alpha Manuals +New patch after resolving some comments +I am really wondering about that +Yeah that would be a good start +What is the usecase? What is special about your impl +This patch contains the test cases I have written +So I think in my first patch I excluded one file +Modifications to the such that if Archive is a folder then the cactus task will execute the tests against an exploded Ear +I did not create the tag yet +The original issue that created the original issue that introduced the arent going to play regexp games with fields that we are supporting direct UI binding for +Verified +We also need test cases ideally validated against R another package or published results somewhere +Hari are you sure +dll +The problem is even our own themes pull from other themes such that simply replacing the hardcoded xhtml with the dynamic theme name wont work +Do you see any harm with the change I made +was used to verify the result of test running +It does not accept attributes which has only the first character as lowercase +Thanks Jitendra! +The different architecture makes it therefore impossible for to use an XML that can be implemented is a which is technically a of but triggers an from +Please check that it was applied as expected +In this was more an issue but in trunk where liveDocs are completely separated there is no need to check numDocsThe checking of numDocs is also no added safety because readers can have different liveDocs but still same numDocs +I will attach them to this Jira +My bet is that kxml parser does not support encodingsThe new stax based parser should provide much better support for that so you could try with a recent maybe +This addresses a sceanrio when the user who submits the job is different from the user who started the jobtrackertasktracker daemon +on patch +flavio and i talked about this this morning +StackIf you feel this patch is good can we take this into trunk or as you said we need to wait for FB patch +Committed revision w the patch above plus some more javadocs +About the for +For other informational requests like dumpInfo where a comprehensive picture is required the index could be built as needed +bq +this test will always be approximate so not sure if this even deserve a warning unless RSE is enabled +Randall I know I owe you comment on the Soft Schema patch Im getting to it I promise! +Thanks Tom! +Closing resolved Release +Hen fixed in r +However perhaps the fix should be in the lexer instead of +Having generate keys can get expensivePlease elaborate? Im probably naive about a detail but it seems like a simple operation to generate the key register it with the RM which can cache it in its secret managerIf this jira is integrated wont later changing the pb messages introduce rolling upgrade incompatibilities +The same bug affects also the EWP on RHEL both combinations WS Native CXFThe server started up by service start on fresh installation +We simply dont have resources to maintain so many duplicate plugins and instead we should direct our efforts to improve those in Tika +added the test for +Seam gen has default values for these fields we should try and use the same. +This submit includes the fix and fixes several unit tests that t pass a few other affected unit fix proposed in this JIRA reverts the changes made as part of +I removed bufferSize from all the Directory implementations and made it part ofI think it shouldnt even be part of? Ie this is private to the Dir implAnother thing we have to do is add into the cache key used in s inner class +Thanks for the quick commit. +this affects patch includes TIMEDOUT DISCARDED in the where clause for purging coord actions +thanx. +See for further details. +Isnt it possible to give an origin type as param to find? Which you can use for a check if youre going in circles +in the larger refactoring I tried to reuse the from the OSGI code for the other variants +Im resolving this one as we now restored the dependency to on the jboss tools maven features +Was added previously. +I have been working with release engineering on getting RHN setup for this already +The trunk is but since has not been released yet this bug doesnt affect +Attached tests for marshaller +Thanx Mark fixed +I strongly doubt that all platforms are going to allow linking from a temp directory seems like a potential security issue to me +This tweeks the release notes as follows Bumps release version to Documents some additional fixed since +Im using the latest rich build +The verbose ant option is useful for seeing the dependencies ivy is pulling in +cleaned up Htable objects at the end of map or reduce +I have committed this +I havent tested hcat and oozie integration but will add as part ofAll unit and integration tests pass +Dave do you care to prepare another patch having the possibility of supplying shorttime as type and with shorttime no seconds will be encoded or decodedregards dont mind doing it this eveningrgds be greatregards as there has been no feedback I am closing this out for nowIf there is interest again I will reopenregards to Ron Smits for providing this patch. +binary is intended to be used only for data +SPRINGCLEANING We can reopen if necessary +Im sorry but this fix does not seem to resolve the problems I am having +All in all Im not sure we should implement this in core +thanks pradeep for contributing +I guess we could be more explicit about the default name of the parameter but I think its reasonable to expect that people should combine reading the manual with consulting the Javadoc for more details when directed to +Thanks for addressing my review comments and fixing other API calls too +looking for a constructor of the target class that take the source class as the only parameter +It means that a connection will not be aware of temp destinations created to jms there should be an exception cannot publish to a deleted destination if you try and use a temp that has been listeningwatching the advisory messages a connection cannot know without a round trip to the broker and getting back a broker the down side is the exception is different and there is a round trip to the broker to get it +This is wrong. +here is a patch and a simple after release. +A lot of code was removed from translator since the time this bug was filed. +think about how many places catch and use Javas +This issue is more than a year old with no development activity +But what part is lazy now that wasnt +It is NOT the dont reopen issues without taking the time to properly investigate +WillWhich version of JVM are you using +Thanks! +I think the Javadoc issues are bogus given I didnt touch the patch looks goodCan I please ask you to add the same feature for the MR AM itself? We can add a config key and change to call the same api to set the logging options? Thanks +Added in s fix the javadoc in +fieldName auslieferdatum type field number class +Should be fine now +No Camel is being built right now as you can see on dev mailing listIt can be added in the next release +I plan to commit this the conf files needed to be modified for test was there is no +is a focussed script i say we do it in binhbase +thanks for updatehow about ? do you able to reproduce this video quality +Is there an issue up for this type of stuff yet +Since this isnt something that is apparently required by a lot of people and it is trivial to extend the existing class we think it would be better to leave things as they are. +Thanks +Cant reproduce or work on this without test project +Ive added a few lines of comments to your updated patch +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Changed the testing appliances to quoted seems to fix the issue although Im not sure if we should really use the YAML comment for any of our syntax +any update on this issue? +were backported to as well adding the latter fix version +Patch committed. +Jiras that are specific to the seam distribution should be assigned to Marek for triage +Only users in that domain will be allowed to create guest in this zone +Methods are changed to accept Configuration object along with Parameter you provide more details on what you are doing +Looks good to me +Attached patch should bring in the changes made to branch as well as some enhancements made to provide project group editing +Well documentation is in separate but the links under Modeshape Publishing are broken +Theres only a single test method and that fails for me when run through ant + +just realized I uploaded the wrong patch +Attaching a quick patch to recover the lost sum and avoid negative cost estimates +yeah that looks good +Attaching a patch that cleanly applies to +The javancss POM does not contain the dependencies required +Please Verify +If yes then that bug is already fixed +Im working on a new patch +But for user point of view I prefer to keep those parameter blank then s +defer to +Please follow JIRA procedures for raising issues and create a +alt requested this be a blocker so I have reinstated it as discuss scheduling with changes are present in ER build +All tests passed successfullyId really like it if this could get checked in to the code base soon +I dont have time to check right now but the most likely thing is that the shapefile is not compliant with the specification and qgis is more lenient and manages to read them anywaysAnother option is that qgis ignores the deletion marks in a shapefile that has been edited with Yet another option is that we have a bug +will have a fix +The patch fixed the leading space problem +This is an important security problem for us +We really need small and focused patches in order to apply them quickly and easily +Id just commit and see how it works out + +already has setget added as part of didnt realize it originally bc I was using rather than +Committed patch to documentation trunk at revision +And neededReplications will have the block with older genstamp still +This solution only provides a more meaningful error message than a +Once this module is more fully fleshed out we can release it but as it is now theres not enough in there to justify a GA said if you think youd like to help fill in the blanks I welcome you to fork the project add API bindings then submit a pull request +Fix applied at repo revision rPlease verify. +Close it if you wish. +Version released. +In most cases it corresponds to a line but as you noticed it would be more accurate to call it a recordNumber or itemNumber +Reassign to a release that you feel comfortable resolving this issue in or leave as is +However it doesnt apply on current trunk any moreHeres a refresh +RHDS must show warning of any sort that tell me that file was not setted +So there is definitely a race condition +I had updated the to use but forgot to set testRun and test. +DBRE was my baby and Im up for the challenge +Attaching a patch that makes the three incorrect callers use the helper method instead +Here are some sample startup times on my intel core duo machine Ubuntu version +In terms of managing the user logs in the task tracker there would be settings +I wouldnt know the other classpath + +For the moment I rely on +The fact that it is not declared in a pom does not mean there is no default version +Hi Bowen are you actively looking into this issueIm kinda interested but would only watch it if you are +Tom Im guessing you want this issue assigned to you since you haveare working on the new http gateway +I think we are almost in the end of the initial development +Of course Im absolutely ok for documenting more precisely that this is an assumption ALTER TABLE DROP does but what I mean here is that we should probably be clear that if you start using CQL youre expected to use microseconds timestamps +Im closing the bug because it was originated by configuration error +Remaining javadoc warnings are not related +or submit the body of a test case that shows the problemregards tom +You seem to have downloaded the source distribution +The client throws an exception to the application when the write pipeline falls below the minimal replication factor + +Just committed this to trunk and +Returns bash cordovarun No such file or directoryRegardsSteve Husting +Please explain how this is a bug +Was fixed for M +sorted out by the refactor of the project builder +I dont think there is such a thing as solrj contrib +Andrea I have attached the amended test in that should be committed with your patch +The IBM Developer Cloud adapter with full functionality and tests +Thanks linux patch was applied as expected. +Ive tried the version and placed the alone in an arbitrary directory and it was able to run the addin +committed +The formatting in is a bit backwardsDont think this justifies a reopen +Excellent so this is an important optimization! Ill commit shortly +I have a unit tests that +Bug has been marked as a duplicate of this bug +It happensSo if thread B is just before waiting for MVCC to complete and by the time of thread A does the check operation it is going to succeedNow the MVCC completes and the put as part of check gets completed thus overwriting what the thread B has written +Theres another issue that deals with the same problem so might be that this one becomes a WONTFIX +The patch for classlib code as well as slightly modified version of patch for test were applied at the revision +minor is not applying any longer + +I attach sonar cismavenmavenbinmvn help cismavenmavenbinmvn e f jreposdevelpluginworkspace sonarsonar sonarcismavenmavenbinmvn e f cisjenkinsjobsdevelpluginworkspace sonarsonar sonar David but we really dont understand your use case +Teiid exceptions generally contain detailed messages and a detailed exception chain +This should help in catching any accidental introduction of code which calls logSync in critical sections +If there is a value in label field JSF message consider the label value instead of the ugly form id +Added test for key object being serializable. +Looks good to all issues that have been in resolved state for more than one month without further comments to closed status +Updated to remove transient applied +added new tests to under review suite thanks because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Ok Oleg its not a we didnt understand each other +SorryDescription is incorrect +I will extract to a method and update the patch +thank Arup for follow upActually I tried before Ashutoshs comment and it worked wellAshutoshs comment confirmed my experiment and also told me another choice of alpha +With Im not sure we still need this ticket +Thanks +The Surround tests still pass with the recent fix in +Still a small issue here in reload we try to maintain the link of a orig name mapping but if the core was dynamically created the orig name will be null and so we must check for that +There are duplicate in each clustering cases +args you are right other suggestion +manually deleting target and running mvn again worked +The patch is rooted at xdocs directory of the site +So we will know this by through the and metrics +RAT check is green +OoopsAnyway I fixed it and I have also fixed the to invoke a instead of invoking theI have also checked in the unit test from +but its interesting your separate tool couldnt repro the problem +If one directory contains an image which has already beenmerged with the edits and another has the image and the edits in the conditions then the first edits file is empty while the second is not but the storage directories in fact define the same namespaces that is they are equivalent +Assigning to Rob so he knows its here +What it should be counting as strange characters are ones that you DONT expect to see in text ever +What about mapreduce? Do we need it right away? Any tracking issue that is already filed +Was fixed a long time ago. +Andy could you try out this patch with your installation + +All of our test development happens upstream I just scheduled an ER build milestone because thats when we would be able to get to it +In particular the parameters should populate the parameters map in the and OGNL should be used for the Action population rather than by custom means +closed +Hello Latest patch is uploaded according to new changes into Lead and Contact like find result form is now same as both Contact and Lead also removed viewAccount from controller and from commonScreens and fromRegardsSantosh Malviya +Will run the trunk patch through hudson as soon as the build machines are up +Things are definitely much better +I introduced which defines how many calls per handler are allowed in the queue +As an alternative heres my after the patch is applied +That linepserveruserserverhomecvs A has only elements so the password is returned as null +changed priority to version is updated to in Railo s great +I also add the regression tests +created a group added all the people you mentioned and granted that group all privileges to content in the climate space +A mistake on my part which has now been fixed +Yes these two tests are supposed to enable compression but they are not +Hmm +Small respective changes in and +The patch was originally against but it needed rebasing after was committed +Ive created an attachment containing the test script +Ie can it be reproduced with trunk and not with apache trunkthe reason I ask is cause we have been selectively merging to the branch and we may have missed something in which case the problem can be fixed by reapplying fixes and applying more we need to work on getting this hour test case down to something manageable once we know on which code base there is a failure +Oh OK +Bulk close resolved issues +so seems the only solution for now to avoid the regression is to make the filter parsing a bit more lax which i have done +Since derbyall doesnt touch JMX I didnt run that +However I took your ldif file added a Unit test declared an administrative point and it seems to work at first sightCan you provie the full Ldif file for complementary tests +But how would the know which to launch for a given requestA more elegant solution would be to use URL patterns to map different requests to different +We might have some in the future depending on how we build HA awareness into the services +Assigned Fix Version +len is going to try and reproduce the original jbpm problem +We use TOF also +I forgot to mention in my last comment that the patch which is enclosed fixes the issue +I mean please give me some time to first fix which will be committed back to +Thanks +Thanks Prasanth! +Ben I do see it only in Tapestry now +IE provide an API that will take two configs do the split and merge between them +My thought would be that the TRS would buffer internally and use the zlib transport to fill a sub block then the zlib transport +Of course by default localhost is whitelisted +Another thing to consider for should we silently skip any updatedfields or should we somehow alert the user +I connected the debugger to B and C and paused all threads to make sure B and C dont get a chance to log the request before they crash +This message remaining in the queue is also redelivered the next time a session is established +Hairong I dont see this in the branch +We wont do this +It currently just includes the first placeholder that closed the cycle which should be enough in most cases to find the problem +that to the best of my knowledge is not going to be fixed Mojarra has been chosen as the JSF implementation in AS +Can you explain the use case a bit more? I dont quite follow because create is only called during a refresh and your clients dont have a refresh token if I understand correctly +bq +For this purpose we need ALL extensions known to be related to the mime type. +patch looks good +This should have been fixed in which included updated native support for Solaris +This patch also contains a partial patch for since it depends on removing the use of the LARGEINT library +In our project we separated the reports and the fixtures and I never tried to do will integrate the functionality and surely your path +Thanks Aaronbq +LenKoen im not following how this is related to tools? Can you explain meSounds like a runtime side deployment thingOr possibly you have added something else to the server deployment that uses hsqldb +The is supported by right +Not to say setting response parameters +My motivation getting clang libc compatibility +raccepted the DRLVM patch and DRLVM test patchIm asking the submitters to open a new JIRA with the classlib specific test change proposal +No problem +Also on to Microsoft and let them know how you are to their software and blame you for the issue in your software +Geoff is not happy with this solution but there it is +Block deletion occurs in parallel on datanodes from multiple threads this put enough load on the datanodes storage device that writes to other blocks on those peices of storage became very slow +At best we have time to check actually works with one Tomcat edition +I refactored and added validation options to specify validcard types +uhtml resolved that because you had to request to get the raw version and Goober to get the rendered fact the pain isnt easily resolvable +It is always good to throw away couple hundred lines of dead code +Looks good to me +This change was tested by Weston in an app server environment +After trying several things today Im unable to figure out how to address it without polluting the code with a ton of casesA better solution is going to be to create a different class or method that handles peculiarities +Logged In YES useridNeed more information about what is deployed +Ive fixed the bug junit test now passes successfully the latest version which will be uploaded soon should thus eliminate the problemCould you please give it a test in your environment and see if it works for you? +Hmmm +This one is performance issue +problem still exists +Due to this reason the is not in the finally generated site +It seems had trouble running on the build machine +Temporarily reopening to update release note info. +Thanks +Chris thanks a lot for your reviewThis patch fixed comments For comment I decided to keep add +So if the underlying logging implementation cannot configure itself correctly as I understand is your situation here cant do much about it +I presume youd just take server code +modified the test to have more fragments instead of single one so that we can ensure last only one last unclosed fragment failed + +Its the same exception but coused by different WST thread +Also added some more configurability and integrated various minor feedback +Patch committed +Thanks +Everythings fineThanks. +Last nights commit of HQL made it so this patch no longer applies +There is a step by step to make run this command with those versions of Ruby and? +Patch committed to trunk +null +Since behaviour is consistent Issue is resolved +Thanks Gaz +The fix is just add that rule to the condition inside +Please close this issue thanks +Sorry I got confused on the way this should be done +Some of the questions I am left with is Post compression the block files have completely different length +And it still gives the random validation results +It looks fine and buildsfine for me +Jason got my account setup on codehaus and I committed this patch to SVN. +I have added in the Methods of the Classthe calling Threadnamejavacorexxx the javacore file of the deadlock and commit is indeed called using different threads +The Algebraic stuff is still whacked some sort of weird memory thing is going on not sure what or why +The uses the first element to keep track of a comparator otherwise it could not differentiate from the first element in the map +Please indicate if not +Sure its mostly simple changes but its still that many changes that could introduce a typo somewhere that causes a bugThe same thing I can say about your set of patches its KB +I have some ideas about how an internal pipeline should operate but its always good to have a practical implementation available +Perhaps your implementation is forgetting to filter substituted exportsTechnically at this point in the resolve there is only one viable candidate for which is the api bundle +Marking this as blocker for since for this release we just deprecate there will be classes depending on correct behavior of it +Ill take a look today +It converts first character of nay give property in to upper case so I have written a method to convert that back +could you provide a patch with this change? thank you +fixed in trunk and in in +Would it be enough to just use sed to fix that +Ivan Gregory thank you very much starts perfectly nowClosing the issue. +It might be a good idea to scan quickly all the places where the getDT api is called to verify null value is expected there +This was resolved by our normal manual restart procedure where we stop thrift gossip flush the node drain the node then restart cassandra but it definitely caused some confusion for nodetool status and nodetool info to report that the node was up and a working part of the cluster when in fact it wasntThe nodes in this state definitely do not make it clear that they are not part of the cluster anymore. +Also you do not need anything else to get it to work since I set the defaults to work for you nowIll see if I can write some more tests +Looks goodThanks Omkar! I just committed this to trunk and. +Additional javadoc comments +Committed. + the test now is not in exclude list +dims +Thanks +session to use socket endpoint to sendrecv with memory issues and try to pass hybrid with m thinking about the similar alternative implementation in to remove the use of corbaTom do you have any idea about point +Adds which will run first the lite tests and then the full embedded and client +Thanks +a brief start at this one +The new language is better but if I put it under a microscope there are stillblemishes so lets visit this in +Web Services User GuideDid not find any info specific for JBT or guide is fully universal +catch Erik +Thanks for the feedback +Prasana Do you think that this might also require specific output values to be selected as default responses for the API calls? My concern with simply tabularizing the return values would be the line and field lengths +bq +I just checked and it looks ok +Committed to the branch +This was determined to be currently unworkable +coat done doneRemainig contracts are DoneTo DoSee previous comment about deletions still to do +keep the config the same +After some tries Im able to reproduce it in production +Closing this issue since WFK docs have been for release. +I am not being very helpful here +Hope for the bestWhen the patch is successfully applied go to the modulesframework folder and trymvn clean installthen step into and domvn clean jettyUse you browser to go to localhost and you can see the web app +Committed the set of changes necessary to use the fix properly when it becomes available +We had a thread that needed to have its own ephemerals so we let it create its own connectionsession +Normal pipeline test case aspects are developed some test cases are doing what they suppose to doI need to find out what datanode successful state means and how to correctly verify if a packet has been placed into a disk file or an internal buffer +Ok looks like youre getting legitimate build errors now. +i assume not for is not actual for now +Hi LarsUnfortunately I dont think we have anyone maintaining the weblogic implementation +with startcode and server INFO Deleting region contentcedccabfeadfcfc because daughter splits no longer hold referencesFor contentcceaaddcbdfe INFO assigning region contentcceaaddcbdfe to server INFO Received MSGREPORTOPEN contentcceaaddcbdfe from INFO contentcceaaddcbdfe open on INFO updating row contentcceaaddcbdfe in region +Also we need to move the change description of this jira to incompatible section inDoes this not cause issues at Yahoo +I had thought that the point of the was to be used to mitigate some of this concern +Heres a sample resource and JAXB class +Should be fixed now +Is this still an issue in + Completed waiting for review +Thoughts +How do these tests pass under AntMaybe this is due to some libraries Ant itself is including in the classpath of the tests runningIll go ahead and this issue so it is as something we should resolve before final releasePerhaps we can include a scripting implementation in Solr at least for testing purposes but maybe also to ship with to ensure this works out of the box on all +your help is much appreciated Bruno +Any modification to the core JDBC class has to be vetted by all three of us +Unfortunately I cant share the heap dumpA server generally has a lot of String objects +An absolute path from the webapp root is required for the location of the tiles definitions config files +Is this already captured in documentation? If not check back with Kevin and he will help write something upCandidate for SOA is a nice to have for SOA +Started looking at this but I am having some trouble with the class not compiling under jdk +Since the code is in a prototype stage can we move the index code to contribMove to contrib is not good +So lets discuss on the forum reference Thanks +You can already achieve this with although it requires using three methods instead +Committed to trunk. +This way when we enable the user will be able to configure all of its close for +The Row modifications are made for our internal branch since we didnt have the lastest ones +This will allow me to continue work on +The patch adds a call to update the isolation level in the when joiningresuming a transaction +with null regioninfos in it by splitting a big table then doing shutdown in middle of splits +Dear all super experts I am only a newbie in JBOSS +Using correct jars works ok. +Skin was Apple iPhone +Yeah those files + see the testcases I created +Id like to check it before release +Temporarily reopening to update release note info. +For Nutch which uses Lucenes trunk version youll want to build the Highlighter from scratch against that version of Lucene +However the markup stylesheets and functions in the original Smoothfeather theme required a significant overhaul in order to give it the best foundation to be flexible enough to support all browsers +Thanks Gera! +Line breaks are hardcoded into the help text +It points always to httplocalhost +Approach sounds good but would that create any instabilities in hadoop +Therefore I set a similar property +Sounds like a plan +uses which uses +cool +I cannot get Cobertura to exclude or ignore +This should now work as Seam now allows you to next parameters in JSF to a message. +Can not reproduce +Attachment has been added with description Make htmldoc work in headless is now in SVNThanks! +Works well Evgeny thanks! +fixes +That seems to be sufficient for the issue. +However consideringthat the really does nothing different than the command lineMaven I see no reason to believe the it wont work +Changing description to my new understanding +has been commented based on this when fixed it +Is something else causing it? Should I do something else +There are lines longer than characters already but Id like to avoid introducing new onesThe patch looks good to me otherwise +Attaching a second rev of the functional spec incorporating feedback from Kim and Knut +Moving to in the hope that the Tools project will be moved out of Tools repository in the next few days +In any case it is really important that when an instance that is active dies another instance will take over +I see no reply on Marks question so Ill close this as incomplete hoping its indeed a duplicate. +A more complete fix might be as simple as extending the change to the handle method +There will be some bug there +The report mojo does not call cobertura if the doesnt exist and paths are always fine also in case of a build +This issue has been fixed and released as part of release +Its my other developer and at least one person on very very hard to nail down because you dont realise its happened until some time after +Just modifying the patch in thats right + +Unzip content and place in sample directory of Cxf binary installation +Original methods are now deprecated +This is not part of the JCR spec and will be possible with the advanced query model of JCR in a Jackrabbit version +Resolved +at least to investigate. +Thanks for the patch Sravya! +but lets see +Aha found a route around the issues +We now have regular releases and a continuous integration process +Moving to +Stuart I think your approach is correct but we should offer this programatically as a method on Manager more like Srinivasan has doing this through a timeout isnt correct otherwise I think what you have is ok remove that stuff and Ill take a look again +Patch implementing the required committed as is. +There is already the issue reported against it many months ago but it looks like abandoned +Let me know when you are ready and Ill turn off the rsync stuff in my crontab +seems fine please add some comments on the top of the file about FLUMEPIDDIR MASTERPID NODEPID and FLUMEIDENTSTRINGThanks Jon +Marking this as blocker as we need to investigate if the performance regression was introduced in Infinispan or it was present in as well +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +The implementation is not fully based on the Sonar Runner API as it requires a to create the project definition under the hood +Isnt it alto el foc +The Stripes architecture is clean Java all the way and it simplifies out a lot of crap we have currently in the system thanks to using just JSP +At first I was also thought the same thing is Mathias but now I am seems that Peter used forms in the menu layouting and I dont think this is a bad thingLets face it is hell to work apparently need layouting for even the basic tasks of our rich client why not just use Forms instead of torturing ourselfs withI ve seen the use of some in our code at some points and it wasnt understandable if this implies an extra dependency on our users +arable land +Thanks a lotBest regards by Richard. +Also those private classes are quite a bit in use it seems +This property is handled by you ask for refactoring its not a bug +Do not know if it is still the same thing in CXF +Once the FAQ is completedBTW We might need to order these question in a suitable manner +Magic +Any suggestion is welcome +will respin release branch to pick up fix. +This certainly hasnt been fixed and if its a cool feature we want at some point why are we resolving it at all? Just leave it open at a lower priority +Patch with fix and test +In non delete only workload it will have minimal effectIts important to have non zero count otherwise it will not be flushed to disk on memory pressure +Attached patch +If we move to then they would go +This works on windows too and I dont need to modify the +In some cases the tables reported arent even the same for example Foreign key circularity dependency involving the following tables calendarappointment calendaradjustment calendarrecurrenceWith your latest changes the issue has gone away for me Emmanuel. +I initially hadnt thought about the issue of cloning too many index inputs +So we will be able make Triggers execute on behalf of the trigger creator instead of the active user +Just merged in trunk. +all are closed so closing the main issue +the problem is definitely mkjk +if it is any solution Please +Closing issue Oozie is released +Either ways fine +Fixed with and +There were some problems with v +Not sure when we integrate it +After thumbing a bit through the code somehow and this is abit too complex for me to venture a guess on wed need toprevent the deletion of the SRs files were deleting from evenif that SR is no longer live +The is a file used by the sreutil file +So later when we tried to search our database for these particular names we failed to find this information +We can use a single thread for all search time limit based cancelations +Thanks again +This patch removed a method from +But at this point Id like to get Santhoshs feedback on your changes before proceeding as he had comments on your earlier patch and I want to make sure your new patch addresses them +Josh sorry took me a while to get to thisThanks for the patchCould we just add a check to ala? That way we can keep a single method in Scheduler +If so then it performs the this open for now since tests are lacking. +Dirk if and when you want to this feature feel free after we get to an Alpha build +is too old +Keep up the good work! +needs to be for Hive rebased with latest trunk +Updated patch from Arun +Catherine Hope has volunteered to mentor this project +We should try this out and find out how this affects users with incorrect applications. +Thanks Daryn +Request Certificate and CRL cause NPE +Logged In YES useridDear ScottHave you received the testcase +When doing the exact same thing via I also ran into a timeoutAnd once I refresh the browser the application is present butOnce you add a mongodb the app becomes functionalIt looks very much to me as if the initial timeout when creating the app is caused by the missing mongodb since afaik the initial create is also launching the app +This was run on a pre designer which uses old SQL ExplorerTeiid Designer uses DTPs SQL Results view +Additionally the change to the last message conflicts with the messages we output in various other places +Thanks +This issue has been fixed and released as part of release +Please dont create bugs to ask questions +I would feel more comfortable in trying to embed the eclipse compiler +These relationships are specified within Spring IDE via the notion of config sets +Duplicate of +I think you worked on the kill side of things +Since there has been no additional action and we are out of time on EAP I am resolving this as deferred. +Will this be ported to? I mean Derby should not crash when updating several thoundsands of records should it +The wrappers expose the new in the component and delegate to the corresponding tree model. +Sorry that should have ended withSame as the other examples just decompress to the Beta examplesjms directory +That way the current attribute would simply be used to reference the composite and the composite itself would be defined as a plain bean definition +however if someone does change the runtimes configuration or home directory the cache will be old and Im not sure theres a good way to verify that +Committed to trunk and +archiva on are in desperate need to have this supported now so I will start today and give it a try to work on it +Thanks for looking in to my patch +Yes already passing alias in the createCriteria +Im running a vanilla AS and I can deploy undeploy deploy and undeploy without incident running the app at httplocalhost once for each ll attach +thanks Stack +Are you sure you are using Camel? I checked and we have type converters which convert from File Source which is used by Saxon +Id be glad to help if you need or otherwise assist in closing the issue +Waiting two minutes to clobber its log wont make a difference the log will be just as corrupt +Thats ok +I checked the stubs from intellij and it just leaves out the annotations so that may be whyThis is killing us +Guillaumethanks for your help and your test setup +Patch for first cut of this new wizard +Thanks for all the effort. +It appears we filter the URL and https was not in the filter +Steven for now please attach as a patch or put into a private svn repo and email me with your JIRA username and I will get you started signing our contributor agreement +I wont likely be able to set up a harness to test with for a few days +I have changed all the exception to handle this was not the case already and done the other thingsmentioned on this bug should now print a warning if the reconnect throwsa or it will continue to retry +This is what I did mvn PASSPHRASE RELEASE clean installThat filled a local directory tree with artifacts signatures and checksums +I missed that +Is that intentional +build now includes the hostname fix so I was able to adjust the tooling code. +Attachment has been added with description The patch as has been added with description An fo file that exercises the patch +Marking patch available +The fix for this is trivial but it introduces the question of when to upgrade to Axis and how to ensure backward compatibility +Thanks will do and let you know. +Thanks for tracking this down +I dont think its working +Closing. +with recommended solution being to recompile the +Thanks for the review stack +Anyway I couldnt get any error message with IE everything works fine running the project with Jetty +Verified fix is included in EAP ER. +Two months later still nobody cares +Ill take a look tonight and see if I can figure out what is causing this. +Hi Tim I have verified +get to get working copy and work with it +not bind the multicast socket to specific interface or bind to the interface which multicast route is set through +Ok I think I fixed it properly let me know if you find anything else wrong +Hey this last comment gave me the idea to restart the Eclipse IDE +s scoring we should just doc that the calculation is based on code units and leave it alone +Yes Richard I also confirm it to be working with OBR so I am closing this bug +I guess we will have to change that back to allow the proxying of the function from the outside with custom code +Patch applied +First stab at trunk patch +hope thats fine with you +has some odd changes to whitespace +Works for me on IE with +Keep an eye on the dev list for more information +bq +I implemented this but only for +This issues will be thought over again +Is v lib something that got left out +test cases that test this issue +Good idea +Removed and also thanks for catching this +Patch file +I believe this should be closed or moved to +Im seeing this issue in +We also have transitive dependency resolution disabled for the compile task so the the listed dependencies actually are those which would be required to run javac against the code +If there is any other better approach please feel free to let me know +We should state that such reserved memoryoffset is for system usage like for OS system and hadoop daemons themselves so that its clear for the admins +Attaching patch for incorporating review comments +especially the code complexity is really high and performance isnt good +If flag is true will shuffle the splits that share a common size +John Should we move this out to or can we get this done for +lets open a new issue if the need for a real fix arises +youre right this should have also been fixed as part ofIll fix this and check I did not forget other places where management resources are used +thanks +You mean all attributes or I guess only multi? Then are we sure the list in ignoreAttrs is complete? What about custom attributes an user could use? +Sorry if this spams things however its unlikely that Ill work on these. +I hope you did not use it! But in any case you are right that I should not have simply removed it +The driver is required for dbre while the shell is running +I vote for halting deployment if possible. +attaching test to reproduce the antonio for spotting this issue +Al can you verify it +Fixed for both and simply removing the assertionJuergen +Could you add a testcase against the current trunk? An updated file documenting your change would also be great +However we can hold SSH a little bitfor the user region assignments +Attached an sql script to create the test DB create the tables add the data and run the failing query +bq +Please see my attached patch +the main reason for the comment is to identify it was generated by hibernate tools and what version since I kept getting questions from people that were not able to identify what files were generated by htools nor what version +Feel free to look into that +Things have been significantly rewritten in the child processes handling +You need to remove your simple project and reinstall it again using Project Examples +Also I agree with harmonising with the ZKFC +At least to me +Latest and greatest +Still not done any changes to the fix +merged into latest cancel waiting for c code from patch implements the c and includes the java part as well +Basically alice requests the role administrator to be part of the issued saml token. +But is the one decided by the masterEvery time Assignment starts in RIT goes thro a set of steps and may be that is why we are not sure as in what step the RIT is and who made that changeWe also tried thought of one approach like can we remove the retry logic itself in assign? Always use SSH to trigger assignment as in TRUNK the expiry of server is very much spontaneous now +Closing as Duplicate. +AaronYou are is a dupe of +Filed JIRA for support +Fixed license header. +Thanks Stack +What am I missing? +The updated specification reflects the semantic change to the permission setting of the automated created directories +I dont know if Julians patch is a complete solution +The has similar issue +Thanks +Verified on. +not blocking does this mean the JEE bundle is missing? Or that something in JBT which depends on it is not declaring that dependency +Corrected to closed. +Closing issues which have been released. +any idea of when will be released +I did not reject null rejecting null was never intended. +I dont think this is a good idea +I think I know what the intention of this project idea was +Three provider problems in DroolsGuvnor current version of Drools in JBDSJBT still has the wrong providers set on at least three Drools plugins perhaps ll attach an updated screenshot +You can look at it when I update the diff +Thanks a lot TimOn which display and which screen is xvfb running? The default display and screen ? should invoke Xvfb as part of your build with the parameters you need for your tests +I may be misunderstanding the context though +this is sets the correct CSS RenYour patch is in trunk at r +The attached PDF looks great +Sorry Paul Ive mistakenly assigned this issue to you +I obtained the same results after about requests +cleanup +Committed this to trunk and +Now its neccessary to check that includes only those pools in which user is a member +Im not convinced that these changes are the best way to do this I just wired some things together and made some changes after testing +Devaraj just did it. +Thank you very muchBest regards by Andrew. +Will it fix your problemthanks and test cases has most likely been fixed by the patch that I submitted for +unit tests +Tedsorry this is my first time to submit a patch +Can actually read in a block +You will need to get a project despot approve your pending application using Xircles +Any objection +Proxying overloaded methods doesnt work then of course +Manually tested! +Closing +Treat unscheduled issues as Native SOAP stack is not supporting JAXWS any more +Yes the addition solved the issue +another possible workaround for people running into this problem would be to try offline compressing the table +Attach a new patch including error messages. +We can continue to use and perhaps throw an exception if is not an even multiple of this +To no avail means theyre different +It would be cool +for the attachment +How can you ignore that +My suspicion is that if we returned null instead of an empty list the numbers would be different since a null check should be much cheaper +We should also update the javadoc for set and get about the new option for rendering both text and imagesDone +At this point work on is stopped except for Blocker issues +Also we may just want to have a report for each of the modules +Oh ok wasnt aware of that so moving I was pretty sneaky and subtle with how I mentioned the transitional nature of the patch right there in the issue description +Bulk closing stale resolved issues +In the meantime lets hear what folks have to say + Username andieveritt +Committed thanks. +Its not a deep kind of XML Schema processing you dont process xsdimports or xsdincludes or anything else like alternative would be using something like XSOM to build and analyze the XML Schema object model +May as well +I guess to keep it short its very important that you be able to scope the start limit to the step but probably a bit overkill for it to be on the job as well +This should be resolved with my patch to I think it is +I plan to commit this tomorrow +yeah we will commit some unit tests performance benchmarking result +This should get right in +The space increase is still likely to happen if a file has a rep of but is requested by hosts the block still ends up +Verified for closure by Julie ZhuoClosing on her behalf as she yet has no rights in Jira. +The Replication Tests do this too +Andrew Im down w your explaination of explicit row rather than explicit region and on your last png +Very but probably not the most API friendly +Patch to allow multiple digits in version components +Wouldnt ensure that +I dont have a test environment set up on Tomcat so please reopen if this doesnt address the issue correctly +Apply this patch to TRUNK build and then run ant test Dtestcase to see the problem +binhbase shell would put up the new shell not HQL +Patch has been issue open until it has been committed. +I committed the patch to and +This looks like a good approach +Much has changed since then +Then I executed a VDB +There are two copies of this documentFORRESTHOMEcontentxdocsdtdxandFORRESTHOMEmainsrcdocumentationcontentxdocssamples left this issue open for a while in case there are some more elements to demonstrate. +Cant do some of the mvn processing +All stuff was backported by me +I found a simple workaround +Im working on adding that and the ability to configure properties right on the Logger +Hi PaulThanks to take set as a reference +I do not call an url +This issue will be useful for hdfs improvement for +Hmm that page doesnt have any content +Assigning to Laura as she is the new lead for EAP docs +In addition we loose all options to selectively grant access to different language trees to different peopleThus unless being convinced with some other arguments I would not oppose this change +Actually I am trying understand what exactly is being proposed here +Done +Committed also this to Federation Branch +We may have to wait for the MS team to produce this but if they need help we should try to give it +I somehow missed to add the junit jar to the classpath as it worked for me locally without itI added the changes in revision added another missing junit dependency and fixed some copypaste issues in revision +Ive added a patch that extracts some more metadata fields in the +Tests and FAILED before and after the patch +It cant because Forrest is not running while the plugins are being installed although it could be since they are not actually mounted until the first request that uses them +Im going to verify this under Linux bit +See is not good reason +It was definately a problem with expression optimization +another is stored in its place in +Awesome +Posting what I have here which seems to be complete but Im having a hard time testing it +patch simply decrefs new if the executor rejects the patch is slightly stale what I will actually commit also propagates the exception right after decref throw e +Since this patch is so damn large already I wont be making anymore changes to this other than criticalblocker bug fixes +Tiziano would you be interested in submitting a patch for those issues +Okay the problem here is the resize method +Erik does that work for you +Its not likely that there will be any further releases from the branch +Ive configured the jbott bot to push to channel and meeting nowI would consider this issue resolved now and other specifics could be opened after it? +Node found node A in DB +Now we move all the variables to Environment for all the containers +We can ignore XML schema refs for the moment Im simply talking about refs to internal JPL SVN urls +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Touching the file every minute or so also makes identification of an orphan very easyOther systems that use a similar approach include mySQL Solr and mongodb +Now the test fails because it asserts that all DIRTY objects should have been checked and included in the exception raised but JPOX stops at the first failed object +Wow this was fastI didnt mention that this problem of non logging fault messages was present both on client and server side +Thanks for your work Im trying to use it right now +Rows that arent interesting are not added to the set to be reduced +If you then backspace and type A it will not the server but will only display the results for querya because it doesnt notice the case m moving this issue back as a bug fix for instead of working on it for +BrettCan you provide more details on what you mean by Dependency Closure Reports +jar from buildivylib is going to pull in all the jars from javadoc releaseaudit and jdiff which apparently have conflicting dependencies which cause the JT to crash and the tests to time outAfter removing this we might be missing some dependencies which well have to add to the specific projects files but thats more correct anyhow +I can say more now +Committed to after fixing a bug and a typo. +However if another moduleA depends on the rebuffed jars at least moduleA can run tests against these jars to ensure basic correctness +I am not sure CAS could do it perfectly +DNS entry added +Jeromy thanks for committing the changes to get us closer on the validation +Then get that property and set to the request. +Although I thought I had handled this particular combination +You seem to be good at finding my errorsoversights +Bulk close for +Putting this forward for as its required for good integration with other specs +Screenshot of all template locations loading script resources with a relative path to use the absolute path provided by the appRoot template variable in Rev +Ive filed +closing as wont fix +Is it really necessary to have them? I cant tell where they are from the top of my head but if it is necessary Ill look for it tomorrow +attach my patch +We have been building a system using jBPM with the plan to cut over to Activiti at the point where this feature becomes available +Which is a facade for which can be used as a facade for another logging framework which can be used as a facade for yet another framework whichCool +Ive fixed it so that it applies to the new directory structure +Go ahead. +I tested your code and its nice for for your fix +Would recommend Raw Comparators for guess Comparable is only required for Key Field which is as of know +Solving this issue also need to take into account the current duplication between groupId springframework and +Sorry I have not looked at the original jira +I was probably supposed to run the build with U option +I think keeping it on top is the right thing +What runtimes did you have in your JBDS workspace prior to the upgrade +has always calculates the tree permissions even if no child exists with the specified name and can IMO be optimized very easily +Maybe that was it I was sure I had clicked WMS Layer in QGIS before but I tried again and seems to be working correctly and not showing the geometryless layers +When Go Back to File View hyperlink is clicked this is the result +Fine to do that +No there is no redirect happening in the example I provided a small file is served +I still opt for letting the user set it explicitely +Alan if I dont think this is going to be that problematicEven if I try to pass in a map dereference with an integer such as mymap would pig automagically convert the to a string equivalent to mymap +Thanks Daryn and Suresh for the reviews +I will test it and confirm +we have completely all the documentation +This was my motivation for moving the authentication related code out of to +I tested it on this and and I saw that all the Google TV apps that use Cordova do the same thing youre doing more or less +Hi Xuefu The patch currently just keeps the params as part of the type string which means no metastore changes but could be brittle in other situations +We currently print out the message of the exception that wraps +Thanks issue is closed now +pretty significant improvement in CPU +Hi Amir release candidate was put to vote on Nov and passed +Im not sure if we need to redo it or not but at the least the current interface needs to be extended to add a column qualifier check +I added a better message there is probably still more work to do than this +Some comments The words absence and set are being used frequently in relation to configuration parameters +Can a brother get a +Those are not reflected in the currently released versions +Now the connect attempts timeout after around four minutesTo me this indicates that the backlog is effective up to around k connections but that connections queued this way will timeout after four minutesI also tried Windows and also found that it failed after a few hundred connections and could find no way to tweak OS settings to improve this +These fixes will now make rc +it is the same as the issue in and has been fixed in the latest SVN. +Moved from to just after merge of old branch into TRUNK. +Why are you running off SNAPSHOT instead of tag +Committed as r. +Alex just to confirm these tests passed with the GA bits and are failing with the CP bits +I only tested this with a simple client app +Single integration tests work as expected with . +Even if all the worker threads were busy still the Runnables can be queued +For all rowsMay be we can just check if there is a +Security is not testable with current test framework so this issue is posponed for nowThis chapter contains few assertions that shoud be tested later SPEC Testabletrue Section values that the Portlet API get and get methods return the same values returned by the equivalent methods of the servlet response objectNOT TESTABLE WITH CURRENT FRAMEWORK SPEC Testabletrue Section is method expects a string parameter with the +Generates the file +This is exactly how I have implemented the LRU +So you get both body headers from MQTT mapped to the Camel Message +Resolved and committed. +Stack you are right +Brandon other than that I this patch +This one should +Is there anything blocking this from being included in Hive? Is there anything I can do to help +Suggest resolving this issue +Patch to fix patched too soon +cannot reproduce on Jboss AS to set proper fix to issues that were set to +I stopped and restarted name node couple times +Martin so this bug was just about us using over ? it does not change anything on the servers actual execution +Seems to be okay now if I click on it it tries to open in Paint +Example of differences +If not I dont see how it couldBut if is released I will try and get the macport updated and see what results I get from it +Additionally for the groupId for cobertura has changed from cobertura to +ant works correctly +I didnt realize the ASPATCH CP releases were on demand +We probably should fix this in +Have a look at of Date +removed deprecated in added option to JMX Agent to only register processors if they have a custom id setWiki page updated with the option nameThis only applies for Camel trunk +followup so closing. +Since we need the default to be production it will always look for production +And Yonik if youre argument is bc Solr uses it I will change it +One can see this network failure by unplugging the ethernetIt would be nice if the build ends up in an error state if some of that information about what caused the error showed up in the Build Error field on the Build result! +Allows for create table +The patch had been tested with the trunk revision +There are significant problems storing indexes inside Infinispan using the Infinispan Directory +Touches the following filesA javastubsjdbcA javastubsjdbcjavaA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavaxA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcjavaxsqlA javastubsjdbcA javastubsjdbcjavaA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlA javastubsjdbcjavasqlM M toolsantproperties +One thing I should mention earlier that it is a mixed javagroovy project +Just a minor rendering issue when the filter returns no projectviewdev the last column of the No results row is not the same color as the rest +If you want the unit test as well let me know if you would like it or plan to do something with the work from instead +Updated patch with feedback from OwenDevaraj thanks +Are you saying people may want more aggressive recovery time for than say s that HDFS may be configured withMy takeaway from your comments is even though HDFS supports this functionality might need even more aggressive timeout +I checked out the other jobs and this seems to be all the logging I can improve given this specific issue and the debug option +I think this may be the problem +Fixed in revision to the branch in revision . +Back ported to closing after release +The idea for having an additional src directory is that you could lay over the binary and source distribution at the same place without messing with the whole binary distribution +offt is not defined on my boxi guess the right way to do this might be through autoconf to check for these typedefs? but i dont want to do that for this patchsuggestions +I have committed the fixes to the trunk +Bulk assign all unscheduled issues with priority major or less to target +include a diagnostic string that is reported back to the resource manager by the node managerThat was clearly for errors which are at YARN layer so lets not use that for framework specific errorsbq + allows to access JMS resources through a HTTP servlet but is a suboptimal solution as it requires to translate any packet to a HTTP requestresponse +For the record Ive never used bulk deletion +So this would be a continuation of not being +which is kind of a painsuch is mavenMy only concern for the flag is that for implementing it every module has to know about and honor parameter which seems not clean to me +if it is what I suspect then it is like this you have the Java class in one class loader and the Groovy class in another one +Ill post some sample code that demonstrates it +Thanks Suresh for the reviewI have committed this. +Hi DeepaIve attached a patch patch which combines your patch and patch and includes a change to that describes the for testers who arent building from source +Roller is no longer being maintained +Reopening to backport to +The module split has happened but not the groupId change +It would be an enormous and much appreciated help ifyou gave me the sequence of URLS you had been hitting that led to the connectionpool lock +It reduces overhead of release gives community more time to review proposal and reserves room for us to adjust according to proposal reviewI will call proposal review again today +Are you able to supply a patch for this? in progress +However I am not aware of any character that acts like an invisible character in unicode that can be substituted for the soft hyphen so the text would appear without any characters and still yet the character count will be accurate +Thanks for your contribution! +It reports the following dependency unused in my project while compilation fails without itjsrjarcompile +Sending assemblysrcmaindistributiontextetcSending assemblysrcmaindistributionbinservicemixSending assemblysrcmaindistributionbinSending gshellsrcmainjavaorgapacheservicemixkernelgshelladmininternalSending mainsrcmainjavaorgapacheservicemixkernelmainTransmitting file dataCommitted revision +At the moment returns that span all the parent querys subclauses +Ill review the patch +After changing this Ive turned up some new problems +I must be more jet lagged than I thought +bq +Is the issue that the JMX config is not set? or is it that the needs to be uncommented for this to work +This is certainly causing the bug but its not clear whether we want to fix or +like we had happen a lot in conversion codeCould you add a test verifies all migration struct fields are accounted for in the merge logic? so if someone adds a new field and doesnt update the migration merge logic it would cause this test to fail +Found whats wrong with syntax should beselect limit from Customerandselect top from you please supply me with unified patches +Is not the latest plugin version +New patch addresses comments +So probably to be closed +I attached a test +I agree with the idea that to the user they should only be concerned about their documents have a Point +I can understand a project centric approach +Another school of thought is that rating something even negatively suggests you have a closer association to that thing than to the millions of other things youve never heard ofLets say you rate Bach a and Brahms a and Mendelssohn a +imo we can only solve this cleanly once we defined how and their Extensions in projects get handled +New version of same patch +to check for the existence of an input bulk file before trying to reading from it +But the JDBC driver wont let meThanks. +Otherwise make a refactoring issuebq +if it doesnt make sense mark it closed +Then the only problem left is DESCRIBE +Done +It will throw a but it will give us more detailed information of the bug +Bulk closing stale resolved issues +Im marking this Wont fix because we recently submitted a patch for a completely revamped Hadoop port +Applied the patch with thanks to Alan I also added some Spring Xpath unit tests +Fixing +should get us this roughly +I should have updated the status a while ago +v lgtm +Marked the fix version as based on Ilyas comment +Cool thanks to you Michael! +He says they failed the first time but were successful the second attached this file which strangely enough looks very similar to the error he was getting on the solaris box +bulk assign all unscheduled issues to next release + +ah what version of TS are u using im usingI cant find the line iniocorecacheve started building on lets face it TS dev streams are generally more stable then most softwares stable stream +To build full site do mvn siteUsually Ill just do xmllint srcdocbkx +Thanks +We could even go for given that it would not conflict with the existing annotation as they are in different packages +Thanks for pointing this outWhile we need JPQL grammar to be extended to get similar feature this one is part of APIThe new kernel level expression and update for function argument introduced by this current issue perhaps will be useful in JPQL case too if we decide to extend the grammar. +The icon is prepared but I am still waiting for the link to the property file +Ill be waiting on your task will be out this week if everything goes according to the plan +Looks good to me thanks! +The attached patch fixes both issues +if a route has been added it wont try to add an event driven route +No new tests added since it is a code refactoringI have committed this. +I think that is a fair game here to return true if we actually can detect a false exception +thanks +rd column is the average response time the second before the last is throughput in requests per secondx affine tx affine affine translate update please try out the latest patch and +Eric I would love to take this on + tools usage just move to uncategorized since no need lock down the specific usage is not part of core so put it just under application development +Which patch represents the best work? I got to apply to trunk with a little love but tests having to do with facets are failing +No NPE and workspace does not crash with +The example submitted on the issue tracker needs to be converted to work with PBR +But our customers also have high security requirements +Damn +patch added +css is OK for your app you need to tweak it +Close. +Now at I suppose you can track this update +opened instead +There are other places where is used +It seems more cohesive to specify the validation and its corresponding control flow in the same place +We also retain some of the changes in earlier patches that sets the default comparator in to validates that the comparator can be instantiated in set and the fix to described above +Marking as Unassigned due to lack of activity. +Actually yarn has a tool +Even if it were covered by the specification Id expect it to say that an exception should be thrownThe test case should not assign the same instance to both an embedded field and a field +In fact just yesterday on IRC one of the subjects was about potentially extracting such useful pieces of code which we could reuse without shipping whole of ORM the conclusion was to discuss these options face to face +Thanks to Bill Graham for most of the text +That complicates the build but Chris Ryan insist +Regrading to I am not sure whether Path is required to be but I am afraid there are user codes depending on it +Im not sure about SYSCSUPDATESTATISTICS that is what Rick reported in a previous comment based on his experiments but I would think the DBO would have that privilege too +That should probably be a separate JIRA +Rough draft against trunk not quite ready for full review yet +Reverted at subversion revision +Maybe influenced by +This would be because I use Factory a LOT throughout the example to support s meaning the page is already rendering when the exception occurs +Thanks +Thanks +The from endpoint is no longer duplicated but we still have some problems when converting the diagram model to the XML model. +Please use wutka as groupId +Hmm that should work +ok thats unfortunate +This looks like a useful addition +I wanted to correct this and then somehow forgot +We need to work on documenting that better I thinkNot sure I like the idea of having to run this markup daemon in the background though +specific situations in has a default of and getrangeslices has required consistencylevel but a default of +Let me state my earlier comment another way +Hi RyanThanks for applying the patch +I dont think anyone tests this paging with multiple sorts +Committed fix definitely improved the parallelism +Adds a step to the docs build +I am closing this one now +I can confirm that these issues are linked +I dont think an extra ifdef is going to pull down the quality of that code much +I had trouble getting screencasting software to work on my box +Do you get anything out of the thread dump that I attached +This is a bit confusing and probably a duplicate of which I have fix for +Given the complexity of the object being copied it would be surprising to find more bugs in the compilers code +Alexey please verify +I must investigate my old project where the failure happen +I will make a patch based on the delete template also +debf +batch update of issues with fix instead of the correct +solved already by +Thanks! +So we can continue to discuss it. +I had intended to open another JIRA to wipe them all out at once +If you have issues with running the same app multiple times on the same machine then you can use the versioning system +Sorry I was taken off the project that uses this to troubleshoot some production problems +Not sure I understand what you mean by installingAttaching my eclipse config +Belongs in srcsvnlib and is required for the attached unit test +Please note that your mail will not be forwarded +Added Camel build output on Jenkins which fails withERROR Maven JVM terminated unexpectedly with exit code we please attach stack traces instead of having them inline? That doesnt help the conversation +I also plan to propose an upgrade to hdfs +hit submit too is now fixed with a system test in place feel free to reopen if it does not work in your ll do another snapshot later this afternoon +Issue closed. +The above commit to branch is also part of the previous commit r the class got committed before the CHANGES entry. +Ill take a look at this when I get a moment Ill wait until I have timebefore assigning it to myself + upgraded to +Im unable to reproduce this using following steps login as admin user with english language set set language to be Portuguese Brazil logout loginResult all strings are in Portuguese BrazilCould you provide your steps if still an issue +This is essentially issue one +was in revision in +Noted as incompatible in +Ive just committed this to trunk andThanks a lot for the contribution Stephen +Committed this to the branch. +just wondering if their might be any unintended side effects given the global nature of converters +checkstyled code artefacts only +for code review +This is the patch that should be applied to correct this issueRegards my broker was too old +Is it possible to not merge the webxml? Currently the plugin is not usableBtw +committed to repoThanks for the review Ashutosh and thanks for the feedback Lefty! +User reports issue is resolved in +So if one user use to read his or her file that would be no different from current way +If we dont enforce that this is an UUID we shouldnt name it like that +Changed slightly the test to be less sensitive to JVM issuesThe change was simply to add the points in a different order +We have the same issue with and of the groups contains the character and the authentication is broken +for the idea +Thanks OlegYour patch is in trunk revision +Well it seems that the problem that I originally saw back in July is not a problem any longer +If we do have a problem then I am guessing Stefan will come back with something that is easier for us to reproduce +back trace copied belowusrsharebin +I guess I can do something with a file of course +New patch which follows Roberts suggestionsAll parent classes are distilled down to classes andAll implementations Solr code and Javadocs have been updated +Added code changes now I need to update the wiki +I remember having tried that at the time I created this issue but it seems to me I met with an issue I cant remember whatSo for the time being lets stick with this patch and once we get some time we can look into optimizations +But adding it nails down the fact that this is a relatively specific operation which fairly different performance characteristics and one that returns a value +Thanks Nathan patch was applied to LUNI module at rPlease check that the patch was applied as you expected +looks good +Before adding tests about the switch between the diffent modes I think we should reach an agreement about isnt it +Therefore is not affected +This patch purely cleans us our dependency tree and makes way for more refactoring +Can you submit some specs to that check for the presence of those constants so we can keep them working in the future + +Also in addition to testing on a physical iOS device I tested the app on the simulator by pulling out and putting back in the ethernet cable +Provided in patch +Justin I believe this is yours +Thanks Vlad +No tests because this doesnt change externally visible behavior +Addressing review comment +confirmed that dev builds are installing OK for users +i have seen duplicate key generation before but very rarely +Fixed in commit +For the time being this is the best will most probably release the installer for Groovy this evening and I would like to ask you to test it +Ive noticed that the regionserver is using select for its io which doesnt tie up threads whereas the datanode seems to have one thread openblocked for each write +For some reason we cannot execute javadocjar against our poms it complains withNot executing Javadoc as the project is not a Java packageI have working examples of javadoc generation for multiprojects so can copy what works from there +Manually tested! +Should I close this issue? +This has been done as part of the types work. +All agreed so I pushed the updated documentation in bbc +So how is this any more confusing +hence it will be better to have it in both the sides +Assigning to Prakash for now +But I decide to contribute it maybe to be not only person who is supporting it +Changes old API as well Corrected about the space after if +Do you have any tricks from this +Need to close the task once validation is done +In an insane build the worst effect will be a NPE somewhere else or that a statement that could have been cached will be thrown away and prepared again the next time +Patch for the branch +I talked with Jon Hsieh last week about doing a more convincing analysis of the performance of the new class from vs the existing histogram method +Created for improved testing for nested WHERE EXISTS would like commit this fairly soon but given the complexity of the issue I would feel more at ease if someone else could also look at the b patch + +configuration didnt use a distribution +Jim this is actually running off ofWere not seeing the dropped snapshot anymore so not sure if that is relevant anymore +Attached by Peter Andrews in Oct post to dev in subversion repository as of rPath applied with minor formatting changes and a parameter renamingThanks for the report and for the patch +The solution of reducing the lock will ensure the task tracker will keep on breathingReducing the lock level in the method get will not target the thread blocking and +Thanks Navis +could this be due to the examples not having a retry on an optimistic locking exception +Looks like Quintin beat me to provide a test. +We do not want balancer to break any of the favorite block placement policy configured on the clusterDhruba I think this jira makes a big change to the core of DFS +Any plans for working on this soon? Otherwise please move to +Thanks for ticking the box Jun because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Merged to branch at revision + push to +Not getting fixed for. +Thanks Mahadev +Pushed a fix to master +is available +Migration is essentially complete + unscheduled issues to the next release +Now sitting with the logfile infront of me +It seems to work in my testing I tested manually by stracing the datanode process and ensuring it doesnt open the checksum file when I run hadoop fs cat ignoreCrc +I didnt look into the output whether previous results were correct or the new one +No further comments closing. +A unit test of a controller should not be concerned with the name or existence of a gsp +Not a blocker since I backed out owner and no interest +Barring any bugs this should be all done +Resolving as not a problem. +Because that doesnt make sense to me at all +Please once a simpler case can be created. +Mobile showcase is not meant to include the notify component +Is this still an issue using +There are no compatibility issues +The mongo metadata logic will now need to be updated as well and more than likely the key constant should be moved onto the class rather than on +I removed all but one of the references to javamaildirI added a TODO to the last call that should be replacedIt still will not be compliant to the mailrepository contract but maybe it will work because of the current usage of the mailrepository bundled objects by jamesIf somewhere in the james code we expect to find a message in a repository by name after having stored it this repository will not work +This will be in the next update. +Thats just costing code space compile time. +Shane suggests that perhaps we could use the file to determine source files to copy +Quite possibly related to. +Yes I think youre wrong +I will be submitting another patch to enable clob cleanup +The reason that I designed these class originate from the discussion between you and Florian you suggested in comment Regarding Even if the source would have existed it would be a runtime check while a compile time check would be possible +Added tests pass for me +As I dont know the reason for that I added bool commented out to remember adding the bool when adding boolShort +Yes that was in the patch just got lost in the conflict +Archiving due to inactivity +Agreed it should be a separate class then + NOTE if this is done then this handler should extend the a modified form of the so it can detect conditions where ancestors of the DN are referrals and delegate appropriately + meant for scoped dont know if this a problem or a JBM problem +Wait its probably better to create internal or smth +A simple test didnt show the same effect +To me the two solutions seemed equivalent except that the starts didnt need to be adjusted based on the user input and the lengths did +Heres a patch that implements this +Davidcould you please give some more details about your suggested way of doing thisI have tryed the following but I am not sure to be on the right pathI have created two one with locale en and one with locale it +The metrics are correctly aggregated and reported for the have to take this issue with hyperic and get it resolved asap +trunk all resolved tickets from or older +Eli Updated the patch +Would it be possible to do it via a hookDo you want to allow a subset of operations? The hook is not very advanced right now and you cannot change the query plan etcBut it might be good enough for disallowing a class of statements +Thanks for the thats ok +You can use external ajjsFunction and call it as a workaround +Commenters can now to email notifications via a checkbox on the comment form +Does it make sense to have more than services on a communication portPersonally I prefer the efficient and lightweight one byte approach its more inline with the vision of Apache Thrift and there is no overhead for messagesroger basically agree but the experience is that people tend to use stuff for crazy things and in sometimes very unintended ways just because they can do it so the may become an issue at some point +Jim how does the patch attached here relate to +Patch version incorporates review comments +Logged In YES useridYep have the same problem +The exact numbers do not matter just the relationship between tables returned +captchas is an upstream issue. +If wsawAction header is present in the inbound message on the server side then theoperation is dispatched based upon the wsawAction header +I am still hunting it down +There are two possible solutions to the extra frame We could make compiled code always call the path which would then always call directly into the implemented call variable in handles associated with other compiled Ruby the side effect of this change is that for core methods for which we only implement the call variant wed end up getting funneled through an additional frame in the same way that throws the block away +Do you agree or instead have further use casesexamples to be considered +Heres an updated patch which incorporates all of Nicholass feedbackbq +New patch where the script is included in package +Lets wait until Android SDK ships with Gradle before doing this +we could scope things under +The value of this entry is a comma separated Job ids +HDFS JIRA +Resolving placeholders is not appropriate for the itself its rather the responsibility of the caller +nice cleanup with the factoring out of the for looking at the patch Dag and KristianCommitted to trunk revision Committed to revision . +I check that it close connections for with Administrator tool +We have finally captured two response examples +You can close this issue +No further feedback assuming fixed. +So either the entities should always get detached or never +And therefore backward compatibility all the way to the first version is not an issue there +Tom this is definitely a blocker +Then execute +Since the +can this patch be committed now +Unfortunately due to time constraints I am unable to continue working on this ticket +The patch looks good to me +Seems reasonable enough +Pushed to master +In our case it forces me to write nearly twice as much code with no added I cant at least use to exclude our legacy DAO beans +A geometry whose holes touch the exterior ring is invalid according to JTS rules thats why it cannot be built it should be represented as a single external ring afaik +Committed revision +Thanks +Using varchar as default column type for sql server and above should fix the issue +Waiting for +Added the antrun plugin to do the clean up in the package committed +for the latest patch +What if we allow an option for the client to fetch a subset of the configuration from the Namenode and then use that and then overlay the +while if only is set with false the deployer will not process any web service implementations provided by the application while it could still use access other web service applications +Its committed +CRUISESPEED +are clean +Ping? I recall someone said something about this but I dont remember who and where +I made the service optional but I havent had a chance yet to find the right places to issue messages when the service is really necessary but not available +Thanks +It looks like I might be experiencing a similar problem under Tomcat +I agree that we should minimize the weight added to the users jobs +I had to look through the source code to find that if you are not in the project directory the plugin silently prepends the project directory to the specified directory. +Hello I have now removed the password +The build works in Maven but fails in Maven +Manually tested +BrianSo we cant even mark as an error a potential parameter passed in via the D parameter method that was unused in the injections in the booting of ASEAP ? someone wants to do a nice patch Ill look at it +All demos work on and with the exception of jsprovider +patch for components you have to attached the file with copyright assigned to Apache +it would help if you provided more information about what your proposed annotations would look like +If I just type x y z then again runs but itdoesnt get the arguments +Waiting for an update +I believe that is probably the case for most Java developers not involved in the Apache community +I noticed that and was typing my last comment when you submitted yours LOL +Hi Tim please review the patch thanks a lot +This may also affect other Context ive changed the code to always inject a proxy for any Context injected interface +For a first step that should be enough although we may need to read the plugin settings from which could become quite an effort. +Heres the patch let me know if theres any issues with it +I was able to reproduce this on RHEL +Edits that occur while the checkpoint is in progress are sent to the new file +Sorry no fix required then. +You still need to run cygpath on it or java will do the wrong thing inside of logj +I wrote it in haste if youre thinking of using this in a production environment please run some tests first +I dont know who on yourteam is improperly claiming this address as their own but theyre mistakenPlease remove this address from your system +Cross form submits and especially this case should work +Fixed in revision We now support offers on a ctor +So great job of yours +Not sure if that information is helpful +We cant support accept directly on Socket but I think the class is a pretty good workaround. +More natural solution would have be just merging between two branchesDefinitely +minor tweak to clean up imports and to improve instrumentation +I have left a red question mark on the appropriate spot in the JDBC Support section +The codebase has been a mix with this patch and backporting the code to what we used in Camel to have a working solutionIt would be good if people could give the SNAPSHOT of the upcoming or releases at test spin in your environment to see if it works for you now +Will wait for the test case though +Patch to ovrride hashcode and equals methods +I can take this up +Setting the page encoding explicitly is not a big deal anyways. +Committed also to Federation Branch. +Of course if anyone in the community has a pull request to address this concern Ill be happy to review it and merge it in +Done apologies for the delay. +What is the difference between the result and projectSymbols collectionsProjected columns for procedures may also include outinoutreturn parameters +By utilization I mean the percentage of disk space used for blocks on a +Thanks for banging on the new stuff Im sure well find some more stuff before the next release +Would you like a version that can be put in the CTS +Looks great JulienCould you just update the class to set the line height to px +Sorry set fix version to. +The goal is to save a hash of the content of a source file during each analysis and to update the blame information on this file as soon as this hash value has changed +For some unknown reason I am getting an error when regression test runs with optimizing JIT +Carsten patch looks good to me thanksFelix we could move the tests to the code that starts the server I dont see a need right now but that might be an option +The size of the archive is reduced to MB now +This could be a separate Chris +It occurs to me that if were trying to help committers merge PRs a simple approach would be to write a git subcommand that accepts a PR and does the merge automaticallyIe something like git Would figure out what repo and commit sha is needed and then could do the fancy git commands required to make that happen +On frequently called function adding call to fault injector in the core code might have perf impact +The fix works! Thanks! +how can we ever provide a classpath container that will be usefulThe jars we pick up from AS is what is visible to all apps on the server so how can users that does not use the IDE figure out which jars to use +Should be fixed there +classpath file in the Git repository says where are you getting from +Do not add multiple concerns into this one requirement jira. +In the meantime WAS on zOS might be documented asunsupported. +it makes the implementation much more straight forward +Committed to trunkForgot new and deleted files in first commitScott could you please sync trunk and double check and I fixed it all correctly with my nd commit +We could potentially change the component name for +The test cases areo o And the corresponding entity definitions areo UniFKAll of the code and testcases were integrated via +Also I found out that if there is an error durring recover it can cause the AM to fail which does not result in another retry +Dear reporter thanks for the patch! I confirm that the patch works for both scenarios +The debug console does the same thing but without requiring any page modifications +Ok so is there interest to implement this who when +Thanks Thomas! +looks good +Only difference is removing the excess echos +maven specifies the use of the default maven javadoc stylesheet while java specifies that the default javadoc tool stylesheet will be used + +The name is listed in the property file bundled in +Also if the chain history is a Linked List instead of a Set we could know the flow +Also please open a separate ticket for the above issue +Thanks guys! +should mention that moving under view will help with the project setup +This is a working version for mounting DFS to linux file system +Bulk close resolved issues +I have Hibernate generate the scripts I use during installation on my application and have to write ant regex replacement scripts to insert on delete set null statements to the schemasAlso I think Im stating either the obvious or plain stupid but it would be useful to allow this action to be applied to associations as well as the collections arrays and on joined subclasses allowed by the existing annotation +Thanks! I was not aware that the second approach described was possible +Not sure +running test of part patch now +I believe this is more like a patch than rewrite +Also the patch isnt cleaned up as the real change should be on the class +Break this into modules one main one and one for each web app module which has its own jandex index +sry Patch with a fix for the test coming +Alexey this patch will not work with my example +It appears to be around setting thread local variables +It doesnt automatically sign the files before upload to the Apache Staging Repository so that becomes a manual step much like the existing step for signing the distribution +All users do X to check trash today +Ill try my luck there +Expecting a in the first few append calls and succeed as soon as the lease recovery is doneDoes it make sense +Attached files for test +mrepository deewu +fix verified. +The patch doesnt have a regression test case as I dont have a reliable way to reproduce the NPE in reasonable time +Attached proposed patch for this issue +It seems they may not be loaded both at the same timeThe correct fix seems to be to not depdend on a specific version of pax web in cxf +Not sure how usecase would be handled as there are not user keytabs in the nodesSimilar to java action shell action should support +I dont understand how Struts legal Java source code compiled into standardJava byte code with Suns own javac is a Struts caused problem +Of course if its fixed I can fix the Quest connectorLet me know if this feature is desirable seeing as Id have to fix anyway I could fix it in Sqoop at the same time and submit a patch +I think its ready to go +I hate to add more indexes to symdata +Closing since no further action are requested +we have applied this patch to the codebase and tested it works for the patch +All old files will have the same defaultreplication equal to the value of of your config +A fix could be to just go in the struts UI action and have it properly populate that we should consider why this is happening in the first place +I wonder how many interfaces your objects implement as they are being given to +So Im afraid it may be troublesome to incorporate only for GEP +would be much harder to verify without it +Itd be great o get those into target +If you want it to check xml files you need to configure the plugin to do soPlease use the includes configuration parameter +Im a long time lurker here and I really would like to see some JTA happening +It was already on x since we cut that branch after this landed. +If it does I am not cleaning up that orphaned rule in CS only on my appliance right now +patch restores the security checks for and Sequence headersThanks the patchChamikara +Use the correct declaring class for the callback validation +We might as well then just do idea which will be easier to implement and less fragile +took care of this +Jonathan you really think this isnt worth being defensive about in the? Seems like it presents an unclear api which is also hard to document since there isnt really a natural point to put it in javadoc or anything +Fair I will make the Forge transition roughly in the end of next week +Committed to trunk and +Patch has been released closing issues. +Oh sorry for not commenting on this +the interceptor is simply converting an optimistic locking exception that comes from Hibernate +Thx a ton Denis and Pierre. +A workaround would be to use the failover transport +This patch is not supposed to be committed +testing the mailing list. +Now because of that port is not open and other CORBA tests passes I enabled back this test +Committed patch at svn +Determining an applications runnability at will only require looking at the size of the runnable list of the leaf queue and the numRunnable in all the parent queuesThen when an application is removed we need to check to see whether any applications can now be made runnable +If it is logging only non consumed messages then it seems okAny way we are planning to upgrade activemq vesion to in production +After all the JPA specification is a major update affecting not only the but also the that Spring integrates with +Hmmm adding defensive makes the test pass +Rick are you able to provide a small test +I hope it is corrected now and there wont be any more problems +Those complexReq test cases are integrated from Andis contributed patches forThanks AndiRegards akiSince these testcases are integrated I assume this is resolved +Hi PrescottThank you for refactoring the +Maybe I should do first if everybody agrees on +see +This builds on what is in branch currently to move the default Alert display into and remove the displayException from +Yes definitely we should proceed +Hiare you sure it worksthis part is totally managed by tomcat and should only work for webappcan you share a sample to reproduce it +Patch to change the DRDA code to obtain the holdability explicitly from the result set or statement as required as a result set may have a different holdability to its statement +I couldnt find the war file with fix +My changes to add ssh ve put back the for the port I just didnt noticed your test in which takes care of opening a jsch session without giving the port number if it is set to + +Hi RoshanPls create a patch and attach that to the JIRA that will be much better than just pasting the code hereThanks Roshan +You will have to exit to get preview to work or close the VDB connect to get VDB execution to work +That nested bean in turn features conversionService and validator would be nice to see this revisited now that has been released +Would you mind elaborating a bit? What is the actual problem? The real world use case? Your proposed solution etc. +We would likely not want it at any lower level than thatIs that what you are looking for +I have not really stated that this issue is for Solr +It causes to fail when it attempts to instantiate a with initialCapacity zero +Example that shows what can be done after insertion vertical center thumbnail +both belong to Eclipse or one of them is oursIt would be great if you could inspect available RA properties and give the options on an extra screen. +Commit option A is required for this test because the BMP entity bean does not implement any persistence callback +Reopen for handling in +Could you post only the relevant parts of the log without the plugin info? This gives too much noise +Hmm what happens if CXF is released or does the means +Cancel patch to fix findbugs warnings +Snjezana any suggestions +I would be for Java +These errors SHOULD be in the log not in the ui +Ive merged the patches provided with a more recent version see patch +The merger just needs to count the number of records it emits +Im currently wondering if I can just remove pool +If forcing a table online is nothing we want to do lightly then it should be exposed explicitly +The patch looks good +Keith could you review this change pleaseThanks comments changes look good. +Patch applied in Rev +for radio buttons there is and Radio components +I am having the same issue as well +applied thanks lester +xsd extension on the cached filename makes it much much easier for users to manually inspect the contents so I agree with your decisionThere are other issues such as stale cache entries and recovery from errors but these are outside the scope of this issue +We do not store the set and use it in another thoughts or suggestions on a work aroundThanks In YES useridAre you using pessimistic lockingI tried to reproduce it but couldn have a Parent and of related Child instances +Alex Sounds good +It does reset readerCache then commit +If tool shows up meantime well and good well commit it +looks like a config nightmare to meSo to add remove a property you need to put that info into two files +Patch committed to trunk +We do keep all actual key names as UC throughout +The attached code is mixing setup with local iBATIS setup though which makes it hard to see where the root of the problem could be +I can give that a try +Thank you for your review Hitesh +Passed fine on hadoop because no pb in hadoop but then fails in hadoop since dfsclient blah blah is pbing +I can see the same problem on Mac +Verified in and my comment doesnt actually have anything to do with this issue. +diff +Running with this patch seems to remove the problem +Nick I assume this patch is irrelevant once Beta is out then updatesindigo will contain the right stuffOr +The attached project allows you to switch between Hypersonic and H by changing line which demonstrates that it works for H +The patch is a bit smaller than the previous one because some of the refactoring of was already done in other unrelated but the approach is still the same +Will open an extra ticket cause Freddy suggest its pretty sure +Thank you very much for the second patch with additional javadoc Bryan +Attached is the corrected web you for the submission +Okay will do +And if you see this again holler. +This worksHow do you handle this in practice thoughATM I see two ways +if the namenode adds entry x and later y the bookkeeper client wont acknowledge y before x to the namenodeAlso does this mean that when the ledger is being read later on x is guaranteed to appear before y +Here the patch +Now it fails to even generated code on Windows +This bug has been marked as a duplicate of +the changes look fine +UI has been revamped for RC +I ran the failed tests on my machineThe following succeeded Running Tests run Failures Errors Skipped Time elapsed sec Running Tests run Failures Errors Skipped Time elapsed secThe other do fail but they also do in trunk +I wrote a simple wrapper to run this example on Linux +Alls fineThank you Mark. +Couldnt we mark it as Cant reproduceThe cause could be a concurrent access to MREPO or simply using local deps from MREPO instead of nighlty +Vladimir are you still around to help me out with this +Also I will ask Tom Schlosser if he can add some more information because he is the maintainer of this test +Steven! +Now its not very likely that any user hits the default threads limit but we still have enough parallel loops for the patch Mark Ive committed it to trunk and +Contains various embedded objects including text boxes and shapes with m new to JIRA so please change if Im wrong +Thanks for the patch and legwork finding a public repository Donald! +However the still uses the private ucp field of the class +Exposing the Map might be be better since it provides more flexibility to the caller +well now I think that wasnt a good idea +What workarounds would you suggest in the if we need to mask our SSL keystore passwords on? Im not sure how long it will be before we can move to once the fix become available +i have changed xarchv to m optionbut getting same error if i use original runConfigure with xarchv also +Thanks Alejandro and Rohini for issue as it has been released as part of release +The OOME is a separate issue +No objections from applied in r +I think rather than tolerating a configurable amount of edit log corruption we should just never tolerate any corruptionWhy and How does one recover from such condition +On the one hand I couldnt be sure that the regex is not a simplification and on the other the Differences from Perl does not mention thisBut yeah if this is an unimplemented feature my request would be to have it implemented and its good to hear that there is work being done on it +The +bulk close of all resolved issues prior to release. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Could you please review +the names of the attachments are somehow lost +Component migration process is partially documented and being followed +I will try to do thatKathey +Attached eclipse config and error well the sources are in the remote repository and not in the local one and I currently havent added it to the Maven Indexs view +Looks great +Hi Galder a couple of questions is there the usual day deadline on this? Im ugly busy at the moment +Because you have a CTAS query and test did not exist in your db so when we first tried to get the table for test we would not be able to get the corresponding objectSeems stages for move operator and create table operator were not executed +My typo rates been too high lately +A pending bug is reorganizing the commands so this field will no longer be availableWhich pending bug +All patches submitted and committed and release note attached +Also I took all suggested measures to generate contents but this didnt help +my primary focus was to check if the implementation was able to handle larger datasets +The old structure would go out and do the detailed and then use it later overloading numTerms all over the place +irb doesnt accept these characters as input +Yes it makes sense to get rid of the logIdent for state change logs +I dont feel its up to Cargo to handle DOS +Bundled and unbundled binary distributions are now provided +This patch is compatible with trunk +Patch switching all Solr testing under Maven to make LUCENECURRENT +Reopening to get feedback from Nick +Moved to. +Agreed with Przemo +Justin I did not try to apply it but from what I can see from the patch file it looks +looks goodlet me take a look at as well it would be better if we can get them both in at the same time +Comments and feedback welcome +Please reopen if its still an issue +released with +It is critical for +I have never wanted to contaminate Tomcat during development +It works thank you +Applied it in revision thing I noticed in the logs the session in is not closedFixed in revision into branch in revision against JR d be good to fix this in JR branch as well +Taking a look at TOBAG the existing UDF already converts a map entry to bag +Both and are missing +The javadoc for skip saysThe skip method of this class creates a byte array and then repeatedly reads into it until n bytes have been read or the end of the stream has been reached +Ok thanks for reporting it +I think its OK behavior for we should verify though that the file that fails to load remains safely in place +The problem is in the that includes only the diff for commons lang +Ive created one manually and if I try to open that I get the already documented Could not obtain project reader error message +open a new jira with test case +Moving to +So the user doesnt have to sift through pages of output +Logged In YES useridThanks for responding Luke I appreciate am using IBM DB Enterprise Edition v +I wasnt able to replicate this again so we have to wait until seen again unless you can find something in the stack trace +This problem is covered in and its related. +on the trunk thanks Rajika for the contribution. +Three files are needed locally to build wo Inernet connectionSet the following propertyhomeuserand have three files availablehomeuserbuildrepodistcommonsdaemonbinarieslinuxhomeuserbuildrepoorgapacheivyivyhomeuserbuildrepoorgapachemaven +A admin might easily bootstrap a number of nodes to same token as it stands +Hardy please backport this to BranchThanks also to branch +Only the package declaration +DavidI ended up committing a slightly different patch to both trunk and +The fix was already in commit you try out running mvn clean separately when multiple goals are specified it seems that maven may sometimes fail to issue is caused by that some of the are not mocked Ive fixed as basis and we need some documentation on how to write unit test correctly to avoid problems like this +Upgraded to in revision . +I have a pair of setup for clustering dev +verified on +Can you describe your approach to this bugIf I have a an XML file configured that resides in my project that suppresses particular warnings will it read that XML file +HiIf no one has any objections to these INFO log additions Ill commit it in in a couple of daysThis helps projects such as Pig Hive etc +To clarify you want these so you can change those streams after startup yes? Normally you can set them to the but I suppose that doesnt help +Please add some documentation to the it section in the book about these conf settings +This patch enables other Issue Management System to be displayed in the apply thanks +Patch for this problem including PhilippYour patch is in trunk at r at r +I tested federated config but updated in this patch to explicitly check in both the federated and HA federated tests +Hi Max can you ping me at ggregory? I think weve worked together before +New patch that fixes the minor javadocs comments from patch A patch that updates all the unit tests that were affected by the virtual column change +Joel +The same is true for list requests per vdb and list sessions +Closing the issue now that it has been resolved. +Later on this makes it possible to associate bound columns with their base tables when the temporary rows which are buffered up for UPDATE actionsM javaengineorgapachederbyimplsqlcompileUp until now it has not been necessary to optimize the dummy SELECT which is cooked up while compiling INSERTDELETE actions +Lets wait to see if anyone else asks for exclusion. +Hi Simon Ive definitely made an error when asking to activate by default server option in Sonar +I dont see why this is a blocker +We actually support this already since Spring The corresponding property on is simply called message though accepting a or a JSP EL String that resolves to a cool I didnt see it in the documentation though +This needs to be fixed inPlease confirm that the fix works in master and then provide a patch request so that I know to pull it into the branch +Im going to let it run over the weekend +Forgot to mention that any class as generated by should not be checked in to SVN +Another advantage of this approach would be that we can use this method later on to create tables directly from scripts generated by ddlgen without writing them to disc +Bundle updated with license section scm and project owner in the jar archive format to classic JAR +This is an update of the previous +I was concerned that the drastically changed css file attachment would cause confusionYes please post your improvements to this issue + for the patch +En un pat familiar es comena a parlar una mica acaloradament de poltica que si els dERC sou no s com que si els de que si el Zapatero +See +it seems Ive created a broken patch and forgot to correct some places +So Im adding and XSLTC to the list of components impacted by this issueIt would appear that it is related to the use of by Xalan interpretive +You are a very good reviewer +Using dfdcdafedcdcbd revision and hal core the updated console is pulled in +Yes No new runtime found message is better to reduce confusion +Could you not listen to the installedupdated events and then do the refresh +I am attaching a somewhat ugly log from my client test bed +Thanks for looking at this +Marking the bug as a blocker +Thanks Matteo. +testjavascriptrun basics +Improved patch +Is there any reason to keep this as separate class itself? Why cant we keep this under dfsadmin commands itselfThere anyway result will be passed toWhat do you say +The JSTLs empty operator supports this feature +I excluded the repository from the search but when you try the build the artifact should be found +But Im still fine committing pauls patch for the sake of CQL once the version of the patch is attached. +on the direction +Ill commit +in the current revision of the documentation +This would be a pretty big change to the classification scheme that was argued out quite exhaustively in there should be time for people to if they wish before anything is committed +reload logic needs policy on failure to reload the file +im not sure if this is the right solution i have never seen that there is a space between the timezone offset and the rest of the date perhaps the pasrer simply ignores the timezone offset in this case +Theres a whole host of STOMP issues that we didnt have time to address for but that well get to in a later release +the patch to get another run from hadoopqa +Thats very helpful! +Hi GavinThanks for that pointer +In the class it uses name to compare which is changed to ID in the patch provided +Also a patch against to change classloading order +The naming clash which gives the wrong impression that we include the recent driver is unfortunate +Committed +Ready for review +I agree that sounds better that was something that I was concerned about I didnt realize that a in the name means it is invalid much better that it isnt a possible valid value +Gotten no feedback +This bug was introduced in beta +Just need to update it few weeks later after the previous patches +done and completed. +Not sure what youre seeing +Thanks Jon! Patch looks good to me +I would probably consider inserting after Chapter Microcontainer as JNDI is a basic feature of the app server +i guess we have another candidate for hbmlint then +But bounds do need to be lazy loading anyhow +Standard buffered images are stored in regular RAM and dont provide any hardware accelerationI would suggest that you disable volatile buffering only if you see this problem +Thanks Chris Giri and Nicholas + source is the full source for and in order to add the iteration optimizations on sorting lists +patches to convert to xmlbeans and new plugin +With Found more than one component of typeSee attached. +oldest ones at top of list since these are the ones most likely to need heuristics in the trunk +Fixed with revision +Updated the patch with a few small changes based on Todds latest comments +Refreshed generated page based on last four patches. +If we could count the number of records at the same time here there is another opportunity to get rid of unneeded data +When the JAXWSJAXB service answers the call forwsdl the resulting wsdl is all wrong having elements with stupid names instead of types +contains all of the following Hive sub projects What are your thoughts on publishing these four jars as artifacts in addition to +try run ant clean from the top level directory and then try the dist is this still a problem for you +Lars No hasnt been released yet. +Is there an opinion among the AMQ devs on whether code will be deprecated in favor of another store +Maybe I can think of an alternative solution +Instances are created through the factory to support both and JDBC environmentsThe patch does not enable any new codefeaturePatch ready for review +I think it would be better for the webservices to return the partial job instead of just the job id +In UNC cases this would be file which would mean that the serverpathtorepo UNC path should render server as the host name which is! localhost and therefore matches external. +Agreed that if this class will no longer be used it is better to remove it instead of applying this patch +Fixed with rThis exception explains why the test blocked in +I used a Linux VPN Client to reproduce the dont think client would make difference unless routing on Linux is wrongCould you help to reproduce itThanks! +How do I do that +Dont know why Ive attached it to the wrong issue +Ive added a regression test case in to verify the fix +ran ant clean test with and alls wellthat was the fix for compilation problem in just went to download the bundle and the whole thing appears to be covered bya copyright that denies making copies outside your own computerIm not sure we should be placing this in CVSCheers Berin +added a devbin directory containing files to start derby command line to manage derby test databse added working test that returns multiple resultsets added new added derby tools jar for derby db management updated to create multiple resultset return scenario added testcomibatissqlmapproc class that is the procedure used by derby +Suggestions? all I can say is hack and propose a change for as its superceded by +Thanks I have committed the current version +This was introduced with Erichs revision changes +This patch provides this functionality tested and is goodAfter patch is applied the following must be done to add two new filessvn add clientsrcmainwebappviewsvn add clientsrcmainwebappviewThanks +Your web application running on Trinidad connects to imap server over imaps right? Your application runs normally for a day or so and then whats happened? IMAP connection starts failing? Can you show us a small IMAP client script excerpted from your application +affected change has been checked in as well as fixes to Click Example demonstration +Hadoop was released today. +Same problem here this problem start happening during upgrade from Maven to Maven that upgrade from to s also confusing the way the attributes are mentioned +Could you test these to make sure I havent stuffed anything upregards code that is currently in CVS works fine +Hey Daryl when you have some time can you give this a quick review and let me know what you think +Just curious which gadget is calling navigateTo? All the others I tried toggle between canvas and home fine because they are container managed +Thoughts +you please add some tests for this in +Please let me know if it continues to fail. +Fixed a few things in +If the call got past the then as far as authentication is concerned the caller is valid +I have a schema of about tables a few of which generally have rows +This was fixed in trunk r r +OOMs are not the only problem +The unit tests are all green now +in files +ogm? no strong opinion but would make it easier for people to figure out in which jarmodule a specific class relates to +The NN and the JT can implement that protocolI can live with +Mina on trunk is changed a bit over +If you believe this issues should be fixed please reopen and add reason for reopening +Part of a bulk update all resolutions changed to done please review history to original resolution type +I told him as a mentor my to move this forward +I have taken a quick look enough to know that I may be in over my head requiring a lot of learning before diving in +Thanks Dmitriy for committing this +Should I raise this separately +Adler is unreliable i had bad experience with it switched from crc to adler for a while +Configuration and test case +the files lengthI think returning null make sense +Thanks +It adds the two negative test queries to the classPlease note that the current JPOX version currently accepts JDOQL queries returning a collection field so the negative test fails +I reviewed your latest patch +The unsupported stuff if it can be detected and warned about then all is well but some requirements you have a network and it is fairly reliable may be things that we can just rely on and opt not to feel guilty if the code doesnt work when that requirement isnt met +Heres an updated patch that puts everything under common +The link from the dashboard should have been workflow for creating a platform is through the agent is mainly because you cannot have a platform without an agent and the process of creating a platform by hand and assigning it an agent connection is problematic from a support me know if you need this new platform for anything +Also modify the comments to match the current patch +It would be better if the applications reused frameworkcommonwidget +Im sort of on this +Also stuff like should require a +Approved +Added patchfile to apply on trunk with some fixed UTF issues and a few translations as Christian I just look at the livd DVD install scripts and the used geoserver version is +Shotaro can you test the patch +So no more extra optimization seems to be required then. +Otherwise it can be difficult to determine the version associated to a row +Assigning this to you as these tests need to be handled specially for DBplease take a look at description and create Hibernate jira if required +It may be that geoserver trunk is already able to handle them +Does it matter if the thread is stopped correctly? taskDone is set in a different thread but stop needs it to be set to true before being called +the first Node output parameters are Type restoreBBType and second node input parameters are ExposurespersplitType ExposurelistfileType and this is wrong to connect Exposurespersplit to which are wrong types +Does that sound about right +working already +Preliminary testing looks good here +It would be nice if you can take care of this I am lost in a little tornado these days +Run in the background pushd deploy start popd +Im still not sure if the multiple entries are intentional but this patch removes the property with the thift value +Thanks! +So if we need another version released from branch we use and so forth +Instead to avoid the issue with conf mutation the relevant tests make a copy of the configuration objects and then set the addresses appropriately after the have started to get the correct ephemeral ports configured +bq +Great work guys +Uploading patch updating toPlease note however that this api is a candidate for deprecation pretty soon +Move the fix to +Closing +I thought does itWill work on the and unit tests later today +Test case to reproduce the that fixes the issueApplying this fix make test I attached before to passAlso I checked other tests in project and they all passed too +I plan to fix this +As long as the server releases the IPC handler in such scenario accessing other regions should not be blocked +t is the original exception thrown by the command execution and t is the exception that prevented the transaction to be rollbackedI dont see how we can put t inside t +Hi Rahul all changes are in please go ahead and start up this release +Im pretty sure this is caused by us defing our atom vars instead of defonceing +Ok I figured out that I needed to modify the +Note that I wrote and attached a test case to +Ill take a look at this! I fully support the idea of refactoring PMF utilities to help in a lightweight fix for this bug +Thus I modified the resolver to first sort the bundles based the number of potential candidates they had +list of phrases is updated +Steven has access. +Closing all ResolvedFixed issues lacking a Fixed Version +Introduced wire format and operation selector on top of current binding http based on what was done for JMS because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +noarch packages is safer choice +The issue looks like its an issue with my desktop machines ability to maintain a connection to our database rather than the code itself +Nate this is your code +BJWhen we discussed a setupconfiguration UI in the past it was suggested that it go in the Webtools component +Why wont that work +new patch fixes isBinary status in typo thats been bugging me +Ive noticed while developing the patch for that break the caching returning for last modified value +I dont need per file +OK So now I understand +The code for this patch lives underIll look for corner cases tonight and commit this in +this probably wont work at all in private cloud +These responses are flushed to the client by the Responder thread described above +is this the same issue +to trunk thanks for the review Stack! + Username letiemble + +Attaching patch with the testcasePlease review and provide your comments +Agree working on it will update soonThanks ArunAs suggested by you I added the credentials to resubmit api +verified all upgrades in ER +So it is possible to always determine the resolutionunit +This appears to have a security setting now. +can you generate the code with u option and see +Cancelling the patch until we clear up the issues Sijie issue Sijie is raising +Protected special characters such as apostrophe and hash mark with URL encodingFixed the infinite retry loop inTried but could not activate the Apache logging +Closing. +Yes was what I tried before +I guess you will add a connection attribute that is used to get the a database booted in slave mode +Annotate action does not work anymore +resolving issue +Drew is working on this +corrected the typo Coud not +Pls apply this +But comparing to the error page I guess that it should be acceptableIf no objection I will commit the latest patch +Most of them are related to concurrent access +This hasnt been reproduced or complained about lately so Im closing +More eyes welcome +I will add a descriptive section in the developer manual +events suffix to the class name for the event dispatch method invocation logging +Patch for review +In earlier cases for the first VM including it used to take about sec +Could somebody please tell me if this is a bug with respect to the current spec of X library +Patch is based on trunk +Ok +This will not apply to trunk as the robot is trying to do because the fix is only applicable toAlso I can add a test case if necessary though I think the fix is fairly obvious in this case +No extra action needed as automatically picks up new repositories from us every now and then +Move +why groupId +From Maniks comments I get that components might be cleaned up so this is not an issue in Infinispan +Rajesh do you still see this problem? If so please post the file from underneath the target directory of the module that fails the build +I suggest you get thelastest version and try that +Proper solution is I guess not to use List as intermediate format +Ive got exactly the same problem and im not using Facelets +Adding a fix based on the above comment +Red box outlines misworded text +I forgot to update those properties +Assigning it to Alena to fix the API and then Ill make the changes accordingly on the UIThanks +Yeah if we developed the Camel model at the same time as the Rider IDE we would definitely have made it more tooling friendly from the very beginning +I will fix this for the next release but for now I am leaving it as it is +I cant see the attached patch +I will send out a document to compliment the code fix talking about the implementation reasoning and possible future alternatives +Denisthe move from to was just a renaming if I did it well it was an overall replacement of by in all poms +On a fresh site and with forrest directly is working fineI am running the command from another ant task +I just committed this to trunk and. +Ok go ahead it makes sense in this case +I will backport to after tinderbox runs cleanly +currently I am trying to fix it myselfI will post here if any improvementmore This request is just to make convention plugin work together with REST +Waiting on hadoopqa +And I personally have never really used this class so i am not confident in knowing if the fix is still good +It looks like the component descriptors are being processed but things still arent behaving as expected as the component services arent being registered as the references are resolved +Let me tried to test it myself but the code your presented doesnt compile +Here are the files to be used with the curl commands +Unlinking from +The general thought was that all modern browsers will give you pretty printing by default +you can open another issue for this +If this is correct feel free to close the issue +Resolving. +see +and we dont accept mirrors without permanent domain names +Uploading version b addressing Knuts comments +And this patch does avoid them +Tests also broke also on because of this +Thanks +This looks good +The plugin support has to be enhance or kerberos has to be added in coreThanks somebody from developer team confirm my understanding in the previous update? Thanks Gary +At any rate as I said the discussion of Trie and when some generic piece of code needs a grant has nothing to do with this particular issue so please if you want to continue this conversation then start one on +Hans thanks for logging this as a bugThe fix should also take into consideration the qty of the asset available +I have used some Lucene but I liked the simplicity of the HS annotations +I dont see an easy way around this +The same workbench but without the libs and without the ear +sometimes can be null like in Eclipse embedded IE +No longer works +Looks good +Done +Hi John I see trunksubsystem also has a dependency on +Anyone can checkout the branch and merge with trunk though its literally the same effort as updating an out of date patch +Only the. +Looking at in jconsole there are metric scopes under with names from toTaking Roberts patch and fixing it to create one static the additional objects under disappear and the number of threads named metric goes from to +This seems to be gone I believe it was a combination of factors including adding some extra synchronization to calls and giving the integration test suite a chance to warm up +Verified on. +Another candidate for +The way is designed all security and validation is done on the server even if it is also done on the clientThe only way I can think of you could do session hijacking is to have access to a browser and admin user has used and is still logged into or by sniffing packets over the networkOf course Im not a security expert and havent had the pleasure of researching these things in detail +Just correctly recording the release note text in the Release Notes field +HelloI am attaching an interim patch with this issueI appreciate your commentsreviewThanksRamin could you explain why this is an interim patch +Just committed this +However the WSRP admin portlet still doesnt work properly pending update. +Seems like there are some tools that ship with the sun version but not with Openjdk until +this problem is due to flag +At least that would be consistent It could easily be done after applying this patch as this one already provides the management portion of itI guess the one reservation I would have is that Karaf is becoming quite large in terms of the number bundles +I wouldnt oppose adding the dep in principle and binary tarballs without contrib would be great but this is quickly expanding outside its original scope +applied denniss the original proposal I added a small javascript that sets focus to the searchbox so that you can start typing right also moved the image outside of the form and removed a paragraph to minimize the space above the image + +Attached a patch +Does this have to be in core hbase? Can it be done as a decoration? As a coprocessor? +Thanks Navis +fixes area all on the fuse branch +I will attach the code to a comment +Thanks Carl! +queueappemail? This was not my understanding from the above discussion +Which means even jstack cannot be executed when running with JRE +Assign to Stale +We are experiencing this same issue +The tree object model reloads properly but the wrong elements is picked up for display +Hmm i have tried it and it works fine for me without your changes +Its very interesting +When a partitionFS fails it will likely remain in a failed state after restart so reporting up and stopping to serve would be better than simply exitting which would lead to thrashing +But without the thread stuff +Paul has an offline comment about it cannot handle Connection exceptions +Moved to to allow more time for testing impact of changes +Any work around for me at this point would be too much of a would be so happy if something was included in to address you +Patch including everything including latest changes of Udai +and attaching deployment and code +aj files out of your source problem is I dont know how to get the +Added to trunk at revision +If this is a dead end after a few hours or so we can just go down the return pojo approach and then come back to adding in reference links etc. +i guess maybe labeling them with ? could be a little better though +Updated patch attachedI think I can just simplify the patch without waitAlso is refactored to remove duplicated codes +Im assuming youve already considered a commit the ideal option for handling the multiple versions +If this is Teiid then it should be Steven Hawkins or Ramesh BarryCould you please reassign to the responsible party and ask Paul to update the component ownerThanks and BPMN are not connected to +We can use to solve our problem +The change is small and looks good to me assuming Jenkins comes back OK +improved by doing serialization before partitioning and collating made the test mock and move to made test mock and move to I dont feel test adds more coverage since there are separate tests to test partitioner and +This is nearly the same eclipse project as in +ACS feature freeze date is November th +When creating a new how should we decide which implementation to initialize? Is it ok to lookup the existing is attribute in the? However I didnt find any code which actually sets this attribute +In the meantime you can find the fix in trunk in SVN. +The idlj compiler in the iiop module couldnt find the idl files when building from the parent dir +Attaching a patch +ML lives in our sandbox and is currently just a fork with some renaming of Maxent +I just point the ant build to my jsvc and Im done +Will commit when tests passNote for next time for patch updates our convention is to number them like etc +This feature would be very handy for master details to put the buttons add and remove right upper corner next to the master list +Some result cant be returned corretly +You could even have it only based on user or authentication status if you wantIt woudl feel very weird to me to have a repository wide encryption policy in place but that should be possible with a Jackrabbit backend +I tried to recreate this problem but am unable to reproduce the error. +This is also the case with Activiti and current Eclipse plugin +you dont update after have been created do you +This is either obsolete or was already fixed +My friend with the knowledge about this is not going to be able to help in time for the release. +the next minor version increment is +Otherwise it would be extra pain forusersIMO JSF did things not correct in the past and people complained about it +Whoever use the Felix framework in embedded mode in a project very likely have the same problem because I guess most of the application serves do set for each application or for some other reason the problem may present +Now I am attaching a new functional spec describing a new attempt +Also changes any reference to the in the comments +Changed all to Thanks for saves much linesRegards Christoph +Thanks Kevin +Its currently passing for me +Ill put this in V +Kieth could you review these changes pleaseThanks look good to me. +If null throw an exception +Currently ti will likely assigned cloned tablets to the same server as the source tablets. +This is an UI issue +Again not a bug +Ok +I am running tests nowTouches the following filesM javaengineorgapachederbyiapiservicesioAdded some utility methods for copying arraysM javaengineorgapachederbycatalogtypesM javaengineorgapachederbycatalogtypesM javaengineorgapachederbycatalogtypesM javaengineorgapachederbycatalogtypesM javaengineorgapachederbycatalogtypesM javaengineorgapachederbycatalogtypesUsed the new utility methods to copy arrays inside constructors and accessor methods +Slip to is not in JBDS or closing Wont Fix. +This is already in trunk +Is there already a feature request that the prios are also exported for findbugs? Or should I create one +See my fix to in trunkThat now shows native comments for all unicode escapesTwo of the otherwise lowercase names were previously converted to the Unicode for upper case umlauts I wonder if that was a mistake +Can be closed I I agree the issue is fixed. +on linux or ant on windows +Presumably the solution they already have in place handles this otherwise it wouldnt block adoption Im wondering what that is +Sorry about the run around. +C both Seam and Seam should be added for EAP afaik +at least your instance +Yep stop using the plugin and write something of our own +As you pointed out Java is EOLd already and SunOracle doesnt even let you download JDK anymore wo registrationWouldnt hurt my feelings to move to Java + +Sometimes there were only couple of rows sent but sometimes from to +Unscheduling due to inactivity +This extension seems outside of that purpose to me +A couple of stray references to objects from the prior session existed +Weld keeps working even with +attached is the suggested patchI think we can improve the pom files much further using dependency and plugin managements +Thanks Clebert +I think this issue is okWe can build real releases now so trunk is back in shape. +I think this is ready to be committed +The alternative would be not to use the extra list since the order of keyvalue pairs is not important anyway +Perhaps me moved the page after I linked it? New link looks good enough +test based on Amros description +Increases test coverage to in CS driverpep checktox coverage checkrebased on trunk this morningshould be clean thanksPlease open a separate ticket for the success issue once the patch is ready. +Werner conforming to EJB does not only mean to support the mapping syntax but also the way mappings are loaded which is quite different to the current Castor way +Because of all of these things we might want to do with the data there is no one way we can encode it in the database. +Maybe we should consider initializing the stores during the initialization of the I think we have the following related issues create method is never using the parameter +Might be related to is just another result of the problem described in +Whos working on this +Pull requests sent +Hi EliAs part of build process these adhoc tarballs are downloaded and deleted after extraction +Use the new async API if you want asynchronous see the seda queue as being very similar to a jms queue in Camel with the main difference being no persistence which is fine +In trunk +It create dependencies because wsdls have to be reached during deploy time +Is it related to this fix +It looks good but add some type testing to ensure feature does not break in the future +Crazy timing +It seems like the right thing to add annotation that would go onto certain methods that would indicate that when those methods are called the proxy need not be initialized +commit if the tests pass +Thank you for the reviews and comments on this patch Rick +I will take a look on your patches +What could be done if the value is null when the table is shown the first timeIt would be populated only after user clicks the button to get some data +This ones bounced around for about a year is there any expectation someone might run with it or should we shelve +Resolving as fixed +If you do remove it then could move it into the test suite? I have tests for the GB bug that use it +Wonderful! The code above is much more better +Thanks! +After Alessio review that this does not affect compatiblity this was added to the badPackages in the test case +Scott can you provide some feedback as to the status of this issue? Should we expect help from JbossI used jdb to see if it would provide visbility into thread activity when system performance degrades +This check removes nodes from the index that are not available anymore through the +from me patch file isnt exactly correct +If you checked you dont need to check s get methodI tried to implement its unit test +It is possible to use the lock manager for a lock that is local to the object? I do not think it is a good idea to enter the lock into the general lock pool +Sounds like a fine approach +Needed in case where cluster is misbehaving and you need to rein it in wo doing damage to your data +Mass closing all issues that have been in resolved state for months or more without any feedback or update +It does look like we only us COMPLETED and FAILED to remove other unused states we can add them when the need arises +Please file a JIRA upstream as well so that it gets fixed upstream and will be in any subsequent upgrade that we pick up +Can I assume jQuery or +FWIW packages are usually separate from packages for Windows +Changes to address phabricator comments +I also think it would be a good thing to remove the deprecated methods which are no used anywhereI will wait a week and will apply this patch if nobody is MartinYour slighlty augmented patch is intrunk r r r rI also remored the deprecated methods and it contains a change Jacopo did at r it was a nightmare to merge else +ps +It would be good if the patch included also the changes where a new class like this simplifies things or reduces the amount of duplicate code +I will state this again in these few cases the primitive model is about the best that can be done and still not astonish the user +I am really looking forward to getting this patch committed before the first release of comes out +so I also added prompting to the I know +Corrected a typo in previous scriptAlso log the number of iterations at which the failure occurs +None of the files have headers and probably arent formatted to the project have not included the in this initial set as it was part of a different module in the original project +Ill be on on as well as reachable by mailBastiaan +Going to use as a guide +But for compactions it gets a bit complicated and any algorithm will become a little dependent on the compaction policy +See dependencies can now be disabled +If you meant something different please open another ticket. +Calendar and webmail can be deployed with the flash portlet However what do you think my example for Portal API is going to be? +Reported prefs problems do not appear in so these will not be addressed as part of this issueAn attempt will be made to fix the concurrent start issue without using the patch submitted above +We will pursue this suggestion and review. +This isnt a problem its like omitting frequencies and positions you cant get them back as you lost that informationIf you have a index you lost this offsets information after indexing time +So I placed the fix in a location that solves my problem +However there are things that can be done without synchronization that still improve the concurrent access situations +I have just committed this patch +The directory structure mirrors the file system it is capturing +so you do not know how many txns are buffered up the tcp stack +everything falling out from this ticket is being tracked in other JIRA +Ok now it works! Thanks so much +Updated patch +It really should be in the too +Thanks I have not yet used it in production since I am waiting for a release in order to do thatBut during my tests on the test environment all was fine +yes it was time to do it! less to do when the migration will be done +Can we the pgp signatures during deploy +committed to X and X +Not to be SVN patch after H is commited to the trunk +Evolving sounds right for this interface +problematic added the patch in revision as proposed by Thanks for the contribution! +bq +Now working on the property sheet and wizard bits +I thought this was written by yourself. +A git admin from infra or anyone on the cordova PPMCWhere are these two lists kept +It wasnt assigned and no one seemed to be working on this +are you suggesting we should try eclipsemeclipse goal? I will give it a try again +Also removed and builder methods which were not being used anywhere +Ping? Is Perfj dead +Hi Ed can you please post this patch to? Thanks +Check the validity of the XML file +I can not understand why we need codegen jar at the runtimeThanks war that i built on trunk seems to have it doesnt appear to be fixed +There are only two scenarios where I see compactions stopping like this with the tiered compaction strategy the server was restarted a compaction failed unexpectedlyThe fact that there were no errors in the log makes unlikely +Verified that the bug is fixedmysqlselect accountid hosttype bytessent bytesreceived from usagenetworkaccountid hosttype bytessent bytesreceived on latest build +I verified that the latest change fixes the issue +Tests show this does work but it would be hard to reliably work between builds and not using in source files. +multinic in openstack parlanceStill testingThanksCurtis +We cant for as people using files and core do still need to manually add these listeners and there are some out thereBut with Hibernate Core and annotations merge with Core we will be able to apply the event registration logic to all systemsWe should make it clearer in t he doc that its a bad idea though +Hi SergioThanks for providing the sample script and the appears the problem was that the Dojo decoration of the checkbox in have not honored the binding values set by the formcheckbox tag +Sorry for the delay. +Discussed the system test changes with John they are included in this patch +Fetch requests per second per fetcher Useful to know the progress of the fetcher thread +i forgot to mention that i see the same problem with other commands and not just the archetype plugin +The key was to do both at the same time +Thats the intend +If you are interested in what lines of code have changed please diff it against the same file fromHopefully this patch finds its way to VFS +I believe this is fixed with the snapshot. +This is very old but +Thanks. +Tab panel with two tabs on each there is richmessage when sending an ajax it says During update vertragsAuswahljidt not found where the ID is the richmessage in the tab +Invalid Issue as per prev comment. +I will push in the button comment from last week and then begin the rebase +Attached patch that renames the Java bean property name on to default to adhere to the implicit standard set by +Can you please explain the purpose of sendresults notification? If it not used it should be removed to avoid confusion +The fix has been committed in branch +Fixed using the new injection framework and xwork configuration design. +With I still have a Invalid content type textxml charset for Axiom and when in the same JVM I call a Spring Web Service using SAAJ and MTOM +Hmm +If there are no objections within the next days I will commit this patch to the trunk + email Red Hat email abhayani vkakmbojanmbojandboerendhuangjmorandlofthoudlofthouapatricimputzltexiergzamarre this ticket done? I thought these were created +I removed the auth handler to address the auth plugability concern +Patch to fix output format for both servlet and servlet +Those methods was part of ISO version published in + +Thats a limitation of the way classes are named in the JVM +The warning can be ignored as the variable is initialized only once and never changed by the scheduler +Since this field was never utilized I took the liberty to repurpose the method +I have your cvs repository root but where is the source for this plugin in the repository +Synthetic will now be when the parser really is unable to recover via backtracking +Even if you add an id to such a bean the corresponding metadata never gets added to the alongside that ID +That gets reasonably complicated though since itll likely require statistics like percentage of input processed output generated for running tasks which may or may not be available depending on the being usedAlso we could move towards getting statistics for individual partitions which can help avoid skew +Closing this JIRA. +And Oozie does similar things but never have we run into such a situation +Find attached the result of resolution imageType png +Merged into in revision after releasing +Publishing to has been removed from the Hudson job and needs to be +etc +Is it possible to run execute testNG junit and junit tests together in the same project? +Thanks Charitha. +ok finally got things setup to take care of added axis dep to the +Attaching a patch for TedI have attached the master log +I just started using Xalan on win +I am beginning to wonder how did I get those messages? I checked the codes and we have to add Dfulltrue argument to view the full projecthelpdescribe Dplugin Dfull in an unreleased version of the seems to have done the trick +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Adding Incompatible Change for now so folks are aware of it + +Please provide more info on the ACL items addedWhat are the ACL items added what is their rule numberAre they not showing up on the VR +Fixed +Even earlier +moving the discussion topic without adjusting the subject +Also this issue does not seem to present itself on the JVM only the one +Nice catch Sriram that actually drops the collision rate by Here is a new patch that fixes that bug fixes the docs exposes the cleaner buffer load factor as a configuration parameter +Note also that doesnt work on any branch except for trunk +Thanks Pedro for contributing the patch. + + +I add it in my TODO it will be really awesome to get these kinds of commands +committed. +jsfplugins +There are around throws in the code base not to mention various subclasses of Exception which will need their constructors refitting +It may be also good to exclude the sample from mobile view till it will be solved somehow +I am going to mark this as not a problem +patch should be updated +Want to retry your patch against hadoopqa again Chunhui? Or try the build locally +with JBT. +Hi Fabrice Im reopening this ticket as the reviews web service doesnt allow anymore to search a review by id + All handlers but channel and channelWritten are already provided by channelWritten is the only handler that is not provided by +The question raised is whether it should be +Itd be better to fix the testuser could configure write thread to which leverage netty threads to handle writes and configure read threads to a sensible number so read would not block writes +I attached a new patch cleaning up the added preGoal and postGoal +The patch has the deleted +Would you be able to produce such code or does it have to be executed within +AFAIK there was a plan some time ago to autogenerate the documentation for those settings +I attempted to incorporate all of your feedback +This patch adds to the list of branches on the Derby web site +This issue causes +Completed work merged from branch to trunk in revision . +Hoang Hung can you test the usecase with this tag +In my head I imagine that reworking Solr to eliminatereduce the use of is a monumental task +This is already fixed here for part of we will not backport this to +In search for html and replace with xml +updated installation to copy the and to the client directory +Couldnt find a way to do the test with +With the dist file it works +Working on it +I didnt VETO anything +I think the delete feature is for deleting files on exit especially for unexpected exit but not deleting the files some time later +I do certainly have lots of different object structures to work with thoughSo please just pure java objects example that I can mostly copy and paste within the ognl unit tests and I might have a chance at being able to do something +I think Ive resolved the issues with the tests +Lets get the basics working first +Nope because the command which calls the hasLock +I totally agree with Jonathan on this one even though we are the once that have tried to optimize the get path +I committed the patch to +Hi may also find the forum issue hyperlinked in this issue to be of as per user request. +Hi ThomasThis patch might serve as a good starting point for +One thing Im not sure about is what to do about the author tag in usually we just apply a patch with thanks to a contributor but Dan may suggest something elsecheers Ill look out for it when appliedRemove the author tagThanksTomasMany thanks for this patch +At revision old resolved issues +all bulk edited issues have in a move issues to and +Yep it works DB tests are closing all issues that have been in resolved state for months or more without any feedback or update +Merged +Thanks for reviewing and committing +In general in distributed systems communications can arrive out of order no? So it is better to develop protocols that allow for this +When creating patches you need to build them from an updated local copy +x to look at in caes JDT have allowed for it but for now postponed +This is probably something to consider for Maven this be handled by just appending to the existing classifier? I dont think Maven prevents you from having a classifier +There is a gap in to provide an option to configure if we should map such methods into WSDL ONEWAY operations +commit efdaccaeeabbeffaafa +What use cases are there for symlinks? The really interesting use cases are since that could make archives transparent and also permit a style permitting a single namespace to span multiple clusters potentially making the namenode less of a bottleneckWithout some strong use cases Im for any implementation that does not support links +Committed to branch and trunk thanks Eli. +Thinking a bit more about the proposalAs you know you have no guarantee that a terminated message will be transmitted in one single block + One option is to simply shuffle the list of files before writing to the split file +Can we fix the thrift shutdown instead to not wait for sockets to be closed nicely +But yes we can move this particular code to other Jira +Im pretty sure youre right +this issue +contining comments there +A new test case is added to +Aaron can you please explain the use case a little more? Why do we really need maps? Why isnt metadata serializable +Id be happy to implement the logic to evaluate the fieldActions and act accordingly +bq +Thanks for the rebuild +Should I just make the Literal Geometry return a prepared geometry +fixes a typo in the assembly file +This happens in the case of large attachments or soap messages sent over a chunked input stream. +This attached diff file is tested on VR and SRX +The goal is to get an online and downloadable VDL document bundleThe attribute descriptions need to be updated and flushed out +There are a few here that are worthwhile +In the second one I wonder if using would be a better betI had some concerns with but it seems the cache layer that caused some troubles in the past can now be disabled so it might be a good idea to take a lookJames can you comment +The error information presented when a repository is invalid will be better when will throw the correct exceptions +Also I need to be able to observe an on change event with fajax so that I can have the value pushed to the backing bean via ajax and rerender parts of the page +In addition to this there also appears to be a scenario where the deleted +The patch has been applied +Thanks for doing this change +And please provide us with information about OS Camel Version Java Version etc +the current state of my effort to reduce to configuration responsibilities only to enable to be the mother of all dwelling and to access in a NON static m still working on this patch but I also wanted to put a first beta of it out to show the direction and hopefully get some early feedback +It works for me on both trunk and branch +There are other open to track problems with using specific logging frameworks +Finally what about the support for VLV controls pagingI heard that does have support for VLV +Both sample applications work correctly on ocj was because the was not included in the guess OCJs error reporting leaves something to be desired here +Back in town after a few weeks in Japan sorry for the relative absencebq +Sorry Chris I was a bit too quick on reading the patch +Thanks to Francois for this important feature and to Sunitha for the thoughtful review + does not use the SEBR yet +Lets target the singularity then +Merged into in revision after releasing +If the proper thing to do is to just create a new jira just +Ah OK it is still strange that this merge exception occurs +So this extra jar is only really needed on JDK and am impressed by your quick resolution of this minor issue +Some conclusions after discussion on the mailing list adding an extra dependency is not good an optional module that uses to decode MIME fields with dates would be a plus although is to be used in Message interface extra Message and other DOM interfaces implementations could be made immutable this will help in thread safety +Unable to reproduce using please provide additional steps if still an issue +I created this patch from the tag of +The bridge between journal and ledger storage is the bookie which uses both +Then put a big bold statement on the web page that says We made some parts of Commons Math Serializable where it was trivial to do so but we do not plan at this time to tackle the parts that are not trivial +There are no other applications running on the machine using my classes +Here is what I got running the block id freeing tool on some cluster images +When does this happen how what version etc +Thanks +Thanks a lot Xi. +note adding synchronized to print does fix this issue +applicationcontainerstderrbinbash binjava No such file or very much TedCould also you post the output of mvn version where you ran the test? Im particularly interested in the value of Java home is there something valid at the location it is giving you +Why does Activiti not simply reuse the open connection it got from JTA in the same way as it does the normal connection? Is there a valid reason why in JTA a new transaction is started? It does not seem right that Activiti has a different behavior in JTA than in normal modeFurthermore starting from the listener all kinds of methods are called in the JPA entities +closing issues for released versions +Attaching patch with the suggested temporary code change +Just added and as related +will commit after tests +Fixed in all the resolved issues +However I think its better idea to have two JPA modules which are slightly different than HIB and JPA which are totally different +That means that applications that use a library that was developed against Avro will not be able to use Avro until that library is also upgraded +Phil thanks for the comment +I have seen this issue before but cannot reproduce it on Mac with branch nor with trunk +Deprecated in revision +Looks are no new tests but this patch is required to unblock more than test suites when running on Windows +Everthing is workingreporting as expected now. +uninstall hibernate plugin before using Juri I think you should add the line above to your quick start guideAnd another question Does it mean that mongodb cant work with hibernate together +may be environment but it should work without any problem OOB +See Russs comment. +There should not be an entry of template on poolid +Fixed +whoops this time it compiles +Sudheer btw your current transaction manager lookup implementation for Atomikos looks dangerous because youre keeping an static reference to the transaction manager +Nevertheless it may be a good idea to have it as one of default to choose +Im going to use as the reference for this. + has been marked as a duplicate of this bug +It does do this +Cant wait to try this one outObviously first Q does this work on as well as? Things where we expect a certain ordering of items to occur in the unit tests have a tendency to break when running on +Can you please attach the full stack trace? The screen shots do not capture the root cause exception which is cut off in the screenshots +The patch file is fully qualified relative to your C drive +However it picks up an alpha release number if you are generating the release notes before you have cut the branch for a feature release +Closing resolved issues. +I think these are useful changes +Closing issue +Thanks for nice work Subbu +I did find that restarting the services fixed the problem though it may take a couple of attempts +Specifically what make based packaging script are you referring to +This looks good +But I think that points to a potential improvement to the tool +Attaching an initial patch that fixes the DOCTYPE and META tag issues for the lib file resulting in content similar to that for other topics the attached file is the output file for Getting StartedI didnt remove the IBM copyright in this patchIll try to figure out how to get the DOCTYPE into the file but thatll be harder +Hi Chris the configuration used by service may be located in a different place than the standalone oneYou may have +Thanks JohnAlso thanks Daryn for reviewing it. +SeeOther procedures in the reference man use CALL +hi SamisaIve put the fix in for this but want the test to ensure that it works OK +Logs from HQ server and console logs from test +I think this Priority should change to the Major since Java EOL is November and more developers has JAVAHOME and Path set to JDK and yet many application must compile to target Java +Fixed it in issue is fixed now +The issue is not reproducible in AS FinalFor backports use Enterprise Application Server +with the latest comments on the patch Can you change to use? Similarly use while building score board etc +Make sure of the tests pass without failure +Ideally users will put their MR jobs in Circus and put Circus in their QA pipeline +Hi SergeyI can reproduce the error will look into it laterThanks for pointing that out +If you finish it before we release beta please correct the version +This patch incorporates Raghus comments +Oops take two +yes it is one major difference from Groovy to Ruby or Perl +The build and repository issues should be cleared up now +Please confirm the current status of this issue and if its incorrect change it as required +RightI do see the potential value in requesting to drop points that are too far from clusters rather than assign to clusters +thanks sijie +Im not interested in changing the design substantially at this point without a good reason +But yum repo doesnt understand amd naming It only understands x +This is essentially a documentation change +This seems smart +I think this would be a nice feature you then have full roundtrip graphical modelling EVEN IF someone just started from roo script command line. +We should also enable the setting to make it send emails to the list whoever committed breaking code +Leaving issue open in case we do a if we do we can commit it there tooThanks for the patch Jeff and thanks for your patience w our slow application of it +Ill report this to the plugin group +Commit away +The default sequencer is submitted as a patch to the of using a default sequencer we are now extracting the info using straight JCR API. +Yes Omega matrix could be generalized to implement Matrix API but for now it is kept specialized to stochastic projection +Note that the attached patch is now obsolete +good catch! Daniel could you swap the value in your committed. +Thanks there any documents that need to be exchanged prior to merging this code or do existing agreements suffice +Initial you create a phabricator entryFew comments +sorry the title should be Memory Leak in reading is in CVS +Possible patch for the applied +This is not fixed in DR +Please verify +Attaching patch This throws a now instead of an NPE +The new release should be available shortly after the sync with central. +No response assuming ok. +Reopening for nulling at least term we should probably null other memory intensive things like deleted docs norms +Thanks issue is fixed and released as part of release +Please hook up the CMS for us +It can probably be pushed beyond +Committed with change so would only happen if keycachesizeinmb option was left blank in the conf + +Too many exception implementations without much value +This is about detection and producing an informative error message +Ive just committed this +The workaround for users who cannot connect to recover a prepared XA transaction because of the lock timeout would be to connect with a different user so that the default schema is not in the middle of being created +I moveda lot of the stuff in the xmlDefinitions package into the base class +I reopened and closed again to set this as rejected because at the end of the day there was no bug in the ws stack +got a patch for that merged in Tycho +Groovydoc does basic processing of those files ll add a and refactor +This depends on +please have a look at with a nightly build. +Hi ChrisWhat is the status of this patch now? dont remember +Hit the add button accidentally +Same thing happens here +we have discussed about this whitespace issue in general in the team we will at a flag to the railo admin suppress any content written before +Generally speaking the code written by Davide looks better than mine but Im not convinced about the way is particular I think that configuration properties like should delegate to the underlying otherwise we run the risk of having the Filter not perfectly alligned to the jcifs classes it better as many jcifs classes use Config to assign private static final variable I think the initialization should be done directly on the Config object via and then used internally by acegy integration do you think? Cheers Fabio +Fixed in +Move to our new and shiny Launcher category +The exception doesnt say anything about what that means +An other point would be to make all pages in lowercase +There is now an option at the bottom of the Preferences Groovy Editor page where you can disable dollar slashy strings. +It is not really a fix I guessFor large reads a direct read would be the right option to avoid having another copyBut for small repeated reads it does make sense to bounce chunks through a buffer to avoid syscall overhead +Hi Willemgenerally it is better to submit a SINGLE patch in a SINGLE jira when all files changes are about the same scopeIt is not necessary to have a jira for every single file modificationThis generates lot of traffic and more committers work. +Since we have planned to call our own and stop using the Java builder this item is no longer an issue +Anything is better than that +Thats fine since that is the very purpose of it but the side effect is that it considers two objects having the same priority equals! So you can register two items with the same Order value +Uma Nicholas thanks for the details +patch based on nightly build SNAPSHOT +For now the set method for scheduler dir remains for backwards compat as the variant is the only one in existence right now +Lets see how long it takes for Hadoop to release the next version + +Example of two autocomplete fieldsOne render results as list second as changes +The attached patch looks good to me +Thanks Rahul +I guess we also need to do something on the consumer side to pull out any custom headers into the camel message +As youll see its a WIP and needs ALOT of work. +Backported this fix to branch committed as svn Resolving +What about appendIs it ok to change aTime to mTime during append +This is possible because of Axiom contains the Axiom CN moduleThanks didnt think it made it through +Should i select under the workflow button the Link Pull Request choice? +Patch applied +Excellent thanks for the update +Committed with revision +jon +Thanks for review +Nothing to indicate Im logged in +that should be in place for the recently upgraded JDK on all linux build agents +on Chriss patch +fix was verified +Why would security be disabled in a production application? The intent of disabling security is to prevent unwanted httphttps handover requests in development +It should have never existed in the first place +You may send it to or vbose +Updated patch to include some test updated to include testing directly from the grouping module as well +a severe and fatal result when the channel is stdin stderr or stdout +Thanks! +Thanks Christian I forgot this part +IMP formatting hasnt been updated for X consists of another copy of the grammar file and another related grammar file neither of which has been ported to docs contain a description of implementation for IMP formmatting +because we have a single notification queueWhich patch is in but not in trunk to make this difference +commited. +This looks like an unobtrusive change +Also only one tab will be available when implementation of preview will be not available +Sure +Added method which uses to obtain the delCount of an info Updated to patch with tiny changes made a few more methods public fixed javadoc warning one formatting fix added CHANGESI think its ready to commit +let me know if you have any question +I believe that property activation is also broken +So I dont see any confusion there +would you please tell me on which version and under the which name or keyword the problem has been resolved +I was hoping this would get fixed in but it looks like the resolution to this has been pushed toThanks for all the hard work put into the releases +If you validate just before the index happens you have a copy in the main CF but the secondary index fails +logj properties file has been added to server configuration to support the logging of the and Cache +bit executables only +batch transition to closed remaining resolved bugs +So the only question that query is answering is did this column keyname exist with these parentsThe only way to query values with the API I proposed is to create a view by the valueEDIT Hmm +Thanks for pointing this outKonstThats already been done in when tasks been assigned any queueThanks Konst for your commentsUpdated the patch with formatting issuesThanks looks good +Attempt to capture a stack dump after the issue occurs +failonerrortrue +or as an alternative include a and include the various adapters +Oops I missed that the patch is for +Tested the patch on nodes and it works fine +See comments on +Its not part of the global operation metadata +Let me try taking off the qualifier +I updated the summary to be more specific and set the to to cover the lazy loading of the of a bulk update all resolutions changed to done please review history to original resolution type +I rather suspect that youre using add generally recommend defining as bean definitions +Combined fix attached to fixes the problem of hanging +Just a guess but does your JDK by any chance contain some version of Xalan that accidentally gets pulled into by +Is there any progress about this issue? Anyway to walk around this +That would mean to remove the test and data wait until the IP clearance is done and add it again +Im pretty confused here what does Infinispan have to do with serving assets? Are you caching assets in Infinispan +Marking closed. +Thats probably the best approach +there is a workaround by adding T false when enter gshell +Jesse you had a pretty well thought out plan for this +Given that decomissioned nodes are provided as the last location for reads very rarely they will be used for reading and hence will not affect the average load +Besides unit test we need to add some performance test in both phase and phase +The blackwhite are not the finished ones they just illustrate the ideas of the symbols +One would first have to get it into Lucene and then get that version of Lucene into Solr +bq +This will be great to have! Thanks +I went through a bunch and they seem to be all restored to their original positions +I hope the patch will be committed +I forgot the exact reason but its due to the recent changes in Windows networking stack +Thanks Robert for sharing I will look the code ASAP. +That code example does indeed reproduce the fundamental issue here +Is this not a problem in? Im confused as to why this patch is marked for only +All bar one done project is not ready for the win build yet please when that happens. +There is now a command in Project Groovy Fix Compiler levels +This is a duplicate of +Where is that +OK now we just need to figure out how best to incorporate it into +I understand the documentation refers to +The version installed on the tasktracker may differ from that installed on the job client and when those differences are specified as final parameters on the tasktracker then the version installed on the tasktracker will win +bulk defer of unresolved bugs to +See discussion in the comments for details. +both work for the delay anyway from now on i have allocated more hours for JMS workPls take the instead of +the patch file is based on trunkusing Li I think both are ok in this case +Reassigning this issue to be fixed in both the next Engine and Commons Auth releases +Temporarily assign to myself for backport +There is a sandbox of bval at viewvcincubatorbvalsandbox which uses some type handling code from the forthcoming release of to improve the generics handling throughout the bval codebase +This has already been Stuart Can you please tell the implementation details? +I added some default tests for the Bootstrap process in general +Please review the code +logs dont have this errorAnd I get to eat my words +Ravi Can you please suppress the warning? No point in increasing the warnings +Ill get these changes committed to Bens branch +This bug seems to have been caused by r +If we do that we have to change several other classes or move them to ipc +Will upload a new patch for the same +We dont check for this today. +I agree we should be providing support for other languages Given When Then should be this while sorting out the Story Scenario will put it back with some examples +It will involve changing most of the files in the doc set I believe but it can easily be done in stages to allow for plenty of testing +Great but apparantly it isnt fixed in +The bug is in which is provided by Spring Integration +Some notes IBM J and Java actually ignore the permgen Java option so it will not hurt +The quota should be checked only when a new block is requestedIll attach a new version of the guide with the clarification +I can create patch if desired +Ill try to take a look over the weekend +I just committed this +A quick fix removing the key is null check thereby allowing null values in primary composite definition of a key is for it to be identifiable +Committed revision +I meant is not effective in S which makes numFile +cannot log in on any of the you copy and the props directory from the embedded version everything works +Also provide the get method so the application code can check it +Turk. +Fuse Camel stable branch has been upgraded to Spring Security assigning this back to you can ensure the fix get into the branches when you setup them +We need to test this and either fix the bug here or if this just isnt meant to work fix the docs to not put any Ruby objects in the cache in its examples +Also the name should really match with the already existing Btw what do you mean with any expression +attaching patch to make the lucene dependency optional again +Do we need a different patch for branch or should this be +This can be rectified by setting the use property to true on either the or the Connection created from the +Theres also some issues that Ive been notified mainly concerning performance when there are over million files to crawl +Closing resolved issues. +Attachment diff has been added with description make iincs wide when index someone please take a look at this? A fix is included and i have a bug that requires this fix +Hi AlftheoCould you also include a test for this so I can apply this already + +I think it a small bug and added ASF license to source looks good one comment Where is the code if the user does not use we fall back to the default deserialization case +For the long term I like Uwes idea better I think rather than restricting which javadocs in core can link to what and restricting which files the analyzers can useSeparately we should fix that Analyzer +Ill look into it +Okay digging around shows that the smooks transformation attempts to handle the wsdlimport by appending those attributes to the schema and presumably this worked you have a working example from a SOA release and a comparison with the proxy version +I dont think so since with operators you rarely need to grab the operators directly +There might be others if everything your doing is in Chinese that commit will fix it for you +With those lines in the cost estimates are different but the bottom line is the sameGood query planModifying access paths to use join order for optimizables with cost at cost rowCount single Bad query planModifying access paths to use join order for optimizables with cost at cost rowCount single Ill look into this more later Ive got a few other things to do first +It seems that it work without problems and i dont think that there will be no API or major changes to the final version +Sorry it has taken me a while to get back to this +Its a main feature of a modern editor I think +This trace is coming from content assist +Address all the comments +Thanks Ryan +So you can see getProperty as a fire first ask later approach while propertyMissing is more a +chop files into chunks before copy and stitch them back after copy +Maxims patch is correct I applied it and everything works correctIm pretty sure it must be included into trunk ASAP because of current Hives implementation is Anybody work on including this path into release branches +There may be situations where setting something in the service client is more efficient than setting an option and passing it to the service client +Thanks for the feedback +I wanted to ask first whether there is a pressing reason for it being in util +It would be better to use Seam BOM idea that defanes all dependencies in the separate ll be reviewing this as part of the assembly tasks +New patch fix conflicts with patch fix conflicts with and updated based on last patch does rename cluster to reset while resyncing create a new file named patch does +Closing again +so seam specific version of mysql is this? version and os +Add to pom as Jeff suggests +When you restart the namenode does this problem get rectified automatically +Closing issues resolved in released versions +Thanks Rajesh +if you wipe the clusterIts overkill to do it just for this issue thoughI think I saw it discussed somewhere maybe in a JIRA related to snapshotsbq +This patch provides a solution of the described issue +When I compile myself I only get warnings not +Will apply the patch soon after lunch +Robert reminded me that there is also a heavily broken custom memory estimator in too +there now +Please try this patchThanks a lot +The problem is that you do not have installed +The tests are very skimpy currently it would be good to add some joins unions etc +But probably this just made the Surefire plugin happy so that it didnt try to throw an exception which thereupon causes the. +This bug has been fixed +My selector fails everytime +Sothe application is not deployed propely +Reducing restart time is important for HA cold warm failover performs all or part of the restart + +There is only one way to run from Eclipse should be bundled with SWT libraries and Java should be used to run itMoved to instead of since it requires too big changes +the facet does not control the validation the validation is done by the schema listed in the xml appreciate specific examples of what is failing +ill try and reproduce to see if we got issue before now with proper fix could you take a look? If not then just assign back to me +Good catch pending Jenkins +Tested with ZK on EC +Closing all resolved tickets from or older +At least it fixes the sellitem example to the point where the pages all render and the flow works +thanks manish i have been out on vacation +Log statement for downloading the cache files needs to move out of the checkFile existance methodc +Well guys + +Important to highlight in the release notes as the input names were s basically due to locale issue i changed the type of the variable to String it works for me after that +And I actually think this is very close to memory leak very unfortunate +Kanna thanks for making all of these changes +A patch against HEAD is even better +Thanks for your help! +Maybe its too early for +The compiler hangs on this test case +By the way is this against Beta or latest upstream +Hi BenCan you attach the deployment so I can reproduce the errorThanks + +Since we are not clear how to manage vertices yet Im not sure whether this feature can be supported smoothly in the future +openejb has to be updated to and to use geronimo +My understanding is that you crack open the Spec and type in the elements that need to be specific and either omit things like comments and the like or reword them to avoid copying them verbatim +the build version appears in the modules jar file is absent from the nbm file and also from the update center xml file +Could you please update it? +Thanks Prasanth +I am attaching a separate testcase for your convenience +This one works well on all the changes recommended by Adam +Patch to add capability to specify a in the http Component +Patch attached same as the last just changes the case stmt to an if stmt +we added a readonly znode under available znode we need to exclude it from available listYes true +Too wide a mask to include aspect files +What is a inode? Asking because im facing a similar issue with HDFS +will commit shortly +I was trying to reach a wider audience but after i got the solution i and link to the jira issue here +Testing now +fine with latest nightly +Thats exactly right the FST got smaller but harder for zip to compress so the JAR got biggerI havent turned it on anywhere in Lucene yet +Hi Andrew it looks that the stack trace that you posted is more like the one described in +we had big troubles in past when the version number was present and someone has done a update of the jarslets say the jar was present and the user has updated the jars and the update collection contains the jar now the user had version of the same jar +Note my test case was just to demonstrate another way of mapping +Patch attached for initial review +See if you can repro I have not been able to as yet +interop tests are still working as per junit because they need to create and initialise their own test instances +Thanks jms +Verified on JBDS. +with something more messagedata centric +Robert I understand your concerns and thats why I would like this change to be vetted by as large of a community as possible my current patch leaves an option of sticking with the absolute path for those who really need that +Part all under Committed revision +cposing as wontfix please reopne if you disagree. +bq +For the sake of conciseness itd be cool to move the SRS List out of the demo page to its own top menu item and add the Add CRS option there just like for the other one thing I always wished is a CRS search by area of validity +Its another issue +Perform VM live migration for duser and admin +fix looks good +Andrey in your patch why did you move from toWhile is used in many tests is used in only in +patch looks good to me +I tested with the latest build and issue is not application context to use Regex based patterns to supportreferring +Rajesh thats brilliant +Brett There is a file named in plexus folder but there is no section for the database configurator +another one of those its fine on my Mac its fine on Johns RHEL but screwed on yours andor others +For example those words are extracted wrongspecificallysalvificfirsteffortdefinitivedifferingSame problem +extends and wires in the assertions into junit +Brilliant Thanks + +Patch Committed revision Thanks of the SQLSTATE applies to a specific type of object +Thanks. +This patch adds info logs to both the c and java clients detailing the connection parameters useful in certain debugging situations +Committed to as well +Could it be possible to keep the changes in decorators in the visual theme in a way or another? Thus we could leep the old Visual Theme +to apply to master pushed as well as master +Committed +I use an to force sync +My fix is in the attached fileThe title of the issue should actually be Boo incorrectly handles overloaded generic extensionsI would be glad to contribute in the project and communicate with boo developers teamPlease review my patch + +Shane Bailies patch and the test changes in geoserver +light of this do you still want to do this? Maybe for your use case it is better to temporarily block the syncer thread from syncing in the backgroundIMHO for our use case doing this blocking and then unblocking is extra overhead +Lets do offline log splitting then +Basically the stack lookup method has to be changed and I think the global resource key wasnt working any more eitherI also think we should at not only hours minutes ago but also the future in hours minutes +Oh true +The problem isnt that does not throw the exception the problem is that it is inconsistent with the documented interface +Why shouldnt the snapshot continue to have those source filesSince those files are transient and wont be useful except for making the system inefficient +Results after the patch has been there any chance for the patch to be reviewed and applied? Is there anything I can help with +Need to upgrade to latest version of and SH library. +Hopefully that is it so marking as resolved for now +HiPlease can you upload this maven plugin in the central repoThis is a plugin to manage license headers of +Heres a test which fails after you loop it a few times +Missed during the initial commit +Attached new patch Fixed comments about caching cloning Added partial clone method that clones all except replaces termBuffer startend offset fixed and to use this method Fixed bug in that became obvious once I ran javaccI think we are close +The ouput from Junit tests using the new Driver +on closing this jira +NULL is probably a fine defaultIn case of HOD default value can be local file system sothat HOD can collect logs at the shutdown of cluster as is today +This leads to the problem that the fails +Fixed in issue fixed in released version. +This jira has been closed as part of a review and archiving of the aj project +It is unfortunate that this feature is stalled +I just noticed that this might not be a problem with Java integration after all but with in general +Closing since its quite old reopen if still valid. +Ok this turns out to be more an issue with how I was adding properties to the advanced list than the advanced property page itself +WGS isnt particular great with long distancesI know nothing on this topic but an error of at km? Im pretty certain that there is a bug not having to do with the accuracy of spatial reference systems here +But it would be nice feature to have +This test case seems to be central to the discussion here and it appears to be checking the wrong resultThe comments in the assertions dont match whats being checked +Test failures are not related to this change +I missed the fact that you have to click on the storage to get details immediately after the primary storage is added +Seams not to work for Seam New Action or Seam New Form in an EAR project +Looks just fine! IMHO we should close this issueThank you NadyaCheers appeared on website and fix is verified by submitter. +Can Owen take a quick look at itThanksA new patch with Owens points and addressedFor point Ive created another JIRA to propose to move Abacus to hadoop mainframework +Looks like broke the feature where the was displayed correctly in the logs + push to +add a test in not very clean but that show the failurein the new handler the saveChanges call is commented so the test will fail +Is this issue resolved now? Id like to close this as Not a problem +iOS emulator integration is xcode is missing the latest ios SDK somehow +Good to go no backport. +This fails because the events created via violate the field constraints +Nice ive been annoyed with the output as well! Pach works as expected Lufeng from me too +Yes I can live with that +A manual control might be an interim solution until we reach a stage where the is capable of making intelligent decisions based on how loaded it is. +Thinking about this error I just realize that I dont see how to reproduce it there is no way to modify an existing OC + this would be a great addition to our existing metrics capabilities +they seem to be responsive and accepting it +Should we add snapshot space used to nodetool info to make this more obvious +Resovling fixed by what was committed against this patch and by. +There have been major versions that are and versions that didnt need upgradesstables +For the +Open the stub and search for the term bugI have given an explanation as to why its a bugPls write back if you need more informationThanks you please attach this as a zip +This fixes two more rubinius spec failures as well +When the spec says text node it refers only to text nodes that are childrenof elements not the text content of comments attributes orTest case output partially covers should be enhanced +I do this for pathtocmsbuildtree cd DocumentsI have my checkout of the cms at this location and also the site I am building +This issue is in wrong Module it should be JBPAPP +I will put questions on the jira +We should open an ant issue for everything we need to get done for ant +The column label is circled in red in this image +In as the name implies it seems very difficult to have a taskattempt in the SUCCESS state that had no launch time +Im going to try and take a deeper look at the testing tomorrow +Indeed I missed this feature if I tick out the workspace artifact resolution the snapshot jar library is downloaded from the right repository +Moved to +Integration tests added. +I also updated the git and svn ignore lists. +From some preliminary experiments we found the transaction log IO to be a bottleneck +I have add a null check to fix it +I need to add setproxy that reads maven proxy settings there +Do we need if REST already has internally a Generator API for iterating over results returned by scanners? Did you partially reimplement that here? What about XML or JSONI am on the changes as is +Check all output formats +For this to work there must maven repository somewhere +Thanks I added the trycatch to handle the exceptionsI found the following coding pattern val idx in this file +minor code improvements +I just committed this +Please run some sample test cases and see as well +As it is more and more common to see users use custom class loaders we should also document the appropriate way to not only shutdown but unload Derby when using class loaders +Good work Gilberto +its disk becomes full then we remove the partially replicated ledger fragment +Juri I think youre preparing a new patch so Ill wait for it before applying anything +The ORC reader already handles this correctly by assuming that all values are present +Thanks +The projects build should include the EULA in projectdistributables ie whatever is the unit of publish zip tgz etc +with JBT. +Why are we passing in instantiated Vertices and filling them in? Shouldnt they effectively be iterators over the +This issue may be fixed in release as indicated and is downloadable +Further efforts with the repository facade are tracked by +Actually this seems to occur also for portlets +oops left a line in +Yes as tanukiwrapper is part of the natives +This was a vote to find the preference of so well rename it to dih +i suggest they let the scientist continue using torque and leave the resource manager for PCS jobs +I also added a test that shows this in action +Otherwise Ill commit tomorrow +This is actually the same patch +This does not really help us though as we presently cannot use LGPL licensed code in ASF projectsOleg +OK super! Thanks for testing Chris +patch for enabling ear and adding m plugin +We could try and attach a patch file if you think it might helpWe were thinking that it this change was made then it might be possible to start thinking about removing the thread local +Hi PaoloI didnt realise that the other projects included +OK I updated the description. +Hmmm +Jenkins failure do not seem to be relevant the job is failing for some time already +Any update on this? Should we try and get this to release +Here is the patch +brings up the IRB prompt with any error message +Dprojectcamelsvn commit message renamed to Adding srcmainjavaorgapachecamelcomponentbrowseAdding srcmainjavaorgapachecamelcomponentbrowseAdding srcmainjavaorgapachecamelcomponentbrowseDeleting srcmainjavaorgapachecamelcomponentbrowseDeleting srcmainjavaorgapachecamelcomponentbrowseSending srcmainjavaorgapachecamelcomponentbrowseDeleting srcmainjavaorgapachecamelcomponentlistAdding srcmainresourcesservicesorgapachecamelcomponentbrowseDeleting srcmainresourcesservicesorgapachecamelcomponentlistAdding srctestjavaorgapachecamelcomponentbrowseAdding srctestjavaorgapachecamelcomponentbrowseAdding srctestjavaorgapachecamelcomponentbrowseDeleting srctestjavaorgapachecamelcomponentbrowseDeleting srctestjavaorgapachecamelcomponentbrowseDeleting srctestjavaorgapachecamelcomponentlistTransmitting file dataCommitted revision +just found out theres another +Werner I do care about the history for the files so Im in the process ofmoving the xml files manually +Thanks +Applied +I will try it in a Linux environment next +Also has been in development in open in incremental fashion +We have to check if this one is like caused by +Not sure what you mean +Also caught up with passes with this patch the same as they do with +But I still wonder why its only failing when tomcat is restarted and not with the running tomcat +Jai Its not you +I think this should be included in branch also as without this its impossible to bring up a metastore server on a secure cluster +ve committed this to +the patch comments the save property declarations in xml ve checked whats wrong and looks like my eclipse project is somewhat confused +Reporter took back the issue. +StevenDo you know how I could do this? Is it just a matter of modifying the POM +In addition to Dougs list Id add that a test plan must be published for this feature +You should use instead +Thanks Andrew +Im pretty sure Ive sent it in before but David didnt like it for some reason +Ill post about and when I get those tests first issue I had was with and +Method hiding is another way of saying it but theyre both method hiding isnt really something you do by accident +If the problem persists please reopen +As requested patch against castor +Let me take a look at this and get back to you +Considering that there are bigger fish to fry for CR Im going to bump this again +The issues describe here sound like they are by other opened issues +What does our own grain mean +No way to change the field maybe we should use a different code is in allow something to override this field +Stable annotation is only in code baseI agree with Devaraj that we should tackle this in a method get to help those moving from old behavior +Can we have it so no changing of readers during a scan +Issue has been resolved in r in the recent commit by AnilRishi applied +This was my original problem that was resolved by the fix given in the problem +I tried with the and all is for the noise. +Looks good to me + with german translations for MarkusYour patch is in trunk revision +The mirror does not seem to have been updated for about a weekPlease check your rsync job logs +Closed as requested. +gentle reminder ping +Im working on a small hack for now to make this safe I will have to come up with something cleaner later on +Hi David I just committed +In this version the representation of propertyids is reduced to about potential test for testing backward someone the latest patch and commit it if approved? thanks +reopened for fixVersion +Should definitively have a solid regression unit test for these kind of race conditions within MINA +It doesnt look like the upgrade step is necessary in order to reproduce this issue +Worse case scenario they get when they didnt need it +If log type is Meta discard it +Tightened the elements and changed font of ApacheIf one of my entries wins please credit the original designer with as modified by solprovider +sorting of documentsExamples how to use them will be given in the documentation when is there one + has been attached containing detail of the error in the test +Yup thats the plan +I think that is more fast for the webui +the posgres dump of my eclipse project without the lib but I dont use postgres +Can you try this again and confirm its still blocking? I just committed a fix for accept that makes it use select like its supposed to +Good stuff robert! I guess we should split it up and have one issue for the basic stuff like base ant scripts and LICENCE files etc +There is no duplication + +Id rename Command to +Im gonna attach a patch try it out +I dont understand why it was closed right after an easy repro was posted +Please verify the fix +Appled patch with some changes to compile with the new public rendering API. +Resolving as a duplicate of +Is there an explanation for this inconsistent behaviour +Feedback you +Hi Pat This is not correct +Im currently out if the office for a while but Im excited to try it out once Im back +I dont talk about users working copy +v lgtm +Patch committed to trunkThanks John! +John is this fixed now in trunk with your most recent build plan work +VNC server will then forward request to the guest +Marking as invalid as I cannot reproduce it and the reporter has not responded. +Ive modifiedBut I dont know how to write bat command so it didnt work teach me bat command +Content assistant is being tunedfilled when the visual DOM tree is being built +If that solution is ok with you we can leave it that way or move the jars to the codehaus repo if you prefer +This is not fixable mapping requires that subclass properties be nullable otherwise you cant save other subtypes that dont have the not nullable on a moment though this doesnt make enforces the constraint if you explicitly declare nullable doesnt matter so much what hibernate does Grails validation is not being consistent +It ends up getting to one under than then sits there +Our unix admins seem to like it I think it is inititutive too +This was done for the release and can now be closed +i guess youve posted your comment to the wrong JIRA +bug is that the return type of a method that is called is not being included in the set of types that must be included in the file that contains the call +Robby +I plan to commit this soon +Thanks Rohini +Thanks Bikas! +Since the lib dirs include and does it make senseto use their SAX features? +Problem is still not resolved +to TRUNK +Committed +I have made a small change to as the recent changes didnt always correctly propogate the namespace through resulting in the wrong code being generatedAmonst a few others this affected the testcaseI have tested this on Windows and is now working nicely +Updated patch against probably needs additional documentation in the HDFS user guide +Also can we speed up this particular query? why is its score so costly? +Not wanting to decrease the value of Ubuntu in any way I always said that Gentoo is the most developers oriented distribution. +We are still missing some testcases here +Wait a moment +Sounds good to me +Green light +upstream issue targeted for Fix in +Also for informations sake Im running solr AllNew to Solr here! I have a question for you all on ghgeofilt +Also the sample job for sources and test data have been removed from samples. +Thank you very much +for the jQuery based componentsso it would be no confusion if several components would do the same thing but with a different javascript frameworkDoes anyone have any objections to this approach + +But the real question should be do we want to support IE? I mean its a pretty old version of IE should we have an official list of supporter browsers for the +It is not related see have committed this +Thanks for porting +How do you modifythe? I would add a print statement in the to printout the key that you are using to validate the signature +Fixed in svnSending trunksrcmainjavaorgxmlsaxhelpersSending trunksrcmainjavaorgxmlsaxhelpersTransmitting file dataCommitted revision +Seems it is since this example works correctly not only on jbas with mojarra but on tomcat with mojarra as well. +Added test cases for two additional schemas +The attached patches should cover the requested changes +Thank you Haohui for the contribution +Good to know +log into google analytics and see the stats go from many to and then backup againQE got access GA +Im still waiting for answers to my questions +Feels like solving a to me and Id like to get rid of the Builder stuff sooner rather than later ll admit this is not a big deal +Committed as svn +Were you testing with transaction logs or message logs when the session view didnt work +oops sorry! i may have some that dont fit under your title though +Adding Fix for as well +checked in r. +Ive got large Monday morning backlog in my todo list today so it will probably be towards the middle of the week +I dont know whether this is still true with because Im stuck with another bug which I just filed +I dont think its different for args +Whether this is or configurable was not decided +See last comment +Currently if a build fails due to corrupt m repo chances are good the repin will be fine +Thanks for comment DavidYes that was my st impression too +It seems this new version is excluding the timing issue finally +defer all issues to +So help is welcome for this issueAre the other issues blockers or as important as this one +lets convert everything in one go in If this ticket is going to go in as the current fix version suggests Id advise not using a json pseudomap because it is quite unclear to me that wont require special casings in the code for each property that needs to be converted +In other words We cannot use that stuff +Wouldnt that potentially cause data loss when they both try to not knowing about changes? I havent tested that scenario thats just intuition but I think thats worth checking +The only one affected not changed was master +Having the above would help determine if its a bug or if this can be closed. +This looks like the same symptom as the second one mentioned in right +mark this as blocking because the tests added there will continue to fail when running as part of ant until this one is resolved +It seems that the signature does not have any cmethod +Thanks for looking at the patch Knut +Bulk defer all frontend issues to +A hint would not be enough +Closing issues + +are unique independent of who is referencing them +I imagine theres potential for some unpredictability in situations where one thread removesloads a directive while a parser in another thread is getting a directive +Mark i didnt realise Aron has been assigned this issue +It looks like these failures are not directly depend on Linux distribution but also on specific set of installed software +Mortensons patchfixes the encoding problem +Attaching new patch with testcases for hadoop history files also +All regular often used configuration files have been updated to include a reference to schema file. +Fixed as per Todds reviewboard comments +Integration test added +I tried at least test collections with itbq +First quick test looks good +Helps that I need this too +Just +Any native support for this feature will make us much happier +Hmm uses for its objects so I see why a CME could happen in that respect +Will it help if module classes are sorted at classgen phase end so that outer classes come before the inner classesAny other suggestion that can be tried +namespace in the generic capreq model in the bundle manifest since the spec doesnt allow thatWe need to put a check that will either ignore such cases or throw an installation exception +We havent checked which are already fixed +Some file were missing in the last patch sorry +Closed after release. +the updated patch to eliminate using the switchCase this can make the the code logic more readable +Here is the right one +I was having this problem switched to the correct update site and everything worked +Here is a patch that calls the method on the temporary block files +Comment If you specify a different macro file and it didnt exist shouldnt it be an errorIm thinking there are patterns +Thanks for the patch Dan +Id love some review of the changes +Also attaching a patch for the archiva site adding in the known issues sectionThanks. +Actually it looks like it might depebd on the version of the servlet API you have as well +Branches GACP and GACP flipped to use Tomcat error with tomcat but the test passes +Ill add a handler to at minimum log the messgesDan your implementation looks remarkably similar to mine we must have had the same inspiration isnt itAgain I dont think we need to do any logging in the bridge +Done +however I am a little scared that the problem exists in several other placesmasaya please verify the error screenshot +More fixes and more tests +This was implemented but apparently doesnt work right now +This is an interesting point and one that should perhaps be generalized to Thrift as a whole +See the updated patch +Please reread the patch and what my comment means +Any news +I think this is resolved. +First patch had extra java files +Can you start a thread about this topic on the dev list +Hey Jura could you please drive an incorporation and improvements of documentation for new features in these releases +Patch that implements this functionality +Attachment undeclaredthrowableexception has been added with description Java test has been added with description Original FO file +must have taken a little to upload. +field number class +Im happy for the content to be added but as I am the reporter I cant vote. +Thanks +Yes I would like this patch to be committed +So what do I have to put in the POM to make the order alphabetical +That is a great idea +Yup +Looking +Added logging to testhlogsplit + can make use of it when we are not able to find the get api from as a else condition +Only the beta version has this bug +postpone fix to work started on this yet postponing to Bobyqa algorithm from prof +Can you explain the other changes? This seems to add quite a bit of complexity to the code is it worth it +Additional patch to move the method for removingretrieving a boolean argument +Or maybe try the dynamic resolve then if there are still missing dependencies resolve only those with the default mode and if things still arent found then fail +Good thinking! Sounds like it will be on the CS quiz +The fixes will be similar +I typed an exclamation point in parentheses +This patch is applicable to the trunk +To override this add this property XXX to the service startup command +SagaraIf you look at the +submitted for hudson tests +Also I can imagine that requirements would differ from page to page +I see this issue is fixed however Im still seeing this issue with alpha of +Samuel please reopen the issue if it doesnt work with Tools +I think the ideas of the logic that reduces cache locking make sense +As stated the comment in file the mime types there are taken from Apache httpd fileI think this list is enough for most of the applicationsAlso I could not reproduce the code to reproduce from this JIRA on latest classlib svn snapshotSo it seems that this JIRA could be closed as not reproducible +hmm i failed to reproduce it on dummy project too +that link i gave for our personal repo is on codehaus +If it is less than upgrading your Jira instance before you do the export would also save a lot of effort +Merged this patch the new one should apply to revision +No problem I can easily change this but this will effect a lot of code +I do not have a suitable test bed to test this currently +Patch for this bug filed in. +How to administer and monitor EDS is in the documentation +Makes sense +we dont care which format as the collection reader accepts either +ThanksI created issue for upgrading to the next release once it becomes available +The change caused regression is bfceaffcebbde which was pushed on May +Ive requested that the original user also validate that the problem is fixed +In other words I think we should revert those changes and leave it up to the user to decide how they handle their bytesIf someone does revert the change please add a comment about why no encoding is specified so that future code reviews build on current knowledge. +Seems to fix the test problem +Would you mind removing the existing attachments I included in this Jira? I dont have the permissions to do soIm assuming that you guys dont include contrib code in the jars that are deployed to Maven +push to +Most important is to be able to do runtime link with other componentswhy you have lifecycle in mapperreducer +So if is not writable a delta import will always do a full import regardless of whether a full import had succeeded before it +Before I could run tests with a single shard on independent hardware now I can do it on the entire index because the offline set is built independently +Please verify and close this Jira or provide steps how to verify it +Add label fro those form accessibility issues about lacking label for those input boxThe patch is for pluginsconsole to trunk as RevApplied to trunk as RevApplied to trunk as Rev + push out to s a patch for this issue +The Runtime detection will find all runtimes and if there are more than one it will offer the user a dialog so that he can choose what to install +No column in a primary key can be null +Actually I should say thank you +Thanks! Cleaning it up a bit and committing shortly +Fixed +I will change the patch to use for Base encoding +Going for a new run early this week +RobDenny is it straightforward to make the changes provided from to the BPEL deploymentproject structure +I was not able to reproduce this on a recent build +LGTM can we get somebody else to and then Im ok to commit this if tests pass +If it is info level it would be good to rephrase that in a way folks would be able to understand otherwise we have a log full of really esoteric things that only make sense to us + +patch looks good. +Committed to trunk with revision This only adds a grammar class which will be used by the test +multiple components are using asynch instead of async not a revert this +I guess another idea we could use is to put the inside a class loader that we can discard at some point +Also the logic of Platform dependent code was separated to a new class called to Alex +is removed +Anyhow this is still a problem in +Since the and dont have weight attributes well currently discard the weights in ARFF files +Rebased for trunk and applied the following changes Removed the private Avro schema that was in left behind the generalization Instead serialization of CF uses Cf Switched back to a column per keyspace datamodel in the Schema cf with a magic column name to store the Avro schema Squashed and to Removed Ready for another look +even if that does not make sense if you distribute something you then want to unpack +Updated with tests +Another issue this would have to be done in conjunction with First Result and Max Results +it could be although snitch is separate from strategyI can just sneak that one in at commit time if you are ok with that since its not anywhere +This would be a case of providing a different set of contexts from the HTTP server at this point it is not related to servicesIt would be fairly simple once being on a slave is detected to display an error warning message if anyone attempts to connect to the console in this mode leaving the HTTP management context still available +All tests pass +Plus we should define a new schema +Hmm isnt exclusivity of test providers the norm? For example you can either use or +I send you patched version of Forums do you use? Im asking because all versions of Forums available in downloads no longer use those because Forums where rewritten to JSFFacelets and have now +You should therefore either create the table by hand or start Infinispan with a datasource +Amanuel guys We renamed the project +The Bridges should be of the sort and should implement its own andI believe that the implementation is fairly trivial +looks good to meAre you folks saying that the approach adopted by this patch is not sufficient and it needs more changes to make it efficientCurrent fix is good enough for this issue +We should get rid of the last argument to the function +There is literature on scalable or dynamic to do this in a mathematically sound wayThanks for the links +I personally test everything on a live cluster because the world is full of gotchas +Vincent your last patch is not marked as okay for inclusion nor marked as patch submitted +on patch +Latest patch no longer applies to trunk. +Make a snapshot version number +its not ready right? Moving to beta in this case +Hopefully the will give us the possibility to compile using JDK +Let me think about to make the region server throw a region already in transition exception in this case which is not accurate to throw a not serve region exception +If you for example do def foo in one script and println foo in the next script then you will get a unless foo has been made available through the binding +I was just writing up an email to the hibernate dev list about dropping for JDK since it has now started EOL with the release of JDK +New Diff file this sould do the job for initial revisionIf not i have to do the diff on the linux system the other svn app does a better job i guess +Ok so just add a reference to chapter +I do not understand how using scrambles the text +Just a reminder to myself to check for async support of this new LB +They have an empty constructor or may have constructors with configuration parameters +Thanks. +It would be interesting to know if still hangs for you on your platform +committed to CVS HEAD Oleg +I also see by default will search for schema files in input directory is that intentional? That will add burden to namenode +Saved Resource +FYI I really dont want to start messing with the host component to get the skin working and i disagree with the proposed behaviour for popup positioning so im not planning on doing any further development on this +Forrest documentation is attached +Each represent the full documentliteral wrapped means that the parameters and return value are to be wrapped in a schema element is what you want +All tests passed on my machine +However IMO we should reuse mimej in the next version anyways so that can be ignored for now +I will report back my another note it would be really nice if the implementation throws an exception when any concrete interface of this repository defines a save method +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Even if were the case the image exists on the producer not on the consumer +Interesting no one commented on this behavior with the current Roo shell +Im marking thisas an enhancement until shown otherwise +After that we just have to wait three days +All patches have been applied library builds on Solaris with Sun Studio for me +Nathan the answer to your question is yesOk sounds fair +Fixes decimalduck typing applied +removing that line is wrong +i am closing this as wont fix you can do it yourself very easily by implementing the interface +Because you would eventually end up with a reference comparison +Was version +These commits add also quickstart +Im changing both messages as part of and I will remove the blanks then +Trivial patch that checks if the log is debug enabled before logging +No further feedback from the user closing this issue. +Its minor and the reporter found a workaround. +Thanks for looking at this Mahadev +Updated fix version to next CP and fix for versions so that this can handled in the next is not shipped with the EAP AS at all +I wasnt able to reproduce this using trunk yesterday +This patch applies to only +To avoid being trapped to the same issue and case of Kill Crash we need to use another extension like +this occurs when running site goal works fine for javadoc searching I found that the problem is caused by or and dont cause this failurethe issue is tracked as +After talking with Justin we decided to impersonate the session with the new user but that breaks down if the impersonated session is closed before the User is returned +I dont quite see how it adds confusion +I just committed this to trunk and +was probably within a different onehttpclient is indeed the only way to currently handle https however it has absolutely nothing to do with pdf as it is about protocols not content +Junping DuI have been looking at some of your patches but there is a lot here to go through and it is likely to take some time +Please open a new bug if you have any more install problems. +This may be as simple as adding Host to the EXCLUDEHEADERS map +Every entry added to that system property slows down class loading for everything and many profilers require different values there +fix with volatile valid for +The term annotation is ambiguous it is either type or reference to it so that it is not redundant in descriptions +Should be fixed in trunk +Fix on the way +If your application is not relying the original behaviour of listPaths then this change makes sense +Code changes were uploaded in revision configuration will be handled through the SOA issue. +Patch applied and tested +All script sin the source tree have been renamed to use this extension +In fact the patch I just applied fix only the timeout used to see if the container is started but not the timeout to see if a deployable is deployed +I cant reproduce this +Both forms might be useful +This bug has been marked as a duplicate of +Sorry for reopening this issue but the new release again overwritten default file +Attached patch to do same + +Updated patch with additional test for release CONCAT support +However this is a reasonable request and we should do something here +I think this is an +Scott is working on a build of branch which pulls all libraries from the build repository +This issue is fundamentally about backing out changes related to since youve rejected that patch +Groovyc by itself is not capable of loading such classes +That means this fix will require a release of Maven Shared unless someone is planning one soon we might want to move this issue out to to avoid delaying +I would rather prefer usage of an existing library then something written from scratch +Will try against dont believe this is a bug +Seems to work fine without sdo so ive removed the dependency +Will commit if tests pass +The example for Flex Plugin does not have code coverage enabled I wonder if it is because you guys are using and not as mentioned above +However the change is extremely important for supporting tooling across an unknown set of rd party tag extensions and is just generally more amenable to extension going m happy to help out in anyway possible but I think a better solution all round it to look at how can build on the new XSD code for all round consensus here is that we really cant change this signature back +I have looked a little bit into it it seems not trivial see I must have been blind not to find the original issue which Ive duplicated +Yup if the znode is still there then theres still a running node so starting a new node is ill advised +sure I am adding this to my TODO list will probably do it next week +Patch fixes wrong handling of proxied entities +java file being saved +Blackhat changes the properties +Just to make sure did you get mvn test work as well + +Interesting presentation +Also flixible for adding additional Spring Social Templates to the configuration for more options +Would this allow the reactor to wake up in a timely fashion when there are interestOps items in queue yet still allow the reactor to sleep when theres nothing to do? Yes sir +There have been no changes in saaj and jaxrpc +Sorry I missed the fact that CS seems to be fine now +Need to update release and include additional dependency section in parent +Committed thanks! +I am providing both the complete Java files and the patch diff files +Not really +If a new session is not created then the subsequent request for the timeout page also has an invalid session resulting in an infinite loop unless the user explicitly exempts that page from the filter chain which most forget to do +And we will not add anything automatically if user not need in it +we got thru put of tx MamtaYes that lower throughput for sequences when preallocation is set to appears to be the cost of the correctness work done in +We now detail which putsgetsdeletes failed and the client code takes appropriate action to retry appropriately. +Patch for trunk +Bug fix patch attached please place in front of your classpath +for the patch +However the inner class did add a small amount of complexity to gain that testabilityAt this point we can modify the Log class slightly to maintain readability +Replace vector of boolean with passed +Fedora Packaging Guidelines require all software to build from source and all provided tests to be run +Good point Indeed should use the bean class loader ve also added explicit bean support to other affected classes within Spring such as got a chance to test this issue this morning +Patch for trunk +Were still dealing with this +When i hit the time ill need to upgrade spring security i will investigate it further and possibly ask to reopen you for your efforts im just trying to make this great library better +Im okay with not fixing this is working fine for me. +Is there something that I can help you with or something that needs more explaining or instructionsLet me know +The patch would not apply if it has the patch like acontribdataimporthandler +Isnt necessary lets stick with current ones. +Unscheduling from put back if you feel otherwise. +It looks like this was never fully done +Ive read the code we should rewrite this class +How about putting the key into the client module which is the smallest +Assigning to Patch Triage issues incorporated by this for justification this is just the container task for the JBC upgrade that solves those issues +these are the three problem issue I see is an in the JSON response +The Portlet App Manager details portlet is displaying but not storing portlet descripton information +should be moved to examples package +Ive tested more since and havent run into the issue +Same patch but done with cvs diff u for readability +Its that is buggy +This needs a careful review before becoming committable +is a much higher merge factor than is expected by the design +mvn N versionsset have asked versionsset to be executed on all childset is smart enough to update all the child modules on its +I have the request bodies from both uploads and both seem perfectly fine according to spec +The issue is that update is now being called after the is initially constructed +Merged to the branch in revision +replaced the bookshelf with a globe +This issue can be fixed by modifying to run chmod command explicitly but it will make the deployment time much this issue can be worked around by changing the default umask of the user +Wed need to see a testcase or youll likely need to debug in yourself and submit a patch if you can figure out what the cause is +I have applied the patch and was able to do mvn clean packageSome cleanup things on this plate are the icons folder in plugin +I also see the messages getting dropped +A Valve issuing new Session ID after visiting a container form +doesnt define support for the GROUP BY feature +OK open doesnt seem to throw anything +Do you have pointers to the conflicting bits of information +Doable for imho +Agree on the remove ROOT suggestion for the current scale +We should at least show its not slower +As I have already stated reverting is not an option +I am for committing this +It can wait for the next major release +It works for meCould anybody please verify it on HOD +Here is a patch +James Oliver Sims Ashish Jain and Mark +Screenshot of the Oozie web console showing job info See the Console URL in Hadoop API for getting job url +this geotif in combination with httplocalhostgeoserverowsserviceWCSrequestformatgeotiffversioncoveragegeodanCorineresxresycrsEPSGBBOX a on this Steven you are NOT requesting the full resolution with this queryhttplocalhsotgeoserverowsserviceWCSversionrequestcoveragecorinemergegtiffprojcrsEPSGBBOXresxresyformat The envelope is not native one it seems that you copied over the envelope shown by gdalinfo but that is an approximation for visualizationthe native bbox is the followingBBOX this a request at native resolutionhttplocalhostgeoserverowsserviceWCSversionrequestcoveragecorinemergegtiffprojcrsEPSGBBOXresxresyformat priority to feedback from gdal folks + +Please give it a good test with Jini services that have a attribute +The Hadoop QA should pickup only the COMBO patch file +Quick question what should I use for the value version +Test Create html output with frames and review it by myself +The immediate problems have been fixed but documents an overriding issue that remains unaddressed. +If that makes sense we can follow up in a separate jira since its not a new issue introduced in this changebq +This also ensures that all writes are preallocated even the first one +This bug was fixed closing this issue. +You can release this now +This is a hotspot issue +Im going to change the status for now but please feel free to further flesh out specific changes youd like to see and how they would benefit users +Can you +still has a decorationName field while that is no longer present in the final model +This will be fixed via that will track multiple documentation changes together in one big patch +Same patch as on +But we are not testing a decimal valuePlease let me know which of these is a good approach or if there is another way of doing this +A few minor comments +But for now Id rather just advertise RHEL is broken dont use it just like we do for +We also tried at one point duplicating the server name widget on the runtime page but that didnt always make sense since sometimes you may be making a runtime without creating a server at all +If I were to change the sorting algorithm it would break compatibility with existing serialized data and it would not be a universal fix either as this type of bug could easily come up in other similar structures +Andrew I tried out trunk and didnt see any problems other than where the broker plugin files go +I have committed this +That said I support that change +Nothing has to be added to get the files to compile +If we want to use thrift we probably want to use it for both producer and consumer +Committed a fix to trunk and x +subject to Jenkins +We are implementing Journal Cleanup preventing the between negative and positives on the journal will also implement Journal Compacting as part of +The issue is resolved +Do you agree? Am I missing anything +My best guess reading this stuff is that Python is having some problems loading +unfortunately this is right now for me a mandatory feature +after release. +Compile classes from tools folder of niochar module from contribution package +Im delaying this as the tests need some polishing + +Cant try until I get home tonight but I am also leaving on a small vacationshortly after I get home from work today and will be gone until next Tuesdaynight +Will post is as soon as it is done +Probably +This is confusing and the reference to should be removed. +bah +Thanks Andreas for your patches test case +I tried the tested in question on trunk and it passes in eclipse +Safe mode will be turned off automatically The number of live datanodes has reached the minimum number +I thought this would be taken care of in the same manner as the support +Thank you Uma +The issue as I understand it does not arise with current browsers so without a unit test I would be driving blind +Youre right this is working +Update your root pom to use parent pom version DONE +Committed patch revision which updates the information in the manual +Were on a december th build +Follower starts following writes some new transactions to its log and is killed before it takes another snapshot +The problem was solved +Good question +Fixed by +Is this a standard? We have to be careful when we change behavior of already existing seems like the right thing to do but their could be unexpected surprises +We just need to inject a threadsafe logger +Should be able to get to this in a couple of days +It is important that the tasks are scheduled and executed rather immediately +bq +perhaps not +Hadoop was released today. +Patch looks good +The workaround works fine if you havent any resource defined +The snapshot of unpacked SNAPSHOT +It works now +Queues properties randomly disappears +Please attach a patch to this Jira +Just updating the target fix release to be consistent with the original bug report +Now remote forms work fineThanks! +there are some other cases mailto news url parameter with space unicode charactersI think it is still far from the full list of all special scenariosattached is my implementation some custom link transformation to handle more cases +Attachment has been added with description console for bad communication +Made the memory tracking changes as described in my previous commentAll tests still pass +PingHow is this going + +The merging of layers would be a big plus +Thanks Kathey +Resolution sounds like itll be pointing to Math in the javadoc of Lang Fraction +Observed the same issue even with the Xen Setup +Lets start by adding the comments here as to what we need to be sure is covered +There wont be a release +ThanksI have to make a new release most likely over the weekend +No problem +Let me mark this issue as cannot reproduce then +Whoops! Thanks for pointing that out Hari +If cookies having the samenamevalue and coming from the domain but different paths are received they arestored as separate cookies and subsequent requests will send all cookies toserver +This patch also adds tests for a few test cases that were missing +runConfigure script as described in the docs? Did its output look acceptable? What version of gmake are you running +merged and pushed to and in JBDS Build id +Please file a bug against the Flash Player atI will see if I can create a simple workaround +Anil in order to guard this a little more defensively I would both check to see if in is set to Y AND height length and width are all populated as it wont work if any of them are missingAlso you put the in value into the itemInfo two times where one is sufficient +I just committed this +VPE has some classes for watching tld declaration on content assistants use it to listen to any changes on these Exadel Studio we used to have a few different places with similary code for itContent assistants used one code but VPE used refactored it and left only VPEs we should have move that code in common place to use it for CA VPE +This sounds like a bug with your JSF implementation rather then with +jliu made a profile so we can run the tests against modeshape instead of jackrabbit +to be checked + +The Gump folks will holler and we can tell them to use Maven instead and that should be that + +Committed +Will the above work if you are not using a users or authorization file? +Thanks Dan +for the patch +We shouldnt remove the floatint methods +If you want I could call you to discuss some details +Probably need to make OS and runtime specific setting for configuring a nearly OOM situation +Thanks for the patch. +In my app users can manually reorder a list and exactly this ordering must be persisted +I think we should put in this patch for and set the write timeout to a value lesser than minutes +Disappearing outline view is a sign that the groovy parser could not cope with the incomplete code +Thanks for the heads up +passes now +Applied the patch into trunk and branches. +The problem is that the parsed path is also used to look up resources within the CLASSPATH +If anyone finds a test case which can confirm any usefulness of the removed code pls revert the patch and add the test case +A patch to the which fixes the osgi manifest issues +Just login and changed the passwordThanks +So this is not a bug but a featureDid you actually try running java jar ? It works fine for me +That should be that the function has a hardcoded interval of . +Ran into this today +does this bug exist in as well? If so can we get it into branch too +Somtimes they are prefixed with do and sometimes there is a suffix +Heres a patch for it +Shelly is the tag ready + +Running updated patch through Hudson +I have committed this +Does this happen if the reasoner is disabled? Then it is a known bug +Ive applied the patch +Thanks Yi Xiao for the patch +Thanks a lot Dennis +I moved the annotation to +Added new version of +Equivalently we could just increment the counter in fetchOutputs rather than have a member reduceWe need to have a counter accounting the number of bytes FETCHED for each reduce the end of shufflingIf the compression was turned on that should be the number of bytes of the compressed dataWe should also estimate the compression ratio of the fetched compressed data and report it somehowWe should also report the number of segments and number of bytes written to the local disks at the end of shufflingAt the end of reduce we should know the number of records and bytes to the reduceThat number of bytes may be different than the number of fetched bytesIts ok to record the counters for successful reducesIf shuffle you should have the numbers of bytes for both the compressed data and uncompressed data +Moreover the registry now contains interceptors as well +It has no need to be updated for this its just a matterof overriding isVisible and using the right message filter +Resolved by +rebased +The b patch looks good to meI assume this part of the patch was unintended thoughProperty changes on javaengineorgapachederbyiapitypesAdded svnignore +Passes local tests Please review +This seems like a reasonable new disagree +This is the WSDL from which the source was generated using the following command uri p is the generated code +Thanks for the patch +We are going to include the necessary file in the binary distribution +prefixed GRIDSIZE key with +There is G or available RAM no configured swap and the heap size is MB +Fixed in trunk and maintenance branches +I am closing it for now as all the bugs I can see are fixedThanks Daniel for the excellent test casesIf the spec is clarified the other way we can reopen it and fix scenario one and three +Workaround to get hawtio backosgiinstall s mvninstall s mvninstall s mvninstall s mvninstall warfeaturesinstall hawtioLooking into the root cause now +Though looking at it like that Im no longer sure on the math +This is a simple patch against source code to make it build for JDK as a target platform +from me but I hope Rick can also vet this +aj files into the +The right location seems to be +Hi All Sorry for the delay Im on leave for few weeks +r Added the ability to publish fb library along side libthrift to the apache maven repo. +The UML identifier domain appears in the OGC document chapter +Committed also to and branches +Do I need a CHANGES entry for trivial things like this +Stack thanks +Hi Nick +committed +Looks like is a temporary fix until the more involved is completed +Increasing vinods comments +I tested it with the entire spring core training workspace and dont found any or out of memory problem +hi Olivier all the functionality tested out +That could work for a might have issues if you had any kind of passivation enabled or storing of sessions to disk on undeployreload on deploy +Problem solved +seems to be no Wicket problem according to the t we omit the check if no maxSize is set? +RTC task bulk close of resolved issues. +This is because many of its connectors require NTLM +Might be nice to add some finer grain times in there for atleast submit launch +This bug is two year old +Applied Juergens suggestion +This was probably just a minor oversight from +is still but the counter value of +A unit test case would be even better +I want to remove the unused parameter conglomId of the method get in the interface Does anyone see any problems about doing thatNote that there is another method of the same name and with the same parameter in the interface +The will be closed when the operation returns here +Tim if you want to take some ownership on coming up with tests or guiding others +Instead Connection thread detects idle connection itself in wait +I would like to close this bug +I was still on the old version +Ray Im afraid this is a type of issue +If you need additional classes you install the specific smaller libraries in your webapp +Here is the complete stack trace +Im not familiar with the JDO side of Castor but having browsed the documentation I agree that this would make the handlers even more reusable +Thanks Dhruba +But alias problem was quite difficult. +Patch looks good Shai! Thanks +Can you elaborate what you mean when you say The specification for the and states that implementations must handle redirects +It should only be changed if the schema has changed +Patch and test ve applied now a slightly different solution where a provided is only wrapped if it cannot take all characters to unread +I just committed this. +Thanks Hong! +Closing resolved issues for already released versions +This patch made the change in Checkpointer as Konstantin suggested +Consider this flak for adding any more runtime jarsPlease dont do that especially for functionality that only a subset of users would ever use. +for with Jenkins for the patch +this is using eclipse think type and classifier and almost the same least Ive found no definite description of both that explains the you are right in the workaround +removing need to access externally to webflow +IMO this is not really a bug but i will leave it open in case we want to discuss it my own JIRA as thhis isnt clearly a bug and we have other more important fish to fry. +Sorry if I added to the confusion +At wich stage did the issue happen? Maybe after having run tests of a previous module +Then we got new bugs from folks using Tomcat which triesto parse the file according to the standard and failed due to our changes to accomodate BEA +Applied +If youd like me to move it back down though thats no problemCheers thing +The cache was defaulted to the which is not valid for use in a cluster +Simple unit test for demonstrating not called +Thanks! so basically I should download the latest trunk on the website right? I gwill give it a try asap thanks again. +In addition it should not be too hard to assign all possible fields when creating a Jira issue +Should it not be ? find the root cause of the problem +I was wondering if I could get feedback from someone on the Seam team about the right way to implement thisWhich would be a better approach trying to get Seam Validation to play nice with Rest or trying to get Rest to play nice with Hibernate Method validation +Thanks for the fix! Committed w minor change to emphasize that subBatch is only created once +The documentation created by DOXYGEN is outside the scope of my document and needs to be addressed as a seperate issue +This has been applied against trunk +re previous discussions +But its name could need some adjustment then Otherwise Jukkas quick solution is probably good for now since this is most likely a little bit of work. +However when treating Groovyified Java programs for s purposes it is unwanted behavior +Upgraded the dependencies in rev must wait one day to let the nightly builds run with the new FUSE dependencies +Inconsistent behavior is being seen +patch with still comment Bam screenlet navigation form to help global reviewI will remove all before merge with two warning in previous adaptation after branch update from trunk +Perhaps its all overkill +ok im reduced update frequency to timesday. +patch +Committed revision +A region that had just recently moved probably has clients that still are stale when it moves the second time +This looks like a solutionRobin Taylor +Thanks Todd and Brandon +This patch changes only testcase +Im prety sure it is not a visual problem like the issue with combobox with there are definitely at least two different issues +I cant apply to TRUNK till goes in +Added unit test that tries to get all Workflow Instance Metadata from Lucene Workflow Instance Repo +Could there at least be a hook somewhere to turn off the backuprestore process? We are developing an enterprise app where we are in complete control of the iPads that are deployed to and can stipulate that there should always be ample remaining space so iOS will never run low on space and clear up the librarycaches folder +A fix was applied in subversion revision +Closing as a duplicate of +But then the only thing I see is that it registers an handler for https which is the same as the one for http +the tabs seem to have gotten worse +Is the issue perhaps be that previous tests have written to the log +As the resident KVM expert any ideas on how this should workThanksDave +Of course just assigned it to me +In the future please generate the patch from the root directoryThere is one minor tradeoff with this patch that I am deeming acceptable +ready to commit +Workaround for non usable HOME in tests +Marking as regression +Migration done +Logged In YES useridIgor Thanks for pointing out the lego script for me +Click +However I discovered thatTotal commander works in the exact same way when configured a HTTP proxy server +Otis is the patch already applied? If not from me +Committed to TRUNK. +Especially since this anomaly is being discussed by the EGTommy open a forum Thread and well get other EE EG guys to weigh in? +Ive been assuming that deleting a group there leads to the LDAP group being deletedSVN and references deleted for ibatis btw +Applied. +Please either make the change on top of the attached patch or make a note here if there is nothing to be done and I will strike the comment before applying the patchI removed the comment in about changing the default option to TRUE in the future +I presume this is a dupe of? +Applied Matteos comments +I see thanks +This fixes help i want to ask you about possibility to store original file name and mime type of uploaded file to some target ve done it temporary for me but this feature can be useful for other developers too +this patch sets HIVEHOME if HIVEHOME is not already defined in the environment +Unloadreload the malfunctioning coprocessor +Add a new class bug in unit test +I think we are moving to guava +Yes I agree logging should be in Weld X +If this will not be sufficient in the future Ill open a new ticket + dlls are in lib if want to reference them explicitlyOtherwise Ive included the tag of release as projects for my individual troubleshooting +Closed again now properly recorded against Fix Version as well +Now that work is well underway with AS all previous community releases are +Thats allTo look at it from another angle a deployment which isnt failed can become failed by removing a deployment it depends on +It is nowhere near the center of my coding world +I realized late in the game that I was pulling out of stale repos +OK committed on as well thanks Simon +resolving per Jackys comment. +is just surfacing this error not causing it. +It appears to be happening in both the master and the region server +Is there any other module generating? Wouldnt make more sense to remove the PDF generation for now +Did you ever thought of supporting fq parameter for the Join like how we support for v for query +Attachment has been added with description example fo file that shows the P open bugs to P pending further provide console output as an output file as well as any output PDF exception is thrown when processing the provided sample fo file as such no PDF is generated +Ill be very glad to help with this issue so please if you guys want to get access to the source code let me know +Closed upon release of. +is a patchfile for sIt no longer uses regex parsing the extract the constraint names +Wont fix +This moves the check for accounting space into collect and makes the spill thread persistent +Thanks for taking the time to submit it +If so then I can probably maddle with how column mapping happens +Available in the Castor for reporting it +patch that adds the variable to the top of the list to assist in forward mergingTested that this change merges forward all the way to masterNote that in the and merge there were some unrelated conflicts in the test code related to +Closing this as dup for + +Also fixed any methods that take tablenames but didnt throw when the table didnt exist same thing Eric did in the version last week +Would you please try this oneBest patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +has been created to track NIX +Hi WillemI will try it and will let you know about the results +Should add regions +Yonik and Doug will know ifhow things can be set up there with EMMA jars should we include this on the site +patch attached +Board resolution has been added to the June agenda +Still no Feedback from the resolving this as invalid otherwise dont mind to reopen it. +This ensures thepatch will get checked in actually +Whitespace removed try again +At minimum it needs to be tested against oracle also +unassign for now +Attached some trivial cleanups to in patch aCommitted to trunk with revision + +I will take a look at your source code and let you know if I find any difference +Patch provides methods to determine if snapshot is runningPatch also modifies client slightly so that startstop functionality is as expectedThanks +Thanks I confirm the fix works. +Your suggestion? +This bug has been open for months now is there any plan to get it fixed before is released? +I think a few years ago that may have made sense but at this point being anal about MB of RAM considering the performance gain seems a little dated +the patch is generated based on +Committed revision hashcodeequals changes along what miller described +But because you have started publishing mvn artifacts with clients of zk should not be having a problem about the number say +any comments on the branch? if not i reckon ill merge it soon. +Hey FlavioHow would I print the content of the synced set? Sorry Im a bit new with zookeeperCheersAl +I do see the leaked connections in and I dont see them in +The latest build now has line numbers in the jar so if you reproduce would you be able to attach a new stack trace? It will make it much easier to debug. +Sorry did a patch in wrong way +Adding the proposed change as a patch +Of course that doesnt mean that equality cannot be coerced to hold +IMHO it is not fair to the operator to have to remember and additional step in a scenario that can happen to anyone specially when several people in chatroom are dealing with tombstone cleanups situations and changing gcgrace has become part of their routine +For some reason the patch linked to changed was wrong but Ive applied the correct one +I am taking the following approach let the RS report opening progress but master will ignore them for the first cut +Options to run and get the log foo Xemopt ctirdumpdotdump am sorry for the inconveniencethe comment is a bit confusingBBP version inserts bbpCFG at all backedges ommitting those which are already interruptableI suppose that in the example the loop body contains only one basic block +Hey Nicolas +maybe we can have more strongly typed types and include things like formula date etc +This has nothing to do with cross site scripting +Derbyall was run for the attached patch on version MPlatform Java Version Java Vendor Sun Microsystems IncOS name LinuxOS architecture iOS version Fails derbyalltoolsNot related to this tests for empty resultset +We add node B then bulk load snapshots with my schemeThe snapshot of A with would send that row to A +Closing. +For some reason Maven couldnt download these from the JBOSS repo +So it does not fit our continuous builds since p metadata get updated on each s why I suggest using it only for stable sites +to set fix version to again with correct fix version +Looks suspicious will have a look +I was hoping you could teach me by answering my questions! I have seen this problem only occasionally and not since I upgraded to the latest build +I suppose it depends on if I work out a concrete plan there or not +Not the prettiest thing but it works for now and well be able to clean it up a bit when is released +I just verified +One comment on Brians patch it would be nice if the use count in could somehow be decremented when references to the go away but I believe the patch is correct as it stands +I was unable to reproduce this issue on trunk or presumably it was resolved during the legend reworking around the release +The other problem is that given X as a parameter if we limit the the allowed targets of setProperty then OGNL would fail to set B in A +Attaching slightly modified version of the patch removing from toolsantproperties and correcting the workaround in +Interesting +If theres a possibility here that the would fail because mbean is not deployed then you should not have metric be collecting by default +Is there something else that you are seeing +In such cases the imported files would differ each requestSo I was thinking about adding a performance FAQ and pointing to some tools or info on performance tuning +Imagine a network failure wherein the can no longer reach the active NN but can reach the standby +The patch looks good to me +Ill search the forum first next time +Restarting TB did not fix it +Id be on this particular change +That would have helped tremendously +I have not seen it fails since my changes I ran it several times successively +Committed branch and trunk. +Thanks BobFixed in a slightly different way in trunk at r at r +The compiler swallows it even when actually compiled against the Servlet API so Im not sure why we see failure at runtime hereIn any case Ive changed the signature to Collection in order for this to be consistentJuergen +Sounds reasonable +Can you try this on AS and see if it persists? I have not been able to reproduce the issue +Committed the changes to trunk with following commit comments ALTER TABLE DROP COLUMN should make use of information in SYSTRIGGERS to detect column used through REFERENCING clause to find trigger dependenciesDerby at the time of ALTER TABLE DROP COLUMN looks for trigger dependencies by looking for column being dropped in trigger column list but that is not enough +Is that right +toestand would be better but still quite confusing since this is really a tech term thats not usually translated IMHO +if the codebase has it on Ill just regenerate a local should be fixed now +Im attaching the This is not a working patchBut Im attaching this hoping this will give some basic idea of how the JCR structure would behere there are two kind of permissions +Flavio I believe that timeouts can be an issue and that ms is maybe a bit aggressive +Moving out of release +The failed tests are not related +Libor so did you manage to reproduce on your platformbrowser too +HelloI added the wished demo of my application +Whoops +I will +now compatible with ruby like the patch was not compatible with ruby was the issue initial tests against ruby look good +The performance slowdown I had was on ubuntu with +Closing rejected issue. +Common components between a Service and its download client jar need to be packaged separately such that both the Service and client depend on this package forming an indirect dependency link between the client and service +And now we are deleting them when they are no longer we need some specifics about when this was happening +Ive committed this thanks Patrick. +new version should work against hdfs +We do not need to redirect them to another pageauthenticationredirect can become authenticationpage to keep the possibility to change it +This injection works in a pojo running BetaMaybe something with +most recent failure with original patch applied +Hi Marko +Stephen would you assume that a call to versions followed by a versions would be equivalent to a single call of versions trueHaving a new mojo would be more readily apparent to people using versionshelp +there was discussion about this feature in relation to the ear implementation see TODO for missing features +Will this be able to make it into +The jar plugin creates a correct classified jar file +Passed in to StagingProduction +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Adding an offset to line symbolizer may work but it would require rolling out non standard SLD +Tested +Hi RahulThanks +previous patch made progress but now triggersFATAL failed assert This attached patch deals with the new non single segment code path +More advanced analytics needs a more robust server side solution +langjavalibpy lib to use this patch +Same as what Jon said +Committed to and trunk. +New formats are handled within dedicated and +For a little background the problem was detected due to a NN token issue +The problem is reproducible with a bean that implements Serializable and contains a or as its only class memberYes I realized that I do not have to have an explicit serialization step if I wanted to use object serialization as the wire format +to update jsfhtmlsingleimagesvisualpagehtmlsingleimagesvisualpage other updates are necessary +I will mark this jira for federation branch +Thanks Jeremy I didnt spot your report when I added +Im assigning it to him now but if you already got something done on them please let us know asap +I am assuming the tests arent run as part of the build script? I cant see there output in my consoleIs there any easy way to run the test from inside eclipseSorry for the hassles +Thanks for tracking down the location of the bug +Committed +Let me know if you get the same behavior with +OK sounds like the current approach is an acceptable solution +Fixed with revision thanks! +SCR has been released +Here is a workaroundIf you delete a bookmark folder you should also remove from your blogs page templates all showBookmarks and showFolder macros that refer to that bookmark folderIf you dont do this you may see Folder not found errors or youll see this repeating folder display bug +was already +Fixed r +Forcing the appClass to be specified sounds like the safest patch in branch commit aaaaebecebb +This patch implements the proposal made by Arun on which many comments led me to believe we had consensus on +Hi Stevensimply amazing! Since you are an ASF committer would you be pleased to be granted to commit to Sandbox? Commons Sandbox is open to all ASF committers I would be more than glad if you could participateMany thanks in advance +dogs is indexed as dogs +I would even say most annotations should not define defaults +Eugene this will likely not be put into the line +Spec was lacking on what to do in this circumstance +Documentation fix for this issue committed under as bfbddbddabbbbfccdbdaa. +This was caused by the message route not being persisted to the back end +part of +This patch adds a crude heuristic to the default range preallocator +closing as fixed please reopen if you still experience the issue in version. +Committed to trunk and +It is good to see that on your platform the defaults result in slightly better throughput for sequences and identities than the previous release +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +patch v after review on issue resolved by +Actually this may be a bug +May be we can postpone it to +Well I guess thats happening on the server side +didnt read the mailing list before I post this issueIn fact Trustin have made the necessary changes so that it works like it shouldI just check and it is okThank you againSo I close this JIRAFrederic +Verified with JBT. +I wonder if its a cygwin binary and your is different enough to not display the bad behavior we are seeing +Thanks Manikumar Reddy. +Todd in your proposal you unlock the row before the Hlog is synced +I made this quick test to see what happens with asserts in spawned threads and the result is that they do not cause the test to fail +Missed some files in the last patch checked that this one actually builds correctly on a clean checkout +Hi SteveI tried Clearspace functionality save and continue and generally it works without any what you told me via IRC you worked hours and then you hit that button +Not an issue +There is a version and an alpha v to v converter tool +Attachment has been added with description SVG Browser has been added with description script file that is uesd for running the svg setup of home has been added with description creation of Home has been added with description preferences file +Since that class directly controls when it invokes I think its safe to just pass the top contextYep +Would that be fineRegards JuergenI have applied the patch on and my use case where the exception was being thrown works fine now +Sailaja can you please point out which command got timed out? Async Job Id would be useful for debugging +Problem caused by Metamer setting boolean attributes with EL +Patch file committed at revision to branch as revision Updated fix version to reflect this +is a patch that manage rewritting of location +In other words classic portal has been removed before initiate a new portal +Itll better if the node resource config is grouped into a repeated field so that you can config multiple nodes in one RPC which is the common case for cluster management systems +Such blocks seem to be removed when the datanodes send block reportsUpdated patch includes the check +Ive attached the modified file +Given is so much trouble to Hadoop maybe that particular setting should be everywhere users to set rather than is an incompabtible change +Attaching preliminary patch in case review board misbehaves +Thanks Tony +Thanks for reporting the issue +We use the helper for launching the scripts I see you are not using it please plan to use for launching shell scripts instead of +Finally found the time to do the refactoring +I applied you patchThanks +I dont quite understand the intention of this ticket +That fixes my concern with the label text +Even though an is set it wont ever actually be triggered unless there is a DTD definition in the document or if there is a schemaLocation in the BPEL document +Itll be more useful if it can print the versions of the native libs as well +The ve tried this patch +The crux of the issue is that was never able to specify stuff like this before so the entire design is still basically what we had when Solr was single coreAs far as I remember shard handler started in and was per core +fixed in for in revision . +For referential polymorphism its not working because of xlinkhref bug thats already logged this issue +Patch for rebase version to +Eclipse test project used in eclipse and to test the fix +Added a configuration for +BTW thnx for the tip with JIRA now I can do more here +Thanks to Dag for fixing this bug +Committed revision . +Not sure when we can fix this +Thanks Junping +Thanks for the example. +I tried this on both the stock eclipse and and it works with at least there is a path forward for using IDE on you have any idea what is going on w Eclipse to cause this solution to work? I can see it works but it doesnt make any sense to me +That should be saved for each target component +Reporters of these issues please verify the fixes and help close these issues +This Sub projects POM inherits from Base which in turn inherits from +Verified OK in Hudson was a CP problem with using the wrong Ant task. +Could you fix the conflicts and Ill apply this one next +Yes if we would do it this way its maybe not a big deal +But this is not something Apache Bigtop can tackle itself +After that is resolved I hope to start the merge process +Bulk close after is released +I could find no other Ruby methods defined in or that could ever conflict with a typical reified method signature +We need another lower priority jira to track itLahiru +But we should just add a simple comment stating that the score from the Query may or may not be meaningful and the Strategy used should be checked for further details +Please add brief javadoc to the methods as wellWith this addresses for the patch with from Jenkins +tnx +Closing. +I agree with you that we need to define the behavior for overriding merging and default values +The comment suggests this API call suggests this is new for +However it seems to have happened a bit more regularly since upgrading to +Can we get a regression test added to the DRLVM test suites +IMHO Commit this one do a second one plan to refactor out common code +Has this been fixed +The log entries are not related to the report you probably have jobs in the queue that are slowing down the system +Pretty sure this is a duplicate of. +The only issue would be if tokenization and value storage were happening at the same time they could not operate against the same stream +Committed to trunk. +Not sure what the test issue is yetAlso looking at the code again I notice the prefetch is already defaulted to +changes are unnecessary +You want to fix +Ive updated the patch file +Hi JonLooks very promising +I will be on vacation so wont be committing the changes yet +The revision is I will add the Axis changes on ThursdayThanks the Axis change + was added upstream in +All passes with versions specified in +In other words ipv would work on windows but not ipv +Fixed some of the issue mentioned in the comment above +Then from me +Which is the missing component in JBDS +Attaching anlyased Konstantin What is your suggestion on this issueDo you have any alternative idea in your mindThanks I think that the RBW replicas on DN and DN should eventually be transformed into corrupt replicas because they should have a wrong generation stampThen you are right it is similar to +Attaching a patch with suggested change +Does the separate zk dir get cleaned up when test is done +Thanks for the review Joey +Make sure that you dont have devMode or set to true +Seems like this has been fixed +So we now have at the Unix level drives full of blocks that are world readableA task gets run on those machines +Looks good to me +This patch executes the token file writing to DFS as the logged in job tracker user +Thanks Edward +Unit tests updated to check for this in all repos too for coverage +This is a patch agains svn trunk version +was fixed with +It should have a the same as the standard datasource types as its the that determines the value +svc files are owned by ducc and are therefore visible by all +test file for the swf parser +Then it only show the first clicking on Documentation and then back again will cause it to show all options in the Details panel +verified applies to trunk +always prompt about possible data loss even when force is specified update so that it talks about to avoid confusion the above jenkins mutterings are the result of it trying to apply the design document as a patch +There will be no reqiured projects declarations in java build path +The second one looks natural to me too +I made these parameters optional and used setter injection instead +So if were trying to do a waitpid after the JVM has already done a waitpid were probably getting ECHILD as a result +frameworks will be used instead +The window is closing on this one the current code is do we want to convert to and move to core? Move to messages contrib? Else it will remain the same +I am using so should i not be okThanks +now handles the expression expansion that was in but there might be places in the project where needs to be used +Additionally it seems to me that the second change from the description above is no longer needed since implementations just return NOMOREDOCS to indicate that the end of postings list is reached +Great patch looks much cleaner +next thing is to get the other stuff like eclipseidea and ant plugins to build using Axis +So i think Change Font button near the area will be the best solution +Attaching a patch to enable the for the existing sql tests in theAll the tests pass except an explain statement that uses a custom property index +However we need this submitted as a patch under Apache license we cant accept contributions directly from external +The patch was not applied in fearing this would break the codeThis will go in allAfter reporting this bug there are so many changes have done for this classI have attached the diff file current code and version has used to report the bug +New component for default XMPP s the License for Smack XMPP Client API +closing issues for released version +Hi Martincan you check this issue after recent upgrade to JIRA please? I have no Mac available to check it +Looks like its failing in the fix logic but Im surprised can fail this way +Sorry Robert as always it was funny to fight to find finally a good solution! +A directory would be copied recursively +The patch breaks code completion +Copy and pasting is bad because it makes code hard to maintain etc +Fixed in commit issue hasnt been observed again +but existing ones do still pass +Thank you for the prompt updates. +This is one example of where new stuff wasnt revisited but will have to be +Im pretty good at generic programming and tend to write more frameworks and libraries at my jobs with applications layered on top +I cant see any +Will reviewcommit the trunk patch soon +Compared to Ruby were about twice as fast and compared to several times faster +The changes look good +Merged to and committed revision . +It also makes it easier to track what changed but then again people shouldnt keep the output of srcgen in source there was some quick way for srcgen to figure out whether the schema is newer than the generated source and thus prevent srcgen execution when the schema hasnt changed since the source was generated that might go most of the way +Patched that reuses Clirrs Plain property to Clirrs plain report +But yeah as I suspected a CF name containing a hyphen causes an sstable write to blow up +I just bumped into this today attempting to set up a JDBC datasource in final +There are many places in the testsuite that specifically bind to localhost +Ciao Fabio +Servlet Configuration Servlet Mapping Load On Startup +Apparently this breaks two CTF tests +Also this is logged against CXF which we no longer support +Patch committed to both trunk and branch. +If someone else can review this for correctnessstyleetc Im happy to fix any issues on and trunk +Okay Cassie making this a bug so we can fix thisLets make sure that UNDISCLOSED becomes part of the spec for. +A similar synchronization was added to invoker generation a while back +Please provide a patch for patch for trunk +Do you have an explicit example of something youd log that would help debugging slow queries? Good stuff +Closing this as a dupe +and are still available in SVN and has references in hibernate ve removed references from feature and updated build related map files +If you have asuggestion for logging statements to add Id be happy to put them intoour system and see if we can collect some more details +Still trying this +Jorg it looks like the version of Derby you are using is too low +Push to a later fix version +Fixed on branch vGACP revision . +That would be the case if it wouldnt break backwards compatibility +Unobtrusive is not the key to the thats the main problem with setting windowUnloadedtrue on beforeunload event +Added live tombstoned cells per read to JMX added percentage of tombstoned cells per read +If no version is known then I think using the Java version makes sense since we will be treating it as an implementation package tied to the Java version +Wont fix because we want to port the from the branch +Looks like no one cares about test failures on Windows +I also get the error when trying to copy nodes that were created with jackrabbit +I understand that the attribute render works now but that this bug here is not fixed by that rightSo this bug effectively impedes to use renderfalse +If someone didnt have time to tell you why your improvement isnt going to make it into Hibernate thats OK from our point of view +Each memory usage was collected after the server started up with the app deployed +We need to be able to peacemeal upgrade our clusters +Fixed in r +Fixing Affects and Fix in Version +I make the changes for the typo +Any news on thisThanks! +Code and documentation changes delivered +Got it working with Java DSL +So as long as the multi itself does not fail were good +you got a test too +Perhaps as a patch forMaking this as theres no patch currently +Calls to keybdevent were replaces with calls to obtained its another version of patch in which native code doesnt look for Win API functions through user library but relies on natural dynamic linking +Not much point doing svn stuff until user accounts exist +This patch fixes the problem so that the generated XML is valid however perhaps the should be modified to support the same logic as the does when constructing qualified names +Thanks Bill! +So if I make that JAR using Gradle even if this uses the latest version of the Groovy compiler I will still get a JAR with a corrupted class file in it wont ISecondly apart from the fact that I get that error in Eclipse are there any consequences at runtime? I fear that when the code of that corrupted class is executed some runtime exception will be thrown wont it? +Thanks Jean. +Incorporated Dougs review comments +Any idea what the timeline for Tiles is? Its clearly not going to align with Spring at this point especially given the backwards compatibility issue +from last GA a copy from seam afaik +The tests are failing at this point but this can give everyone a jump start on looking at the code +get is not static so it can pass on the parameters the constructor was called with to any it returns giving us option that just requires documentation changes +Committed to trunk. +Try to use the instead check out the latest svn +Fundamentally I think this boils down to simply having a special trash state for a table before it goes through the real delete +Auth provider may be candidate for admin api through interface itself +I know I talked about many things in this jira but one of the outstanding issues is something I think REST API client apps will find important having the response from a GET request include the in a response header +Patch applied to svn rev +Ill commit this on but if a python guru could also review the change as well that would be great +Somewhere along the way we started do attempting to prepend +TO +As for VAS Im not really sure I careIn the case of what is the point of mutexes around a function? Why not just call the unsafe one and avoid the extra performance hit on machines that follow POSIXIn the case of Ill be OK with this if we also add a huge warning in the task log telling the user that their system isnt POSIX compliant and to call their vendor to get it fixedp +I am happy now Ive also tested this +Comments are welcome or even we open separate jira for thisAs for the configuration file itself we could extend installer ant script to copy custom config files into right location +The diff is a tad hard to read so here are some notes +This problem is found with Xalan release +Setter methods return void not String. +I wrote another client from scratch and its been written from the ground up to work well in a environment +So youre basically suggesting that we allow the to propagate out +Looks nice +No automated tests for this +is soaking up the sun in Hawaii this week but should be able to provide his approval next week +sorry about that then! The only thing Ive changed is my primary email address as well as my forward +Sorting by Long type only applies to columns so the bug is notIm going to change it to documentation bug +It seems to be trying to make the directory even when the trace directory is not specified and we need to put the trace file in the default directory +Id use this issue for resolving how blob creation should be handled but add a sub tasks for the Microkernel related issue +Are there any other characters that need to be fixed? Im worried that this is a point solution for one character rather than the right solution +Manually tested! +Could this possible cause a situation where page controls would be null in even if already in? Im experiencing weird errors about control suddenly being null when trying to access it under heavy load server load +Jervis you might be the best person to check this +I completely agree with the idea of providing extensive language support +I dont see that it adds significant complexity to make the data and checksums appear as separate input streams +It may also be worthwhile to factor out the code that encodes the special characters and put it in a helper method instead of repeating the codeI think we also need to escape amp in these strings +It also would let us not worry about testing and supporting all these other containers +library is also in JBOSSHOMEserverstandardlib and JBOSSHOMEclient directory +Should I mark this as resolved then +bq +Thanks for that +I believe the failures you are seeing are due to the change +Attached is my dsa public Mojo Codehaus Access +v takes into account the comments from the review board +difficult to assert the stream closes alone +In my test time went from min with sec to min with sec time out +If you want to create a blank wiki page and link me Ill write up the documentation for it +Unit test that demonstrated the problem and a proposed solution +Only two bugs are actually fixed between RC and +Hitesh corrected me on jar names Updating the patch accordingly +Note FUSE MQ is only used for unit testing it works with any version we have at FUSEFUSE SF is used by the component so that is a real please go ahead and start release +amended patch with AL header in the new correction in patch to cover the case there is no op parameter +Ya that is defs the easier solution but it does require users to do extra work which I would eventually like to avoid +Yes nothing changed starting nodes generates also mbit multicast +The only downside to that is that when there is a type mismatch it will be the only message displayed to the user and the other validation errors will have to wait till the subsequent POST +I opened it +Sharath can this be resolved as Cannot Reproduce +This seems less important because the majority of regions will be moved off the RS and the few new regions can rely on the automatically fail over to other RSs +I cant seem to repro the original problem +Committed revision . +There is a fair chance this issue is fixed since I vaguely recall fixing something about file attributes in plexus +So closing this JIRA but opened +We are still waiting for the drools team to release the GA release I believe they are behind schedule but will find out +UDF framework has changed +the think it should be consistently ordered by what is displayed in the dropdown +Yeah this is my concern with all these schemes +Patch that corrects the functionality +This is a duplicate of. +I dont know if running on a cluster is the issue +Deferred until +Committed to trunk and +You always have to first configure your Fuse installation to have a user set up +I normally work on Linux dont even have an up to date dev env on error occurs on a bit system +Presently we do not publish snapshot artifacts to a public repository +It assumes that when it returns true theres potentially another running task that needs to be considered when deciding whether we can assign a new task to this tracker +Provide a interface and a possible implementation such as +Sorry about that and thank you for responding. +Added bundle state replication +Need to release the rest is done in +custom plugins support is implemented +Thanks again! +I like itId be inclined to prefer reserving a couple bytes from our existing bit clock noting that we have several centuries of headroom in in bits +I made a good attempt to but I cant accurately depict the problem without including the YUI stuff because it creates DOM elements on the fly +I just committed this +Will look into it +When I tried to add a party relationship I got a foreign key violation error +Cancel patch since I need to add one more utility class for saw many commented +Updated in revision by Tom +If I move the set to after the start of the xa transaction the isolation level gets set properlyIt would appear that with embedded if you set the isolation level with SET CURRENT ISOLATION URit gets reset on start of the xa transaction to the defaultand that causes this problem with set with client +Attaching patch that just do that +Changed coding style of statements +Waiting for jenkins +what is that for? We need patches not builds +We need this one +uses long for this limit theoretically users could use a number larger than an int today +Switching to java regex is fine by me +yes it was a fresh is due to missing files in packaging +In any case I need to backport the improvements on trunk before +The test still randomly failsIm digging +My original patch did not use this flag it just collects the list of relevent providers for the dependable and passed this list on to execution where it does the appropriate checking with respect to the current user +Noise it may be but in the particular scenario where I noticed them testing ATS with extremely low traffic to verify initial functionality its confusing before identified the bug I spent a good few hours trying to track down where all the cache misses I was seeing were coming from +We tried to fix this in but had to give up due to missing definitions in EE itself +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I have a GSP that I was using for testing and it was lines from a page that I had copied and pasted +We have application dependency and have to limit ourself to Axis since we have a big application in which axis is integrated +No resources allocated for this. +All the unit tests passed save which is not relatedI committed this +Apologies for themess I createdThank you guys struts is such a great thingPS +Thanks! +Rebase +Thanks Sumit Pandit Rishi Solanki Parimal Gain and Chirag Manocha to review the patchArpit Singh Arpit and othersPatch is in trunk rev +I have encountered the same issue +Why not as an extra utility in contrib? It manages logj by jmx +I can reproduce this on using the default tooYou can see this also from the fact that if you quit the installer the title of the messagebox whether really to quit is displayed with the chinese characters from the translation +lets see if we get other feedback before i commit +I have updated the patch as per the above comment +I have no problem adding the sources and attendant artifacts to the release build rather I just want to know what it is you are trying to +Added a maven project which demonstrates the problem +Then I suggest making description required + +Committed. +Found another memory leak see +Currently the hibernate config with monitoring is set to update the schema when the mappings change but i am curious to know how this works +More patches are attached +Note that there was talk about syncing the mirrors in some more efficient fashion rather than letting them try to catch up on k revs +A schema part can be added to the element as a workaround +This was closed long ago but didnt seem to get merged into any release branch certainly hasnt been applied to or +The Solr farm is setup behind our mail servers and it is never exposed outside +The file is the working one +not sure what option means but there is not going to be separate branches for eclipse plugins that messes up updatesites and usabilty +This is still there in the runConfigure +Maybe if he shape it well it can be contributed back to Apache +Can this issue be closed now +Most of the time merge from trunk should work +from me we go +Committed as svn revision to the trunk +closing as clean up +But again if you are working on a fileand it is inconsistent and it makes it easier for you to work on the file i say go aheadI have to say eclipse is pretty infuriating in this regard +We can fix it in the current upgrade script or later +Thanks for the explanation Vinod +Committed in trunk for rev +Solr is a separate s erver like mysql and should run in an isolated process +Pig and Hive folks want to radically change what MR is +Attaching and with the correction of to +bq +committed. +One way is to find an alternative way for runtime detection results to show up +During a compaction or flush wed lock the object for the duration of the compactionflush +The code was modified in Apache to make classes less public +Sadly even with rev my confidential file doesnt render correctly so there is a different cause than the regression +People will see the feature has been added but might not realize the downsides. +The queue size will be +It does all of the communication with the and to access the blocks on the nodes that make up the file you are interacting with +The DSL should be kept stable +A snapshot of the table UI is attached +Do we need this +Field names are only ever present in the serialized DDL +Im not sure Ill have the time to fix it but I do hope I can port the WFS datastore to the new design made for +I see outstanding related issues has been fixed +Thanks Mark +Thanks for implementing that idea so quickly +Entire log file attached +Well want to revisit this at some point in the future +This seems to have helped so Im resolving this as Fixed +Created add function to handle full screen widget view instead of altering addOverlay +Otherwise just default to the first one in the passed list +Its nice that theyre documented but being that its a user guide I think it may confuse people +probably although I guess I need to document what syntax defs do what so the user can modify it +Typo the related enhancement request is not as specified in myoriginal post +This content type is then used by the editor and document setup participant extensions in the ui plugin +Moving it and the jnilib somewhere else fixes the problem. +Please check this has been applied as expected and resolves this issue +Fixed the Issue Description which had reversed discussion about the +Uploading the patch again +Yes I can connect to the FTP Server with +This has been corrected and Woden is now passing tests and +Thanks Christian. +But if there are no concerns regarding binary compatibility we can also fix this right now +On hold until the new authorization concepts are coded +The new releases of EJB Tools depend on the version of EJB that your AS has. +I have updated the patch for this +verified on docs +or the bug was fixed in the branch +Nested get and set getting and setting attributes of private variables and such +Kedar Im having the exact same exception +The fixes for should be backported and targeted the next point have disabled this particular test on the branch +Attaching which contains patches and test classes for the missing tests from the earlier attachment +Patch to remove and the associated example dir +We might change the structure of the eclipse modules and the use of resolvers +A patch for Hibernate final +Vitali could you attach your JSP file +Checked in the fix to trunk in revision +Committed revision in HEADThank you to everyone who commented on this issue +Eclipse starts up JETDPGO thresholds only a +Otherwise use SRE and make an extra request later if it looks like one of the minimal set isnt going to respond in timeNote that performing extra data requests does not affect just makes it possible for the request to proceed if it gets enough responses back no matter which replicas they come from +When I tried to initialize a object from class the program just cant stop + + +I think we should be consistent about this completely and not override any methods with default implementations in these Filter classesSo that means stuff like too +This patch also contains a fix for the bug where end transitions didnt have null next tests are included to test both of these fixes +Im hitting this problem as well +This problem is no longer available in the trunk +They are the same failures that were recently reported for and others +Though perhaps I am mistaken I dont believe that either Sun JDK or Oracle JDK are or ever were official product names but rather wereare common shorthand ways of referring to such products +Just to be explicit the other failures are this issue as the problem is not in the product. +There is the ability to have labels with models or attribute modifieres and both are fully java based + JuergenDo you think these classes will make it into the sandbox before? While it will be great to get them up into the main source tree checking them into sandbox is good enough forNow that is released any chance we can get these classes into the sandbox? They have been performing just like the previous Hibernate code +Yes! Please close themId love it if youd like to do them +is changing a target called +Updating the widget caused the error in a check if the checking is for archiving and if it is then skip the update because it is not needed +Not everything has a find page especially when creating associated data and low volume data + +The problem you are running into is in fact equalshashCode problems during loading the Map +Thats optionsreadonly not dfscontextreadonly +I just committed this to trunk +Hi JacquesI think Sergey was able to build with Oracle JDK by applying a series of patches that have already been applied to trunk +This screenshot demonstrates the railo admin scrolls horizontally for the new coding tip feature +Also the warnings are from thrift binary files and jars +Thanks for the confirmation. +we should deprecate the methods in a first step. +My agent had been shutdown for a few hours now +The attached patch corrects an issue in the connector that caused it calculate an SNS index of for all nodes whose SNS index was not supposed to be +zip file with all Docs was uploaded to ftp +The Tibco EMS Connection factory has the same problem and I think the same adapter would helpBut what is this issue really about? Is only documentation necessary or do we have to change some camel code? If you give me a hint where to look I could provide a patch +Fifth patch for the new style to the Return forms +Me too +Documentation added +Thanks Siddharth for committing this +done let me know if this works for part of patch still necessary with your changesIndex srcorgjgroupsRCS file cvsrootjavagroupssrcorgjgroupsvretrieving revision u srcorgjgroups Jul srcorgjgroups Aug return receiveblocks? case SUSPECT return case AUTORECONNECT return autoreconnect? case AUTOGETSTATE return autogetstate? case GETSTATEEVENTS return case LOCAL +Quotas are per specific entity user directory and so on +This sounds like a good change +I think it makes more sense to clean up the dependencies now then it wont have to be repeated times when the project is split +to both Alan and Olga +And I am trying to use Eclipse I help +When using ROP I could only see postLoad preUpdate postUpdate never the prePersist postPersist +ready to go with that for + +Try that and close the ticket if it works +Excellent +The same patch but without cascaded work units which can be replaced with next revision lookup +I just committed this +Another patch to make it easier to see which jars have the invalid versions +In most casesit should just work +Applied to svn rev + +Postpone until AR clearing AR for release +Badly sorry for the delay +This spacing between the lines is commonly called and is calculated from the property which fop sets by default to +I missed to read the bold not and assumed it worked by default for SOAP wish it didChanging it to bug +Did you activate PMD or SQUID +When for example hdataTable is put into facet styles wont be picked default ffacet renders in div container +It was just a little bit of context change in +Hi StephanCARGO has three possible deployer actions deploy undeploy redeployThe third is very probably what youre looking for it will try to undeploy the application and then deploy itCan you please tryThank StephanNormally undeploy is expected to fail if the application is not deployed yet as a result the behavior in Tomcat is not the correct one and the one on is correctIf you wish we could an option so that the undeploy action would ignore failures +Using a annotation is no option for us +by Richard. +to Shazs statementAlso I think we should first document exactly what it is we are going to implement +Its part of +And remove from list removal please resolve as other changes looks good +We feel that the Session object is being created in the where as the Driver is created in the thread +Examples def abackslashb ab def aslashb ab def adollarb ab +I believe that they are platform dependent hence I do not assume that they always have same compression ratio +Simple patch fixing the issue is attached could you please check if it is ok for I am afraid the patch needs fixing +it is working + has reportedly been fixed +committed the patch with the test for this issue svn +Thats what I thought as well however win a couple week of pulling it in there were enough testsprograms failing and people complaining that I reverted it +Please verify. +Please try latest CVS +batch transition to closed remaining resolved bugs +thats a very good idea! I think we still need to testadopt this thingy to make it rock solid but hope it already does the fundamental things +adding second part of the patch from Christian Vazzolla in revision r +Designer should update the model default to and increment the index version as well +Otherwise if the session does not exist already the attributes may never get written +Patch to skip site generation for YARN and Karaf +yes I in rev +Launcher should be using JDK and not JRE +Assign to myself for backport +Thanks! Patch with test case added committed in r. +Could you provide more information or some links to that issue +Odd +Patch applied and I also added a unit test to verify fix +Do you mean push the listener stuff in a different JIRA +Ive observed this same behavior with groovy beta although I cant reproduce above +I applied the patch as it it seemed reasonable to me +The wiki connector description pages have been updated +I can even help to debug with your instructions +If you disagree please explainIm fine if you eventually want to try to change web ui addresses to match web services but in some cases it might not make sense as there isnt necessarily a one to one mapping and one might expect something slightly different from a UI then from jsonxml rest api output +Actually to clarify there is no bug in the code ignore last comment +The GC fixes from master were merged in AFAIK so if they caused a regression we need to know +Hi VladimirI finally got the benchmark working again your fix is correct thanks a lotIts now a bit costly in terms of object allocations we can think of other improvements for the future for now Im glad it behaves correctlyMerging! +Note that the fix missed Final +Im reopening this since Torsten has discovered problems with this module +Thanks for the pointer +Committed to branch with the nit addressed +Anything were waiting on to commit this issue +Same problem +Freddy could you please consider closing for this ticket? +search thegenerated SQL String for occurrences of JOIN etc +Attached is a patch which implements this +fixed in encoding is now properly detected from its content +The class is also imported into a number of other classes but its not actually used there +Implemented viewport switching to show all columns in landscape mode and hide some of them in portrait issue hasnt been solved +Besides next feature release would be a good moment for tweaking the behavior of compatibility cookie spec given the scope of changed planned for +Thats pretty much the way standard Spring beans work +Here is the patch from +Your latest run came up with a slightly different kind of error a checksum error rather than the consistent array outof bounds from previous runs +Actually no you have a look +But can we pinpoint the commit that causes the issue? Regarding Selenium tests can you please provide some more instructions on how to set it up? Maybe we can even include them in the source code and run as part of Hudson build +Groovy does not have to support all rare cases Java supports +Annotations added assertXXX are changed to +That would be great Tim +XSD for ESB configuration in have you started to model ESB configuration file I ask this because I have created EMF model from just for reading ESB configuration file in Testing tools if you have worked out the modelI think the testing tools can use the model of yours and I dont need to define another model for ESB configuration file +Any chance of a fix for this in a release for +They are not currently included to avoid classloading issues + +Neil thanks for the String vs thing you pointed out Ill have to change my code +This appears to work fine now probably due to cleanup work in tracing and better IO handling for Kernelsystem. +I was hoping it would get us halfway there but if it does not we should do both together IMO +When you say do you mean the last release of the major version +Care to share an error message or even a way to fix it? I believe Prototype and Scriptaculous are supposed to be compatible with Opera +I would expect it. +Patch looks goodMinor comment should below be renamed below for consistency? However I am ok if you do not want to make this change +related to this issue +Adding that would allow a lot of mess in the domain model and I wont do that +same issue probably not going to be fixed in cli use agree w Andrew +Added more descriptive error messages that tell users that they should not be declaring functionsrulesdeclared types here but should use the corresponding guided editor features for it +I suspect that the permissions are ok as it can move the top level tiles to the directory +Failure was due to a bogus URI being used by that should have failed originally but in locallog dirs arent handled properly + +We have Windows that are being used by CI Services such as Hudson and Buildbot so you can of course use one of those services and specify that the Windows VM does a build for you +This link points to the actual +Youre welcome +Will take it up with the Facelets guys again +This is a clean compact bug fix +Once they hit the should have been translated to the latest TS found +Ill write up a confluence page on how to use it more and some things to consider and post the url here was able to get the instantiated by adding the plugin harness test dependency and update the maven pom reference to that is going to be the biggest limiting factor to this approach reconsiling the version of dependencies +Fixed +bulk close for the release +plugin jarfeature jarproduct zipand I dont think you can do this multi mapping with one the product lifecycle does not to generate a build xml file so this is taken out if its is more duplication but it only affects the us the end users dont see any of this they choose the packaging type to match the pde project they are working and everything is bound correctly +Need to merge into works for me too on trunk +Can you please specify if attached pom was at the root of some project or if it is located somewhere else +The fix is simple changing to use simple name as the identifier making it the same as trunk code +Alexs Confluence username in Xircles was wrong updated and he is now in +Actually we would like to get in also and this patch +This might need fixing for too +Claus does this work now +Ill report it tomorrow need to create good test case +UpdateWrapped up results page to display the summary correctly +Cant reproduce +Cheers agree +Thanks Doug for reviewing. +SInvalid reply from network server Insufficient data and then test cannot start the serverSo probably if you wanted to reproduce the circumstances you would need to run a small suite with both andUnfortunately can really affect any network server test I think it is not usually the test that hangs but rather the next one that tries to bring up and ping the server +Ive applied this patch to the trunk out of curiosity +The window is resizable +Sravya would you be willing to propose a patch +sounds nice +mrepositoryjavaxactivationactivationCDocumeUSERNAME +bq +Thanks for all the hard work Nirmal! +My first approach is to modify the exception itself to print more information in the message and add the first exception as the cause +I havent been able to reproduce the issue with a simple example and my application is too big to attach +Blocks Ill take this one +This is probably due to the platform we are on +In other words you should implement the interface yourself using the building blocks provided by Spring Security rather than extending. +Here are my findingsWith IBMs SDK only one file namely modulessecuritysrctestorgapachegeronimosecuritynetworkprotocol fails to compile with error import can not be resolved +And since in your case doesnt describe any beans there is nothing to bind to the you try describing EJB bean in EJB you should get an error during I am going to close it as wont fix +works with current svn head +Patch AMQ so it works for them +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Personally I am not sure of the difference so can see that users would benefit from this +works fine for me with the attached project with functional testcase + +This is because setting the debug property causes havoc in the communication between the client and the server +the schema is generated correctly +Anoop thanks for pinging on this issue. +If there are no comments I will commit this +The problem is when the database is in utf with a non UTF database until their bug is fixed or apply changes to the as detailed in previous tested and committed +I checked it carefully and the problem is all facelets components requires to be excluded for the registration done with +The previous comment relates to the linked issue +Ive added Action for seam wizard and updated shall to be added for wizards and found these icons for the wizards and perspective switch to be added to Welcome they are to be reviewed +Ted So you are +Ive added the table for the Data format options +Setting to Resolved +Great to hear! Thanks +I have updated the for details on doc updates +Thanks Uima AS starts up it prints Setting default totcplocalhosteven though the brokerURL attribute references a different broker +sorry +In this case the memory manager started a minor compaction before the tablet was online +Fixed by Ales +No changeNote also that the link in the above comment is truncated while the link in the corresponding email is OK strange behaviour of JIRA +fix and tests committed in trunk and branch +No just mvn verify etc + revision +Thank you! Its working like a charm. +It shouldnt be noticeable of any performance impact +thanks +Thank Brock for the review. +Sorry +trunk +Local write only if not hinted msg +Otherwise Ill just continue building with the method I described above +It works for me! +second patch is committedbackportedjust remains to add the random test to all remaining tokenstreams. +I just marked as DONE assuming it does not need to be ported back to branch +Added the link to issue to mark the relation between the two. +These tests are running mr pig hive oozie etc and in the back ground jobtracker would be restarted +If all gets are in the first region it will be faster than the sort I will at Sergeys solution and try to implement it shortly +The user would need to make an invalid filter before we would violate the specification +One indication that nothing changed is that the patch size is the same +The question is the wordset is initialized to be empty if missing +now maintains a member deadNodes +Good on you AA +Sorry I still dont quite get it +everything is ok now convention used then calculated default values match + +Intentionally didnt set a fix version because doesnt make sense to have this show up in the release notes for . +On my standpoint its more readable too +Create page called testCreate Exception called that points to view Now try and move the link from to the test page. +is +fixed rThanks! +To my knowledge sharepoint only provides correct cmis support for lists which are document libraries +Time to prototype a read ahead client +Updated unit test +New unit test for the partial stream reader +Please review the patch tests usage doc will me a thumb up i will apply up +Cool! Thanks everyone +This patch changes Network Server to pass the holdability to the prepare method of when running under JDK +The new patch looks great to me +Its easier to handle the ORM complexity explicitly in the Page or Service code or even in an application specific manner. +danielsh checked with the community and folks are OK with the change +I still like fs or something like that +Unfortunately I cant +I think we should just mark this closed +Also note that we should support configuration and it seems the only way to do this with maven is to actually use profiles. +where should I document this Max +the solution is really remove targetos windows and ant would assume the current OS as default +slfj dependency has compile scope in Wicket pom how about remove it from the quickstart pom file +Works like a charm +Wrong versions are referenced in the generated +Not reproduced in to change resolution not Done but Not reproduceable. +If fixes the problem +See what I mean +We will need to create COL COL COL when COL appears multiple times in a shapefile header +Such a deadlock can occur distributed between DB and the Jackrabbit locking +Problem compiliation from that jdk dont autocast Double to double +Either will cause the transaction cycle to be loggedInternally TDB used thread local variables for a transaction +Modifying the proposal to support only whitelisting would require many additional changes and do nothing to simplify the current changesAs I said if everyone else participating agrees on these additional changes I am happy to implement them +Hi can someone add it to the trunk of cassandra? its really a minor but needed patch +This was addressed as part of the AST Transformations +Hey Jens the problem was that Windows paths werent being interpreted +Patch on svn for pig exim loader for cli changes for outstanding patches attached +Isnt this the use case +PS review request i mean at +Although spock here is junit based even execution of tests by had the same issue so I corrected that as well here itself +Thanks Vikram! +I dunno if we go for would that mean? Get then eventually get new body then the server send the deleted revision And if compacted and no revision a +Lets keep and close +Lucene is a library for build applications that include search functionality it is not itself an applicationIf you want to down load an application that you can easily configure to automaticaly index a website you should consider a product built using Lucene +If this doesnt change well need take into account our functional tests which should be modified to to conform to both app servers +This is not a forum +For the async case we need to have the connector track memory usage and not exceed some max window size for the producer and wait for the broker to send producerAck messages to let us now that messages have been consumed we should block when we exceed our max send window size and unblock once enough producer ack messages are received +Sean can you fix these links and then replace the doc bundle for? In order to get the correct release information you will need to merge changes to tag +now reside in org +BradHave you looked at the sandbox focus component? It seems like your work here covers some of the same ground +Attached is a with the changes included in the correct d be better if this was a documentation for the new properties and how theyre used updated test case to check the new properties are being processed correctlywould really to as a +patch will be ready in future +Also since the detectors are now automatically loaded and used theres no longer any need for the explicit option and Ive turned it into a +Patch applied thanks! +No additional on that topic as it is unit tested and already has covering the backup restore feature. +BTW the various errors feel very much like concurrency issues to me +For or other clients you need to perform an explicit ADD +yeah it is not part of the default build for that reason +added the information in section Key attributes and ways of usage +But this doesnt change the behaviour Freeman already described to you +Attached file can not be retrieved +or on which agent exemple if I have agents on windows on OSX and on Linux Id like to choose to run on windows linux OSX platform without to choose a specific agent for each platform +I wonder if we could detect this reliably +Hi I just have tested the patch with and service everything works wellI had a quick review on the code seems good to me just some english mistakes in xsd documentationThe service name use to resolve parameters +Empty file was deleted at r +Review board entry updated too +I have submitted first version patch for this issue +Note that contains more information why the target must be injected into the proxy +Marking closed. +Google have now announced the dates for +Have fun +I am thinking that is not the appropriate place for them +I actually meant and not in my previous comment +Reverting this change until after the release as per Adams request via IM +a better solution is a better solution +Excluded resteasy integration from because you dont like doesnt mean you can break it. +Forgot to add in the description that only the sets this attachment but not the have merged the pull request so this should now be resolved. +The ugly message only happens when the click is on one of the arrows of the menu on the was triaged for SOA and since it is not in EAP at this point we will not fix it for SOA fixed in EAP even SOA takes the latest Admin Console release its fixed for SOA too. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Thanks! +I think there is no need to specify this specifically here about encryptionAlgorithm because now we are explicitly talking about unencrypted database +Moving out of +So on the client side it doesnt look to me like there is anything wrong +verified at subversion revision +For me this feature is available since see +until reproducible test app attached +This patch includes changes based on comments from Suresh +Try +Thanks Kazu and! +It shows it as being added in svn stat but not sure why it wasnt includedIt goes in srctestIts just a copy of site Im still getting the exception even after adding the fileDid you test with ant clean test or with a GUI +Id also suggest to change binhadoop or the underliying java code to accept options like additionalJars +So suggested code has potential null pointer exception +Definitely +Except for some whitespace issues Im here +At the bottom of the screen we can the javascript error +It is not clear why we even write a META entry for flushesYes +Patch against is released +Added example project +Username ari +Maybe we could try making something in the color +Well if we want to do the resolving process right we have to refactore it and change the API ll start discussion about the new resolver soon next week probably +The pings increment the session number as seen by runtimeinfo +I will use a String with as a separator +The features definition only list bundles to be installed but its the metadata inside the bundles that determines how things get wired and which packages are visible to which bundle. +on the latest addendum pending jenkins +A simple way to do this is to dedicate a single thread for all timed locks +Martin is this JIRA issue a showstopper? I can take a look at this based on the priority +Merged the patch into camel and branch +ok thanks Roman regards fine. +Because values are the same it isnt real race condition +huh sorry Ive misread your comment +Set Fix version to +scary +Is this sometihng we need to fix? Probably too late changes out out now for M so moving this to have attached a patch with a fix for the TUSCANY issue +Of course this is of no use whatsoever for community helpers who have no access to the vpn +You guys have saved me a lot of timeThanks for fixing this +However Ports and Server Details tooltips are always empty +submitted a patch taken based on source code from patch contained other changes not for this should not do anything special in commitJobabortJob api as it is doing currently as this special handling makes assumption about underlying storage +Attaching a patch that fixes a problem in to do with handling of null host in the URI +I was unable to apply them maybe because the diff was produced by git? When I try to apply the first patch via this commandpatch p i patchFile +as more people contribute then we may go for the more modular approch +remove this set +note jboss uses tomcat inside as its servlet engine so just need to know what it +Sketch of a patch +Thanks +Classes affected +Wont handle this +You may have to refresh your browser or again. +fixed for both backends in r +This is not a problem in trunk +So I guess simply creating one dummy entry should also be ok +Had a deeper look into this +Good point +We will see +Final patch including CHANGELOG entries +Packages moved +Li Any progress to report +One of the interfaces in my admittedly complex hierarchy of interfaces wasnt marked as public +? That way we wont conflict with potential properties from Hadoop and its a little easier to analyze stuff when looking at the jobconf +The current ESB project wizard doesnt require the user MUST configure ESB runtime using the ESB runtime definition preference page the wizard can use AS as ESB runtime if the AS contains a valid ESB runtime +Trying against Hudson. +But I do not want to +Please describe steps to reproduce it +this message appears on each page change and very is not a big issue we could create a template to skip object tags +Must have looked at the wrong place when staring at the repo +This one just copies the files from the new deploy target to the new one. +Thanks Wei +Hes OK with Chris taking over the patch +We can call set reflectively for direct support in our but it doesnt seem to be feasible to expose the buildSettings method speaking itd be great if Hibernate was more in its configuration style +The poor handling of a null calcresult dates from that timeSince the calc result is a strategy for handling whatever values are being summed averaged whatever this could be handled with a new class called which knows to return null +Ok so general consensus is to add and use an anonymous implementation of it instead of the classUnless someone disagrees in the next few days Ill commit a change to that effect +But i agree it will not help in case of serious hack investigation +Attached patch is the change that embedding API wants currently +I am still unable to reproduceCan you try the same svn revision as me sans any cloudera patches +A web console plugin would be better of course! +It is force so do we need to do the unassign first? I would not pass silently +The issue is actually reproducible really easy on betad build any application in Express Console view Click Create A Server Adapter When a New server wizard appears click NextRESULT Empty wizard page with Internal error Max Could it be part of this JIRA or shall we open a new one? What fix version +Also Ive noticed that the report links get generated only for subproject and not for the other subprojects in the plugin tests +Oops I forgot to style the links +for handling in this going to be fixed in a release? +Sould it be T extends and the abstract method to return the typeThanks is the groovy implementation loader now loos like +Moving out of +Messages are defined in map so parser have to retain the order +The log does not seem complete +Committed to trunk +Ive changed the assignee to unassigned +even beans could not be reliable +Feedback welcome +Doing this causes +c on an HC register different handlers as per below Changes to the subsystem management operation handlinga the describe handler throws an +We could have a different default for weme but then we may run into portability issues when moving a database created on another platform to weme so I think having the same default on all platforms is the safest choice +api +Based on Vladimirs last comment Im assuming this is out of date and am closing it. +same proxy and local you can specify location of your repository in and then use Maven proxy to unify remote access to the world for everyone +Hey Keith + on +Not to mention that lots of requests will be basically halted which causes user visible errors due to timeouts waiting for a new uuid +Thanks Eugen +Changes into manual and XSD should be done after merging dependens on issues +I am escalating this to critical +Per request this is a version of the test that uses ISPNs existing test suite +was the output like above or was there some indication the tests ran? did you clean first +Closing resolved issues for already released versions +I tried using the simple front end but there is something else missing from the docs +There wont be a release +Or do you mean that the widget creator should be able to set the default height in the gadget definition +So know the exchange is number +Patch attached +Hi Amila in addition to duration and dateTime we use stringanyURI I tested these and as far as I can tell the stubs compileThank you +Committed revision . +Im waiting to check it in until I know what JDK version were targeting for JBT JBDS in Eclipse +No specific import needed just be aware that you might have issues going back to JBDS with the same workspace +just commit a fix +The Java build I am using is +I think it would be helpful if you could add the following lines to the faq use the JDK XML SAX parser am wondering if it is possible to load JDK SAX parser and XML serializer if the castor failed to find the Xerces ones +Hello It would be better to have installer check for necessary features of the profile rather than name +The problem is likely because the is in a DOS format remove carriage returns by executing dosunix before trying to build +That one balances purely on the basis of the from the meta +I suspect that the idea behind this is to force people to exercise judgment in where they do this +Currently the API will not permit one to write a Schema so Im a bit reluctant to permit reading themDo others have thoughts on this +The needed test is in issue +Example schema and data to demonstrate bug +I agree that we should make this extensible to allow people to write their own rules for handling enumerations +Its not pretty +On the client side I have no clue how to solve the problem. +Attaching +Tom do you only see this with ESB content? Or with regular JBAS content too +Compare to section +New changes look good +The failure of unit test does not seem to be related to this change +This has been fixed on the Commons CLI trunk and will be part of the upcoming release +Im wondered that none has still committed the work +Please close this issue if it solves your problem +Works fine +This looks somehow impossible to me The only thing I could imagine is that the hashmap internal algorithm is somehow disturbed by accessing filling etc +In looking at the needs for me team I see they are more specific than what a generic than what the module should provide and that to make a more generic implementation would take more time than we currently have +invokeJNI helper in interpreter has incorrect handling for floating point parameters +I committed the patch that passed hadoopqa +The XML merger is tested in coreapimodule and also in extensionsmavensamples was not available yesterday +Thanks +Thanks Daniel +Reproduced with ie lastest IE headers working buddy how do you fancy a career as test mule? +Patch updated with the updated with the comments +Committed revision for IPF fix +That way we dont need to reference in +Review board is updated +Search box back inAfter that were talking about content only and my vote is to put it live with mostly the current content and revise it over the next week or so that were all happy with it before the official release date +So far we are in Beta and we can use the Singleton hack for a while so this is by no means a blocker +Im still concerned about the number of folders at the same level quickstarts guide template dist think it makes it hard to distinguish quickstarts from other things +Fixed util +The tests depend on the v of the patch attached to which addresses a few bugs that were found to be existing in thecurrent Clob locator implementation +Revised the material in the Getting Started Guide to explain how to run the demo but moved all the other tutorial info to this new guide +I would suggest to separate them by package as belowsrcexamplesorgapachesrcbsporgapache +Just neglected to close bug until now +Something like this? ij connect jdbcderbydbreplicationslave +Fixed in revision in +We repeat the license in LICENSE or elsewhere in the distribution which covers this attribution requirement +If this bug is still present and you can provide more details or even better a test build thats failing then reopen it and Ill take another look +Thanks for your insight and code submission +Thanks JakeReviewed and committed! +mvn clean install works t one of the jobs set to use a clean repo +From the post seems we can process the input and output streams before calling waitFor? +I have applied this patch and run ant clean test +In distributed mode grouped is only the supported format +I will add some codes to check whether the destination size is equal to the source size +Revamped code so now things are simpler than it was before +All the other handling is built into the class itself ll commit this today so you can have a look at it in tonights nightly snapshot +Now that we have as the minimum requirement we couldadd thisCan you please make a correct patch out of it against thecurrent cvs +We shall leave parallelism as in this case and then the later code will find it and use defaultparallel value instead +Combined patch for and for branch. +Ran these tests times on vmg without seeing the failure +I am thinking I should back out the JMX portiion of the fix for to correct this issue to get rid of the regression and file a new issue for the potential classloader leak that can occur if that fix is backed out +Heres the final revision +Thanks Todd patch looks good to me +Patch Vivek for applying the patch +This will help get the test failures for with hadoop downThanks for the contribution Jason +Attached is a patch to update the default and template policy files to contain +Otherwise the casting is not reversible +The latter routine only does WSTX initialisationIn the file this is configured by setting a property with prefix to have value instead +I have downloaded the nightly build of Jan and was able to compile etcI have added a news item to the news section of the website and have committed the change. +In Hadoop the serializationdeserialization is implemented by Writable interface +Obviously there were a lot of dependencies and I had to resolve them by crude means by commenting out irrelevant but functions +Reviews and comments more then welcomed +So I realized my changes most likely would resolve this as well +This time its of my own creation +Patch to disable constant caching +Exclude that getter by applying +Thanks for contributing! +If you do not want the extra carriage return to appear in your output document then you must not put such a character reference in your input document +Are we discussing disallowing sub queries that already work? Or are we discussing more stringent syntax checking +using in ER. +I named them badly though +The use case is that scripts may want to run where of the dirs is an error +Tests above are passing for me locally +Switching to AJP surfaced scalability issues in that connector +Lets store numeric types with longdouble +Verified with +h +That way the annotator could easily say that it cant handle a certain document +Not sure what happened but sftpjbossqahtdocsjbosstoolsupdatesnightly shows no featureplugin jars anymore +Ive added new unit tests to try to reproduce your problem but I didnt succeed in reproducing the error + +I just committed this. +feature for central instead that includes the central feature the needed plugins instead of having this different mix of pluginsfeatures between product and communityIn any case none of this can be fixed well in an update afaics so for now not moving it into Beta +to migrate the fix back to the service stream +Thanks for the clarification I have removed the extra test in revision the other tests do indeed fail if I remove the dummy service registration in the bundle +curious +The keep property will keep all the transaction log files so after the fact one can see the exact history of the updating transactions during the test run +Hi Gavin Does Hibernate internally use? Because does maintain a order of the elements +Not able to deploy the VM when I dedicate the Zone to Domain Root and Account Adminbut when it is dedicated to the Admin account then I think it should not be visible to the other domainsaccounts +Effectively those are integration tests which test Jibx itself in the context of a Spring application more than Springs own said were trying to use a test group for it now where those tests with custom compilation requirements can be easily includedexcluded in the build +So I for one would give up on the Java requirement +Thats no problem +Looks to have gone through +Pls find JDs comment I see what you are saying and I understand the deadlock but what escapes me is why has to go touch all the classes every time tofind the locale as I see threads doing the same +This diff fixes the Eriksyour patch is in svn rev +For the properties dialogs I guess that Properties would be it +for the change +Documentation update +If processing started already then adding a transform to an operation before that should cause an exception +We would need a way to pass along signal data to the remote JVM and the nailgun process running in it and then be able to kill that process +bq +Same thing for the standalone +fixed in SP and +The changelist is on my dev machineEither commit it or Ill do it tomorrow +Since the test case is not valid I simply removed it +Fixed with commit abfdaaccefaafabff +Inside the visitor it creates a which IIRC does a DNS lookup when creating the +I realize that would probably be a more involved patch +I think the ability to change the style class based on some server state is very nice and adds flexibility beyond the default styles +If its already an IOE dont wrap it in a new IOEGood stuff +I will recommend to do both in conjunction +May be specified from properties file or from commandline ant execution with true +A better solution would require changes to the API +I could not view any real connector data +headerClass footerClass still present in attributes list for richdataList +We are still awaiting the outstanding information for the Release Notes on this one +Ive walked through the code looking for this and it doesnt seem to be an issue +ehhh where do I find them? Really its just to get started +Thank you very much! +This parses protocols but doesnt appear to be able to make RPC calls using a protocol +Hopefully this will prevent future blowouts. +coretrunkrepositorywebstart +Does your Task based on some tests or just JIRA issues analyze? If it was tests show logs and examples to simulate problem +Thanks +I reuse the same test case structure of my previous fired whole workflow located in the +Thanks Chris +The other files in Jakobs list arent part of this patch +I would prefer to file a separate JIRA to make common between AM and JHS which will better address a bigger code redundencies +I have used the templates and you can run the test case but I didnt manage to set the SVN tell me if its ok to use the application inside the ZIP a lot and best regardsGonzalo Abollado +Given the double quote issue only appears if you mix TAB completion and not using an option name it is unlikely to be encountered by many users in the meantime +However made a lot of various improvements to the underlying code +The filePos field is named chunkPos +Dont think its so important to get this fixed for +Ive just committed this +Thanks +The second attachedfile is the correct one it shows a class that is definitely in the Repository +The plugin will run any defined in a script provided there are no other classes defined in the same the case of many classes defined in single script and if only one of them is a the plugin may be smart enough to pick and run the testcase +We cant do anything from myfaces point of view +I hope it wont take longer timeRegards am providing an improved patch having following improvements +Reflecting fix version as well as per Lukes comment he applied the change to the. +not relevant in branch +Removing pull requests +bq +Fixed in subversion repository as of rThanks for the report and for the patch +Now Ive fixed can you retest this? It should be fixed +I dont know how realistic that is but it does solve all the problems if it can be done that way +no i coudnt get it to work without this fix +Hey LewisI didnt end up mentoring this project bc the proposal came too late and the Apache folks mentioned the program was already over by that time +just the error part if possibleRelated whats the reason that we use XML formatting for junit output? Is it for hudson or something? Its more difficult to read by hand +If everyone is good with this patch I will commit tonight or tomorrow morning +So we cant fix this at Axis levelThanksIsuru +Hence postponing the issue to post is fixed using ref counter techniques +I rearranged my into a plain Java project and get the same stack trace +The response String is normally returned as JSON for the client +i just committed this thanks! +a comment +With the CAS class loading improvements this additional serialization is obsolete and the code can be changed +It is about x faster than the version based on the microbenchmark included in the patch +A lot of the unit tests failed but I fixed them +See some of the changes in corresponding common hdfs and MR +Please reopen if you attach a test case +List volume also fails due to same issue faced with list vm commandl after delete the vmExecute cmd listvirtualmachines failed due to errorCode errorTextUnable to execute API command listvirtualmachines due to invalid value +Username djencks +Therefore aforementioned code should throw +prior art inAlso tag as nightly? Weekly +Apis to look into helixadmin clustersetup dataaccessors multiple paths to access same data and its confusing config accessor setting constraints messaging apiThoughts +Committed. +maybe test or just test +second thought on +The assertions of this issue have been implemented together with the assertions of +This patch is a baby step towards on flaky tests +Ive finished porting the new Failover code to Windows and Mac Ive resolved several bugs that I found as I did so +Since this patch changes the class so much submitting a separate ticket for the NPE seems futilePlease keep me posted when this makes it in! +Aslak said all we needed is the nose we have a nose +Even if the ajaxjsf filter is not configured the example does not work +Why did the region server crash do you know +Its only been in for a couple weeks +Shouldnt these also be marked with TSDEPRECATED so theres a compiler warning +Seriously bummed that were not seeing stress wins +I think we should fix at least for and maybe raise buffersize +looks good will commit if the tests +Proposed patch +Just making it clear here +thanks Jeff! +I am not entirely sure how it affects the rest of SOLR +Ted thanks for the patch I just attached a v with the corrected commentsBasically there are two things herea The is independent of whether the namenode chose the stale data node as the primary DN to do the recovery +Unmarking for because patch has been committed + all updated +This archive contains the initial code of the SKOS plugin for Apache Forrest +The patch contains the bits needed to include the new bundle in the assembly and make it part of the default startup +Hi Fabricedid you also fix this for lineEnding? I just tried the latest and lineEnding is only applied to files and not toDaniel +However if there are concerns that this fix may be to broad I can always add workarounds to the upgrade tests insteadThere are no official releases with the necessary fixes for or +This has been fixed in which has recently been tagged and will be in the next build +Running full build and committing +the Icons as requested by John X +Please confirm they have been applied as expected +The flag is about resuming builds not the pl or am +I was not able to reproduce this again +Vitali just open hibernate configuration file with Hibernate Configuration XML Editor expand Session Factory Properties and on Connection or any other item in that category +i download the source code from svn and compile it all by myself +That said with a certain argument value the system might be exercised to the point of failure +Thanks Norman for the quick Can you checkout the latest seam src from SeamAP branch and test this example with EAPWe can then pass it on Release Engg for an official build +Gary Lukasz is going to discuss this with you to find a suitable solution +Thanks Doug for pointing out the problem +So the functionality of the button is as expected +Hi Eric I attached a new patch +Martijn can you triage this for? commit if you think its ready otherwise remove the fix version +The Chinese localization of has been completed +ok i fixed it so that the that are used with the constrution of a page if you give the params through in your constructor super call are remembered and used when generating the stateless linksI added an extra constructor in Page and and improved the javadoc a bit so that it is know that you have to give the pageparams through to the super call if you want to remember them for that page instance. +I read through the patch and dont have any improvements to suggestThe patch builds cleanly in my environment and the tests run successfullyTiago do you have any additional work planned for this patchI am intending to commit this patch please let me know if theres any reason I should hold off +Updated patch for DRLVM with changes required for GC to work +Its a problem using reduced dimensionality and these original points have almost identical terms +This is the last bug to be addressed +A couple questions more +MuraliPlease provide a link to the functional spec and the feature branch name that you are going to use for this featureThanksAs per original requirements posted in the bug will orchestrate route health injection for route advertisements when EIP is transferred across zones +Marking all current RESOLUTION LATER issues to be against the Nightly Build +Joe how can I detect what the images orientation is in JS so I can send that information over to the server? The exif data isnt available server side when I send the image via base. +Thanks for the quick reply +im confirming the version liferay is using +I think that the main benefit is to upgrade to the last version of asm since cglib is considerate deprecated +This is a just a suggestion thoughI was also wondering about how one configures the sequence of middleware modules +Thanks +Attaching patch that resolves this issue +perfect applied cleanly now +plainin testsstrutstestresourcesshowcase +Sorry the above is not sufficient the will return the reference to the end server not of the reverse proxy +Heres a patch that should fix most of it +Thanks i applied your patch and also fixed a minor typo there and in the cocoon source. +Update patch slightly due to Nicholass comments in Jings comments with addressing one NPE issue and remove some redundant code +But what happensif a is built from configurations using differentdelimitersWhat should be the actions following the modification of the list delimiter? Wecan either re interpret the properties thats turning the lists back into strings usingthe old delimiter and parse all the strings to split them into lists using thenew delimiter +Andrei please create a new issue for your suggested improvement and mention this issue when doing so thanks. +Doron feel free to work on this I wont be able to do any work on this in January +Try hudson +Ive just committed it. +See for testing information +This is not necessary the case for a distributed cache +This is possibly not related but I came across this issue with Glassfish V using weld and hibernate and Seam faces module +Committed to branch and trunk thanks Stack. +so used currentdatabase instead +Probably we should do something with the testcase itself +If anyone finds anything this ve read some things on manipulating confluence via tomcat through Ant but it is quite tricky and not a quick task +Applied Patchthanksdims +The value defaults to Ive not set the initial compaction log message to debug and update the nd one to display as requested in the issue. +Having this get overridden by Java code coming into basically breaks Java integration in nasty ways +it starts with a stopword +Yes there is an M version but it is a very basic one and possibly not much different from an older EAR +I think I will open up a new issue +It was an interesting experiment and could probably benefit from an upgrade to the new version of Lucene +Should have been fixed with +Given that CVS has horrible security leaks and is practically unmaintained since years isnt? +With the PDF Im worried that many people wont have installed and it will be difficult to write an ant task to compile it but if having a PDF in there is a problem I can remove it +Scheduling to fix in +OK will apply same solution as +The feature is now available on masterTo configure IP reservation in a network specify a valid guest in the updateNetwork APIList Network response will display the reservediprange that can be used for non purposes. +Strange +the problem seems to be in the reading the file +appears to be no MS project for the example in the branch +bq +Added a missing privileged blocks to the save file into the backup +Im still not sure what is the impact of BLOCK compression on random accessRegarding the sizes parsetextrecord size is larger because for small chunks of data the compression overhead may far outweigh the compression gains +Please fix this! It has been reported a few months ago and is a real show of the people out there use Subversion and nearly everyone has to release a piece of software somewhen +adding the maxminaverage in the UI +Thanks Nicholas! +I implemented pluggable package loading in xwork in such a way that was preserved +This will also affect if we decide to allow dependencies and is also causing grief with extensions +Committed with nit Thanks +note there are no groups created for users authenticated against +Its still an ongoing refactor but the FTP part should be mostly done +Or you can email it to me privately if you dont want this public +Ok so I got a plus one from stack on the review board that I posted +Changed key from INT to STRING in definition of concat in +I wasnt asking about the ivy settings files I was referring to the ivy configurations +In this case I can point you to information which shows that traditional blocking IO performs betterFor Yes this as well as quite a few others since the release of Synapse +Update doesnt work under MPI or PGAS Sockets either +I added some comment up on RB but this is close to committable revision After another fishy and not reproduced error on +Id like to reopen this issue now that Atlassian has introduced a Webhook functionality as a request to deploy this feature +If its not taken into account its either a typo in your pom or a bug in Maven itself. +Should this one be closed +Im assuming here that negative boosts with bq was simply overlooked in the past +Chris also added you to contributors list +I have created the repository +I would encourage you to keep the maven archetype up to date for the quickstart because I think it will help with getting people to try it out for the first time +This is most likely related to. +An example POM would be nice to make packaging a bit easier +The would contain all classes and as attached artifact it would provide the transitive deps +So perhaps we should just leave it all alone and just commit the patches that add missing license headers etc +Thanks! Can you fix the key creation on line of so that it is created with too +does it crash after the launchdoes it hanganything in logsconsole when running eclipse with debugif hangs can you get jstack thread dump +I reopened the issue +I am still waiting to get the log file from the customer that should have the full stack trace in it +The invalid subscribe causes the problem +So maybe this should be either a configurable number or something that is derived from the number of mount points hdfs is hosted on +But IHMO forcing users to change their regional settings just to have some software reading CSV properly is not an option is it +Uwe what did you use to generate the override +One hazard with doing that is a corruption in the middle of a log might be considered an EOF +Ive now created two different printer modes for zip zipbin and zipjson to distinguish between a binary file and a json file provided to the think we have a first working version now therefore setting this to after release +mess +Just that the invocation command does nothing +bq +I have no clue WHY they fail +Done thanks for reporting please check with the latest snapshot your machine is too fast? I mean regarding Ive messed up two different topics +AndreIt might also be useful for me to replicate if you give me full steps for recreating the issue starting from an empty workspace +Tested with multiple key cases works fine +This is the patch fileSorry for the previous wrong message +This is not a bug Im open to suggestions for improvement though. +It looks like the bad news here is that gossip makes many assumptions about a node being alive when it hears about it and has no provisions for keeping removed node state +Corrected example +There are still some bugs but I will be on Gran Canaria for a week so I upload the not quite finished version in case somebody wish to take a peek while Im gone +So you loose properties transition methods dependency ids almost everything on both components and handlers +Otherwise no way to escape the variableFor some strange reason my +Doh I hadnt realized encryption was implemented now! Spent too much time on old versionsVerifying IP addresses might still be useful as per the but its a minor point if proper encryption is in place +This is already done +Close +done +Any comments on the approach are welcomeIve written a test for a cluster brought up by Pallet but its not passing yet. +Thanks for your understanding +Ill try to ping Artem for reply on my comments +yes of course you are rightSo you prefer going back to the view tab instead of the info tabregards prefer going back to Attach tab +pushing off remaining issues in order to release version in JIRA and create release notes +Will commit later in the evening unless there are any objections +I also added a small test new SNAPSHOT has been deployed please verify that it solves this issue +Note that this will also remove the github version too +This is good +Thanks Knut! Attachment includesthe braces to improve readability in andalso adds comments to a number of the code changesI intend to commit this fix to the trunkThanks all for the reviews and testing +Modified to import another module so that it is exercised as part of current ee scripting tests +Your data may only be nominal and thus have no need for sorting but I think the changes would require more than switching between a and +Youre right Ortwin +Committed to trunk +Theres very little rhyme or reason to any of believe what we need here is to make some real decisions about encoding and look at each place in where we transcode to decide how to support those decisions +INFO Running for rows took ms +Fixed +what version of the server adapter are you using +Im closing this as Wont Fix because while its possible to catch syntax errors when compiling the view functions its not easy to know whether a syntax error may have been thrown by a user calling eval or something else trickyIn other words its hard to disambiguate between errors caused by the view code and errors caused by the input dataAlso this hasnt been updated in a long time +I did not try the . + this only ever happened to me with bulk imports and then being able to quickly drop the incorrect table for a restart +So Im good with your proposal as it isthank you very much +By putting ourown in we are down to +Attached class loader test seems to work ok now +pass this time or not +Might have to sit down with Jon to make better test for the blockcache after hi is ready and with Ryan and look at Blockindex since it is kinda hard to test as it is right now +still fails +If you do then you get the SQL standard behavior a Jar ids are mandatory +Only Apache projects can be added to Bigtop +Reopening It doesnt appear that you actually pushed that doc change. +Hadoop has a property so I suggest going with instead of Pigs +batch transition resolvedfixed bugs to closedfixed +Assigning to me +scottWe find that the only persist the state of competed and dead jobs in DFS not including the state of jobs +Patch applied at +Hi Glen can we get back on this please and try and resolve the outstanding issuesThanks +See for full discussion of the issue +Didnt find any regressions +Visual part always updates to the correct state +I got around that using a persistence delegate and then hit the same issue on another object +Thanks Shelan +Fixed for release with minimal changesIdeally we should pause or stop the timer tasks when the transport is pausedstopped +Correction only is affected +user requested to close the issue +Please apply the attached patch file to fix this bug +Updated which does a better job of testing that error fieldsaction errors have changed when deciding whether to shortcircuit the validation chain or not +Id prematurely set it to +Changed Target Version to upon release of +The MOP will not contain custom meta classes anymore +Ill look into it a bit more +How about enriching message reminding user that reference should be returned to the pool +But because of this Embed APIs mode test started failing +New patch is ready for reivew +Now the for logtail could be catch after ms sleep during this sleep interval the logtail could still output but the will take effect after the ms sleep +Changing affected version to and fixed version to the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Marking this as resolved since we have a build now with the fix +I add configuration option +Passes all local tests. +Its OK +bulk assign all unscheduled issues to next release +Ive also now changed this so that the Advisory topic is not created until the topic is subscribed to +Max +Can you explain steps to reproduce or post an example app I cant seem to get init called twice in my thanks for that guys +I may miss some words on the sync up meeting +Also I think that the JS API in chrome apps is quite better and than the one provided by node +The plugin defines a dependencies on the mysql jar +This does work we have unit tests asserting the behaviour it is likely a problem with your domain model or you are trying to use the side of the relationship to note I tested against the latest code in SVN so there might be some only difference I can see is that I am running this within a service not a +And it is related to but the fix for this one will have to be inh the where values matched by identity constraints are recorded +Was applied over the weekend thanks for the hard work everyone. +Patch checked in +Thanks JessePatch applied to NIOCHAR module at repo revision rPlease check it was applied as you expected +I will be validating this with a real cluster before committing them +Now the dev studio launches fine again +You can upload files under More Actions Attach is my simple project +Oops sorry indeed I saw html and not safe LeonYour patch is in trunk at revision +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Now Im much closer to the real problem and +Great is there a patch available? Or can you describe the simple change specifically +Resubmitting patch since failures do not appear to be related I cannot reproduce them +I think this would be a specific hack +To use this plugin run mvn emmaemma +But in any case to commit +And in the jira was mentioned editing a local file which to be honest Im not sure what exactly is meant +The config files do not have to change The have to start talking to the new machine using the old IP addressOption Another way would be to start the a new instance of the JT on a different machine that has a different IP address +Of course it could be part of another major issue. +Martin in fact you can always execute the plugin in the root directory? The problem for you is about speed thats why you generate only the project where you changed the pom + +the same properties are used across the board for all components +I do not know when will be released +Attaching a fixed patch +Change zk config to allow unlimited connections +a lot for the patch + +the latest patch looks good to me +patch looks good +Skimmed over it +Attached is a sample process deploy archive to demonstrate the issueThe instance cleanup is enabled for ALL +to look for typos that are user facing +Fyi I just looked into this issue briefly and the noscheme errors are from the inets http client and from attempting operations without the httphttps portion of the uri +If theres no handler in the map it means thatone of the followingbq +Thanks +Hmm +Chunhui Pardon me +Then you can just scroll down and see the view for each collection +It may make sense +Looks like this has been fixed. +It is an one of yall test this please +I could include the constants in the itself +buildivylibcommongriddevbin + +Thanks it is in the code base but did not pass QA checks for +Will make a fix shortly +Shouldnt in this case it be at least default cache and not null +Please let me know +When we finally move to Wiki it already has RSS feeds for every page +Just to let you know after converting inner classes to separate class files I have not seen any of these issuesHave not upgraded since Nov +I am uploading new updated patch for sameThanks +Thus the data will be true XML instead of big CDATA sections +Advanced users can set this value to something else if they need maximum performance +I will update the patch +Also make sure to click Submit patch when you want to mark that it should be reviewed +This task is now complete. +Committed to trunk +Please fix issue +Applied to and to trunk +Goods news indeed +In the grand scheme of things this bug isnt really any more important than the s of other open bug reports +I believe that the problem with responses is the same as in for which I am preparing fixes for and +close issues for. +Thanks Amr +Which quickstart fails due to a missing dependency +Its got my vote +It is working now +Sure +My point above is the core issue it has nothing to do with JCA per se its a general problem we need to solve +There are lots of other things that are confusing there also but we cant really fix them +No that would mess with other I can do something with I hate you +Duplicated issue +Accepting as weve already started working on for done and instructions added to script commited by Scott +Thx +Comments on the coding style Indentation needs to be fixed to conform to hadoops coding conventions +Thanks Florian. +Patch to applied on for in Tools. +Anybody know how to determine threads? Oracle JVMs dump daemon indicators in their thread information +Thanks Nicholas and Kan! +Deferred to defer of unresolved bugs to +This is just localized to the quickstartsconf file +Not sure why the failed tests here +Because we CANNOT MAKE INTERFACE CHANGES between minor revisions +Small fix remove versionId from response and from handle methods in +Thanks Carl +This is a clear method signature and forces a custom strategy in the implementation +Is this a standard behaviour for PDF files created with or with usage of ligatures activatedI dont know +So I think we should stay with names only +EBAAnd in the patch I also added EBA portlet into jetty buildPlease review and commit the patchThanks +Will add the right one soon +So in every mini batch we need to know which all Mutations getting applied this time +For best performance use it with the Journal +I am working on retraction +The failed tests are unrelated to the patch and I think the current semantics are fine so feel free to commit Chris +Houra! +thanks Rob! +Will do some work there +Its an error in the plugin for windows phone +We could fix this by introducing a could new kernel exceptions or exception types but it might just be easier to revert to having the validation done in the persistenceAny objections to fixing this the easy way +Username alexbloss +I was very surprised to see this JIRA closed as Wont FixRalph I understand that development on Flume OG has slowed but in the absence of any discussion in the community that Flume OG dev work has stopped Im reopening this JIRA +Well done Marc after so many time and tries you get it +So should we unlink this from the release +Were certainly open to suggestions though +Thanks for the contribution Jarcec +This does work with ESB Server the error above is with AS +ASA v can added with no issues +Please check they were applied as expected. +committed to trunk. +Could you not do a checkout and create patches then +Updated the description of +The work should be complete though +But thanks! +Passing a better error message back to the browser would enable better error reporting to the userIt would be preferable to see the error along with a line number for where the error occurred +I fully agree +We should not get this situation now after handling the specialized exceptionsI will close this JIRA. +You cant use single column instance in multiple tree tables +New patch addresses this +Jason +. +The timestamp should be promoted to a first class citizen +We may end up simply switching to a different launcher at some point in the future so this may become moot +Seems like we need to return the sender and not the recipients +The IT test fails for me and it seems pretty odd that the test poms generally identify it as rather than +I think that there is a tradeoff between that and the current attack of consuming memory by leaving an accidentally booted database up and running +It was supposed to be checked in with the patch with until I found out git doesnt check in empty directories but since the main repo is svn then that shouldnt be an issue +Here is simple solution to that problem +This patch catches the exception calls destroy on the aggregate and rethrows the exception +As I dont have the Nazgul framework in my local repository the plugin installation currently fails + +You are tricking Hibernate by mapping two classes to the same table and mapping the same collection twiceWe need to fix Hibernate to handle this situation +Just my two cents +Patch applied +The bug causes to enter an endless busy loop +Hi NigelThis certainly sounds like a valid issue +What I committed to trunk and amend to do both coming up +Default assignment for build goes to Martha so to Paul +Wow thank you for your contribution +it was decided that the first milestone of JBT JBDS will use the alias Alpha instead of M to adhere to proper qualifier guidelines and to allow use to then do Beta CR Final and GA releases each being updatable from the previous wo first having to uninstall +This is actually a client issue +We can set it smaller +Actually it is contributed by Samuel Guo +But here again I do not see different policies for different files +Jeg jobber ikke lengre i JottaI no longer work at JottaTrygve Hardersentrygve +We can consider restoring this issue when we have some more compelling use cases +what version are you running +I modified test because otherwise it fail since by default RESOURCEMANAGERCONNECTWAITSECS is too long for this Xuan +I have a local change that before setting and resetting TCCL in asks whether the TCCL is already set to the contribution classloader +I have added quotation marks into JNDI attribute in were you able to check this issue +both was fixed by the fix to +We will be adding some more descriptive output when a user needs to refresh a bundle to pick up optional imports and may add the previously mentioned options to +I think we already have it +I actually considered that approach but I know how the developers here are so I didnt do that +so i dont like the dual mode where we sometimes make our own but you can set a different onewe dont clean up correctly at all wrt this in today in my opinion +Hi Luothat will be done when the serves out the resources +Attaching the log +Here is the stack trace for the exception attached +According to the stacktrace you are getting this on releaseprepare not release at the code I would hazard a guess that the distributionManagement element in your POM either does not list patsh to the trunk of your code Does not have anything in common with the tagBase specified for your this dont think this is a bug but a better error message would be useful. +Ill commit shortly. +Click Install Packages +address Yongqiangs comment +Resolving for +No by mapping the filter to a specific position in the filter chain you can delay Weld from accessing the request +Hi Suresh Could you please take a look for committing? committed the patch to and +Merged into the branch. +Oops +I confirm this is fixedThanks Emmanuel! +Sounds like a good idea. +Or we could add a config option to disable the check thus enabling longer than chars for where this is supported and the default would be check enabledAdding config option in this case seems unnecessary to me +Option to the java generator sounds goodI like the idea of generic directives at the level +furthermore i see no use of either or instatefulSessions or so basically all stateful are referenced directly not via keeping them forever if not disposed explicitelyseems like a potential memory leak to I will probably property +wiki moved +actually i did find this in the logs INFO nntpserver NNTP Service uses default thread group +Tests passed cleanly for me on + +Make it throw instead +fixed by r in trunk +Open a case in the CSP as an +The only thing that does not work anymore is that the exportregistrations are not closed anymore by the +This patch implements changes to the driver and mapper to utilize multiple reducers +Editing an existing one is still not possible as I originally reported +Editorial pass over all release notes prior to publication of. +Ive modified it as per your suggestion but only just noticed that the calls in the format method can interfere with the extra rounding code for some values +project file by hand after mvn eclipseeclipse run +I found an important bug recently +Also why is a XSD schema being changed? When it has been published then IMHO it should never change +There is skimpy that tells you want to do to your schema solrconfig files +That is you should not include my earlier suggested fix +Code integrated +newest patchs output of ant Dlibhdfs Dfusedfs Dtestcase +I couldnt get working in Cygwin either so I thought Id edit the code to give me stack traces and see if I could narrow it down which led me to try to get it working right in without having to keep going back to Cygwin +Question where ismrefcount initialized to zero +We can upgrade to a new version once it is fixed there +When we move the leader for a partition one at a time we also remove the broker being shutdown from the ISR +The testcases run fine on Windows +I have running it just in but got the same result with the compiler levels of andWhere to locate intl resource please? I have search the printed message in the derby project but have found nothing +And providing something like that is IMHO outside the scope of Jackrabbit +Nicholas you good for us to commit thisLets wait til for a day +Please prefix summaries just like the mailing list +Even thought an index is still used thats not what I wanted +The possibility of permission problems is inevitable if root creates the files and apbackup reads them +An even simpler and more general fix is to modify the Base utility class to ignore whitespace characters +When handling the findbusiness request jUDDI was purposely not including the businessKey in each serviceInfo element because businessKey was available within the businessInfo element +workaroundUse a java class that works and just make calls to it +resolved trunk issues +Hi HedererWhere are you seeing that the summary attribute is mandatory? Warnings in a validator does not equal noncompliance or that a tag or attribute is mandatory +Can you try patch attached there and see if that fixes your problem +Adding the Release Note Needed flag to keep it on the release managers radar +Patch checked in. +Seems useful +Thanks +The child seem to have completed +Ive only made it through a few examples +Early dev builds will be availble in a couple of months +I just saw this email +Yeah we need to remove the the after the application context is stoppedThanks for pointing that out I will commit the fix shortly +Besides similar files in HDFS and MR dont have it + objects are proportional to the number of files created +I get the expected output + +Its expected the are protected since they leak other sensitive information +Further research indicates that there is actually a ton of excellent stuff we can do in this arena also this optimization will be useful to the JBREM project. +I have added some short methods to allow the test to check if a new bookie is running +In release for AS this release seems to be broken again as although the JMX console and the Administrative console both report that the queue is in a Paused state the subscribers are still able to consume from the queue +I dont have much experience with multicast so please let me know what additional operations are requiredWhat do you think about adding transport? Then we will get to know what is missing now +Ive committed this +I need to make a business mobile app for now and text field do not work +Hadoop isnt a compiled dependency +It looked like there was no test case for setrep so Ive added one +A config editor requires a way of loading and saving the config file + will fix should be fixed via the updates to use autotools. +However we might want to do a small amount of cleanup on that stuff in the future +If you can get the setup similar to ESB project I can help on the details for some classes +Any idea how to configure Maven to support the use of this tool? Is it something you have to put in +Heres a patch that corrects the issue +Hi Svante yes a pointer to the release notes in the README would be fine +ESB should be used +Im reverting this change because it makes me nervous +Wouldnt one of the other levels ie Runtime have sufficed +It would be great to have this ready before we cut Camel to + to hudson +committed it to master +The patch looks good to me +Heres yet another try and +Anyway creating the journaldir and ledger directories will be done by bookie if not existsThis should be called by admin before any server starts + +Looks to me like a problem with primordials +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +in last site commit r +Looks good thanksbq +If you are using the Fuse New Project wizard it will always create top level projects +Converted to groovy into trunk rev +To workround this we change the foreign key to nullable in database +pom says pom to say version turns out the had a byte order mark at the beginning that is making maven squawk +thanks +Instead I just get rows inserted +I guess this is a good opportunity to +All done +Ive committed a first version which adds two attributes one to add a comment and one to add an exclusion comment +I think we should remove the rollback there is no reason we should keep and it works perfectly well assume you are referring to mode right? Lets get the jee mocksin and well discuss this then alright +Thanks SianPatch applied to LUNI module at repo revision rPlease check that the patch was applied as you expected +Hi ClausI have added documentation or the Camel SIP component and also updated the components list to show the new componentCheersAshwin +yes temporary until applied upstream though it was noted that it may take a Very Long Time for such a thing to occur so it could be a fairly permanent see no legal or social barriers to the project supplying a patch file to a third party component that is then applied by a user. + +I would prefer Kilobytes buffer size +Updated patch +Default value is date +Simply changing this doesnt work because relies on the AEs metadata to build its annotation style map +May be using some generic type with our SerializerDesrializer classes. +But I will refrain from boldly going where speller checkers have never gone before +Sorry but it seems the fastest and simplest way to demostrate the problem +thanks simon sorry about that i will check for merge help next time +as an +Still randomly see different conversationManagers invoked for same session +This url returns a need to know exactly when the value is set inside the because i need to stop the listening of the messages for my specific customer +patch to fix the note configuration property changes required in when it has to talk to SSL enabled haddop HTTP Web end pointsetcconfshould define values for keys I tested with the following values for the keysetcconftruststorepassjksYou should have imported the certificates of your Hadoop Web Endpoints into the truststore. +Closed after release. +Patches regenerated from revision Remove asserts on and start using Proper handling and replace wrapResult methods with copy +Testing in progress +Noticed I had screwed up the part of the patch +With this patch I dont see any significant performance difference +Apologies Robby my last comment was incorrect +Probably a incorrect inline usage change the extern inline to static inline looks fix this see the patch +I like using the console output in firebug to see what my ajax responses actually are +Sorry if I lead you to think so I was merely raising the question +Tomcat is using Jakarta commons DBCP and do not suffers from this problem +Paging and bridge code are identical when in and standaloneThis strongly points to a versioning issue +Please update this ticket with them if they exist +this didnt show up in my env +So we just adapt from that +Previously would accept messages regardless of the content type the addition of content type detection introduced this error when an alias for a content type is used instead for this reason it is not possible to write a test with a client as clients all use the recognised namesThe attached can be used to manually test this with SOAP UI just set the Encoding to UTF in SOAP UI before sending the request to reproduce the failure without the patch and verify the message is accepted with the fix in place +People with serious production environments dont just throw new Lucene version on top of the old one without carefully reading +Will commit shortly +Is this possibly a duplicate of +Ill go ahead and commit this as it is +Commented out the store list and moved the issue to +Should this issue be reopened + Elliott +I have commited some initial version of xpatch support +Uploading patch +Also since then the word list from Lucenes was modifiedbq +I can take a look at themThanks +Hmm. +ThanksanywayJoerg + +Yes +I have committed a patch to remove the default URL completely which means that OBR must be configured now to have a repository +Backported to with revision and to with revision . +JBT comes out with and the fix is not in +Someone please fix this issue and release the new version +the test case just hangs there +Move also to srctest classes revision isnt possible to that classes as theyre in use by revision +Especially on modern bit Linux hosts the overhead of virtualizing other Linuxes is negligible +Added bindings for serviceservicesnmptypeadaptor to variasrcresourcesservicesbinding +Oops I already took the liberty of doing something like this in revision for +Maybe I need to open another issueThanks +I get the same error with or without an abort query included +Thanks for the changes +This new patch includes Stack suggested changes Stargate code changes for +Ive just committed this. +Backported patch b to the branch with revision +patch to fix some comments +I just think since in practice its shared so we should move it to the shared directory +Maybe appending pathPrefix to the context path would be a general solution +I will give a patch this week +For the st issue we have a workaround is that when we deploy the exploded make sure we stop the Server +trunk moved to relevant parts of trunk are now in the init target of trunkbuildresults and trunkbuildresults I broke anything we should be golden here +I modified a bit so the has a custom prefix in its eclipse project name +Ill probably take another look when I get some spare time Thanks Neha +Sorry for the delay in respondingMy set up is as follows root project srcmainresourcesless child project srcmainresourceslessIve configured the plugin to pick less files up from targetclassesless given that the js import plugin will copy there +Sounds fine with me. +Nick please check the issue I have update Beta branch +I will find out from our client if theyve been able to work around this fornow +thats confusing +Marking as fixed. +It doesnt use the transport layer so it doesnt have xmpp jms or local transport support +Approved for +Could you show us the thread dump for? I guess doesnt get affected by this issue +New patch fixes to mark patch available +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Attach a new patch +not sure +The Dictionary is an Entity too +Stepan please get these EFL corrections committed +Hopefully the changes should also fix the problem in bug +Eugene please verify that the issue is not reproducible anymore +Attached patch augments to test for the fix +I didnt spend any time optimizing this code at all. +Andr please review this PR +Thanks Karl +See comment for further information +Committed at subversion revision +I currently have the setup ofcsrcbuildcsrccsrc +Confirmed the same bug is reproducible with lazy resolution turned off in JET +patch for have been applied long time ago +And we want to share some of the components which can be common for others +I noticed that too +I think the other patches in flight are either small or otherwise straightforward to rebase as they dont have significant overlap +The issue is that we need to have in Operator because uses it +It detects logj and tries to use it +The problem was fixed in in +As I originally suspected you have two separate bean definitions for the +Am I right +Jon has said in the past that we should use our state rather than go to + info about Dialog is should be added on css class adding storage and new section Pages Styling is created +Getting the response content entails getting an from the instance member +OK I tried it with the release and I do get that exception +bulk defer to defer to +Assigning to Henry Z +Well be moving these templates to the Wiki so feel free to add your thoughts there +Hello GavinPerhaps you can check now +But dont mix interpolators just pick one +I believe we are good but thanks for following up +Thanks Hiro. +I confirmed that the patch does indeed get the docs package build back to workingIm going to commit this momentarily and also perform the two svn rms that Andy suggests as well +patch file that contains the fixes +Fixes patches variables bug +I am afraid this is intentional +Wish we could figure out the OS env variables or something +Added license informationAdded the package target to copy hive distribution to hadoop tar ball locationI was not able to complete the javadoc intergration as there is some generated code which we want to exclude and also it needs changes to the hadoop which is not a part of this +Verified in CR +the version of our patch applied against current trunk +I need to investigate how some of these changes will impact geoserver +Ill do a little more analysis and if I figure out what the problem is Ill log a new issueBut I think this can be resolved +So a Table object seems to be an overkill +Still seeing failures with these control operations +Hi LenThis issue is marked as solved how it was solved and were can we get the patchThank you +I just committed this to trunk and. +Great + for the patch +Thank you +since block is on pendingReplications NN does not schedule another replication +I have modified to do this +I think we should commit it as is and deal with the slowness separatelyCould one of the committers please review the patch thanks +I will try to understand why the interrupt is having this delayed impact on the IBM jvms +I tried declaring the version attribute but it had no effect +Hopefully this one will work for you +Seems like a reasonable compromise to me +I will take a look and get back to +I committed an alternative suggestion. +So the patch is not complete +is already fixed in +Integrated to branch and TRUNKThanks for the review StackramkrishnaIn the future please concatenate test file patch to main patch +Thats how I do it +Unassigning since Im no longer active on At revision +facing same issue with KVM agent agent never comes up even after restart the this is kind of urgent +For this purpose now Im reopening this ticket +Richard if you use Eclipse can you provide the code formatting preferences you used in the attached server +I recognized problems with the messages myself days ago when working on bug +Regarding the interface presenting integers to a user as an interface seems wrong to me +Committed. +Where can I download this server from +However if you feel comfortable with Branching I have no have tested this and looks good +New patch for hadoop +I reverted this because trunk no longer changes to src and seem unrelated to this issue +It was working in rooWhy has this not been reported +This is pretty simple to reproduce with mvn site +Add note to Getting Started on how to disable clienttrace logging +Since this was back ported from to branch have updated the hadoop dependency toAll the unit and integration tests ran successfully +Same for good though Im not sure if we even need to depend on any longer +SVN commit +See comments on this issue +Thank you. +Sorry that I havent had a chance to put together a simple reproduction case + +Will commit in a few days if I hear no objections +tested new patch on the Mac Mini we have in QA +Thanks +i will look into a first step we will compare from which point the context defined in with the in the body is that the context does not exists yet +This patch backports both and +Just think it is wiser generally to require few pings missed to mark a connection broken +updates on this David +The two argument implementation also uses the auth tokensign only has one three argument implementation which takes a timeout value as an argument +This button would imply specific code but would avoid the hard code in the pick list +Marking as a dup +Actually SD does not have the concept of a managed type so it feels like youre somehow activating the JPA module as well do you +The max is K which is smaller than most anyway +Some support for tree strcutures is provided +One of the nodes ended up with two index directories today +to to avoid test case +use hibernate toools jira for reporting hibernate tools core issues please +I agree that being able to extend the API without breaking is important +Committed to trunk +MVEL has been released with a fix for this. +There is no reason not to make it if there isnt a performance impactIn my general experience with recent hotspot final on classes almost never make a difference +I do think that it could be done patch to get improved error messages is also coming soon +lgtm +Yeah I think youre right +This was triaged for SOA +However this is the extreme case +Verified thanks Stepan +is clearly not ready for us +conflicts with in both cases is still flaky +Thanks +Grantas I told you on IRC I need a list of confluence usernames to populate that group +The other way around people will always presented with actually internal methods by their IDEs code understand the compatibility concerns though I wonder what you mean when you say to leave it by now +Thanks BenThis is compiling fine on my end +Promote this to a blockerIs this relevant yetThe client is stuck in an RPC to +log verbose of run +I think my vote would be on halting all processing on the socket until either the delayed response is ready or the connection is closedI hope browsers dont do pipelining of AJAX requestsI dont have any preference for BETA or BETA personally +This makes the project to generate an empty JAR but at least its visibleeditable from within for the assembly projects there is no easy way around here +This should be resolved with upgrade to PLIDM +I plan to commit later todayI renamed the new methods to checkIndex deprecated the old ones and fix their javadocsI also added javadocs for plus a warning that this API is new experimental and subject to changeI also changed thew to be package protected since its type is which is not public +is state of work when this issue was because this issue was moved out of +Though it does not pacify findbugs it ensures that kvstart is accessed consistently and makes it easier to reason about the spill logic +corrected mistakes in test case to cover some additional conditions in which polymorphic binding fails +is yet to be published to the maven repo +Might be needed to canonicalize if we dont find another way to make them smaller +The only possible drawback I can see is the need to send longer strings between processesAnother approach might be to make the protocol stateful where the first time a counter name is sent in a session a String is sent and thereafter it is only referred to by numeric ID +In general we rebuild all build dependencies +I dont see anything jumping out at me +I succeeded to write create custom type Money from and custom converter to Integer multiplied by in mongodb and the read by dividing +If is not there how does the WAL entry happen in Accumulo +Checkpointing a work in progress for review +The patch as attached now implements the DN recalculates to match the checksum format +handling moved outside of the if statement that only considers it is tests are present +Attached as a sample how to use and what is output +This is exactly the one of the usecases that covers I have discussed this with Martin and now he is also convinced that covers his needs +Fix should go into ER +With this implementation there is no shared buffer and each thread has its own buffer and writes that buffer to a ledger which obviates the need for synchronization on logSyncI understand that there might be a concern about the semantics of the file system but in the way I see there is no such a problem +Also ensure that there is deployment to of snapshots +Is this a bug or is it so by design +shuts down the scheduled should fix the problem as reported in your stack trace +So its impossible to programmatically play the MP file +Seems like stuff is not hooked up +find the cause then find the solution. +duplicates +Closing old bugs. +Closing was resolved +Sorry didnt notice someone created the same issue a bit earlier +But I would expect that the objects are stored within a transaction and then it should be possible to have with +Bulk closing stale resolved issues +If the ae or tae or any other unknown option is used in the command line appropriate exception is logged and the utility is terminated. +Fixed all the above commentsAlso the retrofitport code in doesnt seem to be necessary since all registered to token are already initialized to storagePort +Hi Rishi I just noticed that the patch includes references to gora artifacts within the filemgr +As it stands Im on the way to stop using axisToo bad good library +Can you review this trivial change please Robbie +I guess is a better will be done +I tested this feature by running a streaming application which loads shared library. +i built myfaces from cvs dropped in liferay with basic navigation setting and it didnt work +Most other aspects are reasonably well tested +If its so great then why dont you write a whole new JPOX based Roller backend and well take a look at itand Hibernate isnt an apache project dude +Maybe we can discuss it on the mailing list i you have some issues? +Attached is a patch to that will get the pom working against that the jcr dependency still needs to be installed manually since Sun locks it down +Actually there is one more maybe preferable solution just revert this back to state before prev fix +on Dhruba seeing this +Tx +Chris note added in the docs for thanks for this is fixed I need to look at exclude which should use the filter system and be compatible. +I let the task open as Im gonna upgrade the documentation in svn +Atleast to the extent of agreeing to a design +Fancy a pull request? That would speed up the resolution of this issue and most likely ensure that the fix is available in +I am going to commit this one unless objection +Thanx Anita +Not perfect but much better now +Fix one logic of opening TFILE for writing +Think this is related to +Therefore I am quite sure that this is a duplicate of which you have resolved recently +Thats the way to guarantee we dont run into above issue +Attaching the new file +Part of a bulk update all resolutions changed to done please review history to original resolution type +The reason for it is because like every other part of we should be creating a plugin for firing the events +If we comment out that dependency we will get the proper +Jim Im assigning this to you given youre already assigned to the incorporated JBWS jira +The JDBC is likely the primary way users will want to import data +I can repro the failures without your patch +reopen for clarification the title says is updated with connection profile infoYou do mean its created with connection profile inforight +This patch generates a POM file based on existing dependenciesA potential wrinkle is the Hive jars +Ill apply both patches and see how it works but I probably wont get to it for a couple of days +Its in cloudearlyconfig on the system vm +So it should be force Bug has user guide in java doc +Maybe something likeWhere the legal values are all none or a list of integers of which reduce tasks to keep the inputs for +Rico you may want to review them anyway +Ill rebuild after setting my preferred cipher suites and check again +Ah I see +Linking the pull request changing the package in the above mentioned file made the WARNING disappear +OK Im going to open a new JIRA and clean the patch +However since there is some need to test this patch more thoroughly I guess we can make a branch and commit it there so that people can test this easily +Care to make a patch +This sounds really cool +We need documentation for this creating a new ticket for that +For reference heres a patch showing the test class that I intend to commit to the SIS project to test the class in +Problematic with pdfbox +I think it would be an improvement and cleaner to have this separated out in its own namespace +pushed +these protocols are used by both NN and JTso we need a way to define serverPrincipal key dynamically depending on where the call is made from +That way there will be something in the server log and not just the console +The counts from the threads even though they are not running any more should be included in stats count + +Any assembly might prove to be a problem but as for the rest Im happy to helpThe requested output is Linuxiaiaand I assume you have to define a new XPLATFORMWould you be best served with an account? You can me if thats the best approach +Hi MarthaAssigned this one to you +Im not against it but since its not directly related to this issue I propose a new JIRA for that change +Thanks! +Carrier shipment methods cannot be specified for purchase orders once created either +Hadoop was released today. +NI too think the problem is still there +just committed this +The file is never actually renamed in this patchI tried to start using this patch and it failed +Addressed by. +gah +Good Luck +This is the freemarker caching feature issue Musachy to wanted to make sure we are all on the same page +PR were merged as the component builds have been published with new features +I can go back to this approach but downside is still we have to use and files +Hi Tomseems that the lookup works again im able to select a locale string in the trunk when searching for contentThanks for reportingSascha +Grammar changes from Jochen and added tests by Guillaume so now we can close the issue. +Hi YipThanks for the quick work and the clear explanation and the good testsThe patch looks good to me I intend to run a few tests and commit itOne question why is changed? I see that it is justa different but what is actually going on there +The tests manually create a bunch of EJBs so Im not sure how well they translate to real world usage +This code doesnt look like a true Geronimo Plugin given the is not used during the build the filenames of +Ok Lars +bulk move of all unresolved issues from to +I suspect the this service is deployed in a HTTP support only container or the service is running in a soap stack which can not handle the chunked request +thanks! +m directory +We can implement a new feature source text preview formating but in a new issue +Resolving +But thats exactly when you do know what columns you haveIn my example when a new column is added to the file and inserted by the loader its hidden from view till someone explicitly adds it as a sparse column +Via logj preferably as we already bundle it +It looks like sysinfo does use at least in but doesnt need too really so perhaps Andrews solution of excluding is a good one but some other minor changes to sysinfo would also be needed +Im not sure however +Or is it ok to go ahead and fix it so that output always remains in the window in which it was executed +The regular process of building will not alloow for that to happen +after putting the ejb jar into warlib the war could be used in plan creator without problems +Closing this issue since there is no bug in Axis and there is also no suggestion for improvement. +Just took a spin on the patch and LGTM +Any new Location implementations will be handled as separate issues +I hope we will add some tests into these +Ah you have to also enter in the ID of the parent container that you want the child created on +I like this idea +This issue was solved by using the original distributed with +This patch should work correctly with anchors +I thought that one should be able to runa struts application just from the documentation +Fixed see SVN message associated with this bug +They have to keep it in for legacy purposes but we have no real commitment +Thanks Chandan for the clarification +done with the latest code +Many thanks great work. +I cant just ignore the rendering bug present in Firefox until this is fixed when an easy workaround that works for me is available +With Groovy how it goes? +Good point +Well include it in our release +Affected to to issues +I need a simplified test case illustrating the issue you have +Could you please validate in your unix env +That appears to have done the trick. +Oops +Where do things stand now +Thanks ChrisYour patch is in trunk at r +Are you interested to do all required testing or provide an unit test for this changeAlso when making patches it is enough to add an author tag to the class file when making significant changes instead adding name to the changed lines +Heres a patch +Committed this +HiCould you reproduce this bug on richfaces +There is no locking in stdstring no contention and grouping makes not trips in rwgetnumpunct +Thanks for your help! You may close this issue. +level stats are coming out good +At some point Id like to talk to you about removing that dependency because there should be a more generic way to trigger project updates that does not require +Could you please let me know +Updated patch for trunk +Hi Tom Ive not looked at the source Im just reporting the behaviour as it caused me to have a buggy rule +I guess theres no point of locking for reads though +So I suppose Harmonys behavior is more consistent same on both platforms +Ill commit a change for this to workRegards I will try to build and use latest version from SVN trunk +bq +Im moving them to so they can be checked that they no longer cause a problem there and fixed if they do +In fact the both behaviour should have been made for but due to a bugonly the one i indicated is have fixed IIRC it for to first search for provided URL to be a file and if not append to +The details of the compilation failure are not logged. +it is not always of this bug in combination with errors in activemq poms +No patch unscheduling from +Glad to have you finding these things +Hi TerenceWhat would be the differences between Municipality and CityI think you did not get my point anyway if you are unable to provide a patch I will do it for you this time +Any chance of getting this into a release ASAP? It pretty much makes using message bundles from with validation a +David can you take a look at this? If not please assign it back to me +sorry dont agree Bard +After the timeout could the client read again +However I dont believe that the is the root cause of the problem +Updated definition +Please ignore comment wrong Jira +That one doesnt matter in this context +Hi Jrgenthe location of the workspace directory is alrady configureable by changing the parameter in configuration +Some other data types may have a similar problem +integrated. +I noticed that hadoop are using checkstyle as part of their build to help enforce code quality +Sorry about that. +This problem was solved using absolute addressing with port number instead of relative addressing in input path +Running locally with the supplied test using the latest trunk code and I cannot get any of these exceptions +Have you tried this with +Patch is available +Theres really no further benefit to include the jars here again +Unassign issues assigned to Nate +Using separate trunks makes it easier for different developers to use different aggregator projects that only contain the modules they actually work on +If that is not going to happen please let me know by way of this jira +Cooking a RC that should include this fix +I also tried building the Maven examples and that worked fine. +I will now go ahead and delete the branch +But setting none really makes the bug practically vanishes. +bq +I called again +Can you elaborate on what you have in mind here? To pick the first two as examples I dont see what tab completion needs to deal with composite row keys or collections +the most likely explanation is a bug in the JVM check whether your uses a recent conformant JVM version or at least the same as theserver where your example works +increased probability of conflicts compared to the previous approach for example a session does foo foonullWhat would be the JCR API method calls needed that would cause a session to do that? after a conflict resolution becomes impossible from within the conflicting sessionIsnt the JCR method supposed to fail if there iswas a conflicting update? If not what conflicts is supposed to resolve automatically +updated with the branch +Bobby or Matt feel free to commit +Was there an issue is that had to change in? I personally loved the way I could get the servletContext in conversions since I can keep the formats I want in XML and use the framework to read the strings directly + +VERIFIED +I have looked inside the bin distributions for and +Thank you +Could you take a look and see +Please use CXF if needed +I gonna take a look on that artifact +Those can be removed again once this issue is fixed +Made the suggested change in revision +Itd be good if you can attach your testcase +So Im resolving this issue and copying Jean Noels latest example over to. +Which is itAnd do you have an example project I can try this on? I could not find it easily to reproduce and understand what it actually does +short log +Resolving as no longer valid +Dup of +Please save your work get the latest SVN make fresh changes and submit a minimal svn diff +IN PROGRESS +You are not really supposed to be hitting the main public M repo anyway so even when using the ibiblio resolver people need to customize at least the root to point to a mirror +Please open one or more if you like. +It looks like windows is mangling the unicode text files during application of the patch +Changed the package names for the xmlbeans and jibx generated stuff to not have them in the same package just with different case +It can also go back to View tab if change to + +The same issue occured when I tried to build the trunk +revised issue summarydescription to clarify full scope of issue assigning to has already started standardizing output time checking +There is already the and I would assume windows and linux would be next +Guillaume can you please confirm that it is working for you +Both should have stabilzed after so is a more realistic as out of date because of month no activity +HiHere is the patch to resolve this issueBut there may be some better way to resolve this issueRegardsAshish is the file that shows the screen error +see attached and previous comment +avoids collection modification while to and trunk thanks +Fixed the issue with the sorter and ORDER BY clauses in +Whats happening with this effort? Its been a while since the last commentreview is this ready to check in Cliff? +Shouldnt it check if length offset goes negative? Or is that checked elsewhere +This ticket covers the task of implementing a general framework for computing and persisting statistics +record a datastructure made of serverid rather than just id +Named get methods will work with clients and any compatible server versionI took at pass in general through our unsupported methods +Can you explain what the problem is and how to reproduce itWhy do you think its a problem with Continuum vs +Changed the title of the issue to reflect this +Closed after release. +So definitely Jacques will be able to take a call on this +This is really an important issue because currently the POSTLOAD lifecycle listeners are really broken +My preference would be the latter for conciseness and to avoid dumb logic errors +state changes on the trigger listener and causes it to ignore subsequent events +Im going to give your workaround a try if it solves that problem Im going to be a very happy and grateful person +Just tried and no luckStill seeing these errors Dynamic Web Module requires Java or newer +unrelated patch +Thanks Christian! +Fixed assertion failure for classes in default package at to make interface at to make interface resolved issues to so we can generate release notes +Closing this bug for now. +slfj is brought in bybuildsrcmainresourcesmodulesorgjbossloggingmain I see your pointLets ask David Lloyd if thats admissible +It changes the launcher so it doesnt detach the main thread before callingMore investigation is required to identify the root cause +Fixed in revision and merged to the branch in revision Theres not yet a fixed schedule for the release but I think we should do it soon as we already have quite a few fixes lined up for it. +please review and apply +The intent is fetch as many rows as specified in the filter or if the filter is instead of rows fetch as many rows as we need on average to satisfy maxColumns +So should this be any different? I think if this should call create then the other methods should as well +is not a supported platform +I tend to to svn status a lot and I noticed lots of unversioned files at the root which I didnt recognizeIf I untar the file with the command line the directory looks likeAIR SDK AIR SDK binframeworksincludelibruntimessamplestemplatesand if I untar it with the ant command it looks likeAIR SDK AIR SDK AndroidLicensingVersionsairbinframeworksincludelibmacprojectsruntimessamplessrctemplates +The preferred way to fix this would be to add documentation to trunk that is updated as the code is patched and replace this wiki content with a link to the official documentation on the website +The stax dependency is needed to exclude xmlbeans +Can not reproduce this anymore +As of Spring transactions will even work with with plain access as long as the instance used by the is set up by +Committed revision +Attaching a cleanup patch v that replaces patch v +Ported the fix to master as well +When I switch to another cache than CACHEAUTO or CACHE CONSUMER I get a dramatic deterioration of performance +So this bug is a major problem for me by now +fixed in r french translations corrected in r +Would it be possible to add a unit test forAlso Im wondering +Leaving Open. +Standard thing to do believe it or not is to just use unix line endings That is what AS uses and also what we do now in Fuse . +Patch file for implementation code and testcase +Integrated to and TRUNKThanks for the patch LarsThanks for the review Michael +Agreed +Actually I do not think that needs to be executable +This patch adds JNIEXPORT declaration for JNI patch has been applied. +Once the are done this can be resolved +Looks an override for the set was missed allowing the Cas editor to change the sofaNum featureIll edit this issue to address the problem in and Sofa can you close this issue? +Have you tried just adding in the startup option to your connection URI? This option will limit the number of time the Failover Transport will attempt to reconnect on startup and allow the client to fail after that number of tries +I got this in actual application code +Ive committed with revision +The same happens when I click on The same happens to be with there any chance we could use the Find Screen Decorator? It would make the UI more consistent +It looks good to me Werner +Im not able to get SSL logging on +See also target gcc release for this fix is which is not yet releasedSo this will have to wait until gcc is in use +The was still deployed but no hashCode in the name +If after posting on the user list you still think this is an struts problem feel free to. +I left Alex a voice mail yesterday and havent heard back yet +I figured out what was going on +This is more or less by design + there was no PASV sentI will attach a test which tests all affected commands. +mvn version +I ran a success full set of tests against this patch against windows XP and ibm +thanks to sean schofield for removing legacy attributes of +Sorry I should have raised this in forums not JIRA +Perhaps that might be integrated more easily +Please see my other comments on phabricator +great! I will give it a try +Im confident your testsuite would pick up any problems +where is the parent pom located in cvs +Fixed in Committed revision Thanks Shivantha +Patch to add broker Maven at this ages ago forgot to click the button. +Ignore the previous attachmentsThere is no need to generate the to see the failfure. +pressing p makes the panholding the middle button of the mouse alsoDont need a new shortcut for this tools. +The shared packages are generated out of the common myfaces shared stuffSo this needs to be fixed in myfaces if this can be fixed at allIll move this issue +See attachment +Thanks JonathanCan you update the wiki with the new syntax +were backported to as well adding the latter fix version. +HiSebastianDo you have a plan to do the change? +I tested it in ACF +Are you planning to drop cglib support in Hibernate in a future versionI dont know much about Javassist and cglib but all I can say is that SpringHibernate is a very common combination in enterprise apps today +Verified +doc reader could peek into the file and return either textplain or application +nice catch RECUDYNAMIC or even is perfect + and hope I get lucky with Hive QA this time +need to install the python modules in thriftlibpy and thriftcontribfbpy +Ive just hit this issue as well +As a compromise Im certainly favorable towards the idea of making Nano require and not merging the interface +When I am using running eclipse as target platform I can start my test plugin from dev environment when I am using our TP I cannot +patch looks good to to snapshots branch +I will commit this patch in a day or so unless I hear otherwise +I have the same problem had nodes with simular names in my workflow like +I have reviewed the patch and I approve +This error was introduced with the resolution of +And assembly kit should include the slfj instead of migrated in rev +I dont think that there is a problem in general with the userids in the uploaded files +Patch fixing the appender threshold problem +This patch deletes any presexisting if the job has not been fully localized and tries localizing the job again +Please make sure this issue gets priority +Sounds good +As for the solution it would be great if you could jump on this +I forgot to update the test when I removed the from the formula +So here for safekeeping and not reviewCertainly agree with removing the old way of creating handlers and annotation is implemented as proposed + +Im reopening because there are too many unnecessary changes with the introduction of xmbean descriptors that break the installer and just change existing descriptors +OK now it is committed +Patch applied at +GarnettIm using Java on Mac Os X and is successful on my machine +Its good to see that Im in line with Seam developers line of thinking +Further testing this morning on real world project showed me fix did not correct all cases my real world case is still broken although attached test case was fixed by itDoes it mean you confirm this bug is a real one? Test project attached reproduces it and my real world also doesIs there some plan to fix it? within which delayIt seems anyway that the stepexecution is modified somewhere before restoring which generates the case +Lowering the priority since we are not that far off and we could not find and obviouseasy ways to fix no longer have code running anywhere so investigating this is not possible +Thanks Stefan +and These will have different wicket sessions since the session cookie uses the domain name for separationYou can check the domain name in your app using the current request information +hmm what is the error +Also an important test that we should add is to make sure that leaving a shared doesnt leave it in a incosistent state especially if other components share the channelIsnt that integration testsWe are creating new using Multiplexor +Committed to trunk +That makes sense +The thread dump of the described scenario can be seen at the file attached to this issue. +no changes I pretend things work the way they should +Just a note that the same issue appears if the with value is removed +The JIRA I think remains unresolved for some of the incomplete minor subtasks +But the implementation is trivial if you know how to construct a specific base URL +This issue is closed now +Doesnt occur on staging only on production +Instead we are focused on reusing existing which are simply based on +Yep +This sounds rather +also contains some code cleanup +If the registered company is not approved by the backend user it should be earmarked as prospect not as customer +Works like a charm of course +OTOH if you are interested in saving user history to HDFS then there is no permanent fix yet +The ANDOR short circuiting is an issue for both SELECT and WHERE +Resolving as Fixed. +Hi Alex I have new code I will commit to Github +Starting local mode leverage hadoop local mode there is no much difference +consider for the first part of creating the MREPO path s something to be decided right now the parameter to specify the idea settings home is left as workspace to keep a standard with the eclipse plugin but probably a better name for the parameter is needed althought IDEAHOME I dont think would be good since IDEAHOME would be something like the directory where idea was installed not where the personal settings are +Marking this as resolved successfully tested against when testing out and. +The dictionaries are used in many places and it is really missingsome kind of name fieldThere is no reason to pass in multiple dictionaries because multiple instances of the Dictionary Name Finder canbe combined in the +Im changing the category from SAX to miscellaneous because I observe that the same problem happens with as well +I think resolving to use current schema as current path is a good way forward +It gave the name of a jar file +Forgot to add both schedulers also have a thread or a thread pool for initializing jobs but those should be easy to mock out +Users can already import a public key using eximportkeypair method. +Can we close this JIRA now? Now that is in trunk +ah ok got it +Ok +Forgot to close the issuethanksdims +User logs in and is redirected to the original URL secure +I think I have the changes for this but there needs to be some provision for still scoping a mat view at the importing level as it could reference things like system tables that can have vdb dependent data +there is a nocommit on Fieldable since it doesnt have javadoc but lets iterate first to see if we wanna go that path it seems close +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +address Carls comments +with grails is OK +Michael To answer your Its the snapshot of the package with the issue +Current workaround to get Build Agents installation is by directly using the Build Agent Web Service +Hi Ivan +Once reviewed I will post the failures were time out issues verified locally on trunk and my branch that tests pass +As Robert points out this is an issue ofusing a single static collection of converters not a thread safety problem +The main contribution seems to come from diffing the property states +The rule parameter name is enabled so there shouldnt be a collision +I would think that selecting the connection manager or MCF name would avoid ambiguitywith this approach would we be able to work with databases +I committed this with some additional cleanups +Make s an array keyed by mapreduce jobid +Static method calls were not being handled properly in the inferencing engine. +Oh I agree there I think this patch still makes perfect sense its brings lot of gains +patch file of the code that I changed in by delegating to Component to test for the annotation +The notion of what role names are applicable would be entirely up to the providerRamesh in keeping with the additive nature of our role permissions you would apply the permissions additively as well +Correct for input done for input +can you try a JDK? jBPM is not yet tested with JDK as out of date because of month no activity +Included the improvements after discussing various issues in the meeting +Patch implementing proposed change to the Extend new Remove all in Rev +Appliedsvn revision Thank you for reporting this issue. +I added the empty getUrls two weeks ago so far so good I did not hear any screams yet. +Patch according to comments mentioned above +A feature that sorts out literals only is not useful outside of the unit tests +The goal here is that things work in the proper way +Also upload a documentation picks the last file uploaded +Thanks Andreas I do appreciate your consideration +Did you change something else besides the format like the bbox size of the image or removed tiledtrue? transparency wise is again normal you did not specify transparenttrue in the request +The behavior I would expect here is that leading and trailing spaces would be stripped and it would not allow you to add what is now a duplicate local repo +a threadsafe version of the this issue be closed +sln and +Fixed the issue and commited the changes to the svn +File status types from a distributed file system +Hi JayI think its fine +attaching new patch +Please let me know how i can solve this issue asap +Same could go for the toolbar and the +The breakdown I have in mind are roughly alongthe lines of the subsections of section Implementation details Expose any interfaces needed by a subclass +If this is for EAP then for sure +sure but to find out how something is working you have to torture it a little bit +BTW Id like to move it to HTTPCLIENT project and update the affected versions field on my own but I have no permissions to do it +If I have understood this jira correctly then attached patch should give more information on threads and dump the java core if we encounter timeout during network server start +This is a serious regression if extension is set to in some config +I see some good ideas for a further issue but lets commit this patch and spin off a new issue +So if I understand this right this fixes a generic deficiency in which wasnt showing up because by default files are public to read on Linux FS and so would not see issues in accessing private distributed cache from the local FSCorrect this is how I see the problembq +The patch is based on as found on so their licensing applies +However at this time it isnot used to avoid licensing issues +Revision Committed by romanstumm at applied patch from Jrg Wamer to fix Security holes in +That would still leave a small time window where the state stored in the would be wrong before read just before acquiring the lockI agree its small but I dont see the point of leaving it open +Group title is now a short label and the instruction is put inside the group above the buttons as suggested +Patch committed +Committed rev faf +Meanwhile would it be possible for you to refer to your schema in some other way +Updated patch marking as +Verified by Alexei. +Please verify that it was applied as expected +Does that mean that it works if you use a version from SVN but not with the published snapshot? The above snapshot should indeed include this fix if not I have just republished it can you try again +Applied in trunk r and r +This would prevent disk exhaustion even if debuginfo are not disabled +move reporter from Steve to defer of feature items from to +Tests now pass at the same level as ve moved the splitting out of into a separate subtask to notify you guys is in patch submitted state +This appears to be fixed +You welcome +I have committed this +This is the newer and possibly final version of the patch for the upgradeI would like to commit this work later today if there arent objectionsIn the meantime Im going to continue my tests +The user can do a explain plan and see the plan +Both are converters that stay registered after undeploying. +Please change to if you intend to submit a fix for. +This patch makes max xceiver into a setting so that users with large clusters can adjust this knob themselves +And when get the log aggregation status from the npde it will forward the log status to related +Why do you think its not working if Im seeing the alert? there is one call to stopRecord and the alert after it does show +Patch applied in trunk r with some minor changes +The proxy is back up and working as expected. +We also need to stop importing those columns because otherwise we would end up with different number of columns in imported file and in hive which might lead to a huge confusion on user side +I went ahead and moved the providers into the module +If not would there be any more work to do for this bug +no planned release on +I mean manually reload the resolvconf +Ok I didnt find that issue yeasterday +Ill commit this tomorrow unless there are any objections +Here is a little problemwhile embedded engine starts silently at users request user should shutdown engine explicitlyWhile first part conforms zero administration claim the second part is in direct contradiction of the claimTypical scenario is application uses a connection pool it may connect to a production database or local testing database +Latest version of the patch with documentation changes and fixes for some intermittent junit test errors +I am currently trying to get it to work for multiple languages in a test F app +Skip parameter implemented in r +Pushed back to + feels reasonable to me +I feel that any BRMS documentation referring to this should be updated in order to talk about using the Process wizard to create a Rule Flow +comments addressed +IlyaIts reproducible for the. +Attaching a patch +I think is the latest one in trunk because the bug is also in the examples every day are recompiled automatically +My understanding is that we observe elsewhereYes we dobq +Thanks for the updated patch JunjieI committed to trunk with revision As the reporter of the issue feel free to close it when you have verified the fixcommit +Attached patch that forces type conversion to the current property type on the property setValue methods +i am not convinced +Im really sorry +nothing done +changes need to be revisited reopening issue +Max Just reopening this to understand if the title is correct +The patch was tested without any regressions +What does your endpoint look like and how is it configured +Is any additional info required for this issue +We need to figure out a way to include the message of Hibernates wrapper exception too which might sometimes add meaning to the underlying ve adapted and to integrate the original message into the Spring point regarding class loading +I am beginning to understand it +Transitioning to waiting for ve added in redirects for the remainder that need to be. +from me + +patch looks good +Should this be closed as wont fix due to the previous comment? +Ill revert this patch and will provide the one that makes sense +Talked to Craig and agrees to close this bug. +watchedDirtmpfoo bar where tmpfoo bar contains foo +I think it makes sense to just throw a normal when an attempt is made to make modifications to an RO snapshot much as an attempt to modify an RO snapshot in WAFL or ZFS just results in EACCESSThoughts +Backported to together with with revision Resolving issue +It hardly seems fair to place the responsibility on the creator of the tree not to release the namespaces underneath the detached node when the tree is created as the SOAP message is parsed and the location of the xml namespace declaration in the message is determined by the other side of the exchange +Looks good to me +with the native gem installed the Ruby times on benchmark drop to more like per iteration +use of jboss tools in eclipse rather impressive Last time I tried to build against Eclipse WTP I got a looong list of compilations errors +As discussed I reverted and added eqne to Restrictions and Property. +I have attached a file with the maven log +The machines CPU and memory are quite normal. +See the linked issue for the issue that will address this +I am looking forward for the documentation to be fixed +Your ldap creds look correct what is the full path to the repo are you trying to commit to +Attached is a patch that can be used to cleanup the original behavior if it is determined that it is correct. +is removed because we have is replaced by and is removed because we have Ates solution with geronimo javamail pom looks better. +Thanks for the patch +The current way of doing file checksum allows calculating individual block checksums in parallel but we are not taking advantage of it in anyway +x to the additional compiler options in the X preference pageWont this do what you want? Or is there something here I dont understandIt seems unusual to me to have xdt have an option that the invocation doesnt have +happened to this issue? I am running Maven and got this error when trying to deploy my site using HTTPWagon protocol http doesnt support directory thats unrelated +Patch applied by Anita +What platforms are all you guys experiencing this problem on? Any specific one or a variety of ones +Shall we make a patch to skip Suns and the IBM altogether +Try this project and use this command to duplicatemvn so this is the older versions of cargo the dependency classifier was ignored which meant that if you wanted to include a dependency that had a classifier you were out of luck +OK fixed +Problem with this solution is when and who should free parent node. +No longer necessary inside Eclipse since were able to launch Equinox Launcher. +Ok yes seems I misssed the in the build +a thrift server actually runs the queries via qlDriver +Akin to Windows development or a JIRA plugin +My solution is that checking the region plan before acting on time out for the region in the case of this with minor reformattingGoing to integrate tomorrow if there is no objection +Looks good +Yes its fine to port get commands from binmapred to binyarn as they make sense +Hi Yonikall that this bug is asking for is to change the default behavior +Applied thanks. +That is also how data source level exceptions are handledThe problem with this patch is that it aborts on any which means that a connect or read timeout can also abort indexing +The updates have been verified in the ER build +No need to have them separately +Anyway welcome in the JAMES community and happy XMAS +Marat you are absolutely correct +my meaning is FOP cant support Bold and Italic pdf format output + +The underlying was closed so I think this could be closed as well +When I ran wikipedia with the patch i got results that matched just india +Daryn this is important change for making progress on this jira +its more of a software design problem rather than a performance issue +David Wheeler wrote The CCLA is not just a copyright license +Thanks Arun! +Thanks everyone on this we deployed the latest patch with the latest Solr build and things work now reliably +Are you using JSP and Clay full HTML or XML viewsThe components that are created when the view is restored are verbatim components which are not serialized to the clients state +They implement an optimization where buffers may be read from the socket and passed to the application without copying +You can flush the table first and then major compact from the shell but I think you want the two linked +Thanks Adrian +Ashish sorry for delay replying and thanks for testing +I wrote a test to use the control insted of using for the test +For our more significant changes a fork is probably better so our progress is not dictated by dbo +Progress +Ive opened a github pull request for this +So which default configuration did you use? Does the class that I attached fail in the same way? Or did you use one of the +The suggestions by david for handling the experimental checkin sounds reasonable to me +This can be done for but is not a target for releasing +These tests work well now +Thanks Robin for providing a patchBefore it can be applied I would like to see it covered by +What should the official bundle look like? Does it look different from the usual tarzip distribution? Any pointers would be appreciated +It would be nice to strategically reconsider what goes where and lay out the expectations for what can go into each level of the server assembly. +The template used to reproduce bug +Moreover as we discussed navigator View might be completely removed in future +This mainly depends on the value the community gives to the proprietary features of the core and xstream partsIn my opinion the core frameworks offers some very flexible possiblities to store merge and transport metadata of classes that we used for providing type and information via json to an AJAX app +Wed run out ofdisk space in about a week and the logistics would be a pain to have hundreds ofcastor branches +I ran derbyall against a clean build and the following tests failedderbyallunitderbyallinderbyallinderbyallinderbyallderbynetclientmatsderbynetderbyallderbynetmatsderbynetAre these tests known issues in? I ran derbyall again with the patch and got the same result so the patch didnt cause these failures +Added change log entries +What we see here is not consistent +verified the patch by deploying hadoop nodes out of different build timestamps +There was another issue open regarding returning back from subflow to mainflow which is resovled already. +If you finish it before we release beta please correct the version +It changes from one release to another +Thanks BliginReverted in rev +The missing methodsfields in Interfaces was caused by special logic to distinguish between how Java and Groovy treat components without an explicit visibility modifier +Is this a xjc generated object? Does its comply with the settingI think you are digging on some new issues here +if the cache were automatically updated then undeploy on the last deployment would not work +I would propose that we split this ticket at this point +The dynamic serde itself support is actually almost done as part of the jira to add the unit tests for this +In typical runs where this test succeeds the there are events in the first file +Unassigning myself from these issues as I dont see myself doing them in the medium term future +will cause the specified s recommended version to be used if the original doesnt have a recommended version +Issue may already be resolved may be invalid or may never be fixed +Whats the reason to cache performance improvement +Itd pass the edits to the first then one by one call current batchUpdate only with write set to false + Patch looks good for me +closing. +I may not get another chance to work on this one for a few days so I wanted to attach the latest version of the patch +what happens when an automatically registered node disconnects and never comes back? will you keep it registered foreverwere currently doing that and have a tool to selectively cleanup instances that we know have been permanently decomissioned from the clusterregarding you idea for you need to be careful about race conditions and see whats the proper way to handle them +Colinno problem at ll fix thisSergio B. +committed +Incorporated changes from conf call feedback on Oct +Raul maybe a discussion on the mailing list is preferable before opening a JIRA next time +This would likely require changes to the Job interface which I was a bit reluctant to do and also make store a reference to +Not an impossible problem but nastyI switched approaches and created a columnFamily date and now access all data via columnKeys +Replaced code tag on the branch with revision Closing issue. +I have captured some of it in a new section of the user guide +Hiros answer to your question is correct you need the gem in your Gemfile if youre using +So with temporary queues it can pull every sec and pickup messages as fast as possible as the received message is always for the reply manager +Committed revision Ill look into the fix to before I resolve the issue +Before we create the closing node if creating fails because of node already exists check if it is REGIONSPLIT node if so wait for some specified time for split to get completedIf not can we trigger manually to process the split? Pls suggest +This seems to be happening again but this time its also affecting retrieval of single columns +Or does the database realize that we already have locked this thing and the statement is a lockingwise? Also is there a portable way to check for a existing lock without being blocked in that case +When I look at the of a Java project I think it beneficial to see total code coverage on the entire project +Cheyenne Thanks for Forth Bridge has the thumbs up from everyone +In other words I think the compiler is doing the right thing wrt my understanding of the current language specification +Doing this to an embedded agent will cause all kinds of problems +Yes as far as I know +Changed config var names semantics of reduce percentage and updated documentation tests to reflect with trunk +should also setup jbossws as the has been done by Grid Web Tool provides a preference to set up default WS runtime and default Server so we just need to set the default value when our runtime plugin is if we also need to set jboss server as the default when create web service +It is also good becuase it allows us to maintain and test backend independence of Pig +So we dont know what mailing lists they have +Fixed in latest that uses bndlib you have an ETA on when will be released? Obviously I dont expect a firm timeline but Id like to know if its going to be next week versus in six months +Mike Ive done a bit of work in the last week with Googles webmaster tools and its helpedThere are outstanding issues +if this is still broken on trunkRC reopen. +this could easily be an integration peice +Part of a bulk update all resolutions changed to done please review history to original resolution type +warnings when compiling source is deprecated and the import of is never used in just to let you Jonathan you are completely does anybody know why this issue has been unscheduled +I did the mavenization of JAXR project so Ill take care of this fix +Attaching a that was build from Branch +Well consider it in future visual instance will require coordinate information to be generated against various states reported by the BPEL engine +Ill see what I can do on short notice +We should add a description attribute to the interface and elements so the folks creating configs can describe these things to the admins looking at consoles +It is actually quite handy to have this information when it is available +Based on the description yes +Good idea! +Updated to reflect that the issue exists whether or not passivation is used +Those Instances are now reached by the garbage collector +At least Maven works for standard layouts out of the issue gets my as well +In the meantime if you have a chance to test it let me know how it goes +From Chapter section Common Monitor Notification Types Sent when the object instance of the observed attribute value is null or not of the appropriate type for the given monitor +If it is a debug statement will be outputted but no more action is done +Im not sure if this is still the case or not but I havent seen any users hit this issue nor have I hit it in many months. +I agree that it would be interesting to evaluate this in the context of which is a very nice idea +Youre sure youre on r +Comments +This patch fixes my problem by assigning a fixed field for the entity ID +awesome! committedwhat about testing just one method is that harder +We agreed this is not a supported case locks are needed to avoid data inconsistencies and should not be skipped issue is no longer an issue when using the Lucene Directory as it no longer attempts to write twice on the same key sipping the lock. +At the very least we need a sample project with a POM that we can use to test your problem +I completed basic software engineering course of Java programming last year in Accenture +Run as Local using JMX Explorer +If you are really going to ignore my patchI did not ignore your patch +This issue was Inspired by comments in. + checked in. +Not a bug according to Dhruba +Streaming is of course goodIm wondering what version of MS office can open the XML based documents and which ones would be cut off +I guess we should add it at least for NNsomething like thisStarted Mon Aug PDT Version rbfdbbecffbbcafccbCompiled Mon Aug PDT by boryaUpgrades There are no upgrades in progress +fixes the date time zone bug and change the order of statements to sync with the result order +Have created a new ticket with the rename of the expression to language +But I only just found the time to update this work to the new +I guess this would mainly be a problem for snapshots +Every framework will benefit from that +The critical issue is that if you make a fix to the framework you need to know that all of the task trackers have the fix +Did some testing on ZK my assumption that the ZKs data version is incremental post node deletion is incorrect +thanks for the review JukkaI agree with your point I will push the code as is to have the mechanism in place then follow up with another issue regarding optimization of the s diff mechanismThe optim could also include the following ideas wont really have to extend anymore this way they could be made to work with diffs directly +for background workTest Failed +Marking as Wont Fix +This is now fixed on the broker and on linux clients +Yes I will fix it tomorrow +bq +Great +please review this +Case select is Joes submission +Does anyone have any opinion on this +Complete isolation delete files and load again yes my scripts do cleanup the page caches +Yep I made a small error when fixing the latch yesterday +I think this is a Tika CLI issue rather than a Parser one +Would it be sufficient in these cases to separate the logging of the error message as a warning and to log the stacktrace as a detail +Proposal Source +heres the patch I am currently test passes on a clean checkout +Hi AntonioThanks for your reply +patch can be committed once the findbugs warning is addressed +Fixed +James Can you take this bug please +It seems more natural to send a stop replica request to a failed broker +Here are the BPEL processes +Would it be ok to have this on the branch as well +Hi Lohit are you saying that we should keep in? Without is quite empty and may not deserve to have a class for that +This is discussed in CLISending srctestorgapachecommonsclivalidationTransmitting file dataCommitted revision +It is just the improved browse dialog +for the updated patch +Maybe just bug fixes which are easy to port +So it seems more natural to make this a new component +to fix null pointer in test is the consolidated patch do I get the unit tests to run? I have tried both maven and maven and cannot seem to get anything to run for weblogic +Implementing this feature is not straightforward as all the links to drilldown services must be deleted when displaying the dashboard of a past snapshot +Marking closed. +DaveI have the same situation here and Im doing some tests +My was for including it in Having no logs is annoying but zkcli not working on windows is more serious and should also be mentioned in the change log +Ill attach the generated WSDL document from which I have copied the WSDL message snippet above in a minute +Edward JQ If I understand correctly the suggestion isCreate a new setting in the Set the default value to the existing values in templateCustomer can override this setting if they want to work with a different storage systemI personally dont see so many changes to this but if you strongly feel that it has to be configurable then please let me know +Please be advised though that a patch for this problem will not make it into the release anymore +Im not entirey sure what the right answer to this is +I havent heard from anyone on this +Also all tests pass but I have the feeling there should be a and a Locale in the instantiation of objects. +There will probably be more things in there than a servlet +Attached patch for it would be good to have a testcase using streaming +It should be easy to see where we are moving to +We can probably address this for M I wrote a quick patch for which handles the situation there +Do you happen to remember if there was any resolution here? +It will mean that we have to bundle the whole JSTL JAR just to use one interface + +Sorry +After this patch the bug is present +Thanks +I wasnt able to run the due to a missing dependency but I have reviewed the code and it does seem to show the correct dialect +Willing to donate organs on this one +output and output also have problems with cdata sections +Check with ethereal whether the flag is really set in the IP tos field in UDP +Closed after release. +CMAKEINSTALLPREFIX is always defined +When I had commented the svn related info wasnt there +In the original proof of concept for I measured the time for extracting for +Wanna commit Ted +I will commit it shortly unless I hear otherwise +Here is the original bug report going to leave both open as they have different approaches of producing a quick bounds +I have the +Yep here is that same changeset patched against trunkfilename which the patch other file missed by the patch patch posted by Zach Bailey +Cleanup of unclosed issues. +I did not follow up the history discussion here for this jira can anyone provide a reproduceble steps here to reproduce this issue? A sample application could help a lot on determining this issue quicklyThanks! +Thanks Pranay Arun Done at rAshish Vijaywargiya +issue reproduced again +Suneel weird +Would you like the contrib portion to be considered as part of this ticket or separtely +So please waitThanks a temporary workaround I managed to deploy the sample application in by using the PlatformStatic option which starts without +If you look at the gem I referenced it merely adds nonblocking methods toSSL + +Please chancel this issue +Attaching the v patch which resolve some of the tests errorsPlease reviewThanksI am attaching a possible fix for the exception that is being thrownThe fix contains changing the order for the tests that are being executed in suites +It has been commited for the release +Here it is +bulk defer open issues to defer to not essential to address for +I ran Ivyde with remote debugging and found that the problem is caused by POM inheritance in combination with Ivyde workspace resolving we use +Close all resolved issues for older releases. +Thanks a lotBest regards by Richard. +In my sample bean was recreated and pre method was called after ajax request +Looks good to me +It would be good to get it in but it is not a blocker for Edward I modified your patch to make it workCan you take a look and let me know any questions? It will be great to add enough comments in the code to make it clear to other developers +Ive seen several more cases where this has happened +I am not sure there is a hook for what you need but Hooks might better place to plug in deployment specific requirements +Thanks for the review ATM +Attaching screenshot with extended data overflow in m confused by the Firefox Linux i you listed in your environment +Seems ok +A simple port of the patches to the branch did not work +The situation is better in with compression working well +GC v if the only use case for this patch maybe it doesnt make sense to apply it? What is the reason for doing this if this functionality wont be used anyway +Looks like a reasonable repackaging to me +Fixed in r +If you have for example a packaging script that runs something like ant buildsource buildjars this doesnt currently require but now it will start trying to download automatically and will fail if its running in an environment without connectivity +So the th attempt gets killed will not fail the job +This way folks who are currently using this feature in the incorrect way are aware that there would be +I have a solution for this problem +Thanks for reviewing Flavio. +Is this still a problem? +Refreshing the page in the browser results in open file handles for the complete set of stylesheets and images referenced by the page +I was not able to reproduce this could you provide a test case or additional information? Did you modify the plan and change the contents of the application prior to redeploy +Should be pretty easy to integrate your stuff! +I was going to test this on centos I notice that there are additional filesdo we use all of these or should I just replace the ones in the catalinalib +After I created it I manually edited the CHANGES section which messed up the lines count +I am resolving this jira as Duplicate. +We see the same problem with a contributor trying to build the Validator online manual +Let me take another look at this patchset first on my plane this afternoon +Sharad Bikas marked as a duplicated task +Nice analysisI like better +Also the latest code in Git does improve performance considerably and I would like to see if you can notice this in your project +Ok +Some other bugs were fixed for tracing between and and caused this one to be fixedIve verified that the text is not lost for +Ok Lars +Hi Kihwal is that possible to commit it into also +Doing a comparison of the snapshot logs reveals some interesting information in terms of what methods are adding the additional time during hot deploy +older issues for Apache since were no longer actively working on these at the moment. +Andrea further confirms this isnt a bug marking as Invalid unless another test can show validity. +It reverts the change from Reader to and adds a dummy Tokenizer in the test that compiles but does not run in trunk +Scheduled for +Verified etchadoop in the tarball now contains +Canceling the patch +Probably you used old project to validate this and faced the opposite error +We use cvs with tagging to deploy our projects to internal testing and m attaching my version of where Ive changed toe clean target +has the former exadel set of plugins now +Fixed. +The patch has been applied +Some magical interactions between NVP plugin and NVP Controller I cant see +applied. +heard back so closing as works as designed +the possible reason rebuild project is necessary Jiri could you check is it happens if you run rebuild before configuration expansionif it so the possible resolution build selected project for console configuration automatically +Bug has been marked as a duplicate of this bug +We can fix that later. +DB works +Ive attached the patch that removes most of the lazy copy codes +but then i didnt +lets move forward with that in this jira +Pushing issues to the next CR +excludeUnlisted does not mean dont scan the jar I think this is what I got wrong +Plan is to make it the target product so EAP WFK JDG of them should be EAP except the ones Karel identified +on defaulting to overwrite at least if the assembly is building a jar or a war +Theres been no additional progress on the ASEAP BOM side so were closing it as WONTFIX. +grep nfs +The class would need to update its method signatures as well +may be Databese tree item should also display diagram of the tables in the DB +It means that of all persisted states were optimized before the save +Can you post a bit of java code or a standalone java program that demonstrates the issue +It affects the build output so it should be configured in the pom +Closing old resolved issues +bq +If we switch to another library we can change the sandbox control which currently depends on Scriptaculous +This is the right one +Hey FilippoWould it be possible to create a single patch of all your changes and also including the documentation update? I would really like to apply the changes but the amount of patches makes it quite risky for me doing some errors during the applyThank you very much and kind regards AndreasI have made cleaning about the patch that i sentThe patch does not include documentation +Are you running on a local file systemCan you give a couple of examples of your row keysThanks +I thought it was better to get this one in given the size. +Just finished adding a type cache toIts always interesting to see the law of unintended consequences +Committed to trunk and +We do have a warning saying this should not happen per nkeywals request +I already removed and +Looks good + +Bug fixed +Sample project that reproduces the problem +David this is why I am assigning it to you +Attached zip for severity +enableTable isnt synchronous so if I call addmodifydeleteColumn immediately after calling enableTable I get an exception because the table isnt disabled +here is a patch for int variant +Assigning to due to lack of Cache this issue should be fixed with the beta build which includes jboss cache verify the issue and close it. +add a goal to maven site or add a goal to run before maven test without maven site or test knowing about it beforehand +After all the INFO level log should not get spammed with repeated messages for such acceptable conditions either +The reason this is solely for walogs is that doing encoding in the middle of an write is it could throw off the sortedness of the kill relative key encoding and doing individual key encryption just isnt performant +get must be synchronized too as it has a reflection cache and is also only used during instantiation +Much better than complexity I implemented +Thanks for tackling this +Hi DavidI think its OK to share the between different web apps +Part of a bulk update all resolutions changed to done please review history to original resolution type +Yes I agree that clearly supports your viewpoint and I will fix my implementation but could you please comment the rule that I used before should we suggest that the CDI team make it correct or is it correct as is and it was my fault to misunderstand it +You missed one in Content +That is something that can be performed by wrapping the thrift deserializer +Well +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Going back to using a far weaker string naming convention to wire things together is a step backwards sliding us back towards the same problems as xml has in the first place + +Committed to branch and trunk for. +I was not able to build those solutions in the contrib directory either +It works fine with cvs so i think its a pb in +For now Id like to make sure the heartbeat is done properly +test for issue +So it looks like an WTP bug! full JEE spec did you actually do with the booking example? did you update the corresponding files? What happens +gtully can we close this one now? it looks complete given the SVN commits +Pls confirm whether the CLOSEWAIT issue is resolved fully onwards +Attachment has been added with description test font the problem appears to be that the windgding and symbol fonts always report that they cant display chars +Mavenized version of the test case that can swap trinidad versions via the pom +will use all columns as the idprobably not the optimal solution for every situation but it at least will not fail on it. +Rough patch with a monitor daemon +Instead of adding a new slaves file we could instead add a flag to getConf to get the shared edits then parse that URI to get the list of hosts +The png is attached +However when the xml document does not constrain the xsanyType element JAXB unmarshals the unconstrained content to an element node instance of a supported DOM API +Added more description of the path expression and the formatting rules +Are the results comparable in both casesThank you +the uses the deprecated get usisng a priviledged block to access the system graph I dont think this should be done here as the result is no of you if the caller has no access to the system graph. +Attaching the diff should anyone has anything to say about it +You have http chunking enabled so wont be send +PFA the xsd and the wsdl used for code generationI am using issue apepars for me +Updating to GA change the fix for if you finish this sooner. +Thanks +Fixed on the branch as well as on the trunk by applying the patch from Amila thanks for the contribution Amila +Ill look into it ASAP +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +I hadnt yet done anything on this issue +Bulk close after release +Thanks for looking +Any objection to submitting +class file +Here is the patch I would like to commit for this issue +Any news on this? Suite would also be my first choice +FYI The same query works in Beta +the documentation change is in so it may not be worth adding and removing this special case in +Committed to trunk +Since off rack writes are randomly assigned these should be relatively evenly distributed +Maybe thats nagging though Im happy this works and I think were done with this issue +I have committed this +Also has a test +passed all unit tests +Also instead of Excluding private I use the +Hi Andera I tested the quick start and dont see a problem +I think columns are a good suggestion though +Also committed the patch +I guess I am done with my review +Please comment if you think that the code may miss notifying the app about completion status of a container on which the app ran a task +This regression was caused by revision checked in as part of +Fix attached +This may have been added later but Ill fix it +As was primarily for another issue and still needs review you have my to commit this to unblock the tests +Attaching new patch incorporating Hemanths comments and Viveks comments Changed initalizedJobs to to map of job id to job in progress +That approach would also make our detection more robust against manually copying data files into a cluster with a different partitioner +That test will also reproduce the issue if the fix is removed +failures are unrelated +This should be fixed in the SNAPSHOT +For the next release we need to do something to indicate that those fields are read only visually +Thanks Paulex +This patch add the two necessary imports to iPOJO bundle +Verified by Vamsavardhana Reddy +Same results with a similar behavior in high loads using apollomqletting DMLC manage its own JMS transaction in high loads a broker gets an ack for a msg but the container doesnt dispatch it to its listener +Is that the idea +Indeed it should not allow the to be DD while editing container not allow user drag and drog while editing container layout + +Start and Stop on the context menu of a module now behaves the same way start and stop work in the web UI for application management +end of commit comment for revision +It tries to solve a very similar issue I think! +Waiting for until gets fixed. +I checked the code of it doesnt setup the source from the marshalled output stream if the fault message doesnt have the detail information +Updata patch against the latest morning Jacques ok i will try to refactor the code and try to make it more simpleHave a good weekstarthere is the fix for the add address dialog and the countrystate autocomplete +As noted on r works for me +Thanks for tracking this down +See if Hudson agrees +Ill do some more investigations +Theres no user load but its still taking of the CPU +Closing old resolved that dont have a fix version and therefore cloud queries for M release +No the bundle packaging will create jars with an additional information +Nice patch that covers what I needed and works like a charm +Attached is the proposed workaround mentioned in the user mailing list is not working the around mentioned in the nable list discussion does not work need to upgrade to aries when its released to pick up the m continuously confused by the different Aries versions but isnt Karaf already using +It does not include a test but I validated it by looking at the logged output +Well a new should be added as well +Substituting revised version of and to add the special meaning of the curly braces in JDBC escape syntax +Committed revision +John Should we do something with this for or move it out to +For example in order to represent in the text file one field containing a carriage return it should be written as n +yes +Ill check this in this fix with the other two bugs you filed +By his analysis I meant the one on linked page +Florian why do you say These headers are easy to spoof and could become a security issue? We just generate for the client that sent us this header its not a security issue the client will use them not usThere are tons of application that generate taking into account these headers +Complete AM log for the failed application +Thanks +I can confirm that this fixed the issues we were having +pattern sufficient to match too +Let me try again +I am not sure about the case sensitivity of those attributes +As stacks comment patch v use zks split info to check whether region is splitting neednt keeping list +patch looks good +Owen Thanks for the comments +Can this just be marked as resolved for +Changing from Bug to Improvement as it is meant to increase current performance thresholds +Implementation of theThanks for the implementation +Michael can you attach the test case you use to replicate this issue in the app server? Thanks +Where is g which is mss +Today I fixed this +This is a duplicate of which I reported years ago and which has still not been fixed though I also posted a working patchPlease vote for issue to help me get this resolved + for patch +dont know why it failed and now passes but yes lets close this +test suite is running now +The preallocation size is not persisted +RTC move of all unresolved issues from to +Thanks for your patchThere are some points that need to discuss +Please attach an MDB s probably because you didnt set the subscription name +with fix to +I was able to reproduce the error using the provided instructions +an Apache license before committing +So the question is are they? If not Id like to close out the issue as Im trying to get things organized here in JIRA so that developers and contributors can have a good idea of what issues are out there that really need some attention +Thanks Prasad. +Ive added and Arquillian to main be in production during next release + +David why. +instead for multivalued fields the field handles this internally +runtime dependency issue in rspecHowever includes rspec +Reproduced and i think fixed with attached patch +would be great if you could apply this patch for and to the latest code fixed my problem i just had a build where the pageCreator was always set to null +Please review and let me know what you think! Thanks +Thanks Stack and TedYoure right +It adds tests for all the mojos and it also tests schema imports. +This issue has been fixed as part of the patch of. +Sorry about that +Patch for trunk now introduces +File URI should be encouraged is what Joe probably meant +Marking for +HelloWould you please try the patch? Thanks a lotBest regards Richard Patch committed in revision +Same comment as in what is the work to be done to get this included +This should be pretty useful to debug slow calls! +Heres an updated patch for where the reader and writer thread pools keep separate reservoir samples using an array of and then reply them into at the end for printing +duplicate of part already fixedwill be available in build all old Resolved issues +When the browser authenticates to the remote Idp and forwards the results to the TAS for Hadoop identity token how can TAS be sure that the authentication result sent by the browser is not faked +Well probably needs this as a component of clustering too. +SureshThis Patch is corrected now +Patch applied on and trunk +Messages for range validator contains the correct text +Sure I will get the dump and post it here +Added a format flag +Konstantin can you pl tell me if you changed your default Configuration? Marking it as patch available to trigger Hadoopqa tests +I will incorporate the feedback and post a new patch +I am using and for my own servlet +no error thrown +See for example +This patch has the empty work directory available as scratch space through environment variable The directory layout is as described earlier +Fixed for upcoming release. +I do like the idea of one less jar for a simpler for to configure web apps +Simple api to determine if async work is completed +Sure were going to clean it in a minute but Ive also seen cases where hundreds of scanners are created per minuteElse this feature is a sure way to get users to DDOS themselves +script for creating additional table and view in the VCL MYSQL database to enable the VCL broker +I fixed the critical issues on trunk and branches so we should at least be equivalent to there +The only thing that worked fine for me is the from the zip file +Jon Hermes is correct it has to be invalid UTF bytes to trigger line patches to change it to Once the value is in the log in +It is not to say that a broken trunk is an optimal situation +The stack trace and use case seems similar to +Incorporating review for the late response I missed thisThanks for the clarification Mayank Tucu Konst +Set correct fix version +Rajesh correct it is all under version control so all on track for making it into the next CP. +We havent looked at the Quartz clustering capabilities if it were set up correctly I suppose it would balance between the nodes at this moment the only Quartz jobs that would be affected would be the scheduled control jobs +It works fine +patch applied thanks. +Im working on the patch for this issue +Small and minimal is good of course but not a must +Making it visible would be similar to adding a tag to each message +Hint TTL will be exactly gc grace for the reasons documented inusing backdated deletes as a form of optimistic locking sounds like a bad idea would you care to defend it +Reporters of these issues please verify the fixes and help close these issues +This is indeed a duplicate of as Dan suspected +Fixed as of revision Sorry for the inconvenience +I already have the patch for if youre interested +Fixed after verificationso close it. +Btw I have a unit test case working already in trunk so I think this problem has been solved already +Ping anyone? Id like to close this out one way or another +Contributions are always welcome +Read repair and hinted handoff could eventually fix this data +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +Verified on EAP ER +But this is what needs to be automated + +Integrated to as well. +Looks just OK on my android phone except downloads +We have to wait next build results +Can you test it Steve +So could try upgrading the versionHowever dont want to do this until is out in case it creates other issuesThe underlying problem is probably that the base path of the templates is set to the parent folder of the selected template and then all templates are relative to that path +In additional testing Ive uncovered a very strange behavior running on trunk that does not happen in branchThis is the only issue I know of that could cause this problemWill provide more info tomorrow and file new jira. +Thanks a lot for the contribution Plamen +I just reviewed the patch for this issue seems most emconf updates should can be merged by SVN automatically because they are not in same context with this one +Attaching the full log of the container +Tested today after filing this issue +Heres a patch with a fix +Everybody can I please get viewsopinions on this subject? As I plan to pushout the final release early next week Id appreciate muchas comments questionsand contributions +A general tab would probably be best in the long run +For your reviewingtesting pleasure +The is just a single script that addresses a single migration incident +The enhancement has been added. +copy from the attachment to buildmakecomponents +core tests failures unrelated to it was also flagged due to missing tests +It is also referenced as Community Download for Spring Batch Admin and it redirects to a landing page that does not mention anything about downloading Spring Batch Admin. +traadumpfirst is the first thread dump traadump is generated later +committed to and HEAD +Just replace the original one in drlvmvmtestssmokestress with the attachment and try to run it individually + has been marked as a duplicate of this bug +only rename this patchs name to so that itll not confuse other users +Display bug fixed indeedHowever no log message is displayed when a project uses a deprecated key +This is odd I guess if the cache store appears in the thread dump its cos you have preloading enabledHowever even if preloading is enabled it should only happen once when the cache is retrieved first time +Meaning if we made this change wed actually break what are correct String done by QP today +It seems that is run before Configuration table is populated with value for thus the internal default is not initialized and is +Properties wrap the existing getter and setter functions to insure the internal data structures are still updated properly +I removed duplicated definitions only +I think it takes some upgrading in the project +but it was just TOO easy to allow the did autowire by name because that is what fit best with the projects I needed it for +However having the delete acks creates a channel to report block deletions also to standby +Hi TimI make a new patch for this problem +Perhaps should retry on calls to +here is a patch converting all uses of startsWith endsWith will commit this later today if nobody startsWith endsWith to +It was my misunderstanding +I created this JIRA seeing some freq failures +We can provide a configuration which will enable this feature explicitly if the user wants it +Validation errors are files with markers are generated for all the guides +Manik I have not done anything since January +It never seemed fully resolved to me +Making patch generated by git work with p +i will write a condition that tests wouldnt be started if firefox binary is not reports are now in correct are now run on jboss portal GA version +we can expect kuotai to release the first version of the patch in this weekend for review +usability issue needs fixing +Now we must discuss about this +on v +New unit tests have been added +SVN Export of the latest code from trunk upgraded with work with Jetty and with the caveat of some test modifications I had to make listed in my comment that need addressed and the osgi properties needing updated. +I am using Hadoop from Yahoos distributionI dont know whether there are differences related to this behavior + +In table cell selection with keyboard and mouse is exactly what I had been looking for and working towards. +Ive attached an example project +Please check on +Happy to helpPerhaps we could add some CSS +Patch attached +Could have sworn I fixed this already but I cannot find a relevant issue +David is so you for the reply +trunk after creating patch file +Not enough information is available in the issue to understand what the issue isPlease resend the question to the user mailing list if you are still experiencing this issueThe user mailing list email address is user +Am I looking at different part of the code from what you are referring to +Merged thanks. +I updated the article +report core for verification need to replace version used in these errors are still present but so far I could not find anything that is affected by that errors +Thanks +Doesnt that work for text annotationsCould you attach a failing test case +has been marked as a duplicate of this bug +html and +From Ians output above it seems only the naming is at fault and thus it is a bug in how the build is invoking javah +HiAny idea when this bug will be fixed? This is quite a blocker one for usThanksBest regardsGauthier be covered by +Ill close this as a wont fix +Biggest problem with this proposal is whether it wont open up pandoras box and people will take this for the default way of doing things rather than it being an optimization +I think we may of already applied this patch +fixed in current SVN +I think I have a fix as part of a fairly big refactoring of +I always hate to be forced to use reflexion and classloaders but I cant think about a smarter way +Rob I guess this issue shuold just be resolved and the problem of restart is separate issue +I pull them out of my build path create a libgroovy dir in my project import just the groovy jars I need and add those back to my Eclipse project build s there in the plugin is the bare minimum to get someone up and running in you have opinions about this please start a thread on the Groovy mail list +There is no way we can take any LGPL licensed code in any form or fashion into a official Apache distribution + is a list of software grant participants +The test looks fine but doesnt need to be publicAlso is this sufficient to close the ticket? We should probably indicate exactly what it is were trying to test in the future +Dmitriy I agree with Cheolsoo +The description is not correct +Verified in Developer StudioVersion Build id Build date +Uploaded same problem exists in due to CP development +will improve creation of streamed object +MichaelUsing indeed solves my problems +Revision . +If we can add the tracking url that would be useful in UIThanks good Thanks it back to myself for change due to ThanksMayank +It adds modpython as both an and a and the and configuration items +Its vpe problem and I think should be fixed for for It seems that you have forgotten to commit a test in branch +Allen thanks for your comments +I will be on vacation for next few days andwill look at the review comments once I am back +Attached in the modified patch +I just committed this. +needs to be put into WTPEclipse as feature move of M to access to svn causes this to cant reproduce this issue in JBDS +Thanks Brock! +You are right that decimal max value is enough but we are having some relations with other companies that give certification to our applications and they try to check all the extreme cases such as this +Setup command was missedClosing the issue +the first patch worked for me +We have built a cxf integration in our product so I need to extract all the pieces in a test that does not rely on our framework + +Committed to TRUNK after adding a bit more text after the link that hopefully will encourage people to click through to the fancy thrift doc of the thrift interface +consider this patch that removes all the code cluttering but leaves the general mechanism as it is +Any better idea +Should we tune this parameter instead for maximum performanceMade that inherits from and includes the patch +Moving issues that arent going to make it into to as old minor and with the Eclipse support in STS its redundant +Snapshot published. +The patch appears to include new or modified tests +Filed to investigate it +Still no I did create a diff out of +Solr knows that it just called into the class loader to locate a class by name +Thats going to end up in the apache nexus right? Should be immensely useful to my current teamThats the plan +Patch for trunk +bq +Looks good to me +Property introspection seems to work fine constructor parameter having precedence +agree +Fixing a typo in the good sanka +I just dont think until we have an architecture decided on that we should create a new project +I just committed this +You can now set this using the argLine option. +Committed with regression tests. +Thanks for reporting thisThe problem is that duplicate events are sent in the unit tests +Waiting for review to Mar +I tested this on my linux box by using a path that had a space in the file nameI prefer this approach because it seems strange to provide an OS specific listing of files to the log generator +So which one to use for map? Map vs Tuple of tuples +The mc release includes updated handling of the tcl +The logic for includes has not changed we will to any stylesheet that is imported as a result of an inclusion +We should definitely keep thatOtherwise the changes are open to discussion Im not particularly sold on my own patch just providing it so we have a point of reference for discussionThe biggest thing to figure out is whether we even expose to the user +we dont want to insert a n between two runs within an SDT +Can you attach the test case then? Or a simplified version of it +Inside method Ive inserted a trycatch to handle the exception thrown by proxy factory +trunkCommitted revision +The assertion error itself doesnt tell us anything new +Seems that we need to cache the value or work it out in the constructor. +The new stat for corrupt block is not required since it is already there +As an admin I need to be able to see in an overarching view everything that is installed across my estate +Ok I think i found it it was an problematic event call the context was bypassed due to a refactoring done between and +In the number of AM attempts is set and not as is the case in +makes my replication test happy +Same issue as described in parent JIRA and as further described in +Ill commit once we branch off thanks +This has to do with the fact that RADIX doesnt install PCS by default +If this ear is unpackaged and deployed then when undeployed WS will undeploy resolved issues for already released versions +On Sunday March at PM Vadim Gritsenko wroteKeep wondering I get this on WK with and without Cocoon and Tomcat +Stability of the outgoing IP address is only guaranteed per destination IP address +Closing all resolved tickets from or older +of them saying +Good point +Heres a stab at the bat file change I dont have windows handy can someone test this out and modify as appropriate if it doesnt work +is somewhat connected to this issue so a good solution here would probably render obsolete +Attaching console log and event files from Jenkins build +This is moved to testing which is the entry point for running test casesant test now runs the tests when it finds python of version or greater specified through +Actually I would vote for HDFS to make flush cheaper again +The SY tooling actually approaches this in a much more graphical way so I suspect the spirit of this JIRA has been addressed even if we didnt take the specific approach suggested by the documents attached +Found out attributes on wrapper element was not being handled correctly so fixed that +Attached is an updated patch for junit +Are there any plans for this issue +Thanks Hitesh +Claus Could you provide a unit test for the new class alsoWorking on it +Yes +The database procedure could be stored in a jar file in the database +Lack of Sleep Stupid things +Feel free to volunteer if it is really that important to you thats what makes OSS go +Hey Roman is this not a duplicate of + +I wish we had a clearer separation between leaf queues queues which jobs can be submitted against parent queues used to manage resources for a set of leaf nodesI think users will get confused by the fact that you cant have a configuration with and +stack has links from master and region server web ui to configuration pageif you would make based on then is betterOtherwise is betterThanks +Please find the sample and the integration test from the attachmentThis patch also contain modifications to Sample as Callout configuration will be changed after applying patch for +I will do that and check everything in when Ive also coded the conditional download of the mysql driver in the build process +Thanks for noting +Portletbridge has used Portlet locales instead of that defined by +I thought this should also be handled in code as well so i modified source code of and to handle url escaping +Would we have a subclass that overrides this thats used as a base class for optimized implementations +Would you please double check it? If so please add more missing transitions inHere theoretically APPACCEPTED should not exist when an is at RUNNING since the event has been consumed before the moves to RUNNING. +called it goal +KennyI appreciate the patches but it looks like this one wont be needed +Id like Cos or Roman to have a second set of eyeballs on it before committingThanks for following the long and winding road on this one +I just committed this +Completed At revision +How heavy is readwrite load +Thanks Ron +Incorporated most of Devarajs commentsTwo major changes from the previous patch Pulled taskTrackerjobcachework to taskTrackerjobcachework +Wouldnt be fixed because of another solution with is found +Hi Tobias see comments in to get the required SQL request to help you clean up your DB +All test executables need to be enclosed in this option and only be build when needed +Committed thanks! +Feel free to go ahead and commit it +I think including the internal Derby class names in the permission doesnt seem quite right although I guess we would need to specify derby specifically vs other databasesSince this would be diagnostic tool and really shouldnt be used or parsed in any production environment we could as you suggest always revert to the old output when running under security manager and not even bother with adding a permission at this time +support unqualified pom variables +Anything left to do here which is not covered by +Updated on latest trunk +I agree with you that the recommended setting for edit log toleration should be disabled +The first is a transient code review the next two are forked source trees and the last is a search page +Merged the change to +diff file diff since last revision Lots of changes due to some Authentication componentXSD Schema for configuration filesRefactoring TestcasesYonik DougPlease commit this patch soonthanks Header supportRecovering component for lucene storagefixed update insert bugs +Added some clean up between tests as tests got run in opposite order +Is there more to the trace why did the job fail I wonder +I will get a new patch fro review asap +This will prevent other people from using this transitory class and avoid documentationThis patch will fail on until is committed +JamesDoes it have to be a DTD or will an XSD be sufficient to meet your requirementsI have a particular dislike of born of having to deal with them in relation to SGML documents back in the old daysThere are plenty of tools that can work with and use them for validation +Thanks! +If you try to run two versions of Castor with different configurations in the same VM youll run into problems +This is a major refactor and Im not suggesting we do it here +I can open files today without problems but now ORM view is completely missing +I didnt describe the patch well enoughThe are not resolved when the cache is filled +Upgraded jaxb impl to that included the fix to this problem +The only minor thing that I saw that needs to be changes is size of a single tuple bag should be changed to rather than +The whole trouble with jsr is that it does not provide you with means to configure the script engine +Now that work is underway with AS all previous community releases are +Initial patch for review +Fixed per Shanes directions. +I ran locally and it passed +RajeshThis is already fixed in the new getting started examples that I provided in the last drop +Changing the log timestamp time zone could be considered a backwards compatibility break +Sorry for closing it too early +Must be something related to your environment but well see if we can produce a better error message than area has since been improved +Should this be fixed in? Is it a regression + +I suspect even a normal put with IGNOREPREVIOUSVALUES can lead to the same kind of problem +Only increment deleteappend has this problem? on patch +Please try it then and reopen this bug if youll see the same issue in that build +And it work for downloadingSo could you show me how do you do in +It turned out that undeploy listApps and redeploy were not using the correct method +patch to fix in commit +heres another version its using an and no scripts or events or anything else but sub proccesses though +contents of reference files and what about rpc? Should it have magic preamble or you think if its broke on an rpc its just broke and the on retry itll be ok? Good stuff Todd +Not for commit here. +Yep debug sounds like the right level if the activate method is not specified and none is present +thanks +No updates in the last year +If your page is stateless there wont be a th row present when the next request comes inWhat type of component do you use for the links +Patch looks good so I committed it with an adapted test to exercise equivalent generated classes +Im curious about how multiple args will map to Message +If the return value of is include or include and seek next column it calls the Checkversions to check against the number of versions +I have added a test to demonstrate the issue suggested below and find the result to be similar to what mysql produces +Thanks lads +I always understood active to mean active unless another profile is explicitly specified +This is a bug +Adding new features to contrib components at this point would not get my vote +Performance is improved in +trivial tweak required for patch +Cannot reproduce as well so lets close this issue +See one of the for solution merged +I tried doing everything as you described and issues to seem to be fixed +tests is an ongoing effort +is not just but its needed by test. +Lets reopen this and try to get it done for +Ignored as issues +Separate task +If you comment out the last line of config per the INI comment you will see the test fail indicating that caching is working as expected +Craig have you tried renaming property to be ? This removes the error for +It isnt the good well have Continuum profiles +the reports are located in the runtimelogshtml ErwanYour patch is in trunk at r +Then put our normal war on it still runs fine +a to diff command to always compare file in text mode +Initial patch for review +MINA is not released yet +When we considered switching to a new strong name key the scenario of two different dependencies requiring two different strong names never came up +Uploading a patch that might helpHavent been able to validate this because I was unable to consistently reproduce the issue +That sounds good to me +If you get the same error message please transfer this bug to Xerces but update the description so its clear what youre reporting +Verified by Oleg. +So as far as I understand it is a path as in a URL +This will not make into +When Automatically Detect is unselected I can still click Customize why +Attaching a revised patch after mailing list feedback +The temp destination is bound to the lifetime of the connection that created it +I just need to clean it up a little and test it then Ill submit a PR +The system catalog first needs to change the compiliation schema id column of and system tables nullability to true +So this is actually a feature rather than a bugAs per discussion in the IRC channel closing this issue +Incidentally how come the test hasnt been failing on AIX as a result of the signedunsigned mismatch? It might need to be enhanced to detect this +Good to see the doc changes +The tests do not compile at the moment but patch will follow +Its best to have all the kids using the same toys +this is most likely the case with any JMS implementation I have seen actually. +Update to http and jaxws to set on reviewed and applied by Eoghan with r +JoeI am happy to have you apply this fix I think I understand better +Part of a bulk update all resolutions changed to done please review history to original resolution type +reverts the changes which I forgot to do it in the previous patch +The change works in Ubuntu as well + Still happens in +Thanks for the patch Chris. +Frank BillCan we agree on something so that this issue can be closedRegards Senaka Im easy +with +Implementation has changed see. +I have run all uts in have passed locally with +Makes sense +Its better if the sleep interval could take the configured heartbeat interval into consideration +This requires that the configurations live in the repository +The method is better than the patch provided as part of since that method will fail for cases where users install their in Eclipse installation locationsBesides the diff there are two new files to be added that are included as attachmentstrunkpluginseclipsesrcorgapachederbyuicontainertrunkpluginseclipsesrcorgapachederbyuicontainer AronI tried applying your patch but I have some questions wrt +Great thanks for the feedback +Closed after release. +You have done a lot in this but the JIRA says it is Minor +Ping? Curator is trying to produce a release but its blocked on this issue +You are using HTD to carry a message used by the create table where you pass lots of regions +Thanks +I am also in favor of living with it +Attaching and which add missing functions to the JDBC escape syntax for fn keyword topic +This patch checks if the datanode registration has failed +This makes things even more PaulYour patch is in trunk r r r r +It crashes on failed assertion quickly in initialization because tracing is done in a wrong suspend disabled state +My two concerns are According the trace log above it never gets in thehandlenocacheoperationonforwardserverresponse so I dont see how a change there could have an effect on this issue +I am actually wondering whether we should fix it +Verified at +we have multiple ASsSeams on hudson so it all depends on how your test is picking up the runtime +You currently instantiate based on the URIs so you may as well instantiate directly the streams based on that +I can confirm as well that the issue is present in snapshot +and List can be used respectively +They are being executed within embedded container +This should also get handled by the EJB container in the Remoting poing +Changes are +Hey Steve +After realizing that you can get the same effect as the classes applying to the current classloader with the current code by depending on a parent config with only the classes specified I decided that whatever the original intent the current behavior is silly and just applying to the current classloader is appropriatetrunk rev +Not critical so for +Ill commit this shortly. +Cool will update the fix versions +As far as I can tell we are setting up initial IV the same way as MRI +Ive deployed many many times and not seen this issue until I decided to try installing and upgrading it +Close as I could not reproduce as well +IIRC a runtime relationship was being created for Customer Address which was messing things upIll have to dig through the SVN history to see what I did with mandatory relationships that I mentioned being problematic in the initial bug submission. +This patch has provided the original exception when invoking the serviceModifiy file and add the cause exception on the constructor methodModify file and make the method setupMethod throws an +and still going for a reinstall +The stat is the same +Issue may already be resolved may be invalid or may never be fixed +looks like we mean to fix this in +Read the documentation and ask for help on the user forum not here +Hi MaximThis is the patch for correct this issue +WARN Error getting nodes version in CLOSING state aborting close of miwengregionHxBXxbMxBPossible fixes +Surprised that you were able to delete the log file on Windows +Can you retest with CXF and if there is still an issue provide a testcase +But I dont understand why has now become an implementer of the manager interface +Well on the second thought this is more likely a suggested way to fix this not a complete fix +The above finding actually should go to closed. +We still need a unit test for though +sync is an asynchronous operation we usually ignore the result +In release +then after the request is synced to disk you process the requests under the big lock in the order that you synced them +Fixed in HEAD and the branch +batch transition resolvedfixed bugs to closedfixed +What do you mean by deploy undeploy? Do you mean start stop the bundle +I believe an Archiva issue any more +Yes I am working from the NG baseline +My Firefox proxy thank you I will try without any chance you tried this and can verify it was because of wrong SOCKS configuration +passes fine +broke it +If we choose to let it be for now we need to add a document somewhere at least +This is the source distribution of our portlet application +Thats what actually seems to work for comment for parent issue + +Also which section of JDBC spec indicates that Derby is not currently in compliance with JDBC with respect to this methodThe method is implementedIt is implemented by throwing a because the database engine does not support the featureSection says the driver must include an implementation of but does not say it mustfully implement the interface +Works for redhat as well +This looks good to me +I could be wrong but the code itself works with both versions but the jar we distribute only works with the version we compiled against +This patch provides an example fix for the APPEND decoding issue +bq +After the error occurred with debug option enabled I can manually invoke the javadoc toolCProgram FilesJavajre +especially if it leads to a memory leak! I recommend that we deprecate this method +The code does not need to be there and you were correct in leaving it out +One tricky issue about using distcp is reliably determining the status of the map reduce job +OK I committed the patch with the updated method names in revision . +Also there is still an ongoing discussion regarding the design proposalPlease back this patch out +Please advise +Thanks Aaron! +Can you provide a specific example? I cant see anywhere in the tag that would mask an exception +Moving to the correct component +Tom I can work with you on getting this stuff into CVS when you feel its appropriate +There is only one WAR that will not start properly which is very odd +just for your information +Log in to admin console +Thanks for drawing out your usecase makes senseClosing this issue. +I dont see how the two points argue with each other thoughWe will need to be able to configure with a user friendly format most plugins will have that if not just to show the user what can be configured +A prudent sysadmin would want some measure of protection for Solr even on a private network not to mention running Solr in the Cloud +patch with comments at class level +ClausAre you looking for some samples for the wiki? If so I could repurpose some of the unit tests to demonstrate the basic functionality +fixed in svn +This is a screenshot showing different color codes on line on a violation +Test case that reproduces the analysis +I think this is an enhancementto the extended datatable client api so perhaps contextMenu isnt as critical forme as Id first thought without the datatable +Updated patch to remove some classes that are now part of jclouds +all bugs. + +The document import wizard uses an empty type system to create a virgin CAS in which the document text is inserted +Hum looks like that didnt solve at all +Verified fixed in trunk with svn rev Executed the repro steps and I no longer received the exeption +Assigning to Kristen dont believe her patch has been applied responsive tabs to all admin pagesUpdated all forms to use bootstraps stylehtml structureUpdated styles and position of items to clean up ui where needed +Oh gosh that solved the problem too just as my workaround did +Domain Id Scheduled disabling account in domain INFO NULL Started Starting job for disabling account +fixed comments +bulk close for the release +Dan Sounds like a good idea to use isRequestorMichael Can you change that +This is a patch to branch that has fixed the findbug warning +Maybe there are more elegant ways to get the outgoing interception integrated +antlib +Michael Ive made http url for simplicity +txt to +Sorry for the trouble Vikram! +I must have been looking at the older revision +I think it makes sense to leave the config options as +patchAlso small update to delete misleading messages +I reverted changes that has been committed +If you cant AddRemove it directly shouldnt you be unable to deploy it via the wizard +And if new groovy programmers use each on a file object they will quickly notice that their code wont do what they expect it to do by simply testing it and looking up groovy docs for the correct other method they were looking for in the first place +Test to reproduce and fix for all + with branch pls +Attaching patch which puts the calls for waitforpostcommit after each commit following a delete +We can also use it in front of the RS of the right table of left semi join +Yes +HiHere is a patch it works for the trunk and the branchBR +Could you please send me an example of text with which this error happens so that I have a starting point +What is the estimate on time required for this task +we dont get metadata version exception anymore +Did you scrub first +The only use of max is to let readLine know when to stop +Done +If using default keystore and truststore than no need to specify location of keystore or truststore +Forgot to mention in my test the val is an instance ofll review and see if its doing the right thing its definitely possible its doing the wrong thing +I generally format my code so that the real stuff takes the most space and the error handling is as compact as possible +Ill take a look at the EPR and post a suggestion here before impl +Fixed in r +Preliminary patch to coordinate with changes Mark is you have cores that dont have explicit name attributes I think that a message should be logged saying that certain API actions like SWAP and RENAME will not be available +I have no idea how many bad call records there will be +Any feedback is appreciated +I believe this is the issue you are currently working on + to trunk +Do you know if their patch has been committed to to cassandras trunk +Can you please attach better illustration picture for this problem +Erm +Committed to trunk +Thanks again btw +afaik there isnt a mechanism in groovydoc that handles class aliasing +Greu que sap de fer o dhaver fet de deixar o dhaver deixat de fer alguna cosa +If you are not the intendedrecipient any review disclosure copying distribution retention or anyaction or inaction taken in reliance on this information is prohibited andmay be unlawful +fyi Ive just created on an hdfs linked issue +This requirement is to do this through +To Stijn Its not always correct that if modules are unchanged they should not be released because sometimes if component X is released and Y depends on it you would release Y as well even though its the only way is to specify what modules to release +In the short term were planning to implement the MQO prototype using multiple Drivers embedded directly in a client program +Fix applied +As a workaround you could try to specify a value for the soapAction in your WSDL +targethdkjdkjrebin targethdklib second failurecp +Attaching two states of a message that demonstrate the change in the body that causes the sign verification failureHi DobriCan you attach the debug log and exception thats thrown on the server side? That stack trace doesnt tell me anything about where the exception is being thrown +But can you fill out the release notes section of this ticket and describe how is superseded by threshold or noconditionaltask +The patch makes enable to change authenticationI ran the test it on windows server +Out of curiosity could we save a FOLLOWING notification as our lastMessage then even if we dont send it? That would give you the current state more accurately than a stale LOOKING notification and it seems like it could solve this particular problem +If we keep implementing every feature that is already on towards Messaging then Messaging becomes. +The looks like it will skip the encoding of the message being written if it is an instance of an +Well Brett what do you want? I even showed that the problem is not specific to the repo creation page but a more general problem +Patch applies cleanly to the and trunk branches +Logged In YES useridI am setting the status of this open since we still needto be able to see when statements are not being closedindependent of resultsets +This test currently fails due to bugs in the class under test +Same problem here more screenshots showing how badly the pages look when the CSS gets manged in the installer +If it works Ill extend it to the main BZ instance +I didnt file a JIRA yet but I also saw errors on my console about the JMX bean where we return instances and JMX has no idea what to do with them +New patch +As you know we love contributions +Committed to trunk! Thank you very much for you patch +Thanks. +Changed get to return an arraylist on top of the +This made the problem go away +works fine for me now with +I will also attach a dev build of gem for you to try out +Note that these are intentionally not targeted at because they dont belong in the release notes +getting a good model in ZK seems really importantThe wiki page is my attempt to capture it +add an it to test in rev . +Can this be closed? +Shirts should be delivered on the th to the hotel in Addison TX. +Patch for the branch +Thanks DevenPatch applied to REGEX module at repo revision rPlease verify +I can not reproduce this bug on the current build +If thats the case you can go ahead and closeThanksMichi + addressing two additional comments from Travis Append to HADOOPOPTS instead of overwriting it +A problem I see with the directives aproach is that it is appropriate only forread only configurations +There are several issues in the central tests fails because the wait method runs in the UI the job also uses the UI thread to notify listeners we get a timeout The cheat sheet test creates a cheatsheetBecause PROMPT is the default cheatsheet action the test opens a dialog and PR fixes the issue +Comments follow There isnt any config that enables sync in +I tried the same test out on and it works fine +stack i fix the bug based on will make anoher patch based on trunk +I retested this with +bq +superceded by bug This bug has been marked as a duplicate of +The attached test case you submitted does not make this easy +This feature has been started and is in a list of resources that support log and config track see the log and config tracking must be explicitly enabled by the user through the productconfiguration +Thanks for catching the issue! +DrewIm going to make a patch to make some improvements to the parent POM nextAt CXF we have one Maven module per plugin but we group them in a subdirectory +The problem seems to be function hide in +Only if submission is stuck for say times polling interval that way it will not be printed at all in the common case but will print something that is really useful when submission is stuck for some reason +It looks like this has been fixed and there are notes on the FAQ for thisIs anyone still have troubles with this otherwise I suggest we close +Then I iterated it and got a list +after testing with ACF this is now a compatibility it turns out ACF does not throw an exception it does a merge of the folders and overwrites files with the same name +work was discussed on +Resource filtering is the job of the Maven Resources Plugin so we would need to know the version of that plugin to investigate +added the release notes as resolved a rule was made invalid by changes to the DSL configuration users would be presented with an error dialog if they attempted to view or delete that rule +Its approaching two months since I asked whether anyone can still reproduce the problemIts got eight votes Id really like to have heard at least one of you say I can no longer reproduce it or It can still be provoked and here are attachments which demonstrate how +Its actually kind of abusive and perhaps not interesting for anyone but us with our special setup +To override this the user would add a new attribute in the dd for the Cas Multiplier +I have already upgraded the parent and probably forgot to close this issue. +I am on this +I personally think that permission to execute the procedure should be sufficient since compress is not really DELETE INSERT or UPDATE or SELECT +The operation names should probably be add and remove rather than assignunassign for consistency with other operations +Assigning back to assertion is not right +new snapshot has been uploaded to the snapshot repository which should fix the issue. +Im only merging code +settings folderRestarting eclipse does not solve anythingTruth is that I am creating simple Java project and then I add ESB file into it +Thanks +Verified on Hudson Stable Branch job +Both of these tests look for a mutually exclusive intent error to prove that an intent has been added to the model when an Requires annotation is present in the Java implementation +Fair enough +Oleg right I did not see isOpen was just a simple wrapped boolean checkI do not like the semantics of that exception either +But I think enhancing the deserialized objects is a more consistent approachIm surprised that the other solution didnt work for you +Two videos showing relative performance using Node vs current DWR version of the demo code integrated into Wookie using feature connector API +since this is mostly a plexus fix the hours on this end are basically just for testing +Merci +When using Ivy via its API there are no wildcards mechanisms in this context +Thanks +Can you please give details on exactly what doesnt work for you? Note that mappedHandlers needs to be populated with references to handlers not with bean names +Udai created a series of howto documents about setting up Castor with eclipse and about execution of Castors JDO test suites based on the introduction from me + old issues +got busy over the holidays with a demo today +this is the wrong URL this one is MB large! It seems it did pass our setting! This page does not return a HTTP respone header if thats the reason it passed the limit we need an additional issue to handle those cases +Jorge Im pretty sure that the issue youre facing relates to On a Maven project when launching an analysis the semaphore mechanism should prevent any parallel analysis on modules +Instead of using package can we create an or to prevent clutter at the top castor package else would go in this new package? Im not sure creating a new package isappropriate for one exception class +WMS +ping Paul this is what I was referring to last night +I added a dependency on for upstream issue +bulk defer of open issues to +So is that a requirement +Added extra checking with warning to avoid the NPE +We also should get findbugscheckstyle back againHave you admin roles for the jenkins for our builds? Or should we open tickets at INFRAbq +defer to +The main question is about scoping the user should not aware that i add new component so i hide it inside a composite +Pushed these requests into core components and quickstarts. +I have factored out the refactor to so I removed the creation of the to put it into that patch +is the goal of this to create a deserializerNo it is purely to break the dependency that the core module has on mapred +change offset and buffSize to INKU type to avoid negative number to be treat the return type of those AIO api functions based on Bryans suggestion +How exactly are you declaring that form action using Springs form tag library? And whats the original request URI for the request which eventually renders that form actionve committed a candidate fix to trunk now +Actually when looking at the current code in trunk it looks like a mechanism was implemented for you to specify a different xmlbeans versionBuildr Could you update to +I know that there others who think otherwise that should be a conprehensive list of all you could possibly change +Could a committer take a look at pushing this! Looks like review is done please commit? Thanks +Rather than upload file maybe something like Replace current file? Just some indicator that this action will replace the currently open file +This looks like a Weld bug see +I just committed this +Just tried it out works exactly as expectedIm going to internationalizelocalize the element of the CMIS connector as an example and then we can merge the branch back in +Therefore this information needs to be calculated so it is possible that the Web Console is calculating it incorrectly +Ah sorry I need to add download link to it +Model tries to update but because it checks file modification stamp to avoid unnecessary update it skips the file +Various tests have been added based on exchanges on the user list but all of them pass in Revision so perhaps this problem has resolved itself in some way +Thanks for the review ErikOn +Add build of srccontribtransactional as proof that this rigging works +Might be related and fixed by this +At this point it just seems like it should be documented as requirement that if the change provide the SQL for migrating from one version to the other +Please close this JIRA it is not relevant to this project +Done +Thanks Dart +sorry for such a question i know this practice but after going through the list i came across that work is going around on many issue still no one is against the assignee fieldso i preferred this way to problem +I assume that one could set the quota on the root directory to something based on the amount of memory available on the namenode as the simplest means of control right +Also if you use a type class and there is only pojo then the pojo is not wrapped in a list +Bonus points if the cleaner service recognizes and acts upon orphaned temporary files from clients that crasheddisconnected during upload +rather than in containers +You need to apply both patches +Marking the jira as resolved as the patch has been committed to +Thanks +jdench setenv DERBYHOME homechaase jdench echo CLASSPATH +In that case a profile approach would definitely be needed as the install structure of the different servers are obviously very this would move the responsibility of handling dependencies out of the EAR packager and into the domain of one of the deployment modules +have fun +scouting around on the internets Im seeing signs that a fully syncd flush on HDFS may not be available in a released version of HDFS Anybody know the status of this for HDFS +Your changes are in SVN rev . +varcassandraseems kind of weird to have build script create those +Branch Teiid Client jar says teeid so now its is this included in jboss tools by now +I need to find a way to add another level in the result set tree so the two kinds of ordering can ideally eliminating the implementation sort if the one specified gives the same ordering +I have attached the first go at my patch +The changes have been applied in r. +we updated the ant tasks for compilation so that all jbpm libs will contain compatible byte code. +Would openseq be better +Removed previous incomplete patch +We can schedule a meeting to have a discussion on the feature some time in late August or early September depending on the interest +Revised script added support and now picks up Hadoop also +Should have looked more closely at the patch file +There were a few issues with this code they are fixed +In a effort to help resolve the issue I have attached which is the last version of the code that does not exhibit the problem +Real social networks should implement their own decoder which can be a little more sophisticated by validating userIds appIds and what not +This would allow end users to fix their mistakes +The file extension does make it through and that is what Tika needs for decoding the document so hopefully this is not going to be of critical importanceYou will not be able to use wire debugging for this purpose unfortunately but you should be able to see what Solr does +Ill take a look and roll it back if I cannot find a way around this issue +Ok yes I see it need region based marshalling enabled as well +Let me know if this is OK with you +Yup no change for users +Attachment has been added with description adds color pclI dont understand the motivation behind this is there an advantage of NOT using color PCL? Are there some printers which dont support it correctly or is the image quality better? Is the result smallerSecond There are some things which are duplicated in your patch in particular the lines surrounding the vW command +never mind i figured out that the case i wanted to test for was impossible an outstandingProposal queue with a NL and other proposals but i realized that other proposals will not be generated until the NL message is committed at which point it will no longer be in the outstandingProposal queue so the double NL is completely innocuous +Thanks for testing this out +bq +Internal this supposed to actually build or not quite? +Theres nothing I can do at this point +There is a tracing flag for the transports true will show all commands sent and received this might help +Hi DavidPlease try what is currently committed if it doesnt work Ill change things around until it does +And came up with the modifications you asked for +I would just leave the implementations as they are +new patch up in rb +Committed revision +Some name issues like misplacement is just trying to be consistent with in existing +Whenever anyone has time to work on this +So for Thrift thats with for transport and on the side got it +Fixed in master and just needs a to applied in branch +Closed. + +This change was isolated and didnt have the ripple effectOn the chance that we need more specialized processing for the Enum type then we might have to revert to my original thought +It is named as being a version of daytrader but it is also tomcat specific +This is a good example of an application where many managed beans are components of a framework that uses class extension to provide common behaviors +thanks +Looks good to me +Of course mazz the description of this task is EJB PluginService MonitoringI will see what we can provide for inventoryavail in the meantime +Problem is still in POI revision + +i dont understand the suggestionWhat would that help +This file contains example source code to demonstrate the bugissue +Great work +Not totally sure which values yet. +Assembled June I believe these changes should work with any version of Solaris. +Added another testD +Create a simple server run it. +Sorry didnt see this was already taken +I do believe its related to +Whwn you told me that includes work for you I tried to use the command line and it worked properly +Modifies to use getRow and print the read time to standard out +Best regards you plz provide any working sample code +It must have been a quick hack to make things work +The retry queue processor was not properly waiting for the cache to start up so it could start handling a command before the initial rehash +Updated patch now including a unit test +this issue. +Files changed svn commit file tmpSending trunksrcadminguideSending trunksrcadminguide +I guess a quick test just to make extra sure would be to put breakpoints in the code you changed and run the WFS sample requests? +If any of the shards return a response the PRH itself should also return a You may notice that theres a parameter +just committed this Thanks! +If you have ER installed you can use the available in the edsteiidclient directory +The only issue pending is that it does not insert default values for the fields that readers schema has but writers schema lacks +Sorry I did not realize that previous entries was added automatically +Patch and release note look good to me +committed to both trunk and branch. +Set next version number +So Ive applied Myrnas patch to disable the test for vms until someone has time to fix the test or product behavior or we no longer support +The answer to this question is awaiting follow up with the AS team +The reported exception illustrates this perfectly +I am pretty sure this JIRA taks has been already fixed n and trunk because we do parse the CRS object of the provided +Martin please commit this change and resolve this issue +mount point mntsec does not exist this is the output +Resolving this issue as a duplicate +I dont mind sure! + excludes from default test running test excluded at r +However the logic for Enumerationeach is trivial too +Because rendered is true the components should be validated or not +I just ran into the same yesterday +This issue and are both caused by problems handling extreme values in Gamma +remove +For right now mode handling requires a lot of hand coding +A big problem with the current system is that the new plugin often ends up in a state +With runtime instrumentation works with my test case +Okay good spotting +Good idea +If someone wants to exchange the implementation of an entity +However the company has decided to not switch to logj as the logging interface so any changes you make regarding this issue is not going to help me directly +Patch applied with some modifications +Just a question otherwise Im for this patchWhy is not an interface same with +Can we get a review for this +Koji The idea seems to be to keep the dangling nodes intact in the plan and do a search for node in every time we are trying to get it from +Committed with minor changes +But again it might be impossible to do in your system +I have just attached the new patch which throws an IO Exception instead +to both you perfectly described my thoughts +Applied to release branch as r +Attached a patch with this applied the build passes +Im working on getting corporate approval to submit patches for both bug fixes and enhancements now +Pending ranges are stored in a which uses a Hashset to store the collection of values for each key +This is an patch +Oh dear +documentation fix +That class correctly was removed in forIt does show the danger of having tests use apis at least if we want to run tests from an old release against a new releaseI think that is a good goal and was one of the reasons I put effort behind converting tests to hopefully due to lack of master files it will beeasier to run tests against future releases +Reviewed by Barry L. +Until the JPA release is live it doesnt make sense to do a JNDI releaseSo if you would like a release I suggest you send an email to user and ask for one and we will see what we can do +Verified in JBDS process created in simple java project +Uploading patch +Thanks Andrew +Is that the case for your example +Thats why Id guest that it is probablydesigned for some reason +Committed to TRUNK and M +usefalse and are intentional so the latter approach should be taken +If you have a submission for the tutorial section in the mean while please feel free to submit it and well try to get it included soon +Thanks Eli +There is a knownproblem that enlarging the font size inline does not adjust line spacingso that the enlarged text is rendered over the text in the line beforeHowever the report does not seem to describe this problem +Thanks Andrew! +If there is a problem please open a new jira and link this one with that. +Wrong scoring is a serious issue in the moment +Do you mean these should be done in two separate commits? I guess that is OK but both should be associated with this issue +In which case I have no idea why youre raisingthis issue as a blocker +Please provide some more information on your context How does your transaction proxy configuration look like etc? Ideally it would be great if you could post a minimal test case that reproduces the test illustrates the the attached zip seems to be broken +Thanks Jason +push to +I committed thisIt took over a year to get this committed! Im not sure how it slipped through the cracks for so long +We applied the patch yesterday +snapshot creation is is a local operation on NN and the latency should be in the order of seconds +We may have to backport both and this JIRA when we merge HA. +otherwise Im happy to revert +Queue the build process until all the sub modules have been +r. +patch provided by N +Please review and comment +This patch includes the source code of and +please dont open until you have first discussed on the Forums appreciate you raising the issues but this one here would probably be better as a simple forum question +Done +Fixed with rev +Thanks AndreaI extended even more your patch a test. +Damn it it is trueYou are rightit is strange because i do not need to add this line when I am logged but during the boot process it must be declared +I changed the maximum open file setting for one linux machine from to and then the passed +Adds fsm +Updated the loading of initial content files and CND files so that local filesystem paths can be used as well +In any case the bug you encountered is no longer reproducible +With the patch if the stream is closed somewhere instead of a NPE we may get an saying the stream is closed +On our project there is no currently no AOP or GWT involved +Can you attach that license to this ticketAn alternative would be to provide our own with just the minimum in it needed for alfresco to run in our test environmentKarl +CDI support is not enabled for your project so you dont have code completion +More info I saw the same symptom while the agent was running + used test dir and fixed some deprecated warnings + is used for writing temporaryintermediate file and it looks only for gz and lzo +This should now be fixed +Thanks for looking into it +This is fixed by this patch +is the followup ticket to this one +Do you still see error +This is a bug which needs to be fixed anyway. +Just wanted to add something based on another problem I was diagnosing yesterday +Thanks for responding +I think is really what should be done unless theres an overwrite flag in place. +While there are lots of explanations about the API concepts and implementations there is no reference whatsoever to the new correct name of the property itself which is +Some work was started on this on a branch but it hasnt been updated with the latest CXF changes and it needs quite a bit of cleanup and testing +but if this works for netbeans doesnt it still need to be fixed for anything else using the embedder? Im not sure it should be closed. +xml files we use for testing parsing of output files +There is a mechanism in place to dynamically register various available SASL mechanisms +I get itUsing the recipes +And here I just thought you liked golf! I guess I have never used the tee command in UNIX +Add a paragraph in the Admin guide about installation +fixed the display to show Keystore Password +This is one CF only still rightFor now yes +But if we think this should be reverted from then lets do that sooner than laterIm fine with those Im fine even with removing those as soon as system doesnt crap on startup +We have been trying to setup in a seprate place for quite some time now and without much successSo the next question is where Hbase driver should go +The section where mloc has been used in the function get is commented out it seems like pools really were created to use in this function but this code commented out +The fix was only in a unit test so this does not need to go in the release notes. +Then we have to manually fix a few of those bad datanode machines and make them come online this fixes the missing blocks problem but is a manual process and is painful +Any chance theres an updated version of this patch against the latest zookeeperIve finally been given time to get this integrated into our product at the company I work for and am anxious to finally pick it up and work on zookeeper again +It might be worth exploring this further for the release to see if we could switch to using it by default +As Roman pointed out usr shall be considered as immutableAlso people may want to tweak these templates +The attachments is from geoserver and +ArjenI understand your reluctance to add a new method again to the +Nice +If you find further issues please create a new jira and link it to this one. +This would depend on how configurable is and if all backwards compatible behaviors can be specified +and by not using a lot of dynamic endpoints you wont hit this issueA blocker is something that you cant work around at all +If we really want to try to use the standard new filter pattern well create another ticket. +Yeah I know the heritage +I think you may actually need to parse this yourself from the query string +Confirmed working on Hudson and buildbot. +Treat unscheduled issues as resolved issues for already released versions +How about adding an option to default to false allowing user to the table +There are now actions and in future more can be added +I just switched back to the original as suggested by Ellis +See canned test there any other such situations that we dont catch? +I had to move the read method from to because the does not have access to the +patch looks goodWhat about failing tests? Are they related +Przemysaw is injection is becoming more and more popular +Please add variants of the unit tests for this feature and for the BDB +Run test case and verified the fix. +Removing it seems reasonable to me. +Resolving as fixed +We could check if node has pending changes if not call save once on node otherwise call save twice on properties level but im not sure if it makes sensei dont think thats worth it +Changes look good thanks Kristian +Ive had my share of cases where the first exception wasnt actually the problem though so I wouldnt want to hide exceptions that happened later on from view if thats what you meant +Tested OK thanks! +I guess it just underscores what we all know multithreaded code is confusing +If you are still having it you may need to provide more clues or investigate it yourself +attached patch addresses the usip concern and fixes a few javac warnings +Ok Im testing with right now +We will upgrade to Xalan for the next ive tested it with fop and and +Thanks for the patch +If you have a listener on a it gets called for each and every row during This makes the model think every row gets selected for each cycle! You are right every row gets selected during APPLYREQUESTVALUES phase +Mylyn context for this Daniel and Karan for testing it. +fixed + protected would not gain much as this class would unlikely be extended +No one is suggesting that we handle it as a single call Linuxs VFS does not handle both open and isLink in a single call the way wed like to +This is not very critical downgrading its priority +So a decision must be taken +The new patch uses the stream directly +The best way to verify the behavior is use a commandline tool like curl to access the aj +We are in the middle of testing this now +So far both compaction and scan are using pread which caused some resource contention +What exactly do you have in mind? Just the plain JDBC check if the credentials are valid +Support for would be nice and I think there is no need to have a separate type it suffice to be smart in the way casts are done +will be providing documentation changes on the SQL extensions and will be attaching a patch to this issue +I found different issues the given pdf contains images which are embedded in a which is embedded in another and cant be extracted using +Thanks for comments and suggestions Konstantin +From there we can iterate on any rough edges on trunk +This SOUNDS like a and maybe impl jar conflict +Looks good overall +The problem w this caching approach is that there is no eviction +But can this happenAnway looking at the DTD it appears that timestamp and version do not accept child column elements so I think is incorrect toothanks schema did you use to reveng that +Particularly connected with +Please be reminded that configuration will only be stored after Save changes is clicked +Wildfly should be ready for this now +Nope I never did this +Rodney could you take alook at this and see if its actually a problem in the test +is thereRegards again +I have been using with concurrent sweep GC +Fixed locally didnt break Hudson. +I did this by starting jobs simultaneously that tried to move the same messages from one queue to another by calling +note my testing was done with all the other RAT jira referenced in patches +Minor comment please keep the lines with in charsCouple of other things to note +Probably worth a separate jira too +Note the space in the HTML representation +Moved to for configuring an object factory via the property style is now ever useful Properties class remains. +Mm cache summaries could be stored a bloom filters since we have to read through the entire index at startup time anyway +But I was not able to quickly reproduce your problem +This is not reproducible +pushed +Shouldnt rather extends and if so shouldnt it be named +see doesnt even do Quickling anymore so I am going to close this issue. +Committed to branch x rThanks for prodding Shawn +Committed to +Fixed +s plus one but the implementation of that isnt very beautiful +Actually ignoreCRC is for disabling checksum verification +Reading the code I kept getting confused on what the semantics of highseq were supposed to be +So I dont know why it fails for you +Yes I have tried +Hardcode Geronimo system artifacts there so that it could be used to avoid extract those console web applicationsd +I will ask for clarification +changing fix version to especially given agree this is important to fixPatch looks good +The reason is that in my case OS is actually a load balancer which has known history of past vulnerabilities related to sessions +mrepositoryorgapachehadoopUserstucu +windows is to blameP +bulk close of all resolved issues prior to release. +please close issue and let me post there +Changing scope bringing into since its harmless +This is a big problem as I am currently writing a story about Displaytag for an online magazine and I planned to highlight the export feature in the article +I agree it would be best to fix the underlying issue but at least this patch seems like a good incremental step to makethe system better until we get there +Everything works fine on Linux +Please suggest a better URL for the certified list doc +Please have a tryApologize me for giving such big test result and troubling you +radio button +Best regards George +This bug has been marked as a duplicate of +Hello VasilyIve worked with JX flow and I had some weird issues sometimes but I never had this one +Adding content +Done and tested +Does this apply without? Please advice +first we will create then if both succeed then only current container will be either assigned or will be reserved depending on the scheduler logic +Anyone have a patch +Tested in STS and Tomcat +the new sample files pls put them to modulesaxis the patch and added the samplesthanks the new patch fix the class loading stuff +The default implementation of this factory interface would just return a Collection with a single instance to retain the current behavior +Sorry Vivek for misinforming you earlier +but i am not sure how useful it will be since it will be receiving a lot of events +Looking closer into this issue it appears to be an issue with the richfaces gmap and its incompatibility with the portlet ajax isnt an issue with. +It gets the task id out of the to make the unique name and counts on the cleaning up the storage if the process dies +I am attaching the patch that fixes the bug +Thanks +The only pure Java thing that comes to my mind is using ANT to spawn a JVM and then write equivalents of command line tools +with idv tags immediately come to mind +AshishShould we still keep this issue open? +Im tempted to make this a WONTFIX since were moving away from CVS +Hi ChrisThanks for your response +What mysql server version do you use? What connectorj version do you useIf youre not running the latest connectorj upgrade it otherwise try adding donttrue to your jdbc url +since ive removed SVN tag could this issure be closeed +Heres another idea +Then check that flag just before adding it as an include +Can someone please explain the purpose of recursionCap? So far I understand that it makes it possible to adjust a recursion depth but Im not sure I understand what it really does +Thanks Jimmy. +reopen to fix version +One idea was to write a transformer that removes expletives from tweets. +Main action is Reopen Issue +However I have been using the grails from subversion trunk but I understand that you are using git nowHave you installed any other plugins besides those included with STS? Have you updated to the most recent version of the plugin +It might be handy to show the current size of the edits log too +WTP guys dont give a way to handle this +Git repos are writable +The temporary file has a suffix +menu +contains an implementation of called XWWS +Investigating +This patch updates to implement +The strange changes were that comments were modified from required to optional +ad see Igors commentad i wouldnt say the mavenish terms should be replaced but Project name and Base Package Path could be put in parenthesisusers will probably have to deal with maven sooner or later so its better to use the terminology right from the beginningad i will add the links. +We can still extend it so that Sylvains suggestion holds but in real terms itll override all methods so wont really extend +This is in place in branchesjmxmcgsoc and will be merged to Aidan can you review this change please thanks. +Right now adding functionality to a project that relies heavily on CDI causes a lot of painThanks in advance! +Please dont use an already closed issue for discussion +I just reviewed the patch for seems most emconf updates should can be merged by SVN automatically because they are not in same context with this one +Just something for future +Without such cases we could easily leave Vertex implementations to decide whether they needed to return sorted iterators or not +Good work Vinay. +Anyway if you are still interested in the code I can send it +Verified with revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +If you cannot easily find an counterexample does it qualify for quite specificI didnt say I couldnt find one I said I didnt know +The last tests run by derbyall require a separate JVM as they test auto booting of DerbyThis is achieved without the test harness by running them as ant junit tasksTo remove them from derbyall will require those ant targets can be run against a set of jars which is currently the ant targets can only be run from a codeline +Committed on branch too. +Weird +attached is an updated patch which fixes formatting issues pointed out by Kristian +As I said before set has wider semantics than it is required for tarset lets you set an arbitrary value to aTime while in tar you just need to replicate the time of an existing fileOn the other hand setting only access time is not sufficient for tar +KiitosJust waiting for the next release +This patch provides working instructions for the code +Hi Pi yes the comment their is a bit misleading +Main code checked in should build successfully and work for new issue will be opened for the transactional support. +still passes +Im not certain the reason for this but it obviously could become a performance problem +just used this attached jar and it works. +Since we have no way to reproduce your infrastructure or error here you would have to try from the command line maven +Oleg when I deploy your app on a fresh installation of I get the NPE +you can set FINEST level to +Hi Olivierwhich url do you want to access and why +My guess is that a search engine is forming random to your site theres not much we can do from the Roller perspective except for ensure that such requests get a +I think it also applies to as well as trunk +Ok i will patch it soon +Increases code complexity +Just using this issue to test mailing list integration +A fix for this would be greatly appreciated +Had a thought +Better you zip up your test and upload that dims +Do you have a real world example where this is actually causing a problem +this patch fixes the null pointer exception +The example above is from contribsrctestqueriesclientnegative +for the patch +This shud not happen too frequentlyThats true +Im open to adding support for this as long as we can support it in C as well and as long as it is added in a way that doesnt impact the efficiency of the current homogeneous collectionsbq +So Im leaving this JIRA open until that has been addressed +If documentation about adding redundancy exists it is often found in a completely different location as reference material not instructionsOur examples encourage setting up separate Solr on the same server with the embedded zookeeper +has been reverted. +For some reason a jar Manifested by an has run time classpath issues +This method can be implemented and throw +Please anyone teach me how to deploy it on +might be to broad What if the references are equal but the resources are instantiated on each invocationCurrently this condition will pass with it wouldnt map a urlBut why do we check the resource at all? If is mapping a specific resourceReference shouldnt an equality test on the reference be sufficient +Keeping open as a reminder to backport to trunk. +reopened these issues so that I can move them to Future and remove wishes +In any event this patch allows a global numthreads for the embedded consumer +I looked for a maven repo with the new smack API and could not find one +Fix will be verified in +Im not totally satisfied a lot of boilerplate is still between the longintfloatdouble implementations but its a step in the right direction +Albert you are correct +This includes the split by time +If there is any more information please say so now or I think it will be resolved as invalid. +I did put some logic in the patch to to handle retries for +Thanks Dhruba! +Patch applied with minor modifications due to changes that occurred meanwhile. +This seems related to which would be good to fix +So to me the models can be treated as source code put into the source repository and distributed subject to the owner of the IP granting license to useIll try to get the IP part stated here clearly I have permission from my employer to contribute the models just as I had permission to contribute them to cTAKES when cTAKES distributions were hosted by Sourceforge +Thanks for the JIRA Christian +I found an easy way to canel validation just add the following to the cancel buttononclicknullYou can close this bug +see attached comments +When resolving please investigate +Hrm seems like something that we should reasonably expect +This is a nasty bug and its already fixed +There is no call in Superstep API +Reported as fixed. +Unless we change how hadoops local mode works we might have to go without a test case +bq +Please check that this fully resolves your issue +This also fails from Fuse Fabric if you have created a MQ profile container +Looks like oncomplete script is invoked outside of component scope +This issue has been resolved for over a year with no further movement +Hey Jacques +The is not up to date There are string constants that do not seem to be up to date Some variables are in camel case others use underscores The login page has a table with over cells to hold username password and submit The servlets arent just mapped to the where the sample apps will appear making the whole thing more complex less transparent I dont like servlets with tons of instance variablesLets back up and think about the best way to proceed here +Ive merged this. +Done no long necessary post Todds change Great suggestion added an UNCHECKED op category for operations Reasonable done +However this reproduces on Windows so I would prefer this to be closed as invalid not Cannot Reproduce +Seems theres nothing that we can do about this in any case since this is up to Burlaps implementation +Probably a good thing to add to the docs will write up a section on the required steps for release +would limit visibility to and somewhat alleviate this problem but we should probably consider a permission permission to enable this new printing semantics when we are running with the security manager +bq +If its the pb you need to resynchronize schedules and database by restarting dont need the database for the moment what you see in schedules screen is what you have in database +code compiles and is tested Indentations is composed of spaces and not anymore tabs +defer all issues to +Its an operations nightmare to let users to run arbitrary web UI on a gatewayThat is probably true and is great argument to get a webservice up and running so the enhanced visualizationsanalytics could all be done on the users desktopbq +We will do the next raft of upgrades to libraries post now +This was in an earlier draft of but I somehow left it out +I will leave it for experts to comment +Have just been in and fixed that +Martin I think you should always attach files instead of copypasting them into the comment buffers because JIRA alters the format +Added additional null test for the values array +This is not a Critical issue like I had set +I dont see any new issues that would need a fix in +Any other suggestions on how to add test coverage of course are welcome doesnt seem worth eg adding a getter that returns the number of calls +The above fix does not work in one special case +please wait +hi suresh are you saying that we save MB on a GB heap size? If that is the case then the savings might be too meagre compared to the code complexity of the patch +tif patch +Addressed findbugs warnings +It is not copied from Numerical Recipes but was developed independentlyNevertheless it is the only reference in English I know of that explains the mathematical background +I think you can close the issue +If you find further issues please create a new jira and link it to this one. +This is problematic if Derby is not being used to testI think removing the setting of the property will fix this but Im not sure if this nullifies the test! +also needs some fixinbq +Ive committed this. +The patch applies with patch p for me +byte array really works well + + +I like Brians idea +These values are correctly shown on but not on later releases +the latest html zip file for these functions +Let me know if not +Patch with changes to get the state in the patch includes changes to as well as is also part of the larger issue of how to deal with for though the patch allows us to avoid the issue for now +Fixed in r +All images have now been updated and a new version of the Security Guide has been pushed to the documentation stage +We just upgraded to yesterday and found the test case count is zero +defer all open issues to to shared libraries made compilation and linking much faster +Shouldnt log contain information about unsuccessful injection on static fieldmethodCould you please add your input for this item +Things match up as expected with regards to fileline. +It will be included in the release as indicated in the Details section of this ticket +To be sure that it is a failure and not a delay open a Java file make an error in it save make sure that the error decoration image appeared for the Java file in Package Explorer and then check again the image at +Right on the file Open With XML is also a similar statement on page and again on page +Now that we have a Im going to replace all uses of the class with it +Logged In YES useridI am no no longer a committer so I am unable to fixany bugs +Since this ticket is marked resolved I filed to point out some closely related problemsYes that is how I remembered this one +This patch changes the interfaces and implementations sense I will commit the with small adjustments +I would place the log in the disclaimer section at the bottom +Actually use compiler set in CC in C test Makefiles +It always returns the same connection object without blocking +This needs clarification this is a transacted session and a perssistent queue +Riding on the fact that all in the current directory of the launcher job will be implicitly part of the distributed cache of the action job we can create a workaround until is fixed +Awesome! Rock on! +But I think Alekseys patch is even better so Im happy to go with that. +Can you check this project out and run the test and check the mapping and domain model? Ill carry on looking at your one +after release +that the old behavior still occurs if you access the unwrapped collection directly instead of going via Hibernates wrapper +It should be a single left outer join with the eager back pointer filled by the initial orderline queryThoughts +Im getting CNF exceptions Ill check a few other settings in my ENV to verify exactly why +Plz someone give suggestionsYes we should fix the +and are not thread safe +Hopefully Ive committed exactly what is needed for thissecret tck project rev G rev I also disabled the apache directory modulesconfigs since Im halfway through moving to apacheds +Do you perhaps know Beyhan CALISKAN? mirror is incomplete no response on earlier commentplease reapply ifwhen things are fixedHenk Penning +we agreed shouldnt be moved +You will also need to make sure apache commons lang is available +I think you need to handle runtime and separately +However this then doesnt even allow circular references between simple bean instances with setter injection anymore which people will certainly complain about +Fixes ruby +The reason for all this convoluted logic is to implement versioning +Jirka can you add more details about what you have done here +Now that I look to fix this I am not understanding I dont need to change the update handler I need to change the update chain +The Annotation Editor can only display annotation which are in the CAS +Proposed patch +This is because the underlying streams are the same within each set of functions +The patch is attached +Oh I didnt notice those all had returns in them +In the later tests Ive unpacked them. +This change addresses the following issues Use static inner classes where instance variables from the outer class is not used Use Iterator for Map Entry to access key avalues instead of iterating over the key set and getting the values for the key from the map Use valueOf for Byte Integer Short and Long to use existing objects instead of creating new one Remove some unused variables +Yiru any update? Anything I can do to help +The problem with is that it doesnt guarantee even distribution for arbitrary keys +Hi MicheleIm not sure what to do could you please explain the steps needed to reproduce this XSRF issue +In any case if problem arises the app schema crew can reportreopen as needed +Mark Thanks for the quick fix +Thanks for the patch though again +IT added +Odd that weve not seen this before but no harm in adding this fix which checks for null before using a deserialized +Can you guys try this out to see if it solves your problems +Yep I meant thanks +Any update on this? Konstantin have you been able to confirm this is an issue +I uploaded the second patch to the review board after rebasingPlease review this patch +Looks like didnt work seeing the same exception againOrthogonal to this we should make sure RM crashes when the dispatcher gets an exception +none of the commands need it but all of my custom commands doi use spring to wire together dependencies for all of my custom components +Now that work is well underway with AS all previous community releases are +The video should be reviewed to make sure it is still up to date +Closing old issues +to trunk +The latest nightly build should fix thisproblem. +the patch is large because it will remove over lines of code +sql tests would remain in place for jdk compatible testingHowever in the mean time the test is not failing anymoreShall we close this +Patch to fix shell commands return message +Attached we check the other HTTP producers in If they have a similar issue? +The use case for this filter means that it will create more or less as many tokens as there were in the original token stream thus doubling the size of term dictionary +Active blocks are the blocks that belong to a file but blockMap may contain blocks that do not belong to any file +Closed. +added check if map contains key to prevent of branchr of trunk +I will see if we can change this +r +Please let me know your comments +This forced surefire to count the test cases correctly and was an effective workaround for my project +Thanks Kyryl! Patch committed. +Closing. +Fix malformed again against passed +That would probably make iterating on it much easierfaster +I retried this on rca and everything worked fine +David reopen if you think its still an issue. +You talk about the actual command classes can you provide a list with the classes you mean? They would help me a lot +Well take a relook at the design to see if we can make it more obvious +Eventually we may want to use a custom logger that does some out of band processing for errors and warnings but that isnt the critical goal +can we move script source to tools like sstabledebug +delimitedCase seems to be the same as schemaCase can we use the exsiting schemaCase? Actually although they use the same constants they are not the same +BobTake a look at how Apache Commons does things +Would be nice if maven would detect this and given a more precise error message +I can see this in Struts. +The elements in the fieldtype files specify the object type the JDBC driver is expecting +bq +Unfortunately I cannot get the secure storage dialog to show up no matter what I do right now +Yes agreed that the memory implication is different +It still leaks but just leaks less +is rather independent task with tests +The attached zip file contains a refactored version of the rest module +Thanks Igor! +Why does Bugzilla move me to a new bug after submitting? Ignore the attachment +Clinton good catch will fix forthwith +I attached the detail log for it +In this release they have been relocated to lib +This patch is same as Sureshs patch with a little modification to add annotation to +Our best idea here is to standardize an abstract helper class that users can extend to make this easier as passing around the really is necessary +Could you show me a conceptual example +Provided that there are no intractable licensing issues I think that this work falls inside the Derby charterbut other community members may have other opinionso Public API The existing are documented in our public api in the package +Trying to make fastpath unsynchronizedHowever it works if most of the classes are proxied while can introduce double lookup for class +Hi ColmI saw you watching this issue and I was wondering if there is some known or suspected issue in STS client or interceptors that hold on to some xml resourcesthanksregards aki +Ive also done editing of PS files +The skeleton generator is completely gone from M and is not getting called in M so I dont think this problem can recur. +Manually tested! +bq +When I do a checkout I get the white cross on a red square on some subset of the Groovy files it is not clear to me why some have the cross and others dont +this isnt a patch meant to actually be committed but just shows the debug setting Im talking about +The spec lists future enhancements which can make these types even more usefulI dont expect that I will have time to implement this feature in the timeframe +after associating a Fix Version +I think this was due to JIRA running as root on the server for a while which caused some attachments to be owned as root and not readable when JIRA was later restarted running as the correct user +Code is already fixed pending a new artifact which I can help get pushed through +Also i can leave your name as the author of the contributed classes then Cheers just signed the agreement +Unassign me +Oh wow thats a sizeable line and thats only for tasks right? I know it makes for a weird logic but could it be possible to not print the whole list again if it didnt change +We dont define a fix version until a fix is committed +is also cluster command is also affected +I could not duplicate the issue +Sessions are only created when a user is authenticated. +Ive been out of CC for a number of years Im entirely ignorant on C so I cannot comment on implementation details +First off James whats wrong with just calling abortSecondly in my opinion there is a better fix for this issue if it does have to be fixed +Come a look at in +I follow myselfwhen test program connect to Network serrver in my previous comment means when test program tries to connect to Network server +For the tx manager configuration I postpone to. +Without a jar id you should get a syntax error when declaring a procedurefunction +Taking +When you say I did the code to make sure are being and it does or else in a environment like ours which the are being what? +Patch because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Well done this has resolved the problem for me +Grant I agree it would be nice +How do we handle the native lib dependencies though? Do we ship that as well or keep a static resource at the NM? Different MR versions may end up needing different native lib sets perhaps +I moved on without thinking too much of it but it makes perfect sense now +Sorry i guess i was getting close happy +Dave +remaining +I dont claim it was the best way to do it but itdid it nonetheless +The nanodist m plugin needs to be ported to m possibly merged with booter itself +The real problem is that wsdljava invokes the JAXB plugin when generating proxies +People need to invest in designing a good search experience taking these factors as well as many other into consideration +With localization time dropdown was breaked +Manually tested! +I suspect its still lurking to bite you +Heres a patch which restores backward compatibility +I did not find situation where the button is enabled +Perhaps it would be nice to use Javas property files maybe we should integrate with the that Daniel created maybe we should use XML maybe we should do all of these +bq +Im kind of struggling to see the use case there Java serialization is so heavyweight vs say JSON which maps well to Collections +and the normalstandard API page is displayed just the same as seen in Firefox +i should have double checked +If you try with etc +Thanks Patrick +When we set the number of partitions per topic to and set the consumer to use threads we dont see this error +I think that would be useful +Also what is the relationship between the Codehaus Confluence and your? why do you have two +Especially as we are seriously encouraging people to use the Catalog Entity Resolver. +Its quite hard with all these patches around me currentlyHowever attached is the patch file which updates compress to the current ZIP implementation of Ant +from the JAR doesnt appear in the WAR project and nothing gets recognized for namespace jt +There is at most entity active for a given primary key at any time and access is synchronized within a transaction +BTW Anoop your analysis is right we just inject an to control the time to expire the lock +Everything works fine +Problem reproduces on. + good catch +Oh and by the way +Ill commit this in a few days unless somebody has an objection +This is now working as expected +Koji what is the advantage of the over + wants the opposite of that +Better to clearly show that its not being handled +We need a fix from them first +now manages begin commit and rollback of any current AT transaction wheher it is subordinate or documentation still needs updating a separate JIRA will be raised for this +Switched to qdox +For yes you are right +A project worked as expected +Renamed Parameter type to for use with Protocol +I was confused +This actually had no other affect on the behavior since I was already using a listener to check for the font whenever ther view becomes visible. + +Output of make C libcsharp the Mono compiler fails to compile due to not being included in of the fix was committed as r which was attached to +I think since our last change we have addressed all technologies in some way which works and we have no desire to change it again + +We have never seen this behavior again +I agree with you guys that the uniquefalse wont take effectBut please keep in mind that on a one the inverseColumn HAS to be unique because else it wouldnt be an one relation! Ive run into this more than once +Reading master aggregation code I need more time +sharelib is one solution but due to known issue already mentioned wed like to have option to upload to temp directory instead of sharelib +Assume closed as resolved and released +from +Add eclipse launch files and modify +Thats just how you install new adapter TYPES +The issue is in the Hibernate Dialect and it depends on the version of Oracle youre using so the custom dialect approach referenced in the comments is the workaround. +classpath and +Are there bits of the code that need to modified in order for it to run properly on hadoop or should I be able to pop them into the mapred directory and go +Ive just committed this +You can hand it over to me if youd like +Hey Keith do you still have that quickstart around save me from having to create one +Ill try to work on these +Therefore this issue is FIXEDThere is a related issue of allowing Fuseki to stream the results back to the client +Committed to trunk. +Deleted invalid results +Patch fixes the test failures reported above +Is it possible to emit the warning in a more machine readable way like +Created a copy of this issue to track it in trunk. +Is this essentially the same as +Attached Modified from +Ondrej are you able to call the equivalent method on the queuetopic via the? Does that return the expected results +Thanks all +Fixed and tested in on BSCSfbay and +I think and may be viewed as in some editors which is the correct display mode +Just take time to try to reproduce it speaking without test case showing anything we cant work on the issue +BTW in my design each task would use different HDFS Delegation Token even for one job +Bulk closing stale resolved issues +Actually to my surprise I havent seen the diff file attached to this issue it is actually pinpointing to the same code i was talking about am using grails the official windows binary version so im not running off of a built from the latest head version from SVN but i saw the code was fixed lately +Closing stale resolved issues +Issue closed with release. +There isnt a bug here the sys function assigns the timeout to and then calls the wrapper on getting a call with a timeout of the wrapper version calls the libC function +If someone else has time to regenerate the patch against a more recent source revision that will speed up trying to add the additional tests +Thanks N +Reduced priority of non critical issues which have current issues that arent going to make it into to +janandith +Roberts solution includes a whitelist approach +Alright now this description would seem that got truncated +The attached can serve as an example +Rebuilt fine the second time +the variant has more than one associated virtual +Ok thanks no pressure and no rush whenever is appropriate. +Thanks Jay! +However excel did not. +I will try this out tomorrow +The attribute should actually be + +same patch with new tests but it modified in order to replace real hostname to localhostI ran the tests locally the patch worked fineWill convert the tests with aop in for trunk +In my case the solution was to remove older asm +After looking at my patch I found a flaw in it in case endpc points to some big address after the code end +Probably one or two more clarifying cycles with them would produce the correct result +I believe we fixed this in trunk by saving to an fsimageckpt dir and then moving it into place atomically once all the files are on disk +Cleared patch available flag as this would need to be reworked once we agree on general error handling for would like to vote for the half of the current patchThe changes in i would omit but he changes in are very useful i think +I wanted to note however that is such a new class +Once the existence of a har file have been checked processes running concurrentlycan create files +It is never the same I ran it on LINUXWindows +One remaining issue might be a stale descriptor returned from +A patch was added to the maven issue +Generics in is not yet supported +Appreciate your attention to this matterthankssudha +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +I dont see any obvious problems with the logicI wonder if you can debug look at the cooccurrence matrix andor user vectors? make sure neither has negative valuesIm trying to imagine wild scenarios like overflowing an int in the cooccurrence matrix? but you dont have billions of users +Added myid file creation +Testing with a different application +had some inconsistencies +Sure hash values can be anything +That would be easier at least for me +It does not belong in the cluster status set +That is to say an emptynode one without the +The overhead would be huge if every split in routing caused a clone of the underlying message +revision fixed +nope works for me in last nightly build from trunk +Thanks Gunther! +For the stuff do you know of any sample plugin? Especially which dependencies to import there seems to be an old version of the context and the new tesla ve committed a first version which uses the the basics seem to work ve tried the latest trunk version of both the plugin and the scr annotations for a simple bundle and the CLI build works as expected including the interaction with the for adding the header +The original patch should have been rejected because it was not +Thanks Owen +I added some code to catch the exceptions and as unchecked exceptions with a hopefully helpful error message +Any news on this one? Im using hibernate +Dependency on itext is obsoleteIt was only used for test at the beginning of the project to check some weird behaviours with pdfbox we wanted not to generate and test PDF with the same product +I dont like tokenfilters that retokenize by changing up the positions i think it causes all kinds of havocInstead with this patch it simplifies what this filter is doing conceptually for each word all of its prefixes are added as synonyms +Is this maybe caused by a race condition where the NM is receiving a container token before the RM registration completes and it receives the secret keys for the container tokens +For this reason JBPAPP can be closed only by QA +Youll always have to do the REINITIATEbq +patch is added and libs updated to velocity +The Searcher supports the whole index but a given was never promised to do so +But why have you added test test? Is this a part of the bugfix? Or at least has some relation to above problem +Should we just ignore the partitioner after first startup the way we do with initialtoken +Reopening +DoneNote that as you provided your apache mail for moderator that is the address you must use when performing moderator actionsThanks GavinMilamber +please let us know if this issue was resolved as well or if there are other issues +could you provide a patch please + fixed +I try to copy from to trunk however it depends on other classes which will break build +HiThanks for the fix provided for this bug it is working fine with existing data setActually I need to this to version and ofI was wondering if it is possible for you to provide the information aboutcode changes that went in to fix thisAny help will be greatly appreciatedThanx again for the resolution providedSandeep Desale +Yes this is still required +Added a test case that exhibits the problemReferencing the constants in a static way to make it clearerWill commit to branch in h if nobody complains and then port it to CVS HEAD +Thank youfor your help +Thanks I think with some mods will do what I need. +In dev mode in still works but in prod mode only message codes are system Ubuntu Karmic To reproducegrails testappcd testappgrails domaintestapp Create name propertygrails controllerstestapp scaffold controller and viewsgrails httplocalhosttestapppersonlist displays Person list as title Hbut grails prod httplocalhosttestapppersonlist displays as title patch fixes the issue by using a instead of a which doesnt know how to deal with unpacked wars +extends +This mapping can be different for different triggers on the same table and I didnt want to mess up if it is being shared by all the triggers on the table which are going to get fired +Good newsVersion would be out around second week of December I guess +is the greclipse code that uses the new patches +We will address this when the time comes and the need to support union data types +Scheduling for for now +Theres no link between problems for now +qt +As I mentioned in the original issue we will not be using maven for future releases +Hi that your point is we should use the syntax this +Attaching patch herewith +Give it a try and see if it solves this problem +Rejected +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Hi Lincolnsorry to have had to put this off +Yes I know that with Jetty and Tomcat works fine but Im using Cloudbees and its working with +It seemed as it needs these files to run +We can only remove the duplicated system requirements page on the web once this is fixed in all releases +However I had already done the work and have just committed most of itThe core logic was mostly there and I had to just move the code around and addmodify as neededWhats remaining to be done is +Agreed its home is better in JDT but probably wont get there without our helpAkazakov how easyhow long time would this be +No we cannot fix that +Messages would not show up and no errors were displayedNot sure whats wrong with my blog you should be able to comment +Reclassifying this as a bug since supporting http headers is a part of the open social makeRequest spec and is intended to have the same basic interface +Thanks Carl for the work +Applied patch to subversion trunk as revision catchI wonder if the initialization of the binding can be done in doStart +Any chance we can just get a version that just has the update to the xstream versionI forced an update to in my superpom an it fixed the issue +It looks like the query works fine and retreives data but in debug it looks like its trying to get value of key column with alias although I have configured a return property +I fixed making the reentrant +It would better to divide it into several subtasks +Im able to reproduce this now! Yay! I think part of the problem was I was testing with +thanks zheng +For this reason let Continuum show differences in files between current and previous for the changes display users that did them +Verified in +Good idea Kim +Either starting a new thread or causing the same one to start and stop seems like the best fix for +The action is Getting Started which closes Welcome and opens Central instead + +Im attach another patchThis is a quite tricky solution but if this works on more large cluster Ill committed this +NathanI totally agree about maintainability versus performance particularly when the performance difference in negligibleNote that the method replaceChar is completely unused +Mamta just backported the changes of revision so I assume these changes now need to be backported too +Ive run into this bug before and never found out a reason why +For the patch Ive run local tests and all the testcases passed +However I forgot the regression testing process +But while running Geronimo in JDK they failedFrom the codes the author expects that the is thrown for there are some codes to check whether the String arrays length is zero but while the string is an empty string an array is always returned +Tests passed for me +Closing stale resolved issues +Im not familiar enough with this to be sure about the appropriate way to backport toStrong if it really should be backported to please create a new issue for the backport so it can be fixed inThanksGail +Now I am able to import example mentioned on +Why doesnt finalizeLog need to end? Isnt there a test that asserts all finalized logs end with OPENDLOGSEGMENT +Attach new patch +I believe if you treated roots more like a root classpath and only loaded scripts from their specified class path locations in the specified directory for the script containing the class rather than scanning all locations it would make the GSE a lot more yes parseClass needs to come up with better dependencies quite easily done i expect I might try myself +update patches to include license in for the review Eli! Ive committed this. +Bulk closing issues resolved in JEXL. +As its provided with I dont know if it should be added +So if we change it there we have to put it into backwards break section and apply the attached patch to backwards +I tried this on my machine and my dependencies are in the lib directory inside the tried it both on and of maven and both worked using of the assembly plugin +Shouldnt there be some log message about that? Perhaps it would be even better if there was a message about each bean where the bean definition comes from. +I tried the code provided with and snapshots but I couldnt reproduce the issue +Release notes updated Ready to release +Last analysis is is already the case when the build has happened during the day as in such case the time is displayed instead of the date. +Please let us know if this is still a problem with the latest release and we can reopen it. +Great! Thanks for testing against the CVS buildCheersLes +Only error I see is a permgen limit reached on +Can you post your testcase to reproduce the problem? There are a lot of extension fixes recently +As we saw the bad rendering we have stopped the test to avoid losing time We just need to know how to construct packed resources by hand for css +Good change. +Committed revision +This only includes changes I want to commit now +Thanks +Meaningful words in figures titles start with uppercase letter +Clean install went smooth cannot reproduce +FWIW +on patch +Rejected +Committed to as well +Perhaps +No integration tests +Committed trunk revision Committed revision +I believe this patch is for is that what you meant +The planar projections should be completed by next weekend +Possibly about thisWe place the EMC code in static initializers on classes following a convention so that referencing the class leads to running the EMC code +Normally we prefer to have event classes so this should not be removed even if the diagnostics is pushed up into the based class +The admin replied with the following so Ill resolveclose this issue as Fixed for nowYes I filtered out wgets a little while ago because of some serious trawling I had going on +I believe solving that issue may alleviate this issue +What if theuser as Tomcat being used for something else +Well each tool should have its own lib dir that would solve the problem +Need to revisit this issue from perspective +apbackup jackrabbit Apr apbackup jakarta Nov apbackup james Oct apbackup java Jul apbackup jmeter Nov Shoulnt be the same +Thanks Rick! Committed revision +It is probably a bug in but that is something we support +receiving a SIGKILL implies that something has gone wrong in the processHow are speculatively executed tasks terminated +splits into and into and to avoid cyclic dependencypython and interop tests passed +It seems we need to not just count number of rejections but actually the unique +linking to as at least some of the growth in the mailjdbc test is related to this issue +it requires to create Handler and Command for +Am I misunderstanding +Maybe we should just go ahead and start in a module so it is easier to modify in the future +Go ahead and apply Im not fully operational with SVN yet +Thanks Kim +Once I moved the injection of manifest data from the shade plugin to the bundle plugin the manifest file is now the first file in the jar drwoodsopenjpa unzip l target Archive target Length Date Time Name +in a separate branchRight now there are no tests for these so any change by any contributor cannot break anything in these hooks +I was just hoping vainly that somehow the code would expose whatevers been happening on Jenkins based solely on similarity of error messages +Not sure I see the corelation to mentioned in the link +Gmf has some nice looking features and better model structered by emf but I did not see a sence to convert Diagram Viewer to gmf use +Not goodSeems to me it would be a lot safer to require that the database exists and fail otherwise +Patch checked into branch. +Tx +committed +Is it still something well be looking at doing in or can it be pushed back to +Force inline of pTLSvmthread and tmnsuspentenabledisable on Linux +Otherwise please commit. +committed. +This patch reverts it since merging will be difficult later on +Username swolfangel +fixed the space issues and resubmitted +Please if you can recreate the problem in builds r +Patch for +also looks related the change for that rev +Look at log starting line INFO Working on standalone example the issue is with serializing the fault exception +Please have a look at it and tell me what you likedislike and if there are other useful methods that can be put in there +this was resolved by the ISPN JIRA fix +I would like to make one more suggestion through and that is to go ahead and use an interface at the back end of the process Interface Existing code Deprecated Method nameDelay adding a forOtherwise for we might go through this renaming againwhenif other schemes are implementedBy doing an interface now we reserve the name forthe concept it represents and not the implementation +This is a distributed database benchmark +Adding a patch that illustrates the errorExamplehadoop jar randomwriter issue exists both using the default container executor and the linux container executor +Is it possible to make couchspawnkillable watch for the parent process death without introducing a new process into the mix? The nix variant is a simple shell script that execs and Id be a tad remiss to make it more complicated +What this normally means is that we cannot use oids in postgis since they change due to practice has a the FID tablename have any thoughts on this? We are forced to have FID as a string but I would not mind the strong typing +This type of failure is problematic for jdk since test order is random +The overhead is array lookup a delset look up vs +CORS will be shipped off by default +attaching patch here +However when exactly this merging and sending happens is in the it happens on every query but I hope the API can support other scenarios as well +No integration test +Not a good idea at allFor the moment boo could probably just define some standard complex type but keep everything modular enough that it could be easily replaced by a more standard version in +if the context was a chained context the parent key could suddendly show ugh +I hope you are all as involved in their JIRA as you were in mineThanks for all the feedback and involvement everyone +Okay got something running now +In the meantime I have removed the reference to Nexus from the and the install seems to go file if you dont have references to Nexus in your +If we include a qualifier and Any then it should act like a partial match and Observes Any Type t should handle any type fired +Is there more research that is still needed +OF course end users may not care about the complexity of the accumulo source code +I liked v better +Could you resolve these please and resubmit the patch? Thanks Kelvin +Thanks for the patch Elben +Assuming its a duplicate and closing +Turns out we are doing things different somehow currently and dont need that commit at all any more to update our commit points +Reopening for handling. +that seems more like what Im looking at +Wise doesnt work on AS because it tries to construct its own classloader for wsconsume and it does this based on the assumption that its running in an AS container +On further investigation the file system has been corrupted so this is not a derby issue though perhaps the null reference could be checked and a more meaningful error message could be thrown +looks good will commit in trunk and if tests in and trunk +Done. +Thanks for the help. +Can still get locality by spawning it on the local machine +But if you approach from the view point of owners of existing hardware that was speced to hold certain size of namespace it can be viewed as a decrease of capacity +Moved to next milestone release +Rsyncis currently going through the entireapache tree +Since my data is broken out by columns if I know the accessed fields I can only load the data necessary for those fields which will be a huge speedup +Not committable +When that has been done we will close the with a fix +There are new modes supported posix and tell me what I need to do next so that my changes will be integratedinto SVN +Niklas Im nobody official just another meclipse user but perhaps you should try the release that was just released over the weekend +But I run into the same issue asThanks Lou for verifying that this is no longer occurs +Thanks Roman +I added one more commitnow only linked identity hash map is provided by so we can find a better way to write a concurrent modification safe linked identity hash map to replace ve added two benchmark results to compare to see why this patch was worth it. +attached solution just need a review before committing SimoneI applied the patch and it looks good to me please go aheadI would suggest to keep this issue open for further changes to weight operations that were proposed in ML maybe renaming it to simplify weight modelCiao ClaudioThanks for the feedbacks anyway I think it worths analyzing more deeply how to disambiguate cases in wich our concrete implementations have the need to implement more than Monoid at the same time before applying the patchI hijacked specific discussion here because in the ML we were speaking about different topics having a spike to speak about should drive us in the right directiongoing offline in a while until tonight best and have a nice WE ClaudioI still dont understand how the rename simplifies the actual scenario design I suppose that users with the need to calculate a Strongly Connected Component on their graph data are enough educated that should understand basic Algebra if even a + +Patch applied +That and the test stuff in your patch should be the bulk of the usage in the testsuite may be some odd thing elsewhere +But if we actually fix the logger to use the correct class name we should also change logging to only happen at trace level instead of debug so that existing setups with debug level enabled for the parent dont start getting noisy messages +Not sure too many people would intuitively link the two +Thanks Christian that explains it +Why dont we rely on the regexp urlfilter for removing such URL? Is there a performance issue +This is completely wrong +small changes were made as per freds critique +for the log format +Just for fun I add another test case to prove you that all is already fixed in the trunk +Best regards George +trunk Added better startup route error all resolved tickets from or older +And done. +I added a new timeout command line switch that supports setting it to for testing nonblocking servers +Rename to and add it to the debian packaging +Duplicate issue +This would improve the iterative testing process +Test cases included +Sams patch for this issue +I remembered there was some issues in that test but forgot to change it back +The data I was using never entered that part of the Ive used the direct access to the feature namespace as a first attempt for a couple of reasons it is quite a bit faster if you have many properties it better matches the interntions of the user if you are looking for name unqualified in a feature youre probably looking for the name attribute in the feature namespace not the one in the gml The example I am working on does enter that part of the code so something about my case is different from yours which is I guess why my map is still not rendering +Hope no puppy was harmed eitherP +Closing. +I think its ready to commit once is in +The waits and retries if the file is already being created +Closed after release. +Patch committed to both trunk. +Drop down window could be more efficient than separate window +I think they could have made the pencil a bit smaller and lighter colouredOr perhaps an invisible edit button! +attaching screenshot of deployment portlet shown providing an upgraded with configId moduleId change and removed +cc versioncc Sun C Patch CC versionCC Sun C Patch +I just want to make sure you still see the problem at that for letting us know +Patch essentially amounts to a rewrite of get on sorry! Ihavent profiled this but the additional recursion now required might makeconstructing a URI every time just to convert it back into a string a bit of anoverhead an internal method that returns a URI might be better if this turnsout to be a problem +Safety nets are good +should always generate flattened columns +Fixed. +Hi DineshNo +Can you post your Java code here +As it stands this patch allows open access to methods that have had privileges granted we need to move the privilege granter to be package private and provided by the caller to Reflections +Check you poms deps to see if something is being pulled in from one of your own dependencies +from since review keeps stalling +this one is about including into hibernate plugin +Closing resolved issues. +Mass closing all issues that have been in resolved state for more than one month without further comments +in part has been ported to a more generic +this is intended to be allowed in a static context and referring to the +patch looks good +I am working on fixing the existing tests for which now should work because of yesterdays fix +bq +DIGY has the patch committed +If we would switch to Java it would not be very difficult to replace Xerces +Sorry about the name confusion btw +Seems an NSRE on first next invocation could cause this +fixed active TLD declaration +I must tune up Jenkins configuration +The best change to get it incorporated sooner is by contributing it to the project +Im ok with that suggestion Although it may seem that using quorums of size with replicas may cause replicas to diverge it is simple exercise to verify that all quorums constructed hierarchically in the way proposed are guaranteed to intersect +Committed to trunk and Patch for issue Oozie is released +I was wrong about the theory that privExists throws a security exception +OlegCommitted the fix and updated the release notes +This is a costly operation +This should fix the test issues for MR tests +With the legacy database that I am working with lack of this feature is driving me to use hbms instead of annotations +mind addressing Larss comment above lads +This requires more tests so move to +Updated reviewboard against branch origin RB is After looking into it I realize that forcing a rebase would be a bit tricky so instead I add a check on the branch hash to see if it has changed since the current working copy has generated +I just committed this thanks Jerome. +Thanks Devaraj! +In this output is shown for in the given pdf. +Without this we cannot populate the table created using the in HiveRegards hardware was gobbled up recently so I havent tested the patch yet +Actually its similar but rather seems to be a JMX issue +Kev Great to have positive feedback from GSS +unit test is the unit test code actually i think this must be done in an integration test because the broker and the client code must run on different computer +Looking at this now with new eyes after years Id say the error message is perfectly acceptable +As we are use CXF in the camel trunk I just committed the patch which upgrade the Jetty version to. +Marking Patch Available to get it into the queue +You should check with him about the status +I guess though that I cant see how wrapping it into an unclosable stream would lead to more memory leaks than you would have should Piccolo not close the streamAlso it feels to me that in some sense Tomcat could take care that a close does the right thing so that the fact that the streams are in reality shared is transparent to programmersWould you feel better if had an option to wrap any streams that it uses in unclosable streams +If you dont have it you will go to dead datanodes during the split and this will lead you to a bad MTTR anyway +If you find further issues please create a new jira and link it to this one. +Can you provide additional details about your environment and jstacks of tservers datanodes and namenodes +Hi LohitDo you have any steps I could try to reproduce this +Id like to apply this patch but cant get it to apply +I still have ignored commands see comment Ive posted Aug PM +If I redirect to a page with a missing required page parameter I get a +The patch was applied on OJB by Thomas DudziakI am therefore closing this issue. +Also INSERTINTO falls thru into DESTINATION not sure if that is intendedhavent tested yet +The other place I was thinking would be the implementation in do the check there +setting was easier than I expectedThe setting is now and defaults to on +Hello PaulexThe fix looks good +Those are the only two values that I could imagine would have an effect like thisIn the mean time I assume its okay if we call this resolvedThanks again for taking the time to report! Your help in weeding out bugs is much appreciated +was fixed by Bowen a while ago verified that benchmark compiles with xc inlining successfully with release +Fixed the +I now encountered the many child nodes issue raised by Thomas +Am gonna mark this one as fixed as I think the optimiseAcknowledge being disabled in will further reduce your RAM footprint +I added the workaround script provided as into the root folder +Why cant we just let the reject jobs that are submitted to invalid queues? That is what is done when the situation happens with enabled +Attaching the UI screenshot API server logs and Management server logs +Ive made the always rethrow the exception no issues heres a bug always calls skip on both reader and writer +BTW are we mentioningdocumenting any where that product has only arabic translation support not the UI just to educate the customer +If you want this one then yes please do +bulk defer minortrivial issues to +As I can see your is not valid meaning you have some unrecognized characters there +From what I can tell so far the same method is being used for both the splitter and whenfilter +Removed extraneous +The configuration file has an error in the definition of the CSV sequencer and is using the wrong class name +should be complete at this point +Put in langcsharplib works for me on Ubuntu +Huh I just double checked and theres almost an argument for this being a bug though its kinda conflicted with the other thing about any valid unicode character +Same file without compiled classes +Marking for bug should probably move to Github. +in the wildcard address doesnt mean anything +Author m being ignored +This patch should fix the errors reported by Erwan and also an error reported by Jacopo when using OracleI still need to use the new fieldtypes +You should use the trunk lots of things have been improved +This version of the patch allows you to declare enum constants by their given numeric value +Ive used code from +tests +Updated name and instance states arepending user asked to start but not yet passed request to deltacloudstarting instance starting up in cloudrunning instance is up and runningstopping user asked to stop but not yet passed request to deltacloudterminating instance terminating in cloudstopped instance since fixes are in released code +for type is only tag types +I think you need to add an epoch number which is separate from the txid to distinguish different startings of the same segmentThis is what I was thinking when roll is called from NN it includes the version from journalList ZK node +Thanks you Kay Kay. +Third parties can make installers for example you can get one from Mergere +ok Jody confirmed on the mailing list +installed it three days also uninstalled the version with Manage Configuration +a region server is able to append and even roll logs! I think should also do something like that +However this method is racy throughout if the path gets deleted permissions change etc itll fail to get further listings +how could i miss this simple onehowever i chose to use private methods inside controllers +Phrases happen infrequently relative to the number of overall tokens so it is much faster to just buffer those few tokens that are part of the phrase and tee them to your phrases field +reopening as it isnt fixed until the patch is checked in +Additionally versions of svn prior to cannot check out working copies using revisions below about R in our existing repository which is a m attaching the conversion scripts as a tarball +Yes this was resolved in appears to be the same issue as. +Please review +I dont think its really such an issue if users CAN select java with eap runtime +I bet if I made the property required wed get the correct behavior. +Therefore Im also favoring the proxy approach +they are a pain in the arse to work with +Patch applied cleanly I see the service triangles will experiment a bit more tomorrow so have not committed it yet +The merge job will check if the is supported +We can consider using that instead of sync set and copy on read when notifying listeners +Closing this task +can this patch merged to the branch as well +How do we divide the content between the wiki and the website? Do we even need the wiki? Do we need versioned releases of the website as with the jars +Also Ill keep TBD +New to this project +Patch adds these methods to and +If this behavior needs to change a new issue should be filed +Agreed +That worked +Were a small team and were working on future development +tableinfo for all your tablesLet me go over the log to see if there is some thing else that related to meta migration +Its only possible to get the wrong interface by looking at the interfaces inherited +I can have verified that this file is no longer in the +Mark +Hi Vishal Thanks for all the information +Thanks +resubmitting the patch with license to AFS +Committed to trunk. +the only way i can do this today is by subclassing and and even to achieve that i had to make a few changes toin general it is always possible for the resource to have additional information besides the backing node that the importexport handlers need to do their job correctly +I prefer to call it Hierarchical as your suggestion it fist pickup a sequential znode from a hierarchical znodes in way then do id generation using this sequential znode +will commit if tests pass +and pushed both upstream +fixed rev thanks Olivier! +Hope this isnt very time critical +Hiranya you may want to assign yourself to this issue +I did some verification on the broker and the consumer side +The log size at the end of a hour run was about +Patch to add the functionality +committed. +This results in reading tons of extra log data +Try adding all the jars from axis dist in your classpaththanks Putting the in the classpath has solved the probleme for meHope this helps in current SVN +Hmm +Im using WFS is not clear to me the relationship between the and branches so perhaps this fix is just not yet promotedI only upgraded from to to get around this issue and I could not find a newer than +Added example of the new directory with the integration testsExample can run in servicemixInstallation example in servicemixfeaturesaddUrl mvnxmlfeaturesfeaturesinstall +Hi EmmanuelI observed that the release duration between to was months Anyway i will have to wait as I am not success in building the snapshot the building process failed because of testcase failedThanks for the help once again +In one of our environments on iSeries we hit a problem with the JVM where the results of the werent always getting processed correctly +You can extend class and instantiate the with that class to create this effectI would still prefer a static method for this +We can open new issue if this still an issue +I just committed this to trunk and +the NN cannot know that the dont have data on disk yet unless we add additional NN coordination? our choice in this condition is to return or let the exception be +Its a nice to have +This way the original timeouts stay there but anyone can scale themup for their particular environment +This information has no business in the pom +Adding the patchThis enables to get partnerlink details by providing the partnerlink name as well as to get process details +Updated tool to cope with multiple Continuum databases in r and also added documentation to the site in r +Thanks Gunther +Correct +no test case provided to reproduce what the user is seeing +Here is the unit test for this unit testApache header and package name has been added +This issue will keep track of junit tests that need to be added once the replication junit suite is committed +I just tried it +Could you take a look +cache entries never expire +I believe what youre seeing here is not that our encodings arent getting set up right but that our ARGV encodings are always defaulting to +If however I change the dependencys version to a version range Maven no longer resolves the artifact +Reading the link provided in the description requires a payment of US +Considering the release timeline +Im not sure what you wantDo you want to be able to disable the building of any JS component at all? WhyWhat has that to do with HTTP? There is no way to use without using HTTP at the emoment +Alex are you still planning to work on this +Since it is supported to redeploy ESB applications then IMHO jBPM should be able to deal with it +Mark incomplete task as obsolete before leaving it after expiration or external signal. +Committed to JBESBGACP revision have shortened the package name of the class generated for the WS +This is not a priority issue and we are very busy right now so dont expect it to be fixed in a hurry +Does anybody have any ideas on how to get these passing in the automated test system +Move to CR +So if the remote server does not support EHLO it also does not support BITMIME by default +Now when I seen the trunk code to apply the fix it is already fixed +Missed adding some new files +Fixed and the test at revision . +only works against trunk + +I do not have the time to repsond indepth now but I never had the intention of detailed validation +Yes we could probably do this for generated code but we still need Paranamer for reflected code +Updated patch adds some more cases R for chgrp and chown nonexistent files in arguments glob etc +Finished consolidation on the frontend peices + from me +This screen shot shows Jconsole displaying the rpc statistics for the name nodeNote the tree on the left it shows that name node stats and rpc stats for the name node are availableAny of the values can graphed by clicking on the value +I guess the memberOf thing could be handled by this Improvement. +The lowercase a is growing on me +Submitting patch for Hive QA to pick up +Done and done. +Run a single qfile test +It is a bit of work making it so replication does Cells rather than serialized +On region move the region is opened again with a greater seqId somewhere else and the client can reason about cache invalidation +As a workaround a check will be added to suppress this particular in the case we have just closed the +No need for IT. +Would anyone want to try this +Updated patch based on review + was just modified to add support for patterns to be added to the MDC Structured Data element +I hope it will be considered for the final release though +Im still waiting for comment on the latest patch +I then copied the file upload code to the source and rebuilt struts and tested my application +extends +Running it as is will result in the +Attaching patch +Hi LoriCan you please attach the service archive file it will be easier for us to find out the problem and fix that +It would be great if we could see a resolution on this one as I think this is exactly why a lot of us use maven +Can you tell anything more about the nature of your data or provide a sample? How much input for the FST to exceed GB? Perhaps there is something we can do to optimize the input to make the automaton smaller? The rule of thumb is try to get as many shared prefixes and suffixes as possible because these compress nearly ideally +GC related changes can be droped if nobody cares about anymore +It the same in sample with richvalidator facescomponentsrichValidator for the sake of completeness that this works in mojarra +Configuration keys will still only be initialized if either HA or federation is enabled in the conf but will always be set regardless +I like it +in my case its comparing the wrong propertiesive applied this patch and it fixes the problem for me +Oracle documentation says The default heap size for all JSE implementations is MB +Hello Morten thank you for providing a patchI will review and integrate the patch in M +Version has been releasedMoving all related issues to the next version +I have client side code explicitly working around this bug would be nice with a fix in move m submitting a patch that implements commitWithin on deletes +To Max Well not passing unsafe argument to is common sense +Nice sleuthing you nailed the problemAnyway Ill check in the fixes for this morning and this will be fixed +This one looks related to +The managed connection and the xa resource have two different purposes but need to be in sync +Request for reviewThis patch changes overwrite to be a generic options arg anddefines CACHEWRITEOPTOVERWRITE as well asCACHEWRITEOPTCLOSECOMPLETE andCACHEWRITEOPTSYNC +Ill see if I can track down Greg +Thanks again Kristian +This fix has been verified by the QS. +Please reopen if it is +We dont have enough to discuss everything clearly last week because the customers requirement is urgent +regardless of whether the access is simultaneous or not if one thread sets a field another thread may not see the same value in the absence of synchRe Config I thought the fields were used as constants but on looking again they are used as defaults so please ignore that comment +See +one thought there is a shutdown attribute on a broker such that a controlled shutdown of the master can be triggered by a slave failure or shutdown +Ok other issue +I dont see this as a bug its an improvement +Thanks Ashish +So I think that the right way to solve this problem is to decouple the query building logic from create a generic default impl that does the same thing as the current code and then create Oracleetc +Thanks +Can we remove those data properties from the editor for now until we fix it properly +It shouldnt affect any applications that dont to using it and applications that do should see an order of magnitude performance improvement for MLT queriesThis cache implementation is tied to the MLT object but can be cleared on demand +We are currently using Wekas that supports most of the above params +I cant get this to work either why nevermind? +The bits xcace in your session id correspond to the time of xcace which is at PM ESTThe bits b come from the myid file +See my patch and comment in the client error is caused by the server ignoring the wantReply false in the request message +Yeah that would seem to fix the problem + +Fixed the problem in Dictionay class when key does not exist +This is not feasible for +Thanks for investigations Juraj we will wait for clarification of then. +Maybe a clean rebuild can solve the problem +to update Ive found that zkpython tests are failing in trunk and I dont want to commit a patch when the tests are broken +Ive not yet succeeded in reproducing the original nasty result +Thus in the situation above where the only log we have is this corrupted one it will refuse to let the NN start with a nice message explaining that the logs starting at this txid are corrupt with no txns +And whats this setter method +Laura can you tell me which svn revision from the repository you used when taking the documentation to convert to EAP documentationWe will need to track this so we can do svn diffs to see any changes in the docs that need to be also applied to the EAP docs +Magnificent for commit +And I was concerned that the current patch set seems to have a different path near the root using srcjava not javasrc +For mapside cogroup or mapside group by though the splits can be combined because the splits are only required to contain the all duplicate keys per instance and combination of splits will still preserve that invariantYou are correct for mapside group but not mapside cogroup +Some producers are expensive to create such as FTP JMS etc +Argh broken link for me Charles +See the +Mina version +Reporters of these issues please verify the fixes and help close these with latest build +Also it would be nice to see this project in Sonar Ecosystem view +This patch starts a thread to report progress during the merge process +Attached v is a patch for trunk i merged it manually lets see what QA robot will say +I was just buried in the code at the time and having stared at your code that ignores the field for each Term I was not thinking from a high level but was instead stuck on the process of ignoring fields in a similar manner + +So Im a bit puzzled as to which part of this puzzle is messing up the +Revert the jar and live with the bug inBecause this bug was already present in it can live in as wellThose who need this bug fixed can easily upgrade this library on their own +Edwin I dont understand your comment +Heres what I did checked out the the jbossidecoreplugins module from pserveranonymouscvsrootjboss CVS imported the project into my Eclipse workspace Fixed all the compile errors mainly replaced with Hope this helps +Closing issue Oozie is released +Here is the revised rt build script which no longer uses apxs to install the module +looks alright to me +For fullAccess users your argumentthat the permission to +See comment on test results couldnt get it like youre missing +Yes +But because of the standard it was never finished +Work completed waiting to establish baseline correctness for C before enabling GC by default +Here comes the patch for hadoop append branchJonathan if hbase needs this for HDFS trunk I will make one for trunk too +to a revert without providing an alternative +Closed for +From a Spring Data usage point of view we probably wont refer to core Spring unless its in an RC state at least +This was in trunk section in trunk and section in +Right now I dynamically add all the keys of props to the ruby class as instance methods so they are accessible from the template +Therefore in the case of webservices support not being able to set the on http responses effectively means we cant guarantee interoperability since many clients will refuse to process responses that dont have indicate an appropriate +Otherwise theres no reason to change versions +Is this alright? +Sorry for the confusionWe can add source for SQL script invocation +It might be worth doing some testing to see how much this actually occurs in practice +Reopening this +to trunk thanks for review stack. +For information to any viewer if you need to activate stageVideo in your project wether you use OSMF directly or the component you need to call true at Application container creationComple event add backgroundAlpha to Application container use a custom skin for component in which you delete the black background Rect that is under with comment background when the videoDisplay doesnt fill its whole spot add wmodetrue in your app html wrapper to get stageVideo in non fullscreen state also + need for functional need for IT +John +defer all issues to +Preserve +Done for both Java and Ant. +Thanks Peter Mahoney! +Because it may lead to other risksPlease see the patch that limit transmission speed and compress transmissionNext I will contribute other patch that check fsimage and when download +At a glance following files contains changesri might have missed note the differences in presence of certain filesOnly in jrubylibruby in jrubylibruby bigdecimalOnly in jrubylibruby dlOnly in jrubylibruby in jrubylibruby in jrubylibruby ioOnly in jrubylibruby in jrubylibruby in jrubylibrubyrdocri in jrubylibruby in lib in lib in jrubylibruby winOnly in lib yamlOnly in lib files first attempt against filled in Date Id and Revision values to reflect which files are used +Ive done some more research and from what I can see is the only place in the FOP source where is used +I do that and I have seen others do it too and I understand that as a guideline. +I could only recommend that you use a clean configuration and rebuild it from you could find out what files are different from a clean configuration and the one you have now and use a bit of thinking to fill in the that helps +Heres a new revision of the patch which addresses my concerns aboveIs there a way to write a unit test for this? We should have a test to prevent another regression +I think there a two things to do as a workaround set the changing pattern on your resolver to changingPattern +using the matchall ve committed a variation on Ryans latest patch that changes from the matchalltrue syntax to a syntax adds documentaiton about the new param demonstrates the new param in the example +By making the method protected we are allowing people to create their own transaction implementations which could provide updates on a more dynamic and basis +verified in JBT +Lets resolve this as duplicated. +for the patchI verified this fixes the double error message +Patch applied +The title should read Castor isnt correctly validating XML against XSD schema upon mapping +Hence thrwing NPEI think we should block the active initialization until it completes standby initialization +If someone has not already started on working on this I would like to work on making WSSJ bundle +You should no longer receive any build errors +Or we could even commit it and then clean up and improve docs as a followup since lets be honest its not that clean as it stands now Committing this quicker rather than slower will also save a lot of time +The shortcuts are now displayed with proper characters and all shortcuts are now created correctly +I currently have autoconf on OS X and that works fine +To my mind plugin proposed in is good for image preview +the issue still present in GA. +Of course TomIn the following hours I will regenerate the patch and test itBy the way we are using this feature and it seems to be effective for usRegarding the installconfigure scripts tackled separately +bq +Now we return the default value +I didnt noticed it because I first thought it was the current time +has this also changed +No modification for Felix or the project structure is neededIt is possible to create a Felix target platform and use that in the launch configurationFelix will use the exploded bundles being developed and tested everything works the same as in equinox target platformsIf anyone is interested I am happy with sharing the source +Good to know +Simple one applies clean passes tests and specs +Is it possible to get the OIs name at that point and only do the copy if we are working with the bad IO +NATIVEcredentialsDBLOCALThis setting means Use credentialsDB to authenticate operations +Initial patch draft +No jokes aside I think these classes seriously need some good refactoring +Committed revision +Also added unit tests for +thriu conventionally we do use void to exchange a variable length buffer of bytes eg mallocmemcpyreadwritemmap So by using uintt we are forcing people to cast from void to uintt +Noted thanks Martin +avro suffix? The best approach imho is to admit any file suffix and read the magic header instead to detect if itsan actual avro file +Any progress on this issue? Its just a matter of adding a few lines to and doesnt hurt anybody +I am not sure this is the issue if there is one? it works vs the stable supported version of Hadoop so Im going to close but if we can identify a change that makes it work in both cases thats cool +Tom I thought we fixed this one +release to get this feature in to ensure we spend enough time stabilizing the core of the RM right nowDoes that make sense? Thanks +the reason back then was to avoid having to update it everytime but now you have it I guess it is fine +Atomicity you are referring to in your previous comment is when users want either all of the tables are changed or none at all +Committed for in rev +Thanks Stefano i did patch the eventing source acccordingly +made obsolete by the switch to hg +Thanks Willem +I would like to find out why we get this difference and not an exactly identical model +It does mean that the Splashscreen plugin MUST be used with which is a pretty nasty dependency +This was resolved in sf prior to apache snapshot +Though I did not try anything just reviewed +well please help to delete this one? +The patch is stale +Hi MarcYep I think it makes sense to allow the user to specify the signature and cn algorithms via for both DOM +I wont bother invent an XML so I can test the XSLT +Attached patch started from Shais patch tweaked it a bit added the one line fix CHANGES entryI see the test in fact failing on +Thanks for taking a look Mahadev +Ifwhen that time comes we could revisit this +New adaptor modules are added for JSF and EL +Now that work is underway with AS all previous community releases are +Finally place the file in the tmpin directory to reproduce the problem +There hasnt been a source code change to dbcp in like months +This specifies which value of a row column to be checked in case of multiple versions available for the cell +Nice Robert thank you +Just to be clear this problem only started happening when we moved from RC to Final +Changing to duplicate. +The attached file includes a test case and the test case runs if Formattable implements clonable or not +The patch looks good to me +Improved patch fixes a bug in the Freemarker template caching algorithm +Have you set the two properties and as HTTP ERROR means that authentication to the proxy is required +This is actually in the wrong component +Attaching a jstack of the process +I attached the file to the issue +We have projects with the same issue and it would be nice to add this fix to our Spring +DirkThanks for reporting and providing the patch +branch and TRUNK. +in addition to the defaults the javadocs that recommend large values should also be finished the M wikipedia here was the indexing time patch ms trunk msSo there is no degradation in indexing speed or searching speed +verified in +Please confirm and close +MS zone zhost DHCP server +The JBR gateway combination can now handle the delivery of MIME multipart messages +Sometimes the workings of Legal are pretty perplexing. +It does this because aggregation writes to hdfs and aggregation will fail if the token is cancelled or expires +Aaron applied the patch to branches but I think he was going to apply it to trunk as well +No patch for releases removing from andShould still consider for + +As Lars pointed out we can work on improvments here +We should not have to spend time designing CQL features to solve impedance mismatches between RPC and query languages and we should not be redesigning Cassandra so every operation fits into a CQL languageWe have to face a reality it is going to be quite awkward for to clients to maintain multiple connection pools for client requests for thrift one for cql and one for cql one for cql etc +If there are no objections Id like to backport this to +It seems previous patch has line delimiters problem this one fix itPlease try again +Just thought Id point that out in case the truncate code path was being overlooked +For example Maven uses we need a visual marker that something is different and then to explain how to interpret the information +method create line It creates the default securityManager instance and puts into default under securityManager and sm +Ah ok +Please close this issue when you have verified that it works for you +What is this parameter used for? Is a passend when you complete a task regularly +I do agree that we should ideally produce the same index entries on the leader and follower but I am not sure if we should choose the first or last message +Moving to +Please forget my question about the renaming of the artifacts there wasnt any +At least we still hope to launch this year right +useful util. +The content type is determined from other attributes and never explicitly set +a minor update to toms patch with getting rid of a redundant line +Might be a couple of weeks unless I can get to this today +adds some warning around style and in we can fix these in if we want but its a lot of little changes for very little benefit so we shouldnt do it for Ivan for clarification +Applied the patch +Also it might be a good idea to create a separate filterinterceptor for the scenario so as not to litter the more generic filterinterceptor with code +The basefixture will be added for all locales +Has the add on had a recent release or perhaps you could point me toward the nightly buildIn the meantime I think Ill attempt Ryans patch though less than ideal. +looks good to me +Committed this to trunk and +Is the suggestion only for separate API in job submission? Or do we want to enfoce the segragation on the too avoiding maps cache files to be accessible by reduces and vice versa +However I still would like to release as is and resolve this issue in the course of the development +Have you got sourceforge account? Are you sibscribed to developer list? +fix on didnt realize this earlier can you check and see if the common utils are useful? I get a feeling we can use them exactly without reinventing them all over again +I will rework the patch and post it again +I am not sure I completely understand the issues being raised here could you explain them in more detailJudging from your title change I think I agree. +Unevaluated problem in OPT mode split in separate JIRA patch with additional fixes of number of bugs +Ah right shame youre not around Ulhas but thanks for all the feedback and investigation +Thanks John we were waiting for the release of the Maven Checkstyle plugin and its before doing anything +Ok so this is applied now in and and can be seen in both the published docs and the for pelt skinMOTD I believe needs sorting out for dispatcherI tested in and got more or less identical results in the followingWindows XP and Vista Firefox Opera IE IE Linux Ubuntu Feisty Firefox EpiphanyTested in Resolutions of x and aboveI will close this in a week if no other issues are found. +It seems I managed to get Run as Eclipse Application to work +code examples have been checked and task is done and can be closed +Committed thanks +The source of the errror is incorrect comparatorHowever I can suggest a patch to Arrays which makes slightly close to RIWith the patch sorting algorithm is not changed but is passed as on RI +The fact that there are multiple application contexts involved is really a separate issue +See PROPOSAL +We pay the price to pass to it in order to release the monitor during waiting +Yes I keep forgetting which it is. +Ill switch to delete and a new project and see what happens +Closing issue. +It also keeps the configuration for the plugins in one except for the typoscompliercompilerg +Except those special cases with order by and bucketing number of reducers for both RS is +Yep good to go +Thanks Kristian +Denis bumped to for now +Moreover you assume that the content is in HTML format which is not always the case + with debug messages for +Even New Web Dynamic Project wizard used to worked in that way before they removed Target Server filed at allAnd you can select runtime +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The filters in the fields NOTANALYZED funcionaram by Issus think seja um problem analyzer +That may be an OK assumption but I would rather see the test not need to make that assumption and instead validate that reading a file with an invalid checksum file will work with verify checksum disabled +Hi yes please do! +Removes copying of ID and URL fields from solrmapping config +Part of as part of +all issues to the CR +Thanks Rishi +here is a fix with an adjustment of stack and a unit test that fails wo the fix being applied +This is a directory listing of the that are pulled by assemblyassembly for comparison to the manifest which Im also attaching which is produced by jarjar for the same project +A different timer service will likely required for EJB and thatmight include clustered support so theres a slight chance existing EJB timer service wont be to verify with Dimitris +Agreed this needs further testing and tweaking + +Why not make the clean target depend on the target +Latest patch attached +or that it is even worse now Would be glad if you could give it a tryThanksCiao this is not my problemI resolved this problem by using jar +Tested manually +Ill run more tests and let you know if something is failing +The JS docs and platform implementations now support this option! Woot. +Thanks for your contribution! One comment +Attched suggested suggested makes perfect sense! I committed your original patch to the trunk. +checked into branch and Arup +A workaround for this issue is to remove the name from one of the bindings to make it the default one or to specify componentservicebinding on the getService call +Moreover it is enough to call Validate action on +a code drop from the head of my stable branch +Will apply soon +I have a class that contains an attribute of type of Object +I dont see a clean way to fix the dropappend race so heres my proposed workaround +xml +But it should be eap and eap + tests are implemented via the three default filters. +I want to be able to add pooling of buffers across requests +Fix the segmentation faultThe declaration order of and selectorQuery is wrong +Simply add an import for and it compiles +I can confirm this is broken +I see that the handling of streams isnt optimal in +kevin ok so it definitely looks like JBN EM thinks there is only a single instance and I were discussing this yesterday you arent using any clustering features here right? You basically just have a single distro directory where you create two server configurations servernode and server you post your binding service config files and your for the nodes +For emphasis I have addeda note to the release notes for indicating that the compatibility code does not compile with. +Oliver yes parsing means that the escaping requirements depend on whether list delimiters are present or notIm not sure how one can properly handle the behaviour change +This is the starting ddl Im using on my tests with +I think there is something wrong hereinside of this project there is no package +Files are no longer used by funder confRemoved Files ii under binRemoved Files +Patch to upgrade all usage to and to upgrade ASM from to as required by the latest source code +It is working fine for me but I definitely may have missed something. +I was referring to Ted committing this to before and trunk. +Thanks dan for your reviewI am attaching another patch for this issue +Hi Ill be happy to make our bit machine available for any kind of testing thats deemed appropriate here +Fixed via +run the Thrift server is fineBy experience usage in different programming languages would be nice to share +What I committed to trunk +Attachment has been added with description patch transition resolvedfixed bugs to closedfixed +Adding a with an empty replication handler section for the testI plan to commit shortly +Tentatively setting fix version to +The patch looks good +push out to +Added to Workaround +Will rebase patch to trunk if required once is merged with trunk +Also verified that the older snapshot I was having trouble with works if I adjust the order of the resolvers with overlapping urls +The URI component of an EPR does not have to be useable in isolation from the other parts of the EPR +The only thing I could think of to get rid of this is to tell the to skip the README file though Im not sure thisreally works +A block report calculates and values only if the block is not already in the blocksMap +Patch that turns validation on +So if a user creates a dynamic web project untargeted to any runtime then later tries to add our runtime through the method above our runtime wont show because we do not support Java which is what the web project was created basically theyll have to go and downgrade to before our runtime even shows up +applied clean patch +New patch that moves most new code to rather than +If you can use struts then definately try Struts RC +Fix of the +Catching up on Closing my reported issues +I guess we could just sort either before we serialize or after we deserializeOr maybe we do not need to bother since nobody will Append to a million columns of the same row +I do confirm that this issue doesnt exist anymore at least with version of Jenkins +changes look good +ClebertUnless I am missing the obvious I cant find the name of the configuration parameter you have added +This has been running in production for a few months +handle this. +But i think in this case we should set this jira as resolved +The patch builds and runs cleanly in my environmentCommitted to the repository trunk as revision Thank you for the contribution +Thanks Richard patched the at repo revision +Paging to the very bottom of the file you will find the failure in question +Frankly speaking in practice do you really need to modify the parameters of the mail server so often that it would make sense to implement this non trivial feature +This patch the MB segment files writes using IO and recycles the segment files once theyre safe to be reused +I expect that a user would prefer to deal with one deploy directoryThere may be ambiguities where we cant automatically decide what to do with an artifact +I kinda knew that was going to bite me in the butt +Here is the patch +Thanks Mark the patch was applied as expected +Couldnt reproduce this one in runs still counting +Issues resolved in +The code committed trigger the creation of wrappers +Were on the same page now +Closing this for now finding this to be too difficult to automate with our current test suite. +Thats what you were trying to do right +It is possible that some stuff is now never called and that other stuff is not exercised by the range of circumstances considered by existing testsWe also need more tests with validated encoded namesI agree +Closed after release. +do you have any insights +thanks for being so quick to respond all the time +Im recommending an alternative syntax for variables at installdeploy time +See previous comment. +Merged into master +I corrected all of themRegardsMariuscan the patch be applied or is there more work needed +Fixed in r + is the BPMN generated by the web designer is what Id shows the feedback when the event is over the border +The test to reproduce compile and erase then run +The class imports the though this is probably just minor because it is not really required to compile the class but only to resolve the javadoc reference +I am not marking this as Incompatible since the previous incarnation was not in any release. +My proposal does present the display faster when the history files are numerous but we do lose the ability to display the total amount or to jump to the last page +I have two cocoon processes running on a NT production system with the configuration cocoon tomcat java +Thanks Adam +In this case it will be vanilla cherrypicking +As discussed well stop verifying annonymous classes to prevent those false positives +Thanks! +this is different +All Since I have no test case I have attached the environment which reproduces this issue +As an FYI I think there is going to be a conflict between this patch and the proposed changes for +The idea is correct +We could either make this configurable and have it off by default or check for srcmainscala andor srcmainscripts or something like thatWDYT +Yes the connections failing in an unclear way is a separate bug they should be able to deal cleanly with the server shutting done cleanly or just disappearing +Thanks for reporting this +Run for example usingjava dummypassword jar DERBYJARS ij +Everything is defined in +I see two different issues hereThe first is that MR needs to copy locally dfs files and it turns out to be using local fs for thatAnd the second is that put and get commands of are not symmetrical meaningwhen a user puts he doesnt need crc while when he gets he sees a crc file +A space in caused tests to fail +I am changing the parsing behavior to the permute behavior and intend to close this jira as invalid. +Update Ive got an implementation for this but need to sort out some issues with the context to use to update the L cache +waiting until this hits trunk to make any other changes +Regarding your comments DRLVM provides its own which provides ALL services required by classlib +Adding LANGstdoff to the linker options eliminated the core dump and the dependency on LDLIBRARYPATH +If that happens with this patch the bootstrapping node errors on startup +Most of the time was spent on source node waiting for dest node to decompress and write to diskIt looks like the only reason to decompress is to compare crc +It was tested by running mvn compile findbugsfindbugs in the and I verified by examining target file for the absence of any related warnings +It actually DOES print the job progress +I dont think so +This is the failures we saw up on? There was a fail in hadoopqa too +I thought that job priority was added into the line a while ago and then it was more or less removed +It would be very helpful for troubleshooting many kinds of issues +Verified on EWS +Committed for in rev +Please apply the patch first and then run the shell script +Resolving as duplicate of the WAL batching work that has been done of late this issue talks about batching going into WAL +Fixed in revision +remove the hardcoded group name from a few test scripts +Not done by a long shot and probably not in a state for review +? The Solr parameter namespace needs some careful management +The unit tests should not fail if tornado support is not being used or available as this will be an optional configuration im about to make the and would like to hold off on this till the next release as this hasnt been tested by more people good Ill hold off until the release is cut +Hi Michaelyou are right +Hey Flo run configure with Ill make a change to have perftools disabled by default given it is unstable when turned on +Only the cluster administrations could set this up +After we added lazyfalse to that mapping it seems the problem is gone +Assuming the vote on the RC passes we can just leave this as is +To fix it I added SYNONYM to the String array of types at +The generated html source code producing line length with IE Firefox there is the correct line is gone using RF of a bulk update all resolutions changed to done please review history to original resolution type +We are testing it with openjpa and will let you knowThanks Ashok for the test +Simple patch with cleanup +All processes terminated +env and and system variables can not be setAnd other variables can be set in beeline Already + +Ill commit this to TRUNK tonight barring objection and also update the website copy +Thats Great many thanks. +Fixed as per. +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old + This patch adds the mavenVersion field to the srcmainmdo file +The mentioned sultion with select +Just wanted to confirm that it should be +Is your mapping for object correct? Should hibernate not give you an error before the database fails for that columnPlease attach a POC mapping file and ddl for verificationI think I have run into the same or a very similar problem but could not reproduce it on a test system +I will commit it soon +Well it would be done as a batch job daily or weekly so I dont think it would be that expensive +if you are using bookkeeper and you want to do some zookeeper operations usually you want to use the same zookeeper object to keep any session events in sync +number +Marking this resolved +in any case +Patch +You can disable chunked encoding by setting http level in the configuration on the server its not something you control in jbossws setupPlease refer to the AS Web documentation for the version youre using +After further thought it seemed to me if by eyeballing I could catch some missing messages there might be more +I currently have a wrapper class around this class to intercept calls that are outside the interpolators domain +A description of what was changed to arrive at the patches and new files +Thanks for doing the comparisonBased on the above I give for Liangs patch +can already be done in the last release with archive Brett! +I finally remembered what I did to fix this +Documented in the EAP and release notes. +I added it to the list to build under Windows SUA Interix with the gcc compiler using the configuration setup script shown below +This will be fixed in tomorrow nights build +Here we go this is a patch that introduces and +However it will be not implemented by default +Any hope of getting this fixed soonPer +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content the patch for this +Id recommend to commit this to trunk as well +Prevents a clean build of Hibernate Tools +Boil down your blog and use that +A can index all shapes not just points so it cant be sure if is supported +It only reports the checks twice the actual downloads are only done when the files are not thereThis is basically a cosmetic issue and I did enforce running the download task before building to ensure that everyone gets the source built +Is this new symbol on trunk +So the upgrade is good to procceed +Solved! +There was probably a bit of a mixup since we changed the name at the last minute before releasing M +Checking this in +Do you think we should take initiative regarding the Wildfly jira and propose path for them +Im confident that the test failure of is unrelatedI should also mention that I discovered this issue because without this fix one cannot view the conf page of a running MR job in a cluster whose is configured without a port +does support GAE now +It looks like its better as now sometimes finishes but most of the time it still fails +Andre Ive encountered this issue several times while I was searching for the steps to reproduce but I am not able to find reproducible scenario +The paths displayed in the report and the links point to directories the first four characters are missing in our case +width is fixed for Panel Menu +the Algorithm attribute does not have the wscwssc prefix as per the example you gave its defined in the spec aswscAlgorithm Thanks for pointing that out +All compute threads that use this get it safely +I filed to track the reporting format. +It was released with the fix in +faces +Didnt see you reply before I posted my last comment +Attachment sample report has been added with description Sample FO with SVG text that doesnt render has been added with description okay this is the right fo file + +Hey OscarCan you explain more the rationale for requiring the different format type for this? The way the multi shapefile support stands now jives nicely with the existing api parameter configure that controls how many of the feature types in the data source being upload to configure +Closing issue +The implementation is not in this patch so more work still needs be done on getting this cache readyThis patch has been tested with the default collector factory with sorting and ranking queries and in distributed mode +then they fucked up +Although I think you are right that its redundant Im really not inclined to remove the ability to interpret historical xml just because theres another way to do it +If it still an issue its in the plexus archiver +Then we can take a look at see what it takes to support something along the lines of the features we added to the aggregator EIP +Well have to do it with SQL queries or search replace on the XML +Mass closing all issues that have been in resolved state for more than a month without further comments +To take it a step further jmx cd typeValvejmx pwdCatalinatypeValvejmx lsCatalinatypeValveM nameM nameM nameM nameM nameM nameBasically you can use Unixy commands to navigate the tree +This was real I discovered later yed there was another config issue but it wasnt causal Ive just been too busy to look into it so if anyone else has it you are not imagining problem is solved in apparently described this issue in a way that got it fixed. +So could you please try again to install new snapshot if its not a big problem for you +Patch applied +Moving to in trunk in revision + resolved revision see close of resolved issues. +So my question is what do you mean by collation setting would be stored as a database property and not in ? In other words what do you mean by saving a jdbc url attribute as a database property +Well we want to use MINA in next JAMES releases +Committed +Thanks +This seems fine to me +IMO we have to include javadoc outputs to Hama distribution as reported by praveen +It is not responsible for the strange serialization issue I was accrediting it with above +At least thats what I noticed when I tried this + Patch checks for windows error message if OS is windows to see if it was a broken pipe exception +Regardless this still requires significant code changesI will get to it but I have much more pressing things to work on +danJunit tests only test against CXF then? So testing with CXF is pretty much done by hand +Damn it +However there is still the possibility that the confusion could arise during a switch to a new control fix has been applied to branch test +But this makes absolutely no sense +Verified with JBT +For the priority calculation we have to take into account the relations between the entities +This should fix the problem +Closing now for impending release +Patch v adopts changes from Lars patch +Sorry +I noticed that in one of my attempts with your previous patches +But this is the kind of change that is only measurable through benchmarking +You should use the individual we distribute +Sorry for confusion. +Arne this works fine now +This is somehow incorrect +tests and cleanups in r. +Christian the warning is acceptable +I mean you get a wonderful and expected exceptionConsider it to be fixed in rc +bq +This chapter is not included in EWP so closing for. +Patch looks good +bq +Of course +But in we will use it for the wiki space +Another mail with content sent to another address but I can not get it for the same reason +Either of you care to provide a patch? If so take the bug and then reassign it to me when its ready for review +Fixes in trunk r update implementation of index cleanup consumer moved and to spring removed old classes and test fixes in trunk r updated and activated tests for index cleanup consumer cleaned up index cleanup consumer +Thanks +I had this problem after upgrading the plugin with an application that used to work +I will fix this soon +To make all necessary configuration available should have more setter methods and delete final from some fields +If there is a problem please open a new jira and link this one with that. +Hes right I made that change +Code reviewed +Digging more I realized weve been testing with slot size for normal tasks and the same for AM since a long time +Things that will need to be added saving the deployment status between workbench resets +Looks good +I also created some abstract classes to reuse things like parameter validation +The applied patch does not increase the total number of javac compiler warnings +I suggest we could build our and documentations on top of that which I found it is pretty common for other incubator projects +For the remaining files it complained I will attach a screen shot for those +Ill release TODO fixes new features etc +Look in the normal repository not in the eclipse repository +Alexei attrirbute should be show +I stumbled across the code investigating something elseand filed the bug purely on technical merits +Does it mean that in the trunk and in the release branches you will leave useless data files containing the WIP word in them +I am going to try this on windows to see if that makes a difference +Tests fail it looks like this might be a issue +Please help to review it thanks +No tests need for javadoc change +Please if you have troubles with gant +The following issue blocks us from upgrading to there atleast a patch that would be provided if someone needs to use Jackrabbit and Lucene in the same app +The RS would just update the state of the region to OPENED and master can infer from thereWe have already started coding some parts will update once there is more progress +Feel free to provide patches and test cases and well consider it +I also noticed that the problem doesnt appear any more with the latest AS master +I prefer hadoop +RC +This error seems to occur when no installed runtimes have been defined +Updated to remove overwritetrue from initial content handler now uses dyuproject Rory for providing this code I have applied in Rev +Ok applied the additional patch this time it worked +If u agree with this I will go on to raise a separate JIRA and will propose what I have in mind there +However if the conversation id from prior messages processed within the same XA transaction are available these could be attached to the activity events for these other servicesBut in some situations the messages that have no conversation context will be received first within the transaction +Windows We have added a in the Java call that solved the transformation have same problem using do not think that setting should affect parsing XML because the XML declaration tells the right this issue be reopened or resubmitted +now sense at least when the bar is called selection bar +I have updated the patches for connector was tested against Netezza and Jarcec +I set the value for each writer so unit tests wont repeatedly roll when you have a cluster you stop one datanode and now your inital replication is +see in +The list is really where you should take questions about getting something to work +Breakdown +Moreover in such case there would be a kind of duplications between Sonar and Crucible which review would have to be commented? Which review would be the master? etc +At some point I subscribed to Thrift issue notifications but I have not been able to find a way to unsubscribe +see comment +Hello it is a Blocker bug and you dont want to assigned it? I think it would be nice to solve the problem +Indeed it is the same problem but this is not a bug in jcoverage because jcoverage offers two alternatives give all class names to be instrumented on the command line give a list of directories in which all classes will be it is possible to use a list of class files exclusion in addition to the list of directories to handle source modifications but im not sure +Attaching the patchContributed by David Lao Bikas Saha Lauren Yang Chuan Liu Thejas M Nair and Ivan Mitic +sdocbook source with an image for the test works as well I just test it +The first time I open a page in because interfaceMethod is a single page in the application work fine. +last post was a bit ambiguous to minilang to existing stuff +I found an easier way to reproduce this +Create a new issue instead. +now that map is. +If you apply the patch then you should apply the Thanx Jacek. +Before we replacing all of the datanode transfers we could do one or both of the following Replace reads +Of course Id be glad to collaborate with anyone on the task +Lets stick to minimal changes for now +added rollback tests for exceptions in +Part of this information should be the this pointer so you can correlate that to the memory leaks that are being reported +I havent benchmarked this but I suspect that it will have a nontrivial performance hit +if we could find a good way for the dialog to work then that would be good since the code completion will always hang around which is also intrusive +This just makes the code consistent +Im not sure we should remove the leading? in the get looking forward other dev point of view before doing anything on this +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +But frankly ATM I do not see how without other expensive synchronization +bug in release profile it doesnt assemble guides to one folder and doesnt build index +Im confused as to which is more appropriate or extends Logging +ThanksHoward +Patch applied +that doesnt scale in messaging! There are order constraints that you wont get on a can have replication to a backup node +ChrisI dont think the attributes will or should be deprecated +When is due +There appear to be no details to go with this other than the confusing cloning of the original issue +The security interceptor can prevent that which will cause clustering to completely fall over which in my tests causes the entire session to become invalid and unrecoverable +Additionally I can see that the hung child vm is on the next evaluation by hadoop +Code that throws and catches exceptions in the normal course of operations is highly questionable +BTW does have a constructor that takes a class loader +Submitted Prestons patch. +the translation of Hadoop single page of Hadoop twiki page has ever been reviewed +Seems reasonable +Fixed +Great to see these coverage additions Aleksey +Attaching patch for and unit test +Added two new classes and to help with Images loading +We do create a lot of objects and GC doesnt get enough time to clean everything up +Many thanks in advance +We are also running with at Camel +The CDE calls the framework code to validate the descriptor before saving it +A patch with modified is now attachedI also modified code in the stargate package that was using +I think the failures are unrelated +I cannot reproduce the issue +You can find more information on and that version enables you guys to test the editor again +Fyi I have been running perf test using and would upload the numbers soon +It is not OK +The test is not failing for me locally +Is there a logical place that it can be introduced without ruiningthe syntax if a user copiespastes the example +Fix other places where master was using its current time instead of that of the hosting s some weird blank lines in the diff that were added +the QS is also in have created to cover the fixed in CP CR the quickstart has been removed +That said I does not comply wih the DTD anymore +because you asked +I just committed this. +Patch applied + there any other method that will be called each time Component is to be rendered if it was set invisible +Integrated to TRUNKThanks for the patch LiyinThanks for the review Lars and KannanHopefully I got commit message right +It depends on how stable the current codebase is +Minimal test to test is only the domain class Archive and the integration test with code to reproduce the nullpointer + fix removed the hardcoded G for AM and used min +Further it seems odd to me that youd want your test cases to depend on implementation details because if you changed those details you also need to change your test cases +The key for the of threads is the reference to the volumeThat makes sense +it does not change anything for observers yetOne comment perhaps important is that Im considering that observers may not have identifiers in the future +Transitioning to waiting for version installed on Ubuntu and workingjfarrellminerva dpkg l grep thriftii Compiler for Thrift definition filesjfarrellminerva thrift versionThrift version on vesta on and Juno now done +Thanks for the help Mike +Taking a look +Hahahahaha BustedWill update it now +Steven another idea what if we simply added the options to so we could turn off the tokenization of title and date fieldsright now you can only control this stuff for the body fields +Committed in rev +Committed a small fix to the build file for package iapiutil which makes ant all pick up changes in files ran cleanly with +Thank you Kousuke! Its committed to trunk. +Revision updates the upgrade section in the reference guide for anyone wanting to take advantage of the changes made for existing projects +But the reason is understandable +As an aside do we have table name restrictions documented somewhereThis would be nice to have in the +Adding is required by +jBPM +Im not sure I want to support multiple Tapestry applications in a single WAR +I mailed Dan Mace about removing the above but I had no answer yet +RPC layers methods parameters etc +Thanks for the patch +The user will be informed about what a quickstart requires +Thanks for the info ScottTrying to avoid putting avro serialization inside of Writables I came up with this patch that tries to keep featureschanges to a bare minimum +Im working on this but having some trouble implementing cleanly which is already fairly complex +this also came up in an internal Salesforce discussion and we need something the leases for some internal use cases +Probably better to make it explicit in the Makefile as you didThanks Scott! +I went ahead and updated the resolution from invalid to works as designed +Hi Simone thanks for the on this sorry been away last week only just got back +cd www release is in both archives mv ls d lucenehadoop +The advantage of this fix is to reduce the code duplication in multiple places and to use the code +Attached patch a a small patch renaming cloneObject to cloneHolder and modifying a few commentsPatch ready for reviewI hope to commit this tomorrow +Im not even sure that this has made it into trunk either +Same problem occurs in apparently +Closing as duplicate of +This will require the next version of +Ah in that case I shouldnt think so then +It also depends on megs of jars and has a complex configurationIf you describe what you need I can create something +With JDK it worksFunny is there somewhere a minimum Java version requirement Ive missedIf not perhaps is better to state it LucaThere is no minimum minor version requirement +We would like to hook our special EL resolver in so that the logical properties can be evaluated +In this case the whole is dependent on unknowables and should therefore be retried until I agree that this can be wasteful and this should not be the only way to configure it +The actual log contains more statements than what I copied here +I guess this could be the problem here? We reset credits whenever a failover happened but not during a close +I dont know if XERCESROOT appears in the docs but messages are posted herewith some regularity due to this particular misspelling +Please check that it was applied as expected +Wed might want to make job ids a bit more unique too +I am still not sure however how exactly the arqs byteman extension invokes the retransformation +container is shownOnly way to get back the container is the Maven enabledisable the bug showed under version of Eclipse and version of the you have time to look at this yet +It also seems to have has the advantage of allowing a smaller and more easy to understand implementation than with the first approach +I am a bit confused if the primary key is an integer then why is geoserver producing key values outside of that range +And thats because I think most of the time a is transformed in context meaning a simple replacement like thisthis is maybe not the usual thing +I looked at the code its replacing a long while back and thought Oh man this should really be unit tested which is why I commented here in the first place +I cant think of better solution. +Ok I will work on this issue if I find my way to the code soon +Remove duplicated code in processColumn +Since the bug is within the newer version JDOdatanucleusWhy not reference what this bug was in JIRA? that way someone at that side can understand what this refers to +Attaching the smallestsimplest of the files showing the innertype parsing problem +Still looks fine to me log outin and try again. +Without this filtering of the descriptors the SPI layer would advertise support for features it doesnt yet support causing more TCK test failures +Fixed with revision of buildersjbpm message Delete +What I proposed is a workaround solution +this should be fixed by maven +Im attaching a second patch and updated +Im not sure yet the best route to go but Ill either update the Readme or create a new documentThis does not need to be done for the release though correct? I certainly dont want to hold anything upThanks point Susan this has been fixed for so moving this out to +I and the Stream in question is indeed closed in a finally blockMost of the rest are in Test code and as such leaking is a +I join a file demonstrating that pointPlease note that for now prevents the use of qualified identifiers in GET CURSOR as demonstrated in my example +Created to track follow on feature of allowing a relation in addition to a file +JulianThanks for your workIn there is +That will be the smallest change that will rid us of both the issues of this be later you can add a test to the codebase for that code +Yes untill commit deleted object is officially being edited so the reference to it has to be preserved. +Please ask usage questions on the users list rather than raise a JIRA are for development questions rather than usage +HB dialect already supports sequencesThe only interesting thing here is that you select from the SYSTEMSEQUENCES table +We will add it to the release +Thanks +It was released with the fix in +Attaching the patch for branch +The Accumulo team is about to release a new version that should have a published mavenartifact +What scope is your form bean +The latest patch doesnt modify the original message at all +So until then its still visible to other session +This would allow to perform the same mechanism as above in all places where filters are being encoded that Justin wants to redefine the fid mapper subsystem in his work the second alternative does not look very appearling +Please see my comments for more details on how things can go wrong +committed to branch +. +Moving to +Can someone please review this? Thanks +once those are addressed +Seems to have worked and other users have given positive feedback in related JIRA for the quick fix! +Ping +Ill change it the way you suggestedThe example test scenarios are a bit flawed I made so they can be fixed later. +Changes in revision Remove unused Sling API reference from tests in jcr in revision Further reduce the installer context this will make it easier to make it public for pluggable tasksRemoved most of the stuff from the installer itself we can now merge the installer and the installer impl and thread make registered resources in revision Separate support for bundles and configs put each one in own packaeRemoved the and updated to latest Sling the statistics interface and the counters in revision the osgi installer task interface and the task generation in revision failing IT in is now a simple data object all the logic like format detection etc is done inside the implementationThis should give us room to make the installer pluggable without changing the api later onDone in revision Im closing this bug for now as weve done a lot of ve done another minor update move most of the activation stuff into the run method and removed a local instance variable for clarification in revision +Ive hit a few regressions so this patch isnt ready to commit +Not sure I can really control delaying client logging in +Grcies +I wonder whether the problem might possibly be on the side insteadThat is Ive sort of got three possible theories so far The child process prematurely closes and +which breaks the point of having a reload button at all +I have reattached updated fileThanks Shameera +Id run tests and commits but I am on reserve battery +Then others can follow up +Fixed thanks to +This means a meta model containing Generated will not work in a Java JVM which seems to be an unnecessary restriction +The code pathway for setting up a Network connection goes as far as calling which adds to this list +If this one works as well I have to think about which to take +new patch committed to trunk +Patch is good after commit Thanks Anita. +Opened to branch thanks Eli + +Might want to extend output and output as well +Removing for unassigned m not sure this can be solved +Changing public interface of Partitioner to add new method calls will not be backward compatible as users who have implemented the interface will begin facing compilation errors if they have not also defined their initshutdown new methods +Post processing in itself an action could fail +Too risky to put it in. +Thank you Joe for your work +Overall looks good +Attaching a jar file with the javacore and heapdump files of the test fixture when it hit the OOM +What branch is this patch applicable to +I just fixed in the attached patch will commit now. +OK heres the final patch with all tests passing +For sales orders the holds a reference to that entity +I tend to agree with Bryan +It not the test application is a is called expandMode +I have question regarding the graph generated how it looks like? Thanks for your input! +Thanks! +Mikhail is uploading patch +not a bug +Patch looks great but I was able to reproduce +Also I havent added a testcase as theoriginal testcase will recreate the problem if run on a machine that hasmultiple aliases for the same network adapter +So we want to know why the new version xalan doesnt work but the old version works well +I implemented this but I also added the same method to Folder to be consistent with createDocument which lives in both the Folder and Repository classes. +I dont see a lot of people rolling their own but if we switch to API control and have the system VM setup documented that certainly opens the door for people to do what they want +If the opposite is true I trust someone will this bug or create a new one +BTW setting it in theclient will act exactly the same +Assuming lazy consensus I committed the patch in revision to the branch in revision . +Ive expanded fixed the Javadoc on that annotation in SVN + +Anil thanks +Ill probably do this since this file seems to be required may be right that should not get full bean status +Hi The fixed patch is fine +I plan to add this in the same patch +Mistyyes i just provided the release note for this the content provided by Strong please review if you have a minute +We cant upload with variables and extending another m pom +Fixes the class cast exception and a related intermittent NPE found while testing the same code +Yes its case our application does batch updates +Tihomir Jiri yes I have also confirmed only the ER have this bug +Please clarify +Thats great news! Is there a timeframe for this feature +I just committed thisYou had neglected to add javadoc comments for all of the new protected methods so I added that. +Requests into this internal view cause skin pregeneration to kick inThe new API is fairly small A couple of system properties for enabling this service +Could you please provide some more information on the steps to reproduce this problem along with a stackstrace of the to aid understanding of the issue and to help others searching for similar symptomsThanks to include in the classpathIf you attempt to deploy using to a remote server you could see this exception +Gregory how should your code look like if you want to pass a ready configureddriver or datasource into castor as discussed on IRC +Ok +Just attempt to reconnect for scenarios +NULL Watfordiscsi NULL NULL ROOT VMFS NULL NULL Watfordiscsi NULL NULL ROOT VMFS NULL The script went for template to destroy but had the same volume name so it wiped out the volume which killed the actual systemVM with new templateFrom here it was a proper VM destroy scenario delete diskAfter this the manager should have a system VM for sec storage as the system did not have one but did not +Patch applied I tried playing around with different source texts to confuse the but couldt spot any obvious breakages. +should investigate externalizing the profiles used to do this into some sort of standard confiugration that can be referenced by artifact identity +Is that correct? Or are you working off CVS HEAD +The content of this page could be really simple by displaying the list of all snapshots in a chronological orderYearMonthTimeActionJanuaryTue Jan Remove Sat Jan RemoveAprilTue Apr Remove Sat Apr Remove +The compaction algorithm may need to be modified as Adam suggested to make it more effective +M released +Let me make sure that old tables do not all of a sudden start have max version of +Hence unassigning myself from this issue and moving on to other issues I am working onthanxNarayanan +I just ran into this with Oracles update on Windows +I wonder why wont fix +changed fix versions cleaning up bucket +Is there something specific missing besides the extended mode for D notion? If so please file a bug report for the specific missing things and lets convert this to a request for the extended mode if its still wanted +This issue is subsumed by CXF and has been resolved with the patch applied in that issue +Chris thanks for looking more into this +Thanks for the feedback Thomas +The culprit is in the last patch I submitted +Just wanted to raise this for completeness +Fixed in revision +This behavior does not currently occur but we have appendtrue +I removed the eclipse persistence jarsIf you can acquire these jars and add them to the deploy folder that should be sufficient to the issue +Much of the current guides were for JBDS +The skinning differences are likely artifacts of contrast or other differences on the browsers usedBoth should be investigated +Fixed with revision +Not important enough to add a pattern to make it dynamic +Youre right I do not like this +You can update them but its a trip to the namenode +Im not sure its even a problem right now +Will fix +We have never been so near to the end +Im going to deprecate Searchable entirely +If the compilation is incomplete then errors like the one here will happen +LGTM to trunk. +As a convenience it would be usefull to add the rootPath to the property of the provider service registrationPurely for information? Yes makes sense +Attach a file containing your patch to this JIRA +Thanks Matej! This resolved my issue + +Wont fix? Thats odd and rather rude +You just wanted didnt youI was planning to commandeer a friends Macbook once the Ant buld was in place in svn and do just this but I know you have easier access to the platform so feel free if you want to +When we have that then we can make a long running system test to mergesplitmergesplit while readwrite load is going on. +Verified on +I have been reading through the archives +Looks like the plugin might be having trouble determining library dependencies +We have to fix. +This issue has been fixed and commit +Good tests thanks +new patch merged with issue kaveh +The rules service would be expected to represent the behaviour of a choreography participant +HiI run SNAPSHOT on Windows without problems +David I did not use your patch +The initialization code needs to be simplified +Thanks Weijin +I like it +It wouldnt make sense to move them to any of the existing groups +I suspect you have deployed it on ROOT once and +bq +ship it +Attachment has been added with description text path patch is likely only a partial fix for the problemIn particular it only catches DOM attribute changes but ignores animated changes +Also fixed the issue in But didnt do a arc diff against the previous commit which maybe why this is a new review +Thanks Milind! +The idea is to log all events but be able to turn off logging for certain event groups by configuration +FYI Im trying to revert the change from this jira at. +where you get the info about apacheThe problem seems that instead sending the mail without authentication it tries to send the mail by automatically inserting the user apache which is not correct +Created to address exception creation in +Branch Committed revision . +Nice +Patch to upgrade to Tika +Fixed in rev +checked in +FOP has a lot of limitations concerning RTF output +No tests? + of a bulk update all resolutions changed to done please review history to original resolution type +Is this feature cancelled? +If a client call the function without have a lock before it may throw NPE +Up for review +Hi the best place to ask questions is the mailing list +A better approach would be to create a DSL for creating messages specific to the application domain +Attaching a patch which makes accept Oracle Corporation in the property +I am going to check in this patch without using the icuj libraryIf we want to discuss adding the library detection in we should probably start a mail thread on list to discuss pros and cons +bq +we need to ensure it deploys and that all of the security issues have been addressed +Please add it be possible to deploy javadoc and sources +Let me ask them about doing the latter +closing it +please wait a moment David Im almost done with repreparing the patch this was a good exercise anyway I found several inaccuracies I like to clean out before you apply it +Support for external overviews has been added a few days ago but only in the geotiff reader would be nice to have it in world image too +Committed revision which solves this problem +Reopened because I had reverted the patch earlier due to build issue +There was no EJB Plugins in my configuration +More symbolic analysis is done for the instruction finding +I agree that doing range queries against RP isnt by itself particularly useful but it seems to be a popular request and theres no real reason to not support it now that we are using pervasively +Hi barrie thanks for this think it should be better to store this file in the root directory of the local repository +I have attached a test document for which the error is it works great +Any ideas what the payload should be? Object +The best solution whould be to first fix and after that change the behavior to update the lastupdated to +This should hopefully be resolved with a consistent solution for all those views with no cacheSeconds property specified displaying basically works for any kind of view with HTTP and except that its not preventing cachingWith cacheSeconds it works for HTML views but not for PDFExcel views? Does this apply to HTTPS only or to HTTP as well? I assume for HTTPS only in which case this would indicate that IE bugI could imagine that the Expires header still triggers undesirable IE behavior here +Ill give a patch soon +what do you mean by couldnt access? no action in popup? or disabled? or some exceptions appeared +Oops sorry +Ok so the implementation now relies on setting LDLIBRARYPATH instead of +Ive applied it to CVS +Folks often ask for this and its usually honored +Definitely exists and not exists is not a great choice because can be confused with exists drools keyword +Attachment has been added with description Patch to remove the PDF has been committed to r +Tom please attach any patches regarding the documentation to any of the three I have just created +We will log a bug against ptp and will post a link here when we have done so +Attached is the project file +To be evaluated if it it is doable and easy to understand for final users +works for me + + string tables and procedures for PHYSICAL and VIRTUAL boolean on PHYSICAL procedures onlyaggregate boolean if yes all below are allowed on a in a UDFFunction modelanalytic boolean boolean boolean booleandecomposable is only meaningful for physicalThe function metadata should be available for the function model and physical procedures marked as functions +We are indeed checking certain conditions more than once +We should show a proper error if this is meant to be +ChristianThanks for the patchI will review it when I get chance +We can close this issue after update to cxf +Its solvable but I think we should open a separate issue for that as its related to a lot of other things that need to be fixed with containers +Ive been intending to get some unit tests written been working hard on a lot of other projects recently though +Logs of randomwriter console output fsck console output and namenode logs +minor detail +the attachments are logs under varloghue the empty logs like is not a comment sbinservice is not well supported in SLES +Thanks Owen! +See which is in the upcoming +As far as I know we are only freezing it to change the references but the io operations are done outside of the lock +Here is the list of files without the ASL headerCRatjava jar cBuild find ? find v +QE do we have a test case covering this already? If not can you please verify the reported bahviour +The +Closing issue which has been resolved for some time +FYI I have done two changes +It evaluates the body of the tag when this tag is used in the jsp page +Its ugly +I also updated my branch on github +Nope +Closing as resolved in +blocked by actually happens to implement this properly +I will incorporate that with any changes I might need to make with the review +Verified in reopening to update release note info. + +Thanks +Resources be replacedicon +a unit test reproducing the check that you dont have bogus mailcap files somewhere in your classpath +Thus a wrapunwrap for each String is inevitable +Yes this is what I need I think this could be very usefull. +But i was not able to get grails running main because of null version error +The general idea is that Mavens is the source of you can use the following workarounds manually add folder used for compiling classes from srceclipsejava folder on the classpath tab in the launch configuration move all eclipse sources into a separate project +I can reproduce this error everytime I run the corresponding am I supposed to do with this one +would be better than better describes the problem +You are absolutely right +Manually tests added +You are doing something wrong if Hibernate executes two +Steve yes its about cleaning up orphaned rows on the association table +Lets walk through an example reading with PigInitially Pig is going to ask for the schema of the relation being loaded +This issue has been triaged and decided to not blockpreventhold the EAP release +looks great! +Yup I was targeting at +attach an example that reproduces the is the domain class and a screen shot +Jenkins will catch these things or a committer will run it +Koen please review the patch attached. +This only happens with the pelt structurer not the default structurer +Im a big twisted fan +It will be removed as part ofI will post a patch w my update shortly + +Assigning to you if you want it +Also I cannot really reproduce the case you described where you would break a deployed app and the app server would still keep the old version running +Its not too hard to make one right +This is about duplicate functions within a CFC +We dont have access to detailed sequence numbers for each piece of the tree all we have is a single currentseq for the entire group +Thanks Tim +Fixed on revision +Tested a bit more and found out that TS was actually sent but it does not go through ProxySo either Proxy object is not transfered or TS does not respect it +Patch was removed from! a microsoft word screenshot! that must be tricky! +Thanks I did the same change in as its also in use with for +Ill add that in +Great now we just has completed +attached is a seam gen project reproducing the described bahavior +Im not sure I would expect any space removal to ever occur within quoted valuesSurely thats one of the reasons why values are quoted to prevent removal of spaces +Any subprocesses that had forked off of the application would be the applications responsibility +Please reviewI think we are almost done +is fictional I attached the correct WAR +cleaned up a few thingsWe still need a clean way to randomly swap in the preflexrw codec +And Can let the two blocking job go parallely +Just renaming the test case to since all the other replication related tests are named with the pattern X +Centralized configuration facility adds one more single point of failure +It fails in init when it does this mkdir +Want to get this in as soon as possible so that we can move ahead with the branchingJoy was mentioning that an additional change to the cli shell script needs to be made for libjars support +This is clearly out of scope for +Shouldnt we add a test? I didnt see one in the commit +Have you attached a patch? I dont see one +I can observe strange things when debugging PhoenixTBC +It provides things enables the hibernate console for projects having hibernate dependencies provides Hibernate libraries in the JPA Facet configuration page +I also attached a patch +I dont think the problem is in the form I think it is in the method of finding all work efforts some of them are being excluded +You could do the catch by catching and dispatching manually but in Java I dont think you can even declare a generic class that extends Throwable +Was not applying I think due to the removal of some deprecated methods. +Just noticed that Ive broken links here to my github account +Should we just turn on socket keepalive +Is that communication to the ML something you are willing to take on? Im new to the contribution scene and Im not quite sure Id know the best way to present this +Or post your code +Changes look fine to me +This is a fix to possible deadlock caused by the previous patch +Fixed in About that JSON +Yes that sounds good to me +There was a logical error in the implementation code +AFAIK we just open the file and Eclipse desides which editor should be used +Lucia can you test this patch and make sure your tests still work? Thanks +Sounds good to me +Verified fixed in ER build. +Still getting to the bits +brewroot +Attached patch implements the above and adds a test case +Marijan please provide the patch +Appears to work fine on trunk. +Thanks in advance +Otherwise some function signatures dont match and cannot be loaded +The statement was written at a time where Ivy was developed by a commercial entity +RichardThanks for the bulk operation enhancements +Captured fields werent being indexed properly +The is John Patch applied at r +Hi Robertlike this much! I have no tested it but I trust you +Issue reopened to correct Release Notes flag. +yes clicks seem to work normalI could create entities but i could not find a way to link them togetherand createupdate buttons seems to always be disabled +Thanks a lot! +I am moving the builds over to the bigtop jenkins +It will throw the same issues that arent going to make it into to +I also do the diff from the top +Thanks for working on it +This is an option to our launcher that causes it to always use and spin a second processWith that flag splash works fine +The version Im using is +looks good to me assuming it passes testsI dont think this is in fact an incompatible change right? The field is renamed but the content is the same and given that its serialized using the same protobuf tag it wont change the wire format at all +While this is an upstream issue we can work around the problem by adding the name attribute as Pavol described +Yes it should be +This bug was fixed in please check out the latest released Camel for it +Includes patches and one is also eligible for backport to +Done for jBPM +Preliminary patch there are still a couple cleanup to do before this is committable +Previously and again lately there have been requests to be able to customize the notification and reminder templates per task +It seems this is fixed in the new release +Fixed on trunk +Thats fine later in the week is okay +Since the original problem was intermittent I guess I cant be sure that its resolved +And I think these partnerlink definitions are definitely a part of bpel process defintion +Thanks for the review +Hudson is not picking up +Something like. +Closed upon release of. +So if possible please upgrade +Please reopen if you find a way to make this reproducable +If you had specified and in a in one location then it works fine but instead you have one under the actual package and the other in a parent in JPOX builds starting +Fixed and tested plan upgrade +Bill Bell would you care to test this and commentI think there is still a loophole for a bare query it gets expanded across all fields as well and is less efficient than a which is more likely to be the intent when issuing a +Pushed feature validation is fixed now. +I think you want just quotes but Imnotthank you for testing +Though it would not harm to have a system property to control that +See the workaround details in I am going to steal this from you if thats ok +so it seems like extracting that part out would be painful +Please provide a test case +Patch integrated to trunk. +rahul we are going to be merging the integration into continuum this week +deletesrcmainjavaorgapachewicketextensionsajaxmarkuphtmlformuploadandsrcmainjavaorgapachewicketextensionsajaxmarkuphtmlformupload +has since moved and a few changes now have conflicts +Ran jenkins mvn test Dtest multiple times +Arvind thanks for following up on this issue +Please investigate an determine if this can be fixed in CR timeframe. +So hereafter we no longer use the custom wsdlsSo right now i dont have updated aar file +However I havent tried that yet +already resolved in master +A micro benchmark shows that this new approach is about x as fast as the other +small patch hacked + on trunk since +Navis Can you test and commit this patch? Thanks +Permaine was not installed in the upgrade +Hmm I implemented some of it but there seem to be problems with the latest svn version of myfaces so I cant test it yet +This will make it into whichever of our releases uses wtp +Set Yifeng as assignee +As discussed on the list BJ Hargrave confirmed that our implementation is correct therefore closing this bug +Improved the test case a bit +This is a great improvement for FVH +Hey Luke thanks for the reviewIll move +Add new jiras with more specific details. +I took the freedom to use the idea of the instead of. +Scheduled for +The improvements as suggested are committed with SVN id +yes and for the member itself too +We are not using counters +Thanks very much JimPatch has been integrated at r please verify if it is fixed as you expected. +See for how to stop and complain rather than continue to get NPE +This issue can now be s been fixed since. + +I dont knowMay be the standard has been changed from Shark version +We are building document indexing pipeline +in this one ive removed the and plexus componentinjected parameters +Tagged for family +This is a consequence of the change done in +HiI couldnt share my project because with simple project I created it was working fine but with the main big project the behaviour was totally unpredictable +Suresh it looks like the issue is rapidly spreading into multiple jiras +Created for the mapred changesAttaching both the patchesThanks errorThanks looks much better +Attached test results also +Links to RF guide from components at are checked. +Marking this as a regression since the user reports it does not occur with +Point taken the tests are not perfect +Group blogging and simple workfow for limited bloggers has been added to Roller in Roller per above comment. +Well it took a long time but heres a patch +committed removing ILP interface since MS is now the only valid implementation. +and tested the updated patch this now works and looks good for me +test fails is unrelated to the patch build on my local box with this patch is successful. +The last option is Configure Class Loading +Documentation is your friend +Corrected for data +I committed this +wbuildrsvn ci m CHANGELOG specpackaging libbuildrpackaging m Include with as includes directories as files when the directory has the same name as the pathSending libbuildrpackagingSending specpackagingTransmitting file dataCommitted revision +Agreed +So I suggest we simply add better help text +This patch also optimizes the locking inWith this optimization memory consumed by is under M for M watches +Ask GAE folks about it? That is a good idea +David Shouldnt this be bugs against +Changes have been rolled back and need to be reconsidered. +Committed +nobody has been subscribed they wont work until we set up the is up but qmail control files need to be updated and reloaded which I dont have permission control files updated still needs to be +Adding a test that upgrades from v to encoded v +If retries failed we should restart Hmastercan we remove the ? If you agree I will make a patch +Stack trace is useful for failure evaluation i +This class may be obsolete by Westport +Only difference is using https for redirectsinstead of http +Bulk close for the release +I also saw this problem +A test case for the fix should include urls like httplocalhostbugaaaaaabbbbcccc +No object left on Whiteboard +No need for IT +Patch for uploading artifacts that works on the branch +committed to trunk and +I just happened to be running some unit tests on windows and saw a discrepancy with linux due to support. +has been changed you in advanceDmitry +May also lead to ability to cache virtual queries which would be a significant performance This has been scoped out of as part of the Pegasus Server Performance project +Committed to trunk and branch +Can you prepare an updated patch for me to reviewCheersChris +Max I was sure I had fixed this issue it wasnt true +I have not the pure test case for this bug currently +how do you feel about adding a formatter which formats the Long as the readable date stringI like that option +But now there is the choose to either use the Requires or not to use it allMaybe I would have to make a second rpm with no dependencies on installed java version +Im sorry Im just only using the data server and this patch just fixes that component +Do you guys want me to update the documentation in the wiki? +I hit this issue in EAP ER and didnt try with AS +Patch passed core test and patch test +? Just marked it closed now +On it now. +It just strikes me as OTT +Its like a timing issue seemingly where we skip flushing seemingly +Oh the project is to aid in acceptence test where we should log the action execution timesUsing both computer numbers and a manual stopwatchWith this enhanced interceptor I can setup a log category and have logj log it into my own fileAnd I would like to specify my own category instead ofThis is useable for a period of time in pre production and production so we are able to monitor execution timesThe times will be greped and used for inspection in Excel +Reads are not synchronized +Moving to next release +But this may still fail if there are used multiple assembly decriptors in one module where the order cant be defined in that wayThus even worse the transitive dependencies are brought in depending on the order their parents are defined in the POM +Thanks again for helping out. +I side with Erik here +Lets go ahead and do this on monday next week +Each gives a xc No main method found error from xc and the +Input format for working with hadoop streaming +s job is to populate the context from the session prior to a request and store it after the request +The work done so far in this JIRA partially implements the needed functionality +But I think it is worth having now because we do have a way of enabling security in unit tests and it is a blocker to bring up datanodes +All done. + fixes this issue I think +Lets keep this issue open until trie is in core +CITE tests have been running on NG for a while now. +Another patch for someone more OCD than myself perhaps +RS startcode is good in that is has some info its a ts +Should figure out what timeout is being used for the drop +Logged In YES useridI only see this if I run the jbosstomcat bundle fromoutside of the bin directory +Retry doesnt work with Clustered connection factory +Getting it into the development trunk is the firststep so that at least derby developers can always easily get up to date versions of the code +Thanks for the bug submission +I had raised a while back for addressing the versioning in the new PB worldLets do the fixup in +Thanks Deveraj! +Fair enough I dont feel strongly about it just figured it would be good for consistency +The first part is resolved because you can now drop a column +hehe thanks for the quick fix +When shutdown is called then the error Could not obtain connection to any of these urls localhost is returned because there is no longer a process to connect to +I tried to create a test case but for some reason it doesnt seem to have the same behavior as the example aboveAh OK if get retrieves the fields from the tlog copyfields will not have been run yetIt seems like we should prevent get from returning copyfield targets when retrieved from the index +Verified at +Welcome JohnatanSure go ahead and then submit a patch. +I did some quick test since a respin will happen any way +Ran into this again seems high time to fix itHeres a patch that has fsck treating symlinks in a similar way to how they are presented in file listings as a file with zero blocks +I know Ive done it before in the past but at the least there will have to be some changes to the schema and wsdl files +At the moment you can use both role or language as the attribute +Fixed on revision +Ive updated the code on trunk to ALWAYS go through the pathways which allows complete removal of the locks and unmarshaller pools +Your commit was very confusing because itwas also mangled with the Windows line endings +How did you persuade the code to use GB? Are you running the entire JVM with the encoding set that wayCan you post a runnable test case that will use GB even when run on a default system configurationAny chance you could use wireshark or similar to get the raw bytes off the socket? Like Benson Im very curious how this could occur +Are you sure you did ant package? Once you do that you should have the following in builddistlibphp find +That was a lot better +I would prefer to have it immutable just like the various Vector classes and in fact I would like to have almost all classes in this geometry package immutable +I see this error when running nodetool ring command +CDPSM script updated for element parameter change in set command +Hi Jrgen please give the patch a try +If ti helps things along attached are the two SLDs in their entirety +I really dont see how it is possible to ignore fields in another way though +IT +This could be and the code should perhaps be able to detect and accept input that can return a reliable result +I agreed with Rick on this and changed the embedded structure accordingly +The default value of this attribute is the value of the interface attribute of this element +Jeez this wssj release process is driving me nuts! First they put up at WSOs repo then they pull it then they publish it again but it fails our unit tests +actually there are methods for setting username and password +It looks good also for RHEL x taken from I am closing this issue as +better but why is javadoc failing for this but not the other patches + old date issues. +I wonder how much thisd slow down the single threaded caseYes thats a good pointbq +Initial steps implemented +Introducing factory call in that will use the new interface to read the classes from jar file +Uwe Schindler Maybe I should create an new JIRA issue out of my suggestion to merge Filters and Queries? In my opinion this is something nice to have inI agree with this tack having taken it in KS +Lets tackle that as a different issue after this one if the need arises + Forum reference should be a forum not mail thread why not have the Channel part of the object? This would result in a lot less number of changes throughout +The Seam detector doesnt work when the AS or ESB detector is cannot update hte seam detector +Configurations would be managed by assembler files in the conf directory defaulting to wed provide a selection of sample files with different configs that people could easily wire up +Attached is a fix for the issue to escape wildcards in the FQN +yeah my error in forgetting the identifier for the enum itself +Hmm this is interesting +I am planning on updating as soon as Findbugs is released +You should be able to see it now that the can show the correct versions of the document +Fixed +Seems like a flag is the most appropriate solution +no region server ends up serving the region even though the master thinks it is not an unserved regionTo fix this we propose that when the master assigns a region it puts the region in a delay queue with a timeout of lease intervalIf the region server responds in lease interval saying it is serving the region then the master finishes processing the event of the region coming onlineIf the region server does not respond in lease interval that it is serving the region the master assumes that either the region server has died and will never serve the region the region server never got the message to serve the regionand consequently will assign the region to the next appropriate region server +Thanks Klaas for updating the patch +This task represents those issues +So Im wondering if this is actually the query you intended to post? Could you and post the exact query thats causing the failureThanks +We need to check all the grepsed commands in the shell scripts to avoid similar issues +LGTM +Playing w Hudson +I replaced recursive method call with a simple loopPlease reviewI also noticed that the default implementation was not fully thread safe +I will double check and post the updated implementation as soon as I do more tests +Duplicate +exclude feature has been added and now includes and excludes can be given withregexps. +Your analysis sounds spot on and the timing proximity in the log seems to back it up +Collapsing of the node is a small synchronisation issue +Im proactively guessing this is still broken in and remarking it as a JI issue. +on Steves suggestion as well +Bulk assign all unscheduled issues with priority major or less to target +Since I tested this out using the example app and it seems to correctly supportstatic javascript Im calling this one fixed +Ok so our changes areAPPROVED getContact get get get getLocales changes made rThe current state of is not clear we are currently using a of ISO and at the moment? Which specification should our UML tag referenceNote TODO task should be removed from are bits of craziness hereFirst MDIdentifier and RSIdentifier have been mashed together into for some time now +Cant build so couldnt do that one +having one useless install method is docs and QE overhead +scoursecoarse +Closing. +Also need to have a public setter on the appropriate fields as well as making them nonfinal +Attachment has been added with description Test to demonstrate issues with float and long has been added with description Test to demonstrate issue with big long values +It looks good +I am guessing two +This is fixed +as it is it looks like explain on is going to get left out in the cold works with Searchables but Searchableexplain is deprecated with the replacement on Searcher +Makes key methods in protected so the class can be extended +I think it would have been cleaner to just add an API method to to get a canonical name for the type that would be a legal name in the target languages +It seems to me producer consumer is a concept for message flow whereas client mode server mode is a TCP connection specific concept that only really comes into play for connection establishmentA server mode socket can consume or produce messages but will listen for a clientA client mode socket will also consume or produce messages but will connect to a server at startup or through some lazy establishment policy +It just just notify the then the should decide whether to notify the user or notJust a thought maybe Im missing something +will commit it +Sebastians hint was right I was using an old logback configuration. +Verified everything ok. +I suppose thats just a piece of sugar and we could rely on the user guides to handle thisIm not clear on what you mean by release directory +bq +Some places it works some places it doesnt +It tests scans and seeks and outputs a summary of performance and memorydisk usage at the end so you can find the best settings for your use caseIts lurking somewhere in my git repo +Please verify +Committed +Unlike the dependency report tree I would hope this tree includes version numbers +Here comes the test case to verify it +This makes things very fuzzy +This patch sends a Bad Request response with the reason Attachment name is not encoded when trying to save a document with an attachment which has characters in the name +I am in agreement with this however a couple of questions do we know all the areas of the code base affected? ie where native bounds computation occurs? does the Generate button from the UI still do the full bounds computation? or the reprojection from +Those classes will be loaded no matter you for now just remove the JDK specific part while I think about a solutionWhat if I made JDK the baseline? Whats the reason for you not to upgradeNote that Ill be switching to baseline in or te latest anyway +Done +thanks Shai +Thanks again +Fixed in Trunk for +Did you make sure that the patch applies on top of the most current version of trunk? I usually get this error when Ive been using a stale version +Again regarding the architecture Im not understanding +esb +Juergen any update on this? I could use this as well +Hi ScottSorry for the late answer +RTC close of resolved wont fix cant reproduce etc +Submitting to make sure Hudson doesnt MR patch Eclipse shows the same number of warnings with and without the patch +Yes it will +Im happy to flip this off by default and set the default update interval to never if that will help get m out quicker and fix it up later but I like this feature if we can get it working better +The nest stage is to produce test cases showing where this is the case and make them work +I have committed this +AlisdairNo need to create another issue +Reopening so that we get the correct fixed by David including test case +resolving +I think these descriptions should also be available in a REST interface to stanbol +Since this is fixing a compilation problem I will commit it without waiting for hudson +Verified during test cycle fixed closing. +The patch looks good +Henry Zongaro has an example to show that the proposed patch does not solve all problems +CosPatch for +is it +Cool actually I didnt need after all +Marking this as incompatible for it might be affected by user env +Ive reproduced the bugI dont see a problem limiting shared flows to directories that dont contain but we do need to warn because the reason for the invalid action error is not obvious +Most likely this means being able to these values in conf prior to starting the app for the first time +The patch removes missing blocks from standby namenode web UI +In order to get this information we will have to open the dbs which of course has a performance impact like you mention +Now that was resolved I took another look at the patch and committed a slightly modified version in revision Instead of putting all the methods in the interface I added a separate interface and all the related machinery to keep things modularIt would be great if you could test this in practice as I didnt yet have time to set up the test code you sent +But doing it just for is probably not a good idea +takes additional arguments in your code +Can anyone to reopen the issue to simplify the test cases provided for this issue +Thanks for the update and the energy put into this +Running tests +Added the lists mentioned above. +Yep working on this right now so that can get this +I also delivered a unit test that shows how the existing code was wrong +Trunk should be the source of truth ensuring the opcodes are chosen uniquely across different releasesHey Suresh version defines format +Initial implementation for reviewIncludes changes for as they are intrinsic +OK trying again! +Looks amazing +Please verify and close +Looks like one jar file in your Maven local repository is corrupted +This is probably the same bug as the one I reported and supplied a patch forin the mailing list a few days transition resolvedfixed bugs to closedfixed +Ive bumped them all to k for good measure +Thanks Todd! I monitored a few jobs and it looks to be working. +See that in both trunk and M? The screenshot above is from a installer for linux based on trunk sources +I guess its a bug in the release that has since been fixed +The properties are described in the JBM doc +Its not related to the actual fields in the recordI thought that might be a problem but its actually not +Since H is labeled alpha we can just release the binary labeled with suffix and document this as not production ready +Should master write out its znode name too? If it crashes this code could bring on the second master faster +down to go +Ill add the session object itself to the method +Toni want to assign this to me +I suggest that I simply watch the maven import and check if it imported anything +Looking back at the previous test you tried I see that was against the DOM too +Text etc override this method so there is no thread safety issue for theseThe remedy is to only register comparators explicitly +Looks like the last one was the patchfile with one that should actually apply +Mike can you attach a zip file with your sample project that fails +by Feng +Its been a year so I hardly remember why I wanted instead of the log +ticket REQ +Thanks! thats awesome +LewisOne way of testing would be to call on some documents and make sure that the language metadata has been set property +This stack trace alone is quite hard to interpret as you did not indicate the circumstances under which it occurs +Both EAP and SOA are recognized properly using Runtime Detection +Well put some specs together at the hackfest on Thursday. +Attaching patch which raises a warning when no privileges are revoked by the revoke statement for a grantee +I looked into it little bit +Maybe Im missing something +Andrew if you see no relevance of this issue to AxisAxiom you can close it +This is now replaced by +Same thing on my side Fabrice I dont manage to reproduce this issue. +Save a file then edit the same file and try to save again while eclipse says Building workspaceThe Progress window will pop up and say Waiting for CDI Builder +We can use the for that +Closing out answered by Greg Stein after posting the message over on the mailing list +Also the basic rpc headers are also in protobufPatch adds javadoc and also changes the names of wrappers +Many thanks! +Is it an authority problem? Do I need to request the permission first? Could anyone provide any information on this? +Merged the feature branch to trunk +fixed the Number Converters of the Trinidad APIsee commit log for more +Hi ClausI have just tested with +This looks great stuff with awesome documentation fancy applying it +Fixed with svn revision +That URL should read httpmysiteblahwhere blah is the name of my webapp +A Patch against trunk that contains a unittest to reproduce the bugThis patch does NOT cotnain a fix +Attaching a patch to this issue +Otherwise the branch is unusable +bq +Ill try to test it this is now available in trunk and it is in the latest snapshot +Well some of the discussion got held up on the list +Bulk closing stale resolved issues +so you can run this in your IDE using run as java application +Rebased patchBut now that I looked at it we probably need to check the MR for whether it still operates under correct assumptions +Thanks Paul and John! +Thanks Sohan! +Actually because the compressor and decompressor perform evictions at the same time correctness should be preserved +I would not go that far for the following reasons Not all the that Hadoop uses are avail as part of the +Marking as resolved since all subtasks were completed +I was not able to reproduce it again when it will occur again I will investigate on it more. +thanks dhruba +I generate mine on tag version +any objection to change to pom to use the latest version of hibernate tool? I can go ahead with the dependecies +Verified on stage. +I dont know which though +Patch available in branch +I have also added test cases including the regression tests and upgrade tests +To put it another way not being able query information about a task when it starts is a major drawbackAre we missing something hereI think this issue should have been considered in the wider context of how jBPM is likely to be used before being rejected +Doing this will have some effect on online altering also because as part of online altering we unassign regions +But I think it can be safely revisit given those annotations have moved to Karaf and Karaf actually depends on BlueprintHowever if we do we need to use the blueprint one not a different one because the commands usually reuse the blueprint converter from the blueprint context +bulk defer to +I still wander +So I added them back a few days ago to make the tests pass againIf you no longer want these platforms to be tested please remove the tests first +Any other ideas +Thanks for the review + May take up to hours for the fix to reach central repo +Included these optimizations +However many fields in is marked private +Patch attached +Fix committed to trunk. +Committed revision in sandboxErik please move this issue to the closed state when you verify the applied patch +I have updated tests +Above comment is for +Thank you very much for taking time to review and commit the patch +I refactored the code now added better documentation and also added a cache for resolved bean proxies +A put for external read that fails due to a concurrent load of data in the database shouldnt force the transaction to rollback +I suggest that you do that so that your proposal is clear to allId also prefer that you change the actual json for objects as little as possible +You have an Update Action which is called by submitting a form on this can return the user to using the dispatcher if there is a problem with the form data or it can take the user to the View Action via a redirect if everything is OK +Bumping to so as to remove version +Sorry Ive updated the script to reflect the actual failure scenario +Although maybe something like the idea in would work I havent thought about it very hard +How can it be both compatible with the existing strategy and yet generate different replica assignments +No I have no experience with unit tests +on the field vs the setter vs the getter +Code looks good to me +I think Wayne is saying that the error occurs when using Solr trunk as it is but when he replaces the Lucene jar checked in with Solr with a Lucene jar he built himself from Lucene trunk its fine +Thanks for filing the issue +That way we would be backward compatible with the old mapreduce format except for the fields which are applicable only to yarnThe intention behind this JIRA was to know the current usage and reservations if any per job +For what its worth theres a in that uses +Do you know whether the patch will appear in the next release of Dacapo? ThanksVladimir +checked in thanks Richard +Bartosz Majsak But adding javajboss makes it jboss dependent and not portable across JEE app servers right +Since its a source code repository I think we should try tohave as few binary files as possible +Thanks a lot for considering +Closing this as this is a duplicate of +unless were willing to break everyone who doesnt add IPE the optional field is a better approach +Still facing issue with Can you work with Rayees to see if we can get past this issue +Send us email. +Moving to to next release +Problem reproduced using XHTML and HTML docbook stylesheets +I wasnt reading directions See under docs directory +Any update on this issue +Apparently I never tried to edit a file until just now so I never saw this standard dialog +Fred Im wondering if this needs to be done in x branch too? +All those Maven options require too much knowledge +Verified closing +Your code works like a charm this really makes me excited +Probably a screw up on my end +Thanks +Only the IT was broken not the real code +Duplicating that functionality seems like overkill and worse will act as a cache against a cache +If they want to there are more interesting approaches +Change was merged onto the branch after approval on the dev list by Justin +We fixed this in and the fix is already available in +Attaching an early version of a Test Design Specification for Replication testing +yes there is no License activation in JBDS I looked mistakenly at some old test cases +An upgrade to could conceivably drop this dependency and then we wouldnt need it anymore + Username lafr + and other stuff was completely removed +This should be easy to fix do you want to submit a patch +Fixed message is There are no available Target Servers for selected Target Runtime +Setting to resolved. +Whats your current status? How can I help +I can be wrong but Esteves patch doesnt work with old generated codeIm confused +Thanks a lot Beat for working on this componentYeah we know it needs a bit of love +Its barely a blocker when a standard installation works. +Thanks +I have also included back into the nio tests to see how the remaining tests fare +The fact that this is happening implies that the two threads are actually in separate transactions +Minor changes +This is probably caused by the work on login timeouts +Sorry but this is implementation detail were not going to expose to the worldSee my user forum posts how to workaround it. +Patch for issue project to provoke the described NPE +Ok +was fixed by in most cases +Finally +SVN revision now has specialties included in the web stack for web tier has been confirmed to now work with for s specialties. +Depending on which provider is called first we get different results for abx and abczWe could invoke all providers and weight the results and then use the one with the highest weight of courseHowever as noone came up with real use cases for c I think we should go with b +Where is the nativebridge tool that was used to generate much of the code for this contribution? +tested! +Slava can you look into this one for Burr also +Updated patch with the ASF header in the test class which is the probable cause of RAT being in sync w up with for the various svn the patch for trunk version. +This is easy enough and low risk to fix in +Everything is OK +Task is in progress because of changes in dev +The loop in should be instead to have a linear complexity instead of quadratic +Nick and not everyone is developing eclipse plugins there are other kind of license info things +I am trying to understand the fail myself +Verify at +Would You please provide us with some more information? For a start we would like to know which file you are expectecting to be deleted +remove profile from distribution and assemble as annotations parent module +This is the patch we used to make it work +The junit tests have run fine with no problems +If its really that simple why has the javadocs evolved into a mess of bad recommendations and informationAnd if that is the proper way why doesnt do that for the user and then throw an exceptionThese are crazy semantics +John Smart +Applied in revision . + in the testsuite handles this issue +Fixed temporarily by copying in the jenkins job +This is what the admin jsp doesI would like to provide a patch to fix the broken ping function in +in revision Close for all issues before incubation +Translations were synd with latest zanata updates +I think this covers everything if theres something I left out please create new issues +Tested after v of the patch was applied to the source code +By this change System property access will return null +Also what are the queries that are impacted by this issue +attach the patch including the patch submitted so that hudson will pick it up +is general purpose formatting tool to replace placeholders with valueWe will need otherwise that strings are escaped properly everywhere they are evaluated on side +If you pool by application eachuserwill get their ownAdrian +XALANJ +Upon job submission the new app is immediately linked against the tokens +The patch applied is a good solution to the problem +Marking this resolved +All set Mamta the attached patch capitalizes Result Sets in the title +bq +Kind of felt like the constructor should have been in the trait rather than the classes too but the compiler didnt agree +HiI just want to know if there is a chance for someone to look at it soon and maybe provide a patch or if we should downgrade toThanks +Kenney i think you patch it to the wrong place +I think that we can use a easier way to solve this issue +And since it is used elsewhere in the very same situation I do not see how we can throw an exception +No longer works in what way? It should have no note that the withFormat tag will always render an xml file rather than the HTML file even when the XML line is removed +Nobody expects that will be aware of java sources in even if it has the structure of a Java project +Today I got quite a similar problem as described in the procedure +Thanks Jim! +Closing a number of old issues that I raised but which I dont think are still relevant +That issue does not pertain to looking up a host name but the fact that doesnt work correctly when it cant regardless of the reason +Nice work! Does this apply to JDBC connection attributes like autoReconnect and connectTimeoutthanks Solr supports these attributes yes +Maybe close this one wont fix and open a new one to remove the ca option +IIRC only sections are supported in the stylesheet that transforms the SXW into +Is anyone working on this? it would be good to have for but we could release without it +patch looks good +Thanks ErwanYour patch is in trunk at r +I assume that since you filed this for release that you meant to do the work yourself rather than leave it unassigned +patch looks good +Changed to we released version. +Verified with last +may be required for this to compile +But we will be practical about this and existing code or call the same library routines in the inner loop +This was originally part of but per Sprint planning it is descoped from what is described in research indicates that automatically registering a user from a FB login is rather simple once the signin part of has been developed +Basically I have added a logic to sleep for milliseconds and retry if lock related exception occurs +There seems hardly any activity around the maven plugin and it supports afaics only svn +Merged branch to trunk and confirmed it passed CI. +How do you delete a wiki page? I cant find the button to save my life +Thanks +Project is now first in tested +It is +In the latest version of the inserts into SYMDATAEVENT happen outside the trigger +Fixed in trunk. +Ive checked the write and read methods and it all sums up +Attaching a patch which solves the problem by ensuring that even though no region info is loaded from meta or hdfs we still keep track of the table and if there are no regions for a table we treat this as a hole covering the entire range +I finally understand the issue +Trivial patch +If still an issue then please reopencreate new issue verifying what works and what doesnt work. +Marking as fixed without tests because Im a baaad boy. +Any news on this? Propose we close this if this cannot be replicated + +Ive completely avoided the whole strict mode thing and so know almost nothing about it +Did you execute your command with the url parameter as +The tree component keeps track of which folders are open and closed so that when you navigate away and come back the tree state is maintained +bq +gwyn can you take a look at this +Does this correspond to Indy or Daytona? If Daytona then seconds would be a new record no +Also I dont start running the connects until the disconnect thread has started +Ive reviewed the code and it looks fairly straightforward bails out early on a reponse! +No +I think we can close this too Emmanuel wdyt +To speed up the test I will try to change the test case to use an database by default +Closing. +This is also in branch same code that handles +They are AMF aware filters that can be used around flex security +I just needed to be able to quick ship an order with a money order paiement received +It should be added to the About this guide section I think + Username thepriz +Closing the bug as it is not reproducible in latest build +The first stage of the process would just read all edit log ops from disk and put them in a queue +I did this because the changes were substantial and included some formating differences +Daniel has logged an app server issue against this but we can look at working around this if it is not addressed in far as I am aware there are only four quickstarts that may be affected by this helloworldhibernateaction helloworldsqlaction helloworldtxsqlaction and jmstransacted +keep that naming scheme local for now +The bug is no bug +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This is fixed in revision . +Attachment has been added with description Result PDF that shows the has been added with description Expected PDF has been added with description Fo file that triggers the has been added with description Fo file that produces expected P open bugs to P pending further review +Seems like the instance vars in can reasonably be made final +found a few more stragglers updating the like the ideas behind this patch but I think I dont like everything dumped into a single module +I think if we set the start level to level l before we start bundles of start level lower to l bundles will start synchronously +Hi DaisyPlease review this patchThanks and pushed. +So lets move that part of the conversation there +What do you thinkIf we decide to go with enforcement also for fullAccess users in and we need make a good release note at least and clarify docsDan I think you also cannot assume that a user authenticated by aDan database is a valid user in the system authentication +We cannot break all the Hibernate applications outthere +If after consultation with Rob it is decided that the patch is too broad we could always patch just the so that they make the localTransport asynchronous +The workaround seems to help at the moment +This is a bit difficult because one condition is a subset of the other +See attached Selenium IDE test showing the steps +Attaching patch with just the YARN changes +Since it is a user setting which will be consumed by the user in his script it is best left to the user to set it correctly +Changes made to in the original patch is now applied to +I think this is a minor issue since hadoop has survived thus far with no protection +Yet more detail in logs +Ill log a bug about fixing the semantics +Emails from SCM to the mailing lists were getting dropped +If you dont change any code in a library the preferred method is to use the library directly rather than import the code into your projectAre you saying you dont have ant setup? If you do then ant from the installer directory will download the swc into the libs directory +At this point I could provide some information is that the user request the APPEND operation before the file is not closed +An approach like that would not be viable as a method for maintining any documentation we wanted to officially publish and release as an apache project the authoritative copy needs to be kept on apache infrastructureI think thats one of the big things about Git that any copy is an authoritative copy +issue not Cordova API issue +Committed to repo +Responding to web requests and just taking a hit on some of the backend services would be a big improvement over how it currently works IMHO +The following properties are to original value after being dont guarantee they are all intended to be are all from expected ALL found NONEMPTYcache expected false found trueblackList expected false found trueimportAll expected false found truename expected found parentFirst expected false found trueversion expected found +You can of course apply the patch yourself or you can get in touch with about a commercial support package which might include custom fixes to older versions +Then sometime later a repair process is run which ensures that all the records are actually deleted +Rick merged this into the branch with change not sure what version of that was so just marking it fixed in latest unreleased version of. +In addition it has the method is which will return true only if its operand is an instance of and its method is returns true AND the unary operator is of the type unaryunary +Attaching patch incorporating Hemanths offline comment +Postponed to priority and moving this issue to assuming in most cases the certificate wont be updated while the server is running +About the please send me a mail of the dir structure changes you want to do so I can verify that they are compatible with the jdocbook plugin and the rest of the build. +Ah yes +Release Note ESB services are now uniquely identified by nodeesbcategoryservice instead of nodeesb in CR sources +I dont believe just fetched the Doxia HEAD built it and the issue is not only still there but for the fun of it Ive also opened and just to show you that this thing still has not just a few bugs. +Please reopen if you find that when using Dojo by itself it works but somehow Struts is breaking things. +Perhaps someone can donate the code +the patch from the chukwa trunk +Which lists were that? I could not find any such threads in either or concern is not about compiling the plugin +Attached patch +Unfortunately the solution is not this simple +We should just do the check early and throw no need for the +does not include the fault injection tests +Attaching a patch for this issueUpdated the test to use qualified identifiers in GET CURSOR + +But then it gets stuck on a scan of theMETADATA table to validate that everything is correctly deallocated from that logger +ot have special tooltips etc one would need to duplicate too much functionality of the xml module +We seem to be arguing a lot over this and its not doing us any good +Thanks Jacques. +I have my doubts about it being optimization at all +In fact what will happen if two separated maven instances downloadaccessmodify the same file concurrently from the local Wendy said either we need to handle more than one repository wait for this related bugimprovement to be fixedIm no maven expert but I guess theres a risk to see a lot of new problems if the local repository is not unique +Hi XavierDo you have any response to the comment I made? I would like to push a release out fairly soon so this issue needs to be resolved soon if its to make it in for this releaseColm. +For this we check whether binary is set setuid and setgid bits by root by a special group to which only TT is a member +I have checked the patch attached to this issue and after some review I will commit with some changes corrected some foreign key violation that in the demo data doesnt existscontact productId VUDUBX product shipment NEXTDAY NEXTDAY changed all the refernce that used to +by Stavros closing. +I took this approach because I wanted hive to work with the protobuf class directly +bulk defer of issues to +Closing resolved issues for already released versions +resolving as we have split brain in +Which existing one? Custom +Thrift will need fixing before release what w all the API changes +I dont see the point to keep it +Applied thanks +please reopen +Thanks +It could have been a span instead of a div but the empty td was intentional +This was a plugin issue and not core +You can download the nightly snapshot to compilewith JDK +Interesting +Duplicates +Is there a reason this is not closed? Is it in waiting for the drivers to be move off of svn? +It is easy to make conf a part of the key +Is there any other place that we do this? If not any advice on how to do this +What steps are you taking that lead to that result? I just dropped the war in the deployments dir to let the scanner handle it +I tested with vint and int and can not see their differences in my testIt seems there is a bug only columns are saved +That is awesome! Were going to shift this to so we can get CR into QE for verification +looks good will commit if the tests +Even after fixing import statements most of autojoin tests are failing +need to wrap post server start in trycatch that runs shutdown code if exception +the components description has been corrected +Hi Guillaume +It is solved by removing the explicit update call from the translators +looking into test again for ant tests +Manually tested! +The operations I am permitted are limited to attach file attach screenshot clone comment create voting and watching +Some clarifications The ZKFC always performs the act of fencing by executing the configured fencing methods +bulk defer of featurestaskswishes from to +Used build templates so I was faster than I expected +Mike I think you forgot to add the class to your patch +Patch along the lines of Bikas suggestionHighlights Move all RM services and components to an inner class that extends Update all RMcreate methods to use only their parameters and not access the RM fields directly +ThanksBest regards the patch of this issue caused two error of roll it back at revision r +I dont see any effort to make a clean implementation. +Thanks for pointing this. +svn commit m Changed state transition table to use UNSPECIFIED where the specification is unclear trunktcksrcjavaorgapachejdotcklifecycle branchestcksrcjavaorgapachejdotcklifecycle Sending branchestcksrcjavaorgapachejdotcklifecycleSending trunktcksrcjavaorgapachejdotcklifecycleTransmitting file dataCommitted revision . +fixed in revision +looks like this is a bug in the container +An example that demonstrates the problem using a modified version of the simplecorrelation example with properties in the headers +It seems that if I revert to version of the plugin the problem with the quot in the script file name goes had been using and then you give the latest a shot +And if youd care to look at the source code for the portlet plugin youll notice that it does use the +I experience the same issue with as well as with Tomcat also was a problem with the Bitronix standalone JTA implementation and its mock JNDI name resolution had to be implemented just for Hibernate +The filter looks good +Thanks +Thank Pi for the patch. +I will also add this to the wiki. +adding it as an option is something that falls into formatOptions which we dont have on +Duplicate of issue is fixed and released as part of release +should we first figure out why create table is misreporting things? I fear that we may be swapping one set of bad error messages with another +properties defined in should end up in the node specified in the path parameter +Doug Lets see what you got +The assumption is that this was fixed along the way somewhere +There is a clear benefit to me as a user of software and I dont think Im alonebq +to commit +Could you check again +Second this +But this changes lots and lots of messages a wholesale cleanup is worth calling special attention to since its very likely that a user will run afoul of it +If you would +Attached patch for discovering repository metadata and some revisions on the and +I will add javadoc for and make backupStore privateThe reason why is public is because I need to access this from the mapreduce package as well +Sure but that is not what this issue was about +Theres not much use release noting this hence resolving and cloning another one is of no purpose +One person one time would change the config to enable this while people tinker with the config files all the time so IMO editing once would be effort well spentAs far as the beginners statement is concerned the cavalier way we dismiss the difficulty people have when they just start using Solr has disturbed me for quite some time theres no way Ill agree that we should not consider an option because its just for beginners +Well look at least adding the ability for translators to indicate support for literal format strings +However we are using grails and GORM should cache these queries pretty well +I interpret general rule cii the same way as you do +It make the site very shaky since the error will show up and once it show up I have to restart it +Thanks Arun and Harsh Here is the updated patch with Arun comments addressed +Can you make a small test script that reproduces this error +I did want to mention that we have this site on an FTP mirror as well +Im sure +It should be addressed there as it seems to be perpetuating a false sense of proper YAML formatting +The has headers and body but unfortunately it accepts an and an cannot be created with an unknown status code +verified in CR + there any recent interest in addressing this issue especially after Oracles announcement to EOL JDK ? Thanks +This was failing because did not have an defined so I have added an instance ofI tested this manually +I made this task a bit more general +I do agree that in some situations it should be down to the dependency resolution to skip test scope dependencies if we specifyBut for now we can utilise the forceCreation configuration setting within the to forcefully create the even if is specified by applying the following patch +Can you also post which SQL statements are throwing this error +For that you should better ask on user or even dev ML you might get a better feedback thereRaj This is an interesting idea and like Jeroens should be better discussed onDont forget to put a to this issue in your messages toThanks for your interest in its time to begin a brain storming about this subject +How confusing will it be to a new contributor to want to patch the Build Environments page and find that its called Profiles underneathWhat do you think? +Im a little more worried about the second issue this problem demonstrate the buggyness of +How about providing a patch with unit tests +It might just be copy paste code for now but it would be good to refactor the load balancing code in the and Sinks to use the same libraries +There are some known issues with JDO Exceptions when trying to delete things but well need a way to reproduce a specific problem in order to work on it. +What are the exceptionsCurrently there should be no exceptions +However skip and retry were covered as well +Example invocation instructions in the class Javadoc +Pretty sure that an identical jira was raised in AMQ anyway. +Thanks for enlightening me Alejandro +Ill get that up soon +I checked to make sure that the new test was there +I apologize for my confusion +SVN remove unused imports +Patch modifying and to use new api +Patch is attached to fix this issue +Fix committed in rev +Runtime does not matter +How long has this been failing for you dims? I hadnt realized anyone was having problems with the test +Working fine +Thanks Suresh for the review +Agreed +depends on a bunch of Hive ql classes so I cannot move it to common and is called from serde metastore and qlI will change both serde and metastore methods to require hadoop configuration so that they can get the classloader from itI will also move to ql +Hi SiddharthA change in is enough forIf a rd party need specific configuration it can support extra new config fields for itself that will be set by the user and will be ignored by other componentsAccording to my tests a change in solves the issue +Thank you for your contribution +Please close this bug when you have a chance to verify on a Windows machine. +s unfortunately doesnt deliver the resource enlistmentdelistment semantics that one would expect +Backported tovGACP Revision vGACP Revision vGACP Revision fix will also be released in the first EAP CP that has other bugfixes for entity CP should be added as a fix version but I dont see it now Ill add TBD as a fix version +Is there any plan to merge this fix into soonNo + +Your first bug or your first fix? +I have attached zips to this issue +committed to trunk Committed revision committed to Committed revision +Id appreciate any feedback +I temporarily ignore tests to restore the broken build in revision +My test case above works. +Please create an new issue. + this as works as designed as its very hard to find a criteria for this to be done +Generally in hive we do not back port we just move forward +Commons conf is now a merged superset of the I think we can split it out further later as neededI tried running with different installation directories and successfully ran a job in mode +Only use this one this is updated so it work from a common mistake for people with english as nd language the difference between to and tooIn the class javadoc for the interceptor this should be fixed reason doesnt take to long the wait reason doesnt take too long the waitto too +Please ask assistance on the mailing list user before opening may be a configuration problem +The fix went in between and and if he can not reproduce it with then it is good +Having said that adding another API to get the segment metadata might not be a bad idea +This time in patch format with dependency added +I downloaded Build and verified that its fixed. +Fixed +Full generic matches are still preferred the will only start looking for fallback matches if the first pass led to an empty result. +This is my first attempt to contribute to Apache so not sure of the protocol +Could you start IDE with a clean error log and try to get it again +bq +Thanks DJ +Ideally I would come up with a unit test that will expose the problem +apart from syncing the definitions theres also some new protected properties with special semantics which need to be supportedimplemented +Closed issues related to Roller release. +So when the first job in the list was job id then the test would pass and otherwise it would fail +VHI upgraded but work still to be completed to discover plugins that have died and ensure that they are reinserted to the +Reset them all at once maybe +If Im interpreting you correctly well just want upgradesstables to error out in that case and mention something about starting Cassandra before running upgradesstables again is that correct +I used it in a project where i have to refresh the IDE several times for getting projects which are not marked with! +it can not be made simple by not creating thread for every new segment merge but to use standard threadpool queue scheme? Lot of libraries can do it easily +Committed on trunk for the time being lets see if this causes any issue in CITE close all issues in resolved state that have not received further comments in the last month +Currently configuration is static using a framework property which when set to true causes servlet context attributes to be shared amongst all servlet contexts as well as the servlet containers servlet contextImplemented in Rev +We have a few very specific queries that we want to build specialized filters for but dont want to keep touching core code +Should this go into M +Currently its only used for but theres no reason it couldnt be used more broadly for basic input source tests assuming they use a Foo as input +Lets leave the word hotspot for actual hotspots in the token space +Updated description and scheduled for release +An additional option should be added for this +Reopened to correct Release Notes field. +Thanks +Im thinking of a case where the plugin author knows that something will always need to be done and doesnt want to require extra pom work to make it work correctly +I think it would be great to have just not at the top level of the project as it is php specific +With the learning curve Im going to have on this one is a stretch +As mentioned please have one of the Mentors fill out the mlreq form. +I have approved it now it just needs to go through community lead +An alternative could be using the Category annotations for some of the tests and then having each test manage its own download of input data +Yes this should be supported as it is in please call me to discuss this +Updated patch add test cases +I just committed this +Unless you mean develop a plugin in JBT for Eclipse to make doing license work on stuff youve built w JBT easier then I assume you mean make it easier to develop JBT itself and since thats all pluginsfeatures for Eclipse my point standsIf you meant make tooling to build other stuff then ok perhaps having something in JBT would be useful +This is not the only place where ftrue is used in properties files +Thats unfortunate because I cannot then avoid something +The horizontal scroll on bottom of page is div scroll +The configId should be unique to a geronimo serverIf that is the case to fix this make the configId different or remove project a first before you do a Ron on server on project a +Couple things +Whats the use case for using a version range for a dependency? To me this looks like something to close as Wont fix with the only change to the core being that the error is already indicated during model validation +I wasnt sure how is supposed to work +may I suggest to split the number case in such with valueOf and those with constructor to handle the constructor version like in before the remaining mnumber cases? Then we can use for int long and such also the full signature instead of having to build it there every time +Pretty sure we wont be implmenting an NFC plugin anytime soon. +Created a new for loading swimlanes and heatmap data into database +Committed to issue as it has been released as part of release +As this is related to possibly can you take a look at this +Thanks Benson! I plan to commit this in a revision for the xml queryparser support +Thanks Ashish! +move to and remove as its not an issue there +I already deployed a new Redback snapshot to the codehaus snapshots repo +Mail sent to list +Wont be done in time for. +Once thats in place anyone can write an extension elementHowever we wont implement the Redirect extension element since thats not even close to being a standard +As it stands currently Im going in manually after building my EAR and removing the file +Thanks a lot +removing logging classes from causing bundle for the feedback. +So in general the permission is needed +Put back if you feel otherwise. +Bulk defer all frontend issues to +si es busca cakewalk Iraq +But as long as the error message is explanatory Id consider it fixed +Okay the issue is that after the poll then the sftp consumer needs to get back to its original starting directory +It might end up with some duplicated code but Im not sure until I see it pulled apartQuestion about your approach typedefing to Did you get a chance to test it under NIX? I have a vague recollection that it caused us some issues building our application but I need to revisit that to be sure. +I hadnt realize this jar wasnt built with the right version +model project set highlighting the patch code and checked in. +On I have been running the test for hour without seeing it +There were some problems on the build server +Interesting only the implementation is wrong the prototype in ts is correct +pending Jenkins +Sorry about the mistake in the patch a correction will follow shortly +I read it incorrectly +Nice work +Ive spewed too much confusion in this issue just going to rewrite the summary +Converted to the description of +Let me take a look at the dependencies it is pulling in by doing a build +max will mention this stuff to google in his mail to them +Please check that it was applied as expected +Move issue into legact component +but that is just a guess +. +The issue lies in how the refinement requests were formatted and how they were parsed on the shard side Ive made changes that should alleviate this issue and Ill push out a patch Jan th patch is totally missing the refinement step so its hard for me to say what problem is causing your issue +Note that this is meant to be an eventual future improvement +Yes for that case too thought Im not sure wheter I can port it cleanly to as has been fixed as well during December +Join Column Attribute Targets +Using the new mojo importCertificate will solve the problem as there is now all parameters available in the dedicated mojo. +Thanks for the info +Would it be possible to use EnglishBest regards for the fast reply +Im not sure all build systems would represent plugins the same way +ASSERT messages werent very helpful Ill update them so at least we get the real info out when you run into this +The URL was basically the name of the file within the process archive but since facelets caches by URL once the process form was displayed it would be cached even though the URL provider changed +The class that would in theory correspond to the Hot Rod server is receiving threadMemcached in the constructor +However I guess folks can do it in their own code easily enough. +Applied the patch +Moving the cache down to the is one option but I think we should also consider changing the APIAFAIU introducingid andhash was one attempt +Though per I think were overloading the notion of a corrupt block eg it doesnt seem like receiving a RBR for a complete block implies that block is corrupt either +I just finished a full test run against revision and all tests passed +fixed related bug contained within so setting this as closed. +Also switching to Hibernate does not solve the these tests do not mean that Roo is behaving correctly +I will look into created to track the test fixing +Can you suggest a way to reproduce the error +This bug has been fixed in Xalan Java +Is there any possibility for you to try the build using Maven? If it works we can live with that at least for this release I guess +I think we can test it by turning feature on and run +Hey Harsh there was nothing wrong with the patch but there was something wrong with the commit I forgot to add two files to SVNNo need for you to do anything differently +Were interested in external committers but would like it if you could first take on some larger tasks and attach patches to the +I wonder if hostname would ever get into this column +Region assignments should never time out so long as the regionserver reports that it is processing the open requestHave the pass down a Progressable implementation down intoRegion and then down int Store where edits are replayed +If anything please just let me know. +ok comment fixed +Attaching a patch shortly +the patch looks good to meIve just committed this to trunk +Can you update the usage message too in binhadoop +Fixed +Has anything happened on this issue? It seems like a pretty handy feature + file being think I have found a valid solution however there appears to be a bug in configuring the solutionThe Apache CXF implementation should be supporting the definition of an XML catalog within the deployment this specifically addresses the problem we regularly see when schemas are used which are not known to the catalog needs to be defined in of the deployment and can reference how to locate any required schema locallyFor some reason this is not being picked up for an exploded war deployment but I suspect this could be due to attempting to locate the on the classpath which of course it would not be for a war +I will take care of this tonight but can I please ask you to provide us with aunified patch in the future +Many test cases already exist for these problemsmatcherr and for placement of xsltemplate for xsl for LRE at top levelconditionalerr to to for choosewhen etccopyerr and for copy instructions at top levelselecterr for xsland many more +My account name is mcpierce there +Tried to submit to hadoop qa againAs to focing reassign it is needed since the region state in memory may show the missing daughter is openSince it is not in meta SSH will assign it only in fixupDaughters +Theres an example project attached to which demonstrates the problem but you need to remove your locally installed from your repo to flush out the problem +mvn eclipseeclipse doesnt pick up test source directories so you have to add them to Eclipses build path manually +Move the fix to +Fixed the staleMillis parameter in r +svn statM javatestingM javatestingorgapachederbyTestingfunctionTestssuitesM javatestingorgapachederbyTestingfunctionTestssuitesI tested the largedata suite on linuxThe entire suite took about hrs on a intel xeon cpu gb ram linux machine with ibm running with default jvm memory heap sizeAlso note the database etc is cleaned up OK if the test finish successfully +Forward references in variables and parameters are now resolvedproperly +I did a git pull when I started this morning abcc +has this issue been addressed? if so in which release? thanks in that work is underway with AS all previous community releases are +This is the expected behaviour when posting form data if you call getParameter first then all data will be read and parsed into the request parameters +I tried usingan array then doing array at the end but it was really slow +The same fix could be appliedNevertheless the unit test is no more relevantIll try to work on that +I can understand due to history that T really has to take care aboutIs it maybe possible to introduce some new configuration flag to indicate if one wants to clear or not? Of course default value would still be to clear it like its the case now so would be kept +Now the changes just need to go into theThe driver part has been fixed so the queries just need to be updated to be on a basis +That should be fine thanks added and will appear soon. +This will be addressed in the Java EE spec. +No toggle name of additional event to toggle what events you tested the toggle withI cant find anything usable except clickdblclickmouseupmousedown did you consider usage for +So what I need to know is what owner IDs are there that are less than zero? If all of those should always be excluded then the rule is simpleUnfortunately its not that simple +I request you to discuss these type of questions in the mailing list before creating an issue in JiraregardsDavid +I think the main problem is in +Robert the problem for was not right value for the constant but the fact that guessing the reduce input was brokenFor now should we commit the logging change while you investigate if we can fix the guess +Otherwise it will become COMPLETE in add +Fix checked in the other day. +Thats the problem with generating code in files +Committed to trunk +Was on +Also hbase shouldnt catch and retry +Also Dhruba still hasnt got around to reveiwing this patch +I dont use Velocity servlets but in my own usage this is what I doWhat do you mean include a page error +Integrated to and trunk +Had to do it on quiesced cluster because under load region count is all over place +Looking at srcmainjavaorgapachehadoophbasethrift the creation of server seems to have similar issueSeparate JIRA should be loggedLogged +Added the test query +test program demonstrating line wrapping. +consists of composition of +Does it sinks RC +Closing for now because were only going to support Smooks in this release + for patch +Lets put all the polish on before putting it in an official releaseNice work +moving to passes using bisocket +Im of two minds as to whether the need for robustness on the one hand should trump the desire not to mask code on the other +One thing I did was added ommits to the xstream perisister for the added properties +Am I right in assuming you want the persistence plugin and the as plugin to play together so that this is somehow automatically inferred? If yes is there an easy way to find out which version is provided by each as version +fixed in revision +is an optional service that shouldnt be enabled by default. +If we make it so they have to use then well have to do something maybe env vars or something to give them the ability to configure it +So even though there are multiple instances of the file even the most brain damaged of tools should pick up the files in the correct classpath orderIn a project only the test scope of the module being run is included in the classpath which means that your configuration needs to be in srctestresources of the actual unit test project being run +we can add a hint to where to looked in teh target +I think the change you suggest is worth making Kristian Ill file another patch +Spring Security z When you click on Castor haveDashboard Measures breakdown Coding rules breakdown +Why do we have entry points for fields that are not accessible from VMEntrypoints? It seems quite underhand to allow access to seemingly protected fields by using our internal reflection mechanisms +Added option for xmRevisited all Javadocs +will do +Readers work as expected +Test will fail thoughThe problem is likely in xml configuration of this perticular regression test not in VM classlib or test itself +The should throw a if something is wrong with the +Castor XML will handle the introspection you just need to write a small mapping xml mapping file to mark the getConnection method as transient +Backported patch for the branch +that worked +Oliver would you be so kind to give us some hints as how and where to implement this conceptuallyI would have tried to put the functionality you mentioned What we needed is the control creation command placment etc +I ran all the HA tests in both common and HDFS as well +I cannot reproduce it regularly +Also if somebody ever had mistyped and didnt fix or used unrecognized attributes in replication strategy or compaction after upgrade to Cassandra just wouldnt start up which is also a ridiculously bad user experience this is why I think this patch should be revertedEdit I think that concerns DSE as well because AFAIK they used custom attributes in the compaction strategy as well as in keyspace replication +Besides internationalization I would suggest including localization in this issue specially number formats +Since examples own authenticator is not used in SSO mode user is not outjected into session context +Packaged the raw test recorder config files and filter the tokens in the test distribution at runtime +But no need to commit it until such a test is ready +stackI am sorry for that i have reformated the existing codei will do another patchThanks for the designing points youve mentioned +Fix if possible for +Nobody has complained about anything broken so Im closing this. +exp files to they are created during jdktools build and removed with ant clean +merged in from trunk to as reopen if patch from Gianny doesnt fix your problem. +rUbuntu smoke and +just attaching a patch that is svn diffed from the trunk +it looks like your client is returning a result as a hash +I pinged the Signavios about it +This duplicate physical network makes it sound like the database is already populated +Should we just temporarily drop the constraint and readding it later with a SQL script +This would speed up the listing of VMsand also fix this bug +Im seeing this running and some other tests +Attached you configure and in conf +I think this would be a great feature +Sooner setup for advanced features post old issue +Since a workaround exists we can postpone it to the next tested successfully with ISPNJBC as cache and ArjunaJOTM as new patch which calls the listener tx aware before the commit to ensure the consistency +How does Castor resolve forward references to a simple type in other contexts? Whatever mechanism is used there is probably the correct mechanism to use here +I did some testing and it seems scmvalidate will pass any URL that just exists regardless of whether it matches your working copy +SMB now normalizes that by calling Charset +Fixed by +I wonder why will investigate more +We should also show the succeeded one or the retrying one as well if it failed before and the failed tasks is still showing +In the end I do share responsibility as I gave you the go ahead to commit the changes +havet you added all the new conf +I will use the dialect and do some testing as well +Bump +OK great thank you +Done I changed the method name and add several test cases +Thanks for reporting these! +Tim thanks for this solution! Could you clarify how is it possible to configure such gateway behavior for some specific in my app using patterns +In my opinion the issue should be much simpler to solve in trunk now as all the stuff is now removed fromInstead you would just have the configurable sweetspot similarity and assign different configurations to different fields via which will be responsible for the indexschema integration +does it include searchable archives as well +What is the use case for this? at where setName is used it becomes obvious that it can not be removed +Ive noticed that some of the classes deprecated the wrongly spelled method should I use this approach also? +Questions in my head Is there we can make the build run tests using these alternative versions to make sure nothing breaks? Does this correctly remove the appropriate dependencies so that common and protobuf jars are not included as transitive dependencies when using the rebuffed versions? Is there going to be problem with being overwritten if it exists in both common and +Map input records num of Features Solve time with Solve time with difference was just a matter of rewriting the algorithm using vector operations instead of element by element math +I missed a couple of Unicode mentions +Leaving this open and pushing to to complete Rameshs suggested change to +Please check patch I made in and see if that matches what iswill be in trunk +Thank you for the help +I was just waiting for it to be set it Fixed status. +Your a godI will try this in the morning and tell the result! there sorry for taking so long to verifyIt works fine the initial tests i have done +If it looks good can it be please included in +So using the client value when some other part of the server has said to use a specific encoding is incorrect +Pushed into +However its a bit obscure to look in the workspace Ivy preferences for itI was thinking you should be able to use the context menu on the classpath icon in the package explorer to select an Order By item and choose an ordering from a list of supported orderings +I want to check a few more things but will check the fix in soon +project reworked to be a plain gwt project that can be runned in dev mode. +I am still very new to release process +Thats a big to makeIll see if I can come up with an alternative +It sounds very flexible and hopefully could be used with Zookeeper but that would be another JIRA I would think +I think we talked about pushing this into nexus so that does in fact happen +It wont affect production environment because it is a very thing +I am not able to reproduce this using the steps provided +Would you please try recreating the patch against HEADThanks created against applied Yusukes patch r +I assumed that they wouldnt +The lack of a unit test is OK since its very hard to test for the absence of a race condition +However I am a bit confused by the attached test case running warinstall in the war project does not generate the attributes just like in the multiproject case +The public URL was hardcoded for certain pages not it uses the relative URL. +bulk close of all resolved issues +Stepan Wow! You were brave to take this on +Can we mark one as duplicate or at least change the subject title on one of them if it really is a subtask +Fixed some time back. +I removed your inline comments about what was changed +Note the page does not have a inline image so the bug is NOT a broken link but missing xmlimage tag. +Cool stuff +Could you please give the new SNAPSHOT a try and see if it can solve this also need to document this in the plugin site +Patch for branch is attached +The class has been modified to dump the deadly fault +En yes +Thanks Stan +Version of patch thatll apply of +Example project +How can I check it? I joined project after it was year old and I may be not aware of some settings buried deeply in the code +It really depends on how soon the committers fix their current JMS address problem +We may as well use to express the pointcut for interceptor +If we add a dependency to the project it would be downloaded but dependencies arent done until the extended element is issue would need to be resolved some way +just needs to be made manifest in the naming and contexts description +Unfortunately the patch is stale can you please update +That should narrow it down pretty quickly. +Creating projects in different +So it picks tier when the test only sets up tiers +All required work now done in current trunk +To be clear this is accomplished in the runtime wizard similar to previous versions +I understood the idea on what to do with Getting Started itemsBut still +All good points here is v with those changes +Patch applied +My concern was just that I might be misunderstanding tcpdump or something since I find this a little +Please confirm they have been applied as expected +In general the larger the company the slower they are to move on to new technologies especially when that technology is so fundamental as a build tool +fixed by a patch from snjezana +Isnt this something that would be better of being contributed to me +Fixed in SVN +Is this resolved +no problem + +The +sorry +Will this bug be fixed in JSRThank youve tested this in and it now works correctly. +Still waiting for a responseUhg +looks like we have a few to fix before this happens +Small state small performance overhead +changed the jira description to actually do something like +Thanks Chris almost sure I did a clean will try again +I thing that they would need to be applied to official documentation and then wiki pages could be marked as deprecated once its available in official confluence docs. +Sorry about that +Most commands such as ssh do not allow it. +Thanks brilliant! +Hudson test results are not public yet so its probably good to have cruisecontrol run for now as well until it is public +It seems to work for me now +When I turn on http with Jetty I get the following in the response headerResponse header Connection close +Vlado if you report it to Eclipse bugzilla I will appreciate it +I dont want to hold up CR +You need special special handling for keys with possible special symbols. +Patch looks go ahead and very much AndreaCommitted in r on trunk and r on. +Usually request wrappers path through request attributes but what if not? Do we really want to disallow thisMost likely the wrapper would be written by someone who wants to make it work with Sling so I think it is no problem to require the wrapper to do that +Its a long time since I found this +Internally it listens on interface where it can talk to the datanodes and tasktrackers while externally it listens on interface where it can talk to the rest of the network +Good to hear that this seems to work +Thanks. +First is that I dont think its normal to have a release come from Apache and go to another namespace ie +Change applied +So would it be right to use two monitors one to run the thread to do the log shippingand another to wait and notify this same threadI am not arguing against the scenario raised I think the JIRA is a fantastic catch +I uploaded my changes to the test cases in a new Bugzilla entry seeIm closing this one as the original problem has been fixed. +Being able to override the action URL would already help a lot +I had to upgrade to the harness for another fix so Im sure I have the correct rev but its still failing for me +Lowering priority as this is NIO and i guess synapse should be updated first and then the change brought back here +IAC clients should not be expected to unwind and that shouldve been handled in the RPC layer itselfInstead of having the throw an Id prefer to define a Yarn which extends Exception and can later be for more specific exceptions +This issue does not occur for me when I have deployed +Lars and StackUpdated the api to take row instead of kvLarsYes SEEKNEXTUSINGHINT is a definite gain but this reseek will help to avoid that one problem that filters have to reseek one more kv incase where we know the row that we need to seek toJust a thought of me and Anoop +Hope its going to be made into is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be am running +But deserialization the problem still remains +Shame on me not fully awake! I see the test bravo! +This issue was fixed in but I am still unable to upload an artifact with the form in +no s buttons onsubmit successfully reached +I will keep an eye on the test runs after I commit the changes +Tested under IE IE Safari Opera +It would be good to indicate that on the home page of +Fixed and confirmed by user. +bulk close of all resolved issues. +Deferred. +new patch which greatly simplifies the trunk file +I just set the phase to event in the context +I also removed the share counter as you suggested as it was not being used anymore and it is much safer without for the suggestions +Doug do you mind if I open a new jira on this idea +Had to edit that +to the depot path for p commands if there is already one in the depot path while still persuading the report plugins authors to change to using the SCM plugin +Fixed + since its a new feature affects version should be nil +Node doesnt include the sstables it repaired so the merkle trees are mostly different and and stream the majority of their unrepaired sstables to each other marks its initial sstables and the ones it received repaired Run repair on and neither nor send their repaired sstables +I havent tried it yet though +Let me know if it works +I think we should resolve this issue as Wont FixHowever I do like Niklas suggestion for Exception handling +That sounds way better +how do I nudge Hudson properly? Does anyone know +In theory we can support the full JNLP file but how do people use it in complex situationsCan someone find a complex jnlp file on the web? One that pushes the spec to its limitsNote If we add this function I now wonder if we should separate the inclusionexclusion mechanism with the part of it +Revision shouldnt this result in a download option for source on the AMQ download page? +Yeah it seems to make sense to only ack if auto ack is disabled +Bulk close for the release +This patch the transaction log in large chunks +The zip file includes replacements for the top tabs tabs and tab buttons +is probably not the best name for it as we also preserve the statusMaybe we should distinguish between just take the newly injected entry and forget the old entry +So there is an issue here +Testing file that generates those warning has been added with description For has been added with description Test file with several pages for each IDThanks for the report and the testcasesStrange I cannot immediately see the problem +To start eve and use your own specific properties do thisjava jar evetarget eve +Maybe a of case should be rescheduled elsewhere more quickly than the minute dead datanode interval while we let a mild case of of sitI dont think this will require tracking our own list of stale or dead nodes just a list of nodes weve already tried for an outstanding request +Without this patch we have to synch the built pieces with the repository because the CVS directories have vanished from the distribution and exploded directories +just committed this. +Any ideas of when this might be incorporated into the released source +We can do a separate refactoring pass separately +Ill try to set the property and see what happens +Theres no description of how the scmBranch is derived its clearly not just the last part of the SVN URLI get UNKNOWN when executing code in trunkproject but trunk when in trunk +Rajesh Assigning it to you for your evaluation +I cant reproduce it. +I certainly cant imagine any other alternatives but if there are any then this code will break badlyAnyway if you could subject this to as thorough a test as possibleand I conjecture you must have a fair suite of class loading teststhen please close this bug if alls well. +Thanks Nick Ill take a look at it +I have applied it in Rev +It was resolved by using Eclipse +I do not know what changes accounted for the second patch +Attached for Teds commentWill try to come up with a unit test too +Karam this should get fixed byCan you close if it is not valid anymore +Committed to SVN revision . +At deploy I get this time too a has no persistent id be fixed once Hibernate Core is out +Its still on our radar internally so ifwhen we try this again well update this bug +Attaching diff and out files for the test failure +We need to understand the use case to decide how to solve user problem +It looks like this bug has been fixed in the latest CVS code +A patch for Tomcat ext is missed +I guess not so dropping it out of content +I have not tested this as I am putting my major efforts on did not installed clusterDo you mind have a look and test the same +ah the title says Preferences +So we just need to read the first tables data into an array and when we run out of memory just dump the remaining data into a tmp file +Denis you were looking on freemarker +Unfortunately the solution brought up by Paul Hanneman does not help in my case +I tried using the brew repo and ended up with the exact same issue +push to +REST API representation is fixed and the style preview in the Web UI and preview page work correctly now +Updated to refer to instead of +I will take a look at this have gone ahead and made this change +Ive tried to reproduce it without successAlso no feedback so closing it. +Good for me. +Bulk close for Solr +I dont know how difficult could be to to thisSource not found groovyscriptFile name it looks should be known to debugger I suppose +You are welcome! +Active when throws safemode exception the client must retryI will post a patch along these lines +Many thanks for you reactivity Claus and congrat to all for this good work +This bug has been marked as a duplicate of +So far the code has not been changed +Screenshot of issue +Otherwise they would get very little to no recommendations +There may also be a variant of this issue which affects identity columns in and all the way back + use for back to simplified patch in r +So it need at least mins +blocker for we dont want to ship this as a client api then change it later +Anyway I really like the way things have been shaping upOleg +I will put it on our FTP and send you a private message +Therefore it is really pointless to force a new constraint on your user that doesnt help your implementation. +The file that causes the memory out of bound exception +Im assigning this so I have it in the pipeline to investigate whether its possible +Please either use the default patter or put the custom codenamebase into the srcmainnbm file +Thanks Jean for the heads up +I have documentation in the works +Turns out we do seek and then read rather than pread +It just occured to me that we might have to switch towards using theExternalizable interface for as we do not want to rely onserializations default behaviour +Correct +This should include examples as well +rubygemsexport RUBYLIBGEMHOMElibThat did the trick for me but Installed my gems in my userdir +Thanks a lot Andrew just uploaded a version with this slight change you pointed out +I imagine this kind of common functionality needs to be pushed up into the maven codebase somewhere but dont have enough knowledge to suggest whereThis should be enough to fix the issue for this ticket I dont believe there is a de facto sensible default ruleset and instead I think that anybody using this plugin should be choosing which rulesets to be using in their think the XML output is a seperate issue for which an additional ticket should be dont have developer access to the maven repository so Im hoping somebody can take this patch and apply it appropriately +This patch fixes an endless recursion which results in aIn addition this patch implements local site testing of and fixes several failures and errros of +Input welcome! +Wont fix the copyright dates are really only applicable to the year in which the source file was authored and updating them as the years go by doesnt really have any leagl meaning as far as I understand it +Probably with debug logging we could see where this is happening +There is a open issue on naming strategy if you want to contribute +The defaults for this parameter should be such that they areacceptable to a wide variety of users +I would be fine with that name if it would be unstored by default so you have to turn on storing explicit +How important is it to get this in AS +Im committing this one +Added the initialValue method and a test case +I mean the JAXB XJC Plugin from +bulk move of all unresolved issues from to +Java output of test script no errors +I will address them in a patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Seconds was inspired from the JDK implementation we had seconds earlier +Sometimes we must find date to everywhere in cource this is not important issue +Can you reproduce with a nightly after was applied + seems to miss out on and which were failing +Intended for inclusion attaching here for RTC purposes +I am sorry i forgot to mention it is happening on a sender embedded in a region. +Added double error messages test on Ant page +mvn build successfully but still need test on cluster environment +If the Validator framework can take care of this I dont have to deal with the required validation specifically in the jsp +r +As per my current understanding of cassandra that does not sound to be right +Sorry this is not a major bug +The end functionality we are looking for is to be able to subscribe to the events that are produced from a jobjob create name hello definition myjobstream create name jobNotifications definition logstream create name my definition trigger jobhelloThis is described in the manual but I was not able to get it to work for a demo +Yes definitely +Problem exists on trunk +Ive just pulled the complete trunk out of svn and I am now going to put the test cases in place and reapply the patch against the current head I understand it right that the location of the test cases in the trunk should betrunkxmlctftestssourcegeneratorFacets +The configuration file has been renamed to I get the same error +There is something wrong with my thinking is a security test +If it is ok to add a new file for win only test and modify the exclude list please have a look at the and ignore the +Bulk move issues to and +reuses so the only change here is for displaying the url +Moving to TBD +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Committed +Will feed this boolean into and then class +Otto if this is still a problem go ahead and commit +Instead of making it I just committed a small patch that adds a new constructor that allows you to explicitly set the root directory +Duplicated on master seems to be a catch all for numerous problems that exhibit the same symptom +Attachment has been added with description sample file basis for a simple has been added with description Result w has been added with description Result w has been added with description Result w this to the version I shortened in the example to fit in one column and then theres no interferenceconfusion with the normal recovery mechanism +Changed hadoop jar file name to +Attach the patch that Justin sent +Committed +It is a convention to allow user defined meta classes wihtout having to code anything +FUTURE hardly possible to satisfy the webservices contract with the current api +Thank you +I am fine with doing it through the distributed cache +I assume that changlog is updated forevery commit question I am not able to answer is if it is possible to define an ant taskto update the rss on the webside? So its up to the commiter to execute thistask for every commit +Id think it would be more consistent with the spirit of an interface to simply require that every classstruct that implements the interface has the requisite property +Well take a closer look +I used a +I have committed this +let me check this if not will revert it by tomorrow +The config layer is now the only way to access the configuration information +Thanks Jason! +If you use immediatetrue then submitted values will be rendered not the values +Please confirm. +I think by default the test classes arent included within a jar. +Fixed in Rev +findbugs warning is Inconsistent synchronization of locked of time which I think can be ignored because fConf need not be accessed in synchronized block only +Patch seen and applied case closed. +level or in each and every code generator +applied path in r. + +Fixed capitalisation issues and changed all instances of release notes to lower case except in the title +Still needs a description of the parameter used by the plugin in Looks OK apart from volunteers to assign this issue toP If not ill pick it up for in the next couple of weeks +fixing in +Thanks for the reviewName change makes senseI was thinking about making it configurable but decided not to introduce yet another parameter +for the delta. +Still doesnt remove the ugliness of Dpropvalue though +Please reopen if not. +Anyways it is a small change and if you really need it you could always take the branch and do the required adjustments +Ill extend the patch to also allow empty while statements tomorrow unless I hear that this isnt desirable +why is it that after this patch is applied that no unit tests fail on our side in commonBecause was changed to initialize the contained in +This looks good I suggest you commit the patch +We will need the proposal to shakeout before we can make this easily discoverable +I agree with youTo classify coldhot data and store them in lower cost is a generic issue and its always related to application characteristic so I think we should make strategy pluggable and provider a default implement +Nevermind the issue was on my side I didnt realize the main jar included all the framework classes and I had backported the fix to the framework jar. +Go for it +action to make imported projects connected to CVS +I dont see anything stopping us from doing this +Thank you +The test failure is not related and is filed in think we should just move the stuff +But the numeric code should fail to work if this is wrong +It also corrects the read logic so it also works with unbuffered input readers +Patch that that removes cyclic imports was applied to branch at r +So we have issues here I think +Guys I just spent a lot of time trying to figure out why I was getting a when running within Intellij Idea +Link to the upgrade +I will do more test next weekwe only have about million used the latest code test againthe final output also is strange +We will have to create new Mappers and Reducers based on the new API in order to support newer input and output formats +demo project for problem with JAXB databinding +So all issues not directly assigned to an AS release are being closed +Some developers feel its easier to maintain one file versus many filesAnother motivation behind this is to have one file that can be swapped out to change s appearance + +MartinYeah good idea +I added to it a get method +map + +Attached the simple XSD that I used +Do we really have to make it more complicated than it already isKeiths suggestion should work well +If there is any case not covered by then I can add snapshot this parameter need to be tuned in conjunction withLet says your snapshot size is MB but the most recent txnlog file is GB +Ted should I modify the patch to match the java class names anddocumentation that you have chosen +except Something +Which meant that it was okay to make id be hardcoded +According to the documentation the only way to watch multiple directories is to create additional cfg files which create new File Install instances +Pattern line please remove code that is commented out the reporter is left unused +Youre right will try to fix +Verified with japitool +Closed as already addresses this issueThanks for the report +DayrnI just replicated what we do in for the same +While I understand its a cosmetic change being a newbie I just wanted to go through the contribution process and understand it +Sybase and Informix are not in the list of supported databases +trunk +Made some big changes to how static imports are handled +So IMO this feature needs to be implemented outside of Synapse using the extension points Synapse provides +i thought i already saw this and fixed it though where I simply didnt close a file in a thread +Chatting with Dave Simpson to protect against different clients inserting rows of different types all into the one table producing an undefined sort order because of the of type comparators the key type for a table should be defined as part of table creation with an illegal type exception thrown if a client tries an update with a type +But we should minimize it and avoid duplication of logic when possible +We just need to make sure that no Log object lookups take place on critical execution paths +I would need a precompiled class that lacks a parent +Dan If youd like to think some more about this let me know +James yes I still have the problem building the trunk sources +Ive just submitted a patch to fix the marshaller thread local leak +I have two thoughts +Thanks +The linked issue of which this is a workaround has been scheduled for fix version so I guess this should follow suit +Sorry this is my bad because it thought that and wont produce the same bytes tested now +Updated to generate schema for test +Would it be OK to have a protected method for this in? That would still require a subclass but would minimize the amount of coupling that such a subclass needs with Jackrabbit internals +Well we cant spit all the system properties in the log even in Debug mode as this would be too long +I think this can also safely go into +At revision all old Resolved issues +Bulk close of resolved wont fix cant reproduce etc +In a real system this would lead to problems +That solves my issue so this ticket can be considered resolved +Committed revision on main branchCommitted revision on branch. +sshidrsa or +I am not reviewing it right now but marking as blocker for so we get it in +Because it is +Also this behavior is the same with Webrick and Mongrel +Thanks +Thanks to Jean for verifying the patch +Attaching the requested outputs +I removed the deprecated ones should I leave them in +If its a big problem though I wouldnt worry about it +Its also ok to wait until to verify the fix +This is because the same fix has been submitted as part of two patches and the discussion was in fact in hence my confusion +This patch fixes the issue by guarding the stack usage +testsfunctionalorgjbosscacheaopcollection +Verified in +Im not super convinced we want to play God with the but its too late to sermonize nowMinor nits for Please make the default to false Move the config key to we dont spread them around now +i will remove the corresponding section from the documentationI disagree +Ive attached a patch to containing the current progress +Closed after release. +And it opens new problem like for example if missingMethod should be invoked or not +xml to the patterns for signing +I guess the best way around it would be to have an API for checking if lease is recovered or not +committed to trunk and. +Can you provide a complete stack trace and perhaps a test case? +committed. +Because this uses a primitive type in the simple method it is dependent on the from patch used context as a field name which conflicts when sending parameters from a screen +Remove VMLEVELTOOLOW which is no longer used now that the Java functionality is available in all versions of the + +Missed a paren +Please find the codeinput pdf file and converted png image in attached zip file +ant clean and then I also didnot get any releaseaudit warnings +HTTPD also appears all right +Could that be done with a dendency or reference definition in? Then plugin can use that plan if one existsAlternatively the Eclipse WTP Server Framework allows extensiblility to the publishing processes +This kind of a problem has been a huge pain for me when doing my labs with customers + +Thanks James +Marking closed. +I made one more commit to this JIRA to remove some accidentally added eclipse +I will commit the patch with the agreed modifications +I saw this on NS +only uses Closed +Some typos were not fixed ajstatus sample status simple there is succesfully instead of successfully richgraphValidator there is typo when passwords are typed correctly in info message there is succesfully instead of successfully +Moved to Drools since this is not a Guvnor issue +Does that not work for you +I will defer to Stack +Id say commit +hadoopqa is not running my patch and I see that hadoopqa still fails on occasion because we are trying to copy libs to dir. +Ill sink the RC +At least one from statement must be specified in route definition +The current way of logging has the advantage that all log messages for resolving a script get into the same category regardless which implementation class is logging +Would be good for unadorned and ascii outputs +Hi Robert Im not really sure that this extension point could be used to manage a link with Crucible +This caused some samples to segfault occasionallyI have fixed these two issuesThanks strongly agree with what you say Bill +Provider using the using soap and setting the soapxml content type +I think youre correct +Looks good to me +Could you please provide some instructions +about I think we should fix IW to use seq ids eventually but I think for this purpose its fine +Done those listed are now admins and can add admins andor contributorsRegarding the up to you feel free to migrate the pages over and then ask us in a new ticket to shut the downIf they are to remain let me know and Ill adjust the ACL configs for them all but for the record please list all your to change and list current wiki names for each that will be admins +Committed to trunk. +Verified in CR +Am going to add same functionality to the validators page in that PR +Hi in our case the problem was found on the server side +The logo source for the one with the gray background uses a different font then the one with the black background is this intentional? Also they both dont contain the letters apache whereas the compiled one does +During controller failover we need to remove unneeded leader that the previous controller didnt get a chance to remove +it slipped my mind until a was reviewing the comments on this morning when someone commented on that issueI wasnt even aware that I had IE Microsoft must have pushed an at some point +or every operation is changed to specify the operation first +I would expect that method to be called whenever the variable is being discardedMy +Patch based on the code Ken posted +Ashwin +Ill provide a workaround on my side. +If you refresh the stats page everything shows up collapsed +I think those have been displayed beforeBut actually the display is currently strange +If the rest gateway is busy serving scan query and become none responsive it is unclear to the administrator if the none responsiveness is caused by admin task or someone had a large scan query for another region server +Stefan Ive added your test case to CVS in the modulescontainer area refactorig it slightly to work as a unit test caseThis issue is now resolved + tests passed after refactoring. +I think that this should be the default behaviour +Because of this issue when webhdfs is turned on the kerberos login fails +This patch adds a diagnostic message to the task to explain why it move from success to failed when a map output file can not be read +Fabrice I let you decide what to do with because indeed with Java we able to verify only complexity in methods but not complexity in classes +testcase works fine in my happen when Variables scope is disabled for cfcs +shouldnt we this change to the master +Sory for bad diagnostic I was reading a modified not the SVN to plugin code this is an expected behaviour as Eclipse before doesnt support MAVENREPO classpath variable for javadoc a property could make use of javadocs jars optional? Maybe an eclipse version property could be usefull for other goals in this plugin and could enable javadocs attachement download +Thanks much Ted! +Thanks Tao +As I said if it turns out that just the fit statistics are useful by themselves it may make sense to encapsulate them in a separate class and have have one of these as a memberThe ML thread I was referring to is the thread on with subject +Incorporated your comments +New patch after editing lines greater than chars attached +Will clean up and resubmit after I fix the blocking issue +Thats a good solution +Ok what way should this to be realized? I mean should it only do not allow selection in popup element or also not allow to input future date in input field? Ideas +Updated documentation on the semantics of CHECKSUMOK +Which issue had you opened beforeThis one! But if you want to give Korusaka credit for it because he submitted a patch then fine +did something change in the runtime or +set proper fix +AnyoneShould I provide more information? Or didnt I write the issue correctly? If the latter is true just let me know whats missing and Ill gladly provide more info. +Committed to both trunk and branch. +Prasad I left some comments on review board +Seems like the code is there already someone just has to cutpaste it into the trunkCould someone do that please +Given nodes X Y and Z start X +All the +if a server finds session data but the data has been reset it should also reset the session information +Thank you Kihwal. +Ive modified the patch to reuse the as much as possible +Uses an Executor to run all the get calls in parallel speeding up the startup in case of many calls +Thanks a lot Nicholas for the reviewI will address them in next patch soon +Backported to branch as well +Thanks. +One workaround is to copy the DOM nodes to a genuine level implementation +Thanks for doing all this work +Some of these little things took a long time to accomplish to give us a great deal of safety inside lucene internallyPersonally i dont think stuff like segmentmerger should be exposed publicly +Do you have any idea when we can expect the fix to be available as a new plugin +Also are these consistent failuresYes but only on jenkins! Ive had ec boxes running these tests all night and I havent seen a failure in over runs +I think it might make sense for Hadoop to become stable before dealing with these types of questions but I like the simplification +Oops yes the error is thrown +Attached thread dump shows the two threads locked +Thanks committedCreated in order to not forget about the dtests. +I implemented my understanding of the specification but we can try to relax things a little if you accepting the first style if none is the some changes in the spec are advisable but the very structure of the sld is a multilayer request in library mode how do you decide which layer matches which layerThe current has a clear way to figure out if we remove the layer name match requirement we end up without a way todecide which style to apply to which layer +We actually do perform encoding detection if theres no other way to figure out the encoding but thats because we have to support wonky legacy HTTP servers +a POM file and parent pom +Once config parameter is finalized for NN side i will include it inJitendra Can you please take a look? test will give you the answer for change in +I can confirm this +When nothing is provided in these fields is it not intended for trunk +Working on them +I cant claim that it is perfect but someone with experience set it up for us so I would recommend using those instead of the core bundle +Its really completely absurd to send Modified over the wire +Verified fixed at svn revision +didnt have any adverse affects when streaming +Just a minor problem with error reporting that came up while I was testing this +we shouldnt have specific version number in the gfx +cordovabuild poorly informs the user concerning how to build projects and what values are required to define a target +Im game for enabledisable flagI did some basic load testing over last few days +From it seems that we do not need this change depending on whether we want to allow negative ts on scan limits or not +For a a doc to make it through to the map one of its leaves would have to be not deleted which would be visible when getting the doc directly +Client calls from Netbeans is not working fails with same error message client from command line is working +Im not sure of the precise conditions that the original poster experienced but there were issues preventing the setting of the spaces for tabs preference from taking effect and being respected due to some odd behavior in the base implementation classTheres a now in the IMP runtime for this and it seems to work in both the case where there is a spaces for tabs preference and in the case where there is only the global preference +Closing. +Here is a patch +The existence of import statement under types element in the wsdl was making call newChild with as the root +Would you mind fixing it Alex? Other than that looks good to me +Try not to take this the wrong way but I could not care less whether or not your test case passes or fails +Please review the patch attached +This is a duplicate of +Lets push toAs the release are starting to get smaller Ill try to have a release every weeks +It was something I did while reading groking the btree code so I knew I knew how things worked. +Hi BenNo concerns just close the issue +to +Thanks for the detailed explanation of the alternative approaches +is the JIRA for the SBM implementation of Richards idea this JIRA would then be for converting any AS usage of SBM to take advantage of the new feature. +BTW heres a script to rm away the newly deleted jars from git since patch application doesnt do that +you can now use Wagon which should be smarter +Comments +In looking more closely at the latest patch I posted I decided to make a change that while in theory should do the same thing in a more robust manner is nonetheless enough to warrant a new run of derbyall before posting +I think lot of programmers would be grateful if this issue was solved +Not able to run the testsuite as at homeThe idea here is the edits till the seq id already persisted can be safely ignoredNow from this issue we are sure that the edits can appear in diff seq not necessary in the ascending orderSo all the edits less than the seq id from the store files are skipped and all other edits are consdiered +The problem persists for a different problem for as its not longer a functionfixer its a matter of making sure the correct static imports are declared +this issue is part of task +Edit for above comment Your approach which does not fix the other issue but is cleaner and a bit simpler +In our environment MB in files went from taking seconds to over minutes +eg so we in blueprint can let people use a non spring api for this +John Christianthanks very much for all your work this is a very important upgradeIve committed it with rev +This worked fine in RC +BilginSorry for the defaultPrice missI tested order creation from order view and yes you are right it works now other cases alsooBefore commiting the patch I would like to know if its normal that when you put the cart content in a shopping list you get the defaut config and not the config you set in cart +fixed as part of fixed in current trunk which will be. +So it will be downloaded at startup but if the user doesnt have any internet connection the download fails and SMX cant start correctlyTo simulate this behavior I have disable the repositories in etc fileA first way to fix this issue is to override the Camel feature XML by defining one in distribution which uses the correct Karaf distributionHowever its a workaround for now +Sometimes the prefix doesnt seem to be necessary +Makes a lot of sense +I would love to see Derby only support for unit testing purposes +And in many case like hive on top of hbase or a bulkloaded read only table I think its quite normal that a row only got one version +If you get this before you hear from me perhaps you can email me a search jar +So I do not understand how flush enters the pictureThe idea was for it to walk through the entire structure of an hbase installation and repair stuff that is broken such as bad entries in the meta and root region extra files etcIf you have a different idea please post it +in rev +It is much cleaner because of your suggestions +Initial exclude list for Branchx due to inactivity +ivy checks to maven central repository + to both lucene and luceneutil patch nice speedups +Yup we already have some base classes that I would use to build on +The lightweight components are no longer maintained and will be not usable in smx +I am bit worried about just coping the java files you have attached +Restore chukwa agent HTTP Sender back one revision +Hi Rob can you review this change please +Attached patch fix the problem +Committed. +The issue is probably in javascript code where is calling Seam annotated method +raised priority and scheduled it before the next this is solved by removing the usage of proxies +ill commit the patch tonight + +This issue may or may not be related to is getting errors instead of readonly +Heres a patch that fixes this by making the field private and using getters without the get prefix to avoid method name conflicts +isnt that really the purpose of infrastructure etc +patch committed to the trunk and branch +I think the whole skin context provider API should be refactored as it contains some quite major irregularities +Okay maybe it is Linux related +Sounds has the changes that we discussed +Unfortunately what you are proposing is likely to break things +This is great thanks +Patch puts an anchor inmonitor and one in! +Done in r +Rather we should probably drop the full build and just use the bare build as the single buildIn this case we would close this issue as wont fixValentin WDYT + +First test fail is because of OOME +we have written a conversion program and will slowly go through all components +Thanks Chris! +looks to and trunk +Congratulations KoenYou have created a th issue +Ondrej does EAP work? If not this is an absolute blocker +This will be done as part of the release cycle which will be early next year +This one should be correct. +Im glad that you agree to that tooAttached is a new patch which removes the method +Using assertEquals to compare arraycontents is now deprecated +I get no PME with the p closure can access x with or without def +I was looking at the wrong bug +Committed revision closing for +for the patch +I have added a wrapper for byte to add private getter for count and buff +This problem was fixed by +ATM had a better idea fixing this by merging to Ill do that and close that out when Ive committed and verified the build is fixed +Note that this is not just availability related +Searching nabble for surefire branch doesnt turn up anything useful +this is fixed in and because we can get all the pages and page versions backBut i thought in we have an option that the ajax request is only handled when it is the latest page on the stackAnd if another page is created that cost a lot that wont be on the stack yet so the ajax request sees that it still can do its jobthen the page that did take time put itself on the top +but no It should be this or Change this to that or even s +Verified by Jia Liang +Closing all tickets +Then I run Maven install on pom project from the shortcut in popup menu in the Package Explorer and it was also built successufully from the IDE and all projects were installed to local is not quite clear what problems you have with jetty +Thoughts +we still use the thrift api to talk to HDFS from some python scripts +Sorry I couldnt be more helpGood luck +Alejandro Any more details on the issues + +Is there anyway to move to grails version without using Streamcharbuffer +After some quick benchmarking its obvious that this will give us very nice performance boost too +Need to add method for creating unconnected sockets to so this socket factory can also be used with Remoting code +Thanks Daisuke +Does that mean you are going to extract it from Behemoth and contribute it here? It seems like it would be a generally useful thing to do +The patch to issue uses the method newInstance in +When another project uses the client that project will resolve the transitive dependencies of the using this shared POM +Sorry the previous comment got messed up a bit +wont fix this in release +Passes mustella for the dgadg +If there is a JIRA could you make the link between the two +What I dont like is that this needs to be maintained +issue has been resolved for over a year with no further movement +Actually running MOTO other transactions without the card actually being presentAnd yes having card present would save probably around on each txn or about on cc fees +this latest patch clearly converts this file to be an XHTML file +That sounds reasonable +Minor changes could be worked around by versioning runtimes that are defined in Designer but I dont think we want to play catchup until there is an expectaion of stability +I added some code to ensure it starts and all tests now passSending tomcatsrcjavaorgapachegeronimotomcatSending tomcatsrcjavaorgapachegeronimotomcatTransmitting file dataCommitted revision . +The configuration entries are a way of configuring JAAS login configurations +Ill take a look +So I am closing this issue. +I tend to agree with your last comment though Werner +This seems ok since rampart is only coming in during runtime +As Juraj said this issue cannot be reproduced anymore. +Hold up a bit +Ill commit this if nobody objects in a day +the patch was already added to trunk dpalmer will get it into branch +How about this? good to me +BenShould we go ahead and move this to be a issue? Given our discussion last week it doesnt sound like you will have availability to look at it until next month after should go out +About be sure to trim extra spaces around the word true Is anything except the exact string true treated as meaning false? In that case extra spaces around false would not be a problem false would mean the same thing as false though not for the reason the user thinksI dont know of any property other than that cant be disabled once enabled so I will document that as the only one unless I hear otherwise +Danso at runtime the cxf bundle has no access to the classes of the application? I guess this is quite normal in osgi +add as affected version. +Using gunzip didfinally work +Tested locally and on cluster +Also changed issue type to New Feature +Output from make on my the wholesale deprecation is only openSSL I suggest we just turn the Werror off only for onI note from Andys log that the compiler is actually clang not gcc so Ill give that a try on my box and fix any obvious problems +Dont have time right now to complete implementation so if anyone want to submit a patch well be glad to accept it +Also need to remove the JCP notice from the docs in to the JPA and Bean Validation Spec from Geronimo which match the just released specsremoved the early access notices now that the final apis have been released. +removing the shape attrib from anchors but thats the responsibility of the application +environment is Eclispe tomcat on Ubuntu bit +There is also the notion of lover of lost causes esp +Tried this out today +I mean That is correct since is not mapped + is is what the issue complains about +it is a pattern too hidden from the user it is difficult to know then if it will close or not +Its a very strangeimpossible situationthis was in CHMput is it possible you were passing a null key when putting Ryan +Otherwise the value is returned +Could someone please close the lists? the lists been told this is happening +changes from review comments +Needs to be in same stylemax line length as existing code +Added fix version see Make sure all fixes on branch are applied to branch +Sorry I posted that comment about the patch not applying cleanly to the wrong JIRA +current behaviour is of the testWhen message is received under a XA tx and when the JBM is prepared and not committed the same message can be received by another consumer in another behaviour isJBM should hold the message until the associated transaction is committed or rolled back +Checked in with revision +Some plugin might simply be there to insert some custom code on startup. +sounds like a great way to make things very very fragileI think you could use snitch since the performance of the node doing compaction would degrade and dynamic snitch would naturally prefer the other replicas + order to be consistent with timestamps elsewhere shouldnt the changes read + +I must admit that I havent read RFC +So now you can use either Serializable or isSerialiazblePersonally I think its fine to do as we are +No issue until now +I thought that was the first thing done +Actually I did not compare the results of my implementation with those of Rs I purely implemented the algorithm described in the paper and checked that the results are sensible +Attaching v patches added more fixes to tests related to timing +Fixed the check on maxSize +I added the patch with adjustments and a rename to be a little clearer as +I think I am still going to check this in as the current code path is broken as well and this is a step in the right direction +The patch looks good +Thanks for reviews Ted +Plus we dont want to abort do we? Just disable Java? I think this fails with ant +I dont see a similar advantage for the which parses the strings directly +One issue I can think about this approach is that more checks may be necessary on the client +Thks. +Can we close this issue now that has been resolved +This issue has been fixed and released as part of release +visibility is now checked in Sander for the hint If HTML allows tbody without tr we should just keep as it worked beforeReusing tbody for header or footer might get too complicated +Please check documents only for set to resolved +removed Affects Versions from ticket this version does not exist yet and cannot be affected +In general users should not use this class directly so it was initially put to internal +If we still need to address something recommended practice is to clone the issue and set that to the new desired release +Ill let you know shortly +Some JDBC drivers accept placeholders in the form ofparameter or as well as the standard +I am from a java app to mainframe +Fixed. +I think it makes more sense to put it with that issue +oops +click create a new attachment rather than pasting it +If you have millions of docs and no conflicts it will must scan through every doc meta record just to tell you thatAnother problem is you dont get any filtering or formatting +were going to start on a unit test but give the document counts and the corpus were testing against it may be a while so i thought id ask to see if anyone had any hintsremoving this patch seems to remove the issue so i doesnt appear to be a lucene problem +Ill take a look +Hope you dont mind +Attachment has been added with description zip containing good and bad svg files and pngs at and sounds like the dpi switch is working fineThe dpi switch changes the mapping from realworld units to userspaceunits +We are looking into this +I will be watching the issue for possible comments +Committed revision +Attached a preview of release which the fix +The exact same scenario +Quoting the same source If no type is indicated the type is string +Added two samples to start with +Completed At revision +So basically Groovydoc has no way to hide annotations should you not want them shown. +In other words classes from jar project output folder will not be available when war runs on a server +Fixed at the null check was added it was determined that wed like to keep away from making that assumption that anonymous authentication isnt adequate for oauth authentication and allow application developers to make that decision at the resource access level +is licensed by AppleLast two are python scripts +and by should be be grmpf today its too hot in southern Germany for bugtracking +It looks like a bug in the backport library +According to the reporter this issue can be close all issues in resolved state that have not received further comments in the last month +Check for null solves the problem +This one should be closed as fixed +Thanks Knut + +Diwaker The usual time on a linux box is around mins or so +See if it fixes the fails +Committed to trunk +It is true that the change is not complex but we are now adding more threads to the story and the change is not strictly necessary +attached migration guide commentsGo ahead and commit it using the same naming conventions as the previous +Emmanuel is working on JPA integration assigning to him +This would be a maintenance problem? I agree that as a developer it would be quite easy for me to add SNI support to a Java app of mine +committed to as part of done by Rick +Update the patch for this to include fixes discussed in dev around the URL template factory class names default value +But thats not a clean solution +Feel free to reschedule it to another version of installer if it cannot wait portal Portal with JEMS Installer +Ill commit this soon unless there are objections +The only thing I could see was a couple of additional newlines at end of files but there were also typo fixes in the files +committed +applied your patch at revision please verify that the file appears correctly on website +so at least youd need to try capturing the network traffic to get a clue of whats happening +I have noticed that sometimes the new client fails to start properly so the test sometimes incorrectly logs failure but this is obvious as one of the members logged will be shown as null +What do I need to do to see these compilation errors +Trunk has been updated in revision +Write Count Write Latency ms +Can you give me an example to test this patch out with? Also Im a little confused it seems like this patch only contains commented out code +No problem Vinod +Failure was in which passes locally +passes based on and +Upgrading to is also much easier than doing all this dirty hacksWe already effectively ship the with our jars since anyway + folder structure is src main +is a complete patch +Hi AndreasPlease refer attached pdfsIn attached file I get blank images while fetching images from pdf +Method get of in RF can return prameter values of type String +Just adding a field should be fine +This would probably speed up the build a little since it checks the local repo first +Thanks +Its pretty small patch so Im giving my here. +So I still do not think this solutions is the right +Well its not that it is of high importance but generally using the standard directory layout will make the project easier to understand for newcomers +There are a few lines in the patch though +rerunning test +When I do this I need to remember to take into account comments in is a note for myself I also need to make sure I touch and +So it could be used for future ant +Fixed in Gwen! +I tried to apply this to my local image but it seems there is a small problem +This is incorporated in +Running as a admin or root is a requirement for platforms EWS supports +not ready working now with the two attached patchesThere is one bit of strangeness with this right now though +Closed at initial requestors request after an alternative approach was suggested. +Did you enable Tapestry compatibility mode as per the documentation? In compatibility mode all Spring beans are exposed as if they were Tapestry services +I dont know the specifics yet but my client is still having issues in spite of this patch +There are also unit test for each of those in and +BTW I dont think it is related to I managed to have it on but it is really seldom it is probably related to some timing or network issues +Im afraid Im busy and dont have time to verify the fix +Two more sections to go Error Responses and JSON Schemas +New code is quite unobtrusive since default implementations are provided so the risk of committing this patch is quite s your policy for adding new committers? Im already an admin for a few other projects here at Codehaus +I think Erics patch for detecting staleconnections is the ultimate fix for this bug +The laptop one could use a SSH tunnel to to send email but it would be used for ANY email forwarded not just for apache addresses +Hey Dejan when you attached the patch you did not select the ASF Granted License option +Will notify when done +If it doesnt then make one cutting and pasting from the confluence page on db its quite good +On the other hand its ideal to store host data +Hence I have provided another patch that does the version changeIf we do the version change then a version change has to be done in the Openejb too +This is one instanceof of the more general bug reported in close all issues in resolved state that have not received further comments in the last month +Ill create another Jira for the Stomp Size Limit support feature +coho does not run on windows tagged via command line git +Now its working fine +Thank you for checking +Hes moving down to the original JRE with the sigar patch +And lots of jobs back to normalIt will be very nice for me that if we could release jBPM as soon as possible +Sean I do not created comments and just made modifications using Track changes in document +Probably can do the same here +There are a zillion options and tweaks you can do in the maven bundle plugin so maybe Willem can help you to improve that in yourI assume you have used that to generate the osgi metadata +Since the benefit of such a check is IMO only minor Ill rather add a warning when a foreign key without a name is encountered +That proxy has the retry mechanism and hence you might run into the situation you described +Its better in hcataloglibs so that you can have it for different versions of hcatalog and will depend on of that version rather than a pig version +The tgz file is accordingly and uploaded +Changing this bug to only be in the php component +Yes I think parameter references will be pretty much required +Logged In YES useridFixed +Ive got the helloworldwsclient sample working with celtix binding with some hacks +This happened since changing and doesnt seem to happen error +IMO since I have specified in the configuration file that I would like to use action I would expect the framework to at least ensure that the element is present in the SOAP Security header and that it is valid +Still doesnt fully test the JMS connector so Im not committing it yet +I did what you said and that error was gone +Is format committed yetCurrently is still considered the stable branch +Reformatted for readability and removed the questionable link +archiving sounds good +The code doubles the output buffer and tries again if there is overflow +It was released on October nd after the code freeze +Attachment follows +Works with JDK need to talk to DTPJohn Graham if SDK is requirement for DTP +Thanks Tatyana Aleksei the patch was applied to BTI branch at r +Update the case only when both Product and Type have value other than the first option +Were seeing the same exceptions and will fix them using the +Fixed thanks for the patch +Long time back Ryan tried to a custom compression before putting stuff on the wire and then undoing it on other end hoping it would help some w latency but he found it upped latencyI remember that too but if I remember correctly it was going through the whole codec stream interface which necessitates extra copies etc +I see three options here someone that actually uses and can do java development steps in and fixes the issue that someone cannot do coding but can get paid support with a company that has the knowledge to fix this we wait for a core developer not using mysql that has enough spare time to burn away installing configuring it and figuring out why the sql views are not two users confirmed the jdbc driver upgrade is effective +Since we are planning to merge LSR branch onto trunk next week it would be better if this patch directly applied to LSR +Cool thanks +Eugene +Fix is in CVS +Removed method getServer from example can be compiled now but there is same issue as it was reported in +updated patch to work with latest good can definitely extend the comments to make clear its a Geotools part is fine same advise applies to the comment +Do you need help with this? I can ask my former staff and see if he is willing and has the time to help out +PATH usrlocalsbinusrlocalbinusrsbinusrbinsbinbinusrgamesLIBRARYPATH and DYLDLIBRARYPATH are not set emptyI just logged in and used echo to print the values to the consoleor should I trace them from the start script +My script using Scriptom works with Groovy now +Josh formatting was never implemented for or so yes we know it doesnt s high on our list of things that need to be done however +A unit test for the LJR will be helpful as well since its very easy to break with runtime changes +ah ok I will check the documentation and in case add this information thx Freddy +Ive updated the patch +v is just this stuff +Thanks Amar +Closed after release. +This sounds like a utility method making it into the API which smells wrong +its in albeit in a bit differentsimpler form please test itif the extra complexity is needed i will add it upon you its in again +The fact that a Job can launch itself without any state is interesting but irrelevant +Updated test patch +Fixed a typo in the issue description +At operationssecond we will take about years to roll overStack points out what about bulk commit this would then wrap up every bulk commit as memstoreRWCC transaction which might be the logically wrong thing to do and also the extra time it takes to insert that much stuff into memstore could also be an issues with RWCC +WDYT? +Replaced id with local variable +Thanks for the review Eli +That would work +As long as the check is performed by the process Chriss approach is simple and portable across all permission schemes +Found a bug in srccontrib +I dont think anyone is working on this if you want to take a whack Sergey +Now extends +Change final public static to public static finalBTW changing IPCCLIENTCONNECTIONTIMEOUTKEY t IPCCLIENTCONNECTTIMEOUTKEY is a good change +is the only jar file that I had to grab separately from the jar file repository +Is this not fixed in trunkIt is +I will run regression testsTouches the following filesM javaengineorgapachederbyimpljdbcRaise an exception of dataEncryption is set but is not set to trueM javatestingorgapachederbyTestingfunctionTestsmasterVerify that an exception is raised if dataEncryption is set to something other than true +Hadoop has something like this and it has worked out quite well. +Remove flags for inline +all issues assigned to +Great work! Very useful indeed. +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Committed to closed. +I guess it may not matter though? Im just worried it will be like pulling a stringthe issueThe test was made to test but that class shouldnt have been made and the methods from it have been added to Spans and its been removed +sets timeout on client socket connection +IMHO you could merge the branch back into trunk we are allowed to experiment and the current patch already improves the API and shows what to do next +So one thread could create and many threads could read +Attaching SA for testcase described above +closing. +OK event is better but how about or +In the failure output there were the compensate both for and + +This condition cant happen in the guise reported above because too much has changed after went in. +There +Currently you have to go to the download link navigate to the right folder see which is the latest archive file and parse the version number out of the file name +Committed revision missed some things in the contribqueryparser when doing this +Teiid Designer has committed a fix +Done. +It seems that throwing away all method stats would truly be resetting it +Additionally has been fixed by marcus moyses +I will commit this and continue working on the code +This issue should be marked as fixed closed +This is important to know because it will help decide on whether or not to do improvements to the existing delegation tokens that were planned +Please verify +Changed JIRA title from to +The same file may be used when it comes time for the update site on the web + +Hence closing the ticket. +this to trunkChris Karthik when trying to commit to Im getting conflicts in the POM have all the subsequent changes to plugin gone in +According to that bug it seems that Cocoon use of could be improved in our input modulesI spent another day investigating to no avail +all issues to the CR +Perhaps a refactoring of some of the writingreading methods to allow for usage of them may be useful +I think its pretty confusing to not have it propagate +META +Can you put forth the use case why we have two different of exceptions and? For example can you pl explain why is a +patch contributed by Sanjoy Ghosh and contributed in r +Newline EOFWe can easily add that if it makes life easier but due to lack of WS preservation we have to pick one or the other +bulk close of resolved issues. +You can simply deploy this web archive to your EAP server but you will be unable to configure +Please evaluate +Duplicates a duplicate +There was a test difference between and trunk +I will see if and can be moved out from later +Will create a separate ticket for this +As per my understanding SYSSTATEMENTS will only has two kinds stored prepared statements for metadata queries and for trigger action plan +Much better thank you. +If you want to change the default port I am fine with that I am just curious what other web services do you expect to have running on the same node +Yes it works out of the box and I am working on the docs +Lets fix this benoit +If possible this change needs to be made in the cce release versiondocumentation +Suggest you chat to if still a problem. +I dont have a windows install handy to try it out though +No need to excuseMaybe we simply misunderstand each other +This issue is fixed and released as part of release +Preliminary patch it does not include a test +This isnt allowed given the state isnt CLOSING or CLOSED +Thanks Philip. +Thanks MarkIll add them asap +Need to complete sample app is a sample Lists in example that demonstrates binding listcontrols +These are the diff files containing the necessary changes +Vinay please confirm +This was fixed in +Then that model was used to restrict how users can addremove items in the Synapse configurationSo it is my opinion that this feature should be implemented outside of Synapse +Fwiw today make test hangs at +It was not working in +Really not that big +Added to Release Notes as resolved is now in CR +patch applied +Username letiemble +So why should the Complex class handle it differently? This is addressed in the C spec and various associated discussion papers +And thanks for your insight on NIO Steve. +Allow component to have width without units treat it as px +Also going back two comments to the issue I imagine the inability to run it directly in the editor is also what stops the Java Application context menu item from appearing +Regarding maintenance I suppose Ill have to diff the file when upgrading in future inevitably there will be changes there +If and when there is renewed interest to pick up the AntIvy stuff this can be reopened and completed +I know that it is a really tiny work in comparison with the great thing that Wookie team has released so thank you for your encouragementTien +So was this goal not attained? If so whats the benefit of libwebhdfs over just using the existing libhdfs with a webhdfs URI +regards for applying the patch. +This is WEBEJB application to run httplocalhostbugs test case produced +the original patch caused fail but RI failed on same test so I considered this was test codes bug and fixed it +I have no idea why we have to send those extra four bytes +Unfortunalty we havent had the chance to get our hands on the original Spring logo +Im not sure I agree with that +Thanks Bryan! +We can just think about how to fix it when you are back +Committed +Updated in ESB codebase will be in next merge +Could you please chech that encoding of srcitprojectssrcmainconfig is and that the french character was not corrupted by making patchuploading on JIRAapplying patch +Third version of the patch which addresses the review comments +No its fine it just needs to be documented for the software grant +The primary purpose of allowing the connection was addressing this possibility +Yeah this basically corresponds with what Jonathan outlined but I think it is sensible +These would need to call back into the Java piece +OK thanks again. +not sure if this is a good idea or not will put it in the queue and think about even bump to is not used anymore +well +can you check this issue with Maven Ant Tasks orIf the problem persist please provide a sample for me to reproduce the problem persists please reopen the issue and provide a test case +issues after release +i still get this error with springide final +i am implementing the recipes in cShould the C version go in a separate Jira issue and patch +bulk close of all resolved issues as part of closing items. +The Configuration chapter comes way too late +Using includesexcludes was my first approach but the compiler pulls in the transitive closure of the classes which means that you get duplicates between the two jars +be honest I would have expected a generalized codebase to not even mention the word +If this is right Daryn Please make the change? is as a version and am changing our to reflect thatThanks! +This will allow us to make forward progress +More clarity for the above scenario May be there should be a provision to update the role for a siteminder authenticated user based upon some input like a siteminder header variable +c A tag that is used to storeretrieve a instance for nonces +But the difference is that the main build doesnt run the clean target where the db builds do +Others are treated as extension elements. +I just reviewd and tried to merge and there are too much conflicts to reasonably merge them by hand +resolved by adding to additional parameters to call to determine +You should should be grouping by your projected propertiesYou can see an example in. +Can you describe the problem better +Ive added a new test case +See for the test +Fixed +AFAIK this is supposed to be picked up automatically by Jenkins +Thanks for pointing that out +Committed to and trunk. +I am using for testing copying across dfs +Many thanks Stepan +yes however if the patch was accepted it wouldnt be available till and we have institutions going live in about weeks code freeze should be in weeks +writing your own number class which can extend as needed can be done easily in Groovy +The actual buffer size is bounded by good to me Simon +as you described there would be only one ensemble used to write at that time checking current is trivial +Its another choose for usArvind collision also can be found at different bucket writers in same Sink +I think this is a problem with the test which has been exposed by Robs performance tuning work +srcorgjbosstoolsjsfuieditordndsrcorgjbosstoolsjsfuieditordndsrcorgjbosstoolsstrutsuieditordndsrcorgjbosstoolsstrutsuieditordnd Struts Diagram Editor doesnt support xhtml files at all +Thanks for spotting this +Writable isnt a of Comparable +Bugzilla has not done justice to my above html renderings + patch from William and I +Do you have the DDL for the table you are using? Thomas +The protocol is to fix bugs in trunk then later discuss whether to backportSince this is an HDFS improvement Ill move the jira over to the HDFS project +Thanks for catching that +Im working on a Linux machine +Proposed patch against trunk +Heres a new version that includes a test case +Deployer has access to the +The include parameter was still retained for the sourcejar goal so that the user still has the option to generate both the main sources jar and the test sources jar without executing separate goals +Yes this has been fixed +Window Show view Other error log +Looks like this is included in +It was due +Solaris gcc compiler +I will document ongoing findings for those interested +I have updated the summary of the defect to indicate that we need general support for multiple executions +jar is on the path when rails console starts up +I have some fundamental issues with using SQL globally to try and factor out the duplication SQL does not use at all it needs to delegate to to encode column names literals table names etc +Perhaps the installer plugin is what you are looking for +I am okay to make it in trunk and then merge it into this branch later +Think I added wrong file +Failures are unrelated to the patch +Let me see if we can produce something simple on our end but it wont be any time soonAs for the number of sessions we believe that we have established by a myriad of test cases that the problem is reproducible with one or more sessionsTake careAntonio +Please use the mailing lists for support requestsBTW your example has two different SQL statements not surprising that the results are different +Im in the process of cutting down a minimalist set of classes from our app that will still produce this problem reliably +calls inconsistently some using and some using some with and some without +Committed +Still need to make the sortProject method public to make it easier to isolate +And using a function to get the substring of char +Closing out as the tests are successfull +Adds includeguardprefix option for cpp compiler +This is the default permissions scheme +This is an issue +Attached the final pull request. +Thanks Joe your workaround works fine. +OK this is a bigger issue than I thought +The patch was built into the and releases +I havent yet come across a use for it. +This is a patch for the Check and java files only +I think we should stay with className +Reverted the patch +Yes I see a reduction in boot time. +That sounds like a separate issue +Please if youre still hitting an issue with bean services on AS +pluginsServers however can be and in the properties page can have the option to be stored as separate files +This looks like a duplicate of did you try the latest SNAPSHOT +I will test it out and update the JIRA with the resultsThanks a debug print +Updated the existing profile in assembly module by providing the correct location to unix and windows source assembly descriptors +M +Right at the onset let me submit that there are not unit tests done since I am only trying to check if I have got this right in the first place +If so definitely scratch that idea +We saw garbage collection occuring but the on the namenode remained within its normal limitsGiven the above I propose that we accept this patch into and trunk so that the immediate performance bottleneck on the namenode is fixed +Hi ClausPlease find attached an update patch for the supportI have modified the code to make use of the and use its APIs to get a reference to the scheduler +Looks good. +Patch looks good +Use softlowerlim instead of user defined value for load position +Please advise when this is t o be offically released. +Thanks Brett! I will add those items and submit a revised patch +Quite an extensive refactoring +Couldnt create temporary fileError Couldnt create temporary font fileCouldnt create temporary fileError Couldnt create temporary font file its a problem related to your installation +I would like to contribute +To fix this class should use to addremove connectors so that the changes are not lost upon stopping the server +Anyone have a sample project that they can upload that reproduces this +Closing thiswe will deal with unstable tests in separate jira once the testsuite build process is stable in ER. +he is the patch that includes two files that were left out earlier +Note what im propossing here actually breaks in a way that i think is goodat the moment if you declare multiple cores with the exact same name solr will initialize the first then initialize the second which will shutdown the first in effect silently ignoring the first declarationIm proposing a hard failure during config parsing before any are ever initialized +But this one is for trunkThis two patches have been merged in order to keep them consistentThank you +why is commit called on closed connection this needs to get verifiedI dont see this being caused due to change in +Just wonder if the should been named +The test uses in which the token codebase is replace by a URL pointing to the outputresources was an extra that confused Windows +However it would be much faster and easier to have the fixed data in the test program and use it to directly populate the +I am not sure if I fully understood the solution above but doesnt look like a clean solution +Nothing in the previous documentation was in fact incorrect but this patch makes a few things more explicit +I did think about leaving the interface in the engine bundle and youre right that its more reversible than putting it into Sling APIWhat made me wonder is that all interfaces in the package are deprecated IIUC thats a coincidence not a willingness to have the engine bundle export no interfacesI this case Ill move the interface to I agree with your point +Closing this one in favor of the more recently reported issue +Try addendum through hadoop +the first step +But the thing is it wont disable the scheduler +Above comments is for patch +Latest patch generated from my github branch +its not our job to be a BPEL tutorial +All resources are merged into a single before the job file is submitted +I think Serializable is necessary +What i want is a simple ant target testjarThat sounds like a fine thing to have +The expiration of the message can be configured by specifying the property within the ESB message the existence of which will cause the delivery mechanism to invoke the set method on the I appreciate this is not handled by default the mechanism should hopefully provide a means to achieve the result you are after +Modified patch to reduce code duplication in the Delayed classes +It is used from target +There isnt any test for this in the current codebase but I verified that it works in my projects +Ive read through your explanations and almost everything seems fine to me but putting in the same package as is not necessary bound to maybe better under util package? maybe better renaming to +It is unclear what the correct behavior here should be +thanks for adding faron as a mevenide committer for following the process! +Otherwise maybe it is Windows XP specific +Only the first screen on the preference page is showing and for now it allows you to set filesets for each runtime type +data length is always growing first so the solution here was to track the metadata length as visible block length +Im not sure if we want to set the target to or at this point +Thanks Russ! +Here no path proposals +Any feedback? Shall we upgrade to? I tried it out for a few days without noticing problems +Im not buying this you never had a problem with a public user database on the Hibernate forum for people sure for some reason people suddenly provide their real name which is great but we dont show addresses to guests obscurity is not ve hidden addresses for so the information we show about our user base is the same as on any other community website +Camel is being cut later todaytomorrow +The serialized filter has abstract methods or it implements a class that now has abstract methods? Let me ask Jimmy the fellow who made the serialized filter +I just filed for this +I have also tried it with the version of the plugin and it also fails +We should have a committable change to Struts head soon +Maybe it was merged to older docs or something? +Please verify and then resolve and close +Sure +Oops meant to leave this open for is janb on. +I dont think we care about starting a core that doesnt belong to a collection we just want to get to the point where you start a solrcloud node and there is no predefined cores +Not going to happen on the ncl cluster +Changed to ignore classes which is expected these changes were actually for +embeddedejb +Ah just seen your second comment +It turned out had nulls in arrays +Patch attached +Sounds very good +I thought the consensus was that this wasnt appropriately within scope for? Can we close this issue +Im pretty ignorant of Internet memes for full disclosure but someone recently showed me the Youtube video behind this one and frankly its one of the dumbest fucking things Ive ever seen to use a technical term + has the list of files missing Apache License Header +The previous action is pushed onto the value stack so it can be available for the action being called +All security scenarios are working +We suspect this was caused by +I am glad weve agreed on this +Thanks Jian +What does DELETE session actually do then +no it is not obvious that this would cause a problem +You dont need to solve them but discuss them here with us +Aurelein Can you please provide an example +Still need to +I would like to get this in but am a bit swamped +Patch that removes deprecated usage of from Lucene Core completely all tests also fixed +v removes the extra description around DSC +That way you can also get other validation errors + change the type +Ive modified the code in trunk to connect to the SSH agent and look for private keys in +It will of course break backward compatibility though because you pass an entirly different path to the binding impl but I dont see any way around this +Trivial patch for saving autoCommit value not applying until I see what other changes John was thinking of which might make it irrelevant +Please submit the patch for verification so it can be committed into the trunk and merged all the way back +We dont parse m. +Resolving issue as it was reopened due to a bug in Doxia +but i do not know the state of this issue +Updated and reran local build +The attached patch gives permission to write to +content is truncated as per Attached the patch +But I created this issue to add it to the examples +slightly updated version of the patch which fixes a couple of missing exceptions think something is wrong with the patches because my local version has already the changes to needed by the and when I try to apply your patch those changes are rejectedI will create an branch and also commit my current state of to Rupert it looks good since Solr has just been released Ill try to upgrade everything should keep working +Change the semantic of startDatanodes so that it adds new datanodes to a cluster instead of starting up a new set of datanodes +I still think should be available without additional configuration for instance just yesterday I had to remind a coworker to copy the properties configuration after setting up a new systemBut this is working and I dont want to be too pushy with my first reported issue Thanks a lot for the help +Recursively encodeeach child of each child +I will try to switch to laterRobert good idea for an extra +LZO is not part of hadoop any more +Yes its a hack way if we change the interface it will effect the users bean which implements interfaceAs we working on the camel its a good change to add get method in theCurrent can see all the beans in the application context it is hard to tell if the s camel context is belong to the beanso I add the checking in the the to skip the camel context injection if the bean already have the camel context to avoid the injection of a wrong camel context. +I press CtrlC and VM finishes +Ted No +Rn +Thanks will update back when i have done with testing +Cool thanks Andreas! +Vlad did this happen with or? Im trying to determine if it was a problem before and were fixed +I will improve and attach a new patch +Thanks Ill hold on until the gshell patch is ready + +Log the whole stack trace in the message ugly but items not being worked on for M afaik out of issue out as part of JIRA cleanup +Please see above +You might have rejigger on commit because I just added some extra logging +and committed to branch. +Committed to and +Thanks +Dont think its a good idea integrating this into Click itself as everybody has a different preference and will make Click bloatedI think a better place to host it would be clickclick +The new plugin plus the update sources seems to have fixed heaps mate! +It is rather interesting that out of all the pairs of methods that I have seen where one has only used in tests comment and the other has proper javadoc the former would always be the main one used in real code +The directory itself has the same name as the database it represents This is a control file with information about the databaseseg This subdirectory contains the data files in the database one file for each heap and indexlog This subdirectory contains the recovery logs needed to keep Derby in a consistent state after a crashWhat I need is a zip file containing everything in the log subdirectory +the current patch contains also an test +Closing as per reporter comment. +Nothing uses it inside PigLets remove it +Attached patch to fix this issue +With the patch all acceptance test passed +The only exception is If you are adding a new parameter to an existing message then remove the message from the other locales +bq +Doug its important to handle quoting + has been created +If you find further issues please create a new jira and link it to this one. +r commits changes to trunk related to shared properties +list Schema works as expectedAnd I dont think its worth bothering adding special cases for every reserved words +Mikhails patch was applied at the revision +bulk defer of bugsminor issues from to +I like that idea but what others would we put their? I had an EAP version created so if we update again well likely need both AS EAP there +For the forum app please not +Bulk closing all of these resolved issues. +Fixes committed in r. +Fixed on trunk as of revision +its bit like reading blueprint of a lunar module or somethingAt least compared to relative simplicity of lzf codec to wrap within frameworkSo I could use some help in figuring out best way to properly embed lzf in there including ability to support splitting. +The validator variables passed to the server side validation rule are being lost +a runnable test case +The attached patch updates the sample to use both the dbreport and dblookup mediators +The issue must be resolved elsewhere +Closed after release. +Thomas I noticed that you marked this a Coding in Progress are you working on this or do you want me to do it +test code this time with granted license +NET use the assembly when targeting ist there any problem with it +Bug in script for tagging is my best guess at this time +Manually tested +this was resolved by releasing plugin tools and reporting impl +created in toplevel dir +I have merged in those changes and run derbyall cleanly +Go ahead with option +And you can not check with java script what the use has selectedFirefox may upload or download up to GB but only when the Server does not keep alive +These Java bindings are required by VPE +patch to allow FTP imaginary files to have children +Damn it +But this is a little silly +My problems were resolved with the final release of Spring please resolve if appropriate +Hi Emmanuelwhich is your pageThanks a soon as you have some feedback about this patch please let me know +Reopen to set fix in +Viktor has added for it. +Fixed in revision of release. +on each nodeNow the cluster is up +As for Closure as super class +And the last part with the fixtrunk +Just tried logged in as sebb and I get the same error +Because of that change the focus shift from main console window to dialog box and back wont happen you please check if you are still seeing the issue reported here +This is an example +Just saw this test failing again inDo you have a patch Himanshu +Having said that if you are wondering if I ever tested this YES I have done so +Yes but the fix would also load the plugin immediately and perhaps increase startup times +Attaching the output of I will have it up shortly +BTW FYI +This is now fixed with the implementation of can specify this tag inside specified it will override the targetElement for feature chaining purposes therefore when chaining the type you should link it by the mappingName as willCommitted revisionCommitted revision +upload a script for mv file and post script patch plus the full patch for latest patch looks good +Well yeah thats not the whole story +JerekI tried your suggestion +Bulk closing for +This should be clarified in the documentation +I would really need a test to make sure I can fix this in newer versions and prevent future regressions could you please try to simplify your entities as much as possible so that we can understand in which conditions this is happening +Also it could not find jetty libraries for +Fixed it w. +settings and created the src directoryclasses is a derived directory that doesnt need to be in an exported archiveHowever I still think that there is a problem in the because it would have to handle old projects +Thanks Navais and issue is fixed and released as part of release +Thinking more about it copying the raw bytes from the page array would probably only be cheaper if the materialized values created by copying the keys were not going to be used later +Vsphere viewing instances login this issue after checking with person who reported the issue +Fixed in rev +I strongly recommend turning it on. +At least with releases it would probably be better if all versions are in sync so the user knows for sure all artifacts are part of the same deliveryshipment +Found this when creating navigation nodes via portlet +Other tickets require this functionality +If you could test it from SVN then we can close this +Jenkins didnt pick it up due to build issues +If you can add a unit test it will be great +The only classes that should be made are and +I think it can be removed easily because new takes care about dependencies correcly +Do you think that this bug could be closed as duplicate of +This patch is very simple still applies and the issue does need to be fixed +gitignore needs to have these settings to work properly for driven projectsWe could improve it by checking if these files already are in git before adding them to the +jaxrsspringsecuritywebapp +Perhaps you could be clearer about what you are trying to achieve with your code +Tijs or Joram can you give a second opinion what we need in order to accept a patch for thisIt might be a good strategy to implement this as a configuration option +This change seems to break testthe problem is that the pendingmessage cursor is reset on activate but with the default keep the state in that cursor about subscribed destinations is lostI think we should not call if keep is set and we have been deactivated +Thanks +Marking LATER pending course of action +Thanks for letting me knowI will update the patch and exclude thoseBesides the Thrift classes are there any other identifiers that are auto generated +In case it is important We are still working with geronimo here +bq +Rick thanks for the patch +The client was working because it was not entering passive mode +The order of columns could be a problem +Maybe related with +The problem disapeared +It looks like the issue is how the associate sites are being added to the p metadata +The last two tests fail although I would expect them to work +The goal of the optimization was just to skip verification of the block which has been read by the client +By simply looking at your log it seems to me that the datanode sent an ack with two fields but the client expects an ack with fields +Submitting another patch which escapes the +yep they should be unpacked does that not work for you +Noisy unclosed issues are a bit of a problem +We should support back ward comparability we should not fix it test caseThese test cases are working fine with builds +And the use of doesnt cause any problems in the production code even if the class doesnt exist in the insane jars +We already have the machinery in place for almost all of thisThe only special case will be zkhostbq +I added a new one with the relative suffix that contains relative directory m going to change the return type of all methods in the to +is available in exporthomehudsontoolsperllibsiteperl should be able to ssh to the box +Not closing he issue +after Eclipse in eclipse the reason could be someone in jbds dispose some widget vice verso is possible but exception is in eclipse +free cancel event only in afraid the v patch still have race in event cancel when the cancel thread set the cancel flag but not set the inthecancelqueue the thread own the event do thedequeuetimed +bq +Im marking this as WONTFIX for now since JVM cant provide a reasonable temporal granularity required here. +push to +to Todd for +Bulk move of M to move from M to have never been able to reproduce this issue during tests +exactly the same happened for us +Attaching when starting the server fresh with already in the deploy directory. +Ive just made a fix to the archetype in support of anyhow and +Thanks for trying this patch +The documentation has been updated with the info on the new type attribute +The groupId is now changed to. +Close after release +Choose whatever is easiest for you +Committed Mikes patch in r +Replace the driver in the +Duplicate of. +Looks good to me +Would it really be so bad if someone adds a maven plugin that can just download a jar file from any http location? I could call it ? +See for changes necessary to enable plexus components to use the new separated container API and standard implementation +There is no response to this issue on the forum +The focus should be to get out quickly as it will resolve a blocker +Mostly involved textual output and temporarily old methods +FFS hid the problem by the embedded fs albeit with its own uri which is wrong +I am of course borrowing the idea from the information we have about when the last heartbeat was received from a TT +one more comment about the I think we should put it into lucenecontribmisclib instead of lucenelib since we only need it in there though +defer all issues to +ok +No API change +Please otherwiseClosing the issue. +Hi Stefano I tried to run this message through with your change and it still failed with the same error +committed to trunk. +However i was not able to reproduce that recently so i left the debug logging in case this should happen again. +However we can update itadd a new sample on how to configure to Wish List as this is not critical enough to hold up the release. +Fixed in head +heres a patch +Committed to trunk and is nice to fix +Thanks Mark +that should really be a test fail +Can you also post your configure settings? In particular your install path but attach everything +Did a quick test to verify this and it appears that the attribute doesnt fix the bug finally +Note that the sslPort parameter was never part of a release so this patch does not break compatibility +Attached MS log and agent log during router restart and also iptables before and after router restart +fixed in trunk thanks for the patch +Thinking about it +I ran it with and it works fine +Maybe eventually we should change the memory tests to run twice +What do others think +I do feel strongly that Solr shouldnt force me to specify a Z when Solr doesnt really have any time zone support +Hence trying this from inside gradle tasks is no longer validPlease refer which addresses the investigation on having platform specific redis distributions. +This means that you will need to be online to use the sourcecheck maven profile +Given the bunch of other related changes I think I still have something thats close to committable but Ill need another day to get to this +Sounds good Ive attached the project I have that contains modified as well as the test case +Thanks +mvn grailswar failsI need to delete the jasper plugin from the plugins folder and then it works for next run or buildDeployment issueEven if I build a war after deleting the jasper plugin from plugin folder when the application is deployed it is throwing the exceptionI prefer this defect is kept open and you may please to Burt who can review for further actionTemp Work out Use the build config as suggested by you for local development for using and for deployment a comment the runtime jasper plugin in +Interesting way to fix a problem by removing the testSeriously though this is a bit of a problem for entities whose unqualified classnames are longer than characters +to commit patch looks awesomeThe change to looks leftoveraccidental +Currently there is single Upgrade object in the +That should I am not fully satisfied with the way the IDE does this at the moment +Was using JON binaries from World Lab +while investigating the test case with large numbers of messages I found an issue related to a mix of priority and non priority destinations +The proper fix might be to use a set instead of a list +If you use a simple editor like VI or Emacs there are many ways to screw it upThe line length limitation is only one thing btw there are lots of other rules like spaces in certain places are significant and empty lines having special meaning +This bug is still not fixed in +Downgraded my ant version to +Actually you still need SCALAHOME for the base Scala libraries +Thanks for the report +I request for this patch to be considered for a commit +Tested with and everything works fine I suspect the issue was with an earlier version of +Double is not a workaround for +Could you please attach complete page source or runnable war file +Ive solved my specific problem by changing it to in that particular I suggest leaving this bug open as the generic ability to define properties for the standard maven build would have lots of potential uses. +patch looks good +Kevin please point me to the docs where it is explained how to build the from SOA is a work of art you need to talk with Trevor Kirby about has already shown you a configuration running on the binary +I like that it pulls the out of the path and can take advantage of atomic row operations for modifications +What about char? Its just an unsigned shortIf we need we could add support for that in a separate issue +Yes I did I just did not commit because I havent had a chance to test it +You can easily test it +Still looking few comments +Do you expect to have a runtime performance penalty if it is left configured to use hadoop rpc +V +port this feature to +HiIve got a feeling that this might be useful for in generalFor example I currently need to know when Ive fully written a clients last response so that I can initiate connection closureFor me a very clean way would be to schedule a listener with aIll have a tinker and see if I can get something built in to +You might find it useful +This would appear to be an issue with the underlying integration in the server +If you need it you can use a decorator. +The two choices are now possible as new constructors have been added and the javadoc updated to explain this behaviorThanks for reporting the issue +ear if you would like to added it a +The test should pass both on Harmony and RI +This error also occurs when one wants to find references of a constructor method in which one has just swapped the two parametersFixed however in GRECLIPSE +That might be a good thing but Im not sure why its part of this patch +All the tests passed +set should be setThese three paragraphsThe names of the scripts that help with setting the CLASSPATH are set set and setYou can set the CLASSPATH environment variable in the operating system either temporarily permanently or at run time when you start your Java application and the JVM +Correct migration +Thanks Bryan the latest patch looks great! to commit +Is this a regression? Thanks Bryan for working on this issue +The attached patch improves the original one wrt improving partial matches and also adds a flag to disable partial matching for the. +Second why are you getting state if there is no state in application you do not implement state installation callbacks +patch to the doc trunk with the rrefexcept file +Thanks Freddy! Ill try it again when is out +Next time I catch it in our production setup Ill try to capture the logsAlso if has been resolved does that imply this is fixed elsewhere +Marking closed. +Another idea we could possibly do this using the decorator pattern +Could not reproduce this issue +As a workaround its possible to reindex the local repository from the Maven Indexes eclipse menu Window Show view Other Maven click on your local repository updateFred againIf this can help most of +Sorry I should have made myself to Spring hasnt solved the issue +thisObject is different from resolveStrategy and we should discuss it different from the others +IMHO simple dupe of +are there more info about this in +Could do a code drop on +FWIW I remember a similar problem with the Sun C compiler when I was working at IBM +In general it looks good +I do not see this code on trunk you are correct that the could be returned +Kinda trivial I know but the attached patch is all the change thats needed to resolve this +Ive revised this with a synchronized block for the case which should hopefully be a good tradeoff between efficient access and properly coordinated caching +Any chance you can try with the latest? A similar issue related to default namespace writing was fixed last weekAlso can you use wireshark or similar to get the raw bytes of the transfer off the wire and see if the duplicate is really not thereIm going to mark this resolved as Im possitive this was fixed a long time ago and a request for more information has gone unanwered for over a year. +I dont think that this would do any harm as the code stands but generally speaking it would be better to not register the callback until the instance is otherwise initialized to reduce the likelihood of future bugs +Here are the first three here are the last two newaffected email to Juergen re +Can we at least get some kind of reply from dev +It should apply cleanly +Could we talk about it on JIRA? I would definitely be interested in making this a top quality feature +The default px is very small +I just came from which addresses a subset of this issue +I tested on a cluster with seems to work +Andrea indeed the works correctly for. +Im testing another patch now +Good on you FlavioPlease create new issues for problems found reviewing this patch +I cant locate whereif this happens in so I didnt produce a patch +bq +Applied patch to resolve defect +Nate can you try applying this patch and see if it works for you? Let me know if you have trouble applying it +I think its not quite right to use blockReceived here since that puts it into the blockmap but not the targets array +Attached is the latest patch which i made sure works for both distributed mode and stand alone modeWas missing the in the conf for the client in my previous patch +Since the feature is optional it is not critical if it causes problems for some routersservers though of course it would be nice to fix that. +The news page was updated and published by Rick +I think there is a little more work to do as the validator I just tried still had some issues +This is probably a regression yet I think is not the most popular format out there +JXR does not +Closed issue. +Struts Tools TutorialChapter +At least once its all in one place its easy to make that one approach do X Y or Z as we like +and now closed for Rifat to attach his contrib statement +Alena agree with your comment should we create a tracking doc defect for for deprecating older API? Also making fix require in order to complete my second cycle testingchanging priority to critical +Please Verify +I can see arguing both sides of this logic +Fixed in +Among other things I want to run performance testsThere is part of a pass at a concurrent version of the current merge policy +integration started on r and r Ill let the issue open to verify Bootstrap doesnt add new possible regressions have been checked and fixed migration can be considered concluded +This makes it possible to flush the buffer to a characters call before calling endEntityIf an exception is thrown in this section this state never gets reset on a new parse +Thanks a lotBest regards by Richard. +Title should be Add comments relating to the use and purpose of and rationale for acctLock +I first thought it was a duplicate but remarked as related as i found out that this still occurs with the fix for the related issue +Updated Info Strangely this build problem only occurs for builds +However registerargcargv is set so this should not have an effect as stated in the settings descriptionI have asked our web hoster for the +Removed the bullet item in question from the topic +Are you sure you installed the right plugin? Statusbar plugin id is not cant reproduce this against masterTo those who are having the issues npm install the latest and try again +in branch enhancements +to RESOURCENOTFOUND in pushed as commit ababadafaacade +Attaching reviewing the links to licenses and notices at Apache I have created a new file that should both go in the java directoryThe file has the Apache license but also has additional licenses in itand it is specified which portions of the project that those licenses apply toThis information was gathered from the various old README and NOTICE filesThe new NOTICE file is a collection of NOTICES of copyrightsThe new and files are a collection of information across the following files which should now be deletedTHESE FILES SHOULD BE DELETED +Ok +Did you check the minimum ANT version needed for this? If the current minimum does not fit we shoudl update the build docs +The question would be what is our goals for the types of archetypes to provide? +I am having the same issue where the user goes out to page the result set changes to empty then user then selects page +I use git command +Can you see if the issue goes away with this +Thank you Martin. +Perhaps it could be finished up without too much effort. +If we understand Service Breach Time concept right it should be be a Service Breach Time when its waiting on the support Tech +Done +Changed params in invokeAll and invokeAny methods of +In fact this was what I proposed initially too and testedBut I was worried that extra checking slows down other calls also the trunk code creates new Map too when deletes is not passed +attached is a patch this adds a dependency task to the task this way any ant targets that need svn instead will fail with An svn checkout is required for this target +I should have mentioned there is a SQL script to create a sample database in the project as well +I think along with removing this well have to make some changes to our usage of zip It seems that Cocoons version of this has made minor syntax changes to it +Please try our final release of which has some fixes for Packaging and functionality and hopefully will resolve your issue +oops wrong issue +Also I have tried to change clientAuth to false instead of want in the catalina https connector config and now Safari will connect with no problemsSo why Joe dont we close this issue or put it as a minor bug not to a major bugThanks this is not a bug of but probably if clientAuth is set to want the browser checks that the site into the certificate is the same of the running siteIn any case you can set clientAuth to false and this issue will be not more present into SafariThanks Marco +Makes sense to me +Commenting thereAlso in the patch we can simply pass a null for the instead of forcibly using +One for classes one for strings and another for all other object literals +Koen +One relevant argument I have heard from Sanjay is that if you are writing a PB client in C for then basically you shouldnt be relying on the Javas semantics of exceptions +Will let you know by tomorrow +I just committed this +Committed to trunk. +I would prefer to work with an existing set of unit tests if possible +Perhaps create an accounting admin section to account for this superuserugh this ultra wide comment box is bothersome +Otherwise no problem +Removing State from the queried attributes in the plugin. +I agree +This has caused intermittent failure of on our hudson instance +The URL is not set right +Can bulk load data to user table +All seven test suites passed on + +Master files were updated by Satheesh. +No good deed goes unpunished +I cant remember exactly what it was +Can you please let me know where i can post my questions? as I have not seen any messages on one of the mailing lists I wonder how to help you? Heres a suggestion +Yes very confusing will be updating to resolve soon +This patch implements the following +When you come around to doing this please write a first to reproduce this +This needs to be optimized for developertweaker comfort. +the bug description is confusing to me +Assigning back to please test this +Youre right I didnt understand that +EDIT I have deleted the content of my comment as it is in fact crap +Isnt this already implemented in +Hi JoshuaThanks for help but you might as well revert that change + +updating patch mainly for backup doc in parent less empty files based on doc comment from parent jira +So do I kick the plugin out of the build? Kick it out of the build for win? Turn off tests for the plugin +the other approach is to grant those permissions to a role that the guest user has assigned then it is available to anyone through the authz system +Assign to Alexey +This affects pipelining while replicating data and not when client writes the data first time +When your code base grows over k LOC then compile time interface checking is a nice option +These are best figured out by observation analysis and human intelligence and code review +Ill do this for the next iterationPlease let me know if there are any problems with the zip file Im using instead of winzip now +The implementation mentioned above would allow all three of these control options +As it stands however this change seems like an improvement to the existing logic +LGPL isnt really an issue for us because we wont redistribute it i remember correctly Archiva dev had to yank out jasperreport a LGPL type license out of archiva is the different if we use lidquibase for the automatic db upgrade +i think we can do without a test on this one + +So can I simply switch this over to a under LangIf I get a green light Ill get on it first thing tomorrow since this issue will no longer be blocked +I dont know Oracle all that well but I should note this is Oracle Locator on g not a Oracle Spatial database +Sort of +These things happen best thing to do is to email builds please for stuck builds etc no need for jira issues. +Are you going to address this in +I am not a fan of having that hierarchy being mixed use wrt Serialization schemeIn the meantime adding a protected ctor does in fact work around the issue so lets use that for now +Closing the bug. +I chose to put the in a span element so that when you mouse over it it shows what that ge is based on +I am attaching a file which provides osgi integration tests forThe tests cover producer consumer idempotent repository using both spring blueprint +Can you please push out a short email on the google group so that Filip can pick this up easily +Attaching a patch vb incorporating comments from Narayanan and ysteinNarayanan You are correct before the stop method is ever called start has been called +THIS PATCH IS NOT FOR COMMIT +Is this something you want to do? Assign back if not +Aaron thanks for the suggestionsJing thanks for reviewing thisI have committed this. +My application hangsblocks forever as I catch all the Hector exceptions and retry when there was an errorAbove log file messages will repeat itself again and againThere are also no error messages in the cassandra log fileAlso Could not fullfill request on this host is an error message I have never seen before +Comment Please make some of the final members static not addressed +address Santhoshs review is for branch +Please review +So I think there is a need for a of location before setting it +Its there when attaching the JAR but not there when attaching the test JAR +Fixed in attach a test caseMarking this as Wont Fix +Rechecked it in the latest WTP +run DEBUG c allThat didnt seem to help +Integrated to TRUNKThanks for the patch Mubarak +Hi AllThis is important issue related to order entered promotion so please provide eye on this issueGreat help for us +The patch is available looking forward to reviews and comments. +deletes at the level of the however +Comitted in r and backported to the for a while +I am going to based on emailYes it will be both user explaination and tech infoBoth will be individual books and included in the master list for Users an TechI may find some time to add some info in both +Im guessing its a of upgrading to Ivy as mentioned in release notes +svn revision +I would have kept the API clean and made Velocity owners complain to the writers of that tool to fix it since that is a bug in the tool +LGTM +seems to be an embedded tests a and tests which is not relevant here +Removing internalwhitespace or removing arbitrary characters sounds less temptingNote that this change would touch a fair number of Converter classes Short Integer Long Double Float If this were to be implemented I would recommend that a setTolerant methodbe defined on each converter class and that the data cleaning behaviour onlyoccur if this flag were set +Nice detective work Van. +Example of how the suggested fix looks for the sample project. +Attached patch adds a separate class as you just committed this +Pull request merged. +The priority list throw a null pointer exception if you remove an element that isnt present +You will notice that was explicitly marked as an incompatible change +In other words a String is not considered a component type! think this level or purity on configuration versus component is at the expense of some headaches +Patch applied in trunk r +The problem has already been fixed in and +Switch it to false and outline works fine + might be due to the use of the ubiquitous queue queue from JNDI +Reopening to change fix version from to +The missing operators were implemented in r +Otherwise if the general consensus decides that this is an appropriate fix it will have to go into trunk +Changing affected version to and fixed version to the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +The call site becomes invalid if the receiver meta class is not what we expect +Here is the patch +I think the code is correct but I wonder if the ejbql test should be modified so that a CDCFoundation specific master file is not requiredWould the purpose of the test be changed by adding a cast to the expression that causes the failure in JME +Hmm this a base encoded combination of user and password which is far less than a char password quite a blanket statement to just say its secure enoughi would suggest that the workaround is to use httpclient instead of but I assume this would have been done if it was something straightforward to do +All suggestions are welcome +In more complex where multiple sessions are used and sessions could be disposed in between a centralised session could be used to manage the timers when the other sessions are disposed +The storage is and can be found in +Removing File Channel dependency +I dont believe maven works with for password management. +I havent changed that example in quite a while so Im not sure why the GET wouldnt be working +We are considering moving to the latest build but prefer to wait until is released so we can stay on a more stable versionBin has been tagged in the CVS two days ago is officially out +This is a simple patch to optimize the case when a server that would restore the quorum is coming upInstead of leader giving up and leader election happening this change will make the new server join the existing ensemble +Still missing are Solaris HP and I am not sure on Windows if Windows etc have different variations but this is a start +I started it so long ago that the merge was brutal for me +Would you please help to try it +Same patch with timeout increased +I will remove and provide in the fix in the next release +I think this issue can be closed now because the use case is supported in another way now +Hey ErikIs there anything to do here? This patch depends on which isnt targeted for branch and since the projects are split on trunk this patch doesnt make sense hereAre you thinking we need to put and this patch in? If not please close this out as wont fixThanksEli +I just added an external try catch with a generic Error while parsing the affixdictionary file in my opinion better than just eventually throwing some unchecked exception +Hello HarryI think you have done a very good job synthesizing various logging paths in and in Java more generally +Youve also patched to sort directories +Nice fix +Not sure though if it is possible to set up the response code earlier on the WS out path perhaps a simple fix is to register after WS out interceptor sets the response code if such WS code existsCheers Sergeyif we want to get some status for that is analogue to the http status code we could set it in the s serialization step and that can be pulled when the gets executedthats how the sttus code inclusion for HTTP was added to theBut is there some meaningful status code for? assuming for the error case someone is already setting the appropriate code we could add the same logic that is in the httpdestination likeeverything ok and has body ok but empty regards AkiI guess so if no is set then default to something like that +If this env is not used outside of binhadoop unset it +Done +So maybe we should just go for it since it passes all of the unit tests wo any changes to them +Once it reconnects if more than the session timeout has passed you will get the +Previously commented on by thirumg but nothing has happened since +Visual Studio complains twicewarning C unary minus operator applied to unsigned type result still unsignedDsvnPhaseqpidcppsrcqpidtypesBuild still completes ok. +Thanks +RobbieCould you please review added test +Im able to cancel long running downloads at once +incorporated conflicts +Can you consider an alternate wording +It could be caused by other test cases having failed before without proper cleanup of resources which would leave dangling resources that the next test case will be confronted with +Sorry for the delay seems the eclipse I had was corrupted and were showing errors that shouldnt have been displayed +Thanks Ahmed +I tested a new variation of the patch that should have better performance +In this file I added two test cases that use allocationSize and allocationSize to demonstrate the problemThanks. +So the failure still wont be immediate as intuitively expected +Last patch was the wrong version correct one added now +So once its out its in maven. +This is out of the scope of the algorithm so the algorithm has to be adapted to consider these methods too +I was wondering if you had a smartphone that was being helpful or if you wanted to talk on the phone +I have verified the fix on my end as well + +Adding fix and additional test coverage +Thanks Andrew +We can discuss thisHow long have you been seeing this problem for? It seems to be working for me know +is this a permanent change or something that is done manually for every build +ElenaMostlikely this is a test problem +on patch v +At the time of the next CP a new review will be done and any new items will be logged in a different JIRA + is mandatory for color images but optional for black and white grayscale and palleted +It there anything I can do to help expedite this ticket? Ive provided a patch already but please let me know if there is something else I can do that will get this patch applied +This is subset of functionality provided by Enable L L services in shared network feature +tests passed +oops downloaded v by +Attaching patch addressing Rahuls comment +I presume this code will change anyway once you have the new rehash events for NBST? Could you please attach related to this +Did you resolve this issue by mistakeAlso do you have a unit test for your fix? I just played around with gcc a bit the gcc preprocessor correctly transforms this snip define metoo i want that too ifndef defined a askdaddefine ask mom or should it be mumelse yeah we know who is more likely to say yesdefine ask dad dear dad would yield even better results +Pinging again +Please add any additional comments to a new JIRA issue with a fixin for a future release +periodically add a sync record which includes thsounds like a good idea especially helps while debugging a corrupt edits log +Thank you +Ive removed the offending file +bulk close of resolved issues. +nothing to do with eclipse helios so moving to currently it look like please set a specific Fix Versions +btw +Or the other way around +New results will now be correctly marked as failures +Attached test case and fix +Tomaz can you please test this +So I updated the logic and the javadoc +Looks OK to me +It can probably be incorporated into the wms serving rasters +Contains unit test and changes to interface +This issue should be addressed in an existing pull request for. +For now I consider the classes in internal so a Java signature change is allowed +I did not understand that all the agent threads are in the same process as the jvm +please remove it before committing the patch +The expected result is an error +whew I thought I was going madOne thing Im wondering is if anything else is being pushed onto the stack within your sin tag that might mess things upall the end of the sin tag does is blind pop if something else was on the stack top I could see a potential problemThanks for looking into this +Patch that addresses the issues described previously +Hi Mirek I didnt find the said in any of the log files +Never mind surely its me +RSB will indicate ALREADYOPENED +This isnt true for Math +Are all included in thisI need to wipe all content in www on the webservers to activate thisYes the subsites are included +Assign to agree with Stefano anyway I think adding the synchronized make sense +Theoretically there is a way to resolve the issue to your favor +Verified in pager and is now working +NET +The bug that prompted this JIRA was a bug +pach created by diff urPp tar has to be extracted in stanbol root and it puts things to the right folders +Personally I think this patch is ready to be committed +ideally we would expect users to install thrift if they want to compile stuff +Martin if you know enough to do this please go ahead +The documentation is definitely wrong as we probably forgot to make the changes to every affected data type when Integer was chnaged from to a few years ago +Hi BJ thanks very much for these patches +Thanks Uwe ill do that +Can you rebase to trunk please +This patch makes all implementations return Path objects in +Would it be possible to have a look at the WSDL +This will allow you to have more control over tickets. +Votes from me and my team +I think this would need a code change but I think theres no need to put it to M +This issue does not appear when setting READCOMMITED locking +G is using the released version at the moment Im okay with upgrading to the +Best regards GeorgeThe fix looks fine many thanksBest regards by Andrew. +Overview figure +Any stacktrace +Drag and drop in RSE loads a transfer dialog but I havent really been able to track down exactly where its coming from yet nor how it works nor whether its feasible for me to make use of it or not +This is a patch that makes possible to use the Whitespace tokenizer whatever the way sofas are set through collection readers +Arun could you update the patch to match the latest trunk thanks +The version of XML Schema supported will be clearly marked in the documentation +For human verification that the error messages make senselook good +This issue is resolved with the latest changes in since its not possible to call an admin page without explicitely specifying a core anymore. +looks good +Further to Christians comment you can lose all flashrequest stored state by the user simply issuing a refresh via their browser +Usually in that case we show as much as we can but for safety make it until xml problems are resolved in Source tab +Whats wrong with an Oozie Java task +We dont need to do this as will remove these flags of +Note that a lot of this was done under +fixes it +I believe that fixes the problem. +This happened in the beginning of some local short test during the startup of nodes +To reproduce +Rob please review my patch for good for now but of course look into customizing +Probably something for the Spring documentation +it is not only not overflowing anymore the possible solutions in the error message are now correct as well +on +Sounds like the editor should at least warn and possibly not allow any save to +if we go a reasonable amount of time without seeing it in nightlys or reported by anyoneDidnt you say you still saw the problem in one out of five runs with the fix +pre I thank maybe before that its been a whileI thank the problem I am seeing is coming from the if it is set to is that counted as request or as when we are scanning +we just havent added it thanks for the suggestion +The theme is closed +We have the same problem when there are multiple triggers at the same time for example a dependency and a SCM change it is skipped due to snapshot dependency build trigger even though we would have wanted the analysis because of the SCM change +added ignoreErrors parameters name is fail. +OK +Extract from log file lines covering the login attempt +I will resolve then + +What are the plans for the second part? I expected that the SolrTest changes proceed now so we have enough time check the new API in general use and fix issues +The connection has been terminatedinstead ofEXPECTED EXCEPTION DERBY SQL error SQLCODE SQLSTATE XJ SQLERRMC Derby system shutdownI am still looking at the changes but wanted to check if you get the same results with the patch +this can happen if the group parsed from the node is null meaning it was started outside of whirr or otherwise doesnt match +Works for me +Closing. +If you think its worthwhile please open an enhancement ticket +Ive added a missing summary line to the issue +Would anyone want to try this +Wouldnt it be a better idea for people to simply get a SNAPSHOT build of as and run the installer over that +Passed Failed +Thanks to Jaimin Jetly. +Any progress reviewing the patch + +The result shows that this bug has been fixed +Done Jopr r +The jobtracker would immediately move it so that even the submitter could not alter it subsequently +fixed +Whereas the new renewal assumes hdfs and as such will force a hftp token kind to be hdfs +This patch locks the version of the used +Now that we are the domain this is set up atdtd where there is a basic +We also need synchronized on subscription message queue which is similar to what we did in java client +Tatyana would you please to verify? + changes look good +FixedDonwload the latest nightly build and give it a try +exec and java mojo use a different will need to find a way to pass multiple arguments from the command line +this example seems to work for me in +Closing bug. +I am not sure why but I was not able to assign this task to myself +Committed to both trunk and +Overloading is not allowed in remote interfaces per SCA Specification +Looks good +The lookupLink test is failing due to problem identified on tests are failing due to the not gettingtrimmed in the xml parsing see is fixed now +Looks like thats the case +OK let us do it in patches but do them quickly +We must also verify what is the behaviour if we comment out the default constructors present in the classes the error described took me to another problem javassist generates two extra constructors +So in fact you can do all the spring configurationMaybe this issue is less important now +Thanks Jason +Ok I added some additional comments in the Writer as Namit asked and all of the unit tests cases pass +Maybe you can submit a patch? +It actually would be nice to merge with the mirror functionality +Can you come up with a patch for this +Removing produces the following error Exception A Map must declare a elementPlease check for further details +Hi wendyThanks for the review +I have withdrawn reference in get since I do not need it in my project +This patch adds an upgrade test and a permissions test for the new system functionThanks for that feedback Dag and Knut +will do that and attach the results here +Maybe there is time to get it into +The breakage is under both Windows and Linux +I understand now +Id be happy to discuss this further. +Can you specify for which modules this is needed and where to add that it only needed for modcluster or for overall httpd and all dependant libraries and on which versions is this needed Solaris Sparc only or all Sparc versions +Ning is going to rewrite the test in a separate patch +EricThank you for taking the time to submit this patch +Is this still an issue for you +push out to +thanks +When the executor destructor runs it will wait but gives up after awhile so the objects resources are gone and that thread finally ends for whatever reason only to find that the parent is gone so the return of the Node walks into memory thats not valid any longer +Mass transition all resolved issue that did not see any further comment in the last month to closed status +setting fix versions correctly herehappy to backport this stuff to if desired +The original sources for these tests create nodes from a custom type whose definition has OPVVERSION +name of the attached file is t reproduce this behaviour +Possible way to get around the retarded classloading problems from jstljdbc +Implement a deterministic memory manager OK these two together make sense as I was seeing the failures under maven with the exact same MOPTS +Just for the record our namenode servers with release got into this situation twice once with a cluster once with a cluster +Tested and now it changes correctly the memory size +We can keep posting revisions of the document to this jira as things changes significantly +Very early in system initialization in the layer if an Upgrade has been requested the entire data store will be scanned and replicated via hardlinks +So the form tag is not completly conform to the spec +bq +Sorry if this spams things however its unlikely that Ill work on these. +switch to modello in trunk rev issue open until configuring xdocxsd generation is done +I will dig the code for it +Remove empty hostname property for HICC jetty server to bind to all IP of the running host +Ive just sent another with the software grant signed to secretary and commited the file to the incubator repository just now +Patch to implement hilight with the match in the middle +I was hoping youd get rid of the huh hah and just start a new minicluster for each test +Sorry I was looking for for webdavlib not May take up to hours for the fix to reach central repo. +However the latest information from Marek Schmidt states that the workaround is to +Fails with SQL Server but passes with a nice day +Tx +As we agreed with Nathan the patch should not be applied to ARCHIVE + +in release +unrelated test new tests added since the change is very simpleI have committed this +The root problem is that with automatic phrase query generation turned off by default and for the textgeneral field in particular the core Lucene query parser is generating a query for the tuple of using the default query operator which is OR by default +at INFO level no? Most users probably dont specify them +committed into revision +If you ask me I would like to be able to setup the size of the buffer +for the patch +Some information is available in the wsrp file on what jars the test case expects. +because of the timestamps it shows as updated every time you get it even if it hasnt changed +Lahiru did you see this error in the recent RCs? If its fixed can we close this issue +The fix there seems partial +Not only is the contract a little strange but it also makes require a job and possibly a jobName for those methods to succeed +Triaged July assigned normal urgency +git is not getting synced for some reason +Entity security doesnt provide equivalent flexibility its all or nothing +Patch to the specs module +Marshall says that he has a setData issue still that is giving a bit of pauseThis is still a very good reference for review +For example check the source code to be inspired how that lazy session option works +This is still happening on coreIf I change the mapping to it works +putting this to so I remember to look again +Ah right I see +and it is not correct in IE +It also applies to? Ie if you try to add a index in a world you should getYoure right +Not much I can do there +Many thanks +Applied +could not reproduce with +page +Sorry about that Jeremy + now logs missing elementary without a stacktrace at WARN levelIt still logs the stack trace at DEBUG level +Do you agree +Latest Kerberos KDC LDAP schema +binAll of the extra ones are for testing +SavasHere is a thing for you new test case as you are using which has no annotation will use the simple front end instead of the JAXWS front end +Do you really require access to both the String and int versions of the methods +is a feature release so I feel such a change is not out of scope +issue should be in AS vote to remove extra UI that is redundant +tests removed in master but still available from other branches. +Current report was generated on +Not a Spring Data Defect! +This can be used for big inputs and is offered for pack as well +Thanks for the patch JD +Resolving +Resurrecting this jira +Some comments would be helpful to indicate where multiple format strings are used as a group eg the used for connection message formattingIn the interests of consistency either the existing statics in the classes should be reused instead of this change or thost classes probably need to be updated to statically import these will help o ensure the code remains properly maintainable +Ignore this comment only testing changed JIRA notification + to trunk. +for +Ive also moved this to a later release yea I know no one is happy with that +I changed to assert only when directory exists +Interesting +I agree that the solution definitely involves making sure SOME kind of error happens on the client +This is done +I have added rounding mode and scale fields in in subversion tree as of rCould you please check if this solves the issue for you +Committed. +GoranCan you attach a patch with the new plugindirk + tested! +Seconds instead of milliseconds sounds fine +eek X and Y intercept a class annotated with A but not a subclass of A +But i didnt know the reason behind addition of check +I will check what combinations of components are already tested in would so clone pre existing validator tests for select value change listener +As I said before workload should be reduced from to if we choose way +There is some fundamental that Im missing here so can anyone give me a clue where I should start +And please add a entry for trunk +Hey MarkThe interface in the zip looks interesting I am working on putting it into +Could you please let me know which class you changed to map reference to Object + While researching standard Eclipse tool bars icons Ive found that most of them are made through with actions +screen pic of offending by removing message label in and setting minimum sizes in sub panels for Driver and Sources wizard pages. +Please provide more information in that case +bq +Would it be enough to describe the format in the Java files that parse these resources +For the time being you can unregister that interceptor or look to debug it and patch it for your needs +Will commit if there are no objections. +The above fix does not work in all portlet containers +Indeed I thought that without a specified length the column would be created with the default +Hi AndrewI think I fixed the bug +UI example in Fedora to write release note issue again now I have modified the release notes as I got two mixed up and need to edit the release notes did you forget to close this +Keep insisting in this issue is just continue with a discussion that was over long time ago +Please let me know if you need any changes +settings as a temporary output directory +Hence the need to assume that most developers and this is what RHDS is targeted at will have more than one JVM and the default may not play nice with +sorry about thatI works fine with the fix you did thank you +The issue here is that prepare is supposed to return the amount of tests that the test type thinks will be run +Marking this issue as resolved + +Also remove the stacktrace from the get of a decompressor +Will check the findbugs warnings +it should not include the wc DOM classes +reducing timeout to K +Also log when moving to the expiry queue +SCR might be available only if at least one component is installedA similar approach is used by Plugin +I argue there are more misdetected russiandocuments where thats the true meaninginstead what matters is clear behavior. +this may influence the build ofcustom Cocoon projectsIf somebody is interested in the current docs please contact my directly as Idont want to publish them ATM. +Id like the users of our build system to be able to simply specify which artifacts are published into the shiphome just be publishing them to a specific configuration in the ivy files for their modules +I see Replicator as a service that besides providing the replication framework also comes for replicating Lucene stuff +Closed by mistake +EnisThe great grandfather should have ideally been compacted right so that it does not have reference files because he was the one who got splitted first +I will updated the version once the scalatest released +Could you try trunk version and confirm that the problem has been solvedthanks guysany idea on any workaround that doesnt explicitly delete ivy cacheCheersIs anyone working on this problem? Its causing problems for us tooCheers Steve +With Xalan and newer it worksWould it be an option to add a dependency to Xalan to UIMA +Please dont set the fix needs fixing for +I dont understand why you think this is associated in any way with +Currently the values are taken from Wikipedia and other sites yet differ from some of the distances in the codedAlso the patch doesnt seem to remove a couple of empty packages + youre backHow very cool cos man Matt we have so missed your eagle eyes but at the same time how very because it means that youve spotted yet more mistakes fumblings and clangers on our part and actually you only got one of the two I reported just then +closed as fixed. +For excel theres also a few bits stored in the main spreadsheet sheet stream tooNot sure if it would break things if we did most of the metadata fetching first +Please if can describe a use case were changing the default will improve the performance or scaling of the broker +Sample test can this be closed now +Over six runs of an explain query with joins seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds +Okay I misread Avro specification validation should occurs at runtimeFollowing this principle shouldnt int be accepted as a reader schema for a long? This now seems inconsistent to meSecondly is there any facility to test whether a reader schema will always succeed in decoding any instance of a particular writer schema ie +reviewed the patch and made the required changes after discussion with Ashutosh +I just fixed a bad rubyspec +The calls are always successful +Would you please help to verify its fixed as you expected +I have found that I can simulate a perpetual check by setting the maximuRetires to and establish to an extremely large value +changed to use uselocalhost +What are you trying to accomplish exactly +This isue related to +The subtraction and addition functions require a space between the operatorand values thus invalid mm mm transition to closed remaining resolved bugs +I was able to reproduce the issue and also created a test case based on the sample provided in the forum ll provide a fix and post the link to the pull request here +I dont think the state transfer manager should throw an exception if the region is empty + +But each structure is locked individually this should avoid causing any exceptions +I have reverted the extraneous change moved the tag and redeployed the artifacts +Really we are on the middle of solve so this one is part of that issue +Im attaching a screenshot that shows the modules directory in AS +Do we want to enhance the tag to support numbers or just leave thesupport at doubleIf someone wanted a implementation they could write one +Can we mark it as resolved and add the release version? +The dropdown is populated from the list of subdirs under templatesdefaultskinsCould it be you have some hidden file or dirs which chokes the codedirkTry putting this line in properties Normally it should fall back to a default value which apparently doesnt out what was going on the fact that x has a string in it was killing my script +With version running with the JVM I encountered issues related to MVEL which went away by my updating to the release of MVEL +looks goodThanks myrna for taking care of this +specifying version seems to work is not the same problem as +pending Jenkins for the review Todd +It was a bug +Refreshed patch against latest trunk +Custom loader function dataset +Ive set the Documentation component because I think this could be avoided if the methods mentionned above were better documented +Hi Eranda I have attached a revised copy ofIt contains a workaround for as well as several otheradjustments and tweaks to the tests +Unrelated +FYI mvn dependencytree would be the same either way +Because this bug is a JVM bug complains to your vendor or use another one which will work just fine +If we cant fix this bug easily then perhaps the best short term fix is to disable inside of +Eclipse patch for changes built on top of snapshot process Lori Thanks for all the effort you put into this +See. +Want me to mark this ticket as resolved or is that something you guys do +As we now have a context object instead of a list of parameters for each interceptor methods we can start shave the useless normalizations +Pushed. +Heres the patch for loadsaveadded class which is mager of matrices +With set to the proper LOG info is seen on initialization with event queue listener and number of worker threads +Fixed this should be fixed in snapshot and beta +Make the member variables final Is default constructor required? listSnapshots may be a better method name for methods listSnap and listSnapshot Some empty line changes in can be removedBTW this patch does not compile +And the problem doesnt occur +This is happening in. +trunkCommitted revision +Thanks Nick +My bad I was applying this out of order +No I tested it with the default KarafIt works with theI will close this issue but we should document this for our users in the release notes +Can we mark this as a as we have a workaround +As a consequence any async delivery of the message will cause the delivery of the response to fail whether or not this is done by a different service +The patch seems fine to me +Otherwise the record could be skipped if the object comes before theplease we clarify exactly what we mean by? If currently we simpky remove some fields then index the rest then this is not good practice +To be verified +See the reference forum thread re forge is still generated broken getset method names when using its entity command line +Attachment has been added with description Refactored afp barcode has been refactored to remove the dependency on barcodej it requires a object to be created and therefore in order for barcodej togenerate native BCOCA it will need to map to the appropriate object and set theattributes accordingly +The import problem was due to class refactoring +patch looks good +Patch attached to move the Multimap computation into only the block where it is used and not each gossip update +Any news on this? is it a big deal to implement it? To work around this I would have to created separate classes for each enumeration I use +Hi Michael Thanks for looking at that logI ran on the corrupt index and have attached the output here +See the following classes to help standardize Command messages in extends the above specialized for jBPM ESB action class that invokes jBPM methods when a Command message is Base jBPM that eases the task of sending messages to ESB services using the mentioned above +So I put the duplicate skipping logic at the place where it was affecting the least just the synthetic MOP methods that were getting duplicates in is certainly better if it is safe +As a side note I apologize for not trying the latest assembly plugin before filing this bug + +Indeed it will not be fixed in version +Good point +Are you intending this issue to cover both renaming a table and renaming column familiesYes but i have no plan yet +This is a trivial fix that can go to as well +batch transition resolvedfixed bugs to closedfixed +Dear Vamshi could you attach your test project +Hi Guys something that could work on H is to use the MVCCtrue in the URLjdbchmemMVCCTRUEDBCLOSEDELAY +Other than that looks good +I was using File Import Maven Projects +I remember this heapsize failing on apache mc +Currently Im stumped as to a scenario that would account for it Ill need to do a run on a smaller set and attempt some forensics seems to me +Thanks Corinne! +As for s immutability I didnt see any final members so setting up the state in the constructor and not having setters just looked more like it wanted to encapsulate logic rather than immutabilityI dont know if it was initially designed to be immutable its just the fact that text was computed in the constructor and that this class had no method that mutates data that made me think so +is a subclass and not a wrapper and does not exist Matrix has bindings directlyYes label binding transmission is +Does this make sense? Is it helpful +Still having some issues in trunk need to in to trunk awaiting documentation sample cfgs +This issue is fixed for the next release Marking this JIRA as resolved +Any progress by any chanceWant to cut soonishWe did lock before but on a different lock +Are you sure about your version information +bq +Sorry I get exactly the same error with instead +If you think the stacktraces can be cleaned up Id be happy to accept a pull request with your changes as long as it doesnt affect the intended think you are left with workarounds +Does it start multiple databases? Does it start other processes that then start Derby? Maybe there is some missing shutdown or there needs to be some extra wait for something to finishAnother thing that sometimes happens is that there is some either intended or unintended dependency between the tests fixtures and the order they are run +The has changed with and later +But now I am unclear as to the criteria we use to decide whether a policy should be preregistered or not +As long as HDFS has marked nodes stale before starts recovery change from is sufficient +If you finish it before we release beta please correct the version +One possible solution would be to not use the any more just for this single bundle +Patch file containing modifications to modulessrcmainjavaorgapachetuscanyscabindingwsaxis the patch slightly to account for multiple child elements in the SOAP header as per the to at revision +Refactored the test runner a bit to share common code when removing annotated testsStill testing whether this works but thought I would put it up for feedback from those who have big machines +Everything worksI made some comment to the RB +Passes all tests locally +what good can come from allowing someone to force lucene to run on? This makes no sense to meThis is indicative of the larger problem in Lucene development recently the tendency to prohibit something just because you cant think of the off the top of your headSo heres your one can normally get most versions of most major to correctly run applications its often just a matter of disabling certain optimisations for certain methods or in extreme cases prohibiting JIT of the method altogetherHence at a minimum there should be a way for users to easily override this behavior wo recompiling applications +I will post a question to the datatables forum and see if that turns up any results +Closing all resolved issues for. +This would fixIf applied that patch you could not solve the worst scenario others may commit on trunk so the tag created could not really be the one that you checked out and ran testspackaged locally +Setting Fix Version to future for issues without a set fix version +namedistinct would be a symbol like enum +This patch fixes both issues +msec transferTo +revision in CP reopening to update release note info. +Thanks for your commentsIt is extremely difficult for me to add Primary Keys to the table because the tables are created by rd party vendor and I do not have any access to doing modificationsI highly recommend to add another fallback if there is a unique index declaredWho is the developer who works on this module +It is not for commit + fixed a bug +For example I tried KEYNAMEhbase and I couldnt login to the ZK server by ssh i +Im concerned about the verbosityAgree. +Rebased +v is a retry w the javadoc added +Alexey please verify +how long should this take +JS is done +The same problem is still present only in this case it does not appear that even an XML file with the error is written +Please rename +Benjamin there are lots of fundamental changes here coming +Lincoln Baxter voted and though he said that is s tough is that its clear we have two really popular logos and +Should this be closed +change reporter as requested by christoph reck +Can you replicate it elsewhere with petclinic or vote +I noticed this last time I was touching the file and I think they should be correct in CVS +This step makes sense in the community but does it make sense in the JBDS version +In this case the link contains only the page id and there is no information which page class to use to create a new instanceBut I think it was the same in with the hybrid url coding strategies +Looking forward to use this in version +Thank you for pointing this out +I just downloaded latest another thing I did not like forcing user to enter the source and view model names why cant these be auto generated? I believe we are taking this route in other importers +Another jira is fine +but for me after changing it to attribute also its still the same +Changing data set thats contained in model between nd th phases can cause data loss so it should be considered a bad see sources for more info +test debug change onlycommitted to was committed to trunk on in May +And I can live with the workaround +Im testing a second cluster to ensure the results are validI also have measured that the NEGOTIATE header is bytes length of hostname +MaxCould you please decide where to place the M version of the discovery xml file +Lets see if we cant at least get truncate test disabled and committed to see if that will fix the build +So disallowing them in the spec is probably best and it would be best if implementations checked this +I will fix it +Thanks for patch v + that demonstrates bug +I thought you sent me some files but I cant find them +tested the with the Observable Maps example in the documentation page as well +I will give it a try tomorrow +As in Hftp requests will first go to NN which will redirect it to a DN +Nicholas since all sub tasks are resolved can we close this? Did also fix +passing JSF info context from web UI layer to Seams business logicOr are you saying you exclude JSF from the +i added the warning on purpose to notify that we are trying to do a best effort stuff by adding the anytype and that users would need to register typemappings for correct serializationdeserialization +For this we should add the logo to git repo +This has been a major problem for a while nowand I think we finally have a good solutionOleg I am also quite ready for a +Logged In YES useridI was unable to attach the +Submitting patch for trunk +Fixed it by using the same approach as in and having a separate parameter name +I ran tests again with this patch applied and all tests passed +Since almost every dependency at ibiblio is going to have transitive dependencies with versions in them not fixing this will effectively make transitive dependencies useless as most projects will end up with multiple versions of the same jars in them +Thank you for pointing it out +Modified the matcher and added a testThings look fine to me with svn revision +Summary of changes exception handler signature modified to throws Throwable so that it can rethrow the argument Repeat template as and unwraps before passing them to exception closed as resolved and released +I just committed this +This is for the Accounting Payments screens +the bundle needs to be installed +Initial patch contains fixes for Jobhistory server and for working in simple authorization mode +Stephen Maybe what youre looking for is a Set +I think that perhaps the old behavior should be removed altogether it was based on an assumption that users dont want by default to recrawl the same pages +Nit please put DeprecatedMore than a Nit +Verified fixed in MVersion Build id Build date to attach screen shot. +should be compatible with S and will be the base forhth dont see any confirmation that was a successful backport +Is fileid in the foreseeable future +If we get this API right it would in what ever web based applications we develop +Great to hear and see how youve adapted and added to the code to suit your requirements +So it should be fine +I have read that the aether thats used in maven is +Passes all local tests +When a CF or KS is dropped the data is deleted +The logic is like this The Normal logic is used to find the first split to be assigned to a mapper +BillQuick browse did not show me where +true is passed to the JVMOh ok property then +Ill work out the patch and unit test later today. +Needed by rejected or a dupe +The attached patch contains modifications to support outer skewed join +i am so sorry i dont get the attachment why? please help me time about months ago we lost all the attachments on this jira system +During a failover without a STONITH capabilityWithout some sort of fencing youre going to have to run agreement on every update +I want the main content available to crawlers via MVC while using GWT for editing data +If I have more time in the next few days I can try to build a more complete sample program and post it +bulk close of all resolved issues. +I just have to make sure all the unit tests still pass and that it can successfully be built +Lets see if the changes introduced with solve the problem +so wait and see +You got a point there +I am reluctant to add this to the code base without abstracting it more becauseit is Oracle specific +would be great to have it in +Updated fix +More details to come stay tuned Sonar Views is not open source I think you are violating this JIRA license +Why do we need configure system property for tests? Eclipse has by default configured jvm +Attaching DDL to work with Derby +Ill give it a try and let you know +Failures are unrelated +I am interested in getting this into for some Salesforce usecases +Changes the signature of the methodM javaengineorgapachederbyimplsqlexecuteCalls that logic for DROP TABLEM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlexecuteAccounts for the new signature of the methodM javatestingorgapachederbyTestingfunctionTeststestslangSome regression tests to verify that the bug is fixed +Providing a price field with the associated products list could allow for such functionalityTo make things more difficult the company only prices BOM elements that are not raw materials +However since you did not provide all the relevant helper clases I cant be sure you wont have to make modifications to it +coresrcmainjavaorgapacheminasession doesnt look like it has this problem. +Im not sure I see an easy fix right now though. +Still the problem impacts apps in general outside of Oozie that set mapred +Wildfly does not supports CMP entity beans thus deployments containing such components should fail +AlexeyI think I understand what Mark means +This should be fixed again in CVS +I believe that the create table has nothing to do with the problemThis problem sounds medium hard to fix +Hi ChibaWere you able to reproduce this bugIm also wondering if you have any idea about a possible time frame for a solution Im very keen on seeing this fixed +Moving this out to please move back to if this will be worked on soon +Im nowhere familiar with Hibernate internals so I cant provide that +All code looks good except the first blockThis portion of code is used when the socket connection is being set up +I also corrected the which was loaded the war from +That did it +It would be possible for someone to implement their own metrics implementation with the earlier approach +To save me some effort did you just copy the file to the new location? Or do anything beyond changing imports +yes it was dont thing this issue is completely resolved +included in +If the user does not use the correct parameters we should always output the full usage data in my opinion additionally we can output a hint about what is wrong +Thanks for the quick turnaround on this +so closing. +Patches for and +I know it might not be necessary but this is safe +I moved the time estimate on this issue up to days since I think fully testing and creating a sample job showing this functionality will take a bit longer than the original estimate of days +Ok this entails a simple fix however what are everyones view on how we handle the? Perhaps log and ignore it? Thoughts +all issues assigned to +Please see the comment I added on the linked SOA issue +API coverage as of July +Thanks a lot guys +reopen for set correct fix change for resolved issues that pointed to instaed of concrete +Maybe and admin should even be able to mark process definitions as active or inactive so new instances of these processes can not be started +Let me verify +Hi Kohsuke +Were also hit by this this issue since the beginning as we have a lot of bundles and using those inside together with Apache Aries Blueprints andSo we also would like to see deadlock problem being fixedHowever I also foresee framework config parameters here for the number of retries and the timeout per try +Ah we also want to make it uneditable so that it can be used as a display component +I will move the issue to DIRSERVER then +Yea I think Rob is right +I forgot to mention I adapted and slightly to also support +Hi MichaelI see that you have tagged this issue Fix For originally but later updated your comment to suggest the issue is only an irritant +Rescheduling it for the scope of this issue to just PK scalars per Daves patch +Could you provide a patch that fixes this problem +Thanks Kim the patch looks good +thanks for checking. +This is because the current language support on a blog is pretty lame +Having that in the CRS helper class could work but to put it in such a prominent place I believe wed need some better heuristics such as trying to reproject wider and wider areas and see where reprojection enough +Have you tried specifying the Java type exactly +Its all command lists which are just dynamically evaling the command files for the command classes containing the actual command method for the action +Also artifacts related to product have a unique version suffix so Im not sure how having the Nexus repository definition in the pom could possibly resolve product artifacts via this that I have that off my chest +Without the patch shell commands hang. +However with the are not precompiled and hence the didnt like that fact that it couldnt get a reference to a understand +After creating new JPA project I had this issue +meanwhile can you see any straightforward ways to split it in two +Hi MichalThanks for the patch I will try to take a proper look at it in the next couple of days with aim of getting it onto trunk +Sent pull request please comment the code. +Therefore this is good for trunk +This patch is almost a year old +See my comment from July the current patch is no longer valid and the tests fail +But then you should question yourself whether you need these limits at all and whether it wouldnt be just better to iterate over the parameters using the streaming API without throwing any exceptions possibly discarding files which become too largeThird no one with a sane mind would ignore the header fields in order to throw exceptions as soon as possible +committed +These errors are harmless IMO but I cant think of a reliable way to track them between releases +added serializable and toString to stats +Probably a bug in the way the control flow graph is constructed for finishasync so the initialization checker does not pick up the definite initialization +Im submitting two patchesThe first is a new version of which adds the proper ASF header to myThe second a test for the class and hooks it to the derbynet suiteIll be running regressions again tonight +The bufferSize may be provided as an optional nd constructor argument to either adapter. +Changed name of bean in closed as resolved and released +A timeout callback method must not be declared as final or static +yes mostThanks for your review +A magnolia page with links to the Remoting Guide and the new Remoting Users Guide has been created and activated. +james works with Java +Link with hadoop integration related bugs +Sorry I looked at the branch for and compared with and didnt see any change +I just committed this + +I cant decide whether a disk at should be considered as having enough space or not +My Java skills are not sufficient +Thanks your the best! +Weldon is this issue still relevant +Sorry looking now. +beginning patch +RomanI understand the impact of this bug I was waiting for Nick to take a look at your patchI dont think mvn site build issue should be tackled in this JIRAI would be boarding a flight in one hour +ill do it +That would make debugging this quite a bit easier +Doug says has been removed +Yeah +But youre right in order to send views totally ordered it should be below GMSI suggest try this stack and see if it works for youCheers +I can confirm also that it doesnt occur in Seam trunk +Our production system has been running ok for more than two months now +Applying patch from Michael +I think this can be taken up later if need be +Search for other occurrences of t apply to as we didnt the interrup in FC and SFC + to trunk +Ill upload a new patch soon +Is it possible to get a test case showing your current issue +Heres a patch that implements what I believe you are suggesting +Hope opponent client use proper secure ciphersuite +See attached patch +Ill commit this to the HA branch momentarily + fix +I have seen that we already have a profile for jdk +Ok ill do that Werner sorry for the applied the patch manually as i didnt have time to learn how to apply a patch using windows plus it meant i could set the property to true in so i didnt have to create one myselfHave built a new snapshot and tested in one of my smaller projects using beans created with castor and it runs perfectlyThank you very much for this +This one looks interesting +Please use this JIRA to addmodify the content specific to this chapter +Hi Andrew I saw that this is planned for the next release +In the statemanager the structure was being saved in partial state saving which it shouldnt be in partial state saving the tree needs to be to create the structure +The only reason for this is that you can register in parent and child components with the same key and you should still be able to have a dependency between them +Right so maybe its just a task to document how to replace the auditing interceptor in order to be able to audit entities conditionally +Bulk close resolved issues +on latest patch +Same problem here +I see +Not sure much can be done thereI thought view definitions are supposed to be frozen when a view is created +Ive fixed the NPE +Im attaching a revised patch +Good to see a quick fix +Toby can you check if my fix works for Toby +Done +The dtd file is now loading +Committed to trunk and +If the goal is to use this to support overlays we need to consider some notion of priorities or ordering +Reopening then closing to deal with migration bug. +Thanks +Can you explain what it is exactly you changed +Heres a patch that fixes this and including a unit test +Derbyall and ran cleanly +I think I have tracked down the issues to something internal that was going on while I was doing the runs earlier repeated on our rebase that has worked for a month and the that keeps all the predicate lock code from Jaehos fix as described earlier work as normal again +Combined with IDEA build or greater +This little patch illustrates how to reuse the JAXB context of in +SamisaI believe this problem could be related to some similar issues weve been seeing with and which were triggered by the introduction of a fix provided inI have just backed out this change on Linux and it has resolved these problems +The default implementation is GIT +So it is natural for him to expect the ability to override the functionlity provided by coremodules +Thanks hiroIsnt it good to import test to testexternasl as well +Yes we need to have rules to check correct use of co and contra +It generates new SOAP message and thus the headers from the request are lost +I see that before +This is imporvement of app and it is visible is tree loading now it is much quickerPlease review and commitThanksRakeshOh I missed to give some important points for the patchPlease note +Im looking at fixing this but not using the provided patch +It will be really useful to have SASL based Kerberos support in +can you try that was quick +Hi DerekThanks for the bug report Im marking the target milestone as +With compression enabled it would also be difficult to figure out which spot in the buffer corresponds to which byte in the compressed file if we were to use the position in the buffer to figure out progress +Yep +TomaszWe tried with and it has the same problem +Updated the war project to remove unneeded dependencies and to rename the tes war to +App attempt is saved after masterContainer is received from scheduler +The bug was in the tests themselves sigh +The latest patch looks good to me +It is not yet clear in the spec what should happen here +I cant change and setting the node alone does not work +Its pretty much unacceptable that we dont pause between retries in the master makes the logs harder to debug and puts undue stress on the whole system +Note it is not a fieldType +They passed and I committed +Are there any concerns with removing from derbyall now? If not Ill do it Thursday +And yeah its important stuff though almost everything you can control concerns garbage collection +I destroyed the storage pool and then deleted via virsh and I also confirmed by checking theStill it has been recreating with same name +I need to run findbugs on this too so cancelling has like findbugs but Im filing a separate jira to handles those as they werent introduced here +Attachment has been added with description generation for FOP TRUNK +Tested against the samples and some tests +This includes updates to the class +This requires general grammar cleanup +Otherwise new added tests would fail +Patch committed +Line of the patch I personally prefer leaving the parens hereLine of the patch please use is not None instead of Other than those nitpicks Im fine with this +Configuration you thank you thank youwith a setting of we still get some failures but it looks like that may be partly because we where tailing the log file at the same time as the failuresWe do have a recovery process so that may not be a big issue +all issues created in and before which are marked resolved +slip all minor and it be acceptable if the module and selecting restart or stop worked +Let us know if this is a problemThanks I have no objection +A start +Look for this functionality in camel soon +the java interface that is promoted as service the interface to does not help still no recognition of the new namespaceAttached is the complete client testcasesrc dir contains amongst others the main entry classand the guybefore compilation make sure the whole srcwsdl dir is in your classpath and also the libs classes are included from the lib dir +Bulk move of issues into after vote that we merge branch into TRUNK up on list +I will open a JIRA for that and try to run the test in loop for few times +In addition it helps to have JAVAHOME set to be consistent with the java executable youre actually using +I guess metric names are confusing +Without profiling its hard to tell +Patch is applied to trunk in revision +Committed +MichalCan you provide a sample project for this issue +fsck will actually invoke this API only once and will print out the list of first corrupted files +Alexei I do not agree German did make a branch that we would like to merge to trunkIf we have no history and start from scratch there is nothing to merge and you can manually compare filesIts better he commits to the old SVN and if things are missing in the new one because the Snapshot from Apache Infra guys was earlier you get the diff and itSebastian +Closing a number of old issues that I raised but which I dont think are still relevant +Hi RickI just tried it with the snapshot and you are correct the problemseems to be fixed +im fine with +If there are no objections Id like to commit this patch within the next hrsThanks Chris +I will have a look at lucenebench in the next few days +Fixed in. +HiCan you try and reproduce this with the latest source in the branch of svn? Thanks +Thanks for the patch Narayanan +Closing to necessary issues for Joda module. + to trunk. +sounds to me like a os specific problem with the file name length +Maybe that helps hudson +Attached a patch against xerces TOT which resolves this issue +Possible fixes include keeping track of whether files or extensions should trigger automatic module restarts or not +If the physical jar name does not match what is in the index appear whilst debugging +doesnt seem to be a real issue also because some documents will always fail +The new patch creates the source artifacts as part of a profile called +What are the steps to reproduce this or is there a test case +Another jar to add +Where did they come from? Copyright in the new files should be Grrr forgot to change that in the last file I created +Code is checked in +After all it really seems to be related to and new GlibcSee this gdb backtrace +Bulk assign on table create is needed because ofI created which doesnt use startup as parameter to ctor +Ive just committed this. +Lease recovery triggers two things lease recovery and block recovery +Found a +Proposing here that HS has thrift workers and async threads +Yes +Hate having to support things no one uses but more than happy to support things people DO use +HelloHere is the code for the shift method described above and its associated test +Thanks Army and Mike for explaining this +Now I am using it +Jay I applied Dans suggested fix +Thanks for the fast response! I will try the latest version. +Id prefer additional review of this area before closing the issue +I kind of like this approach with hundreds but we never used it before +Patch in is applied +Ian mentioned he had already forward ported it but perhaps forgot to push +of course you still can do your own cfdumpwritedump by do your own build in tag note you can attach screenshots to tickets. +I found that indeed if I commented out the sleep that the test failed as described in this issue +Added documentation to the wiki component pageSee unit tests for samples how to use the encoding parameter. +Information for classes that are already enhanced are taken from the previous report +frameworktesttoolssrcorgofbiztesttools is for completed scripts per componentthose that just want to provide just the IDE results can use this one. +While making the patch for trunk Is it intentional that the roles of and are swapped +Thanks Jakob for provide this patch +Thanks for the patch +Rebased and merged into the branch. +Can you write a comment there saying that his is temporary and will be removed anywayonce moves to the walker frameworkI will commit it after that +That said if you find it more important that the changelog is in sync feel free to add the fix for version again. +Is it possible you are using the same vdb instance that came back to check the status? You have to make another call to get to get a updated state +Differences might notbe significant on small indexesAlso have you thought about keeping track of which scorers matched so thatthis could implement boolean logic? For example if we kept an int or longbitmask with bits for prohibited andor required then this couldfully replace +Putting old code back until I can figure out windowslinux issue. +Correct ive checked the fix for but that doesnt apply to it seems. +Thanks for the quick response +yet Ill take a look at it this one still valid in +looks to mention the previous test failure was unrelated +Consider this for the release +Try jruby your Java side uses the default encoding of Javas own which may be unable to handle some characters +I have fixed this in the current is the service archive +Slight changes and additions to Dans patches +For me it would be a bug in test case if its assumed to get features in the order they where added +Verified by Vladimir. +Reopen to fix it for too +I guess it does come down to the question about ME plugin what do you think +The Teiid plugin for JON now contains the Teiid client jar. +Or is it good to interleave data and checksums +There is no README in libthrift that states where the included binaries and jar came from +more importantly this should be a straightforward fix +applied new patch and i think were really closeThe problem I see now is that if I apply your patch and svn add the files whats in public looks goodBut if I then run rake build or bin I end up with extra files that shouldnt be in public? public? public! publicA publicblog publicblogA publicblogA publicimages! publicA public! publicdocsen! publicdocsjp! publicdocskrA publicdocsA publiccssA publiccssA publiccssimagesA publiccssimagesA publiccssimagesA publiccssimagesA publiccssA publiccss publicdownloadspsdM publicM public? publicjsA publicStill also seeing docs and directories being deleted by the build +This patch is based on +Now comes a rd consumer +Upgrade performed on the trunkRevision . +Verified. +My personal preference would be though to try to move as much of that logic out of into the connection managerI think the most efficient way to proceed is to create a feature branch off current trunk and commit your patch to that branch +Index Key Optimization false +Raised for the sameThanksVikram +For example if a VDB accesses tables in a backing database and the VDB is deployed to a server before the backing database tables are created then the VDBs metadata will not automatically be reloaded +Don does work for you to resolve this? If not can you attach a patch for this? Then I can apply it for you +Hudson QA bot timed out on the last test run but seems unrelated to the patch +Note also that is a deprecated class in the commons logging apiAnyway Ill make a patch that makes it static and updates the implementationclass in this case the need to sometimes put in sessions andsessions sometimes requiring serializable objects seems to justify the change +You raise a good point although I doubt there are many users of the oldHow about making the take an optional second parameter which has all the options and maintain the first one as the list of columns to pull +Are you planing on changing the value once Cassandra is runningI guess my question is although one shouldnt do this manually are there anything else we could force a flush after memtableflushaftermins is removed? The memtableflushaftermins and memtable through put settings didnt work for me withIt will only flush one CF at a time and that will be the ones with the oldest segmentThe problem is I have a hot CF and the data is coming in uniformly to all keys +Thanks Guillaume +Works for me +It worked with Java SE but garbage collection went crazy +Are you saying that you still see the same problem with Maven +We need to investigate more before submit new patch +Here are some comments based on what I understood from your question +Perhaps we can revisit this for +Hmm yes +It seems like if seam validation could replace the Rest module validation entirely if done properly +Applied +m having the same problem +Would look odd though +I think we dont handle correctly the error returned by JNDIWhat message do you get with GQ and Luma +some simple removal of code that isnt used +All unquotedtable names +Ive committed this +Thanks Dag +What now +The actual cause of this bug is an update in the wstx parser code where it outputs a null for the uri of the empty namespace rather than an empty stringHence this has two solutions +How is this a path? Most of the use cases are around things like aging off old data which isnt at all +I am looking into it +since the latest version is i dont think you have to create a new version +The last version I looked at was pretty close to being okay and the test cases look good so if youre jonesing to get this checked in Im fine with you committing it now and Ill take a closer look when I have some time +This is just for consistency because Henry has already reviewed it +Do you need some more information +BTW already caches canonicalized hosts +Turns out its already fixed in the trunk +If you can reproduce this it might be worthwhile to check if a server reboot really is required or just a couchdb restart Id be very surprised if it was the former +But question is is this extra information is really needed to all other classes except for the maybe there is no need to modify the itself +The test output to the browser is XML and the browser is probably interpreting it as HTMLDid you try a view source +This approach seems to work better +Thanks Sean! +Thanks for taking a look Feng! We can followup with moving files out of the in a future change if we want to with that. +The simple fix is basically to allow the compiler to chunk the population logic for literal hashes and arrays that contain only literal values +otherwiese youd get a +Anyway Ill keep that component in mind when filing new EWS jiras. +Fix should make it into the next patch release +I added a new class in revision based on the previous classes mentioned aboveBoth of the other classes serve a slightly different purpose than the new one so merging all the code into a single class may be a bit difficult +the cocoon source resolved relative URLs simply byconcatenating the strings +Because getting a Status code of FAILURE or SANITYFAILURE seems not possible in this case +Runs good for me with +patch attachment for nutchgora +Im blaming my jetlag Anyway I think what you submitted is probably OK +Please dont use patch +A preliminary patch is attached +What kind of environment you use +Thats okay +Thanks Francis +Yes the API contract is a little different than the original skip +But one of the failures might be an actual problem +and again i changed some things test testIt is perfect now with your last tweakThanks Igor +Both of the last two cases will display Resource not found +Thanks for letting us know. +Not throwing at all is fine with me +Thats a mischaracterization of what is happening +The suite will not run with the regression test while building but can easily be started in an IDE +Im closing this +Closing all issues +Note that this test has been converted in trunk so these errors should not happen in versions after + +Committed +But it would imply that we merge back into Weight? That seems like the logical break +Im marking this as wont fix because the test has been rewritten. +Lowering priority as I am currently not able to reproduce this +Anyway its a very interesting idea and I would love to see how it propagate and its outcome +Just a note on why we need this improvementIn our application we actually have areas of code which refer to external LDAP groups for authorization support +it also adds it to the hidden classes in srcmainplan im not sure if that is really necessary +For the bundle repository plugin this has been implemented as ofAs for the rest of the core web console I think this is a duplicate of rightAre you ok with closing this as a duplicate of and retargetting for the next web console release +a little think that we can put this one to rest finally +BrandonI believe this patch should resolve the issue +There is a javadoc bug that needs to be fixed have a choice in defining the inverse cum whether to define it the way we have or to use and inf rather than a sup +Totally overlooked that +The fixup code in Tika is still a private method in so in this case the upgrade to Tika wont help we still have to apply the above patchIll commit this shortly +Just change the location on filesystem where the vendor jars are supposed to be and create some phony files with touch or whatever +Hi Paulex The fix is fine thanksBest Regards by Jimmy. +In each extension or extension point impl can be extending so that it will be called unpon startstop +Filtering the documents being specified in a feed based on whatever criterion is currently not part of the RSS connector +Should I reformat the whole file +The file is not imported into the JSF project in HOMEWorkspace +Updated as per tech review +This is not a bug +test does not need it passing null +excluded files created by webhcat ee at failures is due to checked in +We should not need to know it when we publish data +What we might try to do is that we also flag parent +this will run all tests for all jcrspi spisomething setup scenarios present in jackrabbit core including theDid not realized that +The issue was being caused by bad jar files in my projects +This problem seems to be more of a general problem using tomcathomelib classes +Works on DB Derby Oracle databases. +This can be further enhanced by adding more patterns +That was really fast. +And it has similar issues such as long lines +Uses the same PDF sample file +thank you +I will look into the issue again as soon as I have some spare time. +if has changed then everything bar external reports needs to be recreated +Will merge to when the branch has been created +I dont see a test or two examples of what you get and what you would have expected +But this doesnt prevent my original scenario to FAIL as it did originally +Could explain your use case? a lot +There is one more idea I want to pitch +Ok I was actually making this ticket with the intention of being more in line with what you have been emphasizing +I like replacing with I havent looked at the patch so I dont know what random string means +Thanks for the heads up! +Ill resolve once that has gone in +Maybe Ill attach the patch instead of the whole file +Committed to trunk +Final patch for review +This new patch adds complete flexibility to the formatting of the email subject +Complete with revision +I have attached the Aperi databaseAlternatively you could have also renamed Aperi to TPCDB in order to use the db attached to but at this point it might be better to use the Aperi db +Its not surprising that you get a cascade of errors as a resultThe error itself is caused because one thread wins at selecting and then tries to set the socket back into blocking mode +Sorry I misread +The previous patch didnt work +Hrmm cant see it in though +An expires parameteris now returned regardless of its cached stateCocoongods please review and rejectapply +Looks like crash is an other bug and isnt related to large heapI reproduced bug on Mb heap +This file generated by WEB facet installer delegate That is related to Dafault Configuration for This is odd it is possible if you have both Java installed in JavaInstalled JRE and Java is default +Anyway I guess that nobody minds since its like that from a while o +I open the pom at the top level of +please +This was fixed by. +This is a tough one but I now have a solution +And thanks for doing the work to get this stuff working on BSD. +First build without running tests build success Then run vpe tests build successYesterday I did the separate build first and separate tests after that +This is the same patch but with a new unit test a new exception type and its handling in the master +Also the change to looks accidental and should be omittedThis is a good opportunity to start with the JSP tests +More testing +bulk defer of unresolved bugs to +Ill make a change to the log +duplicat of +Scheduling this for DR as the upstream issue is fixed for AS this issue as it is fixed upstream in AS and will be available in DR build +EJB dont use annotations so annotation processing is skipped +Moved from to just after merge of old branch into TRUNK +I saw your comments on as well +Im not sure what the effect is of disabling journal locking though so Im not sure how practical the workaround is +Couple of changes I made we actually need to build both Ruby and packages at once the release task needs them both +Could you please test with the current trunk +If it turns out to be a jBPM issue then open another issue +Kevin I had some distractions at work but I should get the patch uploaded today +Just thought Id get some more action on this bug so theyd realize it was a problemI had to refactor some of my code in order to make it compile with the default compiler so I definitely dont consider it a fix +Closing. +Closing issues that I resolved in already released versions +Fixed with r +This issue has been resolved for over a year with no further movement +You should see only problems in the editor +Done long time ago. +I noticed theres a branch for but this got committed to trunk +TCP based stack configuration used for all +Heteroskedasticity What if one passes it the quantiles of say an exponential distribution not randomly but in sequential order to the quantile estimator? How big would the quantile error be? Is it significantly more than the iid case? +This is not going to be done now but is a nice feature to have +I am not really sure how simple that is to do with Protocol Buffers +It built successfully for me with the patch. +Allows you to start Synapse without any issues +This looks good now +Richard can you please specify your problem or use case? I tried idmconsole example with and it works fine for me except for deleting groups where Ive opened new pull request while ago +The current layout is fine for tarballs +Running it on my box +bulk defer to +One thing though that isnt soved yet is how to register these modules +So if it runs successfully I would say Mission accomplished +Fixed with regression tests. +Merged in changes with new location. +It changes several key places in the IO subsystem and Im not sure we can safely changeIn any case its in master and were looking at a within about a month and a half and then well have this work in the wild +This is an area we need to tighten up +Looks like its not directly related to +permitting providing info to callers about whether the config is allowing callers to getdump the topology map for diagnosticsAs far as I can tell nothing calls these new methods +With this change I think the patch is good to goCan you please run tests with the new common jar file with both HDFS and projects and upload results of those also here +Filed to address the root cause +The proposal in above comment is a simple solution but it has an unresolvable flaw +Added a simple fix to allow +I also fixed some issues since computeNorm signature has changed +I finally finished off the last bit +Thanks for the backport Rajesh +Yes thanks for noticing itI didnt change it because included features are not annoying when the oncluded feature is part of the same component and will be delivered on the same by the same buildIt is working fine like this however If youd prefer using import here I encourage you to do it to avoid future issues +Is there any workaround for this until or comes out otherwise we have to rollback to +It has been tested against Suns and and IBMs and +On the server side we need to add a new handler in that can handle the new requests +to running the cheatsheet within the Project Examples wizard would appear when the example is can add one or another or both of are you suggesting +Fixed as per Dougs comments +Just a note that this is not related to the original point of this issue +Attaching +Then they can do things like customer maurice and other advance search options +The proposed patch. +Skipping not very important there +Tom is there any way that this can be handled automatically +binjruby case so I reopen this bug +I could reproduce the error while changing the daylight time flag in windowsAnyway I think that this issue is not a showstopper for a release since its a testing problem which occurs on specific environmentsIt good to understand when exactly it happens +Now it is +Thats all! The patch was good +Thanks for the info all +This shouldnt impact JDK since doesnt support exception affects and fix version to +A warmup could help if I knew exactly what to invoke to ensure everythings warmed upIf it were IDEA sources Id just replace with as we have one +Cool +guess I am the lonely decenter I had to turn on my browser java script to see what I was missinglooks like I will have to do more manual appliation of commits from now onwhy do I keep the JS offa lot of sites are showing up that use js to override pop ups and install applications that are usually viruses or trojans +Im pretty sure this was already fixed as part of +This fixes the PierreYour patch is in trunk revision I did not backport because this is really a trivial change for an improvement and not a bug +Thought of another possibly useful payload convenience that adds the Token type as the payload +Thanks +Youre right please ignore my workaround +See my first comment on +Thanks Aaron +There are already some development around thatBuildr reuses a lot of maven conventions and brings the power and flexibility of a scripting language +If there is a problem please open a new jira and link this one with that. +That is strange because the code at is only invoked if it finds a valid cookieAre you using your own +Thanks Dag +See Cleberts comment +Patch file containing the svn diff +What is worse is that it always updates the metadata for the default skin even though the metadata already exists in the users repository +Im new to the codebase so apologies if this wasnt the right approach I essentially duplicated the junit provider into a new provider where I added the support for havent done extensive testing but there is a basic integration test which passes here +I was taking what I saw too literally +attachement is test the changes to trunk At revision +Many thanks +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Thanks Clinton! Looking forward to the new release +The last comment seems pretty close to a for the patch +Thanks for the reviews. +Fixed in +Retrying hudson +Is there any update on this plugin responsibility being transfered to castors team? Getting newer versions would definitly be nice and since the castor team is ready to take charge of this plugin and they are already hausmates it shouldnt take that long +Even if both references are specified in deployment descriptor and injection is not used deployer refuses to deploy application with similar message stating that A is waiting for B and B is waiting for A +made private as the second type is fixed to String +The number of maps is controlled by the slicer +I couldnt find a nice way to go about it +I will create new Jira if portion is not checked in by Monday +Tests will help ensure the correctness of optimizationsI suspect youll find that implementing the copying solution will not be much easier than a COW solution and so seems to me like implementing the copying solution will result in a lot of wasted developmenttesting workTo be completely clear if it makes sense to implement a copying solution as an interim development stage then thats fine but I dont think that a solution which copies all of the filesdirectories that are being snapshotted should be merged to trunk +Load indicator is just one example there can be myriads of functionality that require this +With the changes in this patch all of the tck tests pass on the latest JPOX snapshotChanges in this patch include added tests for configuration and added tests for for the new Map Map +How about we shutdown these threads after a node decides to be a leader or a followerThe threads get created on fly +I think the already checked this +The useful one would be dependencies that will be included into the classpath for scope scoped to dependency graph and hierarchy pages in Maven POM editor +is done +Just to clarify I realise that there was no disagreement and that the choice of words and phrasing was inappropriate. +Anyone? +The sets property based on its own configuration +Just added an based on the tomcat implementationYou can configure it with a servlet init param or a service providerThe fallback implementation try to setup a Processor base on the of Handling annotations wrong order +The service starts the plugin +Sorry this bug was my fault for using the wrong reader +Its part of the build system to help users translate their applications +committing. +We can add that to if such a need comes up later. +Closing issue. +In the meantime Ill take a look at this patch and run it manually +is this roadmapped to be fixed in the future +Attached patch +While Im confident that I wasnt halucinating Im going to close this issue until I can reproduce it reliably. +from me +I couldnt find a pattern +Does it mean message is in memory in the only replica or do we have to wait for the message to be flushed to disk +Anyways thanks for yourgayatri. +Henri does it solve the issue for youPaul I think we can maybe just add this +Simplified the proposed solution a bit by removing the lock and readded the that was missing in the last changeset +Patch +We can support other file formats just not all of them +The pools is being shutdown just as they were already but now uses the for that so we have a central API code for handling this. +or you use the string constant versionsanybody have strong feelings +Updated unit test and new bean showing the crash +You are correct +I just committed this +Thanks for the POI updateA commons project for mime type detection seems a nice prospect +Thanks JingThanks also Vinay for reviewing this +because they are required more changes and separate thorough regression is required with MAPRED when i get the time i will work on them separately +Apache Felix Web Console has been released +The problem exists on have the same problem and all the suggested workarrounds dont fix it +Should we really be adding this to Common or might this be better as a new Herriot subproject +verified. +I tried with and according to Bela it should be fixed in CR +I will have a closer look tonight. +Upgraded dependencies +I kept it running for a day in a loop +Attribute disabled hasnt been implemented yet because there is another bug with it +Thanks all for the review. +A small addition to the fix the path +Closing. +The default value is zero so it does not introduce any problem in already existing code if you want I can create this patch against branch +Most of these changes conflict with the patches from +Moving out to move back to if this will be work on soon +Could this problem be related to +HiFrom that issue its seems that it should be ok in mode am I reading it correctlyI think we can probably move toThanksChris +Its unrelated to this change +Would anyone want to try this one +Youd have to check with them about whether or not they ship to Slovenia +I tried to apply the patch but it doesnt apply on a clean checkout of the repository +Maybe some sort of tiered thing where the default calls the originalYesSo if the CP overrides this its done thereIf not the base Observer class that are supposed to be extending will call the original +If I understood this properly does return different exceptions depending on the status of the transaction +In order to make Helix ready we need to do the following +So please give it a try and let us know if you run into any other compatibility +That should work +test case fixed +Latest patch doesnt compilepAttaching one that does +Fix for standard without Grails +I added a quick hack with rev to make sure the lucene index can handle empty node types +While committing the changes I made two errors in the last two commits in this should refer to not do you know how to change them +Looks like the fix worked +Hi LaxmanThanks for taking a lookI will remove the flag and perform the test again and let you know the removed and performed test +patch was applied at the revision +bq +Thanks Jonathan +On a related note should we add tests and run the same instructions from the wiki page nightly to ensure our wiki pages dont get out of date? If so does it make sense to do it for just the latest release or all releases +Hmm its somewhat hard to decide +Thanks Jothi! +Committed v +I verified my patch on my favourite one +Note that in the latest version of UNICAST are removed on a new view so is only used as a second line of defense when there are no view changes in a long was implemented with an which has a list of members each with the time of insertion +Suggested fix for the issue +I see why +In fact you can rescue an edit history from any known revision +The developer would have configured the spring context or camel registry with a bean mapped to that key +It seems to work for me +Ive modified the original patch so that the test data corresponds to the current expected output +Point me to the source Steve and Ill give it a shot +These patches update older tests to now expect two additional export links rtf and wpdf +Since this problem affects the branch can we get this fix into the branch as well? Thanks +Should be fixed now and added an integration test to help keep it working as expected +Hi RobertLucene dependency is the only change right? Or you also upgraded Carrot from +The connection manager has a shutdown method if that is the desired result +bulk defer of issues to +memstoreUpdated should be set to true before we call apply so we roll back any partial memstore updatesIts only really important to set this after we call checkFamily +DoneIve commented out the mailing lists in the attic section. +Those are the actual items that need to be deprecated in EAP +Right I reproduced the issue +Hello Evgeny I am actually seeing this scenario in +Thanks for looking at the test and the further comments Knut +The Java backend currently cannot handle this because it translates to Java generics +That is a misunderstanding +If it HAS to be put out past release I would put it in the weird exception handling that was wrong you were right Lucas it was purely an artifact of the way that used to have all that key generation and recovery stuff built in +Attaching management server the issue after verifying the fix with the latest build +Which branch? Already committed to and +Thanks David! +I committed the patch to +Okay +Thanks Oliver +I think we should patch this as well functionality was broken +Changed the test to have jobidappid +Waiting for functionality to be added to JDO +Patch with implementation of this feature including a test +Though we have to review that usage in general +Agree that rebasing is painful +Ben firstuser can choose if remembermeif user choosed remembermehe can choose how long time not fixed daysfor exampleuser A would like to autologin for a weekuser B would like to autologin for a month +Updated the FS with API updateDone implementing +So this patch will improve things +whats your committer roll +Ill must do it in few weeks for a customerFlex project +It seems to me that just porting the one line change from will fix the problem and less risky? The unit test in the current patch is still usefulConsistent issue should be fixed in trunk due to +Thanks for that detailed note Tom +Providing a patch without the lines added by IDEA might helpoh sorry about that +Im sorry but I have to reopen this issue againWhen I closed it I only tried the jmx connection with local servers +looks good! Die die! The new attributes are more +Thanks +Too many browser tabs +Dont know when this will be fixed if at all requested by popular to be fixed +Karaf branch svn ciSending webconsoleconsoleTransmitting file dataCommitted revision +I guess no conflicts with geotools? Looking quickly I dont see any so yeah I would say go for the upgrade +Agreed +Thanks Dan for the patch. +Excellent analysis work Benit +Ronald van Kuijk This really needs to go through the Facelets community this is not in Seam +AntonioIll have to close this one as cannot reproduce unless I get some new informationDid you run to confirm your version? Did you run the latest code from CVS? +Reopen if problem still exists with latest nonoss builds. +FMCs karaf logs with stack traces on Windows and you still see the after a restart? Im finding its appearance is a bit sporadic here which is making it a bit more challenging to troubleshoot +Looks good +Clearly you do not understand the concept or purpose of the hilo algorithm nor your responsibility to not insult people who are giving you free softwareThis is the end of this conversation +or as Sacha would say I need you to FOCUS +The two new functions added will call the appropriate canonicalization method for the given algorithm +Assume closed as resolved and released +I rejected the Pull Request even though it was merged +But as a feature its pulled in closing +IE is not a supported browserIf you are stuck with IE I recommend taking a look at Chrome frame as a means of obtaining a compatible browser +Reopened to set the Fix Versions field to JDO beta. +Only failed for ant test but it is not related +The http response code is from fault that means the server side successfully delivered the fault message to faultTo address and faultTo endpoint response with http OK +It might be that there are some specific file systems where this works but its far from the norm in my experience +further all bundles that were active are activated. +Am I missing something +JasonSeems very reasonable to me having a stalerelaxed optionWhos in favour of this option +That was YonikI made the issue a blocker because I want to prevent index corruption +The process for patching the JMS DB configurations is done differently for a better check for applicability is required for and forward. +Thanks +Updated James test reportMy tests run clean after fix for +mevenide was missing some functionality +The other alternative would be to make it available with something like couchdb s which would require some other voodoo magic +Hi Brianseems filtering over version is not supported in JQL for Quick Filters +Ive just used rvm to install branchrvm install branch I get the same error message when running bundle install with the attached against the branch +well get it applied for in with release +There are some test cases that have class strings I need to update +Committed the change to as well +Please also add this change to release +That said the last commit on trunk that I see was in January and there was very little activity before that +Then in your case what i feel is before even the call back for CLOSED came to the master the SSH has started for the RS that had gone downCan you attach the entire log +Code from Maciej Lizewski which may be helpful is included in +other approaches would be using reflection like the layer does but I personally prefer not to use reflection and rather use delegation in favor of inheritancelooking forward to see some comments +how important is it that we use Spring? We talked about changing Struts to use it but I dont know if itll make +Any thoughts +Lets go with the traversing but I think it should be done in the rather than in Configuration +A number of other features also rely on +Did pass for patch v +I attached missing file for tests can we please apply it +Yuki there is an index sstable with keys and that one could be fixed as you suggest but the actual data sstable also has keys +submitting to a new v +Screenshots taken through the process +Open to track Dmitriys comment. +Additionally isnt using the get methods the Right Thing +The run was running on which is a machine with GB +The right approach IMO is to fix and close this issue +This has been discussed before on the listEven picking only common varieties and assuming consensus can be reached on what is and isnt common the final number of combinations the product of the common versions for python java c compiler and OS is just too large for committers to support +Wont persue. +The current design clearly implies that a single handler is invoked to control the navigation after authentication and that is its main purspose +Closed upon release of Hadoop. + +Feel free to comment this issue +so i am not sure the best way to proceed +I believe there are some communication errors hereI dont think its technically doable for us because were using policy based VPN rather than route based VPN and its almost impossible for policy based VPN to add an static routeIve updated Manan on this +This has fixed with the help of the team +RustemEvent reporting is now optimized in the whole JVMTI implementationPlease check that the reported problem is fixed now and close this issue as FIXED if everything is ok +Hi guys someone can take a look at this patch and give me some feedbacksThanks in advanceMichele +Actually the correct way to do this is to check +It seems to not be a Jbpm problem +Forgot to mention that I have run the Junit All suite with no new failuresI am currently running derbyall +I also noticed that archiva didnt delete site descriptors +We could add this at either layer +Which version of Solr were you using +Set it to via a destination policy +The Javadoc states Date formats are not synchronized +I have committed this +I thought opie was a one time passwordBut yes I think something happened during me trying to get opipasswd towork so a reset will be fine and this time I will let winkey generate theresponse several times to make sure I used the correct passwordSorry for the inconvenience but these are new things to meJan I +Fixed in r +Attach a draft patch and it works in my local environment +The one referenced in the will always be remember me authentication There is also an interface named and an implementation named which can determine if an Authentication is remember me or not +Part of JIRA +It will be triaged option is to use the toolkit from the project +This is Axis behaving exactly as expected +Even though we can configure to let soap message pass into the JBI container it should be cxf bc do the schema validation only let the valid soap message for the clarification Freeman this makes sense. +This was causing problems and has been reverted in +This patch also handles this per DWPT and not global +attach a patch for this jira since this ticket is requested for quite a period +Eberhard can you please elaborate how this relates to Web application initialization? How are your ideas specific to Spring MVC applications but not GWT or JSF +Ideally it should be a static method of Tuple however Tuple interface do not allow me do thatYes a static method cant be overridden +Only if it is not too much work as the standard deviation will be high on EC +There should be a file in either the JVMs directory or the applications directory +I suppose one could argue for elasticsearch Sink +bcf +the IT seems to be fixed by upgrading to doxia theres no error during the IT but the postscript returns false +Also Ive temporarily been using this in our test environment for weeks and no issues there either +will be fixed in +Heres an example xsd +javalicenses +Maybe a good problem to have +OK I found it +This seems unlikely but is possible +That way one could start allocating the rd stage as soon as possible but not so soon that we have the above problems +No need to spend time on building it on windows. +This adds a paragraph to the JDBC Reference section of the Reference Guide +Patch for default values handling +This patch fixes these issues and also reduces the log level on reconnects from WARN to INFO since this will happen on the first connection attempt every time and is normal behavior +can you post the wsdl that you use so I could test it and figure out what im doing wrongthanks is still a problemanybody have any thoughts on this? does the XML parser affect this? What does Resin use instead of Xerces Crimson? I would say you need to file a bug against the parser that isnt adding the namespace +Socket readtimeout has been moved to +Thank you everyone for your suggestions and input! +I also wonder if we can extract a bloomfilter class into utils that builds an this could be useful elsewhere +I will take care of this update +Thanks a lot Dyre for making these fixes +Great idea! Perhaps call it info? It would be a command yes +Fixed the issue mentioned where implicit casts are not being inserted for +In particular service configuration objects are not too hard to create and debug +Screenshot attached. +Yes + +I am looking into it +Im more wondering about what type of I should write? Implement a Filter that uses it and write a test for that? Should I put up a preliminary patch without a test for the time being? What I have now appears to pass existing tests +If you have any suggestions on things I could have done better in that patch let me know +Marketing Brazilian Portuguese at rThanks Luis +Hmmmm I need to check with Tristan but Netty upgrade required changes to Infinispan itself +Committed to trunk +What I need is the concept of the native SRID stored inside the descriptor metadata and access to the in order use it for encoding string literals. +Heres a draft patch +Im hoping to work on getting a release out this weekend so if youve not got something by then Ill probably throw one out +Attaching +Its possible for TM to get nulled out after we check it +Snjeza could you please review the patch? I have changed a couple of things after speaking with Max today +Tried to upload the patch on RB but even with some text mangling to hide the gitsvn differences just bounces it back to me with a nice errorNo clue how to solve it sorry +need more details to reproduce java stacktrace create statement for the invoked table cfml example additional info here? Otherwise we will close this ticket within the next days. +Incomplete Patch work in progressMissing I have to learn Ivy and make it pull the Pig dependencies +Not logs or test cases given to reproduce +Broke out the requirement for unsubscribing durable consumer into separate issue +Ebay category code may have more than character so removed condition of ebay category code length +Ive encountered this behaviour in several other scenarios on some other platforms as well however I havent been able to reproduce it deterministically +I dont think you should refactor solr around broken assumptions and misleading benchmarks. +Actually for cloud environments we dont plan to hold such files in memory but rather in a virtual file system one that is ring fenced for each tenant +have been committed the new revision number is This solves problems in subtasks and +However rename to +ejb stuff so it may be easier to track down this bug. +I just committed this +Thanks for contributing. +edah are you interested in making up a patch and adding a regression test for this issue +Same for root module +We always chose the mainland strandard computer terminologies during the translation I will check it over and try to get a list of them +QA Closing code change +If you know for sure that youre never using lifespans or maxIdles then disable the reaper thread. +not a dupe of +I think you are conflating tiny with metadataIm saying we should declare its metadata thats all +There are some moreFurthermore I dont think that this is that slow when using reflection +Put is a list of +Ill look at using your entire project +bang +Sam + +Format classes to convert evalita format to +Attached you find a patch for review +This one is not intuitive to me what code is creating empty? If we can just not do that or maybe create a empty object that might be preferable +Added size checking to reduce the number of copiesThis patch does fix the output corruption failure we were seeing +bulk defer to defer to +As the are created with the exact size and dont need to grow you can use fastSet to speed up creation by doing no bounds checks +Move to CR as not critical to fix now for CR +Just got back from dinner +There was slightly more to this issue than the contributed patch added patch +The downside is the extraneous memory consumed by the hash map however if we use thisll be reduced or would it? Because wed be writing the term bytes to a unique byte pool per segment? Hmm +Im looking at this issue and I have to agree with some comment on your post to the Atomikos forum that the most likely culprit is a change in how the Hibernate session is managed +Fixed r +functions and added +BTW its a brand new deployment from the latest fuse releaseI also attached the configuration dump from console and hope it helpsThanks again +Id like to remove from beehive by deleting the filescontrolstestinfratchcontrolstesttoolstchrdpartyIn my local checkout Ive updated the build files to use the beta under externalxmlbeans insteadant drt and ant both work fine with the updated xmlbeans jar +When the real PM is closed and its cache cleared +In this case I return Since I only do all this stuff if the is this should not affect the performance for normal use cases +so I finally found that bit of time I was looking for to wrap this up +The test passes if you use just pure Spring code only because the test class is in the same package as the service interfaces and impls +question is why does it show up only with hadoop? +Closing this for now +Unless all strings are translated using the same WPTG rules we could easily break cases where doubled quotes need to be retainedThere are customers that currently have their bundle strings not just in these resource files but in other places like database tables libraries etc +btw +bq +Where is it meant to be applied +Added to replicate the have added a Maven project to replicate the is what I didThere are three channels and two endpoints linked likechannelOne bridge channelTwo transformer channelThree output only logschannelOne channelTwo and channelThree It should at least log the transformer which was definied using the annotation +to trunk +GregoryIve tried to disassembly bigger part of code but failed maximum stepping back is Regarding illegal instruction both times generated code is looks similar and rip register points to illegal instruction +In the light of discussion I am asking why the cache information is not kept in individual files in the artifacts location in the local repo itself +all the drivers have been uploaded and the build properties have been changed +Confirmed still an issue in Apache Flex develop branch. +here are the steps +Something like that explicitly loading would be a good way to ensure the script continues to work +Thanks +It wont break any existing pom files +However in a fashion similar to the job log history it would be very nice if the scheduler can output structured events and metrics into a separate file so that its easier to parse +If we think about later release runtime should give theinformation or a interface to us +Thanks far as I understand all these properties are part of the so as long as you can use it as a payload that would satisfy your use case +Javadoc in refers to +Added Rossen as a watcher as well as hes currently investigating async options in that will certainly be one part of the implementation +Its not suitable for tens of thousands of cores +Looks good let me try tomorrow + +Pretty please +bq +i believe this addresses the problems in as well +Hi Kathey +It looks like this has been fixed elsewhere already +The action keeps being called +Stevens patch synced to trunkI plan to commit shortly thanks for the configuration tests Steven +I just checked in a new Glassfish example in examplesjeeremoting +I assume this means well be able to switch to using NIO for the distributed search +on this patch to fix the tests then +what should happen with the file after the exchange has been processed? should it be deleted moved etcThen it would most likely to leverage a for the GET command and retrieve the file that way +There is no corresponding object in metadb +Bug has been marked as a duplicate of this bug +Any update on? Do you want me to look into it +If the column value has indeed changed he wants to read the latest value back and MERGE it with the local value he has and then reinsert +I could not upload the patch today as still some test case is failing +I think it should be written using unicode escape characters to work on all platforms +Could you please take a look +Patch applied in thanks +The documentation is still outstanding +Jesper do you agree +My impression is that you are referring to some component +I hope so +May I know the solution for the above problem give the guide lines to fix this issue. +trunk patch v is in +This is working fine and I know we added some tests for visibility issues like this in the past few months. +Anyway I am clearlygetting the result I statedSecond what I was trying to say is that you have tied the to the application in a way more consistent with aprovider than a factory +need to revisit if cache properly works for the clustered cluster specific test which passes. +what download link +I suggest that input has not only sequential read but also random read feature +Sorry should say human readable so instead freezeconditions use freeze conditions +It looks as if comments coming in through arrive multiple times +As far as I can tell none of the broken or incorrect links have been fixed on the site +FYI the problem is that the field is marked synthetic which seems to let Hibernate ignore it +Thanks AkashYour patch is in trunk at r at r +Nicholas note that the patch does not implement the solution that Owen and I discussed and agreed to +BrunoCould you please try the that I just deployed? It includes a fix for +release bulk close +We cant reproduce this issue on Mac and Ubuntu but anyway it seems to be an upstream bug +It would be great if we can get this into trunk as a starting since I guess will require some more discussion and work +When you use viewer tag render report +I changed the priority to trivialReason is not the triviality of the issue but because this issue is caused by package split so I need a way to filter the bugs that are typos caused by package split and real bugs +Similarly while writing data +Please check if the issue is still valid. +This is due to bad TLV or LDAP message composition errors +Patch committed in trunk rev +Open the file that you want to review +When we use the HDFS local read we should set property to true +Butsince these are really exceptional cases Id consider hardcoding them in theAs for the use of as said before this class only works forprotocols supported by the JDK +Hopefully we are in time for the GA +And well be able to test it +also has the same failures +Either way the real dependency on this feature is on a Descriptors project which is pending resources to work on it. +Unassigning in case anyone else wants to look into this but Geronimo and includes the udpated txmanager fixes required to enable failover support in the connectors +I think that Lily is going to very close to fixing this +As requested +. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Most linuxbsd system readdir return filename and typeI mean linux readdir in my comments not the core API list +Kindly rename property file to +As Snjezana pointed out by email Mylyn for eGit can be installed from JBDS Central into Juno EE bundle +Typically converters get used transparently when you declare or method arguments and there is no need for you to call them +does not package GWT libraries. +Convenience + agree and wanted to mention we shouldnt limit ourselves based on packagingfor example we can have and but actually shove the into the jar for simplicitypackaging purposes if we wantbut this way you could still use the analyzers without the lucene core if you wanted +AdrianI have please look at revisions and +The map of past failures to sounds great especially the bit where that are not in the failure Map are prioritized +That is a great idea! Feel free to provide a patch if one of us here cant get to it first +I just committed this +Committed revision to xI will mark this resolved for now I think we are mostly there +Just out of curiosity who do we intend the INFO message to be for +Another Linux run ran into a exception but that was after running out of disk space +bq Can we discuss the client side dependencies issue over in? name of this JIRA to reflect the specific issue Im solving with the patch +In my mind the JSF solution works well +Looks okay to me +Can you attach a full project that reproduces this problem +I can seem to edit my previous meant What happens is that jersey SPRING uses the same method to register the resources as beans + good idea +I think this patch should be included in the new RC +Delivered fix to both branch and trunk. +I assume Heiko made these changes +Now it appears that some things that were previously logged at levelFINE are now logged at INFO when they should be DEBUG +Close it when the release will be in r +Prachi can you check on this issue in PST to Prachi to see if this can be addressed in PST back to Kishan the upgrade steps need to be documented clearly in it as major since no code changes are required +Fixed checked in at revision +Added it also as a issue since its present there as well +Thanks Bryan +Unfortunately this means that we can not completely remove sax codeOn the other hand using sax does not mean that we need a dependency in jdk So I think we can close this issue +I think the call is being fired but something is up with the way the validators are processed and i have yet to figure it out +thats what a quickstart is +while changes in visibilityenableness should not effect processing within we cannot guarantee it +eps files for this version of the logo +Separated main form elements from display content +As you can see when you run the unit test with the last patch +ids of the components to update on the page encoded into response body as the meta jira has been closed as part of a review and archiving of the aj project +Len this feature is not done yet +Such a value cannot exist in Derby so its not applicable for derbys documentation +The name of the method is changed to get +Im working on a change now that will mean the jar files are no longer copied in lib during development +Thanks Hitesh. +Patch applied +This completes the work of this +Any update +Committed revision to trunk +Commit +I just committed closer inspection of the code are the fields sort and prefix of used anywhere at all? They dont seem to be referenced anywhere in the code and just removing them doesnt seem to have any obvious effect +Cheolsoo can you upload the new patch when you are ready +It means that I cannot verify the error if I cannot remove the global +But its doable and thats whats most importantThe multiple projects problem is another thing +Thanks Chinna +duplicate of +Tapestrys mixin is a thin wrapper around Scriptaculous +so console config and code gen launch config is fine in this dialog +Or you can give me your email address and I can send you the binary +I think we should avoid making a decision solely for license purposes +So maybe reproducing it after priming the default page is dependent on other factors such as system ran the test without priming the default page and was able to reproduce the issue +Hi ErandaAttached code change is a possible fix to pursue +User can also provide custom converters for specific user domain objects +fixed in trunk along with some fix about hazardous merging of configuration setCould you give a try with the last build +modulesfacet which has a separate manifest for its example jar +Afaik we have never seen the nav tree being empty after this is probably not a big dont know yet whether the tree will load using AJAX +My patch uses this information rather thanthrowing it away the way the current build system doesIve tested this patch by building with and g on Solaris +Test failure is because ofI just committed this +Mario and I have discussed this and found both some problems and opportunitiesAs a result of our discussions Ive pushed a lot of Orchestra config down into the now making orchestra less dependent on filters for JSF requestsThis will hopefully simplify configuration for people doing JSF and might resolve some issues in portletsBe warned the code hasnt been well tested yet +snippet of full to unpack eclipse unpack jbt pluginsfeatures and run eclipse in a clean workspace logging to like p picked an older version which it thought was compatible +There is an option Convert to without BOM +but I did need to increase the delay on the client side +It was released with the fix in +bat file as well +Or at least I tried to avoid smiley faces +Im OK with doing necessary tweaks in follow upId suggest we merge and to trunk +commit for branch at revision into fix versions +By default RAT is no longer run on the build process +we add a commented out property to the default file +If getters are used instead then they should be used everywhere in the class +Committed wupdates to +t Maven check which JDK is used and abort with an error message if it is something later than? Im sure it can be done with Ant. +I kept listsnapshotsIntegrated to branchThanks for the review Matteo and Jon +Updated time estimate to account for general testing and handling of rejected tasks when an invoker is being rescheduled prior to removal from scheduled invokers +FWIW I added a HAVECXX autoconf define you could key the unorderedmap headers off that +think thats confusing +Or at least another manual build +You must work on some large project +Sure +This issue is currently blocked byI was reading through the and Myrna saysI think you believe a is a tool to convert the test to junit and it may have started as a quick way to run certain +Patch is applied and committed +In our case this includedBundle is only honored at creating the site using our scriptI cant remember which solution Fred chose to get latest version +I had a bug in my build scripts and there were old groovy libraries in the maven repository +Set the Fix version to R so its easiear to keep track in fixed in latest code drop +Fix the failed test +Looks +bq +I used second option and its have only one remark +ran the script from various directoriesprojhbaseprojhbaseAlso faked removing protoc from the path and it reports nice message +The failure is probably due to some problem in the databinding framework +Sqlserver is currently untested though so cannot confirm it worksThis is a qd fix since I cannot confirm if this solution breaks other databases +Many thanks to all involved. +Have you tried asking on the Tomcat User List about this? Tomcat Have you tried asking on the Tomcat User List about this? Tomcat Bug looks similar also maybe Bug Niall looks similar also maybe Have you tried asking on the Tomcat User List about this? Tomcat Bug looks similar also maybe Bug Niall? that this helps to solve the problem but Im not terribly surprised to seethis kind of problem +Ive added explicit checking for this so that usage without will clearly lead to an now suggesting that is might revisit this later on but for the time being there doesnt seem much incentive for providing a version of support +Marking fixed issue as closed. +Does the REST unit test really need a with slaves? I would hazard a guess that there was some going onThere isnt any single thing that can fix the build but I still think this approach seems like a reasonable start +At most it would generate warnings +I just have got failed several times +Fix for trunk +bq +What do we do? Do we take or the plugins I have added in the sandboxWe also need to add the ejb and par lifecycles to artifacts Ive just replied on the user mailing list about this problem +We can come back to this after a discussion in the mailing list. + +why dont we just remove it completely? What is the point of giving user options to name the classifier in the descriptor? +Hi PaddyThanks for the nd report +Can you post the javascript errors youre gettingAlternatively try to upgrade both implementation and component set to a current nightly and check what happensregards this seems to be a duplicate ofregardsMartin +Probably others who rely on schema code with deterministic ids +bq +is attached to merge the final code and some fixes on example on top of trunk commit I vote definitely +Then we should decide going forward whether this should change and if so for which classes and how +I am able to download the branch from github site +apologies Adela +I just had to make a clone of so that doesnt require the lastest version of IvyThank you very much for your contribution. +I am not sure this is any more stable Consider the power loss just before the sync callMaybe loading preferences should be more lenient in that reading an empty preferences file actually is just silently ignored and considered as the file not existing +Err +Hm the debug mode also works when starting ofbiz within eclipse no problems with that +What do you think +Remove that if you want snapshot still doesnt work +Also setting Fix Version to as we have a lot of stuff to get done before as it is +The remaining error is legitimate +properties +With it stopped am under Win XP if that matters +I thought that best would be to add as dependency libssl against which was mesos built because its statically linked +This patch has been applied in the revision +Im neck deep in geospatial Lucene extensions. +Worked so far +for line and following when the eviction happens +Bug reports are fine of course but discussing couchdb belongs on a mailing list +this is a typical convention from Maven since it configures some directories by default and still needs to work if they arent present +push to +I have bit machines and bit machines +Just the replication resultset cache name is not default in that HA add we should be mind what I said sorry about confusion +Fixed an error in testAdded an Order Customer example +hmm +Going from to caused run times of a few minutes to balloon to days because of this issue +Substantially done except we should still add beforeafteraround facets +My brain must not on the right place when writting it +seems to work now +And it uses less lines on the screen +When the created topics destination is the attribute shows a zero value +I havent had a chance to check out the rest of the patchissue but for this specifically what about a convention? Anything under the persistent key in the commit data is carried over indefinitely +Is there any ongoing or proposed SQL standard work regarding user defined aggregates +please dont schedule anything for unless its a if it is on the way +Its not necessary short term imho but we can add a hook in case someone wants to do it +bq +needs to used with EAP alpha or libraries fix the issue upgraded the to use latest EAP try it in these I am planning on a release tomorrow on this so if you can check I can include any fixes that are seems to with EAP final and TEIID. +has no control over the process of port assignment to individual connections +Best RegardsPhilippe Van patch provides extra warning logs in case of preemptive authenticationfailure due to missing default credentialsLooks good to me +There is very few to include in this chapter today JEEbooking example is working only on AS and JPA example is working out of the box on both major versions only with one difference loading PU by container or directly by Seam +Patch applied at +bulk defer of unresolved bugs to +We could call this a new feature for it to migrate new columns +Verified the screen does move once in a while and is NOT completely locked +Thank you Chris and AlejandroBTW I could not get the merge to build in +So I fixed this by returning a Server error No authenticated user instead of a nasty get a server error No authenticated user no exception even when I access using url portalauthdashboard this happens only for the dashboard +Included +to this patch +Fixed for both TCP UDP to master +Given the way the test case works if thats the broker that you happen to be connected to and you kill the broker the client might not notice so it could appear to be hung +Ran into this recently +on v + push to +The fix has been applied to the branch and should be seen in the ER build +Everything seems ok +Its an odd one to be sure +Thanks for the helpI appreciate your efforts and comments +Added javadoc for the parameters and methods + +can you precise the svn rev? I dont see anything either in svn or in is still at the change is in rev +This issue was fixed by added workaround to plus by again refreshing from current after restore as is already released +mbox files are zero bytes +attempts to prefetch invariant values into a cache +Zone was created and email was sent to grisha at +on applying to trunk and too +Can you modify your test to make sure that indeed gets printed + added DISABLEDtest demonstratesthe problem +C backend does not have this problem even with OJava backend without O also does not have this problem +The zone has local storage enabled as well +This contains the and the firebug error report screen you mind attaching your war as well? I want to give it a try. +And this patch makes many benefits with traffic for to reproduce in +JSP compilation has been in the latest nightly builds +Ideas which have been discussed in the community seem to gain acceptance the quickest but if you provide a method variant of the Closure memoize you probably will be close to what will be deemed most useful +But this bug title is insufficiient for us to do anything +The massindexer was not designed to be run concurrently with normal activityWe should lock the or make sure that we share the same from the backend +bulk defer to +removing the workflow interpreter REST service reduced the size of the package to MB now +I didnt run the code but looking at it yeah it seems like it never parses the arguments so it would def +I didnt see this on EC when I was doing perf testing or just in exploratory test wLTT +Thanks Omkar! I committed this to as well +I only wrote theproposed patch the way I did because its doing the same thing the original codewas trying to do +a little bit of cleaning up today looking good +how many nodes you have? We need here about seconds to stabilize ring +The feature Remote Replication of data is implemented +Yes it will break all of the existing plugins + demonstrates the issue +Dan is this for before GA or something to put on or +This file can contain highwatermarks for partitions that live in that log directory +Changed makeQuerymakeFilter to makemake respectivelyI want to address the method javadocs before committing this +Each instance where we are doing this requires a new issue +it would be best if Stripes guys can using Sonatype service if they dont their users can still do thatSonatype OSS Repo Hosting service require users to deploy artifacts with pgp signatures so the user info is tracedwhy doesnt Stripers guys use the service? ask them +got an idea for a fix +The logo and text themselves look great to me +A general solution to this would be great but its a kind of bootstrap problem how to find info about the appropriate snapshot repositories for a project without looking in any snapshot repositoriesFor the specific cases of sandbox components it seems reasonable to me to add the snapshot repo definition +Derby does not build after that port +mavencacheElement +Thanks Patrick! +I checked out and replaced the custom +Suffixed field name with an underscore in these use cases. +Both Facebook and force a renewal of tokens instead of using refresh tokens +uses GA and throws exception +I greatly appreciate your inputs and Ive incorporated both documentation and stats +Stefan your reasoning looks good to me +What happens with a runtime exception? What do you think SHOULD happen with a runtime exception +That method can be protected or private +We are using Jackrabbit for creating the content and reading the contentPlease find the attached screen shot and config files +I hope it is really final this time +The problem is on other place so I close this bug and create new one with right problem description and some project for simulation +These days I tend to keep things private or package visible per default until someone asks to make them protectedOleg +This looks right to me +Parser and Lexer now dont depend on and I threw in some too +Updated the patch with a minor change to fix a test failure +proto files are source files so for putting them under srcmain +Fixed in revision +Leaving open for Dave to double check the changed assembly code +You can always override is to provide your own +JacopoThanks for reimplementing itI am getting error of Java Heap SpaceI am running ofbiz from eclipseShould I need to specify something different in Eclipse for Java Heap Memory ProblemSee the attached screenshotRegardsAshishI dont know how to set it in Eclipse however these are the JVM parameters that you can use to set the memory settings What database are you using +Please check that the patch was applied as expected +I will have a look how to debug this +There also are code example repetitions as we show the same stuff on ORM and then on JPA +If I persist an object graph where some entity in the chain is transient nulls out the value then throws an exception since the attribute is set to +Looks just committed this +Committed code that fixes the type cast issueCommitted code that fixes the output format issue +But this was introduced with to prevent deadlocks +Tests passed for me with the latest patch +Closing this issue as more specific bugs were raised for other cases. +I wasnt saying that we definitely shouldnt use inotify just that using inotify is not the goal of the JIRA +Yup that looks right +unit test added +So Ive cut it back to the bare bones and modified the file accordingly +Screenshot is sexy +I didnt want want to sound petty but since you guys are using findbugs too I thought you want to follow these coding conventions +Committed revision Trunk Committed revision . +Ill file a infra ticket to get it resolvedJarcec +Reviewed the code looks good +When the next request was received by the previous request thread was still hung and the conversation was still locked +That should be avoidedThe above is not dependent on fixing the cglib issue on Linux is it +The current scripts in +A simple maven based example to illustrate the seems to be present in also. + +First attempt at updating file to improve processing of Heading Heading Heading Heading +Sometime down the line the need for will disappear. +Ive implemented a fix for this and committed it in new has been deployed which includes the fix +So it sounds like you think its the former + +Aah +Venu would you be interested inhave time to look at this +Needs testing +Looks like your pull request is in +This patch insists that the implementation returns not null +Just to make sure this is clear to all this is not a VFS function +This isnt applicable to Android we never had a plugin. +Any thoughts on the above anyone? is not checking for the content in the vector that is being produced from the but just the size +There is no need for you can construct a with any Comparator implementation so you just need to do that with a flavour +Verified upgrade in ER as has tagged with all patches incorporated +Alberto Please file a new JIRA for this issue +We will support something here eventually Im snowed can setup build artifacts eg + +First the EPR does not match my wsdl +Both Client Server side logs for the same client one +This is a difference in file systems with different operating systems +We could close it with a note that it does not affect +They are now defined in think this is bogus as it leads users to assume that they can deploy +Marking this closed as duplicate. +This feature is required for AS testing +Thanks a lot Kihwal +Hopefully well post it to in the coming weeks +This is probably fixed but we should confirm that +In this case I honestly think our is a good release even though this isnt enabled yet and we should just target +This patch is not syncd to the latest in the trunk +Services exposing with the property set to true are registered as components +safe mode would make cluster startup longer especially after a critical issue caused cluster shutdownI think it just make some admin operations unavailable during safe mode but not affect data read and write serviceIt will make SSH longer but its a small probability event where server died during master is initializing +At least a junit test to illustrate your problem would be very helpfullOf course a junit test a patch to fix it would even be better +Sorry but I think that doesnt solve my problem +I commtted the patch for now to branch and trunk +Errr dont we just need a schema definition and we can do this with any number of xml tools +Apparently there are two if statements that have the same condition +Thanks SeanPatch applied to ARCHIVE module at repo revision rI just needed to add an Apache block comment to the impl classPlease check that it was applied as you expected +I had missed merging necessary changes to component this patch adds that piece as well as resolves a good chunk of nocommit issues +Although the files were in the same packag they resided in different is not a to waste your no problemJuergen +I would be happy to incorporate any further thoughts or commentsCheersAshwin +With configuration I meant the JAR Plugins configuration in your POM +Moved to please move bugs back to which you will work on in the next weeks +I dont follow your bug report +I think this of import given we want to talk up use of lzo +Investigating again to see where were at with this +I have fixed append but from a quick look I think other existing commands need to be fixed too +That way anyone could define their own set of entities for escapingunescaping +I also dont think Ruby objects will properly serialize in a cluster +I added the missing piece of information +ugh +Ive got this confirmed in the build I do not have this confirmed yet for the upgrade found a workaround when searching for available updates must disable those for Eclipse Mylyn and all the SVN stuff +some tests dont make sense on mini cluster because they test scenarios for large perf +Will await Sean C +Thanks +Tests passed cleanly on trunk +on such a test which would prevent regression +It is part of old testskript I am not sure how to rewrite it correctly. +did ask for people to test the patch but I must admit I didnt and if we had this might have been picked up and then fixed straight away prior to being commited +Yes these still have to be replaced manually +It also needs changes to compile and run tests in on patch submission +with appears the issue is due to that your table does not have a primary key +Fixed in subversion repository as of r +snapshotbinstartvnc +For the particular error I think it is because the is not present in the build directory without ant package +The Shell interface provides a way to determine the completion shortcut that should be used for a given implementation +Reopened and corrected example in description +Different combinations of Index Hints checked for both Items table and Values table in FINDPROPERTIESBYPARENTIDANDPATTERNCQTEMPLATE query for Oracle DB +Attached patch makes the following changes Change to update statistics and tests for the new functionality Converted to junit test Removed unused methods in +This is interesting +Thanks for the updatekind regardsbob +It is not easy to start up the servlet container in the system test +indeed id been waiting to find a few minutes to add a test case +This is a update to which fixes a test diff I introduced when I committed +snapshot comment as before you need to get the snapshot plugin +Thanks Yu +Failing to do so will cause Derby to hang instead of throwing the No current connection exceptionI do not plan to backport to unless someone makes a request to do so +Maven still wins there +We will need to downgrade mysql to unsupported if this is not fixed +Yes optimistic locking may help in this regard +I tried doing some other tests could not replicate the issue +This issue is fixed +patch appliedThanks +Assigning to duh weinre wasnt connecting. +Asked for a testcase over a month ago +Applying the patch and running the test suite reveals another problem with schema handling in +Resolved in EAP + +thanks for checking. +FIFO is trivial and fixing up the MRU is also trivial +Right the actual fix does enable for subresources see their parents consumesproduces but you right that as a subresources do not see their own ConsumesProduces now +something we need to notice is that we will have a in the lib directory which should be avoid laterbut I can not find any good solution for now +By time series relative column keys are you talking about the row timestamp or are you storing the data as familynnnnnnnnn where nnnnnnnnn is the timestampIf the latter there is no good way to search for the last one because the part of the column key after the can be anything +Should we also credit Bruce Jay Nelson? +Im a little confused by the thread dump does not call +I just posted patch there which this ought to apply on top of +add the new field payment into entity +The majority of namespace users will be using the web module and the small minority who arent can explicitly omit it if they wish to omit the jar +works well +Remove the bad element until I figure out why it was there +Can anyone comfirm with latest trunk? See also. +We can split to smaller ones to work around this issue +Sounds like a free lunchBesides speed wed need to make sure Murmur gives us as good a hash distribution as Murmur so our bloom filter false positive rate doesnt go up see which runs with ant +So all issues not directly assigned to an AS release are being closed +Reporters of these issues please verify the fixes and help close these issues +Great job! It apparently improves task scheduling with better disk load balancing + validates the lastModified and content length as well as the assocated date and int header fields +In CR will match upstream spring as listed in +Thanks Myrna for addressing my comments +Ill check beta branch +We are talking over x slower for symbols +Against which branch did you do the changes? Please note that I did a quick commit removing some unused imports and fix a few on the +WSDL as generated by current as generated by the fix to as recommended in the bug description +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Never mind the Marshall part I did not see the get Marshalled either and it works fine +I agree with you Stack lets start thinking of how we change the heart beat +Good job Andrey it was not easy to reproduce +Thank you Xuefu +im no longer sure that itll be easy to create a simple unit test for itthe method could be improved by preforming each startsWith and endsWith just once and assigning to a boolean variable +This means that initdestroy will be called all of the time which I dont think is really the intention +We probably dont need to include new property files +ConfirmedIts now fixedAfter a few tweak here and there Studio is now working perfectly in both RCP and in Eclipse modesGreat job AlexThanks! +bug brought forward to this to minor jesse so far only you can tell what is intended by this bug report. +oh am busy right nowIll debug the problem next week +So this is more a remindertask to verify it +Commited +Installing in SMX works fine +sounds like the best idea +The conditional logic is complex but AFAIU you got it as clear as possibleOne minor thing In test and test are duplicates +This issue has been fixed and released as part of release +We measured some of a perf counters and found that with the patch we see less which is significant +What about tests and contrib Uwe? Shouldnt we just hit them alleditscratch that sorry missed your comment that you will do contribtests for the patches it also fixes whitespace where broken +Dart I think Tom is thinking about having three a three column approach each column being scrollable first column would be the Source second column would be where the mappings would appear and the third column would be the Target +Rejecting the issue wise quickstarts work when using the JDK. +We also tried topics with and they also seemed to recover. +I totally agree that we need to find a more scalable way to manage an increasing number of services +And now it seems the instructions are no longer visible this is the patch from hell +Agreed +new patch find comments in the dev list shortlyagainst r +This is definitely a bug in the juddi console triggered by a configuration file pulled in from Riftsaw +But that has changed now you can choose from several languages +Plus we are right around the corner from gt so should probably heir on the side of caution +not yet processing processing repl queuesToo many processings +This is mostly related to DT in History Server +In my haste to get the patch submitted for the RC I forgot to make sure the tests coding style was correct +request is legitimate +Thanks Tim the patch was applied as expected +Part of this issues fix will fix +Patch for revision on tried with BranchEAP revision with the same result +Disabled formattingScheduled for and assigned to Travis + with a readability nit it may be nice to explicitly state in a comment that the tests are needed to expose otherwise private methods for testing +Its an awfully useful working directory + +reproduced and fixed +Committed in ys branch as well +But we of course need to try our test there after the installation +benoitc Matt mentions that it might not work with IE but I havent checked +Revisit issues that have of date for all Tasks and Feature Requests created before +Ok agreed +SPRINGCLEANING we can reopen if necessary +Cleanup only include jars from the specified directory +Hi Sandy Thank you for your attention +Thanks Kevin +attaching +this is not required +If we would simply try to execute the call method regardless the type then it would work +I read your wiki and I tried +Configuration errorSending to ndaleyfinished FAILUREWhich makes me belive ethat it is a problem than the patch problem +Thanks Tim! +This idea was suggested when discussing +Then HTML and eventually iOS +Attaching +Unzip it and run java jar settings in the dir with spaces directory +Added and checked it gets activated without issue +in operation map has the similar caps issue + +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I modified the unit test and run successfully the as well as the Junit test on trunk +The patch has been appliedPlease verify +If there are no other places this fix should go then I am on this +Attached patch fixes build now that extends +Hey Harsh +To expand on Gwyns second point another potential issue could arise if you have a component with id field that has a child index plus a component with id at the same level as field due to the use of as a hierarchy separator in markup ids +Here is the output from running mvn jbiservicemixin the examplescamel older issues for Apache since were no longer actively working on these at the moment. +CR is just for bugfixes so if this is risky to do lets keep it as is IMO +union used to removeThanks! +The tool will not attempt to connect to a running server if offline is specified and it will not fall back to the offline behavior if offline is not specified +ll look into this issue +Since Hudson does not run tests can you run and make sure it passes +bq +Two things I still need I dont feel good about the default search logic +will commit this unless there are revision in HEADThanks kaveh +TedDoes this still apply since there has been a release +Merge to from trunk toCommitted revision . +Im stumped +HelloI am looking over the different issues regarding paging with large resultsets +No it wouldnt +On the other hand some admins might be ok runnign these two pieces two togetheris it possible to enhance this patch so that the server can be run separately from the NN +Actually we should just create a tmp directory under build and use that as our scratch directory +incorporated Chris failure is unrelated +If Im not mistaken this comes from one of the timers + +If added any incremental model creation mode that can enabled from the workspace preferences +Hi Ben no reason at all the version was the latest when I wrote the component but you can upgrade it of course. +Thanks +In the updated patch in view is created twice with disabled and enabled aggregation respectively +The issues are now closed. +If a client solely depends on this state from NN it wont get satisfactory resultsIn addition to providing more hints from NN as Suresh suggested I believe the client has to be smarter since there is a limit in the freshness of the health state that NN can guarantee at scale +I managed to implement this issue without using the Oracle Georaster Java API +I have marked this issue NEEDINFO since for it to be included in the Release Notes I need a definitive statement +short comments I dont see any possible synchronization issues with the patch? I really want to go further with the refactoring and want to add the once we arrive at a state that we all agree on +Committed to trunk +allows it cause the list only represents the actual live services so we can get that information +Going back to your very early question any thoughts on whether there are any reasons why it couldnt be in +Patch that fixes this issue +Patch for last the patch in this issue be closed +Not currently everything is a clobber although with I will put that in since we are not using Xcode templates anymore. +Feel free to play around with it if you think you can get better results. +Patch committed to trunk. +It just checks selected project created by Seam Web Project could you post project structure here or attach example project without libraries +Stuart the is set in the very first DUP +Tis patch supercedes the first one +Backport on log with Milamber you need to update KVM host jar files +commit log and data directories still get created although nothing ever gets put there +The code change is clear and +This does seem like the right fix +What happens if you try a different keystore type for example PKCS +Rewrite implemented in sandbox. +I think this is a false alarmI just found out the same thing +This is correct one to use +Im assuming that is a blocker for is marked as Duplicate of and resolved +If there are no objections Ill commit this shortly +Removed executable property from new file with revision . +I think it doesnt take a lot of time +Closing as is already released +Wasnt able to build anything that file recognized as a proper shared object +The problem I take it is that builds are hanging +Tika cant do extraction without a correct mimeType thatmatches the content of the input stream +OK I forgot to backport from master toThis is done now so can you try out master or the branch +Thanks for the patch Committed revision . +basically it was updating the hiberante mapping you dont see the changes by tomorrow morning let me tom. +xml +then i and i are merged and replaced by ii ion next access to the above mentioned it will realize that the reader that was used to get the document number is gone and as a consequence will be forced to get the document number again and adjust the reader reference to ihope that makes sense +emailing infra + requires about hour so it should be used as separate task. +dev forum link have committed a simple UT for an HA Singleton with a Pojo bean injected without a JMX annotation that TODO remove JMX from HA singleton election policy tests and investigate why tests are failing +Select option Overwrite existing logfile if you want to overwrite the logfile +I checked in but will check locally whats still broken and fix itRohiniThanks +The wildfly console provides the web based interface to tranactions. +Ive applied this with some minor changes +Benjamin Havent tested it myself so I trust Jasons observation that it doesnt work +This is not what we want but its what there is for now +Maybe I do misunderstand something or my configuration is wrong? But the test doesnt fail with jackrabbitMy test doesnt clears the database before copying the filesystem content to the database again +Thanks Musachy +I believe namespace and namespace are two more instances of test cases +Someone replied to my post today that mysql XA is working ok +its fixed already +Committed to trunk. +I am closing this as wont fix because the main suggestions in this issue have been around getting better DSL support for Groovy +In that case they can do that manually very easilyIf you want it for a generic overview that is fine but that could also be info within the specific processdefinition +Well I imagine its worth finding out why Hudson let through a patch that failed a test without reporting it +I have committed the trivial change to ignore the voldemort +Java is dead and new code should IMHO be free to required Java if only for the Override annotation which IMHO is a big plus in code stability by supporting the compiler with its static code analysis. +Thanks for the review everyoneChecked into trunk and +Duplicate of this issue has something to do with +The Reported probably has either the other process having access to the index or other. +bulk close of issues as part of making release notes. +Patch that completes the query timeout support for DB and MS SQL ServerNote This patch also contains the patch in but Ill reattach once that patch has been applied +Thank you for contributing on this work sounds this one is related to servlet new feature right? So I make a link to +Ive just committed this to trunk and. +If you register your jar Pig will wrap it up and take it along +sry Iam new here can i post my question or is that more a really bug report placeHow i can edit my post +If the resource manager is restarted while the job execution is in progress the job is getting hanged +Committed +With a new event is created each time the quality profile used by the project has been changed +Once weve got by the RC Ill test and commit +My question has been answered. +This looks really nice Markus +Sorry that my earlier patch didnt +Chris states +Either the current semantics for REST paths remain the same that is tablerow +Thanks Bryan! +In your testing if you come across any corner case which has missed my testing please let me know +The patch was created against Fuse MB tag version +So it seems this is the known issue and will be addressed in future releases right +if that is the case then do we still need this jira +Just ignoring the suspected recipients is not the concurrent case the originator only contacts the primary owner so if the primary owner leaves it needs to retry the RPC on the new primary transactional mode the prepare command only acquires locks on the primary owner so if the primary owner is suspected we dont have any locks and we must retry the prepare on the new primary owner +Missed new files sorry +This has been resolved by adding name attribute in +Note that the Teiid sequencer will probably need to cache the extension properties and node types it has seen before to prevent unnecessary of the node types +This issue has been fixed for EAP as noted above +Fixed project example descriptor +Event should be fired in both cases +Closed after release. +Each PING sends the list of all current client leases which is used to update the list of clients leases in +Looks good +About no idea I believe it can only support one authority or not? +After fix this issue junit test JBIDE has stopped to work + +Since you are changing the filter API it might take a few go arounds to get it right +In addition to removing DESTDIR it seems to have fixed the problem from what I see in my tests +Looks like already solve this issue using windows version of dladdr +Thanks Ravi +If the client with the lock is so burdened that it cant service a thread then it will miss its heartbeats to the server and the server will drop the client ultimately causing the session nodes to drop +could you tell me the exact javajdk version you are using? bug related to this cant delete issue +Ill submit a patch soon +The shareable node feature was implemented as part of +reopened just to set the resolution to fixed +you veryAndrei +I like short emails File New Other Tools Web Portlet Java Portlet Next Cant fill in Project Name so cant get past this New Other Tools Web Portlet JSF Seam Portlet Next Cant get past next so cant get past this in Error log +Lets discuss it more before proceeding with that remains to complete this issue is the resource optimisation +The plan is to replace Base with the codec version in + Username user +Please reopen if this is still reproducible. +Heres a patch that lets you configure the url scheme the same way you can configure socket timeout and connection timeout in distributed search +Im glad you noticed that field I was just about to comment on itIf you think a relationship is needed lets discuss it +Unfortunately it looks like it has to be done for every folder that contains a +The default one is hashed form so if you dont specify explicitly to use plain form the password will be hashed +Possibly the results are consistent for a while but it may repair itselfI just want to note that this is being done against a cassandra server +Maybe Bob forgot to set this issue as resolved but this is already with JBTIS +Tweak completed +Extending the scripting to include load and store is a logical next step +Yep Closing this as it is duplicate to. +I would start from there +But warnings are a sure sign that things can break in future +Patch to fix the javac warning that should also fix the tests issue +Thanks Rakesh +Thanks +If we do with forcetrue and the region is in assignment i think we will end up in double assignmentAnyway other than this +Sorry found the license on my Atlassian profile +added to the SOA release notes as resolved performance of has been improved +WSI Basic Profile analyzer errors also about this +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I scratched my head about this too +adding new profile to the pommvn package assemblyassembly Dwebtrue +You should hit the submit patch button to mark the JIRA patch available so that the automated Jenkins tests run on the patch you just attached +You are welcome mr Jacques cc AllHeres our Patch for Vietnamese translation please review for us and remind me if theres any mistakeThank you very muchps i also corrected few keys for s language key because of CRLF +Thanks. +At the very least it makes the results consistent between Java and Java Before the patch heres the results in Java jsFeatures rpcanalyticsflash skinsviewsminimessagesettitletabssetprefsIn Java jsFeatures flash ifpcanalyticscajarpcviewssetprefscoresettitleminimessageskinstabsAfter the patch jsFeatures analyticscajaflashifpcminimessagesetprefssettitleskinstabss the plan to fix this? andRight now weve got a lot of balls flying in the air we should get better coordinated +Ended up fixing etc that is not written for ubuntuDuring reboot a new set of kernel patches was added. +In a menu situation there is no need to pass parameters within the slink and it would be better to have actionnewSearch in the slink instead as the defines that navigation outcome will result in the search page being displayedIn terms of a general template for replacing slink it isnt possible to create something like that due to the various situations that slink can be used in a non portal Seam application +Also instead of using the command id it uses whoami and groups to get the current users name goups list +please verify that all works ok +Good catch V should address comments from Colin and Junping seems i am too lazy to do a grep before make a patchFor refactoring it doesnt repeat more than twice and the code section is pretty short to me it still ok for leaving it unchanged +skh pushing to +Verified as fixed in +Having said that ill change name of component once again +The meclipse plugin does not work as well as the which is why I said to use mvn eclipseclean eclipse just wanted to make sure there wasnt some other debugger or method you were using that I needed to know about +Anyway if this problem occured only because of the abstract class treatment why no exception was thrown to expose this issue +The main proposal here is incrementally updating FC values and how to continue to use for readers mixed with for RT readers either in TOVC or somewhere else +srcmainjavaorgapachewicketexamplescompref +I read through the discussion and verified that the property is indeed not used anywhere yetOn getting this implemented at this point a patch would be the only certain way to get this implemented +proposed patch includes fixes to +commited +lets add disableGrouping attribute to queue and attachQueue. +Although for the default task controller we can set execute permissions as done in pre +bin server startSqoop home directory homeusergitsqoopdisttarget +stack please note that the timestamp is pretty much what the earlier logs showed +Retest with the release or move the the wsee version in +Tested against java and python servers with php client and all behaved as expected +Bulk close for release +If repeats need support for this basedate to it will be done then. +Sure lets add the test with +properties file looks good in studioeclipsepluginssrcorgjbosstoolsusagereportingAccountAccountJBDS +What kind of tooltipsYou mean like showing TLD docs like Javadoc is shown as tooltips for java code +This patch makes things work but Im not sure what needs to be at the top of this document legally speaking +Stefan i just created a new project project at the top level of directory for your testsuite +If someone wants to create a patch or something in the future to support it we can as an enhancement +With Martins commit use is working again +Attaching a supplementary patch that adds this fileI am committing these patches so that future documentation can build on them +If you believe that there is still a bug here please reopen this bug report with any additional information you can provide. +min and max was the same +Applied the patch and added the URL parameter to the exception message +Did you have a patch to fix this issue +Now I attached a stripped version of the backup here and also put the full backup in my dropbox +Can you please check that you run ant clean testand that that xstreambuildhtml no Mauro +should be fixed in further feedback assuming fixed +That would make all the instanceof checkes unnecessary and would always have access to the yet I think we should like those issues as related though +fixed typo in patch is updated which also comtains old report directory removing +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Is this issue resolved +svn ciSending srcmainjavaorgapacheopenejbjeeSending srctestjavaorgapacheopenejbjeeAdding srctestresourcesTransmitting file dataCommitted revision . +HTTP Analyzer crashes freezes and I cant see the headers +You have the same issue where Hibernate tries to call JDBC methods on fully JDBC compliant drivers? I dont understand how a driver could not implement a JDBC defined method yet be called JDBC compliant +All TCK tests pass with this + contains a test case for this +What will this imply for the final user? No more due to retry no more read on write? Any disadvantage switching to those locked RMWNo more overcounts bc of internal replays +bq +and that my pull request will be hold off until thenThanks for the info +This issue is a duplicate of +Yes both are related +javascript implementation okLabel and cancelLabels in +Could be better and not output null but its better than what was there before +Ok I agree that commit should not wait for merges +Either there needs to be a test case that can reproduce it or you need to debug it in your environment I have never seen this +In it makes more sense to save the length instead of the +Sure whatever you think is the asked about using Netty and thats not an option simply because its not supported for a I think Id just go for creating a for now and get that working well +your CVS repo object above which is data only +With DECIMAL the results of divisions could have a fraction part +grep tcunning find +Jim thanks mirror will be soon HPP +I do not think addition of an attribute to a schema is needed at this time +Pull Request? core issue is fixed in beta +It is possible to reuse Exec and +Committed revision . +Thanks +Maybe using a lock as Steve suggested is a better approach +Should be fixed now. +Forgot that were normally in +Perhaps this was incorrectly +The same should also hold for the dfsclient +KarlWhat shall we do about this issue? Do you still intend to do some work on it? OlegI am currently in the midst of release challenges myself +The rest work well +Alex I pulled the branch and tried building it +Closing. +Can you try with Camel which has been released + +James please set correct affects version +Trivial fix Ill have it committed with your test soon +Fixed by Trustin +but its very much a prototypeI am able to run the accumulo continuous ingest test and the verify without a name node runningLots to think about like permissions balancing replication and even basic schema is open for reviewchangeThis is based primarily on Aarons prototype with ideas taken from Giraffa for cleanly wedging a proxy in for the +we should probably track cores as we do searchers +Thanks Patrick I will take a look at it +The only constraint is that a split will not span across multiple files which is true now +minor changes +Thanks Ahmed +This case didnt work for me +Id propose that we fork a small process off that runs the script for us every N minutes and reports status to the TT +Thanks for commenting Jeff +Ill run some additional tests and plan to commit this patch soon +ACS feature freeze date is November th +AFAIK this is little tricky to do in UI by keeping the OLD way still available +I think this bug says that if the NN is configured with multiple and there is a bad fsimageedits in one of the configured directories while the fsimageedits in the other directories are still the NN fails to load the imageThis feature is somewhat by design +I think I fixed it +Patch for the sameThanksAshish Vijaywargiya +Thanks UmaYes your right +Attaching DB on data dump it is still listTemplate api issue reported in +If a transaction is supposed to be rolled back the timeout is set to seconds onlyI ran without failures on my box +committed to trunk +My apologies for false alarm on codeline +And well also want to have a VM to run Allura the plan is to tickets and some other project tools +Agree +The standard build will not use bnd +use Javas existing mechanisms to solve this problem rather than stuffObviously the benefit would only be seeon on JDK or later but I dont think thats a problem +Do you want to do the data conversion yourself +committed rev cbfdfffaaaffcddce +This issue has been fixed and released as part of release +By the way the version is the one given with struts beta which is according to the manifest file +Link this issue to to keep track of required changes +I was following the principle of least change +The filesdirs that should be removed from the existing lib directory are I am in the process of benchmarking this on a big cluster +In the case when only topics are specified we do make it mandatory +Thanks for the hint +Nick your latest test case is that a ZIP file again +Ignore my last comment +The easiest to deal with are minResolution maxResolution and num +Verified in JBDS L +Verified in ER. +To me MUST only appear if translates to must not be present unless but maybe it means can be there but if the record field is set to then it must be presentAnyway unzip simply ignores excess data we can do so as well +I like the functionality but I would like to have it hosted in a new class something likeclass +Could you please add author tags where possible +Can this ticket be closed or is there still more work to do? I wasnt sure if this patch encompassed everything that needs to be done for this ticket based on Kristens comment +The quickstart script for kmeans algorithm +we might also need to add this function to karaf shell +Sorry youre right disable them during the job +Im really stuck on this +This seems more or less a duplicate of and is IMHO a very ugly bug not an improvement +Also not sure about the correctness of current in COMPACT STORAGE case +any plans to apply this patch? +Hmm +This link brings you to the product page not to the page that contains information about the item that was already added to the cart +We can consider this done as the tests are passing properlyThe next step is to look into +Thanks +So theres an easy workaround right there +Attached patch removes the behavior of rming the old storage directory upon restoring it +branch Committed revision trunk Committed revision . +Thanks Konstantin. +I cannot currently run a full ant test on Windows due to a bug in antjunit on windows yet tested on windows due to +addressed comments +When the socket server accepts a connection it registers the READ interest for that channel +Just a pom file with all the projects as modules +Yours sincerelyCharles Lee +Thanks Doug do you mind providing it as a patch +Itd be great if we adopted a similar technique +In orderDo Do +Then revert r to close +Committed Thanks +FixingI hope to fix them all tomorrow and commit the changes +So I think its a problem of jsp editor in eclipse +Please use the user forum to ask for help +The black line hint will tell you where the layer is supposed to get inserted +Would you mind to post a working link or explain what you mean with this issue? Or a reference to the commits you did +There is no need to cherry pick what errors to log when you are throwing it back +Could you post me that logthanksSubbu +Revamped the page fingers crossed it makes the next build if not then itll be in the one after that. +listening on all interfaces specifyg! telnetd ip start +For some interop senarios it wont work if Sandesha sends an sequence offer +As per our discussion earlier today well have some wizards for the RC in the form of scripts + +For we should focus on running on maclinux systems that developers can easily test with +While I was in there I fixed a couple of things and it was good to take a nd look +On iOs the file grows without a restartHow would you suggest to solve this My app is in the stores using the bytes database which was fine for the time +Patch relative to mathtrunk +mrepositoryorgapachemavenreportinghomesonar +Checked into SVN as revision . +Thanks Ashish Nagar for your contributionYour patch is now in trunk at revision no +Time fields added to patterns +Fixed problem introduced by +So maybe we should stick this to only +I have a few successful builds coming through +Postponed +It wont be a pain I think +Here it is useful for the client id to have the hostname of the fetcher thread +What other options are there? It seems a bit alien for +bq +This feature request is not valid anymore +Updated the procedure templates +In this scenario I see angular as a much better fitIf we go the route of using angular I will end up backing out backbone and handlebars and pass control of all state management client side templating over to angularIn terms of branching I can go either way +more information is me plz the link to spec where the attributes should be attributes must be deleted only by dev team from tables of attributes +Since this is a somewhat large patch Id wait a couple of days to check if people have questionsa the new bloomfilter is any fasterslower than the old oneb they have some tests comparing rates of this vs old one +Jetty support should be done although Im not sure how to test +Test base on revision in trunk +Let me update the patchPersonally I like adding a keyword better than adding a property +next iteration fixing some javadoc issues and adding a changes entry + +Hi AndreasBecause of security reasons i cant upload the documentError Could not parse predefined CMAP file for problem was resolved in the version but Error Could not parse predefined CMAP file for error was present theirPlease check this errorThanks +Could you please upmerge the patch is out of date +I looked briefly at the comments and what you are doing and I feel removing Searchable in is a heavy change and will most likely break back compat +therefore we have to set xhtml as default +Thats right +Using a fresh Tomcat instance on Ubuntu x I started up a fresh Railo WAR applied the patchupdate restarted Tomcat and then used the Restart Railo feature in the Railo Server Admin +Wanted to close on thisthankssudha +Thanks Jeff for the patch +It should be marked as warning if we cant found a class type for such components. +Thanks a lot for the review Harsh +defer all X issues to +Verified on JBDS +The diff and formatted page look fine to me +Thanks a lot + +verified on mac +Looks was applied as I expected +I always forget that about the +We will only be able to aply this patch once testingtomcat is ported back to the trunck probably in a refactored way as being discussed on the +I agree and thank you for the observation +Im gonna create a separate context class to hold the data which need discarded +Some other issues +I recommend moving this to the backlog and focusing on key issues such as where would we be without standards that manufacturers dont adher to? Im just worried that as is slipping to the RC and given that it is complicated it will not get the field testing it needs +The change from Jaikiran is looking good for the other CL leak +Thus each daemon should have different HADOOPHEAPSIZE configured for different memory limits +closed +integrated. +Marcos in general current gateway support does not cover all possible cases so I opened new issue for it to improve thatAs it comes to your case you might get around the current limitations by including the XOR gateway inside the loop no to create the OR flow again in case the XOR needs to be reevaluated +I was having the same error but when upgrade to cordova it was resolved +Here are the files from my portal see the screenshot of my tree +you are creating dependencies for the validation workflow but we do not guarantee the order in which components are validated +soon +You can produce the same error message by uninstalling any system module that another system module depends on +The attached patch contains the fix and unit test case +It should be above +This patch also slightly refactors the code +Hi Laura +Regarding we wont release any anymore or very very unlikely +If all agree I would make rife optional for for +Fixed patch so that it includes all files +Proposed solution attachedThe solution allows the Camel JMX enabledisable capability to override the new configuration parameter and uses one instance of the Jetty to prevent Jetty name collisions when running multiple endpoints with JMX enabled in JettyNote that the Karaf test changes were needed for Without depends version specified mEclipse does not want to import the project +Ill link the corresponding jira. +Thanks Bill! +Just a thought to verify if it is not easy please submit the patch as it is +thanks we can take a look +Im pretty sure you just have some classpath issues there +So for other peoples benefit see that issue for the answer and subsequent discussion. +same with mvn install Dbootstrapfalse which I ran withmvn ceclipse +this has been committed +Whereas on an infrastructure CPU is associated with the number of cores reserved for the virtual instance +It would be interesting to run that and a few other queries that we expect would benefit from this to measure the performance improvements +Mass closing all issues that have been in resolved state for months or more without any feedback or update +There are a couple of additions to Localizable +FYI you can update old reviews then I can just look at the bits that changed +Here are some reports from that a fix with magic number to verify it on modify changes now it work without errors on my machine so I mark this issue as fixed +Committed the two patches as svn resolving as I didnt find other cases +I have not had a chance to give this a test +Also isnt currently supported +Sure Vinod do you suggest I create subtasks for each of the items +I am working with the Red Hat Portal team to get screencasts into the system much has been done on emails so will continue to work there on this +Hi Jacquesyes if you have time at disposal its great if you can look at this issueAt the moment only the patch remain to be reviewcommitThanks a lotMarco +So theres plenty of chances for interested excsbq +Version Released +Comment race +We need more info like browser error message javascript thats failing etc +Often its pom file or similar issues +v adds better debug logging around creation which governs that +I set JAVAOPTS ran and uploaded the output +Can you also write a specific test which launches a container that very quickly exits turns around and launches another container with same ID and token and gets rejectedbq +The other people speaking on the issue are talking about correcting the fundamental issue +Upgraded clean and checkstyle plugins to latest versions in rev +Forgot to include thisThese code contributions are licensed to the Cargo project in perpetuity and the code may be licensed redistributed or otherwise used at the Cargo groups discretion +the modules in trunk is functionally related grouping + +I will try to post an alternative to this jira too +Yes +to me means a few things +Previous patch used JDK Override support +Thanks for pointing this Ted +In Process object classloader is stored its set as during deployment +available in phabricator and jira +Fixed the parameter order of holder parameter the INOUT parameter should replace the IN parameter instead of remove and add. +However we will ask themAn other solution is that we implement it in eclipse and support it +In the basescanner would try and do this fix up everytime it ran so I suppose itd fail and then come along later and probably succeedSo weve just read +Unit test that repoduces issue +Hi AdrianFix was trivial so there was no point in taking lots of time +I havent entered a bug yet into the webtools have checked their HEAD +Setting to resolved as subject of JIRA cannot create Rule Flows is resolvedPlease raise any new issues under a different JIRA. +I will create an additional test case to repro the nonce issue. +Id be in favor of that +Thats supposed to break the cycle +Good points Uwe +Im not sure when or which one resolved this issue but the testcase now passes successfully and is showing green when I run the WC tests and generate the Component Interchange test report +The problem is that our customers demand high uptime +Fixed by rev to during tidy up prior to the issue now that it has been resolved. +New rev of the spec. +This patch makes the exception handler catch Exception which includes so that they will be handled correctly +Closed for. +and are not defined in Haddop will make adjustments to the shim +Logback is under EPL +There seems to be an workaround in place +I will apply this to EAP EAP x trunk +These are public methods so we do need to document what happens when unexpected objects are passed to them +Recently Rysiek upgraded magnolia so all static files has expiration to half hour by default +Its acting as if other rules are activated immediately as each action is taken in the RHS rather than considering the final result of all the actions +Attaching and +So Ill do it and let youothers know +If you dont have time thats fine I will wait for your revised example and try it myself +Cool I am open for queries and requests about the API and can help where applicable +We should document for the EWP release that this example would not work and remove it from seam examples for should be documented in release notes +What version of Eclipse and groovy compiler are you using +It is specific to what it is describing +Nobody except original author cant change any atribute +I have abstract base classes being inherited +It would be nice if included such itself as pointed out +Updated source code for the benchmarks +We can use the logic in our s Superstep class where you can halt the computation by overriding a functionAnyways we could translate this to the superstep bsp right? So once we have working we can test it with the graph algorithms +All other updateschanges to the View for should be written up as new BugsFeatures +JIRA kinda sucks but oh well +Ouch +it seems this has been not reproduce issue now +To reproduce the problem use attached testcase and follow instructions in revision to fuse for that revision +This issue is correctly categorized as an improvement +And isnt the SDO api jar part of the geotools oracle jar? Or are we not using the jar and are back to using Oracles SDO jarI dont think we can include the jdbc jar with it I think the user has to download it on their own and agree to the user agreement on oracles page +RightThat sounds reasonable to me in theory +removed restlet dependency in revision +to the patch +The previous patches supported +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian applied correctly +to trunk +The checksums under orgmodeshape in the Maven repository were rebuilt this morning and it looks like it just finished +Use underscores or any other separator instead to avoid conflicts with the property notation +In the event of error we would currently exit the NN edit log roller thread + +Yes this sounds similar to the upgrade problem we came across in + +rev deployed the attached jar file the beans were detected show up under Stateful Session EJB +Committed with subversion revision +However it is licensed +included missing filesadded new Integration Testchanged inheritance strategyfixed inheritance of attributes in not correctly set +looks good +Here is a hack patchI will add optimization tests and factory +I get Blocking updates messages but no flushing happens so everything gos on hold and no new transaction happen all my threads get blocked +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +fixes this issue +Updated test case original was using the old servicemix kernel package for how the test case initializes its login context and using a separate callback handler for Karafs to call onto appears to make this issue go away +If you find further issues please create a new jira and link it to this one. +SandroIm assuming the patch will end up in the trunk sooner or later +New patch move back under FST +If you launch lots of these in a tight loop it is likely that you are still aggregating the logs from the first few while launching new ones very quickly +Konstantins PR is pushed into master +Reopening as the pluto container implementation actually requires that at runtime the descriptor model is portlet based +Rather than upping OKRELEASEAUDITWARNINGS I think its better to add this file to the list of exclusions in in the configuration sectionGood point Todd +Good luck +Fixed by r on the trunk and r on the branch +Patch doesnt apply cleanly anymore +Patch applied +Jimmy youre currently the assignee for this issue +Thanks +I remember seeing some logic around inserting a +If a message needs to be delayed in the future first it goes into the queue and bounces there times +Great job guys! + +Reviewed the patch +Yes the constructor must be public +verified on +Patch committed in wanted to say thank you for awesomely quick work on this +I will give it a couple of days before I commit +Hmm I dont know +Resolving back +I will write a small test case to see whether the +It turns out we had a bug in that wasnt setting the native method table correctly when you go down the read path +For example we could use the docbook or wiki help for an on the fly web manual +Closing it. +I will have a look on it and provide the wiki page today in the evening +no allen its just a configuration parameter thats hard coded +People just have to select Refresh on the project after importing a project into Tools and it works fine +Priority in bugzilla was normal but there is no such priority in Jira so Im changing it to Major +ok cool +Yep just grep skip inBy the way why only of all calls in this module is fully qualified? It is artifact of? +When we add support for storagevolumessnapshots to the vsphere driver we can revisit this fix +Im not sure how that is picked up on your end +In the case of popen ruby e sleep we may not +Thanks! +putting other fixes for other than but in uimaj project here too +Fixed with the referenced change +Sebastian +Thanks! Patch committed in revision . +Yes somehow that didnt make it in the posted patch but its in the code running on my cluster +But from a user point of view your approach surely makes sense to me +This is the bug only in CQLIt was always setting STCS regardless of the valueSimple patch attached to fix +Looking at that the apr were from build of +Galder assigning to you so you can and accept the contrib from. +Thanks Konstantin +Vertex and could be a common class +Sorry I meant even if the input row type is decimal two rows can in theory but unlikely +There are continously people running into problems trying to use another axis version than the one we distribute with wsee +but using set makes the intent clear that we cannot have duplicates which are possible in list +This is a good time to address cos otherwise we need to carry using regular expressions to filter things out of a stack which is ugly and could break very easily +Proposed variant as explained in previous comment +I tried and deploying multiple WAR files still doesnt work after trying what seem like the obvious things +The build is still failing +srcjavaorgapachecommonslangenum +Does this also apply to +into trunk and +Thanks Alexei for a succinct bug report. +I or someone else will comment on this issue when the patch is available for +bq +mistake +Will likely just reassess as an xml specific enhancement based upon the existing logic. +Merged patch with latest trunk +fix pushed should work now +Thanks UmaIf I get no other comments in the next day or two Ill commit it since I have a already +Filed. +Thought there would be no harm asking +Could you let me know whether the changes I originally logged are now in SVN trunkMatt +Well looks as though Grails doesnt like or the jre I had +I will commit the patch shortly +groovy extension +This is a Mylyn messageIt cant be changed without changing Mylyn code +Weird +Looks like this was fixed as a side effect of fixing. +The issue with existing applications and usability undesirable to fix this. +If you set the version to either or in the POM you will avoid the problem under +So how is it different from setting these properties in the? that seems like a good addition +There is a thread in Eclipse that listens to the s output and handles messages which it prints +Yeah in fact you way is to still copy the JAR and override the file in the system then from the shell do a update xxx where xxx is the bundle id of the JAR and do that for both of then restart the ESB +Some ideas Can you browse easily to and with a browser? Do you have a proxy enabled in Eclipse that would alter the connection? Do you have a proxy on your machine that could alter the connection? A firewall maybeAs a comparison it would be interesting to see how Tools behaves on the same machine and compare its performances with JBDS +No comments yet +Note that Ive patched the pom in my local repository to use and I had no problems so far +I will try it locally and commit in a day or two +In fact when things work fine I see no reference to eth at all +Bulk close for release +Users might add javacompenv or not to keys within Add toString for for make the debug easier +the cli will still use the yarn acls +We havent seen this occur again in a while resolving +Thanks. +Requirements posted somewhere and Ill close this issue +Note that the issue is and occurs in cca runs +Perhaps more of the test code can be sharedCant we use s helper methods +Thanks Steve! +Hello Lukas I tried last week and I copied all the libs in to my project +In this case we have no choice. +and moving them to +A simple password mechanism just protects Bob from from editing Saras session +I think this needs to be in +Does it mean that this vulnerability is present only in version? +mylyncontextI dont understand the problem with Mouse In the moment this combination is used to open the definitions of elements like +myfaces has been moved to thirdparty with the commons jars factored out +When the directory of a renaming place did not exist it modified so that a directory will be created +Wow that was a tough oneThanks for trying so hard to reproduce it Mike +I have committed to +As I am not an employee of Sun or Oracle I dont have the authority to contribute Sun copyright code to the Apache project +A simple version of this test could probably be added to the existing ipc project perhaps bound to the phase rather than the test phase +Martin can you move the third item of this issue to a separate one pretty please? To answer your question about whether theres a test for this I dont know +Logged In YES useridAlexeyI know we had a talk about this but dontrecall the result +while trying to respect the cassandra code style I think I accidentally fixed some other tabs converting them to whitespaces +Also take a look at what can be closed here please! +calls Mojarras with render option set to component +Note that it may be a questionof preference and we would have to discuss this on the PM list +Your solution sounds reasonable do you already have a patch for itIm planning to cut the release early next week +Notably takes minutes to fail +We created executable using library which is with ICU option +Can someone help review and provide comments over the patch +Please set it again when the patch is created +I wont have any time in near future for this +I will put in under instead +Bounce job trackerSubmit a sleep job with mappers +Hi Claudio! Could you please point to these classes so I can give them a look? Thanks in advance +Feel free to use it modify and compare against other took leonardos solution did some modifications and profiled it +log file so we want to show the + +Committed +We can move some logic to static functions to make it more testable +Final version of the patch +stack I dont think it didIll take a look soon hopefully this weekend as Im not completely sure +Building on top of component scanning Spring Java Config and flywaydb organizing into feature modules is a better solution and more practical solution for integration testing you can easily start up subsets of the whole application just run the tests for those parts of the system +Any update here +There was a mix up of the Solr versions on my machine which caused this errorThis tool is great +They shouldnt do +It cant have nested doesnt mind other files are there were two problems Error inflating file when jar xf You can name the bundle whatever you want but not as a contained file because it will get overriden by the contained file when extracting +how about returning a http Accepted just before the shutdownbtw its helpfull to have it i just recently had to use it to work around +That looks like a standard effect +I hope Ill have enought time to work on this problem +So it is fully backwards compatible +I dont think this damns the get patch but we may need a new get method and go to see which one should be called +I finally get HQL query result in JBDS buildId to fill Fix for version is this resolved if the problem still exists +The older issue is marked as being fixed for and commited to Branch on the rd got a clean copy of Branch from CVS last night and the bug is still present +Well I was thinking that in general tools would be secured but I think in general there may be a use case for making particular components of an extension secured and others not +Now they are always added as new resulting into +Will this produce a conflict when building the drlvm bundle from all the individual bundlesThat was the main reason for me to move mainpage from to a separate file +We have made two important changes in the recent past to get rid of FlexBison dependency and to make it compile natively on Windows using Visual StudioIve not studied this patch deep +Also sometimes changing these classes leads to schema gen build failures in I had them yesterday! This looks great +Fixed in Git ID fceadceddafebafabedeb +Any idea of the perf difference? Does the PE thingy help here +We should consider this for problem here as I see it is that the forms action attribute allows you to use the extension attached to the action name + +I was actually looking at this recently and considering proposing it + of a bulk update all resolutions changed to done please review history to original resolution type +Thanks Ahmed +Not having a default will break standard stress runs unless people specify the RF flag right +Im curious why that hasnt happened substantially there but there is this post on this issue today +as they are merged? Would we need a for that which initializes the CLI per the segment versionIs it even possible to move payload data to DV during segment merging +would you please try it +Deleted the datainstance directories and the lock started Fuse ESB again +This file is saved after a successful job completion because of +What if there are multiple SOAP ports for instance? Favor HTTP over any other transport or just pick the first one we understand +I am not clear about what is really needed here +Maybe this isnt worth the bother? How about we split the job up into hr pieces and defer this task until we are in a position where we are using too much CI time +Backporting is probably premature +Patch is probably too broadhandle should just handle and decode it +I suspect this is related to the is also a failure blocking this retryyum install y tmp line retryyum command not found service start unrecognized like yarn is using retryyum which hasnt yet been ported to trunk +The methods I am concerned about return a few MB of data either as baseBinary or as attachments +If this is still a problem for you please reopen this bug report +if user used such uri xmlnslurnjsptagdirtagslayout he will be able define and etc possible templates in web project +Apache Directory Studio has now been releasedLets clean Jira and close this issue. +Same issue with M +However Ill look into ways to write few test cases for thisthanks +bq +With no pageSize specified the count will be correct +Because it may spend a lot of time +OK has been merged with the new users listThese lists have been closed As per talking with dkulp the Muse lists have been left alone for now +fix committed should be in next build later today. +after I have added exclusion to all hibernate deps I still get old hibernate and jta jars +Also could you add some basic unit tests for the go client +The struts package was regenerated to pull in the jar file +Attached is my patch +Problem doesnt happen with using no key generator +Otherwise we wd have to declare this a limitation of the Java backend and the progg wd have to rewrite +The problem was because my spring context was also loading and this file could not be loaded alongside the jetty transport +Not sure why it didnt show up on the but please take a look at I think this commit added a findbugs warning +Well get it fixed for the next release +the setting of HADOOPOPTS would have been ignored prior to this patch +characters are needed for the prefixThe prefix can be in a separate field. +If there is no object I will commit the change +Just a guess but would this do it +Duplicate of +Boom +Can I add this to the POI block? Or to the scratchpad block? IMHO POI block is abetter choice since there is already the POI jar and the excel stuff +Currently only the leader proposes since broadcast messages are requests transformed into idempotent transactions + +Patch applied at +Sometimes you only want a setter +Why is credentials fetching secrets from the system env? +Code looks good +Hi BobbyHeres a patch to do that +If such scenario exists it does needs to be consideredAnother choice is that we can provide all the access to the system user by default by instead of by +by making just a dummy extension of an that doesnt validate NamePathUsage of within core and jcrspi got replaced by a that properly covers the extensions made toName and Path within the scope of jsr . +Hmmm +javadoc False positive +I can do the commit myself now but I will wait for your feedback +to trunk and. +Just to confirm you are seeing the same error with ER yes? If so can you update the JIRA title to reflect that? Thx +All the appclient failures are resolved now thanks. +Fixed within. +instance of page rotation now has the correct position +It will fix the git problem +AFAIK hive auth model is not turned on at most places so we can get away by providing helper script instead of adding checks in code +The VM went into stopped state earlier since we didnt have enough resources to deploy the VMWhat we could do is to make the test case better handle this scenario and give an appropriate error message +Thanks Paul for providing the test suite +This is due to the spaces in the file path Documents and Settings +Here is the v I have so far +It just felt wrong to only have certain types of failure abort the batch +Tested on +Francisant clean jar test on parent succeeds for me on trunk now +The higher the number of racks the higher the latency for the write pipeline +All further concerns related to this issue will be addressed in Remoting in conjunction with +The errors reported in the log are fixed and as I stated above they could not have been a cause for the crash. +What do you say? the patch for version +Thanks Chris +Here is an updated patch to use the same technique in both overloads of checkDir with common logic refactored to separate methods +Just to let people know that I started work on this a while back and hope to get a patch together in the +Here is one use case where setting app retries alone might not be enoughConsider jobs being run via Oozie +This bug was introduced in when row level purging of btree rows during post commit was added +Apply the patch to fix the issue +looks ran the tests locally and they passedI just committed this. +What difference does it make as long as we set MOVING before streamRanges requestRanges +that is why there is still Activating Subsystem message while starting server +A patch for the fix +Patches for x and I just committed the fix +Those contain HBM files as well as the Java domain classes +rebase against trunk instead of just to make sure test failure is unrelated +Please carry on +Patch is trivial +We need to change this if the want to use the native implementation +Hello Would you please try my patch? To appy the patch please +Thanks. +Bulk close for Solr +I dont know if its directly related to this issue but I found the same error with people searching for foo AND AND bar +As per Adams suggestions this patch removes lastValue support for +fixing affects version is ok its and that have the problem +I agree with everything except the separate test war +Ive upgraded and tested things out and the basics seem fine +new icon is designed for open seam componentand hot keys are available for its opening +Since I can observer the rewiring occurring correctly with slashes Im going to assume that you are seeing the tomcat issue +Attachments test AAR debug trace of a requestresponse debug trace of an request implies that the problem may be specific to databinding +User forum has to be used for usage not usage request usage. +Thats whatdecode doesThat means you would have to decode the first byte of the vint twicebq +Raised an issue with second one is that communiction is required between client and server and the server and the client +This illustrates a concern I have with making the iterator unmodifiable there is no efficient way to remove a field while iterating without also running into a CCE +Verified in EAP +Hi Stevenok thanks for your feedbackIt would be nice if you could post a patch im blocked a little bit with this bugCheers by. +Retargeting to and +gitattributes to prevent git from converting unix scriptsalso added equivalent batch +After parsing the value both and are unescaped and returned +Committed to +Hi just did a quick test on ever thing looks feel free to try out the test case that I used +Reopening issue so that we add some broker side protection against clients the behave like the client attached to this issue +This patch simply removes the code that creates this directory and looks for it +To be clear contact categories are currently NOT implemented in Cordova for iOS and this issue is a request to implement themI have not checked the Android implementation to see if this is implemented for that platform +As far as i know yes its correct +Just run gant +Is that OK +I dont believe we push that back to bindings +Im not allowed to do thatDid you mean that the error exists? Or did you propose a workaroundThanks anyway +Coverage is better now all failing tests run +Can you describe the usecase with symlink in language of JCR operations +I should say the messages dont appear immediately +Issue is fixed +Fixed in CVS please test +In that page I noticed This page allows you to manage the attachments for a particular issue +You said you are opening files to scan it +the test case again. +I have had a look at the Web JIRA and did a search for any resloved or closed issues for but it did not return any results so I am still in need of the relevant information for pertaining to this Web release +Impressive +I intend to commit the image files and the HTML filesto subversion +probably depends on whether we want to support both SOLR and SOLR +How about if there are hundreds of columns +The failed transition basically does nothing right now +Yes of course it has to interpolate between overviews its very unlikely that you can make a request that matches exactly the resolution level of an bicubic is being used try setting a nearest neighbour the results should be quite should also check from what is starting from that is the overview inside the tiff +However at this point the module still was compatible with version of viewable because of that it still uses the Interface with the old location +Committed to trunk. +From the war spec it should be in classes + +Hi chrisI tried to include the new jar file in my classpath when i compile the program but i get the same errors with jdbc you tell me y this is happening even though i included the jdbc jar in my classpathThanks +the types that define the message parts +I am curious though why it is necessary +The session form object will indeed be removed for every request but put into the session again by showForm +We expected that the +Yes please provide single patch file its much easier to handle +Fixed so that message property is now optional. +Maven POMS are out of the jenkins issues have been cleaned up + +When message level policies are attached the client fails but does not segfault on Linux +Hi Danie it seems that testperfpigmixlib is missing in trunk can you check it in +Revised and merged two earlier patches +Nice idea and great work! I think I will provide some patches too +is powerfulI havent seen concrete numbers for +works project the issue happens is much more Friday I debugged the startup of the application and the converters seemed to be registered properly in Monday Ill look into the other end into the +Thanks! Issue can be closed. +What does EJB embedded is no longer supported +Shouldnt there be a hole in +So I think this combination is very uncommon +Thanks Brett Ill try your patch on our Archiva installation +Fixed somewhere somewhen +I will test updating to this version +I dug into all of the files and removed the lines causing the warnings +Kathey traditionally Java hasnt offered much in the way of inspecting file attributes +Could you rebase please +Johan please also raise an issue in the glassfish issue tracker and link here +to permit a Avro string to be represented as null or something crazy like that +Bernd please review the extension checkin and resolve this +I expect that the existing session will be invalidated only when get returns non null principal +I agree that the protocol wrappers are a good pattern but Im not convinced theyre required before this could be committed +done as a rename and also renamed the to fixed the Jira but didnt update the projects yet. +yes please sync +Patches for SCR Component JB +Thanks for the explanation +So far this is only the basic data format that is not integrated into Java DSL or Spring XML +If the category has no image it will recursively search up its hierarchy for a category that has an image stopping at the root. +I think everyone will be happy with this version +really a dupe +I think that by design get is supposed to be called only once +im most for handling the definition just as class on the tag to follow our normal syntaxordering I guess is all ok if creates are done first and drops last +The one related to snapshot created volume is already addressed as part of +Comitted by zwoop. +It seems to me that Browser implementation draws a smart image of the checkbox over the standard image but misses a couple of do you see that thing ever when Label is empty or some changes to window size or scrolling browser or settingremoving label text can make it disappear +The issue can be closed here +Thanks Stepan patch applied as expected +This is the first go at the patch +In case of I think the host domain and possibly user namecomparison should be case insensitive enough +New test passes locally for all +If anyone can tests this further that would be great +war that can be installed with or without Solr in the same container +This is caused by validation framework using the classloader of the Message which will not include our classloader +Give a try and well see if we can reproduce on our end +You add some code and then is visiting a code part in its visitor for the first time and will cause for example a meta class to be loaded +Marco in we prefer to close task when done thanks. +still fails for me patch necessary but sufficent +If any of these need fixing urgently feel free to keep bugging me until theyre fixed +Committed to trunk. +I dont have a vmware environment handy can you tell us the mount pattern for devices on vmware? PrasannaI think you can use devcdrom as mount point for iso device in case of +The issue reports the error in relation to durable subscribers which requires that the client reconnect with same client id and subscribe to the same subscription id which is what my comment was referencing +Uploading MR component of on behalf of Zhijie +Added formatting +Perhaps reword the info as well to make it clearer +Oh one little thing please add a method comment to add +Please look Attached and +It is obvious the change should be applied only to the with a scope attribute. +Fixed the hardcoded keysStill need to use deprecated config key for other option will be to define the new way fo defining the queue configuration +Looks fine to me +but doesnt say anything about the activities and doesnt say howAnd we think we hope dont transmit much confidence to the readerSo in my opinion the following paragraph would be betterWelcome to Derby! This guide highlights some of the more important features of Derby through a series of activities designed to demonstrate the use of Derby in embedded and configurations and to help you get up and running with Derby as quickly as possible +Hey LewisThanks for noticing +Ill submit the patch to only contain versions for plugins that dont have any and also are not specified in the parent pom +Id be interested in your comments and observations on the dev list about how we can use and improve these tests +To get npm installed you also need to install npm version is the one that it picked up which is fineNote I installed via Synaptic Package Manager rather than the cmd line but assume cmd line will work as well sudo install npm +Fix to srcwebappsjobto avoid NPE when invalid queue name such as null or queue name is specified +The manager would be harder to deprecate and remove since there isnta simple migration path +This is automatically handled by Forge +JaquesI think you are right the diferences are light between terms used in the different arabic languages and it is better to make a generalization before to specialize on a locale +I just checked it and was not able to reproduce the problem. +committed to CVS HEAD and +You can report sensitive issues to aleksey although Id note that Cassandra is not designed to be exposed directly over the internet so vulnerabilities in that respect are not really interesting +I cant reproduce any problem with having asm libs on the project class you attach the output of the Spring AOP Event Trace View for a build you observe the error? is the aop trace view +Taking suggestion from Jorn a one liner change wasrem Set default JVM optsset UIMAJVMOPTSUIMAJVMOPTS fileUIMAHOMEasconfig This seems to fix the problem +This seems to be a bit different kind of limitation +just uploaded the mac dev jars +The code base had diverged quite a bit +Im going to check this in +we will need to update our wiki documentation to reflect the changes in this patch +Reopening the issue until a new System VM Template for is specifically generated to be used on master branch builds +Thanks for reporting it. + +It should not affect the original cases for the code in which are intended to unwrap the jarfile URL to a file URLUnfortunately this will not fix the issues with not liking running from a classpath URL but I will go forward with this change to ease stdlib webstart useIll go forward with this fix as at least a partial solution for problems in webstart and applet apps +This is impressive +We add the reference by country and determain with zhCN and zhTW then we get the different display item at the language. +Instead of validating at the end when the user clicks Finish we should be doing validation when the web page comes up to provide more immediate feedback +Changes to implementation in line with the metadata changes and publish response changesAll tests passedTests run Failures Errors Skipped +Where would be the right place to do this in woodstox? Ill try to provide a patch then + parent pom with dependencyManagement section where versions can be updated for CarlosI badly need this pom to be released +After a discussionagreement with Suresh Im implementing a quick fix for the NN with and +I think it is OK to leave it undocumented for the release but it would be good to have it documented in the reference manual some time +HenriThanks I was wondering if it were just the classes or the classes and the resources +I got bit by mixing the access type between properties and fields +Ive removed that cast from the test +Proposed solution goes against the principle adopted by the Maven project +This patch attached to also solves this issue +pushed fix available in JBDS L +Not sure what the default should be the current behavior is obviously to not confirm so we could keep that one or go for a safer one I dont have any strong opinion on that + replace block length with replica +I took a look at what is passed to both when called from parsehtml and when run by the parser passed javascript files +now its just a minor feature as a last resort im ok with removing it +Bulk closing all of these resolved issues. +No more error thrown in log when installed from +It concerns me that of the two times Ive done an actual migration from to trunk Ive run into nasty bugsIf this isnt easy to test for Ill try to add tests in as well +Most of the questions we get is how to get to internal storage +Ignore this request. +But without the patch on this issue the check in happens first and errors out +Wont be fixed in time for +Probably it should be done in commitrollback methods +If we are not planning to change the semantics I will ask Corinne to document for included in See +I would prefer to leave the as it is now +batch transition to closed remaining resolved bugs +all bulk edited issues have in a move issues to and +Thats why I suggest to only abort for the issue we know +The crux of the issue seems to be just the creation of schema not dropping schema +you patch was applied reviewed committed +yes please do a PR then its much easier to seereview +Probably its a bug within Maven +Ivan committed to windows. +this replaces the first added a note on useprocess +Included in release +We have cftraces littered in certain areas of our code +should now run under POSIX shells such as dash +Jon and Todd Nice find +Also fixed so that can reference that same external policy file more than once and will create multiple links to it +They should be tracked separately +Makes sense to me and is obviously the saferbetter way to do it +Should be out tommorrow +So yes lets try going the CMS route +Uploading a patch that applies cleanly on think one thing that can be improved is the structure of our ZK metadata +rev for trunk +An updated file +Committed revision . +You can probably remove that dependency +I just committed this +As written above the attached patch makes it possible to use ICU as a message loader +I realized that the main datastructure updates happen only after a task failscompletes +on the latest patch from Mingjie Lai +Closed by schema is added to XML catalog in JBT AS plugin dtd think this should be addressed in +Although the concept is slightly different and it is implemented differently it reminds me of the request processors of zookeeper servers +bulk move of all unresolved issues from to +One is that there is no reason we shouldnt be able to handle recursivetrue +Im in testing process +Johnny thank you so much for reporting the test failure +bq +Thanks Jonathan +We might also eventually want a HEREDOC variant but there is another issue for that I believe. +This will be the use case in Cascading +If I misunderstood your comment could you elaborate a little more on what you meant +Bug has been integrated but it does not provide a list of objects in thecache +xslt file used for transformation +I am planning to commit some of the RPC and protocol changes to branch by the end of the day today if there are no objections +It adds a checkbox for Regression below Patch Available +If you use X youll see a fair amount more information and itll be a little more prominent. +system test frameworkYou probably have some problems with gcc there +Image is too big Figure +I feel that the best trade off is having each jvm running a single task at a time but reusing the jvm for the following task +you can not run jaxws with JDK you need to have java or higher +We had verified that the server stops reading in your test case but had not checked the result with unmodified MINA +The module info lock was being grabbed too eagerly and that conflicted with another thread that had a file buffer lockThe stack trace that you have in the bugzilla issue does not show any deadlocks and I see no codeIt is interesting that two of the commenters on that bug were using STS +Also Ive made changes in this patch itself to include the missing documentation for +That is appreciated +Thanks! the file attached is the file using a jEdit plugin +Me explaining new getslice so now you would slice from with whatever your desired limit is then slice from the last name that returns to pageIan Hm +Caching means less trips to region servers +Hi Mark the zip file is missing +Could you please tell me if the is the latest version of the patch +Are you still having this problem with the and the mystery formappearing in the colsole output? I am using the nightly build and am notexperiencing this bugIt might be helpful if you attach the and +API Coverage as of July +all this add on does is modify your and copy a couple base files +Test failure is unrelated +Right now they sit inpackage But there are already several things in that directoryMy worry is that the code that uses these functions is untested and thus we are likely to not notice when something gets bunged by the code generator themWould it be possible to useOr to relocate all former residents of to before generating the code +youre right this works fine for and EAP it can be reproduced by using if I look at the Environment field + on +Ok this turned out more complicated than I expected +created host with an IP alias and specified that as the config for the RM and verified requests failed without the change and then that requests succeeded with the change +The bug itself is not fixed because first test fails on interpreter +HiThis isnt a Karaf issue Karaf by default didnt install any spring bundle at allThis is caused by camel project feature pull in different spring version which cause conflict this should be resolved in camel by tune the featurebundle version range +Modified the pom file closing. +Committed +What would you suggest the name of the bundle be? I dont have a problem with where it is placed +Thats a good point +Has been closed +This is one of these cases +fix is also shown in examples do I find the build that have this issue resolved? +If not mistaken even in JDBC they took that routeThis was mainly useful in ODBC to implement batch mode to ship results to client +Thanks for the report +All the dependencies are in targetdependencyIts set up with H The test creates a new database creates a table and inserts a rowThen it launches openJpa and retrieves the entity mapped to the row +First draft of a release note +resolves this issue +So for the security examples its because the and keystore etc are not present +If the application being submitted is an unmanaged AM then it should be an error to specify more than retry since it cannot be retried by the RM +Are there any objections to moving this to its own taskissue so that I can close LUCENENET in the next couple of days +Shai will this at all invalidate any benchmark algorithms that are already out there +Good luck +Closing +I still need to investigate persisting to the task log a bit to know for sure how to handle thisAM restart is achieved through reading logs +patch updated +Fixed and merged to trunk and for a while +This does raise a good point +Attaching a screenshot highlighting the couple points of error +If it does not exist or it is not in your path then it would return null creating the problem described in the readlink is called in Ruby and JNA is off it will try and call readlink from the shell +Sorry that should read not true for Hadoop Paths +We do want to allow multiple remote replications that only vary by hostname port and user +Attaching +Its committed to trunkThank you Joe for the great contribution! Thanks everyone who reviewedtested the patch +AlejandroI think there is a bit of a disconnect about the version of the design associated with some of my questions +Please provide full stack trace +Are you sure that the initialization of Configuration would catch all errors? I need to check it again but I believe that the errors are only thrown when the query with the error is called +I checked a large sorter benchmark run there does not seem to be anything close to this exceptionOne possible culprit suggested in this case is lack of space under tmp directory +Attached Eclipse project with unit test covering three recreation cases +Great! It will be helpful +This sounds like a good idea to me but why only the server administrator? If you have a common class loader installation each web administrator handles a separate context and that means a separate server scope correct? I would expect to be able to override server administrator keyvalue settings in the web administrator per context +What about completing versions of libraries? Larry had proposed some weeks back pulling the connectors further out of teiid so that they could support multiple versions and I approve of that direction +changed for HQL nested news support +Its wiki you can edit it as you wantWillem +But then you could theoretically construct a use case where theres a group access for some files on DFS and these are localized per user on the task tracker +so like this +I think the name concept is fine both internally and as a name it gets confusing when its also involved in the path rendering for the we could make it an explicit path property for that +We should also try not to clutter the +I also ran into this issue when trying to resolveSince the or so new methods are required by are new for nowgenerated them as stubs +But this still would cause a lot of overhead for a trigger that never references the lobs +Please check that the patch was applied as you expected. +If you float an element inside a container that has a background color set IE paints the background color OVER the floated elementIm thinking we may have to put IE CSS fixes in a separate style sheet that is loaded conditionally +The primary interfaces to get acquainted with are +Im running TomcatIve followed your directions exactly and I cannot seem to find thecontext link +There is a workaround for +So now origin or origin or some is stored in the server +WDYT? I whould like to have comments before finishing it and integrate it in MINA +I think we should limit the scope of this JIRA to making the default processors configurable though and open another one for supporting aliases +But you should be able to use entities to resolve the product version as long as you are consistentI was out sick yesterday but I have allocated some time today to do some more edits on the GSG +Have done manual testing to make sure these metrics work +It should probably be done one the branch though right +If you get some free cycles maybe get to review it Alexey +Axis was released today so it will be part of the next release +This directory should be removed from svn repository +Attach patch that uses a different field for the id for the following items organization cimanagement issuemanagement individualsThis will also fix will also probably need someway to update the database schema if you are upgrading from a previous version +Looks like your second failure is unrelated to the bug +Highlighted dir needs to be changed to jars +So inferring these properties should help in better plans for downstream queries +Oops +Im doing some testing with the web UI +Marking LATER pending a course of action +Just tried reproducing this and cannot +It was a workaround for the same problem this addresses too many reads blocking the write path on couchfile +This was helping more people cases and is unrelated to appending those request parameters +So that warning is to be have also performed a by mkdistro and running examples +This is a good one to start with +Changed sizing to be percentage of max heap size +But we are updating none of those on a daily or weekly basisI certainly hope were not updating the JDBC driver that often particularly in ways that would impact a simple application like a shellOur experience using the JDBC driver as an application dependency shouldnt be any better or worse than it would be for other users +We should start to use JIRA for documentation tasks. +be it an idiocy or not +Fixed at r The service exporter unregisters services only once +The original idea is to have a timeout when we encounter this error +This is something thats beyond my expectation with my limited understandingOf course this is a different topic but I thought I share my thought here as well +Turns out I would need to modify surefire to be able to pass in these additional classpath elements. +bq +Here is an updated patch +Thanks a lot Aleksey +I almost did it that way first and then blindly copied the conf pattern forgetting that artifact names might not be uniqueThis code is just as nice +Thus Castor assumes it is a and assumes an object exists to represent that using a correct definition for the attribute bug is am marking it as INVALID since bug already mistake this is link to bug in order to mark it as a duplicate of bug +On closer inspection this patch causes to act funky I have a fix for it +We can improve that code later on +I mean that I want to have change locale on the login page and get error message on the login page accordingly to this trying to implement my own decision for this problem but it always get messages in system youPavel +Fixed on master and See commit message for details about the fix in above comments open if it fails for you +We dont generates image now +I had HADOOPHOME defined and pointing to Hadoop +New navigation new JBT views for sync to public Jenkins +the slowwrapping in the sortingatomicreader must be removed before committing as well +I have not had time to really look into this working on Broker issues +Bug has been marked as a duplicate of this bug +Ill rant a bit about my experience from language detection though +newLongFloat +Simple version utility plugin for maven +That is a valid have you tried dragndropping a xml file onto the project in the Spring Explorer? That will add the file to Spring IDE as well +Triaged for +PS If the PMC thinks going straight to a makes sense that is fine with me too +Kabir what should be done about this request +Good points updated patch addresses both +Id want to set up the mailer with jndi like we do it for datasources because its generally available in web app. +Alas they must be because that is the only thing that changes between the test cases succeeding and failing +also needed to adjust the error message +Changing localhost to now lets me hit the service using any valid address +like me does I suppose everybody doesPThe only way to evaluate the if and when condition is IMHO listing at least one graph algorithm that requires a different Monoid to perform its execution otherwise no needs to model something unused +bulk defer to +The fact that it is using JMX means that this is an older version of the script +for version committed the ported fix to +I find it very unlikely that something is wrong with the antlr module jar but wanted to ask anyway +Thanks for the explanations! +Problem as welljava jar clanguageslib ctemp morning Backtick nap startingKilling the main thread at Terminated at java jar clanguageslib ctemp morning Sat Mar Backtick nap startingKilling the main thread at Sat Mar Terminated at Sat Mar +Can you try with a checkout of Seam from CVS +It should be something similar to the data format documentation we have +svn up works without complaint for me +release bulk close +I cannot see a case when this can cause data corruption +so it should always be there +Any chance that you have a patch available for as well? The cluster where I can reliably reproduce this is running so would like to test there as well as looking at it on trunk +Without it not only this action but some other features will fail +It includes a test that demonstrates the problem as well as code to repair it +are these reproducable +Issue closed. +cloudbr is being set as your guest bridgeright? See etccloudagentIs there a procnetvlan? If so can you send the contents? Itshould be building guest bridges off of whatever is listed in the Deviceline in this fileThere should also be some logs in the varlogcloudagent filethat might indicate whats going on +is a issue has been fixed and released as part of release +They also had have checked filesCopensourceorgapachewickettrunkwicketsrcmainjavaorgapachewicketCopensourceorgapachewickettrunksrcmainjavaorgapachewicketexamplesforminputCopensourceorgapachewickettrunkwicketsrcmainjavaorgapachewicketmarkuphtmlformuploadCopensourceorgapachewickettrunksrcmainjavaorgapachewicketextensionsratingCopensourceorgapachewickettrunksrcmainjavaorgapachewicketextensionswizardall are right coded in +Its not as robust but it would be quicker and simpler than scanning all the module dependencies +for above comment inexport HADOOPJOBTRACKEROPTS a findbugs warning in the patch +Initial patch for has more to do with installing and running hadoop daemons with just a single jar file installed +I dont know what the answer is to that one +Patch fixing +It must be something about Windows +Jukka could you take a look at this too please +I was able to run Kafka with the following commandjava verboseclass true cp +The codes original author was not willing to assistTo include an external projects code in an Apache Release we must have code provenance which means tracking down the authors who must be willing to sign an ICLA or SGAPerhaps someone might be interested in creating a clean room implementation of the Surrogate specification. +Reverting commit as it takes changes days old +However I cant imagine any usage cases now +This breaks validation and I get an exception before I can do anything in the controller action +The check is done when the file is closed +I have been using on the commandline for a little while now and havent noticed any problems. +Id suggest an appendix in the language spec should list all the typedefs from the assign all unscheduled issues with priority major or less to target +Patch committed +Ive verified that ESB should not be affected +Please change to if you intend to submit a fix for +i will give it a try +You said three lines only Nicholas? Maybe a patch for also could go here +I plan to commit in a day or two +bulk assign all unscheduled issues to next release +All the changes were made by Gleb all links in PDF are underlined displayed after them are color for note important tip boxes is words are is corrected except It seems that there is a bug in the +classpath by making it +HelloThanks for your comment +Thanks for having a look into this issue. +I thought it would get logged but I looked into the code a bit last night and the code that reflectively sets up the instances and calls the appropriate setters is currently logging the method call failures with debug +Thanks for pointing this out +here goes the patch +Thanks Philip! +Closing all resolvedfixed issues of already released versions of Roller. +Since the check now throws the flag isSupported is unnecessary and removed +We like contributions! +It works if you dont use these classes in your code +great review +screen site layout example +Emmanuel has encountered this before and the fixes I believe have been encapsulated in in the package +Revision . +OK thanks Peter +patch contains the fix for the problem with an attempt to compress unmanaged is the corrected it has appeared that memType can be Object while src type is unmanaged ptr +I think the error is clear enoughParse Fatal Error at line column Content is not allowed in prolog +Downtime wise upgrade is already quite disruptive hours with verification is not going to be much worse than half an hour with out verification +Since this is a little API change to we should get it in soon before its more frozen +The should be a method +This patch seems to work ok for me too +Okay its the autowire attribute on camel context causing thisIts kinda auto installing any beans in Spring DSL that is a route builder and thus we get our route x installedBut the first time it cannot run the camel bean post processor that injects the Camel annotations +Not just for Portal but for seam in general it seems. +Treat unscheduled issues as Native SOAP stack is not supporting JAXWS any more +A producer should generally be thread safe to allow concurrency if not you would need to mark it as so so Camel knows this +If you could have a look at it thatd be great +Patch to always use for hint keys +Closing Out of date issue + +Thanks Ted +You can apply multiple documents to a Hibernate Configuration object +Our experience is that even if we set the total memory threshold to a high enough value would still kills a considerable amount of tasks when there is nothing wrong with the RSS memory +Instead just register a roll of the log and just let out the exception +Thanks much for the patches Ive committed on trunk and branch +Hi SimonI dont have the commons libraries in my JAVAHOMEjre ls latotal gabriel gabriel Oct +I have fixed all the commentsto test private cache files I have updated the test +Attaching a patch and an html for the updates to the IN predication transformations +Thanks for checking out the possibility for a backport Kristian +New patch attached +Guys Ive added support for in the trunk +Turning white pixels intorandom color appears to be a symptom of a mangled color mapping +Fixed in the wiki +Deferring to Please find attached the fix for this issue +Does it makes sense to have a reserved namespace for internal use? Thenwe literally will have nothing reserved as a system namespace +Thanks for the nice patch SamuelWe might consider adding this to hbase +groovy files with to execute themSometimes I want to open these files with to edit themWhen the script opens files with relative path Its preferable thatworking directory of is same as script file +Updated db persistence file +has been marked as a duplicate of this bug +Thanks +I should have commented on but when our users hit this issue those affected records tend to contain a as part of the data by mistake +Just doing a quick grep audit of other settings I think the following is also mismatchedIn the format I believe has the correct has the correct values alwaysCan you submit a patch for this +Thanks Ill look into it +Changed documentation to note the fact that the binarys group ownership should be that of and cluster users should not be part of this group +works for me +Have done as much as we can within the project though +Move issues to next release +I assigned this patch to +Why you think they are needed when +Chapter section Changed +Ok I retitled the issueMark AFAICS Welds CC impl is serializable as long as the Bean implements and the bean impl is serializable +pushed +How would this functionality relate to the actual Blueprint spec +I agree we need the volatileyeah looks great Shai! +Need to update trunk to at least the version used in include moving to in this task +I agree with your comment related to this kind of trace logging but I am happy that the performance impact is now gone +It would be nice to get this confirmed by someone with more intimate knowledge on the topicIm fine with making our use of setDaemon consistent +Thanks Ill push this to Google Code and publish the details on +I am far from a CSS expert +Im keen on doing it early as a an additional means of vetting the page fragments we will ship in +Fixed in . +I was able to get the apk files +I solved in a similar way +Ah now I see it! Youre specifing a version that does not exist +Test should check or fail if the file deletion is not successfulb +Thanks Paulo for reporting thisI committed the patch +This works fine in jruby for the clear reproduction! I would guess this is another issue matching encoding with what the server reports we should not be transcoding in this case since the file is obviously not text +Cancelling patch to unconfuse Jenkins +It can be easily workarounded do not use inner class as page fragments +Ah! For some reason this specific node did not get the upgrade from December th to the th while all others did +Lars collaboration would be very welcome I do not yet have a design in mind for this one nor have I started actively working on this +If I am not mistaken Ive already done that at least the attachment management says so +still running tests but heres a bit of a cleaned up version and I pushed it to of Lucene +Hi Ive added an extra null check in Revision +Thanks! +Dont forget to delete the following files from svn repositoryvmvmcoresrcthreadvmtestsunitthreadvmtestsunitthreadvmtestsunitthread +Eli got it on the assertion thingy +Isnt that the maven repo zip that contains the AS artifacts? Why do we need another zip whats the main problem with rebuilding from source? In fact we should use the sources from the source distribution for running the AS testsuite +Fixed +For review +Attached a new patch +Not going to make this one in for +Reopening to close as fixed +Can you test itHow many memory do you allow to the jvm? On Solaris bit i must to define MAVENOPTS +Perhaps its that the former already has a value and is being overridden in the properties rather than any intrinsic difference between the you set Base Base basein basebase and make sub inherit directly from that instead of base then youll see that retains its default ibiblio value +Ill have to look into that some moreAlso now that I have the images theres no need to submit the patches as a ZIP +Am I rightSent from my iPhone +Dev version worked +Since this wont work for I think I need to back out the commit I made +That class is gone in AS trunk +We can version the two features so people building on top of Karaf can pick either version depending on their own requirementsbundles +Lets do it for cr. +Atlassian have come up with a workaround and this has been enabled +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Received no objections to closing this wont fix +Removed from docs +One thing to consider is do we require that Composable functions be total and +Will they timeout and undeploy themselves or exist as long as there is some connection on it +Patch adds configuration properties to the Spring bean via Cocoon properties +Lot of good work a few things need cleanupI assume youve tested this with test Jiras +Tested with sample mails from my gmail account + +But again since I get a orgspringframeworkdatarepositorysupport I had to disable Mongo since the JPA is actually the core of my Mongo I was planning to move the error logging database from JPA to Mongo +Patch for review +All the things you did I thought were in there +The Netty replaying decoder internal buffer maintained a buffer of data read which made server side memory consumption duplicate +Closing as is already released months. +Therefore some plaintext data might pass the validation even if it does not pass the validation +No problem +and no its not working for meOS Ubuntu Its the only place I see that behaviour +heres the issue report +It allows users to check that the resolution is correct +Hi ClausI think we should provide the management for the temporary Queue as these endpoints are exiting in the camel contextClient may want to ask managed information about these endpointIf we shutdown the camel context all the camel managed resources are gone and it will not introduce any further burden +Thatll solve your problem for you +Updated the issue description according to the latest Trustin! +I was asking Leonard Rosenthol through an associate at work for the specs +Please check latest version of in svn and let us know +I was trying on Juno not Kepler +for the patch +the first thing i thought when reading the description of this issue was that it seemed like a strnage bastardization of the class +Bottom line we need to solve the problem of how to handle conflicts +impelementation insertBefore and removeSelf code seems to be repeated in implementation? Add a summary of test information to the javadoc of test methods For commented tests can please add TODO and a brief description +Im coding some projects now that use hibernate annotations that require this in the classpath but it might not be such a bad thing to make annotation support an option that requires an extension +Alexey please verify +I dont think the time is lost +Attaching a patch that instead allows the user to set the session timezone explicitly based on the proper oracle database timezone nameThe existing code also tried a fallback codepath of setting the timezone to UTC +Added TEMPDIRNAME back to of mapreduce and annotated it deprecated +Some support for this has been added for Hive and Knox provides support for that +Ive only found commercial offerings that mention support for Apache Derby but I know nothing about this in general. +The view doesnt implement the interface +Heres a better patch. +Ugly and wasteful but effective. +The sample works well and is simple to understand +Whats the testconfig youre using here? Whats the cachecachemanager name? Is it AS +Committed at subversion revision +Its been a little while but I believe that was the only line required to the life of me I cant remember what difference that actually made or how I arrived at that solution. +Commented at line due to missing constant in Eclipse +Please look in history for what the actual resolution was. +all issues to the CR +BTW Solr does not use Servlets +So let me make sure we are on the same pagewe should call out an official vote on the dev list though +In this situation I would gothrough as a pair more than onceThis is why I feel its important to use the same mechanism throughout +Closing +Hi MikeHave you reached a conclusion about whether this patch should be backported to? Thanks +Sorry for the guideline mistakesTo be generated correctly i had to remove the newlines between the foblock foinline and the text inside the +Will implement that approach +The issue was that the Service had a business key set but the business entity itself didnt set the key once I fixed that then it saved without an issue +Once it gains attention we will move it to incubation or adopt it as sub project +Heres an updated patch which takes out the delayed shutdown stuff and the health check +Resolving. +Yes we can remove unnecessary code +Alternative is to skip the test as a whole if is not available +Have investigated this and it seems that you are using attributes +Ill design this with the intention to make it contributable with pluggable cache implementations +Closing resolved issues for already released versions +However is resolved in and that revealed the configuration problem +the tests pass on the hudson machine now +verified in +Why cant we transform a modification ofchildOrder into appropriate JCR events +Public classes that are not marked with this annotation must be considered by default as Private +We decided for having the style with the operator on the line before as a guide to ensure there will be no problem and that works +I also added a few tests in to check this behavior +let me take a look +The two last points have been fixedThe two first are improvment and they will be in the roadmap forSo I consider that this is not an issue anymore but an improvmentThanks again Van +QE would you please confirm one way or another the status of disabled in select and inplaceSelect components? +mouseover event call toolTip component instead of onclick +Thanks +CommentsIm ready to refactor this ala the other as soon as this one is in +It is attached +Thats why I am against it +I see what youre saying +Alexey please verify +Since I went into JEE I have been engaged into EJB architecture without any Spring etc for most of my working timeNow I just renewly come into Open Source Big Family +First by moving the widgets to the first page it means theyre not accessible when adding the facet to a regular Faceted Project +makes it look better but it is still a little bit out of place +When the collection is actually created it will find the link +It is all in the sandbox along with a test application +Thanks +Cancelling the patch because In the patch pipemapred does a arrayCopy to get byteArray of the Texts length +hint from Gurkan When the application is removed you have to remove also from cache othewise create a leak +Great! Glad it works now for you +Nice +imho advice chain is not important as it is applied per attributes applied to the type or its methods are definitely part of the cachekey. +Check if the given source event is a jar file and skip it in that review commited in trunk waiting for approval before committing in looks fine only concern is if element ever can be null if it can then the code checking it can NPE +on the change +implemented optional inplace webapp or war deployment user can choose now which one to property for better recognition of project type +Thanks Erik for the fix +The Execution Plan View is updated with each query execution in DTP +Great glad this one is fixed +What do you sayNow Im creating the patch to add info about the doc to the sitemap andto the DRLVM index page +cumulative issue for to is submitted as +Closing. +Setting component fix incorporated into codebase + +See also under r in branch and r in trunk +was for a primitive version of whats discussed in +Glad to know this +bq +The easiest solution is for this to be supported directly by the transaction was a possibility at the time but I cant remember the reason for not doing it +Much easier than I had thought almost ready to commit work +Done in CVS setIve tried to use this feature but I dont think its quite there yet +Fixed with a quick patch to to build up the attribute list. +Hmmmmmm it sounds more like a problem with my Custom naming strategy +Do you also plan to make a URI +And then it is only wrong if there is the usage of null as value as you did here in the second invocation +the test case fix version for bpm to expected inclusion +Issue is due to branch translator could create situation when true and false branches point to the same basic block +sounds like a great addition to a FAQAssigning to Ben to remove the sort and document appropriately +Here is the class modified as per proposition +libgriddevbin +Perhaps only if the process exist with an error +Then of course you could also add more but this would be the basics +And for this is a right +Uploading patch for branch patch for trunk +code committed to new columns need to be added to the createdrop SQL scripts +Patch +Probably will never happen. +I just dont want why the author using the Text instead of String directly +Updated to apply to trunk +Applied to branch and trunk +Im struggling to determine why it is happening +could you generate a new patch with indent? I saw some places are two spaces indent +Looks like Keith Donalds code from fixes this +Applied. +Attached is a proposed patch which merges the db and column routines in the parser to provide a single ALTER TABLE ALTER COLUMN statement that provides support for changing the NULLNOT NULL constraint on a column in a table +I dont know why findbugs didnt flag it in prior patches +No DBCP does not use hibernate +I cleaned up my repo and source reextracted applied fix for got a successful build and the deployable zip file is Meg +Hi Peter thanks for the update report +Some potential syntax approaches are suggested in the forum thread linked above +I tried to keep the databaseType variable but that requires to define Driver in standalone configuration file or create it through admin CLI or Web console +I feel you +So if the plugins are properly loaded and activated it is impossible to get here +Fixed in nightly build +One last question it might be reasonable to move and friends to +Woops sorry +The attachment contains a unit test exhibiting the problem +This is fixed now +i guess because i have not removed the files in solrzoodata right? Yeah exactlybq +There appears to be no value in passing this exchange property over jms +Updated assertion number and text with revision +Ill implement it as a independent functionality it is correct to wait till it is stable +If you want the let me knowRegardsTim +Thanks Stack +target was gone +bulk close of resolved issues. +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Thanks for the review +This test should be fixed in ER on Windows. +Hi Tom Ill take a look into it today +Once you declare a dependency as provided all its transitive dependencies are upgraded to the same scope and they should not end up in the final artifact +Adding a variable sounds like a good idea +we should get this into as well +I put in some javadoc placeholders that point to the wiki etc +That also can be solved with made a patch and committed changes to trunk right now should be injected into please take a look on and class for more detailsRegardsLukasz +I opened a Java file and the method definitions there was red +Stack The build is failing again +Dinesh Ive manually tested this on Windows XP and Windows however would be nice if you can verify this too +Obsolete fix as class has been removed. +It could not be reproduced outside of the environment shown below Java Information Java Version Java Vendor IBM CorporationJava home optjavajreJava classpath optderbylibOS name LinuxOS architecture xOS version Java user name builduserJava user home homewsbuildJava user dir optderby Java Platform API Specification Derby Information JRE JDBC JSE JDBC to track the +Added a unittest to reproduce this issue +Test bugs and findbugs warnings should be gone after this +Attaching two patches which apply to some version of though probably not the tip of it +should we reserve the commitpoint if after an indexversion command +sergey do you guys have access to a Vista +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Then we added all of the limits to try to curtail this but its a bit of a stopgap rather than a real solutionIs this better in MR because the state on the RM is much smaller? Or is there still state being retainedOne thing we should consider is to adopt an interface like s objects that are going to be retainedcached need to implement it and then the limit scan be set based on memory usage rather than strict counts +Issues appear with. +This patch adds some tests to verify that you cant use updatable to corrupt generated columns +Dianne can this be marked as resolved now +Attaching updated patch +But DB connection is gone the server keeps retrying and the client keeps waiting for the return of the call +replaced standard backtracking with SEL merge function +Martin doneRgds and working. +java files from my build including only +So whenever you want to storeupdate such a value in the repo you use the registered prefix slingThe prefix is used by the sling post servlet only to mark special commands implemented by the post servlet +I still stand by what I said on the review board +Still how do we enable their testingRegarding the service loading +This is the WSDL as provided from the server +I suppose I was assuming that you were using the now removed COPP but since youre using just the OPP that should work +I have another question related to the NCDFE +Thanks Jerome. +Added a defer parameter to Submit and that defaults to trueUpdated the component documentation for Submit andWrotes unit tests for Submit andAre there other components? In the WML framework perhaps? +scenario but I wentto look at the problem and I dont get one so where to from here more testers please! and quote your Platform Tomcat version Java VersionHope these comments narrow it down a bitLet me know more specific testing requirements if you want me to do more testing +Appears to fix the problem +This simple schema can be used to test solr integration +Try persisting your key before the master +What are you trying to achieve +Sorry can you clarify what the expected behavior is? The docsDir attribute is used to specify the subdirectory in which the documentation will be placed relative to the root of the webapp when the webapp gets built +I think its ready +the first example should be a duplicate of the other issues but the second warrants more is a duplicate the path parser is called by to that is called by the purge routines. +Please install the version of AJDT first and then Spring IDE +trunk has moved on to geotools nice work everyone. +Good stuff. +Attached a fix as patch a which asks to clone the value as long as materialization isnt forcedRegression tests just startedPatch ready for review +The features would be a superset so the capability wouldnt go away but the details may change so might involve some rework at some point +The test case has been running happily for a few days now and on a bunch of hardware via fuse tc. +In Web Services a WSDL message part of type xsdanyType can be mapped to a parameter of type +Hello BindulThe is an interface +Alejandro want to give this a try +I have some code which I will check in eventually but not going to happen for M +The validation error message after clicking the Finish button attached +Fixed +In addition to Romans patch hdfs script will invoke jsvc from PATH instead of using bundled jsvc +The patch preserves full backwards compat and provides a clean way to move towards a consistent codepoint based API with would still provide the flexibility to make use of the old and buggy behavior even if the char based methods are removedI consider this patch as a basis for the discussion how to solve this problem +closing this one since we got entities and comments now +Strictly speaking isnt a incompatible change +RC was what I was using +This patch adds comments to the latest version of the and classes +Thanks + +Hi RickI will modfy the commentsThere is no problem with its behaving perfectly fine ie returning networked connection in case of framework is set to and other wise +I havent included new tests in this patch since an existing test covers the change in this patchPatch committed. +Ill make a patch +Trying to fix it without a clear plan about how it works is a waste of time in my personal opinion +For the moment though I think its okay to keep it together +Can you please rebase? Thanks. +Test cases result will let you know tomorrow morning as it will take time +Nice find +why is this a bug? does something not work properly +You can cancel the passphrase dialog when youre prompted you can have invalid keys and you can also have keys that are removed from +Fixed for MSending modulessrcjeeschemaTransmitting file dataCommitted revision . +However again DST is NOT supportedI fail to understand why your code doesnt hang with the change to the response filter above because DST doesnt use response filters with! +a script to svn mv +But its worth asking if anyone really needs compatibility +Rado please verify and close +resolving as duplicate of bc that is actually a different issue. +Thanks +I used a panelGrid wrapping it and I rendered that panelGrid for solution +Corrected the testcases +Ill be resolving this as a duplicate of that issue +I just used to create a app and it worked +Id like to see this facility as controllers are not always as transparent on the example for one application I am working the url mappings look like +Full thread you attach your test program please +Hi Ive found a bug in this patch +However rendering a view is not possible as the flow will not be available to resume on the next render request since it has ended +! +Hi Milind as Sanjay pointed out we are going to keep hftp and add a new scheme say httpfs +Thank you for the review Nicholas +Im on your other issue the test +Patch as per defect description +gitignore +It should not stop execution of subsequent tests if test system property is set +James there is no UI for this but you can do the followingYou can change the media type manually in the GEOSERVERDATADIR file +This may be an improvement of RI +Integration tests already exist +jboss as trunk has upgraded JSF and it works fine now. +Also the additional typecheck at every putgetetc +Each store file we open has an index of long to HSK +Re opening this as we still see the issue and was also observed in the customer environment +So +shows the memory consumption after a load of approx for a detailed description and for raising itI usually ask users to have a discussion on users forum first but something with this level of detail is appreciated. +More thought and better design is probably needed +The test will get failed on Jenkins +I think the current code is lenient but Im not be fixed for some tie now +A number of documents state Copyright Conferencing LLC + +Ted I attached the patch I pushed in for all branches although its called trunkIt is also good for other branches too just some line difference which patch can handle it. +Also contrib was missing +Sure will do. +This is actually intended only properties are autowired meaning anything other than primitives primitive wrappers +Im sorry for missed document +Thanks JorisAt this point I have two Concerns can not use and web factory for the we do that we get error with synchronization issues with Save or Update have used +sender is concurrent but there could be a race between the get put and if initiateConnection is called concurrently for the same sid +Okay great Just making sure +on the committed to committed to the committed to +Is that correct +Contributor License Agreement be included in. +Fixed in r +Keith can you please verify the issue Bill reported +christian is right TABLEPERCLASS with the root class made abstract is the way to map your requirement. +Updating this since it looks to be complete marking for instead of JIRA cleanup from above commentaction. +Because the is used to create a unique id for Execution +Agreed +done for orderinList task wiil be done by Vladimir Sukhov +They ran fine on my windows XP machine with IBM jdkAlso the changes for this jira should not be backported further back than +Does this issue can be closed after recent improvement on siteThanks is the when we update the site some years ago so I will close itThanksMarco +I need someone to give me a +Im sitting in the lounge now in Taipei waiting to get on the plane back on with time for this +k researched more the issue +Good startStill planning to post a patch this week a few subtasks to wrap up first +Menu buttons visually distinguish themselves from list buttons by not drawing a divider +Ill post any conclusions here after Ive looked at the results +Sorry for thisGrid +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsermenu +Can you post an r +It should be noted that the changes here are the first effort to implement the feature and have not been fully tested +miquael please put this in a new jira this jira is not related to itbtw +The minmax gram code I added only applied to the +Issue patch is supposed to fix at one place breaks it at another place +Actually i fixed couple component builds before new year by adding requirements component just didnt notice seam has problems as well +That particular message is created by Maven Embedder and we have no control on it +It seems to be by designTags are stripped if room option Allow Font Styles is turned off +should automatically close the connection in this case though so that the unread data doesnt corrupt the next request +I think we can skip tests if necessary on old platforms that dont support it or write a subprocess compatibility wrapper if we really need to +Itll probably be removedIf we used plain Configuration Enis how would we ensure that hbase +NET client +weird +However it should be quite possible to deploy this with your plan +Ornone map reduce an offline discussion with Sameer on this and we felt that this issue didnt require any fix +So the information is very important to us + +If running embedded the probably shouldnt do a separate login from the +If thats the case then the individual attributes should not be used for optimistic locking comparisons right? should only be relying on the Version column to determine optimistic lock exceptions +Thanks Nicholas Googlepatch committed +It is pretty complicated +The Security class could use some thought also most of it is unused at runtime +This would make installation on OS X easier +Ill try and break it down more andisolate the problem w another example before reopening. +Going to look into it today +Doh +I just committed this +Thankssuresh +Justin would knowfor sure + +Please open a new one for if this still applicable and has relevance +I volunteer to work on this issue +Thanks Brock! +Marking as closed. +I have uploaded a sample PDF file and the HTML output generated by +uploading patch with +Thanks for the patch Chunhui +Ive added a force attribute in the config and adjusted the generated WSDL to workaround the Metro bug +We can cache the value of +If its significantly larger there are jars inside that shouldnt be there +Please review and commit +If you have so high load on your server buy some more of them and distribute your searches better to more machinesYou should maybe limit the number of acceptor threads in your servlet container +couldnt we go with this as an iterative improvementYeah thats been on my mind just didnt want to give up completely yet +They are at +version +Move back and assign as necessary +For reasons stated far as I understand t and t are used in coming up with canopies +the class defines a new private method is which isnt used anywhere +However Im not sure it would make any difference in practice since we allocate row keys into the regions as well +Bin could you please confirm if the problem has been fixed with the recent patch? bug is essentially the same as This bug has been marked as a duplicate of +So the effect is essentially that shims optionally depend on hadoop but ql which depends on shims would not have hadoop in its classpath while compiling +The ability for displaytag to be more flexible in this arena would benefit Im sure a lot of developers +The standalone environment doesnt support or +Mylyn ve added a server property to which functions similarly to the one in +Updating patch Added copy semantic to checked into trunk and branch +Attach a patchThe idea is when ledger handle encounters metadata it should reread a updated metadata and compare it with the old one to check whether it can resolve the confliction or not +bq +ThanksJ +The pull request should be merged into and into master +The failure is in lang and it seems like that the trigger action plan is getting regenerated with wrong column numbers +I was not able to test on because of very very slow access to the plugin list +Hey Daisuke these descriptions look pretty good to me +Thanks for the reviews +As discussed on list this functionality has been removed from Orchestra +Thanks +OKhere is an integration test that verifies that when you specify the skipAssembly property on the command line no artifact is produced +I was unconvinced it added anything at first but looking through code it could save us a couple of lines in lots of places +N Runtimes found please move that up after the Searching runtimes finished so Press Ok +Mark do they fail before applying the patches or after? +SPI version of +Ill try to again +The new docs look great Dag +checked on demosite on todays snapshot without any page code changes +How does one generate them? Can I use any UUID generated by our internal service? Right now the patch uses cooked up that need to be replaced with correctly generated before commit +So we penalize native calls in favor of ruby calls +We ought to remove the javadoc comment in that states its an internal API +Attach method simply set existing component to a parent and deattach method removes a component from a parent +Look at the portable extension code is CDI specific +Please check this code in. +Sorry about that managed to attach the patch to the wrong issue twice in a row +Doug expressed some concern over that idea in I guess since data can sit in memory for an arbitrarily long amount of time before getting checksummed but I would imagine most people are using ECC RAM anyway +Ive tested with a clean local maven repo +Fixes malformed XHTML in Fixes bug in that caused new webapp compilation to fail Updates test to hit the new URL +Implemented hashCode and equals in the Option class so can just keep a set of the options +I agree moving out +Going to keep this open so I can test that the installer uses these + Dependency Version Fuse SF We need a new SF release first +Just to confirm Vysper does not currently support virtual hosts +Thanks +Having looked at you proposed code I think it is definitely beneficial to include the testSinglethreaded test +So not an issue. +I suppose thats OK? IIRC the log is rotated after N edits rather than after its grown to a particular size +bulk defer of unresolved bugs to +The behavior mentioned in the bug is the expected behaviorWe should NOT be programming the IPV entry of the non default router +Ah I see thanks for the heads up +ignore the bit I said about Whereas in bash the z checks for existance of env var in tomcats batch file the not exist is used to check for files like or +the patch works with the example provided however it does not seem to work with the original that the patch author provided +was not really usable for any large projects +java PWD Please see the for more details +So we can close that issue and start preparing a new release +I will try to revert it and pull out the columns that it does not supportI will also make the other changes you suggested +Thats unpredictability I would not expect from a build tool. +Heres an implementation that extends and uses for keys +However that complicates the protocol and makes it unclean +Again I dont see a bug in the current impl +this is not done yet +ed on Reviewboard +Havent tested this much but it fixes the problem for me +Ive no problem with the idea of it but it is file specific in a class that knows nothing else about files +I do not use classes from Eclipselink at all in my bundle so it may be during deployment time +where durability is guaranteed by registering copies received in memory in N serversWe had considered that option in the past and that doesnt work as well +What are you trying to accomplish? Whirr is not aware of the EBS drives +We catch the noNode exception in case the parent dir is not created yet +Perhaps thataids in comprehension or perhaps it just crufts up the code even moreViews triggers procedures constraints and other deferred compilationobjects in the language seem to be thorny +Best regards Younes TERRIRESS +Closed after discussion with Ellis +Now all the native libraries will be extracted to varnativejansi directory in the building time and they will be loaded from there on the runtime. +I did search bugzilla and I did look at but my first impression was that it dealt with absolute URLS where my problem dealt with a relative URL + represents a specific instance of this issue and prereqs a solution to this JIRA +I just committed this +Id like to move it back to derbyall and open an issue to covert derbyStress outlining the issues +Im currently using and going to upgrade to the latest release soon however you said that I could use the aegis mapping file only since so Im a bit in advance! Im developing under JDK is jsr still a possibility or is this only possible with Java +We will support Turtle as defined by the standard using the newer Turtle parserWhile spaces are allowed in RDF URI references this is the only place in any standard they are not legal in or +Sure Martin +Brian reviewed and approved the change request so I merged this change request into master and am marking as resolved. +We want to fix it for enums though as they are special and Juliens test case is good there because it models the specialness with the isJVM flag +My thought was putting the defaults in the logj conf in both of the +So that aspect of this problem will go awayThe second part is that this behavior happens on all of the rpcs not just map output transfer +I tried it with EAP ER +Didnt get a chance to tackle this tonight but I will try and fix it tomorrow +too. +I meant you cant register two items with the same priority +That issue seems to be around supporting filters using the new REST downloader +Thanks Dhruba. +Ill copy from the old re probably right on that one +This allows the common patch to be checked in independently of the HDFS one +Move it to the compile phase and have a separate execution for? Add srcmainctest +Thanks for putting this into jira +This issue still has to be worked somewhat as the wiki is not in order at the moment +I have to change back my domain model to the Extension Object Pattern where the extensionrole objects are not identical to their core objectI agree with Ilya +UpdateI was able to create files greater than gigs on my system through Java BUT I was not able to ask for the length of the file using on files larger than gigs! The exception thrown was Value too large for defined data type +Note that the issue tracker flattened some outlines that were two three or four levels deep to a single level +The current rev of the functional spec needs a correction The spec says that the varargs +Great work! +Sergey I cant supply this very project since it requires quite an extensive setup import of data etc +Further testing has shown that the problem Im experiencingis not well described by this bug report so Im closing itIt now seems that my problem is related to using in an overloaded method in a base class. +Suresh any comments here +Thanks Arpit. +In order to open the file youll need to uncompress it if youre on windows Zip should do that just fine I just tried +Daniel can you provide us with some sample data and the SLD? Adding null checks may just hide the real problem youre facing +I just committed this +No the other way around the one that introduces an extra warning +Good point +Hmm out of curiosity what result do you get when trying to parse this? Itisnt clear from your remarks +Well then + +Is this on +require admin privileges to make the request +I assume the strategy is Per region sorts with a final merge sort +For the above test case I get iterations and evaluations +All unit and integration tests pass +Thanks Mikhail +Im not sure is it correct to reopen the issue but it turned out that the supplied patch breaks class unloading in some cases +I think you mist be much more precise on the suggested behavour of the system You can actually install local files to your local repository You can always rsync a repositoryI cannot see what you mean +So it may be better to fall back on simply using the new low value as a single value and skipping reserving the whole increment window for that go way the interleave would simply skip and make some go unused which seem like tolerable for a rare caseBTW if you have time Id find it very useful to hear an explanation of why you regard the existing code as safer despite the flaw demonstrated by the reporters test case and despite the evidence that a case has apparently been surviving well in both the test case failed by the existing code and in a production environment +Verified in +Compiled and installed without error messagesFuton shows a little message and provides a Link to add admin account +Attaching updated patches with expanded +There is already a bug open to backport certain missing in the mapreduce family to Hadoop and it is almost ready to commit +Oh well! +That way you should be able to look at the nightlies Tuesday morning and make sure everything is ready +Fix the simulator for and +I believe I have that fixed but well need a +Because of the bad internet connection to the north american I have trouble to create the JIRA in Apache JIRA so I submit the patch into this the table name should be low case if you use the postgresql patch is committed in the Karaf and Fuse ESB fuse kernel trunk. +Note I could not get svn to delete the EAR file that was accidentally left inside one of the folders +Is this patch moving forward? Is there a target release +I tried the change over an ssh connection and downloading the data was noticeably faster with gzip compression enabled and no. +So being familiar with the overall test harness will be important and converting existing tests to and fixing existing bugs are both good ways to become familiar with the overall test harness +Attached the broken class Yes the cache must be synchronized externallyNathan thanks for the suggested changes +This patch fixes the error by changing the list to contain mutable entities by not getting the list from cache +Ok Alexei +bulk close of all resolved issues. +Is there something which needs to be discussed with the EGMaybe we could start to define a couple of unit tests and see what happens +Pavels data has a third coordinate that is always null +The would be the feature of being able to specify some fields as attributes and some as child elements +Ive made the Deployment Repositories deployment keep count of the number of concurrent requests they are serving +is designed for a single proxy +Use the same cache used to save information related to Resource classes doesnt sound good because asking for non existent resources it is possible to clean that cache and that is not the idea +What is it you want us to do exactly? The subject line isnt particularly detailed +Applied thanks! +Does anyone have benchmarks where Avros object reuse of Utfs demonstrates a significant performance advantage +And most importantly the original part of this bug report is fixed as well +Redirecting the gratuities to him Thanks Luca! +Ill look into solving this logic. +I created which tracks out of the failing tests +Once I finish this will be fine +sorry I miss resetI want to split with +If it is agreed that release notes are good for regression fixes as well then someone can check the appropriate box to have the release note picked up by the generator tool +Andrew There is a possibility that the problem is happening because the perforce plugin is not installed +Looks like this is duplicate to +reopening pending decision on upgrade yuicompressor to +r +In my view with QMF this should be modeled with a QMF user schema and then if that object is supplied by the broker or something external it makes no diffThen all the permissions can be applied to all the methods on that schema using the METHOD object +Initial diagnosis shows that this should be a blocker for Hadoop as there is a deeper impact +Heres an updated patch that properly handles token streams in a number of places +Many thanks GeorgeSorry for the failure +I prefer calling these parameters final rather than force +New and Noteworthy is the responsibility of the developers assigning to Dmitry for Hibernate Dmitry assigning this to you as completing new and noteworthy sections for the whats new page is a developer responsibility +Oops +btw I tested the attached patch manually by running hadoop namenode format and trying the various possibilities +The current partial list implementation does not observe the config variable +Thanks Ivan! I tested the build on Mac and Windows with this patch +Just adjusting fields for Release Note purposes +Merged into AS upstream master and branches. +Its JSF so it depends on what version we bundled in +Any more comments on this +No obvious time dependent use cases present themselves at this time +Please find attached a completely rewritten patch from scratch namedI believe it does address Elis concerns since It will only build fuse if the property is defined +Closing as later. +thanks Prasanth +Im not sure this is a good ideaI think Arun is completely missing the point here +It seems more adequate to have an OOZIE topic and let subscribers attach with a filter +Yeah +At any rate it has xalan in it so removing it solved the problem +This seems to be a change in +Does search requests against the lucene index +But as I understand the goal of the app derby is never going to give you what you are looking for if it does a sort rather thando sort avoidanceYou have choices assuming PID is a unique key first issue an order select to get all the PIDs of interest and then as you fill each row in the UI do a stored prepared statement to look up the data associated with the PID +The java tier is ok but the browser falls into an infinite loop as it keeps requesting +Reflection is no longer required so is removed +The attached patch updates the installer to handle changes for ESB as wellas fixing the following check that profiles exist before installing to them copy the JON plugin to the same profiles as the other files dont comment out the username in in any profile as this stops working in some cases apply changes to instead of overwriting the file to handle differences in the original. +Sorry if Im a little late on this and if Im nitpicking but would it not be more consistent to call the attribute instead of parameterIndex +We should warn them if they attempt to do this +Your suggestion makes sense to meEven if the sendid should be unique it wouldnt hurt by removing the timer first from the map and triggering the event next because it should remove the timer in the end anyway +Robert Benoit your issues can still be addressed +Id think Have perfect Hash in cluster storage so mark this issue as invalid. +Update comment about what parentdir is but otherwise lgtmAnother minor suggestion is to actually print out the bulk load command line in the logs also so that one can grep the output and execute the commands afterwards instead of having to figure that outThe test failure doesnt seem related +tests included +Well it was clearly as much surprise to you as it was to me wasnt it and I dont recall any documentation on what exactly the WTP feature does enable +Fixed so closing. +Changes look good to me. +Will try again and update on that +Closing JIRA marked as wont fix. +Needs to be resolved for +This patch will not be included because it relies on Java +Some generated files depend on multiple source files +Snappy in trunk needs work though + +Thanks +This one is essentially the Adams patch from some passing unit tests I wrote to exercise the new range functionalityIll commit this by the end of the week unless Adam spots a bug +Yes there is a way to add custom request mapping conditions. +They can even hack the CSS if they like a special color for production +Also I noticed that there are tests for and something else called +Example couldbe some issue if removal and passivation timeout where the same and both kicked in at the same time +Sorry +All implemented in mercury now. +Diffs for this wiki need to go to a new mailing list rather than spamming any existing list +Marking this resolved I created for the port for JIRA organization purposes. +The release will be in a day or two so this is mostly just a remote check that its fixedToo tired to write anything of interest so heres the commit messageFixes error with reducesApparently we never tested reductions on? As far as I can tell they never should have worked +Rather moved to CR to be for clarification Nick youre right +Ray was this fixed by +Committed to trunk and +Our fault but it needs work again +They are all crunch example jobs +Hi DavidCould you please provide the documentation in xml format as the other cocoondoc isI will then apply your patch to Cocoon +to and also moved the entry in trunks. +Thanks to you and everyone who helped fix this! +bulk fixing the version info for and all affected issues have in cool! It would be great to have single dataimport handler that serves multiple imports at the same time +Adding another analysis but unfortunately this only contains the faulting thread +The production config has already at INFO as against to what is mentioned in the description +Created for paging. +So we have further confirmation of what we captured in wire shark on the noisy vlan +Looks good +Moved Abstract classes to extension +This needs to be investigated. +Giving him something to do +Attached demonstrating the problem +and Im going to try to understand this header styles s reverted at revision +Fixed in revision +Please change the password ASAP +thoughts +OK as a people I didnt read all the document before thank you to point thisAnd could you help me to understand why the attachment not print method name and line numberthank you +Note the patch was created from the branch +I added a vote for this to get fixed +A minor change in to that handles nulls consistently +Attached patches for trunk and +hmm +Well in the case of this class I would guess that typical usage patternsarecreate startstopreadsplit readIt makes sense to me this is how I find my team using the classHowever if one of the underlying constraints of the framework is topreserve binary compatability then it makes no difference +The Saxon bundle version in the is no longer valid +Please retest with Beta version +net stop serviceName command exhibits the same behavior as reported in the first comment +Where were you thinking of adding them? The directory perhaps +Done lost issues on the ruling IT thanks to this all were indeed false fine for the Comments should not include code coding rule but not for the Avoid Html comment coding rule +Can you reattach patch for Hadoop qa +If you want to open a new ticket for the lack of logging feel free but we could also discuss via the mailing list +Hi JimIs there any benchmark to test these claims +I have nearly done with this feature but need a test case before I can resolve this issue +Youre a light construction company and you RENT an earth mover +right this is to keep nodes from thinking they are responsible for more of the ring than they actually are in case of planned or unexpected cluster restarts +sql file resulting in parameter replacement not happening +The attached patch changes the class to accommodate the requirements of the slfj frameworkI ask if this could be tested and if necessary I will update other classes as per my earlier commentsAt this stage in time my thinking is that we should separate the removal of from this issue as they fundamentally they mean different things for the Nutch codebase +Rsolved +the get is not synchronizedso we can get the buffer queue fast and put the put request into the queue fast +I applied the patch the header is not parsed yet when the file is loaded +i run into this problem when using the jackrabbit jcr implementation +I recommend you pursue this with him or her and theyll make sure appropriate issues get filed to make sure necessary error reporting isnt suppressed +Resolved in branch at revision +for the patch +Committed all of your patches and contributions only modification I made was to put Base as a private class in rather than creating a new module +However clearing ALL cookies seems to help! Cant explain why cookies would have an effect on the cursor position in autocomplete text field + +Committed to trunk thanks Jonathan! Also dont forget to update the wiki. +the patch looks good to meI believe that according to the Hadoop bylaws all patches require a from a committer before they can be committed +NAD Alaska EPSG available in the CRS. +Lets have this in as well +Feel free to go ahead +by the way do not commit such RFC without forum thread opening. +Hopefully we can do something about it in the future but its not likely to change for a while +it is like drools +Cleaned cache of the browser and it for you reply! +We generally only support a full upgrade and no mixing and matching between module jars from different Juergen HoellerIt was really clashing with old spring social problem +Thanks it is working for m closing this issue as fixed +compares nodes based upon the qname definition of these nodes +Contrib test failures are not related to patch +yea this was fixed as part of the commit for both JBW and Undertow. +close off release +Patch are not required for this change +Fixed patch which was missing the removeConnector call to propagate to the est of the broker chain +Pending review from who introduced the feature with +We can just remove it +Could you please tell whether you see this problem consistently +and in rupdated +If you want the original patch it can be found via the link you posted aboveSince youve mentioned it how would one go about adding themselves to the? Ive only seen entries with one author and it might be useful to know what the expected format is for the future +Committed to branch +If retval is a instance of and the text is ChineseThe text will not be decoded correctlyBecause the text is with encoding null +Ok thanks Alexey I thought I would add it anyway just in case +Havent looked into if this is possibleI suggest that it should be possible to set a locale on the class annotation so all the fields etc +from me I thought this had gone in already +Also the job conf is being set twice +Ill try itOK that was easy and yup it worked +Juergen +I know about this one always thought its just an annoying entry in the log file +This is to prevent introducing new bugs since we are trying to stabilize the server at this point for the +Thanks Bill and patch committed breaks +Submitted a fix today +Patch w failing test case for and +There wont be a release +I dont recall ever having to do this +Thanks +Generally guarantees of immutability has use cases in the legalSEC environments +on the hbase version bump +svn ci Deleting demoscommandNOTICEDeleting demoscommandDeleting demoscommandsrcCommitted revision +Its priority is already trivial. +Thanks for the report your suggested fix worked fix will be in +Mind resubmitting w please Gregory and then doing submit patch to try it against hadoopqa +I understand +Can u put patch on rb +If its really safe and easy Ill submit a patch + ctor has the same as well +Closed upon release of Hadoop. +Why you didnt use the forums to discuss this first you would have saved us a JIRATheres a which is defaulted to seconds +It works fine for me +Revision attemps to fix this provides a property switch default value for this property is this case poms are just readrewriten by modifying theirs groupId artifactId version partent and modules preserved poms are treated like normal text files and suffer a complete modification of their contentBrian can you please review if that meet your needs and close the issue or comment +Serializing a Configuration only to make it persistent certainly does not make much sense +Yes this is the same issue as +bulk move of all unresolved issues from to +Diagnosis repaired. +Patch applied +moved to streaming package havent renamed things because corresponds to STREAMINITIATE verb +Does Apache have any control over the apache organization on or should I contact? We do control the apache account on Github though the mirroring setup is managed by the Github team +class file into my jar +Looks this has been resolved. +looks good to me +If you want to remove it maybe it would be best to do it in trunk then I can merge the change back to the branch +Submit patch is purely a workflow state its confusing I know +Duplicates in JBT +Done +We could add but maybe leave it out for now +Please verify using latest Mb builds +commit dccecbddcdecbadca +So Peers must not leave barrier until synchronization is complete +I would prefer not to rely on a shaded impl into our jar system +The timestamp nanoseconds gets truncated to milliseconds +I think this is due to a bug in +Hadrian could you take a look at this? Could be something related to the release process +Its easy to add stuff +Parent job build must be merged with target platform build +This would be nice to fix but I dont think its critical to get into EAP or AS +Since nobody commented or reopened the issue Im now closing it. +Look at the timestamp bad habit from hadoop core +I have verified that the patch works on the code with tomcatx +Fixed on revision +We save notification failures into a list stored in the mbean and thus exposed through +Ive cleaned up some of the noise and added a bundle for Xerces as well +bq +Could you please attach your configuration details information +the file and then trying to use the delete button +I will open a separate issue for looking into threadsafe requires +Thanks for reporting and many thanks for the patch John. +I attached the sources for now since I dont have a workspace setup for Apache Flex in order to export a patch +Webflow has dynamic lookup for referenced beans so removing the def cardService DI field fixes the problem. +Can you please provide a test case for us to replay your problem? Otherwise were not able to help you explain how you set up your project? Have you started by generating the code from the xsd then created a mapping file and now use the generated sources and the mapping fileHow does the code sequence of marshaller initialization and execution look like +Ready for QE +if one DWPT can not flush why would be continue flushing the remaining ones +Thanks. +yet but that will be coming in the RC phase +Thanks for checking Grid +After a few days Im still getting the same messages +For problem these files need to be corrected by somebody who knows Brazilian PortugueseptBRadminguideptBRgetstartptBRrefI checked all the other files and they looked good +Thanks +The new extension is Ive also enhanced the clustering to remember what was deployed so now it effectively denies multiple deployments and also allows bettersafer undeploy of them +I think we are confusing with shipped jars. +Ready for review +Yes you where absolutely right +Got itI am compiling my client with jdk but running it with jdkAs for now MTOM cannot be enabled as we do have various enabled clients using the serviceIs there some documentation on which jars are required for a clientThanks for your help bug is due to usage of wrong jars +Youre replacing subclassing with a switch statement +Thanks for finding and fixing this +Also some modifications are needed to the junit framework +This is version of +latest JBDS doesnt have UML in it so this defect may be fixed by Beta version has the right Eclipse Bits so this exception is NOT reproduceable +The fix looks good +Good catch +Now it always returns an empty list but no information about what deletes completed failed +I added support fro STARTINGSTOPPING events +Patch applied at r please verify whether it is fixed as you expected +Do you have any idea what might be happening in testsample and testsample? Similar issue with file sorting +Im doing some really unusual stuff and this very well could have been a user error. +A fix was made for that issue that is the same strategy as the fix provided here convert a to a String before lookup but the code is slightly differently structured +For this jira the affects version was set to but wo a target version then the affects version was unset then the fix version was set to when it was committed then the fix version was changed today so it was hard to see where it was goingIf you mean it is hard to understand what the intended target is I sort of see that +For example we dont include though Karaf supports itWere still planning on enhancing the features maven plugin to allow an easy creation of custom distributions +Applied +They we TO both my gmail account and the dev list +No tests just moving options to config file from looks good +of course already closed +Are you including in your project +The geotools issues Ive just linked is a likely cause for the problem reported +I forgot to mention This is around the of case is previous comment was for a different unit tests pass +I agree this seems helpful +As soon as the patchesland in the CVS code tree I suggest we call CVS commit freeze and hand mattersover to Jandalf to take care of long promised releaseOlegPS Do I sound too impatient? +due to stale data in theory each thread could see out of sync versions of some of the data structures +Yes +Initial authentication screen using IE +One additional use case this could also be used to prevent binding of that are not useful +Username donatasc +The patch has some changes in and and in files like which dont seem to be intended +it seems the logic I built in is not working all issues which got resolved in M +unit test attached +Same need on the Ant task side +Because the code goes through a check permission with access denied it appears the code is going through a priviledged block +How to Remove the name of the jar file from or the jar file which is +really since will no longer be shown in shell as much since it will no longer be in +Patch Applied +I reviewed the patch +Heres a subclass of the Servlet which uses the Spring app context to get the instance so it can use the same one as the would just like to add two things +Code here invokes a lookup method that returns an object of type I where the actual concrete type is A +MichaelIm this issue since well be supporting through the life cycle +I took the liberty to reassign the JIRA to me +Awaiting respin of installers from trunk +. +Committed and published +However the Developers Guide says this in the section titled Accessing databases from the classpathIn most cases you access databases from the file system +Doap added to svn and is now included in the project listing. +Hence it does not require a package management system to operate if the customer choose to use tarball only +I see this test failing on Hadoop We need to backport this patch to branch +The patch removes unused code +I definitely dont think thats a safe assumption to make so I agree we should abort if there is an exception on the application of an looks straightforward +I have a little cleanup to do on the indexing one then I will repost +my motivation of selenium single test job was the case when developer is working on some single issue +Please consider these issues together when trying to nail this one down +Arun Murthy provides which fix a large amount of failures related to and counters +This makes obsolete +This is a duplicate of that you can check as there seems to be a workaround +I guess the worst case is if there is some bug causing the istat daemon to always fail its unit of work +Any pause whether it is a kernel pause GC pause safepoint pause etc +Ah I see thats awesome thanks had no idea +In other words when the redirect chain is broken the list of blocked locations is reset +Are you suggesting that this wizard only support +From the stacktrace it is occurring in a user query and not during the load of a mat view +These files have already been added to the svn ignore list as they are not to be under version control +Also pressing tab or an extra will navigate you to outside of the closing curly +Attached patch fixes the build error +I just basically added the splashscreen bean to the context and called the +Have either of you been able to verify this +Last patch missed a few patch +Any objection with this patch? Otherwise I will commit this +Thanks JM +BTW I noticed this just changed the home page and not the services page +remove sortign attributes from general header facet to the column itself and it becames sortControl +I did a new snapshot build a few days ago and have not seen it happen again +When I set rollbackOnly on a transaction the data is still persisted +Alternatively you could have decide the checkpoint and use the checkpoint call to simply learn to what point the ledger storage has been checkpointed. +We need to change how this behaves +Committed this one to trunk + for the patch +Tests are currently failing on this so we need to bring it up to speed for the release. +bq +Cluster should be manually specified +I committed this to the branch even though we have the failing tests with negative scores i think it will prevent the patch from becoming hellacious +I think the patch reflects what the doc describesMeaning the direction is goodRPC spec mentions in several places +Could you check against the trunk +I vote for a commit on trunk and the branch +Once the guys have a fixed release out well evaluate getting TB to turn invokedynamic back on. +Added release notes +A good lesson not to deviate from the spec if nothing else + committing this +Santosh Can you please attach your patch when its ready +Simon I believe that detection already disabled for all javadocs +It is intended for illustrative purposes only +Well need to figure out which machine we can run the cron on for the planet as well and figure out how to copy the stuff over to publish itYes we also need to figure out how to point people who are currently using thanks would be willing to help with this +Thanks OleCommitted revision +They provided some reasonable implementation that seems to work for everyone and they dont expose ways to tune it +But that leads to a fairly slippery slope of throwing more and more process environment into the config +The NPE was fixed but now there is another problem with inference see +The is to prevent content assist processor updatingacquiring when the editor is closing +Made some debugging on first page total rows is on second page returned value might be the problem +The patch has NLS for debug messages +The connectivity in Vancouver was terrible enough that we didnt have to do this +Tedwill update spacingUmawill update On the assignment not being able to throw an exception I agree +Tried this CF out and it works in the ESB +As I said I can not reproduce it I did add a local test with an empty form payload posted and got no exceptionYou are seeing NPE but I need to see more details because adding an NPE guard would simply hide an issue existing elsewhereCan you help me with reproducing it? Catch a TCP trace and show the method signature and may be we can do something otherwise Id need to close it as Cant Reproducethanks +Merged to the branch revision . +I havent looked at the patch but can we make signalTask additional to the existing +A square matrix is fine with me +I suspect we need to change each null for a private method that checks for null AND empty +Removes all tinkering w retries +i reviewed the latest patch and it looks good to me +New patch implementing Aarons recommended changes +Actually I developed on the problem for my own and found a solution by extracting and using the inner Class of the Package that was implemented in the +Here is the new patch +Hi MarcioI think its better if this goes first and then the docs one cause doc is a reserved name as well and would be a nicer complete commit at +Usually a patch p should do it +I would like Alex to review this list to make sure there is nothing that really should be fixed to help portal +Ive finally made the changes suggested and discussed on the mailing listOn resolvers allownomd is now descriptorrequired optionalOn buildlist skipbuildwithoutivy is now onskip fail warn tail head +Closing issue fixed in released version. +rewrote description for dataList +I modified the test to allow junit to handle the exception as this is the Harmony project convention +Adding plugin descriptions +In that case the current way seems fine to me as long as its documented especially with +Therefore only the default permission template will be present at first +Let me know if not. +Document it in the SI docs? I think Ill stick with a poller for now +Now it is possible to resolve itSome important patches for other issues should be updatedWell update these patches +Everything else is the same. +Not sure if ios has set this to some default now +Patch also fixing validate to validate modules tools directories +Those html files predate the user manual +Is this a new framework for executing tests or just a way of having junit tests that test replicationFrom it seems it is the latter which is goodAny reason why the tests need to start with no security manager set as a property instead of using the no security manager decorator +I see no harm documenting that a Codec does not support index backwards. +Other people prefer to start a L and increment each time the public interface changes +Patched plugin is available prepared to be part of next out to the production +Then we can remove the class from and cutover all uses to this new one + patch +Unchanged copy of Modified copy of names with PATCH on the end look a bit ugly and imply the file is changed +I think we should fix it for GA unless fix is not simple ans Slava +Reading the doc I had hard time translating this to functionality will be removedThe explicit way to state this is to say mechanism will be deprecatedremoved +The point of having this test in is to not to say the test passes so the bug doesnt exist but rather to say the test fails therefore the bug exists +fixed an issue related to SSH authentication failures fixed a bug in the display of servers it compiles with java only +I saw the same error when I run the test of cameinaction appendC without clean up the directory yesterday +latest TCK is working shipped in +Fixed the anchors are now correctly updated when opening a drools process file. +since i never selected anything +There is now a onPrepare option +Patch for lsr branch also committed thanks Dmitriy! +Ill go ahead and check this in +the jira had been backported into EAP ER +It does not handle failed authentication gracefully +Patch which reformats most of the code moves over to Java loops sets appropriate visibilities on fields makes proper constants and does some genera tidyReady to go +portions of this were committed under also by You always seem willing to test some stuff can you verify that the runtime and server work? On my machine environment Im now running into really weird errors but Im running into them for and as well so Im wondering if this is just my environment or if the server adapters are in seems to be a typo see added the following lines to within AS Application Server I can configure the server when creating an EJB module or a web module the are not available +Connection A is still open ij show connectionsCONNA jdbcderbylocalhosttestdbcreatetrueNo current connection Set connection to connection A ij set connection connA Try a sql ij select from tERROR A network protocol error was encountered and the connection has been terminated the requested command encountered an unarchitected and condition for which there was no architected messageij +push to +I will verify that scenario once again as part of +to unbreak it but we should also figure out why exactly this happened +Committed the patch to the branchThank you Ivan +Can we do this in a short time period? Thoughts +screen shot of eclipse id what I see +Any update +Sounds very good and thanks for the update I would suggest starting a discussion in the developer forum we actually also have another related issue regarding lack of configuration for this http interface so we can maybe cover both there +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +It seems to me that optimizing this is extremely unlikely to make a dent in overall performance +I am working on this issue +This can potentially cause the log corruption described in this patch Logs append API truncates the to validBytes before appending its backing byte buffer to the + +Tara nope I cannot the data Im playing with is not to be published to the wide public +Looks like a duplicate of +Verified in ER +You can do this using the where you can combine your predicates +All other jar files in directory contain signing information except forQE will continue in testing after all jar will contain proper signing information +Ive started working on a patch +I have moved this to CR and will create for each lib + is related +Sorry I have been caught up in other development for the past couple of months +I think this one is fixed and it is included in the release notesThe code reuse patch is not included yet as I do not think it is related to this issue +Committed WIP code to trunk in Implemented in trunk +I have updated the patch +For yes Ive run the tests against different commits and appeared as the guilty + is null in the offending situation +We introduce no additional escaping cases on top of the escaping syntax defined for the policy file +should be great +Will post to RB in morn after I give it another run though + +Remarkable difference from previous patch Remove copyright notice from license header in modified files +This patch implements most of what has been requested for +There is a separate issues here which needs to be resolved in documentation +If i face with this issue again i will attach the test case +Would it be helpful to post them or wait till some of them are fixed and try again +JBDS was installed in new directory +Thank you +Update the patch to adapt to the latest code +Updated patch it moves all changes to and doesnt change the API +on fixing the issue with Trash +Even if there are no wires it should be an empty list +What you did falls in the category of New Features and therefore should have its own JIRA +Thoughts +Ill try and reproduce when I get home in a bit +Code looks good +ThanksIve erased needless one +Also thanx for the follow up email address jean +Please include the JIRA messages at least of the action for CDI happens in JIRAWhen you add JIRA search we could always filter out JIRA messages then and reindex right +All fixed! +committed to trunk +Committed to sqoop +We can remove it +also improves the performance of the tests +Thanks Emmanuel really appreciated. +Does anyone else have an opinionbq +The alias group may be misinterpreted as a reserved word +And the Bundle is getting deployed as the last bundle +Bumping to M as this shouldnt affect M at all and isnt in the JBDS Core editor no longer being devd +Reverted and with proper message +See comment at on how to properly configure test suite in our EAP productization branch +I successfully installed Apache Directory Studio on the latest Eclipse version the day I released the new versionLooking at your screenshot some plugins seem to be missing indeedCould you provide us your log fileAlso could you give it a try on a Classic Eclipse versionThanksWhere would I find the log file you want met to attachMike +StackSee my comment Sep Version combines the changes from +We should get this in lest some other property get introduced which again violates this +If the handlers exist after the dispatch phase they will be invoked only for the engaged services. +See for followup on increasing the transactionCapacity default values. +I have added a note in the faq. +I looked at it and I dont quite see the difference between the setgetHeader and the setgetInclude method? Also on the save I see how you include the stuff into the writing but I dont see how it loads? Are you sure we need this instead of using the include stuff +Help in resolution is appreciated. +Thats good that it turned out to be the same issue + +I even filled some strings into field Age and got the exception as above but it means you cannot put string into int variable But on the user side there is a nice error message see the attachment +Added. +done +It becomes impossible to meet the audit requirements of EAP if you allow changes to be made that are never persisted and it becomes impossible to rollback changes as well. +Then it ensures that the proper is added regardless of whether one exists on the project already +Thats pretty clever guys +is that what you want to worry about any time you are creating a componentif you dont want your images to load synchronously then instead of using lazy ajax panels add markup that points to a shared resource + Copy different versions into filesystem locations +you can install from update center in latest daily builds it will be part of update center as well. +Since sonar takes care of the data migration after upgrade I dont really know where to look to fix the problem +Thanks for the bug Victor +Ive been very tempted +Just double checked it is spinning at CPU. +I agree that getting synchronization to avoid blocking is of primary importance +Sorry for lagging on this +ack +Hi Todd Michaelhave you found some time to look over my issue +Updated the patch to be against the mapreduce svn tree now that its split +Itd be better if they caught this event and shut down relatively cleanly +If this is referring to the Graphical Installer then I can easily add a note to Chapter +Apache JIRA is for tracking the development of confirmed bugs and feature requests. +should help with the more general issue of client recovering on cluster restart +looks ok to on and trunk. +has this issue too +Itll handled later. +Sometimes it is better or worse by +I planned to use the version check in readExternal at a later stage to be able to handle a heterogenous collection of Derby version code +Simple fix for the issue +Will pick this back up for. +I added a test with two recipients and also one message is received then by the server since only one message is sent to it +This makes sure the version of the Apache Felix Http Jetty Bundle is included which also supports forwarding the Http Session events to the +Would be very gentle if someone could commit this still for the not I guess no biggie +unfortunately the way the branch and the does things have diverged a lot +I looked at the proxyREADME today and some things were unclear to me +OK Ill do it +Hi Bill +I expect it to be done in a few days +Tests are working after the change of the remoting version +The attached patch adds a few snippets of code to the base class which interns all namespace strings that are passed +contributed by Ravi Palacherla +So in this case its the producer that has an issue as well +Please help me out with issue as soon as am new user this is very frustrating effort it consumed my Hrs or more +The problem was really the! +one is full WAR which included all the jars in WARlib the other is the minimum WAR that did not include any jarsby default we generate the full WAR Ive tested this WAR against TOMCAT and JBOSS both worked fine just simply copy the full WAR to the deploy folder +No client code should be able to detect the difference between the versions before and after +Olivier I removed you patch from the because it broke the site plugin for people +Its already been deleted +Thanks Ravi for working on this +I get the same error on this full version JRE IBM Windows build works on thisjava full version ll see what we can do as others have reported it +Please use the associated forums thread for discussions +Please fix those before committing + ignore also have encountered this problem +Experiencing the same issue running standalone sonar analysis on Jenkins +Clearly until a more well defined use case can be provided this issue will go nowhere +Explanation seems fine to me +remove +This message is for V patchAfter some tries I find it is not easy to align Loop Pealing with DABCE and it needs more work in following passes after peeling +Ive just moved the configs into in +Use to get stack trace +This schema should recreate the indicated by the user +trivial in +Nevertheless ive fixed it so that this should no longer be thrown +Oh + looks revision . +I committed a simple version of this in revision to make my backwards compatibility tests in sandbox work +The fix should be relatively simple but wont yield the exact same behavior for the list of docs as it does for the list of view keys +Oh got it now we would need to do away with a standard Writable implementation since it wouldnt read everything from the same streamBut then every vertex computation involves some lookups even in the standard implementation +Is there a reason you have selected +A new property in the would be added which referenced this directory +A fix for this was merged in at some point after this bug was last updated +Your patch seems to be reversed +Rob this should probably be part of your AS core reorg. +My guess would be somebody already deployed an artifact with this groupId artifactId and version +I think you are not using the latest revision +It addresses all of the outstanding tasks besides integration with the split brain fix +ok yes I see it now thanks +The matrix is really exactly rank regardless of the exact values of the elements as there are only two different rowsThese rows are at indices and +Added link to the related issue. +At revision +Can you file new JIRA issues for the save functionality and for the workflowtask edges +Lets leave the properties out +Thanks very much! +You dont even need that thread if you dont need deferred log flush +but it does trigger an assertion if assertions are enabled a sequence number will get skippedI do not understand how this would happen if the duplicate packet also gets sent to the pipeline +OK Ive committed a test case that will fail if this condition happens again at least. +I think I got them all out with XML attachments +With this master zk and META are still shared across different groups with little isolation guarantees or +If I cant do it Ill send you the jars by email +is there an eta for the release that will contain this fix +Well wait until to support +Is that correct +The problematic patch was reverted and a new patch was committed in. +Im ready to resolve this issueWe may want to consider a timeout when waiting for the process but then again we are expecting the process to terminate normallyKeeping this open for a little longer to see if anything comes upPatch a didnt help with Myrnas cascading errors issue +Just realized I should have used a object in the test +Yes that sounds like a good idea +There is an in srcjava that is correct +I would like to contribute to myfaces but I must admit I have not yet been diving into the myfaces code so it would require an additional effortThanks for the suggestion that will work of course +But they are not related to the +Porting it to YARNtrunk +I changed it this morning on the train +Concurrent reading threads needs to be testedEach element consists of ints +Seems to work for me but will delay committing until after debugger demo +So basically the problem is that its evaluating two expression languages JSTL OGNL +But I didnt change the jdk for projects my set the compiler settings to use java and now it Snjezana +Subsequent runs worked +If you just cat devurandom devnull it will eat entropy +I suspect if you take Spring Web Flow out of your app the error will remain +This is a table describing the different configuration properties for the JBR gateway +This should be documented as an FAQ. +do you have a reference to the JDK bug that will not timeout idle threads +Hey Tom +Shaz do you think this is a issue +Fix confirmed working for all reprs under Python and +Resolving as Wont fix based on comments +mvn cleanmvn compile +Committed revision . +Well just have to test this +Fixed perm on all RHEL zips will be in CR +that there should be a routine for which we need the google map +What error do you get +We should perhaps have all these exception extend some class to make this more clear +And AFAIK they are also the only ones available so the issue can be closed +The following property files are neededkind regardsbob +I think this case will become more common as more utility libraries will be packaged aware +Weird when I added that annotation in Eclipse and Ant complained +Alexei please apply these patches +Could you create an issue for this interesting feature pleasureThank read through this bug and wondered why I ended up here because we are indeed talking about two different I remembered I ended up here because of issues which seems to be what I am talking about and I dont see why it was linked to this issue +In you are using eager am trying to avoid lazy loading of collections is STILL FAILING as mentioned above. +also like to change the name of this tool to Label Manager and implements some new features but I didnt know when I get the time to complete those steps change the name from Labels Info to Label Manager add the standard Find collapsileexpand features to this new tool add the possibility to add new labels directly from Label Manager add the possibility to write the new inserted label into the corresponding xml file check if its possible to see the new inserted labels without restarting Thanks for your good ideas Wegner submited a tool to sort labels in +I was just pointing out it is a slippery slope +I had to take some liberties with the JSR style interface to be able to enabledisable and reset statsitics in a consistent manner between different containersAaron Please take a look and let me know what you think +agree with Harsh no need for a config +the latter maps to a enumeration in the +Updated +Thus a implementation is to create an Ant filter token for this +Ive asked him to make some changes in it +Yep I in our configuration we are not using +Created a new that uses the new Journal API +Everything is fine +I guess the easiest solution would be to turn handle +As long as the client vm synlinkactual cert file is correct then connections work or if the cacerts file does not contain an appropriate public key then youll see a somewhat better messagePKIX path building failed unable to find valid certification path to requested Ramesh thanks a lot for your comment sorry for verify so late the link you attach is useful acturally its caused by the wrong config of we put the new one to server and it turns to noraml again thanks a lot. +fix for branch At revision fix for branch At revision + +We definitely need another one then +Patch applied in revision + +proposed settled on +you for the contribution Owen! I have committed this to branch. +With a big change in code there is a lot of risk of introducing new bugs especially with something as complex as the oombo box and menus. +ok well apply this asap +bumped to good point thanks but spy wont tell me anyhing about the window +KeithIs this fixed alreadythanks the to Axis +Castor would just skip the binding if the syntax is not correct +We can then when the documentation has been added close this ticket +As soon as I get update from Sun I will post it here +This issue can be closed. +is still under development +Ill create a pull request shortly +I got the put option working now its returning and the data is in the table now so thats goodBut to use the put option you have to save the data to a file before putting the data via php curlThats an extra step on inserting data in to the tables I would like to skip and use post option +the process itself is explained in file +If nothing is there then no impactPls provide your comments on this i can prepare for trunk too if this is fine +A con is that it isnt something anyone else has worked through or made work +Increasing the size of the queue would significantly reduce the probability of getting an exceptionCheersDave +I cannot reproduce it anymore but I echoed them outLDLIBRARYPATH homeuadminuimacpplibhomeuadminuimacppexamplestutorialsrcDYLDLIBRARYPATH homeuadminuimacpplibhomeuadminuimacppexamplestutorialsrc +which made me think about the o in previous versions +Ive reviewed my patch and attached updated one as the last one seems to be little bit crappy +The patch doesnt change all lines +url added to cli added needed files for your review +It has not had heavy testing my main concern is to verify that memory reclamation is fast enough +The first version of in public repo can be +But if we dont then we dont try to force natural sorting on the underlying data right +Changes look good +Patch adding interface and replacing the strings in the classes accessing the parameters +Rebased patch is attached +This patch is NOT commit ready +GaryYeah maybe a multiplexed background thread could run and acknowledge the messages if the consumer has been inactive for a periodI wonder if we can use a single thread for all optimized ack consumers I would assume a thread per consumer would bee too much +If delete the file the editor isnt because the file path was changed but the is the old one when the editor save will happen +sladYeah if we can get the attachment file name to appear as nice readable instead of cid then that would be greatIn terms of unique global it may or may not be a bit overkill +trunk +Yes package and protected scopes are supported as options but currently ignored +Do we really need to include a JDBC reference section? Would it not be better to just refer to the javadoc of the JDBC specification? In addition a table showing which methods are supported and a section calling out any Derby specific issues would be neededIf we still want to include a JDBC reference section there should not be different sections for JDBC and +Uploading a trivial patch +I received this from from the eclipse SMILA project +log filtering on +Although not all our issues regarding graduation have been addressed there is a status file now so this one could also be closed agree +test case +Shirshanka is working on this and will provide a patch +Tested on Postgres and Derby +Army Ill look at the html file and work it into DITA today +great finish to a great GSOC project thanks Andrei +How about rewriting to instead of +I was mistaken on the behaviour of the informalformal parametersThanks Norbert +Thoughts anyone +I will test the concurrency of this solution +Ah yes youre right we reorganized the whole project and it seems that we forgot to adapt this pomI just fixed it in svnThanks Brett + +Hi AaronIts probably better to bring this up as an email on to see what people think +to add documentation and fix the test to verify that a leader in fact exercises its role as a leader +Raised. +We can do that migration laterLooks like you are already splitting the MR changes +When we bite the bullet it normally involves something difficult andor example This project needs to be finished by tomorrow so Ill have to bite the bullet work all night +Perhaps could have found all similar cases and arrived at the solution +Thanks Xuefu! +Sounds good +em hence our test fails again +I file a bug report pointing out that a part of openmeetings is not working for me as it should +nit from existing code comments in cmd file speak about sh fileDo you want to respond to s comment +HiI am attaching the xhtml and bean let me know if anyting else is test and and RegardsI have attached the requiredThanks a lot for the is working fineThanks and RegardsAmit +Oozie is fucked screwed +There is a side effect as the links are no more updated at all even on latest snapshot +on the idea of tearing out the current CLI and replacing it with a dynamic language shell possibly with a blessed Thrift wrapper client imported +Closing the resolved issue. +the problem with modifying the view name is that it complicates the use of the frameworks or any view resolution you elaborate? Are you referring to the fact that mobile views may be introduced gradually and that there may not be a mobile view for every view? Or something else? As I mentioned before this only works for view technologies that can actually check if the view exists +Its quite same as Cedrics last sample +The patch uses a deprecated api +Im attaching the updates to the +If there are still issues with this functionality after the release please with details. +Looks just committed this to TRUNK +Agreed theres not enough info in this one to understand +HiThe workaround works only when you are using fo filesCould you please provide me the solution for situation when Im trying to generate based on xml file eqfop xml xsl pdfregards the author of the XSLT file you are responsible for creating an XSLT that meets your needs +Atleast you must have had a reason to use a lock for search in the first placewhich I guess is to prevent indexing in between +Any thoughts on how we can go about doing this Mikhail + +In the case of acksqueue I have replaced the data structure we were using from to +If there is no value in config still you retained the same behaviourChanges looks fine for me +Thank you Mostafa +For some space reason it also diff on two other different language for getstart manual +Latest patch for trunk and +Not sure whats up there +Fixed in commit edcaccfcccdb +has changed quite a lot with the codec stuff +I guess Ill have to switch back to an older version until this is fixed with documentation or by undoing the deprecation warning +the security manager and JMX but it seems that any lack of permissions is just silently ignored by default at runtime +Btw I think removing initial score arguments and merging scores in is a good idea overall +Ive instead added a link to our plugin guide. +All tests have passed now on platforms +However there is no test for it yet so I am leaving the issue open +There are still more problems with this +Ive committed this to trunk and. +There are several test suites outside of the project that might be affected +The patch changes the code to do a traversal of the DOM tree limited to nodes only in the given scopeNode +Is it still valid or does it for us +The bug probably is burried around the method +I think we should remove this assumption +Note that in JBDS it works which version of the SOA tools did you install +It is a simple patch but it allow adding and removing of search servers on the fly without having to shut down the search website +Thanks. +About time make way for +Also updated my geocouch branch to use the new priv convention +to the code changes +Updated with latest changes +on JBT. +Committed revision +before pressing the pressing the sounds strange that you cant reproduce the bugI checked it in different computers and the result was everywhere the sameThere are no any tricks with reproducing just press the button +I think this issue requires failure injection to be tested +Also it would be a great idea to add scroll for the dataTable this is very useful for a large set of data +v just changes a comment not the code itself +The new JIRA has been setup in existing issues from the platform has been migrated but we still need to transfer the rest of issues from ANSTO internal JIRA to codehaus manually. +Would be perfect if will implemented patchfor this issue too +It works. +Added test for ignore these attachments Ive made a mistakeIll try to create a correct testcase and upload test files please ignore previous zip +jar in liboptional as not needed by Camel +But you can check out as soon as you get the unit test reproducing this with RR. +Closing this issue since cant get rid of lib directory until can fully switch over to new jboss build +Which version you are using +Reopening based on recent comments +Provide a better message +get the v of the patch +See patch attached to patch available as part of on fixing the version info for and all affected issues have in all issues not touched in a month to move issues to and +Some more details +So this really is a bad idea +This is a major issue for us +Ive looked but not been able to find a way to move an issue between projects +I added itfirst because I thought I had an error requiring it but that turned to be something else but I kept it mostly as a documentary thing +Please check if patch in has fixed the bug +Patch +If write access to JNDI could be optionally limited to clients that would be perfect +There are some wiki or documentation pages which should also be updated +Also include the status option in the Usage section +Fixes and to update before generating patch +as is already open +Actually Ive fixed Iterable to behave like a List meaning Option andStrings might be beneficial to the developer but not to the maintainer +I think if we use object creation we should use same strategy for default constructor and constructor with arguments +Ive created a branch called to allow us to iterate on connector code changes +I think that would be a better way to implement this with the issue on our side of overlap between Collections and Lang +hmm sounds reasonable but isnt what we do now what doesWhat is the link +Any IE version on Windows XP +Is he still in this project? Do we have to officially support Windows right now? Until we get more active developers maybe we just have to focus on only one platform +And also implicitly commit the backup transaction at the end of the backup +Tested! +Also the socket needs a timeout for reads on the obtained from it but I guess there is a JIRA for that already +The changed but unsaved domain object does not have the last implicit save do we change the JIRA to GRAILSPLUGIN? I know that on my own JIRA issues I did not have the ability to edit the JIRA once posted +Committed to long back. +New class confirming to new API is needed +Doing more investigation into this issue I found the problem is within +Hi All Can one you update me on this issue this is very critical for us +m + +Determine if these are already tracked and enter new issues as they are determined +i make workaround fix +Writing tests all the way down to the wire is going to further delay this patch +Hope this is a right way +Waiting on getting the website updatedWill send out the announcement emails to dev user asap +Yeh the finalizer thing is pretty naf I suggested it because that where the current shutdown residesI had also considered that if the pool were exposed perhaps via a new eventlistener interface then it would be possible to register a listener to catch the shutdown event and programaticaly do whatever you like on shutdown +A few moths after I did I did a much better implementation using Compass +It changes things to use proper only types etc +With hard upgrade the disposables get dropped properly and index statistics does not kick in +Damn! Sorry for that Ill look into it today. +otherwise please commit if tests pass +In the meantime you can try Roo from sources and provide feedback +Seems good to me. +Moreover the NP exception was thrown while creating an invalid parameter exception object +how do i fix this +So this is not a AS specific problem +seems to be broken +Thanks Namit! +This issue not any more relevant for current trunk that was replaced by branch +This bumps the last digit of the version id after generating the beta +this is on the clientthe logj isDEBUGThis is new in the last few dozen commits it didnt use to spew the only difference is grabbing your new client code +will store any exception into that ignored future and thus the thing will stall +Currently the code just looks at the trailing http +bQ +fixed in revision is actually a bug in Netty trustin is aware of it and AIUI there should be a fix in Netty for it spoke to Trustin about it and he says he changed it because of some performance reason and it isnt a bug +Derive is probably the wrong description of what youre doing +With this update the need to update documentation is less strong +Line A semaphore lock is acquired outside of a block so if any exceptions are thrown the lock will not be releasedLine A semaphore is unlocked outside of a block so if any exceptions are thrown the transaction is not resumed +Mass closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +but its still very much work in progress +Thx. +patch file showing the main ideas add javadocs to clarifying how its different from add javadocs to document that if subclasses are then they need to do something useful with Similarities make complain if there are similarities configured but the global similarity factory isnt new test for hte new error checking in fix existing tests that had were configuring per fieldtype similarities that were being ignored simplify example and link to wiki for more details +Changes in the the new xsd Creation of file should be in else block +An exception is thrown on the closed socket and then the client hangs +What about andDo these files fit in that scope or should we open another issue +Other than that interesting findings +Here is a version of the dihwriter patch compatible with todays entity patch update +Actually I dont know what youre talking about +Im on removing delete family at least at this point +So as Kurt was saying the UI should not allow business keys to be created without the right format +Are we set for the upgrade +Please commit the patch +thanks +I agree Ive just posted the patch it would be great if you could post some results with this one too +Closing resolved issues whose fix targets have already been released. +Unless theres a tested patch supplied this will be considered for a release after +Thank you for the review and the commentsThe attached patch should cover the requested changes +If not please feel free to disregard it +If possible the script could help to register the Geronimo as a service depend on the platform so that it could while the OS is bootingDoes anyone have other comments +next week sounds good enough Andrea thanks +I reran with the tests and it passed +Wont this still be an issue for jenkins runs because even loopback addresses are blackholed +You can list Texeltek as my org +Applied patch as proposed above and committed changes +Im not sure how to best incorporate this information into the page as it is +on the approach +Unassigning until there is a newer release to consider +Linking to +What is an incremental snapshot? If youre trying to make things more complicated by not linking sstables in new snapshots dont +did you correct the bug in the repair method that I proposed aboveYes +Also the popup messages prompt for the forms internal variable names instead of the display name +Yeah sorry I couldnt figure out a quick and easy way to test the method in question +Closing resolved issues for already released versions +This should be pulled prior to the pull request for +Small patch to fix that false positive +Why would another MR job be needed? a similarly simple change to the reader code works +The addDocument call is necessary to specify the transforms that are required on the reference elementColm +news on this? This is still broken +Is it really necessary to try to remove them right now +if the first attempt is SUCCESSFUL and second attempt too is SUCCESSFUL then it marks the first attempt as OBSOLETE and updates the tasks successful attempt to the second one +Musachy Im not sure which version youre referring to as +Foundations of test classes added. +Release is different case in this case the limitation is on the whole row length and requires database instance configuration at the level of command line of started instance +It isnt a big issueIs there any way for me to configure Karaf JAASPax WebJetty to not invoke the JAAS mechanismfor my servlet whilst keeping the JAAS mechanism for the Karaf console? If someone eventuallyneeds to monitor this system it would be nice to have some way to suppress the exceptionthanks againGareth +Hi I hope you enjoyed your vacationAre you able to say when you expect to take a look at this and my other open ticketsI believe that I currently have seven open issues but the good news is theres a pull request associated with each of them with these fixes my cluster is currently withstanding all of the stress that I can throw at it +Thank you for your sharing +Since there is no danger of XSS and the CSRF issues arent related to this ticket on merge +Will commit if tests pass + +Close after release +Reopened issue +thesame file is working fine most of time and some time it is failing +Added defaults for vmtypeid as to reference the default vmware type created my the install SQL and imageid as to reference the noimage image created my the install SQL +Fixed the deadlock and corrected OOME handling in the loopJNDI tests seems to pass now +Committed to trunk. +Trunk doesnt seem to need the patch due to where killed tasks dont cause waiting tasks to be incremented +Removing +LiborI talked to max and we decided to offer no facilty to the usecase where you create a key outside the deltacloud tools +Commenting a year later another option would be for the LICENSENOTICE file to explicitly list each of the artifacts and describe the items in each +To be more precise its possible to update branding plugin in order to be able to inject html before black header but plugin based on Footer extension point and thus all injections are dynamic and occurs when full page loaded +Thanks for report suggested patch +That latter part doesnt seem to be true in your hot redeploy scenario +The file is duplicated here as a workaround for this maven behaviorIf you have an alternate solutionworkaround for this problem feel free to this issue +I was just hoping for it to get a little more review +Backporting to be compatible with would be a serious undertaking +Committed to and +Stephen can you attach the patch +I think here we should abide by the SLD spec and thus stroke a polygon only if the SLD says I noticed tasmania borders and Spearfish restricted areas do have a stroke whilst in fact toppstates does not get stroked + +Scheduled for not because I expect the details described here to be done in but because the improved management functionality intended for should better handle operating on lists. +It would be great if the behavior you described would be mentioned in the s javadoc +Yes I had not checked the autoflushbuffersize option +Is there any way configure this property through castors API like XML Context? +I know that between and there were some changes in the way Axis uses objects +Im going to guess that the active core devs arent using this and the current asignee is as he just got a job at Google +Awesome work KarthikThanks to stack todd kannan and everyone else for all the time spent reviewing +tmp patch to make failure show more +thanks for the s comment illustrates once you permit timeouts you need to buffer query results before streaming them to clients +Applied to branch MATHThanks. +nobody said it was a solution but its a worthy workaround better than deleting and recreating any dont think it will be needed once the state cleanup is there on startup +No for all the actions the values from input fields are not propagated into the request +a plugin that wraps the lucenes summarizerIf everybody is ok i will implement this +I hope we have not been working on the same issue +Committed x revision I will now try in a second step to do a clean checkout of tests again to minimize changes made in backwards tests +How do I get around the findbugs issues +Applied the patch with a bit of cleanup +Dont want spring only. +Test as expected +If my memory is working right youre correct about how Struts picks a default LocaleWhy not do this with a context listener? It would be independent of Struts and you could use it in any web application +Committed trunk and +Fix the consumer iterator to synchronize the consumption of data with clearing the current data chunk and committing offsets +Committed with very minor changes. +changed the name of this generated file on purpose +Good the plexus dependency has been updated for this release which is where the standard I found printing the excludes useful for debugging but yes listing the ignored resources sounds useful +oops closed wrong going to get done for ve added a file created by a colleague of mine +the comment shows how this should be done and it can be applied to any kind of resource that is generated based on the mapping filesIll include it in the docs. +OS X is generating too many warnings to fix at this point in time +Attachment has been added with description Same document as original poster but doesnt break the infinite loop persisted on another document that had no list at all but just the block with the aforementioned attribute +OK in is all saajjar were removed +Please can you update the status of this Jira to indicate when it will be ready as we intend to release M on MondayThanks failing test passes with a timeout factor of see httpalbanyjob MikeI would definitely prefer a stable build which takes an extra minutes to run to a build that is prone to these intermittent failures +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +im not sure who knows resourcesnapshot well sudha can you assign it to someone else first +The proposed patch should fix the bugUma and Eli since you were able to reproduce the failure before see if you could get a chance to help verifying the patch by trying to reproduce the failure with the patch +Id like to see this as two separate message ids I am pretty sure the DES key and shared secret key are not the same and the second message includes the length of the key as a parameter which you have removed and which might be very valuable informationThanksDavid +It has been replaced and is now in the process of being synced again +The problem is not the upgrade right? The update statement above throws SQLSTATE as it self triggers to the max depth and thus is reporting the error correctly +Once for hadoop and once for hadoop +The attached patch is just the Facebook photos part without Picasa functionality +It will be nice to get rid of the exception we currently see when the GUI runner application is closedFurther I assume we dont need a read permission because the runner reads the file before the Derby harness has installed a security manager +Please update using the snapshot update site when you can and provide feedback if this is working for you +The reason for this is that it is expensive to setup the entire context such as the session and a request cycle +Heres a patch which adds a test case for this scenario +Committed revision +Also on my test at first we thought it might have been but this was tested on one node built out of source +Its required in order to set the home page +it might be that it is adding +Depending on how the test cluster handles the closing of a region the test may succeed or fail +Since we removed the Avro related code now resolving it as wont fix +In todays disk the highest throughput is MBsecond +It was a problem with how the processor keeps track of whether the parent of the current element being processed is an xiinclude element +The current script generates a package level annotation which is actually simpler and less error prone then the reading +Rename to isr Rename to isr Im not sure how useful it is to have a count for leaders and under replicated partitions +I think Josh was working on something but I havent seen a patch +Its Js thats patching the main class nameIts better to be liberal in what we accept I guess +Hi Manik Yesterday morning I deactivated the nd level cache to confirm that this issue is linked to Cache +Any progress on this? Would be nice to have the initial HA release support the balancer +Johno it should not +If there are plans is there a development version out there can be tested with +Issue is fixed +In the spirit of the previous example here is a sample wsdl file +u is still generating a wrapped StubChuck +The issues URL now points at the thread +added a tests locally except the ones in +Not many variables are explicitly passed across hadoophive +following jan suggestion renamed keepsendingchanges to keepsendingdbschanges to avoid confusion with couchhttpddbkeepsendingchanges +Code looks good +Thanks for the headsup Knut +Ive installed the whole Activiti environment last week so +OBE +Arquillian support is slotted for so moving this issue forward +If the blank and archetype were essentially the same application it would be easier to maintain both over time +I am still experimenting with different threshold values Vs this patch using different datasets +Instead of Cordova the iOSAndroid SDK could be used +As Dennis mentioned one solution is resolve property substitution in jgroups code simply remove the flawed implementation can make it works fine +Please feel free to reopen it if you disagree or if Ive misinterpreted your report of the issue. +I had read about the Java performance upgrades but did not expect them to be this the java version the change on fibj is insignificant but fibj drops to ms +This is used to add logging for the class for unexpected elementsApplied in Rev +forEach if and so on +and why none of our tests catch it But they do? There were a number of failures when the number of closesopens mismatched and failed the build +Sorry! I got confusedI committed this +lets create follow up issues +Should I raise a new issues for that +Could perhaps use this to implement consider whether to automatically switch to using NIO channels if the and are both instances of File Streams. +This issue has been fixed and released as part of release +Heres a patch that fixes the problem and improves the fix for to avoid calculating the default fault action twice +About failing tests i made a mistake in db configuration i used the same db both for our tests and its ok +The issue with the delay before was primarily caused by the notification being single threaded +I think I prefer the latter for explicit typing purposes +Thanks Ravi +What do you think Roshan +Committed +Im not too sure we should do this one +bq +As there is no issue remaining at this time I think everything should go quickly for nowPlease also note that I didnt manage to make version work by adding the dependency element in my +Thanks Chris and Eli. +I think saying we have three different ways of talking to via JS doesnt really make sense as a little library API could have different concerns from a larger UI framework such as the one Futon or Fauxton might want to use +Ill open similar jiras for hdfs and +Applied patch in r with few changes +This Axis C implementation +Ill make a trunk version for +Patch applied. +Ive just committed this +I am not sure if minimum block size is really required +I figured youd know better than me just wanted to check +Selectcreate any order +This patch is for viewing the stats and is exactly same as for webservice servlet servletStats +Both the calendar field and the textfield were correctly setPlease post a minimal webapp showing this bug and reopen the bug if you still have this problem +Find the patch attached to issue Markthis could occur one to several times per day under heavy loadwe use ok we will check whether the app really obtains connectios and then adjust somes configures for more teststhanks for your help. +will change to the above as seems cleaner +This one seems to have sneeked in on adding query caching for Criteria queries not needing recreation +I am agonizingly slow at understanding anything but quick to move anyway +Other errors to sort out though +George are you still setup to recreate the issue? Its been a while but wanted to see if you could verify the fix +It would be a security problem to run the distcp on the insecure cluster because you would need tokens for the secure cluster and they wouldnt be protected by the insecure clusterOf course that means youll need to use webhdfs or hdfs rpc to perform the write but that is far better than encouraging users to open up security holes in their secure clusters +Suggested attached for enabling webappUI to +Thanks JonIll try it out +So while I strongly doubt compression can have anything to do with that in any way I rerun the test against a bunch of time but I was still not able to reproduce any errorSince you seem to be able to reproduce easily would you mind sharing the scripts you use to reproduce +DirkThanks for updating this perhaps some docs on the wiki would help those of us not as familiar check it out +Yes +I realize this is premature but I wanted to bring up an Accumulo instance and happened to have hadoop running so I gave it a try +See attached for a patch to a new install +Fix an error dir name in test case +fix is included in the patch of. +This tempbuilds have many bug fixes and feature enhancements +addressed comment by Suresh +Style is a bit different from most project sites but that isnt all badI havent inspected the HTML yet +That is the version I am using as agree this may just be a glitch that is why I attached a screen snap +For example choices will disappear +We really need that demo app to verify though +Revision exit dialog is moved +To add further I dont think this is strictly necessary right now due to the fact that we require updates to hit all replicas before returning success or failure +defer all issues to +I am just one developer among others in this community project +Committed to trunk +However for scheduler I think that just a class name may be enough +That is not how things work +We have not been provided with enough information to reproduce this issue +Fixed on revision +not really sure like you said its hard to tell. +The legal values are option dataSource default and nullIf the string contains invalid option values there should be an error +The spec shows library but I definitely found the lib version when googling arounds a crazy beautiful world +The core of the issue is the inability to match on more than one object properties in a relationship +In general this could be quite tricky +In the EJB world all shared references need to reside in the JNDI environment which of course is considered as from the applications point of +Perfect +Executing this in guestfs takes about min on EC large image which is not acceptable. +Im guessing that it is a Hibernate Search bug +Tested +I think the Write permission node will work but it needs a small modification to ensure that in the time period between deleting and acquiring the write permission and creating the using the ledger and other node doesnt come in and do the same +I will keep digging +Maybe someone else wants to do the +thanks well try to reproduce it +Currently we have to use the JBM Remoting SP +Currently the demo test does not work +a blocker +Added an exhaustive set of tests +Thanks HyunsikIve just committed it. +See the most recent set of patches Im attching now to +The values are integers but are stored in the form bean as strings +Ok Brian thanks for the note Ill remove that only on AS shouldnt be removed at all on the is no Fix Version AS in JIRA so I scheduled this for in the expectation that at some point that version will be renamed to AS . +open the file and begin using it in a directory named for the tserver address write references to the log into theMETADATA table as the log is used tablet server removes references to logs as tablets flush to disk gc asks the tserver to remove the file when it sees no METADATA table references the tablet server ignores the request if it is still using the log master will assign log sorts when it finds an unassigned tablet with log references log sorts need to recover the lease on the file to prevent stray updates from appearing log sorts should be monitored perhaps made into a FATE operation once a tablets logs have been sorted the tablet is assigned by the master gc will remove sorted log entries when all references to the logs have been removed as always checks against theMETADATA table have to use the special consistency checking iterators +We need a Tasktracker registration mechanism that validates a tasktrackers version when it attempts to first contact the jobtracker +The patch is applied into Camel trunk the fix version ofWhen I changed the camel version to the test passed +Patch looks fine +Committed to trunk. +Unable to reproduce +Thanks Uwe for the pointer +Patch appliedThanks to Alex for the submission and thanks to Lukasz for his feedback +It shows nothing at all +Im having trouble applying the patch +Compression will help on disk but we do have to decompress to do anything useful and I dont want to cause more GC on every op to deal with what is a pretty rare corner case +Amazon S which would be very nice +Do you know of any public proxy server that exhibits this behaviour? +The new patch is goodThanks Tim. +I hope to have a patch for you to look at tomorrow +I see Willem is working on a test to show this working at Apache +Then we exported the snapshot directories over NFS to our hadoop workers and ran the MR job that way +OK I found the problemSolaris binaries need a symbolic links +Its not consistent and I believe it still fails for some we could determine that a given input would cause too much processing we could certainly throw an error as MRI does +The description of what to do is not clearInvestigate later +Hey HariYeah I should have noted that sync is a noop +This might mean that creation of resources is broken on Developer Cockpit Plugin side +Does anyone still see this? If those that so it before no longer see it then perhaps we can close this as cannot reproduce + +Swift driver is now available in drunk. +Sorry to have stirred up a bees nest but hey at least this issue is closed now +So I asked builds for automatic test hook a week ago but they have not handled it +Theinvalid auth response generated by was due to a deprecatedconstructor I used in my test cases +No rush +Yes I took only the jar into account. +Im also experiencing this issue +Shared instance is used insted of creating new instance in setContent method + the patch looks good to me +I verified in my sample and in my real project that this appears to be fixed in +Just assigning this over to you since youve submitted the patch +arghh the problem here is that does not have Configuration in the classpath thus we cannot check if security is enabled or not +It is trivial to have that many running on demand but when it comes to physical infrastructure it surely feels like an overkill from the management complexity side of thingsNow what would be completely awesome is if we can have something like be managed by ASF infra and provision for us just like EC does +As Ive gotten no further feedback Im closing out because I dont understand the requirements behind the request. +In response to a comment on irc from mfojtikrequire nokogiri in irb returns true +The loader used for transforms might be a but this loader has the ability compile scripts into classes on its own +Vinay thanks for the hint +So you may be on the correct trackI notice that isType is never initialized and is never set to false +I have to setup the env again on my MAC +It should beCAMELCLIENTTRADER +I feel that this implementation is a good solution for integrating Caja with PHP shindig +This appears to be the case as different versions of the are used for compile and runYes I knowCan I ask what runtime dependencies you have on groovypp? I guess you are using something other than the static binding +Spanish localization file without grammatical unicode char error and updated to rAttached patch file for easier integration +Yes is one part of my proposal +Modified features page to include info on cactus goals page to include info on cactus using page to include info on cactus is my thoughts on the cactus sampleI was going to make a Servlet that calls the the and print the results +This will fix one of our failing demos +You just specify the exploded parameter or a JVM parameter true and the exploded deploy will happen +the number of failures was reduced about times but there still are some failure +Thanks +It would appear that IF a tile includes the deg line then the tile is incorrectly rendered +Sorry for changing the summary again +Move to +works nicely reused it for beanshell +I didnt even consider the version issue when submitting thisI believe well be ready with a submission later today +Patch brought up to date with trunk again fairly invasive because it kills replaces with the two subclasses optimized for different use cases +I can probably do that using the on the command line +on this patch +Hi liugangCan you verify my fix by running the test case with the latest CamelI did the test last weekend it works it is +from me too on this Dennis +Maybe the console doesnt work properly with IE? Can you try another browser? It might also be useful to take a look at the javascript error console +Other items as mentioned in above comment are remaining +I will write an overall design document and post it here soon +Many thanks to EdvinBye +Current grammar doesnt support defining generic types +LANG PERF JDK PERF Length replacing length with length times +Good newsStrange it didnt strike me that I should try the latest snapshots before reporting Sorry for the noiseCheers is already fixed. +to this patch +Just waiting on a commit on before I put in the last piece of the trunk patch +s in the comment please +You can check that by opening with the text editor +I would just deprecated it and suggest direct use of scanner instead. +released issues. +It it in some cases also sent over the wire in response to a legitimate requestedit this is actually triggered by something else +Closing this JIRA as the original issue has been resolved +i think we need to identify exactly what the problem is with the current behavior and fix that since the ability to connect to a zk server as an unlisted follower or observer is useful +Attaching a repro for this problem +The caching is to prevent the unnecessary dns lookups that are a multiple of the number of datanodes typically just to view a jsp or query json or for other internal operations as well +Kannans patch fixes the issue for me and I also tested with a smaller base scanner sleep and twice as many regionsAssignment seems a lot slower tho +Assumed that it was valid to pass in a String to a Date field which it is not +The supplied patch breaks existing functionality as no new patch has shown up Im closing this. +The work I did adding widget boundary comments was a lot harder than it should have been due to the existing screen widget architecture +According to one of our recent patches we were always reseeking to the next column or end of current column +suite time hours? Alternatively override on a particular test class that you know lasts longer than that +In r +Can I get a second from a committer +for marking for Does it also make execution of these queries any faster? Sorry Im not very familiar with and related Lucene classes +However with we could simply instantiate a and it would start without additional configuration +Its quite the listener will include the parameters needed by the action class inside the message +A good interface is needed there +For the most part the basic C libraries have worked for a few months +But I have a question around having multiple files in that you have a root pom and serveral modules +Thanks for the reviews +This passes unit tests and doesnt seem like it would have adverse effects but please comment if you know a reason why it wouldIf this approach seems sound I can write up a unit test +Makes sense to wait to commit this change for trunk +Closing old resolved issues +I have to add the tests for the cases when the exception are thrown +Perhaps you were trying to reapply on top of his apply +README uploaded for configuration file settings to enable and disable feature +I have run the basic conversion now have to tweak and apply the templating +Stuart that change looks good to me +HA +Imo its almost the same as with other +Ill debug +I just committed this +The attached file removes changes I introduced in the to test this patch +What version are you using +button in the Remote Java Application dialog opens our preferences pageIn that case we wouldnt have to change the Eclipse Remote Application configuration page that now causes Eclipse to write the following message in the log file when startingENTRY MESSAGE Tab group extension for type for the mode has been replaced by extension Maybe we should change the name of the preference page from Remote Debug to Debugging Remote Java Application +Well because of the code snippets have to be updated +page was attached +Popping up one level even if this bug should be fixed I dont think it will fit in anywhere +To recreate go to from menu +They also do not display the pdf correctly but they do not crash +Can you please that your code is set up correctly there so that your JDBC access code actually receives a Connection proxy from the? Check the Connection handle that your code receives there Is it a raw Connection or is it a Spring Connection proxy? The raw Connection will not work with suspension but the Spring Connection proxy should be able to handle +Very strange +Now i see it +Im also a on the idea +Please can you be more specific about which code you dont think should be included +another test needs update Constructs make sure the spec is right on this +I cannot find mvn install in +Thats not possible with the current Envers implementation. +Could you try with the current trunk and see whether that makes a difference? If not I will have to investigate it more deeply +Is there another view in the same design doc that does have a reduce function? Ive never seen a report of that error being triggered accidentally and your view doesnt do anything that should cause it +Hi JulieThis should go in tomorrow I got backed up with other dist work +This is really strange the listener logic is the same for all events and it works well for other events +Problems like this usually appear when the number of connections to the broker exceeds the number of open file descriptors that the OS allows +Merged the nd pull request in and in master. +Thanks Navis +I have mixed feelings about that +and good luck +Have upgraded this to Critical +Talked to Lars whenever we can make a better interface that is also fast we will change it again + +Unable to reproduce with any scenario similar to what was describedPresumed fixed +So far I dont see anything setting the type field so not sure what you are expecting to happen here +Maybe is a good already exists in order to represent a query block in a SQL query +It will definitely help me to understand the code +Only partially fixed +It looks like Pete just fixed this with r or at least the quickstart started to work with trunk. +Patch to set the dependency chain i prefer we move it to trunk since its a development thing and not a bug for users +pushed to. +Here are the examples of the callshttplocalhostcommanddeployzoneIdtemplateIdhypervisorservicenetworkIdsprojectIdhttplocalhostclientapicommanddeployzoneIdtemplateIdhypervisorservicenetworkIdsresponsejsonsessionkeyzDprojectid to documentation issue +Any click on them will just take him to the login page +Removed markup since this Jira doesnt support it +StackYes for some reason in the original code I was being a bit too clever and registering the protocol handler directly +Thanks! Howd I miss that +This comes from the client lease feature which we will be able to disable in +Also what if the user saved the html to a file? Finally what if the user leaves their browser window open and walks away from their desk? The users password is then easily retrievable +I do not see a split here +It turns out that there is a configuration option that a can set to skip Clover instrumentation +If no objections Ill commit the suggested solution soon +Sorry this error is not present in I forgot to recompile after changing the gdk +I think to get anything meaningful wed need to run many more iterations and get a better statistical understanding of itIm on switching to either v or vGood work Randall. +Maybe can have a protected should with a default impl that handles the two known fixers and otherwise returns false? Then the user who provides his own fixer can provide his accumulator too +guess the expression wins if not null +We actually plan to submit Mesos to the Apache incubator soon so maybe it will become Apache software eventually! Ill get back to you with an updated patch +In the past this file has been bytes after the run but this time it has of the print statements in itPlease let me know what additional information I can providethanksbryan +this patch has the to publish hadoop jars to the apache mvn repo +Patch reviewed and commitedRevision +test cases relative to srcbugs for those withing to help me out +Whats left +There will be no release. +Nice doc. +I added a since can only be created from simpleTypes +If only generated code contains Java artefacts there is no problem from my point of view +Its unclear if this is a bug in or specific to the project they are using it in +Thank you Andy and Slavik +It gives the new field package level access using get and set methods which should be enough for our purposes +most projects will end up with multiple versions of the same jars in them +bq +canceling patch while comments are addressed +Scott Boag thinks we should issue a warning but allow the usage +Please note that is about tmessages not about h but Simon is right about this the code is in shared module so theyre both effected. +bq +Display both master as well as regionserver coprocessors in master Web UI +if the value exists it is returned that part was working properly and was not modified by this fix the fix was simply if a value does not exist at the servlet container level an empty string is returnedfor some reason the original code had a Replace method call that returned a value for when it should have been emptyif is an issue of encoding then there should be a fix that will address both issues. +Geir has committed Map creation code to CVS which will be part of the release +Open to reproduce the error +Doesnt Objenesis itself have an option to cache or not cache? Is it enough just to turn this value onoff? So we dont have a static reference ourselves in the +Marking as fixed when you run on hadoop or later +Changed to prefer local over JAAS +An example of international +Geoserver with WMS Rendering Nearest with WMS Rendering with WMS Rendering Simone any idea about whats going on here? Maybe the new raster renderer code has issues with intepolation settingsGuillame one question do you have overviews embedded in your raster data? will use them so the starting point for interpolation will look just as good as the overview you have +Doesnt have to be plugged against when we open write barrier and mount disk with dataordered we could make sure that the data have been flush to physics disk after we call fsync system call +It seems there is a fixEmpty the tile caches for the two layer groups +it is done in was resolved during or earlier. +for missing source and script files +Committed +Please close this +Should we also set to a smaller value than the current mb? We can override for all tests in srctest +Changes look good +Yes the security provider isnt ported to felix nor as of now +Is this a record +bq +Ok but I do not agree to drop dependsOn in while there is no real alternative solution for plugins +I suggest we still post something to the mailing list regardingthe changesMike +Patch looks good me +Cause as it turns out SAX is a huge friggin pain in the butt +I just realized there was a bug in the original fix as well +this should be finally it +There is no way this is a blocker bug +It will reveal more trouble The demo request is using WFS and for that Geoserver is listing these GML GML CSV and JSON +looks good to me +New patch +I tried it again still no luck +Dhruba this is not about making users write good code +REST API spec for the IBM Developer Cloud +patch to trigger Hudson QA tests +Checked in ruRU nlNL and fiFL +I removed files at Adrians demand +Is this something we still want to pursue +It looks like the new generators cant use sequences and I absolutely need to use them to generate +I went with the idea + for the patch +Thinking back about it wouldnt it be possible for the parser to split those keys that contain periods and to allocate new hashmaps to their first key partThat is if we have a form field name then would be a hashmap containing a bar keyThis method seems much cleaner than the patch I submitted to the core +it should fix the problem +OK +The first message component used in a war was taking precedence on all the others +Vladimir can you comment on this +Tried on the latest build problem still existed +Should the shutdown process start with setting the serverStarted attribute to false? Or should we introduce a server attribute +For xhtml htlm files there is the same in JBT Version +given that we dont have and Configurable Writables +Also if user want just edit device definition for not used skin he has to remember that once he finish editing he has to select previously selected skin in Device list before he close Preferences dialog otherwise skin will be changed to last edited skin +Change looks good +All have similarly massive slowdowns +Dirk This latest patch doesnt apply cleanly for me +Shazron it is those commits that cause this issue +Handing over to Geir since my compilerlink refuses to build even with all the fixes +Beleive me or not I prefer grepping and sedding Gb of logs than stepping through K lines of codeEmmanuel may I quote you on the above +Thanks +Thinking more about it even logging at DEBUG would do +Just a quick update +When I came back the cpu was at and I wasnt able to make the dialog go away +Thanks for the review Devin! to Svante for further update +One task can set one bit other update another in the same byte and the whole result is unpredictable +Patch to fix in related issue +This piece of code seems correct +I will try and merge them later today +As it is done in T we can dynamically decide which zones to update +Fantastic +Cheers +The other file that is effected is since it references the service code class using the full package path +Initially a newly created UUID could be used as the system identifier +example for how to handle this has been added to the tonr application +Advance zone cluster with hosts +Should not be a problem with V of the plugin. +I noticed one thing that would be nice to improve in the patch to nicely format the config file when new properties are addedIvan is the patch ready to be reviewed or does it still need more work +More work would beperformed for an invalid request thus consuming more cpu time on themachine +I relalize that I have tested this property before but not by adding a new file but by changing the property directly in the jar +Logged In YES useridWhat is the full jdk version and operation system version hereLine is the setting of the value to null nullwhich simply implies that the tail ivar is null appears to be handled at a higher level sothere needs to be a check that tail is in fact being lockedcorrectly +Attachment has been added with description Demonstrates the for the bug report and the really simple exampleTheres an error in the border resolution but since I already have changes onthe track its not easy to commit a fix now +Ok +Closing tickets +Fix counting the diff on rb +I was going to export them to disk and then remove them from Confluence when I got a chance. +Fixed this in current trunk +Possible fix +I just remembered that at final migration time well want to switch the repository over to readonly +I am not sure how it works for you unless you have somewhere else in your classpath +And thanks also for the specification of the representation of parts +I havent make a diff file because the attachment is only a proof of concept +Moving to +Is this only about preserving the comment or is there something else to do? If not this is fixed with +. +i guess this will resolve some of the compatibilty issues previously discussed relating to the jdk code from google +Handy if you know more or less the number of groups upfrontbq +setting the system property to false should do! hours working and stable GB merged. +A slight flaw with my previous suggestion of throwing an exception at PMF creation when the isolation level is unsupported +The Hadoop default is to not compress and we follow that +pushed +can you please commit it +I just applied the patch +WDYT +I meant doesnt marked as Done +Right removedNote to self dont post anything when youre in a hurry +Currently in we have case JSE return Java SE JDBC Is that rightThanksKathey +The new patch looks good +All links for JBDS and JBT books were corrected for JBDS through the QE process +Hi FrankMy apologies +That info didnt make it into the defect description but it was part of the discussion on the Google Group +Theres no transaction factory in JBI +Is there any reason why the new methods added to which are and Sync cant be moved to? The reason is to keep the code in par with java Lucene as much as possible +So for a a kv put on root youd get structure modified root structure modified data in initialized with empty map data in root initialized with empty are not currently designed to work that way +When using the default Java mechanism you have to deal with the keytool utility which isnt that easy to use +this version checks both current releases and archives for the hadoop change Ive had to made is to add the memory option for the processes the cluster launches in the that gets generated +TedThis change is not applicable in +This sets JDK as you suggestI found that setting +Looks good to me can someone else commit this then? Thanks +Will attach new patch soon +Will try removing the public from the new classes on commit so they are package private +classes directly? Im just wondering if I need to export these dependencies from the jaxb module +reopening so i can properly set this to fixed +In case someone wants to override it we store it as metadata +Will need to parse out those params and pass them on to the underlying long field +Jared have been reordered on the Magnolia author server +this patch to sendjson puts Connection into the response header if it exists in the request +bq +Approach looks like the better one to pursueThanks for taking this up Hitesh +It would only take a one table difference and the entire would be off +And examples of what types of warningserrors might be applicable in each state +Or perhaps someone might wish to nest a sequence file within a sequence file +So doing these incremental updates is a bit messier now in regards to testing +Should be a fairly simple visitor to write +such that parent and daughters are always in the same region via split policies +Why not fix they current one in a way that does not require spring? +I initially tried using a instead of a which would only partially work as it wont handle the correctly wrt to signaturesI guess the only way is to do the same as Equinox +pull request forBetax is pushed to branch and closed +will update soon +CR tagging time +Youre right was to late yesterdayIndeed we use in etcpasswd as an test source and thats bad +Im also processing hundreds of documents per second annotators rejecting documents for various reasons is normal and have a very measurable performance impact though I cant be sure its the exceptions alone that cause the performance impact Im sure the logging also plays a role. +HiAny resolution from technical experts? +I committed this to branch. +Checked in the patch. +Will open another ticket with module patch. +Thanks for reporting this +Changed Target Version to upon release of +The thinking is that hint type messages ought to be directed towards new users since theyre the ones who dont have the operational experience to know betterDo you have an alternative patch that would satisfy both the new users and the big cluster operators +Thats why a M url segment has lots of separate fetch lists +Basically the classpath may safely contain as many JAXB implementations as you like as long as the proper implementation of is choosen +Lookt at Graphic card as examplehttplocalhostsolrselectqcatgraphiccardversionstartrowsindentonfacettruecatI will try to maintain the wiki page +Let me know if you like it +Additionally the servlet can be registered in using the interface so it can be refrerenced from other bundles which will then wait for it to go active like with any other serviceI think we could make this even better using the whiteboard pattern +Moving to is including the updated plan files in the generated cars now. +I want to add an idea how to solve this issue +I believe it was +I may get to it this weekend but next week is more likely +please review the reason and risks +Hi! Ive the same problem but I solved that rebuilding the index of the local repositoryChanging the build order dont make any effect for me! +Under it probably shells out to stty to disable echoWill rename this bug so its a bit more clear what needs to be done +and changing this jira to be specific +The temporary solution for me is to delete the in grailslib problem seems not only on Windows as my colleague also got the same thing on Linux. +who wants to set the touchy fix version +What do you think could go wrong? +What is left is that if you use a youll need to use a with type Number obviouslyCorrect me if Im wrongregards commited a number converter to tomahawk sandbox which will automatically convert the value to the beans expected type +Kathey also reported on IRC that her run also got no errors so I think this patch is ready for prime time +Cannot reproduce it though yet +the problem confirmed with +I agree with Niall that this would also be a problem with the regular tag but Im not seeing how the original URI isnt set +As soon as that project is created I will submit a patch for this issue +Ill move the JIRA +this is a regression fromSending CHANGELOGSending libbuildrpackagingSending libbuildrpackagingTransmitting file dataCommitted revision +Sure I can add an example you want it in the one at +Scott looks fineI need to log a doc bug in fact to document the new behaviour +I was able to get it up and found the problem +From what I recall I may have thought that all the tests in the junit need to use for update clause but I dont see the original test using for update clause for all the test cases either +Made the changes and I get a Build Successful too +I am a bit confused by the message mentioning Calendar as there is nothing in that source which references calendar +Calls make recursively until it hangsBUILDDIR +IMHO this is something to be done by the webserver only +logging a warning seems like the right course of action for now +Thanks Austin +Much carnage still to resolve but this task is complete +Waiting for JIRA privileges to set it to resolved +is not working with this patch +I tested it and it seems to work +A message was sent to Hiram Chirino identifying the problem +We have also rewritten the test cases to be more explicit about how the cluster is configured +Marc Ive just added repeatCount property for job classes and uploaded new snapshot +If I cant fix this I think we should back out +I added this to the Camel release notes and updated the WIKI pages with the correct version which with this option is available +Initial patch +So that each plugin can still be independently turned offThanks for the workaround though +The change in the makes sense still I see it not being generated +Please close this issue as a duplicate of +These were caused by another patch and they were corrected by with the corrected trunk +Thanks Todd! +This is still an issue when using +If you move your server you can relocate it in the server would invalidate the drools classpath container +I now did the changes and tested within Juno looks like works fine so far +We could provide an ant target that builds a binary tarball but still release the compound tarball +Looking at the JMX info via I can see that under Broker NameProducer there are hundreds or thousands of entries +Let me know which you prefer +The issue with the quickstart repo having this is how it is merged with the quickstart repo +Response to Thejas +Fix version is left blank +I ran ant test just now and it was successful +bulk close of all resolved issues. +xml file +Rick Thanks for your commentsWRT exceptions +I updated the option descriptions and made the a static inner class +Everything that I plan on being available is now available and ready for the build. +thanks +Thanks for the suggestion! Ive added such a message property of type Object accepting either a direct or a JSP EL expression that results in a +Looks perfect ATM +chapter is sentence and several links are added into the fourth chapter. +Patch that adds the above one second counts to hregionservers metricsIt still needs some more tests I just want to get jenkins to run these through the test suite +Committed update the documentation tomorrow. +Thats how I imagine it could work of course this needs to be discussed Ive send an to about this +best solution is to get apache version from the customer and compile the snmp module for them on another linux machine that doesnt have these problems +fixed checkpointer copying of xml configurations from hdfs to local and local to hdfs it works in pseudo distributed modeI hope it will work on distributed modeTestcases are working bench in pseudo distributed mode too +Username lqd +Edit at closer inspection I see that we were proposing the same thingIll see if I can get this agreed in the Aries community +Comments? +Updated patch with few formatting issues closing crossed an issue with last patch anyway nobody cares +Ah +Much appreciated for the incredibly brief turnaround time from email into a JIRA ticket! Thanks a bunch +This version is not released yet +This will ensure that the pkgs we deploy as part of will work for everyone +Can we verify what was done with the previous issue +Cool that you took this on! I was going to have a look at this today! I shall grab the typo version of this ticket needed Fix version for the branch +Requested DDL irrelevant fields and indexes this is because of is generated from a hashmap iteration which does not guarantee an ordering. +This has been fixed for both GOP and GKOP based on the patch attached +Make sure you do not have a faulty service in your services directory +I like having the expected output of the test under our complete control +I am hitting this issue +The error message does get printed to the console howeverAlso there seemed to be a lot of unnecessary import statements leftover +That is correctRegular statements are initialized with poolable false but that doesnt matter for the internal Derby JDBC client side cache as it doesnt cache those statements in any caseAttaching patch a which addresses this issue +And the failing test tests that +Latest patch from reviewboard +Can be committed once jetty pushes another snapshot +The file contains the functionality for loading included and excluded plugins configurtion information from files found in the classpath defined by the configuration variable +Because the patch had ASF granted license flag but it was not right because it was totally based on sebastians patch which was not granted to ASF +Patch applied at +The ability to call renderComponent on a specific component I think is already there though you do have to write some sort of engine service to control how to pass in the writer +This looks greatFor some reason the changes to didnt apply cleanly but I was easily able to fix that up by handI also moved the interface from toAs for I think this patch mostly subsumes it +Tilman Hausherrthanks for your reply and your helpIll download the newest source file of pdfboxand try againreally helpfulethanks again +I think we need to actually pull the subsnitch value out so we know which one its supposed to have affinity to then see if the affinityd score is over threshold larger than the other and if so override it +Weston can this typo be corrected or is it legacy already and we cant touch it +Of course that does not preclude the need for validating user and group names in the future instead of having it hard coded as it is now. +But is fixed a month agoOr do you think that this is a regression +Moving out of +Then if you run on another node it copies the root to the local system cuts a fresh private key off of it and sticks both in the default locationsIf you have time I would really appreciate if you could post example commands for initializing a node cluster using this processbq +Upmerged patch to latest changes on trunk +There appears to be two issues +Thats it +Some fixups were required to tests as the way Namenode exits from bad edit logs has changed +Thanks Sam +BULK EDIT These issues are open to be picked up +I just committed this +Derby does not try to determine if you have other privileges that can replace the privileges that are being revoked +I think I copied +We need a way of including them all +This patch will fix all problems for UDF and UDAF +Resolving this issue as cannot reproduce +Now in Extractor the basedir is something like CTemp and fileName becomes CsomepathI think that it cannot work with the fix in +patch attachedOne bugfix also included for the driver where the wrong variable was being written out to the +I also removed end line bad space for nl patch to correct missing quote +Also the sample parser for json format can be checked in +I also agree with Al I had good success using the method solution to determine which button was pressed +I tried on windows today and the UI was completely frozen +Can you change your patch to share code and tests so that both the old and new API behave in the same way? Also the old configuration parameter should be deprecated but still supported in the new API +Thanks for taking a look Tom. +Moved to we wont release a has been releasedMoving all related issues to the next version. +Just verifying the write access to the resolved issues for already released versions +need to make the JCR one default and clean up a bit I can do that by the weekend I m working on having this as default +After reviewing the specification the default value on a property definition doesnt do anything unless the property definition is autocreated +James I am unable to recreate this problem +Ill look into it again when I have some time since a couple files wont get hot deployed on change +Rohini could you check if this is going to be an issue with federationI am mostly concerned about commands like cd for which we do not have equivalent in the fs command +Left info for a while +we can reopen if we get more feedback +I have also been able to determine that the same occurs when using the distributed reduce phase no matter what size the data in the cache +Dont clone source pages permissons to destination page permissions + +Do we need to make any changes there +I havent done anything about the licences that need to be advertised yet. +Noted +This should be fixed in drivers. +Committing shortly +Revisions other than trunk are removed since coverage fixes are accepted only for trunk +Postponing feature request. +bq +is my application code +Lets make them save +failure is to be sure Im not holding this up Im a because Im not happy about normalization but its important to fix the NN corruption + for usefulness and agree that figuring out the API is the hardest part here +youre wrong about it being unused though its used extensively in the unit tests +Moving back to for now +I have updated the Strus Wiki to remove the reference to this feature + +This just moves the messageID to the dont need logging for infostreamMost of the time someone just wants a simple configuration to dump to a file or they are a developer that wants to dump to stdout for a testPlease lets not make this difficult for the of users with some complex use case +Integrated to and TRUNKThanks for the patch HarshThanks for the review Ram. +Nothing seems to indicate a problem in Solrs code +jarfind GIRAPHHOMEtarget name giraph +Ran ant test +Deactivate waits for the tracking number which never arrives +It differs from inthat it implements the more aggressive variant stemming derivationalendings etc so it gives users a choice +Only one patch is allowed +John would you like to test MACLinux compat for +Actually I just realized that the patch I proposed doesnt include that last recommendation about using PRIMARY for compatibility with other databases +Anyways I hope to spend some time over the weekend reviewing these parts of the changes and trying them out in geoserver to see how it worksBut great work to be sure! This is definitely work I want to get into the release +bulk defer of unresolved bugs to +Our config scripts now assume CC is the compiler on Sun Platforms. +Could you please attach the java definition of and the contents of to this issue +just assign to me if my explanation isnt sane and ill do a PR with the changes instead to make it clearer +Im working on a native unzip in CR. +When using Apache Tomahawks attribute visible the comparison is +finally submitting it to altering the code noIts a constant expression +your bug is a result of a conflicting antlr version from maven +Apply methods are now deprecated +Would anyone want to try this +Note that there are changes being made in to to use a path +Move back and assign as necessary +The recursion cap shouldnt actually exist it isnt found in the formal description of the Hunspell algorithm +It fiddles with the optimizer in order to work around a problem +I did this slightly differently. +Will submit it as a patch once Im done +Also pls create a separate JIRA if it is not related to this +Is that correct? We dont want to be invoking this when Python or a Python control flow +Not unless someone has a small sample ready I would have to take a while im still trying to figure out which part of my apps hibernate statements causes the error +Changing the fix version from to unknown since we dont have the resources to work on this issue in the short term +Thanks Brian fixed now. +We should fix this for +insteadI think and are essentially the sameHow about +Improved patch +Carlo apparently there is also an upgrade needed for Is this intentional or a mistake in the bomIf it is intentional what is the tag for it +explains that we do not need to set the CLASSPATH as it is pointed to and if we started Derby from java this is actually an observation of me +Actually need not be limited into taking Map can take any deserializable type such as Map and be delegated +SVN next +Hi CharlesIf you want to share the port between and camel route you need to consider to use componentIf you want camel router to listen to the other port you may consider to user component +Thanks HansSo you agree with Franciss proposition to have a global one in from where to handle all created right +No biggie +Not a big deal either way lets go with what you have +K +The already contains this information and it should be straightforward to display via the web UI CLI or REST servicesNote that is already attempting to do this for historical resource usage so it will be easy to see which jobs have taken large amounts of resources and could have slowed other jobs in the past +The reason for the large numbers is Im storing UTC times for dates far in the past +fix to xmldbtransformer should solves issue +In case of LDAP clients should talk to the LDAP server directly +Toolstrip which are bound to different resource addresses are treated as if theyre bound to the same resource address +One rob this happens on large s definitely a case of wacky html +In this patch the testcase is an testcase +Excellent thank you! +Ive fixed this locally by extending the and adding a list of +Dumb question why are we comparing host name rather than host ip address +I dont know how. +configuration is encapsulated in a Configuration object which is injected into each of the interceptors +It might be enough to move the isResolved check inside the synchronized block +I just looked at the code and I agree this is wrong +This is actually an issue when using any Dialect which reports that it supports for update of columns the Oracle dialects being examples of thisEmmanuel could say more definitively but I believe that does not map to the enhanced table generator +just replaced with correct one +Basically every concurrent update is assumed to be a conflict by default so this results in a lot of conflicts and the number of retries for runs out pretty quickly +It looks like there are some script issues with several +but not other Advices +If changes will be going into please open a WFLY +doesnt run by default but I ran it on my host where it passes +Note that this bug was filed against RC which I presume was newer than RC +Seems fine for variety of smile tests +The fix should be ready for the bug fix release and of course it should go into the trunk as well hence my choice of fixin values +There are still a lot of issues to work in this area +The image should be in srcmain! +Attached a wrong file last time +Never had to deal with it on a Windows box! I just chalk it up to one more way linux is different from Windows +This is in +thanks would you mind including a test casealso its important that the diff be taken from the level of to make it easier to apply +thanks +added an entry in to exclude them +Sorry whats the status of this issue? Im especially interested in geoserver problems since were about to release in has been modified to return for a FID mapper with no FID columns +Maybe we should also take there the counterpart from guavaNo ideaIn the meantime Id rather add the checks in the way that looks least disruptiveIf performance was a significant target for this method maybe there should be an associated test then we can test alternate implementations +Why? Because hadoop is about moving computation to data not the other way around and without colocation this is exactly what youre doingThat said we understand this is a problem that needs to be addressed but it is hardly critical. +Yes that would be nice indeed +With the same setting there is faster than which at this point makes sense +Ive committed at the revision +Forone thing that page is not under the control of developers but underthe control of the whole system +I think is now replaced by and such +I commented out the test method at r +A link to the improved Maven Archiver documentation has been added in the home rev . +I had the patch a couple of days ago I was just in the middle of testing +Im sorry I saw some comments about the but I couldnt fin enough documentation +pushed +Would it suffice for the help output to say This is an inefficient operationThe intention is to let users know what to expect +ok i was wrong again connecting to the queue but not consuming messages does not reliably work around this issue +Committed to +got everything compiled and I can verify with you that the Null pointer exception was fixeda while back +The patch looks good to me will check it in if Jenkins comes back +Ive committed this + does not resolve bundle properties at all +Im looking forward to using thenew release even more now +The API change is deliberateIve added Javadoc to note that the performance may not be as good as using skipIm not sure that it is possible to fix the code without breaking the API and existing applications that depend on the guaranteed skip behaviour that the methods provide +What is the browser and page source code +fixed in rFix confirmed thanksColm. +From Andrew RyanInstead ofFAILED Error in semantic analysis line In strict mode limit must bespecified if ORDER BY is presentmake itFAILED Error in semantic analysis line In strict mode if ORDER BYis specified LIMIT must also be specified +Yep because to satisfy a take we need to write to disk +Do you mind backporting the fix on the various fuse branches which would need this +HiHmm these are both supposed to work but I only documented the prefix return type +Naokiyes please create new issues for your points and the updated japanese translation +system property +just committed this thanks Ari. +Perhaps we could create such an abstraction in the Core Spring framework or Spring modules +More general solution is to introduce a method that allows templates to create multiple nodes at once +This problem was not specifically with queries though thats what I initially thought +I just committed this. +works for me now +Updated patch to address Carls comments +Ah I see +Ive been running it overnight under cluster loading and nothing untoward +It works fine after removing the xsd reference +From Bretts commentI would suggest the best way to do this is to allow assignment of roles as a step in repository creation rather than introducing a new flag that is tied to a role +Patch committed thanks Ying Alan! +Marked Optional because this is not yet triggered by a TCK is also a difference in behaviour between and +Patch applied at +I cant reproduce this issue Marek could you please test it again using trunk +I see +or binary GML instead standard serialization is not going to work across versions +Compiling source files to svnasfforrestbuildclassessvnasfforrestmainjavaorgapacheforrestsourcetype package does not existimport +This rev makes a couple minor changeso Further clarify that a UDAs input and return types may not be XMLo Add package declaration to javadoc for AggregatorAt this point I have reached the limit of my ability to test this feature +Sure +Attached v patchIncludes test case +This seems to have been fixed +Committed to trunk. +Patch ready for review +Could be seen as artifactsmodule tests depends on module test api depends on moduleBut module tests and module test api are merged into one artifact reusedepend on and run are two different stories +Thanks Thiruvel +Attached is a patch for the assembly using the component instead of the component +Hello Devaraj I am unable to reproduce this problem on the node cluster that I have +I dont think wont fix is the correct resolution +So that may be something to try here +Mmm +Anyone care to contribute a patch +Short of copying the entire class theres no real way for me to override all the private methods which expose this you remember the main tab was added because a user specifically requested being able to change the main class +In this case Im linking to the umbrella jira for cxf update check before release. +of the plugin also works +Just upgraded to the latest Camel release and found that Aries specific class usage is more prolific than beforeThe following classes reference Aries +EliThis patch looks like a reference code +Not urgent but usefull +On old versions of Sonar when a plugin was removed all rules associated with this plugin were automatically removed +I agree Mark +However I switched to and it all worksI confirmed that and this patch DO NOT work for schemaupdate +Thanks MarkPatch applied to at repo revision Please check that the patch was applied as you expected +Committed to the branch +Part of a bulk update all resolutions changed to done please review history to original resolution bug appears again in +I also have to add that it only fails when we launch the build from the parent project +Verified at +I am hoping toset up a faster machine for running the tests sometime over the summer but fornow I am like you it is quite hard to run the tests +Hi DanielI browsed through the code but I didnt found a class named +Possibly HTTP proxy related +If you can submit an attachment then I can see the problem +Closing issues for +Works! +Im hoping that we can treat the JME issue like the other issues address it in a patch +thanks +Shift selecting multiple days and then control clicking multiple days out of the middle of the selection would sometime give incorrect election. +Latest eclipse with both and groovy eclipse plugin seem to suffer from this bug +Chinna Please followup on the issue raised by Ashutosh +The website has been refreshed and the changes should appear in an hour +The two counters are for counting the number of records read and the number of records written not for determining whether the records came from diskmemoryIm confused +Patch against to provide a character fix +Set correct fix version +So why do we have so many junit tests that we arent using? Maybe its time to delete them? What good are they providing if they are never executed? +sectionText Start the debug operation on your class in the exampleAfter the debug run completes the workspace gets updated with new results +When the sample creates a stream source object it doesnt set the system ID which it ought to do for good practice +Btw the retry queue might end up going away as a result of reinstating state transfer so its better to be left as is for the time being +With me to tell let the customer know. +If you want you can post to the dev list to ask for the communitys feedback +Ok thats weird and incorrect +There are a couple of places where you can make things one line like beta +Id also get rid of things like progress and instead focus on allowing clients to do their own calculations and display +More information on this problem This problem occurring on IE and not on Firefox +attached +I have attached patches for both and +I will review and tested it with on my keon deviceIt had failing tests for mobile spec but I dont think they are very important +enabling the tests jar to be built from spijcr +patch makes handle the and throws from it instead of doing it from the stream class +But we cant do that for Maven I think? So yes we need a releaseI hadnt seen the Sonatype thing before that is certainly worth considering +To test this ticket Analyse a project Exclude some source files from this analysis with the Sonar exclusion mechanism the same project The coverage measure must be different This is the case! +so does this prevent users on linux to create +Sorry for the mistake here is the correct the fixed patch work? Is it going to be included in a future version +I see you have done that + +some files dont have the header +There is only empty page +I think the patch has significant value for the Hive codebase so Id love to see someone check it out soonThanks Carl sorry that we have not got enough cycles to review this patch in time +a I think you should also remove the znode as master has to rewrite it anyway +The fix +Assigning James since he is making translation pull requests. +Moving issues to LATER and unassigning +Thanks Subroto! +You are right +I can only see out of any given subset of reviews using the web search interface +If not imo JBTM needs to support this +Please indicate if otherwise + +I could not figure a way to cast this directly from a smart pointer so I left it as a raw pointer and reset a smart pointer for initializing the broker only +Just because I dont have a use for it right now doesnt mean that I wont in the future it would make it a lot easier to hand off some admin responsibilities to untrained personnel if I could put docs and links for common tasks right in the UII filed the issue to eliminate the incredible noise in the logs +I dont think so +I could not reproduce the exception +There was supposed to also be a flash alternative transport but I dont know how that can be enabled. +DBRE docs also need to be updated accordingly +I checked it in JBDS vM CR +It should be mentioned somewhere at least in Release notes or New and Noteworthy +Whats the problem with having our artifacts mirrored to central +left out of +I appear to be seeing this with +I just added one little comment about why were doing the fake write of four bytes at the beginning +It would be really nice to have this feature working for GWT component as well +Missing from the previous comment Current articles containing working examples will have their respective version in the documentation has been reorganized and migrated over the new wikiFrom now on there should be one new space for each new major release of Geronimo +Three changes to the first patch +renamed tx to the existing skipSample to use no the no property and added additional checks in step for special exceptions in tx attribute +please ignore previous submission +We passed with flying colors +If JD could please verify +That extension for Project Explorer remains unchanged since its implementation nothing degraded in the latest versions +Looks like nobody objects +Migrating bugs marked as resolved before and tagged as version as fixed in RC +Given that table validation is generic for all statements today perhaps a force flag for alters and drops may make more sense here +Brion Im pretty sure this is a completely different bug +No commit needed. +We dont need this feature anymore +Attached a nice screenshot +Answering these questions helps to inform us where to look +Bump please fix this issue +I think we should test scenarios where write and scan operations interleave +Committed in branchSending srcmainjavaorgapacheservicemixjmsendpointsSending srcmainjavaorgapacheservicemixjmsendpointsTransmitting file dataCommitted revision +fetch default EAGERI dont know the exact reason for it it was settled before I joined +I have fixed compile errors so that all the code I have written till date gets compiled under the current patch +Do you think you can run your test with this again for comparison +Resolving it as fixed +Please see the attached zip +Thank you +Hmm +ThanksSteve +Didnt generate patch correctly for moved file +See commit comment for rationale +Shouldve also mentioned this was reported by mattopenstack on IRC +Uncompressed in ER. +Therefore aggregators should be used +So ignore my first bit on +Created for EAP to be default in JBDS. +Bulk close of issues +We have a production system that is using Solr +It seems Weld missed the check +The changed behaviour is that the will throw an Exception +I havent had a chance to look over the patch as of yet but I have one concernIs there a plan for deprecation in the event that HDFS itself achieves similar performance? I think having an entirely separate FS implementation that only differs in performance is not a good idea longterm +Thanks Surndra and Nijel for updating Jcarder usage doc with respective to BK. +This is a valid problem and shouldnt have been closed +Its worth noting that with I dont actually see any increase in throughput on my machine +The first one is a known bug +Move fix target to works in the distribution +closing it. +Anything else is really bad +But we can indicate that a workaround is needed to get things working properly +Attached +Jochen any feedback on this +gorkem looks like is hardcoded in of assume its an accident of using same version as cordova it selfThats kinda that you have to version the next version at least or cordova tools will need to handle multiple versions versioning in with cordova also seems wrong if the version does not change p will not install updates thus you should use +Good idea +There is indeed some Java code in contrib I believe the uses classes +You must not have looked at the last patch very closely that got committed this is an entirely NEW checkout process thats being created in hopes to make it MUCH easier for people to customize this process in the future +I just saw that you commited Code that you did not write about here +So duplicates. +So we could remove it but that means changing many more files across the project +If nobody beats me to it I will add two separate issues for thoseYes please with as many details as you can provideAs for this particular fixCommitted revision +Without the patch in the database is getting created with duplicate logical conglomerateids and hence after the upgrade to the database still contains duplicate rows and hence incorrect results +Have a base build with a parent would help this +Can you provide some additional details like a stack trace steps to reproduce the login used etc +Fix for should apply here. +Feel free to reopen the issue if needed. +And Ruchith did some modifications recently on this issue as we were creating without getting the builder from the parent +Thanks Stefan! +scripts amazon +now handles specially as it already does with and header get of returns the script without its event handler +Ive left this assigned to Dave only because this was the last known status +There is xerces in my plugins directory +Reopening +simple test that fails now but it would be nice if it would recommend that the header params should also be a list +Implementation of my proposed syntax +Unchecking Regession checkbox as this is not a confirmed product m marking this as cannot reproduce because I cant remember having seen this again +should we just move those documents into the CONTINUUM space? Are they even relevant +That would not only be more performant it would also be easier on in the eyeIf you have time to do a patch here or on thatd be awesome otherwise Ill pick it up next week +The is in class. +Zhenxiao please address the review comments and then resubmit +Bulk defer all frontend issues to +Committed Kevins patch. +hi siWe couldnt access web service successfully from Flex before applying for your patch +We should probably think of a way to avoid thatOff the top of my head conceptually it would be ideal if the connection knew the trackingId and the audit logger would simply log it if not null +Updated the patch with format fixes +Do you agree with my proposal + I have only reviewed the patch and run the unit tests +If the target doesnt exist the shell assumes its a rename of the src file +Thus it is very possible that the SBN stays in safemode for a long time +Rather than just fork the Scala mode by copying the source at a given moment might it not be better to use a DVCS such as Bazaar Mercurial or Git to take a branchclone mirror and to continue tracking the Scala changes and to create an X mode by branching from the Scala mirror +Always adding return to tpc can lead to bottleneck when EJB invocation returns are big +The fix is not feasible to implement in CXF EAP Wont fix Rejected and Out of Date items as CLOSED +Remove the xmlenc jar too from lib Jon +Forgot to mention what I wrote above works with the current trunk soon to be released activiti +Patch against trunk +Sorry for the inconvenience +Nevertheless it seems like the normal and test phases should be symmetric +i dont like that the rootPath and currentPath are now stored in the object since theyare fundamental to the post processor +Responses added in review board +Committed the think Fix Version is +We leverage the jbpm task management API +Should be in Maven central in a bit +Id be happy to test any suggestions here. +In the case of rolling upgrade +Im working on a patch for this shouldnt be too long to get it up +Duplicate of +It looks like I changed the logging level during the test and caused a bunch of requests to return incorrect results +attached a patch for branch security +Patch to update build files to refer to +Yeah upgrading Saxon also resolve a concurrency with Camel +xml using nodeA +If the underlying web service returns response wsse security header part is added by rampart and returned to service client +This causes the early initialisation of the sessionFactory bean which via its property causes the early initialisation of the dataSource early initialisation of dataSource means its found in the bean factorys singletonObjects map when is being as far as I can tell it would appear that an is happy to be initialised early if its referenced as a property but will not be initialised early if its needed for +The patch propagates logging update to the documentation +So there is no evidence about in which case it fails and no patch so I close this one as invalid + for the merged your code changes into the Romangreat! I just started testing your implementation days ago and it worked quite well actually +Forget my previous post +To move forward the generated project should probably be based on imho the project generator should be based on maven if we move the binary distro build to maven too anyway I think we should evaluate this seeing what can be done about the project generator when moving the native stack to maven +The first patch seems to wrong +But I think it is much more efficiently implemented now +Andys analysis is correct +For whatever it is worth this used to be something that the ops team at a previous job required otherwise they got all +Updating with a new patch this seems to work for me with make have now fixed the first pointsThe rd i am unsure on whether we should use git archive because bigtop is wierd +However most users will not realize that is not serving the correct file because they did not recreate the launch config +We did include a hack for this for preview that turns off the rich terminal for WindowsThe bug is actually in jansi +This is great Id be happy to submit patches for a final type modifier once this is ready for svn +Logged In YES useridDescribe the test scenario in terms of what is being lookedup by who and where the binding exists in the cluster +This issue depends on the separation of +Perhaps this should have been brough up on firstHow does one deal with package private classesmethods then +Please close it + committed the patch. +Incorporated mahadevs code review comments +Cant you fix the bug by not adding final to the name +How will you address this issue? eg RS A starts splitting gets almost done goes into GC pause +The issue occurs due to lost jobs +Notice that the logo is still in there? and the leftside nav? that all shouldnt be its the right patch this time +The current checkTable logic checks zookeepers get whereas the checks zookeepers readonly path +Committed to trunk +I opened related project issue +Fix the warning generate in findbugsThe failed test is not related +propertiesThis will exclude items starting with not items in buildproperties +How are you defining the system property +Which behavior is better? I feel we ought to always count what is given in +James thanks for your submission +Isnt it sufficient to have the streams closed in the finally block of the constructor? Is there any reason we need to close them in readAffix and readDictionary as well + +Doug if you want to commit it please go ahead +Thanks for the review. + +Reopening so I can the stylesheet file +Etc + for the patch +Use broker id as replica id for each partition instead of using an explicit replica id +Let me know what you think +Im not sure that this is a valid issue and the assumed contract on might break as a result of this +I will restore the description body from the history sorry for the confusion. +Modified file svn statusM srcref +I think I may be onto something with this +I honestly do not know myself as the have been provided by the is from Sun is +Gert nice first stab at itWould it be possible to add samples to the wiki page of the input and output format +What is this bug about exactly? It already exists on Android therefore why does it have a tag +Denis for you? how many servers are we up and running with now? Is it something we got documented +Can this be resolved +Committed revision . +components there is only one and this must be at the end of the first component if there are components and the last one has two characters then check against bad What about server +false true kinda stuffIm not yet convinced the additional +Yep I must have not hit the ant clean button hard enough last time +added a patch of the easyrcmet for this catch Andrew +If I understand correctly you are suggesting we expose the counters in JIP for updates from TIP right? I feel conceptually TIP does not own the counters JIP does +Leaving this open +I just applied it to trunk and branch +Applied v to branch +Typo some problem OOME was obvious the tasks get executed in the groomI dont really know what we can do against this superstep hopping +Bulk defer all frontend issues to +By localized I mean strings like changed this page changed one line and so on +I agree it might be a bit too highdefaulting to sounds +I dont know of an actual problem here I just have a funny feeling that this change is going to be rather tricky to accomplish +So I let you see if you can cope with his remarks by changing your code accordingly +The proxy currently loops through all the and attempts to replace the value if its found in the pattern +If you dont want that behaviour I suggest you set count to +Sure +AndyThis looks okay to go but can you rebase it please? It does not apply to current trunkAlso is instanceof better than using a specific catch clause for? Ive seen clusters where timeouts are logged very frequently under load and we dont wish to slow the DN down just for loggings sake if instanceof is expensive to doThanks +Thanks for putting this in bugzilla David +FYI and are the eclipse style files +Thanks Amar! +Patch for s not confusing at all Why are the using as the input parameter in the first place +Send in a patch and click on grant license option +Committed revision I changed the javadocs of to reflect the changes +You can work around this my simply using raw solrj apis to set the params yourself so lowering this priority from criticalHere is a patch with additional testing + is class providing shared functionality for +mark the revId of branch commitsIm not sure I understand what you mean with mark +Sorry for doubting you +Thank you Scott +I do not know how to make a proper IT for this but the was added as part of this commit. +On windows there were some minor tweakage to make sure all +Unit testing is impossible because it is not deterministic +The debug message is always printed in my log +So I suggest that we keep the message printing in +Good luck with coding Patrik Ill happy to help with review! this mean there is no effort going into +The only improvement can be done is to throw a proper exception +Is this going to be fixed in the final Maven release +Thats why Im in favor of using the old behavior or deleting the method entirely so that instead of creating weird bugs for users it will just stop compiling +I think this is more of a usability issue +Remove Fix Version since there has been no response from users +Further assertions refered in this bug will be attached as i go along +I just continued the approach +This can be used by on and by UI too +Embrace its ugliness +Nevermind +After dissecting Joakims test cases on IRC I think we basically came to the conclusion that his usage of the class loaders was wrong +Funny enough we already have which seems better named +Just to say that I went over the rest wiki page and tried the curl commands up there to ensure still works as advertised. +Moving along +Well revisit adding our own expiration service in the timeframe +TWO test cases failed during the unit testing testinsertoverwritepartitions testpprpushdownWhen looking into the query as well as the output the plans generated were found to be the same while the query results changed +Tag property was missing in declaration +Relating to with commit of. +I havent tried a beta yetDo INFO messages get shown by default when deploying to a vanilla AS +Not sure if the issue is related to Glassfish or our application +ensure that compressionCodes is initialized in isSplittable before dereferencing it +Hopefully we will have a fix for this problem in but that will only prevent the problem from happening again +AFAIK this is now out of date as the jee example runs on it is +Thanks John +Rafi rethrow is clearly wrong since it is not rethrowingit is creating a different exception +Thanks Raymond and Ant +verified at r thanks Paulex. +Fixed the JAVAHOME problem +Ive put together a small project which reproduces the problem for meI built via maven deployed manually into GAWhen I open the app the Login page gives the dojo errors +Updated the code example to leakDoron what do you suggest? Do you want me to attached the search index libraries and a main routine? It is pretty easy to execute this code Im simply showing the simple logic +So once we actually start hitting the database theres bottlenecks getting in the way +Added the policy in rAtm I decided not to use inheritance for the as I am not sure how this affect backwards compatibilityAlso the has been adapted and javadoc has been completed +For information the server must be restarted only if the property is changed +But it does do the following +If you think they are critical for CR please them +These changes along with the derby version update apply to both branches and trunk +Thanks Juergen! +The atomicity is a great point thanks hong +New patch that addresses Hairongs comments +If you do apply this patch and it works correctly you should be sure to make similar changes to and +Simple maven projects fails with maven +Note from the code the test case uses only whereas my confirmation was only about +We should move the content around +In fact it appears that the tests can even run faster using Selenium and FF +see templates cloning be fixed in latest git HEAD +JIRA +Can you please provide an exact bug description +How about this addendum +Links have been checked with the maven plugin but the docs havent been fully proof read yet +Encodings and Static values are maintained for each classloader +I also ran the autoincrement script against DB to double check the results +It is my understanding that this security issue was reported to the Apache POI project by email but a bug was never filed +I have to run +A new issue would be good can you open one? The idea for the patch is almost finished I can attach a patch shortly +At the moment the version is set to +And I dont reproduce the issue locally now with or without the patch +that we use in our project +Ugh +I will fix this for the next release +Any modificationsto Logger will make the updated values written to but these values will never reflect in the Logger at serverstartup +Dimitris Add idle timeout setting for invoker threads cant be what you mean +Thats another thing on my mental bank that I need to prioritize and push for +Resolving as duplicate of +conf should also change the conf you want to do jndi over an orb to injects orb into the naming environment otherwiseSuns cosnaming implmentation will try to create a new orbrather than using the one configured in deploy DO NOT EDIT THIS FILE UNLESS YOU KNOW WHAT YOU ARE DOING needs adding to the FAQ +I get an exception xfire fault unable to write name of object that I set at runtime Type is unknown +Thanks Chris for the correctionsG +If you expect a single string in your installer and get four strings you can concatenate them +Some parent and main poms should be optimizedOne easy example module known as examplesportal or +Test failures are the usual unrelated suspects +I was thinking about just making the deployment algorithm search the service registry for available to see if it could find one associated with the current type +Patch and screen capture for the custom period featureAdding a custom period in the period selection drop down +Having a gem would also be and also probably be once we get a few larger projects using it. +At first glance it seems simple enough to just patch the following file and then we are done +Everything else is in +Ruby MRI does not produce this behaviour nor does with Rails +Yes you will get dont agree using other framework than JSF +Changed javadoc +Look at what happened to JSF +I pushed a possible improvement to this logic to trunk in r I still have found no way to reproduce your issue but the new logic is at least a bit more kosher in how it generates the monitor enterexit logic including wrapping it appropriately in an exception handling ll be around all day hopefully you can contact me on IRC to work through any remaining issues +TedThanks for committing the addendum +This is a great idea +I checked Pigs test file it runs fine with the patchSo resubmit the same patch +Can you attach the CREATE TABLE stmts plus SELECT which fails +Createing an empty one with the test put your style overrides here seesm good +Looking at the system configuration also entails some security considerations + +Altered which used to test configuration of number +for the new patch +This issue has been fixed and released as part of release +And this are patches one for jruby branch and other for master at +The cayenne directory is createdLet me know when it is ok to move the directories +What about my previous comments about changing the default in +I was struggling with this but I found that the selection service holds a null selection right after pushing the toolbar button +for your support +I prefer that you create another issue for the first point +I also chose this combination carefully so as not to be network bound +Print out value passed assertion and change assertion to allow that may or versions of cell instead of just asserting that there are in testCompaction +bulk defer of open issues to +Hi Nathani tried the latest version from trunk yesterday +I created this bug many moons ago before I understood what the TCK compat test looked like +However when I stop and restart the shell +Some refactoring +As you notethe various changes are intertwined and I dont foresee anyone wishingto use only the server changes or only the client changes so I think itsfine to include that change hereThanks for all the effort and hard work getting these changes in shape forthe upcoming release +Please let me know if this was intended as a more generic WAL solution +any chance this can move into a patch instead of waiting for +Not sure if that is what you mean by a clustered setup but that particular setup still worked +Hmm +I think that i have done a pull request +Yes I agree with you Dan +Straightforward patch +Hola Frankthanks a lot looks good I am going to test itPS for future patches can you please mind the original code format? TIA +See. +This is very doable as can be seen in JBPM and further +Minor nit we should remove since it isnt used anymore +Closing issue fixed in released version. +I think everybody is happy to revisit the issue and add spatial tier support to Solr if we can come up with a clear test and underlying implementationI think whatever work we do we should do in a spatial moduleSimonPerhaps we should deprecate the contrib in x nuke in x and start up the module +Headers +It makes senseI havent started any code for this yet so please feel free to start +This is this schema that I use for this JIRA so as to make it easier for the course this schema is also totally empty +why not apply the patch as it is for and fix it better in a later versionI dont think that the current patch is a technically sound solution to this issue +Please reopen if necessary +Posting the sending of the but it feels like playing as I saw different NPE but then I refreshed and never saw it again +I think you are using Connection details in so what do you mean I am getting connection object from a third party +If Im holding you up please feel free to modify my patch as you wish to get things done faster +Thanks +Duplicated by. +Test cases are great! Ill take a look as soon as I can +Are you saying that your observations are also taken after the patch +FYI the windows script for create already checks if the +Regarding the staleness transformation it did seem a bit weird to me as well +Could you consider to replace static selector with will solve a lot of problems and will make whole think Multithread ready +Are those the standard deviation calculations both the actual and calculated? Since there are so many decimal digits in standard deviations usually I had to write the test to make sure that the calculation is almost identical to the real value because it was usually around +Im a volunteer to investigate this issue +Questionsa Will you build the binaries yourself? b Can you please generate code fro the above schema and attach all generated classes to this bug +No problem I appreciate your questions not many developers take their time to actually fully understand things they work on. +The recommendation is to not use timeout library against threads that will call or blocking Java code. +These are two separated issues +Introducing other query languages later however is a viable re right it seams to be a lot of think it should be possible to write a short discription how most of OQL is transformed to aritmetic operator tree and from tree to SQL + I like this idea +I think we can move it to +start from changes will be made to earlier server adapters and AS already has a solution in place for the port offsets +Re the trick this was actually recommended by a Google engineer to us as a temp workaround +Existing class is using which leads to inter dependency between and project +mingjian Are you running trunk? That test passes for me +There is another option I made a CGI script for another project that let random people add themselves to the appropriate preference is to set cargo up as a contributor project +I copied your implementation for to the EMS and MSMQ providers and made appropriate changes for their specific stuff. +Ship it +will change JIRA description too +osd across all tests +It also includes web UI updates to display the per tasktracker setting +installed the following executables usrbingmcdonaldvestahomehudsontools + +Now the follow up question why things go into a hang mode and make my testcase to timeout? That seems wrong +I hate those excluded folders +bq +Fixed with patches applied to +closing +Moving to since push component was not working correctly in and I was not able cover this example by testsAccording to my info push example is going to be and then Ill add tests. +Move issues to next release +Move the fix to +Thanks it works now +The patch fixes the problem mentioned in the description of the JIRA +The script is for Windows but it is easy to create it for MacLinux +testcase for input +Now backspaces always only delete the char at the cursor. +For fix I will attach a patch +Reopening. +I also changed begincommitrollback handling a bit but these should not have any effect whatsoever +Fixed the is close but the job client needs to get the file system from the system directory rather than getting the default file system +Vincent has started some cool stuff for eclipse see +I will add a helper method for checking file existence +I would say that unless we can confidently isolate the problem it should be a issue +We could implement this as an option to the compiler so that folks can choose whether or not to generate interfaces that return Void for compatibility +Modifications to the model also need to be made to suppport addingIf someone is interested in looking at this Id be happy to help answer questions +New features really dont fit in the JBDS timeline +That is IMO the expected user experience. +Way to be fast on this one for the efforts but have talked about this before and I would gladly approve a patch to emit warnings on those specific cases but not errors +Christianplz give me something to work with +In the standard case there shouldnt be any memory allocations at all now during normalization +The test attached +If you want to have the the checkbox HTML rendered as described in the documentation use the simple theme not the xhtml theme +from me +I dont think that we should change the it has other job to do. +The above are private +This file should copy clienttargetconf to debiantmpetccloudstackmanagementconffind name homedownloadworkcsawsapitargetclasseshomedownloadworkcsawsapitargetclasseshomedownloadworkcsclienttargetclasseshomedownloadworkcsclienttargetclasseshomedownloadworkcsclienttargetconfhomedownloadworkcsdebiantmpetccloudstackmanagementhomedownloadworkcsdebianetccloudstackmanagement +Tested on cluster. +Applying I got out of hunks FAILED saving rejects to file srcmainjavaorgapachehadoophbaseprotobufPlease rebase +I believe this started happening with the switch to the +ChrisId be happy to do so but because of the way the web apps are structured there isnt a lot of layering or centralization of input output handling +This is afair and it is fixed in ER +Looks like a bug +The are are normal errors +The problem is to avoid clients being connected to a server which can lag behind the leader +I dont exactly remember what was wrong with simpleform +Great that one is now crossed off the list! I believe JBC has done the same and the task for me is just confirming that +Ive changed the template to make it commited please reopen if you run into problems +If you require immediate assitance please contact Sylvain Gagne at +Alejandro The curl commands in the HTTP API pdf file are truncatedFor the createappend file commands do the response urls actually contain the options part +Mark Nope will start working on the new scenario tonight +Thanks Jason I put this into trunk and. +Attahced depends from an unvalid commit of patch and fixed in revision +Big enough change that its probably too much for issue is no blocking the verification of Tuscany +bq +First version implemented in revision work in progress not added to the list of standard bundles yetTo test install the enhancerbenchmark bundle and open httplocalhostbenchmark displays a form that runs a simple example benchmark +Correction TABLE ADMIN gets the same privs on their specific tables that the GLOBAL ADMIN has +Any of the Spring Hadoop jobs would need the ability to deploy things like mapper classes pig scripts etc +mrepositoryorgmortbayjettyjettyUserstucu +Added shutdownHook attribute to the configuration element +Doesnt work yet +Also the content type is set to texthtml +v is what I am testing now +An issue should nearly always only have a single +Unit and end to end tests pass + issue with the package name +I may wind up removing much of the existing refactoring plugin as it is now largely redundant since search is working now +David +However if the namenode is Cd while writing it will have to wait for the write to finish since threads would not stopDo you think we should address that issue +New patch lgtm committing +Hi Brucewhich version of youre using? Ive checked in a unit test for this but I cannot reproduce your case +ThanksPatch looks good but its missing a test case +I like just Case insensitive filter +If you think otherwise reopen and discuss +The gallery fix appears straight forward enough but what about the Camera +The patch looks good to me +This is convenient because it provides better compatability with maps and lists just set the list or map tombstone on list or map elements to mark them as deleted rather than the previous system +This issue has been solved after the ugrade to Karaf. +The wiki doesnt seem to explain that clearly +We are not sharing credentials but I did ask for the account holder Justin to add authorized ssh keys akin to treating the account as a project account since he has largely stepped into an emeritus role leaving us with a build we need to updateWould it be better to request the account is renamedRegardsKAM +More than half of the mappers finished before the first reducer started and thousands of reducers all launched within a few seconds of each other +Yes youve expressed my point better than I did +But funny enough +Yup that is the short term solution +One comment is that the patch caculates the number of active blocks to be equal to the blockMap size which is not true all the time +ah why I change it state to be resoled? sorry +Revisit open bugs older than a year as Out of Date +This will probably need its own JIRA but Ill post here as I was working on this when I noticed itIm a little concerned about the amount of duplicated code between the various driver classes specifically the run methods in and the runTest methods in and and the various setaddwith methods +I am having quite a bit of trouble since the tries to connect to the agent before the starts up +This generates the plugin cache so the plugins become available for the eclipse mevenide then on I was able to pick all goals in maven Run luck +i dont see any code for the insert overwrite case gen generates a load does not mkdir on parent for loadthe changes in only seem to cover loads into tables and partitions if i am not mistaken +Nice one Willem much appreciated +Works locally +Did you test the jar to see if it is validWhich maven release are you using +Well we dont yet have a live central site to download from for Im assuming that well see the same errors then +Updated on Android +Well bump it to Beta +Whats the latest with this +attached applied thank you +Thanks +So there is a event race condition between tasks in the same tip and if the last event is the failed one then the reduces get stuck because that map is marked as never completed +The goal here was to shorten tests +Works for me with unix patch on a fresh checkout +Im marking this JIRA as unrelated to as Im not sure if I have the bandwidth to get it over there +This looks like a genuine exception caused by atleast one shard not writing out a response completely +What do you think about it? Dont we already have the IP of connected host when we establish socket connection? Why do we need to add IP in HTTP GET request? Is it because initially connected host IP and fetched host IP may be differentThanks +Since that class directly controls when it invokes I think its safe to just pass the top context +In this case date is used to get currency rates +it just changes messages and doc +Committed this patch at subversion revision +Is there a planned release date for or +Why dont you change the project name directly in your pom +in the numberguess quickstart as installed via Central +Manually tested +I am auditioning a fastdirty fix for this that caches a pointer instead of the string pointer if the host comes from the +All issues have been fixed. +AOP hasnt been wired into Maven yet so it doesnt pull the latest common jars +I didnt realize those fields were returned by defaultbq +Im specifying which was released May + Username user +Ouch I really thought I managed to convince the spec lead to get rid of this reference to and I thought they did some work but it stayed really bad idea IMHO the filters will see the HTTP request attributes set before the runtime even executes +I have also included a junit test in this patch version. +Flavio you can run this locally asant homephuntdev some reason my setup doesnt report the same issues +Thanks for the test drive and comments Rick +I think we are all agree that javamail is not the way to go in the future. +defer all issues to +If the child doesnt use that data it just gets lost forever +Joshua is assigned to other will implement the highlighting they already know about in our docs +Took some but not all of the changes to this class from trunk. +we can use autotools chain to resolve RHEL vs RHEL etc +Where we can configure the deployment automatically without requiring users to do any workaround +Some comments on the patch thanks for working on it and addressing the issues should be moved out of the package since it is not part of the external api how about? Since the format of the directory path in the create database permission matches s format why not use to evaluate it rather than repeating the logic? Customers running just with Java Security but no Authentication No compatibility issues +No general need for this component. +removed fix versions we dont know when this will be sure how we can fix it for environments that do not support proper serialization +all tests pass and patch looks good to me +do you mind if I take over? for right now +But this is better +Here is a simple patch +Switch implementations to not return singletons +Patch for branchx +Its also true that calls to the are synchronized which will access to SYSGHOST +I also cannot make this fail if I add Xint or ve sent an to hotspot dev asking for some guidance in how we can try to provide more insight to them +Otherwise indentation is not correctSince this jira is about using the config variable for append I think it is simpler to incorporate here that removes the tmp files only when appends are disabled +main issues fixed resolving +Updated in v patch +I can reproduce it only on RHEL x RHEL x +A single bookkeeper setup can be shared by many writers +Can you take a look Dave +Thanks Jakob +bq +Sounds like hes at least using our bundle the splash screen doesnt show up in the ALL or standalone plugin have you installed anything on top +I merged this into . +I decided against declaring as optional as it is really needed at runtime +The issue has been previously notified to a different project +uploading a new patch that ignores pingInterval when doPing is set to false +Latest debug info on relating to the dual entries can you confirm that the DEBUG line is a standard line that implies that the message has been processed +d also like to check to see if anyone is going to patch this +If someone can fix this in the next week that would be great +Is there any reason this was doneAnyway I will increase the max to g or g from m and run the test again +will commit shortly +Later can then be removedAlso I think the new system procedure should go in SYSCSUTIL schema where all th other utility procedures like SYSCSCOMPRESSTABLE SYSCSINPLACECOMPRESSTABLE etc exist +Hi JacquesThere is probably a small bug in that we should set the maxlength attribute to for the field that takes the category id +The only module which depends on is CDI TCK runner +Updated patch using reflection within +I didnt notice a difference because the grayed out style made the entire icon disappearYep yep go ahead and make those fixes and well get this committed +All Sessions created on that connection will be in Async Mode so if you need a mix of Async and Sessions you will need a separate connection that omits the use flag + patch +We are working on a moregeneral approach in future. +I cant reproduce the second error at will Ive seen it happen a number of times but not in a pattern that Ive been able to figure out +verifiedon CR +The Visual Page Editor section is updated according to the JSF support implemented. +fixed in trunk +Ill check before my next contribution +I think at least part of the problem here is which JVM version you invoke ant withFor me junitreport works with Sun Java SE but fails with Sun Java SE Since Dan has reported this error with IBM it cant be just the VM version but maybe some component the VM is using during the transformation phase +Falcon picks up the theme from just fine +This patch fixes an encapsulation problem in ijs utilMain class +I think doing a reverse lookup is fine as long as it requires a flag and is not the default +New patch implementing cmccabes suggestion to presume stable search and shuffle first +Weve done a lot of testing on this issue over the last monthsI have no outstanding bugs on client bug tracking systems so I will also resolve it here +The repro I just attached is the actual test case Im running for this but it has a database dependency but I will attach a repro tomorrow that can run to make things easier for you +This is still an issue with Final +will still take a more detailed look before i do tests and commit +but Im fairly certain the transactions were added because some databases will complain in production without them +We currently use seperated mysql dbs for achieving an isolated CLI environment which is not practical +Thanks Luca +It seemed better to standardise on buildwork so both Cpp and Java Brokers utilise the same directory and can be agnosticIve committed the change +sources to the modules directory in SVN. +In the meantime we found one more workaround +Issues resolved in +From the history it looks like it was reopened because you want support for recursive components that would be a new feature not a bug. +lucen then all of the termswill be in both halves of the index and so the query expansion will beacurate and the query will work +Since tokens are one of the possible authentication methods during RPC this was included as part of the change for. +Yeah lets add this to trunkWe can always open new issues regarding improvements +I will close this issue now. +Idempotent operations like get or addBlock are retried and the client will eventually complete such operation via when SBN becomes active +I have now filed a new to cover remaining two minor tasks including this one +The files changed arem attaching a patch that shows what I changed +Verified on JBT +Verified on Trunkjobvpe +Indeed the get method does not honor the validate parameter properly +wsrprewrite as real parameters this operation fails because producer url encoded with unexpected characters and consumer itself cannot encode URL back to admit rules. +bq +I have a patch awaiting approval. +Its just annoying +Thanks Ted! +Hmm looking at the code this looks like it would best be a web console plugin I think the could in fact report more on its working like what locations are being considered what has been found there whether that has been applied or not and why not + +I did not add test code because I am unable to get the rspec stuff in the SVN repo working properly +A few of them sound similar to what were seeing here +It looks like is available on RHEL I assume it is not available on RHEL +Give genericity to the the method in two accepting also the requestresponse pair or the. +changing status to closed as has been released +Reopening since this is not fixed in beta and a test project needs to be attached +There are several additional places that need to have this fixed +extract the file archive contains patch and a few files to copy in placeNo i use a static class with a language file mapping should be better than a real time file checking +Ive reduced it down to half a dozen errors in that I dont know how to fix +I think it is basically same with youYoure right +I reckon its pretty clumsy OTOH in my particular case Id just want to this while webwork final is not released because it is likely to be on ibiblio once its released not sure what to think now +what version were you using? It would be best checked against a release instead of trunkI just tried this on using the release and it worked as expected +Only other JIRA for is a minor logging change and is done so we can release at any time +The patch looks good +This patch removes duplicate calls to the doc target +As far as I can tell CCC is working OK at least when used from Windows connecting to on UbuntuI also tried vsftpd and but neither seems to support CCCIf the current code does not work for you please with full details +Thanks all who contributed to this issue Todd Dan Yoram and in particular Ryan +Thanks. +Assume closed as resolved and released +Try it with the version and let me know the +Doesnt seem to be any movement on this one +Committed to trunk. +I have tested patch a and it solved the problem in a scenario where my previous two attempts resulted in a deadlockHowever I notice that while work previously was always called when switch was called now this only occurs for one out of three calls to switch +Might be a limitation of P so downlevel the unintallation of P doesnt remove the jar files from file system +so it might be a difficult condition to track down +Lars note that the operation you describe is operating on a double and converting to a long at the end so there is no loss of granularity +looks good to Matteo for the quick review +and the key is bin file. +Committed to with revision It is my understanding that nothing else of this issue will be backported toThanks for the feedback Dag +After your explanations indeed all the reported as fixed reports are working again +consoleWelcome need to be accessed manually for test purpose +Patch warning on console and test cases for possibility is to have as around under the test code +It seems that this issue is not solvedRunning the attached test program twice results in a constraint violation because the same Id is given twice +It looks great +k make sense thx +Sorry for this checked in the above change with Git ID ecbbbcdffebfafadf +No the exception is correct and this is the wrong JIRA for Transaction. +Hmm I wonder how long bunzip would take on the TP +I changed it over the web interface +Since these properties are specific to a given host this would be appropriate place +Im sorry if I was unclear +The Classloader has changed so it cant get the loaded factories and add all the necessary libraries listed inSK +Im not a big fan of silent failuresWhy prefer an approach to optimizer hints? Is it because we hope that we can run the same query against Oracle and Derby by simply seeding the query with two sets of comments one for each database? I dont understand the value of this flexibility +I added a test case where URI has an authority part in itUpdated patch to the latest trunkChecked the test failure failed because of some lease recovery issues +LarsI will commit the version as it is +is there any solution implemented on solr? I have the same issue as OP +A highly uneducated guess is that an sstable was compacted and deleted while validation was expecting it to be there +Are there any full N implementations in Java? Sesame and Jena both parse the RDF subset but not the full N grammar +Most of those properties are pretty low level +We dont want to use as that then break with convention +I converted the to junit test using the attached configuration and when run it produces a simlar error indicated in the jira +The patch file processed by jenkins does not contain binary files +Unfortunately I think well have to reopen this one +This is an update of to solve the issue on view entityI will provide a update as soon as I canThanks patch that can solve this issueCan a framework committer take a look at this patchThanks MarcoA simplified version or your patch is committed in revision +Removal of the condition was deliberate as I noted in my earlier comments point Oh cool! So we can actually fix a bug with this JIRA and its not a simple refactor +Can still reproduce with latest Xerces from CVS running from the test directory +KanThe reason why I just uploaded patch on trunk is that we are going to branch +Yes this has been updated and now works +Reopening +I agree w you that its not a strategy in the code that Im the most proud ofThe AES changes do need to be more invasive than it would first appear +This is a really tricky issue because the bottom line is that people are definitely going to bump their heads into it before they go looking for the option if they ever do +We should add a test case though +Test attached + in +The original authors of Aegis considered such a feature but never filled in the codeIf youre interested in trying to add this feature I can provide some pointers and then mentor your patch +This incorporates feedback from Andrew and bumps the version id in the release notes to +Can not reproduce with the trunk build. +I was able to recreate this with a simple module test case +The new patch follows sysvinit script recommendationsNote there will likely be some updates related to configuration that I fold into kafka for example adding additional zookeeper servers however Illdo these a separate JIRAs +Im on it +Created based test matrix for this issue +Sorry but Im not sure I understand the issueCan you test this page httplocalhostfacesoverview of the demo? It seem to me to be okay +The last column header should not spill into the small area on top of the vertical scrollbar +Thanks Kim + +Can you try to clear the cache + output author +OKthnx +The repository is quickly becoming a magic box with huge responsibility +I am back office trying to catch up + did not make it into the JDK could write a behavior a lifecycle . +for replacing it + +We should check if it also affects dont believe that it does +What do the patch do with +Usage how to reproduce the bug and version informations of JDK and Groovy +With set to false it is notified immediately after the failureSorry +For the record I was able to recreate on a single node so Im taking that out of the new VM successfully completed world buster +Messaging is currently Below is a stub archive configuration for your project +Ive tested under IERF to reproduce the issue +Any chance you could take heap dump? Perhaps adding the XX JVM option will let you take a heap dump whenyou start running out of memory +Anybody knows why +Paul DavisSo it may not be possible without the overhead of the old hash or something similar +could you also review this patch? Thanks! has a from me we have reviewed it internally and are already using it +Fixed by other refactorings and changes to client cnxn +Additional implementations new classes and unit tests +Like you may need the xml namespace prefix +This should probably just be a new issue for refactoring the relationship between Poller and endpoint + to issue as it has been released as part of release +Thanks Enis +Thanks. +It passed nighlty and I plan to commit if there are no more comments by tommorow +You did no respect one of our requirement to build the file from the root +You can connect to the database and execute the statement select from to find out if some transaction is holding an exclusive lock +Thanks alot +Hi Michellejust found that bcel is available on yet they have changed the groupId +Ive used generics before but hadnt explicitly thought about K and V simply being placeholders +It is a which leaks the reference but the chain goes through the deployer classes mentioned in +dooh should have looked at the file before uploading +Unit tests and Integration tests are now in that tests this functionality. +v is a less conservative approach that skips the presumably already mapped row for the key used to restart scanning +the failure happens in both gccc and gcgen +Since Thrift supports a collection type why support it only partiallyThe user can only put those types in the collection which are defined in the IDL or the basic supported types +Marked as blocker for +Bulk move in +Attachment has been added with description Look at left bottom on page Hinweise zur is printed over ther has been added with description Fo file to reproduce pdf dont believe the table cell in fofooter has anything to do with itWe regularly see that last row of a table on a page overlapping the footer welloutside the pages and even over the edge of the pageIt looks like the cell is included on the page whenever the top margin fallswithin the +It is possible for you to provide the xml and xsl files that you used to produce this problem +Also RM doesnt inform NM about all the running applications in the cluster +I dont think its related to this bug +this is expected the ZIP contains sources as well as SRPM the RPM is not supposed to has any source file we publish SRPM too dont we +Ill leave it up for comment for hours and then commit it if there are no objections +Cant really make head or tails of the bug report +FIRST doesnt exist in cql +I suspect this isnt a case that arises frequently except when a user grabs an artifact from the maven repo +So it looks like we should be mapping to orgjbossutil rather than javalang +Hi Nikolay Patches committed in revision +this is a common javavm bugissue. +Setting it in the state is a signal to the user that there are no guarantees that the change results in a properly configured server and that the affected server should be restarted +The current implementation seems in keeping with the standard +They removed the ability to get in a CMT +Yep turns out the Server and gbeans for the module are not having their ports updated +From your earlier commentThe verify failed many times after restart HmasterI wonder if retries are enough +Thats why QA failed +If it is committed to trunk why is there as a fix version +Hmm works for me Ill get others to check alsoSee attached file for how it looks for me +Removed trailing spaces from Test program +See the method for an example of a polling waiting when using an Admin connection +I see its the headersso I have changed the commits list and the issues list to both the dev list let me know if that it what you wantedThanks +i have fixed this issue but yet to submit the patch +With the blueprint bundles installed in AS +Sorry I have typed as in the previous patch +Perhaps we can add command line switches for creation parameters? Or a switch that says dont create +True it doesnt work for multiple invocations of the macro +thanx for input +or I will close the issue +I checked and it seems not +I currently have installed Groovy Compiler Feature Feature JDT Patch Sources Feature Sources Feature JDT Core patch for pluginm trying to install Groovy Compiler Feature Feature JDT Patch Sources Feature Sources Feature JDT Core patch for plugin Im getting the usual conflicting dependency failure +Here is a fix including some asserts that should catch stuff like this immediatelyI will commit Simon! +wow this was fast thanks +There is a script hanging around which does this +Last update of patch to match what was committed +If you run tests wo this patch do you have same failures +Sorry for making this private but the project is not public yet +Patch regularizes exception throwing in HRS +Committed revision in trunk. +The normal procedures described in the readmes for the projects do work. +for the patch +Sorry for the false alarm +Added a patch for this +font could contain multiple fonts +Patch attached for heapsize +Thanks +Fixed by issue this duplicates? Is it +As mentioned previously the sample I produced appears to work fine so I will await a sample from you illustrating the issue you are seeing +So all issues not directly assigned to an AS release are being closed +Thank you. +Setting Fix Version to future for issues without a set fix version +Closed upon release of. +Yes that makes sense +This patch removes the dead unused code and fixes a code patch checked in to trunk and branchx +That is the procedure as i know +This will create some ripples Ill try to keep it contained on trunk +GaryYou can do either I dont really mind +Forgot to include the patch for the earlier +OK attaching a new with the changes you mentioned +Suppose you start with a configuration and then wish to make line so you add an environment block +The POC ignores connection reuse +we allow counters in normal batches? thats a bug waiting to happen since you cant replay a counter update safelyWe do +I just checked out the old r revision and it installs fine +A failing test case would be awesome +I was trying to compile the hellocomponent sample program which came with Apache ran the maven command for the same but it threw me the error which I had given +committed to trunk +The fix records declarations added inside a not in a hashmap so that after the not it can removed them from the declarations hashmap +While its possible to fix pathological queries after the fact the frequency of cases and their impact on user experience is greater than we anticipated +Forgot to upload this patch +applied in master and according to given steps looks ok JBDS L +Chris I do the same thing with terminals on OSX and the test works just fine +Tracked this down +Please provide stack trace +is to blame for the whitespace formatting it really screwed the pooch with respect to the jspc stuff I was using the feature of and thats what comes out appreciate the tidy up +Thanks to Peter Mahoney! +Closed after release. +Patch against the current branch that updates the logj version toLet me know if youd like this as an additional attachement to instead +got a patch +defer issues to +A lot of tests are shared the new just requires some different odd setup that made it easier to copy and change the test file +set proper fix +JBNEM jmxweb console should NOT be madesee interesting but dont think it should be turned on +run ant Dtestsjdtcorecompiler +Something has to be done here +We cant resolve this issue until we dont add Seam Solder to tycho build +Looks like this bug is a duplicate of +Thanks. +Closed accidentally +I am guessing the issue is that some other test creates +Gleb do you have a patch you want to submit to resolve this issue +add small replaces in and Benchmark am for you to commit this Feng +this is a huge issue +Its a good improvement +Keep in mind that you can also put files in the EAR as another option +Author samindawDate Sun Dec New Revision Revision +Jamie could you check that this issue is fixed after a mvn deploy using maven +Did it start happenning just recently +Since this is a integration issue for multiple SOA projects Ive moved it to from +If you finish it before we release beta please correct the version +Tom recently added fieldreaderfieldwriterfieldaccessor to allow access to the Java fields +Committed to branch +I will add streamingsockettimeout and add documentation in the next patch +Was incorrectly marked fixfor +You should use ejb and configure the ejb plugin to generate an EJB artifact +updated with trunk +in section it says the required type parameter is an actual type the bean type parameter is a type variable and the actual type is assignable to the upper bound if any of the type variable orand in section is saysIf the producer method return type is a parameterized type it must specify an actual type parameter or type variable for each type parameter +How are you deploying your JDBC driver? Does it properly export the package? The bundle does optionally import but of course only gets actual access if that package is being exported by some other mentioned in the comments the issue lies with the configuration and packages being exportedimported. +Hi Sam is quite old release +Gavin we are also deploying a mirror in Europe should actually be live in a couple of days more or less same specificationsSteve +Many for being patience and sticking up to my bigger egoGlad we got to the bottom of this and actually discovered it as an issue with the. +verified by Vladimir. +In the instance items are loaded automatically if the view is visible and the cloud is selected in the combotreeWe therefore decided to drop this feature request +Ram What do you think? You think we should commit this to and build fixes like on top of this or Maryanns issue on OFFLINE? Or you want to hold off? At the moment Im thinking that fixes for will be big changes not easily backportedJimmy I added review over on rb +Right removing them is a Good Thing thenNo one will be happier than me when we get rid of this persistence stuff +Its quite late here so Im probably asking something stupid but why is thriftspec a tuple? Using a dict keyed on field keys would allow the fastbinary extension to support negative field keys IMHO +Looks like partitioner info is passed through but distinct doesnt use it +passes +thomas did this resolved issues for already released versions +I found that by changing the font it works +hmm +The mappings for the missing charsets were taken from a page on Mozillas sitesomewhere +assuming you tested this manually +Uploading a patch which corrects this mistake +relativ autolinks are always resolved against the Page +Thanks Steve! +wants an Enumerable Change get to return only userspace table attached an addendum for that doesnt change behavior but makes the code look more similar +Marked for +Awaiting the notice with the account credentials +I guess Jenkins wont be able to test this since the patch has both HDFS and changes +See SVN rev +Most probably this is caused by a bug in integration where the injection is done statically +Thanks again for your patches James +any update on this +See section of the core spec +Thanks +Zktreeutil anirban you need to also to grant the license to ASF when you submit the patch +Added sections to and can someone review Stephen P RuflesrufleHHYahoo IM stephenrufleAOL IM stephenrufle + +Thanks PrabathregardsNandana +At Bolsena people kept on complaining that by default we list too many SRS +The call should not return until runtime is updated +method signatures changed in Utility class based on Iyappan review comments and and it effects to this uploading the new patch for it +Please cross check and close the bug +It has a dependency on +Committed +on v but on commit make it a Large test instead since now it takes minutes to run +WDYTBTW I like the CMS its quite easy to addmodify stuff +Ill check it outThank you for the report +Working to complete this for all Types will take dayRegards applied under revision +That feels very much like optimizing for the wrong scenario to me +has CRLF in the bat files. +Committed revision +Perhaps that will get it more attention +In one of my local tests where I get this token locked error it is fixed by the checkins in cvs +updated the error thanks I can see the problem is that the mail folder gets closed and the mail message is accessed afterwardsYou can set closeFolderfalse to fix that problem +If not can you please clarify +in the rare case then the fails that transaction will be lost but other readers would have already seen the new value because u released the rowlockStack +Is there any reason that fails? It also failed in my machineDue to +Struts uses standard Java property files for resources +This issue is fixed and released as part of release +only substantial code change was to +Hey there arent any benevolent dictators around hereI always suspected we were being guided by selfish dictators +This can be changed by specifying another configuration file +Thanks Tim! +Resolved as part of +Note that column group is already marked deleted even though the deletion of a directory failed +Why there is no Jenkins report before committing +eclipseplugins +I have now a problem a simple +cant seems to add attachment to the issue +Ill try to refactor the code a bit so this logic can stay at one ve just checked out the project in a folder with spaces and that causes several tests to fail so I need to make those tests too +The transaction type is handled by a backing system +Closing +Thanks TedStack Your comments on the patch +Just renaming issue to reflect that its about a requesthandler my badwhat about something more descriptive? dismax is kinda cryptic +This change simply removes these scopes and any other dependency information necessary only when specifying the scope +No I think its a good change +Closing stale resolved issues +Closing since testcases applied with latest patch. +Ive tried to run with e flag but got no more significant message +fixed +on trunk patch +the current class is indeed quite oddcould you submit a patch? is now derived from ErrorFurther the generated js exception code sets the name attribute to the name of the Struct +Is it what you are looking for +Done. +Assigning to David as he has already looked into this +An then we also have an? weird +You wouldnt create migrations from HSQLDB but rather once you switch to Oracle etcThis works fine with and as I described in the previous comment its easy to work around the issue with the mapping. +I am unable to reproduce have just deployed a new SNAPSHOT +UI labels fall back to a less specific locale if a particular label doesnt exist in the users locale +Ive not provided them because Im not sure how they would fit in the existing test infrastructure the tests validate only the output of the annotation processing the tests use but they are not unit tests the tests do not need a Jikes RVM configurationTheyre also badly written +Please give it a try with the latest snapshot build +I am working through this patch now +any help is very much appreciated +dmytro can you pl run the Hudson tests manually and post the results here? Thanks +The doc looks good when will it go public +Sharing the JVM for small tests saves us from a lot of fork and makes tests faster +this is how it would look if the code for the background of the logo was changed to white +Anything else that is different should be a mistake +Some new exception classes defined for problems that wants to just committed this +Let me commit this one for +as far as my intentions with MAXVALUE i was thinking of the difference of a Priority problem vs a Priority problem +I dont see why you think this is a problem +What kind of addition +Im in the process of pushing this all out into open source so we can then pull it back in to our based branch +Thanks Jun +i am looking at how to fix that +Committed revision +bq +The patch looks good to me +So the infinite loop is in the method where it sets the title bar color which sets the Label color style which causes a repaint which goes around and aroundIm working on a fix so assigning the issue to myself +These use to encode two Arabic characters in one glyph +So this change will only give more flexibility but will also allow the current setup where ADMIN permissions are set only globally +com extensions didnt fix the cannot execute issue will need further investigation into whats going on +Relevant patches applied +There is now some discussion on a forum item I entered about this is not supported is +IMO configurations are supposed to be independent ie resolving all confs at a time or resolving each of them separately must lead to the same results +Hi OlegI think between yourself and Mark theres enough good workaround suggestions above that this isnt really an the Transformer with a Filter or Router or just using a Service Activator instead would all work perfectly for your timeGraham. +Kris what did you have in mind that isnt covered under documentationguidestutorialsexamples? +Patch to branch +But do not reduce the number of real seekIn both case lazy seek will add overheadsOf course when there are a lot of hfiles with different version of rows and you just want to get the first version out of it +I exchanged a few emails with Nicolas who was the author ofAll interactions should be on JIRA or on the dev or user email lists +Ive attached a that does not include in the form directory +I hope its not presumptuous to go ahead and call this shelved +This bridge code requires updates to IBM UIMA code so this needs to be done on the IBM side +Opening text in Chapter Overview now correctly references six modules in the diagram. +Patch for trunk +once these are addressed Recommend you add a javadoc to get similar to the other methods +Bharath are you planning to make the same change on trunk? The change to ipc should be covered by a HADOOP jira right? What does this particular change address? Should the exit value reflect whether the start was successful +Heres some more info +The first column for EPSG datum the second column for ESRI etc +my +With strict interpretation such dates will cause an exception to be thrown +I removed the previous content and added the page contents of the domain manager to the ACE +The client should just fail and report the error +You are right +Would our logs or the activemq logs be helpful here +note that nothing special is done just that jboss AS install has a space in the name +I cant see any patch attached to this issue +In that way we wont be influencing any scenarios that are working fine as long as no exception is thrownI will only make a change in to avoid this NPE so that a server fault is returned to the client by the default fault handlingregards DanI guarded against NPE and added the following warning text at setup for now +The explains how to use the example and how to reproduce this issue +Id like to add to the list +We could easily add a separate component for this +Btw this is likely happening because WFS store is not using the simplifying filter visitor before encoding the one has been fixed by Mauro and Davides work +Based on discussions with Brian and Carlo moving this to our next release +uploading patch for Yahoo! distribution +That was the only improvement which jumped out at me +This patch checks MHOME if this environment variable is in system get maven and then get repository location +Definitions for all style attributes are updated. +If you have a lot html or xhtml files and want to convert them forexample with you get problems with every table +on which host? neither nor have a +lets check for bad effects for a day or two +I just committed this patch. +Ive fixed the issue +if you want it in so close to final release please provide unithere comes the testThe provided patch consits of the proposed change the tests for itmf +We can close this issue now. +Im having this error too +Fixed in Spring Core +generated a straightforward patch to decrement entry length when error out entries + Username lqd +The latter property is one that checks internally and uses to override any xml configuration +Im not sure it could apply +I wonder if you have not used the newer file +One of the subcommands should probably be to move the outputs from somewhere like taskoutput to somewhere like ttoutputjobtaskHaving a plugin that lets us switch between the current implementation that doesnt change user ids and a setuid implementation sounds reasonable +We need to break this out into individual tasks with will be part of for in progress by Pawel Rysiek a bit different from the original concept. +Small simple patch to fix +Werner done as well +Discussion to follow from Ken +Im of the opinion this should not be exposed to users for the same reason we dont support user selection of the interface its too easy for a error to screw things up in interesting and difficult to detect ways +Any more feedback on this +More new files srcorgcodehausgroovyeclipsecorecompilerinternalantlrsrcorgcodehausgroovyeclipsecorecompilerinternalantlr new files files parserlexer with antlr some commented code added an exported package to the and added a dependency to since it is imported by the runtime +were discussing now +Add the CLASSPATH only when running hadoop command +Elvisso is your completely wrongWhats your problem with it? Do you get an empty command object or just a wronglypartially populated one +The problem i have is that i need to run builds from the parent or the module and get the same result +We will the problem with new credential storeThanks we it now as credential store is available now? or you want to do it later + +This seems to be very closely related to work being discussed in +It would also probably be good to not include the database connection details with the issue +Assume closed as resolved and released +I will fix it today and commit the fix +Can you recommend a namespace thats OK? It seems like we need a namespace which belongs to Apache but not to Derby +I am ready to help on this in the sample contains a jsptaglibservlet future release will contain a application as well. +Committers can you please review +Mark thanks for clarifying +Refractor the and use as the package name +When I tested locally it passed x without fail +Would it make sense to use a similar approach with removeApplication? If not the submission should probably be modified to use whatever mechanism is addedIn MR did the administer control who could view jobs in a queue? My impression was that they only controlled who could kill jobs and change their priorities +Dont use the +I was able to use the I just attached to use Tiles with Tomahawk and Struts +Looks good as far as I can tellGood for if you like +I have nothing to do with the attached test cases so I dont know if theyll recreate the problem Im the assumptions I described correct? If so Ill see if I can put together a test case +I will upload the patch shortly for review +When looking into the TARBALL of Lucene I missed this great guide in the facet contrib directory +I have removed the ones that you pointed out +He has to recompile them at each access to the application after every restart of the serverI give you the catalinaconf in attachment where those configurations should take place if I understood the documentation correctlyAre ther other configurationsconf latest version of WTP has been I cant reproduce this problem with latest WTP +It contains BLAS wrappers +Thanks Stu! +Working on a batch version also but thought I would go ahead and post the shell script so that people could review it +I need something to try now and use official release in the near future +thats nice! The patch looks good Emanuele +Thanks a lot Laura +The lists need to be moved from incubator not created as new +Well in the meantime I found my own workaround +So we would simply pull down the release versions of the Seam examples and run them +At least after my fix to this works fine for me +fixed in revision thanks! +Not had any feedback on whether any existing parsers work if they do then the fixes for the related issues will allow a suitable parser to be chosenIf a new parser is needed please with full details of the listing format. +It could have been done easily by override create but it would be much cleaner to use composition to prevent other unwanted methods from bleeding through the class +Logged In YES useridFixed on the current CVS Head as well as on Branch Username schaefera +Hi Tom all my accounts are tied to the same email address +Since the only way to build the Windows assembly is on Windows and the Mac assembly on Mac setting EOL in our SVN would largely accomplish that +Fixed. +does appear to be an issue in how I was constructing the request over the proxy +Do it in the opposite order applied +to branch and trunk. +We have the min vendor option now this issue has been solved time nope thats not it +Should be tested no next Ive updated only windows shortcuts first time +It seems that all the replicas are in state RWR and we cannot get the length of the block while the replicas are waiting to be recovered +Hi Justin can you review this patch + +Yes it does not contain at the first place +These docs have been applied to HEADMike +With these two updates the patch applied fine and the build succeededHere I apply the updated patch +I thought that this inconsistency could have something to do with it but it did not +Committed to trunk and branchThanks for the review. +This will probably be true for registerFilter too +I have made a test with Camel on Apache Karaf and the problem is still there +I just tried with Maven snapshot which includes dashboard snapshot +Rather than force all such jobs to update with the release of v it seems better to leave this cleanup task for the indefinite futureThe API has been deprecated and can be removed whenever the community feels it will not cause undue disruption to users +The Java server of which I speak is the Axis server +cancel patch due to need a new version of new version of patch v for a follow on issue to fix this class of problem and for all at. +David any clue about what might be going onSimon can you look in the logs and see if there is any exception logged when the scale line disappears? If any please report the fullstack traceOtherwise a data sample style layout definition woul be useful to reproduce the problem locally and so debug it +If this feature is no longer availabe how do I implement this in the source codeThanks +I know that to build an Eclipse plugin from ant we can use Eclipse ant tasks +Groovy over Java be fixed in for arrays lists and strings +The States example project snjeza attached needs to be committed along with the attached file and pushed to the build +thanks Dan! +Tests screwed up in that I didnt notice because of failing my build. +Automatic code highlighting added to Getting Started code highlighting added to code highlighting added to task was done for RHDS guides RF has its own task +Yep all seems to work as expected in thanks for the fix! +Personally I dont believe that is enough as if there are multiple issues you need to hover each of them in turnTo me the purpose of pop up like that is to say clarify what an icon actually means when you have forgotten rather than to provide probably one of the key pieces of information +A patch implementing the behaviour as recommended above +lowering severity since this is not a severe issue +What do you guys think +when this work is done +Hi Ferenc Im sorry but Im stopping my investigations on this subject +Sorry Luis I hadnt noticed your comments till now +In CP ER there is neither a or file to buildrun the examples +Chomeniklasworkspacesapacheftpserversrcjavaorgapacheftpserverftplet Chomeniklasworkspacesapacheftpserversrcjavaorgapacheftpservergui Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserver Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrctestorgapacheftpserver Chomeniklasworkspacesapacheftpservercoresrctestorgapacheftpserverusermanager Chomeniklasworkspacesapacheftpservercoresrctestorgapacheftpserverusermanager Sending Chomeniklasworkspacesapacheftpserversrcjavaorgapacheftpservergui Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpservercommand Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserver Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Adding Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Adding Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Sending Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Adding Chomeniklasworkspacesapacheftpservercoresrcjavaorgapacheftpserverusermanager Sending Chomeniklasworkspacesapacheftpservercoresrctestorgapacheftpserver Adding Chomeniklasworkspacesapacheftpservercoresrctestorgapacheftpserverusermanager Sending Chomeniklasworkspacesapacheftpservercoresrctestorgapacheftpserverusermanager Sending Chomeniklasworkspacesapacheftpserversrcjavaorgapacheftpserverftplet Transmitting file data +For this to work s loadState method should return an unpopulated object rather than null +Cassandra version have no such problem +Perhaps we could propagate the change up to them +mrepositorycomsunjerseycontribsUserstucu +I did not add the folder so the service listing is not working +Ok Nick +The default could be Text or or whatever you choose +committed as described above +In I don find the ngram filterstpeuss cygdrivecProjects grep ril ngramfilter tpeuss cygdrivecProjectsThis was a fresh checkout +The code appears to have been there since the numberConverter was first created it appears to be an overlooked when later fixes were made for max +try RC and let us know if there are problems and we can reopen the bug +That is a reasonable alternativeThanks +This is a problem with the configurator system +Great! Ive applied the patch and it appears to fix the problem in my manual tests +just committed this. +Looks good to me +if you wrote about this case cause I did not dig into details +handles overwrites adds mirroring when files dont exist +Im having a look now +Does this code require Scala? Or does it support Scala also +Some of the viewfs tests fail under certain configurations of home directory and testdirAttached patch fixes the bug and also correctly set the wd so that if tests are run under eclipse they do not create the temp dir in thehome dir +Minor update on the last one +Therefore since this issue is fixed can someone with the necessary permissions close this issue and mark it as fixed +Lets do that for issues to the latest version. +Thanks Tim +This issue comes up when there are a lot of plugins installed in Sonar so the URL parameters make exceed the max +Thanks Gavin! +I will try it also tomorrow +This is also an issue when trying to serialize using since it violates the spec +Regarding the problem is that the exposed interfaces references classes that are not exported +If it locks up well lock up +That ends up changing behavior for POSIX builds and mingw builds which wasnt intended +is initialized just near and is later used to set might as well do that from the get go and drop +Still need to test the date related functions +Oh no +Its a bug in include for template page was changed also include template was adjusted +Im still seeing this problem with snapshot +Youll get a null exception +Thanks Christian! I will investigate these usages +Reassigned to Keith for review and applying. +I did the first part of this and cleaned up the Message SPIOpening another JIRA issue to track the second part +This does not happen in the new build under Spring and for Spring versions beginning with will be ommitted altogether to avoid these problems. +Thanks! +We can provide link to pdf help file is fine as a solution to command line help +I just committed this +Damon it would help if youd create test project for this +I create the project programmatically so it is by default the faceted project here I do not want to have additional dialog the dialog hang on the faceted project +Note that using WSSJ via configuration is essentially deprecated +I hope someone here can help to look for the problem or potential workaround for this issue +Would you be interested in taking a stab at this? issue type to improvement given there is technically nothing broken. +Sorry that I broke the build +Need to learn a bit more about Maven anyway +Running tests will commit when tests pass +trunk and +We need to switch it over to use journal manager to find the edits and stream them out +I suppose hibernate team has been made aware of this right +Does this sound reasonable? Perhaps we can break this issue up into several smaller issues and have some dialog on each matter on the mailing list +This is on RHEL using bash shell +Committed revision Thanks for the sample program +i think i may have been confused here +I will add a test but would also have to add JPP as a requirement +Thanks has been committed will be picked up in Fuse ESB. +There maybe another option +Moving out of +My simple jQuery checked with Version service the same problem +on this now just about have it fixed +as it was committed to trunk and someone please update the Fix Versions fieldThanks +Closing Robs internal JIRA Please consider labeling such as task so QA do not have to spend time on it +will commit this later today +Fixed please help to +Certainly they factor in to the equality check but what is the real harm if two perfectly identical schemas with slightly different properties end up in the same hash bucket? How often will this actually happen? I cant imagine this would significantly impact hashing performanceEqual objects should have equal hashcodes but equal hashcodes dont imply equal I question whether properties actually need to be part of the hash code +excluded the scala archetypes from problem is still there for example why the projectes are of different version? There are different guess you are referring to the repository definitions inside the pom? that is no problem as the GA will move to a public repo +Well fix it. + SHARE PRINT buttons has pictures this is normal page look workspace name is solution isbq +Ill apply this in assuming lazy consensus +the patches with grant done +anyway +I think this is what Brian tried to explain as well +The only suggestion I can give you is to try to install fresh copy of Eclipse and then try this plugin up havent I told you that Maven plugin is NOT using Lucene plugin from Eclipse +Patch applied to trunk and and branches. +It is more an improvement instead of a bug +Submitting patch for first +has this change been commited to svn +Correct +Alessio the source document is part of an internal repo for the product docs so probably notBec can you reapply the missing changes +Just for clarification You hold references to Class objects loaded by your in static fields in other Classes loaded by yourwhy would this keep the from being garbage collected? Is this an implementation detail of the Sun VM or am I missing something? I dont see where the reference is leaked outside the context of the +Let me know if any other sections are not showing consistent behavior since otherwise the other table views should remove the rows +The changes are in SVN +Yes I would agree that the converting to looks weird +For admins may assume that more than one special user could be part of the special group to which the group ownership belongs say hdfs and mapreduce are two user accounts belonging to special group hadoop +So tying the two together is not bad and would reduce code duplication +Sub task created. +Moving all issues to the CR +verified in CR +not blocking M so moving to Len for M +Hi AlenaI havent reported the second one as an issue +bq +Closing. +In r I reverted my initial commitAttached my changes and postponing until consensus can be reachedMissing things unit tests hashCodeequals implementationsThe feature may also be better suited for +adds a new unit test +When Instances is expanded Loading is displayed for about ms then Images are collapsed again and I am waiting till results are loaded +some day we probably should just raise the default internal max on a major release boundary to reflect increased default memory for most users or maybe come up with a better zero admin auto config for it +better covers this general topic. +Your patch appears okay to me +Ill transfer that documentation across to the reference guide for at some point +Moving to patch that wont display the up linkimage when it is the first item and wont dispaly the down linkimage if it is the last item +Okay I think I got it without dependency to spring +And here comes the clueBy not providing a scope bigger than per WAR we DO NOT HAVE THOSE AFOREMENTIONED PROBLEMS IN CDIWe could go on and define the behaviour for scopes larger than per WAR in CDIThe EJB EG would probably need to clarify the restrictions for injecting into shared stateful but indeed that is not so much a problem of CDI +The rebalancing feature will use the same chooseTargets method that implements this rack awareness and so the replicas in that case will conform to the logic +This is just a dev snapshot things havent been cleaned up and there are no new tests +The status of is still in progress +Hence the push for modules +Bruce et alias it looks like no action has been taken as a result of above discussion +Presumably fixed as is out now +patch has been applied thanks +Accepted for +Fixed in r +Tell me not working URL please +I agree a shared batch sender style of class should be usefulDo you want to provide a patch file plus some tests or should we wait for comments from one of the commiters how to proceedIm happy to provide a patch file though I did have difficulty building the camel distro +MR adds the port information to which is used by this patch +Can you revise the patch to be adapted into the current trunk please +temporary so I can migrate the reporter has been migrated closing again. +Steve this seems to be correct wanna fix it +Its only missing from +If were seeing tests fail on Maven or but not on Maven I think we should get some guidance from the Maven team about what to look forIll send an email +The launch config will automatically update itself with data pulled from the server and runtime objects + support is already implemented in Cargo +committed +We can this when the reporter confirms the error. +Moving all issues to the CR +thank you all guys for your time. +Is this anything to do with parsing command line and providing help on optionsNot so much command line +Reopen or create new ticket. +I havent tested it extensively yet and its not very elegant so you probably want to rewrite it and Im not going to put this into patch available yet +Thanks for the patch +Without this fix an event or could be disabled in JIT when another one was requested to be disabled +I move all bugs to closed after the release is done. +Rebased and attached +I have a fix for this and am testing it now +I just verified that will fix this issueThanks Aaron for the patch marking this as duplicate +Thank you very much for pointing out this problem +I expect the full jenkins run will identify a few tests that need to be updated here +Ive committed this and merged to and . +You said Youll see that the text box left of Add to cart is still enabled +We could have an action that changes and anything else required at once +Removal of the extra copy of is just cosmetics +I will commit this later today unless objection. +Ive never seen this +Hello Debraj Can you Please give more details to produce the bug +As mentioned Im using it for visualization +Updated patch +Ok now I see what you are getting at +Was there more outside of what was described that was in thescope of the jira +ChengThanks for your persistenceI will run patch v on LinuxCan you tell us more about the bug you found +Good luck! Leo +Closed due to the release of Tiles +this patch fixes the issue + +Can u check with the latest hudson build +The test contains workarounds for those two issues +I experienced the problem with Maven on JDK +Will commit after tests pass +Im running tests +qpidjavabrokersrcmainjavaorgapacheqpidserverqueue has some commented out code it should just be removedLooks ok otherwise +Im closing this if you have any further issue dont hesitate to concact us back +I just walk through the open issues +opening again to attach diffs for latest emma patch against base emma is the diffs file against the base emma relase for the patched sources which contain both local var and stack map table patches +Performance is improved on both widgets hotspotmetric and tested! +Slightly modified version of patch that moves Hive version numbers from to checked into trunk branch and branch +Just goto the website and download the dita output in one of those three formatsThe nice thing is if someone wants another format to output then download the sources and add the output wantedDocs are always a pain and I see the current solution a heck of a lot nicer than anything else I have seen +This is a branchx issue but not a issue if its the same problem I complained about +non reproducable. + +All unit tests passed +This is a only change so Im comfortable committing it without +There are no new test cases +We had some such issues ironed out last year +one possible solution is a service that includes all the products in the store and an ecas that creates an automatic catagory like when the first product is added +This patch adds the session id to all the mapreduce related metrics +I think TTL and persistent status should actually be configurable +I wouldnt ever recommend adding the eclipse files to the repository +does it integrate as a report +Thanks Tom for reviewing +Please provide a quickstart application that reproduces the problem +settingif someone volunteers to comes up with some smart and patch thats cool I just cant visualize it now +No we can close it I think +Postponing to future version +Need datanode fixes first. +contains changes in GC +QA has verified that the patch to the branch fixes the problem they were seeing +is there to cover the retry policy injection so Im going to close this one. +From looking at the history am I correct that we dont have a reproduction for this issueMaybe Mani could check if it occurs with in the hopes that the fix for fixed the issueOtherwise maybe we should close it cannot reproduce +Thanks for the info Ive updated my patch to use the right things +Would it make sense to just migrate the subproject pages to the logging wiki +on CTR for test only fixes +Nice work Horia. +The patch looks good +I just made a change in my own webserivce for that +Patch for this issue +Else the release will take a long long timeAgree for the +Marking as resolved. +Please cross check and closeCheers cross checked. +For my own thinking Id still liketo have a queue of deletes per SR and for the ram doc buffer +Thanks for enhancing to show user specified optimizer hints are being usedTransmitting file dataCommitted revision +Having discussed with Frank will fix this along within is fixed +Since we already have a keep alive connection to the bus the server side yields HTTP because concurrent polling is not allowed +Removed the initial listing of results from the Issue Description not relevant now +This will force the macro to render its contents immediately and therefore the correct data will show up in the exports +Whats happening if you type osgishutdown +I can think of two possible solutions for this +I am marking this issue as duplicate instead +After talking with jdcasey we agree that it is best to fix this in +Could it be related toIn the ivy console does ivy complains about failing to load that file? If it does which file path it tries to loadAnd can you test that is the bottom of the issue please start eclipse while your are in the directory where your settings belongs so the current dir will be the ivy settings one and probably Ivy will then be able to find your files +It was supposed to change but you are right and it is not there +I should add that my biggest complaint is adding a new class in the mapred package that is only used by a library class +it might be bound up on ns table but because this one doesnt timeout it goes zombie we do not get surefire output so there is no log to examine +bulk defer minortrivial issues to +Make sure that when we create app clients we have them point to the right port +Would anyone want to try this +Currently its EAP only issue but this might change +Your network is probably more stable than ours or youre not using as many workersThis patch should really help with tracking down network issues which is impossible now +Committed revision . +Please apply the patch for which solves this bug as well +This is complete for +Here is a patch +Would that be a good way to test for the optimization? Or any instrumentation you are already doing from unittests +updated patch not bzip patch suppresses the warning for the entire classIs it possible to be more specific +Ignore the changes as changed the so that its specific methods could be invoked prior to the more generic methods +I guess with inheritance we wouldnt have to implement the forwarding functions +Hm +As soon as Spock is released well cut a release compiled against that and drop support for Spock which will resolve this issue. +Yes there were a few more +Based on the commit resolving this questions from David be answered +Verified on build confirmed the patches and are licensed for inclusion in ASF works +It seems to be a problem if the jobParameter passed is a long +Access to the internal storage of loaded components per Bundle is now fully synchronized and flagging of components already loaded has been redone to ensure components of a single bundle cannot be loaded twice even if the LAZYACTIVATION and STARTED events are fired in very close succession such that the component loading takes place quasi concurrently +If not reopen. +Marking closed. +my excuses to the development team +But I think we should get a rule from jboss ws team for the naming method from a porttype name to a java class name +Proposed ve applied this patch please close if satisfied +Fixed by SVN revision +Bill and I diverged a bit since I posted a different patch to will post a merged patch soonish +It forces to set iphostport in the service +They should actually be modified so that they can run in as part of the jdbcxa suite so that setting the property will take care of testing for the case as well and we need not do this explicitly in the testthanx at subversion revision +If privilege is not found at then Derby will look for that privilege at PUBLIC level +We need to check if this is still an issue in +Thats it! Thanks +We dont want that rightYes this is a problem given that we use get + patch looks good +We should probably be able to support java andWe have to decide whether are just related to actions or other extension points as well +If we could figure out how those windows could in fact be exploited then we might be able to talk about a solutionAs you note if there is a way to exploit this window then it can be used to subvert the value of today +The Netcdf code hasnt made it into geotools yet +Moving out to move back to if this will be work on soon +Applied the latest zip service injection and refactored code so that sample can be run concurrently +I verified that the new test runs correctly in both a standalone test runas well as within the entire regression suite +New patch addresses all comments to date +Hence closing it +Granted this isnt as foolproof as a runtime OS check +I was going to look into it again and the original JIRA were it was failing if I could figure out why +Replacement patch +I would like to reopen this issue since the problem still exists +trying this with the latest snapshot the one on the tag works correctly but the next version goes back to the old one +The calls were in the original code and got converted during the autorefactor +I wont say I remember for certain because I dont +Thanks for the explanation! From what I saw in the source code at least the methods will have to be changed as they have a hardcoded version check that only goes up toAnyway it looks like the release is planned for June so I guess Ill just wait until then +Please see my wire log attached below +Access violation during broker cycling appears to be resolved with latest SNAPSHOT +Yes the matches the latest version which in this case is isnt really a bug as there will always be a problem if you edit against the web version and add attributes from the most schema and then later attempt to use them with an older version +I think we should synchronize fsnamesystem but not blocksMap +Indeed thanksMaybe this should be set in the default Android project generated by the Phonegap scriptI still think work can be done to display the splashscreen fasterThanks added the to the default +Thanks Kostis this is greatThe way of submission here is sufficientCan you add a link for other to read up on tidier +the issue is that it will call stat and readlink on the path heirarchy to find the canonical path +Please use the second version of this file instead +OK will close the issue once all are done. +Note that this actually tests and hence the uni test wont pass without it +Ive enabled reuse of local addresses for active data connection sockets in rev +I dont think I would mind removing final from most if not all converters +Futhermore DOES work as expected +We were sending error codes only for and but if there are other errors while fetching metadata those were just logged on the server and the client will not know +Good job! as you knows the product of the eigenvalues equals the determinant +Patch against trunk removes max handling from all the code +not sure if it still in now +would say to keep reporter information that we should create new jira accountsI am not sure why you need the extra Work Item field please explainI would use a bogus suffix or use if any users actually then use the account we can change it more appropriateLet me have the csv when you got it so I can test it myself on a test instance +The fact that the current filtering retrieves a object first is an implementation detail and it is conceivable that a different FS implementation may be more efficient to support filtering than based filtering +At the very least it ought to be mentioned somewhere so ops teams know to expect itThird this is spawning long lived processes that arent looping on exported functions +Close it. +If it does your cookie willonly be sent to port unless you specify otherwiseI hope that helps right +Assigning to Abhi either himself or any Xen hacker may give it a try +I have a working implementation runs zookeeper itest just missing destroy +We tested manually and compared the performance as shown beforeI have committed this +Going to take a look at this one today with another connectivity issue the client side of a hang is still present +Done! +I just committed this. +Updated reference to the API and Engine bundles in the launchpadbundles module in Rev +Not unless theres any objections +Ive committed fixes for this problem +Please verify otherwise I will close +Please add additional issues if the bean binding doesnt work out for some reason +Current patch is based on trunk +Thanks this now also fixed and we upgrade Activiti to Tijs RademakersWhen release Activiti? +Hi Suresh and othersYes I agree we can close this +Testing on a different machine shows commit r has lowered the BURS overhead in jython to under +What wasis a design motivation behind of this? Why Hive cant be instantiated at will by the client so different instances can be independently for query analysis job submissions etc? This will make Hive much more flexible and extendable from Hive client applications perspective wont it? +GrantRevision to myfacessharedtrunkcoresrcmainjavaorgapachemyfacessharedrenderkithtml fixed the NPE all Selenium test passPaul Spencer +bq +Thanks Mikhail Alexei +This is large maybe pages +Wont fix without further review +MaxI have done what you suggested. +Thx for reporting it +resubmitting +In hadoop the namenode does not contact datanodes while doing lease recoveryIf the namenode is rebooted the filesize continues to be displayed as zero +Pushdown is just another functionprocedure with special validation in the source modelIn if you want get rid of the function model then you need to introduce function to relational metamodel so that one can define a function in the view layer with similar semantics current function model +Ill try to tackle this today +FarahDo you want to try tackling this one? It should consist of just adding an additional property to the s descriptor +Not an issue at all +Thanks Owen! +Ive tried cleaning up the repo again and building from root and from +All the jars in these libs are included in Hives classpath +This patch adds line numbers and file names to import error messages +Since I am currently at a conference this has to wait to next week unless you attach the bytecode of the class for all three cases here +Not sure why +And the println was there for consistency with the other tests +ThanksI will get something going so we can have a working VM soon +Change docs so that they reflect location of in srcsite and add some report config +Hmmm +I committed the patch to +The jar was downloaded to my c +Docs updated according to JOSSO changes in. +Local filesystem might be better as default but looks like more work needed making local work again +Charles will the remain in the root of both archives? +Thanks +I broke this behaviour in a previous patch and have reverted. +Both approaches are not very appealing +Cant assign back to Len for comment +lots of choices. +This bmp appears to work fine in other contexts but cant be read by Sanselan +Sending srcmainjavaorgapachecommonslangTransmitting file dataCommitted revision +The change is too trivial to have a unit testIve committed this. +Thanks Harish for review! +pull request verification is trivial +update Do you have a patch file generated from svn repoThanks will provide the patch next week +Bulk closing stale resolved issues +to protobuf tracking for the release +I see a couple of in the patch +Maybe we can handle that as a follow on. +ok for I suggest you take one of these examples you found and implement and get feedback on that one instead of doing the whole set without feedback +I also dont get the long to int mapping and vice versa +Disappointing that this doesnt work with HTTP Auth +be EDT not EST +Thanks Paul for help testing this patch. +This is resolved +generated with spring roo without repositoryservice layer +Added warning to the cli documentation instead in r. +Many people have found for instance that tends to be over aggressive with locks when the isolation is SERIALIZABLE +Only java identifier characters are allowed +Oles tinderbox test has failed sincethe of this patch +See the Contribute guidelines on the Spring Social wiki for github tips +Doesnt matter if you install more than one plugin problem persists +Works perfectly +Ill carry on looking at Spring to see if this is reproduceable +Ill send out the CL later +Unfortunately I wasnt watching it so I didnt know about your comments +suggests me that you are using RI shipped under the sun +Thanks Ning! +Thanks Rajesh! +Thanks Sanjay for the reviewI have committed this +The issue is easily reproducible but not always +Yes +Is there a manual test that be run with MR for example to verify this as well? how did you discover this issue +Sorry +I believe that operator precedence is a parser issue so Ive added that component +The parameters in the query might be taken as strings instead of an interger +Would you please attach the project? Use the the Roo backup command in the shell +This is a workaround in the platform build until ESB has been updated +After updating jbpm from to the error occured +The Relationships implementation needs some work to make this actually work thoughI was going to talk to Heather before finally concluding the right thing to do +I never thought about that +Attaching test program to reproduce the issue +And I think Robert might have removed the dependency when he moved to a module +patch applied +Attaching file that generates the NPE +The Locale stuff in is of little use to localize an app since it only gives youthe correct localized name for months and other words it does not formatcorrectly +Fixed the forward slash. +I have not reviewed changes to the documentation needed due to my recent changes +Thanks Arpit that is good information to have +What version of the library are you using +makes closing a call abandon on each file that was not finished +Not having Infinispan source code available in my IDE makes it very difficult to efficiently code using Infinispan +The keys are host and host and values are the corresponding NN addresses +added into all places where is drools and JBDS works fine with for EAPEWP ER with JBDS +Thanks Chris for filing this Jira +New screenshots for modeshape publishing wizard +The Datanode then reads the header determines the offset and length where the specified blocks checksums lie then reads the checksum data and validates it +Generalized the ticket a bit +I sent mail to the developer list asking for help +Thanks Ive added some unit tests for that feature +Thanks. +and +Rewriting in JNI probably explains the change in suspend mode state +Id just like a second opinion on this +pushing off remaining issues in order to release version in JIRA and create release notes +Also foundfixed a bug wherein files with special characters could not be viewed via the web UI +Actually Im wrong this isnt a text file but looks like it wasnt checked in with a binary preserving can you take a look at this because its definitely broken AFAIK +Was already fixed +For TRUNK +in +A small and test case would be usefulThird please upgrade to and make your test against it +can you send in some example code for this? just come back to this bug and use the create a new attachment function +Then please to Corinne so that she can update the UDF manual thanks +I was hoping to be able to change the spring example so it could be deployable anywhere and not dependent upon microcontainer or jboss but that would require changing the example to use only Spring managed transactions +in any case please attach sample file since the used headers is what decide what is enableddisabled +I could be convinced to make it a warning but in my experience warnings get ignored +We are only fixing bugs in is unless you provide a patch for the feature request yourself. +Is there a way to get Darwin to output the correct time? Or could we hardcode a string representation to get around it +Is it allowed to have full functionality if both projects classes are available but limited functionality if they are not? Just an idea +This probably can be refactored +Resolving obsolete issue there is a getting started page + +Thank you. +Fixed some more issues +In the mean time attaching a reduced version of the patch for which just removes exception in double caching in cache +This Ofbiz component implements tool set for Maintenance + is the same bug but in branch demontrates this patch fixes this problem and adds for version of patch attached it uses Configuration instead of deprecated in junit test +Tests passed cleanly for me on +It must have been a bug specific to openjdk +Any update? We can just move this to if you dont have time to get to it +should get a clean checkout up to date so that the m build will applied +probably related to snapshot dirs include a timestamp perhaps we should just move along if it already exists +I suspect that it has something to do with the removal of support but Im not certain +surefire log +Use a DEFAULT value in your database schema and on your class +See the attached screenshot of the project dashboard +It is just that you want to control the grouping to the reduce +Michael Yes please open the new issue as a separate bug and well track it independently +Leaving this open until that piece of the issue is corrected + Will commit shortly +on the iterator based method additions +Closing old issues +It also seems like WHENTOKILLTORECLAIM should be set to a multiple of heartbeats rather than a percentage of the expiry time +When will RC be released? Is it in any repo +This dependency has already been removed in CVS +Im running now with latest trunk plus +hmm +Thats usually Spring MVC but may also be your code for example during testing +It remains here in rc +but otherwise to commit what a great simplification +Ahhh well that fixes thingsusing top NEGATIVEINFINITY POSITIVEINFINITY makes all the tests passIf someone has some time I think this is ready to get a chance to look at this? Id like to commit it soon +So wont fix please +Committed the patch into SVN trunk with r and r and into the branch with r and r +Actually it looks like something is broken on that build platform +Changing fix version to +But if you alter the memory properties of a VM with tools and then create a template from it it will retain those settings +The patch should take care of the problem +into +So that should reproduce the problemIf you have a windows VM Im sure thatll fail for you +Add a junit test to my patch +Thanks for the pointer! Ill call a vote to change the list policy and report back with the result +Eclipse requirements target env etcAlso we should be able to build via using either Ant or Maven? Cheers +patch look goodI have committed this +Still looking for a good workaround to that problem +Thanks for the patch guys +For now I included temporarily invalidation of replicas with wrong generation stamp +If you andor others feel strongly that we should break compatibility for this we might consider changing that default but I wanted to get this committed before it became stale as it touches a lot of files +There is an extension type based on the base type above containing attributes on its ownThen the attributes from the base type get lost during code generation +Committed new version of the Get Involved page at revision +For lets do not retry approach +Do you use git or svn? If git you can just create a pull request on github otherwise just upload the new diffs to this issue. +adding a CRC to each log record should fix this +will commit fix shortly +Its in trunk +resolving as fixed in both branch and master +Will this VM be possible or should we look elsewhere +I just tested that this error occurs in Cassandra too +On a mac all slides display as blank +Instead the API has been changed slightly to correctly mimic the scripts +Thank you +Sorry I found that a prefix is still missing in the outputNeed to continue +From my perspective the Project Manager behavior is wrongDesigns should be driven by use cases and real needsLets say Im a maintenance supervisor who is involved in a project to upgrade an assembly line +After this issue is fixed can you attach a link to the unit test case +Date Wed Jul New Revision +At the moment things will move a bit slowly because most people are focusing on the new branch +Committed patch in revision most important part of this issue is not fixed yet +That default was chosena number of years ago +I would like all rd party jars not in the main remote repository to be just put in a lib directory we have in CVS and declare these rd party jars as ones that need to be overriden +Can we change to? I still intend on supporting so that we dont break anything +Yeah I dont want to combine the NN and the DN into the same patch this patch is more about who does the recovery and who are the participants in the recovery +I have moved these errors to their own category. +for the patch +Another use case for these is dummy views which have no actual +Trying to get some meaningful dump in order to compare pass vs fail runs +Rick committed the patch according to comments so turning off the Patch Available change related has gone into trunk before branching of so it has been fixed in those releasesClosing. +If the reader is not sophisticated enough it can simply read the byte count and ignore it +This is a platformspecific issue with the headers used for logic generating +I remember an issue related to this one xmlns property is set on attribute map as +bq +This would definitely help as expect tests are more complete +I amheading to sleep now but Ill review ASAPJessica T +Thanks Giri stack Konstantin for helping with this +committed with suggested change to TODO +Please make this a high priority for +For atomic types we could extend the mapping directive to allow for three options string binary use defaultSo where we currently have ab we would support abstring and abbinaryThe default would be set in a separate serde property with a default value of string forThen something similar for compound types but with json and delimited as options? I havent thought about all the combinations and what to do with column familiies +TristanRadim +The only pb is modulesrcmainjavatestmodule error Body tag missing from HTMLTry to bump to for your feedback +This was committed in revision +I dont know if you guys are migrating it and if not +Removed confusing comments from vjoin that were copied over from vjoin +Diff of changes needed for filetoolswsdltofrontendjavascriptsrcmainjavaorgapachecxftoolswsdltojavascriptto add directory if one does not exist +It still wont be compiling the generated source but if you know what to look for in the results it will make it easier for you to step through your code +The original requirement was to allow the use of a Blueprint implementation other than Aries although doing away with all Blueprint dependencies would also satisfy the requirement +Im not sure about if not done carefully it will cause an extra memory copy on all the writes which isnt free +its time to benchmark it again +Attachment has been added with description Example how to reproduce the issue +Why is deleting the directory necessary +no longer exists and the new should fix this. +I have copied the following from comments on Derby today morningAnyhow the client will be made so that it also handles receiving when it had requested locators +I agree that previous versions probably need this added but this script is trivial on platforms that support bash scripts at the very least. +You should unzip and copy all files under the +I checked +If it does then A will not be deleted until the iterator is done since the delete is no longer done by compaction but by the referencequeue +This seems to make it possible to do the sort trick without the invasive surgery to the +Im more and more convinced that this isnt a solution +i will commit this first i need to rename some test classes and so on and dont want to screw it up +Oh OK +But hearing nothing is somewhat frustrating +svn repo temporarily frozen +I think that this issue can be closed because the work is happening on and +Thanks John! +We should add on the higher level may be directly into mapred checked is fixed by adding to raidIt was failing in the beginning but turned out it was due to the ulimit +So Ill leave it that way +Pretty sure I spotted and fixed the script issue here +Im using Fedora with KDE but Ive also tried it on Win b and the issue appear there as well +What attached project +This tracking JIRA cant be resolved until arent resolved and closed. +Looks like this usage of Introspector and can lead to a deadlock +Tests passed except for the heisenbug in +pulled from repo and integrated into our build +Its relevant just not definitive +we should get it cleaned up ASAP +patch looks good +people also use and that pretty much says that people want to filter rows out +I wonder why is different? +Bugfix Handling of null values improved Namespace of xml schema only one time in the envelope is defined for niltrue attributes for null values +Thanks Sean. +If anyone can think of anything else to move please feel free to just go ahead and move it. +LGTM to trunk. +Building via Maven from still works but not via just saw that that didnt compile at all and was excluded in the cdh pom +Duplicate of and +If this is needed for a layered product we can look into it +there was a mutable static field in this piece of code that got in there by accident then we removed it from the trunk the branch +Ive applied your patch adapted to the latest code base and added some unit tests +I didnt have the opportunity to investigate more so thanks for the work +However it looks like I will have to migrate from to before your code could be committedAnother thing +This results in potential data corruption but is not a regression +I think Pat has a project in called able that standardize the way of getting a generic webapp running box +this would be the framework pathUsersSharedCordovaFrameworksUsersSharedCordovaFrameworksUsersSharedCordovaFrameworksUsersSharedCordovaFrameworksCurrent simlink to Of course the file included in your html would have to agree with the framework you are linking to but this could potentially be an easier path +Added the comments by hemanths comments +I can not use TLFBut I has been completed project and work very I not want to ModifyI think the spark should use IME support logic not now logic +btw I think is thread safe and we should definitely use Weak +Great idea +Sorry hit enter before completing the text +Instead of asking a single replica to stream data from the range a new node is assuming it will have to ask all replicas that may have rows for that range to make sure it doesnt miss anyAdditionally each source replica would have to deserialize each row in the range to make sure its actually supposed to be streamed which would slow streaming down significantlyStarting to think this is not worth trying to support +So you want users to clear flows via some JMX management console? No I need a programmatic way to do it +fine in EPSG in I understand well current solution skips all user defined projections from file +The correct fix version is +Please make protected only those variables that you absolutelymust have write access toFeel free to the bug if you provide an alternative patch +This makes more sense also in a way that the now is what it should be a factory which generates the correct request type and enqueues it optionallyI have to run a lot of tests afterwards so expect the fixup code to be committed sometime next week probably +Heres a patch that fixes this Moved user from into Container RM gets from client connection as part of sends to AM as part of Container and NM validates the from the Container using +Good +Please apply this patch also to the branch +Looks like nice new test functionality +Reviewed all these issues and marking them as closed. +Thank you for the quick patch +Upload the current patch for review which includes both the new unit test and fix for +Ive committed the latest patch +Mikhail after applying your patch Ive got new failures under Debian Linux +This is a mere reference though as indicated by latest heaps Ive seen so it shouldnt take up spaceThe second optimisation would be to clear and shrink the underlying buffer of the cumulative buffer used by the replaying decoder +As requested here are some selftest patches +best to keep heap setting lowThe JVM doesnt generally use more heap unless it has to +why this is not a standard eclipse toolbar? why it look like original vpe? I see too much empty space +Some confusion on my side here +Ive just committed this. +Patch for the code +Xing this patch doesnt apply cleanly with the latest trunk +BTW the intellijidea project files are also checked in for the same reason. +Added some text for all the value of the +Gary Helmling also talked about this issue in the context of how users might implement protocols for custom Endpoints +Its not the synchronicity per se thats interesting so much as the part about making them rather than so we can guarantee consistency via the commitlog +ant passes just fine as my normal user without even a setuid +I have not observed any other issues with the special build other than ones already covered by other tickets +Absolutely pardon my ignorance but how do I go about findingbuilding it? repo doesnt seem to have rake task for building that gem and Im not sure what the right process is to get it out of the repo +verified on +No further comments closing. +to +Now that work is well underway with AS all previous community releases are +I had a look at this some time ago and the upgrade is not just a simple matter of upgrading the dependency +I havent had a chance to play with this yet +oh and the security exceptions that get raised if the class does have a constructor with the right signature but its private needs to get caught and turned into something should also comment that the initialize method would allow the plugin to start any additionalservices that it needss flush is a method that should be controlled by the sorter and not the framework +Very simplistic but heythis whole class is one single hack +Seems to be a Windows specific issue +Seems to be ready for the verification +Ported from trunk to branch at subversion revision . +Retriggering hudson now that was committed +Good morning + to trunkThanks for the review Nicolas. +Update another URL to hudson I missed before +Scheduled as a for +targeting for I am moving entirely too quickly and posting comments to the wrong jiras +Would it be benificial if this is reported in the issue +Against svn +Upgrade to CXF revision +Patch that adds the fix to patch tests for the length but its not necessary I think you could just use the addHeader all the time instead +changes the word grsser to greradds in r. +OK to be clear can we do this all with interfaces? All clients extend which gives you get and get and we also generate new factories that extend which has the two getClient methods on itI think this would be a great improvement and a solid step towards cleaning up the generated service code +I get the same problem trying to install other components as well +Slip to Beta which is when well have out first Early Access build in +Seems like the error I reproduced was slightly different +Closing old resolved issues +is going to add dependency RMAHS +Could we make some of the member vars final too? The reader in or the defaultAnalyzer in for instance +What your patch do is check whether the mep is correctly set for the operation +if it is put under launch shortcuts it can automatically create new jobs that you can save and connect to and it can even find the best match of the existing ones +update to receive data from client +Thanks +Thank you Ill get it sorted out +Currently this commit has triggered a deadlockSee for details +Verified in +The async version will be in the next patch +I found I can get the exception stack traces by setting the verbosity level to VERBOSE +The Comment data input is still added tho in both data inputs and the data assignments in the bpmn. +i know its large +Grzegorz Continuum is updated +HiId like to have the solution to my issue pleaseThanks +generate segment +Im working on it +Looking at s Debian packages the change didnt make its way inDigging deeper its only on the branch +cout should set to be right +I adjusted the description of the issue as it was inaccurate at some point as Ryan pointed out +I think that if you remove this some people will not be happyNote that that is also the reason why hot deploymentupdate is not supported obviously you cant hot update services and modules in a repository embedded in a JAR +Could you share that wats the wrong approach you were having with designer and how you could resolve it +Please try with this patch. +Hi Shawn I personally prefer small issues +I must have forgotten to mark them as keep forever I think this was duplicate of as it would affect java also +I dont quite follow why this would be a blocker as there is no runtime dependency on javacc so it doesnt matter what other components are doing +The code I checked in last night may address this issue +Hmm svn diff is really strange +Can you recreate with a recent build of Java and attach the full exception output? Java is more descriptive about these errors than Java +Ported unit test to +Its existence means multiple +The bit about and was probably fixed in which is part of surefire release +I think the solution here is to serve neither nor when the plain +I suggest replacing with listener that handles both facets and +Check it out and tell me what you think. +I cannot reproduce anymore +They all pass +spring file in it +No it is tested if a resource is used as released! +Ive tried to get my head around how to write a testcase to parse a static response with a schema pattern facet restriction but I can only understand how to use the WFS client code with a complete WFS +Many thanks for the patch +At least it is not a good idea to reuse the same message ID +I suppose this should be questioned as to whether its an intentional or accidental change +Still the problem is in indexing while indexing the all the characters are getting garbled +Thanks will disable the feature on metastore itself +Looks okay to me +Done +Lahiru made contributions for all three messaging components to integrate integration is complete +Lars I believe the pluggable addresses that +I cant remember right now but there was something that detained me from doing this +Thanks for your help Andrea but we do not use Weblogic +bq +Tested locally and it works +In non ajax page both portlet window in URL Test should behave the same +I manually verified the patch on a secure cluster but forgot to check when security is turned off +Thanks Army! I dont think you have been blocking anything I just noticed that there was a patch that could be applied independently and wondered if it was considered ready for commit +My preferred solution is destination level and I am hoping to add that as soon as I get the new addressing patch outPlease be kind enough to review the patch and provide feedback +Still need to either do away with the need for the committers group or modify the account creation procedure +For easier maintainence in future created a utility method in map which will do the mapping and return the legal Java variablenameModified Param and to use this methodRegards bug has been fixed in Xalan Java +No worries +for the to trunk and. +qpg +Change to class to recognise the special WSA addresses and throw a SOAP fault committed in resolved issues +None of our tests catch this problem so it cant be that frequent that it occurs +So this issue is finally ready for closing +mailing list svn area with restricited permissions would be nice +Logged In YES useridLooking further at the logs the Transport class should have been loaded from the repository cache but for some reason it is not +Thanks Jason! Committed the conflict fix to branch +binjruby case if JRUBYHOME is +Assume closed as resolved and released +Looking at the finalKey implementation it looks like you think top is the first half of the file the half w the lowest keys rather than the half w the higher keys +I checked installer and here is what I found I didnt change any port during installation Standalone starts with native management port bind to shutdown script works Domain starts with native management port bind to shutdown script works I changed native management port to in installer Standalone starts with native management port bind to shutdown script works Domain starts with native management port bind to shutdown script DOESNT work It expects management port on Caused by JBAS Could not connect to remotelocalhost +A few problems with the patch it undoes all of the Java work previously submitted +I have a custom built kernel on my laptop +Sure I will put the TODO FIXSo please go ahead and take a look at +Plus deprecate usage add to wiki as well the deprecation removal date +Thanks Christopher +Updated version of fix +Marking this for AS version theres a long list of users affected by this bug +serveralldeploymanagementclasses +If we can find the root cause of this we can try to add some checks to improve the error theory debunked running maven under jdk did not resolve the issue +Today you cannot do through because it will not route the request to the correct replica It seems like cannot do is too strong? There are plenty of where going to any active replica should be fineIn your optimistic locking scenario there should be no correctness issue but just a performanceefficiency issue rightThere are certainly use cases where hitting only the leader with RTG is needed so having the option is good +See attachment with my comments + +small draft is decorated by By Oracle driver +I just committed this Thanks Arun +Thank you KevinI have modified your patch so that the common case of adding a collection that is not the same as the receiver does not copy into a new Array +archivesTransmitting file dataCommitted revision +Applied patches thanks +Fixed headroom to take into ac user consumption not just app consumptionWriting a unit test +Upping the priority and assigning to did you get a chance to take a look at it +I am half way through it but I was trying to get rid of the static nature of loading the system function library so that we can control through the config +Umich copyright statements removedRemoved author and revision tagsCleaned up code and reformatted using Solr Eclipse codestyle +This is because I have the patch for applied +Here is a first cut at a patch for this functionality +Verified thanks +Jon +Patch committed to trunk +and closely followed by this info in the logFailed to retrieve stderr log for task attemptranother info there were not any stdout and stderr files for this attempt under logsuserlog +I use Spring in axis just fine without JAXRPC and IMHO I dont see the need for this +The version with all normal blocks and no block arguments works correctly because at each call level theres a new blockBody that yields and yield does not check for the situation is this +For some reason my svn could not add the bin jar file into the both are empty classes +Attaching relevant output of ant test from a fresh checkout +Ill be happy if someone can review and approve my patch or give me appropriate comments + proposed will be available on forge plugin and jdf site +That should tell you whether the freeze is happening in the SAX source thread or the processing thread and if theyre detailed enough they might tell you what we were doing at the timeIf its the former you should also check whether events are really being delivered from the parser or other +I tried ever combo of the x and h tags along with request and session of the managed bean +His name is not Sudha +No need on trunk +I eventually discovered the same thing and submitted an Axis issue +Wrong fix fixed a long time ago +Fixed it +This should be more documented as a scaling and tuning item +Im on adding the methods for boolean properties but would rather not create a new instance on each call +Ill check it out +Whatever I select I get the same project settings +Thanks Keith +Thank you The s should be thread safe by definition but most of them arent +Will you be using the CMS for your new website +Please verify this issue is fixed as you expectedBest regards Richard Verified +Is it possible for me to achieve thisRegards entity retrieved from Audit table is not a hibernate persistent object so we need to use hibernate to load original entity copy data from audit entity to orignial entity then save the restored original entity. +The killJob and killTask method should be implemented +has been accepted and committed +Other than that no code changes from chris work +From seems to be in the render Macro +Just downloaded mounted fine installed and brought up the expected examplesCan you try and download again Dean? Maybe you only got half the story +SAAJ complete implementaion +In my case it was quite misleading error message +No idea if or when this will be available since its you cant replicate it because your are running on a different OS than meLinux typically times out faster than OSX and even windows +seems this patch broke the now committed a new fix +Question +It automatically changes the web service name to Click works for me +Postponing this issue +While its not semantically necessary to preseve the order not all storage systems have a optimizer so preserving the order is to me desirable +Im fine also with renaming the parameter +the warnings for the transaction manager is another issue +There appears to be some issue with the cygwin environmentThere is no issue whatsoever for wsdljava code generation with the wsdl component of axis releasePlease mark the issue as resolvedclosed +Daniel could you post your test class so I can try your same test on my endcheers thought I would add one more comment +TedFor? We need to apply this right +I also applied the patch to and it works too +Thanks stack. +Im uploading a new one correcting this +As far as I can determine the problem is now resolved +How does a hacker do that? If he has already this access to manipulate the cgi scope he can do a lot more i think +use the basic authentication instead of form authentication in your war project +It checks to see if the statement is closed in addition to checking if it is null before using it +The module now contains only code +This is because the server that is running my and is resolving its FQDN name to and consequently the services are then being bound to and are not accessible from other machines on the networkThis could be addressed by adjusting the way the FQDN is resolved however my hosting provider would have to make that change and in general this seems like something that should be configurable in Hadoop +The contains samples for rpclit doclit wrapped doclit resolved issues for already released versions +orThis means that an operand in the code may be accessed at a point where it is yet undefined +So generation is not triggered without Errai IOC +q tests are currently failing ontrunk +The framework will be safe rightbq +I continue to be impressed at how quickly issues are addressed when they are raised. +Im not sure but at least the issue should be rechecked after the fix of discussion with Alexei we decided to stay synchronization as is and add special flag that allow fix problem +This could open up opportunities for UDF to interact with each other or possibly with other hive components +Best is to create a branch for this issue with failing tests and push it so we can look at see whats going on. +is fine Ive workarounds in place for. +However the thread that was responsible for the transformation is still alive and holding on to resources +In Java I was able to make thousands of requests per second using HTTP +Verified by Anton. +There is no automatic refresh of the views which makes the use of AJAX IMO quite useless +Regression tests added +blank Chris Thanks for help and sorry for my uninformed study. +Committed a patch for this +However is this patch final? I still cant get how thread safety issues can be solved by adding an extra initialization +Attached +But Im sure I remember it +I made changes to +Looks like were on the same page overall +This needs to be validated by more then one person +I cannot believe that such bug may be caused by the attached test +Youre a geniusVladimir could this be your problem tooThanks Mike +Is it a JMS or a JMS implementation you are trying to use the bridge with +LGTM +I need to create more JIRA issues for others to work on +Closing issue out as part of JIRA cleanup +Ill try to get it fixed in the next day or two +Any clear suggestion? +Now all complete +Calls targets to copy model and files using token filtering +tests remain failing due to fact is not available in the kits because it is NOT something we want on any platform other than AIX +bq +Bulk close for Solr +I hope this answers the question +Fixed major issues preventing generation moved generated source out of javadoc completed in revision +Provided the fix updated tests javadoc and codestyleNote the NPE is revision . +I see what you are saying actually it makes me curious why you would want to register the same class with different aliases? Although for map events i suppose that couldnt easily be you suggest we are already using a special property to get around this for now +Marking it critical so it keeps our attention +Ill open the JIRA +Patch to run jUDDI in embedded mode +Do with it as you choose +Maven works as expected. +Fixed with attached patch for opendapp Victor thanks for the patchPlease leave the issue open until its committed +Service xmbean +Thanks for the review Hitesh here is a patch addressing your fb +Thanks +p +Does that make sense? Same goes for since has stuff I didnt want to include it from within the section of the codeDoes that make sense +I will fix it still seems that having the field alone in the ITD causes issues with +I can understand that +Introduced a basic split into and +For rebooting a routerwith multiple connected that should be significant +The problem arises because Highlighterhighlightexcerpt is being passed thewrong array of Spans +If someone takes the time to fix this issue I will definitely review such a ll cut the release this WE if you can fix this before or need some time or help please I havent had a chance to work on it +It can only handle string +Move issue out as part of JIRA cleanup +Koen can you look into this? I know you and Slava worked together on this bit of code +if one is a string literal and the other is a bracket the literal is better regardless of if the other is much longer +added patch that mark race by special API for prevention of further that mark race by special API for prevention of further alarms should be was applied partially +Thanks VladimirPatch applied to LUNI module at repo revision rPlease check that the patch was applied as you expected +AdamFair point +OK got it I think +Actually the and zip files are not installers but simple archive packagesWhat youve done by launching the script is launching the serverThats itIt should be running on port +This bug has been fixed in Xalan Java +A thought this seems to dovetail w in that the dependencies we ship will vary based on the core hadoop package were buildingrunning against +Are they done +The coordinator issues the commit message but only replica A receives it +We could add this as a feature enhancement and maybe include it in +At this point I am not sure how to resolve the type of the closure parameter +jsftests +Borrowing from that theme of ideas we could also consider which would be more consistent with our naming conventions +bulk close of all resolved issues. +Even if it turns out there is no overlap we should still try to use a common terminology when it comes to naming the configuration parameters. +by this I mean closing stale resolved issues +Given that the Castor release build produces for each of the modules it should be pretty to have those downloaded as part of a Maven build. +I have committed this to trunk and +Then rely on the fact that the location of the domain configuration directory itself can be set via a system property +Ill do it +SVN revision +I will try to look into this asap +This is still VERY much needed +In your example it looks like as if there is more padding at the bottom but again that is expected as room needs to be left for characters extending below the baseline +I spent about two minutes looking at this so I may be offbaseThis appears to be a classloadingisolation problem +I will studythe tests this weekend and try to get them fixed up to all run +That is the Hadoop team who knows where their code lives +Encryption is typod as encription in many places in the first patch +It could be re using a SNAPSHOT thus I could update it on the server and your maven could download it But it wasnt the case +Is it reproducibleIf not its quite possible that its transient corruption +Yes I realized that a web interface is wanted +This feature is useful if you have multiple that use a common shared library of where in contrast is used to share within the modules of a single EARIm also looking at which is an enhancement to packagingExcludes to allow different selectors like regexp +I cannot complete this jira until we have support for either EDS or Modeshape in svnrequirements +I have not +Can you please replace the with the one that is in your Ganymede package download +Hey Doacan will you find a spare minute to commit this patch some time this week? Thanks a bunch +Done +Modules are designed to be self contained pluggable components +With missing plugin in this list it led to wrong default generation without false and that the reason of all error messages in eclipse log after jbds is started +The patch fixes this problem +Duplicate of +what counts as extremely large stores in triplequad count? million +Thanks Kim +Works some times fails s a stock distribution with our apps deployed as ear reproduce +Lets get this in if it works for you Sergey +We need to make sure that the XTS unit interop and recovery tests are also covered by this. +I like what I see +Thanks Ian for pointing out that one! Tests now pass smoothly +avsc is more recentPerhaps we should check in generated sources anyway because if we dont do it then it will be difficult to compile in without running Ant first +Jessica can you check on this is fixed with latest build +This is exactly what LEFT does so seems like a nice way to go through that state but we can of course do that in bootstrapping +It adds a HADOOPSSHCMD and YARNSSHCMD environment variable to select an alternate remote shell command when launching the respective Hadoop daemons +I attach another patch on +It doesnt fail on my local box has nothing to do with this patch and has failed spuriously in other builds recentlyIm going to commit this momentarily +I have committed this to trunk and. +Any objections? If not I would like to commit later todayThanks +the proxy properties file and proxy README are missing from the binary tarball. +Id be fine with adding counters but Id like to know why progress reporting isnt solving this +Please verifyAlberto +Ning can you generate the patch again there were many many conflicts when I tried to apply the patch +I should have an update by tonight +The resulting keys work fine for me all SSL unit test pass but they fail when run on Hirams environment +committed +Verified on +Thanks Andrew +just put lib jars in serverdefaultlib is working project under other application but no one attachment impossible to download after is it right +The background write usually updates multiple documents the same way +I had entered a couple other Jiras against false positives coming up in the LCOM score but entered this ticket as well as sort of a attackOn the other hand your response implies that when done correctly there will never be a false positive among the LCOM groups +rebuilt my project this morning and I still get a hang after I try to export DBSee attached +This cannot be changed by us and really we have no reason to care where they keep this small server file +the idea is when encountering on replaying journals we need to check whether the ledger is still existed or not +Its the approach I prefer tooThanks in SVNThanks Dan. +Thanks good call +Should be simple enough to do +What is coberturacobertura supposed to run? I assume it instruments the code and then calls surefiretest +Originally arguments after the first one of elt the can only be strings but for more widely applicable automatic conversion is added +Fixed in comments in for the fix +Markus whats the charset associated with the Sonar schema is it as specified in the documentation +I removed the unnecessary dependencies in the wiki +Do we expect this to fail reliably without the patch +Next could be the change to allow resumption with the same database then the database size method then fixing the insert logic to avoid primary key exception then adding delete cascade for the attachments +This seems like a separate issue from whats being discussed here though and so should probably be done as a separate JIRA +I did have a clever solution at one stage but I cant seem to find it right nowDavid G +for the patch +LEADER node created but with old session is a bug +Ted Husted Jan PM Setting to future pending a patch +Entry will be the recipient for the method call if it is an instance methodnull if the invoked method is static +The patch doesnt fix the second error where the application is rendered in the popup +Is Maven even released yet? Latest I see is a +Scheduling for thats great +Shane as the issue reporter please confirm that this is fixed to your satisfaction in the release that was released on Aug then we can close this issue +That would also be consistent with what users expect from a parameter labeled weight rather than sigma +Hi Markseems that you have fixed this already with shipped in +pull requests pushed upstream. +Brian Can you grab the latest Axiom code and execute your scenario with debug logging enabled for? Just before the there should be a log message indicating where the node has been discardedconsumed +It was a copy of s and very different from the one in +Proposed workaround is feasible for me +it seems it is outside Windows to me +Fixed log and rebased +It conforms to the XSLT specification +oh I dont know how generally applicable this solution is but my end solution was to apply an XSLT transform to the incoming XSD before it goes into the WSDL +I need to add more strict rules for indicating when its possible to consume some text output but it looks sort of ugly right now +ready becomes unused with this patch can we please go ahead and remove it +Just to elaborate has somebody already measured the gain from this one +And yes we have a case for enhancement request no portable way at the moment to figure in the custom mapper whether it was a in or out validation issue +The test order also changes when we move to JDK +mode has to be enabled explictly and then sane defaults for isolation and holdability are added +Extremely frustrating but I dont see this! I tried the insane jars the insane jars on my machine the on my machine and downloaded the and tried thatIm just running the individual testsWhich distribution did you useWho else is seeing not seeing this +It could be interesting to add a section that explain the login behavior and some examples of login portal instance authentication +This would help to reduce copying of the message content as well as offers potential simplifications of the architecture +If there is an en exception in the middle then it doesnt print information collected so far +Ill see if I can get one of our integration test applications to reproduce this so I can track down and verify a fix. +I was going to reply that it should be obvious that removing Clock from the API is relatively easy before we add a monster patch that introduces the only real dependency on it but talk is cheap +If we remove group commit then we dont need this patch +Null names are currently used by RPC parameter list record schemas so we cant remove them for records but we I think we can remove them altogether for enum and fixed +Stuff that implements interface +I was not aware of the class until now +And repository browsing does not show the proxy or remote the command line it works fine all needed components are loaded to my local just tested with your instructions and did not find your problem any more platform Linux meclipse Eclipse you check if the new meclipse version solved your problem +I have to do the release now so dont have the patch handy but if you want to go ahead with this you should at least have a look at the work I did back in the December it probably has all the basic machinery to build http response codes and already touched the areas that need to be modified in order to add the http is the jira I was talking about it already adds support for proper enumeration of exception codes adding http headers in exception responses based on the exception codesYou might just take it and adapt it to attach custom headers instead of the standard http ones +The attached patch adds the missing call +It would be great if you could have a look at this at your earliest convenience since its essentially just about intended fix improved to take care of annotation matching. +Without the actual db it is hard to know what is going on exactly from just the stack +The patch fails the test +It could very well be the case in recommendation applications that this measure is not appropriate I dont know +on Dougs last comment +Update the patch file and remove the previous one which is not in diff version +I think this one is actually a dup of with fix for +This should be pulled into the branch +The only feasible way of doing this is to have the cron run on eris wherewe can circumvent auth the protections on the files in question +modulessrcmainjavaorgapacheairavataxbayacomponentregistryjackrabbituser +It change the set invocation to called the private methodThe patchpluginpublic patch must be applied on the project +Okay Ive found a bug which is the likely culprit +i just cant imagine an intuitive way of selecting which command to select +It was a regression introduced over the last couple of days +Derbyall and passed +The has fluent builders as well but hey its nice to press ctrl q and have javadoc with links to wiki docs that end user might not find so easily +We wrote a Hadoop class that could read files directly completely bypassing the Cassandra server not that hard to do as the file format is pretty simple +Since activemq means to be used as standalone server +The average size of jobs we see is actually much smaller but it is not the average you have to worry about with memory +I would like more details on possible implementations of this feature +deployed it using the tomcat manager and caught the exception on ll attempt to debug it with a fresh checkout and see what I can find out +While you are correct you almost certainly shouldnt have throughput set that high because if you are tuning things correctly you will hit your operations count limit first for of workloads +Just wanted to clarify if this is ok +I agree when no length is specified it should default to the current MB not . +The downside is the whole job will re run however the upside is Users dont need to resubmit the jobsPlease let me know your thoughtsThanksMayank +Instead using new utility class named which will have all you need to test hbase +Defer to +add timeout for some existing tests +I will have more details have fixed the issue with a nullpointerexception when schema was specified as part of load +Would it be from Hibernates point of view if SLFJ targeted JDK in future versions +Ill try to mock those in the as well and see where it gets me +This issue is still valid +You can find it at srcexamplesorgapachehadoopexamples +Rebased on tip of looks good +Works well David could you just Remove the link Delete when there isnt any value +I guess youre both right +Also there are other factors like the highly controlled OSGI should be fixed +bq +first modal shown +Problem is with a loop counter in +Ultimately if somebody else gets a quorum the node should give up trying to become leader +Old versions of the installer have been removed +As per discussion with Eric I just committed this +Make the server side patch ready for this jira will commit it when there is a proper time +Were experiencing this exact problem +you ok with that? will move IF to another jira +Logged In YES useridfor some reason nukes overwrite if the files date is different than the one it overwrite +After first checks everything seems quite ok Ill try the dryRuns for RC and go deeper to make sure nothing has broken +in projects where I know I need to use another version of Maven I put a batchfile named containing a direct call to the desired you have a better or simply different approach to this Id really appreciate it if you share it here +Im not going to have chance to look at this til Tuesday +Pushed in . +I built and verified the generated docs +Just added another post +Thanks for the followup! Much appreciated. +In your case as I can see that is not the case for each request you have new service client +I also think that the format string approach is quite generic and nice +verified by Vladimir +Yes its because the oldPath could be from a different file system +Hi Scott +Thank you +The tests are still a bit weakI should rebase and commit itWe have a real world release this weekend though +What rev was this against? I get failures applying to system the patch to have the rptester method use a instead of a + +Houx canyou confirm that result? If so I think this may be the best resolution for this problem for nowLastly I think that we should expand on the comment in the test indicating why we are excluding thistest something like invokves the IJ tool in a manner that makes it hard to control the locale used by the engine for formatting error messages and the error text comes out in language which causes test failures in locales see +Going to commit today or tomorrow +It is quite safe because length and offset are guaranteed to be positive at this point +Please let me know whether that is good enough for comparing +It is normal for usrshare to be empty +Category is removed and entry is now moved without error +How do we reboot and do it right? Your requirements as I understand them are structured in such a way that keeping them is the only solution +The same problems as before +The patch fixes this behaviorSamant thanks for debugging this +Adds a class to mapred package the path filter which filters the paths containing logs +I have made the constructor of to private and written a separate public method to initialize and get the instance of +Theres a version of the plugin JAR that reproduces attached. +is the removed code basically a workaround for not having CFOF is that why its unnecessary +Are any of the projects developers in New York City? You can email me directly at wboyce at panix dot com +Committed revision +Will revisit after they are applied +The documentation has some good nay great information but at the same time the experience comes off as amateurish as though like with many open source projects the documentation is an afterthought +What do the tests of bug included in the patch have to do with this +basically after I fixed the bugs in xerces I makeclean in srcxerces make +On the contrary I see this failure on WS generated code only after this commit +I think youd better copy etccloudagent etccloudagent on kvm host then restart +How is netlinksafeonlinewicketweb linked withDoes enabling debug messaging on yield anything +The attached screenshot shows the content of the which contains and object +Better error messages? I can try and improve this +If you just want to run it out of the box the only path you need to create is varcassandra and nothing else needs to change +Whatever solution we go for should allow users to control what they want to usea They should be able to select JRE packages if they want toCan this be achieved by using provider selection mechanism +Tests passed cleanly for me on except for the known Heisenbug +batch transition resolvedfixed bugs to closedfixed +This is the patch for additional EJB information +I test it by set configure a gateway which was not accept connections +I agree I will add a comment about that +a simple fix +Crrently you do have a problem inserting data from thrift to a CQL cf even if you know what to insert exactly +Done. +bq +This ahs been finished for quite some time. +I havent been able to reproduce this defect +its better to just think of running the solr serveI feel the sameIn any case Ill go ahead and commit this as its unrelated +Furthermore we are deprecating in x so we do give users time to upgradeIll close this out now +Juergen was right in his first comment picking up the the parent propterty was a of type and not string +Fixed +Patch verified Thank youOliver +Thanks Julien and Eugenie. +is happening with a mutable and this one with and immutable natural id +Committed to trunk. +the two project I have provided patches have accepted them in their codebase now I need to have release containing the patches. +Right after newly installed plugin on Workshop this error shows up in the logENTRY MESSAGE Could not install bundle +Fixed on by revision and on vy SVN revision +Therefore moving to +So far no one else from Wicket development team thinks this change is needed +Correction to and accidental include of +are relocated to commons module and that contains etc +New versions of patches to deal with +All tests pass locally except test which I also see on plain trunk up on our Hudson. +Thx +Could you combine two patch if you are able to solve this issueThanks +Not sure what is an easy way to resolver the issue but a warning along the lines of Missing remote object library in get would probably be the best solution +But AES does NOT send the empty list so an target will reply with the wrong dataIt will reply with all of the data +slip to next version +It seemed straightforward enough +Done for IDEA +It provides also a mapping for +Minor changes made based on QE feedback document. +Should be fixed in M due to a Jetty upgrade +Modified the output handler to enable character escaping by default for HTMLand XML output +Changes ported to the branch as wellOleg +This is a regression caused by Metadata changes that should be fixed +Does GWT also support this +If you can get this working against a standard Apache mirror download I can start working on getting the files hosted on which has had better availability in the pastFor the refactor lets do it in a followup and also talk with the Hypertable folks to plan it out since I think they had to copy a lot of code also +Hi PradeepI applied the patch with revision acefdabaaThanks for your contribution. +So related to hibernate I would guess +Shown on first page of the installer +You might try to search the android bug tracker or report an issue there +Folks could then run the target from within a release if they want to build a tarball +But as you rightly point out well be constrained by the Apache initialization as well +Its what I tried first before looking at the code +Should have a patch shortly +Hi seems this issue is not very popular Can someone please review the patch andor comment thxcheers you are right + should note I wouldve just committed this but I want to add a test to the etap suite so we dont accidentally unfix it in the future +will commit if tests pass +Its great to see that there is some progress +version number fixed as part of releaseIm not changing to be updated from +This patch adds a bit of extra condition checking around the gathering of source folders for the project when were inserting classes +Another vote for updating ibiblio with +Two round trips to get the list in the worst case seems fine here because the list shell command runs on a human timescale +Moving out +But we need to change the readscan logic +to commitFixed and committed +Yes +It enables includes and excudes defined in the etc file +thanks Rupert that looks good +never tried to reference the class directly this was a class not found exception coming from inside Spring MVC classes on you use at all you will get this exception +Hi Abe +bq +Well maybe the server could call other servers cos it knows the endpoints but itd be repeating the same code as in the client +Hi JeffJust check with Jim who is jboss ws runtime guy as if the wsdl file is not good enough +It will be slow especially if its run as a single delete at a time in serial from the shell +Lets make this work as Malcolm describes + and great issue description +Fixed in you can use instead of on service level. +Fixed in revision +Closing all resolved tickets from or older + +Noted + +to trunk + +Patch correcting the issue. +An alternative would be to have some kind of standard Ajax request response text like Unable to complete request session has timed out or Unable to complete request user is not logged in instead of an empty response +Assigning to Benn to help integrate the latest plugman implementation into +OK +Perhaps an older Xerces release has its own in an endorsed dir onyour PC AntonioI have even more problems but I suspect theyre due to incompatibilities +I added a patch that removes any badNodes from the list of goodNodes +I committed this +I have tested the area where the failure occurred and we can now update cmr cmp fieldsWhew +External issue +Works like a charm Verified in JBDS Alpha +Lower level tests can be useful to either test more combinations quickly or for an aid in debugging +There is no global UDF function model allowed in the current version of the JIRA addresses both Larry and comments from above +Adds documentation of shortcut +Please let me know if anyone wants to change anything about this feature before committing it as an attribute +I understand your comment above above about the future of this functionality +That is essentially what I was experimenting with is rather serious issue and requires investigation across all available implementations +Thats well within scope for logj and a self contained application +Anyone know it +This certainly makes it quite useful +It doesnt work if its last but I dont know where else it may not work +patch submitted to upgrade from to dont have that information other than I was able to build cleanly and all unit tests pass with the new version +Yes you are correct in that things will not work correct with usrbinlink Ill apply your patch soon +Fixed in the SVN by make symlinks for the remaining tools beside htcacheclean rotatelogs and with the CR +Thats the idea actually that this is what you want to do most of the time not flip it on every controller method +Also ZK session timeout exceptions present +bq +Created for handling this +Can you add me to the group sometime? No hurry at all I just wanted to mention it +I recommend we back out first since thats the cause of of the failures and then work back from there +We should go through all the pulled in dependencies to check for cases like this +change default for edit attribute +It should cover quite a bit +So either youve made things a lot more efficient in terms of packet sizes etc +Conversely the difference between reading and writing m records versus m records would be substantialOwen I think the number of spills is more important than you thinkConsider these two cases +This problem doesnt exist in because plays the role of RM and AM +fixed the isoid inconsistency +The main sticking point is that a generic solution also requires a reference to the session registry in case concurrent session control is in use +Fix applied to branches remotingGA and test for cruisecontrol results +Attachment contains changes only +I updated both pages +the same for Eclipse and this is definitely Upstream there are several workbench windows are opened in eclipse Window menu shows them at bottom so you can open any from minimized windows +Closing the jira again. +This has not been committed as I previously thoughtIll get this sorted out asap +so rather than wait for someone to write that code Id rather get the signficant startup time improvement in +Its closing for +But if there is no ftl then make it so the SLD does in fact use the label color. +Would Captain Awesome be so benevolent to allow me any budget to hire a consultant that can help me with this incredibly important task +anyway its just one issue theres a problem with the calculation of the dimensions +This issue and related issue are same +Please verify +This robs the performance gain from the patchWith the path cache whys this true? It seems that so long as youre accessing a relatively bounded number of blocks the paths will all be cached and you wont need to unless a block moves etc +The last remaining issue is that libtool refuses to build plugins with . +The patch adds the described interface as well as an abstract implementation called with documentation on how to subclass it +Only those whom have created reviewboard accts can be actionedPlease provide a comma seperated list of rb usernames +Had to take out the absolute in the file refs on the patch but after I did that the patch works and fixes the compile problem +Initially I had an iterator over the list and the javadoc of that method also said list +r resolves the issue +But we might have a few interoperable releases in the meantime +that is be sure and indicate that you wish to assign rights to apache +It is disabled for now but should be enabled when the bug has been fixedCommitted revision +Committed to these branches at these subversion revisions trunk +Committed by Ted a while ago +The s inner class has a service that isnt being shutdown +otherwise just close it wontfix +Closing old issues +Thanks Sarvesh! +The implementation and test provided +ser from file name which cause confusion with testing serialization capabilitiesI hope i had myself explained clearly +This is one of the bugs that Paul fixed in hadoop for HAR to be able to work with CHIF +pushed to master and +So do we need to fork Doug Leas concurrent API or just depend on it +When Weld deploys such an application it first needs to introspect each class and build complex object graphs that provide class reflection this case the majority of beans extend framework classes which in turn extends other framework classes such that each bean contains many methods constructors and fields +UI changes are ready +Now its possible to return a Message instance and that will no longer be modified by the base handler class +This excludes list was actually duplicated with the task responsible for the projects packaging +Verified on revison +update project to use common publisher add property for vm args add dependency on drlvm the spec +Verified the test passes with the patch and fails without the patchbq +But what we do when use is true in is use the last xml resolve report which leads to no actual resolve process +Targeting and trunk +on the Can you please go aehad and commit the patch to trunk and branch +Please use the forums for configuration questions. +In addition the patch at the moment just pass unit test +But the proper property on a real standalone war file running in an app server but itself seems to be working now +We could possibly use this class internally and remove our samples we could modify the extensions so that a default zip url installer is created using a url from the if the user doesnt specify either a zipurlinstaller or a home only issue I see is that the could depend on the location where the use is located to benefit from the mirrors +Test case supplied so I can now reproduce the problem as reported +We are tracking the update of mochiweb in +If its minor in comparison to the original issue we tend to create sub Changes entries see some previous examples in Changes +Would it be possible to run this patch in Geronimo environmentThis patch +Overall I think the Ref +Thanks I think the new patch still has some problems for example the memory allocated for slash may be leaked and some invalid directory name may still cause VM crashIve reverted the original patch with s at revision r +If we turn out to actually have issues with incorrect linecolumn numbers we can and this one or open a new issue. +this will be fixed in a future release +what is the differents to my example? can you please modifie my example in a way that it breaks +Will try to do this ASAP +I see it is implemented +Even integration tests dont work +Please clarify +I like the callback idea +This has been committed to trunk +Is it punitive or political? Please explainI dont believe that Cloudera ever sold a copy of Desktop and shortly after the initial prototype was demoed it was renamed the HUE open source proje +I think theres enough interest in for us to start pushing in that directionThere are a bunch of issues closed already with does this so this issue is going to go the same way +I have a question concerning your unit tests The session is in In that case the messages shoud automatically disappear from queue after receiption or am I wrong! Anyway thats the behavior Id expect + is alternate approach for the same logic +Patch provided +It may even be regarded as a major breaking change if we cant think of any workarounds in which case we might push to +Also when a job does not define the tasktracker sets this to MAXMEM number of slots while localizing the taskThe patch contains some additional log statements that I will remove after the review is completed +We have decided to incorporate it into proper for since it seems like export restrictions dont apply in a meaningful way to crypto stuff + should never leave the project in a state where you cannot initiate the it can only update to a parent version that is resolvable from one of the current reactor the local repo or the current remote your example you are specifically using a single project reactor and the parent version you want to update to is not in any of the three achieve what you are trying to do I wouldcd +This patch conflicts with which I just committed +Would the same approach work for both languages? Would it be worth solving this problem with a generic tool rather than? +When creating the analysis engine from the pear specifier a must be used in order to propagate sofa mappings +clientExclude will override your clientInclude if you specify the same pattern on both +Patch +Thanks KevanI completly forgot about the jpda stuff sounds good now I have it all will test soon +Marking as wontfix since people with super latency sensitive workloads are increasingly switching to so if nobody has been motivated to work on this in the last months I doubt its going to happenIm not on the idea though which is interesting +The stack traces were also the same as expected from the bug. +an examplefoopackage foobarimport varName +Attaching a patch in case it will be helpful +djencks? +only if no injection takes placeWhen does the injection take place +I think that timeout should just be for emergencies +Hmm +Phillip if you do a full recursive search inside all files for in the environment where you run this it will tell you which jar or config file it is specified something is creating it and its not JBM +I am using the branch from git +Hey Aidan can you cast your eye over this again +These diffs show corrections to a few files in the introductory material to both the user guide and reference to remove Rogue information such as directory paths that were correct for RW installations but not for stdcxx +That sounds reasonable +well document that having a doesnt imply that a release is officially supported against that profileDoes officially supported we ran our test suite against a version and documented anything special? bq +Its just to have a decent default +Approved +I can understand whats going on and will commit a change that hopefully fixes this later Kenny this is indeed the same bug which is +I can only speculate that your browser is sending a different encoding for the body versus the form +No sense adding complicated loops that you have to explain to people when a simple System method call works just as well and brings clarity to the code +Added a check to see if the access node has been moved to prevent further optimizations. +Heres the trivial patch +while I receiving the above error +Initial patch containing the IBM SBC driver +Correct patch applicable to trunk +worth trying out +Wow making an example of the only class in the whole package which is missing it +moved to in +Status update Resynced to trunkFixed Cleanup synchronous completion case in line Creates a request for on each completion need to aggregate theseAdditional work item forgot to addo Replication of new queue thresholds for clusteringo Documentation updatesAnd still TBDo Performance regression testing tuningo Sends completion on every else hangs +Fixed in commit f thanks +Wow + problem lie in +so now I close it. +fails easily with or without the patchFailures for and should be fixed by sir +Adding a more informative javadoc as Josh suggests couldnt hurt +In the degenerate case you could end up with a lot of your maps having no inputsJoydeep would a grouping comparator like the one we use to group the reduce inputs work here? I assume it is the case that youd want to group on a subset of the fields in the keys since that controls the sort +This is a runtime issue +It skips D and ends up on A +Should be in next jbt build +Chapter look good now. +Attaching a second draft of Developers Guide changes and with changes as followsD srcdevguideD srcdevguideM srcdevguideM srcdevguideM srcdevguideD srcdevguideM srcdevguideA srcdevguideD srcdevguideM srcdevguideA srcdevguideM srcdevguideM srcdevguideD srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideD srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideThanks to Rick instead of having examples that all do almost the same things we have just one that illustrates NATIVE authentication and SQL authorizationThanks in advance for further comments +Let me see if I can figure this out +The two tests have been failing even after the commit of the patchSince no other bug has been filed to follow up with this I am reopening this jira +Is there something I can do +I see a big problem here +This works for me with the current Web code + using version and appear to still get the cross talk issue with SenderReceive Idv belaban Exp +Bad version number in +I could make the implement the interface so value change listeners could be registered on it and be notifyied about page index selection changes but im not very sure that this is ok +srcmainincludelogcxx +was an issue on BK but this one is an issue on Hedwig so it doesnt sound like it is resolved by + Ive moved this from the checkstyle plugin as it works in Maven but fails with +If you use Dtestfalse then it works +Fixed a few tests that did not need to create dangling links +I do not know the corresponding file in struts to check if the same issue exists +looks like there is consensus that we dont need for either or so rebased and committed +So far so good for me +Thoughts +But these bugs have been nonetheless fixed +The is redundant +Committed this +Need to look at how this fits into the jbossretro work +Committed patch with revison +It would be nice to move beyond version and at long last +I have started a vote +Sorry filing this was premature +NET which does NOT generateuse methods +Depending on the solution to then including the same file multiple times might become illegal +The man hours involved in patching all the plugins will be tremendous +Perhaps those tools have additional internal debuggingwe can enable to gather more information about the instrumentation and reporting phasesto see if we can figure out where the methods are being dropped +Howeverwhy did not the validation checker not catch this mistake +Do it before commitI think It is a minute work +Hey PeterThe correct file to change isThis file will automatically be used to generate and +Sorry about this +I think putting it under the would be most accessible for users +Running tests +Would it be easier to fix the failing tests if we just push this change in now? Then people trying to look at the failing tests dont have to manually apply the patch +Ok ill update the dont think taking compression as a model makes sense +Stddev function is now includedAVG and STDDEV are implemented using online algorithminput flag is now set to false is the reset functionI will now focus on the unit tests but where should I put them in the filesrctestjavaorgapachedrillexecrefropsAlex +Thanks to Yev Bronshteyn + is a duplicate of this issue and has a partial patch for this +Instead all Users and Roles are stored in LDAP and I just want them transported appropriately into + on idea after we have discussed Enis Jeffrey could you guys please share the approach for this issue which is currently under testing +Patch looks good to me as it passes all tests and have only negligible influence on performance tests +No unit tests are included because all changes are trivial +The log is from a previous version of +when copySucceeded is called +The recent change in was designed exactly for that scenario +If we want the compiler to be able to do that then please fill a new issue and mark it is a future request or as improvement +Removed debug log message +Id like to see if this addresses some of these vague concerns about slowness +The original patch turned out to be problematic and applying the new new one to would be too risky +Check also your other files +Could you test with the latest snapshot build +Need input on thatAlso moved the wire format negotioation stuff into the interface need inputMade the use the instead of the openwire +Indeed Marcus willing to share? We could refine it a bit maybe and get this into master and maybe evenId love to see what you came up with since we started playing with this yesterday as well +Ive split this idea out into since evidently its not a slam dunk that everyone agrees with and i dont want it to distract from the work already done here that i plan on finishing up +On a clean workspace before lcosing the welcome screen or doing anything else select layeradd and open the postGIS all the catalog plugins and none of them blow up any more. +Could you check the patch again +Teams like Oozie Pig and Hive are going to be directly impacted and need to be treated as customers for a feature like this +All other listed tests should be working now +I just spoke to Bob he isnt going to be in a location to do this until Saturday +I am not yet proposing I want more to see your opinions on short of being able to just specify the same args I would specify to javac is going to be confusing +Issue may already be resolved may be invalid or may never be fixed +After retries we were sending the query every s +is the real cause of the problems here and not part of AS but will be in EAP that concern to be covered in in +Hurray +HadrianAs the flatpack is not in the central maven repo so I had to use the servicemix wrapped bundleI just update parent and for it in revision BTW because there is no critical bug of flatpack I revert the the change of import package part +Why not just make it an explicit tmpBecause tmp is not necessarily created on a fresh HDFS either +Upgraded in revision . +This has been fixed in marking as resolved +I have applied these patches +Are you still working on this Ram? Do you think this needs to go into +Calling this complete +I can fix that when I check it in The test failures look to be caused by a runaway process +It doesnt recalculate element size if its content is invisible and content size is changed. +Please let me know if I need to change this +Passes the manual sanity tests +default to the module name makes sense to me + request send to synapse request send to tomcat by synapse definition of the proxy services +is not per se so closing this issue as wont fixIf desired a new ticket can be created for requesting means of supporting synchronized configurations. +It may or may not be a bug +Ideally we would want move these protocol classes to common and then derive it in HDFS and History serverIm planning to file a JIRA to do that +do you have a small testcase? It works for most of the projects I have at work and I am now having trouble finding one that fails +oh +background gradient on the header styles in org styles on nightly doc page in is broken on header in nightly builds +Ok sounds good +This hasnt been added because Seam Drools jBPM and AS can be executedcreated in any order +We should make the discussion in the mailing list there seems to be no Bug here. +I am running tests +Related issue fixed +Ok seeing as everyone agrees this version of the patch is a good thing but might still like to discuss the tunable parameter could I commit this and if you feel strongly another ticket could be opened about the tunable param? Id like to get this in and it feels like maybe the good is being sacrificed to debate the perfect +Keep partitioner and encoder pluggableSome detailed changese +Also x bit is required for directory to do ls on it I think +Let me try patch against hadoopqa +Patch is applied at +This is a patch for version +I just committed this + +I committed the patch to trunk +I can handle it in my own pom file with an exclusion +Not me +codegentargettest? Would that be an adequate place +The Xerces dependency might still be needed for has been superceded by which in turn led to so this is not an issue here. +Thanks Hari! +javadoc +Drop a file in farm and it has a state of unknown +Changes to core to check if state machine is internal or committed through VijayThanksraghav +Because the solution can be much simpler IMO +Or it could be a simple case of mismatched runtimes +i will take care of that as wellbtw the jcrprimaryType property is defined to be mandatory by ntbaseso having the check for mandatory items enable in the type validatorrenders that fallback pretty useless anyway. +But yes our average use is not sophisticated in that their average operation to local disk in our environment is just to pull some experiment results locally +fix applied +Another concern on this one is how to handle per field options since we dont really have fields for functions but still it would be useful to have multiple facet by function options +I mean if disable false second attribute should be ignored at all. +open a new if something got missed. +My maps are generated with the generics as expected +Here is the patch +So for the initial stats page a patch is attached that avoids enumerating the terms for the fields +ok so this does the same as before except any static initialization is not called before the first instance is actually realizedinitialized +More fixes in r +Fix attached as a patch +is back! Id like to apply it with this issue +This might not be an Ambari issue after all +Is that a bug or an undocumented limitationNice work guys! We should definitely get this into SolrI probably wasnt clear we are sharding and collapsing on a email field +When I look on codereview I see line diffs for new files +Youre rightThey are here in the Maven repositoryBut I think the initial purpose of the JIRA was to have them generated inside the target folder as a way to easily copy these things to in the case of a releaseThe extra step of looking for the correct files in the Maven repository is not that complicated though. +bulk close of all resolved issues. +Fixed the Camel pieces +Defer all unresolved issues from to +the user guide is also version specific +thanks! +One of the patches includes that so it will apply on top of trunk +Not sure if this will be desired in a final module but hopefully useful for running perf tests +My bad +I executed test again this time with repository configured to use DR tag on MEAD central repoNo change still failures +OK thanks +The first part runs all tests in the one jvm +If we dont do this then theres no guarantee that the jobs at the head of the ordering L will be the ones that actually require preemption +A UDF name maps to a class via reflection and Java is +Made the appropriate changes +Many many thanks Aleksi KallioIve committed the patch I put it into a maven module called and attempted to add it to the assembly Ive not had a chance to test it out thoroughly yet +Committed patch. +I will be taking over from Mark Niland after this month and would appreciate some feedback on your progress with this issue particularly considering your comment of th MayPerhaps you might consider providing a download link for the missing code? +I say ship it +Integrated to TRUNKThanks for the patch DougThanks for the review Stack and. +We hope to use a schema from as a sample test case for this new test type +that would do it +If we could get a plugin implementation for the scope description above it would make a valuable contribution to the wiki +If I see this right then we are basically talking about a groovy side implementation of JSR +I am in the progress of updating them to +I dont really like changing contribextraction to be contribtika since then it is not clear what the functionality is and we also may have other lang +Sample consistent hashing should sort regions by the creation time of regions +still looking at this one +This has been committed +Other changes Currently has action + tested! +I agree +Patch fixes binary compatibility tested this against the latest site plugin and all seems well +does not work in opera only in Opera +Any news for this issue? There is a public Maven repo where I can upload those artifacts +I dont see this as a test problem this is a broker side issueAs per Keiths original comment I think we should be unregistering the before responding to the client that the connection is closed as occurs for clients +I didnt see Herbs patch until after Id found it as part of bug fixes to unreleased issues that a non committer contribs towards we should add credit to the issue that caused the bug +That had slipped my mind againRegards PierreYour patch is in trunk r r r rI did not close because I guess we will continue here to fill the helpWe maybe need to think about the components order inI will fix the title tomorrowRegards for clarification capitalization for section titlesCorrect Edit Manufacturing RulesIncorrect Edit manufacturing rulesI will post some style references in the umbrella site later +I do not agree what is the use of swimlanes then at all? Same thing can be achieved with direct assignments making swimlanes an unneeded special feature +I confirm that everything was applied as expectedThank you +Subj Can I specify the range inside of fuzzy rule in +Yes especially if there are aspects about your content model that are outside your control and introducing a fake hierarchy only makes things more complicated at an application levelAs far as I can see the other source of is in the which is being created by the with a reference to the so that there are no API breaks necessaryMaking a lazy list writable is certainly hard but not impossible +The Hudson build failed without the change it depends on in common available +Thank you very much for providing the patch +Have anyI confirm whitespace is being added in the case of an element embedded in the middle of a text without surrounding whitespace +xml files makes sensebut users may want to modify the files +What are the results of running the new test with the old code +Ive never touched it and have no idea what changes have landed in both branches and it can take me a long time to figure it outReverting the change is not going to help move this thing forward +has been released +I really like the idea of getting something that can map to the native types in most of the languages +Preliminary patch +is about a general solution to this problem +Perfect +Once I have a better handle on these changes we are going to need to decide whether to introduce them next week +Unnecessary scroll in listWidth should be the same as width attribute +Ive found a use caseGoogles search engine is big into giving you result relevant to your localityIf I search shoes in Paris they will try and prioritize the local stores matching my request but still display results that are not within the Paris vicinity +thanks. +Modified files svn commitSending trunksrcrefSending trunksrcrefSending trunksrcref +Oleg +is working fine +Moving out of +Im sorry was not included in +If you want lets add documentation to the website instead on how to setup a pluto dev environmentFor isntance for intellij you can use maven to generate all the project filesm ideaideaI think the same is true for eclipsem eclipseeclipse +Dear AlanCould you please commit the walker patch? Im relying on that +The object has fields so that the caller can find out exactly which column was truncated and by how muchUnfortunately the server isnt capable of sending chained warnings yet so in the case of multiple columns being truncated we only get a warning for the first one +Philipp could you try this with the version? This version contains some fixes related to the path of the working directory +Cool will do +Committed the patch to the trunk as revision I intend to merge this patch back to the line in the next week +Likely configuration error +I will bring the disccussion on mailing listAbout your build issue +Ive tracked down whats happening +This allows the override to work even for embedded and does not introduce issues with vdb imports. +the RHDS version was downloaded October and appears to be beta +Youll see it in the root POM +See and for example both of which I posted back in April +Thank you comments encouragement and helpful advice for mei will progress steadily towards the completion of requirements +Committed many thanks Travis! +set fixVersion to since this is a new developement which sould make it to master not to to since theres no time left for there are more critical issues to tackle than this feature request +I added a new topic for it and pointed to it from the syntax descriptions for INNER JOIN LEFT OUTER JOIN and RIGHT OUTER JOINIll commit the patch shortly so that at least some documentation for the USING clause makes it into the release candidate thats planned later today +fixed in revision thanks for reporting this issue +I dont see a reason for this be more than an Optional + +Part of a bulk update all resolutions changed to done please review history to original resolution type + +In every request context or just one time? Where is the actual request when we have this setuser trick going on +thats kind of the same thing as preventing HTML since it prevents any tags entered from actually being rendered as htmli can fix this one by escaping the weblog title in the pojo wrapper +Avoids a new master for jfoundation +Patches to the core to enable the config checked the patch and fixed some parts because was thought jus as a workaround +Im gonna submit a new patch to cover all ports in use +Hi BryanI am sorry for bit late to submit the patch +bq +I wonder if we should bind it to webstart keystoregenkeyAny preference +That operation should probably be synchronized regardless +Verified +It is now magically importing all my services now once the agent started back up +The header and footer supplied for +Note that the expects the platform to be linuxx +Unit test showing the problem +Thats it +Slides from yesterdays meetup +sent from XML RPC using SOAP using SOAP using SOAP +Code looks good +additional info here? Otherwise we will close this ticket within the next days. +With in that size we can do zoom in out etc +Good catch +I fixed another bug that I believe to be the underlying problem the stack trace you presented appears to be a stack trace from after it does not look like a stack trace for undeploymentIf you can reproduce this under please this issue. +But Im on a Mac +Thanks RaviIn Long term I think Rumen should be changed to use Configuration directly instead of parsing conf from Properies. +I can do an ls in other directories and see the project files so Im sure its not just that +jar! style patterns as of Spring +Thanks Steve. +I made the experience that building UIMA in Hudson is a bit tricky because of the Ant script that downloads the for building the docs +Simple authentication failures with ij do not cause the issue so may be more of an edge case than first thought but I will continue to debugI think it is an important issue especially for IBM jvms as leads to the javacore in addition to the thread dump in the that would print with other JVMs but I dont think it needs to hold up the release +Not sure how this can happen and cant reproduce it either +Reopened to address Maxs concern of needing more clarification and context surrounding what a resource is and what a workspace is +Thanks Bill +Black magic assuming knowledge of field layout was removedOne question left here which thread group should GC threads belong to +s endif +Thanks Sean +Even though the test is still valid for this issueRegardsHoward +which paragraph in the spec defines this behaviour +Amendment completed. +I verified contents with zero and multiple datanodes +The same unit test can be used to verify that the patch improves performance of this operation by a considerable amount +I have just found out that the test is buggy anyway +Unfortunately I had started implementing the fix when you provided the patch so I could only use it as a reference to see that we made similar corrections +Add I tested these steps and it went smoothly +Single bad value. +Also down the road we can use it at other places +Derbyall runs cleanly for me modulo the wisconsin cruft +Were seeing this problem in version +But even then no where in the code operation context is added to the confctxopctxmap +since is represented internally as an internd string +Verified that the fix is present in +One line fix to test code +A new local variable which track the number of most specified method +Got a chance to run some test and Found max files to compaction at one time not working correctlyHas just the changes to make it work correctly branchHas just the changes to make it work correctly on +We have finished implementing also did some enhancements on it +I had a look at the patch and Im not sure how you take care of the scenario you mention above of receiving fewer zk events +Brief writeup of namespaces +Is that functionally similar to using an exchange pattern + +Since it is in both and only need to tag Fix Version as +Seems like the issue is splitted into existing difference for RI for IBM VM So perhaps this issue could be closed as duplicate +Thanks +The name node logs the error and alerts it on the http UII will use the existing data node protocol errorReport for the error reporting but with a minor change +The is not tested +Implemented default deserialization constructor generation in Java backend in r +We changed our repository layout hence this issue no longer relevant to us +is served from pages web server +New fileLength property that return a long +Carrol would like to review the code hes been mucking around in this area very recently and might have input to make +Fixed in + for patch +AFAICT infra is ready to go +Id have to look at with more details but for this ticket I intend to keep those node identifiers strictly local they will not get gossiped +Unassigning issues on which I wont work in the near term. +IMO the API is garbage anyways because it should be using an interface like the JDK classes doMMWhat should be returned if a MultiReader has embedded MultiReaders asI dont like this approach and the comments seem sound like over engineering a simple solution +For The delete is meant for deleting docs which are already in the index +Hi Sanzd I do not know what Dted is +I say pick your favorite of these last four and well call it good unless someone else out there has some helpful feedback to throw into the mix +I agree this would destroy performance +Actually just verified that this issue is still an issue with +This is done as part of +In this case we dont need to qualify Inner in extends list +the same patch since the unit test was not stable +At startup Eclipse is by default configured to only do a refreshThen with the project already open a resolve do use the retrieved jars +pending Jenkins +One additional recommendation support a option +OK I logged into and copied the and to the srctestsxml directory +Thanks for all of the feedback and testing! +If done by friday will commit +Ill update our wiki page on this as well +Just to illustrate what this looks like +Approved for if done by due date +Cancelling patch to get the latest patch through a problem with s getNode method + +Reopened issue to change Release Notes Docs Status to Needs More Info +thanks for testing it Wendy I missed one change should be good to trunk +Whats up with this one +Patch from trunk fixing upper bound in from incorrect to correct test revision Thanks Tim +Similar things can happen for spurious responses if a lock was granted and we get a response the lock will be denied although it was actually granted +Ive posted few suggestions in the review board +Like or or . +If you finish it before we release beta please correct the version +Companies are often years behind the current version and not everyone is constantly updating their environments +Closing resolved issues for already released versions +That should typically be the case if you add the Spring jars to your lib am using version but am still having the same problem +The patch is against but should apply to trunk as well as the onlydifferences to these files are header diffsThis is a proposed modification to and wich create to use them as +Thanks for the quick review JimmyI will upload another patch addressing the above comments once QA result comes back +Would be a great way to contribute to Groovy Colin +Thanks +Thanks for the suggestion interesting idea! However note that we have introduced a query method in Spring which returns a instance that is analagous to Suns standard but doesnt throw checked +See for combined fix. +Didnt know you could do that +JIRA wouldnt allow me to delete the original to reference attachments +Is the problem still present in Tika? Only there were some mp tag related fixes between and that may have solved this for the hint Nick +I cant reproduce this issue on I updated to current Phonegap version +The downside is that for the duration of the heartbeat interval the node may get overbooked but that should not be a problem in practice since the container would already be using a lower value of resources before the AM asked its capacity to be decreasedThe same problem does not hold for increasing resources +Not sure what you mean by not allowing the processing namespace prefix to be configurable but I think a scenario that should be supported is embedding the wtkx XML inside other XML files +Closing stale resolved issues +This is especially true for maven based projects +As you said I am also not sure on why it is loading the Jetty classes when I am using Tomcat +Thanks for the infoAnyone else with success or failure stories? +Are you using konquerorI think its because konqueror posts the forms in utf +Fixed in r by reinstating the Numeric to Java wrapper conversion logic that got lost along the way +Unless objections tomorrow Ill mark this issue as resolved +Also see my comments on +Check all the checkboxes and click update all the values should be set to true +Looking at the code that makes a lot of senseI think its fine to retry forever as the cluster is not functional until all meta regions are onlineWe have a plan Ram you wanna take a stab at it +So as I understand would be removed and tombstones would be marked for deletion once a repair operation was successfully run +I think it will need a new tool and Im not able to spend the time to do a good job. +Can you tell what you did before in which way you builtDoing build clean war I get the following startCopying files to DbuildclassesCopied empty directories to DbuildclassesCreated dir DbuildmocksCompiling source file to DbuildmocksCompiling source files to DbuildclassesSo the copying of the files and the compiling of the mock classes seem to missAnybody else with these problems +here is a patch that adds asserts and beefs up a test that fails without the patch +I think you can safely revert the code in because the newly added following code resumes region transition after zk node deletionI really worried about +Committed to JBESBGACP with . +I have implemented a rough version of this feature using David Spencers code +but if dih is spewing exceptionsthe test runner buffering could be slowing you downI think we should open an issue to fix this bug in dih +incorporate feedback from code review +This is not necessary anymore and it prevents from being used +Note that the serialization of on is not implemented and there is some work to be done in the generators to treat properties of as special cases +can you whip up some js to handle this +That result was against the same version as I had tested beforeIll test again +Call for review +Okay well since I have some tests that pass and this doesnt mess with single node mlt Im just going to commit it +in and in now that it has released we should update the version. + If forceCanonical is true subsequent accesses of vfssymlink resolve to othertmp which isnt found in either the permanent or real cacheIn what way are you accessing vfssymlinkAs I have again hard time reproducing thisWhat if both tmp and othertmp are added as permanent roots +VE script looks good to me +Anyway everything is working fine nowThanksVincent +starting of this in a bean and +If you are familiar with SVN svn diff command will produce a file in the unified diff format which you could attach to this issue +weeks +Please check and close if this works for you +So it may be just we forget somewhere to refresh the db state +After I click on one of those links the session parameter is missing and clicking on them again makes the New Message button disappearThe eCommerce Forum and Blog features are a part of the Content Manager application which doesnt get much support +Is the summary correct or is a newer Kamikaze intended or +has separate get and get methods plus a shortcut get method +Is this intended behaviour or an errorApart from this Id say this issue can be closed now because the original error doesnt exist any moreThe FIXME text was removed during one of the last commits +This error makes sense if the web service was DocLiteral Wrapped which is the default in +A few comments The patch does not apply due to conflicts with recent commits The current patch seems to include the patch for but the comment before this one says that it needs to be applied first +Move it back in if you think differently + +The two approaches seem complimentary to me + Apr INFO start NA seqNo wiId Apr INFO queued NA seqNo wiId Apr INFO dequeued NA seqNo wiId node pid Apr INFO ended seqNo wiId +This has been fixed via +Here is my current understanding on how this part worksThe client receives types of responses ping watch notification and request response +Tony patch applied at revision r thanks a lot for the enhancementArtem please verify the problem is fully resolved as you expected thank you. +Ok Ive patched my instance by adding the into the MDBs interceptor stack however the MDB always receives a null default subject +Updated the release note of Camel +I cant find any shifting whether the text is aligned left or rightThe report is much to sparse to make anything out of it +We can even exploit the sorted characteristics of the know that a table is sortedbucketedbut never make use of it +On the other side and related support can be an enhancement to specify fonts +Use sequential read rather than position read +Fixed as suggested +Its the compile log for and it contains the two compile errors listed above +This is a biggie +Good catch on the title issueOn just a few thingsIm glad you caught the problem with the term library +I think Ive confused JIRA can anyone direct me to the magic button to submit my new patch to Hudson? Thx +I like that form +A stack is introduced to cope with the implicit push and pop of writers +Attached new patch with minor changes and added apache for all the help Andrew +Annotation and Type completeions in java editor are supported by Eclipse JDT +I have committed this +This was reverted because of a fatal bug that it caused +If you have a place to save drafts you can then make workflows much more lightweight +This can be done in the d part of the screen and then hit Recalculate OrderPlease use rather user ML before creating a Jira issue for such suggestions questions etc +The patch contains a tweak to the Makefile which Im going to leave outAlso I will fill in the reduce implementation for PAMI so that any X programs that run on PAMI and use the newly added reduce will work correctly +Any solution would be okay for me But if I try to implement Ill try to not break any current API +Greg Are you still seeing this issue and do you have a logcat? Id like to see some movement on this issue since I cant reproduce it on this end +Im doing a reg exp replace during the built to fill in the correct path but apprently its not working on windows +Closing the issue as it seems to be an alternative solution +Youd need to submit the change as a patch cant tell whats changed above +Hi Ted thanks for your patience and help +The if and ip options are replaced by vifeif and vipeip +The list isnt completed so if there are some needed mappings we have to add them to list +fixed in version of the plugin +Thanks Paul +I can confirm that the solution proposed by Titi and Julien work +I probably do not know use cases where now can not mean exact now and it requires the granularityCould anyone tell me why now can not be but needs to be tweaked to or +Its unlike any number should reach e in iostat +Fixed in several places with revision +does this also affect +Im not sure why the alias for SORTKEY is necessary +in avalon is a but for plexus it will probably be a string +Your file CWINDOWSProfileshohwille +They all are DD related +Fixed in bfee +Im using version of the plugin and does not appear in +Thanks for the report and patch +Moving issues from affecting to affecting or before deleting version. +When thread waiting on monitor is interrupted all threads that wait on the same monitor are waken up +Patch doesnt apply to latest trunk +This makes time out apparently it can no longer contact the masterI do not think Ill have a lot of time in the coming weeks to track that down +Also updated a bit +Great thanks Somebody working on this issue? +Let me know if Im missing something. +What happens if you increase the nscd buffer size +Added a combo box to choose the format of the created CAS files +Kurt can you take a look at the SOA platform and advise Trevor on this? Is it safe for him just to remove this sar file +This should fix the findbugs errors +I will upload a new version of the patch after fixing it +I think the issue is that is being called which is actually adding every class in the archive not just the ones that have requested to load by this method +Committed revision . +Ive also included some relevant svnignore settings +Im sure the mail client does some syncing before closing but its not that I even HAVE more than one mail in the mailbox +Makes sense patch applied in revision thanks! +Example to failing cases etcThere may be others since there are a lof of references to that functionDIGY +Manuel Garcia Thats rightIm using in my applets and it works +Anyway to chec if a split request has come for an already going on split? Because currently every split request creates a new split transactionNo +PFA recursions errors produced when using injection +This is something that could maybe be considered in the +Perhaps Ill have to go poking through Thrift to understand it a bit better +Few applications depend on token types noProvided the token text issues can be resolved Id like to see replaced with this +Here is a patch that adds basic graphs to the tracing view +docs and docs removed from the SVN trunk as no longerup to dateOleg +Wonder how that happened. +The attached images show a scrollable data table using an ajcommandButton +Correction does NOT represent a status +The pom files have specific versions +Completed document source files and Generated output files in PDF HTML multiple file and HTML book formats +This patch specifically targets support in hadoop for filesystems other than hdfs +Also added default pass through for all of the HADOOP YARN and JAVAHOME environment variables +This issue is about silencing the not displaying them as error level message since its not an exceptional situation that a node is suspected during topology course currently may accompany other serious errors but those should be logged separately. +I will commit the attached patch now to trunk and +Thx JacquesYou and others are very supportiveThx for your supportRegardsAshish Vijaywargiya +Patch committed thanks for your contribution +Done +Thanks for the clarification Jonathan +Then they should be fixed in this jira +Committed revision +It happens when debugged JVM exists and generally is harmless +The EAP team produces the RPMs for RHEL and and thats why you wouldnt know anything about it +Sorry for that +Do we want to do this +I need mutable granted authorities and the collection in the User bean is unmodifiable +Another area thatmay be problem would be network testing in addition to embedded testing again in the area of a newtype that just defaults to the basic java externalization routines +I did some research I think problem is in in get +This issue is the same as the one I reported during testing of CR which can be closed now as it works with. +Hope this works for youDavid +Ive committed this and merged to thanks Matt! +But yeah I think in this workflow it would just be another option you could select a directory and then youd input a directory and youd be taken to the dialog to select which featureType for that +Changes have appeared in Latest Alpha Manuals. +Hi The read and write operation can be deleted +To be honest I dont have a clue +a sharded configuration for testing binary responsesAhh +Markus can you hold it until is released? This is a substantial change and we should take time to review it carefully +To handle soap headers you can write a handler as well +Committed Carlins change with revision +Thanks Michail for catching this +Applied the patch into master and +Manually control NN to enter safe modeTest with zk zk server is down before RM start or while RM is runningAll the scenarios client is able to wait and retry +bq +Absolutely correctWhile is thread safe it is not friendlyWhat I would propose doing is allowing multiple concurrent updates and that the lockid be the identifier for a particular update +Thanks for your helpI think that the defaultAction issue was resolved as part of. +The relevant text on page one now readsYou can configure and to listen to a Queue or Topic by setting thefollowing parameters in the Service Configuration file will need to ensure that the client jar files of your chosen are included in yourclasspath +Attached archive now includes both printf and print versions for looks like is corrupt +Such a tiny fix sorry to take so long to make itAttaching and with this correctionIll commit in a few days unless I hear that further corrections are needed +Suspicious +I think the Extended Attributes is orthogonal to this issuebq +client classes under the deployments directory of the app should also note that builds and deploys done through Maven will always work fine because we simply exclude classesclientlocal from the WAR file +Not that Im aware of +It is not really database specific functionality we are testing +So basically the errors differ between jdk and jdk +However its pretty much useless to humans +Tests ran cleanly for me on the patch +Yay well be able to remove and use this +CLI does not have sophisticated anywhere else so adding it here is not required +Sorry I meant to say type handler callback rather than type mapper callback in the issue title +I dont think there are any pending issues Im going to commit this in a few days +Similar to +A feature can include other features only +It would be helpful if you could attach either the exception stack trace youre getting or an example document that causes this problemIt looks like support for these file types is not yet included in the Apache POI library that we use for parsing Microsoft Office files +This discussion goes beyond one vs +Verified that I am now able to tab through the trees in Firefox and expandcontract them from selection. +Hi Bhushan Mandhani is no longer at Facebook so this email address is no longer being monitored +Thanks to all who helped with this! +Bulk close for release +OK thanks for the docu update. +Flip spent several days in the user IRC channel trying to deal with the load spikes +Created another patch for the test changesUpdated to run test in mode as well +Vinod thanks for committing the patch +Forgot to mark as fixed. + +Here is a patch that implements the old blocks purge strategy +ok but what about the rest of the tasks for tlp graduationunix groups mailing lists svn issue tracker wiki etcHave you had a look at for reference +fix for version Madhumita +As mentioned in the test comments this test case creates a file with one block and replication of +so I would keep them as they are +I removed jsf version at all +lgtm +The assumuption is that keyType and keyClass will be appropriately set to Tuple and the collect here wraps the key in a Tuple +This patch can be applied to branch also + This bug has been marked as a duplicate of +Ill make sure I enter a component version for new jiras. +Therefore I left the remoting source under the module for branch so could do minor bug fixes if needed +Note that the attribute called encrypted is a bit of a misnomer The value is actually the db encryption key encrypted using the boot password plus a hash of the db encryption keyBtw Rick even if we introduce a new method for this the issue would remain unless we change the behavior to generate a new encryption key and +Apply after vote to commit pending resolution of +Thanks for cleaning this up Im always pleasedto see patches that remove redundant codeIm not that familiar with +From the command line test are running fine for me as with a fresh launch config and this is what gets calledSystemLibraryFrameworksVersionsHomebinjava classpath UserstcurdtdevjakartateststargetclassesUserstcurdtdevjakartacoretargetclassesUserstcurdt +I cant run the example with an IBM JDK and it needs to be investigate +I cant seem to find any information on how to set entity manager properties from theDo you happen to know off handif not no worries Im just running it with out the constraint right now as my work around +see branch documentation is part of the install this record since is now released. +Thanks perfect so its solved +Thanks for this patch applied in r. +Thanks for the information +Alexey I mostly agree with the update +Primarily it is worth looking at whether GCJs could be merged into Classpath allowing general access to this code +A summarized version of this list should be later included in the A prioritized Summary of changes should be the foundation for building the Geronimo documentation +Closing issue as it was included in version which has been released +Moved to the JBAS project +Thanks +Could you take a look at the current and and see if I missed anything +We patched wsdlj to allow resolved issues for already released versions +This is an incorrect statement +Reporters of these issues please verify the fixes and help close these you using the latest Agent codeI had verified on with latest kvm agent jar copied to kvm host and tried +The new System were I have imported the backup file is rI am not sure about the old version +Adding the source file for review +as this issue cannot be reproduced +but no reason we couldnt already up and taking writes in the meantime before the log replay compoletes +I rebuilt the artifact with version +But this should wait until we have a performance test +No problem and the speed difference when not pumping through the exec pumper is significant +If you remember the location of your project in the list then you can find it in a few Libor says I do think we can order this alphabetically though so we can improve it +I assume you meant that Roos class doesnt have support for and +Moving to +The log on connection and disconnection are both INFO so it made little sense to have this one as DEBUG and it was annoying to have to enable debug logging just for this output +Just realized that is not needed +Thanks to Bobby and Vinod for reviews. +Patch updated to failure for is not related to the patch +Thanks Tim! Verified +Am I correct +Any direct reference to the environment variable should be capitalized any reference to the general concept of classpath can be lowercase + +Cant do much +The changes went as part of revision into trunk + for the patch +It seems that private is not workingI send a mail to that but did not get it mind readers we are not you provided us your namit address so that is the address subscribed to the private list +The SIXPT branch contains an addin for both VS and SD +In the second select box with multiple turned on the list was incomplete +Consider writing a quick test in should be pretty easy +So in order to update this number the user needs to select the next page result +Initial patch removing all references to from srcdoc +If I dont hear anything on this issue today Ill check this in tomorrow +Smirk +Hi PrashantLooks like some of the class member variables are still public +Ive reproduced the problem but didnt managed to reopened it so I made clone and attached sample projectPlease notice affected version is Ive discovered that method is not working good when artifact is instance of specially without jar file at all. +Youd sure never suss that out from looking at implementations though. +Verified at +An overloaded implementation of is used to a override the tests root dir as the home dir b validate that there are no delete operations executed out of the tests root dir +rebased patch to head added and logger statement changes +Added which will enable deploying the Infinispan hotrod client as a jca datasource wrapped in a +This includes parser changes for the same +unless Seam requires a JSP container now +Ill keep you posted once I have something more concrete + has been marked as a duplicate of this bug +Attached testcase. +will open a new JIRA + is a preliminary fix +pushed upstream +Work in issues to the latest version. +Prachi once the phases that you mentioned were added to was invokedAnd since the source file is correct i will mark this as a packaging bug +The patch contains a fix for and +its very difficult for us to investigate what might actually be the problem with just these traces +So not related to your changes +Resolving as per Jimmy suggestion above +What do you mean by starting files loaded by a URL? Pardon my ignorance but wouldnt WWSAF ignore this page and return it as static html +Closing stale resolved issues +So this is an important fix for and should be harmless for trunk +Fernando can you provide us with an example which highlights the issueDavid and I have both been through the code and cannot see a difference but perhaps we are missing the obvious +In general multiple kinds of advices should work +However since Java doesnt give us any way of lazily evaluating the arguments to a method this check does seem to be a standard way of working around this limitation +are any changes needed to ddspring for this +verified on build +New patch looks good +I think these methods would be better added directly to so it can be managed in a Spring applicatoin I usually do in a test environment is to rely on s default initialization that is on a file in the classes directory typically with console logging only +Im having some troubles downloading the patch +Is there a simpler way +When I last looked for a replacement I didnt find a cache implementation that didnt push items out if the cache became full or had some kind of expiration +I merged this to. + +Looks like a variation of this patch has been applied to trunk so close +If a component throws an exception in the activate method the state is not changed but of course you dont get a component instance either +Now we also have a in the package that is respected by the +does the inaccuracy happen even when you are zoomed in close to the layer +Should apply to as well +Took snapshot of both the data disks and root disk and it was successfulDo let me know you can still reproduce the issue and if I am missing any steps + provided slfj because we have bundles in the umbrella that import and use slfj +Based on what I see in your code I think the interface can be very small +I created +This is not needed only pages in need this +Hi Milind as mentioned earlier either case sensitive or not is fine for experienced users +Ill remember how to make the patch +If you can implement the desired behavior via the Atmosphere API I suggest you do so because that should be much less expensiveWrapping the Atmosphere api is not what is for it is a bridge not a wrapper +What query is causing the error? Could you post a stack trace of the exception +Test remove the fsubview the stamps are correctly created when the subview is not there +Correction +Jonathan Is your patch being applied +so I guess this was not a valid parser bug +I may be missing something here but I do not understand your comment +Applied patch into the trunk +Patch fixes the issue +I guess theres something wrong about Socket IO handling of +Also if you have any code that would be of general use please feel free to send attach it to the jira and I can see about getting it into the base classThanks +However items should not cause memory issues Ill investigate +What about the other case statements in that return or but never +jars folder in flex +Need to support proper movingrenaming +Hi David that Ant bug strikes again +Hi Even though this issue was marked as a duplicate of a solved issue Its not solved yet on surefire version +For each transport there are four kinds of tests a +What kind of options will I have available for retrieving statistical information from a remote Solr instanceWhen I fired up the multicore example I found that the cache statistics pages for both cores only had fieldCache and field +Fixed in SVN You can use the title of the plugin to navigate automatically to the correct tab +Yeah Geir had the same feedback on the list +Hi Would you please try my patch? ThanksBest regards applied +Thanks Guillaume can you give me an example document too +Thank you for your helpHope newer version of ODE could be released soon! +Closing this bug since all of its dependencies are done. +Ive reverted the commitsSending srcmainjavaorgapacheservicemixcommonxbeanTransmitting file dataCommitted revision +Scenario and scenario of verifies encryption only scenario for Symmetric Binding and Asymmetric Binding in the Rampart build +Mass close all issues in resolved state that have not received further comments in the last month +Proofreading review completed good job! +Bulk update to fix version +SishenI edited your patch to remove going String Text toString +The problem is that I need to test with heavy weight components +Fixes problems reported by GuillermoI just committed a patch that could configure timeout via system property and reduce default timeout value to secondsThe revision is r please check if it could resolve your problem +Last sentence is Is there a workaround +Updated patch +Also see which is probably the internal representation of the same! The above should say +Having said that Im resetting the Interrupted flag to be on the save side. +Please stick to the possible through reporting +AOP Advice chainsI dont see any reason to create new issue for that very simple bug +Im afraid this bug is not related to Flex +A Java double can take the form or whathave you +This is too risky this late in the RC cycle for so its bumped to +Patch for review implementing the proposal +In the index the fields id and should be presentAre you saying that this is not the caseCould you let as know how the class is annotated +Great thanks for the heads up Ben +Looks like the fix for this was checked in with the fix for. +If you want I can make it so that these emails are sent toFor now this provides a sort of s continuous integration +The conversion probably is not necessary since we are dealing with nulls +is inherently unreasonable the question is whether we can get to fast enough with +ConsiderTwo files +This discussion goes beyond one vs +The patch looks good to me +Testcases are running in trunk +Please comment on if you continue to have problems +This was because we are mostly using the NIO connector +Reverted and at revision revision +changes how the workaround Stack describes is done within the shell in lieu of +So we could try to remove xml extension assuming that we can have mix of and +Thanks lads +Hey LesIll get to this as soon as I can Im off on Thursdays and Fridays and I work weekends so today is going to be a little sporadic I may or may not be able to find some time this evening though +This patch the resolves the cascading merge problemThanks available in following revisionstrunk svn r svn r +I cannot test it ATM because is not startable at themoment +Thanks! I updated the patch on your suggestion +The prologue setting if implemented should IMO apply in all task contexts streaming pipes or plain old Java +I am also a little suspect about the mapping in general +Okay fixed in r +Also the prefix is not the problemI expect the error deep in the automaton +That makes some sense I manually installed cucumber in jruby and then found that I got the same failure for hoe manually installed hoe and got the suite to run +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content fix for this particular JIRA has not been done yet +svn diff +Attaching and adding notes to topics that discuss the security mechanism in the trunk documentation +Closing old resolved issues +thanks for your contribution please verify if it fixed as you expectedNote that set only has effect for the current year and beyond +Hi Nicholas I think both way have its strength +Same problem + +I think youd be hard pressed to claim diligent search or orphaned work status for a file taken from an Apache distribution +User can use Wordpad instead Notepad +The v patch does not correctly apply to the latest changes on trunkThe patch itself looks good to me simple is better +Fix it so hdfs exceptions dont fail tests move the verify of mapreduce job before shutdown of cluster to see if makes a difference up on hudson and add some debug output so can get more info on why hangs on cluster startup +Sigh +It seems already defined in interface +Thanks +Need to figure out what to do with existing projects if anything can be done. +This has an impact on the upload servlet with regard to authentication +It works +Uhm I had a long discussion with Max about this once +Notice that this option can be set on component level so you dont have to enable it per endpoint +Moved to SPR the new home of OXM +I believe we have other jira issues that this essentially duplicates +Consolidating to so we can then rename to +Fixed in in trunk +fixed function in r +Not a compiler issue more related to our core classes not having backtrace frames in all cases +Are you running trunk or an older version +Hello GeorgeThe fix is fine +AmbiliI am unable to reproduce here so the issue must need your disk or configuration +Dirty work +Committed to trunk and. +Attachment has been added with description Minimal xsl file +It may have just been something I caught in the middle of dont get any errors with rev using the nd script above +This patch updates tomahawktld to work with the first patch +Are you using for Mac +Is it possible to get a list of files that was changed in order to patch this? Im working with an old snapshot this week and would like to fix this before I throw it away +I used the binary installer on +AFAIK uses and notAnyway if you think that the tests are good and has better performance over feel free to change it. +Resubmitting to Hudson should do +Committed to trunk. +Patch attached to fix issue +I believe the straight benefit out of this would be to keep problems with the application and problems with the framework would be kept separate which is a big thing +The passes the is test so I cant figure out how this exception was ever thrown +We need to make sure the client injections are somehow modularized possibly using the interceptor chain +Created +It no workie is generally not an adequate description +Please take a look and let me know if there are improvements changes that must be futher madeSome points about this implementation Does not generate schema for static sdos that have been generated from an XSD +Our training and courses use Maven because it is the only viable option +The parameters should be passed to the constructor and private should also be final +Just keeping the book keeping straight +I would like to suggest to change the history filename to +I made Iconv backtraces better in it seems we really need to come up with better way +Will somebody try the patchBesides pls export the resource directory of beans as source directory or else eclipse will not put the new ser file to the bin directory +Have we reached consensus on this? If not I want to offer my opinionWe started using ivy for several reasons but the big problem I thought it was solving was avoiding license management +Is that correct observation? or am I missing somethingThanks in advance +Not to per Stack +Committed on trunk backport on been committd on as well but guarded with a system variable enabled with issues being marked closed in conjunction with the release of. + +Although Ive never used it so Id probably have to lean on Brian since he wrote that code to describe what use cases it was trying to satisfy +modeshaperesourcesruntimedeploy +So when will the hardcoded api be upgraded to api lvl +Applied patch with thanks to Bela +In this case we will only see one of the two entries +As for Move action it seems to clear +I agree that processing war would a different beast +I came up with an idea how to implement the validation using the existing features so I had simply modified the existing code in other to create another marshaller that will be configured with and used for validating the marshalled not sure if this is the most efficient solution but based on my current opinion the simplest to implement +The routes used to deliver the message to the correct instancechannel pair are not being removed and the engine gets confused +We could put it into a if wanted since this is all internal +Maybe this exception could be related Provider not foundAndy you are right I just run some old commands but turtle should be better +I have committed this +Ive applied just the test patch in r +This comparator is part of only the default tuple implementation used internally within pig +Included an if statement to the create statements to allow for proper datatype usage based on DB used +Email sent to Yourkit +I couldnt reproduce this issue +Patch attached +There is a duplicate tooRegards Urugn +The windows installer dont have a newer version native launcher so this is still the old problem and has been solved. +Fixed double click selected the application and closes the other issues except the are not fixed +Javadocs warnings are bogus +This was great for the situations above the open retries would happen in the logical nodes driver thread +Verified in the doc +Feng +the patch looks good to me +Eureka it works! That was a most excellent shot in the dark Andrew! Works fine now no problems! Drew you should give it a shot +Closing all tickets +These changes are merely updates to ensure that applying the patch goes more smoothly and to pick up changes in the output for the generated classes in the test caseWould a commiter please review these changers? Note that the three exclusions listed in the above comment still apply and this will result in three of the test cases failing even after the patch is applied +From what I can tell they seem to line up with code in the C and Java libraries +I will do a seperate change to fix +If you have any patches please provide them +There seems to be no issue with Metamer onThe rich sample behaves same on AS andexcept the constraint of which is addition of and thus doesnt work onBe sure to do not include when deploying Metamer otherwise Bean Validation stops work. +this patch adds an isLive method to fix this For a simple column live is +I clsoe it and set to wont fix +If there were an easy answer we would have done this already +Bulk close resolved issues +Because the airline sample has a dependency on the samples project you will have to run mvn install at the root of the samples directory before you can mvn install airline +Maybe we could create some subtasks like one speciallly for automatic validation +If it is public then there is no parse error +I have updated env +Im having the same problem trying to load a with EPSG in Geoserver discussed this during FOSSG in Sydney and I thought it would no longer result in an error but only a warning +I think we should allow clients to set parallelism to as well the same case where splits are computed by the client instead of the AM + The only thing in question is the copyright notice in the header +They had a real release +Im not familiar enough with Lucene anlysis to know whether it should be exposed in the factories but in order not to break existing factories I think it would be better to make this new method throw UOE by default instead of being abstract +Also please reopen the bug if you think its needed +Any hope this will get fixed +These issues are linkedThey would greatly reduce code duplication between and as well improve the readability and maintainability of the code +We should try to do this very early on in the development lifecycle so we have ample time to fix any issues that could arise from this +That seems scary +Lars should I commit this +Thanks Gunther for contributing. +Dunno if they changed that recently or not +Had a quick look the patch looks generally good the only thing that raises eyebrowses is that the test does not check the chines chars have been echo Andreas comments +Anyways I like the idea and your code looks good to me +Yeah he had a fix +i just committed this +Many thanks MarkOleg + +Thanks to David for the sampleIm going to debug it and update with progress +for agree should be supported +Of course if you can provide reproduction steps thats even better +We have recently migrated to wicket and than started seeing this issue +We need to add a parameter for repo as well +The bread crumb on what page +Do delete Cocoons working directory every time you addremove any library +This adds a unit test for this ll get some change in for + +VERIFIED +seeing a would have given me a hint towards what is actually happening +Wont fix for +Marking this issue as done +Happy to take them out thoughWe can open a JIRA for the testObservers hang and I will take a look +looking up registred editor is something that is almost instant in eclipse for any other editor so much be the mozilla init +Already done +Fixed on and closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +thanks michi for testingif nobody objects for whatever reason i will commit the patches within the next daysfor the record as far as i know this maven feature that has not yet been used within the jackrabbit project +String is not yet solved for core TRQ +Can you please check if we have this fix in our EAP build? Thanks +Attaching component documentation in for the docs and Willem for adding them to our wiki +See untagged the todo jars however the was used for the RPM creation in the stable build environment and thus its waiting for the previously mentioned jira +I like the normalizer approach +Did you try +Great work +so I assume its a random Solr cloud failure due to the fact that your machine was very busyThats my assumption too +And it would be nice if ant eclipse was triggered in that case as well +personally i would prefer if we dont have a separate script for changing the maven filesIm not really sure what this tester is currently doing but in my opinion if someone gets Lucene i should know WTF they got regardless of whether its from an FTP site or mavenSo if it doesnt exist now at least in the future Id like more logic between the two things to ensure they are consistent with each otherIts scary to me that different build systems are producing different artifacts and we donthave this today +Cannot reproduce +jcs go ahead and resolve +Practically speaking it never happens but its thereWhat do you suggest +Zlatin Yes +Basically this will ensure my coding governance at the CI server end +Patch there is no unit test included I manually tested with two failed script both pass without error message. +Why double slash in myhostnameAdministrator? It has been ages since I touched anything Microsoft but if my memory serves me well it would be myhostnameAdministratorAs far as I am concerned DIGEST auth exchanges look correct and authentication fails due to invalid credentials +does something similar but I dont find the syntax very readable +pushed upstream +updated the test still the same result. +Seems strange to implement a deprecated api +It makes it a lot easier to review issuespatches if the patches are submitted to the issue that they belong +This is the default setup the jcr provider is mounted at all other providers are mounted at some paths ootb these are bundle resource providers and the servlet provider +We will make this decision based on more testing and user feedback +consider backporting to Currently all of my work is targeted towards with the caveat that some trickier stuff might bake in x before being back ported +Suddenly there are three classes that have the exact same functionality so naturally you refactor them into one class or start abstracting out interfaces and concrete classes +Canceling patch until the unit tests are working +Thanks +Could there be a problem with different transactions overlapping and causing such post you want any more information +Vikram Im unable to reproduce this +Ive applied the patch with some trivial cleanup at revision +We will leave it out +Attaching v that does the optimisation of checking for last version +Thanks I committed this fix +But something obviously needs to +I think that signing is probably a better mechanism overall. +Perhaps we should ask if Jarkko would like to continue +And esb module reused these codes too +If all goes well Ill into x in early Dec +The last patch breaks consistency with the existing components naming so I think that it should not be added +This patch depends on +Dave Can you confirm if the sts step is required +Also removed the extra assert as suggested +Applied those parts of the patch that were still applicable +Thanks Raghu +Ive tried docRoottrue but it does not work for me +Fixed title. +Mark Newton wroteHi CheyenneMaybe you could have a go at this since you were responsible for creating our wonderful Microcontainer logo in the first place cheers JustinHeythanks for the Mark just informed me that were only getting onesIll try to find somebody here in Slovenia to create me full colored one this means I would need your original designsCan you send it to me put it somewhere where I could get themWhats the best file type to distribute this fwd to Ts creatorsCheers is in the mail +Closing for now. +Expand the scope of this ticket accordinglyYes lets solve them both here +Out of to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +DavidCan you be a little bit more specific please? What font did you use? Please attach the resulting pdf to this issue +Otherwise it will either be bumped to or closed as Cannot Reproduce +Brian formatted xml output also involves response in xml output? I am asking because I have been trying kitchensink example and response from WS is not well formatted +If we had executed the Client side multiple times for a Remote call the Remote side would be invoked multiple times as alternative solution would be to use Arquillian Spock Runner +Im running trunk revision and it looks like is included therein +and head +Patch that takes care of review comments +Yes try the example and set up your browser preferences language to something else than english or french +move to for please set fix version of LATER if you want me to see it +Fixed in +First there is no issue with setting channel as a String +Once again I think it will help Seam option is to say Eclipse or workspace +Another reason for repository synchronization +Why would it beYou were calling a method recursively and Boo doesnt introspect recursive methods to see if they eventually terminate or not +Maybe an American going to live in Cuba or North Korea could still be said to be defecting +The setup I have is operations are routed between different and each JVM is readwriting to the JCR repository +The description is hidden on line and should readStringencode withinvalid replacements still raises an encoding error for binary to conversion +Hithis is definitely on the TODO list will try a look at scheduling this into the next releaseregards Malcolm be included in as a impl +Ok If you can verify this is no longer an issue with the code in TRUNK I can close this report +Hi Tony I am not over familiar with Jail setups but based on the information youve provided I assume it is viable to go down the route until we progress through the incubator +Do you know whats wrong with directly closing the modal window +You need to remove the codehaus sanpshot sorry for the oversight +Issues are linked so I wont forget +Fixed in trunk patch attached for +This already seems to be at TRACE level +Then Ill apply the patch to see if I can update HEAD that way +Looks pretty good +You were right it was remoting jars that were needed +Its hard to immagine youre getting the same error as above as there is no longer a method remove in +HarryAre you seeing a build error or a runtime error? When I run ant compressedjs all looks fine to medirk +Thoughts? Trustin Stefan shall I close this issue +Please verify the issue on the latest build and let us know if you can reproduce it anyway +Other than my previous comment the patch looks good to me +I think if we just added a JMXnodetool call which resets the hint state for a node that would fix my mental block +This is really an improvement request rather than a bug +Its the whole point of the attribute! This attribute provides the best user experience possible with no validation round trips and page refreshing required +It could easily be moved into the getter as well though +Thanks Erik +The incubatorooozzz directories are still out of synch README should have been replaced by by now +Closing as requested +Do you know of MTAMDAMUA already supporting the MTRK extensionI dont understand german and I see the RFC reference in the pdf but IMHO MTRK is part of a really new RFC and none of the mail server I know does support MTRK +One issue he ran into was that uses auto flush on every connection +Can this be prioritizedthanks verify this thought I will check in a temporary fix for this to make this particular kind of to execute in sequence +And than make sure that is serializable +Yes happens to be quite closely coupled to the version it uses +I dont feel like spending the time to script gitweb via chef at this point +Cleaning up Seth request to close it +The same as +It would be nice to have somebody look at the tests firstthoughI also need to write a small functional specification to sum up the behavior and attach it to this issue as well as a release note +Please address my comment Aug and attach new patch +I should finish what you suggested within days +is there no activity oni found that project useful since now i never want to use so is anyone able to submit that patch to providing popregards +Ignore the README attachment INSTALL is the changed file +Please attach some code or a small test app that demonstrates the problem +You can reuse everywhere +Attachment has been added with description my local fix for the transition resolvedfixed bugs to closedfixed +Ahmet Alfresco is released in two versions Community and EnterpriseThe Community artifacts are publicThe Alfresco Enterprise artifacts are private and are available only to customers or partners of Alfresco that have an account to access the Maven repoPlease wait before removing the version tag I would like test this change on my environmentI hope to confirm this today +Straightforward patch + for patch +Is this ok for you Benoit? +There is a JIRA which describes the bug this patch was intended to fix +you never stop learning Java +Ill submit a patch after the rest of my tests run +Fixed + +Sorry local configuration issue +I can reproduce this but it wasnt exactly easy +half of the previous runs detailed above until it had an error +Stan do you think we can close this now +There may still be other problems of a similar nature +The server application is both a client and a server +Completed and Checked In +I just committed this +Issue is fixed +Corrected logic in SVN revision . +The attached patch updates the list of compatible JRE versions to andAll the regression tests ran cleanly +When support for quoted identifierswas added in they would have had to change the searchbehaviour to to distinguish the two tablesabove +Moving to next CP quite a number of issues to address not possible for this CP +Hi I would like to try to write some JMS portlets so that we could do some configurations on the Broker +Both querys do the same but in the query im using an implicit join in the from clause is this invalid in HB? if it is so i think the exception message should be diferent and it should be a warning of this in the migrations guide +Deferring all remaining issues so I can release in JIRA + +I dont have a reproduction for so I would appreciate someone who is seeing this issue could give this patch a tryThanks ran fine +Closing as this was applied correctly. +The variable name has been changed but the issue still has not bee addressed +class files +Ive added support for the old name + +and to think I was quite twisted about that leave native option I thought nobody would end up using it +Stacking pbufs on Thrift means two two marshallingunmarshalling layers and glue +Patch applied thank you for your contributionResolving fixed. +Moving it back to that +just committed this thanks Steve and NigelCommitted revision tnx +Where can I find the fix for this issue? Can you please post the linkThanksKamal +trunk +ScottQuick question have you tried this patch with JVM enabled? On my this patch doesnt handle that case I dont know if its a real problem or notCheers based on Philips suggestionWe obtain a initial CPU cumulative time when task is initialized and we subtract it when reporting the CPU time +oops specific should be specify in that last sentence +Previous patch failed to handle templates which didnt specify a +We havent updated jacorb +Yes +Clebert Id like to work on this +Juergen +nightly builds + +It looks like need to use multithreading +Metrics are collected successfully for both also. +I dont have enough insight information and I dont want to change any code without a unit test for reproducing the issue +In my case the Logj client can be configured with multiple agents +If we force an interface getHosts or getLocations for any file status type we can simplify the client and server API for creating and listing file status objects +Any remaining problems can be opened as separate issues. +Patched submitted for the patch file be in this ticket or was it submitted somewhere else +Implements and support as discussed on the Tuscany dev list +Im confident that this doesnt work correctly +Indeed Romain Im really sorry but we wont integrate this feature. +It is NOT ready for reviewIn this patch I have prototyped local sessions and readsThis patch does that for all sessions and not just for readOnly sessions hence it is not correct +Demand around Struts features and integration is pretty low as most folks have either already integrated their Struts apps with Spring some time ago or have moved onto using Spring MVC or other web doesnt mean that its not a useful idea but probably something better developed as a separate project +checking permission in one place +Activiti needs at least two +Yea I dont think you need to strictly sequence this after the work +I built and installed Teiid into jboss and tested the fix it works +have been around for a fairly long time +We also need to reword because EE simply has missing areas of coverage for CDI +No extra NOTICE or LICENSE entries are needed to use this codeBased on this tar parsing code I added a simple class and the respective configuration entries and test casesResolving as Fixed +So the intent definitely is to ensure that Jackson can be used on it should of course also compile on Im just wondering how much effort is you specifically need to be able to also build it on or are there problems in using resulting jar on JVMIf latter I definitely need to look into that +When read handles it needs to getPos seek seek and report most of which are part of interface +Please confirm they have been applied as expected +I just committed a patch that modifies start level handling such that we always process transiently started bundles synchronously which means theyll either be started on the calling thread if they meet the active start level or they will throw an exception if they dont +Committed revision Im leaving the issue open until and have been fixed +Im not sure that fits better in +The property tells Remoting to not destroy some of the resources it has setup immediately doing this allow to reuse the SSL Sessions +When you are ready with this option? I am waiting for this +site changes reverted for discussion +Another solution is to HIDE by default while running +See for further details. +Agree that this should really be a separate project as its at a much higher level than mapreduce per se and it should be able to utilize features from a framework say an alternative BSP implementation. +If the former we are not broken for this case by returning a dupe +Once again this will be punted mostly since became more of an incremental and release +Try to be accurate I think it is the method findResources +Explore the possibility for the binary I didnt mean to imply anything about tokens Im saying for just add the same handling you have but also for I will do that will switch back to on it and open another JIRA for the token probably also need to validate again that the cxf binary disptach is still inadequate +We cannot execute usrlibjvmbinjavaCan anyone tell me the correct JAVA path for buildbot +Applied +Thanks +pm files that are attachedThanks SQL is working for me though I am setting SET foreignkeychecks in it which probably isnt the best +WhyAnyway I fixed it using both and issue for callbacks +Ill dig +Attachment Contains the test failures +Can you try this on trunk or branch? Its working ok for me under OS X and I dont have windows handy +However even today Ive seen people claiming a configuration with a seconds timeoutI understood the first part +We should explicitly check the exception name to be instead of the message +At that point it seems best to approach the Maven community for adding the plugin to the Maven core distributionAs for JAXB Id suggest to develop the plugin for only +Verified in documented as a known issue in the release notes and in the Getting Started Guide under upgrading applications +Forgot to mention restarting the datanode did workSo probably this only happens when it has a temporal connection problem to the namenode +Well the absolute time reaches the server correctly which is the important part + +Sandy can we get a test for this? I see only checking memory resources we can modify that +Sorry Drew for the typo in your nameEric +Create +But Ill happily sit around and watch people get pissed and have a good laugh as my admittedly poor skiing skills start to look pretty good against the drunken horde barrelling down the mountainside +committed +Unit tests included +Committed to +Closed issues related to Roller release. +and +this is exhibited by the dev be fixed when is fixed +This is already fix in weld upstream +The service should be removed from the file as this is handled through an ant task which Kurt wrote +But can grow dynamically so that later they must be returned to another pool which contains the larger buffers +Manually need for integration tests. +Projects could use svnexternals to link their status file from a location in incubatortrunk into their project tree +Create new Windows native IO functionality that solves the above issue +Eduard Can you try to run the sample with cxf default http transport firstThe Message Address is in the soap header it should not be affected by the JMS transportI will dig your issue when I have time next weekWillem +Thanks for the patch Bruce +the add to hadoop qa +JPDL and BPMN process definition +Verified in JBDS L +i was already the admin so i already had the observer role +This has already been fixed and merged why was it reopened? Which QS shows this +Faizan was released on Aug +Add an assertion in that checks that indexes have no compressed fields +Will add more as more features are added to the console +Hence assigning this issue to him +Caching is always a trade off between processor and memory +We are going to provide shell access as well as programmatic access +default server profile has unsecured +Heres a working patch +Could still be a bug in Camel though +Still not working in also doesnt work in +UIMA AS source LICENSE and NOTICE files are stale +Are you logging in with the username The server is case sensitive so make sure you have your capitalization correct +Its a backward compatability issue +I have exactly the same issue in Weblogic m also getting this problem using weblogicThe issue looks like was reported in issue tooIts not actually possible to remove all the other JPA implementations off the weblogic classpath as weblogic uses the kodo provider for some of its internalsThis means we cant upgrade our application to use the JPA api from the hibernate implementation at present unless can find some other workaround +The content assist engine prefers to show property proposals over field proposals +I can cd into any one of those +I missed thisI looked at adding an override of start that took a +So the Qpid specific stuff needs to be written in C and interface that to AxisC with C +looks like I was wrong about those tests failing just because of the same Reader looked that way and the expected fix worked but doing things correctly as directed by yonik now all the tests pass no problem +This will be for not. +postponed + +But where is also a link under header of modelPanel +Thanks for the patch +Next few days Ill add more items to it +Bulk move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +also interested by a review of inclusion possiblilities for m very interested in the bug resolution on RFIs there any workaround till then? +confirming that I am seeing this in versions and with +Simplified how the implements the JNDI Referenceable interface by also implementing JNDIs and eliminating the inner class +Could you attach a minimal test caseSort is applied at the Lucene level before any kind of pagination so the behavior you see is odd +If youre esoteric you may argue that this change may break existing users that serialize the exception and lack the class on the deserializing side +Created new issue in TAP +Hopefully this will find bugs in Scorers! +setting to since theres no this quickie disabled if several adapters are selected! enabled if only adapter is selected +Adding sequence generators for Apache Derby +Im not sure this needs to be fixed for and +However it was very hard for me to follow the application logic of all these multiplexers +Michi Any updates on this +Patch which fixes missing endif as well as failing if both async and asyncctp are specifiedTested on Mono +This is a patch to add these two properties to the documentation +Thank you for work on the Hibernate project +I did not try to test this in +fixed +But isnt that what a classifier is for? is fixed now the linux also using +You want SOURCETRANSFORM if its not an internal transform +Fix now in CVS. +I think you can use one of the methods to add the plan without changing any apis +Note that with zhengs work on lazy serdes we would remove the dependency on thrift for serializing and deserializing data +Fxied in cvs +It would also be nice to be able to configure allowed content types file extensions. +I understand your point but this worked in Axis and passes the check in WTP so I suspect it is legal even if it is logically suspect +verified in JBDS +The patch was send me from Kay Schieck +Forgot to add zip file + patches works great committing to to branch +Updated patch attached +The patch should be updated after commit commit +Patch is not applying cleanly on trunk +I have updated the code and the test as well +Did you try v patch? That was the latest one and has no feedback +Usage extract the attached into your Seam projectss action directory and build and deploy together with your application +Sonar does not use the Checkstyle maven plugin anymore +Here is the patch having some product content type +Fix Mikail for pointing out the flaw +Typo on data Do we really need to make the constructors public +good catch +postponed to want these issues to get lost so I think you guys should investigate them first +Ok looking into it +Thoughts +Because of source editor +Linking to +I meant that the contents of is now large because there are a lot more edits on the working table +Sorry Andre +Never mind I am using a private Nexus server as a allows rd party Artifact Upload without programming any POM is really sad that after this only I can can build my open source Maven built very open source any more +There might be others that Im missing but those are the ones I have touchedIn addition we will probably be implementing our own metrics classes +Ive reviewed your solution and think you should apply it +Hes the only paid developer on this project +Logged In YES useridRequest id In YES useridThis has been fixed for +Thanks for the advice on this Ive created a filter that wraps the request with a that overrides the get method to return an empty enumaration if it was going to return nullSo its all working now. +Eugene Should we just incorporate into? What do you think +Let me know if anyone sees any issues with it +The issue was a static reference to the old version of lucene so that it wasnt possible to update lucene without compilingWith version Ive updated the from version to the current version. +The patch is updatedThe peerClusters is still because making a concurrent version would require modifications irrelevant to this ability +The patch failed core unit testsI dont think this failure is related to this patch because as mentioned before its a build change that is used only for eclipse environment issues and this should not hinder the running of ant targets that are responsible for running testcases. +Comments added on review board +trunk +In practice for any role other than an this leads us to mark the plan as user specific +Can you add an r? some minor comments +We are currently on and I am still getting this problem +trunk Initial project added with contains a +I posted first cut at a patch that adds the timestamp + +Hi AteIm back to my project and do now need the struts bridge in a productive environmentDid you find some time to commit the fix? I tested the latest head and this issue still appearsThanks in advance ThorstenSorry I let this issue slipI still have the fix laying around somewhere but I need to review it again as it has been quite some time since I last looked at itIll try to allocate some time for it real soon maybe even this or next weekRegards for your reply +Is it possible to create a test to cover this code +Test war that displays the in AS and access httplocalhostjsf to replicate the problem +and the guide still mentions it. +In that case a may be being used to serialize a which wont work +joseph the zip file is not an updatesite +It works fine +I was waiting a bit to think about some issues some of which were just discussedSince the collapsible screenlet was converted to we now have a problem +In fact thats what this bug was originally complaining aboutThe lib directory is like that in war files +Lets keep tracking the DBCP issue here since the basic problem is the same +Counter intuitive doesnt even come close to describing itIf Im reading the spec right it talks about the capitalization of inferred property names set In EL we already have the property name +Bug is missing the commit showing being updatedCould you also comment on whether how you addressed steps before marking the issue as fixed +Thanks Mamta +Greg Ill build a patch against the file and will attach it shortly +A bug needs to be left open until it has been completed provision of a patch is only one stage in the processThe patch may be rejected or may require amendment +proposal for implementing basic dispose long have had a dispose method after this discussion +are you using as well +I have a question about the right operands of the in list +This patch causes Derby to lookup VTI optimizer overrides using the sessions classloader +If you need to them you probably will have betterresults if you use matrixvector operations instead of explicit loops +So yes without writing something currently or patching Maven its not possible to block +No tricks to reduce the size +Reran all test suites and fixed another several wrong test results +I did not realize there is an alternative to the Autowired it mean that my classes can be completely free of spring imports. +bulk close of all resolved issues. +I think Im fine with your proposed solution +Benoit is correct Constraints contained in your validation Rules are only ever applied after type conversion +Verified +Making patch avail now that is committedRebased patch updating as the constructor changed signature +I created new workspace e project copied my entities e went ahead! Thank you +I can develop it if you are interested +including the correctversion of digester in your ear and setting the classpath property in themanifest of your? OliverI ran the unit tests with and without the patch failed inthe same way both times at testLoad +This doesnt really affect any version +Changed the to throw instead of for which this log message was originally added to expose any issues with serialization +for the patch +call anyway +looks good to me +More to it after having the same problem with Mojarra and Tomahawk I tested with other versions of Mojarra and both and solves the bugI recommend changing the status to closed not a bug or something similar +It will work however as long as underlying objects hashcode is based on value which will be the case for primitive types and containers containing primitive types which is the case for Hive datatypes +The servers SPN is always the authentication principal in the UGI +This patch applies to branch and trunk +This task has been completed +r use cout and fnescapeXml in fnescapeXml in prevent xss attacks in build agent local repository and purge added for XSS vulnerability fixes with UT and Selenium patch in r with some modifications fixed validations in project group action and build definition action fixed selenium move validation to xml files remove regex validation for description and just escapeadd validation to prevent xss attacks in added validation in xmlrpc fixed validation of artifactid in removed regex validation of build agent Merge changes in trunk to branch +The lib directory were talking about here is the one in KARAFHOME +I thought timeout means that never timeout +Not sure if its the real culprit or not although its possible we have more than one problem +Moved to next milestone release +closing issues for released versions +Saminda fixed it and I just tested with latest trunk +What all non working phones got in common is a sync with Exchange if you turn that off and try to import all works fineHave you had time to test with a phone that got exchange sync for your contacts? +Thanks Vaclav btw what is the ETA for +Messages are truncated +with a patch that fixes the problem for meI shall resubmit the stuff with the Apache License later but surely not today +sorry for the late on thisNicholas I think both patches may required to apply +This is documentation issue +Go for itShould we move it out even? Should running on and hadoop be a target for the next major hbase release +I took an easier root as mapred may not be on every system I knocked that little test off +Can you make sure that you use +Feel free to come up with that issue on dev list so youd be advised how to resolve it and simply let others know what troubles they can expect +Also it is not efficient +Now the has failures +Linking the correct ticket and in the correct way. +This patch in addition does +test added in svn patch addresses the issue you discovered and should fix the same issue for other simple XSD types +Is there an auto move from the Maven to the Maven repository or should the Maven POM be used to publish into the Maven repositoryI guess this raises the issue as to whether it is sensible to have three parallel build systems Ant Maven and Maven or whether to restrict this just to one +What is the finalizer java finalizer? or you mean garbage collectionIgnore this comment +Marking CR to make sure its in our radar but not necessarily our problem +I am glad to hear that the next WTP Eclipse version will ignore it +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +batch transition resolvedfixed bugs to closedfixed +This problem is really annoying! Please fix it +Add the tests to the jdbc suiteM javatestingorgapachederbyTestingfunctionTeststestsjdbcM javatestingorgapachederbyTestingfunctionTeststestsjdbcAdded the methods to the Blob and Clob implementations in the am interfaceM javaclientorgapachederbyclientamM javaclientorgapachederbyclientamTests Runran derbyall and junit and no failures were there +Let me know what you think to work for me +didnt see your comment and already committed to trunk and prob +Implementation of the Nysiis algorithm +Thank you + +Thanks Amar! +Its done in such a way that most of our users get going with it asap +Attachment has been added with description demonstrating the odd page content overflowThis patch to in conjunction with the patch attached to Bugzilla seems to resolve the issue and passes the unit tests +fixed in rev +Alexs resolution to also resolves the remaining issue highlighted by this Jira. +Sorry did not traverse through the full thread properly +I think that is what you also meant +Once the complete implementation is done we can look into more accurate estimate of number of tuples etc +Wayne the latter sounds like the best option since it would at least let us run correctly on all platforms +Now that work is underway with AS all previous community releases are +Damon can you please verify the scripts now work as expected +Minor update to the patch should continue to be an and not a +lib +Or you extend Script then you still need to pass the binding but all other things are done for you +I thought the point was to the conditions which caused +Fixed in branch in +Also should we close as duplicate +There will be no enforcement of whether the target is a directory or notTo be clear under your proposal if one does link hdfsfoobar to sbar and sbarbaz exists if one tries to list hdfsfoobarbaz one will get a exception +Improved patch code cleanups added +I think these two could be kept in the patchAgreed that trying to speed up test execution +Thank you Dave for these tips +Microsoft SQL servers dont even support +Pl +How would I annotate that +Should the app contain multiple tiles for each website or should it have a common stream? For example if a person allows Picasa access he gets Picasa photos added to the already existing Facebook photos on his list or should Picasa photos be displayed under a separate tag? appears to be a great project +Maybe Camel should thrown an exception and fail to startup +Im closing this issue as the original issue should be resolved. +i will close as will not fix +Hello BilginThank you for the patch! I made a few changes on it and committed itI also updated the WIKI page +So I cant really consider them as separate patchesSome bits found in second cleanup patch seems to be the preferences pages and their hookups and hookup code for the semicolon remover into the +Thanks Robert! +I mark it resolved as wont fix then thanks. +However since there was no reliable reproducer the fix is based on my assumption as to what the cause might have been +Changed example to avoid web context the project name and WS suffix so projects that also include additional services that have under the project module name dont also clash. +Indeed +So can you attach the original project and how the view model was created? Deb +Backlog is correct Fix Version No Release is for stuff that logically doesnt belong to a release +Having the ejb reference the tss prevents this from happening +Are you running +applicationstarget +Checked out and built the branch +Patch to introduce sharded joins +I see two issues Coupling to Service annotation +This issue happens because of a libxml issue which is out of scope for us to fixThe solution would be to either use a version of libxml that works or use Guththila parser +Jason you can as you feel appropriate +Done +Yeah +When it pops up at an interop we should take care of that. + for this issueAfter this issue is resolved Tajo can have date type right +Junjie if I am mistaken please reassign yourself +I thought that was guaranteed for static fields but perhaps that was not the case before Java In which case it might be worth using temporary variables to populate the Entities in the static blocks and assign the static field at the endIts a pity that the static Entries are not immutable once created +Before uploading it I wanted to double check that it really was the problem but I can no longer duplicate the problem without the patch +I found that problem too I have looked into your patch and now I know how to deal with it +see resolution of linked issue +I have setup a testing env for enabledisable the logging objects for times with no connection waste anymorethe codes may have some crashing issue in the other parts of the logging system we need more care of the logging issue +If returns not null exceptionEncountered is also falsewhat about naming the boolean alreadyClosed or closedBeforeor other similars +Ill take a look at this +Thanks boss +Fixed in Branch and BranchSP +So I am asking whether you feel strong about making it an undocumented config parameter +This is because I added code that transfers the jars to the distlib area into connectorscmis a while back +That being said I dont think this is a jQM issue +As far as I know we dont provide any Eclipse plugin implementing project facets axis +I have no idea if we even want to implement everything that I had in the list +Since is committed to I think this should go into as well + initial patch for s unlikely that would go in to +The query handler is not fully transactional +I dont understand why there would be two notifications for a single ll test it again but it seemed to me the email I got was the normal build notification Im used to it just had this new hardcoded subject line +Menu items are added +Well let interested users subclass because I am reluctant to add additional Spring API dependencies unnecessarily +Ill have a look at this +ThanksSean. +add new timer metricsunit tests passed and the new timer mBeans are verified +I think we should try to limit Nutchs core code to a single XML parser and logging API +This issue has been fixed and released as part of release +Now that work is well underway with AS all previous community releases are +This is a different bug +just committed the latest patch to trunk +moving open issues in X DT to X DT as discussed with Bob +start the repository instance and run your orderBefore code againIf you still see the there must be more code involved to reproduce this issue +Will submit patch and tests by end of the week +One of them should be changed to indicate the ASF JIRA usesReference relates to is related to +This issue was migrated to +If ok resolve this +Just need to sift through the code using these extras extended operations and make them use the codec instead of manually access private implementation classes to do the work of encoding and decoding these requests and responses +You havent changed anything in the patch +Whats the consensus on this? Do we allow spaces or notEither way lets update the docs so the decision is clear +I have the same suggestion as Etienne Bernard in comment +Added documentation to +This problem occurs also in someone actually describe what gnome shell crash meansIs there a log is there ane xception? does eclipse itself crashis it the JVMis it the whole window managerWeve had many fun moments with Eclipse and Gnome Shell so actually getting some info on what kind of crash it is we might be able to find the duplicatecause +I think all properties and permissions should be maintained +We should have changed the name to guinea pig instead of removing pig +ok sounds reasonnable use caseDont miss to create an issue in MASSEMBLY too for upgrading the lib version +Verified in reopening to update release note info. +bulk move of all unresolved issues from to +One more comment +it should be fixed now +fix cool Ill play with it this evening +Yes +Does is make sense to support multiple listeners? If yes we should implement the observer patternI do not have a use case for multiple lister +I am updating this issue to mark it as closed for future comments as it is now past the deadline for incorporating comments +Dan added this. +We tried to solve this by changing the +Tried tinkering with some settings but didnt see particularly good s test case if folks fancy are already reusing a compiler instance to speed up subsequent compiles +This has not been fixed so the question is really Layers View +Alexey this is not debug information this is the information which a user expects to see if there is something wrong in users code +am i miss understanding something +Committed +Weve been hit by the bug whilst attempting to upgrade to release plugin +Do we want that? +Done. +These annotations such as Label Description and Category reference the In messages by string value rather than In that a helper method was added to the that is used in each module that uses In constants +This uses a CI version of batch admin on the th of March +However what does not work in IE is that the menus are disabled on this specific example +Hi WillemSeems the xmlbeans isnt in maven central repo would you please add necessary repos in your patchThankss my bad I will update the patch shortly +This may be too restrictive since is a legal identifier part AFAIKAlso I have thought a bit more about the problem +adds a merge tool which can be invoked from +This method now basically does nullb leading to the nullnull you have observedWe can of course do a special case for nullnull but you actually want a NPE for the cases of any of the operands null or not? That would break codeSo I kind of tend to close this issue as Wont Fix +Harsh J do you want to update the patch? Im happy to do that as part of the commit +This fixes the problem simply moves the attachment to after the object has been put into the cache +Sergey! Im planning on committing to trunk tomorrow unless there are objections +New patch created +The error occurs not on my development machine but on an server +Open to suggestions here +Closing +Thanks for the contribution Eugene +As Mike stated Id think that theres something unique about this test environment otherwise wed see this in or runs in JEE environments +Jan to me this seems like a matter of preference +Im trying to build a release using Eclipse and the Eclipse build is failing because the signature of has changed +This JIRA is partly resovled +Thanks +Hi Ivan can you comment on what testing youve done of this patch? Did you perhaps modify a test to artificially induce an OOME and then observe that the heap dump showed up in the test output? For that matter where does the heap dump show up? On the console? In some test output file +I was hoping it would be fixed with the release of Maven but unfortunately theyve removed support for profiles in which uses +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Thanks for the patch AshutoshThanks for the review Stack and Ted +anyone got a chance to review the patch +This may be either issue of my environment or plugins working flow dont archives work on another machine I would not insist on further investigations just in case youare curios about it I may provide some additional. +Xuefu! I have committed this as I dont want people to get used the old structure and then change it on them. +This patch is just refactoring so it doesnt include tests +I suppose other suggestions could be made but I dont have any and we discussed it with them for a long to appropriate version I guess the ship has sailed for WTP so cant be fixed in is suffering from a lack of imagination on how to perhaps fix the icons to show vendor icons while still able to show the state of the server +A with the reasoning too +There is something I dont understand with this patch +Ready to go +It first fails in since it doesnt find an instance an instance with id doesnt exist in db table I added this into db then it failed in which wasnt authorized for but thats probably a continuation of not being properly initialized before. +Verified by Sian. +no need to backport this one any more as the login timeout feature is new tocommitted fix to branchsibmsvn commitSendingSending javatestingorgapachederbyTestingfunctionTeststestsstoreTransmitting file dataCommitted revision +this looks good to me +sometimes things just fall through the cracks. +First release done +The patch is looking better but you are still checking one of the arguments for null but this time you are returning null instead of an empty String +if someone wants to work on this please go ahead I still plan to just no time yet +Thats the thing Id really like reinstated +I changed to another server for test +Thanks Knut +If I still fail to get this running correctly Ill send you one by email +alexey this looks like a duplicate issue to me +I deleted the index file +Hi all Will somebody try this patchThanksLeo +I hear you +Web Tools Platform Version is the truthyou both use correct version the reason why it is behave so weird is that Slava have no plugin in development environment but Anton found class that add that supernatural behaviour here it is always return true and thats the cause of exception +Please try again with patch and didnt experience the failure up to now +If you dont put requiretrue then svalidateAll will to correctly validate missing fields later +No test case added as it is a script change +bq +is not application specific user can also explicitly say get should not be stored along with app infoOne question is does ZK support directory removal meaning if the parent node gets deleted will its child nodes also get deleted along +thats pretty much unavoidable +function was indeed kicked out in PR during all of our its trivial to do in a listener I wont reintroduce it +Yeah I was wondering after my last comment if that might exclude the fix works thanks +in +Scheduled for DC +war in a child directory of deploy and attempting to delete it from the UI now works. +They behave like before +But in fact it upsets my SOAP server +You might find some issues when trying to do it via the apis +Well look at fixing this in it is a hard problem to solve as we almost need to maintain a full AST graph of dependencies to figure out which classes to reload for any given change +Sending srcmainjavaorgapacheservicemixcamelSending srctestjavaorgapacheservicemixcamelTransmitting file dataCommitted revision +Great stuffOnce I have moved the java connector framework into its own SVN area Ill add your PHP and we can work on it from there + +long markup differs between IE please view + is called first whether it is able to resolve schema without internet access +I logged about the x evaluation per exceptionI cannot reproduce this error you report here on trunk +Rolling back local change and saving the experiment for posterity +Declare depenencies we were using anywaysM M M M M Purge unused slfjs +Nice results +test with legacy pooled invoker which uses the JRMP invokerproxy ha so dont foresee any issues +But like i mentioned before adding clean up of keys is something not necessary for then well consider this one closed and finish the rest in in +I cant imagine something from still being relevent +Should we close this now that and are in? Or do we think there is more to do +I did get some and had to increase the memory for applications +It would be good to figure out what commit broke this +No IT +From the looks of it patch is ok +But I wonder if that couldnt be done later as work? Would you say that the changes Ive proposednamely see if we can push the predicates to both sides and skip the push if notare incorrect or just that we could build on them to make them betterYour changes are not incorrect +you did rerun +Unlinked which is related only to +svn ciSending Sending siteSending siteSending siteSending siteSending siteSending siteSending Sending snippetsSending snippetsSending snippetsTransmitting file dataCommitted revision . +Fixed with +This was fixed in a recent revamp of the web site. +I think Godaddy is just a load of crap +Dominik is the reload operation broken or it doesnt work only using the testsuite framework +I dont see the line Exception during image upload anywhere in the log you uploaded +I will take their code for a spin and do some benchmarkingtests +Hadoop was released. +It turns out Camel context has a little that discards older endpoint instances +Can you attach the server logs and the esb archivetest client to this case please +much better +Dave let me commit the patch! +upload a patch store the version info as a separate file in hdfs and separate znode in zk + +However no comment or response so think implicit arrays are a quite common use case +After this user can do what ever he want with the templates +The basic idea is that you can bolt together a header the default header replicating the current header +Issue seems to have been caused by an older version of H installed with the ESB that was overriding the version bundled with not been able to reproduce the issue since the update but if this then please the ticket. +Completed the central service catalog along with a registration servlet for programmatic registration and a console page that allows manual registration +In the example above the atrributeN of my action form is a Collection of bean and attributeA of is a String +Many many of testing went into this +Thats right CMS using Maven to build it +This can happen from time to time since the timer is invoked every seconds and shutdown can take up to several seconds when running in cluster +I have Printing of Maintenance implemented I am waiting on this patch to be applied so I can prepare other patch and post it +As the first fix wasnt sufficient +Retested with xerces appears to be resolved in this version. +I think the system property named with as prefix is OK I will head on the JMX wiki document right now. +hi KumarI guess you are using on client side and the version of axis on server side +The way we found to avoid the case in one of the jiras I mentioned above was to resend the last message the peer dequeued for a given destination +Removed unused imports +I have applied itThe quartz osgi test passes +Gem dependencies are right on the server but it looks like my local rubygems cache was fubar somehow +The patch does not apply anymore +The page attribute type is now template just for consistency with other tags +reopened just to set the resolution to fixed +Please set the components in the issue thats why they are there +Minor changes to earlier patch mainly added additional tests to test submitting jobs to non default queues and also included the queue name property in +Corrected get as fix causes regression for clearcase on Windows +Turns out this test wasnt really disabled was weirdly calling our disabled test on the side +The admin user created by the CLI installer has the time +Should I attach it here or open a new bug +Forcing people to build bit on bit machines is a no go +duplication of such a small amount of trivial code is far less evil than yetanother dependency +Yes this is a bit inconsistent but this only happens if you have two stores that shares the same namespace +You have forgotten to add void primitive type +Attaching a new patch for trunk +to too. +zip with existing logo is rather ugly so I decided to play around with some ideas for new logos +I am inclined to check this in soon unless you see some problem with it +It keeps the scripts simpler. +privileged should be in command line for uninstaller in trunk and branch for running uninstaller from Start menu +Committed to trunk +final notePerhaps the logging level is configured differently +Since is private I cant override it +In those cases the engine would declare new concepts and discussed in the list due to our full support to class hirarchy and interfaces there would be no gain in creating s in advance as we would not know if the nodes to create should be from the concrete class or which classinterface in the class another approach was taken with the same final result +Is configuration property is set to true +You can now pass in jbossx to keep the task from adding xercesImpl and xalan which cause failure under +Since a single Pipeline could map to multiple MR jobs we might want to have some settings apply to only one of those jobs but not others +You dont need to use you can use libjars to push jars to the cluster and have them in the classpath together with your jobYes but you have to be able to find the jars which the jar finder does +New version +I suggest to resolve first afterwards the typehandler can be returned as subresource the http method annotation and the call to extensions becomes obsoletereturns now only typehandlersto apply the patchsvn cp parents trunk issuespatch p i think the synchronization on line isnt needed. +Was there any respose to what David saidIs there some compromise or change in stated scopethat would allow this enhancement to take placeIf there is no action on the listBugzilla in weeks I will close this as WONTFIX +Glorious +This patch looks good to me +so we can repopulate ZK in cold boot scenarios +You should ask this kind of things on the Hibernate Search forum. +So if a data dictionary cache mode switchoccurs which has the side effect of clearing the cache and then addcomes along but there has been no intervening call to thes will be null leading to the NPEIll try to work up a more clear and coherent description of all of this processing this isjust some rough notes to record what I think I see in the heat of debugging +Im going to commit this momentarily +If nothing else can you describe if youve gotten it to work if so what steps you took and if the steps are not obvious what libraries for Grails Spring youre using? Thanks for your help +Im worried about cases where we over schedule the first half of the cluster and leave the other half idle or remote scheduled +I did some playing with the config in question and was able to fix it by sorting out the namespace decls +Got it + +Using the patched jar it happened less but still happens + +The issue with handling updates with deletecreate is that the updated object will have a new uuid which would then break any any prior references to the object +Please check with for a resolution to your problem. +Please let me know your findings so that I can fix more +No I was just going through some open issues to see if I could lower the count of open issues +The finalboundary must be followed by to terminate the content +In boost I believe its all preprocessor based +closed for final release. +Is this behavior specific to iPlanet version +Fixed by r +I dont know if we should consider this to be an error +Pushed to master +Committed revision Thanks Julien +This is my try to fix it +No need for IT as the unit test exists in the Checkstyle project. +I was thinking there would be a ant file that has an +Fixed +Committed to branch too. +Need to think it thru +dont look too bad +If the target install directory should be cleaned can be specified with an additional parameter when using the PEAR API +Matt I have created a partitioner and applied it at assuming the valid cluster Ids are the latest and sequential +CorrectThanks are correct +So the ability to attach a transaction and rollback is that a feature or a bugIts kind of cool that JMS can be used as a remoting protocol in this manner it basically becomes a way to cluster remote objects without having to set much up +The alternative is that we need to document that any statement may contain a label and it would be up to other tools +Attached patch fixes the bug +Its not a perfect solution so I think we will refactor it at some point in the future +HiI just wanted to tell you that this problem also occurs within the version Here are the attached sources modified for +In my personal opinion this config change was useful and extremely unlikely to break anythingBut thats not the criteria weve agreed on for stuff that goes in after the freeze +However other databases systems are suppose to have real admins and consider that only admins or superuser can delete a database +In case when there are many clients it effectively worses the situation because it will keep connections open and lock down the server with idle connections +Hi Grant Regarding there doesnt seem to be any change +Bernd can you please try +I dont think thats what youre asking me to do anyway +GillesI like the changes +Suggestion is put to new issue. +Uploaded the patch +Patch implementing above proposal +This defect will be fixed as part of. +I never delete cookies +Thats not a matter of naming its structuralThere doesnt have to be any trace of vertices and edges in the resulting output and whatever system consumes that output doesnt care how it was producedSo to be honest I think this is a but maybe Im still missing the point. +things likeReally you are right if we have a dependency we should be testing that its workingBut its good to check the inverse at least we know we are not depending on things we didnt know about or the wrong versions thisonly complements that +I will give it a go. +Not sure whats up with Jenkins +However those operators and seem to need more optimization and cleanup +Wondering if there is anything about the Jenkins system that is preventing execution of binsh properly +Im fine with startstop and other discussion happened while I was in vacation tooWe also need to make sure all the changes conform to what we are discussing at +Closing resolved issues for already released versions +The obvious disadvantage is code duplication +Thanks Mark! +Check your spam bucket this is the most likely location for these emails. +The log files are equally long with at and MB +Alex I am in the middle of making a release +But the circumstances doesnt allow for it +It seems that the work for this issue has been done. +Should I submit a JIRA saying that sconversationPropagation should only be propagating parameters and therefore it is a bug that it propagates page parameters +Can you give some more details about how you intend the new design to look like +It comes from a predefined set of places but point taken +The second click is ignored just the first one is triggeredThe implementation is more or less the following +Has anybody looked into this +I have added code in Git ID fecfbaefabaadabeb to trap this error and display a more useful message to the user +Hmm +thats an old versionI have fixed some cases with that kind of message +please help to verify +I think a problem with the current is that it has too many responsibilities +In my mojo ruby script I want to require another ruby file +Is the issue that its already present on those platforms and requiring our header breaks it? Could we simply ifdef asprintf and only require it if it isnt already available +note the use of quick code review indicates that we are infact handling things will need to wait on feedback from the community against before closing this bug +I showed Joe the relevant verbiage from the specRemember the problem is about not seeing the declared notabout whats in that +I am going to commit this sometime tomorrow unless there are objections +I didnt use the but network trouble isnt completely out of the picture +I dont know how to get the source since Im not an git expert +bqPerfectNew patch w that +on the patch +Cannot reproduce +Please check it was applied as you expected +Yes it works as strings in +Just a question +Were under the impression this might no longer be an applicable bug against CR but rather a misunderstanding between RVM and yourself? If so wed like to close this +Richard I will agree with your above comment for package import + +Not a general solution but something that works for similar cases at least. +You need to look at the jsr mbeans to view all components in an ear deployment. +sets b to disable passage length normalization or something +I will append the flag in front of the regexp instead of using a delimiter +Thanks Amar +but its not something i though through to hard as i said work in progress +complaints here +Hey Barry adding you as the Reporter so I can be accountable to you for this JBDS this issue since the book is ready to it needs updating before release please this Isaac +In this way no information is leaked to a potential attacker +sorry it was mistake +Added more test in and incorporated Devarajs off line commentsremoved unnecessary synchronization in Jobtrackerhistoryfixed javadoc in +I am waiting for these two things from you +The attached may be deployed to a Tomcat server to test exposing a remoting service via the Service annotation and the included may be run to test linking the service in a a client +Which version of adb will this fix be included in? Is there a tentative or certain date when this fix will be available? If not how can I get this patch +Fixed in CVS +Tagging and lowering priority to minor because there is a viable workaround in place +Thanks for reporting sorry it took us so long to react +Attached patch to validate content of list without assuming certain order upon database retrieval +Cassandra is testing this over in +This was resolved via IRC was a configuration error incorrect host name cause problemFollow on jira make this error easier to detect. +Great based on that I am resolving this one +This overrides the class in the mina library +I would like to see this fixed for branch as well +Closing this as there has been no response from the reporter. +However the TEST +Testcase that reproduces the problem +I am sure there is benefit to updating all the required libs but then I dont know how important each lib is to HS +Added unit tests +bq +Alternatively the onlyGlobal option should check for the presence of either keysAnother good point since right now i only checked since is deprecated +I doubt it a widely used feature if its used at allDo others have an opinion on this +I am not sure the workaround applies unless you are using Maven directly +shouldnt there be a or something in the patch +Final patch +Maandag en dinsdag ben ik niet op kantoor aanwezig +I just committed this +Find attached a packet which holds the wsdl and its schemas which WORKS fineBr attached the wsdl and its schemas which cannot be processed by wsdljavaBrJust to make sure that you know which packet is which +to commit to branch and also trunk +If the user is messing up with non EE applications in an EE environment he needs to provide the EE + trunk +The rest are the resulting changes to the files +like my patch will fail some tests +I dont see any warnings anymore so this can be closed. +Does Rebuild button fixes this? +Patch with testcase alone to illustrate this bug +If you upgrade to the latest version in the repository this problem should be fixed +I just committed this to trunk +My bad. +Closing. +This code would only be enabled temporarily I dont expect it to ever be committedIn any case moving the current method would be a good incremental improvement in my opinion as it gives you better information when the test fails +Please Verify +Well depends on which depends on +Thanks again +Ive just committed this. +Imo we shouldnt simply remove it from our repository +Also annotation gets ignored +bq +But there should still be some eviction algorithm or at least a so the JVM can cleanup the cache for unused fields +This has been released with +Verified on revision +Do you mean the fix wont be included in version of tomahawk? Id really like to because Im currently working with and I would like to have my patch included +Externalization of the tomcat authenticators is done via the Tomcat sar configuration +HiI have attached a zip file of a striped down version of our maven project so you can reproduce the erroneous also did some more investigation and found out that this error only happens with Maven but not with Maven and +Closed issues related to Roller release. +Work on this issue is now mostly done anyway it still needs more testing +pushed +patch adds some javadocs makes use the default connection manager configand moves date parser formats to a configuration item +I agreemaybe we can put the just in the tests folder and not withthe alg files in conf? do we really need it or maybe we can use something smaller +is generated by ANT class +Committed fix as svn closing. +An evolution of my opinion on the matterwe can create groupsregistries on top of current like etc +I should have it by your use case the fix approach looks reasonable But what about the normal put flow? There will be perf degradation as we have to deal with Set now for every row in the batchWe should test the write throughput with and without patch +Added simple unit test +Wiki is aproaching +The NPE is likley caused because of failure during the acquisition of the hence its not establishedremains null +Should we prefetch in multiples of the HDFS block size +Whats the warningMakes sense wrt the Parameter class + I fixed it +With Wicket the problem seems to be fixed on Mac OS X +Now that work is well underway with AS all previous community releases are +How do you want me to combine these? There is overlap but Pavels dropped a few things from Geregorys and when only Pavels applied I seem to have the same problem +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Apply this patch so that Steve can test before we put out the next release +Merged to the branch in revision +log file please? with the exceptions attached +No I meant I tested both +moved since jboss tools does not have an installer only assume you find this when you start JBDS on an existing workspace +test it on +Patch option to use distributed cache for B matrix in AB job +Josef sorry to bother again but is there a chance you could post some of the discussion from the customer support portal here or summarize what was discussed thereAlso is it correct that this patch will be generally available to users in AS and EAP? +Committed at subversion revision +So everyone should be lucky! +Paul I have done my first review now +Please confirm this fix +apply cleanly to as well +This hasnt been tested yet +These blockers and critical issues are resolved but not verified +That said it is frequently said that ASF projects are a those who do the work get what they want or needAs for the SMTP Appender I really havent had a need for one +I also agree this JIRA is in the core of MR and merits for extra attention it gotThe current approach which has been done following your recommendations has been already thoroughly reviewed by you and myself +Actually here what happens is requests without a content body are not counted +Here is a patch that fixes the issueThe included unit test isnt optimal though it does tend to fail on the unpatched code and always passes on the patched code + Looks good although include should probably be include multiple keys can have the same token +JPA processor currently ignores those facets with default values that are not greater than +Reopening as an enhancement +I have added an test which uses the servicetest channel +Closing this issue as suggestedCreated issue to fix testcaseadding DBSybase to release notes in as Wont Fix +The only practical difficulty I see would be that Id have to add a Level class so that you have DEBUGTRACE instead of JULs FINEFINERFINEST but thats not too big a deal +message senders have an accept property which by default is true to allows for gzipped responses +It now has the ability to have a in the bridge implementation +In fact the problem is that Cobertura internally uses even if its uselessJavancss is quite buggy and cant be disabled +How about this patchIt tries to guard against closed +patch cause these tests seem to pass locally +Running a sample pi job with nodes with common prefix followed by results in the job hanging at map progress +JacquestYour doult on the is right +I think this is resolved now Ive added a test case to try reproduce it and I think its fixed now +Does this still support +Updating translations could be done within that can also give you direct access to subversion so you can commit updates to the translation directly when you see fit +Anyway patch appliedthanks +r I did a significant update to the test cases failing bit more work and some performance tuning is the regression tests are added back to the rubinius rubyspecs +Will focus on this again past the got a LDAP security plugin +so with the patch you end up writing half in tmp and half in the testDir +result using a client with axis build from th april own build th april +But Apache does claim a copyright on the work as a wholeTo Larrys Downstream users still have all the licenses notices and copyrights they used to have +you have now repeatedly made the claim that mkisofs and genisoimage takes different args and that genisoimage as shipped by the cdrkit project is not a drop in replacement for mkisofs as shipped by the cdrtools project +Am I in err to think this behavior is incorrect? +I reran the applicationworkload where I saw the reported issue +This bug is not reproducible with current charset impl so closing as Cannot Reproduce +jar and +Thanks Bruno! +Committed on newtrunk. +According to Marks suggestion indexing is enable by default +closing as WIKI is deprecated and EOL +I remember times significantly less than that for the baseline confirming that now +There is +It also includes the same modification to that I had made toThis time I tested the target and got the expected response since I do not have credentials to upload +By Alternative to this I mean implementing for these caches +You mean File Settings File Encodings IDE Encoding does not affect your test +It makes expressions in WHERE clause use column names instead of aliasesMichal sorry for the delay +The change for should not have incremented oriPort +I cant see any issue with pdfbox and the fact that adobe provides the same result confirms my assumption +Can anyone explain why my latest patch isnt getting test on Jenkins +docsupdated +this is my first patch +on this patch +A fix was committed for that yesterday so if you pick up the most recent version of the code then I expect the problem will go away +We had seen this deadlock problem in our tests +See for more information +dirk +We cant fix it at this versionI have talked to Rajesh about this and my understand was he was ok on postponing it due to this reason +and +upgrade should be without problems +Perhaps if each path had a qualifier attached to it but thats not possible currently +The command line option is working fine +Im not so sure about making these changes just yet +the default value is binhadoopalso fixed the binhadoop script +Refactor in feature managament +After upgrading to the latest version from Portlet app is failing and I have no clue how to solve that Maybe a new Portlet example is needed anyone? +Thanks a lot +No problem Committed revision +Basically moving the client to another machine is the same scenario reversed +This one should +All we can do is disable this for Windows XP to fix this enabled for all windows versions +User guide is attached +The latest patch updates that as well +The patch fixes the problem in where the same file is counted multiple times in the worst case for each directory +Theres of course possibility that well run into some TCK problemsFYI our preferred mechanism for updating the version of would beadd the new openjpa version to the servers repositoryrepositoryorgapacheopenjpaopenjpaThen configure geronimo to use this new version +Disconnected after restart +FlorianTxs for pointing out that was still using +Ok heres another shot +Still got the same behavior is the new code under the SNAPSHOT buildthank you +Cocoa libs also assume UTF encoding +The patch includes a source code change changes to the unit tests to provide test coverage for the feature and changes to the documentation to demonstrate the new let me know if Ive forgotten anything or if you are interested in something being implementation in a different way +Now I updated my patch with your comments +build the sources from cvs +This is fixed with latest builds +ear +will remove +Suresh +Test with latest build from gump +I could build all artifacts without this dependency +Bard this is a language design question right nowChanged Priority to Major and Urgency to High +checked in the trunk. +This is subtly different than the left join fetch of HQL and and the set on the Criteria side right? All you want to accomplish is that the properties are loaded during the initial fetch wo having them specified in the main query as a left outer joinIf so Im all for it since we currently have to touch our entities after theyve been loaded to ensure the properties well need are loaded +B sounds good to me as wellonce we deprecate support for older versions of hadoop we can get rid of this special case code in execdriver then +does work +The tag central has been added to all project examples in reposjbosstoolstrunkjbosstoolsexamples in trunk +The problem may happen again with other jvm implementations +Im ready for kittens +You can monitor the dependent issue +Able to configure autoscaled LB. +You can probably also use the this application for the bugThank you very muchNote Ive reduce the application to only xhmtl files +We required the audit functionality to go to PROD therefore the clone of this branch +Documented as a known issue in the EAP release notes +Addendum patches for all branches +Youre using something with known issues in IE +Passed all regression tests and randomWrite with region servers +WeirdFortunately will obsolete the web ui RSN +This should be fixed now +Redone fix attached +I modified Seam builder to scan only Java sources and jars included into class path. +No bug for the OOM yet unfortunately cluster was before grabbing a full heap dump +Thank you LouisI updated the patch into following manner +A token greater than is invalid +Two initial test cases +Glad I could help out! +Moving it for next release +There are edge case limitations but non PK target is supportedIf you have a reproducible running test case fine but go by the forum first. +Verified on JBT +I agree that stalerelaxed doesnt really make clear what it is aboutPerhapsstaletriggerstalestalestalestale +sure the snapshot counters that Jesse mentioned should use metricsbut this is something calculated from the disk unless youre keeping a map of snapshotstats in memory and you dont check every N second for the corrupted state you cant obtain this info +Yes thats correct +Done +In stream partitioning all PEs subscribed to a stream will be on same node so sender will send event to only one node without worrying about which PEs will consume it +Rather than add the new initialize API can we just add another argument to the conventional constructor for the interface? ie we already require that the implementations have a constructor which takes the interface lets just require they also take the URI? It seems reasonable to meAgreed +btw thanks for these changes to improve the authorization codebase +Since GWT SDK is available through Maven central it might be better to make GWTHOME optional and download the SDK like any other dependencies in Grails +Changing the title to what the JIRA really stands are building a Distributed Hadoop Event Notification System using pubsub model +Sorry I wasnt using Designer correctly +Charles That would be great and also stand chance of getting bigger audience to your question to get more answersreviews in general +Patch is committed to trunk. +Hi David Ran the full build with the change suggested in +are a few issues at play here other than the null constraint on +A search of the cattrack database reveals that this error has not occurred since October last year +Uploaded a fixed. +No regressions were seen in the tests +What a silly database +The solution proposed for trunk is Get the jobid for a job in constructor itself but this contradicts with which creates the RPC client lazily +The current circle causes problems beacuse independently of the startlevel the Provider will never be actiavted before the leving time for mgraphs to be created without using the externalizer +Ill jenkins for sanity +Arc is unwieldy again +Closing per request. +If a particular resource is using properties instead of exposing fully defined attributes please file a specific to get that corrected. +Im attaching my startup and shutdown console log from inside eclipse but Im definitely not seeing what youre seeing +Allowing more flexibility and evolution of implementationsThis does make me realise though well need to load into the applicaiton class loader for some time +Currently in our logs dir youll see both + + +Can you give an example of a case where evaluating an expression at render time is preferable? I dont see any case where such a thing would be relevant for performance improvements +Howard +Is there a point in keeping this issue open? +Thanks Ramesh +All you have to do is specify a WSDL URI and Request Body and it figures out the rest +Attached log and db in +We should ensure that the transition from SPLITTING to SPLITTING should happen only after the master has set the watch on the znode and we should be sure of thatIt would add a dependency to master but may be there is no other solution +You mind finishing the job Martijn +I never said remove More I meant moving methods from up to Configuration and deprecating. +For example was opened by chtyim +I consider the issue resolved +It is really not providing much value +This is not a Grails specific problem as there have been similar problems when using Portlets withhttpsupportuidswgThe problem is related to the use of Jakarta commons logging but there are loads of applications that use this now so IBM should really fix their web server +There is a possibility that upon first boot when we do a bundle install that will grab gems that do not match the CI build of the virtual machine +How important is it deemed by the teamI cant answer for the team but can tell you that personally I dont consider as a big priority +This issue has been fixed and released as part of release + +If any problems remain please reopen this JIRA. +The test was passing the wrong URL to that classloader +Punting issues from RC to final +Issue closed with release. +I have the bytecode transformation approach mentioned earlier mostly implemented +I couldnt reproduce test failure on yesterdays VMLooks like bug has disappeared at all +Assigning to me since was one I was looking at for a community partner +It is not currently used by much elseHowever it would be good if the element type is accessible via the Type system +values are written to and read from the databsae with the scale specified in Column +This issue looks different from the one in in this issue we never reach convergenceI lost the logs as I reran the server after implementing my fix +Producer credit based flow control is now implemented by leveraging the session replay functionality and blocking on a semaphore when there are more than x credits unconfirmed. +It does make the java client a little untangled +Now that work is well underway with AS all previous community releases are +I could have thrown together the same thing in minutes but thought it better to post tested codeIt would probably be a good idea to treat object arrays and primitive arrays separately with different iterators rather than using this iterator for both +Its an issue in +reopen to change CLOSED LATER status +Any chance for a test for this Bing +Attached the updated patchThe test failed in setup but I dont see any error info in logs +But I dont know if this is reasonable +Personally theyve had years to fix things and havent done it so I dont really think they need more time to fix it +Any luck reproducing +Hence this would not be fixed. +I probably see this as more generic keyvalue serialization knowing about the future direction with the schema +Id love to be able to add a Spring DM to my Grails application that would allow for the use of Spring services from bundles deployed in my dmServer context. +Is there any way to move that to configure or makeIn general I found that trying to cram other build systems into our autoconfautomake setup leads to trouble and people have better luck running antgemphpize directly +Are you able to load thisThanksBob +To make sure it applies the same way on the list +The will not be generated for SDO types that have originated from +Preliminary patch not intended for committing +Thanks Todd and ATM +Hii thereDoes this work in solr? I am using solr and want to apply document level security +Is that the case? Not doing checks by assuming that will hold is a slippery path IMOCurrently when calls then the scheduler can throw an exception about invalid allocations which get returned to the client +Incorporated Amars review comments +it would be interesting to see what the wfs cite tests do thoughPerhaps we could add a property to control this +Olingo Web Builder at Sourceforge is not related to Apache Olingo +Add the test to sync which +For example if key Hello already exists the name will be Hello and so on +I think disabling cookie parsing altogether makes more sense +Good idea +Castor source generator does not complain with the given Schema since no class is created for the simple type +i agree with igorwe shouldnt add the to the base but just change to do what that doeselse we clutter the apii think that is also what most people want? so any objections to that +since hbase never close mapfiles in normal usage timeout of should always be set +I did not test it on but the patch for got backported to +will mark as solved as I removed the Executors call as it was not needed but you need to optimize the test to not page +The fix is in and the trunk +I think this is going to have to be a wontfix since its terribly involved to get the batsh scripts working with a bunch of bare Ruby code +So maybe were opening ourselves up to concurrency issues +Committed to the branch +Patch to correct the problem +Error in configuration file path for resources doesnt exist in it +Andrew please see for faceting supportI suggest we first commit this basic field type for and then add faceting s the patch Added with hardcoded rates Added fieldType and field for mock provider in Added test case validating mock! Thanks for the cleanup work +Ok +Deployment in those situations should also be as easy as possible +Functionality has been added to Installer so this issue is redundant +Is it is standard to do package version in the package name for HadoopI prefer to have the code in to avoid package version in the package nameOtherwise on the patch +Then we have best of both worlds issue fixed but still backwards compatible opinions +Thanks. +I just committed this +Revisions and for branches vGACP and fix version to EAP since entity manager will not be upgraded for EAP +It should not fail again with this +Fixed in current SVN +The scary part is that its been around for some time and we both independently hit it today +In fact mail servers receiving mail seem to take the IP address directly from the socket and that was where the problem was +since you have vs bit env can you can compare java factory source with the env bat file that comes with vs and submit a patch +Not a problem with calling skip at all just trying to use A in scanuntil and ilk + +Hi TrustinI tested against same tested against and it worksCheersPaul +open httplocalhostmetamerfacescomponentsrichAutocomplete +I tried it by deploying a an artifact like which created the entry in localhostarchivarepositorysnapshotscommycompanyapp +It is worth a try anywayThanksMark +That will definitely help in the short term +Thanks KelvinPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +Backport to and +Groovy Shell command which appears to just open the console Other than that it looks good! Thanks for your persistence on this +this a metric nameYes thats a metric name + +Committed as r +Thanks for fixing this any idea when will be released +This is more or less documented in the developer guide +Hi JacquesHere I mean without applying the Stephens patch things are already working fineIf payment is not selected then an error message is shown +Closed until more feedback +using default behavior +what clarification do you want +srcjavaorgapachecommonslangmutable +is missing +after fixing some typos +Some classes are using Windows line endings others are using Unix line endings +Closing +In Commons Math they added a that doesnt overflow +Sorry +Can you please analyze what classes are using this structure +I have this coded btw so will take care of that issue if you dont beat me to it +Moving to reschedule if that work is underway with AS all previous community releases are +Whenever you report a wsdljava issue please make sure you attach your wsdl and specify the options you used +We need the before copy of the blob +It also fixes the null checkThe next patch will fix the other two issues but I thought I might upload this one anyway for review +this does not happen any longer. +This introduces a boolean flag that prevents errorstacktrace output if error occurs repeatinglyFlag is set back once activity succeedsIn general i would like to have a more general solution to this but this requires more changes +Im wondering if its somehow to do with the slight mix of and plugins currently in there as a result of adding the new launcher for the linux builds? Not really worth looking into that though when the fix is so simple plus it should all just be updated to anyway including the win launcher +I will try to test it better later but it seems patch is missing class +Thanks Yongqiang. +Is this a strong enough reference +Andre +Please open new for any individual issues that you believe remain +A URL to access the default version of a service would look like this httphostnameappresource +How about thisApplication Changes RequiredApplications that call the setTime method using the TIME data type should use the TIMESTAMP data type instead +Fix for PERSIST cascade loads unilitialized elements at flush time when the deep level is or more +Thank you for the solution Andrew but how do I fix this in my system +I think we have all places +Updated in the project will be in the next merge +Please close if you are satisfied thanks +We can add that later if we really need this jar +Will be done in the newer fop block see +A small value could have more burden on that are already busy +Thanks Enis for bringing this up +Early I have realized my own key binding but now I have found that we use Ctrl via extention point in changed the functional and I think now it works as you expected +GIT formatted patch +John Im looking at the patch now +Hum +Based on Gunnars comments Ive made the following changes +Here is the patch for the issue kindly have a look on itthanks +Its from an entirely separate module and I dont see why one should correlate to the other +I will take the action of converting your colorful image into a more version and attach it to this issue for review +The problem occurs since I switched from MRI to +Thanks +Committed revision +Again having an file still causes this issue which makes the Any annotation very difficult to use +Ill get to it Monday if no ones touched this by then +Fixed it on the JSF trunk as wellThanks for the caused fix was not valid. +Im using and encountered the same problem while trying out +jsf mapping should be removed from the think the testcase is in proper working order now although the surefire test fails because the Resource injection does not appear to be happening. +This means that TEST have to be excluded for derby also in java source and the pattern in pom should include all tests +An alternative to passing directories might be to pass a list of files +Yes seems the is null expression is missing in the API +Closing a duplicate. +patch looks close for release +I will review this issue laterDoes it the only class that fail + +Whether the call really does the work need not be s the status on this one +I will create new task if required. +I checked what ES does with it and we always pull the core key here anyways +Just realize there was a bug in the previous patch introducting +As this isnt a pdfbox problem I close this issue +Resolving as incomplete +Is there a way to control this flag from Solr? Would have expected I could add something in to control this flagAs I typed this I realized this might not be the place so is there a way to control this from command line in +I agree completely we should definitely proceed with Mikes improvements +Which shows that unit testing this feature is indeed important +I have attached a patch for the +And now it works +Carry on. +Even the last is bounded by the length that the user passed in +The patch I proposed and which has been moved to is only targeting client support so far +If we upgrade to or later I would need to this files +add new testcase to test the the patch based on the latest testcase for constructing response to create etc the test case as last comment test case to verify status of all containers in every much better checking it in +It takes a lot more time to reach the peak +Why should this move into the Struts project? Why isnt the projecta good home for this action? I admit that this seems pretty neat but Struts hasa large dependency list as it is +Good question Ill test it +search users in admin interfacesearch users in role members in admin interfacesearch in subtasks are resolved +Only needed by client side validation which will not make the train +Marco this means you can only apply this to jBPM master in this case and not backport this to the branch as that one is required to stay on Guvnor. +Applied +mrepositoryorghamcrestoptcceAUTOMATIONeAMS +Label could be Note or something like that +It might be an issue of my localjenkins maven repository +Hi MichaelI spent last couple of days on this issue and came up with a solution as described in in my previous commentThat is to retrieve the FK name from DB while deleting the FK constraintEven though this solution works fine it could be costly as it involves going to DB to fetch the FK nameSo I am still researching on a less expensive solutionCan you commit my attached changes on this JIRA as it fixed the testcase and I will open a new JIRA to handle the actual foreign key issueRegardsRavi +would it be possible to release with the fix and push it to maven repos +Added you to the cloudstack auth group should be good to go +on patchThanks for summary of changes aboveIts missing doc of this new way of doing things but can do in different issue +At a certain point the servlet engine needs to be notified of a series of changes through setAttribute +If I understand it I will have to update my version of Jquery? Is it +files and if that passes for me well call it a day +Resolved defect by applying patch to trunk and branch +Hello Houx have you had a chance to study this problem some more +Fixed at SVN revision . +There are a few more cases that we would like to have handled +Im afraid I do not have a Windows machine to test this on +New patch looks good thanks Uwe +Brett we saw the change in the link eventually so we also tried to request a new password this would be the solution but this triggers a mail action WHICH FAILS now we are really stuck +Please check that the patch was applied as you expected +Owen Hudson will not be able to run the mapred patch till the changes to common patch are committed along with the common jar committed to mapreduce +I you can provide a patch to update HDFS and link it to Ill give it a look +You can try archiveClasses option that may chance this bugfeature could be officialy fixed? As it is very interesting and that patchs are provided it should be quite easy to release it shouldnt it +It seems that hangs in a similar way +Other developments that help with this issue are that you can use session based materialized views and possibly the somewhat related global temporary table functionality +for trunk and branch +HiTo help integrate the changes more quickly I suggest to split the big patches into multiple smaller ones +On Linux it wont build +will contain a fix to the HTML text extractor +Grab is a annotation and we currently cannot annotate what cannot be annotated in Java +OlivierYes this is something that I could not figure out how to solve +ch error +The first attachment is the Scheduler Class of kahadb +It lacks some tests and code polish however the logic is there +it lists the web project instead so I click browse button to change it to the ejb project but after clicking browse it only lists the web project +Not a final patch +Now in the groovy editor will initiate an extract local variable refactoring. + setup script with configuration location +passesReady for review +Simple Forrest site that contains the pipelines and locationmap entries necessary to serve XHTML content from the Lenya default publication +However the smaller project still gets an error +I am working on this and will probably submit a patch today +It happened a while ago for us +Please explain what you mean +Here is a list of other files that are in the build that we need to consider +I can try +Thanks Rodrigo! +Ive committed a first implementation in revision only part missing is the keySet implementation with revision which delegates all map method calls to an embedded map on make this map is simply copied into an unmodifiable mapNot sure if this implementation is the best approach as it makes sub classing and lazy fetching of resource metadata way think we are hit by the fact that we made a poor choice when doing extends How about renaming make to lock? +sample used for reproducing the issue +The threads should be accessing different znode path to write splits to all the time +How does shell perform w the patch in place? I do not see a note in javadoc in the patch that explains how this patch now makes us behave +The idea of Agnes sounds goodBut please post a patch If you want to discuss further +Thanks! +I changed some things like your callable definition and it worked for me in if I compile using booc and then run the exe I get an error that the types cannot be loaded because has no RVA +This would help for fresh installations +This applies without conflicts +It would be good to confirm and are covered with an http subprotocol connection the subtasks on this issue to remove them should be closed invalid +You are right I am using +I have done some changes in existing code and tested on Example component mentioned in It is working finePlease have look on for the further changes made in this review +Evan A bit late but as youve noticed already yes those lines should go into the cluster properties +foooolks can we cleanup all such issues +The data is proprietary and huge so no real way to give you access to it directlyAlso the bad test case actually used the best indices +However failed + +the patch with minor changes to branch as well as to the trunkThanks for applying the patch +Could you rebase +DND pops up a dialog if action takes a long fix version +If you want to contribute a patch that improves that platform I dont think anyones against it but a simple config switch is certainly easier + +There is no way you would want to use the generated UI to establish such a relationship anyway +or does Jetty export some cool stuff via JMX that we might want to point people at for debuggingIf we go with JSP we should just remember not to use any specific features yet +I think its more intuitive for users to be able to use the normal CLASSPATH env varIts also more error prone +Can you think of a way that we can run test queries against an S filesystem? Maybe we should think about adding a? Does anything like that already exist at Amazon +Informing others wont hurt +The problem is Id still like to have it for a simple delete of a node +Patch coming that includes a testcaseThe patch also includes fixing some missing Overrides and unnecessary +I am just trying to understand what it automatically verifiable as disjoint +So all the variables which are modified during that is threadsafe +You simply pack thenecessary together +Hi JamesCould you update the deploy script to call the maven deploy with deploy profileThanksWillem +Applied with slight changes in rev + to the branch thanks Chao! +if an event says that a node is added the index assumes that the node does not exist in the indexI see two ways to fix this issue The query handler does not automatically the workspace but rather the to get a valid index +Ill look to commit the patch by end of Friday CET unless there are further comments +Actually I dont get it +Then you have to add a wait in your mailet initializationAdding a wait to fetchmail is not a solution +It hangs up quote Start processGC use large pagesWARNING Cannot cleanup application +Forgot to grant think the point is running tests in memory and not against a machine +A configuration change to from its default of to something much larger has altered overall system behavior on my cluster sufficiently to stop the apparent split suppression by scanning mapreduce jobs that I was seeing +One option may be to introduce a new method with your suggestion +Is your Perforce repository located on your machine or on another machineCan you issue other scm commands? Like this one mvn scm is like any other SCM where you and have working copy of the project locally on your development machine +Thanks to all the reviewers too +But java cant be used to run Sonar analysis whatever is the language of the project. +Under certain conditions this error will occur +I think automatic has a lot of allure thoughAutomatic is great + actually that I fixed in +Alan how do I validate the patch +Patch appliedThanks! +The list seems like its going to be very easy to accidentally change as youre tabbing through the page +Im going to assume that this has been resolved +See this long and unpleasant story for details +I think the OBR spec definitely intends to support other resource types but so far has only demonstrated this fact by calling the base unit of description a resource +looks good to me +Guess they dont have the query rewrite issue Derby doesI also attempted to fix the problem some time ago +on the implementation too +Hi Keith can you review these changes please? Thanks Robbie +Would you please search the jira first and check whether this is a duplicate of another issue on this same problem +There we can take as many stabs at it as we want tocheers Roland +I understand you added some print statements which would skew the line numbers +The patch you attached is about useless as I have to merge all changes by hand which is wast of time +Same patch modulo the commented line in in so jenkins will run +com and the under your codehaus account. +Since its only one week from Ill probably wait until early june before release but I can be convinced otherwise if there are blocking problems. +This should be sufficient for the use case described and it doesnt break the encapsulation of the class +Abriged means core subset +Have not been able to reproduce this one created a unit test but no luck +Well get that cleaned up +patch applied thanks Greg! +Added pull request for branch +Ive reviewed the reproducer in and noticed that Ive hard coded value in script on which ip should server be killed +Patch to fix accidental deletion from may have overestimated my ability to write even an test on this oneDo you think instead of trying to process all those PROPOSE inside the sync loop we should just have the follower take a snapshot immediately after receiving the snapshot in sync? That would probably be much easier than trying to do an log in sync and should still provide the same guarantee at the cost of doing a second snapshot soon after the first +Im adding Owen to the watchers list for he had a very strong opinion of why suites have to be used instead of say framework annotationsThis whole minutes test list has exposed the problem with suites vs +Rick thanks for the patchI messed up the commit comments the changesets are r and r. + +Snapshot is deployed feel free to give it a try. +Running regressions +Has been fixed a while back. +Gord is working on fixing it and its nearly ready to rock again but it isnt working right nowKeep an eye out for it to be supported in a release of CLI well announce it in the blog post +After some discussions we realized that the patch is too big to handle +This one tests the existence of partitions in a simpler way +The patch no longer applies to trunk +Im willing also to contribute to the project as need be +Extend to work on multiple join keysLet me know your thoughts and what are your thoughts on how the user can specify which is a fact table and which is a dimension table? Or are you using storage based statistics to infer that information +I also created a branch of the build module which contains the masterRight about the JAR +commited to branch for +Im still not sure what you mean by this ive tried to reproduce with No controller action or view A controller but no action A controller and action but no viewAnd in all cases i get a was fixed in I think but the issue left open +If you can come up with a better solution please do +Commit +Many thanks for your explanation i didnt recognized the asyncthing. +Validating a project CDI validator will ignore resources from other projects then each resource will be validated exactly once +actually this is because the same seed data is loaded into each tenant databaseas I commented on the dev ml there should be separate seed data for tenants from the base seed data that gets loaded when ofbiz is first createdI did this by adding a tenant reader and separating base seed data from tenant seed data + Author Group Red Hat or Red Hat Inc +I was able to capture log messages of this in actionCreating instance of bean Finished creating instance of bean Creating instance of bean Creating instance of bean Creating instance of bean myBeanFinished creating instance of bean myBeanFinished creating instance of bean Finished creating instance of bean Creating instance of bean myBeanCreating implicit proxy for bean myBean with common interceptors and specific interceptorsFinished creating instance of bean myBeanIn this case myBean doesnt get proxied since the two advisors arent finished being created yet +Yes I take your point to stick to the specific topic +committing in a few then Ill get to using it on my own projects as well +This call gets the hints from the bookies +confirmed they work in it be this little oops in? +This is still WIP in progress +I saw appropriate error messages line numbers etc +Thanks for your Help! I tested the changes and it works as expected +I am using the jnlp goal and version with the mojoalso look at the examples undersrctestprojectsand related BTW I am running the build using mvn clean package not +Thanks Paul! Ive just committed revision +ifn you dont mind + +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks a lot Matt for taking a look! +I hope he will answer me this weekSimon Ill check maven tooBTW I didnt find the link for assigning this item to me +Ive gone ahead and made the change in almost all the files +No other comments so Ill close this as wontfix +Please cross check +No remote gets or cache loader loads are performed. +can we get this into? This is a super benign patch and people are hurting without it +If anybody has any clues please let me know +Ram I was also wondering and was looking more into Trunk code alsoYuan can you pls test with some test cases +somehow that got lost when moving the code out of +I applied the patch on and get a Could not find artifactpom Its referenced by and I couldnt find in mvn central +So there may be some differences in +Changed the class name for the udf in other tests +I am not sure is it done or not +Thanks also uma and brendon for reviews +for null the most general class should be selected and that is Object not String +Draft user guide +I agree with your analysis Konstantin test works as expected however OOME thrown in threadB isnt getting caught in threadA starting the formerI have removed all AOP code from the test and manually added OOME throwing to run method and see exactly the same behaviorThe only modification I can think of in this case is to throw OOME before new thread is spawned but this defeats the purpose of the test completely +Hi Suresh To clarify this needs to go into trunk +Done. +Committed revision Committed revision +I commit the changes to trunk at rev +Lucas is a pretty old version +remoting final release and updated within repository + for the patch +With number of network elements implemented it does add little delay but not worth fixing at this stage +I created this JIRA file mainly because I wanted to have this issue related also to JBDS +cool I will work out a way to ship izpackexe and izpackapp together with standalone compiler to codehaus snapshot repo then you can work out from there +passes my tests fixes and presents a better API +The weblogs are fetched to my client +The protocol may have truncated the content in which case perhaps we need an metadata property or something but we should not be overwriting the HTTP header nor should we trust that it reflects the length of the data actually fetched +Ran ant test and passed +We found the root cause of this issue +Thank you for reporting thatthe build was indeed out of sync with the fact that the jar was renamed +At this point I think it has to do with the memory management settings that Samsung is shipping in their latest S line which seems to be causing Android to get really aggressive with reclaiming memory from apps that are not in the foreground +Cancelling. +Also noticed in your pom that you reference ispn ver +I will try to work with these suggestions and see what comes of it +The issue Vinay has found is in that code +I just deployed build which should at least fix that problemI have not been able to reproduce the negative numbers +this test will pass if you enable Nagles Algorithm +Generic type +x +Hi ChrisI developed a sample service called Version to reproduce this issue +does this issue affects the too +Can you consult with the Red Hat expert to see whats going on +Is this issue something we should consider for the C client as well or is it java specific +for the committed the patch +Since people keep coming back to this issue for what is actually Client JNDI support Ive raised +However this leads to a larger issue after talking with chris i think the output format should move back to look more like the old delegatE api +it will take until mid next week the meantime CVS access by sourceforge should have beenfixed by now +HiOk +the fix for the issue is to have the service state in the list api then UI can pick up the param value and show itWill work on it and send for it as critical as currently the code displays the state of the VM and NOT the application which is being LBed +I am setting this issue to resolved now since ariel does not answer but I am not quite sure if the missing content length was the entire problem Sylvain upgraded to recently without upfront notification and I had issues with ie under a certain condition with the dojo releaseBut since ariels problem was not reproducable here I am leaving this as is because my dojo is very seldom and no other case of this issue has been reported on the dojo mailing list so far! +Also we need to decode if we include it in module or if we create another build module +Fixed in svn revision +Fixed in rev by immediately returning a after shutting down the server +This is hard to test for explicitly as GC is and covering all the cases is difficult +Issue resolved as of r +OlivierIm using a version of that last got a blue light + +Hi you would think if you follow +Alexander did you further pursue this issueSorry for being so quiet on this one the last couple of months have been somewhat stressful for me day so I didnt find any time to work on this issue +That said you won me over Im fine with switching it back +Theres a flushed per DW and a global one in IW +But I dont think this is something critical which should be fixed in. +So it would be environment variables would contain everything that is required in theAnyway just fyi +Great stuff Willem we should really make sure we get this documented as part of a Camel users guide +Good point Igor thanks +Anyway your patch makes it much better and actually ensures correct behaviorLets commit unless therere objections +Please note that I replaced the files in the lib directory with files for shrinking purposes +still getting errors in the ant built when I run the zip target +I already added a rolling to the log in my as well to verify that old logs are not replicated +See comment in patch +Thus moving this issue further to as blocker since this is important to get out tried and test +My first preference would be throwing of due to the above assumption +Whatever +Woops youre right new patch attached +In the past Ive removed the node but other users complained that after removing one jar they wanted to create another +Option is less restrictive than option +I would commit this +How do you intend to enforce that an AM is dead +i have attached a wsdl Any idea +I do not know what the real values should be +Changes to the config file need to be atomic for this approach to work correctlyI agree +Might be a conflict in the location where we are creating the SNAPSHOTS +This problem resulted from a and that were both using the same instance +Havent be ready for review as test failures in. +Added Patch forPatch applied +The test failures were due to a kernel left behind by test failures in the system moduleThis issue can be closed now +Committed to trunk and. +Leads to a regression in task tracker this patch isnt the root cause as the exception was already being wrapped +Can this issue be related to? Adding a link just in case +Problem seems to only occur with circular like in your attached this issue since Lari provided a workaround and no further feedback provided by reporter +Minor tweaks to the build scripts to return to calling +it was just the update of CHANGESI am not sure yet that its properly squashed but if you can try that would be great thanks +If there will be a need in the hierarchy above then it can be easily created +This is not necessarily wrong just an unrelated aspect requiring its own discussionI am not applying the patch yet as well need to ensure that all our client test cases are switched to the new format and all the tests pass +We believe that the customer had no problems after getting the patched sigar libraries. +iOS scrolls it with one finger +The current version of the PR only includes support for products not products +I dont see this behavior for ages now +You may want to look at the target in +But what to do about this up to the JCA team so Im passing it to you +The reason for running maps is for having sufficient history size so that recovery can be tested +A complete stack trace would be required to shed more light on thatThe most likely reason is that the problem is in the Map contents +Thanks Amareshwari and Enis! +Is this still a problem in or trunk +If any comment please let me know +Patch for s not clear from the commit record how this patch is in and not but it is and thats fine so updating Fixed Version to include +Then consider contributing a test case as asked for or trying out my suggested fix +We can get properties from propertied file +studioeclipseplugins +removes the validation call and updates the example to set the thrift contact point on the correct configuration v and merged everything to trunk. +This was fixed for final and was closedWhat is the version of Axis C you are using +If we change set then we might as well change get at the same time +I guess you can close this issue +In the browser there were a couple of problems I hit a immediately +Rob something we can fix or WTP API deficiency +Looks great thanks +I deleted the first patch since the second is better +Ill give it a try and see what I can come up withHowever I think the discussion which was initiated by Pete is very important +Committed branch and trunk. +If its policy to separate them it would be nice if it warned at least +I am working in r and I get message The Build Environment cant be removed it is probably used by a build definition +I just committed this +Please back the old system jar if you want to rollback with the change that Claus just showed +As I said above if you dont edit examplePortlet before then clicking the already accessible link gives an error +Unassigning +removepeer +Done not retested as Freddy is not available. +Rollback done revision +Thanks for the contribution +The path is defaulted to so if that does not resolve for those vhosts they will need to be changed manually +Since does all the necessary work for identifying all found problems it shouldnt be too hard to implement this featureIn addition I would like to see it implemented in its own mojo as it is also useful in the default build lifecycle and not only in the site lifecycle +Applied to the trunk as rThanks for the patch +Dont add them to the default messaging RajeshI just ran the tests and it all passed +No tests are added because this only separate datatypes and existing tests cover the functionality +SVN revision + fixes both +All DoneFrom now on then just commit to your release dir. +fixed +I am able to use both fine with Facelets +Need to determine if this remains an issue post the refactoring of concurrent session support +Please reopen if you still experience this issue +yes this is duplicate of +One workaround might be to use the overload and specify priority +This is probably upstream so Ill assign it to myself for now so I remember to create an Android bugpatch for it +Yeah that was a change in but the error would be to figure out why JS is passing for the file length +Thanks David! Awesome work +I fixed the Ant task to check for OPTIONAL references +Example implementation of IPA you have a very small ipa file we could use for a unit test for this +Added Code to put Oracle Queues in JNDI +Reopen as the same fix must be apply for others scripts debug run in testScript +A few more comments +The plugin classpath is a requirement of the PMD configurator as well +giri I think what ben meant was run jdiff on svn trunk and then run jdiff on svn trunkpatch +I would like to suggest to keep the current menu structure however to add a field organization which will be shown with the relevant menu items like invoices and payments +Ive verified mvn clean install +Committed r trunk r close for +For some unknown reason the build in one version of the OS generates bytecodes at level and in the other at levelThe sources were confirmed to be the same but with the different bytecode versions a comparison is not possibleHere are the versions of Java used on and on The difference could be as well due to a different default in the maven compiler plugin but the same one was used +happens if user continues using directly +uses the version of the indexing server +Please confirm whether it still is an issue with +ve retired the plugin +This issue has been fixed and released as part of release +Can you hold off committing this +The workaround to this issue requires to be Issue Test description with standard unix socket behaviour +Ive applied modified patch to avoid loosing events that must be processed +No need to clutter it up more and we should just do the cleanup automatically +duplicates issue came first so arguably duplicates this +Given this level of abstraction I wonder if Jakes original suggestion about replacing the enum with a classname makes more sense +Verified +for the stack trace and files +Thank you +bq +I ran your script back to build and every other test has a random hang usually different each time +The PDF issue is resolved +Will try to track this down +Guess it boiled down to RTFM. +Regenerate patch for trunk the nd patch applied cleanly for me +Attaching and the resulting HTML file with a proposed correction to the topic +Ok +Committed revision Added minDepth and maxDepth if you work on the ticket then assign the ticket in JIRA to your name +There appears to be some inconsistent behaviour with the field +This says that an application that worked with JSF plus facelets is supposed to work with JSF without facelets unchangedIMHO setting the context param should be the only way to disable the integrated facelets feature +This is a problem and my vote goes for method +Jesse this now has the patch from +It will if the imported project has the very same name in the pom as the project in your local workspace +So my temporary solution to this problem is to add code to deal with the exception and return null if it happened in order to trigger the hdfs checksum mechanismIt may take too much time to find the real cause of this problem in hbase checksum mechanism +Switch to has been done in the trunkTo close this ticket there are two tasks left update the doc add the new rules added between and should be rules added. +application works fine on AS . +This issue was meant to make MTQ single pass so lets do that first +OK closing the issue then +Closing based on lack of response. +Reduced priority of non critical issues which have current issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Replace and if they already existed +Please let me know if there are any other issues +committed to branch and on +not enough info to set a good component setting to misc +Thanks for the patch Jack +I havent removed the old code yet. +I dont see any reason why wed want to deny the DBOs granting the execute permission on this procedure so to keeping it in SYSCSUTIL +I think using both makes sense +There are some changes in detection which havent been fully tested +Can some committer review this patch and apply if considered kosherBTW the patch also fixes +What is the alternative Whirr not supporting Puppet? If Puppetlabs doesnt want anyone using the gem right now what are our options +Linked +method is no longer present in the signature of am blaming lack of sleep +Sorry what I meant to say is that the conditions for these to be called only occur in tests +Is this the correct way to locate the relevant logging code +Robbie can you review this one please +Boundary comments are still enabled with widgetVerbose set true in the parameters MapJacquesThis patch includes the Example component context parameter in. +The user of is blocked from using this plugin since at repo is picked up first +bq +Without it this work cannot support major use cases +Thanks Kristian for great work! When does the invalidation request be performed now? Does it control in? I run the with the following comment in windows java client true true true dateboth true true both XXM XmsM cderbyoldreleasejars tee As expected some tests fail because of standard out is different than canon +The bulk of this support has been fleshed out +Trivial backport to complete and ready to go +i have deleted my whole repository to be sure i was in the same configuration as lambda have used the last snapshot of dashboard and no problem all works finei used under windows and youdo you always have the same problemso no can I close this issue + +Any issues from findbugs we really ought to fix should be filed as bugs and patchedreviewedtested as normal +Thanks +If you want to see those comments you can still look trough Change History tab +These will become cluster level params and will be removed from add and update clusters +Aim at this for +Now that we always run in multicore mode its just more prominent +Should I go ahead and collapse those classes down as well +Sanjay feel free put back to blocker if you disagree +Dont know if this is possible or has a precedent in Apache +Reopening to reflect port to +I still dont understand why in order to transform a Java object to a SOAP infoset we need a content type and a builder +From the patch Sets the ENABLED state in the cache and Creates or force updates an node to the ENABLED state for the specified tableId modify the above to be Sets the ENABLED state in the cache and creates or force updates a node to ENABLED state for the specified table +The parser is just loopingaround looking for the next component and keeps comming up with the XXXX and ignoring it rather than erroringThe old parser was correctly able to ignore this statementI could work around this if I new how to disable JPA validation or how to uninstall the JPA validation components +Please do I think you can set them to final and remove setters +That is an interesting one +this works fine in AS but not when I switch to AS +Im just trying to make the system more generally reliable in the presence of extremely unreliable clientsOkay just wanted to make sure were on the same page before you started writing a patch +So if you have a chance to test it thatd be tried for half an hour to replicate the but was you also upgrade to? It looks like youre using beta or something? Theres an off chance it may have just been fixed +Using to instead since this still sufficiently reduces the probability of a collision +No performance gains so no point to use this approach. +Reopening to commit the issue +Thanks again Dag +Please take a look at the included sample output and let me know if there are any other changes that need to be madeThanks +Weve fixed which seemed to be the root cause of this issue here +Couple minor comments on Phabricator looks good +Attachment has been added with description The has been added with description The terrible has been added with description XSL Why not close this bug as DONTFIX +Then we can have a separate interop target which actually runs the tests against the real endpoints but isnt required for the main build +As the number of dependencies can quickly explode once you take into account different versions of a dependency which is next to impossible to test we are a bit restrictive in Camel and make the assumption that future minor versions of a project will stay backwards compatible but we dont necessarily assume that major versions will +just relies on zookeeper instead of the master +sighWill look at this to see why theres a difference between clusterdir and clusterdir +I also posted a reply for as well +Fix in all old Resolved issues +Reproduced in attached +This is done +Headless parsing essentially requires a lower level configuration change plus a few modifications of the higher level parsers +In when it catches an exception it wraps it in a and then throws it this patch will instead output a small log message +Good point +Im not really convinced by a Clear console button +What is the desired target release for this? The code change is quite simple to allow null values for columns unless we want to preserve the exceptional behavior in which case a new TEXTTABLE switch is needed +I believe we can close this issue now +Can you please recreate the patch and resubmit? I tested by manually applying the changes on my test machine so missed this +Can you replace them with spaces please +dont forget to remove it later +The actual exception should be as per the JCA specificationAPI +Rafalcan you please try attached can verify that the latest patch from Anurag fixes the issue as long as you apply both patches together +Good stuff +And upon receipt of the message get the reply to property and send to thatLet me know if that works for you +Attached a patch that fixes this by copying the signature to the new +Although it adds up a different way of wotking with mocks which might be confusing somewhatHowever I still can see failing test if a user umask settings are different from expected +Will add the stuff to as well +Closing this as this has been fixed +However since this effectively boils down to a matter of taste I am fine with doing it your way +With this patch I was able to run in a batch of runs with no failures +released! +While I think its useful I do agree that its a rather niche caseWere holding out on embedding Cassandra mainly because we dont want our code to bring down Cassandra and its a good feature to have the isolated processes for thisI guess the other question is why have the fat client +The patch has improved but there are still some thing Id like to be solved differently +We actually need the fix only for trunk and above where will no longer be supported +Fix has been submitted for sometime. +bulk move of all unresolved issues from to + +The patch has been applied +In release I have seen the long standing map jobs getting killed consistently when they exceed the time. +changed report code to ensure release slot happens always by calling releaseSlot in finally block +may it be related to the keyword languangedictionary work +fixed in trunk not sure if we really should enable this in since it is not a showstopper fixed in shipped in +Gary glad to hear you like +I would check trunk if I were ever able to compile it +reopening needs to be merged from trunk to from trunk +javatestingorgapachederbyTestingfunctionTestsmasterjdk and +Its not but the Specification dependencies or provides notation +It looks like this may already be taking place in the latestversion of the The DTD registration should beApache Software FoundationDTD Commons Validator Rules Configuration ENorgapachecommonsresourcesSince James had talked about using a common DTD between struts and +the patch looks good to mebq +I looked into this issue and udig is updating the georesource info +Thanks agreed that thats a better title +Thanks. +Here is a for the patch +Sending srcmainjavaorgapachestrutscomponentsSending srcmainjavaorgapachestrutscomponentsAdding srcmainjavaorgapachestrutsviewsfreemarkertagsAdding srcmainjavaorgapachestrutsviewsfreemarkertagsAdding srcmainjavaorgapachestrutsviewsfreemarkertagsSending srcmainjavaorgapachestrutsviewsfreemarkertagsTransmitting file dataCommitted revision +this patch solve this problem by implementing a button on the gallery panelwhere on the st click it changes the slides for every secondson the nd click it changes the slides for every secondsand when the user clicks on it the rd time it will stop the slide show +Here is the simple program I used to recreate this is the database that cant be opened + +New patch addressing the test timeout +This step happens after building the distributionsIn addition the standalone write target is now more capable +Unassigning myself from this issue for now not working actively on it now +How about introducing something likeI attached just an idea and draft level patch +Pull all operations for JAXB model into jaxb model plugin +This looks like a dup of +thanks for the heads up +WIP patch +Feel free to reopen the issue when providing more information +I think this should do it +Why login is a method of but logout is a method of? Does it more natural to put both of them inI think it is possible to put logout in +As discussed earlier in this jira its not possible to mix strict and allocations at the same priority +srctestjava could be a good place to start if you want to try to and help by creating a real test case +Closing stale resolved issues +This issue is closed as wont fix because no advance can be done from this point +Primary use caseWith explicit packaging export HADOOPHOME mvn clean package Pdist run HDFS run mapreduce Test thingsWith implicit packaging export HADOOPHOME mvn clean package run HDFS run map reduce Test thingsBuild time are the same for both cases because the directory layout is required to run HDFSMR for testingSecondary use case for projects that depends on HadoopWith explicit packaging mvn clean install PdistWith implicit packaging mvn clean installBuild time is seconds slower for implicit file packaging on a Mac laptopThe secondary use case is less frequent than the primary use case +tag files +Right +cylce was removed from activiti project +Thanks for the quick responseWe can use the latest from the trunk againRegards reverted the custom socket factory support in revision based on concerns explained aboveLets look for an alternative solution to be included in a later Jackrabbit release. +I think the problem is you dont have Freemarker configured for the right encoding correctly +Will do that in a follow up jira +Im still learning my way around the code base so any pointers on which direction to look would be great +Latest patch that converts all streams except the single onesIll commit this soon +Works OK after nukesusers table migration from to +I am reopening this ticket for tracking the effort to make the list of envs to be configurable +We will need to create a sort of search actor where we can register to listen for tokens +to trunk +Weve talked offline about the generic YARN history server and a payload being sent from the to the RM when they unregister but I dont believe thats happening in a hurryThe multiple job case does not exist at the moment but may in the future +Thanks for the contribution +Should work now +Backported the latest tests from +Any comments on the patch at all +Hi Steve Why is not it a good ideaCan you add an option can be activated or not? +I am going to see if I see a memory leak using Eclipse I am using with little or no trouble +no menus no banners just plain html +Integrated to snapshots branchThanks for the review Jon +Verified at r. +There is a caveat to it however + +WelcomeI also assigned the issue to him. +patch looks good +Committed revision Ive logged for the IN subquery anomaly +Unless im missing something its absolutely critical that we include these licenses +If thats no good I just want to put them in a reasonable and visible place so that they arent hidden from folks +commit it +I think we should try to hook this into such that we can still have many of its options available just done at bigger scale +Both the issues and shows status as resolvedWhat about fix? have you opened a separate ticket for that? +Nodetool requires max MB memory to run and it shoud be limited to that value no matter of default limits +Hi FrankThanks for reviewing and patching my patch +Please respect my opinionI do respect your opinion in the large sense of the phrase but I dont agree with you about not adding the k jar +Year is not neededSure I have removed the three instances I foundbq +If this is to be an improvement on performance then by definition it is not a bug but an enhancement +thanksAn updated version is posted +My first guess is that since is this is related to the fact that is not the concrete implementation but an abstract class it is possible that registration in uses incorrect registration mechanismBut it is odd that unit tests do not catch this as I could have sworn I added tests not too long ago +I marked this issue with labels mentor and gsoc +The fix looks fine. +ThanksSandhya Turaga +This issue has nothing to do with JAAS +Here is a sample for this behaviorThe is and the bean Please have a look at this +So you can compareIMO does pipeliningKlaus +Fix verified +What about existing classes like? as far as i know there is no JCR constant in there +Because if there is some problem with the State transitions of the regions then we need to first clear it out and then enabledisable the tableI remember something related to this was done as part of HBCK +Patch looks good +No error messges are given now but in my opinion when you go back you should not be able to finish the wizard unless you are in the last page of the wizardThis issue can be closed now +I have committed the fix for the of Core to CVS and Im waiting for some time to perform a full build so I can test if it performs the way thats expected +I guess I was looking for similar functionality to when I would override a method that Roo manages in an aspect but the files in question for removing a static view controller arent like ITDs theyre actual project files so it could end up being destructive if Stefan +So we create ejb web test projects with the same compilation level and java facet now. + for the patch +For submitted for with this fix its very compressible data +however is a container for a Matrix which is Writable +If you are creating a new class that needs to implement this interface dont forget the default constructor +on my wishlist targeting this for move s +This test not part of +Thanks Jitendra. +I figured out that this will only work with newest Mongo and PHP Driver +Some applications would prefer having hyphenated tokens together +we have to create a branch for JSF or decide to make the current trunk before committing this patch! feel free to start a thread about this on the dev list Martin +This patch should only be used on a Portlet Bridge +Perhaps we should be relying on integrating with Seam Valildation Bean Validation for this +Isnt it same with +Im glad it has worked out well finally +Most of the demand is for Simplified Chinese and Japanese +We need to be very careful to ensure this doesnt propagate elsewhere +Also fixed similar potential problems hopefully there are no more +unfortunately we cant sort entries in the classpath container because that is the actual order of those entries in the classpath so it is critical and cant be changed +Patch for +After all scrapbook pages are so dynamic and I cant exactly expect you to be able to that an arbitrary snippet of code is Groovy or Java +I will respond on RBI will backport this patch to and after it is applied to trunk +No problem. +Once done these should move to the website +Since comments are already covered in attached patch making to patch available status have just committed this to trunk and +The patch is fine +Im still seeing some issues with this on +ughUpdate Im not able to find it +IE FFt reproduce bug under IE of a bulk update all resolutions changed to done please review history to original resolution type +This allowed to resolve the images via cocoon sitemap and drop the workaround that relied on the file system. +VishalFlavio can you guys please comment on the jira when you add a review board request and when you review it? Its easier to track on a jira than the mailing list +It got introduced by +The ones I uncovered were The file srctestcomclouderasqooputil is missing package declaration which causes compilation failure The ant docs target is failing due to some block placement problem +Thanks EricFixed in +Changed pom file to use the existing test in change looks good to me +Committed at subversion revision Touches the following filesM M +Stack Ryan Its a bit early +and works fine with Chrome. +Thanks +Cant login to portal when click username on should check this problem and fix them after have merged navigation controller branch to trunk +Connection forcing Force the session to obtain connection from datasource before it is returned to the user code in this way we know which connection is used for the main transaction +Let me know if there is more information needed +Note the updated patch fixes a bug in the original patch which tested the stack meta info for the presence of HCFS instead of the cluster services +i removed the instanceof and getter from mockanalyzerthe bug is simply that the flexible queryparser didnt enable position increments by default +Are you proposing we issue a single compaction submission when streaming is done insteadYesPatch attached for branch +This issue is fixed with revision r +bq +but leave the the way it currently is +i have just signed the contributor from germanysebastian +After doing so they will have to edit some configuration files by hand +Sometimes I have up to seconds of startup time of looks like it depends on the load of the machine whereas when using the new scanning stuff startup is always under a secondIt would be great if we could add this to the codebaseAny objections anything I can do better to make this happen +Committed the patch to trunk andThank you Arpit +Moving to as tests still need to be added. +Um are you sure you tested it with the latest framework trunk? The fix for your second issue has been for framework not the +I dont know that corrupt sstable is really that much less scary than a stack trace +I just checked all the JMS examples from the EAP server config guide and they are working fine + +Would be very good to have it in order since we should be working with JAI with no additional downloads +We should only require target when the Xpath is ambiguous and this should never happen if there is a relationship between type names and element names +Modified text rawyes to rawtrue +Just a quick question isnt this controlled by the file with a delay of +I had the exact same behavior as Tim with fedora core am experiencing the same issues with Eclipse and Meven IDE on Mac OS X this is the only plugin since switching that no longer works +Or if there is a way tokeep that where its located now but get the path to the to resolve correctly in either case that would be ideal +Changing Issue Type to Improvement. +Thanks for looking into it though +Now that work is underway with AS all previous community releases are +Ill leave out that part on RHEL for now so that we can start the server without any problem will further investigate +Ive just committed this. +upgrade incorporates also some pom file fixes in jbpm dependencies +Yes this was done +Unassigning from myself since this is open to community +since the command line specifies svn execution mode +I didnt run into that problem and I thought that the connection pages are loaded in the same order as the plugins +I also added an additional test to check for pubsubnolocal +Incidentally calling length on the result of inspectbuffer will overestimate the size of the actual payload by something like xRajesh Ill commit your patch for fixing the inspectbuffer method +Too bad they didnt have that idea for Javacheers Roland +I have made the changes in createShipment method in +I am going to test validity of the report +Its OK +I think it should use the approved Hadoop elephant and with Powered by in textIts useful for a logo to be graphically distinct from the official logo so theyre not easily confused +Refactored the code to use a new interface + +Thanks Pawel! +committed +back port this issue fix +Tsk tsk +All looks great +Patch attached generated via svn rm the three package directories +Is there anybody opposed to commiting this +Not sure if the benefits are worth the trouble +bulk close of all resolved issues in preparation for release. +Patch that will do the if the hadoop append configuration is set to true in a place where hbase client can see it +I applied it thx! +Integration test added +Fixed as part of the overall client refresh in Tapestry. +Thanks for the link +can be removed +However I thought I remember reading this in the for and +Youre accessing the JSP and the new Web UI through the same port so I believe that this is a +property file changes committed +I just committed thisHi Konstantin the patch that I uploaded at pm on addressed all your review comments including last two that you commented after pm +Fixed the test failures +The slow startup time comes from having to inspect many classes at startup by reflection +Are there any ideas for improving this situation completely from the Spring Integration side +This means that multiple instances of the fetcher thread can be active at the same time thus inserting duplicate entries in the fetched list +One was applied to the trunk and is no longer necessary +This patch is now compatible with the current revision +That was my perspective particularly given the current SSL discussions +I changed a dependency in the projects pom and started an eclipse of the compile error maven does not update the classpath +Is this issues linked to +This sends false for all unchecked can you review this patchany other places in the UI where similar bugs might be lurking +Itsounded like everyone agreed with this change on the mailing list +So the counter is just a count of how may maps got picked up from the task cache +Patch attached addressing the real small task of making usage comments in the class consistent when a command is not properly used from the command lineIf no objections Id like to committ and clean up At revision . +HiI think there is a confusion between the and the version of the up your project and use the same version find attached a spring configuration file which show the object definition using new syntaxLine object definition using old syntaxWhen choosing line definition by commenting the other definition an error is thrown during when adding the line you suggested before reading the configuration file stopped this and now it worksHowever the online documentation forgets to specify the id attribute for the rll now close thisValentin +Oups fixed in commit dbbdeffafdaace. +Putting the version range resolver into a user plugin sounds great +This updated patch corrects this little issue and now passes +Not submitting patch until is in +Hi Subathra could you provide a bit more information about the issue? Any idea how many messages is in the queue? What are the conditions to reproduce itThanks + +This does two things Checks during writeTo in to ensure the underlying view has not changed +Are you sure there is no on the classpath? it +This looks to have fixed the issue in TRUNK going by fact that this test does not fail in TRUNK anymore +I dont know what causes this in your caseCan you please try this with Java? Are there any other Fedora uses who can reproduce this +I will try to put together a page proposal tomorrow which will be as much as possible +Moving znode from OPENING to OPENING or CLOSING to CLOSING ensuring the sequence numbers are as we expect is how we check we still have ownership of the region before we go about making alterations in the filesystem +What version of JSF do you use? Where do you see a problem? XHTMLJSP? Please descibe the issue step by step +Sounds like an interesting problem if you are willing a patch for the module maintainer to review would be very cool + +This was implemented +OK thanks Matt +Ill upload a partial patch that only includes enhancements to with better measurements right awayFor me using your first patch that only caches the resolver I get ms million entriessec +Will add the are planning to move to JBM but we still have some integration issues to solve +So the pojo that is used in the testcase is the older one and the patched file is the new one +Is there a way we can make the patch available through injection? At the same time isnt there a JSR or something coming up that deals with this issue +We need to make sure that all the javadocs are generated and collected together as part of the build + +wasnt it just recently that people were still running? +Ill check +Ill tidy up the links and add in the missing ones as well +Attached patche from Robbie Gemmell and me fixing the issue +With r the will not be deployed anymoreI didnt find a way to disable the deployment of +Thanks +X toolstestingderbyPerforming status on external item at toolstestingderby +And if it is noon on the th of the month does it round up or downOr does it do bankers rounding? +Not a bug with +May be there is some thing more complex I didnt see but at least all the unit tests went wellNothing complex hereRegards test on a real cluster not necessary on a patch this small +So maybe itll be the feature request +Added patch for requires +Bulk close Solr issues +I attach a patch to this bug but the patch should solve bugs and as well and provides correct escaping behavior for fields and terms that are not quoted +I have now merged this code from the code line to the codeline so bug is fixed in both versions. +Please find the same attached as +v +Assigning to Kelven per Daves comments +This issue was the very first issue for gwt which depends on course I had to convince the other members of the power of maven so I saw I nice challenge in this ve started on it and Im very its time for writing a lot of tests some code cleanups must be done but I had to share this little success +I just ran into an exception today because I tried this +New patch attached +Moved to Bugzilla as the documentation issue. +Write an equivalent of under standalone +I have committed thisSee Linked Issues for the followup issues. +Doesnt actually compile +Added patch for that shows the strange behavior +If it is however then likely well need to open this up to the Trinidad list to get some comments +Is there anything happening to this defect which is several months old by now? +State that precSteps is bits and trie is the entire data structure +Ill look into this +Heres fix +I will write it up +During all those patches its hard not to loose the overviewthanks for your workAre you trying to sneak in the alleged fix for as well +Targeting in updated books parts project sets with cleaned up. +The same could apply for the plain File component +Remove the next line about PATHThe rest of this page should be structured as follows set DERBYHOME include the content formerly from tgs mentioned by Kim +HmmmmSo the underlying assumption here is that the enum type is restricted to being single valued +Committed thanks Tom! +XavierIMHO Stefans right with his comments +When the reader reads data from a stored table we read a compression block at a time +Any other people here with an opinion +The code dealing with the header settings has been completely redone as far as I know +If logj is fixed in paxlogging can we just extend in a fragment to add our custom thread safe appender +Thanks +Milos this issue is NOT a duplicate of it is just another incarnation of the same problem +No you still need to check if nodeId exists before you create else will bite us and CS will have wrong resourcenode count +This is really something thats always useful to know afterwards even with snapshots not only for releases +Added testcase and patched affected readers individually item buffering implementation consolidation moved to closed as resolved and released +Same thing works from commandlineTried with bot hadoop and hadoop Also ran smoke test successfully with static dfs +Version Released +Thanks Josh. +Please tried to verify this fix but at Step I get the following error when accessing httplocalhostgrailsauthorcreate +Paste any relevant errorsIf you still see problems please package your sample code in a small project so that I can try to reproduce your exact scenario +This was alleviated when Velocity was upgraded to Velocity Engine after release. +John suggested two places to add synchronized blocks to resolve this issue these blocks have been committed to the repo +Correct code shouldnt rely on read readFully +What about removeStatement +The patches are unfortunately split because I do not have a single checkout of jbossas +Might be the same for see +I dont see any reason why we might wait instead of publishing with H as long as it is clear that it is not tested enough +I emailed the person who had this error log and Nick on the email +I recommend that this bug be closed +Hi KatheyI noticed that you assigned this bug to yourself a year ago +Attaching +The base set of tests appears to run successfully after the rename but I didnt find any other references to it and there doesnt seem to be any unit testing around this Strategy so Im not certain Ive done all that is needed +Or it should be +bug fixed in rev . +Have you looked at related which touch compaction code +Patch looks good +Your first idea is myfavoritet understand your last comment but at least for the moment you seemed to besatisfied +Please confirm or comment +HiI have a few questions regarding this issue and the implementation been updated after the initial oneI went through the code from the checked out version around r of OODT and it looks just as it is described by Davids comment +I would prefer to leave these changes in +Ive encountered this in a Rails environment as well +Im close but something is going wrong when calling in the background to do the RAW QCOW conversion +releases do not contain this fix +has been marked as a duplicate of this bug +I do a forward search for FAILED and from this position a backwards search for running to find the name of the actual test suiteAre you perfecty sure the failed tests were not run as part of the +some of which have changed in +can anyone think of a reason not to change the default? +Ive started working on a fix for this one will submit pull request shortly +The client will decide wether to send the message to the partial rcpt list or close the connectionJavamail for example has a parameter sendPartial that manage this scenario you can decide what to do when you find invalid recipients while sending a new messageIf we implement the standard behaviour then the client will have the power to decide what to do +patch applied +Youre still sometimes placing else and catch clauses on a new line rather than on the same line as the preceding close braceThis could all really stand to be replaced by a but thats for another day +Marked for content +The plan is to upgrade to as soon as it finishes getting releasesUpgrade to newly released +I created the attached +Ive attached which adds argument checking to isSet set and clear +revision +I committed this to trunk and +It is working now +With this patch Ive been able to successfully process GML schemas + Removes the content before cajoling so any subsequent failures never allows unsantized content to be sent to the browser Removes calls to console which does not exist in all browsersA subsequent patch will format cajoling errors to return to the user +These are not productized artifacts and never will be +This patch adds a command line option to generate JSONThe patch also improves the binary representation to include allowed values for enumerated types +Updated version of the patch removes need for isset in cases where there is no default value provided +OK all clear now +Verified in Tools +Silence +Might be because we use some logging bridges inbetween who knowsBut I profiled it and the difference was really huge +The generation of the Maven bundles is now part of which generates all of the release artifacts as well +on the issue and I cannot recommend to support more than in the ajax requests for now +On the IE Tools menu click Internet Options +Mamta A +There is only JS in file I link this file to the simple page to reproduce the problem +I agree with your above comments +If you upgraded from we didnt bother ripping them out +This version of the release note urges the user to enable user authentication and to customize the default policy file as described in the user guides + +Jenkins server has installed submit patch for build test +I think therere some unit tests missing here that Ill try to add while I fix thisRegarding the data structure let me get the functionality fixed first and then we can discuss implementation and data structure +agree with Laura on this one +But the only way I found to solve was to create a new project and overwrite Domain and Controller files adjusting the config am using but the upgrade target from to not solved that for me I needed to move the files manually and make the necessary adjustments +Something like that should do the job +Why should they bebq +Nobody is disagreeing that we are inconsistent but this has no dependency on EBWS +The issue is definitely associated with the jars used by and overriding them using have recently modified the ESB server to include those jars but there may be something else needing done for the can you add some more details of the errors you are seeing? Is it on the serverclient side? Can you show the outputSaying wont work is not very helpful to me +Uploaded for the following change +Closed without integration test +Thanks Bowen +Fixed on ER. +The translation of the associations seems OK to me +I see that adding purgetrue actually deletes the sld file today +But I still do disagree slfj itself is not unnecessary +So the reason i noticed this is because we bulk loads werent showing up in the stream stage mbean +Thanks Bryant Luk +I wonder if we should limit the number of bits for replicas k seems is overkill and this would leave space for an additional field +Closing issue out as part of JIRA cleanup +But you may want to verify it renders correctly before merging +But that would mean that we dont support pure Python so if you dont have the build environment for some reason you cant use Compact Protocol +That would not be an issue once the scripts are cached +Yeah if you can stop reading a sentence whenever you like you can probably make the spec bend to any interpretationThe full sentence isIf no source is found then return true if the registrant bundle is equal to the specified bundle otherwise return false +Navigate to the Available Service Group page in the admin console +My pom had xfire dependencies +Ths svn script I actually used modified from shawn I had to modify the svn command script extensively in general its better to do svn mv on your local machine unless you are movng or copying an entire module +OK its all right. +This is the root cause of what I am explaining above +There is definitely a bug in Axis code but as I can see there are couple of xsd in the archive file without having a targetNamespace so that might be one of the reason to have import with empty string namespace +Thanks Arun! +I gonna launch a Jenkins build +Oh and Ive had Bamboo restarted which fixes the problem for about hours +The alternative would be the you spoke of +However I will not have my on method called until after all prepended javascripts are already added to the response +Changing spock dependency fixed the problem +Will create a placeholder for doc in the wiki wed appreciate some documentation +submitting again +All done enjoy +Switched to using classpath entries which covers all these types of things. +Thats the problem we cant replicate it easily we cannot find the pattern +See also +HTH sorry this took so long had to update some of the scripts we use for this appreciate your patience +Sure let me try +Let me double check it I may fix the issue +But still there can exist situations where the data in SI cahce will be different to CRI ve committed the doc update in revision +Do we care about calling zkcli from with? Should ZK users call ZK CLI from their separate ZK installHow about removing +Add more Overall totals +The wiki page is very useful! Thanks for that +patch applied thanks James and Michael! +Changing the fix version from to +Attachment has been added with description patch created using cvs has been added with description FO sample file to test the behavior of fop before and after the has been added with description updated has been added with description proposed patch to has been added with description test fo file words with punctuation marks and +But somehow I think I should try to understand the naming scheme implemendted by fully +revision the UUID lookup query is the one that is bothering me at the moment +If you are not really familiar with this and time is limited I could try to upload a custom build to a selected area at Codehaus +Thanks Aaron +tnx for again! +Looks like the sources have changed since then + is the patch to the main class is a new test added in the ve added the patch with version Ive slightly changed the patch +This happens with the seq numbers mismatch +Fixed in commit d +Assign unscheduled issues with Blocker or Critical priority to +From my initial testing I couldnt tell a difference so I need someone more familiar with it to help +The attachment is added containing the error messages when the enforcer plugin has the additional rule +Looks good +So can you query for the user with referral resolution enabled and bind to the resulting ldap url for the user? +I have attached a stubbed layout of what Sqoop workspace might look like +at +The difference between the classes extending and the classes extending isClasses extending are standalone classloaders whereas the ones extending are decorators of otherclass decorators such as asynchronous or singleton store could one day become standalone cache loaders and stackchaim them updown other standard cache loadersIll open a discussion in the design forums soon +I think its sufficiently dissimilar to this one to warrant its own issue +Hi CassieI have updated the patch accordingly +Now that work is well underway with AS all previous community releases are +Fixed at r +Lilians are you sure about this? Afaik below changes only apply to releases and later +Address Lukes comments in v patch +On a related note I just saw that Mock documents for missing classes are only set up for the simple syntax not for the labeled syntax. +Also the generic task depends on it already +This will handle both the secure and nonsecure cases +Thanks for the patchOK it looks like youre replacing two cases in where we are manually copying arrays with for loops with calls to instead +Verified fixed for Action Form and Page Flow scoped and Range rules. +Ill fix +Oracle hsql and mysql this time +this will be fixed when be by +Hitoshi ok I see thanks for explaining +Ok +This check must also be done when following shared dashboards +This assertion can be reproduced though if or is removed +What is the status on this bug fix? It would be nice to get thisfixed since I think it is one of the more common intermittent in thenightly tests +It would be helpful to have the code that is emitting this error +However Im not sure how to trigger this +Right +Added an additional check for Bidi numbers +Has this been investigated yet +In the patch from Hamlet it is a public method +Again since derives from code should still compile in the case ofmethod signature changes +Sorry I may have chosen the wrong security level +To prevent IP conflict configure different subnets when multiple networks are connected to the same the doc and it reflects the changes required +Thanks for the reminder because I lose track of that issue +Presently the zk apis usage are scattered in Bookie Cookie +Are you sure that you have a session created before? The encodeURL transformer does not create it +Any chance I can get the basic setup used for the RAR example above? I would like to get it replicated +i did verify the new version fixes the issue when i submitted the patchSee the link attached in the description section for the fix that was made to surefire +Pull request contains changes that add allowed values for all onoff attributes +Mind reverting this and fixing again with that method of exclusion +Moovida updated for these changes passes +Thanks +One thing we should do is to make the constructor package private to reduce the risk that it the widget map is replaced with a list that we dont want in a runtime environment +I am the bug so that bugs gets attention of right persons +I have recently run into the issue with and am looking to fix it if I can +Ok +Cannot reproduce this +Fixed sshexchangeidentification Connection closed by remote host issue +But Ill talk to the team tomorrow to be sure. +Right now I am only aware of two JVMs that run on Windows which are Suns and BEAs +Thanks for helping Robert! +They both remain about this issue being present on the new GSE anyway +So it seems pretty doable +Not sure that this JIRA was intended to be for the AWF project +I have created a renderer subclass inheriting from the existing and added that as the renderer class for a Navigation pane for PDA in the PDA Render Kit. +Hi Nicholas did you commit this to as well as? If not I believe the fix version should be set to +Since the PR has been merged into that repo Im scheduling this for +Only the virtual dispatch is unimplemented +If it comes to pass that we need a new patch Ill open another one +Hm +thanks patches were applied closing. +generateBenchmark generates results from the source of truth +For now I defined small as but that needs to be tested more +So conceptually it all looks goodThis patch literally includes the latest patch you posted on +N trunk version of the patch updated because of merge over +Guys any comment on this one +This patchset depends on the refactor attached to +Last TCK ran perfectly from start to finish +Mapping reader should use but that namespace value should be corrected first to prevent stuff from breaking +If they need meta information it should come from IDE specific files +Committed to SVN. +Also we are planning to drop the top Apache Foundation logo and put it in the footer +Attaching a summary of localization changes needed byAlso attaching the programs used to generate this summary and +Ill take this +for trunk only +Therefore I am not sure if is a bean type of Office +Patch implemented using the simplest description +Lowering the priority on this as there seems to be no activity perhaps the workaround given above is all that is neededAlso there is the Subdirectories tool which may help +Done +But that return value is never used by +IRC discussion conclusion tests artifacts will be removed from repo current test wont be changed to detect if any file was m resolving this as partially complete because we have about of the source jars +Can we create a accountErm +Fixed +My current plan is to leave the server side as it is and upload a new patch that hides the client side SASL code into a custom socket +Forking has worked +Potentially we should provide abstract classes that have those conversion things already implemented +For the describesplits call it makes it returns sub splits the first of which is a wrapping split +As you mentioned a more recovery would be nice but for now simply disconnecting a node from a broken version history seems good enough +The opinion is that this should be done as part of a implementationWhile not optimal this can be done +This issue is closed now +Amareshwari could you please move to a separate? It is currently in and it shouldnt be a public class eitherI removed the class +Once you go over that size we cant properly address records in that file +Fabrizio BottinoSure Ill take a look at it +This pull request only has a single commit so I can just use patch from that. +Yep +Seems reasonable +I applied a slightly modified version of the patch to account for common errors in redirects +Workaround only supports containers that implement which is not the case of sSo you can use mx instead which is supported +Let me know if some of my assumptions are wrong +Updated patch +Patch to unwrap the OOME so the main in will shut down the server +This should be solved together with to provide a consistent way of configuring settings in OJB +It is a sufficiently obscure case that I recommend moving this out or closing +System test passed +The attachment contains changed files. +Is the length parameter a random value as well? You have to check that before you start looking at the characters supplied +We cannot reproduce your issue with if theres some way you can help us reproduce it well gladly reopen this issue again +Fixed committed at r. +minimum length chars +Tracked down the issue to the SERDEPARAMS table which in the mysql create statement has PARAMVALUE default NULL +Thanks for the input guys! +Sangeetha is it still an issue? I think after enable acceptra in the interfaces file its OK +patch to move check if name is null above accessing Jaime! +Note for commiters +Thanks for catching it +And also fix makefile to suit new structure +Patch applied thanks! +no output is generated +This allowed the sonar maven plugin to deploy to the sonar instance successfully +Committed revision I think this issue is wrapped and we can spin the other improvements off +The problem is a legacy client I have which assumes that leading spaces are trimmed in the matchedDN value and I am probably not able to change this +Otherwise Id like others to feel free to work on them too +Added targetblank +Thanks +We have discussed a lot about how we can support languages code in ADS last year +m are licensed by Apple already +Its just one of those things where I wasnt doing the proper null check +Not very user friendly in my opinion +Thanks +For my education these are the features are that supported in the container? Doesnt affect the gadgets +Sound goodSome comments Throw instead of +Yep all of that should be fine once we fail on a mount we mark that log as corrupt by rolling the other logs +this patch is for for the for branch +You probably have lots of TX data in pagingIm removing the executor call here as I didnt really need it what will remove this message +Please provide a reproducable test case +It looks logical and niceHowever Im leaning towards keeping it very simple +All steps show that the jbosstools code is properly updating content +Please verifyThanks +Good idea though youve got some extras in that diff for and +I translated files belowexamplessrcnetsfclickexamplespagepanelexamplessrcnetsfclickexamplespagesecurityexamplessrcnetsfclickexamplespageexamplessrcextrassrcnetsfclickextrascontrolframeworksrcextrassrcnetsfclickextrascontrolcalendar Ali will integrate this in regards in trunk. +My bad! I thought Id attached the test case +Thanks JB +javac +Closing resolvedwont fix issues. +my local ant test will not finish in time to make the cut i think +And extremely +Hive needs not of it + +on the patch +I should also mention that this is running on Windows Vista SP with JDK +turn on the debug option for streaming test cases in is the just committed this +We should add a regression test for this case +Bobby if you havent gotten started working on these yet Im half way through a patch for and +I only wanted to ask before commit because it is assigned to you +We do not need moderation on this issues list +I committed the patch +I just tried edit some posts in test env +Does anyone know why we are getting this javadoc warning? also got on javadoc and Im not sure what the problem is +Thanks again Earwin! +We could have a thread pool on the side just to deal with such requests but this would be effective only if not all events require blocking on IOMatthieu and I also discussed this and realized that it would be equivalent to having a PE on the side doing such requests +Maybe it would be best to leave command line utilities to a later JIRA once it becomes more clear if there is a need for additional support +The problem went away when I added my new layer in the empty group calculated the bounds from data and saved the new layer is not tolerating empty layer groups +Help section is still missing +Fixed again with implement fix in rather than in c +So i guess the proper fix would be to let all cases where the Batcher hands out a Statement having it keeping track of it and alle cases where the Batcher is asked for it let the Batcher also clean it up +lot of tests are failing +First draft patch attached +Well fix the recent changes and review the original thread again +srctestjavaorgapacheflumeserialization +Agreed +Attached RAR i am using to create a new JMS a new JMS Group deploys successfully if there is no plan bundled with RA +Done. +Can we keep this issue open until the fix is provided and included into a new Derby release +They work on EC and on standalone machines +Test in release + file that goes in contribbenchmarksrctestoalbenchmarkqualitythe test is setup to use it in zipped form +This issue will not block the EAP release + +Then it is a classloading or module dependency this issue and will try to change the component to Jboss classloading + EAP is also concerned by this issueIf its the cas can you update the Affects Versions attribute +Closing this as Fixed +partitioner inputformat inputsplit etcplease provide a javadoc link to at the first mentionOverall comments This is extremely useful +It adds sourceleave function but in general quite similar +The current versions seem to work ok even though they are in some cases newer than ones used in struts + +That way we get correctness and we can still interleave syncingRTT in the pipeline +The section Increase textsize Default for Sybase contains the updated directive +Adds rolling logs at N edits +Fred I mention in the initial post this behavior can be observed by just creating a new Maven project with packagingwar +what a coincidence Just recognized that was just reported today + +Username mnewcomb +Thank you for pointing me at after the bug fixes introduced by the mojo javacc plugin developers I was able to use it +I just committed another change +Big thanks for finding this bug and providing a patch +This is really a duplicate of +Anyway we are doing more testing +I did see an error similar to this while investigating but when all changes were completed I did not see issue is related to the final change in this jira fixed. +Not sure what other datatype to useDo you have another ideaWe would love to have something more structured but this fit our current needs +If varibales for some specific properties are not working please create a specific issue +Description Please explain me this exactly have to doWhat should be the name of the TOOL +Descoping items not being worked on for M into Unknown Fix Version for not +iBATIS test dataExchangeisThis is a very serious bugThis bug crashes our business system many times on a day +Cannot reproduce with latest trunk +No output format yet just converts the loader to not require gossip +There is an issue where client can currently pile up on a single region because of the way we do client queues by regionserver +I tried the patch and observed that it fixes the issue +Done +Are reducers informed about new locations for map outputs on the flyThat is my understanding otherwise no reducer could be launched until all mappers had finished and all reducers would have to be relaunched if a map task disappeared on a bad nodebq + set via dependency find schemaLocation relative to baseUri +Example of a python script that submits jobs to workflow and checks the number of ESTABLISHED connections to the workflow manager running on localhost +I think youll need to roll this back +I guess I dont have permissions to close it can you close it +Local commit ready +I dontknow this part of the code well +partial fix via need parameter in the JAVA number of retries and time between tries. +commit this later today +Stay tuned +not to verifychoose itI meant not to verifyuse it +My plan is to put ASM in the class loading path for coprocessors from the outset +ah! good catch +Hello MamtaI tried the repro as well and I see varying results +Closing this issue. +daniel I was thinking the same thing over the weekend +v grabs CF reference before removing it from map +Good idea! I will take that approach +A simple testcase shows that this is the case +Hem this patch has been out for almost two weeks +Please use thrift to generate these classes as per the above cited link +The purpose of having a test was to encourage and set up good practices +defer all issues to +Manually tested +Max that is true but one thing which is odd for me I did not touch project at allI just open from ZZ project and got exception ten times +I was watching that item and didnt realize there was another jira that went in for serverside trash configuration on v +Ignore the patch +Still we need to get our nbexec executable every time we run clean nbm +Youll have to post your extension or detail the method where the modification is being made for us to understand what is happening +Existence of styleClass will be very helpful +Made this criticalAnd looking again the thread doing the readInt is not stuck rather its RUNNABLE +Since we havent heard anything to clarify on this bug Im going to close this issue +Thanks Vinod +Thanks for catching thisNow fixed +working on which uses the socket to call back the response +Question Why dont you load the file through a provider and give the actual value as a ctor argument +Please test close if verified fixed. +Incorporates Yoniks patch moves params out to and adds a unit test for the exclusive operator +The result is the spi does not depend on impl anymore +Changes Add and delete requests are processed after local operations but before distributed operationsThis patch makes no changes to the versionAdd versionDelete methods and is much less invasive +There should be two repositories for community and product binaries +On the other hand the simplicity of the checksum makes it more certain that weve implemented things correctly and will properly detect corruptions +I verified ant clean jar test still works +Andrew thanksThe patch is applied as expected +I dont know if this is the best way to do it but at least de class should allow for bundles +what will happen with this code once IE comes out or some other browser that accidentally matches this hardcoded ruleTalking on the IRC channel it sounds like is the way to go. +Just to make one important goal more explicit this patch will as a side effect and what a nice one remove all staticness from the existing code + were two unit test failures in the last patch had a failure which was because joins describe now prefixes the outer relation alias for each field corrected the test case to update the expected result +No problem Mircea but signore Berindei told me he already has a solution for it +Any chance of these changes getting into RC +package +This is an example webservice that I deploy on Tomcat with Axis +Please check should be possible to specify on package level to avoid conflicts with plugins +Committed to trunk and +also see +ill be doing a bit more testing +Ive just committed this to trunk andThanks a lot for the contribution Ivan. +And speaking of m I took a look at converting the build to m this past weekend but gave up +still looks nice +Attached is which differs from in that it renames LANGMODIFYCOLUMNPKEYCONSTRAINTto LANGMODIFYCOLUMNEXISTINGCONSTRAINT +Thanks! +Added on Cave trunk revision +I guess that as long as the SDE java API can use the datastore will be able to use it as much of the datastore needs to be reworked thought I have no welcomed +update with the new configure parameterCurrently flushPeriod only deals with application not system +Fixed as +I tried it and it works correctly +It has not been brokenCan a interested person please take a look at the patch and if everything is fine recommend a commit or give a commit + +Thats what my worry is if thrift cant handle nulls cleanly then we have to litter our codebase with everywhere +Please verify it +it doesnt address the fact that we still have to partition by dynamic partitioning columns and that requires the same operator that requires +imagine trying to fit initializations of generics into chars +See my comment +Have just committed the final changes for this stage after some sorting out and testing +reopen after files into CVS +Its believe this duplicates +SYSTEMSEQUENCES is a bad choice wont return one row +Not fixed by I see your point +I thought about this too so Ill try it just a bit later +For both cases you can use a generic message such as You do not have access rights to this resource please contact your administrator +The patch did not include tests since there were existing tests in which I confirmed work in the branch with this patch. +This mainly benefits high concurrency random reads lets get this working and then think about the write pathI can devote some time to benchmarking and reviewing this more thoroughly but likely not until the end of this month beginning of next +The project I contributed is an entire piece of functionality that is not intended to be broken up which I will continue to maintain and add on to +Tests passed cleanly for me on +JohnThank you for the clarification +already fixed +Hey Brandon +Thanks for your response Owen +Bulk move issues to and +I will attach a patch to this issue when ready. +It is actually important that the editor and toolset know what ports to interact with the server on +Ill commit within the next day or so unless there are objections +Arguably if the tests are ran with TRACE level of logging youll see everything including the process outputs +if the above works as expected +isReadable should return false if the method is not readable. +I added under and modified accordinglyI also modified to reflect the license change +If a large build with a long Sonar analysis phase runs and a small build is started during this analysis phase the small build will get a timeout while trying to acquire a lock for the step +Default false +So there is message per batch and each segment has about messages +The gap threshold can be configured with G +Wise requires the Sun JDK to import the WSDL and compile the proxies +Im not sure if it would be possible to receive a nicer error message +lets discuss the future of once is out +Not default replication max replication +EN and NL localization fixed in like Dirk fixed these so marking it fixed. +Can you please try this with the trunk and let me know if you have any issues +Sorry for probably stupid question but how and where I apply Dans patchThank you +Resolving as fixed. +Only added failifexecutionfailsfalse failonerrorfalse in to prevent build from failing when incorrect browser is found on host system this would confuse the user since the demo setup completed succesfull. +Hudson is set up to build with Java +could not duplicate +Did i overlook something? Implement it differently? comments on this change of functionallity are very much appreciated +So I suppose its related +If you do create a patch in the next few days reassign this issue to +patch committed thanks! +Is it in nowLooks like asfbot did not pick it like this has been fixed in +The exception is when takes an information by the http request header through the parameters file and folder but nothing can change them except modifiing the inputStream of the httpRequest +But then what would unit tests for Solr be +As described in the discussion on this bug the described scenario works finefor me if I install JWSDP without the old early access version of JSF thatit included +It might be obvious right now but later on it will require an extra effort to find out why ed patch has been committed +Heres a patch to change the comments to include a remote JMX setup by defaultThis seems okay to me as its not enabled unless uncommented and the comments point you to how to enable password access as wellLars what do you think +The fix applied to breaks rendering of +Hence Id propose that somebody updates this issue to have the same status like +Sounds good Ted +Here again support would be a large undertaking +Reuploading Dheerajs patch to trigger hudson +Thanks Ted! +Heres a test that fails on release but patches with current branchxSo I think this confirms its fixed by Robert! Ill try it out. +Hi Maxthanks for your commentAttachments are stored in the same file as the documents but they are already stored apart from documents +Its my bad I am fixing this now +I suppose libSVM is in a similar situation +The fix would be to ignore the initialization you do in the class initializer and just set the unmarshalled value +I totally agree with and +I just committed this +Removed the sleepTime inNow the delay is managed only in the test case by using a as suggested in the list +Though when this issue was open the cause was not known +But an insert that overwrites an older value implies deletion from the base and therefore a potential deletion from the viewOne disadvantageadvantage Cassandra has is that an entire row is not available at mutation time so we need to defer all deletes to the view until read time +Closing all resolved tickets from or older +FWIW see also fix version to unknown version to allow release to progress +Im going to attach it as Java files +Updated to use a configuration +In these case we can think we just put the processor in the middle of the routing rule and the processor will take care of the more detail EIP verb actionSo I dont will to use verbs for the EIP actions in Camel DSLAny thoughts? think we need to align the Java and Spring DSL so we use the same verbs otherwise its confusing +Changing fix version to to match parent bug +This prevents the jira from getting littered +Thanks +Yes this is performance relatedThe code previously performed a JNDI lookup of the connection factory for every session retrieved from the pool +So all you add to is this I think we may need to just add a little strategy method on the so you can reuse as much code from itAs all you do is that if its that bundlelocator type then you use the +Is there some reason to believe it was fixed in RC and final but not RCSadly I dont have an environment to try in at the moment +Closing. +Manually tested +I just updated the tutorial and Daniel has a bunch of refactoring which among other things changes the handling of + +I also reverted the default behaviour to that of so if your Partitioner is not expensive there is no need to implement the new interface so you shouldnt need to change probably are the only one on Earth using a different naming strategy but dont feel bad +I still have interest in seeing this issue resolved but I guess I am the only one who ever attached a into an Email programaticallyBest of luck in the next releaseDave Cherkassky +Updated possible Jason I prefer waiting for to be committed +Anteater script is committed stops for me at runs so I can confirm theleak is still there +People who want to check in their entire project are going to have to learn to do it properly I think +Deferred to defer of unresolved bugs to +fixed +Id definitely like to see a patch made to ACEGI though because the modifications are simple enough and AJAX is becoming standard enough to warrant a patch rather than maintaining extensions over time +You should make the abstract and pass in the wrapped postings format and delegate independent of the field +Good idea! I initially thought of the growableoutput but i didnt want all the resizing +System +I think going forward we may add an extra column for the DB native type in addition to the JDBC type +Have you considered commiting the test? Is there anythingh I couldshould have done better there +Moving these all to a Doc release version +Please use the forum for forums you says to to Jira and on Jira you say to forums No support needed just a question regarding design and if the implemantation cares if the session is from different entityManager other than whats being passed from the entityManager property. +And xsddateTime seems to be the most capable comprehensive +Doh of course +As Gilles mentioned changing the package name for was discussed and voted on for +sorry about the typos +Removing severity blocker as this is not related to Spring IDE itself +I committed the patch to branch +Patch checked in +The attached patch called is an update +for your patch +sorry for not complete comment beforewhyBish can we see your demo of the roudtripping perhaps in video or screenshots if you cannot publish the code +part of release. +Changes look reasonable no comments from me. +But regionplan is one thing which will tels me where the assignment is actually going onIf master wants to start a new assignment its again the regionplan that gets changed + +Andrea can I backport this patch to for consistency +fixed in brew looks good in EAP CR +I only moved existing code to get a meaningful error message +Thanks now I understand headers is the map and is the header key which would refer to the data I am resolving it as invalid +Verified in +You will get an binaries to SF +Added license info +Attaching patch +Addressing Stacks and s comments +Shall we use add instead of manipulating uid directly +For the record I dont think this is ready to be committed as is but I should have something ready before end of week +Duplicate of. +We should look at adding a sample webstart app like this to the distribution in andor +The keystore should work with Tomcat but havent integrated the keystore manager with Tomcat yet. + is out and stable time to upgrade +Hi SandakithThanks for your reply +Tested this on Samsung Galaxy S running +Ron you can apply the patch and commit it you are a committer +Yes there was discussion before release and this was determined not to be a blocker forFix is on trunk now +So far I tested only of them and fails because of this there any work around +As mentioned we had to modify tutorial + are you using MANUAL as default value? In the JPA spec the default value for flushing the entitymanager is AUTO. +Thanks CharlesPatch was applied at r please verify +Thanks +If you post the diffs Id be happy to spot check if they just look like bugfixes +So I am on having a configuration option to use UUID +gadgets name is unique and add a name do not set the Fix Version and this has just been fixed +reopen to set fix version +not reproducible under what should I do with this +Reporters of these issues please verify the fixes and help close these found with last runs +In any case it means leaving zookeeper null in the constructor and checking if zookeeper is null at a later point which is what my first patch did but i think using null like this is ugly +Technically this is a regression upping priority to +I can certainly agree with that Leonardo +Got a Hudson problem at the moment +Well I think this is a totally valid issueWhy not let the SPI implementation do the check +I baked in the config to the web grid instance and put the JTDS as a mvn dependency on the inherited project +It looks like the function live is misspelled +The timestamp issue is a low hanging fruit which can cut +This causes file descriptor leaks under at least fedora for both the sun and open jdk vms +This was probably fixed as part of the release but Im leaving at since this is where I verified the problem as fixed +I will try to investigate this for but its nontrivial +Thanks Adrian! +Nothing generates that event +is a new part of the serializerwhich is copies from as well as a slight of some serializer classes +change +Ill bring the patch up to date and look into proceeding with some final testing and commit +But this has no effect on +Is blocked iff we want to use the provided unit test +Hi Willemas discussed here is the svn diff on cameltrunk +MRI appears to do this by ignoring the return value from fseeko and since the primary goal of seeking to current is to invalidate bufferswhich happens regardless of the errorignoring the error should be d appreciate a test case for this ideally something that can work on several unixes but at least one that works on linux against an unseekable filedevice all linux versions will have +Sweet graph +Sory for delay +Maybe it something that I could request on the projectI havnt come across an open source solution yet +Sorry +Thanks Sebastian! +The only solution is to store all the relevant properties and parameters in the service context since there is only one service context for lifetime of the serviceSo when you deploy a service in application scope Axis can not make set thread safe +Why not call log after get returns true +Please check that the bug is fixed for you too +Initial patch. +And I was wondering why this would not be done in the other order? As the http character encoding may be configurable it would make sense absent other information to identify the same character encoding in the http header and in the xml declaration +The underlying problem was misunderstanding by the user over whether to use or to delimit multiple services +So all issues not directly assigned to an AS release are being closed +In general there are two options either make textValue property persistable which makes sense from search by variable value point of view or use in getter for textValue +We dont want to encourage using standard and guessable usernames and passwords +thinking about this then im on this if QE is ok with it and our branch CI builds can pull from this new repo instead without too much fuzz its fine by me +Well we need to get all of those over to central eventually but its not a blocking issue +You can use guards and only where it is not possible waitings are only other should wait for page to be fully loaded +I think a bug implies something not working in the plugin +Time to go update the book againThanks very muchSeth +you mean remove from? then the ejb integration is not activated afaik +Rafael is for Richfaces dont mix that with +adrien I deleted the jenkins job for this +Would you try upgrading to the latest Mono shipping version and? I will try testing this on Mono for Windows to see if I can reproduce +I will close this bug as I see +Looking at the hunspell source code I might be wrong but I suspect it just skips that specific rule with some warning +Vladimir do you think backporting this change is feasible? If so please add as a targetCheers +I will respond to your emails as soon as possibleRegardsManan + havent changed since last is updated to latest version +As defined by the bean is a resource bean +Kan thanks for catching this +this problem will not be fixed in the development stream +checked in +Thanks Jacob I applied your patch +wicketsrctestjavaorgapachewicketresource +will be the release where we first go out modularized +Thanks +Wondering if the fix can get into +Im not sure how you lost the module since it seems to be present both in the as well as the default that my server build starts with +Its simple to replace timestamps with latency values and then use an arbitrary value when calculating the phi score for a host +Im even tempted to just close it as INVALID since nobody is following up + +The rd patch has been applied under rThanks to during tidy up prior to +It is possible to get the version information via the class +Were running with M for the JVM heap size which should be more than enough for this question is why is the directory scanner caching references to files so aggressively? This seems very inefficient +batch transition to closed remaining resolved bugs +wondered why my workspace was broken after a m eclipseeclipse +so I dont think the currently attached patch will work +Bug confirmed after deploying the attached process definition and following the steps to check for null transition name in page fix by going over the steps again +RTC Task bulk close of resolved issues. +please provide screenshots of what is wrong the palette should show content when it actually makes sense to show it +This seems to be requirements case not supported in the current implementation reopening the issue +bulk fixing the version info for and all affected issues have in after release. +Now when it executes the it does not meet the criteria to go through visitResolved since its state is linked and not resolved +Do not file bug reports when your custom javascript breaks in a particular browser +No part of a primary key can be unknown +Probably it does but havent given it a try +Committed in revision . +Whatever works +If you print FS arrays things become very confusing and its hard to see anything +The problem started when extends from so the code of get becomes incorrect. +I am looking forward to hearing from youRegardsAntonia have accepted Antonia as a student to work on this proposal. +Both issues describe similar problems by their nature +Can you check in the nd region that API command is received +journal flushing latency supposed to be very very low since it was used as a separated disk +Id like to close this issue but Im a bit bothered by the Descriptor Format page +It got notified that the heap will be running out of space soon shortly before the other thread starts parsingThanks for that tip +So i have just made mycopy reflect the list of files in your Okay it works and i am ready to commit it +hi SaschaYes we do need a wish list and I think JIRA is good enough for thatAlso I have created a new issue for my wish thankswith the jQuery branch being merged into the trunk new issues have been identified in the onepagecheckout and they are the remove item link does not work in the shoppingcart panel and the cart is not updated if cart quantity is changed or promo codes are enteredRohit +bulk defer of featurestaskswishes from to +If someone needed to use a in multiple threads after this patch would the recommended approach be a GDR for each consumer? +testing the streaming job again +Since we support SKIPWAL model the is still needed to main the write order because there is no wal sync operation at all +New committed to trunk. +for +Wouldnt it be an idea to move this issue into +Updated the patch. +Hope to take a look at what you guys have been up to again soon +I looked into this for a day or so and there appeared to be some bootstrap issues around the Provider that required it to be setup before the could be initialized but I cant remember what those were now +Also I believe that Package in the context of maven already provides something like that in the jar +push to +Part of a bulk update all resolutions changed to done please review history to original resolution bug is not the same as since this bug is related to the pickList component whereas this bug is related to the itemDisabled attribute of containing selectItems inside the pickList! Which doesnt seem to work in version +Moving out to for the moment +Adding new jmx properties +With this being a rather important part of initialization and no time left to test the fix I think we should move it to +This is with file is on an NFS share +property has to be set in openejb properties +Solved +Do you have a stacktrace handy +Need to fix log messages +Thanks +Please do not target a specific version if we dont know that we will really work on it for that version +Daniel is right defaults to all methods not just GET +the User library is a combo box +Can we update the unit tests though +I was recently convinced that it is not +Hi FrabriceNo Im not blocked but I have no much time if you want to contribute a patch it will be welcome + +shift to next last option works for me +Len mentioned that the update site had to be called SOA Platform examples but Brian it seems now like they are just going to be apart of the generic update siteThanks nick this is not about update sites its about project example sites +Although I dont think we should be doing more moving around of files for AS GA lets discuss it after to related request by Sacha +Attached a patch that implements the proposed dummy parsers +Huh no clue why youre getting that error +Done +I dont see how this change accomplishes anything useful +Great work guys +Can you elaborate a bit more open a bug report in the Bugzilla for the component bugzilla? Thanks a lot. + to and trunk! +First version of this fix +unexpected server error in chrome when trying to save data via gwt scaffold app +shouldnt GCJ implement that class? AFAIK that is the standard API for accessing a we should definitely have other alternatives in the bootstrap +Ugh +OK thanks I confirm it works in IE! +JDBC tests ran cleanly. +And if the procedure is used for the collection where do the values come from +Will check ASAP +r +Martens suggestion that I quoted above would be perfectly adequate in many upside of parallelizing bean initialization in the Spring container could be significant for a minority of applications using Spring while the downsides the inevitable bugs added complexity and unintended side effects would affect potentially every application using Spring +Youre talking about downloading the latest version of the I hope your whole copy of pdfbox is up to date isnt itAs pdfbox works for me displaying printing and converting the pdf I suggest to ask some of the fop people +Added very very end of error log to bottom of as link issue is closed +made a commit for that pushed link +Ive deleted the two previous patches I submitted to avoid any confusionIve also renamed the project to to reflect that it contains items that are specific to the project and are not ports of Java contrib items +This has gone stale +I think this is a general issue not specific to code generation +It was replaced by which has been applied. +Nice Ran the clean then test then site +I guess I stumbled a bit when clicking around in my IDEI am still waiting for Globalspace to accept my request for a trial version of their product +Unless I misunderstand the HAS JIRA they would probably be more appropriate thereThis is a great start though +The problem is theres no possible comparison between development and production systems +Manually tested! +Thanks for wrapping this up for me Andrew +follows complete errorLog maven version and mavenconf thanks alotsThomas +For M make existing groovy refactorings easier to find and potentially remove Java refactorings +Due to the script refactoring this got lost in the merge sorry +dont see a patch anywhere +Todd Do we need a refresher on this patch for? Looks like you fixed up a few things subsequent to your last version +The local repository is fundamentally tied to the Maven tool therefore Maven should manage it +the docs should contain the description of how to run that script and produce the db schema creation scripts +Yes I just noticed the same +bq +Actually just ran some tests and what you said is not strictly correctAs long as the file is parsed before any mappings files referencing it this works +This patch effects the profile section of the pom only +This is required due to EJB only affects a clustered SFSB +Now all I have to do is figure out how to put all things togetherD +I have reported the problem with nexus and will close the repository when the problem has bee resolved +I want to put this in the object +so close it. +indeed my demo works OK without specifically deploying this helper bundle thanks for a tipJB I agree it may not optimal to have all the JAAS modules up on the startup but also agree that it also makes it easy for users not to worry about setting such modules in the first placeAs far as Im concerned I consider this issue resolved as wont fix or not a problem etcThanks +Attachment has been added with description fix the problem in WMF can you confirm this solves the issue +With Random it will be just broken because you may see someone elses sessionCan you explain what is the current problemIll answer you separately in +Moving to +This would be useful for specific stuff like people confirmation links and things +Hi could you please email me your class loader factory implementation? I would really like to give it a shot +Yes I was thinking about CLHM +Looks good +Ill need it for the next version of my software +We were waiting for a long time to make this switch +Okay moving this to Camel so Willem can fix it when CXF has been released that has the fix +trunk patch patch with a few modifications when ran in secure cluster +Test reproducing the in it slipped of our radar. +Btw I have tested this against both Jakartas and ORO packages +I think this is a good thing to have +implements parsing of some other formats provides test case for patch for latest cross check and close this bug +Thanks for your work on this +so well do filters first. +bq +Ok it works just like I described it above! And with a nice little tab too +Id like to add some info in the ZK + +I live on the mentioned CST zone and never had this errorCan you explain a little bit more +Oh +Add the ratings when counting and hope that better ones will stand out even if they less +Should this be closed as a dupe of or? + +The client must not send lease renewal to NN for a while so its lease expires +batch transition to closed remaining resolved bugs +despite what some might think Ive actually wanted more things to break inbetween releases than what most of the other devs have wanted +This is not being unset when a new test is ran +An updated version of the user schema prototyping work +to the approach outlined in +Test and a patch +I hope to have those tests fixed by tomorrow +To follow the default test profile file and some ones +Thanks +Attaching a patch with two simple workarounds that are harmless to the itself but extremely useful to the downstream +Thats a good question +Logged In YES useridI also tried to remove the finder it complaints about andthen it just give the same error about the next one +Should be fixed +I have removed the changes pertaining to that was in the previously attached patch +youre right i reread section and agree with you that we currently dont handle IMPORTUUIDCOLLISIONREPLACEEXISTING on import of jcrroot +an AE in their script +Removes the queries +and I think the bug is fixed can you confirm thatThanks for the reportBilgin +The patch didnt apply for me +Has there been any workprogress on this +I just committed this. +Thanks Luke. +Stefan convinced me that it is easier to track these changes in one issue +Just changing the assembly to include the connector projects or referencing jar classifier artifacts doesnt seem to work +On an API level there was a way to set the execution environment +Technically Creation of a column whose values are DEPRECATED BETA or null Ability to define a status for a rule through the Rule annotation or from the XML format +Ive made a commit here so that new servers get all required attributes set +thanks +Anil what is the status on this task +Link this to Better Datanode handling +So running eclipse clean stopping and running plain eclipse afterwards showed the old faulty behaviour ie eclipse am sure I get you some steps to reproduce this issue locally since +If false is set then the all the query stages work OK +Thanks +Might be ok thenOn rewriting +Once I rectified this I ran threads in a continual parsing loop with no leaks +has been committed +This will not happen with the db translator. +Just look at where is the test enhance to use this new functionality +To clarify my earlier comment about different EC kernel images I thought those were the actual linux kernel images I was getting but they were actually just different builds +So a that didnt rely on Writable but only on the Record base class would be great to have +Dont implement the Adobe specification at all thus denying ourselves whatever value there may be in that technology +the community version +Does this get called anytime throughout the test? Is it even requiredThere is also the question of what is the type of that you are using that reproduces this issue? Are you expecting all of the messages to pump through the network connection between the brokers or are they also being consumed by consumers that are connected to the broker that contains the queuesHow many consumers are you connecting to these brokers? What protocol are you using? Are the consumers coming and going or are they permanently connected throughout the testHave you been able to reproduce this issue using the latest releaseClaudio +Patch attached. +I guess it proved to be stable and can be ported now +Also someone who is a commiter should decide if this is something we really want it limits hbase to always write under honey +Elliot Metsger you can still submit your test cases if you want. +This should be done and is related to the tasks for updating to git scm as well +For nowthough the present simple implementation will do +Hi Davidafter had been removed from the trunk Ive been experiencing an exception during container startup which Ive been able to track down to your file +I just commited a quick fix +Function is provided by updates in the module committed in processing of the faults declared by annotation in +Eric I am not saying your idea is bad +After xjc compilation I end up with more than a generated classes that cause many PMD violations +Asking people to work on CLI simply does not make sense here. +Do you see the same problem opening the builddocs? This has no use of font tags or style sheets so I dont know where the redmond and thinice references are coming from +Storm would be really good candidate to add. +We should consult Justin +however for this case ive just rolled something myself since i dont know how to access that class +Resolved at +The changes that went in for this jira seem to have specific code changes for and higher codeline +Let me submit it again +Thanks for updating the component Mike. +While Im moving this to M we should get this completed by or revisit our approachFor future releases we will create a more automated process. +In need a mechanism that triggers an alert when theres a single test failing ignoring i have another tests that succeed +The tests test both modes depending on which is selected +I suggest the workaround Markus provided above +I dont believe it is a serious error that prohibits functionality in any way +Committed at subversion revision +You should use method to get list of in file upload listener +Francisco can you provide some examplesAnd btw we have changed proposal sortingfiltering since beta and have a few issue in jira for future versions +Im now able to run goeserver for hours with only mb of heap and with threads sending wms requests on two layers in a continuous patches require changes in gt thought +Hi DavidDo you still affected by this issue +I dont know if that can be wrapped up in a jar yet? Or if its possible to do that yet? I hope it is since if not I guess this idea gets killed to But Rest support stuff got in for noThen we need TimAndreas to package up the files that should go in www Hopefully can make use of the same openlayers files we have their now but I think its ok if we replicate +Patch updated + +Thank you Ravi. +One more thing +Patch to add which writes out vertices with as Arun described with a configurable delimiter +Initial commit to trunk on Adding assembliesAdding assembliesAdding assembliesAdding assembliesAdding assembliessrcAdding assembliessrcmainAdding assembliessrcmainassemblyAdding assembliessrcmainassemblyAdding assembliessrcmainvarAdding assembliessrcmainvarconfigAdding assembliessrcmainvarconfigAdding assembliessrcmainvarconfigSending assembliesSending configsjettyAdding configsjettysrcmainAdding configsjettysrcmainresourcesAdding configsjettysrcmainresourcesAdding configsjettysrcmainresourcesSending configsjettysrcplanSending configsAdding configssrcmainAdding configssrcmainresourcesAdding configssrcmainresourcesAdding configssrcmainresourcesSending configsSending configstomcatAdding configstomcatsrcmainAdding configstomcatsrcmainresourcesAdding configstomcatsrcmainresourcesAdding configstomcatsrcmainresourcesSending configstomcatsrcplanSending configsAdding configssrcmainAdding configssrcmainresourcesAdding configssrcmainresourcesAdding configssrcmainresourcesTransmitting file dataCommitted revision +Please let me know the Git URL +The following changes have been made has been renamed and made an interface has been renamed and now implements the interfaceThis change eliminates the weirdness we had around the relationship between and and also results in having to create one less object per IDM session +I have heard that one of Portal guys had similar problem +An instance of the master keeps track of time for each complete FATE op that the client has not finalized +I refactored in to two separate test cases as and +Looks like an issue of EclipseWTP +Thx Hoss Ill upload another patch +Generally looks fine to meOne thing thoughYou have to add the license text on the top of the created java files +I found the problem +Ok should we delete used by index or let them be +I wanted to know the exact scenario when this may happen so that I can add a testcase to verify +I was able to manually unzip it +I will incorporate it and add more tests + takes care of the problems this patch resolved this. +Well lets see what we get now that has been committed as well +I think warning here will be a pretty good option but as you think the best +I have no opinion on which are better as both are equivalently expressiveIm on the current patch and would not object if the indices were adjusted to fit the former semantics +Note According spec this feature should be implemented in the future version +Then lets open this under a more apporpriately titled issue +The goal for the patch is to enable a faster cross project dev cycles +Thank Adam and Claus for the work +Uploaded to jboss +I have some codeI need to test that implements it +Adds documentation for commands and overview in HDFS user guide +This commit should be included into for eXo Platform. +I dont think it was a mistake in case of Hmm +What would be the point of setting the system properties in the VM that is not the one used by the tests +Fixed string declaration +Thanks for review Jon +It does however result in the masterUrl parameter being consistently used on every call to fetchindex +So if is a local directory what properties use this path to create an HDFS directory +killing this old one. +When I tried to remove lib it successed and now Its validating server classpath isnt now any +Documentation looks good +Thanks + +It may be because dollar signs +patch that removes initial from all calls to getPath including the one where the test breaks +Allowing an individual Upload component to override the global but this has issues as well because of the nature of the way that all have meaning to the stuffMaybe some more thought needs to be put into this one +Now that work is underway with AS all previous community releases are +SaschaTake your time and keep up the good work goingThanks +So projects like wiquery databinder and spring can with less hassle +then we can cat gzip file like dfs cat does not handle bytes correctly +This fails to connect as builtindba but sucessfully connects as fred +uses separate set for userinfo and fragment now behaves the same as URI for unescaped fields and generates the same escaped in most cases apart from arbitrary queries +Patch that adds the variables MXJADDRESS and MXJPORT to conf make configuration for mxj obvious +Because i saw the bug on the demo system but not in my local instance +ws +I just committed this. +TO be fixed in and and tested +Paul KingAny hints for Chris +I am now just waiting for Maven Central to come online so I can test this +Should be fixed in hibernate core +Ran in test +Im closing this one +Sure this patch applies against trunk and the modified test passes. +I encountered this only during the removal of the node and then saving the state. +Where is the problem? bug in HSQLDB +And made all the SQL keywords upper case for consistency +Worked! Thanks! Im going to read Thrift Scalable Services Implementation paper before I start converting into just nits remove +Will also require that the following are built using new layoutmain runjarserver t this moving to the next point release +Also would it be possible to add the following tests +was about urls in camel web console +Fixed in rev +Ill be back! +import +Looks like related issues +So its no longer there +Fixed in r +I added a test case for this and in trunk it seems to be fixed +rd update after svn plugins directory was moved into new may not be the place but can anyone tell me how to use this functionality? +But as long as its not letting bad strings to go its okay +We seem to have come to agreement that we need to support in addition to + has tightened query parameter validationRelated test cases have addressed similar isue +Grails Metadata fileThu Jan GMT the moment the security interceptor logic is commented because is not used on the current version +bulk defer open issues to cant reproduce this bug on with a development build and the source code from the description +Thanks dhruba +Did the Kafka brokers come back after a certain amount of time or were they gone forever +Patch applied. +We dont actually use them because we use guards instead +The way to go is something like pig or hive if Java is not an option +Patch that adds a close method to +We can still introduce the metadata though but well have to come up with an alternative mechanism +I will have to investigate this further but passes the functional tests therefore I would try to run that +bundleplugin it looks like this feature was already added by thanks Hiram! +Commented in in +here is an updated patch with the test +Sample war which demonstrates this problem +Added doc bug which needs information after verification of this bug +utilincludeplatformsunix has a USEUTL compile flags that can be used to enable Libtool loading functions +This could be raised as an AS issue perhaps +Will commit once the tests pass +Also FWIW the error in the case had the unexpected jobqueue status being STATUSPENDING +Hi Eugene I will need an example POM to debug this issue as I have not been able to recreate this locally +bq +The local catalog is not preserving a catalog info we could consider holding the information we care about members member title icon description boundsAnd then issuing a REPLACE event when the real information is available +The Import and Stylesheet classesalso had to be updated to set import precedence properly +Duplicates +Hi FernandoThanks for the contribution +Thanks for HELPZeb are facing a similar problem while loading collections from +The above stacktrace should not occur in the above scenario +Moving this to +Thank you for the comments +CLHM includes an manifest if that helps +Tests passed locally but I dont really think there is any test testing that +Russian ln bundle and flag image are attached +Thanks Doug that makes sense! So it looks like this is a safe change +All MRI +For some reason the apikey is coming up as blank in the and its returning forbidden +Id be ok with changing the element to be named rather than port +Velocity like Java wont do automatic casting between integers and floatsIf you are looking for a solution using a tool that takes an int and returns a float should do it for you +If it works with trunk Im inclined to ignore it +So we do not know what type this is and have to determine this by calling get and then get if the first call returns null +Possible applied the patchPlease check and reopen if need bepaul +Nice to see you back with a commitI have backported the fix to the branch. +Okay I can work on that too +trying to add features to it and maintain the patch fresh is too unwieldy +This is way too specific for. +Ill take a look +So far beside JDBM we are able to manage LDIF backend In Memory Oracle and +New version with fixed uploaded. +Is there an alternative or a compiled working jar available for download somewhere? +Resolved in SVN +bq +The demo isnt required +fixes all to in the description +So if we design it as a common module we can improve it later if neededWhat do you think +Thanks Remus! +Im no longer using faceted projects they dont fit my use case. +Mark I think we should commit this for +I have fixed it on both and nowAlso added it to the knows problem on the release notesPlease test it on your end +was some bad code in +Whoops ships passingbq +sorry for the late reply I were on the chinese Qingming festivalok I give some examples +hmmm +We need to move towards beans and injections as much as possible +We should probably somehow escape the submitted value before logging by Metamer to the page because that is what is failing. +I made it into new Brainstorming issue type. +I think this behavior would make the framework a little easier to use +Also draggingdropping should result in triggering a move refactoring +simple change the status to Patch Available if this is ready for review +Please have a look at the patch I attached and vote for or against it +No you cannot try to further avoid the negotiate +It should stay predictable. +Patch looks good +Two Things +As described in please attach or describe the AST transforms that this project is using +disabled the test in trunk if base port is setsibmsvn commitSending javatestingorgapachederbyTestingfunctionTeststestsjdbcapiSending javatestingorgapachederbyTestingjunitTransmitting file dataCommitted revision +The connection profile will not be usable by any other project in Dev Studio that Teiid Designer +New iteration of the changes +Todd yes I am using it +Basically the script referred to in the upstream bug should probably not concatenate JAVACMDOPTS to JAVACMD +Are you saying that It can only be done if requested by its project members? Thanks! Anita +I unset HBASEHOME and HADOOPHOME and the tests pass +Yes I meant +What about simply specifying rollbackFor? Or do you really intend to not trigger a rollback for runtime exceptions? I was going after was to only specify the exception types we want developers to throw from the business layer of our application and we are trying to enforce rollback only in the cases where we throw one of our exception types +This change actually conforms to what is currently available on document +I would have said the standard place would be the class but the class seems to expose a public constructor so we need to change it there as well +I tried to prettify log messages and it looks like my prettification bled into directory making when splitting logs +This patch add and to a good EricYour patch is in trunk at r at r +applied the patch r +The only difference between add and is the operation name +What do others thinkOn the patch itself it would be better to access these via accessor methods rather than protected fields +Seems like we want to keep this as it? Should we close this +For the design discussions and implementation feedback in keeping with how its done on most in Apache lets keep it focused on the JIRA +This happens with galileo as well as with ganymede +jar file with the changes to to protect against NPE with mouse locations +This is an updated patch +Patch for +Attaching an updated patch synced with the latest codeline +I reran staleplans and it passed +the difference between cleanup and destroy +Other than that this looks much better than the previous messageThanks a ton +Thanks +Theres some performance penalty in looking through a larger buffer but that shouldnt be too much of a problem as long as we dont make the sniffing code do too many passes over the data +The patch for filter aggregation which hides extension point etc +I as resolved as noted in earlier comments the core functionality required by this issue has been implemented and the remaining work will be done in the JOPR plugin for Teiid. +After the binding calls on the operands bindExpression in checks if the right operand is a parameter and if so it calls setDescriptor method on it to set its type to the type of the left operand +I need to investigate on this issue a little bit more +applied thanks much for finding this little fixed up the other locations you identified +Like the autotools build we grab the library version from the script +Probably easy to workaround this by implementing a subset of the functionality at the X level so not a blocker +Issue has been resolved for a long time waiting for potential confirmation from the reporter +just changed title and subject properties +I committed the patch to +Whats the fix? ACK when all messages have been received +Resolving +As there is no bundle available for version I think the easiest is to revert my changes in the content loader which I just did +Committed to trunk with revision Closing issue +I had a look at the wording in the standard and I agree with your interpretation +I just committed this +just the performance tests which Im doing now +Ive been testing newest ehcache where was needed as i remeber but using version it can be optional as it was before +I think people are under valuing this request as a useful tool in production code +Appreciate if anyone can clarify the type of applications that could benefit from this +Generate Postscript instead of PDF and stream that to the Printer using problem is there is no lpr in windows +This was commented out as it caused the test to hang intermittently +Closing since changes have appeared in Latest Alpha Manuals. +Oh +Thanks! +Of course it was a copy and paste error on my part +In rev +will try out the IT thanks +Checked in a fix which seems to address this +A couple commentscdevcsecurenativeautho First bullet While youre in there I noticed something which could be clarified now that NATIVE authentication is turned on by storing the DBOs credentials +Bad patch last time my apologiesBumped the version on this one to to avoid confusion +Reopening +The JDK does not allow opening a directory as a normal file so we cant provide this functionality. +Issue will be closed after successfully nightly testing and merging the regression test in branch. +I wasnt implying that we try to usurp BND from Peter +Thanks +So this clearly is a distinct issue +error message in the Apache log is caused by the Basic header being forwarded to AS management which is not supported so the server throws an Exception. +Given this I am at a loss as to what we might inform our users ofFurther discussion of should occur on that JIRA item. +Closing tickets +Seems this only happens with gmail +Thanks a lot I will forward it to the exjello team and try to patch it temporarily till they fix the issue +Updated the description to reflect the use of setters to pass configuration properties into the implementation +IIRC that test doesnt run during the CI build +A patch is attached +In fact both cores were damn near +Should I try an even older version +This would seem like a bug in the ESB detector to me and it seems that returning an esb runtime itself is actually a workaround this bug +I think youve done a great job disentangling all the various suggestions Bryan +I think it is not an issueThe problem is that the return type isJAXB provider does not understand textplain payloads so it can not unmarshall it does not know how todeal with populatingI believe the solution is to register a custom with jaxrsclient and set its consume to textplain textxmlThat will do try it please +Demonstrates that without the decorator enabled the unsatisfied dependency exception does not occur +Jon if you dont mind can you please update the docs as well? Thanks +also uses it to store the state value +The new build did not retrieve new artifacts so it seems that the access to the local cache has the same problems as described for publish to the repository +Attaching a repro for this bug script Bash script to drive the repro +Moved to upon release of +isolated change to where config is parsed + +Did you build from source code or downloaded a published snapshot +The structure looks good but the tests need more dependenciesCurrently in maven all tests will failI will create a patch for the maven poms that includes the dependencies for the maven test lifecycle +Javadocs JXR and PMD reports are hidden +Can someone confirm whether the Hadoop jar file will indeed contain the term dev in the release version of? hadoop +dtd files stored on disk but I can store dtd files onhttp +To have a one option instead of bunch of and lot of defines seems good for me +Id looked at wrapping the but decided it was likely to be much too much workIn terms of the temporary files we know that usually the img tag will appear just before the nested parser files but I was hesitant to assume that as a rule +Thanks +Matt if you dont mind Id like to merge this into since its been well +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +Since the is bypassed only file level permissions can provide protection +You can configure log level using JUL configuration. +Fix for epel repo +It replaces any prefix that appears in the expression with the corresponding URI causing itself later problems if that URI begins with something that looks like some other token in this case +Confirmed at least passes again +Andrew have you got an updated implementation to review +The class would be able to initialize and clean them up once in a fully synchronized enviroment cache them in a LRU map and properly manage their disposal when the resource is no more neededOne thing that worries me about using the LRU map is that nothing prevents us from disposing a schema that is actively used in a current request Im wondering if using a soft or weak cache might be better the way they are setup makes it impossible for the reference to expire if someone else has a hard reference to the same object +I think the correct behavior should be to match the usage +I have a sample implementation if you need itAre you saying you have the code modifications to do what you want? If so and you are willing to select the option to grant license to ASF please do attach a patch file to this issueRegardsm attaching two files that are both valid xmi +There should probably be a JIRA opened against Java implementation. +Would an unchanged client still be able to talk to a server with PB headers? Or would we backport the PB header stuff to? Or something else +. +Changed to a feature you are describing is an + +This patch is simply updating mock data that is used by Ambari Web to reflect the latest format returned by the API +When we respect both things why do we need the wrapper at allAh because query must extend That solves my problemSo the rewrite methods can only be used for real how about moving those rewrite methods out of MTQ? I know then we must make some classes more public +The same error on my envAnyone looked into this +I understand your concern with finalizers being the flawed thing they are +I try the same query without subquery and the result is order +Done +Ok got it now agree with Z +groups +Was checkpointing enabled on this slave +Dont think that the example is relevant +Or make it a protected method +I see that some things have a corresponding Attribute but others dont +AlejandroHeres a patch and also a generated new site HTML for xref. +RTC task case is not working yet but otherwise there +Patch with wiki page with documentation +stack yes the problem was occurring before the restart +Please correct Fix Version to one without x like depending on where it was wasnt fixed +I also havent seen anything suggesting it has to do with the META tableAny ideas on what would cause a split to fail and retry? Is it more likely caused by some internal regionserver problem or the region being moved during the split etc +Thanks SaschaThis makes sense indeed your patch is in trunk at r +However in the general case I guess including the endpoint id as well is a good idea +From looking at the masters it appears that failover is attempted before the master has actually been able to establish its master role by talking to the slave +Python is an example of what Jim is saying here +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Perhaps we should also upload as part of our release process? It will become more relevant over time +Great thanks for doing the investigations +Hi Koji Noguchi minor thing calling toString on Path might be redundantThanks Prashant +Ilya what do you think +So are you suggesting that we default the scope to Global +But at least the class data will be consistent across threadsThe parent classes arent either and have data publication problems. +Because we are using a prepared statement here to perform the creation of the table this obviously gets transformed into a stored proc on the SQL Server +Patch for review +done David can you close it +In other words a node may be totally fine running more tasks if the box is actually being hammered +SDR only checks if the current entity belongs in the list of exportedID entities +This should provide a better mechanism for adding dynamic panels +I think there might be another way of setting the timeout without using a custom socket factory +This is big enough so Ill put up a vote for release at the end of the week barring any suprises +I have yet to figure out an effective way to test changesHow about you make a server implementation that intentionally waits longer than the clientside timeout +But the related views or xhtml files for that flow is never loaded and it fails with context classpath resource is required exception +Aside from writing my own custom implementation any recommendation on a framework to use that will support and is similar to enunciate +should not extend the current implementations copies all the triples of the wrapped graph in memory As I already commented on May as writing generally goes together with reading I dont see why we should have a for reading and for writing +Closing issues that I resolved in already released versions +Verified on EAP checked in +Changes merged to the branch +change to improve the codes readability +This looks good +Compiling with works fine +If it is we execute the function otherwise it skipsThis solution should solve all javascript dependent controls not just Calendar +I understand that you are talking about a long term implementation but I could use today as well as many other users a mobile version of the doc on my tablet +That wouldnt be hardSince theres been no motion on this Ill have a look to see if I can do that quickly and easily +Verified on work properly. +OK Im calling this done +Wed be happy to try it out in the wild once its available +Committed in r on trunk and r on stable. +as I said before this shouldnt have been holding this patch +Otherwise the patch will be quite involved again +I think we need a thread in that method tooI agree this is a similar case +Sanjay for the reviewI have committed this +Thanks a lot for finding it out Ted +Do you remember the steps that you took to get problem? This would help me reproduce +This diff was run in the plugin minor changes to the page and quite a few changes to the WAR Bundle example page +After looking through the code some more and seeing the attempt ids likeattemptmThere are problems + +Thanks Suresh! Will verify and publish +Also I noticed that file is copied to several places in the repository +How those services are configured or deployed would depend on the use case for each +Good point we could allow configuring the method names for classes that return the event property values rather then automatically deduce them from the a workaround one could of course wrap the legacy class if feasible with a minimal POJO that acts as an event container +is considered as private in Spring already and will probably be officially deprecated in Spring + +The files in srcmainjavadoc do get copied to the build output directory dont theyNo these files are only copied to the report output directory +Address comments by Hudson +Committed. +Ive applied your patch in SVN trunkIll try to merge it in the branch tomorrowThanks a lot for the contribution! your patch into branch +But anyway I see that in both examples the Validator is called to validate each inserted string so works as expected +Ive been having problems with this as problem is that in the file theres lots of public methods for constructing projects +Committed patch +These changes conflict +I dont see anything more coming up for this issue. +The final concated file wont be in the previously created snapshots +The file is the xfdf file and the file sample +The latest AMI build is +I didnt find any JAVALIBRARYPATH in trunk so Im guessing its been obsoleted +Closing resolved issues for already released versions +In certain other deployments we need to obtain the service url by writing custom code +Whereas what we want is to filter to come up to and dont get pushed beyond +Replace O with N for stat and acl objects passed to Py +This has now been committed to trunk +Patch trunk. +oleg I dont agree with you but it is good to have different ideas +I used the type for default attribute in the binding think the other side of the same issue +Its not in the public git tree yet but it will be +Integrated v into trunk and +This jira will be used for removing the back slash based path support once the required testing on Windows is done +Is this a new guideline? I found it hard to debug tests like these and always ended up removing the timeout what seemed a bit odd +Thank you for the description of the option but I need more to understand what you do +We need a place in apache infrastructure to store the template before we can actually modify the tables +Thanks +Im attaching proposed solution +doesnt work correctly on Windows +Im sorry for this delay I think that we only need to add all the needed dependencies to run these integration tests dependencies Lucene dependenciesI hope during these days to finish the implementation of integration tests +Thanks! +This has been fixed verified by running the coreWeb misc test on the specified environment. +New rev number +Ok lets wait till the autodetection issue is resolved +From John G to Brian do you have suggestions on what these dialogs should beFor the Properties dialog title it seems a bit odd to put Guvnor Resource Properties in the dialog title since its already the first label on the dialog the Error perhaps simply Guvnor Repository Error instead of Error +This allows us to incorporate any wire changes freely +And I guess meta is not there anymore as were closingMay be we should not stop the regionserver holding meta if a split is in progressLet see first if disabling the split helps +Another thing to clarify everyone can reproduce these bugs now even on Linux +Setting rd party cookies might not be desirable for all containers +verified +Youre right I can actually add other folks to the watch list not just myself +Updated cluster test revision implements distribution of urls to clients instead of rnd +This would not be true if they were all under the same hierarchyHowever if we keep both the implementations and interfaces in the same package logically and visually these are going to be equivalent for the developer +I see this is about OCM +It really should have a separate method abstracted out for merging the flags but since the old harness is hopefully dead code I didnt bother +use a object only when initializing the +adds lseek to and the lseektoavoid +Thanks Jeffrey! +committed to CVS HEAD am perfectly aware that the proposed fix is not pretty +I know this isnt quite the right place to ask but is there a timeframe for releasing? +I would love to have it right now for storm too +complexity of this should be compared with what is required for Or perhaps this can be used as a template for the generic version +Thank you Jing. +Hi Michael Im afraid I cant share the jar source due to IP stuff with my employer +is it safe for +I suggest to have a escape quotes setting in de the decision table +Patch to add a loop as described and use it in conjunction with +Didnt know about this practice +I dont know what we can do in terms of logging what actually does +However there is another being tracked in +Not fully tested yet but xalan passes with it +Can you please suggest a query to create a unit test for this? to hive I am willing to bet we only hit the block merge if the input is larger then hdfs blocks size or the defaults is high and it normally does not happen during the course of the unit tests +Im not actually working on it just considering options and only intended to attack this on th side +My problem was with static inner classes +bq +May somebody pls try this +Consolidated patch compatible with trunk after creation in tests +Any idea if the related shrinkwrap bug is solved already +I appears that throwing an exception tires to set ! and in turn when SAFE is set it again tries to throw this is the cause of recursion here +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Nicholas should this be ported to too? Leaving open for now +I actually dont think we should make the writeTo method return a long since we use byte ints in the protocol to delimit size of request so we really cant take a buffer +the anchor tags werent working in Safari because the margins were still negatively indented as a technique for making the main and right col align +We brought this up because we actually saw this in our live clusters +not in the framework +Could it cause any problem at the client because we are terminating the lease earlierbq +This patch checks the cause of the failure when setting up a RPC client tries to connect to a RPC server +To repro use a Beehive SVN environment and be sure to adjust the Ant variable in sf to point to your beehivetrunk directory +Overall I wonder if there could be better approach +Will we support any stripe type other than count basedsize based? If so probably we need to change how stripe type is configured since it seems that we can support only two types now +We can dodge those errors but sooner or later somebody will trip over them anyway +We will continue working on it for Camel and beyond +Admittedly the doesnt support that either but I had an idea for how to do that and I cant easily see how thatd fit in with parsersIll apply your patch shortly then carry on with my work on making the office format embeded resources available but using your new rename makes sense to me Ive done thatNot sure about the best way to do the strategy object but hopefully someone else will doRecursion was broken in Ive this but Im not sure if its the best way to handle it or not? See r for how Ive made it work for nowNext up images in word +Seems to me that each regionservers are asking for the same lease +Id talked to Rich a bit while about this before he filed it +I can understand how valuable it will be in a Tech Support situation of a product that embeds another product that embeds another product that embeds Derby to have sysinfo with additional jvm data show up in without having to figure out where to set a propertyBut I do think if someone has already set true and is used to doing a lot of starting and shutting down the could now grow quite large quickly +Have we been able to verify this with an Apache SMX kit as wel because all the information in the issue is about the kits +This patch was in flight for a while will push this in once Jenkins blesses for one last time +Funny thing is that with latest hadoop from SVN testGenerator worksAlso no matter where I put this property test failed with again tested with this propery test passedIll try my url list if this was the case +Do we really maintain version any longerRegardsLars +During the last changes callbacks for and event were missedIve just added them +The goal is for the entire replication to use persistent connection +I updated the testcase patch to include a check for spaces in the generated URL +Because I added s to Server +So the have acknowledged the problem but say they wont fix it because theyre to busy +However there are many places where displaying INFO messages on the console also makes sense again especially in the local mode +Merged +If an authentication filter is installed it will block hadoop daemonlog +Forgot one comment +reopening since need to commit to on +idx at the same time when starting replaying journal +There is still one thing that needs to be reviewed the lock is acquired before the decoding is done and released after the message has been processed +Looks like under clang is not defined +Please fix this bug it makes real customization of appearance impossible +One test from the javadocWith lenient interpretation a date such as February will be treated as being equivalent to the st day after February +However I did encounter another exception which only manifests itself when a timer is created while another timer is executing +Closed after release. + Bug has been marked as a duplicate of this bug +Removed and from need list +No this is not planned for backporting Spring s AOP capabilities will remain this also applies to the upcoming release +bq +In that sense yes we dont need this separation but will do that anyways so we may as well do it herebq +Committed Clements patch although I altered it slightly to use the setting as the repository prefix as this is more consistent with the other goals +JBDS Project file containing the example Web Bill +The methods you have mentioned are being used by the builders which are in a separate package +please reattach these as a patch +In this way store just maintains the invisible columns as normalcolumns and no special support is necessary drop column support is already available +This is actually expected behavior as long as the application is nevertheless started up this occasion Ive refined s get to log the message instead of the full stacktrace +Ivan please check that it has been applied as expected +Excellent documentation is now updated with this information as well. +Fixed +On the main trunk Committed revision On the branch Committed revision . +i did move it in subversion and updated the scm connection releaseperform goal fails while it tries to find a dependency it is supposed to mentioned workaround in list mvn releaseprepare clean installsolves this issue so i can release again +Ah so the motivating is authentication +This concentrates handling of REFERNCE values in a single place +Applied svn diff on launching project +Please provide screenshots as well if necessary +So Ive deployed a nd datasource called and it was discovered properly +Just tried what I pulled off the trunk from gitI meant just changing your to use as the version for Web Flow artifacts +i havent had a chance to look over all your changes to trunk workflow but is there really task queues? or is Queue Manager a queue of workflows +Looking at your benchmarks I think the number will go down to for in general because of overhead and since not all task reads are local +Just needs to be deployed +Given the low priority pushing to +run on Licene Corpus creation took msSpan queries took msTerm queries took ms +A text change listener would introduce a lot of events being fired +To fix this issue we need to replace XR with new one and check it by running a x build +It is the responsibility of the portal toproperly namespace these extra parameters to avoid name collisions with parameters theportlets define +Hi ItouI believe this is a duplicate of +The only thing is that no endpoint is added with an actual wsdlsoapaddress +Is this JIRA about extending the event mechanism introduced with or implementing a completely new one? If the latter is the case is it supposed to replace or coexist with it? Id like to see only a single event API in Wicket +Greyscale version of my logo contest entry +People can always put the flag back if they want. +I could have sworn that I had reported the exact same issue +The TBODY end tag may always be safely omitted +I dont know actually we have an issue about the Alfresco REST ClientIll let you know soon any update on this +Committed to trunk +mark the node dead does what it says no more no less +Agreed with Stu +Patch looks good +Four Threads running and eight Packets in queue +Issue closed with release. +Changes look fine to me +Or maybe this is the problem requires bytecode instrumentationHow to do this +The attached patch also forces autocompletion and DESCRIBE to use +Since there were no objections on the list fixed failures in modulesawtsrcmainnativeglwindows at rI have rerun the tool against the latest source and I see no new failures +I have also conformed the unsafe isnt statically linked to native Malloc by adding a printf in the malloc c code which basically counts the number of times it is called + +Ok nice to know +This ticket has main points +I think it should beNOTE The ESB Management Console has been reconfigured but must be for the configuration changes to take sure about the Tech Preview part +This seems to work now +I applied this patch and ran derbyall and it it passed but I was not entirely sure about the permissions changes at this late hour +I now see the problem resulting in the findbugs warning +other ASF products +If we are there is a lot more work to get done before the work done in this patch +I suspect it is because the initializes itself as Map type +So the first fix was to use +above test failure seems transientunrelated +Time to implement it I guess +Fixed in trunk . +Hi Ramesh unfortunatly I cant give you a date for this issue +I did one or two refactoring at the same time +This issue can be closed +Attachment has been added with description cyrillic glyphs has been added with description corrected version of const inthanks for the patch +Is there a chance you can try a newer version or current upstream ones? or master +Already done. +Patch applied to svn revision . +It worked as below shows the matrix how the driver versions work +I agree that it shouldnt be inIf you look at the zip file I have it in a folder and perhaps we could borrow from Wicketpivot Pivot Core WTK WTKX Pivot Extensions that are direct extensions of WTK and WTKX under the same meme more panes widgets skins +One other note +I think it has to do with text like overunderthrough and handling of as a breaking char somehow conflicts with your workI am also uncomfortable with the triple breaking being done First FOP natively breaks text into chunks based on sp nbsp lf and the like +The only alternative I can see is to disable keep alive on the connections so that the output is not chunked. +A workaround doesnt helpThere need to be an implementation independent resolution +Vladimir patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Patch to simply clamp the progress value to before we try to update the countersNeeds a test +Woops sorry I did indeed see your new patch and applied it but then failed to svn add +In fact could be wrong there didnt experiment much with that +I was hoping that these differences would be resolved within the camel community either on this issue or within the developer list +The same for that also have this dependency we can loose +However I did not find anywhere in the specification information about the namespace of the type attribute +After getting the confirmation from Panagiotis Vretanos that the schema for WFS really is not correct I made a change request at the tried to verify my suggested correction in by changing the file in the jar archive of which only partly worked because it is not possible to change the base type from to string +Updating fixed version since Groovy has been removed from the scripting demo +I was asking why the class javadoc only talk aboutreceiving messages from the slave which you confirmed meant that theclass is only to be used by the master +I think we just need to add an overload that accepts a +Im concerned about changing the default behavior especially without a mechanism to get the old behavior +Brett any chances of supplying me with a unified patch +We can either use the proposed patch or fix the problem in Xerces +After your comment I realized that my coworker never tried to run ant in that screenshot +Not before Wed PM but I can look at it then if you need +So basically the burden on removing duplicates is passed on to the caller? How does the caller know that it has them all +I defer this new feature to the next release +make the scanner ignore a URL the modify the content then eventually resume scanning with the option to consider the deployment or not +hadoop uses mds files as well which are very difficult to validate but I guess they do that to work around the rsync exclusion +Thanks Isuru for the contribution. +Not so easy because of +Thanks Mark +Of course this can always be found and reopened by the contributors or anyone that wants to revisit it. +can we proceed without it we will update it in our cvs when it get bundled in any of the binaries which just bundle? Is it used in the unix shell installers +Im trying the test wo the patch and nothing happens +A new will be created each time a Step is run similar to +I end up with a patch against things like src in the root directory which doesnt exist nowFrom manually scanning the patch I have two minor questions already has methods to close Closeables surely we want to reuse thatAnd we are bringing in for its eigen decomposition implementation is that duplicative of whats in Mahout + to agree with Mikhail that we should follow exception throwing order +Itshould have been able to load them in a single query +I did not mean to fix this myself +I believe the patch I attached last is sufficient for this task +We use a String as a uuid exactly as you did in your person class. +Idont understand what that means +Thanks DavidVery warm welcoming from your sideFor a long time i have been considering diving into the interpreter i believe that its a really awesome projectHopefully i will find some time for that +Updated susbsytem to use into the master branch +changed all methods +Charles I tried the patch against both my reduced test case and a real world app and it works for me +What are the problems +can i close this jire without having it implemented +This is normal +Using a simpler fix in patch V aborting flush if server is stopped +I think all EAP bits need to rebuild anyway +Ive got about developers mostly junior using this eclipse config and so Ill get a fresh as a last resort +And the actual uri is the attribute +Otherwise I havent changed anything from the patch you looked at +I agree that the wrong solver property is being used +Again the name is temporary +No one has done anything +Just make sure that the XSLT template have good indentation that will cover of the problem right +xcf files? Unfortunately we dont use gimp +If I hadnt found this bug with the correct information I could not have set up the server in a path other than webroot +Devdeep Can you look at this +Has anyone experienced this before +Oops noticed one more in there +At least thats the best I can think of regarding this issue +not for Tools +Otherwise this is likely fallout from and the test needs to be updated so that the Java method signature matches the SQL routine declaration according to the ANSIISO rules +Please verify +Assigning back to Jean as I was unable to commit the code changes +Petter do you consider to work on a patch +This would be really good to have the errors right now are totally cryptic +This bug is a regression that was intentionally introduced by the fix for as a necessary evil to prevent return of wrong results for some queries +My understanding is that each managed ledger has a single writer but can have many cursors +What I am saying is that I think we cover that case +users can disable jsp validation for this project or exclude specific files +Ill report this via WC bugzilla for now +Pushed to branch and with and OSX +We dont allow true concurrent access we just holdqueue requests while somebody is using an instanceinstead of throwing an exception +BrianThank you for the prompt fix much appreciated +the and seems duplicated to me +b Remove trailing whitespace +Ill need a full testcase +It works on Tomcat and Resin but not is that correct? You can try stepping through with a debugger +garbage collection implementation for the global data garbage collection implementation for the global data and refactoring improved the also prevent the to load the binary property automatically when you get a node? I have posted this a few weeks ago in the mailinglist an jukka told me that this problem will hopefullybe fixed with the +how do you intend to use itWe are planing to use this on +Please cc me by email for updates on this issue just to err on the safe side +No because the offending repository declaration might actually not be relevant to the project +Making elegent and strict following the Sun standard is a right practise suggestion +committed. +The attached file containing the query result has a double asterisk in the first column for the other discrepanciesI had to modify the query a bit to get all the output into the third columnIll file a patch +No there is a different issue +To fix this issue we need to do a major change in the way we implement the wsdl generation logic +Consumer redelivery happens on the client side until the consumer is closed without acking the messages or committing the TX +And we are using hibernate +I will do more tests and report later +Nice work Ate! I dont have outstanding changesI still dont get the difference between and +The does not make the correct calls to create all the necessary objects +Thanks Shai +Dart can you look at this for trunk +Do we have data to confirm that intermediate compression is only useful for reducing network traffic? It seems were also reducing disk IO which can be a bottleneck especially when the coredisk ratio is high +Are you sure the steps include selecting a Web project +no problem Ill take a look at your patch +Thx updated +But this is have been able to port some of the tests to oracle without hardcoding values +As far as I know the Package class wont even return you a package if it had not been loaded previously so its not really suited for inspecting packages anyway +DBRE can now cope with database refactorings quite the all the large enterprises I have worked in the development schema when requirements and design are done is replicated through test and production +We could leave it out and have something like and as long as you dont think it would be confusing +Will be waiting for the release because we are stuck +Done? +For Teiid behind JAAS is sufficient once a user friendly mechanism is provided for specifying the data entitlementsXACML was what I originally had in mind but it bears more investigation before pursuing +I havent looked at it in a long time but I concluded that it wouldnt do any harm +I added a capability in file to skip the disabled option +It works with but after applying the patch it fails +Patch looks good +Reviewed by Mike +Something is off still fails the final rollback it does causes a to be left behind when we try and close a directory + requires to upload the image to a location accessible by HTTP after this it can be downloaded by the platform and a template can be created +Heiko Specifically what would you like me to move? All tests under EJB webservices +Moving back to for now +Havent tried any container besides OCJ so in theory this could be OCJ issue +Actually there was a confusion between Dev Staging +A couple of small comments posted on review board +thats why I posted it +Where do we go about getting this from? This looks like the last issue required to close should the DOAP be OK +The key issue is that you try to persist the properties before a Save is issued on the +Somebody may already be looking at such a for what it the most appropriate default setting +attaching again +Bjrn Gustavsson from the OTP Team replies We will not include the suggested patch in OTP +Please take a look at the code committed in SVN to see if it addresses all your issuesI have tried your example above and it works for me on the latest code +It will be nice if RickDaganyone else can also review it +Tim thanks for explaining! I was just checking svn annotate to understand what happened +Marking for so that it can tracked accordingly. +thanks for your reply f we hava a plan upgrade the hadoop to next monthbut there now we need to do some test the version on the production is modify base oni think version will last at least for a monthbefore this time the server work very well more then yearsnow i can do is monitor the jobTracker logwhen found some datanode was crashthen add this datanode to hostexcludecan u give me some idea or why it happenedneed i turn down the socket timeoutfail time? just noticed an old issue which appears to be the same thing +Fixed by a Fedora repository update +Thanks Leif! I agree that it makes sense not to bring this into. +Agreed that it is problematic if the test should fail +Any new crashes? Seriously this is a shot in the dark +Please verify +. +It is not in SOA to be raised through Red Hat support. +minor modification to move srctestorg +Bulk move of M to feature seems not to work properly +This issue is not really replaced by +ChrisIf its not too much trouble could you post your improved file on the dev mailing list? ThanksI tried to grab it before Jacques removed it but I was too late +bq +Statistics are needed when tables have a large number of rows and it does take time and resources to recreate them +anillocalhostsecuritysrcmainjavaorgjbosssecurityintegration svn ci m handle the NPE Sending integration file data +BTW TC is using a separate do you know? Ive just looked at the mapping file in srctestsjdo and Icannot find any reference to being set to true +It wont be done in +trunk +The performance and benchmark tests should run in an automated fashion as part of an ant task +This was due to some local references to and commons logging in the tagdoc plugin +I dont know what the licensing issueswould be but its likely to be a nightmareSo back to you Daniel +The second one fails utilizing batch write for the same input keys +No Tomahawk issues have been reported by the community since the change to +Thought I would not have looked for a mailing list under contact us +No +Does it work for you now +Ill let you handle the meta ticket +If the create block succeeds but response is lost due to NN crash next RPC from retry will fail +Thats why I included the whole text in a comment above +This is basically a back port of the changes made in truck via with one additional fix to forThanksHeath +Patch attached that updates the Build Install page in the main site to include information about the Maven artifact +May be once you restart try to poll for the admin connection until one is successfully made +Attached take file as diffs along with updated regression test files +Consequently perhaps a listing of the plugins and features Im using within Eclipse would help +As far as Ive seen even the codehaus continuous integration build ignores I appreciate the work the others have done here +Please change back if you feel otherwise +Unfortunately ProjectClean does not fix the problem +Is this a duplicate of? Or alteast forward compatibleLots of ideas floated around there itd be good if these two issues are in agreement +is more about to load the filter class dynamically +Hi Jorgif you could provide a patch I would be pleased to applyMany thanks in advance +I investigated this JIRA in detail + +Adds toString method to and logs the errors in the configuration file after the parses it +fixed at with the release of. +This is working ok for me +But there is another problem in Axis and withSo from my point of view there is no version we are able to use +I thought the idea was to introduce connection pooling with minimal fuzz +The JIRA title is formed as a question currentlyI think we should use more descriptive formRelease notes are needed as well +I was asked to create a new bug in the conversation about +Mmm Ive seen the assembly plugin I could reuse that somehow to define the do you think +It should be a simple config tweak and it will save us and our users a huge amount of really is a MAJOR issue for us and our users. +Rather than embedding meta to the list containing the counts perhaps we should bite the bullet and add an additional level for the countsyeah +Which parts have already been applied and which parts have you decide not toapply Rory? The business about numeric date formats can handled by the new system +I made the initial changes but Stuart is the one who actually made it work the Arq stuff is just over my head I guessThanks Stuart! +There are tens maybe up to a few hundred messages a second +Attached sloccount output for the stdcxx snapshot +Updated patch accordingly +Mind rebasing +Same as BK +Is this one good now +ok! thank you. +houtputStylesheet and houtputScript are also available with the attributes name and library +I was always approaching the issue from the perspective of two acceptable types with the same quality factor but differing specificity since that was my problem case +It occurs when the properties or are encrypted on server side +Duplicate of + +A first version to get post put delete as well as getFeed was implemented + +after release +Since the confusion with the JPA API provided by geronimo is resolved there is no reason to change the dependency. +can you please provide more details the description is to general setup likes was already made times without a problemwe need more detail here how does the cfstoredproc looks likehow does the storedproc looks like +Patch applied at +Done at rev +svn +This is expected behavior of HDFS but while this is happening utilities like and fsck produce high numbers of underreplicated blocks and the node is not on the deaddecommissioned nodes list +It is antipode issue to +Im using changes have the same problem with changes goal in multimodule project +The XML importer needs to behave correctly with Java and Java +new snapshots have been deployed using target Carlos Ive applied the rest of your fixescan you let me know when those snapshots are released as it wouldbe good to do a new release of the bundle plugin in the near future +The last curl above is to get the network address and of course has to be done after the instance has fully started +What do you mean by Now it would generate? At the moment it is generating the code I pasted above +I dont think so +Please reply to devopenjpa where this discussion was raised this JIRA issue is resolved and isnt the best place to carry on a discussion +Reopening for updating Components +Ok improved runtime type coercion to handle cases like the provided mutable attribute let me know if you face any additional problem on for reporting and providing test case +This required adding state for each tree as part of the initialization of the tree + +But it is not so +This issue is addressed by the latest patch for at r. +No response assuming ok. +Hi ArpitI do see etcpigrpm qlpetcpigetcpigetcpig +Marking as resolved +Note that this breaks compatibility between any python serialized without this patch applied +Reviewing that code change I can see that its going to these tests on Windows +See patch This fixes a typo in the README script for creating the PEM files +That is a problemBut Ram the problem w your patch is that get is being called with forcetrue and you are NOT forcing a new plan right? which is going to confuse no end? We should break up get into region plan getting and then another method on whether to proceed w assignbq +The mentioned in the description got fixed in +Because the connection is in chroot its hard to create the parent node +Thanks for the patch +works +screenshot is all screenshots for new ones +Thanks +On review quite a few of them are really just links to other parts of the reference manual +Hi Chris the package has no full set of Requires for the lot of EAP components +I have fixed the problem with the XSLT but I was hoping that the code would detect the infinite loop within the XSLT and return an error rather than going into the infinite loop in the first place +It also looks like the quickstarts assembly needs to be updated to include these quickstarts +Schedule this for support needed for having this was actually quite easy +Let me know if you manage to track it down in the meantimeAny idea whats failing? The automake stuff or the gem stuff +AshutoshNewly added test case seemed not deterministic +Excuse me I must have some issue with scripts because it works nice trying it manually. +It turned out that the shared instance introduced in was actuallynot as we assumed +Committed. +applied onto as of r will apply to trunk before closing this bug. +Deleted an attachment as it is a file whose canonical version lives in a repository +testcase always change that even it is possible that it could be slower in some situation because the overhead of declaresetting a variable into the local scope could be slower than for example just calling times the length course get that nano as well +Reply is never sentI see my request make it to the onEvent within the process and subsequent log entries from activities but I dont see the reply activity get executed +Patch integrated to trunk. +Perhaps using whats there already limits could be imposed on core count and total RAM +Here is a patch inclusive of a test that fixes the issue +Lets assume this is fixed after years and let the creator of the issue reopen it if the problem persists +perriman two possibilities +Maybe someone else wants to step in and finish what I have done +The attached file illustrates the problem + +but now KT Ucloud API returns api name that include Upper Case +It seems like we should remove and only keep +Is it in a node like ex locktokens or a file +So in your scenario is it that containingBd is? Are you using inner beans that and experiencing lock contention for those? Or does alreadyCreated for some reason not contain the specified bean name there? how are those bean instances created then? Using s configureBean? I guess we should populate the alreadyCreated Map there as well +Thanks +Can you not issue a warning when moduleA has a dependency on moduleB and both share the same paths +The stack trace tells that was used the obsolete extractor based on XSLT which is no longer maintained. +Test failure on Hudson was because of the shell specific statement cat devnull +If you try to run getSequence on the root data object getSequence returns null +Resolved with the patch +committed just like i should be +Attach a patched jar to the JIRA issue and I will tell you whether it fixes the problem +Yes I installed your public key that you have on becuase you had one there I assumed you were logging in with a key and not a passwordIf that key is out of date then please provide me with a new one we are not using password logins for the jailThanks +The parent pom with the war plugin attached a sample parent pom which contains the declaration of the war plugin and the build number child pom does not contain anything special except inheriting the parent pom +ThanksCould you add the commit log to CHANGES too +I have committed this. +It would be similar to other overridable properties with default value like +I dont know if this is the best way to handle this the ability to inject the annotated type for a given bean would imho be a better solution +Also a new object is instantiated every time adding to GC load +I think we should get this one committed +please that both bugs are fixed tests passed +It would also be less efficient because until we have compacted the streamed sstable each read will have to call the cleanup over and overThis is true but compaction is fairly likely to trigger soon after a big batch of streamed files arrives since they will trigger compaction thresholds +I think its fine to use the current approach that uses sequence number for tokens and key id for secret keys +EVER +Whats the status of this guys? It seems that the JDBC part has barely been started so I guess theres still some work to do +I had to umount and HDFS using fusedfs again +ldapsrcmainjavaorgapachedirectorysharedldapcodeccontrolsreplicationsyncmodifydn +I do not set to true +Couple of questions If start time is not defined does it just search from the first available record? Same for end time if not specified will it search until latest? Correlation key if the activity records the values are enclosed in square brackets possibly because the underlying representation is a list to handle where multiple key parts are defined +And I also think that many others users will appreciate this approach +Same thing for the store side +Committed revision fix causes errors reopening +Appears to be some issues here still +Patch v ignores a brokertopic change listener callback if the state machines have not started up +Please see +Turned off in the JIRA now. +package +This bug has been marked as a duplicate of +Ive found that using XHTML Transitional is a better fit for Click +Ive been trying to recreate this with the latest Apollo snapshot and have been unable +Looks like this is still open +In this case the impression is that the files are not copied but they are in fact copied to the root of the web applicationThis happens because copies all files disregarding the directory structureLocally I have fixed it by replacing this call to copy +so much for your consideration Sbastien +see test comments this issue out due to work done on vmpayload to be used when floppyinject is not available +Now that work is underway with AS all previous community releases are +Identity from one table to another +So the may not be needed +I confirm this issue is fixed in +If this is looking like what you had in mind I can add a command line option to set the encoding manually and maybe were good to go +New patch addresses comments from Jakob and Nicholas +If it stays like this for this week Ill close the issue +This issue seems to affect only topics +Can someone tell me how could I assign this issue to me +Thanks Adrian! +Upload the correct patch +I will try again later +I had a look at the rpm side and it looks ready to commit +Still an issue +Ill let you know as soon as I get the chance to have a look at your patches and provide you feedbacks. +Assigned to Jason to come up with a servlet listener which we can use to initialize Seam when the first request comes into rather than when the app is deployed +attached patch fix this for +Ive attached an updated which you can use to replace the one in the lib folder of your Apache Flex SDK +Should we close this issue? +will look at it +for the patch +That looks great! Based on this work we can consider some advanced graph partitioner based on the number of on graph partitionsI need about one more day for more investigation because the patch is somewhat complicated for me Besides for the deeper review I would like to execute the some tests and trace them +This issue cannot be worked on while isnt closed +My first try my Java skill are not that great but it seems to workTest seems to fail with this patch I couldnt find out why exactly my dev environment is not setup for Java debugging +It is no problem to reopen it of dont think the three small patches covered everything in the original big patch but we can certainly cover future changes in a subsequent issue +If it is your itch provide a patch and unit tests! Otherwise wait +attached not present in v +My goal here was to implement the minimum needed to work on andAgree +Net code quality projectsSonar Plugin for A Reactor Project for all the Sonar +Classloader issue which has been fixed with the new IT testing technique. +Please update this report when your mirror has been updated +Or we could use svns new branch merging capabilities which I havent tried out yet +this time a bit slower and in some steps +See the related forum thread for more information +I think it should be the same rules for both versions +No problems connecting using jar in gt or closing all issues that have been in resolved state for months or more without any feedback or update +ouch you are for your understanding +Does it work on your system +st part Authz interface and class which implement Authz part +Set JBOSSHOME appropriately +Regenerated and attached as +Cancel due to incorrect patch generation +You are just clicking it and nothing happens button not pressed item not moved +I will attach a patch that I had written as part of +Im going to work on a wiki article of a checklist to look at if there is not one already +All set +Also it sounds like a more general issue to Hadoop +I was thinking bothBut is this going to forbid us to release a version +Thanks +Bulk close of old resolved issues. +Since we are not directly dealing with cache and instead relying on Hibernate to do caching related stuff it sort of became a bit tricky to test out the steps mentioned in +If alternations are expanded no matter there is an embedded or not we could use an algorithm that requires only one pass of scanning +My suggestion is to create a of this bug to cover any other memory leak if present +Committed revision . +Yes setting it the way you had here +Closing +bulk close of issues as part of making release notes. +Please explain further so that I will understand +Added a verbose and xml based report generator +Committed to trunk andNote that I replaced s with relation names from the new test case in because it isnt supported in +so is this fixed orI adjusted the fixaffects version +Ill update the documents +Add exclude attribute to testmacro default exclude pig test +ThanksCommitted revision +The CSS class logo is defined at the top of +I just committed this +NOTE I did not add the section on the history commands that Gert suggested +Since the reporter did not upload a test case and since I suspect it is a bug on his end I close this issue as incomplete. +Do you have application with WELD servlet? I just cant understand how could the be registered in Seam application +Thanks Roman +This works with but doesnt appear to be that useful +Attachment has been added with description updated as the previous only with the addition of creating only one URI action for each distinct URI +The patch looks good to me +Yes of course +Applied to our unstable cluster will update ticket when further testing is complete +for the other points you are right these transition states would be a different state view from the point of the consumer pov not from the engine pov +Please comment how you feel about thatI also addressed the issue with the header array reference +There is at least one bug reported due to these hard references where a call site references a methods which references a singleton class which references the object it is associated with +groupId should be the metadata needed and Ill copy it everything by handYou may want to move the plugin to well setup there the infrastructure needed to sync directly to ibiblio without upload requests +Being fixed in parts via. +Agreed about looking later at how to optimize thisOverall then Keith does the proposal make sense? Seems workable +Added sbyte type to RodrigoI neglected to add an routine in and will leave to you to add +Yahoo guys please test this and see if it works in your setup +Some of the recent underlaying changes has made the tests in this patch flakey +No seria possible a deliberate travesty of justiceUna pardia deliberada de justcia +Let me know how this looks +Not fixed in is still asking to install Flash plugin +Thanks Sebastian! +Cool +I am moving this to release since its a feature request +If there are still issues with Rails integration feel free to reopen this issue with additional details. +I will take a look + for the change +That may not be a huge problem but does add losses +Thanks for the review Suresh! Ive committed this to. +being able to print out which of possible many available persistence units on the classpath that is relevant for the current entitymanager +Look at HDFS stuff +Now they are the default Maven names with version number etc +If entity maped by annotations then Source file opens in action on the following line +Please remove the attached files from this jira +Do you have a USB cable hooked up to this device when you do this +Test case executin is completed checked test results +bq +has been marked as a duplicate of this bug +This is an issue with the console independent of the ESB plugin +EARlib works in Eclipse M +CP fixed in revision fixed in revision . +I added a SLSB to the war declaring the EMF as a resource and both deploy correctly now +Taking off the fix list for for now +Custom subclasses might but the standard classes that we ship aren also object to calling this a defect There is no requirement for a proxy or other such adapter to be since they just serve as proxies for a target which in turn might actually reside in JNDI +To make the test pass we need to add the setsgets in the class +I have reverted the patch for which broke the link creation for most svn web presentation systems +The issue came up in the classlab at JBW that a student had an enumeration drop down in the UI +Resolving because this is old and doesnt appear to be identifying an actual SDO problem +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +I have resolved this as invalid per your previous comment. +Thanks for the quick fix +Yeah if is not fixed in we should fix this in +Looks like this was closed +alan junit doesnt persist the logs into the testfiles in case the test hangs +Updated Patch for never mind me +bulk close for the release +The state table is slightly different for reset anythingmay be reset so Xs all the way across +Heres some more unit tests +Add visibility tags +That looks pretty good to me +yes the same issue +All seampay tests passed tried several times +Xu has a patch which fixes this could you try and let me know if it solves this problem? Thanks +this is a major implementation change we will be doing it next release +Probably +RobbieCould you have a look again into an updated patchI added a description of property into docs updated test to extend and fixed compilation issues caused by recent changes. +Please check that it was applied as expected +This proting to and +What I see is that at the client side some remote calls are blocked packets not being sent to the server some others are timed out waiting packets having been sent to the server but the response didnt get to the clientWhen that happens I found the CPU usage usually is very highHoward +Ive also got and + trunk closing all issues that have been in resolved state for months or more without any feedback or update +They use services that are delegated to them from the portal itself which in turn run within the context of the Jetspeed portal which has the definedHTH by recent changes described in can be closed. +This new patch passed unit tests but currently there isnt a test for UNION +will handle specific remaining cases separately + LGTM +contains a little bit more info +An alternative may be to change the externalSevrice to use instead of however that results in the infinate loop in that was reported in related JIRA about using in +For existing duplicate conglomeratid work needs to be done at the database upgrade time +Much of what has been suggested here Im looking at incorporating in but perhaps its best to do it in small steps +Using the examples provided i constructed unit tests that are now succeeding +Upgrades libthrift to latest release updates generated code and now extends +Closing out patch applied. +Done +The second datanode sets a timeout of minute and so on +I think we have have forgotten to port that feature in thanks for spotting thatI agree there is no need to delegate to Objects lets keep it simple +The fix in this jira has been included in +to and trunkThanks for the reviews +Per discussion we will use HTTP to transfer finalized journal segments instead of RPC. +Lets experiment with before moving it up at the spec levels. +Hoop is rw already btw +Reopened to fix fix version +Logged In YES useridPlease provide a testcase +merge ops against the IWMaybe even MP could be +This does interfere with some experimentation that Ive been working on so I would appreciate a fix +To retrieve the location the search strategy would need to load the CAS file +hey Ondra please verify this one once the corresponding JOPR jira is fixed and a new plugin is available +Suggested patch simply sets HTTP header Expiresnow in Stage to Martin Koci for provide this patch +In the utilities repo the historylog of the master goes back to the Aug utilities is not my svn root but even the called is back to the Jul so I guess theres a problem of history here too +HelloWould you please try the patch? Thanks a lotBest regardsRichard +applied though I was unable to deploy a new snapshot due to a permissions problem on the snapshot repository +It will help us see what exact change you made not loose any of your changes +This should fix itTested this on a single node secure setup but I guess that doesnt prove that this patch is better than the previous oneI dont have access to a bigger cluster +Verified on. +Also fixed on other branches. +Must need to think about when a parameter validation interceptor ought to be executed in the interceptor stackBefore or after transaction security? Some interceptors can change param valuesIts likely that transaction and security must go first +Reopening the issueTo clarify the assertFalse test was intentional +It was not a bug in XALAN +This topic comes up from time to time +the in puts the downloaded artefacts in a strange place +This patch contains a few classes to replace those in the plugin +In addition I believe that adjusting code that works with the output of the call will not be odious +yes you are rightthe first is not valid in XHTMLI noticed this when looking at Once that is fixed this is gone as well. +Should we also consider that +could this be private and finalDone +Can you please test if it works for you? +Robert et alI am not entirely sure there is anything to be gained from splitting parser and dom into separate modules +just did it myself. +Can you grant ASF rights +On windows this means CDocuments And Settings +Ive quickly worked up the attached patch to +I would have committed it a while ago but I dont remember my password +Infact if goes in we have to do some rebase here +Should this not be the caseI have a potential fix but currently no time to document contact for more +I use the jsvc debug option toget some meaningful outputIf the string debug Waiting for a signal to be delivered is foundon the last line of the file then the loop is exited and a return code isreturned +The sent attachments based on the snapshot will allow Xalan tobuild on Solaris +Java doesnt see that as a valid main from the command line works only because you are running off the uncompiled script +Therefore go ahead and commit. +Hi PavelWhen I first updated the test failed with the original value of and passed when I changed it to +Thanks Navis +i like the way it looks now +Patch applied at . +Committed to trunk +bq +With these fixes in place all the reproduction cases appear to hold steady in memory +The second patch should correctly pick up the and use it for sequence purposes +Having all the descriptors in the same form looked really messy so I separated them out to different forms and added links to them form the form +This is not a major issue +Yep has been committed for a while +Thanks Devaraj I put this into and trunk +It seems like a dirty hack especially where its traversing Action classes and looking for properties +Thanks Peeyush +Thanks for the report +Also running all samples does not mean all features work fine but right now I cant see any option other than samples unless Rampart developers spend sometime to improve test coverage +Thanks Suresh +Same patch but rooted in the correct directory this time +xml files extended in +The attached patch can fix the issue +We may look at expanding the crypto implementations in the future now that the are there but for now Id prefer to limit its scope to providing the canned options +Dupe of per comments above. +I believe the AMs credentials sent in the AMs container launch context already contains the same information +Ive push forced an updated version of my changes so that it no longer includes commits from the master which I had merged +Nigel said the release audit failure might be because of lack of apache header for and I could ignore it +I started looking at it for for but its going to take a bit more work than can comfortably fit into a patch releaseThanks for the report +Now that work is well underway with AS all previous community releases are + +This is a perfectly acceptable way of doing things imo and in line with practices in this communityExcept that there wont be a regression test which might be very handy when a consequent change will affect the fix in question +But with it does not work as all write to testeepigtestdistbuildtestmapredlocallocalRunner and parsing it fails with Content is not allowed in trailing section +Before this patch gets merged in I would like the one in to get accepted +Thanks Elliot +This was beyond the scope of the original enhancement to allow sequence values to be obtained without using a +Dinesh about item newer JSW has a support for UAC which is not simply declaration in manifest as we did but an API call when admin privileges required +I couldnt not render items conditionally +in JBDS differ in content +Ok +File is missing again +Were nearly memory efficient except for terms dict and term freqs +In the opt compiler inlining runtime services always grows the code a lot +This looks to be a of Cosmin +svn commit username kazk message Fixes to add a serializer for MAPSending srcmainjavaorgapachegoracassandraquerySending srcmainjavaorgapachegoracassandraserializersAdding srcmainjavaorgapachegoracassandraserializersSending srcmainjavaorgapachegoracassandrastoreSending srcmainjavaorgapachegoracassandrastoreSending srcmainjavaorgapachegoracassandrastoreSending srcmainjavaorgapachegoracassandrastoreTransmitting file dataCommitted revision . + +I just committed this. +Fixed at revision The super constructor invocation can use the bundle context given in parameter of the constructor +This is the modified code of echoclient code Have a look +Is this an issue using +Hadoop seems to work but zookeeper get overridden anyway to be is now failing to find the tests artefacts +Moving to as this is a nit +The issue should be fixed in trunk in the meantime +Well here comes the code in a more readable fashion. +Patch Includes jQuery dependencies +Git commits are done once during publishFinish +Yes the birthday problem is similar but not the same +Hi Paulex In the new patch I have used security property to enumerate the names of all the security providers which include the jce one that should have been signed +you must pass your globals specifically as a parameter. +they are only guaranteed to be compatible within the same release +Some responses follow Ill change the comments in the policy file to note that the code is generic to I was simply testing the autoloading added by JDBC +It was reported years ago and is still hurting productivity significantly +I think this is an issue with the code in CXF +This exchange parser would be responsible for parsing a specific message exchange type +As I understand this issue can be marked as fixed for +I just copy the paste the soap payload in the text area and then click send button +Why dont we try to use as a way to monitoring calls to the Java method calls +Fixed in r. +Need to understand how work here first too + +where are we with this? I patched the other one for the fields. +However I can replace the running binaries and test them if I have the patched rc +We may need to merge it in Cayenne test suite at some point + +Thanks! +This duplicates +Ah thanks. +Thanks rmuirCommitted revision xBackport to was clean ill commit there when the tests are done +Im not as pessimistic I think pooling will still be useful for most workloads but if we add a configurable ceiling on memory use we could certainly add that zero no pooling at allWould also be useful to track the pool hit rate +pushed in cfaae +Yup Ill take this +ok thank you for your great report and help to fix the issue such attitude helps a lot OSS projects even more than direct codeFYI you can override dependency version of a plugin in your pom so you dont need to wait for exec plugin release +This patch adds an ee test that tests the same functionality +Thanks. +Are there any subtleties I am missing +Just have saw your movie +Will commit soon +Sorry I missed that looking between the two agree with Dennis that this is not a release plugin problem +I think my points have been addressed +war before it enters the create fix this issue by inventing a new +Ill see if this is something that can be fixed for +Requests like this keep on coming in from time to time +bulk defer of issues to +Maybe the blue color can be reused for the with revisions so far +cool! thanks +on the original proposalWe should copy lucenes priority queue implementation to Hadoop and remove from lib +removed extraneous text noted by review comments +I am planning to upgrade to and I will close this out after I verify the timeout issue is resolved +The asynchronous version propagates the zookeeper error code but the synchronous one doesnt it simply logs the error code +Patch committed to trunk. +I will go ahead and close this jira. +Coool Enjoy the has actually been fixed FOR REAL in thanks to +So the simple feature view of the world represents falls too short even for types declaring only simple content childs cause the supertype cant really be represented as a +I will commit later today. +They are leading to unsuspected behavior and name crashes like the one above +Nope Im good +lots of people beginning to use hadoop the earlier the better +used MIN and MAX macros from it +attached has quite nameIt fixes the build and I hope it manages comments in the proper wayFeel free to rewirite it as you wish +It is however for reasons amply enumerated above not a generally useful mechanism that is appropriate to include in Mahout +Bulk close for is it something that was ported to or not really +The Javadocs say that will return trueIn the standalone this has turned into get but this is an older API +If you really need the class to be loaded this cant be used +This appears to have been a simple coding bug in the where we set the idScope but didnt return the value so it wasnt being output or handled properly +how about a patch +This is what I got backIt looks like this fix will be in it is already in and and here are ifixes for and ftpsoftwarewebspherecwsupportfixesJR David can you please close this issue off +I did not see any test failures due to Server Already Running or during testing cycle on any this. +Nice simplification and a good illustration of the benefits of moving to a modern Java platform +The patch looks good but I am wondering whether it makes sense to add a default conf and add the default there +Of course it has +Ill propably write some questions to the dev mailing list on +Without setting copy to true usage of is dangerous could be a bottleneck depending how long streams are kept openIf streams are not closed hangslocks +Attach the patch for +Thank you Chris for porting this change. +And should not have reopend this original status sorry for the mess +Implemented and is available in setting of per page theme make the theme choice global +removed non relevent cant reproduce this failure +Another option is to switch an entire DCs snitch at a timebq +I am experiencing the exact same issue with Eclipse and Maven t been able to solve the problem +Move to to minor issues for next because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I just committed this +The class must be wired into the engine classpath in order to be useful that is accessible to the table function machinery +Need to investigate unit test failures +Patch turning on the asserts in +Committed to trunk. +according to JIRA meeting June . +Should be a separate JIRA to add this into? Or does it need to be part of this one +There has been a lot of changes in the code base since that release +One workaround I have found for this is enabling the globalQueue and using it to detect the begining and end of ajax requests aparently it is immune to region course the problem then is that IMO if other named queues are defined the global queue will not continue to be an affective workaround +Ive updated the SVN workspace on minotaurI think one needs apsite unix group membership for that which you dont have +So did it work by replacing the JSF JAR +I was not able to reproduce it and therefore I removed the commentPlease note that derbyall and both passed without any errors +bq +Top level sounds reasonable +Not a bug youre just asking that the first argument is among the next +Please clarify in the comment that it is there to ensure that statically configured one is not used +Sorry javadocs were NOT being built by gump on the maintenance branch and would have failed if they had +checking this in +WST is copying everything over nicely +The XDT can certainly provide such an option if the compiler supports it and I can see a case for the compiler to support it as well +Improved and final patch. +Several minor problems were corrected so that it is possible to set a WEAKREFERENCE property on a node and have that property remain a WEAKREFERENCE +Ran ee tests exception is gone +was about groovyDoc using wrong packages while resolving classes +Passing back to Julie +Some of this was already done while fixing bug Ill attach a patch against the current cvs of whats left +This functionality was added the agent as part of the patch +My Im fairly neutral either way but I want the current duplication issue resolved and it seems like a whole bunch of work to get to where we alreadyDid you get any response from the legal discuss list whether we are required to have LICENSE and NOTICE files present in SVN +Removed docs flag so we dont effort in triage +The patch and the regression test are attached +This issue should be if it is intended to get into +Duplicated by +But you said its a wrong way to solve this problem +As a follow up RSA V reports an error when editing the facets thatEJB Module and Utility Module cannot both be selected +Most should be addressed except for how to deal with file permissions +I think its because reference to should change to but I dont know exactly where so I need some help +The FAQ says M too need to change this +Sure I meant using the column headings listed units to guide perhaps an or switch to do the correct conversions for some degree of automation +Clearly we dont yet have and in RM as of now but Id like to have those eventually so you should create them and putem along with +Aaron thanks for the review +After all a potential client should have a local copy of the wsdl and not need to contact the server at deploy time at the most +The first seems to be broken +Can you double check all the and make sure we have all the scripts reinstated +Its one line change so Im giving my here +If you think this is a issue can you create a reproducible test that shows the issueRe logj do you know where the blocking happens? Id like to take a look +See for install problems. +Where do you read that I like to do thatWhat I was talking about was that you always see only exactly slice of a context for each Thread +I agree that this is not the most common case for large existing clusters +I took the trunk from svn and this issue is fixed +Im happy to brainstorm on a better name for it +A patch is attached +This change upgrade the sequencer to also use adds several unit and integration tests and also makes a number of fixes to the way the sequencer worksAll unit and integration tests pass with these changes +It should be test +Can I now this patch? +Evgeny as I have no experience on Eclipse plugin development and you are saying this is not a simple patch I am not in a position to provide oneWill check with Steffen Pingel of Mylyn how this could be solved +I didnt even notice it when applying the patchIt shouldnt be a big deal though because this ticket and the CHANGES file still correctly attribute the changes. +Thanks for the feedback Daryn +Thank you PlamenPlease generate proper patches next time. +Thanks for your help with this one +bq +optionaltrue was broken by the fix forFor now Ill back out of the fix for +The way it works is similar as how the Python script works +Punting to next version +The testcases pass when run as on RHEL but fails when entire testsuite is run from hudsonI increased the the open files size for root and hudson user and tried to run the job from hudsonbut build internally uses ulimit to reset the size of open files to +Ill buy that testing it in redback now and Ill put it into continuum if everything thanks much +Is that correct +Sure +Confirming my comment from this now WORKSFORME. +It reports no errors +Patch to xercescbranchessrcxercescvalidatorsschema revision +Great sleuthing Todd +The classic example is rolling back a transaction on any +That same instance is stamped out for every node +Id also like to figure out a better way to write the test too +You can use your own Uberspect in to achieve this +Also add the new parameters in the subject of the jira this helps in searching for these parameters in future +As I just updated in richfaces archetype for the jee profile the should exclude this +on +And trunk +No more reproducible closed. +Toni can we close this JIRA? +Adding this feature to too nowI will commit the synchronization of this feature to trunk shortly +Closing freemarker is not delivered with EAP +Also the test case should not use but instead modify the existing file directly +Added the benchmark feature +fix is to remove the affected entry from the db +Need to verify with reporter that things work as expected +Well done +Its checked into SVN now if you want to look at it +Any that has started but not closed is probably useless recovery +Iterate utf symbols ve fixed my patch it was a bit wrongI run it works +Yes it is still there in +Stein am I right that the only difference between your final patch and the JDK patch is with the use of vs +If need a particular feature or bugfix to be included in you are welcome to get involved and contribute. +This is a lateral move that can come after +As far as I understand dbLocked Castor should not only maintain a internal lock but also lock records on the database +mcslee can you give us some insight into why the current version ofpeek does a read instead of calling the underlying peek +But with my Version the openmeetings users are not displayed in the invite people card +I suggest that most people want most featuresplugins they want Tools without having to spend too much time thinking about what pieces are vital +Looks reasonable enough to proceed with +Trivial patch that causes build to fail when libhdfs compilation fails +bq +This issue is fixed and released as part of release +It adds a new method to to obtain the TLD for a given URL +As a user of the I would expect my to be by type in the same way as and instances are by the +Ignore the other files they are there for historical and reference purposes only +Log on some datanodes would show more info +A sample for bookmarks and continuation invalidation +Just to clarify we dont want deployed as an EJB jar just a regular old bunch of as it was in also valid ina jar file in the lib directory of a WAR file +I ran the capacity scheduler tests and they passed +The XML both and SQC find the schema plans to fix this for +The patch in source form as promised +I think the comment in doesnt apply any more after this patch +What does it take to get this issue fixed? As noted by others theres a proposed patch submitted on one of the other duplicate tickets but still no one cares +Request to committers +Closed. +One problem is that you dont necessarily want anyone who can connect to Thrift to be able to invoke heavyweight ops tasks like scrub or repair +For example there are warnings in which were committed as a part of Alfredo without Jenkins +Another pass at refactoring the package and class names. +This should be it +Hi GuysIs there an update on this? This seems like a fairly severe bugCheersAjay +Set Fix version to +this happened in places was already fixed and ive fixed the other +The header items submitting is also something that we are working on but in some ways I wonder if this is premature +a patch that reduces problem a littleMethod get is in my tests the producent of + I kinda cringed with this one myself but its a basic one that seems reasonable to add +I dont know if preserving this comment is important to anyone. +Hello a few notes on your patch Nutch uses double space for a single indentation not tabs convertIndicatior seems to be misspelled doesnt look like Solrs supported as its missing time and timezone Z +theres an attached patch and it would take a few minutes to review +Before doing this we want to see what solution is put in place for EBWS +Documentation +How strange + +Yep working on one right now +closing since nothing for QE to verify +quickstart demonstrating the is invoked before the render process +I will put something together for consideration +Thus we cannot fix it on our side of the trenches we would have to file an upstream bugzilla Im setting this to since theres nothing for QE to verify +I think this optional value is something we should consider for most of our config options +Thats a pretty good idea +OK sorry you got it from the TCK +Where would these go +For x we just need to run the distribute task on the abstract deployer +I think this patch can be applied to branch also +It is not critical bug if we have time to fix it we will do it in atVersion id +It is this N x N configuration setup that is the reason for us to keep things simple and then simply explain in docs how to changes are not exactly trivial to implement in something like without introducing a big dependency and version mismatch +In the time for Chuckme to dig more into this you can downloadfetch the proton sources and try with that in place +Could it support both applicationjson and textjsonBy google this it seems some frameworks use textjson by default and theyll just fail to work against accept could be complex Accept textq texthtmlq texthtmllevel texthtmllevelq q only expect Accept textjson should work +Thanks Tim. +Modified POM to remove parent POM dependencies +Reviewing the code now +fixed in the trunk and branch +Personally i prefer to fire a major comapction request +One option is to transition to Standby state but it is unnecessary the same Namenode may be asked to become Active once the connection with Zookeeper is established again +An update on this regarding HIVEPig interoperability would be appreciated +its faster to linear scan than to seek aroundYeah and I think this makes sense but you will notice both the Lucene qp and Solr dont allow leading wildcard by default the perf is generally bad enough on a large index that we assume you dont want to allow it and force users to turn on the option +I think we may need to update the instructions for the EAP version of this quickstart +Fixed in revision +The fix is to set to none if we truncate the segment to size +The entity definition doesnt specify Double for field types only numeric etc +It was to rebase the patch for I just attached it +Tried to use nodetool removenode node was in UL state so wouldnt run +Updated to address the issues with on Linux +Im going to code up this pruning to see what I can come up with +Done +Sorry I missed the t at the end of the line +bq +Any committer with further feedback feel free to comment here or open a new JIRA against this branch. +Attached is a help section on migration +Im getting an error on +Marking this Complete since we actually did add support for with the addition of. +Attached final patch with details +It depends on how kind we want to be to stragglers in the Groovy ecosystem +Would it be possible to check this problem still occurs +Hi I am facing the same issue and not able to fins a work around +I wanted to open a new ticket but when I looked over my code I thought that its not ready and I want to work on it over the weekendThe attached patch adds several classes an implemention that uses a to request authentication against configured +Closing this issue +I am changing this to an Improvement type as this is not really a bug within Roo it is more so about defining what is expected with multiple front ends +rebased onto trunk +However Im unable to login to been getting the following error No supported authentication methods left to tryAny assistance would be greatlyll create a new issue since this issue has already been ve granted commit privs to Mike for Pico +I was wondering is it a good idea to check set a flag on that possible allow check about database status somewhere between generating and the time we check for exception in for store lcc tx or something else? Could it help the situation we are in? I did run the test suits +It was also discussed that we might want to make these plugins use the class loader that can load from an HDFS url +Im not sure that would be a good idea but it would solve the problem too +Resolved in SVN trunkOleg +Lets be careful about the semantics here Grant +Fixed in Apache Karaf. +Same error +To get a stacktrace in wed have to wrap the exception hiding the original type of the exception +If you have not done so yet please check the menu widget code and xsd to make sure you are following the same pattern solely for sake of consistencyThanks in previous here is a small patch for some labels scottwhen I finalize my patch and add information on form widget xsd I show the recently code in menu widget xsd +I dont think theres anything left for this bug so I am resolving it as fixed +maybe set the version of the dependency as a property in the uber +Looks reasonable +adjusted priority as i dont consider this a major improvement +back port this the CR +Hi Arpit Thanks for review and comments! I can deliver a split patch to fix in this jira if you need +I do not want to introduce an incompatible change for nowGiven that this is a new API I am not sure handling concurrent modification and ignoring deleted directories is a real backward compatibility issue for mapreduce +Looks like a good change to me +Jesse with FOSSG incoming Im left with very few time for bug fixing +BTW for patches that include deleted files I dont think that can go in the patch you really have to issue an svn remove from the command line. +Hi BilginI agree its better than before but I wonder now if we should not only color the link in tables +The Yarn tests are busted yeah +We do not have time to roam all over servers lookingfor files that are not being processed properly + +Thanks Pi. +That should also fix itIn anycase Id appreciate a verification +So far I noted no side effects of the patch on other webservices I am running +see it and woks! big thanks both and John +This means you could clone it yourself regard to processing done sequentially +The module is not a typical community based project as I am aware of but controlled largely by a single main contributor +latest patch changes DEFAULT to DEFAULT +maybe render these in a different color or something to indicate they are not there or maybe have an general option when it should render visible and elements + +activeio transport moved in trunk revision + on making code of type short +With secure mapreduce the streaming classes get copied to te same directory along withAdding the org directory also in the expected contents +Which can be ok but must be optional imo +sigh +Remains a low priority open issue +I dont know why I missed this one +I have added some comments to with regard to fixing the issue that this change addresses in the caller but it still seems reasonable to bulletproof this routine even if in a running system we dont expect the case +Not sure how it got assigned to Guillaume initially +The only class Im aware of that inherits from it is which we needed for grammars +since it can be solved through appropriate +Thanks Denis for the comments I agree +every unixy OS +on patch +Should this be a blocker for and now that is out in the wild with +Same thing here +How much more pain would it be to implement that wayThis looks good to meShould we also have a method to set the default mapperOne note here is that it is auumed that all the mapper will genenerate the map output values of the same class +We agreed that guessing wasnt going to be implemented +This is the cause of +Oooh then were doing evil things +Solve this problem will not going to be easy +Then Ill ack your patches here and push them ontopthanks patch +All the abstraction is done in anyway +In case the above is not clear +Often type expressions cannot be completed +I just committed this patch to the branch and to TRUNK +Btw as mentioned to Dan yesterday to get each component to log to a separate file we could use a similar trick to whats suggested in +This issue was resolved by using alternate tr +Does this make sense +Thanks Matei! +This commit updates the EMF models to support the Camel Quartz binding adds a new create feature to the palette to support adding a new Scheduling binding adds support for editing scheduling bindings on the binding property +Added patch to fix error during first start +Attached WAR that reproduces the problem and logs from app is requested on httplocalhost the snapshot works with inside the app +I have looked at the code many times and do not see a way this can happen +writing early in the morning not a good read the description you provided +The collective softwaregrant is complete +And the reason we have THAT is because the are really controller pages that link to the template presentationBut with Stripes we do not need to be so rigid +Ive checked the patch in + +I dont know where to start +It would only serve to confuse the new app attemptTherefore I propose the following The current job token remains primarily for the authenticating of the task umbilical and each AM attempt continues to generate its own job token +HelloWould you please try the patch? Thanks a lotRichard +How about a big edit box together with samples for each connector that can be cutpasted and editied +Thanks Owen! +hiCould you post more details such as the version of JVM Axis version and a minimal test case to reproduce the problem so that I can help youThanks +Thanks for the patch Dag with its clear explanation of how behaves on various releases +With version Ive added the provided patchSet this issue to resolvedThanks to Johannes for his contribution +for the patch also +We can probably continue to support the existing Writer constructors no? If not wed put this in rather than +For the usage scenario custom filter needs to be writtenClosing after the agreement from Davis +This is a minor bug problably not worth fixing because of all the other problems with standard canonicalization algorithms when used as the serialization for XML Encryption. +Note that this is all primarily presentation layer changes +Looks like JIRA is using a relaxed consistency model today +Thanks also to Michael and Robert! +Generate using wsdljava allModify the URI address inUse this address httplocalhostefileThen run this attachment as the client driver codeObserve if the client created successfully from the WSDL + +This is a duplicate of which was already fixed +Note that along with and version the configuration object is now called instead of. +ack actually I think it doesnt have anything to do with recursion you may have just noticed that in that context +The nd screenshot shows the Edit screen and the rd the erroneous Create screen +It basically sets up a onto xjc to automatically detect and handle the collisions +Fixed +Rev +If you describe the problem I will take a lookI will submit also new patch for the issue because of the changesRegardsRashko Rejmer +The showhide logic has been completed and is available on the top level issue +See in the quickstart +Note that Perforce also needs to use this +Fixed in Jive +for me that isit works with SUN JKS and our own DBKS keystore +The three referenced classes currently have no other dependencies +Committed to trunk +Hey Sanjay +Whats the work around for the problem you mention +And a select operator is inserted before the groupby operator +I guess it shouldnt go without saying that the super optimal best case solution is to have the user always see a helpful human readable stack trace error +The code in my example here is not being fixed +I am running ant now +Ah that is good to know so the comments gets awareness since we see those mails on the commits listIf thats the case we could add a similar link to Camel since CXF already do it +fixed in svn +Hi HenryAll of this content will be changed to ASF and Apache Gora including the graphics +Done in +Simply I wrote my own componentsbulk closing issues that have been resolved for more than a year. +also suggests an override attribute but with conflicting semantics. + +i have no idea if any of this works yet +Running run Failures Errors Skipped Time elapsed all API and Pattern issues +This should be resolved in upstream now. +I need to dig into source code first and I was stocked on other issues +Can you check if this is the case? If it is not please tell us the code that needs to be fixed +Fix of memory leak from rththrowlazy runtime helper +Closing old resolved issues +PatchesFirst for the dummy test case which I created to check the strategy annotation Second for the suggested change in class +trunk closing stale resolved issues + +If you enable the CMS GC it should automatically enable for you +If Brian is willing to do releases using autoconf or greater when he does releases heres a patch for this problem +Go ahead and shut off the cron +Attachment has been added with description Screen capture of Adobe Acrobat showing the duplicate transition resolvedfixed bugs to closedfixed +In contrast with the way I proposed you dont need any such estimation +Great! API suggestions make an awful lot of sense although the JSON array API over stdinstdout is exotic by itself +I hope people try this with many different use cases +Version +Mass closing all issues that have been in resolved state for more than one month without further Trunk Completed At revision GS Trunk Completed At revision +bq +yes just those +I think this might be fixed in +This is causing a penalty +Issue fixed and verified +GregoryI think we should try approach firstThanks for your persistence +seems good to me +Please cross check +complete +Added this control to examples only for now. + +The removal of quotes in fopexeccommand worries me theres certainly a good reason why they were added in the first place probably to handle paths with spaces in them +dnsEntrypost may may free itself but still there are some operation associated with it +Ill commit this +The point is that it builds indepednent of any code changes and even if the previous run was successful +So the time is not being spent in getRow? This scales linearly with number of columns? Time is elsewhere +Ill file a new JIRA with a patch then +Transitioning to closed. +In the case of tests running as client the database is not fed +Today path actually contains path numPartitions is part ofSince I have changed it to be path partitionInfo numPartitions is going away +In theory the unit tests should pick up any regressions but it would be a good idea to review the commit logs and issue reports for this class before ripping out the bracketing step +Sure I think theres enough there for now +More investigation is needed as to why compactions cannot keep up under high write load +Thanks for the patch Igor +I let come up with the using a standard algorithm for all my bundles and that results in a consistent behavior +Eh lets leave it alone and we can edit again if we need to +Committed revision Jeff As per your last comment I assume this has been commited? Can we close the issue? committed as noted above +Samuel your patch seems odd +META +I cant recreate the issue with the getting written when it wasnt there in the original SCDL so i wonder if that is coming from somewhere else outside of Tuscany? Can you give any more info on how to recreate the problem +I downloaded the patch and ran the tests with and without the it and got the same number of failing tests +reopening as resolved by mistake +Should we send cache report only if some thing in the cache changes instead of blindly every seconds? This could happen in a later patchjira +The logback appender is something I will need for a current project +lower precision and rounded up when the first eliminated digit is or higher +Phillip please verify fix. +try this at homeIll upload a revised patch shortly +Sorry about this +I would really have preferred to see some automated tests included with this change as an incorrect implementation of this could result in data loss from snapshots +Missing guava is for jbdevstudio target which does not include the GDT +You can take a look at that +it doesnt help w the posttext but it does stop the bleed into the signature +Implemented using eclipseLink +Sure blockfor is better +Small typos fixes in the description +fixed from an annotation point of view +Updated patch to take into account Stus feedbackI punted on synchronizing around the get method +Thanks Amareshwari! +The proposed changes already exist in additionally as Brent pointed out the provided patches can not be integrated due to license issues. +Most UNIX daemons use syslog for logging +The patch has been applied +Default wms path is not used anymore one can create layer groups that expose their inner structure and nest layer groups within other layer groups to create trees of layers. +Thanks Jian! + +A reliable mirror will not make the internet connection reliable but I guess the solution is good enough for now +Cache info time count hit miss timing my previous comment another way to look at the is not that they are nested but just that they are +Committed. +Thanks Namit! +Is this an option as wellClarification BC doesnt need to be added at all means doesnt need to be added to the list of providers as one is selectRemove the dormant Apache Juice provider from WSSJ code as part of this refactorThe solution I commited is causing interop tests to fail with WCF so reverting it for now. +Thanks Navis! +But this goes only on +Yes i just check and it has been applied to the trunk +Our default is really scenario though we do have people relying on that +Should be fixed through the jboss common update. +This is captured by so lets get rid of this one +I think disallowing constraints on annotations is reasonable +Moves the setting of writesEnabled before we go into the wait +Vmware support may come in think this need to be cloned for support in and move the rest as resolved state so this story can be closed in currentThank you for the messageI am on vacation for a week therefore expect a delay in responseRadhika +Closing no test case or additional info received. +While the lives some other Region X may have split and been removed and then the RS crashed + +We are on the way to update the jdbc drivers soon But every update needs a deep testing had for example a lot of troubles with the latest h jdbc you please write future tickets in English so everybody of the Railo team can pick up the ticket thanks +The fix works fine in ADS but has a bad side effect on shared LDAPDN tests + EricIve tried on beta it is really great But I do not want to save the attachment in the databasehow can I configure Thx JamesGlad to read beta works for you with OracleIn version it was possible to configure to store the metadata in the database and the mail bodyattachement on fileThis is no more possible with version +Resolved now its first checking if JAVAHOME is set and uses it if not it falls back to the javaw as before. +Punting +I wasnt talking java objectsYou also said that Perhaps just being explicit about would be nicer +in case of a forward but still aims to let a Portlet request in place if with a exposed it basically wont touch it and simply leave it exposed right? Would you rather want it to build a instance for the request that it has been invoked with? ok to me +The latest patch looks good to me +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +I assume this will go into the release +These icons are basically renamed icons from Apache HTTPThe icons are in the public domainExpand the tar in the base geronimo build directoryThis creates modulessrcimagesimg +Could we make current a local variable in makeNext +So I think we can do one of two things to correct thisThe first way is to always use the correlation id of the request if it was setThe second way could be to add a config option that controls if the above behaviour should be activedWhich solution should be done? Are there other possible solutions +Note If you build the head of the branch it already has the fix because is already fixed in that branch +Youre probably right Richard so just to be on the safe side I started a new vote on this updated version of the contribution and will continue the IP clearance when that vote closes +I copy one entry and paste it into an LDIF file then the second and the third and so on +Youll be in the next updateIf youd like us to display the logo we need you to host it and provide a URL as well as a URL for where youd like us to send people who click on the logo. +Suppresses duplicates but doesnt merge overlapped chunks +fixed via Thanks again for this patch +Committed revision +Resolved in. +What Im saying is that is doing just that and actually solves the problem not only for atomic batches but for writes in generalSounds like were in violent agreement then +Closing. +The tools jar is required for the schemagen stuff not the stuff +So if we wanted that changed to id feel fine with that change +Great stuff JustinWith my Sling apprentice hats on I was thinking to a totally different approach See my attached pathThe two approaches can coexist though please give a look at it if you have time +Thanks a lot +I have added an it test and cannot reproduce +Thanks so much for putting a patch in! +IMHO retry count doesnt make a lot of sense esp +It works well with the src distro where we dont provide the file +Can you try the latest gem and let me know if you still see this issue? Im trying to reproduce but am unable to trigger the out of memory m on Mac but also using with default settings +Yes Thanks Jean I was trying to get a volunteer to help get these documentation changes into some DITA realm so that they can be committed +This is like saying we shouldnt have EJB because everything is possible without it +In some cases specifying a different name is desirable but in this case I see no reason to have a JAR so I think overwriting it is fine +Sorry that was intended to be an enhancement +backported to +is old and never it coud be bring back later +bulk defer of featurestaskswishes from to +All have been implemented and tested as part of the certification +Agreed +We can load them from at the beginning instead ofhandling them per table +bulk move of all unresolved issues from to +Solved by addingclasspathrefto the target javadocThere is still an issue with the tutorial project in java +I guess we just would have to document it better as to which component to use in which case +A quick fix for +It seemed to have resolved the issue +Hi Ianthank you for looking at this problem and providing a patch! Ill take a look at it asap +Not been a problem for quite some think this was fixed somewhere in the version of +So the setting of looks right to me +Im attempting an upgrade from and Im having the same issue +dont ultimately disappear from as a result of the fix +convert as well + +By this we automatically do the checks in all modules without taking care which actually need them +Sorry Raghu I missed your comment aboveYes please test this against earlier versions of Thrift if you can as I have only tested it against and may rely on features of that version +In this case please beware that pigpen is not actively developed or supported and is looking for a new loving owner +ZFS has quotas and refquotas +But when someone who didnt delivered the patch comes to just say hey! what are you doing you just have to fix +Looks like this has been resolved? Cannot find the line Warning This channel expects that only immutable +This issue is closed now +Id thought couch was targetting and above but if it needs to run on previous versions we should provide a module that wraps re and regexp and uses the appropriate module on the appropriate runtime version +This passes all unit tests +I will be writing and uploading unit tests in the next few days +The Deletion by DN will still fail with the exception specified in the previous comments +Carl isnt there a problem in having hivejar in ivyI am not sure but will it create a problem if I am trying to compile different versions of hive on my machine concurrently +removing or keeping them but now I feel like removing them will help keeping the code cleaner +Its the root resource for the actual running m not sure why its not showing up in hostmaster +I dont think it has been called for quite a while nowI should talk to my clients and see if they still need a solution in the toolbar +Heres th and percentilesIll look at that stress patch again I seem to recall it not making a lot of sense to me when I last tried it but will give it another go +Though its not explicitely mentioned it seems to be the same +closed +I need to rebase this patch now that has been committed +Is this reasonable? The user dont uninstall the the application all its config information will be lost in the +the cache is initializedPlease discuss with amc and myself before implementing +I committed the new patch to trunk and branch lets see what Hudson says +Bug still present in hibernate +Unblocking this +Both testcases passed on my local machine in Y! S with patch and fail without the fix of this patchSame patch has an issue in trunk +JIRA doesnt show changes in the issue by default and there was no other indication of any work being done on it +I think the simplest course of action is to make a shallow copy of the Configuration in the ZKW constructor +The patch is strictly meant to fix findbugs warnings and doesnt add any additional code hence the lack of tests +The fix has been released with the web console therefore Im now closing this bug. +Closing this now. +ok +It depends upon the nature of the config change or fix +interesting catch +The code works in but not in +Thanks Eks! +Fixed through defensive checking that a provided class actually implements +Marking closed. +Patch applied on and tested against tests +Thank you Bikas. +On to the next one +This is the patch to fix based on xbean +Shouldnt this be changed to now +Its with +I merged back to branchx +Will post a new patch +If Hadoop community dont make integration easy then it depends on third party to make the integration happen +The JNDI specification states that the callers bundle is determined when the initial context is first created +Thanks Stepan +Sorry I meant to close this ages ago. +I am running against +Stylesheet file. +Committed to and trunk. +Adds test and fixes some comments in test +Verified Fixed with SVN Executed Repro Steps and now Im able to run the target. +It defaults to name and schemehelloworld which are the current defaultsExample usagemvn archetypegenerate DnameAwsome Dscheme for the patch +Do you think we have a need to do something like checkstyle in the future +No need to keep this issue open. +Futon should reflect the true case underneath so just fixing this in Futon would seem a confusing step albeit one that works for many people +I tried the patch from august and that seems to fix the problem +close off release +MR token auth against multiple clusters will be hard to do without it +so everybody can understand how it is been fixed +Its been a long time but as far as I remember this isnt supposed to be a problem anymore +well keep it in mind when this gets worked on. +Sending the number through +that doesnt seem like a good idea + +Ill check this in +Hi ArjenWe can work for a future release +Pub Sub tests have not been change of shipped and resolved. +I had to up the child task arguments to fix an OOME +Can you remove this patch +nope forget about it +I think we should only do this upgrade in trunk +It will allow to saveload tags templates to external xml file +Resolving as fixed at since the main parts are in place +You are right I am using snapshot compiled from svn not logical +It would probably have to be somewhere other than in a totally separate document that they may not even know enough to readIts a real problem +During the build time the hdfs usage script should modify toolsservicerun +Committed in in . +a few things The reason it isnt saving is that the service method isnt annotated Transactional The annotations for the Joined should be for both The UI currently doesnt provide feedback that any action has been taken these methods should be in the person service not user for finding theseI have added these methods in userservice because opensocial doesnot provide option for adding friends and this feature is a part of Rave and rave users +All these features are by now part of the current Hibernate Search release +Brian Thanks +yea i debated that but I want to clean up the queueOptions with one commit thus put them togethernote that ring queue went in as a separate item and will most likely put replay queue multi consumer queue in as their own +This is not done +Ok i will try if it work i can prepare a patch if you want me to do soAnyway ill keep you informed +The master then transmits this to the slaveThe basic steps that will be performed when a stop command is given to the master are Stop adding log records to the log buffer Flush all the log records in the log buffer and send them to the slave +added a shell script and areadme file +Anybody intending to look at this should waitAfter further testing with postgres It appears to be causing race conditions under some circumstances probably related to the updates and triggers happening due to order line status changesMy plan is to move the support lower so rather than call the cancel line item routine N times Ill change it so that I can pass in all the line items to be deleted in one call and submit a new patch +Committed to branch and to TRUNK +Definitely not a biggie but then not a big change too. +Entries are migrated +Dave Ive raised to address the Hibernate issue +I changed the text in but forgot for the trunkJust attached the updated version +The code works as mentioned above but with one addition +Yes they also carry information that can be used to authorize an otherwise authenticated client but in some cases they are also used for authentication itself +So if there is an implicit mechanism and it isautomatically set for the dispatched mode I think that would besimpler +I have attached a sample app that works with Grails that accesses a package scoped class in srcjava from testunit as per your this is still an issue for you please attach a project which demonstrates the fault +Hi RandyThanks for the awesome patch againWould you mind having a look on make check C libjstest? some path changes might still be necessary +one exception is the pgsql driver the new version of which is not on the internal server so we get that direct from pgsql web site instead for this done we can now run all the JTS QA tests on hudson +The one passed all the tests +Its an ugly thing and lots of small places in the code need to be updated +No worries good to hear that it works for youJuergen +was committed in enables weakref and finalizer processing in semispace word for semispace is finished after committing patch at r. +you can go back to sorting against fs only and the performance never recoversSystem Ubuntu on Phenom II Java So my guess is that this is caused by the ord lookup going through and the JVM not optimizing away the indirection when there is a mix of implementations +Inside show revision +I would also like to start creating jiras to start transititioning to this class +Can anyone provide more information about the original report or should we close this report for lack of additional information +This exception is expected as class is not available in this environment + +See the devel on closing all issues that have been in resolved state for months or more without any feedback or update +applied for jozef as he is not avail right now. +Having a way to point at one of my own existing databasesseems to address that problem adequately to me +publish and subscriber is easy to create new API +Committed to trunk. +Its a little tricky either way +Ive just committed this + was changed to completely strip HTML +The normal behavior is to put up with the deficiencies of the data and try hard to deal with it anywaysIt would be nice if the complex feature path would behave the same way being lenient so that has some behavioral consistencyI fully understand that not having this kind of validations may make debugging harder +Ive sent a VOTE on dev with copy on users +Patches welcomed +I think we can all agree there isnt an easy fix staring us in the face so we need to think in broader we need to have the KVM agent start thinking in terms of poolsclusters with some quorum logic then thats whats neededHA is a major component +do you want to create a separate branch so that we can work on it together? because all these tasks are very +class files get only time Ive a stack trace similiar to yours is with a specific version of the plugin and a snapshot prior to JSR +Ill do a pass to clean them up right before we merge +Later on if we need to support a lot of mutating operations we could add a new mutator api +No need to be public +The constraint gets applied as well +Most are ready now we are currently working on M components For M we are will wrap up with Wrapping up Client side validation Calendar of these things I believe +Go with +Let me know if theres anything I can do to help +If you want a new module we can set you up with one +Its running the tests twice again +It is about potential issue as well +support for this has been added in roller trunk revision +Closing the issue. +Tested on as well seems to be working fine +Do you want this backported to the branch or just on trunk? If no backport is required then we should close this +Here are some MAC related code from CXF please feel free to use whatever realistically can be borrowed from these text files +this for the time being since this is probably an XML parser issue +I think should be there +Thanks for the good in resolved issues. +To be clearer it matters if you are asking for the column name or the column label +I have no idea what is going on there +I like the simplicity of this patch and agree with the tradeoff of logging in the state transition handler versus the transition itself +That user node should only be added into the tree after the credentials validation +We are migrating from to and we do not see a way to fix it +I ran the javadocs task and found other mentions of as well as fixed some more javadocs +I see from the stack trace that this exception occurred while in the compare editor +Another optimizer can be plugged into the code when availableDeprecated classes must be removed when preparing the next major release. +This is already implemented with which takes the ones we recognize from parsingand it will only apply filtering of those included in my new MC updates +So the problem is that OUR implementation of Cipher reinits itself at doFinal which MRI doesnt result? Different IVs are used in subsequent encryptions with the same object +On linux I dont see any error so I managed to get a few others to test it as well and there was no error +Also out of curiosity what were the main complications w mutable +Attachment has been added with description Fop configuration xml has been added with description An executable batch file to show the difference between pdf and awt has been added with description Example fop config xml that works with the patched files +Reverted the previous commit since the whole fix needs to deal with in whichs going to take me quite some time that I cant invest issue is gonna be rescheduled for on and trunk +Extracted code into +Additionally the hcatalog tests write files all over hell +Closed with the correct fix version. +When I looked inside I found that somehow it includes all +If you think its safe to commit it without further comments maybe someone familiar with committing docs could do that +Patch is simply the changes from from + should fix this problem +Your prior proposal was much more straightforward creates the Mapper +Debug output with stacktrace reattached from bugzilla +When I realized the dependency on JVM I installed that JVM and switched the dependency +Format so table is on one Migrating ESB +Yes this is similar to what I suggested for our own import existing project +if a user wants to submit a complete dtd we will be happy to include it +strange +Sorry I did not notice Alberts comments +But all EJB dependencies should only get pulled in via the plugin and shall remain only +I dont read it the way you do looks like a must to meIn the FAQ is says All releases created and distributed after November must comply with this policy +Verified by Anton +Thanks for the fancy patch though +reset on its last try calls setup with allow aborts which will call reset again +I didnt have declared as an extension. +Thanks SumitYour patch is in trunk revision +it looks like your strategy is pulling out a bean that does not implement accountservice interface +need additional work to add dom flags +Ok I am now able to log in as admin but while trying to click the Configuration link I get the message belowAuthorization are not authorized to access this page +ova file does not exist anymore the size will be +Ive cleaned a lot of this up locally as I rewrite and Session +michal +Moved to is up and running on dev +Cool thanks +A patch against current Git head to generate the separate were at it I think it makes sense to pull out all the JMX interfaces into their own jar too +But not sure how long exactly as it does not only depend on my part +I think its better if the sys ops did this +Assigning to Tom and he can resolve the issue once a released version is available + could also be made immutable by eliminating add and set +Assigned to Wayne +I determined that it did not + +no longer relevant +Removed the folder +wicket is not the one transforming the headers it is tomcat as youve stated +Is there a plan to fix this in Axis +Ales do you plan on doing this +Thanks guys! Sorry for the long wait on this one we should really get this in +Closing for now as all the main stuff is done. +AdrianIve noticed that the tabs of the party details screen are not highlighted when selected +These blockers and critical issues are resolved but not verified +Screen shot of the new project configuration page in the new project wizard +Hence the plans kept failingI have the same changes in the above patches too +This patch looks good to me +Logging is on and eyes are peeled +If you approve and would like me to add documentation to the wiki I would be happy to do soCheersAshwin +Attaching a new patch that fixes the test cases failures and few corner cases detected while fixing the tests +Andrea do you have in your local workspace an uncommitted class extends +Added in the wrong place not a subtask of this parent +Hiplease discard my mails of October thI finally found the bugAppareantly the xml files that gave the error contained a space in their filenameAfter removing the spaces it worked correctlyApologies for any inconvenience +Im not getting the NPE and the WAR file looks like its getting generated correctly +Andrew can you review the changes please? Thanks +This is just a test case patch +trunk Now Camel is better at picking methodsSpecifying parameters is not implemented as its not the right solution +Can you check the patch? I cannot apply it on the head or on rev num +Hum works only with EPSG authority so we get whatever comes with as is where this is gonna be fixed +Have at itDont worry about modifying the paths to reflect the live repo Ill be running your script against test in a second migration before hitting live so Ill only have to the file +This frees up the jcruuid property so that nodes can have the dnalock primary type and the mixreferenceable mixin type if this becomes useful in the future +This patch allows the C files to compile without errors for me +The makefile is not used and it didnt have a license so I just removed itCommitted as r. +Thanks! +In some cases some cookies have not better be because some websites can have strong assumption that its cookies are stored for the root path onlySo I think the cookie path rewriting could have options for includes or excludes for cookie names +Maven is faster compared to Maven this proofs it +Instead use meclipse +This seems caused by +One question why tie this to Sequence files? Why not let the user specify any input format that can return a key and value +Seems like a release note might be appropriate +Ill investigate +Thanks Hairong! +is responsible for checking if a partition becomes available again +The actual logical commands currently use bunch of POSIX commands that may or may not be portable +The cache has fixed size and element would be expired in minutes +General consensus seems almost complete at least in terms of the format for messages +The changes make this very close to being possible +Marking this as resolved as a fix for is included in CR be verified with EWS CR +Im not sure if this will help with the decision or not but I should clarify +Im a bit surprised that it should be possible to deal with this issue in the kind of patch you wrote though as opposed to a complete redesign +At the moment I leave it as it is +Fixed +Each scope has its own persistent file +This plugin has been copied to the antlr repository and is being maintained in conjunction with antlr proper +When jvm reuse test invokes before sort test the sort directory is not created and thus test fails +It works fine for me +If you find further issues please create a new jira and link it to this one. +How ever you dont seem to covert all for exemple you seem to be missing +Here is the trunk update +Commit first step changes to trunk at revision need to update once we begun to use +this is a doclit wsdl and the bindingoperationsoapoperationstyle said that it was rpc instead of documentthanksdims +Jar is fixed with saxon in v and Jaxen in V +Patch with right VERSION +Transformers are not multithreaded so we cannot use just a single instance but each transformation can be a CPU and process +The blocking behavior should only be in use if the options has been set +What Java version do you have in path? My guess is that you have Java in the path and its being used to compile the code or was compiled with target option +also looks related there was a place in Configuration where deprecation was not being handled +for proxies I would think it to be desired that the raw object hold as equals to the proxied object so long as the raw class definition of equals holds and the raw class should be unaware of the proxy +You can do this in Django Apache tiles and I am sure there are other countless examples that I am not aware ofigor the abstract method technique where each child class provides the custom components is not a complete solution to the problem +I had a look at the link I will follow the recommanded way +Simple patch to check for fields as well as properties when parsing the URI +Cannot create a inner struct +In this patch there is a concrete subclass of that rewrites links on a pattern matching basisThe patch also contains an extension to with the purpose of showing a sample usage of the transformerDo you think that this solution could work +I just ran into this today! I was about to file a bug +Having such a jar to send to the JVM developers would get them to pay a lot more attention to us a bug that they cant reproduce is too easy to ignore +Ian Sorry I did that several months ago and I dont remember the details +So maybe we should just close it as WONTFIX +It could get lost that way +Hmmm +Agreed closing as Wont Fix. +Steve check revision r and closing stale resolved issues +Verified in is sent after initialization issue still not fixed when restarting +indefinitely until a document is transferred +Note the scope of some dependencies was changed from test to default so that mvn exec can execute the job +This is something different from Affected Version and it is not clear how and when this needs to be fixed +Thanks +This is a cosmetic issue mostly +Add also patch for branch. +Added a test case to fake a dead region server restarted quickly and region open on the dead server should fail +Maybe the server should just handle it as other unknown attributes just ignore it +That might be a solr issue as I would expect a retry to fail with the same error as the first try + +Fixed within +Something about saying Its live! really focuses their attention +Since I have those files sitting around anyway I would love to publish it in any case even if the community decides to stay with the old one +Which is potentially a critical issue +Sounds good to me +I have the same problem in +Fixes this ticket +Committed revision +I think you mean not well formed not not well formattedThis is a regression relating to our use of +In any case the most important part for mixed smalllarge type of situation is that concurrent repair is possible even if configuration changes are needed +The van helsing input string should be exercising the alternate branch +Test Cat HBM put in package example +works as expected +BTW which version are you using? Between and the trunk version weve added a LOT of +Include unit tests +No problem +What had previously been done was if defaults had not been set all args would be set and added to the launch configuration +While agreeing we have to fix the hard coding I am interested in knowing what sort of handlers you are interested in invoking when REST is in useHandlers usually deal with SOAP headers hence when doing REST my understanding is that you hardly need any handlers +Thanks Matt +This is already fixed in CVS +Its a everything is implemented as it should +I will look into trying to pare the zip file down to MB +maxandersen please review on +looks great! Thanks Uwe +Thank you both for your help +Furthermore what is the reason to avoid it being serialized? I find the reason being the cost of support kinda ridiculous seems this reason can be applied to any bug fix because this at the end of the day it is a bugThe implements Serializeable was added to support +Sorry for the last attachment this new one should be used +I have not tried oss build +if you have a patch that you have written and tested Id be happy to commit it + +i think this is a ve searched through open feature requests but did not find matching one +For instance serialization overhead is present and if backed by a slow disk performance would suffer +Okidoki closing this issue then +Leaving issue open in case any more files need the disclaimer. +this is a test case where you can view the problem +I uncommented the test in revision +The error can be most likely ignored. +a comment is mentioning the behavoir +We had to back this change out for GA since it can be the source of a leak when the Spring jars live in a common lib directory but operate against classes from a specific web applications lib directoryThere might be other ways of achieving a caching effect for Spring MVC there trying to cache objects in MVC processing or passing a common cache structure into newly created objects similar to how we do it with the alreadyRossen Ive assigned this to you for the time being since it primarily affects Spring MVC +Thanks Alex for the great work! +Unfortunately youre late for the train the release has already been tagged +At the absolute minimum they should refer to one another in the config file but Im wondering how we might unify the or different reasons for flushing in our monitoringlogging somehowAlso we should start making a plan to deprecate the settings or convert them into fractions as mentioned above +This is already fixed in release +Attached patch +I get lots of node type related errors when running tests in +Lowering priority but this issue should be explicitly specified in release notes to install quickstarts by chosing a valid maven repo is not possible via installer without this fix version as we do not know if we will have an ERJyoti please put correct fix version before next build. +Sanjay Id like to make a suggestion to accelerate fixing these token bugs +This is actually a Hibernate dependency Hibernate itself requires on the classpath Springs Hibernate support just follows +Im not sure whether we will be able to resolve this issue before AxisC release +ok but depending on the looks very weird to me since I dont really depend on it +Regardless it should fail with a +Thanks Ate +stan can you take a look at this +In my patch I need to override create and convert because I cannot find prefix and namespace mapping information from EMFsPlease review my proposed changes +push to +Looks like this JIRA is dup of as dup of as per Ted. +Same problem with JPA persistence and without installing the fckeditor +So I reopen it because Release and Nightly builds contain think I am not able to resolve this +Why do we need keyLocks in Memtable? is for writes too right +This patch should pass both with security on and would you mind taking a look at this updated patch +File content is stored in appgadgets describes the steps to my case the link to file is httplocalhostrestjcrrepositoryproductionappgadgetsappkhakhaappdataappresourcesHowever I dont know why this problem didnt happen in acceptance server only in local you like to confirm whether can you reproduce my caseActually Ive already tried to written some test cases for JCR but I couldnt so I had to reproduce with platform +Fixed a nasty memory leak +Added pidfile and to make it more easy. +can close pull request but its usually because the work as been merged in master +bulk defer of issues to +The following files are still using and as the conf resourcesI think it is ok to refer as we support it now +This does not mean the same thing as +Would it be not more efficient than two ints? If not lets close this as invalid. +The followings page have been updated in order to focus to the input field when opening the modal window Quality profile pages Create Java Profile Rename Profile Copy ProfileMeasures pages Save as Manage Copy Manage Edit +We may need to replace this test data if details are not forthcoming +Thanks for your attention +The downside is you have to wait for writes to complete on other rows which is unnecessary + +IMHO this is far away from the state initially intended by the subject +I agree that it makes sense to check in a specific version of the whole library +The fix has been added to only since doesnt have such a property +by rev I have the activemq plugin building ok +Heres a new patch now that Avro is in Maven central + contains a complete PK class which is inherited by the AB class +While trunk depends on is still with so I need to include this class as well +To work properly the EJB Server portlet should be used on newly deployed EJB modules +We can do this for too +avdl file to a +Fixed +httpd updated is not fix added +Thanks Vinod Kumar Vavilapalli I think is okay to me. +I think the parameter is getting interpreted ok but its not also peeling the argument to cp off the arg list +Thats what I meantWere all in agreement then +It should fix exceptionThe user define for AM container such that we need to update the resource of according to the real allocated size +the same patch as Amith to trigger Jenkins +is to ensure the roll up for previous day happens as soon as possibleIf we change to then the roll up for previous day will not occur until AM of the current day +Jpeg version of the pdf file attached earlier +Thanks Aaron +Actually Im using Firefox +Issue is moved to Hadoop Common because changes need to be done in common for this issue +Committed to trunk thanks for the review Stack! +Hey Luca yep on the Apache OODT wiki +Committed to branch to include in releaseThanks for the patch LN. +Ah I see +Bumping from to this to me and removing the target since I found out that that part of the code was added in I forgot to mention that I ran SourceManager times each and they all passed +the latest patch looks good to meVinay can you comment on what testing you did of this patch? Were you able to verify manually that the ZKFC now doesnt retry times during a failover? Ill commit this patch as soon as this question is answered +Sry. +This contention is still quite severe +Impl +Dropped overhead of boxingunboxing of parameters would be a pleasing bonus +Thanks for the patchA good solution to fallback and use the interfaces for private class beans. +I thought I did +r added error checking to the Place constructor so that an attempt to create an invalid place will now result in an immediate exception +I prefer also to keep the blogs together +Youre right my mistake +Yeh but not without deprecating first +No you can resize a visual content using these items +On jdk the proper MTC version is selected. +Patch committed in trunk rev +Verified on +If you carefully look at the current implementation only possibility that this could occur is if some one tries to add an already started transport to the listener manager without letting the listener manager know that it is started +That should be policy +It goes to the first case that matches and from there it falls through all cases till it finds one with a break +I fixed the performance issue by caching the hashCode until a coordinate gets changed +Lowering to Minor +possible workaround is to use native sql +Im going to commit this tomorrow unless I hear objections +i cant think of a reason that functionlity should not be commited aswell +Hopefully it will make it in EAP will appear in EAP +just store the default root dir plus projectname and source the pointed to directory +Waiting on OK from jgray before commit +This patch fixes the examples for environment variables and +Should now be visible on +Reopen as required. +The header always takes precedence +Committed to and trunk thanks Todd! Leaving open for since it was mentioned in the Target VersionsTodd Will you be doing the version or want me to handle that +Sounds good Ill look at the compression bits and work off that +Message we dont accept changes as full files a patch file is the minimum a pull request on github is better + +for both the trunk and the patches +Escalating for dev analysis. +We are experiencing the same problem in our environment and or now considering removing all our from the application +However in my design i plan to support JSON to Tuple conversion +This review has made it clear that the Getting Started guide was in really bad shape +Hi This is the patch for deleting method in in luniThanks! patch applied at r thanks a lot please verify. +the trunk patch so hudson will pick it up and not the version +will probably try to fix for ve been thinking about it some more +Will it be OK if I write unit test in a patch to this issue +I dont think subclasses would need to override those methods so marking them as protected final is probably the way to go +This is entirely done in Tools nothing needs to be done in the Forge core see +Keeping a copy of the schema with source code that writes data is a good practice the schema is part of the writing code and should be versioned with it +Defer all unresolved issues from to +project for at least then? Thanks +I have no objection +It was produced with a combination of photoshop and adobe illustrator. +pushed +The cost of creating a branch for each version of hadoop seems to be too high compared to the cost of the Shims approachOf course there are pros and cons to each approach +This is worse now that bulk upload is popular all get low seqid so likelihood of clashing more likelyMade it so I had to do manual merge doing fixup of user cluster +Nice I was just looking for the same thing +I have already removed the duplicate code from and am working on the compiler warnings as we speak so you can either go ahead with the smaller patch or wait +And everything seemed to work nicely +Im not sure if we should consider this as a bug +I opened to try to restructure our current Codec idea so that we can put more of the index underneath it +We have time to think this over +There is still a significant improvement to be realized in terms of the photometric interpreter +I have added a failing test in the patch provided +Theres also the JAR signing issue to take into account +dims +This fix should not go into any previous releases. +The patch is working well for us +I havent seen any update on this ticket in some time +I also realized that if the working directory was the file system root the script was broken +Thanks for spotting thisFixed smoke jvmti kernel tests at r. +The difference in the looks different +Hmm this problem does not occur when to So perhaps it is dependent on the version of sshd at the other end +There is no IO tab the Data Inputs are in the Event tab +Thanks for providing tests. +you are right +I am intendingto adjust the comments in and and then proceed with committingthis patchI think you should open a new of describing the nextclass that you are working on +After debugging the source I realized that this not a bug of the net package but a problem occurring when vfs is sending many requests to the specified server trying to resolve theMaybe there is a different approach to this problem +Are you still not happy with my patch +Turns out as it stands Scan objects cannot be copied as the are collected into the Scan object +I was able to reproduce this problem with the Interpretive processor in but I believe the results are correct with +I suspect some other class is on the classpath causing the original problem. +This file contains a patch that activate the use of the links with ve added a test +At the very least the method should log a message saying a null attribute was attempted to be written +fixed views +I eliminated the need for an enumeration entity to specify the data typeThe two screen shots show how I used the user preferences feature to implement a collapsible masthead +The upstream issue is Resolved with WELD which is included in +I dont remember what all changed from openwire to but I dont think it was all that much +Attaching the diff for the patching of this issue +Is this still an issue +Thanks Owen +It appears that we do indeed internal and external encodings based on flagsThis shouldnt be hard to fix +However such a is too verbose +Thanks for the comment David this is this kind of little notes that helps to better handle. +Please critiqueA contribA contribudfsA contribudfsjavaA contribudfsjavasrcA contribudfsjavasrcevaluationA contribudfsjavasrcevaluationstringA contribudfsjavasrcevaluationstringA contribudfsjavasrcevaluationmathA contribudfsjavasrcevaluationutilA contribudfsjavasrcevaluationstatsA contribudfsjavasrccomparisonA contribudfsjavasrcfilteringA contribudfsjavasrcstorageA contribudfsjavasrcgroupingA contribudfsjavatestA contribudfsjavatestevaluationA contribudfsjavatestevaluationA contribudfsjavatestfilteringA contribudfsjavateststorageA contribudfsjavatestgroupingA contribudfsjavatestcomparisonA contribudfsjavalibA contribudfsjava +Making this available through the API as Jukka suggested wouldnt increase complexityMaking this available through which API the Oak API? No that would be fine I believe +I was going to make the changes discussed but it looked like both Sun and IBM nightlies started using the existing format so I dropped it +I have committed this +Dear KeithMany thanks for that +Fixed +AFAICT there is no bulid problem. +Sorry for the delay +I cannot reproduce this running on linux +I dont mind having it on Schema directly like applyAliases +Closing out the issus for the release. +point the queries seem to assume a field firstName in Person yet it is firstname +Closing all resolved issues due to a successful release +Thanks Devaraj +The patch +for the to trunk. +I dont know if it is the or log it depends on your configuration +Can you post a snippet of your route with the recipient listAnd what version of Camel are you using that fails +Ill look at cutting a release today + +Go ahead and deliver it +What comfort to me is the fact that the connection is open and valid at thetime I call isOpen when it can still go stale by the time HTTP method executesI remember that +This fixes typos and adds to the example MERGE statements +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +HiWould you please try my patch? ThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +The previous assumption that there was no need forholdable btree thus the ASSERT +Marking closed. +JIRA project created +Thank you. +I fixed it + +Any chance we could close this now +Now fixed for all lock types etcThanks for reporting Joetrunk +this one was open for far toooo long. + +Once this is done we can close the issue again +There is no automation but you could have a look at which can convert TO confluence syntax I dont know though whether it can convert FROM moinmoin syntax +Fabulous timing Alessio I was hoping youd see it on Monday when you were back from PTO +Also this patch doesnt add any tests to verify that it actually fixes the problem +Also when we do that I dont think it should be by default in the tomcat server used by Oozie server but in a separate one +One last mile Guess autoreconf should not be part of clean phase +Instead wouldnt it be better to come up with a set of procedures to unwedge the cluster starting with setting key to false resetting and using manual failover +In our case now its only enabled true or false +Functionality added but theres some comparison failures based on results from different marshalling methods if difference bigger than x fail +There is nothing that we can do to fix Stock Android Browser issues sorry! I think your issue may be related to jQuery Mobiles broken handling of touch and click events +Thanks Arun! +I just tried Windows Vista JDK it worked +BrianIf we have a volunteer for this then we can mark this as a blocker! no? thanksll reduce the priority until we can confirm someone to do this +verified in +Works well Simon could you just Add a description to the widget Basic welcome message used to provide links to the most valuable Sonar resources documentation support +move to breaks this with head and URI has more detailed reporting of bad directories but it does not raise any exception if the no of data dirs +Actually Im glad you rose the issue because Ive tried to trigger it but in vainIll look att your patch today +An administrator might set the sharedFS globally while if we use the shared prefix idea then the user has to remember about shared and make a decision for each job whether its a local file a file or a DFS file instead of just the foremost and latter +I sometimes use spring to inject configuration values too which might be of spring but works quite well +Are you stating this as a concern about doing this or just stating it +I think it makes sense to sleepretry when the reset is not successful and I like relying on for the retry +Can we trouble you for a test that exhibits the behavior +Do you mean that xsitype would hold the complete java type? That would also be is sufficient Thanks +I dont think there is any other way to represent this +Thanks Mike and Sheryl for the conversation! +This JNI environment is used in JVMTI processjvmtievent which is called to signal JVMTI that thread finished +Jon the patch has not been granted ASF rights +Attaching a patch to remove caching from which I plan to have no problems with this +The map need not be sorted +This can introduce some changes in the core API so lets resolve it in +Of course I ran all the unit tests locally +We should clean up the dependency and build structure to reduce duplications +Thanks +And Im pretty sure that all of these commits were in +trafficshell online help was implemented in +I will likelyfile another JIRA for the code inspection fix and link it to this issue as I still cant repro this for myself +DaveI meant that numRegions could be negative +Added a progress display since the files are taking a very long time to download +Can you provide the SHOWPLAN DEBUG output? This will show the planning state before +Will wait till Dan is done with this +I did not agree that this issue is closed +Implemented running total functionalityThanks to Jacopo and RishiAwdesh Awdesh Rishi JacopoDone at r +In the current version of this I would need to create when lines selecting only one at a time they were hoping that even if under the covers this could be managed from a UI perspective they would just have selected and unselected cities in that list of without needing to create s of when lines +I wrote a client that uses the new read API +Closing resolved issues. +I used the xmlbeans databinding btw + +Fixed +Will likely work on these two together as I have similar changes already in can you separate out the stats piece into a different issue? Ill fold the quoted stuff in with and then we can deal with the stats in other just rename this issue to just be the stats email processing additions moved toEnhance statistics to assist tuning classifiers +DVF will use the Locale object to construct Collator if user requested territory based collation +OR Do we really need to change the behaviour of distcp +RAD is round the corner +Can you validate a current to be sure +Checked in a fix at r +Quickstart +Especially if all upgrade related commands are going to be issued by the namenode +It could be that streaming is broken with thisStreamingrepair will only be broken if at a particular instant in time replicas are not supposed to be holding the same data +This should be enough to get most of whats needed +I ran for hours with the latest patch and no problems +Any chance you can check this in with an Ignore + iteration done. +So FOP is out +Though Slow start and Nagles algorithm are related they are not the same +You can use it with you just have to supply a Comparator +Working on it +Hence removed localThe methods of the Database interface are used in two scenarios After getting a connection to the database +Just allowing empty or null parameters can create problems like clients setting null value but user consider its a mandatory for the application to run +This was related to +Committed to trunk thanks for review Nicholas +and that the failure policy be explicitly stated +Not for +Works quite nicely +Dont conflate their versioning w the suggested WAL log versioningRegards PE data its data is not amenable to compression +deferred resultion to since still awaiting response from EAP team +I will probably look at it next week please push all your changes to the branch +The case that reproduces all these problem is when a client tries to open a ledger that has not been properly closedAfter fixing the original classcast problem and chasing further ran into couple more issues +I sense heavy committing in flex branchs future +this was my first try at using gant to install the demo views +Ilya review this bug please +Need to include this in as well +This patch should fix the problem note I didnt alter the formatting in as this would have made the patch very hard to patch from Scott not an easy problem to fixrevision . +The simple way to fix the problem is adding waiting time +Read it created only FILE in the specified location. +Thanks MichaelI decided to use a shared cached thread pool on this is to ensure each startup task gets a dedicated thread to run as they can potentially take a long time andor block +Is there a more permanent solution to this so that the patch command is able to delete files +multiline is false for textDisplay sking part. +The hashcode will always be additive to the previous hash and dependent elements to reduce hash collisions +Hi PaulexThe patch looks fine +Supercedes the previous patchIncludesAdditional methods removed from the test mocks as they no long the patch and the regression tests pass. +Will apply it soon +Will need to check that next time Im infront of the code but perhaps w all the other changes that might not be necessary +Martin is right JSR saysThe element contains a that is the fully qualified name of the Java class primitive type or array type of the XML root type or anonymous type the WSDL type scope the applies to and the set of variable mappings for each public variable within the Java class +Thank you. +no caption at all in s necessary to remove captionClass of a bulk update all resolutions changed to done please review history to original resolution type +Would appreciate a review by someone +might give some clues if it touches +Jobs fail because containers classpath does not contain the clover jar +A lot of people seem to have the same problem +Please check that it was applied as expected +I perform a release of a java project with continuum and I end up with a truncated file +The selector on the main thread is used exclusively to open new sessions while IO events of running sessions are managed by selectors on separate IO threads +It would be good to have thisI am hoping that even without this there is a user who will work with us to get the connector developed +Any updates on this +Roman were not actually using the Advice instance hashCode there +redone to adjust for changes since last BobYour patch is in trunk at r + and are done awaiting email response from EPP and engineering leads as to product and project compatibility with the tooling provided by EPP +But youd end up witha lot of the same functionality in both sets of classes +Getting a run just to be clean +Veirfied by Sean at rThanks. +This appears to duplicate Jira bug report which has been fixed in CVS +How about collapsing these two into a single loop? Attaching a patch for this what do you think +Verified in +You should really be discussing this kind of stuff in the forums +As is parts of the really needs externalizing from such that it can be changed without having to recompile the ant +This is a second attempt to fix the truncated download page +interface sync async impls not done yet unit tests for async runner +Scuse jacques for my latencyThe ui present in patch was an error +Otherwise Ill leave this to you to defer +Committed revision +Worked completed and implemented. +committed to trunk +The low level could then stay simple. +Appears to be working now though the folder structure it displays is insane and does not reflect the actual state +Sorry. +Example project attempting as simple insertBeforeAfter +this should all be properly handled in the branch. +What is the syntax +At the very least it will require us to put another on our brains next to the note about implicit schema creation +Thank you DimuthuWith snapshot everything work fineUnfortunately WSDLC is not a single +im closing this issue then. +Closing +Checked links on it. +Thanks Virag +Thanks for the patch +mavenconf is not relevant for embedded usage in netbeans +EMF doesnt think its important enough to fix immediately maybe we can use silimar philosophy +I took a stab at fixing this +Fixes +the case has not been closed so probably no. +You have a new patch for us RR +Having to browse through the properties escaping the characters and all and all was too long +please or create a new jira if you want to add additional features +Ive updated the class to also use the branch when a is passed added a testThanks for reporting this +WDYT about removing from jackrabbit and you just add it to your particular project? Is that an option? Seems to be a quite customized installation anyway +Users can download if they wish to run the Reuters example +I tested the change to OID and typeblob is still broken on existing postgres JDBC drivers +Let me know if everything is ok and Ill close the issue +Thanks for analyzing the relevant jsontags required to be processed by +Ill pull it together +i have build a example based on description but this example work as expected teste on railo not reprodcue problem please see attched have changed a little bit the examplei call now a in custom tag instead of b and to avoid infinity loop i have moved custom tag call in costructor of the falis in railo and adobeCF +Isnt it possible to reduce the number of stylesheets by creating bundles? Wicket supports bundles out of the box +Fixed in although there is some doubt about whether these are valid absolute +We currently wait about seconds at most to load the app +You are correct I used the Tika command line tool which worked correctly +applied in Chinese translation of translation of IN resources +after merge it will do a scan on the repositoryThanks didnt read the previous comment on the audit logsAttaching an updated patch that includes audit log and reportsThanksGwen +Mark any chance you can provide a test case to replicate? I know its been a while since you filled the bug +Ill go over this issue more carefully +Im also curious what you mean +cannot reproduce this failure on windows after running tests times +I understand the reasoning to make the MAXSLEEP lower but can you explain why we need one more datanode +tested confirmed that it didnt work on basic auth +Candidate for SOA +This should be fixed as the has been synched up with the version in passes locally. +Cool with me +One with the old path and another with the new pathSee screenshot where I was renaming from seamproj to seamproject +Methods such as setName and getName on are synchronized +committed to ant branch. +I understand why you guys dont want to include itSven thanks for the pull requests Ill certainly take a look at what you have and probably include it +Thanks Knut +Follow these steps to reproduce Create a database by running the script in ijjava jar ij Put the database in a zip filezip r db Run the script in ij +This TEST page add dinamically a Node to a Tree +A minimal testcase reproducing the issue would speed up the process if you have some time for creating it +OK So the test looks like it is bogus +Do we still do a thread interrupt anywhere in mailjdbc +Attaching patch with a testcase +Change default port number to for Hicc +Deepal I think we are done with this +May somebody pls try this one +Do whatever you like +Attaching a screenshot of the the the patch +Hence closing this issue as not a bug. +I understand making it package private but why remove the test? Maybe you want to take it out of and make it standalone +Reverting should not feel wrong so here is my to rather than revert +We do want to improve the isolation of dependencies between framework and user code which I believe is related but really orthogonal to this jira +Could you guys pls update me if you have found the m using the same version as Stefan Recknage had used +Not sure I got the part about where first letter of second word is capitalized +Fixing the findbugs warning +Could you please give out some more details regarding the time constrains of such a task +Two years later +We need to push thisWe cause many many major GCs which let the application hang for nearly of the whole execution time +Until Ive fixed it set loopbackfalse in UDP or TCP +How would you like to proceed +Patch for orgapachecommonslangtexttranslate +This old attribute was replaced by specification to be consistent with other service dependenciesGuillaume do you see another inconsistency in the +Id wait another couple days before trying to build the branch things are a bit wonky on it due to the +Yes I am able to pass properties between filters without any problems it is just when I try to pass a property onto the request where it is failingSo how can I make my filter add a propertyheaderparameter that is then retrievable from the resource? this is your requirement inject into the filter and then access it from the resource +Stand correctedThe is an escape code that can be followed byN for nullB for blank for a literal four hex digits any other unicode character +Note that Im rather unfamiliar with s infrastructure so Im not entirely sure whether this was the best approach +If you finish it before we release beta please correct the version +I mean this automatic stuff should be a conservative default that gives you reasonable performance +Please show us how the complex schema type is mapped to Java in +Much cleaner. +If you have a system where a user can setup an import from a database the limits they might want to put on that import can be very different cause one imports something from a one from oracle one from a clustered db one from a db wich is in other use as well etc +Removing the target dir manually causes the test to pass onceThis happens on winxp dont know about other platforms +So what about I create a class that has the native SRID handling provided in to be used as a superclass for JDCB filter to sql be it prepared or not we eventually try to make the two code paths a little more similar we close this one as wont fix +Thanks Brent and BrockRev cfaeedcbdadbce +choose some by default but let user have capability to change +A is to go back to the previous version of the plugin ll create another issue about making this configurable +Test failures are not related to this patchThey were failing in several of the previous builds as well +However I am a bit conflicted about where the parsing of the job history file is happening +I have similar problem on Firefox when I using Two Columns Layout +Stefan this looks much better +Hold on I wonder if the script has been updated to take the latest from GIT instead of SVN +Tested it on a dataset with snapshot files and it works good +Updating with +Here is a patch for french translation with much more fixes than this quote the quote problem I just wrote two simple quote which leads to reported problem both quotes are rendered in the report +Still have warnings for some of the proprietary connectors which Ill have to look at so I cant close this ticket yet +please with rev +Moved to please move bugs back to which you will work on in the next weeks +PawelCraig pointed out that is following the specification wrt to property names and I see nothing in your post that persuades me to change that decision +Updated summary since the described behaviour is arguably an error in the default setting +This way in future if someone adds it back inadvertently we catch it + verified +He may be open to switching to using the Sonatype OSSRH repository or at least deploying releases there +The HOD test suite does not test that options are passed to the resource manager hence I have added not testHemanth can you review this patch +I am hoping that we can pick up the changes in without upgrading to or can either browse the SVN log for the branch or look at each of the issues included in the release from Jira +My problem was that when I sent message via SAAJ it gives me proper SOAP envelope from server back while axis gives axis fault +Hossman closing this as I dont think it requires a change + +Thanks Paul for the feedback. +If you find further issues please create a new jira and link it to this one. +By making the class be functionally outside the scheduler class and thus unaware of the latters data structures and by moving the generation of the display strings to the concerned and objects the synchronization problem is easily addressed +How would be useful? If the two are different then rpcaddress is the only one thrift will be listening on +Its not that big +Sorry I have no idea why but it seems to be working tonight even though it wasnt the last few +AFAIK the webview enables CORS policies when the location of the page is anything other than fileBest bet is to download your remote assets to the device and access them via file +cool. +It should not be difficult to port to trunk +I guess I should have searched for duplicates before raising this JIRA +Im looking at too +What version of Explorer are you usingcan you locate an entry in regedit for GUID +It seems the current patch is missing some files though Ill reroll it +Sandro good news! This is issue is not existing in Alpha once we switched to new MapReduce impl +If we are going to do this lets get this in nowThe test failures look unrelated +In Y patch I see that the user who started the is added to the admin ACL +Hi ErwanIs the patch still appliable as is +Thomas any chance that you could have a look at this for still? At least for an initial cut that works across AS and ? ThomasThanks for looking into this so quickly +It is hard to imagine these headers introducing this problem +updated the maven plugin configurations to output the pdfhtml to targetsitebookrev +committed Revision close for release +Jonathan you updated the existing tests for DISTINCT so it is covered +In case youre curious the also uses a file which was originally a WTP class that was liberated +I guess for now I was just looking at us doing things consistently +Fixed +Switching all issues that have been in resolved state for more than one month without further comments to closed status + +Looks like there may be a Subversion issue +im running continuum on mac os do you mean by how i run itvia terminal +Patch depends on works great +Taking issue +none of the bundles is installed then you installresolve one of the bundles and then install and resolve the second bundle +It would be a step inthe right direction but still only a baby step +All fine +I decided not to delegate because of issues with components including without having a template +It used to use the masters file +Bulk close for the release +Ill spend time on this now before you pull your stuff over +From the Resolution looks like it is still unresolved but from the comments looks like it could have been fixed in some versions +Definitely tell users to upgrade +Thanks for everyones patience as this issue was fixed. + upgraded to and now all tests executes fine +I assume the nightly build does not pull in the snapshot of Forge +But This is a sample code +Attachment has been added with description description in I submitted this bug I modified the charter itself so this bug is CLOSED +Harder than anticipated +How would one do stemming stopword removal etc without some extension like thisPerhaps we can add some validation to the solr indexer component in order to limit the detected languages to the supported ones +I can confirm that Stephen is leading this for and the PMC +Ill commit this shortly +The problem is that out of the box Hadoop uses +Thanks Jim +Could a message be logged instead of letting the exception +So I committed the first patches along with the cleanup patch and I backported the simpler integrity check from trunk +betaCR users wont be affected by this the tptp plugin from the as feature could break update +lgtm +It replaces the with exception +xz support is there a need for +Thanks for your feedback OwenIll give a try today and let you know if it helps +in Todd +I dont really know if its valid to say JDBC assumes ODBC version +John this issue should be the same as Rob is working on it +A second addendum +looks good +I plan on committing this shortly +through JMX or through custom management code hence a cannot really be relied on +Thanks Clint +Christian I agree with all your points and am somewhat persuaded by your argument +When this adaptor is more matured we can clean up and create as abstract class and expose as something that third party can implement + tested +Why dont you allow updating existing ones? Is there any technical reason for thatI have the use case of trying out several different schemas +Fixed in . +Fixed at think we should make very sure that Vijay wants this behavior even for the short term +Ill leave the JIRA open until HBASE can get this working properly +If it is not needed then you can remove it from the MRIs parser as well +Patch fixes the delivery count issue but causes the following tests to fail test test and your patch fixes the deliveryCount problem and still keeps all the test passing +Latest patch has not improved things much as far as this specific problem goes +According to their online doc it seems as allowing CCC is a configurable option on the server independent of explicitimplicit FTPS +As Im checking right now it looks that it checks existing crawl folder and recrawl all the sites again but is it possible to filter them out? So to only crawl sites that we setotherwise I think it very useful patch +Many Thanks for all your help +Or did I miss something else +I fixed portlet title handling issue at Modified files arecomponentsregistrysrcjavaorgapachejetspeedomimpl componentsregistrysrcjavaorgapachejetspeedomportletimpl portalsrcjavaorgapachejetspeedtoolspamanager portalsrcjavaorgapachejetspeedutildescriptor +is now working for us but also was changed to be more alternative library friendly +Hello Oliver Thallmair This trick definitely works well +In fact this is Sonar Runner that prevent us to pass projectBbin because Sonar Runner wants a list of files or patterns matching at least a fileAs a workaround I pass projectBbin +Ah see I did not see that page either +all default values in all attribute tables should be meaningful +Assigning to Jonas since it is closely related to along with see there for further details +Attachment has been added with description SVG result for test has been added with description small test program used to transcode the test case +Well actually the are clearly and precisely +Added fix for failing test +I havent tested yet against chunked encoding +So there should be nothing to be done in the backend +looks good +Sorry I wrote the correct one but had attached the wrong file still doesnt pass all testsThanks for having a look Im committing it since it is anyway completely independent from the rest of the library. +The idea here is when the region is tried to move away from the RS that holds the META the region is actually closed +We are not limiting to attributes +But the initial version of the patch should give you an idea +bq +so I would suggest to use that solution instead +committed +The other issue still remains +This has been merged +Consider this done. +Patch a contains a test for this issueThe test has not been enabled +Do you know whats happen? How can we exclude these jarsThese are transient dependencies ofbq +The bug report is accurate +Accounts merged. +patch was applied to but not trunk +and if you mean a long series of switch statements then those are certainly more expensive especially if each if clause involved a significant compare or preparecompare operation +The repository eventually took care of itself automatically +Thanks JasonI put this into trunk and +If that are not in the buildResponse when errors occured in jsonValidation process my Ajax call return only success moreless It should return errorThis issue appeared for me since strust remove the and in the buildResponse +The equality check depends on the type of working memory +noone available to work on in right now. +Anyone oppose +The patch failed these unit tests in bq +ie +The problem stems from the fact that the class only allows deprecated Expressions I am adding a check to allow Expressions +The demo contains a protocol two filters and a handler +I think youll need to specify this behavior via an additional parameter +Both fixed I think thanks for the transition all resolved issue that did not see any further comment in the last month to closed status +In a freshly restarted Eclipse I get this error but after a while working on other files the problem disappears +Thanks Milind! +Patch looks good to me Pavel +Same doesnt fix the problem +This is a DB v bug or limitation as this does not affect DB vso those test case failures should be marked as failure expect +So none of your changes to archiver are actually in tested a fair amount of downstream projects with their ITs too so I do believe were in a better shape now than ever thanks to your thorough will publish a this afternoon with your changes +Shouldnt the C client docs state the sasl package dependencies? +I think i still changes the macro renderer for now +Its currently change would be applicable to the and the. +Ill use the storage location and the schema on LOAD AS to generate the SELECT statement +Also I use Spring integration as specified in docs. +If users were not ever expected to leave Recommends unsatisfied it would be the same thing as Depends +When another Makefile includes the path will be invalid +patches updated patch for trunk for for ve updated patches +Its important the is independent on any other frameworkAlso classloading should work with JEE app servers and whatnot so its not trivial to doThat said has a classResolver which is runtime agnostic +Since there seems to be an agreement towards keeping it and you documented it +roughly moves things forwardIve added stuff for Long Short and Byte parsing changed getAuto to a static cachkey method switched core Lucene from using the API to the new API added some javadoc and roughly have things going with the reopen methodIn short still a lot to do but this advances things enough so that you can apply it to trunk and check out the goodness that this can bring to sorting and reopen +Mitesh You bring up a very good pointI need to investigate more and see if there are alternatives other than doPrivileged in the openjpa code baseThanks for you insight into this matterAlbert from a performance standpoint it seems like we should make the methods create an interface and provide access to the instance via the object +Resolving as fixed. +If thats tested and proven then I dont see an issue with using the code snippet above +The entire tomcat dist structure is back in the bundle +If it is really needed for let me know +System Property set on the console +It is confusing +See for the result issue. +I will disable xhr level entirely for now until I have debugged this out properly +Ive changed the target version toOne question I have you say in the description that youve developed the patch set against Hadoop but that youd like to refine the patch set until it can be committed to Hadoop trunk +Are you sure this is what you want? To be clear there will be no lucene downloads available if we proceed +This fix is available in Scout which will be shipped with jboss resolved issues for already released versions +these goals always move the version forward from the pool of available versions OR leave as is +If you want to create a that does that it would be great +Daryn Linux and Unix allow users to continue writing to a file even it is renamedmoved +interesting because i thought you were injecting an which case a jdk proxy is used instead of the cglib one + +in any case a file is written to the repository home which holds all the parameters that were as in the +attaching a new patch that solves the previous bug that failed in grouping partitioning columns +If the bug is really Trivial then at a minimum some documentation in the installation instructions on a workaround would be extremely helpful and not such a hinderance to adopting the DBThanks! +I know Queries are a pain because of Serializable +Combined trunk patch integrated to trunkThanks for the patch Francis and MatteoWaiting for Lars decision on patch +Its a DEBUG message +Any objections +The additional patch applied to DRL VM at rPlease check that the patch was applied as you expected +If its a zip file can you try it and setting the contenttype to zip during the a lot Jeremy +This is a mess I for sure helped makeHere is viewItshadoop Master +Christian Schneider is currently investigating if it would be possible to keep setter for release but without clearing collections +What we find is that when we get an we do not have the required linked exception +Attachment has been added with description transition resolvedfixed bugs to closedfixed +Updated In my first posting I pasted the error message with wrong line numbers now the line numbers should represent the current JAR files +Thanks. +I modified the test cases to be a bit more thorough +fixed now with regression tests +It was which timed out during getting a nw server connection +Apologies for delay all donePlease allow time for our mirrors to pick the release up +Here is the version information from the version version name version for the comprehensive bug report +Patch for using application New MIME Content Type constants in Replacement of hardcoded content type strings in files affected by +Also it would be good to have an elrte editor there rather than plain texthtml +Firefox do notThanksTorsten +Attached file is sample strace output +Fixed in make that r committed changes to fix borkedcommitted in rI attempted to add an assertion to this in the integration test but could not get it to pass regularly +Im renaming the issue because I think the problem in the description is fixed and I want to open another JIRA for different failures on the same tests +Downloading the poms you mention should take less than seconds +If the user wanted system privileges but still wants to use LDAP or custom authentication the credentials database could still be used for authorization purposes for the system privileges +The reason this constructor exists is to allow users to plug in an alternative normal distribution implementation to be used in computing normal approximations +Since the patch for was created first that will block this issue +if app crashed attempt will receive CONTAINERFINISHED event and then attempt goes to FINISHED stateIf the node crash attempt should receive EXPIRE event and attempt should go to FINISHED state as well +I am not a huge fan of +From the other side conventions themselves are more than just proposed they are agreed on +Nothing in the logs when i clickYes was running a snapshot build from February +some of subtasks are for beta so issue itself is for all moved to remaining subtask is pushed for since its intermittent +META +the link above should be of closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Added patch for branch vGACP +They both can be committed to DRLVMDIRdoc directory +I second that I dont think such a change would be useful +All tests passes on my box +Log attached +What would happen when two clients connect within the same nanosecs? Whatever time unit we use we need an additional counter +I tried and the issue is gone. +has been marked as a duplicate of this bug +renames the method as Jing suggested +eclipse file that crashes when i open the file using the visualsource this log doesnt exists errors related to vpe does in eclipse install dirrectory exist any crash file +This JIRA is related specifically to documentation +Thanks Saminda for this patch +Removing a line in as attached will address the issue +Some things that should be verified before it gets committed It would be good to testemulate the case where the bin scripts are run as if the project split has happened TODOs for things that should be done when the actual split happens should be minimized overall Does it make sense to have symlinks in the current HADOOPHOMEbin directory that points to the new HADOOPHOMEcorebin files +But once data is converted into vectors the general code you are creating should be able to take overBoth implementations can write the whole matrix and take the same approach to +I think we could put it under the Labels I did not intend to make a blocker have done some changes but still not been completed please do not commit its only if someone want to review itThanks patch but still something to be implented not ready to be commitedThanks screenshot +The Wiki examples are also pretty much out of date the rar adapter and JCA are the preferred ways to access foreign queues +Im fine w removing it for now but note that it may be restored in the future if a real need arises +Attached test passes +Just deprecates the various ctors etc that are removed in the trunk patch. +In this case thread basically waits for thread to finish and then returns whereas it should start its own new commit +bq +Attachment has been added with description for reporting the bug +rpc server is very busy and start dropping connectionThere are caused by permanent failure conditions +Revised version. +an addition to this issue the unit test was testing the order of in that they were applied in the order they were provided +Forgot about this already pushed out +Would you please attach your script and ideally some sample data along with any errors or other relevant info that might help us troubleshoot and reproduce +When I apply this and try ant I get No module named setuptoolsShould the build get simplejson setuptools etc +From hard codes +Patch to clear Counters cache for trunk +backport this issue the fix +The roundtrip reduction hack is a feature that can be extended to any sasl mechanism that can initiateNo it cannot reduce for SCRAM or Kerberos while my patch canbq +This was part of the debug output for but this is a separate issue +I still need to test the patch will report back once I am done with thatThanks Allan again for the work +Done Editing goes back to Screen +Thanks. +Matthew +But Im pushing it to instead of closing so we can investigate further +Reopen to edit do not set with an Unreleased version +Marking for +Hey RobbCould you please try with Forge version and let us know if the problem persistsThanks manually deleted the plugins from +Ive been just committing as I go so I dont think we need a lot of review for these things +This error also seems to happen when you return from a subflow WITHOUT persistence context to a parent flow WITH a persistence context +Sorry I catched you at the start of your working day +Im going ahead and closing this as invalid since it isnt a software issue. +I created a doc bug +Let the messages drain and one or both of the queues should now have negative counts +Here is a patch against the latest SVN screenshots to illustrate the problem +I checked this with Eclipse plugins but you are right Ill try the build +I feel that this issue is more likely caused by +The Beehive tree has moved to beehive +With this change we get something on a GCL that configures the event though the is supposed to be aprt in the configuration of the GCL +the overhead is particularly large if i only invoke projectN probably want to move this discussion over to mailing list because it sounds like your buildfile is orders of magnitude bigger than what everybody else is using +Sorry missed this yesterday +I would rather go with creation time rather than with modificationExcept that only implements modification and our creation doesnt match the semantics of any posix concept +We cant +Isnt it the very definition of a Bag allowing unordered duplicates +I just committed this. +Thanks again RickCommitted patch to documentation trunk at revision +locking at the code I guess its easier to use the special prefix ujax Ah you are right +William thanks for bringing this up and submitting the patch all the same dont take it personally that were not incorporating your contribution. +Small thing in the new unit test test it would be good to actually check the resulting list of streams +String value combination +Yes +But as a developer I am sure you also understand the unworkability of maintaining bug fixes for all versions you have ever released +I will submit the report patch tomorrow +The change I made was to accumulate remote repository definitions starting with those passed in from settings and other initial locations and proceeding down the POM hierarchy as each model is scanned for extensions +BTW Id like the patch committed if possible since thats what were currently working with +they assume not empty which is different. +With the upcoming Hierarchical Mode of the Schema View I think you have what you were looking for +Chuck did you get a chance to look over the patch +snakeyaml approach +I have update the patch with file fixes also +as an optimization it could copy the logs from the shared dirI dont think its necessarily an optimization might actually be easier to implement this waybq +This patch makes it possible for a aggregate to have different input and return types +No we removed it +The ugly thing I see in is the single huge file stored in zk +YARN app kill API can be used terminate the DAG +Ive checked in some changes to and that implements the WC rules for locating a start file +Actually looks like will need and before its up to speed enough for this change so can probably hold off on for nowIll commit your patch for trunk and lets follow up on the test timeout in +To resolve this issue we should allow the test to pass if this occurs +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +push to +the v patch into and trunk branch +Would this be difficult to fix +css styles was improved marker should be improved it seems a bit rough +Oh heres a question how do you change these properties after the collection is created +Only diverges in meaning +In our case if the quickstarts from the project run on platform then weve done the goal is to manage all of those users who might start on the project then decide at a later date they wish to be on the supported bits +It does not matter which one of both is currently selected it will always resize to the size needed to display its largest entry +we should egt the message returned by the namenode and fill it in the exceptions created +Hi Jonasthe patch is in Trunk Sascha sorry for the late replyFinally got a chance to look at your latest fix for this but it still doesnt work properlyI see you changed the mapping which is great but it still misinteprets the dates +I think youhave some good ideas in here and will be taking a better look when I get some time +This illustrates the issue +httplocalhost +is unpacked after installation and that makes +I dont think weve changed this file since yes it is the ESB project readme +If users update dom tree by DOM api it is very difficult to update feature tree synchronously +In hadoop export was exporting to a random dir in hdfs while import was expecting data from the hadoop file system +Verified at the EAP release notes +It does not expire sessionsthat is still controlled from the Leader in +Thanks! +Comments in the forums +I will try to reproduce this on directory. +This is failing quite a few builds now we have eliminated several other minor memory issues +I recommend looking at X from maven in the perform run to see if you see any clues. +I think we should revert this patch from the CP branch +We have moved this to future and will review for. +Ill try tomorrow to merge trunk into the branch +Either you define an action using the method or XML but not a mixture of both +This issue is caused when the package on the repository is compiled with a different version of MVEL as your process runtime +Nick +Rebased to latest trunk +I cannot propose any fix without a test case to reproduce would need to study the root cause of the class loading errors you encountered +Heres a patch that updates the build to create a separate jar +overall +Hi SebbHow would I do thatDenis +Test fails when executed in a ve commented out the failing testcase in the test until it can be fixed +Thanks is not a blocker +adding the path to the request context so handlers can use it and we can make a nice debug +Modified to not synchronize when updating content since this can clash with the synchronous event listener being called during a store operationFixed in release +Looks good overall +Obviously your new patch is better than mine +not really doable on unfortunately +Uploading +As far as my thoughts on knowledge needed in a looping context Ive got nothing to add that I havent said in the mailing list thread +Which method is it that is causing the problem? Is it part of a public API or an internal method someone found? I cannot believe a public API method got removed there are deprecated methods years old that have not been removed Sun and now Oracle seem to believe public API methods should never be removed so why mark anything as deprecatedI have now stopped using Groovy for scripting because of this Grab and JDK problem +Im happy to test and review a patch if youd like that +is now used to access remote repositories +Perhaps this is already fixed +Im testing a fix +OK new SNAPSHOT deployed with a possible give it a try and post your toRelative +R Compendium sources have been committed with initial M POM on commit . +Assigning regex bug to Ola since his Rej engine is so +We can close it I think +committed the patch +I have no intention of porting this fix +Fixed the url property for jdbc based binary stores +Thanks +OK thank you for your great work +Not sure if has been deleted or not but if you create that dir the tests should passsudo bashsu hdfshadoop dfs mkdir useroozieLet us know if that works +I actually want this change in +Thus for fixing this +But hopefully will make it +Generating compact is not a trivial task +It could be an issue caused by hbck I tried running it a few months ago and it didnt work for us and Im now thinking that it actually put that invalid +Ill update the patch to include changes in docs and some tests +Kind of think we should push this to +Looks like the issue happened during the standbyactive transitioning process of the namenode +I believe some topics need more information particularly since it describes what all the clauses do +The result is used for recovering a pipeline +How we use +Although since this is a mistake that a user could easily make it seems to me that a more error message would have been preferable +Just a fresh deployment +Thanks Kim +bq +There is mention of XPCOM error +bq +Fixed in trunk +It lacks binaries and has wrong +I didnt say every sentence was relevant I just said some of the additional information might be helpful + to trunk +There must be some solution in hbase space for this +committed thanks! +As I mentioned before Im running in all my environments so thats what it will be easiest for me to test against +i am looking on how to override the domain class of pluginthanks a lotcheers. +This report is for two separate issues related only by the fact that some logging is done +Move it to the right fix version +One more exception is occured if if include the whole schema in Table +I was more thinking of a declarative way of doing this in the war plugin itself a la of the dependency? +Same problem here tooCaught Totally agree with other comments Groovy is awesome but come on there is so much potential here unfortunate we cant get it to workHaving a rapid dynamic way of invoking complex SOAP services with Groovy could be very slick and efficient especially compared to the traditional JAXBCXF +First attempt at a patch +Forgive my ignorance of the development process I see there are several log entries of people making changes to version +Fixes attached +This patch adds an option of specifying as a path relative to the location of the executable itself +Tests passed cleanly for me on +As we have a workaround and that it is not related to an issue with the framework of felix I propose to close it. +Looks good +If you find out please let us know +pwdusrsrc wget ftpblablablabla tar xvzf ls export XALANCROOTusrsrcc export XERCESCROOTusrsrc cd usrsrccsrc make Maybe it should be wrote in the this was a configuration error. + guides where looked through and no one screen showing content assist with incorrectly sorted proposals was found +can you attach the wrong and correct pom +This new Scanner API is still +My own interest is mostly in matching the documentation and the unit tests though thats how I discovered the differences +I read the ppt that published on the HBTC It seems not address to this issue +Thank you +Thats interesting stuff thanksI kept on turning in circles in the docs always returning to the pages I had already read not finding the information I needed +It was better for +The mysql dialects have not changed between and +is related because if maven version and comparion were implemented too it would also cover this use case. +This is because if the user goes back and then hit the Deploy button the existing cluster is deleted and a new primary key is assigned to the cluster +Superseded by. +It uses the Java Preferences API thats already being used elsewhere in the Console to persist the font size changes as well as the height and width of the input and output has been tested successfully on OSX but I believe it should work on other systems as well with the encapsulation given by the preference API on other platforms +If it is broken could I file another jira to make it work for IBMs javaAllenYes kinit will regardless of R unlink the file openwrite the principal openwrite the TGT +Please apply the patch +There were no commits on this ticket +I think we still have these memory leaks in the current CVS codebaseWe should be fixing those before had a look into this and fixed a leak due to non allocation of channel objectsNow I get defnitely lost bytes with the valgrind profiler +just testing notifications to the list +This patch makes the changes Knut suggested +If anyone disagrees we can reopen. +Ok I did some exploration on Peters machine and Im sadly no closer to having a solutionReproducing it was pretty easy though it wasnt consistent +anything else will be handled by the existing interpoliation done by sling properties +I suspect the culprit is +Closed after release. +First why does the number of trees vary like thisHmm + +Beautiful work +If thats so it opens a stream to the resourceThanks +Not sure this is really an error as the point of the RAW transformer is to encode the entire incoming message into a binary payload inside a JMS +closing this issue assuming its fixed +I havent put the eyes on that before +I will file a separate JIRA for that issue once completes +Can you describe what the problem is? There is too much stuff in the output and I cannot tell which pieces of data Im suppose to be looking atThank you +Ive fixed as well as some tests from package to make them compatible with new model since these tests start to fail +The same thing with that the join does not even have to be on the same table +Thanks really nice now +Your problem is similar to +did you set the includedir flag? please post full configure s automatically set by the build recipe if I claim its an Autotools based buildconfigure build host target prefixusr execprefixusr bindirusrbin sbindirusrsbin libexecdirusrlibthrift datadirusrshare sysconfdiretc sharedstatedircom localstatedirvar libdirusrlib includedirusrinclude oldincludedirusrinclude infodirusrshareinfo mandirusrshareman homefearlgitdcimagebuildpokybuildtmpsysrootsbeaglebone no no no no no no no no no no no no no Problem is includedir for a cross compile when youre supposed to use a specification for is bogus +This is a user interface issue rather than the tool +notNull install finder class +Gotcha +The patch does not contain any author tags +Fixed in CVS +Do you have an example? If you are thinking of the sqlj artifact this is because the jar is closed source and not avail on centralallowed to distribute +Thanks MarkNew math implementation integrated in repo revision Please check that the move was completed as you expected +I think this can be closed now. +Dupe of see solution there +Do you have a preliminary patch to look at +All tests pass except for test in the streaming contrib which is an unrelated local failure due to a forked ulimit v not returning the expected output +Also includes the change of to Map +js is that its scaling all background images to their container warping them horriblyLets remove these scripts for the release tomorrowAfter the release I found a jQuery plugin that works for background images called Superslieght that Ill give a try +Update in kit release notes +However the RM should not crash if the user input is bad +Broadly the patch moves the internal class from to FilterThat means that to add a Filter to a as a clause the user will need an option whether or not to have the Filter take part in the scoringSo with this for an option of scoringnon scoring in the API for adding clauses will be neededwhereas without this the possibility to add a Filter will be needed +Making a class final might have had a performance improvement in Java but in modern it has no affect +Closing. +When it comes to code running it cant find +So it may be not valid + +Ill take this +Verified on JBDS. +Commits are in ER +Do you get that error when the source tab is maximized +Mind OL is doing its computation using single precision floating point numbers is using double precision ones +Checked in the fix to trunk for last comment +Yes +classes can be shown +r +Moving to as it is a new started to review this patch but decided to wait until I had a chance to review the Conventions plugin in the sandbox to see how well it jives with what we are doing there +configure CFLAGSmarchcore CXXFLAGSmarchcoreThe g commands will come out a little redundantg DPACKAGENAME +Thx change applied in r +Fix +This is a neat new feature but Id have expected it to work like query string parameters work if they are not mentioned in the request mapping then it matches +Hi FilipWhere can I get the said jmssecured example to reproduce this issueThanks FilipThanks for the info +Verified against SVN +Merged to BRMS branch as a side in BRMS m assuming that this is related to the checkboxescheckedunchecked are now dropdowntruefalse issueIs this JIRA just related to the sample being broken? Can you supply more details about the fix +Regressions passed +sln has the setting whitch contains but that file doesnot exist +Committed to trunk +Last time I checked seriously at that I remember giving up because that would require a substantial refactor of the repair code and I wasnt really sure what was the best way to get started +The recovery is not as sleek as it is in +This has been committed to trunk. +Ive used the first approach as Seam already does a similar thing to set the +If you try and launch the Job with the old parameters the will simply throw an far as the original request goes the is a interface that you are supposed to use in your application code +In the meantime we cant let this block the release any longer +This always needs to be the first Advisor in the chain in order to provide meaningful was simply a bug before where the actual order was even dependent on the order of +Some products come without the Console +As I see it The is caught by the and turned into a +drools components in the tools directory removed as a consequence of the old console being old drools jar left is in seam which comes from eap see linked jira +Also I want to make sure that sstableimport isnt broken in the same fundamental way +Hopefully its good now +You need to add this jar to +Thanks! +All the other packages need to audited as well +But I dont think my test program is same as the case described in the release notes +When I discovered that I stopped chasing them +Lets change them in +Right +Updated patch based on Adrians that adds changes suggested by Leon based on ve tested a combination of Adrians approach and Leons comments regarding rel and uploaded a new patch +Sorry for the trouble. +I can confirm this bug +I have streamlined a number of the versioning operations and made sure that all versioning changes are persisted strictly before they get referenced in normal workspace content +Will need to think how to address this. +The trouble with your repo is that the hard links apply to the git objects not the checked out tree so merely creating a second copy is not acceptable for you no matter how its done +bq + +I just tried with a fresh branchx checkout on Windows +Patch containing work to date SQL parser and JDBC driver +Thanks Piotr sounds like a good idea +Since the implementation may choose to use the active fetch groups to prefetch data the test cannot test this behaviour +We should not start moving back to the world in which we have a mix of XDOC and APT docs +This is a knownproblem +Patches appliedSuccesfull interop test sendreceving tofrom java and +Bug has been marked as a duplicate of this bug +You mentioned this is your first iteration do you have changes already planned for this patch +If youd still like to take this Sachin please let us know and we can discuss on the forum how to fit this into the new API +Im happy to announce that this was fixed with +but then I dont get why the test fails strange. +Thats interesting +Hi ClausI did notice the nabble issue yesterday but I hesitated to create a Jira issue and start work on it I was waiting for Johan to create the Jira issue +The aggregator project should be treated as a special case +Changes should be visible on the web site within an hour +Are there any news on this issueIt seems a lot of works has already been done but unfortunately we only have the initial patch from attached to the issueWould it be possible to have the current status of the code even if it is not complete yet +do we need a reportingManagement element +Please reopen if you have an example file you could share. +Thanks a lot Todd +For after putting something into the log about the exception I check if the server has been shutdown and if so then we should leave the otherwise we simply close the client socket +personally I am ok with changing an inconsistent file to consistent fot tab vs spaces especially if you are actively working on the file and it is causingproblems as above +I dont see a Derby policy file which fits this use case +My issue is that the code that is generated does not compile +Updated version of patch available to replace thisLooks good +A very raw patch that I was using during the performance testingWill clean it up and add tests if any possible +trunk +I also updated the example to break it up into two trycatch blocks + tests decorators for request and session scoped beans indirectlyI did not find any appropriate test for interceptors going to create one +No need for +not blocking final. +Also can you set the svn to native makes submitting patches a PITA for windoze users +Nice +Well of course the myspace record is wronghost t txt descriptive text krsa prxgcnhThe is wrongly excaped twice. +The question is if the properties are set soon enough +Found the bug +Only the first of these is a failure +Nicolas great catch about that lock! That may have led to a deadlockThe access issue can be ignored as it was just a trivial thing +Can you provide a concrete example? For instance show me perhaps an example htmlform and the intended URL you would want. +I committed this to trunk +It was using the default +Done in trunk +The idea with listener was to avoid the need to cache the view id itself after updating the address cache since this would two operations to the cache and would require batching or transactions to make it atomic +Looks pretty complete to me now feel free to add stuff. +Code looks fine +We use maven in the Andromda project +Looking at the XSLT spec It is an error if a stylesheet contains more than one template with the same name and same import precedence +resolved in issues for released version +What are the suggested workaroundsIm pretty sure this is now resolvedFor point that wiki page has been updated with the proper form for the For point the wiki page has also been updated to reflect that the proper name is endpointName not portName +Ive found the problem +Update test to expect different exception from +Hello +The issue is not reproducible after the issue is fixed +Ill take this zip and unpack it over the eclipse folder and apply the rest of your patch +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +tim Looks on trunk +lgtm +think its unlikely that what youre seeing is related to the issue of this half your messages are getting stuck most likely cause is you have two consumers on the same queue but youre only consuming from one of them +Until then we could do more harm than good if we start fiddling with the code with no way to actually validate it works ourselves. +That doesnt matter +There is an option upgrade Buildbot Master and this sorted request will be the default action +Typo in the name of the param in send API docs +One problem with the time is that it is start time of jobtracker server which wont change until restart +Done +Committed to branch +Mice need your help on this issue + +Starting the second NN +provide a testcase with the full mapping for User too +Id like to sugegst two simple changes that I think would really help the community out Rename it to Friendship factor Provide a link to a page that describes the metric every place the metric appearsStephen agreed. +Glad to hear that it works for you nowJuergen +Why? Cant we just treat it as that upload field failing validation +Awesome +Thank you guys for your ongoing endeavor trying to solve the mystery +If the module is actually unmaintained would you consider becoming the new maintainer +Manually tested! +In this case the content is not encoded at all and directly streamed to the repositoryWe should improve the documentation but Im not so sure we should always buffer +The issue has been fixed in M +GiriI added in the jackson dependencies twice because avro was in there twice with a different conf each time compilemaster and compiledefault +When built against this version of the POM all unit tests pass +Yes create operation is also working fine +It was only found after writing the security random walk test and running that on a cluster. + +The patch itself is just one char change but it is not trivial to create a unit test for testing the log +Do we expect all required plugins to be in the IDE when this link is clicked? Or can the link check to see if required plugins are installed and prompt the user to install missing plugins +if the writer dies how to make the file available for writing without waiting for the next block report +Once this facade works it can easily implement the proposed tracker interface +Zipped files from Capps +Eclipse and other IDEs are much better at this +I cant avoid having added to my interceptors even though it causes the credentials provided to the constructor of are private so I cant access them when configuring my +Which attribute? Can you show the place in where this attribute is used +Thanks Hairong! +But I think this is wrong since the referenced dft uri will be made relative to the server which could be different +I dont think anyone is thinking of that at the moment at least I havent heard of itBut could you be a bit more descriptive in what you mean what kind of scenarios +It should be withCan someone with the proper authorisations change thisAnd concerning my original comment this still leaves us with a regression in sonar for which there is no fix or work around +I forgot it is the same issue as youve already spotted uses the The contentType property will be ignored in this case as a JSP will always use its own content type either from a contentType declaration or the default texthtml charset +Probably distribution +hmm do you think its correct behavior that we only support flags at the front of the args? Linux ls for example is fine with ls l +Also I dont like s combine functionI dont either! It seems like Lucene should rewrite then search against each searcher rather than rewriting all up front and combining + dedbceffebbffbbcbranch since is now released +MridulPlease try the attached patch and let me know if it fixes the problem +Logged In YES useridIt should +Any progress on finding a real fix +There are optimizations vmware has for windows this bug has to be kept open Ill mark the priority downward to minor for now for there is little we can do from cloudstacks perspective. +Sounds good to me +Issue is overly general no value in keeping it open +I still recommend that be deprecated and would say that the RI should do the request attribute as an internal implementation dependent technique +When I grepped I noticed that we dont fix connect in +Added a at upgrading to The suggestion box now appears at the left top of my page A list of possible options is placed underneath the input box but as a flat list The scrollbar of my div disappearsI will attach some before firing suggestion after suggestion box is provide page of a bulk update all resolutions changed to done please review history to original resolution type +The patch seems to be ready for failing test is unrelated to the patch it has been failing intermittently for some time now +Hi StevenApologies for late replyI do not get an exception in my scenario but the request sometimes gets blocked and we do not get a responseI have noticed this when we are getting a connection using theIt works fine if we get a connection using theIs this a new bug +This may be related to i found the same mistake in a few places of the code +I encountered a NPE in case of multiple root entities when deleteQuery is not specifiedCommitted revision + an issue in staging or prod +It ought to be possible to get fairly close but itll certainly require someone to spend some time preparing documents and testing themFor the colours we should probably take the discussion the poi dev list as itll almost certainly need some work on POI to expose the information before Tika could use it. +Are you sure relative paths in property constraints work in Jackrabbit +tomcatx and tomcatx now go through the standard tests. +What do you think +The chokes because of a missing service that gets installed later as part of a transaction +The release notes look fine except for the issue listed in s leave this JIRA openresolved until we get the final bug list in +looks good +This seems like a very good idea I would like to see the components messaging and take this into accountI am very interested in seeing how this develops +Uwe I think you need to change your grammar to take advantage of the new features + Simplified the code + we want to apply this patch before or after we release version +Im dont know how we can handle the shell scripts this way +Using instead of as per earlier discussion +We need to think about whether or not it is worth it for us to do so. +I am trying to link Analyze the Job page to and see that the job used is different +doesnt cause the locking issue that exists anytime that you use a single Connection for both send and receive operations +Running the latest patch through Hudson +Does that fit your schedule +This will change the API a little too but all use cases only require an iterable collection +Julien why do you have ant there if it is not to be used +Updated in ESB codebase will be in next merge +thanks arun for contributing! +Krasimir Would need some more details before we can help you include your whole filesystem then there will be time problems +Was problem when deleting the connection +Fixed by adding weldcdi dependencies to and file +I will attempt to attach a subset of the xsd but I am not sure what I will be able to include due to business restrictions +ij +would seem to imply that theyre the same for all instances of a single job +Attaching patches by module +Rakesh Id suggest either update this patch or create a new jirapatch with your contribution +Sounds like youve been bitten by +Here I try to minimize the change on NN and refuse the same methods used by journal sync +Will do +Note the workaround disable the audit for the JDBC persistence adapter +I think the other issue can be handled separately +Remove HTTPBASEURL from expected values +The original fix was backported to so we probably need to backport the fix for the compatibility issue too +The tests store and retrieve that serialize to around K and K bytes +Sorry for the screenshots +But lets update the roadmap when its done you never know +I think theres another JIRA wrt how we scan the classpath and I wouldnt be surprised if this was related +push to +Could you provide a specific patch +Other user having similar problemsNot sure what is causing thisCurrent common denominator is Windows OS possible proxy configuration +Thanks Nadya verified. +The IP may be often that of a load balancer or proxy that sits in front of Solr +Committed the backport with revision +When you use TSDC its easy to fill because its already known +Tests are running nowTouches the following filesM javaengineorgapachederbyimplsqlcompileAdds method to look for the offending referencesM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceAdds an error message to flag this forbidden conditionM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangAdds regression tests +dont know why it failed +Ive verified this brings us down to warnings and committed this +Thats what I figured +Arent these types and states defined somewhere already? Seems like theyre a public API and therefore shouldnt have to duplicate the definition of them in taskTypes and taskStates +so lets draw some picture to editor is no longer being developed +Actually creating this as a new a few chapters will be included in the product doc should be less than a dozen pages Currently creating docbook version of complete community guide as publican compatible maven project +I have removed this test for the time being +merged with the last sources +This fix is working fineWill this fix be part of Servicemix or will it be updated to the earlier versions as well +duplicate of. +Hi AlexanderCould you try if you can reproduce the problem with the attached distro jarThanks DavidThis seems to workIs there also a multibundle file for thisThx +I also like the idea of sorting before flushing but Im not so sure how much it will improve performance +If we let all flushes go then we are overrun by store files +Good catch LarsI think the fix should go to +defer all issues to +Changed version constant in the codeFigured this was better than adding a version number field in the conf file +Look at local list of plugins that match our nightly list of plugins and do a replace with +Verified with daytrader sample of the web context root appears in the EAR list page so close this jira. +how can you still be auditor if youve lost your zk connection? Yeah its true +Therefore at this point this is being rejected. +the new patch solved the problem +But this is the policy we have +needs to be fixed as well +verified in to add release note text +In what case cant you make it easier for your users by putting it in a repository? Whats the case that made you ask for this +this just as a side note +I see two different issues hereMe too but Im not sure they are the same twoOne is a bug +The was getting incremented twice +Included is a small maven war project that demonstrates the problem +done however the issue is still open +patch added test in. +Let me know otherwise +bq +Agreed this can be quite confusing +Does anyone know what license it is +Since the delta will be zero its now only byte only a bit worse than bytesOne more thing here since the tiis are loaded into RAM that unused proxPointer wastes bytes for each indexed terms +Our users especially production ones are not very keen on changing their processesI think this probably an ok change since our users do not have schema files right now +SenakaThis is great news thank you +Andre Youre right in Betae this issue does not exist any more +Verified with RF closing. +this should do it +need a test did not respond to request for test case +FYI Michels moved on and is no longer contributing to HadoopIs there any chance that we can get his code posted as a patch even if its incomplete +to it +Can you run it from within Eclipse afterwardsI am using the sysdeo tomcat plugin +Closing all resolved tickets from or older +We use our own Hyperlink framework that extends Eclipses are three main types of objects Recognizers +Maybe that can be used here? +Was this reproduced in TeiidEDS + +Is this reproducible on +That way the beforeUnload event is not triggered +We need it to fix issues and have linked these issues as dependent +This may be related to +The client will simply backoff a bit and call get again +I canNOT access the EJB with test context set from outside the JVM +I just verified the procedure successfully on RHEL +Brian No I havent done any progress yet +Philip as a workaround you can specify default JVM arguments in Window Preferences Java Installed +If you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +Tested on windows +Sigh +users can make map reduce jobs run on one group and another group hosts the MR result for online applications +The extension point is used in JBDS only to install Seam Open On handlers for the +However is not a production release +id say max Ah OKI just tried russian w whitespace analyzer used to split and did get a good size savingsUTFsize +This contribution was made with the express intent of being integrated into the mainline as requested by my direct superiore I am not aware of any s licenses or other restrictions that could applyIs this OK +Google Chrome is an alpha product +Well we can augment the model to express if its literal etc +I forgot to stop this +Thats why I just didnt go ahead and do it +Please try the latest snapshot. +Closing all issues created in and before which are marked resolved +If no one really uses hardlinks why dont you get rid of this altogether +But Im trying to get rid of a separate Interface specific to join +But will take more simplicity in implementation for some reasons like dont have to remap to additional virtual node layer structure is easier to extend in rather than leaf node etc +Should this be closed as a WONTFIX because Gant behaves like Ant or is there a behaviour different from Ant that should be preferred + +he explicitly suggests the property i used closing the ticket as resolved based on that +I checked DB on Gentoo again with new knowledge and it looks like destructor appapp is actually executed +In addition anyone can find out who the actually contributed to what through a few simple svn commands +Committed revision . +Could you check with the latest snapshot build? Ive committed some changes that should improve performance +After turning replication off on the slave cluster the +bulk close of resolved issues. +Original patches wont work +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +let me know what you thinkBen +Fixed in commit +Done +Hum this issue is apparenlty affecting all users +For the second I will add JSTL to the archetype and Jettys JSP support as a dependency of artifacts +Slip non essential backported from upstream. +Could you please check that you have only one copy of groovy jars in your classpath +Thanks hossman ill get to some of this shortly. +The underlying issue here is that the port forwarding dialog is not sharing the very same code as the ssh keys wizard +is it actually ASM version conflicts for example? What else +Yeah I suppose +We can talk about this as +more investigation isas far as I see is already enabled so some issue resolvedconfirm the patch applied correctly. + +Closing resolved issue no response from reporter so assuming ok. +Im going to commit later unless somebody else wants and has objections +Some of the situations were currently trying to help application teams with are things like call took msec to process call took msec to process +The solution consists of a custom web client built with Wicket web framework and a possibility to deploy wicket components alongside the workflow +An updated patch for bug fixes +it either extended or shrunk only one contiguous vlan range +You can find everything in the Runtime window component now +The scope of is broader though since wed like to review our use of ZK there +the trycatch should be around each aux service method invocation so a failure of a given service does not affect delivery to other services +Thanks Jim! +It looks like the cache should have been one where the values are weak +Thanks for the reviewing Ivan and ChrisChris I noticed that and I tried to restructure the patch but the conflict seems unavoidable so Ill leave it as it is and rebase laterCommitters please consider committing first +sounds reasonable +Also updates the testcases +patch that fixes xxx tests +Reproducing the issueThis is reproducible but not with a Unit Test case +Ill have some bandwidth this week to work on and issues +I think a better way to do this is to add an encoding property to so that we can encode and decode the data correctlyThat also allows us to do codec changes if neededDoes that make sense +In our case the name node actually reported all the as live +So at this point I think we cant do too much regarding this m closing it as wont fix for now crossing fingers for the problem not to be bigger than what we expect as having two libraries in different classloaders is a pretty common scenario in app servers Im ll reopen the issue if a fix is needed that no workaround would solve +There is no posibility to separate both operations of loading and removing + + +Changing to a more appropriate summary and setting fix version to +I saw this magic version before and that comes from out of nowhere +Youre ignoring all the other suggestions and alternatives closing discussion then asserting that the competition is between new the point about which the disagreement turns and that its time to write code +Does the wait between attempts need to be configurable +This version provides a template for derby embedded derby network hsql embedded mssql mysql +Fixed in +pushed to master +Ill do that tomorrow +if you have remembered exception throw it after in r +looks good +I just tested this on my macbook running leopard and windows XP +So I have no clue to give you +YeahNN does not know the block construction stages +hmm +That means Kelvens disk chain sync checkins were not into the buildsetup +howeverI would like to note that in the ant version systempropertySomeString will not work as it will actually look for the property named systempropertySomeString instead of just SomeString +Will commit shortly unless there are objections +Just as a quick side note when submitting patches please remember to follow the Velocity coding convention for braces +There is a test named jdbcapi and initially I thought this was the right test to use but I can not find that this test is doing any calls to setgetThere is a test called jdbcapi that does a lot of calls to setget which are disabled when running with the network client +Thanks Cedric! +That explains it +So removing the filter is the first step +Thanks for the help and quick turn Olivier! +Hi Markuscould you please attach a simple test projectFred Java resources are part of the projectclasspath but web resources should not be in the classpath +Fixed discrepancies between the and the docs +To the best of my knowledge this will improve things but wont solve them +It is small easy and just makes all builds failing. +It would it allows you to add a descriptor to a rd party deployment without touching the archive +I disagree +Both should be part of the projects developer documentation not the product documentation +Please see comments above at Nov +Is that by design? Is there a switch to turn some debugging on +Just got the permission to work on this task +Moving out till questions answered. +But it seems like theres probably some dependency confusion here like one of the libraries loads in MRI but not in +distcp initiated from and permissions turned OFF data copied from to WORKS +Ive run times on my machine and it passed on all +Committed +It isnt usable as is but with a little help it could solve the problem +The patch incorporates Raghus comment +Well obviously currently it is not +Im going to try another snapshot to see if it is back +We noticed this before and are in touch with the Tomcat team to do something about we should present this specific case to Filip as well +Fix checked into the branch. +Bertrand are we agree about the testcase +Posted patch to wrong issue earlier +Oleg +Can you please upload a patch for that as well +Could you possibly repro this WITHOUT dOSGI? I dont have a good way to set up and debug this possible Aegis issue in there +fixed in r +Lets see if it passes. +Note that the repositories content is still fully recreated on each test +Attaching a simple patch for the patch +I will shortly be creating a new bug to cover this and will link it here +This is the backward compatible version of Chriss patch with this the object can be created using the old relative paths or the new component sintaxPlease review because Id like to commit it soonThanks +It works. +Committed for in rev +So how do you get around the for each warning in your code? Every for each in your code will give that warning unless you dont use for each +I had cleared the cache +Would you mindproviding a test case for the bug or shall I take it over from here? test case suite may server as a good starting point +Thanks Ralf for importing the comments +If I edit the dbf file with openoffice and change the boolean values from FALSE to F the exception goes away +Uploading shortly +Possible should pay more attention to bug reports +Hi Jeff Im guessing this broke just recently with the latest BPEL build from +Handling the payloads distinctly is the best way to do it as the interface doesnt have to change +Once as a declaration and once as a reference +This patch adds support to +a stack trace + +To avoid problems keeping community enterprise docs in sync please submit for documentation issues +Fix it +thanks for the clarificationmy concern is that how do i in a code review determine if this doesnt introduce infinite loops that were once broken out of by the IE +Doh might want to limit maximum size of stringValue that would get displayed without being truncated +If not the patch would need to make changes to not only Kill command but also other user allowed commands like Suspend and Change +Victor please verify +Please verify +You should also look into some of the issues the reported +For VNC you would need to connect to that address instead of guest IPhostname +yingDo you think your patch can solve my problems +Implemented a check on the possible error markers present for the project +I replace plugin all work fine now +And thanks for the issue titledesc clarification +If is excluded from Jersey bundle where does come from +this one is so old that if the provided case works Im happy marking it resolved +hope I am reinventing bycycleI believe some databases do that +can you please attach a screenshot of this? I am running FF on linux and dont see a difference between maximized and normal sized with Chrome cant see a difference +Reverting line in makes the exception go away +Fixed at bfe +During the recent updates this UTF search broke +Closed on request +Sorry I screwed up with the various mojos there +I dont like returning null in getLanguage but I couldnt think of anything more elegant at this time +Could you apply the patch above to the latest Santuario trunk and let me know if it fixes the problem? The current code is a mess I created to refactor it for the release next year +But if there is a bug please continue open bug reportsThanks. +Is that covered in the patches at this time? +Also many of the comments indicate that the API is subject to change so I didnt think much of backwards compatibility due to this statementregarding the user seeing the benefit the common wont but from a developers perspective it just keeps the code more organized and cleaner +This case cannot be carefully tested using Selenium framework +In this connector it is the users responsibility to register the application and come up with a context tokenMaybe the right way to do this is to provide a script that does this kind of registration for each connector that requires it +I removed the code from the description sorry for the confusion re stdstring +Thanks for the review +Their icons are always shown what about the second situation when after selecting another css in package explorer toolbar icon becomes disabled? If we decide not to add duplicating functionality then we can leave it as is but to my mind it leads to a little misunderstanding +The default value for SEAMRUNTIMELIBRARIESCOPYING isnt defined correctly +Sorry for the delay I missed your comment somehowAttached test case demonstrates the problem +code checked into cvs HEAD +This fix changes readbuff to do nothing if readCredit +checkstyle +I am a bit overwhelmed by the codebase +Do you mean merging indexes of two solr cores? I have thought of exposing that as a command +No +I tried to use the emory dcl which does implement a fair lock but AFAICT it would require exporting from the framework which I dont think is acceptable +close off release +exceptions also need to be changed +I am thinking of creating new blob using and getting the stream from this blob to use in the test class you have written +Should we now be looking to use the new version of mina now that it is approaching a stable platform +Paul yes that may also be the case TBH that is why I kept the description vague in the issues I have raised today some of the schema inconsistencies I have spotted could be a missed back port to EAP if EAP does actually support the changes. +Yell if it isnt! +During RSs it tries to assign the region by setting it offline first +! +Michal can you have a look and if you cant reproduce then just close it? Note this was tried on Tomcat + +Please try the following patch +idState Bundleosgi Havent tried the updated branch +OK well for issue is still a bit too big for + set fix version +Right if the head cmd was run before the ulimit info was captured then it will only be in the log and not in the terminal +Congrats Charles Thank you very much to have to reopen this fixing it to not try to shut down the child process causes some specs to hang +Also your fix ends up being cleaner than mine was. +removed site target as this was causing errors as it tried to locate the artifacts on remote build timeout to to use maven hoping to prevent those build error with the text Processing failed due to a bug in the code +Can we get more info about this bug +Done as far as I know + is a limitation of +Another technique might be a job completion URL +Of course multimodule builds may also exhibit the symptom found here but they also obscure the essential bug IMO +Is this code still interesting? That is are there newer tools in Lucene that handle this problem? I have found searching movie titles product descriptions etc +See also is a constraint system issueToo complicated for me to fix +Where can I find the snapshot to test +Thanks Bao +How long failover takes is very important +thanks Henry +Thus far the only thing that changed is in the where the is no longer required +Preparation for Gora release candidate +Ted and nkeywal Thanks for your comments and patience working with me on this Jira +Attached file +When working across multiple it is very difficult to make sense of the Build Time +This isssue is always a real problem +Ok I will try the stress toolJust to be sure You are talking about and not the stress? Because I was trying the java stress and it did not accept the I parameter +Were getting a NPE instead of a exception saying something about corruption or an inability to read from the store +One final note If you are testing this out yourself remember you need to install the proper fonts on your systembrowser for the characters to display correctly +BenIve seen a couple of updates go by whats the current status of the namespace +Vaibhav The test passes even without your patch in place +I see why youre doing it but its unusual and potentially suspct there is a way to call reflectively but I prefer to avoid reflection hacks if we can possibly avoid it +So likely allowed you to execute the command which hangs +Your test case does what? the problem +This issue is not critical forResolved for only +Reverting status to NEEDFORINFO as this was not imported from +Move issue out of the release +content has been added to BRMS Admin guide still need to add clarification about using password encoding optional content has been removed from the BRMS admin guide will ship as a smaller doc when the modeshape files are made available review +if it were to be left as it is it would also be helpful to put a comment there saying these will be directed modified +This seems a pretty severe bug +Will commit shortly +fixes the problem +Bulk moved to properly on my Ubuntu +Please note that I suggested a Future mainly because Spring Integrations async gateway returns one +I thought I was running +Andres want me to apply it +The bug is not reproducible in trunk +Works great terrific job! Also refactoring the package name is fine especially if that aids in your development efforts +It really is the best solution +Eventually the Start page concept would be deprecated in favor of a Index page. +Its good to know but functionally it does not matternightly is for sure not staging staging is something you use before going to a releaselatest doesnt say anything about what it isnightly at least has a clear signal its not the latest release nor is it something stagedI agree its not a perfect name and even though snapshot is a bit too much like maven its probably not the worst name to usebtw +I just committed this +Im with Mark on this one +Yep but well keep the issue open until we have unit tests to prove that property selection works for +I think the tricky one is unregisterRegarding I dont think it will be done until end of July at the earliest +will support bundle deployment through deployer +Although were still using a Snapshot of we actually made most migration effort and we can consider this issue as closedWell use release as soon as its available. +Committed revision makes some initial changes for regular and embeddedI hit return too quikcly on the commit and so missed giving credit to Knut Anders Hatlen for the changes to +It will affect performance +Either way is fine Ive just never like using Pair to hack multiple returns In keeping with the style though doing the Pair return seems more consistentYeah a new patch would be great Andy! I was just thinking about backwards compat it would be nice to not lose it and just add a new method with the new return that does the correct delegate method calling in +This will be it from me for now +The changes are OK for meThanks Ivan +something likewarning more than one found using sounds fine to me +Please check if this problem was induced by the last fix to atVersion id +And that is that the compiler does look for the class but does not fix the meaning of the name +Thanks this to trunk and +I have purged my maven repo and this fixed the problem above +Thanks Aaron! +I only want to retry when the client hasnt started communicating to the server yet +Thanks Namit +linking to the issue I found IPC and the security manager +Please check that the patch was applied as you expected +Just as a +I have the same problem exactly with the same is definetely not at least using single processor does not resolve the ll attach log file where you can see that artifact is succesfully downloaded and just next can not be resolved +Yes but using a synchronized block maintains API compatibility by keeping themethods as instance methods +Skipping outright seems a little extreme without first determining if the image that is being requested to deploy exists or not +Thanks Jacques Scott and Leon for your contributions +Not every patch I review should look like the code I would write +Closing as wontfix. +I forgot about one thing +I think its ready +The replacing happend because of my way to insert my translations I load the whole XML file inject many translations and write the file back +I came across this just about a week ago when looking for the next task to do +Thanks for the report SparkThe fix and the regression test for the second issue was committed to SECURITY module at rPlease check that the fix works for you +Its likely to be very slow +Have changed a number of files to depend on updated versions of me and but Im not sure that will fix things +The builds the message passed on to the persistence manager +Those problems existed before and were solved before ivy +Andrew please verify that this resolves your issue +I already work with different coding styles in different projects so Im passed the concept of the right coding style meaning that I can live with these drawbacks +fix in the axiom with revision +Committed to +running thanks Syed! +You should use cforEach +However we need to be careful with this +Its not clear to me where this should be added its not our normal patch formatAny suggestions +Seems like an issue with either the zookeeper client or Java NIO +Yeah I can now drop a xml file +Thanks Nicholas! +Adding a new configuration option which basically is an ACL in a servlet is not the right approach +it should be there for sure but doesnt solve the problem of many quickstartsAbout quickstarts in AS then lets move that to a separate issue we should actually try creating such +Fixed in trunk ce +to commit from a docs perspective +IT also seems the API in is insufficient to properly handle this and will need to be changed +Sergey this is great informationSo the crux seems to be the classloader reference if the pool is to be shared then it must not refer to an app specific loader +Dmitriy i should be able to test your patch on my issues by tomorrow +Running in packaged app works fine +Implemented for configuration support in Rev +Commit the changes based on the patch from Xiao Yi to at and trunk at Thanks Xiao Yi! +Users can still configure the default authenticator since goes through anywayIs there anything else in ARQ that uses HTTP and needs to be updated to go throughIn the meantime I will start adding a new documentation page explaining the new functionality +It is probably ok we just need some time to review and apply it +bq +That is certainly more flexible than a simple toggle switch but I couldnt come up with a reasonable use case where it is used for specifying something other than +Still need to change sleep job to use delayed submission +tried to fix this pre yarn move to branch +Removed minimum component length +will be done in a few. +change log for code results of the CVS +Heres a more elegant fix up for review that doesnt require the initial build and allows the Jenkins machines to take advantage of this fix as well +Committed to trunk. +Tentatively making this understanding that if we dont find a solution Myrna will this +The RFA rewrite should fix this issue +could you rebase against trunk +Im hoping the change to file names is acceptable +get goal didnt existed by the time I made the patch +Isnt the plan just to remove these in +I could not reproduce this bug at my environment but I think may fix this bug as well +Just to make sure before I can start diving into this the patch for needs to be applied correct +reserve commits in indexversion ArtemYou have been quite accommodating w my requests + +Thankyou +Thanks for the explanationPatch is good to go +Is the right thing to integrate ivy into recipes +The cause for this seems to be so replacing with in issue title +We did that and the runtime penalty was really noticeable +Attached is the patch which waspasted as a comment and now is attached as a diff fileAlso attached is a standalone programwhich reproduces the problem for me with the currenttrunk +Still have some Shale stuff to review and update tonight +Commited +Dunno maybe that idea could translate to this situation is the distance metric is proving problematic +I mean we could provide a SLD that has a point symbolizer by simply uncommenting the central part of the style but will that help in any way? In that form the user is supposed to know what hes doing as he has to hand type the an alternative we could provide a blank slate and give the user a drop down with a set of styles to choose from +fixed in SVN just needs an integration test +Ill work on it tonight +Haha JD still wants me to try on Android +WIP in +Hi Nicholas Thanks a lot for the patchLooks findbugs are related +That would remove the need for locking on it entirely +to trunk. +So I tried the test with and without the patch to and it behaves as expected in both cases of the time. +Sure thing please open a new issue when ready I spotted some going on yes +Im under pressure of this issue very often too fix will be very welcomed +JrmeDo you have a simple example portlet that has what youre trying to do that you can attachThat would certainly make it easier to determine what the problem might beThanks version are you usingWith I dont see that problem + +This is also the likely the cause of much of the duplicates +try and set default catalog to lra which seems to be what is read from mysql +I dont know why I thought the screenshots were a problem. +Marking closed. +Rebased merged into and into master. +As Bobby mentioned and Ive mentioned to Owen offline Ive really been wanting to configure the login context to use an auth other than kerberos yet still use a secret manager to generate and validate tokens +I didnt even try to test nested tables because I didnt have a problem with it +My gut tells me that this doesnt seem entirely bad and Im not seeing a lot of situations where people are thrown off because presumably they are very familiar with the testing framework of their choice and at least enough of the configuration parameters overlap to get them started +MarthaCan you please build this in brewThanks in CR build +I can do this but am awaiting feedback regarding issueNote A problem with loading LDIF files was found in which extends the + configuration should be done in +While this may look like errors of unsafe concurrent access on our side I dont think thats possible + +Reopening then closing to deal with migration bug +sure it was +I apologize for that Nick +doesnt have this problem +As per the Sawzall paper the field ID is required for identifying the field in the binary representation +Maciej please check this out. + +This issue is fixed and released as part of release +Further I pulled in the domain model from the original project description on +The original reason was to avoid showing intro first time to keep VPE working. +Any help would be greatly appreciated +Created and sumitted a patch for this +Hi Ronald No I think both this and my original test project use +I added the in the maven build but it still adds excluded appenders +I would summarize the problem here is exposed by the need to select alternate rendering scripts in this case via slinginclude +Thanks for your response +I have this on my local machine now and will look to commit it to we can take a look at it +Ok I do take that back +Adding new patch generated against the apache trunk +Someone might want to check on that +I mailed the openshift team and Im waiting for explanations +Unrelated to this change we should be using here +Please verify +Verified that ouput of hod command has been updated +This has been completely refactored a long time ago +Viola the root cause is that the start command could not start server correctly under IBM jdk at server side +In the case where there are multiple types the maps will be joined +I tried to apply the patch then I got few test failures so I fixed the bug getting some data from patchThx for the re welcome thanks for taking care of it so quickly! +The patch attached does not evaluate WHERE clauses on expression columns in the materialized subquery resultset correctlyThe predicate of the outer select is flagged as pushable and ends up being pushed into the subquery for evaluation +Ive checked in some changes to bring cargo more up to date with the jetty have not yet been able to test these so if David or someone else has some time that would be is still some more work that could be done +If binary compatibility is a concern please make note of that on the contribution page +Re my previous comment entry sorry it was Werner Punz who made that comment in reply to Laurie. +Obviously the scores das das das das higher than das alte testament +Changed from depending on apache commons lang to our own implementation in +gsp and the controller that is passing data to the view? If no can you at least copy relevant snippets of those files into the issue +I dont think weve tested the server on Java just yet +No newer ones +If the issue is some thing other than this this JIRA needs more information +It would be great to get this fixed before goes final +It has to be exactly the same email address as the one you use for your and confluence accountsIm in the process of identifying an owner for the module +Moved to am having issues with this again +Though from a practical point of view the one is the most usable one we focus on the one + +As per the descriptionIn what case will an be thrownBy that I mean in what case will a different as opposed to be thrown +Sounds OK to me +on second patch +Assigning all documentation related issues to final since they dont represent a change to the in packages cannot access classes in the default package this is a Java restriction that Groovy doesnt help with. +Details are on the release would like to get this automated when maven builds the release +They may be related +weird decision though the web controllers are among the most utilized classes at least for us +patch looks good uwe! from my will commit this together with +Thanks +Even if the dropins part works I get exceptions due to missing plugins from the p chain +The former represents a lists of integers with their differences the latter elides leading zeros in integers +We need make sure that the working idrsa key which is usrsharescriptsvmsystemvmidrsa +Therefore its bug in Harmony +Search for facilitate upgrading to prove itBoth show all changes so it seems not restricted to RSS +Yes that works! I will add a test case for so we dont break it in the future +This functionality went in into I wonder if it wouldnt be cleaner to close this again and perhaps any further adjustments should be headed under a separate bug +I just create a new JIRA to trace the related issue +Fixed under r. +Hi sure which line break youre referring to +There was a bandaid in I think so you could try that +revert the Im perfectly OK with your suggestion and it would be great if you could provide a patch +Ah your too fast for me Emmanuel +Thanks that worked! The artifacts deployed has confirmed successful deployment +Thank you for the explanation i think i have to dive deeper in Apache CamelMany thanksWith best regards Dario +Only the optional SDK doesnt work +sources in zip +Without storing things forever theres no way to tell them apart in any scheme +I tried applying it but it handles only the source code the tests and the sample need to be upgraded as well +bulk defer open issues to defer to not essential to address for +I dont understand the urgency panic mode though +Closing out the issus for the release. +I think this has timed out? +Anyone have an reproducing test case to attach? +Clearly not breaking function +Thanks Paul Gyuho Song. +xml files and checked that the binwhirr in the new +So if just add richfaces capability to such projects then we get conflict of aj have to remove aj lib from project before adding richfaces capability +I just looked at this problem and it seems the resolver is able to pick up the schemas +Both Jetty and had issues so moving this out of. +Why disallow a user from setting a lower limit? If someone wants to set let themGood point and I thought of that right as I was uploading +If you find further issues please create a new jira and link it to this one. +Not sure if you intended to include changes in your patchBTW here is another version of the patch with no need to serialize the +Is this what you had in mind? This code is completely untested and probably doesnt quite work but Id like to get some feedback before I clean it up +I was able to fix it by recreating the old regions folder moving the right store file back and moving a +seems like this is a bug as who really has a network that never drops a packet? +dmg I was using qt MB +note that the attached file has? instead of spaces displayed +is perhaps too generic how about +Sigh +The existing is slightly refactored and can now run one recommender against two different data models +links have been fixed. +This therefore cannot be fixed by the teamIf you want the community to come to an agreement please use the user listThank RobertReading the again it looks like that the root cause is the fact that the non is not set properly by MavenAs a result revision has just been checked in and the Maven proxy settings are now used for setting andThis should make things clearerThank Savasconcerning I actually had the same guess but wasnt sure +in other words they will be ignored as subtrees for the transitive dependency calculation. +Please review +updated patch for +Pushed to master +Its fine I used the new markCan you test the new code on your systemAnd if you have an unit test that you think is good to add then you can attach itYou can check the unit tests I added by the commit log and the new added tests from the source +Patch created against src synched to CVS on +Added a unit test which fails without the patch and passes with the patch +Yeah unfortunately the directory is scanned on datastore is really a datastores wide issue a few of them do cache infos and theres no general way to force a refresh +so far you have not explained clearly what concrete usecase would require thiswhile you think this request makes sense to you you have to consider how it will impact thousands of other users +Still left to do deployment chapter repl chapter last bits of web chapter +You might right Deepak! I havent yet tried to fix the code it may required additional effort to fix the issue completelyI have put another comment to bring the attention! +Ill reassign it to myself and make sure i patched properly +Doron Thomas Ive opened that covers site plugin getting the module artifacts from the would be nice iy you could provide a sample testcase project for. +I am looking at committing the backport +There are now releases of the xml beans maven plugin for each of the versions of the xmlbeans library +Issued against master and +We have the same problem with Websphere Portal +Looks good +Supun +Does it really need to waitRemoving KILLWAIT is quite a bit bigger change than the current one +It will be better if you can take it from within your project itself gsoc +This is also addressed by +Retracting the earlier comment +ok +I missed the boat on this one as we were focusing too much on actually getting Any moving +The section on document management is labeled as Important This is still an experimental the jBPM Version bits are the same we should keep the docs the same too +just wanted to see what this is about and whether its still pending +Ive upgraded the Turbine dependency to without any ill effects +An alternative to fixing this issue would be to add a null pointer check in as described in +One thing I noticed immediately is that you did not implement get for the driver and effectively disabled all calls to this method +Please note this would only fix GML not the previous ones unless required +It should be possible to do this by just extending the generic I think itd be good to get as much of the implementation into so that boilerplate code can be shared between implementations +Tilmann could you eventually create a reproducer? It should be ideally in a Warp functional test suite? +Patch addresses the findbugs warnings +Paul owns JBTIS JBTIS owns Teiid Barry opened this issue +The stack traces are attached in +Wonder how this carries over to eventuallyThanksRobert +You can see from the comments that lots of people have suggestions +As a workaround I moved all the configurations from that file to and deleted the file +Dima wasnt th einitial version of this supposed to be in alpha? I cant see it in Alpha build I only see Generic and +Thanks +Probably something in the other code +The only possibility is that an Error happened and this was certainly not a as it was working for quite sometime before it stopped +Thanks Stack! +I think this patch would be relatively easy but requires a nice test +When doing distributed mode then you need to edit the ZK config +When Ive applied this patch I noticed some problems with closing tags in +IMO there is a category of config +This has been completed +NB is bound to GML syntax in the filter will be using WFS and support for this is a separate issue +i have an RPM that could be released after a little polish +Cool +Sorry this was invalid +Closing all resolvedfixed issues of already released versions of Roller. +Apart from unnecessary storage and retrieval of the same type again and again in this column noticed that every object load query that happens also has a beantype ? condition as part of the query +Already fixed at some point +The purpose of these is to prevent larger container requests from starving +Confirmed! The bug is fixed! D +I guess that it would make the most sense to add that test when that issue is fixedI think we need new bugs after this one goes in both minor or even trivial One for the ij exit code when the file is not found and another for the if the resource is not found with +Sending srcmainjavaorgapacheservicemixeippatternsSending srctestjavaorgapacheservicemixeipsupportTransmitting file dataCommitted revision . +resolved conflict for the commit for CREATE FUNCTION +Patches applied +Heres a patch which implements the changes described in the original issue filing +Developer error. +Cannot reproduce with r +Changes namespace to urninfinispanconfigNominee for the most trivial patch of the year award +Odd +customInput in the example is a compositecomponentIn other words a composite component inside of a +But it does no harm as far as I can tell +Attachment has been added with description the original docbook transformed with has been added with description the FOP transition resolvedfixed bugs to closedfixed +Hey PatrickIt depends on the scope of as there are two stages where plugins are needed first to go from Avro IDL to +I looked but I didnt see one +This one does not provide any supporting jars +Thank you Jing +This patch looks great! is so simple now +So I am with what you guys choose as long as we silently do not ignore failure to stop proxy in non mock cases +after completing item +Clarification this is a only patch +If you flush and compact before the last put you will see your timestamp cell +Your configuration is wrong +This issue was fixed as part of a SP on Remoting +Attached the same patch but for the branch. +diff u based on code base. +Applied the test +I was using it to understand how the api worked and to confirm that it would pick up modifications +This property may only be used to rebuild the past in a chronological orde +Server side support for seconds pattern was added +Brianthe zone creation is blocked until we receive some working memory to get it in hyperion +Can you add that message here as well +Patch applied to branch in SVN revisions and +It works thanx +All set +Does anybody know if this issue is still open with current groovy versions +Attach the final version of the patch +Attaching the request by Denis project structure +Note that I added only setget methods for the context +This needs to be in release notes since it changes the reported the SYSTEM TABLE type and return all table types as upper case +The change looks good to me +Fixed ages ago +Is it not possible for you to do something similar on your platform +I havent researched this but I assume this applies for only? I have no idea how this functionality work with respect to collated indexesIve also seen examples where you can modify collations using SQL. +Wow that was fast! Lets see how the goes +Patch for looks good and applies just testing and then will commit +I update a diff of file changes created with Tortoise in . +bq +Why are our tests not freaking out with this bug +We at least accept four params now but something is not quite right +Patch makes senseCan you integrate it into? attached to retry only on +Obviously the pull requests from Magesh need to be pulled in before this gets added to the tooling +bq +publish maven artifactsLooking forward to see your proposal +Done in trunkbranch closing stale resolved issues +You could also argue what the added value of this check isSo my suggestion would be to drop the acceptable method + +I cannot replicate +Thanks for your reply +I am wondering if we should apply this to so at least the problem in is partially fixed in trunk as well before land and we also get more testing +Elad Maven always uses unique snapshots +Based on current implementation other members will eventually think this node dead and perform failover for itAs we know performing failover for a live node may cause duplicated messages so the failover should never happen +are you proposing that there be one thread to renew lease to all the namenodes +I still see the same problem as noted in comment +Just waiting for to be released +Changes to address this bugEmbedded JDBC driver at the end of updateRowdeleteRow the rowdata in the resultset is reloaded from the cursorClient JDBC driver removed the use of update at the end of updatedelete the rowdata is reloaded from the cursor Tests testcases added covering both this issue and theNote by reloading the rowdata from the cursor this fix also fixes issue therefore these tests are part of this fix +messagedigest uses for theseRight it uses the destructive get which is why we need the duplication +backport +You are correct +removed KEY special case +Interesting +we check for this by checking the discarded property and write a marker if theres room at the start of the discarded regionAh right +Our unit tests look pretty sparse in this area thoughLet me know what you want to do +Still need to check in the change but I deviated from your patch +For my particular calling is actually undesirable +But of course not on Windows +Viktor then just resolve it +starting the server by giving h option with ipaddress on ipv machinesjava start h afaAll of them works fine +See the test client that uses security +Let me know +Ultimately being able to manage multiple stages starting immediately with slow start would need some changes to the way vertices think that they have started +looks good +Do we need a way to pass other properties into the EMF? I like what you have here Mark seems nice and easy +Oo spreadsheet for sizes with Axiom +Therefore Im differing it to the release cycle +Im quite perplexed on what the interplay might be thoughAlso the SM fix I did the other day seems like less and less of a fix +Applied patch by German +Im going through several transactional services but my mongo repository is not marked transactional +If there are other cases which need addressing please open a separate JIRA or a subtask +Version that handles used Scan attributes to setrestore the reverse flagStill need to fix the MVCC stuff for backward scans like done in trunk +I have opened to track the duplicate NEWLEADER packets +Logj was modified so that wherever filtering is allowed only a single Filter is used +When a property is found the always uses a instance to create the value for the property +The attached file has a test case in it that reproduces the clarify a junit test case for ant test +Although they produce different hashes in the result coming from checkschemaagreement but since the client does not know how that hash is calculated it cannot decide which one is for which change +Done for that I havent tried to reconcile the contents of plexus redback repo and redbacks redback reporedback symlinked redback into plexus the old redback repository that lived in plexus old file listing new file listing +The change was applied in revisions and +I cannot find where this was ever do not see anything in the suppose you could have true and configure a war to use the default loader repository for individual wars that you want to behave like this would not work for wars within ears +but this task is too generic +O forget one thing +Working fine on fresh workspace with JBDS +Resolved +I will share a complete class diagram ASAP so that we can review this +Its not a or common enough requirement to keep an issue open for when there are more enhancements that need addressing as well +Hi have updated the class level Javadoc for the PPHC to highlight the issue you raised Adam + + all resolved issues due to a successful release +It doesnt matter what the max redelivery count is set at if the consumer stops the message ceases to be redeliveredThis seems like a big problem with transactional acknowledgement mode +Thanks Prajakta Ashutosh Brock and Alan for your help and support. +needs to be discussed this is currently possible using the functionality but this might make it easier +in revision + +The problem probably is caused by the positional code which just traverses the parents components position a separate CSS positional code which also traverses the css info has to be added I will if time permits look into it this weekend +I just mean that the problem might come from the same cause thread visibility memory management +With the form proposed in the description its not clear how Derby would have enoughinformation to know how to convert the values in cases where a simple cast wasinadequate +It is nicer to have config file generated to HADOOPCONFDIR without storing a copy in cwd when invokes + +This is cool +Attaching simple test program used to output the diagram series +Patch version includes changes for review comments from John +Hi HadrianHere is a revised patch for this issue it fixes the failures +Installed JBT nightly build +Hows it looking so far? Are you ready for the switch yet +Itll need to be done for to go out the door +Proposal discussed with community esp Rajith Carl see doc links on parent JIRA for detailed proposal output +I need to install svn and DBI by myself for commit +Hey Scott pick any unassigned +If I comment out javascript include for then inputCalendar worked fine on IE +example to be shown in the UIThat would be awesome +Any help would be greatly appreciated +Closing as a. +Unfortunately we have had no on this and cant reproduce the problem +Attached is the test case requested +Just seeing how it fares on dont accept more than you can handle +However the Configuration in the cluster comes from the cluster deployed jars for hadoop common and that does not have +The error codes are never ignored +Ive tried many manifest changes but without a you +Im happy to look at that tomorrow +After switching to there is no such problem but last step is empty +I hope one fix would address both components +This issue has been resolved for over a year with no further movement +Forgot usage example in to current trunk patch generated from a Solaris Express box +If user typed file with different extension manually it should show an error it preferences dialog titleThats why I couldnt use this yes +Heres the patch that for me makes the difference between a green and a red one +Was committed a while back +This was already completed +Try shapefile and you will see that it behaves properly +I am particularly interested in a solution that can correctly set the id of inserted child records +But that doesnt solve the need to have a reference to in order for the target file to pull in the latest version availableOr have you found a workaround for including xulrunner and in the TP +It seems unwise to swallow exceptions tho +The static class in for seems unnecessary +It had some methods which where newly introduced in JSF to be overriden +release +Resolved for a while +Would you take into consideration sparse graphs or dense graphsWeighted or unweighted? Directed or undirected? Or all the possible combinations of these two +Closing all tickets +Thanks Hairong for the review +Either way I think we can do the code review since were already using the old BTW the second patch should probably be submitted on top of the first instead of as a separate thing +Hi Jaspercould you explain me how to apply it? I cant find the right file to apply the diffThank you a lot +The CWD of a daemon is used for core dumps java profiler output etc +Is it planned to fix it for the current release +Something at the Jira end must have changed +RalphI attachedIt contains an Eclipse project with source and binariesThe top directory has a file and two scripts to run performance tests +What is Bug ID ? I looked it up on the old bugzilla system and that didnt seem to be the bug you are pointing to +something we can do after the feature freezeWe need to run jdiff etc +hudson seems to have missed this +Added check to prevent classes from being created in Git ID eecdbbcedadfaf +oh sorry got the wrong logBut this log is even more weird +kit NPE check +I think it might be related to the fact that we had to use recently +settings for ejb project +The problem would occur when users are upgrading their cluster to from +I know the multicast discovery is fancy that like you do now you dont need to specify any url and the client node will join an existing cluster with same name but I think I would go for the other type of client here the which is more lightweight and just sends requests to the configured urls in a fashion using the internal binary protocol that elasticsearch uses for communicationLet me know if I can help more Im certainly willing to get my hands dirty here if you want +But I would like the documentation component to stay on this bug to make sure that when we do fix this bug we update the documentation to remove the current restriction +Tests passed cleanly for me on +It seems that a better fit would be in Part II +hmm I did try move s logic here it yeild to about compression rate did you think it is acceptable? I will see is there simpler random way to gain similar outputs +I have verified that the emails notifications are coming through now +Test Case adjusted +The current unit tests adequately cover the testing of this internal change +db parameters etc +the latest patch looks good to me +Additional question is when are you going to ship next stable release +Can you attach your patch files to this jira and select the appropriate license grant option +Correct me if Im wrongbq +Resolved with +XDT items targeted at XDT not X +Doug is this acceptableI trust Alexander on this one +Merged to with JBDSVersion Build id Build date +Until than no to be done clearly +we dont need make changes to DRLVM right how especially if these have to be redone laterIs it ok +Heres the newest version of +defines appropriate methods and varialbes +Ive misspelled affect version +Perhaps it would be better to expand that sample it is up to you +Im not objecting to choosing the exclusive option either Im just saying both options have their pros and cons +Thanks for reporting +so I delete one project +I confirm that workers responsible for sending notifications stopped working from Sep AM EDT +Why not just call in stop? periodically check health state since because entering an doesnt quite parse +Applied this in Rev +Removed DSL and Field on latlong implementation does not handle meridian caseShould fix it quick +However we can always have antto for I guess +admitting the possibility that such evidence could exist +Tried to put marven dependencies before JRE System Library +As part of the work on the exception handling has been complete eliminated +This is the only time this test has been seen to fail so closing as a spurious failure +But i would definitely not be happy to upgrade and find it blows chunks because of the attribute name change +TEGRETOL LEVEL WAS AND DOSE INCREASED FROM BID TO TID +Hi Jakob should we have the avro change in this same patch or in a separate patchThanksre welcome to if youd like +I think regression test will be ready soonThank you Mikhail +I looked into the feasibility of this and while the serializer is nominally pluggable it doesnt allow the calling classes to pass in any information that could indicate a desired format such as a format query parameter or an accept heading +so not sure if we should detect texthtml to send html page or not +Must have Etag support to demonstrate this feature in +The patch isCould you review it +updated doc for +Feel free to continue using this issue to exchange patches etc. +Our options would seem to be Dump Fraction and recommend Math +So I submitted the patch with comments but without a unit test +The solution proposed here is Raise coercion errors when trying to bind an expression which needs to cast between and types +I am now trying to override the properly using my local Nexus and will report back when I got it working +Im calling it before if an exception is thrown the change wont happenI think the spec is not clear and this makes more sense +You could do that from the SP side +Since the CLI is suppose to be universal Im thinking that we should have a single guide that documents quirks when a platforms CLI does something unexpected +Should send close out to any it finds still open +Timo can you please test and report whether your problems are fixed now? Thanks +If you dont want the warnings to show which are only warnings about possible issues with AS migrated code then dont pass them to the compilerAnyone else have another view +So Ill simply add second one +It needs some further testing before I can close it +Changes of the comments according to the suggestions +we should simply change the id to be the url retrieved from its the ony way to have it be unique +It will be inAlan can you set the fix version to and? I do this when I resolve issues. +Sure I can make the change thats fine but of the populations going to be inconvenienced and if they havent been following this theyll have no idea why +Theres very little changes in classes +With the fix applied the test attached to the ticket pass for me +This feature can reduce the impact of the that issue describes and is another approach to making Continuum easier to introduce in an organization +Thanks Yonik +Sorry for the noise +Im running on Mac OS X and Java on trunk +get on isRebooted needs to be synchronizedI updated the patch with this changebq +Committed to master +Agree on all accounts Ill update the patch with your patch that incorporates Dougs feedback +for removing from the branch +Changes from previous patchResolved trunk file conflict +extends and is an example where the original constructor is being called with a instanceCould you elaborate on why you believe this doesnt work on sequence files or provide an example +Yes this is the first exception during deployment +I probably did something wrong because I this and now it seems to work fine +Is that rightPer the JIRA triage meeting on Oct Kevin C +Also removed author from the file +Copy Nings concurrency patch over here since is supposed to the changes +Seem like weve hit this problem +Thanks +fixed by other fixes +Id prefer to limit to trivial or issues at this point +The gradle build is ok but I have not managed to use Tomee Pluto and Arquilian in a gradle build +once is resolved then we can add the build file to build this contrib module. +But I noticed that it changes not only the tests but two classes in the EJB sourcessrcmainorgjbossejbmainorgjbossejb the problem is not only in the tests may I classify this problem as a regression? If so we should apply it before CR is released +Better suggestions +Please apply +has been filed to improve the documentation in this area +can de will get the same +Attached a patch that implements the requested feature +It always uses bit read write operations for bit fields +Perhaps a simple starter task for someone interested +bq +Also tried with a colormap based style changing the opacity in the dem style to here is the result +Patch for upgrade +reading Sateeshs comments looks like more of an enhancement than a bug in the software +Feel free to the issue with further instructions on how to reproduce. +Other case could be that maintain only renecnt block report with the greater genstamp +This patch against should fix the problem of unreferenced subclasses not showing up in the generated wsdl where the only reference to the parent class is via an array. +At revision o Test Enhanced EAR containing WAR EJB JAR +Closing this issue as its pretty clearly an issue with their server at least imho. +cany ou check if this is still true +Already resolved by and is a dup of +Great job Mark +closed shipped in +Added null checking to and to +A small test project would help us to save some time +Hi Petar can we have test? Writing test for this plugin is painful but the written tests have kept this plugin health so far +For some reason the batch contains data rows the first one contains a change and the other one doesnt +I am not able reproduce with the latest svn build from the branch +Ok thanks a lot for the background especially during vacation! Feel free to ask if there is something I can do to help +However these are missing so that Logging and cannot be used seriously modulesorgjbossloggingmain modulesorgjbossresteasymain modulesorgjbossresteasymain modulesorgjbossresteasymain JuergenCan you point me to some docs for how to install resteasy into as? My version of the installation does not have it +Claus and Jonathan Yes camelrun had this overriding possibility but cameldot wouldnt use it +All tests but the deletes failure in pass +Yes it works for meAm I not allowed to close the bug or am I just blind to find the button +I didnt realize that that was the slow part +The reason Im slightly wary about adding the factory is that it is really hard to design it in isolation without being completely biased about the +I think this should be moved into the website dir as a template for automation so we do not have to manage the version section by hand +The repository contains a jboss release which goes together with Tag build of in Branch has been repository version is used instead. +Should we document those too? If you run Gridmix without any parameters you would see these options +We use xstream for marshalling and unmarshalling of this configuration files so xml elements map the java object properties +Im sure this is because groovysh has some magic in it to try and allow for line by line parsing where groovyConsole parses fullcomplete source buffers +Attach a preliminary patch to see if it works +Just came across this again +Additionally JSP allows programmatic access and customization of the EL evaluator through a set of standard interfaces and abstract classes +Setting Fix Version to future for issues without a set fix version +All the code is done but well need to build new platform that pull HEAD +Ron states above the connectionWait is in ms and defaults to but the commited has value which seems to be a very low value +The schema for module name is xsstring which also can allow for characters that need to be escaped +Cool Ill apply it now +Ok I was working on a similar issue regarding the fileupload it seems that I have missed one more part where the this can occurI will investigate this early next weekThanks for reporting the bug +And all are working fine +My vote would be to do nothing on +Recently we also faced this issue +sure ill include with the cache features +At least if it is then I cannot find itSo we are both blind as bats +Yes this wont affect +Should be fixed by the JBOSSTSSPCP release +It broke in +No problem Maarten +IMHO we can close the issue nowCheersSveta +Only affects usage commentspatch p i and committed Georges patch. +This bug still exists in that should behttplocalhostsamplestest there a patch available for? Id rather not use a releaseright now +How would you define small +I dont know if it should be WARN or INFO +attaching the same as patch +David Lloyds solution is excellent +The Maven site does not even link to version anymore +Its returning +Interesting +Thanks! it looks broken to me +AmilaAre we there yet +Thanks again to Hal for reporting and identifying the source of the problem +The current mode of operation for data files is to perform a operation for each record present +Thanks +As for the querying parts Im unsure wether this is the best approach +The patch looks as expected +There is a question concerning guides styles could we use one stylesheet for both enterprise and products? if not then why? as its easier to have everything located in one place and all guide use it +This will be addressed later when we have a new implementration +Could you add a testcase that would test this please +Im submit this patchI tried to use the but it related with the job scheduler and FT mechanisms +I have similar problem with all tutorial applications and my own application as well on +One straightforward way to accomplishthis would be by adding another parameter similarto the already existing parameter that allows controlling the port numberAttached is a patch proposal which adds a hostparameter to the file and enhances to use thatparameter if present to set the host name parameter +It should at least update the stored block length and add the replica to the also if two datanodes report inconsistent block length in their blockReceived confirmation of the same block NN should be able to handle the problemAslo triggered the problem +This passes on my current snapshot of available in my maven repo +We are not using internode encryption but compression is used +some info to self for better understanding the main Karaf launcherDescriptionResourcePathLocationexecuting initialization scriptsrcmainjavaorgapachekarafshellconsolejlineline creates the Karaf Shell Console Threadsrcmainjavaorgapachekarafshellconsolejlineline +Hi againcan you show me the code that adds new contacts? Ive been unable to reproduce this +Either that or figure out a way to make protobuf RPC not mandated but rather merely the default +Ok It now works for variable in pom but it still fails to replace variable that are defined in the parent pom +Can someone review and commit if things are ok + will need to be checked in before this patch applies additional of this is necessary if is you please add a reviewboard link +long as were refactoring Id like us to look at setting up the appropriate JNDI so that we can lookup the necessary components +Ill fix later when we release +Now I tried that on Mac and Win and it is not enabled by default +The next step is to get SDO Datagraph from this DAS so that the changes can be tracked +Closing. +It is already possible to configure a database connection using the classes +QA fun on different than sun +I just also noticed that whether I do cmake i g Visual Studio Winor cmake i g Visual Studio It is always saying Building for Visual Studio CCSourcesQpidcpp +Ive added the test case into itest +Hi JimmyThe patch is for java only should commit to java branch +Oops I should have classified this as a bug not an improvement +This is one of the things I have been working on +So by the time of next snapshot this file should not be there in any case +Ive committed the junit test as wellThanks again for the contributionMaarten +Todd can you please hold off committing it +This way regardless of whichever is lower well sleep for the minimum time needed +Jake this looks good +Targeted for inclusion in JBDS Beta Designer has checked in to fix the issue +the Setups the bug +Number sounds great but wouldnt you always want the latest scoring document since that should reflect the latest updatedb and rank of the page even if its lower or higher +Thanks! Best regards verified by Andrew. +I checked the current implementation in details and saw that perform does a first REST call to retrieve the instance to operate on and a second one to perform the action + on patch +Chris Thanks for posting the proper patch +I thought the plugin was supposed to have an embedded versionI do have a proxy configured in as well as Eclipse preferences +Previous version doesnt reload values from TLS in case of possible GC +please do this + milestone added to incubator site December report has graduated from the incubator +Context menu structure was updated From Palette menu item was removed to reduce nested menu levels +A good example is +Id like to take this issue may I? you for your participation +Sorry its not working with def middleware Ive got it after private keyword and default stack with transaction was used +I checked with Owen this is an issue in hive as well +fixed in r tested in basic example and working +The line dbCreate updateshould be removed from theStill you need to change the comment on that line to include all available options including to ommit the parameter +Further changes will be considered improvements or bug fixes +So this should become a wizard then to set the decoder and parameters and just happen once like we do for the Java Mapping and Apply Template tasks? The properties view would then just show settings for the decoder type but allow editing of the is that what youre suggesting +Dear Raimund sorry Ive been kind of busy but this is definitely on my radar and I think its great work +upgrades rdpartylibprocessrdparty to make use of +Would that be good enough +Thanks for catching this! Ill update the files and commit a new patch +Thanks +Fix was merged to fuse on Jan but it did not make which was cut on Jan +Attaching patch to add a log msg when reading a struct if an unexpected field ID is encountered or a field has an unexpected type +Are you using any custom components +partkuntil added to TM +Closing. +Discussion on was negative to adding these methods +to commit +However we didnt register the event tracking number say until after the ungetServiceNow has no references so it can be deactivated +the same lines as this is further locking in a client having one and only one identityPlease see my response on it should still allow for support of multiple identitiesve often considered having subclasses of UGI that were specificGood to hear +Youll probably have to run the script manually +Your logic is correct +What you used was an hack over a bug to try to have portable I could recommend instead of trying to have an ugly fix in the plugin is that you use profiles in your settings +But I guess that those are pathological cases that we do not have to take into consideration +The attached patch gets around the issue unless someone can think of a more elegant solution +Cleanup will take a file that has N other files pending and will remove the dead records on it +Must have looked right past itv coming up +Parameters are taken in account when specified from ignored when using +I think it is a bug that could happen when by casuality two threads are building the same composite component at the same time +Please verify +This seems like a fairly risky incompatible change +to trunk and branches +Maybe put it on the onInitialize method is +Possible related to in wsrp merge to branch +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Are there moderation controls for the list? Yahoo group had something like that I think +I just committed this. +Since I cannot read russian does the DBF still contain the same data you cited aboveIf not what strings should I check +If we were to process work in the order in which it arrives you will be th in the queue +Thanks for the patch Scott +Please reschedule this accordingly if we already know we cant make that +and have made it into the release yay! has made good progress but is currently blocked on someinteractions with the new GRANTREVOKE feature +quick reminder is defining a method which was created to plug in some counters in the future +That way the issue is solved in one place and does not rely on everyone who wants to use maven having to monitor their entire dependancy tree every time a new dependancy is this involves the introduction of a redirection system then so be it but it just seems like its creating work for project maintainers just to save some time for the repository maintainer +Updated patch for cdhb. +Will upload a new patch for that +So this new fix makes on SQL Server work ok for you +ThanksIve added the user as a committer to the issue can be marked Resolved. +hmm interesting problem +Fix applied to trunk +Doesnt make any sense +No need I think weve got enough info on I see it for both ant and shell projects and will be adding some Selenium tests. +Marking as incomplete description since its not clear what if any enhancements that are not covered by other should be worked from this +Though I wouldnt argue to forcefully against preventing it if everyone is on board with that approach +Ok thanks I will look into it +There is a mixture of tabspace indentation +Logged In YES useridApparently there is a problem in the jdbc wrappers since they are intended to change the underlying driver autocommit setting only when you actually do some work requiring the new setting +It seems it did not like that I used instead of standard in one of the tests +Thats strange Ive merged the exception branch locally myself yesterday and I dont have that Problem +Attached modified version of the test one that invokes the query parser to create an MFQ +What is the version OS platform and database vendor and version +Failures on was the final cut to throw out the module completely and move the necessary code for toAll dependencies to inij have been removed there are no longer remaining dependencies on inij at all +Thanks + think the select should filter for typeSTART right +The conflict might require a user jar that is not compatible with one needed by the framework either order breaks something The user might override a system jar and alter functionality in a way that breaks the framework or subverts securityBoth the host container and the user code need to be able to be certain of what code they are executing without stepping on each others toes +rodrigoramkumar can you please post the result of ant test +task needs to wait on the outcome of do not reopen issues against an older release unless they were truly not fixed +See +Why not just use a implementation to handle the current logicColmColm O hEigeartaigh commented onI tried to minimize my changes to cxf code +Project attached zip archive with some of project sourcesThe full list of used libraries is in file. +We confirmed via email that this fixes it for him +latest version +The attached is the patch +Already fixed! As reported +at least that is how it looks like until I start clicking things and realize I can go selectunselect entries and click install again +Waiting for wstools issues to be resolved issues for already released versions +Id suggest that it be part of a separate issueIve opened to add support for disabling the cache on a basisIve also regenerated the patch for this issue +Test case that produces the discrepancy between and +Test case with fix +looks like its solved in GA +Im slightly busy atm so I will do this a bit later +Please keep up the great work on this front +The dummy jar file I have created to keep Weld happy with my EE app under Java I see the problem +I did a demo upload to the staging repo and everything worked fine +Yes this was confusing +Santhosh can you provide feedback soon or let one of the other committers know what to look for so we can move forward on this +Move javacc related warnings to. +I think this is a suplicate of which has a patch attached +if the return of null does not cause new looked at all users and they check for and handle nulls +I approve this patch +xml without being installed as the default editor. +Thanks Matthew. +The tests will be ran with the fixed patch +caching of layouy template locations was introduced by this issue +Thank you for the review +I have to admit I dont totally understand all the complication so any health debate is welcome with the modify solution inspire by Ramin and Dan +Would it be more consistent to forbid deletion of files that have active leasesIt might be but the change should apply to all files that are being written or read +We already provide public static final fields on to represent valid primitives +This is another xml and a new code +srcjavaorgapachecommonslangmath +With the exception that these are not matched against the projects dependencies +Oh yeah Im sorry I missed this. +Attaching a prototype implementation of basic machinery for generated columns +and the only way I could find to produce a null hyperlink +It passed all junit and ee test on a clusterNew code has been documented +In this case it is not a bug +Heres a small update to use retryyum +Thanks Jon +The possibilities are endless given our big configuration. +This is untested but seems straightforward +Rob can you please have a look at this at your earliest convenienceBut just wondering since the issue is easy to reproduce shouldnt there have been a unit test to verify that is was really fixed in? was but due to an error on my part it wasnt comprehensive enough +A release version of staxnav artifact should be used instead of SNAPSHOT +Max please can you decide or assign this issue to somebody who can decide if we have to change the year in copyright text of would prefer to not search for changed plugins latter than in and change the copyright text only for those plugins +This patch exposed some problems for stylesheets because errors were no longer thrown for unknown xsl elements +deactivated which is not the case if the default level is INFO instead of WARNING +passed tests committing to trunk +Committed by Sateesh BandaramSending javaengineorgapachederbyimplsqlexecuteSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +If these changes look good please close + +And lastly this requires that we be very careful not to accidentally refer to classes from the wrong compiler level in a projectThis is all possible but a big deal of work and something that is unrelated to this bug report +Can you please try again +Substitute the more general Mutation type for Put and rename the preXXX and postXXX functions here from +Please provide more details as well particularly how you configured in your cache config file +Ive added the audio alias in r +with +One last question is it really necessary for the shell to be instantiated under the root account? While reviewing the patch I havent found anything that would require privileges +Alternative approach is based upon support in the initial commit +Still to resolve we are not pushing the event callbacks through in the same order we used to +Also all the integration tests pass +Points can be clustered into one when they are too close I guess wheter this is done or not depends on the generalization technique used +The parsing of value is a bit loose +I am not going to move source directories to srcmainjava etc as that would break point in your list +Marcuss patch was pushed to with latest RPMs +In fact since weve taken over the Maven repo duties for Spring we dont have any files at all +Let me try to find a good example +Hi Julien could you attach your little Maven project to this Jira ticket? It could highly help us to reproduce this FreddyHere is the pom that was Julien this issue is fixed in Sonar and occurs when the project is fully empty +It picks up Python though so Python is not on the default path +on patch for and trunk +Version Final is released +Suggested patch file also includes an additional method to facilitate subclassing the for other nefarious purposes +So reduced to minor since there is a workaround +Could you please exactly tell us which hibernate GA release got this fix? +The latter Data classes can be shared with the producer side as much of it is common +This is more correct than simply updating RELEASE when there is no release +bug brought forward to for the version. +Fix the problem in by using a regex to filter out the problematic that there is a potential conflict with since they both alter +GCC bug was opened for this issue +Confirmed stack gets blown with k sstables still +accept null +Resolving this issue. +The is from the src repo and the is from the examples repo +This has the advantage of working in singleton beans and serializable beans as well and to always obtain the current session handle even in case of invalidation between requests or between multiple calls within the same Juergen this is fantastic. +I didnt mean to set the priority to Major +I meant getter +A potential issue here is that the deprecation is done in so if somone is using just using the Configuration object the deprecated config wont work +But on subsequent boots we check the stored digest against a digest computed on the padded value +Heath is this considered to be fixed? Your commit made it into the branch +Huh something went seriously wrong when sending this +Ok I thought it was meant to be included in a future release and with the binary encoding +This issue is resolved now. +gert please define a user on with ths criterias and then assign task to me not to be a MSSQL server issue +Let me try that againIn JBASSOURCETREEtestsuiteimports there are two equals sign errors +Request Headers Request Parameters and Request to in trunk +The issues is about removing features from Categorized View now all features that has no category appear in Uncategorized category +Return the default behavior to blocking +It will not be used for heartbeats block reports etc +Here is a fix for the failed parse of Reference files +Corrected fix revision . +Thanks Anupam! That was a long drawn affair +Would you be willing to submit a patch with a fix for the problem and possibly a test case for itll work on that the patch is easy the test case may take a few days to get to given my current schedule +No there is no public page to view the received SGA +Due to a rework needed on I was not able to get to this task prior to the end of Dev phase +Any progress +The Glassfish and other proxy repositories have been added to the public nexus you Paul. + +It may actually be sufficient to just store the max timestamp and not the min +If we build in this feature we must be able to switch it offthrough a parameter +I also set the to false +Closed after release. +Actually I dont think this is a messages holds a buffer +you might think that you want every hit back no matter what and ask for like the top million hits and require a huge pq and blow our your ram and oh how confusing +Im not sure what output I could get for you +Any takers +Good job very happy to hear that +This might put you back in a good state and allow you install the latest version +OK I have just merged the branch with the trunkPlease let me know +See if you register in a child pico a component with the same key as in its parent you will never get the parents component although it might have been instantiated and is of a matching type +Was able to reproduce if I start my server from inside JBDSLooking at the console output on the Teiid Admin URL is being created as mmlocalhostIf I start my server from my file system terminal it creates the Teiid Admin URL as mmslocalhost which is the intended default secure URLTeiid Designer was changed to use the mmlocalhost as default URL +passed all hadoop tests +As are deferring most of the functionality for the to the we actually have little control over what it does in relation to SSL we would likely have to largely it in order to do this +Since this is restricted to some IBM JDK releases +Thanks for your interest +We can add your link as soon as your homepage works +I am inspecting and applying we resolve this issue +It seemed to make more sense because thats technically the last version it supports +was just reading the docs and noticed the redudancy +updated the long stacktrace message +Thanks Espen! +Those services may be unused byDoes that make senseIf there is any interest in our implementation I would be happy to discuss it furtherIll try to answer your question with our implementationAn user has a role of OFBIZUSER +thats intended behavior +doenst seem to be a problem anymore +Why no link for Amazon? What is the difference? Both are S object stores added using the same build using the same steps +A trigger is present which should return generated key with insert this mode is only supported for Oracle but it is not covered by a cpactf test yet +Does operate withmultiple NTLM proxies and if so with SSL as well? +Updates to complete the new look and feel +edbbfe on master branch thanks! +Seems that this is not the first time this issue has been encountered +Can you try it as a Spec within a Rails app +Glad to hear that +Committed to trunk +Ok I did another set of testruns the error is not reproducable anymore with the default crypto settings +To be honest I dont like creating a load directory in some kind of temporary folder in that case +Thanks to Myrna for pointing it out. +Looking at more it is probably better to require that the collator be of that type it seems to have a more flexible api than the pure CollatorAnd for the performance stuff I agree if you can get the functionality working then the performance improvements should be somewhat easy since the code will be fairly in these classes +So it turns out I was +The implementation also needs some work +Not sure I understand your problem though +Excellent +The difference between and is that adds a line in to disable additivity on a logj Logger instance +In fetchResource you check if resourceStream is not null if it is null the whole function will return null +Applied changes as suggested by RolandOleg +Grails is not currently Nexus enabled would you like us to perform the migration to Nexus +That was the reason for this Jira issue I think +Thanks for the comments AlejandroReplies inline +If we want to run heshe has to run ant again to get containing all libs +I think your categorization and my comments above are telling the same thing no confusion there +However Im not sure its right to ignore in this situation so wish for your opinion on these two patches +I dont recognize this structure +SDK compiler and a representative Flex application passed all unit tests and was able to run +Test casesThe way we have done this might need more improvements. +Sure I can easily create a branch from the tagbeginldapwork CVS tag +Also this particular messages implementation lies with in error dialog implementations +the require also applies to the runtime but is just required for the testing during the build +Thanks for that report +libUsersjoaDocumentsworkspacescompanyworkspacebin +Sorry Ignat +Thanks for taking this on with this final bit done we can start using mvn for the Azureus bittorrent client +am in the process of reworking the patch for Credential Management Framework in order to support accessing keying material for this issue +Youre right sometimes it does just say Fault and we need to fix that +This is not related to the patch +it is included in this patchI would be really helpful if somebody can commit this in to the trunk after receiving so i can move ahead +Committed in revision . +perfect +Will update notes on to indicate that a patch is available on this issueThis is part of an effort to close old issues that I have reported +The patch is small but the consequences arent it imports common and at the top level so I cant predict all the consequences really +The RAM buffer skip list writer probably requires two additional parallel arrays +Here is what I committed +Moved from to just after merge of old branch into TRUNK +Fix attached +in get it should be pass false for isReading rightYoure right that the test isnt reading from the returned streams but changing this will result in the test failing +Its worth keeping an open eye on the EnumerationIterator transitions in future patches on this issue though as they may indeed cause CME if were not careful +Im open to patches +All I can recommend at this point is make sure no deprecated methods are used anywhere in your application code and to collect more details about the issue I could act uponOleg +turned into that the main logical piece of work from this JIRA is being reviewed in if you want to do additional review please feel free to +does the above Owen patch and addresses the first Doug comment +Through Junit test I have found that is reopening on the clustered files +Isnt this behavior a bug? +Its not really a big deal to make it work with the sequential classifier but its much more complicated when it comes to the distributed oneIts strange that throws an exception I tried that same example a week ago and it run well +Have you used the user forums first +It is not neccessary to user have ability to resize images on search screen +Even though that would simpler in the number of line modified I dont think an existing stage would fit the bill and creating a new one for that doesnt feel right +Oops I forgot to clarify in my above comment that the error proneness would still exist for the user configuring the manually not the user using the tooling +Could you please elaborate on what kind of files are you expecting to be dropped into etchbaseBasically if you could walk us through a scenario of using this proposed feature that would be greatly appreciated +No need for QA as this is internal build and infra Do you have ETA For RHEL instructions to be addedthanks there latest update for this task +Currently my research indicates the problem is that we should use the terminatesm flag but it doesnt work +Moving out of +Changed singleton documentation. +Committed Also accidentally committed a few additional items that I had been holding back on +If we push all these down into the context partition then it will be specific to one implementation +Thanks Bikas! +Im not saying there shouldnt be a download button in central but the core extension point feels more sane to have in runtime detectionThe Download button in the central plugin enables downloading only required runtimesplugins +Attaching +The old can no longer be reached from the console Message alerts Text contributed by Tom Cunningham in The JON integration for message alerts collects the alerts and displays them as JON events or if the administrator wishes to do so as JON alerts that are fired triggered by the message alert JON event +Victor I put this as blocker since it sounds like a bad regression to me +There the problem was with the DRDAID column +Any ideas on this yet +Patch for submittedCommitted by Ryan +Im attaching updated version of the patch that is removing the long line violation +Can you please try this one +Tnx +After further internal discussion weve determined that this is a critical issue for us +Let me know if you need any further clarification +tasks were not related to this change +HiCould you extend the documentation with some info about how to setup the database connection if the application is not standaloneThank you +Philip TomIve updated the patch to fail of available memory and added the Assert to the unit test +So this must be configured only possible solution is to use more unique paths for the Servlets workdir but this has significant side effects. +Im working on this on my github branch +Went ahead and committed to trunk and +committedThanks for the review Mike +to what Stack saidId be happy with mins +Should be backported to issue fixed in released version. +Thanks Jon +You wouldnt need to register the with thatll happen for more difficult part would be where the code that generates the agents certificate uses a bunch of classes that are only in the Sun JSSE and have not javax +committed to trunk Thanks Andrew +Ill look more into this and see if there are any differences between the two approaches +I have tried following your steps but cannot reproduce everything compiles fine for me +I did commit this to trunk +and hook them in to common build targetsMy lib directory maps onto the existing one as does my srcmain and srctest dir +Joe thanks for this +However now we see duplicate message consumption by the consumer within the same group as side effect +I suppose its a difference between having one step with multiple Tasklet implementations or multiple Step implementations +code has been broken out into the plugin +Can I make the changes you suggest in another JIRA so that branches and are kept the same +Note you can add only inVM listeners +either you create a new node and then you know all the mixins you want to setor you alter en existing page and then you read the mixins beforehand +What is the Oozie jira that fixes this fundamental problem with OozieWith the correct settings this may be a problem only on rare occasions for an Oozie workflow when an node crashes +I shall verified this component fully after CR release and make a new issue for full update of this component +the connection is almost certainly no longer in a useable state +Ported the solution of to the Struts branch. +Marking critical +In other words use IDM as the jBPMs identity think I will close it as invalid request +Initial patch attached +Thanks Tom and Daryn +It is true that theoretically state machine logic doesnt change much over time but in reality state machine changes would need upgrades due to improvements in the logic or fixing code bugs +This should be deployable in most AS releases +Is that what youre arguing? Nor is that URL very friendlyI agree that is a problem with this proposal +It does work as intendeddocumented though +I have done the refactoring in Netbeans and created a patch +What version of are you testing with? When you set scope to provided these dependencies will not be copied to +push out to +Is that feasible +Where is the info for the page coming from? From +Committed as is. +Jeff can you look into this and fix it if you think its appropriateSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulessrcjavaorgapachegeronimojettydeploymentSending trunkmodulessrctestorgapachegeronimojettydeploymentSending trunkmodulessrctestorgapachegeronimojettydeploymentSending trunkmodulessrcjavaorgapachegeronimotomcatdeploymentSending trunkmodulessrctestorgapachegeronimotomcatdeploymentSending trunkmodulessrctestorgapachegeronimotomcatdeploymentSending trunkmodulessrcschemaTransmitting file dataCommitted revision +Here is the fix +We should use JAAS and JACC if we need to cross utilize secuity between components. +Can you provide readable pseudocode + +Im a PPMC member of the Wink incubator details mailed to nfischer +Changes for were reverted back in svn nothing to add to RN. +While were at it we should also consider renaming the script +bulk assign all unscheduled issues to next release +Thank you for your feedback +Why dont the RS act based on the znode +Seems like just the implementation needs to be tested out +see last comment inP +I update the title to reflect more the intent than the target version which could change if we maintain backward compatibility +I have commited a fix to also take environment variables into accountJose I dont understand your point because Skip if environment variable is defined is only available with post build action and not with Invoke Standalone Sonar Analysis as build step +Araon as usual this is going to be committed only to the branch +This patch contains implementation of different ack modes plus completion of transacted jms sessions +conf is not used by default if you want to use it you need explicitly say conf +Good catch Bikas +push to +I vote that we close this as. +Fixes the issues I found +waiting to apply until current junit test failures are provided backported patch to trunk + +Thanks for explaining that +Accordingly I created branch remotingGACP for future CP releases for EAP CP releases +svn ci m Adding a state check as per +Seems the fix is not sufficient +Attachment has been added with description used to produce the issue has been added with description original has been added with description resulting pdf after processing with FOP let me know if theres anything else you need +fixed in r +Makes sense +new version of the patch archive containing the binary docs used for version of the plugin uses Tika +OK I think this fixes the problem with using constants as default argument values +I dont have them to will update my home Maven from and leave it running to see what shows up +So if youre going to add hibernate integration then definitely create a archetype and keep the current archetype in place +Each has their own copy of the logj jars and each has their own configuration file +Also if it is supported in Freemarker I dont think an example of what it would look like would be such a bad idea +Closing this issue since I have not heard back from the reopen or create a new issue if this continues to be a problem. +Many thanks. +Thanks for the hintAnyway I will see how useful it is and get back +the patch looks good to me +I fixed this in revision Thanks for the report +I just come to home +Ok +there was a problem with the library not a phonegap I dont have problems with new projects created with the same +BTW the issue is reproducible on Debian Linux as well +Volume snapshots work as expected. +Thanks for the patch. +Patch was applied and German comments removed +I just modified so all the product plugins get our specific plugin classes as appropriate +related the merge policy stuff off into adding randomized CFS usage to every test that can use it and laying the groundwork for the other issues under the parent taskstill running some iterations but plan to commit soon +Fixed with resolved issues. +It was failing before +Fixed in rev +I added them back +Failed test seems related +added Connection member to the Handle to be able to record attempting user +AIUI we do not have an Apple account and thus will need one +Should bitIt works +This fix had to be reverted reopened since the additional runner indirection was causing severe KristianI had a look at the fix and its revert I saw that you reverted only a very small part of it +Hi BryanTherere nodes and all boxes are displaying only thing is its too long so in the screen shot latter part is not visible +Bump this feature would be immensely useful +The popup should focus on the tab violations but not sources. +I just committed this +Eclipse project prototyping the new feature +Instead of Wont Fix havent we already fixed it? I know I caught a bunch of this right after I built +OK I will check if the app can work if adding I add the file and find the test result is successThank you Ivan +The patch is called The change changes to always look at the data scale and precision when sending decimal values to the server rather than using the parameter values which are never set +Its nsync Stack! I guess youre too old Sorry I confused sync and flush i guess that variable isnt volatile in my mind and I still see an old value +Too big a change after all is said and done to put against a patch release +Patch committed to trunk and branch +the implementation and added documentation for the API and the command line tooling. +Hi have not tried this +This workarounds the problem on windows that recognizes stdout in different differentlyPatch applied at +One day is NOT equal to milliseconds as you can see from the result +Here are possible ways to ensure that the autoincrementValue is preserved during the drop column Add autoincrementValue to +This feature is complete. +ive committed nowcould you provide a test case now +This is a FAQ about and it should behandled by answering the FAQ not by failing to conform to the specification +Yes I did +I would need to to get those logsCurrently the set up is running performance runs +Verified on revision . +Ive applied the fix to the remotingSPCP branch where it is included in the version tagged original issue is now closed and theres nothing left to do except include the newer Remoting version +Will look into this issue when I get a chance +Closing tickets +bq +PierreAs explained above I am reluctant to applying the patch to +Parentheses characters can now be used in substrings which should solve this issue since it will no longer throw an which means the dynamic import will fail due to finding no providers and a CNFE will be thrownPlease close this issue if you are satisfied +Thanks for your contribution xieguiming! Hoping to see more +I think this is a good idea +I thought I had caught them all but under unix in the trunk dirgrep r i author src grep v +Gokhan please feel free to assign this to yourself and bring it to Release queue if u think its feasible +Patch including a test +Please use the forum +These classes are anyway aware about the soft upgrade as they allow or disallow creation of unique constraint based on it +for private is a targetted fix and test +Closing as fixed for now +exclusion pattern is a patternexclusionPatternSpecifies a regular expression that is used to determine which files and folders in the underlying file system should be exposed through this connector +Itd be great to avoid a Tomcat compilation dependency so Im inclined to hang on to the reflective I was in the process of updating our for anyway I went ahead with a at the same time Its very similar to the current now capable of handling both Tomcats new own and Springs old +It seems this was just an issue of a missing dependency +At the moment the cache name is embedded far enough in the marshalling chain that is not easy to carry on reading the stream with a different marshallerIf I resolve this JIRA and this is not in Ill create a JIRA for future improvements in this area +how to modify exist comment? Find no way to do it while it seems some one could modify their commentYou need admin access for that +I guess you will have more context about those changes +I do have an alternate solution to that in my back pocket +Curious that the test scope is apparently trumping runtime and excluding the logj dependencies but I guess a bug in assembly that is ignored because best practice is just to include one module anyway +Should we wake this issue up again? I think the general debate was on whether our string types should be explicitly or not +Only issues that are actually fixed should have a fix release. +Rockin! +version of the patch applied. +find a bug open a filesystem when create userbut this filesystem instance will not be closeand it still cache in i comminited this svn patch i hope it can help +For unsupported should be treated same as private +The illegalargumenterror is thrown when the formatter is unable to format the particular object so it should be in error in my opinion +Allowing an old task attempt to authenticate with a new app attempt seemed like it would be a problem waiting to happen +Sorry for answering so late +Ive been running without the workaround for a few weeks now. +btw given that there is probably no advantage in terms on contiguity when we are reading across a block boundary in the same file compared to reading separate files we can prototype by having some sort of that contain multiple files +Today we try to build an implementation of the just by a stupid refactoring of one of the other JNDI DB adapter classes +Created and tested the fix on the branch and pushed the changes upstream to. +Moving to or RobThanks for the suggested fix +Included in. +is dynamically generated when server is built no source change for +but are you still seeing this? Id like to get stack traces closer to correct though largely theyre always going to differ somewhat because of optimizations I dont think we want to give up +Is it actually used by the App server when it deploys the EAR +WIP patch +I agree that the dialog of saving is not required if user edits only one file +New patch soon +This is not a bug the decision table is formed incorrectly +Great stuff! I wish I wasnt so backlogged and this out of my area of expertise Id love to help get it in +Jieshan This patch is really old +I just committed thisthanks andrei +This is one of the areas where HA gets tricky and wed need to make sure its if we intend to store edit logs there +It is ufs unix file systemThanks +do you expect null to be inserted into the template +This has cleared up the problem +Apologies for my error. +Patch checked inOleg +Heres a very basic sample project using maven to demonstrate the problem +Ill take a look later this week +Possibly related to view sharing try disabling view sharing in the configuration +I dont see any crash but incorrect exit code only +Assign back to me if you need +The checksum patch did not overwrite the method get in +The fix looks great. + Added swap usage collection Added UI descriptor file for swap Added cluster swap usage for file +The following tests should pass Reason The handshakecompleted listener had no time to populate the cipherSuites because the test would proceed to check for the cipherSuite +Updated patch addressing Matts comments AND fixing up the testP +Added +Assigned back to Emmanuel as it is not a problem in the test +Brian Denny said he just commit the codes to trunk not branch +Inspected JAR and closed. +Document was completed and committed to the documents codeline on May +Thanks Jarcec! +if the information is on dev list seemed to think this was a good idea so we can apply this for the coming release +We want to move such specialized capabilities away from into their own interfaces +This is a clone of +Thanks for the reminder Sascha done at r +even jboss is not even giving any exception +QE please verify this issue on an iOS mobile device andor a OSX desktop device. +Think about when you have two projects one can migrate to JDK and another has to stay with lets say JDK +Committed to branch at revision +This is currently working in both the compiler and the interpreter probably as a result of my complete redesign of zsuper logic +Move to startupshutdown component +There is an outstanding PR against will update the commits against later today or tomorrow +Fixed in rev need to revert changes made to tried this in the branch and using a single checkout directory for a project fixes this problem +Withdrawing patch +That means it is fixed already in. +I added a new method on instead of on to delete the destination +This first draft contains the skeleton and the interface part +closing issues for released version +Sorry about this a cut and paste regression bug as a result of refactoring +Is it necessary +as to not make it differ too much +The JAWS license was added in NOTICE and LICENSE +Please apply the patch +This doesnt seem to be fixed in Hibernate either +Meanwhile you can see if it is fixed if you build from branch +A proposal for changes to Iterator Lock and Transaction implementations for Jena +This is an important test and we dont want to lose anything in the translationOne feature the old test had was that if testers became hung it would wait for some time for them to complete dump the thread stack traces and interrupt the threads +It seems like we currently create way too many proxy objects +of course its a high priority +you are using firefox do you get the same issueThanks +Maybe use a chroot fs over local fs to lock it down to +apparently this patch from last year has been applied so Ive closed the issue +Does something here need fixing +Hopefully the next Seam release will include the tests pass for me locally so Im not too concerned about this for M +Any chance you could attach a simple sample application showing this issue +It does not invalidate anything in the book it only makes it less complete +Ping +Committed revision in Nutchgora branch +I have been swamped at my day job + +Fix the and the to use the junit temp directories +ChinnaI guess you need to mark it as Patch Available so that committers know that its ready for review +Edison do you mind checking this one out too +extra list entry in the xml code block for the in the peroidic recovery section +I just like to know the status about BSH in widgets and minilangFor UEL I must admit it can be very handful in some cases less verbose than even Groovy +What you are asking is way to create Base Layer based on source model where you can select choice of foreign tables +Is there another solution for the class cache? ms is much also with profiling overhead +I would strongly recommend you dont use it at all +I did look for docs on this in case default settings changed +Reopen issue due to test regression +The only thing not so critical are what kind of line delimiters because of CRLF canonicalization. +bq +Is this Issue a Blocker or Critical priority for EAP GA? If yes +am attaching patch which eliminates magic strings in the JDBC factories for both the Network and the embedded clientsM javaengineorgapachederbyimpljdbcM javasharedorgapachederbysharedcommonreferenceM javaclientorgapachederbyclientamThe jdbc tests run cleanly me for except for the known regression in +This information would help that effort +same issue with after disabling JMS inside and changing classes to use javajbossdatasources as jta data source and removed RESOURCELOCAL +Yep if the test passes no output is printed unless verbose is specified +Now that work is well underway with AS all previous community releases are +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +Id been getting around this by fiddling with the parameter +The workaround I have now is to use a ton of include tags instead of inheritance and then to maintain simple copies of the main templates that are primarily filled with include tags +etc mightve contributed to it +I thought it was the OJB JIRA +Merged to the branch in revision . +Hmm +The problem with multiple files on the endpoint uri is that it makes the xslinclude much more problematic as now it can be relative to the location defined in the endpointStan there is some recent code which uses the single location from the endpoint as the base path for xsl include resolution +Are the show invisible tags option available in the editors drop down menuI would assume this is something users would like to toggle while editing without having to go and change a global preference +It sets the context attribute at the end but that shouldnt cause a +See attached test you can toggle the mkdir flag to make it pass or fail +Struts has nothing to do with jdbc so this is an environment issueSuggestions Try creating a statement without the type and result set parameters +Hello this JIRA project MUST not be used to get support but only to log bugs or features already discussed on the User Sonar mailing list +What and have in common is a bad CD section end delimiter thatis not matched by a CDATA section start delimiter +Upgraded JSF based on Mojarra +Look for a forward in an location +Mixing old and is ok and was obviously usedI will commit shortlyJust one question Does anybody know why there is this extra payload cloning +This is reason why its available at websiteOk this text is a comment inside XML dont cause any deployment issues but if a customer or community member read it they will notice this text error +All uncompleted issues are now moved to the new To be determined for future releases category +Which distribution did you find this reference in? Im guessing in one of the old packages since I had to produce those locallyand as is now apparent I didnt use a fresh checkout +Either the patch of replace with or the patch of replace with the performance slightly degraded +Im attaching a revised patch along with a new that mentions both the error and the exception +Thanks +Closed +the find button is not working in the googlebase application +only Ajax fallback expected in Final and then we could work on Client part +Missing this cleanup to the next release to coordinate with the jQuery UI part of duplicate code has now been cleaned up and collected in the class. +Dominik please provide steps to reproduce the problem if they are deferent from description above +Agree with you on this one been meaning to do this for a ve improved the performance of the script well implement some server side code to aid this process once we have the infrastructure in is fixed in the branch +Nicholas can you try to rerun your example with latest trunk once and see if the problem still exists because moved pidFile to task directory +I committed a fix that seems to catch the current problem +Programmatically run wst project can this be added after after every export to db? Or some toolbar item synchronizing this? Current steps are not very comfortable +Thanks BrajeshYour patch is in rev +We now also have an integration test that runs a similar scenario at build time for verificationDushan feel free to reopen with more details if the issue still persists +Somewhat curiously Im having trouble reproducing this locally +Hadrian can you speed up this one or we are going to defer this to we wont upgrade on planned to work on it this weekend +But I recall you fixed this before that right +The documentation also says Each thread holds an implicit reference to its copy of a variable as long as the thread is alive and the instance is accessible +Im in no way married to this so I am game to change it to a null check if that is what folks want +Ok all events should be working now +Ive committed this to trunk and +I just committed it to under. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Makes it hard to see what changes are related to this issue +If Aperture is generating that nobody can use then we have no reason to duplicate that approach +Your change above makes sense +Also a little bit of cleanup is required especially to naming variablesfields in the classes +The problem is not within xplanner +Then start geoserver and see if it fixes the problem for you +Closing as this needs more info. +In the select list and select clause can factor out KWSELECT as that is the common part in the two options in the selectClause rule +I attached the file with everything in alphabetical order +With this patch those jobs recovered successfully +Place it under under deploy dir + of a bulk update all resolutions changed to done please review history to original resolution type +Was fixed for M. +Would be cool +yes that is all fine but apparently multiple foreign keys are in there which i noticed mysql can do since it keeps the foreign keys around even after deleting the tables and when you create them again voila the foreign keys are defined twice +Maybe for in Eclipse this problem doesnt exist +Verify the following content in this chapter +so you have to run the bootstrap commands again when building from trunk +Agree +will attach a new patch +Many thanks for your effortsIf you can help out with testing of earlier OS X versions please see the thread on the dev list +Hmmm +The bug is in plan generation when is followed and is followed by +Thanks AlexeiPatches applied to BEANS modle at repo revision rPlease check that they were applied as you expected +If a bookie starts replicating a ledger fragment because it has been added to the ensemble of ledger but it doesnt complete the replication process because +Passing to hudson +I think the Lucene summarizer requires more CPU than this one +I really appreciate that youre taking your time to revisit this part of the code +And the scrubtest data is there in the patch +I worked with this on Danushka and the following fix worked for us +The issue is more pronounced when trying to use Hive with Hadoop +I confirm that the problem exists and is serious +nulls aremore universal then empty strings +Added patch that fixes this issue +Ive made deprecated and added now both are supported +Attaching patch again +resolves this issue +Problem is that we cant expect users to increase their open file limit can weI am not sure if there is any other solution +set to LATER since there are very little requirements that would justify that we go for this big change +Working on another approach that should let us throw away repair handlers in the expected case that everyone responds quickly +Ill try to take a look at your patch early this week +I was able to insert hundreds of documents per add using the REST API and a single collectionWhen I my collection as as a cloud instance with It fails whenever I try to insert more than one document per add request +Please open an issue in the JBAS project +To replicate the issue edit and replace the import with import +This patch should fix the Lee Thx for digging into thisYou mean that only the first split works well and that all subsequent splits fail with the current code +Whenever we discard compatibility with older Java versions and get rid of the threads there will be noIts probably more appropriate to throw an than an +Failing tests seem unrelated +Hi Aidan can you take a look at this please +Damn +Additional tests committed. +I also think it make sense to have the contain the overwriting interface +updating title to properly reflect this as a potential attack build for looked good and web related TCK buckets passed on +Add a new patch to keep up with trunk right now the autoconfig is not supported that is if you want to update the health check config file you must trafficserver restart the trafficline x cant do it +Apply that and fix the IT manually please +testing to see if assignment works +More folks will see your question +thanks for nice fat patch kay kay +To me it does not matter where the repo sits I just cannot have it growing reopen this issue and listen to the community +See the first subversion commit for the implemented workaround and follow Davids instructions to reproduce the error +Example Regex credential validationMust be at least charactersMust contain at least one one lower case letter one upper case letter one digit and one special characterValid special characters are turned off by default +thanks amila for the contributionLahiru +environment issue +is changed to take File rather than and +chunhuiRoot is fine +It is referenced from man +It seems they are the same problem +The patch introduces an issue in the +ok the solution is to rename jdbc to messageSource +Err +What do you think +Verified fixed in Version Build id Build date +Could we add this to the new etc +We may need to add a similar workaround to older branches too +If you managed to provide me with samples that do not include the issue discussed here even better +Hello Laura The behaviour no longer occurs clicking on the PDF link now takes me to the PDF as it should +Uploading new patch per comments above +But shall we make this post +Could we select size either along abscissa and along ordinateI think it would be a nice improvement +The patch will also disable spawning if an output is configured and log a warning informing the user +I think we can close it I cant reproduce it with dont have tested the issue described by question is addressed in other issues +Ill close this issue as it is moot +Where I can download this adaptive scheduler +yes you are right +See the error message +Thanks DagCommitted revision +Same patch but granting for ASF practice is enhancements dont get a until changes go inIf this is inteded for the branch then will most likely be +I am all open for a suggestions however i read your thread and cannot discover where you like me to put his information +For RC +Will report back the resultsThanks again +no History and cant edit commands on Windows progress here? Losing history is a real pain +Tim if you have modified benchmark to work with various formats of older TREC collections that would be really nice +Im open to patches but as it stands I cant reproduce the problem +I have added a patch with a unittest and a possible fix for this problem +The debugger should be able to find the source code for debugging wherever you installed it on your computerThe you got appears to be from a version mismatch +Is there a chance you could isolate it to just XML marshallingunmarshalling testcase? I would be much easier and will show whether its a problem with XSDXML or really jbossxb + +revision +Committed! Thanks Namit! +thanks Ivan as well as Jitendra for reviewing! +This is not a cxf issue it also occurs with the not related to the is property which initially seemed to fix the problem but seems like it is a timing issue as adding slightly more debug statements caused the problem to to investigate what causes the ODE engine to initiate the process cleanup before it has finished sending the final response +Thanks Chris! +Harsh how does it sound to you +I have been struggling with how to implement multiple inputoutputs as I dont see a clear way to implement then in today +For the new log line it can happen today so it seems better to put it as a warning? The application will not stop from working at least not immediately +It no longer occurs in specific manager code +Putting my application in right now +I am this issue because I think we there are similar problems in the other datetime xsd types +Christian wanna look at this +Andrea I dont see a way to set a timeout +Yes I just found this too Knut! Seems we were both chasing it down lol +We rolled back my solution was wrong +One other thing in addition to Marks note the step where you link the config with zkcli isnt necessary and at that point the collection doesnt exist so it cant be linkedThe scenario is that the linkconfig step isnt doing anything at all +Fixed on revision . +Attached design doc and any comments are so welcome +Obtain guide from s not possible to include this change in the GA moving to +Fixed for EAP verify in EAP documentation on stage + that need to capture restore state should also be more efficient since they only need to capture attrs live at their point in the pipeline +What is the failure when uinsg the patch +Apologies in that case I just saw that I had exactly the same symptoms assumed there was a common Youre right my case is against Ill try upgrading to or and raise if need be thnx +binsh is always better then plain sh for security is a beast by itselfPLeaving it as sh lets users setup their installations as long as required stuff is in the PATH +I think it has something to do with the Resource Name +If a user has CREATETABLE they are granted to the tables they create +I changed the version from to CR and the exception disappeared but still have other problem Thank you Vineet Reynolds +An updated patch is available as part of the latest patch attached to the parent in parent JIRA patch. +I dont have the time or knowledge to really incorporate this +I modified the karaf script to start geronimo in rev +Should I create a new issue and link it to this one since this one is Resolved? +I only want to take messages off the Inboxunread queue that meet my criteria +number is to branch in SVN rev +Obviously there is a lot of undocumented parameters like you mentioned we could add these into the same section and create a similar section for export parameters and common ones +The host attribute can now be set +Please verify +BSP uses the communication phase to communicate between the nodes and not the nodes to a master node +Works for all properties whose key is starting with sonar +Should we keep checking whether kfsImpl null +Strange + Exclusive Gate Way does not store Default Flow in Designer would appreciate to have this solved as soon as possible just as muchThis is a critical bug and deletes important information from existing process definitions every time you edit them +No I havent reported it +Looks like the findstr to detect if a server is down is matching any part of the string in +I also ran ant that is mapped to ant dist from what I can seeI guess the problem is that when using cygwins CVS upon initial download it wont convert the linefeeds since it isSo I guess the should try to detect whether it is cygwin or not and if so it should tell checkstyle to expect unix linefeeds even though it is on windows +And I dont want to use additional jQuery plugin for that so Im working on reusing our previous code for that +Closing +That would certainly be possible but I think an annotation would be a better option since you get validation and it doesnt conflict with any text you might want in the javadoc +Thanks job RickThank you for your timeI think its good manner to provide a test case +The refactoring work turned out to be much bigger than thought but its almost done and Im running tests against it +What makes it tricky is that the ability to search the Lucene index is critical in that system and recreating it takes rather long +This bug tracker for tracking bugsenhancement requests for the core Roller software +It is based off of language agnostic bits extracted from and it is my understanding that it has been TAG approved +I do not believe hte archives nature is automatically added to all projects anymore +Hi martin everything else in this cluster of bugs is assigned to you +You basically have to write a new kernel to do this +Instead these attributes are taken from the directory that contains the linkhmmm +this was unintentioanl i didnt think s yui was upgraded and didnt bother to check +Sure +After the default is changed all existing tests except run with DFSNAMENODEEDITSTOLERATIONLENGTHKEY +Building from source is the recommended way to get the latest snapshot +A test case as a Maven project is preferred +Then the async trhead will start prcessing items from the queue and delete the block file +A quality team effort +Yes good catch +I am reviewing this as part of review Patch Available backlogMona any update on this will be greatly appreciated +We will disuss this at our meeting this week in and should be binary compatible with +Thanks +I can open the cert with notepad and I see the begin and end so I assume its asciiIll rebuild as soon as possible todayThanks +Please see that for examples and Ill update the wiki spec appropriately +I have the identical problem but this to add to it in the web server directory it works fineI can either sayjava cp +quick update after this fix also the other mentioned problem disappeared +If you cant go to the frameworks that would be the version to use +HelloThe screenshot was taken from a locally running version seen on the screen the selected row does not fit to the highlighted rowin the +This has been done in trunk +Ive committed this to +Ive just committed this to trunk and +different indexes will be useful in different scenarios +Oh its very annoying +Bulk close after release of +I didnt have updated +In that case there is no more line number neither source file path in the generated bytecode +I have copied index page only +Is that what you were asking about +Thanks for extracting createGenerator +s tricky to simulate the commit case +Ooops +Returning String makes sense +Could you please run tests or post results if you did +Jing please keep your DEVENV running so that I can try to connect to it +Also addressed comments from Santhosh +Updated the headers in the statically generated files +Working on this patch Ive unified how retrieves the Long values +Thank you Brandon +I agree wholeheartedly with the earlier comment about needing to document this in the Ref +Committed revision . +Thanks Arun! +The concept of error listeners was implemented +This could use a before committing +Assigning to raising severity one for speed +There is also a bit of overlap between what can be accomplished by a within the same context and these that would be applied only to the bean definitions created by a particular you provide an example use case? That would be very helpful. +Note that the file is correctly populated +Layout was updated +Sorry for the delay had to send in my laptop and therefore couldnt be onlineAnyway it seems that the problem is fixed I had another older version of struts lying in my path and that caused the conflict +deprecated comment should say what method should be used instead of that one or why it was deprecated if nothingThe rest looks good +What do you thinkYes it is a workspace patch And I used subclipse to checkout the projects because it is easier to just let it also create a patch and had never problems until nowP The fact that svnpropset are inlcuded is because I had to add generated folders to the ignoring list +The reference to this book should be removed +So done and is useless +Committed to trunk +You might want to file a bug if you have thoughts on what such a policy should look like and probably should do so anyway so I remember to add a version field to compiled output +Ken could you take this feature request +Can it not be an extra operation on the +Patch moved to code with test Sang for evaluating and giving your inputs +OK fixed it for testmri at leastThe way is not really ideal but seems the only approach anywayThe source of the problem being there is no exist? for symlinks so it might return false because the target does not exist yet the symlink be there +Okie +Still saw the same problem described above +patch looks good +Seems innocuous enough to me +Thanks for the however when I try this it doesnt seem to affect the feature it still loads the saxonhe I will try testingagainst the +patch looks good +Committed revision in trunk with patch from Erik. +This includes text classification recommendation engines and clusteringThere are definitely better ways to represent a Lucene index if you want to preserve the data but if you want to use other Mahout stuff then a matrix is really required +I think problem is in commit Regards +Marking as a documentation enhancement +My earlier patch changed the input causing the testcase failureHere is the patch for Y! S distribution with the correct input to testcases +so the part that really matters in garys fix is adding the nulling out to shutdown +attached +In my particular case I use ant as a driver for some ETL jobs +IIRC there arent any classes in that extend types +Aaron this patch has gone stale +Use Dimitriss work around to get rid of exception +Applied to trunk as of been resolved for more than months +RayeesWhich testcase is failing? I see that testcusomhostnameinstancenameflase is passingRegardsGirish +Otherwise In the warning message still refers to active segment +My next attempt will be to set forkMode to never and try to just run mvnDebug and see what I see +Maybe we can draw from both of them and create an even better solutionThe only thing my prototype leaves on disk is +Its already done by the base class +This patch fixes a couple targets in the template for the distribution links used by the tool +Patch has been committed +This should be done in nodetool from the raw information rather than adding a new JMX methodAlso changes a lot for +Maybe Marko can take a look +The community cant access Citrix do following configuration to help capture detail information configure etc to enable slow SQL query logging following is an example line varlog in files append following to prepcachetrueprepincludetrueincludetruelogtrue +The links are useful for getting hold of the information that is lost in translation when moving to JIRA stuff like comments for instanceUsing a suffix of seems like a good way to do it +Thanks! or anther solution for me the subproxy keep a minimized schema the schema can fit for every katta node +Upgrades HDFS to and includes the guava upgrade required by this version +If such a stream is at the beginning of the document it will read too much data and the parser would notice it while trying to decompress it + +I can add in an about page without too much work +Hi Olivier +Should be an easy fix so targeting the above reported issues and a couple of into the master branch. +A bit over my head but Ill look into that +Manually tested +The cause of it is someone adds new commits after I opened that pull request +what if you have multiple warehouses or facilities and stores per organization +You do not need to get an acknowledgment from the same collector you sent to +Maybe add a boolean to the memstore to track if it contains edits that were not written to the WAL adds code which tracks how much data is in the memstorebq +Attaching the patch again +And as a reminder to check what the ASF does for us + +committed with tests and other changesIll work on updating docs shortly. +Symlink attacks are very difficult to identify or prevent +This could be improved but would still be clumsy +Itd be better to use something already without need to override Result just define methodparams and you are done +Both in dev and production +not to mention mavens all about best practices and that location no longer is +FredCould you please explain a bit more why we need to access trasport headers in a handlerWhat advantages would we have over doing the same through Stub API? +I need to direct instantiate the same classLoader you use on the application server +Includes a few significant changes to core +I saw that too but if you look at the implementation of you see that this is currently not supported +Thanks +Escaping of double quotes is now replaced by that of single for reviewing the patch Cedric +I dont see a reason why having a factory in contrib makes any sense it can also reside in core +This really smells like an issue in Spymemcached which is both present in and +Is this something we need to fixDavid is this causing problems for you +The damage done is the situation described in where a compaction would end after a RS comes out of GC and then it deletes the old files +try module name try method name and try modulemethod +Hadoop QA will pick it up if JIRA is marked patch available +bq I think there are some frightened enterprise software companies looking to spread some FUD +The default value in the patch is false +it looks like the following may change the final state from what was originally intended to be the final statewhat is stored as the final state when the attempt first moved to finalsaving stateFixed +I started up a stock AMI in hadoop mode so that DSE would create the cfsdsesystem keyspaces and also created an additional keyspace using the customers schema +Given that modern versions of Pig support boolean and that Hive supports booleans +Can anyone else confirm this problem +The test fails in the RI +r jwaldman linesChanged paths M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadresource M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadwebapp A myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadwebapp M myfacestrinidadtrunksrcmainxrtsorgapachemyfacestrinidadresource When a resource fails to load add an exception if the cached temp directory doesnt match the current temp directorythanks to Mark Yvanovich for the patch +I dont have time to put together the docs and juint stuff right now +maybeMerge has the ideal name +The thought was that we could create an interface so that we could pass the send buffer and a start offset and have the encoded data be copied directly into it +Surprising that the issue is written off like thissupport for nested sources is an eclipse feature not a limitation +Done as described +Should appear in the next beta build +Can someone please confirm my actions Downloaded latest ant binary +Though if we use profile and in the future branch isnt trunk wed have to update the poms anyway to remap profile to branch or whatever +Ive done some brainstorming with Stack and the result was +Again we need an intuitive public XSD API to use +Thanks Hemanth! +Peoplesoft metadatawhat information is there extra in there +Therefore the line parsing logic itself will probably need to be modified not to stop at a CRLF followed by a LWSP character +One possible implementation of the interface involves writing on an fd as is done now +Do you have a patch for this Thomas + +So no worry about external my opinion it should be possible to configure if this exception should be sent to the client +Please close this issue if it works for you +Fixed in rev +The test should have its own log +Set correct fix version +And one more how can I download version of Roo? Please help +Updated patch that includes fixes for some failing proxy tests +fwiw Im getting a failure in the Cassandra itests but Im getting that with or +at projectRoot remove the line of +Grgory do you already have a patch for this +The improved Build Test infrastructure can be obtained from Issue ve added the suite at r +but before that let me do my home work first. +Java maybe +KeithCould you please review and commit an updated patch +Theres been no feedback on this for quite a while and it didnt look like a real bug so Ill mark as resolved +will fail in the build since it will not have the tar file +That usage of transient may be a substitute for an Atomic objectWhere did you see the field +I thought I saw that but wasnt sure +Another factor is that the NPE is occuring far enough into the render that the response is already committed +I am not seeing this problem +bq +Because of this problem we ended up using RMI for our message instead of Camel +The use case is fairly limitted +You should be able to use jarinstall or jardeploy and a correct pom would be generated in your repo. +I was ok with either way +I think were better off just implementing writeFully and using that +doesnt support multitenancybut yeah if these settings are missingmisconfigured then Hibernate should throw a better exception instead of NPE. +Oleg we havent yet received a config example from the reporter but can you try to reproduce this? Based on that check in the parser Im curious how this could happen +Scott do you have any examples of paths that are wrongly mapped +there are a bunch of queries and filters +Ive used this for a while and Id like to donate this code to the commons +Many thanks Sam! in trunkOleg +With open everyone is empowered to try new things without asking to get their code into the core project +However It seems wasteful to remove that functionality since wecurrently have created a way to extract themShould we breakout viewnames and childnote text entries as separatemetadataAny input is appreciated +Ive attached a new patch for this issueThe problem was that messages requeued from a subscription into a deleted queue were not through the alternate exchangeThis patch fixes that problem. +This issue has a helpdesk case assigned to it but no Release Note Text was provided +Why was it patched +OK I definitely see that we have an issue storing property values +Resolved with revision of Commit message In do not create a separate +Thank you Mike and for feedback +Added a test case for reading and writing an UUID +reRender fixed +I have seen it slow the environment down some +Any comments on this patch? Should I go ahead and commit it? cant say I have thoroughly reviewed the patch +this +Thanks a lot +Merged into develop branch and will be available in nightly build. +Here is a scenario that has been happening to usHave developer check in a change to a properties file +committed +There is very little point to change anything from the current behavior +that falls under the byzantine header +new patch that includes a servlet to possibly replace due to the fact that it writes out invalid xml +If we rename to will this kind of convention be respectedWhat about simply keeping as bundle and extract from here the stuff to create some otherI fear splitting the bundle without renaming packages will lead to some ambiguity on which bundle provides which package +Use Template button states saved in eclipse preferences from its last call +Attached a patch for stopping SENDPARAMEXECUTOR when all clients are stopped with reference counter in +Resolving this still Id like to include more tests in the code base for CI to run automatically as soon as we have the new CI machines. +Thank you for the great work +There doesnt appear to be any equivalent of in DNJDO +Dave any comment +Ok +Thanks Jeff! +Ive committed this to trunk and branch +Thanks for the discussion! After talking about this here and with the other Ubuntu developers were going to solve this by splitting the couchdb package into couchb and +Ok to close this one +Could not resist and run the test on SUN and BEA +part of +The addition of integer seems over the top to me +Have not tested but reviewed patch closely and it looks good +ok checking it +Closing as per reporter. +We should include the classes from but not +So therefore I wouldnt see the need to use IPC via socket +I wonder if the complication of keeping state on the server side and substituting binary values into a new tokenized structure will really be fasterI suppose nothings a given but hash lookup is pretty damn fast +Manual instructions updated +Created a separate issue for refactoring the factory beans. + for the to trunk +Wrapper scripts are common easy and well behaved +In most cases the Fixnum construction will outweigh any minor perf tweaks to the logic +Libs with Carrot weve just released + +For this issue I think we only has to run the tests affected since this is a test code refactoring +Thanks Josh! +this looses at least no data if somebody messes up their types +No response assuming ok. +I suspect a class loading problem here +Node changed event processing was updated +Thank you very will try it make up +Lets be hype! Mx Emmnuel +Will be addressed in. +Initial code committed in revision + has been marked as a duplicate of this bug +Also I think we should allow regs on servers where TLS has been disabled similar to plain auth +The comment by Rick from th of May says Looks like a solid first increment of support needed for this feature +Please consider both and variants +Committed at subversion revision +Right now it just does GET PUT and POST +Thanks Siying +you want to assign this issue to yourselfFine +To completeexplain Jacopos request we decided some years ago its better to discussshare new ideas wishes etc +It seems to be blocking on a call to async while waiting to acquire a monitor +VERY cool +We are discussing this on the dev mailing list as far as whether we want to support this +I think a simpler fix is to have create only when needed +Do you think they will release a new version +Closing this issue since insert functionality using JDBC has been implemented for all the types of resultsets supported by Derby at this point. +stagingjob does not existBut I think should be already in classpath +Please verify that you use the server name when deploying +I was playing with your patch this afternoon and am starting to think this is better +trunk +good point I will try that and let you know +Without further details I mark this as resolved +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +This error seems only to occur when using the is fine +Thus +I polished it a bit. +For the cache I guess the choice is between keeping the invalidation or leaving the expired tombstone in cache +Thanks +We should use standard HTTP header strings like as opposed to +Works well with latest works +However still seeding. +Thanks! +Using link local addresses for initial testing is asking for trouble as they have special problems +Fixed in +Thanks Stephen + the most naive solution could be separate shared in the same classpathclassloader as pluginsYeah I considered that briefly but I had two big reservations about attempting that approach +v attached +Oops I missed the revision number it should be r +unit test modified to reproduce the issue +The dispatcher does not +That class uses describe ring to generate the input splits for map tasks to work on which obviously doesnt work if it returns the gossip interface instead of the thrift interface +Not sure how that didnt make it in +Id say this list is fine with QA +Likely it was the same as. +Hi JacquesThanks for persisting with this +I agree with Thilos analysis above +Unit testing looks good so far however more tests need to be added and some real testing with cable pulls etc needs to be done +It tells me that the domain name is already in use +Uploading the patch shortly here +Please feel free to any of those when required. +I think this is an issue with the transitions in the workflow +Also note that doesnt contain SHA sums this jars classes +HelloPlease try my patch +TedAddressed all your comments except the support for having more than onepair of values for the update request +Given that I dont care too much about what the description of this says but Im pretty much opposed to doing anything here that would make much harder that it needs unless there is a good reason and I dont see one +Wow didnt even realize there was an open bug for this +If the stub constructor throws instead of Exception then again no client code needs to change as all existing client code will have to be catching Exception +Im happy with the option +I have a patch pending in my sandbox +But I still think we need further clarification on what the actual enhancement request is +Reopening issue was reproducedThe patch that fixes degradation declaring classes as there any larger more strategic fix that can be looked at with regard to the use of +TODO Review the rest of the retirement list to see if anything else was missed. +Most probably caused by a WC corruption which can be only fixed by hand. +Please help to review it thanks +Merged to in rev . +thanks maurizio thats exactly what we need one question is this going to make its way into a formal release +Sorry for so many updates +I will take care of the comments posted by Knut +I think Ive seen this too +Remember to update the release notes about the API change on notes updated too +When finished reresolve with the new fix versions or label backportrejectx as appropriate +I think it affects many character sets not just russian Probably anything that is not ASCII +The Callable Statement returns a as an OUT parameter that you want to map to objects as normal with ibatis +Perhaps one of the more recent patches added in some test thats not exiting properly and leaving something running +batch transition to closed remaining resolved bugs +James Sorry to reopen this later on but can you look at the jsp handling in that example as well +Ideally the previous test was covering up the bug +IMO it should be a separate to trunks headscript patch version i wiredfindbugscloverratcheckstyle work +You would know which module because they wouldnt depend on each other +Fixed a couple of bugs together with stuff +Remove it and everyones gonna end up writing the very code you remove in seven different ways five of them broken somehow +Fixed the unit test failures and the findbugs warning +So we could easily inject the properties into the endpoint config +Switching to +Fixed in branch need to determine if trunk needs fix and apply there +ve committed this +I assume you already have that downloaded could you please verify that everything was committed correctlyalso please verify that all of the source files modified by this patch include copyrights it looked like the patch might have removed a couple of the java style copyrights from the xmls without inserting new XML style I believe I rooted out all the xml files that had incorrect headers and fixed them +It looks like a compatibility issue between visual c and +Use for concurrent threads +Although I had reproduced it many times within a minute on another faster machine it took me near minutes of clicking refresh on the metric page before the timing was just right at the moment that a discovery scan was being performed +ok this lgtm +Ill have to check into it +Added +I dont see a defined at all in Also the second thing isnt ideal +There would be no need to duplicate that +OK silly me +May be closed as duplicate +Michi +there were tabs in the patch +Word of caution I tested it last on GAE version and I am not using anymore since my requirements have changed +I created patch for this issue +Mladen WDYT? + I got it and will commit a fix in about my mistake +and seriously someone tell me why if central already include the AS feature which includes everything incl +Done +Attached is a XML file can you also add the log from the server + Alphas should track and hence update to JSF +The changes to correct the broken detach logicThe changes to fix a related issue where you could not detach and then storage for managed primary storage +I think this could be a further improvement implemented at a later time +this was apparently discussed in the past and there is no need to merge this patch into be merged into the CP by has been merged to the CP for inclusion in the. +Checked with build of installer seems to be fixed +Thanks atm! +The test is too brittle and time sensitive and has been giving us some headaches for a while so removing it +It works fine when I saved without importing the class Eclipse eventually complained the the Organize Imports save action was taking too long +updateseq numbers are included in the new changes feed +This is a test project I used to reproduce the problem +Instead manually verifying the WARN message should be good enough +No sense adding another variable into the mix when our current branch has some problems +sounds like a fine idea to me and an easy fix +doing it now +Reopen if this is not the case +is the HTML I get +to do that I will have to change all test setups +Im also under Mac OS XAnd is definitely not running from under X here by default +If the report isaccurate I would agree that it is a bug +Works! Could we get the help lines to reflect this +This is not a confirmed regression that might affect existing applications so unchecking the Existing Application Impact checkbox +Hey any progress on this issue? I have the exact same problem +It can be rare +This looks like a great start to C support for AvroOne question Do we need a separate COPYING NEWS or here? Well probably always distribute this with the rest of Avro which already has equivalent files no +Will continue to test +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved +Broke out second part to a second JIRA +If you still have a problem please use the mailing list user and I will be happy to help furtherPeople are working on improving the docs including which improves the interceptor docs a fair bit +The old implementation did not write participant band shapes to the DI which means that during import the participant band shapes are not found and therefore not shown +Phabricator doesnt let you download a patch +Thank you Danushka for the contribution +Hey cool how does that work? I only know the same like Robert said the format template has to be imported globally and can assigned per project +In that case the fix version would change to +Quite frustrating +This also applied to Change TreemapThe solution is to not display the savecopy actions when editing the visual representation of the filter +I agree that it can be simplified a bitMy original implementation was just about registering file extensions so probably Roshan can explain his idea with classes you mention +I am planning on letting this JIRA stay open until either both of those JIRA make the problem go away +Closing issue +Ive removed the from the beginning of all the links listed +Btw is just waiting for JAXB to be released and used in. +Suggestions +Committed because build is broke +I just interpolated from the exception message that the placement is wrong but I dont have deep guice knowledge +No problem maybe next time try to upgrade one major dependency at once? It will reduce time you will spend looking for a problem +Committed revision +It looks like this was just a matter of the cloned stream not being sync if the original stream was sync +Closing resolved issues. +Feel free to reject if you dont want to depend on spec API in community +We still need to push on +what exactly would it be confused with? If you downloaded the sources and playing around with them then maybe there would be some confusion since the folder that holds the different types is named adapter +Committed to branch +The virtual is this model project can be executed with US as the required physical can be execute with the followingselect capitalcityresult from capitalcitysoaprequest req capitalcitysoapresponse res where and US +Ive just watched how it implemented in the same class for different property and do it that way I think its not our task to take care about additional overhead it must be a part of the Interpolator logic +Are you seeing different behavior? +Zhihong Yuupdated the patchthis patch for versionRamaI will check the TRUNK and version +ant clean all +Did you declare in your project dependenciesFrom the information you give above it is still not clear what you are doingwhat is the goal you ran that lead this error +you are right +The subsystem is available and can now be used to provide a better user experience when deploying applications using the documentation is updated with more information about how to install the modules subsystem and configure deployments. +restoring orignal owner dont plan on backporting any farther back now. +Its broken on a higher level so we cannot do anything against it anyway. +Ah +can you please try running the test without the patch and see if it passes or not +For example extending the property format to plug handler on matching factories only +Ill put the statement back in but instead of checking for filterProperties! null it will check to see if any filter operators are present +Bug contains a first small patch to enhance the error text maybe this Bugis a duplicate of Bug ? contains a first small patch to enhance the error text maybe this Bugis a duplicate of Bug contains a first small patch to enhance the error text maybe this Bugis a duplicate of Bug +Prior to the patch is ignored if is zero. +example a counter like Bigtable allows cells to be used as integer counters but with this type of check on a update the column could be anything not just a integerso I will on voting for this +It is quite surprising that this issue while flagged in various over the past months seems of rather low priority +So works with Servlet etc +Mircea test passes with my latest pull req for +Please SjurAttach a plain xml file with the encoding where you have problems on this bug report so we can test it +The code is in the repository a prerelease is coming soon to expose remaining issues +In many cases there will be no streaming columns and get will never be called +The build subdirectory is only for building the distribution +Michael thanks for providing a clear response +Sorry +Ive removed your changes to and fixed conflict in +You are trying to render as a polygon whats in reality a line SLD says we should close the line if needed and fill it +Version Released +I am going to see of other replication related scripts have this problem and fix them once and for all +Just one more thing if I put an outputlink inside a menu item and I click in the menu item but not on the text of the outputlink my click is ignored and users find it very annoying +Verified in ER build. +I have attached a proposed patch to that includes the GOP fixes here ports them to GKOP and also includes the sync fixes proposed in. +If my srcmainresources file has an odd number of chars in it none of the variables in it are filtered +Issued resolved via the new interface which supports all application types +Working with so it looks like it has already been fixed +Because of differences I question the benefit to porting all this to maven simply so that a single component can bootstrap itselfHaving it in Maven makes ti more readable and maintainable by anyone who know how to read a pom +If nothing else it will get rid of my own scripts to do such things +Its something that sun and joshua do for the sun libraries but even they have internal stuff they dont like you touching and which moves about from time to timeThis is why the idea of having an Internal tag appeals to me something you can put on an interface to say this is internal track SVN Head if you dont want to be surprised +If I were you I would use a the module and set up the correct and to go to your oracle db and try to run one of the unittests that creates a business or something from eclipse +What do you make of adding a test dependency for an embeddable database such as Derby or HSQL such that the test doesnt require +and you cant run JBDS w JDK anymore +Perhaps the easiest thing from the users standpoint is to have a separate Xcode template per version +Thanks for your reviewCheers +I can only assume no response means it is an outstanding issueWhat do you have to do to get support from the Hibernate TeamWith all the time that you I and Gavin have wasted going backwards and forward like this the issue could have been resolved a week ago if a meaningful dialog was established +An earlier attempt might not have removed the directory +Thanks Jason +I guess most installations are intranet so no need for SSL if you have on the internet you will have it listen on port but you wont run Tomcat as root so you use Apache and modjk and Apache does the SSL stuff +I say the Unix hackers win this one +Regardless it should work +Even if we use the RI tools to genereate the WSDL the runtime fails with the above stacktrace attached by maomaode +I will send out the patch for +It has not been back ported to release. +It has already been committed +Yep this is intentional +Considering Im not free recently I reassign the issue +Id make configurable and leave the default to the current behavior +Perhaps moving the setting of master to become method will solve the problem +Could you find out if append is synchronized? Once you know the answer I can commit the patch to +patch to make log thrown by the courier +This patch also combines the fixes for GRE and +The warning message is also wrong because JRE is required to run Geronimo +is for reseek only as we know we only scan forward in that caseSo it looks like is working as expected for reseeks +Think through the right thing to do for each call to the deprecated methods +Test case refactored now it will be launched with all other tests and it decides whether to skip itself or not depending on version of JSF +Please make sure you post any results from your conversation here so others can follow +Sorry I didnt notice the absolute paths +No thank you for fixing it +Released with. +Closing old issues +I trigger that queue at the very end of the whole response +Could you please provide a class that is a simple test case that demonstrates this? +I was able to remove the hack I had in my code to work around this and all of my tests still passed +No this is not tied tobq +notNulladd field string class +Closing all resolved tickets from or older +Whats the status of this feature will this ever be implemented? votes and still no action +Please review and comment +Potentially related affects functions which throw I commented on this is an invalid option. +Currently references to println. +Werner Ive made a mistake while refactoring concerning svn mv sorrySo heres another patch for deleting the moved . +il +Allowing the null values allows the junit tests to behave more like the production already have enough problems with formsuccess views with redirects being different +Fixed in for and trunk branches +Id like to see your patch +phew +Interesting approach +But I just took a look and it appears you did exactly the right thingSo the next thing we need I think is the devuser mailing lists so we can start getting conversation happening over there and I need to update all the lists ws about current move status +After an official release it will be much harder to improve the handling of temporary resources as its part of the public API +Please close this issue. +LGTM +In fact was adressing c c issueBut you still have somes issues with such convertion ctest cTemp +Thanks +I dont want to bloat this Jira +The process of getting from an URI to a losesthe credentialsIve tried patching it myself but this patch has a lot of impact +Thanks Dan for patch and keep them coming. +Since the review for trunk had relatively minor issues Im going to work on this to the branch +Attaching a patch which addes status metrics +Yes reducing the job latency in the presence of reduce failures is the reason for having multiple wavesBatch transfer may not be a big deal before because there were a lot of inefficient places in the shuffling code anywayHowever it will be a significant improvement forBased on my observations on running gridmix with the fatching part becomes a bottleneck of the shuffling phase inI think it is definitely worth to revisit this issueBased on my understanding of the shuffling code it is not hard to implement keep alive connection in the fetcherHad an offline chat with Devaraj about this issueHere is a proposal based on the chatWhen a reducer need to fetch map outputs from a TT it sends a request to the TT with a list of segments not just one it needs from the TTUpon receiving the request the TT builds a response consisting of a subset of the requested map output segments up to certain upper limit on the total size +Im just looking in the system properties for the three properties I know Im using +Is a duplicate of +Ive read the README and I know that it is currently not supported thats why I opened this issueIMHO it would be a nice feature to not be forced to unpack the +I was going to implement it if it didnt exist but since you have made such progress Id like to team up if you are amenable +Ivo Ive corrected the typos and am rebrewing for +Ive added a fallback in case the uuid is +If so any ETA on a fix +This allows the classpath to be resolved for local mode +Here is a packaged war that is only missing the Tuscany Libs to be used as a test war starts up fine with Tuscany release jars +We kinda use that name for similar optionsI added a Spring XML based test as well to test that it worked and we can use that for example in the wiki documentation +Alan sorry for forgetting this +I prefer to keep it that way to avoid them againbq +It does not mention the constraint that the context classloader cannot be changed during the lifetime of the application +Our migration from to was immediate so im not sure about whether i have tested it inAnd no its not the side effect of that issue since its a reproducible error from your end +Here is a simple fix for there was a wrong assumption of no +Please let me know if thats incorrect +Doesnt support range query +The local states are pushed blindly to the shared states potentially overwriting a previous changeFixed by locking the journal in instead of +Yeah I didnt really know where to put this as it is a tool but it is also a replacement for fetcher +Though it wont be a well do our best +A bit more descriptive for Chapter The optional discriminator attribute specifies the column to be used as the discriminator for the embedded object +Marking for so that it can tracked accordingly. +Updated to use Attribute not abstract Property descriptors +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +i dont agree with the resolution Incomplete +Commit first step changes to trunk At revision Next step +Fixed by shuffling the internal test a bit +I dont see why this is marked to the logic there does not seem to be faulty at a glanceDo we want to add a test with a quorum or do we think the unit test on the predicate logic is enough +Need by the artifact plugin +We will get a EAP release to fix this and will base SOA CP on this +So it fails on both and bitAbout the test chart panel then its close but the real useful chart is the progreschart that shows how things changed over timeIs it possible to have that for this view or is all just latest numbers +Not sure how much this is possible +I will do the needful for using in my next patch +fix in rev problem is that quotes are placed around the CLASSPATH variable on the evocation on the second line +Updated patch for +If I send the email to our company I can reproduce the errorI try differnt email encodings and my patch works fineIn my opinion its an ms exchange problem so you can close this issue +As I commented before javacc warning is because we use which is marked deprecated +This jar was coincidently used by Drools andAfter upgrade of Drools version to the disappeared from seam distribution Drools changed dependency to instead of +public SSH RSA key wants his user name to be rwilcox +resolves to resolves to + +We are implementing better way to achieve that in using system table +They are also much less of a hack +Would it be too verbose to go even further and suggest the remedy in the exception message +bq +It is fixed now +I did have to run eclipse with clean after altering the file before it noticed the update though which isnt mentioned in the linked forum post. +The functions of and are also replaced because they are said If the string is not in the Host Portable Character Encoding the result is implementation dependent +These appear to be the two models on the table In this scenario unsecure is a generic disabling flag with more or less the following meaning Let my legacy apps run unmodified +Id like to take the oppotunity to say THANK YOU to Joe and the test of the infra team for the smooth transition of Ivy. +How about explaining the situation as it is on the page? +slightly the way this works +Sanity check on j jdk and jdk look good +What does that preference do exactly? It was set to false since I upgraded to but was still working for me +Thanks for the review Andrei. +Take a look at the test assert two facts outside of the rule engine which should cause the rule to fire twice +Patch for trunkIncorporates and is not for commit +Docs will be ready for release at GA and not before +Yes I can confirm this one +Therefore sonar is gonna run +Patch is generated using git diff text HEAD +Attached patch where the is removed from the OR row in add maven project +Looks like it works now +Keeping this open until we get that reported in eclipse is right but the eclipse bug gives no context at all not a big chance they will look into that +Stuffing SASL exchange in seems to be hack to me to avoid fixing the bad version response +Modified Chetan I applied your patch at r. +The fix has appeared in the Latest Alpha Manuals so this issue could be closed +Add one more to the list of IBM JRE users +Updated patch with new changes +Even more so when the lcrypto is already fixed but not yet backported into Hadoop afaikWe need to figure out what happens for each of these see if they are alreafy fixed and get them backportedcontribute a patch if needed +Files seems that this specific version of pdfbox is not on central repos +Now I have +on your patch looks good. +Yes solves the problem here no more exceptions on covariant return types +downgrading to minor since this is only about a label that disappears after refresh +I also tried the patch in but that has causes some end of file reached errors for us +Im going to wait until all the unit tests are passing and functionally it works correctly +But the scope of artifacts are not known to jsf and better not to support from jsf +Now that Im checking it seems save did not have the GPS data +The new patch +well +I have been watching this discussion casually on +SVGs publication the svg images still break see template definitely wasnt tested for svg images. +I prefer Ubuntu +If however the answer is that this number is used in some other calculation lets figure that out and put that case out here for discussion +TOM is a safety net to protect some region from stuck in transition +This is perhaps specific to Hadoop branch but it would be extremely nice to split Hadoop packaging into HDFS YARN and MAPREDUCE +Updated patch with test case +Password sent to private this is also the email address in case a reset is needed +Hi guysjust my additional cents the EL implementations behave differently with return values +moved the rest can you remove from pao wwwrepo you have ownership and i cant +We can add one more trigger Application state change +Since it is only for demonstration purposes we have assumed that you have installed AxisC in your home directory which was the best option when we do not have a proper axutil library that can load modules given the relative paths +The patch for the improvments +Fixed in the SVN trunk +the in wasnt needed +The issue can always be if someone wants to contribute more. +The stack trace looks similar to +Ive thought about it but never done it because the query parser does not support multiple statementsWhat about just using for the seperation +reopening to change the component to Module API +do you think it would be possible to get a cheap implementation with a single mapper performingmultiple dimension joins one after the otherYes + +Changes the logger to +Apparently the field value is not compatible with the grammar defined in and also likely to violate the strict interpretation of RFC +Also single diff should have all the files including the testcase and its resultAlso add the both test files and the outputs using svn add and then take the diffcan you drop the new tables that you create at the end of the testcase so that if there are any show tables test cases will not fail +Before closing this bug I need to make sure that it is not reproducible +I have fixed the problem in the patch Im attachingThere is one tiny thing that is bugging me in this patch and I need to fix +If we are happy with it we can merge this branchA new jira issue should then be created to remove this workaround when the maven shade plugin is fixed +This uses the optimial buffered image type on Windows +Thats the approach I intendBesides you can leave any comment or express your agreement and disagreement on jira or mailing +This should make the next Continuum build progress abut further as according to the last run adaptx has been the last remaining unresolvable dependeny +Patch from Bug allows to build and successfully run testssamples +If there are no live nodes capable of handling the memory asked for that should be looked at in. +Resolved by adding name atribute during mirroring for final JBDS Update with and related Update Site. +One is an annotation and one is a regular class +just committed this +You are welcome just next time please use the mailing list instead of jiraCiaoMario +If it works with a newer jossl version then that is great but I would still like to understand why we have this problem only on windows with the current jossl gem +Moved back to as this is a critical Deng You just forgot to attach images Ill check tomorrow +Can you test again +Marking closed. +Trunk and already have code to relogin before download in that should take care of this issue. +This patch is temporary tests are still in progress in Karaf +Jochen seems like you forgot to add +If you configure it it will kick it once it is fully configured this depends on the order of your bean disadvantage is that you wont immediately notice when your transaction aspect is not active at runtime +Thanks for the patch Koji +It also renames three methods in those classes to make it clearer that they dont encrypt the passwords but rather hash them +Thanks for spotting this. +dims +Verified in build +The EAP docs should include thisThe standalone distribution already has the XSD +d of master branch +I tried also a other project where i had similar problems with and meclipse With the project was deployed Oliver +I was under impression that to only read didnt required admin permissionsMaybe the only thing to do for this one is to make the message better and clarify that the x and not the x is the one required on the path +I saw the errors only once +Im moving all bugs out to and we can move some bugs out to to get some release action going +Try avoiding a NPE by using a bit more ifs +This updated patch replaces and reflects changes incorporated by patch andpatch p i patch p i You need to set in proxyconfig to a valid interface like ethYou should be able to bring up the trafficcoptrafficmanagertrafficserversudo usrlocaltrafficserverbintrafficcop To shutdown trafficcop you can do the following on linuxsudo killall s QUIT trafficcopThis will currently only bring down trafficcop +And what Ted said +Merged to +A quickstart example showing the problem +I will have a look at the code to see what you are talking about +attach patch +bumping to the next seems ok with in offline mode +I wish it was really more openMaybe switching to a different serialization format as suggested in is the best bet +We might be able to optimize this further Ill start a post in dev forumRe If you can find any other alternative impls let us know! Its a while we havent looked around but it seemed to be best around with a decent licenseNow I suppose the difference between his previous measurements and now is the fact that loading happens still within the same thread +push out to +resolving. +Note that the code doesnt compile because the packages are the packages of my project change the path in the source the unit tests require my Seam testing infrastructure but can easily be adapted to a different system +Take a tip from aviation Unix and Java time values should always be stored in UTC because are viewer dependent +Thanks issue Oozie is released +OK the key was to get the latest file +The ability to turn it off and on cannot be done per job based on that implementation but could readily be specified per Solr connection +Changes are in we now use from jUDDI v. +yes I will +I have tested the patch on Ubuntu using Oracle and the patch seems to do the right thing there too +Start ij and run and +There is no package any more +I dont know and dont care if the problem is in trunkThe changes in Hadoop always first go to trunk and then to older releases to keep the trunk +Thats not to say different approach which addresses some of the comments in this thread wouldnt be commit support +compute MD checksum when saving an image +I will investigate it now +Juven please attach your Eclipse configuration details from Help About Eclipse SDK Configuration Details +Is there a plan date for fix this issue +Please verify if this is working with embedder defined in doesnt seem to work in embedder I am also seeing this issue +Im working with the latest a look at trunk and it is the almost exactly the same only minor changes in that class +The other one I am thinking of was specifically withing Configuration itself when trying to write out the configuration to a stream and the reader of the stream in the same process was trying to read in a configuration value +Now also gets that +Sorry new patch forgot to change tests to LUCENECURRENT as Uwe mentioned +I ran into this today too thanks for the workaround and fix +Common stuff between the server and client codeFor some time while we move around thing I dont expect much changes to the patch but a lotsa changes to the script +Right good point +The max in Oracle is and this is right +Looking at the packages in all mailet projects I dont see any packages split across jars so there shouldnt be any issues +I dont know this file well +Most should contain these classes in their +But it should work +Further infos If I do the configuration with vi or another text editor everything works fine +Looks like a duplicate of of mistake have not checked already resolved issues. +Great research EvaI would opt for the second option given the time remaining for this releaseCheers think there must have been something wrong with staging this document when was malfunctioningI will repush the document to the stage shortly +The finest level of granularity for this parameter would be at the table level since a region split affects all the columns in a particular row would like to see this option added too +Attached patchI tested it successfully against and +I suspect this means that emm addressed this issue somewhere in the last year of work +no subject line +Will revisit for so changing fixversion. +This patch removes the xml declaration from the final outputFurther it add a default namespace to the html element tag. +I am in the process of refactoring it use the simplified extension support classes in core +Indeed in this case the listing feature was listing the modules in the target namespace instead of system namespace then trying to install the modules and leading to the message about inconsistent information found +Consolidated patch for integrating from federation to trunk +Ive avoided the exception +Hello RuthThe patch has been committed at revision r +If it still happens after I commit Ill track it down +Please review your items if not ready for please move them out to future release +From now on is set as an attribute in the http session only after creating it +Should I open a separate issue for the build problemI do not think that is necessary +I just committed this +They take some time to reflect and in transition period no trap will be generated +Fresh copy of the example configs + patch dirty and not optimal +The Affects Versions should include +You can go ahead for this one as well +Looked like tweaks +I fixed the last issue I had in revision . +One aspect of is the ensuring consistent lower level test coverage +The src folder is all the Java files and the war folder is a completely self contained WAR +The patch is ready for review +Because you explained what the problem was and the change was minimal I thought it was sufficient to treat your initial entry in this JIRA as an actual patchPerhaps Dan or someone else can help you with the Maven issue building CXF locally would help in case you spot some other problems and decide to submit a patchCheers Sergey +uploaded patch with you also take care of avoiding lines longer than characters guide line for the changes. +Improved patch cleaning up XML test suite classes accordingly. +Finally able to finish this one +Hi GianmarcoMailing list and Jira should be a good place to discuss your ideas +Perhaps you want to do it then +Im dieing +You could turn your configuration to get this working in the IDE import the three specific config classes into the base config class and use the profile annotation to distinguish between the different settings +StefanoYour proposals are more verbose and less performant than the current approach but I really dont care enough about the DOM API to continue this pointless arguing +But this is wrongI dont see how you are able to do project building in eclipse if the validation of the project fails +No +hcommandButton can get form field value +batch transition to closed remaining resolved bugs +What you are seeing is different issue as we already explained in forumsYou are welcome to open new jira issue to address this new problemAs main subject of issue was resolved i am resolving this +Patch Small cutover thats cutover w patch fixes lucene javadoc errors +Basically because code changes should always be accompanied with respective tests +Are you sure about plugin and Idea version? See and +The result is the expected list of closed issues +This patch cleans CLASSPATH variable if or options are available but Im not sure if its enough +Also edited title. +Thats a nasty bug +changing minOccurs from to solves the issue which element and in which schema file +if you would be so kind to point me to the bzr provider source code ill have a look +Lets try to fix this for +The patch looks correct to me +Im reviewing this patch +The mail api was not able to send the binary attachment for some reason +Added cost for reion count skew Added cost for moving Added cost for locality added base class for costs based upon region load moving average added cost for read request added cost for write request added cost for storefile size added cost for memstore size +Tried to reproduce and mine was fine +So to be clear this ticket is about the branch +Looks like this is not happening forClosing +I wonder if black and white might not be better since then its more identifiably an orca +OK Ill work on it when I get to Austin on Sunday +Linking related issue +Updated the summary to reflect more the problem of this be ok now. +is pretty old +Ive applied your patch in SVN trunkThanks a lot for your contributionMaarten +Hi the workaround didnt work for meHow soon until this issue might be fixed +fixed the failed this is a patch for is not closed by +Fixed one of the failing tests and another unreported oneThe failing test is under investigation but is cosmetic +NET client +Ok Im working on downloading the virtual box instances and getting that stuff going but am wondering if you can ping Alessio with this issue to see what he thinks +Hugo before creating issues please First ask questions on the user mailing list Second search in JIRA if the issue does not already existThanks +I reviewed and it looks good to me +Test case to show the errorJust run to get the +I think making tomahawk for JSF with JSP upwards compatible like Werner proposed and using the new resource API for tomahawk and facelets is the best way to go +I have never seen the error on +Turns out my understanding regards overloading in java is incorrect you can have different returns as long as the signature varies +Fixed in the Committed revision Thanks Suran +This task depends on railspackage and takes the environment variable PLUGINSDIR which is the path to the vendorplugins directory to which the user wants to install solrb +Closing as incomplete +At least thats what would make sense to me +patch with test case for this issueIll commit this just now +Eric can you evaluate Susans issue and plan an appropriate place in Doc to put this information +DBIncubatorApache TSIKWSewsIncubatorJDOWSkandulaIncubatorlucenecWSsandeshaIncubatorwodenWSWSSJWSNew singleton categories forForrestHarmonyThe only project is no InfraApache Geronimo Apache and Apache all changed to not have the Apache branding to bring in line with the other categories. +I think I know whats going on there now Youre using the same Configuration class for all of your all running in the same +For some reason it got saved as iso encoded file +sorry about my ignorance but is Glassfish v preview a release ready production enviromentsI thought Glassfish v prelude and preview were still beta +It was filed as. +This looks broadly reasonable to me +For example the reason I didnt use Histogram is that it is impossible to implement current recent metrics whose value is cleared for every access +Please let me know if that helps +All occurences of in the whole section should be replaced by The paragraph following the Warning is out of the page borders in the PDF +So I understand that it is not solved yet? Is there a deadline for this to be fixed +Hi DavidPlease help to review the patch it should fix the failures of the console testcases +Provides interface and implementations based on only +But to be honest I do not see such a goal brings much benefit comparing to letting javac do the compile job except this way gwt can handle the build classpath explicitly +Patch looks good +Feel free to try it out and comment on whether the improvements meet the requirements of this issue. +I also added a spec. +committed +Heres one more svn commitSending loadgenerator file data +render fixed +Just a few lines of code but I still managed to submit a bug +So Im closing this issue +were backported to as well adding the latter fix version +This loop iterates through various security mechanisms +Weve rolled a release w this now so probably best to make a new ticket sorry. +Closing since the code provided did not demonstrate any issues with Spring Security and no updates have been made +Thanks Enis for the comment +Please attach a sample pdf +Sorry missed the change +How can i upload my source files whether as a +didnt spot that should be fine then +Have you tried updating the statistics or using a later Derby version where that is automatic +Sigh +Please check if this is correct +In the test i was building on pre commit was screwing with things +Can you provide an example showing how to configure httpclient? Thanks +The intuition should be the other way around right +Please contact your administrator to be granted the appropriate while clicking Appearance link it brings me back to the log in page although Continuum appears that I am already logged the way it is time is fixed +looks good +Ill try to look at this patch this week +eTags and are one of the requirements of the parent JIRA +Ctrrl is occupated +I think we must also test the validation with the code base in SVN where the bug may have probably been solved +our unit tests I thinkIn any case Ill try to do a small test case to verify it +Implemented the hemanths changes and also added testcases for high ram jobs and user the suggestion from just committed this to trunk +Ok for fix just to know what will we do add it in during releaseprepare set required in the mojoNote its not really a bug but just a user convenience +Hi Lars could you post some preliminary benchmark numbers for hsync vs hflush? Thanks +Thanks Nathan +With the new codebase this now works +I had a discussion with Devraj and realized that the issue being addressed here is to enable JT to renew tokens with a namenode running different versions +Here is the second diff I think this uses the value stack properly +HiThe explanation of seems to be old as well as +the request is not served because there is no data for it due to the premature end of streamIn order to fix this I created a function apacheapgetclientblock in apachestreamread +This diff is against the trunk as of now rather than as its not a bug fix +Thanks very much! Ill try to find some time to test it out soon +Is that the case +Maybe it was not enough to just modify resources? Should we also change versions in manifest pom etc +Yongqiang added the following virtual columns as part ofINPUTFILENAMEBLOCKOFFSETINSIDEFILEThose need to be documented and we may still need a row offset. +Can we please ensure that this patch is working OK as I manually installed the two maven plugins so I am unsure if all will be working after the exculusions have been added to modulesThanks +Since this was a minor change to the test that fixed the problem and this test is a bit of an annoyance for testing I merged Fernandas test change to with revision . +Invalid conclusion is incorrect +However we cannot do that for dynamic partitioning as we dont know the directory names +Adds the action to the composites button bar +hereit is the diff my netbeans did something strange THI IS THE you attach an example of how to reproduce this using only the Style or the Relation componentsCause this may very well be an ognl issue so changing the getters is just a workaroundAnd also get is a totally acceptable way of creating boolean getters so without more info Im inclined to closing this as a Cannot reproduce +thrift files +Will commit if tests pass +This improvement could make the interactive shell much easier to implement. +Is there a current version somewhere +Manually tested +Please ignore the first one +you can enter it into the form now +However all I have been completely sucked in by one of these projects +And here is the patch fileCheers guysok i will check today and to fix the issuesThanks for reporting + +of for the report anywayI marked the issue as a duplicate of and changed the fix version to +forgot to update my eclipse classpath with the latest snapshot I got the right output now many thankswhy do you have null value for the returnTypes argfederica +I have already tried the clean flag +The interfaces are bound to evolve when actual implementations are done or become different from what is in this patch if we end up finding cleaner abstractions +It would be difficult to write a test for this +Implemented the functionality of Accounts +which is is now deprecated and maintained only for purposes of backward compatibility and will be removed after about more releases +This patch makes all the minimrcluster junit tests except the use multiple directories +Try using edismax for more leniency +We need a component category for the sandbox so marking as Test for now +to commit +I dont see that this has been fixed so if its still an issue can you look at. +Attaching patch against current trunkPatch contains fix for the issue and new unit tests +Closing since this appears to have been fixed almost a year ago. +Geoffrey De Smet added a comment May AMIn jboss AS its also failing but diffrentlyThen it must be sth we introduced to master this week +It uses a default configuration which holds Elements in memory for minutes and will overflow to disk +We have identified a patch that we believe will fix the issue +The dev version I used came as an attachment on one of the bugs +It wont work if CM runs as master but wants to kill hbase process running as hbase user but that might even be desired +some bug fixes in key derivation calculation +That means I want to be able to assume that things will continue working if all data is lost in the disk for one server +Stock install using the all configuration +Youll need to run the example with the Flash Profiler or Scout to see if there is a memory leak or not. +It does not include any tests the process for testing entry points is new to me +Its a major rework of legacy code +We have to check the proposal see if it applies to too +Thoughts +It was not clear which test timed outRunning through Hadoop QA again would be wise +As to the Jira title either way is fine with me +Just to fix the correct culprit here is +Im also not familiar with the Tika project structure is there a place I can drop the POI jar file and maven will recognize it or is there an environment variable or PATH I should set somewhereIve noticed in the that it mentions POI but just as a comment and version numbers for a few properties +Closing issue. +There was an issue in generating wsdl due to not properly closing the out stream +and tests pass +Want to setup a windows build on jenkins? Is there one up on apache jenkins? There must be +bq +Are there any updates on this topic? +There are a few tests failing but these are due to other Weld implementation problems +I assume it works for both and MSSQL? your change in Git ID cfbffcdcaabecdfdad +Im not looking for doc just the additional knowledge needed to use the pluginAn example? Do you test it? If so provide the test +Just wasted hours because of this one +Please test beyond if possible on Windows +It fixes version to allow arbitrary strings in addition to numbers so works +This shouldnt be allowed to is not a gc causes schedule can trigger a gc +bq +I can remember about a discussion showing only tablesvies with geometries and we decided that this is not a good solutionOpinions? +As a first problem Harmony does not have such file instead they have +So are we on a consensus here or Ill mark it as Wont fix +Remove request dispatcher options parser and unit tests moved to the this now everything has been updated and adapted. +Why is this reopened and assigned to me? If this is about instructions for how to install it needs to be assigned to Fernando or someone on his team +And the script does work on Windows provided you have cygwin installed +Min Is the suggestion looks good? Can you close this one for releasethanks to and this record since is now released. +Can you grant license to ASF for the attached file +Or leave him ignorant +It only passes if optimize is falseSo thats good news for youyou know why the problem is occuring +There should be no manual steps required by the user to get the latest jbosgi release installed in the last stable ASIf you look at installers from previous versions youll that this is how we done it with AS too +Some changes to the last attached spec file I think the package name should be as it differentiates an rpm from something that might be provided by Riptano or similar +The newest patch is should use it to match the server sides change +Removed all abandoned branches. +Uwe Im so glad that you are so keen on stuff like Java serialization! patch +Now it is users duty to add this import property when using the Teiid Connection importer to properly import the metadata from the Hive when using this particular version of the Hive driver +Ive applied the patch in r +It passes unit tests and +We could release an installer for JBDS +Changing to so that it gets in the next release is already out +I love weird special cases +But now it looks like the failures can be reproduced reliably on a Mac as well +I suppose yeah for now Ill disable it since dont have time at mo +Since this is a feature that could help people use RF as a classifier even more than it can now i guess you can keep it for with some documentation ofcourse +In that case a subsequent restart involves a recovery process thatis causing the db to go into a recovery process the next time it is started. +These files are my changed versions of the files from the srctestorghibernatetestannotationscid folder of the +Im going to move this down to trivial +Thanks NicolasYour patch is in trunk at r and at r +Right +Ive added a note to this effect in the wiki page +This patch is for branch +has tests is probably +Thanks Caleb thats good news +Thanks Namit! +derbyall ran okay using Sun jdk on Windows XP +Patch applied with minor changes +Failure to support leads to nasty silent failures because you expose developers to silent capture of stale test jars in their local repo +This is not where the problem happens +The documentation is comingIll explain all the pros and consThats the reason why I havent sent a mail to the mailing lists with announcementIts interesting that you ask for docs in the same day +This patch has been submitted to review board as well +This one DOES +Sounds like you incorporated + +Yes you are right when I use the testcase I am able to see your resultcfafbbfa Linux amd Thank you for responding so fast to my requestsHowever what still bothers me now is that the testecho n aP mdsumon my linuxshell still producescfaafeccebffdfeMaybe we should check this on an alternative? +The user will be able to see why his command doesnt working in the first place by using lineage tracing that somebody is working on +But it is not an efficient way +Thanks Nigel +Test passes locally with this patch +Stacktrace from maven log attached but it only informs about the fact that it could not reach that HTML page +Should be also includes Project Undertow +Requires a configuration to be set to point to location of tez jars and dependencies +Harsha thanks for providing improvements patches +Thanks for the update patch and commit Vinod. +The process is if the current directory does not contain a directory then the script will start looking in parent directories until one is found that does contain the dir +Cating the regioninfo file youd see binary garbage followed by text representation +Resolved in commits through ffebdcddbbaefecec +Thanks a lot Nicholas. +This patch should fix it can people give it a whirl to make sure it works on their systems +The patch section may need some additional info at some stage +Marking this in the tag file. +Any objection to applying the patch to convert the method checks to IODThat will remove the overhead for applications that dont use stripAccents +If the field is nullable in the constraints then the conversion error does not get created +thanks +Andrew Thank youIf you point me in the right direction Ill send over a patch +submit the patch again +I will attach a patch +Im pretty sure this should work +I did some poking around and found that this support comes from the PROJ library +Also it will automatically heartbeat with the RM so that the AM does not have to worry about it +I am fine either ways nice to have +Yes unwrapping is still TBD + edit the patch as a file +The only reason I have not gone ahead with this change is that it introduces a new interface that will make gem incompatible with earlier releases +a test case to verify exact type of exception being thrown for optimistic consistency violation +An exported sql dump is MBFree space on the VM is GBAre you saying an JSON dump of the DB will exceed GB +But in trunk its resolved +Thanks Arun! +Scott would you be willing to have this committed as a compatible extension of the existing templates? If so then I think all this lacks before we commmit it is mangling of getClass and getSchema perhaps to be getClass and getSchema +With Drews patches it will be even better + +My thinking was that I could save a lot of development time by extending the rather than implementing one of the interfaces or extending an abstract class +Closing all resolved tickets from or older +After setting set TMPCTemp before install It Worked! Its installed now Thanks a lot +Please see the comments in that issue if interested +Now the handler classes are properly loaded both from root lib folder and service archive +Will check what that patch does and understand it and then come up with a fix +May somebody pls try this +Fixed the metadata at apache Vincent needs to look how this got in central sorry +As per the description I think we do not need to change the underlying assumptions at this point +Fuse MR trunk is using CXF now. +Thanks Tony +Any new issuesenhancements related to package manager will now be tracked in new +Oh +No it throws exceptions and then outputs a message to the effect you saidI must say that on the surface I completely agree that it seems very odd that having no logging binding on the classpath causes the entire application to fail to load +I can do it if you like +Looking at the patch this failing assert was not now passed +I have now attached the failing code as the file +Is it possible to release a new version of the plugin? It would be better instead of using a SNAPSHOT is good practice to generate a new version when bugs are corrected. +I have committed this +with Elliot sorry about the delay on getting back to youI dont think theres an issue with sharing an executor +Closing. +However the problems are visible only when CDI comes into play +Thanks to AllAnkit your slightly modified patch is committed in trunk at revision I simply changed the ecommerce request map in ecommerce no needs to be authentified to be abble to remove a promo code I also added a French Jacques +Interesting + have attached the service aar file and source files and the tomcat console outputI invoked the service several time and I didnt see any more service instances creationSo pls try with the attached service and pls attach the log fileOr provide a way to the problem +Test patch is more +Indeed the JMS package is not modeled analogous to the Hibernate and JDO support rather follows the JDBC package The root defines exceptions a core package contains template and callbacks a support package contains reusable stuff thats not tied to template ORM package simply has a different subpackage design strategy with generic root and support packages plus while naming and co are always subject to individual preferences I believe that the current JMS package does fit into the Spring analogous to jdbc package rather than orm subpackages. +Please review your items if not ready for please move them out to future release +Verified in. +Thanks for merging my fixes Owen any update on uploading a patch here +Interesting +Any comments or suggestions are much appreciated +The algo returns a collection of list of vertices and its possible find the connected component for a specific set of verticesI am available for comments and tipsciao even if related to the Boruvkas algorithm thats the wrong locationPI ll fill a new issue just to track the type of patch we are applying +Please note that mentioned may be masked by Resource NotFound page but the logs should reveal the truth +But if you can think of a case where the algorithms fails please let me know + +version of patch attached this changes method copying for from to use the +Therefore instead of saying what to exclude is it would be better to say what to include is +Looks fine to me +Despite this both the Java and Groovy classes do compile correctly and run correctly +Resolving as wont fix apparently the controlstestperf directory is obsolete so I just deleted it in SVN +This issue seems to be a duplicate of I noticed that the logic in is a bit complexI think we need a clear documentation about the type checking and some code work +Depends which one you put it inSomething Im trying to spend time on helping to get organized as its still abit of a mess +Thank you +The patch. +Ill start on that today +The fix is according to what Flavio suggested in with the right have you looked at? It seems to be a duplicate of this issue +Will be in if I can do it today otherwise +It looks as if the M change might be that M changed maven so that the setting in the is honored +The WSRPJ has been terminated because of lack of interest. +Applied with minor changes to svn mv the file rather than delete and it +Currently there is no difference indeed +perso I would prefer Map of parameters but I didnt have time to work on that. +I just got some time and reworked the patch by using tmpDir +Failures occur in some of the unit tests while other unit tests which construct and pass reference parameters in exactly the same way as the failing tests seem to succeed +Shouldnt this wait until we have in place +Yeah Id say this is even more important for coverages as they can be even bigger than rasters and this option makes it so we dont force people to move the locations of their data +will create seperate issue closing so M can be marked as released. +Will try to create an independent test to recreate +Verified in reopening to update release note info +Its not a blocker for +should export the right types. +I also think its a bit dangerous to do without admin involvement +Updated for latest trunk +Right I tried and +If you dont want the caching then just use +I ran into this one trying to get the default planet working +Committed trunk revision Committed branch revision Committed branch revision Uwe for adding the messages and committing. +Allen committed a fix for this one +Ok tabs removed +Thanks guys for working together on this! I am going to commit the combined patch which includes the docs +I did some quick tests on the prototype I am developing and it seems to work for me +The issue can still be seen if the is modified for example to use datanucleus and GAE up front +Committed revision . +Leaving that for later +Thanks +Andreaplease attach the patch in diff format so it can be applied easilyAnd remember to check the box about confirming the use of your patch under the ASL +The objective value should at least be but it is around +This is likely the same as which is resolved in. +Ill commit later today +Also extended a bit to use dedicated builder to build results so you can define your own builder without messing with the whole great thanks +In Tuscany if there are multiple under the same namespace we return a facade WSDLJ Definition that use wsdlimport to import the physical WSDL definitions +I brought it up on this issue because i dont want us committing anything to the ant buildfor this issue if we dont regularly test it in jenkinsSo that means removing javascript from our jenkins antlib or whatever it is which wouldhave the side effect of reducing test coverage for that solr test +Right now its more complicated because of the other bug +It surely differs for workloads but in our case got much worse along the CMS dimension +There will eventually be an async appender in lognet +Yuanke this question would be better asked on +Ok could you then repurpose this JIRA as you suggested and do this for and? +Patch breaks the new functional test +I have done this work and I have also added a min option defaulting to or five minutes +Unfortunately no Seam version was empty before for such kind projects +I prefer having a coherent API where all the variants accept the same parameters. +Solved in in jbt build. +Changes can be made to fix the issue +Not sure if youre able to get it running or not but Pig will run on OS X if you setexport JAVAHOMESystemLibraryFrameworksVersionsHomeFor previous versions of OS X you have to upgrade to but for Pig at least do this and it will just work +Reviewed and tested looks good +I will go for option It looks like it adds the test classesBut neither resources so that should be addedI actually doubt that we should add anything from the test? Is an option needed then afterall +Devloped by Sumit Pandit And in second aproche we made seperate screen for SFADeveloped By Awdesh PariharGuided By Santosh MalviyaKindly suggest us which approach is good and efficient so that we can do our further work in right direction +We will have separate Jira to track the javadoc jar +Are the failures because of this patch +batch transition resolvedwontfix to closedwontfix +The method calls look identical yet localRepository gets configured correctly and ploymentRepository causes an exception. +Issue to change Release Notes Docs Status field +Thanks Alexei the patch is fine +Against variable this synchs and syncLock +Taking a crack at this +Thanks for reportingYes there was in fact a glitch in the project setup during the release +it skips +Code has been tested as before +Ive checked that the producer properly passes the using as expected so it does really seem to me that the problem lies on the side +Updated patch +Can you duplicate it in a dev environment? Can you turn on DEBUG logging for the or perhaps the whole +Is this relevant yetI have tried running randomwriter and sort on a cluster +classpathTransmitting file data +tried latest CVS just now +srcconf +Can we reopen this issue please it is not fixed in CP thought was built on top of and therefore include the fix but it doesnt +I was very tempted to apply it but it makes slightly too big of a change for me to be completely comfortable applying it now +I dont see how these things relate +I have a couple of comments on this +Anyway the important part is that becomes an interface so that you can plug in whatever JDBC datasource you need +I admit that it was not a normal use case but potential tight loops without sleep seem like a bad idea IMO +Mice can you check on this +Changes in the attached patch has been applied to the trunk via JIRA +I apologize for not finishing my thought on the Subject Associated Products are assumed it was meant to say Associated Products are assumed to be in the same Category +Can someone explain how come a customer issue that I reported back in Octdidnt make it in time for EAP CPEAP CP? months of development was not enough? Do you really think this task is the only thing weve had to do in the last months +Doc row attribute was should use one of converters or implement their own +Since the replacement order has no payment preferences associated to it payment preferences of the original order are used to process the refund +pretty trivial goodWhy did you choose to return sometimes but sometimes? Id rather see for errors consistently and for success +However I see this has been marked as fixed so I will investigate further +I will check the log of the most recent failure and try to figure out the cause of the NPE +snapshot support this issue is caused by add symlink support to HDFS and but not handle consistency well +Also it turns the constant strings into static constants +Ignore what I said two comments above +v Renamed to Endpoint and rebased against trunk again +Changed defult collection table +Your latest patch looks good but you forgot to click the box that gives license rights to the ASF +I think this is a duplicate of this issue and are resolved as duplicate without any fix version +Because of alignment and issues these documents become unreadable in +The jar included the jar which had a class that conflicted with class in the jar +this looks reasonable to me +Or did I get this wrong +Hi Sergeyyou did not include the check for null +Heres a version that handles repeated fields and that caches some expensive operations +So the fix here is a documentation fix we should describe these parameters and recommend increasing them for large clusters +Its not just a blocker for a release its a blocker for contribution +on new issues for the two probs above +This issue got closed as part of the commit for . +Committed revision . +two out of each three times when I run osgiupdate on the bundle I get the above exception on the +unit test patch applied at revision r and I fixed the implementation by check the directly seems all the failed tests you mentioned now pass for me please verify the problem is fully fixed as you expected +Or is this defaultFactory just a for Resteasy and not another application +I dont know how that affects JBM +Im not sure what tools our currentwiki technology has for describing graphics +Ref guide was updated according to comments of Max Rydahl Andersen. +Gershwinou I wrote the Security chapter for Spring Roo in Action dickens and have written steps similar to your comments above +Thanks Jarek +I think this has gone fairly stale +OK Ive added a check for the missing as you suggest +Ok +However in AM such RM terminations mark the task attempt as FAILED but in this case we need to mark them as KILLED +Fix regression that caused CSS files hash to change between server restarts by using stable hash codes for the enums +in rev +For the record I am a little surprised that it is deemed ok that a null pointer exception is leaking through to the end user user interface +I cannot reproduce message loss howeverCan you describe your setup in some detail or submit a test caseOne other thing have you verified this with trunk +this is an updated version of the after the commit fixes an error in the previous patch and contains the fix for ditched the property specifing the directory name src +defer all issues to +attaching again the tests I forgot to add the commit method +Really just curious but why would a linked list be slower? The only drawback I could see is the count of child nodes +Depending on which class is loaded first something will break either talking to the HS or AM +isnt included in the patch but I can guess +Closing Jira. +Though this patch was not used a similar scheme was implemented. +I had a look at the last week and there are a couple of issues with it +I change it to Major +Closing tickets +This one requires some pondering to make sure we can achieve something deterministic +Hi Govind +Those who want to use another value can use writeByte with their own value instead +I thought Ales had the locking issue resolved +RC compiled fine tests changes will be made in. +The ticket has not a major priority +Mark I am fine with and but seems clearly wrong to me +Also I was looking at the central sync requirements +indeed +I think this further decreases the reliability of the product +Removing all those unused methods +catch around the cache puts and cache deletes as well for completeness +Remaining captured in subtasks +We know that the way the query is parsed you cannot really ever do phrases inside phrases +assume that the total maps and total reduces you are talking about is on the main page of the history server and not on the job specific page +Or did you really want the actual value +patch for this applied with thanks! +Could be done by whoever adds all the other args such as I assume that services dont do any DUCC logging but presumably agents ws do +which version of hadoop did u see this problem on +Categorizing bugs related primarily to stylesheet elements etc +Hi VenkateshSend an to to subscribe to the development mailing list +Thanks David John! +Wrong version of imagemosaic was being used with latest release of geoserver +As for the duplicate atomLinks I know that doubly registering a resouce causes an exception only in the case of subresource locators +If any more specific changes are required a new JIRA should be created to detail +This patch is just an initial draft to describe and discuss the snapshot operationsLars Yes the patch should be just the ruby files for the shellTheres also the protobuf since you need to be able to run it but once weve all this code will not be hereJesse yes restore code will not be here and I agree to add restore to the table classWhat do you think about having mount and restore as separate feature? From a code point of view are almost the same but for the user is a different operationuse case +How would you expect it to work +But set appType to null so that does the replacement with default name no need for duplicating that code hereHitesh can you please file a separate ticket for your proposal? Tx +Enjoy +Fixed for upcoming release. +Among our many customers at Cloudera its far more common to target the execution engine with a higher level language rather than the Java API +Where can I download this jar +Quick question Simone have we set this up correctly? We have been unable to find any hint of what is wrong in the is present +I think this has nothing to do with IIS or Windows +Ill commit it soon unless someone objects +and because you think this isnt an error I digged deeper and deeper +Yeah you should use wildcards eg activemqtopicintegrations +This might need to be revisited to maintain sanity Simon tells me that GML in in development. +At the point where intersect is noted there is a missing right may be a way to improve the error message here but Im not sure +Added patch to skip known variables rather than attempting to load them as a class +Hmmm it works on my Windows boxI will ask Babak to have a look on it +If not then Derby should throw an exception +Ill apply it +Now theres quite a bit of duplicated logicPS +The patch for the property file is not correct unfortunately as it will break on windows boxes +Normally we would have removed it in it is really only be there to support old indexes so no new features +Thats a good idea +On the point of duplicating Collections IMO a small duplication is better than introducing a dependency on Collections +Why not just post the ips to the mailinglists and see if someone recognise them + file attached which is not compiling into a proper Java class with this name +Woah! Dont make me switch all that againSorry Im just tossing out ideas +I tried looking a little bit at the hadoop logs but I dont really know what Im looking for +I should also mention that I am using you test in testunit or testintegrationThey need to be integration tests for these methods to test is in testintegration +Unfortunately when I tried fixing it a different unit test failed +Call sites start off abstract and are optimized as they are encountered +It for sure has to do with the number of files in the directory that has been added multiple times in the you attach a simple Maven project that displays the problem? If so I can use that to turn into an integration test and make sure this doesnt regress +Heres a patch that passes tests against Hadoop +The current checkpoint code will either wait for current checkpoint to finish or just return if one is already in progress +Is there bugs which should be fixed in initial patch? Im asking because I need his feature but Im stuck with for a while +as soon as I got james running Ill ask andor investigate how much work this would be +In the worst case this is an Accumulo problem and is released it can be fixed in +Bulk move of issues into after vote that we merge branch into TRUNK up on list +Committed revision Thanks Martin +I think that will work I have one REST endpoint that is basic auth secured and one that is not and with you get https for free +quite sure if this will be completed this quarter I believe was helpful in handling a similar use case? +I need to kill the scp process manually to make continuum continue running the queued builds +I believe the functionality you seek is already there see +Point taken +If we fix it we should do it in the +Reopening as the token signature validation needs to be moved to the request handler. +Thanks Dan this is really a duplicate then +Item has been fixed with r in trunk and r inThanks to Tim for figuring out what was missing in the manifest files to get this to work properly +HiMichael The problem is a lot! +Defer to all open issues to think this is now fixed for everything except the boolean operators +Im not sure how good of an idea it is to include code not intended to be used by the open Apache community. +also fixed copyright dates in runtimeDEPENDENCIES and annotationDEPENDENCIES +The changes look good +pom and cargo plugin logs looks like this is a bug the configuration object is actually cached based on the configurations home value and even if you specify different options for start and stop only the first is usedAs a workaround could you create a separate profile for example run which you would use for manual testing and debugging +For example when store getsa serious problem while logging it raises a database shutdown level error and for safety basically nulls out the ability to do io to the log or the database +This failure again occurred on the internal build +An excellent advice! Thanks +Found more stuff we can hide +I believe the issue here is that byteaoutput is a tunable in so using just the version is likely to be insufficient to know whether hex or bytea encodings should be used +thats much better +Pull request applied +This is a known issue see +Is that the right commit +unary minusDonebq +When the milliseconds from is converted to a date object on the iphone it automatically sets the offset from the phones system setting +In general this looks greatThere are potential problems with the license attached to In my opinion this needs to be replaced however Im taking this up with the project community before doing so +Id prefer the latter because we can fixpatch it ourselves if we ever needed to do so +I HAVE NOT run the patch with security manager enabled +I dont see what is so bad about puttinga second line into your configuration +Thanks Matt you beat me to it +The cumulative patch substitute the previous patch Ive attached. +This did lead to poorly tested snapshots +class as well as +please null checking to all applicable committed. +Split out the bandwidth tracking +Will open new issue for this +After commenting out all of the footnotes in the document the processing of the file runs to completion +Another benchmark + +thnx +You already found the workaround. +and makes sense to me +Because most of the strings are internationalized now well need to coordinate with the various committers +This is an issue with a specific version of Tomcat and a workaround has been proposed +Robert are you proposing not having the Cassandra service automatically start even on boot? That seems like it would run counter to the interests of most users +Also is there any performance difference between throwing an exception vs returning a status code +Patch that compiles based on just committed this to +Small tweaks to the number of mappers and the size of data produced +Yes Maven Update project did the trick. + +This patch is basically a straight refactor +This patch depends onI will link it +Cool I look forward to seeing this new sample app +AFAIK this is what the current Jackrabbit implementation does +Verified by Marek Baluch on build. +I have done this several times with the same result +How do this fix work with that? The maven installation will be ignored +you mean checksum chunk size right? not blocksize +I can confirm that removing the annotation work w as wellI should also mention that the reason I am having so much trouble with this is because I am running my webapp under +Yeah we would need to check in if you have configured Discard and +I found that the message generating code was only in in the authentication +has been added with description Test case showing issues with is the status of this bug? can it be moved to resolved +However should point out that the doxygen build of pdfs from docbook xml source is unsatisfactory in many cases. +I attempted to fix this in CVS +Is this integrated in release? +submitting +Seam uses this scope to manage his extended persistenceContext which is one of the most convenient features +This issue has been resolved in +Sound right +Is this still the case with +Ok Ive got it to work by adding a pause when stopping the container +Code has been fixed but additional tests must be written +note having links into svn trunk is not something that should be used in docs +The is not tested +Normally the user can bump the heap size of the client to work around this issue +Ive attach the source you canintegrate it into Collections if youre interested +I think that number of retry times needs to be configurable and be set to like times by default what do you think? So on insertread fail we will retry current operation X times and then continue with next or should we exit and report to user after those X retries failed +Perhaps a is the correct way to go +someone should investigate the scope of this dependency +Here is a patch + +There are currently no plans to proceed with this +Thanks! Works with the latest build from svn trunk. +Also marking as resolved if there still are issues please either reopen or create new issues. +Committed to +Bulk move of M to move from M to smooks editor is no longer being developerd so moving to LATER as agreed with +Attaching patch a which removes the HTML Book entries from the Derby documentation HTML pageI think this is ready for commit but whats the conclusion on the appropriate action to takeGiven that we have in addition to the HTML pages version shall we err on the side of caution and remove the HTML Book version +This patch is based off code from Kelvins which in turn is from patchThe patch introduces an optional reconciler setting that allows the user to control how reconciliation between two conflicting columns is done +since this depends on the EMF upgrade I assume this is trunk only? Definitely for the trunk but I would have some reservations about the stable branch at the moment +Launching succeeds but killing it fails +Thanks +yup +Stack I dont think this was committed properly +there must be a different changed to not swallow the and to properly restore the system it seems that with the current approach it is not possible to test the consumer because of compiler class path issues +It is most important to keep the export versions the same +applied +The Stateless is needed for things such as getting an instance of the etc +This is not a bug +Yah make a new module and well get the new module working or include by default in the short term as a work around +Well in this case please let us know +Hi Milind in the use case you mentioned heshe can write his own put the jar in the import list + push out to +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +All available options can be seen on the Settings section of Sonar under the menu PHP. +You need latest you need you run the ear deployment against trunk +Vasyl Im afraid this combination is not actually supported at this point +I only wonder how that is supposed to work + +committed to +Since this is a separate Geronimo it will need to go through its own release and voting process before it could could get included in Geronimo which would most likely end up slipping the Geronimo release +The section Prepare and start additional management servers still refers to +Patch to resolve the issue +Either way I will debug into it +Could you comment on how you fixed this? Appending something? +For simplicity it should apply to all instances of the service started by the deploy script so could use the pid of the JVM and a special signal but may not be able to rely on the shutdown hook since the logs may have been closed +To disable advertise you need to set advertisefalse +It works with my patched version +Brian Its weird +Paul Cooper is the champion of these projections among the polar communities and is responsible for getting them into the official EPSG database +There is only one path to a folder +The relayout of import statements in is inadvertentThe was a that I felt was an improvement so that I didnt have to pass in an empty params list to s constructor +Threads usually have their host as prefix check out other places in server where we name threadsWill do +Bumped to major affects all stores +Patch to generate the reference list in the correct order +Makes my head hurt as the code becomes even more complicated +srcjavaorgapachecommonslangmath +Submit a simple runnable test case so I can reproduce +Why does it need a separate thread +What are your ideas +I have decided not to upgrade Toy Store to +Attaching a new version of the release note +Xavier Ive just run my swtbot tests related to this JIRA and they passed +Thanks Amar! +We dont really have a way to find out if theres an accessible ldap server on the network +Thanks Philip! +OK Ive applied the patch I suggested and renamed this ticket to reflect that change +Theres virtually no impact on memory consumption hfile sizecomplexity etc +I will do today and give feedback +this patch looks fine to me however one nit to fix theres no longer a need to check osMbean! null in the conditionals given youre instantiating the object on the previous line +At least for Hadoop and I think we need to be able to find the public hostname +I am ok with postponing any jar coalescing until later +We should use exactly the same metrics +Are we good to go then +It also uses the workaround discussed previously instead of using specialized +There can be possible fixes I can think of + to and trunkThanks for the patch AnoopThanks for the review Stack. +reopening the issue as I think it may not be a duplicate +The process is killed while doing findbugs analysis on one of the modules + has some work and more discussion than this jira +Manually tested +What I think is needed in coercion code examines the existing property to determine the scale and rounding and apply these to any created valuesregards Malcolm Edgar +Actually it looks like all methods belonging to Object will not be intercepted by invokeMethod which may be a design constraint not to mess with the basic class but definitely warrants another look when the MOP is revised +for the Mean time I have attached a patch which I hope you can review +Another question do we even need a? Cant wehave only the Builder and that returns a +Thats why I split it up in arguments +Patch not committed in time for +Thanks Dag +I made a env mistake running it it +Bumping to release since is being released +Rahul since we cant run this patch itself through hudson can you please make sure that is functional when used with projects like and then I can commit it to unblock +Hi MikeI see that you checked into the trunk at subversion revision +What were you thinking a long with the number of milliseconds since UTC? If you need more precision than that you are most likely going to make your own type +Anyone tested the patch successfully? Any feedback on this? Any better idea to fix this +A clever way to solve the problem +We shouldnt be encoding for Servlets when we explicitly know that we dont want it and then ripping out the results +I anticipate this moving to PoolDetection of using pooled objects after they have been returned to the pool is not currently supported in Pool and isnt on the +Change has been committed +Committed to trunk and thanks Stack +This issue is fixed +Can I go ahead and start developing my first Wookie widget or it will have some major roadblocks later +node where all is with server side just look like client disconnected unexpectedly exceptions +I stand corrected +Will somebody try the patch +We can make the later configurable or can mandate operators to kill containers explicitly in that caseClosing this as a duplicate. +This is from a consultant colleauge I met at a customer in Summer in NYC +Does that make sense +Perhaps a volatile fieldBut I agree the easiest is to avoid all these problems by leaving logging to the calling codeIn general it should not be necessary to log in low level libraries. +IMO this change should be applied in RF not in +I wonder where fits there thoughPersonally if it was all only developed by committers in a public repo I dont have a problem with being practical FWIW though Im just one guy INAL +Attached the changes +Unlikely to fix this for the issue is targeted for but has not been updated in +is mentioned twiceFeatures are missing +Here is the test case patch for service setRegardsSurya Surya for the patch I have tested it and it is working fineSumit with mods in r +If both were part of the same group a duplicate ticketgroup entry would be made +TrueStill adding new CLI commands is out of scope here +Is there a way to find out exactly which document chokes on? Ill open another ticket soon if the errors come back in this latest crawl +Change to bypass target installs directory when fixing line endingsAdded trace feature +Maybe we could try something with the Builder pattern and some runtime validations but it sounds like a best effort solution only +Nothing to verify here. +great thanks Joel +It could be +Could this be solved by implementing the writeReplace method to write the collections class as a standard java collection? When the collection gets to the client its going to be disconnected anyhow so the extra hibernate data is not required +Attached original version +Fixed it. +Thanks for the catch +In master this test requires this definition directly in XML descriptors +I am getting the same error +Verified in beta +Committed to trunk and +Andrea more from the new reporterThe log window appears to be totally quiet when I do this actionCheersOllieOriginal MessageFrom Ben Sent Monday January PMOllie did you see anything in the server logs when this happenned +As both patches touch the same code base shoudl be committed after +It seems that the roller user needs to be defined in in such a way that it doesnt matter what host the connection seems to originate from +It just increments the count if an unseen group has been detected +ok +since remoting has been archived off and no longer supported this ticket doesnt applyInstead rich viewers should probably be based on the restfulobjects API +Ive just committed this to the trunk and branch +Can you confirm and if so please resolve this JIRA +Thank you Haohui. +Not committed +Im with Arno +Without the fix the test fails with an trying to load an empty image file in the checkpointerI also took the liberty of reworking the fault injection we use for these tests +But I think if you use a correct way to create jboss bottomup ws you should have no if you may send a zip of your projects to me I will take a look for the error +That would help us to not change stuff for every similar use case +bulk move of all unresolved issues from to +Done Freddy +bq +need to modify at least one review and commit if youre happyThe broker release now includes BDB if optionaltrue sys property is set +Thanks for looking into it. +Ive even tested styling using layers of the same dataset. +outside help is always welcome +A number of files in the samples directory also have this form of clean targetThis can cause surprising results if the target directory has subdirectories then only one level of subdirectories will be deleted each time ant clean is run +I will provide a fix later this week +Since this ticket is marked resolved I filed to point out some closely related problems +The prepare and prepare events should still take place first before the canceled event +I will submit a new patch in the next few days which uses your proposal for automatically switching to the new api and avoids adding pipes +random placement where each data block can be placed at random on any machine and +See the file +Im assuming those are intentional +Testing the attached patches now +For now you just get whatever a pure Java environment can provide +Commit to fixes reading metadata from this image too and unlike your patch it can be disabled with strict mode so I am going to resolve this. +Forgive me for the perhaps stupid question since Im not a native English speakerI do not understand the comment of Dan GravellPerhaps I should add more infoWhat i did was java jar options the property file contains the desired informationWith version this works fine +Lets fix hadoop failures doing as Andrew suggests by making adding in hadoop and hadoop compat modules. +All ejb testsuite regressions should be blockers +Roller doneContinuumArchiva requesting a move from +Fixed in revision complained +I would rather be defensive in that sensitive case and make sure that clear type password isnt logged inadvertently +is out the door adjusting on all remaining unresolved issues to by default +It may be a duplicate +Applied the patch +What I wonder about is how did this ever work with parallel compactions which are possible +Verified in released beta build. +Heres a patch for queuestopics Ill create a new issue to cover aggregated metrics for durable applied thanks. +I think the issue I had was with the Customer and Customer relationships +Thanks Isaac +Well its not case as I checked in sources +I updated the Envers module to have a dependency on and that addressed the CNFE +It seems to be quite informative report so I would think about turning it on in core or super project as wellWell probably limit the amount of reports that get generated for the public Jackrabbit web site but enable quite a few to be generated as a part of the nightly builds Im going to set up +Closed upon release of Hadoop. +Jons implementation of custom per schema tuple is better well go with that instead. +use the previous result as the next seed +Ive attached two patches for this issue +I wonder if NIO classes could load the native library with as all good behaving Java code should do +Patch to make s Iterator conform with the Iterator applied thanks Kre. +It cannot be run standalone since the Main class needs to be passed in +hbase has schema right? it would be nice to fill it in +Make senseIm with patch v +Using Avro here could simplify maintenance but might add a significant amount of work +the issue only happens with omitTF my guess is that its returning something strange for total +Proposed interfaces and stuctures for crypto abstraction for Hadoop Core and Map Reduce layer +but the doc itself completed. +With path specified and patches from and from this JIRA applied eclipse starts but do not find Harmony JREBut I think this is different issue which is not related to starting with standard launcher in any wayMission accomplished +instead of +Very slight tweak to the release note +ThanksManpreet +I have a version ofthis running in so I can provide that if necessary + Test Case is added for the to update on. +Assigning to Shaz for further investigation +I think this was related with bug so this can be closed or reported as duplicate +Committed revision to trunkI will test this against and upload a patchThanks for your help Uwe +That said do you have a case where it happens. +Christopher please reassign yourself if you would like to continue working on this issue +Patch to fix this issue +It reads the language from the userLogin and if the last locale field isnt empty the language file will be loadedto add this feature +Yes I unfortunately forgot to rename the directory +Updated patch to revision from +Attached patch for checkingalso cleaned to tab charactersPlease some one take a look at this patchIts a simple patch shouldnt take much time +Hi UweThats it +Need to study the test case design +The opposite +The refinement specification in the grammar is for specifying a restricted set of objects +This is a native x all issues to +Yeah I would say that whilst its nice to have the granularity to have mappings behaviour differently from the general settings this would be the exception rather than the rule +That do being said I do like Billies suggestion +According to some reports it is supposed to be x faster +Many thanks As I was doing the conversion I noted that they are syntactically only slight variations of one another +Added some log output +Hide the in trunk +Fool that I am! I just understood that enableassertions is a property of java not javac +failures unrelated +Move to Jackrabbits Webdav API should improve the situation on testing +Ideally the code could print deprecated warnings but those are not available to runtime + +All unit tests were passed except for and +Already fixed in via commit aebeacc +One concern with adding a contstructor without correlation id is that most users will use it which will hide the correlation id information +So it will need to be implemented +Ill work on it during the week and will keep you updated on it at the end of the week +There is no extra stuff +Bill I see one problem with your approach +Otherwise Ill do it +Am working on porting this patch to trunk +I think we just need to focus on things that are always inlined currently and are possibly causing us size andor performance +Retries are now subject to error handling and thresholds defined in the deployment descriptor for each delegateReduced Spring recovery delay from ms to ms +You can select this option when upload attachmentThanks a new patch including both annotation based and spring config based soap handler osgi tests +nd paragraph in list of features a mention of Farming would be good +It has since been fixed +Patch to resolve issue +resolving as all subtask are now as is already released months. +See the itext example +Jody? Opinion +so I resolve the general to check this workable Open Mapping File should open mapping file for any configuration type Core Annotation or JPA. +ChristianYes I did this changes with the help of eclipse it finds the reference in Java file and the scala file automatically except the partAnd all the camel projects are opened in my eclipse workspace +Thanks +So multiple characters can be consumed by the reader even though the receives a single character +This will reduce potential double work meta migration +Removes set of split and region flush on compact call +Overall the changes look good to meA very very minor commentLine import got removed from and got added to +Closing this issue +java client now also disallows in application names +Run testdone scriptThats all +Imagine a company using spring and hibernate wanting to upgrade to rc +I think with this may be the may passPatch looks good +I found that by setting ldap to one of the available field in also allows to workaround this issue +Thanks yes you are right thank you Simon +last one had svn rmd for some reason +Updated the patch because the previous version contained a copynpaste error causing jvm memory info to always be hidden +let me know when the apps are up then Ill do the proxy stuff +Given that Im still in favor of just requiring the admin keep the excluded hosts files in sync and call refreshNodes on both from +We also use it in subqueriesI second Aris opinion about fixing the bug and not disabling the feaure + should fix this issue transparently when is integrated into the the latest version of the AS Version is not shown at all +hasnt failed since above fixthanks for reporting back the failure you reported was due to a reset call at the wrong position +Upgrading to Maven first might still be a nice stepping stone to getting to Maven +I just committed this to trunk and +Stefan as you remember we had a discussion on modifying the fetcher and specifically changing the thread mechanism into a +Now that has been resolved Ill proceed with adding the needed factories for the Tokenizers in Solr +The implementations would not be in error rather the schemas would be +If you compiled the test with the earliest installed +All observed problems are fixed +How true is that? The default scheduler and the fair scheduler dont support queues +Thanks Amila it works fine now. +I agree with your comment about CROSS JOIN and I hope that we can get that resolved by fixingCommitted revision +for a fix in +Marking as a blocker for targeting a patch for a given release please list that release in the Target Version field +tried with my own mosaic of tiles and it worked fine. +all things where portability is a concern but java jar +There isnt any need for this to wrap the in another exception +the servlet api should be set to provided no matter what app server or web containerThe jboss problems are due to other jars that shouldnt be included on jboss and should be included on tomcat. +looks good +For refactoring once the module and other bits are ready we can move select tests from Bigtop to +Thanks a lot Dierk for your work! +Ideally they would be separate from the other jarsIm going to apply your patch when I can get a chance and Im sure I will have more questions +Therefore Ive updated the docs to that effect and have changed this ticket to be a feature request rather than a bug +bulk close of all resolved issues. +Thanks Parimal Ankit for testing and Arpit Singh Pandya for commentParimal your slightly modified patch is in trunk r r r rI simply added in label that no productId entered would return all the orders also in French +Shouldnt we wait until Lift version is released +There are some duplicate constraint ids in your fixed them and migrated to the current m schema +Marking for move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Note that code for the latter was committed as part of the former +I think this patch takes care of it I added both GML and JSON +With the version from SVN this is no longer reproducible +Alternative C Allow to plug in variable change listeners which can iterate through the trailing chain only if needed +Please could you verify that you are happy with the and that only the previously discussed deletions have been made +The only contract is that when a feature id is published by the server it is persistent so that the clientc an reliably make queries with it +Logged In YES useridI having a tough time coming up with a test case +Well that did not fix it +fixed in with latest buildAfter upgrading from we are able to stop and start the Vms that were deployed in. +Im ok with the heuristic of only assigning local tasks after the first of each kind +Thats what Ive seen in testing so far +Ill have a look tomorrow +At first glance this might seem to be a duplication of the existing templates +current trunk is too different for me to use at this time +Where should this be added? Im thinking that adding it to the Task interface would be the best place +Changing to fix for since the new tree is very usable now and provides lots of new useful features +does not work on ie in all browsers the Get Flash Player icon still appears even if flash is already installed +I cannot reproduce this issue + +This implementation adds the conversion logic inside because this allows us to totally behave as Strings everywhereSomething Im not clear on is how to make this conversion optional +Fixed in branch +Verified by Oleg. +Just so you know +Currently it is impossible to create a Path with a slash in a name although colons are permitted +yes it looks like having module configured in fixes the issue without configurator +If I can work on the method Ill try to refactor and minify the code using lambdas + Use SAX methods to output XML rather than printf equivalents do not the parsing code +Proposed patch added +Attaching a patch to refactor the redundancies discussed rerunning regressions and my runs +The same isue in FF and Opera reproducible if the newest jars are used. +Changed version to CP fix will be verified when a new version of JON is available +I had a problem whereby I couldnt actually change the C client runtime library version in my deployment to incorporate a fix to so I wrote the attached which patches the Address object by iterating through until it reaches the bindings and explicitly sets their encodingThis is actually more of a faff than the changes needed to but at that time it was easier to fix the user code than the library code +Hi KevinThe max setting has to match on both sides of the connection +Jenscould u plz help with a patchthx check latest svnthxdims +Any objections from anybody +Would a sample implementation help? If so give me a pointer +Ive tried to look into it but i didnt find an easy way to figure the initial constructor arguments out when get is being calledalso isnt even algebraic +I am attaching the patch for trunk which is trying to backport changes made to to backout changes +A simple test in my mind is for larss point if you write instructions on how to upgrade in the future assuming you have the same configuration files which would be shorter? Ideally wed choose as the default the approach that ends up being easier +defer all issues to +Sample project to demonstrate the problemRun mvn and open httplocalhostIn the console all logs should go through SLFJ. +What JDK are you using btw? If you are running anything belowJDK do you mind trying to reproduce the problem using the newest JDK from SunOlegI am runningI dont mind running with a nighty build for a while but I hope you can find the time to put it in the final releaseThanks for the quick feedbackAnd again if you need help or extra info please do contact me +is intended to be a API and is not really suitable for performant work on the +Jasonthis might be related to setting the as default bundle after the actual configuration is loadedSo all custom are added to the front of the list during initilization and then the come in as the last additionA possible approach to solve this might be to add the properties within the as the first message bundleAfter this all custom properties will be added as expectedCan you please take a look at this +Applied with revision +METAThis would make each test behave deterministically +hopefully will be part of is the ant should be mostly technical stuff and a Setup should be mostly user stuffI am not sure what you mean mixing Ant and SetupMy goal is to move way from loading ext data for setting up a Business as a normbut have interactive screens that this data is enteredI am thinking in terms of a type operation where once the base system is setup someone would purchase from the Host a package in the ecommerce that would setup their db and seed data on the fly +So it sounds as if is attempting to process the SCDL entry for in the context of a classloader which does not have access to since I believe Tomcat segregrates its implementation classes from application classloadersI havent been following the changes where jars have been moved to common in Tomcat but if the above is correct this may be the culprit +the option proposed in adding JAVAOPTSJAVAOPTS to solves the issue +Extended the monitor to enable to plug in customer listeners +I also dont like that it does not take a jar and pull classpath from its manifest which means wed have to duplicate the classpath arguments or write a custom plugin to generate its config from the + +With a this does not really seem possible since we have no way to know when a soft reference is terminated +In fact if it is okay with you I would simply point people to your project for YMSG support +reopening since if jslint was removed this is a rejected not done +If you can find a way to do this declaratively without changing the semantics of please reopen +Thanks for the review Daryn and providing the backstoryanalysis +The code was missing an end section and close the sink +I will upload a patch as well +And it is really a problem which should be solved because it takes lots of time for downloading +Sounds tight coupling you can now control the includesexcludes per overlay +Thats what users always wanted really to restrict this job to run say only one task per node +Thank you Chris. +I dont think objectsList is never with threads making actionCompleted method synchronized doesnt help when other threads are simultaneously updating the database in run method. + Cleaned up error messages +It may be recognized when the main class is created +There hasnt been much demand for so I think we should close that as WONTFIX and apply Emmanuels method above +Fixed. +Henrik you should now have the necessary karma to edit the wiki +After removing and clearing out +But what are chancesNevertheless Im ok with your patch +Thanks +Ill provide a when time permits +So either improve the current ones or trash them all together +I fixed that in this patch too by simple confirming that TOKENS is actually present for the host in handle and if not just handle things the old way since its a bootstrap +This patch fixes tests in +No worries Pritam +fixed a long time ago + to trunk. +Me Stack Ryan and JD all took a single car across that really safe bay bridge last nightNo more talkin smack once you take another ride +Thanks for pointing that out +You are welcomeIf you like you can subscribe to the dev in order to vote for future releasesI will open a vote session probably tonight in order to release version +If we fix this one the other one will be fixed to was already done with latest work with server entry +Would you mind your internal review to the ASF JIRA? This will help others understand the code and will make it easier to merge this branch into trunk when the time comes I think +Cool the effect is boost on the microbenchmarks on both platforms! My respectApplied at r default is ON on both platforms as agreed +We should check if this node is valid or not before trying to unRegister in NM unregister and hence the SHUTDOWN event can happen at UNHEALTHY state too +The problem of course is that I need the temps for proper app functionalityI also tested this scenario with the current and can confirm that it does indeed resolve the issue. + is the workaround for this to CP for now but it may get done in time for GA. +Remaining issue is to port fix to Hibernate code base +Thanks Christian +Moved to next milestone release +Rats! I meant to assign this to myself and to Farid +So if minthreshold is and maxcredits then a receiver sends credits whenever a senders credits fall below Wheres your config file? I didnt see it in the Hudson job artifacts +same trouble here we need to rollback to and even then we are a bit screwed since it does not support shared assembly correctly with maven +Rob +Nice improvement! +patch committed to trunk Committed revision issue has been resolved for over a year with no further movement +sar also a build target +Im resolving itmeyerdp EAP is not available yet and wont be for some time + +Revamped the original description to focus on the more significant problem +Thanks +Verified by Paulex. +Cluster will help +Can you keep me informed of exactly when we should stop using CVS and then when we can resume using SVN +Committed r +Closing old resolved issues +This was fixed long ago. +Add replication factor for Directory need correct the is there any chance to commit this soon +Thanks. +Works well for the action plans but not for the review comments which are not deleted +Patch for earlier version of hadoop not for commit +Is the system info handler accessed via the adminsystem URL? If so it doesnt seem to have the index sizeI would prefer to have it accessible via adminmbeansstatstrue so that I dont have to make more than one call per core to gather stats but you should do whatever makes the most sense for Solr not for me +moved out of main jar in rev . +bq +The hudson failure can be ignored +I was letting Eclipse get away with default behaviorIll regenerate the patch and will include additional detail on getting Eclipse set up correctly +I suspect this is due to the exception being raised during atexit which executes outside the normal flow +Its very Java Help like with a license that should work under Apache +all tests pass +Simple fix safe to apply +There are many more warnings than just these two +is the place to put some very basic bundles which can do for container if we remove jaas bundles from we also need remove almost all other karaf bundles from the which I think not doable think about without features related bundles ready at very early stage who take responsibility to launch the features fromIn summary the jaas related bundles isnt be optional its very fundamental thus must be available at very early stage +However the specific on VM bootstrap should work fine +Would it be reasonable to ask for a better diagnosisYup +Confirmed by compiling branchx that this fixes the bug I noticed in release +The vertex id however should never be null or nullwritable right? I think Giraph would crash in that case since it uses those ids internally +Uploading patch with change to fix thisThejas can you please check if this works on your mac machineThanks +TODO Move styles to CSS Manage multiple dialogues opened on the top of each other managementHow to achieve that without system manager Make sure that we disable selection of the children underneath the mouse mask div Managing focus within the popup implementation changes in to make it subclass from popup supports drag behavior on header by using bead +Ive committed the latest patch to trunk and branch +see how we have this in case as arquillian Im now a bit overloaded with my work so perhaps on the weekend +The thing is to make sure that the WSSE namespace is correctly added to the +Instead you have to set the property to emptyThis has been done for the Struts +The attached script works for me +OK greatI think there were no commits in these areas +Have they tried a machine with a fresh JBDS +closing issues for released versions +This issue is not a blocker for EAP moving to TBD EAP +As per Ryans comment during the conf call yesterday Im assigning this to him since Im not sure who to give it spoke to Javier about this issue yesterday and he agreed that its something you guys will want toneed to fix +Ive just committed this to the HA branch +I think I am being smarter about it now +Committed changes to and trunk +The only one who can relicense a work is the copyright owner and Apache doesnt own those copyrights in the components +The issue is that the Renewer code has no direct interaction with +If you just fix that before commit no need to post a new patch in my opinion +fyi has a plan to address gaps in +Simply clear the FILTERAPPLIED request attribute and acegi should work correctly with situations +Charlie if necessary I can make arrangement for you to access an environment that can reproduce this errorI will be up in my time zone for another or hours +forward files allow to be passed to shell scripts etc +Was just wondering if the error also occured within that specific node or in +Until thats done we should leave the jit max and threshold alone +Just tried it +This will be available in Spring +the failures I was seeing are gone +If apt generates any errors stop processing +Since is defined as Serializable and clearly is not Serializable I would say this constitutes a failure to implement the contract defined by +I have recently resolved some problems in this area can you try the latest nightly builds to confirm if this is still a problem. +patch against for for extensions for and +unassign it if not +Thats likely a ripple effect from localizing those strings +We add it to Adam which has Gump on it +Is there something in the near future? I need to make a decision on vs +Sounds great +It turns out Eriks problem is not caused by the merge but by the partitioning itself +They have a set of channels receiving JMS messages from an AMQ broker processing these and then forwarding them on to another set of queues +push to +Tried again today on the hopes that the problem would have been resolved but it is still there +Marking this as fixed since the original scope of this bug is no longer being discussedIf we think migration from is still important we can file another issueIf we want more fine grained customization of cloud backup and file api setMetadata is insufficient we can file another issue. +Resolving the issue. +Patch fails to set the initialized state in the corrected in fix +There are comments in the patch from Chiradeep on reviewboard +My guess is that there is some data type that is tripping up the code that introspects and builds a model for the GUI +Looks good Izaak +Looks good on my end. +patch looks good +the sharedsrcmainresourcesschemaouschemacnadsconfigouattributetypes file has been has been releaseClosing all related resolved issues. +do we at all need it +Resolving as Fixed because this already is in and trunk +if someone does that he will likely setup a full dev environment and get the sources or use a dependency management tool +bulk defer of issues to +Also DEFAULTYARNAPPLICATIONCLASSPATH shouldnt contain mapreduce jars +Can you give some more details about your i hope you tried it with the latest trunk +Its enabled by defaultSimilar to the use feature +Thanks +Those will be forthcoming soon +committed the patch +John saw this with Tomcat which uses the class to determine availability +Ok I have refixed on I didnt see the leak on the user said they did so I wanted you to confirm whether or not there is a you are absolutley sure there is no leak this can be closed. +Can you please verify and reopen if that still an issue +IMO listing things as important as Xmx twice is a bad idea regardless of the actual behavior +Thanks +Committed to trunk. +Fixed in SVN trunk +I have applied the patch to hdfs and tried to build with and without network connection +Is there any technical documentation available about implementing this? Are there any guidelines +svn revision contains a change based on your patch. +bq +for the latest patch save for the findbugs issueTrying to fix it myself +Thanks to Gert Vanthienen for this fixregards by using s renderChildren to encode the child components. +Yes Ram the locks are correctly used AFA I have seenIdeally should not throw IOE as it is clear from the thread dumb that the roller has not even near the point where it closes and reset the current writer +Do chgrp and chown support the option R? It seems that the option is not tested +I specifically called out for help earlier in this issues comments looking for someone to explain what it is that I need to changefix in order for that to work +Im going to have to dig further on this +I dont think that the package rename was a step to far +Bulk close for Solr +Currently we look too far ahead and tuples starts to accumulate faster in heap then we are draining +He said he will fix it +Since we already branched this feature will not go into branch but would instead be committed to trunk and released as part of release +Is that too heavyweight for what you are looking for Himanshu +If Weights have a searcher fieldthen this could be quite expensiveChuck +Thanks Myrna and Ole! +Can you provide some more informations where the sources are +The docs team can incorporate the changes into the product manual afterwards +This patch addresses all the concerns listed in the description +I already gave up with mvn sitedeploy and am using mvn sitestage something like this is the best to do in this case right this is going to be fixed some day +This means the slave instance can act as a consumer and master can push the notifications to slave +In light of that I added and +Let me do that and update the patch +Resolved by importing DFKI source code into the includes an and combination including services files so there should be no need to have those files in Marmotta anymore +So no +Resolving committed patch as svn If nobody argues that this should be backported I will close the issue in a bit. +Not a of a bulk update all resolutions changed to done please review history to original resolution type +Jonathan too slow is actually going to make it completely unworkable for many folks while the CL issues only remove certain classes of applications +Im not a Python programmer but will attempt to fake it this afternoon if no one else does first +hmm sure you attached the right patch? I dont see the arg in the latest upload +why does use +Havent checked the other issues in the comments yet +Queries for tables with a VOID type fail at execution time rather than compile timeSome potential solutions Treat a VOID type as always being NULL +I reformatted the codeI hope its OK nowThanks +The TCKs ejb module passed +The patch incorporates Nicholass new review comments +In the with JBDS. +fixed in +Can you create another JIRA for thatThanks! +OK +Committed but the deletes still appear to delete files not in SVN according to svn and +target does this fail under? bvt or +Committed +A true team effort thanks all! Committed to trunk and +Id rather keep the code clean +. +But that makes more worries Tony +I have solved my problem I have forgotten to set GWTHOME setting GWTHOME it seems to be I have problems with accessing grails service with GWT but I will work on the way what is the easiest way to update the plugin to? Currently I have installed. +I also tried parsing directly with Tika and everything went fine +Rice is it the same bug you were talking aboutDaniel could you take a look at these tests +My research indicates this is because many of the possibilities end up with the same score so this is not indicative of a new bug +Uploading a new version of the specification rev +Mike what JVM are you using +Issue in PBR as well. +Working on a launchpad based pom in a whiteboard area commit coming soon if I can get it to a stage where all the packages are resolved +Fixed in FUSE and will be available in next patch release. +Well Im not sure its the best thing in all cases not all are academic research papersId be for making this a param and allowing it to override the default value +It doesnt seem possible to run directly from the source directory anymore after the maven structure change +Calling the main method of the Client class gives an output ofThe The Other Messageindicating that only the method implemented in the subclass was intercepted +Hi CristianIm glad the issue will be fixed in +It was fixed in. +This proposed patch takes the path of least resistance +It needs to be merged with branch +Grant Can we close this now please? Or are you still blocked +Deferred updates are meant to untangle the problem of a single statement writing to the same data container that its reading +Eclipse didnt find but here is a patch fixing the issues it did findThanks +Works for me. +Yes it is a little redundant but thats okay better redundant than inconsistent +This could result in duplicate request mappings for +I dont think this is a weld bug +To give more options I fixed which should stop mandatory usage of URL classloader in Jasper +It still has all tests failing and marked as known issues but we can deal with that in followup issues starting with +Feel free to add your comments thereSomeone from the community with an interest in the Maven Repository needs to step up for these tasks to get done +Was in the wrong browser tab +I still dont get the point but never mindThanks anyway for your clarifications +Now I find it inIn general how do you plan to fix this? +just a matter of moving out the bodies +I thought a few days and did not find a good way to slove all problems +Thanks +dev branch +its much easier to write an import handler that can call resource methods directly than to jump through hoops with intermediary context classesultimately im considering some more extensive surgery to so that it doesnt use the chain stuff at all +I tried actually Hadoop will ignore files start with +If you dispute please with use case +I think we should fix it as its own bug it can go into when this may not +As a matter of fact Id make backups with timestampsYes timestamps the same format used by the snapshots N days old can be periodically deleted This is a good idea +Siying Have you attached the most recent version of this patch? I thought you were going to modify the IDL so that the force parameter is listed as optional and has a default value explicitly set +The file has to go under srctestresourcesorggeotoolsfactory in the metadata module +I just wanted to say that it may make sense to provide both steps and an ETA but just an ETA for the whole process and not for each step +Seems like too much was made per Tims comments +Yep Spring I must admit that done it before in struts but I cant see how this will ever work in as validationAware is persisted between requests and never cleared +Here we wait forgot a commented out bit +v I still have to convert the tests +Hope this is okay +I tend to agree with Hairong we should be very cautious about modifying the heartbeats +This is fixed with samindas commit Thanks the resolved issue. +My fix has been pushed to the Apache repo and I believe the iOS repo has been updated as well. +Its too bad the Java developers didnt think of this kind of pluggability for their hashes the way they did for their trees +However this is a bigger change as Camel end users then must change their code +However if it is not possible to do it with one command then we can still have two commands but have them use the same underpinnings. +The affect by was wrong sorry the version from title is the one where the issue occurs +Guys the patch is not really good +Committed to branch and trunk. +David After your comments I now understand +Also made the same changes to createSnapshotrenameSnapshot +I cant help but wonder if the reason none of this works is because the main codebase is now in DSE and its all modified to work with DSE and not with hadoop +Could you review +you dont want to close your wizard get to the console add the user and launch the wizard again +Well thats one define available in when NPN is available +thanks resolved issues for already released versions +ok fixedThe patch was not entirely correct since it produced an empty render or execute parameter the spec states that nothing should be sent down in case of a none so I modified the code accordingly +I think the most important is that there be a profile that includes all of the elements supported in jBPM BRMS runtime +It appears because of the problem in functionality +It was already fixed +files and directories blocks total +We do the same with and the generation of the META tagLet me work on adding that support to the patch +It working for me +If the user really wants to do this they can configure the output format to accept larger keysShould this be a fix +Thanks +Can you suggest a better way to do it? We can definitely change itbut just cant figure out how right nowI guess one other way is to configure that in the form +Could we consider make creating table as a transaction +But this can wait until we get them working +Yes its set to state of the snapshot is and it just stays this way +Child has identical but transient dependency +Product specific examples have begun to be added to the soa +create is changed and get has code removed that restricted which colorspaces it would returnIn addition lookupColor and set which are never called in are changed so they behave correctly with the same signature previously the meaning of lookupIndex and componentNumber was reversed so I changed their order in the signature so any external code that successfully uses them can continue to do so if this patch is applied +batch transition to closed remaining resolved bugs +jira issues that have not moved in a year +Twice weve encountered test hang took placewe always saw errors on balancers stats page +At this stage RAB is now the only extension of AnalyzerPatch includes a few unique changes Adds which is used by Analyzers which wrap other Analyzers +Updated patch addressing Jitendras comments +this error would most likely also happen if you import the plain project and just added jpahibernate to it +I might miss something +The default action on being synchronous previously while now being asynchronous +On the top of the file the errorAt the end the same data but packaged in a +Yes it can be simplified +It works! +Thanks Kai +Yeah I get your point will change it accordingly thanksFreeman +I wonder if it could be the issue with the rebase or with changes to trunk after the last patch or if k rows are needed to show improvementI get similar results nowNeed to add some logging to see if the new code paths are actually being run +andor +If trunk versions are no longer target for this jira it can be resolved +This patch is making registration of the adaptors async +Added the +Should we be doing this for all tests? Only has failed as a result of the new lower prefetch +Thanks for reply +Could somebody please confirm +ChunhuiCan you include test in your patch to show that the problem is fixed? test case in the is what I committed +Created into the GACP branch as revision . +I see this too when I follow the steps described above +Well see when we need this +Can be made deterministic +we should either remove the Rebooted Nodes column from the metrics since nodes will never be in the REBOOT state or add code to support the REBOOT example wed have to update the in to transition to the REBOOT state instead of the LOST state when the REBOOTING event is delivered +applied patch from issue and tested +Btw about the comment above then we need a and a modules too that separate the GUI classes from other stuff +As javadoc says is a readonly view of the job provided to tasks +So there are no leak problems involved here for case just a use case in a environment +Dmitry are you planning to resolve this quickly +Issue closed with release. +The latest from zanata is also merged +txt? This may be a convention Ive never heard of but the other examples all have +With the new mediator +It does seem more reasonable to me right now to let the local node distribute the updates +basanth Slices are definitely supported in CQL the query I gave you earlier involves a slice for instance +Should be an easy fix +In my opinion its too Look +This is a language design issue + changes Hdfs removes the use of deprecated +I somehow set the resolution to Unresolved +This is trivial change and doe not need a test +Not sure how this is possible +But why do you need to pass the headless argument? I run CC on linux and dont have any issues with it +Attaching a screenshot showing broken of issue does not occur after import for site acme +I was unable to get unsynchronized state for VisualSource as Vitaly said +Parameterizes the ending copyright date on the preamble notice +See the forum reference for possible workaround code +Sent to list dev and QA for review +Thanks for looking at itI can the statement without any pbm in TOADI have done some more testing this afternoon and I have the same kind of pbms with leading tabs etcI also find some interraction with logging +If any of the properties in the are set to false set optimized false +this all affects select statements +You are completly right +for thatCant blockSize be effectively set via D optionId also like to add a sampleSize option at least because it runs a looong time if youre running it against a table of any size +Yes please checkin that modifications +It does not take into consideration the plugin dependencies which I suppose should also be managed via Maven +This patch will add the necessary classes dependancies config changes for activity streams classes +Verified on built on +Patch looks fine +has fisheye +Set up the right indentation two spaces and try again +Attach mockup file for popup image for link egateinportaltrunkwebesrcmainwebappskinportalwebuicomponentviewbackground +Seems like we may have to disable this plugin +The link would provide a popup screen to select the language similar to the lookup screens used elsewhere +I have a test environment that runs a set of automated tests with our latest code every hour +Closing resolved issues for already released versions +Thanks for fixing this Fang +My opinion on this +Logged for LOG level change +Should the SDO classes generated from Tuscany tools have the code to initialize it? I assume we should have the code generated in such a way +committed to TRUNK +Ive updated this JIRA to more accurately reflect what this is about +i would try to those these simple api changes for to next milestone release. +It is likely to be related to the Request is no longer valid warning +zip in the same file once we start publishing to mvn +This might need the users to flush the data to disk +A simple global exclude would solve the problem for usIt looks to me the code I did is good enough to handle my use case +Creating the binary distribution is complete now +Closing resolved issues for already released versions +I did get a clean mvn Psourcecheck clean install but not sure what all I need +would be better than latest though hopefully the API between and isnt massive like the change from to +For a few scenarios it might even be faster than plain arraysGetting back to reality my own findings indicates that using for sorted search is not at all faster than plain arrays +Diff against SVN Revision is no problem with the code included +reopen which fix version snjezana +scratch that last comment it should just work and there was a reason for how it was done +Should the fix be merged into branches too +This patch does not alter the code change +bq +Rebased patch for hadoop QA +This should be an easy and nice improvement +contenttype matchThese two steps could be simply implemented in a custom Comparator for the +Basically we have switched to a mode that the runtime will try to recognize the databindings by introspection without user annotations +Ill leave it to Andy to close this one in case theres anything else planned for it +Is this already finishedWhenever i try to add an Item with fselectItem with itemDisabledtrue the item will be always treated like every other itemI am using RF issue states it was resolved in the select and inplaceSelect components +Thanks for the patch +And now I get to go reread that patch because I mustve missed something in speeding up single writers +Im currently working on editing the source files so Ill just add it straight in +I meant the later one +As i cant disclose the original module name here i have renamed to module module +Thanks verify +by the way this affectsentity code generationAND believe for now this issue should be in the main Forge project +It should be added that this also happens with maven could somebody with edit credentials update the issue +Good idea! +Missed renaming a test +Please provide some more info +Patch added +src +This is how we actually use it exept that with Websphere it is rather easy toconfigure using the admin console when installing the with Weblogic we need to modify the script that launch the server +BPEL project artifacts that demonstrate this missing on. +I just committed this. +We do need some tests and documentation before we are release ready but the patch looks good to me on the first read through +Retitling +I dont think it is critical for patch so that the automatic tests run can occur +Assigning to Justin that developed support for workspace specific cant access a workspace specific style through the global endpoint by design +Patch merged +If this issue is against the Quickstart now then close this issue and open a new one +Major responsibilities of have been moved to +Could you provide a test case that demonstrates the problem on trunk? +Feel free to backport the patch probably applies minus mavenization. +The new tests in this patch focuses on the the atomicity properties +Please review +It should be easy for both server and client to set a timeout if it needs one. +Thanks +I did have one written but it looks like it didnt get added to the patch +Not a very friendly thing for us to doIf we want to have a rollover state how about a background change? Like a warm outer glow +Doesnt detect deadlock or starvation +should not have been closed +I think we all agree that Solr does a pretty good job of hiding some of the complexity of Lucene +The more I look into that issue the more I have no idea what is actually wrong +reopen if needed +We have also such a problem on the Java side see We have to find a common way to solve this issue which could the solution that Freddy suggested on +HiI think has already addressed this +The last commit included files that had already been updated to the new styles so they really didnt need to be changed +Verified the fix on my system that was consistently showing the problem +What i like with jcr based one vs +Guys committed patch causes assertion failure error in Struts test on Windows +Redid the attached patch as a removal of over in +Artifact cleaning for party component +For example your default logger could be a log file but your logger for a particular class could go to email andor log file +But you are saying that it should be close to its first useThere are semantic differences between these two options but different situations may call for different usesI did try a similar scenario in Java and the result is putting the local variable inside the closure +bq +patch in git! Patch applied in revision . +Such extensions should be much easier if the transaction management is separated from the actual resource management +Thanks +patch doesnt break backwards compatibilityHandling of coprocessors is kept +This looks good to meA test is required before this can be committedConcatenation can and probably should be done as a separate issueMetadata would be nice to have from the start but could also be done as a separate issue later +By now issues only get fixed if is security releated or a big big bug +Because Sun likes to rebrand every couple years +Apologies for the earlier comment +You can download and run the examples yourself if you want to do so +The groupId and artifactId for these are incorrect +ok i see +Sorry the issue is incorrectly named +I tried this on and its still an issue + +Clement please close this issue if everything looks good +see attached took care of it +I was inspired to extend your ideas and refactor the code so common code only appears once +is out but this ticket is still unresolved +If you have not yet started working on this I would prefer to pick it up as I already have some code written for it +I dont think the merge to worked as intended +was not the bug reported +Still need to fill in the findReplacement method +So perhaps it would be enough to call the functions without validating the results? Any comments I missed three more functions that are part of C cbrt hypot and second patch for attached plus an updated test case +Commit Id say Enis and just leave it open +report properties because without it the tests in the dashboard failsWe can add a dependency test on the dashboard plugin to check that at least the checkstyle plugin is Im not sure its a good idea because a user can use this dashboard plugin with an old checkstyle plugin if he redefine some properties +Hold on I think I have proxy issue that is screwin stuff up +Is it time to mark this issue resolved +Updated with correct Path to whole build system is currently in redesign phase +The differences between the two environments are +transition resolvedfixed bugs to closedfixed +Please report back if this fixes the issue you were seeing +I think this message is misleading because the profile actually is activated its just not used at all in the pom being processed +No configurable via properties resource or no named PMF or persistence unitnamed found using resource loader ebe +It looks like is not initialized when calling +Instead will be with latest builds +In fact absense of command means absence of data replication +Adds the serviceThere should only be one FIXME in the patch which refers to +Could anyone review this patch +Sounds good +Is there any fix for this +Theres also writeTimeout which turns on Remotings write timeout facility +Attaching This solves the issue even though I had to keep the antcall for itI will refactor it in a separate ticket +for +For slices its always going to be an overall win +pdf on said attached the extracted text +It hasnt been setup after yesterdays bdbfsfs migration +so doing the conversion yourself +This was due to now returning a Date from the datePicker +See if that fix the problem. +Any attempt at addressing this issue of the infinite looping should address both the cases mentioned here and in as wellPossibly a simple solution is to track whether or not FOP tries to position the same object on a new page more than twice? The moment an object is placed on a fresh new page and cannot fit there is no need to continue no future pages are going to fit it either +what about using a brighter color to show the match values then the current grey +Seriously +Fixed by r +Please add further comments to that issue instead of this one. +Fixed with revision +Patch version again +Committed revision +Refactoring individual tests to belong to right project can be subsequent issues +The Apache Tapestry project could use this feature +A potential migration issue caused by this change is seen if components are accessing the ejb local homes assuming they know what the global jndi binding is +I believe this worked +i used jboss with the specified settings and deployed a ejb application and redeployed it severeal times without reproducing the problem +If this doesnt get picked up before code freeze we can deffer this to m picking this by the end of the day Ill say if it need to be deferred to or not +Burr could you provide the updated template according to Kris comment +Im anxious to cut M which updates to the latest oauth draft so Im pushing this to the next you please provide some detail as to how to reproduce this problem? sparklr head is currently on but I dont see that error +i am the seon person at maven user list who is complaining this bug again for multi projects with non aggregated javadoc this does not work +The model contains a file which describes how the feature generators are constructed and the features generators are instantiated and put together by a standard dependency injection framework +I havent looked into this but suspect the needed jars will be there when running in maven due to transitive dependenciesAs an alternative to changing the by adding mfcp entries that will normally be invalid perhaps a new jar with only these mfcp entries could be set up +bq +I do not think any one is working on it at the moment. +Thanks +Committed in revision +Updated diff files are not sufficient when renaming files +trunk +This is likely the reason why the bug was not detected during the BETA phaseOleg +Simple patch no direct tests included but the test in uses this new feature and will fail if the config doesnt work correctly +Please let me know if you need any more information +Merged on branch as of revision +Actually after careful reading of maven user guide this is not a bug +Manjula just an FYI and it is not a biggie but if you use svn stat q then you wont get following kind of output? test +When the was not set to use fully qualified name teiid was able to fail immediately upon finding a duplicate +This updates default to reflect changes since the original contributionApply from srcmainresources +committed in rthanks +for the patch +this follows kevins suggestion and provides simple instructions for the synatype eclipse plugin while keeping projects dependency isolation in tact +Carl can you pelase regenerate the patch yet another time? Ill try my best to test and commit ASAP to avoid conflicts again +Resolved with the new internal mapping scheme +Ok suggest a patch to make it a warning instead of an error +This would be really valuable +patch should be fine but it looks like you have wrong char in patch +fixed RAWPATCHANALYSIS +Perfect JulienTested +I thought it was working but I was injecting the service into the domain from a you please attach an example that reproduces your problem +There should be disabled hook set one for unvisitied one for active and one for visited stations else the result might become strange when the user wants to make the active step disabled +Good but it should work fine with Firefox there is a test case to check the supported browsers that pass without you using a firefox plugin like User Agent Switcher +If the jar command is used the JAR file itself is added to the classpath +It looks like the is available in +This is still present in the columns in the filter are mapped to the join table +Im not a RF commiter +What is wront by this configuration +test case that uses H as the all tests pass +I didnt see the note on the dependency on would be great if you could look at that one tooApologies for the missing headers and Ill look at the checksite output +Created put and test method +Okey +Autosave is always enabled and cannot be changed +Ever periodJames currently does but it is important to realise that James implements the Mailet API using Avalon but there is no dependance by the Mailet API on any APIs other than those provided in JSE JEEThere is a plan in hand to implement JNDI in James for service access this could be used to register and avalon resources in a non vendor specific way +Waiting to merge these changes until after changes are merged into the codebase +That works in this case. +This means you dont have to configure them in Spring is the autowire by name the default you want to use? Please advise +I think it would be useful to in any case support the latter option so perhaps thats all we needbq +Awesome! +I verified the patch against trunk and they now match +I am sure that the good designed software and normal processing is working perfectly until something unusual happened or a bug in code +The failed tests are unrelated to the code pathI run again on this patch +It is fixed in already +this is a regression introduced by lets handle it over there. +Sounds fine +BrianThis is a widget bugI attached screenshotsthank you +I will start with fixing tests once the API is stable as these are huge changes +Probably more useful if we remove the unhealthy nodes from the Active count since you can already get to an unhealthy node list from the web UI via the unhealthy nodes count link +bulk defer of issues to +File removal now works correctly +Is there a benefit to being able to change them to something other than CustomAlso will your average user know to check the Java level for the facet by clicking the Modify button +Sean did you try mvn up test +uses JMX for a few of its dutiesIf this is important we can look at making mxj an official dependency instead of just +Whats the issue? If youre trying to compile source or target this wont work with the HDK code +This will supersede functionality seems to work just fine +Sorry for that +This still works after rolling all the modules to the last commit of +Go ahead and close this issue +Jacob I could really use a patch file attachment and a test again +One thing I noted is that the node name included in the error message contains a different cluster ID than the one of the actual cluster node +so why this is not resolved +Fixed issue that caused tests to fail +Committed as r +Fixed in fe. +I had to modify the patch quite significantly to make it work for test class in any mode which it now does +Fixed on +Has already been fixed before the issue was raised +So maybe when is called it can in additionto saving the before also save a clone of the befores activationor perhaps just the resultDescription so that later when is called it can arrange forthe befores activation to point to the correct result descriptionThis would confine the work to trigger context management which seems likethe appropriate place for the work to be occurring +AlexIf you try this out can you please let me know how it goes for you + +Not a bug + the change looks good +Such methods are usually utility methods put in that class because it makes sense rather than create a new utility class for a single method +I dont think we want to expose the pool since this is actually a linked list of idle objects so exposing it is not that useful and would really tie GOP to the current internal implementation if we allowed subclasses to depend on it +We need a change that fixes console but doesnt regress irb especially since console IS irb +Sorry I no longer work at the location where we had this problem so I dont have access to that anymore +However it might be possible this error would still apply to the latest examples +I agree with Ahmed that we need better documentation on these commands +Let me try +I tested the same scenario againstI was able to reproduce the errorIs it possible the code you test on trunk was different than what is in the snapshots +This is it thanks issue can be closed. +Ill take care of it later today +It also should be fixed in trunk too +Don go ahead +I did a quick smoke test and gave Ryan the green light on this +I received sets of images today Ill hand them off to Brian tomorrow +in +The build has been switched to remoting waiting for test results +But only a small minority of people will run AS from an RPM +of a bulk update all resolutions changed to done please review history to original resolution type +This is also I think +moving to think you should store a soft reference to the WSDL instead so that this cache will not leak memory +does end up in the +Just adding these to the classfinder should be sufficient +null is no longer considered to be a valid value for a primitive during the method dispatch +Implemented new configuration gui in Rev +Initially tested +Can you get the error every time? What is your route isnt there any destination from the timer? +One problem with approach is that in some backend systems users can set arbitrary custom properties +Should be ready to use pending any new requirements +I applied it now +Heres a version of the close for +Remained would be startkey +several whitespace changes made it into this patch on purposeI do not like changes They occur because the line contain some whitespace and in the development process they are removed +Looks good to me test cases pass as well +It is really unprofessional imo +Juergen if needed please reopen this JIRA with more detailed information. +Log file for the error after submiting the forgot to it +Oops +This is fixed in the trunk at trunkdocsforrestsrcdocumentationcontentxdocscontrolsIt is not yet fixed on the live siteBut we wont have a new process for updating the live site for a few weeks +Since every connection is actually a new physical connection setting clientID works for us +Duplicate of +Closed upon release of. +this one will actually work +I have run for both and multiple times which all passedPlan to integrate in the afternoon if there is no further review comment +Id put the list into an unmodifiable list to stop concurrency problems because I knew the risk adding children while adding children existedIt looks like that isnt enough we need to take a snapshot of the list and then iterate through it +Here is patch that dont fail completed maps on lost tasktrackers if reducer NONE is specifiedThis is tested patch +It is declared and strongly type checked +Other CC projects use shell scripts or CC code for testing +A exports a package p with version +All JBT components selected and installed +Closed upon release of +I did not use maven to create the bundle +But AFAIK you can add new params into existing thrift api and they will still be backward compatible +Ie we just have to insure at indexing time that docs within the same group are adjacent if you want to be able to count by unique group valuesThis means that in the same group also need to be in the same segment right? Or if we use this mechanism for faceting documents with the same facet need to be in the same segment? If that is true it would make the collectors easier +I believe the fix I just checked in should cover it +This requires examples to depend on tools +I will remove the crypto aspect and refactor the design to accommodate versioning of keys under a given alias +NET needs to be accessed +Added a patch to which moves the logic introduced here from to +Then we couldalso just make a new codec for fields and maybe then wedo not have to introduce a new Field APISo I think we should try for to overhaul thedocumentfieldfieldinfos +in my environment the problem appeared when a int type column contains null values +If it is more jars then better drop the idea +Backport to Apache Camel branch and synced to Fuse and branches. +Will post the test plan to +Hi Paul +Found that the with nameserviceinvokertransportservlet is registered in the class for version of remoting but not in version TimAnyone in JBR I can ask for help on this +I am working on fixing this issue +Quick StartSmooks Design Page and Smooks Source Editor Page sections were were updated according to the new Smooks Configuration Editor +PatBen Can we mark this as wont fix +On second thought I think the right solution is to disallow KS creation when the number of live nodes cannot support the replication factor +A bullet per each foblock inside an fo dirty hack to get around the problem of displaying multiple fo blocks as list items. +This only lets you change fields in one directory and only works for TIFF +committed to trunk +findClass is not always executed +verified +Make sure you always release the when youre doing using think this patch should workCompiles +I think for now its a good way to create a separate JSON Object which contains the param +This effectively disables document order +This still happens in +Thanks I applied the patch. +The fact that to fix the bug is an easy thing doesnt mean that finding the bug is easy +However we never faced such an issue with the previous version +Could somebody kindly do a check share update with the Test Case with compressed input in and recheck MAPREDUCE please refer +thanks for the update Nemon + tested! +We are moving to soon no need to waste time on this one. +I measure QPS out of cache on KB small objects +It took me a while to uncover this issueA couple of suggestions +defer all issues to +I hope the is the correct syntax +In that case filebased project is missing bnd embedding instructions +Problem is that some code is doing instanceof where it should also be checking for +Best regards at +As reported by the user a previus commit has applied a similar patch that fix the problem. +Alexey I looked into removing the shale features but it seem like this is all spread across +Should be fixed now +Ill just reject this one then +Im going to blame it on sleep deprivation +Sure +All thats happening is the background loop is unnecessarily trying to go around and around and delete a node it KNOWS doesnt exist +at this stage +Committed with that should be removed once clerezzajena is as resolvedNOTE the temporary workaround for comment in the code +My testing isnt rigorous just a few basic test cases Ive been using but I dont see any unexpected resultsThanks for your attention. + +any ETA for the central Maven repo +Im going to keep trying to recreate the lockup +Aleksandr thanks for investigating this but your report seem to skipped right into the cause but does not say what is actually broken +I disagree +I suppose that means it will not be ported to? Is this because it is not possible just too many unresolvable conflicts? or what? +Fixed in nightly build and will be included in +My DTD skills are fairly week so may need to be looked atCurrently you can only select on value for cascade property +That will be much simpler +but it doesnt seem to work +Inside that directory there is a maven project with a file has a goal called testPlugin which is the goal that plugintest testPlugin goal and its that goal that exercises the functionality of the plugin and performs assertions on the I can help there let me know +on the direction +Tom is this still a problem +Moreover the works only at classinterface definition level so your example is not correct +Formatting. +Cleaned up the initialization so that we create and validate directories before locking them +No not yet +crating a JBRULES jira for this one. +I agree lets just push it to since its noncritical for the majority of users but it has very visible differencesGo ahead and fix Arraychoiceshufflesample using the current random logic and push it to master for RCfinal +I think +Ok I will recheck this +Patch uploaded +Thanks Hairong! +I am closing this as Wont Fix because at the moment Groovy does not intend to create binaries that can be sued without Groovy +thanks +All the SGA documents have been filed +closing pending response from the jetty we still do anything about it as there obviously are already versions of jetty out in the wild that shutdown in reverse order? +I have a number of comments not necessarily blockers for this patch In the case the digest of an entry doesnt match we will discard the entry although we have heard from the bookie +Just havent had a chance to do m full time on Codehaus from Tuesday next week so well be seeing a lot of actionI will try to get to this later today though +We are already sharding counters +I should RTFM first Please note that this diff contains fixes for this bug as well as +This is a useful utility +The same for rich attribute was hiddenpassedMarker facet removedMarker facet names attribute was removed but passedMarker facet still present plus passedLabel and passedMarker present in example code +It seems like there is enough of a workaround for you to get things going in the shared environment but please explain to me if there isn you are still around and want to chat feel free to ping me on skype werdnagreb +Resolutions other than done should not have a fix version set. + +you are correct +Removed formatting from DescriptionSet Original time Estimate +Hi are there any objections to putting this in? I know unit tests can break if applications incorrectly modify the schema +But it is available for left outer join right outer join +I think simply default to trueYes please cons up a patch +I think thats already fixed in issue is already solved in tested that for IE and IE. +bq +My solution was a misunderstanding +Yes useStream true does the trick tooI tried useStream true for Tomcat too but this breaks umlauts in TomcatI think should be rewritten in order to make it work on all JEE servers without having to use server specific hacks +The same thread could check the some of these conditions too +r fixes many of the problems +If we want to extend the scope it has so many situations +Based on an offline review this may increase memory we need to return thepartition names periodically to put a memory posted on the review Namits comment +Integrated into trunk and +There is no initialization overhead creating the test classes +Now all tests pass +The framework has no way of knowing if there is someone out there listening to events and relaying the ERROR events to anyone so it doesnt make sense just to swallow them. +Will submit a consolidated patch by tomorrow +use free to ignore if this is not relevant to trackingfixing this issue but am simply trying to run the attached test case against the trunk which I have built and installed locally and I am getting the + +Sorry Brent I led you down the wrong path +Attachment has been added with description Generated FO has been added with description OriginalSorry for taking so much time but I got some trouble on finding the portion that was giving troubleI attached the docbook file and the generated fo file +Doug The boolean accept method would be fine +Clebert I suppose you are not working on this +Boot that is the missing piece +I will redo the measures +Hi Jeff could you attach the source message that resulted in this problemIm assuming that all of the zips are Eclipse projects that need to be deployed from JBT and that the +JonasIf you get a chance could you check to see if this issue is still vaild against the latestThanks! +attaching new patch fixing Hadoop QA failures +I think you have probably misunderstood my proposalI understood the proposal and looked at the patch +Patch created against the modules directory of +Finally implemented +Refactored to imply to imply. +The ee tests had all succeeded when I ran them for +see also TODO comments in the current user configuration and in close for the release +Will commit this today then and roll a tarballFor the future we should see about adding a Release Notes field to JIRA Hadoop does this and then when they make a release its reasonably straightforward to generate notes beyond just the svn log +Done a reasonable default impl so this is lower priority now +Bulk close of resolved wont fix cant reproduce etc +I was just looking at the exceptions that failed +Implicit parameter validation and conversion is outside the scope of the core container and should be provided via extensions +I may try to add a unit test about it if necessary +Derbyall runs cleanly on this patch +Would you post the result of test patch +Removing +Best practice of naming methods that check statuses is in favour of starting with check +test case added to thanks for looking at this +There shouldnt be any user code that calls HDFS code directly should there? So I dont think is an issueDoug +Attaching a quickstart that doesnt reproduces the problem +Its fixed I think but I would not mind a look by Sanne on the deduplication algo +works ok now. +The time format really should be accessible to the user as a configuration dont know enough about how is configured to format timestamps in the list output +Good thing we have those tests! Now all test cases in both test projects are passing except one suite but this suite was already failing before this patch was applied +Pretty cool +Patch with updated test defines its own MR job independent of other tests +There also had been main methods for testing in in the types package +You are right that each blockInfo of the linked files needs to be updated when the original file is deleted +The problem with is that the setting is universal to the grunt shell or scriptIn cases where user wants to read from read from multiple sources with different loaders it will be inconvenient to be forced to use absolute uris for all of them +Whats the issue with paths +get the qa just committed this +When it comes to cookies browsers do a lot of crazy stuff to maximize compatibility +Looking at patch they created a class that uses reflect and chooses between IBM or Sun jvm +Do we expect to support HTTP URI style in requests? Is this done in a referencing plugin +Weve resolved by making testcases ensuring the issues no longer are present +I did not see any deadlocks +mrepositoryorgapachehadoopUserstucu +Attaching patch that should set the max to zero for connections toNo automated tests still but did some manual testing by hacking a test putting in sleeps whereever needed to take thread dumps of +Also verified that works as expected on +Temporarily reopening to update release note information. +Really do expect to submit this week +I do not see why Unix like boxes should be first class citizen while Windows boxes not +Ignoring missing data is not a scenario we want to support +The tag is now validated +Invalid issue. +Closing as wontfix +After fixing the problem is less visible but is is still here +If you create a service client with null argument then addressing module will not be engaged +The first apache release of the libraries fontbox and jempbox are out now +To the best of my understanding this is more likely a Hibernate Validator bug than a Spring MVC bug +With there is a new API that can be used to write your own handler for SIGUSR. +Should this be resolved +This is the policy that causes the failure. +I sent it +However I still think this is rather an improvement than a bug but I agree with you that Axis release process is quite mysteriousLets hope someone at Axis developers side will give us some feedback on this now. +I will commit this patch and put derby +Committed to trunk +I cant do much about this until the Application server also add support for multi tenancy +Then the starts up and it wants to initialize so the whole set of contexts Ive loaded are wiped out +Thanks. +verified that and can run fop successfully. +s app feels like a native app while not being a full native later a more native approach can be examined or something like this is coming from someone who really believes that native apps are far superior to mobile web vote for Android first and foremost since there is already a Spring Android project I would think that would be the easiest to do first +Moving those methods to make sense to me as it seems as unnecessary code repetitionI thought of this as well but if you look at the exceptions they are different +If some log needs to be discarded it will be the head of the log +I will do and +Reopen to test new +Ubuntu +Leaving open till I can check trunk for similar problems +Committed revision +I think this is an lots of terms pass thru checkindex fine but then the problematic term happens to have df +The cache used for this is a and this is unbounded +Verified in JBT Linux JDK to +I cant find variance in + svn ci containerSending containersrcmainjavaorgapacheopenejbcorestatefulSending containersrcmainjavaorgapacheopenejbcoretimerTransmitting file dataCommitted revision +The compiled class is identical to one compiled without specifying static it doesnt stop me writing a constructor in the type +Suggested implementation of the suid programLeo please verify the attached program +sorrythat was a wrong information +There is not enough demand to support it +familycolumn scan the majority part of extra time cost is on Create fake key and compare keys between scanner +Adding support for what in particular? Creation only +The only potential problems that might arise are some amount of additional exception overhead for unrescued exceptions passing through rescues but Ive been unable to show such a performance hit +Hi I got the same problem +Patch for svn messagingtrunk +Fixed by code change to +bulk defer of open issues to +Closing. +Thank you MayankDo we need this for trunk or other versions +What does tampered with mean? Thats exactly why we use XML Signature When you modify a signed message the signature breaks +This one is also a bug +Our product is built on Windows with Ant javac from JDK +Attached maven project +However I am now able to build the udig product into a zip file automatically using mavenMy hope is that the udig community will be able to use this work to automatically build nightly snapshotsOver the next week I will be adding links to my Git repo containing the prototype work and some instructions on how to merge the automatic build work with the udig master branchMy current focus is to convince Tycho to respect and package the required eclipse language fragments +Or return a Message object where they can transform both the body and headers +Patch applied on applied on trunk changes for this entry point to the wizard need to be made +Commit de la revisin . +Classpath is missing file extension +I checked this in Eclipse do you have any validator which allows this +The solr support for numeric fields writes the stored value as binary vs the lucene Why is this done in two different ways? Can we fix it so we only have one numerics encoding +Ive just committed this +A postgresql developer confirmed that a prepared statement must return the same types with each call so theoretically there should be no obstacle in only returning types once +The regression was made by estherbin after his modifications in See the lines after comment Added by estherbin fix was not correct +Actually Ill commit this fix first and open a separate issue about not using a boolean here +Committed to +Hardy just applied it +umm +Tested +Try the latest update +Bring back the node to healthy reservation status should reflect accordingly +I configured nothing for that so I believe thats the default behavior +Committed to trunk . +It may not be ideal but its not an errorSo Im a little confused why does solr still work fine +might be touchy since any file in a webapp not just in or could potentially be sensitive +desc DAG these are the information that needs to be propagated from compile time to execution time +Has someone fixed this for CN or does it still existAlso if you have test datasource please attach them here +At least there are objects related to the references instead +The directory name is only tied to replica id and wont change with a mastership change +I dont recall if I was the one to set it as Hibernate but I will fix it +I this issue because was closed due to not being reproducable +Thats correct +Thanks for the link Ill give that a Layer Preview etc link is something I want to share with the internal development community here so folks can play around and see what our can do and hopefully build some apps that use WMSWFS etc +Leaving open for potential performance enhancements +Please move to TBD EAP so that we can place it in release notes properly +simple syntax oh! Should have noticed this myself earlier today +This is the modified based on which resolves this issue +ThoughtsHowever as you suggest I will try to modify the function being called so that it can handle other types +Actually this can stay open +For example if you need to patch something in LOGCXX just put the updated files in the appropriate directory under HarmonybuildpatchescommonLOGCXX + around the postgres bug by not pushing timestamps past the year . +If you are using the tilesConfigurer bean you would only need to indicate the bean in one of Springs xml files +Is that a general comment or specifically for the issue regarding +Verified on Tools +Im not quite understanding what the problem is with the Comment annoation +Perhaps the final release build instructions should disable assertions? Or am I being too paranoid? +done. +Im talking about just using to build your sfsessions to get autodiscoveryIf you choose to use pure JPA or Hibernate API after creation is up to you and irrelevant for this issue. +Hi MikePlease can you take a look at this for Ivo? It looks like more CR failures on slow nodes +Otherwise it conflicts with the legacy namespace parameter forAlso added a bunch of unit tests and fixed a couple that werent failing before just because namespace were ignored +I think we still need to release the connector so Im keeping this open with a target of in before the branch in rev +After related sanity tests Ill try to make a commit for this asap +it is fixed already but unfortunately we cant go back and corrected +same for MSN I fixed. +Andy I think the Maven metadata is in sync now +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +reopen for to in cbdcfaeabadfafbc +images was aligned left +Attaching patch +Can we use that forYup I can rename thatbq +Although persist works after persist the objectId is not set properly causing subsequent update to fail +Thanks for the new stack trace +Also they can update the state of the replication under that aquired lock node +Theres also the problem that the namenode has historically been very fussy about what hostname was used to refer to isnt focused on servers with multiple interfaces +If I export or print the map in the printed dimensions should be correct right? I mean the printed km grid in should measure cm +i have deleted the revised patch form this jira. +Yep deletion exists in and works +Im seeing a Too Many Clients error on Postgres and Oracle how many minmax connections are being used here with the schema script? Whats suggestedTried this on and saw no errors +Half the modules was assume by now we can close this ticket. +Need to do further testing there to see what can be done about multiple WSDLSchema imports scattered throughout a complex WSDL +Okay but then we need synchronous lease recovery for in HBASE the patch v makes us skip lease recovery +In the CVS version there are no include after the namespace macro +Of course I meant +Any consensus on this? We have the disableCoord and now the option to use a constant score mode for all the queries +Thanks Mike for the commit +for the patch +Ill check it out and contact you lets say till weekend. +I think we should defer the namingversion discussions to another JIRA +The new build procedure assembles the entire update site all versions from jar artifacts +Java struggled with the same thing prior to Maven +Im curious if this fixes it too but changing the locking to pessimistic is a big change Im a little hesitant about this +Github mirrors from which mirrors from Github pulls dailyPlease let me know which mirror is out of date +Attaching new version with Teds comments addressed on trunk +AndrzejCuriously enough as Roland pointed out the patch actually should not make anydifference +Can we have a reference on what should look like? Widget spec is it +I used downloaded from Customer Portal +Given the discussion on which talks about removing packaging from hadoop does it make sense to wait for resolution and close this as wont fix after that +Done +Might also be nice to have something on that wiki about how to use and Cascading with Eclipse in an context +theres a lot of weird directories in some of the configstartup scripts +Regression patch has been appliedPlease verify +It happens on the latest official release of the designer +Committed the patch to trunk +I thought we can find there asimilar convertor from MS Excel to a XML format. +This was fixed with svn rev with. +Performance degradation was introduced in fix for bug +Merged the nd pull request in master. +I got a bit confused with all these classes and interfaces +This is a behavior of the security cache would need to be flushed by JMX to invalide the subject that has been how can I flush the security cache? Can we do it manually or programmatically? +The changes in this patch are good and have been rolled into the patch for +If the table is already disabled the disable operation is aDo you find this only in +Fixed some typo and adjusted font and color +quote should not be cut +If the intellij thing works we can post that link on the Accumulo web page +I created an issue to track the implementation +This patch removes the REMOTEREAD scan type from the +Thanks +Ill hook the macro back into our page and let you know how it goes +Actually maybe we should just close this one and create a JIRA for thirdparty upgrade of JBREFLECT and track that for M +One a new jira you cannot set fix version one that has already been released +As long as this change ONLY conflicts with some unit tests we can make it work for that case +Generally the clients OMS cannot easily strip out the orders that arent ecommerce so we just throw a skippable exception and move on +And from another offline suggestion putting a placeholder like addbulkfilterpart in the for upfront indication on reading it doesnt pass JPA syntax test so I have skipped it +I patched it up so that it works on my local far the changes is local to a single file for simplicity +JPA configuration is working for this seam gen example after some runtime configurations modifications and after moving into appropriate location now. +Naive implementation work in progress +How can i assign this issue to myself +The only downside is that I need to set and retrieve an extra execution context attribute. +The WBC patch is reported as working from GC team the WBJ could be tuned in a separate JIRA +If we want to change it lets file a follow on jira to change it +Obviously because the query is broken +We dont do any crazy stuff like compressing tables or checkpointing but we do add lots of rows quickly and I dont know maybe once the computer gets sufficiently fast this issue surfacesThe machine our user are using is a monster with GB of RAM although we allocate less than GB for each individual process which each have their own databaseWhat is the underlying cause here? IO not being fast enough to update the database so it has to write to the log instead? Is there any way to avoid it at all +Will this fix be applied to the branch? Were running currently +There may be some changes to based on this which could remove the interface +Its the equivalent of raise and rescue +Patch committed +whats the URL I can use to view the new content +Attached mslog captured during restart observations Restart with clean up set to false is not behaving as expected either +With the last update now the bug is changed +Thx +Revision good points on the last commentRegarding Was thinking just to specific you are right about that will be fixedRegarding I was going along with the configuration where the file is set +Thanks for the detailed bug report Sebastian +You are welcome to submit a patch which fixes this in the Velocity templates in x would be nice +small nitsremove the line as cc isnt available in what you really want to do is say dont parse cluster amis which is this moreover if you always use explicit image ids then set tooI hope this helps +Adam how are you packaging the and the servlet? Is there a war file +Fixes the issues in the drivers and adds a run script button to the view of running instances +Oracle i Oracle g +Closing as a duplicate of per comments. +Please attach a Roo script and a zipped up project in Git ID baaadebeafef +Please reopen if you can provide reproduceable test this issue could be already resolved in the last dev build. +try patch tried apply patch inside Flash Builder looks like it doesnt support that format +Verified at +patch looks goodI have committed this +Committed to trunk! +Yes issue still exists with issue exists also in the current trunk +cargo plugin is now available for use with jboss as +In my opinion there are several other issues of higher priority that this one +I suspect that the cause may lie elsewhere not with theSince Tom is on vacation this week I will do some should have said not necessarily with the since this was not an issue withWe need to first understand what has changed +It should be possible to provide the replay capability based on activities events generated directly from the execution of the BPEL process or from the BPAF representation of the execution trace +I agreed Flavio that we should let admin guys be aware of why the bookie cant start and how they should do when encountering such cases +You need to set you log level to a less verbose one +It was fixed after adding customer rule to the partyBTW I verified that this issue is fixed with my solutionCheers Kiran sorry for the late response i dont forget you I will review your new patch ASAP +I dont think at this moment +So if were adding a filesystem format can assign the GUID to the image +This is not a hard problem +Merged into master +Hi AntonAs far as I know Axis got upgraded to HC some time ago +I attached a fixed version of that lets the test pass +Both JBDS and JBDS have +it is not a but but as i said on irc today current in eap is not from brew build but prod team import it from community release thats why they added todo postfix to indicate this is to be built in brew so reject and for further hibernate related issue please just assign it to me closing. +A simple config file that uses a and committed to +Very much looking forward to this feature +JSESSIONIDSSO is the name of the cookie used for SSO +Changed priority from critical as I think its not really a bug might need some clarifications on the docs +EricFor what its worth it looks like after the fix it is not retrying AT ALL with the regular schedule of remote delivery +I noticed in the in SVN that there were still a few Dutch translations missing the attached patch fixes those +You need to encode it using +Therefore we will prepare a special document with a single checkbox will shows the error +Since the patch includes changes that affect classes I regenerated most of those classes except for the classes generated by the +Would it be useful if I added that kind of switch to the +Although that was previously considered and not done because can simply be used to integrate a different for a given URL that needs a different still dont feel this enhancement is necessary in Acegi Security given there is an acceptable alternative way of achieving it +Great +Please verify the fix in the latest. +I am using IE browser +Will any one try the patchThank you +you have completely eliminated the redistribution of your picture by simply consuming from a single point. +It would be nice to tidy up these remaining cygwin issues +Ive committed some base code to the cloud branchThis will make it easy for anyone to play with how things work +Thomas I will check this fix by Saturday and report my dont close until for above comment on tag +Is there anything else to be done here +Seems like a name collision problem because renaming the variable content in the innerClient and innerTemplate to content makes it work +Additionally if the artifacts have been downloaded and unpacked by ivy before the upgrade of Groovy to things continue to workThings fail when the dependency is not in the cache +Now it beats me as to why the comments say that this has been committed whereas this bugs state is NEW +Isnt this implemented +If the user gets a Multipolygon instead its because the database contains geometry of type theres no much we can do about this +So you want this +The file should never have been included as this is a hidden file created by some of the software I use for editing the books and the Hibernate Search and Hibernate Validator Reference Guides were included with the FP releases but since Hibernate is now the same for both the CP and FP releases these two docs are now to be included with every CP release from now on instead of with the FP am unsure of what files is referring to sorry +Here is btw the for the old project where the exception is thrown +I restored the fix in revision to prevent the deadlock +Rebased and merged into the master branch. +I just realized that there is a small problem with the patch +So its easier for me to test just have to mvn install the codegen module and run mvn antrunant in the xmlctf module +However when it comes to the numbers of votes including non bindings Type B is greater than A +This functionality has now been moved to the wiki plugin still need to add a sample and address the issues above +Looks like this might be fixed pending verification please +Im going to submit a patch for this issue after some class must be updated as well +And since it is always pull you have both oldnew and new old +Dont imagine that it is correct or even all that sensible +Attaching file for this part of fix +Nope lgtm! +That should work. +Same to me +i think if your particular case requires ordering then make a wrapper layer around Metadata in your code which maintains a metadata field that is an ordered list of keys +According to last Nick comment PB should be rendered in initial state by default and not in started +I wrote this mojo just so I didnt have to clean up a few of my pom dependencies by hand +That definitely helps with debugging as I can mvn eclipseeclipse and import the project and start debugging immediately +I use logj but its not configured using the default configurationmechanism +Note to self When verifying this JIRA also verify which is fixed in Beta but cannot be verified until this JIRA is fixed +fixed in r +Fixed on trunk. +Thanks guys +Ive sent you the files by email id rather not see them online +Thanks Jakob! +Initial QA skeleton has been created +Thank you +In production environments where the server binds to a specific address other than localhost this present a potential security risk so the adapter has been removed from the deploy directory and we do not recommend enabling it for production usage +Attaching +Add logging around setup of assertion scanner in in same place in +The first time I have seen it on a pure windows clusterBut this log is from a linux machine that was running aIt should be noted that the was running on windows thoughBut it is not producing any logs while the loopingso I suspect this could be a general problem +lucene tests are passing +Since this could be some kind of standard library interaction I thought that worth mentioning +I agree with Robert I think it is fine to be left as a +I think the interface should called because this interface define the extra tobago write methodsThe should called and the Wrapper for the none maybe or something else +As the servicemix spec is released we can close this issue +Please work with Tom F when he gets back on this one + went final on Jul it possible that full support for JSF comes in the final spring release +So it seems to me this issue is resolved wit the latest changes +fixed ia by switching to Iterables if that doesnt scale then the problem is deeper down the stack +defer all issues to +That is a flaky test +I should also note that Tim is finding the same with the vanilla data directory +Can you clarify the aboveSorry it should be +any s? +sorry I mistook affects verstions +CommittingModifying titleJust as a workaround for people not having this patch they may use Base codecs to get their delimiters in and out from the XML +Verified in in order to only change the fixed versions to exclude the FP release as this update will already appear in the CP release for customers of the FP +This has nothing to do with the web site +Thanks for the feedback +Could you clarify if we still need this for JBPMI found JUDDI also needs the large page size table space to create tables should we also move the to CP? jBPM require k pages or not now? If yes then the requirement should be merged with and the documentation to be updated properly +This time was the good one Thanks for the patch! +Our cluster is smaller two of nodes each +Findbugs warnings reported by Hudson are also not related to this patchPatch looks good to me +No code touched at all +It is not clear if the one in is intended to become private or become deprecated +Based on what Simon just said Ive updated the JMS binding to unwrap the jaxws exception and set it as xml in a jms text message +committed to in JBT +I agree improved exception handling and more user friendly exception should be fine +Then use them all as a source for your payloads +Catching up on Closing my reported issues +Triaged for assigned normal urgency +Uploading a patch with a more comprehensive testcase +fix in trunk at rev in at rev t continue the improvements for dojo based osgi portletI opened for new pure html osgi portletwill close this jira +The hsftp port is fixed across a cluster +Deprecating serialization entirely needs to be taken to the mailing list as there are quite a number of installations relying on itNo we make decisions on the mailing list +I believe this issue has been fixed within you reproduce it using JBDS trunk +committed to trunk and branch. +i think it shouldnt be done since jndi browsing is pretty expensive and doesnt match a lot of can definitely make jndi not a scannable config source +I cannot recreate this bug on beta see attached screenshotVersion Build id Build date Did you install EVERYTHING from Central? I installed only +Ill put up a fixed patch shortly +Committed revision +has some decent broker restart examples and there already a virtual destination test case for durables you could add to have a test case that can reproduce this now +Perhaps the best thing to do would be to put it in its own whiteboard package for now +Heres a first version of this +Maybe some of your test data will illustrate the case in which this happens +I am not sure whats your definition of the latest file length +Just need to do bunch of work making sure we always get from value not from key +Its about half done +Should we make extend for this JIRA? Or we can only make extend +Thanks for your VERY fast answersI will give it a try with fix things Kai +Ok Ill tell you what +The reason being is that evicting a member of the ring is more complicated since the needs to be notified to update the in case the host is a member of the ring +Having the exact same issue +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Yes but we need to wait a few days until is officially released +Now also this test failing in your local boxThankst see failures any more +Thanks for the review committed patch to just about to comment perhaps it would be good to rename hasStarted to isRunning like in +I meant via +but currently it can handle periodic high bursts of writes just fine +This patch allows me to pass classloader exception shutdown and util subsets except test with XXfalse in OPT mode +to master +Its definitely not Seam validation +In order to use Thrift you would need to generate code for your dynamic language and then distribute it to all of the nodes in your Hadoop cluster either as a library that you update for each Cassandra version or as a JARd script dependency +Vlado could you provide a screenshot with the waybrowsersim position where this bug is reproduced +BrettWe could not find a way to work around the issue so we started using Nexus +Find a way to support also added a annotation +look at line to wsdl with emitter fix +I guess what can happen is New block is allocated and then the client immediately dies without writing data The namenode crashes and needs a restart +Plain simple page with link to modal window with javascript patch for workaround +So all issues not directly assigned to an AS release are being closed +Thanks for the feature! Would you be interested ina getting support for this on the client sideb Getting support for this on +Svn complains about missing argument password +patch looks good +Same with except that Puppet is a much cleaner way of guaranteeing a state of a system IMO. +Fixed in latest snapshot +I am having so technology issues and am without a development PC at the moment +Closing Metro stack related jiras previously marked as for community contribution due to lack of interest no action. +suppose we want to test user limits we would submit jobs to a large cluster and want to measure the exact counts of the running jobs to verify that they match the expected countsTo automate this one could use the job client to get the required counts +Marking all current RESOLUTION LATER issues to be against the Nightly Build +Thanks Mike +Thanks for the patch Bryan +Victor maybe look at the bandwidth module might help +Then I committed it to master branch +Hi JevicaSorry for the late reply +xml extension does not work because the handling the +handling the case where jobs instance id could be null while getting the UUID of like a charm Murali +I renamed the file to and made the appropriate build changes +I miss that partAdd the verification to test no more allocate calls are made after SHUTDOWN is received +Thanks. +DavidThe method signatures changed recently on have you done an ant clean and rebuild since your last update +In the cases Ive seen so far the restricted users could not acces the USERSDO at I dont know why thats possible at both of my databases use an another user name which unidentified role has select grants on the use Oracle Database g Enterprise Edition Release with Partiioning Spatial and OLAP options +Please login to mysql and run following command and then keep the management server idle to reproduce the problemSET GLOBAL logoutputTABLESET GLOBAL generallogONAfter you see the problem is reproduced login to mysql again collect output ofselect from and then attach the log to the bug for analysis +Added fiddler am also surprised by the fact that Firefox serves cached data following the even though all pages include and Expires headers to prevent this +If it cant find it we fall back and try to see if there is the older listDeployed operation +Committed r +Let me know if you think this would need a new test +This patch seems to fix it for me +Jon You just argued for a metacolumn no +This is a good idea + +I will be adding test cases with my next commit so that should clear some things up +A few commentsCan we call the config RMSCHEDULERINCLUDEPORTINNODENAME instead of RMSCHEDULERUSEPORTFORNODENAME? The latter makes it seem like were only using the portAlso like in can we use dashes not periods for the part that comes after scheduler? Also it should start with notIn the get doc diferentiate should be differentiateThe whole test added to needs another space of indentationThe finally block at the end of the test shouldnt be necessary because we reinitialize with a fresh config before every test already +Install issues are being tracked in a this as testing is in progress issues will be tracked in defects +Actually constraint checking is usually done on inserting and loading rowsConstraints on reading data are just views noAny concrete example you have in mind +The admin should fix the codec issue on all region servers before moving on +Merged into main branch. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +I wanted to bring it up just in case it was a spring bug +Can you take a look to compare to your use case +Regarding implements queues with to deal with this problem +Finally added after repeated requestsJuergen +Should we rename it to to be more clear and takes file name +Of course it is always a good idea to leave plenty of disk space so you can still run but not doing so should not corrupt your databaseAs Mike said we had one scenario where we theorized that runtime file system compression violated Derbys assumption that it will only get out of disk space error when growing the file +Thanks Amareshwari! +I and put this issues number in the commit and did it again while merging the fixed changelog entry into branchx +Curt what version control system you are using? Updating from the renamed trunk folder should not be an issue with svn and cvs +Sure and that is based on dialect +Patch updated and ported from branch +Its hard to explain to people so example code and books should never tell to optimize +Thanks Omer! +Hi Mano thanks for your points I will investigate it closely +The merges to and look fine though as mentioned in the behavior of GC stuff changed +We should create new packages qlexecvectorudf qlexecvectorudflegacy and qlexecvectorudfgeneric +Ive got a patch for option +The v looks nice run passed on my server +With plugins of version it works fine +Build is blue +Oh want I mean is that the apply method would be deprecated in the future +Ive submitted a patch that is simply a rename of the testschanges attached to the filename may be different but not the code +Thanks RegardsAbhinav VaidiLabs L T Infotech LtdMumbai +The test has now been converted to +I think that logging with Debug level is enough +Not able to reproduce on a clean setup +The issue is no longer being seen as of build +Reopening for clarification +Ive posted the patch for in case somebody is interested +Moving to +According to the JWT specification the JWT should be rejected if it contains duplicated header parameters or duplicated claims +class file +cannt reproduce in ve used incorrect resolution at +Closing this since somebody provided a patch in duplicated +This is one of those features that I have to wonder isnt better done as some kind of plugin or tool +Patch applied thanks heshan for the contributionRegards +Im attaching the test code I was using to compare rawCamel JMS performance so I wont lose it +This patch is dependent on +Heres the fix which should be an improvement on the previous fix +same result +is copied from tomcat documentCluster BasicsTo run session replication in your Tomcat container the following steps should be completed +Thanks a lot Marco for your quick review and apply +Could this be because some dependent jars are missing +Setting affects Im beasting the test on and to try to find seeds that fail faster +by +I will update our default in the plugin too +The idea is to have a full addmodifysearchdeletemodrdn set of functions into ldapStudio using those clients functions +Yes sortProperty behaves in the same manner +The current version is now. +Nice! Yes itd be nice to do this refactoring in trunk as well +Did you check the byte size and did they match? Were you able to open the Zip file in WinzipI downloaded Tomcat and I am using jdk +String index out of range seems to happen on a new projects that havent had any objects +Patch applied thanks! +The best solution would be therefore to go with Mageshs original HTTP codeThe attachment contains the code changes made so far +Use reflection to maintain the original error type and added pertinent info +Does this patch has any performance implication for large clustersIt appears to me that when a region is split the has to inform the master +Is this an issue still Kev +Also on counters We need to make sure we get the counters from jobhistory when the job is done +Work in progress patch needs +Already fixed +This code is now in trunk +Regardless I think the flexibility vs +This looks good! I recommend merging it into trunk +It is time for project to bring format and checkstyle standards and refactor whole existing code +Justin is it possible to get more information about the fix that resolved this issueAlso is that release notes draft text even approaching accurate +Thanks +I think the most of the arguments in that issue still hold even if we move the consistency to the protocol level +I have not followed the work closely and plan to take a look how this will be handled there +Suggest pinging him if he has any inputs +Fixed in. +Are there any opposing concerns for having too many small modules +This was fixed by +Thanks Stephen +I wanted to use and +Added for declarative support +Thanks a lot +Scott where should we put a test for this under the new layout +Just felt it would be cleaner +I did some limited testing with local filesystem and it made literally no difference in the speed of or +Thank you for the quick fix +I dont think there is anything that can be done on this issue +so if you enter for example abc what is c? do we create a parameter called c with a null value? or do we create a parameter with null name and value c? these strategie are made to work with parameter PAIRS +the container must validate that thebean truly is passivation capable +Im not sure if we want to support loading multiple validator plugins or not +can we plesae revert this until we get it worked out +Here is the valgrind command linevalgrind full +Add configurable upper limit to amount of links well read +Ill commit it but it would be good to upload one that applies cleanly +I hadnt noticed that Luke had d this Ill commit it +Let me rerun the patch +And can do what it doseThanks a lot for all the help +ant clean package ant test Dtestcase Dmodulebeeline fails after applying this patch. +Thx Dejan! +Porting is required for fixing this issue as the JAXWS Provider needs to be overridden to properly set the thread default bus +I committed this to all long as there is no lengthy exception I am fine +The purpose of the extra in the source file is to keep the code in the member function from modifying the actual parameter value +Can anyone review it? +We just want to steer users toward standard syntax which will work in more cases +You dont need to test with verifyChecksum off I think + +This is our initial draft of the format version design +I bring the discussion to the mailing list +Ill post the updated patch soon +Committed to trunk! +patch added in JBDS +Im not really a groovy neither a SWT expert I just needed the features +AdrianHave you an idea why its like that in English sites +Would you please try this patchBest patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected. +Rescheduling for +Closing stale resolved issues +Do we have a GUI installer +Jan can you create a new jira for this and link it to this one? I prefer not to move issues around after they have already been part of a release +Is this issue still current or outdated? +I wonder if they are even supported? It might be we need to use different SQL for maybe +That works for me Tony +This is my Hackthon projectNow the can have multiple ZK servers +fixed the issue +Thank you for the patch Adam +Marking as critical because it can be confusing for the user waiting min until the plugin is installed even if the cancel button was clicked +Adds a doc method toRemoves the choice of a compact underlying data structure from contrib filters +Candidate for but it may be such a corner case its not worth it +I also had the same problem with the original bug report before I figured out how to quote codes + +attaching a new patch that gets rid of the find bugs warnings +Point me at it anytime David and I will look into it +We had the same issue with JSF +Cloud Image and now have a refresh command in the context menu that is the default refresh command +It was pretty cool to see the test integration between gadgets WC Widgets in Rave +The framework doesnt even use it everywhere so open it for extensions can lead some inconsistenciesEven if you want to implement the URL logic per component inside its class would make more sense to provide a custom request mapper testing components for some interface like +default headers can be defined at the host or agent level Header class made immutableLet me know what you think good +Assuming that Gary will fix this in the same way that he has fixed other bugs of this kind + This bug has been marked as a duplicate of + +regarding documentation +I think the net effect is trivial but it may effect some usecasesDo you agree? see what youre saying +oracle adapter already had one setup +means a group without direct not including threads in subgroups +I wonder if we should use symbolicName instead of bundleIdIt could be that the symbolic name is more unique out of the already got it as symbolicName in the pattern +For example homeApostolisworkspaceApache Gora mvn clean installWithout the patch when I execute the gora script I get the trace that I added in the descriptionI hope this helps +Needs to be merged into as well +Definitely weird +In the of the plugin Marcos Fabio implemented a better solution for the security issue than the previous think one can close this bug because the fix is included in the +But this sucks in performanceSadly there is no other way but refactoring the messagesI wont do that because I changed it in the last release and it is currently quite useable +One improvement can be done in the patch now for every file in the user log directory it is getting the jobs which are to be completed every time and checking +HiI am facing the same problem with Weblogic +Log file displays updated to show observed behaviors of port and the randomports +CraigThere will be no fixes for code line barring critical and security issues +by installing zookeeper library I meant doing a make install and installing it in standard path of your system +Use a special exception handler that is aware of the retry +No +Paul I guess you can resolve that issue +Patch applied in r +Works for me +Although I cant find a specific issue that refers to you test this with Maven and verify that it is now resolved +In fact that needs to be done to make pruning work with select from T where xyz kind of queries and I ran into that a few days back +The output is not garbled not no content type is ever set due to a piece of code commented outSecond the sample request does a band select and the result for png is an bit png with no palette that various apps cannot open but Im not sure whether to consider this an error I mean the request does not fully make sense either +Sorry when I say the doc specifically I refer to the idm wikis that what we have today +Verified fixed as of rev. +Our trace of the management of the memory pool +Open a forum thread or file a support case if you have problems configuring your application correctly +Im not sure it makes sense now to move the other ones. +Youre right new service couldshould be mainly inspired by what exists into screenshots +We use this exactly for as you were trying with the PDE plugin. +please rebase +Lets focus on the idea +Also using the word drop is confusing +it was a bug in +We did not use d like you had +It should be possible to just wire his test into the jdbc suite +For completeness yeah sounds like daughters should include events that happen their parentI think the only parent event that would appear in the daughter log would be delete? This would happen after both daughters have let go of their references +Verified once again while testing +Tried on final same problem still ve fixed this one already as part of another jira +Attribute attached is considered to be deprecated use attachTo instead +committed to trunk thanks! +The instructions on the wikki are not quite rightI did not rebuild cocoon but put the jdo jars in the lib dir +but this is slow because a lot of time is spent on the lock while it really should not happen since the lock is sharded into ones +Andy can you give us your opinion and flip it over to Patch Triage Team please +Here is a new patch that uses instead +In order to get the tokens working I need a change to the code +Reopen issue to backport to +As long as I dont have jlib nor set things work with ibm compiler and either way with ibm compiler +Take a look and let me know what you think +Do you provide a p repository already containing the most recent version? I would like to try it out +I am not able to reproduce the latest problem +That was why I logged this bugHowever redeploy setup third time did not reproduce problemSo closing this defect for now until it reappears +Thanks for the prompt attention! +Closing this issue as per Shengs comments. +If you need me to resubmit please let me know and Ill do so +bq +Committed changes to branch on revision . +took the liberty of rebasing v patches here +Sorry I forgot to mention no there are no errors in the error log +Changed the title to better match the resolution we came up with + +Run it against to test for Issue +How about a new that writes solr xml rather than using + +It appears the resolution behaviour is different on the two probably due to Map ordering + logs killing and restarting them +Please add the following description to the widgetReports on code coverage by integration tests +There is commented out code that should be removedfixed +Need to wait for next Common publish before this will compile +Heres a static method to work around the problem +implemented in trunk but did not add an extra configuration flag +Please provide exact steps to reproduce this issue +I spent the majority of my time on the server end and unfortunately its showing in the code +So should not it be really included +Please upload your patch here and we can begin to review +TomasYoure correct this is a bug +bulk move of all unresolved issues from to +Thanks NaveenI will have a look before tomorrow morning +If the running user environment contains HADOOPNAMENODEOPTS then it will trigger this problem +The CLASSPATH is set within that shell +This patch contains the C examples for some reason the README and Makefile arent showing up in the global svn diff +Patch submitted to issue is no longer relevant for it is fixed inChanging the affected version to. +pushing off remaining issues in order to release version in JIRA and create release notes +CR still uses the community documentation +Whole point of Object Inspector based design is object reuse and lazy evaluation +Patch which adds parentheses and removes the warnings +use the patch can fix the to use this patch +a key element to change +Build with mvn package +Ive verified mvn clean install +bulk defer of unresolved bugs to +I did merge this as well. + for the patch +Attaching a simple stateful mapper and a test which fails due to the bug in +This is great +Lars you can take this into itself i think +I understand this is not a new issue +I will file a separate JIRA to deal with the locking issues and a separate one to look at how we can update the web service to reduce the load it puts on the system like cursers +This is related to duplicated by plugins from the POM over those from the plugin groups in the settings wasnt fixed until. +Bo +If it is a quick deal Ill get it done tonight if not Ill update this ticket with a timeline +Well then the patch isnt quite right but the current code isnt either +The problem was the firewall setting on the mac +Sandrocan you please open the issue we still see this problem on macox windows linuxthank youAndrei +This shows up as red highlighting on the review board +a new patch attached +Thanks for pointing it outIvan +The patch looks good to me +Openly lobbying for inclusion of new Cordova API demo app for Android inPutting a more complete demo app with more example code in the hands of Cordova users may result in a better newbie experience and fewer basic questions on forumsThe pull request has no copyrights and all instances of phonegap changed to cordova +Hi JimmyFor the record maybe you could add a relationship between this case and the one for EAP if it exists add some links or information about where to find that EAP repository +Ill get to this update in the next few days +No worries thanks for looking into it +ok so what you are saying is that we need to go through the test suite sample programs etc and uniformly apply the workaround since the fix is going to be a while +Who knowsKim where this could leadKim Kim If someone knows a reason why Derby should not document support forKim bootAll and no please speak up +The M release will have NH RC support +Could be an RVM problem or a build problem. +Thanks Jon! +Specifically the problem Im concerned about here is that generates asupport with onblur which breaks as described in the forum thread +May be once this conversation dies down and we have an agreed upon approach Ill close the subtasks +Verified on Hudson +But if we need to patch then we will have to go all the way from trunk +Would you post it +a console command dstat that will query and display a tabular view of topics or queues +probably the most important things is to get the TCK tests running for starteam clearcase and perforce so we can run them in continuum and make sure they dont all of these commands be implemented? following task was done by me and Brett all providers need to utilise the TCK where possible +I reversed out the changes since I dont want to deal with package dependency and this is a rather obscure APII added the changes to I didnt know it existed but I like itI wonder if there is going to be a? It may be to soon to answer this but Im thinking probably not +Closing this bug bug was fixed in Xalan Java. +fix committed in +Maybe it doesnt make sense to create a without a Model I leave that up to you to decide +Allows the user to specify a description string of up to characters at pool creation time by displaying a for the purpose +will wait for remaining fixes and more comments on other files before closing the issuekeep up the good work +If there is a problem please open a new jira and link this one with that. +You are welcome LaurentI have to add that after putting that solution I came to realize that managing users only for identification purposes was becoming a maintenancedesign headache and so decided to look into plugging some kind of logging where I delegate the authentication to someone else plus all the benefits of not having to provide basic user management such as passoword policy user account policiy password recovery etcI hope some easy way will become available in Shero for those very usual cases. +The s memSize never shows up on the statistics page +Yep the interoperability between widgets and gadgets is going to make a really nice demo +Ok Ill try +The only feature which we dont support is correlated message groups +The attached patch shows the compile error messages to the eclipse console +I applied the patch in revision as proposedThanks for the introduced a regression +added to release notes as resolved The can now safely refresh its associated whilst a new stateful session is being created on the by another thread +this patch cannot be tested with hudson as it patches the script itself +I am not a user but it seems itd be cleaner to remove these since any code using them will have to change regardless +bq +Actually an improvement request that will be error message has been improved but the shell still misbehaves in unexpected ways and pluginManagement is not supported. +i agree that string to number casting could have some gotchas but i could see it working +Same functionality totally different implementationThe new patch uses entirely and does not use gnutls at all +Please update the docs to reflect +jconsole snapshot will look like this +The only thing is the TCK test does exclude this as a valid mechanism for detecting changes since it defines the schema and there is no available column for this +Uses DOJO for an interactive interface. +If it is significant I think we probably ought to drop the messages counter and just stick with bytes inout +SinaAs discussed on the mailing list this patch cannot be applied until the correct legal credits are inserted +with minor changes to the explanation note +Hey Florian could you check if your patch conflicts with this one? Id like to take only one of these two +Marking as invalid +a dependency on the configuration admin is essential though +What would be the point thoughColm +Ben did you make changes to allow the max worker limit to be configured? If so can you either update the patch or create a new jira with the changes +There seem to be a number of bugs concerning constraints which are getting in my way +I didnt add the new classes +FWIW I havent seen the issue yet with a Xint JVM +Otherwise it is nearly impossible to figure out whats wrong and where in your code or in +I am thinking to remove these two streams +updated the test +Test failures are not related to this patchSince this is not included new tests +The basic differnence is that I moved a lot of the duplication up to the basic tag and still used the static modes if they are available. +I verified with Jennifer that this is fixed. +Thanks Thushara! +Um sure. +Id rather cross the API bridge when someone requests it but I could be convincedI agree with the caution here for now +I think I fixed this now. +And how to express this is a problem of designer +It has been tested with java and Thank you Id say it was a significant change and hence major seems more s a major change for me too +My point wasnt that its valid but that its commonly used and may cause the same problem for other usersI apologize for being unclear in my first posting +Attached the first clumsy attempt to fix itWill supply a test case if the fix is ok with you +will commit if all pass +Also it doesnt happen all the time withmy code +Sure lets try this out +A cron task starts collecting code quality information every Sunday. +So untill we do load test again to show its existence I am moving it out now +and writing the deadnodes to a file and kicking namenode to read said file so the deadnode is excluded when allocating nodes for replicas the decommissioning facility wont do in this case Id say +It is copied by mistake while testing jsessionId +Updated variable names +Linking this issue with to record that someone requested for this feature explicitly +apply the patch +You tell me then +Thanks! +It is just that if I make a new patch attach it to this case get you to review it again and then committed to trunk I think would take more time than if you progressed this +rhotrod p cacheconfig +maybe not today +The goal is to ensure that two server instances can be started on the same machine with no startup exceptions +Is there anyway to check thisAlso this file works fine with latest ivyde release and eclipse eclipse resolving and cpu have added a picture showing the progressbar inside eclipse when it was stuck +Thanks Colin +Protobufs encoding is neither bulky nor slow and Stargate supports multi column and multi row gets and puts with it +available in CVS will be included in release +the issue is that a pyton executable is not standardAh my mistake +One can totally disregard this property. +This belongs to the specification JIRA component +There is a class in that is closer to than +Considering that Day is just like family I think we safely can go ahead +Patch has been attached please help to review +I only wanted to have maven to make it easier to setup my development environment and to do the dependency managementIll try to come up with an improve pom for the if that is not enough please let me know +I have not had a chance to build it yet but I had looked at the code in some detail and I like everything I see +Please change to if you intend to submit a fix for. +Updating the patch with no whitespace changes +It does appear to be reliably reproducible with added to java in srctestresources and running mvn clean verify Dincontainer f +Fixed in both and by adding timeout when trying to retrieve message invocation result to avoid blocking +Will this patch be able to deal with that condition +NEW are for my suggested with Suffix +This is actually wrt viewedit inventory properties in UI +The one thing I dont like about the patch is that the mcomponentInstances field is changed to be protected +We can decide to ship the JDT ui plugin in the feature patch for JDT +The input file +But because the test is concurrent you needed to get an unlucky assignment of documents to a certain thread to expose it +Im changing to How do we publish the jars to Apaches maven repository +I see no reason why this reference is needed so please remove it if you agree +I suggest resolving with link to the feature request implementation Jira +Assigning to Fernando so he can add his special brew +formatting all the code +Reapply to by the chnages in. +John could you please provide a patch for trunk as wellTrunk is a lot different from +passing namenode rpc address to datanodesWill add new multiple namenodes tests +on patch +Each app data and attempt data is stored in its own znodefile +Tested patch +Ideally this would mean trinidad goes maven as managed to run the Trinidad build from hudson environment with maven Now the build worksThe attached patch the correct maven settings at our hudson build machine are the fix for this ticketKeeping it open for a bit +That will handle the great majority of stuff +the patch for trunk looks good +You are fine with the logic that converts parameters into requests and result objects from response +I will improve the implementation of the component further blocked only with the OSGI stuff +Just added an acceptsEnter skin property for +It will be tested in JBDS yet work item for DanAlexey to ensure works +Please let us know if this solution is working fine or not +to building these jars as part of a normal build +Yes I also have the feeling that is an aspect of this issue appearing because of exhaustion of ephimeral ports on the client side due to the masive number of operations on the physical connections +IMO the way client side validations are done by type is wrong anyway +This would effectively fork the codeI lean towards the latter +Thanks +Branding Icon removed from standby in +Also added documentation for the chart component +Thats the thing Im afraid it will require some changes but I might be wrong +Will investigate this later. +I have meclipse workspace with maven source projects +The empty string passed here should be fine since in trunk for the same test the string passed is also an empty string +Ive searched and found a couple of places that need fixing as sure about the maintenance of the aspects though +thanks for the report +Thanks Roshan! +Prashant theres nothing really that ties Hue to CDH except for the fact that historically CDH has been the only distro that shipped it with Hadoop +security related extensions to to core necessary for adding security implementation of User and this case I would strongly suggest to modifiy the patch to comply with +so the full scope of the problem is more clear +Forgot to resolve. +There is plenty of progress here though much code still uses old Hadoop and specialized parameters +should have been reversedIn addition to javac warnings this patch introduces a new AND already deprecated class +Iwein since your write access is working now can you apply the patch directlyAlso at this point do you still feel that an issue should be reported against itself + +Move to which will focus more on supporting the switchyardjBPM execution environments +which properties can I try to see it working +How to do you add a PO to a +I assume you mean tried but maven dies silently in a antrun task so reverting back to cant use either because of other bugs +Hopefully they will be of use +It does not appear as if activemq is doing anything wrong in this case short of reporting an exception that says it cannot cleanup an already closed or shutdown connection. +I committed the patch to branch. +Or are you trying to test something? Please give more details +Any update on this Jessica +I suspect this is a Maven issue to support the latest I had to enable the use of some snapshot Maven plugins +The framework doesnt actually have much sychronized code in it which might cause a deadlock in any case +I have noticed one thing that causes a hang it may be related to the AS embedded hangWhen I run AS without any new bundles in the deployments folder the standard JMX bundleState bean doesnt come up +This is slightly over and above what Mojarra does in its SCI +Okay apparently the protocol for is to close this once it is in trunk +Fixed in both and branches +However we cannot use this patch because while attaching it you did not check the checkbox next to Grant license to ASF for inclusion in ASF works optionIf you reattach the patch with the license grant we will be happy to take it forward +That said the code can be guaranteed to work since the edits log will be guaranteed to end with x instead of maybe being filled with xs +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Patch for to as well. +This enhancement went into. +GA I looked in the changelog of +Ill attach to this issue once completed +I recommend considering the following pointso The actual value of can be a little confusing depending on whether the database has been +jstack of faulty cassandra + +The average id length must be over bytesBrian if ordering isnt important for some of these big bulk queries you might want to consider directly querying the shards +Move to M as Ill do this with the M move to Infinispan +Looks like only messagealerts and publishaswebservice are ours +The specific part you point to is specifically existing for tests alone +Closing as no feedback for a month. +the building process doesnt check the for the parent project therefore will miss parent projects that are cached from the current builds +warning format ld expects type long int but argument has type timet warning format ld expects type long int but argument has type timet warning format ld expects type long int but argument has type timet warning format ld expects type long int but argument has type int warning format ld expects type long int but argument has type int +The point is whether it is ok for the compiler to be so that conflicting def String s is also ok and ann s is also okStep is to decide whether to see it as a faulty behavior or not + well for me. +So if we +Clebert will you have this withing the next weeks +Not changed to Stopped until click refreshPlease have a checkThank you +Updated patch that also removes some classes under sling havent found out how to hook a script to POST with a delete selector tried and various other options but that didnt workLooking at the code I think that might not work at all and I didnt find tests related to this +Attaching a new patch with some minor offline comments from Ravi and Vinay +Rob Stryker do you know if anythings changed in the code to register runtimes? I think thats whats throwing the validation error but cant verify that since I dont know that part of the code at all +Do you now when the release is planned? +I confirmed on and +Reload the patch by minor changes +Should be fixed by the use of Sonar Runner API s assume this is fixed thanks to Sonar Runner. +KristianPlease see attached pom from output +So since there needed to be two different implementations I just went with making the monkey factory +should we close as a duplicate + +Hopefully well havesomething by the time you return from vacationI just had a quick glance on the patch +Looks good to me +This one should work +The issue does not appear to be related to clustering configuration or farming +I might be missing something but the patch moves it down into the while loop from outside itCan we do something simple like log every of shuffle progress or something like that +This patch just converts the Fields storeindex type to be the same as in java lucene +Here is a one line patch for the issue +Added the methods where one can specify the database to perform getTable or dropTable on +Logging is done in JSPC +Here you go +I committed this to trunk and branch +Ive committed a first cut with only one constructor implemented +Test case with in as has been released. +Thanks it works now. +The documentation might mention want to mention any issues around string lengths +Could have an export command that creates special files along with a list of files to distcp +You may close this issue and splitoff a separate issue to support the identityBehavior functionality +Agreed it would be better to move it outside of +If I can narrow it down to the field that is causing the issue in my index what information can I provide to you that would give you more insight into what is happening +However we still need to have access to both the repair session and the future from the executor so the implementation returns a pair of those two objects +dot in itBug on parameter with +Then a committer can move the artifacts to the whirr bucket at commit time +Id rather or Confluence +My concern is that this much flexibility makes it easy for a to hang a different wo any clear indication that such action at a distance has taken place making it hard to track down what happened or wherebq +Ideal solution would be to exec this from hbaseexampleexec the snapshot command with the correct args +What you think of the idea of using supervisor or any of the other babysitting programs instead of writing our own from new? If you need to have hbase regionservers dump out their servername so you know what to kill up in zk that can be done easy enough +This is to be applied over the final patch on and assumes is in +This patch adds the attribute hourNotation to the schedule component +Thank you I will have a look at it soon! +The patch still appears to save an empty +The statement about the was independent of the proposed patch +This facility enable business user to define product url for a product according to the business requirement and locality of the business +The discussion here seems to be about the issue of using an +Let me the YARN only patch so that Jenkins can pick up YARN changes first +Weve been having some issues with that part +then we need to parse the http request data verification of +It happens twice this morning during the load of some data in a rails controller +It might be a problem at your end +Based on the profiles Im seeing for bloat after r it appears theres an additional speedup of prototype images on running bloat to be had by implementing this feature +It may helphaving a look to see if you can find a bug report or a fix for that +We now ask the for the list operation +marking fix for so its not overlooked prior to code guess it safe enough to commit this patch +Moving out from ELB is not proposed feature for +If you want to add my request to the TODO list in CVS from here that is finebut I would like to track this through this requestPier Fumagalli has set up the Apache Bug Database +Doc update to correct FQCN +decided to leave numerictokenstream untouched and simply modify internals of +The baseline compilers GC maps are more restricted than the opt compilers in that we only get gc map per bytecode and the typical assumption in the maps is that the operand stack at all GC points is the same as it is at the beginning of the bytecode +These classes have no dependencies to Excel parsing and could be used for similar rendering purposes also by other parser implementations +So according to the latest report its broken with even for master +Ive just encountered issues with myself and am looking to apply your patches in hopes theyll address the problem Ive encountered with s nested within a +Try again with different format to please the checking script +integration test showing issues +Modified impl to synchronize access to object during initialization and shutdown of the client to prevent concurrent access by multiple threads + +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Ive added to parent and srcmaindescriptors +this is issued is marked Fixed but i still see with tomcat deployment and there is no traces in SVN it is fixed either +This is now fixed +If the intention is to support adapters and the base protocol offers enough compatibility to detect protocol version which we currently have then it seems straight forward +Minor fixes r Mention ldconfig in cppINSTALL instructionsr Remove obsolete options from example file. +Do you have any ideas on this one? Its a minor nit but one someone may think indicates something worse +I did not see any directions on where it should go so assumed attaching it here was best +This change doesnt affect the exported symbols or anything else +Should be an easy fix +Fixing capitalization +I think youve describe the problem completely Michael +If replication factor is then the are set to active +done +xml file without the BOM caracters +Yes I think you do have a problem but as far as I can tell it is not in Commons IO more likely inAre you sure it is catchingAnd if so what is it doing with the ExceptiionAre you sure it is not reporting a fatal error which is then being treated by as requiring shutdown +Tests can be added there for the three rules above +can you elaborate on the initial description? do you mean the filesystem location +Thanks. +If this has been resolved then go ahead and close the bug +It is bound into JNDI at java value of the transacted flag is irrelevent with Username swaldman +Perhaps if I stop commenting and start coding I will find the time +To disable hostname verification one can set the following system propertyallowall +When jetty runs out of handler threads does it not accept new connections or does it accept the connection and wait for a thread to become free to handle itNot sure if this is still the case but I believe Jetty did just use the standard socket backlog queue and set it by default to the number of service threads so you can have that many threadless requests queued up +We do have a lot of synchronization machinery to make it safe to access a connection simultaneously from different threads so in that sense its supported +Verified +Committed at svn +a patch for the pom to fix builds + +fix patchFinally got around to testing this under TRUNK and the patch seems to fix the thanks +I agree the configuration updates should be controlled but it just means that the developer should be able to control what will happen at runtime we dont need to force multiple phases for bringing the configuration live +RebasedUpdated as per review commentThe test class was changed added new test case for larger layload +The attached PR will disable on Linux if the SWTGTK isnt setThe plugin hasnt been removed so users that want to use the VPE editor on Linux need to start Eclipse with GTK as followsexport SWTGTK +Assume closed as resolved and released +Ive enabled heap dump on out of memory error +Use the configured auth type instead of just kerberos or plain +We should look at fixing this for we agreed on the output I close it as Wont Fix +It looks like Springs is using the Threads context +Upon general request we started with grid support +Thanks for the PDF +This script is useful for duplicating the problem +Ill investigate. +After all its easy to publish your own WSDL you just need to put it under the webapp context to allow people to download it +Any info on thisAs to the proposed changes theres no point in caching on Class keys with soft refs +public publicStarting up serving public on port Hit to stop the serverAnd the site is up Yay +Hi Savas thank you so much +This is indeed a quite simple class but also quite dependent on Commons Configuration or possibly even a specific version of the Commons Configuration package +I took all of your suggestions except I didnt add null checks the variables in question are never reassigned to null so theres no way they could be as far as I can tell +May be Sebb should be the assignee +Also the syntax for access to arraycollection size +This is working in didnt notice what Emmanuel mentioned in the logs but if there are additional missing files it can be a separate issue related to the Struts migration. +I should have marked it as resolved +Ill try to post shortly +Verified in +Is it possible to channel these messages back to client side to make this logging more useful +The following action should be performed before releasing archetypes the archetype catalog is not up to date the http and jms archetypes use old endpoints it should provide new endpoints definition new archetypes should be added like be compliant with file or ftp archetype the mail archetype should be split in two parts mail poller and mail sender +Ive just committed this! +thought has occurred to me that the difference between your and my runs may be not in the Linux distros but in the installation settings +Committed to +Hi Eugen Ive read the article and I am interested on this subject i am currently playing with the James Server and will look for the way search should be implemented +Updated patch and I updated reviewboard +I see +This is the patch we came up to fix the issue +test patch +So we would not have the requirement anymore that the servlet must start first +Tying in YARN would definitely be great +So I am afraid that I can not take the wrapper solution +Verified on. +Without headers for million messages I got a speedup from msgssecnode to ! Ill test this on my node cluster home and if I get similar results I will remove the headerslock altogether +hence I am the issue +New patch has been attachedThanks for notice +Thanks very much KnutCommitted patch to documentation trunk at revision +May be after commit we can have this JIRA open for couple of days +You can leave out version info from a plan in which case it will use the most recent available +There are no released artifacts to move +patch is for t fix as this is for an unsupported version of Wicket. +You should bring this up with the Jetty group. +By my count its being added two times at least and probably threeGood stuff +It does correctly handle the page numbers starting with . +Discussion initiated on mailinglist +The idea is that any bean can actually come from an object mapped with Hibernate or any other ORM tool +txsrcmaincpp +First commit should be set to resolved instead of closed +IMHO using for creating a parameterized URL is cleaner than creating an directly +Test project attached +Tha patch has gone stale could you please rebase the patch with latest +pushed +Ill attach the full log in case it helps +Should also work from the command line if you get the classpath right +Ill see if I can distill it down to a small reproducible test case +Maybe the right thing to do is to have a single constructor and change other tests to specify defaults +Bulk closing after release +I Have the Same Problem +Are you looking for validation of the API or the Xalan api the jaxp apis +Many of the seemingly blank lines contain space charactersthanks +This is no longer needed +post commit hooks are now installed +Attached patch implements support for test exclusion and provides corresponding documentation +bq +I had some difficulties to apply this patch today +Freeman do you mind if I do this bundle? I would like to submit the bundles release to vote tomorrow +StackTedI have given my word to Liu on commiting this +Any future changes to the URI structure should be opened as separate issues. +Wed have to also getthe Seam folks to respin as the AS guys dont touch the Seam component in we should clean this up in EAP . +We dont need accounts on the project +This is reproducible +At cleanup time method tearDown iterates the list of registered oids retrieves the pc instances using get and calls to remove the instance +Its a little unclean as we dont have a nice way to isolate the java methods from the ruby methods and absolutely zero formatting from the output of the different commands as that is currently handled by the individual command implementations +this should be an issue caused by this upgrade +I think it would be better to spend the time fixing the node API +Heavy committed at revision is it possible in a current implementation to update the doc values fields without a whole document? +Thanks for the clarification UweOut of curiosity and for reference are there any reasons for the abstraction having to overwrite setPivotcomparePivot? Using the implementation in my patch would actually allow to get rid of having to overwrite these methods at all possibly being faster due to removal of some calls depending on JVM optimization and possibly being slower due to a few more swaps and branches in the code +last path doesnt fix problem this one can fix it correctly I thinkAnd after this path on my machine the performance of harmony is better than riSergey Its very kind of you and very helpful if you could use measure the path +So then the server fails to startVerified in JBDS H +Itd only be useful in the case where youre always looking for EXACT matches on rowcolts since you cant do any behavior there +if it is part of a larger change include it with that change. +AXISSTRTOUL +Thanks +I guess an alternative would be a placeholder +I know that we have historically gone back and forth on even allowing them to do so +Can someone estimate when the next version including this fix will be releasedThanks +The heartbeat recheck interval was set to be minutes +Thanks +If you find further issues please create a new jira and link it to this one. +Hopefully this important piece will not be ignored by users +I assume not +Has been incorporated to the latest ydist patch of +Revision number to the fixed version so it can show up in the release notes +I understand that using in memory this is likely the worst case +r in santuarior in wssj +What you should get if you grab cvs head is Type does not specify the required attributes to be a ComponentI have updated the Spice cvs so that component now has a descriptor +Thanks Ram. + +Youre welcome to take a run at it +each block could be present multiple times in this array +In any event +It seems it is just being too aggressive +Obviously if entity is involved in associations then it has to be indexed at runtime. +Please try to reproduce this it worked for you before maybe something today +Radim Igor If you have failing scenarios please try to include some testcode +Any new issues that are found will be logged as new issues +I would also like to know +Here I discovered null as a response for restconnectorschemalistshowalltrue when handling H connections +I doubt that there will be any failure but I am just being careful +Criteria seems to be safer because you might want to be able to protect yourself against like select from Cat cat inner join +bq +It has been confirmed on Firefox Google Chrome and IE details Weblogic Server Facelets of a bulk update all resolutions changed to done please review history to original resolution type +Duplicate of +patch applies cleanly on my box +As a consequence we have to rely on the specified path to indicate where to apply a relative path at least for the general Im afraid the best we can do is to document this in particular the difference between s have added corresponding javadoc to s +I just added the log file for a snapshot +Please tell me if you need further versions +on patch for +I think v for verbose is more common so I wouldnt be in favour of changing now in +The filter would raise an exception if it encountered an tuple +contextMenu rewritten however not super stable right now need to fix it when I know how +Build test checked on Lin iaemt +Trivial patch so skipping RBCorrects the default documented setting for and also adds detail about the possibility of sinks not being balanced if it is not enabled +As we dont have any Powered by logos yet I guess we are done here +Chapter +Marking this one as closed since throwing that wrap the checked exception is a valid workaround. +Committed the patch with suggested changes in revision +Applied with minor changes thanks. +Straightforward patch +The first is consistency +The goal is easier maintenance for anyone +Committed small changes +Changed scope of this issue to include index and any other auxillaries such as bloomfilter dump +I still have the very same failure +If genavro evolves enough all the user facing ease of use stuff for managing schemas could go there while the individual language implementations can all be as simple as possible and only use the JSON +I am changing this issues type from Bug to Feature +Smooks editor is no longer being developed +Hello Gianny Please find a patch for g branch +Damn +I really would like to replace that with a script that at most needs a role account for accessing the information. +worked on my test +I was able to reproduce this NPEIll add a junit test for it fix have in the code +I have prepared a preliminary draft of the new Web page for Documentation +Excellent! Looking forward to it in +Fixed with revision +Works fine thus I close this issue +Thanks for catching this +I committed the patch +I dont know what I was smoking with my previous comment +We might do javascript controlled auth but we have no plans to introduce state in a stateless exactly is DIGEST non compliant +this issue should now be resolved +Profile included in activeProfiles in +Hudson builds are now successful +Though sometimes choosing a balance of the regions can be an expensive operation so the balancer should provide an api for a cheap balance +in the rhs features directory dpropped beta shale jbpm +If that the case I could easily build a remove duplicates function on the reducer side +Please close this issue. +Yes archetype was bumped from to +You will probably need to add a set and get to The is a Split factory no? Who else would need to create SplitsSplits are passed in and the RPC mechanism supports polymorphism +This should already be fixed in trunk and I just released of the jruby FFI gem with those Mikes replacement methods in it which should fix it for and earlier +Attached unit test with Cookie and Host headers assertions +librecords +Let us discuss this as part of a new issue or on the list. +Thanks for the reviews DagResolving issueI dont expect to do more work here +Will investigate when I have time to try and create a test case then +Reopened to set the Fix Versions field to JDO beta. +The occurs more often. +Issue workaround Using the default id lookup scripts instead of the legacy lookup scripts +Thanks for the test project guys +I think they were left over from what I did in Seam before I fixed a bug about adding the same handler more than once +Ok got it +Bulk move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Screenshot of error message that stops the update process in jbds to and results in broken jbds +Since is not syncrhonized and Vector is Im wondering how much gain is from removing the syncrhonization as compared to using the hash +Ok strongest hypothesis this is a sign ofJoaquim was some row cache enabled on the incriminated CF +As long as these mappings exist the server believes that the compiled exist at the speicified path and tries to serve them from that pathAlso to bear in mind is the fact that the path that Jasper uses to place its precompiled JSP classes is Jasper specific +I think its better to write an edit which might result in a problem at next reboot than it is to actually crash a production cluster +I tried some simple changes to to reproduce the problem and could not +I figured youd start the discussion on +Further investigations led to the problem of bug +is still required though +So whats the suprise? There are no zombiesThat is a really strange problem +The link is available at this page in Bugzilla References +Reopening to backport the patch to the patch for as revision +Still a problem in the CP CR build ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun was a miscommunication between Trev and myself none of the changes associated with this merge were pulled into the build +ok for now +proposed priority to minor. +Lets open new issue Ted to do other balancing algorithms +eclipse this is another one related to bot blocking M Len promise to look for to Andrej since the QE responsibilities have changed and he is now the new QE person for BPEL +Did you try multiple runs +Thanks Andrew. +Duplicate of +Retargeting fixes for SCR release +so I add code to remove the znode first +now you got it +Using the latest sandbox snapshot its removed by default you have to reenable it by setting a specific policy. +Fixed the committed the patch to as well +The change has been applied to all active branches . +committed +What about if youre not running on Fedora +I hope so +Are you starting Jenkins build +Small note we should use openssl genrsa out The number is the size of the key in bits +Cant reproduce this so closing +Sorry guys missed this note +Committed in rev +Fixed with blockers and critical issues are resolved but not verified +Im not real sure how hard it would be to add but I agree it would make sense +see +These include test Ubuntu +It is a little more flexible than the text connector but lacks a way to set multiple parsing formats for dates +I have a convoluted test case that exhibits this issue definitely a bug in juddi +Unfortunately there is not an immediately obvious fix for this so you may want to consider some of the workarounds discussed in PaulI see that you recently added a link to your project on +There is no planned date for yet +Stupid me +Looks much better thanks +Since its a length and not a position I guess the exception thrown by the client is more accurateIf we throw exceptions for length and up I think its reasonable to throw an exception for length as well even though the wording in the spec is a bit unclear +Surefire is already doing filtering based on the filename and category annotations so why is this any differentIf you dont want this code in surefire what about adding an feature to allow for the inject of custom junit Filters into the classpath scanning process +Verified during test cycle looks fixed closing. +Syntax highlighting would also benefit from the preview button since the code is not highlighted in the editor. +This avoids having which uses the term integration in two different contexts +Patch provided in +I have the same problem also with the possible option parseContenttrue doesnt seem to solve the problem +How about making it an option to start on an ephemeral port as in production it will only complicate firewall setup +Mass closing all issues that have been in resolved state for months or more without any feedback or update +To fix this a needs to be implemented that supports wires +We could probably eliminate this dependency but I think both bugs are useful to land and clean this mess up +Erick the github repo was only used to expose the new structure +The behavior of is defined by the JSF specification +Thanks +Not really an issue for writerloop and readerloop here because theyve got only one receive statement apiece with a clause for every message they expect to receive +Ive added it but havent had time to debug it so it probably wont work +BPEL samples succesfully resurrected +added setting of target attribute to of unclosed issues. +SLFJ bug was fixed moments ago and is part of SLFJ version +Defer all open issues to +Fixed in Geotools for and closing all issues that have been in resolved state for more than a month without further comments +Adding to fix version as this fix was ported to the branch. +All of the map projections except the Transverse Mercator have been tested +After thinking more about Im not sure now whether its a valid approach +I cant execute it successfully but it appears that I need a license value to pass +Pushed upstream. +After talking offline decided the only thing to do for this ticket is to correctly chown the etccassandra there was confusion before +Applied the patch into trunk and updated the wiki page +It is possible that an implementation that is being called via the serviceConsumer can return a Response object so yes this should be taken into account too +Manually tested with Chrome and integration tests +As discussed in we can do this with a simple groovy script again +I greped through the source yesterday it appears in a number of places +strictfp would help but it slows down computations. +Jim it may be critical but it cant be critical for M since it is a fix +my patchset above does not noticably improve insert performance fwiw. +The most important one is that even if someone is hitting the web service a lot the history server should still be usable +None of the class files for the ant tasks are there entire orgapachemavenartifactant package is missing +That is the code change I am proposingI would like to hear some thoughts from others as well on this +I have not seen the entire code but it might lead to some cross dependancies during startup +Having said that I agree with Ralf that we need to some of the old assumptions in the light of multiple JDO configurations +As far as I know depends on Xerces for marshalling but for unmarshalling any sax parser can be used +I am working on that point +I will fix alfredo jar issue +isnt a very joyful activity +adds expclusion list +Attaching a repro for this +I may modify your code and use it as an example +Patch for review +Which JVM do you use +That is because Mockito applies its verification only after the test has been executed and cannot see previous changes to a reused instance +It was released with the fix in +in the statements looks like you are omitting irrrelevant stuff +My initial findingsconclusionsEither +No response assuming ok. +Ok I will try on Windows machine +Now tests work for me +Tried JBMAR and the issue is still present so it must be something else +I can reproduce the problem +We should instead note that it was redirected and to which URL in the fetcher output +For trunk we should do the right fix the hack is appropriate for +But the was failing if we remove the balancer in theThough the changes are many passing in constructor should be fine +Im still not a fan of these JIRA issues that are multiple things all in one. +Its been a requirement from the start due to redistribution issues +Ive checked s ns for ns it works correctly it changes during selection start and end offcets are correct +Also it seems to be necessary to include appropriate XML files into because otherwise wont work at all +I dont think it has anything to do with correctness +Im closing this issue for now but please reopen if needed. +I am not sure what happened but it looks like my latest build did not make it into the MEAD repo +Patch against Hive with unit against with unit tests +Closing the issue as version has been released now. +no problem for has been fixed +Maybe since May or some other time near that +we can use it to do other end to end test +If you switch back and forth you should clean the target first + and are not consistent +So this is strictly a JIT issue +There seems to be a mistake with the attached patch or at least they dont correspond to the description +I mean classes +Strangely enough this problems still persists and needs to be addressed. +The solution is to add a judgement for default servlet before throw a exception +This patch is slightly more fix doesnt work for signed think this may be the wrong need to create a separate pom for the nested jars and depend on it. +and +The attachment contains all thats necessary to back up the reopening +Running tests now +for the patch +Proposed you please create a review request for the patch? thanks. +While testing I discovered that the Threshold field for Mtom support of SOAP bindings was not available in the tooling +Cool +However rolling out a non final to prod would raise a few eyebrows around here Im sure +Changing default mode to merge best mimics previous behavior however you run the risk default config overwriting customizations +Updated in revision . +Ill use a workaround of splitting up my large arrays and hashes and stitching them together at I did search for this issue on JIRA and came up blank +YES manually tested! +This will include sampling +Patch looks fine to me +Thanks G +On second thought it could probably remove the fetch part altogether instead of just the keyword because fetching shouldnt affect the number of results +Though thats a bit subjective +Reopened to fix the bug for +What we normally have done is use a build tool to minify js when we do a production build for our own js +Attaching full debug output +I broke Validation preferences when fixing should be fixed verify +I shall keep this in mind while working onthe coding for the Network Interfaces +I agree +The test fails with sun jdk but passes with jrockit +Changing to enhancement since this is working as documented +Im afraid that this is how the feature has been designed The default value will be applied immediately whenever the placeholder is being parsedWe generally dont recommend multiple to process the very same placeholders +You should probably make the set and get methods synchronized here +There are two different files under cppincludeqpidagent and cppsrcqpidmanagement directoriesWhich one should I use +So far so goodFranz +That explains the action could not be found problem +Attaching the patch +The change you removed right that was introduced in? Is it ok? Or are we going to leave the change and address in +good other than that lgtm + +WTFPL +But also necessary in append I was thinking of that other similar JIRA about leases + committed the patch to trunk +Is there an MR design doc? A couple of people have asked me about this would be very useful to share +committed northwind database file and switched to sqlexpress for now +Bulk move from to to to +Totally agree with Rober s comment +Updating InterfaceThanks patchThanks it good to rename as for consistency with +As for changes Im referring to lines such as etc +Maybe a Rampart issue +Column pruning should still to submit patch state +has been added +and can configure one as per their environment needsAlso I wanted to add that since the new hadoop security is kerberos +the unicode module is new in R +If the remainder of the abstraction can be implemented right away then it might be ok to postpone in the interest of progress +it also contains a test case +If you think that it is not like to change the value of the multiplier then we dont have to make it configurable +reflects why is getting blamed +Since nobody has claimed this as yet and its critical I can take a look +I need this feature now so I had to recreate the patch for current SlingFelix +You are unable to state report plugin dependencies like you can for build plugins +We are seeing lots of permutations on this issue up in mailing lists +Do you mind opening a new JIRA issue with this information so you can attach a patch to that? Thanks +Richard please provide information necessary to document this issue +The patch is missing the class +This is going to seriously degrade transaction throughput +Exclude Commons Logging from the + +You mean the upgrade to Guava ? Yes probably makes sense +failed on Hudson but not on any of other machines I tested it +Good point +amended to accept proxy to allow HTTP and SOCKS proxy is hard for me to tell since this issue contains Java files instead of a patch but I believe that contains these changes +Attached is the source for the three services +I dont need primary key but Hibernate forces me to define one +Migrated to snapshot +frequency of means it is probably a typo in another this means nothingI think typos are another problem though +The change to the speculative execution policy is a change and should be only +Lukas is not the only one that will be working on this issue he is just the one that will tackle his builds first +Patch applied +Patch to fix this issue including test +Issue seems to be fixed +Thanks Steven +polygons +Implement layer B streaming at +I will upload the patch +Thank you +I see that this issue was fixed in beta +Making this available in the UI is cool IMOAs soon as this is committed Ill blast a message out to the Solr list to see what people think +We should give a option to user to register service IPs or alias +when I run the nbm goal I can verify that files at targetinstallerbuild and at targetnbistub are both the same with the file in srcmainresources +If someone is interested implementing this we would certainly help and give pointers where to start +I also dont feel too strongly about MapperReducers as inner classes +Sorry for not checking for existing issue +Applied patch r +OK new patches look good Uwe! Thanks for fixing this +but it appears to have dropped the time spent in Thrift struct constructors for me by +The above code returns +has been marked as a duplicate of this bug +Verified by Mark. +No test case no implementation +Will be in +Figured my confusion +Fixed on trunk r + +JeffreyCan you attach your sample stylesheet and source documents to this Jira issue as files and grant a license to the ASF? I would like to get this into the conformance suite and the bug into the regression suiteThanks! attach page states that test cases do not need to have a licenses granted to the ASF +Patches applied +This one will just cover the plumbing +This isnt the perfect fix since means download Maven dependencies but now youre downloading from a Maven repository +I also verified some of the functionality as its the primary consumer of Avro within Hadoop +the refractor work for the web security b +Could be useful for uploads like certificates for Apple Push Network +This looks great +fixed. +Closing issue fixed in released version. +But the Palette is supposed to help you to use tags from which are in classpath of your project isnt it? +However +The local tests suite passed all except for which has a known issue at the moment +Attached Maven project required for repo steps and screenshot showing threads +Closing as Cant Reproduce Bug +Fixed a few failing tests relating to serverside unit tests missing Webdav matters of style for future patches match current formatting we dont tend to use final parameters we dont tend to use +Can you please explain the individual changes for the sake for records? Couldnt see what this issue is addressing without looking at the patch itself +For example it can be optimized to use less JVM memory and minimize Java garbage collectionSome of the concrete use cases for the plugin are hash aggregation hash join query etcThanks +I am inclined to revert for now and add Alex test at the same time to guard against this in the future +I removed some nocommits but I think added more Still in progress +Even when compiling with the script doesnt find or in the Ivy cache +Patch committed +To be more close to I committed +Shelly was handling this for EAP +fixed with revision some reason this bug fix although fixed in subversion is not included in any binary distribution until nowAnyone else noticed that? Are more bug fixes affected +Trying to apply the patch now. +Hi Rado Where can I find the standalone +This is +The query result consists of one element per groupI think that automatically adding fields to the grouping clause is likely to change the expected result and therefore wont be what the user expects +replacing the JSF JAR in the AS should be enough +This is the pacth related of the fix Ive done +what JDKJVM is thisthanksdimsPS either way it is not a blocker +I hate it all and may just remove all logging whatsoever +I guess the hook is just the same as and its Protocols +How do you plan to remove the current directory from the patch queue once a build completes? Do you have a patch +What happened to the tabletransposed syntax +Appears to be a compiler problem. +Thanks for the reviewsPlanning to commit this afternoon +How about working on to convert the current ones to forest? The patch may already do this. +Chris if you have the code running is there any harm in integrating it back into the branch +thanks +Tests pass locally +New feature backwards compatible +This problem appears to be gone in Xalan. +A fix was made to scout reducing the impact of the connections so we are deferring this to a later release +I would like to incorporate this into +BTW Im using the Eclipse IDE let me know of any other tipsplugins that are useful +When RET has displaytrue it looks like it causes TLF to think someone pasted in the which might take TLF down some code path that allows the keyboard to show upI dont have a device to test it on so I cant help much at this timeIt occurs to me that a possible workaround might be to always force that paste operation after each keystroke +If the modify also adds a new objectClass value and new attributes then you increment those schema entities etcThis way we know if entries in the server depend on schema elements without performing an expensive search to find all entries in massive directories +Certification is going away in +glen I know there is an oracle domain index implementation too so maybe a database category isnt a bad ideado you know of any others +Since the parameters didnt do anything I deprecated the accessors and removed them from the docs +Please check that it was applied as expected +Commited this +Im wondering if my future work on composing the netbeans installation from nbms on remote repositories could eventually render this issue obsolete +This patch addresses the proposal on and this issueAnd jsp file changes for findbugs warning +Do we still need on and initLeader if we keep handle +Thanks Devaraj +Im still having the paste problem with the latest patch +The test is currently commented out +Theres a third bug involved that can hide the second is calculated incorrectlyPatch adds tests for the two main bugs and fixes them +Thanks a lot Chetan for integrating the patchThe json support covers the special use case of a headless operation +I just took a quick look at the source code to +just committed this +arguments added to both afterStep and afterJob listener missed out the onError methods though +Yes sorry I will be looking at this week at some point been busy trying to figure out s fine sir please take your time +With java version the ratio is almost about so the test fails sometimes +Deepalthe problem appears only if you use the same callback to process all responses +ill send an email to the dev list to see if someone can submit it +Using Codehaus Bamboo for CI and site documentation deployed. +Is this going to make it into? +bq +A logo is no longer used for branding +I actually committed now some code which would use the latter approach +Ill post this patch when that one gets committed +So the question is what to do if a lazy implementation of file status tries to fetch the permissions for a file thats been deleted since the was created? I guess returning the default values in this case would be reasonable + +Rob I think this has been fixed invalidated by your refactoring is that true +An ant script or a batch file is fine as well though Ill probably still evaluate it manually +OK I spunoff +This is fixed now +Doesnt look like a lot of code but turned out to be tricky to resolvethis without substantial rewritingRobert +Bruce Keith are you guys making any progress on this at all? in other words is this still an issue at all +I have tempted a fix for OSX +by the way if you wona to use JBDS betta on x lunux it should be working fine on openjdk +Worked it out +Did you clean everything out +From Knuts comments this is only an optimization issueAs resolves this hang we can the shutdown in the test jdbcapi +And I am currently working on providing the UI support to this functionality +I think brett was fighting to get the extensions working but for now I would say just toss it in the core +Suffice it to say erlangmonitor on a random process in the write path does little to assuage my fears that were entering dangerous territory relying on our own file writing buffersNot sure where there this is three abstractions away but I gotta have to defer to the experts Filipe and Damien here +Im not assigned to the Pivot project so the JIRA workflow wont let me assign myself +Submit patch for review +That is why I went to the Redhat Linux system and ran with the same query the same JDBC driver and the same JDBC connection string and the same java runtime environment + for this issue +Patch to add test case + +I dont think completely fixed this issue but it should allow about twice as much output before it gets truncated. +Weve run into the same problem +Checking it inRegarding the CLI stuff can you look at the latest patch at? We will need to add something like similar to there +Does anyone interesting in thisThis issue goal is to figure out improvements and implement new ideas of Message communication +Checked interaction with contextMenu tooltip modalpanel true +Im seeing drastic reduction in retransmits and a consequent big improvement in goodput +Based upon your last statement let me just say that I am not sure whether substitution groups will work with with classes generated in type mode +Any thoughts +the related issue is now ll pick up the fix when we move to a problem or at least the original comment is still a problem +Regex support is desirable as well +Fixed findbugs warnings +See the However I have no idea why the validatorhandler seems have no effect +Thanks for the suggestionsI applied Solution Move to srcjava delete dir srcjavaresources delete dir classesresourcesand updated theTed +Moved to please move bugs back to which you will work on in the next weeks +This issue still exists in Ive added screenshots for IE and firefox +Whats bad about using? Whats the implication of not backportingIf youre happy with this being backported and promise it wont break existing configurations +Using max of subchain selectors is pretty added regression testsoriginaloriginal fixing it +It is the intent of the patching feature that it can install a minor or micro version upgrade as a patch +what do you thinkI like it! as long as it only attempts the JNDI stuff if does not require extra libraries for a lightweight embedded version +Attached two files that change to use a random port roughly using the algorithm proposed in the email discussion +Thanksm working on this folks +Thanks Joel +I really appreciate that +Sorry to say my fix didnt handle the thresholdWindow thresholdCount case +Moving to +I just committed this to and trunk thanks Allen and Matt +One way to test this out would be to create a Solr unit test that tries to create a Solr instance on top of HDFS using an +bq +Pls +It works finePhil +ITYM +The fix is to update the test not to assume localhost when comparing the output +Here is another attempt to make it clear Move formatting exception message into a static utility method Option Subclass the well known javahadoop exception subclasses +Any suggestons are welcomed +So in direct mode we have the running no matter what +It probably does need a QA task though +The fields that causes the error is the NAME field on the ACTGEPROPERTIES which imo should not be limited to since can get quite long +Just a wishanyway looking great +The dump shows that the error occurs while buildingtesting the jcrspi module +Initial implementation provided +Okay! Ill be committing this in about hrs or so +Unfortunately it is a really really really hard thing to do correctly +Please see continued conversation on +However i cant really find anything in the spec or the schema +Investigate license issues +Thanks for looking into this +The patch can keep the URI normalized which is required in and allow the a space in the path +Yes its the same installation +Maybe its an ignorant question but can this be a security hole +The handling code is correct +bq +The description of this attribute see in the component template +Igor it was a pleasure to contribute a little bit to Tapestry. +I just committed this +Updated patch has the Realm changed in to fix the test +The only reason I have not already added Communication Platform is because I only added Communication Platform to the download code in HEAD so I did not think we wanted to put it in branch +The above message is triggered manually by me +Fixed by Clement in revision +yes they should all be using absolutely +Its just the configuration is doing not active monitoring of the instance +The scanner concepts are very different in. +Only if all we care about is the JDBC driver +I cannot reproduce the problemI started the app added an image file for upload started the upload +Hi Randall I liked the concept so I took it a step further +I wouldnt recommend to run zookeeper using ramdisk without having an additional method that ensures consistency +Still the fact that it actually changes the converter for further messages doesnt strike me as entirely right I would simply have expected empty output to yield an empty string instead of a null that magically changes the behavior of the component going forward +Chiming in +Duplicate of. +the video sample file to be added in the images folder of the sandbox examples +Attached screenshots from issue test case for and and also screenshots of one part of my application where Im using it so you can imagine for what Im using it Lukas thank you now its all much clear to me +Caleb could you profile some on which I could verify that? Thank to looks ok here verified JBDS L +The code uses a Phaseinterceptor +I will start researching this issue using the CCD class asIve read through the and I have some initial questions Do we want to refactor both the sequential and the version of the buildClusters? How does the outlier pruning relate to this effort if at all? Do we also need to refactor the tests for this class as well Im guessing not since the is abstracted away? refactoring is for clusterData phase and not for buildClusters phase +However despite that confusion what I have described is still an issue +Website including pages contain the following links nowAbout Apache Apache License Sponsorship Thanks Security +I can send you application exe with DB which is around MB as sample is not able to corrupt the same +Ales is not going to be fixed in CDI. +I have a project +Closing as Wont FixI hope to contribute on the solution for the next release. +Applying this to mavens will work somebody changed the issue value at some time +Code to reproduce this s the proposed to add Regression thanks for all the patches +Dont know if it should include output path in tests classpath if it is different than target so close this bug if you think it works as intended +but I still get the exceptionIll go ahead and try to replace the node +Individual builds take a few minutes each on average +I think Adams reasoning is sound about it being a pretty innocuous but important part of the code to have running with native speedups +I dont see a reason why an error has to be displayed +Can you attach a complete stylesheet xml document and any other files we need so we can try to reproduce this? Thanks! +Is this what you are objecting to? Colm + passed +bq Master determines which regions need to be handled +This patch also removes the unused class +I have just committed it to trunk. +Interesting +Miroslav you probably forgot to set a not null relationship +verified in JBT Alphaa +screenshots +The fix seems to work fine closing the issue. +Thanks +Ill resolve it for you as well +Assign to Ales to look at +revision +Woops I fixed first and am having trouble applying this patch I would really appreciate it if you could update the patch against SVN trunk +should I still send you my sample Rainer? I finished it because I didnt see you already closed it +bq +Shall there be an option like vertex and edge +In the entry has to be implemented too and the two things have to be connected via the +I am considering moving filesCompacting into and have a patch but that will be separate +changed to get from zip result before transition all resolved issue that did not see any further comment in the last month to closed status +Instead it uses the current objects values itself +entity class +Trunk has now switched to Commons Lang in revision . +so I suppose that this type of graph would be well also with Findbugs +arg +OK I found the problem +Hi TedThe value of Y is always fixed which is independent of beta +Given that it may be that this doesnt have to be fixed in but can be an that provides the override of CDE thats needed +Thanks Chuan for the patch and Bikas for the review +Error reporting like that is hard in JS +However this meant that we could only do generic exception translation and so were no longer throwing a or +Im sure there is a better way of doing it but it at least works +is mine are yours + +Where are the existing templates? +Both variants are specified within JSF of name and library attributes is new feature +Fix in r +And we havent really decided if we should be using at all in the long term so the cookie might be a moot issue ultimately +When YARN does have additional feature for this we can always change the way the information is pulledIm not a big fan of the artificial delay either but for now I think this is a so that clients can pull information out from the AM specifically counters job state can be handled via the RM +In development branches we might have we might also commit warnings and fix them later etc +Hmm +NET which people some time ago reportedJust for the records Ive never experienced such problems with +As puppet region servers are defaulted to slaves +Looks like a reference type could never be used as a this argumentIm currently trying to write a patch but as this seems a major operation Im not sure Ill succeed. +Assigning over to Tom hes been working on it. +Please test thoroughly I had to change a few things to adapt for the new event handlers +did a svn up and generated the patch again after resolving conflictsZheng can you commit now +Attaching +JayWe cannot change level because messages are emitted by we cant affect them somehow +You change this should belong to POI project +Second version human readable with all changes and split in like separate one will apply with havent actually ran the code of the proposed patch yet but from a conceptual point of view it looks good now as far as the php side goes +Hi Thomaswould you share some sample code with us to reproduce the issueBR are trying hard to reproduce this issue in a small amount of sample code but no success yet +But as the sonar analysis configuration includes the DB password everyone who can see this configuration can also access the source code of all teams +Verified on Ubuntu with +Phil It looks like theres already a in JFFI +Finished build +All test scenarios pass +If theres a significant performance decrease with Id like to suggest the Serialization +If it is still an issue then please reopen. +Pushed to master +you are using which versionrevision of OpenmeetingsDid you import any backup data? And what kind of backup data? What was the source version of the import +Its fixed in by commits for andI am attaching patch which can be applyied to EPP branch to fix the issue +If it suits we might discuss the integration into the apworkflow as a next step +If further discussion is required please use the mailing listsSven +Dory has been successfully patched with the SQL update +I am finally getting off so I should have some time now to devote to xplanner +Test for PASSMDSUM for regressions were found when running the patched version +I am certainly open to enabling new hooks to customize the behavior of these classes if youll describe what youre trying to do +Wont fix since we are going to remove annotation +Well it didnt work for me and I didnt see the Not opening daughter +I got the source code for and will compile it soon +agreed +Evil +Attatched a patch +Likewise the PMC groups. +credentials are stored at the gfac backend and xbaya doesnt have to send those data over the wire back to the gfacLahiru +failures are not relevant with out patch also they are failing +Unable to reproduce using the info given +You can already remove the outlinks from a page in a and change the status of a page +It seems this method is only called from the so also printing it in the master log doesnt add much value +While I originally thought this was a good idea I now have really strong reservations about it and would prefer more discussion before it is implemented +not an ASF community project +Hi Darranas I know content of JIRA Summary field is shown in hover popup when you stay with mouse pointer above issue key for a while +One comment is that we dont always now what the line number is +Thank you for your comments +The cached unmarshaller with its b should be using the current thread context classloader +I think extending it should be fairly straightforward and we can probably finish the integration in a couple of days + +Ill also removing +cant promise anything +The class has been replaced by the interface +Closing. +Attaching the correct file this time. +overloaded to prevent any Codec etc from being used +uploading a patch created from svn committed the patch in r but well leave this open until the specs have been updated +Thanks violas path and I have verified the latest build it works fine. +Were using an enhanced version of that now supports transactions and are deprecating our own jdbc connector code. +Edgar say that could be something more related to a Pivot decorator and not specific of so it would be better to take a look in +Attachment has been added with description Precalculates string buffer size to improve s a patch that improves the efficiency of constant pool generation. +No longer a problem in trunk. +Confirming problem resolved +Adding patch for the issue +Committed revision . +Yeah works like a breeze +However they are not retried +I dont really have a need for more than a single lib dir ever pretty much +Likewise for +Thanks JonasThis is a nice improvement +Sounds good to me +and List types should now be supported in the CVS version +Hi Luke Thanks for the patch +for trunk and I can add doc to book from script head once committed +So if the user specifies some order in the schema then Avro will make sure it loads all data using the same + for patch +It also affects + +Seems to be an issue of WTP +Attached diff file +Also fixing a Javadoc cp error in is and is +I cant see any sane reason to have two versions of the same client +Scott the rpctoken fragment is being added by Shindig and its used for quite a few things +Many thanks for this enhancement + + +At least it forced me to write integration tests for the plugin +Is it just me or do local docs not allow attachments +The submitted fix ist just a not working workaround +Streaming tutorial Tutorial Tutorial patch +Carl Can you hold off for just day? I will definitely take a look today +Also I did a svn update on my checkout to verify whether the patch has been broken by a recent checkin +is already done +This basically forces the existence of an interface for any ejb +Should this issue be closed +Bulk close for +sorry for all this trouble people +Bridge + +is a part from apacheds rolling log creatred during session that imports ldif file +Rebased and merged into the master branch. +The patch address the issue +This issue has already been fixed in +EnjoyMike +Any hints on workaroundsI currently create the database manually +will you be able to send a patch with the current axis trunkinstead of setting default values can you think of setting checking the defaultValue when setting the object values. +The fix is not complicated +That can be done using the standard recipe for leader election using forgive me if my questions are trivial What I mean is the election of the leader can be handled by ZK but what happens if the threadprocessmachine with the coordinator goes down? Does every consumer group listen on the path in ZK and issue a new round of coordinator election ala how itll be done for replication leaders +My use case is wanting to host recordings of a series of Hibernate developer discussions from our recent meeting +Cest la vie +Presuming it is that simple I have applied a patch to do thisCaveats You should apply the patch from first since it adds proxy support +On a side note I have unpleasant feeling that some real problems in TM are just being hacked around +I have committed this +Fix applied and tested in trunk. + +I also feel if this patch is fine we can commit the patch to trunk also and work on and to solve such scenarios mentioned in the above comment +First pass last are filtered out +ill start to take a look again +The root cause of this problem is that the pageflow runtime isnt using the which has all the code to do the event handler wireup +This should work properly in all cases avoiding the potential +Let me know if you are interested. +Can you point some of those toolsscripts +That is the right branch +Therefore we have no way of knowing by default whether the developer wants to capture a time or simply a date +Very cool +We are reseeking to the start of the passed row? Is that what we want? I thought we were trying to go to start of next rowIf reseeking the start of passed in row then this patch looks fineI defer to Larss opinion though +We use it in our monitoring web console +The change of maximum block invalidation commands per heartbeat is for Konstantins point above +Validation now includes XML Schemas +Hey very nice thank you +Its legacyThe code and doc is there its in Apache Excalibur and we can work with it +Any remaining issues with this should have been resolved with changes made available in the nightly build +Please update the examples start up with development and go to httplocalhostajaxwicketbookmarkablePage and click the exception link +The errors are causingt he numbers to go up +duplicate of appears to be related to which hasnt been addressed +Looks good +Uploading a new patch with a small new test and a fix +added to livesites thanks +What keeps conflicting every timeIll make a new patch +There were some apparently forced line breaks in the patch that I needed to fix manually before the patch applied cleanly. +Moving to +New rev of the patch addresses Aarons comments above +Fixed in Rev +If you specify the version for to will get your project workingThis issue is about filtering on a single line +A regular update of the HTML files was needed. +To execute the cppunit test useant Dcppunittrue target runs the autoreconf and executes the testPlease reviewtnx +Updated patch for branches which includes the latest updates made to trunk in r +Looks like our verification frameworkSlava what we can do here +ICLA signed and sent off +Strictly speaking this might qualify as both generic and extendable but probably very painful as well +Hi Brandon since there is no new tests could you manually test it +Ive pushed some changes for this in that cover most of the codebase +OK will add quirk +This property is recommended for all major features other features may omit it +Will be resolved by . +Basically we keep a special counter on attributeTypes and objectClasses called +regression tests +But we should first try noIm not convinced that your really handles all of the cases that might ariseits pseudo code that hasnt been tested yet +Just like that we wont mention that scheme and authority are case insensitiveFor delegation lets leave it for the moment +Thank you! +The only thing wrong with the headers is that they contain the copyright notice +Ok I removed the exception from the logged informationMy original confusion was that I thought you were saying that Felix was printing the information but it doesnt since Felix default logger doesnt print stack traces for debug logsI am still not convinced that this isnt a lack of configurability in the log service impl but I removed the exception since it probably doesnt provide much value any wayAlso make sure your logger doesnt make calls back into Felix since we witnessed deadlocking which is why we disabled that by defaultPlease close this issue if you are satisfied +Maybe tommorow well have the CheyenneIve asked my members about the logo +and jars uploaded for Spring +in this case its the FTP Notifier +the problem is that waitFlush was removed but its still referenced in the format string used by this clientim going to try and get a fix for this into fix was trivial but i spent over an hour trying ot figure out how to modify to reliably detect the failure in its generic way before giving up and writing a test for thisstill doing some more exhaustive testing and then ill commit and revision +it seems odd to me that one apparently cannot use a package level constructor but one can use a protected level constructor +Resolving. +Maybe you were referring to +Currently in the README it saysAlso insure thatit should be Also ensure that +For example if the configuration has defined but not then will be ignored +Still blowing up on branch and presumably master as well +When opening the all projects are loaded except thoseemhythrinterpreterjitrinovmcorevmiand those projects were originally on svn +bq +It is quite obviously an out of memory error +committer should just the first then apply the patchNote testing I did was done in conjunction with all jira referenced in +If any encoding problem happens then it is already in the string +Not fixed in AS console for EAP GA +I actually have no idea why we dont support raster legends in wms decorationsDavid and Simone should be able to provide in not sure when it got fixed though +After iterations IE is using over K of memory +I was surprised to find out that the maxlo was a multiplier not a starting value +older issues for Apache since were no longer actively working on these at the moment. +This is the first rev of a fix for the problem +How do I assign this to myself? Or am I unable to +It has some affinity to the class +bq +We are also looking at digesting roo script and outputting xmi for it +Ill try to patch in your latest fix +Wrap and to include the required params to make orchestra work +MB still but with MB Im able to scan a M rows table whatever the CACHE value Im using +I will test all plugin updates before I commit it +Once its perfected a bit more Ill integrate into trunkE +If this is okay I can make this change in my next patchPlease reviewcommit this patch +This is an important problem but it doesnt apply +On second inspection it appears that s output is actually more correct than MRIs +The change made by sometimes causes the main write thread not able to exit therefore not releasing the resources it holds causing too many open files error +What about configs for clients +Hope its not too late to sneak in some more ideas +Do you think you could also update the docsSomething liketype The FQCN of this class type The component type name needs to be http +Also I do not have an Internet Explorer browser to test with so I am not sure if this will work with IE +Create a duplicate view to list all packages files methods involving the duplicated code +Grub is showing up and machine is freezing after that with blank screen and high CPU load +I dont think I have ever run this test +Its a good idea to do a mvn clean first +This has been completed. +Hi PieterThanks for detecting this issue +the better way is to set flushed to false after you add something to entry logger noI dont want to notify the offset until we really need it +Ive attached the diff between the current trunk and p +Unfortunately this happens for shapefiles too +Can you describe any manual testing youve done of the script etc +Not sure why eclipse variables is needed for that at all +catd out for my AMbinbashexport YARNHOMEtmpbin +Confirmed on Fedora here too +We should come back later and reexamine repo order +Im using The Flex plugin error was actually due to a from the pluginsNested classes are not allowed in so the plugin doesnt need to support this +bq +Needs to be verified with the last snapshot +it is probably not harmful but Im not sure if it would also be a best practice not to use is the use case you have? If the executions are structured appropriately I would think the mojo just runs or not and the configuration can be used to alter behaviour +bq +this isnt a blocker and isnt broken anymore. +Verified on EAP CP CR. +It might be better to have the requests as a of ftplet. +Thanks I see whats going on +created and added to CVSNeeds more work really though +ok pending test results +to check if already known filters are DanielBut I have applied only your patch for the unit test and it doesnt failId like to have a junit which failed without the other patch and successed with the itTo ensure the bug is fixed +We dont plan on doing any more work on Seam affects added a condition if a component is the same name and class not throwing exception otherwise for component with the same name and different class the exception will be thrown +I generated files this wayscomp compiler Cbin jar Cbin mx m src src d classes debug verbose out is interesting it seems that allows two enumeration values that are identical of course Java does not so we have to check for this case +Let me know what you think +This is because the failsafe plugin is better equipped to handle the use case. +I verified mvn install +Affects as version of Tomcat shipped with this version of AS is not +I saw it in my source build and I thought I was looking at the binary distro. +Ive added a flag to enable the async indexer in rev +Tom I will take you looks good as a +this record since is now released. +Ive also verified that it works as well in and dont know if this is specific to I think it may have occurred in as well +Thanks +Shipping ahocorasick in the asSource for Flume is problematic as that is being shipped under anApache License and the licenses for the entire project would conflictMy recommendations is ship the binaries as a Jar in the lib directoryand put the license for it in the lib directory along with it +Fixed in trunkCommitted revision +example code has been checked into HEAD to the tonr application the demos how to handle this case +So no problem there. +I believe there is a JIRA for this +Issue resolved as all requested changes made and checked by independent party +Hibernates Criteria API does not provide a means for cloning or for resetting so we cant prevent against this in Springs either +Do not save leases into the image +But I dont know the date when it will be released. +Do you mean that its a PMD bug? Sonar API should avoid creating several times the same violation on the same line +Lemme look at this and see what the issue is +so this would be fixed when upgrade to wtp work with the next build since we upgrade to in +wiki and doc redirect through a proxy and end up on the vmNote that the docs wiki is already receiving spamOk to close this issue now +we can leave javax +Yup Committed! +I should be able to rebase this patch next week. +Patch and extra test file it version of the patch with custom metadata! Patch applied in revision PS +So unfortunately this situation will be going on for a while +I hope this fixes the revision Thanks Fergus and Noble + allows to use stretchable markers as the plate so this specific use case is served by using a square closing all issues that have been in resolved state for months or more without any feedback or update +HTML attributes are optional now with behavior +Changed to useAdded a comment about using in test environment onlyCommented out servlet defs and mappings in comments moved the to a sample package under the server tree. +Much of the Windows user world is still in XP land +As Brett suggested Im just going to make RELEASE as the default recommendationVersion so that it will be returned for conflicting +Ill probably take asjs external and falcon tonight +OK +Attachment has been added with description bug fix for fixed in SVN +Looks fine +Splitting into arbitrary parts so far required multiple passes over input data using the approach of teefilter codecs its possible to do this in one pass over the input data +Duplicate of. +We are using more than session beans in more than a dozend ejb modules +Attaching a patch that fixes the problem in these two topics with the two output files +I was just following the spec for Retrieve Items as that is the functionality I needed +If you dont already have something ready could you can assign this to me? This fix also depends on compiled specifically against hadoop +thus no real added value in the current implementation +name +Closing issue this has been languishing for some time the problem seems to bedue to the fact that the group is trying to recompile codec withdifferent Ant options +I generated C code with the modified version of jute and verified it fixed all the compilation errors +message is still displayed when an upload is with r +Additionally a test should be added to ensure changes are written to the file +I dont have contributor access to Thrift so I cant assign the bug to myself or mark state as patch available +This works items for early release of Java Eclipse Mgt Console +Sorry for the blocker but we will soon decide whether logj will be dead for the next few years or not. +At the classpath is determined by +Its unclear to me if theres a better idiomatic Scala approach to the class composition of the choosers +Mark can you take a look and see how badly I broke the Windows build with this patch? I updated but I dont know if the build system actually loads the correct paths into the +This issue is no longer occurring +I dont think recovery mode is the sort of thing that well be putting into scripts any time soonAnyway if anyone else has any input on this please post +If copyLarge should not accept a buffer size parametermaybe we should use the copy method +in progress on +Simple cases worked well but needs more study +r implemented for Native XIm going to write some additional test cases then close the issue assuming both backends pass them +Cancelling patch patch did not contain change to as well +this method is called when parsing the string in typefredjoe so we end up with an extra fred and joe when this is called +Is the above change intentionalNot at all! Updated version attached +But I do get it when connecting to a RHEL boxHas anyone gotten around this? I would imagine this would affect a lot of people +Hi JulienIt does not take +OK checked in related code and ran unit tests to confirm they still work +However the underlying problem remains The Oracle driver does not accept its own Connection handle passed into its own method +Ok then Pete + for mispelled pathIn of course I use the following path ormilocalhost homeetjavaopenejbhomeetjavaopenejbconf NaderIf you want to attach a small sample eclipse project that we could try wed be happy to workout these issue and get this running smoothly for better if youd be ok donating small sample we could check it in under our examples directory so others could see how its! have created a small demo eclipse project which was used under September demo compares POJO and EJB development +Merged +add new test case for this problem +Does my explanation above make sense +Fixed in CVS should be available in tomorrows builds +NMR upgraded to woodstox +The attached patch removes all explicit specification from public instead package names are retrieved either from the schema or from a Java class +Thank you very much for your assistance +This allows for replicated caches to be configured with async comms cos the possibilities of landing on a different node to work on a key are reduced +Cloudscape probably allowed you to specify conglomerate properties in CREATE TABLEINDEXDo the manuals talk about other constructs that were supported by Cloudscape but not by Derby? If not should we simply remove that sentence +I think we should close this bug and create a new one for the not supported param problem and attach the code snippets to s reopen the original bug +Snjezana could you look into how we could do this best +I think we need to open a separate bug report on it +Any suggestions +I plan to commit that last patch soon +Lets evolve this a bit in and dump a patch here laterTip When uploading patches give them the same name every time +Simplified the code +The issue is resolved +Changing that to a query parameter would have quite some impact across the ASFIt looks like this bug has been fixed in Safari so I think we can close it again now +Fixes this NPE allowing deployment to patch fixes this allowing deployment to proceed +updated directory to point to removed reference to examplerewrote slightly to make it is no longer included + should be modified anyway the nasty regexp in there should be replaced by something faster +modifiers method processing etc +Committed to trunk. +So what does this mean for us as customer? Will the requirement be addressed? If so in which release? I need to publish the answer in our internal tracking tool +Pre it was assumed that all the trackers will atleast report once before a job gets removed from the system +I had to tweak it because I was getting all sorts of java dependency files in + +I dont understand what you mean can you elaborate +We arent about to start writing Seam in python or something +bq +Might be worth considering +Its not that I doubt that theres in issue with the code +When I take that SQL and paste it into Query Browser and execute it it runs without error +If JHLA does analysis across set of MR jobs over a given time range it can be added as another offline analysis tool extending vaidya framework and possibly some of the existing code +Hmmm this is something that needs looking into +That would mean we would have to store the executor instance as static field of the base class and hence we would have to make sure that all the deriving classes of this base class are not run in concurrent threads in the same JVMWDYT +If you have other problems wich are not related to the situation reported by Vijay I do Agree with you that my code is not suitable for your needsTaking a first look at your code I could not fully understand your problem but Id recommend that you do not generate nested combos with the same ID as you are doing since if you would reuse my code anyhow you couldnt get the proper nested comboIf you whish me to have a closer look at your problem please attach a runnable example w source of your problem or post the complete code like Vijay and also the stack trace of the errorKind Regards +is a long standing bug how is that Bigtop never hit it before? What am I missing here +Still potentially painful and ugly +Hope it worksregards +Attaching +The release audit refers to files containing test cases which are in +Daniel please verify +passes all unit tests locally +And for answering the question about the schema +It is included globally +thanks MarkOne of the lessons Ive learned from Damien is that to get people involved in a project you have to be willing to let work go undone so new people have a place to jump inThat said implementing this would be a good way to learn hacking on for anyone with an interest in itIf you are interested leave a comment here and we can maybe give you guidance on how to proceed +I have added the jclouds and spy repository to nexus and they are now included in the group URL +Applied the patch +Setting component +I will do that when I will get a chance +But irrespective of that if i fetch an entity at a revision there will always be a row version mismatch and I will not be able to save the modified object obtained at that revision +This would make developing a new compression scheme hardOut of scope for this issue +commit in logic based on more manual testing +I did also file that the client doesnt gracefully block +In fact I wonder why we cant use for token renewal +The KIND defined in should come from the in HDFS The place where the delegation tokens are registered for renewal is not correct +fixed code in svn r +Thanks +INFO Running for rows took ms +Gabriel sorry I didnt acknowledge your comments above but I have already integrated them into my patch while I was making the configuration changes suggested by Justin and Andrea +Tom could you please take care of this +The attached patch for is ported from patch for +We experienced a similar issue when Diskeeper was defragmenting a drive that contained an index that was being updated +I applied this patch +distributed cache should be mapred +Its working with the latest code from trunk on my machine +ha no longer presettings +Obtaining an for a class +I think writing such a test is a little bit of out scope for this JIRA but its definitely something we should do in the future. +Ill open a new jira for the above discussion. +Be sure to put those into the appropriate folders Make sure that the +Sure Ill go ahead with your suggestions and let you know how it goes +By including src entries to the referenced plugins this results in an m beginning to think it would be useful to provide a flag in the configuration to better handle the case +Thanks! +The bug in in the JXR library +Not sure if Im missing something here though +Unfortunately I dont have time to investigate this much. +Committed. +archive +So why not just fold all of this into the ARH? Seems like all of these features work just as well as input parameters and there is no need for deprecation etc +A timely quote for what I take to be the basic sense of doubtfulSeptember th percent of US citizens doubtful about official version. +At the face of it this seems like a cyclic lock order with a potential to create a deadlock +Have to pass null for the hostname property since hostname can not be obtained from a component resource directly +Attached patch which implements the changes described in the issue description +I think we can not change this behavior easily +By the way your test is precious as I could see a very similar stacktrace but I was unable to reproduce it yet thanks a lotLooking at your thread dump it seems to me that after the NPE error the cachemanager fails to stop Hibernate Search or is Search failing to stop also when you dont see the error +I know there is a ton of other useful security related info we could output +Ha +Some additional comments +Please feel free to review +Is that the preferred approachWill post a patch addressing the other comments +Apart from that I have no preference to one of the proposals just please dont silently ignore config errors +Closing as cannot reproduce until we get user input. +Actually ORO should have been fine since both and are shipped in the system dir +I knew whats going on +Erik using simple string substitution always sounds like a good idea but Ive never found it to beIll post some benchmarks of REXML vs +patch has been checked into trunk alsoGeorge +Could you post any design thoughts about how the ant script might interact with Sonar or possibly what conventions the Ant script must follow in order for the planned interaction to take place? In light of this new feature were going to evaluate our legacy build scripts +Regarding memory If Im not misunderstanding things you also have to materialize all field values youre sorting on for each doc in each PQ +The proposed solution is not working in since this feature is broken +iocorecluster +What oracle driver are you using + +We have a similar issue with property files in our Ivy environment +See more details in all stack traces you posted in the issue description related to and there doesnt seem meclipse code involved there so unless there is something else need to be done here I think we can mark this issue as a duplicate of as a workaround for init java tools issue you could try to run Maven disable dependency management and then Maven Enable dependency management actions from the popup menu on broken projects +Vinay thanks +dont forget to add the include path to the build line BLACKTIEBINDIRbin +Also need to change how we read so we read the whole block in rather than piecemeal it as tfile currently does +I think this is a bug not a feature so tied it on +This is because Server Admin cannot resolve mappings for His Kontext only the webadmin can +superseded +Latest patch review item has also been updated +Tested +Stack trace for Windows m slightly puzzled by what the stack trace has to do with eggs +thanks HenryCommitted revision . +The only real usecase I can think of is being able to say copy all of these user libraries into this dir +On one of the clusters a map attempt was expired as a lost task in thread but it was not removed from taskid +Ryan can you see if this reproduces on +meta +Main method is on please comment here if you need any more information +Ben will update forrest docs to have small icon +Just tested with Safari works too +Im leaving this open as it is a bug however Im not planning on fixing it at this point +I think we need to make sure as part of our tests that upgrade works and we dont have any compatibility errors when running metadata tests +Your workaround works for my projects +wasnt needed to backport this to Branch as the logic is simpler on resolved issues. +properties file no more error prone than that +Mladen Im not sure if this is something that will be in EWS can you please make sure this will be fixed in EWS if the same code is usedAs theres no further release of EWS planned feel free to fix that in EWS if time permits +Verified by Ilya. +We encourage adopting a fix at the earliest opportunityIs there any workaround that we can employ when debugging in existing environments +OK I moved them to +Thanks for the patch Naryanan +Why would you reopen itcreate? Upgrade to Guice is done in wicket +Have you tried the new Roller RSS feed with +Satish have you fixed it already? Send a patch on review board or let me know +this patch includes tests and fixes issues pointed by ben +Jeff assigned to you to add the runtime install details when worked out and the installrunning samples document is located at cdldocsconversationConversational ESB User task should be finished now and now we use jdocbook to write the user guide and samples guide. +Which version is this with +This is very helpful and insightful +does such a fork makes sense for the compile goal as well the code is essentially there since calling the of the gwtcompiler was the first implementation of that plugin +And finally does this represent a common need or a very rare need +In which version are you going to roll out itYes for us this is necessary feature +Still need to talk with Peter about the modules organization and know the impacts about having a single module for each seems that we can remove the build modules and configure the picketlink with the assembly configuration +Superceded. +Patch for source +Those caches are the key to performance and if you are trying to overburden the cores with work then its the cache miss penalty that kills the jobsMR dodges this by having most tasks regularly blocking for IO operations but other workloads have different characteristicsAlso as Timothy Points out noone is going to be releasing with less cores on them the will only increase +The patch looks good +Please see for more info +see +It did not pick up my previous for a long time +file changes also ran with default to check if any other newly added tests need to be updated +That issue will not affect next release because it going do not use prototype at all +Committed thanks Avery for the review! +At best it can indicate the start of a stack trace which is in my experience not particularly useful for any but the most trivial of issuesIf we start using error codes to indicate code points well quickly churn through hundreds of them every time some piece of code gets refactored since the refactoring means the old code point no longer applies and thus the error code doesnt either +Any chance we can have these changes as a patch against trunk without reformattingD Thanks +Mithun Are you going to make this jira patch available? I think we should commit this patch soon +The patch looks good +I have a prototype of this in my portlet branch and Ill backport a cleaner version to the Trinidad trunk hopefully today +Yes your simple is looking like fine + +Now how does the user express his desire for lenient name validation? The user may not have access to the source code that invokes the parser and so accepting a flag in the parse function may not work +For the record to reproduce this I ran the attached thusjava XX cp CLASSPATH +In this case comteradatacrm is in a different Ben i havent done much to try and confirm that this issue is same as +to jump to the selected controller method +does not matter i dont want stuff in eclipse plugins that could just as well be in the core +patch and unit try please new patch and testA lot of thanks to Andrew Zhang for pointing fix incompletness out +Bulk Close for all issues before incubation +Original bug is +will ship with your provide a patch for? Or instructions on how to make the change on our endThis is a blocker for production +Since we no longer bundle them in which versions we use doesnt matter +Added a proposed patch to also see forum reference +I pulled this into +Can you please open a new JIRA and post the entire stacktrace so that we can inspect it? +From my point of view this makes things more complicated why not just use the extension id where would be the advantage of aliasesMay the aliases would more human readable but in the end you have to define the aliases anyway and need to lookup the extension ids +see also fix of course requires that we add a artifact to central +Eric Thanks will try it out to see if everything is working secureunsecure +Sent a mail to the list +Good point +Please let me know if something is not correctThanksPlease remember that if someone of you has a custom application that extends order and ecommerce application hat to replace the following labels with the new ones +version to be used in the descriptor +Attached patch derived from Dougs Chucks patches from the thread +Im in progress of wiring up the process of identifying validators based on the configuration +Thanks Amareshwari +bq +Accidental clone of +Do we know why its adding every class in the archive +Once we agree that the code looks good we can write some additional regression tests for this problemTouches the following fileM javaengineorgapachederbyimplio +I go in and my password and the build works +Those using the handlermappingadapter can set the flag to false explicitly if they need the old behavior +I dont think anyone in the Derby development community uses Maven but if the necessary files and explicit instructions were contributed deployment to the Maven repository could be added to the Derby release process +Thanks for updating those github pull requests +Thanks for the patch! +Thanks for the quick turnaround Juergen +Thanks for the work around +There is a unique name check ahead of the option check which overrides the option +Without further investigation I cant simply make depend on runtime +No it looks like youre right +add secondary comparison on real key value after hash compare in RP +mail consumer commitrollback logc do not allow handover to ensure running on same thread as polling to use the same mail session. +files and directories that havent been accessed in months +I originally included this only because I was too lazy to write up an entire for the entire server +Some simple changes to the test cases to properly set up the when creating the documents and removing the unnecessary ctorThis patch is dependent on the contents of Nicks patch for which fixes most of the failures in +Also working on another issue that I wanted to verify that need to run quorum using config +Given the number of things that have gone wrong with loading this file so far I think it might warrant this idea +The patch +OK +I havent checked to see if this is still a problem but since no action has been taken it probably is still a problem +Note One possible scenario to consider would be when the out of transactionlazy loading needs to take place in a different JVM for example when an the master object to a servlet in a different container or physical server +Im also attaching a few more logs with different stack traces not sure if its all the same race or not +I was working on some other issues recently so I hadnt time to test it again +Attached This is an early version of revision c + +actually I fixed this as part of my recent changes to the metadata handling +working on fixing C codegenruntime support for implicit casting of a struct to an interface +I have committed this +Heres a version of the patch that also fixes things for repeated enum fields +LGTM this to trunk and existed before +I make some more tests with our tools before submitting the patch againIs there a unitary test that can show the bug you mention +We have similar hooks at Master side while table creation + +Ill close the original jira as a dup of this one since the work occurred against this duplicate +I could submit a quick fix patch that unmarks synchronized but I dont really know if that would break other things or if it could miss other deadlock pathsAnybody else know more about this code +Each incoming message is processed in a separate thread provided by the +Committed in! Yes I think we should remove then in the next major release after. +Actually I am not sure whether a FAQ entry issufficient +I just noticed that revision that contained changes relevant to this issue was incorrectly labelled with instead of and that I therefore missed merging it to the branch for the release +James could you whip up a patch for this? Thanks +Cannot reproduce this in revision Mikhail +Thanks +Color issues are related to the branding that is not currently applied to Fuse HQ the colors from the table +Reopening and deferring +Thanks for the heads upFixed on rev fixed +We can always do the egie swap later if this issue crops up again +Can we keep the github mirror +with the javadoc on the method +Not yet +Thus this is the expected behavior +Could you post a code snippet +All of the feedback Ive heard is positive +To return to the original issue can you confirm whether the problem is resolved +Checked in an integration test in revision . +True normally we dont store the zip in maven +btw +We are now using codehaus jira as our central issue tracking repository +This was reported before but since I dont see it listed in the thread extends should not be unbounded +will be included in +However this is difficult to solve given the current framework +Hey guys this is exactly my use case +Can you describe the situation where you are missing the date picker? Just in case it is something that we have missed +Looks like this wasnt the whole storyHama INFO RESULTS INFO INFO INFO INFO INFO INFO INFO Hama INFO RESULTS INFO INFO INFO INFO INFO INFO INFO +This will get the CI server back up and building and we can look at it we want to go to using boost uuid within all generators rather than just the delphi one +See Fix field +is closed for both core and contrib +Also I will add the heapdump dir to the artifacts +Closing them out is much easier then having to research the two or three dupes a ticket has or trying to figure out when the issue was resolved +Documentation updated +However Im sure theres a way to resolve the issue? Ah yes might be helpful to note this only seems to happen with one project so far +I probably should have commented on instead and it looks like thats where adding the checkbox would do some good +Heres what I know about message levels in generalDEBUG intended for developers of a library +And that we reserve full future rights to break back compat change the change semantics of the methods remove the class entirelyEG Ive done this withIn this particular case Im inclined to simply make an exception to back compat and simply remove the class for +Logged In YES useridThis has been working since at least the release so Idont know what version you are referring to here +Will come up a temporary solution for this as it is related to some other aspects such as how the applications are kept in Geronimo etcThe idea is that a +The issue in the forum is harder to replicate since was fixed +Thanks for your feedback Adrian I appreciate itThe caches arent intended to be singleton sources which is what youre saying is the problem +Is this not true? I ask for my own education +Its testcase source code for that bug +Thanks for the good catch +Sorry for the delay was out of town + +If it is not true please let me know where this is done and assign back to me +explains the idea correctlyThis means that an operand in the code may be used at a point where it is yet undefined +Any suggestions on which one in JBT to use? Or just use an Eclipse default one +Is there a workaround? Not being able to run the mapreduce client tests is frustrating +Ive provide a review board link shortly +The problem is that if DST transition occurs then theres an interval of invalid wall time +See what Hudson thinks +This and a few other issues have since been resolved and another release candidate shall be prepared. +Itd be cool to have a setting in or pom saying which level of verification poms you want to redownload all converted +Running by Hudson again just in test is a known issue +Patches added +SureshThe addition was just a log line and I ran the compile locally before committing +Attachment contains patch +Test program to show the problemPass a URL to a Solr server as first arg and a PDF file as secondThen search for id and and see the difference in stringtxtstringtxt between the network dumps which shows difference between and ve downloaded from trunk and recompiled the tree and verified that the fix solves the problem +Actually I dont know much about ICUJ yet but it seems its worth doing it then +Fixed Spring dep +I dontt really care about the name however the code shares significant synergies with the Java escape so I would keep it in that class +New patch moving fixed block and hardwiring it to For int block encodingStill need to do payloads +Actauly I have done it +I feel this kind of sleep based approach in tests causes them to be brittle and prone to intermittent failures +Were on Java EE and we were hoping to get rid of the transitive Spring dependency by replacing with +Issue may already be resolved may be invalid or may never be fixed +JukkaI my case the problem were the newlines of the multiline description +Works fine in +Thanks Ill think about it and post some comments there +Since this upgrade was hidden away in the underlying implementation of the test case I think were still OK +Not a Shiro bug +I think a key reason not many people have looked at this yet is that the example is not setup for it I really think if we are going to support it we need to move the example and tests to it and force peopledevs to see it +Applied conceptual changes from patch +should probably support resources particularly if the symlink overlay is enabled Yes I was thinking about this +Let me check this out today +Marking this for looks good to me +If a user decides to use one of these implementations then only majority quorums will be available Mahadev suggested to change the LASTZXID message to FOLLOWERINFO and send both the lastzxid and the server id in the same message +Attaching the in Andreas Your point is correct +This is definitely a limitation on jnlp dependency generation a fix will be greatly improve the flexibility +Hello Stefancould this issue affect MD notifications between and also +NOSASLNONE transport layer test failing with latest trunk code +No new rubyspecs failures as stacks also were compared visually + on website solved +Looks like this issue is fixed. +Yes it is not accounting for isnt accounting for proxies is there a workaround and is this intended to be fixed? the patch part of the comment comes from linked +yes so what are those +Looks good thanks Tim +Also fixed the last javac warningThanks Konstantin +We could try and protect a user from these silly situations everywhere +Applied patch to my +A fix I made for another bug last week caused this and a number of other compile errorswarnings to be masked +The attachment contains the missing files +Ideally wed update the to check the configuration and tell the user that ant deployant runant runtest will not work for the minimal configuration +Another option is to modify the hadoop to something smaller than the current value +Checking it out now +This would be a great change of behavior + +Its useful to let the clients know +After fix transactions doesnt work at allIm not familiar with Jackrabbit code but it looks that end method is not good place to clear because it is stil needed by commitrollback methods +I would double check that you have your classpaths set up right too +There was just one minor issue for which we already have a workaround. +Ive refined the log message accordinglyAs for the log level those checks do work out fine in scenarios where an advisor is obviously only meant to apply to beans other than the ones that it depends on itself +As you just submit runnables +If I read this right Jeff youre done here? at least for purposes? +componentssrctestjavaorgapachecamelcomponentrmi +set a global targetJdk to workaround a bug in the maven pmd it in there could cause confusion +I didnt see any good reason for why should have a reference to +Running patch v through hudson +tests pass for me +apologies it should say while execing sql with parentheses around the stmt +the problem with making the default to be joining is that it will mark the server upAh I see +There is a thread dump on a testcase you will need several threads loading and instantiating several aspects using that has been written to reproduce the deadlock. +OK in this case is better place to go +Yes the issue remains in issue still exist withIs there any workaround to avoid this very old issue +It misses the connection management +Otis can you finish commit this +Guido I aded you as a contributor and have assigned this issue to you +Other than that test the regression tests ran cleanly for me +With your current patch we dont need to explicitly store the infoPort anymore +This patch adds support for Delta CAS serialization of modified multivalued feature +The comments are not all confusing and there is no value in this issue until we have reached a conclusion on the mailing list + is even though partition is dont think the ejb threads have anything to do with this problem since the pool test war doesnt use ejbs +Please review this patchThis patch uses a different flag UNSPECIFIED to bypass running the test +Acting on that configuration and producing a set of ports seems like a operationbq +Thanks issue is fixed and released as part of release +It confused me today for quite some time especially I learned the pig for a short of time. +Im finally able to reliably reproduce this one should have a fix shortly +Please if theres a replacement API that should be documented. +This patch doesnt include unit tests since its just adopting to new logging api +Could you attach a code snippet to show how you were using in to encrypt with aColmMarking as wont fix due to lack of information. +alexy was this the same behavior you had ine xadel studio? there were also yellow warnings or +strip the relations to +Over in Ram determined that this always happens when the RS we abort does not carry META +Back in November we had a node cluster of EC instances +Ah sorry +Mostly looks good +I removed that and now it all works fineIll put in a pull request for the fix +Yeah ought to be fixed as Camel is pulling in ruby using the and its now using +We justhave to clear the hashmap with the authentication attempts prior to executingthe request redirect +Actuallythis is not a bug in +Btw the can be run with mvn perthread true testto test this issue +I plan to commit soon +Hey Libor +Is this a firefox issue or a gwt issue +resolved merge conflict for +It allows reads without having to copy the workspace +Well none of the examples you have given are even compiling with +Note the issue is resolved in trunk via the mavenization +Hi StuartStephenThanks for the detailed explanationI will try to run the code as suggested by you +This isnt quite right +What if Java is not installed on C but on Z? Your method seems to me like trying to kill a fly with a bazooka + in +I looked at the stack trace nothing obvious +It this functionality already exist I apologize for my ignoranceThanksTom +I really wish it helps because I know this topic is something Hibernate Team has always wanted to deal with and I also would like to see it solved as a potential dialect provider +Thanks +I am using the existing +From section of the specRemoving a node is considered an alteration of its parent +It seems like there is already one out there and that wont introduce an option that isnt forward compatible +Thanks +bp files +If is added to classpath then project runs without I think it has to be fixed in way it will add these jars to Drools Runtime + contains the code to implement this change +Is it normal +From now Jackrabbit dependencies are used only in the unit tests +I added in a couple of assert calls where it seemed appropriate +Patch available see there are reason why the method does not also copy component attributes? Maybe we could just outright copy all of them +Ive attached an incremental improvement for the error reporting as patch a and committed it to trunk with revision I dont know how hard it it to use the programmatic but I do know that using the command line tools is +merged +The trivial patch fixing the is trivial patch avoiding a crash and fixing an inmemory binding spec implementation bug +We also need to modify to expose injection point for this strategy +Added all the functionality to make factories work as expected. +Reproducible case is in +I was unable to adiquately depict the same view that I am intending hence the lack of some features within the mockup +I guess this is to do with the validating parser we are using +Thanks Steve for this patch. +Erick should this be pushed to +Hi MarkSince Java is just around the corner we will have customers attempting to use Java with their Entity classes +fixed in issue after bundle release. +Im open to alternatives provided the build works out of the box for anyone trying to build on or bits without any manual tweaking +Thanks Max +Now that work is well underway with AS all previous community releases are +I can provide a local. +Thanks for the clarification for now I think its best to modify the mirror settings to ignore branchestags for the whiteboard area +In the testcase you could define your values in unescaped format and use to escape it +is next version that would be delivered +See also the subclasses with standard deviation +Updated openejb patch which fixed compile error in +This has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Jyotsna Rathore Committed your patch in svn rev +Its subjective +I am returning to duty today I made the patch with the trunk +Regarding bulkdocs response timeoutperhaps adding an heartbeat option like we do for changesSounds simpler than supplying explicit timeout values for the ibrowse connections +Customer code is expected to synchronize access to those objects +So that comments referring to earlier patches still make sense +What help can i extend to get this fixed +Instead I added logic to skip them if running on Windows +This one is correct. +comments on according to comments on phabricator +closing this since its a dupe +So we should Change the no good message to something a tad more descriptive like assignment of region is invalid Enumerate the known ways that a can fail to serve a region trap those problems and figure out what responses wed like to give to those events +Weinan can you check on the status of this? Should it be closed +Even with the base configuration model in place its impossible to track individual changes since everything is updated in one big fell swoop +the fix for this is now committed +An updated patch for earlier version of hadoop +Just changes the default setting for two parameters +New constructors for to provide all components used to create a default Mapper chain +The failed tests are not related +please wait a little bit +The autoform information is clearly visible in the screenshot +If has to be a certain way to maintain the wire protocol that seems to me to be a respectable reason to make that class private and make a new public class that can be transformed into a properly behind the scenes + +Thanks for your patch +Sending containersrcmainjavaorgapacheopenejbconfigSending containersrctestjavaorgapacheopenejbconfigTransmitting file dataCommitted revision . +I added the accept headers everywhere I could findr +My diff tool doesnt include large binary files in patches +Is there by any chance anything else in the error log that could lead to this attach a complete error log +All these options are being supported by the new version of the plugin. +Perhaps for consistency with the annotation could have attributes which indicate where the response body should be taken from specified as an enum defaulting to the method return value +fix now in CVS +Cant figure out what might be wrong without the original data set as you say the gdalinfo output is regularI have similar geotiffs that appear to be workingWe can make progress on this one only if you provide the original data set or a portion of it that still manages to trigger the I send you a zip file with tif file and tfw aux file regards +Thank you Kim for filing this issue +Raising this to critical +not a blocker +But we already have these! they are createOutputopenInputyou mean for each directory I want this in I need to subclass the dir an pass it in? That doesnt seem reasonable +I committed to trunk. +The patch will be welcomed take your time please +the dir pathCaused by Illegal character in fragment at index ftpusernamepasswordftpdir +It looks like the YAML subsystem isnt handling file properly +Ive changed both names the pointcut name and the annotation binding name and I still get the error +Thats why I thought they should be provided by hadoop. + +It seems cannot be used because it is required to support null values +Zeta Components wiki is currently restricted to committers of that project +Hey AllenCan you point to places on the net where the hdfs alias is used? It hasnt been as common in environments in which Ive worked +Ill submit another patch later removing the sbtSince you do the releases any info on or changes youd like to make the next release easier +reopen to be able to target correct fix version. +Thanks a lotMichael Xiong +Just because it doesnt do exactly what you want doesnt make it broken +Sorry I missed that while regenerating the patchThanks Todd for noticing itI submitted the patch please check +Attaching converted +Yes It seems to me exactly like Len wrote +Can you run the stlyecheck again with the new filesThese are the files I found where the changed file still contains lines characterssrcjavajavaxjdolistenersrcjavajavaxjdosrcjavajavaxjdosrcjavajavaxjdospisrcjavajavaxjdospisrcjavajavaxjdospisrcjavajavaxjdosrcjavajavaxjdosrcjavajavaxjdoI didnt find anything else to complain about +Here is a patch to the Maven pom that will build all of Pivot using Maven installing binary and source packages in the repository +Thats correct +Thoughts +If we can not assume this then might be better +Thanks for all of your help Daniel +Now that a separate dedicated module is in place Im going to try to align the attached implementation to our this now complete? If so can we close it +Default is which is synchronous channel however if you configure it with than it would become Async +verified by Nathan +In other words could the reduced running time symptom be reproduced always even you run patched version first then rerun the original version? Itd better if you can issue echo procsysvmdropcaches to free pagecache between each test +Keith can you please evaluate this patch and commit it +Derbydoes reuse space in the table when subsequent inserts come in +May not matter much since we do expect extra copies with this patch for larger responses +This is close to what Ive been looking for +I had problems applying the patch +Changes include moving init from joinDesc to to make it after column pruner +Bulk close after release of +Closed for incorrect DR release already fixed in upstream but not until DR +Sorry about that it looks like the circular dependency check cannot handle some cases where there are incorrect dependencies +Hi HudsonI tried to download the latest axis jar but I am unable to figure out the changes done by you for this particular problem as well as if I try using this jar then my build seems to fail +In trunk it is not even used in core! Only in s +This should be done in the was fixed by. +The graph chart would be interesting to see. +Fixed loggers to be transient and DONE +How could you not be impressed by Jackson? Its fast effective and full of cool features! +Ive had no luck at all with JIRA search lately +IMHO the strongest this should have happened is you propose that the current discussion is not pertinent to committing this patch +Until Tapestry itself is fixed users can insert this code into a custom as described in the mail thread +tests +In code he class had its use I guess but now it no longer changed the title of this issue because this reflects much better a this be fixed now +The survey was still hanging out in the old mail util +Logged In YES useridLets not change that it seems much more logical for me to first read all directories sort the result sort dependencies and then deploy +this is directly extending Tool just with a static inner class +Bulk close resolved issues. +Vesta and solaris work fine +I tested it on Linux SLESSLES and Windows on both xx and all jdktools tests passed successfully +New patch that has testsBecause the original test was developed in RONDHUIT and it includes test codes for not only but also synonym dictionary the attached tests may be redundant in terms of +Once those releases are underway Ill take a look at this +Its already very late here in Poland so Ill have a look at this tomorrow. +Sounds like this bug can be marked Resolved right? Once resolved and we publish itll get moved to closed. +you should modify the ant fileDo you actually use ant to compileIf so please signal this issue on the dev ML +I have done I hope without to much mess +Yup should be all good. +Should be supported in latest version of the designer +Verified Fixed +Prepping +Partly fixed the doesnt pass null a the default configuration +Unit tests check that the code does what is expected from it +Please reopen if you still see this issue +committed +No unit tests available yet Im not sure how to test the URL or failure cases by looking at the existing tests +It is getting the raw Spans produced by the Compilerobject which may contain overlapping Spans +This looks good to me though I wonder if there is anyone who knows this code better than can take a look +Updating website svn ciSending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexcommunitySending srcmainwebappindexSending srcmainwebappSending srcmainwebappsiteTransmitting file dataCommitted revision +Patch for added. +In a module +Thanks Myriam +without my patch to try and get a baseline +See also the comments in which describes exactly that +adds a test in that tests that temporary tables are deleted and not visible across logical connectionsAlso set the createDatabase to create on the data source for some of the tests so they can be run individually and not fail if the database hasnt been created yetCommitted to trunk with revision and to the branch with revision +Can you get some generated classes? For example some web service interface classes and so on? You only cant get implementation classes or any classes +Ive added a FAQ entry and added jta inside the to m marking this as fixed since the problem has been documented in the FAQ +Something like this +The more sure way is to use this plugin and specify the older version. +The changes have been applied to branches and for hudson results +IMO this should be part of as its an enhancement request +Ruth patch applied at r thanks a lot please verify +great removal of another synchronization point +On the other hand we can always move files around if the top level generated directory starts to look cluttered +Addressing above comments +The problem is that the default behavior is only to collect those metrics which have been changed +Because when attaching the component we dont really know what components really are attached +Missed the model role field and could add a check box for same role simulator in which case the field does not need to be repeated +Ill commit it presently +Nod I got confused. +I prefer having slightly less but more general SPI interfaces +Please follow the steps provided in my last comment for the upgrade process +Applied patch from Joe +I have the same problem +Moving unresolved from M to M in preparation for M release +Attaching which expands the javadoc as you suggested +Ok thanks +At least Neha expressed interests in taking another look at the latest patch +why do we have a copy of in the test resources? +Updating the document with minor updates based on comments +Thats typically unique to each project and should be declared in the pom +Yes the EL name not the value of Named for injection as a qualifier +I have committed this +Calling mvn install on the project will do what you described +cant see this happen though +Patch to fix this issue +Looking at the code it seems that is the case +Added smoke tests based on Phoenixs suite of end to end tests +can you please expand on what you are not able to do +Jared can you either move it to the correct one or work on it from here +IIUC the modularized allows to use the functionality of all on one file but this slow downs the processing extremelyWith the current implementation you have to select the type of the before the processingThe current provides all the functionality of the and extends it by providing the XML functionality +not enough info to reproduce +Attaching the output here +Committed with the fix! Thanks +Is it better to treat both errors similarly +I investigated the problem a little bitIf you change something while you have DFM hold on a breakpoint at the beginning of run methods mainloop then the addition may be registeredI suppose it has something to do with the children property of the ObjectsIt looks like the in this map are referencing the new so that the check method cannot detect themOr there could be timing issue before +It might be the case that junit allows you to add some convenient notifiers and make the screen shot taking conditional on this +Nevertheless most diffs seemed to be reasonable +Looks like duplicate of +Sorry it took some time but here is an updated patch with some improvements that simplifies a little bit moreI also added some documentation but I guess this should be reviewed in another patch when this one could be committed +added proper class finder that seems to work with jars refactor code out of a test class into separate fixes the master reconnection issue +Thanks Stepan +Can someone change it in documentation +It is not worth a deprecation. +well need to doctor the original code in some ways +Except that scriptable does not describe the purpose of the flag in any meaningful way +I think I discovered the cause of the problem +Some wrapper scripts to make it easyI regularly use some wrapper scripts for most operations +In case of Xenserver you go thru pool master or any connected host in the poolIn case of KVM there is no concept of pool so if a host goes offline there is no way to determine if it is disconnected from MS or if it is off the network +needs a minor Paul! The access is now fixed on master. +I can fix thatBut I just thought of something the way Im doing the test of a region in transition over on the RS is wrong the stuff I added to address the Ram feedback +Change priority to major since if we are going to do this it needs to be done sooner rather than later +is it correct then to say that should be run if there is a significant change in the number of rows or the number of unique values in an index +I only tested that the pagination mechanism keeps on working with H and +Here is the fix +Now there was both a bug and a missing schema item to configure this and both of those have been fixed upstream because I needed both fixes for my testing on +You say that there exists a workaround and invalidate the report +Plus it is analogous to Lucenes Max Field Length which is token based as well +I am considering how this would compare to having a separate config variable for expressing this as a percentage +not an issue because meclipse is installed with FUSE catalog thing on eclipse +The patch was committed into SVN trunk with r and into the branch with r +Notice that it is already possible to write applications in boo +Would contribjcr be the right place to put it +Can this be made a priority as it will help with scalability +And I do not assume it would cause any problems if the socket is already blocking +When you get request level errors the response is empty and the error code is set +I had planned to cut the release during GX this week but other realities prevented that from happening +It should not be used in secure deployments except as a compatibility hackIm not clear what youre saying here Owen +Could somebody apply this patchThis patch improves the performance of init getset operations and also fixes the issue +Please ask about this on the user list +del file of buffered deletes +In this case Im expecting that such a sample wont even compile properly but whichever +Ex hdfs yarn etcWe can remove the explicit call of from scripts other than will resolve all the issues +Oleg for the quick response! +Should be fixed in is indeed fixed. +Moved to M +I confirmed it +Yes we need to update the doco +Ooops +It appears that the fix to the caused some other issues and thats what is about +Theres a JIRA to remove retry count entirely and just supply the maximum retry time and also add the ability to override it to each request + committed. +I perused the patch and it looks reasonable since we are just getting the from the Core there anyway +Can you attach a test case and a stack trace +Youll be able to upload a file and reproduce the error +And while I can see it being useful to periodically litter the version in the logs I really think it should just be logged once on startup. +Barry please use the email list for user support +bq +Im marking this as invalid as the unit test has already been applied andit looks like everything is working +It is a relational source model that has a UDF contained in it +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Im not sure how setting the fetch size is going to help if the entire table is being fetched +The specific issue listed here has been dont have any files in +We just need some agreement that proposed names are okay plus a consensus on how we go about deprecating old names +The class in Mahout does the calculations you want +Is there some kind of mechanism that protect it +i could find it +In the example this could leave an entry with a mail attribute value of +Something pretty trivial but for consistency can you clarify whether the name of the property should be as this tends to be used for indexing filter functionality whereas is used for the solr indexer properties +Checked. +Hey David could you post the exact command that lead to this errorThanks keeps all the table names in lowercase even if the table was created with upper case +I think we should fix this in btw since its a clear bug and can be fixed easily +That sounds odd +I will integrate many parts and make focus to database persistance of a this day August th as +Lets live with this issue being open and the test case failure until Jerry creates that fix +Also see thediscussion of the three functions in. +Yes I agree with you maybe you can change the title of this issue +oops thanks for fixing +Looking at the code it seems it doesnt use the maven archiver at all so I dont have a patch for that at this time +Not everyone uses avro IWE dont +Or you could use the forumsAfter all there is no issue here that needs a bug report +correct calendar attribute boundary description the first sentence is misleading +Agree with Dans comments +RalphThe last attachment is based on current trunk +This patch cannot be committed until mustang build posts +Shaun can you please specify book and chapter names when you locate typos as there are more than twenty books in the EAP doc comment with the link when this doc is built for verification +Encoding test page from Tomcat FAQ and descriptor with security options +Its one of the major features of the plugin +I think all we need to do is implement the fix for and then change the view group process to check for the presence of a +Plus some minor changes on the testsThis fixes the problems with Int as well as improves the buffered transport impl +Patch committed rev cfcacadfbaadd +batch transition to closed remaining resolved bugs +Created reviewboard against branch origins probably because Apache reviewboard was down for sometime in the morning +This change will not work +This could also be addressed with an expansion of the translate feature such as translate update set +This was the nail in the EWAH coffinbq +Dan did a change which should fix the issue that both services are visible in both contextsI think we still have an issue if two services specify the same context as we currently create a new servlet everytime +Marking closed. +Verified fixed in BRMS ER +It should be easy to extend that test case and share the code thereAlso please take a look at +bq +However I need to check once more with a restart and reinstall of Karaf so its plain vanilla again +Fixed with r +This patch improves the handling of the content in multiline text boxes +VFS targets Java +and it made sense back then +Since we have blacklisting via health check scripts in place we should enable this option to print the correct list of blacklisted nodes +Marked as resolved +Is this ready to mark as resolved for and trunk? No activity since July th. +a lot for the review Eli +Since skins may install additional components as needed there is no way to establish a rule by which containers themselves can reliably manage the of inserted components +If I dont hear anything from anyone by next Tuesday Ill go ahead and commit this +Interesting +Flavio with the encapsulation comment basically if you look at the interface there is very little that is about it append open close roll operations all basically take hbase constructs and could have any implementation behind it +I remember i tried to use +It looks like a new issueCould you tell me the network type? Basic Advanced Advanced With Security Groupselect networktype issecuritygroupenabled from datacenter +My schema was incorrect I didnt declared an EQUALITY matchingRule for the DN +Adds Scan Writable +bq +is in path element and when it is not correctly substituted it causes CLASSPATH to not expand correctly +Hopefully the need for this kind of bridge code will soon go away. + LGTM +I can confirm this one is fixed but unfortunately another referencing issue is blocking the geoserver release now +It seems like patch is applied successfully +Scheduled for +A lot of tickets upstream are requesting replacing forrest with something newer and not needing jdk Ugh. +Seems more like a language design issue so to +RolandWhat should happen if one attempts to get a token iterator for a complex header such as? I clearly prefer to have a generic way of dealing with headers of ALL types something that works in ALL casesOleg +However when this was all deployed on jenkins the tests were run from the command line + +Just submit a patch with the changes and a committer can review and commit it for you + dialog forms completed +Thus we need an engine branch contains that adds an engine defaults configuration that by default uses the and can be set to use +I think this more correctly reflects the original unlock semantics + +I dont like this too much +This runtime data is kept living by its classloader which loaded the rules +I have not handled serializer error scenario in this patch +Removing the tuscany hierarchy from the repo and rebuilding SDO and SCA fixed things for him +is attached to show the with the correct names +I didnt see a method like this in Math Utils yet +Still present on the master +Thanks very much DagCommitted patch to documentation trunk at revision I dont think theres any point in backporting this fix to or earlier since these docs arent regenerated once a release has happened please correct me if Im wrong +Profiles will do! Thanks for looking at the issue +Its happening in a different place but its still happening +All other ideas for the monitor page should go to which is providing more functionality to the monitor. +That also tests transitive thnik this test sover the issue not sure though +Thanks for the patch! well look into it and apply +However the test hangs outside the debugger in my git environment +I am not sure of the implications of changing it +Resolved with revision of assembly resourcesCommit message Add migration script for +Expected result Success in console log when any file is present in mntsdcard +patch looks good +Resolved by Chrstophers all tickets +There were no issues migrating from the as I already migrated to the snapshot forward to the configurations +Ive deleted this comment because the error case it describes is based on a completley faulty testing method +I have done the implementation to this and attached the patch +Rather than raising a bug report against Struts could you please post problems like this to the user list firstRaising Bugs is not the best way to get help asking questions on the user list isThanksNiall +Is it now actually hooked up to ext? Most users dont run on XFS in my experience +I dont think we should merge this to no rather not +Mailing list tested +Dmitri if youre still out there can you please acknowledge your intent for the content of your patches to be available to be contributed under the terms of the Apache Software License versionThanks response +Thanks +Months have between and days +Thanks +And rely on thread local on storing resources +addressed recent changes to switchyard core soap hornetq camel common rules bpm rules +I wish I could just submit a pull request +Implemented the following types as standard types in the String Numeric Decimal Binary Date Time types exists as an enumeration in the where Custom is a special type decided by the user +The old one may work not very well in scenario. +Returning the data in this format allows us to use the same in both implementations and no other modification is necessary. +It also adds a hashCode function and test cases +Sounds like a bug will fix for +Generally when ZK loses the connection youre session is dead +Are you saying that the code worked correctly prior to RF? IOW this issue is a regressionCould you provide some sample code to reproduce this issue +Dillon you out there +Yup this is true + +can get this to work if I configure the datasoruce JNDI properties properly +Patch applied +Thanks for the replay +Thanks! +Yes its should be selected in Visual part +Craig is there an bug report that would address supporting VFS urls from which an in jar format could be obtained? +So there must be a missing or or misconfigured dependency +Need to verify and add unit from the release +This is duplicate to I suppose that s scope is the reading part +Click on server Add and remove projects and here Remove In order to reproduce your posted problem would you attach sample application that can be imported as project on eclipse platformIll try it too +File additional bugs for related behavior thats still broken. +Do you have a test case? In particular can you demonstrate the problem with a failing expression? Generally if you access the pattern API directly its pretty easy to violate certain preconditions that are true when starting from a full evaluation +v contains what we should be able to removeIt breaks a test in currently +Since we have implemented statistics gathering which is a runtime operation we cannot fallback to naive cubing if we detect a small dataset +Would you be able to work on a patch for this +By returning we would be able to serve user requests for start time on old jobs as wellIf you think we should go the route I could redo it that way +bulk defer to defer to. +David youre right this is an svn checkout +just using it as a reliable storageWe probably wont use ZK as a pure storage + DV work is suspended because of no volunteers so closing this issue. +Right now unit test code would need to inherit for instanceI am rewriting for our needs +I investigated this problem a little bit +Walsh? Does it allow us to license under AL +This patch works perfectly for my application +Done see versions component. +In small applications you just dont realize how much its costing not a problem caused by Roo as such but the current Spring beans implementation with Configurable means you cant use it in a domain layer for anything other than a trivial app where the cost isnt important +For now TYPEERROR messages are used only during the initial exchange while exchanging UID versions +Andrian if you reproduce it again then reopen please the issue +This is hard +Fixes have appeared in latest alpha manuals +issues after updating fix version. +Thanks Ingo +Ill and fix this issue if the prevayler module will ever go out of the sandbox. +While it is an interesting ask I am not sure changing the PL semantics and guarantees of is a good idea +Critical seems inappropriate unless there is some functionality broken +See you and Jerome in an email thread a week or so back mentioned that you were planning on releasing a second Chukwa patch +attached patches for and to in look good to me +One more note I get a response even on subsequent not just on the first one +I just committed this +I believe this is a Struts issue rather than Commons Validator oneThe messages for javascript are generated in the Struts also Its probably not just a javascript issue but server side as well +So what do you thinkBest regardsSascha +The patch attachedUpdated Oracle and Postgres UPGRADE scripts added ON DELETE CASCASE for foreign keys in ambarirca schema +Sorry but I just found out that is the issue of our application in stead of CMS +Not approved for EAP +Yes good point +Closing because this was not the problem. +Its not just mm +Setting descripton to be to since we ran out of time for to since theres no urgency to this +Somebody with more git skills could merge my pull request into the X branch +I thought the master could do it before that in the bootstrap +If we backport then we need to backport interface and two other data structures zk and regions +Thanks for the fix Scott. +Thats ok for me +Making a note +Ive noticed this before on a single node monitor and gc also didnt stop if memory serves +Patch looks great to me appreciate the refactor of the Server classes as well +So should this persistent property be added to in the same way that the destinationType message selector etc areIll go through and check the pointer where we are sending JMS messages +synchronization using +These issues where all marked for we no longer maintain the branch in the community +Assign to Vicky to complete the deployer changes and rescheduling for CR +Can you provide a full test case I am not sure its related to post load and I need to see eactly how you play with your entityTo disable the checking use +HTML design has been changed to use style attribute instead of css +Ive a question +please not find any attatchment here so I come up with something similar +Ive pretty much gone through the entire code and made sure the PROTOCOLHEADERS are always set to a that is set for case insensitive compares +I want to fix this before next release +Flushed Unity gets into a bit of a knot if you rename a page +Patch to be used for Yahoo which fixes an infinite loop in and for yahoo hadoop +bulk close of all resolved issues as part of closing items. +Closing for now. +Fixed the output handler so that it only escapes double quotes insideattribute values in HTML +This is already fixed by +But if it helps I can test anything you want + is done +Here is the correct patch +I implement solution that its mentioned here but still get error that get and Franois Le can be resolved? +While the correct file was dated the upload at used an older version of the patch file +cannot be verified yet due to will be fixed can be marked as verified and set to CLOSEDThe needed setup of HP Systinet should be properly documented for customers Kurt can provide the information what should be added to Systinet. +And I think it was a bad idea to mix up two subjects in one ticket The message header is the really important thing for me +The first concrete use of this improvement is the implementation of +What does this mean for adding jars to load or adding files to the distributed cache to run a job are we going to create folders under this HDFS folder +Just wants to avoid such wrong usage patterns and gives clear notifications +Oops They should state that in the release notes +Gregory Also please update the reflecting to the current findbugs OK count with your patch +Feature verification is completed including Upgrade +My for figuring out a better way to solve this problem +Thanks for your noticeThis patch adds Incubator Logo +of course Id use that if I could +Built as +Im changing the IIS module worker to use this new function available in +Id appreciate a review of the definitions plus impl +okey dok I cleaned this up in r r r and r by moving the Benchmark and its classes into srcjava instead of srctest +Most of the difference because of somehow taking nearly minutes less +Id like to see full ARFF are you tagging this for? Sounds like a fine project +Cant possibly be a blocker the footprint parameter is not exposed for a is the reason it is not exposed +Instead add schema includes +Sorry was a typo it is +Upon further investigation of this issue its looking more and more likely that its a kernel bug so going to close this one as wontfix. +I meant use the HQL Lexer in color coding +LeonardoThank you for the fix +We might also want to extend this to cover the parser class loading done in the class when dealing with explicitly provided files +thanks for all those bug reports which are very important for us to keep on improving the plugin +Bug has been marked as a duplicate of this bug +Not as clean as having branches and tags dir per project but i can live with for the help +I ot the error every time I execute the jsp in the attached get the error create the tables in in a server EAP and launch the ll get immediately the you can also find the ejb involved in the test caseHope this helps to fix the issueBest in an attempt to clone and move the clone to WFLY +Please review my simple patch +for to trunk and. +All Done. +AaronMade the suggested changes and added some unit tests +When anonymous editscomments were disabled this had the affect of everyone not specified in the Permissions area also not being able to edit +Committed +I have fixed this defectKindly find attached the latest filesI am sorry I am posting multiple times +Will wait a little bit before to close this task in case some new issues or bugs appears +When this issue is and the Spring dependency is as a minimum we can remove the workaround as addressed in +Scott please close this issue if not Acegi Security specific +Thanks Pradeep +Then we should log a message saying what we wait for every X minutes so that user doesnt have to use jstack +duplicated now by and would also resolve s a tangent to the original entities should be resolved if they are supported +oh welll. +I know it is possible to generate code in the test sources +This means that this patch does not fix +Jeff best to ask that on the Archiva lists +implicitly converts all unlogged operations to logged mode for the duration of the online backup if they are started when backup is in progress +Enter key has to be prevented onkeypress. +Yep +Matt Is there anything I need to from my side to make sure this patch gets included in +Yeah I should take my own advice with test cases trying to with all new code now it really helps +Can you also verify that the other instructions on the page are correct +patch committed +Your patch does not solve the problem I think the management of locale in forms has to be deeply will try to provide a patch ASAP +That fixes the reported problem +The only possible solutions here are to have project settings which can be overridden by the server but not actually modify the project file +Marks hack could waste huge amounts of memory due to the complete fetching of the clob then converting it to a byte array then reading from thatIm not sure why using get doesnt work +This causes a number of the tests to fail as they rely on the s earlier fix sorted things out in dev mode +The problem exists only if empty values are inherited which can be quite always eliminated by avoiding to rely on empty values +Also there is not need to notify the AM of the decrease taking place as the AM requested that +Patch has been committed +Some of the commons projects are already mirroredNeethigitgit +batch transition resolvedfixed bugs to closedfixed +Well +bulk defer of open issues to +This is solvableLet me try it +Sorry my notice that it was resolved was premature +I cant reproduce this in trunk. +Weird will try to investigate a little bit more +Well have plenty of time to go through the bundles later +Right now OVM directory is not present under ASF. +Checked in the fix the by adding the storeeds to the sequencing paths +Replacing all spaces with non breaking space would do the trick but that is not easily done using a stylesheet I believe +Committed to trunk. +I have discovered that the must be the +Patch to OGNL to distinguish situation when property source is null +I will start incorporating your comments and hopefully in the meantime some python maven experts will jump in with additional comments +There are more specific solutions for certain roles and processes in the specialpurpose components but in the general application components we are trying to keep things pretty generic in order to avoid huge confusion due to a large number of custom that are irrelevant or bad practices for various user organizations. +thanksI think what is missing however is the creation of a code review item +There is still a lot more to do at and how are general i like the idea to specfiy package or class based mapping files +Committed the patch into trunk with thanks to Michael I will back port it into other branches once Camel is out +Oh youre indeed right +I will attach a patch +Will update the jira with the result +Many thanks Paul +Hi AlexeyYou are quite right about the spec +Ovidiou are you working on this +Adding Seam Portlet to Seam Project +Thanks you for the test case should be easier to resolve now. +We will investigate this problems +Yeah definitely there are good use cases interoperability often needs wrappers for convenience it would be possible to implement this in the main problem is just that its the input factory that needs to automatically and dynamically do it so it may lead to hard to notice problems for end user +I corrected in this new patchThanks +Grand +Was there a reason this didnt go into? +Marking ready for review +I think YARNAPPLICATIONCLASSPATH is appropriate since it sounds like the classpath to use for YARN applications vs +Remove reformat of apache code +Yes when a value is specified for the parameter the operation completes successfully + +So Im glad you found even better solution. +broken said in the forum It has been tried again just to make sure in CR +Attached is example of code and to resolve this is example of code and to resolve thisThis looks useful to me +This will involve lesser changes external systems and testcases and also all the changes that will be needed will be during job submissionrecovery only +Verified in +It order to fix this request two files will be tweakedFile name yarn I have added a new OPTS called COMMONLOGGINGOPTS in to turn off the unnecessary log info messagesCOMMONLOGGINGOPTS is used into the yarn shell script to turn the log messageSuggestion are welcome +So could this be added to the TS to execute automatically +It seems that you can change the name of the datastore to be different if you want to upload two shapefiles with same name +Im going to hold this off for awhile +BrettIm not happy with that change + +There could be some invisible garbage inthe command line +There is a boolean PRINTFAILUREEXCEPTION that can be turned on for debuggingI am curious +Thank you for giving useful hints +I wasnt sure if it would always be easy to maintain order in future implementations +Bryon Jacob re namespaces on propsthats also probably a good idea Bryon Jacob at first my reaction was why would someone ever set their own properties on +This allows User Managers to action various requests without significantly changing the API +Move this for it requires a bit more work. +The filter will throw if one tries to write something while theres a stream being worked on +Still a lot of boilerplate shared between Append and IncrementI think we should take this opportunity to reduce duplicate code +I have run derbyall on latest trunk with v patch and found failures as belowderbyallderbynetmatsjdbcapiderbyalltools +I just have cleaned and the test still got the same output +Since extends Configuration could override the common codes +In current implementation in the first loading even though we can know the correct class name but there is not a method to know the fields in the correct class yet +And now I get the with no additional information in the consoleCould it have something to do with session invalidation after reloading? +Localization ditto +I recall now that I have occasionally seen these beforebq +First page is now muuuch betterNow the second page is the issue Its dull an ugly and as a user you kinda loose the context of what that page is actually some additional description and add the same labels as used on the first page so one can see it is a class etc +including email pdf jbpm etc +Adding the updated lib into our source path +committed thanks Robert Surwka! +the patch looks good +The question is whether the module is meant to be visible to applications without triggering a log WARN +checking for the null value if null the properties were not found for some reason +The previous patch doesnt apply properly +Initial attempt to build the Maven component failed to resolve xulrunner +It will not run on a Mhz machine +Fix checked in at revision +Note that wiki comments do have a preview but forums do not +Sebb keep in mind that creation f a file might be involved +Google App Engine is not needed +Yes these are ordinary Hibernate properties just like any otherTheir names are prefer to use audit over envers in any method names +However we could easily set this as a runtime configuration via nodetool instead +I thought derby was meant to be able to handlesome missing pages at the end of the file so this would be one case where it might be useful to try a more recent version of theDerby software +thank you Ted +Removing patch flag since the patch is in +We will see if this changes in the future but right now all we can do is add an exception when we detect this +done +It seems to be a problem of specific FF implementation and version. +Hence aLet me commit +This trivial patch bites the bullet and removes the dependency on newer autoconf and on ACTYPEINTXX macrosThe logic here is this on all platforms with conformant C compilers these macros are useless anyway on platforms where C compiler is NOT available libhdfs is unlikely to compile even when these macros are present +After reverting and keeping the other fixes repair seems much more stable +I dont understand how youre using a to parse XML +What do you think about this? A other idea is to use soundex to show keys that sounds similar so you get the messageSusa was not found do you mean susi +Debug Configurations Remote Java Application +If I get updates to cause a split again will this fail if the region has not finished compacting all the regions reference files from the original splitOut side of the above concerns I have not noticed any bugs in the patch while flushing or compactions all seams ok in that area +Proposed fix and test cases +As far as entity enclosing methods are concerned RFC is convincing enough for me +Since it doesnt make sense to reuse bc it assigned null! Also I cannot know if the caller is done with theIt is ready +Im sorry I didnt have the time to get back to this if anyone else wants to pick it up feel free otherwise Ill try to update the PR in the next few days +RaviFor your first comment I think we could try and do fault injection +Some discussions should probably happen but the initial thought is to calculate the next rollover date based on the file last modification date +Docs developmentmaintenance for the EPP release is Thanks for your feedback Ill include it in trackers for the next release cycle. +Assume closed as resolved and released +Checked with the patch and the current Mac OS and was unable to reproduce the previous behavior that I had noted where semicolons were no longer accepted a statement and it was impossible to get out of comment mode +Attaching patch. +It makes sense to use the maven properties +Thanks Jason those are some excellent points to considerJoep +If the test failed on due to this shouldnt we revert fomr trunk as well +Right I see no worries then +I did a basic investigation it is possible +Is this trueA Jody +I guess this could be implemented by placing generic last on the list so that it accepts anything that has an avro schema with specific and reflect picking off things that have classes loaded +There is also the possibility of having a switch NO CDATA LESS CDATA MORE CDATA etc +Single combined patch and this time against the SVN repo instead of our own git copyPatch is against reposasfthrifttags IRC discussion with Bryan Duxbury Im going to make some adjustments to this please dont commit or do any more reviews until I upload a new patch +Released. +It worked indeed +Just need to modify the test so that it actually tests something meaningful and passes +Yes we are not explicitly setting the its generated by based on the first EPR in the chain +Todd could you please upload an updated patch for? Jonathan is asking me if I could commit this +Any further issues with gpp should be discussed in a new bug. +bq +I just tried with but I get Caused by orgapachetapestrybeaneditorWith that simple sure you are picking up the correct version of as well +Im this patch without a config option to turn this behavior off + fails due to failures not related to this patch is handled by +I think it would be nice that do not introduce pollution in the first place +Personally I have a work around thats independent of putting the distance or pdf in the clusteredPointsClearly there are larger issues regarding the refactoring that have to be considered +Updated the mylyn and tasktop discovery listings to the latest versions +Should this be done in backends or +Since you were building from within mvn pulled in the from the apache mvn repo +Hi JacopoIt might be yes! I will have a look +Im not really up on how the FID Mapper works within Geotools +Can you post the Camel route you use +Great to hear that its actively being discussed +Ive committed this to trunk and +Feel free to reopen if this is still an issue. +Anything else I can tryUnfortuately short of installing the MS +say once every mins to viewing logs its obvious where its atThanks for taking it Nitay +It will need to be removed later +Thanks +Simply invoke mvn test to reproduce +Verified by Richard +committed in rev +Attaching patch that fixes the problem +bulk defer to defer to +Seems to working in Pivot Windows JDK cant reproduce this issue using the latest source code. +Attached to the trunk +Suresh I have ran the full unit test suite locally as well as operate a distributed cluster which uses the API in custom filters +The iterator goes over datanodes which have the block so there is likely some timing involved +I prefer systems that do blow in my face telling me whats wrong instead of ones that hide errors in the log that nobody may notice for months or years +Can you tell me the value for configurationSee thanks for reminding Ted +added please check and close +Jerome youre right we need to meet soon has been fixed now so no need to vote any more +war plugin build time avg +You are correct +An anonymous tagSo it should also be fully safe ie theres no time when you could do a checkout and find the back compat tests are failing +You may want to keep the old configuration option setter around as deprecated for at least one release cycle to be kind +To me it seems a far better solution than converting the stylesheets to a stylesheet with many imports as it also reduces the number of requests the server needs to handle and it works with AJAX +New patch with use of to the choice if bitset for volume of content + +No need to pound that point in further +Hi AswathDid you get a chance to look at it should we keep this open? file but with prefix in order to bettereasier follow the Aswath still there +fixes the javac warnings +When I was navigating the site with enabled the only actions that got invoked were Are any other actions actually being used in production? Should the +I can confirm that we get the same on relations with s when using lazy loading on and Hibernate +Im also on Windows XP +Oops happened to use some markup +Also I dont know if this will help or hinder but I did whip up this image that kind of highlights the way we store information as well as a mesh that hopefully connotes the mode shape of the information +New change to delete the blocks if the number of blocks is largeNote the part that Uma is talking about is from the step +Adding notice and license files +Even though the approach is more optimal the implementation is a bit hacky due to compatibility constraints +Could be used for this etcAnyone interested in working on this + solution in rev +Can you have a look Max +Frankly I think thats ok since its the only way to know the client will know the data was written AND the trigger was executed +Only two minor comments + +I think that problem pop up again in version is probably a duplicate of this one. +Applied patch as attached +I didnt notice it was the reload component classes that causes the failure. +Ill email you so we can discuss how to divide the workId like to post a question for everyone What are your thoughts on using Java velocity template for C code generationThanksDona +looks good to me +Can you attach the benchmark tests +use optimizer hints to force the use of the index in your query +fixed by adding profile we have another one in the +Right now this can only be StandardBut will introduce other kind like a Version that can contain +I just committed this. +I added safe close calls to all occurences +Marking as Wont fix. +sha too? or are those useful +Thats what this patch does +Apologies accidentally added this as a duplicate of +Also improved the displayed Hostname if all have the same protocol its skipped same for ports and directories +Thank you Benoy. +this there was a remaining issue with the descriptor +Thanks +See patch for fix and tests that verify all of bad parameter construction return codes per last paragraph of and E +Its still not fixed in the last snapshot of Now its OKThe plugin gets the value of skinconfBTW the solution of Gavin was a good workaround. +obvious patch is obvious +Im trying to lift that piece instead for consistencys sake +Here is a patch for review +Release on Oct +Im not sure if it is going to work for JBDS installed from installer built old way +This will fixIve applied the changes in to cvs + +FWIW I have the same issue with zkCli on Fedora JDK +Wdyt +Added fix version so we will deal with this issue in the next release either refuse or fixverify +I dont understand the compatibility argument since this is new APIEven though this is a new API but mapreduce already has some logic in their code traversing the input directories and getting file statusblock locations +Since the hadoop support is a client having more detail there is invaluable +Can someone running Windows try and replicate it +Dupilcate of +Koushik can you submit the patch that was used for +NETCF according to the MSDN docs unless Im missing something +cancelling the prev +Is it possible to update this info? odd +I could see changing the names of README and to and +Also removed synchronization where not required and did some code cleanup in transactionmanager tend to commit this and move additional refactoring at and to a new issue and not to back out + +Main motivation was that my laptop isnt very good and it was much much easier to develop and test it as a separate project but of course thats a pain to merge! +However svn diff does not include the imagesI just copied over an existing beehive test image netuitestwebappsdrtcoreWebindexto the urlTemplates test image netuitestwebappsurlTemplatesurlTemplatestestAlso svn move does not seem to show up in the diff +We agree that there needs to be enough training data for a recommender to output something but you believe the cutoff should happen in the evaluator whereas I think the recommender should figure this out by itself via some sort of a threshold on the expected rating +The performance should not be different from earlier since the actual operations on a container are still identical +We can fork it and commit to the code base as long as we keep the original copyright attribution +jar artifacts. +I can look at approaches to this later but lets get the warnings under controlThanks Patrick For warning suppression thats my previous understanding too +As far as I can see that might be true +The example also dedupes data +Those are part of but I cant find them in the +You can use LOG instead +Now all issues are retrieved +the read operation needs to create the ledger description since we have entries in ledger +Just to clarify my previous comment lets make sure we clearly define a contract for the durability guarantees of every channel that ships with Flume +I committed the patch +Ill look into this some more +Also seems to affect pull request . +previously we returned null for the JDBC remarks. +Mark is there anything wrong w the patch? Did you get it workingI got it working it didnt apply cleanly but perhaps that was just me +Bulk close for release +Thanks +Do we need both fixes? With this fix do we need the previous fix +Im still on an own private branch of and merging tested changes to but will use in the near future for newer versions of our company project +This issue is resolved for me on all my build targets +I dont understand why an exception cant be thrown +When apache builds switch to jdk this will be sufficient +Duplicate of +Now that work is well underway with AS all previous community releases are +This is it is a requirement of the EJB spec +Provide scripts to create the keystore manually +My guess would be yes +This fix seems reasonable and works for my installation +Is it because you create too many instead of using a shared instance +Jacobs suggestion helped +Is it possible for somebody to review this patch +All finder results are placed in the and all finder results are placed in the +Could stand to have some testing done if you want to try out a trunk build and see how it works for you +Few initial comments Unit test is good but It would be great to provide a in form of a hive query to demonstrate the intended usage on how this feature should be used +This patch was committed fine to +Closing this now. +Can ask DhrubaJoy +Will test when during migrating to wicket when it is released +This bug has been marked as a duplicate of +I am using version which I believe is the most recent +Fixed +In my cca year history of being around Lucene I think I saw GCJ mentioned only about half a dozen times +This does probably not work yet and is missing tests +Just been going through some of the changes and noticed that is also an issue +Also note that we are planning for a release of in the next few days +fixes all the issues with tests +I believe it needs to be per core as you could have multiple layers and levels and each core could belong to different levels +From chat conversation confirmed that does exist in the instance created with +Nice work around +Same patch review in RB +Referring to the is check +I am also not sure if the specification specifies that this case should be possible but JDO implementation do support it and therefore it would be nice to have it in +pushed +I will redo it after once are committed so identify something reasonably optimal but I wanted to put a first stone there +Patch only delete files if they have corresponding template files +correcting release note issue +The code freeze is good since we do need a couple of days to test the changes +trunk The interface have been renamed to and having a more precise and simpler interface +Do you have any idea how I can get around this problem +But at least we should replace by new + +Can somebody please sum up the current state of this issue? Can this issue be closed? Do we need to create a issue +It was tested with ER ER was not available yet +Do you think this patch is valid for the defect +Its only used in convert +I changed the to send the request aswsdl but I did not change the +verified by Mark +I would like to see some JIRA to rip out and but I can file those myself +Seems it destroys logj config as well +Only the supportsSorting method is copied but the two copies are not equal +Note that the patch is developed against a of +Thanks Nicholas. +The sample directory needs some suggest to Move the following filesdirectories into a subdirectory +the attached is a fix to the attached is a patch to the patches in r +Fixed for on +After configuring the checkstyle rule Commant pattern matcher everything works fine +With it there is no need for instead make the default the +The patch is in the trunk now +stack on the client or the server? I captured the above on the client which does not have debugging on yet just on the regionservers and master +Ill resolve this issue as Works as Designed +Are the proto file and the generated java file in a different patch +done in the jbm Dev branch. +Patch with changes suggested by the documentation from suggusted by Arun +FrankReading over the doc everything stated that GK is responsible for is already completed +Its not related to the maven build ofit looks like a problem in the old ANT script build of you got any idea what that script is doing +Initially the C client would be marked experimental until the frequency of changes to the Java client API drops to a maintenance level only +Also is the examples with some of the youtube and other real content going to get included +If someone could test tomorrows snaps that would be great. +Because once the increment rows memstore is flushed if there is no Delete it will be no affected if there is a Delete the data will be wrong undoubtly +Aaron any update here? this is working fine for me +This must be check in updateProfile wont use the patch because IMHO this must be managed in core and not in the UI will add an Exception called which will be throw by updateProfile and in rev in rev Wendy for catching thanks again for the coming review +This patch addresses comments from JIRA and RB including rename blobmid bloblast to blobvar blobcopy refactor blobvar to not require a term byte in most cases extend documentation around blobvar encoding rename Numeric +as long as were in there optimizing this stuff we should make sure that the ddoc signature is generated the same regardless of the order of the views in the JSON document +Its very likely that the VM Metrics will go into a tab under HostsServer InstancesVM Metrics +Looking at the issue no it isnt +Yeah this is definitely going to require more digging around on my partI believe that the synchronization is not necessarySorry for the trouble +Thanks Ive fixed and the web pageCould you please update the ldap to remove ccustine and put me instead in the groupgnodetminotaur ldapsearch cn grep ccustinemember uidccustineoupeopledcapachedcorggnodetminotaur ldapsearch cn grep gnodet gnodetminotaur +This error occurs within the step + +This fixed the issue for me +Ive added a regression test at the same revision as well +The problem is that passes strings to strings without doing any conversion +Any update on this issue? Note that this will help with edismax handler configuration. +Resubmitting patch now that Jenkins seems working again +Why do you think is break consistency and related with relaxed sync or oblivious sync +Comments are much appreciated +BTW Im not removing the DTP SDK feature from the branchs TP for reasonsa too close to GAb would cause cascade of builds toward CR and we dont need the delay extra churn right nowc no need if no one else depends on it its simply an extraneous thing we can remove in trunk andor after GAI am however removing it from the JBT TP in trunk + +Moved to almost never makes sense for these things seems to be a badwrong fix +that is the problem configure with users and group options that will triger the cache init fail during the regression test +This should be fixed in the eclipse plugin. +Jody want to review it before I commit +Thanks Jason +Fixes warnings in and +Any further enhancement will be tracked with a new issue + +Are you sure that confignameair is one of the compiler options +So yes I agree with you +So there will be extensions of the abstract CFRRCFIF TCFRRTCFIF and ACFRRACFIF +Bulk close after is released +I think the right thing to do is if the proxy value is set to true to return the type of the target class and if its false I guess its a little bit greyer +It is just an additional option that would be good to have + +BULK EDIT These issues are open to be picked up +Your patch wont hurt it +Making a blocker on Only pity is that I cant reproduce on machines here but Daniel Leffel reported this yesterday up on his ec cluster + will commit +html it already will be read live from your deployment +the new patch looks good to me +In that case we should exclude Setup the Environment and Import the registration example because the example is already imported and the environment is already set propose to have both the solution standalone cheatsheet cheatsheet within the Project Examples is the point of the duplicationWould you do that for every example +Thomas could you take a look and comment +Attached the with the offending transport connector commented out +this was fixed in and the integration all test with open a new issue with more specifics if you are still seeing the problem +What I seen there seems a lot more repeatable like a kind of feature and I found some interesting pointers +I have written a draft Release Notes item for this issue but Im not sure if I have described the issue accurately +This sample can reflect class methods implementedextended parent classesThe servlets and filter are registered and mapped in a servlet listener when the application is starting not via the as java ee didIf the geronimo plugin deployers are not ready use this command to build this sample mvn clean install Then deploy and run the war revision +I dont anticiapte it depends on more than hibernate swf Spring logjSee the javadocs for info and example config specifically the javadoc for ll post a comment with general issues I ran into +The pdf plugin is defined in our pom but do not work because of some malformed xdoc files +Shouldnt run in every deployment say run in any configuration since deployed often means put in production? also have to describe how a user deploys a new +agree we can do that in a different JIRA though I would mark this one as dependent on that new JIRA not to commit changes that will be undo right the way +If no format is provided it takes the default THHmmss format +They fix as little as possible in maintenance branch +Thanks Mate! +bulk defer of featurestaskswishes from to +A modules class loader is created using the builder API +If anybody will advise me on how to write tests for these methods I will update the patch to include tests +Thanks Marios Trivizas for the patch +Hi points +Tracking arent available at the time of request since there is a delay between job submission and container assignment +The runtime engine checks if there is the bin file +I made the set ranges method write all the ranges to a file instead of adding them directly to the configuration which is the same thing were doing with the password +to commit +That seems to run fine in +Dup of my bad for not searching hard enough before opening. +The patch is commited +And thank you for your reply. +Hi SamindaPatch applied! Thank you for the contributionLahiru +The debug from commons may be mapped to info in which will by default show up on the console +Just go to the root velocity directory and typesvn diff +Settings it to false solves the problem +should also allow me to get the linenumbers that go wrongany hint of what schema you are trying to reverse would be nice luck getting a stack trace just got this from the error log view +Ill add a CHANGES entry +Please add the file to the Derby web site under the integration sectionAlso please put the file under the srcdocumentationcontentxdocsbinaries directory since will reside under srcdocumentationcontentxdocsintegrate and expects to beunder the binaries directoryThanks revision +Should we find out whats the direction the spec will go before applying this patch +I suggest to resolve the issue with a wont fix +will work on this after the subversion by +and let us deprecate in favor of in the +Fix has been ported from to trunk. +Necessary for AS CR +We can run as this PDE via eclipseThe one thing that is missing is setting the order and export for eclipse so that it does not show compile warning +Nice work +Hence we would have to go back to the timer every iteration through +Closing all tickets +Ravi I dont think this collides with but please check +Forgot to add +Thank you Brandon +Usually I get comments from Jochen very fast +continues as before but now does not bother to ask the deprecated class if there is a definition stored in the request +Technically doesnt have to be present Spring simply only supports the beans namespace in then +Committed to after fixing the same problem in the other command line tools. +For the is and isIBMJVM methods I think all we want to do is call the methods andverify that they dont throw an exception +Mind taking a looksee boss +Any progress on your debugging +I will close pull request. +Of course +Everything happens within someones transactional context as such I see zero issue with asking a user to get a database object in order to get the connection +I can take some time tomorrow afternoon or Friday to look at the ordering but feel free to make adjustments as make sense for phoneME in the meanwhile +Not optional attributes are not reflected by this since they must not be nullThe patch only affects the code generation for Javathe first patch for did not work with this one does +Thanks for the patches though Viorel! +Thanks Ben +The fix is already in trunk and will be released withThere is a workaround specified in the comments of +Dan i saw the way it works and thought IMHO it would be more common if the file is not in a subfolder of the classpath +That code is available in Git RajeshYes Anoop +Considering that everything within the +this issue need to be closed? because its marked fixed Im not sure its correct +awaiting uploaded +This has implications in all dependent projects +Ok well just have to make this a FAQ +Thanks for coming back! Set to closed +IShould also mention we could put this in contrib as it is self contained +Lets try to reduce the amount of sleep statements in the test +I would like to keep the servlet codebundle in the repo +initial dump for the issue show io para for a service update registry api +I think theres little doubt we should find these classes load them and create metadata for them on app startup +Thanks! +Anyone have problems to make it configurable +removed mavengroovyjars The maven folder just redirects to maven so Im looking there and there are without with jarwhich one should be changed? does the jar need to be copied +Assuming that this daemon will be started from the same installation and configurations as of bookieAny thoughts suggestions inputs are welcomeThanks +Update the client with null checks +Ugly hack but its OK for now +test and fix in svn for +Heres an updated version with that change +I merged your PR which is correctFixing Java version on the job is probably easier but feature provided by your PR is a. +Manually tested and rule description updated! +Can some one please add the environment field for JIRAActually my dev environment is in windows so i faced this problems +instead of creating a brand new system bundle does it make sense to just load the desired keystores under +I searched on struct and equal looking in summary comments and environment and didnt find it +Ill try to get to it soon +When you put a newer version of Xerces in the endorsed directory it cant reference the right method +sorry the issue got burried a bit but now it is in +This a good change +Still to be evaluated as I dont have an Android device to test the issue and as I dont have the full code of the client and server to reproduce it +It is my fault +So there is no need to submit itI wanted the Jira to show up as patch available +I have pushed this latest patch to trunk +JanDont quite have the energy to tackle but Id be glad to comment on the sidelines +There are two variations one with the bare minimal changes and you need to run aclocal autoconf and automake and the other includes the artifacts of these commands. +I am getting some compilation errors while applying this patch URI and not foundCan you import the relevant packagesfiles and then resubmit the patch +I cannot try it out right now until I get back from my vacation +defer issues to + +Will help in debugging where the issue is +Adding the property to the distribution profile seemed to have fixed the issue in the hudson continuum environment +I had to refactor the way resources are managed internally so there may be some new issues but at least the Temp Destination integration tests are passing now +I have a deadline this Friday +Hibernate doesnt need a discriminator for so why use one? More detail would be good +unit test passed +And an exception from the optimizer will ultimately cause a fallback to interpreted m not sure that these are really memory leaks per se +Please pardon the spelling error that should have been not works in on OS X +Its a bit different working on code for public reuse than it is working on code for reuse within the firm or for your own useAs regards to the biased vs unbiased thing I feel that Variance does it right start it off bias corrected and allow the caller to change it if its appropriate for their useTheres also an upside standard deviation which is the same thing except you accept only the data elements above the cutoff valueI should have it ready might make sense to introduce another boolean constructor parameter to indicate upper or lower semivariance +If someone depends on a non documented behaviour +We feel it is indeed useful to distinguish that case along with unexpected errors not relating to state from those in which some other state applies +with me happened the same problemIllegal use of mappedBy on both sides of the relationship +patch testNote that this issue fails on linux system also because the lock on linux is advisory +Defer to +We wanted to only run whatever is minimally required as root the code to bind to the privileged ports +htaccess file in both SSVM then Copy template worked for me +this is IE bug with overflowscroll and position relative workaround set for richpanel position on on IE +Also documented as a known issue in the EAP and release notes. +I have completed the support on the Tree for tagId +Please include some more documentation about the functions and the files in general +bq +In addition there has been no activity on this issue in years. +I tested it with the Datascroller from project +Possibly add an attribute to choose what to do in case of conflicting specsI believe that its more useful to do this as inclusion rather than as inheritance +It passes and +Except for maybe two Shows how to setup a datasource Shows how to start multiple containers deploy remotely and the use of pingURLBut yes I agree these samples would be better if put as it in the plugin module and called using the +So I think this bug is fixed as discussed above +RAT during build and fixed configuration. +Which version of are you usingWhich database vendor and JDBC driver are you using +Then we modify the code in one of the bundles run Maven to get our jar and copy it into Glassfish for redeploy +Documentation patch with changes to to the content of the attachment Trygve should be the one to handle this patch +thanx + +As has been released moving this issue to future version for fixes should be done in +Deleting unused skin is not reason for changing of skin on runningAs i wrote before problem is that once changed selection in Device List its changing used skin what is not obvious for user +This is especially annoying when using the in anothersubmodule the other submodules are still affected by this issue +I have uploaded the dependency to so should be available from there now +It would be nice to have support in Arquillian for this this feature +Thanks for the report +The idea of having a constructor that accepts a params object was to avoid creating default params that get replaced immediately afterwards +Small fix committed to branch and trunk +Patch integrated to trunk. +Darryl please move any patches or outstanding work from this issue over to a new PROTON issue. +Artyom the patch does not apply well after Maxims changes in r +It looks like fixes this issue or at least Im able to extract text for +Testing might be tricky since we dont really have unit testing for internal methods +We should try and this as at the moment we dont have a lot of information to hand over +In retrospect however a patch probably works better than an uploaded file combined with transcribing comments +Applies to both and +I tested this by loading a table exporting it dropping it recreating it then reimporting then running rowcounter to confirm +How would this work in practice? Would an upload to a HTTP URL become a PUT request +We need to lower Xmx parameter to a value which allows the test to proceedMy feeling is that this route needs some more work +I only found this leftover from the page you most likely copied To remove Fuse IDEI assume that text should be something along +Also the standard deviation of a task progress is what is used for determining whether to speculate or not +Your contribution wont be lost +is used by +For WAS CE support youll want to check with IBMHowever this problem has been around for a while in Geronimo +And the related issue is required for proper usability IMOThis issue is open to point ASEAP team to issues that affects us +I am not super familiar with the mini cluster so I am not super sure where to look to fix this +Afaik we dont build this one for EAP at all so thats why it has the upstream version +It makes the text look odd +If it isunder half the available ram it will to go the ram dir if notthe new segment will be written to disk +Thanks Rick +Just to make it clear hajndi is available only on master node but I kill it before I connect with client to backup node so I need to be able to get lookup from backup node +The patch included adds a dialect property to the jdbc filterset +This diff prevents the overwrite from occurring +Here is the codes I got +Note I am NOT suggesting running the Hoop daemons on each of the datanodes but to put the actual code inside the Jetty servers in the Namenode and Datanodes is in a sense a replacement for hftp except that it is RW +successmark and +Please apply once Jay sends the all clear to commit to the branch +Things are much better now +Reproducible on for resolution of version was NOT PORTED TO for review to +Defer to all open issues to +PR applied in in JBT +Resolving as fixed +Weve just run into this issue running tip of trunk without this patch applied +Your code is using an not an array. +This is consistent with the format produced by Findbugs xdoc option +Copy file from and svn add before committing the patch +Perhaps the problem is only present when I use the +Some of this work was pulled out into other bugs some was rendered obsolete +It seems we need to modify the API like what does +Addresses some bugs in the earlier patch +This was developed already +Looks good +See for the resolution comment should refer to. +is this occurring have mentioned +Everything works fine +Bring one node into the ring +I was hoping create subtask would automatically link em +Enabling JNA didnt make any difference to the test I made +Verified by Tony +It seems that the NPE happens because of a programming error that is hardly a bug in groovy +Committed all except +Bug still exists for for this misunderstanding +after on +Do you want to have a lifecycle interface that you will have your managed bean implement or something different? the approach I proposed in might be applicable here as wellRegards this since there doesnt seem to be concrete demand for such lifecycle callbacks +A middle of the road would be that we could say that if the folder name begins with seam then a hypen is inserted after it +But that aside the log message could just be improved +for urgent issuerelated to my jobemail eugeneyumobile no RegardsHao +Perform merkle exchangestreaming flag tmp sstables coming in from streaming as repaired +The assumption is that if the server entry is already in meta we should obey it and it is there because of a previous roundrobinassignment +Ive committed the same changes for all plain numeric types +The and +I agree finalize may not be an appropriate choice here +More cleanup in this patch +Issue may already be resolved may be invalid or may never be fixed +Thanks for the review Colin new patch attachedbq +Just to be clear though if we want to clean up terminology that would be another issue +Linking as this was originally about remote not starting +is the original issue created. +This is diff archive that contains all changed files by me in the original to get it working under java +Patch combining patch v and Jimmys patch +In the generated HTMLthey both had name attributes matching their form bean nameIf you still think there is a bug reopen this +It should probably be closed unless multiple people are having this problem +Im always very careful when touching the code which could affect security but after giving it more though I think this change should be fine and it doesnt really change the attack vectorOne change I would to to your patch is to only use file provided by this environment variable without failing back to other common paths if the provided one doesnt exist is not availableWithout doing that a typo or missing file could potentially cause a security issue because we would fall back to other common locations which could potentially result in a more permissive CA bundleIn any case I will still wait for feedback from more people before merging this patch +note No tests are needed for this patch because it only changes comments +Test makes use of the jmxinvoker +Maybe it can be added to do after all +check before writing the generated pom to the target file and ALWAYS attempt to verifypatch the checksums for the target pom. +I think the essential semantics should be explained in the reference manual +we need a Jira issue for each of the subprojects or can we simply request that they also be addressedWe are actually wondering why we have more than one wiki +As we didnt released we still have an opportunity to fix it now +the analyzer fails if there is no target folder +Deployed to snapshots. +Sorry I meant +The change looks good to me +If I can support you please tell me +How do I download Logj from source? +Hello Any news about this? Has anyone been able to reproduce it? This is a showstopper for our migration to Grails +How does hte toolset automatically detect the ports? It uses xpaths on specific files and these xpaths are stored in an xpath model in the toolest +KernelOnly this particular line was found incorrect +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Weve been bitten by that didnt think to change it thenEither way is fine with me as long as one can tell from the command line that the heck that process is doing +I think this is solved by +Cleaned up Ivan +If we want to use the RM project classes in AHS we need to have the dependency AHSRM but this is not possible +Ill roll back the site to the last release +Please confirm the current status of this issue and if its incorrect change it as required +There should be a behaviour in this case +Patch applied but we really need an automated test to prove that it works +That is why a response files was +Good point +createScript is responsible for it in the end +fixed in revision Dmitry Ive slightly modified the tests please check if its OK for everything is fineThanks a lot +Im currently trying this +Sounds good +Done +Fixing it myself +all issues to the CR +Is it ok to impersonate the test jobs with those users instead of using a user which is specific to testing purpose? Actually i thought that there are some users which specific to testing and those users can utilize for impersonating the test jobs +Thanks for fixing that Scott. +Why commenting out assertions in instead of removing themPatch looks goodI ran and it passed locally +GhaisDo you want to provide a patch that fixes this +Thats not possible since the pointer is passed by value +Added checks for invalid queues and fixed test that was previously expecting an invalid queue to be ignored +As pointed out I think using a real mocking approach is better than our custom mocks +are some other file need to update such as the file in the resource which Claus mentioned and the in the +So itll probably take a couple of months at leastJust a quick question Is there a way to programmatically change orientation in javascript? +But I currently dont have the ability to go testing older versions +required artifacts missingconnectorjarmailjaractivationjar the artifactjar optional dependencies shouldnt even bee assumed silently I believe +Should I upload this with a plugin before the test? Should the test go to systests uncategorized +Ivan I just uploaded a new patch on +submitting to hudson +That was exactly what I was looking for thank you very his last comment customer is OK with provided answer and behavior. +On a happier note the previous submission did fix the broken links to the distributions +When the order of the elements inside the payload List is important then there is additional work to do +jobCounters should be volatile because it can be accessed from getCounters +bq +Thanks Todd +This copying will be done only when are created +bulk defer of issues to +If you can try the commit mentioned above thatd be swell +Ive committed a workaroundIf you use sconvertEntity with a select many you will need to set the collectionType attribute on the selectMany JSF tag +Wiki tickets +Could you post a sample code of your Action +Anyway for implementing this and commit them to trunk +My analysis of the problem was incorrect in fact this was most likely caused by problems in Generator + +And interestingly it seems that the is also a bit messyboolean timedOut false on line is never set to true resulting in a loop that will be infinite unless an exception is thrown which will halt this testThis suggests that this test is not properly implementedRegards I have excluded this test +Hi MikeI still cant do a full checkout! I have no permission to read package +The introduction of cachedServers is for fast lookup when clearing cached locations right +Please do reopen if the issue occurs split as per recommendation in comment please reopen if the issue happens again +All the regression tests passed +Ive created to discuss details about a new API +closing this issue which can now be dealt with by the infra team. +Patch against trunk + +and +Rakesh +check out the subversion commits tag to see the changesIt may not be enough for you if you cannot augment the environment to add the xatrue propertythere is also a property for the connection where you can provide the list of names you will expect +Rayees can you pass on more details +My guess is that its the method calls you may gain much faster reopen but you appear to lose quite a bit on standard sort speedCould give the choice of going with the arrays if they prove as fast as orig but then back to needing an efficient merge +Can anybody with more NIO knowledge review this +Thanks for the patch Prasad +I didnt change the dependencies of the geospatial part +I have addressed the issue in the release by setting all osgi modules to use latest Felix release and removed the itests that depend on the SNAPSHOT version +The reasoning was that the old gml encoder was battle tested enough and that introducing the new one would have caused performance regressions so we did not touch it +Hi HaraldDo you want me to push my work in progress to Github so you can review and help me out +Please report back on this about position info +Run tests +Hi Dawidthis is a good addition +JanneI dont see what you mean +Thanks for the update Lars +It feels a lot snappier now thanks +Great job +Would be good to change this so I dont have to check stuff in on trunk +Committed in rev +Added the ability to create the program rather than the properties file for testing against current trunk +Currently there is no means to change any property while the core is up and running +Ist his the problem? +Thanks Hemanth! +Rolling your own class might be better but if this is something very reusable maybe it could be its own class apart fromTheres too much discussion in this issue to follow Maybe you can bring this up on the mailing lists if the issue is still interesting +Logged In YES useridNo and I need to see the full log as anyone may be loading the mail classes +Committed to trunk +spot pricing on a is pretty close to being cheap! we are waiting an extra for spot to fill to save penny eachIf we want to use spot pricing we should set to somewhere not and probably makes sense for larger instances +So all issues not directly assigned to an AS release are being closed +of course since blocks are on multiple racks pulling blocks from overloaded racks may be a bad strategy +Raising priority as Maven have now changed their code to disallow any plugin named maven thats not in the groupid +The phrase message to flume event converter seems more correct than message flume event converter +The patch looks fine +Will commit it tomorrow if no the run some tests +I will investigate why it doesnt work but if anyone else has any ideas feel free to share +takes longer then expected in which case the thread is stopped and true the ends faster then the time period then false is returned wo interacting with the thread. +Fixed in trunk +But that solution doesnt spare a necessary modification of the Solr build with regard to patched dependencies +Package location needs to be changed + +This is the list of classes considered to be public API in +NOTPARALLEL to the automake DavidI added a few more hot fixes for this topic on other languagesr +In version you should be able to comment out define OPTIMIZEDREADS +Sorry I got ahead of myself +attached an xml file just MB that takes several minutes before it fails +I have reviewed the class and found the more than a language is given in the configuration file then the user gets a dialog to choose a chosen language gets then propagated and set as the default language +The problem is that when the search index receives an event to add a node to the index it does not take consider the persistent indices +OK Ignore the last one I didnt update the JS Ill check them out today +thanks +Is there a central repository url where we can manually grab and install the plugin bundles +Hi Xiao want to update a patch +bq +Being verified at still verified at environment. +Closing this ticket because there is a workaround in Tims last commentIt uses special proxy header to check the context paths. +Yes it is replacing spaces with line breaks in the text which is displayed in the editor +Input send as initial messages Coordination between client and BSP Input has random read featureInput and are managed the same way as it is in Hadoop +Here is a copy of the test in a Java file. +Hi HariLeslin mentioned this topic so I just added the JIRA to the Blocker list +PaulDo you want to propose standardizing this? I think it would make senseAlso in the interim we can make proxying honor cache control headers from the origin server if theyre present +But as with Hibernate it was dropped in favour of javassist +YesI dont see a need for it anymore +I and several others am working on approach initially one in which the edit logs are accessible by both the active and standby +Thanks for your patience in advance +Please update v patch +Hi DavidWhy do we deviate from default maven practice of including additional resources in srcmainresources? I suggest we should consider resources folder as the default location to host Ill try to come up with a patch to point to srcmainresources folderHTHThanks and Regards be honest moved on to a different project but your suggestions seems reasonable to in r. +Very cool +fixed in r + to trunk +This job that requires the reducer per machine configuration runs as part of a much much larger workflow where every other job benefits from having multiple reducers on each machine +Cant wait for this feature ready. +Thats the price of being on the inside +all issues to the CR +I believe that this is fixed by the change in. +Committed to branch +Applied patches for and to branchWaiting on successful build in order to test and resolve +Excellent my hack has been holding together fine +Will the fix be applied on as well? Or is that no longer supported +Hi somebody could give me the version for of given patche I still have not passed on as it still not stableRegards Daniel +Isnt this a duplicate of +waiting for details from not quite resolved the auth issues with google +So if pig depends on and it should work fine +Thanks for the historyIf people are feeding you malicious data its not like switching to JSON or XML will magically fix the security issues anyway +Vincent you didnt documented the jdk +Lets say I do agree with you for external links like +Just because something is in the reactor doesnt mean ANY version is applicable +I added new Bloom Filter related options to +Maybe we should try to close the view +Do you have any local Ruby installations +batch transition to closed remaining resolved bugs +The problem has gone. +trying out hudson to see if it likes the patch +changed severity to Major +Current patch is on this JIRA its is rebased on trunk +committed as r in branch. +Thats nice +So for example if you genereate a new Drools project but with Drools code compatibility the generated import errors +Id believe that more of the stuff and the stuffI meant to say I would believe that more of the stuff than the stuff. +The attached patch which Ive applied fixes the typo and adds some more create table examples that hopefully will helpIf the new help text is sufficient please close this issue +However Im still unclear on what exactly it is you want to happen +Feature freeze for has past +Please verify +Resolving +Initial patch for review +PingIll assume soon youve changed your mind +Still some more work to do +I need to test JSR well under custom classloader not to break other stuff +the second third attachment +Ditto for post processing directory +I agree that a logo have to be simpe +Closing all issues. +Documented for release notes. +See my previous comment +This is yet an issue although the fix is very simple +Would it be possible to check if the fix work please +Kanna You cant exclude or from the distlib directory since these are required by different components of Hive +Anyone that is calling a Java method will no doubt know something about Java or at least be looking at the method signature and it would be nice if what we used was the same. +Chris Nauroth i run the as other but i have the same? +The real fix would be to injectweave it in but we worked around it so far and will do that later on +The environment more specifically is Centos running Apache is attempted to reproduce the issue on Fedora with Apache and was not able to reproduce +cool now I understand +Any ideas +Thanks Jesper +Thomas can you please have a look at this forve added some logic to bypass any INOUT parameters without a name. +I will not post comments like this one on Maven JIRA again +patch file for Solr users +perhaps the suggestion from carlos in is the best +I would prefer to offer this on than as a util I believe this will offer greater clarity +Crud +Ill submit the patch today +Bulk closing stale resolved issues +It seems to me that a reusable CRC system could be used by HDFS so that HDFS doesnt have to it all +Attaching patch with the in CVS +Inserted blocks are recent by definition so it is ok if we dont get around to them til the next scan +The executor is configurable if desired. +modulesorgjbossasservermain now includes the following in its change I found seems to be sufficient for our needs +Yeah this one has tons of plan diffs due to the virtual columns +Yes everything else is inherited so you definitely have a problem with the banners there +Thanks for getting back to me +I see the same interface in too which I have concerns about +Imagine using it on the JDBC disk cache +I never got any dlopen failures +It also broke assign all unscheduled issues with priority major or less to target +The jars listed previously do not appear to be required by our integration and are being removed +Derek thanks for the patch +It would be good to support the lock timeout the way it is supposed to work +Revision + update the version of to be used as default from to +is not updating s updating only when we manually refresh the page or click on the Sebastian Have you checked my attached images? it is okThanks for the reportI will remove the time part from this clockThe datetime shown in that box is actually nothing else then the selected date in the mini calendar on the bottom of it +New patch which makes a couple minor improvements Minor code cleanup in Use BINARY encoding for the read buffers in and +Basically I think the decision of whether an enum or union is valid should only affect the enum or union itself +bulk defer of open issues to +Access time is a great start but I can imagine a lot of potential innovation and experimentation in this area +Then you are ok to turn nightly into snapshotand about what about something like +You should definitely try that first as the version of some classes conflicts with the JDK ones +Will commit fix as soon as precommmit passes +Id like to do this task +nice Xinyu have fixed most of it +it was purged after each call to project resolution +Previous version put all tables into one +we are using Httpclient and statistically to of the sites were acessing return +Fixed! +Let me add that to release note +I am on it +If it retrieves the container from the RM and gives it back before any other NM can heartbeat it will also by fairness receive the next container that the RM allocates +bq +So if you access the link httplocalhostecmdemojcrrepositorysystemjcrsystemjcrversionStorageecacabaeabjcrfrozenNodejcrcontent everything is OK +So I say go ahead and fix. +Added full deprecation part of this implementation that Im still wrestling with is the storage of the computed distances +I was able to rescue myself from this situation by restarting the node so that does sound like a plausible situationIf I run into this again I will get you a thread stack dump from the running process +Removed unused method +It fixed that error but still failed with a different error +I think this would be confusingI sugest not fixing +I have some doubt on +is set to have bytes per second permits and when tries to read the data of one row it acquires s permission based on row sizeThis way we dont have to wait rows to be read in order to throttle read in compaction +That is odd it worked when I tested it on my box +See comments in forum +Committed +Can you rebase v Jake +right we have optionsa +Please take a look at this patch +You cant see it now because I did in fact downgrade to and rebuild sorry +Changed patch to eliminate conversion of to +This attachement contains a wsdl with three and the sca module definitions and a loop back test case to test it +marking this as issue needs more input from is annotated screenshots of the short Waaay to many technical details exposed to the user for such a simple thing as a port we cannot find something simpler to present the users Im all for just removing the Customize part since users can simply just set the port number explicitly +Should I wait on this until you figure this out +I am fine with both +conf +It doesnt solve this problem fully but may be helpful +fix minor bugs +I verified that and Container in Apache Flex have the correct code +Please provide a testcase for this issue +Fixed with revision +Please verify +Im resolving this issue. +classes log level. +Rev Controls are now enabled for +wasnt this improvement optionalWe also need to consider changes in the xsd should be somehow marked in the xsd version +Done as part of made for doing the synchronous compactsplitflush mentioned above. +Honestly I could even deal with the delay if it wasnt just something so silent from the perspective of the client +Added XTENLANG to the test suite +Initial patch contains a simple registry to get list of users +I also think we should document this new cool feature and have onefew examples in our demo webappAnd should the +. +Rechecked no reproducible any more + has been released I update NMR to use it +Logback would log a warning if no appender is found when a log message is received +no it is not an xmlbeans issue cause it works when using plain xmlbeans +when maven is embedded in another application the System classloader can contain also other stuff from the embedding application +Thanks Paolo +One downside of this is that it doesnt allow the erlang install dir to be overridden on the command line +Is there already a test case that covers this or do you need to add a regression test +Please let me know the change to be done to to log Debug IMAP messages +Thanks Knut! As for your parenthesis comment I wonder how other databases treat this case +Patch coded will be rolled out in next into PROD +master and mastertest cases and older outfiles are untouched +URLs are fixed but public render parameters no longer work +Committed to branch and Andrews suggestion +Defering until we have issues with the templates again subsequent issues will be raised in AS closing deferred issues +just need to check materialize library is available on classpath containers +in this patch random matrix generator also moved +To solve either update the demos in that zip with the new ones as where Martin pointed out +Closing issue out as part of JIRA cleanup +There are problems in this jira one is the one you mentioned to reverse the order to kill attempt first the other one is described in the description that RM may potentially relaunch the application if client doesnt wait for the application to be saved in the state store +Asokan? AlejandroCan you pls confirm if we need this for? If not we can decouple this and do it separately after merging to trunk +why is this issue marked as blockersince its not a customer requested fix it should probably not be in a corresponding ejb tests are broken so if we dont fix this problem we have a major regression +First take on a Vim for +seems the code in components project is not sync up with the lastes one from the servicemix trunk which cause several tests trunkSending trunksrcmainjavaorgapacheservicemixcxfbcSending trunksrcmainjavaorgapacheservicemixcxfbcAdding trunksrcmainjavaorgapacheservicemixcxfbcSending trunksrctestjavaorgapacheservicemixcxfbcSending trunksrctestjavaorgapacheservicemixcxfbcSending trunksrctestjavaorgapacheservicemixcxfbcwsrmSending trunksrctestresourcesorgapacheservicemixcxfbcinterceptorsTransmitting file dataCommitted revision . + +Thanks Rob +This is a great idea I wanted to add support for it too +It would also be very helpful to allow for loading different sets of plugins for different am sure I have seen something on the mailing list about the Dependency DSL being able to exclude plugins by environment +Thank you for addressing my comments Dyre +is working on thisFeel free to take it +doAs is used instead to avoid a new context class +CL Leak working theory is that the CL is held as context field of which is not shut down at end of invocations or upon undeploy. +That breaks the patch it seems +It also changes Binary and Compact protocols to use this method of encodingI havent benchmarked the overall performance improvements to the respective protocols but I did perform a of string encoding and decoding which showed the new methods take about half as long +Patch commited on and trunk +Any chance of considering a on the Context as an alternative? +It also would be nice if you file a JIRA about differences we have discussed in devharmony +I cant see a difference +Quite a big part is done but especially the ones depending on third party interfaces are still not fully parameterized +I see +Preparation for Gora release candidate +Avery Im sorry for delaying the review +patch to add to of impl build +Ill fix it as part of the next patch which has more tests +Rather than confuse people by having this done for Beta but with subsequent changes Ive reopened it and set it to the GA fixversion +is under package which cannot be used by classes in package as the latter is compiled before the former +Leave this open for now create new issues for specific compatibility changes and link them to this one. +It can be argued that it is not fair to compare a channel to a source but I think it is important to alert ops that there is a critical problem rather than just filling up the directory with a bunch of files which indicate that errors have occurred in the past +Ive posted snapshots +Thanks for reporting this +instead we should use the profile which enables the aggregate site instead of the composite one TARGETPLATFORMVERSION and dont have to be set they are inherited from parent pom and there values are correct +Improved version +This patch incorporates Konstantins comments +I believe you are mistaken +The patch looks alright to meIt passes our unit tests but we start the server programmatically and we dont use JMX so I dont know how much my vote counts +FlorianThanks for reporting this it will better if you can attach a test case first +Ill make one as well +Sorry to break the build +It should just invoke the code generation tool when appropriate to keep the generated output files in sync with the protocol buffer definition file +OK now I know what happened +punting to B +Its not an issue of understanding its just our different styles and preferences butting heads +Assume that this is OK as we have done releases since this was the name +See attached stacktrace +Chris can you create a new issue for this this one has been resolved in a released version of issue isnt really fixed as the end result is still an NPE exception just from a different issue created +The critical mass of Javascript requiring projects is just not as big to see the impact inNot even ANT would work LOL +you continue to out pace us committers RyanIll try and look at at least one of the open issues this weekend +Tested locally and seems to work fine for a shell and a minicluster +This will be part of the integration effort with . +Applied +As another issue we should calculate the requirements of the machine based on the roles present and their minimums +This can be seen in the UML Diagram +Did a quick code review everything looks good to me +Recommited changes for this feature and fixed Vitalys broken test +Im sorry Javier but we dont manage to reproduce this issue so Im closing this issue with resolution wont fix +screenshot to +Escaping datefield value is inappropriate for localized dates for example French months dcembre or fvrier +Hence closing the bug +tmp file even then we will get the same problem +Verified on revision +Ok now I see that we dont use by default in code but to the one initialized based on algorithm and an emptyI will commit a fix +If unconditionally forced early lookup is not acceptableCurrent patch could have been much simpler if was always created +I was just doing everything from the commandlinevim. +This needs to be fixed in Polyglot and thus will require a new Polyglot release +Any news on this +I leave it to the WS team on what they would like to do +VERIFIED +Sorry for the noise. +Indeed checked. +The fixes were applied but this issue seems to have been left open. +Mike saw a couple of these and laughed a littleparam matchVersion Lucene version to patch this is passed through to +Elena thanks for the patchTried on SUSE no more failures with the patchApplied at r. +Do you have SLFJ on your classpath +The performance of our program doesnt have to be that fast so we chose eliminating reliance on other tools as one of our priorities +This version manipulates Tomcat through reflection to allow Tomcat to be placed in a class loader entirely unrelated to Cargo +Heres the modified form of the test that works there was a glitch in spy storage setup +ThanksBiswa +Please if the problem is indeed something different. +Claudio once a committer s something they can commit it can commit on behalf of the submitter +But do you ever sort across fields +How about using isis instead of introducing new flags for logging if this logging is useful for debugging +the server uses +Logging should occur and when the issues in have been addressed the if clauses can be removed as there will always be a set linked exception +I will completely remove the tests added by Charles as they fail. +Once the script is complete that shell exits and you are returned to your shell +Rolled back for andThanks for the pointer to +Im going to provide a patch about it not sure if its going to be enough to close itIve to see which effects it might have +delete is called only once and I added delete just after this call +With border as mentioned aboveI can reproduce with changes and ultimately Apache license if this is a direction we wish to take +Tested the fix +Yes Im going to update the example webapp config files to point to the Commons Validator DTD +Test case patch with thankstrunk old resolved issues +Thanks Chris! +If only move is specified the operation is +please just commit your single stage encoder should always be more performant than streamscopying +This bug is eqvalent to bug +Dont worry about the NCDFE Im ok with writing that off as a build problem of some sort. +Stefan David Im pretty sure that sync havent happened so far due to is a new groupId +I get the MD checksumcbbfbaabc +But it still only looks at the identifiers it doesnt do a database search to guarantee a you close the JIRA once you verify my commit worked? build works fine on my box and on hudson too +Reading online documentation for DB Sql Server and Oracle seems to indicate all of them return NULL in such cases +still wondering why it doesnt fail equally on all OSs though +Yes please write is up +The resulting framework still passed the CT our internal tests and the test suite +Include test case for +Here is the new patch including the unit test +Updating patch +What is lacking is an owner for either package +you mentioned removing the section completely +Not much help sorry +Thanks Sid! +The locators needs to be sanitized manually. +Marking this as resolved +It still needs to be done for the CP branch when it is unfrozen but Ill do that under another jira. +And the same would be true for HTTP headers on the response +Revised patch +You can see why Im confused +this is because there are now two I need to get that into a report format instead +Fixed with commit bdcbdaceefeabcabee +This will be included in the final release. +Applied patch with thanks to William +Boy is my face red +This may not be an error as one of the alternatives in is to allow prefetch customization for a remote delegates queue +Hi AntoineDid you install a fresh sonar or did you migrate data from a previous release? If its the last case is it the first time you execute sonar on this project since the migration? Antoine did you experiment the problem with Sonar version or? Antoine for your response! +For +ThanksDiluka +Polished the Derby front page and added a pointer to the release announcement +Update the not to disable the consoleLogPatch applied +I thought you proposed to get rid of directly reading the Configuration properties all of them +Was accidently committed with merge +Should be part of the snapshot build tomorrow +Since the connection is reused the declaring is neededWill better document this in the code will change the jira title to do this +In the case A left outer join B right outer join C A must be small tableI have a test case to test the backup test +corrected fix versions +If I change the component to an hcommandLink everything works fine +Ill let you know what I find +any chance? What needs to be done to bring this closer to fruition +I think this has some adverse side effects for the rest of the world that dont want your behavior +Todd If you dont have any further comments on this patch can you please commit this +I happened last release of. +Committed to master. +bq +Example to be decorated by +The areas Ill be concentrating on besides basic cleanups will be in the tidiness around the parameters and in getting some basic tests completed +If you are saying that calling the list method on a domain object doesnt return anything that is probably because there are no instances of that domain class in the database +Will you be able to address both and? What is your expected ETA for RCThanks! +I disagree all of the Good reasons for making that change to the RPC back then hold true today whether youre talking about RPC or CQLThe keyspace is equivalent to database in other systems that use SQL its the namespace +In that case we might need to change the tests andor the default list of types to index +Attached test passes now +Thank you very much for your assitance +Oops sorry when I regenerated the last patch the bad ivy dependency crept back in accidentally +I attach a patch which uses POJO Schema and create a simple echo service on localhostPlease apply it to airavata accepted the patch as it has good test cases +The still keeps a list of jobs in the jobs variable so it is the ultimate owner of the job list +bulk close of all resolved issues. +Thanks! +But is there an direct API in PDFBOX that takes the certificate to sign without using the bouncy castle? Regards dont personally need that library but Id encourage you to proposeit on the PDFBOX developer mailing list +I dont think this is lazy as a first agree with both Noah and Jan + +Rollback should reset not only addsdeletesdeletes counts but also cumulative counts of them +need a patch for trunk +Hudson build updated will be in next build +Anything on this issue? If this is user error what is the correct method to fill these fields +Yes this patch is a quick solution that allowed me to complete a crawl +Mabe he should send you the project so you can reproduce it certainly is a blocker issue +Substantial work was put into protecting Derby from interrupts in +Please feel free to any of those when required. +Committed +some coding flaw found +woo! prototype script to merge pisoa packed plugin update site w jbt uberbuilder update site works! See attached screenshot +doesnt really add any benefit +The refactor ended up bigger than I wanted so splitting it up makes sense for easier reviewing +It just annoys me personally +However this patch is for release version +Also ensure no locking around waiting for searcher after commit or optimize +Say every min we get deltas +Current behaviour is as fred says treat it as a standalone war +only files and when +Reviewed and committed +We need to somehow include it +Setting to null before or after close does not matter as it will never NPEI will commit soon +this is with the width removed using the runtime css debugger Firebug as u can see it lines up nicely now +What cassandra version are you running? Did you try to recreate a schema on all of the nodes or just one of them +scaffolding would be good to see in will vote for Android and HTML +I am working on it since I am interested in implementing them for Tiles so expect a patch soonI did not assign this issue to myself since I am not a Shale committer +I like the idea of making this a single property +Changed builder to make sure model exists before deleting problem markers +Patch checked in +In general the style I prefer is try to declare fields and methods as though each class were public +Any suggestionsThanks! this was fixed in trunk and but not in +I tested this with a big image and I could see that this now reaches the time of lsr +Unfortunately obtaining the spelling of make from files is not as convenient as obtaining it from the host languages configbecause the files may not be available when some actionsare called that need it +Done in EJB Projects and reflected onto an AS installation using the installer. +Thanks. +Im just doubtful that it will fix anything +remember to handle multi property naturalids +Disturbing the public as little as possible is probably best at this late stage in the game +Theres a test for that +I madeTurns out a bit missing +Will look a bit tomorrow at adding docs and throwing another component into the sample route + Looks good to me +Plugin upgraded to latest release +Each node need only have the RPC address of the other node set to an actual address for everything to work properly +Configure load balancing using Apache and modjk Task Configure Apache to Load modjkSpecify the URL to forward +This update fixes a small bug in the original supports only the functionality already in and Ive tested the updated version and it works as advertised for me +Because thouse wizards need some templates from Seam installation which are used by as well +Resolving the issue after the deprecated classes usages were removed as well as the deprecated code itself was removed +I think a singleton can be used here we just need to update the comparators staleIntervalTed I am not sure this is going to be such a big issue +We need to relook this issue in detail and create a fix accordingly +there is some situation where we would want to change the value of the containerclusterTracker Im fine to have the single reinit method +Fixed in trunk as of r +So youre right my mistake both journal should sync by default as before. +Oh Myrna Im sorry for my mistake let me post the patch first +never have classifiers. +Sending a token to an older insecure cluster is compatible because the older server kicks the client out of SASL mode +If you like it you can even post a patch +The affected version was a release candidate so its pretty much possible the problem is already gone +Thats sounds like a plan Adrian +I think that is acceptable as long we document the as an potential issuefix for the input format +I just committed this +I use the c Broker only +Thanks Ratnesh Upadhyay and Awdesh Parihar +This will serialize undefined as true +Thank you all for putting effort to solve this issue +implemented an exception handler to at least write a sensible error message if the class file format is not supported +The above pull request fixes it for me +The trunk job you set up previously seems to have vanished and is now replaced by the Nutchgora branch +Thanks +The code change in this patch has already been ported to the branch in preparation for +It might be related to because that issue is also reproducible only inside switchable panels +We dont need to rush it I think its fairly contained to change we dont even have to deal with this for if we arent happy we can deprecate just have it extend in a future release tooI am against this we should change this before Lucene +Richard I merged request please test +Checked Repro attached again +I did not know that this feature existed and I agree that it is a better and more flexible solution +You are right Sir! Done +BTW Graham after I wrote the unit test I remembered that your patch should not be needed +Any suggestion is welcomeRunning and derbyall now +Ive renamed the bundle to +is supposed to be a more lightweight locking mechanism but in there is very little contention for the lock so its possible that heavier mechanisms may actually take less time +File Attached patch using nested wwparam in ww in other way just including projectMenu +I attach the patch file +You are most likely using a version of CAS prior to which did not strip the jsesson from the URL +patch was indeed helpful and just in time for the upcoming which should finally be released this weekJuergen +So the threads are being reused by the application server +To me because thrift uses slfj is not compelling enough +A little updateIve picked up the latest release from FUSE and this memory leak is not presentIt seems they merged the patch provided in this issue +Added the revamped code to trunk for the first time +did an update and clean build so youre using the version bundled with Sling rightI suggest to try latest webconsole form Apache is no official release yet of the sling web console so thats why we still refer to the old version in sling + +But in my case I have several wsdl files to generate so I cannot use a global configuration +Pull it back in if you think different. +Fixed now. +What can i do next WillemCamel bug not a CXF bug +I think it makes sense to get as many of these new features into the API we have now so that when we finally make the new api with the annotation we wont have to patchwork the features anything changed on this issue since the last comments? Im having a pickle of a time testing many of our jobs that use multiple outputs +Hi JamesUnfortunately Commons does not permit any releases from the Sandbox so this is not possible +Dont know however if this is in accordance with original intent of this build target since the fix relies completely on what is built in target and therefore identical to nightly buildSee attached patch if you want to review my approach +So will add a test case and update the docs +Merged the second into the master branch. +Any harm swapping the order in which we do the delete and RIT remove +Sorry I meant documentation not configuration +Patch looks fine +OK Ill start applying the restructuring changes now +Sergei How are you going to remove the SNAPSHOT for your artifact? I mean when we are releasing SNAPSHOT artifact we keep SNAPSHOT in version and when we are release artifact we remove SNAPSHOT from version +Can you post the entire xml and xsl files you used to reproduce the problem here? It will help us identify the problem easier +This should be fixed by resolved issues for already released versions +Approved for. +trunk +Patch for to fix encoding when opening temporary files for Writing and Reading +I taught pmd to notice these and heres a list +I have pushed revision and closed the issue. +So even in case ofvmware shared network should work just as isolated if your observation is that mac address is not supplied there must be some thing going wrong in core +you have to set your own cookies as well +The committers +alignment within a cell please include VERTICAL as well as HORIZONTAL alignment +Er I mean patch applied +As regards to CPU cache misses given delete had to touch so many objects and namenode in general has so much active memory garbage collection etc +Cleaned up the problems Stack noted in +Heres a patch for this +This would be useful to me as wellI tried extending the in August and found a couple of limitations +Which together implement region assignment in doc hasnt been out for long just clarifying anyone interested in providing feedback for part ? Itd be really nice to start working out implementation details in part with some confidence andor writing code +Looks like writing our own impl of shouldnt be too difficult +Resubmitted the patch under +Nigel pointed out that sorting M might not be better than sorting M so minor change to the config for the rwsort +We need to check the existence of the membership before popup exception message if membership was delete +Ok I see it now please add the header and footer files back in there is important information on them we would like kept on the mirrorThanksGav +Instead perhaps add an integer member to the test class such as num? Then in one of the new cases you can set it to and in the other to ? +We do need to open a new discussion on the purpose of the and its usage of the caseSensitivity flag +change jws to jws +can you upload this patch to? Something this big should have a reviewThanks for the +Thanks package adds a new dynamic wiring has still problems +Fixed +Sorry +While that reaches some conclusion we can use this ticket to just mavenize the MR src and test jarsIn a way I am rooting for the v patch and script set +The row length of the system temporary table exceeded the limit that can be accommodated in the largest system temporary table space in the database +This can be considered as closed +Could you attach some example with reproducible code to this Jira issue +Fixed. +I have an implementation where I split on any change in character class withspecial treatment for alphas depending on case and position relative to otheralphas of different cases +All Ajax is broken right now +Ive noted that you have to check the javadoc for the part Im not sure what other detail could be offered feel free to propose anything if you can think of what youd like to see thereIve added the following to the class javadoc The following is the approach taken +If there is still interest I could post this codeWillie +I committed r which adds the ability to add arbitrary +Remember that each has a target recipient and it is removed after the redirect +Hmmm +If buffer ends in r and the following segment starts with r it looks like this may not separate those linesIve modified tests in to do additional beating on readLine +From the latest code to run a server you need to do cd modulesassemblymaven server +Column Family Memtable Columns Count Memtable Data Size Memtable Switch Count Read Count Read Disk Count Read Latency ms +Hi Paul +Were currently using of and I think this has been fixed in a more recent version of that plugin +This looks fairly straightforward +Ashwin! I committed to trunk and +There is a flag already for is so this looks like is +Actually the split has happened based on the code dependency and not on the package +mkdir mkdir in r +This patch makes two more changes +I also think it may be affecting a range of file oriented operations +can we do this for all sharelib components as part of this JIRA +Follow instructions in +I am closing this as wont fix. +now no matter what I do I end up with all pass +Patrickis it not possible to reverse engineer the java namespace from the namespace on the wirethanks sorry no can do IRCYes use the JWS dont forget needs hacking to set dotrueThe tests should workPosting updated version with less bugs and some explanation ASAP +Patch looks good +Stay tuned +batchId is not set in currentJob because we set batchId after creating currentJob i change to set batchId because this issue has a new patch that should be committedWait +Yup JSP tag file changed the items attribute to itemAttribute and changed the code to use to look up the collection of to reflect changes in the code namely to pass the itemAttribute option rather than the items option +I just committed this +Im not sure what differences exist between and so we need input from Anil +I will create an issue to run next tests +bulk fixing the version info for and all affected issues have in Im going to put this in +Applied patch sent by Heri Bender +Hopefully this will have a similarly significant effect on the startup when using a remote database when the total times are longer +If no definitions are injected in the tilesConfigurer bean the factory will go looking for definition files in some default locations +Once we do that any AM even in secure mode can talk to RM on both the protocols +The js was updated +For each client connection datanodes have buffers and the last datanode has buffers +Im missing something and cannot understand what is itI may guess few reasons of having the numeric localization entries one has localization tables for other languages which stick to this format one has a tool which operates with these numeric entries making it them maintainable this is just a bad style we all dont like but bear since the migration would hardly be completed one dayIt would be enough for me to specify explicitly that we dont want symbolic names even due to some unnamed reason and I would be happy +Looks like it defaults to hsha if the config doesnt specify but otherwise + +In any case I dont think this is a to send this around the horn to get input +Looks good enough to me +if nobody objects i will assume lazy consensus and apply the patch +I need a review of a proposed solution before implementing it +My email mraqaoui +Screenshot of preferences page +Web browser will get opened right away without waiting opened too soon +Please check that it was applied as expected +Also like startstop better than startupshutdown +Looks good to me as well +the redirect has been established kudos goes to the ASF infrastructure resolved issue +Patricks Day +Fixed rev +I want validation just for one ort two methods of an action class so I define one or two actions with method parameter specified with validation enabled or solve it with approach setting right excludesI really think that an optional includeMethods brings next no benefit but more confusion for users and some serious logic messup if someone has the nice idea to specify both includeMethods and excludeMethods on a single case which we cannot avoid technically +Also by JSF Im assuming that you mean the Trinidad renderkit and NOT JSF itselfIn general I think I would prefer to have a new exception rather then making exception logic more complex +It seems like it would be a little confusing +I fixed that in this patch and also added a testcase for that +To be fixed in actually +Thanks Kim +sections from to are converted into xml files three folders are added in images chapters +Basically oozie deb is pretty much empty +I think you are saying it has been fixed since then +Do you have information ona compatibility issuesb performance +Patch attached +Also please see if theres any workaround such as closing the view and it +Updating this issue to invalid as a part of bulk update +Some commets about the modificationjoin is pruned by seeing which of its output column names is used by its children and removing its unused exprs +Not quite done +Please read carefully +do +OK Ill investigate further +Could we separate out the cleanup and leave it in its own target? Either that or have two different build targets one for development and one for casual builders +I implemented it a bit differently after a suggestion from John Firebaugh so dont need to switch on versions. +In this case we know the source so there will be no deadlocks +A ledger fragment is the sequence of entries an individual bookie stores for a given ledger +The example file doesnt have a require in if what I fixed is a separate but related issue I am happy to create a new JIRA issue for what I fixed +BTW can you verify that setting key to null and not an empty writable yields the same output when using? Also just found +Also all the fetcher threads were trying to push data to the queue of the thread that died causing the tool to stall +Closing resolved issues for already released versions +I will fix the NPE though +Attached the junit report +Now its looking even less necessary but perhaps a slight improvementSorry it took me so long +Max is this ok from a Tools perspective? I know you told me that using java code from freemarker templates isnt always great for could use rather than +Thats true GavinI think well need not all issues urgently +This issue is related to +Some details would be nice where is the component placed in your EAR structure +Is there a branch that is still being maintained with the patch? I am getting errors when trying to apply it to trunk +I think we need a note for users in +What I meant was that the CLEANINSERT didnt work +That patch looks a little noisy but Ill take a crack at it +btw you know you actually need to remove the java dialect to avoid this? Otherwise whether you use the java dialect in a rule or not the java dialect plugin is initialised and thus the need for the dependency +bulk close for the release +Fixed by correcting the path to the keystore +Could you please update and try to see if the issues you reported are all fixedThanksTYPO +I removed the exception because I saw the problem differently if two nodes have no common parents then their comparison is undefined and thus irrelevant +waiting on access +Just one query before I commit it +I mark it resolved since Scott said the doc has been updated +Fixed ordering +I am using Apache CXF As far as I can tell the definition of is coming from +The problem here is that scope annotations are not consistently used in The default prototype scope does not have a corresponding annotation by design +is related to via which will be upgraded. +Due to the above changes we have restored the clean operation for AES +Stephane its probably best to open a new ticket for +Me and Sammy plan on opening another ticket for better shutdown support in Cassandra +I think so too +Well thats actually the case we have and thats why I commented this ticket +Another future idea along the lines of autocommittruefalse is supporting clients with different timeliness needsFor example a client could send an update request telling solr that a commit should be done within minutesmaxwait or something like that +Indeed when the SONARRUNNERHOME env variable is not defined I get the following error was unexpected at this time +Im going to apply this patch and wait for Gianny to review iti think that since we dont believe in singletons but rather gbeans registering in the constructor is more appropriate than registering in a static blockBut the question still remain do we want to register or impl in or not +This patch corrects the iterator logic to deal with cases where term offsets are stored in the postings lists +I committed this to the append branch. +Thanks Roman! on getting this into +Ill add comments for server +Resolved in SVN I have corrected the default value for the maxLength member of Text member type so no problem would occur if a value is not set for the member +please teach me how you share your connection so that sub process doesnt have to start from scratch +Marking it as a blocker since I am not able to update the physical network through API which blocks me to deploy a to master +Trying the description again this time without catching JIRAs deleted Wiki formattingReferencing a classname qualified static property within a static method of the qualified class itself should access the field directly at least within the propertys own getter and setter to prevent infinite recursion +over the weekend I finally got the source onto my machine where I applied this fix +Just needs to be deployed +I also got lost trying to find Struts documentation when I firststarted using Struts +Id rather see a return early in a function that a nested if branch that includes the rest of the body of the function +OK can you suggest a patch? Were on Lucene here as far as I know so feel free to assume that +We can build the C client but we cant run tests +Trustin I presume you will this JIRA to suit the results of our design discussions via email +Before any AWT realization or does it have to be before any JVM bytes get run +As of we started using the code under newplan as our default logical plan +If thats what you guys wish then can we make it happen? Our client would then need to fetch stacks from two separate files and I assume simply group them into a larger array +I had to move the mentinoned file into the resources directory then the tests seams to work +If you change the accessor methods to use the primitive type youll at least work around the problem +Leaving this open only until it can be tested by others for a day or two +I cant seem to reproduce this problem +m working on now +this is scheduled for monday +was released about weeks ago and there are no urgent blockers or huge additions to justify a new rc right now +If the reader of this message is not the intended recipient you are hereby notified that any printing copying dissemination distribution disclosure or forwarding of this communication is strictly prohibited +bq + verified +Aha! Found itOne of these should get closed as a dupe of the other tho +Still an issue on master. +I can try to get it out if people are interested but I havent tried yet because it was never extensively tested and there are no plans to use it in production +Checked out the trunk but get build errors +Not sure it would still reliably do that +classes +This bug has been marked as a duplicate of +If this is a problem please let me know and Ill find a resource to pick this up +Can we also have some thing like equal split which will try to split the token ranges into perfect halfs? this will work well for bootstrapping ring of sizes is what auto does with the caveat that nodes need to be started minutes apart so they dont race as in Jakes example here +Please dont close they are closed at the time of Release. +Screen shot of the access permission tag +So verifying that the compiled artifacts do still work was a good exercise which makes me happier having done it +are serialized independent of their container +can likely be added via when the shared context +the fix for might fix this issue too try in a few in build +I have committed the patchtrunk +bq +Just commandLinks are not working due to some method in +Are you using the latest version of itest? +Awesome looks like Gert backported these changes to the branch as well as the branch! Thanks Gert +It seems that it has something to do with the hibernate support +Can anyone comment on them +Is there a speed advantage +contains the +Merged to the branch in revision . +Its necessary to add informaion about row ajaxKeys etc +mdsum has been uploaded to. +Seems this is not major issues I have difficulties to replicate this on current CR +Many of the FTP mirrors are broken as well see belowftppubMirrorsdist FAIL CWD james Failed to change directoryftpmirrorsapache FAIL CWD james Failed to change directoryftpmirrors FAIL CWD james Failed to change directoryftppubapache FAIL CWD james james No such file or directoryftpunixpackagesAPACHE FAIL CWD james james Permission deniedftpapache FAIL CWD james Failed to change directoryftppubwwwapachedist FAIL CWD james james Permission deniedftpnetapache FAIL CWD james james No such file or directoryftppubapache FAIL CWD james Cant change directory to james Permission deniedftpapache FAIL CWD james Failed to change directoryftp FAIL CWD james james No such file or directoryftppubapache FAIL Cannot connect to OOPS failed to open xferlog log filevarlogftpmirrors FAIL CWD james james No such file or directoryftppubapache FAIL CWD james james No such file or directoryftpMirrorApache FAIL CWD james Failed to change directoryftpUnixWebapache FAIL CWD james Cant change directory to james Permission deniedftpUnixWebapache FAIL CWD james Cant change directory to james Permission deniedftpUnixWebapache FAIL CWD james Cant change directory to james Permission deniedftppubapache FAIL CWD james Failed to change directoryftpapache FAIL CWD james Cant change directory to james Permission deniedftppubapache FAIL CWD james Failed to change directoryftpapache FAIL CWD james Cant change directory to james Permission deniedftppub FAIL CWD james Failed to change directoryftpapache FAIL CWD james Cant change directory to james Permission SebbThx for the feedback +updated patch for marketing component +We will need this in for at least we need to document this for the users. +in case of codi it isnt an injection for static fieldswe just have and it isnt about logger injectionit needs to be discussed on the list please start a discussion thread +Welcome to open source world! applied on behalf of Iris Ding with thanks +Breakdown of the startup time in from from in from in from than SQL requests by table are done on rules and rulesparameters has already been improved in version. +But if you have a better algorithm lets implement that in Axis +works for me now! +bulk close of all resolved issues. +One minor change to required for compile resolved issues. +Verified in JBT. +If struts wants to support safari then this needs to be fixed +ted The last tests seems ok on even if Iam sure that we will discover some issues +I opened about the stupid hack that is inconsistent and confuses people. + test fails because of this bugAdded to exclude list for DRLVM at +Please and close +has since been opened to track this task so I am relating these two and then Ill close this issue +This looks like a configuration issue +William looking back over your comments I am not sure why you call using explicit parameters messy +Removed use of number from since its no longer used in socket transport +MaryannLatest patch looks goodIn the future please increase revision number for newer patches +There is definately an issue +New patch updated after jQuery Patch for language file loading +Im going to commit this tomorrow if no objections come forward +So this can be worked around but its definitely in an invalid state initially +Ive committed a fix to trunk in SVN revision r its a very small change Id like to merge it into the branch +I agree timeout behavior will be useful but we should somehow tell the AM that request either expired or was rejected in allocate response and not expect AM to ask again after time has elapsed +Sorry Im wrong about +Since uses the Hadoop HTTP servlets it has a conf for service config dump already +esbserver is not meant for deploying of ears and ejbs +However Ive noticed that the new Hide lookup field link doesnt maintain the lookup parameter for example I search for the events of a facility I see them then I click on Hide lookup field the new calendar screen will show all the eventsPS you should really use the data model for your custom calendar application +An has been implemented and added to the list of filemgr versioners +Thanks for the reply I updated to Xalan but the problem still exists +It would be possible to use XML schema to do the enforcement +Edwin Can implement this for at least the tar and jar providers tooIm applying the patches but not closing the issues until the rest of the providers are done +META +Fixed typo and test in previous patch +resolved on change of shipped and resolved. +Can you do this Doug +Why? My comment was from before recent sorry got my dates mixed up here +Its a little obnoxious but we can turn it off just before release +Do we honor it? Parallel it? gwt log +Patch forces inlining for both x and x Linux platform and for WinPatch contains following changes Share private structures of hythread +Christian do you have a link to that load algorithm being used? I would like to add a note where to read more about it from the source code +I just committed an alternate fix that works for me on Sun Java but does not require an unchecked s fix works as well on my Apples JDK I dont change the issue status anyway to resolved until Elijah doesnt confirm that tests pass on his compilers ve just verified that chain now builds in all of the environments that I tested with previously +I raised for the same which would be duplicate of this +This is my that seems to fix the failure +Phase part patch has been patch removes references to and replaces it with +This would allow things like including or something similar +v +I did not find direct dependencies for Commons Net +Just a question wouldnt it make sense to add this to too? Or is that for static into only +move resolved issues to so we can generate release notes +patch for trunk +Im out of ideas. +i expect something to happen +My only worry is that it will depend very heavily on the fact that we have the old job status +Im not worrying about using a conditional for the +can i get the task now +Changing the message type to use SERIALIZED instead of XML makes a dramatic difference in the overall time spent to process the ll have a look at the serialisation and see what can be done +Im going to commit this later tonight unless anything comes up +In particular they can be used as in the example above. +Ill update the bug details to reflect that. +Try using xD instead of xAD +These scripts are executed with binsh and not with bashSince all scripts specify bash +The packages should be gone now would be great if you could try out it should be in the maven repo in the next couple of hours +I did a quick diff of our rubytest dir and bergerspec and there are certainly a number of changes +A couple of points +Patch with the changes as suggested by Alan +This is the test project wherein the menu and menu have dataTable and both have references in dropdown and menu is called the List object of Table and table are called +The problem was that the was closing the output stream and the wrapper code was trying to close it again which would sometimes generate theThe new servlet is more careful to handle errors including setting an error response if something goes wrong +I seems like I cant edit my reported issue any more +Comitted to CVS for further comments +bulk move of all unresolved issues from to +The retry handler is configured by specifying the parameter on the action +Please Ray may you have a try and let me know? TIA. +Verified that is set after running a sample job +To support in a clean way I added a new interface called Routable which is implemented by both and +Issue still exists in EAP thanks for the update +From the error it sounds like youre using cqlsh not the cli +This needs fixingThis should be fixed by +But thats okayI am +Given that this reads proc this patch likely wont work on anything but Linux as written +I assume namespace masking isnt something that is done or happens a lot in actual production docs +Vitalii Not sure the is valid +But youll still need access to stuff eg +Ive changed to ignore the finish times as they are unpredicatable with of tasks now +Interestingly the patch applied almost cleanly for me just with little offset +In fact I only found the problem after taking away the remove in the for your help! It is +Sorry for bad screenshot quality +So if you wanted the form errors to appear along with the appropriate form on the page there wasnt any way out of the box to use it +Modules has a redirection facility that were discarding in +Move to to +what is the isIsolated mode stuff for +We were tyring to improver performance of castor so thought if I can try replacing the parser will xml pull parser since they have the will try it out and post the results here +No looking at the download information it stopped downloading then checked the checksum it failed so it retried the download and kept trying and only got as far at the G each time finally I just need to find a way how to test that +However if you do use the two patches should be compatible as long as you do first and then +Another try at the patch with a unit test included and a method refactored +You may want to run it through a final test but I feel that it is ready and if there are larger problems that arent apparent we can always roll it back +verified by Vladimir. +in this avoid copying the entire key just to compare it +Rebased the patch and fixed +This said I still think itd be nice if we could somehow disable those tests that we know produce a lot of noise and only enable them on a build plan that would run with a specific prefix or with an entire different sink +If you have some other patchsuggestion in mind by all means do try it and let us know if it helped +I will investigate if there has been a stdlib update that solves it and if that solves it for us +Oh well the version of the RI i used did not include the fix +Julien fixed the problemThanks for your support. +Please review +The patches the issue in a relative ugly way but effective +That code should be updated too when service tasks are added +canceling due to findbugs warnings +Same problem here +If you are concerned about long stack trace I can remove it +The applied patch does not increase the total number of javac compiler warnings +If there is something specific that cmislib isnt doing please create a new issue and specify those details but remember that cmislib is spec compliant so if your request is repository or vendor specific it will not be implemented +The same logic must be applied to the and the to and the same logic in Rev +Will it still be possible to manage older Geronimo versions with the plugin +This patch should fix the bug in v +I feel like the feature is nice to have and the patch OPs is apparently too hard to follow +We can do the batching optimization after this +But the total code coverage shows a wrong value inHas anyone faced any similar issues +Regards sync aint there some in the layers above this method? Thanks +Closing as fixed +I guess we can close this. +LGTM thanks for tracking that down +If not we can start with a new Jira and iterate thru problems +The directory listing case was a clear reason to do so and I can see s thinking that we should match that new behavior and also Javas convention +We remove this memeber and make member ofOne instance will be created per at start +The patch was applied at r. +Therefore the change is backwards compatibleThis PR contains commits one which fixes and implements the direct skipping mechanism one which exposes the chunkSize as a JSON and AS configuration property allowing clients to their ISPN binary storage if needed +here is the jdbc driver I tested with +To be a bit clearer The comment i would leave it as is was about the question of whether or not we should enforce lowercase urls for everything +Thank you for your quick answerWere using the JBOSSOn my side Ive just made it work +HiThe interfaces from the super classes are already included I thinkI did mean to include the super classes one way or another +I am not sure which method is is there a better place we can add Macro documentation within Maven itself? Most people are unaware of Doxia +Werner I perfectly agree with you +Cache COMMONTERMS in configuration instance +It would be nice to have an error rather than null if possible though even if null is just replaced by Error Check the server log +No particular requirement on this JIRA at this time +My account has been activated +Fixed now all threadlocals should be null or contain soft references +Bulk close for release +the problem is actually fixed in +The failure was an environmental fluke +Perhaps thatll encourage people to read the docs about it not being analyzedNot sure matching vs searching is a commonly understood differentiation +i can reproduce that the default setting inside the administrator need a restart we will fix cant reproduce the problem that the setting inside the is ignored this case work as on multiple osrailo you see the inside your debugging output? pherhaps it is not invoked because it is on the wrong postion or you have disabled inside SettingsApplication +is what allows you to then send in values by system prop which is much easier to do for the getting started demoAnyone deploying to production doesnt need to pass those sys props +Hi Masato thanks for the patch +Encrypting it with some key would not make it any more secure IMOIf we need to encrypt the password itself it seems like the key used for that encryption should be configurable which would have to live in a somewhere and we kind of end up back at the same problem +Missed change to pom so that javadocs can build +Addressing the topic is possible but Im not sure its just the static initialisers that we need to consider +Actually that is probably fine +Thanks I may look at this over the weekend but dont count on anything +To further clarify the dependency ideally should not be on the implementation but rather on some kind of interface exposed by the shuffle service +committed +The bug was in version of runtime arithmetic helper which did not handle the corenr case properlyFix regtest is attached +This report is incorrect. +Committed patch +so existing velocity users must such down the freemarker jar just to use the config browser? +Im concerned about the potential impact on users of this methodSwitching from throwing a to a runtime requires users of that method to revise their exception handling strategies +Screenshot! +Please take a look at the others and check if we have to do something before cleanup has been done +The large tests fail because a interrupt status flag is set in the thread that the hbaseminicluster is shutdown from +likely environmental but moved to correct place +The trouble is that were not using results of the build +Tested this with the b option and script service now worked fine +already answered in the other JIRA +Clicked on default then IP ranges and here I have this optionSorry for wasting your time! +Thanks Pranay Arun Done at rAshish There is some problem coming in order completionThanks Ratnesh for reporting the issuePS Arun please provide the fix for this ASAPThanks! is a patch to fix issue of order completion from order detail page +All requests are using the unix domain socket +its been fixed +When I ran the tests I noticed that scriptbrokenpipe was failing but this is supposedly a known intermittent failure so I didnt worry about it too will take a look +Why two web apps should share the same http session? I dont think the difference can be done by having those inside the same ear If you need to share some obj consider the usage of a distributed cache or a datagrid. +No I do not have any secondary indexes on any of the column family and I have done the fair comparison and seen some performance hit in utility +Also and are involved in this process +Hi Vamsi the patch looks ok to me +srctestjavaorgapachehadoopyarnservernodemanagercontainermanager +jdcryans it looks like was not included into your trunk commit of this JIRA +Still in it seems. +I agree with Emmanuel +Can this be in Andrew? Id start a candidate immediately after release +Superfluous records are now ignored but a WARN message is written to the logCommitted for in rev +Merged into master branch +This is a dynamic guard check generation bug +Should we open an issue to jclouds to discuss or this is the expected behavior and we can move forward with this patch +But its just empty +These issues have been resolved but never closed +We used to eventually run out of RAM heap analysis showed this was the problem +Test failure was from an unrelated change trying to fix an issue I ran into while testing +Ok ill check this later after i prepare my Linux enviroment for testing +To clarify the historical record we reverted which was causing the problem +So my manual adjustments were doubling the offset +If there is any commands we need for usability reasons we should just have a specific issues for them +Have been job hunting and will now be relocating to LA from LondonIve just rebased will attach a patch +It deals with RPC stuffin future it could be used to support chained request processor to support like authentication handling +The problem is that when the original page comes up after processing the action and adding the record to the databse the form is repopulated with the original values +This is actually the session is transacted the acknowledgement mode is is not allowed for ve tided it up so the debugging is not confusing and it throws an errorfor an unsupported acknowledgement mode. +For those following this JIRA darran will provide a version with his patch applied for testing +That is out of scope here this jira is only to build the dynamic functionality the code is still unreachable because the is not available from the switch statement in this patchSeems fair I created for my request above +I do not believe its right to automatically coerce Symbol to a Ruby String since they are very different types +derbyall and junit suite passes +The patch sets the auto in SYSCOLUMNS t to new value provided by RESTART WITH +Yeah youre right! The text option only seems to work on large documents and even then the tail end of the text gets clipped +If you do not set this property ij will start with an error indicating that the database was not found +Verified in +I dont know why the form content is being written out at all +Done +Im surprised that there are so few headers that require config parameters but there are still some +Feel free to submit a patch here +do you need an test that casts a string to a timestamp? EricThank you for your reviewing +Attaching a unit test and aThe net of this change is that in dont check to see if a field is null if is a field as the value will get populated later in the flush processing +That is a very buggy feature and we really should be discouraging usage of it +does actually give us a error +I have some questions to this project can someone help me? IRC or instant messanger? Thank You there anyone who can volunteer to be a mentor for this issue +I will sync the latest version and redo the operation +Agreed +would need to be updated and would need to be enhanced to support arbitrary ID strings but still support integer +sresubmitpush the submit patch button again for +Heres a patch that includes a test case for this feature and also changes to the source code to make it work +The bug is here missing breaksYep good catch +Thanks! that was fastI should make them all this easy. +Marking closed. +and i assume that there is a way to turn that off as it should be possible to configure a custom error page +Some of the work done in will likely need to be backed out of the plugin when this is completed +I dont have an SVN version of so once I am outside our corporate network will download and submit a svn diff +Updating issue title to reflect symtoms better +Thanks Tim! +This has been fixedfor. +Thanks. +May be we need to update the doc with clarification +So is there a change done so that types can either go into package name based namespaces or to the wsdl schema targetnamespace +I looked into the log and the timeoutmonitor picks up some regions because the region server died and these regions were opening on it +Mike I guess I mean id prefer UCA order which isnt just the order codepoints happened to randomly appear on charts but is actually designed for sorting and ordering things +Just curious have you seen substantial performance gains because of this +Im using and jdku on winxp +bq +This patch logs the error when checksumOk is sent more than once +patch Updated to include +Patch is missing the unit test and overall the patch has issues +This should do it +Verified on JBDS. +Good idea Ive seen this bite a few users before +And the regression still happens on trunk and we cannot accept such state on version +Its actually kind of broken at the moment the last round of development on it left it in a sorry state +The failures are recurring issues unrelated to this patch +Matt or Dave +Committed to trunk +As has been closed +The attached codeif you run the separate method will crash the master about of the time +The patch at RB looks good to go +What do you think +We can try to add JSF facet but not sure is will work well everywhere +Not yet commited on the branch +Networking is not supported in version of the Enterprise Platform +It includes a regression test which fails without the patch and passes with it +Thanks Stuart for the detailed answer +on all your for community module breakage i take it this involves geosearch closing all issues that have been in resolved state for months or more without any feedback or update +bq +It also makes both the ij invocations mentioned above print Portuguese messages +Thanks Eric +Jay please test and close if this fix makes things work for you too +No this totally alter the semantics of +jars for test guysI have uploaded the jars for those who believe they have this issue +It looks like things could be easily improved by renaming booktoc to this is a known issue +I have a good solution to the problemI reinstalled the original binary package of subversion from so upgrades should work without conflicts or other problemsInstead of compiling a static version of native SVN I downloaded svnkit extracted it to the folder and used our ANT coniguration to use the CLI versions of svnkit to do the tasks +This new classes were introduced by but I ran ant which didnt seem to catch it not sure why +HiramWith help from Holger Bruch on this issue I think can be resolved +I just committed this +When local CRC files are enabled a get followed by a put of the same file should not fail but apparently it does +Committed. +should be able to comment more on it. +It is commented because it currently doesnt run to the the limitations of test is also commented as it doesnt run until is committed +Several minor bugfixes are also included in this patch + +Tried to see if adding total number of Containers was trivial change but looks like there is no notion of application max resource available to resourcemangaer +Therefore any xml tool or parser can be used to validate against the generated schema +This is overridden by cloud providers who either dont return hostnames or use images do not assign the hostname properly +just picks up the INFOURL and adds the taskLogs links column +Trunk patch +I can upload diffs with the code however I think the solution would be to not have to resort to creating interfaces for everything +Done +Isolated test case in a zip file please +Will try to get an updated patch out soon +Please review the patch +Please add this to the appropriate Fix Versions +I got it working now +Computation of the hash code costs almost as much as copying the data +This will send the full stack trace +Verified in Affects and Fix in Version +Alessandro Do you have commit rights for Stanbol? I dont see your name in the list of committers here in JIRA +The failure is due to commit job not being called +Another javaee folder is not necessaryThe trunk is aligned with Geronimo releaseAnyway its good to see these new java ee features there now +I didnt implement a customcomponent +Used the snapshot +Im of the way through the patch but it will take a fair amount of testing when Im done to asure that it works with both andIm convinced that if we dont do this now its never going to happen which is why I think one of the exit criteria for needs to be either a providing wrappers and a clearly stated EOL timeline for the namespace or b changing the package names only +Ignore my previous comment your change looks fine to me +A month later it is not out but is still imminent +AlanIn the patch I created a constructor with a second argument indicate whether to overwrite +Hmm after last Solr upgrade my application stopped working +would it be possible to use the jstl el evaluator for this? +Why not log the exception? You keep a to save on construction cost? Could you use a instead of aOtherwise this patch is great +Yes please I have the code written and I even tested it a little but Ive been able to spend no time on it in a few weeks +If so then maybe we could resolve this issue as a duplicate of although we should still commit the unit test +Note I only wrapped the test which was hanging on DB +If the priv block is used more than once in the same file then I agree it should probably be a methodFor a single use I guess its just a matter of preference having the code in a separate method means that its harder to read imho since one must look elsewhere instead of the code being right there. +It seems to be caused by having a mirror defined in the the attached project for an with no mirror defined works correctly mvn package s +With regard to testing with had to modify the compatible versions in jbossjbossws to avoid versioing errors during the build +Let me know what you think about these +Right now the communication is JMX believe these should also be split into separate responsibilities +xmlsrcjavaorgapachecocoonutilsrcblockslinotypesamplessrcblockslinotypesamplesstylesheets +I think we need to calculate setting height of scroll area when list has more IE CSS problem overflow auto the table inside must set style width +Just realized that this is a patch +this has been fixed in MVEL so should be fixed in drools when MVEL is released and we upgrade. +Can you post an example of a filename having unicode characters +My tries manual recursive folder removal method rimraf npm package setTimeout to let files removals endsNothing resolve this issue +Arent you supposed to use a user in that case? Like anonymous +Thanks for the quick for the hint setting seems to be doing a better job cleaning up. +Standby will not be writing to edits directory +Changed to normalized paths in revision +Attaching patch a which makes sure the transaction is either committed or rolled back +I was just making you aware because it might just be easier to do this as part of the larger change. +If we can verify that this issue is indeed fixed in supported bits as stated on then I will close this JIRA as well as +I agree with Darran +Example application +feel free to close +I would do that git bisect Gary suggests but would have a low probability of finding the culprit +Closing on job successfailure to release connections +Sure will do +This is tyipcally a configuration issue or a problem where you have an older conflicting version of groovy please query the mailing list or post your environment the GRAILSHOME variable +cant remove this class until post but we can eliminate its use from within our own codebase. +Fixed the javac tests are not related to this patch +Patch checked in +This cache loader exists adding unit to be documented +This looks fine but again I dont think this makes much sense in Avro indepdendent of Id prefer to commit this and all in a single commit +Can you please try again with the attached patch which should provide a more informative stack trace? Thanks +I do not see how possible to enhance this +moving issue for better resolution in +Fixed by +Can a new jar be created with this file in it and can I grab this file from somewhere else in the meantime? You were looking at the is what you need +added test for and one? can you be more specific +On Spring Security side to handle the CAS feature the dev lead asked me to build a solution based on timeout instead which is still pending waiting for the new CAS client versionTo draw the big picture CAS server will be added a LOA support this year so after that it could be the right time to build something more complete on Shiro library. +This fixes the equality and adds a test +Also Environment should really be on the Basic tab since thats information that should be on almost every bug it seems reasonable to me + +Jason you are correct +Lucie can you attach to this jira +Moving to +That being said do we really need to keep this issue open? It seems more like a user error than a bug +lgtm +Need to add a few more retries since first few happen in a few seconds +I have attached a simple class to show the issue +NET clients aparently send Date Time values with digit millisecond values but Axis replies with the millisecond value trucated to to what i heard this issue is common to all the java clients +Feel free to as often as needed +I dont think the comment block is a good idea it doesnt cover new languages so the merge problem would still be there +You could try deploying a large number of snapshots in rapid succession just to see what happens with some volume of SSL connection attempts +Im going to try to get something for +You may want to use and use instanceof to check if it is the expected exception +I commented with one additional question around performance tuning but I think this can be closedThanks for the swift response +Its OK +Whether they are actual snapshot versions or not they clearly dont fit the standard Maven numbering more to the point they break the version number suspect some of them were accidentally did not find any ones with recent dates so I suspect they were uploaded before Maven Central was better protected by staging software such as Nexus +This is actually a larger issue that Ive finally tracked down + +The issue was probably introduced by the update on this +Thanks MattI have applied this patch locally but since it requires a changeto both struts and Validator to work I cant commit it just yetIf I apply it to will no longerwork with nightly build of Struts +Basically im not really sure that and are about the same problem +Unfortunately Ive been caught up with other stuff this week so I havent been able to implement any of it +Now we compile +We may adopt that approach into James but temporary files in James created by are cleaned up during normal operation as you could see by watching the temp directory +I found working revision was updated and original version is more old than Seeing records I started translaton at OctI will see log around that point and find exact original version +When ajax request is performed inputDate field isnt submitted +Im almost done getting an initial patch for this just one issue remaining +bulk close of resolved issues. +But it would be better if we change the name to +QE will continue in testing after both files are in place +The source file needing patched is at sitesrcsitexdoc for the patch +The issue is that the Resource and the actual file get out of sync so when it uses the Resource to determine if the jar is accurate it gets a truncated file exception +Ran into this earlier today while testing with Confirmed that your workaround fixes it +This actually isnt documented anywhere. + +I think it would be a better idea to just check +Use of a instead of a concurrent form +needs test categoryRight +Attached is a patch which explains the situationNo new tests because this does not modify the underlying behavior of the system +Quite subtle only happens on a secondinclude to a page where a closing nestedroot tag puts the wrong nesting contextinto the worksThanks Tibor +This is known to work in every situation Ive tried +This Error is due to WSAD +kudos to Sebastian +Ill then create a new patch +Marking for instead +This was fixed by. +transition resolvedfixed bugs to closedfixed +No test cases but I verified by starting up a cluster and doing some reads to make sure the counters incremented +I wondered whether there has been any news WRT this issue yetFYI Weve been using JAMES along with our own SAR inside Phoenix for a couple of yearsBut now we started to use and Eclipse for our own components and servicesNaturally weve some interest to integrate JAMES as an Bundle to have them all running under the same hood again +The file should probably also be updated +Sebasitens workaround caused some regressions on Windows +that work is well underway with AS all previous community releases are +I attach to replace This version incorporates an version and should be applied after the patches for and +At least the ones could be fixed +Sadly Reflections has no logging support so I am going to have to add it to see if is the source of this strange behavior +intowicketupwicketup we do wicketupwicketup then i can already hear the complains again how can we change that +client reports to NN once NN finds that ALL replicas are bad NN asks to verify +Closing stale resolved issues +Resolved +Patch applied on branch and it been fixed in Betad build or shall I wait for Betae for verification +Added Utilization widget for displaying both HDFS Usage report and Jobs Data report +Test case adopted to the M a lot for the test script and ambitious debugging +On top of this to aid with debugging and development of the fair scheduler in the future there is a scheduler event log which is disabled by default but creates some event logs in format in fairscheduler if you turn it on +when there is a space in java path there will be an error fail to start if using but not +Let me quickly check. +If someone has time please review my approach +With the fixes went in for Disk Fail Inplace we can restart datanode with failed disks until volumes tolerated is reached +This patch should fix the bug +I just committed this +for the to trunk. +I just sent in my contributor agreement to Apache +Out of date +I had to apply another patch firstCommitted to trunk. +This jar file can be exploded and imported into eclipse as a project +Nitpicky code formatting issues a The pig coding standard is now to use spaces instead of tabs +but otherwise table looks fine for me to have in docs +This was fixed in xwork although I would love to just get rid of the profiling thing. +testcase for this issue +Anyway thanks for pointing me in the right direction +That seems to have slipped in during the upgrade to metadata +It works by adding an empty line between each sentenceI guess something must be fixed here + +I remember speaking with Jan about the problem of providing a custom name iterating schemeWhats the suggested fix? Would introducing an interface help here +postpone after beta seem to have done a good job on this issue its not clear to me what is still not working +Now actually has everything in the patch +classpath already contains the lib part so I only need to add the dependencies to the jetty runner project +Yonik if you arent planning on those from scratch would you prefer a patch or an explanation of what I needed to change +Although there is a workaround Id rather not have my clients deal with a mess of shared in order to install two on the same server instanceThanksIll take a look at it forColmCould you try with this jar and let me know if it works? Im just trying something out from one of the otherColm +This has some fixes +but in general i think youre right +Can you point them out clearly +Really nice catch thanks Xian +trunk +Hmm +Tests ran cleanly for me on +Ashutosh Can you please post a review request? Thanks +Block branch too +Needed to touch this code when working on other issues +In trying to track down the bug I have found that the OracleThin Client will give back the result message No Suitable Driverif you cant create a connection to the Server +Then I would recommend to update installation guideThanks for help oh! The README is actually replicated from the one in SVN so its better to fix the SVN source +I can reproduce this problem without binding the plugin to the verify simply invoke mvn and I get the aforementioned exception +Its fixed in Hibernates Branch and in Beta the fix will also be in the next maintenance release of Red Hat Enterprise Application Platform +Could anyone confirm please +That doesnt scale +This is obviously not an issue in mahout itself but a hint at the may prevent others from running into the same problem +I will apply it +I dont think it would get longer from inside Im afraid that an external script would be more complex to maintain than an Eclipse extension that iterates on all examples to load them. +The change looks good to me +Except the Autoscale one other files are checked in +I havent gone through the code so maybe its obvious there but why would rpc solution change the runtime of vertex input superstep and of setup +Todd would you like to work on this with your idea +Hi Dennisquick update could you post this to +I filed to make compaction policy pluggable and configurable per table +If the attribute is set to false a base class will not be scanned +One thing I could not quite decide on one way or the other was discriminator column handling +Pi sounds good +Nothing should block a release short of failures and security vulnerabilities +This is very nice improvement Sergey +Thanks Yongqiang +For the runtime setting of partitions we used to estimate based on the samplers input this patch will instead estimate based on the next jobs input as for they are different +Same issue with +patch looks goodsvn is currently down +We should do some thinking about the nature of minimummaximum occurrence constraints relating to this bug I understand this correctly we need a notion for This child group is optional but if it is present it must have at least n elements correct +My point is that the cassandra users home directory is varlibcassandra so it should already own it when the package creates this user +This is a new patch with some fixes for using including more consistent hashingIt also obviates the need to fix and guess the patch missing file +OK go ahead and make that change +Its a lot easier to read the code now and make sure that its behavior is consistent +The choice of the start prefix was reflecting that the cluster nodes were being started but obviously your choice is fineOne thing that might be worthwhile adding is a HADOOPHOME setting in the AMI perhaps with a PATH update too +and hopefully get to it for +Still there are issues like that may still not work +Thanks for the patch +This version works particularly well as a companion to the solr flare version +Ok so this is +That makes sense to me +For the purposes of this definition submitted means any form of electronic verbal or written communication sent to the Licensor or its representatives including but not limited to communication on electronic mailing lists source code control systems and issue tracking systems that are managed by or on behalf of the Licensor for the purpose of discussing and improving the Work but excluding communication that is conspicuously marked or otherwise designated in writing by the copyright owner as Not a Contribution +branch Committed revision +is modeled after filterFair comment about +Figured out how to register event listeners by retrieving the from the parse stack +Dennis ignore the above comment +I updated the and I added the fixes that were already done to the quickstart with the same code base +Actually you can check the latest build of in both browsers popup panel appears stretched to the whole width of a page and its the right behavior +The issue can be if someone wants to pursue itThanks to Knut Magne Solem for the prototype +Sorry for autoformatting I will upload the patch without itSince we use we do not need any more because the text will not have such a code point +It looks like this is fixed in the release candidate +You wouldnt +a figure used in the doc +the search field and the correspondent label have been made +I am not using Googles App Engine +classpath fileWill update that and attach new patch tomorrow +Committed revision forI will leave this issue open to investigate whether we need a change in README in the next release +I added a log printing the bind address with the scope feAnd this is the address tries to bind to feIf we remove the scope it tries to use a different one and it fails with Cant assign requested +Justin the attached patch is the same as the pull request +Andy thanks however I dont know where I can see these commits + it seems that meclipse is messing up with eclipse configurationThat would be easy to fix but unfortunately meclipse dont change any Eclipse settings including proxy +Just a question you have provided both +yep Michal had similar observation when he tried richfaces portlet based on portletbridge archetype +I dont think this is a webwork related bug but a config problem with DWR +Hi Stuart any updates on this +Updated and verified for GA +Im not sure why generator needs to know encoding +The other could be a run script now command +Heres a new version +Can you please use svn diff so that it includes revisions against which you are patching +Thanks Joe +derbynetclientmats running no failures so far +New patch addressing the comment +Im not a ruby user +Hey thanks a lot for doing thisLike I said I get the same error when running tests on trunk in mode so its not related to this patchAs for the token selector I might be wrong in deleting it +I have solve the questionIf someone wants maximize on edit mode he has to change property maximize of bean from false to true in assembly. +Hey sorry spaced on this one +The patch contains the followingM toolsreleaseM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTestsmasterD javatestingorgapachederbyTestingfunctionTestsmasterD javatestingorgapachederbyTestingfunctionTestsmasterA javatestingorgapachederbyTestingfunctionTestsmasterjdkA javatestingorgapachederbyTestingfunctionTestsmasterjdkA javatestingorgapachederbyTestingfunctionTestsmasterjdkA javatestingorgapachederbyTestingfunctionTestsmasterjdkA javatestingorgapachederbyTestingfunctionTestsmasterjdkM javatestingorgapachederbyTestingfunctionTestsmaster + +Alexey the new patch applied at revision r with patch for please verify the problem is fully resolved as you expected thank you very much for the enhancement +Merged into x r think there are some issues to look at here there are a variety of non thread safe accesses to shared variables +Tomaz your last comment indicated you planned to do more tests so I didnt want to resolve this without your ok +I should have a unit test for it already so youre probably in a specific case +Ive committed it and fixed it as resolved. +Test case attached as requested +Looks great to me +Oh didnt think about it +Its just not a good idea for exactly the reasons weve been discussing +Closing resolvedwont fix issues. +It seems that fix is ok. +Reopening as per issue workflow decided once the issue is fixed it goes to the Resolved status once we release a version we close all the issues related to curent version +And the serialization fails it again casts there to +Ok I did the same test on a jetty server same issue it failes on the mac and on the singlecore linux based one it works +I agree with Anoop and also now we are making the clientscanner also retryable +Like Shankar said we should only add a dependency on a logging package ifreally neededWe could consider making more use of the logwriter on the datasourceOr optionally provide a aspectj addon to do more detailed tracingloggingBut thats another issueenhancement +Bulk closing stale resolved issues +The question is whether this bug is bad enough for us to polyfill it +I change and enabledisable beforeafter each test +And now writes a warning to stderr for the default crypt key. +mrepositoryorgspringframeworkHOME +Stefano already committed other versionsIll leave it up to someone else to put this in if anyone dares +Sveta what do you think? these updates arent scheduled within the next month +oops I closed the wrong issue +done long ago +We would want to make sure that indexes dont break if you lengthen a VARCHAR column using ALTER TABLE ALTER COLUMN +Removed IDE files +Can you please clarify what you mean by stop a node +Here is a first draft of the Design and Feature paper on hadoop file system +The only difference between this run and my last run was that this time the JSF lifecycle made it as far as INVOKEAPPLICATION before the thread abruptly stopped logging messages +Code snippets for test case and sample implementation of you saying that it is possible to shoot yourself in the foot then +Logged In YES useridYoull also need to submit a testcaseotherwise this processing could be brokenwithout anybody realising In YES useridThere is no change to the existing behavior so I dont see aproblem with including the extension. +patch file has a change in it that reflects a change to the constructor of in my code base +Patch applied to branch thanks +Patch for enforcing strict method and property first glance i see two problems I recently fixed the issue of if statements being ignored because one of the sides of the comparison is null or +According to your log file and to the creation dates of your snapshots I suspect that youve in fact encountered issue Sonar doesnt support parallel runsanalysis on the same project + in r +Verified. +t see any problem +This is very distracting +Patch to fix the specified i think the issue is that we dont know what the remote client is so we dont know that the remote client will have the specific exception class available in its classpath +Could you try it again? Thanks +Attaching and making content formatting more consistent and removing reference to Java SE Will commit tomorrow unless I hear otherwise + build for +Any upgrading of a bundle is unsupported +Correct me if I am wrong but it looks like these are only useful if I have a Task in my job which needs a container and the returned container will only be used by that task? If yes That is a very narrow +Requested an update for this in the +Thanks Navis +I think the problem is that it all queues behind the prepare build steps +The simple example with calendar doesnt work anymore as well +Spec changes are not a one off thing +Publishing to the mvn repo requires some more workI will be workin on this soon +But from your explanation that would not be correct +It appears that the contents of testscpifuse does not contain a servicemix entry will create Fuse based upon Fuse general design +We use it with Continuum and if thare are failuires in maven the correct erro code is transmitted to must set MAVENTERMINATECMD env var to on before to call mvn from ant +Thanks +Submit again +It was not reproducible on CI or CLI. +I guess i will try in production as soon as we have a window +I pushed the fix + +except had to be nested in try +RM starts retries after the expire time has elapsed +This updated sample now produces or more executions of the sources jarThe only difference between this version and the previous sample is an additional inheritence level +I patched a similar issue with corruptmissing EOF flags in but that patch doesnt handle missing in the middle of a file +These changes seem to fix the problem but I would like more verification before I closeSending modulessrcjavaorgapachegeronimojeedeploymentSending modulessrcjavaorgapachegeronimonamingdeploymentDeleting modulessrctestorgapachegeronimonamingAdding modulessrctestorgapachegeronimonamingdeploymentAdding modulessrctestorgapachegeronimonamingdeploymentTransmitting file dataCommitted revision +Also provide steps to repro in case you are able to repro +Created in Prepared data in Prepared data in at rev Thanks Vikas +for the patch +Attaching a unit test the demonstrates the deadlock is still present on trunk code. +Its fine with me if I can somehow manage to export the confluence space to a more user friendly format and host on my own web space at I wouldnt mind a vhost like either please let me know what you thinkpreferusually please use was in the wrong format +Yes I think thatd be perfect +I thought needed the same fix as well +for in rev +Enabling dns caching in nscd etc +Passes fine +I was thinking wed implement it as follows but I am very open to other ideas Have a digester open for each file and all bytes written to this file are added to the hash +sql and adjusted the +The improvement you suggest sounds like a good idea if only because the code is puzzling +Applied thanks new snapshot pushed to repo +But I guess so is the data block encoding +Suren the patch looks good to me +This builds and tests cleanly with maven with an empty +Also useful when upgrading from to +If that would be Java source directory Sonar will not analyze JS code if that would be JS source directory Java files will not be compiledEven the change from sourceDirectory to script which should be relatively easy to implement would be very helpful for the situations where we want to analyze JS and skip Java code for a particular module +is nearly comatose in terms of contributors +patch excludes jute package for jdiff +Hi luca now you can click assign to meand then attach you improvement patch thanks luca +Initial patch supplied will open secondary issue as resources that are no longer present in a repository should not be proposed for future all issues raised by me +Im testing the patch for +When the get is called the value of permission at that time is returned +Attaching latest patch incorporating Hemanths comments moved assignment of within swith statements +Resolving as Wont Fix for reasons mentioned in previous youre using a profile to provide alternate implementations of a service it makes sense for that config to be closer to related bean if the profile contains a property placeholder definition it feels wrong for that definition to be at the bottom of the config when it is normally at the top +With an int ord everything you add is going to collide and you need the ords in the queue added to the double lists and you need to fall down to the subord much more ofteninterestingI guess Ill go with it for now though +Thank youYes but I think it should +has actual stack trace +Thanks! Patch committed in r +Im not certain about the translation to think for pareixer +Absolutely +What I meant is that if we were to add that result boolean we would not return vnull when the row doesnt exist +It is currently not impossible +The intent is to declare such parameters as Boolean allowing for null values analogous to how we handle this in +Nigel and I discussed +Maybe if you could show your filter implementation it would help +We have reviewed and discussed action items that will be taking shape in the components added to corresponding wiki pages +I will change that to make sure this case cant happen +Reviewed the patch +Daran I created the linked to review the code and catch exceptions from that could kill the shared timer so we can keep the for this release +Committed revision in trunkThank you Eduardo for reporting. +The expanded argument schema for the udf gets used in typechecker +Lets see if I understand this correctly +Verified on master +Ill also revert the part to prevent cpactf to fail. +For reference this patch was tried to cause the INPROGRESS response in the case of a install when no other components would cause a responseThe response was sent but a subsequent query from the UI to retrieve the task list gets an exception +Then we should figure out what logic jclouds should follow and request a corresponding change in the project +All paperworks have been registered now so Gilles you can now proceed with the inclusion of the patchThanks for your included in revision . +Code is duplicated in both confirm the bug and the bugfix works as expected +Lets visit for the backport +This should be handled in the processDeletes method +I dont think its a big deal myself if it can be improved without adding complication thats fine but you can still simply view a as information to be used for core creation +Sure Ill have time at some point to post it we can use the one from Lucene +Best regards GeorgeThe fix looks fine +to issuesThat is where our JIRA notifications should goI am sure that the absence of the issues list and the request to send JIRAnotifications to devlucy was simply a glitch +Assuming that it is fixed and closing the bug. +Why should this be supported +By the other hand classes not implementing Serializable should have super classes considering their fields. +Updated the to use the new interface +I will try to look at this bug when I find some time I will not be working on this issue regularly hence am not assigning it to myself +Here Import wizards comes into Ill postone importexport implementation and create a configurable location to store now templates are stored in the workspace metadata so if team members share workspaces theyll have the same templates installed +Applied thank you. +If I have a pom say with both a plugin which is a SNAPSHOT and a project dependency which is a SNAPSHOT I cannot release until I manually harden one of the versions +Ah sorry I misunderstood your question +This appears to be a duplicate of reopening for release note status. +The links to documents are too code +There should probably be a configurable way for users to register wsaAction Service methods so that the server side will know where to dispatch given a Action URI upon request +We have second of unavoidable slowness so more per task is OKMy response wasbq +We are using String replacement to provide dynamic content +shape +It is not mozilla +The test passes in interpreter modedefine class firstdefine class second to get expected Illegal attempt to redefine class The test passes on RI as followsdefine class firstdefine class second to get expected duplicate class definition +Dont worry thats exactly what I had in mind when asking you to attach a test case +Re +Just going down the list of patched files +Im not sure this is an issue +Ted +Personally I would have preferred for the these logs to be at debug level +Making a true hierarchical configuration is certainly the solution because in the next major release of Commons Configuration all configuration classes are intended to be hierarchical +The resolution of componentnamespaces is an rats nest of issues that your report brings to light but I dont think this type of feature will see the light of day until at least Tapestry + +One way I did a test was generate X sstables upfront copy to temp place restart copy them back over monitor page cache trigger compactions with a real workload +Thank youBestGeorge +Still an issue in wonder if this was fixed for SOLR with +I went into the file andchanged the default transaction timeout to min and it imported thedata without any problems +I feel its better to keep in mind the use case +but Ill leave this open pending further discussion and testing +Can you upload the patch for the maven configs +New patch fixed the typo set +Or an specific symbol which has a default value of +Added sample to Metamer and updated Steps to Reproduce +fixed the issue +I cant reproduce the issue but have fixed a possible NPEAlexeyCould you please provide steps to reproduce the issue +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Thanks a lot for the patch and the testing Colin +But thanks for the fast response +The default constructor may present no tangible value but it isnt a bug to offer it +something in the www folder +another patch that adds the vertex to the aggregator calls + +All other tests passedThe test failures are in trunk are all existing failures +TomI think that the resolve cant handle even if their qnames differ only in case +screenshot of see the attach for patch +I think what you would have to do to deterministicly replicate this is mockup a file system that just gives one byte at a time +I have left out the jdbc driver for Ingres which is the file +Applied patch to to catch +I agree that they are less useful for the trunk once it is on +this one reverted the previous patch will post a new one +The only workaround right now is to uncheck Automatically detect for both Web and Management port and set it manually to and +Oh +Is that right Anthony +See the attached tick list +Ill update the issue once I know more. +fixed in r +Thanks +Is there any workaround to resolve this Issue +So it sounds like your approach is a partial solution but its the best we got so far +thanks joe! +Shane this looks a duplicate of can you confirm +on patchBefore integrating to branch please run test suite +Weld could either scan the dependency graph in advance and refuse to start the application or let it fail with at runtime +FYI patches were committed at r +Thanks for your detailed instructions BryanI think is not suitable to convert by extending even not suitable to convert to junit as its name contains ij which shows its expected to test as a sql script +If the number of rows of Ai is much larger than the size of Qi then that is very very unlikelyRegardless we will still have QQ I and Qi will still span the projection of Ai even if Ai and thus Ai Omega is rank deficient +One for and one for normal Javadocs with different configurations. +i also noticed all the mave hosted plugins and sf plugins has extends in their pom +Patch for y +This issue is going to be addressed in the next release of Derby see +Im not sure if i was very clear on the issue i runmvn coberturacheckeverything runs fine and my tests are only run if i runmvn installwith the check goal configured in the executions section of the plugin just like the example in the site the tests are run for the generation of the report itself and once for the check goal +Cannot reproduce on RF please this carefully in branch +Committed a change as indicated +Thanks again +Take care Folks Ive been testing the Messaging examples on a standalone and clustered application server with Messaging using the file included with this post against an ingres dbms all on linux +Ok i have a test but is not complete because the broker is remote and i need restart the broker manually let me see if i can imbibe and restart the broker inside the test +It appears that there is now no way in Maven to add a entry to the Main section +Patch to fix the +Hi Joerndo you have some experiences in this area to help us with that issue? Michael +Patch added +r +it happens only when two threads have failed to load the font properly which is seldom enough by itself +looks goodNit NSDSQUOTA and expectedSize should be the st parameter to assertEquals +I cant add ea to the JVM options because the Run Configurations UI does not work +Waiting on CI to clear so it can be merged to trunk +Moreover ordering might not be +OK +See further investigation in +Looking into the schema of SOAP faults to find if these nodes are actually meant to be there +bq +Done +Commit of did not fix thisRunning a test what Im seeing is that the root cause is the reported empty other note if the deletes are actually all in place then there will be no need of a repair script to fixup horked +Tracking work on Spring Framework trunk +HiAttached a simple project to recreate the problem as requested +AndreaThe best thing would be to create a utility class and put it to a package in the contrib module +another dev checked and chased it down to query plan changes caused by the update to patch since +lets defer to get fix version again. +So this certainly isnt specific to erb and isnt something new. +was a token +But here at apache I think the generated files also need the license attached. +looks discontinuedReopen if necessary. + +schemaName etc +Attaching patch a which removed the method and its implementationsRegression tests ran without failuresPatch ready for reviewI plan to commit this patch shortly +I just committed this to trunk and +i looked in the wrong testsuite darn +This looks like its related to against axis version +Not sure what is better +That patch has been applied in the dist +Martin can you try to reproduce this in the QA lab +Thanks for pointing this out +sorry I have no idea +If an envelopenew protocol between server and client we could fix. +Is there a standard place to put this and do we just want to check if the URL starts with a slash and throw right away +Thanks. +Im reassigning back to me till it is decided ifhow we can do this without duplicating the code. +It is is a workaround for this issueUse one of the following depending on your need for http or https instead of the does not seem to be fixed for RC +This refactor while itd be nice since as you say is a mess seems low priority but high risk since its something that cant be messed up since this class orchestrates servers in the cluster +Nope actually in this setup the default would bind to the interface returned by +can someone help? +Commit fcafeaadbffeaefa +Can you please close this issue i wanted to create a clone of this JIRA in but I didnt knew that I have to do it manually as cloning of this issue will create new issue in Geronimo project onlyMy mistake +one more deadlock problem must reopen it +to I just noticed that you didnt grant the license to ASF +This issue isnt a + related changes are now commited to trunkMore enhancements to generics will come to this is an ongoing effort + +I will be returning on Monday August Kevin +The other solution is to update to pull data from ZK anyway when is null or empty which means the contract is different though +Is there any updates to this issue +Patch looks good nkeywal +Agree +Its fixed nowb in the jenkins job +We do have that level of granularity with the Dispatchers fallback mechanism though so it should be easily resolved there +This will be a major stability improvement release so perhaps Mahout community can consider latching onto it and reviving the work on this patchWhat do you all think +There are no known outstanding problems +Just looked a little at the code the is then also a bit simpler because it only needs to use one flag to determine if the compare is case sensitive or not +this race condition is triggered when tried to shutdown bookie encountering +Here it is +It was released with the fix in +If no internet connection is available a message is displayed saying that the test has not been runFixed under revisions r and r in the branch and under revisions r and r in the trunk. +This means that I do not want to write into my progams hadoopisms +Im on changing the compaction thread priority as opposed to using a solution like ecapriolo has suggested +I cant follow same project from what? And what exactly is the relation to that shows a completely different error output +I will test after the next release +JeanIf I understand correctly youre suggesting making configurable the key that used to specify active this key is named +Do you still need this to be created +And the app looks to be application managed so the library may be overriding the library without ill effects to the containerTomcat also includes JPA persistence API classes but only a small subset +Committed to trunk. +Give me or hours and I should have it working +With plain EJB CMT no Spring transaction synchronization is active wed have to fall back to plain JTA synchronization through the +Im working on it and I noticed that in the online documentation the imagemosaicplugin tutorial is missing in the Tutorials module +This gives users greater freedom to handle invalid UTF bytes in their code if there is any and allows them to write orginal text records back to output files +This may be related to the lazy state but I expect its something more fundamentalIll dig in and either comment here or open another bugThanks for the quick turnaround on the patch +Load and youll see one input text box for updating the in false to see first problemthe update button will in d or any other single character to see second probleman exception will be thrown +I could confirm that this bug is fixed. +I agree it looks like a dup. +after migration completes +Comments comment issues addressedThe thing was due to Macs compose key Id typed a there +may want to look into getting the correct languages installed +Vladimir what ant version do you use +But when creating more than onethe exceptions occur +Are we going for the upgrade or not +I think every component we ship needs the change +huh? Ive been working for the last week with HEAD and havent seen the buganymore on a setup where it was often visible beforeWhat placebo effect on codeIf you tell me where to look I can check if I have your corrected code or not +We can still aim forPull back if you disagree +Have emailed Darrin and we will revert to direct svn check in to avoid problems in the future +Are you finding that using faceting for the total of groups performs better than nGroups? I was under the impression that nGroups was designed to perform better than the faceting workaround and in fact am in the process of converting an application to use nGroups from the current implementation +Also pooling will be optional +Hi MikeThanks for thinking about this proposal +IMHO is better to report MOVED PERMANENTLY as errors so the user can change the links +There was probably some deitrus laying around that it picked up +Attached a patch with file +This is weird indeed since the called lookup works ok and returns the values but then the st lookup does return any values +The portrange config can control how far it increments looking for a portSetting the default portrange to is another way of saying By default use the socket configs the admin specified or fail dont open sockets on unexpected ports +Note that this increases serialization cost for any tranfer which is pretty much most RPC +which makes no sense +Does this still happen? Is the assert from running regressions or from some other setup +The symptom was not caused by the checkbox. +AppliedThanks! +Marking as duplicate and closing unless there is an objection. +Logs now give meaningful information +Also since the official name of the mobile platform has changed to Java ME shouldnt we use that in the returned String instead of JME +This patch adds a new set of non core specific urls you can hitlocalhostsolradmininfopropertieslocalhostsolradmininfothreadslocalhostsolradmininfosystemlocalhostsolradmininfologgingIts a quick hack first patch +The problem is a rarely used path through store that does not do the conglomerate number to hex conversion +Attached patch which also solves an identical problem in toString +I have moved the ticket to AMQ as its a bugissue with AMQ +Alceu please create a new Improvement issue regarding your suggestions above and mention this issue ID when doing so +As per previous comment +it might be better not to have a mixed view of the world during the catch up period this could cause application problems if they need to assume a single arrow of time of edits and not wanting to see a partial world viewWith multiPut catching up should be quite speedy +As such it does not seems to be a priority for the existing CXF community +tmpdir can be configured to be in the working directory using build profile +I will append a patch which honor compare between principal from amq and principal from karaf yet not introduce any dependency between amq jaas and karafAs you said your code will honor Principal from any other jaas implementations Im not sure Activemq would accept it as it MIGHT cause security hole +patch looks goodI have committed this +I think write is needed for the newly created not being put into page cache +Logged In YES useridI think removing the warning entirely is the best plan +I see the domains there +Yes me to! The problem now is that I sometimes forget it +No problem +That potentially even eliminates the need for the simple access control provider. +This is very tedious and tends to get out of sync rather quickly +There was a bug in the patch we postedThis was fixed in commit effdcebeceefccdcecf in our repo on Jan thI can send you a patch privately or post the updated patch here +This class is gone in so when is eventually phased out this is no longer a problem anywayEither way make an addendum Andy? on changing all debug logs using this logger to tracealternatively on reverting with comment +I will be performing an upgrade on more environments +Increasing the updates per interval to k and having the scores calculated every ms didnt produce any adverse effects on a dual core machine but it increased the snitchs responsiveness by a large amount +Please test this patch using a couple of different unoptimized readers in the constructor +Kicking Jenkins to see if it shows any particular issues with the patch while I work on unit tests +And when I switch it back Im now able to see your output +Hi Daisywould this prototype work for you? Could you review it? Perhaps we might integrate it already to base further validation work upon itHappy for any! Stahl added new features to the patch that I have started to review internally +Thanks Vivek! +So says its fixed in + to trunk and +me trying it out again +Could you create and share a simple app that demonstrates the problem +Works fine. +Dropping it +Regarding this I think it would make sense to add pusetrue to the defined in it will also make it easier for users to start using message keys in their application without immediately breaking things if the resource bundle doesnt yet have an actual message for the given key +However iirc this popped on the mailing lists a while ago and I think the idea was to not put something like this in the http component because it has broader useWillem could we discuss this a bit on the devlistRoberto many thanks for the patch +In addressing this issue of whitespace around the ampersand being lost while parsing xml the versions of and were loaded in the lib directory and the problem still persists +WDYT +Thanks Alexei the patch is fine. +Hmm +Attached a patch for footnote positioned below the end of the page +Here is the server +Ran system tests against the h patch and none of the issues are found. +Applied to all branches +This has been +Modified avro schema which allows fields to be to generate the java code using GORA +Unfortunately the code is complex to study and I dont have much time to devote to code t there a code sample new layer sample or tutorial wizard or anything? the Web Terrain Service still in use these days +Shall we simply commit it given its been tested against several RDBMS +Yeah lets undo these dependencies +We can use these to compare against the random id that comes back and breaks the test +Closing the importer has been incorporated +jar files? Or the webtools type project +actually already fixed in shipped in +bulk fixing the version info for and all affected issues have in for the patch Stefan +This should probably be considered a doc bug +Hi Bela I am well aware of since I opened itHowever this is a different though similar leak +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +There are a bunch of places in code where we deal with a by waiting and retrying +How do we get a dump of the demo data +This look easy to maintain and deploy +configchanges was always exported in osdaemons module +Patch passes unit tests andAddressing review comments from Richard pointed out that that hdfs Path class constructor can fail on valid Uri like the format used for jdbc +Donethanksdims +Ah do you need the jar in a repository to run your test +It makes the out of the box experience better because users dont need to find much less read the relevant part of the documentation +this was fixed in by comletley eliminating the jsps and making the entire admin UI use request after release. +No response assuming ok. +am going to commit the patch since I could not fix findbugs issue from the generated code with a filter +I forgot to add this status should prevent the addition of further recipients but should not invalidate the mail the client must be able to complete and send the mail for the recipients that have already been accepted +I will provide a proposal for a later discussion. +Final chance to comment on the proposed fix +also passed after hours and GB heapI think its ready +Alrighty thanks for the fast response! Glad to hear that you are working on thisAlso noticed there are some annotations on the hidden files +Its not very efficient but I guess its the only way? New patch is thanksI think its generally a good idea to only attach patches for the issue under consideration +Done +Mind if I take this onnot at applied the patch +And it worked properly our cluster is repairing again and performing normally + +Sorry Werner for not having reacted to your comments but I forgot to configure notification for this issue +Ant tests passed on my box +Closing the issue as he main work is done +cant we fix this in the way that you fixed it earlierThanks I changed the default behavior to preserve addressing and if you need to remove addressing you could use the REMOVEWSADDRESSING property in the flow before sendingThis is how we handle security as well and it is consistent this way +bq +Starting with this patchQuick comments on the patch Lets try to avoid adding one more server +Finally +I got the openejb build but cant find a settings file +It would be better to go with themetastore approach +simple serivce class should be copied into srcorgjbosstest project s even odder is that you cant AddRemove this from the AddRemove dialog on the server configuration but that it tries to deploy it nonetheless when you go at it via the WTP is another one I havent got a clue where to start looking +It appears this is related to the solrconfig query option use +If the dependency plugin is the solution then also this will be solved +today this constant probably it will be called HDFSSNAPSHOTSDIRNAME since the idea of reserving it is because that name is taken by hdfsotherwise well end up in +Ive verified mvn clean install +to Weston since this is in the JCA JMS outbound adapter +Hi and sorryI thought that Samisa and I had clarified the situation with autobuild in the past few days on the mailing lists With hindsight user error was a very poor choice of wordingThe system is being superceded by ANT because no one users or devs have the timeskills to keep the system maintained +Verified in JBDS +Not sure what is the most RFC conform +I will keep this open for a little longer while I try to think of a way to make it better if possible +Anything we said would just be ignored +Yup were saying the same thing +fyi in a nd step ill provide some javadoc for it +This issue does not only affect requests that have a empty element but all requests as a wholeI have raised a JIRA Issue proposing a fix in the SOAP Action Based dispatcher so that whether or whether not you have the url of the request having the operation at the end your generated code should work if the soap action corresponds to the one in the WSDLRegards Senakasorry for the late reply but Im on my Christmas vacations +But echoing what Eli has said already we should open another jira and explore some of the good ideas that this jira has generated + Adds field validation for createDB and sqlStmts +I not sure exactly what the problem is you should still be able to chain multiple advisors together in the manner you describe but without creating unnecessary double probably duplicates where the report also says that multiple no longer work in for the test well have a look at it ASAP +Committed revision +Maybe even configure a static destination and kahadb for that destination that will be responsible for locking +bin fixes the following fields checked into trunk. +Range and State are both missingThe range is completely missing as we dont default it whereas the state is defaulting to ACTIVE +BTW I uploaded the test log because the last patch is blowing up for me +As requested by Oliver. +It should be a part of camel. +Will do once the ballot is through and the final version is release I will upload them. +Ive checked both patches +Change build files to use the artifacts changed the to pointto jUDDIs new and changed one like of so that the database is stopped rather than started at the end of thetest cycleWill close the issue out once the are available to ivy and the build changes are made there +I suggest that the file be removed not fixed +Billie remove it from where exactly +Sambavi is working on another patch that will do the two Alter commands +We then added a dummy call so that entry for employee is included in generated wsdl it worked schema was generated properly +I will make some tests in the application I created those days and give feedback asap +Moving out of +Merged into the master and branches. +Galder I am verifying the question in regard to the background timer thread name and the distinction from the worker threads +This feature is not turned on re waiting on IT team +Thats not easy to communicate on several possible and say users Choose your favorite one +Originally assumed to be a light restriction users of this package have demonstrated that it is a immodest obstacle to the problems they hope to apply it to +An easier temporary solution would be to provide a script with proper settings so that the users do not want to bother about setting the class path +Konstantin I thought it to be fine for that test case but I see now what it could cause if there were no defaults +Made the error case less draconian and allow subsequent lines to succeedOf these Im least sure about should we simply stop parsing the file as the original code as doing in an error case? Its hard to judge the differences between a warning case and error case +I merged this patch to. +If there is a cleaner way to plugin the lifecycle support to a let me know +Does this and the minutes +First issue is solved +This group could develop hard for the rest of the year no problem by the looks +We need to reopen too then +Hi Raghu thanks for the info +rev Let me know if this works for youThanks for the heads up +Here the query terms extracted from the bean for which we are searching for similar beans includes terms with embedded colons +We have two options here either set the bundle in to oran remove the configuration in +defer all X issues to +This wasnt meant to be in the build and I havent received any saying the build failed +I think the basic Idea behind creating AM is that any new applicationproject can implement that and use the YARN framework +Ill try that later +Can you produce a simple test case or better yet an integration test project +bq +Sounds good to me +Thanks Mick Ive applied this patch to the extent I think it should be applied but it seems there is still some work to my first submission I incorrectly included this hole patch in the description sectionCorrecting that now +Sorry I read the comments in incorrectly +Fixed by the fixes for . +Ive been looking at the security info related interfaces in the patch and prototyping out an alternate implementation that avoids passing around these empty interfaces for the sake of security checking +Attaching patch that seems to work for me but the issue raised in is breaking the build for me too so hard to say for sure +Task is finished +This probably could probably be marked as a duplicate for +Fixed on branch in feee and on master in fcd. +Ideally the major compact again gets the PUT that was deleted and the same is written into the compacted file + + +Please provide steps for reproducing the problem I think we are missing some detail there. +Instead of creating a history file and then testing whether it adheres to some schema it would be nice to catch the incompatible change early and let the developer know the effect of the changes made +Both variants of the substitution rule above break existing tests +Still not fixed forin ER and as the corresponding runtime issue is not yet fixed +Related bug +I can install webrat in just fine +Thoughts +I think Simons patch is a good improvement even if in practice this syncd getter should be a tiny costI agree +Currently I can just switch the jdk to to run all unit tests otherwise I cannot +Please check that it was applied as expected +Committed the patch above which also implements the requirement of +Please check that the patch was applied as you expected +Did code review and we no longer experience the problem. +Objections +Id think if we did something like this wed need to also provide some sort of wire format negotiation mechanism so that a client can tell the broker whether or not it can handle typed properties or not +I think just disabling WTP JSF validation for resolving EL variables is the best solution +These items are still not resolved for Please review your items if not ready for please move them out to future m still working on the last parts of this issue +Robbie can you please review +this is a performance issue so its very difficult to add a test for it +i will take a look on it so i look forward to create a good patch against the deadlockgreets RobertDou you have seen the warn message Unable to set the while a other one is associated with a different with this +Im not aware of anywhere in Spring Batch where we rely on ordering of primary keys +i think this might be a duplicate of thing I forgot to mention about the logs I have reduced the number of compactors to one when enabling the debugging +Galder First I used your project and was unable to instigate any NPE +Yes have respect for the existing codeThe state can be stopped shutdown etc +Dump or Snapshot +This module is converted to m. +Can we change the algorithm to check against component names service names and reference names as well as fail fast? We should probably keep a list of all registered named elements and check as the each of the elements are processed +svn commit m fix svnignore properties revision . +Weve made a few fixes in relation to that may help this issue +Already working on it +We can discuss that in the next Forge Weblogic testing issue will be tracked in +I think for now we can disable this check for the HA caes until we get a better solution for that +Thanks fixed in the branch and in HEAD. +I would prefer option to switch between modes +Jun made a good point when we discussed this offline +I know that jira is near end of life but Im not sure about +This worked fine in + +Including and +When I contributed the previous patch I didnt realise that the +Im not sure how you can give them meaningful names while keeping it pluggable and not tied to DSE but Im open to suggestions +Please supply profile activation based on whether or not this build is a development line or not +Again thats what were doing now so I dont see it as that big a deal +Quoting from my comment Can you provide a buggy zip filegenerated by cocoon? I will be glad to test it on Linux +It sounds like the same issue +The updates to Translator and should successfully address this issue. +The real solution is already implemented inSo what I mean to say is there is no duplication +This sounds good! +Hi ManoharI get the expected out put for this test against C server +Based on some research it should be sufficient to embed the JDOM jar and define the dependency to as optional +Have a look at revision if youre interested in the gory details. +Both files were located and the application ran correctly +Now mkp is always required +I will send you a new patch this evening +Causing major headache +Attached initial patch that supports rudimentary grouped field facets for single valued and non tokenized string fields +If this is not fixed for I can clone it to. +Oleg appreciate your help +It seems more convenient to follow Ant behavior and not to execute a target when it is declaredIf the user simply wants to execute set of Ant tasks he will not need a target +ve committed it to the master branch. +Navis is there any way we can unit test here +Actually I dont think the prefix apache should appear in the artifact names +So what can Jetspeed people say about this waiting to lock for infinity time +Defer and lower priority of work items for regionarray +This will still go to and +I tested a build from the branch and it passes my test +Test failure is unrelated +So are they stable in the sense of not raising problems because there are no problems or they are not used? Perhaps one criteria for stable would be that there are examples and a GUI for their usage +Attaching and with changes to the following filesM srcdevguideM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefI hope the text now clarifies the situation as far as it is known +Can anyone add me into the contributors list so that I can assign it myself +This patch adds an example application template for building with the standalone iocore and the publick interfaces +Verified by Aleksei +Sorry I see that you say the same Mike +bq +My sample project is between configurations +I tried to see the acl table in mysql DB +Bob can you look at this +Canceling patch to address Sureshs comment +However this test case is passing on my local box with change +Ill backport it to the trunk +Thanks +yepp +Had to roll this commit back because connector and Elastic Search connector both still had dependenciesr for the rollback to reapply wed need to roll back the roll back +at least deploys it with some dummy entries in but so far I didnt manage to get it deployed on Geronimo without having to comment out all the affected resource references in At least some mode which would just give warnings instead of refusing to deploy completely would be greatSomething like Strict Manifest Classpath processing mode which can be altered by the option. +The server seems to leave the system property set for the bootstrapUrl after shutdown +Perhaps there is a way for the int array work? Or is it best to create a separate file that is very similar to the term dictionary file but only stores terms and docs +verify the advisory propagation via waiting for a failure to delete +Also configuration of the and other parametersDidnt quite get this +envers version used +Flavio the index is also flushed in grab +I fixed it at ASF +It is only recommended that Comparable provides the same equality logic but not required +I am using git through IDEA which makes it hard to say which kind of patch is produced +The problem related to having multiple epsg factories around is still valid but is reported in another issue. +Reopening then closing to deal with migration bug. +even though an assembled karaf is not reproducible this way I guess generating a kar file on the fly might be a nice idea +thank you Nicolas. +AdrianOk yes fineo Actually I think it was my error I forgot to put some jar patch now includes service to get an external website HTML element +I dont feel it should be in included as it opens the door to basically any utility module and I dont think that is the point of DS +We should consider wrapping the in a if information about the vetoed state is needed to handle the access denied event +Committed revision +This patch updates with the latest list of bug fixes which will be included in +No worries +added new which take care of the transaction bindings too and also rework the whole process for special bindings +patch patch needs a small unit test showing that the problem exists and that the provided patch fixes the problemAlso this patch makes the code more confusing consider + +wow i seriously cannot consistantly replicate it +Patch so will download emma jars if needed +To be verified in newer build of JBT +If so switching from MERGE to MERGE may fix the issue. +Oh and probably it is best you do not use the patch but rather move the classes in eclipse yourself +Resolved in +I second the vote for table +The update site should address the dependency issues by downloading and installing the required dependencies +If I hear no objections I will disable per default in in SVN trunkOleg +How about we close this issue for trunk and create a follow on issue for? This will unblock this for and Greg can address this when he gets back +cacheThese changes improve dramatically OGNL performance under +I will look at this today +Dup of off release +Patch is complete and ready for regression testing +No I think my patch is somewhat better so Id like to keep this around until I can publish it and have Igor review it +Attaching these jars for others to easily accessEric whats the plan with incorporating these jars into the build system? Should they be committed +Three points With the new changes of I did not find a way to write a unit test that simulates a DNS error +attached is a patchafter applying it run ant genrbbi which compiles the rule tailorings to binary for faster loadingyou dont have to do this +The second alternative is great its short but still easy to understand +I took a look at the file now and It seemed to be ok +I will to install now that Ive been able to remove the Oracle Spring ORM tool. +There is very much a question however as to what methods this interface and imple should actually expose +I left out of the Environment section Oracle i using the g driver +Yes no in this case +Reviewed by Irina approved for. +Didnt make so moving to for after the release +We are using load testing need to through all the load testing after all of the changes have been completed. +Also note that doesnt get run as part of as far as I can see +Just in case of what? Functionally it is identical +Note NOT exposed for +I just committed this + is the main file in this archive +id of enclosure is now stable in is still problem with unique id ofThe introduced counter is field member of but instances are not shared between markup files +limiting for a specific resource +If you find further issues please create a new jira and link it to this one. +My apologies for not responding earlier + does not include the stack and is not meant for using it hence the wsconsumewsprovide scripts it comes with do not include that lib +Tom I have changed the codes to make the test case provide the correct keyboard layouts configuration automatically you can update the bot tests project +any update? I cannot reproduceIve tried to reproduce on linux without success +Ill mention this on derbydev to see if we can resolve it +Ive given up on test the test setup is seemed too expensive to do +Images were process and uploaded to the DMS under the Hibernate project +Done. +The latter two files are read into the esb engine and they have no xsd since they are pretty much free format. +Will commit when tests pass +Linked to issue where this should be addressed +Does the ? mean that this choice is up to the implementation? If yes I would strongly vote for only implementing autocreated for all those props +This patch fixes the problem and the testcases +Please check +Shall we close this one? The request for a language selection in the gui is already in Jira actually one of the oldest tickets in the system +Ill wait for a few minutes and then Ill commit to and +HiI use Window XP SP with cygwin environmentAfter run ant clean command it still remain some hadoop scripts in NUTCHHOMEbin directorynutchbinnutchbinnutchbinnutchbinnutchbinnutchbinnutchbinnutchbinnutchbinnutchbinnutchbinrccnutchbinnutchbinnutchbinnutchbinhadoopThanks +How can we order it in the compaction list at the time of insertion? That would require editing the in place to set its seqid no +I also have a version that replaces with +Crec que no +as discussed it is not a real please close it +See for example issue can only be closed not assigned not fixedAnd it doesnt use to be like this +This could also include a callback for when a Callable produces a result or the underlying response expires. +All necessary bundle refreshing job is done by and there is no need to perform any additional work on removing a bundle +When I do the commit it will work in concert with published maven artifacts in my own personal space +I think that the intention was to state instead of +Ill create a separate jira for this since its an easier fixFor this jira may I update the error message to suggest typecasting before the union? ERROR Received a bytearray from the UDF +from me +I was thinking about that it would be easy to implement as the current code would simply be moved toShould I add patch relative to whats currently committed +See notes in new pdf file +An exclusion also seems to prevent a direct dependency from being honored +Works great +reopening for clarification we should be creating a new for JBDS shouldnt we +For this problem we made a patch earlier for +I think we have to check also Job Scheduler. +Updating jar to incorporate resolution of +The reason for the class is so only the locale being loaded will be locked but I see little performance benefit +The annotation assumes that roles start with ROLE otherwise theres no way to differentiate between role names and expressions. +Id like to provide some additional support for fixing thisWe have a situation where we have some hardware crypto modules connected to aLantronix EDS device +Does it really make sense to add an option in a configuration panel in while we can do it in the ivysettings file +So Ive added the latest schemas to the current implementation +Ive done some work on mail clients and servers with Java in the past I would strongly recommend moving to JDK s Charset as it saves all kinds of headaches like thisIm not sure if the current API could be preserved so it may be a breaking change reserved for but attempts to handle charsets with tools is a truly Sisyphean taskReleasing a with current fixes is a great idea +Test remove sessions after iterating through the The failover code to remove transacted sessions is not safe since it removeselements from the when an iterator is running on it +Yes actually its in your patch inSo at the end its a good news the test caught the issue. +I look forward to checking out DW after this change +No need for a snapshot release I can build Castor myself +Hi ToddThe question I had was how is the info znode creation prevented when the client does not have the ephemeral lock znode? Is this ensured in the zk client or at the zookeeper +When there are more than one function it will use just once per record +FYI Around Axis the array code was changed to not allocate the size of the array as specified in the SOAP to protect against a DOS attach where a malicious person could POST a small amount of XML that claims to need a size array of something +Youd need to break your installation pretty severely to get the VM to throw that exception +We should add commons validator to that list too +Also could you include another file with the output of svn stat q Right now the patch has this hardcoded path in it so it will not apply +I will open a new issue for the +Hi DanThanks for the patch I get the point of this patch but generally we create a new bundle but not modify based on a old bundleI will commit the new bundle soon +This is duplicated in +If not specified the current notation is usedThe patch includes changes to the patch supplied on +working fine Hence closing this bug +similarly jcrours childnode definition uses the reserved and should be renamed to repours +Only when a new release is made are all included fixes marked as Closed by the release manager +Im going to provide complete testing for this patch and new inlined version for this helper +If you always run on the local host then add the following line to the HOSTS file activemqhostBecause I dont run on my local machine and I cant guarantee or force others to do the same then it is better to create a generically named host machine and allow each user to resolve that generic nameThis was the intent of the naming inside the configuration file +You sure are lucky +Also if you have turned assertions off please them +Guess this is just curl being dumb +Patch +ChristianHave you been able to apply the patch meant to push JAXB handling completely over to the +This cant be easily folded into if is in File +Removed worng jar and simplified poms +but id recommend using Axis for the service as well +If it cannot be fixed within wise then we should add something yes +Cool on that strategy +Of course if you need them the change is welcome +Im resolving this issue as the problem seems to be with the tools used to untar the +Ive just committed some changes to the sonar ant task +any update +Weve upgraded JIRA and Ive tested it in Safari and it seems ok now. +Inside of a trigger action USER CURRENTUSER and SESSIONUSER return the name of the user that performed the operation that caused the trigger to fire +It looks like is a minimum resolution for our feedback logo image is created as the less image width is needed. +out files? Any scriptV bugfixesThis patch successfully passes tests including long SQL and SWING testsIt seems like I was too optimistic about cachine is property in +I did a local test with production data and it looks fine +Skips Validating when rf I havent gotten a chance to test this on a cluster but Ill give it a shot later this weekend unless jbellis beats me to it +Ive manually compiled Tika with that BP +The other failing tests are not related to this patch +AaronIm pretty convinced now that it is incompletecorrupt wal logs get stuck in the retry state machine +no more changes for would require a change to the +So my question is why are collection elements being proxied? If I read the Hibernate docs correctly this is only supposed to happen with fetching +I think you forgot to attach the patch +Maan this comment list is infiniteHow do I currently get the +Committed revision . +Is there any update on this +Feel free to reopen the issue if thats not the case. +I think current behavior of Engine is reasonableWhen setNull method was called type information of the null value is specifiedThen Engine can determine whether the type of null value is acceptableOn the other hand when set method was called type information of the null value is not specifiedThen Engine does not have any other way other than takeing the null value as incompatible type error +I think you said something about junit making this base class obsoleteWhich way do you prefer +The here is to expose the netty apis into the applications +Ive committed to trunk branch +If we can move it right after patch download and before the javac warnings build in setup then we can see even more speedup +Thanks for the additional information +Yahor yes please attach that failing workspaceI have imported project to my workspace and created its copies +Gerald Thanks for adding your workaround +Merged the into the master branch. +committed as r +Youd need to provide some test cases that demonstrate the issue I dont see anything here that I can go on +Well it keeps happening +Thats probably a decent workaround but Im continuing to research this implemented this part of our library similar to the real +Changes to trunk to remove all the sun crud and use the proposed spec schema jar +For the record this was Major +I guess we have to revert this +This issue will need coordination with other commiters as ther may be merging issues as its likely that there are extensive changes which may also impact the sample application example +Is this an accurate summary of the evidence so far? Thanks +Before sending split message should check can or if fail park the message somewhere where it can be picked up when master comes back +Of course purchase is generated so the data service needs to find out what this generated value was and return it +It works now maximum workers is and more than a java heap exception occurs +cleaning made on revision . +Do check in before allocating volumes for the new block +All that is missing is a bit of exercise while under heavy read load +listKeys is implementedNext step is to adapt the key management UI +files need to be patched to eliminate errorsattached patch file by manually deleting the other stuff from the nd patch file I attached +Looks good to me +Can you provide a test class the demonstrates the issue please +Does that sound okayto you +Barring no issues I think we can get this into +Let it be container first +Attached patch was committed in r +At least for one release we could keep the old scripts and then remove those in Overall we should keep one script for this as proposed +to chiahungs comment +The stylesheet is completely out of sync in and +Ive just committed this +Updated patch. +Checked in commented itest in +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +This issue is out of date and will not be fixed +push to +Sanne Ive created for the subset of work done already +on this as a solution +Feels very unwieldlyShould this concept of is in range be applicable to the concept of a Durationor to JODAs concept of an Interval Duration or Period? How far do we then wantto go into this +Fixed. +This used to be null inMaybe the disconnect is that hostname is used to define what is published in ZK which is always required whereas youd want to define the bind interface as well +fixed of a bulk update all resolutions changed to done please review history to original resolution type +What files should I attach +Now byte for the class type and the size of the type itself this means we can have only byte keys and byte values in the map + XSLTC defines a table of possible mappings between XSLT types and Java types for the arguments and conversely a table between Java types and XSLT types for the results +Merged +But then how to get that into the area where the schema validation is turned on +This appears to be a significant regression +I will watch these issues to get more information feedback on the above proposalSanjaySuresh Id particularly like your feedback since you guys have a lot of viewfs expertise and some of this is similarrelated +The previous one was generated with svn diff +Also checked the riptano rpm +Where is the nullpointer +Fixed. +The redirect issue unfortunately may not have an easy fix I am afraid +This needs the released version of Apache Shiro +Thx +By virtue of the amount of good work you do for the project and the complexity of areas like this the occassional mistake will get through as do misakes from all involved in submitting changesThanksThanks Jacopo for your good work +FrancoisI would like to resolve this bug as closed as we root caused it to an issue with Feel free to it if the issues is reproducible on blockers and critical issues are resolved but not verified +Tests passed cleanly for me +Lets do it +This is how the system takes back orders and when inventory is received balance will set it off against a positive item +Ahh right the reset is pointless good catch i was just doing the bare minimum to avoid the race condition +Thanks! Ill fix this issue as soon as possible +m I dont believe they belong there since they are project scoped variables and are not applicable to my plugins project for example +NET server andor ASMX file if possible +Looks fine +So my comment is not relevent anymore for the trunk code +On the other hand there is still apparently a possibility of getting an exception while getting peer certs so now I am not sure whether to work on a solution to this problem or ignore it +How about updating this jira with a release note then? Well file a Pig to update the tests to work against +Please confirmHiranya +Do you have any suggestions if either BLOB or TEXT would be more appropriate +Thanks for catching these +Closing all tickets +Thanks Nick +Can we generate these docs by massaging the output of help or usage +Havent look at those source files before but the fix and the patches lgtm +patch applied +available as an bundle revision I add a required xfire bundle. +We also have this issue when we analyze scanned news paper articles + +Jira notifications changed to go to devAll done I think let me know if any problemsquestions +Added unit test to trunk where it works nicely +Sorry for the delay +Punting issues from RC to final +I fixed that by converting them to DOS paths +Sorry for uploading with the same nameHope this patch addresses your concern +This would be the ideal combination of great language libraries speed documentation community and TOOLS +Patch in file form +the FINFO and URI constants are special as there is no equivalent in Apache and they are only relevant to modpythonThus contains should therefore be +A fix for resolved types in the last version revealed this problem now in +On a second thought does not need this since does not have append +can you give the context of this error? What query is additional information provides +Heres the script I usedproject top setup provider DATANUCLEUS database GOOGLEAPPENGINE applicationId myappentity class +Attached patch ensures that Pig does not error out when the error log file is not writable +Reusing a corrupt repo would cause subsequent builds to fail +I am pushing this in +that will kill optimization +I forgot to add a break +I was not configuring the object with the proper settings from the project +There are no platform restrictions in Feature in case of target platform is not supported only xpcom is going to be installed +I have migrated to the new style the Oagis application +It will contain all of methodsand attributes +Maybe you just havent done a clean build +Anything available to get around this little issueThanks assembly plugins should allow the use of stub when these sorts of are encountered +Indeed all seems to work here +Instead of storing the index we need some trick to remember the selected datanodes such as setting the bit when the datanode is selected +Will upload a new patch after fixing that +to the new API and backport of the logic +Suresh Nicholas if you have a specific suggestion for something that needs to be addressed in this patch let me know +We can edit out the long waits caused by +wonder what to do with this issue since based on JTA API it cant be solved and you should be able to get the exception through logswhat do you think +The best way to dothis to have autoconf do it in or get autoconf to define a ISMACOSXmacro in the same place +And what is the problem with that url? it should work fine in every browser +Or if you can give some pointers on how to implement it that would fantastic +See for the fixing in the script. +Whoops I meant will test this +We should verify that this is fixed in +the jobs superstep count should be not +Really sorry only saw this message now +We dont need incubating anymore now that weve graduated +Assigning to you Geir based on Bretts comment +Hi BrunoHave you made any progress on this? If not Id be interested in giving it a shot +Eli If you dont have any comments we will commit this patch today +Thanks! + +Generally most teams will have a single custom implementation of a custom but there might be cases where they need to pull sources from multiple teams and need a custom reader +IMO the current Trinidad code base has some similar code to finally support this +Additional benefit of the real unit tests is that it is easier to debug and errors dont go into some black hole like it happens with +Looks like the best solution Ive seen +Deciding in favor of XML I thought it being advantage to be able to look at the result + +This will needs a protocol version change since this wont work with prev clientsdatanodesDoes this also affect fsimage version +I ll look into it thanks for the info +first iterate on trunk and when we have it at a reasonable stage we backport it to +If I am mistaken maybe there can be a change to the documentation explaining that a detach wipes previous updates done to a pojo +We may later find it a problem when some downstream context tries to do something we can only imagine now +As suggested moved to and and assigned to Noel our expert +Therefore we should be very careful before introducing any kind ofcentral or decentral configuration +Juozas any idea of why this issue is starting to pop up know? Why didnt users see this previously +Im fine with downgrading the priority +This happens to me as well and is a major performance issuemy account on softlayer had about virtual guests with all kinds of services installed on them +Ive taken a look at the example portlet you provided and Im quite confused as to what technologies it should be provided it doesnt build so I assumed you wanted of and that the Seam dependency was not needed as there isnt any Seam code terms of Beans do you want to use JSF Managed Beans or CDI Beans as what you have at the moment mixes the two technologies which opens the door to problems +Running regression tests +we should do all we can on each compaction that would be less work on the next compaction +Fixed in Release Branch and SVN trunkIf you want to backport to please refer to rev +Ive just committed this. +Ill try to create some patches this week and attach them to this issue +completedMaps includes succeeded and failedInstead of incorporating completedMaps in the calculation it can take succeededMapssucceededMaps completedMaps failedMaps +If my task got scheduled to the same machine with these CPU intensive tasks my tasks might run or more hours +First pass at investigating this would probably be easiest for Nick since he did the original connection pool work +But when I run it through ij the way Myrna mentioned it still has the same problem +Addressed by adding the virtual group update type to the generated proc command +Attached is the patch +seems like a valid capabilities document to me +Another note Ive been using the trunk version to build the jars for both creation of the objects and the application itself having initially seen the sequence issue noted in JIRA case +The fix for this is nontrivial so Im marking it a blocker for RC and will look at it today +uploading the patch with the findbugs warning fixed +bulk close of resolved issues for release. +We need unit tests and a base implementation as is abstract +As for facets they are equivalent to sorting in the aspect that resolving the actual Strings can be delayed until the very endAhh OKbq +The Interpretive processor should always gracefully handle error situations improvement should be reserved for new features. +for the patch +I just installed the latest JBDS CR installer and followed these steps Created a new Enterprise Middleware Enterprise Application Platform runtime +However im a bit reluctant to change the scoring code until I understand its history because its a key functional piece +Updated patch accordingly +The abstraction helps to keep the code clean +Remember Spilling Buffer is a circular array of configurable size that contains configurable sized buffers +Closing. +varaccessorfunctions is renamed accessors +The resolution of also fixed this issue +On what test did you check your improvementDid you test it on +Committed compile fix to trunk. +postJSON acts just as you describe +The text and the Decoder are stateless +There is a similar problem with ejbthat name of the generated in the target directory is changed when the is I assume that the problem will be the same withejb +I think this is the reason I try not to create an atomicLong and insert into hashmap +I am out of office until Aug th +Raghu do you think that the relatively slow reading speed per block poses a problem to the datanodesSlow reads are same problem as before + +s +It would be great to be usable in a production environment if this flag could be switch on on per job basis +Probably it might still provide gain when using Mongo results indicate the same improvement +Should be all set +Should I put it in Michael +The workaround is to load ivy downloaded libraries if they exist +does that mean it used to work in +Only the new snapshots after upgrade were not getting created +I manually tested these cases default not set set to a valid port set to invalid values such as or fooI also changed logging levels and checked for proper output in the logs +On my machine bin is executable both after fresh checkout and after running ant jar +This is a show stopper for me +I have added a test as part of the fix +Thanks Adrian Ill fixed itErwanI will commit the changes to the +the docssr tree is issues for released versions +Added code to use the queries cache also when the session filters are enabled +Also does there need to be a regression test for this problem? Or will a regression test appear when abcd is rewritten +This is a corrupted set of settings which may be fixed as followsRemove file Remove from +Hi Stphane about your use cases The first one is not covered Users involved in the violations But the second one will be receive a mail with all the violations introduced in the last days if they are still there at the time of the mail is sentAnd to cover the second one were going to always work on period whatever is the value of this property lastanalysis day days +I was thinking about adding an osgi test +Changes from patch was case of creating after creating in the same thread was added in. +Right +I think having the system providers be a lower priority is more natural +Resolving this as cant reproduce +Changed scriptfile parameter name to scriptfile +This is a bug in PMD rule pmd which is now marked as deprecated in favor of squid +You can even npm install it +Im sorry! This was just an early fix attempt +Fixed in a local branch to wait for to be released to push changes into SVN + for the patch +Dave Parfitt is correct + needs to be the URL we actually use it can be backed by something else but we need the updatesite to stay available under since that is what is used everywhere both in users existing installs but also documentation and as a way to separate community bits from product bits +Committed with a warning in NEWS to drainGo ahead and open a separate ticket to fix sort order +ChristianAs mentioned in the original description be sure to implement both well as forgot to show you the latter in the code review but if you do a coverage test while running both and youll see that coverage improves for +Ive sent a pull request to merge these changes upstreamAs for the WARN messages you mention under heavy load if this turns out to be an issue in the AS code feel free to either open a forum thread or a JIRA with the details. +So its ok +This is by design but in any case this behaviour will change with the new HA workI recommend this issue is closed +Lowering priority on this issue as there are available +RC patch does not introduce any new failures on tests +Need fix up the mediator after completing Axis configuration changes +Still if you feel strongly about it start a discussion and call a vote +Still I could see the issue even with very clean build +see also attached havent committed I think weve got two threads crossed in this ticket +has the new replicator JSON in C snappy compression +Well I already attached the patch to so I suggest we close it again +Escalating this is very serious +As this is definitly out of scope for myfaces implementation I think we should find a way for jee containers to define their own implementation for this +No response from reporter +I just change to true and wait for is there valgrind on the linux node +Fixed +A good ratio should be arrived at experimentally +My apologies for the noise. +It seems like it is useful that updateInfo takes in a Set instead of a Seq +Thanks to Curtiss Howard for provide this patch +Selected state is now stored in the component like the expandcollapse state. +Of course not! Just encourage them to upgrade or zoom out! Lol. +Will fix +Initial version checked in the branch +Thanks for the feedback Konst +I dont think theres any dependency on XercesXalan +bq +working fine! thnx! +So I didnt discover anything new here +I tested with trunk and it was fine +I wish I could fix this but the truth is that the parser doesnt like line by line compilation similar to how python works. +Verified on Hudson jobs for trunk and stable branch +I should add that I didnt change the script at all that had worked with +no longer handle the encoded length +Actually our is at apparently +I would agree with NiclasMeiers point that this can be solved outside Struts but since we are settingthe contentType anyways it seems logical we would allow the request encoding tobe set as well. +already updated to final yet the problem is still there +Fixed and available in trunk for release +Fixed and commited +Im not sure we should remove the license linkThe hover text is clear in that it says is GPL licensed so if a change is needed change the link to that as well +Reopening because the solution must be different +patch for includes the linked jiras in hdfs and updated to merge with latest trunk +When adding MPUT consider recursion. +Just committed this to trunk and +Not all repositories are completed. +One more update +Or is this normal that two parts of a deployment are seeing different class pathes +If these exist then I apply them before the tokenizer gets called +yes I need a sample content and commands used it seems it is easy for you to create it thats greatbut for the moment I personnally dont know how to write a project that will have a property when archetype created from and I would loose time trying to write one if you create a sample project it will help me to focus on fixing the +the easiest solution for this would be to add setset bundle methods to +Ive tried to find the balance between semantic and simple. +Commit as you see please +Assigning to you instead for +Saschathere are some errors when displaying fields you can see it in the attached screenshot or go this page httpslocalhostapcontrol Erwan thanks for the commit that looks strange i iwll have a look tomorrow morning +jdf plugin released at CPI +But even with that patch if I lsof the java process as it runs I see it hit or so localhost TCP connections in ESTABLISHED state while running this test case +IT yet to be added in +No testcase in yet since I have no idea why the db test worked before +history is used for figuring out when bugs occurred why things didnt work this way before finding bugs and being able to create patches easily and simply to keep the source where it is instead of having to always go back and can also build out of git if we choose to but having that as a requirement would mean we would have to do even more work to set up this migration since we are changing the layout of the projects +test classes that use old stuff for testing in fact I am working on the tests first +Good find soft referencing them and letting them disappear does appear to be the wrong way to handle this +with mixed puts and this gets trickyhow to do wo a different read path piece of code +HTML Clean up +Patch to add configurability to Solr +Theres exception raising code that needs to be outlined so we can inline the common code paths of key methods +we recently updated our webapplications to and had to rollback the latest release cause the server crashes every time it got under some load +In the test case at hand Im trying very hard not to create any new tokens or so I thought +Otherwise the user needs to set to enable the use of +Thanks for bringing this up +I will commit on trunk tomorrow if there is no objection with s for the timeouts is not safe especially for you need to close your resources and tell to the masterzk that youre leaving + as previous patch but with cdh as default REPO in +HiIve created test programs that reproduce the reference count error on the bufferAttached in the zip called Theyre based on the exampleEach program logs any errors into its own text fileAfter about a minute I see the errorSystem detailsJREIve produced the error on JRE version andMINA version OSOS NameMicrosoft Windows XP ProfessionalVersion Service Pack Build Processorx Family Model Stepping MhzProcessorx Family Model Stepping MhzRegardsBen +Slow consumer test with multiple connections +reopening only to change version number +Will think on it over the we +Its crazy world that its come to this point +Olivier if my understanding is correct the issue youre encountering relates to this know limitation +Thanks for the fix Dag +Fixed the bug mentioned by Vinod +I talked to Marek Novotny and the situation is not ideal +input to unit test probably not +fixed + +when the bookie with SSL support it registered itself as bookiehostportsslPort so an old client could connect to new bookie while an new client could still connect to old bookie with port +Thus it looks like the existing code for trackingusing previous jobs results has a bug +This is a reasonable changeNit the change to is not needed since HADOOPPREFIX is only assigned if it empty or unset +Sure +the latest patch look goods + +What do you think about renaming it to connect or something similar? Do you have any idea +Alright committed then. +sebastianyes it should honor scoping rulesbehnamyou should work around this by changing URL normalizer order depening on your scopeHowever we may also change the basic normalizer to disable reference removal via configuration +Fixed in closing all issues that have been in resolved state for months or more without any feedback or update +done +The fact that the shift is entirely points to it being a problem with the scale factor +Can beans be registered after the context initializes? Can those be tracked by PPC as well? +I think this reflects a bug with the intel at r by declaring private memberPlease verify it resolved your problem +These scripts were written to help with setup for rpms being generated +fixed +I can create simple repro if it pleaseBTW I think you must use something like mvn install tomcatrun +Can this be resolved now +This issue is an umbrella for ongoing work now move out from any specific release +This is because they are copied from the MR app master tests that also need something better so I am going to file another JIRA to update the tests +We should address this at the same time. +Added patch with style changes and adjusted licence header +I would suggest that you try getting rid of the init servlet and putting any config info into the Logj config file instead +HiTo deploy to the apache snapshot directory you need to have server in your file with the correct permissions and make sure the directories created in the repository have the correct permissions after as deploying from windows as there is a bug in the pscp +now and +As per Martins last comment. +is this really a noob task? im on revisiting the structure here but shouldnt it be part of the larger retry facilities etc +This works fine when a Set is used as the child collection type +We have seen too many instances of issues with ST on different platformsL +This is caused by the backticks that escape the password field in the generated user class +Thanks for the patch it greatly improves theThere is minor problem with the attached diff seems that it removes changes rather then adds + +SOAP Action in SOAP header has eliminated +This issue can be closed down +this patch for branch only +I made a slight modification I moved the declaration and initialisation of positionBack inside the AIXZOS block as it is not used outside and I also added the declaration of the result variable as it was missing +Ive changed the repository url +Attaching patch that fixes luke and adds support for binary fields +However I donot have sufficient access to edit the page +Your solution above seems the ideal one as it only affects this single app launch +Patch for MTQ to not use null in get for backwards compat so null can have some other meaning +My schedule has slipped a little though so Im encouraged to hear someone else show interest in the parallel strategyI wouldnt be put off by the performanceThere are some very valuable use cases that make the parallel strategy valuable +Patch has been regenerated thus the verification +Ah sorry I just noticed that the sender is the commiter not a general user for the script +Simplified log trackingStill based on patch for +It is possible to sort by score +As for th December following results were found during sign and encryption +Please find this improvement from the attachmentThis patch also contain Indirect endpoint support +Fixing the test and the find bugs warning +To alleviate any problems related to classpath length and Windows OS we could implement and use the same mechanics WRT classpath as the surefire plugin does +Thanks Tom + +Thank you +Move issues to next release +Eclipse gives me a warning without it +The option does not appear to work in Jetty +Hi Inder thanks for reporting this issue +However orthogonality of the returned eigen vectors should always be ensured because I cant think of any situation in which orthogonality is not required +Fixed the issuethx for the patch +bq +To ensure that we need a prototype +Jira causes mojibake +This is my testcase demonstrating the problem +Otherwise we will end up having two copies of the same actionThoughts +Look custom components in VPE +Lets do that instead +sorry forgot +tested and makes the compilation thanks. +shows doc changes +I am using the felix +Lowering priority as this is not a required to be in the next release but would be nice if it could get in +Username charpr +Cool sounds good checking out your branch +seems to have broken other testcases in MR not just the compilation of the +After further consideration I concluded that it was not desirable to associate an in the connector framework since there is no way to tell when it may be committed or when the thread association is over +Im using bytecode instrumentation to get the lazy loading working as mentioned on the hibernate documentation. +Let us know if you are still having issues with transformer. +tarscp cattrackbackup + then other use cases might want to make tradeoffs that are bad for luceneI cant claim to see the future but I think this datastructure is already so specialized and solid that the chances of it evolving much further are slim +Please reopen if you have any further issues. +This patch includes access methods forTo CollectionCc CollectionBcc Collection CollectionI have also refactored a few of the tests to leverage this new functionality and not rely of methods from the test mock we had in there previously +The exception is correct +Committed to TRUNK +For importing from jdbc use merged +Test updated +Periodically we could merge these updates to the master image fileAn alternative approach would be to update the image file periodically batching updates +Thanks for the regression report Rod +Discovered in the process of writing it that streaming didnt allow you to specify an output reader independently of an input writer +We still need to test it works for remote deployers +Sorry for the inconvenience +last patch they changed the API from to just when it exited CTP +For example for a bunch of create operations if we do it one by one once we get a node event we will watch it again so we will get another one for the next create operation +This is an apparently better fix that will allow at the very least do dynamic modification of Oozies poms +Simple patch to the locality counter +Updated patch with correct indentation +Please confirm it is okayAssuming you are planning to carry on with this process if not let me know and Ill add it to my todo list my preference would be to do one of the difficult modules next one with four jars +Those valueOf methods are only available in Java +Id rather open a new issueIm still not sure wether it should be in or nowhere +The bug and arent related +More update to fix code style +lgtm thanks! +Interesting +Now committed. +Hi Freeman thanks for submitting these changes in a single patch +Fixed to branch fix this specific problem after the console testsuite is working correctly. +Please update your contacts DB with my new email address +Patch containing the bug fix +the test plugin only compiles classes which are not excluded in the sourceModifications excludes + fixed in and parameters taken into account in thanks for the patchr +add code the original reporter of the issue didnt say otherwise I close the issue as Wont Fix +Excuse my ignorance but how do I find thos out in Windows cmdIn Linux I just type locale and its all displayedI typed chcp on cmd and got Active code page if thats of any helpApologies for my lack of knowledgeThanks for your continued FreddyThat worked a treatThank you for your helpPlease feel free to close off this issue as it is not a bugIt may be an idea to improve the error message for this if you think its necessary! news Grant! youre right were going to improve the log message. +As pointed out on is a very thin layer around Jetty rather than making a change to Avro well access Jetty directlyThanks +Could you please either check if the problem persists with the latest version from subversion or submit a small test case that makes this easy to reproduce +Please provide asimple reproducible test case +Fix in CVS +Cosmin +or use the svn rename I corrected this already for youRegards HansAnother patch for some more online help screen documentation +the was there and thats the convention i followed when i added the quote +Committed to branch and trunk. + +This bug is addressed in +I reimplemented the +the patch looks good +Ivan thanks for reviewing and posting your feedbackI think your suggestion is reasonable but I need to understand it better +This is expected behaviourYour esbs define two instances of the same service +Will this issue be fixed in a future releaseIs there no other way than overloading the +The decision logic in about which gets selected for the encoding leads to unless one of the following is overwritten somewhere in the class hierarchy of the concrete Component class with return false the Page that contains the Component happens to be known as stateful the Page that contains the Component is not bookmarkable +After fixing we also need a slightly modified version of this fix here it is +My patch is ready to review commit and scale testing +Verified by Nathan. +will revisit submission of a brew task when a m or m release is complete +This patch still applies to and fixes the problem +Thanks to Douglas Campos +Test case added. +Tom if this has been committed to trunk please close the issue otherwise it wont turn up in the release notes +Im sort of tempted to think that we should namespace all stuff into a THRIFT namespace or something for example +Attaching screen shot +Both patches should be Both patches appliedPasses on Ubuntu testcase smoke kernel +Ive fixed that now in r and it all builds ok now +However I agree that leaving messages unmarked in store after an attempt to process them fails doesnt make any sense +The easy way is to just create a qualifier in the serialization protocol +The latest change was a simple one +Ok thanks +Eclipse me project with Enable Workspace Resolution fails to find any of the projects that I had listed as dependencies in the +Extract and in myplugin folder read some issuesclean upAdded a patch for Subversion +for having a sortablesearchable conf page +I think Douglas Campos is going to try to take Brian Fords specs in this area and apply them to logic +Let us try and fix it now to have time to study side effects +Then you change your configs and try again based on the errors +looks like sami fixed this in after release. + +Bulk move issues to andOdd that the pjoin cache is making things slower +I can work on the REST calls no problemIll be happy to help on the topic too I think the first step should be letting the run inside an container and then refactoring things to let binding of services from other bundles. +Same as patch but patch has more comments per a review by Yash Talwar +Might be related to the collections no longer being cloned behavior may work for inline evals but what about for standalone evals? I have some rules with eval as a standalone conditional element that worked in and but failed to compile with this same error in CR +doesnt use any of them +However the infomation which BN has decides the latency ofThus should we also decide which information BN should have for different HA requirement? Such as Persistent data Namespace information which has been logged by Backup Node with an administrative issue BN could swith its Role to NN file status information lease information which could help the service continueous during NN switch +Awaiting release +I am rerunning test +IMHO to make it pass on interpreter we need changes in GC and classloader to align objects on bit boundary otherwise we will have problems with objects crossing borders of cache lines +on commit +Please have aclose look again i just removed the variable assignement but left the readstatement +So no yetI think the final decision should be made with an official vote on +In the opensso file we dont need to ask users to download the as we already contain the in the sample +So if there is garbage in the test directory before the test begins its very important to remove it +Reorganizes code to srcmain +AbhijitThanks for attaching the war but Im afraid Ill need a much more minimal reproduction of this problem +What is the rule for trailing zeroes in the release number? Currently we have not or +Im not sure I fully grasp the dequeue idea but I can see that there might be some simple more efficient ways to do it without having to scan all ledger metadata for example by keeping a table of deleted ledgers +Ported to +Im not going to investigate that + +I was wondering if the requirement for the file to be there was because some piece of code was statically initializing a new configuration without getting passed the properties +bq +So should we deploy ESB to the same MC on which JBM is running or have our bootstrapper start the JBM MC in parallel +Perhaps we should also put the Overview under svn? If we did that then other people could improve the following parts of the pamphlet By editting a JIRAs one line summary you affect the JIRAs summary as it appears in the pamphlet +I committed this +Sanjay I used container too loosely +Thanks Ill test today +If you find further issues please create a new jira and link it to this one. +Mickael no I had to remove the source to work around the fact that Tycho cant handle two copies of the same IU in a feature +interface needs to be updated how did you change the without that +I experienced the problem in IE +The patch looks good to me but I have not followed this issue closely so confirmation from an expert in this area would be appreciatedFixed conflicts with latest svn version +There can only be one underlying msg provider in jbossas so there would not be two different destination managed objects + already provides for the unquoted JSON element names +Many thanks for your help +Not sure if thats correct but lets leave that for now since the same thing could happen for Linux and the Java code is written with that expectation too +Though thinking about it now it might make sense to change it as well for consistency +Passed jenkins +Any tasks including setup and cleanup tasks would need to sleep at least seconds to finishThe patch make the wait interruptable when tasks finishWith this patch and and outofband heartbeats is about faster than +Fix looks good to me! Youre becoming quite the expert Pushed in dcaa. +I am attaching the changes i made to serde as a separate patch +Fixed overall bytes failed should be in ER +Patch to add definitions for and to the Query Hints section +Thats a good point to trunk +Sorry it was a problem of the TCP Monitor +committed in r. +Looking at the lib directory in the WISE core distribution were talking jars and about megs so cutting that down would be my first said though it looks like the WSDL functionality in WISE is cool Im not sure how Id use it to even generate a sample request message +can apply this patch only after gets committed +Marking this issue as resolved for the release +Whats the URL of the SVN repo for it +It must have been fixed at some point +Lets set the default file size to a GB by default +Any interested parties please watch the branch and the subtasks of this JIRA +As described in I do get an exception from addColumn but no exception from if the disabling for some reason didnt happen +Ruby assumes always +sets the setting the src address in UNICAST I want to avoid that because src is always set either in TP or +Reading on few pages in the tutorial STS provides via the Run As Test a more visual interface which I prefer but some may still prefer the Roo Shell +Starting to think +Thanks flavio! +Did you verify if it is broken or just based on the eclipse bugs still being openWTP definitly should work with this +Thanks Ludovico +Its unlikey this patch caused the TTMR failure looks to be problem going to hdfs but just in case +what alan described I feel that we should commit the patch since it will improve common case +Patch integrated to trunk. +Microsoft is notorious for ignoring web standards in its and has done this in multiple ways in +FWIW the handling of relative was far from complete in in some cases outright erroneous +In a very strict sense it is possible to extract odetrunkjacob but you would still have jacob under the ode repo as well +Log when I only define the when I only define the privatekey +Attaching full server log +One reason to care about routing orderThe batch methods allow inserting to multiple at the same time under the same key +Now that work is well underway with AS all previous community releases are +So you post the duplicate form and get back an HTML page which you probably do not wantFor duplicate submissions I think its best if the client side handles it ie show a loading indicator which blocks the form and duplicate submissionsI dont object to the change in code though just not sure how folk will handle the on using AjaxKind regardsBob +The proposed fix is to create the split with all nodes that are replicas for any of the blocks included in the split +Committed revision thanx david! +The section reads very smoothly +This is going to be a problem however coordinating the stuff across multiple users +Attaching the patch file +Uma that approach seems fine +But master still does not read in the full information of the new table so it is not REALLY enabled +v should be ok +If there are objections to this way please let me know +Verified in CR +Use the correct fix version. +Thanks +Some memory leaks have been fixed in the upcomming release +This invariably happens with Subversion repositories converted from CVS but also when various svnadmin hacks have been applied with older versions of SubversionGiven that you are not looking to preserve and entire correct history Id go now with git svngit svn has always managed to correctly enter all branches and tags for me as long as you have a canonically structured Subversion repository or you give all the correct Git command line arguments to offer up the Subversion branch and tag directories as well as the trunk directory + +all isues related to new p enabled product build are moved to after local build is by replacing in inside package for in +make some user errors by not providing the correct options to +Would be good if someone could review my changesWe still need to convert +Proposed patch to link can see your point with regard to the client library changes +sounds reasonable +We should file the bugs against EUT and look how Eclipse answers our alerts +patch committed thanks Santhosh for this significant improvement! +But dont know if it make some cracksIf it will make cracks Brian please ignore it and close the issue +In the services unit tests we only count the code exercised in the src of that service while it also gets into code in rosetta +Making it a blocker as it blocks test development +However the split could be triggered by a single store reaching the split threshold +Thanks! +One experiment I tried was to remove all the secondary indexes +Have you tried with upgrading the version in plugindependencies of your pom +Thanks in advance +For right now were just aiming at cleanup so we can do an initial baseline release so Im worried more about making sure its not going to Cloudera when it shouldnt etc +More tests could be added on the management sideI also want to add test cases to cover and once I fix the issues +Exceptions added to prevent inaccurate use +OK Carlos Ill use as groupId for release we close this issue +DUP of and still NEW +why are we targeting instead of in +Summary separation works ok in my Chrome version see attachemnt +Alexey Ive fixed the patch +There is no way to save memory be eliminating isset when you know you wont need itAre you suggesting that there shouldcould be an option for turning off isset if you dont require it +Thank you Plamen. +Yikes + +Postponed to not yet available in AS +This bug has been marked as a duplicate of +Thats OK +Thats the intended behavior +This means it is NOT using a url intended for the OTHER project and it is NOT confusing the two projects together +clearing the CL wont help because the HH cf already exists so its not recreated +fix committed in r +Instead of storing a new field data we could just store current +If you find further issues please create a new jira and link it to this one. +Recommend you provide a test case to show the issue and protect the fix into the future +Actually even my tests are no longer showing this issue consistantly +Carl I am having problems making the build tests run using ant +Moved to please move bugs back to which you will work on in the next weeks +This was caused by configs in the reduce +Note the patch attached isnt quite complete either +Having a hint override this value is now to implement but it would take effect for the entire user query rather than a specific union or join +As a matter of fact we just fixed a huge bug in the server yesterday which was producing the exact same error than he one you postedBottom line we had some conflict when some search was conducted when some modify operation was done at the very same time due to a wrong usage of some locksWe are cutting a new release todayThanks for the report +Delete is not yet supported +Ive added this test to JBC trunk +And will backport this ticket as well +Now it works fine with +Hello Aswath Please try to reproduce it I hope that it should be fixed by now but try with instead of the reason for that is since is not associated with Company +I ask Dima to help to reproduce the issue he cant reproduce it tooWe use for testingCould you provide screen cast video +Actually you can workaround the exclusion issue by repeating it in the dependency section of the for the artifact in the jarResources which is the direct dependency triggering the transitive dependencyBut is would be nicer to have this fixed so that we dont have to double check that were getting all the correct jar versions and not getting ones weve excluded +Sigh +Is this fix suitable for backport to +Any in parent coordinates makes repository content fluid which is a bad thing IMHO +No response assuming ok. +Editorial pass over all release notes prior to publication of. +Sorry I did misread the questionFirst thing to try See if an explicit identity stylesheet works better than thedefault identity transform +Ersin what is the status for this issue +I havent really hacked on the hadoop source code and have almost no java fu so it would be a pretty steep learning curve +The index of a seed site validates with this patch but I did not test further +HiI just checked the code it looks like Quickfix doesnt supports to let you configure two logger at the same timeWhich means even you switch the code let the work you still cant get the work for you at the same timeHow about you just update configuration instead of changing the code +what should i say +Ive created a few very documents using facelets tempating and did not encounter any problems + +requires changes to how the xwikiportlet deals with the request object +Also following letters will result this and possibly everything else local think this is a bug because packet names should work also with all these letters +Fixed by catching exceptions from point where there is problem +yes I read it and it doesnt change the fact that it is minor and will get there faster with a user contributed patch +Similarly there are other AST transforms that might behave in different ways when finding a constructor +I need to make sure to check these issues with PNG as well as +I think in this case it is OK that we let continue to run +Committed to both append and security branches +Looks like you have put in a great effort in thinking of these menus and also drawing it up +Here is a simple patch that lets you use a rate limited directory wrapper you can set the rate for each of the io contexts that lucene provides +It has now been fixed it in the version +Again I think that internally mapping the html DTDs to their XML counterparts is a viable solution now that Daniel submitted the local DTD patchIve got things working locally to pass through doctypes from templates to the client following a first doctype encountered wins principleIll submit a patch later +I just filed a bug and attached a fix for the bootstrapping code inOnce that is submitted this can be retried on the jenkins againThis patch did pass on my machine with the modified committed the patch to trunk andThank you Raja for the patch. +Moved from SPR to ROO for further investigation +Is that ok to reschedule this for? I dont want more changes in Branch before we release a community that work is well underway with AS all previous community releases are +My understanding from a conversation I had with Sasha is they have found their problem in a network when you receive this notification if you remember anything relevant please reopen the case. +True but why risk dropping something you dont have to when you can just use a deque +Expected UsageI imagine that large users of Hadoop can port their jobs to work in this framework and add Circus to their QA pipeline +Each run of a workflow has a globally unique ID and it has a name that distinguishes that DAG from other +Thanks for reviews +this looks like ant code that gets executed and ant has a debugging option from the command line +If we internally implement the clients of those specific concerns using listeners that would be understandable but not radical +seeWhat about the encoding is it really +any additional thoughts or comments on the patch +Revision now includes automatic execution timing information on exit if in development mode +Dump a pile of HTML into the log then scoop it up with a regex match in the Build Description +Snjeza any ideas on this one +Community has requested ready in so bumping into M +Im planning to commit it next I have some free time. +This is likely to have been fixed by constructor splitting +Ignore the Phabricator comment and patch for D that was intended for another JIRA +character in generated script names is now replaced with dot +Setting and +RECEIVEDCOMMAND which might possibly cause problems for code that only checks for DODONTWILLWONT as per the Javadoc +I am glad this was accepted +That JIRA goes about this by making detect the right branch and take the appropriate action +Claus there is one more issue I am working on +to be compatible out of the box +This column mismatch is not recognized by released jars +We recommend you use if you need this. +After further research I believe the problem is more widespread that just relationships +I have corrected that line to recompile and orientationchange event still not being fired inMissing some steps +Attaching patch which implements support for wildcards in facet field specifications similar to +I managed to run a specific test but when I run I get an error saying The input element of the launch configuration does not exist +This patch creates a mapping between column index and variable label so that when we can read the final solution out of the tableau +I just said the behaviour of in is undefined so there was no reason to support with NRQ at all +call should go now as the datetimes in facets never seem to come out in that pretty format +Will postpone this to and +So you have to click twice Does this cause two requests? Or is the Ajax request only sent with the second clickIf it is only sent with the second click this has something to do with your form setup. +Sergey can you fix the compile issue and reupload the patch +Well that would be unwise +Remove the since flush is a noHup operation in Refractor the rotate mechanism Compute once an hour +As these methods are protected a subclass would have to do this in another thread or make the methods public composite service policy is switched to stop all services that are STARTED or INITED +I cant see a reason why this would be Ubuntu packaging relatedA NPE is something in the code and I dont see a class or so due to a missing libraryI tried it on my setup and I have the same issue though +Xufeng isnt functional at the momentPlease clarify whether you have run patch for TRUNK through unit test suite +We have the ability to introduce guidelines gradually so implementing each guideline corresponds to a single commit +I am attaching a patch with test cases demonstrating how the plugin works now +In addition theres now a unit test that demonstrates the bug +I think there only minifying option that makes sense is to separate iocore librecords ectThis one mostly does not Sent from my phone +Hmm Ill specify the id field info +Is there a reason why this is only? What did do to fix this issue +This is clearly a duplicate although Id lean toward my solution of using slack instead of a hard coded k +Thus I suggest we push this forward +Thanks for the patch +mrepositorycomgoogleprotobufUserstucu +Updated now that has been fixed +with svn svn diff r and hack out the change in the middle manually from the outputwith git create a branch rip out the offending OTC change and squash the other think theres deep surgery to be done here still though +The main reason I posted it to jira was so that you could take a look before it went in since youre our resident JUEL expert +The servlet at testing is responsible for creating a node at testingfoo +there are some extra jars which i have added for infinispan +To avoid any bias to the existing code and make the same mistakes we should just delete trunk completelyAttached it is a script that deletes everything +Please see response for the SVG example +Does anyone have a good workaround for thisHow do you activate a profile based on the existance of a file whose path is relative to the pom +contact and take help from B +This is not controlled by Arquillian but +Will start working on other queries as well and upload complete patch for review as soon as it will be ready +Ahh +Is there some reason why that isnt suitable +From my point of view I would like to get rid of them by merging them with the current base classes +At the next restart issue will not be presentThis means that sling restart is successful when xml file is not empty +Rolled this up to close this when is closed. +We need to be consistent. +bq +and what happens if something goes wrong in that thread +Sorry Brian +that matches this region first before we do the loading? Otherwise patch looks good +Can you attach the whole javascript that fails to get stripped successfully +Is this something that should be part of lesscss plugin? + you submit the patch again with a longer path than just the file name? Also can you include a patch for aThanks +I see three patches from you +I have a patch almost ready that does just this +What tipped me off is that in the mobile spec you can start a watch on the compass accelerometer and do a page change and you still see logs from the callbacks +no nodeupdate events in CS queue +bulk move from M to CR +Finally a third stream is created which starts with a to restore the initial graphics state without any modifications especially no changes to the ctm +reproduceble in +org portal +I dont understand what you are trying to do there +Resolved. +If someone wants to provide a codestyle for something like or another IDE it would be welcomed and we would try to make any changes in tandem +Hi RaulYes XSLT and Validate are failed at the same place that the exchange is passed into the converter +Update +for to trunk and branch +Could you please check +Fixed it by saving the content instead of the and recreating the every time the pop up comes visible +I set up a merge branch and began looking into jar dependencies +I agree with the approaches youve taken proposed classes are added +Moved to +There seems to be several ways to do so its unclear which is the bestI use the one above beacuse it is the cleanest most easy to understand but I have no idea if there are any drawbacks +If you have to grant permissions to all there is some sort of problem with a privileged block missing somwhere or something like that +i read the comment but i dont found a +Closing all resolved tickets from or older +That way we could avoid CAS and also support timestamps by just appending a unique value on the server side +OK thanks Ill do my best +I suggest you put the test instead in and use +just realized I have my +m misc fixes to bugs flushed out by Python implemented as a wrapper over the V C libraries. +Apply font parameters from updated of a bulk update all resolutions changed to done please review history to original resolution type +Attached is a workaround that works in many cases +Screenshot of havent checked before + +I have checked the patch and it looks great so Ill committed that one +bat files +Im a bit confused in the above code theres only one validator per field so how would you know if was working or not +Any other hints? Im resolving this issue as the snapshot is now available and Im just having issues connecting to the server +In our environment running as the real user is important +Thanks Andrei. +some refs to parent missing the relativePath this was fixed under a different Jira and I dont remember which file it was +Hopefully it will save someone a lot of headache in the future +some were still present and then generated some problems in the interface like when viewing the content of an entity +Could it be that we need a new Interface for this? The first is used throughout the code and it does not make sense to implement a in all the places the interface is used +Main thing is that bug will be fixed +As long as you dont provide further relevant information as I asked Im not gonna pursue this any further +Cannot reproduce on Linux even with the same versions of Nokogiri and libxml +Have you reproduced this in a version +Any progress on thisIs this proposed to be fixed in +r. +But I really like the idea of a daemonWe should rather have these options saved when creating a db via query parameters or headers +maybe its fixed +The problem is that consumer may have +The argument that is passed through to other generators should be passed through to JSW and used to generate a wildcard in the configuration file +Why cant you use the to implement a fail fast iteratorThe second issue is the killer of course +If your XRX was with NOCHECKS you would get that behavior +The merge of into is running into locking issues +Im confused now +Ok Ive reopened the issue +provides both get and get so user can determine which port its bound to +EAP turned into EAP so Im changing the summary of this JIRA +Bulk closing for +After trying everything I could with my limited powers I only managed to make it work by downgrading Roo to version in which does someone had the same problem and found a more elegant alternative please let me know +HiIMHO this deadlock would never happenThe start level for is as specified in and the bundle is installed within jbi feature which means its startlevel is I cant see any chance the bundle could precede the bundle if follow our designI will close this issue first feel free to reopen it if Im wrongThanksFreeman +Nigel and I had this conversation earlier in the day and that was exactly his idea of the modification. +Revisit open problem +Thanks +Some codes are added to +Only issues that are actually fixed should have a fix release. +Then it works +Can you give more details on what how does it fail? Is there an error? If you set debugtrue in head is there any error in the log at the bottom of the page +So all issues not directly assigned to an AS release are being closed +Instead of adding a dependency onWe added a dependency on +There are probably separate for those but still I would prefer to verify this JIRA in M +So this indeed seems to be a bug but a very risky one to change in the short term +Hmm +Now it works fine for meTested on two different servers with the rev +So somebody please actually CHECK IT IN this time +Need to investigate whether there is a real use case for this both in as well as +Is there a way to trigger solaris build for uncommitted patchno. +Attaching a jar with the output files and file of yet another run +This particular issue might be fixable but theres a whole host of other issues that could surface +I was going to compile the +as hbasemeta where hbase is the hbase system table. +Ted I noticed you marked this Reviewed +Documented in the EAP release notes +Thanks Enis +Thanks Daniel all tests pass for +Moving unresolved issues from to +Dhruba is especially interested in this approach +does the runtime check actually use the user specified configuration directory and not just a hardcoded default one +I also updated the docs to reflect the update +Good catch but the sequence is irrelevant for the cache key +This looks good +We changed the client apis again so I have to change my ycsb +At some point the compiler needs to figure out everything that is being captured by an at or an async and whether it is the lval or the rval that is being captured +If we shouldnt include quotation marks in translated messages we should change similar occurrences as well for instancebq +This adds Oracle and tests as wellThe purpose of this patch is not to ensure that all databases import data in a identical manner +Ive committed to trunk and but there are conflicts on +You have two versions of each patch and +Oh well I guess thats the reason why why nobody is trying to ask me to do these kind of custom patching +It looks and works great +Enis got some distraction +Smart tree row keys comparison implemented to skip rendering node if it is a part of the bigger subtree that has been also queued to render +defer all issues to +Yes the cardinality is accurate +The most interesting solution is here as already covered by the Spring ll revisit this issue for Spring seeing whether there is anything to provide beyond is not considered to be within the scope of the core Spring Framework project anymore +Anyways it can be done later as a new issue. +see rd map +I meant to type rather than +Everything is working fine for me now so it appears to be fixed +Internal politics at make it hard for me to adopt packages later than those in the current debian distribution but having said that I can certainly lobby for permission to create my own httpclient debian package for s useSecond let me put on my LCF hat +Hi TedCould we do that fix in a JIRA? The script is already quite useful in its current formFeel free to report in and Hadoop and link together +I just deleted that job from running and have not been able to reproduce the error. +I like the way you handled the optional inclusion of the classes in the jar file +The proxy service instead may do that effectively provided that you lock it down to proxy only specific calls +I tried the test code in and the result was that I did not get a hang but rather got a blizzard of diffs involving parameter conversion errors which seems to be the intended behavior of this testI also got similar results from writing a trivial line standalone test programSo those experiments seem to indicate that maybe this bug has already been fixed by some other change along the way over the last monthsIll my results some more to make sure that Im not overlooking something +Remove the Thai website +jar to our SVN repo now Hudson CI will take it from there +Missed couple of files +Looks great SimonTheres just an issue when displaying a rule description in the resource viewer the rendering is not as good as in the QP pages or in the issue search page +There is a lot of code so cant send it all so will try and find is not a forum so wont ask a question but related to you comment lazy cached even if it is lazy cached it cant afford to be that lazy that when an object is called by the Rule it still returns nullo Could that be the problemFor now we are good +r +This is not good +I changed it textxml and it started working. +The latest patch look goodI have committed this +I didnt remeber that correction +Assigning for by reading and compiling with tycho applied to M branch and +Fix attached +Hi HenryI am facing a similar issue related to the ant version +Hopefully this is the last addendum +ditto +Modifying my code to manually encode as did fix the issue +Thanks Knut Anders for committing this +I like the standardization with the output of + +I will commit thisHowever a test case for this would be appreciatedsee or test Ill keep the ticket open and wait for a test case based on testall the bestrogerr +This implies that the NN becomes a limitation at to node range +Something like thisComments +Running into this in all my tests +I know that threads will be removed with but nevertheless it is helpful when loading tons of data on a multi CPU machineI try to provide a unit test as new issue with reference to this issueStupid question if threads will be removed with does this mean that the redesign of dataimport is by default so that no threads parameter is needed or will it be sized down to? +Only that it not cause major problems +fixed as per better to put those in seperate issues +Please give me a hint wich one i should use +Reverting from patch available status +I would rather say this as a bug instead of an enhancementThis is one of the critical issues with large files for clients behind routerfirewall +But of course and must be included otherwise you wont be able to use them successfully in the Wicket components as you will have two copies of the class in two different classloadersAnd about Wicket making your usecase fail suddenly is because the internal handling of classes has slightly changed recently but what you encounter is just a side effect of the misconfigured includeexclude patterns in the first placeGlad it works now and that you use this feature successfully +Updating patch +Also adds some javadoc to these packages +patch + +Yeah I thought about to +I just committed this thanks Ari. + +We have not yet gone to production with but have seen other messaging brokers lose their connection +Just I wanted to point out that splitting is going to happen soon so we do not do much rework +I will look into this and also the tests +So the reassignment will not complete +Dont think its critical enough to put in +Question to Barys Why It is impossible to setup c client to create new session while reconnect failure? can you please elaborate? You mean when automatic reconnection fails due to the case above you run some commands attempting to reconnect explicitly but the client fail to be connected? Can you paste relevant part of the source code please + +addition unhealthy healthy restartedreconnected removed +for example mine is eclipseplugins in that folder shutdown eclipse backup the existing unpack the patch jar xvf apply the patch jar uvf org restart eclipse full build and see if there is any difference +svn ci Sending admincommandsrcmainjavaorgapachekarafadminmainSending admincoresrcmainjavaorgapachekarafadmininternalTransmitting file dataCommitted revision +I have done brief testing and will test with manually deleting some blocks from the datanode to trigger new code + +You can in fact create tables with column families that conflict with HQL tokens merely by enclosing the name in double or single quotes +I think this is fine now! Thanks +If you implemented this its because you think its better so its should be the default +No need to backport this feature. +Hi ThomasThe documentation does not specify that this is a connection to an Apache Directory ServerAs Apache Directory Studio can be used with almost any directory server It explains how to create a new connection using the default port for the directory service +I needed the seelctabletrue +Im not competent to judge if it should be committed or notIts nothing to do with the removal of deprecated test classes so should be in a separate JIRA pleaseOtherwise the changes become too difficult to followAlso it would be helpful to have a test case which exposes the faulty behaviour +No doesnt look like my bundles been released yet the latest one on is all resolved tickets from or older +Fixed in the trunk thank you for reporting this +I have uploaded the new version of our movie on XPart will introduce how to run our test with eclipsePart will show us how to fix a bug of harmony with eclipse +Currently is targetted at JDK so I think this needs to be looked at ifwhen beanutils moves to knows its running within a environment it should be able to do extra work +Removed getsliceprefix and added getslicesuperbynames +We are using our own version containing not yet commited patches anyways +Also in many cases asserts help tests fail early +How difficult do you think it would be? I think this would touch the store file scanner maybe memcache and for the client something like during scanning would have to handled +Merged to +What MP did you test withApply my patch it changes to usebq +Quick note havent forgotten about this one just been very busy hope to verify integrate soon. +Heshan Sorry we have not applied your patch in a while +First crack at this +when tests are executed from maven should be added to let maven install required up to M. +We got a build system now +In Html Editor all works right +Andrzej could you please test your application with the latest SVN snapshot andlet us know if the problem still persistsFeel free to reopen the report if some new data becomes availableOleg +Im running Google Chrome on Windows +This jira tracks a change that originated in +am anxiously waiting for the fixRajat +Committed +Ive just committed the patch +Do you have something more concrete to propose or perhaps a patch +So there is no complete memstore insertMaybe the special ICV group commit makes sense +That would be the most straightforward way to bring support for new skins +was a special case that took place over years I care more about the practical effect of the change than I do some rule about when I should easily screw existing users +Daryn increased the header buffer side to K on the +I think the exception should be more descriptive saying what type did it expect and what type it found +Removed usage of the database from the session service +Perhaps the page could progressively render the like Jenkins console output +I wonder if this static metamodel would be suitable for different jpa providers +Regenerated patch +Has this ever been testedShould I reopen this one or what issue has Hugo Palma created that might still be open? I would tend to reopen since this describes the same problem I still have with. +Keith This is done +Thanks Tim +Ive attached an updated version that has some slight changes and there are a few things that I think need to be addressedFirst there were a lot of unnecessary formatting changes to the java code generator +We can do something in addition in this case but I just dont see any downside to provide a way to disable input counter +Grid heres another one to look at as far as tests go +Please open a separate jira to track this +The trunk branch has been replaced by version +Marek will know for sure +Will commit it once I get for the patch +It breaks if the DN listens on all interfaces and does IPC to the NN on an interface thats not the bond +Thanks Xavier. +Thanks Mikhail Roman +Good to know + +The new public class needs some javadoc +Patch for branch that includes relevant portion of trunk changes plus conditional exclude of for + duplicates in the Application Server project where it will be resolved +to Roberts reply on Oct +I have one suggestion though +Wanted to get confirmation first that this really does need a patch +Attaching another patch after changing get in and KFS +Are there any more comments from others? Oleg Jeff? I will commit this onetonight if I do not hear any complaints +Thanks will take the time to review and make suggestions on the weekendDo you have any plans how we can automatically test this +Did you update the jenkins jobsNope I didnt +Tests should be run by the surefire of tests can be done by configuring the surefire plugin properly. +Wiki syntax is now enabled for comments and descriptions in the LDAP project. +This is in master now +defer all issues to +William thanks for the missing unit test +Passed +I agree with Simeon and Filip that there is a problem with extracting the hostname from the object name +POST PUT DELETE query string parsing and POST parameters are now implemented but not tested much +This seems like a bug even though your problem solved +I just didnt want to spend too much time on this +Can you provide a sample input document XSL and expected output documentBetter yet a unit test +OK I just gave a at itHowever can we have an updated version +There was something missing in my shell script +Problem with blocking UI was fixed in rFor authorization problem Ive created separate issue +Currently statistics object is passed from the client to the file system +Just tried +Was able to change most of from to TOPPHowever the scripts directory has David Winslow as author so I will assign this issue to him +It works perfectly well with my Firefox beta and Safari on OSX +integrated to trunkThanks for the reminder Sergey. +Id say we simply cache the child node names as well at least up to a certain number +Never mind +IDEA keeps on insisting on putting these in when I create a patch +This is at least the third jira to fix mounting home dirs +Patch applied on and all issues that have been in resolved state for more than one month without further comments to closed status +Okay +yes go on +The devil will be in writing a reproducible test case for this madness +Putting the generated files in removes that headache from people downloading and trying out bookkeeper +Thanks for pointing that out I had missed it +Deploying Hadoop in Karaf kinda require for ease of use at dev time +I see your posting to Forrest and will apply it there andmaybe also in Cocoon webapp docs +cursor +Nicely figured out +Implement todds suggestion reduce to +IMO ephemeral nodes have to be deleted on session expire otherwise we lose consistencyWhat I suggest is to allow ephemeral znodes to be parentsWhy would you want to have pseudo ephemeral znodes which dont get deleted after session expire + +Ill mark this as resolved in version still needs to be fixed for +bulk defer of issues to +Trunk only is fine but then we need to change the tags to instead of +we dont set tick until after we receive the first message from the follower after the leader becomes active +Committed revision +Namespace added in rev +Manually tested! +Did you actually get this to compile? If so what other changes did you make +Ok will try to do that +Thanks for considering the issue +The types seem right but there might be some issues surrounding Runnables that wrap other Runnables +Looking at the stack trace just execute hql query. +If the problem reappears on the test machines we can reopen this issue +This is a blocker because looks like an issue in replication and could be one of the reasons we fail a test copying test results +No concerns from me ok to proceed +Link with +Patch file +bulk defer issues to +Thanks Colin. +I guess the reason for thesse problems is the class in combination with the lazy init mechanism that I implemented +Also while working on this code I didnt notice any code that shouldnt be in that file +well is just what is directly bound to ant so that explains it +So the right offset is not relative to the first byte of the array but to the last one +Sorry the path to the web app was invalidBut the deployer should throw File not found instead of Invalid JAR file +Then in we can expose this feature as solid and official with supporting user documentation +Are you sure that the rebalance doesnt require all partitions to have a leader? My experience earlier today was that the rebalance would fail and throw after having stopped all fetchers and cleared all queues +This should be handled on the coordinator it should use its own view to prune leavers from the consistent hashes +I didnt mention it here before because I had opened a separate JIRA for that and didnt want to confuse the issues. +Could you reopen this bug at the Maven project level +Bob is an ASF Committer +Assigning to Alexey to either suggest any workaround or decide if this is expected time Dominik first open a AS dev discussion if this is really JIRA bound at least state why this is JIRA issue not just simply stating the obvious wo any explanation on why you consider this as an issue +I would like to start on this +Can anyone have a look at this? Its fairly simple to reproduce and I provided a patch +This is a fairly old issue +Leonardo thanks for reporting back! Good to hear that it is working for you now. +Fixed by Grzegors Kossakowski. +Revised error handling completely in cddaabbcdcafda added several ITm still seeing this error in latest surefire project has unit tests that exercising quite a lot of rd party libraries +Bulk close for Solr +The in your patch looks very familiar +Attaching test log +However the plugin manager is coded to support build plugins +clean up srcc +Duplicate + +Actually the reason I wanted to have a jar generated was for Eclipse to use +Committed at subversion revision Touches the following filesM javabuildorgapachederbyBuildRemoved the requirement for documenting the release managers Java and Java environments +And how about missing this issue if i use +Btw I am in GMT +Fix pushed to master in cad and to in eee. +Thanks +You didnt it still works Thanks for committing +Close after release +No tests but all the code is ripped out straight from Hadoop Streaming +Feature implemented +upload trunk thanks +This time Ive changed the title to Update docstrings add requires annotation for required arguments +Both xdoc and fml validation is now implemented in the corresponding m plugins +Marking closed. +Patch applied as proposed +George did you perhaps miss the file in your commit +Is there any movement on or workaround for this ticket? I need to use Oracles SYSTIME and cant maintain config for the number of tables involved +was OK +I cannot figure out how to make the work though +Hi HowardI havent done anything with the components in my +Sorry about that. +Retested and this now appears to work +I like Retos proposal of rearrangement +Thanks for the patch +It will be nice if this Japanese tokenizer uses them +Turned the visible display nodes in Operations Editor to always be Visible +It changed the cached classpath and used the Hadoop profile because binhbase doesnt know and had no way to specify what profile was used for building +I also addressed this in +The problem with onException not kicking in is because of using Spring XML +Linux +I create them now similar to what we do in Apache Commons but I cannot test the result +Hibernate throws concurrent object modification exceptions if i lazily instantiate a collection in my domain class i ended up using and using lazy lists in them and mapping nested back to my domain object graph after form post just to workaround Spring MVCs data binding restrictions +Modified to use would you mind testing this out please? I dont have a Windows machine into SVN as revision +This patch changes to be the identity hash of the Subject and equals to check between the subjects +parakalo. + +The failing tests are unrelated to committed this +It is applicable to one case and one case only the use of connections during authentication +Has anyone encountered these failures recently? If so can you hare the directories for the failure? If anyone runs into this in future can they please attach the directories and exact failure to this jira entry so we can make progress towards resolving the issue? test rig where this happened is no longer in use and the problem hasnt been seen in other environments +Here is patch for fix for the problemThanks to Arpit Singh Pandya and Ratnesh UpadhyayChirag reviewed the patch and I think it is done exactly it should be done to fix the issue that is to just remove the existing adjustments from the database and add the complete new set of adjustments back to the databaseWhile adding adjustments existing code does not acknowledge the existing adjustments and neither it update them +The portuguese one is actually ptBR the brazial variant of portuguese and I guess activates only under that specific combination. +Things that we intend to fix have jira issues and thats enough documentation as far as Im concerned +If tag is on method omit this unless explicit setting is provided +Ok but generating inavlid HTML is still a major bug +We can discuss there +Dcoumentation updated in r +Hence changing the Bug description +This way the design is more generic and cleaner +This failure appeared for a long actually and cannot be caused by recent CU enabling +Implements most common pattern for solving such problem in Ofbiz +Let me pull in this exact testing scenario +I do not propose to close this bug +Is it the case? Does this comment make sense to you? Do you see it differently or have a use caseAssuming that it still makes sense cant you implement it as a layer on top without modifying the API +The method names being changed have not yet been released so that should be fair game +checkVersion is looking for minimum version of +Somehow I missed this bug +If so can you close this one as dupe +I prevented setting a null scmResult which is the cause of this inr of branchr of trunkSometimes updating a working directory returns a null scmResult maybe we need to file this in. +Thanks Joe Crobak! I just committed this. +Dougyes it should solve this issue as a side effect +I tried copypasting the file into srctest but had the same result + +This version will be made available to the user when creatingediting project details +On my Macbook Pro I saw a throughput boost in this caseii The throughput gain is less impressive if your transaction grabs numbers from a single sequence generator +comments on your findings Thats a good idea actually it will make the code much simpler +yes +I would like to suggest two suitable names call invokeThe method method should be called before the init method would be called just like it is in the case of the setters +I was being overeager in determining value equality for dependencies I dont think we actually have a use case which mandates we match on anything other than GAV +Committed above test for null. +how about this we require the elementType when we serialize a non enumset? if the elementType is not provided we throw an exception +I think a few of them are full programs though +Thanks a lot +An alternative is to configure the in samples and then launch jobs from command line like this mvn execexec myJob +and are just errors on my part +issues will not be resolved in time for this CP +warsignerinfostoretype filesignerinfostoredirectory certificatesattachmentstoretype diskattachmentstoredirectory attachmentsaccountstoretype fileaccountstoredirectory accountsdeltastoretype filedeltastoredirectory deltassessionsstoredirectory sessionswebsocketmaxidletime websocketmaxmessagesize adminuser welcomewaveid listenerexecutorthreadcount waveletloadexecutorthreadcount deltapersistexecutorthreadcount storagecontinuationexecutorthreadcount lookupexecutorthreadcount include waveserverdisableverification truewaveserverdisablesignerverification truedisableregistration falseenablessl falsesslkeystorepath sslkeystorepassword changemeenableclientauth falseclientauthcertdomain disableloginpage falsesearchtype luceneindexdirectory indexesanalyticsaccount thumbnailpatternsdirectory seems like you are using a modified WIAB version instead of +I am new to using this tool +Hi RalphIn my cp CUserseweiwan exists and jar files +Incorporated Joydeeps review comments +Its easier if the number of instance types is limited a la AWS +The changes were prettyI did include a to bootstrap the configuration in so I have attached that fileBTW using Spring as an object factory rocks +test application attached +Sending UsersclrapachejdotrunktcksrcjavaorgapachejdotckapipersistencemanagerfactoryconfigTransmitting file dataCommitted revision +I think I have a fix for this +Fixed with +Tests rewritten to selenium . +This looks like a pretty serious issue for people who will want to embed Tuscany in their runtime environment so Im changing the priority to MajorDo you have a test case isolating the issue withRuntime Class Loader Contains all Tuscany classes Application Class Loader Contains SCA application classes and has Runtime Class Loader as parentIf not then well need to create a test case and add it to our integration test suite +I have a fix that I will attach later today +I just committed this to only +On other solaris platforms it doesnt show the prompt but you still need to press Enter +I missed that +But thanks for the suggestion +The error youre seeing is actually caused by the fact that two of XHTMLs schema documents try to redefine the same attributeGroup in different ways Xerces is right to throw an error here +This bug affects the weighted average view and not other views +Hmm not really but this is the only explanation that I could find for something that happened a few weeks agoIs this included in also? We used until recently and honestly Im not sure if it was or when it happened +Thanks for tracking this down I have created to track fixing it. +Not sure if it should be this front and center and should be moved to a tab or a collapsible section but the HTTP Rest and SOAP reference bindings will all have something similarAny feedback would be great +Hi Raman This is fixed +I agree with Gordon that the is a separate utility +Committed revision . +Similar yes excepting that I need to define the mapping myself not just depend on a particular charset +Turns out only Xerces needs to be upgrade +Wouldnt editing creating a inside the deployed module be sufficientOr would you consider this a hack +Please close this issue if you are satisfied +Thanks for reviewing Sijie. +Patch after incorporating review comments +Thanks Kanna +Marking fixed issue as closed. +Work well but there is a typo in the rule description Replace Ruby by PythonCould you also change the violation message This statement doesnt produce the expected result replace use of Python SQALE model has been updated with this new rule +I think we should remove the call inside IW to? This is a holdover from when was the default +Please upload a complete wsdl +patch for for lower was already added to HB +Well include this in a larger upgrade +bq +Can you double check if this is the script you are running + +Framework +I take that back +turns out this was all just programmer error +jar of the java program that needs this native libIf the source is checked out in a directory without spaces the build just works +Resolved +Thanks for your replyBut I still dont know why is different from others. +Here is an updated patch I also added time difference for phases along with finish time stamps on +Vladimir changed the port range to and the test now always passes +following commands will be skipped duringcommands in commands in commands in commands in netapp related commands are removed from and contained in a new properties filefollowing commands are also removed from add them into or external firewall commandsadddeletelist external loadbalancer commandsadddeletelist +With the chown enhancement we can make it so that an MR job doesnt have to be launched for importing hfiles +Revision was a bad idea +I have reopened the issue to make sure I check this again with the FUSE distro Ive only tested with Apache you provide me with another memory usage screenshotRegards another screenshot of the memory pool when using JDK +I have commited myself to a project until December and so I figure Ill be here until thenI got a perm +If you are building with msvc then you should not compile +But plugins references the jars requiring an really this is an example of how another layer of indirection not only solves a problem but increases reliability when it is used properly +Dmitry ive committed a new batch which fixes it for the collection types your patch lists +JinchaoCan you prepare patch for TRUNK as well +I have removed thatThe idea here is when the region is tried to move away from the RS that holds the META the region is actually closed +Did some minor tweaks for passes +I have commented why it does not work and how to make it working in that issue +Or one should inherit from the other rather than repeat +If making the effort to change the preview behavior depends on and that is slated for then well have to let stackholders weigh in on the prioritization +Sorry about that +table +Still the third container did not get guice installed +I tested the patch and found that my private callbacks do get called as intended +Added the latest patch making a wrapper only +The only thing missing is the negative test which Ill add in a bit but I wanted to get this posted sooner rather than later +Working on a fix right now +Max The errorwarning issue should be fixed by the duplicate issue +Closing resolved issues. +Yes I am aware this feature would be Groovys specific +Perhaps a finer level might be better if the errors are set to be ignored + +What is the size of small table? and the number of rows in small tableSmall table has around million rows +so much for ever warming searchersIsnt this a problem for certain unit tests? After adding a bunch of docs and then searching to see if they are there must you pause for a bit to make sure enough ms have passed +Richard do you have time to look into this? I think at least we should return HTTP status if an exception was thrown +I fixed it in a different way +some stuffs been movedaround +fyi the last vowel in leverage is an a. +Applied a modified version of the patchThanks Scott +Ill change labeling according to your recommendations thats easy! Thanks for your quick response +It can be handled in a couple of ways +and the file serverdefaultconfprops is the one that I saw was not working +Attaching the patch +Thanks Arun! +No problem! Thanks for getting back to us on cos I am adding some extra docs to address this issue so that others dont run into the same problem +New request is created for solving this purpose +Instead of repeating that code it is better to have the api which simplifies that +The patch tested on windows and Linux +This is improved inPlease try and close the ticket +I have also found where they suffer the same problem +Recommend you try and modify the test case srctestjavaorgapacheactivemq to show your issue and open a new Jira ticket to track its progress + +required no changes +Issue is fixed please verify +SendingSending rttransportsjmssrcmainjavaorgapachecxftransportjmsSending rttransportsjmssrcmainjavaorgapachecxftransportjmsSending rttransportsjmssrcmainresourcescxfSending rttransportsjmssrcmainresourcesschemaswsdlSending systestssrctestjavaorgapachecxfsystestjmsSending systestssrctestjavaorgapachecxfsystestjmsAdding systestssrctestjavaorgapachecxfsystestjmsSending systestssrctestjavaorgapachecxfsystestjmsSending systestssrctestjavaorgapachecxfsystestjmsAdding systestssrctestjavaorgapachecxfsystestjmsSending systestssrctestjavaorgapachecxfsystestjmsAdding systestssrctestjavaorgapachecxfsystestjmsAdding systestssrctestjavaorgapachecxfsystestjmsSending testutilssrcmainresourceswsdlSending testutilssrcmainresourceswsdlTransmitting file dataCommitted revision . +Implemented comments posted by Rafael Schloming +Dave no idea at what version we are but definitely were behind version which according to the release notes brings significant performance gainsIm trying it out and see if no issues maybe update trunk with to be applied in ordericuejsonlibcurlsnappyincludes +is it happening randomly +Applied to branch in commit af +Shoot this really should have gotten closed with Beta and ongoing work done against a new issue + +decided not to bother with this for just trunk for now. +Thats strange have you renamed the tests to and also checked that your test methods are void test and not def test as per the release notes? +I think Ill make sure that I have this setting set to false then ignore it for now +I assume this is superseded by +or is the right place from mypoint of view +bq +Hi +here I have integrated a dojo progress bar +Turning int SOCKET on is one example +I also verify that the format of the connection strings is correct in the test codederbyall passes all testssvn status outputM javaengineorgapachederbyimpljdbcM javaengineorgapachederbyiapijdbcM javaengineorgapachederbyjdbcM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapi +I dont think so Dart +Will switch my mailing lists back to a personal email so I dont miss anything on the list as soon as I get there +Hearing nothing this is marked resolved. +It can occur in mapped mode but is much less likely. +Fisheye config builder a project edits project and then requests that the fisheye config is ve configured xplanner and done the bits that need to be done +patch was generated using branch using trunk this time +Theyll show up in the nightlies for this platformIm sorry for the confusion +Sounds like the JAR may have been present on the classpath multiple times due to WAREAR packaging issues. +What do you think Jrn +Im wondering if adding a value attribute might not fix the issue +Verified in. +Attaching +closed +Unfortunately time really is against me at the moment +Clover indicates good code coverage in this area and unit tests pass but I havent checked the unit tests in detail to determine if they functionally test this particular piece of code +To configure for SSL support would need the following in the configurationattributes the JMX Object name for the to be for an example +Essentially were looking at many of the same tradeoffs we have with the distributed Lanczos decomposition stuffThe mappers and reducers do not use much memory for side storage +not pulling this change in many other Hibernate modules Envers will need this module +Closing. +Lets get this in and then worry about getting fancy +I attached an improved version +tested on windows xp sp and macosx +Made a mistake +It sounds like possibly this happens during log rotation and the disks can not keep up with logging events +I am hitting this in Version So is it really an issue in greclipse or notAny workaround +no do not need a random inthow do i test this code how do i get your branch +Sorry FreddyI found some of the error messages I was getting in that thread +So this will require an upstream change in GWC to make this configurable as well as a change in the GWC integration module in to set that configuration option +Attaching patch with fixes as per stack +Adding a rpcTimeout parameter makes this easy +Im going to evaluate the behavior of RI how tolerant its to unmached localvariabletypetable elementsGould you wait and not commit your patch Ill prepare a new one +For now though Mikes suggestion covers the two major JSF implementations and allows for others to be plugged inWhat do you think Leonardo is it doable +if SICS is configured to have temperature sensors but they arent physically connected display not available +So all issues not directly assigned to an AS release are being closed +perhaps +This wont get fixed for sadly we need to send this to the future +Provide a config option whether to send a notification for read error or notJust to clarify this a little more +I had a feeling this would happen some day +I have added fix to remove accesskeyxxx and secretkeyxxx from api query string and api response also hide accesskey and secretkey from STO printout in all internal agent command logging +I think Matt Massie has some work on this somewhere do we have an open JIRA for RPCAlso SASL support is a strong requirement for this to be used in Hadoop +It could either be a data store or it could be a coverage IO plugin when we get done defining that +I am reviewing them now +Created Velocity attribute renderer +to trunk +I just committed this +If one of the files output or rejected are empty the job deletes it on we restart the application the mentioned error occurs +Commited on trunk and enabled by default commited on as well but there it needs to be enabled by setting true as a system closing all issues that have been in resolved state for months or more without any feedback or update +Isnt this a very serious regressionIn my builds fail cause the surefire plugin is to new is not yet feasable +Its content provider gets a list of all the active extensions and THEIR content providers and asks them what to menu action handler gets a list of all active extensions and passes the request for menus down to the is no support for the main view to pass the doubleclick action to the fileset extension or to the xml configuration extension or to any of the of right now all thats part of it is content providers label providers and handlers to get a valid menu +I was thinking that for this CNFE to disappear with jboss we would probably need to change the listener or add myfaces jar as well +improved context handling and added SvenCan you explain better what is the improvementI read the changes in the code and I cannot say that I understand what exactly is the improvement +Let me know if you cant resize your dialogs but I was able to resize it in my browser +Going back to the previous editor or selecting another due to licensing is another issue. +This needs a lot more thought +I just committed this +This is missing in the new layout +cluster on xenvmware without DNSSame as aboveOK +in this case railo only allows call from other components not from cfml files like defined in documentationavailable only to the component that declares the method components that extend the component or any other components in the package +Otherwise well have to make up database specific statements for these type of queries I guessThe other option would be to simply stick with what there is currently or to just raise the limits to the maximum on all databases differently by keeping the database type and their functionality +Im certain that would eliminate the ability to run multiple solr instances in the same servlet container because of the way the and much of hte information is implimented via Singletons +As the revision introduces only tests it is safe to commit them into release branch +Added to sandbox because i cant test it on windows +Trying to figure out the best way to approach maven packaging to support both Hadoop and Hadoop One idea is to employ a maven module that uses a profiles to specify which hadoop dependencies to use and those dependencies are inherited transitively by the other modules +The Seam guide is checked acorrding to the issues no the Figure +In the future we could perhaps do this with an implementation of +Does not affect Wicket as this is not an Apache release +The then uses only the ordinal from the to reference a which prevents us from creating Map instances for each term +Adds null check to set rather than where its calledI guess theoretically someone could override this class override set call it with null and set the member var to null then encounter an NPE in no which they couldnt fix due to scope issues +Close it as Not an Issue then +really its not that its EOL for if there are serious bugs I am pretty sure we can roll another release to fix those +batch transition to closed remaining resolved bugs +You have to pull it out of the test logic but after setting it up you can submit your jobs quite easily and debugHave a try +Standalone Tiles does not currently support being run in aportlet environment so there is no reason to include this JAR +revert the commit for license issueProbably I can add sample using openJPA implementation which is also apache license later CharlesFreeman +Check for null return value before using + fix based on does not work when Synapse is deployed as a WAR +with and so closing this issue. +However I noticed that when I click on the project in Hibernate console it says Persistence unit not found null +on change but it breaks +This is now implemented in the trunk. +It looks good +It attach a view listener for and a component listener to viewRoot to make the view listener be restored each time +When focused on build definitions then the show projects page will show build definitions instead but each build definition will have its own status name version group and build now build history working copy and delete links +The other jobs happen to be so a task takes more than one core leading to even severe CPU contention +I believe the two steps I had to do wererename tables name fromMETADATA to metadata in ZKcreate namespace in the accumulos instances ZK directory +for patch +So I think the patch is good to go unless there are issues on new chanes or the issues left open +Implemented in rev +Integration test added +Thanks Lohit +It makes the assumption that if ENCRYPT gets notified of a TMPVIEW the node is trying to become a coordinator so it should assume keyserver duties immediately +We arent worried about fixing this issue here any moreAutpexport tool is now an Atlassian Plugin and any issuesfor Autoexport should now go thereIn addition as Confluence is being phased out as a CMS again we are notworried by this issue it will of course remain as a Wiki but that does not needthis issue fixing for that to continue. +I dont think delete true is a valid commandThere is deleteIndex and I recently added delete and delete as other options +Ive been using the following patch locally to work around a similar problemJava takes the last xmx opt it can find as the actual value so moving the arg to the end of the opts seemed to help +This causes an invalid JSON encode because +linking to ticket for tree for the dep mgmt you want to resolve everything in depMgt to make sure the versions exist? +This may contain a very primitive interface for managing concept provider which will be refined in review +Can you comment on when it will be fixed there? Or is there a patch release available for the EAP +Please if you disagree. +Seekable and SyncableThe definitions of the methods are in the Seekable and Syncable Java interfaces +Im not sure Im understanding this +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +thanks +applied to both trunk and closing all issues that have been in resolved state for months or more without any feedback or update +Yes Id love to play around with a patch for Christmas +Right +Although technically possible I dont see why a bundle would need a dynamic import on core packagesHow does Equinox behave WRT this? Id say a bundle that uses a should fail to resolve if it does not import It should still fail if it uses At this stage Id rather be strict +Since is public and is public static we probably could preserve the existing API and make it call the new one with status falseIn the interest of time I will just upload a patch with those changes on your behalf +I think theres probably a bit more cleanup and refactoring to do but whats attached is fully functional +Thanks for providing the patch +A new file was missing and is now included in this patch. +Wickets could cache as well Jrgenwe run into the same issue under load using the with Spring have a workaround for this using a custom +This should be toggled by the configure button on the application bar +Let me run some experiments and post on the jira +See +It also includes everything fromStill need to update for deprecated options and new and optionsAlso will open a JIRA after this to make watch master so that when master moves they will get the updated location +generate patch after svn up and in root there is a diff in +Still not fixed in hibernate still not fixed inAlso it doesnt seem like a minor bug people are obviously running into this quite a bit +better set the initial value of max to be MAXSIZETOMOVE +In my opinion the subclipse patches are fine as we can ignore leading path names when applying the patch as I told Edward after you have left yesterday +How it interacts with libhdfs I have not looked into +I this patch to address recent changes in the code and to preserve backward compatibility +Patch added +Added extra spaces to URI examples so that the angle brackets display properly in JIRA +assigned to me by accident +Do you have a short tagline or description for Capedarf? This will be used on the back of the card +The web services are deployed in the context of the classloader associated with the but when the activate was being performed from the BPM console the subsequent activation of the web service was being performed in a different classloader context. +Changed to the arraylet style model in +It looks like solved the lazy field load issue but you want to extend it some more so Ill leave this open +This way if someone attempts to execute an slink action after a conversation has expired then this would check that for the old viewId before the slink action is again for the original page has already tested before this method is called +Uploaded the patch that uses instead +If I back that out everything is fine +As far as the ticket comment the submitter says that installing can potentially interfere with a preexisting sun JRE install +The case sensitive flag should be final and it should not be possible to change it +I think we can resolve this issue for now +Please see the attached screen shots +I was thinking about basing its version number on the JDT version number but I have the feeling such a change would cause problems +I think this issue can now be marked as fixed. +This should fix it +Issue fixed +Bug has been marked as a duplicate of this bug +additional information is added +NTS is for multiple DC cluster +Elliot how would you suggest we pass the from through the ipc? The region server will throw an actual exception wrapping the message which will be understood by the ipc layer or the region server does not throw the exception but set the exception message to the response +Either someone needs to tell me what Im doing wrong or a fix is needed +ms inv +Thanks ill let Will take responsibility for the unicode escapes patch if hes willing +Yes in this case we can as the string does not represent a general URL in which port is optional but always represents a hostport combination +Oleg I based my changespatch on the HEAD rev +removed the warning onThanks is the in the folder +Otherwise looks good +It fits in with the design of that are under development alreadyCan you attach the native code that goes with this? Then I can create the and refactor the regular IO to use the package +thanks for taking this one the AS runtime detector checks the attribute in the manifest of the this attribute contains EAP the server is recognized as EAP +I just did a clean checkout of castor from CVS and patch rejected the previous patch +Spec issue is +I think this is fine as a default +Eli this broke MR tests compilationAre we sure we are not breaking et al by changing signatures for +This patch adds basic support for the UPDATE action of MERGE statements +I added missed check of super class name +Not test case or other logs +In other words shut down everything and only open up the exceptions +Would you like me to testverify the suggested fix +The permission checks were checked for value createorupdate but such a switch case was not present in +Alexey it looks like you missed the test +Attaching a patch for this issue to explain my approach for fixing this issue and to elicit feedback +Added support for however possibly locale encoded by some other means such as a query parameter is completely ignored theres currently no mechanism to let Routing module know about the alternative locale encoding strategybulk closing issues that have been resolved for more than a year. +Im not sure it would actually break anything to slip a byte array into the message properties we could just make it so that a user that tried to set a byte array on the used for Message properties received an +is this similar to? it was the reason the tests were disabled on that setup +Lets see if fixes issue +latest versions og logj UIMA logger implementation and to allafter long time I uploaded our lates version of logjs UIMA logger implementationI hope it could be useful for you +Changed default formatting in. +Thanks! +Finishing an accidentally submitted incomplete issue +I think there might still be some questions needing resolved and code needing polishing for the authentication stuff that might still affect the proxy but those issues are more general issues and not necessarily +Small clarification why does there have to be an invalid message? If its a separate methodmessage when the new client attempts to send it to an old server it will get an easily recognizable exception back rightFor the old server its an invalid message +It was never clear exactly what the point of this contribution was. +Even in Java we end up doing gymnastics to support standard types and it is making interoperability difficult and +Good to hear that this worked for you too +Bumping this to unless Jim can get a functional patch in +Will create patch to commit the completed work when repository access is restoredFiles attached The DITA source files for the entire document Archive containing the Generated document in all three formats The Generated PDF document for easy download and COMMITCodeline patch containing the completed Working With Derby files revision +acts as wrpper on top of bytes stored in memory and the temporary file on the file system +Oliver apparently sometimes I miss the obvious +Please provide a Roo script to reproduce the issue +Closing this bug Based on Prachis and Alenas comment +Tweaked this slightly to support +Tools to manually load sequence file to database and aggregate +I am trying to find out how to isolate it on Solaris now +leak +This is resolved in the BETA installer +Update the patch +If you want to do it yourselves that would be great otherwise Id be happy to do it in January +Does the editor ever need to know the rsversion part? If not why does the EMF need to knowIf EMF need to know then I thought EMF were supposed to be extendable without causing big changesEspecially for such a thing as a namespaced xml element it should be possible one would thinkIf this is the only change we do then we should be others are just bugfixes right? No new features beyond this field or has time shown otherwise +This patch removes the above optimization and sends the leader and isr request for all partitions in OfflineOnline state Note that in the first request the partitions in state are not included since they never had a leader +Ill add that method in a separate jira +Im not sure we should do this change +Should be done now +An updated patch for the trunk +We cant skip just complete empty rows because there is no way to tell if the complete row is empty based on a result from a slice predicate +assigning to Igor to possibly to someone else in +does this need to go +The problem is kernel relatedI upgraded my kernel to and the bug is fixed +Javadocs class has an execute method +Accesses to them are already synchronized and the norm case is that writing without reading +Includes views simple spatial tables point line and polygon classes +Resolving +Note in this case is not on the classpathI think if myfaces is used we could include inside our but if mojarra is used include it on a separate file as part of tomahawk and add it inline as shown +Anyone want to try this +Perhaps you are trying to make light of the matter but I hardly find it funny +that should fix the issue +this task was done a while ago +I think the reason for this problem is that the create components at renderer timeWe should try to create the components in the or in the tx alternative +this patch is an addendum to the last patch +This problem should have been fixed by r +patch solves the problem nullifies latest initialized instance of refires static initializer for patch committed in r +The feature works partially +Certainly fpm is enough to get by for now. +We really need something that can follow requests across servers not just reporting what the coordinator sees. +I am fine with throwing this record away and put a warning to the user +Done Running Tests run Failures Errors Skipped Time elapsed sec +It can be removed if reviewers want it to be removed + +I think that issue covers this functionality in common doesnt it +The fix for has introduced this issue and I feel that it should at the very least be documented + +After the posted explanations its not an issue at all but an improvement over the old behavior +as it is in other osgi smx components +Committed to and trunk. +Let me know how the tests come back +Crystal clear +Would it be fair to move the into the regression suite and the regression suite tests that pass into the master suite +added wiring for and fieldbridge support on and documentId for discriminator implemented +It works perfectly now +Fixed with a doesnt currently recognise a root page named anything but ve created a dummy Home page that includes Project Overview for fix this soon weve got all the metadata to do this Ive just never bothered to fix it +Uwe Big thanks for taking care of this issue! Hoss Man Thank you for your input +To gain a slight performance improvement all Equals codes are dublicated +Simple patch that exposes the data member variables to child classes +to branch and trunk thanks for reviewing Stack. +This fix is being included in this build. +Updated the patch to address Jasons comments use instead of synchronized methods wrapper not to use for updating so quickly Karthik +You are asking to probably ignore it or something like thatSo let me ask you something +We have a separate issue for updating the docs. +Replacing this patch with one that turns of the debugging code withboolean constant set to falseI did leave the debug code in the patch since I think it may come in handy in the future +Im about done just need to fix one last unit test +Additionally I formatted the file using a definition list which makes finding the functions easier +Patch does as described +Another option would be to say that the application gets an error if the Primary datanode fails +remove the leading when generating iframeUrl +Fixed by throwing a on the deployment if the root it references does not bdaf +That would pinpoint whats failing +Hi Christianthis is the patch after solving some svn conflictsThanks Christian could you please this issue open after reviewcommit the second patch because something more has to be doneThanks patch appliedKeep up the good work Marco +do not have any external references normally so they are getting cleared earlyFix should be fairly easy though. +added jboss target only for deployment to jboss as +Thinking to add remove inside so that once MR unregister with RM RM goes ahead and cleans up the app state without waiting for the actual AM container process exits thoughts +Thanks for the unit tests +This would eventually resolve these problems +As I suspected this is interference between tests +If somebody is pulling data from a secure cluster into an unsecure cluster the data is anyway going to be world readable +People can then use the tool for models that arent executableBut I see that you are talking about a view in Modeler that would only show the constructs that work on the engine +Can you please verify this on Linux as it should be the same +I dont really where the jars end up I just want decent logging out of the box that pretty much leaves you with logj +It seems returned a random dom element Stevens javascript fixed the thing in my case. +Repro for this issueThis repro needs a local namesservice on port +Every connection definition has a Max connections parameter +Im not opening scanners at the client side +approximately of Popup Panel is fixed in context of. +This should be resolved in the release +Sure +Done I was able to merge the patch into current source ll be included with you +IT has been corrected +Hi thanks for the contribution! Ive added Justin and Mike to this ticket they will review and apply your the presence of a couple of java clients around Im tempted to close this one using straight java code as wont fix? +bulk defer minor issues to +Because the Context doesnt actually behave like a Map +Also test is very flaky and almost failing in every build +I just committed this +Closing tickets +In looking at the source code for version it seems there is a race condition that may cause the server thread to notify the client before the client waits +Im trying to get the unit tests rewritten but they are failing right now +which I cant +If we want people to use this extension point we should document it +However I think that this project is ready for nowThanksRick +And the code never goes through the usual persist logicBut outside the test environment Ive manually tested both create and split and they work fine +Basically CASretry for files. +Set the XALANCROOT to that absolute path and just follow the regular build instructions +works on in the apientity started the work on apientity branch but now have another assignment so this work has to be delayed +This can be cleaned up later but the docs for a replacement API should be well proofed for the releaseget is temporary? CanisFile andisDirectory be removed? These are redundant with get Overall though +So I voted for this to be fixed +After consulting this issue with a GSS representative the conclusion is that we dont want to include this information in our documentation +I will do that +Active Directory is still buggy +I think we can let it go for since timeout monitor can handle it and there is no better way to fix it because the region state in is not so reliableFor this one is not covered yet +Is it ok if I test this issue with the next release and then close it? Would this be too late? Are you the one who will close the ticketRegards you using designer from its master branch or a previous release working with Guvnor? If you are working with master you can just fetch the latest master and build but if you need the fix asap to work with Guvnor I will have to backport the fix to the Designer branch and show you how to build it yourselfLet me know whats best for you at this point and we will do that +Ive added iWork support to the container aware detector in r + +The same problem exists with other SQL objects such as functions +v vThat was a bad idea I thinkDCheers Paulno worries thanks for the feedback +Heres a patch which implements read repair for rangesFor consistency ONE the header is sent and the responding node performs a consistency check on each row after sending the response back +i think it should work yes because we only set the params in the Resolve step and i guess then only one is being made +Whats the point of thatCan you explain some moreIt doesnt make a lot of sense to me +Thanks Amar +This was fixed with. +The changes sound reasonable +Thank you sir for removing the unnecessary counter size checkI actually see the inconsistency with renaming the method to addCounter +Another issue reported by someone on IM is that an entry in hbasetables is causing an problem where you cannot create a table with a previously known name +How could I replicate the problem +Any suggestions +Next time I will try to use the +This has been amendedAs for the error message I disagree with your interpretation +I would imagine this is something we can request name translation on the INFRA ticket as well +These were velocity dependencies that arent used by Solr otherwise +cp +Originally my pom was a copy of the version +this is fixed in version could you try it out and let me know if it works for you +to and trunk +Ithink only uses this API if the caller calls seekExact? So Ithink the current location of the assert is both more costly and lowercoverage than if we moved it to +This scenario should be fine for rename +trunkCommitted revision +Also the stale classloader detection used in the UMD doesnt hande classes loaded by a delegate loader +any updatei will try to make the serializable for my current project maybe this workarround will work +The task for CR is to validate the concept and if valid to make sure it can be implemented in a later release without breaking compatibility +Ok let make it clear what do me want to be the out come of the ticket I thought that we want to change CQL parameter behavior from to real maps for commands like create keyspacetable and alter table? Im not a fan of keeping old syntax because that could lead to confusing mixing problems where users would use old syntax in one place and new one later in the same statement +For instance given a set of search results it can go do recommendations based onbq +I understand why there arent any tests for this but worry about the sleep statement why choose three seconds? It seems one of those sttings +I thought EMFs might be a useful helper class to implement several methods including this one +Looks good and tests pass +The failure will be fixed by +Bug should be closed as Xalan has already been released. +if there are any suggested improvements or feedback would be happy to resubmit an updated patch +all issues to the CR +In order to not break old installations we can keep the old data and not use it +Committing this +The upgrade of to fixes the issue +Ivo or MartinWould you please attach your full server log? The above is not the root cause +Patch which uses the correct but slow approach of synchronously committing the checkpoint each time we truncate before fetching restarts +bulk move of all unresolved issues from to +Moved writer and output stream getters to mime response +So we can access the directly from the without implement Servicable +Im also bitten by this nasty! I think it should be taken seriously +Apache Facelift? had problems with my mailbox so I think it is too late my proposal is Apache about Hatchet or Axesince it is similar to a Tomahawk +Alexey just would like your confirmation that this then just not an issue and expected behavior and should be rejcted +lck files do not get deleted +Any issues should now be brought up in their own Jiras +Marco could you try to run the query the tracing enabled +orgapachehadoophbasereplication +I have posted the i got +The only thing that does not delegate is the attribute +Apache on Mac OS possible please attach a example and reopen possible please attach a example and reopen what else more do you need from him? He already attached a CFC to the ticket +This is an important issue to a few of my projects so Im eager to get something in place +Tim I can still get the test to fail even after the change you put in to close the replyConsumer +I need a way to get at the Servlet instance registered for a mapping Ive had a go at this but couldnt find how to get it could you provide an implementation for this? Ive added the method but it just returns null +Closing this issue AS starts and stops fine now +Fixed that bugReplaced deprecated scan API +If you can please update the patch because a fair number of refactorings took place in the meantime +This is not a bug +Perhaps you had to check for SMTPnot sure +bulk defer minor issues to +Things have changed quite a bit in the docs but all the ones I could find are fixed closing as FIXED. +DanielThank you for working on this! I tried to apply your patch but it generates an error +I did the partial backport of Knuts changes to my codeline and with that backport the full upgrade junit suite and derbyall suite ran fine +SeanSorry for the late reply and yes +Does it make sense to come up with new content or just link to the section? +Of course I would like to use if possible but as it is not Springs original intention I already hacked it not to close until the EM throws an exception and it seems to work fine +The conclusion was protobuf was the decision to use +I am closing the issue as I cannot reproduce the problem +As of and there is aggregation support +I feel wired if it only prints out a bunch of new lines +Thanks for the feedback! Ive submitted a review request as +Patch for in patch file with full paths previous patch was taken from the mapred subdirectory +WFK Seam is cleaned up from Embedded +We were previously requiring all cache entries to have bodies although this obviously doesnt make sense for caching a +In fact I am skeptical that OSI would approve such a licenseStopping there for a moment does anybody disagree with either of the above? Continuing on +So the first one is usually justified +please comments verified at rThank you for applying this patch +Spring upgraded to Notes marked Not Required as component upgrades are not included according to policy +It seems jankier than it was earlier +Btw I was using binaries built from trunk r +I think you last statement is the issue +well it works now +Can we try and make Todds work? It does some nice cleanup +Note the unsupported version error +So now the session is live on server +In order to return the same object wed need to cache the hash value for an object as well as caching he Long value that goes with a specific Fixnum +Attached is a patch which instead logs the length of the byte array +Hi KimThanks for the patch +Your fix is correct +And of course you can always checkout Cocoon in branch. +Thank you Jing for the review. +Dan Since a plain text client talking to an server will receive some gibberish youll get some DRDA protocol error message from the client +This is a duplicate of issue is now closed +Do it the following way Shared networks Always send speciftrue Isolated networks +Ill commit in a day or two +Sowmya can you clarify what it the baseline here +Please close the issue +Never mind I got confused with the comments +We could pass both of those to the or we could move the timestamp and index name generation to a default andor abstract +But if error must appear then please reopen this issue. +committed to as commit dccbecdfaecfbcecedf +Ive been able to run the tests +David which view are you doing it in? got a screenshot +Straight forward port of patch but shouldnt be committed yet as a couple other patches went in first +Patch to change and to be static nested classes +It will default to off and we can continue to work on itSo unless someone steps in Ill commit what Markus has put upMarkus have you tried this out at all beyond the unit tests eg on a cluster +This is to make sure that reproducible builds are reproducible +How did you come up with the x factor though? Is it something off the top of your headI wanted to differentiate between a significantly bigger unallocated heap to force a grow if it makes sense so factor seemed to be a good start +Thank you for the patch Liyin. +Thanks NathanJust waiting for the CLA to be approved +Test program +This willcause cygpath to convert the path in pieces and avoid length problems +Committed a fix for this issue and added tests. +If we get to the end of the consistency check with no errors then we use the copy of the permutation where we removed all the candidates that caused conflicts for multiple cardinality requirements +TCK module really covers this. +We really just want to check that some process has received notificationsIssue The connection will eventually timeout if not established so setting a different value should not make a difference +Patch attached for +Thanks +Honestly DOM is just truly brain damaged here. +Hi Michael any update on this? It would be super helpful to see how far you got +Verified that the body tagId inherits the scope from the html tag. +I am trying a couple of methods +It is hard to imagine what at line could be causing the NPE but I would guess that the that Remko got above is because the webapp has already been undeployed and the class isnt there any more +This is a tough one +Modified the issue summary in order to better describe the issue +Seems like a low risk new feature +If you end up hitting the compilation problem thats the biggest help you could give me now because I cannot reproduce it +Amar Would you please review the patch? Thanks +Fixed by commit andPlease verify +I can help with any merging problems of +This is definitely a +Not sure if we still have this on our roadmap +bq +Dont you want to move the test to as wellWoops right Ill do that before committing +Definitively possible but slightly messy which decreases a bit more to me the appeal of this ticket +But I think we should throw an exception in this case since its a horrible trap now? User will get results but thats flat out silently wrong +should CHUNKSIZE be a configurable settingIf I knew how to access Configuration instance from withinbq +Applied at revision +Hi ToshiyaI tried your test and the subscriber seems normal after mysql has been brought down +Pls check this issue resolved by then we can close this issue as duplicate issue of issue is resolved +The plugin must be installed at start and can be modified subsequently + +I think a better fix would be to change to avoid sending a kill to the RM if the reported job state is terminal rather than just checking for KILLED +to going to fix this +if we stored seqnum for each KVWe have talked of doing this in the past a few times +Resolved in trunk +Fix applied in trunk. +Ah I ported the perl compiler from the php +Im not sure why we need to register others here +Reopening since this needs more work and needs to work just like seems this has been solved on all is required is the port to trunk +Can you provide your example? I know how I would do this but I want to see what youre doing in CSS to do this +I might also recommend documenting how to telnet to the nodes and part and make sure the network is not in the way of a cluster starting +Can you tell where most of the time is spent? Should we create the in parallel +Hi MyrnaI believe that both and are resolved +This is patch to bring Tuscany to sync with the Axis RC release +Should this be marked blocker + tested! +resolving this as wont fix! +Thanks must have been the recent renamingregardsMartin +I believe this is expected behavior because the eXo JCR that uses is not a true JCR in terms of the spec +I was reading the contains check as being done on the zk list + +Deleted discarded patches and attached corrected updated patch fixing errors in +If you can attach a sample project that demonstrates the problem that would be even better. +Thanks John for the sample +Vladimir this is in develop +It is either a problem with fieldsreader or deleted docsSame as previous with +for linux will become available in rhev which is scheduled forexternal beta aug and full GA in December +Resolved in ER. +Passed all unit tests locally +Im definitely seeing a problem with a couple of real plugins +I just committed this +SUPER FAST +Reading through your source code it seems that after can be pushed in but install cant +The analysis above is only correct up to stage the real issue is a deadly embrace between two faulting servicesThe sequence of events is as follows we have sent message form jBPM to ESB The message is processed Reponse is sent to jBPM callback service Callback command is executed and exception is thrown in jBPM code This exception is handled by pipeline processor The pipline process will send message fault message CHANGES HERE The fault is received by the original ESB service targetted in Original service is not expecting the fault message but attempts to process it This processing leads to a NPE This pipeline creates a fault message and sends it back to the jBPM callback service Go back to As the callback service is receiving a response from the ESB service it should not send a fault messages in response +Sample Implementation for CDI and Singleton caching only +Lets take a new to +see it checks the number of clauses in the qualifier against max +Pleas close this issue as an invalid duplicate of issue Kind regardsUlrich of PLUTO +So I think we can resolve this issue and handle related improvements in a separate issue +Ill address it +Also testing as this needs a good few tests to verify the functionality +Say they reconsider with this auto submit in place they would have to fill out all the credit card information to go back and change their choice +I just committed this +Last patch was as r. +Write wiki see that this ticket was changed to Provide initial support for +completed testing and found no new issues +Attaching a patch with the test mentioned by JasonReading one more record if the split ends between the delimiter bytesPlease review +There are many types of incompatible changes today +eclipse Project with quickstarthttplocalhostxyquickstart shows the pageIf you choose a file and click upload the file will be uploaded and show a feedbackpanel messageIf you click save Mainform the Mainform will be submit and redirect to the same pageIf you click categoryChooser and inside this dialog click on a button you will get the exception +Updated RN be documented as Resolved Issue. +The attached patch file was created using svn diff +Changes backported and hupushed to server. +I dont know why this happened for me these tests pass +Looks straight forward enough +This is due to the limitation of the method that renders the header title which only runs if title is not empty +Patch uploaded +bq +They are now called when there are no authenticators registered to handle the authmeth +repeat entry for uniqueField times or until it this should mention that im using grails with JRE on Windows xp. +attaching patch that removed them +Why dont we expose it as a configuration with default so that any one can change it if they have an env as stated by your usecase +Thanks Shreepadma +Still i am getting same error +Maybe turn it off vi logj config +Also tests which do not have much value but add non needed rigidness to the code +Please looks good applied on trunk. +I see that there is which implements and so intercepts master ops to check counts +Its not ready +I went with a New Host +I will have to finish work on that issue and wait for you to commit it before I can create a working patch for this oneI guess its about time for me to start usingRegards rush Bene this issue can be closed later no need to have it closed before applying other patchesThanks +Confirmed still an issue in Apache Flex into develop branch +Many thanks for the patch Nikolay! +So I get the nextBoolean but not the othersPresumably the sequence goes up so that explains nextInt +Should I do that +Right now this is a working assumption but I will use your test case to validate this +I am forgetting how concurrency will work in this case ieinsuring visibility due to the JMM +Also note that my first entry shows how to reproduce the problem without ANY consumers and producers +We should think of the two as disjoint in this way +Can you please test this manually on a real cluster before committing it? Just running the various hdfs haadmin commands with and without security should be sufficient +Whats a use case for flush iterators? A user can never be sure whats in memory what has and has not been flushed +Essentially Roo supports input text boxes and text areas through the JSR Size annotation +Fixes the problems so that ant jar completes succesfully +Patch to fix BURL string in JCA examples +The other central class of this stuff +Will pull it down and prep it for commit +Antonio Suh was joined to this project +implemented. +I dont think a should be possible improvement should be added in a release but apart from that your solution looks fine to me +The only way to avoid the error is to remove entirely +Itll probably be a couple weeks +Patch with a quick fix +Patch applies to and trunk +Not sure if this is repeated elsewhere in the migration code but it might be worth further investigation +Thanks very much +Stefan Ill go ahead and mark this ticket as resolved +Ill see if I can get to it today. +tiv files left open +I also need the data to graph in +Sorry to hear that its worth checking though on the mailing list whether user impersonation wasiswill be supported from the +I think dblevins said he had some code that does this in openejb +I dont think so unfortunately +class is empty in the constructor +The patch contains four things More extensive testing of processing Username Tokens +If I explicitly set HADOOPHDFSHOMEHADOOPHOMEhdfs then it works fine +Thanks Keith just checked it out and the problem is indeed fixed +Hey +The! This was a very bad bug introduced by me being an idiot +for applying the patch to both X and trunk with unit tests +If the property is not assigned to the component just the component name is used +Please check them and give us feedback if something can be improved further. +v +This was resolved in revision . +Thanks Markus +Updated the patch +Enjoy +This sounds good to me +org makes this a priority we may revisit +the of reformatting problem +The common patch removes the files and mapreduce patch adds them to the mapreduce project +Until we migrate to using the same theres no way we can support this +Please use this patch instead +If you have any objections or further comments pleasereopen this bug entry. +Ill wait a full hour cycle since my post to make sure no one raises a red flag +Clause to check for general keywords key was in there however it didnt handle null or blank key which defaults to general keywords +Fixed in SVN trunk and branch +Yes a very odd use +Zero means use another replica when the pinned one is any amount worse than another which we optimize to be the original behavior of sort by score +To be clear Im just against deleting because I dont think we have an adequate replacement yet +Thanks Mike +Committed revision +The token with its attributes is encrypted and signed +Also for future reference a context diff is the preferred patch format +the trick Ive used is simply copy toIt works for nowIll add that hack in the next rev of the snapshot +please refer to to Create a JIRA space for complete +Fixes a bug where a response for a closed socket can throw an uncaught exception +Thanks for review Stack and Ryan. +This convert to lead is defined in the requirement of itself +Bulk move to +Hmm +The patch passes standard loads may be I should add more tests for stressing VM creation and destruction +Thanks Aaron +However this will suppress all Sonar Violations on that line +Fixed at revision +After restart the service is gone +works for me now +Gaojinchao I dont understand your comment +But the tutorial and every use I have seen in years uses the new format. +So I will work to get one of the other committers to apply this patch +Perhaps but not so great for stepping through with a debugger or viewing in a hex editor to examine the bytesAgreed debugging is sometimes simply just painful thats what Im facing right now trying to get to the bottom of thisRobert your patch indeed fixes the problem for me +Thats why we have to tweak the so that classloader can find ll sync with team to see if we could add an alternative way to inject directly interceptor instances so that AS would be able to load them modulary and pass them to at runtime. +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Screen shots of VPE looks browser patch and looks after pathc was applied +It takes a very special application to require perhaps times more often than MS Windows does to keep it running smoothly +Leader started with zab not set or set to false it listens only on one IPThe log shows address XXXX when listening on all and the specific IP address when it is listening just on one +can u please attach your service archive file alone with the wsdl tryied to regenerate the problem but I could not able to do that +Attached fix for Tomahawks trunk +The problem is that the timestamp cache should be replicated asynchronously but its replicated synchronously due to state transfer being enabled +See SVN no one is working on this issue i can help porting from +Could you take a look at this? thx +This has been fixed and the button now removes the global as expected. +I am for the patch applied to the Yahoo! hadoop distribution. +bulk move of all unresolved issues from to +Problem fixed in svn will be put into next in CR +According to no should be handled on micro release branches +Closed this gap for now +Cancelling patch while Sids comments are addressed +Sorry for confusion in above comment it is patch and patch instead of and +We should just use the kernel for this when the mount is in mode it automatically intercepts all write operations before they even get to us +Thanks for that advice. +Erratum deletion changes s by purging deleted documents merging changes s by purging deleted documents +Since I was modifying the targets to a more appropriate I also took the opportunity to change to use macros since these support mandatory arguments which makes them less error proneNote the above patch also contains a rough version of a common ant file that the depends script now uses to check the property +As a workaround I removed package and eclipse started to work ok now +I wondered how long it would be before this bit me in the butt +well at least I figured out there must be something wrong i appreciate the review +Luke thanks for the quick response +Right patch this time +now im trying to hack the necessary SEAM interceptions and SEAM stuff in could also be used +Also can you reproduce this issue for Xen and Vmware +This has been verified in Alpha but verified again with JBDS B +Fixed mapaggregate jumps to views referenced by the part labelsThis is in accordance with the compiled engine behaviour +The current flow requires us to tell it to close specific regions by name +Only when a specific bug is found in an example is JIRA a suitable place to raise an issue +It replaces dash variable value with new one +See note that in your test case you shouldnt expect to see onRow as no rows are returned +These are probably all obsolete +Without cmake the compilation and testing also pass +GytisThis usually happens due to the dependent job failing +Situation was better next days +Not sure about the performance impact of this patch but it should do the trick +Applied the patch from Prasad +Wouldnt it be better to only append HADOOPCLIENTOPTS once in rather than in each bin script? Each can have a variable like Yarn if neededIt looks like handles the case where doesnt exist which is perhaps why all the bin scripts set it but I dont think we need to worry about not existing post +whoaa! oopsthat is meant to sub in the proxycant believe noticed before +There will be for example no setter thus no need for the special handling in Closure +A stack trace corresponding to the error +The URL of the servlet is followinghttplocalhost having had a look at the issue your fix seems to make ll create a subtask so that this gets documented somewhere. +Uploading a patch which fixes this issue +Currently the project is hosted on site +Ill try to get to it this weekend +Fixed page in wiki markup +fix paragraph and are known to be failing on trunk see and +to trunk +For deleting a user see +The transport is looking pretty stable now its a bit of a resource hog at the moment but that can be fixed over time +The upgrade seems ok so far +anyone wants to be brave and run java lets let them go ahead and test it for uswhat is magic about versions and ? should comment thatbq +Verified with JBT. +Ie you could simply load a FST from a FST suggester into WFST but the results would be bogus +I need confluence user names of those that should be administrators for that space +Paulo this sounds like a different issue +A release candidate is available +If you determine that there is a bug in please reopen the report. +Instead it should probably consult the snitch to get the DC so they all work +No IT +I cant judge whether its correct or not with internet explorer on windows +Patch changing the prefixed parameter definitions to use the FACET constant +Dave Can you please take a look at the findbugs issue +I like this algorithm +This appears to be working in. +it will be a mess if you depend on snapshots that continously can upload whatever you want under your group +Task status showing the completion of instant schema after went in +Exactly Todd This is what we have done in our internal branch as a work around for this fix +Ive updated the documentation to have an example of modelling ant using CLI +A third option is to statically import those constants but Im not a big fan of static imports +Updated patch includes improved file from +Lets give the robot a run on this this actually fixes a trunk problem +Can you provide a reproducible test case? We have not seen that behavior +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first. +Committed a patch +I have to notice that clicking on lines is a little hard if they are thin +Do you have any actual heap comparisons that you can show us? I just have a difficult time checking in a performance fix without some test manual or otherwise to show the impact it is having and if there is still more that could be done in a follow up JIRA +This is how the C works +This is a big improvement to the tag and several people have been wanting it +Thanks for the patch +CP reverted in revision updated in revision . +Heres a patch that lets you override the script that writes the version file +Since its a static this will affect any binding using Axis ws binding +Update the patch to make it compact +Closing since no activity was reported for more then a year +Obviously the newer Tapestry is doing something different maybe just swallowing the exception trace? but probably not worth your or my time to investigate further +the jvm property did not appear to be inherited properly by the the javac and javadoc tasks +Executed Repro steps and I no longer see the warning. +I close this issue for now and will see what happens next time the weekly schedule gets run. +One single view definition file really was owned by a different user +Is it possible for us to upgrade ONLY Avro with this umbrella issue? Is it absolutely required that an upgrade of happens at the same time? I was hoping that this would not be the case but can live with it if it isIf we can begin with the above then we can move on +Yes thats where it is sorry I didnt clarify +patch attached +Thank you for your quick reponse! Will be released on the th as stated in the road map? Sure I can test against a RC but I am hoping that you can supply a fix on the line because we must be ready for production in the start of December +path Denis +Integrated to TRUNK after renaming hbaseThanks for the patch Subbu +Excellent thanks for reporting it. +The first part of the software update for this is checked in +Works fine with NB +Retargeting to fix in since has been shortened and made more focused +push to +I have the same problem +Now everything is passingCommitting shortly +I am facing the same issue in my project where the wrong response document impl is being loaded for the response document +The highest bit of each byte is an extension bit +Sorry Ive made it automatically +Here you are required to place the operator on the end of the line or use a backslash to break the line +JuhaniWe worked around this by interrupting the handshake thread after a given timeout +I second Will on the use of Clover +Then assign the remaining regions randomlyhowever it is being done now +The test doesnt assert anything +Leo thanks a lot for your evalutationI agree with Leo on these two points so it seems theres nothing to change in the patchMaybe its time to commit it +Part of a bulk update all resolutions changed to done please review history to original resolution type +up later than I wanted to check +sample XSD file for reproducing issue output of command requested above + +Note that was for the client runtime +Is it possible for a podling to start off using GIT? If it isnt or if this isnt the way to request a GIT area please say so +Preserves the initial goal of passing renewcancel directly back to the issuing objectMore cleanup is needed in hftps mangling of services +Uncomment the last lines to make the task fail on trunk +Leonardo this is not a myfaces trinidad issue only +Eclipse likes to tell me that I have collisions +bq +We can use the next release to start warnings that the default behavior will change so developers get a change to guard their buildfile against future updates +Assuming platform or application issue since no similar reports have been filed lately. +Id recommend that you try and crate a test case to show how this is happening +In remove is called when there are no appId nodes children +Still the same with +Closing this issue was rejected in +Attaching a patch which fixes this issue and touches all scripts to check to make sure DERBYINSTALL DERBYHOME and JAVAHOME are set and makes this handling consistent across scripts +Ill update the hbase jiras. +Working on getting the tests to compilerun now +Also I think testing if the removal of the two subrules does break anything +See for the original issue +ant broke png filesbut yes committing to the installer should b efine +Committed the README +David Ive now committed your patch +Testing indicates this is not a related problem though the error is the same +Closing this record since is now released. +I have components A D on and B C on +Committed and added test case to trunk and branch. +One where we need to allow duplicate correlations is for so its not just about +Improve RPM release number handlingIncreased release number to +Committed +Please reopen it with the latest stacktrace if you still see this issue +No updates so marking this as RN Not Required +On the Advanced tab you can see we moved Channels to the top changed Audit Loggers to just Loggers and changed Properties to Drools Runtime Properties +Please review patch and commit +Verified in JBT +LenI will need to discuss this one with you +closing per comments +Hi Ronaldthe intention of this issue is to provide an easy way to test processes in isolation keeping the number of required dependencies as low as possibleBernd the code is committed I still need to finish the documentation in the user guideCheersNils +Patch to follow based on this +Thanks +BrunoWhat is the status of this issue please +I am looking at this for commit as its been a real pain +Please confirm that it has been applied as expected +It might be the root cause in factSince the Master relies on the HRS to report it if it never does well never really close it +That servlet needs improvement for navigating and selectively executing tests of course +keep up the awesome work! +Tests are passing locally for me so I guess they are flaky +I had it discussed on mailing list before creating this Jira issue +Instead mirrored the hierarchy with a matching unchecked hierarchy and now simply translating to those much like Spring JMS. +Now any of the binary distributions do not contain a file in the absence of that file all registry related configurations should be read by file +Issue closure will happen before release +A quick test i did on at least version does not seem to validate the statement +Can anyone give a pointer to the downloads for this fixThanks in has not yet been can build it from CVS if necessary. +I currently cant reproduce this issue +As far as I know there are still varying opinions about this support among the committers +still working on build +Sure worth a try +Its a web application and the piece of code above comes from a struts action class +This does seem like it could be Glassfish specific +Ive tested with the latest release it works fine from commandline +Would anyone want to try this +Vladimir can this be resolved? I believe it was taken care of with the upgrade to +This behavior does not fit in the cases Richard mentioned +Committed to trunk +Pushing to due to some uncertainty about the best approach as well as the desire to get out the door +We should commit this ASAP +I removed that in src also removed references to it +And alsoFor each request directed through the portlet bridge the Faces environment is initialized the Faces lifecycle is executed and the targeted Faces pageview is processedrendered +Fixed in revision the change also the the branch in revision for inclusion in the release. +you for your contribution Xuefu! I have committed this to trunk. +Lets log an issue for above +Karaf itests are failing for me and thats because needs guava +Any reason we only do it in mapreduce plan +When they are true they skip adding plexus default patch also fixes the building failure when surefire tests fail due to including +With Cloudera Search and Blur being available plus I think we cant close this as Wont Fix +Four questions Is there a minimal version of the patch? It seems to include some extra clean up +Actually the umask should be specified in octal +Its quite possible for the end of the file to be longer than the maximum transaction length +What should happen next +Values in the config file are clamped between and but this also exposed via JMX where there is no checking of values +Was taking milliseconds component of the time and not total millseconds +We submit a Job with tasks jts memory usage as follows +Im working on the patch and will post soon hopefully within the next hour or so +Sorry I forgot to close it. +Could you please try with the attached jar file +sample seems to work correctly now +Thomas could you please have a look at this for? you think you will fix this behaviorWe might either change all our to or apply one or your fixes +I think they end up in target +I take the latest code from trunk svn info +a font that rendered a capital E so that it looked like a W +The CL config must take place before the Resolver impl is instantiated +And there is now a better unit test to catch and ensure headers returned is in the expected case +What do you thinkSorry I meant not verify method but scan method +Looking at the source it appears that this bug will manifest there as well +What is this class for I think using is acceptable which is linked to +I have committed this +Sorry James for committing this so quickly I thought the original direction was an oversight so just fixed it without discussion +The patch looks good Knut +Yes Max you a right I found the code in and decided that this means that this fix was done in the branch too +Thanks. +Im going to close this issue until more information is provided +Anyone know if theres a way to avoid it +Rather than remove it any after all should be equal to any other I fixed it +I would support Jeffs proposal +I would like to link this issue as it relates to Stu Hoods previous comments that bytes will be returned as encoded bytes with the client not caring what data is encoded within those bytes +Marking this as resolved +I have read the helpdesk ticket but Im still not clear what is causing the bug and if the issue is being fixed for EAP +Is that more feasible? At least on failure there will not be an UNKNOWN status and a restart will be possible +This should be a self contained test case +It looks like an excellent patch to me I am for commit +we should really get jenkins to do the builds too to catch regressions! Sorry I put it in the wrong place wasnt sure how things are organized +thank you +That needs fixing too +It should be ok for these jars because we dont deploy them into ASfor the samplescommonlib I agreed with you we can just reference samples lib from the root libdoes it sound good +This looks like a duplicate of which was fixed in the latest release +has been exposed in TRUNK does someone want to take a shot at using objects here with? Or do we need another +Storm has its own configuration object and for better integration with Storm I would set up a Storm conf for that although in reality I would probably just never set it because I never want it to go down entirely and that is how I would get the maximum number of retries allowed by the clusterI can see other applications that already exist and are being ported to run on YARN like to want to set that config in a way that is consistent with their current configuration and not in a Hadoop specific way +We sacrificed step scope for this and I dont want to go back without a fight +Thanks Stack +But it sounds like its fixed indeed +The fix was to close the Git repository before deleting the directory housing itThis is a proposed fix +Ill go ahead and resolve this issue as wont fix given the discussion above. +Test passes on jenkins now +Youre right that for any such basename needs to refer to the classpath whereas for it is a Spring resource +Any other test application from. +Can we mark it as FIXED +Note that this patch should be applied before the inposted changes on +Ill change the key again so that its not quite as ridiculous as +Committed revision Thanks Tom for reporting the issue and reviewing the fix. +Verified during EAP test cycle +This works +When you think of the classBoolean then there should only be one instance of andMaybe those problems can be solved with Java using the builtin enumerationtype but so far I will have to provide a big table with all sons of my Enumclass +Patch for which implements feature +I committed the attached doc +I am relatively new to Faces though so I could be mistakenI agree with you that if there is a validation error then the incorrect value needs to be redisplayed +Out of +an admin method getPlan and added plan information to the at a trace level via the getPlan method +Please reply to this mail as its an important issue to be solved for me +The suggestion to configure the multicast port requires +Sorry I kind of lost track of this issue +That is correct +Ill have to look at some code to seewhether I have done anything at all or just analysed current behaviour +Fixed in Rev +When uploading an artifact via Archivas the user expect to see the artifact listed +then I dont think its too great an expectation that the Tools should be forgiving in this senseIts from that perspective that I thought Id report this behavior +Renaming +I observed inconsistent results before r +Please reopen with new info. +It could be the case if the QA account belongs to less than two groups +Its scheduled for M so yes. +good idea +Looks like this is from your mailing list discussion about so I have linked the two issues +The close can get stuck +Ive create the which contains and hope that what you expect in the bundle. + that was fast! Thank you for fixing this! Im really looking forward for the final Release! +Can you make the Quickstart even quicker for people unfamiliar w like me? Im thinking heres how you insert a bunch of data and heres how you read a bunch of data +This design is intended to preserve the status quo of nested iterators by adjusting their contract slightly +see if that fixes this test +row stateVar was ignored in saverestoreState methods +This does look like a timing problem with load balancing +XavierCan you post instructions on how to generate the XMLNothing is ever simple and importing the XML into a test instance fails with action is invalid and A user with that username already exists +Will do that later +I dont know of any high level overview of the JDBC changes similar to the overview you cited +No problem at all with adding Experimental +Are you trying to use logback exclusively and bypass the container logging system? If thats what you want to do youll have to set up your deployment to use its own classloader which filters out slfj from the container and include that in your own JAR +hive patch for generating maven artifacts for hive +The test is passed on revision I suggest to create a regression test and close the bug +Heres a patch for this +Reopening for cleanups of +Done the error has been fixed. +Yes the last on call was made in a transaction that was actually committed while the first two and the skip call was made in a transaction that rolled back +class javadoc should be more clearer of an original file that is snapshotted +Please revise +links are all checked in and just awaiting the new svn stuff the mailarchives guys have been contactedStill need to do redirect and custom search pages +Okay this is held up on which is a GPL file +and no we didnt try to remove any of the built in point line polygon or raster +fixed in rev +This bug has been marked as a duplicate of +Thanks Marcos for the patch we appreciate that +After more investigation it appears thats the test fails only when launching test via cobertura +Glad it works for you! Will close this against then. +The sample still references the variable as CATALINAHOMEwebappsin docswsm +How many regions per server? G or RAM? I should add a note to important configs +Patch modified from Barrys extensions have been added to the new model wizard to create these procedures + This patch dont relate to this issue but itll improve in the future +Restores to the default value when loading from the same property file +It will save us when the hbase recovery starts before the datanode is marked stale +And figure out the best way to register custom serializersdeserializers +Updating this issue to invalid as a part of bulk update +The problem is a client that implemented executecql cannot recover from this +See. +Hi Marcuswe are using the patch you provided it definitely fixed our problem +Weird. +Further to my previous comments we should have this behaviour Applications see all messages provided by plugins without plugin authors having to modify your application source code via a script Applications can override any message provided by pluginsThis is the correct and common sense behaviour in the same way that plugin views can be overriden +Might be out of date now. +Whats the use case some files have a timestamp and other dont? The one that dont wont really ever be extracted into the output if time is activated +Hello MadhukaraPlease attach a proper diff that carries only the changes you have done +The method now just forwards to this factory methodPlease close this issue if this resolves your problem +Fixed in Branch and trunk +This isnt being ignored its just that for the next little bit it would be kind of good to keep the codebase as as possible +Possibly related to happening in +If we everyone agrees then I can do a quick hack to make it so +I will review the patch and commit it +Didnt want to be pulling the original authors name if it was allowed +csv not good +While it doesnt explicitly use the same list query it does not treat web services any differently between SP and SP +Screenshots for Wicket for Wicket there anything under the advanced section in the maven configuration +looks fine +Tested exchangequeuevhost pages on both Firefox and Safari on Mac OS X +Nevermind my issue was due to a bad JDBC driver versionPlease close this issue as I dont know how to do +My assumption why this happens to me most of the time is that the order in which analysis occurs is not predetermined and that this order influences the outcome +identical to close for Solr +The differentiation between hostdomain? could still be done within the object +This attribute is missing on +Added patch for my test +No changes from +That should help if your connections are mostly idle +Also not happening on Windows Tomcat Java with the current roller trunkWill try current roller trunk build on Linux later today +The idea is simple just let the algorithm be executed on build view time form tags inside fmetadata and handle fevent because it is an special case +If I can use the same across requests with a fixed thread pool of something sane like or threads then I can be reasonably sure that this particular class isnt going to spiral thread counts out of control +Same with the other parameters +bq +Also I found that moving to the end of the config file will make the failed test pass +There is a third situation where cache and database result is out of synch +we already have a hacked up SLD xml to avoid validation errors for text symbolizer extensions +Trivial docfix patch +trunk doesnt afaik since it will always override Mx or Rx +If the roles are retrieved from ldap outside Geoserver and sent in an http header attribute no ROLE prefix is added +Example of bootstrapping guice putting the realm in the Security Manager and placing the injector in the that uses guice to return a that configures Jsecurity guice didnt look at the actual code yet but how do you add multiple filters to the same chain with that syntax +patch attachment for trunk +If I did the steps accroding to the description I could not view the content of the tables content via the admin console +Please go ahead if there are no further objections +Sorry I hit submit but did not intend to +Thanks Rob! Yeah I just hit another failure around. +Latest rebased patch +Tried it again and yep it works +The editor is adding a file to the deploy folder which is undeployable. +A test hung above +Hi what happens if a field has a double quote in it? Good question +The idea here isnt to solve the worlds spellchecking problems its mainly to get rid of the extra indexYes definitely +I decided that it would be better to change Jackson side to resolve this problem + just committed this +Like Tim I also get this error when trying to access a method +Applied the patch in smx fuse component trunk . +This behavior equal to behavior of new HTML file wtp wizard here you can add html templates with diff contexts +The main reason for own implementationinterface is to introduce methods which are not available in the interface from Selenium +copy directory recursively not sureWhen consider semantics and compatibility these should also be considered +Thanks +Sanitizing user input is the domain of yourapplication and not some library you are using +Patch applied +Because it will destabilize branch +It works fine now. +It looks like you forgot to commit the Disposable interface though +The test still passes because of the padding on the names +This looks eventually fixed. +Tests that failed last time passed locally lets try again +fixed in r +Since we do not have to deal with the complexity of a buffer cache I think this would be much simpler to implement than what is possible in traditional databases +Closing internal item xpp is not delivered in EAP . +This is a very good idea +Im working with Jonathan on the caching client +I think its ok if we introduced IFU for DOCSONLY at first throwing exceptions otherwise +Woops remove that unused variable +Patch for submodule that clears almost all warnings +I shouldnt file bugs until Ive had my coffee +iterations isnt to shabby considering where we started +Linking since there is a possibility will impact the deployment support in the the majority of and blocker issues to the next point release +defer to +Please verify +I have a parent POM that specifies a newer version of in dependencyManagement +Here are the changed files did a bit of code clean up and initialized the crowdClient in the class constructor we will now use a cached object instead to save that bit of overhead from creating it every time +Committed to branch and trunk. +Only allowing services is a reasonable trade off hereIf you dont want to use a service you can always hand wire things up. +Might be a straight solution or workaround could be the following Set a smaller timeout for the purge operation or check for existing locks in advance if possible Do not let the build fail if one of the purge operations fails just skip them seem to share the same problem here We have around projects in Sonar +This is not an issue for CC +I was wondering shouldnt there be some tests using the sqlAuthorization mode too +Oops +project and the +This means that those variables have to be maintained separately inorder to have a common path +But if nobody has an issue with that Id say we should do the refactor as you suggested. +great debugging +This is a possible fix that takes the gc threads off the transferQueue and creates a special slot for gc doing this we introduce a new mutex that ensures that even if we are within schedule it should be safe to schedule the gc am not sure of this fix and want it particular I cant seem to reproduce the failure reliably so it is difficult for me to test it this corrects it +Im on holiday until Tuesday so I may or may not be able to look at it this weekend but if I get some time Ill try to debug +Thanks Ning +Thanks! +Also remove deploy from all to close again with further removal efforts. +Fix checked in revision +This patches backports the numeric support to xUwe and Adriano please review it and let me know if the backports looks OKThanks +n movement so tested this again its still fixedWe should be using the BPEL version plugins which replace version +The attached testcase looks OK with current code +for keeping this patch Filipe +sure but considering there should be a mean of knowing the number of actually received messages we need first to understand if this is due to an actual issue or not +Im dealing with a hard drive failure at the moment but will look into fixing the as soon as I recover a decent dev environment +All other IP addresses are IP addresses are routed to that box +fieldName creatorName sizeMin sizeMax notNulladd field reference jpa class +see also move of all unresolved issues from to +Thank you for the feedback Dan +Thanks Andr +Remove ssh session checking at the action level since it is already done in the wizard itself +Abhiit looks like both issues are same is fixed now +I realize we may want to the example config and call this out or not +Denis thanks for your contributionI have added the osgi features and updated the json data format wiki page to mention gson as well +The main thread can send the requests of saving the states to the dedicated thread without blocking by using queue or something +Another suggestion for maybe a future development is to move this into the core query parser as it does not change any behavior as long as you do not explicitly specify any extension +Just committed this to trunk and +I briefly review the patch looks good +committed without the logj changes +Weird youre absolutely right but I could have sworn we fixed this already +I am trying not to change the and expected behavior which will allow us to use the current unit tests +Proposed patch fix +Moving all issues to the CR +Ill try to prepare the patch the next week in case its late for the release you are planning I think we should commit this patch as it is with the correct license header +Guillaume said no for it to go in in one of the comments here +Dont send HTML mail it raises the SPAM score significantly +Still to find out whether RAT recursively filters through project modules therefore there might still be work to be done here +Proposed changes +That sounds like a case for an optional request interceptorI personally see no problem with converting Host header to lower case per default +Harsh The reason of the error is the warning by the test codess using or some deprecated methods +Configure the maximum number of splits that a job can have +However thats a red herring in that case +OK heres a new patch +Verified that the HTML Books on the website now show the correct titles +we just need to set the error message somewhere and redirect to the page +Fixed in rev +Thanks Kim +The callers of the security layer need to be assigned appropriate permissions +Ill add a topic for itAnd I will revise the docs to indicate that traceDirectory and traceFile can be used together +Ive got a file this one improves +Heres an updated patch which includes a unit test +Why not humanize the block cache count hit and miss +to Brunos comments +I have addressed all your check it and let me know your comments +Thanks +So either provide a patch or consider commercial dialect property asc nulls first property desc nulls first property asc nulls last property desc nulls last dialect use the minus sign trick property asc property asc desc nulls first property desc asc nulls last property desc +I know its a big embarrassment that I make this wait so long +Just your exact JRE +I would want to index locally and push the index through a separate grid FS access manager +That was a bug in modcluster it used the original uri instead the one provided by modrewriteNow it works the user doesnt need a additional ROOT context to have the rewrite rule working. +Update to the error message +The error happens with but it doesnt happen when I use ve kept the caching but reworked it into a attribute which should work fine if the changes since the attributes will be cleared then which leads to a new getting created and cached +My assumption was that this is rare and I optimized for in my union patch +v of the patch +rb files gets a proper licensing header +but I think theres an issue with the linked hdfs jira expecting the wrong output so this shouldnt be committed until that issue is resolved +What is a repairThis is a separate feature that we provide that runs the AES on tombstones exclusively +Is it possible to expose the redirect flag in the next release of webflowMaksim +Leaving the issue open until the fix has been merged to +Priority minor because a workaround has been applied +Built and tested on both Windows XP and Linux +Add it to class enum and then update RECORDTYPEMAP in the zerigo driver +Closing. +wrt to this or the previous jira do we support the option of the client doing an INITIATE without the NEGOTIATE +Closed. +Thanks again Gwen +There is not any tag yet only so I can not rewrite tag to because that is apparently the demo which uses AG of version which is nonsense to propagate on site +Rod can you please try it out +Since indications are this is a container issue we can visit in the timeframe +But I thought it was small and simple enough to find its way here +Still missing a part of the snapshot repo +A No + +Thanks. +This feature is mostly about bringing atomic batch guarantees to set of of any size +There might be other issues +Are we using latest ASF master for the tests +Incorporated hemanths good +Also they would not be as easily configurable +I filed against Stargate for this but set the priority to trivial +Instead the current bean value is shownThe attached example uses the described workaround +Yes if you use the online CMS browser plugin you can navigate up few directories up to contentctakescss and could modify the css as you wish +what version r u using? Works fine with recent I think is still an issue with Trinidad and trinputFile components +The workaround does work but is very much a pain with a large object graph and inheritance involved +For example expose some of the metrics could mean different things to different individual +The problem is that if you wanted to do something like +Then the helper would strip the files lines and provide them as inputs to the service +Package should possibly be moved toSyncing with merging processes currently would be too complex because of the way handles deletes and updated documents in RAM sometimes before flushing +When press CtrlF for the second time it would show a in the combo of new window +To completely alleviate the bottleneck wecould allocate a group of keys for each thread doing inserts then they would only bottleneck when they had used up thegroup of keys + to trunk. +I am working on this +I am not sure whether we need to touch this we do close the connection properly but we dont necessarily close the if it fails +So in theory it should already work +Is anybody working on this issue? not actively +As I read it the spec gives no indication that modification to the super class should affect sub classes +Kids are good at generating such +I have tested HS and zk quite a bit via the jdbc driver and it does close connections to zk +That way Regex loader can work with any regular expression and then there wouldnt be a need for a different loader for every regex +Fixed in abeababeaeaea now uses JDK manifest class +This has been excluded please provide an additional patch for these files +Marking this report as invalidOleg +net performed +I ran test and findbugs on my machine it all worked fine +No changes have been done +Error is logged and process instances is aborted since it is not recoverable +It appears to be a namespace naming for the quick +My was minimal see below +If you forget Configuration on your test config class things appear to half work and its very difficult to debug what the hell is going on +There are some fixes in AS and above that made this issue disappear +Can you please take a look +The problem is quite the same for C projectsWhen using the SQALE plugin if test coverage is part of the requirements having a blank instead of is really annoying +Much of the code has been cleaned up and will continue to be cleaned as such situations are discovered +Applied patch works fine +Previous version left off a few required files +removed merged and clearer test output +and i am PC he is if you are french you cant write doc in Feb August and DecemberWorkaround apply for US citizenship. +Awaiting minimal patch for reported issue +Thanks +Anyway ive tested your latest patch with and and the problem is solved with my eclipse plugin and also with the supplied test the patch is not yet part of the i will patch the for our project and wait for the next version to join the stream a lot +Id like to see it working on trunk before applying changes to branch +Committed to trunk +You can test by running mvn clean package in the directory +Actually we should even reconsider whether the m relation between transaction mode and databases instances is justified +Sorry for the late reply but I have absolute no idea why it fails like this in your environmentMaybe it will be work better for you with the next release as the updatesite layout will change. +What settings are you using +re think you need to quit thinking in terms of JAR files you are right will doexcept I do not see how can I resolve the following even if my host slicesdices exports events api current felix event admin bundle still will not see will not use my host export +It seems to work out just fine at the least it turns out a lot better than just stripping the commentsWe should be doing the same thing here with the starting comments contain very helpful information for manual editing and users may add their own commentsIm not sure why we do any XML writing manually I have no problem replacing that +This likely indicates the source data is corrupt +I think its best to either ship with the simple runtime or the jul runtime to avoid users needing to do anything specific on the logging front to use Wink +The code will eventually be changed to fix the issue so if the solution is good enough today why not? It would give users more assurance that there are fewer open issues and they would be using what the code looks like eventually anyway +Committed revision . +So where do we sit? Is it still a problem or does it work as advertised? If it is an issue will anyone sign up to fix it for +It seems that this regression exists only in and do not exibit this behaviour +I ran the sample code works fine on +Uploading a fixed version of +If this issue has been resolved by can it please be closed +Great new feature with the protection scoping +thanks Charles is this commit going to be applied to branch also +Will dig into this further so I can understand whats going on +The patch is committed into the apache servicemix branch and it will be synced to the fuse branch. +Indeed there is no point having man pages in usrshare issue seems to be not valid anymore ls usrsharedocflumemanls cannot access usrsharedocflumeman No such file or directoryI would like to close it if there is no objection +This is indeed a bug with Jspc in tomcat +svn ciAdding downloadTransmitting file dataCommitted revision +Can theory things should appear in the archive pretty quickly after being sent +Really short patch +Yes I agree with you Jun +Focus is now PBR + +Wow youre having the same problem on Hibernate with a feature completely new to Hibernate? Sweet! +This issue is more than a year old with no comments or activity from development +We can have some splitmergecompaction action which can be used in the data ingestion tool which does the verification +remove the use of MHOME is test +bulk close of resolved issues as part of JIRA housekeeping for release. +patch committed to branch and trunk. +This issue is not resolved +to revert as this needs to be investigated more +runs fine +i dont have any strong feelings regarding absolute or notbut i do think it is important to have it consistent throughout thecode base in particular as it grows quite fast +David that is reasonable for development however this came out of the UX Workshop where the support guys need this info +kowsic thats another topic +Affects also version +There is a workaround use noreadline know it is unsatisfactory but it works anyway +Submits a very small job with maps and asserts that the job completes within s +Now that work is underway with AS all previous community releases are +You can implement from Spring MVC and inject it into a in your application context +The concern however is that it changes the APIwhich meansthat it will break clustering so we have to get cluster changestestingbefore committingjohn +So I thought that manual installation of these plugins are not needed anymore + +Nodes down for long periods of time would have a bigger impact on the nodes trying to send HH dataCan you clarify arent we sending the entire row as well? We might be doing paging but if I modify column and it has M columns M columns still get sent? I would agree to having it be tunable but I prefer consistency of performance over storage when it comes to IO +Can you please look into this? Thanks +Looks good to me +Again its technically easy to fix in but we just need to decide if should just support escaped terms instead of backwards compatibility stuffVersionetcPeople wouldnt be escaping wildcards and expecting them to still be treated as wildcards so IMO back compat has no value here +Thanks Todd. +reopening does not merge cleanly to trunk +Ok I was mistaken this is implemented +Thanks Tim +It would be nice also to be able to select logback instead of logj as the logging framework +Maybe the file should go into srcmainjavadoc but I could not make this work either +Verified on EAP ER +Wall clock sUsing default theme +I was excited to see the doc on this but then scratching my head why it didnt work +Good thought +I think we should split it right now than later +HaikalThanks for raising this issue +By calling super we get that behavior without duplicating code +After starting with an empty repository after doing mvn builds of spec and sdo the repository contains the following jars that include the constant javaxxmlstreamorgapachetuscanysdoorgcodehauswoodstoxstax +Im going to continue to work on it +Given that every thing gets recorded in it I decided to add all the entriesOn a separate note we should rethink our policy of adding every change to +Cool +Looking good go ahead and remember to resolve issues marking the fix for version so that the change shows up in the release changelog +When you generate a patch with git diff please use git diff so that patch applies with patch p command +Thanks for the reminder AveryAlso added some whitespace formatting for consistency +Available in production +I am convinced this bug is not a bug in eclipse or the tools +Fix the test failures +Lets do that as soon as is released +There is a Configurator interface in the API module and a Configurator class in the core moduleIf we used and Configurator it would be less confusing +There is overhead to this and likely not what the request is looking for +Invoke mvn jettyrun and click on both buttons sequentially +So theres very little gained by sacrificing correctness +Hi NaveenDid you solve the problem? It seems that this is a problem is in the Server side Can you please send us the AxisC log file so that we can have a lookSupun +The provided unit tests work fine with the latest trunk code +I agree with your comments about documentation being part of the wholeMy only problem here is ARE an unstable feature +Looks like an update needed for Hadoop that was missed + latest patch looks good to me + +Fix needs to be backported +I am seeing classes in the source code +Sorry small mistake in patch forget to option the last part. +QA Yes they didQA +I dont think that a good solution is to simply eliminate JSTL why not eliminate OGNL some will say +I have a plan to cleanup and by adopting +Im attaching and with changes to the followingM srcdevguideM srcdevguideM srcrefM srcrefM srcrefThe topic didnt seem like a logical place to put this information +Version should work with Derby +So you should find everything im talking about ve just seen the admin console fail on IE with and without compatability mode +I cant change the to fix it because i havent permissions to do that +defer all issues to +Id think if you need this it needs to at least require some configuration rather than being the defaultWould it make more sense to have a command that can be used by plugins and clients that need it +I am +Resolved as described resolved issues +The title attribute of the calendar image icon +Is this mean that cascade is only possible with ALLThanks +Submitted this fix to trunkSending javaengineorgapachederbyimplsqlcompileTransmitting file dataCommitted revision +So basically the default behaviour of the classes is to return naked values but the property accessor gets around it and still returns a wrapped attribute not breaking the code that relies on that feature +For many of these applications the DBA owner is APP +Ive attached a cut of my attempt to fix the problem based on the threading analysis in the previous comment +Im not able to reproduce this on master or +Thanks for taking an immediate interest in this! Im very excited to start using Groovy for my Spring MVC controllers but this is a show stopper for us as we inject way too many services into all of our controllers to want to explicitly define them per assume you posted the thread link as a reference and not as a solution +If constructor fails it checks if it caused by a disk +However I agree that we should improve our algorithm and Ive created a new issue for this. +Hi JosephWell I was able to generate client code in a folder just like before +forgot explanation of rule +bq +Sending srcjavaorgapachecommonslangenumSending srcjavaorgapachecommonslangenumsSending srctestorgapachecommonslangenumSending srctestorgapachecommonslangenumsTransmitting file dataCommitted revision +Yes that would work +ThanksIsaac +verified by Ivanov. +bulk defer of unresolved bugs to +In my current project I have data sources +Someone with magic powers please mark this as to be fixed in Solr +At this time it isnt intended to be threadsafe either but it may be easier to make threadsafe if folks are willing to pay the performance penalty for that protection +I changed the pickList css classes recently I will confirm whether this is a regression and fix or accordingly +Not sure why it worked for me the first timeAny input from the community on the appropriateness of this workaroundfix? +do you not see that as a real problem +Hi for the moment it will probably make sense to map xsdinteger to long given that xsdlong itself is a restriction of xsdintegerFYI requiredtrue will lead to long but dropping this flag will lead to LongI also added a mapping for xsddecimal to +GaryMy Bad please ignore the previous comment +Implementation based on RenaudI really like the idea of having a mutable datamodel +We need to add more test cases like that after M + +Marek thanks for the link +See please Window Preferences general it be closed +Will do +Looks like there are some unnnecessary formatting changes +Check your database and code this is most probably an error from your sideIf you already solved this issue please post a commentIf not please post details so that other usersdeveloppers can reproduce the problem + sets the on the request object +Resolved +My attempt at simplification works in one location but complicates at the other end +For simple scaling you can just pass in a differentwidthheight to the constructor +Please watch and vote for that issue instead. +Banged my head against the wall for an hour or so before I found this issue +Your changes break the build when compiled against any version of Hadoop that does not include the new security +This approach seem more secure +We are trying to fix this issue without relying on or fixing so this patch wont be too large or complicate +At least thats the issue that I had which caused me to write this JIRA +I already fixed it but the fix is not clean enough to make a PR yet. +this is not relevant in our case since its using neither autoack or dupsok its JTAThe JMS spec does specify what we should do on rollback of a local tx but says nothing about JTA +Hi Sande could you please file these issues as new issues? Seems that theyre not related with the README right +Logs are attached for both test runs on both is this issue seen recently +Itseesm like I sent a patch from the projects root folder rather than src +I will file another jira that is more suitable for and will make that patch available +Oh also removing Thrift would imply that no one is using it +Also note that this can happen when client terminates the connection rathen disconnect +So this is redundant +Commit is separate from IP Clearance and you cant commit until the clearance is accepted +I have run test from JIRA with TC and have located points of code where the problems related to issues have occurred + +Could you please if this works for you too? Thanks +On the way hand Id say well lets not rename the package then its not that important +These latter problems are easily fixed by editing the prepost scripts and rebuilding the rpmAny idea how I get the correct smackx and geronimo jars to build? Am not a maven expert by any means +However hang problem should be fixed for it as well +StackCan we commit this now to trunk +Yes I think we should do this via a symbol NET or some such +this already exists. +Changed the API base on Thorstens comment +Integrated to TRUNKThanks for the patch Jimmy +Ill provide an analisys and a patch for this one soon +Closed upon release of I misread the code in trunk +bundled to JBDS wia plugin still has duplicated entries in directory +Cant reproduce the bug using the latest CVS code +Thx for the patch Cdric. +I decided that I apply this patch toAs you said I think its good if we can set all streams nonblocking +committed this +Unit test were tested and worked well on my end +but I have review the source code of cxf component of +Hmmm it also works in my minimal test case +To be honest I didnt know about and I will study it further +Xerces has no logging facilities so Im not sure what youre referring to +Could you possibly provide a test case which demonstrates the problem? The full configuration and whether you are using an embedded server or some other external server would also be useful +provide a simple rack aware policy using hadoops network topology library to build rack topology choose the ensemble based on rack locality which ensure there is at least two racks per write quorum +We will know more in the morning and hopefully be able to close this issue +There is no HTTP client code +Sources for Profiler are not avaiable so finding out what exactly is needed is tricky +verified in +Thank you Nick! +sample file to use to reproduce the patch solves the reported issue +heres a prototype fix but we need tests +Need to be normalized for Lucene project conventions +This issue validates this functionality in the mentioned releaseIve created and Im closing this issue. +You do need JTA for any kind of distributed transaction that spans multiple for the time being since the current transaction synchronization is considered an underlying base mechanism that management can be built +bq +Bulk closing stale resolved have exactly the same situation as in description a which combines two tables by +Do you suggest to call from initialize? +Hi SergeyI used your patch as the basis of a new Crypto implementation that supports loading KeystoresTruststores from a null input stream +First theres the preemption which works as described in the issue jobs may preempt others if either they arent receiving their guaranteed share for some time or they are at below half their fair share and negative deficit for some time +Apologies! Ive removed my previous comments. +Nope I dont disagree +and both implement the Preloader implementation +uploading the patch with changes made to mapred script +This file defines the storeCustomer statement. +There might be books saying so +svn info please for your but mentioning his jira username above in the manner I did means he gets an email directly from Jira about this issue +Hi XiaoluThese features sound very interesting and useful +Both jars are from +I deliberately didnt put it under the section in those commits since I hadnt committed it to that branch +Great +Luke did you get a chance to finish reviewing this +Is there a URI where the old documentation is available? or does the used need to download it and then pass the location using D to the diff target +add some addition and leaving this issue open until we have a final solution +For each block it reads the uncompressed block length and then reads the chunk length and compressed chunk +While digging in the code for this issue I came across a problem in In case an exception is thrown the decompressor is not returned to the poolThe attached patch fixes this by enclosing the content in a try +bad jar +Ok I hate to do this again +It would have been nice to have all the elements in the same namespace but given that is not backwards compatible its not worth the headache +One thing I was wondering about should we add a shutdown hook to the launcher? In the past if someone does a we just exit and the framework doesnt shut down +Attaching patch for trunk +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +If there are still issues please attach an app that reproduces the problem against snapshots +If the version is specified this change is not needed +Please try or starting your application pass in the system property which will avoid weaving for this class +This type of thing is damaging to the whole community and its not okay +Next tasks +We are in the business of accepting patches from willing contributorsIf someone wants to see these features implemented Id suggested writing new patches +imo the mixins should work the same as other MV properties +Removes unused imports and added License headers and unused variablesStill bzip decompressing wikipedia dump +Previously lack of proper comment had made it difficult to understand the search processanurag +Is this a bug or or an improvement +Good point +So please an other one and attach your patchThanks. +Yes it should be pretty easy to grab the code from github and do an ant jar to generate the jarjs files +As Jochen points out this is working as designed +But the NPE is confusing +Container protection more methods +Thanks Alex! This is a big help +Thanks Myrna +encoded with an Integer size prefix or just a String type which is encoded with a Short size prefix +Im going to work on this for +I tried that and still have the same is the content of before clearing the settingsThu Jul CEST ApplicationsApplications this is the content after clearing and changing the runtime to my local maven installationThu Jul CEST ApplicationsApplications one is bugging me +All tests pass for me +It would be my opinion that we keep only the +I had the boolean return value of join backwards from my comments and thus wrong in the implementation +Could you post your as well? Ive seen this symptom before and each time it had to do with the getter and setter signatures +In that case trafficserver will open the sockets as needed +reset the retry counter in set so that it cannot overflow and to get again trials after is reached +The patch fixed the compile error on both HPUX and Solaris. +toForgot password Forgot usernameand changeIf using a public computer please Log Out and Close your browser when youve finished +lovely +The surprising effect is that it might work the other way round that is if you start initialization with the other bean involved in the circular reference as long as it is not a general recommendation is simple actually Avoid circular references completely +I was doing some blogging this weekend and came to the that just provides a simple low level API +Added an access entry in the reference +Do it +For example migrateVolume +committed into revision +We had added these before adding get +Then I add to the project build path Runtime and get the problem The project was not built since the source file could not be read +For those who dont know RBAC stands for Access Control +I believe this is the same issue +the patch includes both VM and JIT changesRana I marked more helpers as and allocation helper as interruptible +It is cryptic but I dont know if it can be avoided +Corrected one thing in Integration Test +The service implementation beans is used if is present on both +Please review this patch at your convenience +YonikThis sort of travels around full circle in the exchange of views between Hoss and RobFor the factory components I have to agree +Test added Contributor guide updated Added elements that Paul identified to the xsd Mailed to remind to update XSD +The class is documented and there are two small changes in the parsing of the query +But its definitely painful for people outside of +Committed revision in ve taken this version check out of the brn in r while we work out if we want such a ridgid restriction +Adds a page describing how to override the servers impulse to install a security manager +I assume this filter would also handle conversation context functionality as well as exceptions +Attached the patch Please for the fix Vinay! So for the new unit test I think maybe its better to put it in +Thanks Alexei the patch is fine +It is a large mb request but Ive snipped out the middle bitThis request is seconds before the stack trace +note that I dont think this would work in a validating XSD as spring uses ID to validate match up allows you to specify a name instead of an id so maybe we should default name to the name of the element? +The checkpoint code updated the source first and the local document was written but the response was too slow so it was taken as a timeout +you see this in and shows the methodmissing error you expect right? So something regressed between and +Ill look at the suggestion you made above +Can we limit it to this jira and resolve it to keep the discussion in one place +The other part that is also needed by the WC Incubator group and of course that would have helped identify the bug Reto just found is the creation of a client tester to test the endpoint +I just pushed this to trunk +Please review ! I reviewed the commit and it looks fine. +No these is currently no known workaround for this issue +From CXF use +This is in a local environment +Yeah sorry +Thanks for your confirmationI will go back to the developers of the other program and confirm what you have told mePlease feel free to close this call! I did not realise Java was not installed I just installed the product as requested on the understanding it was already thereApologies for any misunderstandingThanks for all your help. +leveraging composite operations +Anil since Jared has completely rebased this chapter into a Security Guide would it be all right to have this chapter refer to the new book rather than reinstating content that is repeated elsewhere? If so please close this issue as out of date +srcorgapachecommonsconfigurationimport +Trivial fixes to inputoutput of fencer test commands +The patch is too old to be applied at this pointWe should have a benchmark if the issue is still present today. +Good supplement Yi Xiao +Patch in trunk rev +to commit +Thanks for identifying the issue a merge to will follow tweaked the test +Attached and ignore my previous comment +Im trying the nightly build +Modified window chrome to make it look like a terminal window +Dhruba we have been testing the of Hadoop release quite heavily in the past couple of weeks as part of an evaluation we are doing +Ill have a look to see how complicated it is to implement +to +Can you rebase this to apply on branch? It does not apply a version of the patch rebased for +We could merge these into one definition by just using the port but I initially wanted to set alwaysCreate to true +Shalin whats left on this one anything I can help with + +What is the error message in logsArent you using user root +Once async maxThreads is upped the test starts to fail +I found the problem on linux as jboss +I just had to remove it but maybe you can upload the example without that dependency so that no quick fix is the jsf problem +fixed in revision +Thanks +A listens of various ZK paths registers ZK listeners verifies system state and state transitions +This can help when processing chunks of a file so we can get parts of a file in parallel and stitch em up at the destination +nethttp does not validate the certificate +Thanks again for your help +Do you mean the undocumented config variables +Now that has been applied Ill rework this patch so that it lives under the contrib dir +It could be considered for if there is another build +Probably +Also specific error messages for command line opetion errors are not example jruby C should producejruby invalid argument C must be followed by a directory expressionBut currently it only prints out generic help usage in progress +Of course I can exclude the import of that package assuming this class will never be used in a production systemBut then why including it in a production library in the first place +Thanks for update Jeremy +Most users will never notice itI dont have a preference about the product numbering +All node lists accounted for. +ThilinaIt will be better if you could create a test case in place of the service and client you have attached +So in this case there could be collisions +I guess Kabir went ahead and added that stuff while I was on vacation +It would be interesting to see the state of the ready selector keys when the selector returns after calling sendPing +changes to trunk at revision thanks for the patch Rex Ivan. +were correctly finding our Generator instead of yoursIf I require generator in Ruby it fails exactly the same way +Added to the SOA release notes as resolvedSupport and DB databases has been added to the SOA Platform database schema tool +Trying hadoopqa +patchAlso exclude list were unified exclude was addedNow the exclude files have following structureFailed tests failed over all platforms tests failed over specific platform testsIntermittently failed tests intermittently failed over all platforms tests intermittently failed over specific platform tests +Patch applied thanks and sorry for the deployed +I have a possibly related error complete with a test project that can also reproduce this error depending on when I start my two Modeshape instances +Adding the capability to pull blocks local on demand would shorten the period of high latency due to remote HDFS reads but it doesnt eliminate it +I probably need to go back to +How does it work? It looks like youll have the decorator location specified in the request parameters +Maybe that should be fixed in though +Here you Andrea +I forgot to mention in my case the data center only had one cluster underneath it so maybe when that scenario is hit just use the single cluster under the data center +Of course it would be nice to have a more liberal implementation but that could be added later if someone finds it useful and feels like implementing it +Should we use another term than tools here to avoid confusion +Marking closed. +The javadoc explicitely says that it is necessary +Proposed result file +Thanks for pointing that out Knut +The system property points to a directory where can load the property files for the custome mapping files and a additional file which contains a property +Carlo can you please have a look at this one too? Thanks +client Xcvsbuild outputclient Xcvs licbuildoutputclient Xcvsbuildoutput +keith +Test pass without any issues with and without persistence +Trying to +Got a simple fix Remove the hardcoded ljvm in two files +This patch includes the contents of Grants Nov patch which fixes the tmpdata path error +A symlink is conceptually treated as a file that contains a string target path +It was based off patch +Is your perforce repository on your machine or some other machineCan you please run the command again with the X flag som that you get a complete stack trace +implements getChild and get with snapshotsRemaining works get other meta data maintain modifications implements addChild and removeChild +The test failure is covered by +and there are no includes to any +A know about RooJavabean ITD +I tracked down which other test was causing the to fail with ibm it was +Please copy the license from bincqlsh before committing +Hello PascalI will follow both suggestions +it could also incorporate options for triggering the new options Yonik has been working on in as well +Unfortunately I cant get the NIO transport to work at all doing anything +I will keep you posted on this level of testing and keep careful watch on what the server does +Im going to pull them into our test repo first and let you look at and approve it before then loading into the main repoShould get this done in the next hours or so +that is strange +But some cannot find symbol errors occur in and classes +trunkCommitted revision +since the test method is an error the Test annotation attached to it is not resolved +Good suggestion +It works fine onSun JDKIBM JDKso Im reasonably sure its not the repo +I can commit it if youre willing to test itbq +I tested it for regular anchor links +marking fix version lets open a new issue for backporting +Fixed a typo +John Good to know but whats the motivation for this change? Was is it covered in the discussions you mentioned above? And is making a public API really a good idea +Until we get an active Windows contributor to be responsible for these its not worth worrying about +Please wait till we evaluate the new API +this patch updates the test to log the threadsstacks if an error occurrs during session est +Thanks Michael the fix looks solid. +Checked into HEAD will try to port over for for this at the subtasks it seems almost all the work is already done in both HEAD with a remaining subtasks for HEAD +Thanks. +Todd Bringing the discussion about this JIRA from to hereQuoting Todd This seems like an incompatible change probably a bad idea for a dotrelease of a dot release of a maintenance series +Thanks +Reopening to merge to to with revision . +If bug is found in maybe its time to roll back those changes until they can be made in a safer way +Patch to use to purge working directory +This has been implemented for the release. +Logged In YES useridI agree this is definitely a packaging issue related to theway the UCL works + +the thinking is that if the source file as a whole is valid then any fragments within it will also be valid +and to point folks to a better JDBC datasource with support for prepared statement pooling in futursam +A second reason faster build because a second JVM isnt launchedCan you explain the requirement to build it outside of maven? +They are in the range +Surelets do that +I have now resolved as well +What is this patch supposed to do? patch wont really apply since it is just showing where the binary file should be placed +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content Services +Those are just generic definitions of authentication and authorization which hold true for lots of software not just Derby +Secondary Tertiary sorting feature will add a great value to display tag library +I think thats a very good idea StefanId vote for your first optionNew New Entry New Context Entry New Search New Bookmark New Batch Operation +Test breaks test case errors for andUnit tests passed for s the status with this patch +However I dont know where to get the information about which test timedout +Would you be able to recode it in +Something included to generate more would be great too +Can this be closed? +A workaround for the issue is to force an access to the by the endpoints +Basically it is bleeding and changes from the session into the second level cache for other sessions to see. +same problem of handling complex arrays is addressed in the Jira +It doesnt happen on Java or the distribution of Java +Do you see any problems when just launching jboss server from command line with the same exact JDK +Also the classpath for the and the tasks +But I will put into QA Testing status so we can check for this. +I have made all the requested updates to the doc source +Is there something different with the package name compared to your others that are correct +It makes it easier to compile only those things that have changed +In the long term we need a proper solution but the current situation is that broker acls cant be used on WindowsThe long term solution is to tidy up the plugin interfaceqmf interface so it doesnt have this issue. +now only contains bundle activator and metatype information registering itself as a managed service factory for multiple proxy configs httclient received client connection manager tracker and configurable proxy authentication support for the transparent proxies is missing after your review maybe youd have some suggestions on how to properly implement that +orm and +Test is no longer failing +Commentsfeedback +It looks that for some reason port was not being closed +when navigating to another page and then returning to the original oneMaking the behavior for controllers configurable is a good idea in any case even if pages generated by those should never really be cached by a proxy or browser +Moving to longer applicable +The code is really pretty simple and shouldnt introduce a maintenance headache. +I think it would be good to do a sanity at some point to make sure nothing obvious was missed which is why I left it open and I think this has to be done manually but if you need to close it I guess that is OK too +Thats me +Added a test for multiple attempts +Totally agree that its silly for us to be formatting status messages in the internals +And hence the type setting rules for ? will be same as for a regular? parameter +Patch v supersedes v still applied to trunk +The main problem is calculating phi is less than free and doing so for each request to determine the least latent endpoint will be too expensive +These are tested with both and +Invalidation sounds like the best option thereIll try and review this more closely in the next week but Im not sure I like the filter as a configuration option as opposed to any of the ideas in the summary +Thanks Michael I havent had a chance to look at the code yet but I agree with your description of all the changes +This is the junit test that reproduces the issue +Thanks StefanIm not seeing the same behavior when I use your exact table definition so the index is probably a dead end +please have in mind that if you define a connection in sever admin then only settings done in server admin are synchronized to synchronize web admin settings you have to do a connection in web admin as well +I looked at the patch and everything looks OK to me +Moved from WW to XW since lives in xwork +There are a bunch of them +Or if I can turn off the password altogether +So either were actually writing empty row keys or were seeking into the file at a strange offsetSince Leonid has bisected it to the global cache patch my vote would be for the new key cache returning a bogus offset yielding to seeking in a wrong position +I guess we missed converting this to use read lock as part of +Resolved in +Its also about the size of the numbers each can adding numbers without infinite precisions already you should accomodate so they can hold +do you want to provide a patch +Additionally I can only really write a test that mocks the module item +attached the Rakesh +Committed branch and trunkThis tool of Lars is a nice little addition to our debugging toolbelt. +I make a new patch to fix the bug however its a little slow to compile the project I will run the test locally and given the test result +So we only have Camel specific features in Camel features files. +When moving things around the source tree its best to use svn move or svn copy so that a resulting svn diff will only point out that the file was moved instead of including a full copy of the entire fileYou dont seem to have selected the Grant license to ASF for inclusion in ASF works option when uploading the patch +Barring objections I will commit this tomorrow +Closed at user request. +change Show View into Show View. +Thanks for the sweet patch Liyin +I wanted this from the very first s annoying having a separate open with maven +So cases like mr +bq +reopening to make as properly in successfully and added +I check it one once the move to git has happened +You can already find older and newer releases of this package in rhinorhino and rhinojs +timestamps are unreliable without synced clocks and would require server overhead +I am surprised this doesnt fail in the java build eitherJSON makes no promises about the order of the attributes in the object so comparing JSON strings is unreliable as the order of the attributes may be different but in all actuallity the objects are the same +let me take another look. +i am working on a future release with castor which is blocked by this issue +This should be working since the poller is a scheduled task +I hope it was the only one I introduced in the codebaseNevertheless the refactoring was needed to get the codebase into a cleaner state for future extensionsAgain sorry for this. +The number of places the code changes is very high relative to the potential value +Added additional checks when sending a message +Patch applied to SVN +Hopefully some otherscan comment on specific junit fixes that have helped in the pastCould you describe at high level what one of the test fixtures that is failing is doing +Could you please confirm that this issue was addressed in as well? FYI the stack trace in my issue is the same as the stack trace posted to address this issueThanks still has the bug +Looks like Im using the wrong mapping file +Still waiting on designer team for an impact assessment and also need to update the system vdb datatypes index but that seems like Ill need to use an older designer and modify the types xsdThis batch also makes another attempt at consolidating type code +Also closing the resultset when the number of rows in set is reached +Small changes in methods +Excellent +I will have a look but it will probably require some other changes since it isnt currently possible to use a custom for web access. +I will cut the branch from master on that day +Or may be we should try and put letters T and F at top and bottom corners of the If component +as for the open files sysctl reports which seems pretty high and ulimit is unlimited +I have other errors but they seem to be permissions +For example my classpath definition changed fromCworkspacecommonshared +it really was fixed +Issue fixed included in the iPOJO release. +As per the bug found by Mike when he entered a dot in portlet insttance name can you please create a test case for it +Patch for with svn revision +The test was invalid +Added code in the local tx demarcation strategy so that it would roll back the trasnsacted session +We need to have a comprehesive test case for jaxws that is equivalent to the test case +Fixed + +Same version without the extra spaces +Thanks Konstantin! +example code +This isnt a problem nor does the statement exist in . +Please let me know if there are any objections to not depending on a particular logging toolkit +But then we may also want to map headers from MQTT to Camel Message Headers +I dont agree with the intent of this issue +Rebased for failed tests +Moving items not being worked on afaik out of M Fix this was fixed at some point. + +I just dropped a region table and everything worked fine +Not sure if thats the right approach though +Added patch here for easy access. +I committed the patch to +I think most tools that let you upload directories end up using flash or some other hack unfortunately at least when its a GUI selector +The question is if this new functionality should be added to the mapping fileor if it should be possible to specify the sort order in the OQL expression +sourcelabs +Resolving as the seems to be working ok. +I think we might able to skip that step becausea a must be really rare +Thanks Harsh. +I tend to agree that the repository should not start if a parameter is wrong +Sorry but this patch no longer applies to trunk +Heres the deal +yes embarassing i know +defer to +You forgot the test usage but I fixed that up for youCommitting now +is it not more symmetric that you can deploy the files that already are know to be deployable in Run on server than having to go and select each individual one +I messed up configuration hereWorking on fix for this +Its obvious +Throttling protections and such however can be added using authenticate module eventually or any other system +Heres a patch for it +I think its a good idea to keep the tests empty any test you put there will be cruft the user will most likely want to remove anyway +Ill roll a new RC and lets test that rather than individual patch +Reconsidering the above proposal OBR functionality should be consolidated on the OBR page Currently we miss functionality to explicitly refresh a single bundleTherefore I take another route I replace the Update button by a Refresh button causing the bundle to be the Update button by a new Refresh button on the bundles page +Please try with QA build as I fixed this the message is sent through it was able to successfully aggregate the message +I think that is going to need to be grails Another thing can you remove the +does not apply cleanly to trunkand I cant find these deps in a m repo only m which does not have transitive depsIm trying to figure out where the new artifacts are published and will update the activemq modules when I find themIm keeping this issue opened so that once I find them I can make sure to get these excludes in there +Working copies are updated as text fields are edited +Im closing this issue as we dont have any way to reproduce it. +So encoding is not allowed in headers +I thought webwork had some built in support for this might be pretty the job for add mm projects +For columns this means the column is NULLABLE and for the element is NILLABLE +Cannot you simply add it to the coretests site? It worked before +You could file an enhancement issue for that and if feeling really generous you could even port the Java one over and contribute the code back to the project +to commit the following patch Thanks Laura +The original reason is because the examples are all sharing the same bpel context so I will update the examples so they are in different contexts and we will see if this solves the database exceptions +So in fact it shouldnt be Rejected but Fixed +Hmm +I still fail to see any benefit in this annotation +GillesI have attached a new version of the EM fitting code +Decoding should be done in also at the place where filename split is done to extract jobId +Since there is GAE archetype no longer released there is nothing to verify for now +Note also that Mike as part of should have been Note also that Mike as part of +Regression +There are few releases where the rows in sysdepends do not match with what the test expects them to +inlined filter +This screenshot shows the problems regarding fml files +Also moved to Struts things I could find are fixed +de and +esb file on the left side for the file similar to the one in the editor +The following builds passed TODO all these jobs should be able to run on linux group of Hudson slaves needs extensive modifications because of changes in svn repositoryDocsMetamer jobs will be started as soon as docsMetamer are tagged +I would guess without looking at the code that the same improvement should be applied to all parts of this plugin and not just esb +Even if it comes in next release Do have any idea if validator is backward compatible with older version of the Java +Updated issue to reflect documentation is required +This work is completed with commit ceaaaedaeffedddbe +Grails defaults to index if no defaultAction is defined reverse mappings must do the same +Version with the small changes that Ted found +Ill check this in today +closing stale resolved issues +Reesh would you like to mentor Ross? Any OODT PMC member can apply to be a mentor? Would be great +Also in an attempt to minimize the size of the patch I didnt check in the new generated thrift code +I suspect it would not work because the implementations would not be compatible with the JDK a good solution for this would be if Oracle updated their package to include the latest DOM! The latest version of Oracle XML Developers kit I can find is dated and this has DOM API versions similar to the one Ive been for your John! I wasnt expecting such a timely change you mentioned did get me past the warning messageFeb PM Obsolete content type encountered textjavascript +changed it so that mapreducejobs will only return the equivalent of a not a +Daniel you mentioned in mailing list to use an interceptor to turn off validation + result after search there is no HTTP server here this is just some folder at disk which contains other folder htdocs +This issue seems confusing Evanthika I think you got the issue text messed up and the request to Synapse has to be a POX and it is converted to a SOAP request to the serverIsnt it? If not I dont see an issue here of this issue is invalid but I have figured an relevant issue in processing faults in synapse and fixed that +Issue addresses thisNote that Im doing this on a stock Glassfish install Unzip glassfish copy over my jdbc driver set up the db pool wire it up to eclipse and deployIll try to put together a simple project exhibiting the behavior and attach to this +As a workaround you can try it out to set the width attribute over css with a given px value +The problem affected both and which is why you seeing the problem on +birttests +And what happens if you execute an HTTPS request via such proxyOleg +Can I use selenium for this test case +So in that case that connect seemed to succeed while NO logging was emitted by the thread running the +A sample project may help but if you needed a quicker solution I suggest turning on debug and seeing if axis is attempting to load Spring before your app has configured it +Is this resolved +Would you please attach here a full Roo script to reproduce the the file from heresimple setup workflow put file in srcmainjavacomexample roo perform setup roo persistence setup roo gwt setup mvn gwt see issue with either or latest code from Git +It should make a blank picture +I dont see anything in the dump eitherIf you can cleanse and send me your I may be able to see something. +Once a company takes a particular version upgrades for them is a very important issue many companies will adopt a more stable version +sure I will review and apply this. +I am ok with the change as long as none of the tests in ant test Dresolversinternal are failing due to +GIT Patch files to apply to trunk and the branches respectively +No answer. +Thanks Konstantin +It would probably be simple to add a check during compaction time of the time range of each file and if the max is expired just to wipe out that fileThats one optimization but only saves on the read of the file +Patch containing a project in slingscriptinggroovy which enables groovy support +Patch checked in +As requested by originator + and committed +On the basis that the collections will be small and there wont be many updates I think this is ok +I guess because I dont see any or reason for using other than some kind of testingIf an index is truly then a real lock factory wont have any overheadBut a real lock factory is a nice backup in the event that the configuration is incorrect and it becomes writable +I think we can track further integration in other issues but since I see you put an entry in were done here right? GREAT WORK +Currently the easiest way to do that is to build the distribution war file +Fixed in Committed revision . +Well look at this post +is about avoiding to clean multiple intraday snapshots but has nothing to do with keeping file level data +Hi LarsI took a quick look and it seemed good +As a first step the methods should be deprecated in a releaseAnother alternative in case we plan to deprecate in favor of would be to move the code into pdfbox so wed neither need the dependency nor break API compatibility +Updated the method to allow for USER query modification and to indicate whether CACHED or PREPARED commands should be instead handled as new requests +Reopen if theres still a problem with more details. +I am going to commit it and update the related documentation in the Solr Wiki. +Hi Apurv this is great newsI suggest that if you have not already done so take a look at +Ive created user without administrative rights to test it under XP SP while SP download is in progress +Updated Awdesh Sumit Done at rAshish +Patch that removes async +I ran tests for HDFS and inculding contrib tests for both projects +The samples in the docs were tried before adding them to the docs +The DTP for this pass the plan around as a String and I need to provide at minimum a parser to convert the string to a tree of nodes for the view they supply +ok that should be finethanks +See for details. +Patch to which removes the premature call to +Quickstart ve changed to not touch full UrlsThere might be other cases still failing with pointing to resources on the same server but outside of WicketBut the supplied quickstart works fine now +With security disabled there really is no such thing as contents the user should not have access to when all it takes is setting an env. +Ok but you cant just look at the hierarchy of the class to know if mapping it as a hierarchy is correctand just mapping it blindly for all would result in very if we do this we should at least have it disabled by default and maybe have it as a checkbox onthe first page saying Add for base classes which if checked would do even then we have to be careful since would have no corresponding good best guess is that you only add it to abstract super classes to avoid problems +It seems like this was fixed somehow +We are still on hadoop and we have already lauched trunk into production +ok deployed as new snpahot version Odin Standal added a comment i reopen this issue to take account his new parameter keep will work at querying data and no at storing data as Odin Standals proposition and version numbers on the historical chart +Cancelled patch for fixing Jenkins warningsThe patch mostly looks good except I just cannot locate the source for +Please also add a test case for instantiating the Connection class with a timeout argument +I like the name choices they read nicely are easy to understand and match what actually happens +Marking now that dependencies have been checked in +It seems like the consensus is to not have the cluster logging going to a separate file + +When you look at Apple products or Ubuntu I think the opposite is true +JacquesIndeed! I read what you wrote there and looked at the examples you providedAnd I still dont get your point +A build must not depends on the execution directory +Interesting this doesnt work for me +If it helps doing an ant main and packaging that up for use as a SDK did not give an error with compiling an application using +changed priority to minorchanged assignee to Peter De checked in +Since the FP ships with Seam whereas the EAP CP ships with this fix requires inclusion in the FP release notes as well however it should be stated that the FP fix is for the different version of Seam so there is no confusion for the customer when reading both the CP and FP release notes +I think that billing accounts are working very well now and your made their logic more simple and functional +Committed that includes Marks changes to tez branch. +Small correction the last line of my previous post should have been trap echo WARNING has received a SIGPIPE output may no longer be associated with a terminal +Whats wrongEnter authorization information for Web API at Username lukaszlenartPassword Error creating review request You are not logged thanks TonyOne more thing just to keep it clear could you rename group to struts and repository to struts? It can be confusing +Amos can you take a look at this issue +It was my wrong assumptionI will fix the injection to use proxies. +Well revisit this in RC +we wont need anyway if we zap the license check stuff +Implemented iyappans comment +Uploaded patch At revision +You could also try using the provided by +Module changes happen quite often as class library is now in active development and bugs are fixed often +Will be nice to see more benchmarks as people try this out +Note that the removal of flushaftermins in is a problem for this patch +Ive just committed this to trunk +Tokens are intended to take the place of kerberos auth within a task +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Committed this to trunk and +resolves this +Instead a new resultCache should only get created each time a new permutation is attemptedThis updated patch simply removes line of code that clears the resultCache map +Patch merged with current trunk +Pushing to when well have more time to get this right +Applied the changes and did spell checking +We dont need a build of under java to integrate ejb +Craig what do you think about the following folders shaletoolstrunktldclaycfg shalemaventrunktldclaycfgmojo shalemaventrunktldclaycfgMakes sense to me +Verified in +Closing this +Hiroshi Thank you for digging in here +Here is a new patch +bq +Both the new thrift file and the test case are called anticipating more stuff can be packed there in the future +I dont know if Maven supports this sort of compilation activity +Yes but the iiop classes are inside so there is no way to deploy ejb without it right? Do we need to move these classes to another jar? Or is there a way to make them neutral +What about a wicketloop wicketif wicketelse? It basically starts putting logic into the markup through the back door and degrades the page object to something even less expressive than a Struts backing bean +Ah but we do in at least one case which is where the misunderstanding came from +Maybe we should rename this saysI am in the middle of rewriting how the gatewayslisteners areinitialisedtracked etc +See for further details. +Yah youre right not a proper random sample +Polish patch to meet requirement +All the config info is in varcatalina +HiI got the working at runtime +Update for latest the patch Martins second example still fails for me on trunk ve updated the patches supplied by pbrant for both trunk and but Im not familiar enough with the jruby internals to be sure about what work if any needs to be done to backport the changes for the trunk and patches were built and tested against r on ubuntu wJDK +It sounds like under I can do exactly what I want +Nope looks like in a CLI script it would work +Version logback addresses this or at least related issue +Ive made some of changes mentioned before with version +Im not married to the idea of incorporating it just seemed like an expedient solution +For now we have to keep them atomicAs a side note Why is contribmiscs missing in trunk +Attachment has been added with description Test FO to has been added with description Proposed feedback received during the last few days so I assume patch is ok +Patch to update the Archiva parent with the Apache Felix Maven bundle plugin +Bundlor M has removed the directive +The existing should also be refactored to make use of common code which has been extracted into +Sounds like a joke to me but I will do my best +The implementation is basically completed so you can expect this as part of the release +Just checked in the fix it is available from SVN and should be a part of nightly build would be nice if you could confirm it +I am not so sure about this +Logged In YES useridI dont know if Suns Specs say anything about this behavior but this change is probably for safety +The activation keys work fineOnly in the parent child scenario it does not goal is to have all profiles defined in the parent pom and every child pom inherits this dont want to put all properties in every pom this is the reason for existence of the parent can confirm this problem still exists in Maven +Thanks Andrey the patch was applied at the revision +In the future we could use NIO for file based caches but this would require upgrade to Java +to commit +Doug Cutting Sep AM this jira is not supposed to be a replacement for Thats unfortunate +Ive committed this +It breaks existing users and implementations just to have a different name +Over time we are improving these positions +For me it would have been the +the same issue can be found in as this is usually not related to Spring Data JPA at all +I am OK with all s in Empty stats +Depends on fixes for and +Please help me +Works well Julien but could you provide the SQL requests executed to display the content of the Ghosts page +I dont have a strong preference for how we present the modes +But again triggering rare case is the goal of our project +Thx a million mate +I understand the limitations and appreciate that all of you are busy doing different things thats fineThis ticket will stay here as a reminder and well be ready to pick it up later +This works fine with. +no need to log full path in many contexts +Now uses multiline text area +And I am trying to duplicate the issue +is a new file according to my source control system that was just created +Language feature designed in +Broke things out into separate tests and problems went away +I dont know the reason but I know it is possible to use Resteasy with Jackson provider and scan set to true and have it working. +rebased and updated to never drop INTERNALRESPONSE doesnt look like actually made it in to trunk so this doesnt compile +Im running the tests in eclipse +Incremental patch against which eliminates the security info classes and replaces them with a single class with static registry +The fixes probably wasnt included in +What did you use to build the text inputs in pure Flash? Feathers +Ill follow your suggestion as to where to apply and try the patchIll be eagerly waiting for the updated trunkRegardsTracy +LarsI implemented and not +Committed to trunk +The original intent was to avoid having to have and annotations everywhere and assume those by default +trunk Now supports reading the system property which defines the factory to all resolved tickets from or older +For all the other packages they putthe file there along with a docs directory +Is it mandatory to have facet depend on expressions? Cant we do it under demo only and use mock for testing facet +fixed ant I have just had a look at your problem +I give it another go let us know if we can help you +Please review and apply the patch +inFlasher has caused crash of eclipse for some platforms +I only want to enable SSL for Thrift and make it possible to add any functionsI only added two SSL related classes and +I will look into this issue closer next week Once I can reproduce it I will write up a custom patch +Latest version of local spatial lucene with LGPL dependencies removedand working unit tests + +Closed after release. +Deleting that download from my local repo and MavenAnt did the trick. +I see +DaveYou can attach a patch via More Actions Attach Files. +scheduled for that the problem is not in the use of table id inside table tag +jar you can use instead of waiting for Hudson to build a new SNAPSHOT at Apache will mark this as resolved for now in case Hadrian decides to build a RC of Camel as we want it to be released ASAP as SMX is waiting for it as well +But even if we only had moved they would have to change their code as the import is from a different package + +SLFJ really isnt need to allow the Avalon Logger to bridge to Logj or +you should remove the pid files +glup +Itd be neat to see it done but its not a blocker for anything afaik. +Although there is no such plugin at the moment there is an that shows you all users and groups is that something usable +Bill if I understand caching mechanism it uses weak references and thats why it never actually but in our case we see Full GC that takes moren that seconds and runs every seconds +Started jenkins validation again +Added Dejans tutorial with an additional sentence for the existing JSON driver. +bq +I just verified a use case and this is indeed happening and causing a bugI also think the pattern of billing on and Invoice is probably old and may be better to be replaced by a billing in +Patches should always be against head but I can figure this one out +That seems to have resolved my problem +The patch incorporates previous patches and adds few new TODO comments +We need to focus on the high priority items now +First attempt at integrating Hibernate Search today resulted in CPU consumption and a complete hang of the appserver +bulk defer of issues to +Forgot to mention the file is committed in SVN rev +Then use those views in your queries +Thanks Dmitry Alexey the problem is now resolvedClosing the issue. +I dont think they will e an issue with changing the license of those classes but Ill ask and yes the plugin works for for the patch I have an example set of projects I use for testing the +it was a little wrangle we get together discuss and seems we find solution +at compile timeand then when the syntax of the SQLXML operator has been successfully parsed the XML query expression is itself compiled + +Trying against Hudson +Can you describe use cases more clearly +Patch attached + added a interface +In the Pagenavigation filter we check only page reference of a node whether exists or not regardless of the descendant dont have a hierarchical structure so I think this JIRA is not well qualifiedArent you talking about navigation node instead +I dont plan to work on this issue in foreseeable future so I leave it unassigned. +Trunk has warnings +Actually come to think of it we will never need to distinguish for the purposes of autocompletion but only for the purpose of validation +The way its currently written the will defeat all caching +disregard that +Issue is related to display but test case is not this one there is some other thing that breaks the finally managed to find how issue showing finally found how issue occurs If I add skinClass to then issue occursFirst go to text input without displayInternational keyboard occurs Go to text input with displaygo back to text input without displayInternational keyboard does not show anymoreThis issue does not occur if skinClass is not set +PR createdNote the change to the viewport for a portal will not be seen right away since it is cached +This is untested code sorryAnd it does not compile eitherCan you revise your patch +structures +Attaching file that reproduces the bug +The test required some modifications in order to be run on LinuxiaUpdated works for both Linuxiagcc and WindowsiamsvcOn linux the test can be run as follows cd make export LDLIBRARYPATH +Not sure what everyone else is thinking though or if its even remotely possible that this code will be useful to us at some point later +I am using FOP in SCO unix and simply using the command lp to send it +Testcases checked into and trunk with code updates to use to determine the expected test result behaviors +Ill go ahead and open a new jira for thatMeanwhile given the benefits of assigning multiple tasks Id propose we go ahead and do a along the lines suggested by RunpingEric +Test also needed a few tweaks as well. +Fabio I modified the behavior as follows If the class in question is a compiled ruby script it will be loaded instantiated and executed If the class in question is not a compiled ruby script it will just be loadedSo you can use this to classload a given class but it doesnt actually bring it into your namespace +Modifying the SPI file does not work because the SPI parser will collect all parsers from all SPI files it found on classpath so removing is impossible +Feature works ok JBT L +reopeningcould you fix it in the trunk toothanks in trunk at revision +Feedback appreciated probably not ready for merging into the trunkThis patch implements and few of the ddl commands describeTable show and show +Hi GlennI tested with the latest nightly binary build of Apache Fop and indeed the issue does not appear anymoreSo you can probably consider it fixed +test was updated +In Maven it looks a lot simpler +Please verify that they have been applied as you expected +as defined by the repository +Thanks for the patch +We dont get any exception just heap dump after application undeploy +merged into hibernate tools core set Tests run Failures Errors Skipped Time elapsed secVerified +It needs to be refactoredreapplied in the new generics modified code +for +No response so marking this as Release Note Not Required +See my last comment +and have been fixed +My revised proposal +No additional comments were received so I am taking that as general agreementJust a bit more historical background on these changes in case we have to revisit this in the future +Interesting observation +Strong Pull request is on the way but should I open a new issue first? +jsftests +fixed now in trunk. + +I think using term hack makes it sound like its bad +Another thread owns the latch sees interrupt and tries to reopen the container and in that process need to call hence deadlock +HiIve committed this patch down since it looked good +Hi KelvinPlease ignore the codetext included in this issue and only consider the content of the attached patch file +Any chance you could try with a newer version of CXF? This may already be fixed with as I know a bunch of fixes have been done to the attachment things since +Yes I believe the IT is the one attached here or its original variant attached in Red Hat Bugzilla. +Anthony Ive added you to the watchlist for comments +Fixes issues with the security and transaction policy pages as well as an issue found with the security configuration functionality on the domain page of the editor +And we have to use Deprecated at the method starting +yes youre right its a good idea to use a JVM think a good name could be there currently a workaround? its quite annoying not having reports on my Linux install +Attached revised patch for this revised the changes as a map in that will contain the error messages during authentication +Updated exception handling during CND import so that a clear error message is raised if a namespace declaration is missing +run runner only once with all tests to avoid report in trunk with a fix expecting Java to raise exception if any test fails complete successfully otherwise +If that value is used it replaces it with localhost +Supported configs page updated with product launch on March . +bq +I did said on last comment that it works per documented +Please let me know if the problem still occurs with latest build +Fixed in r +I think that the javadoc should explain that you can get verbose output by setting true +The test attached +If you want to delete a document it would be nice to be able to send one request to both sets of shards and get all the required information with a single request instead of having to query each set individuallybq +see pauls comment at aug +Before this job would have dumped field b as chararray instead of failing at the middle at runtime +See for screenshots +For example theres no listing of files or directoriesThis component exchanges messages wrapping a that represents a File in the OFTP server +Ill look into committing the is another review package with more code comments and little reorganization of the code +Added regression test for this issue. +Thank youRegardsYour patch tested without any errorsRegards thats great hopefully I will be able to contribute more to the project +Thanks Sandy +My only suggestion is that you have the ESB team produce a template XML config per action that the tool can use a template for the +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Attaching doc +Hibernate is obtaining the information from the correct context classloader but this defers to the repository and therefore picks up the which is already just happens to be the drools one but could in reality be any resource exposed globally +may we know what it is +The session was not new the session has been created few minutes before the failure +Ive tried the patch but the test still failstestsapijavabeansbeancontext +I will open a subtask JIRA accordingly and let you know +Final patch +Thanks for your feedback Claus +thanks a lot Simon +Release Notes for will not be at this time +It is expensive to do that in here because the component may be interested in only one variable and we end up resolving all variables +yup +The build works fine +It is not pointing out specific errors in the code etc +The code is open source and nothing further should be I got what you mean +Well it rolled in just fine and appears to pass tests and specs as well as did before +applied pedros patch with a small bugfix +It could also be a bug where we are passing in the wrong alias to the check +Hi GavinSeen this issue and think i have cleared it +Seems the same problemIs there any solutionThx +For an example this class should be able to invoke with the http binding as wellSoap headers only defines at the binding level +I was having problems getting eclipse to recognize the slfj libraries the fix was to start eclipse with clean +Sorry for the inconvenience +Is any progress being made on this issueCameron Ross +Claude specifying a negative number is infinity currently +We agreed that that would be taken over after merging MR into trunk and branching +Include updated screenshotscaptures tools and commands +The Xmx switch can be used if needed in JVM args to specify a threshold that triggers garbage collection +Forgot to mention that you should also add the new tools modules as dependencies of s POM so they end up in the tools TAR ball + +Ive added a event category and ordered ticketSafari Mac OS X Lion on both master and the version deployed on a maybe related note I see no bot button under the monitor tab on. +Will investigate this more closely in the next version +Good luck +Updated to catch the at shutdown +Fixed under The bundleplugin now installs OBR metadata as part of the install phase for additional OBR functionality use the +have no cache +I believe there are still quite a few unvalidated cases + +The is never generated? Is that stillWhen would this be invoked? Only if IW is pooling readers? Maybe weshould hold off on that for a separate issueWhy were the added checks needed in? Oh I see itsbecause you compute the size of the merged segment before themerge completes +This simply replaces the N by true which makes sense for a search Form +zip then new project will be immediately consistent +I think the first phase of shuffle might also gain to some extent +Patch looks good apart from one tailing space areaSo fixed that and applied thanks +Issue can be closed as far as Im concerned +Thanks to Alex for the patch and to those that commented! +There is one thing +for the code it is a good improvement anyway +My sql is a remote serverWhen git pull all these fixand reinstall cloudstack manager server +Is the suggested fix safe? Would it be possible to get this JIRA reconsidered? +Can you please have a look +Right looks like the workspace specific settings stuff is causing this +Need to see the Duplicate issue. +Hopefully we do see this much concurrency on the block cache +Trend Micro on machine caused the problem +I think it is reasonable to suspect this is a bug in the Hibernate mapping for the Derby dialect +Was that for two separate guides or for creating a part and part in the guvnor guide? There was some email discussion about this +Note that this is only for the new interface +Are you going to create a separate issue for the compiler bug +The algo I was running sends the same of messages every superstep it looked frozen at superstep +I though the new website was built and pushed into production but it turns out it is not the case +It will be better to move included tests to these classes instead of creating multiple classesId much rather keep the unit tests for in the same package as the code that way when the code is updated Jenkins will run the tests to catch errors +I agree with Jukka in that we should get rid of all config files that dont have a sound reason to stay where they are +Closed +This sets you up for sudo access +The code for this enhancement is checked in and has been tested in production for several weeks without issue. +You are confusing chunked transfer encoding withHTTP pipelining +Now if the return is not received within days then this replacement order will be automatically canceled by scheduled service auto +Hi Felix thanks for posting the code +Thanks +patch you for supplying the patchApplied patch to CVS +Im closing this issue now but if Ill find other deprecated operations Ill create a new issueAgain thanks. +Thanks +Keep up the good work +Thank youBest regardsAnton. +His name is Yajl +Need to close this defectthanks to since is now released +Preconditions and results are as described above except I am using artifactory +Okay I will give version a try +Curious +Yes I like this idea can follow this in +So the bug does not have an effect on the dfs block allocation +Update to incorporate the classname changes +I committed +for hudson +If you truly feel that these log messages should be why not get rid of logj entirely and use +Look through all chapter and sections add proper keywords to them +If you suppose i need to use schemaLocation specified on the xsimport please explain this +If someone had the classname in a comment it shouldnt be replaced for exampleOtherwise Im a big fan of being able to shade avro when necessary +Patch with workaround Bricon mentioned the high priority break other things +Closing tickets +We are skipping phabricator +Just pushed a commit that added the new option to the model in the this allows users of the XML DSL to configure it as wellThat is a rule of thumb on the data format +Agree +This bug is specific to DSE and is not present in Apache Cassandra +My solution is consistent with how handles crc option now +Doing it properly is a minor problemYou might also want to get some practice in shorter development cycle patches +Since there is no active connection thus NPEI verified the error with and fixed for the next release +You should consider Cloning the issue for each stream if a fix is required for and streams +It was probably an attack and I set values to those which should be approx +Add to the +But have discover another problem as it in +Martin please confirm +The portlet uses the default embedded Apache DS configuration +We now have buildbot coverage on Ubuntu Solaris and for C and Ubuntu for Python +Im planning to do some Selenium dev with Mike W but Ive managed to wipe my private key so I need to upload a new public key +just use the first Note that the kml output format sets the filename to be just the first when you request multiple layers +We are trying to implement a standard +Or even keep a round robin buffer of the last such operations? It would be a miniscule amount of RAM to track this right +Updated patch that includes the updates to identified by Dennis Kubes + +fix checked in under svn trunk r r r +Updated the test file and committed the patchthanks shravan +Added into thanks again +Fixed and deployed to snapshot repository +The jar deployed to central is mostly empty it contains nothing but +Patch for branchx +Mapred tests including contrib tests passed except for two timeouts and but these are failing even without the patch +Patch attached againstIt is basically rewrite of but I moved that static method to and renamed as get because I felt that is more suitable place +Waiting for javadoc release before making this final +is not final +Hey forgot to mention that +Hi Jacquesfor the rendering there is no problem as it is a ftl macro which converts the docbook code +Here is a sample +No exceptions in the log in. +Is this still an issue using +Fixed +Restart Eclipse with a new workspace +You can give it a try if you rebuild AS trunk +Closing +Ive just committed this +Intel still expect the stores to be aligned +JaikiranThis being flagged as a regression could you please look at fixing it for the next release +Does this need to be in every log output? Or just do it once? When we switch to multiplexer is this still necessary +I think that creeps the scope enough that we should push this to especially since Yuki is concurrently working on a new Streaming design. +Assigned. +I have not run it since +You additionally need PGP signatures so we can validate the origin +Thanks Jason +This patch also adds generics parameters as far as no other code is affected +The new patch proposal adds the master file update and issues svn del on the master file rather than updating itIll mark this issue Patch Available because I think its ready to go now +Support added for tag classes and DTD versions of the TLD files +Merged accidental commit from import branch +That doesnt mean we stop proposing new projects the ASF is about building communities +There is most certainly an encoding issue +Here is the patch without refactoring +Zero +As tries to map an id to a class but that mapping is not read yetSorry if the description is not very clear +JAXP +No response from nevermind I say close hpp +nd patch applied at +Ill fix that asap +There is no way for to know whether there is portable IP range at region levelSo cant change UI like you described in this bug +Both the errors expressed by the validators and the BPEL archive issue are caused by the creation of a +Cannot reproduce +And that why annotation exists +Thanks Vinod! +Different reviewers of this document will do the spell check differently according to the default language setting of the that readers useIf we set default language setting at least we can make sure different reviewers of a same document will do the spell check same though default language setting may be incorrect according to users requirements +Trying to extract real test case still +Unsetting Fix Version for unassigned issues +Ping! +If you still have problem please upload the log as well as the config fileHoward +This is probably an issue to report to yahoo though no +That was quick thanks +with JBT. +JonIn the scenario you described above do snapshot and restore happen on the same table in the same clusterIf so would you be able to utilize the following to control concurrency? ZK based ReadWrite locks for table for clarification JonTo my understanding the restore action normally happens after a crash +appContentsIf Java Preferences configured likeUsing one of those two option it is possible to force eclipse to use Java or +Works now +Dear Mister KingI do know the patch isnt perfect +If its as simple as it doesnt work well then Im not able to replicate +METHOD +It now shows up an error in the test +I also have moved this to Common +FRAMESET is an illegal HTML element in the Trinidad supported doctypes +The updates have been verified in the ER build +Fixed a lot of bugs added a lot of comments tested it with a zooKeeper cluster of nodesselect currently performs a dirty read we can add a new parameter to change that behavior if need be +This will help reduce a lot of page config weight on my end +This should be done needs testing though especially with portable installer. +Ive just realized that the plugin doesnt work as expected when put in the aggregator pom +Still deciding which entries to evict could be expensive +Please apply the fix to the +I dont know whether the connection pools from Torque could interfer with othersCould you give a few more details Can you reproduce it every time? Are you able to successfully login after this? is there something interesting in the streamdump? which client? Does it happen after a bigger number of connection attempts? Imap clients open often additional connections +In the code there is a test case which uses this configuration and the consumer receives its messages without issue +If that is the case why cant we create the file with overwrite flag as true +Resolutions other than done should not have a fix version set. +push out to +Updates on this? Could this be there in version +I remember changing all the user limit tests and test to accomodate thisOther than this I verified other changes are fine +This is certainly a bug in the C multi code I wrote so im taking this bug +Could you clarify what you proposeSave char strings with stack information in C heapCreates java Strings without constructor runingTurn off lazy stack creation +So i do not really thing we should just do it to be like hadoop it it makes peoples life harderIf anything please keep it as +I think we should stay focused on the issue of credentials management and not veer off into the management of authorization ids +It seems that Ive managed to fix the bugs +I tried it on and tip of tree they seem to work the same at a certain point you cant zoom anymore and it rebounds +Im not aware of any reason to distinguish between the two +Patch appliedThanks! +So I am inclined to close this with Wont Fix +The other place I could not set it using Spring were the connections that were created through the +However this issue needs to be fixed mainly in RF +I rewritten the issue a little +The only major refactorings that could happen would require too big of changes for so Im moving to +Can you svn blame the code in this area and see who the most appropriate reviewer is +my second patch still seems to have some flaws +I basically discussed with Emmauel to rename this issue +ASF License the patch is on the file itself again will grnerate from root next time +snapshot present is true in the config Force the snapshot cleaners to be in and enable the feature is false in the config The feature is disabled but if cleaners are not enabled and therere snapshot in the +This patch does rely on from lucene contrib +Thank you Arpit +You have most of this logic already just dont need to write the other information +Ive put an addendum there +Stack trace of hung by Gordon and Andrew +Since your patch only specified I had to go and hunt down the appropriate file to apply it to it turned out there are actually two files in classlib so I then had to inspect the files to know which one to apply the patch to +Updated to remove change +Actually I think it is simpler than that +I had the same thought as you not every context scanned the complete set entities +I just committed this +application is doing row delete and then row +FWIW the default java lib as it exists on branch doesnt seem to be running the test cases as part of the default target +teiidadminshelllib +Just wondering why is only half the array is memcpyed into pNewwhats the reason for zeroing half the array pObj? Why zeroing at all +attaching to allow to parse a string for the port number +Unit test that causes for the test case Peter and for the fix MarkI attached a patch that merges the test case into the existing merges in Marks fix fixes the whitespace in that test and adds a CHANGES entry +ChrisI see your point about the years +This would not be necessary anymore if this issue provides the ticket directly through the RPC layer +Sorry for my last post actually there is no change in SOAP Response it was same as you can see in +Thanks Colin +The bug seems no longer reproducible +You can use this project to get started +Incorporated Yongqiangs you explain what are the comments and changes +any updatesThe bug is still open does it mean the patch hasnt been applied yet +Should I specify the version somewhere in a configuration file +Please create another issue for this and assign it to me or Yahor +is failing intermittently +NET library +Printf is certainly not the most natural way to log things to a jsdeveloper +Initial testing results test failures known issue see known issue but no jira report yet? I opened a JIRA issue see I opened a JIRA issue see need feedback and or resolution on failures on or prior to releasing +Just wondering if Curator wouldve been a better choice here instead of writing our own? +Its me again +Hi Jason we are using this patch in its full capability for quite some time +Moves and from to +This already works +Do we care + +barry assigning to you for now +Thanks +I still see leak now in +Minimal patch that leaves the responsability to make the copy on the codec sideI created to address the documentation issue +It would be nice to document this in others looking at this ticket the command lines arejava cp true cp true props to specify a bind address if needed +Committed v patch including deprectation information. +bulk close of resolved issues. +moreover one minor compaction can handle a bunch of small meta data storefiles +I recommend that your patch rename row to row as you suggested and create row +Pretty please add the license to trunk +bq +we will need test it +I have rev as the latestIm still getting errors in applying the patch +I am attempting this with and getting the error Exception in thread Main Thread orgcodehausplexusutilcli +Attributes are written back to all descriptors in all regions of the table in META by the Master when shutdown is first initiated +nit sounds to me like it might make a substantial update to all the slave nodes +this also fixes with following differences Moved main to +No use in laboring and redefining for all that information again when it is already provided by JMX +RobertAre your comments applicable for version? The behavior of an segments and deleted files is very easily reproduced +I just dont think these use cases are necessarily core concerns +Patch +Ok here is the Patch commited and tests performed +No problemI have just tested itThis new patch works fine tooThanks +can override this though +I have a patch that fixes and is also compatible +Committed the javadoc fixes to the trunk as revision +yes they must be separate because the jail wont allow you to see anything now the builds run chrooted but I was told the user could escape the chroot jail in the chroot call should be followed with a su username with an user that has no privileges +Not enough time for maintenance +I take it back the issue is still there +The system should retry an attempt upon hardware failure or YARN error and set this default to . +The alternatives Im alluding to are the use of to do an efficient bounding box style filter instead of the tiling system +the server and client are some how out of sync in terms of the WSDLThe Axis documentation on how to build and the architecture overview were helpful in resolving the issue +for the patch +The attached diagram illustrates the current encoding flow for and two possible options to solve the problem +This bug is inaccurately filed as fixed in milestone its actually fixed in RC +Jeff Davis +But your approach seems reasonable to me too and has the advantage of stirring up less code so Nit it might be good to add a comment to the new that explains what its used for and makes it clear that the token cancellation comes after the deletion +The fix here is for +On the good side of things I cannot think of a single reason why this would be required +Im one of the undergrad researchers +Hi PavelSorry for the misprint Ive updated my patch +Cleaning up the code will rarely hurt +This is okay as we have a farm of CI servers which would occasionally fail and we frankly dont want the servers to fail from time to time due a tiny issue with some port number in use issue with +I can surely include start and end times and any other metadata for each nodeFirst cut +We can continue to check out their master branch Id imagine settling on a suitable stable SHA or I can make a point release that is with just this fix +I just checked in a patch to parse the query in the correct core +Whatever time you have to spend on this is much appreciated + +When a selector matches only key are required to match all of their fields +Can we resolve that one at first +The timer for SLA compliance should be from when the queue has load that isnt being satisfied +The issue with the latest patch is that you return one last client pipe before closing and throwing a transport exception +Please apply to release branch as well as the trunk +Believed fixed now +Sorry for opening issue Ive just sent an to given addressThanks for your support +Found issue by code inspection fixed and backported to all affectected branches +Is resolved in my view +The problem is not the display +Hi BryanYes it works with me perfectly +from me otherwise +Also I will try to modify the test to catch the case I fixed in hopes that it displays the problem locally +Thanks Andrew Well make enhancements accordingly +Temporarily reopening for release note status. +It will be boring but worth it +BTW Id encourage you to try JSF +Applied to branch and trunk +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +Ok +First crack at a patch just to get the idea started +Not testable +I think you are assuming that the string is a Unicode string or a string of Unicode code points +Hi Mladen do you have an idea what may be causing this +I think is now in builder and in impl +Incidentally Ive tried running this inside bash as well as on the normal command line and I get the same problems +Ill give it a try again sometime later this week +Similar issue in original further testing +If the source of low performance is the delay between polling requests then the observation holds only for low rates of changes +phew! +this was done when we found that forge snapshot runs fine within eclipse on windows +Ideally wed need to add something to the script so that it determines where the nutch command is located +This is the first patch to fix the test when run with jars on Java +I try to use default +This is the proper patch +Ive just committed this +I have spent several hours on it this evening but have failed to identify and resolve the issue +r +Also the API doco lists methods such as get under the heading Functions introduced in DOM Level + for thinking about where to place thisBut please note that just filling the injection points doesnt yet give you a fully working CDI beanThe filled bean will not have any proxies nor will its lifecycle be managed by the CDI container +Let me know if it works for you. +Ive just committed this +What we can do +looks good +On trunk the fix is easy dont make it a singleton have it be instantated every time +isnt this a duplicate of +Your understanding of the binding service behavior is not correct then +a workaround has been documented in rev will be fix better with using embeder when will be released +If the RM is indeed telling the NM to kill the container then we would have a race with tasks failing due to the notifying the AM before the AM sees the container status from the RMOh I didnt realize that +This issue is fixed within +Im just explaining why I prefer it +but the jar is actually named so why is this an issuebtw +So even if the fetcher queues are cleared before restarting the fetchers to fetch from the newly assigned partitions the consumer iterator would continue returning messages from the current data chunk it is iterating upon +sure I think we can update documentation in a branch a new chapter of the user manual would be great yep but default only the compactions are traced you can do a trace from the shell though +Hey Joydeep +Henning changed the build file a while ago +Weve never heard back from him on if he has something more complete +thanks Ill apply this logic to a few other multitermquery tests i am familiar with and upload a new patch +Ill add tag uri or tag name fields +This test demonstrates the issues Im talking about in this ticket +Ill close the issue then +should resolve the issue +Please note that I just submitted the patch I already did for myself and that I will maintain as long as necessary if no correction is given in future versions of HibernateBut as far as I dont know all the constraints Hibernate bears I suppose it would be better if an experienced people of Hibernate made the patchThen if this issue is unrelevant or if a better solution can be found please close it. +Its a shame that JSON Ruby etc are dates but not worth the effort to battle it at this point. +I guess it wont hurt to close this one and open an improvement ticket +This went through couple of iterations of review +bq +In fact the issue Tom has outlined for multiple driver inside same ear wasnt a problem also in old code + for patch +Thanks Doug +Minor changes to address Lukes recently comments +This needs to be looked into a little deeper +Its very rough and I tried to go the least intrusive way as possible so the implementation is mainly to show the main principlesfeedback welcome +Im currently on reverting this change in but Id like to finally nail the exact semantics for this negotiation once and for allAs I said in my first comment I think the only problem here is that we are ignoring the value of q +Oups made a mistake with license +Thanks for the fixregards Ivica +Thanks for applying the patch +I committed it to branch +Closing this issue as the duplicate name issue is more or less handled +Hari I still think you and Todd are talking about related but slightly different scenariosCan you confirm that both your were configured to be able to access both of these directories? If so I think that would explain the confusion hereCan you also confirm that the logs you posted in the first comment on this JIRA were all from the NN +No tests included +Thoughts +Closing as per Matt. +I have tracked down the cause of this problem the class caches the XML after it has been the debug logging is enabled the first time is called to generate the XML is when the classloader of the EJB is the current classloader this means it has not problems generating the XML +the patch not persist the bundle id in GEP new patch at revision . +on the latest patch +Using air with it +please verify of a bulk update all resolutions changed to done please review history to original resolution type +ThanksOn a related note please file a JIRA ticket before creating a review request and make sure to update the JIRA ticket with the location of the code review once its created +Thanks Rayma +Thanks +The tests succeeded with +Committed +Search tag elyogragspringclean +start HDFS and zookeeper init start Accumulo created a table and insert some data kill name node kill all accumulo processes restart namenode restart accumulo recovery succeeds + applied to rev +The numbers of javacore files depend on number of the same unauthorized shutdown error +Oops! That was a mistake in the patch +Note that therere two possible cases herea Reading from data put by a Hot Rod client +youre right +The DN puts these blocks to the head of block scanner to verify +Great stuff Ashwin! Im closing this issue now. +Here is the dependencyscmgitgithuxiYou have to build Sulky before you can build Lilith +Resolved with changed the model entry to the form described in the resolved issues +I think I could generate them fine with buildr +never mind the problem is that there must be a missing explicit dependency to wtp since the plugin requires it for its consider it broken because sensible installation choices lead to an unusable me configurationme and do get installed in the process as dependencies of IU +wonder if basically we should just make a goal of specifying what the allowed field names are +I think was accidentally excluded from the commitIndeed this file appears to have been missed +Thanks for quick look +Merged patches for Indexer and Crawler with v +Elliott Clark To what end your parameterizing? There are two master interfaces only and both inherit +Committed +Gordon has reviewed the c patchI agree with Rob about making it an option to not set the userid +Turns out as you say that probably isnt the caseTests with the changes in the latest patch are running just fine at least once +BND is asl licensed and as long as we have appropriate notices in svn and released artifacts we should be able to copy source code into asf svn +If you find further issues please create a new jira and link it to this one. +So I think SOA runtime should be better for ESB project examples since SOA runtime contains ESB runtime but the JBDS doesnt contains SOA runtime there is also a problem +cannot reproduce using a fresh machine with the specified JDK on Windows machines but could not reproduce this +The best it can do is to wait upto X seconds in the hope that things will work out and at X redistribute resources to meet the deficitAt the same time I can see that there is an argument that the AM knows best how to free up its resources +refreshed and also attached the new Namit running tests +Tools is just a tools +We need more discussion on whats the right syntax and semantics for the comma seperated path names + rightSeems to have been applied to trunk now +I was also thinking about abackslash but I found it inconvenient that in String literals the backslashitself has to be escaped +Whats the specific test case where we need this? Cant we work around it by making the test setup easier? Having dependencies like this is not a good idea as it complicates the build +as per IRC discussion path looks everyone +It cancannot happenBTW i donot understand why you are avoiding live to live delivery like Jboss messagingPlease let me know what you are proposing in future releases of which can address this problemThanks alot are avoiding failover +just committed this. +Yes I agree OGNL would answer my needs very well +This class probably ended up a little more standalone than needed it makes no assumptions +Thanks Doron +No please dont open any new issueAs Ive told you if you have questions please use the mailing list + +I removed the conflict and now its all can ignore this +Since this is a Jetty bug I suppose Ill have to hang fire until its been fixeddavid +Ashutosh I actually tried out and wasnt able to get it to work +Ok Ive added a new git mirror at crunchgit will mirror thisLet me know once you are happy with the mirror so I can delete the old github mirror of thx +Triaged for +Committed revision Ensures archetype built after jetspeed application dependencies +Proposed patch Enhance JSON output to support wrapping plain text data Support selection of JSON output method +There were all kinds of issues with the Nexus yesterday is this still a recurring problem +Pull request sent +Patch for review +I think when we init we should do a check and alert the user if they have an issue +Thanks Ashutosh! +I have operational experience with Git and reasonable but rusty python skillsIt is understood about the CMS +uploaded well see if it working on a repository short on time this month +pat the changes look good but again you have a change thats reflected in as well? This is regarding zoo pid file +Once you have it all ready as a module Ill create a space for it on and add you to the team for it. +OKI wait for to be released heartily +I believe this was solved in Tika +They simply are not shown on the WS Custom Form screen when you only have one Fact and one Field +This works on the and linux systems Ive triedI will attempt to try the trunk version also to see if it works +However this does not apply to unknown extension or xsl elements +works with tables against Derby fails against mysqlinnodb with foreign keys defined between tables thats the case Im hoping your patch would solveLet me know if you have trouble setting uprunning the tests +and lest I forget one more +Add new implemenation thatmakes use of the RETRYBACKOFF over in and that has a maximum amountof time to try +so maybe we could use there code in the framework as well +Updated copy of the Drools S Integration. +They both implement the Interface? I am not sure I follow +The default would be the current behavior +This would have to be synchronized somehow too + is fixed at Apache we just need to make sure it gets in the next ESB release +Fixed on trunk thanks for the patch +Sergey any update on this? +Which then makes the and password unneeded +Disabling the plugin by configuration is IMHO not required because the Web Console is targeted at admins only +We can take another look at it then +cant remove Map from ofMap i added it there for a reason because the compiler was getting confused +we dont do ip hostname resolutions in C +Updated RB with latest patch +So have defined a new interface with implementing singleton classes to execute queries +done for dataTable drop menuGroup menuItem updated for calendar datascroller +the general inconsistency that led me to file this jira was most likely because some of the files checked into trunk were produced using thrift v you do anything differently +There is nothing wrong with your XML + +Thanks a lotBest regards by Richard. +but it definitely is noticeable in the setup +Still thinking about this so Ive just posted on the forum about this in order not to pollute the jira +to patch looks good to just committed this thanks Ari and Jiaqi. +The change has slowed that down to years +Arrgh +I agree with Sergey +Will be released with version s clear from this is still a problem +slight modification to response this issue fixed +Hmmm +The patch removes the jetty dependency from the plugins as it isnt necessary at all +Dmitry are you sure it affects the current branch? Can you update your working copy and retest since now can be given a in the constructor +Its not necessary to force marshallerunmarshallers to return a preferred stream +The Javadoc needs to be very clear on exactly what the escape character is used for +Contributions welcomed +Thanks Jesse hope commits r and r are applied as you expected if not feel free to speak outFrom a quick glance this microbenchmark doesnt appear to test changes tothe selected keys set +Ive attached a bug template to demonstrate the behavoir +This unit test occasionally fails because of short session timeout period +If a committer feels this is wrong it can be reverted and discussed further +pending the results of your test you described above +It would be good if you create tickets or add to tickets already created to mention the associated issues and why you think they are unimportant or immaterial +The same work has to be done for all all YARN really cannot fix application issues in case of split brain problems due to network partitionsFor the launcher point of view clearly there is work needed in oozie side to make the launcher itself not restart jobs from scratch +Thanks Harmeet committed in rev +Thanks Prashant +added javadoc for the deprecated tag +Attached log with screenshot with error message +Yes Ill do the transformation table +Sorry please disregard the first attachment which pertains to Foreach not For +To test this scenario i will add a test case to the framework using +Sorry about that +uses enumClass +We should discuss it in the dev mailing list again +Are you using namespace in ZK connection string? If so the typical problem is that the namespace is not present +If I were to do that Id need to start at the head and pass the octal representation around the whole way through +Once a tunnel is fully established data flows from the client via the proxy to the target fully encrypted +This should be pretty simple to do since wed just have two expected libs files and compare against the Hadoop version at build timeAn issue this raises is that is actually Hadoop since we bundle the shim for the version of Hadoop you built with into the jar +Ill do a quick test of it this evening and post results +Otherwise they would have to specify all the depends and that defeats the purpose +ZIP file of changes and support files for this issue +I just tried to it and Ive had no problem with +Fixed in commit ecd on master. +An additional check in that matches on baseUrl page +A release will be in the works once all required features for SMX testing are complete +I needed to modify some of your changes for the CTF tests to run without failures +Your suggestion looks more consistent with other system properties +to add back in but with a small performance penalty but the penalty is really just hit for binding files with no schemaLocation so not a big deal for most use cases +pertinent tests aretest closing stale resolved issues +Done that was an oversight on my partAs mentioned before there are a few to Map changes in the code that strictly dont belong in this patch let me know if you want them in another issueId be grateful if someone had time to review this patch +Ok let me add it to review board +But more than of the startup time gets eaten up by loadClass and doPrivileged stuff where we cannot do much atm +because of a restructuring +So they are fixed by parameter passing +Fixed not to fail even if the boundary is missing. +Can you also work out a patch to fix this one? Otherwise attach enough sql to setup the data on someone elses machine and wait for a developer to have enough spare time to look into this one +This patch fixes the issue and adds a unit test +Patch changing to +Commit at subversion revision +Removed the flushBuffer bit as it was on and trunk +Yes sounds good to me +Thanks Eks this looks good +for the patch +Attached patch puts the system messages in a mini content area just above the main content area +One to check if some surprising migration needs to be done and provide linksoptions +Tested still working fine on Chrome Safari and Firefox on fixed +For already partitioned files +This patch disables the switch back where goes disable enable disable delete and replaces it with a disable delete for now +Alternatively you can tell me all the classes affected by this change and I can override them in my class path. +reproducable on todays manual +Obviously documentation on any all of these is welcome though documenting best practice is preferable. +Small patch that solves this problem +I think this is good to go +Look for it in a day or two +For instance an order manager application would have a Find Customer or Find Supplier screen and that screen would constrain the Find Party list to customers or suppliers +I dont think reverting will be enough. +I missed app master recovery local mode and test mode + on both JSON Thrift +I suspect of the usefulness of that page can be had with two or three words anyway +Test case for clarified section in in r by switching to the less strict behaviour which got defined in +Organize imports is not yet supported by the plugin +Sorry I moved by mistake this issue to the TILESSHOW project and then I moved it back with a different issue idI hope this did not much harm. +It would require additional research on my part but hibernate must have the information it needs to perform this operation. +I dont see how it could work under any circumstances +Committed +changed the jdbc metadata to return the jdbc which is a native datatype mapping between jcr datatype and jdbc data type +closing +Ive just uploaded the latest patch superseding all previous patches +I think the idea of providing a seperate Maven classpath container for compile versus test versus provided scopes is a great one +Also added unit test +Attached patch fixes the issue +Release my finger outside the control +Is that sufficient +Patch has been committed +For example if there is a need to validate pdfs containing then the error code could be checked to deny or embedded if the current PDF syntax would continue into the regular validation process +The error was triggered by which can be enabled by set true +Thanks Gregory +This is not important issue but can be a start point for other optimization exploiting ordered traits from multiple bucket filesAdded configuration means number of reducer for lastly MR stage +It appears there are unit tests for the but no integration tests for resources and probably other aspects of the Karaf integrationTomcat support is probably valuable to some new folks but this change makes the server unusable for existing ones and I think we should move back to until such time that Pax Web has better testing for existing use cases. +Update patch added test case for trunk. +Thank you Suran for finding this issue. +Hopefully it is compatible to svn + +Im on vacation now so I dont really have time to do this myself. +Improve extracting capabilities executioniii +loaded the wrong URL even without the java portlet created +Whoopsmissed those sorry! Also cleaned up some casts in two other +Thanks +I suppose you could call with a parameter +patch swaps described parts of code inThat is my first use of jirasvn and diff please be patient if smthg is wrong +Im asking because most tools default on nn interpolation when creating the overviews +Ah I got snagged on this as well +Li asked me lzma some of my logs from the wild +The NPE thrown for this issue is not fixed as they come from the runtime module here the fix provides a meaningful warning message to the user so that he comes know what went wrongAlong with this fix i have also included the itests for the validation messages +There is a simple patch in which solves this issue +I am +looks good to me +Maybe the version in has a problem +duplicate +Making it would make things much more flexible which is a framework goal in general +Ben just to confirm that Jeff Brown is to get developer writes for cvs jira and confluence for frobbin a DSA key +I certainly didnt add a dependency for it. +Not sure this grid is telling the truth nor is understandable by humans this grid states only one version of RF but multiple of AS and Seam +Ill work on getting pg binaries built to allow for the imbedded database choice +Attached fix for this execute action in when adding project +but Im also not sure what are the side effects if we comment these codes +All HDFS and Common unit tests passed except +in My should be selected by default as well as in Shared +Thanks Simon + +well I guess it is ok now +I dont think adding the XML files to the test jar is a good idea +Please give it a shot and let me know if this is what you expected +Make sure theres only one version of the core Spring jars on the classpath +Im having a look at them +Did more test on this issue +SureshI have deleted the file name appscriptscontrollers and changed the name of to in this patch +I have a question +Verified that the test fails without the code change and passes withJust committed this to +I will commit the patch as is +Temporary queues is only associated with a jms connection +It also indicates theres an SCM component to be fixed which isnt in scope for the next release plugin release now that is already out. +svgsCheersc +I have run the build twice and it failed both times +Still not seeing what the issue is here +The color scheme for buttons part is terrible +Thanks Joern +not sure why we havent picked up this fix when picking up the groovy versions +Please consider this patch +I agree that the final fix belongs in ant +Consider opening a new issue for follow up work. +Preliminary patch for Decommissioning testcase +add to images directorymodifications made to axis web admin guide to increase usability and accuracy +The patch might be useful in the future although looks like there are some rd party solutions already there + for the patch +It reports the uptime of the monitor program on the master summary page +branches +The previous revs had long summary sections +Tom can you check this our while looking at +bq +Please if you can update the apache Jackrabbit OCM example site +Splitting the MR changes into +And its in any news about this issue? Can anyone fix it? +We should add this patch and benefit of the perfomance issueThanks is the patch in GNU unified diff form for slightly easier reading +I just have to make up my mind whether we traverse the DOM tree or whether this should be done upon encountering a namespace prefix where no namespace definition is available +Ive attached a patch that moves everything from the into a class called +I think I started that somewhere in JIRA a while back +As per A +bq +For each security mechanism the loop starts up the Network Server with that mechanism runs various tests then shuts it downMy findings so far are interesting all the security mechanisms in the loop didnt affect the bug So it doesnt seem to have to do with the security mechanism that is used +To reduce the risk of problems what about creating new classes instead of patching the existing classes? Like that you could concentrate on one database type first +With round robin dispatch we rotate the consumers after each message memory limit of a queue is consumed by the storequeuecursor to cache message but also as the messages are dispatched to consumers they are still held in memory and so consume resources +It can be done in later in +Please review and reopen if an issue is found with the fix. +has a special meaning for that term which doesnt apply to any interpretation of resolve +We are going to have quite and awesome CLI for version +I mean Axis does not know which handler capable for what so Axis can not take the decision at the run time by looking at the message context +Not enough code here only the Map one has any code and isnt worth much +any chance of having a textarea for lines to contribute to +When the tests are run with grails the assertion fails as expected for meCan you provide any more details on how to recreate this problem +See for details +all core tests pass with this directory +Moved to upon release of +hiI just ran a test with a sequence of requestsThe deadlocks does not occur as frequent as when I reported but is down to about of the request +Reopening to address Lukes were changed and is now an interface +This is highly probably fixed on it isnt although the problem occurs not as often as before +Thanks for that additional analysis Bryan +Looks good . +These files are deprecated on the platform by SSO +Ideally MB or at most +The should be able to continue to provide administrative functions and file access even if file creations must be delayed or deferred +The STP deploy component has become inactive and been retired but weve fixed it in FID that you can do this with the WTP Server Framework +Such as MILLIS and MILLIONS +Fixed with FOP viewer enhancements in +I also observe this failure often +Use at instead +I tried using this artifact from two different locations and the file appears to be corrupt +Logged incorrectly to wrong to wrong project +the latest patch looks good to me +Deployed on Feb +Can you send the exact line this exception thrown in the code I have sent to youI can not reproduce your problem here +It is present to make manual assignment more clear +So whats wrong with richeditor richpickList and richorderingList +I did not use because I needed trickier comparison logic to handle names +I got the resolution thanks for your replies +Sadly no +This is a test project containing only one +Slava please verify +I suppose I will have to try to build the plugin from the nightly build and see if that makes a differenceJohn +bulk defer of issues to +Hi AndersSorry to have dropped out of the conversation for more than a month but I had to switch to other activities and had not the time to investigate more +Marking this issue resolved based on jira comments +Then this vector must be incorporated to Bt job to produce offsets there +Added a test case that demonstrates the issue using and +is frozen and we are generating a image for a vote and release +The first version makes gcc complain about initialization order +Upgrading to blocker since we cant have a bootstrapping server instance without this +If no one else objects Ill apply that patch +whether disabling classloading for the latter or creating a module for the former the results are the same +A patch to fix this condition +Menu item was added to the Source Context menu +Your understanding of partial join order is correct +The EAR to reproduce the issue +Mass closing all related issues as the module does not exist anymore +My mistake +Debug file attached when running mvn site from parent project +Hi nathan I assume you want us to apply this patch to activemqIf you want us to do that you will have to updated the copyright headers to match the copyright headers in the rest of the project +With your proposal one would have to create artificial classes to satisfy the dsl scope restriction requirement +Postponing to the next your server side method signature to debug this issue provided details are bot sufficient +Please let me know if you still have problem +I havent had time to figure out how to set up the spring configuration files to use this class instead of the normal J one so it is really untested +Was not able to reproduce on Linux however reproduced by throwing an exception in the referenced code +Were kinda hamstrung by Javas type system here +Due to Jira sanity check fix version was set to CR +And yet two H failed tests magically passed in H +Committed to trunk and branchThanks a lot Ravi for the patch. +Hopefully Ill get it sorted out today even if I need to forcefully convert the path myself somehow. +This needs to be part of the migration information +Will update the patch after fixing the bug +The suggestions for future improvements all make sense +Mark as patch available when the new test is up +Following this principle shouldnt int be accepted as a reader schema for a long? This now seems inconsistent to meThis would be reasonable if we wanted to check that every value fits in bits and throw exceptions when they dont +Moreover when saving all the default dashboards that have been unfollowed are flagged back as followed +improvement committed +No IT +Has been incorporated to the latest ydist patch of +flushState in DWPT is unused to remove it +I uploaded the patch as an attachment I didnt mention for the patch Matteo. +moving on req fromPostponed +Submitting to hudson again to get a proper +Please find the patch attached providing one of the possible solutions to this +trivial change +I dont see it yet but a comprehensive proposal can clarify what you have in mind I suppose +This patch renames the stdout logj appender to console and sends the log events to stderr instead of stdout +Repeating an old question is there any reason that this patch is only for Avro? I think this could work for all storage types +Hi SagaraPlease review attached patch which contains my progress so farThanks +Attaching an updated patch +thats why I was suspicious with that catch clause +Thanks Todd +Other paged tables work this seems to just be an operations history thing +Otherwise all memory will be escaped after a time +Thanks for the quick replyYes I have tried various variations of catching the but being a less experienced developer I have not managed to find a way to retry thereload command if the exception occursAm running standalone mode btw. +See the label for the progressBar which shows uploading +ech +Setting the value to true or false in plain text wont work +they will take effect even though they should notIve been in austin for a recruiting trip so dont have the patch ready +Will this ever be fixed +Duplicate of +We should probably maintain an original value in the when and resend that on dispatch as opposed to overwriting the vendor supplied one with out own +this should be done in user code ie +rev adds a separate datasource for the plugin farm controller +Should be resolved due to the upcoming JBAS as well as in AS +Reassigned for a higher version needs +Another suggestionWhat if buildNumber was only appended to the versionSo the current version on the branch is and adding a build number givesAnd if the version was adding a build number would beMy reasoning is that it is nice to consolidate the true current version into one value and have that be set in one place and not constructed +Test case +This seems to work in +No worries +tmp dir so we dont have partials if crash +Committed at subversion revision Touches the following filesM javatestingorgapachederbyTestingfunctionTeststeststoolsM javatestingorgapachederbyTestingfunctionTeststestsderbynetM javatestingorgapachederbyTestingfunctionTestsmaster +It must be some modification of common navigator +as InvalidAlexanders concern about a potential clustering issue would only have been valid if your configuration would have been correct +Perhaps the problem is that that jar is corrupted in your local repository +Please file a new JIRA issue if a problem is found. + +No you cannot escape in Unix within single quotes +Actually I have a minor bug in the second code snippet +on patch +Committed. +For binary using keyb in the columns mapping should work +This version allows for pass likeif true fixes typo with pass true pass print okis disallowed +some sample code +It will mess up our green lights on test rigs? Or do you think it reasonable to expect that hbase user have sudo on a test rig +I am not sure why why chose to make them serializable in the first place +I am convinced that Im not using the Counters Counter or Job API correctly here +This is a workaround specifically for jetty and harmless for other versions +I am not sure whether Spring Modules is the right place for this to be honest +Feature already function problem with JIRA +We will want to make sure the jira comment is posted to notify the user similar to how applyPatch is being used +This bug prevents proper hosting of services inside a container +The rule qualify ambiguous references to disambiguate is clear enough +Committed +If there is a specific need for this in we can merge it to the fixes branch then +Thanks SparkPatch applied to LUNI module at repo revision rPlease check that it was applied as you expected +Can someone please attach a test project that can be used to reproduce this problem +oups forgot to test that functionality you have to open as much prepared statement as possibleI was stressing a single functionality of my application without any resultwhen i opened the scope of the test and created as much prepared statement as I could then the bug appeared. +Does it belong there or should it be reverted +Will patch shortly +Added on Cellar trunk revision +This issue only seems to occur if the is removed but not the sample MattYour issue description is quite confusing you talk about war files and ear files in the same JIRA +has been committed to with the examples found in the site +Id be fine if that part of the patch was ignored but I personally think that it makes for a more complete LICENSE document +This could be caused byIts an ignorable warning as the system couldnt update the journal during a shutdown +Im using Postgres as well +Yes hopefully no update to come on this one anymoreThe only exception would be a committer noting something too odd or confusing +I found that the test complexity is difficult to remove without breaking the reproduction +I also forgot to mention that this probably only affects use cases with something like vm transport where the message stays internal to the same JVM and is not marshaled across the wire +Not going to happen for done if we uncomment the block in. +OK to commit this? Would you like to do it or shall I +Therefore Im removing the fix version of from the JIRA +But I found one issue in the patchJava Method cant be started if exception is raised +should now finally be fixed +Patch looks good for me too. +The idea is similar to popular accounting packages that have you log in to a company +The test fails intermittently it passes about of the time +Although Im wondering if this release note still applies or if this information should just go into the server and admin guide +I am unable to replicate this issue in wtp +Closing old issues +Fixed in trunk in revision change into branch in revision +Attached patch +Marek can you investigate please +I cant think of any likely problems offhand but it does give me a sense of foreboding +And we want to do this for since local lucene is not yet released and we have the freedom to make such an otherwise drastic change to the index formatIll update this issue to reflect its new goal +Thank you it works with an another look and feel +I am working on this ticket +Thanks for turning this issue around so quickly +If an alternative is pickedwhich is finejust keep in mind well need a tool built to go with this change +Thanks JasonI put this into trunk and +Convention over configuration ideals suggest fixing this somehow +So for passing the to the add a line pathtotherepositoryto the file. + +This happens to me too +From our experience that is not the case though we are usingHere is what had happened +The other related parts included in the previous patch are too big and out of scope form this issue so Im going to work the parts in a new separate Jira issue +is not explicitly imported but is a result type of a method call from module util +BTW you can look at the code of or to see how the distance was stored +My miss even if blocks size is equal to desired size partitioning should be executed. +Ok so then can you please attach a screenshot showing how GC is reduced with this optimizationThats would be a more factual argument that memory allocations dont you think +seeked around +In which case how about we create a feature branch for remove the relevant changes from trunk and put them in this branch so if someone wants to pick it up and close it out they can this way we dont have to spend cycles in other development maintaining code thats not being actively developed +Tim I would not say this was fixed for M I just added the tests to the regression suite +Fixed +move resolved issues to so we can generate release notes +I applied the patch +Looks good +clear push for Lucene to happen soon so well use it +Maybe if we record them in SIS its simple enough and we can keep them? If they are meant to be used by Codecs only then maybe we can force Codecs to manage them themselves but if +The key is that youre using basicTypebeginningdatetimegroupmonthname +Working on that +Container Table +This will help us identify the bottleneck of the reducersWill it? The shuffle phase can be delayed by the map scheduling the sort phase is often a noop and the reduce phase takes whatever time it requires +I need this mechanism to work and have a working patch +I considered it +Requested that the original reporter verify that they have the proper file with the NMS assemblies +Thanks John! Please open a issue and lets get the rest of the TODO items done +I am sure there are numerous reasons why this was done and I wouldnt expect the default behavior to change in such a drastic way +Here is the patch following are the steps to test this patch +I think this is an improvement not a bug +Havent had time to work on this yet but assigning to myself and triaging for so it doesnt fall off my queue +But I dont really think its going to bea big issue for at least for our web application +Ruel is it okay if I assign this to you? Lets have a call this week we need to discuss which versions of can support this config +Thanks for the patch! +I would suggest you run them as I havent done a build on Windows at least not recently so Id appreciate knowing if there are any test failures +As Piotr doesnt want to provide the patch Ill do it for him Simple patch adding a new constructor that allows to pass additional parameter the recursion cap +Please update your address book and resend your message +Update Sites names still contain version +Issue closed with release. +classpath contains that causes the plugin still to have compile errors when JDKJRE exists in the workspace +Please let me know if whats up there is stale and Ill check on the copying process +Trunk HEAD Parimal I have maybe missed something else I will double check I will check and change that I let Jinghai following Jinghais code I will rather use debug warnings in most also trunk HEAD JacquesJinghaiWhile testing I got couple of more question here I have some short of answers for these but it will be good to get answers from you guys +v incorporates the disconnect patch for and removes a redundant endpointstate lookup +It is questionable whether anyone is ready to add support for transfer encoding to the current framework +To run bvt only do netui ant +Absolutely agree on that +Before I remove this line is it possible for you to determine if there are other asserts in that also fail for you +Perhaps my previous builds were somehow differently damagedSo the patch seems to be working correctly +How is this related to? Could you attach a sample deployment that allows me to reproduce waht your are seeing +Ive added a unit test and an integration test to cover this. +Theres still the possible locale issue when checking but now that code is limited to a debug section so it shouldnt affect any production use +The reason that it was moved out is because the to cleared on a Callinitialize before serialization happens +fixing comment and think patch is missing a small fix I added to +It has Print preview as well +Today the suggested invocation is ant clean test which runs unit tests javadoc and checkstyle +Thanks Patrick! +Thanks! +Removed ampersand and replaced with amp in all URLs +I did experiment with polling under C a while ago and this caused virtually no CPU at allYour idea with using a queueing mechanism sounds very attractive to me However I have already spent quite a bit of time for all this so I dont know if I can spent even moreAnd when I talked to the JVM support they explained to me that the behavior which is described above is indeed needed just in the way it isThanks for your feedback Cheers MarcI tested your patch against and here are my numbers from two rounds of a iteration simple testTPS on a laptop for simple proxying with SynapseIBM JDK Patch Sun JDK Patch Sun JDK I have been following the recent updates on this from Oleg but do agree with him that ideally this should be fixed in the JRE +As Anthony Juckel said disabling invokedynamic by setting false fixes the problem for me however so does setting true without completely disabling invokedynamic +Why didnt findbugs find this? Or did it +Thanks +Seems to be the same bug +I had a look at the patch and it doesnt look to me like it implements the suggested workaround +patch looks ve committed this! +In revision updated srcref in the trunk with sqlstates and messages as of derby code line +Closing this minor issue as wont fix. +Removed most of the obsolete code in revisisons There are still a few references left +And has also a reset method which is used for example by the Parser to read the data in again we would need this method on an Iterator like interface too +Also not to have the same package names across jars +Then Ill raise some JBESB JIRAS to update those items if necessary +Sounds like this might resolve NPE in Solr if more than one +Will need some unit tests to help prevent people from breaking it since hudson doesnt have a parascale backend +Resolved in Git ID abdbcbafadbeddc +Today I made expenses report and run it through mvn gwtrun + +bq +I thought about changing it to +En tot cas podrem en col locaci de la primera pedra +Search the Autoconf Macro Archive for prebuilt solutions +Deploy to AS +Fixed default behaviour is to trim +You meant true otherwise false rightNo +Im not sure whats wrong with the patch +You are right also needs to be migrated +By the way switching back to make it work of the time so there must be sth with +Thank you Claus for your diligent and work to help isolate this issue +In trunk for is still not working correctlyWhen you rename a seam project including the you will end up with two different available for deployment +HilelThanks for the patch! +I dont know if its in the correct format +Eyebrowse has been replaced. +Im using RELEASE too and still have this issue too +Some notes The GML file is invalid wrt its schema +It seams that in contrast to your opinion persisting none leaf nodes is supported except for dbLocked mode +True enough the XDT needs to define the characters for IMPs sake +Without a workaround we wont be able to use Sonar I am afraid +Is your install directoryDo you have any relevant errors in your error log +Pushed in ae +If your extension use case is common we can consider adding it to the framework and then you wont need to fork or extend +OK I just committed them +Thanks Deepaknd Commit +Since we do put a properly formatted in the classpath we thought should be loaded +will be used instead with elevated privileges as referenced in +And yeah with a gb heap itll take a long time to get an error seeing as it took over hours with mb +The one way to do that is to have the TT send their passwords in the response to the map output requestHow is TT password generated? The same way as the reduce Task password? They cant be the same password since otherwise TT can simply read the password from reducer request and send it back as response +works as expected +META +It shows a couple of simple tests to show how you use and tests the couple of major interactions so unless anyone objects I would like to just resolve this. +I added support for project groups and authentication +I think it is too late in the cycle for this change +Marking as resolved since we cannot reproduce the problem and have gotten no additional response from the original reporter on the forum. +Closely related to being applicable to only xfire or cxf Im going to close this out +But a developer could also define and use their own provider supporting that interface +New patch to fix and add license statement at the top of +if match just continue +As they are they refer back to the original documentation so I like keeping the namesGood reminder on the copyright block +I imagine the first two can be examined and dealt with quickly +Committed as proposed +I dont think its a problem that the procedures are visible since theres not much harm a user can do with them +Please the patch and upload it +I have a theory that the issue also has to be closed to show up in the release notes +al get their hands on the code +Applied patch in revision Merged into jsr branch in revision +I havent heard of problems recently so think this is working ok. +The sample uses a large amount of code from main that wraps the JMS API and makes it difficult to determine if the issue is related to the wrapper code or to +Such a feature would be useful and would be completely consistent with the Avro specification +Verified in dist build from SVN +Refer to +This issue is closed as out of date as per the middleware process rules for aging of bug reportsIf you believe this issues should be fixed please reopen and add reason for reopening +Yes and if we dont use the lib classifier then we also need to go back and correct the scripts which assume that is the jar names +Aziz any solution that requires mvn is not in applicable to the maven ant tasks +Had to regen the thrift code as the patch didnt applyOtherwise . +Done +We can move it out later +Its a simple approach but I think it worksIll be running the overnight +Id also prefer the approach like one should delete the inprogress znode on exception if he has the lock +patiently awaiting +I am still seeing this error with Axis in Tomcat +This issue has been fixed and released as part of release +We added a strategy and it is initialized in the same way as messageSource application etc +So this soap connector can be used in and also when CXF libs are already present +The reason is that every filesystem needs to implement the raw stream interface in order to provide a data input stream +Ignoring the hostname the datanode claims to be is blowing up tests that are checking rack placement +Approved for +I just committed this +Yep Juergen a test case would be very helpful here in isolating the problem so we can get a fix out +To get Exact to work properly we need to resolve +This definitely caught me though +If someone can figure out how to get it to run faster please post a patch here and I will resume my work +and pushed both in JBDS Build id +Some processors allow this use ofa variable +This cannot be implemented due to limitations in what can be done. +Verified. +Confirmed the patch fixes the issue in my case +Your patch mixes LF and CRLF line endings +Actually this profile is needed only for build server +Committed revision +But I was under the impression that we want to have a generic +Application is using hibernate under managed env using sun app server +More information is needed as to why this is a bug +Committed to branch. +It appears this isnt a framework issue the related SO question is answered with an incorrectly implemented Editor implementation Gareth I didnt get a chance to come back to this on Friday +Attached patch handles the following env variablesHADOOPNAMENODEOPTSHADOOPSECONDARYNAMENODEOPTSHADOOPDATANODEOPTSHADOOPBALANCEROPTSHADOOPJOBTRACKEROPTSHADOOPTASKTRACKEROPTSHADOOPCLIENTOPTSNotes There is no HADOOPTASKOPTS +I just added a dot to the regex so that it will match and not fooxmlJust out of curiosity do you think it will be better to use a Pattern for performance reasons or is there a logical difference besides ensuring that there is a character before the file extension +What you might see is that is fixed in +Was that the rationale for not using the transaction implementation last time? +YesI am aware of thatThe problem is that the HOD still requires the user to explicitly specify a UNIQUE cluster directory for each hod ob even though it runs a given scriptThat was a mistakeRemove the changes in just committed this +If you find occurrences that I missed please reuse this issue so next time we have a place to look for all the files that need to be changed +fix committed in r +If we did that would we need an extra knob +and can be run as x cp +Added a test using the old mr api +w commentsI find it odd that we have to build geronimo times but this is an artifact of our circular depdendencies rather than a difficiency of this patch +Uploaded artifactscomsunxmlfastinfoset activationactivation activationactivation activationactivation activationactivation apachecxfcxf apachecxfcxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachecxf apachegeronimogenesisconfigconfig apachegeronimogenesisconfig apachegeronimogenesisgenesis apachegeronimospecs apachegeronimospecs apachewscommonsschema apachewssecuritywssj apachexmlbeansxmlbeans codehausjettisonjettison codehauswoodstox mortbayjettyjetty mortbayjetty orgmortbayjetty mortbayjettyproject slfj slfj slfj resolved issues for already released versions +Hope thats okay +we shouldnt throw back to the client its an internal error and either means theres a bug or a hardware problem +We really need to get you committing your own patches. +rUpdate +uses the which require at least Maven +A binary format would be fine +This issue has been fixed and released as part of release +Do you see this issue with autocommit off +The test will write the generated page to stdout and it is very helpful to know which page the test corresponds to +What already works in the branch consumer for creating a Nexus format index search implementation which uses Nexus quick search partial integration of new search implementation with webappWhat still needs to be done search implementation which uses Nexus advanced search searching within search results complete integration of new search implementation with remaining things to be done are already fixed and the branch can be merged to trunk now +Thank you for sacrificing your posterior in the name of science +Moreover some people or organization cant use the release plugin so their packaging would not contain this particularly interesting informations +part of this issue is that the currently and are not really designed to be used from the grails command line script since the grails script isnt aware of Ivy at should however be able to do ant run And this should use IDEA for Grails development and because of the incompatibility of the Grails Maven plugin with Grail we switched to dependency management via Ivy and build management via fine +This REPLACE the previous transition to closed remaining resolved bugs +Have used for generating it +Add button was added edit dialog was updated +I agree +Technically the shuffle handler is a web service +And this is the diff for the branch SPCP +Now RS does since so this condition should never arise going forward +Documented as part of Websphere reference guide chapter +orm files for mappings wrt +Okay I will add it +That way a simple listWill provide the list of all the existing performances tests +log file or into the will probably take a look into this but not soon. +I believe this is now a dup of thus resolved by revision . +It appears to work on but not that +the client would schedule the next partial scan ahead of time +Hi Kristian +Those changes can be ignored for the sake of reviewing this patch +EAP T E S T SRunning run Failures Errors Skipped Time elapsed for EAP CR +will be reevaluated have tried to reproduce the bug using JBT without any success +Are there API changes that need to be covered by the upgrade? or is this just a examplesdocsdefault compile v update +Can you take a look +Delete some code from patch that never should have made it version of patch +modify the class loader to parent far it seems to work I hope this is useful. +Hi Tom I have verified your sample and it works properly after fixing thus Im resolving this one as duplicated for the sample! + could be used to argue that is not a path mapping +Yep its the best approach +I meant Thanks Milind! Sorry +Here is a simple fix +I have some reservations about making it even easier for new Lucene users to write inefficient codeIterating all the way across Hits is normally a poor way of achieving something in Lucene because of the repeated querying going on in the backgroundNew users already fall into this trap too often and adding this feature will undoubtedly lure more people into writing bad for loops +Thanks! +Can you rebase andWhen you do please tell us which branch the patch is for +Committed to trunk and branch will kick off another build +its interesting for a contributor ready formatting for on though +Does NN keep running if all the storage directories fail? Why should it then enter safe mode? NN should keep running even if the last storage directory fails +TxWhile this is a quick fix that should help we should think of more long term solutions specifically looking for correct exceptions etc +solved this issue. +Workaround to configure client with like your code from patch is a better approach then making createClient protected so closing issue +Ok how about this patch +Cannot reproduce after fixed layout version issues. +It seems that commit number caused a regression +Seems like the incidence is low +I pointed to this one as they really address the same problem +I also added some useful links on the welcome page the copyright +needs support for compiling SQL into operators so if you have that working Id like to point the Harvey Mudd folks at it when I talk to them tomorrow +The javadoc in does state anyhow that the class is not intended to be subclassed +This indeed needs a fix I think it will just turn down to using java jar as the command rather than trying to open the Jar +I meant to check that box +By correlating them we will be able to systematically identify actual failures +Hence I think we should just double check the documentation clearly states that we dont support this and then mark this bug as resolved wontfix +Jon could you please look into that? patch is provided with permission of my employer with rights granted to the asf +listBox hselect for your example +COMMIT and ROLLBACK are not part of Derbys SQL syntax but ij understands them and translates them to the corresponding JDBC calls +Hmm +Initial draft of the conversion of this test +The documentation is now attached to the category +So although I agree we should be spending effort on the Maven class loader functionality this will remain a concern beyond that +has made it pretty clear getting adoption would be an uphill battle one that I dont have the time currently to fight +If you want local test server packaged as a separate artifect please let me know what classes packages should be includedOleg +Updated the description of the jira to reflect the sameAn updated patch is also uploaded +Committed to branch at revision +So I agree that from the Spring point of view this is not an issue is why Spring cannot be used as a JNDI provider is there some reasons not to use it that way? It will completely remove Spring dependencies in the user code +Any ideas how to solve this? perl Cant locate Bit in INC +For QAVerify it in Open With dialog Open with Others Customize Perspective Dialog tooltip +Committed +Im going to close this one +Upgrading to blocker as we prepare for +The last one was causing entries in to get booted too soon +The srccontribstargatelib folder is empty and can be deleteddoes not need to be created at allI found if it was not there then ivy would error out +Specifically the existing data stored in the last block being reopened may not be replicated +Uploading concept patch +that did the sure if this ticket is the right place but could we upgrade to the latest metrics v in cassandra +The patch contains the updates discussed in the description +im trying to use it and I have to iterate over the full list to get the artifact I want +Closed for. +Hi Slava +Lets get a run +Futon API clients will simply wrap this functionality in anyway so I think its only people using cURL right? So Im undecided on this +if the datanode dies in the middle of a rename we can have a partial file in the destination as well as the complete file in the source isnt itYes thats true +BTW afaik the goal was to also enable other plugins like +problem is that the metadata are only read at runtime and not at compile time for the other settings read at runtime was early enough reading the cfc at compile time is not easy +I agree it should be documented and have a way to disable it +Bulk defer all frontend issues to +Do you have a patch to propose +I dont think we will be backporting this to since all future releases there are only +Please review tests +If the case is that is there a way that we can convert other data types to varcharThanks +I remember we had a discussion right when we moved to sourceforge about moving to maven and we ended up agreeing not to do it +Hicustomer did have an special security context implementation so wasnt an option +You are right that using an entity would make far more sense that copying and pasting a literal x but should we add a new literal named mavenVersion +Jetbrains that projects which provide paid support are not technically allowed to use the licence but it is a little unclear as to what that means to us here +Does anyone have a patch for this +I am OK as long as this gets committed +It pops up at the end of the packaging phase even though packaging was successful +Thomas your patch references wich is not included in the patch and I couldnt find it anywhereRegards +Is there a way to get a list of running containers assigned to the app with their start times without tracking it explicitly +The dependency is only cyclic in that is a static inner class of +Still keeping this issue open to get the user feed back comments issues have been resolved but never closed +This is a corner case but the code that appends the file extension to the view name does not reinitialize the view name properly so if there are multiple media types requested the second extension would append to the first etc +However as far as I know the artifact is effectively excluded during build time by Maven +I get this behaviour too with Tomcat b W JDK ExploreroI used to think it was due to resource contention since the browser and the server are on the same machine but now that you have such a different configuration i wonder +Found a better fix +Ill add it in the next rev + for me Using this PR builds and installs without issue +Updated version of the in rev +moving to +Once approved Ill commit begin release process +Patch is fine +Reopen with more details if we can it in a later release. +One place that puts up memory pressure is in memcache scanners each outstanding scanner carries a copy of the state of memcache at time scanner was taken out +It would be good though to have the multiple port based configuration and higher level dynamic vdb support for wsdl operations +I think what happened at iteration was that dropping of keyspace failed so the subsequent attempts of running the driver would fail as dropping column families and dropping the keyspace were part of the same block and attempt to drop non existing column family would throw and so the code would attempt to create an already existing keyspace which was not being dropped because an exception thrown by column family drop never gave column family drop a chance to executeSo the issue was not in dropping the column family but rather the keyspace +Can you also provide a patch for English +With lots of concurrent writers small buffers are ok and you make up for it w group commit +I can look at this latest patch soon Noble +We also see this bug in Lucene but NOT in RC +Ported XSD sequencer to into the branch +Different subtasks can have different fix version +Thanks Patrick! Resolving as fixed +Thanks for the update. +Committed with revision . +Hence if the user is trying to set the attribute with SOAP version before a method invocation which is the typical case there is no way to validate the version used +This doesnt work well with generator methods since they are implemented using branch instructions and its not possible to branch into a try block +I hope soAbout the P link it should be a link to the exact promotion like L link is a link to the shopping listBut getting product from shopping cart item is not very easy and this P is link to the same page instead of promotion detail screen +trunk +AntThe jars are the same version as I have +hmmm +Youre taking care of of unclosed issues. +whoops changelog cleanup w w tabs in changelog +Ive attached the ant build file +add a javacc element to d like to submit to svn treeIf you are not sure about committing this patch let me know +It happens alot especiallywith browsers that save form information +want a little more time to review this though this fell through the cracks +groovy files +But is ugly for the poor user +I should add that the race may also be a possible explanation for the missing file. +Reproduced see attachementBut failed in only out of runs +It really didnt seem worth the risk of introducing a regression just to get workingInstead were just moving away from that involve on +I stop the VM and i try to start it +I am planning to add more senders and receivers to my test tool and see if we can catch the problem +Usage of requestDelay sometimes is a must +bq +Am I missing something +Removing omitNormstrue from the fixes this issue +Committed to trunk. +This patch fixes what appears to be a bug in the origional fix +I thought you wanted to write this and submit your implementation? +renamed setGet as set etc +Suggest you leave this one on the shelf for a few days +That is it is not enough to check just +Will upload the patch with all itemspolicies as soon as I can +But some of the changes cannot be included due to API changes +Should probably also test for MINNORMAL which used to be out of range but should be supported now +The issue is caused by excessive locking in the bundles activatorThe problem has been fixed in revision in camel trunk. +the build automatically ran the integration tests also against whenever a local test database is availableThis was implemented a while ago +closing! +Thoughts +Any updates on this +The acid test is as followssudo su dcontroltouch homeprojectsnanocontainerdistjarscheeseThis should NOT report Permission denied +It could be better to use the latest version +Applied my last patchThis is probably it for now +patch I am able to get all tests to pass in except +I think the backporting is still a good idea for others but want to clarify the current situation with my deployment +It worked for me yesterday but not today +Regression related to committed to trunk as svn resolving. +Pull request polished and merged +This issue should and can probably be resolved as part of the refactorings +The ivyretrieve Ant task can now create a path or a fileset containing the retrieved artifacts by setting the pathId or setId attributes +LGTM +Im trying implement a generic batch job launcher to be used to schedule jobs in Quartz +Im not sure how to solve that problem +But isnt this easily solvable? we just wrap a around that +Can be applied to both trunk and branch +Made a number of changes to the code +Bryan which container do you run your appAnd how do you restart your app +Patch has been manually tested together with of +Thanks +Im closing this as cannot reproduce. +Yes I did see itbut the original report is from June! Does no one on the Maven team really have the minutes to fix this problem? This is a gratuitous pain to newbie developers just starting with Maven and for corporations debating their usage of this product having dead links that dont work on a vital introductory page reflect poorly on the product and how well it is supportedits like getting a resume filled with typos +I committed this. +The properties filename is configurable from on a basis The testcase is cleaned up +How does it work on now? I had assumed it works there no? This really feels like one of those Doctor it hurts when I lift my arm high above my head If we cant support running trafficserver directly from command line on its not the end of the world +A later enhancement would be to generate the images but every approach has drawbacks with confluence. +If anything runs at all its a lucky coincidence +Thanks Lars and Billy for figuring this patch. +start and stop in Unix Format as Attachments on this post taken from for input +Attached the latest patch includes a symlink implementation for and more tests +Still getting errors +for the patch +is a patch that makes also write the stack trace to file when an error occurs +Actually I can also reproduce this with a two node cluster +Will try to work on the description but have a issue with the priority reduced to minor +Id suggest bringing this up further on the dev list once you are happy with the work +I initially reported this feature request in which Ive now linked to this issue +Patch I think its ready to go +Generic Exporter is set as entity +Jeremy can the cleanup scripts be modified to gzip the old logs rather than rming them? That might allow you to provide us with the full logs for reproducing the issue +Verified in JBDS Teiid Designer. +there is a problem with the EJB test case will fix that and send a patch +First this hasnt been commited to branch +Wow +This new fix takes into consideration the nested jar scenario +My hope for all of this is to get this all integrated together get the wiki page updated and complete with information for both JSON and XML for the REST interface and kind of tie both of these implementations together with a more extensible inputoutput interfaceAt least thats the plan +I just committed this +Committed r +Bug has been marked as a duplicate of this bug +I knew the other test case how to reproduce it and after the patch applying everything worked fine +But it does not tell you that there are deprecated methods declaredI will post a patch with the rest removed +Native support on Windows x should be fully functional now. +can somebody help me restart the hudson test? It seems there are some transient errors +Closing this issues because partly my own dumb having trouble with the plugin but the issues in this case are not relevant +I will remove those title cases and of cause apply the typo and grammar errors if there are any in the patchBTW documentation section for the JIRA seems a good suggestion and will add that the patch with modifications. +Docs updated to say is the minimum level needed +Ok NicolasI will follow your suggestion +Overall though Im not convinced it scales well either to large files or large clusters the latter offering the IO and network bandwidth this policy would fail to exploit and the highest failure rates +And fixed of the Servlet Resolver to only get the response writer after setting the character encoding and also reset the response before sending back the error message in Rev +Too bad so much needs to be cloned in the case that closefalse +seems to have lost my original attachments +One group can see all the portlets the other one only portlets and the last group can see only portletsNow the page should appear normal I mean without any error message to all the groups of users +This is the cleaned up version run just now fresh on EC and works fine +Template and java files used to reproduce the issue in +This is not an EJBTHREE issue +Reopen for to myself for backport to +This is a limitation of the YUI library +This patch updates all references to use the new representation +Made the streaming parameter check backwards compatible if using older metadata. +Move open issues to CR +Fwiw Corey this is more an exercise in the classloading than it is thenative mapsSent from my phone please pardon the typos and brevity +I added the links listing recommended crypto primitives +The only difference between the you sent me and the one for final is the version of the branding feature +The javadoc warnings was introduced by but not related to this +Hi there I believe the contract with the C client is that the application manages the memory allocation of the data it passes to async calls so Im not really happy about copying the data +With the help of Paul we are using now at. +It seems that there are much more examples that we did not revertI havent done a thorough survey of commits which have broken tests so couldnt say for sureRegardless thanks for looking into it promptly! I agree theres no need to revert this patch. +Thanks for working on it +Corrected version number for report +I commit your patch to trunk with the following changes +This will fix this will fix + Tomcat giving me hard timeReplacing with old file and reload resolved issue. +Changed to a blocker since we really need to have this soon +ArpitNice catchI assigned you the JIRAIt would be nice if you can attach patch for trunk +Sorry that comment was for another Jira too many open at one time +fix checked in for and does this scenario work +Hi HansThanks for reviewing our patch +Maybe Ill try to add a profile to run that command after testsUntil then there is testsuitetools +Only dump the system properties from the at debug logging level +Hi Arul ok Ive checked the javadocs it says that the config location can be null so I guess I can close this issue +I thought it would be nice to have it in configuration so that youd not need to boot the node that runs overseer with the system property on +Resolving this after my last change since I think whats left is covered in +It might be viewed as insurance against integration of some annoying component or another that is in fact completely uncooperative in this respect +It is similar with I feel +Closing issue which has been resolved for some time +This validation isnt fixed as far as I know +Sure Ill try to get the patch to you by EOD tomorrow +It doesnt seem to work for me +Thanks for the contribution Cris I added you as a contributor in Jira too so we can assign you issues now too I merged your fix so I will mark this as resolved many thanks for the contribution +If there are problems please reopen +Would it be possible to get a stacktrace +Please check that it was applied as expectedI also added svnignore property with valueIf you create more files which should be ignored let me know +Yes on both comments +It is incorrected for you to think that you are not using XSLT but +bq +Could you clarify the problem please? Perhaps a description of what you are trying to achieve would help +I just committed this to trunk +A lot of them will have their own import MR jobs which converts processed data from one format into and then call complete +This patch addresses the original problem by using the Partitioner to ensure that subgrams and ngrams are processed in the proper order in which to avoid having to store values in a hash when calculating frequenciesSubmitting this patch to capture the current state of the fix +Look at my link comment +The black dot disappears when I use a normal Form so it is linked to +The fix has been confirmed +I suppose you could close this issue now +In trunk version everything is correct because of JSP tld +get does not suppress deletes because perhaps client is interested in deletes +Propose a fix on the Dev mailing list +Just a comment this bug is critical for proper creation of workflows for Liferay since enum are used to specify possible outputs +Thanks Mikhail +The latest patch looks good to me +Fix dependent on fix +Items to be fixed in multiple releases should have one issue per release version or we cannot accurately filter for the release notesPlease edit the fix version field to only one version and clone this issue for any other versions that this issue should be fixed for +Not sure if we have tests that show the delete case does not work +The opening of the first resource could easily be moved to read however restart isnt as simple +Fixed in revision +that way we could also but more experimental implementations in without risking default installations +My perl is a bit rusty and I was offline so I didnt have access to docs +The code for parsing hosts is spread in server and client code bases so there should be a more extensive patch in the future +is fine +Patch works on vesta +Part of a bulk update all resolutions changed to done please review history to original resolution type +JBPM doesnt work with Seam bug has been fixed in jBPM trunk but not a released has a bug in it that causes Seam to break. +perhaps I missed some bits in the branch but theyre not jumping out at me + for the patch +That might not be the case +Steve as the original reporter of the issue do you have a system that you could test this on +Moving this task for the next release since the linked issue left is out of the scope of and the only remaining remoting issue seems to be is tracked here +Additionally as we put more stuff into core that people could use independent of cassandra the server like Hadoop Pig etc +The patch is entirely code copied from pagedrawerve added Andrews patch with version Thanks to Andrew for his contribution. +The key had to change +Are you sure this problem still exists? If so please report +bq +run nrofthreads is the much awaited new version +So once weve made the next Hadoop release and upgraded Nutchs trunk to use that release then we can introduce changes to Nutch +Reupload to get Hive QA to run +Currently exceptions use STLApart from that client is not returned STL strings. +Set suggestion +Thanks Boris! +The issue is a known problem and already documented on the wiki +Yes I thought about making this change +WindowsIE +Ive the link in the code but it would be possible to add a property in the update center to set a upgradeNote on each sonar release +bulk close of resolved issues. +There will not be a moving to for +as per my comment of th of other thing I meant to add was to capture the fact that we must be able to be able to define a instance to be used with the load schemas option so that the newly found types can be registered in the context that the application desires +think ive fixed but need an IT expert to review before committingemailed ahmad awaiting his responsesent links to mark didnt find any issue in the two pages I didtested on IE on XP Pro SP in to Ahmad to review on his XPIE machine +Not sure its a scm issue but a continuum one +When Ive discussed this in the past my response has always been what happens when theres two implementations of Hello? Im assuming that the reason for their being an interface is to abstract the API for multiple implementations +I also changed the decoder switch statement to use only small label valuesStill to be tested out of bounds situations for offsets sizes and input element valuesand performance comparisons with and +This patch reflects your comments +If you register a Converter to the converter should be an instance of and should return a valid pattern from the method getPattern. +Not for commit here. +bulk cleanup of Jira versioning +The problem was your WSDL which was incomplete +You are even authorized to ask specifically for their legal opinions which they can give openly or under privilege +Im wondering if SSO has introduced a race condition so Ive added more loggingAs before this patch will result in an enormous memory leak +I was hoping for at least a code reviewwhile the changes do look harmless id have to invest a significant amount of time to analyze andstudy the potential impact +Heres the graph of the lock cycle +Sorry for the repeated noise +all issues assigned to +This was not any problem but client one +JimmyStackPls review the patch +I did that during separation of the plugin in different jar +similar problems also happen with the C backend +You then CANNOT have the file in the parent but need to have it in EVERY child +Thanks Carl! +Have you checked to see if you have this problem on the latest code +I think this is a major change in Runtime Behavior BW Compatibility but I only see this listed as a bugfix with almost no info attached in +Replacing the which is in that two java files into DFSHTTPSPORTKEYOr should I just only replace without removing +well thats why I called the function normalize Collating is about comparing strings so I think that does not fit well with a function applied to only one string at a timeIn one can specify a collation URI in the order by clause +Commited that fixes this issue +This has the Output side of shuffle changed to match the newTBD for the Output Generating events to go back to the AM Token handling Verify the output file generation is compatible with the consumption when thats done Other left behind in the codeAdds several to the TezContext interfacesOnce the Input side is also done will upload a patch for commit which deletes the old code should provide better revision history +So the plugin will have to be modified to actually parseunderstand the have started thinking through how this could be done but I do not foresee it making the release +Good catch +None that I am aware +i found some more while running helgrind +mrepositoryorgcodehausplexusoptcceAUTOMATIONeAMS +snk file only comes in binary form sorryI cant say I really know whether it is customary to include private keys in the +Attaching and the corresponding html output +the web ui shouldnt include shuffle sort time columns for map tasks +Attaching new patch with the correct fix +This is my first time so please let me know if Ive messed something up +Committed revision +Sorry if my words were too strong and thank you for your jobHibernate is great +You cannot input the original value of on the Datasource +Mmmm +eg +So shall we resolve this as invalid? +I totally suck +Fixed for next release +I use Tool Suite try Roo standalone and edit the file inn a text editor to see if the getter and setter is removed while the shell is running +Ari rc +These are two separate issues and so it would be better to raise two separate issues for them +Is this ok or do you want some other format +is public whereas and are not +Please note that dynamic imports are needed as jms replication is doing deserializtion +Didnt post it today +Please check that the bug is fixed now +Many thanks to Rick +Everything works well with the patch on two different versions of intellij +could you please let me know what driver do you use exact version of it +But if really a compaction is USER requested then what is the priority +svn ci m Adding note that using switch is not srcjavaorgapachecommonslangenums Sending srcjavaorgapachecommonslangenumsTransmitting file dataCommitted revision +is now decoupled from the adaptive layer but I would like to refactor it even more so it is easy to replace the with any other single token suggester +Verified in JBDS H issues is not fixed for Hibernate and be checked in further in Fedora +Transitive does need to work so that features like have their configuration pulled in correctly +for the merge patch available at have been moved out of this epic. +I have a CPAN account +Manually tested! +Thanks for tackling this Bryan! Looks like a clear improvement to meInstead of dot you might consider dot or source bash allows both use of the and the source keywordAs for the example do you presume that the user has done a cd to the directory containing the scripts or added it to the PATH environement variable? Would it perhaps be even more helpful to write something like thisUNIX Bourne shell DERBYHOMEderby +Ill fix the versions +This came as part of the changes in +In this test such messages are also expected for the embedded framework but in that case the name of the exception is replaced by asterisksFAIL unexpected exception I have not found out how this is achieved but maybe this is an alternative to separate canons +After running with the test fix for an extendend period of time the user saw the with the proposed patch so I wont be committing the change to +Modified internal locking strategy to mirror strategy used in Session objects. +feel free to provide a patch for this +We wont have the issue in production directly in Karaf as the spring feature and features descriptor are presents in the Karaf system repo +Thanks Gunther! +I propose to close this issue as the logging seems pretty explicit and just fine to me +We started the namenode by ignoring the save lease exception +Fixing Affects and Fix in Version +please can you provide me feedback on this or I will close of this JIRA soon +Tried it on a VM running Windows XP and +I forgot to say that the stack trace from the error condition was from beta +I still think the use of a checkbox for a one time action is not terribly comprehensive +This patch removes incubator disclaimer incubating suffixes and fixes links +to commit this is great Ross +Thanks +not is guaranteed to be the same object for each test class +Some warnings are still produced in release repository build however they require some changes in I could not effort +The zip you reference in this description no longer exists and there isnt enough information in this bug report to reproduce the problem +Changing affected version to order to complete the skeleton structure of the angrapa first of all we need at least dummy components that constitute the angrapa +Attaching the patch that uses rather than Statement to get current isolation level + +Username djencks +caching generateValues in Inserter +This is the latest fixPlease help to review it +Well i am not listed there yet +failure is because of +Yuk +Did you review orand test +Ant and maven build works +that I just committed solves adds this functionality +Version Released +See and contextzip +Adds the Promote actions to the button bar +Reported against this issue as dup because it is older then +There might be another reason for Xmx not working and that is that the bat file may use Xmx already and java does not like multiple of those +This is not a bug in Ode itself but rather an shortcoming +This should be enough +added to svn on svn? And whats the plan to avoid namingversion collisions? +I suggest that we fix them now +Would the fix also resolve that weve seen with +Ted the patch has a test which brings up a cluster from an actual data +Applied to the module +resetting P open bugs to P pending further review +As to order of reading the reason is for verification +Probably all of the source files under javastubs will need this improvement +I committed this +See +I think it will be far easier to debate the pros and cons of this kind of change once that is applied! +I am pretty sure there are few meclipse users who are using OSX. +Ive implemented the Ed Hill Display solution + +Is it still actual issue for AlphaDaniel could you verify the issue +I changed so it uses eth to find a nodes self address and the test passes on Rackspace +No theories atm +This means Dalis developers removed thewhat the exactly Dali version you are using +Test moved to +I have merged this to. +lgtm +That is right Pete has no business setting it is autoset for the confirmed to work dialects +This should not be a problem since all child executions should be ended first +How do you interpret thisMy take is that it seems like a rouge client swamps the server with Presence messages in extremely short intervals +This test failure seems to be caused by +Thanks Hairong. +Patch committed to both trunk and branch. +Enter defaults to the Connect button and Escape cancels the dialog from any location in it +This updated patch has the Interpolate function ready for testing +move resolved issues to so we can generate release notes +Reviewed no comments +Hi Would you please try my patch? ThanksBest Regards patch applied at revision r thanks a lot for this enhancement please verify that this problem is fully resolved as you expected +Do you want to add itThere is a rogue s on the end of presume it should say does not have rather than does haveregardless if i click yes I dont get the nature in my +The client probably has to declare the intention for a channel to be transactional when it is created +change in enhancement completed +Hi Jin Can you please let me know what version of and did you used to resolve this? I am using the latest and and I am still getting this error +John +Looking at I see that it kills an RS and that everything needs to be rejiggered within sI suppose it can take slightly longer on a slow machine and if the RS happened to also carry +Forgot to unassign +OK thanks Fred. +On at PM Paul R +The file changes are now as followsM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefA srcrefM srcrefM srcrefM srcref +Im attaching a simpler patch +By default it connects to the embedded Apache DS but if it fails to connect the user will have the option to connect to other directory servers +Can be blamed on meas I just checked for the log not the content when rebasing the has induced property name was distorted after it got it in the patchActually you need a pat on the back! I was trying to figure this out nearly one hour but couldnt +Yes +You mean we conceptually create a new ? where the part is basically the overflow ability +really likely affects all components +so you can remove the delegator completely in trunk +related +only fixes the LP blocks but it does need the changes +Hi againDinesh I dont think that AXISUSLEEP has any issue during compilation as it is being used in line of srccoreclientapiRegards applied +aside from the one missed svn add this worked perfectly on linuxLine endings are a pain. +However I also think the default behavior for when a client calls flushCommits should be that the buffered commits are successfully committed or an exception is thrownWhat do you think about using a separate set method rather than overloading set? +No it will drive you crazy if you know you have a working implementation on the class path but another one is failing that is found first +Fixed with r. +I just figured that for generating sources for we need binding support in place as theres a file in srcmainresources +Even though we might find out there are issues we cannot resolve and have to revert the patch but this can be done by any commiter and is not tied to anyones vacation +Im aware of the risk of evaling untrusted strings but Im not sure how this could be a problem with a Solr response +To minimize change heh I left them in +So deploying a module into a first container foung does not fly +The Patch part is now split into generic tips on submitting issues and specific tips on diffing and patching +Branch JBPAPPGACP switched to the new revision to Darran Lofthouse for all the work in getting this new version tagged and the help on building it. +so all statistics required by file scan are not collected +The result of is often a full queue of dead jobs and the admin is not able to see the really important things +Right now that will only happen at startup but in the future +Transitioning to waiting for Infra +I have run them in the project but I am not sure is the right place where to have them due the dependencies to the moduleThose Junit tests might deserve a separate projectNoteall the original tests have been migrated successfully but testattributesfromsession and testresolvewithslingaliaslimitedaccess +relates to class loading of the application server if not being able to find the class specified. +Thanks Cheolsoo. +there either +pushing out rolling +Yep the data is stored in a buffer inside the sequence file +Theres a project setting for that Sling Folder to sync +None of the changed methods are public however it would be good to compile Hdfs and with this change just to make sure +Unless the issue is already fixed in trunk that isThank you mistake +back kernel is and jdk +Im going to be looking into this this week maybe Ill see the propagation issue but if by chance you ever see it again please attach the logs showing that problem +Added documentation to Draft +Please supply profile activation based on whether maven is in offline mode +Verified very nicely written. + However these cases are well specifiedThey are well specified for runtime but not for toolsHow should tools recognize implementations? What if there are a few classes that implement this interface? Only particular CDI container knows what to load to provide spec required servicesAnd one more problem +shipped in +So it too has a precedent in Javas standards +Sorry I was wrong only one file has been work has apparently already been following codes are considered Success in and providers +some work started with rev +Committed to trunk +Xavier says this is now done at the user level and per eclipse session +Mhhm +This has been fixed recently in trunk not backported to though. +its really only intended for hadoop which parallelizes the queries Id like to find a way that puts clients on more equal footing if possible +added environment variables to advanced for the was just trying to document some progress with the placement of the action actually wrote a referencing moving the button to the other view page to better associated with the Mockup discussions that have newer is the one that I am going to work agree that having two expansions is not a good idea +Rather I suggest to the test cases maybe along the lines of to make sure all things being configured corrently the tests pass. +But I used sun sdk and I could not reproduce thiswarningYou can try to add the following line to the befor startingstart Name Server JAVAHOMEbintnameserv PSDo you receive message about my degree work +Which will require either saving namespace before exiting the or restarting it one last time before upgradingI think this will make things even better +Jonathan how goesI am wondering if we can make it into +In order to not lose this change I made a patch with revision it is not a type the signature L means the class uniquely specified by that name +Its a lot more than just a new name +fixed in JPOX CVS +Your idea? Regard! Mike tool namespace is not displayed on the UI anymore +defer all issues to +bulk close of resolved issues. +Marking closed. +Thanks Aaron +Wont check in until at least tomorrowAnd thank heaven for local history in +markus +It is very simple and localized +There are a few empty lines in class hence the line number difference but it doesnt matter +Thanks Christian Mller for supplying the patch +Thats what is supposed to do but isnt in this suggest separating this issue into two +I have included Christians patches on endpoints changed unit test to avoid failure as I dont have commited SMPP server mock yetCommit done +reopening to change version. +Patch for the for a comment for setIt should logicaly calls set for the conduit sendSince we do not take the message from the exchange it is harmless to set the inMessage or outMessage with the send out message to the exchange +Do you have a sample jpg you can provide to do some testing +This is a lot less work +ah yes I forgot how p can be at selecting bad mirrors in this case though my assumption is this coincided with the outageoverload registered at servers. +Manik Im working refactoring the tutorial to fit tutorial into it too so I might as well deal with this JIRA too +Can you confirm +pdf version +They work great for us and are a definite performance gain +patch created from root +I put it +I think we now have all of these issues addressed +Thanks for fixing this. +Thanks for reportingAre you considering working on a patch +XML attribute now contains spaces that are NOT ESCAPEDParse into DOM escaped spaces are normalized away DOM is now differentSerialize DOM to XML +Sorry for not replying earlier +I am unable to reproduce the exception +Talking with Tim we agreed it would be sufficient to have a single profile profileminimum that only lists and the native srs of the layers contained in the capabilities +I will test and commit +is a quite old issue which got originally closed without a real fix +in +I like the second option since this looks pretty easy to add the other option to this work +usernamepassword generation what do you mean? just disallow user to use default adminadmin + +Was suggesting we use the Interface rather than the implementation +Seems the in has some problem with findbug? By the way I run ant test for the latest patch and it passed all the local unit tests +How can I make one using the keys from Avro data? Still trying out a few options +In my case a server restart didnt help but a grails clean solved it +Creating a custom and specifying it as the encoder on each Grids seems to work around the problem +I have not tried it on or yet +Probably this will be fixed with close this issue since there is no activity and I cant reproduce feel free to reopen it if you can provide additional info +Fixed bug with realm format in Basic authentication request header and added a test to verify +Will commit if it pass tests on eclipse +files for these +Both issues are related to how passwords are handled +A couple commentso tadmincbdjhhfd We should capitalize the lead word in the sentence beginning for information on user authorizationo tadminnetservbasic Same commentThanks +We now always create a new target when converting a collections you mind trying out the latest? This is also scheduled for backporting +Additional YARN information added +Hi TrustinThanks so much for the fix +IvoDid the AS test guys get back to you about this +Added testcases and the patch. +The disadvantage is that you can produce code which will fail at runtime +Upgraded to Ubuntu reverted the patch to Trunk HEAD rebuilt +Updated Henri +See where James introduced this behaviorToday the exchange ids is based on AMQ ids +Please post the issues up to jBPM BPEL +The same issue exist on EAP. +This is what I applied +The instances created are actually extensions of to ask around and see if would be more appropriate +Thanks Jakob +thanx! +Actually you could probably keep all the method signatures the same but if it is a then recursively call extractType until you get a Class +Dont mix attach a test that isolates the issue. +Patch against rc +This should solve your problem too I think +Just moved to new package +for bugfix release +on on trunk +Thanks for those good suggestions Knut +If we get IT tests for these before is final we can take a no way to replicate bumping again +Most web applications I have worked with or used do both +I can get the https support related code as wellAs to proxy users per method the issue is some security concern +Ill make another attempt +Lets move our night builds to it +Here is the patch for above solution +Your description isnt clear here it says disable the scope if not KVM ? Also make cluster as default which is already there +I am fine putting this in but just be aware this fix has not solved the root cause +I agree that this should have a much hight priority +The next move i will followyour last advice and reinstall the serverThanksAntonio Nobre +Our first projects going into production are all facing the same major performance issues under IE is mostly due to the complexity of the generated DOM a few hints that I found above helped a lot for instance packing all the set to ALL reduced the response times by a factor of +and deployed a new SNAPSHOT +We use Axis to publish the but it removes all extensions from the coument before publishing it which effectively removes all the partner link type declarationsWell follow up with the Axis team +ok +for +Rohit precommit does not run automatically +Hence resolving the issue. +Please ignore the first line from my post +FYI +Please help us to correct this problem by following these simple steps +This is a great point! My annoyance is that this file is created in the conf directory when I think the data directory would be a more suitable location +So all issues not directly assigned to an AS release are being closed +I committed the PMC updates a few minutes ago +Once your mirror has consistently been updating for a few days let us know so we can check the logs confirm and then add you as a mirror assuming everything else looks ok +See problem does not exist in the jQuery branch +Patches for fix was published on assignation change and fix version linked Bugzilla item is unrelated. +wonder if there is a workaround for this issue +What exactly does your corresponding invocation look like there? can you please try to reproduce thisd be willing to share a source tarball privately if that would help +The now uses the and to read data and calculate splits this makes it alot cleaner +Note that we should implement this in such a way that the user can determine if the variable was found or not based on the return value +Fixed w the release. +This patch fix this bug +Fixed +Resolving +Signing this issue off +I propose to close this as wont fix a continuous integration deployment can be done in many ways and I dont think that we really achieve anything here other than having more code to manage +Client is one area where we are lacking but Im so busy trying to move things along on the server side and fixing bugs there +Ugh indeed +And the tests were simply disabled by an Ignore that somebody temporarily added and forgot to remove when updating the tests to style +Added a unit test for +Also checking they are different doesnt seem to be sufficient because the name could have been cut because of the length limits +The first bad build comes with revisions meaning that caused this +Patch applied. +I tried latest nightly build thru Apache Flex installer but still dont have dpi bucket supportI saw from dev mailing list that this was finished in a separated branch a week ago +Lets retry m running this locally too Enis to see what results I get so we can try and get it in whatever hadoopqa is up to +Upayavira whats the status on this one +Is this an issue with the new zone +In all the other cases it should stay to current value and hsync resets to false if the value was set to true +includerw include +Let me know if you have ideas +This issue has been fixed and released as part of release +Moved due to waiting other issue +All of that markup is being sent out with every page request +i tried with jdk but didnt solve the problem the weird thing is beta works correctly with jdk +ah ok +That helps clear some things up +Jody if you look at the logs the size of the requested image is taller than what is large and the returned image abides to it +contribute to Apache Thanks +Lets investigate into this issue more and release without resolving it for now +That would explain it for me +Please feel free to take it over if anybody wants to help. +Provide the patch +Having now filled out that form I dont expect therell be any issues +Alexey everything is correct but the Exceptions strings should be internationalized +Tried to verify this using a cluster of size but could not reproduce the errorWill have to buy a Ethernet cable this afternoon in order to test with two hosts +However firebug reports content type of PPR responses as texthtml +too late to add into moved to relates to SOA so I will leave it with you +job file +Thanks a lot Cedric Ill check those issues. +fix the calling sendrr in need another integration +Work in JBT for this think this issue is kind of misguided +In the meantime having this knob is better than no knob at all right +Have you change something or modified the code +Why dont we kill before we sleep on WAITFORSTOP +address Ashutoshs review comment +Done +DmitriyI agree +Do not JIRA issues resolved against released versions file new a new JIRA instead +closed +Yeah itd be great to have a test and if it were at all possible this could cause a data loss Id insist on it but peoples time is limited +We have a application that is affected by this issueThis issue breaks the connection pool when all connections are used and a thread times out and triggers release of connection associated with a stale thread +Got it thanks for the explanation +Im using the bundle version available in the standard EclipseAs its a class loading problem due to the ASM way to work with COMPUTEFRAMES I think Ill have to provide a that could look into dependencies output folders JAR dependencies etcIm open to any other suggestions. +We cant fix it in our side +Tests still fail but they seem to be related not installation related +In fact much of this was copied from the JSF module to build out the structure for beanval +Downgrading priority after further investigation +Is the change in really the right solution? A better question is do we really need to have the copyright headers in the test files? If we dont then we should just change the test files +ping for a review please +but of course a test case is missing! And we should test if the issue is present on as on and +Update since in with in Linked task +Interesting I dont see a failure in there +reopen it to set fix fix version was set wrongly reopen the issue to update the fix version +this probably needs more vetting before we jump and make them public +I verified that the patch resolves this issue in our application +Thanks for the background +Apply the compile and run the test and see whether it changes anything +Rototilled tree under javaengineorgapachederbyimplstore at subversion revision +Thanks Dmitriy for committing this one +But since this issue does affect also Tika users lets leave this issue open to track progress. +I guess we need to check for an active context before running the it would be nice to be able to disable security for the current thread +I am not saying it doesnt belong in Commons or Commons Lang but perhaps may be better suited +Indeed it makes no sense to catch the exceptions and report them in the pipeline result +cheers +Sounds interesting + to trunk +So Imattaching a patch which changes the logging level from warning to infoThe text of the info message has also been changed to state just that itis Running without URL template descriptor +I could not find any more specific testing around this either but put the new test in an already existing class +However I think that it should not be seen as the indicator of success for the setup or initialization of the Connection instancethat property socket is not null +r Punting on the unnecessary else fix. +Thanks Suresh +Thanks Raghu +Looks good +My problem is solved. +Just to note this is probably an issue as everything works correctly with RVM +The attached +patch for the unittest to clarify the for the test Frank +is in +Is this doable +I think this is close to commit +Successfully deployed zones with local and NFS primary storage to the point of instances then rebooted KVM host and everything came back up +Hi Dan I had got a MUSE zip file corrupt +I just committed this +from me for clarity +moving to BETA in rev +Sigh. +I have been looking at this test and it does not seem to be related to my changes +Thanks to Matthias Kster for this improvement patch. +So this means that any plan that is sorted on OR fulfills the order by requirement to be sorted bySo in the original queryplan it turns out the outermost index scan isTENKTUP using index TKUNIQUE which because of the predicate is enough to avoid sortingAgain the question is whether the code is meant to catch this case or did it get luckyOriginally I have to admit I didnt understand how any multiple table order by could be a sort avoidance plan since we have no indexes that cover multiple tables +Patch looks good +Integration tests were updated +This is a bugActually its a new feature that doesnt yet workThe workaround is to put your split implementations on the classpath of your tasktrackers +Maybe we hit another thing Christian +Thank you! +This patch includes the missing which is expanded to better test the deleteAll behavior and it passes now +fixed in you sure this is fixed? Im still facing the same issue with serializerI have the following error in Nexus The artifact xalanserializer and its remote checksums does not match in repository central! +I tried hard to get this to work with CXF in a similar fashion to native but every move forward exposes another issue +Ive started to do this and also plan to use it workout how to do has now been done +The problem i was having was i was running this with the wrong packaging type +probably easier for me to build off of SNAPSHOT hacking the to use my line class +In fact before the code of this issue was added attrstreamsize showed null +Would u please mention your start and end memorymine was mb and mbis not bad +This is a prototype implementation for handling large uncompressed streams +Any progress +All solr related unit tests do not delete artifacts after completion to allow the inspection of results after the completion of the tests +Moving to component deployment +In my last commentin Instead of +Hi JacquesI tried many of the command line arg suggestions and flavors of policy file on Geronimo and I duplicated your results as well +If we have an idea to use input stream to identify the operation then we can do this inside the +See +From +The best would be to set a relative path to this file so it dumps the file into the varlog directory +What is needed to get there? As for the examples on the site that too is something that would be good +It seems like it is more generally useful than just Behemoth or Mahout and fits well with what Tika does along the lines of Tikas command line tool +btw since the project listing is located in the project group detail area it isnt necessary to have the project group column in the table. +I have also included in this and tried to include but was unable to get to work +Thats different from testing for regressions but maybe I can do both +Bob yours is a separate request +Ok Ill look at this issue asap Im a bit busy with the other work at the moment but Ill do my best to get it fixed for is a sample maven project to reproduce the error +I know about a workaround because you can have in a war in JEE but this is not what I originally wanted +Doug heres another patch incorporating your comments +Lets trigger the precommit tests again since the queue is short +But at present I dont have an idea whenBest JohnThe patches of this issue are fixed from camel and cxf separately +If styledocument then the body of the input and return message must be fully described by schema root bare means that the operation takes input and has output +Attachment has been added with description The AllI agree with Jeremias +In owens table I assume is a backSLASH not a backquoteMore substantively I think I would rather see an escape for the zero byte perhaps imaginatively enough added to Owens tabledk +The schemas look good now +Fixed issued above so ant main and ant work +Mark looks like this was a false alarm according to your clarification message on the email list +is this still deprecated? the test is in trunk and passing +You can choose the way to startupAs for the log size growing its another isuue and I remember Kevan opened a JIRA for it +If you look at the patch this is what I recommend use forward slashes +We have the same problem as MM +Thanks for the patch Glen +Do you have the version you are running installed? Try a plugininstall the patch should use mavenget instead of the deprecated pluginVar but Im hesitant because we have some initialisation issues with that at the please take care not to mess up the formatting of existing elements +In addition simply depending on modules would not solved the problem identified in the issue +I ended up creating an branch +We should add this soon but after +Here is the list of missing AT and OC found inattributetype nis nis nisDomain automount automountKeyobjectclass nis nis namedObjectI suggest to create a new schema containing those elements +Provide a unit test if you think theres still an issue +Verified by Nathan. +Im on the patch for and trunk +Fixed +mail sent to with proper contact mail. +settingssrcsrcsrcsrc +Completed in revision +Um wasnt the main problem that ftp HANGS not the +Please provide update on your items as a comment +This issue is related to which will be fixed as part of RC so Im moving this issue as well +Then the Invoker can check these +bulk close of all resolved issues prior to release. +puppet. +properties for is done no +I would check if your stylesheet is working correctly with your XMLinput that is producing a valid FO document when processed through your XSLTprocessor +Ill commit this shortly +Let me know if you want the code for it +It is allowed in so far as for a given Indexed class we look up the hierarchy to find Field properties. +I think this issue is not relevant any more +Tested the various communication security classes and version incompatibility errors and updated exception message propagation for better user errors +It reads a at init +User was able to get me plugin working using different me configuration. +to branch +A primary purpose for Avro is as an interchange format +I see this means we should remove the method check and only use relogin because it is redundant and confusingNo need to change the patchCould you please file a jira to get rid of check? for the patch +It build fine on my machine +Ah no its simple really you need to ensure that anyone can buildthe site for any purpose you built into it +Closing at reporters request. +Ive backed out this patch in r since it causes failures in Ill raise it on the dev list +Stepping back a bit +fixed in beta +Fixed in revision the bundle now starts on Sling launchpad Note that this wouldnt be needed for trunk as the bundle exportsversionBuilding Sling from trunk should work if not please give us more details about what fails +Ran and +This was turned on in the originally and then we had to turn it off due bugsThe bugs were fixed a good while agobq +Can you update this ticket with the duplicate issue +I just committed this. +The patch is against right now but Im starting to wonder if is not a more reasonable goalThe basics of the patch is that the CQL timestamp type now default to that new type +A couple of comments Using java I have not been able to generate the expected output using the option Using java it works as expectedSo I am also inclined to say its a Java problem on the Mac fixed with JavaMark +For the startup script just use the argument to pass in arguments to the JVMsu USER c usrlocalresinbin verbose ROOT conf ROOTIt seems server works +Sorry about previous post posted to wrong jira. +Other issues need a clear description of an issue to address. +I tried the approach you recommended before I used the +added new distribution policy and changed the API to make this simpler revision . +Thanks for looking at the patch DagCommitted revision Yes I intend to do something similar for the replication tests but it looks as if more changes are needed for them especially for the code that supports running tests against remote hosts +Hi Would you please try my patch? ThanksBest Regards Jimmy Patch applied at r +Requiring a static marker file breaks the Servlet spec +No unit test as this is very difficult to access in its current form +If theres something youd like to improve for ORM can you submit it as a pull request? +But the distance computation does not exploit the sparseness of theSuppose you want to compute the distance of two vectors both of high dimensionality but with the majority of entries equal to zero +Hi Junping I committed the patch to trunk and tried to merge it to but it failed since there were some compilation errors +Reopen the issue if the mail isnt showing up for some reason +bq +I am working on it +Thanks Joydeep! +no worries +Verified on. +fixed +Dan suggested I try to make a write up of the current behaviour and what is being modified +May also be an issue for events which are immediately actionable possibly in case of streaming +Because SATA controller is default since the bug report is correct and needs to be fixed working on it +Or you could try to getaccess to a bit Vista system +A separate issue would be good +svn ci m Moving center left leftPad right rightPad mid upperCase lowerCase and swapCase over to input +Hey I even gave patchesCan somebody commit these please +Then view the page source and see how it was done with CSS alone +My guess is that most users would have upgraded to the new channel code but there would still be a small of users running and we need to support their data upgrade +Hey JohnWelcome! Its certainly okay to tackle this issue +reason you are using notCan you verify if this occurs in +This has resulted in many tests getting stuck at shutting down zookeeper and some times the port is not freed by the previous zookeeper so the new zookeeper never starts +Glen thanks +bq +Works in cant replicate in either +While I developped a big chunck of the scripting package Vincent was the onethat developped the security part +please verifyfix and close +that could be a problem +Thanks Namit! +This approach would be easy to use indexes that RDBMS provides +Dyre and Rick thanks for reviewing and committing the patch +thinking of ways to factor out the code applying the rendering transformation and wondering if creating a new type of layer that holds the sld source marking it as transformed and offering methods to run the transformation could be a good fit +Catching the and printing a log message when the race condition dont think the fix is right +bq +The ability to externalize the request mapping would be of great should be the focus not necessarily the definition of mappings in could also imagine a approach similar to s or confroutes in the Play instance if the framework separated the mapping configuration from its execution and provide a convenience configuration builder to be used from XML we could then just roll our own mapping configurators as needed +Attached patch incorporates Aruns review comment and checks that the returned task is a setup task +Excellent! Much obliged! +Resolved with application of patch f from BJ Reed +and files modified to reflectthe order in conf +put in critical to make sure we figure out what and how much needs have migrated to extension pointbut we have to migrate actions to handlers as well +Note that the assertion of the data store operations would have to be written for cassandra +The failed test is tracked by a separated jira that filed before by Eli + +bq +in general is more efficient for lookups but has bigger space overhead than +But I feel that if thislibraries need to be in the directory then they should be movedCheers Jrg Sorry for the my mess writing your nameThe question isMove both jars to libendorsed or notI am on moving them to lib go on with it +The only solution here is to fail after a certain amount of time or to completely rewrite the resolver +Yes it happens without the profiler also +Brett I created. +QE can you please verify this issue and also establish what browsers are affected by it +Could you please have a look at it +Changed everything to use the default timeout mechanism +Ok for the comment tagsBut why the response of the Ajax call is only success evenif there are some errors when we tried to validate Form in using jsonValidationWhen there are some errors the return of the Ajax call should be error nor could I ask my question to the user mailing list? where is the link +Yeah I checked with arc too +The current encoding algorithms on github throw off a lot of garbage which tends to indicate that theyre not using the cpu cache very wellSome data sets would be able to disable gziplzo compression and just go with prefix compression so you save some cpu thereIm working on a benchmark to compare a few more factors including compression ratio and sequential read speed +I rather suspect that XA transactions simply cannot be fully supported on to begin with as long as there is no full XA support in itself +Thanks for the patch it has been appliedOliver +Awaiting a last check so this is close to being closed +shx and +Could you post your comment on the mailing list +Isaac before resolving this can you include this in your release of docs +bq +Any way thanks for the patch +It seems we are ignoring a critical error message when the server is not present +The main changes are Reverted changes to and the mavenant plugins because well always generate the callback API +Mamta I think that a new topic is needed to discuss this behavior +the test is to verify that the patch actually fixes the issue of not having defaultschemadefaultcatalog the test must pass otherwise something is wrong with the patch or build +Kindly review and let me know if it is acceptable or if there is a better way to do the same +Integrated to trunkThanks for the patch AlekseyThanks for the reviews Anoop Nick and Nicolas +So if it is can you update this issue with some more details and we can see about helping you out +If we are introducing a mechanism that allows for unregistering a listener I wonder if we need the Soft cache +Im sure someone on there will have more insight into the problem and be able to help you out. +Heres a patch with a proposed READMECan someone check the one line explanations for all the demos to see if I captured their purpose accurately +verified by me and Xavier works for both of usMax please approve the looks sane if name is entered and it exists as app it will load data if name does not exist as app it does not tocuh anything and just errors in validation afaics +My wild guess is that the if condition around that line is wrong or does not cover all situations +This is a to +to review +Please make sure this is the last change as it does get a tad tedious having to redo something you spent quite a lot of time doing before +I fixed the compilation error in it and there was no crash +But they care when things stop working +Its more precise and shorter than noreturn +Reopening to be able to edit the release notes text +And by here I mean +This patch adds back the import cleanup that allows a false dependency on in to be removed +Alex assigning this one to you +I thought so when i was not able to reproduced it twice +After a good clean and redeploy the rule in worked for both and +so the separation is used albeit seldomly +But between I am using the Struts the handle of my Ajax call in Jquery is success +Id rather change the shell commands to not need start code and have the port as optional +This is a solution so that existing hdfs installs continue to work +The gold rule should be use UTF +A condition to check for local or global transaction manager was +source code in +Committed to trunk +Patch for the backport into +This is not an intended drop of Would you be able to create a patch with a profile for +Issue was with wrapper script not CLI. +Richard patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +The EJB as well as the provide the same transitive dependencies +The Total Documents Processed field in the raw DIH output appears to go missing when it switches to the second request +So I think it is not necessary to create a separate item for this. +I am going to check +reopening this issue so we can resolve it in code back to resolve this issue. +lock rows +Its a little more overhead to have separate objects but it saves you from having to inject arbitrary special names into the handlers or worrying about default implementations +Ooops true I didnt realize that +Scalar support +we have to discuss how we handle dynamic lookups since the el allows even more constellations. +If there remains only a few characters to finish a statement I prefer a longer lineWhen the line is really going to be too long I always try to figure out a layout that will not be too unnatural +fixed the few things Konstantine patch is not merging cleanly with trunk +Thanks Arun! +It would also be very beneficial when using a remote logging host to compress the stream to the other host +But still yellow. +I am apt to give this a WONT FIX unless there is a solution to apply here +Jukka Zitting added my user to the incubator group. +Ok should be fixed nowThanks! +No integration tests +Aside this is one case where having a version attribute outside of the namespace would have made things a tad easier +pnsslinit can then set the sslpeerhostname field +But I agree that it must not be part of the component +I think did not exist before version +Then the race would only cause a redundant chmod +I removed the execution count as part of lastExecution is still there though but could be removed +Tested on ubuntu and fedoraGeorge +Install the tar first then the patchThis patch changes modules so that itinjects the icons into the appropriate places within the generatedAdditionally the patch changes the XML so that fewer icons are usedand disables the look and feel extensionsThese patches resolve this issueThe Looks look and feel is licensed under the BSD license andcould possibly be used +Upload a diff file for patch hope this issue be fixed more quicklytks +I proposed an idea on how it could work and if they are positive about it Ill go ahead and write the plugin + PrebenWe would need an bundle of the Splunk JAR so logged as then our build can rely on this JAR which would be in maven centralWe have bad experience when we have too many rd party maven repos which may not be up so ppl cannot build the code +AS no messages displayed AS messages displayed as expected +Works for me on both Mac OS X and my usual Debian machineCan we commit a patch that turns on Wall Wextra pedantic and Werror assuming we take out Werror for releases? +Good I will try to ponder this solution a little more to make sure it makes sense before committing it. +I let a script run for many hundreds of process launches and it was result every timeCould be specific to Linux since Tom could not reproduce either. +Thanks for this advise +Thanksbq +anything except recalculate bounds +Im on this +Its wonderful to finally have support for this OOTB +The fo and itext sink should take care of this +Attache a patch +CarlosYou are right +trunk Committed revision x Committed revision +Ive attempted to resolve this issue +The difference is in line endings +Maven reorganization finished. +This replaces the existing funtionality +Chekcing if PING as well need to check from last talks with Frank we are no longer using PING we had tried using PING in the past and it was a hack +Thanks Claus! Ive updated Confluence have a look and let me know if it looks OK +Committed the proposed patch in revision +The expiration of the certificate only prompts the user attempting to hit the URL via a browser and they can choose to ignore the warning and move on +Verified in EAP ER. +I added you as a contributor in jira +If you can reproduce the problem please attach an example of your code and describe the environment youre +The dependencies should be part of the and distros +The style skeleton provided is not validating on purpose because its empty +should fix this +Anyway I gonna take a look +Feature merged into trunk. +Im not insisting on nested conversations for natural ids +SorryDONECommitted revision +We said this was out of scope for CR as we would have to pull the property default value into the interface and this was not going to be done +Have you triedlifting restriction a so that the background cleaner is allowedto clean more than one page? Is there a reason not to let thebackground cleaner itself +Fixing this in. +Give me some minutes before starting to compile I already see what I have to do +Need to wait for things before we mark this one as resolved Itll be good to know if Peter is not seeing the issue Paul is having some issues with working on his Windows machine +We specifically publish SQL scripts for this reason +Heres the file with my basic Ajax Test example which doesnt work correct +note works fine under has been added with description fo file causing transition to closed remaining resolved bugs +defer all issues to +please dont do it this way fix hibernate tools instead +no we cant because Im sure some SCM need a files list but for svn we can use targets argument on svn ll fix it in in +If this is as easy to reproduce as it seems we should be able to track down the actual cause +patch with refactoring changes only +mdo files which isnt going to happen until especially since its not a ll revisit syntax for and apply lessons learned from then +Fixed in rThanks Marc! +this one is strange +Thanks for the simple examples +Maybe try to run your test on and see if it is a problem there +Maybe someone else can help with thatIt all depends whether you actually want to learn it +Oh I messed up +Thats not to say there is never a reason to do so +You never know what a user will choose to do with our classes +lib you see if it reproduces with all machines on u +I do see Keycache working in my tests +Will check if I can reproduce it. +Robert It not about me wanting it or not +My only question is whether this pattern exists anywhere else or only in +from JucaIt still happens in Sybase with the following message from the databaseALTER TABLE Version failed +This issue has been very satisfying +Attach version of the functional spec reflecting the fact that these restrictions are only enforced in GRANT databases +But Esc +Not sure whether there is any general solution to this issue +I will look into adding a configuration property for this variable so that it can be easily changed externally +WDYT +I can put in config variables to disable fsync if needed +It looks like there are also timing bugs going on +Looking at trunk there should not have been any wait since active threads is right +So I revised the test configuration and found out the numSegments attribute was set to +Closing all old Resolved issues +Theres little chance of getting this into the modeler indexing logic +I posted it originally in November so its entirely possible its out of dateCheersChris +resolved in r +PFA changes +I tried to verify this behavior again but it did not occur anymore +I dont know if that requires a much full featured test framework ala Local Job Runner or if we can fulfill that requirement with +raised its priority as a prospect was asking for you can now set the according preference in the editor preferences +verified fix using a local build which includes directives from external +Hi VictorThis is fixed already +Already had a quick look and just one question +Hi well that sounds promising +The transformer should probably just log a to fix the test case I dont know at the moment Ill have to see exactly what fails +Please confirm that this issue is now resolved +Verified +applied. +We are still trying to pinpoint what the exact problem is +changed pgas code published new binary tarball for bgp and updated Makefile to pull new pgas binary in +Initial commit changes are needed in +In connect we send a sync call across the cluster to update a service view +This looks specific to the because we dont create a bean for the +Seems more like a language feature than a bug fix issue +cant edit this wtf? I hit the Create button by mistake +We also shuffled the paging stuff around a little bit +Made some modifications make column typed string and others +If it was set but got skipped then its just as if it was unset +Thanks for looking into this +assumed not done for beta punting to beta +Im trying to put up new RC for +The patch does not contain any author tags +I havent been able to find the exact JIRA where these changes were madeIf this is still an issue please. +I think eu had a crack at this +Moving the semaphore update to commit +Thanks to Fernanda for fixing the problem and to Bryan for reviewing the patchCommitted v into trunk with revision and into with revision . +This is a start at addressing the immediate issue of updating Avro libraries there are only three Gora classes which will not compile +After the host CS worksI will close this issue at firstMy colleague will test the upgrade from to to see whether the wrong rule will appear againif it appears again I will reopen this issue +As long as we do not regress on the current Hadoop I think we are good +You can verify this at your end +looks good to good to me too checking this in +My application is not that complicated but it has to have a database connection and is restricted to those that login to the site +Tomcat +closed for am seeing a similar issue when using from Ant when specifying the freemarker in theWhen Ant is invoked from the project directory the freemarker value srcmainresources is appended to the project directory but if Ant is invoked from directory X the value is appended to the path of directory XIt seems to be using the working directory of Ant +Sure but can we issues like this in AS project do you think +The error affects version +Yes thats exactly what Ill do tomorrow just need to make sure I get all refs out of the build scripts etc +Niall I dont agree with you +Without the patch only clients will work +Patch updated to looks good +Attached is the picture associated with this defect +Thank you for solving the issue Roshan I clearly missed the point +Hi guysI am attaching the diffs for the annogen fixes in order to gen inner classes to work +To be assigned to Yash Talwar backup Brian have tested this defect +HiI encountered this issue +Enjoy +refine the patch +My colleague Daan Hoogland mailed about this on the dev list a while backCheersHugo +Hi Joe and JustinPlease see my comment on do you know what I did wrong with svndumpfilter? Im happy to fix but need a pointer or Cheers a few this is now complete. +Ive applied both patches in r +Closed as difference +HiThe double converts the double value into IEEE double format bit layoutFurthermore the XOR operator prevents returning for values less than This is the hashCode function used by standard java implementationI was noticing unexpected delay in one of the operations related to double data typesAfter some debugging I realized that the puts and gets were extremely slowThat pointed me to the hashCode implementatoin in which turned out to be the cause of slow IOThat is why I propose to use the standard java implmenetation of for double typeThanksVaibhav +I just committed this +Committed to trunk and +branchesJBESBCP +This means Jesper and his team need to figure out and provide the corresponding management operations or hooks to verify the connection is configured correctly +This has been fixed with the r changes +The old obsolete faq page has been removed +Report your findings either in this issue or better in +Closing and reopening is the only way of getting The hanging tests is what I wnat to investigate but the problems I am having with my Eclipse installation is a bit of a blocker not technical there is clearly a workable workflow but in terms of my patience enthusiasm and energyI guess my biggest problem is I am not sure how to get useful data that might give a clue as to what is happening let alone what is wrong +The services handlers are not added to the pipeline in +was recently released and there are just fixes for so far +I think it is essential that maven produce a jar that is functionally identical to the produced by the ant build including the relative paths for all the classpath entries +Thanks Yongqiang +Since no longer has a stop method we have to do the check on execution lower down the stack +Switched attribute names to horzOffset and vertOffset +Waiting for a little bit of experience before to close it +I have committed this. +The only change I see is to use the version of stax which doesnt version the package +I think we should address that in a separate jira +I also want to add new method to where I can pass a HRI and a for the case where a region is out on a RS but not in +At the moment the test from patch are failing because of thanks for the patchThis is an API change so we cannot backport. +Actually all that is really needed is to make the state names in a flow unique +Bulk closing for +This bug is on the list of bugs that I am going to fix before the next release with any luck but definately before +As for the classifier it should be fine to set it to null since its optional +modulessrcmainjavaorgapacheairavataxbayastreaming +Should the config values be expressed in MB or GB or KB or just bytes? MB sounds good to meThe other parameter we have in hadoop related to memory is which is specified in KB +Unit test included +Regardless of what the script does the user still has the option of generating a certificate themselves and setting the configuration properties manually +Results are the same Please wait still loading as is the result if I bring up a Roo shell inside STS +fixed stale resolved issues +Switches default resolution mode to lazyNow there is a remaining bug for calling lazy resolution helpers from geninvoke +Thanks Bernd +It is no longer a problem +Ok I integrate your doc into the book +fixed in r +This must have been fixed along with the other problem. +IMHO we should try to avoid throwing as much as possible +We will implement this distinction in Processfile and run different decoder functions accordingly + +Rebased about the delay in looking at the patch +Not approved for EAP +a lot Savas! Will try the snapshot in the morning +Thanks Uwe for committing this +Can you move the so that it is under the module? Remove from the file and remove the call fromOtherwise the patch looks good +Downloaded the tool and generated theseMD CFADBFEF progress +Im personally doing all my development and testing on linux and using Ingres +solution works +Merged fix into branch in revision +Thanks Raghu! +Since we are now using new Create API to force sync there is no need to enable the sync any more +In this method the fieldStream object is writing the field number the bits which sets the field type and the data itself +Please report any broken links that Ive missed as individual Jira issues. +So whatever the problem is it seems like it was introduced by version of the plugin. +Making it a configuration value will require changing oozie settings + +is gone from the language so no longer relevant. +Thats why we fixed and thats why Im closing this ticket Wont Fix +All please verify that it solves the initial problemIt should also be mentioned that in spite of the fact that Hello World works ok now on clean Win it is still impossible to run graphical apps that use AWTSWING on such machines because our implementation of these libraries depends on even if it is built in the release mode +Trying to edit +Thats because we do the zk transition to open before put it in the online region list +There is no such constant values created nor automatically assigned +The point of contention is whether to stop the retries if a competing is detected to be inprogress or whether to stop the retries when a competing has successfully completed +However the problem itself should have been fixed +Thanks Hardy +Regression of +Heres a log excerpt showing one of the regions that got would not handle any regions on the dead server if they were already in transition +Non shared transport is an easy case +I would like to ask you to open a separate Jira issue for that +Logged original exception if rollback fails as suggested +bulk move of all unresolved issues from to +And when I say dog slow I mean they were double dog slow to an unusable extent As far as I remember it had something to do with the piped streams jsch has to use internally in case of stream based operationsI dont know if this has been fixed now should give it another try +closing this for now +Hi Richard thank you for your comments +Same as release randomWrite works for me in G of heap +Added patch for the same test in trunk +Surely we need to have a good story on how to get app info after the app is done +Many installations of Atlassian products are used internally and do not require SSL but do require secure reasons for implementing this feature are pragmatic and have nothing to do with compliance per se +fixed in revision Mark please verify it fully resolves your good +deeper changes would have to wait for imo +I was using ll try it one more time with a mvn clean to see if I have the problem again +committed +Still +Thanks Vinod +I have a good idea of what has to happen here especially for Windows +Committed to trunk at revision +bulk defer of issues to +Ill use this workaround until either a patch arrives or someone has a better idea +Sorry for taking that long +Thanks Adrian for reviewing +work is completed with revision +Another problem for you Russel I know you likeem +Secure bulk load would not work based on current hdfs branch bq +looks like for I needed to do a mvn clean install before mvn site but other than that looks good +If there there is at least one java library thatsupports it then it seems an unnecessary API complication to split the create into two +Nicolas I do not think that this is good approach +Then when in lazy seek path it need to construct two fake key and more compare while real seek path only need one fake key and less compare +Please let me know if something is wrong +Yeah exactly +Can you please point me on the places where its broken? Thanks +I will now this issue +This issue needs to have a component set +Patch works fine +This makes sense though since it is often needed to skip validation in order to apply the rastersymbolizer extensions +Also no updates here in a long time. +any comments +m reviewing this now +Once we made sure that the patched version was loading everything has been working just fine +Please do not set the fix version yourself +What is the status of this issue? I see that there have been multiple commits +Trustin could you please explain further how one could accomplish supporting multiple allocators in the same running MINA app and still use the static methods? Using different classloaders? Using will be hard I think because of the asynchronous nature of MINA +This way it is cleaner imho +i got this exception a number of times before when i started a numberof repository instances on the same repository configuration +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +my humble contribution to the cleanup use CFS executor for +Thus reopening it and rewording the issue description to match whats currently missing. +As you said I commit this patch without passing hudson. +I cant reproduce this problem on either a normal blog or a blog deleting an entry always causes my caches to get propertyly cleared +I should note that i didnt suggest any course of action because im ambivalent about the state of this +Would this be different than plugging in a distributed cache that just happens to be distributed across multiple running in the same jvmAs long as your replicated data has the same primary key on each shard it seems like this would work great +I couldnt quickly see how to convert the string to an URL and get the escaped version to pass into File +Oops +Thanks for the suggestions +Thanks Owen +Whats the ip. +Added various timeline and mentions methods to. +Hi NicolasIm working with Jan on same project and face same problem +With Roland gone I do not see a way of coming up with a reliable fix for the problem without removing disabling the connection garbage collection code +But I am still puzzled why this has started to show up recently +finally closed thos one +When I realized how it worked my idea was to supply another parameter in that lists the that are to be used +But I think we can do that as a separatesubsequent change some time in the future +Emitting the document into the view is not always stupid +Changed config options from nonormalize and nofilter to noNormalize and noFilter which is the same as some other Nutch tools +I dont have an issue with the patch but perhaps someone else could comment +Thats why I tried to recreate the problem with the echo sample +Can you add comments on why you need a +About one out of ten times scsiadd wouldnt find the new drives and wed have to reboot the boxThis was years ago or so so things are likely improved by now +Do you have multiple versions of the Oracle driver in your classpath? One at the application level on at the system class loader level? It seems so since the different seem to pick up different versions of the class +Seems arbitrary and like a pettifogging change by arrogant devs +Yes you are correct +UGI and SPNEGO will now instantiate a that explicitly uses their respective security configurations +I hoped that would work for layer groups too but it was not checked since it was not part of the actual I was mistaken in the blog post +Useful as they may be theyre not a solution to a scenario that causes corrupted databases +Sample removed as it s not ready for release +Updated patch with some unit tests plus addressing Rams idea above +My version of with clearup method added +do not allow deleteput method for invalidation and committed thanks for the patch! +You need to change the client protocol version +updated +A format parameter can still be introduced later to transform the empty strings into null when parsing. +I have reviewed the patch and I believe it resolves the problem +png file with the same name as the process is now saved in the same folder as the process definition +From section of the JDBC spec? JDBC API changesThe following changes were made to existing JDBC interfaces? ConnectionAdded the methods createBlob createClob createcreate isValid createXML getset +Please feel free to open and Ill promise to be very attentive +Already seems to be supported closing issue. +We should document this for at leastInteresting about +But I think we can move the javadoc for the use of old api to +But renaming the class to makes sense right? YepOf course I was never quite sure why it was javabin and not just bin or binary +Thanks for the patch +The Hive build currently satisfies its Hadoop dependencies by pulling down a complete Hadoop tarball + +Better to do this before the release I think +Hi James any update on this issue? Were discussing cutting a Karaf release soon and Id like to know if this issue may be worked on resolved soon otherwise we may push this one out to +weve added the taskcleanup thread long ago for precisely the same reason +Build with and you should not see this task run +Dave was released Aug and should have the fix +Im using cygwin too +In the long run if this becomes popular it will of course generate a lot but at this time considering that people would just be curious about it and be experimenting with it it seems to me that creating a sub project is an over optimization +Merged to doc branch at revision . +The NM should stop for REBOOT also this will solve for now +You can only create EAP runtimes +and +Please svnadd it tosrcmainjavaorgapachewicketexamplespub andsrcmainjavaorgapachewicketexamplespub foldersThanks +I am running tests nowWhen the table functions is created and bound we now stuff a into each +Im currently using dbcp but have no problem updating to a newer release +I recognize that the original comment says that this is happening in your production app +Resolving the issue +This is the part covered by and second issue certainly in relation to the WS usage is that it is being used in a light weight manner when it is heavy weight +integration tests are passing so this works now. +This is not a problem with DR I am closing this jira and creating new jira for test failures in this as Out of Date since we didnt really fix the regressions yet +This was due to not setting the property and only using the default value of +Trivial patch attached +I guess there are scenarios where adding a stored method to the class path to get this info may be awkward +Id be happy to see in code what Noah suggests down the road +This reproduces also on the intercept mock to endpoints that didnt support the async routing engine properly +Addition of algorithm methods with an Iterator tests for newly added Dusans changes as a patch as its hard to work with old copies +but I believe that it also adds a buffering +Thanks Paul +In order to keep up with upstream component upgrades we have to import upstream binaries the ones that we dont have time to rebuild will be todo jars +Where I think this really belongs is on the new menu weve been discussing when hovering over a service or reference +I took a quick look at it looks reasonable to me wed use this option only while actively debugging I imagine +As I said if we can wait until Friday then Ill get right on it +I will fine out a change to implement itBest RegardsChatree have no pb with an iframe do it as it seems better for you looking forward +I committed the patch to trunk +Shuai Please see my comments on phabricator +You example uses a select which does not have a reduce phase +to make the switch we need to make your svn read only also you ready for that to happen +Created a component for this resolver +Youre right it works but as youve written it would pollute with messagesWhat we need is a solution for libraries to be able to contribute global default messages after the clients application properties +Is this is a bug in? As far as I can tell no released version of Ruby had this bug but it did exist in trunk for a while +This issue also exists in the remoting codebase as our class is a copy of theirs +Thanks +Will commit sometime tonight +Hi GavinI was able to get ortpassw running howeverjuanpablo sudo ls la jsPasswordjuanpablo is not in the sudoers file + + +Pushed the fix +I was just suggesting we make Cocoon less painful by defaultAdmittedly this is a bit of a hack but its not that difficultto understand and trust +The disclaimer is no longer needed +will revise if necessary depending on how goes. +They get executed when the engine sees them because they are supposed to have the ability to control the generation process. +patch provided did not have the proper fix uploading a +Uploading new patch that combines changes from all files +I would like to take a look at it +These patches represent the subproject work to incorporate HCFS functionality +Marking PA again for Sijie +Looks like a fairly large patch looking now +tracked in the toplevel issue and resolved except of +they also talked about bundleizin dmr +You still need to use SVN + OK for LARQ because its new +Thanks for the patch +Lets try not to make API changes at this point +Good point attached the change to +The new chance of getting this in? Im carrying it as a patch in my own local version but this is obviously not ideal +Alex I left some comments on reviewboard +Hi Apurv sorry I lost track of the whole issueBut I had a bit more time to code it to a configurable designAre you okay with this? Or do you have any suggestionsAnd was there a specific reason you picked the snapshot of snappy instead of the releaseAnd I added BZIP Codec from Hadoop. +Fixed the javadoc warning +This last piece will take some more design work + +I might not be able to update a tree and look at this right away +Im not entirely happy with the test case though +On the browser master UI was stuck in refresh +All files and diffs are included in this ZIP package +See +So only permision SETPROPERTY is enough +Sure thing +Just walked through doing this and it is more than a script can handle even in docso going to do a wiki on how to do themes then put a link in the script +actually maybe thats how to structure it +Debug and version show same behaviourI will try to look at it now +When finished reresolve with the new fix versions or label backportrejectx as appropriate +Ive gone back to the EJB group on this one and apparently agrees with me that aliases should not be supported in DELETE and UPDATE +About the constructor I would say it should be changed to just accept the interfaace +code reviewCommitted to andThanks Ramya +looks good +The version here fails silently to rename the new file. +temp file gets left behind it means the log splitting failed and the +Currently its not possible to order results by entity using Criteria API +somehow the Loading message disappeared and we need to put a progress bar or loading tree node back in +I think that your solution is fine +Its now possible to compare views and developers + trunk +The new mojo also allows interactive mode +This mirror doesnt seem to exist. +Alas we cant do this in as wed break custom filter that delegate to any of the existing filters +The bug is not reproducible on Windows and issue is not reproducible on the latest vm buildsfrem and drem instructions return correct results +I will prototype such a thingy +Thank you Suresh +You must set a dependency on a UI plugin otherwise you wont get the appropriate View template +Verified fremarker version in component in Fedora to write release note issue now that release note text has been written +Once we enable a few apps we can then streamline the process of adding implementation to the jdbc interface so that there is no duplication of work +Assuming it works. +Maybe it was something else +The syntax was incorrect. +I agree on the SqlJdbc renaming +Any chance that this will get fixed? Clutters up with the exclusions +Thanks Amar +Integrated to Thanks for the patch RajeshThanks for the reviews Ram Jeff and Lars. +python script used to calculate frequencies of sql queriespython pathto less +I believe this wont be an issue on EAP +Ive tried but this is actually surprisingly difficult to reproduce failure in a test on +Issue closed with release. +Dont log Ignoring call to stop during normal shutdown +Complain to Oracle! You paid them thousands for their product and it doesnt work. +Resolved in incremental build . +for +The patch didnt apply to trunk because of so Ive created a new equivalent one +But there may be those out there willing to take up the? +Use the mailing list for asking questions like this. +I believe you are correct that this is a bug and needs resolved +Is there any documentation on Mumak or is this JIRA the current place to find out about it +v attachedIn this version can run in parallel usingCMsubmitBackground submits task when submitted CF is not currently compacting and it fills up threads with compaction task when there is room +True +That log entry is still valid because we still delete the region from meta because daughters do not hold references +The hang is a cold hang with nothing spinning and nothing in the log +Since is a superinterface of Closeable in JDK we can just implement the s not the which needs to be closed but the +Minor issue to the next is just the way it works the list can point to either some directory whose content is to be scanned or a packed upacked module would need to have a trailing slash to indicate it should be treated as a directory but then the scanner literally thinks its a directory and scans its content which is not what we would be many changes required to handle this special case for which the workaround is trivial so its doesnt worth the effort +Attached r with small changesMade public the methods of dense and spase vectors +They have already been included in release +Raghu could you do us a last favor of commiting this? Thanks +not for beta and as nick says shouldnt remove things from here if tests rely on them +James any update on this one +Im trying to fix the issue together with +Hi JamesIm not sure if its been fixed or not +Why not keep the code in a single place and have the test +Users can always include parsing libraries that they would like to use but we should very clearly document how to do this +Good to go for resolving +This script induces methodnotallowed bug on trunk induces a functionclause in writestreamedattachment in and speculates its an unterminated chunked transfer encoding its missing the final byte chunk for eof +It fails on all versions I have tested including the branch head and master +This sounds like a duplicate of which was fixed in and +Please with any specific instances that you find +Done +Unfortunately I cant send you my workspace because it contains projects for my the plugin compile out of the box from the SVN code? If so I could try to debug the issue myself since it seems to happen more often for me +has the correct file in it +As a consequence the custom jBPM configuration is no longer available to the and the default configuration which does not describe the nonstandard converter is used +HiI am able to download samples from github and run this successfully +Thanks! +from my perspective the request timeout MUST end the request if it is possible to catch this and to go on with the request a request timeout makes no sense +There is always errors when I try to use the patch +good catch Ben +Modified the title to match the investigations made by RyanFix version for Raised priority to Blocker +The first choice isnt always skip sometimes its truncate +There are two parts to the patch for this issue +Needs to be rebased +Thanks for looking into this! +Please confirm they have been applied as expected +If you want accessor support you can try lobbying those compiler maintainers as any enhancements they add will be exposed in our consequence. +These tests also succeeded +Thanks for your information +Hi Jrnany updates for the documentationSince this project is in the UIMA sandbox please provide some information about your project on the UIMA website +If you enter a custom state type youll be expected to do similar logic yourself +Could this be commited for? +bq +Please pardon its being a bit of an omnibus +Bulk close for release +Nit should be + +This should be resolved in with the component upgrades for Infinispan and +Ill download the latest snapshot after New Year and let you know theThis seems to work fine +Will get back to it sometime later this week +reverts moving updates the script +Also it can coexist gracefully with the existing localwriter or with with base +Committed revision Thanks Shai! +It seems like the stochastic load balancer gives the locality awareness when balancing +Uploaded the one pager +added +I mistakenly did not grant license to ASF +please review +So all issues not directly assigned to an AS release are being closed +in Rev +The test fails in my environment also if I just run the fixture +Im in favour of +Added one more test to test update able reader over large clobChanging the exception to in and is a good idea but i will prefer to do it as a separate patch as that going to effect code unrelated to this issue +This will only have effect on validation of parameters no new functionality will be introduced +I can see that a number of other adapters also throw it though when there is a problem instantiating the would expect that any Exceptions thrown while constructing instances should be rethrown +Stefan please check and remove the redundancies +See also also also which concerns the same problem with the properties tag +In addition well need a number of test cases for testing the fix +If this is Ok I can easily add it to. +fixed issue with default +Nothing more to do here Olivier +In true style Id like to that patch until further discussion +Google Collections everywhere +Bumping to +Are you explicitly including the Cordova plugin when using this API or are you just using the API based on the documentation +Switching all issues that have been in resolved state for more than one month without further comments to closed status +If my consumer goes down seconds into the minute I want to replay the entire minute over +Yes I am +I have a patch that covers about doPrivileged blocks in about files +I would try +Closed after no further comment has been given. +No big deal +extra jars were getting included with previous patch fixes jars are also not required at runtime by Hive so can be excluded +Original class was throwing exceptions where read methods were not found +It could take a little effort but we can provide the patched source to get to the actual line numbers +See for a proposal to control relevancy from within a DSLD. +Unfortunately I think its not only related to this part of the work but also to themes +canceling patch to rectify testcase s weird that this patch is against a jira tracker called Zookeeper logs are filled with INFO messages since it isnt about that at all +Pull back if you disagree +Hi grantedCheers +in just committed thisShould I apply this to the branch to keep things in sync with +CSDM is dead long live not having CSDM +looks good +Spark patch applied at r thanks a lot please verify +Updated patch that addresses noted concerns except for security manager changes +Tomcat build updated so that the latest modcluster jars in the tomcat zip +Im going to reopen this because I think I understand what Aaron means by roundtripWe cant produce Time objects with nsec granularity but we could at least preserve nsec if it is already present +If possible please give me a call at to discuss this +BTW what versions of Grails and GWT are you using +Migrating bugs marked as resolved before and tagged as version as fixed in RC +The original patch has already been included in CM since +Thanks Amila +orionx currently builds fine in jdk from samples +how many levels up should we consider? If I fix this by adding support for two levels deep should I add support for three levels deep +These changes look good to me +New patch uploaded +This task is mostly done with the exception of some duplicate interfaces that exist in both and +Traditionally weve avoided introducing dependencies on the compiler packages in the XSLTC packages just to allow for the possibility of building a smaller XSLTC though that hasnt been done recently +bq +PhilCould you please review changes +Oh yes it seems that I have been to the wrong place +Ideally we dont want this because kryo classes are anyway packaged in so downstream projects should not see kryo as dependency of hive +Do you have any other changes in mind that you would like to see happen in? I was just about to start cutting GA release but Ill hold it off until all your changes get incorporatedOleg +This change includes verifying that the current imageslogos used work with the new skin +HiWe are migrating an EJB application which uses SFSB remotely +Are your Image and User classes entities? I get the correct output in this case with your sample if the Image and User classes are annotated with +Is this something you would recommend potentially delaying EAP to fix? If not it should be moved to TBD EAP +The release process I followed and I think Robert followed was to tag after a successful VOTEI dont think that is a problem as long as the tag matches what we released +Rick can you point me to the SQL Standard document to which you are referring here SQL Standard part section? Thanks. +Stefans patch didnt apply cleanly for me on svn revision so I itThis patch fixes the illegal XML characters and prevents opensearch clients from choking on that bad XML previously emitted +If it is okay with everybody I will check into trunk today +The patch also addresses way to is hooked in bothers me as it is going to prevent any other from being used +All the matching must be done in the Arquillian container connectors imo +Eh its just hard to allocate the time to test with the various client software there is a number of higher priority things to address +i am attaching a patch for additional junits tests which were collected from this jira entry +This might make more sense for cases where we have two or more waves of reduces +Could this be the same problemIve attached a quick hack that makes things compile but as a user I have no idea if they actually run +Ill continue working on this jira +Everything happens through the Ant build file. +Opening a separate UI bug the issueClosing this bug as shrink is not supported on mentioned VDI formats. +Updatd patch incl +The patch seems to be ok +set location of the configuration to null +So in this specific case theres no way for me to propagate the error information +Please submit a test case that demonstrates thisbug if you are still getting this exception +I think this task is not valid any more since we dont have jBPM in our SOA intention of this task is to implement jBPMs Identity SPI by using IDM project +At this point we know of only the s uuid which can help us determine all the rows in SYSCOLPERMS for that given table uuid +The patch is synced into Camel Fuse this feature need to back port the lookup from camel trunk and we dont want to introduce new feature to camel +Ill upload the patch for trunk next +Yes this is a bug +This issue is similar to am looking into this issue and will revert back with a solution to this issue before I leave for home today evening +Assigned to Saul since he is the maintainer of arcsde module according the file +Implemented simple JCR that uses DNA Graph API +This is not a bug report and should be posted to the mailing list. +I just took the default and didnt pay enough attentionHowever on the second point the reason I want this change is that I am implementing another implementation of +SVN Rev Checked two one using synchronisation one using a concurrent access queue +About access controlTheres no access control at this level +Everytime a new connection is opened the spatial extensions via shared library have to be loaded as well they are not persisted in a single db across connections +Committed to trunk. +but we should also release soon +Thanks. +same principle as the one used with stylecop a new attribute is used in the configuration files +Patch looks good except that should be moved to second thought realized that all classes related to should be part of same package so its fine +Ram You should add it to trunk too going by Teds reasoning above +Removed ops from removed macos from above files and and issue +But what about the server side? Will it work if the annotation is not there? +Attachment has been added with description Corrects Logger in several LM extended to all which use a logger and applied +Can you add a paragraph or two in a comment explaining what the code does and maybe first steps to customizing it for real use +Youre right +It worked on every single browser and newsreader I tried it on +closing this outdated we dont have a clone project action anymore +The Apache equivalent is not released and is codenamed Falcon so it may be a while before we investigate this issue +To do this properly youd need to row lock and make a copy +There is talk to add a local namespace only for the CLI and a var keyword as a follow up +stateless apps are not secure because you need a mirror token on the default crypt factory is now which uses a randomly generated encryption key stored in httpsession +has both and +this looks like an improvement rather than bug fix +Yes that is the best thing to if you dont want any action to be produced by createLink? How can I disable actions from the link creation? +youll have to replace with or whatever the latest downloadable version is +I mean whether it should be a parameter or a configuration variable +So if the core needs to be increased then I would increase the queue size as well +Might very well be that in the end this would be better than the current wayBut even this would not remove the need to handle old application state correctly +Resolved in Savara which is aggregated into this release. +This is the Apache way +The auto restart is a pain after upgrading since I need to merge my and and dont want the node to restart on its own +Absent any evidence that there is an actual pool bug here I am leaning toward closing as not an issue +So imho there is nothing to fixMind no envelope transformation method directly changes the envelope not even they always build a new one +Didier assigning this to you feel free to discuss the issues +Ive tried to verifiy the complete functionality of rereplication module +Fixed in SVN repro +Tommy any update on verifying this fix +Its possible I can convince max and others that adding this ejb jar to the defaults is worth a rebuild but this would only take affect in a new workspace +doing the YARN patch I puzzled over some of this logic I think this could be another case where maybe some of this code has evolved quickly and isnt doing what it used to any more +Its the cleanest solution at hand for this kind of issue +Marking as critical since BPEL is one of the corner posts of the release and needs the additional support to help our users +Ill check it in +However I think I might have found a decent compromise +No regression breaks +isnt in our avro generation file yet +Its a plus from me mate +Setting fix version so I dont forget to make a call on this for +Fixes and this NPE +There still arent global inout handlers but I think were going to try a different approach where we create our own spring xml format. +I submitted the patchI think we should wait until the patch applied +bulk defer to defer to +it is working on Mac but Im not sure if thats the issue +So it looks like I may have fooled myselfThough insists that our actual persistant storage is only about MBVery odd + +See +I think thats a bug if something goes wrong during recovery we definitely shouldnt be deleting data that hasnt been replayed +If analysing a graph of the results for n iterations any period would be included anyway and could be understood as is currently the casebq +Sorry. +Still missing to add ASF banner to Any Service landing page +Git commit eeffebedafdd +Wow that was quick +One of the more dangerous isfinal class implements from packageThe first application invoking that class will force the configuration of all other onesPlease remove final from classes definition and from developers mind +I feel these are better suited as rd party packages or under projects like Mahout where they may be utilized as a utility classHowever feel free to reopen if you feel they add good value to Hadoop common itself +I dont think there is much further to do on this one +to commit +Heh it was a long two weeksAs another step towards making the API prettier Id like to rename the queries And maybe add an that just wraps a and a +Now Deployable is default and it does not show child modules. +Hi RegisWhy we have to create a new class rather than integrate them into one test class +Yes this page almost crashed my browser right enough! What can you rely on if the HTTP response header is not present +Implementation is done +Assigning this back to before I close it. +patch that matches reviewboard +Im adding fix for this +I will work with him on this issue +Ill close the case nowAs for the interesting questions you raise Ill rely on the time honored advice If it aint broke dont fix it +Typo I didnt want to sound like I didnt know what the hell I was talking about +This is the last piece to the puzzle +Can you detail what you are trying and seeing +I gave it a shot and I think it works fine now both for interface based security metadata and implementation based security meta data +Lucene has been released so maybe its time to revisit this issue +what parts of the configuration are you editingneed to share +It has at least remained unchanged since the code was donated to ApacheIf you enable the and run the test you should hopefully see a reasonable failure messageNow the most annoying issue Ive found so far is the fact that the support files directories are deleted which means that the trace file itself goes awayHas anyone solved this problem before +My tests on various Linux machines all worked too +I suspect this is a separate issue. +In the absence of that the command on the URI is quite elegant +The schema Validator class may perform DOM parsing but it can handle the optimizations +Patch add delete options +Brians patch works +i think we should have a rule that the logo is never used with a black background if we are happy ill get george to generate all of the relevant versions +thank your four your IonutI dont think your cronjob is ok mirror is now days oldPlease check again crontab lRegardsHenk have changed in crontab file the line root rsync rtlzv delete homeapache +Should be CDI support enabled automatically here too? Because it is not +HelloWould you please try this patch? Thanks a lotAfter applying the patch should pass on HarmonyBest regards patch applied at revision r with minor modifications thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Patch applied thanks. +It was always like that I just thought I would create some build hook to update that automatically for AS but that wouldnt be stable enough +Not yet creating one +I left some comments on reviewboard +I am on incompatible change until next major release +August Hi MarkSure thing +Personally Im but suggest thats its deprecated instead +Then connect to the nely created testdb via the HSQL database manager. +Im attaching the source code +were still at the experimental stage here for sureSuggestions welcome +Thats pretty unusual +this is a typical business analyst feature +The isnt very efficient in reducing the number of instantiations of objects +this was added in after release. +Should we commit this now early in the cycle so that we can address things like +I would enhance test to add files instead of the first file should be localized correctly the second should fail after the reference count is incremented and the third will not be localized at all +This is a bit on the large side anyone want to run their eyes over it before I commit +name change for review +Which Couch version is it +I forgot to add this to on +And if yes do you think its a goodbetter idea +How do you want to go about testing? Are you in a position to build Castor yourself or shall we push out a new snapshot release with your patch included? +fixed in has been fixed. +I have a vague recollection of something similar to this being discouraged in the past but as I cant find it right now Ill just go with my opinion +just reopening to remeber to get any migration issues clarified +Awesome! Thanks for taking it over the finish line +Verified with JBT to clarify description +Confirmed committed to both andUpdated to be consistent +Please let me know if you have any questions +Dhrubas comment makes sense +You dont want to shut down the job tracker +How about and? I think we can with that name for the total counts +Mazz did we get any more info on this from our customer +I just committed this. +Marking as resolved please if the issue still exists as I believe it was fixed by the issue in the previous comment +How about that? because the same logic applies just because its not in the similarity table doesnt mean it doesnt exist in the model. +the main idea here is that the response should match with the generated wsdl +Unit test coverage for this rather subtle parsing code is quite poor +And just say thats normal practice +At the end of restart two services are bound to the LB rule and the instances are stopped and not restarted +The simple story is were creating manifests that are valid +Thanks a lotBest regards by Richard. +to trunk and branch +This would probably implicate that the release cycles of Forge and JBT would have to be aligned +I have applied patch files and +I have also added a test that triggers an exception using the well known golden ratio which is a kind of extreme irrational number. +Which ticket am I missing? A new Wiki +Trying again +I think Required Software is a must +Will add this to the fix +Resolved +Id really appreciate if the attached XSD could be simplified as much as possible so that it contains structures relevant to the problem only +The Standalone Configuration is only about configuration and hot deployables are not part of configuration +Update title to reflect new info from needs more investigation. +How are you setting the username and passwordColmCan you supply aColm +Committed as revision It can now be closed but somehow I dont see these actions available in JIRA panel +Done will be available in Designer +Lets remove it from as well +v does full coverage of cases no Snappy SO no Hadoop SO Hadoop SO but no Snppy SO no Snappy Codec class Hadoop SO without Snappy JNITested with all these combinations +That sounds like a good idea +Yes I noticed the other sporadic failure too +Although I used in the commit message. +Just leaving a note that this is a and working fine in the below mentioned commit cabcdfaaafff +for hint Tong +But at least it lets me press FinishSteps to reproduce the current error +Patch attached is causing build failure so will be attaching a new patch +I fear this will only work properly in the mode +I think that the test should do this for me +Thanks +Run sort job +Therefore I have input element +On a related note Id rename the parameter from length to count +This worked correctly in the unit tests as they were run locally and the path did indeed existWhen ran in MR mode access to the normal file failsThe included patch uses get to retrieve an to the script and copies data from this to a temporary file that can be fed to the Groovy Script Enginev of the patch uses do you mean normal file? I still didnt get why copy a file to a tmp location works can you explain +committed it to master branch +fixed it +But a lot of things happen before that could depend on the repositories specifically handling of artifacts +The VP is merely another participant but does have the additional responsibility to send boring reports to the boardI miss Henris participation here. +Hi OliCould you review the attached patch? It fixes all three issues on trunk +I think it is a good idea to add it to documentation +The biggest problem is that this caching gets completely wired with indexesThe rewriting is done on the reader +This test coverage is desirable + I need to create a JIRA for this issue? +The aforementioned is an example of an interim XML interface listing for which was added for copying while we work on better protocols +This could be a complicated issue that unfortunately has been overlooked until now +Added support for dropping from the tail or head of the queue when they are full +An additional fix that enables transport options to persist connector restart has been committed with svn revision and will be available in +That is something we should discuss on another JIRA? ? As a global filter on a job we think it heavyweight and limiting +Our kitchen sink quickstart should follow the existing kitchensink quickstarts example imo +That will have the same effect you would have to set compile scope explicitly everywhere in our project +Add method hashcode unit test +Needs to be tested more thoroughly +In spite of the above tiny shortcoming this will make the Daytrader app a lot more and still open to work on all the different app servers +dragndrop has been described. +Resolved after switching to the new EJB resolver SPI +Make sure that when we go to get data from a znode that if doesnt exist because its been hijacked then make sure callers do not try to make use of null dataAdded test for null to +Maybe you can move to and use it +Patch containing suggested extension +Of course it would have to do the right thing when the second monitor is absent +Please close if it works for you +You may want to close the issue +If the goal is to load data from FTP servers into HDFS or then a FTP implementation would permit both using existing tools like binhadoop fs and distcp +Hmm +Did anyone else have this problem? It seems to be pretty common after some Googling but Im wondering why I am running into this problem and no one else on this list is +Using the lack of a coverage metric as an indicator of a build problem while potentially convenient seems incorrect to me +I have tested the attached ldif file and it seems that no space is all issues created in and before which are marked resolved +Probability of data loss increases with cluster size +Conn Factory props still need to be done +But now we see there are other issues +This issue specifically asks for supporting which I am definitely against +Pull into +The problem here is that theres a tight loop so this will overflow the stack +This issue isnt of much use unless you point us to the actual links +Lets push data format change to + +The objects are never killed and stay there forever +When youre learning Pig all the log files piling up in your working directory is undesirable and confusing +I cant see why has been changedWho made the change? Ask the committer why he made this changeAlso I really sorry hate the all over +So yeah +Attaching the code as a diff file +I was waiting for comments and after the last batch of messages Im wondering if there is anything you need me to do or if I should just leave it up to you +can you confirm its ok for you +I also remove the superfluous at the end of the inner class +I dont believe it is a packaging error though +Unless there is a bug it will not throw uncaught exceptions or get deadlocked +Thanks for the review! Ill update the patchThe equals method was generated so that is why it isnt very readablebq +Slava please add a spicial flag to the model for camalcase component make validator recognize such components +If I removed the exceptions went resolved it as removing all uneccesary libraries from the artifact +That patch caused this test to fail +modified patch according to recent Commit Revision +returnPrevious boolean renamed to closePrevious +It seems the old fix for was not goodThanks for the report and thanks to Thomas for the identification of the wrong maximum element detection. +At least being able to configure it to work would be good because Im not sure everybody would be ok to see that their build is successful in this case +Added a few new comments on review board +Going to check in the current iteration of this shortly +WTP wizard dosnt have Traget Server field +I have generated a patch +The code is correct +Great idea! this not already done Dan +Additionally the exception object and message will contain the current stack of components being created when the problem occurred +That may well be the cause +The simplest solution is using WST validation builder to validate seam it means that user have to enable this builder in the property page for projects to have seam property our seam validators work the same we have to use seam builder or create new one for validation. +I am submitting patch for this issue +It is not a bug +oauthtoken values from oauth requests will get intercepted here and will cause failurescurses! they should have made it oauthtokenI can fix this quite quickly by checking for the oauthversion parameter +Thank +Nothing about the chart is being parsed at this time +The dependent versions are verified and updated. +This feature be implemented once it gets into the specification. +General Windows problem I think because what it fails at is in file paths which is nothing that cygwin would specifically add rather to the contrary +There is a similar query that sometimes returns the expected result and sometimes returns a single rowThe difference in the query is that there is an extra clause in the WHERE clause AND orangeThis query is expected to return three rowsPERSONID PROJID rows selectedbut sometimes only one row is returnedPERSONID PROJID rows selected +The issue needs to be pointed! Thanks for taking care ofErik +Looking forward to seeing this land +name +you please attach client and server side code as well then that will be easier for me to regenerate the issueThanks that a client issue seems like client does not close the connection +Checked the original bug reported and that is fixed by Bug +Im not discounting the possibility that Im doing something that should not be done luckGreg. +Configuration file for to look into this and see what we can file for new editor in +The s triplet seems to be incorrectly increased by +Ill get it fixed to make the return statement +No problem +Ok but is it intentional to + +Thanks! +Any chance yu can check that against a current release or snapshot or provide a test case? +Java and C queue recipe +Thanks Ate Ill fix that now + +Im hoping this will make it into that release +However thats already the case in HEAD so it is fixed for +Running a cluster on my phone. +Theres problem in Native Core + +Setting this property to a value will automatically set the exclusionPattern property to null +I took this issue over +sample document for for reporting thisThis issue was caused by the class using a special pattern for detecting entries inside Microsofts OLE container format +Committed to trunk at revision Committed to branch at revision +Thanks for the order to authenticate the request I must read the request body +Anyway just change the header to or something like that +note I committed +You have been quite on Camel lately so I guess CXF got your full attention nowMakes sense to wait for CXF before we do the first cut + +The permgen for info was in the aggregator phase. +Yes + +Another workaround is to set true this works because NIO doesnt use a dedicated reader thread and reuse of the socket works even though it was timed however detection of no reply is better handled using instead of setting the soTimeout on the connection +I think adding JDO will make integration for new providers more difficult +Committed to issue as it has been released as part of release +xthanks again Roman +When is this being targeted for a fix? To me this is a huge bug and Im wondering how it ever passed any sort of testing when it was released in Beta +Fine with me +Included patch for. +KBIll post an amended patch in a second + for is fixed in CR +This has been resolved in the ESB plugin compiles against embedded JOPR as part of the ESB project build +I sure do + joint compiler also compiles it fix IMO should be just to exclude them from the generated Java classes since it would be impossible for Java to call these methods even create a method named like that? I am curious +Cloud not apply this patch Need to double check why +Hey how did you do that +You can run the Quick Start with embedded Derby in a mode that allows external connections +composites +Were simply releasing ALPHA as a zip file. +Itwould be nice if it could be checked in +Repro for the hang +Works for me +Then wed never have to search the column name for the separator +Debugged the patch for Linux platform +Nice work Tom +java files have a few tests each that use +Regards +Hi FrankFixed Issue +So you will also have to change to ask the Cache to resolve the dynamic revision again maybe with a special parameter +Will doOne task complete Roller has a new license applied +Even if its bogus you need to either figure out a way to avoid it or add it to the findbugs exclude file +For completeness sake heres the planned optimizations referenced in the spreadsheet +Im hoping we can have a fruitful design discussion on this JIRA then perhaps get a branch and start development +in SVN trunkOleg +Attachment has been added with description Screenshot from P open bugs to P pending further review +I committed Dawids patch to trunk r +Patch was wrong working on a new tried this when Paul first made the request and there is a thread on core +This is related to which deals with a similar issue in the test suite +Can the priority of this either be increased or at least a manual process done to sort out to get the published in line with what specifiesUnder STS for example even if the xsds are specified in the schema locations is broken as there are additions in +The other images that are related to these files are related to a previous grant to the ASF and are thus not included in this grant +this issue is already reported in +add some tests where the new transaction sticky lbps work as the default ones when not running within a transaction +Alessio would you please confirm r properly handles these API changes? If so assign back and Ill remove my comments from the code +The class might exhibit an NPE if the current threads context class loader is null +unscheduled issues to the next release +Indeed as was said by Freddy without this file we cant investigate your problem deeperHowever as a workaround you can switch back engine from Sonar CPD to PMD CPD by setting property to pmd +We can still use it if we go change to something else and update all entries in db for +If you define your error message to include the minimum number of characters then it will display that +Can you provide the according please where you use the +Working as designed. +Thanks! Itll take a while until this will be reflected on the website +They all work correctly on OS X with Java +Sorry forgot to add +But it is some duplicate effort already in hbck +Hackathoners if this is too specific of a component feel free to close thebug I wont be insulted +Robert you are not fixing a bug in you are working around a deficiency inThe only way to know the latest sequence id is to make a complete changes query +The version of I attached previously included my copy of the target folder which meant just doing mvn jettyrun was actually running the wrong code +and btw reported bugs that are considered so super critical that they are marked blocker yet do not even come complete with a simple testcase immediately go into my yeah right pile +The original motivation of porting the X runtime from Java to X is long gone + attached with and Committed minor changes +fixed in +is the only one Im not too happy about since it can be misinterpretted as reduce stages +Theres an additional issue that we currently dont have fuse enabled at all +Thanks for reviewing the Release Notes Andrew +We just need a testcase for servlet run as identity in being correctly propagated into the ejb subsystem. +Is this run using the build with Kelvens MS memory fix or not +Lukas If we rewrite the Web Service wizard with the ability to create a Hello World class we dont need the other one so it would probably go away +The patch uses and Tier Compaction Policy interchangeably +Ive enabled support for exporting the same package more then once +It appears you have already looked into JAXB DOM bindings and didnt have any success and due to time constraints switching to was the only option +When one tests changes to our projects download mirror page by using the browser proxy trick before the rsync publishes the pages +sigh I will try to take a look and get back to you sometime this week +Were always open to patches and improvements +They both are designed to work in a environment and are reasonably adept at it +Jira lets you attach multiple files so it is not necessary to use an archive even when attaching multiple files +Ill double check and report back before I attempt any commitsThatd be cool +Favourites metrics configuration page was removed at the same time as measures view +However Im getting an error with MAPITEM +Issue closed after explainations +This is similar to though that patch only does this behavior at startup and doesnt track the peak datanode count as youre suggesting +Thanks for reminding us that we havent had a proper JIRA issue for it yet JuergenI am so glad to hear Spring is going to add a high level cache API +Alexey is too old and or even seem to not have this +Worked on Windows +Attached the patch +HiOne option is copy the same +The name would than again be ok as it would only contain the stopword convenience +I guess this would cover this task. +But did not finally fix it +Done also changed an INFO level log message on batch! +I vote for closing it +In this time the vm stands and a heartbeat cannot be confirmed so the node got suspected + +Fixed in Branch BranchSPCP Branch and trunk +I made the following changes I looked at the code change suggested in the previous comment and the code context and made a different change +I will try and see if I get the same issue so that I can find out it is really a jdk issue or a new code issue +Im assigning the task to him +It already supports typed columns by defining a Map event type and using the alias for that mean +Gary raised a valid point + +QA regression testing completeddeaaacfd someone upload to it toCustomer needs it +Thanks for your patch Francesco +This time guess what its a class named thats missing +The scheduler is perfectly able to translate requests as described above into the current model +Kicking Jenkins again +The above bundle is generated using maven felix bundle so we have no maintenance overhead due to new dependency inclusion exclusion or updateThanks and Regards AndreasYes +So if String is used it is wrong here +Now I am able to reproduce it on a test system at our site as well +This was never a problem for drools AFAIK there are no hibernate specific annotations in Drools +Further investigation still required so deferring to resolved issues for already released versions +Changes appeared in documentation so closing. +Thanks for the pull request +Patch for seems to me that the bug was fixed in the recent OPENJPA build +Yes the bug is specifically about the ability to deploy the war file to a specific servlet context + +Not really urgent +When embedding the embedder code is responsible for closing streams +and files are added +This will make it in. +found out that Clean button was also removed from Jboss Server View +when in doubt +I am uploading the patch to solve above problem +Yep same error here +Attached is a patch that upgrades zkclient to +PYTHONPATH is out of date in as well. +MR may also need it when doing state conversion from Yarn state to MR state +I leave in the hand of the commiter to review the choosen name for the tag +This is similar to the way that an INSERT INTO +Backed out that offending line at svn and for and respectivelyJohn maybe only devs can do that so Ill close it then. +If you can post a repeatable test case then the jira makes sense +So add table tool makes regionstinytinyuserThis seems right +Not an observed problem anymore. +They need the th bit to switch to character encodings +Thanks Mark +Hi Davidthis issue is an improvement of the website of OFBIZ but we have not the access to understand what to do here so if you can look at this or close itThanks in advance vote for my old favicon thanks Hanso +Ill retest. +There is now a successful build + Cucumber tests for IBM SBC driver Workaround for invalid HWP pushed few weeks ago +I tried using the filters to create a category metric to allow filtering on categories +OK +Return an empty Cluster instance if the instances file is missing +Im testing it on local target platform to confirm that repo mirrored that way is working for us +However I havent been able to reproduce the problemThe problem is consistent in my ejb project and goes away when I revert to the older version of the myfaces jars +Resolving this issue since the build is not failing +Hibernate core lets you plug in your own batcher which could be a wrapper +In the Sonar team nobody currently works on this support for weblogic. +These combinations would need to be tested +bq +Bulk change fix version changes + +Yes this should be possible +Arvind Im not actively working on it so please go ahead +Review comments from MilindDhrubaLooks good +is first time is being supported +In this case seems we can see NPE in +The implementation of fixes this issue. +See the screenshot +Heres a patch that should show how sync and async libraries can be created and what the async lib looks like. +I will investigate the failures +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Its a bugThe table is designed to be automatically expended in ODFDOM +New version of the patch which includes test cases +Actually Ive already done of the change +I have run all the failed tests locally and they pass +Patch for wrapping the code in in a trycatch block returning null on failure thus falling back to default +that makes it much easier licensing wise for us to integrate that patch directly +I think regardless of the testing strategy this issue can be closed Invalid since it is a jvm bugMy testing preference would be to use runwithfoundationfalse for this test as it is good to have some JSR testing to make sure we dont check in something that breaks it +But I think we need some help here from a openwebbeans committer or someone involved with CDI to make this part properlyThe main issue is how to deal with the passivation of the beans +is starting you try with Eclipse and a fresh install +Thanks for the review +This ties into the issue of adding modularityplugins to the application +It seems Vinicius is ahead of schedule with this project is that corret! Is there anything else to do after documentation? If not I would ask whether its possible for you to change the way you use config API for numeric +Added patch +thanks for finding this Markusfixed in rev +Fixed in both branch an trunk +Sorry that was a typo in the patch due to a which ended up only inserting the first value for a IN +And committed toAhmed thanks for taking care of this little monster +Ill add extended to handle component +All entries in the map are incorrectly made dirty when a Persistent is deserialized +The problem has disappeared in my environment now +If there is still work to be done on this issue please reopen. +Uploading a patch that makes this construction work running regressions +The changes will appear in our next release +Then is best option to create a new filter? should have been named and should send only the key +Would that work +I added a more general solution and implemented a slightly smarter solution for this in the base class. +I guess this is the best I can do for now +We could introduce an nant task change from dev references to release reference locations to remove this tedium from our perspective +This bug also seems to apply to relationships +I would deprecate those persistence managers soon as they are much slower then the bundle persistence managers +Issue closed as the guide in question has been published and made available publicly +Merge s master branch +Im taking the liberty to close this issue since its been around months since the last activity and issue seems to be fixedThe reporter should reopen the issue if the problem is still present. +No I have not yet had a chance to look at the memory usage with the latestcode but I did with the D release and found that the majority was fromthe DTM +committed +It seemed to work but with a couple of drawbacks saaj only ugly configuration and a bug in the handler with suns saaj you like to see something like that in? Can I contribute it +Also thinking aloud now that the tool is called hcatcheck instead of HCATPINGDATABASEetc we could make them HCATCHECKDATABASE +Id prefer if venus ran on a machine that can be casually given access to so seems like a good place +As noted in the docs you need to be case sensitive. +Popping up priority one is not supposed to hit the filesystem and modify the catalog entries by not reproduce at first +I can create a Jira project for you in minutes no probs +But I updated Seam Faces to and now everything works fine +I just applied the patch to and trunk. +set a so this doesnt get lost +Are you looking at the same project I am? I dont disagree that those are nice features but the namespace is already in use in the trunk +This would be one way to solve the bug +Resolved by taking away the variablelist with the duplicated info +The same state as the last patch upload but with the new files added +Sorry no idea whats causing this +We know that the messages are only send out ones because we control this ourselves but sometimes Railo sends them more then once even to times +lets resolve this issue and try to use in order to detect any remaining issues +Sorry forgot to add that the time limit for validation I guess should be configurable in a property file or a database table if something like this is already configured in a database table +IMO most configurations are required +What is the introduction of a new make tool for? can whatever is trying to be accomplished not be achieved with autotools currently +yes exactly +Not sure if that is doable in a consistent way +Well I just wrote a novel as to what I think is going on and then the jira server crashed before I could submit and subsequently my firefox crashed as well +The issue is after +Thanks for the contribution +Patch to create an internal text bag Property +Accepting +missing files in rev +Closing issues before next release. +This patch seems to basically work +svn folders shall be removed by default without user configuration. +Would be great to see where to use which font size and color etccheerstsuy +And all of them read exactly as many bytes as set by the length argument ignoring any additional data +Patch fixing the javadoc to configure hdfs good to me +The disadvantage of my proposal is that it will also effect MapReduce +Emmanuel as you asked me Ive tried on and the issue is still reproductible +Closed +but I thought I fixed it +I think because the source generator is failing but is not returning any complaint or error or exception +How else would you like this count to be accessible? Would you like the retryRequest to have a different parameter signature or would you like the connect count and the total count values to be accessible throughOleg +I feel silly now +Nice find Edison. +fixed in revision Paulex please check that it fully resolves the fine thank you Mikhail +Attaching the which i am using for my application +will create the log path by prefixing their +fixed by +So at this point we should just say that TIMESTAMPS are not supported with thrift serde +Please if possible supply an alternative have no clue about java on Macs. +even if it has lots of in it. +Missed changes from user guide +I think this will still be an issue if the binary value passed in does not have a length associated with it +Stay tuned and hopefully well have something for you shortlyThanks +using classes where applicable INFO Got compressor reading all sequencefile entries +updated patch committed to trunk in r +Hi PrakashThanks for helpIn out Hadoop cluster some datanode store nearly thousand blocks and Diskswhen the disks IO with high presurethe file getlength funciotn could be slowAnd it will hold the lock dataset too long +No this is for Sandbox +BTW thanks Tom for the initial investigation thats precious when you say in the commen above that later the unscaledHeight where does this occur? in Grid unscaledHeight always has a positive value. +change behavior IWho uses velocity? mostly Java developers who know the concept of null values +ScottIs your sync up to date? I fixed a bunch of these issues over the weekend including check for copying it only once per call and adding the classes to the test classpathSo and should already be taken care ofThe next checkin will have and +please reopen if you have more information +Patch attached +With the SVN co mentioned I eliminated all the changes mentioned which area Disable authenticationb Connector port number changesc logj tweaksand I still get the same resultsI killed my AMQ broker to accommodate b +When gets installed with Maven configurators then this error should occur anymore +It looks like you are suggesting that customers opt into security when they turn on authentication +Since the patch contains only disabled code for creating an empty Clobthat is locator enabled I will run the following suites jdbcapi jdbcPls do tell me if you think this will not suffice and I need to run more testsI will do soI ran jdbc it passed +Closing +patch compiles against current for jdk on Michael For an example look into the from LuceneThe ant file on this Lucene module does not follow the lucene convention and it uses a complex implementationSo I fixed the problem in a different wayI renamed the contribqueryparser to and I fixed the to include when a jdk is presentI tested default all work fineI just uploaded the patch v with this fix +We do notcompensate for wrong user input +A runtime is not the complete answer +Bulk close of issues +The fact that s can be cached makes me even more hesitant to switch over to them because unless clients are sending the etags from possibly unknown previous entries it would seem like a bad possibility of masking a document +Committed the updated patch to the trunk as revision and am marking this issue as resolvedThanks all for your help +James Turner submitted a patch for this but for some reason the mail didnt go out. +reopening so I can modify fix target to +Since the implementation changed I also needed to switch some test code from using mocks to using real file system interactions +Is released? I dont see it on the site +Note the intended use case is that you point it to a directory with a runtime directly in root or a set of runtimes directories directly under that root +I would LOVE to do this +This is not a bug if you are running the test local to the AS instance the local authentication mechanism will be triggered before the username and password needs to be usedAt this stage individual users do not have their own permissions once we do have users with different permissions I will this +I am with this +committed patches in and +can we write the current size of the ledger along with each entry in the ledgerThis might not work very well if there are multiple clients writing to the same ledger concurrently especially because the Bookies do not keep state for the clients +Thanks Jakob go for it patch is looking good +The sneakiness is there to allow order scoring when wrapping BS otherwise CSQ would strip the scores but with the cost of turning itNRQ is out of scope here because query is already rewritten +Greg how would you feel about getting this in for the next milestone build +I can commit if there is no for the patch coming up today +So I really think the analyst should be able to access models and enums etc as a package wide context by default even if the analyst has not been assiged the permission to these model and enum type assets explictly +Obsoleted by property scanWars should never be changed for any reason +Turns out that my previous fix only does half of the story +This is an abstract application for a demo +You could attach it on the CWIKI and put a link to that +We currently have SOAP services in production that still provide a when malformed XML is sent to them this is a very difficult bug to deal with and there is currently no ruined a better part of my day +Fixed for build +verified in Developer Studio is an issue again because the annotation was moved from to in Seam with. +That way once the creation of the is complete its ready for use across all fieldsWhat do you thinkYes I agree with your comment and updated code as wellI didnt understand point are u not able to see initFunction member of struct inI am attaching the latest patch with only three files +Can you provide more details of your environment and ideally a test case that reproduces this issue so that we can see what is going on +The patch seems to do the trick +As this was one of the issues this test was forAgreed an additional check of messages still in the broker would be prudent +This will then includetransaction isolation levels being used and indexes vs +Patch applied Thanks Ron +Junit is still widely used and I dont think its going to disappear any time soon +The attached patches start the process of making the security builder pluggable by turning into a gbean implementing a interface and moving stuff around so everything still compilesNext steps are to make the selection of the security builder namespace driven and to the interface in particular the default principal builder used by the +Ill give it a shot tonight +The does not carry them +Once the internal process provides the permission to commit the fix needs to be then committed to appropriate branch +But maybe we should revert for now +Let me formulate my thoughts and write them here +The old has been deprecated in favor of the new. +Found due to the fix for +Are you thinking of a that allows you to use primitive types without a Tuple wrapper or altering itself to support that? It seems like the latter would break a lot of contracts in Pig many of which were put into place for a convenient reason +Hmm +fwiw both cases pass on OSXAlso I tripped across this attempting to do some work with so it might smack others +Then you could treat an index that only has a segments file set to isCommittedfalse as a index +Each such KV block could have several and the protobuf structure is going to have corresponding number of records that dont have but have for example +Anyway since it is already there we should support it +From this information we will write the final release notes for all issues + Bug has been marked as a duplicate of this bug +Changes are in the trunk +I added the license to the +A common example for this would be the memberOf attribute of an system +We should only schedule plugin issues if they are relevant for a core release +Please delete this from help for wsconsume to not confuse users +is the orginal description still relevant or is that what you wanted to edit? Its usually better to create a new issue and just link back to the ones that look similar +Isnt that just another example of the generic problem with the current layoutthat it doesnt handle changing properly? +my nd point above was for the trunk +Patch applied in svn revision +mergeAction class is added +META +bulk close of all resolved issues prior to release. +Basic site added mvn properties to configure and source encoding in reports. +Do you have other similar binaries in usrbin and optlocallibexecgnubin? If not changing order in PATH should help +The latest files changed areDocFilename Title Summary of updatedevguidecdevdvlpDerby limitationsAdded XMLdevguidecdevstandards Added XMLdevguide Added XMLrefrrefexcept and error message referenceAdded XML error messagerefrreffuncxmlexistsXMLEXISTS operatorNew filerefrreffuncxmlparseXMLPARSE operatorNew filerefrreffuncxmlqueryXMLQUERY operatorNew filerefrreffuncxmlserializeXMLSERIALIZE operatorNew file +One possibility is the case that theres no leader but only followers +fix version needs to be done +Ill look forward to the Beta download. +Applied with modifications thanks. +srcjavaorgapachecommonslangmutable +also all parameters to invokeinvokeHttp should be nullable +Avro container files do give an option to get the offset of a specific sync marker which we could use as the offset to specific events +In that case it made sense to have the +At lease it will be available +There are many extensions to the service within Facebook to support more general information +Committed revision in trunkAlso added a couple debug log callsLasantha thanks for the patch. + +Logged In YES useridI downloaded the file again and it is definitely downloaded the file and it worked did not try the file +With version a suitable key is used for caching as Daniel suggested +Any better ideas? Patches welcome +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +It can be compared with directly which introduced in +I like the new log message on how much assignment retained old locations + +Use this one only +Tom NicolasShall we revisit +FYI this is because the s string handling is based on the java parsers string handling +Committed to trunk and +The program seems to have a problem +I am marking off unresolved bugs from to +Ive used setting logger level when debugging tests in +Grep for the WARN messages like the ones posted in the description +I have removed the conflicts and generated a fresh patch +Using the proper component +Please look at Websocket for the wire protocol it is way faster than HTTP and also HTTP does this matter? As far as JDBCSQL is concerned a resultset can be from many tables or even none +I just committed this +Yep this patch does not fix that behaviour on theThere is one setting for everyone for the connect retries and timeout and AFAIK that setting is not configurable on a case by case basis I am talking about the hadoop ipc client +I rollback the property back and find another way +Tested with DBWould love to see this one merged +the patch looks good to me +No there was no error +Did this also resolve the richpanel issue +I refreshed the patch to make it work with the changes introduced in the Bug +User guide will be updated inRemoval of deprecated package cannot happen until. +Currently it is implemented for a variant of in mind there will likely be some minor work to port to stock and significant work required to port it to trunk +thanks for doing this +Concat also needs to make sure that all files are not under construction +Future specific cases of this or similar defects should be recorded as a separate JIRA. +Sorry for late work +Can this be customized to accomodate those languagesMaybe but we have to do work first +I just committed this +Whats the verdict on whether this affects any of the trunk code paths and the secure launchers +until we add a unit test and implement a proper fix +Thanks for the suggestion Im hoping to put together enough interesting ideas for the next Lang to specific +See alsoWDYT +update your project with last ajaxjsf version +Applied the code +Voted I also think that the drag indicator should not react to the drop area underneath the popup modal panel +I can confirm that with this patch applied to a dev build but not without Maven execution leaves behind no threads as expected + +mvn clean assemblyassembly P distributed true For assemblyassembly verified that hadooplibs contains correct version and wars do not contain any hadoop jars +This patch fixes two issues missed dependencies that are happening because of changes in the Hadoop jar files layout too narrow mask for lzo jar file the name isnt always start with hadoop +This would also be an appropriate place to say something about content based filtering +lgtm +Another reason perhaps to introduce it to the bk client library is that we have to make modifications to introduce the length of the ledger so we might just introduce a timestamp as well +And its agreed that this piece of work is not containable in Stage One and it will be pursued for Stage Two +This is pretty uncommon so I think well just live with this +Michelle are you saying that you suspect that it would only appear if you use an old workspace with the new installation +Makes sense +Committed revision +Moving all unassigned bugs out to to please move bugs back to which you will work on in the next weeks +added a deployers module to aop for trunk and the not closing this issue before i know if the deployers in the as repo should be removed or not +Hudson only runs tests against the trunk so we need to run the manually against other will publish tests that failed +I will probably implement access times for files and directories for getblocklocations RPC only and then run to determine its impact +It works when running as root the error occurswhen you ask jsvc to change user to +The should carry different Project objects even when they arenestedJust food for thought +force discriminator has nothing to do with whether the discriminator value is saved or not +I think this PR is good to go +These tests are green on my box +Moving fix to +Then I will try the same on CR +Hi Gavin Has this been done? I still see the old repo that hasnt been updated and no new one +Thanks Suresh for reviewing the patchI have committed this +Please confirm the current status of this issue and if its incorrect change it as required +another version of the patch which shows how to hook in and log things like node patch only to the class +Time we get one in keeping in view all the recent history of these tests +Cannot infer schema for Collection +Thanks Mithun for the patch. +Please commit this changeWith Best RegardsDiwakar +Have you started prototyping this at all +Need to since this commit broke the Eclipse +difference is that I kept sorter version as and long as +Seems as if s schemaType is null +If anyone has local modifications in an affected module it could be confusing to merge +The initial functionality been implemented +Implemented basic support. +Deploying that change now Ill report back w the next crash +Please review before committing +Slight update to patch I had forgotten to implement release properly +I guess we could start from the end of the URL and move backwards until we hit a +Was just curious about the issue being +Postgres does not allow using aliases tooselect id as i from atable where i causes errorERROR column i does not does not allow aliases in the where clause eithermysql select id as i from atable where i ERROR Unknown column i in where clause +ThomasYour explanation makes sense +think private boolean endshould probably beprivate volatile boolean endIt looks like the variable is referenced from a public public methods of are synchronized so access to the variable should always be protected by the instance mutex +I noticed test hangs when running the new test case against a server without the changes to +Reopening to address the chained +I added the proposed patch in revision inlcuding some minor tweaks and some optimizations +Wow thats a tricky one +sorry wrong issue +Fix for setting default value of in r +Davids fix for addresses this issue +is appliedDIGY +Can you try again +Better patch which fixes frustrating javadoc issueCommitting now +Does anyone else disagree +Uwe I dont think you understand what I am sayingif my text is instead ISTANBUL versus your STANBUL it will not work +Same issue +Please close it when and have been the fix version of core issues to just have and for fix versions. +I need to have a solution for this +The way to rebuild pig is to edit ivy and bump the jackson version to then run ant +Hibernate seems to pick as the default size +Add README LICENSE and NOTICE files at the top level of the projectok Use srcmain for the LICENSE and NOTICE customizationsthanks for the hint +Tentative classes have been coded +Awesome thanks +Do you guys think this has performance implications? Looks like a local flush after every write +Sure you can place an atexit anywhere +Renamed package to +for the patch +Sorry was moving attachment from bugzilla it must be xsl +Done screenshots describing directselect attributes and Style classes +JuergenI did a couple of things today to further test this +Thera are examples in the eip wiki page +right now we dont verify the volumes listed are actually available +Still I think the javadoc to the class should mention that +IIUC Arun wants to make off of not so right now this isnt slated for GAI agree with Eli though that it would be really unfortunate if symlink support wasnt in GA since it means punting it out until the next release +If you specify DISTINCT then there will be no duplicate rows in the result +Can you check if this issue still occurs in your setup +Will elevate priority of that issue +It seems like adding the catch logging the exception and following the flag setting would add more predictive behavior +Resolving. +The patch updates the ref manual CREATE TRIGGER syntax and examples +you can test your patch following the attached data script +I will move all AJAX in response XML instead of use HTTP response headers +The final jar + committing shortly +Thanks JB thats not corrupted artifact issue its the issue we need address in OPSJ Base project I will take care of itCheers! Thanks for figuring this odd one out +This enables caches to share the same executor factory for things like eviction vastly reducing the amount of threads on environments such as Hibernate LC +If so we should refuse creating a table by throwing a kind of access control exception +Thanks very much! +I think both Pauls and Roberts concerns are valid +Patch has stack changes UI changes and a puppet change for hive module +This bug has been marked as a duplicate of +The chains are updated independently there is no replication +Please report tickets with more details so we know what the issue is +Added as of rThis uses the default schema and includes the happyjuddi test pageStill need to test with the JAXR client +Ok I didnt know about the ejb tag +Provided a PR to resolve the plugin dependency resolution failure issue that invalidates any cached build result to reattempt dep resolution +Verified inFedora to write release note issue now that release note text has been written +A thread with keqqq suggested that we do a dot release in incubator and that is a far simpler idea +the overhead would be an issue for publish requests for subscribe it is OKI misunderstood you proposal before +We should probably patch it in both trunk and +Can you rebuild the patch on the latest code? with with latest Sandesha trunk +For one they have completely different SSL context initialization logic +Work should still be done at the connection to allow for ROCG authorization that seamlessly stores connections in the connection repository +Instead of using the meta class of it must use the one of +It requires more test +This will be a breaking change but given that the previous behavior is incorrect I think it makes sense to change for with a prominent remark about this in the release notes +Including the drools asset +Old API supports this and is used extensively +Hi DavePlease resolve this issue if you are through with itThanks version was this fixed for +Since these retry operations are all over http will it make sense to let them share the same retry configuration properties? In that case we can rename the new configuration from WEBHDFS to something like HTTP +I dont know the cmake stuff quite well enough to review but one question why does this affect us despite not having any calls in libhadoop that use float arguments? Is the calling convention different even for args +First we should test system VM template before we change it in can you revert the template change? VHD format is a little bit different from VHD format normally it will work but sometime it doesnt workEach hypervisor has tool to convert raw disk to its own container format we should create raw disk system VM template and then convert it to different virtual disk format by using tools coming with the hypervisorif original issue is fixed and you are running in to a different one pl close original one and open a new one for new issue +We are starting a follower replica +Thanks verified. +Mail from Myrna onHi John Weve been seeing an intermittent failure with this test with some of the jvms we run against on a slow machineIt appears the test is not tight enough to always pass on slower machines sometimes one of the queries results in a table scan sometimes an index scan +Didnt look too difficult but I dont entirely understand the implications for the rest of the code that expects to only take one argument +FYI I believe that David is taking care of and as part of his jBPM console integration work +It has no effect in a Portlet environment +Great thanks for the update David +LikhitaPl provide assitence to Pradeep on how to package this +Release note looked clear to me +My question is this could this popping and pushing overlay state information +Closing as wont fix given the above clarifications +Havent seen any test failures since Robert added the patch so Im marking this resolved +Improved version of the patch +Implementations should use the t this be added to the project pluginNow there is a project groupIdartifactId already + +I think this was pulled in already +Resolving +However I think Lao had zero tokens output so if you just import the Lao test from you should see the issueOk I will take a look +Unfortunetaly the commitlog still has all the data in it and I do not see a way to drop the commitlog data withoug flushing allWell I guess I will make a patch for our internal testsuite +Looks good +Suspend Resume added to the deployment process +a little foxed the metadata output is not consistent with the state of the +Corrected patch +Check out the screenshot the config version with the is not the latest which may explain why the changes are not shown +The idea of parsing expressions with new additional flags controlling the variable reference handling and the key function handling looks good +clients issue the same scan in parallel against the same set of G cached kvs +Looks good to meI would make the getHandler method synchronized just to be that there isnt going to be a second handler created +i guess to be most portable xpaths should interpret its paths as relative to server or would configuration be betterthen no variable needs to be in play or +I did a bit of poking around and every thing that I checked seems to be in orderIs this the functionality that is needed? If so I can further test the deployment and check whether the deployment of in geronimo works properly + +Sorry in the previous comment I was intended to write When I am using AA ElenaThe fix should be in the next build +batch transition to closed if someone wishes to restore one of these to resolved in order to perform a verification step then feel free to do so +Perhaps I could point them to this ticket and ask for more informationFor now they will be implementing their JSON API client by hand but I thought Id put in this feature request for the benefit of others +I include the additional source code directoryusing the which is included in the parent pom When I run the commandmvn javadocjavadoc Daggregatetrueusing javadoc version the classes in the additional source code directoryis not included in the aggregated javadocs +New key installed. +We are exploring what an ideal value for this is +this way it wont break with the current use +Id appreciate fellow committer reviews and sI was hoping to nudge Daryn to add optional client initiate +Note that the supplied in the root of the file must be installed see it working without the bug comment out the checkstyle report entry from the in the mavenbuild this helps +So calendar doesnt work +Apparently I didnt read Owens proposal carefully enough Chris +Checking in modulescoresrcjavaorgopenejbcorbanew revision previous revision Checking in modulescoresrcjavaorgopenejbcorbanew revision previous revision Checking in modulescoresrcjavaorgopenejbcorbanew revision previous revision Checking in modulescoresrcjavaorgopenejbcorbanew revision previous revision Checking in modulescoresrcjavaorgopenejbcorbanew revision previous revision m sorry I misread your ealier comment +Moving out of +All files intended for distribution should be checksummed +also in the data directories the accepted and current epochs are all b but your zxids all have epochs way less than that +Solves the NPE +Slipping GAE isnt top priority and conversations need a rewrite +Bryan this seems to fix it +It would be really nice if this feature could be expressed dynamically in the respective descriptor +Thanks ErikBut how about this patch? It unifies the value handling in the wrap method handling reference values +Do you have idea why does not have such issue +The docs refer to being able to do this but no mention of additional tools +war to demonstrate this would greatly help +If its meant to use that is a JDK method +SVN migration done +configure make thing just to have a quick look or to get started +This is a little tricky +I have extended the analyzer class and use some rules of Bengali languages grammar +Thanks Dhruba! +Committed to with the suggested changes. +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +The code is fine +I tried to understand the above explanations but I dont see any dependencies to class from Continuum +CR bits no longer present! find +more tests pass with it applied +We can extend this class in feature if needed +is fixed in so no I do not have it. +What version of was this logged against? Assuming andor if not please correct +Itd be very nice to have an AM which is capable of processing workflows +Ive read about Retroweaver before but never had a need to use minor question I assume there would be a new jar available as part of the distribution along with the existing jar for +There are two Gadgets category e and Gadgets created by in case Import Application +Makes a little simpler +Patch renames the Vel into Jexl classes +Thanks Viola the additional patch was applied in trunk problem is fixed nowthanks a lot. +Also removed lot of unused imports. +This issue has been fixed and released as part of release +libs defer to +If you want to skip the tests as they are in fastinstall then you would need to pass in the trueI believe to do what you are attempting though can be accomplished as followsmvn PsourcecheckfastinstallFYI I am also uploading a new patch to fix some errors I found while documentingBest RegardsScott Fixed exports for the Connection Resources +The patch also has fix for +Debugging switched one for catching problem that occurs repeatedly on hadoopQA patch testing environment +The great day when ivy moves out of the incubators happenedI have made a change in the xsl in order to reference +However what you mention about multiple hops is a valid pointThe distributed searcher could have an option that returns the shard set +The problematic file appeared in +From my perspective the test mode rollback needs to be fixed +This should be possible even when sending back a serialized object instead of an error page +That change is correct +svn statM javaengineorgapachederbyimplstorerawlogM javatestingorgapachederbyTestingfunctionTeststestsstoreA javatestingorgapachederbyTestingfunctionTeststestsstoreA javatestingorgapachederbyTestingfunctionTeststestsstoreA javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestssuites +KrishnaWhat would happen if the is as a file instead of a directory? This should be considered as a and the should bail out +Issue seems to be related to this one +Jira created. +Sweet Ikes looking goodMaybe a bit too shiny in the forehead? We dont want to emphasis his bold spot +Fixed the first one but cannot reproduce the second one +mike has done several customer driven fixes that are in +confirmed this is a regression and does not occur with +I thought I had recorded this comment before but I guess I didntThe very first error in is a +All tables have a different name +the News link in the left hand menu does not jump down to the News section as it shouldI just fixed that +Reducing the Priority this is not really a blocker for itself although I agree it is something nice to have. +After talking to Jeremy and others I decided to leave this as it is since the spec is vague the best approach is to be consistent in the event of multiple calls to define +Patch for initial skeleton +verify handle short tests run all the tests and also +It uses a method that was developed for +yea no need for this anymore. +Fixed isnt the correct resolution as theres no code change so JIRA SVN wont sync up +Suggest we revert this patchI agree +Support for and annotation scanning should be considered as please add any comments to +So just to confirm as long as both RHEL and zip installations use the certificates generated in EWSetc +Thanks Markus! +As far I know a generator should not be cachable per se +Ensure that the data structure is also initialized in as we do in other classes and apply the same recommendations for local creation and resident option overriding in +This refers to the new start page +Before submitting the patch to Abdera please make sure to carefully read the Javadoc of internalSerialize +I found the cert folder and the cert is in binary format when I open it with notepadEncoding it after the fact sounds greatThanks +Resolved in is still complaining about timeval struct gettimeofday and not defined result +Thanks for posting this well get it fixed soon and let you know when a new snapshot is up +uuhhh not sure if we like to do that +We dont have that luxury nowWill file another JIRA to make sure that work at too +Patch checked inOleg +I think the should be removed and the should be used to signal interruption uniformly +This appears to be related to has been applied to CVS previously and appears to have resolved at least one example +lemme see if I can help with the test +Ill pull it into the three branches +it will be reopened when we can find how to reproduce +It is integrated with the patchI have added the configuration for the three components in the under search request handler +I also changed the maven output at the end of each configuration to include the configuration name +I am running a JEE build +Checkstyle will throw warnings +I tagged build into so the next release should pick it up +There are no tests included with this patch because in reality none of the real functionality should have changed +After we restructured all of them there is no reason to keep default ctor or allow null at all +and it wants a version I do not specify U I still cannot am unable to get thru your advise +Even if its not meant to then a seems like a weird errorIn any case the underlying difference is in the regex module so any further examples should be expressed with the underlyingThanks +done for Server Manager for jBPM Tools Reference Web Tools Ref guide FOP updated to version problem with out of memory is fixed +AndyDBRDF stores the graph information in the database tables +Looks good to me caters for all possibilities +renamed to searchAfter added a little to the javadocs and improved the test coverage a bit +However the initial work done as a pull request referenced this issue instead +for Teds addendum +Thanks Martin +Thanks Jitendra! +These sections havent yet been ported to the docs. +Im not against this even though I thought you had plans to move for example the into a separate hibernate core module +It also hung when I tried running it from a dir on a NFS filerPerhaps you should do a clean lucene checkout and try that +bq +Ive also included files for javaxxmlbindutil and javaxxmlbindhelpers +That said I wanted to clarify my proposal +aar file in order to this issue +Fixing directory limit problem for public cache +Both and are tested and all Lucene and Solr tests pass +The server I was running that code on is a production server so unfortunately I cant easily mess with the JVM for fear of breaking other apps +Ok +so it should be something that would include any possible compiled c artifacts? +Right now the test doesnt test much because it wont fail even if there is a leak +What do you recommend until the next releaseI suppose I could download the source and apply the patch myself +I thinkthis is actually knowledge that was included in many of the specificpreviously? Im not completely sure thoughstephen +Patch for the regression test is attached +Even with volatile messages I still have the problemIn my sample is possible to see +Its how now new commands are and its just about adding a new file to support new functionality. +The cursor reader isa bit different since it needs to keep it all open +Id like very much to add German locale support to the rolodex example but the patch uses a custom loadBundle component that is not in the RI +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +generated from our internal branch but applies cleanly to all open source branches with patch p +bulk defer of unresolved bugs to +Updated to reflect this as a feature so that the impact can be analyzed against the other containers we to CARGO issue and postponing to version undefined. +I downloaded the BIRT sample database and ran the second repro programThis bug appears only in versions and +the cause may be a null ip address or a null netmaskThis may come from a code bug or from missing guardTo further analyze it would be good I can repoduce itI suppose it occurs when you send an email to a specific domain for which the dns is only ipvCould you post a scenario to reproduce itTks + of a bulk update all resolutions changed to done please review history to original resolution type +Thanks +the current solution is just helpful for but not forany real life experience with a oak content repository +Ignore the attachments to this issue +Can you please verify and close the issue if everything looks OK? Thanks +Thanks Dag for the fix. +Thx for all your help Michal! +I think it would be OK if Derby just throws an exception indicating EOF here since we probably wont support using the same stream for two placeholders in a prepared statement either +Tested and closed. +svn upAt revision However i tried making a new compile and install to make sure but it still generates erroneous selector names thrift versionThrift version would say that it should follow the +I suggest trying this cd JAVAHOMEjre find +This patch makes collect rss +jj file a bit +One will likely result in an NPE as the entry is being removed by the Thanks for the patch +There is lots of backwards incompatible changes in Lang theres no point trying to sugar coat other changes and putting the pain off until later +It would still be good to get reassurance that someone else was looking at the issue even if the review is just a patch looks good Yes that is project policy +I will commit it +Closing it. +Marking for so that it can be tracked accordingly +I would like to propose for your consideration a somewhat different approach based on the strategy pattern +Reassigning all open bugs to the development listSorry for the mass mailing +yeah those console issues are all windows specific +is the updated patch +a that the unqualified identifiers should resolve to the current schema for the SQL session and the definition of routine invocation has no language to contradict that and simply says that its a qualified identifier + is now longer a part of this issue as it now works +Any update on thisAny fix? +Please verify +Unit test was not added as it compiled with errors due to changes to that class in creating version of it. +Under the circumstances this hack looks like the most elegant solution. +I just signed up to report this problem and suggest this kind of solution +And your suggestion is super +when will be releasedThanks +We tested it on RPM installation without the noauth patch according to the description of this jiraWe did not test it on zipfile +Thanks Nirmal +fixed +to trunkThanks for the review Jesse. +There was not a single line of code change needed on the addendum after I pushed the patch on to the branch. +That patch to the forrestbot works for me +I added the unit test to but it is disabled for nowtrunk +However a littlebit more research makes sense. +They should be metioned in some chapter like Using Messaging +I will try this out assigning this issue to myself +Ill rename the unchanged class to +Paul I totally agree with you that this is mmm lets just say not an apporach I would recommend +Nod +So I dont see the point of that +The other consumer will buffer messages until theyre you post a test client that demonstrates the problem you are seeing so we can investigate +bq +This is a very simple program that illustrates the problem +The testcase creates a simple tree of nodes and then iterates over the parents queries for the child nodes and then moves one of the child nodes to a different parent +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Verified with ER +defer all issues to +push out to +Revision +Once I get some free time I can assist in any way I canThe approach sounds great +This adds some randomness to the static sourcetarget datanode match +this is the correct one +included in. +Jenkins should automatically run because the patch has been marked as submitted +Sorry to take your time Juergen +Thanks for the patch Andrew! +Kindly commit the improvements i have made in RecruitmentRegardsAbhinav VaidiLabs L T Infotech LtdMumbai +I introduce the service interface avoiding this unnecessary cast +The main advantage over symbolic links being that when the original link is deleted the nd one keeps the actual data from being deleted +When Error happens in Java usually the applications cant do much about it +Yes doesnt check whether the state is available in the log at the point where removecommitted is called since all we do is write a remove state entry to the log for later processing +Might be a couple of weeks before I get a chance to look its extra busy at work at the moment +We wont be able to get this into though since hasnt been released +I will delay it for the next release +No problem I hope this is okay +In the current approach even use method binding and variable binding interchangeably +I want to take a quick look +The generated test +This issue can be closed +supports closing all issues that have been in resolved state for months or more without any feedback or update +here a patch are some issues with the test case for Java +Since this would consist in an important refactoring its better to work on it after GA +Several times your work has introduced outright regressions including this one that you have left to others to fix +So this issue can be closed +I confirmed that this problem also exists with current builds but not with ibm +So well call this fixed and look toward fast and similar optimizations in the future to improve it further. +Moving to +If someone can determine what Sencha does to cause this error we can this issue +Im not sure yet but I dont think its a replacement of the functionality here + would be the ideal spot to catch the exception for failoverHowever its only safe to do so if the method call is imdepotent or you can guaranteethe request has not been processed by the serverThis method is used for many different calls many of which are not imdepotentAnd the may happen after the request has been processed +Try this rendermsgs this +Applied to branch and trunk resolving. +As now its not only one advantage over standart grails approach is no need to place annotated model outside domain you interested in my patch mail me vladimirfx at gmail dot comThanks +I will experiment +I will leave those tasks to someone familiar with the official Apache distribution process. +Updated once again as per Anoops comments +Unfortunately it stung me just as I went to deploy the latest Surefire today +My understanding of Elliotts findings with is that youll need to add it for now because something will instantiate the enum which will try instantiate the using reflection +Applied the patch +that is the code of the activity indeed is incorrect and i fixed created tests based on your process in the test suite. +Assume closed as resolved and released +you provide a new patch +The plugin is tested with the latest nutch SVN and seems to workfine +Ill run unit tests now +Here is a new version of my patch Id propose +Addressed findbugs +Committed on trunk at r +Oh yeah +That is best way to goThis issue is just for the meantime +Providing and patch files that fixes this issue and the POJO inheritence issue described in the linked issueAdditional documentation about the fix is inline in the code +If the corresponding store is configured for string serialization of the key and java serialization of values the client API transparently them accordinglyObviously the thick client and the nodes share a great deal of common code which means it is +bq +Looks like this will be an issue in Weld itself so could you write a test for this? +But no response +Are there any dbs that allow compression indication at the colun level and if so what syntax +Above Correctly raises No more than commands may be added to a single batch +See groovys antlr grammar for an example of generating custom errorswarnings during parsing +Oh good and interesting +Centralizing selection logic has always been a goal of mine but we never got around to a formal cleanup +I guess what Im proposing is that when given an object newRecord would not bother to validate that it is the correct class +ysteinThanks for reviewing the patch +Updated the patch +to enable it by default +ok by you now? not sure what you wanted and why you didnt update it +I know what you speak about I have a good pile too +I think I see what youre saying +The branch was already in RC mode when I fixed the bug so I left it to the release manager to merge it into +A feature wont get committed we are actually moving to remove such features from trunk to get ready for a release next yearYou still have not said why you believe the Java approach is better than using the existing in a script with your postprocessor +Print out more debugging info +The logic behind this whole JIRA is that shouldnt be mutually dependent +Fixed in revision +And I have configured the output path of my eclipse project to be docrootclasses +Jon do you think you could have a look? can try to get to this sometime towards the end of next week perhaps but to be honest I havent had a ton of time for coding recently +Who becomes the initial Moderator for the list +Hi I was working on the bug as Vadim suggested +Think it out loudly reading and writing implementations can be separated because its more secured that RM only owns the writing interface while only the reading interface is exposed to the users +I must take back words serious troubles and lower the priority +NOTE this is having a problem in the service engine with losing the local dispatcher some weird stuff were looking at now. +I left the JIRA open until the issue resolution +Steves patch +Thank Sergey I will check it +Should they be up +Can you please email me under schmidts at vmware dot com so we can get in touch? with commit accbdecdcfaaeffbc. +HiAny idea on when this will be mergedreleased? Or does it need more workThanks! I am sorry about making an erroneous assumptionYes if you could provide more tests that would be much appreciated +Works as expected documentation was just lacking. +Hey fred can you clear this up for me? not sure how to read your last paragraphAre you sayingTomcat is a pleasure to use and our server adapter is far from automatic to use on an mavenized projectORTomcat was not a pleasure to use since opposite to server adapter it was far from automatic to setupIn both cases im curious to know more about the issues found if unrelated to this specific one please open a forum thread instead ok? I meant displeasure! This report is not my finest momentNo the other way around deploying to when I adjust the poms to take into account the naming issue Ive tried to report here works remarkably well +Dmitry and VinodWhat specifically do you want to do with the code once it is factored out? Vinod There is significant amount of vectorization work in expression evaluation for example arithmetic expressions or logical expressions or aggregations etc +Ready for does this change the file channel format? Would older channels have to be completely drained before starting an agent running the new channel or is it backward compatible? I havent looked at the code you posted yet I apologize for that +Please ignore the previous comment +Closed after release. + has been marked as a duplicate of this bug +I will try if we can prepare in classpath resource so any step necessary from user would be configurating could you please look into packaging document how user should do that? +Only security related fixes go inLook in the URL coding strategy that is in useOr just use Tomcat until you upgrade to Wicket +Need to switch back to a pure SVN workflow I guess as I seem to be getting the git one wrongAs for the Version thing I will try to get to it today + Application Server Cache Hibernate Portal Transactions JTS Transactions Web Services Messaging Seam platform above portalPortal platform are using the following for product names Application Portal pid Application Platform Portal Platform Hibernate Framework jBPM Framework Rules Framework ON ON Monitoring Developer Application Platform Dev Tools Application Server Embeded Transactions Embeded +Committed revision in This also covers I also added the correct mapping for the host table in revision in trunkThank you everyone for the input here. +B A copyright notice statement where the copyright is actually held by the Apache Software FoundationSince the Apache License and ICLACCLA involve granting the ASF a copyright license for any contribution is there any case where B would be true +Im using +Thanks Konstantin for the ReviewI noticed it while uploading the patch +Keith Im closing this let me know if it doesnt work for you. +Added a KI to the release notes +What about retrieving user fullname from LDAP? It seems that with the given patch the user fullname will be set with login +There was a COBOL program that contained years of business logic that they did not what to re write so they made a COBOL stored procedure in DB on the mainframe that would call the program to retreive the data +seems to be part of the spring framework not +I was swapping File transport with JMS when I was testing jBPM integration +If you say you want to keep does this mean you want to keep a copy of the in camel named and mark this class as deprecated? I think this could be a good thing +There is a Protocol Buffers Maven plugin under development that could solve this more elegantly but it hasnt been released yet +Should be possible to test after fixing +I found the issueWe pass the MAX into the message +It seems much better now but I still got value that is not what I expectedFollow the same steps above after step I check the counter on +Yeah thought about it but but then sticked to it dont think we should have separate exception for this scenario +BTW currently changing project facet Seam version from to isnt also possible +committed. +For example all of the tokenizer and filter factories moved to Lucene from Solr +Also description in should mention the value is specified in seconds +I deleted it from the issue and now adding this new patchProblem with earlier patch is when running self describing tests it fails in first attempt +This caused the class version incompatibility +The patch is the complete patchThe phabricator entry does not contain the changes for test results compiler files +We did miss a doPriv block when trying to access the service +Hi DawidThanks for working on this +This patch brings the cluster up correctly so that the client can communicate with it however running a job fails with a errorNext step is to set to something large to allow inspection of the logs to see why the classpath is wrong +Could you look at this as it seems likely that the changes that caused this were part of changes you this still an issue +I will apply this patch next and will close this issue +I will wait for one more before committing +Applied in fcebbLooked but did not see similar problem in Harmony libraryInterface codeDecided to not change meaning of sys return code since it is called from more than just this one place +Thank you for reviewingCommitted patch in revision to the branch in revision . +Actually strictly speaking we are modelling a Date not a so the format should really be just without the time and timezone on the we ever want to support Time or then these would be new answer types +The best solution is to split the property on several linesI can work on this Oliver if you havent started yet +bq +attach the patch including testcase +The original problem was resolved for additional problems please create a new issue so that each entry describes a single problem. +Transitioning to waiting for as wont fix as this use case should be entirely avoided if people actually read the text and do as it asks +Add commentary to Fraction pointing towards Math and its +Will be resolved by Single Sign On which correctly remembers back url +release in version +It is not what is eventually written to the doing would make the groovy wizard aligned with what the jdt wizard does +ThanksUma +This always has been implemented except for the delayed registration which I will report a different issue for +Done +fixed by patching hadoop +See. +The only thing I think we may eventually do is to make this optionnalBut unless there are many votes on this issue I dont think this will be implemented soon +New patch with proper logging and configuration files +I have not included the static SDO as this topic complicates matters and should be discussed seperately +Sounds like a totally sure case for removing synchronized best regards removed synchronized from all methods because this is not needed +Highlight statement show schemas and Right click Execute Selected Text +I see that you are using the Struts Tiles plugin that is currently incompatible with Tiles +Im leaving alone since its a little broken and doesnt support Numeric encoded fieldsEverything compiles but some tests fail +I tried to write such an ANT task that shuts down the db +Ok Thats s from PMC members +created +Ok Thanks very mach +Lets close it as not reproduced +I suggest that we split the JIRA into separate ones as they are technically unrelated each requiring patches for trunk and + +Where in the documentation this should be added +Markus I need a patch file upload with an ADF grant please +David can you attach a new version that is modified and tested +Version has already been released therefore moving to thats not gonna help +Stefan can you remind me where exactly this returned erorr status code is used +Thanks Jieshan +postgres has the follwing configurationport maxconnections note increasing maxconnections costs about bytes of shared memory per connection slot in addition to costs from sharedbuffers and maxlockspertransactionsuperuserreservedconnections +I am unable to reproduce that behavior +Completed in current trunkDirk thanks for the insights on upgrading SeleniumHarry changed download for Jetty and Selenium to avoid downloading anything from during build time +Alsodid not fail when run individuallyError Not related to changes in this Jira +The fix is to provide the correct path of the output directoryWith the new patch now all tests passed for me with mvn test +See bdfccbaaaccbdc and surrounding commits in the repository. +the passesso the original issue fails on javassist because of some hibernate code +This is running on RC +Patch addressing Lars commentThanks +Not a problem not a bug +But you have to come up with a heuristic to know when to shut off LCS +This bug is trivial to reproduce set up an thatresolves a simple panel do not add that panel explicitly to a page butrefer to it inside a wicketenclosure tag in the page markup +Now both files below contain the tokens +Thanks for the suggestion! +Stupid of me not to notice it +Anyway Ive added there to fix it after upgrade to Java. +and if it is the child of the ear then it should remain exploded +BTW is it convenient to write a test case for concurrent allocation? Like testyeah wrote onebq +We need to check all occurences of resolved issues for already released versions +Ok I double checked everything before I reported this but I forgot to clear the cache +the patch looks patch is an change that was analyzed and tested on a real cluster +Thanks Todd and Sid. +Well actually things are working as designed +If deprecated in then remove forOtherwise set the fix version for this issue to +Patch v is based on the assumption that the list of dead region servers for distributed log splitting stays the same across retries +Elements weld and weld are emptythere should be no actions creating on them + patch for trunk revision +New patch the previous one didnt include the on the files being copied +Unit tests pass +As James said the trainer should recognize that the training data has only one outcome and then report an appropriate error message. +Snell is that you +If you have a lot of lowering file count does help reduce tooIm not sure I understand what you mean +Am I not combining these correctly? How else can I do a complex compareIf Im wrong on the first part of the clause please let me know +thatindicate this is an XML parser issue and has been fixed in early versions ofTomcat . +Test project to reproduce problem +pushing to is basically already accomplished +Note there is a workaroundsh rm rf tmpsh rm rf tmp +Attached patch to resolve issues preventing compilation of generated c codeIssue regarding nested union described above remains +I think we have to remove it and integrate validation to WTPs framework +attaching the believe this is the same problem as in +The patch fixed the problem for me +With and User has files libjars and archives options for streaming to add filesjarsarchives and libjars are available in classpath also +just committed this. +I think that since this file is actually a simple application and if resources are register there they will also be loaded we should set the name to be instead of This feature can be use also by application providers in a way that a developer can pack all the resources and providers in a jar add the and thats it +revision +Verified at +To run the cellcounter with default rowcfqualifier separator for table planet with a prefix filter rowjava cp +Thanks Kim +That was fastThanks alot Paul +Fixed in the nightly buildA new thresholding output stream implementation is used to switch from inmemory to when the size threshold for an item is exceeded +Remy can you give this another try now that your build is fixed +Lets continue onChris I think youre right +Fix also addresses +As we did in other feature branches if you miss out on reviewing individual patches you get time to review it during merge time + is marked as fixed in IJ and upstream master and are both using that version so can this JIRA be resolved +bq +Also its clean and easy then to have a big Split button in the master UI this way +Cleaned up the repository factories as described above +EDIT After a little more consideration the caching bug had nothing to do with our hash function +Please retest with a CR build +bq +This one probably deserves a reviewboard review if you would please create one +Now I just have to figure out what that is +The gateway proxy approach works too +I didnt exactly know why but with my newest patch it seems to be fixed which means in the future unsupported operations are now throwing an exception at +Could you attach a small test app that demonstrates the problem +on that change +Patch rebased after a regression via distributed tests +If I recall correctly though the delegation tokens rely on a secret key that the master daemon periodically rotates +Please see my post ont he dev list re this I dont grok why this isnt a workaround for a problem elsewhere +Test failure is +Will do that in +I know this because is busted +The library only uses for signal in order to approximate a Unix kill +We can do this through the Snowdrop example but I am trying not to introduce too many variables in the process +Class file version is wrong for closure classes +I am running derbyall now +this code ought to be considered a Public Record produced by a contractor for the commonwealth of massachusetts and would therefore be part of the public domain +Im running tests and will commit this patch shortly +My mistake +The fix is in integrationtrunk +What are your thoughts on thisWhat are you thoughts behind allowing the using of a pattern match for both the Consumes and Produces annotations during content negotiation? For example allowing matches for vendor specific parameters similar to applicationxml when applicationxml is passed in +Done artifact present on central Im updating +I compared the eval numbers and ensured they didnt change from the original +oops sorry forgot about this +This can be used for security or other purposes like replication or transactional support +I want to apply them in this form to trunk and flex +Hi sagara Finally i could able to resolve class problem +My Jconsole doesnt know either how to call a method with array as a parameterKarl how do you invoke the method +I made it for convenient quorum testing its more handy than current and I plan to make a special case of it +Looks good +moving to M will fix in the is part of the issue I think +The latest release on the QA labs for is in opt +Patch that adds check if the exists before trying to us it +As so many of you seem to need it maybe someone has time to look at the Criteria source code and contribute something +CLI should be ready to tag for proper. +When a user is delete doesnt cleanup the account it belongs to +Hi Jacquesi removed the second Decorator +Ok thanks a lot Tony +Can you do a killall java and provide the java stack traces for when the server is in it hung state shutting down? +Source and binary deployment package for problem verification +When run with jars against I do not get a freeze but I do get a lot of stack traces in the output and a failure of the tests +Thanks for looking into reason I raised this issue was that I didnt see it with within the rabbitqueue element +Looking to see if the scope of the lock should be more than just the call to +The jencks project tries to run geronimo components in spring with no geronimo infrastructure present thus it needs bean factories etc for the geronimo transaction and connector components +Improvements +Also the quickstart uses its own action class to send messages to wsmq rather than use the notifier +Right now the FIFO mode is global and ignores pools altogether it was meant to be a way to switch back to FIFO at runtime if for some reason the fair sharing is buggy not as a final feature of the scheduler +for dropping it in favor of. +Hence is more readable as or something + +Fixed in CVS please test! Thanks. +Please? Pretty please? Having to manually wrap them in every application is such a bear even with Groovy to make it easier +I was expecting that your launch configuration using the installer would also result in the same bundles in Equinox +An additional property is added in file and checked in to svn +Created to track this issue under AS issues +The same issue reproducible with any iteration component including hdataTable +we use Maestro so I originally made the changes to the Maestro source then migrated them to the codebase because it looks like a lot of the changes made to Continuum for Maestro havent made it to the Continuum trunk +The template is still valid as long as you run ant package instead of ant jar but I agree that this should be fixed +Thanks for any comment +I think so +bq +Not null check that was mentioned previously seems to resolve this issue +I verified the changes in the patch for Branch to see the shared data structures are accessed under the JT lock +The fix for the memory leak is in CVS please verify installed the latest snapshot and the memory leaks seemd to have disappeared in the test program and in my project +a more robust way to hook shutdown is to use was fixed by Fil as a of the switch to native threading +These log configurations need to be separated from Hadoop client in any case +Id prefer if this patch didnt revert other changes +This is really a bug as depends on a lib that isnt designed to be embeded in a webappI did open an issue in because I expected developpers of fileuploads to be concerned by hot deploy in web servers more than Could someone of the support my issue in by voting for it in and add a comment supporting my description of the problem +This was certainly a good thing back at the time when we discussed this +Code changes look fine to me + test case capturing few scenarios +If the response does its own thing and does nothing within our impl the queue still processed so the queue handling wont be affected by some external override even if the function is overridden to +I think the language in snowball must be English not en +I cannot reproduce the error with an app running exactly the same plugins and filters but the problem persists for existing apps +Is a set of regexps too expensive to iterate over? Would we expect people to have a lot of these? Is a regexp specified on command line too tricky to pass through shell logic? Perhaps simple is best +No problem Rich we can proceed with the current HQ end user license +on the v versionConfirmed compiles successfully and new unit test fails before the change passes after +That didnt fix it +Attached patch using the expireBefore idea +When drop down for autocomplete text field is shown +Can this be committed now +Thanks for the patch +Ive moved on. +Code compiles without warning with r +Any suggestion is welcome +m as the build was on blue ball in some nodes +It sounds like we should just handle this as part of then since theres no need to make connections to one machine +Importer completed including support for update. +This is a nice addition on purchase order cart page as QOHATP for a product will be visible to admin while creating an order with just a single click of button and he can adjust the quantity to be purchased based on the QOHATP levels +Im sure this is not a bug +full name local name +if is able to handle relative paths containing +So theres no really a way currently to have a cache flag that only applies to Hot Rod +This does not represent a schema migration feature only something similar to what had +Fixed in rev df on master trunk. +patch modifies to add footer modifies to add footer +You can close this ticket. +This may be an opportunity to job submission or it may not beFor example perhaps your implementation will use the existing mechanism to provide the new capability + +I think bnd should not include this header at all in the manifest +CoversDefining an SCA Module with a WS Deploying a new service in XML Schema type to C type mappingNotes on supported WS styles message types +Its because since the javadoc ant task in Cocoon is using a fileset to be ableto get block stuff the command that the task gives to javadoc is too long andis truncated by winblowsAdding useexternalfileyes attribute makes Ant collect all package names in atemporary file and give that to Javadoc thus goin round the issueThis fix has been committed in CVS please test and if ok mark this bug as closed +BTW partyboy if you really had just cp as a dependency without then you were never having Hibernate use cp +Todd Vinod Do you have thoughts about this version of the patch +lib and +Maybe overridden again by specifying a different package name in package. +Needs to go in first +Already done +There might be another way of solving this +need to tell what kind of query you are actually executing here +This patch is for code refactoring of some Screens and Forms as well as created services Screens and Forms for the entityThanks Abhijit M +If most people think Knut Anders suggestion is better I will change it +Uses phased +Moving to next release +Good to hear is working for you +Closed upon release of Hadoop. +The UDF and are both included +Although this might be an incorrect use of the API if it works in m and the it may be worth to check for a workaround in would be most gratefull if you could do that +from me too +Committed to trunk +Description of this patch Replaced encoding of eol in javatestingorgapachederbyTestingfunctionTestsmaster from CR LF to LF and Set svn as nativeTest Executed jdbcapi and found no error +Im having original problem in that the three tables still do not sort independently +Well this error happened again we used the schedule component in a Critical Production Application everything was fine until today the appl just doesnt work but if I change the server date to yesterday the appl works fineI hope you have an idea that what is going on I tested the appl and the it isnt going to work fine until April stI am still debugging the appl if I found anything Ill notify youThis dcument show us the problem with the sandboxs schedule example +Attaching the patch that puts back heartbeat code which was prior to +test that verifies the correct operation of concurrent snapshot flush +Yes it would be nice to have something like the policy for ASYNC caches in REPL mode +The changes pertaining to the addition of checksum types to does not necessitate new dependencies that are external to Common +Youll probably notice that it is a bit awkward to remove one part at a time +Does it work overall? Well enable the fields by default and then disable them unless the specified Context Mapper class supports em +Re versioningI think it is better to inline the differences in the reference guide for different versions rather than branch it out +Punt again +So that might mean that we will change this behavior +Could we continue to call it for finalize but just with that parameter set to falseAlso I would appreciate input from others on this +Stay tuned +Failing Hello world project +Hi Suresh Does unicore submission really need all the extensions in +I know that it has since stabilized but I am still nervous about such a large change +In general the naming of compiled classes based on input files and JIT methods needs to be cleaned up and unified +regression test applied at rI think this issue could be closed now. +I take it this is in HEAD? beta lets me navigate the relationship as well. +I forgot to update the position property of consumer so the client wont receive any requeued messageWhen I try to add code to update the position property I found out that I have to insert code before invoking requeue +Should we plan this feature for Solr release +Reported as not a broker problem. +Ive also found a problem in your proposed testcase +A welcome face lift +At least it dissapears after some time +Not that I could tell +Task can be closed because the explication about the html code generation has been added +Retrying the patch +Thanks Ram +The code has already been fixedNote that start and end is specifically allowed as a special case +Configuring provides a workaround for the case where the application contains some pages and some pages +Decreasing this value may improve performance in situations where many retransmissions occur due to the overrunning of a network adapterSetting this parameter does not guarantee the size of the TCP receive buffer +Whats the likely schedule for its release +trunk +Marking as resolved since this should be done. +Note on possible improvement of the unit test +First cut at ESB schema and JBESB Configuration schema +Ive reordered the issues but other than that it looks okay to me. +I wanted to debug into the source of the beans but eclipse wouldnt let me +Dont let this scare you away pleaseDont hesitate if have anything else to contribute at some point +Committed trunk revision close for +Just FYI with the default size I calculate that the empty ring buffer alone will take up about MB of memory +We are already acting against. +From your attached wsdl file Yes we just need to comment out the empty porttypes +Added scale method. +Use the intead of +This was definitely not intentional thanks for fixing those testsPatch looks good to me +Can you please review it and? I think then we can resolve this issueThanks the second patch +There are talks about object serialization based on Serializable and serial to be deprecated altogether +Especially because its being tested quite extensively by thrifttestpy command. +tried locally seems to work now +On the other hand it is a shame to lose that apparently useful debug info +Patch includes the required second patch almost as is. +Do we feel this is necessary? Rubys timeout on connect seems perfectly adequate to me + of a bulk update all resolutions changed to done please review history to original resolution type +If someone wants to document the style for posterity the wiki is open to those that know or can find the answer +In any case I agree in general with Matts point +I am not sure whether this is really a good idea to accept this input silently +I committed this +We can evaluate both approaches +Please can you fix some code style inconsistencies +Ok I am convinced that this bug only appears when you are using tapestry in cachingfalse mode +So any cite compliant db will need to do booleans +To copy the files it should use the directory abstraction lay and not use directlyId use Channels instead generally much faster +I just checked in a fix for this +Stick with the simplest +Meant Uploading not Uloading in my previous comment +Still what is the best practice to calculate Q matrix? From AQRSounds right to me +No unit tests added for these logs +here is a Simon! +Or you can find it on your file system directly by going todocsspecswsrfdocsspecswsndocsspecswsdmOnce you have these files in place you can run wsdljava and the WSDL will be loaded +so it can be marked as Closed. +Bugzilla cleanupSeems to be fixed in the current documentation +Patch for routing all reflection access through privileged patch looks good but clearly we need to deal with the prevention of misuse firstIm still concerned that this approach can leak permissions we need to get Anil to review too +Unfortunate side effect was that you need to be careful with Tomcat now. +Using the e seems reasonable it is the right version for kepler after all +Do this patch need to be applied for as well? If that is the case I can report back with the performance numbers I have for the GB input data scenario of +Daryn alternatively could the expired tokens be ignored and not written to fsimage during checkpointing +Note this is a TODO item it will never block a release +pushed +Do you know what loggerlevel I would expect to see messages at? Many thanks +Hmm not sure exactly what kind of test you mean +Fixed in trunk +Closing +See the new patch file on. +I think have problem thatthe message of exception should be shown to the user +Though I dont know if such tools will really care about push eventsI dont know I do like the idea of making it harder to do the wrong thing but maybe in that case its not worth it and good documentation would be good enough +It appears that none of the other portal properties could be specified as empty before which is why it hasnt come up sooner +Would also make dynamic analyzer selection easier +The patch has the changes to the protobuf generated file as well +Iirc we had some discussions on this and limitation on based fileupload was part of it +Fixed NPE while handling exceptions that dont contain Cas Reference to attach a patch for release +Still doing development in a somewhat awkard manner with examples +maven I guess users could select or +It did require one test change as simulated failure by only marking the directory as failed so a subsequent write from edit log would not fail +bq +Agreed regarding comments on patch +Ive also found a bug in the thrift PHP implementation + in revision Vladimir please check that it resolves the thanks +Its against the ASF policy to distribute packages containing LGPL code +Hello I am interested in this issue +Applied patches in r +If Hadoop is throwing exceptions because weve used too many counters lets catch it and log it and move on +Please take a look +We could support them all IPADIC NAIST JDIC and +for the fix +The builds may need to use some other option because the maven community was not very positive about making the change. +First cut at Solr packaging +Thanks for getting back +I also checked in a change to the releases page to add information on getting Pig rpms and debs from Bigtop and getting stuff from maven. +Yes now I see the transparency grid me too +Thank you for your messageI am out of the office on vacation for the remainder of the week +bulk close of resolved issues from releases prior to +In patch creates only one instance of +John a patch is very welcome +I should be more careful next time +This one is pretty old and weve done several iterations of improvement for backtraces since +Hi David Which JDK version do you use? Thanks a lot +Florian As Daniel says we dont typically allow the use of custom domains for things like this +This patch has been rebased after the commit to for the rebased patch +Thanks Eugene +Im not convinced DNA needs to ship with any JAAS implementation but we certainly do need to have tests that verify we do work with a JAAS implementation +I forgot this handy process. +Excellent stuff thanks Pieter +unix shell account is not linked to the JIRA account +ETA is +Sounds good +Can someone else commit? +Sergei Ivanov Whats your alternative? Could profiles alleviate the need for multiple? Is there some other wayThe other point Id like to point out is that Mark et al did agree that the previous behavior was theoretically correct +In our case we have extending Block and we dont need to define equals in +Applied patch in r. +Thanks everyone! +Verified the workaround with the modified +Uploading a preliminary patch for MR +Version Final is released +I will write up a few regression testsUnfortunately the greclipse svn repo is closed for commits right now +If it ends up that there is a need to remove data behind external table a question can be asked why do you define it as external tableSaying that possibly the proposed syntax and semantics are not consistent to external table use casethanks +Cool idea +By comparison if we break the cycle we can compile common test common compile hdfs test hdfs compile mapreduce and test mapreduce +It still fails in the split code but it should get you to new ground +Lets target this to be done for +You encountered two bugs Typo in the manual +The default criteria approach has a serious drawback in determining when the appropriate criteria has not been set for example if the isDeleted column could be used in an function or predicate +this must only occur in some situations because i dont recall anyone noticing this problem before +In another world if we were starting over all hbase components would have state and even history a visitor could iterate all registered components and then dump the output to UI or to monitoring tools via JMX etc +fixed in both m a little confused +Cause was found to be some classpath manipulations which were subsequently making network calls +for the ve committed this. +There are no obvious problems in the code that can be easily fixed to improve performance this can wait till next about version was it faster +I did however look through all the files I changed at the findbugs warnings for those files and I was unable to find anything + +Please fix this bug it makes AXIS unusably slow for big XML files +Please check that the patch was applied as you expected +Id like to remove that +or +TakeshiRegarding levels Im not wise enough in the ways of loggers to comment started a thread on which hopefully can kick of a discussion +In terms of implementation I have no idea which of the two will be easier to do +See +Edisons comment implies this is only required if you are using Xen hypervisor +Bulk move +What is the direction on this? It seems all the patches regarding Windows platform are not likely to get in +verified by Vladimir. +reopening to edit the desription +Makes sense to me near everything is an IOE at its base in hbase its hard to distingush the retryable from the IOE that is fatal +Added patch for release tag is affecting +I will try to reproduce it +Lets see what we come up with and then consider whether the result should be merged back or perhaps copied into a new package in trunk +Will commit when tests pass +For now this patch disables stress tests +This shouldnt be the case anymore +to apply the patch as is +The context will be closed if the token is expired +commited thanks! +Run as +Perhaps we need to do better with our welcome wagon +In the case where my code needs the custom impl and the current thread has not passed through the HSCIF this then fails +the package names of the classes need to be for the decoding problem for branch and thanks for the great patchFor me still fails on the branch once I figure it out I will commit the patches +We serialize the object since the JMS spec states that changing the body of a received message shouldnt affect the sent message even in VMJMS spec SectionAfter sending a message a client may retain and modify it without affectingthe message that has been sent +but only handles a single validation case +These are usually just for code that only one line ever reads the config but it is still thereI am for anything that will help to clean up the configuration +Contact addy changed so dont worry about that one +Hence we are deciding in favor of returning null in the queueWith this context I looked at the new patch +I have attached a screen shot for reference +tests in jst and javaee components dont show any new editors related issues +Very minor change +warning +I picked up the work from a contractor who told me he was using the latest release version but it was not the case +Here is the patch for this issueStep to test this issue After apply this patch create a order and give your email id +Minor update to the patch that did not apply cleanly to latest +be fixed in according to +Could be slow if many version are to be retainedTheres also one other options Use upsert if VERSIONS is set to and use the normal otherwiseIn both cases Omid type transaction engines could still workIm going to take it as a given from now on that we want to keep the upsert logic as the default option +Will be in +Ive raised to track this +Typo correctedThanks for the report +Seems to have been a similar problem on Windows +By default it will use your local repo as it should +Looks good you should ask jzb Radhika Jessica for their final review for syntax wordings styling etc +uploading a new patch that applies to the trunk +These optimizations were added because they made big improvements +war until the XTS service has started but as we are using a that may is fired as part of you could ask about this on the AS forumPaulCan you provide a link to the discussion? I couldnt find +The applied patch does not increase the total number of release audit warnings +At this point there are still some outstanding issues with certain tests but these are being tracked under and +putting on next rev to fixexplore options +The setters mostly look right although some of them are missing the assertion that the job is in the setup phase +Woops heres the patch to add +We just truncate the active segment to the start offset and the data +on I still cannot due to the freeze Ill back port as soon as I on trunk was at on as well +Im just curious if you have tested this approach with and without the? The kernels fs scheduler is already performing so my expectation would be the improvement is largely or entirely due to the fadvise to drop pages +BrunoCould we change the CSS class name to and then use that class instead of H not in addition to H? Just make the class a duplicate of H in the other themes +Also removed the duplicate label and its references +Then I can also get rid of the ugly SVNREVISION initialisation which was only needed to track the version of the listSince that was your only irk Ill commit the changes directlycheers Roland +Thilina can you please have a lookThanksIs this fixed yet? test with Axis RC if the issue is there then please the issue with the necessary data to the issueThanksDeepal +for the patch +Attaching a patch for this issue +enhance to check that this problem no longer affects patch to extend to check for this bug please review and commit if youre happy + +We know +JAAAAAAAAAAAAAAAAAAAAAAAAAAA +To be specific do you agree with the following Client checksums the data to write and sends both data and checksum to DFS +Also has the support for saving the data +and is a release thatis allowed to break things especially since checkparam is available +Did you redeploy the plugin? separates view fields from view attributes +Id still like to have the +This patch implements the proposed improvement +not blocking alpha imothe PR from snjezana does not fix this issue it just work around it once more by adding more data to the projectexamples instead of actuallly using what is in afaics +Only small thing to add is support for the isMap entry +Yes its my fault +So Im puzzled where to add the test casesPlease give a further advice to help me advanceThanks a lot +Hi James I think we should add a entry for this fix +I mean that I think having a Comparable enum for is great however I think you could easily support Java as well right out of the box +in anycase will it going to be a long term solution or well see the same problem each time a new version of is released? +This has been reported as still happening +A flat more contemporary graphics nice! Bravo +start udig +I dont think youve demonstrated that using a generic checksummer will make things significantly more complex in HDFS +Could we name the scala file? Also the message in should probably be something like topic is not allowed +I encounter the same problem with RC +Other hunks appliedPatches applied to branchesSending applicationssrcwebappSending applicationssrcwebappSending configssrcplanTransmitting file dataCommitted revision trunkSending applicationssrcwebappSending applicationssrcwebappSending configssrcplanTransmitting file dataCommitted revision . +adding quicstart will try +Sure go ahead +Oh I believe it +Bug explains all the observed behavior +Cleaned up test changes no need to disable asserts simply allow buffer size to be passed in so we can use something large than existing default +but now it proves to be a wrong idea +Looks good + Bug has been marked as a duplicate of this bug +Click the check box Mark Viewed in orderdetail page +Its working as expected +This is basic patch +This patch maintains API compatibility and gives access to coreconfig via +I committed the patch to trunk +In Chrome too +Thanks for fixing thisWould you mind adding a unit test that demonstrates the fix +WellBefore going full blown on statistics a simple extensible audit trail would be nice pretty much every commercial workflow toolset has this available inone way or another in both their standard clients as well as in their apis +Is this good for commit? If so can someone please commit this? Its failing a lot of builds +HelloWould you please try my patch? ThanksBest regards Andrew +NOTEI have a typoerror in the attached of private static boolean winner trueUse this instead private boolean winner trueIn other words the static keyword must be removedSorry about that! targets JSF +Hi i dont look at virtual memory but memory +If the stream is not closed somewhere why is it null? Never initialized +Hello PaulexThe fix looks good +is not an exact duplicate of this issue but since the main concern raised in this issue has to do with getting the error message to state which row number in the input data caused the problem and did just that it seems reasonable to consider this request a duplicate of in that respectI agree that it would be even nicer if the message additionally specified which column in the input record was problematic but just having the row number is a major improvement +Looks good +My review comment about hiding and handling of clearing parameters etc from is my only comment for now on this patch +Do you want add this to release +I will test one of the nightly builds but unless someone has checked in a fix this would still be broken +Replace all URL space substitutions with the original space character +my specific usecase was to be able to run what was in a mvn execjava goal without having to wait for maven to do the launch externally +Please fix your software +Yes I will give it a try with codebase today and report +If a user doesnt specify fs name it will use the default name by the config +That can possibly be addressed as a followup +committed to trunk which I assume has not yet branched even if it is frozen +But before this issue is put on back burner allowme to say that I find standards compliance of mainstream browsers in severalareas simply appalling +Attempting to run the jruby specs just sucks up all my CPU with yes +I tried released GEP with Ganymede SR JEE edition it just works fine +You forgot to use the jsonPrefix variable in the writeRaw function! +Updated patch adds j option to produce JSON output +Be careful currently Camel feature use Karaf one and Camel feature will use Karaf oneIt doesnt match the Karaf version included inIt means that the Karaf version required by Camel will not be present in the system repository +The Annotation Editor is not sending it our currently we would need to change thatIt should be sent if a piece of text is selected in it +Patch to fix moves the logic to a method and adds comments +Also please attach a test case that reproduces the issue so we can track down the for the time being + +It tried to note each transition a region went through +Im not sure file uploads are even supported for ajax submits +I will wait for Toms anyone elses confirmation on this +Assigned for review to Martin +This class defined what fields will be displayed on the tool options page and is responsable for updating the preferences on change +When were indexing we also dedup them by domain which is much more expensive than creating an iterator +Attaching a patch for review +Those can be added in with the improved middleware benoit is proposing +Leaving open issue for HB +ok I did it +The best way to fix this is to make use of the FQNs of the so you can skip the imports +This bug has been marked as a duplicate of +Sure +patch adapted for current trunk added test and fix for visitTree to Martin Koci for provide this patch +This is being handle via another jira I think +Please give the new build a try and report if you still see this issue in that build +Thanks Nathan. +true already set still i am getting this is an UI issue so assign it to Jessica to take a look +LukasYes IIRC Ive already commented on that Metamer behavior previously +But I just got that resolved using a nd lookup using a pseudo lookup +doesnt support Java yet +good to committed +The changes look good to me +combined querymetadatawrapper and transformationmetadata +Jonathan Whats the easiest way to find that out +Thanks +This is a check that the groovy runtime performs +A lot of methods have to aware of the reloading functionality and if a synchronization is missing this will lead to trouble +Adding the Patch with latest rebase and change Container status to Container state and add container exit statusThanks toString optimizationThanksMayank +I understand the desire but it would definitely be a lot of work to add this to all the supported Thrift languages +In the end verification failed with lines +iPOJO Core was released. +That should make it easier as you can just keep the edited version around and copy it over +This might be a side effect of s path variable support where we revised the pattern matching code in the course of it +The fix will be in the nightly shale build +In this case I have now a testcase for class level validator but generally the more the merrier +Alex I still face some problems with this one +Unfortunately just moving the stuff doesnt work because the test componentsfor the CPE stuff use the same classes as the test cases for and other parts of UIMA +Changing the priority of these issues which do not seem to be major issues +attach a jconsole screen snapshot to demostrate what we can get from the CXF endpoint! +The mentioned test with now uses a fileURL +So it does not look like a jboss issue so its ok to close it as notabug from my part. +Committed to trunk! Thank you for your contribution Edward! +It currently relies on the Rackspace javacloudfiles api to communicate with Swift +we should probably try to fix this before the new release +Think that should be changed +Ill try to convince Dagmar to do it +I forgot to remove the quotes from the option values so I added the fix for that. +Please use the mailing lists for help +From what i can tell this bug was fixed in Solr as a result of other changesfixes to replicationthe attached test patch passes on trunk and x but similar assertions added to the test fail. +Id love to test for this since its obviously not super commonly used but has had a couple of regressions +should take in consideration where clauses in case of sql query and parameters passed in case of stored procedure. +It would only affect very specific cases when someone has extended these mojos +please ignore the previous one +Maybe we should add a release note to give the users a heads up about this change +Yes I will provide a ZIP file with an XML export +I will push to the jbosstm directly +this problem does not only apply to hmessages but also to hmessage fieldsThe main problem with the spec as I see it is that it defines what has to be rendered for the standard components hence I said this is a spec issue as wellIt might be that by rendering additional code we might break the TCKBut I guess others have more insight into the issue here +It should be not +Closing all issues that were resolved for. +switched from improvement to bugFor our current build process a lot of builds fail because of the missing synchronization between different ivy instances acting on the ivy repository + bbrowning lanceball I see these as two separate things bbrowning one is handling case where instance was stopped externally and the other is handling failure scenario in our stop request lanceball good point bbrowning one is a permanent failure one might fix itself by just retrying again later lanceball bbrowning perhaps we need both lanceballstopfailed andwtfisgoingonwecannotfigureitout bbrowning I likegone orunavailable for thewtfisgoingon case bbrowning andstopfailed when we encountered an error while trying to stop it tcrawleyunreachable lanceball ohh I likeunreachable lanceball pretty clear what that means bbrowning havent seen an instance left running because of failure to start in a long time +Entering on the find field will return all plugins that start with that name +The didnt have this problem so must be something wrong introduced in the CR and Final release +Youre right +thx fr the patch venk! +I meant launch configuration doesnt changed in this case +Need unit tests +This means the bug is fixed isnt it? +Thank you Benoy and Laxman. +incremental cache reports This is listed as a maybe in the design doc its something that we want to evaluate before doingquotas metrics and TTL are definitely I think +The password entry is invalidated at the same time the session is invalidated +Were getting to the end of cycle here so I dont think this is going to make it in a real form but Ill add the catch anyway and well come up with a test case +This could end up being a complicated change depending on the kind of configuration information required by your providers consumers +vof is correct +thx for testing itim not sure if it isnt just an issue with our +we probably do need to try and fix that somehow +Closing Jira. +Theres no patch available yet +That would give us enough wiggle room to switch to an alternate implementationSeems like a good way to go forward +Not the mixture I thought but still my fault +We dont seem to have any file named for this build +This is plain compliant with the LDAP RFC +Since such a feature might be of interest to other people I thought about requesting it as a feature +Nice! Thanks Alexey +Is this issue resolved to your satisfaction? If not whats left +as this is clearly not part of the specs +Jimmy the removed code it is not needed because if region we already transition the state to offline state in the exception handling partthere is a possibility that state can be changed to OPENING in handleRegion after setting OFFLINE in exception handling part +I am not an expert in regards to that aspect of Lucene +is now in including applying our patch to fix +I guess the problem should be fixed as per as fixed +Heres some incomplete doodling on this issue +And why did you remove it +Thanks for reporting this +Pascal or Camel case for whatThe logic of using adaptor as a marker for adaptor is that its consistent with the way Hadoop marks its identifiers it looks like job and task and so forth +Steps documented in the JIRA were followed and the resulting changes committed +build the latest Harmony jre +The UI for this was fixed previously +It is not related to shared libraries building with static libs reproduces the failure +We need to close our issues first and finish our refactoring workWe have not discussed it deeply at the moment but my mind is usually thinking somewhere by autumn or beginning winter this yearPlease join up the mailinglist if you want to discuss this in detail or have an opinion about what to include or not youre welcome +Affected widgets have been changed please test if everything still works as expected +I have checked in the patch with some modification + +We made an exception for regular expressions because they are well understood by a wide variety of people +BTW had a chance to see if some code sharing would be appropriate? +It shouldnt be +It solves all the issues in one fell swoopQ How often do you think you might update the zone? I ask as one of the current public NS hosts can be temperamentalA The zone updates with rule releases and those are getting more automated +Hey Kihwal +Also it does depend on the shard approach since some shard by date +Well yes but my point is also that the only negative behavior is a lack of exception translation +Ive committed this and merged to +I killed it and reran the buildProblem went away as a result of thisProblem still occurred but less often +Interim solution +However I hope that our release manager will be willing to delay the release a bit if it looks like a little more effort could fix these outstanding issues +add settings for read op sizes make unconfigurable ones more t we just use for +integration into the code branch a directory layout b ant vs maven +This patch changes the meaning of the resource name argument for an +Patch fixes the test failure reported +This happens when using +What are some cases where a value needs to be overridden where using an value as the authoritative value will not suffice +Attempted to correct the invocation of this specific test case using multiple surefire test executions with different fork modes but then maven failed to launch the single test invocations correctly +Sent DM +And maybe I may add you list of issue keys where participatedYou may then send Thanks emails by some standard waySecond option is create more sophisticated plugin which will perform complete mailing but this seems rather heavy for nowSo is report with list of emails enough for now + + for patch +Current CVS HEAD seems to be fixed. +Thanks! To that end Ive added a new patch that now Removes dependencies in typessignatures to the implementation Moves many comments to the interface and leaves Javadoc references to the interface in the implementation Removes compareTo from the interface definition and instead has the interface extend Comparable Adds the ASF license to the headerAny further improvements +Can be closed after is done I need to check if everything is in place +not sure why jira didnt list this but the precommit passed on jenkinsThanks Thomas! +If that is the case the fix will not work +Reopening to change fix version. +I am a bit confused about the options to use either ICU or font info +The new patch looks much cleaner thanks +A couple minor tweaks to the patch including adding some javadoc for the new methodfield +I upgraded the Tika dependency to in revision +If we elect to abandon in favor of and since all of the underpinnings of this issue have been reverted perhaps we should now close this as wont fix? +We dont have that many so we try to take care of them +Verified in CR testing cycle. +for now I added a check on named list in assertion and will now skip over result set processing if there are no +In the case of JEE Web project it is about entries +Hi Pavel MirekIm taking a look at this issue +but maybe Im wrong and it might be also needed other types of Beans being tested +or I suppose it wont really +Thanks Steve for the inputbq +Make sure Hudson agrees +I was just looking at this today hoping to finish it and up you put a patch +Fixed that and +I reproduced it while messing with spring security which is good +Example diffs showing the simple change that will cause the code to compile when included into c code there may be other files that need changing also these are just the two I had to use +please try with latest snapshot and if still doesnt work reopen and attach a proper quickstart +It does support COW but if it doesnt include changes to how files are named it will still need this fix +Id not be happy with this being released tomorrowCan you give more details here +Any Haskell users want to weigh in +i think that the may resolve the thingJT doesnot monitor the status of the every job and tasksbecause many TT must RPC to the one JT every few secondsand many clients get the jobs status through rpc the one JT every few secondsso the majority nodes of the cluster connect to the JT discontinuouslythat lead to degrade the performance of the the number of cluster increaseexample Klike hdfss Federation Branch the MR project must create a new branch for the +Its the default exception in Java impl when none of the declared exceptions are a match +I wasnt aware of thatIve not modified the POM for the XML module +Manually test added. +Committed to trunk and Thanks for the patch Anoop and to Stack for his review. +We use JVM and JDK to avoid wrapper problem in bit PapapetrouIve just successfully installed Sonar as a service on bit Windows using bit JVM and it works fineAccording to attached log file you have problems with deploying checkstyle plugin +Yes + does not have size +you could base it on latency instead of pending count +Its the reason that I said I dont find how to configure js calendar to support this functionality +running +Interested in comments on this +Thanks anyways! +those were the ones related to is this progressing +All passes +Is this only happening on vmware? I tried the steps with xen server it is working fine +attaching a booking example fork which uses session beans and a applied into SVN renamed to jee am closing the issue +Fixed in r +I mean any problems in the implementation of just a reminder +When I change the file to have CRLF line endings the comments are generated +I agree that it is most interesting to see eligible beans but it also might be interesting to see which beans are easy to become eligible especially when it turns out that currently there is no eligible bean for the injection point in questionList of eligible beans does not include disabled beans and alternatives that are not selected and those beans are potentially ready to be made eligible +This is not a bug as stated during discussion with Lincoln +pinghas anyone had change to look at the result set iterator addition +Looked like only branch is supported +Ill quickly roll a patch for that now +Create patch for branch new patch for all Yarn test cases patch looks good too +Koji not a big deal +Thanks Christian +This should be present in ER +Attach the code with an example project that you claim is not working +bq Ill commit to x and it will be easier to work on thisIm working on the Maven build and am almost there do you want to wait for a revised patch? I also made some minor modifications to the Ant build removed dependency from resolving test deps to instead of lib +I also tried to run a profiler but I couldnt get it to workI give up +neither of the suggestions makes me happy but that might just be my state of mind having to find a better name +If the user wants to use standard they can enter it +Please check this helpsThanks for the contributionCommitted revision Andun could you please check if email sent has correct From field +With that patch I attempted to get around the need to have the get permission at all +Chapter is Virtual Deployment dont think this information is useful for administrationconfiguration of the server +patch to hackaround xdoc problems by tweaking +Anyway you can usepackagesexports packagesimports and dev commands in combination with grepAnyway Im not against new dev or packages commands to have a better overview of the bundles packages resolution +Re Extended label mode +AnilYes this seems logical to meAnil if you want to vote please use the link above +Well I dont think doing a null check here will solve the issue since mwires should always be for a resolved module +The initial use case is definitely solved we now detect classpath prefix in the DTD in Core +My suggestion for the predefined drivers tell the user what dependencies are needed and require the user to install driver jars with exactly the repository id the driver gbean wants so no jar selection dialog at all +The start flag can be used to override the load flag on dependencies that happen to be plugins +for with this was what I wanted to say with my complete nonsense comment +Torque only has queue +Hansplease enclose the test case for recreating the problemthanks ran the testcase against the latest build and the serialization of the exception still failsI would greatly appreciate it if it this would be fixed forIs it possible to mark this one as a blocker +The bug description says that a new blank database was created in so it would appear to be a vs performance issue and not an upgrade issue +changes for winutils to launch task as domain user +Marking as a dup of. +I had failures in and +I have a patch for this +Thanks Todd +Or javac should be possible too +Thank you Mario Martnez Snchez and Alan StewartWorking fine. +Meanwhile weve possibly missed an opportunity to save a block from early death since a copy of it is still actually availableIll probably open up another issue around this later one once the current one is closed just to clarify +Ive added a printout to the Driver to note this in my next patch. +I will try to come up with an ACL matrix with expected and actual permissionsHope that helps in identifying all issues related to ACL +This only mattered if the stats file was missing completely if it existed but was a version earlier than h it would EOF out as mentioned above +you very much Gopal! I tried it this weekend and it seems to work now +The conf in running balancer is not initialized well so left as a null value +Concur +Hey CarlosIm trying to run the tests and its failing because TAPHarness isnt availible +Committed to trunk and +Ive applied itWhile the Activate annotation is not really necessary as that method is using the default method name signature Deactive is required +Now looking at the patch attached above Id love to be able to run a test case that allows me to replay this issue +Interestingly enough I think that the Acegi client used to work with programatically established usernamepassword on this class something may have changed in either Acegi aor Spring since I first tried it back in December will follow your advice and look into creating a remoting proxy programatically with unique credentials to ensure that clients are handled in a thread safe manner +I just committed this. +So maybe the maxbatchtime is not handled well in that case +Ive implemented my own Giraph jobs about times already and have only been able to recycle IO formats once or twice +The changes do not affect the quickstarts just the product and qa build can use the target in qa to create your server instance if you wish but you will still have to setup the properties in you choose to use the embedded server then you should use the following config localhost esb esb +UI wording In the code its but the UI says +Im not clear who has the power to update these snapshot releases +I committed this +We are running +Downgrading to major if their is disagreement let me know +I will post the new numbers with two changes +after creating the symlink according to the direction mvn install succeeded with the patch on my Mac against Java +And I meant to say I knew my batch file doesnt have relative paths I was just copying and pasting to get it working +How about or +When the new build is available could you install and try it out? Let me know what file is causing the problem and maybe there is something about the ast transforms there that is a causing it +first +fixed in close for the release +That will unblock us for failover testing +Needs some fiddling + +I agree. +its a later issue is being addressed in +attached to hibernate search was also upgraded to +Its fixed +and the screenshot clearly says AS not switchyard +We can try to extract components that people feel would be beneficial if there is interest +Feel free to in case theres something that needs to be added. +This happens even though a second namenode is active and providing service +Waiting for metamer and ftest for archetype to be released +I think it sounds like a reasonable restriction to only allow LOB values to be accessed once +So it will help people who have or more DC but not with DC +The problem is that base revision commit can be invalid +This can be closed as I cannot reproduce this w what is in trunk +Patch contributed by Vinay Shukla +It will be easier for us to commit this if someone will create a patch and attach it as a file to the issue checking the box +so while you are most probably right about the linear performance you do not have the same logic as before +Thanks for the quick feedback John +Im just going to let it finish running +Patch looks good +committed to trunkThanks for the contribution Deepesh! +The appropriate templated +Max allowed age I Current disk usage +Dear Oliver Nice to hear your voicesI suspect I meet the same situation as Mark and blocked lots of you please review my post at I updated some feedback there +for the patch +Patch was generated with Solr looks like a dupe of see that issue for more history +v passed all those three cases on my box now +Tested the patch with various combination of maxCap userlimit and queue capacity +If you want to include this with an upcoming release I can raise its priority +Im sorry about that. +When is this project going to start addressing some of the major outstanding bugs that have been around for almost years? This is the sort of bug that makes Hibernate unusable unless you have full control over the details of the structure of the database +Another thing I wanted to ask was whether you tried to simply subclass +Ie we could throw an assertion error here the catch clause here is just to work around the weakness in this particular API +It will be problematic because this code is property of our customer and Im not able to share it with you directly +implemented the review still see some inconsistent formatting +Unregistering this instance +Added Lance as a watcher Lance please see my previous comment +v also updates the added for the extension actions +Thanks again Herve. +Committed a few changes with revision which removes the warningsAnother patch will follow later +the class was refactored yesterday +threads will not persist modules and classes should persist only by name and be looked up and symbols should persist by name and be loaded from the runtimes symbol table to guarantee atomicity +All the projects that depend on you wont notice t that mean that every project that uses maven has to run through the entire dependancy tree for the project to ensure there arent duplicate jars being included due to differing groupIds referring to the same product and if there are some every project then has to maintain their own exclusion listThis seems to go against the goal of reducing the workload of developers by removing the need to check for jar the right thing to do is to fix the repository to ensure there is a one to one mapping between groupIds and products +Not to be committed. +Thanks xulei. +I committed this a few days ago +For messages are NOT recognized on my environment +So virtual router element returns true +Moving all issues to the CR +bq +Im not quite sure I follow +Tested at passes fine now. +I guess it has been fixed sinceYour patch is intrunk r r r +Anyway now fixed + sounds good +Will this make it until Friday +Theres a tld stylesheet that does all the work +I do not have access to a server in Europe to test against +So basically this is include always requiring one argument? Weird that this has never come up before +Almost all this code is different now. +so the needed to be copied into the build directory where the unit tests are run from +Axiom works fine with +Yes but I am using since I want all versions of a column to be deleted +other places doesnt really help us fix anything +If there are no objections Id like to commit this soon +We still need a way to handle atomic operations for bit that dont have a modern gccWe have the option of using the linux glib or apr atomic functions +I can add this but I think it would be better if the jbpm guys could create their own schema +verified on +H test all pass still and I have ran all postgis tests and only lob tests fail which failed for me before the patch as well so I believe those are unrelated failures +Will investigate that but it has nothing to do with this issue +This patch still depends on now that has been implemented Ill the patch so that this contribution is placed in the contrib dir +Steve Ebersole I can confirm this with JDK and Hibernate +Pattern check against null now throws NPE and is consistent with Java +There is a test case attached if you want to test it +Ill let my tests run a bit longer +Patch to force usage of supplied client certificates r +But when the feature is not there they cannot decide and they are forced to use update site or other installation way +Alex awesome +orphans should be set on the outermost block inherited by the first level and reset on the second level to trigger that particularAlso reformatted the source file to reflect the eventual line layout +line linem on it. +Patch applied with modifications +My plugin after upgraded to release requires the class in the jar to work +I am playing around locally with creating a copy of the ast that will remain concrete +Committed to trunk but nothing in this patch fixes or modifies the situation for RSE +But it still knows what nodes its on which it switches to a nested plan etc +Hello Igor Since you discovered this issue any comments? RegardsKiran guys dont repeat useless information already present in previous tickets +Closing stale resolved issues +I think well be talking about this in detail when you are here in Vienna +Sends appropriate PUT to host endpoint to remove a site exception. +Would it be useful to just create the executor once and reuse it for all compactions then get rid of it in a shutdown hook or something similar? Ive attached a patch that does roughly that if you want to take a look +Funny that you guys would get this error at such a low number of workers +A workaround for this issue is to connect to the database and make the database metadata calls that are needed once before jaring up the database +with this fix all async tests passing +I just committed this +Im sure things were clean when you wrote this +It runs stable for almost a week now +New v patch which has all my changes from as well +It would be helpful if you could attempt a test case that demonstrates the issue +My approach is to always use the most restrictive qualifier in the first place and only make a method protected or public if really required +Thanks for the helpKelly +Alejandrothat the test passes sometimes is unfortunately not good we document that the functionality covered by the test is not available with HSQLDB or we fix what needs to get is fine with Done means it got fixed it makes it into the release Wont fix cannot have a Fix Version it does not make it into the release notes +Thats excellent Ram all good stuff +Removing from +Its the java code +Thanks +Verified in with +We should split this jira for each release so we can track what was actually fixed and what remains +even if we convert the to an array and serialize it we can not sovle the none enumset problemSorry that I was not clear +My take at this is to drop the mock classes completly and use something like jmock etc + +Seems like the plugin isnt being copied into xFactorPlugins folder +Attaching revised patch +Because the only reference to the is held by the there is a risk that it gets garbage collected at any time + thanks +Hi RobAssume that user is authenticated using non + Username tjiiii +Indeed looks promising +Yesterday the only feedback I got wasException in thread mainToday the tool decided to be more verbose withException in thread main Java heap space and a stack traceIncreasing the max heap size to GB solved the problem although this is a lot of memory for parsing GB file. +Thanks ChuanI already reviewed this patch so from meThe change builds fine and your test passes +Please investigate on your windows machine. +Lets fix everthing that we can immediately and leave TC for another patch +Hi AnimeshYes I was planning to look into this in the meantime I will lessen its priority +patch works patch including both the code change and the unit test +were backported to as well adding the latter fix version +With ASF inclusion +Patch file for correcting that one of the tests got caught in an infinite loop with my patched version because of extremely small distances +Attached updated patch as per with small changes to whitespace at of release. +Dmytro task over? if not what else do we need to do +I do not understand the problem eitherTo me the problem is rather mixture of internal API with implementation in the same package which should be solved + Username srivatsanp +Ive committed this to and +patch applied in rev Thanks +Hi Knut thanks for the comments and review! I kind of lost track of this issue but wouldlike to return to it and get it completed +Tested +Anyone have an issue with me committing this +Phase ordering doesnt seems working properly where I made my first attempt +Graham I need a confirmation for your particular test case though. +thanks vinod +Please let me know if it can be closed or if youre still able to reproduce it +It adds a thread for each which periodically reaps from thread which dont appear to be alive any more +Looks good +John have you verified it again? is it working now? +Proper label was applied on master and in JBDS B + to Rahul who has been working on this one +Duplicate of Sorry I meant to post this in. +That doesnt make use of +Thanks Wang for your contribution +Attached is the zip file containing contribution forThis zip file contains a package called which contains implementation of default required by the package +Ive verified that this suggestion reflects best practicePending my getting some time to work on this if someone prepares a patch Ill commit it +Committed to branchx in r +Yeah my intent in that post was a way that works but most likely does not solve this problem in a standard manner but perhaps some or part of the ideas might help form a solution + +Yes since Rob had already posted a patch on I continued from there and must have forgotten about this ticket +When I set to it works as expected +I agree with you +Its worth filing a separate bug for SPDY push support and discussing how you would want it to work +Has the syntax changed since? The namespace shouldnt be updated just because the Cocoon version changed +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Fixed in and on the branch so if we decide to release another SP it will have this fix +NPEs we dont like could you attach the log from the new buildmanaged bean issue? ok new jira issue for that please +Committed at subversion revision +Can you not just have your client specify timestamp of I hope this can be done in open source hbase and can be pluggable +In fact it is not solved Being able to specify a custom delimiter at startup is not the same as being able to switch delimiters when context +Fix is for VERIFYSUSPECT to listen for view changes and remove timers for any members that are no longer in the viewIn the scenario in the description it would be removed when the node is originally kickedand would not be kicked again +The publication is not as straightforward if Maven isnt used to build the artifacts you dont get the same checks and guarantees that the product is consistent with the POM +We propose to remove this feature and do the work in the programming model as per EG agree here +Tika is still in incubation +Implemented by adding +Ups test jar unzip with jar x. +Kan could you please dig a bit deeper and get the explanation +has been released closing issues. +on the patch +and for the Helloworld one +This was posted by me but apparently was logged with an anonymous account at the of? +Verified in +Fixed and test case added. +Thanks for clearing that up +As discussed this issue is beyond just and actually requires some updates to the JSF Spec +Works for me now! Thanks! +reattaching trunk patch since Hadoop QA wasnt picking both of them due to same timestamp +Should add saving parameters and clean code +Resolving as the dependency on has been resolved as well +Thanks! +The current widget test testing results. +Moving out +Since this seems to work for you now I close this issue for now +Timmy I ran your and it seems that the cypher results contain column but then actually elements per row +I dont deploy from Eclipse but I do compile and run from Eclipse +BTW could you please also ensure that this pattern of code is not repeated in any other places +We can upload the izpackexe and izpackapp as zip classifier stogether with main jar to maven expand izpackexe and call the python script form there +Good one though the thrust of my message on the dev thread was that we ought to do argument checking across the entire codebase this way or not +Thanks +We now provide that in Roller +Thanks +Any inside the property is escaped to before loading loading into the properties. + +Great info Guillaume thanks for thatSo at the moment if you have a set of jars you can work around this by copying services into your own bundle would there be any other way to work around this problem for cases where a customer doesnt have access to modify a third party jar? I think thats the problem here figuring out if theres a low touch solution to support customers that have legacy jars that arent aware +furthermore import jsf should avoid to exclude existing files like +We also have quite a few articles and videos on the site as well as in the transactions teams blog +Do you think that it can be related +This JIRA depends on core +Note that the dfp library is released under LGPL so cannot be included in any ASF releasesHowever it can be used as an optional dependency which may well be sufficient hereIm not as well versed in these copyright issues I as I should be but my understanding is that as the copyright holder of the dfp library I could dual license it +Committed revision +I will be continuing the work in to make the reduce side sort pluggable +I dont understand the issueskip parameter skips the plugin execution not dependency resolutioncan you give a sample output and explain what you see that should not be run +Hithe attached patch fixes this issue for my comment here This patch fixes the issue for for the patch! Will apply it shortly +is it possible to have the right to create components and versions in jira +Adding inject method back into resource class +Victor can you explain what kind of problems do you have in this issue +The extra check is probably not that expensive so Im fine with either of the approaches +because of a UDF +Closing this issue. +Should these be processed and moved to done as the the collector starts up +What is peoples feelings on moving away from the Lucene logo? We arent a different product like solr nor are we a loose port like Lucy +Hi Benjamin that helps +Committed patch c to trunk with revision +Ive lost the motivation for this +i can not reproduce this problem any more at +Thanks Jaehong Choi +Patch applied in revision +Also were not terribly good a being consistent as to what information should be storedIf you are referring to hosted stores +Manually need for IT. + +The tests are all commented in the code +Here is the patch +The patch looks good to me +Can this issue be closed +Many thanks to Ernst for providing this patchRegardsCatalin +Fixed in latest +Unless I am overlooking something and theres an overhead even for objects +The comment for the last patch should read Tests and to insure that equal objects must have equal hash codes +Not for update fixversion to +Interesting Thomas +I just would not even bother with the parse flag though +This was caused by the new DSL grammar parser ignoring the n escape character +I dont have any direct experience with building eclipse plugins but I think I should be able to find my way around +I was able to deploy on tomcat server also from command line after making the change described above +So in this case i would saythat your explanation in the Jira entry is very helpful and clear and I dont think that anyadditional change to the patch is necessary +ok I found an XP box and tested the script +I committed the last patch by Olga + about this one +Attaching a patch fixing paged column iteration +I had Hessian at the server end and at the client end +Thanks +Related to +Closed upon successful release of. +Committed r +The implementation is built off of the apache issues resolved in released versions +and +This does not patch + +This patch was applied to and tested against +Unfortunately those variables slipped through the cracks +I looked at the patch too and it looks good to me +we can use a to construct them +actually MRI has the same behavior it iterates over the range until it gets the value +update the author and authoremail this is displayed on pypiIdeally these could be Apache Avro and +Actually I said it should be the oldest unreleased version suspect its counting archived versions as the general issue Ive previously talked to Vlastimil and Anne Louise about this and it is the way the process is designed that all bugs and feature requests have a target release set automatically +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Thanks for the review Ravi +We need to put together tests for the most important functionality in exec especially something as hard to detect as the limits working correctly. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I looked at the code that sets this and it appears correct. +After that I was able to view from +HelloWould you please try my patchThanks a lotBest regards by Andrew. +This is by design as noted +Denenche you sound like just the person to make a patch +Difference is in the inspect instance page or the normal tasklist +I thought client compiled against would pass int instead of Integer +When NSL is destroyed CSLbecomes current again +Cant reproduce this bug on our richfaces demo site and on the attached pages +This is an interesting idea and might not take too much effort to implement +Fixed by +No +If anyone else plans on reviewing the latest patch let me know for this work I am counting on the review of others but will verify that patch passes full set of tests before committing +Also trying to figure out and figuring out where the lose ends areJoseph fix version to from IvanThe actionURL is needed by Pluto +Yes indeed I was also talking about the barrier only sorry for the misunderstanding +Suresh I created to address your comment +No headersrcextensionoracle all issues that have been in resolved state for more than one month without further comments to closed status +Link to Sling issue +bq +Not a huge improvement but a bit more diagnostically useful +Configuration file used by the instance showing the reporter comments error was a namespace mismatch +This patch fixes my issue but im not sure if the change is correct +We need another patch the is missing from the kit +Yes resolving this issue +The epoch is very similar to an index generation only it is incremented whenever the taxonomy is or replaced with another +Thanks Hairong! +Sorry missed that one +once reconnections make it around to the beginning it will get reset to again +If it turns out checking for +Differences in the validation algorithm Dependencies to Trinidad classes which is not possible from the packageOther than those differences Ive tried to leverage as much as possible but its not muchIt also contains a JSP Tag class and Ive added some annotations to the Validator +Can we please commit thisI am using it in PROD for last few months and it works greatJack XML PHP Ruby dont have this issue since if the field is XML and you use wtxml you get XML normally out of it +This issue is fixed and released as part of release +Verified fix on pig branch + +After first test i dont get it to work +I tried exactly what you said in webapp but the problem does not show upI close it as cannot reproduce +Thanks Viknesh for the contribution. +Copied from the Mail list +I think this should be fixed in +Thanks for your +Sample application to reproduce problem +Thanks +I will add pool support to this change and move the cache requests themselves toIt probably does make sense to create a unique cache ID for each successful new request +I am convinced that the same mechanism should be used in FAULT messages and possibly also in ERROR messages +Please verify +There are some ors in the examples which lead me to think that the list is not deterministic +In SVN +fix the NPE and anded Unit Tests for to me that the simplest solution would be just to assume that the average entry size is and not bother with trying to use the actual sizes at all +correct thanks Ill fix it +This Configuration should be stored in the Session so it should be serializable +Thanks! +Its fixed now and in line with the description CASE A indeed uses more memory +the last patch had a bug which is fixed in this patch +I recommend that people who NEED a Camera work on a Camera Plugin and concentrate on maintaining that since that is currently beyond our scope +If we have the latter with async then we can adjust the createasync to have the same help text and support for all split types +The scale denominator i am not sure about i am not expert when it comes to this so there may be a better way but i think the best we can do is assume a standard screen size +short names +name ls grep sar +bq +Changed that too see +does not seem to be the jira which address the file structure layout change +What were your stepshave you compiled sourceshave you replaced any files +I have to spend some time to grok the whole patch still but if we can prove that that model is better we might as well let that overtake this issue +Does the relative path issue reproduce on or is it a regression? Especially if it is a regression and maybe if it is not it might make sense to file a new bug +I agree with PaulAllen that the current scheme of things do not work well when data directories are not homogeneous +I dont have a workaround yet for that one +When did you commit that? Because the new runs for clover and nightly were passing after my commit raising permgen so should I revert that to try out +No idea what happened there. +Thoughts +If you run each JVM in a different directory it should be fine or explicitly configure the directory to be different for each ignore the last comment +Great news +Thanks for the review stack +This number is less in reality because some ports are system reserved others are need for thepublish calls etc +These were duplicated patch needed some cleanup +The newest patch is the file +I think this is unrelated error retrieving current directory getcwd cannot access parent directories No such file or just committed this +I remember Eli was working on mocking may he knows how to it betterMy first patch for mocked up its doable +Therefor I close the issue +Bug has been marked as a duplicate of this bug +if using the then honor annotation if its can understand if others would prefer the opposite default setting +Resolved for +not a java plugin bug and recursive properties dont make much sense +In discussion w AS team on whether this constitutes a true API +So its a bug in the test +All right I correct it and submit a new patch +Is this even an AS issue? I see as a class in HQ and jacorb but I have a feeling this JIRA is talking about an AS class +Removing them stops internal browser working +This one will also be good to look at Kris +Reopening issues with +The main purpose of the patch was to create a dependencies module responsible for creating the file to be used by all tests +Please let us know if this works for you +That should be fixed now I wrote some more unit tests and did some JSP and tests m going to work on multiple selections over the next few days while Im travelling Ill try to get it in before RobMaybe I am doing it wrong but I extended your Input field to provide a simple and javascript to the default text field onfocus and onblur +So this patch also removes the local result variable +Its hard to use a servlet without having that provided and we shouldnt mandate a particular source for the servlet apiAnd shouldnt the other Jetty dependencies be for tests only +Closing released issues. +This looks like it is working again locally running using SBT and Jetty +Closed as this issue duplicates. +Im going to commit this momentarily +Anyone working on a patch for this? If not Id be inclined to resolve this as Wont Fix +We should though add a debug line to the debug log stating why the pushdown has been prevented which is typically done for criteria pushing but has not yet been added for all other situations +simple eclipse fix on the this to trunk and +Some commits around that before the ticket +New version of the patch which includes the same fix for the Export Connections wizard +The updates that you are requesting are reasonable +That way Logging uses it Weld X logging uses it and in can build on it as well +From discussion that I have had some weeks ago with Guillaume this code should be place in karaf project of camel instead inside karaf +I never used them thought so at the moment Im unable to provide you with instructions on how to configure them +Find attached the configuration the log showing the security exception and a JMS sample to send a message with TTL into the queue +Any ideas on thisDuring the testing I found a bug in the because of which it fails to activate applications when resources are added to the cluster +Patrick currently FAQ does not exist +Please try this patchThanks a lot +Now you can create your own singletons +As a result the configs builds dont work after a couple of configs +but I must admit I feel that my code was inappropriate to use that property namespace +Sorry +I will link a pull request when I submit one but unfortunately I cannot test it with my own Forge build +Sean Does this happen with default configs? Or was it result of local customizations? Thanks +N has taken care of heap size issue +now it uses maven and the commandtarget is different couldnt maven as a minimum backward compatibility upon the error I just did alert the user that she just attempted to access a maven plugin with maven and possibly provide the URL to a FAQ entry giving further insights +I have created the numeric unary functions +So all issues not directly assigned to an AS release are being closed +The binary from module should be sufficient to support both the and code bases +CP fixed in revision fixed in revision +properties file in the correct place +int long void etcCould compare against a list of valid names before delegating. +My initial idea was to add some button to check application creation state so user can manually check again whether the app was created and try to import it +Let me know if its ok to commit +For the same reason jobs within the reliability test are run one after the other +Thanks David for your explanationsI am currently checking how Weblogic container can force the connector to retrieve an +Proposed patch +I think about a boolean properties use which could be put in near the Google blockBut to demonstrate the thing we should also provide OOTB some real addresses else Google will find nothing +Issue could be resolved but may not be see for further details. +Get the latest version and make sure my changes work +for the change +Resubmitting to see if Hudson picks this up +Then attach patches here +Committed +Maybe the code paths are not different for the switch of an upgraded database and a database? In that case we may not need this testIt would also be good to know what kind of test coverage is expected to test the switch +the code is already included though to my shame I didnt manage to test is throughly +After the bug fix in the is no longer removing it from the filter chain check if everything works fine +Thanks Dylan! +This is a patch using my simple replace before plan generation technique +Added an SVN diff as a patch file as well as tests +pending Jenkins +Reopen if think otherwise. +We shouldnt have this prob +Chip Ill post a screeenshot to this ticket +Thanks for the plugin solution weblogic deployment happens without any exceptions +Before starting on this i want to be sure that this is the right way to go delegate jmx calls to the Kernel first and if the target is not found then delegate calls to the Server +I had a look +In the case at hand I was looking for the first of a VERY large set of objects to satisfy some condition so I spread the work among several threads and broke the search into blocks after each block each thread checked whether something had been found +is only used for the client api +See namespaceerr for a simple testRelated bug exposed by namedtemplateerr +Good point Dan +Let me comment on the above mentioned blogpost and exampleIts important for a framework to have this feature as a native implementation because you can see with the other frameworks that these are patched or hooked into the existing +Not working in also +back port this issue fix +released closing. +Yes the original ssh command ssh hostX VARexpr +option strict would check this and also that you declare variables before using them. +to to make sure the json being sent matched what was expected +Unfortunately anything I do here is likely to affect all of Springs pointcut resolution not just this particular scenario +Tested today with code and walked through in debug +What is the status of this peopleDo we make this change +I have no doubt that with all the strange build environments out there were going to find some other issues +Oh I didnt notice +Another thing we should probably do is add a key for the meaning of the colorsoO didnt see this comment yet +defer all issues to +done +I guess you need to debug it more to find out what is really happening +Please review +Hmm +Ill commit the fix +Patch removes groovy lib associated license information and all callout structures +I dont see a +Fixed +Yes because the web server of the staging area does not execute the cgi script. +Converted this to a issue per Todds comment. +If a directive implementation sees its parameters are constant it could evaluate them at initialization time and substitute a more efficient version for evaluation at runtime +me only has a Maven Repository View +issue closed +Ive talked to some of admins and they are saying they never do this +Paul is this resolved as a result of your recent work +I will open a new report +Fixed branch and trunk. +Perhaps the user error is assuming that backslash is an escape character in paths +Thanks! Our current release is already pushed and now available. +fixed in current CVS snapshot update comes +All of the unit and integration tests pass with the changes in the file so it was committed to trunk in r and to the branch in rVan please update this issue with the results of your integration and testing. +Closing because deadlock has been fixed will look at alternatives suggested by Eugene for speeding up index download. +Its very annoying to have to modify the original distribution for each user who want to upload artifacts using + +You are a hero THANKS +Can be +ORDER BY also not conforming +nothing for QE to verify +If there any concern regarding this functionality or the patch please let me know +restores API compatibility by making extend again and undeletes the inner classes +expands fully qualified pathnames for export and import args +Viacheslav I tried several times using your steps to reproduce +That is definitely a bug +The patch has been applied as expected thanksAlexei the test is unstable and sometimes it fails even after applying the patch +Hm I cant reopen either +Dan thanks for the info. +Is there a chance we can have a look at those +I forgot one critical piece! I modified spechelpers to add a helper method +Please consider upgrading to m aware of s older status but is not yet available via the stable debian distribution so until it is available thats the version Im stuck withIn any case the goal of this submission is as much to verify that does not have the problem as it is to request a fix in +Close out issues. +Since Cassandra sockets are Im fine with having being the only option +Looked at the patch looks good overall few comments The isnt useful in the general sense other than debugging this issue we can remove it +Oh +if someone else has a better idea please jump in +Zoo Keeper files deleted Chukwa docs committed +I was thinking about our options and I would like to propose using feature branch in this case +Testing with a recent might show if the fixes for the concurrent store and dispatch functionality when using message groups is your issue or not +bq +uploaded pom + was moved to be an integration test as a part of +Patch for context lookupYou are right the CT doesnt seem to test for this but it is wrong +aggregatefeatures +Tested against Wicket trunk and with latest works +I suppose people using lower isolation levels might disagree but for me the lock is essentially an artifact of my code not being able to know itself whether the object is really in the database or not. +Thanks for help Stephane! Ill cleanup my code together with tests to resolve this issue +Im unassigning myself from this issue as dont plan to do more work on it +I cant reproduce so I am going to assume this is fixed with more recent versions of rjb used in buildr. +Sounds like a bug in Maven +Committed an updated file +Blanket replace of +Unscheduling to simplify release plan this is now linked to the umbrella task +target file so there is no target folder +it is a race between a flush and set +Im fine adding as a utility class especially if a we base a tool on it +Every time it format my code like thatI will change after QA result +The class has been improved +Resolved in svn revision +It could be tied into some of the work for fsync and fadvise behind the scenes so that we try to keep hot data in the disk cache and only run fadvise to flush out the cache if it is not really hot +Hi please get back to me on this issue can you still reproduce is with say cxf +I have no problems on Solaris although with somewhat fewer d look more at what is happening on the box +This issue has been filed against WAGON for the next release +batch transition to closed remaining resolved bugs +Just add the appropriate filter for your SCM +priority is absolute ordering and beforeafter is relative +added a comment to the above mentioned subversion commits tab its already fixed +I applied the shutdown patch in JIRA issue and this appears to make the problem much less frequent the debugee most often exits cleanly but the above symptom still can be seen sometimes +note the hudson test failure was unrelated to this so i think this is ready to commit +Attaching a backport patch +Fix in and test cases in +Also I suggest to clean the hbase datastore before testing +can you tell me the parameters you used to generate the code +Fixed +When I look at the patch file directly in windows I see symbols +Patch applied Rob many thanks! +Solution looks simple and quick to do +project +Submitting patch for review +Opened ticket for the aforementioned issue. +lol +Add form param support to via abfaefadbdcfaaca +This has been fixed and will be in the next code fixed in Hyperic drop +Checking for error code in the fetch offset response +There are some wording variations as well +As discussed We will do the changes as part of +Covered by resolved issues. +Yes +I have no specific requirement for a List +committed fix in close of old resolved issues. +In general the handler counting is there to trigger device listening of system events only when necessarySetting to Wont Fix for now but if we can figure out a way around these issues then I am all for reopening. +bulk move of all unresolved issues from to +Accents are handled when the message is shown in the layered popup thats why I did not want to waste too much time on it +No response for a test in years +New patch that fixes lint warnings +Thanks Uwe +What is the issue +Closing. +Now accepts ip or ipport does not accept null ip etc +Merged for the issue encountered with a single browser. +Bulk close for Solr +The issue is addressed +The proper fix is to ensure HADOOPCONFDIR as part of the class path hence we do not need to maintain any change for hadoop configuration change +There is no fix yet +Regards Issue +Thanks Colin. +It is same +created though renamed scm to commits and ppmc to private per! I am still waiting for the committer accounts to be set up to add a list of initial members +This patch addresses the above mentioned issue +HelloI originally logged this ticket and was very happy when you guys solved the issue so quickly +attaching proposed patchThis will follow the suggestion from Michael to have a commonThe it became and the get is handled in the implementationsIt introduces as well the concept os +Fixed width indents are used throughout and should be used here as wellAs to the actual test in that ACCOMPILEIFELSE I finally realized that its relying on an error being generated due to an invalid preprocessing instruction +Lewis +Here is our new informationmavensynchomescalatoolshudsonwwwrsyncsshJosh SuerethadminmavensynchomescalatoolshudsonwwwrsyncsshJosh SuerethadminmavensynchomescalatoolshudsonwwwrsyncsshJosh Suerethadmin Do I need to do anything else to be added back to the sync +No feedback as to whether the problem exists so closing +I reopen the issue for +Increased stack size for maven scala wonder why on OSX I dont need to do this. +Sounds newer patchbinlogreplay has required mode normal or forced as before and now the option to identify a maxtimestamp after which mutations from commit logs will not be second thought Im not sure we really have a use case for this anymore the main one was addressed already by +lets change to warning and keep as in +moved to quick sort. +So this can be closed. +So that looks pretty good + +Ill fix it when I get into the office this morning +The FTP configuration is already mentioned in the Programmers Guide and includes the sftpftps attributes +You are no testing correctly you need to set interface in the and the corresponding IP when using Advertize +I also dont have issue with this test in our internal branch as wellWhen this assert fail it means that observer didnt flush file to disk +It appears that the problem is the stashing of the on the class +This violates the naming conventions definition of a property and utils like dont work properly with it +It is a problem that transitive dependencies are added that are not needed +There is a relation with Nicolas you have a look at Saschas proposition? Jacques for latency I look asap this +HiI had already changed the documentation to say the default is never in revisions and Changing the default to actually be auto may suprise some usersDenis cited incomplete instructions on +Thanks +Now that work is underway with AS all previous community releases are +Comments on phabricator +bulk close of resolved issues. +The emulator on Windows is far too slow to run this on startup +Do we need a PB for URL +Verified with +This fix works +This feature has been checked in + +Some variable has been renamed + +Please post the EL which causes this to Patrick Madden for pointing out the solution to this one +Paddy can you try again? I have committed a new fixCommitted revision My tests was having a starting directory configured where as you URI option above does notAbout the absolute paths +Thanks +While hand migrating these changes to codeline I experiemented with synchronization in and checked in the changes into through with synchronization in it +Marking resolved +should work with that version +Even then IMO I dont think theres a lot of benefit +Filed +Resubmitting the patch because the system tried to apply the screenshot +it is a standalone instance however rather than an ensemble +Rates are can be calculated by the client by sampling the jmx metrics on the queue object +Committed to trunk and. +It is not passing unit tests yes +Teiid messages are shorter and more descriptive now marking done. +If it were published to maven which is currently is not in any reliable way I think mahout can have an optional dependency +Yeah I think thats a reasonable solution is there a chance this could make it in a patchThanks vm +Very coolIt seems that the parity is computed by xor +This shows a screen shot took by me for a hand written XML file after applying the very basic XSL stylesheet +sounds good will try things out later the day and respond again thanx a lot for if the code is not in but only the changelog entry that would explain why its not workingJust tried the nightly build +Lets see what happens +A slowdown would have to be some major breakage somewhere +Removed the invalid widgets until the as notion of configs can be in +Patch against +Maybe its because of the compilers extensibility frankly I dont really know +HiIll resume my work on this issueBut Ive a small doubt +Seems to be good to go +It seems that they are copied into distribution by test suite +We technically can include incompatible API changes in a release but Id prefer not to as they make it hard for folks to update to the latest version of Avro +Actually I would prefer to change so that the default dash array is null +I added an additional function on +Forgot to mention that your example wouldnt work as you dont have a constructor to accept or more references only for or +I think it is a duplicate of and now fixed. +I have been having fun with Wicket the past few weeks whilst going through Wicket In Action +That stacktrace doesnt really seem to tell the whole story +Read more carefully the comments on +Your solution sounds reasonable to me Ive already tested in the same direction +Better +This patch contains an alternative to test +Do you have time to do that? If not Id be willing to help out with that +Resolving committed patch as svn Marking with existing application impact because the patchintroduces two new reserved keywords CURRENTROLE and NONEThere will be one release note which covers all of the roleswork as part of +Yes this is an unfortunate leakage +Sounds good and thanks a lot for your help with this +I dont think its in the best interest of the project if the setup instructions involve changing a bunch of things away from unsafe defaultsFair pointbq +a unit test annotated with Ignore for now +It need not +Other than that Looks good +for example +Isnt this a XSLTC issue +Addendum integrated to trunkThanks for the reminder Xing +Will update online docs to make this clearer +Michele said I think this bug should be fixed ASAP +I think this is an issue in the version of the load balancer which is no longer in any active branches +Youll need to use it with the version of from the latest drop +Sorry this issue can be closed this is due to two different jars having under same location so its getting conflicted +The result will be that is an empty shell +Replaced this example with REST based serice to Twitter that exposes the tweets as relational table based on the query submitted +I got one failure in when I ran the tests a network server failed to start +Ive sent a first request to the hosting company and they are looking into it +Yikes! I think this JIRA is getting out of hand +Im not sure though if this will create a security problem +Ie +covered the improvements that have been discussed here +to have just one we consider splitting this ticket? One for vendor object stuff and one for modulesThis modules approach seems great as its tied to more useful features than just version reporting but it also greatly extends the meaning of this tickets titleGreat work Randall +Patch applied. +I was fooling around with this earlier without complete understanding made some progress but didnt have enough time to really test things out +mark fixed part a new one see attachment part fixed part like everything is s examination pas curriculumcourse nurserostering and travelingtournament examples in environmentMode TRACE do no longer found any memory corruptions. +to trunk but fails +We should remove the deprecated methods in filter too as part of applying this patch +Thank you for the patch! would be nice though +Closing resolved issues for already released versions +If you do let me knowCheers +Ooops +It is also a problem for +Theres a whole lot of test than can be +Ill change the patch to just put in the task tracker name without the link in that case +patch added +Needs to be properly implemented in the +Making that initial scan part of boot is something we added in. +Cant be a reference right? The file name would give that away in the logs statement no? Also I checked the config and the compactionThreshold is unchanged so set at +Im sure there are reasons to object to my code so object awayThis also adds a getWriter method so that the current choice is available to client code similar to getParser +I am using the reflection library to look at the generated java classes and building Castor schema model from it +this patch is my working changes for adding quotes around the paths created by the rpm plugin to have rpmbuild deal with spaces in r +defer remaining issues to +AlexHow about instead of forking you create a patch? I dont see the point of maintaining it elsewhere esp +Only one request should be rejected +Lowered priority since this is probably not the most popular platform for stdcxx +The part If the string is empty then the of the attribute has anull namespace URI +Committed with a simpler fix for the readonly listIf you can find a way to reproduce the remaining bug in a setup that will make it easier to debug +At least such visitors are only invoked on startup so their overhead is arguably not as relevant as +Ron can probably answer if the intent is to reuse the context param +Thanks a lot for the review Eli +Hey Nick could you take a look at this please +Thanks Jan worked great! +The signenc sample works perfectly fine in the latest CXF release +I verified your patch and it looks good to me +I see no duplicate Teiid plugins in the latest H Beta build +We need a hcat server package too +Ill commit this soon unless someone objects +So we introduced a breaking change +Closing resolved issues. +Thank you for your information Schindler! +Thanks Kim I agree +I think there many potential features here to implement +Against svn +Plan to integrate to trunk this afternoon +I think it would have value that we provide a way to have users get access to this content similar to what Eclipse EPP packages does +Sounds reasonable maybe we should close this with Wont fix and create a more general one +Im not sure which way we should go here +whats better +Removing the to the release notes + +heiko i have an oracle installation locally +This issue should has been resolved after looking into the code of cxf +See my comment from June +I manually touched one file in the srcconf directory with a timestamp and found that the archive did reflect the changed timestamp +Thejas Ok I get it now +This is yours Simone +My bad I fix thatSorry about that. +service creation will be valuable +Sergey do you see a stack trace or the error mentioned above +Wasnt this a problem on Common? EAP uses Common so does it applyAlso if we needed a new Common we have not received it by component freeze date there is no Upgrade Comon to +Removing fix version until we have a release slot for to Amos as a Maven plugin for detecting Byteman script errors is a high priority to the transactionsThis issue seems rather general +Hi PaulWhat is the prioirty of this issue? Does this need to be escalated to the bug queuePlease adding bundle cxf is now was not enough before maybe something was missing from the features or maybe confused the issue either way now the problem is fixed by adding cxf +The test has been running in a loop for a couple of hours without failingCommitted revision Im keeping the issue open until weve seen that the test runs more reliably in a few full test runs +Any thoughts about how we can fix thisI cant think of anything elegant but a dirty workaround should not be too difficultFolks are there any objections to committing the patch? Odi have you had a change to review the patch? Are you still planning on reviewing itOleg +Am I missing somethingIt seems to me if you hit an OOME due to creating too many threads you shouldnt create so many threads +Dont set the fix issue is still open so that I unset the fixed version +There is some strange behavior restarting a running server in debug mode but if there is a problem it is outside the scope of the plugin +to commit +I have made the parser change which Satheesh pointed out +Updated to use a thread local date formatter +I will retest tonight and post the error if it appear when the video is being reproducedThanks +Help in resolution is appreciated +Thanks! +Vladimir applied this patch + +author +Fixed +The template approach sounds overly complex +The attached patch is svn based and it should hopefully work +The default implementation will work much like the implementation in that it uses java beans style syntax to resolve the keylabel properties of objects in a list +No I ran the reproducer +See Fabio +on patch v + patch +Ive put that change into the code Im working on now +can you please attach a sample patch you are having an issue with and also which project is this for or are you having issues with all projects +Every Component has a master typeEven Link has its own use for a model the people example above demonstrates itIf you cant engineer a Component to use a single Model perhaps youre abusing Components competencesFinally the good compromise youre looking for is already in Java +I dont agree with whats in already +Didnt get a chance last night Kyle +We need to work on this a bit more to get this right so that it doesnt conflict the original concepts of synapseI will try to modify the implementation when I get some free time or can you propose any changes to make sure there are no conflicts in the semanticsThanksRuwan +I just committed this to branch +Good thanks for letting me know +Whats cardinal number of CAS in each Sink? Here is a failed point +We have a lot of tests that cover all the injections you mention +Better call it maybe +This one is from the CP sectionThis section contains details of changes that have occurred in the release that will require additional steps to be performed when upgrading from the or release +Hibenate must allow to limit indexes size or hibernate must limit indexes to size for tables itself +Thank you for the information Michal! Ill look into the issue +Thanksm closing this issue because it cant be reproduced +works fine +AFAICT I fixed this in +I am sorry for my English +Since BSF uses one single we may use a configuration file that allows you to turn off explicitly a that is normally delivered by BSFoffoffThis would allow the upgrade to newer artifacts that deliver such an engine on their own now while it is still possible to use the remaining engines provided by BSF +Patch was rolled back for now. +This looks good to me +Committed branch and trunk +If reproducible this is serious + +It just seems simpler to me to avoid this whole issue by leaving our jar file names the way they are +QA verified +It just means we need it to test and compile but it will be provided by the user at runtime +Thanks for the tip! +didnt get asked for where the patch file was when marking the bug as patch available +jettyNot a fan I think non Java guys should be able to open the package and quickly see the folder they need to move to +Elliottethis should be fixed in +When we can consistency run the same system across OS X Solaris Linux and Windows well leave this open +found the problem +The issue was basically the in the path name was mentioned earlier by someone +that is odd +Explicitly setting to value makes sense so losing data doesnt come as surprise for applications +The operations that use can then just use that +Bug is fixed nowwhen i tested and reopend then in step of your mentioned steps Normal user is able to see network offering for which vlan is set to true +Without reproducible environment we cant fix it +I will see what I can do +I have added changes to hide zoom for non moderator usersAt revision you Maxim +HiWorking in the IT department of a bank we are studying the means to make our applications accessible the subject of accessibility is becoming a major issue in many waysOur teams are using mainly technology for our applications and the recommendations from WCAG doesnt seem to work properly on them +But it has to be multiplied by the number of join operators in the query and the number of concurrent users to see the potential peak memory usage and its probably better to execute queries successfully than have them failing with OOME even if that means the performance could be suboptimal unless you do some tuning +Thanks for catching the Analyzer final assertion test issue +My conclusion is that uses default encoding of the environment so in my case encoding is +Can you provide more details +thanks please reopen if you see the problem again +Note that Peters proposal makes it work for any type and any definition of nullThe only drawback is thats its more subtle cryptic than an explicit param or annotation +Thanks Emmanuel! +The default value is empty +Once stack name was updated repos worked +Committed to +BTM is still fully functional minus the monitoring of course +test failures are unrelated +Did you see that only one thread is running when you run this new hbck +Reassign to a release that you feel comfortable resolving this issue in or leave as is +After authentication hes redirected to s no reason that in the second scenario the user shouldnt be redirected to the page hes trying to access +Struts Tools Tutorial was checked and updated +To me it looks totally fine now. +Really appreciate your help! the NEW FIXED +Oleg +I recommend upgrading to +This is implemented in the patch. +Revision +reopened just to set the resolution to fixed +See last comments. +Im testing it with the standard jboss portal bundle getting many errorsbut the culmination is this +The code checks whether the time is and if it is waits until it is +Reopening for x backport +Decreasing priority since we know of a workaround and there is not much we can do to actually fix this +Please reschedule if necessary +Bigtop also has this version of Flume +Thankspatch updated +If you want to provide the text that you think should be included in the documentation that would have helped you out wed love to add it +Certainly without adding a complete testcase +This has been fixed in so Perl and above shouldnt need this workaround +the issue result of which still in progress when will be fixed this issue should be not reproducible +Mike is there any reason why you didnt use and insteadWoops no I forgot we had version info in! Ill switch it over +But both end up adding nodes to the plan +The FTP consumer should also skip +This patch contains the aforementioned optimization +The patch looks good to me +re fixing this can be enhanced with a more specific version of ORACLE and add a corresponding SQL statement in the I tested now on a RAC +This was a decision so it should also go into one of the current release +Note that the current approach to estimate the completion time of jobs is quite simplistic it is based on the time it takes to finish each task so it works well with regular jobsPolo Can you please expand on your definition of regular jobs? Are these for +Attache a patch file and a referenceType is assigned with each jndi reference so that we could hack on some codes in the injection process +Still a good idea to look at trunk. +His fix should work for the new stuff +Im talking about the p repository created using the mirror tp mojo +src this package packages transient directories such as coretarget and which should be is missing! Even though module jars are in +Since the contains Felix framework implementation I would assume you would launch a different framework by putting its JAR file on the class path first +The attached patch is for trunk but should also apply toAlthough the environment may be set by users or by the admin the framework including what may be needed by the hadoop code base in seems more reasonable +fixed in both branch and trunk +I feel that reversing the assignment of timestamps is the safest approach +So I agree lets not do that nowThe patch looks good as is +Merged to branch as a side effect +But its a part of backup relies on and but it can be not a case for current implementation of JCR logic need to be checked to avoid usage of and +The simplest solution for this is to modify the constructor and remove the check for GSON on the classpath +And so most probably youre facing another one. +We have provide a workaround in the CP which is installing the opensso as ear in the +this appears to be fixed +Attached patch to fix it +counts you me Andrew and Dirk at least in the top contributors +If a developer is building a development version of their module theyre running it through a compiler +Druba its your baby please review the issueAlso patch fixes v issues removes commented out code replaces with proper logging +App use plugin +IT added +if we break every existing hibernate users existing it is not something we can put in +The patch from has been merged in Web CP branch +Hi WillemI just noticed that you did not grant the license to ASF for inclusion +Our goal is to have a leader that has enough supporters so that we can make progress +This issue has been fixed and released as part of release +Its an intermediate issue reducing priority +Is there any progress? Attachment is not enough +Juergen we both arent happy with this issue not being in M but the issue exists for all other servers too and is an unfortunate caching bug which for M need an eclipse restart once you created the server adapter to take effectI would rather have a working M with this restart kludge than risk accepting a patch the last minute that can potentially do worse +Unfortunately I cant see the change text stays as JBOSSDISTserverconfdeploy in chapter +The is much betterAdding a transient field that is lost when going over RPC is problematic from a maintenance point of view +Can you please commit thisThis way we can check if you have been set up with your Apache account etc +Note that I had a fresh workspace and a fresh unzip of EAP when I tried it today +LGTM +That should be handled by the part of the script starting with patch up so the driver will loadThe update is very simple unpack from then cat the needed updates to the end trimming the line endings as necessary +I just checked this with the latest CVS code could not reproduce this assert using the version of and we ship with any of the samples +This patch has been applied in the revision I also added a postClose method for closing the other stream after the cached stream closed to fix the rm systest +Thanks for the clarification. +Committed trunkOutstanding then is addressing these leaks +Please feel free to reopen if you can provide a reliable mirror +Changed the patch to do cleanup after calling stop on all services in +This patch looks good to me +AgreeI havent been able to get to work on these beans either so Ill have to start looking deeper +The only security risk would be a denial of service attack by one client being able to render other clients connections unusable +But it doesnt contain any details on how to actually add the repo to your maven settingsWith the article could also describe how to use this new functionality to add the repo rather than just manually editing your +I am getting the attached error when I try to configure the xmijndi I got below given error is it related to some memory settings or something else? Please Help +it is also in in trunk r as part of a set of commits for +Flight in the AM so packing! +mvn dependencyresolve Dclassifersources does the same as mvn dependencysources +A shard is the logical unit +In both the cases the table schema is copied into the partition storage descriptor +The work on the guide is started +Thanks jacques +Thus its also logical to explicitly state the reception refers to bean impl +Is there a WARN or ERROR log line if the timeout is exceeded? If not lets add it +A first overview looked good I think that can make itI will try to import it in the next couple of daysThanks +Dimitris do you know how to disable this in the AS build +It may be helpful for RAID cases +I will provide a patch for this issueThanks and Regards patch resolves the issueIn the current implementation output location is just appended to the workspace root +My alternate workarounds also hit with the massively nested objects but an alternate unit test that just makes a larger object works beautifully + for to trunk +It does this so rake tasks rails console and similar work +I can confirm this is a problem after two hours of trying to figure the problem out I found this bug within the Struts JIRA +Also committed some udpate at revision in +I was thinking of commiting and then making some further modifications +And nobody it seems is clamoring for this feature +That waswhen I opened +all to merge all the regions by and allrecurse to merge all the regions by and recuse until there is only one region remainingPatch for ready too +You mean that all of the clients must share the same? I think that there is some misunderstanding here +Moving to will do it in Apache all resolved tickets from or older +Thus the example can only be used by itself +May be there are other procedures too for which we do not do this check +You need to write the full path for one of the Enum classes +On the other hand returning a Writable may have performance implications +Since this size doesnt halt puts and can end up with a slightly larger I add pad +exposed by this test +UEL functions are very easy to createYoure right the approach does make for prettier syntax +I am planning on working on this issue +thanks +Patches forrestbot to use plain svn +Ensure Hudson agrees +Bulk close resolved issues +These files are missing Apache license headers +This might be a red herring but I recently found and fixed a thread safety issue when validating connections in DBCP +idont think that you know about the first visit when looking on the absence of the value because the value is set in the context filter i agree on youre second point maybe there is a better solution But i think at the moment its a better solutionCheers SaschaSorry maybe I didnt do a good job at explaining my feedback instead of trying again Ill just commit what I was talking about and you can see what I mean +Deal +Producer and consumer I used +Applied along with the associated SCA changes +Due to the upgrade complications the configuration approach is a practical one +I have committed this +After the first transaction commits even though the second transaction is not supposed to see changes on the first transaction under the level the message route is visible on the second transaction due to the update lock on the route table +I added a little bit more info the the resolved issues. +One last question +Thank you Adamis good to know that sometimes older Jira issues are addressed as well! +WanbinThe patch is a diff fileCan you produce patch files for andGood job +runtime package will containutils package will containThese classes are currently in the top level package +Ive committed a new maven module that better explains what i have in mind +Awesome! +And some small issuesAt action information parameter Default location extension always +hang on is this just the removal of +Verify this section in Admin Guide and close issue +Deepa does this issue need to be addressed prior to committing any new trunk changes which include update functionality? For instance there is a patch to ready for commit +Code looks good +Ill rebase against for the commit +Thanks Raghu +Reschedule for the next same problem occurs on AS +fixed +There are no unit test cases +It seems this works only for JEE no? Whatabout JEE? Have you tried running an ant release from the top level to see if it works for all cases +the code does not work +When can we get our hands dirty? +see it for a test case and for two mojo implementations that use this feature +In the second one Im not able to select any item +I was caught up in the moment and immediately jumped and created this issue without much further investigation +Patch added +Attaching a patch which adds a Javascript enabled link next to toggle explain which expands to show all fields +fixed including the wiki page. +So in some of the tests it passes if the message is transported within ms +Attaching correctly formatted patch +Thanks +What do you think? +Seems useless. +Added a comment explaining what might be required to have this addedagain +In particular I believethe call to close the output stream at the end of the response is the culprit +Main point here is the casing of public fields not parameters +I agree an error message should be output for the naming collision +Committed to trunk and branch. +When indexing files empty set is also required to set in. +JNDI bug on theRegards +If a resource is used by more than one stylesheet why not just promote it to a common webappresource anyway? Because a plugin developer has no way of knowing that a stylesheet is going to be at some point in the future +Yes please change the +Weve been using it since I reported the bug but the new version isnt slatedto go live until some time in Aug +Ive posted the question on the forum when you asked me month ago +PatrickI applied the patch but I am not able to see any new contentCould you please give some information on how to test it +Following Brunos comments in Im using the following guidelines while making the install script consistent across different packages Use spec and rules files install sections to determine what variables are passed to the install scripts Keep PREFIX out of the variable definitions in the install files to reduce duplicate work in spec and rules files Use common variable names +The upstream issue seem to be fixed +Are you sure that your environment is clean? Cant it be the case that your class files have changed between serialization and deserialization +My only comment would be to rename the new class to something like since thats what its actually testing and we already have +what is it good for +thanks Mike +Our partners at another company have the same issue +I too see this weird icon +Denis Can you give Mickael a hand with this? Ive no idea how you hooked up Emma coverage in the past +Closing issues in RC. +Ben Alex is back as lead of Roo +And while I am at it I should document how to set up and run JET +Hi SanthoshYes there is inconsistency issues with and +Because most of images outside css +Opened re right Todd it looks like with the new job retiring process the JT lock is always grabbed before the fair scheduler lock and neither of the two deadlocks I pointed out could occur +Hi FrancisIts in our internal repository atrepositoryorgapachedirectorystudioTry to run mvn clean install first +taking a look. +To be clear the error report is that the file becomes big not recovery log files located in the log directory +Thanks again Rick for your help +In the case that its a null CL it fails because isnt on the bootstrap classpathA workaround which fixes it is to temporarily change the threads CCL to match the UGI classs then flip it back +Probably I can use this information to set the state of the form components +One of the issues i see so far in configuration is that to use a remote server DB its not picking up + on. +The error comes from +Thanks Bernard +Yes that makes sense +updated patch for +I dont think there was any good reason for it +Please add a few tests cases to make sure the overwrite flag is honored +Pushing to as discussed above +Brian No not for CR +Closing all resolved tickets from or older +Is there anything keeping this patch from being committed +and it is a guess based on my knowledge of that code +As far as I remember on design stage we discussed the inputs inside the menu and the result was that it isnt normall for the component +I had the same idea and I am coding it currently +This is passed to jetty as it is started up +Reopen if not +Applied the patch but the integration still need work +The ant target in head is named not so I beleive whatever script is created to tag the release in ALPHA didnt do the rightWheres the script that generated the file in ALPHA +Yes users cleanbuild solves the problem +Not possible. +Commit at subvesion revision +The session remains untouched only the action is affected as it gets a new map that it thinks is backed by the session +We have the VFS working +Does that sound like a good plan? Whos willing to help test the SNAPSHOT +It appears I restored this output in ea but I dont recall the reason +I think its important in that case to clearly identify the modules that expose the public API Pauls work to reorganize the source tree could help in that regard +Adapted version tag to be able to remove versions +I see +You are right that the class was taken from jUDDI and has not been updated +ResolvedNow devices have pixel ratio property +Defect is in patch which is committed in current and update latestI am using ubuntu and bash shell +The granularity of is sec +Committed patch to the trunk revision +Then we would want to warn users to always shutdown their databases gracefully in order to avoid leaking the longer preallocation range + +I cloned the branch but it seems that now there are some compilation errors because of recente changes on upstreamAre you planning to work on this one any time soonI think I can help on this since its needed for the project Im working right nowThanks for the great workDavide +Closing all resolved tickets from or older +Optimistically resolving this since Nahi believes its fixed and theres been no update in a month +Patch applied at +The patch included a foreach type in the substituteType function however substituteType is already called for each type so thats double double +Hi Lionel did you try turning up the logging to debug? Looking at the logging code the next line should have the exception message if you turn up the logging to DEBUG then the whole exception stack trace should be printed +Updated patch to apply cleanly against recent trunk chnages +Maven project to demonstrate the Plugin output logs depending on the for reporting and especially for the test case ChetanIve fixed this in revision The code was a little bit strange and I guess we had a wrong idea of the maven api at the time of writing. +Wondering if we should have a separate binary tarball for it +I forgot to put small changes in the specialpurposeLICENSE file this new patch contains All put as resolvedincomplete to facilitate searches with My resolved pending own news thanks for the effort Jacopo +Yongqiang can you please add the name of the new property to this ticket so that people can find it via searchAlso do we lose anything in terms of functionality or performance by ignoring the HDFS path informationIf the answer is no then we should probably just ignore this information by default instead of adding a newconfiguration property +Jody I can confirm that all tests are turned off in the pom +Making a daemon thread as this was the thread that was hanging the JVM +In all use cases we defined our searches and therefore indexes need an order by clause as well as a query criteria to support the paging that most applications will require +Thanks m inclined to settle this issue on showing times in the server time zone +At time T HUB released its ownership +I think that most suggestions can be fixed right nowBut I have some questions +Ill send out another update later today +Just for the sake of amusement results from ApacheThreads Total speedup in trunk after patches KBsec KBsec KBsec Should make folks happy +But the population is +Verified +The stacktrace will please in the future can you please give a more detailed name to your issues as they are used to produce changelogs of version and isnt that any case thanks for your interest and help +Allows upgrade to new stack via ambari server +Ill commit this later today if there are no objections +After being hit by the same problem I fixed this in revision The config admin update event is now handled async and does not block config admin anymore +ll look into this issue +Patch implements my suggestion +Thanks a lot to Charles Moulliard for donating this new great component to CamelIts committed to trunkCommitted revision to tickets +Im also looking for support +See it anywayI cant read +Thanks Jonathan. +patch applied to branches and trunkMichal thanks for the patch +And this caused a Invalid Event called ACQUIRED at KILLEDIn the heartbeat response to AM this container is present in newlyAllocated container +added some more assertions to unit test +I think grapes is also where custom compiler extensions more naturally belong too but the current compiler extensions mechanism was put in place as an interim solution for groovypp +Works create +Please note that we also applied Jaka Jaksics patch for to theMartijn +Only operation is to remove this dispatcher from theAnyway from Axis you need to have jdk or higher. +Going once +I mean +I just committed this +Robert patch applied at revision r thanks a lot please verify +A trigger cant be dependent on another trigger +This rather is a CMS featureDo you agree +Will update patch soon +zhoubq +Patch adding failonerrorfalse to the delete in the example ant task +The failed tests require slfj +Need to merge to branch when we have this branch setup +Theyre pretty small and ought to be easy to understand + +Thanks thanks for addressing Todds review comments and added unit testsTodd could you please review this one more time? Thanks +If distcp supports then this shouldnt be a problem +I agree Id be in favour of changing the Schema mode to follow the specIf no one objects then Ill make those changes +Ive updated svn to reflect this and updated with your changes and my additional changesAll good now +Ive just committed this to trunkThe patch doesnt apply cleanly to because that branch doesnt yet include the work +Just got it again with postgresql during EAP testing +Ok my mistake +Need to do the same on and just fixed this in and trunk can we close this one +Verified on windowsia. +Is there a plan to include this patch in the next versions? will be implemented in the UI plugin for the Spring Security Core plugin that will be released soon. +This implies porting changes done on branch GACXF to a new branch obtained from tag GA +The data loss was in in both andAnd there is this clue Thats another configuration item to try but I have to switch back to or +opened to track exposing the table building via v we already have a API for the client +But in the refectoring I use the Exchange so at this time it became important +It now writes the process ID to instead of tasks so other kernel threads started by the same process stay in the cgroupSo it seems like there is some reasoning behind it and having other threads started by the same process stay in the cgroup is important +Sadly I was away on vacation when this flurry of activity happened or I mighthave been able to avoid a lot of unnecessary hassle +It seems like a good idea to move it to core +However I dont yet have a set of points to test +Is different from the hadoopqa bot for patch testing +If We maintain the state of stages on quitting We can definitely quit and the wizard polling on same requestId +Imported the attached project and generated the WAR without issue +Cancelling patch to address Raghus comments +The used for activation is the of the Bundle registering the plugin servlet service +Cool thanks +but there should be a conf to disable it as there are some exception situationsThis question should not be relevant now +I updated the overlay with the version retaining our customizations for and +You have the eyes of eagle +Sounds good Ill prioritize getting my tasks done for the release +While we could certainly include morefunctionality into the it makes more sense to include such functionalityinto the AJC tool d recommend pitch this idea to the project. +Also for float values I wouldnt expect too much of an improvement. +When such a case happens repeatedly then I think the test cases should fail and we should reconsider the leeway given to each task to start +This likely has to do with the runtime leak I was seeing in but was rather confusing when I observed it +You would get mine but I am not Abdera developer +This issue unfortunately got to Beta release +the following files are modified question do you know of the blueprint stuff has similar issues +to change the release the fix is in. +Committed to branch +Instead I defined a new state named SAVING between NEW and SUBMITTED of +We need some kind of notification so people know what to use for script and are no longer present in +There is some life on Surefire for there wont be anything before Q I guess +Patch is ready for review +And selectDesc is a member variable of right nowI think should have those descs plus the children information +The parent pom has already been updated to the new mailing list addresses +using for indexing. +on. +Id be cautious with committing to a particular for of +Works fine after solving +Fixed most tests in rev +protected field and its setter is protected there should be no injectionA Only in the case of protected field without a setter method the property should not be injectedQ +Since you have already applied the patch Ill provide a new patch based on r +updated get to use and snitch and updated BLM to fix units on RINGDELAYrest lgtm committed +on Hudson +Thanks for the review and pointer Gary! Ive updated the patch for windows +If the discussion leads to a decision to change and require escaping then we can make the required changes +There appears to be no way to reproduce this +This issue has been fixed and released as part of release +Filed for adding +I can try this out a few hours tried BranchEAP and the behavior described above not confirmed +ide please verify +The above comment came from myself a new member of dev team in Inc +Straightforward patch + +just went into RC +still open +Some explanation in the file for how the are to be used +meclipse doesnt use maven but QE do for WTP in QE and meclipse is very very far from what you have with the eclipseeclipse for this info +If we were to address that problem too I think we could do it by always sleeping for a random amount of time on authentication failures +Hi Andrew thanks for the prompt fixI verified it fixes the regression on OS X +Thanks Brian +However according to RFC the value is supposed to be numericstring which should not exclude values like etcSo I guess such values should be supported in future releases +I can replicate this deadlock in a simple test which registers the same exceptionlistener on two different connections obtained from two different connection factorys +keep all the raw methods in the interface +cn so close +Therefore marking this issue as resolved. +Youre right about the memory usageI see that as a limitation of the implementation more than the queryResults though +Should we change the name to something like fileID +Hi LorenzoAs discussed on the user group the HTML select element does not support readonly attributeThe workaround is to render a readonly Select control with the disabled attribute and render a hidden field to submit its valueregards Malcolm checked into SVN +This patch will remove PHP closing tags +If on a return value of a overriding another I do change the group conversion I can end up stop validating things that I would if the superclass was to be validated +Much appreciate +im seeing both XX and in for the mac version of on mac at least it should be fine to add +This script corrects the problem +On the server where we dont have jruby in PATH it does not run its closer but Im wrestling with CI at the moment +Attaching patch that adds the Job Counters NUMSUCCEEDEDMAPS and NUMSUCCEEDEDREDUCESAlso fixes a bug in updating NUMFAILEDMAPS and NUMFAILEDREDUCES +Strange that it worked in Java I wonder if there shouldcould be a test for this kind of problem +Environment is +Hopefully with the new icon the problems Tim for the graphics +One line fix +NET plugins +Merged into the branch. +A more realistic sequence of events is this sends to with a of +Stephen I didnt behave correctly before in +So that would break backwards compatibility +Ryan is the contact person for installer issues +Documenting a property that follows a naming convention probably doesnt add a lot more entropy than adding a new property with the same function would +It would be great if the build works on cygwin but this has proven tough to maintain for Hadoop since most developers build only on Linux andor +bq +BN has some bugs as well introduced while the code evolved +The changes for this jira needs to be backward compatible +Good idea and Ive created to do that +Please be gently first time I contribute and translate +I think we could add +Sorry I forgottn to execute the mvn clean install packageBuild Failure +Remains known under EAP by Strong adding to my list of changes +Lines should not be too long +If this is within easy reach I dont find arguments that YARN is philosophically opposed to it or that the additional freedom would allow to shoot themselves in the foot satisfyingI realize that we are rehashing many of the same arguments so Im not sure how to make progress on this +The nameTypes argument is mandatory now +closing issues for released versions +Modifies frameworkmodulessrcmainjavaorgapachegeronimosystemmain frameworkmodulessrcmainjavaorgapachegeronimodeploymentcli frameworkmodulessrcmainjavaorgapachegeronimodeploymentcliModifies frameworkmodulessrcmainjavaorgapachegeronimoshellgeronimo frameworkmodulessrcmainjavaorgapachegeronimoshellgeronimo frameworkmodulessrcmainjavaorgapachegeronimoshellgeronimoPlease review this patchI thought it best to keep this one simple and require the user to specify the full path for the if they wish to define it +Looks like this configuration was part of the security documentation but was removed by to reflect the combination of and into in trunkI think this is really an issue with having the generated ref guide on being built from trunk when everyone using it is likely to be running or earlier +Then I submitted this new patch +seems this si also a result of the class sorting bug in the compiler +Also marked as Optional +Patch applied on branch +unzip the patch to NIO module +Given that Im inclined to close this bug +A nested spider trap test page has also been added to the fetcher tests +and ported over from Java contrib +Im attaching a patch to solve this problem +Ah yeah I am using and it got some tis a fabric issue and not IDE +Well that makes me sad +Problem in hive was it was missing classifier bit in +I wasnt able to see the but I saw another one on getI added some synchronization code to the methodsFixed in rev +So I think we can close this jira now +An obvious step would be to invest in a well linked Wiki section that provides a directory of such projects +Ive merged the fix to and changed attribution in so that its listed under release +We may need to generate edits for naturally expired tokens just like we do for explicitly cancelled tokens +Patch is updated which also includes patch for +Looking +Added to the archive parsers chain +This is the Undertow portion of the change see for the sideAlso quite likely there will be a corresponding change to allow JAAS to work with a. +is it still applicableits been sitting in bugzilla for a while +Fixed in revision a wrong variable was used for the cookie handling +change block merge task too +No it is new and no it is not keyed for id at the moment but Im struggling to think of a use case where you want to do that +Thus closing the issue. +Shouldnt it? The pom I looked at did not have the profile added +Patch to remove invalid on and there any maintainer who can commit this trivial patch? Or the project is already dead +Looking goodMinor pointsI think the setuid exe should work if the file is readable only by root +its basically just a backport of to fix the problem +Thanks Dick! +Attaching +I left it in to prevent an infinite loop if some other bug later introduces the same kind of problem but Im not sure which is the best way to go from a software engineering point of view prevent the lockup of a critical system component and hope that users will notice any infinities on the calculations page or allow the lockup to make sure that any calculation problems are noticed +Fixed +Do them as pending perhaps +The last comment was appended somehow everytime I refreshed the page +Committed the patch to trunk andThanks Jing for the review and Allen for verifying it works +it doesnt have any option to submit patch or close it +Bulk move issues to and +I committed the patch toThank you Chris! +I will update the test app in Jopr to reproduce this +gitignore +Committed revision +I thought you were going to commitGiven that the imeem guys have already sent in their paperwork and the triviality of the changes to the generators I think we can leave those out +Thanks Enis +How do I run the mass service +I have not tested this +Verified on EAP ER + introduced +This explains Frederics in current SVN head of AxisJava. +Thanks SibelBir ey deil gzel gnlerde kullann +If you can add such a mechanism Ill create french buttons for you +What do you mean with itI do agree that we should add the activiti extensions to the Activiti Modeler for theFalko it would be great if you could start that +Attached to this patch there is a modified solution using the proposal from Ingo Hofmann +Thanks Sharan for reporting issue is fixed in trunk as well as in release branch on rev no +I just saw your I can take a look at that +adds username and password options to +I dont mind adding a component in specialized apps folder +Need to investigate +Thanks Rick! +The closest wed be able to get would be if Date and Timestamp were implicitly convertable to one another and they are not +For the sake of discussion Are perhaps http connections not being used? A connection should effectively achieve ugi caching too +We also inherited the message tag from the original validator package so theres backward compatibility issues there too +Changing to use get instead of getHost is almost sure to break other peoples applications + to mention you can now use int and double etc +This is a bit messy +Please ping an email over to builds sent email and have gotten suggestions but I still dont have access +makes sense to me +Done at rev Thanks Erwan. +Are you sure that this worked in in the context of a Portlet controller? support has only ever been implemented for Servlets +lgtm +This issue should be closed +Ive created a ticket to figure out whats going on here and fixing itSecond the RSS test did not even finish one time in the allotted interval even though it should be a good deal faster than the file system test +Verify whether HTML docs hold any information about how to add custom cache classes +looks good to me +How can be finalThis was just a comment about the class but its not possible because it is extended by or similar classes but the implementation of methods should not be alterable +Pushing to has been pushed +ThoughtsOn the renaming of the finish I do agree we should change that +Btw avoid any commit and rollback calls otherwise you haven no chance to use a global transaction in an ejb container +crashes on interpreter +Could you give me an example of these linksAnd couldnt we just link to the stable download or documentation main page instead of always having to update docs? +see comment of Carlos about composing test to check if groups attribute value is correctly updated when inheriting groups +This issue has been fixed and released as part of release +Now that work is underway with AS all previous community releases are +Any scripts that rely on the location of should be updated to reflect this change +Its really a just a limitation of the Maven plugin which pulls down all the modules into the classpath +Moved to common as the patch is for common though the test is in hdfs +bq +added validation by means of getting a reader for the provided url and return to the in case of success +Close after we do rewrite +Patch applied in revision +Hey Xavier the problem was that the plural is being capitalized in but not in +Part of a bulk update all resolutions changed to done please review history to original resolution type +The queue component in should be removed as well +My own testing indicates that tweaking down the dechunkedbuffer as noted in my previous comment has measurable effects +Obviously there is some benefit but the user still ends up cding all over and then running a mvn command and eventually going back to the module that was being targeted +Reviewing then commiting +This makes it very difficult to support past branches or maintain distributions since it appears this was fixed long ago but in fact lacks a major part of the JIRA +I would be happy to create and commit theZhijie is planning to work on this +Reverting the commits from this ticket resolves the NPE +Let me take a look +Since it looks like a release of Maven is on the horizon it would be REALLLY nice to get this sorted or at least a clear directionplan of a proposed resolution +I think wed go crazy trying to divide the metadata up into categories and that theres not much harm in stuffing it all in one bag +A changed to THRIFTINVALIDSOCKET since you generated this patchMajor bzero isnt available on windows +on getting tests working again +I have committed this +Thanks for the patchI improved the search producer a bit to fallback and use the keywords configured on the endpoint if no keyword header was given +Added which addresses the issue Mahadev mentioned good catch +this should not be needed +The test passes both with and without the fix +I noticed in the one you have the path as d but in the working one you have d +Can you point which EE spec document says something about being a location for classes? If its not a standard one then this issue shouldnt be fixed +I think we should simply remove them from the FUSE MR distribution as well +This patch depends on the interfaces introduced in +I do not see any reason to make the buffersize configurable from outside +The last digit should be bumped on the branches when this is backported so that upgrade gets triggered +Finish edit +I need a more complete spec than make it good +Why does simply using a Geronimo class force the installation of its log factory? +Pete +that way if the recipe needs to change because of a bug for example it is clear exactly what implementations need to be updated. +So we need two thing to do +Any other suggestions +I do now understand that the Introspection stuff is more than a new handler mapping +The repository is now available +Hi Mamta This patch looks fine to me for the targeted case you are fixing +I appreciate the fast response from you guys +I forgot to mention that I ran the tests that failed manually and they passed for me + applies to the trunk and the branch uploaded patch also applies to the branch reviewed the patch + +This will be available in tonights snapshot please give it an early try and let us know whether it works for you +Because WALA cannot handle inner class well either +How does this handle two file system implementations with same scheme? Currently it is handled by picking the one in the configuration +It doesnt happen on every split just this particular one +Patch was applied to both branches. +So this should be a +Hiis it possible to attach a sample PDFKind regards added the processing of in revision +Gleb please review for Final and let us know about risk and timing +Make sure the runtime is using the exact same drools jars as guvnor when it built the rules +The paths in will have to be modified slightly but should be an easy reproducer +At the same time this string allows values like a so xa will you require to have a value for x +When someone tries to select with the distinct keyword will can get back dublicated entities +Assigning to Juergen since Im unable to look at this for the next few weeks +Is there a way I can kick off these tests on the server next time so you dont have to +eg there must be this apache feather logo on the attached file +IIRC were operating in a mode in +Rb does not let me submit this yet since it cannot see the last commit I guess +IMHO its overkill for a single line of code being changed but I suspects others may disagree +Thanks +Patch applied to trunk r. +But it assumes that a term is a string yes? This is an incorrect assumption for Chinese +Interesting feature +Then I will be glad to test the latest versions +Alex Yes you can this is pretty common practice +Applies to all branches and patch makes changes to the test and adds the startedOn field to the too +Closing tickets +note that def x creates no binding variable so the is what I expect +Fixed resize to autosize to first page of dialog or to fixed size if specifiedStarted fixing data picker to use popup if configured also to remove frame redirect there also +I assumed it was a requirement +this looks good. +I believe this is fixed via another bug already. +Those jars should match now +Dont put them thereHow about adding a switch that allows a developer to create packages with the libraries? Developers can get kits with the libraries whenever they want existing packaging scripts will create installation directories without the protected run time libraries +Sure it is not pretty and it could not work out +I dont see any artifact dependencies in any poms that reference the +Agreed +Ill try to find some time to create a patch +We dont want to loose an important current performance optimization to add block compression if we can avoid it +This should have been fixed as part of the cleanups in transition all resolved issue that did not see any further comment in the last month to closed status +It passes with both interpreter and no optimizing jit +The patch looks good to me +Feel free to comment more or close it if you think this is sufficient +Ill commit this shortly and then we can deal with sorting in distributed search through a new issue +nope dont think so +What was wrong with the old approach that used a barrier to force ordering +Resolving for QE verification +To prevent mixing up features of and EJBQL in the future we should work on them in seperate issues. +We often ran out of memory while trying to spill +This patch moves old classes to a new package and its impossible to see what changed +So whens out +Patch applied! patch fixes the saving issue exist error issueapply patch to trunkmodules +This bug does not exist in trunk +Looks good now! Maybe well update the description if the Doc team has a better suggestion but for now lets consider this is done +This failed the spec because it was no longer iterating at the time it called rehash +I dont recall the icefaces app ever working all that well in Seam +Bryan there is already a JIRA entry for NULLS FIRSTLAST who is in control of this ISSUE! Can someone please push this issue to be resolved PLEASE! Great pain to build software solutions for structural weaknesses which leads to bad software design! This issue is now more than years old and fundamentally one of the most important! Just tell me who I should sleep with I would appreciate some ACTION! all of this issue have been marked as resolved and since introduced OFFSETFETCH which provides the functionality requested here I suggest that we mark this issue as resolved +This is perhaps what Vivek intended +Should be ok +Yeah I saw the test failing but I did not get into investigating yet +This seems to work ok for me for an existing oracle table of mine with a single record +Applied to master +attached the command and a printscreen of the shell clientThis was reproduces on and do you install the command? Do you use the blueprint namespace? Can you provide a full project +all these stats should be able to collected automatically at insert time +fails +SVN fixedPlease validate +This is setup in JIRA +I misinterpreted text editor cursor as a mouse cursor +Ill investigate where the version is coming from and will eliminate that. +After that if I try to browse the INBOX folder using IMAP I see that messages are not deleted but when I open them I receive an error server returned UID which does not meet the IMAP standardsThen there are some values +Seems ike Scriptaculous autocompleter is not all that great +can you provide an unit test showing a different result when this option is used and when it is not used +Great catch Liyin! And patch looks good +David So there exceptions are thrown when is running? Am not sure why its exiting so many times +Ooops +I think that the DB documentation is most clear about the danger associated with this privilege +This patch adds some missing getter methods in filter variant +Part of a bulk update all resolutions changed to done please review history to original resolution type +We have examples where increment is not always +verified all the test cases +patch that addresses this issue +Excellent will try tonights build tomorrow +Its for developers not users +Committed to trunk +Thank you very much for taking a look +It gets pulled in through +Gerd any update on the issue? Is the existing bundle okay is Eclipse to blame or bothThank +Its not currently in CVS unless the Attic files are current +For example to test if all abstract methods overwritten +In my case I could work around it by disabling the Combiner +Looks good Ben +Maybe that wasnt happening in the daysWith state saving the JSF view is held in the +Thanks! +checked in + +Quick reminder If you want to retain the ability to run under Java you may have to access the context class loader via reflection +Thoughts anyone +If anyone will decide they want to use the code Ive explained that the class has been copied from Trinidad +However this is not universally true across all the object models JAXEN supports +For my scenario under test the input CAS to a service has a single named view +There is more that could be done here to assist in building the pks and fks by specifying in translator properties and that can be added at a later date. +Two comments Nit In the test just let the unexpected exception fall through +Yes I will update service interface as well +To setup a node the user can manually remove this line +I just created and attached the change that needs to go to COMMON there +Heres a better version + +Deployment of JBI artifacts can be done since a few weeks already so I close this JIRA and well open bugs if needed. +Ill the patch now +Closing resolved issues for already released versions +Release Notes not required as the issue affects only internal Red Hat testing +Same error here tooI was working with FOP and everything was working fine except that I couldnt delete the images after embedding them into pdfSo I wanted to move to FOP +This behaviour happens as a result of the fix for and the broker being a variant +there is no more filter multiple check boxes have been replaced by a combobox so no OR conditions for now +Im attaching a new patch +What version of are you referring to +closing this issue as these requests now work +I should mention that this extension is based on work research done by Jean Deruelle to implement the Tomcat Embedded container for Arquillian +Patch supplied +All its passed +I have committed this +John The Jenkins build for this keeps getting aborted +Ive commited to a +Dominik do you still see the deadlocks +Verified with JBT +prior to generating release notes +bq +In jetty the source file had to be present anyway +Thanks Jason. +Patch to fail imports if the data config supports delta import but the file is not writableAdded a test to verify +Ideally I would like to see this removed earlier rather than later as I suspect it is related to some of the other issues we are experiencing +I believe I have encountered this problem with a user using svn import to pull in a cvs exported repo +Looks like parser bug to me +Plugin version are being handled inThe comments about oro and logback are from a post I found commenting on unavailability in recent versions +Username ioparra +peerdevelopment can now be used in the jmeter sampler. +Reassigning to if I can take care of this please reassign and let me know +Has anyone had a chance to look at this patch +Thanks a lot Vinay for the patchI also wanted to review this change +resolved by a install and an restartThis should be part of the installation +Usagecreate database location pathalter database location pathAfter alter only newly created tables will be located under the new location +ThanksHowever is deprecated in favor ofAny way of having work with Oracle +I opened with this idea +Ive tested and using it already +Isnt this Jira the same thing as +this is a diff between the original and the version containing the changes i think that are required in order to solve the problem +Thanks Colin +I remember back at a previous job having the software literally cripple drive access +The same code work fine in an integration test +Setting up a framework extension is annoying this extension must be made available in the target platform! That introduce much complexityHow difficult it would be to reimplement this stuff with normal Java +The s wasnt checking for connectionFailure and was waking up the continuously which was causing the loading issue +Closed upon release of +The existing move assumes that its part of the ring so it almost certainly needs some code added to handle the this still something youre looking at doing Yang? Otherwise we can close this +Simone is this one fixed +I dont know what happened but the code you sent has a right after ? in the URIThis injects a null parameter and then it fails the testIm now setting to use twitterj version +can only be handled correctly if we analyze the pom and find out about the maven name of the application +Otherwise Ill need to spend some more time understanding how to generate a patch suitable for application with the patch command +This was fixed during the last update. +bulk close of all resolved issues. +They are not really related +But right now its just a rough idea and needs some more thoughtbq +committed to trunk +Latest patch for trunk +It adds calculation of a summary on the server +i dont know what is meant with sandboxing +I will upload the updated patch that applies on trunk and with trivial edits to formatting and using in the tests + first complete patch for reviewing +Ales do you think we should have an admin guide for the MC as part of EAP doc bundle? Do you see the administrator configuring the microcontainer in any ways +Further I think we can assume that the streaming tasks only use the hadoop command line as a client +moving off to +this should be global to dependencycopy mojo however should have higher precedence which makes it difficutl to implement since dest got reassigned at the base so it is always available +Mainly because I had to move to a graph different library for my actual workresearch +Submit to review +I agree however Im not saying that Seam should provide support for these types of edge cases but should just try not to get in the way if an application decides it needs to do something out of the norm +Any idea on when is this going to be solved and with which Cordova version will this be releasedThanks +Just cloned this issue from +i was accidentally on the wrong jira + +thanks for the patch +Please provide a simple example with that shows the tried to remove as much code as possible see the new attachmentI cant make it any smaller the extension needs to replace the so we need +Updated to new plugin now opens up fine on windows and linux. +Again I believe this functions correctly in or laterOf course reopen if you find otherwise. +Have you looked at the code you are talking about +Many thanks for supplying the patch. +r makes the current IT fail stating incorrect number of files in archiverpm +Removing some files +I attached file to recreate this problem +Thats true as long as your client app does not need to access HDFS +Ill open a new issue if it does. +This will be done as part of the model refactoring work taking place on the branch. +instead of fixing test failures in I introduced so many more failures + committed to branch and trunk. +bq +Ill close this issue and sort things at Goras end accordingly +Under what circumstances might you not know whether a listener has already been addedremoved or not +DanielThis would create too much overlap with the Header interface in my opinion but I see no problem with adding a getter method returning an Iterator over all parameters +if you have multiple action listeners your solution wouldnt you please attach just the snippet of the jsp tag code you are using and explain your usecase where you need multiple action listener +Agreed +Any test that does not require the killing of a process can be convertedThis is probably all of the simple tests and almost none of the stress testsvines is right the testsystemtest need to go +Works fineThank you Mark. +The patch adds the license +duplicate am unable to reproduce this problem in current cvs head +Will fix for its affecting other people as in gt +The docco says that the properties will be managed in a nested contextBut anyways the patch does nothing but add a few more checks so I added itanyways +As already mentioned i have created a release for Cstor at the Codehaus snapshot repositiory + +I choose a high number so minimize the performance impact on stores while maintaining a delay +Your approach sounds good to me +is currently failing for me on trunk +Indeed but it is maven bug I think +The attached patch generates a simple Hive website with the same look and feel of the rest of the Hadoop subprojects using Apache Forrest +I noted this in a comment I think +Please attach a sample that demonstrates the error you are having +Hi JacquesThanks for commiting our work and we will continue to contribute more intoThanks RegardsPradeep RuhilL T Infotech + +Merged and committed +screen shots to show the much advanced XSL features +The patch was tested on LinWin x and Lin x by build test +I will update the patch to keep the original code if it is desire +At some point generating must have changed to use the author name rather than the ID and thus this change isnt needed anymore. +There is no default file name in New File wizard +It seems to be fixed in trunk as it throws the directly rather than catching it +That makes sense +There must be something that has changed in the new export classes that is different from the old for Decorator thanks for all of your great help +Thanks Arun! +ill keep an eye on this ticket I guess +Set the needed info in the descriptor and modify the filter to sql encoder +Not an issue +Anything special in your config +is the problem that its guessing for of utf space instead of keys actually on the servers +A patch for orgapachepigdataparser the patch as Hudson was not able to successfully apply it +New patch +Done +Please feel free to if more details are available +Thanks. +applied on JBT. +Fortunately this has been fixed for already to be available early next +Thanks sgoyal now has the correct perms to administer Kato project. +If this is still something we want to do then we need to at least upmerge the patch. + quickstart unit messages during render time are not supportedeg in a regular request the feedbackpanel can already be rendered before you even call info in populateitem. + was well tested on Windows before the release +At the very least if its moved to a place outside of the classes weve identified as the public API thats a reasonable alternative also +It definitely is the case that we are considering the exported packages from system bundle as in use +The project server is reformatted +I think I fixed this in SVN r +Patch applied in revision +I submitted the same bug a few weeks agobut I mailed the testfile to the mailinglist and did not includeit in the bug submissionwhy is this one still new +As you can see from the current implementation of borrowObject if there are idle objects in the pool borrowObject always succeeds even if maxActive is exceeded +It would also be helpful to be able to define query restrictions without value bindings at all +Verified by Rustem. +yes that would be ideal to evict depending on the DB +I checked the latest patch and it doesnt seem to compile without the patches for and +checked on works fine +If someone could provide instructions on how to set up a functional test for this provide a test that can be included to test it and then provide a patch Ill gladly apply it +Merged to doc branch at revision Merged to doc branch at revision +Can folks vote on the following names for a generic powerful agent? Choices are +Completed. +Please check that the bug is fixed for you +All other repos were unaffected +Will let you know what I come up with. +Applied thanks Dennis! +I will kick off a hard upgrade run to see if we hit any other issues with hard upgrade +Should we make the limit configurable? Also I think it would be good if the original job name will be displayed as an +Im vary of adding more knobs without more thought +Please correct me if Im wrong +More than likely its an issue with patch applicationAs for resolving your issues with the patch for now just get running +Thanks Raghu! +Ive committed this! +What youve done here is definitely prudent and we can debatebenchmark turning it on by default in another JIRA +Can you please take a look at the and check if the name of the service is indeed and check the wsdl in the folder of the aar and make sure that the service name in the services section matches thanks try nightly +prasad the patch seems to have about copies of each file in it +OK +What work has been done on the vfs component? Any source for cifs writes or readsThanks in do not think Ramon is working on this anymoreAnyone feel free to take over and implement a is a dead all resolved tickets from or older +attachTo is ignored if attachedfalse +and will be merged soon +Afterwards it is destroyed +If it loses the race it fails even with minute timeout +Committed. +Note that it is against Axis branch not the trunk +why? With the end of the body is also +Im not sure the edit command works the way you think it works +No exceptions or errors are seen when run against trunk +One thing that does seem to mess things up is precommit +It would be great if someone took a crack at fixing this +Logged In YES useridAddtional information JDK OS wink Professional Edition +The patch for this +Thanks for verifying John. +Dereferencing a collection is misuse +One possible solution which adds a button to the Bundle List panel +Done. +I have a deadline on Thursday but can get to it after that +Committed revision . +I am working on a fix now +patch applied with rev Thanks Viola. +This is fixed but an additional refactoring and simplification is possible namely to create a new class that includes the info in the inner class +makes sense thanks! +jBPM asynchronous continuations are a related messages due to their asynchronous nature may be a related problem as well +I am looking to see WHY the is needed and see if I cant code around it +A bit tricky to catch +Unless of course its an app client in which case we must always inject into static fields +Fixing the signature of overridden method get in would have fixed this problem +Thanks pass over all release notes prior to publication of +I think it depends if leaks the pid or allows killing by external handle +Moving all unscheduled issues to +applicationssrcwebappdataM +Moving issue to a next issues to next release +Derek if you want to work on this I can assign this issue to you +Since this is just a minor cosmetic patch I am just planning to commit the changes to both the branch and the trunk without tests and review +Can you reupload the patch for jenkins to run again +Well +I can do that +Ive added the dependency to the build but there will be some problems with transitive dependencies etc so the remaining Spring dependencies are still at +I just committed this. +Quick patch to clean up unneeded imports +Well giving users the choice to disable the feature sounds reasonable +though this is applied when I set the jar plugin to use the archiver it still generates the the patch didnt actually remove anything it just made it possible to override +Note that unlike the SP branch this does include the changes since the Reaper on the CP branch does not have and changes that make this fix obsolete in the SP branch. +Note that I also suspect that even in the Portlet case encoding the resource URL will add more context than we really want +Thanks I found in and +Increasing the timeout should be addressed here +When binding a queue to an exchange the queue name is set as QUEUENAME as wellRA +Good catch +youre right +Hi Kim just uploaded version of the spec to cover this addition +Hi ScottThanks for looking into this +Somebody else could help you there if I wasnt available. +You should be able to +It makes job automation more difficult +I took Uwes we can not automate this comment as applying to my request to handle it automatically within +Set Fix version to +In the description of this issue there is written that the Executor is now a web listener NOT A SERVLET +DavidIve noted in that the problem persists in +Youre missing some of the options if you want to be very strictSome of them are activated from the default +Logged In YES useridAttach a thread dump of the deadlock +Hi AndreyI dont know what could cause that error +Patch for ve integrated your patch thanks for your contribution. +In case you can live with that Id appreciate you closing the entry +This code path is changing in to fix this and other issues so we should probably dupe this JIRA to that one +Once we implement well ensure that shared dirs are always required +Resolving since the new is complete and is verified not to suffer from this issue. +I am starting using it now that we have help pages thereI think that having the high level Error pages here httpslocalhostcontentcontrolnavigateContent could be usefullIn httpslocalhostcontentcontrolweb we will have the site specific error pages +Closing issue as it has been released as part of release +Fix committed in revision +But this is another story How about postponing this issue and turn it into something like Drop DOM for performance and memory footprint reasons +Thanks for resolving this issue +Need to get things working again which should happen today +Its there now rev +ChrisSorry I think in my haste I may have misunderstood your intentIf you want to implement another version in minilang and show us how it could be done thats greatI however would strongly prefer to keep the Java version of the cart methods rather than outright replace them with new ones written in minilang +Ok Stack +My fault +Id be in favor of throwing an Exception rather than trying to complete everything inside awaitUninterruptibly for simplicityAdditionally since we dont necessarily have control over the Executor we might not be able to determine whether we are in a deadlock situation at all +Can you confirm that the same memory leak happens when devmode is disabled +Checked in on trunk not yet deployed to live site +This looks ready for trunk and Id like to see it in +I see you grabbed the issue after I started progress on it earlier today +I think the approach is to simplistic for your application try the way from the testcase! +xml I can double check that too but FYI the test resources in the original commit to did have +well probably wait until later this YEAR to make a move +This includes javadoc fixes and shim correcntess fix +This is part of our compensations API that is only in ATM and will need consideration when that is productized. +They are piggybank which we have javadoc but no forest doc +I just committed this +Oops I mean module archive not luni sorry for that +sbt works so does +Looks like this was done already +This will not be addressed in will not fix more issues in EAP RN fields +Looks good so commited it +this one forDo you think I should go ahead and apply this one or should we wait till the NTLMproxy stuff settles down? believe these two problems are not interrelated +OK I have submitted the IP Clearance to Incubator +see for that subtask +Those were recently refactored +Attached is that makes the changes for client RDBNAM USERID and PASSWD +James +Im working on that but I wont be able to have this behavior consistently as I dont have any code that could handle the new +Bulk assign all unscheduled issues with priority major or less to target +There appears to be an unnecessary diff change inThe request I think that is rather similar to this in that both cases involveincorrect handling of the intermediate nodes in the tree that is constructedfor a INSERT statement with a VALUES clause +While you can populate and inject them if you control the validation point you cannot do it if the validation is triggered by another framework like JSF or JPAWhich makes me wonder if we should rather focus on the integration between Bean Validation and Inject CDI where the contextual information could come from the injection framework insteadWDYT +I have checked with code base some time back and I was able to find some of the configs only referred in code but not in My thought was that we didnt have to do this that we only had to add ones we thought folks might want to change +See the annotation for example which has a required flag and a defaultValue making a request parameter optional +Committed to and trunkThanks for the patch and your patience gaojinchao. +And theres nothing on the geoserver documentation about this limitation either as far as I can see. +tests right now +But I assume there is a bug in the scope search algorithm unless JSF is placing the variable in the wrong scope itself +We should remove these constants if these are not suedbq +Alexey please verify +Chris indicated in IRC that this does look like it only occurs near the end of the commitlog where fsync hasnt yet guaranteed that data is safe +Can be +My understanding was that the was on changing the format now for please correct me if Im wrongIMHO this patch does not introduce a new format +I ran the tests locally on my machine and committed the change +makes the error disappear +Clearing patch available flag as it is my impression that there are no patches that are ready for commit. +as and we use will store the object according to +We should definitly think about mirroring the target platform on a fusesource box so we dont hit such issues again +Never mind +Suggest the website link be moved from nav to the Jakarta Retired Projects page +same issue in default and all configuration +Will replace it with a new site portlet by updating to the new frameworks bridgeNOTE this new framework bridge isnt available from an online maven repository yet so you will have to build and install the frameworks bridge yourself firstDavid could you maybe update bridges and deploy the latest jars to your bluesunrise maven repo? +Thank you Deveraj and BenoyWill close it if not needed in other version. +Now that all other dependencies are applied I can finally commit this patch which was done in version . +The failure frequency was very high due to conflicts from unrelated services being run on committer dev boxes +We were using a constructor of without the locator information +Maybe we could do an Apache redirection for these files since + says this is a required feature and applications do make use of it +Why does lets use the location of this node cause problems +Jacques can you check if the patch matches now your requirement +After this last patch is applied the only difference between the main canons and the j canons is the +the versioning is still in a prototype development phase and those issues are well known +Well for the ids have to be unique for all components inside the repeater onlyThe toolbars needs globally unique ids because its not sure in which dataTable the toolbar is going to be put inRegarding long running systemsActually youd have to create a lot of toolbars and put them in a single dataTable to run into duplicate ids +It shows this diff even in a clean subversion client without my patcha a Test Failed +Mass closing all resolved issues that have been closed for more than one month without further comments + +Fixed both the test and the code +Thanks +I think the trick would be to maintain backward compatibility +both statistics and restart data can now be stored as in the table BATCHSTEPEXECUTIONATTRS +Ill check it in +That a project root is on disk. +AlanOperationally looks like everything is good with your new patch on branchx +This issue is still present in GA +Dependency on is now optional +Can some one help me in resolving these types of issuesThanks Mathias I also come into this problem have you deal with it? And btw how you get these beautiful logs +Once it is resolved the cluster test for the Translator example can be created +Sanity checked on mysql and pg also let run for now +To confirm your proposed API changes do not break compatibility since it adds calls to the API without changing existing calls +StackI will be adding test cases for the rollback scenarios +It appears to be intermittent too +Jackson supports applicationjsoncharset +Committed revision . +Hi RolandAll the values in need to be uinque so its no harm to override equals and hashcode inIam going to apply your patch in a couple minutes +jffi needs to be able to unpack its native binary to an executable location on disk +Putting the exception stack trace togethre from within the debugger failed to write bundle caused byUnable to read index for string versionStoragecaused byBorrow prepareStatement from pool failedcaused byTableView VERSIONNAMES does not exist +lgtm +Please have a look +Im seeing this test fail with a timeout when I test everything +All you need to do is change the url and jmxPort in and Add jar files Compile and execute +Not that I know of +Rich could you please review and submit this patchThanks with revision +will ship with console works! +It isnt clear to me why they are treated differently +Moreover top location seems more standardized way see Jira Nexus Jenkins +Any queries youll fine the MR engineers if you join fusecamel +It would be really cool to have different files from the same region on the same machine but on different drives +If this does not work in WSAS then please contact WSAS support team to find out how to ensure that correct file is picked at runtimeI will mark this bug as resolved in version +But do you want to analyze each time all warningsMy recommendation is to make the code always warning freeWhat is the problem to initialize all these variables +Any steps to reproduce +Even after adding the new logic Im still seeing random differences with the expected number of documents on the order of or so +Patch v integrated to trunkThanks for the patchThanks for the review MatteoLarsDo you want this in as well +We need to be able to reach even years old bits because of patches +Have fun +gems build with a totally different mechanism than extconf extensions and dont build until first called +Thanks! +Use site plugin +I will follow up with Vijay but cant see commits for this +Stranger than that it seems like the file is not being generated on pizzashop entities but the file is +Len as per our IRC discussion you might want to have a second look at this now that I have backported some of the rewritten parts from to lift the standard of grammar and spelling +yes the exception and path are wrong they were already fixed in Archiva s true it is failing to convert the path to a reference that would be because it doesnt exist and it doesnt proxy directories +But I notice it does not contain a flushA major compaction issued by does force a flush and hence stuff thats in the memstore will not get compacted away until it is actually flushed to disk +Removed the statementsvn revision +Once the data reach peta bytes then it makes more sense to have a time leased index where each part of the index could expire and remerge more easilyBy using KATTA it may be possible to have the linear time index partitioned and updated on multiple server and search will broadcast to all index and retrieve the result more efficiently +Rado when was the last time you saw this in a test run? If you havent seen it Id say its safe to close this and the related JBPAPP jira +Ping +Hi GuidoI generally agree that comments are good thing to keep the reference +Is there anything else youre doing +its not but since patch is trivial and you ship thrift library binary with cassandra you can ship a patched one. +I could try to look into this issue +Let me know what you think +see my description on this bug +The blackfeather extracted garbage text is a result of lax byte matching rules on handle +become runtime exceptions. +Was fixed for +I will commit this shortly +Thanks for the extra legwork +Without java classes adding a runtime provides very little value +Theres no point in removing the xalan URI references from WSSJ as theyre needed at runtime anyway in +Nasty little bugger +What about for the case that just one or a few files have the small block size? You wouldnt want to put this on the NN web UISetting default block size smaller and having all the files affected by is a big issue +could you also make a small unit test for this? and maybe one with a? Then Im prepared to check this in +Check out and build the current head trunk +Yeah +I created a testcase as Michael described but it DOES NOT reproduce the problem + the proposal sounds good +I will add testcases tomorrow and generate it without CRs and with real svn +I found that it woks well after replacing texthtml as textplain +Besides imeem many of the files include Facebook copyrights +I move it into +The rest configuration API already works this on and transition all resolved issue that did not see any further comment in the last month to closed status +bq +Ill work on this for completeness but I dont know if its a significant issue +The should stay as fixed until we actually release the version they are targeted for. +Thanks for the patch +Fixed in close of issues as part of making release notes. +Somehow the first half of the changes were missing from my original patchFixed in revision +Thanks very much for your comments on this OS +required for class unloading +Im still seeing the same behaviour from though +Updated patch that changes the constructor for to take a StringI will commit this shortly +All commons projects changed to dev +There was no response +Thanks CharlesThe patch was applied to trunk r please verify +Used the pom from +caN wE diTch tHe camelCase toO +Then when we print it out later we can see the call stack that caused the thread to get started useful if trying to track down a leak + get with nn address fix updated with the branch +Okay thanks Brett +However Ill peruse the SWT bug list and see if anything matches what you are describing +I have attached the methods added to the class for this feature +The system is hard enough for users to understand without asking them to do this! Also if the default is this will effectively not be used +Thanks for the report +Please also make sure that you have id of all all html elements +Okidoki! Good to hear that you do get proper proxies thereJuergen +I used ver and ver +rev fixes similar problems in the tomcat integration +Whats keeping this from being committedDo struct fields support modifiers as well +Now that Apache is out and modpython close to release maybe its time to have a look at this for modpython +Its resolved as far as Modeler is concerned +rollback of r and rtrunk +Moving to EWS +Updated patch for Y! version previous was missing a public declaration. +From MS interface we can take a snpashot for a storage device create template etc with the patch +this patch has been tested and I verified all the methods using +By the way is it envisaged in grails to manually put the plugin into the folder and issue the command? I think the issue was of an earlier version of jquery zips lying in the folder +technically persistent objects remain attachedvalid only within a single request and should not be stored in session state +Please verify in ER interface minor version major version public class extends minor version major version major minor Java platform version +I will have a look at itGood luckLeo +Ive just committed this! +As commented we do not include the zookeeper jar +good question +Now that work is underway with AS all previous community releases are +Fixed in bbebfadaeeacfbedeffcd thanks for the patchSorry for the delay but when I first saw this patch I was almost sure there was a problem with it and it just slipped day by day until now +Patch for Ill merge that to too +scommandButton does not have a way to control the processing lifecycle of the inputs AFAICT +Sergey how can I obtain latest snapshot? contains for working onPrzemek +Moving out of +I updated the pull have browsed the codes +Logged In YES useridThis has been fixed by maintaining the principal seen forauthentication in the web session level cached value andpassing this principal as the security context principal onejb calls. +WernerIts a good point +Kan suggested to make the return type of the API to be that encapsulates the new GS and access token +Briancould you verify the reasoning behind this erroneous event while adding new primary storage and checking out its detail view +Thread dump of the hung process +assign to me for and checked in all Appears to be resolved so closing. +This is a small test case that modifies the formsimpleXML sample in the block to use a cocoon resource instead of an XML file +Its unlikely that I will get a chance to work on this +JSF requires proper to render all action URL beans should use binding to the current src if they need reference to the included view. +No the original writer was killed at and never restartedLease recovery started at by another process on the same machine +I guess the underlying issue there is that theres no corresponding mapping element for that +I just committed this +I dont think this is your problem I suspect the problem is elsewhere and this is just a distraction. +Wont fix +HelloWould you please try my patchThanksBest regards patch metadata with respect to moves +on correcting the typo in this issue +Moving to +The use case you talk about is highlighting +If the factory doesnt need to do any complex stuff to create an object instance then why not let it call on its ownAnyway its just a thought and your proposed solution would solve my original problem so Im really happy with it! +Thanks for setting this back Ben +apache hive is no Hive driver in the Apache Cassandra project +The SNAPSHOT dependency on Karaf is just a maven plugin used to test the features xml file +Can you please post comments there +Attached another patch +I will work on developing an alternate persistence layer based on JPA for the SJC petclinic sample +Removing from onwards. +Marking this as resolved as spec file has been updated to include apr newer than or equal the specified build and committed to +Hi scott we are not using on our cluster +But is totally unrelated to docvaluesThats clear +Right sorry I misread my own spreadsheet +Resolved in latest changes to the plugin part of changes to +The stack trace is the same +Unfortunately validatedocupdate functions are called before the attachments are processed we dont know how big they will be +The new comments look good +Comments are welcome +I am not sure why findbugs is complaining that these issues are new +The idea was to introduce as little custom code as possible to accomplish the wiring of the job parameters into the job and step +Werner are you busy with something else? Please let me know if you can spend some time on this issue +Also write commands for renaming resources +John thanks for reporting +bq +This is the argument for having our own or using rd party libraries + unset fix for has already been released and its too late to get this in for +Hi Ranjanplease attach a small testcase stylesheet and input xmlRegardsBrian Brian stated this doesnt appear to be a bug +That looks a bit weird +fixed on and trunk +Currently files are added only when the descriptor has loaded successfully +removed the assert will enable asserts in a separate patch +patch against jwaldman lineChanged paths M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadinternalio M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadinternalio M myfacestrinidadtrunksrcmainxrtsorgapachemyfacestrinidadinternalresource Trinidad doesnt support surrogate characters +ping here the pull requests associated and Ill help review them +Im not even sure I understood correctly how placeholders work can someone help me out with thisLet me know what you think about my patch +Close and move the patch to this or dicuss in +Rebase on latest trunk which now has post +Im arguing that that the jobhistory code should use the right blocksize and buffersize I dont see the value in making this configurable +The index is not corrupted +Patch which adds missing deprecations +I forgot to fix the trunk itself +Marked that its also in +VERIFIED. +GreetingsJohan +We were under a attackfor the past day +Thanks for clarifying +looks great +No more alarm bells +Is there a way to enhance the test so that it triggers negative lengths and negative offsets and verify that they generate the expected exceptions +Moving this because it will hit after the release and get pushed into the platform +Log file from could not reproduce this on the IDE version of the bug in +Screens in JSF Tools Reference Guide were updated +rename package when proxying java +Thanks for reminding me! +has been created to port the fix to the EAP branch. +Fixed in commit cbaf +Awaiting word on whether the fix can be provided for dont provide official support for the community project you can however simply switch to the new XNIO snapshot and run with it +Check the code coverage report +I just committed this +Closing resolved Release +Closing all issues +I wont say yes to inclusion in a release until its running in my live service Im going to start doing that testing tomorrow it survives my service baking it itll survive anything + +Yes I believe that is correct +I dont know what was amiss. +I got totally confused and misread the test file name in the patch +Im going to assume that the related issue has the fix +I added a as well +Committed at subversion revision +The incorrect setting of statementtimeout is the likely cause +pathc from thx +Marking as done as we dont know why it was reopened. +Ill ask stefano to check if its regression on the within the subsystem +It will be available in AS as soon as we releaseTo use it one can replace module manually +msmith how did you get this NPE? From the stacktrace I can only see that you pressed the cancel there any other exceptions in Eclipses error log +So all issues not directly assigned to an AS release are being closed +Attached screen shot to describe the issue +The second table of contents is unnecessary because Forrest kindly builds a table of contents which does work +That is it can check if the value of a primitive was not set in the bean definition +I looked at the patches and they look quite good very simple and direct and creating what I think is a very useful and important abstraction between the network server and the engine +Grab any youd like +yes please open separate jira for each problem +A patch for review +quantified comparison +All the tests for this code run except for the regular expression test +I have tested it with version +Edmund could you give the latest snapshot a try? We are now adding args and method to the evaluation context for those header expressions in addition to theMark +Close all resolved issues for older releases. +This issue one provides more info though. +This issue has already been resolved +Which compilation error? Not sure which everyone is talking about +Set the component to Build System +When all applied to their respective branches the jdiff tool will be at parity across all +Few updates to the archive Possibility to change maven version added to scenario automation +Yes it works. +fixed in r +BJThanks for bringing this up +Hi I would like to know if this new Parser is integrated or eventhe standard in current releases +Whats this Result FAILURE message + +Hence the cluster concept is already set in stone in Chukwa +We have to revisit this way of making it asynchronous as it might not be acceptable for this patchIm still to check why is failing for Utkarsh +ETA for fix Thank you very much for providing an ETA for the fix +Reopened to set the Fix Versions field to JDO beta. +no longer depends on anything produced by AS trunk +Committed the a patch with revision +Patch checked into branch and trunk +The Specification formalizes this mapping +Shouldnt this go into the Release Notes? I believe that is what Dan has marked on the issue +The test bucket runs without problems +reopened for fixVersion +Made an error in the attached example file the newer version at least doesnt have a XML parse problem +Let me know if I can provide more data +b On loading the site youre prompted to log in using the same registration used on + tested +Lets remove from language +Yeah thats the idea +I think that it should be safe to backport this fix to earlier branches +The importer logic already covers adding catalogschema to the name +Removes the POST from the a lot Jasha patch applied in rev +Tagging for family +shift to next response from Sean and in is moved to trunk anyway +So Im going to close this issue. +We are doing the sublist after the get eligible files I see thats a separate am going to move it to apply +Ok slava add it in if you think its safe and generate code that runs btw +Theyre not related although they are concerning +Maybe we have to close the issue and keep as it is +I will commit the changes in early January +Its inThanks a lot Matthew for your work +In what sense did things work +Should be an easy fix +LGTM +Apprently this is not a problem as another bundle expresses the mandatory requirement on the same service +Why should I do that? English is not my native language I do not talk English and possess only strange writing patterns from scientific papers +See the attached file +otherwise the same logic as u described +So that others can help you in parts when they get time +All the jars that were already signed lose their manifest information +I am taking the patch file off the JIRA issue since these changes are specific to our university configuration and do not apply to all environments. +emptyFile does not appear in part +Please reopen if you have additional detaisl on how to recreate +The behavior is coming after a decision but Im not using swimlanes forks or joins +Patches for both the openejb code and the related Geronimo plan changes to leverage this code +I saw same behavior on a different machine on a different high speed network against a build version from sometime on +Removing project archives support completely cleans the project of related artifacts +the settings in the web admin always overwrite the settings from server admin the settings in server admin are only default settings for the web admin in server admin you can define what the settings are used for a new web admin +It shouldnt be too complicated but Ill be out of the office next weekThe interesting part is the rationale behind the change in the config file since the version works for me +Not for commit here. +On linux and in intellij the resources are being loaded as UTF +Also had to chmod startup +Im putting this off until later +The way my application is currently written it has a number of pages created the standard wicket way +Thanks Carl! +Daryn decent point + +Jay taglib has to be OK in. +and unable to proceed any furtherpl take pity on me and kindly fix it at the earliest +Albino Biasutti Neto biasuttin at gmail dot com +Basic functionality of CND editor verified +OK fair enough +To me the will hardly be useful for confirmation scenarios but rather for creating a file as output from something else +NOTE you will need to use JBAS r or later +Well technically speaking you could do that on any of the transports but hopefully that random data would generate a marshaling exception and shutdown the connection +Im going to update the attached jar just to double check but as far as Im concerned it works for me with just the replacement +Fixed by backporting fixes to found this issue in latest builds +This is done to check that no place in solr uses the old methods anymore +I guess its more a Spring DMLC config issue +Sorry I could not get to this sooner +and so he does more than his name tells us +Studio works correctly when connecting over or when using startTLS +I have update the rails app to generate a virtual host for the projects website under the the project is public there will be no security applied to the site +In addition to embedding plain you can also now embed Files directly. +The only problem that remains then is for Android because IIRC it used to return something like as specification version +A lsr on the same cluster took minutes +target in Target editor and click on Set as Target Platform in the corner +Trivial change to remove some commented out patch applied patch changes composite and JDK invocation handlers to only set fromAddress in the message in the forward directionIt passes all test cases in core and spi as well as the supplychain and simplecallback sample +Waiting to commit. +unzip the patch to NIO module +Now we have to fix just fails in junit tests. +Committed in r +Hasnt failed lately +What do you think Greg +defer all issues to +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Larry this isnt just about coders being lazy +bq +Which makes a lot of sense to insert the interpretation is still open +A patch which changes class to behave as described in the issue +to Apache License you are requested to do so before making any ASF releases based on this code +fixed in these events to input in +I will retest this issue using a more recent release +myNode +Letme try and get a basic implementation in and then call this patch quitsNow if you enable ipc logging you see the complete client side request and over on serverthe complete receive and then what it writes on the wire +The user must close the app completely reopen the app and then the localStorage is backEDIT If in step a user starts putting in new data it will get wiped out in step +Few doc book documents are modified however the connector developer guide is not modified as it requires more still left to be done is to modify the Web content +I looked at but the error still occurs +Hi +Changing it and seeing if this fixes the problem +They may implement Comparable but they dont do the widening for us +IvanIve refactored as per the comments please go through this +See +I also did the same thing forSorry for having multiple patches uploaded Ill see if I can delete the old ones +While the metadata is being updated in memory by one thread we get a response from the bookie and ack the entry +If you need any help to test it dont hesitate to contact me. +testloadpkey in test is a test for this issue. +It is mainly due to the remove of aggregator a mojo with aggregator should resolve this this change be done to the? If it is will someone let me know so that I can try out the with this fix +The prefetch value is applied to network subscriptions but if the remote broker can accept a message it will take it +John is now a committer and he has restarted the work in his sandbox +Stephen As always if you can produce a reproducible case we can probably fix this quickly +Very much agreed +Change to just make a copy of the cmdArgs array without modifying position + apply this patch to see the build to fail fix of the bug which works for in rev +This issue was fixed in +It would be nice to have a wrapper that adjusts these criteria in an adaptive way depending on the problem at hand +The problem appears to be that is used to set the path to the stylesheets in. +Thank you for the patchId like our examples to use the preincrement operator wherever possible its more efficient for iterators such as those of deque +If there are duplicate attributes in the TLD that would seem to be amistake on our part +AttachedThese patches add a connection to the following tests +It seems I have been a little careless about it +Two possible solutions +bq +Im not sure how else you correctly handle errors if you dont do this +Thanks +META +We intentionally kept Springs Hibernate support rather minimal in order to stay close to the native Hibernate ways of doing things +Created to follow up +My allergy to it grows +Fixed +I can open it +Did not handle properly subsampled images too +Ive committed this +Had a look at a looks good +you so very much Ulrich for the fix! I will be out next week but let the Derby community know to reopen this issue if they see any problems +Also i dont believe i have commit privileges +There is no change +So it might make sense to give them the basic capability and then start to layer on top of it with test awesomeness +Since there is already a JIRA entry on problems with volatile long becomes a part of Thanks for fixing the problem +updateInput is used internally to keep compatibility with the old way of using mrunit +Committed to trunk and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +JAXRPC tools issues may be fixed by community contribution +If you want you can try the attached Oracle jar which has been patched for and too +Updated patches against latest svn code +Is this bug still valid +Heres a test case +it does not do any emit or progress other than sleep +better thanks +Fixed on the last commits this should be fixed on trunk as well +Fixed in r +There wont be a release +Please explain what are you doing and when exactly problem occurs +Thank youHowever adding Bootstratps CSS to a bundle seems to fail as the bootstrap icons are brokenProbably because the path is relative to Bootstraps CSS not to the resource bundles cssAny idea on how to solve this? +Id suggest to go this route for For may be we can provide some class which has all the magic removed and since it is not derived from Border the exceptions in Form etc would not apply +It is just that didnt handle filenames properly for certain edit log entries +However if there are no skipped lines that will be empty which the repository will attempt to pass in as null when recreating the context +Updated patch slightly modifies the error message to look more like normal Linux commands +Ive just committed this +It was released with the fix in +Will address this and other minor details in the next patch +I dont want this implying API here so I have chosen a different name +Attachment has been added with description patch for misc javadoc should have noted that this patch was built for the maintenance branch +unmodifiable some have a protected constructor to allow some have a public constructor that does not validate the inputThis needs to be further cleaned up in a reasonable way +closing in bulk all issues marked as Resolved +to trunk and added more on comments on phabricator +Thanks for testing the patches +Based on a review Im not sure if I like how all of the failureMessage stuff is handled but thats not because of this patch +Thanks TonyVerified at r. +Here is the sequence +My only difference in environment is JDK +In the longer term we may consider other ways to do it more efficiently for with a small number of characters +I have updated to reflect the link into but it still needs to be assigned to Kurt +Jon said that the new location ispserveranonymoushomeprojectsprevaylerscmHowever doesnt even seem to be a valid hostname and is actively refusing added to DNS +Added a check for the compilation unit being null and trying to compile to help solve problem of outline view initialy being blank +Thus as a general rule and best practice I would recommend that anyone using the source generator choose a package that they are not using for other code for their generated code +Updates would include Put Deleted ICV etcI can supply a patch that does that +Do you think adding a configuration property is a valid enhancement that you might consider adding to a future versionThanksDaniel +Justin I read your patch and it seems to make sense +All you need to do just type java jar when downloaded +I dont know what project from existing ant build feature is +Ad +The patch is applicable to all the main branches trunk +It will help to keep the graphs up to date +There is another in that needs to be fixed +Retargeting to +If you think it makes sense you can include that in your testsuite to make sure the tag is fully tested +Let me redo and add to the patch +Ive tried to reproduce the problem without success +. +Please upgrade the svn +Changing to ClasslibLUNI as proposed patch belong there +Thanks for raising the issue which brought to light that with was not documented yet! +The observer would also need to be slightly different because it wouldnt need to look for a leaderI wonder if this is something we should pursue in this patch in another patch or simply forget about it +KapilThis is a place for bug reports and fixes +fixed in revision . +Actually I have a fox for this almost ready + +Looks like were building up a few patches there +Good point regarding saying the broker sets those +Committed +on removing a test that is too general especially since it has not yet been included in a release +Thanks +I AM SO LOSTWhat source code repository should I be looking into for the GERORNIMO? I this actually in the Geronimo source then +Fixed +Ill open issue for the nasty trap. +Fixed patch for consistency +Dependency on wtp features removed from in trunkResolving as work is done +First patch for portlet Thijs what sort of percentage complete would you say the support is at +The generated data file will contain a database of bug reports with and release numbers the stylesheet will select the first bug in each release using Muenchian groupingFor instancejava IN XSL PARAM numbugs OUT XSLTCjava IN XSL XSLTCTry values to get a sense of the performance trend +And then it ran fineLooking again at the OOME from an old mail message of yours try setting the HBASEHEAPSIZE +I think you meant related to not itself +Tom suggested that the problem may have been data from previous testsinstalled versions in the DB +This is not caused by objectstore refactor +r +rfeng has applied a change to fix this problem. +Thank you +Includes a new test case in +checkstyle +Changed the VC projects to create +The cause of first Exception is adding temporary files to before save +Attaching a patch for branch +updated to reflect the actual problem +and show some examples of bad request templates causing problemsI dont think having this as a JIRA issue is being of any benefit +Should be fixed now in the commit above +No patch +Set correct fix version +If you want this issue to be backported on the branch please poll the users list +Git rebase worked without conflict but here is the current patch directly instead of through arc +updated the issues summary to match the use of m deferring this to since we should consider the options more carefully and also consider how this may relate to new features planned for such as EL and annotations for return value binding +Updated the quick start docbook to be less designer centric and to update the links +Would the best way to handle this be to extend runtime exception with a custom error class +I can approve this case +FWIW I have run BB with its two webapps on seperate machines +We should maybe point this out in the Javadoc +Ill it in a few days stay tuned +I dont think it buys us enough to be worth the hassle of dealing with the extra port +related and describes more requirements for types of connections +Which should never be null but may be null if a persistent store loses the publisher information +Even then it is really up to you to tailor your security policy as required if youre running with a security manager +That is because a was being raised when the engine tried to fault in +One issue is that the relative references to the JSP are often different than the relative references to the action +I did look at the wrong rule +Screenshot of the Pluto Admin Portlet Application home page +as patch has been shipped as part of. +Another good thing is we dont need to change code on both as and hornetq sidesThanksHoward +Seems like this is another case where users might want to define these outside of a VDB and them by selecting them in a VDB EditorPossibly +The encoding used is did not compiled with any c flag can you please elaborate what this flag is +What if we went back to the old scheme where we added a delete for every cell deleted so if you deleted a row itd add a delete marker for every value in that row? Deletes of rows or column familiies would be expensive but so what +If you are using overwritetrue this should work rightOr perhaps someone else has some ideas +Attaching a patch which handles following Invalidates the replica with lesser genstamp which would have been created during write pipeline failure +I think the former but Im not sure +Its growing fast and gets too complex for me to foresee problems. +returns a null +Posted from Bugbox for iPhone +Could you create an own issue for this and commit with a new issue number that we can distinguish the two things +Having a lenient mode is crack if the value is set it shoudl be checkedOne question how does this patch work when we federate messages? Messages sent across a federated link will be of different to the connectionit would be nice to be able to run with the option on for client messages but still be able to federate messagesCarlRajithThat sounds like a good approach to try out +upload a quick fix to let hive always get column names column types serde parameters from table metadata +contains spaces but the source code has tabs for indentsIs it a problem to make it with tabsThank you +Reopened +I can fix the resolution directive +simplefs is slow too +Without ordering things you cant really guarantee that thoughOf course the same problem exists if you add multiple documents with the same id you dont know which will end up winning by being last +Whats the alternative semantics? Ie +bq +What you could try to help is reverting the commit and see if it fixes this think that the Facility tab could simply be removed from the Store +Merged into trunk thanks! +Please try the latest release +Great Im eager to look at this +do urls have now been moved to clean url is not being used +As per the patch it will report error +Applying this patch the error message in your case will look like error Function not supportedIf every method in is implemented by its +to both because I think were rushing this without a significant amount of thought as to what a directory structure should look likeIn my mind symlinks really dont seem like real symlinks to me but a sort of hacked version of autofs support +Sounds good to me +currently in NEWSUBMITTED on recover event new app will be created + +Christian Does db have some sort of implicit key structure on tables that is being picked up? Like the oid concept in postgres +I attached the updated patch. +Its not intended to be checked in just to be useful to people who want to run types on instead of +No Im seeing a timeout rate +Done +The only problem is that the name is now identical to one internal Java class but we cannot change it without BW breaksFor this special issue we should fix only to use Javas PQ which dynamically grows when new elements are added +About the only thing you can really be sure about nowadays is that can be ignoredThanks for the report will work on getting them fixed +to be committed +Verified in dist built from SVN . +Testing the parser alone doesnt seem very useful to catch regressions +There may also be a bug defaulting enumerated types when they are instantiated +A workaround for thisis just to ignore the errors +Patch in fixes this +convert into a date +But note the tags it is in are the classes you want in? That would explain one problem but not why the files were removed +Thanks. +Closing all resolved tickets from or older +Other points If Arquillian should handle deploy and undeploy ? If the deployment should be enriched for upstream master +fixed in in +The same would be done for theThe and JCP NDA dont have similar text +We might be reintroducing a problem with scalability when we start putting snapshots in one directory s hold off on this for the moment on the upside related things are more robust now +Attaching the patch removing the waitingtasks column +Tests GET and POST but not multipart POST +Sorry doesnt appear to have fixed the problem Im having +Second patch committed to trunk and +I must have deleted the correct version of this scriptSorry Lars for the trouble +Sure +You are right thanks for the patch +Done though I do want to finish up some related refactoring. +Too bad +However I can still define a component wo wiring the reference when I define component +And wed probably also need to implement the equivalent for generic data +Lets make it a task so we dont forget it +I still have to run a install first +Trying to figure out why other situation had problems +sorry +Committed to trunk with revision . +the client will be charged until RM gets the report +I agree that adding type artifacts as modules by default is incorrect +is fine without this dependency because it just pure java libraries for bridgeIve fixed fremarker in trunk +What matters is actually whether my patch breaks things +Ok Robert so Im closing Wont Fix this ticket +Patch to address first write when appendtrue +I had considered that when I created these exceptions but wanted all path exceptions to derive from a common class +regarding key change we can deprecate the DFSWEB is good to allow the keytabs that need to be shared between services such as spnego to be shared +If you want to update a nodebased on the contents of its neighbors you would first read the list ofneighbors from the node then read each of the neighbors contents and thenfinally update the original node +Moving to to add the ability to leverage acknowledgementModeThanks for the heads up I moved the ticket over to the component to expose the capability +mvn deploy works but you cant get the artifact back out of the repo and Search function cant find it +This is not related to the patch +esb archive is deployed zipped then it sees updates to fileif the +Ill check it out soon +Unit test +Sorry but I think there are far too many changes in the patchIts really difficult to follow the patch and it will be difficult to follow the SVN historyI think the patch needs to be broken down into more manageable chunksAlso there needs to be a vote on whether it makes sense to change the behaviour of an Expression so that it behaves like a Script +metadata +Can reproduce +Does it matter that we dont allow server principals like a and insist on ac? Does SASL insist on it? It is certainly the standard practice but we are forcing it as a requirementIMO I do not think Hadoop should force it as a requirement +Could you take a look againOn a side note Do you know what is the matter with? I still always get the Error message when I try to submit my patch for reviewThanks +I think this was somehow related to discussion on compaund columns we probably should close this as wontfix because I also dont remember why we wanted that +Go to project Build Settings target Architectures DebugRelease Any iOS SDK +OK the issue is actually clear +For example if they extract the zip to a hosted Nexus instance the index information will be automatically generated +BTW it would be better to generate the patch with option which might be easy for users to apply +Thanks looks like this is now fully implemented +No problem I did exactlly the same when raising an issue elsewhereHowever I happened to notice the create so I raised a JIRA issue to get it fixed. +basedir is a global and is only set once relativebase forces the task to reassign basedir for every fo file +Fixed by new Relation component and related changes in +This is probably some error I made but I havent found one yet +I tried Daves combos above and it seems to do right thing +When calls the it passes either DROPCOLUMN or DROPCOLUMNRESTRICT depending on what the user specifiedTests and test output is also updated to match the new correct resultsThis patch is not ready for commit yet because I need to address the interactionswith GRANTREVOKE but that will require investigatingIn the meantime feedback on this patch would be very welcome +Also the accumulo proxy command should be run from the installation directoryI also got a chance to look at the proxypythonREADME file +Ok i tried on linux and it applied OK +from me Tejas +But Ill bet if you take that slash off it will work and thus this isnt really a blocker +Marking closed. +Whats the motivation for more than one wave of reduces? Typically applications are better served by fewer output files so that cant be it +Thanks Igor the Bug seems to be fixed! +Checked in fix to for this extra code changes for the fix to compile under it! Part locking makes sense and part boost namespace compiles on windows +to be subsequently fixed with proper covariance support hence reopening the issue +I guess the codehaus lists use UCEPROTECT blacklists as a way of blocking email +slight updatesAdded some license headersDropped the util package just easier to keep all in for the errors +any chance of a stripped down testcase we could commit? this one is pretty long and depends on multiple things not included in Qdox think this might be related to Ive just filed this one also contains an enum that is not at the end of the class +VCL Image inventory will have to be expanded to include additional fields specifying privacy and security of the final Broker Interface is uploaded +Ive occasionally run into the wtf? problems but just fall back to +Would someone on Java please test the current Forrest trunk SVN +I attached and? at Aug th at and AM +When completions are in insert mode the replaceLength is not used but instead only the replaceStart and actualInsertion point are used +hmm looks like I took trunk code for some reason +Closing stale resolved issues +Tim I dont develop in Java +I see what you mean but any simple http retrieval has a chance of requiring authentication isnt it fair to throw an for attempting to connect without credentials when they are required? Or would you prefer to just roll that into a +Hardyyour argument makes a lot of sense to me +We can commit it and keep looking into it +not currently relevant and not sure if it ever will be. +So we are free to go ahead using the logo +It is unfortunate it was encoded this way +The error on Mac when using Open with browsersim +Im sorry for my English skillYes autocomplete dont workIf I tried to get parameters of method that I get message No parameters +fixed in r +Verified status of this issue sure I am not sure about this issue leave to open +Thanks for assigning me +A average like that will quickly become meaningless + and generated source for protobuf is not compatible +The authenticator now returns usernamepassword one time and null on the next call +Thanks Shi +I know this part has been in flux +Default is zero meaning a connection will be kept open indefinitely or until the OS decides to close it +Why dont you do it and assign this issue to yourself? We had a chat today about this on IRC which I posted +Added patch to keep references to start of chunkeddechunked buffers to avoid blocks from being released before buffer readers are allocated +No Apache has not made a release off this branch +I cant reproduce the issue using either the latest version of JBT from the update site or using the JBT trunkWhat version of JBT Central are you using +I have attached a screenshot showing the output of an exposed service +Forgot to close this +can you please make sure youre working from trunk here +Having looked at the patch by the time it was made ready for trunk maybe of the original patch would remain +Heu it should at least throw a correct exception instead of a NPE no +After a try with plain Maven CLI invocation it worked properlyI will report this to the issue trackerIm sorry for the noise! +Hi Yasin +This patch fixes the crash +Same patch with the MR changes removed +In supports index and only a portion of the index is read when the file is opened +And default it to true on Linux and false on? You can use uname to find the operating system name in the script +While the pull request was closed the issue is not resolved +You can disable optimization to get around this issue +I just want to make sure there is not another part to this before resolving +Lars Yes +Hi Jarek Thanks for the review and commentsI agree with you this is a temporary solution so I did not commit it to trunk +We can add an API in that takes a timeout +Then Id say see rd party documentation for more details or something like that +Not that I know of +Is this actually an issue with Springs support? I would rather consider this as issue with itself or with Axis as the underlying provider +This is a gross violation of the HTTP spec on the Microsoft part +Seems Pig core is not using isNull so not dealing with it in schema tuple implementation is totally fine +Committed revision +The patch is not ready for review although much of its functionality is covered by the new regression test included with this patch + +This results in the oldest entry remaining the same causing repeated scan verification of the same blockI will review the patch to see whether it will address this problem as well + +AnjanaWow +Implemented in SVN trunk +I used the zip of as and the zip of aop will probably need to run the installer and pick the EJB option to see the issue. +not sure why I have the impression the lower level tiles gets reloaded even if I have http caching headers enabled +This improve the manipulator efficiency when annotations are ignored +Assigning to the EAP Maintenance lead +You are seeing a HEAD because it is a HEAD +Hi JeanneIve noticed now that there are a lot of aliases that dont end inalias +SSL is not my area of expertise though so any help is appreciated +In fact Evgeny with both Jenkins and Hudson I dont manage to activate the two first options Skip if +The is the latest patch based on feedback from +ok no merge needed. +This should probably be downgraded as a very minor feature then. +Committed to branch and trunk +Tried on the build sample revision +This ensures that we dont block the merge if there is no possiblity of more data being available +Will be marked in RS. +ivy dependencies file with dynamic revision +Peter C +Verified in and +push to +I totally agree with you that logically related to the build target +Why cant you just change your appender for this category +Fixed the bad committed the fix to the javadoc. +Patch to remove commitoptimize query args since the request already contain an commitoptimize XML message in the POST body +. +I felt sec is small period +This can be done as a resource or a URL + +Where do you think if should put thisThanksDaniel +I have committed it to branch +Of course using as few threads as required is better than using more +was the fix intend to avoid future problemyes +the case for custom parser registration via a new ctor arg to appcontext is a seperate jira issue. +Partitioning by token ranges is functionally equivalent to virtual nodes no? Which in the OPP case means you now have to deal with load balancing +I ran usingit worked correctly +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +More Actions Move +Patch looks good to me +So all issues not directly assigned to an AS release are being closed +hmm signing script modified maven artifacts +Sempre he sentit lexpressi en frases en les quals lexpressi constitueix prcticament la frase sencera +patch creates stub if deserialize fails that assumes every mutation needs to be still could cause a problem if the cosmic rays aligned just right but since weve already rewritten it for and since nobody has actually hit this in the wild lets leave it alone for the series. +Another option that occurred to me was to create another logj logger which would be used specifically for logging the traces we need +I tried on Windows XP and found it picking the right environment is this issue reproducible in? Or has it been fixed? +Right now a release of Core is on hold because we cant update the site and in that way we cant send the announce mailHelp is most welcome +on +As a first step one could make choose inSo that if a block are over replicated it is removed from the machine with least percentage free spaceThat way all we would have to do is to copy the block to one extra node with lots of space and the rest would solve itself +Thanks this right now +Thats a pity thought that jsch would cut itYou welcome Ill try to provide help as much as I can +but ultimately wed need a more robust comparator factoryThats the idea +Since its better to relate these two we need to make sure that we dont report complete in case of a failed execution +The test case shows that this fix stops a plugin getting a stop request. +Patch for is the internal class made publicThis patch is a on a real issue we have yet to find improper clean up of rows +When a SOAP message is sent and if it is being converted to a POXREST message at Synapse level and then if a SOAP fault occurs it is converted into a POX fault +Im certainly interested in taking a look at this +Sounds good to me +Seems I cant closed it +Exposing things from the vs things happening in the balancer is totally that just needs to be made manifest in the naming and contexts description +Please provide to reproduce this issue +WHERE updated that runs the last two combinations WHERE EXISTS +James have you already started with a fix? If so could you send me the source code so I can build my own deploy plugin +It fetches delegation tokens from namenode only if security is enabled +ouchhhhh +Committed to +Derbyall runs cleanly on modulo the wisconsin diffs +they still reference incubator etc +Sorry to say but this issue is not working yet it +Only project with and SDK will go wrong +This is the second attempt to update the Derby website +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +All looks the same to me +d like to see this configuration added to with some documentation also it should be documented in the book +Thanks Arjen I appreciate Your help. +Its more an integration issueIve opened up instead. +Thanks Sriram +Manual testing verifies files are packages correctly +javadoc generates without any errors and is added to the openejbrelease goal checks out the website module generate the docs and adds it to the distribution. +I for got to add that I did not change any tests yet +I still dont know what the special conditions are for it to get into this state since the simple case doesnt fail but it sure doesnt seem to be a MINA issue +It is configurable actually I added that a while back +No this was a description created using a previous version +Technically do not need clean every time mostly you only need it when changing the hadoop version or updating one of the libs +Since this stacktrace has in it my guess is that somehow it got confused by the renamed app +what if you use cobertura from SVN +I hope it will fix that bug +Thus attaching the v patch along with the javadoc HTMLRan all the Ram +Issue may already be resolved may be invalid or may never be fixed +Im not happy with its location being in the source tree and not in a build directoryideally we would put this cache in the build directory +is not the same issue that is related to multiple deployments sharing classes with one of them being undeployedIt looks like a jbpm job is executing a task using a classloader that has been undeployed so this will be some sort of race. +I believe this is not an issue +Any part of the distribution that cannot be created using assembly plugin will need to be created using a custom plugin or by calling antrun. +will then be modified to query the completed applicationsattemptscontainers from the AHSs implementation instead +Is it evented via a reactor pattern or similar or is it thread based? I guess I trying to get a feel for how many web sockets could be supported simultaneously? +has the patches +Yes please create a new issue with all the required steps to reproduce the issue. +Does something less with the SPI for JAXP in the bundleThis is definitely a parser bundle issue in the osgi container +All unit and integration tests passFor the record this patch removed a total of lines. +Let me rerun with a clean checkout +The reload method was added just for debugging to force a reload of the management objects even if there had not been any change to the profile +for the patch +merged to x close for release +HiI had the same problem during my development on my emma plugin +Im not sure if we should add jars to the dependency to avoid these warnings +Please netuibutton can now be used to render an html button tag using rendertrue attribute +Committed revision . +Ive checked the way those PDU are created and I think that the best place to merge the chunks could be in the encoder +Oh sorry I didnt notice you were not the initial reporterIn that case Im going to create a separate issueGood to know that the generated LDIF was perfectly importedIt should not be too complicated to combine the two actions and link it to the Execute Online buttonThe ETA for is still unclear +If we want to add features to wecan either this issue to address those or create new issues +Ive been a bit backlogged with the holidays and all so I havent got to cutting the first RC +The examples and your input really made the difference on this one +Castor is transactional meaning that once a transactionis commit the change is achieve it Castor always write back all changes to the database during the transaction prepare +Here is the patch for the described functionality +Ideally it would be set to default to and still be overridable via command line but I am told this isnt possible with the current way that configurations work +Assigning Tatsuya at his request +A context menu is exactly supposed to show actions related to the selection +option for language files only should be options are all old Resolved issues +This way doesnt involve too much code change to pass the effective user from client down to the RPC layer +Jesse Can you make a combined patch +I will push it to my own fork this evening and let you know +Yongming You have a patch for this? If so please attach for review and commit +Jean and TedMy mistake with typo +contains the patch for this JIRA too +I sped up three of the unit test that were using VFS and HDFS +This can be considered resolved +Looks like your project is misconfigured +I have just committed this +we can modify input field name in wizard to make it less confusing +Thank you for your contributionIve committed a modified version of your patch into SVN trunk and added a junit testPlease verify that its working as expected. +Heres patch for +Thanks Erik! +Yes it sucks +Added the descriptor file in Rev +Nicolas Committed your +You can go ahead and close this error ticket and I apologize for not catching the actual problem before opening it +Hi ArminAttaching the zip file containing the sources for the tool +I have this exact same problem with FOP and Batik +There is no update on these items and are in open state if they are not ready for please move them out to Future +moved common type to runtime type and runtime type to desired type conversions above the connector level into connector work item +However from a JMS compliance point of view this wont affect us +reopening to assign component +Running tests now +The patch containing the JSON protocol implementation the tutorial which demonstrates its use +Thanks Daniel! +A next step would be to make it easier for people to use XML which has far fewer invalid characters +but rather hard coded in the plan andor java code dependning upon implementation +okay +I think you are reading the text wrong in the spec +Rob should we then move this issue off +The quickstart demonstrates the described issue +So are you saying that DBRR should be a class? I dont have strong opinions about this +It also seems to be necessary to bind to logically since its really just per segment metadataYeah I was going to do this in the realtime branch because I dont think we can get working correctly without having private +I tryed create +Mattpatch is commited to CVS HEADYou can try either from source or get the updated jars from the ivyrep at opensymphony. +Jun thanks for +Sorry guys +Removed xfer info and useless semicolons +DB query split in for increased performances +I think its very modular and extensible +I tried by deploying a service but did not get exception +Please provide more info on desired keyboard navigation handling of disabled nodes +Also the LCE uses in most places instead of the used by the DCE + +Patch committed to trunk. +If not try cleaning local repo and try again +Part of a bulk update all resolutions changed to done please review history to original resolution type +I was just going to hand write the native stuff into the template and upload my old native jars into maven unless this is fixed the meantime I applied some of the patches that were posted here earlier but I think most of them stopped working under maven +Added in trader example to reproduce simplified behaviour and allow investigation +This is not a blocker for BRMS moving to future +The reason appears to be that if strutstomcat is just returning the content ofthe landing jsp and thus the browser might not aware of the URL it is displayingThus landing on a target anchor only works if the application always does aredirect and then the browser learns where to go including the target anchorWhen doing that it also does work Please document thisThe problem that a bogusaa needs to be put between the action the thetarget per comment this only works with Firefox not with MSIE MSIE still opensthe correct page but doesnt jump to the target anchor +The has a set that is unused and can be removed now +I dont see anything here that points to a issueNone of the error logging here is coming from +Note that in it is noted that this particular randomness causes problems in the unit tests +When doing that it occurred to me that I should probably remove the special handling of SYNCCTL in my sanity ASSERT at the bottom of the +How did you configure to use +A proposal add a strict mode for counter checkingBasically we want to have an error if a counter was observed but not expected +Filed for the dots vs dash inconsistency seems like it affects a number of keys. +I will need to think about this for awhle +Commit away +The findbugs warning is spurious +bq +Any ideas from folks more versed in Rules etc +Well Im glad you askedI made it work on trtreeTable the javascript code needed a bit of tweaking +In this case the user can simply reupload the file +Looks like its mostly import statements that were causing merge conflicts +For now youll have to create a custom ngram tokenizer factor with the desired settings +Periodically flushing the memstore seems like a good feature to me +Thanks patch looks good to me what do you think about making the checkpoint interval methods package visibleThey are not documented and dont have a real usecase currently +the connection succeeds and is in the default schema but does not wait for the real schema creation by the other thread +Sudhir you can try +Hello JuergenI completely agree with your explanation given but my scenario is slightly differentI would expect the behaviour like described by you if I had nested transactionsBut what I have is a BO layer with declarative transactions which are bound to any call BUT with propagationREQUIRED +I just mentioned that as part of some related discussion +Jacques thanks for committing it +Im not sure what brought out the issue in my case but its worth noting before applying this patch +Basically we would use fixed package names for example +Ahh I see so the PG layer is failing to convert that value to a timestamp +Let me know when it has been applied to master and Ill to +Any other updates on processors can be done in separate issues. +Sorry for being unclear. +I really have problems convincing my colleagues to switch to Maven when such a bug is not fixed for month +I didnt see any tests for so therefore didnt add any for this +This is because we read the file in with encoding and then send it to the parser with the systems default encoding +Apparently Fop fixes some bugs but breaks other things +I forgot to say next time you provide patch could you generate it as a diff not plain java files please +Do you want it backing up anywhere else also for easy access +This patch includes the new and configuration +Updated as recommended in the description and tested. +Apologies +It also compiles the complex pattern only once. +I see the code for the timeout and thus the differences +Manually tested! +Ill fix this properly rest you are right but you have also done some fiddling which destroyed the failover mechanismIt worked in the old version before your changes so i reverted your changes to fix it +Sample WSDL with ant files First level WSDL does not define the service which is in imported WSDL +Tried to build a admin war based on Building your own applications section in user guide +However the package may configure a modprobe file somewhere that automatically loads itFang I replied to your other email about agent logging Move etccloudstackagent toetccloudstackagent +Maybe in the future for any administrationI first of all interested in the presence of engine API features a list of potential initiators of processes +to trunk and +attaching a testcase to test the fail scenario while loading a erroneous resource filePlease review and apply the patch from stanboltrunkentityhubindexing with revisions r and r +Pushed the fix and a test +No need to keep this one here. +This patch will not work without because hdfs does not currently support files without blocks +I think this might be corrected already in +I will add some docs can I get a on this +Nick please this going to make GA? If not well need to mark it as a known issue in the release to Martin to verify fixed +default use SUN version of JAXB and it took as default encoding but sun default stream api took machine level encoding +excellent job John. +The messaging server was started with a Gb heap +IIRC there was a couple of commented test parameter let me investigate +The attached version returns a tuple of instead +Need this to work to use XDT for writing X code +Would really be nice to build the reports together with the rest of the code +Thanks Avery committing. +Fixed at r added reg test. +In the AS environment the caches are treated as services not just the cache container +Nothing in the logs of interest and not reproducible +Oh well + +A last minor small petite issue on the measure filter the modal window does not scroll when theres an error and so the Share checkbox disappears +Fixing the URL made all the jars we deploy there visible to the public +on the patch +Ive already fixed this yesterday in the course of my own decorator tests to be committed in a all of this is only as far as I see and we only require in Spring +Rolling on the actual size with compression turned on could make system behave differently for good or worse Im not sure which +This problem appears to be introduced by +Thanks Todd +I think we ignored in our recent performance fixes +I just filed to track this problem +Anoop +Caused by the compiler bug described in + +r r +Does it need an RPC version bump? This changes the wire format of a Get right +This problem has been resolved earlier +Ive updated my issue subject +There is obviously a risk that I stuffed up the backward compatibility for old options but hopefully not +I still dont agree with the change in ordering +Add a new node type and corresponding componentscontent +Whats the state of this item? I suspect we might want to close this major task and open additional Jiras if there are still problems that need resolving +Vladimir Alexey thanks a lot for the fixIt actually fixes the standalone test but unfortunately doesnt fix the original issue in Geronimo +see +The intent is that will be on the local network or otherwise under the control of a developer and so these fixture configuration files will not be portable to other developers +It makes the build process much too brittle if HDFS depends on and depends on HDFS +Folks can track using twitter google search and whatnot today. +I suggest we set it as normal priority but do not close it till someone give a better solution than just bypassing it. +Applied the patch to trunk in abea and to branch in abfdThanks. +Id like to see it in detail in weekend if I have time +Archive commons jar needed for reading arc files +removes from the exclude list as well as which also passes successfully +You have two options +Juan Nope I fixed this in the trunk +war file + +We need some kind of if there +Yes as Chris said this is not a NN call +etc but as the application grew and some features were stable and other new features were being rapidly added the organization of the code would not scale +Sample data from user was never close all issues in resolved state that have not received further comments in the last month +Here is the full HTTP response headerHTTP Proxy Authentication Required NEGOTIATE NTLM Pragma texthtml charset close PathConnection close Of course I understand NTLM is not supported by AxisC core transport however it should not crash and report an error +Im not sure why there is a jira +Not sure why this passed when I firs submitted the patch but I have a fix that adds the assembly to ensure the type is properly created +Whoops +Excuse the use of tabs instead of ugly spaces +Looks like there is a bug in in version +Bulk closing of issues that are old and have been released or marked invalid. +could you give more detail on your web server? And maybe also attach a debug log +push out to +I believe this should be applicable in have you tried and had a problemBTW we decided in the latest release to instead remove XMLRPC support as we now have a complete replacement in REST +Hi MatthewIm having a hard time understanding this patchCould you provide a high level overview of what its doing? Many files are either removed entirely or renamedThanks I thought I had updated this earlierThe meat of the patch consists of the updates to Constants and +Would address your case as well? It sounds like it will +If more information becomes available please feel free to add a comment and Ill be happy to take another look +Thanks Johan +Lin Sun can you close if you agree please. +Yep +cli interoperation with CQL is not supported see if a column family created via cql client with one composite column of boolean type Its metadata will be of or? +Actually I am new to this PDFBOX +Do you have the same values +Also made members and type public so that they may be used in testing scenarios as it was discussed in this issue earlier +This way the two fields will always be kept in synch +Reopening to accommodate corrections +This was not there previously +I have applied some more fixes to the tool which seem to work nicely on Windows. +updated patch to work with trunknote this does not do anything yet but points towards a path that actually works +Thats how the spec is written +Assuming that Ive understood the problem properly then this appears to be fixedin CLI +Yes it was! After deleting the dev and rebuilt it successfully started on Apache Felix +However we need to that SELECT for UPDATE actions +I should be more explicit about the bounds Ill fix up the docs +It would have been good to flag the tests that do not have timeout +Reschedule for issues to the next CR +Patch can be committed to CVS I hope someone will do thatGreetingsJelle +can be considered as +Ive added an error message to the web page if there is an IO error +I will file jira to fix the other part +Ah! So the point is to concatenate the rows themselves +AAR file +It resolved b instead of b +Attaching patch with the fix described in the previous comment +The correct parameter for the client bootstrap is keepAlive rather than +trunk +DoneGoing to deploy for testing +Ok as Stefan has double checked the fix I mark it as resolvedThanks Stefan you did a perfect job testing it and double checking the fix +Please commit this along with the patch uploaded to JIRA +Thanks +I have to setup a distributed mode for myself then +I applied just the test to TRUNK +s for both have to go into JBPAPP because of workflow issues in verified +I think I can help but in any event I might need to thrash my own vision out on source to accomplish what we are trying to do but I dont just wanna maintain our version of so yea whatever i do in the meantime i will sure bring it forward +bq +not done as Creation and Issue Comments go to both the dev list and the commits listAll other Jira Notifications go to the commits listNo clue where you got the idea most apache projects do this because they dontYou need to add to the dev allow list a moderator can do this with to a moderation mail +implemented via a user patch +Great job! Just a few minor corrections API and Core should be and respectively +One last update to include fix to request as well +I can take care of this updates next weekCheers Sergey! Sounds great thanksAndriy +So I am not clear how this should be integrated with Woden whether I have to introduce a new module like maven converter plugin or is there any other place to include this one? Please provide feedback on this +That value implies the query executed as mentioned in SYSXPLAINSTATEMENTS table +Im setting version directly in the code yet its not generated in the output +Opened to trunk the issue. +Have you set to true? If you have this may be the same issue asIm able to reproduce this error on the development trunk with a fresh unencrypted database when I have enabled logging of query plans +Fixed in Rev +If you think this is a problem attach some testcasesource that shows it +Ill close it after doing few more tests +Done +So maybe its better to just throw an exception when validation is enabled and schema not specified +Thanks AndrewAdded the dependency to and committedRoger +Hmm +on commit after addressing Jesse comments +I can put a sample code on would that be fine +committed. +Also for users who use or the likes for unit tests and will get bitten by the issue of the issue we did in camel tests +This issue only happens at different Sinks +Hi againBtw I havent added chunking in the above example +Im now initializing the status page +Thanks Lars +I agree with Tobias not having a trunk for contrib is not a good solutionI still prefer the first solution moving the core module to a subdirectory +Thanks Myrna for offering to work on the harness changes +When I create a new seam action using our latest bits I still see POJO class Class name is used in WAR Seam projectsLocal interface name is used in EAR Seam projectsSee the screenshots +The correct class is now used +This patch modifies srcc and srcc files in order to not override user values of CFLAGS and CXXFLAGS environment variables like it is expected and defined by designSome C and C options are still in +verified in +Filed to address +It seems that the checkboxlist does not use the checkbox tag and so the fix to the checkbox tag didnt carry over +Its a trivial change but its an enhancement Ive avoided eagerly performing until a project considers it a worthwhile feature +Patch +As far as I know zip does not contain the flags + +Once Hudson approves +I think Ben wants it to be removed +Scheduled to a version that has not already been released. +Ill create a JIRA issue to create more tests to test out the block of code +committed patch in fecc +And SOAP messages are valid XML stuffYou can probably try using SOAP attachments or base encoding etc to send the binary data +Should make this a. +patch is the wiki police James remember to update the wiki for the mock component if you have added new options etc +A proper error message saying that the package name is invalid is given as belowInvalid package name +Since usually we dont commit unless all unit tests pass getting this corrected will unblock other patches to get committed +I am wondering whether it is a issue when using the policy attached with the generated client +Use reference counting for a zero copy solution +Thanks! great product great support +our testcases are still passing can you please do testcase for ll try its pretty woven in the application +We would like to get the benefit of though without having to rock the whole cart +The issue is that will force sorting if it does not find a column reference in the s result columns expression tree cf +Updated patch +Given this I think we should include these projects that we dont want to break in the limited private set +This is superseded by +Moving to per s been a year and no one else seems to have run into the problem +we already had some discussion for this to also let validatiors be behavioursI guess for we need to make a discussion for this +This should make my cache patch much more efficient + +Hi Hairong you may revert and +Logged In YES useridThere is no zip file with the attached to this bugonly the +includes updates of generational and some other functionalities +sounds promising but still not ready for forge integrationjust my opinionCheers +lgtm +Is this facility out in the fb branch +Can you check whether it is a dupe of +It is still substantially identical to the last couple that have been posted so reviewing those is valid for anyone who cares +This will only ensure that the list of input has unique values as we do currently on the output with the Serialization class to have a constructor that provides the confcommitted in copy convenience methods to call copy convenience methods on all setadd inputoutput methodscommitted in testscommitted in +fixed bugs +If we make that file only readable by the hbase user it should be fine I think since he has access to the credentials anyway +Were in the same boat as Daan our searchingsorting code is also abstracted away and does not know the types +error message is fixed +Thanks we will try to keep our use down. +Yes I had done some of changes in the other patch +Michi committed to trunk +This is OK for the XML exports we generate but not necessarily for things like the DOM to SAX conversions used by the webdav component +RegardsHeshan +So you can freely type another letters +VerifiedThanks a lot! +Attached the next patch regarding avoidance of the raw types by the generified classes number The patch also includes many checkstyle violation on the trunk +Uploaded the desired log messages +Its still a problem on my machine for what thats worth I have updated to head and selected Clean and then Build All in Eclipse to see if that was the problem but Im still unable to selectadd features normally +Duplicates +Isnt that the standard behavior of all jobs? Maybe we should a generic option that makes a job delete the temp dir if it already exists +Attaching new patch +Hi Mike yes please! I have not had time to work on it +Got windows server installed and AD configured +Patches welcome +I do not think the metadata caching is surviving between restarts so they load again and this exception occurs some times +Sending srcmainjavaorgapachecamelcomponentfileremoteSending srcmainjavaorgapachecamelcomponentfileremoteSending srctestjavaorgapachecamelcomponentfileremoteTransmitting file dataCommitted revision Also should mention that Dave Stanley came up with the fix for this +Sorry for the delay this bug was submmitted as a hibernate plugin bug so I didnt notice this bug until today +Id also want to turn it on and off an a per xstream instance basis so if I use xstream for message transport Id want all the fields but my loggingerror xstream instance would have some ignore fields turned I file a separate feature request? I can think of a few other related features +Fixed SOE support in JNI functions in stress condition of insufficient stack sizeFollowing changes was addedException callback sets exception if stack size is insufficientException clearing functions which are called from JNI restore stack protection +Is it possible to use a more accurate log name here +Indeed wasnt this Findbugs annotation usable with the latest version of Findbugs and without having to configure anythingThanks +what exact steps are for reproducing this issue +The changes are in branch +A better solution is probably to enhance the call to in the dominant Map + config file version selection page +I think this is a dupe of +Client is created with the classes Core has not got themWe then create Client without the classes and Core with the classesWhat happens if an exisiting Client Core is upgraded to Client CoreWont it cause a problem to have two copies of the classesOr maybe we can just document that certain combinations of versions are incompatible +Another idea is to set the INVALIDOP code to be and the ADD opcode to be a positive number +added notes implemented +Thanks to Oliver Bayer for provide these hints. +Tested and added a short configuration snippet to for +I am surprised this hasnt been addressedShould be done automatically and not by some web admin console etcLater JMX monitoring could be added where we could control it from there when is mandatory +French error message should be reviewed by a native French speaker +So Im looking for the details +Is this fixed? +It adds about MB to the download as it ends up duplicating most of the jars +Enhancements for evenodd support will be addressed in a release of Struts after +But definitely for something thats consumable by programs and wont need to be actively upgraded when we add or remove fields +Any comments on this +a patch for this issue is attached to the corrections that correspond to from the patch attached to +and again throwing back a server exception is something that wfs forces us to do +The is used to represent a response entity of a generic typeFor example if there is a provider that accepts List of a specific type the will be the solutionIn your case the doesnt expect to handle any specific generic types but only supported raw types +JAXRPC tools issues may be fixed by community contribution +But I see that my does no good step in that direction +Maybe a weekly job given the machine resources we have +Then youll get a very wide select +The copy does not even have tests for itselfBut I cant imagine sharing code if we begin trying to filter or resolve symlinks given two implementations one that handles it and the other that doesnt +See for sample will definitely be enhanced between now and GA +BavoAny chance that you can use instead of? Or you can download an eclipse version directly from +bq +Everything in should be reverted after the the core is complete +thanks uwe +I tried converting the images that were odd to jpg and they were rendered properly +Eventually we shall switch python script and no worry about sync +Created to follow up +is usually we face and it could handle most of our cases +The process would just restart with its last know sequence number +thats nice since you retain typesWould the rowkey and columnkey be automatically included in the resultIs that a proposal for CQL +Targeting to and upgrading priority to critical +Thanks a lot for this contributionI applied the patches for and +I fought the good fight but the IBM people appear to be winning so far +A fix is coming your way soon. +Targeting revision . +Change the option Treat wchart as Type to noMake sure you do this for all of the targets you want to build +fixed for the observed problems added test closing for now +I will commit this tomorrow if nobody objects +Wouldnt it be same as hard coding this role for SYSADMIN model then if user has the role then have access otherwise not? So we only need to tell the user what that built in role name is +In particular these affect developers launching from Eclipse for testing with paths +I agree this does not look worthwhile to backport. +There are a couple of packages in those two jars that might need to be added but I believe those are only implementation class that would not need to exported by the system bundle +Ive seen some solutions based on relational database clusters in activeactive or activepassive mode support DDL sync and DML high availability +Can you try the latest snapshot of the plugin? Either build from source or use the codehaus snapshots repository +Its up but someone needs to actually publish the content . +They should be part of the application using the opencmis library never in the library itself +So maybe access to the Policy should not be pluggable. +You can see the patch attachedIt breaks the long property value support for multiline but it is not a stopper for us as it shouldnt be for many people also +Bumping this to and repurposing a bit +but no bugfix release has yet taken placeI will resolve as a duplicate of ok +JasonI added a unit test using Camel Error Handler redeliver and it provides equivalent better capabilities than your max code so Im reverting your code for now +Added a basic fix for this in r This isnt a complete fix for whats described in the OASIS specs as the design of the code doesnt really suport the lifecycle as as the spec describes for that i think well have to do it in the code all this fix does is try to ensure destroy is called if an exception occurs during init processing +This means that the user has to be aware of whether she is using cmd or powershell and depending on the shell use or not +Surprise +If its the issue we discussed about earlier then I think I cannot do anything about it and youll have to certify with a different Tomcat build for the time being +All major features fixedMoving following good to have features to a new ticket support for pygment based colored output Fix fuzzy param completion to a much better implementation Scope for unicode and zshbash this record since is now released. +Thanks MayankTx for the reviews Zhijie! +That said integrating with Spring as the backend might still be an option just to simplify things +Yes I know they are separate and it will be in a separate contract what I am saying is that it still belongs in the footer panel becuase that is where you want it to appear in the footerlike it is with skins +Can you please take a look at this methodThe data path is set correctly but it is not usedWhats wrong +I encountered it several times on windows +Closing. +Currently my line of thought is that we should remove this example altogether seeing as it is based on such a technique +Complex benchmark test was improved +Attached sample code to trigger the behaviour +Marking as blocker for easier tracking +Thanks. +a weird decoding exceptionI think decoding exception would have similar effect as the PBUF magicWe can catch the exception and provide more meaningful message +I wonder if you instantiate that Comparator before inI was thinking of making the static assignment in fixedfiletrailer an instantiable class +We can further optimize the later +If it starts to seem in the way we should go ahead and jump on this task +Not sure anyone would bother doing this +Other than that it would depend on the lock factory impl +needs rebase with recent checkins +Been getting Gora sorted in Eclipse and your right Chris this is a real pain indeed +Attached patch makes the following change all confirmation prompts accept y or yes as a positive response n or no act as negative responses anything else will repeat the prompt until a valid response is were editing this code the fact that it cant be forced on the command line and the fact that it treats an empty directory as one that already has a namenode image on it are also usability issues if youre scripting around namenode formatting +The ICU is choosing wchart for its data type but is using unsigned short +xsd +Im not sure if thats the right way to go how often are those going to be called? If regularly profileoverloading seems the right way to go over having to specify a bunch of goals even with helper scripts +Thus resolving. +How about getting rid of the commas an colon and come up with something a bit more realistic and short +But it always fails in the activitiformKey +Juergeni simply stated that there should be tests that check the hierarchy in a queued components oninitialize method +user puts text caret in front of existing text on a given line hits Tab expecting text to be moved to appropriate indentation but the right thing doesnt happen +Introduces http and http configuration parameters is created per core only if timeout is specified +Please push the patch attached and set the issue status to verified +I am not sure what to do here as it is working as intended maybe just document it +has been marked as a duplicate of this bug +we would need to create one +Thanks Zheng ShaoI check the failed test +When theres contentavailable let us know and well switch over the live site +It also explains PIGCLASSPATHs other uses +patch adds a new test file and directory +I committed the patch +Fixed in rThe ID v header parsing now tries to do its best if not all the data for the header you NickJust looked at the diff to learn how you fixed it and it was nice to see that you have also wrote a test caseBuilt the latest version from svn and it works now java jar v m srcpluginsample Author The White Stripes audiompegresourceName title Girl you have no faith in medicinexmpDMalbum ElephantxmpDMartist The White StripesxmpDMcomposer nullxmpDMgenre nullxmpDMlogComment engxmpDMreleaseDate xmpDMtrackNumber +I do not see why that would break if the bundle location was set to a URL +The only problem would come if you had failures on an attempt and then had a JT reset +Have we tried just changing binsh to binbash +Also Im fairly new to Spring Id like to know if there exists a release scheduleroadmap for Spring Webflow where I could see when is due +Max was talking about differences in how java on Max OSX has migrated +If using JCR then I wonder why the data store isnt used in the +Yeah Im running the full mvn install now will only commit when that passes +committed to trunk. +I think were on the same page now thanks. +I will attach file soon +is only called in one thread rightyoure right that should be finebq +Added related recommended practice in documentation +assinging to mark to assess if this is a blocker for or should be this is part of a commit of a bunch of issues Im going to do shortly +Patch is applied +So although it would be nice to have there be a more general that has the check in there as well as checking in other places like the paths in the workflow and coordinator configurations the check is not currently done in oozie +Thanks David! +RHEL builds are done please update the native zips in for when hudson builds are done thanks +Any further questions comments objections before we commit +The assumption is that the properties are only used during initialization and are passed into other components as simple constructor args +OK +Yeah I felt for debugging the chances of Turkish upper case was less of an issueUploading a patxh to make the behavior as +I set the intial retry interval to be milliseconds and the max number of retries to be +Andrea if you could look it over for sanity that would be great +Thats what i too feel +I mistakenly tried to use dll from net and not folder +Ill try testing the updated patch over the next few days and report back +Confirmed that this issue is fixed in EAP. +This is similar to which would not normally be subclassedThe problem with the is a duplicate ofI recommend this bug be closed at the same time is fixedThis issue is superceded by. +Sure let me do that change and upload a new patch with that change +samples now have new package structure Dave please take a look whether you are happy with it and the issue can be closed +See for followup. +Thanks all! +I updated the page and I will verify later today that other instructions are old resolved issues +Please use target for tmp dir +commited patch in Closing this issue any additional missing cases will be handled via new issues for specific missing methods in the class library +So can we rename it from leaders to allBrokersSame for and to leaders at one another place +By the way I have run local test earlier and all the unit tests have passed +Paul is there more work to do for CR +It turns out this is actually used +Why did that I read those as comments! Whats happening to me +In the file line breakpoint is copied to the local variable to save it from being modified when breakpoints list is unlocked before calling the callback +This functionality should also be introduced in the new class +We are now testing using the as well so if something like this is missing we will also see it +An easy to understand description of this new architecturefunctionality would be really nice. +if notthen FOP cant output correctly pdf when content include bold and italic +Looks like this never worked before at least down to I this is not a regression but I agree that would be nice to haveworking +You have a patch +any reason why we chose such a looong default folder name for the content folder? why not just src or content? look at how other editors works in eclipse on this +It seems that you got it +Thats a good point +i tested this yesterday +Jim version is needed to control what AS WTP Adapter will be created for particular server so it doesnt matter do you have EAP +T is busy interacting with the user and is NOT stalled inside a JDBC call +Committed +Cancelling and resubmitting as the tests Submit again wouldnt run the testsThis is because Hadoop QA remembers the attachment Id of the patch for which test suite was executedPlease attach the same copy of patch for TRUNK again +t see it trunk +batch transition to closed remaining resolved bugs +Trunk patch that goes on top of v +Logged In YES useridYou can always post a RFE for that +But it also says is a list of all implemented by the framework so I do think it should be too +Please reopen if it does not get resolved by analyzing the still persists even after launching a new project analysis and clearing the cache +The attached class shows you how to use the trick I hope this helpsI suggest you to rename the class in IMHO sounds more intuitive for users +Or feel free to improve the wiki based documentation +thanks I think when we generating using given WSDL if we create WOM and using that if we create we are going to have one unnecessary step I agree that it keeps a room for supporting both WSDL and WSDL +The SERVICEPROPERTIESDONTEDIT message with id M is similar to the messages we want to add in this issue +Committed to SVN trunk +I really like this feature +Open issue isOPEN ISSUE What name should the interface be given +Updated the patch to work with the current trunk +Bump +Please give couple of more minutes and it will be done. +It wont affect anything since the server does not depend on the jars +but do you know of anyone who actually has run a cluster on? Probably not worth rebuilding RC we can just tell people just kidding for now +Thanks Ravi +Im setting up a with windows to reproduce the issue and fix it +I havent finished running the junit and derbyall suite with it yet junit suite is running right now +This version is a crucial design option that should not only be maintained but promoted to developers +A new event could be chosen as well if needed +This change fixes my problem +I find the same problem in the updated code also on clicking node link +This wouldcertainly often lead to the error my +The latter step may be automated by CI +The biggest reason I wanted to fix this in the controller is that its a very small change the patch only changes a handful of linesHow would you feel about doing the controller fix in and something more principled for trunk +It would be good to have a test case or two to verify that it is though +push out to +It was possible at least mention in the comments that POM is incomplete and appropriate versions should be specifiedNote that experienced Maven users hardly would use your POM and inexperienced will be bewildered +Close. +maven needs to be updated to a released version once available +OK I opened for the nd part of this issueIll commit the first part shortly. +Later the hint finally comes in and resurrects the data +I am not a committer on the maven project +Patch updated against trunk no changes from the last version by Carlos +Roshan for getting this done. +Why do you ask +Thanks! +defer past of binary compressed serialization supports having the two type systems not be identical so the need for this may be diminished. +bq +Yeah its not setting for me either +Attached a patch that changed the cache pooldirective iterating a little so that it uses directly which supports NN failover instead of +Consolidated versions in the parent +If we reach consensus on that definition I can take care of that issue +I am going to convert the remaining groovy files into Java +I just vaguely remember that was already proposed and the resolution being that would be too much work so having the option to manually clear the caches was good enough at least for the admin to be in control +Well adding such information to the manifest should make the issue less likely but the code remains fragile since it could continue to fail when the classes were packaged in various alternative ways +HansSome users prefer to leave turned off for security and privacy reasons +Console needs to use local task service first before a JAAS callback can be JAASJACC based callback implementation is provided and should be used for as default oneMerged into master +blockSize that datanode sees before writing is always + for the patch +After closer analysis this is not a duplicate of but the fix is the one suggested by Carsten which is already included in +Should that back to once someone else confirms it as fixed. +patch Greg itd be great if you could add a test case in for this +lib +Thanks Jacques +BTW I am not sure if proposed patch is the best idea +We have a flag in the with this change forceAttach which on disconnect the client can check on channelDisconnect to decide whether to reconnect or notIf you explicitly need notification at the application level this is a different matter and I think it should be separate to the reconnect logic +Ill take a look but renaming to just disables the plugin since you get a and you have no secondary datasources just the primary one +Attached a small patch that renames the file to and changes the prefix for the main namespace from dna to mode +Does it mean that you need more technical administration features in UI? CI tools that schedule code inspection jobs are not enough +Or make it a large testIll leave the test running in a loop for the rest of the day before I conclude anything +Even this last suggestion gives me pause a little +This problem of compacting really old sstables with newer sstables is mitigated with a restriction that an sstable is taken out of consideration for compaction if it reaches a certain max sstable size +Thanks +Therefore it is up to the module to maintain references to specific data and clean them up if it is not intended to be cleant by the core +Butit is pretty large so attached unit test code that reproduce theproblem +The info is added +Dont currently have a reproducible test case the failure injection might be slightly tough +I notice it does not display when I start with no +Have you tried against a relation rather than aPerhaps it was never working against a +was figuring I might work on this on Thursday at the contributor meetup +Also did you mean to post h +This bug can be ignored if is fixed with the currently proposed patch +One thing when normally the force unassign will be used? What are the use case normally some one will use +No integration tests +OK Ive fixed this in Yecht now +Please fix this issue in the branch too +I will update Java project configurator to take the highest version from plugin and executions config +fixed by closing the Conversation at the end of the Request if it as issues are fixed in +nope +Then he fails over to B +a patch to make getStatus a little more reliable and fix the Serialization problem +Providing updated patch with some refinements +Thanks for your transition resolvedfixed bugs to closedfixed +I independently went through the looks pretty accurate as to what hasnt been committed yet +Committed revision working on backport to revision trunk fixup of an unneccessary change i noticed when working on the revision +Maybe the archiver doc may include a section for appclient modules. +Joram is going to check the userguide reference as part of +Shouldnt have resolved just set the Fix version to AR +Stack requested a provisional patch for +A new issue as Jochen requested +The next new tag will solve this problem +We need to decrease the number of output messages in case of failures during class loading +it does have the bits +I am going to leave the statement as is until we actually figure out what it should +To create a new project archive I have to edit +I have a fully Groovy project in the same workspace as a Java project that contains Groovy sources that Id like to be used as +New files have been uploaded +Patch looks fine to me +I should not have used the words Use the forums for help above +But given that youve attached the as a correct example I think its finally dawning on me that youre not talking about the schema youre talking about the example XML that Enunciate generates +Yes +Rebased patch for current branch +the fix +Ill add unit tests tomorrow and close out when they pass +unpack and add +Youre right that warning looks like it was missed +Its been over minutes since your reply and now I cant commit at all it hangs without any error or timeout +In that case fetching only a subset of topics should help +Thanks Joe +Can you take a look? Examples like You can use ant test Dtestcase Dqfile to reproduce. +Fixed by Greg. +This issue is that would like the ability to operate on rows and records in the +I thought it did not matter +Committed to trunk. + +Im sorry its a mistake +Attaching avg and to CP +Create standalone Seam Web Project createnewseam change the name on the screen no capital letters createnewseam change the name on the screen no capital letters createnewseam change the name on the screen no capital letters createnewseam space is not recommended in project names seamwizards change the name on the screen no capital letters seampreferences change the name on the screen no capital letters seampreferences change the name on the screen no capital letters seampreferences change the name on the screen no capital letters testng change the name on the screen no capital letters +I was able to verify that behave in exactly the same way though so this is not related to per se +Im pretty sure that each module has its own instance and there can be multiple modules in a servlet +Just noticed the concat function I made is for strings not any atomic type +Brilliant many thanks to all! I just fixed a pitiful bug at JacquesI checked at the demo site the State dropdown disappears as the page loads and no result of selecting any country too i think dependent select is not working here or may be something elseAnkit Jain +Only about of the message make it to the broker that was brought back up after running the test again +Sun will fix for the jdk release. +Thank you for your comment CheolsooPrashant said that should be piggybank but I think its natural that is a builtin UDFCan I add that as builtin UDF +Hey Du +Thanks Stephen. +Fixed in trunk now wtih wiki documentation at Error Handler and Dead Letter ChannelSee also the unit tests for details etc +FYI the forum thread notes that this is against +The test case uses the asynch capabilities of the to easily ramp up the traffic +Only if Foo had no such propoerty it would work +New patch fixes this and separates them so it is impossible to use fields as methods and the other way around +Attaching patch test that treats score like score for distributed since thats what was done in the past with +I cant quite follow the patch +Can you provide some more details a fix is more then welcome +Thanks Ben! +Not to mention the option on the war plugin to spit out a separate classified file which can then be consumed by shade in another project. +I dont necessarily care about being able to deploy a file with a space but I think the error message could be improved to make the cause more obvious +Cool +This should have been intermediate complexity it doesnt look like theres any way for me to change it now +No longer a requirement for the platform so am moving this out for may revisit this before if we have time. +Committed new to resolved issues. +this issue is resolved. +I created a test case and didnt reproduce the it has been fixed since version +For the attached that differs only by including the attributes inline instead of in a group the generated stubs are complete +Most id tools Ive tried silently fail and claim there are no tagsIve fixed Tika in r and itll now not break even if the tags are likely to be largely gibberish since theyre corrupted +Very happy to see interest in this workWe are planning on doing the initial in a week or so +The output is still built sequentially. +No response assuming ok. +My bad sorry guys. +You are forcing the junit provider and using junit? +Looks good applied thanks! +This will give yet another boost on checkpoint performance +But in my opinion this would be a very helpful feature +believe Bob fixed this +This looks rather similar to the restriction identified in which was summarily closed as by Justin and is slated to be fixed by the resourcepublishing should this be reassigned to Justin at the least to make sure the resourcepublishing split supports this use caseUnless of course I misunderstand the problem +Server Server +The cost of gettimeofday is huge when put into the inner loop like that +Thanks Michael for the update +Applied patch to CVS. +Weird +revision +H +update should just update the counters by using the enum values explicitly +I think this has also been addressed +Rather not +First patch that fix orderentry is a request a not a view view is not existing but it isCan be committed +Interesting +This is a duplicate of which is now resolved. +You can use logic similar to the attached patch to print a schema +Look like it is a dead end at the some one advise a quick fix for itRegards Hai were currently working on javancss but I cant give you any dates of availability +I really do not want to rely on s perl step would be to set up a development environment and the initial CSS stuff so that one of the interns can jump on these or so pages of content +but major enough for me to eliminate the feature from the app +I dont have the same behavior +Not likely +Every client needs to do the same tracking +Or are you planning on doing something like dividing by the min allocation from the config to continue faking +The one caveat to that is that if youre not saving restart information the step isnt saved at each commit point which saves you a table write per commit +I forgot that the JDBC store still does its own thing for the metadata +Im going to close this out as I have been unable to reproduce the problem +Can also be used to verify that demo content is gone from a production systemChecking the status of internal requests is almost as good as a check as that can check that content is present and that the components required to serve that content are active +I find the naming confusing where setProperty adds a value +Same error is in mentioned in the mailing list it looks like this is caused by a Groovy bugBut it is wrong +Please go ahead and commit it I was wrong about needing to fix Gump that just needs the patch to remove the copy of the files in lib +One big difference between and is that the EC driver was completely be pushed and resolved +Please make this work with ajrepeat or provide some kind of adaptor tag. +The RMI test case failure issue. +We can also look at adding detection logic like the other jdbc translators so that the user doesnt have to explicitly set to trim +revision . +Im just trying out the localisation features for the first time though so I may well have missed something +Thanks for your help! +Will try to get to that this evening or tomorrow then Ill upload a new patchWill also be adding some more rigorous tests now that I know more about how I can test this more thoroughly +tried to apply the results as a patch +do you have any sample project +I looped times locally with patch v and they all passed +Provide a small test that shows demonstrates the lazy ability of the two modified XML elements +I didnt look at runtime partitioning code closely but I think itll need to be optimizedIf a lot of data exchanges among peers are expected they should be transferred through +Can someone please confirm if the draft Release Notes Text is accurate +You want to do logging in the child process so the api and core jars need to be in the classpath of the child processLets start by finding out what the classpath of your child process is +Frankly it is an unviable option and leads to umbrella projectsThe point that seasoned Apache Members like Chris are making is the same do not have umbrella projects +Michael Im going to test it on webOS I currently dont have access to a webOS device +Yeah exactly +not clear what Rick was asking for +Feel free to jump in if you have thoughts +Further changed the serialization format of the deprecatedFixing these were originally planned in this patch itself +Any further feedback or should we close this issue +Thanks for the report and the logIt helped me found the problemI was not using the correct in the configuration folderI generated a new version and updated the download with itFeel free to test it when you have a moment +HiActually forgot I was going to be OOO last week +I really have no idea why were building everything up lazily it doesnt look like it was this way in Seam +So yes what youre looking for is supported +Verified Thanks Tim +We also want to make sure that hint TTL is less than gc grace period +can be used as a write barrier implementation for JET so that a generational GC can work +Yeah this is my fault +JeremyBTW if you are interested by this issue you may have a look at JeremyThanks for comment +Built with maven as I dont have m up but It hink it should work fine +Integrated to TRUNK. +Just to clarify I meant the WSRP schema elements for your new property which isnt in the test WARWSDL you attached earlier +to mark as private. +The names of the modified files have been changed to comply with the webhelp license and include ASL +I just committed this +Sergey mail tags might be invisible if you render them in a browser but that is not what you would use a facelet page with mail tags for +Does it matter in which entity it is defined? Are the names globally uniqueI would think in terms of naming that it should be with the name of the entity that it is both defined in and for which the Repository is parametrized +The last time I ran any of them I wasnt aware of any failures and what Kris has found here is that some UI tests are exercising pieces of core that the core tests arent +But Id obviously rather fix the underlying issue here before your fix obscures it. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +The war show up in inventory +Yeah I bumped ino it before +HelloHere is my patch including test case +Jervis screen shot proves its fixed. +bq +I will reopen the issue if I see the mentioned errors again. +Net do output these by default +Modified library dependencies and locations so that does not require a network connection +bulk defer of featurestaskswishes from to +Please review the patch and let me know the comments +Please review modifications to the +SergeyYes of course please may you attach the jar file with version in its name +Closed after release. +Cant see a way to reopen the bug but I still see the issue in Continuum +Personally I think its nice to have a line for each file installed +All versions of iOS simulator up to as well as iPhone on +We have since I posted this migrated our builds to luntbuild so I cannot look give you any more information it is possible it was because of a misconfiguration on our part +Thanks for the investigation +It makes more sense to keep a config param to enable this behavior upon the users requestSo I reopen this issue again. +Latest Merged patch +I am still running the unit tests and manual tests to be sure that the leak is plugged +Do you think there will be a way to remove old Visits and keep the tables +Yes that is what I believe hes talking about as well +Its a goal of the Maven project as a whole to have people Mavenize their the dependencies in the POM is the point of Maven +Correct since is GPL we cannot use it to generate our +Was already fixed +Changed monitor to print one additional report after the job completes +the patch might be a tiny bit out of date now Ill check in a bit when I get to work +I just committed this +steps and any material to recreate would always be good to have with send abort detection and +Does anyone know if Apple is rejecting apps because of the images not being sized in the Human Interface Guidelines +Use the unreleased version of the plugin +I see a lot of copy and paste code to break the parameters into subsets +Not facing the problem in though. +bq +Confirmed bug only on branchPatch attached to eliminate last comma from hash representation +iterations it not that much +I am fine with your idea of merging the filter chain +trunk thanks Shenghao Fang for the patch! +released! +version disallows empty column names everywhereversion disallows empty column names everywhere but in the system disallows at the parse layer + requires a special configuration to run under HSQL which is unnecessary for other databases +Gary we use certificates coming from tens of around the world so changing the way they are created is unfortunately not an option +I committed the patch +the correct diff +How about merging fix to the branch +They are no good + +Since your tests passed without issues Im for commit +The only place we generate bag schema without tuple inside is +Also the post execution hooks need to be run. +perhaps its not appropriate to implement closeableLots of people seem to have problems with the analysis workflow and I think this adds confusion +In the branch this test is failing about of the time now +OK our docs are located herehttpreposjbossasprojectsdocscommunity assign the commit priviledge of the community documentation repository to me well start translating doc soon +I didnt realize that Jexl had this issue so I appreciate you letting us know about it +I would rather see customizable images than random text and we should do that instead +To +thats where its coming from +Unless you want to contribute a test case to the module codebase or have any other thoughts as to what might be going wrong Im going to have to close this issue and move on to other things +That way we can see how sampling stuff can work with this as well +I can see in your link command you have lxlpgassockets where it should be now lxrtpgassockets +Deactivating the column menus lead to improved performance of the rerendering +Take it home Noah +To me it seems errors are coming due to old lucene jars as like error is coming as cannot find symbol symbol variable TOKENIZED location ckass +And have other maven repos set up as well but that shouldnt make any difference +I agree with you +Ill wait to see if there are any checkstyle warnings etc and commit it if there arent +so youd want to use ts host etc +The service attached runs fine on Linux but fails on Windows and Windows trunk +Also confirmed with greeterrest demo thanks +It just makes futures patches simpler +It turns out to be impossible to modify a filter within a object so that wont work +I will close this as duplicate +There are few more files to be moved out of include +Hi JayIf its not splittable how are you going to gain the benefit of using hadoops infrastructure? The infrastructure would be busy making network to build up the binary file from replicated sources and passing it a single mapper +Added escaping for comment content in both RSS and Atom versions of comment feed +Realized I never changed the method and field names to reflect it no longer takes the app submitter name but rather the client name +But Im not sure I understand why +This is an additional patch for the testXSLT executable +This bug has been fixed in Xalan Java +Patch applied. +This patch is clean and can be committed +batch transition resolvedfixed bugs to closedfixed +fixed +Brian I think this jira needs to be split in two +I have applied the patch and did some updated of my own to get the files to compile +Tests passed cleanly for me +however all the data dir had was one non default workspace with local settings and wms settings and I hit reload getting this +Nick could you please update the patch once more as theres some noise atits top? And Id really appreciate a reply to my last question be it yes or no +Also it does not look like the this config is was ever read by the server +So it should be safe to backout part of the change wo impacting any queries negatively +Hi AlanHere we need to replace key to +I will generify my sampling sort and create a patch for it. +Lukas if this becomes reproducable again please let us know. +Verifiation Close it. +bq +Are and issues really blockers for using newer version? Are there not other features of springide that are useful for spring users? +Unless Martin you can think ofa way of introducing this new required functionality in a meaningful way +Because before sending the message it is serialized properly without these additional NS inside synapse +Provide the sample in war in order for examination +They are listed in the repository +I agree that the link to the README is enough closing +bulk cleanup of Jira versioning +Move cleaning code to new method thats being shared by timebomb and this oneIt has two configuration options to enabledisable and set the minimum pagessecond to set the number of times allowed to drop below the threshold to prevent a few accidental pauses from immediately killing the queueIts tested in a production cluster and seems to work nicely no more long dreadful delays when finalizing a fetchPlease comment on usefulness and implemenation +Its probably OK to leave the actual message as it is since the freezeunfreeze approach is very likely to have been used if the situation occurs +However it might be worth looking into moving the lock acquisition out of and putting it into the application +After some further thought and testing the above issue is not described accurately +I just committed this to trunk. +The patch provided assumes too much about the error and its definitely the wrong place to fix such a problem +Jessicacan you check who is best fit to fix this issuethanks with fine Hence closing this bug +Starting a file doesnt make sense here +Committed to branch with revision +Ive reproduced on another machine +So why would the client abort the connection? The timeout setting in Railo admin for this job is far longer than the time it takes for the abort to occur +Heres a patch does the above renaming +Assuming that being fixed and resolved and as has said anything we can close this one. +Logged In YES useridThanks for the +Ill backport this to if there are no objections +new patch lgtm committing +Pritam NicolasIf no one is working on this issue can i take it up and port it to trunk? Thanks +attach my patch +Would that work +I have temporarily reverted this +A unit test would be much appreciated by the way +We dont want the Namenode to report missing +Hello James this was an error on my part we had two different oozie web applications pointing to the same oozie meta store this is what was causing the issue i think this bug can be closed. +I might also recommend something like See LICENSE and NOTICE for more information +top bottom up down screenshots in accordance with new examples +Attach the patch again for Jenkins +Perhaps we could check to see if removing testDelete changes the code coverage +Thanks for the patch EdwardFYI before committing I changed the version to align w hbase version rather than run a separate hql version number +I will try out some stuff on some older though and see if I can trip somethingThe difference between CCL and JDK concurrency is locking +Only is present. +The container uses the mapping of to when determining the return value of the TESTABLE WITH CURRENT FRAMEWORK SPEC Testabletrue Section the element does not define a element the container must default to checking the element argument against the list of securityrole elements defined in the deployment descriptor of the portlet TESTABLE WITH CURRENT FRAMEWORK are also two Propagation of Security Identity in EJBTM Calls assertions but there is no need to test it now +Fix is in revison of srcorgrollerpresentationFix should appear in release. +I aggree to your setupreopening to attach a file +Updated patch with following changesaAdded javadoc for the methods that adds paths to the b Rename fields DOTXXX to PRENSDOTXXXmvn test Dtest passes +So sorry +Verified in JBDS. +works now +Regarding step will the block report processing wait till the datanode is resolved or will it return an RPC error to the datanode +Personally I feel my code would be more readable if I got rid of the boilerplate this +Patch for smx and smx trunks unset the CDPATH prior to trying to capture cd output +public key DSA bit you please create the account and install the key +charithamarking this invalid as the configuration is invalid +Additionally it changes the javadoc to actually match the behavior of the code +Its a Swedish coordinatsystem based om and i dont realy know if this is normal or not its that gets and for what i can remeber have it always looked like that in when i register i new my data has been writen to the database using FME and it works in older versions of so i dont think its FME that writes it in a stange way i cant exectly tell you which version this odd behavior came it be fixed so geoserver saves the whole value without spaces? +Let me know if you think the comments in the code could be improved +Sorry are you expecting a proxy to override the authentication? The purpose of the proxy is to do just that proxy the contents through to the endpoint +and the test should also do should be a quick change +and the docs as noted + Good catch +Or use a w fixed capacity +The login code in is repeated in many places +Forgot to do it +Im just putting the thought out there not making it an official part of my proposal at this time. +As usual a very good patch I like the new processing of xml much and we should work on this as basisI also feel that we should do a release very soon +Attaching the config files that I use for each node +is now the front controller for SWF even in a JSF environment +none is probably the most sensible default but it breaks backwards compabilityuIm definitly for none as default if old users code expliciitly fails with this default +Until we understand the requirements for this and evaluate any alternatives this wont be No schema attribute for slink instead the specified view will be used to determine which scheme to use. +It will enhance on accuracy +The problem still exists with the latest CVS +Resolving as. +committed patch for it +BTW this is not just a Linux issue it fails on Windows too +UI should have provision to enter nic ip addr to acquire specified ip +Wow I didnt even know that page existed +changeset details the actually changeset Thanks +Only focus on getXXX apis for the now on the inquiry service +I also see the same issue with ee Scripting tests where Jython UDF scripts are not found in classpath +Ive just tried this file with the Apache tool using the option +What are the dependencies for building this +There is a lib that gets bundled on the platform +I filed to backport then found thisI took a look at the patch here but Im not sure if it subsumes the changes contained in +Just a thought may be a better name than +trunk +The only reason I say that is that I attempted to actually use the strict setting to ensure a key got promoted and ended up running the code through a debugger to figure out why I was getting the. +Thanks for the quick responseClosing this issue as cannot reproduce. +I also noticed that AVAILABLENODE is duplicated +This makes sense just committed your patchThanks this a bit late but Im concerned about this change +Still not working Ive updated my file too +The attachement is the test mentioned in the description +Syntax is like Nialls examples above +Also the patch for seems similar to the patch for trunk +I believe weve got all this functionality in CVS now +and are like the same thing +jar doIf you print f you getforfinUsersnitayjfbgiraphtargetWorks but entertainingAnyways I fixed that as well +Probably should work if you just change that to public +Moving to +Im going to commit this soon +You may need to recompile with the new overloaded methods from the previous xb version were removed and for some reason the JVM expects them to be implemented +Someone pointed out a particularly easy fix dont reuse files after a restart +Attach a patch to flush entry logger before flushing ledger index +Ive copied that script back into place havent tested it yet though +Therefore tarball release can continue as it is but rpmdeb packages uses system provided libraries +It turns out I had TRACE logging enabled and that slowed the test a lotThese are the throughputs in the DIST mode simulation n yn yUNICAST starts a little slower than UNICAST with nodes but its performance degrades much quicker as the number of nodes increases +The hard part is to extract the v id when parsing the xml +Think both issues address the same problem +I was just stating that it shouldnt be done apart from full input of the developers +Using operations would result in faster implementations +Added references to accept parameter to Remoting Guide on branch and to Remoting Guide and Users Guide on branches and. +Let me try to get some trace logs now somehow +Hi StefanIm preparing a new version of the patch +Actually this patch should initially go into the FOP code base before beinglater propagated to xmlgraphics +Both of these issues concern the propagation of metrics from the tasktracker aggregated at the jobtracker +additional mods to fix remaining test case + +My recommendation add +Ive tried to make a patch for this issueHow is this solution +i think kannan is correct here +Were nexting through the tfile block that has the wanted key +Changing from blocks to related +maven based test casesexecute mvn to be very similar to the fixed several more tests tried to make a work around +Please give us a component in the next release +I am not going to block committing this change +Ive verified it on two more platformsworkspaces for and it seems to be working just fine +Yep thats +Thanks! +Sorry producers of course not providersAnd yes I red it all above +Now that I see the error related to Ill fix it as well +A workaround for this behavior is to first create a DOMDocument then invoke create on it. +bq +Task failure would mean losing processing on corresponding inputsplit +This issue is closed now +The patch improves the console output whenever a streaming job is run +I dont really like the fix because it requires all of the jersey methods to be modified in exactly the same way but that is the exact same fix as did so I guess it is OKThe root of the issue appears to be that is trying to guess what the content type is based off of the URI of the request and always setting it no matter what +IMO We should avoid referencing chukwa in these patches as theyre just generic to send metrics as JSON even though theyre originally from chukwa +The intent of this metric is to be abstract although RS compactions minMax are the only two uses currently +still open but this might not exist with juddi +That implies that JBM can recover from its own failures +you can use cargo to set the jvm memory is this a bug with cargo +IMO hooks needs more fleshing out with context about node potentially unfamiliar to target audience +Anyway I can live with that +Thanks Tsuyoshi for your first patch to Tez +Looks fine! Though I dont need is it possible for configuration to be null or notIm still struggling with setting up Camel project within my Eclipse the reason why I didnt submit a patch myselfThanks +Thanks for the patch JMS. +However the same problem may occur when changing the target of a method callI will extend the test and open a pull OliverI have pushed another update if you see any other issues please the JIRA. +It is currently done as iteration. +Where is this feature documented +How about cluster node jobhistory and mapreduce respectively +Stefan you are right +Thanks +If we dont here from Brent well take up the patch +Doing this retroactively for all of commons is on my todo listI did a test run with stuff and while the javadocs happily moved from the maven repo to the maven repo the sources didnt +Would be good to not replicate setting the env varsThanks will fix this +You need an and an if the original intent was to question AegisSimple I guess I could add a test case for that too. +So IMO we should reschedule this for the first CP +Ive reviewed the Errors Reports section + +To test this script the following commands can be run before and after its use to see that conversion actually took placefind +Basically at the moment we give the repo location as a parameter +As the dojo classes are part of the infrastructure the scripts are provided as and not copied into the repository +Will attempt to reproduce the problem but independent tests have passed +This patch fixes the off by one bug that was messing up the locality of +optionals property is deprecated now all control over optionality is on nullable constraint +Ive got it fixed now +but it is not confidence inspiringJetty have been around since +Its just a matter of updating a file for cmsFor buildbot I can probably update tlpreqLet me know +Trying to find a solution atm +I cant see why these are issues +changes from arent covering extended selections properly +Which tar archive are you talking aboutAbout the README and the lack of docs in tarball youre right +verified on +This is the real bug +Olivier your example looks like what I had in mind +yes your are right +Please close this issue if this now works +this as first paramter yes +Hi Roger this is a good additionIm thinking for if could be good to integrate that part directly into Pivot but we have to contact the original author ask him to sign the ICLA etc +Thanks again Suresh! Update the patch based on your comments +To my knowledge the RPMSRPM release of EAP has never been available except via RHN +I see no way to leave them empty except not using auto and then typing them the same for every submod +Fixing the race condition between two thread external thread tries to shutdown the cluster when a condition is met main thread invokes After method which checks cluster reference for null and then shuts down datanodes and then the cluster itself +Its been implemented for the taglibs +So for instance if I click on ArticlesTools and MigrationTools I am taken to the Blogs and articles about Derby page but I am not positioned on the Tools section +will commit if tests pass +in JBDS +Thanks Sravan! +Hi Olivier thanks for uploading the log +some localization in action error messages is not working some error messages are not showing at all in the more fixes in rev and +Works great Ive tested individual geotiff individual shapefile shapefilegeotiff layergroup +Hi BabakI can reproduce the error without try to release the camel in the trunk nowDid you try to reproduce the error by just using mvn clean install +Heres a version that has less code duplication +Applied thanks for the patchThe implementation of the servlet within the blocks fw are rather uncomplete as so there are still plenty of methods that needs top be fixed to do anything meanifull. +Ive confirmed the fix is working! Thank you +There is a test class namedcontaining tests that apply to XA environment using simplistic +For Autocomplete the problem comes about with the use of the selector attribute +Unless + +If it takes calling it to make you happy thats doable +It looks like the important bits of the might have been mistakenly removed at some point over the last month or two +I personally just rank items that has known ratings so that the relevance judgement is always known +This patch also fixes since it was simple to fix while I was at it +Given that can change the types being written by output formats would it make sense to include some mechanism to select configure an output format based upon the ptypes family +Fixed at revision to the branch in revision . +If I dont hear any objections I will commit it tomorrow morining PST +For a stress test too much sleep time just means less testing each day in the stress testI think we can close this one but I think we should make a new issue to reduce the timeouts and eliminate the deadlocks and also track down why the compress is needed to reclaim space used by the test since it seemed to be something different than +Label newandnoteworthy is added +The patch looks good for me +I added an integration test for the max call limit check which in fact revealed the issue the method. +Theres nothing in thats jumping out at me at being the culprit +I doubt we need the Overview page section since the user is already there +We have to add a follow up issue for the other scalability bottleneck +wondering if this is regression +Now we only acquire the primary owner lock instead +Committed revision . +commited in trunk r r rThe improvements are in trunk only at r Next time for the sake of patches history create rather Jira issues with improvements and bug fixes separated +However I would add the socket stale check to class itself +Remove both the forms and screen definition that you have created +the new PDF input stream implementation which caches read data and allows seek had a quick look at the changes made and I think that this is a very good step forward +Offtopic Agree completely Evgeny +bulk move of all unresolved issues from to +Ill add some tests tomorrow and then think about forcing one reducer +revision +Sounds good to me thanks for the summary. +Now of drlvm classlib tests pass on +Closing all resolved tickets from or older +adds this to the +I agree with most of the items +Stable patch that fixes some of the issues commented on and mentioned previously +Hi screen shot shows changes done to the toolas it shows SORTTYPE and NOOUTPUTROWS details ofSYSXPLAINSORTPROPS table only for sort related nodes! I forgot to remove the id sorry about that +changed in release +Users issues seem to have been resolved by reviewing the attached sample +Fixed by +Hey HenriqueSounds like a good approach Ill try to put something together later this week +want nothing to do with Maven and I think thats just fine +Thanks. +What kind of SRS handling was chosen? The streaming renderer should not be trying to render this file the feature source should be wrapped withing a force coordinate one at the catalog level +Please could a hibernate developer comment on why this was rejected +What Ill apply +Since it is internal I can pass on the connection url to you +Marking closed. +Thanks for the sweet patch Kannan. +This patch should solve the issue +Attaching first patch +Reopening to take a lookI have a vague recollection of stuff not being closed down if not all is read out of the socket +Sorry but it doesnt work +This bug is very old and does not appear to have had anything done in almost a year +State diagram for new HDFS WAL +Thanks Peter +Theyll all get bumped to or whatever +Ignore older patch attached to but new patch should fix things +Make sure the is NOT changed +i didnt mention this in since its really just an implementation detail of which is already close for +Patch just basic GC based stuff for now +Although a check to validate the filename would make sense +in one pig script two macro definition both dot mask any builtin UDF +Upgrading to Alfresco solved the issue. +I looked at the RS logs +Only people who need to modify the lexer code need to run flexbisonThe python script is a code generator to build the serialization code from a schema and its not strictly needed to build the library +I suppose the region could have changed since the fail but this is way better than what we currently have +Please reopen the entry here if it didnt work for you. +Client certificates are not required its just Basic Auth using LDAP as the auth account were you logging in with and what client were you using +no there are two one embedded in EAP and one standalone +No test I assume it is covered by some of the exiting tests +I see it in CVS. +q file I think you can add a testcase to qlsrctestorgapachehadoophiveqlmetadata +I do agree with Mark and we should close this issue as WONTFIXBut thanks to matthew for his proposal which may have helped some others users with the code he made available +Unassigned +size attribute was added in nightly build +And sure enough just before getting stuck we ran through the new codeSo Im pretty convinced that we should tracking this all under the same JIRA it does indeed look as though its something about the new fix that triggers the tight loopIf theres anything I can do to help eg test fixes or gather trace please just ask +Papaki is Greek slang for the think that Banner v Logo v with the dark blue as the text white background Logo vare the best you could make those two banners if would be greatBest regardsll go with banner for all your workRyszard I havnt created a ticket yet for the site busy with other stuff atm +The agents will not have client certs but instead a CT will have one master client cert it uses to auth to the agent +Resolved in commit out issues. + +I did a lot of things in the same patch including ps ps support and refactoring of parsing to a separate class and adding test cases for boosting +Unit tested and so far it works as advertised +patch that solves the the value encoder is going away then at least this applies for the future +Its tlipcon now +Version of the patch consolidating feedback from Suresh and ATM +Thanks to Paul Brian and Andrew for commentsreview +Im not sure we can do this without introducing a new state flag though +I checked a few things before porting including testing that files were correctly handled if present +hopefully this week +As this is a part of the API promise it should be considered as blocking +proto then java code will be generated etcIf you want this to be postponed until after its fine +Looks like all the CITE WFS feature types are being loaded as disabled +Good luck +Now I will try to create branch and commit those changesIn future it can be generalized to all output connectors +Affected files projects +If you still wanted to do such multipleattachments to the one issue is fine +And correct file would be good +This is probably out of date and can be closed as such. +Thanks +I tend to think that no one will want to have a query string parameter with an empty name since the rest of the code doesnt allow it anyway +It seems ok +My problem is that I dont see any benefitsYou dont spend enough time on a laptop or small monitor then +you didnt commit +This issue is resolved in the recent Embedder build +We have very less time we dint know about the issue and we developed the webservice using Axis +Updated patch that fixes table deletion problem and adds more parameterization +Jon Noted thanks for the clarification +Hi RobbiePlease find attached proposed patch for this improvementcheers good to me patch applied. +Right +This will also help simplify the growing number of was fixed +This patch applies to the trunk codeline +Commit logs available in for tracking purposes +fragsize works only with the so if we want to ensure one fragment per field value then we will have to work with the fragmenter +I will update this issue when things are at a stage where we could do a test import +Hi AnkitHuho I now reproduce on trunk weird! OK I will get back to this later +Im reassigning to Gary as the latter issue is assigned to him +Speaking about EAP Affects Version is EAP ER because there is no version for +WS CXF +Yes I reproduced itIt happens only in Jetty +The patch +So I think we need to upgrade for MCF + +Joram checked in the attached text into the cvs +why +I thnk this is same as +yeah only trunk +This required no changes in the plugin itself only the tests were updated +Sliding times are much less useful if they come with a big cost +Also please include management server log file as well as the API command that resulted with this error +Kill the master to exit without setting shutdown node in ZK +minipatch +Ill get to this soon Po +A very good comparison would be memcached client +So just like all of our other schema entries you will declare them at designdeploy time +Adding an xhtmlVersion attribute probably makes sense. +Damon the latest patch looks good +Will revert back with test results +whoops typo +Thanks for such quick turnaround. +Feel free to create an issue immediatly if you fear to lose the context +JonathanCan you test with? We are not maintaining now +A fix is checked in under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Swathi Are you running in secure mode or mode? Are you running metastore as separate process or embedded within +Actually I think you probably need that patch opened for part of this +And Eclipse generates the diff in a different order from git commandline so I cant tell whats changed by eyeballing the diff eitherIve attached versions of my can you attach your fixes as a on top of those +Jason the test helped tremendously here is the fix +Could you please suggest how to make BCEL to be thread safe directly in our application we are not using anywhere the BCEL related stuff +lgtm +Any update +i dont see anything in the logs that would help us identify the problem +extends by adding some of the attributes of in particular the messageId and the arguments arrayThis will be handled in a uniform manner by always setting a object as the cause of theAs a result of this change I expect that the classes and will be removedThe classes are likely to become generic and therefore will get moved toThis change will simplify the exception classes and remove some of the unnecessary dependencies between the modules +Done for the AS guide +Ready for review +sorryit look like source codeit sound that dynamic taglib wrong +the short of it is if you want wicket to work then do not have +Thanksyour patch with some minor changes is in rev +I am fine with making the spec match what we are currently doing if this is the case +Thanks +jcodings objects arent serializableI made a slight improvement to error messaging in de. +Lets say that I understand the idea but wonder of the complexity and how this would work in the am not a JNLP expert Ive only written simple applications and the plugin clearly reflects would that work with complex requirements? How would that work with transitive dependencies +Attaching patch file which should fix this issue +Is there someone I can talk to in order to quickly rectify this? This has become a real pain and I have had to work through JIRA to communicate with others +Some additional notes on this issueIf I look at the container using jconsole it looks as though the undelivered messages are still on the inbound queue +I have created another patch to avoid patch applying problems +We could consider just using the cache control currently used for the Expires header +resolved the issue with revision +Thanks a lot for this patchSending trunksrcmainjavaorgapacheservicemixsaxonSending trunksrcmainjavaorgapacheservicemixsaxonSending trunksrcmainjavaorgapacheservicemixsaxonSending trunksrcmainjavaorgapacheservicemixsaxonSending trunksrctestresourcesTransmitting file dataCommitted revision . +Please do so if you know how +I also assume triggering GC here is havent verified exactly what is going on here and if it is the caller or callee at fault +This goes against convention but is more flexible than putting each page property in the context +using aop in trunk +rebased and fixedwill probably need another rebase after sigh +I probably missed them because I was using newer version of surefire +John and Eric you should close the ticket +Opinions +What dif you use as a base for the diffs +After removed static please rename it to no +where will the datastore be stored on the filesystem +But it might be nicer to keep the one Changes and have that link to a landing page that would list all the releases with a link to the full changes page for each release +This simple patch fixes the file descriptor leak +Yes it works fine now +bq +resetting P open bugs to P pending further review +It might be worthwhile for you to follow the conversation at +What are the chances that the DB becomes corrupt? IF you are quite sure that the DB will not get corrupted We can tryImplemented THE setting on THE server +However this doesnt happen inside probably because were using an older version of ll check the status of the and try to upgrade it +Hoping for it to get some committer love soon + +Ill keep an eyeAccording to the docs Apache should receive an email requesting the auth codes that C Michael has +bq +Restored latest comments +Plain bean factories allow for programmatic registration of applying to all beans created through this factory +Thanks and Jim for feedbackWill add back in row locks to operations +Where exactly should we put the required dependsOn in order get around this bug until the official fix is in? +To me it smells like to refactor the single class in its own assembly +same patch with white space adjusted +Yes I believe that HS will leak connections but is deprecatedI believe you should switch to HS +Its testing stability of the sort and it needs to have two different strings so that it can tell that they were not rearrangedbenson +Thanks Henrik! Im pulling your code into my fork +with updated +Now that our focus is on and in the future support our plan is to not fix pluto issues +Review edit and run the scriptSorry Windows fans +If you read the recent dev threads on this you will see how we are considering moving away from commons as a publicly released project +No more instances of WRE +The contribution zip +Then the test closes a producer and creates a new one and sends a message +Hover over the comment you want to edit there will be an edit icon in the top rhs of the grey box +Improved the fix and attached a new patch +Templating API changes made +Removal of a share must not be ignored otherwise the index will get inconsistentSee attached patch +added a test for this +patch files for +This must be some kind of magic constantI have no idea how Novel eDirectory works and how it handles Everyone group +min +At any rate its the core classes that will represent the vast bulk of effort +It would be nice if we could support the notion of consistency level here but its not as easy as setting a directive like other output formats so instead we just allow defining a maximum number of hosts that are allowed to fail before the entire job fails and default that to zero +I will do that and commit if there are no objections +We have a ou file and a cn file in the schema partition +Fix javadoc warnings and add unit test in v patch +Closing resolved issues. +Ive changed it to bug and made it blocker as priority as it affects consistency of loaded data as well +This appears to be handled in trunk already +If theres some other aspect Im missing feel free to. +Ive changed operator for method equals and other file types started working as expected +About the schema locations in included files at least Eclipse does not take the location of the including document it will complain with a cannot find error +seems to me all currently available patches have been committed +Solved on trunk and closing all issues that have been in resolved state for months or more without any feedback or update +I see Todd is working on it +one finalised JIRAs categories should be aligned to it +Fixed a problem with cfs files +Thanks for the background +would you please try this patch? thanks a lot +Removed in +A duplicate of +Pushed to master +These were previously verified to be incorrect. +tested on windows xp osx and rhel + +I fix the problem with this patch +Merged to resolve conflicts with changes in VM part more compiler independent +I think the plan is fine +Did the tests pass for both +because theyre the only relevant scopes forVerified in JBDS +Dennis I think also that Benjamin is right +Right now theres no way of knowing the txid of the downloaded image +this new call name might be syncConsume or other better name +I think that can satisfy everyone +BTW where is the activeness of a vertex? this is totally not included currentlyThe Pregel API dont include this kind of convergence this needs supervision of the graph and cant be handled so good with it +Thanks Ted studying the design and code thinking that how the rs group assignment w seperate conf and w table priority would benefit different use casesclient app also if load balancing is involved. +I thisWe can always reenable it when people want later +I mean Preferences General Editors Text Editors Annotations Sonar +The second was that even if we do that you still have to block for one socket buffers worth of data +Are there any objections to closing this issue +implementing support for openshift flex is covered by issue was done in M Flex not relevant. +Hm +OK to removeBinding for logj version a widely used logging framework +I am attaching a yahoo translator from my local build +Rocket science +Committed revision Now trunk revision +rebased for +proposed patch which fixes the so patch can be reviewed and considered for Final. +You have to set the system property +Thank youIt would nice to have the ability to set this globally +Hi EliI wanted to have this change in the same Jira as but those were reviewed and committed +update to resolved the issue +Closing this issue since there have been no further updates. +While examining the session object and try to resolving loops using hashmaps. +In process the state in the shipping and billing panel is mandatory field +That will help cluster administrators diagnose the case when one of the has gone down or having disk issues + I always thought this was a feature +Thanks all for the help +This completes the feature so marking as resolved. +Id rather not muddy the waters by having different fixes in different releases +Well apply it to trunk and it will go into the next release +I started with a new workspace and added EAP with its scanner set for +Seems like was missing +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Workaround for the second test case was committed at r +Looking more closely at the failures two of them seem to be because the recovery could not find the jhist file to recover from +A proposed fix for this issue +a neat enhancement for the future +Ant Runner should be fixed for the cases when it runs or By the way maybe the problem is larger than just Surefire maybe team and Junit one should try align their way of thinking if possible +I have this working locally exactly as you would like to use it +Marcos there isnt any way to recreate some CQL tables in CQL +We have decided to focus on environment variable support for +After much tweaking and polish and refactoring Im done integrating this for now +not a blocker for tracking in +Is this jira still valid? +Closing since it has been fixed for simple checkboxes and probably wont ever be fixed for checkbox lists +the changed with implementation for different phases +good stuff you were like seconds faster than I wasD +Patch to set the port to match exactly whats specified in please review +Carlos do you need a separate attachment for each jarbundle or you need separate JIRAs? There is over jars poms maven plugin in a bundle I already a past one bundle for all jars was enough +Moving out of +Interestingly it doesnt have much influence +I spoke to soon it failed on the second trial +for to as well +Perfect +JAXRPC tools issues may be fixed by community contribution +HiI dont understand exactly what you mean do you want to have all comment entries in the reportThe fact that we filter on the matching between a grammar and the comment is a feature o +Then I will close this jira +generalized type signatures for map reduce and scan and updated xdoc comments in r +Basically under load it would allow the indexer to fall behind but would still allow for mutates to occur as fast as possible +This issue might actually also be contributing to the trouble were seeing in +Sure there is a schema in webservicesrcresourcesschema +There is a similar problem in too which Ive fixed in the same wayHowever it may be better to fix these problems by using a concurrent collection rather than synchronization and copying + +With the switch to real annotations inner classes are not supported at all Im not sure if inner classes are supported in general by DS but dont see a reason why ve added a first implementation to handler publiuc inner classes +The build first checks if the HTTP is ready +Fintan do you have a zip or a location for the example +May be I coined a wrong word here +Logo added in revision +yes i think this would be a good been thinking about trying to switch out strings from sh and use instead for a lot of the output +Closing Jira. +the link doesnt load for mewhat example are you using +i found this appender also takes time +and I am using maven +but Ill take this issue and get it applied first +Cleared some extra files that got added with the patch +The methode factorial is not available in the class you declared +It needs a module add command included in the script +Its quite cool although I must admit it requires quite a lot of cognitive processing to get it right +But a copy of the database is also interesting especially in cases of non reproducible errors where this may have been our one chance to gather info or to allow someone to debug a system specific problem when they dont have access to that type of system +I upload the patch to fixI confirmed it was succeeded to execute generated script on derby +Have a long running replication setup and its up for a few days +Any remarks on my questions in stack Sep PM? ThanksCan you check my comments in Hong Tang Sep PM? Did I miss some of your questions +Added support for virtual hostsTests updated to use the virtual host test which is created by default by the rev . +current could be null while findClass is executed +Verified in JBDS +When doing so please keep in mind that if you specify multipledatabase instances using the same name is still a +Oh nice I didnt even know we had this script Thanks WillemCould someone with the proper privileges run this script now +So this is both line numbers and editor menu +this one got past me +I benchmarked this some more and it looks like it may not be a big issue +Testing now +I fell sick last week and then got busy with a conference +Also I do login with admin at the end of methods +This solution can also been used for a clustered environment +I assume that one could set the quota on the root directory to something based on the amount of memory available on the namenode as the simplest means of control rightDoug yes you are right +In the dialog click Download and install +Thanks for another patch Stefan its applied think we have a working implementation now therefore we can close this +ok one more thing +This is from autoconf on RHELCompiler specific flags should get added after compiler detection +Added fix to trunk by modifying in moduleAdded new testcases. +Thanks for commentsIll fix the patch +no probs. +Right +Deprecated the methods without dbName in the Hive enoughPlease file a JIRA for the rest of the cleanup +Proposed patch for the issue +This should go to the jai stuff soon anyways though when WCS branch rolls in. +It would be nice to have some details on what triggers the problem which dodont exhibit this problem +Thanks seems to be fixed testing with revision . +Once I get past that Ill let you know if this is still happening but just updating to for the client has not fixed this issue +It doesnt need to be a WARN +And since we are going to move PHP to attic after release I would close this as fixed and tag JS and JavaIf anyone has contributions for the PHP side we will use other ticket. +On srcmainbin should be updated to call like it does on trunk. +I didnt configure any are three directories under the server directory alldefault and minimal +This class walks through all the indexreader and warming each field as needed or configured after the index has been opened in a shard server +You shouldnt need to exclude anything + +This still has MR classpath dependency +Also features if at all possible should not impact those code paths that dont use the featurei will address the specfics in another comment +I have experienced the same exception as in the original bug description with wicket on tomcatIll post a stacktrace if necessary +I attached a new version +I found a machine on which I could reproduce the errorThe submitted patch does fix the problemWaiting for a from Hudson +Manually tested +this one seems to not have any interested committers for a long time and the security issues are addressed by the HMAC changes inFeel free to reopen it if if this issue is important to you. +Im having yet some troubles with replication and im investigating +Update website svn ciSending Sending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexcommunitySending srcmainwebappSending srcmainwebappsiteTransmitting file dataCommitted revision +I generate client and server stubs from this WSDL using the followingjava cp AXISJARS v p o stubs java cp AXISJARS v p o stubs skeletonDeploy true and compile the stubs withjavac g classpath classesAXISJARS d classes stubsarrayclient +But here it is code from a closure and that code indeed goes through the setProperty call +I fixed this problem by modifying two methods of Both equals and hashcode methods are fixed +Ive logged a jira to get the default changed +Im pretty sure this works with AS Resteasy integration +How to configure the drools runtime correctlyOpen menu Window menu item Preferences tree item Drools tree item Installed Drools Runtime environmentsAdd the dir binaries from the zip as a new Drools Runtime EnvironmentActivate checkbox of that newly created runtime environmentRestart eclipse. +Trying again +If you finish it before we release beta please correct the versionThanks +I have events a week and need to be able to see month vs month quarter vs quarter year vs year +closing as cant reproduce +Fine with you +Issue When adding new hosts in add New Hosts wizard no matter what new components we selected the review page saw the Services field emptyReason In the Services field only selected and uninstalled services will be listed +Thanks Michael +For dynamic failover the initial failover string used to configure the client is really a bootstrap because it is most likely not complete +OK so it was a design bug too +Please have a look at the second by we miss the data between week and when you look at the filled area in week you can see very the apparent between normal scanning peaks and overblown the way we do not have any other software running on that server +we shall fix it for +So this functionality does not work according to JPA spec +TrunkCommitted revision branchxCommitted revision thanks Chris +If N is the maximum number of jobs that can run in a particular pool when all the maps of first N jobs are done and only the reduces are running the map slots sit idle and the maps of next N batch of jobs cannot yet be scheduled +It reads from thequeue consumer attachment added as per comments above +POJO serialization has worked since +say +Committed the change above in Git ID cfbccebbdebaefaefa +Currently names are +Yet having methods instead of one was kind of ugly to implement and I didnt want to add switch case statements to visit methods just to call a different method +Classic parser cant extract metadata from but can extract metadata from it +Also same class name in at srcmainjavaorgapachehadoophdfsserverdatanode +But if average norm suffices that can be computed on demand and cached in the Searcher without significantly impacting performance +gcj is not used by standard fop build transition resolvedwontfix to closedwontfix +passes locally +In any case it may not be worth the effort if nobody needs this +Applied to trunk + +Thanks Konrad +This seems like a simple useful additionShould this be in or +Given that the service name is from the AS layer this may be an AS issue +Attaching an early draft for converting to +What exact failure do you get? Where would you expect schema locations to be defined? saw the same thing in Rational Application Developer +So why do you need to reinitialize the service and port here? What does this give you beyond simply retrying an invocation on the existing stub? JuergenMy first reaction was exactly the same I tried to subclass the and do the magic there +Fixed thread manager not to hold unused Thread objects +Its in now. +Ive verified mvn clean install +Patch applied as provided +This Jira covers the technical review now +This issue is also resolved by the patch attached to +Im experiencing a error as a result +I was waiting until M completely finished before addressing any maven distribution license issues +Is the INFO logging not only happening at startup? Or do they log quite often at runtime as well? Choosing the right logging level is not as easy +so we will need BRANCH hudson jobs in place before TXF can be added +should IMHO mirror these capabilities and support requestresponse bodies for all methodsShould PUT also be present in the subject? Both PUT and DELETE can return with no content or with the updated or deleted entity in the body which may not be identical to the one that was submitted +So if the isnt thread safe that would be a bug in the JDK +Once the new byte block pool that records the slice level at the beginning of the slice is finished the skip list can be completed and then the basic functionality for searching on the RAM buffer will be done +Also a bunch of fixes +Will do will add something like nt repair patch will stagger repairs if snapshot option is usednt repair snapshot pr +If further work is required will open a new one for +Clustered translator now works on AS so I believe this issue can now be closed. +Doesnt need it +Would someone mind confirming that the error messages actually go to stderr? They cannot go to stdout +Ill leave this in the list for +Oops you are right! Fixed +I think traditionally companies were giving resources to the Apache Infra so they can be managed and controller by ASF +Yes You are right +The project homepage usually contains some static contents like an introduction to project how to get source code and so onWe can follow this convention +Did you try to drop and recreate a keyspace with same name in the past +entry for this new featureWe both agreed that we can deprecated later once we are close to getting rid of it +Besides we should most possibly attempt to get a Maven plugin into the Maven core distributionIn other words Please be so kind and have patience for the next two or three weeks +if we were to do this then the tests would behave differently based on whether it were in a or run from the individual project +Could you add a test for this also +Attaching changes to address this issue Added a list item saying that you get when you run on Java SE Added a one sentence paragraph describing driver autoloading under Java SE I dont think we need to amend the since there we have separate classes depending on the JDK level +That should be the only diff between this and the last rev of the patch +Ill try to chase up if theres a DM plan for avoiding this +looks good +Ill try to keep more +The fix for this problem was bundled in with the fix to for bug +currently thats because you have an int and an Integer which we dont have evalutors to handle that +Also adds an expert public String for debug purposes +It takes some time until everything is up an running again +Hi Luke +This new patch has the changed nameSubmitting only for trunk this time +Thanks for that revised Dag +Need a test program that replicates this to investigate further +i hand copied the files for the zip so i hope i didnt put any in the wrong place +yea I will cut rc from trunk before the end of the week with a bit of luck and commit the deadlock fixes once I track down some further regressions with the unit tests +We will automatically parse all the host names from all the certificates that ATS is configured to use. +Standardize on binary and have easy ways to pull it out directly from luceneSeems option has the most flexibility and least impact on existing pure lucene users +FilipeIve just committed this. +Closing the resolved issue. +Thanks Torsten for the feedbacks and thanks Thomas for having a look at itI mark the issue as resolved and Torsten has to feel free to reopen it if needed. +Will commit this small change later today unless objection +Hi DigyBefore I apply this patch I like to know if this is a port issue or an existing issue with the Java version of Lucene too +Looks good to me +I guess the page has been changed because the link on machines is now visible +Can you attach a new patch that places the test file into the right module namely +It really should work with other stax parsers as well +Cacheinvalidation does work +Patch with test case it failed here +Here is an updated patch against trunk +Could you point me at what has been done here as Im just about to attempt a Quartz implementation +Would the timing issue still manifest itself with the cacheLevel set at thisSen. +So a value of retained jobs would work fine for small jobs but fail miserably with OOME if bigger jobs were submitted +slip to next version +I am assuming this is intentional as for default store that means the store is blockedThen why do you want thisbq +Trunk and x? Id go with both +Committed revision This should hopefully make the test failure go away +Do we have any updateThank you on the case now +It is NOT performance critical nor is it run on substantial data volumes +The reason for asking is that the Perl package installation would be much easier to maintain on the Ubuntu boxes +Anyway PDE marks this line as unknown attribute match +Taking it overThanks this the only thing in Counters? If possible modify the title and fix all things related to counters +I have this applied to our internal branch slated for testing next week as well +Rob out of curiosity whats the status of this datastore in? Does it make sense to add it to? I havent seen any activity on it for a long time +I removed grammar doc generation from since it was using jjdoc +Closing all resolved tickets from or older +The trunk patch looks good + is a required property +The groupId is missing on the pom for this one +I just committed this! +Bump any progress on this? Currently we are having to repack the jar to strip out the Aries specific classesimports +on patch +For example in a I cannot run the validations in different threads and also I cannot run in different peers and join the results after computing +More jUnits will be provided +committed to trunk +Local config issue not a grails problem +Closing issue. +Committed the patch to both trunk and +It seems that a given javadoc param has no type +Its unselfish work much appreciated and very good to seeAdrian +Closing issues resolved in released versions +Then it will be easier to find all instances of this workaround when we decide that its safe to remove it +i will post it in about Minutes +Sure why not +Thanks for reporting back +I simply editedSureshs text into the admin guide backup page reformatting itslightly for improved readabilityI also attached the page for easier reviewingPlease let me know what you think +by looked at the paper Enhancing Applications Performance and Scalability by using Adaptive Compressionand some online tutorials about MPIIn the project we would be looking at two areas +I dont think this even works or is enabled in the Broker any more its probably best to just make the option depricated and ignore it + +Yes I see your point +Which version are you running +Could you upload to review board for easier review +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +thanks Thomas! +part of release +I have committed a fix for this +bq +Ive committed this patch to trunk and. +Great +Can you do a mvn dependencytree to see where the wrong version is coming from +But we could just some incrementing numberAn alternative would be to instead allow the of the META edit using a check which might be simpler but less optimal +Good work Sijie. +I think you are right that the patch wont work when default port is not usedThanks thought so +It is a bug +Hmmmm actually having compared values I tend to start AS from the server root +Regards tests against Oracle without failures +It doesnt help clients and wastes bandwidth but if it makes it more compliant +Another updated patch +I assumed underlying library already does some caching for hash value +Thanks scheduling this for +Fixed at some point. +txs your passion Rudy +Should be fixed now. +Yep there is a minor number of were found +Too late for but it could be nice to have this patch included in the next version +So please let me know what you thinkThanksCheers Chris +When it comes to open source software my knowledge is weakIn order to have a sufficient safe encryption we should also generate random keys and store them in the database +Im sure they would really appreciate the contribution! +unless you use offline mode locally in which case it skips the TwitterJIRA stuffBingo I was looking for this option but couldnt find it in the doc anywhere I could see in the perl code there was some offline option that skipped trying to do the but I had no idea how to turn it on +If its not going to be fixed by at least make the error message not completely misleading +When a datanode confirms receipt of a block log it +Apply patch from geronimo rootOne question I had was if the attribute values had to be specified in both the configuration plan in order for them to be in the server +This is the cause of the memory leak that I discovered earlierThis problem also happens in the standalone compiler but the affects are much less bad +Even if this issue is not fixed can you please add a check after compiling that the expected class files were created and if not issue a warning? I had to put print debugs to find the problematic file + thanks Sami! +I was also using the when I encountered this problem +Im not sure if this is still needed but it seems better to at least not clear reporter until the message is rendered. +thanks JeromeNote that this patch also alters the display of test results +Updated the code to use the relational extension namespace for and as part of +Will notify list later today if derbyall should fail +Ok then consider this a feature request to make the context menu for the two editor parts similarsymmetrical +Out of date if fix is required. +Look at the DB Browser demo sorting is implemented thereits just not implemented in the administrative portletseasy enough to add +infra ping any chance someone could have a look at this issue +Patch applied svn revision +Anyone who needs more can build their own +Yea was running the wrong test last night to many hours starring NIO is not good for the brain +bq +Oops and revision . +I will add these improvements to the next patch +Defer issue to +Will be in the next code drop +See the attached issue cannot be reproduced with attached smooks plugins +Not sure about that the SRID is used in a number of to commit +Thanks KristofSending CHANGELOGSending libbuildrideSending specideTransmitting file dataCommitted revision . +Just go to Editor page in and wait while editor is loaded +looks good to me +What does the IBM thinks of this + +I am setting up SSH etc +Download source code from svn build and then use generated jar instead of the one provided with distribution +Performing a clone on the project left the Data folder out entirely +Havent reviewed it +Verified able to launch system +Uploading +Its due to the P manager limitation it will update something automatically +passes for me consistently on trunk ran times +thanks. +You mean the feed connection timeout value in the Manifold scheduler right +As per issue it was decided to remove the corresponding classes for +Your client can not ensure if the transaction is successful or not since it cannot receive a response of the commit request from the broker when fail over exception message Transaction completion in doubt due to failover has told you that the transaction completion is in doubt and now its up to you to decide what to do next +everyoneHas anybody solved this issueCould anyone share the knowledge of the solution with the communityI thinks this is a blocker feature +Thanks Tom! +If not a bug lets move it out of so we can get the RC out +Raised to blocker +In fact I think I put it there myself a long time ago but somehow I forgot to do this on my own siteI plan to fix this by introducing a that uniformly sets the request encoding to but it must be used in combination the Connector configuration aboveKojis suggested fix should also work for this specific bug but a filter applied to all is more general +Committed to trunk +So one option here is just to double quote all generated column and columns alias names and let the Teiid parser remove the quotes that arent required +And so we have a component that lead me to report this ticketGood to know you guys are back working on commons VFS as we have reopened a ticket in Camel to work on a component +This verifies my theory POM file attached is a slightly changed version of the pom for and looks good to suggest that we accept Michaels submitted files into the central repo +m +The test FO uses an image file +If I understand correctly this issue is about a test setup of Hadoop not recovering after a suspendresume power operation possibly a laptopSuspendresume is not a supported action and I doubt it ever will be +works great +Ive added a test for this +The patch I submitted is ready for review +committed to trunk and backported to + resolved issues for already released versions +Attached documented proposal +It is possibly that this may be fixed in EAP but without a reproducer it is unlikely +Guess its been a while since I tried rake with a fresh source build. +Not a bug as it isnt intended that the namespace reproduces every feature of the underlying bean classes +Applied patch in SVN rev Baselines are from my system as SVN does not handle binary patches. +Maybe just notavailable +Closing jira. +blocker for +Helloin my opinion this issue is critical not only major +Thanks for the patch Mamta! I am going to read through it as as soon as I find the timeIts good that you have added so many comments throughout that is a great help +Thanks so much for looking into the issue and the quick +My last comment was rather extreme +The one exception is if compression is on I compress all styles that have a in the name like afinputText but if it doesnt have in the name like +The ticket view for issues with Closed status is treated as an open ticket +Fixed earlier today. +I committed the second patch +Resolving as someone did it in some other jira without closing this one. +No +Removing fix version of for those that are not in branch. +Tested +What happens if we replace its functionality with ant? +Reopening at Allessandros request +Seems like at the very least we should display an appropriate error message if we want to require the sys admin handlers for the UI to work +Indent braces in +What is the status of these items? Should this be closed and new Jiras opened for any missing items that can be worked on individually +This is the same patch but only the incremental diff from which has been committed +We might not end up implementing all my ideas and thats OK +The new patch are to be applied after the patch fromI confirm that modification of code at r have resolved the problem +I was seeing this no matter what my methods returned but of course that does not have any influence on what Tapestry is returning in other places. +It is very simple to define a factory object using standard configuration schema so there is no need to define a custom one +Attaching a patch to delete the old XML binding test and the old XML suite +OK I just tried but our internal build doesnt seem to be synced with the trunk yet +It looks good to meWould you mind adding a simple unit test in +What additional features does give? +branches contains the documentation changes which will also be committed to trunk when that is ready which I presume will take a whileHope this helps +usrbin +Also the workaround mentioned for is not intended to be a method to set the query timeout but to increase a lower level timeout that should not actually apply to query timeoutAlso if the same statement is used for additional executions the timeout defaults to and not the connection property +Manually tested +Closing as duplicate of. +Attention JaredSince this is marked with a Fix Versions of TBD EAP should it be targeted for EAP s Release Notes? Its a issue +Closing all resolved issues due to a successful release +Thanks my fault +I think this is currently being discussed as part of Justins work for citewfs as wont fix since its superceeded by the new dispatch system +do you agree +test +Test for the server build of +Please dont mark the issue Fixed unless its been checked in to trunk +Actually I want to only sign the response message as specified by the policy +batch transition to closed remaining resolved bugs +Thanks Sijie +I was searching for the wrong string to determine where I was in the test +Thx Stefan +This makes the documents unusable because the rendered font chars overlay the char images +I think this bug requires a fix in +bq +No me importa +I also agree that its not a showstopper for +We were unable to do that today due to some technical issues with our Hudson server +Move to is now in its own project. +It would be nice to have a definite answer here as most people who encounter this problem will find this issue +Yes you have interpret my last comment correctly +internal package build is resolved +The priorities Ive been given are helping get ready and preparing materials for a clustering course in +poms for added the jars to the jboss repo +Not for me +java file Press do you expect the next step isI would like javas hot replace of java classes to just work and not see any restart of the application since that will possible trigger concurrent triggers or similar behavior you dont want in a debug we do what you suggest debugging becomes hard dont you agree +Attached patch fixes the issue +I built maven from source after changing the version of aether in to +Ill create a new ticket for a future improvement. +Back to the drawing board +I assume each of these threads is using its own instance +have tested that using +batch transition to closed remaining resolved bugs +Would this issue also be responsible for ignoring a argument provide this flag in batch mode the branch plugin doesnt use the version Ive specified. +this seems to hang together by a fine thread +Jon this is great +Mike this is going to creep up again in Resteasy +We will handle the server side for thisIn short we will provide service interface that can decide whether or not to allow a new download an implementer of this interface will be asked by the deployment service and will send a header to the target +Nice strategy for solving this one we should be thinking about viewing this feature request together with one or more of the linked issues maybe especially patch which applies atop and exposes the testPreprocessor as a property and adds an integration test for it +Hopefully there will be no collisionsWe need full control over this from both javawsdl command line and from +The were started up again and they all went down again because the were retrying the tasks +Patch for reducing memory churn attached +This occurs when an expires header is set to and also when an expires header is not presentAdditional Comments I would expect topic and queue destinations to have the same defaults with respect to how messages are removed +The OS shouldnt return a read error unless it has retried several times the drives ECC has failed etc +Trying to add to RB faced some issues +Verified with JBT +This happens when there is one sstable for hints since it wont currently compact it away and tombstones remain inside it +Dirk what OS typeversion and Java version are you using +Thanks for the review! +We could just patchchange that file for now. +The logic to do that is more complicated +sorry this issue has to be moved in GTNPORTAL maybe +Maybe hudson will notice it then +Because of clearify here the steps to reproduce the bug on any platformStep Run login with Remember my login optionStep Stop Run cleaned againStep Login without Remember my Login option Login OK +How about resetting the seq numbers in a test setup methodThomas can you do this so that we can finish this? Tx +Im guessing the donotreverselookup along would do almost all of the speedup youre looking for +we have seen this problem in our cluster too +Hi jacquesNo problem Thanks +Unfortunately I did not have any time for backport +I mentioned with that notion in mindAlso you are correct to say that wed integrate with the workflow classes +seems to be doing something it shouldnt always updating the repo in some wayThis isnt a problem with the repo though +revision resolved in WODEN branch +Maybe some time later these tests may be revised +the attached example is flawed as i mentioned in my comment +Apparently the NPE was handled by either Weld or JSF and not rethrown +Both of those have sources and I cant find an upload request for either of them with bundles +bq +With I still get Expecting a stackmap frame at branch target +Please reopen if you know the steps to reproduce. +The value to delete exists so an nosuchvalue error is not valid +OK her is the would change hashCode equals and add subtract multipy divide alsoShould I initiate a poll on the devel list or fix a pacht the way I want it and ask for commentsArne +How about making the role names unique to +Does it pass for others when run with the ant target or with +A well defined set of error codes make sense +docs updated +So the user has more freedom to customize the context and we have a clean way to determine a context even for more than one dosgi service + +Fixed by setting temp for very large input +on Dans proposal making this flag a separate like it +Sorry for the duplication but Ive got a better fix and test over in. +thanks. +ThanksI actually meant that we dont test this automatically we dont have a replacement for with in the test suiteAnyways too late and too small a thing to worry aboutOTOH some patches started depending on this in +Thanks Todd attached patch that takes care of all your comments but the last one +Patch committed thanks Hari! +should this be closed +Backwards compatibility +See +Fix and unit test committed in revision . +I cant test them since the current version of soylatte apparently doesnt run well enough on Snow Leopard to run mspecI am going to close this ticket anyway since Snow Leopard ships with Java +bq +I committed this +that will surely not be a picnic +All the issues are now fixed +Bulk close resolved issues +Attached test scenarios which we have performed with BKJM and BK in NN HA. + +But what about xml as aregards applied +Include into convinced by clone feels like a very undefined state to get into I may have a clone +the project set will result in validation errors which will need to be fixed by wrapping table names in +remember if you have cached creds anywhere for committing to svn be sure to clear them and use your new pw. +The cause is the interaction between Accumulator UDF and binary operators +HADOOPHOME exported warning +Just took a clean workspace and built every thing from scratch + patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Thisll be my last nudge +Committed to trunk and. +Sorry +I then deprecated all the Fieldable methods in Document +The patch does not work +Beginning of a memory increase of in the memory +Committed revision the thread you create via this start method be a daemon thread? I couldnt really tell either way from the diffIs it possible to add some comments to the code you added while its fresh in your mind? At least on theserver variable and its waitnotify +Could someone submit my latest patches for this +It might make sense to have a single thread sending out all the messages +The components are then part of user had to modify the and because by default jbpm creates new instance of class loader for every delegate object requested +Verified on In case when verticalPositioncenter and horizontalPositioncenter default is used. +Committed to trunk and x +Otherwise everything will be allowing for expressions +Id rather have an actual instance created somewhere. +Thanks ArpitThanks also Jing for reviewing the patch. +I have left comments in areas which still need work +I think that this is an important platform and I would be reluctant to require a Java compiler at this time +by having the of my most used Maven version always in the PATH and +queries instead and those pull only a single D per Term I think +Adding the imports manually works so I would ask you if theres still something essentially to discuss probably with the other developers too +Just replace the INVALID to cant fix because we dont give a damnI guess youll get more of this issue from others because you dont care but make it so that others do see the issue and know that it is an issue and there is a solution to it and not have to spend hours tracking it down +I will have to play some dependency games in my pom to use the updated resolver +Thank you stack +All tests pass it would be good if people could test it against their indexes +Verified by Dmitry +even RC would not work +we could look up the Observer at httplocalhostserviceshelloworld instead of httplocalhosthelloworldserviceshelloworld +Your constructor should be where you set up your plugin +This is +Which version? Let me retest it +false alarm on that +Thanks for the submission Ive incorporated the generation of the documentation We use the doc +In order to help ease the transition from to it may make sense to make this configurable such that a different prefix can be used or it may not +And included in the as it should +Committed the changes + More granular methods will allow to change behavior in Modifiers changed from private to protected getters introduced +Otherwise there is no reliable way for the plugin to determine the location of the site descriptors of ancestor projects for inheritance. +bq +The current implementation seems to aim at the third approach +The failing test is passing when run on my machine +Example word document used by to expose issue +I will fix how it has no explanation +multi function tests this. +Sounds fine to me to do it in a separate JIRA if we later decide its necessary +I have a basic test case which passes but have yet to write the test for the part although it works for the build for the project I am currently working with +I have upgraded and moved the prototype into trunk under r +That field is for which should never have long unstructured data in itIf nobody objects this issue can hang around until someone decides to take a crack at fixing known limitations +I completely agree +Still need to decide the best way to handle the notion of a default +Thanks for checking in my test case! +fixed +Product versions keep on coming as it is the product life cycle +Bug fixed + +At that point the xmlns logic runs and because it finds the default declaration up above in the tree it outputs xmlns which obviously is incorrectWithout a total overhaul I would have to assume that your original patch is closer to the mark even though I have no way to know what interactions that might causeAlternatively something would need to be added to the stack interface to allow the xmlns logic to use it even if the firstnext methods dont return the default declaration which would in turn fix the extra printing youre seeing. +No the stuff youre doing with has nothing to do with Hadoop but is all to support features youre adding to Nutch +Starting migration +Tested to work on my local install +In the EAR deployment assembly we have these jars being deployed to the standard lib folder +In the future we can do better +Yeah if this has been tested +Attached is a patch to change the py files in it shows the affected files +Makes no sense to do that in the +I dont think thisll be available in as it uses some heavily refactored code in should now all work in the extensions is not mentioned in the becauseit needs special treatment +In console for can see that file descriptors are indeed now at k rather than at +But does that mean users have to run a hadoop instance to see the document? +Attaching a patch +will do the trick +I figured out why my patches were not applying +Awesome thanksAs a reminder always commit to master first and then anything you deem required for a release into the release branch +The initTasks should load all the classes and create an object per class +For the time being we intend to preserve Derbys current behavioras we believe it to be and adhering to theSQL standard seems important for this case. +This is fixed with next build after which has AIX fixes for CXF +The plan is to include in and then hiveservice will just forward metadata calls to the metastore code +I committed the patch to +Can you try once more +Remove the ear from server modules in server view +How to over come with these locks +simple patch decoding file applied with one minor modification use specify the charset rather than relying on the JVMs default good point! Now the top line in says Starting with the minimum required Java version will be so I guess we can safely use this method. +Hi MikeThis bug is assigned to you +buildivylibcommongriddevbin +I sent an email on the continuum list about this and I believe its better to continue discussing that there +This was never applied to the branch +Fixing Affects and Fix in Affects and Fix in Version +As these are only different roles of the single +Why do we allow commit if doFlush issynchronized? The main cost of commit is most likely in theflush as it doesnt wait for merges? Theres a todo abovedoFlush indicating we may want to make it in thefuture to not block merges how tenable is this +It can also leak a Channel as if state is set to CONNECTING then DISCONNECTED another connect can start and two channels are created +Attached is which addresses the new variant problem that Knut found +what target produces the problem? We have seen something like it in the past +class does not set schema option +Basically the endpoint is registered as a fault handler and hence have access to these information after a delivery failureFurther we could extend this to persist the dead messages but as an start on I would like to do this on the in memory level +Reviewed looks good! Interesting to see more of the mechinery involved hereMinor notes make javadoc misses param for parameters This comment in bind is a cutpaste palimpsest bind the check condition verify that it evaluates to a boolean Shouldnt finally clause in bind also reset the current auxiliary provider list? two spurious blank line diffs toString should include AS keyword after ALWAYS +I may also get rid of Cursor in the process +Please reopen if this is not complete. +Introduction +Happens oftenId like to try the snapshot by taking out the SMX stuff Im not using and making it like kernel but last time I tried that my app wasnt happy with that setup so I had to back down to plain Felix first then migrate to kernel so I could get filewatcher and sundry other niceties +Pushed upstream. +sorry to put my problem here do I need to open new issue +I am almost absolutely sure pulling configuration from other plugins is considered a bad practice +on may still be broken in the sense that calling uClibc stat may end up trying to populate our Linux stat structure and then blow up because uClibc has a different struct layout +All patches applied. +would like to nominate this as a feature +bulk close of resolved issues. +No I have not tried it with +Updating design doc +are not exactly light structures and they are not thread safe +That property was set across all tables so config was returning a when I executed the command + +I think the valid time feature could solve my problem but I was wondering if there is another workaround for itSupose two entities A and B with A having a FK to B and both Adudited and both with a property named NameHere is the scenarioday the entities are inserted +Thanks Jian! +Good idea I wouldnt have liked that as well +As recorded in measurements performed for perfect forwarding in the facet interface actually improves performance in environ +I dont usually use system properties for this so this part of functionality is not well tested +Ill eventually need to redesign some stuff in here based on this and another feature request for Criteria queries but given the current code structure was nicely done +Sorry for the delay on this one turned out to be minutes of work. +That would make it easier to debug problems like this +Hi DimsI do not consider this as a blocker if you do not use moduleConfig then you do not have any issue +Thanks Suresh and Alejandro for the reviews + +The issue tracker is intended for managing work on the project +Attached a patchThe patch makes sure that the +what do u think +Jun let me know if you have any further comments and I will. + +If it is ok I shall just make the changes and close this issue +Flagging most recent released versions as affected as well +Looks like Mark has already applied the patch +Here are a few example runs of the test python getrangesliceebbdebeddcd not marked bfaadabceaadaccfaf not marked python getrangeslicecdcbfbbbfc not marked ecaeaaddd not marked python getrangeslice python getrangeslice python getrangeslicefcceabdfcccaca not marked python getrangeslicebccabdfaf not marked python getrangeslicedbedabcfeaaa not marked Sigh I just realized that under trunk my nodetool doesnt work +ETA on this patch Jason +hiverc as part of the build and then verifying its effects +The script to remove duplicated indices was not provided +Closing resolved issues. +Looks like a for me + +It should of course say overview in the issue title +Actually I ran it ok against after I added the and the updateRow call so I think this is not a bug +Sorry for the bugreport +is the current stable release branch. +But this is a bit counter intuitive +Resolving as WONTFIX +Can you post the stack tracethanks be fixed now. +The patch contains the actual localization changes +If a change is required that will affect all clients or all endpoint why not just edit the Standard Client or Standard Endpoint definitions directly? Why does a new configuration need to be added then a system property set to specify the new configuration is the new default leaving the existing defaults unused +No longer an issue +Thanks Javadoc looks up to date now. +sure +Release NotesIn certain exceptional circumstances remote EJB invocations can hang indefinitely blocking both client and server side threads +I would need to force handleerror to get called to bump up connectindex +Both the captcha servlet and remoting servlet have now been incorporated into +Cool thanks Dhruv +I did spend quite amount of time and then tried the same build with Maven it is working as desired which is all the transitive dependencies of the dependency that is declared as provided scope are not ending up in the lib folder of the WAR +Yes this is not very but it would be nice to know maxminavg verticesedges per partition exposed in the master logs for when we experiment with different partitioning schemes +If we want to encourage folks to use the API in place of the API directly then we can deprecate and subsequently remove utility methods in that make it easier to use so that it becomes mostly a set of abstract methods +resetting P open bugs to P pending further priority for bugs with a status from ASSIGNED to NEW for consistency +Chunhui is a flaky testSee are lines longer than characters in patch v +Can you elaborate on why this may not be a good idea? +I will fix them todayI bet the pb is also occuring in so I will fix it tooThanks for the report +Thanks BJ for the patch +Thank you for quickly tackling this! +Thanks JB +The linked issue seems to be linked at least by looking at the kind of parameter passed to the method +Alejandro I understand the of Hoop as a proxy +Regardless of the solution we should build a self containedtest case and not rely on a telnet server to run the tests against +Hi BJIm not sure of the reason for the problem youre having but note the following two points +This work is now being managed as part of Tools. +Please review +When des it happen? are you sure youre using maven? we need something to reproduce the IS a type in the xml +We already have a mechanism to toggle any cache to start eagerly which will form a cluster on startup +It might make sense to rename YARNAPPLICATIONCLASSPATH to something like YARNFRAMEWORKCLASSPATH? Any name suggestions welcome +the JBM destination metrics which Farah has seen getting updated +u admin p admin s localhost query jbossserviceinvoker +Added test cases to test library deployment loading and unloading +lgtm +Thanks Aaron +It would take a huge effort to write source view that uses WTPs DOM and provide an adapter that keeps the two in sync +. +After messages are sent the master is stopped +heres a patch that makes the test work in reproducible looks good! +This was added to remove the confusion when a Web application was deployed as regular Java EE application or as a WAB +Scratch my last comment +bulk close of all resolved issues. +It works well on branch. +The exception is expected because the code tries to use an after it was closed +and failures are known issues +Great work Kaz on brining this module bang up to date +Applied my patch in Rev +Like Shawn we use one or more fields as the business key in the equalshashcode so that we can use a set both before and after the objects are persisted and not worry about their identity changing and breaking the contract of a set +So we certainly helped not pollute cache by taking in cold data we also effectively removed hot data as well while compaction runs +We cant the eclipse artifacts after modifying them because the missing certificate from eclipse +Unfortunately I see the same error after applying the patch +Thanks Sravya! +Will submit it when that component is available +Yahor exactly +Nuke this setter! +Tested given example on my system returns with messageEncountered at line column +Part of a bulk update all resolutions changed to done please review history to original resolution type +Good point I was having a very similar thought when writing the patch +spaces and style with last comments + +It sounds like youre asking for instructions that say if youre installing JBDS from update site make sure that +An amendment or revert and commit of an update that restores the differences in how Delete writes out the family map versus others could fix the issue +Thanks for your feedback Steve +The source file of the program that tranformed you +Since this is a large enough featuring will move the current commits against and their child task to a separate feature branch and merge them once complete +Committed to trunk and. +Is there some reason a more targeted fix wasnt chosen? Only the first line needs the ! syntax not every line +My question is whether there will be different config items to specify the logical name for the rm? In ha case its and in case its? Or should this jira just use and not +Patch applied on and trunk +Attached the patch which removes the above mentioned builds and tests were successful and Ive committedthis change to the trunk as revision +Whats the downside of adding a format flag to the datanode and having format pass it along +Ive added validation code to maven itself that should check for this setup error and should present a more descriptive error +Thanks for the feedback +Christian I got some prototype code working now +Please either debug further or provide a test maven based project +Thanks for the patch Anirudh. +Okay so the problem is not the retry per se but when we have a split decision on the nodes that reply +In the new way you only iterated it once +Will test on master and submit a separate patch. +I have not followed this closely so correct me if I am way off baseAssuming does all the number mojo needed in lucene should it eventually replace the existing number implementaionsIn solr would it make sense that intsintfloatsfloat etc are all implemented with? Obviously we need to keep the existing field types for If this is true should we deprecate the existing Number implementations for? perhaps justShould changing the schema version to trigger using the classes rather then the classes? Becides supporting existing indexes is there any reason to keep the solr number formats rather then the Trie version +Still the problem is presentIf there is only one server there no errors everything works fine +defer all issues to +Ive attached a test case with the same issue running under cxf +Both examples work on my machine with CVS HEAD +I think its ready for trunk +All unit tests passed +that the test actually passI am really thinking that this issue is related to a environment any idea how I could write a more complete test? +Bug has been marked as a duplicate of this bug +Added artifact information in the you cant replicate it I close it +Here ya go replaced LOGFILE with stderr throughout and removed the option parsing codeUnfortunately I wasnt able to really test this patch in a cluster since I dont have a secure cluster setup based on trunk but ant works + +jQM waits for domready then manipulates the content to give it the mobile styling and then fires a pageinit event which is what you listen for to tell your app that jQM is ready +contains both the new and the old versions of the replication scheduling algorithms as well as the benchmark to compare them + fixes nt exception handler to return before trying to instantiate java exception objectThis fixes hangs onIt also seems to pass smoke cunit and kernel tests on Windows XPHowever Ive just seen the test crash with an assertion in GC which hints at some enumeration problemI need to look into it furtherPlease review the patch it also incorporates nt exception handler cleanup that was discussed recently on a mailing list +Can you check the bug +Resolved +So it seems it might be Mac OS X specific +Ill be more drastic why do we ever do this? ie why not just wait for the NN to do a checkpoint like usual +You may also want to check out or solrj does not deal with this functionality yet +Will enumerate +And with this patch we deal with deleted columns in isRevelant and then in removeDeletedTo be clear I am not saying I have a clear idea on how to achieve that and that might in fact not be so easy or even doable +Can someone please help me understand the relationship between this jira and +First Ive attached an updated Interop Grid +Committed v. +Closing. +I have an instinctive feeling that this may well be related to another problem Ive seen here but cant recreate reliably on MS SQL +Both Abort and add are synchronized methods so the intrinsic lock on the method should suffice given the new code in enlistResource +This issue has been resolved for over a year with no further movement +OTOH we dont want to be constrained in moving the project forward to a more sensible structure +I manually modified the patch file before uploading and mistakenly caused a conflict with +Yes +changing issue type to new feature +add security +Intended see comments by Sven for more details +the searchtime was around ms whereas the same query without the code changesBTW this doesnt apply to the first search +Woohoo! +Unassigning this issue for now +Though I am curious how the url monitor service is implemented +The conf object will be modified by the and the HA conf setting will affect the following tests +start from screenshots please for new and noteworthy +Temporarily installed the JAR to the snapshot repository on +I dont see how you can do it without use of a stored procedure or with use of two statementsiBATIS already allows you to use simply the solutions +I am assuming by default lenientValidation would be falseregards Malcolm Edgar +Part of to port +Nice work! +predictors duration protocoltype types numeric word categories features you help to commit the latest patch to trunk please +Ideally structurally similar and functionally correct +Sounds like the fix will also fix what this ticket is about. +However no longer uses it just uses the actual charactersI think I can make this change as long as no one thinks Im going to break any browsers +Please resolve with Wont Fix and I will close the issue +I found the same as Tarun +Great thanks for that Tony +Thanks Billie! + has been marked as a duplicate of this bug +Andrew the problem occurs with the release candidate which is slated for final release soon + +Reopening in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +H SergeyNice addition for enum +Ill submit a revised patch shortly + in +The management server should have access to ssvm using rsa key no +The second variant does not require significant efforts and would provide basic type safety +This is a good enough +I just installed the latest NB and the initialized variables look fine but on one of my Assigns I was not asked to generate an Initializer and on another it kept asking me every time I added a Copy but then did not add the initialization copy +Is there an easy way to do thatAlso at the root of this is shared data stored inside OGNL using a static variable +I didnt see enough configuration information in that thread to determine whether they were reporting the same bug or some different bug or configuration issue +This shouldnt affect anything besides having a guarantee that the object wont be able to be used in a long I see it now the throwing of an exception does put a somewhat artifical limitation on using Timestampable objects +Regarding module samples you can +HugoTemporarily I have created my workaround in order to make packaging successful for ASF master +What is the status of this on your your end +we cleared the problem +The missing part is that the URL is not updated after the container is finished because we need to access the aggregated log via AHS web which is not completed +Patch looks good for the most part +yes this will fail the sync +Siiiggghhhhh +Feel free to close this issue if you want since Im not going to be coding any unit tests this week +I modified the code for unit test such that it takes timezone offset into consideration when comparing the test resultsThanks +This may be out of date now +If you have data on Postgis too you can create a view in it and then use pgsqlshp to export it into shapefile if the wfs request on top of postgis datastore works do a and ask for the output format +correction at rThanks Olivier +will commit if tests pass +Retargeting to and. +Ill try to work on a test case tonightThx +fixed in SNAPSHOT build +I reproduce and fix +This is s tough one to review +Defer all unresolved issues from to +Attaching the patch +Ive heard that request from some folks about having more cookbook examples +Ultimately it may simply come down to doing better documentation for the interactions between stemming and wildcards +Indeed the root cause cant be the same and we need to figure out what happens +If we do this it can break the compatibility since the old proto util methods cannot deal with the changes +Ill rephrase on commit +Ive updated also classes in SQL module that extend otherwise the build failYour patch was applied at repo revision please check that the patch was applied as you expected +So can someone pl +No modified added tests but my coprocessors are behaving normally again +All tests ran cleanly except for one which is unrelated filed for it +Thanks for a good evaluation of this bug +Patch to directory +Just a minor comment I would move method local after method main +Adding support for supporting custom headerDecorators in Sinks through this issue has been resolved by can it please be closed +It remains and continues to get fixes by maintainers at Yahoo! but all of the mainstream development is currently occurring onWhats stopping you from using at this point Radim? Does it not work on? Cause as I see it is now a subset of +Thakns for filing the bug here though. +This is not really a bug but an improvement to what already exists +Thank looks very pretty Christian +Does it make sense to push this toAlready committed there +Sorry but I not understand how can I set the swimlane to null in the execute method of the +Im not really sure how to do that since the service has to implement +Thanks for the patchI made sure to return empty map instead of null as that is what the javadoc contract says +I assume its still ok for my test case to execute directly as opposed to? It was easier for me to do it that way since that maps directly to the test in my +contains remaining tweaks in findbugs dodgy category that can be dealt with without stepping on the toes of previous patches +So since its going to break users compilation either way I dont its worth it to keep the old versionpatch close resolved issues +Something equivalent for Grails would be great whether it is an extention to eclipseeclipse or +The idea of that sleep time is to represent the time a follower takes to declare the leader dead +This patch has been applied to release and should be in Shindig trunk as well +Siabout is it possible that you have set the information for MATA and MATB in a facility different from the one in which you are running the MRP? In the first screenshot I cannot see the information for the two materials +Release the +Trying negation had a though +Sweet +Ive also ensured that the magic and magicdetector both have the right mimetype on them which helps with the sorting and avoids confusion when debuggingIve added a test that shows that you can detect a test excel file repeatedly without getting the wrong answer +Editorial pass over all release notes prior to publication of +Thanks KevinPatch apply at r +Work done for this issuer r r r r +I considered binary searching the extents but like you said the common case is only one extent +HiWhen is the targeted release date for this patch? Oleg +Unless we found there is a problem in farm deployment or there is a good reason to use this deployment we can this. + is still possibleThe problem is that you use +After restarting the server files are there in the tmp directory and permissions seem to be oktmpjsporgapachejspjsporgapachejspm using archiva btw but also happened with a previous version +The naming is very poor +I have returned my laptop and now have a Pro which I am more than happy with and does not exhibit this behaviour +You are asking me +I agree its nice to work on Hudson but I want to confine that to the test code + +There are even tests +Working on it now +Patch applied with minor whitespace diffs and doc +svn ci container Sending containersrcmainjavaorgapacheopenejbcorestatelessSending containersrcmainjavaorgapacheopenejbcorestatelessSending containersrcmainjavaorgapacheopenejbutilSending containersrcmainresourcesSending containersrctestjavaorgapacheopenejbutilTransmitting file dataCommitted revision . +We cannot add release audit notes to golden files +Ive reverted this commit until the library support is bumped from to +Attachment has been added with description fo file that fails property name is invalid +Did you restart the weblogic server to see if that cleaned up the issue? I noticed also that the URL in the error message doesnt have the port specification +I was wrong some of them are in so visible from the client +only fixes the the failing tests of the original checkin +has been released +I think it would be a good idea to also introduce the property on trunk +Ill update the ee build file add the dep checker for too and repostRegarding ee tests Im still not running them as they require so much setup and I havent been able to get them running on my local machine +Committed to trunk +can we close this issue +Tasks +Almost every jar we depend changed in that moment so it is difficult to know where is the problem +Well spotted Bobby +when the instance schema is not to the previous instances schema but that would perform poorly with unions +on discard unrecognized tag and use time partition in the archive key +The attached file contains changes to do a wrapper on the wsdl defintion object that is saved as a Parameter on an object +You do not yet have the fix for are you certain that this is not a duplicate +Argh +Could be the status of this issue has changed with the recent update perhaps Dan Diephouse could check if matters have improved +to see if get a thread dump every minute if stuck +Thats great +the feature loader doesnt deal with cyclical dependenciesrenaming views to and then adding a place holder so that existing gadgets that require views dont blow up should work +Good eveningIve just tried it +These menus have different ids +appears to be fixed with no further comment +Yes I can change the batch size as well +One thing I would recommend to update regular expressions for host names and paths in the spirit of andAlso need ASF license header +All reopened issues should be looked at as a part of and evaluated. +Thanks GavinYes you can exclude tooBrett it looks some snapshots are deployed to homecontinuumdatacommons and homecontinuumdatarepositoriesIm not sure we have to backup this snapshot repositories +Task is completely done. +In response to this there are other tests that verify the matching capabilities +Was this validation change reverted +I have no time to work on it so a patch is welcomedThe +Kris thanks a lot for all the improvements looking forward to +Tested with the plugin issue is not valid against it +To get the full functionality of semantic search currently we need to remove the default index through the felix configuration interface +CP fixed in revision fixed in revision +If possible may be we can have one test but not forcing to add here +Looks like there was a compilation issue in Hcatalog +and committed +Fernando confirmed that our patch process does not overwrite config files +Im using Weblogic as per +Moreover if this is caused because of an admin concurrency issue can we find some way to make it? IE is there a reason that we cant make creating and dropping tables atomic and blocking? If youre deleting a TB table I think you should be prepared to a wait a few minutes in the shell +and agreed about adding a commented line in setting the timezone to as well as a pointer to where these timezone strings are documented if anyone can find some +patch does not apply anymoreone wish +Committed to. +Fair exposed +Does the community have any suggestions as to how I can narrow down where this bug is in jrubyIf the stack trace were more intelligible I could perhaps be more helpful but it appears that jrubys optimizations are complicating thingsThanks in advance +This relies on the transport changes that are needed in +Applied in issues +In the previous comment I stated something wrong +Otherwise individual authorities will have their cache lifetimes conflated +its been deprecated in the previous version of Pig +Rick! Committed patch to documentation trunk at revision +Compared to other operations like checksumming this is just noise +to a patch release if we need to Jukka let me know +Otherwise I think its a wont fix +test with scm. +What does dpkg l and dpkg l firefox tell you? Does anything show up in the javascript console at all? This is probably due to the detection of the server ID being a bit dodgy +Administrators can then enable any plugins they want to transform the rendered commentThe end result is that it is now quite easy to disable html in comments and then enable any plugins that are desired to reformat the plain text comments +I am wondering what the community thinks of going ahead and commiting the changes for switching from statement to prepared statement for get and creating a separate issue to find out why the code is not getting used anymore after the commit for went in +Sure thing +Thats perfect Ben thanks. +Simple testing in Felix seems to work fine. +Fixed within the SIXPT branch +Sorted for choice merge but recusion currently has other issue +So set the dialect explicitly +I thought i remembered giving it my but I guess not +looks clean and tests pass on +Just to clarify by history I mean historical data +they should only be run in the continuous integration environment on databases that support proper isolation important part of this task is the systematic approach +There are many lines in that are over characters +The good thing is that return addresses on stack cannot be merged so if we come to already marked subroutine the return address is already in local variable +The problem seems to be that uniqueness checks were being processed before the full set of changes was processed +Thanks Ben so everything is ready for deployment? +Do you mind if we move this conversation to which I created to track progress in publishing artifacts for? Ill respond there +Committed to and trunk +Then the next thing to check is that your XDT builder is actually using that runtime jar +Bad bad me I forgot to mention that this bug is against the associated patch file was generated within the directory +Its all due to the same underlying JVM bug as far as I can tell +ThanksThe patch has been appliedPlease verify +Would you mind renaming the generator and lib directory to cglib +As per Jochens hunch fixing the compiler automatically fixed the joint compilation issue. +When it hides it needs to shrink it +The first thing I have in mind is a static map containing all loaded dictionaries with some kind of unique identifier so that the same dictionary can be reused between coresBut my question is is there a mechanism to share object between cores in Solr? Is this the first time someone needs to share something between multiple coresId like to hear your thoughts +I would prefer to have a click facet on tccolumn +Teds proposed patch for trunk +I just realized this is even worse than I initially thought we will have to create a disk for each instance in a group +Ive taken the liberty of resolving this as its not really a bug just not usage. +Can a committer consider this for or before you roll it? Itsglaringly simple and is a needed addition +Please post page and bean code and reopen the issue +What appenders are you using? Maybe you have an exclusive lock on something that is still being kept after the restart +If a browser fails to handle a particular encoding you should upgrade it else throw it away + for the patch + tests manually +I was trying to stay away from enforcing the order in which the user calls the static configuration methods but maybe this is unavoidableIt does not have to fail fast it could fail when its trying to create the input splits +I dont understand what this patch tries to solveAlso could you be more concreteWhats made you not confident with the current implementationEverything is configured from one single file instead of an hidden file embedded inYou can take a look at conf Ive put some info on how to use it at the same time of the JiraExampletmptrueThe UUID flag will add a timestamp at the end of the filename +Ive deleted the old and replaced it with a new ve redo the patch to use +Simon Can you be more specific +We now have a welcome page +However it seems the persistent hash map package is under GNU which is not compatible with apache license +Created issue for upload +Unfortunately it is not possible because the parent component will always be so the clientId will be the the same id +As it turns out recycling pools does not result in memory usage reduction however can result in instable behavior as some synchronization objects are allocated out from these pools +Added the projects to a fresh installation manually +One problem found was a field was added as a static field yet appeared to need to have different values for different instances. +Cool +Applied the path to the trunk +Added unit test for this fix to the patch +should be in its an important setting there seeems to be some duplication between this class may be some stuff could be shared +Getting the artifacts in maven central is important +Note that for I hacked in system property replacement for +bq +Yeah only PK columns are required +Closing all resolved issues due to a successful release +Tools doesnt filter anything out +Feel free to if you still experiment issue after issue was not resolved and is a bug and is continued in +Fixed in svn revision thanksdims +If you have timers in the same context they really should have unique names in my opinionEasy fix Ill code it up tonight but get more feedback before posting anything +Ive reopened the issue +A datanode restarts removes all blocks from the tmp and blocks directory and moves all blocks from the blocks directory into the real block directory +This is really a checkbox feature of little use and not very showey + +Reopened for backport to +It would be possible to be more exact by having calculate positions and append multiple entries to the index if necessary +What about my previous comments about changing the default in +Hi WillemI am seeing this with the I did a hard reset on my branch and I still see the issue +Ideally we should not set status if param is OK +I did not understand though from your comment if I should do it differently and post another patch or is that a hint to how someone can still apply the patch +You need to return instead of to get the step scoped bean automatically registered as an item stream +Hi GaryI am using jdk you have multiple in your classpath? If not could you check the version of you are using when launching your program +I am using perfdb as a db server at it reaches more than CPU saturation +Final change that was committed +Added myself as mentor but havent republished. +This can be closed for now as the Jackrabbit API is provided over RMI with using the latest Jackrabbit RMI library snapshot. +I have reworked the code hover code select and code navigation to use a completely new infrastructure +Were they added simultaneously +we discussed at one of the first meeting to include just in examples +I do not know what caused it +Wiki page ticket should be closed +It show that if you have the error with a single log file out of youre still paying a huge price in terms of added delay +Since its abstract and I cant find anything that uses it I think thats another issue altogetherThe additional compiler warnings I cant find them anywhere +It worries me if likes of a Laxman gets confused over this so I think the commentd be great Mr Hofhansl +I guess Ill wait for the next patch to work on something +We should also put in place some code to make sure that it is really killed if it doesnt respond to the graceful shutdown request over and use kill if it doesnt +release note as since this is the file name which the release notes generator looks for +This is not ok +All your points are valid +Duplicate of. +Problem added after fixigI think you are correct but we clearly have a missing test hereIs this functionality exposed enough that it would be susceptible to a unit testCould you write such a test +Incorporations from Clebert now in guide +Bug still present under IE +example of application that doesnt work because of the is really the only separator we can safely use on any filesystem for separators and because it is not supposed to be used in classnames i dont think there is anything wrong with that +Please review it +When processing idempotent messages they cannot be processed twice +did this get documented? +Or assign to proper person to perform merge if we shure this fix will not cause any problems +This patch contains three files the public review javadoc it is missing one method but on latest javadoc it is there. +My fix may have fixed this problem +Logged In YES useridIll have to think about whether this can be supported +Ultimately it depends on what goes on in blackbox of GC and phantom references +Some code is asking for Traditional and TWO services are marked with that +This issue talking about rwdebugiter wraper class that used in debug builds in all container classes when RWSTDNODEBUGITER macro is not defined +Thanks Ravi. +I will look why it is not copied in brew occurences of was renamed to this affected only and was fixed in Im removing the release notes flag +end to end test which wire SE and BC together to handle external onewaytwowayFault request +It may be as simple as you say but I remember the impl being super super finicky and it took months of changes before people stopped having issues with sleep +the new file needs an apache license comment at the top +Mostly looks good some comments The removal of apps from causes a memory issue we have to remember hrs worth of events in addition to what is remembering +Assigning to original ll have a look at this tomorrow and try to get it done before the release +Oleg +Authorization has been established to the SVN repo for Todd Greg and Chris +this doesnt make any sense why would a war contain source files in the first place +Committers seem to agree that the proposed addition should not be added to Commons Configuration so I am closing this ticket as wont fix. +Any idea if this with be ported to the branch +You must allow Tuscany the priviledge of reading the properties file if it is to succeed here +If you were to pull the code out into a utility class we could make use of it in the mailet +The XSD schema file in needs to be updated to include new option change in the XSDAnd for that we need to run the generator and checkin the updated source files +Not quite as easy as the command line I suppose +Hence this issue is invalid +Thanks for reviewing the release notes Knut! The release note for may be correctly worded because Kathey says that when she ported the fix to and she reworked the patch to remove directory creation +Committed at subversion revision +Thanks Knut Anders I dont know how I managed not to include the semicolons in the patch +Good job everyone! +Code that demonstrates the bug +Oh sorry +Pushing to Teiid the first item above by requiring a CREATE permission on the targeted schema +This was due to the fact that the GET request processing method is deprecated in the axis version and the alternate method introduced by Axis does not fit to our use case because Synapse should get any request without looking for the operation to be present or not +Closing the issue now we never got a response from A +ambiguous target might be because of a space in the remote repo url instead? +Is there some configuration we should change before running ant +Cloning this so we can document it with Datasources +Instead of restarting the whole cluster you can restart just the job tracket using restartDaemon with role as JT else the code looks good +Discussion on list feels that this is outside the scope of dbutils +Manually tested! +Committed revision in branches +sorry +Fixed in r fixing. +My tests resulted in calling progress and the progress was reported +This patch makes sure that the initial block report is sent once and only once +Fix throws an NPE when using Embeddables + +The userid and password in hwi is just a nice way of naming the session +Closing. +Previous patch had a minor error fixed +i think this one can be closedThe datepicker is for and completely replaced and i dont think we will do much about it for anymore +Moving into the CP fix in revisions and +Can someone provide them in a patch +Update the patch based on changes in +I didnt know that either +We should also add the inverse is +Ill work on implementing Kens Parser in Tika and get that end going so we can integrate laterCheersChris +uploaded second patch as should also be picked from stack instead of defining in web client metadata +I will wait for word from Shelly and James before I check them in since there may be another solutionUpdate per IRC discussions Adding the dependencies to the POM files is not the best approach because this means every application that runs Arquillian tests would neeed to add the dependency +Now the keyboard shortcuts are available again. +Im not sure what am I supposed to do here +Should I break them out into individual plugins so there is one per file type + This issue is intolerant as we could not delete the membership hence could not rollback the previous state of the Membership Management component within a session +Spring uses antlr +After playing with JVM options a bit it looks like XX helps to solve this +You are right +All tests ran cleanlyCommitted revision +And WTP doesnt know anything about JSF tags in xhtmls +It will never be removed +Why do you want to have both implementation to have the same result? As you suggested RomeVelocity is so simple +The test +In project export the project into a jar +Patch applied at r please verify +Tarball of Makefilespecfile for building activemq rpm +It looks like the month is incorrectly updated with the value of the minute field +wont be good enough for the big grids without this +Bump up version to +This works jettyxtomcatxThis fails jettyxll look into created to keep track of the enhancement needed to the m plugin +Martin can we not just replace it with If a bean which declares a passivating scope typeIll check on the rest of the wording +Good work on cleaning up the type checker. +Patches to suit the project to +Currently checksum is not verified when a block is read to serve to clientThe checksum is verified on the client and failures there are reported back to DFS +Refer and +Closing this please open a new issue if the GSE is the way to go! +Why does the default setting affect the import +Attribute removed for dataList is absent +I tried to reproduce this with a testcase but so far after doing few thousand iterations of clean add commit check no successStephen can you provide a testcase that show this error +The attached patch changes all uses of hdfsConnect to hdfs and hdfs to hdfsThus since we own the instance we can call hdfsDisconnect and no longer leak the handlesThis requires either or +This mirrors the MVC namespace and for more advanced cases extending should be possible as well +I would say that is not exactly the same thing you are using the fact that the model is shared under the covers and renders correctly +I also changed and interceptors to not overwrite an existing Date headerThe example no longer removes the Server header it is not a connection headercheers Roland +Thank you Junjie +I think is locked lets target this for the release and get it into the stream once its open for general again +heres the before and after sqlsvn executed onceSELECT TY FROM HOLDINGEJB t LEFT OUTER JOIN ACCOUNTEJB t ON LEFT OUTER JOIN QUOTEEJB t ON WHERE executed onceSELECT FROM ACCOUNTEJB t WHERE ? optimize for rowCORRECT svn executed onceSELECT FROM ACCOUNTEJB t WHERE ? FOR READ ONLY optimize for rowexecuted timesSELECT FROM HOLDINGEJB t INNER JOIN ACCOUNTEJB t ON CCOUNTID WHERE ? FOR READ ONLY optimize for row +I think it will be easier to understand if we saidTo lock a table you must either be the database owner or the table owner +Ok I will tell Peter to change it +Failure case attachedThere might some incompatibility with checkstyle and the configuration but Ive been using that configuration for several years without issue and did not see any items in the release notes for about incompatible configuration items +Here is a patch +Suggested alternative. +problem introduced when fixing in r +Thats better than an ugly stack trace +Glad you were able to reproduce it +I have not really tried it but I added unit tests with a basic coverage of . +m reducing priority on this issue as we are awaiting more information from Matthew +Basically dispatching to the task component and getting notified when task get completed +Rev and add support for jndi configuration in moves around jndi configuration in openejb so that the jndi enc context in openejb is an immutable federation of the global app module and comp context constructed by geronimo and an additional writable comp context that openejb can bind stuff like and into +Its not a bug in our code but the driver +Oops not sure how I missed thatI agree the fields should be made final probably best to document that sync +added css to push the dropdown over to the right on the sub and project navs to align with left side of highlight bubblewaiting on qa for a refresh +Here is the patch for the sameRegardsAmit AmitYour patch is in trunk at Jacques for committing my patch +This can go to installation needs to be added in the SEP entry but only in prod its not necessary to have it on the qa server +bq +model jar compiled with java table that can be used for the uploaded a better error message to the pop problem happens if classes and jars that are uploaded to Guvnor are compiled with a newer Java version than what Guvnor is running on. +Candidate for SOA jdbc guide is in think we can include this in the SOA Getting Started Guide dont really have another place for itNeeds review to determine how to include it +coverting character set seems to have resolved it +The only problem I can think about is people parsing the log for some reason but it should be easy enough for them to fix their parsersscripts +If the type flag is given the resource that is specified needs tobe a resource type id +Options i have selected in Wizard are Generate both Asy and sny generate Server side code generate Client Code generate interface for Skeleton Data Binding Name xmlBean +Committed revision merged to for +no failures in this suite in last runs +to with version to QA +Initial patch for discussion +I went for Paulss solution and not changing the order the activatevalidateNew behaviourIf activateObject throws an Exception and the object wasnt newly created the Exception is now ignored activation errors on new objects continue to throw +Anybody working on this +Im not sure ifhow that would affect existing JCR applications however +I am publishing a new +Bulk close of old resolved issues. +Please supply an example of this +The change was to remove the synchronization on the statement cache from the removeStatement method of +Please also mention that Storage layer provides data streaming +Marcel you need to run grails primitive data types will always be problematic and are problematic even for regular Hibernate apps +Lets start with reading and do writes in a separate ticketWhat is the advantage of preadMemory over reads into? You still need to set direct mode on the fd right +Can anyone help to review this patch? And I will try backport this patch to hbase trunk +Pretty straightforward fix +Dont worry I have to admit that the attached zip maked me a little frustrated +Can you provide a couple of sample mail so that I can add them to the test suite as a proof of the bug bugfixOtherwise a full stack trace for the conditions would help understanding what the condition is and how to synthetize a couple of messages to prove that +Need to either fix this issue or workaround in DMLX +This is user data and if we lose user data due to resource constraints it is appropriate to have the task die. +Isnt this a configuration option +Closing issue. +Hihow can I send a patch with this feature request? issue will be resolved in the AS project. +Now use the ll need to use the workaround even after hibernate core guys will fix the problem as users could use older hibernate core jar with this problem +This looks similar to we need to make sure content stuff is still working on the lineTom you have content working in JoprJON right? So you should have all the appropriate facets implemented +done +Sure lets move it +attachment output of after the first failure of then I attempted to execute the commands inoriginally in etcufw ENABLEDnoI changed etcufw ENABLEDyes service ufw restartTried to execute commands inattachment is output of after executing commands inSorry I am neophyte to ubuntu googling to learn how to work iptables in Ubuntu +Thanks for the patchI noticed that there seem to be public methods named testXXX and privatemethods named testYYY and the private methods appear to be helperfunctions for the public test case methodsI think thats a bit confusing I think it would be a bit clearer to use a differentnaming convention for the private test methods to make it more obviousthat they are not test cases themselves but rather are helper methodsfor the other test casesPerhaps the private methods could be renamed from testYYY todoYYY or performYYY or some similar naming convention to helpdistinguish them +The info logs generated by the maven scm library will only be displayed when the logging level is set to debug +bq +It then sees that the remaining bytes in the buffer are fewer than the size of themessage so makes a call to asyncread to get more and the function end +Can it be that the waveid can be found in some other way without using? +Something else must be going on here if the parent region is in the set of regions to balance the split message must have been missedChanging this from blocker to major +I just committed this +If this is not OK please fix it or exact tell me what to do +Closing all ResolvedFixed issues lacking a Fixed Version +If you can create a test case that shows the problem please reopen and attach. +PS I assume participants is null +Not bothering to this Issue because describes it betterRegarding the comment above the does now include the referrer uri. +common or +The executors should only specify the ARC slice they want to execute and use the top level searcher to do the searchesthis patch is just for illustration purposes +Cool thanks for the confirmationCommitted to and trunk +I agree with Claus on this one +should be Correct and in a new minor version release +I have since moved on from the project and am no longer in a position to test this personally. +Actually I reverse my position +may be some problem with QA? +Yes +I was looking at and found that this is slightly opposite to this issue +Patch updates srcjavaorgapachesolrhandler and fixes the facet count problem +on checking in the patch seems large since it removes the generated java files +Joining the voters encountered the bug while trying to use Glade with Boo +Submitting +classpath and +Just attached a patch that catches the logs the failure at trace level and then continues to pickup on any more matches found in remaining javascript +It will be surprising if some tests break without that change and might indicate some other bug +Im attaching the new patch which will take care of the closing issue and drop functions +The only bug I see is that host is not set correctly +I was able to workaround by setting false on my connection URII think the root cause that the xE bit mask should be x in the marshalunmarshal methods but Im not sure exactly how this code works +Sounds good! I think I understand what this capability will do but the ticket description is somewhat vague +Aim is to just stopstart to move to +The only way to have a unit test for this bug is to bump the version of Avro that is used while doing the junit tests to version fix for this bug is very simpleIn order to have a unit test that actually fails without this fix we create a dependency with upgrading the dependency with avro to version I leave it to the committers to decide to just do the fix or to do the fix upgrade to Avro and add the unit test +cant fix an already released release +I will drop s the benefit of having all those converters +Tested! +Andre if you already have a +current status of the +But Ill revert that changeSo two things I need to do revert the changes to those tests add warmer to +New version of the patch Fix misspelled words +Agreed it may be less cumbersome to change the API doc so that it documents the real behaviour +Applied at revision . + to file URLS always have response codes +Right I dont want to add a separate proxy repo for each sourceforge snapshot repoI think Hudson uses mirror settings which overrides any repositories defined in the +Best regards works fine +Problem should be fixed in the current HEAD of the CVS +Surprisingly no the CRC files stay at least on my boxYes it is cleaner to wipe out the directory +There are people using current codebase with jdk wo issues +It includes the followings +Would I be able to see the from the +Forgot to menttion that the diff also includes a fix for where the timeout option is not being converted to a String +Could you attach screenshots please? If its confidential you can send them to me at AT +You should use your debugger to find out what you are doing wrong +Makes sense +If you like this idea you may want to just add the feature directly to your implementation in the trunk +Ill just update the docs and apply the comments above +Please change the existing constructor to use the new constructor +Also there is the fact that each update must be sent to all shardsreplicas at the same time which can be problematic when half of them are on a connection with high latency +Closing all tickets + removes unused code dealing with tokens and hashes to make the upgrade simpler +I think the spec is being a bit draconian I could do without those lines too +I went ahead and implemented this for you +Here is what I thought but I may be wrongSince Tools is just the aggregation of some components and we want to provide this theme in Tools I thought it would be fine to put it in what Tools adds to independant components +Do you jave a loader that wraps the input format + +Are you requesting the content stream and dont read and close the stream in your code +fixed in r +Will let you know how it goes +Testcase seems okay but when you work with streams please use try catch and finally +All the loops for processing paths including a try loop that catches so interrupts are always converted to exceptions +I was getting my wires crossed thinking this would be locale specific not currency specific which is OK +pushed in Developer StudioVersion Build id Build date to add newandnoteworthy after having added newandnoteworthy label +Binding to does not fix misconfiguration +This should fix the issue +Let me resolve this +waiting for an answer and test turorial modifications. +Ben you talk about a splitter? But the route you post dont have any splitter +Make this a separate bug maybeAs far as solutions go I dont think that the filters is a good option it feels like a hack +supports both syntax +At least from my current personal perspective this would be a good solution +Still not resolved so moving down for further investigation. +The issue has been fixed and is marked for inclusion in the milestoneI do not know the planned release date but we should keep our eyes open and consider upgrading when the update is out. + +set nulloutputformat as hive output formatset option to omit setupcleanup for hive submitted jobs +My intention is to remove these and do a clean up well before the changes should be considedred even close for inclusionI am working on the Invoice creation at the moment and want to get posting to the GL before I come back to cleanupIn the mean time at least anyone enthusiastic can see the direction that the change is going +Works for me though I wondered if we wanted an editorial down the right hand side like a newspaper typically has +Moved to next milestone release +Ideally we need to figure out a a few things to narrow down the problem +If encoding information has to be specified it should be expressed via the connection URL as a property and not assuming that a application will be started with a collection of D options. +just tried it with maven +Did you do any tests with the RI to see what exception is thrown when a provider fails? Also Id be curious to see if it is thrown during static initialization or at some other time is the factory loaded at class init or lazily +The configuration change should be done as part of the SOA build process +Ill dig in and try and figure it out this patch is pretty useful for +Getting very close now deserialization is implemented too just need to weed out minor problems +I will make sure the rest of transaction is using transaction manager +files contain additions and edits to Javadoc for documentation +The note would explain the divergence from the ANSI specPlease let me know if there are any strong objections to this approach +Well make a separate JIRA for it +Hopefully this will make the link to it work in a couple hours +More of the existing test could have been modified as well but I wont do thatderbyall passed with the changes introduced by the patch +So although it is the same process the address changes +Please retry after copying the tool and see if you are still facing the problem +Now it has been backported to +works according to unit tests and the initial sample has some wrong code +We are putting our main efford into JAXWSJAXRPC tools issues may handled by community resolved issues for already released versions +Wendy pointed out the velocity site and how they incorporate dynamic content specifically rss feeds into the website +Marking as patch available +What is your concern +By default messages with a level higher than INFOwill be logged +Frank how is the new translation system proceeding +More work for the inputoutput types might be required but Ill create a separate jira for them +Committed revision +ON Android we switch to using our shim because permissions do not allow us to use the +why not just give it to me? I dislike methods that return an iterator because you can only do one thing with the iterator while there are other things that I might want to do with the set data +The Lucene implementation of is backed by an array of buffers incontrast this implementation uses a single buffer +set it and they will auto deployto compile easily I also set JBOSSHOME variable in with correctincorrent usernamepassword and EJB create EJB creation EJB Call +Can you attach the jsp please +you met the error you should undo your operation before you do any other operation +Heythanks for the new patch +This was fixed for Beta +Committed. +So do you want me to remove only the host option from the dropdownThanks +Thanks Owen and Brock +Only the following tests failed +Resolved in master +Oh Gerhard it would be useful if you could include the log for that request +I did NOT see it until I uninstalled the tomcat plugin installed the jetty plugin +Thanks a lot for your help and quick a lot! +If you could test that would be great! Looking to do a release tomorrow morning so hopefully youre around before then +The bug also occurs with the Apache app +After boot if a op attempts to introduce this error the op will fail with anThis logic applies to the interface attribute on as well. +Consolidated patch file with and demoing using Velocity browse interface +Fix going to be committed in a short while +Ill look into this issue +Done +The XSD if referenced in the WSDL with a relative URL +Attaching this issue which describes the problem with altering a table and using. +If there is property to be injected In for a component like the application will now fail because CONVERSATION scope is not active until after appears that JSF restoreView is now sometimes called in different order like it did in JSF version +does one thing and does it well and can be composed with SCM tools to achieve similar objectives with more flexibility +Hum can you also paste the WMS request youre making? You should never specify a style for layer groups always leave it fine in +Dave any news on this or should I bump it +Hi DoronIve created a new issue for some improvements to the formatting at +Dependencies are automatically handled via Bundler and the +Leaving as not supported unless the underlying JDBC driver supports the feature. +Hey folks +Now passes in jet due to fix in +Yup +Bharti is looking into this +I also moved the interaction between the two around +Duplicates +first check the PARALLE in query if not set then check the defaultParallel in if not set do estimation of reducer number +bash binnodetool should work +Also the javadoc should mention that the configuration options are only used it Metastore ledger manager is enabled in the other config itemsAlso theres a failing test in the precommit build and something went in recently which is messing with the patch application +I mean a CF ttl of X is useful only if it lets us reason that an sstable written more than X seconds ago is entirely expired +I will close this one and continue the design discussion on of +RyanNo worries +Close pending Jareks comments and changes. +But I dont see why this should prevent it from being applicable +Simply make the DBCP config options that cause an eviction thread to be created be and provide a way to enable the erodingPool decorator that will be in Pool +Hi Vicky will the tag for Common be available soon? +skimmed code nice docs and looks pretty clean +Basic installer created for Fuse GA still needs initial customized splash screen for release. +Thats why I used a separate issue +I just verified that this fixed both the fidbugs and the junit issues +Could you please svn add those files and the patch +If Im using the standard talking to a database for the user details service it doesnt occur. +At the risk of sounding prudish I have some reservations about this feature inasmuch that the architecture as it exists today only has communication +The problem is we dont have control of whats inside the compatibility test thats a black box and Im sure they are not setting project stage variable to test +readme contains deployment info and there are wiki page for demos deployments. +We just need a JBDS theme based on the by Red Hat branding +Libs exist when there is no variables in do not exist when there are variables in I understand the issue is not fully solved but part of it is fixed in +bq +Marking as critical +If A doesnt tell B that region is inactive B may respond to a failover request or something with that old said though agreed on not propagating the inactive region exception to B +What is interesting is that if I change the container in Karaf from Felix to Equinox the problem goes away +Components should flex in ratio with their boxFlex +any additional updates on this? Robbie what direction did you decide to pursue +Looks like the full solution will take some time so I will leave this issue to fix Jeffs problem and split off the issue when mixing super interfaces with classes having methods with default params. +Add space before With in With +If your fix works properly for you over time it would be very nice to provide a patch +JorgThe logic of stage is fairly obvious +I cannot reproduce the problem with +There is a second method for gaining metadata which is not compatible +Please give me your feedbackI like the patch even though it makes things more complex than absolutelynecessary +Lets run with increasing to m and watch out for hudson slaves that dont have enough physical memory to handle this +WaldemarWhat are the steps to reproduce this error and what version of the JVM are you running? Waldemar do you maybe have a unittest to reproduce the issue? I can commit that with the fix to ensure it wont get broken going forward +Thanks for your workPlease have a lookt at this issueIm working on many PHP projects all are linked to sonarIt would be very useful to use the eclipse pluginThis plugin has been made only for java projectsThankas sonar supports many languages I think it makes sense to adapt Sonar Eclipse also +And also to try and provide sample code that demonstrates the problem +Moved the blowing of interpolation limit over to its own issue +I can see how that would really help and what a pain it is to try to guess which stream were dealing withUnfortunately doesnt have any way to know the name of the file or even if the is a file +I am available to help out on this if there is anything left for me +a constant does not guarantee anything have a number of entities that should not grow extremely instructions operands nodes bytes etc all those may cause overflow of some optimization +All bogus connections would have to time out before subsequent valid connections would be negotiated +I meant that it is correct and in the differences I found were just because I was checking all files including those pointed by symlinks +This is a testtest execution issue and not a product issue so it not a release blocker +Verified at +It looks like the notification and response page is not coming up +But I get the same error if I setup up a simple Hello World application in netbeans with ant as the build tool +Is this gonna be fixed soon +Solved in the head revision +guess it was bit more that I bargained for +jar files in usrshareplugins will be added to the classpath +Discuss usage questions on the forum. +no more stuff. +Retested this issue with groovy and it is also affected + +Progress temporarily suspended while refactoring in metamodel is performed +bq +Any progress on this? I can pick it up if sorry for the late reply on this +For this reason Id rather we stored Date objects in the various implementations and do the formatting in the SOLR indexer +to the next depends on since a single config option for multicast port only makes sense if the multiplexer is used +Now I delete the last version and unzip it newly it works well now +I tried dot using and it doesnt make a difference +I managed to open it last week +If there is further review feedback Ill continue to address it inThere is a bit of a conflict on the merge currently because of a couple patches that went into trunk +Fixed a bug in the test +ok got jenkins to be happy +The problem with redundantly adding the repo to all projects is obvious it will not get automatically removed when releasing the projectad +Please provide a patch which shows only the changes in visibility you would like to dont put any functional changes in there as well as that will make it harder to integrate if you want to change functionality create a JIRA issue per change +This has been fixed in commit Jens in commit cfcadceffbedeeb it is now possible to use either or file separators when specifying an application to deploy +I have not thoroughly studied Heritrix classes you are referring to but I cant help thinking that you would be much better off decorating Socket input stream the session buffer binds to rather than session buffer itself +Will the same thing work for cdh for specifying different releases? +Its do only for svn that its the only one that store in scm path the tag of the release +that would be a problem +attached snip of to see if resolves the fixed with with issue +I see now that this system property is used by Lucene and not SolrJust for my understanding now Solr uses automatically but only on platforms is that correct +one day +At this moment six platforms have built successfully included among them some of the previous failures +I committed this to trunk! Thank you very much for the contribution +you can also avoid a recrawl by passing a refid of the same fileset to two tasks rather than constructing a new one in each +Will commit in a day or so unless objection +So if you push in fields then you dont want Roo to managed them any more thus you can modify the field annotations as you see fit +Now you should be able to edit port of a connector even its still online +Well track and resolve during the to the right JIRA component. +I dont think this quite works +New database introspect command now working +Dmitry I think it would be good if you could review the API to make sure that it fits your needs +Attached adds a test to w the fix as Mike proposed +Typo in the Summary should be Sitemesh freemarker handler sorry about as per +Thanks for Rick and Mamta for your comments +on the patch +From Jiri Pechanec +Koji could you try set to or min? Its good step in the right direction since we are going to essentially remove this timeout in +I just committed this +The behavior you are seeing is consistent with a flawed servlet container not a problem with Tapestry +Resolving as wontfix +Thanks for the contribution Alessandro. + for patch +it already works on also done as part of fixed and has been a duplicate to begin with sorry for the hassle +Thanks Shai +I also agree that a separate pluginmodule will be a cleaner solution +I would need the replicator in order to fix itI would advice the user to use Paging though +Oh and Ben or any admin Could you fix the title of this bug? ouput +Should print Servlet loaded from classes to when servlet class loaded from classes +One other thing if isnt set we should use hostname there are standard Java to determine this +Thanks Gabriel! +Thanks Todd! +Ron this fix will be in you new Remoting right +Adding as a migration issue in the SOA Release NotesThe Enterprise SOA Platform requires that JAR files in ESB archives are placed in either theroot directory of the archive or in the jars directory +The patch for a different problem addresses this HA issue also +Postponing it to is the same problem in Linux version of Chrome under Win has no such you please confirm that still valid and if valid check please if only for linux environment +I have to think about that +Agreed just pointing this out for anyone who expects this to be a simple cp from the output +Ciphers retain state so our previous name was not the best +It is reused only then old thread is fully collected +Obviously I had to modify some paths and names in solrconfig and log to obfuscate important details as per my companys security policy +If in the we populated the key rather than the value then one would not even need to specify by default would simply partition and sort Avro dataI like this +Thanks David! +Todd we plan to do that for all the jiras +This was on a standalone cluster in mode +All this sounds great but lets make it focused with concrete examples since beta is less than weeks away +Attached are changes necessary for the client side +Attachement internal and Index are pretty much the same although I did not look at Ross work not sure about citations but might be related as well +Added a registration check for andIf NM request doest satisfy minimum allocations RM returns +did you test it with or? Maybe this should be closed as cannot reproduce +can you check if that patch makes it faster for you? It never takes that long for me even in k runs with random to branchx trunk +Apologies for wasting your time. +Could I have you put together a simple unit test that recreates the failure +I think with this change you can specify data centers +trunk to +It seems like replicationFactor better expresses that its a target and not the actual number of replicas +Thanks AnkitYour patch is in trunk at r +I cant determine the cause by the exception Unknown node instance type +Well use a separate conf file +committing shortly +Thank you for the patch Roman. +Cant reproduce in or +in +Attached is a patch file for allowing for a domain less ldap user to be created +Both the cacheable and reloadable parameters can be specified as variables which will be resolved each time the module is called +Hans Ive also added a test case for your use case and its working properly with the current state of the plugin. +justbytes patch submitted for code review +Ill see about using for that too +Thanks David +Help me make it smaller or lets defer it +Enhanced version of previous patch +Thanks Dag! Ive adopted your suggestion of leaving the topic as it was and adding the production to the revised syntax topic +Is it secured Sonar +xml that can not be read correctly +Will be present in. + Datablock Encoding would using binary server in blockAnd I think it is impossible to do binary search in block for no Datablock Encoding or other Datablock Encoding changing the format of block +I also agree that users need to be careful that resources dont change between restart with any item reader but any help we can give them in determining what the error was is a good thing +I am closing his as a duplicate of fix for contains a test that will check the binary expression and the constant expression to have the right information +Perhaps this config does not make since any more now that reduce tasks can block map tasks from running +Enforcing process exit code with cmd c exit b n because exit b n does not do that +Bulk closing stale resolved issues +Committed with the partitioner range fix +same underlying solution will work just a different page to put it is fast +however the impl +preferably one without meclipse involved +This issue still exists in trunk +though I still have a problem with the javadoc +Thus I would like to know if this would be something I could use for Apache Synapse in the coming weeksthanksThis is a fix for Apache VFS to use a provider that allows writing of Zip files +Just a start but it think this is the way to go +note that the image loader in the needs to be enableddisabled +The problematic PDF document for page to image issue is solvberd in revision see for details +I was thinking more along the lines of how we do the current cacheing of design doc functions +I have just committed a few patches that would imply a lot of manual work when trying to apply your patch +The docs indicate that parameters can be bound to a CALL SQL query +trunk Marco feel free to try it on your all resolved tickets from or older +I also added three commits to existing test cases after the three calls and confirmed that the three tests failed for network server without the code fixI am running tests now +Nope when you doesnt select template it inserts an empty file instead of previous code the template selection button state stores in preferences from last wizard section +Is there a way I can create a confluence space myself or do you need to do it +Not worth fixing since Forge already has a event +In general bank accounts numbers are public and there is no need to encrypt them +Done +The failed contrib test is well known +Added description to the parameters +Please review and provide comments +It may be a couple weeks before I get around to adding it however +The cumulative fix is attached +The MR parameter that was changed is final and so the test cannot override it +Yes open a JBAS feature request to look at it at some point +Thanks for the patch DD +Am I missing something? I dont see a in the jrubytest archive +If we want it against the branch let me know and I will back port it +Nearly there +Closing all resolved tickets from or older +Also normally such pages do not get a high OPIC score and should be not in the top search results +If you have a code formatting configuration for Cocoon project please reformat the code after patching it and submit another patch which is formatted according to your settingsRegards this was copied from my other project I was working on by mistakeYou can remove it as wellKind regardsKrystian Nowak +made deprecationon the default in the build +If we go down this path we should probably consider renaming to something more appropriate that could also allow semantics like adding removing nodes based on changes made to the +Actually do it +How about Resources class? This has addsubtractclonemultiple etc methods that users will need to work with Resource objects +No path pbDid no test only reviewed patch +Given that the reactor build is not sharing state with the dependency resolution provided by the implementation the best solution is to change the releaseprepare goals to perform clean install as explained above +yep but this is why I said the we cant either revert the logic it would be as flaw as the original oneWe have to find another way to handle those cases +Confirmed. +This can be easily implemented as followingWhenever you try to add a child to an first you check whether they are from the same type of factories +The that I am using is a developer release from Apple +like i said it is fine by me +Just kiddingCan we document this and probably other scenarios that we may face this problem? +Agreed re breaking change +Patch committed revision +Thanks for taking a look at this issue +In fact this is really partially a dupe of and partially a feature request for sdecorate to have a new attribute prependId + for this approach +You are right +no response. +Looks like itss just not my day today I will fix keyRestriction thing and nits asap +Heres what I did +Commit r broke DRLVM build on Windows x I did quick fix at rPlease verify if the fix works for you +This means this might still be a test issue and not a bug with grantrevoke +Precondition is that every NIC should be properly configured +Laura please reassign yourself if you would like to work on this issue +I believe this is resolved so closing +verified by nick on hudson my local builds look fine as IP has changed +Luke could you please take a look at this ASAP before we release final on Friday? We need to ensure this works correctly given its a refactoring +Thanks very much +Im assuming something like a class that starts with default settings plus getters and setters and then passing those as the constructor args where needed instead ofI can take a shot at a patch for that +So Im pushing to get that release done first +bq +Grizzly definitely looks simpler and flexible compared to Mina +updated +Rebased against branch +Another jira should be opened to finish this conversion +ant pass +The spec is silent in this regard +I know that at present the Eclipse plugin is dependant on some code from Burrokeet but that project has voted to donate the code here +Picture a scenario where a user looking for road layers examines the capabilities of a WFS that just so happens to be serving VPF data +Its working fine +The attached patch ports the relevant code from onto +How you suggest we proceed +However its not under synchronized and its not marked volatile +Thanks Sushanth +Decided to resolve this issue in. +JeffJames should this issue be closed +Add the provided helo to the log if it could not be the provided ehlo to the log if it could not be config option to allow to not use helochecks for the option to check defaulting to falseApplied the code in the main HeloEhlo hanlders but I will work to refactor the whole handler things issue fixed in released version. +Please attach a quickstart app and we will reopen this ticket. +Uma Can u pls review it once again +Edited all NDA information +Solr is being released as we speak will hit the road in a few days +Woud it be possible to take advantage of this to further secure the Apache SNMP agent +Attachment has been added with description Fop patch for AWT Viewer has been added with description required by has been added with description New for fop has been added with description New class required by awt viewer has been added with description Updated patch fixed stupid fopexception error +Ill commit shortly +This task is currently marked Critical on which I understand were shooting to get out this month +Hi chenWhen you first released the plugin did you then check it out again? this problem occurs when your plugin is not a working copy of the SVN repo +Working on it +Preferably you would use the system to get the artifacts synced to central since they are well apache artifacts +It makes the migration very interesting some people report problems whereas the other dont see them at all. +beta reqiures the from betaalso features the first junit test case i ever wroteDthis is really simple code +The URL ctor doesnt throw an so is the try block needed +Yep +Closed without integration test covered by unit tests. +committed +Next build should include fixed provider name +How is FDALL going to scale to very large clusters +No companies involved +I will post a comment as soon as i have newsRegardsMirkoIs this still an issue? Any updated informationNo real updates or response in months +was committed with revision I disabled this optimization in all modes exceptPlease resubmit the patch that fixes EMT problems. +Thanks for the reviews Anoop and Ram. +Just tried to reproduce this against SVN trunk but failed to do so +If its binary like applicationpdf the filter should let it go do you think +a port for trunk +Addressed Chris concerns +Or perhaps have a list of aliases for names and +This works but its verbose ugly and brittle +Just a note that we get these errors with Maven Java Suse are seeing this bug consistently in the build +with JBT. +Isolation levels are only configured to be weaker as a performance tradeoffIn this particular case forcing Infinispan to give you strict READCOMMITTED for deleted entries will result in a large performance penalty for what is weaker isolation +With your release scheme you pile up any parent in the project you have released ever because parent version x will always refer parent +Modifiy this issue to not create new interfaces but to extend the existing method definitions to the Sling API in Rev +Sorry I was joking + +Any workaround +This seems valid the included code removes a +We would have had to make changes to the way slices were handled down to the level +Patch looks good +fixes for typos in to the are applied +Theres currently no itch to scratch here so Wont Fix this issue. +There is also a method that needs exposed on +nice +Its a shame cos I use these windows on a daily m now running F and the same issues are still there +The current API can be used to support all of the use cases I mentioned but in a highly inefficient manor +orgjbosswebtomcattcjasper +Otherwise keep the current logic +For a working example consult the updated class mentioned earlier in the discussion. +Updated with the correct patch +Erwin provided got the bug for in JIT off modeRTTI param value are updatedjp target instance etc as wellBUT it appears that the RTTI also have a reference to target and this and this one is NOT add to test suiteneeds to think is it goodbad to NOT bind RTTI to JP +No further response in nearly months +The value part here is directly the object put by the clientb Reading from data put by a Cache client that used a +same for is the bug described by Andreas fixed by now or should I open a new issue for it? +Actually the ui plugin should have both ui and doc plugin together +Pete ah I did a big typo in the bug report Before the artifactId was instead of the new now its fine I guess somebody must have corrected it after I asked on IRC +The machine I checked does have binls and works like usrbinls +Changes look good +Elle est dclare pour le package Utils +I really dont have clarity of vision for how to best handle adding Groovy to a project +I intend to commit by tomorrow +Well finally Its fixed in +They would be backward compatible +Please post when your patch is availableRegards Any progress on the patch? Im an engineer at Eucalyptus and Id be happy to discuss anything related to EucalyptusCheers no longer have an environment and the time needed to work on this issue +in the described scenario is directly removed and not through am using Oracle +What is the purpose for wanting this script? Why do we need to know who hasnt committed in the last X months +This probably isnt a big issue when dealing with region location stuff but you never know +I had this error when my frameworkVersion was set to but the stated vendorVersion and frameworkVersion asIn my case there was no error with the file but with my pom configuration and a more informative error message would have saved my day +Eric yes we are trying to move it to the jboss repo which should have plenty of BW +Could someone translate the hudson +This means I cant move the jira +Against svn +Maybe we can remove them or return a heap pointer instead of a stack pointer +With your patch you committed some changes to special characters like or in CHANGELOG file +Hello Would you please try my patchThanksBest regards applied at revision r with minor changes thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Applied fix at r thanks! +Pretty sure I fixed this in. +This patch applies against trunk with the patches for and applied +Hi Gayathri the problem is that were not yet able to understand the issue +So would you guys be fixing this issue or I should useI replaced with and I also got the correct results It failed with Appender as well +So the fix I applied was to disable the remote recovery manager when running crash recovery tests with the hornetq store +May I ask you to try applying that change to etcoozie on your existing installation to verify that its the same problemIf it is the same problem it wont be specific to Tomcat as we were killing the environment before making it as far as Tomcat +That should be OK for now +Although its totally possible to develop it in your personal repo and post a giant patch at one shot the bigger the patch is the longer it takes to be reviewed +issues for released version +I want to display a portal like homepage with portlets containing some personalised data for the user +Owen please check the design document attached in and the patch attached in its subtask for what have been proposed and done +Not really since the new journal based objectstore is faster anyhow +Using console when i deploy an application and it throws Exception when starting the the console still showsThe application was successfully deployedThe application was successfully startedThis patch catches the Exception and prints the following message along with the Exception in Show Details buttonDeployment Failed Application failed to startShow Full DetailsWhen user hits Show Full Details the Exception Stack Trace is shown +Issue was resolved several months ago so closing. +Closing resolved issues for already released versions +Merged with trunk +Does this mean the problems affecting sendRobust have been fixed +We need ZK trackers in both sides isntI dont think so +Agent will call setTagI dont see why we should build yet another metrics framework +Anyone could apply it +Tracking dependence on the filter parser Gabriel made the effort to port CQL parser to it would be nice to include this fix in on still need to port onto at r +It wont be on every request as until now +Yes it should. +default core is going like they where using it like an alias? We have alias support in now but the single node alias stuff was previously deprecated. +Moved log excerpt to attachment +Please verify. +The docs are talking about something else +I wasnt aware they could resolve the issue +please move back if they will be worked on asap for +attaching a patchdiff which adjusts the skipping of the test for to be only done with ibmNeeded to see the javacores with ibm +Turned out I had a bad installation of some of the npm modules +Just a question that logically follows from having a coordinator what happens if the coordinating process goes down or becomes unavailable +Patch applied +I think we should revisit this issue at a later date +Marking closed. +Though svn Revision is not working for my commitsShould I resolve this issue or keep it open to address more convention fixesIt can be taken care as we move along and refine the code +constructs +Im also missing having a useful DEBUG log level for pages so some solution is very desirable +Patch committedOleg +Sure I can show you on a hangout tomorrow +overall +If region has been closed by other thread just abondon the region should not be online again while rolling back +means that this is a block copy for replication +but here is a patch w the thrift changes bundled +They are working for me +Hiranya I dont think we need title cases here +This is finished. +Hi Jaredsorry for late reply +batch transition to closed remaining resolved bugs +description was added to a lot of elements +As discussed on the PR the intention is to port the legacy functionality from into the new Enhancer +Passes all tests locally +OK new clause added +Updating this issue to invalid as a part of bulk update +wed also need to add in locking primitives for all config params +Sorry for the noise +OlegSorry for the slow response +Of course if one was using good database practice one would beusing parameterized queries and not have this problem +If I had some suitable code I would have attached it +A is kept as long as the Thread is alive +patch applied +Project is imported +defer to +Could you do something like binext +fix out the log files to ensure they need to be there +Small name change suggestion +Confluence has switched to using a javascript syntax highlighter to format the content +Still a problem with Debian package m not sure how to proceed on this +Couple of things Do we have anyone who would directly make use of this? Without that these things tend to rot +We dropped support of AS in jbosgi +Indeed the test passes on Windows but fails on on how the external command is being executed the error is either there or not +Good spot Jason +Also encountered this problem and I came up with the same fix +Bumping down the priority since a simple workaround exists +I have a couple comments Thanks for the diagram +Looking at the code again +This was partially done in time for CR +This patch has the following +Moreover the language used in to indicate a bad version which seems fairly standard between implementations isnt intuitive +If editor is eager to do so Eclipse creates new editor input for this other file and puts it to the old our universal Red Hat XML editor can open etc +batch transition resolvedfixed bugs to closedfixed +I dont recall ever having to do this +Internet Explorer says that it cant show the when stylesheets are used +Sounds good Matt do you wanna take a stab at implementing this +Only conditionstocover and it +In such cases subclasses should take care to chose unique integer level values and to ensure that they maintain the Object uniqueness property across serialization by defining a suitable readResolve methodendWhat do they mean about the Object uniqueness property? And does that mean something else needs to be done in the patch +I would also like to be a moderator please +As noted in its currently not possible to fix this due to limitations on the st version of the protocol +Replace the Weld item in the Carousel with Central and add in the mobile browsersim +The interface is a better candidate for extending Serializable +bulk close of resolved issues as part of JIRA housekeeping for release. +You can ask them to be able to manage several classpath or just to be able to have a different order in the UI and in the projects +The annotation classes are apparently in the jar files +Would you please file a patch to complement the existing unit tests in wicketsrctestjava so that we can work on this bug and find a fix or provide new unit tests as attachment +Changing title to better reflect request. +It generates m records for the pages table which is too large compared to the available disk space on my cluster +I havent browsed in other languages so I dont know if that changes things +However I cant be sure that my description is accurate +Happily reading + Could be disabled in the admin section if the admin doesnt want it Could show the latest info from +Could you explain the error handling in the new scheme? The old code would call pad and that is now gone but I am not sure I understand how the error handling is handled now +Sure! Thanks for the explanationsRegarsLeonardo +References were removed verified closing. +Attaching patch modifying the log statement +I can try doing it next Sunday +The should not use the symbol but the a m p +The regression test passes for me without applying either of the other to patches +Ive just committed this +Verified at +Applied patch Fixed in revision +Changing the codec to figure it is what am also thinking +Ill make the fix with Andrew and cut a release today +Here is the patch to the code which I used successfully to fix a corrupted Derby database +Its just an absurd stance only engineers would design a system like thisNow that we are also becoming a viable option I think the whole example thing is just stupid and a long term huge negative for Solr +Now that were in better shape it should be less visibleThis said I feel that using a simple implementation without any weaksoft reference would be more efficient +Sending srcmainjavaorgapachecommonsfunctorAdding srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorAdding srcmainjavaorgapachecommonsfunctorAdding srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorAdding srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorSending srcmainjavaorgapachecommonsfunctorTransmitting file dataCommitted revision +Im okay with the proposal for supported but I would make two points +h as Introduction HTTP Services EJB Services Hibernate Related Issues Changes From Previous Versions References but maybe thats what you intended +Do you think this change will make it in forIf we are going to change behavior we probably best do it on minor release boundariesI agree and I hope to get back to this within a week or so but if others have the time to look at please do so +After review the patch is ready to go +How does one create a ticket with a user name? needs an id +We could but itd be a lot of work to do on branch +a running on the ml +However when I selected all in the list box including the Build Queues list heading I couldnt save the changes +Hi KimIm a big fan of belaboring the obvious +Following the steps in the description reproduces this issue for me in and later CVS builds +As expected thanks +committed rev +bumping while waiting on details +attached the log file with debug +apply to trunkmodules +Filed as +The split is handled by the sequencefile and not by the filesystem +Just to be clear Im fine using launchconfigs to implement this just saying this launchconfig shuold be smart by default so you dont need to adjust it in majority of cases +Please let me know if you have further questions orAshwin +Sorry +Wei mirror looks good again +I am not totally clear if this is something that exists already or is a suggestion for a solution +I am using bit Sun JDK on Fedora +okey so this is a blocker since that is not supposed to happen +version file we should create it in a tmp location and then move it into place +to work around for now I just dropped back down to +i agree with romain +Since we already have rules for valid user table names why not just apply those in the directory listingUser tables are not allowed to start with +committed in deecfbdfaceccd +Sorry for my oversight +I have made changes based on the comments and uploaded the new patch +bq +Were already asking users to synchronize their server clocks now the client clocks must also be synchronizedThe only gain we would have from setting it clientside is so the client can be aware of the stamp +I may toy with this some night this week +forward or +Can I have such a login too +Patch appliedThere are possible problems with not released connection to +Let me know when the Gant scripts and archetype are fixed if you want me to test again +Lets just load disk contents to cache with a warning if the settings are different that the setting is being ignored +Im new to alot of technologies here Spring in general Roo and JPA among others I guess +Thanks + +Hey Hari any chance you were working with an old version of the branch? This exact bug is covered by tests and should have been fixed by +That would be a very bad behavior +Demonstration of the issue +fix for JET compiler +Thanks Eli +I dont want to close this ticket until Ive sorted out how to do this correctly +Generate java code from wsdl is build time action it doesnt mean will build servicemodule from wsdl during runtimeDid you check logs for your client sideYou should see the output like buildor buildwhich will tell you how the servicemodel buildNo I didnt test it on or I dont have those jee container at hand and I dont think its a jee container specific problemHow you use client side code? And how you specify wsdlLocation annotation? If you can append a runnable testcase I believe it would be very helpful to figure things out +We dont support inverse classloading in at the moment everything in an ear is in one bundleclassloader and whatever we do end up supporting will be osgi based +Closing. +thanks again Ill make sure I help out my fellow fullsailersSincerelyMatthew RichtermrichterMDVBS +Same for us with RF makes this functionality totally unusable as you dont have to cli that quick to crash IE r get random exceptions +Patch for AM only +seems that libxslts method is just a heuristic based on the recursion depth which is needed anyway to also avoid finite but huge recursion depth +Tomorrow Ill check out whether the patch fixes +from me as well +on my linux box it does indeed work with and and fails with I was not aware of the jruby library until I ran into that issue here and reading about itthe problem here is not similar to does not work with newer rubygems versions since depended on mutex which newer rubygems versions do not require anymore +What I thought was not a bug is +Originally was for internal use only and was correct for all the methods used internally by the groovy codebase +You cant create a schema which begins with SYS so you cant create any objects in such a schemaij version ij connect jdbcderbymemorydbcreatetrueij create schema sysmineERROR An object cannot be created with the schema name SYSMINEij create derby aggregate for intexternal name ERROR X CREATE DERBY AGGREGATE is not allowed in the SYS schemaij create derby aggregate for intexternal name ERROR X CREATE DERBY AGGREGATE is not allowed in the SYSMINE schemaThanks notice the Documentation section of the spec mentions changes to the dblook documentation +Easy enough to add Ill do that fix +They can either be redone or a different workaround will be needed later. +thanks +comitted changes revision and in the installation instructions +If Ive misunderstood the situation I definitely welcome corrections +As a result of the the correct Exception message class can not be determined at the stub if two exception classes have the same element +Committed revision Thanks Eric +Same problem for me +fat fingersI have seen exactly the same problem with Mozilla too +To demonstrate the poroblem run in mvn install and then do the same in +Thank you! + +Im adding a small patch that adds a simple for any field that fails the instanceof String typecheck +Can you try with the edge version or at least with? We have location tests in our test suite and they are working correctly for me running with or the edge version of Cordova on iOS +I will test and commit if OK +I am taking this to see whats the status for. +I put this into trunk and. +Attached a patch to port Chuan Lius changes to +What are you suggesting an appropriate URI would be for your file INTRANETA class such as will undoubtedly have portability issues since its unrealistic to cover the cases for every platform in the world +Patch could be committed imho +Willem the binary operators should support type coercing so if you compare the LOCALADDRESS with a then Camel will covert it to String as wellAnd I do think the regexp will convert to String before matching as it uses String by nature +Done as makes sense. +Dan cant this internal API change be handled via reflection so it work on both pre and post versionsbtw +np +Does this patch actually fix a bug with long revision histories or just one regarding attatchments that nrstott was bumping into on IRC +No problem +Good catch thanks +Assume closed as resolved and released +Fixed in the latest in +Works great with our without server connection. +Will try to reproduce. +Its not clear to me what a committer needs to do with this patch +I think using one known value is recommended +But Id never promise to keep it that way +If there is any assistance i can get from the developers on where i might look further that would be great +verified at revision + +set to invalid so it wont show in changelog +Added paragraph describing jmx information about modified deploy that uses dynamic classloading +See +Thanks Suresh! +Thanks! Yes that explains a lot +We must be very careful about this and be as much compatible with other components +Added combinations for the client and server The derbyall diff tool was enhanced to fabricate an empty canon file if one doesnt exist in the codeline +These statistics could then be pulled by a report job later or just used by operatorssupport staff to understand what the job is this isnt just a problem with Statistics the same problem exists with Restartable +That said what additional functionality would these other secret managers provide +Ive also uploaded your patch for convenience +If you are an ant user chances are you have the older jpa persistence library in the classpath +Then do +to include default configuration with a default transport configuration much like CXF does +applied in r +Ill try to provide a patch +It modifies files and are uses of the regexp module in the mochiweb source +However given the fact they have been unable to make up their mind about the use of LGPL code in ASF code for years I would not be holding my breathWelcome to the wonderful world of ASF bureaucracyUntil this matter is decided upon by the ASF legal people I personally will not touch Microsoft specific code with a barge pole +The same is true when B send messages +In fact logging is there it just needs to be activated +test by. +So as a workaround I should include a temp file in every empty folder +If you think they are critical for CR please them +Committed the updated patch tx Zhjijie +The new attachement doesnt contain the java files anymoreHOWTOUnzip the attachment in you folder this will create the needed foldersunder srcblocksNow apply the patchYou have to move from srcjavaorgapachecocoonserialization into srcblocksitextjavaorgapachecocoonserialization and into srcblocksitextconf +The work has now to be done on the Taglist plugin side. +Update the patch The and are still be wrapped by thus currently we have to reuse the original hack +The test server starts +Look in the hadoop net code +Couldnt locate Jamshid in Jira who is the author for the very much Ive created a jira account +find the updated logging guide really cool +Other than that v looks good +So custom synchronization it is +You are right it is +Committed +Thanks Kristian for the review +If i left it blank i have an access denied you try it +This issue has been resolved for over a year with no further movement +Misread the query +Trying to see if theres a workaround +Many votes etc +So srcdst pairs +But need unit tests for sigil bundle generation to check format is correct will add new issue for this +On large clusters Im seeing bulk imports fail so the master moves the files and yet there are entries in the METADATA table for those files +Perhaps the new jar could be called +This could be an architecture issue +I remember we dont support volume migration in kvm across Zone fixing this NPE This happening because when unboxing Long to long and Long object is fixed with commit this on the latest build and it is working fine without any issues +So the token cache is blown away but the content of the table is still using pointing to the old token cache +This version fixes a typo in a target name +Managed not to set the grant button on the original +If you move these remember to move as well. + +I can then go ahead and check it into the repo +Its not strictly necessary because this bug doesnt exist in but merging this restructuring of the code to will make future merges in related code easierIf I dont hear any objections Im going to do this late today some time so please speak up +Owen Ravi thank you for the feedbackWill getWrt are we worried about this case? It doesnt seem like something that would happen in practice as disk failures arent typically recoverable and per Im thinking once a local path is identified as bad we should remove it from the config +Ive published it on the mevenides AU for along with the dependency graph module and the local repo browser +HDFS test case is failing due to +Workaround applied to the example config +The good news is that Im able to replicate the problem with a new integration test +The error is happening on the server the type of client OS and browser should be irrelevant +I generated docs for NMS today and uploaded it to the site dir not sure why its not showing up yet though +Updated NS configuration as per your suggestion still this test case failingTest case not failing while adding NS test case failed to raise exception can you please check again +The new JDBC data store series should handle properly this worth committing +jvisualvm or visualgc is a free program in Java +Good catch Himanshu + issue +This is broken as of Maven probably due to the change for +Fixed at revision +I had a chat with Jonathan I am going to revert fully and then remove the duplicate tags that were present previously. +Thanks Jukka the git repos got the latest updates now +Code changes look unchanged and okay +Thanks Brandon for offline review testing and bug fix +Im sorry for attach the patch so late +Good news and bad newsGood news This is not due to the recent refactorings I did to change to a comet conversation timelineBad news I know this because Im seeing the same behavior in the releaseContinuing to investigate +If you find a way to get it reproduced let me knowJeff +I literally meant types not values +committed +Parser still relies on reparsing without BP for it to obtain all outlinks +See last comment in could it be related to how the is implemented +svn revision . +Its successfully tested by new tests and running the JPA TCK +I modified that test and removed the equivalent test I had placed in my new tests +import modeler +When copying to target please let new timestamp be configurable I dont want my files to have a new timestamp even if they are filtered +Close listener socket connection +Ill create another ticket then +ive added the completion of properties the editoradding more dtd driven attributes to i think would be an option but changing existing attributes semantics are notassigning away since i wont be working on extending the concrete have the completetion now adding dtd specific elements will most likely not happen +Please review it and push to main branch +OK Ive rethought through my design and youre right +I was wrong about the client file it is indeed the version from FP +but when i see it in the morningconsumer is not getting any msg from producerproducer stopped publishingFollowing is the producer log DEBUG Broker Started DEBUG Simulator started +OK Jeff thank you for your supportJust another question about properties updatingThe cmiscontent is writeable? I can read it from the repository but what is the right way to update the name of the content stream after updating the content stream of the document? use the not JIRA for questions +Its hard to write a test to very this issue please test either or within your setup and let us know your outcome +Thanks MarkusYour slightly modified patch is in trunk at r +you possibly resubmit the patch using unified diff format? You can just type svn diff in your trunk root directory +So the getRowgetColumn vs viewRowviewColumn merge exercise is turning out good +I have to revert this and close this issue wont fix +This was either solved in the jbossws integration layer or thanks to the latest cxf update +Thanks Corinne! +There is no other wayBut my intension for this Bug report is the increasing without a reason +i doesnt make sense to start looking atyour request earlier +There is no need for this to be a fate operation if the op fails it can be retried wo concern of the status of the previously failed remote op +Sorry +We would normally encourage the use of withWriter which would handle resource cleanup automatically but I guess there are times when that might not be what is needed +we need build script change to run multiple outputs test cases? Are they any different from other test casesI took a look at that part of change in the patch seems nothing to do with the multiple outputs feature +Twas a simple fix +I have now seen this problem on a completely new project which has never referenced another so I dont believe that the diagnosis of is correct +This may have been fixed in a recent commit +Patch for generate rainbow color heatmap +Thisorder can be different across jvmsOSs unless you specifically program the order in +Just a week ago there was pornography listed instead of Apache and content stolen from other websites +know we went through this before but just making the volatile does not make the implementation thread safe +Confirmed successful local build +Im not sure how people want to approach this with regards to M +Cheersre still getting emails from this address +Test after will be updated in to M +Fixes to committed in resolved issues +Im pretty sure the is identical in and +I can write its details on a wiki but I dont think we should document it for users +Moving fix version to awaiting patch for this enhancement +The docs say it works on a basisFor requests I think this makes things pretty simple +This would be a nice feature otherwise I have to write a custom source to do blocking. +Trunk currently has mapred part way ported to the new API in both a not backwards compatible and a not functional way so I just finished the work along those lines so I could run jobs and put up a patchWe can do this in a backwards compatible way +So Harmony doesnt have such issue now Probably it was eliminated with fixes to other problems +Found this behavior back to for forward only result sets in result sets were not updatableFrom and later it applies to scroll insensitive result sets as well +You might want to add a link to the forum and clarify that you want to be able to invalidate duplicate user sessions +I just saw this +So have nobody an opinion to this issue +Should it be mounted somewhere for it to work properly +Thanks for the update KarlLet me check the felix version used in Pax Exam and see if an update works +Unless you mean our own client would be the sole way to access Stanbol and that they would hide away RDF making it impossible for instance to use RDF tools on the enhancement results then having our own clients would just mean that some additional clients have to be updated +Another last try +It seems I identify the component I state this is VPE I can reproduce it only after creation simple empty file then VPE open it andI have got the screenshort picture +Vinod for failed disks did you mean to say and not ? denotes invalid container status whereas denotes container aborted +Good question here is my brief digressionThe code was originally written with an model in mind recommendations happen in +This pluggable patch is not include any new unit test but run against current test +Now we both worked on the same task +It is even better when there is a plan for further progressI say ship it +Alright thats really a separate issue and its bothering me less over timeIll commit as is without further input +no problem +Fixed in CVS HEAD. +Test looks good +Ill look at it tomorrow. +The new delegate should take care of this now +Thanks is really cooking these days! Massive increase in bug turnaround since the old days +Committed. +The error page could do with being far more specific +Is there an expectation that the jvmRoute wont be configurable at all in the future? If so I understand not adding a config option for it to our management API +I think the has to be set as well +Could you also post the changes you have made in the persistence manager and the complete configuration you have used? Just to make sure we are testing the same thingRegardsThomas +Because of the way the Grails test code is structured I had to copy a lot of stuff so its not using the same codebase +Emmanuel Venisse can confirm this is a problem using IE +Thanks for contributing to. +Fortunately we can all help the real developers with this by testing the requests and reporting about the errors +Also for faster reads data extra copies of data could be done in memory storage tier thus facilitating caching of the dataThis proposal is very high level +buildclassesthriftcassandrabin +Release notes commentPreviously remote clients using nested transactions would get undefined behavior as this was not supported but no nested transaction checking was performed +Updated version of previous patch for review + +slip John let me know if this is not true +textedit box color is changed to white possibility to easily edit the page names is added +Joe thanks for working on this jira +A might be generally useful +Thanks Jingguo! +cannot reproduce on last buildit will be uploaded for QA on +DIGYI have not done much with the Highlighter code or any of the contrib code in a while +What I was doing before was to have my annotated component use the to pass the domain object populated from the call into Mule but with Mule the doesnt pick up the from a static variable but rather it needs to be passed in via DI +Oops couldnt find this one before +Committed to trunk +See linked issue for configuration files +I assume it works within version as well as the base functionality of the class was not changed Please consider using this patch +Committed revisions closing for +I could be wrong though! +Can we disable the buffers? Just the keys for the merge shouldnt be that big of a burden on the jvm +Your second jar happens to be first in the classpath. +Just to say it worked really do appreciate that SirJan G +FYI working on creating a patch against +Can another committer take a look at this patch? Id love to get this fixed soon because its broken for people trying to use smallinttinyint columns +If you apply the same sequence against trunk with no branchmerge you get the same failure +writing a manual scheduler that may or may not do slightly better in the short run is not going to change thatwhat we want to do is slow compaction down so that instead of short violent bursts of CM activity you spread it out over a longer period of time +I guess the dependency of ivy cache in tests is the cause +Attached the log for sybase +Thanks Stepan +I updated our SQL Server JDBC driver to the latest provided by Microsoft to see whether it might be something related to the JDBC driver that could have been addressed and we still experience the same issue but instead of it blocking between and it blocks between and rows +Applied the patch +Marking resolved +Thanks for your comments +Adding missing check +One other note in terms of backwards compatibility users handling their code with onSave will not have any compatibility issues because the same order is preserved for the case of context value client valueThe only users experiencing issues will be those who implemented onAction from the component +The patch from Jochen solved the issue I had +Heres an untested patch +I think it is important to have a tokenizer that does not try to be too smart +Committed as r +Yup lets try without the repackaging + +Interesting am I correct when I say this means the caching was not working correctly for the internal Freemarker configuration? That would certainly explain the massive difference in performance we noticed when rendering UI tags +also has this have you closed this issue without applying the patch? +This will be as costly as a transaction commitThere is a chance of a false positive on the redelivered flag if the consumer fails after the delivered ack is persisted by the broker but before delivery of a may not be necessary to track updates in the index in the event of failure we could always rebuild the index if this feature is enabled as it will be a rare is fast recovery vital in this case +Probably a fix similar to is feasible +I think the site docs will also need some revision +Providing a fix +Is really slow? The dates are in JIRA and on the roadmap web page though we are about a week behind and probably there will be another week slip for final +Could you please provide me an information about how this isssue can be fixed or how can I get round this issue +Seems like a previous Hadoop RPC using the same bucket writer taking too long can cause this issue because several threads end up being blocked on the same method +It should however work in most cases +Patch attached +If it goes in we can resolve this issue against hbase +Patch attached +so far it is working sweetly I will report when I finish my tests +Replaces previous attachment entirely +merged by on EAP release notes for eap +Otherwise assertions will not be triggered +Plz correct me if Im wrong +Each mapper could also take its ascribed range of keys break it into subranges and randomize each of the subranges +A new patch remove synchronization +Maybe commit this and then look at moving all ad hoc Writable versioning to use +I guess this is good to go once is out +If the main method body is wrapped in method test doesnt crash +I know that I can branch and I will +Note the build instructions will make use of jars which result in a compile error based on not being present +The ER release has stopped until this is the package to its former self in version and trying the upgrade from there +Ill enter an issue there instead +Should be ready on Monday +This is corrected for the next release of JBDS and. +Is there a reason +Thanks for guidance. +for internal central repository cleanupcorrupted poms causing major headache +update looks spying I did on trunk didnt work for +I have committed the new config file with update property definitions. +If I just changed the file and try to save iterror will pops up but cant see the detail message +We havent received a new JOPR console by component freeze datte moving to next cycle +look a bit sad on Though if you want really fast dates chosing hourdaymonthyear as precision steps is vastly superior plus it also clicks well with ranges +Youre way ahead of me on this one +As a result if the executorschedule hybrid fails the job wont finish +And it imports the wrong versions of +h seems to be failing because h is not accepting new connections +No local mapping +This prevents invalid months effectively +Assigning to Brian as he was addressing a similar issue reported by customers today +benoitHow about we match the rest of the file for formatting? There are zero cases of indentation matching the argument column +Ill try to translate some more components in days +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +A lead party could be associated using the using the lead to model thisIf the requirements around associating leads to campaigns become too much more complex at some point we might want to move those to a new entity +Didnt put two and two together +Ken can you elaborate on thisWhat exactly is not working +Take the patch using the single commit on the branch as your source that single change set should be equivalent to applying all three of the trunk changes in order +log WARNING JSF Unable to find or serve resource from library +Do you mind opening a separate ticketThank you +Yes I know how the method works +I think you are mistaken +Alexey is this essential to know which file cause exception +patch +Let me know if this is not the good way to checkprove the memory leak +For example it means that deployments will have implicit access to if exporttrue is causing problems in your case you should just take it out which is exactly what you have done +I should add that is already included in SOA this has been fixed on the ESB side is there anything that needs to be done on the platform documentation side to pick this change up? +Thanks +The patch passed all unit tests except in my machine +Fixed for + +Patch fixed a bug and added more test cases +When you report back Ill commit +Does the remote service send a Connection close header? Can you post a trace of the response received as see attached file for proxy services request and response +It doesnt look like this is going to make it for +For me is about inner classes in general +Closing. +I do get that point +disregard the old to account for converted class to use Assert class rather than keyword per Pete the patch +No problem just keep it in mind +Personally Id prefer changing defaultOperator in example schema to AND but Im fine with OR there if browse gets fixed +Maxim have you fixed it already +Thank you TonyVerified on. +Hum since WFS is referring to GML schemas what are they using to deliver GML? It seems to me having the bindings alone would not do much good wed also need whatever version of WFS uses that GML +See +Anyone know this issue +I dont think its a good idea to simply change the default wire format to javabin as things stand now because is incompatible with Solr serversThat incompatibility only exists when you try to use the javabin request writer +Ive reverted the change only +In this case every message should be sent with its corresponding superstep number to avoid race condition +I just discovered other resteasy examples that have dependencies that are not covered by resteasylib +Looks great to me but lets have a think about how this miht affect users of Castor JDO that upgrade to +assumed to not be targeted for move to +Regression tests passed +I ran the test times in a row and I did not see this error again +Ill create a separate issue for this to be fixed in the next version. +I was trying to update the logging lib in hbase and noticed that there four versions of slfj being pulled in by our dependencies and that these four versions cannot be replaced by one only as they are incompatible +Works on IBM SDK on and trunk +the patch for this slightly modified patch in r. +Contrary to my earlier statement there is no easy workaround possible +When adding the first solution to the Im wondering what Index value I should provide since a value of represent the first step +If Mahadev agrees I will dupe this issue to +Touches the following fileM javabuildorgapachederbyBuildCommitted to trunk at subversion revision and ported to branch at subversion revision +Java has been a requirement since Hibernate. +To fix this we may be able to clip the region for which the WMSC bothers to fetch tiles to that provided by the? I think I want to make a map graphic which displays the valid CRS area for the map +Patch attached +Maybe its simply very slow because network is overwhelmed by all hose move? Are you running on real servers? Or on +For example tomcat support in it contains more libraries for seam project in war then for as server and ad special archives with configuration files for embedded wee need +seems to have been fixed in trunk +. +Resolved as fixed for JackrabbitSee for the request to add support for properly importing such base encoded values. +the fix moves the error checking from constructor to the setters with minor twist +May I apply the attached patch instead Miklos? It reuses code that checks returns out of meta scan results elsewhere? Thank +I think some of the handling of stored prepared statements is causing the memory usage seen during the junit tests +Hi JochenTo answer your other question No code was copied +Yes I wonder too +It would be great if we could get some input from the community about whether it is desirable to create a quartz component vs simply updating the current component to use Quartz +See +I can install thrift tomorrow and create a patch for +this seems to have been fixed +Use bin instead +Thanks Tom HDFS builds correctly now. +I am building with Visual Studio +rob looks like we are adding and as user entries in the launch configuration that intended? Shouldnt these be part of bootstrap entries or somehow be in the user entries when restore default is pressed +We should not make a release with all this stuff in the javadoc since the published javadoc determines supported +Type and message definitions in the same order as defined in the avpr file +Also I didnt see it mentioned in the perf doc but you should use only the locally attached instance store volumes as datanode storage volumes to avoid variance introduced by EBS +I think you r using build +An excellent solution would be to support loading the from the classpath + think for this one I prefer to wait for Uwes refactoring of onThen mmap is simpler and i think we can even use the same indexinput implementation hereThis would mean no slowdown when searching CFS +But it along with several other configurable properties is not in the documentation +Hi Chris thanks for looking on this +Thats nice FlorianI think people in the project who can sign within the web of trust is nice if you all agree we can close this one too +Everything else should be OK +Fixed. +Fixed. +Today I think there are only clients in python +If you want we can also portcopy it to the book or have a comment in the book about looking at the package info +Im moving all bugs out to and we can move some bugs out to to get some release action going +Attached patch that fixes this applied in trunk r +Well +And the new python native driver isnt official yet +Currently there is no patch available for this issue +I have a fix for this and went back and did all the same tests I did using the with the HTTP +If a block is in pendingReplication it means that chooseTarget has already been run and a Replication command has been send to datanodes +Seems reasonable to me +Is Derby planning on providing Japanese translated manuals in PDF format? If not should we close this jira? +I understand that the size is important for this map +Fixed configure on IAFor use +I suppose you changed some you now everything works perfectly. +Yes +to will check the hot fix +Using these excludes might be required if a SCM system keeps the original file extension for the working copy files +Hi Craig +Now we have a patch for and for trunk +We can supply a small library to generate the PB headers for others to use +You need to remove the argument checks +if execjava aspires to mimic a invocation of some java program +BPEL editor is now on extras staging site +Different connection strategies could be implemented as instances of this interface +It was caused by an incorrectly defined return type for a function definition in a custom dialect +Jeff submitted a patch based on Satheeshs comments +committed r + or for notificationsshall I use your user as the project admin +I committed this to branch. +Patch +Please verify and then resolve and close +Thanks Koji +An RTFM is OK of the manual exists somewhere I havent found just tell me where +Lets see if there are more comments from othersFor its ok to resolve this in +The error displayed +We should update the javadoc with that info +Looks like the plugin installer is attempting to write to the output stream after a response has been committed +The error message has been shown as below + is a test script which demonstrates the problem +Already done +The release note of also applies to this issue +Its a good idealWhy not displaytag use it +Thanks Virag same failure is to trunk! +EmmanuelI appreciate the fact that still shows interest in logback despite the licensing issues +I have attached a release note to that issue. +Should not be hardlong +True but we are talking about notification or a callback here +Maryamthe text got corrupted +Logged In YES useridOKI have attached an example of a custom principal in thisfollowup +Im pretty sure it will fix it +A separate issue should be filed for implementing a way to generate the English names for the parameter markers in the doc +One side effect of previous patch is after we run ant test the we get contains no hadoop jar and is not runnable +Further we no more need the recreate boolean variable because now it is provided through a method and every time we use it we need to update the value of by calling that method so we can directly use the method inside the conditional statement +Andrew took a closer look at the code after Riley brought this into our branch +Added to codebase needs some tuning for perf +psWe did apply direct connect patch from Mr Schuller in order to user directConnect +Patch to properly synchronize the abandoned trace patch as before but no whitespace inclusions as the previous applied +I say punt on this +Finally the solution has been foundD +I should say its an existing a week ago I was testing Rider +reopening to fix status on close +Fixed on and closing all issues that have been in resolved state for months or more without any feedback or update +Same for app schema. +Thank you for the review +Id close it as wont fix +The addendum was added to avoid many of the restrictions in the original spec license +Thanks Oli the patch was applied as expected +The inputStream needs to be closed before a new response is read +Yep we do +The latter would be better I think except if what you mean by forcibly terminating by the Erlang VM is that a separate process handling javascript is killed in which case theres no cleanup to do +Accommodate spaces in paths +However Im going to close this bug as invalidDan +Please verify this causes no harm on other platforms +The same setup is working totally trouble free on a Continuum installation on an Ubuntu vmware image on a Windows box +Please verify + +Wow very fast! Thank you +Is this still openThis improvement request has been open for years and has stepped up to implement it +I agree +Would be great to be able to install camel and cxf without changing the +This issue is fixed problem was with the postgres and oracle jdbc driver implementations not handling properly within the context of a Hibernate typically does not flush changes to the database right away but optimizes database access by flushing data to the database later during a these two drivers if not flushed right away result in loss of the BLOB dataSolution Manual flushing is done the moment the BLOB data is insertedupdated or removedSolution tested onPostgresql g please use the forums for such questions. +Logged In YES useridThis is actually a issue +thank you for the URLBut in my environment seems to not work FIXI tested with IEI also built its own Wicket jar did not workAttach a project in Eclipse WTP that I used for testingIt has its own jar file contains +this looks good +This page would extend the same viewtemplate of the serviceinstaller configs pages +Thanks Nick I dont know how I could miss these +The annotation I proposed in uses counters to report on runaway udfs that get killedI think the question isnt so much why would you expose them as why wouldnt you expose them +Or do we want to let our users set their artifactId +For sure you need these dirty bloomers exposed +Changes to spec +Thanks Bill! +but it is not +Testcase showing the Unbound object is the Embeddable and is fully bound by the Embedded reference to it +Added forktrue to for the integration module +release notes done +Class I used to backport the functionality to do batch updates with named parameters +We could make that optional leaving the current behavior as default +will get control during web app deployment and set the context as a attribute +Thanks for your contribution +Thanks Army +Yes Im using the native protocol +Works in Mojarra and does not appear to be browser specific +We have been waiting for the fixes on the MMC and since they have been open for quite some time we have built a MMC component +I just committed this +Closing per above comments. +I noticed that has been released +Im trying to make it possible to allow end users to run appfuse without further modifying their system +OK and me given all the commentary +The HTTP Server implementation might also override that method and redirect the logs to the Log Service +Pavel can you try to set up local SSL w a ccm cluster based on Garys instructions to verify +Please ignore the previous comment +See for further find attached patch as mentioned above +Hmm Im not quite sure what to make of that table +no tycho uses the in in JBDS +though? +If this issue is still actual please specify the version of EclipseWTP and then reopen the issueAnyway its certainly issue of WTP +Okay sorry I didnt realize there was a delay before new commits appeared on the server I was pulling from. +With code drop I removed a server from AI portlet +is out and it looks like the should be using thoseAnd no the was some split between and so some of the jars that are available in the were not there in I found them and I updated my project to use the jars and all is well +done. +Toru Along with the query processing also add more info to response like highlighting and more information +Not invented HereAll right +ok then the compromize for now is to get rid of windows encoding only to replace with for now +pushed robs metrics update +Both List and Set for or relationships will cause compile error in generated gwt scaffold complaining about can not resolve +Patch committed +My patch wont work well in since it depends on other patches which are too big to backport +SAAJPassed Failed Error Not Run Filtered Out Time Failed Error Not Run Filtered Out Time Failed Error Not Run Filtered Out Time Failed Error Not Run Filtered Out Time resolved issues for already released versions +These both abuse the fact that their results are already resolved. +If installed properly everything works like a charm +You can increase stack size to kB by specifying Xssm as a JVM option +Because and are just methods of the services +differences between and the current error from the list of sure if there is anything to do regarding the this can go into a waiting state until we are ready to do a for restoring CICA compat and I think youre right we must continue here with the next RC +I would personally prefer a before the lines to improve readability a bitThis is probably a matter of different style +I opened and attached the logs from a failed attempt there +Id love to see this happen as well +Chris You cannot modify the list of moderators +reset to clear buffered you were to say that its the same concern and invoke this mark method at the same time for reading and writing you need to either mark when a transaction begins perform a noop write on an empty bufffer at the first mark flush the writing buffer on stream close? somehow that last buffer has to get written +Can somebody please suggest how I can get to the underlying logj object +This counter is incremented is incremented for every iteration of listFiles when listing files under a large directoryThese statistics are collected in job history for analysis of how HDFS is loaded by map reduce tasks +not needed as itll be GCed if not used anywayIs it possible that the classloader referenced by cache eclipses another classloader which would load className class correctly +Once optimized there doesnt appear to be a way back other than causing the softreference holding the call site array to be GCd +Once for my scanner and then another for the actions of believe this specific situation has been solved as it now polls the remote server for a list of currently scanned locations before adding so as not to add a duplicate +I updated the main README to emphasize the use of Maven for builds and removed Ant from an additional ten projects there are still some legacy ant projects and ant references that will continue to be cleaned out over time. +Committed this to trunk and +I think this should be considered a bug in and +This issue is fixed and released as part of release +Do you have a central repo? Can you remove this from the local and central repo +In this case streaming app might have to increase Xmx value or make no prologue by none +Fixed in this bug down. +SimonKeep in mind that this additional header would only appear if you asked for it via a request parameter like true +Eclipse version? Eclipse XSDecore plugin version? What elseSomething must have changed +Is there a workaround available for this issue? I really need this functionality for some SEO +It really depends on the intent exact version like family version like propose to interpret as family version or a plugin depends on the exact version there is something wrong +such an array should be deserialzied to either null or empty array +on the new patch. +Woops sorry +To be honest I do see a couple of timeouts and attempted rejoins but they all eventually succeed +If anybody has a better suggestion for its location please let me know! +ThanksI just signed in at your JIRA because I liked to recommend my idea of an logj logo +Would it make more sense for DIH to farm out its content acquisition to a library like Droids? Then we could have real crawling etc +Library provided by Tools +I also have a plan to test it against ether Grizzly or Havana +The switching of logs should also serve to help balance our log writes +Note is removed but are the same thing a developer would need access to so the issue is still there +for hudson +AXIS +Ok slight change of plan +Dont know how to upload the files +Thank you for your nice finding +William Ashley same no progress on this issue? +Yes thats much better and to adding to Commons IO Ive created a Jira ticket so that it doesnt get forgotten +works finethanks for me +Christian here is a patch for +Siying can you give an example to show why we need to restore the jobtracker +I have also found that if I trigger a full GC on a shard so it becomes unresponsive for a brief time the searcher that is coordinating the distributed search logs the error in communicating with the shard then the rate of increases dramatically for a time before recovering +bq +This has been done closing +Onto +Sorry to come in late the patch has gone stale +Until now would happily ignore any CDATA making it impossible to specify HTML as a property valueNow treats CDATA as text +The ability to have the correct start day based on locale is important if wicket should be used with real world users +My first inclination is to try and match the root cause and then all of the exceptions it extends from before moving on to the next exception in the +Used wiki for spotting this its fixed now +OK I commit that one +Im also glad that it worked +it will take me too much time to get to a new schema that includes descriptions +This bug has been fixed in Xalan Java +Go commit +Looks reasonable to do for the next to work for me +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I think we can unload from the cache using and +Because of Java Element is validated each time a String is validated the number of calls to validation of Java Element is reduced +Documented as a known issue with workaround in the release notes. +Patch for parent is mvn eclipseeclipse needed? Cant you just open a from Eclipse? Would make more sense as maintaining the eclipse stuff is done by Eclipse itself and not from eg Camel +After this patch has been applied there should no long be any warnings when the previous testcase is run +is fixing multiple issues +This is just a standard technique for managing dependencies +Hope that solves your problem +Actually it was walking two hierarchies at the same time both the inheritance and the page containment one which is quite ve been using the patch for half a day and found out that the containment walk is actually necessary so I reinstated compensated the loss by adding some extra exclusions in the resource loader so that trying to access localized html page or properties or xml files is not even attempted in the resulting time ishome m attaching the second patch +In revision committed a web extender that can actually deploy web application bundles with simple servlets and jsps +Bulk close resolved issues +Just raise a warning? Why and what should it say +I fixed to get env vars +I just provided the parts for the container module yet as I cannot build the whole project due to test errors +Hi Alan Ive been testing this failure on the following databases DB Derby and Oracle +I applied that patch in Rev +This patch includes all of the embeddables sample code +This was my first shot at fixing this issue for an IBM jvm but I want to make it more generic as Myrna suggested +Attaching the revised patch. +Thanks Doron +Reverted at subversion revision in the branch and at revision in the trunk +There might have to be some check if the should be set or not +also we would love your feedback based on your experiences there +I am attaching here patch +Until then I think that is fine +Committing shortly +Not for commit here +Enabled new test and added patch to fix itBut we should still consider if is really necessary. +Tests pass +Can you propose that to Lance +Id like to find a way to incorporate this in the short term like with the maintenance release +Is there a possibility that something could be done to either adapt this functionality to respect sorting before collapsing +Fixed bug need a test +set is a duplicate of +I merged the two classes from contrib because of this +Ported from trunk to branch at subversion revision Ported from trunk to branch at subversion revision +Note that there is a reason for the current directory structure we need to run tests from inside Eclipse +But on my local machine at least I do not have original problem of ant clean package failing +I think Gert fixed to set the TCCL as well +Can you create a test case showing what you are trying to fix +Should itCan it be changedIs this the same issue that Im having in my code for a component parameterabstract method name +Allen fixed this +Thank you Nicholas. +Tested in my build environment + update +RajeshIt was my question +Amongst other things the Jandex index can even be saved to disk in which case on following startups no reflection has to occur at all +finished enabling all the services and created an internal wiki page describing what to do on new channel setup +Hey MaxId rather just implement source scanning the proper solution imo +Wish there was a mapreduce for ll commit still seeing some code inspection warnings but we can look at it later +Yes +Thanks again +Many thanks for addressing this annoying issueId vote for the second approach indeed access modifier does not do its job here +This is enough for geronimo +Currently BTI supports only one SVN repository per test suiteThis patch implements the BTI support for multiple SVN repositoriesWith this patch it is possible to specify several SVN repositories on which the test suite dependsThis patch also updates hdk test suite to use this new feature +Ill check into this and post a fix +People could have done the work with their lucene hat on or their solr hat onThe history behind this particular patch is that it was for a customer using Solr and I scoped the time based on thatIts simply preposterous that an improvement to Solr be blocked because some might want a module too +Found the problem will post the patch up for review after running some initial tests +I think this patch has gotten too complex + Ive observed that when result is displayed using +Once we have the spec ri and tck suite under control and if we have some runway we should try and make it simpler for new comers unfamiliar with the toolsetI am thinking of creating a mega bundle that contains Maven the Git repos people are expected to clone etc as well as maybe a few scripts here and there to limit the numbers of stepsIf we can avoid scripts though Id feel better they are hard to make portable across the gazillion platforms +will now report any exceptions generated by the Translators. +I dont see this happening in either FF or youre correct the line breaks were caused by whitespace in the xal file which textareas dont ignore. +I tried to reproduce the bug but the Grid was loaded correctlyAre you sure you are including all the dependencies in their correct versionI had to remove most of your beans because I dont havethe classes and resources but any exception during the initialization might disrupt the grid creation +I seem to have broken something in there for windows +that was one reason for having the keylets take a look at that as well +Thanks! +I dont know of anyone actively working on this +Is there a workaround by using any instead of a to retrieve information about the messages? +If used as a datastore for services that temporarily store many records such as a messenging service the disk space used will grow to be much greater than the amount of data actually being database administrator should implement database maintenance plans to ensure that unused space is reclaimed +HII think this would be difficult because I suspect that the concurrent modification exception is caused by one thread writing to the registry and another iterating +Oops +Add cd NFSPATHTObin into bin script and you wont see any error related to pwd issue +I fixed it accidentally in +Newbie to Axis and are unable to proceed +If anyone wants to work on it feel free to take it +Please refrain from making invalid conclusions in the Description of a JIRA at the time of opening +As a side effect one test case was modified to catch instead of +I used the one Ben wrote for the old leader election as a basis +bq +Ive just checked in the change but JIRA isnt giving me permission to work on the issue to mark it as fixed +Moving to R since I suspect the solution is +By default provides a list of all tables since people often want to publish geometryless tables as is a black list of tablesviews that should not be published geometrycolumns among them geographycolumns should be be closing all issues that have been in resolved state for months or more without any feedback or update +Thanks my bad +Right I havent seen that this was indeed implemented as a task and not a datatype +Uploading a revised functional specification that adds more details about upgrade privilege handling and dependency tracking +the patch includes fix for source AWT doc and resulting AWT page on website +Any updates on this patch? The last one was in May any chance this will be committed soon or are there outstanding issues +I just need to figure out what gets called when the save button is clicked so I can put my patch there. +please check again the formula in the merge part +Closing old resolved issues +So I added a finally and always close the stream +Since the File EPR never really fails the entry in the cache wont be removed +Applied patches in r +Resoliving as Out of Date. +This is because I believe that withouthaving to consider item turnaround time would just be a couple of do you think + +I sure can +If its not its a featureOr are you asking for that featureIll have a look later to check the sample + to trunk +Thanks Mathias. +Yes and hopefully you agree we cant fix it since its upstream from us. +For requests the methods take multiple parameters but we wrap them into one +Looks like a deadlock scenario can be reproduced this isnt specifically the agent unless the agent is doing something silly +I think that most of work related to annotation and webfragment scan are finished. +We could also use this layer to do protocol inheritance +Seems reasonable to flush the memstore if its had no write activity at all in N minutes + +But I did not find how to use it +This has finally been done +This is done by the class but it has to be done for the tooLet me check that +Just log in on Xircles and ask to be a Cargo project contributor and let me know +Please if you see it again. +would you mind trying it again with the latest? Weve fixed quite a few bugs in the past or weeks so if were lucky the issue may no longer recreated this problem the steps are stop the app wait for a success message to be displayed then start the appso far its for so it is ll try reproducing again +When I ask for the top terms using for our document indexes it takes about minutes to an hour +To have a deterministic sort it has been implemented programatically +Moving from due to feature freeze +some changes to be made in the patch +Sorry for my delay in responding +even when the constant is already a double? Wouldnt that result in multiple decimal places +I would say it is incomplete that it does not update the namespace prefix at the same time +Please attach eclipse projects required to reproduce this issue given description does not provide sufficient guidelines in order to reproduce this issue +Even changingserialization to svgxml and using a SVG browser works for meSame configuration except OS I have Win k +Daniels branch has a annotation that I believe relates tosupersedes this +Can you give me a precise set of steps that will result in such an exception +Maven try to update MAR plugin first because its SNAPSHOT +It does look possible DEPENDENTIDPROVIDERID could be OPERATOROBJECT of new system table +As pointed out in the performance degradation is significant +Team is working on finding a workaround for this. +Actually before I jump the gun on this can we verify this is the same issue +Ill merge it into branch +Verified on a cluster writing a unit test for the same +Best regards George IBM is an updated version of the contribution zip which has been pruned of some unit test code that should not have been included in the original upload +We run nightly builds that update our website documentation for developers +Ive committed this to trunk and merged to. +Looks like they mix up some xml regions during editing +When is +Should be able to knock this out by mid Monday +bulk defer of features from to +I attach it as an aditional test for the bug. +What does Maven ask for +here is the zipped file of the foo project I generated +Or if we have directories we could make those readable only by that user providing greater privacy +Done. +The CMIS specification provides a mechanism to get and apply an ACE but only if referred to a specific object in the repository because it is a protocol +resolved since rev . +First time contributor figured this is a simple chance to go through the motions +If I look at my logs I see nothing between exchange and the beginning of the following supersteps +The link will be fixed once the old Forrest docs are converted to APT and the website is regenerated +How about this patch to fix the sort order? It should make distributed search lex sort equal to what one would get for search +Fixed this issue now too +Neither in +I am working on a fix +Feel free to reopen if you see it again. +If one connection per user we may have many connections if there are many users +Incrementing the full vector of new values is definitely the safest way to do it +Can you take a look to avoid duplicate work +As far as I can tell this has to be a UI issue there are already explicit tests for updating the builder configuration +Concerning this is just a recommendation there hasnt been done anything in that scope +Is there any need to retain +Hmmm +Hope thats OK now. +Im guessing Andrew is suggesting we add this to our list of hbase metrics +Patch committed as is +If the view for the iFrame URL has the type URL then it will add all the Javascript libs for this view to the URL +Id echo Andy that long term we should refactor checked exceptions away but if its needed in the mean time then we gotta add it +Ill wait for feedbackThank youMirek +Committed rev fdabebbbeaacedcff +I think it is may not be that bad that we kill the taskLike I said the problem is that there is no predictability +Yes I checked it and the problem is the component resource facets are marked as transient +not needing to lookup the web application context but this is bogus +Hi Fay unfortunately your patch doesnt seem to work for me +Ill try to do that soon but meanwhile forgive the large base patch +Thanks for the report! Indeed wrong matching can happen here if the argument value nevertheless matches the parameter type which it does in case of null +SPARQL Update +But really we can leave them unsorted +Yes will look into it closer +I know its a problem I have a workaround in mind I will try nowUpdate due to infrastructure problems we wont be able to provide this logging on client side atm +A patch for this +Candidate for EAP the in the jbossasclient directory is the unpatched one +Finally it doesnt look like you close the file you open for writeGood catch I didnt close the file I opened for read +Only if it doesnt break COPY FROM +eclipsetemplates +try to make it as like to the other MR jobs in the project? Most everything tries to use the imperfect thing which is a good rallying point +Could someone post the link to the state transition documentation in the branch? This would make my life much easier +I gonna check if branch has been correctly updated to. +The issue with KSSL is that it requires that you enable weak cipher support for all of Kerberos +However I note that this change makes the same assumption that youre debating in so we need to work out what the answer to that is before closing this JIRA +should work as expected takes only two seconds +Another factor is the coming HDFS caching +Linking with the already existing issue +patch has been pushed close this issueDevin +because axis is using the system property my temp files are being created in the tomcat temp folderso if multiple instances of my project are deployed in the same tomcat all instances write to the same directory +Sorry I should have read FAQ first +Markthanks +If this should be reconsidered please add the label +Eclipse classpath should be compliant with maven classpath +Path that prevents freeing a watcher in response to a session event per the feedback in +Whenever replication completes they can delete the lock entry and failed bookie entry from ZK +Thanks for the contributions Ruth +Will commit if tests pass +you for your contribution! Ive committed this to trunk +Grrrrrrr +reloc +Thanks Suresh +The targetserver reads the request headers fails to authenticate the user andimmediately drops the connection without reading the request body +I just committed this +Marking resolved merged to branch with revision +But yeah the jars should be there and all. +using to do get opened up to subclass in different package seekCeilFloor now returns a to prevent comparisons fixed some tests to not rely on a default values that is a general test problem in docvalues not related to this issueI still have some nocommits in there +Hi Felix I did two changes to the eventing test the job area in the repo is now always removed before each test and the drop test is now better syncedCan you please retest +Patch v results in a warning +By the way in the method convert in my object I filter out any of the requested scopes that are not in my list of approved scopes from my clients +Is that a bug or should I modify the test to allow Lucene and Lucene to write postings +Thanks for all your input Daniel +Unexpected type is different because thats an error on the level of the top struct so it makes sense that it affects the top structWhat do you do if you see an unexpected enum value in a list +Fixed in verified as fixed +It should be obvious that you can separate aliases from queries and mergecache the latter +Slated for issue has only started to appear since +Also such a configuration only makes sense if the multiplexer is used +Has this problem been fixed by or? Unless we get a clear no we will close this issue +See work log for recent patch +This was fixed +The attached patch causes to fail every time +So Im willing to switchAnyone else have an opinion +Vlad please investigate what izpack does +David awesome thanks a a bunch! It looks ok to us got confirmation from a few of the other committers on the tread linked above +Thanks for the report and pull request! +Ive added a target to conf named which will check the version of for the quickstart securitysaml this is not enough as resource injection does not work with any version of jbossws on +I have a test that I am writing for that tests for this scenario as well. +oh and I am using Grails with the plugin and the plugin on a experiencing this +Our unit tests that contain split do not exercise the code path in patch for with changes to address Santhoshs for slicer related changes revision of patch which includes +It makes sense to name it for the edit. + Bug has been marked as a duplicate of this bug +Be careful when you reply your browser seems to be defaulting the back to each time you reply +Hi CheyenneI circulated the images around the team and have now got some feedbackThe general consensus was that image was the preference although it would be great so see some more variations on this if possible? I think the abstract nature of it was what people were appreciating and the general tone of the icon as opposed to direct relation to something like an idol or beingIn terms of the typeface I got a positive reaction to typeface but it would be useful to see the icon with a mixture of other typefaces perhaps something like typeface but with the effect that you have on the icon where you can see white through the text and roughen the edges slightlyThanks so much for your continued assistance on this! Cheyenne I will forward them on to the team they do lookgreat to me though CheyenneSorry it has taken so long to respond to you would it be possible to see Option but with the darker colour on the icon as well as the font pleaseAt the moment the decision is between or this other choice that I mention aboveThanks for your patience! CheyenneSorry I did not respond to this earlier! Please can we go for option Thanks so muchTom +bulk close of all resolved issues in preparation for release. +State saving is really necessary update on when this might be included in a future build? Thanks +will only solve it for the default instance +Why is this issue still open? +In the connectionloss case usually everyone should come right back in the normal case +I think you can do that from the HWI interface +Ill do a full run of the test with the lockless branch of jcarder before committing +needs to be patched one more time +closing this i dont see why it was left open +Were now adding it if it is localhostjonplugin svn commitSending jonpluginassrcmainjavaorgjbossonpluginsjbossesb file data +I have spent a few hours trying to figure why I could not compile on OS +James IMO your work on this is done +Many first timers to Solr will follow the tutorial and itd be handy to have a example of a search UI with all the bells and whistles possible +Same error here using +Sample correct script that call mvnecho offcmd C hif ERRORLEVEL goto errorgoto enderrorexit endexit We can perhaps add it in a faq +I am not sure if it is absolutely necessary but it is the recommended way to run the bundle pluginWith the several packagings we may have some problems if a package is exported by api and full +Namely +This was implemented a while back. + with eJOPR regular higher spikes in memory usage and CPU usage +When finished reresolve with the new fix versions or label backportrejectx as appropriate +sweet thanks pquerna. +Closed upon release of Hadoop. +Changed Surefire execution to not use UI windows +Hearing no objections Ill commit. +The solution was to deploy the updated CAR as a plugin by using the deploy option. +if the instead logged a warning if it gets used with clauses? And if assertions are enabled then fail +Of course the object use is really the important part so Im glad you found that wasnt being used it explains a lot of ugly and not so helpful messages. +Attaching the new patch +Somehow we have to locate the event that causes the truncation of the filesCan you enable s infoStream and then get the corruption to happen and post the results +extensions automatically +Usageant rpm creates rpm packageant deb creates debian ThomasThank you for bringing my attention to the official Debian builds +Would the bugfix also apply over HDFS or S +First patch Changed return type of getLinks from Collector to List +Mikes latest results are more ambiguous let me run the new stuff on Linux too +It should be an easy fix but it is a bug and not a one +Im not sure if this is the right spot or if s run method should have this as a body +the Eclipse support is really extremely poor +this could have something to do with my observation +That would cut down on a lot of spambots +Bulk close for +Leo can you take a look +Its not in the All cathegory +I have following doubts +Please revert this and remove the verbose comments and the leading tildes +yes the other stuff is unrelated but I didnt have the patch created to start anew issuethe patch should be ready in a few minutes +Patch applied +Adding srcsiteaptguides srcsite file data +The change is slightly differnt from the patch and It also helps me discover and fix another bug that the output wrapper was using the input wrapper +Of course I wanted to do that with not +The actual fix in the patch is the same as in just going to pull this one in +I concur on both pointsThis just in Jira is awesome +when there is timeout the it is deleting the entry and upon ssvm restarting or coming up then it again rechecks and the download now +One more time +Closing it thanks Ivan! +Reopened to update fix version +Issue may already be resolved may be invalid or may never be fixed +the patch looks good +What should be done to make the dangerous singleton less dangerous +Looks straightforward +I have experienced the same error +Attaching section of from node that showed the problem +Want to fix this and send a patchThe JPEG image you attached works fine for me BTW +Running unit tests etc +This patch fixes the problem along with someadditional code refactoring +Tests passed cleanly for me on the previous rev +Thanks for the patch N +Implemented this proposal in Rev +Will review your patch soon +Can be incorporated in +Will the wired bundles change if dynamic import is used in the bundles meta data +Username nobody +I have committed a fix +food for thought +Which means this makes this the same problem as with typical equals code that checks private fields +Marking as WONTFIX. +bulk defer of open issues to +FYI there was another issue opened for an version Im working on a new one for soon will have it up next week in a new issue and will post here +The question is should the Lucene index have its own parser? Maybe its better easier if it doesnt +Cancelling patch as this should only be applied to the branch and not trunk +Were these also broken for this problem? Is the code to do the syscat update shared and thus this fixes all of them? Probably worth adding some test casesSo the current overall status of this issu is Need to write logic to drop column permissions when dropping a column Need to add test cases to see if the regression affects DROP COLUMN +This looks good to me +I am actually more inclined to solve my problem using. +the multithreaded connection manager of yours is still a sort of a blackart to me But from what I can tell the patch looks fine + +We will try some more tests tomorrow and let you know if we find anything +for distributed hbase and that adds requirements section from overview with some extra fill +My suggestion is essentially true for readers false for writers +Hui please file a new jira ticket for your concerns about +committed +Therefore this is a configuration or container problem rather than a Trinidad bug +Added associated patch there. + +Christian Id say go ahead +why groupId is? that site doesnt seem to existwhy name isit has no dependencies +Though this goal is partially realized with it will be more completely implemented in with the creation of the jbpmjsf component library +Changed to read may be migrated back to it manually does this fix the issue +Good idea +Too complex an API is a bad thing +I added step here just because I upgraded on the standby NN last timeLets close this issue with not a problem +Back port of the Apollo version of the comparison filter +Now here we have a parent class method overriding the method from an interface implemented in the current class +Stack or elliot mentioned that there was a concern that would be an unhappy character for some file systems and browsing hdfs in a web browser so lets pick something else +I think Johan is right if the conversion is interpreting zoneless dates as localtime it should set both TZ and DST otherwise the result will be inconsistentShould this bug be reopenedthanks +I dont know why the discrepancy exists but I am happy to see them on the maven repo. +bq +Here it can happen that a higher revision commit is successful and later a lower revision commit fails +Updated patch with check for starting slash +I thought Id done a rake clean but maybe ant clean was also b stuffis because I set TEMP and TMP to bThanks! +What platform are you on +Ill have a there an easy way for me to run this locally myselfJust wondering whether the other plugins have tests that are testing protected methods in other pluginsWere the other plugins this difficult to integrate? If not any idea why pisoa has been so problematic? Want to find out so that I am aware of issues to avoid with future plugins targeted for +Please use forgot to move changes to gc part +Ive added a line warning in the migration guide +Then the VM wont start because its nic no longer has a broadcast URI +Thanks DagMost of the function topics dont use a title for examples but its a good idea so Ive added one +Thanks Jothi +Merged the change to +Yes this is actively being worked on +Looking at your changes against my branch and gotta find what went wrong +I also find the limitation quite is one limitation that often results in XML hell with Spring since it is difficult to isolate different parts of the application and use basic composition at runtime to assemble the app +JacquesThat will be nice thanksTom +Im going to close this JIRA off and shortly open another one as looking through the conversation functionality now it seems that our timeout handling is not optimal at present and generally there is complexity in conversation handling that could be reduced +Given that we are getting really close to the branch unlinking from +But your problem setting is same as mine +fixed in branch revision to trunk r +Updating sabdb after your update +We ARE planning to remove them +Run the script specifying credentials for builtindba on the vm boot command +Can you post a new stack trace +This object can be used again for another transformation just not concurrently in other threads +Includes with semi transparent buttons to reload local file and webview reload +If passing the wrong type to a foreach is really wrong then an exception should probably be thrown +I think I got all the dependencies +Only those commits listed are valid. +This way we will at least look at them again prior to GA. +And perhaps that is part of why it is more difficult for me to reproduce in a test +Attached are selected logs for different databases and look for deadlock keyword +Thanks I will have a look at that. +Should we put it in branch also +I would like to avoid a design that incurs an overhead of an additional RPC everytime a link is traversed +However I still recommend either EJB CMT or Spring transaction management Simply turn your to BMT if you want to demarcate your transactions in Spring +The history is all there but it will no longer appear in the latest checkouts etcOn iv I presume this means commits dev +Yes good point +I didnt manage to use to get to sure if this is just for branch or also so assigning it to you brian since its bpel related so it doesnt fall through the cracks accidentally while im gone +this avoids dups +Reopen if you have issues still +It would prevent the user from moving or closing the host frame +In fact we just received a request to make this same change on Oracle +It looks like youre connecting to stompnio connector and not ws +I have tagged it thus +Fixed and applied changes in commit revision caebdffdbcaeeef +Bulk move issues to and +Placing the product type in a config file is one way +It does the same way as yarn script +Thanks Giri +If so would any commiter change the component to differenceBest by Spark. +I dont see where removal of value is testedaccounted for? Am I missing something +Dont see any obvious issues in the patchsetShip it +Allow applications to configure which tools show up in what order +I wish there was an error log or something to see what the value is +Diff for adding WAL log added Michael Allen as a contributor for this patch +sounds like a good idea to me +Please commit this only after is submitted +seamdemopart is remade and is remade and corrected +Fixes the regionserver lease and removes unused default from old shell +This is looking good +Is this the case earlier too? for this patch +I might just round out the producer and consumer the easy way and let someone who knows what they are doing do the fancy version when they get time +So this patch is good for testing +i parameters +KevinJust checking if there is any progress on this one +Closing old resolved that dont have a fix version and therefore cloud queries for M release +Did it not occur to you to like mention the fact that this happens in such a nonsensical case +Im not sure maybe calling a method that often isnt really a use case? I found this bug while working on benchmarks for +Removed wssj +srcjavaorgapachecommonsconfigurationimport +No problems go for it +Reporters of these issues please verify the fixes and help close these am able to invoke find API without any issues +Moving out of +Thanks! +Can I get a diff or a link to the commit so I can take a look at porting this to stableIts causing us some rather bad production issues and we cant roll back to due to another issueCheers +Tried the new processor it works wellA nit For the processors not supporting the delimiter oiv should show an error message +Instead how about looking for custom assembly level attributes on an assembly event +Without that I would have to run the test that isnt present in ve committed changes as a result of the review commentsAssigning to Andy to resolve the failuressvn commit srcjavaorgapachejdotckpcsinglefieldidentity srcjavaorgapachejdotckapipersistencemanagergetobject srcconf srcjdoapplicationidentityorgapachejdotckpcSending srcconfAdding srcjavaorgapachejdotckapipersistencemanagergetobjectAdding srcjavaorgapachejdotckpcsinglefieldidentityAdding srcjavaorgapachejdotckpcsinglefieldidentityAdding srcjavaorgapachejdotckpcsinglefieldidentityAdding srcjavaorgapachejdotckpcsinglefieldidentitySending srcjdoapplicationidentityorgapachejdotckpcsinglefieldidentityTransmitting file dataCommitted revision +Thanks for all the reviews I just committed this to allow me to roll +The patch move the map merge function to run it when a map is given +Much appreciated. +Im surprised you are getting any work done at all vs just recovering when not in baby mode +I tried it +doc included in the last patch +Available in trunk. +Committed for Donald under revision . +This is the main reason for the eventing test to fail +Thanks for committing my fixesThere are still typos I thinkThe first one is about the default output for the +Thanks for clarification I will post patch soon +So if myfaces would be more kindly to an osgi environment would be a big step forward +Sorry updated wrong version of release notes +But because the file lives in PROJECT +Brilliant! +Thanks for the pointerWe will see how we can add them already for but currently we are focusing on getting a release out. +Same here is coolIMHO it just needs something to make the letters stand out a bit more maybe a bit of thin white spacing around the letter shapes +Thanks Tony the dev list is the correct target for notifications +I do want to name my classes as I wish without them to be executed as tests especially when I use junit annotations to specify what classes are actual tests or youm a relative newcomer to Maven and this one bit me too +Sure +However i wondered if it would need to append this setting to any existing JAVATOOLOPTIONS then reset at finishI have applied your patch +As your proxy looks to require authentication could you try to add into the karaf startup script +The could use this for the and +Its all in the server project except the classon the linked task which depends upon the new TM classes and obviously needs to go in its configuration +Ill be reviewing this patch +This would make designer only runnable on AS +Is it a set of colors where each one has its own transparency or do we specify a transparent pixel +forgot oracle explicitly commits for ddl operations +And any advice are welcome +Hi guysCan we add days to the drop down too +The attached zip +Awesome it works againThanks heaps Bryan +My mistake +thoughts +A fix is checked in +That lone collections business is pretty silly and removing it looks good +Not sure why users need to pass in generic options to given that we the hadoop options but at least this fix will allow us to pull the options out of this class +well whatever I am +ivy +In order to investigate this issue were going to need more details on your scenario +Thanks Matteo +Committed revision +Thats it +Example we have etc +Marking invalid until quickstart is provided +Yes and thats why I cant understand the problemHeres my library pathjllesieurusrlocalaxiscbin echo LDLIBRARYPATHhomejllesieurlibusrlocalaxisclibAnd heres the content of my lib directoryjllesieurlib ls latotal jllesieur jllesieur +Looks like this finally made it into the newrelicrpm gem +See the discussion in the file +eg like the JMX explorer where you can are now able to use the context menu of the fabric node to connect disconnect. +Patch applied to Java Javascript Python language +Create a testcase in the testsuite to investigate +Matching first on the exposedName then on the underlying nameseems like a promising technique for this issue +The new targets dont seem to be show up in the output of +dont know what happened but this issue somehow reappeared + +patch v check region is splitting through Also I fix another a bug when region is disabling and split is rolled back in current logic if parent region is roll back and it will not be closed if the table is disabling +Should we create a different bug for this? +You are right I mixed two things in one sorry +Thanks +This latter guarantee cannot be given for circular references thats why circles can only be resolved on a best effort you cannot get rid of the circular reference you could implement your in a manner +I have gone through the state machine for and I think I have found a few issues +I dont know Stripes in detail but I guess this switch will change or remove most +Removed checkThe hangs since it was updating the offset for the wrong ZK path +I have applied it but I think this getting started should be much less chatting +Minor change + +Paul took care of this after the release. +Tested with both Maven and Maven +I again suggest to make changes in to make sure that it creates service directories with correct permissions +We can replace by our implementation in our HTML Configuration +Ill close again once Im sure it was something to do with my dev branch +Added further text and images to demonstrate making a change to the sample BPEL and redeploying using the full publish capability +I hope Im not discouraging you from pursuing this workthats certainly not my intent +Perhaps its because these machines are too fast not too slow The only reason I can see for the task to get rejected is if the executor is shutdown +I felt this was a more flexible solution than the original request of using a system wide property +This one is much less +I have tried to use jing it is not suitable cause it generates some unnecessary windows events and it is not possible to demonstrate the play this avi file ffdshow Video Decoder is necessary +Hi Olga I have added the parse error into the log info in the patch +Applied thanks +Would the following be a solution? Make sure that a jBPM installation always has a deployment servlet That way the ant task and the designer can both deploy to a URL So there is only configuration of the database +Yeah it gets complicated quickly if we want to do anything more +I have now done some testing with the and so far it works great +Could you check the server log possibly attach this to this case? It should be located under standalonelog +In the future these strings should all be externalized NLSd +Ive committed some code in revision which I think fixes this looks like we were creating threads and not killing them properly +We first had the DROP limit at based on a suggestion from you guys when we originally had problems with this we recently increased it to on your advice +Id love to see a patch and some benchmarks thoughSee for some of the view server performance work +Im not sure what you want me to do with this +Hi KarlThis aborted pretty fast +defer all issues to +svn directories in the Axis for the patch I got the idea from the patch and change to ignore those files +I missed thatIm full for having ddoc as root +Good idea +Fixed thanks for spotting this and for the patch +To build +Backported patches a and b to with revision and to with revision . +Carlo could you please comment +Yep +It used to happen at startup with a workspace already having projects too +Thank you for your report +Setting useJmxfalse in the camelContext works around the issue that this test case exposes +Latest patch with comment as requested. +Should it not show the options like list and create and search? It will only do a search using rooaddon search +here is a patch +I went ahead and put the new version into place and updated my in the same way that the patch updated the example now I have begun a +I have encountered the same situation +lets close this as Cannot Reproduce +Docs Q Struggling to gather the gist of this discussion +When you open sourcevisual tab in visual tab does it resolve +Im also curious whether jars could be used to read indexes created with Lucene My gut would be to punt on this for but the patch may be handy when we decide to itegrate with the api +All in all probably something that wont be fixed without some funding. +Added IT to confirm this. +Its strange but it was tipelement CSS class which had been the cause of widget malfunction with Chrome +as will be default in core this will be fix only in +I want to thank Carl for pointing to the appropriate PIG issue which provided the idea for the fix. +New patch generated from the base directory instead of +For now we do a compare and swap update and throw an on failure +What is the use case that you envisage whereby someone would want to use a custom in combination with the namespace +Problem exists in too +cloned from fix of this issue involves too many API changes so I dont want to take risk to back port the fix into EAP if anyone run into this issue free feel reopen this as a Known Issue for Hibernate in the EAP Release Notes +I created a patch thats from the trunk and uploaded it + Rohini nice catch +Rob any updates +Do this for M release. +I think the urgency should be set to normal as it is a matter of incorrect error message and not a bug which may cause data corruption etc +Hi ClebertI have this test with very small load and it works fine +servicemix upgrade to woodstox as has not yet been releasedAll components build work fine +Fixed. +Perhaps the test is too synthetic although it does capture the essence of the changes +Thank you Brandon +Anyone here that can check this one +HiIs there an example code of this I would happily incorporate it into +Should be fixed both in and issue has been resolved for one month or more without complaints +This should fix it +but not the norm +In Beta I think this issue still could be clarified +I attached a design document and entered linked issues for task breakdown +Ive renamed them to the unique names and improved them +One more thing +Updated to build revision Opened unified multiple and jbds targets and clicked updateAll have errors so see my attachmentsThese have been tried in a Juno IDECannot currently understand why these targets should not work in Indigo since the point of a target platform is that it is independent of the IDE it is being used in thus allowing plugins to be developed that could work in multiple platforms without having to open a different development environment +Perhaps my question is bigger than this issue or shall be asked somewhere else but looking at Hive code I couldnt help noticing that it is essentially build as a singleton +Logged In YES useridThank you so much! Well try it again +waiting for feedback from Mythili +If more things to refactor in that area apear we will create a new issue for them. +I merged revision to with modifications because of the removal of the changes file during the release work +But you will still need the file in the includes the new file +This issue is resolved on the types branch +Agree and thanks for the explanations +Indexer that sends to output to the above patch disable solr indexing +Thanks Arvind I am still unable to apply the latest patch to trunk +From river dev mail listPeter Firmstone wrote Lets add the constructor it wont break anything and commit the patch Ive read the patch its straight forward then well run the tests again +Have you verified it? If so which patch should be reviewed and committed at first? Or you want them all be reviewed and committed together +I think this more straight forward and with better scope control the old way could possible have some strange side effects +Missing licence added to new file +I spaced stating for clarity that I have not checked the change in +Marking as resolved. +to Bikass comment on separate callbacks for the allocation and completion sets complex may find that the order of callbacks create problems as each set will be handled separately without an overall picture +Can you share a example recipe on how to reproduce it +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Andrew So sounds like it doesnt drag perf but how we know it addresses the issue? Should we do your patch commit against another issue altogether +I was able to get simple extensions such as subsetsum to work but homerun pg and mysql all killed the process with SIGSEGV +Sorry the original patch does not compile due to a few typos +Im talking about the maxThreads parameter on the connector +As Amareshwari pointed offline trunk patch would probably need more changes +patch submitted to fix failure of build In LIBDIR moved back to BINDIR and added back MANDIR to comply with the spec fileIn spec fileRPMBUILDROOT removed as prefix in as BINDIR DOCDIR are prefixed in +Well now its fixed for real +The indentation of a few statements in is incorrect making it hard to read +Opera is not supported +Camel is whats being used in Fuse Enterprise release. +Make sure all licenses are on commit and add some class comment to new classes saying what they do on commit +Even if there is an exception occured in eventlistener it shouldnot rollback acquire +I must play with module excludes and extra step for removing duplicate entries from. + testcase +I have attached screen shot of the archive created using Compressed Zip of Windows XP +XHTML pages via the XHTML File wizard with different options +ToddThese timeouts look reasonable to me +If we change the format we can do PREFIXCOMPRESSIONV +This even does lead to nasty when using parameters of wrong typeIf it dont mind you I extend the current implementation to consider during +Thanks for patch Kannan +uses both the raw region name and the encoded region nameUse new static method avoids race conditions on log splitting by only doing them in if the servers lease expires while the master is running +I cant think of anything +Can you put some comments into the code +to Jake to this issue +Stack I agree +Also consider the dropby variant whatever that is +Closing. +My point is that the user should not need to manually make sure that the is present in all his for distributed indexing to work +The difference is that in Maven also incorrectly omits and some set subtraction is going on +After every child within the list the node level is up level rather than which cause the whole json output strange +If not that it shouldve run all tests and reported all failures in the nightly builds +to update +Excellent! + +no i think we need to back it backwards compatible with so we need to be able to process txns without the parent cversion +Is it ready to release? please provide staging repository for seems that this new groupId is not being synched with Maven ll ask to enable this synchronization. +Looks good +yeah +fixed in next nightly +This doesnt look right at all +plugin seems good enough for now +Please verify +Thanks for all +Would need to be tested on a cluster +Committed to trunk and +However this feature should work along side conventional XML configuration and not be required +The memory limits of and can be used to control things on a basis +trivial sure thats the same issue? Seems to be referring to issue with needing quotesFrom what Ive seen its that having two argLines doesnt work need to have one +I believe that optional dependency in +I do not think so there are no other types of errors +It is still possible to build the project byant +But we definitely would like to see card present functionality available +was an example of having too many copies of counter data structures in memoryGetting weakly consistent views of counters is consistent with their current semantics during the course of a job +Looks like files got moved with mavenization +Sorry this fell off of my radar +Throwing Exceptions may not be a good option +ok i upgraded to the latest and get the same thing +This is our real case that is failing +However that will break the APITo allow existing users of the API to migrate to the enum API I think the approach should be Create the enum class and add API +So its odd that the first node never complained. +batch transition resolvedfixed bugs to closedfixed +No need to file another issue +x in. +Ive committed this to trunk and +From Jesse +Closing resolved issues for already released versions +jvmargs filtertrace tempdir jvm +Yes you are right Robert but I dont think DC make sense for writing however people may have use cases for that as wellI will create the patch and update here and we can discuss more about itThanksMayank +CXF also used to generate code for a wsdl which had no portType element +classpath to the root of the annotation folder and then adding the two +It makes sense to me +patch looks good will commit if the tests in also +I reopened the issue yesterday because I felt your closing the issue and marking it Resolved and Wont Fix was predicated on the idea that my changing the options to enable false would fix the issue +I was specifically referring to statements that are meant to evaluate to a certain value +Im not sure I follow in what way it is really a problem as you dont use the variables from the binding if your script is empty +Moving back to for now. +Marking critical +Thanks for the quick turnaround +Combo with all dialects is added after DB type is changed default dialect for DB type is in in +If the cassandra node where the TT resides isnt working then throughput is reduced regardlessRight we want it to be blacklisted in that scenario +Fixed +I actually looked into this a fair bit when I raised it I think the easiest way is to just make it a Configuration parameter then you can use the standard Hadoop Configuration object to get the array of connection parametersPart of the problem with this for Sqoop is each connector has to implement it so while it can be fixed in Sqoop it wont work in say the Microsoft connector or the Quest one +this pom is just to define the repositories but why not just sync the whole glassfish releases repo +Rerun the app in the debugger and step through the put method +To shutdown down trafficmanager and trafficserver you can do the following on linuxsudo killall s QUIT trafficmanagerTraffic Manger will come up with the Web interface disabled but you can execute the sudo PREFIXbintrafficshell command and see some cache proxy and http statsPlease review and let me know if there are any issuesGeorge +file with broken links in photo album guide is attached +Weinan I agree with test since an Exception is thrown in the +My apologies Ive been out on vacation just getting back today I tried but cant seem to repro it locally with small dictionary files will try to use your patch and larger disctionary files +Bumping the urgency of this issue to clarify that these problems are a blocker for +Shouldnt its priority be raised to blocker as mentioned in the comments so that it doesnt get missed out before +Please find integration tests from the attachmentPlease apply the patch given for before applying this patch as it contains the sample configs of the automated tests +After looking I noticed that several files had commented out require openssl which probably was the source of this + deletes header files after successful replay handle short entries and garbage size to add a test for the corrupted partially flushed minus the BRAF change +Any help +maxActive each thread holds connection for ms threads runs cuncurrentlyI got the good record of this patchdbcptype throughput Ave +Just for information it can help in moving files in SVN moves the following files renamed vmportinclude vmvmcoreinclude renamed vmportinclude vmvmcoreinclude renamed vmportinclude vmvmcoreinclude renamed vmportsrclilemtpiminclude vmvmcoresrclilemtinclude renamed vmportsrclilemtpim vmvmcoresrclilemt renamed vmportsrclilemtpim vmvmcoresrclilemt renamed vmportsrclilemtpim vmvmcoresrclilemt renamed vmportsrclilemtpim vmvmcoresrclilemt renamed vmportsrcliliapiminclude vmvmcoresrcliliainclude renamed vmportsrcliliapim vmvmcoresrclilia renamed vmportsrcliliapim vmvmcoresrclilia renamed vmportsrcliliapim vmvmcoresrclilia renamed vmportsrcliliapim vmvmcoresrclilia renamed vmportsrclilipfpiminclude vmvmcoresrclilipfinclude renamed vmportsrclilipfpim vmvmcoresrclilipf renamed vmportsrclilipfpim vmvmcoresrclilipf renamed vmportsrclilipfpim vmvmcoresrclilipf renamed vmportsrclilipfpim vmvmcoresrclilipf renamed vmportsrclil vmvmcoresrclil renamed vmportsrclil vmvmcoresrclil renamed vmportsrclil vmvmcoresrclil renamed vmportsrclilpim vmvmcoresrclilstackiterator renamed vmportsrclilpim vmvmcoresrclilstackiterator moves the following files renamed vmvmcoreinclude vmportinclude renamed vmvmcoresrcutillinux vmportsrcmoduleslinux renamed vmvmcoresrcutil vmportsrcmodules renamed vmvmcoresrcutilwin vmportsrcmoduleswin +It defaults to ignoring associations which is the current behavior +Fixed in +Also we need to update document +waiting for to be resolved +patch applied thanks! +m s would contain definition of mirrors on Jenkins +r release bulk close +But we cant determine that because all events come from current datastore right now +This adds a failure in and +The problem seems to be something related to different implementations of threads between the JDK on Windows and Solaris +Closed again as fixed after correcting some mistakes in my test environment setup. +I am waiting for a fix +In the past I tried to configure it for but my effort was unsucessful +Hi ChrisI would like to work on this issue but dont know from where I can fetch all the supported fields by the container +Adding a fix for so that this issue is not forgotten I think its a serious problem +Hi markus I am not sure if the committers committed it +Might there be a need to standardizeIf you wish soYou could start a discussion about this on the dev MLIf you agree Ill set this issue to Resolved +C on the namespace also allows table creation in the namespace +Maybe we could think of another way to handle that the jbi bundles could start a file monitor on another directory where pure JBI artifacts could be dropped and the JBI layer would handle the transformation to an bundle and install it through the api +It actually works out better leaving it the way it is as we can then use the same Connection yet set compression onoff based on a configured threshold +Great! Ill keep an eye out for that +I recommend adding one in the CP tag to set it rather than update the file +Indeed the patch is against legacy interceptors so no MVCC +You should not have to reboot to clear the problem however +At a quick glance it looks like the subqueryFlattening diff might be OKand perhaps I just need to update the master outputBut it would be nice to have more eyes on this +When will this issue be solved? Or how can I implement a workaround +As bug is a duplicate of bug Iintensionally added its test cases hare after cleaning them up +Basic docs for new features are doneAlso i have provided some basic examplestutorial based on iTest on separated page +So I think this is not a bug +Bard once confirmed please add it to the spec +core tests +This really simple Java stuff that Im surprised is raising concern here +This is a bug in the rebalancing logic which disallows us from horizontally scaling out consumption in the presence of multiple topics each with a small number of partitions +I think it helps other devs to get into the internal and hopefully transports knowledge +The now provides a basic implementation that uses plugins for policy decisions. +bq +No this has not yet been implemented +Attached is a revised version of the first patch that leaves the buffer doubling logic intact but remedies all the other issues I described and addressed in the first patch +Looks good +We probably need integration tests for this functionality +If I understand correctly I dont understand why you need to write your own file +There was a problem in the MEMDAO implementation where the method was only removing the routes associated with that correlator and not all the correlators +patch for this and thanks +This mutex is internal to glibc and there is no external access to it +This has been fixed and the fix is not available in higher versions +were not using the test harness for the assembly plugin now that its refactored +Thansk for the review Pi +I have not updated or addedcomments and I have not tried very hard to make the names ofvariables and methods meaningful +And yes I kept it at K iterations because I am getting OOM at M +The build is blue again. +Sorry +Pushed in . +Component set to tck +Fix sounds good to me reassign as you see fit +Without it only outlinks from the extracted text are collectedThis is a bit nasty and id appreciate if anyone with a bit more experience with Tika can shed some light on this +Turns out this caused a Java deadlock in one test when running with the automatic index statistics update featureInvestigation revealed that this method can actually be removed which makes the uuid variable final +servlet redirects it to the portal +Fixed with revision +Data size is just Gb +reattaching with correct patch failure due to cross module patch contents +Alexey please verify +Be aware that this extension can be developed to support both SOAP and REST +I took a further look into this in particular the code where the parsing occurs +Linked in the issue where I first introduced the approach +Works for meOleg +Thanks Bjrn Andr. + +Deferring all remaining issues so I can release in JIRA +Fix patch attached +But dont worry about that +Proposed titles are not always bad but empty titles are +committed to branch and trunk. +I think that it is safe to backport +These are of great help +The issues were imported but many times they cant be reproduced +but then the issue is how to know what to put in the model do we ask the user? or look at the typevariable names etc +Look at that we have in a testsuite +Verified upgrade in ER build +Although my mind twisted like a pretzel thinking about the correct condition here +Should we open a JIRA to move the config key to Yeah since Andrew has already committed the patch +These are the anchors that can be clicked to invoke connection creation +It is unlikely that any one of them will be close to gb otherwise your cluster is very badly unbalanced and you should rethink what youre really doingIMHO it it much better to have subranges by node than +the file is just downloaded and never actually used +Of note this patch doesnt fix all the eclipse template problems that we have currently +Is a duplicate of this of closing stale resolved issues +Should this test do the same? For now it prints to stdout but Ill remove that laterThe only real problem right now is that the threads run for a little while then stops and wait for almost exactly seconds then runs for a little while again and keeps going like thatIts almost always seconds +Seems like move should just fail if token is null +Hi Xavieryou are right property editors are stateful so they should be gotten from the application context and declared as prototypes +Include the here suspecting it has the same code path +I think better to rename Style class field is empty why its empy its not our deal +It facilitated adding but also is more realistic because an admin client is likely to connect change perms then disconnect for each set of changes +I changed the static links to https +This issue is no longer relevant for it is fixed inChanging the affected version to +Everything looks to be in place and working. +buildivylibcommongriddevbin +Derby query processing performs WAY better whenever it can use hash join vs +Is file the only scheme needed? Or should this support other ones like say ftp +OTOH you should fix your ideeditor settings +Thanks for reporting +junit test case to Reproduce the issue +Nice fix for shell functonality +Fixedwbuildrsvn ci CHANGELOG libbuildrcoreSending CHANGELOGSending libbuildrcoreTransmitting file dataCommitted revision . +The support for connection related HTTP headers was implemented and is in CVSDifferences between HTTP and HTTP was taken into account when implementing thisThe implementation was tedted with HTTP Connection close caseConnection headder support is also implemented but could not find a practical test case to test this +We do not have access to the system name that this actually representsI do not have access to Solaris at present but can you confirm that Ruby on Solaris reports sunos or something similar? Theres no reason we couldnt report the appropriate version in both and modes but I want to know if this is a change or a SolarisIndiana change +No sorry i investagted but this one is from my IAP +tomaz did you attach a patch? I can only see your compiled in the attachments +Merged to the branch in revision . +Hence size of the plugins will be reducedWDYT? If this sounds reasonableThanks and RegardsHarshana +I am not sure if this is related to the balancing algorithm or to roundings + is the incarnation of a similar issue for WAR files +Anyway I condensed my ideas about the issue if there is any in my latest comment +deleted on the Emmanuel +that could introduce the need for extra synchronization memory barriers + +More environment infor local repo no mirrors is central repo plugin version the thing it found for log was a taglib that I didnt recognize. +Pushed to master +Committed revision . +Same with aspects beans orm +added +It would be good to only include the the end to end testing with real cluster +Committed to trunk at revision +Yes as long as the clock on the master doesnt act funny +As previously noted is final because is an interface so no one could have had a different implementation they were using with Lucene hence it would not have broken back compat in a meaningful way to add some new methods +for the with revision +Closing all issues fixed previously after a brief review of each. +The changes were not difficult but I only changed Cocoon Forms v +I was thinking this also while looking at all the xml elements available on the whew! Ill do up your suggestion with this ticket +Committed revision The referenced test was just a simple problem with the header filters +but HDFS should not depend on that for any important functionality since this Datanode is already in bad state +Ill kick it off again once is in +It still has an obligation to process the file to find annotations and to properly identify all the +I added a test for this too and the test seems to pass I didnt set up a project to actually test the parameter +It is better perhaps to use a special map for these headings +I think we ran into that issue before and the JAR should not be packaged in the WAR +one is at the other is scripts are kept in the qa module in svn. +Instead it should use a new property like +Will work on this as soon as I finish up what I am currently working on +What do people think of making span highlighting the default behaviour if the query contains phrases? It might be better to have the default behaviour that which people expect even if it is technically different output from +Anyway we should really discuss that with the team on the dev ML I let you fire st +It this bug still valid? Attached test is not failing +Again the question is if this is needed for or can be deferred +Furthermore its quite unclear why its required to hardcode the servlet context in either of the described situations +Bulk close for Solr +Delivered on . +added new links to Users Forum +It is very easy to reproduce on my servers +Thanks for pointing that out +Hence the necessity for retaining the instance for the event handlers +but checkout the permissions on folder etcconftry ls l etc grep confwas i able to explain it or i need to be more verbose +I logged back on today for another reason and noticed your activity +Yes with the bundled with the RHDS beta +In case your service functionality depends on some external resource which can be unavailable the standard way is to have a dependent service that will either check or start that resourceYes but in the real world networks go down fail etc so to work on the assumption it cant fail just isnt possible. +With so many watchers and voters there should be at least one person who can provide a valid right? Would be great if somebody could provide a which reproduces this case just to be sure were working on the right issue +Please find related design doc in umbrella jira +Folks will help you outBTW the problem here is i think the is not able to connect to ZookeeperCheck if there is any firewall and your configsThe first time when you do JPS the process starts but it gets killed which means there is a setupconfig related issue +Weird! Ill poke at it a bit +Makes sense +if profiles define in mconf then property and jdk and os triggering profile not is the correct behavior +Will do +consider JSF optional validation of empty field +patch applied to master +That looks good passes all the existing specs as well +Im assuming so but wanted to confirm +btw im getting the failure on a coreduo laptop running ubuntu +disabling caching entirely with the given patch didnt solve one of rvalyis issues though it did solve the other +to trigger Hudson +I found a few more things to fix +Some missing classes seem to be Can you redo the patch one more timeThxKurt +Start time and finish time are both computed by the scheduler +Care to put together a patch so we can start moving with it? Thanks a million in advance + Ive just committed this +Yes that did it all runs perfectly now with all tests passing +That is why I manipulate the zookeeper by explicitly inserting a znode for the new table +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Fixed by +clean test output to from am going to import to fix this issue. +This is unbelievable +Thanks Xuan. +is a concrete while is an abstract parent class +Committed to trunk and. +Oh great! Youve gone a step further and did what I had started inI think Drubas comment about performance measurement makes sense and it is something which needs to be tested +bulk defer to defer to +the commit method appears to have the same bug +I think its not too far from ready although the core reload bit is kind of a hack +Updated patch for fixes blocktree not to seek on write changes are backwards compatible we can deprecateremove appending codec we should just keep its testPersonally I think i like this better +Manually tested! +The costs are year for years +DavidThanks for your time. +it was handed to us +Thanks Yongqiang! +Thanks Jason +See resolved issues for already released versions +Thanks again +Picking this up again I havent received any more comments on this and it seems a net improvement to me even if it may not be a panacea +Because of this submission you will need to upgrade your to mustang revision or higher the signatures for the wrapper methods have changed and the jdbc support will fail to compile against older jdks +Thanks! +The forceSync isnt used anymore but its still in hflush we lose the ability to set flush and theres refactoring around log thats out of scope +Please check it applied as expected +Its reasonable to make TT run without DN and +Simone Im curious about how we create the palette for png +Close all resolved issues for Engine release. +I have prepared the first incomplete implementation which checks validity constraints on minExclusive minInclusive maxInclusive maxExclusive +Now the error is seen as the folders in the secondary store doesnt have executable permission for world +resolving +no error indication to AM +Should this now be a different JIRA issue as this is marked as ClosedLater +I dont have issues with having an issue description message in two sections if that makes things more readableprecise +Please reopen if it persists for you in version +But mostly based on the assumption that the user could just configure it and at worst we could drop a note about itThen I saw that there would be some side affects to that based on this patch? I didnt really look closely but from memory I guess it would disable being able to set compound file? And you wouldnt be able to change the tiered policies settings? Or does Solrs config any given method voodoo work in this caseEither way Id be fine if we didnt respin or if we did +Code looks great so far +code is in patch for contributed the implementation of this patch +Please attach an example that reproduces the I wanted to make an example I saw that my server returns wrong +No longer able to kill catalog the loading of a project has also been isolated from server availablility. +A solution of this problem would also be helpful for the requested feature of saving back changes made at a composite configuration. +But is not moving forward +Im afraid I cant reproduce this issue +How do other committers feel about upping our prereq to? Another option would be to disable Erlang support if is run with an older autoconf +I also think mocking out the storage directory is worth the effort because there are lots of scenarios this would allow us to test I dont think storage directories are that tightly coupled +ok +You can reassign this to me forpost +Good catch Srimanth! Thanks +Ignore comment about pinning I am not sure what I was babbling about +Hi Knut Concerning the query timeout failures on Solaris has the frequency of the failures gone down? Thanks +So the workaround until this is fixed is to install the required pom locally +However i think the flow of the howtoxml and howtoconfigurationfactory has been really messed up by myself +Temporarily included Eclipse BPMN plugins in Savara local update site as trunk code works just no stable release available yet. +what about licences +Not for commit +simple fix by adding the option to the mysql committed to trunk +I rolled back to and I got the same results +So this patch will not actually apply to Hudson yet +Simple patch that drops a few unnecessary Overrides +Patch adding a new distinct retry interval field +properties files and sync up with across them all where appropriate +Various example jobs jobs through pig oozie etc also failed because of this +managing displaying infos about memory interface for manage modified for getting memory usage I attached the files without any javadoc for continuing contributionDI feel like well have JMX support earlier than we expected thanks to your work +Assigning to Peter to verify RHEL and close +I am trying to assemble a zip that contains all the libs and source so someone not using maven can still use the package +Ran ant test in mapreduce successfully with the core jar containing the patchAll tests passed except which passed later when run individually +Many thanks Shawn +Of course I know there are other people waiting for this +Otherwise feel free to tell people they need to download it +Added an odata configuration property that can be set to the desired replacement string should an invalid character be encountered +guvnorinfo is deleted the association with Guvnor is lostThis also implies that if the project is checked into a source repository those files should be checked in as well +FYI is still though important that users can override the b for backward compatible point worth checking is how WTP decides wether a server is remote or local? if it is done per the host name then it will affect hotdeploy. +Perhaps I am goofed up somehow but this bug does not seem to be resolved at least not in the latest milestone with and I observed this problem on both I upgraded to for both libraries and the problem was fixed for Facebook but not for Twitter. +then close the second should get a failed send to the temp +mrepositoryorgjbossarquillianextension upstream. +One question I have is how the results frommultiple Searchables would be collated into the correct order +introduced a property that is set true by default and causes the fixture to registerunregister the context for each test. +It would seem that Class proxying is being used on a class with no further clarify that last comment we can proxy the result or we can proxy the FB it simply depends on your configuration +Ive created these followup issues and marked them as related +defer all issues to +But this is not directly related and thus would better go in another issue +Yup more precisely for all partial matched split only unmatched part will be processed by following pairSo if i have pairs all using matcher only match first recipient and send a mail with recipients +A test case that can be included in the unit tests would help in terms of investigation and in ensuring this stays fixed in the future + +So in other words if an application gets submitted to the RM before any NM registered the application will be stuck in pending state +Heres a more polished patch which implements the changes I previously described +It also works on Windows beta are you compiling from source or do you maybe have spaces in your path? +name +Related to the above comment I moved to +As well I think the logj logging needs a separate section +Ill try to set up a test case and Ill verify with the developer who had issues that is the only version with problems +But why we habe a standard? Its only an advice that there is something whats not working like defined in the standard +This addresses Owens one and only findbugs warning can be ignored +Verified on ER. +BTW If you can write a patch before tomorrow thatd make my life easier since you already know what the problem is +I think this patch broke MR trunk compilation +I think I just fixed this issue in my last push +Moving to +H insert statements fail as well +Typo on line the exception handling filter should be the logging filter +block info might be sync asynchronized +Committed at subversion revision +Ive confirmed that the files listed above existCProgram FilesApache Software FoundationlibprivCProgram FilesApache Software FoundationbinCProgram FilesApache Software FoundationsharecouchdbserverAny other thoughts? Thanks again +I just added a comment into the mail thread Please dont put the servlet transport configuration and http jetty configuration together they are not supposed to be together + +Even though the patches dont use drag and drop associating items in adjacent columns works fine so Im closing this issue +Did you see the log file attached in the second attachment? Is that enoughSamsung Firmware Version Baseband Version TVUVKJBuild Number log file attached +This bug is in the test code and we dont push large amount of data over the validation framework +But I think that is a good tradeoff +The problem is two lines of missing code to set the request attribute +This means hte easiest course is to simply change the runtime into a runtime +There are several problems with the distributed domain support and the apis to sort out so maybe would should just say domain support is differed until beta +Fixed since build version consistency is important for upgrades +to trunk +Im basically in agreement w Guillame +Integrated to TRUNKThanks for the hard work N +Hi BoThanks for uploading the patch +Im trying it to see if I can get into same state +This should work in both modesUsers arent currently asking for an open method and its possible to work around no open by keeping a state variable to see if youve been called before +Sending the improvement implementation and the test case +Have modified to filter out vdb source VDB source models from being added directly to VDB in Add action and New VDB actionbutton and dialog to show user list of importButton is disabled if no import vdbs exist +Toby isnt this fixed for already and only needs a backport toIf yes could you please have a look at this otherwise close the issue pleasetia we sure this is a bug in xwork? Ive never come across it +Does your fix rev r mean GEP will only work with a Sun JDK +Can this be done via the new annotation approach? for both the dynamically deployed endpoints and also services deployed into +the pagemap itself should be created when the link is clicked +Not sure why we suppressed them +youre right +bulk close of all resolved issues. +Change the message to remove unique constraint from the messageProblemDuring soft upgrade run unique constraint cant have null able columns +I just havent had the time to upload the new patch yetIts a good test case to have +Id guess its that they are loaded by two different classloaders +This came up on the Tika mailing list so Im attaching the current state of the patch I have +I noticed that every time I update the issue the list gets spammed sorry about that +The reference manual should contain at least a synopsis of this topic +Other projects Ive used have thisfunctionality and I was surprised when using velocity that I could not find itSo I submitted the patch and expected some debatecomments about the API theimplementation or more convenience methods and would be happy to modify thepatch to suit the project +Same thing with repositories list +Cant +Could you please retest and eventuelly provide a new patch? +But if I can get some color around those concerns Im ok with the patch +Verified using IE and Firefox +Thanks for pointing this out and describingclearly the problem and a possible solution +Ben as long as youre ready to give a I can commit this +Uploading a patch that applies to +The issue is more widespread than just html sources +give me a feedback when you got the answer so i can close or resolve it does not appear to be fixed that I can tell +JBESBGACP +If we have the metric we can display it in the UI which will allow users to see if shortcircuit is used or not +Why cant this tool backup the disabled table znodes too +I wonder if itll break anything not to send the UPTODATE ACK at all +logs and db reproducible +Can you please help me in debugging this or give me a hint that what could be the issueRegards Gagan yeah I will take a look at this and get back to you +I could do this pretty easy I think for response entities of type etc +bulk defer to seems to be a general Eclipse issue +What version are you assigning to the added packages? If you dont assign a version then it is which wouldnt match your constraint +The difference in the qstate is the Undefined gets mapped to Unknown +bulk close of resolved issues. +Sian patch applied at r thanks a lot please verify +Thanks for the patch Schubert +proposed due to results of discussion with proposed some logs in revision +Makes sense thanks +break the Make to NOT blindly heartbeat irrespective of whether previous heartbeat is processed or not +We are using re running out of system resources +do you have a quickstarttestcase? the testcase outlined in the summary of this issue works fine for me +I had to run it from the ANT from the RC disto exploded zip to let it find the libs needed +Yes it isnt in upstream it needs to be tested carefully. +Marking this as a blocker until we get to the root cause +Please note that the actual output I tried to show didnt display correctly because Jira changed the amp to an inside the description +For the print may be eclipsed by successive logs +Otherwise we decorate and stick in ours to the one provided by the customer +I believe this patch is ready to go inThanksJeff +Im not fully satisfied with it because shouldnt know about or node types +Basically an individual user is limited to the natural size of the queue they submitted +I committed the patch that I attached to trunk and x branch +This will be addressed through a separate Jira issue. +change to not use +Until then you will not be able to commit or use https to checkout +And after clicking on Approve order link its working fine +for +A major rewrite of the bookkeeper client for the delay Ill be reviewing this shortly +See my last comment and have a look pleaseOf course feel free to ping me for anything you need by me +In addition nobody has been able to reproduce the +Is there anything I can help to fix this? Where is the SVN? I could not figure that out + +Created to fix it +Hadoop was just released. +Looks like using htop I was able to identify the memory usage as caches +The project examples plugin is compiled correctly +Ok closing this bug according to evaluation. +Heres the patch +Although this could look as if it were a CLIRR mojo failure Ive reported here since it works fine with Maven +Thanks will commit shortly +Will first test then fix +I changed the name of the option to parse to be consistent with the naming +It may have been another problem I was having a shard server ran out of memory during the query and threw an exception to the distributor +Will add to Release Notes as a Known Issue +Why not to set them correctlyThere are examplesexample artifact domjdomjusing vesion EAP bom version example artifact using vesion EAP bom version example artifact using vesion EAP bom version example artifact using vesion EAP bom version example artifact using vesion EAP bom version example logjartifact logjlogjusing vesion EAP bom version example artifact using vesion EAP bom version example artifact junitjunitusing vesion EAP bom version example artifact using vesion EAP bom version And Im not sure about hibernate example there we have hibernate related artifacts with version +this should go under missed the resources directory when we created the new site +adding an artifact in VS but it is listed in +The above code is part of generic Dao and entityClass refers to the Class object corresponding to the entity being queriedupdated by that dao +My apologies for not cleaning up my mess +Going to commit +Committed revision +Verified on revision +Thanks Filipe! +This should be complete with +One should be able to invoke an RPC but not block waiting for the response instead using a selectstyle aIsnt it enough to require that an asyncRPC is more like a sendMsg in the sense that it sends a message without expecting any response? Such a design would satisfy the requirements of doing readwrites of user data +Sorry for thisRegards reset done for dejanb onLeaving task open as Dejan also needs a reset on +Verified in CR +And fileabsoute is renamed to file to be consistentI have added it as headers to the IN message using the Camel style patternThis allows us to let the ftpsftp components as well use these headers and thus the file language support as well +Thank your GiorgioIve checked the patch in +Triaged for +I reform it to string and split call on itIsnt it possible to represent such data serialized in byte code with call to unserialize method +After applying the modifications to the code base Id be able to verify that the test does not fail any longer +was a cautious change to solve a deadlock situation +when is this change going to be committed +I guess now this could be done using the event system. +Put these on for review and at least inclusion of the plain project source lookup which seem to be useful for any kind of project and without affecting users in a bad way +Elmer I couldnt reproduce your exceptionI did the following I checked out the tag and did a mvn clean install true I started a Whirr cluster with namenodejobtracker and datanodestasktrackers I cleared all my local MAHOUT and HADOOP environment variables and I pointed HADOOPHOME at a local Hadoop install and pointed HADOOPCONFDIR at the conf dir with generated by WhirrI then ran the following hadoop fs mkdir input hadoop fs put input binmahout seqdirectory input input output output charset binmahout hadoop fs cat output binmahout seqsparse input output output This work without problems +thanks for the report well have a look into Ive applied the patch +This would be a great project for someone interested in both zk and python support +Even if provider can implement by checking if there is target or not and return null object Tuscany will use the result of the bindingProvider invoke to inject a value to the components reference +to commit +binsh based script to validate on sunos script working on sunos on trunk will be in additional script in the distro +huh +Its not a null value but an empty valueException is correct one +the queryparser warnings were trivial just raw types that needed to be generified see revision +Is the fixed scm component already included in a version? If yes which one? If no when will it be +That looks even better on balance of readability and conciseness. +so we can close with an Assigned Field! +In this updated patch I made it the background for the HTML element that way we can move things around if we want +For what its worth the code for this was never in the code branch at all so didnt make it into a beta nor the release candidate +Linking in a new feature request for a real not matches operator. +Per the JIRA triage meeting on Oct Santiago agreed to review the patch and see if it is robustIf the patch is not good then it will probably not be +I guess I just cant keep myself from running a build one last time to make sure it works +NN will never exit safemode by itself +Im not sure I follow you Ted +Yes it is a problem since they are not doing it manually but EFFs popular HTTPS Everywhere browser is doing that for them and they often install this to have better security and privacy and without really understanding how HTTPS works +However it doesnt look to have made it +Ill take a look to see if can be made clearer andor reduced in level where appropriate +changes localized to inner types so will only affect any users working with inner types in groovy. +Im not one for rushing discussions but were getting close with which this is a large chunk of +Stop making this about CDH +i think this should have been closed long ago + +Now the targetn namespace for the xsd that includes these types will be derived from the package name of the type and appended with xsd +Probably need to configure a basic terse logging configuration as well so that logging doesnt affect the perf numbers +That can be done already providing you manually create the XML Catolog and place the common schemas and in the right place and define the appropriate JVM system properties +Graph of current data of axis memory use compared to memory use +Are you going to support these also? Email or forum thread please +if i could touch a small helps this would have made happy me thanks for your motivations and supports special thanks to Lewis +Thanks Chris committed to branch. +Patch adds Debian Packaging stuff Uses setuptools and Generates the Documentation on the fly Adds samples as examples to springpython doc folderFolder packaging will be reserved for more packaging related stuff eg +The classifieds are stored in one table and the validity data in an inner joined other table +The patch adds the facility to set a parameter in the same way as all the other parameters +Ive mopped up the outliers so that the tests run cleanly on solaris on and +As for it being a list we needed multiple qnames to grab +Verified +Is there any workaround for this without manually editing generated file +I moved the files to their locations in trunk and included all the dojo files in the patch +I vote for option aThe patch looks fine to me +This may be due to a misunderstanding on how the alertdef add command works +ive updated the systax to my knowledge +Regarding the ZK failure that is effectively the trace Im seeing on my end as well the log makes it clear that the ephemeral nodes get deleted but the test still fails when creating them afterwards +Attaching patch fixing the testcase +Patch updated to trunk +Hence it affects +then can you elaborate a bit on I wan to replicate only part of entries what else are you expecting to be replicated or +If dont believe it just benchmark it and make sure you have at least two threads doing the update so jvm doesnt elide the locks + +Good call let me look into that +I would propose that this be considered a specific case of the general problem of form fields in cases where the form is not being in response to a validation failure + +Garren or I will get on this as soon as we have time +Hi PlamenI quickly gone through your patchPlease find my comments above + is now only in lib with EAP +This is still a problem for Tiger but not for Leopard +Group membership should be good to it workedthanks! +Should be ready to rollI talked to Camille last week and it sounded like she was on the verge of committing it +war that I used to recreate this problem in geronimo +This is a flawed that is doomed to fail +All patches submitted and committed +Some coverage of sorting for key names would be worthwhile but the reader can make the assumption that keys are sorted in Java sort order for Strings +Ill wait for Henrys comments on this patchHis preliminary communication is that this patch is OKbut it imples that we drop runtime support for and moveto and up +bq +Yes Ive also set it to false same problem +HiYou cant remove whole +Looks good +Updated patch +Closing resolved issues for already released versions +how will the old extra bugs be cleaned up +Oops! I resolved the wrong issue +when sending a new customer notification it should come back to alink that lets them change the password + is a separate classloader that is used to create an arquillian archive or discover an arquillian uses the arqullian directory to avoid the Windows jar locking PR moves this directory to the +Curious in taking up this task +kar package? +Need to add tests for JPA annotations +I just tried it out I dont see the issue either +Looks good +Please verify against using the provided example or the your own tests +I have reused zk as waiting period for zk connection establishmentPlease review +I had forgotten I was a part of those discussions Paul +is too long to do a comparative run +For testcoverage I added a codec that always uses noCFS +After resolving the issue I was facing when is trying to apply the LB rule even though it cannot provide the service it returning true +I dont need tokenizing at all! What all I need is to separate the records +This isnt the big IO reactor rework Ive been hoping to do but its good enough for this case. +Ive got the exact same problem +Thanks for the update +was adapted from of Hadoop for writing +I confirmed that removing the instanceof checks in constructor dont appear to have an effect +Does this issue have a fix version? Like +Unzip to a directory +Just invalidate weblog cache at the same time the entry cache is invalidated +Maybe there is a way to find the physical WSDL defintion that actually contains the WSDL artifact +Fixed in r +Added comments about locking order assumptions to methods and +commit fabdfaddfebbacccabb +Good community getting started issue +Additional cleanup +Where I almost always use emacs and so anything longer than characters is annoyingI generally like space indent but it isnt a huge deal to me +The whole parametric array implementation is screwed up and needs to be fixed +Screenshot showing web console with messages listed out of order +Should apply to trunk and branches identically +The correct method name for the DST bug is get in +This patch removes the restriction on select on hitting prefixed request handlers +trunk I added a test to show how to send back a reply to a which is based on a temporary all resolved tickets from or older +The test failure is unrelated +So if were closing in on consensus on the syntax the next question is how do we represent this in the schema? Do we even bother trying to make it fit the Thrift schema api? Id take at this point if that helps us get it into +Whats the problem with vs +Yup you hit it right in the nail +Release Note text edited closing JIRA again. +condrestart works as I expected and the other commands all still work with offsets provided from all sources +Tested in an environment where I reproduced that issue without the patch +The intention of the license file that comes with the DB Universal JDBC Driver is not to restrict use to a personal or home computer but to allow you to use it on a personal or home computer +Would it be possible to factor this out? For example just do a catch in readPagewritePagewrite and do the processing of the in a separate method +Looks good to meRegarding implements Iterable since thats only in Id consider it an implementation detail that we can change without problems if theres a reason to do so +They are actually integration tests and running them involves many steps building target eclipse installation with all the dependencies start eclipse prepare test project wait for background processes to finish and etc +The first option Stop copy process does what you want +The is loaded using a and there is no reference to from +This automatically solves issues and +Could you create a trunk patch +here is a first patch that rewrites to if possible +We welcome any patches or test cases you care to contribute +merged both PR in JBDS Build id +Resolving as invalid as the regex code is not in the repository +Having a default precisionscale might allow to continue to work with decimal types to some degree +Fixed in r. +I attached the changes I made to your code +If youre recompiling the code wo checking in a fix to bump the revision number then there be dragonsThe motivation is that some users dont run the same version of linux on all the machines in their cluster +Were also planning to use the opportunity to restructure and rewrite the existing documentation form +Thanks Zhijie for looking into the patch +This passes all tests +Check the order manager application +How does this work in? There are no features to remap in either +You should also include your platform what version of the installer youre using and since you know there is an issue open the log file and paste that into the bug as well. +I should have time this afternoon to investigate more and find a fixFor moving ivyde to svn the problem is that still uses a build which can be used only in jayasoft environment +it was just to log in this issue my comment as a reminderrecord for future reference +Uploaded patch for trunk +Yes the file will be uploaded to thetemp dir no matter how large it is +Just a Im running KB and dont experience the perhaps my source was hosed +Thanks for the feedback +upped priority since it was a hang but I there has already been a fix checked in for this issue +add solr authentication +Then whats the syntax to access it? +better unit needs to be changed on the server side +Thanks for getting back to me. +The resulting type alternatives implementation can handle all the production rules in the grammar stated in the spec +This worked for me with and without the patch actually +I just committed this +This is an issue only with and odd numbered release of Hadoop +There was one without Consumer and Producer but with counting +I will try to get a test case when I have time but here are some steps to prove this +applied patch to at r and trunk at r +I also think that it is a separate JIRABut lets suppose that we decide that a server boots up requiring usernamepassword encryption also +from me did you want to test Brandon +Command line argument for origin property is the feature id like to get +cloned to and closed original issue +Bug. +In our case every reduce task failed and therefore the entire job but I can see a situation where only one TT machine had a bad hostname and therefore only a subset of reduce tasks would fail and the overall job may still complete +in +I havent looked at the code yet but the feature seems very useful +I think that a site rebuild will solve +Patch with comments incorporated +Let me know how it goes +to take a long as an argument so they can operate directly on an input vector array element with no type cast. +It is a test problem +Confirmed on master +Does it make sense? If so what we are discussing here is not an issue unless you want to have a mix of and servers for some other reason that is not a rolling upgrade +Hi Andrew Ive committed the patch to trunk +fixes a bug for shipping with the job +Therefore its usually synchronous + +We need to remove this patch temporarily because the changes to ties to the new container which prevents us to release version of the think this can be included back now that has its own lifecycle +Nicholas could you please create an version in Jira if you have permissions +Some of our costumers are using EAP and our application needs the TLS cipher suite TLSRSAPSKWITHAESCBCSHA which we implemented in an own security provider +So in we can be sure that deletes are to be applied after updatesI guess my implementation is not safe Shai Mike can you please take a look +Seems to apply also to reopening +Hang on let me download the plugin Ill try and let you know asap +I think there can be some parameterization done in the and classes to avoid so much similar repeated code but that is for another day and not within the scope of this jira +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +This is no longer a problem +and the parsing code was developed outside of Grails w Groovy and the top jars mentioned above +Hi Arpit we already filed on failing +files for every class within a package +that code is committed +Thanks +To be reopened if any further issues +Sorry David I should have tried that +the timeout should start after the threshold is reachedIf namenode is started long before datanodes then as soon as the threshold is reached superfluous replications will occurthe timeout should be short Id say less than a minute and should start as soon as the datanodes threshold is reached to allow stragglers to connect and just committed this +should have a outputDirectory and project field move the loading of the reports into the doxia mojo out of the expression evaluator and core due to the way locales are done this will moving the output directory creation into the abstract report where it can concat the the separation +What level of Java are you usingAre you using Maven? Have you upgraded any of the Maven pluginsAre you using a user defind or for MavenBTW I have also tried buidling on a Win Server w Sun wo any luck so farWill try building on or SLES later today +Fix in place everywhere Ive been knowing was string is now correct version as. +Testing note the data for tests copy and copy could easily be expanded tohave outside the document element + review found this problem +I think of something else now in the printer to include first a summary of the threads without the stack traces and then detailed dump +This sounds serious until we triage it lets mark it as a blocker for +Note that this is a regression of in looking at the code it does seem to support both method and value expressions properly +Heres the third iteration of the patch +Roo does support composite keys to an extent however the MVC support is very limited +The test result is the followingCaused by Duplicate entry for key userid +This is extremely problematic +This should helpfor a poorly configured environment +Fix in r. +Did you modify this the patch I posted? If so it would be great if you share that change +Nota bene that I think you may run into surprising gotchas when you examine how all this interacts with Hadoop serialization +The suggested fix requires you to directly manipulate the lang is not working properly +patch which resolvesimplements the described in r +I was looking for a solution ans I found about split the page sequnce how can I set the for pages? transition to closed for remaining resolved bugs +client commonlib and resteasylib +Also the code is completely different. +I apologize I send you the wrong key this is the correct one +The vote count has remained zero on this jira for long enough and there has been almost no community demand for this jira to be resolved +Thanks for the patch +After all this it appears my actual hibernate mapping files were not correctly formatted! Arg! You can close this bug +Is this needed because we have separated modules +Thanks Todd. +seems more important and hopefully not so hard to fixI move it into the final release +I committed this. +I will try that out and provide a new patch +Thanks Eric +Run the node with again and the schema converged +Yes +No response assuming ok. +Hi Daniel could you please attach trace log? Which Build version you use? was a typo in one of my post install scripts I didnt read the log correctly +Is there any update on this? I was looking to create a Spring set of classes and stumbled across this +From looking at the configuration options for TODO it appears at first glance all that is needed to support this is adding Groovy to the type list +testcase is still happening with a slightly different constellation +Committed to and trunkThanks Lars Bijieshan Ted and Matteo for the review. +I dont mean to suggest that the Lucene Query Parser should know directly about the structures such as the Solr schema Solr params and Solr Q Parser plugins but I am suggesting that Lucene could declare and support abstractions for those sorts of interfaces and then the Qparserplugin would supply all of them and pure Lucene application could optionally use them as well either with additional query parser methods or subclassing +Ill try this again but as an end user I shouldnt have to know what version of BIRT I need +Wei + please committed +Any objections to committing this patch? good +Reset assignee so mails go to list +For large chunks it is equalThis version is simply a term expansion of the previous one and it operates four bytes at a timeThere is a way to nearly double this one more time for large chunks but it is proving tricky to nail down correctly +I ll try to take a look sometime this weekRegarding matching classes this method is to be called on and objects and hence couldnt use instanceof +In the output for the original there are some bare n s and perhaps that is what svn complains about +change the method name +Working on is not so easy but Ill try to get it checked out and buildingAlternatively if you could just look at the recent patch and apply that to TRUNK then manually apply the equivalent changes to the branch that would save me a lot of time youre obviously already set up to work with that branch +attaching the the fix +Bulk update to fix version +Is this issue reporting a new issue not covered by and? If so please indicate what the new issue is otherwise I will close this issue as a duplicate +All client software should work with replicas but if it knows its going to need a copy for a while pulling one over makes some senseUnfortunately Lucene cannot work with replicas +should also makes this issue less likely +I have reverted the move and created a new httpclient directory under Jakarta +Fixed on the SVN trunk +I ran the performance tests over all Jackrabbit trunk commits since was branched +Could anyone attach simple war where this bug is reproduced or video to see how it reproduced +Some small fixed to avoid EQ warnings on +has been marked as a duplicate of this bug +Thanks for the suggestion +Retested with latest release +apbackup svnwc B Mar +Illcheck this +trunk Thanks Christian for the patch and unit testI have applied the fix for the other components as well +Attaching patch to address this issue +But in the project that I generated using the maven archetype I dont get this behaviour +I cant see from the schema how id of element is mandatory +We ran into this problem because we have classes in both a package as well as an package and only one set of classes was being file isenunciatecoresrcjavaorgcodehausenunciatetemplatefreemarker attached that will use string comparator in the case where string lengths are equal +Issue cant be considered closed till this issue is resolved +I have applied this patch +Currently I have wrote some logic and trying to come up with an API +Committed to trunk +I have tried to replace them with simple expressions which are really constant in the hope that the compiler would the constant expressions and substitute them in the query tree +Clean up thread was dead locking with journal checkpoint since the checkpoint is done in BIG transaction +If host and port are omitted they are assumed to be localhost and respectively +The and PreviousNext buttons have been adjusted for these deletions. +Closing issue as the version of the bundleplugin which contains the fix is now released. +How about attaching the Severity class as well as its being referenced +Ripped out some code into another protected method and tried to abstract out a runnable +Kris suggested change does not work for me so +The user vectors have nonnegative values and so does the matrix +The fix is to use Jansi for reading characters in WindowsThis is the default jline behavior +created for tracking common change +LGTM. +Logged In YES useridWell its have take the sources that I compile with a and Ant works fine including the bug is closed but I think there is better to put anadvice on the README file of the binary release. +Will leave open for now to see if this actually takes care of the issue weve been seeing +Hence I had intended to look today +Ok thanks Alejandro +For tools that dont allow static invocation +Yes but going off of Kojis comments we also want to be sure that if the previous attempts edit log does not exist we dont know what state we were in and we should just assume we need to unregister and exit +One unit test fail resubmit the patch +With is no longer in need of the missing messages. +That is not something we want to make any guarantees about from release to release +Im finally back from several trips and will inspect this tomorrow looks quite bad +Resolved in Savara which is aggregated into this release. +Yes the datasource is configured as type XA datasource +Perhaps this is a case where the and friends can come into play in Solr finally +ok +This is good enough for me +Change to and toMonth methods to simple static arrays +Add the visitor interfaces and accept methods +Thanks for making clear. +This will be an updated and javadoc fixes in the code tried to add your patch as well but it still complained +All of the above is now supported and available in SVN trunk +I think it better to remove it entirely +That is the base file +I will make an issue for this shortly +Do you mean the resources that a tag library requires such as css and js files forIf so that would be covered by in effect it +Can you check this workaround? In a can you set falseYoull have to use reflection to do this as this field is private +We just lack of control on the retrieve job +If no runtime has been created yet the next page shows the runtime information and allows you to change the runtime name +Thanks Amareshwari! +Because this is a hibernate issue a candidate to start searching in is the Hibernate Core User Guide +Thats all Ive got for replication at hte moment as I really dont remember how to fully run project archives outside of eclipse +No more NPE while displaying or saving deployment plan +Capturing the scheduling decisions involved when we are not able to find a task from a Schedulable because of lack of resources on a givenRegarding the latter the current patch just returns null which is similar to the decision used to take in previous versions +Hello GeorgeThe fix looks good thanksBest regards by Andrew. +Setting tos worked +I am going to do another more look to be sure +I see that you didnt commit changes in trunk +This in theory would break any browser +Hi ScottYes I agree why only removing Person since its tied to Party anyway? Chris? +Updated in revision . +It looks like I haven made myself clear enough +Attached the third version which fix the issue when pass the columns to inputformat and there are no clustering keys then the composed query has null as one of the column +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +What I mean is that I think that the mocks should be implemented in a way that they also work when mocking the getMethods explicitly vs +I read it again now and I still have the same questions +switch case still broken in Subversion Head is issue is a replica of and +Ambari needs atleast a way to get app state counts +JMX Dejan +Now I cant reproduce it +Less than hours between this issue being opened and committed on issues seems a little short +So I will post a patch this week +May I ask how do you sync? What does your cronjob look likeRegardsHenk rsync twice a day +Yep like couchauthcache does +Please add the output of unit tests to show success +We cannot to encode a special character using amp +same in the system folder. +Ill see if I can come up with something +Working with Steve to get this resovled so I can commit the new build files +thanks for noticing thisI think broke this will have a patch shortly +No probs didnt mean to offend! I just didnt want to see lots of work duplicated +Nicolas this is only for +Right now in trunk the way we initialize felix has been changed and now the container will properly shutdown when you leave the JVM +I will try to factor out essential classes into a Maven project though +Fixedboisvertsixtinesvn svn commit m Improve error message when JAVAHOME points to an invalid JREJDK installationSending CHANGELOGSending libbuildrjavaSending specjavaTransmitting file dataCommitted revision . +It is implicit currently lets make it explicit +the class claims to be immutable which it is thanks to the header array fix +this query does a lot of things right now maybe we can break something out of it +This will only have an effect if convertType is enabled and will only affect date field type +So we can consider it to be rich enough to be usedLet me repeat my second pointIf we try to codegen for the built in schema types also I strongly feel its +Thanks Matthias +scanned this issue again pre and it looks like the exception thrown from the RA cleanup method should be handled by the RA contract and the connection should be removed from the pool +Committed r + Environment configuration Running application as Denises flash provided Running upplication using Screenshots making chapter The CRUD Application Walkthrough writing +Thanks +I have to catch a flight +My concern isthat it needs to be discussed somewhere +Updated patch attached +Hey Mr +Removed workaround in revision a package has not been found during a class loading attempt and a bundle is resolved which exports such a package wereload the classloaderChanged in revision ve improved the reloading behaviour of the classloader in revision to only check if a bundle is used if a bundle unresolved event occurs. +Its been a long time for this one +all nodes have possibility to run cvs +Patch v removes some whitespace from An else keyword is removed due to the return statement in its if block +Committed to trunk and +Wherever possible we should strive to reduce the amount of code in the base mapreduce system +Theres a definite cost to keeping the components in sync but theres a strong benefit to making it easy for people to interact with Avro through Pig right out of the box +Ahh thank you I will check out CVSHEAD +Tested the patch manually by running committed this +Patch which adds static final Object as a lock. +The cached read transaction view also had a txn associated with it +Here is a patch that only includes +Yes it will affect +from me assuming we get the python driver kinks worked out and pending Erics +Relates to as this would have prevented such an issue from This was also pushed to but the wrapper has never been released in code yet +I have changed the code to useclosesocket +Imran thanks for finding this fix commited at r please verify if its resolved as expected +But by default we should keep the existing behavior. +I also think it is bigger picture than javadocs and spans multiple packages so it doesnt really fit as nicely into the Javadoc model of package level documentation +Does this mean that there is leaks +Woosh PIG is in so Id love a more thorough review of this specific proposal +I think we should should add them to +I agree with atm and others +I can take a stab at this +The combination most likely wont work with users +new key installed and verified to work. +This approach puts the caller in better control of when the refreshed data is acquired +Updated patch +I managed to wrap this up +Heres a patch that fixes the problems with the testcase XML master and regression test suite both work as before with the changes is only a minimal change to solve this additional work for cleanup andor design improvement is still necessary +lukeall JAR embeds Lucene but an older version +Also deleted old patches to avoid confusion +duplicate of +We will get it in CamelHowever we could also get it into a separate component on +Yeah I think I like this ideaWhat do you think +I couldnt reproduce exactly the same exception but numerous in and in were solved +So all issues not directly assigned to an AS release are being closed +bq +They allow adding aspects on the implementation which should rely on a concrete class or a if needed +So if we cant guarantee the correctness of the data we shouldnt cause a state transitionGiven all of the above at least one of the fencing methods must succeed before the ZFKC can reasonably cause the standby to transition to active +Mike good one! Would having a doc id stream per thread make implementing a searchable RAM buffer easier +If you have RC plugin you just need it to extract them and put them in the eclipse plugins directory and it works on eclipse +We will write a FAQ describing this +Im attaching and a revised version of with the correction +This is no longer valid with SCA promotes JIRA shouldnt have a target release +But this is just a tried passing a JSP to the input field but whenever the validation error occured all attributes I had put in request scope were gone +This can be fixed without regenerating the signatures +Id be willing to open this bug again if someone is finding it impossible to do as suggested here +Is this something we have to do on both the maven server and in the source? Let me know what has to change in the source and I will update the patch +Added tests in test fails when it is built in unix environment +This has been observed only on sort Ive tried to reproduce this on smaller clusters +With this patch I was able to run the job on a Hadoop cluster +For most cases we could even do neat stuff like wait for all scans to cease for a tablet before we migrate it awayCurrently when a tablet is closed it interrupts running scans +Thank you for the feedback Knut AndersAttached and committed patch c to trunk with revision +do we need test casesYes please +Patch for renaming applicationsecommerceconfig to applicationsecommerceconfig +I finally got around to try it with snapshot and it +Going to delay this indefinitely or until we find a nice in memory implementation +Hi MirekAbout the optional attribute it means this attribute can be absent in the configuration fileHoward +for added mockito to ivy +Clearly the function getvalues needs improvement for handling invalid inputs +Alright so I can just use the ivyconfigure task for now and just ignore the deprecation warning +Can you help? The current solution only filters in the version directly into the default Velocity template +Yes link other issuesYes patch is available +Setting for move to if its not urgent for the next release +patch applied +Attach is the patch for this JIRA +This is causing some of the unit tests to hang +Patch for in +bulk close of all resolved issues prior to release. +Transitioning to closed. +Daniel I think the problem was addressed so Ill mark this as resolve +The patch +Updated the overall patch for convenient review. +Ideally would be renamed to and moved to the new util package and would be moved to the util package as well +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +In occur the same in same issue is a duplicate of +Gilles will you be able to update itCan someone familiar with Avro C please review this? Douglas? BruceThanks! +I think this is still a problem if you have to a class and you are not able to put the embedded class at the top of the file +It seems full implementation of get will also require at least get and overall support of multiple resultsets per query +As discussed with jbellis I will try +This issue has been resolved for over a year with no further movement +Fixed in trunk +Types defined via typedef will be generated as expected after this patch is applied +WOW thats really a great improvement +so putting this one on later until we have more infosolid plans. +cloned fieldInfo to fix overwrite issues +So is this an issue with EAP now or are you filing it against the product as it will be an issue +SureshAccording the original forrest bug disabling sitemap should be enough but just to be consistent I uploaded a new patch which was equivalent to the original patch +Closed after release. +Jon my bad on the label +We are in a case where this kind of supervision works pretty well +Could you please resolve the conflictI do also suggest renaming the test case class from boot to +Patch committed to SVN trunk and HTTPCLIENTBRANCHOleg +Attach a patch +Its a small change which doesnt affect functionality but very useful for EL extensions +Ive tried it using the snapshot and have attached a sample project with no URLs but with an inherited menu that now seems to work +Add patch for the branch +It may be that Jonathans original examples I a hores dara no veiem cap mena diniciativa per part del seu hores dara el codi encara no est disponible per sanuncia que es publicar what Esadir calls s impropi +Interesting +Nigel in order to make our job not fail we had to continuously monitor the job and make sure that any bouncing did not come back before any reduce task ran out of attempts and failed the whole job +The proxy should refuse to boot if this parameter is undefined +I believe the patch completes the future sorry for the confusing comment +I thought that it wouldnt be much of a problem for the moment +May be some change made to wsdlws tool fixed the problem +sigh +Please check that the patch was applied as you expected +This was done as part of. +trunk all resolved tickets from or older +Hrm +Finally cleaning the Configurator and of. +Puts are already done and are not that complicated Deletes a little bit more complex but also done but are maybe going to see rewrite + +bulk close of all resolved issues as part of closing items. +This issue was fixed some time ago in the commits fbfdfaed caaedccabbccbbe cfdfcdeadbdceac and cbfadfbebafeac +I remove old path to avoid at +Did we had the vote on the contains or begins with defaultI would also opt for the contains default +For example there should be a DTD for a core namespace which defines the properties of all the builtin components +Regression as of r +Im not quite sure that we need to override that +That way any env variables set via will only be applied if they arent already set in the environmentThis should allow apps to override the variables I believeI thought it was a little goofy to have put specifically called out for certain env variables and not for others so I put them all in and made all variables put +so that is what this bug is about instead of just template for uiparam correct +since Axis is going to base on java or higher this issue is no longer valid. +How to solve this problem +In regards to recovery from a user perspective I think that different people may be responsibleauthorized for recovering transactions in the management system versus in the application +The Javadocs have been extended to include details on using localizednotations in resource bundles +Implemented the change in Rev +The pom which allow to build the Base Driver with the cyberlink library require the linked feature to be POM +I dont understand what you suggest here this sounds like you want to change how eclipse works +I personally like solution recompute the checksum for the partial last chunk +Marking these as fixed against rather than against. +please update the refdoc related to integration +merged in masteralerts were sent for alertType so changed the condition to alertType this bug after verifying on the latest master build +might be feasible +Tom we can push this in the release if you think its good enough +Have you ever asked the alfresco guys about your problem +This was triaged for SOA +Hopefully it will help. +As you can see from the output in comment number this affects apache maven +Should the documentation team be notified to propagate the info further? Or how is it supposed to get into the docs +Thanks Ashish! + is indeed the latest officially releasedversion of Validator +The second page is inherited from the that our wizard extends +The changes to make options nested breaks these cases +ps fo file i used produced this problem because the bug is in theI also attached an extract of the generated +Attached simple repro +Looks like Sid is not seeing these failures +This isnt just about Demux boundaries wont this come up whenever a user tries to visualize a job +Ive tested with Jackson and this kind of configuration works +I did not find the Jgara licence yet but for liberation we cant embed it in as its GPL +Thanks +How about making a separate page for these sectionsSounds good +I have installed on the server I hope it will help +Parameter name and changes are in testing +The test case was very helpful +Attached patch updates the metric names to conform to just committed this +Thanks Nathan +Committers can you please review +Issue in PBR as well. +The old deployer just deployed into some random directory I believe? Or did it actually integrate specifically with the WTP provided tomcat deployer +About second warning we need to depend on JDT because we invoke stuff like project rename from refactoring wizard +That effort was being held up pending the Math release but is out and so should be moving forward +Also there is another JIRA issue I submitted earlier that this is a duplicate of +Its not surprising that the tests store to be trueI was referring to your test environment +Heres the patch +Did this ever get installed +Because if after updating the META entry to ROOT if transitioning to OPENED fails and even if closing the region also fails any way till the META is available the system is not going to function +Sorry about that its a really busy quarter for meCheers and thanks againSteffen +I added a unit test to test the toString generated when using named parameters in the constructor of an Immutable object +My fix should take care of the current requirement and I tested the change on IE and in nightly build . +Im getting fairly good results from the test suite +Manually tested +Also a necessary unitest for the has been implemented +However when invoked from other locations such as the outline view this is not working +Pushed to master and on JBDS Alphaa. +Ill check later +Seems that DEFAULTDIRECTBUFFERSIZE kb is the best option here I was playing with it for some timeCan you please apply v patch and test it on your machine so we can see results you will get comparing to your previous comment? Would be very helpful +It is not a bug +You can do this yourself using the Configuration API +It seems likt this bug only happens when you use Connectors not Acceptors +Modify the deploy location Add the module to server publish and start server +Thinking some more about this bug I think that a possible workaround would be to disable JVMTI events down the stack when a Java method is called by JIT compiler + +What if when a HRS came in and if it has a MSGREPORTPROCESSOPEN then we do not do balancing +So I am reassigning this bug to JIT to complete the work +I tried to put it in sync with +I understand the pain in the workflow but I think allowing to load it will lead to other side effects + fixBefore After fixbefore after +Update for solr Download Download most latest mvn package will generate files under target foldercopy them to solrlib cp target Downloadssolrlib call ant clean dist to create a new file under solrdist have installed complexphrase parser and followed all instructions for installation +About the write buffer Im not sure how we want to do since I think that we are gong to be able to mix calls get puts and deletes in a single multi call so we have to decide if this is something that makes sense and in that case maybe not use the write buffer +Reporters of these issues please verify the fixes and help close these fine +Thank you Chuan. +bumped to to LATER since its very unlikely that well have spare time to implement this even if this would be very handy. +The todo is gone in Chromattic release which should be used by Beta Can you check which version you are using of Chromattic +We dont use internally +Testcase to reproduce this problem +Sorry I didnt meant it to have priority Major +FWIW I have never understood the distinction between the NOTICE file and the LICENSE file +The real change that broke the pooling behavior was +thats exactly what I thought was missing and what I needed +Jason! +Markus please assign this issue to me that will serve as a reminder that I need to review it to Julien for review +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Committed to trunk. +with rev +This is patch is actually for flume and not for or Just uploading it since i have it and in case some rare soul needs it +If you examine that stack trace you will see that at no point does BCEL try tolook for the class in the Repository +I hear no complaints until then Ill commit the patch tomorrowcheers Roland +Thanks Zheng! +Im still getting used to the gitapache workflowIve rebased against trunk again and tested this set with git am +If you can yes thanks a bunch +Closed after no further comment has been given. +Owens patch +They dont need to be long running +I can probably play with this tomorrow +However in that version of the set of required dependencies changed quite a lot so we decided not update until Maven Embedder is updated. +Hello Haraldthanks for your interest in +Will append patch shortly +Here is the complete log of the test run +It is not a full patch +Probably caused by +Finished in branch of course +Baptiste any news +Hi Suresh +or the should not stop at the Exception but ignore the problempom and continue the build +According to the docs for get adding the group is optional +If not +Why are you catching and throwing the exception as an? If you just let the method throws Exception shouldnt Surefire do the needful +Applied patch into trunk and also update the jms wiki page +Patch with some aesthetic changes to the code but more important SNIPPET comment to start documenting the Scala DSL on the is applied with thanks to Gert +I havent had the opportunity to look over this yet +Sorry to waste your time on this issue feel free to close the issue +Patch v adds the proposed methods for testtest now creates table with multiple regions and verifies that each of the two custom coprocessors is loaded once +Can it extend +Hmmh +Here is the patch file in diff format for this check out the unit test that I added. +If theres a problem we dont say I cant uncork this bottle we say I cant open it or I cant get the cork out +Closed by baadccaeeeabcedd +Ive reviewed what I had previously done in my patch in light of the additional comments and the only question I have is whether the should have a call to validate the endpoint as well? If so that the only addition thats required +Can you please verify? Thanks! +The only difference is the OS and java version AFAICT +from me +This should be fixed in Site Plugin +For example to be sure one from Europe and one from USA +all passed in the latest run +Thanks Charitha +If you need to split a message you use a splitter for that you could then route by sequence number +I would assume that you can debug issues without having to reproduce in your own environment for these types of issues +Roo does this by adjusting the JPA and the configurations +is basically a dupe of this but both have good comments so leaving both open for now +package names it looks like these were built with an old version of CrunchIn any case that doesnt take away from this being an issue +For now as long as the project server support jboss web serivce even if the project has not jboss ws facet it is ok. +A middle or goal to get rid of the inline scripts in Wicket would be the better way +flagging this as relevant to compatibility and documentation since it involves changes to the Tomcat instrumentation customers have deployed +Changed Target Version to upon release of +Ran xa tests on Windows and ZOS +Though adds a bit of new code it is advantageous in that administrators are informed very earlyAfter discussing with RajivHemanthSreekanth the later is agreed upon +Committed to TRUNK +For exampleclick would override the file deployed by Click itselfWith this change the in the Click jar would be deployed over my changesAnother thing to consider is IDEs +Anything happening here? We are using Axis and still seeing the issue +I thought I had submitted a patch to update the dataScroller docs when I had the same problemIll make sure to change it right now +Can you? I will not be able to do it today or tomorrow +Is this going to Hbase +I am quite sure that it is an excellent piece of code find the new skript hope we will make it to ibiblio now. +Broadly looks OK to me +This is a namenode GUI change and does not seem worth it to add unit tests to this +I just committed this thanks Cheng. +Im still seeing when I run with JDK and EAP may not be a problem after all +Quickstart to demonstrate that POST params are ignored +In both cases youve buffered IO because of the OSs file system buffer +Until then we need to skip these tests. +Thanks for clarifying +I meant that we need to create a new Evaluator which can escape query characters +Freddy just in order to let you know during work on seems that Ive found an explanation of this issue +v version and check Grant license to ASF for inclusion in ASF works with minor changes in executeDescribe method error messages and help +Hi Denis could you please set this property in the build so that usage reporting does not get triggered in our tests +We should look into the new functionality in that Mylyn and WTP Server adapters are using for downloading this type of thing to see if we could adapt it somehow for downloading a runtime +Until we deprecateremove the setting we may internally try make the conf setting a final to see what happensWill address findbugs after some webhdfs firefighting +Perhaps open a ticket for it +Now the complete geronimo supports the Simple type web serivices +Sender will keep the file reference until receiving ACK and when it receives RETRY instead it then sends the file again +These tests fail wo this patch +Please let me know what you think. +That wont be this is possible with IE by opening multiple instances and only interesting for development can understand not allowing multiple logins but can we at least clean up the error handling a bit? +Running a full test run again +even without the interface annotation support which would be a great enhancement +In a previous life I used systems which has multiple ports open for the same protocols and relied on the both hardware and OS queueing to make one port a higher priority than the other +Wayne It definitely sounds doable +Implemented and committed to all resolved issues due to a successful release +Agreed Earwin lets first see if it make sense this is just an experiment and might not make sense for single threaded writes. +And add a note in the AS documentation saying Note that the previously supported HAR deployment has been superseded by the standardized EJB Java Persistence deployment model +Ms too close to do that +I have fixed the enumeration supprot +Fixed in revision This issue is a regression introduced when I fix +New features and some javadoc and formatting up initialization of I dont think Hudson will give this a because of the javadoc errors already in trunk lets give it a shot just for grins +Split out the sunosx platform into so this issue now covers only Solaris on sunossparc +has been marked as a duplicate of this bug +class property on link tag +Ive just built from master and there is no JBOSSHOMEgateindeployments and no JBOSSHOMEgatein your scripts +Will mark as closed but may need to reopen if there are any additional issues +if observed again. +This is the basic example of +Most any situation where something is loaded in memory is cheating but a very useful speedup since it works fine up to merely huge numbers of items like M +No nothing special is required +yes sounds like a good idea now needs some ressources to be done. +Cumulative be happy to commit this +However trying it on the Availability metric did cause the error message described above to be code was programmatically passing parameters as resources instead of resource types +I think all we need to do here is dont run upgradesstables if the kscf heirarchy doesnt exist already for the system tables +Uploading updated spreadsheet to describe patch for +my mistakedeploy doesnt should be able to runcleaninstalldeploy in sequence to get what you want +Because many of these systems roll their logs to avoid filling up disks we will probably need a protocol of some sort for the container to communicate with the Node Manager when logs are ready to be processedAnother issue is to allow containers to out live the app master that launched them and also to allow containers to outlive the node manager that launched them +We were able to use the to enable support for our date fields however its not very +Still I think separating out single and collection values would be a good move +I think javadoc is run in a child process and does not inherit the parents classpath +Its a problem in general but virtual nodes makes it more likely that the stealing happens + +The relatesTo value is not recognised and the message is not a response message to a operationAlso included are the tests previously described +It sets both the original value and it also sets the deprecated value +Now it doesnt need to do that it can simply lookup in the phy + +All the tests eventually ran on the sort API +a couple comments +patch looks good +Patch applied in +Commit r has a fix for and testcase within the AS testsuite +Proposed catch +is configured with annotations +That exception is due to wrongly returning a singleton list of the empty string rather than the empty list but this was fixed in +Ive applied the patch in to add version to project names as thats the main use case for this jira ill closeOpen new detailed jiras for other requests +It you authenticate on a single node all apps on that node are unlocked +A test to test getBlocks +What does this compare +so my code being told to replace the replaced all instances with its new value +You can then attach your patch to the newly created ticket +Where possible I have recreated the issue running under both gdb and valgrind but in some cases this was not possibleTest Application Files MakefileIssue Issue Issue Issue Issue Issue Issue Issue +ve tried to apply the patch but I end up with several compile time errors after doing so +We should move to it is only a minor bug fix release +hrm moving convert and get to seems somewhat misplaced +Yes that was exactly the intended functionality allow for a bitset filter queries and a regular query +Nullability is handled correctly. +Marking this as low prio and moving it to since it doesnt impair functionality as it is +Thanks again for Tedstack for the detailed review +I think we must change it to limit itself by number of docs insteadI agree lets fix that +Added debug snapshots as of the snapshot posted +That does not prevent the passed kv from being changed though +studioeclipseplugins +Hi what is the state of thisCheers have resolved it instead of just writing a comment +embeddedDoes it happen with a standalone server also +that was fixed in Smooks which is the version in ESB +This is fixed +But if we should be using start instead then I suggest we make the change +Fixed in trunk +fajax is tag of JSF tag libs +Hi Would you please try my patchThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Here is a patch which I believe can fix the problem +As a result we are closingthis bugIf you can reproduce this bug against a currently maintained version of EJB please feel free to reopen this bug against that versionThank you for reporting this bug and we are sorry it could not be fixed. + fixes this +Integration tests updated +one test passes so things dont look horribly broken +So I close it as wont fix +Unassigning myself from these issues as I dont see myself doing them in the medium term future +Sorry for the false alarm with the bug report and thanks for the help +Only issues that are actually fixed should have a fix release. +Ive figured out the problem +Atleast it is valid for the current stable version Have removed the previous button from the installer frame for preventing the user from going back but i couldnt get the warning message to display +Patch which moves the implsCompiles and tests passId like to commit this in the next day or so +I just want to reiterate that an Eager load with Select works! +Ive made all of the changes except for the +This returns include if column is of interest seek to next column if the column is not present and there are more columns of interest after the current column seek to next row if the column is not present and there are no more columns of interest + +cfg extension is added to the of the property placeholderIt means that for the file will be etc +Postponing until after. +s Virgo which appears to be based on Tomcat +It changed in r to process the shebang line looking for additional options +Hi JanJust to clarify are you reacting to the patch for the parent issue or to the codeLooking at the patch again I think there is a bug for the case where it just takes the principal of the provider parameters instead of taking the principal from the received token +Its more a settings problem IMHO +It was getting set to the number of default options which Davids example was +So we need first to define the outer then the inner class +Updates should primarily be done by upgrading bundles and not war filesI do agree in absolute terms but we dont want beginning users of the Launchpad to have to care about so being able to replace the Launchpad war by another one from a newer Sling release without having to reconfigure anything sounds useful +The file contains a fix for this issue +Do you want to apply that patch to Beta +batch transition resolvedfixed bugs to closedfixed +For now I am running Junit via the ant plugin but this is not the point of maven +All the unit tests pass except +see comment LanceFrankThanks for working on this issueAny ideas on how we could make a junit test to show the problemThis would make it easier to evaluate the patch and possible to prevent regressions +Ill update and close as I make progress. +Please take the time to look at it +Trivial patch attached +This should be fixed now +The above patch also changes the test method +We will check our logs laterCan we use mirrors instead of mirrors +Option to only allow local access to control port +My code is up to date +Moving to +and increase the Google ranking of the blog +both of them. +add HADOOPCONFDIR to the looks like a good idea +last patch should get us down to checkstyle complaints I dont know what to do with these any ideasall unicode issues should be resolved +The first integration is committed with versions +flavio can you upload a patch for as well +Getting same results +Maarten do you still have the warwork from the build that failed? If so could you attach it to the issue +Not sure TBH. +Any chance of picking this up again? Id love to have this included with +There have been refactorings in +However the problem here is not that MVEL cant access the property its that the nested class is not public and since MVELs runtime exists outside the scope of the nested class it cannot access the nested class +However these transient files will be remained in the system until the snapshots are deleted +Low activity +may be better than nothingGreg small suggestion from next time you can use instead of +to trunk +Todd if the feature is scoped to then an alternative would be not to expose it in at all at this point so users have to cast to +to Coordinator Start Time must be earlier than End Time +Output file from failed unit test +Also verified that this patch allows a from hdfs to pick up a change from common using a local repo +BTW query parameters have the same bug +Times like this I wish we were using git +for update instead so it passes itNow skip it for +admin suggests an administrative connection +failed test +Of course thats because its not running at all but I think the reason for that is sufficiently covered by we dont need two issues open +The XML does not get equal attention and it is typically treated as subordinate to the I mentioned that Talends route editor supports the Java DSL Scott Cranton said something to the effect of makes sense the Java DSL is the one that everyone uses +then a patch for the script would be nice +EBM was needed since we could have time sorted columns and could still access per column data +Hey Suresh And RajuPlease review the patch +I think the problem has to do with how the page boundary is determined in thecode +Tests passed cleanly except for a heisenbug in +This import process created the Cars node and everything below m marking this defect as Rejected since it was resolved by proper configuration. +But maybe you were talking about the SSL certificates instead +Arun I dont feel strongly where streaming lives +We could always deprecate saxpath and leave it in for now but stop using it I suppose +VikasSharad +Can you try the attached patch +Like everything else in the project those users who encounter problems using with turned off can provide patches to fix them +Failing to understand how your system not being able to delete a Hibernate jar file is a Hibernate bug +Assuming that the zk instance is already connected seems reasonable to me +It would be better to remove the error on overflow and generate some useful output for the user +It looks like there is more classpath flexibility with Maven but that would involve making the plugin work only with Maven and forego backwards compatibility +Attached the patch for easier evaluation +you guys check out the attached patched file +Issue closed as there was no log supplied by user +I committed this. +smack doesnt have a special connection for google but only a preconfigured tried thishost login google loginpassword domain name google recipientSSL connection falseWith this configuration i dont have exception on connect but i dont receive messages +Thanks for the swift response +It looks like youre right that we arent putting them in syncList at all whereas we could put them in syncList in the case that they have length the calculated minlengthHowever its still the case that DN might be shorter than the good replicas and not included in recovery +I am splitting my patch attached to into parts because people want this bug fixed in and but dont want the import checkpoint feature in these versionsThe patch contains also the test which fails on the old version but does not fail on the new one +this ticket should be closed now. +Will this be tackled any time soon? There are a lot of projects out there and otherwise that rely on Xalan and are currently broken under JDK +So you see I think its dependent upon language and how you want the application to workOK indeed the issue is not simple thanks for the examples +A unit test just cant stress the namenode in the enough to exercise this race condition +This is correct behavior +Fair enough thanks for taking a look! +Web Service Binding Specification Version Section uses soap intents that are lower case +I have added this code as bench for future exploration +When expires then consumer fails and is not reconnected and if is set to then hangs +great! Too bad the documentation doesnt even mention how the behavior that most people would expect from a sequence operator can be emulated +This is useless to discuss this way +I dont know how we can only use a seperator not adding other palette entry +Used the snapshot +When a new user signs up it is logged to the Updates table +Cannot reproduce anymore. + +So its better that to replace all the jars +Ive seen this +Hi Piotrthanks I had received your CLA so integrated it +resetting P open bugs to P pending further priority due to presence of a has been added with description better FO test a bug behavior is according to transition resolvedwontfix to closedwontfix +did this query work before svn revision and fail after this commit +As discussed in April contributor meeting the best way to do this is by using recently introduced hiveobjectRef +You dont have permission to access rickmcguirejavamail on this server +Triaged for no changes. +Ive even considered making the client the default for all versions of Android +Here is a descriptionI have been trying the several settings for running junit tests in parallel +Corrected estimated fix version. +Perhaps a final bullet reading Keeping deployed content in sync on all nodes in the cluster via the Farm nd paragraph broken link to Determine a unique integer Scenario D switch sets system property should be capital D D switch sets system property Scenario nd paragraph brokend link to Install Enterprise Application Platform on all your servers Scenario pathconfbootstrap is nowconf changes have been made and links fixed thanks +Ouch that one was nasty +We can call the interface in the RVM identity hash maps and its also extensible if we need the identity hash code in other places in the future +Correction the patch created a performance increase for random +In that case we need to invoke security handlers before body based dispatcher +and failures are caused by +Still need to fix the file with snapshot case +Make sense some time the real error message gets lost in the usage delugePatch looks good. +From my perspective applying this patch resolves this issueAssuming that the Felix Shell bundle is installed a new tab Shell is displayed at the top +Still Ill make sure to do the best we can from Springs side for +Is it a fresh install of Tomcat or did you add some libs to shared classpath +I think this happened because the vmops plugin in the test setup is not upto date +No harm letting hadoopqa have a go at v before commit +Please note the inconsistencies in the assigned roles +Hmm +pushed in Developer StudioVersion Build id Build date +Please feel free to any of those when required. +Damn I didnt see that +The release audit warnings are false alarms they are pointing to javadoc api changes +These are minor cosmetic issues but requires retagging and work from the seam team +Intend to commit soon +Is there any difficulties to release the old branches from the new release strategyThanksRex +A common modeling convention is to label gateways with questions and the outgoing sequence flows with the possible answersAlso the naming of events is of great importance to the readability of diagrams +All packages listed in step should be in either or +Also to the current branch +I think I may have a fix for this but not having a test case to recreate it at the moment I cant be sure +Lukasz I no longer have the original source file to be able to do this from +This is the contact emailjamieand Ive added the optionThanks +I admit that the code may seem complicated but it is a complete clean implementation quite well optimised for compact storage and fast lookups +Deploying the jar in the server deploy directory should work +we already generate a timestamp and I dont think we want to support uploading ones +Committed to dont think this is a this would affect all of our MR tests that use more than one mapperreducer +Good catch on the bypass and loaded +bq +Thanks Nitay. +Also needs porting to txmanager trunk +It contains a highly redundant text file of exactly GB compressed down to K with bzip +Mildly related if only one partition of the big table is used as input to the SMB join there is no need to prefix the file name with the partition spec +He was going to try and look into it +This seems to because of a change to how Groovy resolves the delegate of a closure after it has been attached to the meta class +This doesnt need further QA +Do you expect the time to load the image and the time to do saveNamespace to go up by a lot with this change +has a better solution +Update Change dont pass in Activation when it can be obtained from another Completed under this issue number +Hence closing the bug +Its an erlang limit which can be configured +But if the user has no permission to access any repository authentication will be prompted +png image that I attached is different than the actual report chart I had to edit the +libavalonsystemavalonmerlinjars File or path name too longStarting Merlin SNAPSHOT +ant Dtestjaxwssampleswssecurity issues resolved in released versions +Weve had the soa profile in tests because we didnt want to buildrun those plugins remember they were mostly in master soa plugins were active in branch +That means that an can delete files out from underneath a cached and theyre really gone no? Stale NFS Filehandle exception +I will add some unit tests to cover the scenarios as well +I thought Adrians VFS changes fixed this +Fixed in r and r +Check latest trunkThis method is already in. +I committed the patch to to be picked up in case RC is re spun or in subsequent release +Thanks Nicholas +Ill run it through the unit tests and +the patch nice catch Koji! +Closing for release +As of patch addedbcdfcaedfaf that test now passes +Anyway put them together and do the comparison is incorrect +Since it is mentioned again and again +planned and should be tracked in +Done using run b c all g all AS traits plus Server home dir Server Name +Found the following issue which contains a patch for booting felix in eclipse +Attached patch that shows the issue fixes it +The master file of this test javatestingorgapachederbyTestingfunctionTestsmaster had same problemI solve it as the same issue for test program +Sounds like that will handle the situation I have nicely thanks I look forward to seeing it. +Fix on commit Id say +If you have a deadline for getting this fix done then you may want to have a look around to see whether you can find the issue yourself +For people using broadcast address there is no BCAlocal address mapping kept by nodes so changing the version stored in Gossipers map is difficult since you have the version stored as one and not the other +verified it on the latest build and found that now GSLB service for VR LB rules will be showing the state as up and effective state as down +If I add the panel with cforEach I cant use the same variable in a nested ajrepeat and I cant modify the number of panel without a page refresh +Are there any plans for embedding the solution inside of a release +Heres a new patch for this issue a bit more complete +I can not think of a reason to allow multiple vdbs +Are you intending to add it in +we call the method so we can replace objects while we are serializing them +We will look into it +Do you think we have to rename java classes too +Under target I found several directoriesI looked atapplicationcontainersyslogbut didnt find much clue +The type of the property in the metadata is annotated as so we are not properly mapping that to a value somewhere +But I dont want to delete it for obvious reasons +for +Version has been releasedMoving all related issues to the next version. +I am on vacation from to If you have any question on deployment and JEE bugs please contact Saurabh Arora or my manager Maruthi NuthikattuFor emergency contact me at. +Mostly looks goodDont like using Configuration as the user payload but this can be addressed at a later point when we work out a better way to configure componentsThe partition range is only sent along with one dependency completion event +But yea I did try it and it works +Once that gets done Ill upload the relevant artifacts including the pom to the central maven repoColm +Adding a modeuuid property definition to the modedefined node type seems to work around the immediate problem +installed w clover reporting as an external link. +In which environment does this happen +All I am asking is if this still valid or could be saved +Assuming lazy consensus I will change this issue into a story later agree +Patch in cvs +You can keep the same name for the consequent versions of the same attachment +It will be useful to have this configurable like JAVAHOME for containers separate from the NMs JAVAHOME or may be extra envs that NM doesnt need to be set withAlso another minor TODO is to verify that the streaming examples I commented about above are working now +Moving out of +Patch with test and fix +Priority in bugzilla was normal but there is no such priority in Jira so Im changing it to Major +just did a commitRegards believe the serialization as of now in Xerces SVN code base incorporates most of the XSD syntaxTherefore im marking this issue as fixedWe can make improvements to this as necessaryRegardsMukul +one that is loaded first and will configure your second configurer +Happy New Year +Those problems existed before and were solved before ivy +bq +With the new moin the actual function used for sending all should be modified instead and likely theres some other changes that need to be made as well +Actually the relatives seem OK +I modified to transparently load and index the super class of the configured types +Deleted the timer that would automatically kill any remaining cleanup jobs during shutdown +Hoss The issue described in also helps to some extend +When I unjarred with jar tvf and then back up using jar xvf the resulting could be used the test did not give the sealing errorThis is very strange does this mean the jars are bad? I welcome any suggestions on how to proceed +Added license to json file +Hope to put in a patch before the end of this day +What to do if backup is started in a transaction that already has unlogged operations executedIn previous discussions about online backup it was concluded that existing backup procedures calls will WAIT for the transaction with unlogged operations to commit before proceeding with the backup +performance gains found across the board increasing with the size of the row mutationEDIT should also mention that the number of segments initially allocated is controlled by commitlogpreallocatesegments in and defaults to +Can you reproduce this in a test case +with the current code path we can fix them through a separate jirai just wanted partition to be not created until data has been movedthe load part and the metastore part looks good to me +I just wanted to let people know +and done +This only shows when I kill the DN and RS at the same time +This should ensure all the base requirements for as are satisified even if the required archives core ui is not present +Thanks for the review Uwe! moving forwards +The description mentions that the Ant plugin has a bug in its handling of stdoutstderr +Building a new one to see if the problem is resolved +Interesting apparently there was a problem with bndtool +Yes this is fixed upstream. +Thanks! +Its explicitely excluded +This issue has been resolved for over a year with no further movement +However checking this Ive found that cascading WMS got broken in the series opened a separate issue for it +Its been annoying me for a while but never enough to open an issue Used to work though +And well add Eric to the admin list so he can take care of this himself +All wiki documentation has been updated +comments get doesnt seem to be used anywhere +Categorizing bugs related primarily to stylesheet elements etc +In testing with Maven they were identical even when the POM specified a range +Once we are more comfortable with it we can expand its use +Alejandro you giving a talk about Oozie during the hadoop summit? amr +No further comments closing as invalid. +Ive replaced the booth graphics with the latest versions +Should we closeresolve this one? It seems to be mostly getting tracked on +Test failure reports from from out of the box and Tested OK with same invalid schemas in CXF so the became more strict in the release +thanksList of things addressed with the patch +Not having a standard API to implement this relationship represents certain difficulties during implementationThe problem is that this relationship is visual in cases of creating templates of complex components and navigation on them requires the knowledge of the parents and children +youll see yourself if you try +This should fix the issue with F +generated in my working copy usingsvn diff +for the patch +I dont have the bandwidth to upload the newly rebuilt jar though so could you just mark this as closed when you get it in geoserver? +Still the best solution is to not create sequences suffixed with seq if possible +I verified the test passes by explicitly setting to simple +Probably Accumulo doesnt use a lot of files and therefore doesnt stress NNIt uses plenty of files and the number of NN write operations is becoming a bottleneck but the number of files needed at any one moment does not stress the NN. +Thanks for reviewing the patch Mamta +We were able to use the dynamic service discovery mechanism of Tuscany in our tooling environment +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks for the test case and dont worry that you havent had time to dig into it +Since the trunk code will be merged out to I figured this would be a good thing to do +mP +Committed revision Thanks Sean! +Can you check on that +I know this is old but it is still have the same problem using grails and +It is possible that we will have other similar reproducers for other events +I see problems with this patch It does not close any of the issues raised in the description aboveYou can see files in the ls but you cannot see file progress and you still loose all accumulated data if the file is not closed +Done +Will you please attach a screenshot of the wrap? What version of AIR are you using + mapred depends on core and hdfsNo mapred should only depend on core +Completed and is available in branch braches and is available in branch braches to branches as the build structure and code structure has merged to proposed solution binds concepts from the jbossws native stack to the general server Magesh Ill have a resolved issues for already released versions +pushed pushed upstream +My description explains more in detail why such an additional attribute is useful +Committed +Im marking this as done but I will open a related issue for unifying the two models in teh future +Is this addressed by the Always Build checkbox on the build definitionThat makes the scheduled build always happen regardless of whether there are any changes in not what behavior are you looking for? Im confused by the use of default forced and scheduled in the original description. +oh sorryThis is an error that happneds on the cocoon site +It is kind of strange to go to compareTo but then to fall back to equals +Committed at subversion revision +I am attaching the test log and the diff between MR and Tez runs for Join NPE in Operators Result diff in Join + +Logged In YES useridOk +jar +Looks good to me +If the target server returns an error response because of an intermittent problem the proxy server can evaluate the post condition and veto caching if the response from the target server contains an erroneous status +this is what you and Jukka decidedWrong +Have verified with this change? class should have set conf otherwise conf will be null right? While you are setting conf please use +Ive committed this to trunk and +Do you know if this will be an issue for either of these + patch looks good +That doesnt convince me It seems that this is a common use case and hiding the information in request attributes is not really user friendly especially as request attributes do not provide any type safety +In the future I will write to mailing list my fault +Please change to if you intend to submit a fix for. +This is fixed +I remove the assignment to me of this issue Joachim is working on a Torque based implementation of a that maybe will be able to provide even a javamail store wrapper +create should just work with no additional steps +Shift which will focus on clustering issues +That sounds like a good improvement to have wed appreciate the patch +Already fixed +Actually with the new server view this is happening also when clicking stop and the clicking start +KamilDid you have a chance to try the logj Rolling? Can I resolve this ticket +No worries glad we know why now +The timeout values for waiting on the latches still need to be updatedIs this coming in another patch +You only need to freeze interfaces that you promise to keep stableRight +Sergey thanks for your reminder! I agree the sync issues is necessary I will fix it soonThanks again for your helps! +Just to retate with the CP bits the process diagram is displayed with these stepscd bpmorchestrationant deployant deploy to the jBPM process view the process diagram +Set a breakpoint in constructor of in +Yes this is not fixed yet +Moving out of +It works fine up until Sept th +The is not excluded unless I specify it +thanks Lars! +Committed revision +Have you thought about how your locking mechanism might interact with that of AMQ +All subtasks are done +Connecting to a container via fabric and installing using the osgi commands is not a good practice since the will remove anything that is installed manually the next time it kicks fact we were thinking about completly removing the osgi commands as we did with the features command to prevent the users for falling into that bad practice +changed fixed fixedPlz review +With the test app you can verify the tried to apply this patch to code base but got source code out of date issues +Still not fixed in release +resolved in trunk and branch for on JBT +for the change +Many Thanks +I removed the username and password parametersI replicated the issue and now I have solved itIm cleaning the code and then Ill attach the new patch +I spoke to soon it does seem to be fixed in r will have to play with a bit more and see why i thought otherwise +Thanks for the patch Army +I think that it should not be error message +I slightly changed your patch I changed the package of the helper classCould you please cross check and then close the bugMany thanks +Ill now correct the Release Notes text +Did we miss Datetime +block the TT till it can be given a task from the job at the head of the queuepool +Thanks Kim for look into the documentation +we havent agreed that isScannable should return true + skin on skin on IE everything is the patched file +In my lines only go up to +Connector development is greatly simplified in and will not initially require a kit +This allows the decompress stream to read ahead so that buffers are available to read instantly +I could not recreate it because I was using +The new patch looks pretty good to me +Hey Len Max and NickCan I get some clarification on which this actually impacts +Now when shortening the method the error really goes awayThanks +next iteration +My screenshot shows this is simply not replicatable at all + +I thought its quiet useful to have this option for user to auto converted a query result to list of user objectsAnyhow let me read up little more on mybatis impl and then see if I can fully understand what exactly is that you are asking for here +This bug has been marked as a duplicate of +Example where function accidentally returns a function instead of an resulting barf +Looks to be inactive and resolved. +Committed revision on to revision on branch +Committed patch to resolve defect +This issue is closed now +does include and whether this solves all of the issue I am unsure but using worked for me +It works well +any chance for something like it for or should I look to doing that myself +sounds good +Cant reproduceThe test runs without assertions +No Xerces doesnt generate such a message +Otherwise there will be more change as you suggested earlier +Will you mind taking those changes out and put on in +On the trunk this was actually fixed by Amila in r +All classlibs vm tests are OK +Thank you for your contribution Carl! I have committed this to trunk! +The issue has been fixed in xbean so we should close this issue when weve upgraded to this version +Assigning to Jessica to make API call changes +Need to fix the messed up branch history +This issue is fixed and released as part of release +Agreed +I should have studied more deeply on this issue +let me do it for you +It is probably a Flex issue rather than CXF +my project with GWT SDK my pom file is correctly writeits just GWT Designer crash when i open a file with itIm waiting the new releasethanks +Dart any ideas on how to improve the popup menu behavior on Linuxno idea could we move it to next versionDart this particular issue will drive many Linux users crazy +There is no additional work when using project example right? I set correct runtime and seam and no other configuration is needed right? I am asking because it says Could not find datasource javadvdDatasource +class file caching but it still needs to be implemented. +there is the nd version of patch without can not put the context directly in the message exchange because all propertiesput inside the exchange message must be serializable else clustering failsUpon redeployment everything is recreated so I dont think how setting the pendingMessagesto null would solve the problem +I cant remember what this was intended to deliver +I will first correct the svn locations and the mailing list names under this JIRA and file a new JIRA for a site overhaul +i was just confused by the comments +Fixed it +fixed in revision Alexei please check that it fully resolves the +Also cloned to a designer issue for resolution there. +Since this would useful for a number of projects and we can copy the file as such +Ill raise a jira to track the unit test failures with IBM JDK +It needs to be ptBR +it might simply slow down other users who have highexpectations for compression +We may have to raise a JIRA against meclpse for example +Were not really sure if well implement this feature for now as it may be a source of confusion for most users +awesome job +Sorry for my type mistake +That is because APP is probably not a real user with a real password +I will try install this on clear eclipse installation and will inform you +Screenshot of the exception raised on scaffolding when it attempts to list the items of one of the hasMany subclasses and finds one of the others +The reducer triggers the resolver in a way that might not happen in ordinary file reading +Ive committed this +can you try this one? youll need to rebuild a local copy +Youre welcome! +But the feature cannot be enabled by default in the AS. +So I think the issue is that the comparator is sorting daughters before parents so we process daughter first then parent + test case with JNDI search ops as described in the issue +By default the checkbox is checked and the application name field is enabled +Is it ok to use the version? And Have I to use it both on client and server or only on client +closing this issue. +Committed revision . +to v patch +Yes its fine thank you. +Immarking as WORKSFORME until I hear more. +In the future consider forcing reply queues to always be allocated on the same broker as service requests +Indeed and if you are not running with debug logging enabled ivy just sets there with no info about whats going onThese s should be fatal or at the very least retry a limited number of times +Yes we did disable the +Can some kind committer please review and commit the patches +Following Markos instructions the build works fine +Should not happend but I will check my ide settings +Does it work when you run it from command line +Seems like we are running into this on +Thanks Bennie! +edell applied patch +I think the patch is fineWill integrate if Prakash has no further comments +Thanks Jerome +I see a stacktrace on stackoverflow but no code +This is a large patch +Could you please add a test case with this patch +review by Brian Minchau and committed to CVSThanks for the patch YashNow both interpretive and XSLTC have the same behavior of going into an infinite recursive loop +Therefore closing as wont fix +How can it be declared at the class level if its applicable only for FIELD and METHOD +For example includes which includes +This is what happens when you execute the servlet +Table screenshots showing problem in see the attached screenshots to see the problem +BULK EDIT These issues are open to be picked up +ill try and digg that out + +Patch to fix this bug +UI s not +The test failure is unrelated filed for it +Reopened because of patch +And that takes a long time of waiting for timeoutsPlease has anybody a for that except placing fake javadoc and sources jar filesThanks +This patch also includes the code changes needed for +I will set this bug to invalid as the bug is not related todirectory generator. +This looks good to me +binhadoop +Great stuff Evert +onmouseover onmouseout attributes on working also +It provides a very usefull overview of your build Maven changes for are extremely welcome and means that I can start using this plugin to good Id like to seeThe ability to turn off the drill down into the fancy Maven dashboard graphs +Changing title and version +Add a new configuration parameter where the do not need a lockThis parmeters default value can be false the current behavior isnot changed +Rebased on trunk and fixed to use option to generate a patch +Commit Subscription Listener the webservice stores notifications in a singleton and the portlet uses a servlet to read the notifications from the singleton +There was also an in issue identified in defect key server requires an argument that is fixed with the of +Yes that did the trick +Ill look into the bug as wellProtocols like testMerge are not currently considered mergeable but I agree it seems like they should be +So the wrapper class is to avoid changing tests? Is that a significant amount of cleanup? This approach seems hackish to me and threading the conf object through the layer looks +There is a file called in the root directories that identifies the SOA version it is also repeated at the bottom of the page displayed at localhost fixed in CR build this text is displayed SOA Platform? ESB Embedded ESB Server Version SOA Platform? ESB Standalone Server Version dot terminated with extreme +Still having issues with this too hereI was able to fix it in osx through recreating keys but now having issues again on ubuntu vm on amazon +Using nowPatch is against trunk rev Peters patch to remove invocationsThat part isnt actually horrible it reuses the instance so while theres certainly avoidable overhead in parsing and whatnot its not forking a process or anything too notable +Thanks Hairong Attaching patch modifying as suggested by your comments +Using them to emulate what is essentially a mutex is ugly and fragile as hellbq +Also modifies to fail on either write or writeI think its ready no nocommits left +for using temporary do you have the rough code available a sketch of the code that youre presently using? It would be great to review this before baking such functionality into itself +need review. +Struts should not duplicate the efforts of standards like JSTL +No problem I have a work around although its a hack +Mike I think the class is missing in your patch +This issue is not reproducible in either hadoop local mode and +That seems to be due to the complex queries so we need to investigate why some methods are so slow with complex heavy load the explain plan of FINDNODESBYPARENTIDCQ changes and the database does a full scan due to the and use a temporary table due to the sort by ORDERNUM both have been removed +Latest version of artifact sources and sample application generated by it +The node placement algorithm takes a specified node height and an svg height divides the svg into lanes and tries to place nodes into lanes so they are not overlapping +I would make them optional dependencies since they are not really required by actually Id would really want to make sure that the is not being used in this situation +I am going to try some changes +Resolved +quickly built a testcase for this though no patch yet the code is somewhat spaghetti +This is exactly what I tried as well and it works for me without issues +how popular is his operator in groovy +Im attaching a new batch file to this issue which I believe fixes the problem +I try to reproduce the issue forcing a failed feature. +I tested it on branchThanksArun Arun Done at at rAshish Vijaywargiya +Thanks Galder that works fine for me. +some people dontwant to have that feature on their portal for security reasons +as it would remove unnecessary code duplications +Its the same problem and if you specify a URL with an encoded space in it Maven will create a new directory with in the this bugs posting is a bit munged by the automatic formatting but all the information is essentially visible ignore the links and the bullets in that markup +put in the best can do for now +Any chances to solve it? +bq +bq +I did not review the actual content of each deprecation string but my perusal suggests they are good enough +Colin do you agree +Its the landscape ones that dont +The translations do not ship until after the English version +I would have to investigate it are heavily dependent on the mechanism because we are having a front end that is waiting for a response +CXF not only uses JUL they require it in some cases when you implement some abstract CXF classes +the TODO you mention in the means that the code isnt completed yet and that we in factshould use a namespacethe patch you added isnt a proper fix for the implementation of the query offsetcould you please file a separate issue for your findings about the namespace mismatch? new issue is at with the same patch attached +I fear it might be too destructiveheersen hersennoors norseen I like this approach better focussed andbq +Moving to is fixed in. +this wizard would be great please doing the check for the existence of key redhat artifacts in accessible maven repositories and if they arent there provide at least instructionspointers on getting it setup or even provide a way to update +I think a solid orange circle would make a good Suno and in the context of the other letters would be an obvious character +This would work perfect in JON to provide history of the levels over time +A GRECLIPSE issue +if youre implementation reads the ACLs from repository content i assume it caches the ACLs for better performance +Changed to minor +Unfortunately its not possible because the persistence adapterstore I tested with is a custom one outside the AMQ code base +In the future itd be great if patches contained appropriate test cases showing the failures prior to patching and success after patching +Fixed in close of resolved issues. +Bump to +Sounds like a reasonable plan +A slightly improved version of the previous patch +removed pasted wsdl will add as attachment for reference +Please verify and close +Cheers Any update +I forgot them +If there are any issues let me know before final is released +It also reproduces on Windows with XjitcountIll try to narrow it down into something smaller for the JVM team to work with and file a bug +If it affects more than one EAP major release please make sure it is cloned so that there is one JIRA per major release +But I guess you can specify the version you patch against when submitting the patch +Hi BillWell it seems that we should rather append the const prefix to the return type as you say +This should be added to the wiki page of Derby Testing when this patch is finally aubmitted +Unforutnantly Icant get the unit tests to run +The NM is the node controller and responsible for managing the node resources +The bug persisted +I also added more test cases to make sure it doesnt break other cases +Sorry for the inconvenience. +thanks +Should we attempt to factor out the commonality immediately or commit the overlapping code and then do refactoring as a followup? Im fine either way I can give suggestions on how to create the reusable abstract bases and where to packagename them +And reference it from this JIRA +Just the usual path for resources inside of jar files +The Field class is the worst of the bunch +Fixed in closing stale resolved issues +I guess these two techniques are complementary and can be used at the same time +For actual commits please check subversion commits +Thanks +minor correction in the test +Patch attached for Release branches and trunk +I think we should change that perhaps in a separate ticketIm checking this in to unblock builds +Verified this bug already +after you modify the file you can run ant documentation from solr and then look in solrbuilddocs in your browser to see the generated file +This appears to be fixed out of the box in Spring +Definitely this is a pain and is easily overlooked when you first come to issues that arent going to make it into to done +Just a note this only affects unmarshalling +Default main resources were not copied to classes folder. +For customers where this issue is a problem well have to tell them to set the local jndi name in their think it was originally considered a bug because the names of the MDB change across restarts +META +targeting for Beta to at least resolve the teiiduml the nightly beta site +VenkatCan you please look into thisthanks you are at this one can you also look at which also is related to special characters +beyond that it should be a relatively quick patch +Your shell command is invoking curl multiple times which is going to prevent that connection from being reusedCan you provide a trace of the HTTP traffic that leads up to the timeout +These issues have been marked as resolved for some time now with no recent activity so I am moving them to the closed state. +Can we see there what causes this crashI use the build services from phonegap +Jesse smoked out of svn will be released pretty soon please test out the changes I have made +openejb patch to remove stax version of patches +ImplementedAdded homeapsitebin to PATH for the ant build created xsltproc wrapper therein +Could I ask you to try using or to verify that the problem has been fixed? If you still experience the problem please reopen this bug report. +Its the expected behavior. +is nice but elephantbird needs to a be discovered and b built +The shell scripts will handle building running then +It will be handed after I access the +Canceling patch. +Fixed that with revision Also created a jira for doc updates couple more checkins to make the shutdown message more readable in I have attached a sample to see how the shutdown message will look for I backported the issue to and. +was fixed for +Reviewed files changes look ok applied to branch +Dont know what to tell you it works +I fixed this in SVN. +Just thought of a counter argument for my previous point +Note that the result is different depending on the root object and that the same test succeeds if the doesnt implement . +nice work +Modified get to have the option to check safemode and throw exceptionA test is also added +Only queues were supported in SOA my bad +I suppose this is their copypaste error +Ok you are right for the backward compatibility +I did not mean to revert the table lock fix +The new package is +Patch committed +I will do some analysis on the pdf issue +Theres a checkstyle warning in +This includes all files +Jonathan please see +This problem now seems to be resolved as the client testcase mentioned is now passing. +AlexeyNo one is working on this issue because it does not have a reproducer +It also includes a test case +This means that datanodes consume lots of CPU even while writing data to a block +But that is one of our requirements otherwise it is difficult to discuss feature generator configurations +Update the patch to fix some unit tests + +This seems to be a bug in ptp one of the components we build on +The put is already synchronized in addId method +As discussed with Kyle custom token validator can be plugined per service to employ advanced validation mechanisms +If tests continue to run I will go ahead and commitOne important note to people visiting this bug later the patch provided is for hcatalog not the root +Guard the exact one token special case from the minimum token a minimum token range should wrap +Thanks OliverYou are completely right +No there arent +Updated patch with suggested changes unit tests and asset file containing the direction numbers for the first dimensionsStill tbd provide a way for users to use custom direction numbers +They are for the mmigration branchHere is the list +also a cap on the total size of all row caches would also be cache and row cache are both notWe plan to make the caches global in. +is this still an issue in Continuum +hm ill try to reproduce it on linux if not i have an excuse to use my wifes macbook +Closing issues which have been released. +You are removing part of the code that does the proving +If the is closed before the isolation level is changed it works in embedded mode but it still fails the same way in clientserver mode +committed thanks Tom +Is there any way to determine the Java version for the Eclipse workbench and set the facet version accordingly +Created in the docreferenceenmodules directory with svn! +before versions we could use other servlet definitons with filterdispatcher configured in in as the above example we cant +Looks good +This kind of failure has happened in the past in the dictionary classes. +I think what Jervis was suggesting was something along the lines ofWhen a service starts up check to see if a logger is configure in the or similar like INFOor similar and if there is then automatically add in the to log to that logger +Attaching a patch for just the samples change to fix problem . +A new target which mirrors the and targets +Uploading a patch with additional doc +Closing stale resolved issues +Small change +The attached patches fixes my snafu +Added to srccontrib clover handling +TedI corrected the AMQP spec version toRegardsRajith +Was done and works like a there something to configure to let the various implementation benefit from the lucene index? +Thanks Nathan everything is ok +Yet I am as surprised as the guy who asked the original question +Yes we shall change to have a different name for artifacts +Good catch tom +With other app integration methods in mind David and Kevin may make a different decision on where to put your class +I have already downloaded the jar file and placed it in the +How will impact this immediatelyId really like to get this patch in first and refactor if and when goes in the benefit of bypassing static comparator isreally needed +It appears that in Grails my project simply is bad +Ignore the performance hit to that appears to be unrelated to this patch and affects trunk without it as well +Feel free to ignoreOlegPS Actually we should also consider fixing which has the same problem of reading one byte at a time +Here is a patch +In addition to df it uses whoami and lsbq +Comments are appreciated +I cannot imagine the order of the two commits has any significance +Otherwise Hibernate should handle this +Tests are running successfully. +I didnt know this existed +Pls reviewTed There is a minor correction needed in the Javadoc for trunk patch +Other comments +Tomcat context definition file +Space created +Hmm +Until the code is fixed the import should be changed from optional to so that people know it will work if it resolves +The patch removes vmvmisrc and makes a header rearrangement for a new loggersvn remove vmvmisrcvmvmisrc At revision +Sure +Integrated to TRUNKThanks for the patch MubarakThanks for the review Stack +This means testing will still work if setuptools is missing. +Im willing to do this but I will need a little guidanceWith the current POST buffer default size of you could include just under boolean clauses if each clause were only byte a highly contrived and illogical query +HiCan I give you patch? Absolutely +No problem thanks for having a look. +Well TBk maps is MBmap +And finally on the patch +you have a very good point pedro +Do you have a small test app that reveals the problem? I tried creating an app and adding the and immutant from the gist but lein test ran fine since there was no test loading I you tried running lein depstree and follow its advice regarding the conflicting libraries? That may solve your issue +Configuration is an embedded with producers connecting via the VM transport and a network of brokers over TCP +patch commited to apache with revision +WIP committed in rWith xc xlib a jar file specified with XLIBSRCJAR will also be included in classpath +If a property returns the empty string value the label and values are shifted and no longerFF the lines do notIE seems fine +All tests pass +The patch is incomplete in two respects The decoder does not resolve schema +I dont know if there is a separate JIRA for it. +Olegno problem +Martin can you try sending mb files and let me know how it goes? Thanks +Committed to JBESBGACP with revision +Because if in a production environment clerezza crashes its to easy to forget deleting these folders and then eventually you run into disk space problems you would not have otherwise +formatting changes in improvement +Id like to see this changed too +Did not manage to reproduce +I agree that it does not make sense to catch one case in the typechecker but allow another to slip through only to fail in the +I guess at this point one of our alternatives is to use but we have spent a lot of time on configuring Aegis for the webservice and from our client within on how we can proceed? +Once the them UI is used there is no way to go back to a simple style +Apart from the unit tests there is no other stipulation right? In all probability we would have actually tested it out in our production environment on an earlier release of hadoop even though the code could out with +the xml schema still needs to be updated +Assigning lars so hell take a look at this issue +It now submits the full classpath to the +Thanks +Dan any idea what the purpose of the Client Test feature is if it doesnt use +Ive committed a fix to the branch +to treat such errors as a regular end of input data + need to upgrade derby for this to actually we are now on derby +Darren can you please have a look at this? this was suggested before but I would like to see the class extend in stead of +Can you verify +and are different test case i think we should not make its duplicate and resolved +I implemented basic support for handling multiple alternatives +Also shutting down the nodes when name service ID could not be found +btw +For sure slave domain controller needs fixing as theres no such thing. +Removed duplicate ASF repo entry from subversion area of jira admin checked issues duplicates now gone +when password are in concern we should pay maximum attention and reduce possibility of password leak to the minimumI understand that we assume that access to Manifold is somehow protected and that there is only one level of security so everyone that has access to all Manifold configuration + +Assigning over to Ben for the patch process +This is the default +Some further javac warnings are deprecations +Im currently checking your branch nice work as far as I have seen +In the listener put a inspection to check if the algorithm is +Which version of Hadoop you are using? does exclude corrupt replicas +I reverted the previous commit +The issue was that Eclipse was using a JRE not a JDK +Enebo examine and apply or fix +When you load the enum java creates every part +Are there no tests covering this unpacking feature +Tom are you still working on this? I can takeover if you want +New patch which fixes failing unit tests +We do need to update range slice to match reads. +support should be in for both and thats great Simone +changed priority to attached +no I meant a comment in the documentation but other than that allow it +Ive subbed jenkins to your dev list +Ignore my last comment about having this target startstop its own server +Whats missing in the current implementation? I just tested a Groovy class that holds an instance of a Java and one of a Groovy class +Ted in the current patch for trunk I removed the append part from the test +Also see bug where is used inappropriately +Testing the new still not getting sharedLib to work +Great! Issue closed. +If you think that would be less of an impact I am happy to switch over to that instead +I think we shouldnt resolve issues fixed on branches before they are merged back +Definitely need to get rid of the warnings about no schema when we dont have one +typo +Whether JBDS is part of the ESB we team or not our customers use them together and we want them to use them together +It should be fixed now and I will let you know when the next build is available +Tested OK. +Fixed Exactly +Sometimes one should read the entire page Anyway thanks for providing the dump +After we get building deploying properly we need to delete everything from SVN at forrestsite so we dont have any leftover junkThen build deploy from trunk from trunk and the site from branch +This file contains a micro benchmark demonstrating the benefits of the commited to trunkOnly tested under winXP on a way cpu under +Submitted by Andrzej Bialecki +What does the EJB plugin produce when you try to use it +moved the binaries away from the source packageIn binary release we can still ship the libsHowever in the source releases we dont which is okay +Im attaching the +If the outputFile attribute is not specified any program output is for input +The block in question is patch adds a test to trigger the issue and a fix +So the number of large bags that cant be garbage collected would be similar to that of single query +Usage of a + to mention underlying problem in Hadoop is already committed via which will be available in and so we may need not to put a workaround in Hive for this +I tend to think that moving the Jconsole support to a separate jar project makes sense since it only has value within jconsole +fixed +Use wsdljava to build your client +This patch +I assume this is just hanging the scanner thread waiting for the model controller to finish executing the task +Then leave it up to the developer and iBATIS to implement it properly +This behavior is not addressed in the spec and therefore is undefined +That issue is resolved now with a patch +this has a from meId like to work a bit more on the test case but I dont expect the overall behavior to change +Typically soft upgrade does not enable new featuresSo we need a pair of terms which capture these distinctions +I dont require this forNote that APP doesnt mandate a particular charset +I was wrong +By submitting operations asynchronously we need to do it one by one +codehaus repo has been synced to ibiblio so parent poms are there s done manually so just ping and in have all modules projects been upgraded +In your project declare the dependency on hibernate before the dependency which itself depends on hibernate +Test failure looks! Its been a while +Peter Abelsson wroteOk my fault +Looks good to me. +Thanks Edward +Thanks Viknes for patch +I let it open then +Ive reverted the commit at +It also makes my laptop unusable + +Thanks +rb script and two +Ive committed this and merged to +Need to add permissions in motion no longer requires explicit permissions in Android. +I think Mladen has created an abstraction layer that lets you do this sort of stuff but it still uses a native APR layer in the bottom +How about a stolen disney dwarf and the mottoWSDL while you work? +Added some test casesThe test cases passed on andThey failed on because of classpath issues that use conflicting versions of slfj at runtime +If it is required then I will do itDave based on feedback +Thanks Namit +Yes +While the developer might want to store other business specific statistics such as Trades processed or even separate it into something like credits processed and debits processed +patch committed thanks Corinne +For now you will have to remove one of the two hrefs in +bq +However I havent tried it yetBut generally Id say for committing it +Brian any comments +In my opinion such a change would constitute an API change and should be done nosooner than or possibly release would also consider this an API change +We will defer this to CP and synchronise the required changes with the jBPM team +jar but this seems fine for now. +This however is misleading since this and other defaults are determined by the properties file +Comments on rev It looks like the Metadata parser needs to parse out the join table attribute so that it can be applied to a mapped superclass +The onus is on the user to ensure that the file system is globally available on all nodes +I think what youve done here is sufficient and the overall result is an incredible usability improvement +I think that is better than overhauling the general resolution approach taken in spring +We need to investigate this issues once we got all process versioning questions solved +It only specifies requires on compile so the other comment abut defaults is irrelevant +There is an merge error in the patch thanks to a report by Sonatype people +With the introduction of in favor of and the move of ftp this issue is no longer deleting files from a local directory see the file specific Transformers +Janna is a female name in Finnish +The patch makes a white space between the end of a section and a title of the next one wider +move the count from the transient map to the dictionary +bq +The Community development team havent looked into this yet but I think Dans idea of using an Awestruct extension could be the most sensible place to contain the logic required to call the various services and generate the desired output +Do a ant +Resolved in the SCA assembly spec +Anything in particular you are waiting on before committing it +Commit changes to trunk at +Thanks Harish for review. +Patch to shorten names of nested files in ears +Thanks for the patches! +Is your code available anywhere +Some of the license files will move to new homes within the maven directory structure +I suppose this is really a server issue as its a bug in saving the server config that just happens to affect you try this against the recently released AS and see if this is still an issue +I am going to take Dave Lloyds advice and ensure I am using the proper GA release of remoting and XNIO +I didnt dig into the rat check +You can commit which ever the code you think it works all i want is struts have to work with Jboss AS And do i have to file a separate bug for the performance issue which i have mentioned in my last post +Is this a blocker for inclusion +The configuration system has been redesigned +May I commit the test into trunk or not +Patch looks good to me +This would probably get us there +After this we do not need as a seperate file so the increse in after the system test contibution can be reduced +Thanks +But I guess from this big patch its harder to split it into two now your working codebase has changed soo muchAnyway you are on a good way here +Please provide a test case that shows incorrect behavior that requires the proposed fix. +an additional advantage of this is not needing to extract the entire enwiki collection in order to index it setting the repetition count to for in alternative or for in alternative would mean that only docs from the huge file are extracted + +So if the time was right we could still attempt a test run but theres certainly no rush on thisFor the links Ill see if I can find a way to locate all those issues and add a comment with the link so we dont lose it +Well they are hard to maintain when you want to go add a new method and you end up breaking users who then flood the list bc their implementations are broken because of interface changes +It is not very efficient to check if a batch contains only one Put which is kind of duplicate some of the multiput logicYou can say the patch is not strictly for single put +to trunk +It shouldnt +Ive asked Brian to comment but you say this a is specific to cxfs javaws? Then we cant rely on it for users that run with the default jbossws enabled +Added additional Camel expression can now specify an exchange pattern in the to and in the Scala DSL now uses camel case to be consistent and aligned with Java DSL +Attached patch against latest HEAD looks good proposing to + +The benchmark code has jira table output +Committed changes to allow people to experiment at machinery has been committed to via system variable activation trunk its tentatively always we still have to decide how to handle it thought Im inclined to add a system variable there too +The code should be anonymous +Yes I was waiting for final okay of the last patch +Please discard my previous patch +Perhaps we should close it +It is purposely breaking backwards compatibility +If a committer wants to these changes I feel this patch is ready to go barring further objections +Unit test and fix patch applied +Thanks + +Fixed. +See +But it also generates the Java classes for the imported type system and puts them in the test folder even though those classes already exist in the src folder +As a result this issue has been moved around to different targetfix versions +Will go ahead and commit this +Bao what if I commit the addition to +patch v + +The should be discussed at a higher level +I had to mention that you needed to install both community features and Does this mean that the Central features should dependinclude the community one tooNo +We use still exists in SB version +to reproduce the problem in a consistent way you need to make sure that the Property View is not open when you select the nodes. +Attached revised srctestjavaorgapacheservicemixcamel unit test that exposes the problem when using camel +I cant seem to create new pages +Will post the issue in Springsource forum too +Thanks Yin +Works as expected with +do you have thoughts on which the lexicoders will live +Added foreign key constraints to the sql +So its not Spring choosing the wrong method here its the JDKs support +The bug is not reproduceable on EUT +This work wasnt complete for the release but you will find this available in the nightly builds +Blocked on for testing +The patch simply check if tasks is negative and set it back to zeroThe included unit test produced a negative number of tasks before adding the fix +And im running into a deadline to deliver the thanks for spotting thisTareq thanks for the patch +this is fixed in revision +Well then sounds like a compelling reason to fork +Nice! +Just faced same situation +The issue is the splitter copies the exchange for each splitted message +failure and findbugs is unrelatedIve committed this and merged to +Your command line is wrong many ways +It sounds like nothing will show up there though +One thing that could be looked at tought in this patch is the name used for the new attribute on + to Niclas for him to take a stab at it +Alexey thanks for good words +Choose one of these extension functions for testing and add a dummy copy constructor +Please attach an example that reproduces the reproduce this problem extract the attached file cd to the extracted directory and start the webapp using Point your browser to httplocalhost to see if the works Point your browser to httplocalhostlangdelangde to reproduceEnjoy you use the proposed you need to make sure that it is configured below the char in +My vote is for the namespace and not +After recent release notifications works perfectly without any delay +Standalone test crashes as described +I think it does work I basically copied the as on my local repository +I think this approach is worth pursuit +IT added. +This patch removes all unnecessary use of confctx within savan module services +So similarly for TCP and UDP transports services get endpoints like tcphostportservicePath which are perfectly valid and we depend on the dispatchers to do their job +Thanks a lot for findingfixing this issue Todd +please check if it is still valid if not resolve as can not reproduce +Thanks Daryn +Its not portable anyway +Thanks. +looks good +Then if how it operates doesnt make the best sense lets make a new issue to fix that +Well try to refactor what you have already provided. +I still cant reproduce your problem even with some very contrived tests +Merged both upstream master in JBDS Build id +BTW this is a very early version of the patch +I would not change anything else +Here it would be a lot more focused so less of a concern +But now that weve mentioned it here similar searches by others in the future will produce a dizzying results thereby leading them to imagine it is positively commonplace OK unless any Catalan speakers wish to regale us with stories of fond evenings playing joc descales I agree we should probably let it go. +Here is a patch that adds the new security mechanism to the syntax +Do you think theres any point in trying to push the support for offsets upstream into GuavaDoubtful +Clover version is not fixed in Sonar +This issue applies to codebase prior +However the core integration work is doneIll open a separate bug for JSP migration +I disagree with the original patch on this issue it is very inefficient since it calls multiple times which is an expensive operation it doesnt handle and cases properly since for sorting is considered greater than Alternative patch is attached for your comment +I eventually get the running with a pool size of connections +we need this fix in modello released before we can fix the issue in maven +I am interested on this project +If the JPA spec ends up addressing schema generation this might be a candidate for a TCK test +That way we can ensure that the response is always handled by the socket thread that has registered the needed socket key for the response +Added link to +so i think the custom should also can be configured with just change the read method so can i read the configuration from my custom +Twice today I forgot to change the output filename when doing the checksum +As we need to move it to +If we want synchronous flushing then wedplace a flush event in the queue and wait for it to complete executing +Reschedule to GA +unit reproduced thanks +it might make more sense to invest in moving the MR framework into user space rather than replumbing the current implementation to be binary firstI agree that the two are separate but I dont see that they cannot be pursued in parallel +Reopen to add fixed version. +So you can clause this issue +Please reopen if you feel its not fixed. +Patch for trunk +Uninstall support on logs needed to be removed from the patch file generated +Sorry for the delay in verifying I have been on holiday +ah you mean the wizards should allow you to override the generation location +In the generate case we do allow to do this simply because we want to know the service implementation class in order to generate the +Please update the patch +Any thoughts on Issue Should I a patch for it? Im still on but hoping to upgrade +There is no ate account in jira +So need to fix the test to pass in the full job to verify againstIt looks like it worked with jdk because it just so happened that the mock job created has both total and completed as +Thanks Billy Ill dig +Thanks Michael for helping to debug it. +The fix implies that the stored procedure calls must be prepared for each logical connection if the LOB stored procedures are needed +Manually tests added +Not sure why the component can not simply check if the class is of type Calender and do the right thing? This is a Major bug +hmm the maximum error is quite large for some functions compared to the current implementations sin ULP cos ULP log ULP log ruled out for Providing a separate based on that code could be a new independent feature +The trade off you make is read performance when nodes come back up +So template syntax and rendering engine are both easy to understand and quick to learn and implementI used it for build my university project +Closing. +The for remote servers is a known issue and is expected +Once weve found them we do a bind as the user using the method described in this JIRA issue +From some very limited testing I just did removing the dash doesnt seem to break anything +because I think the is valid +Ok thanks problem solved +Andrzej mentioned backporting in an older post +I tested it and it seems that the is invoked and passed an Exception which contains the line and the error that was found +So thank you all for developing this frameworkGreetingsTorsten +Ankur any thoughtsI looove consolidation of code +I would like to wire a force flush method into the but before I do Id like to get some feedback on the attached patch +At best the client will fail to initiate GSSAPI with the wrong principle +It removes almost all of the options and sets the values directly +WhyThe simple answer is that as good as is for basic evaluation and production it is limited by the constraints of derby so the configuration of LCF with is considered superior for production useSo yes will be used but the configuration will be used as well in other situations +But putting it on this patch would be just as useful +For you need to apply manually +pushed +As us far as we understand its not a common way to use it for things like validation +based on documentation it is supposed to handle any object that is serializable but it does notI have made the business object as little as with three instance variables of type String +No test how should this be tested +There are test failures but I dont think they are related +commited in rev +I will try to get a functional test case +Changed to improvementOf course the change is not absolutely necessary but I think it would be nice for users and also quite appealing visually +Example using the new hooks +It may make sense for Roo to replace the standard JDK properties API with a simple util class which preserves the formatting and order of the properties +We can make it protected that way its expert level and a userneeds to inherit from to use it +Ive check the behavior in Eclipse the problem still Exception is thrown. +I agree with John +Otherwise itsCan you please provide more info? I cant reproduce it in my environment +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +Thanks Shai! +Not caused by the changebq +May be I am missing something here +jira chopped up the code snippet a bitLooks like this issue occurs in places line and hmm +Sorry for delay done. +Requests an optimize operation on an index priming the index for the fastest available search +How is this planned on being addressed? Given the changes in to this feature my disable implementation is quite broke +Whats the trick here +Should probably log this in a system table so its easily queried +I have updated the and the +Since doesnt guarantee order that could be the issue here +Pushed to master +Not a problem here are the patch files rebase for a new into trunk thanks! +Given that for trunk +Hence I will fix this as suggested in the second alternative by catching exceptions with messageId +i guess its for unification purposes we dont usually indent the first comment line this is not a strict rule but whichever style you choose please follow through the whole file +Task is done. +Ive removed the user filtering at each level in a path instead the filtering is done on the full path name after the globbing step +I have no knowledge about +this is the outputDBMS DBLINUXX SQLJDBC Driver IBM Data Server Driver for JDBC and SQLJ Database jdbcdbjbossqaUser gateinjcTables +which might help you +But since the error message is descriptive and you can indeed use searchZone i dont seewhat else can be done about this now +with Don +You also previously instructed to use libvirtd d l instead of service startThat was why was in Stopped state +I doubt that you create so many cookies in a single test to run out of memory +Verified at rev +Fixes added to handle invalid entities and treat them as +Or if you know for sure that code is serving some scenarios by being there then we can let it be there and just add the change in that is getting rid of both the issues reported here +Attachment has been added with description patch to font short hand test that shows bug in has been added with description Fix problems in font short fix two problems in font shorthand with this patch precedes unquoted font name list begins with unquoted font name but has quoted font name in transition to closed remaining resolved bugs +It is only on master currently. +Dave did you ever get any more information about this issue +Builds local +I dont touch any other Sandbox project +Take +If thats the case its a bug +I will look into it. +Ill look into this later today it would be rather nice to be able to do this +This might create a problem for unit testsDo you want to add a configuration parameter for this for unit tests instead of it might be better to overwrite patch is outdated +sure Ill file a JIRA and submit a patch +Implemented by commit. +Yes this was against r +It also ensures that the offset assignment on the leader and followers will match +Committed revision Thanks Sebastian and Steve +perhaps usefull to other users however not supported in the trunk +Ionly wanted to put a simple comment somewhere on the Hadoop Wiki that saysthat a line must end with an end of line delimiter +bq +This is the same patch as for issue +there is need for a manual merge to get it into but that has not been completed yet +Perhaps that can be part of the checks done when we move such plugins to core. +Should have workedHave you set HADOOPHOME and HIVEHOME environment variablesAlso a dump of your environment would be great including classpaths +This proposal should be revisited for family. +So went out without the migration layer +This looks like a dup of bug which I submitted and which was fixed in +Sounds useful as wellIll take this one on +It looks like this got fixed some time recently since your spec now passes +Investigate for SOA let me investigate a little bit more +Sorry for raising a false issue +Dont we have separate handler for opening rootmeta? Where was META on this cluster + +The down side to doing that is that now the two tracker items represent parallel versions of the code +the new representation would be three new pieces such as +So the patch will be applied to other files +Can you please check that Maven plugin is correctly initialized and you can actually see content of the local Maven repository index in Maven Indexes view +DIGY +Thus this module is the only one being able to feed modified configuration back to files +Go for it +Attached regression test fix for this issue +It could be a good idea to refactor the class by splitting it in multiple classes to improve ease of debugging and in need of some shape up and refactoring +right its not supported now +You need to provide the license as txt not html +Closing resolved issues for already released versions +z files +The runtime was updated with so this issue should be fixed now +IE still exhibits the problem after successful login the user is treated as guest and shown only three functions +bq +Eduard does it work if you do not start it using from an existing EclipseBasically I am asking for does it work when its deployedinstalled as it should be? The code is working fine if you start it from a usual +Im not doing anything different +Fixed with revision +We could have spent time modifying the existing CLI to work with HS but we decided against this approach because a the has a lot of bugs and b we risked introducing new bugs in the process of modifying the CLI to work with both HS and HS +So might satisfy your requirements for a whilethanksdhruba +I kept the tests you submitted so I just amended the commit to leave the author tag on them as is but format the commit according to the contributor guidelines +Thanks for addressing the javadoc commentIll run this against smallTests and report back +Thats better than nothaving a test at all in my opinion +The issue may be the QE tool that strips out the signed information from the jar files as it may not repackage the jar correctlyThe jar file structure assumes that the manifest is the first file in the archive or is only preceded by the directory entry as such the plugins in the QE build are invalidThe contains the following entriesmavenmavenmavenorgorgrhqorgrhqpluginsorgrhqpluginsplatformorgrhqpluginsplatformcontentorgrhqpluginsplatformcontentyumorgrhqpluginsplatformwinWhereas the correct structure should start with the manifest such as the following from one of the original files +That is more an option for an OCR engine to readwrite PDF +in to do exactly what you want to have +Yourreview is appreciatedThanks like this was committed some time ago +Then this bug may be hidden for a long time +We can call createComponents in create without depending on class user +It was like a one line change to. +Instead when the Method is required it is passed as an extra argument +Confirmed issue with +I think skiping foos revision when asking with since seems correct since that revision corresponds to sequence Like this the replicator does not receive revisions it already received beforeAm I missing something +Marking resolved +With the previous interpretation remaining will be zeroIf we want the latter then we need either do du or maintain disk used to be sent with every heartbeat +Attempt a at runtime and if the engine chooses the wrong type the user can force it with a property +Thanks Markus I try the patch and can index the document success +fixed at daa +Marking as Not a Bug +Ive tried the newest SVN version and it works like a charmIt looks like I will be able to implement my use case +nochecking js usability +If they do not caching may not work as expected anyway +Plus test case in Rev +For each class it makes a summary comment consisting of everything in the class header comment up to the first period +This ispotentially unsafe when the same user request multiple pages simulteneouslyI propose to put the next token in the request context only and change thehtmlform tag so that it gets the next token from the request context instead offrom the session context. +verified on +These delete operations are merely not in the patchPlease review this testcases are fine +Attached service you are going to use RPC MR then you have to either use wsdl which compatible with RPC or not use custom WSDLThanksDeepal +table? Am I reading this rightI am still unclear why we need auditing facility in the regionserver? I do not understand how you expect auditing to work going by the patch +Added JVM info to about dialog +Now it works as expectedThanks for your great work! +This change has already been made +If the user is using JSF they will still have to specify the spring in their so that spring beans will be found during +I assure you that is not the dialogue that comes up in the installer +Id call this unexpected behavior but for others who are looking for a workaround this info might be of use. +enough the Spring version of the class does check whether the value attribute is a String in is already +Hi Stepan Will you pls validate the patch and apply it if appropriate? Besides I made some modifications to the original testcases which have problemThanksLeo +we should delete the old code first imho +i would rather refactor it such that the optimization is located where itsactually required and where we would have much better control on it in terms ofconfiguration changing implementation etcin other words it could be reconfigured such as the number of iterations the algorithm and so forth +How is this a problem +Lets try and get that in beta mid october +Very good idea! Do you have an idea for the syntax or should we have a separate editor for itIn we can add arbitrary metadata to all pages with or without separate editorsAlso how would keywords relate to tags and should they have some sort of an relations? I would like to hear Murrays insights on how this sort of stuff could be handled +The only case Im not sure about is normal finalization that might occur during the program +Unfortunately the default implementation of doesnt support catalogs +The window is just closing +Since we are still caching privately I think only expected degredation is on first call for booted db though I know there are some applications out there that boot very often +Ill try and get this fixed but its kind of a bizzarre case +We cannot add a readString to as the Input and interfaces define a low level interface that is not to be aware of anything other than primitive types classes like etc dont know and shouldnt know how to read those types of objects thats not their purpose +AndyIts a matter of perspective I guess +Ive been working on implementing the document reconstruction feature over the past week and have created an additional admin page which exposes it +with this is not a blocking issue so lowering priority +Thanks for this testcase although it does introduce an interesting problem when extending the notation with assertions etc how this would be dealt with in a merging situation +checking that it is a string first. +Jeff can you send some example code for this +Ive just committed this +Also we need to provide the proper diagnostic and custom +This version is consistent with the currently shipping version of karaf and smx +Also I think this bug is a dupe ofThanks the concerns Suresh had +Only isredundantrouter then the redundantstate and some other column would make senseClose the bug +Also we should add a disclaimer that and will be off +Ive been trying to get Giri to get me access to the new build farm +Please take into account that we have only a very small number of tests that test fail situations +seems you pointed the same +Solve this jira on. +If you find further issues please create a new jira and link it to this one. +Also the new results for are wrongWhat is the problemEssentially in the current code the nocondtionaltask is not enforced for a +This testcase fails on trunk but works fine on +This file would ideally be built at build time of the jar +Patch attachedAny comments welcome +Accidentally attached the v files without granting license to ASF +All right +Please unzip the avroStrings fixtures and change the path constant +Can you provide a test case +This feature was added to the current module +Fix confirmed. +Alan do you think we should still support this +Something like a label in the extension builder +I guess youre right should make it bit more user friendly +Sorry diffed against the wrong branch +Hi Michael +A workaround which worked for me was to add this snapshot dependency to the projects is something dependencyanalyze issues warnings on anyway +Ive attached the compiled class file as well +Im ok with this change but it did expose the Loathing I have for and its locky mclockster locktasticness so I raised as a result. +Since we are just doing copies for the rename this should be fine except in cases of cloningexport which should recieve some thorough commenting to avoid contention +I verified source code of distcp +Oh I see its an ant task within the projects trunk + +The attached patches are already applied at revision r r and is already integrated in all modules that contain tests +If the fix to implement full Holder support involves changing this assumption then we might need to reconsider some of these instancesThe example I have in mind is the support in which we have cases on the input side where were concerned whether or not to serialize an array or its contents +I also added toString equals hashCode and new constructor to Token so that I could test the new code +I think this is a duplicate of which is fixed in SVN trunkCould you give it a try and see if you still have that the same problem +Im going to take a look +I tested it with the latest trunk version and the result is The relies on the Beta function which got several improvements in terms of accuracy since the releaseCould you please test yourself with a more recent version of CMThanks ok my fault but adding more regression tests is definitely usefulWill add the test case later on +Feedback is very welcome +ERROR No Host matches server name +pushed in components +New attachment with corrected formatting and. +After updating the tests I have been unable to reproduce the original failure so this does appear to solve the conflicts +Ive subscribed the watchers of this issue to. +OK Ive applied the patch to both x and trunk and it looks good +Problem has been fixed. +Aaron it doesnt seem to populate the queue does it usually happen immediately or after some time +Turns out all unit test failures were due to broken unit tests +I tested it with mvn compile openjdk and it compiles hadoop fine without that dependency +war +Waiting for a more detailed casesample to illustrate the desired behavior. +Suggestion Components should be changed to or View technologies +This can be optimized to avoid the extra copy of data +bq +is a writable and in case of multiple searchservers distributed in a network it makes sense to minimize the network io since getting details should be as fast as possibleSure StefanI will take this into account of course +Please correct me if I am wrong +Why not just make a codec that strips tags from keys on the encoding and a decoder that can read though the tags are stripped? Or will clients be sending tags +Committed revision in HEAD +ok fixed the comments i had and also removed some unused variables in the code +My confusion suggests to me that I dont know how to read this deadlock traceThanksRick +Updated you for the patch committed +Qmx let me know if you run into any issues +We need to cater for both options xsinil and not specifying the element +If the problem persists please reopen +Ive attached the patch file +And MPING too if it works in UDP +The very first time a component is reset the value of PARSERSETTINGS should be true +And IIUCit can not use Cache as this is completely different thing with clear semanticsof storing Cocoons but not arbitrary objects needed forcomponents internal workings +Since this changes the wire format of shouldnt it also increment the version of protocols and files that read write? At a glance I think thats just +Thank you in advance +Makes sense +I understand classloading in AS has a roller coaster history unified to hierarchal vfsSo as of classloading is no longer default hierarchal wre to embedded jarsMany have grown used to embedding jars within war ear and esb artifactsI see nothing in your release notes +NPE for for non compilable getter cause it doesnt work. +Thus far we avoided gossiping any ports because using different ports on different machines is a misfeatureI agree on the misfeature but forbidding to do that is probably harder that allowing it +by the means of automatic naming resolution and the available resolution strategies. +GeoffI realize its been a while but do you still have work you could contribute to this effortId like to see this project fully mavenized as well and I was planning on working on it +There is no need tointroduce more config options. +The JVM may make this complicated for integrated so perhaps just for those hosted in external processes +But also now see typeB in the results section for the operation from +There wont be a release +thx to Hazem Saleh for submitting the +new patch that implements a ref counter to avoid disposal of still used binaries +Anyways I have made some changes to so that it can load any avalon component implementing the Mailet interface as regular mailets +If this is the problem I wonder why Solaris does built throughThanks if the integrated patch works for you please close this issueThanks +We need to Implement to hold the list of sessions and then use it to resolve +I have seen a similar problem with in the past +Why did we remove the call to from srcpackagesPreviously refers to PREFIXlibexec to source existing HADOOP environment variables +SESSION based performance tuning is done for at NTT Data +Also all test cases for my application now pass +but if we created and opened the stream and arent exposing it toanyone else we should clean up after ourselves +I made my fixes on an obsolete codebaseI merged my fix with the codeIm attaching the including the changed files +These items are still not resolved for Please review your items if not ready for please move them out to future sent one more email to the list asking for review including a link straight to the XML file +Getting farther +Attached a screenshot the version elements is missing +For questions please use the user list +French properties file for the Kaptcha you. +I ask as there are several other asserts similar to this one +mrepository help? Am wondering if youve got some old version of some jar in your local repo? You definitely have the latest from trunk right +Assuming of course thepatch is good and doesnt have any bugs +Marked up documentation as well as shown on screens describing outstanding issue. +Yes I was able to reproduce +Wait how are you seeing megs? Its only megs +Florian Closing this as you are pegging this issue on the end of the other is not a good plan +Wont this go out of the scope ofIsnt health monitoring something that the loadbalancer itself should doFor HTTP those loadbalancers should just do a health check every X seconds to all the backends and see if they are alive +If not you return failure +There are multiple cases on which this information will be stale the region moved again the server is dead +But dont ask me why it does not matter when using Java instead of Java +We can also recommend it to anyone who comes along with such a question. +if as proposed the read requests load might be not balanced +Uploading initial and docs uploaded to +The patch is formatted with git +Patch attached I hope its easier to understand +Thanks +original filename for the note +Ill be playing with WMS filters again next week and Ill also add support for CQL filters Chris talks about on the mailing list +Also I needed to comment out the override commands for some reason but still everything looked good +In particular the master needs to know what region servers are running and what their start code is so it does not try to recover a log out from under a running region server +Please attach a test case we can use to reproduce this thanks +Ok I was talking with Tom about this some more +I added a catch for and used the patched code inside that +Bug has mysteriously disappeared +Given this required source archive I dont see the point in also putting the source into any binary packages and I would not support doing so +I believe there were some eclipse bug about this one but since we now got others reporting it I would like to track it +that should be almost ALWAYS means and not almost means of course +Tellurium will not be used for now because it is missing some features that we need +Sure but keep my point in mind +Perhaps it is the better for user experience to display these options and disable them when the key is locked +Sure you can take this bug +If there is its something youve made up +Perhaps we should use whatever the JDK claims as its default platform encoding for our default mode encoding? That should be the correct encoding based on the current platforms LANG and other settings +So you just totally ignore objections from other committers and commit anywayPlease revert +I thought you said that if they were typed correctly they remained versionless +The accomodation Map or Spot already contains the accomodation so there is no need to add that to the workEffortI hope this answers all your questions +Can you produce a test case that shows the issue? Without a test case I really cannot debug it +The Replication Monitor catches the and signals the namenode to restart +All mapper tasks completed +There should be no new features added so not so much to break JBDSWTP has bug fixes so there should be also some improvementI have installed JBT to Eclipse Indigo SR without any problems but with Java not tested with Java +I generated the html and it looked correct +Otherwise Ill assume that was the fix +All my jars are inside my EAR project and I get wsjarfileCworkspacePath! +Moving out of +all work perfectly with version but with version there are troubles +I was not aware of these tests +just a parameter +A couple fixes a test that catches at least some of this +New test cases for +Gokhan I applied your patch and ran the TCK successfully again +But the document isnt display completely +Heres the revised accounting app patchRegards to mention Im moving on to the frameworkbase took care of the file it now has the ASL header and a credit to the source for certain portions with a compatible license +its quiet the opposite no? I think it cannot get any more simple then give the last messages within the ERROR level +while we are on it should we maybe add a th type that allows multiple values +Assigning to Max to review attached as hes told me he no longer sees JIRA mentions emails +Resolving +Close as duplicate of +Right but there should be +Updating the fix Version to reflect trunk issue has been resolved for over a year with no further movement +The Twitter link to a page displaying all core developer channels and other community channels would be on this page. +works for me + is a maven project where I was able to cause the NPE using the right breakpoints +Updated patch with the current trunk codebase +I find it very intuitive that I need to specify that fields that I want to load rather than the fields that I dont want to +Have been able to load up a node cluster w B rows using default heap sizes with about regions per server without my MR loader failing a single task +that Lars and Ed figured. +Kurt can you remove the WS patch from the zip +Updated patches because I realised the license header was missing from the new test version of patch with fix to sstable max against branch +Today I was able to receive and this one worked better +The bad export you can try to import it +Looks good +Having said that I am not able to reproduce the problem after a full rebuild of the projet +Dear Zookeeper committee I have generated diff patches branch and trunk and all UT in both Open JDK and SUN JDK are passed need I provide it +Id give him karma if I could could anyone who can give George karma its been weeks now Ive mailed the Geronimo PMC several times to nothanksdims +Seems like an easy fix +I think your fix should close that hole +Thanks Brajesh Rishi SumitPatch is in trunk rev +The XSLT spec does not anticipate this scenario but XSLT does +Yep with allThat can be useful if one will want to transfer part of documents on changes especially when full versions are big. +What is the bug in Apache CXF +constant name constant name class +lgtm to get rid of the extra logging and to kill the child process +If you would take a quick look to see how I use +But if I dont want to provide alternate uris and just looking at a single broker say my activemq server dies for some reason and I restarted it the message log should then be picked so that there are no event loss +Is there any chance to get it in some JDK update + +HiHere is my first step to contribute to Aries Documentation and project too +Having or not havinggetHost will not significantly change the situation +Hmm +added intermediate page with a note about configuration. +I also combined the merging tests so that testMerging test share code +Hehe race conditions for commentsI think we should get this fixed for +But this has nothing to do with having our own clients +The behavior is a bit different now +I will try to get this site deployed soon. +I agree that adding code to correctly handle the wrap around condition would be better than just expanding the number space by switching to a decimal typeWould you be able to contribute a patch to support rolling the cur back round to +Good idea +Duplicate for +am attaching where i moddified the pls see that and tel where i made attaching the screen when i started the weblogic server i found the error that can find in that screen shot +I just committed this +I might be wrong but I dont think so +How do you want to handle resolving this bug? Should we consider adding release targets +yestrue is indeed ll so some more search on this one +Meantime this has been committed to trunk so closing this out. + to trunk. +just with servicesShould I post this to the maiing list or do you think its useless? +If anyone wants it backported I can backport it +I tested this with a map having an infinite loop which has a sleep inside +I think you are the only one who can address error shows up also calling mvn coberturacobertura from the command line not sure its sonar related +This issue was not filed with the expectation that it will fix the other issues +aggregator was added with originally thought this may have been fixed along with by reverting the change from that added aggregator but now I see that it was reported against +is acceptable for +That is there should only be one possible cascaded compaction at a time and it should finish immediately before the next compaction becomes possible +Fixed for +What JDK are you using on Windows? We support as a minimum +This has often held me back in the past wanting to express yourself in code but getting bogged down in basic setup stuff and then just giving up +Thanks Sid! +Note that Modeshape offers its own subsystem kit for AS which starts and binds repositories to JNDI +AS uses aop +No Check the INBOUND A ACLINBOUNDeth j DROPA ACLINBOUNDeth s p tcp m tcp dport j ACCEPTA ACLINBOUNDeth j DROP This is wrong +Another common scenario is incompatible jars +Yes it looks like support for JAXB XML annotations would meet my request +If IDE is not compatible with Eclipse then which version of the IDE should I use +What Im saying is its ridiculous to expect the user to do this +it doesnt include their libs +Thanks +The patch looks good but Hudson is right we should have some basic sanity unit tests here if for no other reason than to catch any mistakes that may be introduced later that dont jive with the assumptions made in the patch +It failed for me just now on local machine +This fix got rid of the issue + +Do you think both came from same cause +Pass the build +Correct Im going to mark this as done then +Please ignore the first is the real one +I think by audience you mean when this class gets invoked? I am not clear about what you mean by stabilityIf you could point me to some examples in the hbase code that would be great +Updating change now not sure how this used to work +Temporally I just followed the choice of the reporter of this issue +Once you finish the change comment on this jira and I will have a check and see if the problem is fixedThanks +Is doc the only thing remaining to be done? Does Hernan know that you are waiting on himRegarding making the configurable per application +What is JPT and where can I get it? Is it part of WTP? TPTP? Some other project +All that is happening here is that a group of transactions will all be applied or will all fail +Do you want me to attache a diff to the changes I am purposing +Patch applied +Reassign to a release that you feel comfortable resolving this issue in or leave as is +What are the circumstances that make throw a? The docs dont say it could do so +You seem to be using some fairly old code +xml folders containing those files do not show up either +We do have a multi module projects with more than submodules tooWe can use profiles to make modules optional but that will lead to a long and complexThat would be nice to be able to say dont fail if modules are not existing +The attachment shows where the relevant code needs to be added +reopen to change fix version +Hi Jacopo as you probably saw in the mailinglist we are very busy here and can currently not help muchWith the current multi tenant implementation I would not make the webtools application available to tenantsRegardsHans +Thanks Vijay +I upgraded from +Juraj I dont think it is caused by Graphene do you have such impression +This is often the first thing new users see when they write their first program which does inserts with autocommit true thus committing on every transaction +Will we just have some which supportdir and others which dont? Or will we add this support to all of them? We only have like or so frameworks supported so it wouldnt be too tough to just run through and support them all but its still worth thinking about +This needs to be with the totally new geolocation implementation of coming week of April th +Struts shouldnt trim the data provided by the developer but it should log a message indicating the potential problem +Still hoping someone will contribute a patch or sponsor the work +this issuewen +To be fair without the generated protobuf stuff the patch is only kStill big and no tests +Patch which deprecates the method and enum in favour of new ones in Cluster +I havent been able to reproduce this NPE exception +With this patch you can remove the line in that explicitly adds guava as a dependency for that jobAre there any other dependencies were not shipping +Itd be dangerous to do into a live hbase +As we have a release pending youll be able to test the new version tomorrow I guess +Committed to trunk +Assigning to fix in may be fixed by updating for me in IE . +RichPlease take a look at this as wellthanksve removed the check for nulls in svn revision thanksdims +We have a client that accesses data from a remote cache server +lets remove this wgetcurl dependency +MikhailI tweaked build to include to compilation classpath please try AlexeyHere is the fix with regression testbuild test passes for me +I blindly applied this in +This is intentional since HDFS is only private +is deleted in database but is left behind in runtime causing the to fail +Because these are also used in other applications we can not add any dependencies to Spring Batch +being used in a unit testRegardsHiram +No problemI started my comment a long time before yours but i commited mine only minutes after yours +The lookup value should only be set if its +I dont think we need to keep this issue open +Future If you search for a table in +Most simple pom that demonstrates that behaviour +I set this back to unassigned. +It is strange that v patch are built and passed tests well in local +For some reason I cant log into the bug site +Sure ok +The string should be copied into the edit field +I will respond to your emails as soon as possibleRegardsManan +maybe we could use jackson which can support multiple format +I was able to reproduce the issue and I have updated the test case to reproduce it as well +If it is necessary then I will investigate options as Red Hat cant provide support for RPMs +HiOk no problem +Can we move this across into trunk then +that you dont handle outputting a usage message that includes defaults for port and address Are the methods get and get copied from? If so should we make it so this code refers to them over there? What would need to be doneOtherwise patch looks good Bryan +It has nothing to do with action or annotation simple disable Spring Security at all and check. +If you look in the PDF version of the spec that possibility is not permitted by a compliant implementation +And it is not too comfortable to live without DI +Soon if now show stoppers appear +If it is important to yall why dont you provide a patch for it? We all work on things that are important to us welcome to the world of open source +Reviews please +Timsorry for the delay I was on vacationI confirm that the patch applied fine +Similar but slightly different wording would be required for illegal keys +just do +The code will suppress the full exception stack and just log a warning now unless debug logging is on. +Clone of the simple stateless bean with a deployment descriptor instead of committed in revision +Verified by Leo. +Wow that was some time ago I couldnt say if it is still an issue or not +seems to be the correct place to have currentDatabase state +serving write request removed the rbw block from rbw directory to finalized directory +Thanks Sid! +We want to test it firstWhy does that matter? If we come across an explicit match before isnt that fine too? mmm i like the separation in funs here easier to read at least for me +We found this by accident as increasing the version number of our software cut the build time from about hours down to think some repository caching would really help here +Thanks Areek! +Its clear but a patch would be very welcome anyways +Are you sure this does not already work +out of hunk FAILED saving rejects to javatestingorgapachederbyTestingfunctionTestsmaster patch for trunk +WTP lacks a delta model so changing this property on the server will only alert me that a new deploy location is set and I will be unable to access the PREVIOUS deploy location to properly clean up +There is a shutdown attribute on the broker that will ensure that a master dies when the slave dies both need to be restarted to continue +batch transition resolvedfixed bugs to closedfixed +Is it possible to be more granular and suppress a specific violation +This is actually a Java backend problem +add conf to svn just committed this +Unless Im mistaken you are seeing correct behaviour +Hi Doug thanks for your review commentsYes Handling of symlinks should be addressed in in +That might actually be a clearer way to express some of the logic in though the serialization becomes slightly more complex and wed add a layer of indirection +I guess I was expecting the editor to recognize the DGM method instead of the method on collection +The biggest advantage of Maven the pretty HTML has been achieved using Forrest. +Please give a little hint on what you are trying to fix in this onthe richfaces libraries will be removedreplaced only if a user adds his own richfaces libraries to the JSF Portlet wizard page +Therefore the fix detailed in is a bandaid to the real problem +Downstream users then arent able to effect the operation of the iterator +As soon as I have a working version Ill attach some screenshots is now an interface this can be done using a new implementation +We have the same issue +Reopened +The restriction on the shouldValidate method as private makes this trickier that it should be +would you please try the patch thanks a lot +naming scheme +The exclusion is in the plugin pom and its working for If I read your comments correctly this isnt related to the xml beans maven plugin at all +Thanks for reviewing guys +Thanks Sandy +Accepting for you also tried this against the CVS version +Im linking this to since regardless of how we store the spatial point field types we should have the ability to output those fields as georss per ryans suggestion +Thanks Army for the good explanation +Pretty sure it escapes but its better to check +Functional code is moved out of the and into where it is centralized and can be overridden by jacc implementations +Im currently focusing on exposing the to the Jetty HTTP server I dont think exposing the information via a HTTP interface is the right approach in its entirety +Changing the component to system controls +Can you tell us which version of you are working with? About how often do you see these exceptions come up +The code is there but commented out +Should be fixed now +Im for that too +Unfortunately the leader logs were not available +I polished the patch a bit +slate blue coloured p and d as per carls other alternative Qs +Surprisingly all java files have headers on this one +Initially scheduling this to be fixed for though Im not sure how to clean consistently in all environments +Deveraj We are trying to get this fixed for hope you do not mind me taking this ticket +Moved to +Fixed now with regression testsI had to make a change to the Groovy AST visitor support to get this working +Tested +f BEEHIVEHOMEant Cbeehiveprojectspageflowtutorial pageflowtutorial undeploy deploy +Rob do you have the log trace when running with AS successfully +In the end if we go with this new idea it would mean that the compactions would be faster and use less resources during bulk updates and allow more resources to other task running on the server like map taskSo my proposed idea would be to have two types of compactions +That would be great if you can +One way or another this will only be fixed in not in where its an NA despite using +I have observed it in one of the patch builds in which patch is not related to the testcase + +Sound good to me time to commitBest Regards +As the issue reporter please confirm that this issue is resolved to your satisfaction so that we can close the issue. +Adding scope attribute in some tables. +Added as a note to in the Hibernate Core Reference Guide for EAP and EWP +Since that information would be available from the log or out file that may not be a mandatory thing to resolve this issue only a nice to have feature +Anoop Thanks for the review +how would a user actually use this +Also there seems to be some issue on OSX with these changes as well +Did not re apply the patch and check +Note that all binaries are not the same +Here is imported from to make it possible to keep sealed +The fix that I suggested for should resolve this as well +Raphaels comment was because of the backout +duplicate of issues in to add link to duplicate issue in description + Added Jetty dependency to the examples module and removed it from integration module Added Jetty conf file to the examples module and removed it from integration m not sure if the patch was sent before Im sending it here again just to be sure. +I incorporated your suggested changes to Dag with some tweaks to eliminate passive verbs and inconsistencies +Hi Torsten This issue was report weeks ago I just added a link for it +moving fix version to and +Hi Sorry to nag but was wondering if this issue got lost in vacations etc +Making classes accessible from a regular groovy script would allow me to reuse all of that code without the pain of a major refactorIf there is already another way to do this Id love to hear about it +Unfortunately the current doesnt timeout correctly for which was created +commited at revision +The WS logic typically returns a data source that is converted to a blob by the datatiertuplesource +Nice work +It does not compile until is committed +Like this +per revision of the most mode parameters in the navigation rules not yet implemented + sudo u wali falcon entity type feed name update file updatecolo updated successfullyprism updated successfullyThis only lets us know update is successful but we wont know what coord has been set endtime and whats the new coordhave to with oozie for exact status +So the JVM waits for for concurrent gc to finish and block the thread until that momentFor my application Ill apply the patch youve provided plus the modifications to the so the server doesnt freezeI understand this issue is not directly related to this bug +Isnt this basically equivalent to the that weve got +I have workarounded this issue +Escaping backslashes is unnecessay on Windows because it is not necessary to use the backslash in the first place +New version of patch with comments to explain some of the has been fixed in Maven core for the release so a fix here is redundant +This is working as designed +Deferring soap service work beyond xml and console issues. +Tx +Attaching and a second patch that adds a link in two topicsM srcrefM srcref +uses yaml literals and has the following top level elements banner displayed when the client starts help displayed for help commands Help for each command grouped by the antlr node from the CLI grammer +Attachment has been added with description the corresponding transition resolvedfixed bugs to closedfixed +This shows the issue +Ill take care of the doc changes next week +Attaching the Ramya! Quick question is fully fixed with this patch? If yes feel free to resolve it as duplicate +We could make sure the writes always happen to a temporary file like or something and then do a rename +It makes sense to fix this but I dont know that we can meet your original fuzzy test +Im also on board with this ticket +There are shuffle methods in +WDYT? we can create a new one or update this JIRA +Reopening to edit the release note text will close again momentarily +Bumping the priority of this issue +for this issue being addressed stable generated ids are highly desirable for tools such has which build automated test cases based on finding components from their id +Ill take a look and merge next week +Hi David which icon is missing in your screen? or a tree node like in attachmentIf the latter that should be part of debugviews pluginThey are shown correctly from my side +No new feedback looking at the latest version of the patch +Most of the information as stated in the jira should probably be calculated at the RM +Set to no +new patch is ok to me +This commit will remove that code +xsd etc all appear to be workingI am using +It is also tested in the coming version of the patch +OtisFirst I am glad that you responded at all +What you suggest we do instead Hiroshi +I will have to retype but its no biggie +In fact its relative to the +Thanks DavidIf the compiler is not supposed to be calling arbitrary equals methods at compile time then it solves the problem because in that case it wont try to acquire a lock on another method that might be already locked by a second thread. +Unassigned +This was actually a Cache Core issue as indicated in +Also removed the Preparator and +pushed on freds seems to be fixed I verified that once I hit finish on the Add or remove dialog the files appeared on the remote server in a blink of an eye +Is this data dependent? I know I ran similar queries in the past without any problem +Adding a pull request for core as the api module has been updated +Hi Raintung I dont think there will be that many threads created although it depends somewhat on how the injected and are configured +Dont feel some duplicate work +Sorry I missed you on IRC try again plz when you are around again. +We see the same issue when we bootstrap a classloader +The message Im sorry but the administrator of this site has turned off new user registrations +bulk defer of unresolved bugs to +However the actual systemProperties bean is only exposed to the in an actual +NET passive STS as an SP for being used by the Metro implementation +A snippet of all the stack traces over the mins the couch was unresponsive +Im going to try to find the bug and provide a patch +Verified +msec transferTo +Any objections for me closing this issue +The cycle check is now. +and put them inside the graph +Adapted version tag to be able to remove versions +Origin of this issue +Thanks Ashish for committing my patch +Assigning to Store component +Is the latest patch the one with the desired fix? If so then please mark Patch Available to run it through Hudson otherwise please post up a new patch first +Mojo has been created. + Test is added for the is fixed +From memory there was a bug about a missing apostrophe in a nightly release of a few months ago though it was fixed pretty quickly +Test cases related to protocol buffers were failingI dont think it is because of this fixIs it fine Ted +What about making this change in behavior configurable but not the default option? Then you dont have to worry about finger pointing but still give people the ability to speed up their builds if they are willing to accept a mild increase in risk. +Fixed some bugs in r +batch transition resolvedfixed bugs to closedfixed +There will be no release +I had to turn off XMLXSD validation to get around this + +This is better. +bq +I ran into the bug while unit testing changes to the implementations +Sure will revist after Iwill add dependency tag +This patch fixes the issue +personal seems like a good feature to have +I attach on top of Jonathans patch +You might want to fix it when you update the patch +yes +Many thanks I can confirm this has fixed the problem +HiIm experiencing the same issue in with the Image Pyramid pluginCheers has sample data to try this out +Added info for a couple of build options +This will work for hadoop and for hadoop Uma +Up to now I have successfully built the source package of derby +SHAbedcbfdeeefed +Upload your change and Browse it prior to CommitYes there is a discrepancy between thejavascript markdown in the Preview and the python markdown used in site generation and thats unlikelyto be repaired any time soon. +First attempt at a patch using my preferred solution removing the code that reads partition information from the HDFS entirely and instead relying on the for accurate informationAlthough I assume the code was put in there for a reason so Id love to hear more about it +Not committed to FYI +I just committed this +In one or two weeks I may get time to provide a patch for this. +ok I did update and merge for you this time +Im able to do it on the command line just fine. +Its unlikely that you can change it in your ajax request +Yep that does the trick +AIUI the choice of license needs to be documented in the LICENSE fileAlso in the NOTICE file if the license conditions require an entry in the NOTICE file +I thought the PFD was the most recent version +I cant reproduce the issue either +Przemyslaw did you manage to test this with debug mode disabled +committed to trunk Ill add changes and back port to as well +I got PE regions into single HRS before HDFS went bad about M rows of k values and byte keys +Added v which fixes the case +bq +has been marked as a duplicate of this bug +Could you maybe add some info to the user so that heshe knows what will be the result of the operation? For example waring that the remote files will be removed or info that heshe might want to merge manually between local and remote repo +Until we develop a better API we should certainly document this as a part of +Thawan I was looking at the patch and it looks like you always have one acceptor thread +Dupe +Mark tested on his Fedora and I tested on Ubuntu with no issues +A different instance of the same target type +Do you have an example of where ugi contains tokens but security is disabledYarn is moving to tokens regardless of security +Until then it isnt fixedOn the bright side Im actively working on this one now and the release WILL come out +Perhaps you could attach a modified patch that moved the logic to +Release is now generally available. +Should be fixed with inclusion of JTS the fix for version the upgrade of JTS to happened on trunk but not on the stable series +in the most of changes I got no problem +I believe you mentioned on the ML that you have data that compacts fine if you increase the inmemory limit but not with lazy modeIf so can you get me a copy of that sstable to test with? I can set up a place for you to upload it privately if you cant make it public +All repositories were patched so that they pass this test +Second choice I like the loggingpen reference +This JIRA may have broken backwards compatibility as testcases that were passing before are failing now +if you want to still use an aop proxy you could perhaps return a proxy from the factory bean +The patch you proposed is a nice optimization too but Ill fix the real problem as wellThanks +Provide a patch to improve help messages of deployencrypt command +Hi AlexThis is indeed kind of unfortunate +I think to compare the patches we dont need to worry about changing the of partitions simply because there is no change in the Log layer in both patches +Just running a CI run to make sure everything works +Not a bugfix wont go into +Investigating on how to best configure mailman and postfix to implement this requirement +Basically I just added a second dir which includes a to the classpath +I just wanted to be sure that it wasnt a problem first +When I added the start to a spark string validator I just made a tests folder next to the src folder for the component in the whiteboard +simply do the edits on the text file and save it through eclipses normal documenteditor saving facilitiesHenrik any reason why you dont actually use the encoding as stated by javadoc of +committed +Minimal test eclipse project and ant output file attachedNote The project uses absolute pathes references to the enunciate libs and the JDK +Merged to as revision . +Added patch that makes the Hadoop nfs port patch looks good to me +GET commandaddClusterzoneIdhypervisorclustertypepodIdusernameadministratorpublicvswitchtypevmwaredvsguestvswitchtypevmwaredvsurlhttpAFFVMWAREDCFVMWARECclusternameFVMWAREDCFVMWARECresponsejsonsessionkeyFzzgDHence UI should skip the following parameters if user does not want to override the public andor guest traffic settings at cluster level publicvswitchtype can you check on this issue? +It comes through and it means hibernate is selected +I have applied it in Rev +Pythons optimize mode is pretty silly at the moment but I see no problem with this as long as it doesnt interfere with the +Attached a log file with the reported errors +However isnt accurate when calculating the changes on the document so extra validation is very much possible on text deletion and replacement operations +This is most likely caused by the lack of caching in wtp and will be fixed in wtp be clear the add remove dialog has to check project heirarchy which project is a child of which etc +I think that this should cover the comparable Derby types +trunk looks like this is broken under jdk +Yes comes from the WS source so this issue has missed this CP would suggest modifying the script using ant for this release and create a new issue for the correct fix in the CP release +but can we add a system test w a ks to help prevent regressionsEdit never mind thats what were already doing in hence the continuing clitest +Certainly its not essential but having the variable interpolation instead of would make it easier to adjust for all these projects handling SVN relocates etc +Switching from LTQ to a custom elimination backoff stack appears to have dropped the to us +Thanks +I think it partially depends on the client technology being used and whether it would bawk at garbage attached at the end of the envelope +I have seen the same bahavior and this is a show stopper for +net client sandeshaC serversandeshaC client +We need to launch continuum in a phase. +Thanks Tim +If someone wants to provide infra with some actual text to use we will happily deploy it +That line contains a bug I should have used min instead of max +So I should consider one task at a time and alternate between task types +Could you please try this patch I am not sure I understand the expected behaviourThanksRana +Note the on event listener is still invoked even if no documents were created and commit was skipped +You had me worriedAlso there are lots of overrides on interface implementations +Perhaps use Commons Collections to get round the Java restriction if required +Fixed on rtrunk closing all issues that have been in resolved state for months or more without any feedback or update + +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Max we already have RSE support and deltacloud +Maybe it doesnt work that way +An example of configuration would help a lotThank youThe same config file can be used you just need to change the portsjournal dirs to suite this is no different from Andy Im just trying +svn and CVS directories and anyIf there are any CLI operations that are still failing because of this reopen this and Ill make sure those ops are using the right utility functions instead of doing things by hand. +To solve it in drlvm Ive added wd to the This is the only way this issue can be handled as defining magic symbols leads to symbol redifinition in and build failure +The changes include constructor changes to class +In fact at one point everything pretty much comes to halt +This patch +This patch adds sample methods to +Ive updated the QpidProtonMessenger class to raise exceptions on methods with return codes +Or is there a use case I am missing? +I dont remember at this point but what Jesse said sounds not sure what I was asking for +scratch that +Yeah I like the idea to keep the old around for a whileSo whether we do a or rename the existing to is well a matter of taste and whether we want to hazzle existing users to change their deps to quartz to keep using the version of quartzThough in the mina case we did a +Failed test now passes on localAlso addressed the code review feedback except for stability annotations these are internal interfaces I wonder if stability ones necessary? Theyre not used on similar classes +I have a suspicion that the infinite loop originally reported may not be a problem for to resolve +Indeed from the original description the folder is Cderbyttstsystem which is more than characters long +I thinkCheersChris +FYI I created identical FLUME just now while retrying the JIRA creation button when getting this error message +Does this work on Java or do we need logic to skip it? Naming of the new classes Would it make sense to drop the prefix and instead follow the pattern of the by calling them simple data sources? Like etc +This is a clone of +The Javadoc for saysbq +Im seeing microseconds for a small info log message with one variable and one function call +After regeneting the model the classpath gets screwed up and has to be fixed using mvn eclipseeclipse on the module is a second attemptIt only adds a metadata map to and ends up touching few filesmodified +The patch +and actually if were saving it cant we go on to add this approval to the pom to streamlinePerhaps Im not all that wise on the ins and outs of licensing and legality +PASSEDThis bug may be closed as a DUPLICATE ofThanks as duplicate of. +Select all works on windows +One question when the plugin version manager rewrites the user it will likely wind up in a different order than it started +Attachment has been added with description screenshot of the bug visible in can see the bug in acrobat when you zoom in to +Thanks Bikas +Weve seen many other tests fail with similar timeouts thats why I suspect its a more general problem +testedI dont know why but the metrics names are still vertically badly aligned on top +Im going to see if we can upgrade to something newer in our labs +but was not sure if it was normal practice or not +Sounds appropriate for a newcomer +Fixing the path yieldsls la which xcodebuild root wheel Jun ApplicationsContentsDeveloperusrbinxcodebuildls la which xcrun root wheel Jan ApplicationsContentsDeveloperusrbinxcrunxcodebuild versionXcode Build version HThank you so much for your patient help Shazron +Yes instead of or besides the greyed out Run local analysis we should have a Refresh from Server +The user can now edit the incorrect classpath container and have a list of runtimes to selectThis is NOT the approved UI for the users use case +and we got the same problem as the people already mentioned abonvePlease fix this problem! Or at least take a stand on this issue +And I cant see any good reason why you cant decorate just a Class +They note it as the canonical name for the ISOLatin encoding +Also if you could avoid tabs in the future that would be appreciatedIm not sure if you want to do more work on this jira please close if you consider it done +How often does it happen +Committed trivial patch +This is a browser issue not a Struts issue +SankaWe need to fix this forThanksCan you please take this upthanks in rivision . +Findbug warnings related to class were not introduced by this patch +Good find +I think everyone is waiting for the new request processor sandbox from CraigBest adress this then +isnt Swing already +Deferring as Wildfly still bundles Apache Client +Thanks for comments +bq +q file +I found one issue with the existing patch will upload a new one soon +If there will be more problems then i will open new ticket + know this ticket is still valid +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +James are you sure you added your comment to the rigt issue +Tested on and +In progress nowAs per ASF standards the repo name will be +This one is good to go for both and HEAD +I think we have to leave the job that is currently completing in memory +Verified with beta builds closing. +fixed by using a regular expression to parse scripts out of the markup on the client is a must for certain user interactions +A summary of the showing how things have moved +This jira issue deals with running a spring project inside geronimo and giving it access to geronimo components deployed in geronimo using geronimo plans kernel gbeans etc etc +Do you want lucene or lucene +Then we do not need to move the file position backwards +Reviewed by Keith +So it may not be doableI have reverted the import change manually +where can I get the sources for the maven andromda plugin +Ill look if there is a new updated derby version +The will wait to fix up its settings until after the childviewport has its measurements +This is because of design decision taken in the past +The changes for were simply wrong and the code was reverted to the previous revision +Seems reasonable and low risk to pull the ID from +The is changed also to first set the proper producing entity in the connState then submit the request +Attached an rebased patch that includes the changes suggested by for taking a look Zhijiebq +This is a really very small patch +Test case +But sure make it a bit also for demos that you can click on the left sidebar on the title of the placemark there and it will pop up for you +Reposting files +Will being able to filter based on regular expressions run against a document URL be sufficient? The web connector uses this strategy but it seems to me like it would be problematic in an RSS situation +wbuildrsvn ci libbuildr libbuildrjava spec specjava CHANGELOG m ECJ dependency now required to build any java projectSending CHANGELOGSending libbuildrjavaSending libbuildrSending specjavaSending specTransmitting file dataCommitted revision . +When trying to distribute a +Thanks! +next time please separate code formatting from the actual dont a single buzilla bug all new reports but you are right too much from the above +Thanks Stepan. +once you import your you no longer care about the import +An example would be saving a value to a static variable this is generally not advised but may be more efficient in some cases +we just need someone with some mercurial experience to write the branch command +Wont be working on these and theyre old +Oneday this will be fixed +Attached patch fixes the NPE though I havent had a chance to test on an actual misconfigured clusterIs this the right approach or should we just do a stricter job of verifying the results of the topology mapping script to ensure that all hosts are the same level? Does anyone have a legitimate use case for locality of varying depths +Really let me know if this fails again and well see what to do +Not sure what the problem was. +Here is a better patch +For anything i have forgotten pleases either or create a I shall redeploy this page if there is no comments +Thanks Shane for looking into this +Escalated to Atlassian no feedback yet. +Jetty builds with Maven which definitely would complicate the build even further by providing yet more build dependencies +Well report our suggestion of generating Java classes is no solution for us because we dont want to be dependant of generated classes as our +After taking several thread dumps threads are always stuck in same placeSo for now we are not taking the proposed modifications in our productWhat we do for now is to avoid creatingdeleting nodes under the same parent in different threads +Im not sure why that would be any slower now than before other than multiple calls +Thanks for the links +I did see not any performance difference but I did not measure it +The references to the columnsOrder attribute but the component does not define the attribute + Here is the patch +At least we know now its an issue throughout the chain and its on our radars nowOut of curiosity did it break the same way even though there were no appends? Or was it a different failure +I believe this is a jBPM specific issue that will be resolved in a future release +war or something to the effect +I have the same issue with that doesnt detect the last term during migrationI dont know enought about lucene to fix have tried to merge the with the but no success eitherShould the indexes used in the test be upgradedDo you know how they are generated so at least we could test them +See my comment on the tynamo user list +To continue Ill SVN copy into the new directory first and continue review then +Editorial pass over all release notes prior to publication of. +Chris very good startMinor issue I dont think that the parenthetic statement multiple calls to makePersistent with the same pc against the same pm will result in a is correct +will be working on backporting this fix to at least once nightlys pass across all systems +and wicket is not inside the projects so it does not result in a broken excel and is not in there either thus this need to be added as a separate thing to do for Seam projects +I do understand the importance of what you are suggesting +Looks like a lot of fun work ahead here +I suggest to not fix this +The intent here was to make the logging of block events the default +Ill confirm. +The console should ignore them +But we are doing it on the older list of regions +And since MRI folks seems to be acknowledging that this is indeed a bug not a feature we probably should just fix this and be done with +I have no idea how or why +However until that changes reflection was the only way +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +get this into the main codeline ASAP +I will change the key of the to File which represents the currentDir of the +Its to late +Hi SebastianIn order to reproduce the issue try to change the timezone of your client osI think the issue appear only if the server timezone and client timezone are differentI tested it on openmeetnings demo server +Hi BryanYes it doesHere I am attaching the log file with this and the test output file with this +I have no problem with the format command generating a cluster id when one is not given +merged with thejas change for manually run related tests and they all passed +To be committed to CVS in a patch to resolve the issue +Summarizing Sebastians scenarioFirst attempt on that cluster was made aboutThen patched TRUNK was loaded +to give the feedback ive promissedIve tried the producer turned on and the lock problem persists +Thats perfectly sensibleThanks for this report +I already implemented it +In my case it wasnt necessary to add scope for full launcher dependency in module +Alejandro raised the following issues for discussion rename should the target path contain the webhdfs prefix since a client app will want to simply take the target path and use it as part of a read operation +Im using version of RubyAWS and experience the exact same behavior + +Resolving as won for the reasons described above please if Ive misunderstood the issue. +The exception is gone if I apply your fixRelated to thisThe exception that I reported does not appear on console when running with the standard console from felix + +Im going to steal the patch and rebase Sams work to trunk +This issue applies to any object used as a key +Even if it is a bug in the JDK Xerces should have a workaround for it +From a code review of trunk this is already resolved by other fixes that have changed the locking strategy for the destinations collection +Do you still get this error Jerry +Committed the patch to +Will file a jira improvement for thisLets move this patch forward for review + is the patch file name +Please check the changelog from the previous release before opening new issues. +The problem remains on Solaris machines it looks like there wasnt applied the patch +set permission of the file to +Ill cancel +Tagging this for since we can redo a lot of the anyway +This is an easy fix that should definitely go into on JBDS to set fix version +Tim Im sorry for not doing this myself +That looks good thanks Nathan +Merged logretentiontestsuite into replicationtestsuite +As a result this patch would have caused a conflict so I deleted the old patch and am replacing it with a new one with the same changes except it was made at a higher revision number +Duped by +You are saying there will be no decommissioned replicas and therefore all this code should be revisedNope +Revison +OK I had hunks fail but I managed to apply them +Tomaz is this on any xml file or just in mbeans xml +If this is satisfying you I will take a look at implementing the optional CONTINUE IDENTITY and RESTART IDENTITY clauses +See for further details. +Schema class has a constructor to take both the XML schema namespace and a prefix +It depends on Spring autowiring Hibernate collections HQL queries and auditing for the bug to show up all of which is proved and explained in the zip +verified. +Wish Id looked into it more then +Using to create records wherever necessary +Am I doing something wrong hereVersionMessagesSend SpeedReceive Speed BK BK BK BK BK CRK CRK CRK CRK CRK Test I mentioned previously the defaults have changed for sync between beta and CRIf you want to do an comparison you will need to set them you just compare the out of the box defaults this is NOT an see the relevant sections in the user manual +tldr +i herewith attached the patch for xdocs for new Synapse xpath extension +How are you passing null to theIts good to remember that this is a special purpose meant to be used exclusively in combination with the Lucene +Repo deletion will take place when I next do OS level maintenance its not supported via the maintenance screens as yet. +Ill take the issue +Jed please dont one jira issue to report multiple bugs +It might be related to the fact that a newer driver is used here compared to the one jBPM is tested with +Do we need extra output in or was it a debug change +Sanne can we close this bug as a duplicate for +Well investigate this as part of a container deployment but we may need drop the notion of a shared installation +Fix and test case look good except please add the apache header to the test case and preferably update the test case to you for your advice +However the text position is every time I +Are we only committing one file per commit? If so +Thanks for the information Florin folks should find that this. +You may not want to load them +components versions descriptions etc +It definitely does not block a release +Feel free to add parser support to that one +Thanks verymuch for taking the time to do soIts too bad we cant set the deadlock timeout to less than a second +Thanks for this Jon +Dead lock issue resolved in +JonWhat is the benefit of moving protocol specific logic back to the? Presently is just a simple bean class that carries state information about a cache entry +It looks perfect for me +Thanks Namit +We cannot yet close it in clear conscience but I believe none of the remaining issues are release blockers + +Hi MarshallPlease review the attached patch +Actually tools using Ivy in multiple threads need to set the in each thread for Ivy to work properly +It will be used only by so it should be in the thrift interface of +DanielWhen do you think this can be checked in? We would like this on branch in addition to and trunk only need to add test case I will try to upload the patch today +Patch introduces concept of closed and unclosed +This issue is not observed +This sounds like a classpath issue +Currently without ooxml support +Patch to add Chukwa to Hadoop SiteChukwa content added to and files +This was implemented as part of conversion of core model to RD but it needs some tests +Now even if the folder exists it will tell me that it doesnt +Each under construction blick has two list of locations the first is for keeping track of locations in a pipeline and the second is a triplets for tracking of finalized replica locations +the submit of this page would put a message in the approvers myportal that they have a request to reviewthe approvers page should have ability to modify days and startdate and an approve or deny plus a place for commentI would leave room for future to let the approver split datesthe Status would be LEAVEAPPR OR LEAVEDENYif LEAVEAPPR and SECA will deduct from The LEAVEACCR for that particular type of leave for that party and add a record to LEAVEUSE +How about doing more aggressively by checking +If that doesnt solve your problem please provide a quickstart or test case +Theres no guarantees with receive if the message is still enroute to the client and you call receive youll get a null back +Can you maybe give some example your link to something related +I have created a testcase within have found more errors related to the NPE Gavin described +Just very busy right now so this will have towait a little bit +This generated illegal code in both C and Erlang which is why it was disabled +push to +Theres two ways to make Harmony and the JDK agree change RI so that headers are in forward order change Harmony so that headers are in reverse orderSince forward order is what application developers would want the best fix is to lobby the RI maintainers to fix their code +Is there any chance the build is being set up with in the +I get the same error +Ive added the repository inside my active profile and still getting compilation errors +A NPE will be throwregards +tests now pass +Trunks code is checking for null in both cases of and +Where has to go the attached file +why you ask? you see some possible drawbacks +If its required to make the outputPath relative to the setupscript we should do this inside the mojo not in the configuration +Stack trace continues to grow indefinitely +branches +Plz review this patch and let me know in case of any concerns +Its the location for global settings +This also affectes the parser +I would think this definitely counts as a issue +Thanks! Claudio +I propose we handle that first +We can revisit in +Closing stale resolved issues +so i guess the whole palette has to be backported to +I have some pretty cool ideas for it to make it efficient in a scenarioGary have no fear +Imreally glad its solvedA big thanks to everyone who took the time to look into this. +Remove VMRTNULLPTREXCEPTION runtime helper for x +error report from rational Purify for echo client +Requires documentation +If you have a state that represents a union whose content could not be read then the code for reading a union looks exactly the same as the code for reading a structure when viewed from the perspective of the caller +Heres an updated screen shot with the domain security bits added +Ioannis thanks for the patch and +Closing old bug reports with no updates the last you feel this close was unwarranted then please reopen or create a new one with link to the old and we will revisit. +I fixed the issue although I put the fix in the URL component itself since returning null is the expected behavior +Youre right it was an existing project group and it picked up the default build def from the group plus the ones from the template I selected. +on patch +yes indeed the problems are now gonethanks a lot Michael +Hi I refix this to use instead of the new gbean +Remove label on restart function since we now have a doing trafficline x. +is kind of an odd issue as its title is Test throughput versus +and higher up the assignment will be dealt with +Hi thereThanks for the are reviewed and applied to svn +The change I was thinking was in it currently reconciles the difference between the block information in datanode process with what is on the disk +Also please add one more test to cover your modification here +Something in the new version must have fixed the problem +Was this realized +One question I had was if there is a particular recommender that would be a good starting point to begin this effort +The failure doesnt happen on every run though +Updated the ref manual topics per comments receivedThe files included in this patch areFilename Title Summary of updatecrefjdbc JDBC Reference Added XMLrrefbuiltchar CHAR Added XMLrrefcreatefunctionstatement CREATE FUNCTION statement Added XMLrrefcreateprocedurestatement CREATE PROCEDURE statement Added XMLrrefdatefunc DATE Added XMLrrefdayfunc DAY Added XMLrrefdeclaretemptable DECLARE GLOBAL TEMPORARY TABLE statement Added XMLrrefexcept and error message reference Added XML error messagerreffuncxmlexists XMLEXISTS operator New filerreffuncxmlparse XMLPARSE operator New filerreffuncxmlquery XMLQUERY operator New filerreffuncxmlserialize XMLSERIALIZE operator New filerrefhourfunc HOUR Added XMLrrefjdbc Mapping of to SQL types Added XMLrrefjdbcrefsqlxml New filerreflimitsxml XML limitations New filerrefminutefunc MINUTE Added XMLrrefmonthfunc MONTH Added XMLrrefsecondfunc SECOND Added XMLrrefsqlj CAST Added XMLrrefsqlj Data type assignments and comparison sorting and ordering Added XMLrrefsqljtypexml XML data type New filerreftimefunc TIME Added XMLrreftimestampfunc TIMESTAMP Added XML +to be used more effectively for DDD based Repository should not contain any business logic therefore should not contain any finder method +r +Ive just committed this +For some reason JIRA didnt notify me about your comment +Fixed in the attached vbq +Once sub tasks are done I will move it to closed statethankssudha +Including top of output indicating this was under +But once I fixed that this test was failing thus exposing this issue +Any deviation can be handled by SPI implementations optional configuration properties etc +I am guessing that failed links arent recorded or generally reviewed so I see this as a place that parsing and crawling could fail and it would be pretty hard to track down +Not sure this iswas a validgood fix +When can we do this? any time +I started working on this +Latest patch now including trial balance report to show the ledger is also balanced at a segment level when posting from order patch correcting problems with quickly add transactionsno updates foreseen in the near future +Its similar to security services you have permissions and permission groups +Id like to have a last look at the patch since I havent looked at this issue in a while +Even if you did that youd still have to do it at the client as the broker cant account for the time a message spends in transmission or the time it sits in the clients dispatch queue before being dispatched +This has been committed to trunk and. +I can give this under creative commons attribution +What about following behaviour +No the session is definitely not expired +This is due to the munging plugin not playing nice with Maven +Thanks +But if you did some changes to demo data and connected the prodcut to Mouse Hand PollYes I connected the product to the poll +Wonder if it is Cygwin only +Before I did this I was getting a new coverage created each time which didnt happen with datastores +I can reproduce the behaviorBR ve tried to reproduce it with the latest but it seems to be solvedI merged the doc with doc to a new document +The stream ticket has been engineering QE verified +current status the following need to be resolved before the release Blocker An Issue with lots of votes where patch seems validAnd since this critical one just popped in with patchI will resolve today and the other two should follow with hasteETA no more than two days provided all tests are consistently passing +Does this affect +Thanks +Added a junit test case for the user data class +Unfortunately no unit tests for this +This patch adds a variable for specifying the location of the eclipse uidoc plugin and signs that artifact along with the others +Can you tell if the jar from the update site is truncated +Looking into test case failure but having trouble reproducing it +It passes on lin and win and original intention was to commit + for the patch +Toni just pushed the release to m central +The reason you have NL instances is because for the same mapKey you have content +It is rather difficult to cause an Error instead of an Exception to happen except with the change in binary compatibility caused by +Do we have a jira open for that +Reopening so that I can update the fix version field. +Would just changing to a local context work? Thats the default for clients so I would assume yesA temporary fix was put in to copy the request context when pushed onto the background thread +Liaising with the Seam team. +I am also caching one connection which will be closed when the resource adapter is stopped +WDYT +Congratulations on the awesome work +Changed relevance categories so that it can be used in content assist +Thanks for the review Bobby +Were doing a region split and it seems that the test doesnt think we will splitIn another may be there is a real bug behind this +Seems okay to me +Thanks Petr! +I made the class serializable and its working happily with the latest snapshot. +On further reflection it may be a resultof using a custom parser configuration +hey guysi have a snapshot on my own rawson directory +So the only solution Ive found is to do the maven goals successively +I dont want to simply give up just because of an isolated temporary network connectivity issueSo it sounds like were left with the staging directory issue + +Thats odd + needs to get redone without +osnon stable com stableI know os projects first fixing thing in os version than early releaisng and later merging to stable commercial branch +The community prefers to retain Rat as the name for the original product and adopt more unique top level project name after graduation +The only outstanding issue is why the viewAccepted event is not triggered on the member whose Network Connection is disabled +Let me look at it again +brings in the conversion also for and removes the in derbyall suiteRunning and derbyall. + is my first attempt to improve native memory management capabilities +If you remove the part then your class is extending Object +Unless theres a problem why not go with whatever we can for x and sync later +If we want to maintain any semblance of It just works then well need to keep maintaining this dependency for quite some time +If the test is designed to have much tighter constraints than a typical production cloud then it might not be something I have to worry about +Please provide an javacomp namespace should only be visible when using the defaultinitial context and that entire context should be read context used is dependent upon JEE context and nothing need to explain how other contexts are involved preferably witha you just want to discuss things use the forums +There is a minor changes in to put the call to within since newer versions of Hadoop through in this constructor +Alexander such details is best to discuss at the camel developer mailinglist +But it still does not apply +I made a start and changed all the single email address notifications to go to issuesThen I noticed someone else in there so I have left it alone for now +This patch applies with release and all Junit tests pass +This will most likely something to work towards for the release +Bulk closing for +Hey GavinYou got it! Hopefully should have some time today to get going on thatCheersChris +Closing as not to be fixed I still think this is ugly but if the runtimes in the list are there because each of them is exlicitly required by an example lets just leave it as is +it is the showing I can not modify in the Tree View fro structured editor only has support for those things that were in the seam XSDs when we generated the editor +Seems a problem in the jquery code in line no +Bounce to Sneja for help advice re missing upstream dependencies +Done at rev +Wouldnt changing the package name be necessary to avoid the potential for weird classpath issues if both ant and plexus were being used in the same project +command +please ignore +Ah OK +reopening to update component +I have committed this to trunk +And the memory limit never goes down even if you stop the producer and consumer program +Sorry my bad +was that new feature with direct HibernateJTA synchronization that caused the current trouble Interestingly noone noticed in RC +I agree there should be a single set but it seems like the ones added by the should be consider the correct set +Thats OKThank you Mark +This patch corrects the shutdown behavior of Kafka mirroring +Screen +My bad! Thanks again + +Please attach a test case that demonstrates the issue +Ill apply this patch when Ive worked out exactly what has gone wrong +Verified on Geronimo plugin . +But I see nothing wrong with this +See for a simplified test that shows adjusting the poll delay +Looks like WSRP related might need to look into JCR mixins to be able to add dynamically new properties to schemas +Contributing guide is fine! +Committed +Works fine for me. +bentmann added as cargo good to meHopefully all tests and checkstyle and everything are passing arent they? +Redirects should be handled in so they should not get to the +thanks. +I agree with both of you that it isnt necessary because it has no influence on the that is created +Patch adding string constructors to mutable numbers and adjusting tests for the branch adding string constructors to mutable numbers and adjusting tests for looks good but tests for the String constructors are needed +Thanks DavidEverything looks good so farSteve +The API should be a separate jira so Ive created +This ones faulty +Its not something I know how to reproduce at the moment +Clearing browser cache doesnt change anything +Weshould provide an abort method that does that +Agreed +It would help if you can manage to create patch or at least capture compilation errors save them into text file and attach to this issue +wfm +Just to let you know the licence is MIT and that the arch switch is no longer required +Were hoping to see a resolution to this one! Thanks +Take your time on this no hurry its already working with themes on and even those where its not working actually can be used in the meantime since the links are there +Issue no longer exists +I solved my problem with Criteria API which always produces full wish my project was not a high priority. +committed to master +I cant even view mb json docs having just strings +were not developing Deltacloud any longer +Im pretty confident this worked before producing output with the layout applied +The use of in the fetcher indicates an unrecoverable error thus this issue is not a bug and in fact describes the actual intended behavior of the system +Im still keeping this one open to address the feedback from Erich and also give James a chance to take a look Architecture and deps +New patch resolves the serialization of commitoptimizecloserollback actions that should not be concurrent with each other +Ok Ill try to update the patch later today +They removed my summary icon +Sorry RajI have no time yet but I will take care of it +I tried on trunk assembly and there are pases and failures +This should be fixed in struts +Would it make sense to do something like this start with a set of bundles to resolve obr figures out all the additional bundles needed for them turn this entire set of bundles into a feature copy the bundles into the system repo and make this a boot feature +Just need to make sure that what that old unit test does is also done by the new unit test +fixed in after release +Now basically delegate its attribute and task to and run s execute method to do the task. +Fixed on EAP ER +When I can reproduce the problem I will open a new issue regarding to this point +committed revision r +Verified Windows bit wrapper install net start karaf net stop karaf and remove on Apache Karaf. +the actual updated patch file this time +Right so the issue is about nonces being not validated +bug is not reproduced since was fixed config reader and empty value of password is stored as empty string instead of NULL in the M versionHowever I think code which is described above shows not good idea for checking user existing. +Well try to add source jars correctly to the repo in +Tom thanks for correcting my JIRA misuse +There is one complexity though +Seems to be cache problem its working now + +Also a point to consider is that the more methods are added the slower groovy will startup since that complete registry is done once at startup +I agree +Committed. +If this comes up then we can split off another test file to do so +These changes are in both HEAD and in branch. +options etc. +Looks good +Looks good +Verified with JBDS +Thanks to Szehon for the contribution. +Note that in the future this should probably come with a modification to and also perhaps a release note +Works as expected with the latest build +I assume that we only know all the lock timeouts on the last suite as we lose the from previous oneDoes it make sense to anyone that the test is not reporting the lock timeouts as errors on the run +And indeed the versions generated by maven dont have which contains both classes not found +Thanks Doug +Regards the Copyright comments +Double check to see if the updated is in your environment and that it is getting picked clarification around the username JCIFS specs is also appreciated +I was actually looking at the installation guideOk so then it is RHEL vs windowssolarisIf you ask me I prefer a single extras directory in CR for all binaries +Bob is running into issues passing lists of primitives as method parameters. +Please apply themI have merged the cmp plans withThere is a problem with prefetch plan +Now recoverLease is called on same file +If a similar fix is required in master andor please provide a new patch +I committed a partial fix in for Float but for whatever reason I didnt also add Fixnum and support +For starters the API can be exposed internallyspi and later we can consider some nice DSL supports so you can attach custom code to be invoked +Fixed +Writing to S is to be unreliable in fact the document makes a big deal about thisWhat is your experience in practice? Does streaming to S work well enough that Mahout jobs do not fail? +Ive tried writing a new and incorporating the new into +Alexey could you please confirm that the change fixed the problem +Sure Ill add the location to the class javadoc of on commit +You can always use custom readerswriters when using the if the provided converters arent flexible enough +In order to help restart creation of appToken needs to move out from +Details on the project architecture and is happening with this? +Claus checks that a is thrown if the endpoint url contains an invalid httpClient configuration like true +Same issue on both and branches. +Im fine with restricting this scenario to only when is present +is loaded first +The most significant issue is that currently released versions of ships with old releases of MINA and many things has been fixed in MINA since +I have the same problem with the version attach your and Eclipse project with that would allow to reproduce issue +At first glance it looked strange because there were transient fields in classes that were not serializable but looking back at it it seems to be because some of the subtypes are serializable so I will leave out those changes in my updated patches +What does this mean? FAILED Execution Error return code from ATTEMPT Execute +nothing special just basic integration into the modulesanalysis tree +Sorry forgot priority not major! +Possibly or +Yes it works the same for Eclipse with the checkstyle pluginHowever there is no automated way of just pressing a button and getting things straightInstead one has to manually fix every single issue +Trunkjobhibernatetools +Here is a first patch with a fix test is visual I have to add some checks to it +I confirm the use of protected variables in ODE is a design choiceIt was driven by performance considerations and the need to change data in place at several levels and avoid copying as much as possible +Only on Darwin +Accept that the same user may be authenticated multiple times longer term review integration with SSO maybe at same time as clustering +I believe that you intended that we update the codebase to use things in the newer version of the jar +Awaiting test of the ts jars t break the quickstarts quickstart paths should be this issue for the will raise new jiras in appropriate projects and new SOA jira to monitor progress of duplicate removal in dependent that was removed +Attached new patch for this issue +Tested manually as well +change cloud users home to version if needed +attached patch +Thomas this bug should be fixed in that was released Aug +Need to wait until is released so that JSF bits of are available in Maven Integration library needs to provide wrapped response objects for JSF bits to work +Hi MaartenThanks for looking at this issue +Its been almost a year any update on adding this functionality +This patch contains a test and the fix to this issuebtw this patch was applied against trunk and I have not tried to apply this on an earlier version +I agree +I also retested with JDK and could not was reproducible in my environment at the time but not now +bq +where should connection info be placed pom or persistence xml files? And does Dali offer a place to map datasource names to actual connection bindings +Closed with the release of +I think Id prefer to go back to the first plugins page where you choose whether to install more plugins export plugins or assemble a server +java source code for details on how Axis finds a compiler to usePlease reopen if you can show how an Axis code change can fix this +Doug IanCan one of you pls commit this? Thanks +Didnt see any objections for days moving to have also run into this using +If you rely on row existence you need to get all the cells back and check yourself +This PO is already compiled and is in one the +from me +What timeframe would the ooo PPMC be looking at wanting to create a release containing these dictionaries +For example in elasticsearch world it allows people to build admin stats applications +reopening for additional patches +Can you tell us exactly how to reproduce the error +Equally issues of stateful Filters has brought to attention that the breaks alot of Filter principalsI say we close these issues and begin moving what code we feel is salvagable to a spatial module where we can work closely with the Solr code +Merged changes from trunk to and +We can also increase the update interval to decrease traffic +Patch for this defect is available please help review it +Speaking of which I am in favour of creating a separate brnach for this objections +Distributed Cache gives inconsistent result if Archive files get deleted from the task tracker +suggests a permission issue of some sort and Id look at the logs for any logs from package classes +Henry you have to add the license header to the new files otherwise the release audit test will fail +Not sure we canneed to support cancelling after sendBatch was called because all the remote messages will already be on the bus and reach their corresponding endpointsId suggest to start with cancelBatch to cancel all enqueued calls and treat sendBatch like a commit but its up to you. +Is it still worth it to anyone? +Raising priority +This is the correct behavior if you are running on jvm orIf you are running on the major version should be +Please let me know if you have a fix forI have fixed the file +If you want more controlled inactivity checks you need a STOMP client with write checks which will detect broken sockets on write +See attachment +Rewriting according to the suggested this is new functionality I dont think it is appropriate for +The exception goes away +Deprecated in +Dmitry where this is fixed? There is still the nd case bug in +However it may be it is not working for everything +Looking more deeply at the word count example because they are doing the serialization in mapper and reducer they do a bad job of it +Moved back to issues are postponed to +This patch will remove the dependency to have title +could you attach the +Patch contains the change suggested by Andrzej +Any volunteers to look at this +I didnt spend too much time looking at security but FWIW the NFS proxy supports Kerberos in privacy mode +I wonder whether it would be possible to use a module like ctypes to connect to Apache functions + implements something like this +defer all X issues to +Having a bulk API simplifies the threading in AM and at the same time helps us to launch containers fasterbq +The commented tests now are enabled within fixing of +mallocs are not allocating an additional byte for +Throttling might simply entail spacing and scheduling when you scan the next block to complete within roughly two weeks +It turns out I forgot to attach the jars +Still broken fix for +An output file showing one execution of each flow highlighting the start time end time and duration experienced on our test environment +More will likely come as a I need them for testing +yeah it is the second possibility to solve that problem +Patch contains also some bug fixes +Server config server log and console output attachedPlease let me know if you need anything else +Instead one can extend to implement a custom that takes care of doing work onIt doesnt look like there are other such libraries having Apache compatible license +Simple fix but takes a long time to debug +Change with a selector with more preference ul could work +Can you close it +Now we create different version of Adapters znc +Updated patch based on review board feedback +I noticed you are working on this in the Velocity introspectorOne idea I had +Here is a patch to fix thisCould it be applied before releaseThanks +Fixed in rev +Im closing this +Also forgot to mention please update to use the defines here rather than duplicating them in both libhdfs and +Pascal pull request applied ready for and +Not a problem +After analysing other tests found that is not shutting down the cluster +Same patch with the correct Hadoop jar now is using the just committed this +Not that I dont want this one in but I havent looked at the patch yet +Cannot reproduce it on the current Harmony snapshot +Committed revision in +Perhaps I should have unassigned it +Milano could you provide more information on what youre trying to do that causes the exception +It looks like this was fixed by +SVN commit and +Ted Yu thats also a bug +revision +Suresh with resect to Change final public static to public static finalI would like to leave final public static for consistency and file a separate ticket to change all the uses of final public static to public static final +thanks Robert! I think DISI as basiscontract for is a good thing so we should not remove that dependency +Can be applied to almost the same +Likewise same function appear at Linux kernel +Thanks! I applied the patch in revision I also upgraded the POI dependency to the latest versionPS +Since the plugin exists as a separate github project I marked this issue as resolved +INFO Running for rows took ms +I already tried to apply the diff of review to the trunk of Shindig but it seems to no longer cleanly apply +Id like to volunteer and test it for Eclipse +The number of connections is not stable either it varies betweenIll rerun the tests on Tuesday with older versions of pooldbcp +pending failed case and audit warnings are not relative with this patch IMHO +To me seems like the best default here +How frequently is this event likely to be raised +Hi JoernI agree with you on case two +Reopening for further discussion and brainstorming. +As you said Im also not sure what might be a better name without being too verbose +It was maybe a temporary issue +Like I said its a bad list to useIm currently using and in addition to an internet security appliance very little spam gets throughAnyways thank you for your time +If you do hear from the IBM JRE team on any updates do keep us informedthanks AsankhaMy intention was never to write a patch to improve performance I simply wanted to make Synapse work for our customerNevertheless your TPS results puzzled me and so I also performed a few tests on my laptop +See usage of elsewhere it deals with this case sensitivity issue +Yeah Does give savings +Thanks +Thanks Sreekanth! +This is because by default the Oracle connection manager will chose the very driver you are trying to pass in +The Stomp php client was update to now use the correctly spelled version +What happened to the agreement on applying this to one or two modules in a feature branch and be able to see how it workswhat we are missingThe parent pom example on top is still pushing down the alpha qualifier this would not be aligned anymore with this approach so that should be removed or at least explicitly overridden by components +on idea and patch +Please put patch on review board it is of decent size +Hi AmosPlease can you delete the old build in CI? +metadata +Closed +It passes the regressions and Im currently installing this on my node cluster +I merged the fix into and ran a test and got a deadlock hang again +When the book is recreated from source the issue should be fixed +Do we concatenate their messages or take the highest one in the stackWhen an exception has this flag set the assumption is that it has the right level of information that user needs +I used the gwt debug mode cant open two windows at the same time +does have a create call? Yeah this should be at server side which will be used by bookie format logic +Hence we cannot check before allocation only when the ringmaster runs +Nikolay actually to add a new content under ecmdemo you need a to connect first to the platform with a form by going to the private part of the site ecmdemoprivateif you want to add content with basic authentication like you try to do you need to use the url httplocalhostprivatejcrrepositorycollaborationChuong Minh I tried to reproduce by trying to access to httplocalhostecmdemojcrrepositorysystemjcrsystemjcrversionStorageecacabaeabjcrfrozenNodejcrcontent but I did not get any exception could you please give us more info to help us to reproduce +How can this be done +Thanks for the comments +If we cant fix anything for just mark it as Not A Problem please +This change is a simple move and no tests are needed +Tabs crept in somehow +Closing resolvedwont fix issues. +I wonder if in Fedora differs from other distros +Agreed though currently there is only one feature! But this would be good to put in and ensure that it stays up to date as we release new versions +Assertion fix attached +Paul can you double check +So the work for fixing this issue is the same in either case and we can avoid adding oddly named methods to a deprecated class +The original format that we maintain should be easy for all to editIs there a reason not to use Forrest and include this in the site documents? Forrest documentation is written in XML but PDF is produced automatically along with HTML +when you create some jira issue attach stack trace in jira comment or in issue description not as a screenshotfirst it will be easy to find issuesecond it is more practical for developersscreenshot is necessary to describe what you did and what you got and what you expected +I really think this seq change is unrelated to this issue and unnecessary +keeping a separate set of levels for unrepaired data would be the way to goCould be although it may be a lot of extra io vsAnyone up to run some simulations +Also please review and comment re +Per Nigel Daleys request I should point out that this bug was discovered using Sierra from to view results +Hopefully is a single collation determination method is created it will fix this issue in fix in main codeline to fix the problem +We currently cannot include into our customer builds due to this bug +Removed all mention to grid in spatial in favor of quad Nicolass updates not resolving yet as Im having another pull coming. +It appears to be a bug in the JDK but it is not +Ive attached a patch which fixes the functions namespace issue +I think using entities will suffice even thoughthere are some minor drawbacks to it +Ive committed this with a few small changes to add documentation and license headers to the files. +Would you please check youre using the Sun JVM and confirm which Java youre using by executing the following commandreadlink e which javaThanks +Agreed I think this is a better approach as well +was reported as failed with the error Forked Java VM exited abnormally +Ive been testing and improving a Hadoop version of this patch extensively at Facebook +Not standard usage is enough to reject m double checking with the customer to get a better understanding of their use case +AFAICT it is not fixable +If it isnt meant to be a long coming in then why handle it +PigHive will give you that but are not real time +The ESB docs carry a similar disclaimer +Please if anyone still thinks there is work left on this. +The changes introduced a compile time dependency from to +Thanks for submitting +It should be tested manually +for the button +Ah i understand! Ive made changes to before adding classes and in this case removing one +Applied to trunk and branch +We still add elements to the Retransmitter though +IMO we should put it in +they are carried in RAM +Thanks JinhoThis issue is resolved due to. +Pleaes wait until the end of the week +Will open a separate bug for that +Thanks for the s fixed on trunk now +Hi Devinplease review and push this fixThanks SvanteReview pass and you can push this fixThanks verified by Devin I have uploaded this patch to the master repositoryClosing this issue +The issues are now closed. +heres what we need agent sends up inventory report we should not schedule any metric collections yet since we dont even have the resources imported yet the server gets the inventory report and populates them as NEW in ONRESOURCE and puts them in the portlet user clicks import to commit resources into inventory we send a message to the agent to say this set of resources have been imported you need to go grab their measurement schedules this can be sent asynchronously the agent gets the message and then turns around and sends a request to the server asking for the schedules the agent gets the schedules and populates the with the new schedules the needs to be told to now collect the data based on the new schedulesQuestionsa how do we identify the resources to the agent? can we send down the resource keys only? the parent resource key and the resource keyb can we combine and ? can the server just send down the schedules after the import button is pressed? it can do this back to greg hes gonna to finish this up and check in the final is complete and checked in +Andre +i plan to commit this shortly +Because of me taking the wrong map was filled and that has a major effect on the fix released in and till the point the jsr module my not really work anymore +Are there any environment settings needed like MYTESTIPMYTESTIP +for the patch +not working correctly functionality is now in as of +Hi Bauke! If those classes are in a jar which has a and qualify for a CDI bean then we have to scan them as per the CDI spec +Ill do some testing +Thanks for the +Arun as youve given the OK before and the changes where of form only Ive committed so Asokan can continue with the other +I am changing the component name for this JIRA and am going to open a new one for the HOD bug +I think something along the lines of convert to relative where possible is that even relatives can be converted into better relatives +Thx for the patch +HiteshThanks for working over the weekendPlease specify hbase in Groups field on review board in the future +I just committed this +I spent perhaps a half hour putting this together and most of that was testing to ensure it was backwards compatibleId generally vote we get that whole automated conversion bit sorted then structure the project around that since that is the proverbial tall pole in the tent +Can someone relate this issue? I dont think I have the rights to do so +Thanks for the look Bikas +I tested jbossws and jbossas testsuite with resolved issues. +Glad to see this committed! +I can now cleanly build SMX Kernel NMR and Features on AIX +should start services as active but appear in toSo that one could configure to report to both nodes +I mean Developer Guide. +But attached to this JIRA and the same jar in zipped distro doesnt contain and files +Thats right +bq +I do not understand why you are so negative about it +Check method was moved +I think the issue here is whether the path is absolute +The question would be is it worth for us take that much time on this issue +Thanks MarkFix applied to LUNI component at repo revision Please check that the patch was applied as you expected +Committed thanks +The question is now whether this is because of your switch to RELAY or because I fixed the FRAG related bug +I rather like the x false instead of x as well its more explicit readable +At the moment I dont know exactly how to do this +Lets simply have the with records as input +It works well just like in +Patch attachedFix required prior to release +The now supports JES Interface Level and The JUNIT test cases have been updated accordinglyHave fun in +duplicate of by +testcase for the input! +Ive started updating Scaffold for the latest release of Struts +Its so similar that I had to to see if Id actually applied the latest patch +This skips the Java language test in case the jdk is. +This patch also cover the situation when user presses when runtime exception occurs +Looks like this has been fixed with the new variable mechanism. +it is a partial solution since it does not move out the transform code +See and +can you validate this issue against? That is where it will get fixed if it still exists +All xsd files are now resolved locally +OK thanksWhen you go to your Solr Admin page today it lists all cores even if there are of them +please ignore it Thanks a lot +Please check that the patch was applied as you expected +Patch missing. +It is an unfortunate oversight on my part +Waiting for your response on HOD thread +This doesnt use runtime jsp compilation but does use a precompiled jsp +Otherwise same as the previous one +Unless there is a problem I am sure that nutch developers will put it inPS I actually have a plugin that uses Rome and I will work on rss index and rss query plugins once this issue is resolved +Well for interfaces the only fields are constants so I dont think it makes sense to use Override with them +Is our goal here to replace compaction entirelyThe manifest information consists minimally of the level information and ranges +Committed to trunk. +and i have test resin on my windows environment +Corrects a bug on the previous version when using a value greater than as parameter +So from the start of the request until the end of onClick for example and from the end of onClick to the end of request +Patch looks good. +The patches went of of sync +I am assuming the absence of the file is normal condition for this testYes +Tracy and I had a chat on IRCI think he misunderstood the ticket and made it overcomplicatedBasically what it is a delayed recipient listeg you have a message at hand and what you want to do is send this message to X endpoint in Y time from nowIts a message so it should not repeat or do anything like thatA use case could be that you process a message and in min from now you want something to happenSo you can send a message to a timer with min with some payload and tell it to send it to a route which then routes that message min from nowDoes it make sense +This will eliminate the need for socket communication between the java and C code when it is on the same I have updated the description as we no longer support +Im guessing its rare so its probably not worth it but as already stated Im not really a Thrift guylarry +Im more think this is caused by some of our db server this as wont fix there is nothing we can do against the hibernate code +This is one reason they have not been updated to reflect the latest versions of the respective containers +This change introduces the necessary upgrade scripts for upgrading metastore schema for Hive to Hive +This is because plugin recalculate dependencies on restart and sets updated classpath containers for projects that have maven nature enabled +This is how procrun worksIf the main returns it will kill the JVM +This solution does not work for installation where the NN needs to support more x scaling +Search for to see which have this problem +Finished review of the new hudson views and shifted around various builds +I think its +I verified the attached patch works against real cluster +All the regression tests ran cleanly +Added sqlite +That is a bit sadHow likely do you think that the JDK version to be will it happen within this quarterI guess i could try to write a patch that is compatible with java but that would require me to do my own parsing of the format that java stores characters in memory so i would really like to avoid having that code in a library +Ok I am working on it I will take the freedom while fixing this to perform a small refactoring which is neededto keep the code maintainable in this part +Is there any workaround for this? Im being bitten by this one too +The is also affected +I prefer separate folders for eclipse output and maven target +Otherwise the warning wont show in other modes. +Thoughts +I feel that the Apache policy is unfortunate but I understand the rationale +the class defines a constructor with a as parameter and a method called getPrev +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content is only validationverification issue +the problem has been resolved with the updated JRE thanks! +Are you sure this is tied to the views location +do we have other sun in our docs that would need updating after acquisition +Yes it is installedWhen I define PERLLIBRARY and PERLINCLUDEPATH explicitly as described in the linked issue cmake finds perl and build can proceed +heheh i meant it IS tricky +Im updating the fix version for this one to +patch looks good +Tried it and it indeed works +Ok now Im completely confusedThe original bug asked for a piece of code which isnt needed anymore and now the description changed on meWe should have opened a separate bug IAC LDLIBRARYPATH is already setup correctly via which defaults to right path for native libs +QE recommends this is deferred to CP as previously discussed +Committed trunk r rAlso committed nightly changes in . +I created also a branch lucene based on the current trunk so that others may join inEdit note that this is NOT a working implementation yet its missing many key pieces +After a long struggle finally the ops struct is either removed or statically allocated. +Per thread profile is printed after merged profile +PR sent with orientation and landscape mode tests +I just committed this +I am not quite sure aboutremoving that reserve code +This is now fixed in +I recommend we turn this into a task for cleaning up the message files +The first we refers to the authors of the design doc the nd we refers to the people working on +Thanks for checking for me. +Greatness! Thanks Snjeza +Is there a workaround? How are nulls handled? +Now name of component is checked before removal. +Letting it match the number of input blocks is a much better idea +Fixed this by adding node ordering to now get the expected output. +it could be +Ill commit this +And we probably should for Maps since it is not appropriate to ever have a map with a key hashCode and equals are not compatible between implementations +Thanks for the contribution Lefty! +Nonetheless the referenced forum entry shows that child elements in the default namespace are incorrectly losing their namespaces +I think I have just fixed it +Well done +Breaking things down more than this could lead to confusion among users as to which component to place something in andor increased administrationPS There are other mechanimsms in Jira called project categories so we should refer to the above as components to keep the terminology uniform +I reverted my change to the test because of that +Yes that sounds like a fine idea +Technically this feature was added to Java in however the implementation is easy and it doesnt break any existing java apps so I think we should include it +a test case to allow us to replay this easily? +Although see for for collection batches +Patch replaces patch which had an unnecessary change of file properties +If you need help using Hibernate go to the user forum +Dynamic meaning that the set of possible source can change at runtime which is possible with the current logic although its not entirely implemented for that as Im highlighting with the prepared plans issue +Marvin do you have an explanation for thisImho coreinit distributabletrue do the same does not work without the patch +Tamas Ill look at this +Perhaps a page on the WIKI and when mature move to CMS +Graham will work on this +can be done in a separate jira and wont break anything alone +If you restore the timer with with one of the programmed schedule after the current time the timer restores perfectly +Should I make a pull request in github +available on r. +Manageable I think +Thanks for reporting +This would postpone adding Bipartite Matching for some time should I create another issue for it +Ill review the patch and possibly check it into a sandbox as the starting pointThanks made a mistake by importing the code into Tuscany SVN without going through the ASF process of accepting external code donations +This is because the test sources are embedded in Branch +I didnt see the exception that you have mentioned +Bug has been marked as a duplicate of this bug +Do you know what affects this has on javascript form validation? Will js still work with an id attribute instead of name +And I dont think that either attack is likely or worth worrying aboutMore seriously discusses race conditions which could arise when more than one user tries to boot the database by connecting to it +Yeah sorry has been for some time +This is a known issue +Hopefully youll get a helpful error message or stack trace at the end +I think the intent of the code is to make sure we dont have an old reader after we update the dictionary but if were creating the dictionary for the first time we get a NPEThe simple fix is to check if reader is null before trying to close it +No IT +This one works +Added test is the corrected patch +I started on OJEC +Colleagues stated that the migration from RAD to RAD is much more painless than from to and so we will give that a try +Could that work for your use caseSounds like it thats greatDo you think there any efficiencies to be gained on the document retrieve side of things if you know that the documents commonly being retrieved are physically nearby +In addtion to the patch files the following needs to be modified contians new files to add contains the patched files instructions for renamesmovesdeletes of existing files +Thanks to Siddarth Seth for the analysis on this issue +I am in the process of evaluating this fix +Like the new class lets put a javadoc pointer to in the javadocWhy use less instead of compare and why does the epilogue here differ from upstream guavaOtherwise looks great +Thanks Uwe! +I optimized it once again in revision +If the system property is set to then the socket transport from Remoting can talk to the socket transport from Remoting +Committed revision feel free to comment +Bumping up Priority and scheduled accordingly +Hmm +The reason I didnt is that the tracing is is optional Ill create another patch see what that looks like otherwise Ill commit what I have +First warning is not an issue apparently this version hibernate does not implement the EJB spec correctly +Also it seems quite a bit faster than which is always very you +I suppose the question here is whether there is an intention to support this interface +Image that work is well underway with AS all previous community releases are +Thechange I made is in the latest SVN and will be in +if you generate everything togehter they get linked +Im working today on a patch and if its quick maybe we can squeeze it in +This is a simple resolution for this issue +Its been there forever +Attaching patch to only open sstableScanner per level during repair +This version adjusts the detailed Issue notes adds a note for and improves the summary line for the Interrupt Resilience feature +I have attached my configuration files +I am looking forward to look at your solution +Hi Daisywhen your team has reviewed and integrated the patches could you take a look if this issue still applies to the new Table APIThanks as resolved +Stay tuned + +The base ejb install group has not clustering services. +committed to both trunk and branch. +to close this issue + the simpler the betterI created for this reason +Turning off schema validation may fix it for you if that is the case. +Closing this since has been released +I dont have the necessary jira permissions to resolve this issue +Worth playing with +Starting with a simple one for patched +I think the problem doesnt surface until you get a sufficiently large set of classes like an enterprise app might beThe timings for the sample are as follows unpacked s packed s unpacked m s packed s unpacked m s packed s +Thank you Dmytro. +The ESB branch has been updated +Until we have a critical mass of developers using this functionality I would not want to add this to the core struts codebase. +My last week was super busy + +I assume the patch has tested successfully with all compilers on both Linux and Windows +Should be OK now +But we shouldnt be disabling the test we should be fixing it +OK will see next year +When the send timeout occurs you are not sure whether the message was indeed sent so an exception might be more appropriate since while it is possible for the message to eventually get acked by the broker its also possible the connection could get broken or the broker could go down who knows +The databasemetadata has been done to support creating a Teiid model via jdbc importer but executing a query still has to be hooked up +Thus wed need a testcase to see what is different in your case +Patch Committed revision am going to spend more time on your suggestion but from a quick look I dont see any recompilation happening for triggers currently +Bugs tested +Hi Michael At the moment were only starting to use too but the need is real +Calling toString on proxies is always tricky as they are mostly implemented wronglyThe toString of can be made more robustNote that this is only because of a statement in +Thanks Ning +Hello AllWe are implementing a search engine based on LuceneNutch and we are also facing problems with caching +So we shouldnt view this setting as a quota at allI agree with the other Jacobs comment regarding the name of the config settingIf you guys think using specific types of exceptions is more preferable sure go ahead +I also ran dblook on a database which contained functions and procedures with parameters that had just parameter types specified +Ive also noticed that these dont report a failure when using the latest surefire which seems a more critical tests pass on my Windows XP box +Thanks Armin for pointing out the dup. +Work around based on heap size increasing causes time of Full GC and the node is not functionalCall is good for MB fsimage but not the oneEven without a explicit the MB one can be cleaned after period checkingBut I think x memory requirement is not reasonableThe rootDir referenced a huge tree in memory JVM must find the exactly object and cleans them in order +The ID of account is brought to the sales forecast screenIt also includes a modification of the that ensures that the id of the account is set and that the default internal organization is set to CompanyIt also includes an addition to so that new sales forecasts can be created from the account screen of an account +If it is your code and you are in the position to relicense these files with an ASF license I kindly ask for a new submission with the appropriate license headerThanks JoergMy apologies my IDE puts those copyright notices in automatically +Anyway closing this issue as the tutorial will be rewritten to use a standalone HSQL DB +thanks. +Ok got this the fonts are defined by a artifact +I am assured this is a saving bug and not a loading +The doesnt support to be called suspend and consume at the same thread +I recall doing something similar way back +If I change my ini to have sm insteadof securityManager all is fine +I think that the same argument goes for session +Oops ignore those attachments +I didnt look closely enough at the full report as it does list all versions including the latest in the detail sections +This looks like a new policy +Ive also rebased the branch in my fork onto master so would be worth running stress test again to make sure we are just seeing the impact of the one commit +The fix is in. +Here is a quick explanation of this solution remote deployment tells the server to download a deployable file from a web server somewhere +Dispatcher xwork profiling +Doh ignore me thats an empty file +Will upload the patch +I dont know if a subtask is preferred +In future as we add other types of Collator objects we can modify get accordingly +This particular issue is an extreme example but we really need to remember that each addition to the lifecycle is an addition to the build vocabulary of Maven +Committed revision I added no CHANGES entry as we are all committers and no external persons involved +I submitted a patch to jaxen to fix that but I must have created the upload bundle before! + +Ill look some more tomorrow +trunk +next iterations +I dont personally like files without extensions very much as its even harder to know what they areDont you think we can expect our users to make the leap from cpeGui shell script or depending on the platform? I think users who dont manage to make that connection by themselves will have a very hard time with the rest of UIMA +Ive fixed them we must submit before we start checking and we need to have at least one snapshot complete for this to be interesting +Im curious what makes qdox require at least +It would be cool if this feature could be added I dont know if Scala has any better handling for extracting the necessary field information out of case classes that might help here +Adding a volume only creates a db entry +The agent is only required as a workaround when launching from WTP and this is something JBT can take care of when it launches the serverThe big advantage of the agent approach is that it will work with any app server +mntvolhive +Part of a bulk update all resolutions changed to done please review history to original resolution type +bulk defer of issues to +but i just double checked +I just thought you were interested in hearing about usability problems that I create for my own potential users just because I use Maven +I had built my project from command prompt with maven and tried to install in Fabric Server and Apache Tomcat +I agree +Using binhadoop namenode doesnt even start at all failing to resolveI didnt test it on Linux yet +Hope this would of some help +Ive checked in a fully untestedIm leaving this issue open until Ive had time to make some testsFull unit testing will require acheers Roland +Updated patch submitted +There is a minor typo Implementations that previously allowed for the previously lax restrictions may need to continue to support that for some use cases +no ones commented on this for a long while so Im going to resolve this issue +Cool! Looking forward to including that branch into trunk +In the meantime I would prefer committing this to be sure it makes as Im sure that not acquire references was bad +Try again this should be fixed now +Ill have a look and come back to you as soon as possible +I dont know why exactly +This works for me +Patch for resolving this testcase failure +Thanks JBDont miss to push the release buttonP +It should also attempt to remove itself from the session before passivation +Hi I get errors when I apply this patch probably because of you have the file +It would be good if someone can look at this patch and commit the code changes if they are okay +Thanks for helping guys! Im working on implementation which uses for buffering and pread native call for accessing file data +for have committed this. +However isnt the same true for TOPICMOVED and SUBSCRIPTIONFORCEDCLOSED? Or at least for TOPICMOVEDAlso regardless of this NOP is a bad name for the subscription event as NOP isnt a description of the event but rather a command to do nothing +New patch fixing the bug + +Feel free to pick it up +Ive rebased the patch and uploaded the third patch +Checked in asThe Windows code has a stub only should compile but will print as the timestamp +Thanks for the reviews Owen and Arun! +sha +IBM has searched their database for problems with nio on AS and cant find anything +The DTD says the type attribute is optional but you get an error if you leave it out +Its the users responsibility to use the right port numbers +Thanks LeonMy apologies for missing the call to update from the update service in my patch +Theres only so much time to work on issues that dont directly impact us +it would have been nicer to have the cleaning of a counter context not change its sizeCan we pad it somehow +If you file another issue note that is not a version +Anyway I am quite out of ideas on how to approach this issue either in creating a reproduction or fixing and would most appreciate advice on next steps +Im not sure I can remember enough about how we ran into it to confirm though +Im building from CVS +The correct symbol in Java code is which will be used for bitwise XOR instead of +Please close this issue if you are satisfied +Application can get false positives result in some cases +Thanks AkashYour patch is in rev +I think it is to their benefits to migrate as well +The two first points are important IMO. +updated patch to allow getset Connectors uses Solrj and should be able to force multipart to trunk and +that exposes the bug +on commit +Awesome nicely done! Ive asked Vijay to review this patch but Ill also take a look at it +I am raising priority of this issue +revised oddness patch according to suggestions +Instead of deleting the tests I added tests to check if targetmodel is instance of. +i analysed the status of the support case i need to see whetherhowwhy we generate this sequence of calls to the hibernate session +tsm ESB project that illustrates smooks capability of splitting csv files to xml fragments that are then routed to other service via error usually means that the content being passed to the parser is not XML can you double check what content is being parsed +Attached sources for the exploratory is definitively +Thanks RobertPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +If we can eliminate the volatile in even some circumstances it would seem worthwhile to do so +Marshalled values are used to deal with store custom classes in binary format +hi DougCan you upload the messaging configuration files of the two nodesThanksHoward + +I think this ticket is blocking some of the security work +Hi Benjamin Please send us an update on this issue whether you be able to resolve the issue +push to +Committed to derbysitetrunk revision Committed location of to infrastructure trunksitetrunkdocslicensesexports revision Minor info point bis Bureau of Industry and Security +Patch rev + thanks for the patch Liang. +Because you should be using the right version of in your +Will let you know when they are ready +Have checked fix into SVN +Perhaps a short guide around the and code +This is the patch with the indentation changes +Can this be closed as duplicate of +moving to resolved state QA please test and close accordingly + +I suspect that this has probably been fixed by the changes to which supports expressions in GROUP BYSince the original reporter of this issue has reported it fixed in I will close this issue as fixed in in a few days if there are no other comments +What is the rationale for pushing these configuration changes back to? If they are required they need a compatability story +End of from for this report Ive found the issue a bad for extracting the test run ID + +If it needs config again please reopen +The exception goes through the like it should and is handled by the code in that handler +made the change now +Thunk you! +We would need to wrap those policies and then remove the exchange from them inflight registry when the rejectedExecutor callback is invoked +The removes deprecated method to retrieve the session space +Just that not every build works out of the box +i hope i didnt miss any artifacts + +Wicket issue is only fix committed +Thanks everyone! +It is very good for dumping complex object graphs and you usually need to write lines of code +Closed again now properly recorded against Fix Version as well +I delete CLONE prefix all times too +No we can easily this one +is traced in I looked into the failed log the timeout is caused by a race condition luckily after committed the race condition no longer apply so there should be no problem now. +From that we can craft something more fancy to support the testsuite but right now I dont think the set of goals we have is the right solution +Its Java compatible +I was probably just imagining the impossibility of supporting all the possibilities of s while only requiring for Tools +r provides the fix for the Woden OM implemenation +Management server log we sure that this doesnt effect as well +good point + Actually this is our deliberate + +Just saw thatbq +Verified with revision +Approved for SOA CP +Bizarre + to trunk. +Thanks for the review +Yes it isChris +Tests pass +more Will need to change to reflect new pom specs +resulting an another corrupt blockCan you pl explain this scenario in greater detail? If the client encountered an error and exited then the write was never completed +Now when we get a excpetion while parsing this DN we generate a where the contains the error codeThis should be done for every places where the DN is parsed +Attached is already fixed on the trunk I just created an test case exponsing the bug on the +We discussed this over the IRC and since Daniel is not planning on active maintenance for the test suite the best course of action would be to start migrating things to +Thats exactly why I dont like the entry +If Im wrong please. +I think Ive answered my own question the wiki contains the description and the field correlationid will allow for this kind of implementation +Err +Ill take it out when I commit for you +Once we isolate the issue then we can look into new sorting strategies if they solve the problem but Im not sure we can say without fear of refutation that the recursion alone is causing this yet +Patch looks good +The problem can be reproduced by running I have attached a small patch to clean up the use of some old filter code +Setting fix version to for consideration there +Implemented for tested seems to work +If readChunk is for supporting only read then we could have readChunk to be optimized for sequential reading +The problem is the tool by default set the Gateway to complex and hence it messes up the whole thing +Hi Markushave a look at +Ill kick out of the trunk build till I can get that fixed. +If they could optionally set the keys then the code would need to support that but should be very easy +Fixed on trunk now +The properties in were messed up for the jdk requirement +Maybe we should only consider supporting a C extension version of the compact protocol for Python +We have a test case in for this +I will raise this query to the user mailing listRegardsVijay. +Hi DennisPurpose of my file was not to remove the timestamped version from war instead it configured the ear to use timestamped versions for the shared libsNevertheless I changed my code now by configuring the war plugin to use versions and now it works fine +I can see clustering Wikipedia articles with this +Fairly change +Built thrift and fb +AlexI did wrong when filling the initial form this is a bug not a feature request +Which JVM args are your using in waveinaboxTIA +Im satisfied with the fix therefore resolving this issue +Is there a reason why we werent doing this before +I am not sure this is really necessary as is a private inner class and as the javadoc states is only implemented so can be used + name of the attribute is misleading +As default accept header I would suggestapplicationxhtmlxml texthtmlq +No youre not being paranoid +or similar +This either needs to be completed in CR or shift to +I get the same NPE in the case where I have a service with no params +I dont put back these alarm clock but Ill finish class and schedules defined in project. +Will open new issues for anything found +Since started from extras I wondered why it isnt already there +The fixes involve a single fileThese patch files supersede the patch files Id previously attached to this issue +Just release a new snapshot release +Update patch to add test case. +This should be fixed I didnt check in a class reposjbossasbranchesSPCP is correct now +No its not a problem with loading +i work on a fix for the that takes care of headlessenvironments and accepts a resolution +About I was passing that so I could get to the configuration but I can easily change it so that loggers get the configuration object itself insteadI think that would be best +Offhand do any other dbmss have trouble with the exampleTake care Murray I have only tried mysql and its been no problem with example +Applied patch +Ill change the comments +New version resets rowIndex and uses to represent state +This will probably make my life harder +Please make sure you attach the files pasting them directly into the bug report can cause problems when lines are wrapped +I retract my earlier redirect suggestion +Make sure you attach the files please dont paste the content into the bug report +A better top object listing using the the newest version of +If we have a patch we can certainly a good improvement for +The remianing issues concerning incompatible jena versions should be tackled by another isse +this can easily OOM if the sstable has lots of out of order rowsIs that likely given that Range vs Bounds is basically an +trunk to remove Cleanable it should either be explicit or using a smooks specific lifecycle manager +No we havent been able to install EWP from rpm on rhel yet because of update issues +Good job I didnt press the button on Friday +No further comments and the writeup speaks for itself. +While I couldnt find any documentation of how populates its rule classes the impression I got was that any field could potentially be populated meaning that tracking state within a rule class would be problematic at best +to master as be. +Sorry one more questionCould you explain the error handling in the new scheme? The old code would call pad and that is now gone but I am not sure I understand how the error handling is handled now + +patch attached +Id not thought of that +Ashwin Thanks for pointing that out I just committed the patch of setting the default on userName and password +Thanks Sijie. +In my opinion should all mapping information be available at startup and not be resolved automatically during runtime +Ping! Given that has been out for quite some time and that is supposed to come out pretty soon is there any interest in reviving work on this JIRA? Has the criteria for acceptance changes or is Mahout still only wants to support a single version of Hadoop +This also effect EAP is a dupe of +I agree with Michelle that we should leave legacy project alone since we have no plans to ship them +Looking through the diffs now +Im guessing that its the wrong version but i need to check further +Parent and release pull requests contains commits for too +I fixed this bugI also found out I could make the test Repeat annotation refer to a constant for the iterations so I did that to make it easy to the testing temporarilyIll commit this shortly +Reopen or comment if still a problem. +From KabirThe AOP release has been uploaded to the repository but we need a POJO Cache release before head can the POJO Cache release is going through QA +Either we should entirely use system ones on all archs or let the bit library also be available on the tree +And at that time we can add capability to process commands like dfs count tmp +The bitsets being so small will just quickly reach saturation and force all reads to hit the underlying +Both patches applied as revision +Yes it does +Yes Tim +I think a response wrapper is passed on to the JSP engine so we could simply disable forwarding the call to get in the wrapper implementation +That makes sense +Just did some research and looks like these kinds of test failures were solved for MR in +I updated from trunk and using the new version of portlet plugin the problem does not occur + +Once I change it to anything even existing the warming Install folder does not exist or is not writable is displayed and I cannot get rid of itWhat does work correctly is that the triangle next to the field is only displayed when relevant +Its entirely not sonars fault. +Creating a new data Coverage data file homejdoecontinuumwork either does not exist or is not readable +have disabled this test if baseport is set in trunk and branches +I dont understand how to better explain this +you are right the closure classes do not get their constants cached +But still please check you have latest JDBC driver +If the value is a Boolean and true then the bridge clears the current scope at the beginning of an event render or resource request +Configuration is setup to look for new config knobs when old ones are used which is fine +But I think that the casts to in and should also go away. +Can you test the patch +Removed dead code in r. +sorry +Please fix in trunk and verify that revised OS feature when installed no longer throws NPE +Dima said what are not related but exactly the same +Jimmy This is similar to what goes wrong a pending delete creates havoc on the next create +In this case is quite useful using a prefix and reuse the bundle across application for example in Label components and is quite intuitive having bundles like theseBookElectronicsMusicAnyway as I said thats not a big deal or a serious problem o we can keep issue closed. +Thought on this more +Mike you set this as fix and but does not have +done because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I feel the fix is not right +only refer to a misconfigured case where someone didnt index their content correctlyThis shows you dont understand the issue +My first fix left two tests still failing so I committed again to fix it +Committed. +Please check the docs on thisDoes maybe have the option to see the source code of generated EJB proxies? That would also help as we would see how an EJB proxy triggers suspend and resume for the generated stubs! However those ties on the server delegate to an support in the package Transaction demarcation obviously happens there in the internals +Hi Rahulwhen authoring complex SCXML state machines I heavily use to source in the content of states and the datamodel +I meant you dont need to patch findbugs itself +Ill send this one in if there are no objections +Im confused about why this cool query parser Ive been using is experimental +Deploy it to an EAP instance and then in the Admin Console go to the Configuration tabs of one of the below Queue ResourcesnametutorialserviceQueuenamedefaultedtutorialserviceQueue Youll see the serverPeer CCE in the server log +OK the case of Embeddable entities is fixed +I will try committing it to branch after running the tests and others +I dont see any request for enhancement or bug report here just an explanation of the javabean spec +I dont think there is an https version of this repository +In my case killing the master did it +EditedI mean the SSL connection to origin hangs while the does not +It has found many bugs that I have fixedIn fact I think Ive finally got that test near +It contains a facet and ext +Just have to make sure it does not include spills +That got me too originally until I started testing out the variationsI had to test out the combinations of splash onoff status bar shown hidden launch images specified for standardiPhoneiPad and the orientations +Could you perhaps try this on a newer JVM and see if it is still a problemThanksKathey +If you have a small file or a fast link no error +Please feel free to reopen if you believe there is different details need to be covered on this issue +Committed to trunk thanks Vishal and everyone who pushed this through! +issue is missing Affects Version Component or both +Reopening to investigate fix +The strange thing is that code seems to be generated in about or seconds but the console is still open +It should be noted that my ant version is to branch with revision . +Lucene H +I think that this is a useful thing to communicate +Basic functionality is working correctly +However it filled the needs of the original author and was betterthan nothing +Awesome +Add all the relational operators +Please review the list of features provided by Marky prior to the final release +That being said if it did work I can just remove the last commit and we can roll with the changes +Moreover options file and files are confusing to the user now +A question which came up while discussing the feature To resolve IDREFs the parser will obviously create an internal Map of ID IDREFs +needed by theIt is not required to add H as a dependency +Committed with breaking early change +The build runs ok and is able to properly encode cascaded WFS features with datetime well need to run the WFS CITE tests before committing to svn +Applied branch and trunk. +Very simple patch that enables deployment optionally to nexus + +But Hudson passed contrib tests and managed to fail on core tests +Almost all other input splits are public I think including +I will let you know +I compiled the native code as mentioned above and ran the tests using it seemed to fail and i have fixed the same with the patch above +Good catch +Fixed in upstream +I ran all the ZKFC tests and committed this to the branch. +Daryn OKXie Would you get a chance soon to address Daryns comments +Done +The real problem I see is that it throws an from read +Things fixed with this patch +Fixed zipfiles for DB and DAAlso updated and to handle the directories. +Hello SylvainThanks for your interest in Trinidad +Gordon has applied the patch +Push to! issue may seem like ancient history but Im nonetheless looking forward to it being realized +Attachment has been added with description Contains the location changes from the previous patch and the new code +Attaching the workaround patch + +In the case where the usergroup does not exist so we provided a workaround in +Scott is preparing a patch for the patch for +Issue is specifically about daylight not being declared on +Plus if we get it fixed sooner theres no reason you couldnt use a nightly build +In case of straight Spring there are of them and the is among them and this handler handles the application has only adapters and is NOT among them that causes the cryptic message Does your handler implement a supported interface like ControllerHope that will help fixing the bug +LucieI think your project may be corrupted +all instance domain objects contain a lastExecution rather than a status or even restart data +Stax was removed from AS to be compliant with TCK in We have plans to upgrade to in EAP in The Stax versions across all components will be synced in EAP when upgrading to the release of EAP pending all relevant JIRA issues which are open are being reviewed +We need to catch the guy who wrote this test +site of so theres nothing else to be done here once its updated in the JBDS target platform +This issue is more than a year old and does not have any recent developer activity +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Better late than never +start and updateAck are called from run in the thread +It is up to youThanks Ron +The same patch but with granted ASF license. +All Apache build servers are offlinePlease run test suite in all branchesPlease list the failing tests is any individually so that people can help verify +One way that an edit log can become corrupt is by ending with an edit log operation that is cut off in the middle +Have you looked at the +btw the applies to and the Revision you very much works for me +Resolving. +Thanks. +changed two files and am closing the issue +Thanks for you patience while we sorted out the admin for your contribution +Unfortunately it fails some assertions which causes the entire test to fail +close issues for M. +I dont see why it would matter which is used however the important thing is that we express the name of the header file required in the build so we dont get burnt by this +Will post an ammendedment in a bit +Ill commit this version momentarily +Verified thanks +Looks interesting but we cant act on this without a current patch. +Ive removed the method from the trunk code. +Marking for backport at apurtell request +Tried this on an Ubuntu machine and had the same problems with Chrome Firefox +As I mentioned we shouldnt do thisIm happy to have the configure script ask the user his option for the scheduler +So how about clearly documenting that using wildcards is slow and going for the combined way of first performing a hash lookup or two and then checking expressions? If there are no wildcard patterns for CLASS andor METHOD this should be as fast as before but if I need this feature it would be available although slowerRegarding implementing this as a contribution do you have any pointers where to put thisCheers Wolfgang did you started working on this? Or do you plan to in the near future? I am also considering working on this since I also believe its a quite common use case +patch containing the testcase added to you for the test Tommaso +I ran into this issue and it turned out to be that was being set to +updated lets get numbers before saying L is bad +Needs as a dependency +Some javascript error message from the console or a screenshot would probably have helped I cant reproduce this +We will do a release and I added this fix to it in rev . +So minimally it should use the latest +Fixed the exception caused by modification ofcollection within for each loop +It would be good to get some feedback once I have a fix for can you test the latest nightly snapshot +If I run maven on the command line with U it does not download the either! Switching to mvn and invoking the exact same command with the same repo and the same settings just knows whats happening? This is bad since it seems I cant use a snapshot plugin with is by design the problem is the reference to the old version of the site plugin which will be updated in due course +I am not a networking expert so please correct me if I am wrong about this +Spec ticket updated +will try out patch on my test cluster +No +Added you as a watcher Andrea not sure if you got the latest notification after the issue was reassigned + +For we can do separately +Thanks for pointing me to that jira +Alexey please verify +Marks asked me to assign this to Tom Cunningham for help +For other cases hashing would actually introduce a significant overhead +Or if anonymous requests were allowed because you really dont care about who made the request or that information is not availableAnyhow I think if we force users to input the from party then that should be done when creating the request so the error in the overview would be avoidedWhat do you think +sorry meant to be should have been implemented in trunk some time ago +This way we could set the setting file we want to use +Hi Daryn it is probably late to change now but whats the reason not to specify pingconnectionContextsasl nagotiationsasl encryption in why reuse callId instead +Doing it shortly +A simple Hibernate runtime configuration setting can cause the application to fail in unexpected ways +I dont know what your test method is but a difference of is very unlikely +Fixed in version +It may be that the correctresponse is applicationxml typesvg charset or applicationxml typesvg charset Im investigating now +I have already created a new issue for this +Updates fixes doc and more tests to followPinaki I answered your questions several days ago but the comment didnt get make it into the JIRA for whatever reason +Im seeing only one failure in the CDH integration tests but I believe its not related to this patch and we should fix it in +That said posting patches against trunk and is certainly appreciated as some people like to live on the edge a bit and pull these things into their clusters +Yes that looks good +If this isnt adequate for your needs please reopen and explain why +Thanks +Cache the users its already a good improvement for a very minimal price +If we want to change the exception type or message its instead of locations to change +Reading mappings is crucial and we have been improving this over other data stores as well +A revised patch is attachedWe would like to commit this patch to trunkPlease provide your commentsThanks +The change log token is somehow encoded in the next link but a client cannot decode it +Can you please try to install them along with the other nbmsIm just guessing here as it works all the time on my side +thanks +Patch for +I prefer to generate new uid when we duplicate A +Committed to trunk. +I am running with the latest snapshot from Jenkins +Patch applied to the branch +The output of make check +Generally tiles and work together pretty wellregardsMartin +Hi Lewisbq +Thanks +Added additional tests where the partialTriggers attribute is on a documentation demo and Completed At revision on trunksrcmainresourcescomponentstrinidadcore srcmainresourcescomponentstrinidadcoreincludes trinidaddemosrcmainwebappdemos srcmainjavaorgapachemyfacestrinidadinternalcontext srcmainxrtsorgapachemyfacestrinidadinternalresource +Pull request made +The only workaround is to catch the exception and parse the message to distinguish it from other exceptions +DidntDont we have something along these lines? See the discussion in +As long its also easy to handle an war in Tomcat then I think its fine too +according testing of now committed +Whyhow is this related to +case that fails behind a proxy +Osoby is pluralb Project is spelled Projekt in Polish +The failed core test case seems to be caused by +If this is expected behavior then this is not a bug +but unless someone is clamoring for it i say we help our users avoid classpath hell just as much as we want to +Attached an updated version of the patch with the three modifications from my previous comment + +This is conflicting to the table schema +This looks like it has been fixed by the changes to the typechecking invocation order +Hi Paolo thanks for taking a look at this! See for background +changed fix version to unknown because this problem wasnt in the last release so dont want this item to show up in release notes. +I also tried the same patch on the trunk and it works too as there were no changes so far since the last change in I am unfamiliar with the source code and in particular with this class so please review closely the code and unit tests +r +Have you tried using a different logger or playing the usual games with WAS class loaders? Or if you think the problem is in Batch can you create a test case that demonstrates it +Install a redirect from to the new make sure that the POM points to Apache +Set Release Notes status to Needs More Info +Thanks in advance and bug use the different way to solve the problem and I think the attributes parent element should be kept there when setValuegetValue so I use the bug patch solutionbug only delete all the setValuegetValue method of attribute class no other need to merge +As per Michael Bs email Ill start on integrating deletes and aka along the way the deadlock issue could be uncovered which test is it occurring on +Should the problem be occurring with you with Roller please ask on the user list andor submit a bug report +Patch file containing fix described applied and tested on Websphere new jar uploaded to bluesunrise remote maven repo +I think we should reuse the existing exception if we can +Willem applied patch to trunk And I refined the patch a bit +But its only used in internally and i have changed all the logic +It is good idea ToodArun Can we reopen this issue for providing the patch +Ill have a look at the logic on master and see if the autoload lock remains in place +was to add a into the manifest +Thanks Robert rebased the patch +Unassigning myself as I wont have time to work with this issue immediately +Fixed on trunk by patch by Ernie by Chuck +Separate issue will file a ticket. +Its unlikely we would rewrite Spring Security itself to use MVC interceptors or provide a parallel implementation +Should the dfsadmin CLI be updated to use the new api +Dependency management is be able to run each class in dependency management will no longer run +Close after release +I did all kinds of regression tests and everything works will be released as. +I saw that at least one other Apache projectJust because someone else does it wrong +It seems this is not a problem can be ignored. +correct +well couldnt it just do the same thing as the other tags does? But I agree it is a very minor issue. +I agree that this is better in core +we now use all issues which got resolved in M +From the unit tests the user query cache hint is working as you provide more on the exact queries being issued +could also be the holder of a centralized list of available fonts +I also think the capability should be in place and tested before anyone ever needs it +Thanks Ben! Ill try to check things out over the weekend expecting some things to be +You need to save and then activate the Consumer +It worked perfectly with state saving set to serverAs you know the usual purpose of replication is for failover +Closing all resolvedfixed issues of already released versions of Roller. +Thats for clearifing I verified that CPFP is using SP +We adopted this approach to avoid any changes to query planner for vectorizationbq +Im in favour of this in general but not in favour of holding therelease up to do it now given that the example seems to show quite a reasonablesolutionJustin If this is satisfactory could you please close the bugIf you have any concerns or suggestions for better solutions then please bringthem up here or on the mailing list ASAP to get them in beforeThanks as there has been no further comments I will close this item. +Refactored the Transport layer a bit creation of a transport is now done by the class +Attaching latest one +Hmm my problem with this is how do i get access to an environment to test on +The patch also failed to correctly detect unchanged protocol files so I fixed that too and fixed the test to remove its files before running so that it can be run twice without ant clean between. + +You can always submit a patch +I guess I am still unsure as to why this exception isnt being thrown all the time on this code but only when during a paste +Craigs solution is fine except when using repository is stored in the folder +patch applied on trunk +It make take awhile +Scheduled to so it should get picked up sooner +The patch contains the interceptor the corresponding test case and the update to the user guide. +Call it from UI +cayennePlease confirm whether this made any difference +This patch includes Peters new test cases and some fixes to and to correctly handle the case when theres no tables at all in +Since this issue is marked Resolved shouldnt the Fix Versions field have been updated to EAP +Verified on EAP ER +Nice +bq +I wanted to use the RSs but pulling it down there is quite messy +I believe the SQL scripts under nist should not have the Apache Licence header since they are actually under the public domain +Darran do you remember which testcase ran into this issueEdit Didnt refresh the page before posting and hence didnt notice Darrans previous reply +Resubmitting for QA +This problem no longer exists +Track for progress on this issue +Also how do I get the command line trick working when the agent should be running as a service in Windows? Something else that needs obvious docs +Issue is Test is added to verify the to master +Adds another unit test for resizing the cache +Ok my fault +You could even use this to automate the testing a bit +CXF is very effective in that it will store big attachments to the disk temp storage if needed +Ive updated Struts scenario automation scripts according to latest BTI requirementsTested on Windows ia and Linux ia +Might not make +You could copy the region and operate it in another instance of hbase +Then we can look at how to facilitate it in the parent pom +Also add XS logs in future +Basically what happens is the framework is calling stop on the scr bundle which at the same time is trying to register a service +Fixed as requested since I received no feedback on my question +Any thoughts on that +At least test executed for the minutes in the default mode and for minutes in the opt mode with CPU usageUse you release build? I use HDK +were backported to as well adding the latter fix version. +Are you running into a different error or is the script running OK +Please comment if there is additional work to be done on this issue +The debug output is the same like reported in see the error now I had been looking in the wrong places +Verified with Seam branch +Id expect those exceptions bubble all the way up to the container in which case they should be visible +Seems like we get the old attachments when selecting the one with the date from april +The simplest way which worked fine is to omit the encoding statement +Seems like this changed M I tested back thenPatch looks safe to me since the only functional change it has is to catch a and wrap it into an +Ive limited modifications to the existing framework as much as possibleHowever some changes where necessary +Also yes this has to be backward compatible but that should be fairly easy we already have a version in the header +Certainly! there should be nothing till something entered +You can use the original Java sources to compare +looks good +The threading issues you addressed in your patch have not been addressed in my commit yet +Any chance of getting diffs here instead of full java files +The client did not see the issue as the string conversion of is not adding the in front of it +Simply changing the external context to an unwrapped version is not sufficientThere are a number of approaches we can take to this issue and Im going to ping the Dev list for to see if people have a preference +sorry looks like the original bundle must have been bad the javacc jars are missing +Thanks that clears up several questions I had +I think i can improve patch as we told earlier +Any idea on when this might be fixed +This is a blocking job of matrix +Normaly collection of map or collection should works but they are not map of collection seems to worksources explains the code modifications +What if we turned it into a real query cache? That way we dont have to predefine filters in the schema we just use the objects from the queries involved +why not use JSP? Its way easier to work with. +Marking closed. +Integrated to TRUNKThanks for the patch JesseThanks for the review Gary +This observation is in line with the general observation that gzip compresses better but performs worse +I think we are close on this but it failed the test +Test failure is a known issue on trunk +Tested OK +Verified by Leo +defer all X issues to +To fix this we can deploy several bundles as a karaf feature so they dont show up in the geronimo dependency graph +In some moment of time the disk was full and activemq failed to dump sync db with disk +I think this is likely fixed but need to verify +Ha +Silver circle logo is very nice +Thanks Tsuyoshi +Simple patch +Not a Seam issue please pursue on Hibernate to a Documentation Request and attaching the documentation I previously promised to write +trunk patch with Teds review comments +attach a draft version of hedwig console client +Members of the unix group should be pulled from group created + +I was wrong the current post in trunk does not contain its dependencies +Thanks crocket for reporting +Patch to invoke termination handler +pull lastest code in branch and can not reproduce Mice cannot reproduce it I will mark it Cannot Reproduce +Manually with integration test. +Its just that it issues only a warning now if it does not find the schema +Forrest insists on removing our internal anchors and adding its own anchors using generated ids +You can reopen as a feature request but I think it is sufficient that you can change the templates yourself. +The problem is that coberturas instrument task implementation fetches the classpath for the command line call from the and not the which in my opinion would only include the classes defined in +The Ip addresses are not getting released only for created from ISOs +It worked fine +Thanks a lotBest regards by Richard. +Unnecessary as we are probably going to stick with the static height layout for the moment +Wouldnt it be easier to just configure the logging level for commons vfs to be Error or Warning in the logging framework you are usingIn thinking about it this message seems totally appropriate to be an info level message +Patch adding analyzer just committed this +I think you are correct an alternative to is the main issue +The new patch now applies to the current also contains some minor import and comment cleanups +Today I realized that the problem also appears under JSF RI +you combine the two patches into one? suggested I have combined the two patches into one +Updated to use an page instead of a dialog increasing the available space +when you pass a null stream uses the name to identify the file +Yes thats right +perftarget +Again for sequential algorithms this is pretty easy to deal with but for parallel ones it really might be better to store the original data and pass in a function that handles the vectorization on the fly +This one needs to be verified against it +patch to fix the +Sorry I misspoke +I attach a temporary fix but we only need optimization part of compilePp +The findbugs warning was introduced by another patch and isnt relevant +Setting connection timeout at theHTTP method level does not make sense as the method may end up reusing analready open connectionLet me know what you think good to me +Failures are meant to be heres what the user did wrong and how to fix it errors are something went wrong that we didnt anticipate and heres the trace +Ill work on the other one. +This is the patch for removing ebay into order and product applicationIm not sure that the service set has to be moved from order to eBay but the probably the request has to remain into order but I have moved to eBayThanks good MarcoId say you can commit this work +Tom C to investigate +Problem is now fixed inProblem seems to happen on the command line with groovy so isnt just a issue +No problem man! Im just happy youre doing the work on the reorg instead of me +This is with Sonar now too +Waiting +Hi UweI used +However turning it on will allow element references to be considered for unwrapping. +The page compiler guarantees that JSP page implementation objects will invoke this method on all tag handlers but there may be multiple invocations on do and do in between +Id say test needs to berewritten to ensure proper servlet endpoint initialization ordering +Im not aware of any more hostnames Resolving. +This patch removed the unnecessary import in +thanks Uwe +Ill keep digging but this is unsurprisingly annoying +That sounds like a bug to me +Maybe I am doing something wrong +I do understand Spring cant solve any case and Im very aware of the code change is in a marginal part of the Spring code it seems that the feature is not much in use but it is a nice feature in a scenario and the possibility to do make the whole idea very useful +The whole URL for the system info is absolute and hardcoded not only the port +for Clob type as using in the JDBC API it will reduce any confusion for access at the same time it will have all the streaming capabilities +Thats not saying there is not a bug here but thats the intention +Steps to reproduce Deploy Metamer and open httplocalhostmetamerfacescomponentsrich Follow the instructions in there. +Closing. +CBEAjdkbin Cbin C classpath Cboot agentlibjdwptransportdtsocketsuspendyaddresslocalhost true B s CWINNT +Defer all unresolved issues from to +another possibility is the dirs are in a cycle +works in DR closing +lgtm +I didnt want to pollute Cargo with Jetty dependencies +Here is a simple reproduce the problemunzip targettargettargetmaven Dgoalbuild multiproject for a possible for for me now too +Ive hit this one as well +Can you help outbuilding on me a pvt email and I will send you the binaries somehow +We can certainly regroup them into independent modules +I also understand that this may be outside of Springs scope and maybe more appropriate in a separate tag library. +No locking at all nor a concurrent collection might introduce contains changes in the locking of the reload data to avoid mainly a potential bottle neck when rules are modified you retest your issue with those patches please +Even if preload size is if the bucket has entries it would load them all +Correct +New projects are now created without any errors or warnings +PSeverity and DescriptionPathResourceLocationCreation Time entry will not be exported or published +Ill commit this shortly +Ive upload a patch for it +Committed to trunk +A final update has to take place when the cache related code is stable +Attachment has been added with description Log of build for the quick response Jeremias +recycling hudson to check change means that the configuration that the map context uses does have the SKIPRECORDS boolean parameter set correctly +So in case people dynamic change it they are aware of this fact +The problem occurs in bothThis is a MUCH larger problem than I originally thought +Thanks for the fix JB. +Fixed in rev +IIRC it depends on language detection +Need to resolve before release +At least to my eyes there was a consensus on this patch and to the relation between the deprecation discussion and the patch +Patch looks good +Please review the issues and create change requests if you feel the issues are valid otherwise resolve them as assume classpath means that I can refer to schemas inside of the exploded or zipped +Not sure if this is compatible though +Waiting till clears the blocks before applying to +this also has an impact on embeddedjsp and osgi plugins too +No problems with SSL connections +FAQ +This patch fixed when any type of server restarted and transition into server +Do you want to do that Chris +Updated the spec with a new example which also shows the feature used with a table function +As this was fixed well before code freeze for Beta Ive set it back to Alpha +Andy I think this issue can be closed right +I agree with Kurt nil makes more sense than zero in this case +I think the comparison notion is built into the range concept +The only difference between the two requests is that the first one contains gzip which should cause no errors +The real reason why it works is that the destinations attribute must be defined +that come later should also be returned no task till the job that is initialized moves to running queue after is done +I had to bump the version to in components to get the test in to compile +Hi Diesthis might be a small shell issue depending on what shell you are using it seems that the first line in deltacloudd might be missing a space after the ! IIRC there are shells who do not like that +Marking this bug for patch with a test case which reproduces the bug and verifies the fixI plan to commit this shortly + applies to the applies to the branch the patches and they look ok +means a subset of all existing +With a test case we could verify the expected behavior against trunk +I was using but Ill back that out +sneaky +because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I feel this refactoring change we can do it as part of another JiraExcept this comment I have made other changes to latest patch +Same sample app in +That method applies to temporary resources only and the scope of that interface is broader than that +I want to spend more time investigating this to see whether the classifier is consistenly appended to the existing one on problem would be because it is being replaced +This is due to the two health checkers running for two topics +After a conversation with Ben Ill try and get to this before R goes out +This format is accepted by Maven +Ill commit this soonish and get to the others at some point when I have more time +its value is the goal you want to run to perform compile mycustomplugin +I want to commit this +It is not necessary to push all discussion to the meetup at the summit the time we will have there for discussion is finite anywayYour questions are welcome on the authorization approach +No I cant start server from the comand line +Needs more info but Ill add it to the backlog. +Attachment has been added with description The block alignment problem with +This patch is for create screen form and services for theRegardsParimal is updated patch for create screen form and services of +eh +looks like the issue is with not correctly stopping the miniyarncluster its using +Yes that is clear thank you. +for the patch + +Could you build me a jar and email it too me please +The above should be Java code +Since the minimum unit of throttling is still a row do we win anything with the extra complexity of pushing it down into? Why not just leave it in loop w adaptive checking based on row size as above +Yeah you won! bq +Its not required to wrap it in a bundle +with ASF grant +My pleasure to lend a hand +That way the set of XHTML docs without netuiidScopeor the tree attributes should at least be able to parsePlan to fix this issue post +Patch for for looks good to me +Committed the patch to trunk +It will be easier for developer to locate dependency libraries across different Apache projects by using the directory structure layout in this chain of jiras +With the help of Andreas I managed run the tests of rightly +I did much of the RTL work and while it is far from perfect I think labels work fairly well for Hebrew +The question still remains why unit tests did not catch this +Andrew Ive marked it against the latest releases of all the branchesShall I create JIRA release versions and for this patchAlso who is still running that can verify the fix took +Added a new unit test and refreshed to the latest trunk +Ive verified mvn clean install +Closing all tickets +As far as I understand this is the same as the existing implementation if you used a that always returned +Ok +Hi Devaraj If this affects as well does it need to be commtted into that branch as well +However my experience with is that I need to manually create an index on this owned foreign key column +The behavior should be the same for all the examples that you mentioned +The only reason I said dropped was just to make it look like tpstats +Corrected link to current coverage report for +Patch against rev applied +bulk close of resolved issues as part of JIRA housekeeping for release. +I think this has improved a lot since this issue was raised +Committed patch to documentation trunk at revision +Miro he had a bad configuration on +OK so I implemented a version check in first we read the length then the version +This needs a more general solution +Fixed with revision +I think the rule of thumb is not to let tests to step on each other feet in terms of temp +I just committed this + +Ok attach the changes in diff format against trunk and I can take a look +You can work around it in some cases depending on the glyphs you are using in the Font +I agree with the principle of generate strictly accept liberally with respect to schema names + has ideas about how to improve the downloads as he already did it for Studio +Test code must also be generified +Thanks for the minimal link too BTW +r and r to remove exclusion of aspx document acls xml at the code its looking for item not Item +Has anyone taken this on? If not its something Id like to try +Actually that depends on how you treat the XML data document oriented or data oriented +Looks like Nicholas and I were doing the same thingTodd this reminds me the fun when we were doing andEric thanks for joining us +Appended processing onsubmit function before call for a poll request +This is definitely a blocker! dims +doesnt handle dependent native libraries since it cannot natively intercept library loads +Assigning to Rajesh for confirmation +Hi jacques againFor relative path using eclipse you can check ignore leading path name segment to overcome the issueBest regards +No it didnt workI think it is because nightlybuildrepo and subdirectories doesnt exist +Yes Ill be quite happy to switch to the branch once is fixed and will attach patches from now on +Screenshots showing some different components of XML highlighting in the Solr admin UI +I took over the improved message for which is now thown as a see above +With the other alternatives provided in this jira and other mailing lists I wouldve lost all dataYes I know the issue is invalidated but I strongly disagree with that +Proposal can be added later by relaxing the requirement of having the strategy defined only at the root level +We have upgraded the release branch to include Apache ODE directly instead of the older FUSE BPEL build we have +Oups! Sorry if I was too optimist about other database to take in consideration is Derby +Right file for version patch +Thanks for looking at this patch Knut +It seems to work +Hey Jeff +Found itIt was in +Ive fixed access to the to be synchronized +Ill take a look at the whole patch as well +patch looks good +Suppose you were only checking JNI methods +Committed +Committed the text under number of regions discussion that already exists in refguide in configuration section +I agree a solution which depends on knowing the source code is not a very good one +there are graphs at the bottom right of jconsole which of them is consistently maxed out? You can click on them to get their name and could try changing the max setting in to see if it helps to something like memory issues were fixed in if this problem persists with that please let us know +Looks like a problem mapping to Infinite +Youve got to be kidding meWhenever possible adding tests to do this stuff is a good idea for several reasons Other parts of the application will have dependencies on this behaving When those other parts fail theyll fail due to a or a which wont get One of the things that has and will likely continue to do the most damage to a running instance of Castor is an unusual situation that results in one of those two classes of errors preventing locks from releasing transactions from committing or aborting reliably can see the reason for not doing it now mostly IMO because the whole test case suite needs a chuck out the window for something a lot less convoluted but not a reason for not doing it. +Patch file for adding the eclipse plugin to the hadoop license files from I suggest this patch in which I replaced the headers of all java files with the Apache oneThat said I am totally ok with this patchThanks +if you need any more info let me know +Thanks Alan! +I would be keen to test this using Ruby can you add this for the Ruby Request as well as XML in patch +the problem with your corner case is that you can end up with a leader who thinks it is still the leader but zookeeper thinks the leader is dead and allows another leader to take overthere may be a way to do this reliably but we need to vet the design first +Can not repeat this issue +Nope +Could you try toreproduce this with the latest Struts nightly or SVN build? This would atleast help us determine if it is still an issue or was inadvertently fixed +File based configurations are now loaded from the classpath if the file is notfound in the base directory +Yossi will upload it after a bit of QA +Id rather flush out any bugs here sooner rather than laterSounds good +If we need to make changes to facilitate the jdk run thats ok but I want a single binary distro +not relevant for uiplugins +Especially useful if we for some reason need to do this again +Will you please elaborate more on how to create Custom Property for Dynamic dasharray Thanks in advanceTrupti +If you have a working updated version then itd be great if you could attach it heread thx we will go over it and update itWe love contributions +Mitchell I think the stylesheet you refer to is for converting WADL to a HTML format +Patch +remove empty directory samples and remove this module from root pom add to samples and modify samples pom accordingly +Confirmed +ramakrishna to me it makes no sense for the user to configure to get such an exception for this reasonable and no harm operation +I would do that in addition to the current that takes requests directly in XMLJSON format +Hi Alan yeah I agree with you about DBRE will automatically adjust the columnDefinition but it occurs only with fields in the ITD not with fields to s a problem for large custom Entities +Thanks againWhen will the version be released +Correct +I was hoping that you could bring the improvements you suggested to the current schema +But overall in better shapeWill add more tests with more split keysMay be some refactorings are needed still +Derived job processor based on Java EE from existing enterprise web console overlay adapted to Java EE job processor +I am no good at python so unfortunately I cant advise on how to best go about setting that upD +Remove Managed X from all issues to +That would be very bad +All the href links in have links toSorry +When we detect the first use of that jar we can then each of those classes +This means that until there is a flush things are fine +Let me see if I can get that done at least +What was the fix for this +The syncHandler also assumes a single outstanding request for a given session since it removes the session handler from the tableI think a fix needs to get rid of both hashtables since it is very hard to do reliably +Would be helpful to be able to specify custom version number for the build in the command line would love to see prepareperform in one step too +Revised according to tech review. +From reading through the comments it doesnt seem like it was ever resolved but we are definitely seeing this issue in AS with clustered nodes +This looks fine to me +do you plan to? to review board +HiCould you provide a simple example that we can reproduce it? seems to be fixed in +The result is that seems to be larger than +so far it looks like the has been taken out of Ecommerce entirelyI am still tracing code forI notice that upadateReqquest is used but find nothis is more complicated that a single patch mistakegoing to ask hans to review since he knows where he was going with this +There is no catalog query to execute to verify the system is ready +Hi RonI agree with you theres no regression when upgrading to the Remoting release so Im marking the JIRA as. +fix attached +on patch +Thank you for your work Rick! +Attaching the properties file with ptBR messages translations. +Patch with changes suggested in previous comment +When you create the server manually then this issue doesnt occur +resubmitting +Maybe Guillaume could give it a try +Yes exactly +Jean thanks for catching that +This way we avoid the problem altogether +I didnt make all the right choices +Ok we now have confirmation that this works fine on Red Hat client from James Williams which made a whole presentation with it and nothing indicates that redhat should behave differently when it comes to using Java compared to Fedora where we have at least people confirming it does not is wrong on the machine QA is would normally close this issue since since there is no identified issue we can fix but I would like that QA confirms that they can get RHDS up and running on Red Hat client andor server edition so we at least know that it can be tested there by QA +Does this mean that I shall stop looking into creating a specificWe will be more than happy to test any early adopter version in our environment just let me know if I can be of any helpThanks again best regards Ill be committing this later today +He didnt specify what it was that he had seen that made him think it was partially merged +Good point wrt the link with +Thats a hook for cleaning data like deleting the message after the message was processedSo the test is clearly invalidYou should modify to kill the VM before the storeAcknowledge was calledIm going to reject this issue based on these facts +No reaction +yes agreed +But then we should retrieve the ZK paths through the utility which uses the separator +Please assign to Lukas once component is added to Metamer +Im missing the friendliness of review support a little bit at the moment +This also included a change in core +If you run grails and restart the form data is blank +Duplicate of. +You are a bit general to really improve something +With this all five ptBR localization files were revised and updated. +Trying hadoopqa +Has anybody worked on this +Grantyour are absolutely right does not block this issue +JasonDo you have a test case for this +version is dead wood. +It doesnt blank out the database name like it did last time +The JIRA rendering rendered the symbols above as HTML rather than verbatim +Have a client with the patched version Call the new api added in this defect for a region in the patched version This works fine +I had resolved the issue am facing while creating the LB rule on Netscaler deviceThe issues +ShaneJust curious if youve made any progress on this issue +We need to track all which must be verified +Broke very long classpath into pieces to fix formatting +Can any one help on sounds as if you need to have two executions for the Javadoc Plugin +Updated to check all the history file name formats and the history conf file name formats +Committed. +Very prompt thanks! +Have to change to change the of addressing module +New updates to codebaseCommand line invocation has been removed since user interaction is required for mandatory license agreement acceptance +Set ts to the time of the delete and not support delete explicit versions but just delete all versions +Can anyone think of a way that a program would fail because of this proposed Derby behaviorThanksRick +To avoid these loops you could switch off the Link with editor option in the LDAP Browsers toolbar +This may have the affect of having them also run for the Ant build +Liang Jia patch applied at r thanks a lot please verify +In poking around in the generator source I have found that the parameter to a custom setter needs to be called value for it to replace the setter that is automatically generated for a property +Thanks Tom makes sense and the Whirr bucket part was just for reference +what about +There have been some recent fixes to the priority queue + to trunk. +Right now it is showing iterator size toArray etc for +I think only major compaction could lead us to this problem which probabaly deletes it +See my earlierMy pleasure +There is now the app assember plugin at Mojo. +Thanks Hao +Postponed to version is dead wood +Closing all resolved tickets from or older +partially done +This will not matter in though since all extended properties will be defined in the properties tab and the key will be a constant +Patch thanks for the patch! Can you create a review at and add the shindig community as a reviewer? I would like you to get some kudos for submitting the patch +Will fix that +Attaching simple patch to fix +Attaching a new rev of this patch with these changes and a corresponding tarball containing the ouput html +version released +The largest I had is million blocks +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be can you produce an example showing this problem and attach it to this bug? +It is being updated as part of the ESB project +Tested +We can always pull it once you get a better feel for how much time you need to be feature complete +Thanks for the George +enabling static nat and disassociated the ip using public IP range instead of portable ip range works fine +Not sure how far people want to go +Im fine with you committing the patch +Hi ColmI am the person who report this issue +old SOAP based verion of testing REST API calls before integration +OK +The only one that failed is which seems to be a race condition +See only bad things come out of +See +I have modified this patch +Committed. +Hi RossenJust spoke with you +Also when opening issues do point to a future release like affect this is not released yet +That solution suites me just fine +I also removed the call to from +I notice this is a very long lived issue and that it is marked for +I beasted the seed for k times no failure and run k random iteration without a failure +Probably in result of the fix for that your second pom has test as dependency and not note that you can also try to close problematic projects or their dependencies and reopen them +Your link refers to the table tag having attributes to set the singular and plural names for the table +As I have been reporting several issues I find it likely that I have been confused by having multiple configurations running at the same time +Its implemented +By pausing the jobs and tasks youll consume resources that will block effective work by other jobs +Dont fill out the next page just navigate away +Could you please attach a complete context wire log of a session exhibiting the problem? Even if this is a genuine bug in it is unlikely to get fixed as branch is effectively approaching end of life +Test failures are unrelated +I did not examine each of the remaining files that is flagging as unlicensed but I believe most of them are either +Michelle please mark this issue as resolved if youre ok with it otherwise assign it back to me and let me know what needs to be changed +This was completed as part of + +Closed upon release of +Attention EvaSorry for the delay in replying but I wasnt this issue so wasnt notified of your questionI agree with your decision and Brians as to the right section of the HTTP Load Balancing Guide where this information is to be added +Thanks for reporting +No need to talk to me like I owe you +Try to add the header to page +This one Im particularly worried about as this may have been donedue to some other limitation that Im not aware of +Copy its source code into Avro +Since I was not able to reproduce the full extent of the regression seen on nightly benchmarks will wait till this patch goes through on such run +We will definitely add parameter count checking +applied fixes to geotools untested but likely fixed. +Attached diff for patching the change + +The chunker and dechunker interfaces have been created +Stay tuned +All the things from this issue will be going into the next ll deploy it as soon as our server allows me to. +Ive got a friend who really need sme help and I had quite big problems helping him due to bad debug good default would be to write out the entire command you have an example on how you want to exclude stuff Jerome? Something like a regexp perhaps? We could perhaps have a regexp excluding + +Allowing this to persist because we have calls might affect the scalability as some other piece of code inadvertently uses the same +There is more to this than the class loader mechanism change though +Passing to dev business is because youre rolling builds outside of the formal release proocess which is what would rename them to whatever version you wish +I dunno +The only bad thing is that the path is not relative to doxygen file orworking directoryWith Kind RegardsOvanes Documentation Project should resolve this for the new software release +All typos are fixed +Anyway Ill be here to answer questions whether you want to pay for commercial support or not +M released +OK Ill cut a new MC Kernel asapIll close the issue once i do that. +but Jakob I am not going to check it in right away +html index pages but it depends on to be marked as solved. +So users should only hit this when they start to test and experiment with the deployment +I agree with Chriss assessment this exception will be thrown under normal conditions and isnt indicative of anything wrong so we might as well not log it at error levelHeres a patch which addresses the issue +AllI spent quite sometime on this so I would like to share this with people +The attached patch fixes this +It still generates the error messages saying with GRECLIPSE from July in my workspace +on Ari only concern is modification time is easily changed by doing touch on the file +Initial fix at revision at revision +The simple patch was tested manually and it works fine. +Tony patch applied at r thanks a lot please verify +When we do the next release we should consider removing this and renaming the directory directly +HiSorry to disturb but I would like to have some news about the progress of issues +Closing this issue as not reproducible. +Great! Lets hijack this issue to address the Atmosphere upgrade in. +I will try to set the id using a +A different colour would surely be a feature so maybe out of the scope for this what do you think +It requires a committers credentials +committed to trunk. +This discussion is very off topic here because this is the wicket frameworks issue tracker and your problem is not related to wicket. +joeluo has confirmed that this is fixed + +this patch is a simple could someone pleasetake a quick look and commit? It would be very nice to get this in for +It seems that a couple of added files were not committedsrctestjavaorgapacheahcsrcmainjavaorgapacheahccodecThe build must be broken without them +But since the provisioning plan is accepted the rest is just downloading and unpacking which Ill leave as an exercise for the reader +Ive checked and it should be rather than +Sorry it took so long to get to this patchI just made some slight refactoring changes to eliminate duplicate code +All makes sense to me +This sounds reasonable and I understand your concerns + +I also met this problem but it only happens when I use GAE +I did not see this exception during EAP testing +please ignore previous this patches are tested with IE and Firefox there any plans to include this patches? +will commit shortly +Sorry Gav +Everything is OK +not in but in for the affected +The other parameters are now available. +Cannot override parameter projectThe parameter project is so you cannot use it in your POM +I think the best way to fix this is to move the JSF service provider files to their own module +Minor adjustments with commit ecfcbcedbfa to enable installation and information retrieval of versioned bundle identifiers. +Patch has been think this issue can be closed along with and since work has been done to support these features on all types of resultsets which Derby supports +Logo looks to be ok now +New patch without any warnings +For instance getMap is a lines of code method that had very few coverage and this kind of test allows to force the code paths needed to assess its it seems what we need to come up is something that cana run fast in order to allow more tests to be written and ranb provide a good enough default config for most tests like the current doesc be flexible enough to allow for easy config tweaking on specific test cases to get the race conditions they needaaime and myself found that a fully mocking approach like in using is overwhelming as it makes for a maintainance headache having to replicate the exact internal sequence of calls to collaborating and challenging stuff lets see what we can come out and pity that gave you so much the speedup in tests would be great +Re new vs new is a Java class that leaks through the Groovy api +I dont believe we had tests exercising lookups from additional packages otherwise they would have failed as well +Id say that on cloning a params object should copy the parameters it holds locally but keep the same reference for the default params +This is because the request has to have mutiple parts and we therefore cannot specify the content type +I committed this to and +bulk move of all unresolved issues from to +Thanks Corinne and Thomas. +When run again everything works +The latest doc patch looks very good + +Changed component and fix versionAssigned to pamarcelot +These look good to me too +Committed the patch with thanks to Brian +Please let me know if you think it should be even lower +Thanks +I will have a look. +I think adding tests for some static resources should be an issue by its own +This can be useful for test still doesnt run successfully but it does get past the aforementioned problem +currently we have just used the runtime that the EAP is closest too since they are more or less the is still just using beneath +In my opinion the only way to solve this issue is to use Equinox Transforms +Yes support for multiple sessions is necessary as my environment has multiple mail servers +There are some Windows libraries which should be relocated in order to get Mb of Java heap on WindowsiaNow with BINDPatch for relocating all found libraries in bin folder +Patch for straightforward patch +This is based entirely on +The error title of Guvnor Repository Error would be fine +Thanks +closing as issues are fixed in +Its a large refactoring rescheduled for going to get to this in moving to +The error went away when I configured the +Done +I comment the code +The calendar uses the entire width of the window. +The most common use case of parameters is to be bound to action fields which cannot have spaces hence the restriction +A quick start which tries to the problem and failsPlease try out the quick start and see if it works or not for you +In the future if it is unavoidable we can work around that +Delivered to and not delivered +One character fix represented in rev of the patch provided +This looks good to me +I attach a patch which I added the test code of environment variable +How does this relate to the servlet filter that david wrote to log requests? This is different? Or about the same +Hi +I have opened and for some of your intermediary comments please feel free to file anything else I might have overlooked. +deploy directory is a directory where axisc install after doing make install. +it requires Java etc +sorry it was dunno how came out +is not dst +It wasnt related to the portlet stuff +work completed for jsr again now properly recorded against Fix Version as well +Ill experiment with this +Resulting in cryptic odd failed scanner issue +I shall fix this bug soon +If someone relies on the current behaviour they should delay the upgrade until the client code is fixed +This is a modified example from Dierks m not that familiar with Matchers right now so I cant come up with more sophisticated patch is based on the assumption that eachMatch is implemented correctly but I didnt find any test for eachMatch +unfortunately jcraft doesnt seem to like either null or for the passphrase +Updating Fix For to Unknown on issues not targeted for issue out as part of JIRA cleanup +is it possible that you provide me with a sample that fails so I can have a look into it? unset class loader problem is fixed +The first four GET tests are of the type of failure described in the email discussion above +Hey ToddI was going to commit your patch but Im getting a lot of fair scheduler test failures even without the patch due to No queues defined when initializing the +The base problem if I read it correctly is that the doesnt expose a method to closekill itself nicely +A should be able to map multiple to a single allowing for full customizability which hasnt been possible beforeSo IMO the idea behind this is NOT to get rid of as such but to delegate them really into their proper role providing the templating engine +otherwise it could be like finding a needle in a longer occurs on the of today so its out of am switching back to trunk +your turn to have a try i played again +I had a similar problem and after upgrading to version the memory leak is gone +Heres a new patch +Brian Im assigning this issue to you +can return an empty Stringlooks like it returns empty string because Distributed Shell registered the URL as empty stringwe probably can sanity check in that if its empty String leave the still as NAThen we dont need to special case the empty string in Also In fact in s comment users are already instructed to use null as default value but in we dont have we can also add that in the comment +Fixed r r r r r +Attachment has been added with description FOP ant task patch for intermediate format file P open bugs to P pending further priority for bugs with a patch +Unless we can get help from users with access to machines we cant do much about this one. +Thanks for checking. +Seems we have our test in Xalan now affecting projects since doesnt get built +Yes I ran Groovy from the command line on the script I attached +I have tested it on the jdbc test suite and got no failures in the junit testsA minor comment utility classes like and should have a private constructor to enforce noninstantiabilit since these classes are not designed to be instantiated +We would like to be able to change the table name so that its use is obvious +verified thanks for applying my patchEverything works as expected. +When I do the same thing here I get the columns you describe +Would anyone want to try this +Changing version to unknown +The changes in the document attached have been applied to the trunk and to my own drools branch +Also what is the class for? It does not seem to be used anywhere +arg no that was not intended +To truly prevent that wed need to make the trash into some sort of special purgatory directory with behavior like no other no +Added a JMX to monitor the callback store +Ive attached another patch which includes your recommendations +I dont like things buried in +Find out what aspect of the hadoop scripts do this check and work around it would be my guess +Thanks Ashish for pushing these changes in the SVN. +by using the Xbootclasspathp command line option for compilation. +Ive ran the cql dtests and all seems to work all right +For the ALLPARENTS case we can compare the pct diff while for NOPARENTS we can only compare absolute QPS for now +On one hand it is much nicer if the follower has identical logs +Fixed with revision +vartemp is not good to be Assembly Target PathWe can set the other value to it +But I am not sure if I should not consider this as error given the first oneAll in all I currently have the tendency to close this one actually +Ok then closing the issue. +It would work fine as long as you werent sending the collapse parameters I dont think youd need to replace the WAR +For this a separate que worker for each workspace can be createdThe downside of this approach is obviously the performance so it should be used as a last resort if deadlocks will still be reported after this fix +This is looking very much like a known bug in the version we just upgraded svn to +Ill review shortly +That may account for the performance issues you reported with it on IRC today +used to throw anexception if a file could not be found +done but we realyl should start cleaning upremoving all those shadow models in its way to hard to ensure correctness and its a waste of cpu cycles to keep them in sync +Committed to +AdrianWill you be committing this patch in the branch and shall I move ahead with removing Anything specific you would like to be considered while creating patch for these? changes may have to be handled relatively carefully +This is still an ongoing issue +Good catch +So its a different issue open a new one with the HBM mapping and the failing annotations one +This patch adds in the new state transition and ignores extra TATOOMANYFAILURE events +Hmm the challenge is that the tracking done for term vectors is just within a single docDuh! Of course youre right + unscheduled issues to the next release +I also committed the fix to x branch if we a new RC then Ill update the fix version here +Makes sense +Thanks Boris for help! +If constraints not supported allow unique on the column +I forgot that the client side isnt working +You have probably similar issue but not the same as this which results with the same error dont use annotation on and +Waits overThis is the Iterator implementation for vectors +Canceling the patch as it is rather old +Closing all resolved tickets from or older +Well this was some time ago who knows whats changedId be more concerned about the proper use of exceptions and removing use of belive this has now been sufficiently addressed +Simple patch +The issue ended up being that the find function in was not able to handle nested +then lets close the issue dims +The was for branchI dont think this should be closed as being invalid. +I will the patch and commit once done +First although the jndi bindings are persisted in the journal it is deleted during server shutdown as part of its undeploy action +SVN revision ThanksRuchith +This is a missing dependency from one of the in CR config file modified +Closed after release. +contrib tests is due to committed this +According to yur explanation I am fine with all changes but not with removeal of the design for extension info complaints + Final is released +If some of the files from the demo site need to be duplicated and deployed which ones would they be +Needs to be tested with the latest release +camillevishal Are we waiting for this to be ported to branch before resolving this +Is this an issue which users have reported? in my experience with Solr mailing list I am yet to see a request where users wish to add arbitrary directories to classpath How important is this feature to be in? Users in general have a lot of problems with classloading +Please update the whole cluster to firstBut I think we agreed that its safer to target this at and acknowledge that concurrent table creation will only be supported then so this will need rebase to trunk +Patch to fix the problem using the class from that a new dependency +I will review and reupload that patch. +DavidIs this done now? can i close the issuethanks it isnt the WG is still working on a solution +From looking at it looks like the Compression and enums need to be public for +Forgot to mention the patch does need to be rebased on top of branch +Marking this critical +Attached patch seems to let all the tests which failed in when I did this pass +I thought it would be cleaner and easier to understand if all the rule logic is encapsulated within one method +moving to M for final updates + +Need to assign version. +Committed +Thanks +Verified in Tools +I think I misinterpreted what was being asked for +Looks like Ubuntu won then did it +Fair enough +I got confused. +Uses NULL for all datatypes +As the doesnt take consideration of the and this time +For me its a neologism that came from English probably +Is there anyone else who can help me? Ramko cant help me? RamokoPlease let me know how to get rid of this error +Similarly a pure JPA implementation can run on a JSE environment whereas an EJB implementation cannot +It also removes the tagged spec +I think its really a bug +If there were compactions scanners or updates in progress these would all need to finish before the region could actually close resulting in long periods where the region was unavailableThe solution is to have call only after all the locks had been acquired and the close was really about to proceed +It also adds missCount checks back to CachingFilters I lost then during cleanup +The old architecture of having composite hierarchies on a single machine has been replaced by a sparse tree model which impacts how this must be done + Username adrianprice +Patch adds to Increment + Bug has been marked as a duplicate of this bug +It works like a charmThanks +integration tests covering both single and multi value cases committed +bq +Its called showsTime on the client side which is a bit awkward +This looks like what Im working on so reassigning to myself +The hostport combination depends on the server where it will be hosted example +I do not understand why plugins are a features +I expect to be able to address all of Karls points +I also attached cmd files that register and unregister the service +Yes +I cant reproduce it today on rev +All set. +The patch is against the line the line is not actively developed +Committed at subversion revision . +House cleaning old unresolved reports +Marking as rejected since this applies to server concerns and is not considered valid as per the last two comments. +Not relevant for V of the plugin. +The plugin central wiki page has now been overhauled and is in a far better state +a simple page to show how it also added to more in messages in the +Thanks. +runs Ok can be close as solved +Except for the obvious case where there is a bug in the new implementation I dont think users would want to override it +I think the usage of cglib for Hibernate is a good at the different patches it eliminates the need for the user to use the additional interface just for enabling lazy loading +ACS uses package to make as obsolete this causes to be unnecessarily installed during agent upgrade +does not support using assemblies from Ruby classes and hence doesnt support dependencies +Doesnt work if shadowDepth is set +Thanks Jitendra +Please see the discussion on +Heres my first draft +Pushing back to since its not really a bug +not works under Opera Beta and Google Chrome also +Appears that by setting the logouty on the securitylogout tag this did the trick +Bug has been committed andclosed already but I have not made any progress on bug apart from someinitial docs within the code +Sorry got the last line wrong +Jar for Colt after removing the LGPL code of and the the dependent classes +Then why did it work before? And why do you pass a Processor in the method argument? I assume it is a fully initialized Processor since it is provided by the framework. +Brock your preference is noted +will investigate +Why dont you have Jervis test with the patch +I created a test case and there is definitely something wrong with this behavior +studioeclipseplugins +Issue may already be resolved may be invalid or may never be fixed +And thanks for commentsBest regards by Andrew. +I thought I had included all the proper licenses but maybe not +Checked in and issue was fixed there +Perhaps we should add a new exception that includes the datum change resolveUnion to throw change Responderrespond to catch when calling writeError and rethrow the datumThis would cause the original error message to be sent +Thanks for the due diligence TucuWith and we have accomplished the original goals of this jira +I just committed this. +If Jetty contains the xalan jars they would also have been OK +Or even a remote one +maybe a warning on platforms this is not working would be nice else it could be all resolved tickets from or older +I just committed this +perhaps a full sort is not necessary just a merge +It was released with the fix in +Inline boo in pages now works +Attaching the +Reopen to check whether addAll method shall return a boolean or the input CollectionFor we can change this now +ahhh findbugs is an external dependency thats not installed on my system +TCK test content is insufficient for full test coverageAt revision test content is sufficient to get full coverage of all node type tests +Attaching comments changed condition for flattening changed to exclude all where subquerys in a WHERE EXISTS subquery from flattening and renamed the helper method accordinglyI have not touched flattening of IN at all with +Proper fix is tracked in but that is under discussion and is blocked by +Im fine with the changes +for upgrade patch +Tests passed cleanly for me on +I still have this issue even in +These services are listed in this section +My team has been burnt by this many times + +This has sat for a bit so posting a quick patch +It may alleviate the exceptions shown in this bug +If this names are true so plz close the isue +Updating description as it was outdated +This replacement only works in case of a limit this bug is probably a duplicate of by +Had to apply a few fixes to the class after discovering issues that were confirmed by a user on the dev list Transmitting file dataCommitted revision . +The generator thought it could use the range from to +As far as I can read the doc and the code the is not supported anymoreThe only reference to an organization spelling is in pattern matching which are only used in the ivy settings files +Updated based on Richard Halls naming suggestion. +i will let it sit here for another day before committing +Closing old issues +Thanks Vinod +ok +If this is still a problem please modify the test to replicate your situation +Forgot to grant ASF license +or perhaps even a different application in the same container +Fixed in rev by removing all use of state info +Actually the suggested fix is not complete +It seems that it requires +which ant release are you using +Nothing except components specified in reRender will be encoded. +good idea I will have the test enable the unmap hack if revisionstrunk +One thing I would ask is to change name of the interface to +I meant you can do that in Apache HTTPD conf which is completly unrelated to +committed a baseline benchmarking suite thanks to Doron and Andrzej +NET on the server +In order to have something to start with +Here is another patch with the following improvements +This will allow you to set which options display and the order of displayBVTcoreWebcoretagsselectcoreWebcoretagsselect Verify +HI MarcoDid you try with want as Joe commented +Yeah that might be a good plan since the code is completely distinct +This issue has been stalled for a few months now +webdav deployment is working fine for me in ask on the user list or if its still an issue +Hi MiomirThe patch seems to be generating additional warning messages can you please verify if it is +Of course that doesnt mean they will be instead have an operation on Endpoint which enumerates connections somehow with some information and an index of some sort? Then operations can be performed by index +Hi EricSupport for jnilib deps is definitely something we should would be really helpful if you could provide a small sample project with a native lib that I could use for ll let you know once I have a snapshot ready forEirik +Changed packages at the request of AndersFeedback would be appreciated +Alvaro logo for web allSebastian attached somes images thatcould be a possible way for the Apache webThe name is a little too long to build a small logo therefore is dividedby two parts that are joined by a littecolumn of cristalPlease tell me where to change it or if thisis not the way you likesThe attached file calledThanks! Sebastian I have replaced some images in resources but after doing that also same old images are appearing +Here are the cleanup changes that I made +willemSorry I was using different logger which was causing compilation issue +Patches for and to keep all these branches in Committed revision +It worked shall work out a beta version of the connector in next couple of days +Closed issues dont get any attention. +Putting on my Reporter hatThis issue is marked resolved and it is but the tests wont come to until is resolvedI will close it at that timeAlso wanted to mention that looking at the fix I am not so keen on it now +I believe all our problems came really from build where the topicqueue interaction really doesnt work +ill try to update this today or this weekend +Cant we just have a new syntax along the lines of mvn plugingoal or mvn plugingoal execution best bet is to open a new issue with your improvement idea +Hi Yang can you build and upload to for us please +Now that my math lib is in lets just contribute the codeWe can make it scalable later on +At this point stability cannot proceed without agreement from the newly joined memberSo stability should be resumed as soon as weve received an ack from the state requester than it incorporated the digest. +Because DOM node lists are live one cant iterate the children list and remove child nodes in the same a variant of the proposed workaround which uses linked iteration +A pull request with some nice fixes in this KrisIf you can think of someone to assign this to thatd be great +HeshanI still dont see the usecase +My argument is thisYes we could hold on to the array in an initialization variable and then iterate over the elements cloning each one in turn each time a form was requisitioned +This often has memory ramifications when using with since most uninvert the fieldOtherwise I like the sounds of Yoniks proposal as well +XUI release candidate B is required to fix the cpu utilization issueComments welcome +The loadringstatefalse directive should probably also clear out the peers table because otherwise the state that youre trying to get rid of is still persisted there +All worked fine. +This can cause A to be before B at one moment and B to be before A at a later moment +Few additional bundle dependence like and are not available on the Eclipse platform feature +Will try to find some time this week +EAP builds locally +This might be a svn issue +The workaround defined in should satisfy this requirement. +But thanks for the feedback anyhow. +revisiting this just a bump to show interestwill be contributing a patch in the next month or so +One thing that is quite powerful in the show variable syntax is that ability to use like matching +But it come down to my fault I want to fix it in next patch I submit +The distinction between and on disk should not be in FCbq +use parameters without making any change to metastore +i didnt checkSo please update to latest sorry for the noise +Im just wondering whether this is something which we should define in BV +My understanding is that Fix version needs to be updated after the fix bit before +Are you facing this issue on put failure case? Can u attach logs? can we have a option for skip events if the event serialization fails +Thanks for the patch +Prevent copying a directory into any of its subdirectories +V has for Ted and Rams review +Hello Brian I cant try this in the production environment but Ill start my dev station and see if the problem also shows without many users +such not suck +adapted to new build process for p +Please could you verify that they have been committed as expected +Closed see comments for explanation. +The real use cases are I want Drill to use bloom filters to work against and I want Drill to use bloom filters to work against Cassandra +Hi LeonDid you try? Have you an use case +Yes Windows +Here scalability doesnt come into picture because this will be used to debug issues which are difficult to find out or hard to analyze with logs +Closing out the issus for the release. +Embedded agent is out of scope coz of Embedded source +Ive consolidated the bundle trackers so that they all share the same strategy for tracking bundles recursively +We do not +Poking JMX in a shell script! easy unless you happen to have crossed that threshold already nodetool is one way to do thatAlso From brief inspection it seems its storing the time of the completion of the last successful repair +This would just be a nice enhancement that would offer better +This model still doesnt account for extern strings rather just automatically detected extern strings +We should probably make this change across the board as part of a new JIRA +added this feature into ajax theme using +I suspect that posting trunk docs on the site could confuse users as to which version it applies +I will also check in the work effort permission services I will leave off the other changes for now but this will get us a good first step +So create a small one for us +The original was inconsistant +Today start on ephemeral ports so RM see the restarted NM as a new NM +Please reopen if this bug persistswith the latest and greatest. +Same fixed theme is applied now. +Looks good to me +code integrated +Only tiny nit I have is that the indentation got screwed up by once space on +First observed this in May +remove copying of row improve performance +I have written a patch to solve this its only small I think that I could be useful for somebody until Keith concludes the new implementation of the location based on virtual fields +With databases there will be a bit more effort of course +In a managed domain the domain controller takes care of propagating changes across hosts +ok added a guardNotice notice here again name of the element was created in its parent but is a concern that should be handled by the element itself +The last image is do you mean by should be the distribution to user? I would say the file is the much preferred user deliverable +I agree that its the best thing but since there is a and the logging framework can delegate to logj if there are more problems in the future with logj and we bring it in for unit tests etc +Forgot to mention that the patch is for the branch but will probably work for the trunk as well +I simply overlooked itFixed in SVN revision . +Negative field ids are deprecated so this doesnt need to block the release +Extract common pieces of into +this is mandated by the JTA implementation +Why not have all of these listed under one category called +DennisIve reverted this change already due to the failing IT +for python after make install you also need cd libpy python build sudo python install +Patch that sets the tracking URL to the RM app page when an AM attempt is killed +At the moment it doesnt just work if you put the native lib in native instead of nativemap so its not just a matter of updating the script wed also have to change the search path for finding the lib +That is exactly the case that Im worried about +This is greatShould there be motivation to add this patch we also need to add the mongodb dependency toDoes anyone else have a use case to add a mongodb injector for Apache Nutch +Trying to upgrade CXF to resulted in finding which has cause quite a bit of instability in the CXF tests +Probably best to split it into core and mapreduce pieces +The project would have to have an owner from the community and a sponsor at +The new test program seems quite clean nowI ran both the tools suite and the old derbytools tests andeverything seemed fine to me so I committed this to thetrunk as revision Thank you for the contribution +You need to use the latest version of RF +Got any time to help develop a patch? Very cool issue reporting. +is related to deque iterator with no matter RWSTDNODEBUGITER defined or not +I think this is the right think its fair to have to specify a snapshot version you shouldnt grab that automagically. +Will come up with a patch in a day or two +I just committed this +Hi ha alguna paraula en angls per a dir camal? +Ill be happy to help if I can +Yes I agree if the inner method was just returning a value thenmodifying the constructor to take in the value would work because no would be required +Ive added a simpleLogging parameter to the tasks which controls whether the tasks configure logging themselves or not. +but if were disabling some tests it would be good explicitly refer people to those tests in the javadoc so they know exactly what the open issues are +bulk cleanup of Jira versioning +Have made the image to text transition here +Now something interesting that I just observed with this new patch is that the previous one was using gigabytes of data of my disk +hadoop fsck files blocks locations provides this d also like to see this information in the fsck files blocks locations filename provides this functionality +patch that fixes inconsistent of files that need their svn property set to +Both of these issues involve releasing LOB resources after we free the temporary LOB then we cant access the LOB this is a change in behavior for the it wont be the default to call freeTemporary on any LOB after added a property on the named release which can be set to true to force openclose calls to be made on any that are read and freeTemporary on any temporary read +Attach a new patch addressing Ivans comments +Are you using Apache CXF or Codehaus? This issue tracker is for the formerMarking invalid as this is an issue not CXF. +Filed a bug with lift +Release Notes marked Not Required as this issue relates to documentation which is still in the planning phase not existing documentation +It seems to me the only reason were synchronizing here is for the increment and we dont need to get our own severity out of gossip so we can just track a local instead +Thanks Franz +Stay tuned. +I havent read all the comments and havent looked at the patches yet but Id like to draw your attention to and so you can make sure that whatever you do you dont reproduce this issue +Ping +Weve been using those nodes for development since cassandra and I guess it held quite old test data so a fresh start was needed indeed. +Unfortunately Struts wont work well as a servlet as it breaks other features such as our more advanced tags +Committed to Trunk r +These tests restart the nodes multiple times +I reenabled site building and packaging up on jenkinsLet me know Jesse if you think we need to add other stuff from this issue to the issue +One node on the cycle has an edge which is not on the cycle testDependency graph has two cycles +Well deferring the opening of the stream until it is time to write the EXTDTA will not work because we need to know whether the value is null when we send the QRYDTA +Two things are happening in that will affect this +I believe this is the last outstanding patch for this JIRA so unchecking patch available +The other members of the cluster will get jgroups notification of its leaving but they havent other way to know the nodes real state than watching its state in the DB cluster table +You dont need to see any source code you only need to read what I dumped above +Regards can do an will require changes to the generated files such that they follow the code guidelines of the projectI would rather see a tool such that we can use the latest version as part of our build and get the latest fixesimprovements +About the maybe could use the check cycle algorithm in futureJarekThe algorithm you supplied will cause an if ABA so still need a flag to mark the bundle is visited +on looks good +Looks like passes this time +Also publishing to cpan would open other issues like who from the dev team can upload updates to cpan so yes leave with ATS +In some cases that works +Thanks again Teddy +The patch diff only contains the differences between the moved files and the results +The new semantics sound good to me +As your patch was outdated I was not able to merge +I fixed the comment on +I will remove those comments from the class +We are creating a modified version of that does the splits as we could split our binary data with Record forOne more query at places our data contains n and r characters as part of the binary data and we dont want the stdin to interpret these characters since its corrupts the data once it reaches the mapperIs there anything that can be done? I dont want to hexencode it before writing it to the stream to the mapper. +I think this is due to row cache +And I also thought it was a good opportunity to fixclean the class hierarchy of the response writer +Chris I checked in CVS contents is ok but description doesnt fit. +Reopening the ticket to track these changes +Other then my reliance on cron to start off these jobs I am pure Java +This is fantastic! Thanks very much for testing and creating the +thanks stephan for his patch +Mass closing all resolved issues that have been closed for more than one month without further to trunk was commit on on trunk along with basic tests more tests will follow +so lets get it fixed +Shows links to field type the fields it is copied from indexquery analyzers as well as the histogram and the top terms form +try qa is getting closerIt is perverse that knows about + +As a workaround I am actually trying to write a custom convertor to convert a Department to its id +I created a unit test for your example in trunk and it passes +Attachment has been added with description FO file used to generate the two PDF has been added with description PDF generated with FOP has been added with description PDF generated with FOP Is anyone reading this? I was hoping for some sort of reply at least a conformation that its a FOP error and not something Im doing wrong +Sorry it was typo +lets use as the modifier if eclipse and use if eclipse are the new fragments +Please change it back if you so wish +Removed whitespace changes and fixed typo +Assigning to Kris since I am not currently able to look at this +It is also not present in the docs as an attribute on task but task query has a nice example selecting tasks and sorting them on duedate +We werent quite reading cpLong correctly or any other bands +Ill log it there and close this? would you rather delete this one +Carlo could you please show me how to access the meta data model from the +version into your geronimo repoThe real solution is to copy the changes that have been made in the branch for the console to the trunk +I havent seen this occur lately +please help +That looks like a viable solution once we get to +I dont think its correct to have a dependency of the shared code on anything but the API codeYour fix sounds fine +patch is compatible with +Close dialog with QR code + +for the patch +Andyyour attachment does not contain a patch but cvs status prints +Patch detailed in bug description +This type of infrastructure is available as in Hadoop +Can you search in mail archives there were issues reported and solutions on windows build +I will continue to track down the problems on vesta after thats done you should move back to ubuntu in general +I also added a junit test for testing Throttling +for +Now uses s static variable writeExternal uses s static variable +Brahma can you please keep the description of the jira short and add the details in a comment later +Ive tested standalone and embedded with helloworld and bpmorchestration and theyre ok +Yes this should work +We didnt revoke the fail LB rule in DB but no vm would be associate with it +The fix is to shutdown the from test +Investigation into this issue has confirmed that there are issues as reported using the implementation on + +if you want to use parameters in scm urls file a new issue. +Ok I see it now +Looking at the code I am happy to throw exceptions for encoding errors return null with a log for JTS limitationsChances are that you data includes a polygon with arcs? If so it is not the kind of thing our JTS objects represent for you +thanks committe +Okay with u I no longer get the Socket Closed error +Screen Attached +Fix committed at r +Just wanted to mention that I was distracted by other work during last few days but have started working on this jira entry again and making progress +Hey Jean Roo configures your project to use the via a Maven plugin +Thanks Kathleen! +stageinstall and stage is empty are totally ignored +Patch attached to disable blob streaming when not running under ThinIdeally we would provide an alternate implementation that supports streaming when running under other servers but this patch at least restores the rest of Deltacloud functionality under +Moving issue w no movement out of +Added the corresponding properties to +This works fine as the jsvn and jsvnversion are compatible to the originals +I dont think I like the concept +Currently we put up whole cluster +After recent fixes the problem is resolved and I can copypaste on both Windows and in rev +Sorry it was solved in. +minor import fix ready to go. +How about implementing using jflex to be minus NUM and ALPHANUMI think as i already commented that it would be best if just implemented the UAX algorithmits a great default and I dont think we should try to alter it +as much as possible +Moving this one to the start of AS as this is going to require some additional workIn the meantime we do now have a that delegates password verification to the realm directly so the realm based security mechanisms can be used to secure web applications +applied the patch with revision +If you find further issues please create a new jira and link it to this one. +If not email me. +They add such a proposal with current attribute value +I guess our options are codify the bad behavior forever or fix it now +Now it is possible to change the parser without subclassing a connectionThe same could be applied to but I dont know whether thats the way youd want to do it in NIOIm through with the refactoring and will work on nextcheers Roland +now the character is showing up as K +May your local repo contains old jars that I do not get from the repo server +In my test environment I see that error as well and the source nat is working +This really slowed down ingest +Trying to replace httphttpsdav providers in a branch and make that it fails on unmodified branch with old providers so its not related to Mercury wagon +Nodes get added to only for exceptions +I am also aware that filters support universal functions +I dont think this is critical +But I will surely like to take a shot at this laterCan we move this to +Assign to Doug +The causing bug was fixed back in please close this issue +raised. +I dont think it is a bandaid +This bug persists in +Another bug noticed on the user list edismax doesnt seem to respect MM all the time in particular when there are negated clausesComparehttplocalhostsolrselectdebugQuerytruedefTypedismaxqftextqxxxyyyzzzmmhttplocalhostsolrselectdebugQuerytruedefTypeedismaxqftextqxxxyyyzzzmm ensure that the the edismax does not have old bug where a query forbar foogets treated likebar fooUsers do not expect the first query to remove all instances of foo from their results +Ive just tried the itests from core they fail on but work file on +looks good note didnt we agree on not having the version number burned into the installer since the version number would be displayed on the first page anyway +Should you add the id before you start process the message or afterwardsIf you do it before then you have the issue what if the processing failed? Should the id be removed or should it stayMaybe we should expose options so end user can configure before after and if id should be removed in case of failure + +If you find out other too wide pages report in this bug file names or +OK heres another update other little fixes tests and matches is added at the top level to give a complete count of the number of docs that matched the query +POM for Oleg for your contribution +Basic XSL stylesheet that turns into a just added a very basic XSL sheet that works if you want to open up the xml document in firefox +Thanks +Patch uploaded +So a clear documentation and a proper example will atleast help users in using it + +Is this really what we want? It wont work with mvn deploy. +Real fix is to have hcat build against local hive artifacts +Taking a closer look I dont think the new test applies to YARN as the Renewer cant be stopped and restarted +I ran this on with the directions detailed in the text file +This patch is nowhere near done just checkpointing what Ive been working on so far +The set of requirements can be that it starts with a capital letter or there is a test scenario for this rule +correct patch attached +Any thoughts on how to include tests for this in the unit tests are welcome +Tests passed cleanly for me on +headers over the wire +This is my listing of automatically inferred Chukwa options along with types +Lets file a separate ticket for the issueChecking this in +may ease the the and there will be no deadlock. +This time activemq restart didnt sorted the problem +General addresses are simply ignored +Attaching code for a servlet as described +xml file otherwise +its taking more time beforee before it actually start i agree with your comments it has to convert it into different format and this need to be communicated explicitly to end userif the size of the template is more then it will take more time to deploy because it has to convert in to different there any parameters to fine tune performance during deploy VM operations +What would be best is if inline validation always fired with onbluronchange and it got set up without requiring users to explicitly set any onchange or onblur on the input components +Otherwise Select a product will not show up +The root cause of the issue is that in order to compress the JPEG file to a different quality level Android requires the entire image be loaded in memory +This also optimized xpath builder under load as there is less thread locals and no on completion etc +I used these suggestions in improving the charting documentation +Looks good +Heretics should be burnt at the stake dont you know that +Now content is escaped and enclosing CDATA is not used. +We intend to provide just java implementation of Tfiles thoughFair enough +Hi MattSince the code changes applicable for also i updated the Affect versions to Thanks to and +This has been committed to trunk +Actually it was just a suggestion something I came up with because of the scenarios we are setting up. +The publish call doesnt do any IO anymore which is good and should not be a bottleneck +Patch upgrades the following dependenciesJackson to Jetty to Netty to toIt also upgrades two plugins to and to +Patch attached +updated in r +It might be a good idea to remove those methods so other people dont fall into the same +I also give clear pictures of what the directory structure should look like after the copy step +add to wish list +The first thing that I will do is rename the package since it looks like we broke binary compatibility +I had meant to say I tested the patch and with the patch checkDriver no longer fails with remote server + thanks +or simply allow the exporter to run but not write the file. +This is still a work in progress but Ill be submitting a patch and more details about the implementation soon +I think we need to rework the way gets its exception trace so it at least has a totally valid Ruby trace and ideally so that it has the extra Java frames in that trace and only shows the full Java trace in debug mode +It is also after the forced token removal of the phantom nodeThe phantom nodes token was +If there is a real need for different serialization type for rpc headersasl in the future we can bring it back in future major versions of rpc +thats super awesome many thanks +the former launcher modules have been merged with the launchpad modules and the launcher folder has been removedClosing this issue for now +Thanks Carl! +Thats funny +In this particular moment both rbircher and imacat are unresponsive +Hi AhmetThanks for the link but were not seeing delays but. +The simple answer is there are too few committers remaining and those that are remaining dont have very much time to contribute to the project anymore +Hi Misty Ive changed the fix version +So I close this issue. +I will give more thoughts and try to upload a new patch +Content has now been moved +Can you confirm this Paul + Code used to produce the problem +In a strongly typed language if something implements then code can reasonably call methods on that interface +Seems a large number of splits were suppressed by scanner activity prior to restart +I try to report it to Atlassian +Thanks for the review Alejandro! +It is up to the implementor to use a dispatcher or not +Ive moved the issue back into WeldMartin I fixed a problem like this in Weld trunk a week or so ago +For it would be good to have an option to cap this value as itself relies on and has a fix timeout +What error did you see +Rick it might just be a typo but svnpatch junk as you mentioned above will reapply patch instead of applying patch +Initially I was planning to use the tmp directory but couldnt find the way to do it in commons vfs in an platform independent way +The steps Im suggesting now are aimed at finding how enabling the can result in corrupt data being written out +So while the symptoms appear to be similiar the cause must be different +This should be raised at the Buildr community. +Yes I will resolve and close this bug +I dont know of a way of doing it it would be nice if we could though +Nope not a problem sorry to waste the bits. +Thanks +Happy it works wellRegarding the documentation I can feed wiki today. +Hit add too said an Object return type shouldnt be relied on unless you can guarantee your back end log manager supports it +Do we want the patch or not +Even isAlive or wait would need to rely on a timeout so we wouldnt have won anything really +deferred beyond release requires changes to core to support comparing type beyond beyond beyond Release +Ive deployed build with the fix. +Thanks +This is not a bug +In this implementation I have introduced the threadlocal concept which makes sure that the object instance should be safe for same thread +pass path to your this work +CorrectIf so maybe its better to check the artifacts and log a warning instead of throwing an excepttion +Sending coreSending srcchangesTransmitting file dataCommitted revision . +domain rerun the reverse engineering and add all the finders in the project +Bug has been marked as a duplicate of this bug +I have not defined a serial though +It is a resource manager and not a protocol handler +You list this as affects did you build the eclipse plugins yourselfThe derby community decided not to generate the eclipse plugins anymore not since +Thanks for the clarification makes sense now +We dont need any other changes +Refactored how events are fired Now the PID and Factory PID for the Event are given as parameters and not as object which in fact may be null in case of a configuration for a newly registeredFixed in Rev +I would prefer to follow the semantics of the jdk namely returning null if the resource is not found +What does does not work mean +Working on trunk patch and unit test now +As you can see the patch is available +Needs the cc patch in +If the this has to be fixed quickly please remove me from the assignee and start work on itthanks +With the sleep the test can fail in the createTable call if the number of client retires isnt bumped +I fixed a typo in the patch the permission for unassign is not UPDATE but DELETE +This is how the loaded +Any chance you could package this up into an example project and attach it +Not excluding for now unless I get asked to +This is why and components have an explciit parameter to connect to their encloding Grid rather than an Environmental +Please let me know if your understanding of this issue is different +Here are two test files +Thanks again! +I tested this on my Windows Vista VM and it went well +but I would prefer if you processed these things using new JIRA issues instead of old ones +Then remove stuff in trunk only +Can you please add that flag to as well while you are at it? +How would the engine know how to bind a specific set of request parameters for my session to the converter? There is no session adapter mechanism to bind the converter to a scoped object for lookup +It is heavily commented but only about lines of code to implement + +Dont see a need for adding this specifically to the recordsIf you are looking for generic serialization Id Sids proposal to do it as a library which converts to and fro the serialization format +Brianthe main stylesheet is xhtmlThe output given is the result of running the transformation without anysupplemental parametersJouni +we dont need anything so complicated as in the patchcomponents are responsible for generating their ids so in this case adjust the repeater to genate idnum +It is found in and uses under the covers +Please fix +Thanks for making the changes Dustin! They look goodJust a couple minor comments I found one more instance of Redhat artifact + +I could get and when I ran rake update +That would be helpful if someone wants to clean the ouusers or ougroups container +The icla and ccla are already on file +We have seen some timing issues with high concurrency with seda queues +Problem fixed and patch attached +committed to ant and master +This means that validator will not now stop at the first null value for indexed fields but carry on validating the other indexed values rather than letting through invalid values which was a serious problem +I just committed this to trunk +What about your configuration? Havent you used the same datasource name twice +For this the proposal is to add a new interface for a function like SUM which is an extension of Function which takes many argumentsThats what I had in mind yes +It is not intended to display runtimes that cannot be downloaded as that completely violates the point +Im running tests locally and if all goes well Ill push +Thanks Zheng +Closing issue since basic implementation is ready +Added you as a contributor Juhani +Probably a dependency is more appropriate +Peel optimization pass didnt check if loop header is dispatch node +So I guess the benifit must comes from the nested inlined region + +Is there any fix available to this defect + +This also hides that generated file from subversion +Infact the fix for all for is the same I just listed all the issues for clarity +should read Include the as a dependency in your web applicationI mentioned trying to integration tiles withI think some of this could help others +have looked into the source inside of RAD has been replaced with and is not available any longer in the API +We can fix that by modifying updateStatus to not overwrite until it knows that the updated status is not null +The need for expect script would go away and hopefully the code will be a bit cleaner once I am done with this patch +verified thanks for applying +This will be fixed by the patch to +applied + +comments on is available +If there is mapping between the doc and your patches then the doc is sufficient +You are not doing anything wrong +This has been implemented and passes the tests +Thanks Yongqiang +Thank you for the review and bringing this up +bq +is not affected +The information required has been added to the JIRA. +Shouldnt we have an eclipse packaging mojo for this that doesnt depend on an eclipse installationCould this possibly be solved by a generic OSGI packaging plugin +It looks really good this way and allow good customization by users +Sorry + now has a flush method Im not sure it actually works as expected though +Hows this going Cosmin? Im itching to fix a couple bugs in error handling during log splitting and dont want to duplicate work +Thanks Cheyenne +I like Tahirs patch better than blindly replacing with synchronized maps +Issue can be reproduced +You may want to give trunk a try and see if your issues are resolved +in salesforce Repository items represent all the downloadable software in CSP these are stored as +i have the same problem +Im a little confused by the closedfalse in the detach +Clearly not test is needed for the patch +Fixing this issue by having a copy of pointer and using it for freeing +Lets test again +While I like the idea of moving examples to a own module and also appreciate the ready to deploy war I ask myself about how much different example modules we may need with this approach +If we null out content and from that and we still have the to deal with +I havent yet integrated it in to PIG as I would like to get any feedback about it +I wouldnt worry too much about this comment +Thinking about this moreIt would be best if we worked on the components in this order base entity service +I have checked and you are right the example does have tried it in Grails though and it IS an issue there +While the only visible dateto CAD engineer would be Commercial extractors do not behave this way +If you are interested in coming some of the way by providing me an interface I program against I will be happy to formulate as a patch what Ive posted here as a prototype +We probably can avoid asm entirely by checking the classloaders for the dynamic classes instead of checking the annotations I will start to work on that very the class rewriting code is dropped we simply check for the correct classloader having loaded the class +It was the first time dealing with the particular data that the error occurred on tooAnyway Ill give a try at reproducing it and if I can Ill let you know +So all issues not directly assigned to an AS release are being closed +Let me know if you have other suggestions though +Proposed patch +In branch the architecture is only determined inside the native code check and this patch moves it out to before that check +unless you have any concerns or at least none that I can talk you out of lets go with this +problem and workaround demo. +There were a couple of test failure in dfs they failed without the patch also +You cant set the initial token unless you set it in the config file +all the above three tests passed on my local windows machine +Should Weld accept calling select without any annotation? BTW I guess could be +for the patch +Which version of AS did you use +The integration tests are in the mpt imapmailbox moduleYep maybe less obvious than it was looking like. +There should be a Galaxy in the Device Library +for the patch looks good to me +A small maven project that illustrates the problem +about having some sort of locking mechanism that prevents Whirr from starting multiple clusters with the same name? Would that solve the problem for you? + for the to trunk. +I cannot reproduce this problem myself +after release +Corrected the BOM POMs reference to the BOM parent +gsp located in view directory +was an issue was pointing to strongly signed assemblies not in the distro +Its due to zone wide primary the issue as attaching a volume to a guest VM in is still failing with the mentioned error +I wrap all this into a jar using mavens packaging +Make sure you do a mvn U to make sure you get the latest snapshot +Tarball includes unified diff for source files the modified files and a test case +The generator is not limited with topNJava regex averages on seconds whereas ORO averages on seconds run time +It is simpler and less error prone just removing it together with the assertion +I cant find duplicates of theses likely that you had an older version of the files where these were indeed other two are confirmed and fixed +Fixed +Commited fix +Continuum tries to checkout from +The fix for this problem on was unacceptable FYI +Expect a patch soon +Still do not see why Ant is required there +not the most elegant solution but it does the trick +Resolving as cannot reproduce +Also note that +Shawn the issue with should be addressed by this as well +Thats not what I far as Im in concern the updates needs to be in order +I forgot that in the patch its in my local checkout +databasemvn installcd +If anyone knows of the history behind state information related to isolation level and readonly mode in please share them +Can you please delete the contents of your directory and try running the script again +Thanks Noah! +A solution is ether move it to product build or separate jbds builder which would depend on richfaces seam and jpdl +Cleaned up the use of c in in revision +Chapter The above configuration shows routing using the beanId attribute +Applied on closing all issues that have been in resolved state for months or more without any feedback or update +However the same problem exists in which implements the identity transform in the Interpretive processor +I am working around it by having my build detect the failure and retry if neededthanks I pulled it into cuantomaster +there is no requirement for you to include pdf zip etc in your plugin includes +It looks like its a class from Apache Derby so if its unmodified I cant see a reason not to from Harmonys point of view +I have similar problem myself Unfortunately dont have a consistent steps to reproduce this problem +So the user can choose to not use it + and both log to +Your original patch is to add lw comment to disable session validation on the leader right? If you could try my approach I think the patch should be much smaller +For example the difference you pointed out between FTP and SFTP was due to such limitation on Windows OS and had to be handled using deleterename +I dont mind taking a crack at this using the +Committed branch and TRUNK +It seems that it wasnt a good idea to switch the formatterIs there an Apache Aries specific Eclipse code formatter? I know that has one and I could use it if I need to create a patch again +Its listed in s pom which is a parent of is already in dependencies list + +Reproducer +The would mean to add a logic in httpd modcluster to receive the multicast from the other httpd server and add a messageresponse to MCMP to request and forward that information to the nodes +I dont think so +Great! Thanks Ill do the other things I had in mind in another ticket +Source View screenshot added +This patch may also fix that issue +For now a simple fix would be to point to common from both hdfs and mapreduce +Attaching patch a which makes it possible to specify true when running ant genrelnotes to make the JIRA SOAP client report issues that are disqualified because they have already been included in a release ancestor +Should release be held until is releasedJames +a possible prepared by Blackdrag looked like the right fix to apply +OK I did disable the cache and restarted the nodes +No test necessary +It is therefore the first rev of JDBC which could be implemented to run on a lower rev level of the JVMSo the short answer to Bryans question about precedents is No there is no precedent +push to +Thanks Navis +Just to be clear The query is already cancelled and then the NPE occurs so the client is not going to miss any resultsbtw thanks for dogging s correct that the client is not expecting results otherwise a would be availableThe fix just refines the current checks to make it clear to exit if there is no receiver and to explicitly not consult the buffering logic in the continuous case +ArunI think the ant tests depend on the fact that the trackernames can have same hostname but different port +Marking this a blocker for based on the decision being reached in +Thanks for that explanation Darran that is a HUGE help Ill add that info to the docs now and then resolve this issue +This problem was causedexposed by my partial defect for defect +Can someone confirm that we can run this on windows? Also can someone write a simple test we can run to verify this is working? Leaving open for the moment +new patch includes findbugs exclusion +Ram Ill comment about the motivation for fix over on +I had a feeling that putting html markup in the properties file might cause some trouble but I didnt see this problem before committing +and simplified version of example code +would be good to synchronize with GPD release as wellI have submitted few patches and still working on few others but it is up to Koen to decide it they are good enough to be included +I have similar behaviour after deploying Rave on a server +How can i change the docs or make a patch to Apache docs +You can navigate the imports of the facade WSDL definition to get the document URI of the physical +Revision thanks Carlin +Ive tested its the pom above the links point correctly to the ve also tested the links areplease file another JIRA issue for that +I think Konstantins proposal seems to be one way out of this problem this keeps hackery confined to the branch while keeping the trunk code clean and streamlined +Ill also ask for some users to provide +Those are the failed testNone of them are related to my patch +Can I take it that there are no questions with regards to the tests attached by myself +Ive got it to work using Saxon as the XSLT Processor and the transition resolvedfixed bugs to closedfixed +Is this an issue a specific issue or specific to both +Patch looks good +Till then keeping status in Resolved state +Fix is in CVS +That changes the behavior though since now you have a single pool of across all compactions rather than one per compaction +I have restored the cardinality check in dot method and also added cardinality check for plus method and so updated unit tests accordingly +As discussed via email we are closing this bug as cannot reproduce as there seems no good way of unit testing this consistently +bulk close of resolved issues. +Closing this issue for the following reasons There is a workaround +The related Publisher issue is missed that Yes by all means! You right method name out of Gateway is something that is giving me pain now but I am sure that use cases for wrt to other endpoints will come so keeping such meta data in the history would be nice. +Using a timestamp is fine for the case of show schema but lets change the extension to +Hi Michael in that case thats what I was saying When you backuprestore the overall Sonar configuration you get exactly the expected behavior When you backup a Quality profile independently we dont know where this backup will be restored and if all related parent quality profiles will exist in the targeted Sonar server +While I think it is the same issue I am not sure that the solution offered in the other jira is correct +Thanks Shanyu for the confirmation! Will commit the patch shortly +Howard Im updating this issue to have a fix version of TBD EAP since there is no definite fix version yet +when will this be released +OK I opened for this thanks! good point +Of course I have tried without a security key in +Teiid has not tested with EAP yet +Can we exclude the client libs from bin prism packages +In my case I use to build the navigation menu so that element would show up blank +when account type is ACCOUNTTYPEPROJECT list API didnt return accountName but it should +I actually read your explanation after you wrote it and had concerns that we might be thinking of two different uses cases +Sorry +has been filed to track work to improve the error message returned in this case to be a checksum error +Design document explaining how options will be modified +Cannot reproduce this error on +Patch in patch didnt apply anymore but the changes were easy to make by hand. +pingCan we have svn path created? Wed like to start import some contentThanks +Thanks Travis +Hey Prashanth hows this JIRA coming along +Yongmin what do you think +Assigning to Rich for review +Rene can you post the workaround +Just so its clear what is being asked for here is the ok to include a test xml file that uses the license within a source release package +Committed to branch and trunk. +I have it installed to localhost and have tried using of httplocalhostmanagertext and httplocalhostmanager both to no avail +HelloWould you please try my patchThe patch includes and ThanksBest regards Andrew Changes committed in revision +Antoni is correct the used value was wrong and just accidently worked for the other examples +Next though is to figure why though all regions have flushed you still do not have enough free memory +Moved from to just after merge of old branch into TRUNK +Enhancing for SWF m +Furthermore we will still delegate to the initial factory where possible +Looks good to me +There seems to be threading issues in libhdfs when disconnecting a filesystem if I create two filesystem handles disconnect one then try to do operations with the second the JVM itself sigsegvs and dies +But +I had files for ICLA nowApart I had verified for this issue and now these methods returned the correct values. +Patch committed in r. +Oopsy daisy +Uplaoding a patch that inserts a +Also have to make sure that the error dir is not the same as the input dir +The patch you have applied will honour the comments even for my directory structure. +As Max said it would need to be tested on different platforms which I dont have time for ATM +No issues +defer all X issues to +This is still broken +Added description of buffersize to Chapter of Remoting Guide on and branches. +Changing the urgency to Blocker +Committed revision +With the interface it does not seem there is much that can be done +The only real thing to do would be to trigger a rebuildof the string on display change so the no longer displayed text can beremoved from the ACI +I assume we should hide by default following icons from main toolbar Preferences Page Design Options Show borders for unknow tags Show tags Show text as on JBDS. + +Looks good +Switching all issues that have been in resolved state for more than one month without further comments to closed status +I will commit in a bit +I think that somehow didnt get reflected in the v patch +Thanks +I have completed a further test with a Callable that returns an object to be marshalled into JSON +Fix junit already applied the last patch to unittest while testing +As far as most people I know are concerned is a dead branchUnless we declare it dead officially it is not dead yet +lets discuss it at todays call +Its expression language syntax is much more powerful than thatsupported by Struts tags +Still need to backport to to and backported to +BTW all the camel ones have fine descriptions +properties files to use as is no longer supported +Yes disabling lock striping will help +I can build the project fine via on the Run As +This is the weirdest support case with the weirdest conclusion that Ive ever been involved inThere was definitely a problem but we could never figure it out +hmmm +Resolving Cannot Reproduce. +I commited a fix for this +Committed to both trunk and branch. +Well I did see this pass once with the upped timeout but now seem to reliably get FAIL Server failed to respond to ping ending testwith firewall turned on and ping time set to I am sorry I wont be able to update the masters for this issueKathey +DarynFor the hdfs mini cluster it is all running in the same process but for MR it is not +Applied patch in rThanks! +Also after the clone the binding provider isnt reset so in effect the binding provider and associated wires are shared between all callback references for a particular component +Deleting the configuration folder as root did not solve the problem +But my feeling is that at least some info should be displayed on the console +Attached is the I use for manual testing +to commit to trunk and tomorrow if no objection +But youre right maybe we should sometimes prune +Patch looks good +Someone with commit access should put this patch in the cvs +Now we can either share a single server configuration and deal with the conflicts or simply have two separate custom server latter option seems easier to me +Ah I now understand the simple comment refers mostly to word stemming not to analysis in general +Ill commit the changes in a few days unless anyone objects +My fault I have changed the number but left ref to the project name +fixed in cvs +from the alpha docs on the web site the endings are wrong +Would you please help to try it +Hey Tony thanks for the contribution +pushed again as we do have a issue is reported against a version which is no longer maintainedwhich means that it will not receive a fix +Hi MamtaI dont know why the newlines disappear +move the work to after the commit +But by then I am hoping to have procured access to my old testing environment formerly at and now at qBase so that I can verify the codeIf your time frame is too urgent for that then by all means I urge you to go ahead +Investigate possibility of extending stemmer to work in expand mode +this is the implementaion that uses services XML to configure the REST based access of web services +bq +By excluding it from being processed solved the problem for me as I did not needed it to be included in the output XML in the end +Tests all pass +slightly changes some code and comments +Multiple arrays? A tad gross is pretty big damn big +An example for a large enough source I will try another way +This is a patch that will allow uses to cache Filters even when using is new and better version of my Lucene patch + +I think this is primarily an issue with multiline string literals both in the CQL grammar and in the python driver +in my case its not a bug at all for the help +Bulk closing stale resolved issues +Still wondering as to how the type cast to old MR API could happen? +Ive found that there is a class which is coded in such a way that is a pretty good fit for +Patch containing updates to various JSP files missing localization +I think this is for commercial version but will be applicable for open source as is no version in Apache +Bug is in ESB and should be fixed in ESB +i just tried to verify the behavior and I got the same result as Nicholas +Reused newly added function +I agree with Guillaume that if we are going to support Maven then we should use the latest stable versionEran can you please test your patch on? ThxAs discussed on the dev list we have a major event coming up and we need to prepare for that +The new code closes block first +phongCan you please submit a diff u against latest CVS sourcesthanks here is the diff to revision Patchthanksdims +The issue of what to name local graphs for remote users profiles is still to be settled but that is a minor issue +Just one question The cache is initialized with max entries +This is what the Eclipse generator generates +Making them volatile would fix this but would not help with concurrent access +Attaching junit that demonstrates this issue +Thanks for thatNow whenever we decide whether or not this should go in to the project my vote would be that this replaces the existing accounting menu structure since its already enough tabs apps for the user to deal with +Ill just use your codehaus emails. +So just want to be honest by saying that we wont make any effort to fix this issue except if a fix is provided by the community +ouch you risked you own skin +thanks +I am also using the M builds +Please kindly help review. +I have attached it +This is something I plan to pick up when I return +Mura and are both rather popular and I found no other reports of this type of error +Made implement and drop the cache contents on a resetreload +people should still use though +Completed the testing for the feature hence closing the issue related to the feature request +All of the packages from the test compose were installed so there were no leftover packages +Should be verified in gatein master and reopened if still valid +Uma thanks for the reviewI have committed this. +The object in question can either be a String or an Opensaml for Attributes of different typesColm. +We might be able to work around this by regularly touching the file but Im not sureCarl are you still seeing threading problems with the most recent patch? has been succeeding for me consistently +Also in standalone mode you should somehow inject your services which is easier with specific test context +This is going to be very hard to do in Seam component type is needed at bootstrap and we dont have access to at that point to abstract out the call to Hibernate to discover whether the class is a hbm mapped guess we could hack it in but is it really worth it +Since adding explicit resource directives Maven didnt copy into the jar anymore +Hi FrederikDid you drop all tables by using or make sure the h database is empty and run the Activiti application using db Anyway I will it for the coming rc release if you cant it. +This cannot be the cause because we are using already problem was resolved in version +The patch of does not change anything for this issue +Hmm the suggested patch is not good for DRLVM which handles version normally +was failing with NPE since it didnt support Cookies that started with Version +The current error message is fine ofr people who know Groovy but is no good for people new to it learning programming +This would only really pay off with a RAID where different parts of a file are stored on different physical devices +Ive also noticed that other services like CIA seem to work on a single commit level so it would be consistent there as wellJukka I hadnt seen that particular project but I have been looking for other examples +Update the testcase to verify it could correct this issue too. +If we can it is actually a problem with the Changes Plugin I will move this issue into that project +The last time there was messing in this area it was a bit of a disaster see and its subsequent fixup issues +The mirror is now up to date. +Sneha Ive assigned this to you as you maintain the SpringSnowdrop docs +Cheers anyone made progress on Christians first version? It looks good but its lacking the ability of moving trees which Im having trouble implementingItd seem that event is not the way to go +Any thoughts +A comparison of source pool or volume against target pool of volume yields the information if storage migration is required for the VM to move to target host or not +Feel free to remove +Technically what we should do is change the request ids but I think we decided not just to kind of start fresh +This patch contains a testcase and a fixed +I briefly read through the Message Transmission Optimization Mechanism specIf I understand correctly as far as the generated wsdl is concerned the only requirement to use MTOM is that the elements that we want to send optimized on the wire are of type xsbaseBinaryIf thats the case then no changes need to be made to the idltowsdl tool +I find most users expect a common sense approach and dont usually take the time to read specifications +verified +is treated under windows exactly how exec launches the process under windows etc so couldnt narrow it down moreOverall it seems clear though that exec on windows does not hand over the state correctly for +Build is now by reviewing hudson job state. +Doesnt seem to be in the incubating maven repo yet but Im sure it eventually will be +Lets commit this after the release of +please merge with branch +I would be interested in taking this up as a GSOC project +They work in both servlet and portlet environments +There is no type conversion +Youre right about the sessions +To test this youd need to bring up a datanode without any namenode try and interrupt it +Merged thanks! +this issue is trunk only. +I suspect Tom Enebo can probably figure it out in no time at all. +Left for community to resolve. +Oh +Caching now helps to make things decent. +I dont want to specify the deployment folder again and again I just want to state that the deployment name should be instead of independent of what the deployment folder is setup to be +Oops Gavin I dont have installed the transaction in +Yun Feng this is just wonderful! Now we can use Abbot and Selenium all within Eclipse +youll just need to test it with JDO +The fix may be to port work from that issue to the branch +Ill open an issue +The query should show up when permissions are checked against an we should consider asking the person who reported the issue what path he followed to run into the problem +So the test should not need to be enhanced +Even when a is not defined for a connection factory like the connection AS tries to create a assume a dummy security domain called other and tries to validate a null user thus gets into this exception +Pretty work as netty made this rather simple +We should have a cache in front of the call +There is the issue of RPC threads getting blocked which may necessitate but otherwise would be preferable since it requires less cooperationcoordination with the +Please clone this so there is one for and one for +Will include this after a new tranql release +the page properly adjusted for each groupThe advantage is that you have only one page with all the portlets on it but still serving to different sets of users with access to different subsets of all the portlets +No need to run derbyall since the change is just in a junit test +all resolved issues due to a successful release +The thing probably worse than only varcachecmdline +I have no time to figure out how to buildthe Tomcat distro +screenshot of the added to scanner is not added unless the checkbox is checked +Out of date +and maybe old lib still installed for old programs +I thought that push was employed primarely to boost performance +Can look at it again later if required +First we need to graduate +Sorry Igor +after further investigation it appears that pats fix is the best +Implementing the exclusing of the s shouldnt be to none of the Maven guys have the time maybe we can provide a patchI got some other things on my mind but if its isnt fixed in a few weeks I ll provide a fix according to the development guidelines +I set it to visible to anyone at this point given that it has been in the public eye for so long +The stomp spec on github contains all those details and any of the stuff is captured in the page. +Will push the fix soon +defer to so I can release in JIRA and get the release notes generated +I need to store the html content somewhere and not a new db change will be accepted +If we plan on releasing for each version then we need to be looking to do a feature freeze on soon so we can roll it out fast and start work on that will tagged for is unlikely that we will implement dual log writing threads for the first release that uses hadoop appendWe dont know anything about the performance of append at this point and writing the is not a bottleneck at this time +Issues resolved in +The BEA lawyers wanted the BEAIBM legal verbiage at the top of each +The stuff I brought up can definitely be iterated on in later patches if desired +If it is not the shell then it can be fixed in code +A query consists multiple each of which is a step in a query +we should push for getting DTDXSD schema handling done better in eclipse WTP they should at least be caching these and not rely on fast responses +attached and which would help in resolving artifacts from the local fs when Dresolversinternal is passed +Its a thread start and stop calls +for the patch +I should add that it is unknown when will be addressed on the Teiid side so this fix should still be considered for Designer +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +To do this I declared it safe that we may delete a few more jobs from the cache then expectedThe class is no longer informed about items being removed from HDFS and the tries to preemptively know that they were removed but it is not that critical if it does not happenPlease take a look as now that there is no locking around the there needed to be some extra checks to avoid +Markus So I shall go ahead and create a separate jira about replacing concats with appends while this can possibly go in now +In addition this patch enables the existing benchmark to use some input dataset on HDFS instead of +The example suggests he is trying to parse XML but lets wait for his clarification as to what is actually being parsed +and will start working on it after we set the design +I have tried to run quite some time ago with the same result +Build passes using mvn locally with this patch +On switching over it should no signs of leaking +I have created issue to address the second point Ihow id you get the NPE you mentionned in your first post? I want to reproduce it to be able to fix itThanks +This patch changes to favor the code that was in the based snapshot branch instead of the based trunk branch +Updated patch +I see it now +This dialog is creating new images at the start and then new images each time color attribute is clicked +My concern was to make the ecj compiler pass the specs properly +style and styleClass is missing also for toolbarGroup component +Please run tests and if you havent +Thanks +is okay. +Cdrics patch looks good to me +Yeah looks like duplicate +Setting the release notes fields +Thanks for your contribution Christophe regardless! +Job completed +Thanks Stu I look forward to your review comments +Sorry Im wrong +Merged pull request test as well as the in pull request with Git commit ID fecfacddfdbbadca +Konstantin Any chance of an update on state of the branch? Should we start trying to use itbreak it? Thanks boss +Also items that fall below a threshold can be pruned +what is your exact environment? AS or Tomcat +Sorry +Fix checked in +I am currently attending training the week of until +thanks that will allow me to fix it +Its on windows Ctrl or SIGQUIT on most unix like systems for the sun based I see what you are referring to reguarding the loadClass calls to UCL +That is asserts are not compiled in +This was not obvious in the code which tempted me to reopen this issue +I think it is still kind of complicated because there are so many abstract super classes and interfacesWhat do you think about replacing the with +OK Ill have a go at thatIn what form and where do you want the documentation +Submitting patch for trunk +Commited to the transports trunk +I think its reasonable to disable this test for bit while we look for the revision that caused this problem + applied in r +In some cases the query could be SO messed up that interpreting what MIGHT be the problem seems like a lot of workMake sense to me to tell the user WHERE in the SQL it finds a problem which is contained in the current parsing errorThe second error message described here is much more verbose and can probably be simplified but that would be a Teiid issue +I therefore doubt that this would be used for serious work +The select transformation is the view definition +I obviously wasnt paying attention +Guillaume not sure I understand +the fix +It might be significant that the namenode is rolling the fsimage file right before the +BTW most of the code are pasted from the Plexus Util s latest version but modified most of it to handle the dotnet tools patch doesnt apply to the which plugins are those changes relevant to +Does that really really mean we cant change itIf we go with a builder approach can we just deprecate the old methods and let them hang around for a release or two +are loaded so its rarely a problem +Would anyone want to try this + +Should this warning also be generated when the server is listening on an address that is not localhosth +Would be nice to have a show schemas command +Bulk defer all frontend issues to +bq +Is this a blocker for? If not we should fix it in trunk. +bulk fixing the version info for and all affected issues have in seem pretty green here now. +The test application I used +github is also synched with latest changes and incorporated review is done and in master branch already. +Thanks Jakob! +bq +SVN change . +Xavier Is the behaviour described at previous comment correct expected? If so this issue could be closed +Just switch the between and to verify the difference in behaviour +After a year +verified on trunk and JBDS to set proper fix to issues that were set to already been verified +im having a hard time duplicating this in other places +Script that offlines regions in the first run and then on subsequent run closes and deletes regions finishing by adding up a region to meta +I am having problems. +So the isClosed returns trueThe new check that is added to see if region is splittable internally checks for isClosed +Ive got the problem with other apps too so Im attaching a file of another app which breaks s loading at the start with this +It is not affected by Java version and Maven version as I tried older releases and they are not working as well +Version Released +We have a few more changes to merge into this patch +HiIm a bit annoyed by the fix you applied +I like the reflection API and am very comfortable using it directly +I should mention that the segfault issue is already reported at this patch should fix that issue and this one +Updated to work against trunkAs always let me know if there is anything I can do to help get this committed +to the errata too +I guess Im thinking of cases where starting with a URI string is not by choice perhaps it is obtained from an external system or retrieved from some storage +The reason tests are during perform is because its part of the site building. +is required for splitting of Bzip data +Thanks I had this same error sometimes and didnt know what the heck it was +I believe you need in order to compile never tested againstI dont support those platforms myself theres been some patches in the past to try and support them but I have no idea how well it works there +Thanks for reviews Ted +I attached a test case which reproduces the problem +useful stuff here +Patches for this issue should be applied after patches for otherwise it may break other pending JDWP patches +If possible the list of displayed in the runtime configuration should also be filtered so that only allowedsupported are displayed +Can you suggest work around for this +Attached patch implements the optimization +Tested with both java and java it works only thing I still havent tried is using a bit version of Java but I have a notebook with it so tomorrow Ill try that one as of curiosity how do you get your fresh build? Gridlock you too +Dan press Refresh when? +Attachment has been added with description Sample FO and PDF transition to closed for remaining resolved bugs +Fixed and will be available on the release +I am sorry for wasting your time +Tables that have been used for a long time are now giving us that kind of problem under a MR +Need some time to check over why it is failing and why in the original patch it was modified +The reason I chose to consider Sets and Maps incomparable is because there is no natural ordering to them +something to be wrong at that timeI meant its ok for close flag not to be set +Doug could you please review and as well +Theres actually no need for it to be done there as will correctly absolutize a location whenever its given a relative path and it doesnt make any sense to prepend a file path to a location that isnt actually on the file system so the decision to absolutize needs to be left up to the Slicer +back on CR since the actual plugin is missing from jboss tools update site +Will be rolled out in our next jira related release +Patch v fixes javadoc am ok with the change +Another patch removing more xml QP build dependencies +I believe this is for compliance reasons +Patch with the additions noted +Well then on the same lines wont it be good to have type support in the tool as well? It would be helpful in case where data has to be migrated between the tables with some type transformations! why u think a type conversion needed while copying wholepart of data from one table to another? Do u have some concrete usecase +Will require a patch to the +Unfortunately the classloader seems different between these two and I dont recall it ever being anything else but a source of grief +Feel free to try it out from Roo master and let me know if you have anyI made further adjustments with commit bbddcedaebffab +I confirm James patch works for me +Yes +Attached patch with fallowing properties added to Doctype Property Resolver +Id hate to make the factory take the super set of all parameters that any of the Writers want +Ive also made changes so that fixed fields come out as binary as well +Patch applied thanksColm +We are using for so long the option true that we forgot it is not the default behavior +Thanks for the patch +Thanks again! +IMHO the issue is that the API is huge +Similarly if disable is set to true then puts to the will be ignored and no Map will be created +One hopes that anyone trying to install or manage would have enough experience with Yum or APT to know this already +Should it not be +They should all be in for it to work deterministicly as far as i know +Issue may already be resolved may be invalid or may never be fixed +Yea I think having one argument here should be ok +The nature of the problem has changed but still persists +Yes thats the problem I referred to in +More enhancement for supporting multiple parts with the same name may need to be coded given that not all form data payloads containing multiple files are +Hi Daniel could you check the that should fix the failed tests +not finished +Any word on whether this will be accepted into and? It would be comforting to know if we want to upgrade to we dont have to it ourselves so as to not break our systems +I have verified from my source code build looks good +Only true false and are supported now not having this issue. +sends which do not have this source +Added the system property that modifies the behavior of the week and day functions to be consistent with ISO regardless of the system default localeNote however that the day function whether using or not is still different than the dow extract in PG +Comment I realize that STATICCALLS will reveal the problem that is not the issue +Sorry we couldnt get to this issue from when you reported it in and the website URL you provided now returns a +It could occur three problem cases in the following teststesttesttestThe following test ensures occurring case problem +This is then referenced from the for certain for the patch +mvn clean install P +This patch fixes a few more compilations errors after applying first +Are you sure the collection is empty before the insert and remove runs +Update to trunk code +Setting false does not work +So it is easy to detect if the reason of the error is due to a lack of connectivity or an invalid session +It would be interesting to run experiments w and wo native maps for M M B and B entries +PS but anyway Id rather see an explicit copy of the LICENSE file in SVN in the mvn module root directory than have it pulled in as part of the build +This is just the same as if we hadnt truncated the log +The prehistoric block problem is still relevant since the s keys blocks by their ids see +Arno I do not have the ability to close this issue +Please take a look +Tucu the patch I uploaded earlier adresses all your layout and build comments +Please remove the previous version +Its quite easy to omit the attribute afterwardsBtw +For some reason a local build of locus from contains the classes in the +done at rev Thanks Olivier +Seems a bit of an overkill for such a simple tasksre facing a similar issue and would also vote to add this functionality +to bothe branch and trunkThank youSaminda +Thanks Carl! +Fixed at revision The iPOJO logger does no more print log messages if a log service is available +The renewal should be improved to avoid guessing the URI of the NN +dn krbkrbtgtKERBEROSDOMAINKERBEROSDOMAINouprincipalsodirectory +Please change to if you intend to submit a fix for. +Basically unpack this archive and run tried to remove as much as possible in order to focus on the failure +Thanks Jarcec for reviewing and also the additional tests that uncovered more interesting scenarios +Now there is a scrollbar in the tutorials sectionCommitted to the trunk +Proposed resolutionpatch to be applied in trunks why I dont want to store the original request headers in the Enhancement Graph but rather in another graph or dedicated java datastructure +is currently one big garbage can IMO +Im having trouble deploying to maven central right now so cant publish anything new just yet +Slava can you look into this +Thanks for Jonathan the selector can automatically decide to parse the content or not +what is this plugin doingA slave broker blocks on creating the persistence adapter which is typically locked by the master +Same issue as which has been fixed recently. +I agree theres something strange or wrong here +I guess there are a lot of other documents which will render better than before +The request to close mail lists needs to also note that private has to be removed from the pmc mailing list +The fail Ill dig into in morning +I have a two questions +The reason is that DI tries to apply the change locally first so if a particular node might not have the latest values in memory +missed deleting some +This defect is fixed in latest build so close it. +I cannot reproduce this problem +When using the Data Direct drivers they seem to import correctly +Verified in repository +then the popup window looks like in my screenshot aboveRESULT OKCONNECTION ldapDATEdn uidcnuserscncmsdcAPPLICATIONSdcDOUGLASHOLDINGchangetype addobjectClass inetobjectClass organizationalPersonobjectClass personobjectClass topgivenName testuid userPassword dsn cn testRESULT OKCONNECTION ldapDATEdn cncnrollencncmsdcAPPLICATIONSdcDOUGLASHOLDINGchangetype addobjectClass groupobjectClass topuniqueMember uidcnuserscncmsdcAPPLICATIONSdcDOUGLASHOLDINGcn RESULT OKCONNECTION ldapDATEdn uidcnuserscncmsdcAPPLICATIONSdcDOUGLASHOLDINGchangetype addobjectClass inetobjectClass organizationalPersonobjectClass personobjectClass topgivenName testuid userPassword dsn cn for the answerI can see from your logs that you managed to create users and group but the creation of the nd group failed +Modified the camel parent to use latest qpid version of the Qpid projectCommented out dependency which was used by an ignored Unit test +Similar to what was done in Continuum Ill remove warinplace and use multiple webapp sources for instead +So this bug does affect actual in +it is done! +I am resolving this as cannot reproduce But if anyone is able to reproduce this failure please post to this issue +Fixed line that exceeded +Unit test doesnt help replicate the same situation +Ill take a stab at this +What do you say we link this as incorporates +The priority of the issue has been lowered as a result +pls +Im not sure if theres a valid use for this or if most databases would even allow it +Any method who names starts with is now ignored. +srcmainjavaorgapachemavenplugin in exception text Id suggest removing the +Lets also fix the typo as noted by this as part of the addendum patch at. +To be verified +Use and include in your classpath +The open call however is not aware of what the application is doing so it doesnt know if the application is really reading only committed entries and in that sense the open operation is unsafe +Weve also been busy with the and releases +The top most +So if mappers still need to use local store then this is dead in the water +I will take a look now +Committed +no problems with patch applied reported so closing. +Right but they all deal with class and property names all of which will be more in the next release. +so i took the liberty to commit since its anyway +Do you think you might be able to add a brief description of what Forge does incorrectly when operating on an EAR? Also perhaps attach a sample project that we can reproduce the functionality withI know this is asking a lot but it would really help us get things workingThanks least I can do is help out Ill try to do a run through of a new app in the next hours and post the results with more detail +Wow it went down from violations to just +Yes except that outputDirectory directory of empty causes the file to not be added to the archive but no warning an outputDirectory of say docs works ll write some unit tests and submit a patch if needed +currently also being called for parts which are lazily loaded and in this case the lifecycle listener will get completely wrong data handed overMy patch is because the fixed handling will only be used if POSTLOADONMERGE gets enabled inIf noone objects Ill go on and commit the patch tomorrow +That way users can specify the type of join they want and if for whatever reason we switch the default theyll still get what they want +Moving to TBD EAP +Thanks I seem to be unable to perform the nexus migration task that disabled the old repo +Do the parallel tests still require a shared NFS mount? A while back someone told me I did not need NFS anymore because hadoop give me Big Datas +I ran tests for this change and all of them pass +Ill create a separate issue for that +Final Logo Set the in the zip file +the txns are supposed to be idempotent but the way we maintain the cversion is not idempotent +Patch and the two binary test files +If tag is on field set it to field unless access is explicitely overriden +It doesnt yet support passing that option for things like rebuilding resizing etc +Part of a bulk update all resolutions changed to done please review history to original resolution type +groupId should be to the fact that we want to do a complete rewrite of it may be that xmlutil will become obsolete +I think we also need to limit on flush since theres no need to knowingly save data we dont want on LCR as well as PCR since enough small rows can still overflow to LCR modeAlso if Im understanding correctly were tombstoning the beginning of the row here? ISTM tombstoning the end of the row will be more in keeping with our advice that querying from the start of the row in comparator order is fastest +Should we make chmod parser also consistent with this? I feel we should either change both creation umask and chmod umask or leave both of them as it is +Updated checkstyle to where this bug was fixed +Reassign to a release that you feel comfortable resolving this issue in or leave as is +For future reference I need all changes completed and confirmed at least a week before GA so I can start adding them to this issue +Please do a case insensitive search for veiw to correct these +No test since its just an error printout +If that does not happen within a file data might still be sorted but when you read the files in order data will not be sorted +Subsequent commands try to use those +committed w minor tweaking +event attributes should be checked once plz recheck all Skin Parameters Redefinition Style Classes event attributes +hi filippo could you please give some more info on what you are doing +endpoint snitch isnt strictly necessary but makes some things easier +Unsetting Fix Version for unassigned issues. +At this time I recommend this feature be postponed until a stable version of Spring is released +see to where Mitosis will be and entryUUID index has been added +I will deal with this tomorrow +Im okay removing the news section until we have some news itemsI also like distinguishing between user documentation and developer documentation +Dont want to annoy you here with all my comments +Ill carry it to the more advanced Groovlet usage perhaps itd be best to turn to a light framework like Grails? +Itll be fun trying to find out whether thats the plugin or Compass +should be changed to allow filtering of subscribed components as specified on the configuration file using nested tags of subscriptionSpec is now applied after default filtering on service name interface +We also used Lzo to compress the intermediate map output data +By the way as this is occurring on Solaris it is preventing a clean build of for the Solaris jdk customer certification +Just a thought +HiCould someone please backport the newly attached patch to? It has pretty much the same changes as with a small difference +This might be related to +Thanks again +I dont know what jars are associated with the docbook so I will leave them for you to delete +No core tests failed +Above mentioned approach will send messages to rest of the members and ignore only the one about to join +Patch was committed with svn revision +I removed the previous two methods checkPosition and checkLength from Lob class and added one method check that checks for the position and length input parameters +Im closing this as we have another issue for RC but lets keep it unverified for the time being so we still easily see issues raised by Ate in RC. +There appeared to be something specifically wrong with the CVS directories in the in petclinic app on the build server +Great it worksAs for now I see no side effects +You are so nice Mathias I am using the Distributed Search of Solr and also let Katta chose a node for a shard +Then traverse the subtree for each node in the subtree +Bill if you can will you craft patches to add the naked modules as well as the modules and the fully unpacked modules +parses any rows with TMPPARSEMARK set and sets TMPUPDATEMARK as before removes the column TMPPARSEMARK so that a later parse wont the row +Not sure that is causing this but that is pretty bad +How about calling the first phase create second open and the third phase transition +Im hoping to finish the stuff in the next day or two and provide a possible final patch candidate +Closing this issue. +even if it can looks quite strageAlessandro +having a configure addremove project archives makes sense to me. +Set version +Since we are at it shouldnt hadoop version be updated as well +The main cause of this is the use of the ant zip task to explode and recreate the jar fileThis patch removes the rather clunky ant task to move the ASL jzlib ognl slfj and springframework into each jar file after it is built +words should be highlighted in blue without bold and italic components in blue bold attributes in blue italic attributes values in blue without bold and italic +Martin or can you check if this problem is still present in? there have been many changes since but I cannot verity if works better unless I have a complete project that demonstrates the problem +Now incorporates tighter integration with shale +The dependencies are so wrong we would be better of with no dependencies at all +Thanks John +Moving this to TBD EAP +Gary do you know if the SOA will consume EAP? Thanks +Testedwould be great to have an IT to test deactivation of the cross project duplication detection mechanism when the property is defined +Is it because the execution path includes java methods and those methods occasionally attempt to allocate an object while a GC is underway thus causing a deadlock +I mark this issue as resolved. +moved to to +The fact that documents are sent as different requests of documents each might explain the higher load on the replica +Change committed to trunk. +New patch all nocommits fixed. +Thanks Nadav +This is true if a CF is sorted by name +I dont think its relevant to log absence of combiner partitioner etc +I am confirming this +Very happy to have this handle just static content as I have my doubts about this being used for dynamic contentAlso happy to have a documented JAR instead of adding a new source tree +Its not really possible for me to strip it down piece by piece +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Add ignore in +patch file attached +The wsdl for both services +The changes appeared successfully on siteIMHO we can close the issue +Rollup fixes for also into the same patch + +Searching with google hints a few problems but the culprint is hard to narrow downI do not think its a Camel problem per see but a general problem with Xerces Xalan and SUN JDK itselfCan you try working with this without Camel and try to see if you can find a solution to the problem +Ive attached the version of this codec for both patches the codec should be defined in the codecs property in adds to comment indentation of java code should be spaces not for the feedback +The cache can be invalidated when a request is fired to a wrong node + +The result of the boot image writer identity hash code method would be the identity hash code +This patch adds a unit test and fixes a bug in the previous patch +Did you download and follow Lorenzs steps to reproduce the defect +When we find that a particular operation benefits from a larger buffer it is usually best to increase just its buffer size rather than the default buffer size for all Hadoop io streamsYes +move resolved issues to so we can generate release notes +Rahul did the lions share of the work here thanks Rahul and thanks for your patience +Value hint should not overlap track or for trunk +Thanks Andy missed that addition +Any ideas? BUILD FAILEDhomeelisrc Cannot find homeelisrcsrctestaopbuild imported from homeelisrc +I have created a patch which alters the setProperty method of sothat it checks for the ERRORREPORTER property +Uploaded a patch that adds a test for the calculation when data is not compressed +I cannot immediately reproduce this on the code but passing to Charles Lee to see if he can comment on the stacktrace +Patch applied +There has been no need for this to date closing. +So it should not be effected by export lawsI will fax the software grant on monday +Corrects two syntatic errors found in anonsvn as of errors have been fixed +Your one thing is fixed in trunk +is the system you can do however is deploy spec compliant bundles that do not rely on proprietary features that are not available in jbosgi +patch looks good +Do you want me to fix the conflicts Ivan +Im making a complete review. +Opened a ticket accordingly +It looks good to me +If it is necessary we can add the test case later +bq +The startmaster code is relatively old Jrgen Lland +Hi I resolved the same using axis core RegardsSubba +Thanks for the comments +Bulk closing issues resolved in JEXL. +Fixed in r +I do not have the source code with me ATM +The dependency graph can become +Done filters must now be annotated with Filter AND implement Filter +Maybe some other stuff conflicted with +Ashley can you paste your configuration? I still havent seen any situations where this isnt a configuration problem of some sort +Would there be interest in a Java version of this tool for inclusion in theYes thatd be great +Im having trouble trying to use this snapshot repo +to trunk and +LG committing +I do not see any slow down Ning +I think another round of cleanup for is in order but this may simply be a case where no solution is perfect +all works nowI solved the internal links issue there links in the menu bar +Can you please review it +case as states above but this patch seems to get the same effect by forcing the first scan. +That is now fixedTo provide more details about the original problem if it happens again the exception message was updated +bmp files has the sequence of changes that needs to be doneThe other scenarios is difficult to reproduce through testcasePlease let me know if the solution is fineProvide your comments and any scenarios needs to be verifiedI am working on the patch and testing it +Im trying to reproduce in my test case +Test case and documentation +We have not tested it with +EL resolution seem to only be half done see related jiras +The patch though still solves the problem +Maybe I need open a new JIRA for it +Straight forward patch mostly looks goodExcept it needs minor upmerge against trunk +After a bit of refactoring jacob does not need the generated Channels and Listeners anymore. +Im to SPNEGO too +It creates a regression fix has to be reviewed in order to manage the following usecase The user marry has full access to a versionable node in the workspace collaboration and has no access to versionStorage +whenever I change a here the logic is obviously broken in regards to what requires a rebuild +But it is also handy to be able to get the mutable collection of redirects not just a copy as a list +Could you do a little bit more debugging to figure out where what conditions gets nulled out? That would help me in understanding what is going on here +could you deploy a new snapshot for us? or perhaps even a release +El segon? He vist aixUn snippet es un cacho de cdigo que realiza una funcin til +This is done for most of the code needs work on XTS +ant Then it will compile finejust talked to Chatree thanks +Patch submitted +No objections so far +shifted as agreed during the IRC release preparation meeting +See subtasks for areas where specific attention was requested subsystemMessaging subsystem subsystem module optionsSome or all of these are likely already done +A more appropriate name isRenamed this and committed the patch +Committed to trunk +The most recent file seems to have resolved the issue +Alison feel free to reopen the issue and submit a new patch +Ill have a look at that for problem persists in +Im starting to stray from my original design concept which was meant to be subtle sophisticated and conservative +Remove the FIXME from the testsuites to enable the associated test +after all this was my first incursionin jboss source +Also as mentioned there is no eviction strategy for the cached data something we worked on w and +compatibility is another matter this will still take some time +This needs to be applied AFTER the +Can you use? It has setTimes +Specs added to in eec. +Dennis Ive committed the change to the batch reader +Can you elaborate? Or even better submitting a patch would be even more helpful +It also discusses the configuration directory flag +Im reopening it so that we dont get a duplicate JIRA task +Dumped content type existence check +Sorry about that +The same change needs to be made to the ant run sample output in the README for +You can always implement a traditional static proxy that implements such an interface and delegates to some setup work in the Ill do +To discuss camel casing of module names and also module option names +Java already made a trade off not to return the error code from mkdirsrename and similar as it would be hard to achieve this in a consistent way +Changed name of test as per Teds comment +There would be for the projects replace the pom for project +Reviewed by Robbie +thanks unbelivable that I didnt find this before +Ive verified mvn clean install +I am starting to suscpect garbage collector has to do something with thisI have not tried because it seems then I will have to adapt my service implementations to use async responses +Unassigning to be picked up by someone else +Committed revision +Wait I missed that you did not check the grant box on the files you uploaded +Sending assemblySending assemblysrcmainDeleting managementSending Transmitting file dataCommitted revision . +Apologies +I hope you get this resolved yourself with the hints of +Resubmitting to test the same patch +The patch doesnt apply to trunk unmarking as available. +Appreciate any comments +Closing as wont fix +I think this kind of feature is best suited to Spring Integration so marking as Wont Fix for now. +Ilene Seeleman to review fix in this patch is now included in a new patch that I added into +I think can cause corruption during a btree split +Thanks +I havent changed to using el tld so I dont know why I would be getting this error +Why not web service REST apiBecause custom proprietary interfaces are evil and are a calling card of NIHId rather change the things we are looking at by changing a few parameters in our JMX clients than deal with Yet Another REST Interface +So just implement hasUrl and list as the subclass of that class did in? removeUrl is also relevant but not critical +Moved example classes to directory of their own under the client dir +Created because this is a more general issue and might be solvable in a different way. +days should be fine to catch up with their emails and start discussions +The deafening silence and lack of whinges convinces me this must work well as required! +No QA required so not creating QA sub doc and QA needed closing it +Looks good thanks + for setting proper fix proper fix version +Patch after svn copying HitHits from lucene branch +BULK EDIT These issues are open to be picked up +Right now the focus should be on getting out with minimal code change +I found the +Done. +can you update the ESB Admin Guide with this please? +Only one concern is before this patch for all the data import block size being used was value of +WellThe code in util is Java compatible even though the use of Override annotations means that it needs a java compiler +Please let me know what further information you need I have provided answers to all questions asked so far +The code moved toFile xercescinternalLine the DTD grammar is stored using the system id we need to make sure that the system id is resolved properly +But maven can +Ive fixed Cargo accordingly +The data on disk is now useless and the Right Thing is for the agent to retransmit to a different collectorCertainly HDFS improvements would help reduce this problem +Ie +just committed this to trunk! Thanks Ravi for helping with the reviews +The original request was for the same url but for http on port Thanks +Everything in srcjavaorgapacheluceneanalysis has been copied from which was originally downloaded from CIIR so if that can possibly be loaded on demand then it should be fairly straightforward to include support for this stemmer in Solr. +Thanks! +changing to enhancement as this really isnt a bug +Its used to create a JAXB context for use in Java to WSDL runtime schema generation +Appears to be because the Input proxy is not being used in Java +Could you say more about the structure of this new folder and how you intend to use it in kmeans +A google search for enunciate operation alias doesnt really help +bq +Resolved in r. +Cant put port in there as its not available + +seems a bit +If you replace the version of web console that comes with the Sling release the configuartion screen will workHowever this might introduce other problems as its a SNAPSHOT jarThis was built from trunk noticed that accessing the config manager by clicking on the wrench icon next to a component works fine its just the link in the top menu which fails +Ill leave this issue open a few days unless someone else closes it unsure of how you do that around here +Might be nice to make the cleanup an option +nor is it usable by JDKMay need to retrotranslate activeio so that activemq can use it +As such no binaries are included in the SVN repository +fixing javadoc patch adds a property with the name of the current namedOuput to the used to create theUsing this property custom can obtain the name of the namedOutput being created and do any nameoutput specific configuration +The project is WSCOMMONS +The original implementation made a mistake to use as reference to hdfs +There is not really much room in store to store metadata aboutthe columns and no existing interfaces to return this metadata back to the callerLanguage at compile time simply uses the existing catalog information to compile a plan which returnsthe appropriate columns and uses the appropriate hidden columns for a select +Attachment has been added with description updated patch in an attempt to address think I might have missed something on the mailing lists explaining these features but Peter can you explain what these new properties do? What they offer that its in the spec +Good suggestion but I wonder if some callers expect the password to be there +The cons of this approach is that requests that have already been assigned to the blocked processor would still suffer from the delay but future requests would be assigned to other processors +Perhaps we can merge this functionality with andor such that Solr utilizes it for reader opening +Extract it into the samplesdatebug folder then run ant server and ant client similar to the other examples +IMHO it should not be limited to the session + +Having a lower bound of on might not be efficient as it opens the possibility for map reexecution on transient errors three different reducers reporting error when the serving task tracker had a transient problem could lead to of the map +Then we vote you in as a committer +Agree with Robert on the additional context flag that would make things most flexible +The common use case is to create state per proxy created which is controlled by the singleton settings of and the target object +This is tracked in fails occasionally in trunk cool +I think I uploaded it under the ASF licenseThe behavior youre describing sound correct +This patch fixes the remaining failing tests and correctly handles compressed messagesThis patch is ready for review +Could it be that XML Schema and binding file are sort of out of sync +Fixed +All Java put back in plus more with some of the changesFixed the static int counter +And with bulk loading restore is also faster +Note if instead of performing an install in step I do a Help Check for updates I also end up with Beta being installed wo seeing your issueSo +Test was missed I have included it here +Thanks for having a look Willem +Roland can you please attach an XML schema that highlights the issue at hand +YesI verified patch in Xalan Java and it works goodThank you very much +This will be hugely helpful to us +Works well but replace the default Hotspots for +Want to do it step by step +Thanks Uwe +The fields that are added using these statements ALTER TABLE cloud +Thanks for the patch Dhruba +code reviewThis is just a test fix +Explicitly kill the active rpcs when a connection is broken +Thanks LohitBTW I didnt introduce javac compiler warnings see +Committed revision +Forgot to mention that the cluster was run with the natvie compression lib +Just going to leave this bug open for a little longer to make sure that there are no other problems in this area +This was fixed by changing the url encoding state logic in portlets specifically I believe +Where would that be to specify the debug to be enabled with krbtrue is passed to the JVM +This patch makes more of an effort to determine acceptable module values where filesystem sourced module are found to not include +flavio can you explain the race condition so that a reviewer would know what the problem is and it would be easier for him to review the patch +I assume this is releated to? +Have you yet had a chance to test it on a large cluster +so is this ready or not +Field evaluation is done at insert time any globals that are relied on must be set prior to the insert +I have been staring at it for a long time and cant figure out what is wrong so am hoping Bryan or other protocol expert can take a quick look and see perhaps something I have not thus farI will attach the repro for the protocol error +The underlying were not using some new API when they should have and the UI was ignoring it in very specific cases +batch transition resolvedfixed bugs to closedfixed +Your unit test verifies half file has this issueYes I can reproduce this in the test environment but it does not mimic the requests coming from upper layers +I see you point Ben +A suggetion I guess you should allow people to reopen issues +This is an updated version of Michals patch +Ill try writing one myself and see how that goes +Its very for me so Im even ready to build an intermediate version from your SVN not waiting for the official release +Since this is a performance improvement and not a functional improvement and we are implementing a new query execution path it is inevitable that there will be code growth +Moving to committed +Patch looks good +I am partially done with run of with more heap size and different log level setting +I am attaching a patch that provides non singleton versions for Cache and and also an interface for thoseUnder the sheets Cache and will use the non sigleton versions and singletons will be just a Facade to themPlease have a look at the patch and tell me what you think +Good point +Is there any intention to fix this +John this might be fixed by the error rework +Thanks Arun! +Added the Assert messages Cos requested and trigger patch submission +If its ready Ill commit the patch +Thanks +Garys busy enjoying himself so Im gonna stick this one to Bailey +Results of running Bug on linux. +This was true in CLI for every occurrence of F +How do we resolve this and move forward? Sean submitted a patch for this so lets go with that unless someone wants to submit an alternative patch or has a good argument for why the status quo is better +Im trying to crawl between andBut does not finish +Im happy to make the change if you agree +Ah thanks! Never thought about that +Done. +The other managers are not forced to null so it seems ok to keep it this way +about setting JAVAHOME +SVN repo now is fineThanks +I had some doubts when creating the jira but created it anyway saying a documentation component +BUILDING part of the Release Documentation please review and Ill incorporate your inputs +Pages in SBS are designed to have one URL each using the format +Uploaded in older +Fix it with for branch and for trunkDue to JIRA the has been added to validate the ejb deployment plan +the testcases are included in this patch which fixes the added javadoc to and even they are +Closing resolved issues for already released versions +for the review +jasondavies suggested to change the header to textcsv charset which solves the issue +Resolving this one for RC +Registering all softrefed threadlocals for cleanup on exit would result in more hard references and no guarantee that the exit hooks would get firedIll see if Wayne has any thoughts +What is the output that you see for the test? Perhaps I can glean something from thatThat said I wonder if derbyStress should be run as part of since it is supposed to run with +The patch is synced in to fuse repo I can resolve this issue. +FYI I asked Sebastiano and he does not have plans nor intention to change the dsiutils license from LGPL to ASLSo the only option left is to remove the dsiutils dependency +Yes thanks Marcel thats the paragraph Im refering toThe only question might be if you have a two receivers A and B for an event +Verified for of a bulk update all resolutions changed to done please review history to original resolution type +Today i had the same problems here +Alternate solution we could just poke set via JMX too +It shall be fixed in in ER. +the patch with small modifications +BTW do you think this should be noted in the release notes for people who are already trying out the new security code? Im not sure how big an impact the changed property type will causebig +and root servers startcode is not recorded in ZK node So I think it is a little disordered for solving the issue +As long as the method can be executed what if anything you do with the method is up to you +While it is a good idea to implement as much functionality as possible perhaps this one issue should be about queries +Applied the patch built the site and committed to cvs +Well maybe not completely because it makes no sense to ignore a query shape so what would it mean to set this to IGNORE and the query shape is incompatible? So maybe no IGNORE? So default FAIL is your preference +There are way to many variables involved +I just committed this +Closed in favour of which is more specific. +A uniform distribution doesnt necessarily mean the old code would generate uniform splits since it leaves fewer options for the last few nodes being processed +Added label attribute for to orderingList +I was thinking of skipping the process but an assert for those versions with the would be good too +Konstantin you are probably right I forgot to click on Target link + +can be this merged to +Though again knock yourselves out folks +Yeah it was on purpose +So when it tries to update in ROOT he is not yet onlineThe behaviour in TRUNK is different +Change asyncexec to syncexec in can hang up eclipse so please take this into account +So we should consider the upgrade when has more stable release +Buit we already doing this with so at least this is consistent +Maven +Noted deleting replicas from blocks is a low priority task in namenode +Moved patches from +servicemix Jboss deployer patch attached +Jehudi is doing the right steps and still seeing this will take a look tonight +So there is no version number I can tell just that it answers standard XML +Let us know if you see further symptomswe did see CLOSE build up until the system stopped respondingI will try to set up a test later today but I dont have any experience developing Axis so I dont know where I will end upBrecht +updated documentation +This no longer happens in r +This sounds like a bad practice to me +I have attached two updated patches incorporating both of Alexeys comments +A small error in the examle aboveThe start and stop methods of the are startDelivery and stopDelivery +The replication steps are simpler Load Processes tab in Explorer Sort Table by clicking on any heading Click Start Process or Start Form Nothing HappensThe issue will be with the events triggers not being rebound after a sort +Committed the test cases +I have not been following your progress closely but was wondering if you could confirm that all functionality being added is part of the SQL Standard or do you plan some Derby specific syntaxThanksKathey +looking at the commits it seem we changed to init the classpath to use the WTP runtime ones insteadWhat does that mean for the existing classpath containers? Are they deprecated? Which one are deprecated and could we please add a unittest to ensure we dont accidently remove the support for them +The issue is not the issue is that your ou mapping in your ldap config is wrongPlease use the mailing list to discuss that with other users about your individual configuration +Same with whitespace with boolean declared as I am going to commit +I have checked my own which doesnt also need and it appears to work fine +Attaching updated patch which includes the change to I left out of the other patch +It ran fine for tests test and testSynchronized are failing for all test classes that extend when I try to run the scriptIm running in a Linux environment using the +You may have limited freedom to create specific users if you depend on external user authentication +Swapping Affected Fix Versions as were not going backwards +Just to be clearIve just tried out the command and it fails for me too but Im pretty sure it is not related to Martijns changes. +Updated fix this jira +Throw in +But not just for the leaf then +Further messages for the same module publication would be sent by the plugin but one of them would fail to be delivered because the connection would be closed +Seems it cannot reproduce after is fixed +Ill attach jstack output for test and. +reopening for discussion before releasing was there to facilitate using wicket validators outside of wicket or testing them +One of these was our pending entries page +So no check needed or such +Probably the best way to go +Thanks a lot for the contribution Hank +Updated as committed only to Trunk and and not in +Patch for reverted at r please verify if the issue still exists thanks +diff signifying the changes madeThis draft is a bit of a mix between calling this version and because Ive not yet modified the actual version not in JIRA nor in the source +This actually already resolved when using SD JPA and SD +looks fine +I verified that this works correctly by applying my fair scheduler queue metrics changes and observing that both the superclass and subclass metrics appear in a cluster +If this is not the case this is a big problem +Jarcec Btw Youre right this usage is not documented and I believe that its a bug that we should fix + +So the truth is that Spring does not support transaction isolation with JPA +In that case I think we should use the first node whose last is no less than the id of the snapshot +You are rightThis is the reason and the solutionThanks! +The does validate that the context object name does match the domain so I dont see how the jsr mbeans can affect this after all +The patch did not add any new dependencies to the plugin +I am fine with the log messages I just would also like to deprecate the term optimize and change to forceMerge +Merged to the branch in revision . +Testing attached +ok fixed +moved to m if deemed local specific poblem resolve against M again +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +From the discussion were fine with leaving thisIfwhen we introduce the manifest well either support the old approach or have an upgrade path +Needs further exploration +Also put the schema in the load +Thanks +Please see bug as well for a discussion wrt derserializing objects that have not been loaded yet +Please give a try. +It appears that we are not cleaning out all the possible cells when a region is deleted +This is a problem with JDKPlease use FOP transition resolvedfixed bugs to closedfixed +Ok +It is an already existing limitation to not distinguish and. + +So having bundleinstall and bundlelist in the same role doesnt really help I think. +was designed to handle this problem +It seemed trivial to get it to a separate patch +That is right Arpit +Trying hudson +It makes more sense to me to separate these two operations which are in fact different semantic operations with different purposes +After fixing some problems in my deployment environment that resulted in the wrong version of the Castor JAR being used I was not able to reproduce this a fork in it! +trunk +I have tested this +Attaching mostly working version of change for safekeeping +Created for common fields annotations settersPlease let me know if I did get your pointThanks +r changed the compiler to interpret MAINCLASS as a request to not look for a main methodBTW the compiler also raised the same exception on MAINCLASS so something appears to have changed in how XDT invokes the compiler + Fix xff confusion +I think the best policy for us is to look at bugs and only emulate the ones whose behavior might be depended on staying the same +But still I think having a manual override here is useful for advanced operators who need to use it in an extenuating circumstanceAs I said above Im OK giving it a scarier name andor making it prompt for confirmation upon use with a scary warning message +The patch Im submitting is intended to be a change to srcwarcss hope this is more helpful than annoying. +Verified in Developer StudioVersion Build id Build date +Additionally A table compress is supposed to update the statistics as well which was added in but I found that running table compress did not create this missing statistics +So far it looks like the LN patch is an improvement +Give me a bit of time to review the code and Ill include debugging suggestions shortly +I will look at writing tests +For the fields that differ in values Show data of two contacts in Tabular two section form +Rather we change the behaviour of the existing option to always send compressed large messages as regulars ones whenever version +Is there any workaround til this is fixed +Im adding the ability for the resolver to call back to the codec +But I dont understand why it is out of scope +The defaultlength of does not look reasonable for CLOB or BLOB +This means this problem should go away +This issue is duplicate of. +Pull requests have been merged into branch +File is having less replicas than expected and not yet replicated +Please post away. +I am seeing this same error for a project that is not a project +Piggybacking in an existing testcase +I just tested this even on my slow machine and I cant reproduce it +Its a contrib srccontribec +Having the new client fixed for Alpha +Will test this laterDo you want to fix the rest of the tests and remove the keep method? At least this method should be hidden by a accessor or if not possible +Bulk close after release of +I will discuss with Daryn and let him file a Jira for the improvementIn any case would you retry the patch +a moved region could be snapshotted by the newly assigned regionDid you mean by the newly assigned region server +I will look into the culprits and try to find out what triggers the problem +Accepting +RightThe setup should only be done once then all integration tests should be executed then +This looks good and would be useful to have +Thanks +In general this maps to domain objects for which many instances will be created +Decompiled the jar and the fix is not included +thanks pat. +patch leaves only the latest aspectj version +this patch adds a test +Please verify +I have retested the functionality in the current version + +Seems like that when those threads are started later due to scheduling mechanism of threads then this happens +Thanks for looking guys +An Subtract should be of r +Thanks N +I like the blue used there +Closing any recent changes are tracked here +Thanks Bruno! +Addendum to the previous comment should have mentioned that there are potential problems with both Sources and Results +This is with maven Beta and CVS head of jellyI think it could be closedEric need to run maven jar in the root jelly directory first +Applied to branch too +Are you wanting to add this as a project? I am already mentoring one project and timework restrictions mean that I cant step up to take on another mentoring role +limit on bytecode size is a limitation of the JVMI dont have any good advice for youYou could remove the facet from the definition for the purpose of generating the xmlbeans +Large scale performance tests including NN benmarks and sort are scheduled to run today +Well the of where to place the exceptions is not +it was also ported to the branch but should not have been since that was a release branch that was already cut +hiMartin OberhuberCould you give me a url about nightly build? I checked out the latest code builded it and then tested it +This gives us the hooks to do log and console out harvesting for Chukwa which lets us close and probably +Need to check this next week +going to schedule for +There is one exception I prefer not to use terminology for users +If a new row is appended the cells inherit the style of the cell +Vinod Ive verified it +is there something i can do to help? can i just try to update the queries in the report directly? could it be just a semicolon at the end +FYI With snapshot Im now seeing for the workaround which makes the work around unusable for that versionThe instance of type class with oid null no longer exists in the data store +This makes it fairly simply to detect if a posting has been written because if it hasnt itll be or some other d value +Adds configurable upper bound to link field in +Feel free to close it +Thanks! +Assigning this to you in case you want to take a look at it together with +looks good to me +We should add a note on the behaviour to the +The fix for this is to introduce new behaviour intoAt present does not understand HTTP formats so does not work with SOAP attachments +Defer all open issues to +Fixed by adding support for synthetic ears with no ear file just external modules in the vendor planSending modulesassemblySending modulesassemblysrcplanSending modulessrcjavaorgapachegeronimojeedeploymentSending modulessrcschemaTransmitting file dataCommitted revision . +xml the context of the needs to include the list of allocatorsThis will ensure that when this component is loaded by Spring the allocators gets initializedContained in branches masterContained in no this record since is now released. +Thanks! +Thanks +I was thinking about the that Spring has the infrastructure for converting all kinds of things to and from Strings I wonder if theres any point having a more extensive schema to for example store integers in INT columns thoughts for the schema would be to allow a hierarchy of would have thought database triggers would be the ideal way to support the dynamic reloadable functionality +Right the current code allows only one definition per method meaning only one Cacheable and +Second patch also checks for missing values +The code I was referring to was actually using digester so I dont think the provider method is there +Is this patch made for for java or for for javaCan Spring richclient be compiled with for java and still run on a JRE with glazedlists for java +Hi will somebody try thisGood luck +Test application from EUT +This issue only happens to the migrated volumes which is missing format information in db +Rafael can you fix this as a priority and do a releaseThanks +Patch applied +JakeCan you test this against the latest codebase set to be released shortly + on patch +There are hundreds or thousands of people with these strings in their configuration files or code or whatever +Any objections to this patch? If not Ill pop in the trunk in a few days from now +nodetool compact with sizeTiered +From another mail by JoergDo whatever you want with code I thereby license it under the +So I catch the expection close the socket and then rethrow +Patch applied in revision +Closed upon release of Hadoop. +You can always go to the Jira website and log a bug there +bq +Ive verified the fix is working for us +I assume we are will still be supporting just that installation instructions should refer to where possible +No need for more than the existing updated. +This Update implements changes suggested by SiRegardsAnil +Patch for the version that includes optional Mojo parameter componentTypes with default values jar bundle on. +Now available in trunk but unrelated tests are failing so snapshot is not available yet +One thought I wonder if fixing the indexing interval so its actually rather than default helped make this issue worse +As it has to be set to in combination with setting close for keepalive to disabled chunked encoding for response +From what I discovered so far the first one ideally can be achieved by applying tiling strategy +There must be something in your environment that is causing our build process to take a left turn +Committed to trunk and relevant branches +But I have created the patch again using svn diff command instead of my IDE +Thanks Dave for the patch and Dag and Lily for reviewing and testing the patch +The closest I come is when clicking on the editor the page scrolls up to set focus at the very top of the editor +out of time moved to. +Oh implicit in the task name is the fact that if theres a chance in SI therell need to be a reciprocal change within the API. +verified on +This is a mapreduce application specific setting and should stay in +Ive successfully built a doc using DITA Open Toolkit +As source of ddlgen is not yet integrated with main castor repository we dont need changelog and yet +Thanks! Committed +Teiid Beta client code contained a bad reference to some logging jar preventing to initialize +This one is older but the other has a better description +moving description from environment to opened for the client issue +Would probably need a complete back trace of all the threads here to see what the issue is as it looks like in the first comment the is blocked on something not related to the Close transport task runner +Thanks +This should be very easy to fix +Can you give some more context how to reproduce? How do you configure a spring databinding for an service + can you please elaborate on the use case for thisIt occurred to me that in all of that massive message I still didnt mention why we need some of our plugins deployed unpacked +The solution here is correct we dont need to reopen it because its test found a separate unrelated bugany shit that limits your values to smaller than needs its own checks to fit throwing UOE because its choosing not to support totally legitimate values from the analyzer. +Thanks for pointing that out Bobby! I had no idea that even existed +It was very helpful to Armys comments in that issue and in the older +Do you know if there is a way to make the plugin metadata and the site pay attention to the jdocs in artifactitem? Ie it doesnt pay attention to the required fields and doesnt generate any site info for it +This is clearly working for other issue is confirmed too +that fixed the issue i saw +This seems to be the reason +bulk close of all resolved issues prior to release +Timing splendid +Ive applied a fix to CVS +Javascript is confined to HTML rendering and the form widgets are supposed to be rendering format agnostic +So if you are having issues with that class then please open a new issue +looks is ignored +just committed this +Oh yes +It passes the tests would you please help to review it +a better way to go would be that we make the timeout part of Message obj so that when adding callback in we can pass in proper timeout to and also in we can drop the message based on its own timeout +This should then also allow jcrspi and to be bundled cleanly +Asking if the OP can try to clear the cache +Ok just realized this is affecting the tck as well +It seems intuitive to me to be able to do so in function args because theyre essentially strings being passed in +Running it through Hudson +Its really an error and with this issue Teiid will not make the VDB active +Listing job instances works somewhat +Make all regular substitutions and depmap adjustments that you are used to +Committed revision +Ill close this for the time being since there doesnt seem to be an urgent need for +set +examples for JBDS +Note that the fix involves minor changes to Eclipse files that will remove capability +It doesnt sound like a issue +Complete memory histogram is attached to the issue +we will switch to when possible +I was expecting a file with a +IoannisThis is the first time Ive done this +Thanks for the patch +Issue may already be resolved may be invalid or may never be fixed +In my opinion using the build requires the user to know the details of the subtypes to use effectively +Here is the jar +What is the lowest value for producer retries that works for partitions and producer threads +I get a similar error on reason +We will definitely add it to Camel +Please double check +AdrianYour patch is in trunk rev +Verified +Close out issues. +this patch only adds actual union code generation for Java right +Clicking Submit Patch so this shows up in the patch review queue +Verified in +Please let me know if there are any commentssuggestions +pushed +Unfortunately I didnt see anything else +John I tried this patch and it doesnt seem to have any negative effects either in standalone or clustermode +Retry after revert of wal compression is not because of this test and passes locally + is already released +why should we serve static resources using a dynamic class and incur additional CPU power and memory? The application server already has optimized code to serve static resources +Thank you very much! +Closing as no longer valid +If you have anything else please let me know +This patch ignores the following assertions for now the count of fds in the count of nodes in These changes will be reverted back in before is released +applied to the module +Thereforethe most likely reason for the problematic behaviour seems to be anunsupported palette format or an invalid or mangled JPEG headerThere isnt much more to analyze without one of the images at hand +Merged into x r after release. +Ill fix it when I fix the multiple fls not being treated right in distrib search around that same code there is still logic that expects this. +I was going to request a port but I should have kept this open until thenI also agree on the comments about the tests for this +Fixed in the libraries +Thanks to Amareshwari for helping to port to the new context objects api +Patch looks goodCan you run and ant test and upload the results +This patch adds the configuration parameter +Emanuel it would be nice if after calling stop or other methods that could change the deployments state if the deploymentManager automatically updated the deployments state in the managementView +does have a field called gatewayMessage and the purpose of the service is to save a not a +Im going to optimistically mark this as resolved +Yea thats why we use the target version. +Im proposing to guard against an in to prevent it from bubbling upattaching patch +We dont need separate Jira issues for adding the visitor pattern to the screen widgets +Yeah but there are more calls to this method which are not checked too not only the author call +This will not be fixed in and Daryn is going to be creating new jira to fix the issue he intended to fix. +If we arent going to fix and Faceting then perhaps we shoudl consider removing completely from lucenePlease try to give the branch some time +running +Passed regressions +This feature has been available in Cocoon for ages storing Template objects in a LRU cache and reloading stylesheets as needed when they have changedWriting a new Cocoon is fine but it should learn from the old one and reuse some of it +I ran a svn cleanup so the data should get updated on the next svn update add svn unlock commands to in apsiteThis would stop it recurring +Remove cast to bits of doclen +Ill run the test more times +Can you attach just one say for instance jaJP so that I can confirm it works right hereIf theres a way to change the locale for forrest alone that would be a better solution +The patch added pkg solr and for yumaptzypper +This is a bug in Hibernate and needs to be resolved guys +I think it will be but numbers are better than hopeSo I wont include deletes in the patch just yet +Since there are many other ways to process the message using header mediator property mediator and so on +I also agree with Dag that we probably dont want to fix the existing SQL violation for INSERT because of the compatibility concerns +Committed to trunk in revision +Tested locally with our class resolver and everything works fine afterward +I want the client test to fail if the RM behavior changes +also needs close of resolved issues. +Trivial patch +If there is a problem with this please contact me +monitoring hard coded text in jsp files are extracted to external resource bundles and translation are added +Tested. +The API is different in xMartijn could you shed some more light on this one? Like is there something in trunks test infra that we should back port to? Maybe we should open a separate issue for this +Stackbq +Thanks +libhomeedwardworkspacebin +Thanks for feedback +We should have a resolution to choose or before we proceed +Sorry +for the patch +information Daniel and other INFRA FolksMy name is Cameron Goodale and I will be the main point of contact to try and get the svn repo migrated to the incubator +These seem like they might be related. +SPRINGCLEANING We can reopen if necessary. +Come the next major version I would suggest the default for this option change to indicating that by default old names would not be accepted +This is to avoid in removeBlock in case an exception is thrown when intializing field checksum +Can some of you create the patchfollowing my description belowGul +Hey dont get nervouspI agree generated code should be as cleardocumented as possible but it should not be under control of QA tools +And create will filter out all the null onesIll post a patch soon +Unless this can be shown to be really a jboss problem this will be closed as rejected since we do not rely on any of the frameworks that seem to be causing the problem + +Patch committed in revision +bq +This is the new google base application to be insert into the specialpurpose folder +its necessary to translate the description of button component provided by dev team and to add it to the guide +The total runtime for goes from seconds to seconds with test enabled +Is this the right ticket to take on fixing that or should we create a different JIRA +I ran into some problems getting the test to work against mock Accumulo +Committed to and trunk. +This would sacrifice the write performance of dynamicComposite column family but at least it allows the cassandra to perform the validation before the actual write +but there arent any errors to be seenFirefox works fine +The difference is just its not yet stored in CVS +As per my last comment +Thanks Suresh for the reviewing +I have reviewed the patches and even though I dont think I was the comitter that Lawrence was referring to in this patch is not consistent with a chat I had with one of the committers I do agree with the opinion that having substitution group members show up alongside the head element in content models is the case and its not worth the memory penalty of keeping track when substitution groups were usedSo I would propose we dont worry about this case and we fix the more common case where you have substitution groups but members of the substitution group cant appear in the same content model with the head +Needs some more thought +Steve you need to download the src distribution the binary distribution only contains testssamples +I tried to apply Todds path got a unit test error on user defined key cooperator +This sounds fine for nowOn a related question if the writer of the ledger dies a sudden death how long will it take after that event before the reader can read all the committed entries from that ledger +I like the idea of cleanup like checked all code and found that these opcodes are not used anymore and marked as deprecated +Make the patch unique so that it does not depend on the first patchSorry Jacques did not respond in a timely manner because me is not Francis +He was in the same concall as I was and heard the same answer given byAll we can do is wait and see how it progresses. +And this failure is because of the same wierd on some core classes thats been plaguing other test runs this time on I could run successfully +out the connection handle BEFORE notifying listenersAt time of the callback the must be disassociated from its previous logical connectionIf not there is a risk that the PooledConnection could be returned to the pool ready for pickup by a new thread +great +looks ve committed this and merged to +I just stepped in to issue and it would be nice to quickly have a look in the source code what is going on +Updated patch after fixing warnings on Windows this patch fails on cpu box +Hold on +Sorry I meant the +Falcon always removes this theme thus the problem +Updated to revision . +Appreciate it +Be aware that mMany changes were made to the language and whilst most scripts will still work you may find that some will need minor changes +osgiuninstall two +Not an issue this was due to other configuration that was incorrect +I have a workspace created with N and with that build it doesnt happen +A sample bundle which simulates a servlet which throws an exception when the data received reaches a configured maximum +Go to Facility Inventory tab +reproduce and didnt find anything worth commenting for now +AlexeyI reported this issue under the JSF component sorry if this it is not the appropriate one +Andrea Could you please attach a complete XML file and XSLT file if any +Ive attached a patch that uses return instead of exit more consistently and returns values that better represent the status and Ive also attached a transcript of my test sequence which I have read through and verified that the output was always what I thought it should beI have only made changes for the use case +Why does jira insist on assigning bugs to me when I comment on them? Assigning it back to you Kris +unit tests is always welcome +Some requirements Allow permission specification for deployments +I need to file this as a dependency requirement for our QA approval. +The important thing is just that the videos are initialized with wrapping layout in the area where there is no whiteboard +I make two patches for and trunkUnit testing has check Thanks +is legacy for any tables +There should be a scheduler the does a check for older then hour and delete them +you can test this issue once you have a copy of +This would be a huge help in terms of usability +A deadlock has been encouterred +Did you see that behavior on an iteration list of storiesAt least in the upcoming the list of stories does not page just for that very reason +Add a line waitforPageload to the testjmsresource cases +There are already versions of it in libwsm and libnetui and Ill be introducing a controls dependency soon as well +The plan is to fix this in AS the AS download page there appears to be no mention of a or release +Done! +I have committed this to and above +Will put up a v patch if it can pass tests consecutively +Not a major version bump but not appropriate for a patch release +I do not think that we will reach consensus on this issue before we need to generate the next release candidate +It passes all tests against trunk except for ttest which I believe is failing against trunk for other users +Can you add export JAVAOPTSverbose and see where the XML apis are being loaded from +Manually without integration test +Use the official EPSG code for the Google projection closing all resolved issue that have not gotten any more feedback or comment in the last month +and it works so well we tend to overuse it +I am calling the same method as for the interface but really annotations on the Impl should probably be BINDING operationsAlso at the moment because the class annotations are encountered after the method annotations any OPERATION annotations defined at class level will not be applied at the method level like they are for the interfaceSo this requires some more thought and a bit of refactoring probably +IMO we should move generated code out of the main src tree too but as I have said before I am not a java build guru +to Scott since this is UCL usage issue. +Confirmed on trunk +Was seen running a patch since abandoned. +The problem got resolved with the above solution. +Should I cancel patch and +Isnt it better to use scopestep for the item reader +Committed revision +The is likely a mistake as that release didnt exist when this issue was created +So this threshold will be independent of disks capacity +This patch is tested in a clean tomcat and the new structure is working for local albums. +Attached a project with simple application exposing the issue +The gateway should be used for allowing postprocessing a file if that is s merely a rename kudos to Alex for the whatchamacallit +It builds with the release as well +Adding some SNMP and HTTP and Memcache isnt related to whether or not all tasks can be accomplished with the same protocol which is what I was afterHow Postgres does things notwithstanding I dont see it as a feature to put everything under one interface like this +should be on drools download page in hours +Hi Akiafter reducing the maximum heap size to MB I was able to load the heap dump into +This issue has been last updated more than years ago has no assignee affects an old version of Tapestry that is not actively developed anymore and is therefore prone to be in the near futureIf the issue still persists with the most recent development preview of Tapestry please update it as soon as possible +works in +Therefore I will first factor out the new interface from to match the architecture of the reflection package +Can you check in a more recent version of Spring is the latest production and is the latest milestone +Error still there in +Attaching reproduction for this issue +Im getting a clean build now in m getting a clean build as well. +TODO Implement Externalizable More unit tests? How to write a unit test for multiple versions +Thanks for the quick turnaround +I consider rev an API break +Can you check if this still exists in trunk and if so prepare a patch against trunk? Thanks +Again not intended for commit just yetI also updated the failover test case to ensure it works with this annotation facility which it does +Danno applied the patch I didnt apply the patch based on your feedback +Updated to Jetty which has the latest Web Socket updates + seems the only approach that would hold up +See and have committed this +smoke tests passed for with on winlnx IA +ugh pressed the wrong button and assigned it to myself I added Sujee as a contributor reassigned this jira +I am closing this as I cannot reproduce the problem +Thow alot of messages at the topic using our sample programs ant producer Dtopictrue Dmax +Thanks. +Second patch with required changes for SWF working +Whats the motivation for changing it now anywayOne challenge with implementing this will be how to do upgrades from existing clusters which already have block randomly distributed across the valid range +At the very worst it seems we could turn this into a finer level of trace rather than removing it entirely +committed to both trunk and branch +Given that Im not sure we have a real alternative to juggling the phase for at least the findbugs call +MODULESPATH modules +The last error output is there only to illustrate that Hibernate Search requires that property +So why not providing a simple boolean switch +bq +Use case is when you would like to run a query against blur and retrieve all the results via a job input +Maybe what you really want is a notion of a task being purged from the system +Fails with IBM and Harmony Javas works with old is still specified in set so http services still fail + +Im going to resolve this one +changes to use the new map and replicas +projectprojectproject shown in the example from Kristian. +is it necessary or desirable to query out the rowcolumntimestamp data in this fashion using the pathCan you answer the Tom Nicols questions BrianAs stated already what the old API has going for it is that if the new implementation kept it up then it could be just dropped in +Removed canned db from +I think we should put it in our docs directory +Thanks Divesh Pranay for the contribution Committed at rAshish Vijaywargiya +Add the containerId to +Please read the file for more information +Tim please try this patch to trunkAfter applying the enhancement of java the patch for java will be providedSorry for the inconvenience causedThanks a lot +then the cache will take care of evicting old ones to make room as neededsuggest making the capacity of heap sizeshould probably have a separate setting for maximum single statement size +Fixed with and which added a quick test to read and write a large number of ranges +Tested on OSX desktop and label show bold fine +for the patch +Ive attached a new version of the page and a PDF that shows it in action +Scheduling for as this is a EAP PRD requirement +This is expected behavior the web and the service run in two different runtimes and you cannot pass Ruby objects across runtimes +good stuffapplied the patch with rev +Code change of resolved shipped +To reproduce copy and paste its contents into a field in a document in Futon +Merged. +Now the following build commands are working successfully with Maven mvn clean install mvn jetspeedmvn Dtargetmin mvn jetspeedmvn Dtargetdemo +Updated the formatting of the descriptionCan you provide the CND plus code that is trying to create the nodes? Without it its very difficult to figure out where the problem is +This bug is getting old and is breaking many things +Thanks Uma I will commit it to the trunk later committed to the trunk + +But there is a race condition that will make it fail anyway failover right after the original server sending out the NEGOTIATE +How do I go about getting a Hudson account? My usual apache usernamepassword isnt working +Im not sure if its an bug or is using some way to set permissions +Is that a bugThanks thanks for the extra info Fred +means that there will be problems in concurrent tests on the same machine +New jobs created t we need these under view as well +Looks good to me +this affects all versions of patch adds the static context tests to the test case +This is the new patch which fixed the exception handling in unit test case +I am willing to introduce a context attribute that controls filtering +Patch committed +But finally we have added a new feature and all the users who use RC need to code gen again + adds the underlying functionality to the STATUS command and has a patch available +Out of the box clustering in AS is not eager any longer +Verified on revision +Should be done in there plans to do something like this in Richfaces? Itd be useful +My choices of names for although they now have a name which is clearer to me mean they diverge from what class initialiser variable names were +NoteDue to the limit of node ID has been changed to . +Closing this as a duplicate of. +A very basic quickstart showing a page with an invisible link that becomes stateful +A few approaches I have thought about are +As mentioned by Lukasz Racon checking the file size will not accomodate files which is a valid scenario +it would be nice to remove the designer schemacatalog concepts altogether as they have no real runtime representation they simply become part of the table name +I agree with that +It is jsp syntax isnt it +The problem do not exist in this revision +testtestsextensionsjava +right +Only affected site should be theExtras site which means this could even be done ifnecessary +Previously application was using default timezone for instance +Thanks Varun! +also disposition on on p documented p needs applying to trunk to +Ok I have added you to the groupHow stupid that a Space Admin can not add Users to GroupsPlease close this issue when you are happy everything is doneCheersGav +OK fixed +Please review and apply +I am out of town but will review this patch sometime before tomorrow +Seems to be unrelated to anything we can actually fix about our build process. +This bug can remain hidden in some combinations of hadoop version and OS where a partfile gets returned ahead of directories or if the temporary directory wasnt created in the first place +There is nothing wrong with the archetype +Is there a dtest for this +The extension doesnt matter but having them named makes it easier to see what Im working with in my source tree +I dont see any failed tests on the Test results pageThe patch for is mostly the same as for trunk +This is interesting indeed I vote for +Just forgot to note a similar bug in the past that targeted into the same direction this is fixed already but not in +fixed a patch with the interface for the +reopening for improper the issue will not be fixed +The issue is fixed in CVS and released as a standalone release +We added regex to enforce per test timeout +Also adds basedir to the property to make it absolute +so both of those warns are expected on failover if an ack is outstanding when failover happens +Anyway the tests are passing now +Named is a CDI annotation +The issue with explains the problem I am experiencingThank youLeonid +But this was necessary to try to load the XCAS +This is a duplicate of so that can be closed as well. +JPA Work as problem is in the way JSF integrates with Bean Validation +That is why it was done the current wayPlus the Compiler class in some of the VMs leaks memory really badly +pending jenkins +This is not a bug but a nice feature to have +Committed +Using seems to fix the issue committed in r +Go ahead and apply it +this is not a bug +for your patch +As it just confuses people and there is a lot more to test and ensure works as expected etc +Sorry I didnt commit the code earlier Now I have committed itThanks +on simplified version of first patch +Agreed. +Tast Case and mailing with Mr +Im not even sure if it still fails in? Can this be closed +Make sure your app version matches the one on iTunes Connect +all committedI wonder if you could provide some unit testsThanks +The problem is how to identify that a request is coming from outside our own + +Patches created are for the array +So its another instance of a KV stored into the wrong store +In such a case the trx can not be committed and will throw CME +If possible upgrade to AS or AS and see if that solves the issue. +Just curious are you interested in working on it +PR pushed to master +I dont know how this can happen but its happening for checks in plexus to fail earlier +I have no regression test this bug was reported by Pavel Ozhdikhin during some experiments +Carl Would it be better to separate this issue for briefness +Thanks for the reviews DougThe other day we have a short discussion about this with Nigel and Ive implemented something similar to what youre asking for in another project Im working on right now +This is the patch that Ive tested in three scenarios Using default true value of the property Overriding with false in overriding with true in +We are upstreaming into the solr contrib as well as into the solr contrib as well as into the solr contrib +Great thanks for pushing this forward Simonbq +Can you please review? andor try nightly build to see if it indeed fixes your problemthanks checked in fix is OK +See for more details. +Attaching new patch by changing testcase a little bit +seems to deal with a different subject imo +The configuration name space is global so most likely you already got a Animals under java or junit launches +Do you know the answerI will investigate Apfloat a bit more to see if theres a way to get it to do infinite precision math and see if the performance is still good. +Yeah +Would it be better to add a gettersetting setting for the data handler mapping rather than adding another argument to the constructor +Thank you for the workaround +Committed version of the classes in revision +Ok lets reject it as not a bug +I think we should continue to add features to the old API and once we are happy with that we can design a cleaner API vs trying to squeeze these features in later +I see +I will spend some time reviewing cvs trees of other eclipse plugins and come back with something OKIll be watching the dever mail list in case there are questionsJohn +Henry +Just to remember The problem still exists in uDig a sample shape file which contains an arabic field +well +for background workTest Failed +Hmmmm +Actually thats what this change did sorry for the noise. +See if Hudson will do the testIn any case Nicholas will review the changePromote for +Definitely a known issue +The attached has been verified on nighly Axis +A useful strategy is to compress the first few k and and then either compress the whole thing or not based on the ratio +I even loaded up which uses this element and everything worked correctly +on httplocalhostsolrsinglecorefieldtext in the Top Terms with the moreless links +Any objections to closing this issue +Thanks Rob +We have to think about that +They will get a deprecation warning in the logs but else everyting is as beforeB Previous Solr users who wants to upgrade config properly +The attached patch incorporates the feedback from you and Nicholas +Sometimes the code will form the tiles definition in code or so better just leave them they are harmless +Ill check it +Ah whoops youre completely right +The snapshot page has been updated +Someone talk me out of it +Ensure the table has been disabled +On a side note related to documentation we should distribute our wiki Installing and Using Qpid page as a README file +It was sloooow but produced documents in the site some point I expect that this will need to be converted to a report plugin. +is an so the will not run any request for the topic until it completes its callback +Clearly something in read is messing up the that is great news +myattribute valueThis is probably worth adding to the base Node +There is also which has a http consumer that needs a tweak. +Doh sorry did not see your comment my jira mails were still going to my previous employer mailboxTara no there is no need for a proposal for that a patch is more than enough +do you have modified notifiers list during the build of the project +Due to Tokenizers cant clean themselves up completely on close which makes this more urgent +Fixed on trunk. +Thanks for hunting this one down Alex +Could you share a minimal script that reproduces the problem? Ill test it against my fix to make sure Im solving the problem youre seeingThanks script demonstrating the use of ipv numeric address I fixed a similar issue with java on this recently looking at the c code now it seems that we are looking for the first occurrence of rather than the last +Im reviewing this patch +bq +No more stuck messages observed so far on highly loaded systems since end of January. +Seems ok to be closed. +If that is the case we should revert +Its unfriendly behaviour as i think +Ah okay I see what youre doing +It is planed to rework the concept of concurrent access to configuration objects +fixed will post a new driver on the Geronimo plugin. +improvement applied with thanksnote for the future every new file needs the apache copyright header +Interesting James thanks for the + +It is a feature request. +This report is vague and anyway does not refer to behavior of Seam itself +there is a broken image in that need some idea of what is happening with this JIRA so I can schedule docs work for What volume of material +resolve as out of date because of month no activity +r treats no type specification as anyType +In hadoop list API is added and both and fsck listcorruptfileblocks can provide most of what I want +Thanks +committed close for Solr +This patch includes the fix for the new config api and all changes suggested by UweUwe you should be able to commit it now I checked out the trunk and applied the patch myself and everything is working fine +Updated by Tom in revision +No failures or errors. +Yes please submit a PR +Patch is remarkably stale after and went in attaching a rebased patch of the original version +Im happy to review the related patches though +At the moment our user docs are integrated with the maven docs as xdocs themselves +There is no way to configure the parser to process XML that is not +Should be fixed +Fix is in spec coming +if user wants to delete more than one column family? How would he she formulate through one requestWhen the type is FAMILY we will delete all the families coming as part of the scan result +should be marked as duplicate but keeping separate as the other was closed and not backported to +Attached test patch +You should see a error mesage if so. +Please let me know if attaching the TTF files used to generate this PDF would prove helpful +this patch attach to JIRA +Thanks still see this issue in +The links are to other issues that may affect the implementation and should at least be considered +gitignore but otherwise we do since they should not be committed by default +Sorry +See of the XSL FO spec Property A value of left or right is converted to the relative value as specified in the property definition +callback site removed cms config removed redirects setup for incubatorcallback and incubatorcordova requests to go to done +Ive never seen D execute the command on unix +Hopefully this is the last patch +If the same code is in use here and there it might get fixed automagically +This number goes down further if we have more regionsThe client should be changed to maintain at most one subscription channel to each hub +But I think transport like SMTP UDP TCP should always configured at the service levelif you configure the UDP in global level how you going to dispatch the serviceAs I remember earlier Axis smtp transport uses one global address and used an header to dispatch service +We are unable to reproduce this with latest code +this feature would also be useful for the file +I suspect wed be ok with migrating the existing code out of Pig into another project if that would be helpful +Sorry I was looking at an older version of +bulk close of resolved issues as part of JIRA housekeeping for release. +looks pretty good +Oliver Gierke I have tested your test case It also indicates could be found +Opening a bug for that +I have created a request to upgrade HSQLDB in +Damien you are welcome to massage the patch into shape if youd like +I updated the release version to +Maybe it might guess it wont work for general method replacement we want to make objects using lookup methods serializable at all? we finally conclude that lookup methods and method replacers are not meant to be +Thats what I feared +Any new requirementsbugs can be raised on separate Sent a test case for Firebird just now to your email ID +I am experimenting with my own nio implementation of Zookeeper +yeah +Patch applies atop and exposes testPreprocessor configuration as property for +build is green +This is fine since I plan to build against Eclipse Ganymede first then move you up to depending on Eclipse Helios +Changes in mrr sleep job to demo usage used for manual testing +Applied the suggested fix on trunk +uploaded which works with the +thanks for a quick reply +Fixed this in our wrapper class +for hudson +There seems to be discussion about whether this is a bug or a feature its a bug plain and simple either in implementation or designPlease please fix this +i rather keep then separated +I think having a single link to hudson on a developer oriented page of the website and making the nightly docs findable through hudson only is plenty +This is not an issue for as the microcontainer deployment has been a bug in +Max what do you think about this +Updated location for loading jars hasnt changed what has changed is the way in which the app server now handles as the app server would recursively locate jarsdeployments and add them into the classpath or deploy them +Theres no need to open a new JIRA for this +Therefore you can have multiple each of which is supposed to be completely independent +So the fact that the correct bootstrap source switches mid stream isnt really the problem I dont think +Could you attach it as a git patch to this ticket +Mask off the LSB and there you have your base type +why are you changing get signature +this is NOT an INVALID issue +I know we are not going to agree on this but wanted to throw it out there anywayI didnt mean that you as a committer should upload a new revised patch that would be quite a workload + on at least a release assert and a sufficient code comment +Unfortunately doc values are not updateable +In any case whether this is illegal or legal and a restriction of Axis it should not generate code that apparently works but does the wrong thing! At the very least it should generate a warning or an error +and interfaces should be tested in and now no additional work needed +I should have updated my workspace before applying this patch +The problem is that the relevant text width calculation doesnt take kerninginto account it just uses the character width only +Slip to not ready for prime time +thanks Mahadev for looking into this patch and fixing it +causes failing testcase is the cause of failures in is the cause of the failures +It happened when one region server died +bq +Attachment has been added with description The used to create the page +I have updated the clustering documentation for both EAP and EAP +I think were going to pass on this one +Yep that solves my problem +have reviewed the patch +Hey Bill There are several parts to the +Its already moved to rao and the folder +Thanks Niels +Just to follow up I was able to create a fork of with exactly the change shown in this patch and this resolves the issue +Changed method to throw correct to closed. +The test passed +Please review +sigh thats what you get for trying to cramp to many things into a afternoon +zipped built should update that I am on can now see the issue it was in extraction from jar SORRY! +Thanks a lot for the reviews Todd +Sorry for the delayare we going to have one index file per hdfs fileCan we also support exporting these index files as a table to some other storage system like or Tokyou Cabinet +I have added some tracking functionality to the code It creates the timer with the timer schedule The rebuilt timer service goes good +sure on trunk +Committed at subversion revision +Attaching Vinods yhadoop patch on his behalf +Thanks Mike +configuration file that uses HSQL as the repository +Actually no this is +for gradually enhancing +For instance I like that you reordered the check after the command line args checks +Fixed for everybody I am still facing this problem in version which I have installed recently +Thanks for the help Mike +This particular issue is about JSF action controls which is fixed in not +Right and right after that is the section I was referring to +But then it will be a good idea to use Mavens reactor to manage build order +Btw how does the dependency manifest itself +how does one cascade javacc after jjtree +Its not intended to act purely as a library +was fixed with but did not show up on the release notes because it is still open +Patch for this issue +yeschecking presence +Added a attribute which gives the list of suffixes that define deployment descriptors which if found in an archive directory indicate deployments which should be ignored by the +Patch applied thanks. +Being tested +maven artifacts etc +Shouldnt have this +Neither way is the one true way it just depends +Thanks I see +mavensynchomegroupsddbdbfithtdocsmreporeleasesrsyncsshPal Brattbergbrattberg +Ive created to track a special test of in the post servlet but it seems thats the only place apart from the ve removed the test for a specific execption completely as this is implementation specific anyway +Looks like our setup for Nexus was indeed getting in the way of Maven resolving and downloading all dependencies +it works the first time then goes one time through all j for i and is happy +The trunk code assumes that this id is always valid and it causes some trouble if the sid is not valid +Defaulting to the current date seems more reasonable to me +I have added the new assertions in sections and to the assertions spreadsheet +Please verify when you have time +Real fix this time + to trunk and x. +Still working on this +Here is a patch that implements the above idea +If fact so does RunDebug AsGroovy at the moment +Tried that on my big instance at work +In addition a copy of log audit log are very helpful too +We didnt go down this route so I didnt fully flush this out +Fix will be available in version of jndikit. +I ran a gatein instance with new portal testes on PLF tung do thanh You can check the last trunk snapshot of tested on with usersNeed to set Transaction timeout to s and migration was done successfully and takes ms +I tired fooling around with it like adding a step of mvn install I have no idea why maven doesnt like it +Ive uploaded the patch I modified w the abovementionedThanks for the tip but Im a bit confused because yesterday I imported a fresh of trunk into my companys svn server so that I could start the process of creating all of these separate patches +OK content you want back in is back and the content you want back in devstudioproductplugins is back there too +hmm I guess we could add a new method that takes a throttle arg Id hate to add another config option just for that +Even though I have two operations both returning Strings the generated wsdl oprearation response doesnt contain return part +Attaching initial thoughts after reading the code +Mockup of if you open the search user dialog and change the menu item the dialog staysThe dialog should disappear if you leave the groups admin section. +is minidump showing the lockKind of exception raised is +As he said there is no reason for not adding yet another but thats not necessarily a developers task +Thanks! +on revision +I am excited about this as weve been using ZK to manage configs for our remote Solr wihtout requiring filesystem access even without solrcloud +without the patch at I meant +After that point it ran wild for a couple of hours and then stops responding +pending jenkins +im assuming this is acceptable. +Duplicate of +buildqltestdataexports and does not clean it up +Reopen so that we can run addendum through QA +Ill do some testing later today on Windows and see if I can fix it in buildr code +Is this actually apache hadoop or or trunk or what? I have not checked the codes in details +Serge if you dont agree please reopen +The patch broke two REST tests so those were modified in the commit as well. +It is scheduled for which should be in a few weeks +Attached is a patch to use the frame from the caller of binding if it defines a block and the binding call does tests pass including the new ones here +Ive posted the charter to the OGC Portal +its now fixed +Danielall logs including loginline one references webstartjnlpHow do you run the plugin +In short term this probably needs a WA but a better long term fix is to handle proxy delegation and renewals for long running jobs +Moreover if we handle FINISHING in the same way as FINISHED we will see that print tells the app is still in FINISHING while killApplication says the app is finished +Thanks in particular for considering a test case +Move the removed code in here to created a service with an independent dispatcher to write history data +Is it ready to release? please provide staging repository for the only version used would be if you use the GA suffix +I overlooked that you use the karaf command to start the route +This is a preliminary cleanup Step one supports these functions +And later when we are creating the we should check if there already is an instance attached to current threadAnyone has any objections +Its the only missing functionality which stops me using Mina Ftp server +Fixed Finnish Swedish somebody pick this up that is working on the wicket contrib projectOr Juha you can do that your self if you ask to be a wicket contrib should go into jira on +So lets see what happens going forward +on the got with util and I dont have time now to figure out why +for some applications this might be a feature because you can put each record on a timeline but for others that dont care it would be an unnecessary overhead +I understand that the Yahoo use case is different and requires assigning percent shares to organizations +It works fine flows are loaded +CommitersAlso if someone gave me the ability to assign Jiras to myself that would be great +Sample stack traces attached +We also see a problem if the userDN cannot directly be used to authenticate and check the pasword using the LDAP bind operation +Patch to display version in Tomcat and method signature +Lets see if that makes a difference. +Not the case in linux machines I used +Sorry +Backwards compatibility was already broken by renaming perTest to pertest +The helper method in handles these cases correctly +John Many thanks for the tests! Unless Ive missed it would you please add a specific test to to verify the output +This proved far easier than trying to a connection to ZK in the running process +Close all resolved issues for Engine release. +Cos please commit it to +Added key infrastructure in r +Even if the messages are obscure and not obvious at least the user knows there is an issue somewhere and can reach out to the community for help in nailing it down +An additional required library for the latest patch +finding the reason is the real task in fixing this we cannot do is reporting the variable name in case of the NPE +current versionJust do svn up on the RT branchbq +Just committed this but realized you couldve done it yourself. +However we need to support an explicit poller even when a locker is configured as well +There are only steps +We are working on individual and need to be able to store these in svn so others can use main issue is that we need to be able to use variables to specify the path to various external jars that are needed by JBOSS to run our applications +Thanks BensonResolved the issue as Not A Problem and closed it. +The is not part of the value +Its going to be a very large patch and the trick here is making sure not to introduce any regressions +Eventually I end up having that is not serving out content anymore it returns bytes +Ill take this one though I currently dont have permission to assign it to someone give me JIRA rightsThanks resolved issues for already released versions +The attached patch rollsback the missing line and I would love to ask that one developer with write access can roll in this in asapTHANKS +For questions please use the dev list +Also John could you please confirm how would the Ss be listed on the dashboard? Do you plan to show it through Capacity Charts +Updated the for the module to override the dummy holder task from and invoke its packaging process +should do for the class resolution error reported. +It appears that theres nothing to do here unless similarity is a completely new tlp +Will start a new thread in a few momentsThe project would like to start deploying their to Nexus as well +Its no big deal to changes that but is probably something else missing? How can I test your patch? Any sample pdfs or fonts or? Concerning your side note +This patch on jdk +I just committed this to trunk and +Lukas how much of an effort will this be? It seems trivial but wanted to make sure +we dont need it +Integrated to and trunkThanks for the patch HimanshuThanks for the review Lars. +Which parser are you using there? Please try a different one and see whether the behavior differs +is scheduler internal logic on how it simplifies the problem +The lookup in copyCache should not be done if the property is null +Patched +Pushing to per Flavios comment +The error seems to be solved by moving the entity classes into their own package +Verified on. +The title problem is a known issue and covered by I tested this with the Availiability metric on the Mount resource this problem did not occur +Can you submit a standalone Java program that illustrates the problem +Ssvd writes multiple outputs and doesnt cram all the files to same output directory +If power users like Ed or Todd want to do that thats fine thats why we made it pluggable but Im on shipping that functionality as standard tooIm more interested in the getting rid of supercolumns use case +Have placed MIGRATIONREQUEST at the end of the Verb enum in updated v +I thought I lost the patch so John started working this jira but I dug it up the patch this weekend +From a quick glance I see no Thrift code at all in the stack traces of the failed tests +I need xerces statically linked as well and will start work on this real soonif no one else volunteers +Different specific configuration data could be plugged in +If no schema is given then user name schema name +When I can find some hints about the correct usages of an API then thats all I need be it from code structure or examplesdocumentation +Thanks a lot for the review Eli +My program does this with a lot of different XML schemas and for me to put special start calls in to special case every XML schema that I have to deal with would be very ugly +Unit tests for all implemented encoders and decoders +Specifically I need a project setup that will display the problem +Too many logos near the top is just going to be cluttered +Will fix momentarily +Are youplanning on changing the btree index implementation which is a high risk change + +I have committed the patch for which will take care of this problem for both OS X and Solaris. +the jar is not only an assembly of smaller jars it changes some jars to a ed version to avoid conflicts with other usages of +I just committed this to trunk and +Wow kudos to BenPatCamille + on patch +See also the linked issue +I will try to create a patch for this tomorrow +re right it does not happen on trunk +Ilya the issue you are reffering to is a wontfix so I just provided this one with a different suggestion on alternative rendering mode as a Feature Request +Whatever the rules I think they should complement the rules for committing patches submitted to JIRA +Instead of setup bouncycastle as a global module can it be setup at application level? Such as at? People have the concerns that a global module may have impacts on other applications which are running on the same ASAny help are highly appreciated +Leo modified patch applied at revision r thanks a lot for this enhancement +Either the server went down or there is something mighty special about that document +Can checksum be returned with roll? Yes +We went with the improved backtraces for +Attaching all files that may have info of the one time I saw when a rerun of the test by itself also failed +Please upload patch to review board and get approvalscomments from reviewers +Actually I have already done this +Though I havent dealt with it +They work in some situations and dont work in other situations like much of the web stack of solution youve proposed might break compatibility for a completely different set of use cases +Fixed on trunk r +The behavior is the same on IE +As Dave points out the optimizations are handled by the JVM so I think its better leaving the code as it is +I would be very strongly in favor of doing a proper analysis and amending the default implementation +Is it going to make +OK +If you would like to argue something along the lines of I see your point but I believe that the percent format will be more intuitive to users thats fine +The new patch keeps the old IPC server behavior while throttling the clients when the call queue is full +I hacked up to generate deletes and it looks like Delete markers are respected +I prepared next patch in which I have done the fallowing modified the Castor version to introduced for createing the renamed the Marshaller Unmarshaller and introduced +Or are there any problems related to this patchI attached a new patch for the latest svn revplease add this issue to the wsdl component group +Ill try to clear some time to post a trivial test case in the next couple of days +If I also do the same test case the latest patchs performance is same as the orig because the increment rowkey is always in the memstore and we do not need to read the store file +Verified with rev +The proposed solution is too simple we need tobe able to handle all situations not just times when there is a space after the +A lot of people using Maven arent +gant with the Groovy content type and get colorized source when editing my + +tnx +Attaching common changes here will open HDFS MAPREDUCE and YARN issues elsewhere +Like I said we can put attribution at the bottom of that page as per the creative commons license +fixed a bug +Thanks for the good work! +Obviously some files are misformed concerning their header +Modify the synchronized block to cover the entire emiteRecord method +Since I suspect were in the extreme minority on this I dont expect to ever see a real solution for this issueIf there are no further comments on this issue Ill close it shortly +This reprojection was correct +It has nothing to do with master restart +on cursory review +tests orgjbosstestwsjbws +Fixed the replacement of identical image now part of +i have isolated it into a small meclipsemaven project with a test case which will hopefully facilitate you being able to look into it +As far as I can tell hes chosen to name his Jikes RVM directory with spaces in it +Guys from ATF use earlyStartup like have tried using plugin with name beginning from com +unzip patch u i p issue is not resolved by submitting a patch but only after the patch has been applied +postponing +to and trunk! +Attaching patch with trivial fix +Thats true but at the column level +On the other hand after reading all the comments I dont see a single good reason againstIts a no brainer to me +makes more sense I was just trying to maintain current behavior where anything other than native returned true +bq +It is not blocking alpha but im wondering why we need this in TP for it to work? what is special about this component vs another +a jopr is targeted for so this is no big issue +The stacktrace might give you a hint +Vijay is working on fixing exactly this issue +war from the deploy directory +This ticket has IMO more severe implications because the analysis is terminated with an rather than simply not showing the correct result for that class and continuing the analysis +For example ltxschoicegt +Patch committed to the branch +Accepted for as resovled as all components have been upgraded +proposed assert attached +Thanks God your eyes be sharper than mine SureshIve removed those two lines +I still need to do some cleanup in the code but the patch is working +In contrast can be raised according to a number of other reasons +Therefore it is normal that Axis throws an when processing the response. +mocks based on the previous implementation of checkDir +Closed. +The class extends which presumes that the database instance already exists and puts everything in a single transaction +In addition Eclipse is offering to Fix this error but does not actually do so +I think we should roll that specific issue into a new JIRA issue +Now the exception thrown from the old archetype code inherit from therefore i need not to encapsulate them. +See the last line of their FAQ +Just so everybody is on the same page let me just summarize how things work today The Central Statistics referred to in that blog posting are delivered by means of a Nexus plugin thats currently deployed on and +Ruth patch applied at r thanks a lot please verify +Another issue that has been discussed in the past is the impact that long lived processes can have on resource scheduling +and its requires support for multirow inserts namelyIn some cases such as in an insert select statement more than one key may bereturned +Expected results has already been realized on js side so we only need to make sure worked in this patch +Is this a change to the API or a completely new class altogether +Is this still the case +But I prefer +Removing maxSize from is a big changeI applied the to TRUNKDougPlease continue with javadoc changesThanks Lars +Throw exception when lookup fails only try the other configuration methods if lookup wasnt configured at all. +wont fix this in release +Command is not present in DR +XML files from ant junit runs for suites that have errors +please provide the steps that are required so that I can reproduce this +this would help us a lot +and please also show that the debug log printing out Hibernate and not some other hidden hibernate version +I am using ibm jvm to build trunk +I specifically mark s added through the SPI and create a long id for them other beans have the same have also added a simple is not finished as I need to do something similar to some of the other types of bean +What is the status hereThe possibilities to fix this where it actually breaks are not that nice but the above patch is actually would be great to have this fixed in a release of the near basically prevents anyone to use git maven and eclipse in a flat project. +activerecord is updated to create oracle timestamp columns instead of date +I feel like this one was previously completed no +However it does not function properly because there are some other issues which block this behaviour +In this case still just the column should be returned and not all columns +Ive unscheduled this issue as its not a issue and there is no decision about if or when we do go with something like! Hopefully the same infrastructure will enable a simple Grailicious mechanism for users to package their application components into deployment bundles +Flag added in revision . +Better to have some auto select mechanism to help with it +Im leaving this open in case you want to discuss for explanation +Moving out of +Any comment why this issue has been reopened +Ive subscribed the following listsuser AT I dont receive mails and cannot send nor receive messages to those anyone help +I dont think the compiler or runtime are doing anything fishy with loading or identifying the class +fixed the check for no lodger in this patch is broken +Uploaded the patch +this is still happening with a fresh build of the plugin with svn at revision part of the site is failing? or cobertura or something else +The conclusions from this issue have been copied as part of the description of each of the above mentioned issues. +I think the best way to create a test would be to run a chosen existing drools test in or even more threads to spot the difference +Patch with Todds suggested fix to simply remove the log message that triggers the fatal exception +mark yeah thats an option as well but I am only willing to take that route as a last resort after trying the approach +So that patch is ready for port to +Attaching the actual patch for enabling validations with annotations +Also see in resilience tests. +Really sorry that we left this patch outstanding for so longNote that the views evolved into the dispatcher and these documents are replaced in with docshowto +Username kongdy +Marked for move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Interesting workIt seems to me that we may need to consider deployments where a gateway such as Knox is between the UI client and the Hadoop clusterHow are the relevant configured for the deployment are they easily configured for a particular deployment scenario such as this +The case for HOD was different there was a LOT to document that each guide would have been somewhat unmanageable if we plugged it all together +Thanks Michael! +Attaching a patch that adds javadoc in the class +Moved docsdocs to docsxdocs Added description to ant docs target +Thanks for that Eugene it worked! Very happyI cant help but wondering though +I tried to modify as least code as possible +However if a database supports some datatype we must be able to verify that we can import it to HDFS in some recognizable representation + +Added +Josh thanks for updating the patch +For the tall bench were almost as fast as MRI +this is really familiar +Properties files and settings folder attached as well +In the file there is a web application that contains a portlet to be used for tests the issue. +Why should I object? Please go ahead +Committed +Hey Dragan would it be possible for you to include some simple tests for this functionality +Updated in revision +It would be useful to describe the and the tool as well +I found a work around for this issuemvn clean should do it +Could you try it and let me know whether there is still a problem? If so could you post the full stack trace and any stack traces that appear in the console +See +The bootLock test is now getting skipped with weme and +Initially I thought that the issue is only for listAccount and listEvents +Maybe put that in a comment over the enum explaining this priority chain +Vitali it is just a few plugins from common I do not see it as a big part of Tools +RegardsAnil Patel +I think it is not the task of the tests to check for system readyness They should assume the system is readyTherefore I think the workaround is not needed and the whole check block should be removed The standard test setup in the launchpad should ensure Sling is up before starting the testsIf the system happens to not be ready when the tests start they will certainly fail and thus we are on the safe side of not getting false positives +Verified +Added a Class to look up data locality +Hrmm then we can use it as is as test in trunk is testing it and since these versions are quite close to each other I trust it works as expected +This is the feedback I got from supportHow hard would it be to catch the exceptions and replace the exception with a static message like Security Manager does not permit reading This will be less likely to raise any concernsI think that this is the same as and I think is a good idea but I think it could go in after the patch for +Attaching +Yes I plan to measure seek performance within one blockI havent implement it yet but I rather expect that it will make seeking and decompressing as fast as operating on uncompressed bytesThe primary goal is to save memory in buffers +bq +You can take a look atcontribsrctestqueriesclientpositive contribsrctestqueriesclientpositiveThese both show how to test a UDF that is not hard coded into the care to take a look +When I check the it has three classes call and +Verified. +Max Can I get approval for this +Would like to get this into +We dont want to hardwire version numbers in as it will break if there are future releases from this branch +The good new we have upgraded to then we can perform the test soon +ilya tells me running without sudo fixes the root userssudo environement on his setup either uses gcj doesnt have gtk fine without sudo +failures Will be corrected as part of + +I have removed the delete statement in and types to overcome the problem +Carousel Central toolsMove downloadcommunitydoc to the side have a front and center Install now pushing Eclipse marketplace installsDont show all options on first download pagemenu move things of to archivesmore downloadsConsider putting a movie front and center showing install usage +instead of saying Command not found user can easily fix whats wrongIm fine with current Double definition if we can fix the error message +Closing this as Wont fix +OK the issue I looked for is of course +fixed. +WolfgangIm having a hard time recreating this issue +Or maybe your Maven POM can use the same tricks we do to generate such a +on samples are moved into example need to be addressed separately. +Logged In YES useridThis has been solved in CVS as now the refresh occurs at theend of the job. +The new patch incorporates Raghus comment +cool. +This is the cause of the problem and it is related to some changes that I made for +marked as resolved there is available a fix in JSF implementation see workaround. +I would like to submit my documenation so that other users may benefit from the information + shouldnt the unit test be inYes the generic bits certainly +Checked in to trunk and merged back to close after is released +Thanks for the hint +Some machine disk or operating system settings can prevent a proper sync and cause unrecoverable database corruption on power failure system or software crash + +I suggest that tests are migrated from one file to the other and that the wiki page is completed with the tested scenarii + well Dinesh could you just Remove the deprecated PMDCheckstyle rules from the default quality profiles Not log some issues on import removal of the deprecated rules will be done in ll have a look regarding the rule suffers from the same limitation the one about so Ill also reopen that one +Ill have an update patch today +I am planning to continue the work +Any new problems should have a new jira created +Can this be an improvement or bug? I think the balancer usage in and AM was a bug right +Patch applied +This results in the kind of errors that the istat daemon is seeingFor an orderly shutdown on the database should the system wait for istat daemon to complete its current unit of work? If so it seems like it should be possible to make the system wait for istat daemon to shutdown but I am not sure how easy that would beThere may be some value to logging information about failures in the background istat thread to but i dont think by defaultit should be enabled +Committed to trunk +new parser change how the following exmple is +NickYoure right thanks for specifying this +I generated the documentation and it looked fine +We dont have any comments from avro generated codeI can see the comments only from +Yes you are right please submit updates in the diff format by updating the filesWe appreciate your assistenceRegardsHans Bakker +It will be a minor nuisance for people running in mode in production so we should educate them well in advance +attach patch for review +It is not required since even if the controllers cache information is undergoing change and we send stale information to the client it will get an error and retry +Oh cool +AndyI am testing this now and it looks good +Everything works as expected +For this reason Im making the JIRA depend on +Moved to upon release of +Thank you for the patch Jan. +Please pardon the long delay between attaching the patch and taking care of itThe patch is now applied +OK pointing to for tracking the documentation effort then +Powell does support boundary constraints. +Im unable to reproduce this +Committed to trunk and branchxSteve could you take a look at this patch and let me know what you think? Feel free to go ahead and commit if you think this looks ok but if you have concerns and think we should go a different direction for the fix you certainly wont hurt my feelings +This issue should be documented in the release notes as a known issue +This is right if you have the all components created in theBut this prevents extending theHere in the how can you add a component as you have no reference to the htmlTag. +I only tested against the with same patch by Thijs Patrick I will try to create an improved patch next weekend currently I have no time +this does not seem like the right way to fix it +If I remove richvalidation the correct behavior happens in the old fashion way +Here is my patch for this issue +on patch + +Same as v +Non committers can still post patches to the branch in JIRA the same as if the issue was in trunk +So Ive changed the issue to a bug since tests should not fail based on the order in which theyre executed +looks good to commit! Thanks dudefixed in r +I think this means that together with the patched JON plugin we need to provide upgrade for JONs backends right? AFAIUI this data is not stored in the s about history? Will that all be available with the new plugin +Peter is this still a valid request +Without having looked at this more than a glance my gut reaction is Why is there a being carried around in the message headers anyways? But I agree with Tom that we can protect ourselves by not mapping headers +Not sure if I did something unintentionally stoopid here but it does exactly the right thing and other approaches like etc +I took the initiative to add Windows detection as it will be hosted on Windows repository in windows subdirectory +Ok not a big deal +Amits review is certainly welcomeWe may add more tests involving deletion and major compaction +This patch lets users specify names when converting an array for use by pig +Shreepadma Any thoughts on how we are going to store low and high values for column? Seems like there are four possibilites We can add two more column in M table of type and +Well consider making this more flexible for Spring Spring will be more flexible +not by much but it strikes me that way +Added description to and added the goal to the for all repository goals +This content could be moved into common +Completed +I will try to get a fix published tonight with the goal of doing this saturday +fix in rev +master assigns the same region to the server because it is still unassigned and no region server is serving it +The main propblem I found with thisis the presence of empty fo elements +A combiner is definitely the next step +Bookie still owns the instance but creates it by passing in the and +Close and reopen if seen again. +Umm +eg Karaf ships with a custom file you should override and use when you use CXF Camel etc +New patch for v +Thanks Owen! +Not a bug +Checked in for I know wsdl is generated with and there is some way to edit wsdl so it can be invoked successfully +Confirmed atIs it just a wrong log message or is something going wrong +Any chance you could try and reproduce this with as well +What would be better a simple property to enabledisable exception handling or some more advance mechanism like interface with default Struts implementation thus will allow to define user behaviour what should happen when unexpected exception was thrown +bulk close of resolved issues whose fix target is either unspecified or from a prior release. +Sure +Thanks Robbie +Committed fix in rUse color +If that doesnt work also upgrade Axis and Rampart to +I have tried the code in the SVN trunk for this fix and it works for my environment +Looks like client can accept with out hex conversion when the data type is described as binary it is defined as binary however client is looking at SQL type as SQLCCHAR which is different than PGTYPE. +Seems to make sense to me makes it clear what the target of the code is +But I wonder if it isnt even worse to have the container disapear when we have an error or even just when theres no dependency +Style and style inheritance still does not seem to be working the way that I would expect +But if it receives an error it will be able to rollback the transaction and the same message will be processed later by virtue of the redelivery mechanism in jmsI think in your use feature you need to intercept the exchange before it reaches B and act there else there are chances that endpoints earlier in the flow may be a bit screwed +Committednakarufelzaderbytrunk sh Sending Sending srcjaJPAdding srcjaJPAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefAdding srcjaJPrefTransmitting file dataCommitted revision +As with Elis comment it happens when taking the pic +But if something is really broken we will see that this test fail every time +So what do we do +I dont think the precise location of the commit or of the sort buffer size matters because the sorting occurs beforethe first record has been retrieved so as long as the commit is after the first record hasbeen retrieved whether the occurred on the th or th record doesnt really matterWhat matters is that the commit doesnt close and release the temporary sorted records thatwould cause some sort of internal error fetching records after the commit if it occurredIdeally the test would verify that the sort actually did overflow to disk I think there is someinformation in the runtime statistics that will prove this +Reopen if appears again. +If it happens to be unavoidable to reobtain the bean names every time the factory itself could do some caching +When a client fails over it will try to reestablish its last known session on that server +Do exact phrase queries rely on somehow? I cant find the path between them in the code base +As a matter of fact this is already fixed in trunk and on the branch +I will send a PDF to Rajesh when the release notes are complete since they contain information about the JMX issue +Reworked the patch to send an event when the completes and have the expirer handle it +Thanks Jitendra! +This is CXF +This issue seems fixed now up on hudson +It may be something specific about my clients application. +is not that +Hi StefanThis feature is very important because databases with composite primary keys are very common in real the feeling is that database reverse engineering does not work properly when composite primary keys are have several upcoming projects in this situation it would be possible to include this feature in version +overall query performance is great +They will come as a separate jira issue +I like the line of thinking that you have using the context to differentiate writes and repairs +I hope to get it into Roller trunk this weekend +svn move scheduled for tomorrow afternoonPlease notify the dev listThanksCarl +Well be putting these jars in prod tonight +I think this is explained in the Axis C help pagesI dont think I would label this as a Critical Bug in Axis C +bulk defer of unresolved bugs to +A logo should not contain a story in words! The logo itself should symbolically identify the companyProject +Change the c and linker switches to equal the Release switches except that now the linker produces pdb files +My fix had an overflow issue and the commit commands didnt actually wait forever. +they made a blocker after I added a comment yesterday so it will at least be fixed in Commons Collections. +Perhaps I should rethink how I use the Modal Window? Any suggestions + Attached patch contains updates for the README and files of the following samples The README file is modified to include the instructions to runuse the samples on Linux environment +I dont think so I had originally though a contributor licence agreement was necessary but one second thought this isnt really that much code to where it would be necessary +Vinu can you put this in the installation guide instead of the seam guide so people actually note this during installation and before actually running into the problem +Where should I look for that? What triggers +He uses same build +I am not sure if additional jars will be required for that or if it is some other issue +Committed in revision . +Do you know how the string values for these items will be displayed? Still need the previous questions to finish up the work +I was not able to find the JIRA issue now +for the idea but your patch does not apply to trunk +I cannot produce any case where spillable list bigger than elements using the current code my implementation to save memory is useless +providers +good +Writing a testcase is really difficult it requires the kerberos setup +testcase for this results of the testcase +I started seeing this last week and I think fundamentally it is caused my some of the new code in localizeJob +Im not sure how we want to release the NMR JBIRight now Im thinking about a simple collection of bundles that could be donwloaded automatically from the Runtime +Attaching adopting BeforeAfter annotations +It seems like it would be simpler to handle the KILL events in the states that missed itThere isnt anything like a missed state that is causing this issue if I understand Ravis issue description correctly +Wed break this batching writes +In facelets work fine +Fixed because were now using tycho to build installers. +Wrong priority Minor is the right one +Sample java code and think the second level cache is not involved here +will commit Niranjan! I committed this to trunk and +bats +bq +UML diagram of what we currently have for an API +Namit thanks! I will test it locally and upload +Charles reckons there dont appear to be any repercussions from that message so there no plan to do any work on this +Thanks for the review Mike! Ill remove the printlnOk I think this patch has been reviewed a bunch of times andshould be ready to commit now +Implicit value assignment still goes the same way as it was except that we check whether there is a value conflict before using it as the final value +Looks good and nice tests +Please add the model project set containing your model project set +Your comments are welcome +Changed Configuration everywhere so were explicit that its an thats needed rather than being lazy taking Configuration +fixed some testparse logs s not clear what adding a new group by mode would do +This patch is per Alberts comments to throw in +They can come in handy when you keep your links +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +offline working is +Applied your patches with some minor changes for now Guenther + Synchronizes calls to in +Just one comment source spells property rather than soapAction +Were there any persistent messages that were delivered when you tried to start activeMQ? If so please give the number of pending messagesWhen these crashes occurred then most of the messages sent via were persistent +was a duplicate of Hadoop +launcher can actually use this API to generate the file at a location convenient for its use and use it from there +I am resolving this. +For example if I subscribe to T with S not knowing that someone else has also subscribed to T with S I will expect to see all messages and this may not be the case +ShashiAm looking at this again and was wondering why it skips the labels when there is less than ids? Seems like this could be configurable or are you saying that it only performs well w at least +Its a nice addition to shell Jie so lets get it in +I think thats fine +Looks like Cruisecontrol Plugin was the former name of the Build Stability Plugin +Unmarking patch available until we get a svn diff formatted patch +I agree that Stuts could do more to verify the preflight status of itsconfigruation files +Does that make senseAlso please submit patches as a single svn diff from the top of the lucene tree +Thanks for the contribution Nicolas +Yes we can only do this for things that have additions +Now I still have to use my own string +Im working on a patch for this issue +Of course in the case of JBDS we really wanted to rebuild not just because of the fact that TP contents are INCLUDED in JBDS installer +This is because java doesnt guarantee consistency of long value if not synchronized +Has this issue been confirmed yet or not? +This is a bug with MVEL Ive added a failing unit test to MVEL hopefully well get the fix for the next MVEL release +I just found the error message of subversion that causes the scm url to be marked redProvider message The svn command output svn CDatenContinuumtarget is not under version controlThis message of course is wrong +I have not yet updated to the latest version of the tests cases so Im not sure if thereve been any changes in the case +Could you provide a stack trace of the blocked program? Then we can find where its blocking +I will fix this in parameter is not a full timemout if we dont have the we will continue to loop until the master is killed +If you want to prove the ability to use something else as well then that should be a separate JIRA and patchThe token endpoints included in this patch should also be a separate JIRA and patch +It doesntmatter on Windows and surprisingly it doesnt seem to matter on Mac +will commit this patch momentarily +Heres the patch as a file upload too +Attributes descriptions were added +implemented in Revision all issues which got resolved in M +I have just built the trunk of the and got this with my simple test above ldapmodify h localhost p D uidadminousystem w a f adding new entry cnKate Bushdcexampledccomldapadd Invalid syntaxldapadd additional info failed to add entry cnKate Bushdcexampledccom Attribute value noNumber for attribute mailpreferenceoption is syntaxically incorrectSo this problem does not occur in SNAPSHOT +Yes I think Garys example has one thing incorrect +we want to integrate this into trunk and? If so I can write the trivial test case? +Also we could perhaps now bump up the MEMSTOREFLUSHSIZE of META now that we have sync support at the HDFS levelWas done in trunk for branch it should be also like that if and pals are there +Merged the fix to with revision just in caseAdded link to another report of the library typo. +While plugin dependencies can be resolved from snapshot repositories the plugin artifacts themselves only come from release repositories. +I have just created a test case for this and it appears to work for me against trunk at least as expectedFelix will only attach information to the Package from the manifest main attribute group not individual groups +Actually dont worry about the reproducer I think I can see the problem +Committed. +I dont understand +Thats really bad +causesdeletes and new segments to be flushed to the directoryThe test cases from test andtest fail when the IW is opened again aftercommit and close +Please find attached xsl fo file as requested +Perhaps move the while youre at it? if allowed smile +patch applied in r +If Hiram regenerates the certs unit tests work on both his system and mine +IMO the documentation was not really clear on this but thanks for the suggestions and explanation +fix in svn for you please reopen this issue +Can doc values be optional? I am looking into merge and see that the logic assumes that for every docId we have a existing value +Thanks for patch ive committed the scopes test from within the to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Rolled back above change +Problem solved with latest patch +It turns out that we only check for a zero memcache size at the region level +Agreed +I agree with Kathey that this is a good candidate for a bug fix release in the lineage +Also renamed to and added to the sandbox. +I will let Jonathan look at it +In essence the created schema conflicts with the business logic of the application +We wish we can get this functionality in EAP this mean that this feature is pushed out further? This is one of the critical features ATT is looking in for long time +So do and solve this issue? If so then we should mark it duplicate or at least dependent +I cant see a lognet namespace in the stacktrace +Right I actually already planned it for this release +Im for keeping whats working +reopened for fixVersion +Yes there must be a sync while building the cache it is a complex structure +It impacts how easy it is to share a jstack of Accumulo +A patch to the new patches has been created and appliedAlexander please verify that everypthing works as you expected +Very strange! Ive tested it in another computer linux as well but with LCD monitor and it works nicelyI will install the google browser and try it out tomorrow +But its not a big problem so far becouse JSF facet works fine with JSF +To preven this Castor internally has to execute an update instead of create and delete +Simondo you have more info? further news from Simon so Ill mark this as fixed because it works for meThis problem may occur for instance if the the default ibiblio patterns are overwritten and the classifier token isnt specifiedBut without seeing the file it is hard to see whats going wrong. +Here is the patch implementing those changes +The same problem could occur if were passed a URI without a trailing slash but this would have to be handled in +Would that be enough or do you have something else in mind +Lets just remove it +There may be errors if the type declared doesnt match but thats how it would work now anyway +But just out of curiosity how would I test whether a warning with a specific message was logged or not? Is there a pattern for that +Found this bug while writing tests for +However this is a very very minor point. +For me it looks like slightly different issue it happens in batch runs only I cannot reproduce this running the single test +Will try to get this done in next few daysAs I said earlier keeping without adding Quorum semantics allows writers that have different policy +So I suppose I was on all along tho Im pretty sure it was claimed to be when I downloaded it +It is default +Its important to remember that is keyvalue not a column family data model so there are concerns and constraints that apply to cassandra which do not apply to +See +This causes a concurrently running checkpoint to fail +OK sounds like there could be some value in supporting DHE for key types +New patch that incorporates Roberts improvements and tests more corner cases +They will have Incubator brandingAs I understand it then this transition should occur when we are ready to release a new TLP branded version of the source code +Committed to trunk +scan since it is still able to get a invalid region location from the +Thanks Tom. +Geoffrey I am today using the same functionality as the assembly descriptor is +probably the script will be removed and support for runtime calculation of concurrent paths of execution will be supported for now we didnt change anything except that we log a warning if users use the script xml element in a tom. +Got it thanks Liyin! Nice work! +The Fix Versions should be but it is currently missing +Committed at subversion revision +its a misconfiguration problem +the WZ library is not suitable for complex drawings +Would you want to do the same thing for binyarn +Hiya Jeff +Fixed on trunk revision +The replicator will quietly start a new connection if the changes feed goes down as of r +has been closed as fixed +Cause this method is expected to be called for every CQL column of the result set and this for every row even if the result to return is null +Maybe would be better +I dont think well support bit systems for the time being unless this becomes an issue for users. +Just come across this ticket I think this functionality has been lacking from from day one itd be a great addition and would save tons of code across an plans to implement +Ie this issue now has two such labels gosc and mentorthanks mike I changed them back to have no this over to this wont happen in move issues to and +form +If you have any general issues feel free to raise them on dev over the comments here and without knowing the details it looks like a big improvement over what we currently haveJason is your concern that this is a fundamentally wrong direction or that it does not go far enough to support NRT? Can this patch be improved? Can it be committed then improved +I have been taking a look at this it only takes the method with one parameter but there must be at least one field with one setter +Re making this configurable if thats really desired please +A patch is not needed just to delete a file +I undertand sometime for classes i need to restart +Im looking forward to trying this out later on +All unit tests passOne question Should we count buffered delete terms towards ram buffer used? Feel like we should +However my primary artifact type is WAR and the current does not package dependencies of bundletype into the war +The server log file showing clearly at line that shutdown was called somehow magically +That mitigation is really the only benefit of and obviously must be weighed against the costs +Thanks Owen +I wonder why I reopened thatIf there was a mail I should have appended it right away +I will add that comment +why the number of message enqueued is more than the number of pending messages +This looks very much related to it doesnt run fine for me +Done +That could be masking another problem of course but I did not see one with my changes. +Even if it is smaller some io threads shutdown will wait for some space to free up +Guys its a pleasure to see such proactive cooperationApplied the patch at rI presume no more general fix needed can we close this issue +just the gui didnt update or somethingThanks for the tip Prasanna +but I mean its a bug if wsdljava generates errorneous java code isnt it +Fair enough +Would you please chat with Norman about it thanks +CLI contains support for a template key in +The application of the expansion occurs in a generic way in the class for properties listed in the config property +Unit test Isnt the default value true here? If so then not having the query parameter should default to true +Committed to SVN +What I applied to rb and what I want to commit +Are you seeing the same thing +Patch modified to be invoke and invokeNeed to write tests +For cluster datanode process is hard coded to hdfs userDoes this mean Ill have to adduser hdfs to run a cluster? If so Id much rather not +See the discussion and solution attempt around Solder +If you can use for accessing such metadata that would be betterOnce this additional information goes into Table Column Family definitions we have to consider security support in a environment +I could find workarounds for my case +That way we can look at smaller lumps of code and hopefully commit the chnages incrementally +Reviewed webservices only diff is a typo correction in a method name +StanCan you please create Maven project reproducing the issue? I have problems with setting up jsfunit application cannot be launched correctly Im getting various dependency errors +I just tried again with the version and was able to repro so something else must have changed that fixed this +So all decks cleared for hbase mvn repository publishing here when it we do a release +Changing this to a bug since the current behaviour violates the ORDER BY clause that might be present in the application specific EJB QL +Checked into trunk + +Still you could be having something wrong there +First into trunk rev +Reopen for backport. +I must assign the cardinality of matrix while create them +One should not create a object using one authority and then pass Paths to that which use another authority +I need to setuptest a jenkins job that publishes the latest quickstarts to then update the descriptors +I havent been able to reproduce the issue so Im trusting you to test this fix +Thanks! +Thanks for the patch +s get returns for such an empty path within the mapping which in turn matches the automatic prepending of a slash for +Phew +This could stall first level batch writer unnecessarily +Make that. +that adds a file so thit test will pass again +In the Forrest documentation max capacity is being referred to as sometimes and max capacity sometimes +Logged In YES useridThere is no attachment +MAXINT seems too high +There is jit mode defined already why adding def? Also note there may be issues with empty switches I recall VM does not receive arguments following the empty one on Linux +Since SDG typically does not use a there is no output in the log that can be used to verify that the configuration is correct +The launcher is not constructing classpaths correctly +If not for it should surely be a blocker for +Remote jndi is working for jms objects in EAP ERER +Looks like the server is a bit stalled today or someone is saturating our Internet connection Ill better let the update for tomorrow and let the current test working +In short the algorithm is working as expected which can be slow in some scenarios +Johnny thanks for these bugreports it looks like both of them have the same rootcause so Im closing this one as a DUP of +Is this just for HDFS YARN MR? Weve already got an existing umbrella JIRA for downstream integration with Hadoop +Committed to the trunk as revision I havent marked this issue resolved yet because I suspect we may want to investigate merging these changes back to the branch +If we put it into the normal build it would fire every time possibly prompting for passwords or failing outright or showing a message or what have you +currently assigned to +Would you mind producing a small test project? I am trying to find a collection of example project to test the plugin with and then devise a way to automatically test them all that doesnt involve me using my laptop +and not to forget that is not always available +What is the suggested patch process in case I want to apply this patch to my broker running the jdbc configuration against oracle? Should I just build the core jar from source after applying the changes to code in my local svn repo +Reopening to assign Tomas +This is reasonable. +Move bulk of performance items to target +Therefore the CS can cleanup its files +The edits look fine except for the footnotes whats the problem blocking these +However this is a incompatible change +However i may possibly introduce the feature of being able to batch file downloads together if there is a list of files from the same site which are scheduled to be downloaded +Removed all references to junit +Hi Gwen are you still working on this +I think the static variable and the usage of it was causing trouble because the reference was visible before the object was finished being constructedFair point +Added documentation for Emma. +I just committed this +Correct patch +Ill like input on a Question I have If we select Agreement while creating an Order Later after we have added Items to Order and we go on to Finalize the order Order Terms screen is presented where we can make changes to TermsThe Question is The Order Terms step is dependent on selection in Agreement screen Can we combine these two steps into One Screen or at least have Order Term screen follow immediately the Agreement ScreenInput will be helpful in refactoring process +from me go ahead +Ill put it up today +Brian Use Case is You have XML and want a schema from that and then you have a schema and need classes from that +They try to resolve value as URI by different resolvers one of which looks for the resource in all web roots of the project +Perhaps it makes sense to combine both by having woodstox throw an that has a nested +The data types must be the same between the input and output of the combiner +bulk defer issues to +Drools is also updated +This makes various changes to the messages to clarify the context which translators need to do their job +if NN indeed deleted the blocks when the files are deleted there is not much motivationneed for this jira +Moving this one back to critical +The logging should also help us pinpoint which threads are blocking each other and why +Is this intended for CR +same thing only specific to android +Forgot to mention that I commited the fix +Please test and verify all is working +Why do we have that limitation? Even by deleting only I can seriously damage things +Hmmm +Sending srcjavaorgapachelucenesearchTransmitting file dataCommitted revision . +Use NEKO or TIDY parsers or encoding +You still havent expressed a bug +That makes this less criticalWhy these are not used often is based on use case +Moved from to just after merge of old branch into TRUNK + +As requested Im closing this issueIts not necessarily a problem that you have and bundles mixed together +This is a request to fix this issue in the AS ESB plugin in this bug was considered for but requires a plugin pack update to deliver the fix +Point versions are really pain +Here are the classes I plan to contribute +Using releases promises you back compat to some degree not using them doesnt mean we will change things wily nilly on you if we can help it +That sounds workable to me +I will first work on getting those snapshots published +Fixed in r. +Patch committed to both trunk and branch. +also has a patch for this issue +There is code in the generator and the databinding transformers +Im using IE with and without Compatibility mode +WHERE wherehere is not a possible query but at least the set of illegal field names is small and with that oneThe bug here is this is completely missing from the HQL documentation +Integrated to synapse +Thanks +Have a minor fix to update +I have been looking at the changes for the last minutes and I think I know where it went wrong +A new build is available +thank you very much playing in fullscreen works nowbut the other issue described in is still therewhen you play a move and hit the done button the in disappears as expected but the audio is still playing +When I forgot I was reminded and told to correct the log +push out to +I found another problem in page for MINThere exists next sentenceThe types comparison rules determine the maximum valueHere maximum should be minimum +The PID is basically the id of the portlet targeted by the last portlet requestThat said the problem of the is NOT properly aggregating query parameters which have been programatically appended to the path to the render parameters which allready existThis should be looked into further +No time for that right now +The other option of prerendering areas with effects is a bit difficultIf the same results as version are required then the effects should be removed from the svg +Please discuss JREs trust anchor handling at +Thanks Stack for the comments! Ill incorporate your comments after feedbacks from your test run. +See igor sorry i didnt find this bug before +Sometime today +Clicking on the username field does not select the whole value but this is a common behavior in Eclipse so theres no reason to change it here. +I plan on committing this soonThe public API is very limited just a threads local param and fcs +Can you put together a small project that we can use to test this issueWhat commands do you use for your incremental build +Have you tested it both with http and httpsI do seem to remember there being issues with aboutblank and https especially on to trunk +The failover transport cant automatically fix those sorts of issues +Please click the checkbox Grant license to ASF +The patch is trivial and does the trick for the Eclipse the pathc the +If you do this youll probably need to mention the origin of the original pluginOr you can write something from scratch +properties file are located. +I made the following changes in master Spring Security will no longer log the error if Spring Version and Spring Security Version are the same +If you then set init paramters on it you modify the only available singleton +Instead of using the weak reference why dont we use a dispatcher callback that wipes out the iterators opened in that thread once the request is done were sure the iterator can be disposed of and we dont have to wait for he garbage collector to run on a JVM with a lot of memory that might take quite some time +I have commented the offending test +And it is enough good even its simpleLooks like DBCP has something like and variants +Thanks Hairong! +When that is the case BRMS should ask for one package thru a is more like a merge feature I guess +Im trying to set both lang and language attributes now and creating a documentation issue for later +It does look good on simple +This has been fixed in revision +in there are two modes in which the junit tests are run +The From address field doesnt seem to be used at least with the SMTP server +This file loaded in the nested load of native MR operator it should be possible to check for this file +An alternate approach would be to have the passed as a parameter by the +There isnt any patch here yet can we move out to +iBATIS only caches prepared statements within the context of a connection +thanksIve just try to login into the web ui and worked fine +the bogus conversion of into connection timeout +If you could update the cdm so that all of the choreos are contained within the top level one and references updated then we can see whether there are any other issuesPlease raise a separate jira regarding dealing with top level choreos providing this cdm as an example +export totalScans from for testing in scan add an early exit if we are already done scanning for the current period +Thanks Venkat for the patch and Akejandro Daryn and Jitendra for taking a look +Many thanks guys! +I doubt we can get it into +Committed changes for this problem in revision +Not only code changes are required in these projects but their ability to work with different versions of camel +You are correct that it changes the behavior of theAt the moment Im concerned that while applying this may make things work in an environment it may cause problems in environments +it should definitely be configurableStill the code for s getWith contained in the patch could probably be applied as is +Attached patch for trunkAdding parameters to the merge tool +Please try again +ya I saw the NUMADDWORKERTHREADS already existed when rebasing +I guess you already implemented it for you mock +New patch which fixes the history server and adds an integration test +But the whole point of the ticket is to remove this concept +See. +Bugs are escaping because of Jenkins broken links we are close to fixing thatI am going to fix the warnings for the time being +May I make a post instead of you +datadatacurrentVERSION +We mayconsider the Part API changes for +Thanks headius +Im running using s Debian cassandra package and using the same shutdown script I havent seen this exception at all through many shutdowns on my so this suggests its new in and trunk +Diff of the +its normal when flushcompact occurs after minutes write we saw such level downgrade when doing long time tests for both withwithout patch +It looks good for me +Well have to verify what is needed with CXF +Make top levelId defer this to a follow up patch +Weve designed IDM with that known limitation of our previous would remove that limit once we remove the former API to use IDM dont see how we can reasonably solve this for the releases +Notes about the specific files added and moved in svnin trunknetuisrcorgapachebeehivenetuitagstree I moved to to and added +If fails to deliver to the local queue then the remote server should not believe it has been deliveredI am not saying that we need to hide other exceptions as they will probably occur through other routes and definitely need to be fixed but that they should be reported while being handled gracefully +To DimaThis is not a production code +First we need to update contrib to the new API +Is this issue still existant in or? is not maintained anymore +Yes this was a quick test using a bigtop install +I get the same result with the new versions of the struts filesWere you able to get the sample to run or not I was not sure based on your comment +Updated to include specific issues +as you mention we dont have support for dates or enumerated types translations +Deferred to is actually a discussion on whether its feasible for bull to go fo api for all process languages or api per process language +Thanks for reporting the issue and proposing a patch +If the user chooses not to include the location or the location cannot be determined then is passed for both the latitude and longitude values. +Load from the CFS constructor instead the way you had it originallyThis adds the preload cache option to the Thrift call but doesnt update systemaddcolumnfamily +then +I committed to the branch too +hm the change to fix this possible failure seems good +Is this issue still open? I will close on if there is no response +Class User has many Resources +mistake I used the hardcoded value in place of have done a great workI really like this stuffIs this JMX integration already in some version +Mac problem is related to and Developer packages +Thanks JihoonI also agree with Hunsiks opinionsIll upload the improved patch ASAP +It continues to allocate memory but its never Micheleis there any possibility of sending us a patch +Im closing this for now but let me know if you need the Download link and License info adding. +I guess it will not apply now +A test with just javas yields only a times performance hit versus with Spring AOP. +That may or may not be practical for youWhat is the status of this issue +If you update the patch with addl debugging and fix the test to actually fail in this case Ill be happy to run it again on my clusteradd some log detail when the followers complete snapshotting so we can rule this outin +classes testsThis patch contains also a launch configuration for these tests its called UIMA Cas Editor pluginHope that helpsJrn +Kens unavailable ATM +I find some other cases will lead to publish errors I will consider it more carefully and provide a new patch +That would probably be a little more difficult for me to do This patch wasbasically copying a little bit of the Linux code for the file io and a littlebit of the Win code for locking +but if this will block the commit ill look into it moreRegarding before Not gonna happen +Given the observations of the reporter it seems highly likely that this is a problem with the implementation and not Xerces so resolving this report. +this patch includes the doc changes that went missing because of the split and docs removed from common which should have been moved to mapreduce +Im seeing a possibly related issue when displaying custom view only the first rows are shown with no next link even though there are rows in total +In the scope of one session we have only one instance of popup +Im guessing since this is not a Boolean it is assuming a result is true +Committed revision . +I dont have an Ubuntu to test this but I think I fixed the problemNorman can you check +This patch Unwrap the Top level Elements used in skelton and skelton interface to make it compatible with the Axis generated code +Is anyone working on adding eager initialization +etcThe problem is that show Locks also comes under the same category no locks need to be acquiredI didnt want to special case show Locks +Do we really need a warning for thisAdding a attribute may be considered obvious enough like adding a static modifier to a class does not warn about implicitely members. +using static inner classes instead of anonymous classes +Looks like this was already fixed some time ago +Is there some check that we can do just to see if sysinfo prints without trying to compare the full output +Fixed in CVS. +is available on both phabricator and attachment +this seems like a really silly and anoying mistake in +Make the Replication Monitor periodicity configurable +needs a new patch since the test cannot be compiled +Added a static counter to the image name and added headers +When the task is finished running the Start button is changed to a Remove button so the user can clean up +Fixed in trunk +Guess your issue is the result of that +I consider this implemented with the basic webear though mentions Arquillian testing but that isnt there +This could also lead to log files which did not get after restart +JD aside from the wait in I dont really see anything else relevant to this issue +get still require two input variables +Ill look into it again but it seems like an unfortunate situation where the data is written but the checksum isnt able to be written due to an OOM +Closing all issues +Have you verified that it works with the current git master +Added completers for config admin and features commands. +mean for meShould I correct something +As I opened it Ill close it. +yeah me too with sfrestricted +Added missing dependency and redeploayed redeployed +I check through there implementations and for most it does return the previous value but I did find one implementation that doesnt so far +One question +relative a path withwhitespaces in directories absolute you did that NAMENODEa path withwhitespaces in directoriesAlso test that all cases work the same with globs +original test case mentioned in first post has been replaced with maven project attached +Uploaded SNAPSHOT build number containing the fixNow closing this issue +Which affects the POST IDP is now considering +The attached patch addresses the optimizer overrides that were implemented with +A sample formatted source file could also be added or linked from Subversion. +We are forced to use that +took a quick look at the diffs and it doesnt look like you are doing the merging +Since It has more than lines Id better not post it here +replacing those lines in this case is of course no problem +Indexing should stay as isconsistent with mathematical notation +Will not implement too big a change for backwards incompatible. +This is an issue already already filed against maven dont know where it is thou +The project was restructured several times during that time frame so it is likely a lost change +will we support it? if we want should we add a PCRE filter +the names are truncated because the horizontal length of each tab is constrained +The cloned calendars are memory waste i think whether we can add on nextYear and substract on Committed revision . +This is great work Bryan +Thanks for the feature imeplementation! +Feel free to give the latest snapshot a try +A screen shot of the stack trace from the error report +is it implemented or not in latest release +Thanks Bilung +secs +Attaching and with the suggested fixesJust a sanity check in one paragraph its the JRE thats closing the connection but in the next paragraph its Derby thats not closing it presumably because its still within the block and handling the transaction error +The user that started jetty doesnt have all write rights on the issue due to new the file is removed by the file before the project is fix this we can copy the uploaded file in to use it in +Added convertes +of it should be a duplication of +Thanks Ben +Ill take care of this +Yes calling parents validation first will solve the issue +Amar pointed out a mistake that leads to +Thanks for the hints Alexander +I recommend that we start to use the prefix TS for everything instead of INK +thats good practice anyway. +We need another JIRA to further enhance the HIVE translatorSupport for JPA is added +As we decided previouslyit was reasonable if version could not boot because of bug that user could could use +Thanks Bruce +All same as youve done Caolan but there is also additional step back in frame for relative paths at resolving start due to at the final step current and id points to module source not module source level so without this step function would generate a little invalid module ids +Fix has been committed with regression tests. +For example wed like to put the Sybase READPAST hint in some of our HQL queriesThank you kindlyJames R +OK +Looking at this this report is going to be out of date as we have already worked on this further for ASDo be aware that the XML schema is not related to the parameter names for the operations whilst a lot of the time the same names will be used this is not guaranteed to see the structure of a command useThe error you report does appear to be a bug in that the wrong name is written to the XML during marshalling however from AS we will have added support for specifying the alias to use from the keystore so now we do have a attribute and an attribute +This fits in well with the upcoming changes for where the first immediate properties of an object are serialized thus avoiding the in viewCollections at a deeper level in the related entities may be fetched depending on the EAGERLAZY fetchType values on the nested collectionsAlso replaced Seam Render usage with +Updated patches for branch and trunk with changes as suggested +The fix didnt get into ER so here is a replacement jar that contains the fix +Hi AytekI was able to reproduce the same situation you described earlier +I just committed this to branch and trunk. +The resulting page of the patched contains a small black circle with a white in it +Its because contain not implemented stubsSo crash handler have no information about native frames and prints stack from Java stack iteratorThe changes required in are absolutely independant from crash handler implementation and could be made a bit laterBut and methods should contain correct line numbers +It implements what was called above as selection of Organizational Context +Well does it work today? If so we should get rid of it for right? If it doesnt work today then lets just go wild on it and kick it to the curb +to the template under the same directory as +I was going by Zhengs last comment +If user needs to send only the Created element in the Timestamp specify a value less than zero +One other item of note I didnt touch the metrics in the because Im going to fix those up in. +Correcting reason for failure. +lib +ill fix the schema to remove Kishan +Inner classes are not the only way you could have mutually recursive data structures +I committed more fixes for the failing components +Attached the patch +Verified +And remove is already synchronized in removeId method as wellBut is better for performance +Oh the Regression suite is the place for failing tests? Some of the tests in the regression suite pass so I didnt realize the distinction +Olivier specially with regards to Tomcat there is a non homegenous behavior of ajp vs httpThe n handling in HTTP connector is mostly legacy code +If you think about failures like that we should also work on protecting data in case of disasters +Fixed +Thats probably something that deserves its own JIRA issue +Proposed patch attached +Do you want to add a sentence to srcsitexdoc +Upon a bit more thought it seems to me that we dont actually need the shim +No longer relevant +Hmmm what should happen when there is a null? A blank entry? Skip the entry? Stop processing the list at the null entry and log in devmode +The design document is very goodWill get back to reviewing first +Hopefully this will be enough for you to test with +It seems like the resulting artifacts are not valid with the latest SDK +Silly +Good question +In that casethe second insert will fail and we will get back a return value of it would be good to verify that only row is in your test can you add a case when the table columns are being used for parameters the current test takes constants +I just tried the zip file snapshots for and dated and the both still have the problemI was able to fix the problem with the generated file manually by commenting out the import for and making the extends class use the fully qualified nameOh +You have a stack trace +You are right that Envers does need the module +This JIRA should be filed against the JCA project not the Messaging project +Will remove it then +Fixed in +that the javadoc comment was confusing +It may have been due to or but we cannot reproduce this issue any moreIf you can still reproduce this problem somehow please let us know details and we can reopen this issue +This looks like as far as well get on this forCommitted revision +Now the default renderer is created after the other renderersAdded an untyped renderer that uses other renderers as delegates +Reuploading Abhijits patch to kick of hudson CI on the right patch +We have a similar feature implemented for attribute class of a custom action +Right that happens because the write to that log is never flushed so the log file is never created +Pdfs will probably work if you can preserve the layers and editing depending on what needs to be updatedThanks +As of right now its kinda unusable for me +verified + are relevant and were recently fixed +cassandra from trunkbin i will have trunkbinstorage +Is this the most up to date JVM availible for RHEL +That must be the problem +joram will review +Not sure synchronizing on the lock is enough +DEBUG No Package Found for No Package Found for No Package Found for No Package Found for No Package Found for No Package Found for No Package Found for No Package Found for jbosssx No Package Found for tjws No Package Found for jaxbintros No Package Found for No Package Found for jbossweb No Package Found for No Package Found for packages for nowDEBUG No Package Found for No Package Found for No Package Found for No Package Found for No Package Found for tjws No Package Found for jaxbintros No Package Found for No Package Found for jbossweb No Package Found for in +Since its a sample and not core code and because its very simple I just committed it +We need to figure out how compressed is delivered to replication sink. +Your test file is correctly detected as a tar file for me java jar target detect tmp applicationIs there something else needed to trigger the issue +which version of the checkstyle plugin are you usingYou have to use with maven example POM would be highly appreciated +Ive reproduced this on X with a base version of Eclipse tarted up to the nines with a pile of stuff I dont normally use +Muhaha! Im quite stunned to see Im held in such high regard! Anyway Ive been told to close this out as the people who really wanted to do this say this will workThanks. +Added licensesHandler constructors all take a and instances nowRemoved all the unnecessary imports from all the classesAdded class comments to all handlersAlso I dont seem to be able to detect these tabs youre talking about +And for checking the +But before that attach +tests are probably incorrect and have been marked broken the TCK. +Retargeted the issue to since is being released very shortly +Thanks for the contribution and wed like to see more from you +I dont think we need to support running multiple Solr instances in the same JVM thats what multicore was developed for +For data integrity CRC probably is the best +These are nodes produced by the Xerces Deferred DOM implementation +I got some simple implementations working that are functional but I never fully went down the thrifttserver side of thingsI think the code I wrote would fit better into Keiths Typo. +Another example of the impact of this startup time is that if one has a syntax error in his or her build file each editcompile cycle will be at least seconds +Patch to resolve RobbieAs discussed +I think that might be a great newcomer is the trunk diff with the test +At least thats how I work +Integration test is in place. +So I added two new counters to show this Combine Input Records and Combine Output RecordsIm not sure if we really need these Combine InputOutput record counters + +have not provided any feasible solution to our problem except for the fact that it pointed us to the problematic code there in the core and was a great example of the way minilang may be extended +Maybe something changed in the related logic I didnt knowingly get tricky with the constructor order +The primary build of the Spring Framework is based in ANT at this time and there appears to be very little benefit for us to maintain two build systems for the code base +Attach a sample that reproduces the have a look at the issue i placed before is exactly the same problem with the file and m affected by this issue as well but Im not sure if it is actually an issue or built that way +This means a or a or the like +So although this fix is not correct hopefully the log and unit test will assist in fixing the issue for the unit test that made it easy to s now fixed on master. +Alexei I tried to compile as you +Thanks Paulex +The bug never existed +No snails at all! OK Ill set up a test case and see if I can make it whimper for me +Spring does not have that interface and so components that need to have auto startup capabilities would typically start within the initialization cannot make any component within Spring Integration implement the interface since that was introduced in Spring and the minimum required version for is still best way to overcome this startup order issue is to upgrade to Spring Integration +Then it fixed backslashes in the command itself and pieced the two sides back together + Preview mode and VisualPreview mode + +Closing stale resolved issues +Isnt it a featurerule? I dont agree that this issue is a major priority since it can be solved on a client side +No reproducible. +Thanks Dawid now shows an icon indicating the affected test is ignored and no exception is printed +But that is a good oneI cant repro the fail local +Having worked on the and Saurabh fixing I now think the collation needs to apply to all charactr types because CLOB and LONG VARCHAR can be used in a LIKE expression +This would make it easy for reads but for puts I would still have to keep it the same +Maybe you should spend some more time learning the things you want to discuss +I guess there should be an in Solr as wellI agree +IIRC all soap message is carried via POST but not GET +Verified on JBDS +If it seems like a good approach we can change the docs to use commas instead of spaces +Once you do that line changes can be removed +an entire repositorySo you can end up with a list of Available components from Available RepositoriesInstalled components from Edell +This was already fixed in to +I have tried to replicate this issue using and RC by calling the via a thread created from inside a servlet deployed in and I cannot replicate with it either committing or to running I have sent messages to the queue if I then run with the committing all messages get successfully I run with the test consumer rolling back then the sames message gets redelivered ad infinitum which is the desired am closing this as Cannot reproduce. +Summarizing and periodically outputting data is a theme +cleaned up codes added some this patch is ready to be reviewed +defer all issues to. +In the case that the unit test fails we could do ant clean +I havent debugged this more throughly to see the oddity +First patch is against +Looks like the new patch is missing +Otherwise were just going to print out the diacritic after the character its supposed to be over +I didnt want to commit given the current release vote +Attaching patch incorporating Rahuls comments +This weeks behavior and bloom filter growth has been much betterIs there a reason why this patch was not applied to? Are there stability concerns? We arent ready to make the jump to and would prefer not to move this table to Leveled Compaction if we dont have to +I think we need to discuss what features we needwant to implement for this cluster of issues as a group rather than adding features piece by piece without an overall plan +While it correctly hides the calendar panel when the input field loses the focus it also prevents the picked date value from being taken as input +It would be good to fix this before the RC is published +Its not to invasive. +Committed +. +All examples works nicelyAdded stdDeviation caching for Kmeans like the centroid +Hi Joe I think you forgot to add new files to the patch +Committed this fix +Closed after release. +So information that needs to be removed is cleared already +for EAP +MladenPlease take a look however we should be removing the examples in both Tomcat and is because the and is missing fromwebappslibPlease dont bother me with such m not Tomcat support forum +The jvm debug configuration is now a property only. +trunkCommitted revision +The failed tests were unrelated to the patchI ran the failed tests and they passed on +To my mind this sort of functionality is beyond the scope of such a simple tool + is now fixed so I think it should be possible to resolve this issue and if necessary open new issues for specific problems +non existing classes missing schema point declarations and so on. +now results in sending asynchronous acks after each message is delivered to the application. +Thanks Stu! +Neil thats way better then the way I do things now +Its also a good practice IMO that if you disagree with someone that you first put your point of view and discuss +Thanks NathanPatch applied to LUNI module at repo revision Please check that the patch was applied as you expected +Deferred after +And the patch looks reasonable though I dont know this area of the code well +What line is MS in your build? In current its the middle of +It would be most helpful if the folks that developed this particular parser could take a look at this +We rename it and have a successful recoveryThe second case can potentially block the view group for a long period of time as it scans backwards through the file +It is great functionality and solves many RIA problems +Its functionality has been replaced by the streaming functionality that was added several releases ago. +outputxyzampabampcd because otherwise the document is not well formed +Fs looks good to me too +Ill do some further testing next week so put this one on hold for the moment while I finish verifying with m quite sure the was included in the war because the application worked just fine in tomcat +verified by Vladimir. +bq +Add patch to use slfj to log info when strating the OSGI bundle of last patch file +teiidadminshelllib +Thanks for patch. +can someone commit this patch if you dont have any comments +Your fix has advanced the cause +All tests pass will commit soon +I will start work on them +Karel +ok this doesnt exist in my patch only existed after Ivan rebased the patch as this comes from Journals change +Once you done with your changes you can get the svn diff +I will look into it some moreIf its ok by you I am going to work with your code some more with an eye for adding javadoc strings and creating unit testsThanks problem +Steven I think there we should just set the property in lucene in the reusable test target +I know its not a big object but it could generate a significant amount of garbage due to being used and the JMX interface already generates quite a lot +This is important for maintaining compatibility between publican jdocbook builds +I was monitoring the repair pretty closely +Thank you Benoy +Looks safe to me to +sorry misunderstood the context. +Closed after release. +Ok thats good +The inserts then dont involve any array shufflingIf this proves a good win for performance I will work more on refining the change and expanding it to +Handling the transition is also part of a bigger question regarding upgrade support + Im actually using pom file for basic buildproject creation in Netbeans +So if this call flow is transacted then this property is installed on it +This will allow applications like PIG to ensure that their shutdown hook gets invoked before the object is destroyed +If AM process is killed at this time when RM comes back it will relaunch the job again even though sees SUCCEEDED +removed dependency from jars inside project and added dependency to bear with me as I display my maven ignoranceThe extra jars are checked in in the lib directory +Hi KatheyThanks for your input +Templates was updated +However any service can have its own TAS deployed to authenticate the identity token +showList and hideList are added to on CR +I was using the wrong default for the validateXml parameter +All the comments incorporated into the latest patch +This is a duplicate of and is fixed in SVN HEAD +Using the current code I can create an ESB project and create a Smooks file in the esbcontent folder without a hitch +Good point +too +Time is flyingToday I must edit some properies versions of depended modules otherwise build failedI use recent production stable version of Eclipse and WTPPatch of correntess is included +Looks like Anoop doesnt have account yetIntegrated to and trunk with Stacks suggestionThanks for the patch LiangThanks for the review Anoop and Stack +Mild panic unjustifiedThe mild panic however did uncover a possible problem for you to investigate +You set a short of retries in the client and hand it off to the policy to react to failures +Down in the region close its already taking out the region write lock +I look forward to some quite focus time over the next month to rebase it and to implement Sams comments +Merged into with revision . +But anyway do you have a fix for this? Happy to pull it back +Suite just didnt run +Dependency relation between and has been reversed. +here is a revised patch that doesnt do the second thing +To me this doesnt seem like a good thingSo in short that is the answer for now although it may change +bq +the suggested patch is not backward compatible +Well need to release of that project to get this in +I just committed this +Thanks Doug for your quick review +Attached file to this jira update to the patch that restores newlines in the doc strings when written by the velocity templates +is another possible optimization + svn ci containersrcmainjavaorgapacheopenejbutilSending containersrcmainjavaorgapacheopenejbutilTransmitting file dataCommitted revision +Peter just to confirm that the changes were done to and without the picketlink jars being available EAP was able to start up fine and the testsuite runs without any errors? If thats the case I dont mind merging the change +Shelly can you make sure we have subtasks to cover all testsuite failures grouped together as much as possible? +NiceHats off to you gentlemen +Its interesting how many lines of code need to be modified just to change g version springframework introduced a new separate jar for the bean factory classes +Ok Shane thank you +The ratio on the same bean with the correct units is wrong +If I am not able to create the junittest Ill at least add all config files java files and versions of all libs and a use case to reproduce itFrank +Ad I think the API names for is open for suggestions for new names +Thanks alot Scott +So calculating checksum while uploading the image from SNN is not reliable at all +This has been done for a while for the process governance work but not in the trunk +If you find further issues please create a new jira and link it to this one. +js files stay separate files +Tests above are unrelated +Would that be correctFor example creating the symlink above is pretty simply by just quoting the file name but if the symlink is something that will be executed then special chars like or even basic things like spaces can create problems if not quoted and escaped properly +implement reusable here too + +Remember the feature freeze date for is and the last day for merge request with hour feedback peiod is +The problem is not present in. +I want to be sure were OK with the design before I implement the backend +Fix submitted to SVN +Add unit tests for selector to python testing frameworkI added two unit tests +As there is already a cache of plugintypes to plugins there is no need to also cache them again in a pluginManager member of the configuration +The Java interface definition is attached +Thanks for your help. +How would the authorization service be implemented as anything but a security filter unless we added a new dependency to Spring MVC? I dont think its a good idea to add that dependency +This is the patch for the improvements suggested. +Cheers Tom +This change adds a new ssl related option to qpiddwith possible values This can be used to set the certificate store location that qpidd uses to find the server certificate it uses. +That results in dir which then should be passed to the classloader +At least if I run my tests from compiled classes +I just deployed new artifacts for both and +comments suggestions please let me more logo proposals +Hi Ericthank you very much for reporting this issue +Fixing this will also allow us to fix + +It does not make sense from my perspective to prompt for a web container to deploy to as youre already specifying the server youre deploying to which at most would contain web container +The signatures in saml messages is bit dependent on has been solved by. +Thanks Chris +The error messages simply point out that the configuration could lead to issues +Wed need to change the implicit name determination to make this work +There is a good reason why they are not being able to treat simple or complex schema types indistinctly +Great news +target verifies that a lib dir exists +I moved the external repos to a profile that I must remember to turn off when calling the site target +Regarding if statements which are never executed my preference is to delete the if statemententirely rather than commenting it out with more comments +Service A has forwarded the same message to Service B preserving the reply to EPR +Please look at ant binary not copying jar files properly +Proposed minimally invasive fix +As ODFDOM and Simple API have some little changes so the patch needs update +unflagged the patch available as there isnt one yet +thanks for the review +Looks like is a p patch instead of p +The new patch from RB +Im a Maven guy but I guess that is out of the question? This needs to be Jenkins ready? This should build Source and Javadocs artifacts ready for Maven Repo deployment right? The resulting JDBC jar file would contain the C dependancies from the build directory pointed to by the properties filesystem property? It would be self contained? The testing part of the build is going to have to reach into the C build area too I think +The query that works is select from ORDER BY oorderpriorityIf you add DESC to it it then fails +We have security related code under security and built with Psecurity +I merged the patch to. +is simplified to be just a cache for sync producersg +Patch submitted +Bit sooner than I thoughtsvn ci m Applying maxmin for Comparables as supplied by David Karlsen in srcSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +Testing Executed derbyall and did not found new error +I dont need to know the internal workflow of Hibernate Validator just debug information to help me debug my code +etc +No need to resubmit the patch just let me know if you agree +Also seconds might be a better default +Logged In YES useridYou dont write the URL in the becausethey do not support it +Odd change looks like it wouldnt affect this but somehow the highlighter test fails as it attempts to access a deleted doc +does not work in starting to think that this never worked and is not a regression +Reynold had run into a problem with that. +More than cheating is separating concerns and fixes +Im marking this issue as blocked by because that would provide support for the Socket transport mechanism which does provide support for request only messages +could not match patch due to import reordering +UghAnyway Im in a compilers class now and am considering attempting again with lexyacc or similar tool to maybe get this just right +must be a bit liberal for old applications. +thanks for identifying the problem +Even I tried with different combination of charactersIts bug in +Help in resolution is appreciated +BEAs Weblogic SPSP and IBMs server can handle such deployment unit correctly it opens file is able to overcame this issue despite an error message had been thrown it pull into classpath ALL content of EAR +If this patch is applied before then that issue is resolved +That test is not different from the others so I suspect an issue with the minicluster? +Hope thats ok +This is certainly a manual step at the moment and were working on tooling to make things seamless in this regardOn processing sources your imports will be resolved and your code linted +Thanks for writing up release note text though this will be very useful when writing the component overview. +So we need to set up the same cluster on different data centers +Hi Andrea I dont know how to write a junit test but I have attached the SQL statements for an example +ThanksId assumed that the mail software knew that the minotaur was to be ignored but obviously notIve now managed to get a response from the address. +I believe you see pros and cons better then me at this point +I ran the migration script and it fixed the user creation problem however my existing users bookmarks and newsfeeds do not show up +I didnt realize the comment would just be dumped in the main area +The CCE is around Proxy +The attached patch allowed compaction to complete and the region to split +Thatd be good enough methinks +Another possible approach is the one taken by contribpig where the build requires PIGHOME to be set and locates the jar that way +I will replace by wherever possible +thanks +mild preference for setkeyspace this relates closely with Ill spend the time logically separating these +META +I should also point out that the does not have +Reopening to set fix version +this patch +So you are using it in production? Thats quite a good news +I think the move of mapreduce to got lost in the latest +Note that the generated documentation is included in the distribution archive +added apache license to the wiki as required by the Gabriel I have gone through the license files and add information to the welcome section +Rob please review my patch for good +Also get a related exception when the entity returned from the query is different than the one that it is defined in +can anyone confirm? I looked an and have expected to see a null close to line but its not there +git rebase +Does this need to be done withink lognet? Can this not be the responsability of the viewer +Hm +It said no test executed which means no test has run due to some other problem +Git pull request looks is the status of this? +No it is not related +Ill have a discussion with him on whether we should just drop MD. + released issues. +has been committed +Need to test in cluster +I plan to commit this soon if nobody objects +for the patch +Closing this I removed this listener quite a while ago. +This is fixed with SVN revision . +Ill look into this ve run the test times on IBM j vm and times on drlvm without any hangsI think the issue isnt reproducible +We can do it by checking the ledger metadata stored in zookeeper +Thanks for the had one question is this problem exists for all the platforms? We have seen same failure in Linux also +Sorry to reask but could you try with the latest build from hudsonAnd one thing I dont understand +Burr I cant reproduce the bugDo you still have this problem or can we assume it was fixed and just close the issueThanks reproduced this happens for both jaxrs AND google content sure why I havet been able to reproduce before my guess is that I only tried using Ctrl level to critical since it is not specific files its any java file +Hope its fine +Closing all issues +Given that JDR subsystem was rewritten to pure java impl instead of jython based +Some notes about the patchAll history information are logged using events +Verified the fix and closed the bug. +Ive just fixed the bug in and moved the transaction manager implementation into Spring Data Commons as it is not strictly tied to graph database scenarios +Its the second read that seems to be erroring out +attached patch adds D to to allow passing of nodes in a file instead of on the command +as far as I remember the only issue here now is the XA Recovery +Sailaja I dont see an option in to change the network offering of a guest network anymore can you please check the steps and this bug and let me know? Thanks +bq +Its a very good project for someone interested in getting their feet wet with ZK development and the Apache process +There are still files that need to be updated +Thats very Pythonic of you +The fix is in +I personally see little sense in encrypting connections to HTTP proxy as proxies are meant to be trusted a prioriHowever the network between the client and the proxy may not be trusted there could be eavesdroppers lurking +base +need a wagon upgrade +It invalidates the skin but the skin is not redrawn update is not called for the skin +And enabling compilation of Groovy files causes problems when your run your Grails app +Forgot to mark it in rev +Milos and I will take a stab at resolving this during the week +This is the correct one +Let me know how the new diff looks +Thanks all +Switched back to +Committed to branchWill address the other as followups. +We have tests for these attributes. +Agree with Jarek +Thanks Suresh +Attached is the patch +But perhaps a warning during the configurewould have helped me check for the existence of a critical file for example. +is new version of the internationalization patch +So richselect and hselect have different behavior +A great deal of discussion has been done on this issue on the mailing list but no decision has been made yet +The kernel client has a better performance than the user client +Mark go on with the RC! +The way I ended solving it was breaking it into two queries first I selected all the keys that was going to be deleted and then I issued a delete query with the set of keys as a parameter +Thanks +Solved when using the snapshot release +I looked at the list and didnt think I saw which is really what I was worrying about +Parse the query with all possible dialects +Thanks Peter +Can you also check if count for failedkilled mapsreducers are correct +Name in Source is available for source column name differences and abstract those from the Teiid models +I like this idea generally +Could you please verify that this meets your expectations +Hi EvgenyThanks for applying part of the patch +The only outstanding problem now that blocks this branch from being merged is the Maven build is broken because it does not contain the proper nativeascii conversion for the properties files +But now that Ive got the script applying patches like this will be easy for me in future I suppose +Patch that adds new converter +I have prevented the exception in revision +Allow me some time to comment on phabricatorMeanwhile do you have performance numbers for stochastic balancer on a real cluster +Thanks for reporting. +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +For that matter this would allow one to see the permissions etc of the root directory which presently I know of no way to view aside from the OIV +Since it looks like will take longer lets move this forward +just one question Did you use the command line installer to install openmeetings or did you edit the manually? +That way Ill pick up any new goodness without having to examine every line including ones I havent changed +some fields in the Context class are protected others package privateShould probably all be package privateIll create an amended patchbq +shakes head +Maybe we should fix this for? It really is an API bug +To setup a POM for a JAR see jdbcjdbc as an example. +EPS contains logo versions A B with TM sign +Need to double check at some point that that rollover works happilyI also just removed the struts sitemap file given its jira is now gone +Anyways the solution as you have found is to point your local m repo at a directory wo spaces +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +add the patch the blog states the can not be read twice so it is a known issue in CXF? CXF can not support MTOM data save back to client +And at some point Derby will support such routines so designing with those in mind I would say is a good approach. +I will submit another bug for the second testcase. +The order by clause then typically references these names not the metadataSELECT INTKEY FROM UNION SELECT INTKEY FROM ORDER BY actually becomeSELECT c FROM g UNION SELECT c FROM g ORDER BY c NULLS only time when the metadata will be needed in the order by clause is with an unrelated columnSELECT INTKEY FROM ORDER BY executed through the engine should becomeSELECT c FROM g ORDER BY NULLS FIRST +Just attached a first draft of the docs of course that we must add browserdevices list to this document. +The Jackrabbit JCA Resource Adapter must have already been installed in the server +is this acceptable to you +Regards SOA TP Release Note could I have some clarification of what how this behaviour. +Christian this is very interesting +committed to trunk and +Giri is working towards a fix for this issue +ted ok so I dont press submit patch then +Ive looked at the code again and I cant spot anything obviously wrong +Committed to GT +How was this one created +Drop it +But it is not recommended to deploy before capacity is recalculated as in some cases this might lead to greater capacity allocation than what is availableIn case of kvm we need to create a separate title and mention that Kvm dose not adjust the guest memory dynamically based on the contentionRegardsBharatAnd if the overcommit ratios are dynamically changed the +Slightly better but still ugly patch + +Especially the new features imply additional processing time +And you can change this default later prefect +and from an API perspective Im not sure this kind of error is functionally any different than that kind of error +assign to Daisy review +Seam has moved to RI a while back +First phase of this massive change renaming toCould someone please review +Claus done +do not pack +It seems that windows is a little pickier than OS X about multiple attempts to write on a closed socket +Please check that you file is saved as you show the hexa decimal numbers of bytes of your script file +Installation to clean Eclipse works have reloaded Update site restarted JBT but got the same errorIm installing not to clean Eclipse but to JBT mentioned beforeClean installation works +In my project I make sure to use +But thank you anyway +Actually the code does look right it creates only one thread per node + +Closing. +the patch is against standard source +That said with the change to allow physical schema to have any virtual construct these assumptions may not quite match your expectations. +Logged In YES useridI was working around here so I fixed this +Now due to the system table assignment its enable is delayed +Apply the patch and then run it from the hbase as +The installer cannot have broken examples +committed. +I just committed this +As long as we dont want to split the API to a separate component things like remoting IMHO fit best within that same component +Will make more sense when we get the gdal going +I followed this convention +Using the XHTML minimized syntax on output destined for HTML compatiblebrowsers will cause problems +Should be targeted at HIBERNATE and affect versions and +Search by time +Wont fix problem in JSP view t Fix +All plugins have been updated with the extra locationAll whiteboard plugins have been update with the extra location except for the followingNone if the above have and so havent had the LM treatment yet +Patch file is attached +maybe at some point we should completely do away with it and change all of HDFS to use Subject too +ve just committed this. +for the just committed this +The complete WARs lib directory is missing +I have coded a fix on my windows box +we also encounter this problem frequently +batch transition resolvedfixed bugs to closedfixed +Chris Patch has been applied and tested on AIX and it works +Committed to. +A corner case where it may matter is a signed archive +Nevertheless I dont like the current behavior +When I debugged this for a similar case I noticed that the pk passed in during the case below will be the string ITEMID CATEGORYID +NN should only allow to add new replicasDone +Will move it further down if requested. +Enable sorting by the parameter truefalse +Committed to with revision +Marking this as Patch Available sothat hudson will run tests +This is expected behavior +Thejas Im confusedpatch sets to falsebut wiki says set it to set them to false I see Lease checker threads created on each query execution +I like the current appoach of changing what is stored in SYSTRIGGERS +Getting the same issue locally and for other tests that also use ZK +I am lowering the priority to critical +Document patch submitted to be committedFILE the patch file Adds Chapters and Updates Chapters and Updates to correct build problems on LinuxFILE PDF document generated from the above patch files +Attached patch makes the test to beNote this is not only problem with this test it also fails because of JIRA issue +Per comments on the thread referenced in I am concerned that this could break clients that try to unpack +Logged In YES useridIts not defined by the servlet spec that the principal information should propagate across the session irrespective of permissions +If you send me your I will have a look at itShould the author tags be removed completely? I copied parts of the archetype from the old one because I had some troubles with the created archetype +The reload operation itself does not work but only on the server launched by the testsuite +submitted patch +Please do not comment as if the current work is not considering it +for the patch +It never occurred to me since then +I should have mentioned that I did run tests at the time I posted the patch Im them now to be sure I didnt miss anything +Thanks for reporting this you are right has a bug where it wraps the readerHeres a patch. +Please use latest snapshot of Metamer. +Fix for testcase and also interface change in +this fix will be in that release. +Thanks for your mail +Oh I see now +I wasnt able to apply this patch because it appears to have been created against a snapshot rather than trunk +this lets you distinguish between multiple solr instances for tracking purposes +For larger clusters we would implement the blobstore optimization later +Ive filed to address the removal of the now obsolete attributes +Assigned to me scheduled for is a gt problem we should open an issue there. +patch until it gets updated +DoneThe svn portion is also ro permanently. +And so I have a question is it right behavior or notAnd the second question is What is the main difference between Refresh and Rebuild Configuration +encoding is not desireable when redirecting to a php page for example +This looks like a ligature charater issue +This is outstanding Chris! Thank you! +does make in the c directory still working with these changes +Looks good thanks for checking it in +closed +This is different to the import on demand in Java +I had not tested this since the upgrade +James Looking forward to your findbugs and ivy stuff +Ive just committed this to trunk andThanks a lot for the contribution Arpit +The bytecode size change before and after inline nextDouble are method new factor +There is a missing message bundle in the following warning message of Quick Fix Tools Knowledge Base problem NatureBuilder is not installed on project +Just forget to address the indent issue +The s changes are fine but using generics all over the place is something we can do on to future this is a long task and it is not something we need to do right away volunteers welcome +Thanks Jason +Committed the patch to trunk +Joerg would you please expand upon what you mean by not keeping the local repo in the profile? While Im sure the documentation has plenty of information it is always helpful to summarize or at least link to the specific text that is relevant to the also would have voted for this as GB mysteriously disappeared from my server over time until I thought to look at the local repo +Right the one is not referencing the event typeAs for the rest well thats the trick no? Its the part everyone else requesting this chose to +This worked well all the time +Im very excited about this addon Id love to see a nice JSF integration with roo unfortunately Roofaces project looks a bit checked some implementations showcases and primefaces beat them all right now + +coretrunkrepository +Minimal set of configurations for reproducing the errorLog files for the steps use a better erorr message its not finding an update log +same patch for our friend jenkins to take notice +They should specify LDLIBRARYPATH in the child mapred jvm env config +The license files should now have the apache license +Can you explain why it should get two events +Closing this as a duplicate of. +the binary release seems to be working fine +Can you submit it to JSW project so well can include it in next releaseCan you send it to me too? i prepare release so i can include it if it works +in my opinion it should be fired almost at the end of the task activity execution since everything shall be already initializedAs you stated before these two are completely separate events so I dont see a problem if both events will be fired if defined in the process definition +Because that part is just for loading + is from the is the server log Released +Marking this for +When its done it should just output done to the browser instead of rendering a view +The patch with the fix is attached +Added serialization to all NMS exceptions +I just implemented a slightly different fix for this that does not require synchronisation on every call +Corrected the title and description +Cannot reproduce issue +This patch isnt applying for me perhaps it needs to be rebased against current trunk +Dont divide by count in But at that point youre very close to outputting counts anyway! So Id advise you to just use the item implementation anywayYoure talking about thats not part of the implementation Im suggesting you use in order to effectively use longs +for making table column mmsubtablelabel larger +You can find me on the infinispan IRC channel if you have any problems +uploaded refresh patch that includes test +If we keep these we have failed with wire compatibility and all the protofuf stuff was for nothingPutDelete is still used as writable at least in these cases +FrancisThanks for your work and patience on this +i dont like shading under because that will have additional classes and in will become a separate module +If you look again youll see its different than the column metadata +Please verify and close +Should we keep things the way they are? I will add a note to confluence explaining the restriction +Turning off the workspace resolver means I can only resolve dependent projects if they have been published +This fixes the issue for now +If you would make that property not final that would help too +I believe you could clean up the code as well +I read the diff and it looks good initially +Thanks Rick! Attaching and with those correctionsAny other changes? Thanks in advance +This is a proposed interface +Huge time saver Thanks again +Verified on CR build +Any triage on this yet +Having the first occurrence in the results not be in conflict would be misleading because it is in conflictTheoretically we could add a check to reject bulkdocs updates with duplicate id rev pairs but that doesnt seem like a good idea for some reason +They were moved to the main distribution site and the mirrors a week or so ago +As we spoke offline there seems to be edge case where not updating ZK could lead to bad things happening +will review and fix after next for the feedback style view was not made part of the default udig perspective and the workflow of the style dialog is more clear. +Thanks for review Matteov adds printing out duration spent retrying failed write +please reopen ticket with additional details if the failure still occur after update +bulk close for +Do you think it makes sense to backport it into instead of keep fixing the damn thing? otherwise +Adding missing file and incorporated Vinods feedback +Hi RobThis seems to be because Jetty by default logs to stderr which CARGO in turn transforms to a WARNDo you by chance know how to make Jetty log to stdout insteadThank revision for a very small improvementNote that this is specific to Jetty and to fix it we need to tell Jetty not to output to stderr but to stdout +Marking closed. +Reporters of these issues please verify the fixes and help close these issues +This applies to both Trinidad and Trinidad +Just to be clear the sql semantics are Arg Arg Arg OR Arg false false false false true true false null null true false true true true true true null true null false null null true true null null null Arg Arg Arg AND Arg false false false false true false false null false true false false true true true true null null null false false null true null null null null Arg NOT Arg false true true false null null +I have chosen to record that information in the file itself +changed warning logging from previously erroneous situations to debug messages +Andreas Lehmkhler asked me to upload an example PDF for the bug described +However now that we patch this appender I would like to patch it good +if I run mvn install on project root then I get no problem +yes +Maybe we should dig into how to integrate with Ambari +Is this what you want? bq +If yes I see no problem with that if pooling is generally turned off +messaggi tradotti messaggi non tradotti +Incorporated this change +Theres some minor style issues with code blocks that dont meet the Maven code style but the functionality is improved +So we are covered +I guess there is a misunderstanding +BTW I dont see Update content ever being supported for scanner deployments +These boxes run about K qps +bq +Its a duplicated issue with +Thanks Ahmed seems a rogue caused the issue you notedI now attached a fixed version of the patch +Just tried our application with +Commit changes it. +Ok we found the error in the template engine we are using for code generation +I wonder if its always possible but I bet you can write a random test to ensure that +Such kind of corner cases will be difficult to work visualizing part one can probably contribute a label decorator or special model for Project Explorer that would do some fancy stuff +In Core we are moving away from checking in versioned documentation +Closing all issues created in and before which are marked resolved +appled and delivered +bq +Can you enclose the above command in back quotes and display its return valueIf we want to display the result we simply need to remove the devnull +While we might probably build Aries and anything else and add to the repo the problem here is that the productized artifacts are having the dependency trees changed +to trunk and x branch +Note that I havent noticed so far XML that properly export the XML service as a service +But it still does not explain the exception to me +I got here because I got the same Invalid char have reread your posts and I dont quite understand what you mean about interpreting the directory path in a require statement +I will think about it some more but I am not sure I am OK with swallowing the exception when an underlying connection that has been closed without the knowledge of the pool is passivated +Still same issue +Hi StefanThanks for the quick reply +SystemLibraryFrameworksVersionsHomebinjarsigner verbose keystore UserswoodjLibrarymy storepass keypass signedjar VolumesJaguarUserswoodjDocumentsprtprttrunktargetjnlp VolumesJaguarUserswoodjDocumentsprtprttrunktargetjnlp woodjexecuted on the CL cause a bus error which empties the jar file of any content +jacques for your interest and followed on this issueGil I updated the path with your correction +bq +A workaround has to be done. +Attached is a patch to move derbyStress back to derbyall so that we can run it with MB heap +Will need to investigate further +The patch makes it fail fast +One year ago you explained my approach is not sufficient but the issue should be embedded a upcoming rewrite +Emptying the statement cache when a shadowing function is created sounds like a cheap adequate fix for this rare +Im not getting CONNECTING state too +Committed +It should work now on +Made use as default if no others are declared +fixed on revision issues associated with existing releases +Marking this as resolved. +Press Apply button +I need to put in my deployment you in advance +to commit +No way I can access a CVS repository out in internet land through thiscorporations firewall +. +So the answer to what the elementType values should be set to is that they should all have the same value as the name attribute +Fixed patch is here +It could be affected by some background task +I saw some patches going into Hadoop itself to disabling the need to have the Oracle specific JDK +Once thats added it should be returned as a and used to determine the home directory but until then I think this is not worth addingNote that this patch does not change how the home directory in HDFS is computed it only adds a method to expose the home directory already implicit in HDFS +Attaching Ms db if the is correct on the host +so it might be something else +Manually tested +v uses inferring approach +I saw the same behavior with svn revision yeah that fix was in revision so it looks like its a different issue +bq +Why is this issue still open if the patch was already committed to trunk +Found other places in the code where that message is thrown +It produced the correct result so Im pretty sure that does this +I will raise a new bug for that. +Note that a couple of the need and Transactionfrom Database +This patch is to retrieve the productId from the the title of the auction in case the sku is not present in the xml document +But it is impossible for the compiler to select the correct method for invocation due to multimethods +thanks Ivan for correcting texts + the patch looks good +we should have a queue per collection and subsequently we can have a queue per shard + updated to have close method +Thanks Tom +Needs to be reviewed by someone who has experience with ASF distributions maybe from a mentor or Thilo +The test passes on RI and fails on Harmony because it throws which is against spec +python libcloudtestcompute see you removed the patch files was this intentional +Hi JamesGreat idea I will do super awesome thanks! Gonna start using it as soon as the next perfectus completes +John did you ever write the documentation you talked about for thisFurthermore do you plan on implementing an easier fix for this? It would be pragmatic +Task is in progress because of the enough amount of RF components +No we should not deprecate those methods as end users should be able to get the Route instance if they needIts just the lifecycle of the routes is best handled using the API from to ensure graceful shutdown and whatnot. +I want to make sure we can support all features Amazon simpledb offers without the need for an API change which we cannot do in a micro release eg Camel. +It will be indented per common code practice of course +from what kelven mentions we did not log events for these actions at all public static final String EVENTACCOUNTDISABLE public static final String EVENTACCOUNTCREATE public static final String EVENTACCOUNTDELETE public static final String EVENTACCOUNTMARKDEFAULTZONE +This bug has been fixed in Xalan Java +Here is an attempt please provide attempt had issues +Bumping priority to original one. +For a Statement object on the client a new is created on the server each time one of the execute methods is called +In there you can choose the things that are filtered +Both of and should receive this fix +I dont know what the meaning of that is or if its significant +yes we need to have recommendations for it +on commit since it does doe DISABLING +Ive circled back to Mesos with a General Sherman approach to the build mechanics and will try to post an update to this thread by the end of the week +good cache peter that is a bad example +Hi RonYour suggestion is very nice indeed +error crashed mashine is jdk i am not sure whether jdk make tasktracker crashed +Closed upon release of +Looks good +I remember I didnt import anything to ssocircle but the version demo works but doesnt works +can we make cell level an exceptionWe couldbq +plugins +This would empower the user implementations to get out of the bad jobs asap and achieve better cluster utilization +it would be nicer to use only if exitCode! remove this comment the exit is here to guarantee that we actually exitalso fix the exit code for Exception back to +Currently this is probably not easily possible without extending the template and adding special code to get the stacktrace. +Hi Aaron it is federated +fixed in BranchSPCP +In and splitting is mostly a client side operation + was serializing the actual exception and sending it over the wire +currently not planning on backporting any farther than +Thanks! was my problem +bug part fixed post improvement request it possible that the improvements listed above will be implemented in the nearby future? +Dali core library required. +Would anyone want to try this patch +rob +FixedThanks! +I am running with all the subqueries are working except for the few associated with this JIRA +Closing issue Oozie is released +This was done so that first time when we try to create the znode with RSZKSPLITTING state if there is any exception rollback was not taking any action +Agreed. +Committed at revision +bulk close all resolved issues in preparation for and releases. + push to +be covered as part of Public API work +Now that work is well underway with AS all previous community releases are +And it is not take into account in the solution +The example can be made to illustrate the problem by commenting out the Entity tag in +Making a checksum optional and off by default sounds good to mebq +Thanks Sean for the pointer will do it later today +This pull request has been merged into master +Hummm +Hi PaulIm still working on changing the BOM scope to import there is a transitive dependency resolve failure that also was why I had to comment out the scope of in +Heres an idea to fix this add a shadeclean goal which removes the jar and let people bindthat to +See and +this is a nice way to say welcome! Giving you a hot potato right on the first dayWelcome on board +This is definitely very strange +Please see for more details and the patch used to fix the will need to review the patch for inclusion in Gary is a patch jira already in progress? +SeeActually this appears to be a duplicate of which was reopened when it was committed due to the same issues +Anything else you need from infra +If any of them are older than minutes they are marked as failed +Is that rightShould we commit this as a step towards a complete PHP RPC implementation? If so it would be nice to see some tests added with this code before we do +And in one of your plugin you have your own xml editor not default as I when plugin tried to get documentProvider from your Editor it got null and NPE +Fixed in current in svn +We can and should maintain a common test directory at the root still for the thrift files and anything else that might need to be shared +Lefty the change looks good +bulk move of all unresolved issues from to +In revision the throws instead of +Updated patch only record offending column qualifier if access is denied +Thanks! +Committed to tez branch +and especially will probably need to be improved to support kerning for all available glyphs +Let me know if theres anything I can help with for the integration work. +Here is a simple patch that sets the default to tests added but manually started a server and if not defined it started with G heap +Renaming Coprocessors to Plugins +Was accidentally duplicated by thanks for reporting +I could try changing the default so that it has a better chance of working right out of the box but thats not critical +Lets add a note to the FAQ to turn off cores to avoid the worst of this problem +I didnt want it up there initially as that tends to find a lot of nitpicks which this patch Im sure has lots of +Bulk update component to Hibernate +snjezanas patch has fixed it +Fixed in master and s not yet verified however +Reassigned for backport +Mark kindly looked into this and sent me an email confirming it works OK on OSX but is due to a buffer size limit +Im going to make a few more changes to the docs before I commit thoseHow does everyone feel about applying these changes to? Im thinking we might as well as it doesnt break anything +Currently tck cases passed +Thanks Vinod! Addressed most of your comments except the followingbq +Sorry you are changes I needed to make for this are in in the linked issue as the archiver does all the actual work jar just invokes archiver +Added a patched version of +Attached file an original of the publish task flowchart in case further editing is necessary +I think these changes should be reviewed to see if they are necessary andor consistent with the discussion of the two appropriate scopes of encoding issues on +I think that I just made things more confusingReferring to the documentation I believe that I am using the strategy in section Enhancing Dynamically at Runtime +Btw some of the information I put into Environment section previously but its not really visible there +v +exception view from eclipse error is not exception it is our message that class path could not be processed +Adding test cases +See my workaround there +This patch should do the were not using absolute any more so this is no longer an issue +committed close for Solr +Neither Rex nor I can recreate this problem in the latest build +my app config s the version you were using earlier? The component was first introduced in JackrabbitDid you upgrade the Lucene dependency to versionSee also the Upgrading to Jackrabbit and Upgrading to Jackrabbit sections in the release notes for more details on the upgradesIm resolving this issue as Invalid please use the users mailing list for support questions +Merged to with revision +But this change breaks the tests spectacularly so clearly it is now undercounting for some reason I do not understand +Saw no problems +Older clusters will speak and not be able to support spnego +Ive attached my +Ben has apparently started working on this + +Thanks Vikram! +We are getting the same issue in our there any resolution for this circular reference issuePlease issue is expected to be fixed because is fixed. +Also you should not setthe state of a bug report to RESOLVED FIXED +I think it can be closed +It will be good to know other peoples opinion on this question +Please check. +Allowing side effects in render actions could potentially lead to unexpected results that would be hard to implementation of would be cool but might be a large change this near to with the current interface returning for the various scopes +or get it off a environment variable +This problem is solved +how about integration test ruleI havent seen anyone insisting on a test for this jira +Is this the file? coresrcmainjavaorgapacheaccumulocoreiteratorsuser +I was not aware of the logic that resets the state when an exception is thrown +Changes to the das runtime for +Dear I tested geoserver again on trunk the latest revision news! I confirmed that this issue was resolved +Note access to the field itself must be synchronized as well but you can generally do this with the same lock +Possible with Sonar +We should move it to Always On once we make it currently some of the functions dont apply to Standby RMAlso the patch depends on +Execute derbyall on jdbrc environment and found no error except for intermittent error of derbynet and derbynet in framework +With a distribution Gant already installed run gant install if you want to avoid all the tests being run then use option true b +Fixed in +No response assuming ok. +projectfilehomemaxrhdevstudioserverdefaultdeploytesthomemaxrhdevstudioserverdefaultdeploytest +Sergey Im going to commit the patch with the modification proposed by Alexeythis modification does not affect default JET mode it makes an extra option to turn BPP off do the meaningful changesSo you can check the initial patch only +I will commit the code after I complete testing +If the patch has been committed sureFriday March AM you wrote +Which means the class static cache is technically unnecessary +Lets see if any problems arise +Actually the condition is testing for jars that should NOT be included in the resulting war so I think what I meant was SCOPEPROVIDED and catch +Thanks Zheng +I have a test case in the main varchar patch but it wont work here since that type doesnt exist yet +I thought was available for download but its not +same patch with a standalone fix is already committed +Daniel and I had discussed something similar a while back +Thanks alot +It is a simple class backed by to check if a name is in our list of datum aliases and to return an array of Identifiers with the aliases for a datum of the datum alias information solves most of this creation this class is used in the WKT parser to add aliases to datums when they are created +The output schema should be a set of records with the added fields for and +Delegate the creation toThe reason of this bug is because Pig use to handle IO in local mode and use and to handle IO in mapreduce mode +No need for Reset +No this works fineBefore the task is run the does a set to the working directory of the job +Hey thanks for the explanation and FWIW! +We should have this as an option for all +Same Poblem here but the hint with the double click on the error solves the problem only at the first time +Yes it actually will improve things indeed +Do you need to checkin specfixturesprojectscordova +I played with the attached sample under XP and could not reproduce any locking issue. +I dont think pulling in full oozie support is something that we should do on the first change +Id love to see the project support webdav well +You are correct I appologize for cheating a +The build failed before getting to any deps verification step or step +Maybe the tooling team can create blocking Jiras on all the underlying projects such that the underlying projects have to completely describe possible configuration options before going forward which could then be used by the tooling the wizards could be setup to have some generic aspects where there is some property or XML file in the plugin that lists the options for a given screen +The clean +Retry hadoopqa now device has been what Ill commit did a rebase and changed one true to the new default that I missed in +Lets see how it does +for got to mention that commit causes hte test to pass reliably for me times including all of the reproduce lines from the last failed nightly +short term and long term are good enough. +AhThe above stacktrace is from an end user on camel IRC +However the application code itself does not do a further check in the POST resource method nor does the application actually use the username or any other security related information +Closing this JIRA. +Will turn the rsync on nowI can only give you the email address for the kahuki mirror it is contactThanks Manuel dont worry about wasting our traffic for a few days I like to see the mirror performing properly for a few days before adding it in +DIH assigns a random name if not specified +leading me to think that the linefeed is preserved but one ends up with a lineA way to fix this may exactly be to insert such a in the code somewhere whenever there are two consecutive linefeeds +META +Thanks for the to Jochen Added to HEAD +Thanks for spotting this hey this is sandbox code What exactly are using for Seth? I would love to learn about use cases for it we need to decide whether we move it over to the main for commiting the patchJust experimenting now with all the different Spring JMS support classes +We would need a new EAP and new EAP this a look +Set several values in the MDC context +Todd I just wanted to post a patch to get Jenkins results and to give an idea how a merge patch looks +It is not reproducible when int volatile variable is usedI run the test on and machines +seems to be a much better way around this problem +toolsmarvin find +This is the case now for me with a bug in the switch security module fixed in trunk but not yet released and may not be before I need to release my s end this discussion agreeing to disagree. +At some point I was relying on the fact that I could know the size of the CLH given the number of +Good to hear we managed to fix it in +Start by telling us which build it is +ve just committed thisThe change is too trivial to have a unit test and the failed contribute test is not related to this change. +So for example Im not sure how well be able to share an or with the Spring Integration project if its got all of this stuff about stream state +Any thoughts on this one +should be fixed with and without adding sys props now +I reviewed the changes and would like to request the inclusion of them into for +We can also extend JDBM to spill only when needed +Ive added a corresponding warning to the namespace ll revisit this for Spring +Fixes newline substituting environment variable +Hi Matthew looks like youre using Sonar and not whereas this feature is only available on this latest Sonar version. +Will create a new issue to address the error. +I should not that Ive modified the Comparator that stores service parameters in such a way that I thought the path parameter was always guaranteed to show up last in an asset URL +Verified on Tools +downloads page updated in r but the KEYS file isnt where it should be in distarchiva +Patch was committed to both api and api repositoriesCommitted version . +For the first one Im not sure I understand what it should do because the metric selector never let the user choose new metrics +If it helps in it the srcmain only contains a sub directory and a +Thanks for reporting the bug I also track issues on just search for the installation script to include pdf works for me +Moving discussion to. +I am investigating this and it seems there is doubled up use of margins and padding in adjacent areas of the stylesheet +patch applied to branch firsttrunk will mike +Thanks a lotBest regards by Richard. +Will submit the patch again +All the references Ive read suggest it for server sockets +Dan managed to work around all the effects of this in Infinispan so slipping to to reduce risk for with NBST the grouping information is now kept in a wrapper that doesnt need to leave the node. +Shouldnt we make and tooOh and thanks for your help Todd +We can not reproduce this error without the wsdl +Looks like a dupe of? +In other words I think we should keep it with warning what do you say +Another option is to use an exclusion filter for files that start with +Hi RobertPatch applied at r Thanks for your contribution +That approach is replaced now the NATIVE authentication service opens a connection to the Credentials DB using Derbys own skipping the need for a nested call through JDBCWith this patch the test case now runs correctly on Java I have run the successfully on the following platforms Java on Mac OSX +Attaching some screenshots of the metrics dashboard for a test case run to give people an approximate idea of how the output looks like +I could find no reason why couldnt be completely lightweight +I was rather puzzled when I found out that uniform distributions were not available while they are so trivial as they correspond very closely to the underlying random number generators +I merged the changes into the branch in revision I need some help for documentationIs there some doc on how to make doc somewhere +I think I understand now why we need here an composite key instead of the single Id primary key annotation +The same error string pops up in where the datanode refuses to play +And simulate a failure in the nd split causing the message in the DLC to be the original input message in the route +Looks like there was an issue with our permission migration last monthPlease try nowThanks +for some reason the components are disposed when the wizard is entered +Sanne through Spring and +is this a new issue or have it always been like this +It is cleanup from a bug that can cause pretty serious consequences +Fixed +Maybe they will be made public in future version +Attaching which describes my findings +I confirmed with Rajit and he does not see it anymore as well +The latter the whole memstore would be layered I think its basically like we do with snapshot right now except that there are multiple layers of snapshot and we have a trickier way of doing it without locks +It was originally filed agains XSLTC +This was expected to go into +Brian is this something you want to do for Beta or not +This is a dupe of +will acutally be rendered obsolete in +Could you please attach your test case? If not Ill try to cook something up +Going to run some tests over the weekend for this and as discussed. +Yes this plugin wraps the ant plugin you can run antrun or you can use the plugin like the file attached to this issue. +I just committed this +Proxy user per context needs to maintain connections per context +Committed to branchx and trunk +Confirmed and exclude list updated +Did you mean to assign this to the stream as well as or just +The failed test is neither related to nor is using structs in its DDL +The URL pattern is wrong only in the console SOA overlay +This is behaving as per the SDO spec +Closing. +whereas the comment in says The query was given a new name to allow the old query to be used by +on what resolved means +Thanks for the great turn around time! +Sorry for the delay I screwed up with my svn commit +OuchThere is a very serious bug somewhere +Is this something that can be integrated into the module for use across installer users +the first would be overwritten by the last +Hey FredIm looking at the code and testing it +corrected new version to and filed PR +Specifically which only packages a dependency if it is not optional and it is a runtime dependency +Yes the fix is for +Cool cool thanks stack! +For this one I restarted the master but that did not solve it +I didnt know if I could trust retryyum yet but if it works it works +reuse karafs bundle hot deploy mechanism +Can you throw this in RB? Either way some initial comments +TODO merge changes in in all branches +made implement as thats where jobRepository is set +If the document is a file url or jar url or whatever url the relative import will still be the same type of url just with the resolved path as described before +Oh +On which AS do you expect this working +For me the easiest way is to add just a little bit of code to the hit collector I already have and be done with it +after trying to get timings and getting failures in a row I did an svn up just to be sure and saw some updates to the testcase so I guess I did svn up in the wrong directory before +bulk defer of featurestaskswishes from to + fix the javadoc links +Wont fix for lack of interest. +The Junit fails because the port is alreadystill open for sure +Committed the a patch to trunk with revision +Solved by u is the JIRA for the upstream project +Contains a tests which do not have parser exceptions +Could we pass in the load balancer to use into the AMs constructor rather than call a setBalancer method +Thanks +Yes the behavior changed in version +You could call it rejected though +There is a bug in the patch for the JMX handler the old syntax is not accepted any more by the handlerI have corrected this bug and added a test case to check the old syntax compatibilityThis new patch fixes the problemIm naughty +I will refactor the code soon +CharlesCan you create a patch for srctestjavaorgapachecameldataformatbindymodelfixsimpleI will take care of the rest +Will as part of that and close both these issues out if the result checks out up on ASF Hudson +Put it on and launch Explaintation is the status of this EmmanuelShould we chat on IM +This issue should now be solved using available since Sonar. +the new version of all tests passThe problem was mine +This should then happen each time when the activiti jar is updated in the alfresco codebaseFrederik is going to ask the Gavin and Nick where we can include extra test dependencies +Removed +I think youre rught to focus on JPAnow. +Merged the patch to. +The patch has classes in the package Derby I assume these would need to move into something in the areaThe new files have the old coopyrightlicence header so would need to be fixe dup to match the current styleThe code has hardly any comments this stopped me from reviewing it any more as I dont have time to guess or try to figure out what each class and method is meant to be doing +New version of the patch SSL certificates were removed from the SSL transport constructor and are taken directly from the connection object +Thoughts +Its hard to see why people would want to join on two different key types when one of those is numeric and one is string +Resolved as not reproducible. +Ideally Spring Portlet Bridge would be fixed by the developers though +Do you really intend to delete the following in the patch? systemtestmigrationtooltestsuiteconfig +Theres still a hardcoded port in the jamon file +Changed the index loading logic to perform only a single scan over the index files then build the metadata in memory. +Also please check if Out of memory occured during uploading processing +NamitThe common is created new for every query +Thanks! +Freeman thanks a lot for all the patches and findings you have submitted recently +Well the one thing could do that we dont have a replacement for that was usefull for is load testing it might make sense to leave that part in but rip out the guts and replace it with using the +Maybe someone will come up with the script +All characters were Javas own encoding +Ok this looks similar +Closing the resolved issueLahiru +Hi Jesse you probably want to check this out? I will wait on the release untill I hear from you +This is interesting +The has been tested locally with this patch and some improvements have been suggested +Thanks! +pushed +Attaching a unit testfix patch for this +There will be header in front also PB so far it will only store the compression flag +If someone wishes to take this on for commercial gains or something else then please reopen and provide a backwards and forwards compatible patch for us all to use +OTOH there really seems to be no reason to disconnect the filesystem weve had mounts lasting for a monthIm hoping future versions of libhdfs will help out here so Id like to keep things factored out into doConnect and doDisconnect +Hi TomasI think that this might be a duplicate of +Run ant package twice and the same thing happens +Committed revision +Im inclined to close this as invalid +now that I thank about it I thank I tryed a older version of trunk my mistake it applys to trunkI will run some bulk import test soon on it and see if the compactions work out Im running tests too +That seemed to have bloated the numbers +Making them optional would leave the users in the lurchSerializations have complete control over how they serialize their metadata +This is not true +These changes and fixes have been applied +If PROJECTSTAGEDevelopment is used loads the uncompressed version on but if is production the compressed version on resources is used +One thing we might want to do is split the versions configure different slfj api and impl versions like camel does +The patch still applies youre lucky +resolving +v of patch that builds against trunk + +HTML files are included for review +I also added change to print net elapsed time of the algorithm +patch looks good +The server instance would then first look for configs with the prefix prepended and fall back to the generic ones +The probleme is taht my installation was bits butthe gallio host executable was +I have this idea that you should be able to reconfigure your cluster by whirr I like this +verified. +This is a small example on how to use those methods to expand or collapse a graphThe case is very simple we are creating some vertices with arithmetic values +Found a copypaste exception error Thanks Jack youre right committed fix to trunk branchx and lucenesolr branches +Thanks for reporting +I believe the content is in the public domain as it comes from the USPTO patent databaseThanks for fixing thisOn an unrelated topic Ive added the exact changes required to fix +Perhaps something on the web side is conflicting with the transactions? I also turned on Spring debugging in my log files to see what is going on during the call of my method from within the web app and there is no reference to the or with Spring +Ok so the actual use case is the union of what I was talking about and what Joshs patch doesLike I said before I definitely like the idea of both of these things but I do think that we need to have a way of saying that we want to overwrite something that has been written in checkpoint mode previously +Thanks your suggestion worked fine +Or is there some workaround for it? +You will need or some other mocking framework to do it +Lots of people will ask for it its something we need to have +But better to upgrade to the latest nightly snapshot +Im marking this cannot reproduce because we dont actually ship an rdoc binary +NOTE Must be fixed according to new rules +Created Karaf release page for websitesvn ciAdding downloadTransmitting file dataCommitted revision +Attaching patch that fixes the problem mentioned repetitive localization of users +In this case if customers dont accept to use driver with server then it should be considered a Blocker +To be fair its more change for the sake of shutting up findbugs which is good if it reduces the effort we have to do to investigate future reports +patch works +Maybe there is something else +To test the quickstart the URL required ishttplocalhostmyprojectThe problem only occurs when the app is installed at the root context so in order to reproduce this IE problem you need the quickstart to install the app so that you reach it viahttplocalhost +We attempt a proper rename now and only fall back to copydelete otherwise. +The patchalso contains a simple regression test based on the original bugscript +I did not visit this thread after my last post until I got PMed today +Im attaching the thread dump to this JIRADespite this I think we should probably go ahead and commit this patch and file a new JIRA for this intermittent failure +I will resubmit the patch +How about the followings +BULK EDIT These issues are open to be picked up +Could we let styleFamily automaticlly be set based on it parent container or some situation +I agree with Adrian put would throw a +Release Note no longer throws a when run on the Windows platform +Thanks! +Ill commit soon +Iainjust out of curiosity what kind of requirement method are you using? Is it working well? Have you ever considered to use the MRP instead +Thanks a lot! and That shows you are so indispensable +The committed code takes into account both syntax so the latest code in trunk should work like with Leonardo +this is a dead project closing issues +I committed the hack to the db pool portlet in rev +We see a similar problem +In Firefox it looks ok +To do so would invalidate its purpose +and remves windows lien delimiters in the file also +Allows the extensionadd to be propagated to the server so no special handling on the HC is needed to stop that +Moved the valueFactory to Pools constructorUnit testssystem tests pass +Closed all tickets + +Closed upon release of. +bq +I think its safe to simply remove these methodsIll commit this soon if there are no objections +am seeing the same problem +Either way its fixed in and trunk. +The two ways fit different requirements and cases its best if we can find a common way that can be accepted by both embedded and server environments +Removed some trailing of changing the overall structure I decided to move the logic to print that context token into the new generate method which prevents breaking existing error messages which rely on that behavior of an error token finishing the message +Thanks. +In retrospect this should probably be Critical since it creates a bug that can go unnoticed in a production environment +supports these operations for bitmap indexes which makes adding it to KEYS not a great use of time +As above not sure we want to promote this. +In fact as you have noted the site is all well right +JieshanCan you prepare a patch for trunk +How about or so +The test failures were caused by Too many open files +Done +I think after Yips changes for the above information is incorrect +Logged In YES useridFixed for +I guess Id still like this to be created +Added setter for context +Please open a new issue with a runnable test case if you are still having problems with thisThanks opened a new issue with a runnable test case. +Also can you paste a diff of the patch youre using +Sorry old patch is missing default charset and lang in this new patch +I will use the +Hacking out iln isnt a good solution +Please set IZPACKHOME +trunk x Thanks Yogi +attached patch per comment to your patch +Dumping DDL from one metastore and replaying it on another seems like a hacky way to synchronizetransmit metadata +Verify on CR Laura I have found several other references to the +Yay fix +are correct and working +. +So the only bug was right? I mean it remains true that fetching metadata is slow because of lack of zk multiget but there is nothing else to do on this right +Overall this looks good to me +The unit tests cover it +patch for on labels goes only on a new patch which should be working +The patch still does not apply need to rebase again +I have hadoop directory with backup before upgrading to TRUNK so i need about half an hour to unpack it and test with patched TRUNK +Sebb great just tested it and all examples are now workingRead about the +Thanks Kang +Have changed the db connection url to allow testing +And a trunk patch +We still use it in FST but mostly because the generics are ugly otherwise +Youd need to explicitly set to false in XML or programmatic configurationIts probably best left as is after making a note wrt to this defaultAlso all these Type instance properties in Configuration and are initialized to new instances +They could even become inner classes of each individual class that uses this fault injection technique +It will throw an exception because there is no Web service deployed but it will nevertheless start the +A patch for this +Reopen the issue due to previous patch broke VPCWould deal with it after Bharats bash script changes for dnsmasq configurator +Patch additional ve forgot to mention that the patch includes the boolean variable with a false as default value +solved +This completes the new plugin for branch +Yeah I think so as well +Please review +Thanks for the quick turnaround. +As things stand now a downtime and cleanup is needed to keep the running well this is less then optimal +Will need more investigation +As per Ralfs comments you should be able to see that this issue cannot be reproduced anymore +Is this correct? I thought maybe USER and CURRENTUSER were supposed to return the name of the user that owns the trigger similar to how they work in procedures created with EXTERNAL SECURITY DEFINERIm asking because the assert failure seems to indicate theres a missing SQL session context and I think USER and CURRENTUSER fetch the user name from the SQL session context so the two observations may be related. +Operations like to minimize the number of ports +This is because of security reasons that CS changes the error message coming from to the generic error message +Thanks Prajakta! +Approved for inclusion in by Chuck +Reopened and attached export of am getting ready for a release this week but if I have time I will try and debug it myself +I thought that when I set Declared SRS to EPSG it would be overridden by this value but the result is aboveMy colleague read your comment and just found that EPSG in the db is +We should also check other areas where we use smooks transformation to see if those also suffer from this +The check aborts the build if the IBM Java compiler is being used and is set +Its not a patch but a workaround that seems to work for us. +EugeneMay be this is some confusion of terminologies +If there is a problem please open a new jira and link this one with that. +Ok I see +Should we extend distcp to use this? Yes the canonical application is distcp +Thanks +Issue is for a long time without any objection so closing it. +Andrew please check that this fully resolves the issue +During the compilation I can watch the resources being written to the targetclasses folder and some seconds later they are gone again + and are both closed with cyclic reference +Fixed r +Maven project showing the problem +I verbosely explained the reasons to avoid this in the linked forum might be interesting to revisit this when the update document functionality becomes available in Lucene +Oops +This should be fixed in the latest CVS code. +Tested with the latest code updates from and the updated build scripts and didnt see the issue +The method getFonts in is sometimes called without a fontcache I didnt check that and finally it throws a NPEIve provided a patch so that this NPE is gone with version +I suspect it was introduced in the patch for +Fixed +Grab the latest activemqconcurrent and give that a try +Thats pretty weird behaviourI guess that invalidates my requestThanks for your time +easy fix +it does not seem that this has been implemented +Im not sure we can achieve that with the current way we parse things +three I installed DB in the work environment it had quite a lot of troubles with NIS +my bad +There is nothing wrong with having a cron job +Edited to remove a suggestion which has some problems on closer examination. +the queue servicing thread blocks while a commit transaction is in flight and when the transaction completes it clears the flagDoes that mean that it is still the case that while a write of client A is in progress no read of client B can be processed just because A and B are connected to the same followerThankss correct +Just hit p and accept the certificate +Just my +committed. +Fixed in brew builds and +The proposed solution gives more problems than the ones it wants to resolve +Can someone who them add the maven coordinates of those? +Can you also add one or few tests for sort merge join? Attached file has tests for sort merge join + +for +His plan is likely to Classes in a package by scanning the disk and jars + +Expect v later today +The has been moved to sandbox +A user with write could issue any kind of write operation to issue a put a user must have the put permission and to issue a delete a user must have the delete permission +This actually fails in web code which is at the same time doing all sorts of weird stuff wrt war assigned to Remy +JoelThanks for the quick review +Committed Thanks Zheng +Sorry for the noise +Juct committed a fix for this +Please remove messageboxresources directory and commit +Ok I see +This version needs the to be included into logphp appenders +started testing +bq +We should leave it as before +Here is the patch for thisNote Just experiment +Ashutosh it seems the review request got a wrong HIVE JIRA reference +for changing +Is there a connectivity change +closing since is now released +fixed. +Please try againRegardsBrianthanks for you patience +Good catch +Were dont know why the thread would hang +Also regarding above +Is there a reason to include it also in this release notes? has been fixed in CP +if there is still an issue please reopen +closed with the release of +If you do secure it will use https I manually tested that it works correctly +It looks good after solving and Im sure there wont be crashes anymore as dont rely on the JDK anymore when rendering TTF fonts. +But I think its better having a build script that checks that the plugins are actually installed +so we need to stretch things a bit. +best thing might be to use wicket and wicket from the behavior instead of wicketHide and wichetShow which will allow users to override it however they want +This is the last blocker I would like tor try and get it inA question AFAICT personsupportedFields is not a regex and its always going to be at the end of the URLIs there are reason that a Pattern is used rather than just saying endsWithIt may sound minor but I believe that a regex pattern match is generally slower than very simple String comparisonOther than that I am happy to apply the patch +If I am setting a value to a collection then I can see how a defensive copy may not be done +Ive removed my latest patch +Incredible that its been open for years! +Im guessing its because Id upgraded before to +Because Tomcat puts threads into a pool the never go away +However I am unable to figure out the failing tests + +It looks good other than Aruns nit about the getLength javadoc being for the wrong method +no didnt try that one will do +It simply looks for the first entirely numeric id above a found clientId and replaces it with the given index +Do we have a fix around this? I downloaded Guvnor and it gives me the same error with that is where I got the from +I would appreciate if is closed as a dup of this one +Patch applied to LUNI module at repo revision rPlease check it was applied as expected +The goal of its subtasks and this jira is to make the minimal changes necessary for clusters to use tokens with simple auth +There are some threads that I ignore in Solr because there is no way to shut them down cleanly +This could be related to +False alarm +Bulk closing stale resolved issues +I just tried it this morning +Im nearly certain this problem is related to the xinclude support that was in previous version of UIMA +Orchestra now does checks for unaccessed conversations only after rendering a new view +The pb occurs only on derby + +Let me see tomorrow if that is possible +Im not particularly wealthy but yeah +Fixed by checking for more specific objects in the Wicket namespace in and quickstart works for meDo you experience the problem with the same quickstart or with your application +Flushing is the technique employed by the old session scheduler service and the PVM enterprise timer session and works well in both cases +is genericized as wellLooks like were going to lose some flexibility in protobuf defined coprocessor service version +Fixed both the issues expressed in this bug AS console is now left untouched and AS console is installed correctly. +Hey Tomas this has an overlap with and it looks like this fix covers the fix can you confirm so we can close them both when this goes in +Please use the navigation to get back there + +can you provide a test project please +Oh one thing I forgot to add in the limitations aboveColumn criteria can only apply to columns included int he results +Sorry got my wires crossed on the different jiras going around +Mina version Thanks Sangjin! +I am about to update it +I checked Suns tinderbox runs and the last runs have run fine with no to add backport reject label as this only affects trunk +Why hold up other developers? The file should be correct for the HEAD of trunk +for the patch +I think those can be on different domains +Youre right +Got it +The trigger for external users is the easiest and should be done first +Total nitpick but table should probably be deprecated in new interfaces +Just worried that it will lead to confusion +Thanks Rob +To me the patch moved the second part of the condition inside the block so essentially it is the same right +Only one region server will get the lock and handle the failover +I like the changes to JDO making jdoConf an making theconstructors private and adding the getInstance methods +Cant we just have them begin with test? The output from the tests when run through ant test is in the wrong order +Patch is attached +Requiring the keytoken conversion on the client side for paging is ugly +Here is the test case for this issue +So I thought its not supported + +I see no need to add such an annotation since it is used in a custom which can be provided in s constructor +The YAML issue is fixed. +If not please. +Thats right +How costly is delete and exists check? I will check on this more +Awesome +Why do you choose to notifyAll when requested is set to true but not when scan is set to null or requested is set to falseIgnore this comment +Is there still an issue with that updated patch vI want to implement a variant in the future but currently this is all I have for my embedded device +If you replace both inside the the default flow should be saved. +may simply not be using the Ruby subclass object +Hey Hive committers can one of you please take a look at this patch? Wed love to let users start using sets + +Unless you did that on purpose I think runtime is better than test for this case +You are right I dont catch it yesterday +Patch for Nutchgora is much more comprehensive and a far cleaner implementation +Thanks for the clarification Reginald +Thanks BilginI will do +This also applies to changes similar change inChange timeout value +Awesome +java directlyis there another way +Patches for and trunk to print configuration and fix typos +Especially if all of the other jobs on this node are done localizing so their events get processed quickly and removed from the queue +Id disagree that using the tag makes apps more likely than not to have an injection vulnerability though +Thanks +You have to either turn this feature off or use value objects that wont suffer from circular references lazy initialization +Hi I am under the impression this problem is still there for https connections +Added a test case +Apart from that we could fix it in and a bug fix release +bq +All is timestampBut Sometimes we might need a convert function as a timestamp to string or string to timestampBecause we are humanSo Make a function packageAnd id like to make a system table when system initializingThen user can insertselect own date format style for convertingOtherwise just hard codingAnyhow it also will be a good util for usersGrammars are simipleId like to some copy the common SQL syntax +seems like theres with work being done on replication +Move the fix to +As per previous comments problem looks like a duplicate +Again I cant thanks you +So it is not XHTML validAlso I trimmed the values in the comma separated list before parsing RI does not do this +What version were you using? There are issues fixed for Putnam on this on the latest build +Then the clients will translate the logical id to the physical id of the instance based on their configuration +last patch with a unit test +In the ADA case we need to use apply up to the segment point when the buffered deletes can be used +implement a pluggable which adjusts fetchInterval and next change and to use them +BTW I am having trouble running the tests on trunk so I wasnt able to verify this patch Ill work on getting my dev environment more functional +Please fix this bug it is annoying the heck out of my customers and me +And now I have run the complete test and Im sure the new patch doesnt make bad effect on old testingBesides Im not sure whether its neccesary to run the complete for this patch as its just to have added a new test class not changed any product code maybe its safe even not run the complete test suite +httpshudsonviewTrunkjob +Closing. + doesnt work with and its refactored distribution to bundle runtime required libs in assembly and Seam distribution just use that and unpack it +I think something is broken about it currently though since Ive heard many people report the same behavior + should be moved from impl to util or a support should not implement DONE +To me it looks like this issue is resolveMarcel is there anything else you wanted to change here +Done at rev +Hell do that for the replies so far +PM decision is we stay with alldefaultproduction get the new jarsdefault and web do notAs for signing This revision was not passed to RE for signing this will be done for the CRs onwards. +Due to not being released in CR was released after the +While we dont have further informations nor needs seems way enough to me +Once I deleted those extraneous +My of a bulk update all resolutions changed to done please review history to original resolution type +I am not at all a legal expert but if the possible TM issue is limited to regexp patterns I dont think there could be anything to claim about + for the to trunk. +This patch has been done on the trunk +Hey RishiI looked at the proposed patch and it looks like a minor change +The only issue that really needs to be settled before the release is whether or not we change the way the Jira URL is specified if that isnt addressed there will little chance to change it in the future +Thanks for the review Robert! Ive made all of the suggested changes +r for update +Slick. +Tomorrow you will get a lot of patchesCheers +Please verify +Event handlers yet to be covered with Selenium tests onsubmit onbeforedomupdate onerror of a bulk update all resolutions changed to done please review history to original resolution type +looks sane for at least +Ran the rowcounter MR job +second patch which also handles conversations nowCould anyone please review it +The most appropriate fix seems to be to move the procedures into something like and to retarget the designer role editor checkbox to instead say something like is admin +Any recommendations on how to provide unit tests for changes to +Does it really require the base class to be abstract IOW when the view is being used do other super classes work for injection +Ah OK +all issues not touched in a month to i still see this +I dont see why we need a special case for this +The same is in so at least it is nothing new +Should that be a separate patch by itself? I think it might make sense to separate bugs with and schema change bugs +please try r should work as expectedsteps sync ant +All looks well thanks! +Test case Sven! +Ive committed your most recent patch to SVN +I see it now +Changed all projects to either remove manifest entry that caused nested jars or changed to unpack the project when libs were involved +You can find fixed problems such as FK names Primary Key with Index Unique Generation and so on +Ive also contacted team but they are not sure they want to do junit support +I will looking into thisThank you Craig for all your hard work towards a Pluto +Thanks for the report +I just dont want that mixed into this issue +Hello Claus I have done this several times but still the formatter is not there +This seems unrelated to the patch I am trying to merge +Bulk closing for +All set +Committed revision +Went through the tutorial steps using GS trunk and it worked without a hitch +Heres an updated patch +On my end Ill see if I observe it as well and will report back +If you still have the problem we may reopen the issue +escape things or how backslashes and colons should be treated etc +Have you prototyped your custom CL similarly to the one listed at Peter Ks blogI wonder if CXF should ship such a loader and try when the initial proxy creation has failed what do you think +Elliots first patch used which allowed passing initialization parameters to ctorMaybe revive that approach +I see with vnodes we have enough ranges that we can have a thundering herd even if each range only clones oncev attached with the approach you described originally +Yash Talwar +Thanks +GMT will be stored as because in GMT time it is GMT A date that is stored as +Jane could you also document that using the context menu Maven Update Project is only available on the project folder and not on the file? I know its unrelated to this issue but would be nice if you also put that in place +Good points Joy +Currently we are building an advanced search capability that dynamically builds a query for Hibernate +When using nightly builds you have to wait until the next build after commit which was in this case tonight GMTYou dont need to configure Ive added a renderer class yesterday eveningThe attribute are not full implemented yet +thanks to Paul Rivera for provide this patch +Anyone want to take a crack at reviewing this? If not Ill get to it tomorrow night +I need someone to review these changes +It seems pretty awesomeindeed this looks good +Upgrade to compatibility to latest snapshottrunk +My thoughts on things +Sorry again +Added improved RDFXML type metadata and test cases to verify that the type is correctly detected +Integrated to +Hi +I really think that not supporting headers is a major shortcoming of the framework but regardless could we at least remove the final from the relevant methods as a stopgap measure +I ran through the manual test from which returns all contacts from the device and prints them out to a DIV +Removed enum in addendum +you need to override the clientExclude by default and are excluded +ctor param what to do with SICA? Im passing a null in as constructors not so pertinent to SDI components +Im confused about what version trunk is at leaving it open in case we need to apply to a branch as well +You removed the cd HADOOPHOME line in +Indeed this is a when using Maven +This patches adds the Xalan NOTICE to the end of the Derby NOTICE file to cover the Derby builds use of the Xalan jar files +Can someone provide us information on this? The claim is that is was fixed in Axis and it was not +Compromised on the javadocs just have the reminder that it is the clients responsibilityto close the datasource when no longer being as of r Marking this resolved will close when uDig has succeed in making the switch +requires classes +Doesnt easymock extend? Then the scope is perfect with compile +The testing files must be compiled only when you want to test Jitrinos internal algorihmsSo this is my TODO +I will close it unless someone objects +s o a +Its not complete +So if you dont see the problem anyway its because Rob fixed it no +So when a new checksum is generated when saving the image to disk the new value is set in the spyImage but when saving the signature into VERSION file using it uses the value set in orignalImage +Resolved by r on the trunk and r on the branch +Done +I will commit this tonight +I solved the problem thusTyped yum install antwhich updated by ant from to version which unlike the version I had installed includes an fileHowever the forrest docs say that it includes its own ant version so it shouldnt depend on having ant installed already +waittime is now fixed to +bigJava vm version +Whether this is a good idea or not is of course up to the AOP team +Were going to compile it into our version because of that +I was just informed in that current implementation is in experimental phase +commit m use name from login scren rather than always using whatever cookie matches the site Committed revision +Removing target release from issues that fit any of the following criteria No known way to fix them Java integration enhancements out of scope for release Other out of scope issues for in r +Someone wants to add this to the doc? Maybe for the cli the threading bugs have been fixed as part of other issues the only bit left in this issue might be about termination when reading from the console. +See for a related problem +Which distro version of mozilla do you have installed +Attached test output +we plan also a feature compareable to static imports but for methods only and the first parameter tells us where the method to add to +Thanks for the quick review MatteoWill integrate later today if there is no further review comment from others +I meant to say the test is not running at all with jME. +There are roughly two important call sites for these methods one is in the and one is in the during responses to leader requests +More tasks may be added as details get dissected. +Simon yes see +Thus moved this and other issues from target release to +Testing patch now +However we dont work on it at full capacity +Its only the builder who knows how to build this message +Probably we should check with Richards about the environment and setup again +Can i commit this +Can you please upmerge to latest +As for the default I need to hear from Dan +This is a complex problem to fix +Committed to trunk. +lowering severity from blocker to critical +Build script log showing errors +Thanks DagCommitted patch a to trunk with revision +We could probably download the plexus compilers we use instead of having one hardcoded in the Not that downloading the implementation isnt useful + +Daniel I applied the patch and I see that history and line editing is working great +Patch applies over patch for +Does address the requirementsFor job limits Yes +Okey so im not seeing this as an eclipse or tools specific issue its just how the runtime picks up the deployed resources am I rightIf it is a bug for tools let me know +Closing all resolved tickets from or older +HiI am basing the converter tool on to parse the tld and write the config file +I will commit it in a few days unless there are any concerns +Thanks Uwe +will provide a better solution and the ability to put in a accepter in etc should be a long time ago but not allowing or to have an Acceptor or Forager instance any more. +improvement Doug suggested +were thinking now of allowing code into benchmark. +All EAP Installation Guide comments applied +The work around is to register an MVEL data convert for Pateint and Strings. +Lets see what Eric would say +Well set context classloader for all threads by default and see how it goes +Verified that ligature in author is fixed +Targeting for the release which will be pushed out the door tomorrow or on Monday +Okay here are updated patches with all your comments appliedConcerning the I dont think they are needed any longer if all dynamic schemas are properly disposed of +After installation there is only file but both launchers eclipse and jbdevstudio +We only expect them to be functional Please let me know if this can be included in tried it +Andrea this one still bites me on a regular basis +Commons Codec was released a couple of weeks ago + +thanks Jochen I will verify that it works against Grails branch +Having interface support with annotations would definitely be better but does work as a stopgap +PS I attached a file and just added a comment instead of saying submit a patch which changes the status +Just adjusting the formatting +well if you use maven out of the box with no mirrors in your it should have downloaded out of the box the right hibernate pom from the central repositorydo you have any special config in your? +the big problem with doing this is the which relies on the fact that the index is loaded lazilySo maybe we need to wait until to do thisBut perhaps for this issue we should make it possible to pass in the term to and deprecate the current methods and then in we could switch to loading +Thanks Stefan +With this applied a remote system error is thrown locally as an rather than an +My assumption is irrelevant in the endIm simply asking the defect owner to help identify which one might be working hereIf the thread context classloader is adhering to the parent first search behavior then this would only be a problem if the thread context classloader did not have the current classs classloader in its hierarchyIf the thread context classloader is adhering to the child first search behavior then it opens up all sorts of ugly doors in a test environment so Im guessing this isnt true. +code change is committed have to mention it on reference docs +Fang can you check on this issue +Perhaps something about the way the handles the file is damaging the encoding of the file on the way in +build and build test pass on both winxp and RHEL w gcc +Test failure is not relatedPatch integrated to trunkThanks Konstantin. +I also fixed some typos and java warnings +defer all X issues to +Ive been trying to create a single record application which exhibits this failure and have been unable to so now Im having to go and try to create a minimal application which exhibits the same existing application can not be posted as is but heres what the initial model relations look likeA B DA C D ABCD represent active record model also exists a person lookup which is invoked from the Layout to put the users name at the top of the screen which is also an active record model class +Thanks Michal for explanation +So Ill finish up the async api work in another issue and then circle back to this. +fixed +The problem has been problem has been in +Also this is unrelated but the Locking class doesnt have the annotation present to export the protocol via JMX +I have the blank image problem still +Please remember to include in jstl and everything should be ok. +Thanks eliminating the enclosing parentheses solves the problem +Sounds good +Except these methods at least for our use case the rest worked like a charm +The class comment says Converts a boolean primitive or wrapper to a String +The attached patch introduces a LINESEPARATOR constant and uses it throughout so that all unit tests pass on alternative solution might be to have a utility that compares strings ignoring line ending differences but this would cause the test to fail to detect line ending corruption caused by the filter itself +there are only zero length files we have byteCount +Attached trivial in parent JIRA +Suggested implementation +Patch byClaus PolankaMichael guys +Committed revision r +I hooked that up with Random testing and all tests the given patch Id commit that to trunk and backport the structure without the enhancements to +The patches with should be applied prior to the Nov series of patches here. +Ill commit a trivial patch to and trunk after tests pass locally +I could go through the assembly if necessary +Ick +Create a new folder under managementsrcmainresources and store the rd party plugin file +As a result when rewire on split is called the call is dispatched to the split output +So you are ok with coprocessors having a different configuration than? Dont you forsee issues with that +Domain classes used in the +If so we should probably modify the javadoc +fix candidate for Anyone vote for including this name change fix in the releaseCustom update chains are very little in use out there so its easier to change the name of the parameter now than later +Please check that it was applied as expected +However in this case the List must be nullified in the onDestroy callback +I apply the patch but it does not fix the tried it again and the test did pass! I might have done something wrong before +All in all its the client responsability to manage correctly Aliases no +However this is has been addressed for the time being +I feel this is not something to be fixed at the Storage class +Verified on JBDS. +If we then immediately ship them to the destination worker and combine we would end up with final message to be stored for the next superstep +Done +does some things + attached patch +Patch is attached to here also +Need to extend the JIRA to cover as signalexit info to details info displayed by WS for andCode delivered. +If you are lgoged in at jira and have selected to show this issue you should see a link Attach file on the left hand side of the page just above the Comment link you used before. +Instead of doing that I created a variable that just said whether or not the expected values matched up and then if it doesnt skips the save part of check and save +Dhruba or Jakob can you please comment +Uwe ill look and see how you do it forif it can make the code for this simpler that will be fantastic +class and id attributes for rows can now be set using a table decorator +Mark Miller do you want this on the release branchIll defer to those of you deeper into the issue but from a high level to me it does not seem worth the gain vs the risk for +Thanks guys +Can you create a suggested patch +When a more general solution is found that hack should be be great if this were a general solution that could be used for any annotation not just in EJB. +See it is syntaxically and semantically valid but it is a bad practiceTherefore there should be a possiblity to detect this pattern usage in case. +Its a of +this code seems to check only whether there is a soapMessageThe behavior of the Dutch government compliancy test can not be explained by this code which would imply there is no difference between the namespace declaration in the header or the envelope element +The change to now tries to find the project root directory instead of assuming the current directory is the project root +The recipient is left to deal with it in this format +The integration code hasnt changed for a while +Forgot to mention Cocoon does not have compiled sitemap anymore and usesEclipses compiler for so this should not be an issue for Cocoon +This issue has been fixed and released as part of release +The find command passes null to the method to print out the bundles which causes it to print for all start levels +is gone +My workaround is committed at +The test actually fails deterministically if you just run it in isolation since its missing some initialization +By default tcp is enabledthe client should use a URL like tcplocalhostfalse to disable tcp on both the client and the server socketIf you guys get a chance could you confirm that this fixes the issue? thanks +fix has been made in trunk and backported to and branches +Simple WAR file generation and dropping is less maintenance burden for us and sufficient for running the samplesJuergen +In the automated JUNIT tests will be invoked to verify if the result has been received or notIm not sure if this is the right way as I not sure if invoking and will hit the same instance +But the FAQs have since changed order and this doesnt make sense now +reasonable defaults seem like the way to go +If compliance is the main goal that focus should be increasing the variety of compliance rules on the source tree rather than including Test sources +bq +regardsNandana +Tested +add to resolve all parameters waiting by a service and complete xsd +Add proposed patch to add svn ignore property +Moved to since its not critical for the release +eos first using the there everything will be logged to tmp the logfile is rotated iff MB not committed added chmod in place on eos with chmodDugorxFugoruwt not committed +If its ok Im holding off reviewing this until after we have out next week +Put a branch release as requested by because the task is not totaly solved yet +Put the common code in the super classWith the copied code its best for now to mark the copied segments and change as little as possible inside of them so that when it comes time to refactor then we will have the easiest time doing it +Patrick I am interested in addressing this problem +One other thing that puzzles me is that your Forking command line doesnt look the same as mine +I am finding I am having to write action to render view fragments and that is all they do +Attached test case which reproduces the issue on an HSQLDB +Also updated jdbc tests to verify +The keystore should go in componentssrctestresourceorgapachecamelcomponentcryptoColm +There is no support for the in the trafficserver script to startstop the TS process stack yetGeorge +I forgot that I added a feature to fix this problem some time ago! Im adding the necessary new option to the and Im moving this test back to the +Please reopen if this not the case +Hopefully this can sneak into +Even if you dont have unit tests I would really love to take a look at your code because we want to do credential delegation too +Its really outdated +OK I found the issue in certain cases the method too should return true so that the formatPrecise method is used instead of the formatFast method +Does someone know what is the best way of doing that? Should it be handled like a resource +Hi Lyle yes this is the same error + +on commit +Let me fix it +Thanks Vinod +Thanks Heiko! +is for branch +Yes we already have update +acked +This has annoyed me too for a long time +Discovered outputsare pushed into a queue from which they are pulled by the copiers +Name clarification +The patch doesnt fix the second problem foo +Patch attached for the patch into master and with thanks to Dhiraj. +Once an executor is launched executors resource limits should definitely be controlled by cgroups! We probably need a better abstraction for the launcherLet me know if you need any help from me. +C is my native so Im crossing over a bit into JavaI will of course your suggestionsTBC Colm +Calls the method in the places after move methods are called +Overall I dont think its a big deal +Just my cents + the patch looks good to me +It seems this does actually show where the error occurs +We can be more specific to say something like is not supported +updated xml schemas documentation and committed to sitetrunk at rev used xml schemas from branches at rev +screenshot of the error be in next code drop +If you can supply a patch file that would be great +Please try them out +patch looks good +Applied the patch as is +Please select a platformRunning Fedora JBDS universal installer Hybrid Mobile Tools released to master +Thank you to Ruslan for his comments and Matthew for the +Heres the implementation +Would still like HQ to at least verify that the change I made to this query to fix this bug doesnt introduce other problems +Path variables are no longer added to the model for +In addition it should be possible to specify a default binding file as part of the binding file +Had a offline review with Yongqiang looks goodwill commit if the tests +I thought exactly the same when I was working with the +Use UUID as the hive session id Clean up the download dir in when CLI is invoked with f option Handle minor error just committed this +Please look in your system logs to see if there are any reports of problems accessing the drive that stores this index +Thus standard Java rules apply +I tried the same test with the branch and I dont see the problem +Ill the integration +Documentation is part of svn as well. +Hi I am facing this same issue +I will attach my patch and a testcase +Kinda hairy stuff in there so give me a few days to scratch my head and review this but something worthwhile getting fixed finally +It was triggering in one of the unittests and is not now +This should be okay though when running in parallel the results may not always be sorted +Do you still experience the problem after upgrading to Apache James Beta? I am trying to see whether this issue has been truly fixed in Apache James Beta +Thanks Tim +Which strategy do you use to eliminate overlapping names of samedifferent typeWe should have a test class before we commit it +The unanswered question is whether this patch does anything useful for the project +Thanks Michael! +It did not seem to have conflicts in any class it affects +This bug has been marked as a duplicate of +Updated now +Committed Sheryls unit tests in r +And I dont know why JIRA put lines of code into confluence code tags to off release +Is this present in other versions also +Not sure we should change it +Therefore it must be treated more carefully +Thus this might be rather JBAS issue +You have to be careful with cookies because the web app proxy strips out cookies before sending the data to the application +attached my example which works. +Setting the local repository in +Andreascould you please attach page snippet and screen shot +Fixed and tested on IE FF and the problem on the email settings page and in turn also on the filter s OK now. +Please feel free to chime in SimonIm proposing that we move Kuromoji to core to make Japanese supported with a useful field type and corresponding documentation on the wiki +This has invalidated +I was using a derivative of te class which was also passed to a that deletes all the files handled by the step in its afterStep was not able to delete the last file since it was still probably the issue here is that the afterStep is called before the is am not sure if this is a bug or WAD +That can go in separately +Thanks Pavol +Re attaching patch for jenkins mainresources still needs an entry for otherwise the patch looks good +I tested the change in our Oracle environment +The upgrade to the new blueprint core should include the fix for this +Otherwise youre in for a world of unpredictability poor performance and maintenance nightmares +Thanks Grant! +This would also reduce the need for keeping both tightly linked because they can both use the same widget and much of the real code is shared. +Configurable +Is it OK +Im sorry about the duplicata I had done a serious reserch when filing my first bug but I didnt this time +these assorted bugs have been fixed via other patches +removed dependencies on Java on +The patch is currently setting envp to null +I Didnt check all the eclipse plugins but about file names works properly think we can close the bug. +We should that add into the user guide +Fixed at revision Thanks Ivan Vasiliev for the patch +If the tasks only need scratch space they can write in +Thanks Gunther! +The patch looks good with the exception of some minor formatting issues which I noted on reviewboard +Patch cutting over test to checkFields making it also check deleted docsThis found a bug! In Lucene codecs term vectors seekCeil +Bulk update to fix version +I think this should only be handled cleanly if the operation is invoked via some helper API in the helpers packages +Script fixed and a manual run triggered. +Did a home page and fixed up nav menus +Perhaps a hiccup due to the recent jira updatesAnyway seems to have resolved now +quickstart tested it on wicket versions from to all those versions have this bug +Youre right +A quick fix +Yes +My last patch had not the change in we can compile this class optionnallywith an exclude unless in the javac task +We should add the delete option as well +If code relies on results in document order the needs to be adapted when upgrading toCommitted in revision think we should drop the QUERYXPATHDOCORDER descriptor +Fair enough thanks again for the explanation. +Hellorsyncd config has been adjusted and the apache rsync module should be available now +Thanks for testing! +Lukas please review and assess before we release Rene could you please open new issue as bug for M? +I see it happening in as well +Patch removes the not needed imports and adds fully specified paths to the Put and Delete links +Im sure most of it would pass the only thing that might cause trouble is encoding of double and float +unscheduled issues to the next release +Thanks a lot Nicholas for your reviewsI will commit it in some time today +Ill go ahead and commit it to and as soon as its updatedthanks +Our customers are in great need for this feature and so far none of them has expressed the need to have different prefixes for different tables +Ill keep trying to reproduce it on my end too +That happened primarily because of the test cases were mock container +So it seems to be possible to set the property values to an empty array for an existing property state +If this doesnt work for you please the issue with a description of the problem +Will focus on this again past the old tickets. +I guess its OK to put it in trunk +The AS discussion was really for AS +OK posting queries to select now works for me and updates are handled by a separate servlet. +If the limitation is not lifted for the user documentation should be updated too +Its better we tidy up this Utils stuff to achieve better reusable components +set the default value of max linear search to +Shall I file a separate issue for this? please +When you say it diverges Im assuming your concern is more at the code level than the functionalExactly code level +! forgot to change the prioritySorry +Considering this fixed then +As of now it says use +lgtm +We have a project that cannot be built by continuum because of this issue +So now we must be sure with your actual is why I asked you to see what gets as parameters in your case +This demands a fix through JON logged to track it +We carefully examined the test import and it looks fine so far +AlexPlease check if this affects ajax mode. +Committed to trunk and x +Not yet will try to find some time to see how it can be adjusted for trunk +This changes the target which copies the docs making it also copy the css filesCommited to trunk at subversion revision Ported to branch at subversion revision Ported to branch at subversion revision Touches the following filesM toolsrelease +Then we dont handle the thrown exception +When we see all nodes in the election are recovering wait N seconds for more nodes to join the election it may be that an active node just needs another second or two to reconnect +Thanks for the patch Emmanuel I wish I had seen it earlier as I ended up duplicating some of your workThere were some additional issues presented by along with a deprecated Tiff parsing method some tests not passing and Ive refactored the but all seems to be working nowThis should be resolved in r +Add a +See you can run the tests +I added a which I reference from the about how to fully recreate the parser +jeichar is the project source of a plugin that does the same thing that my app does when I reported this bug +Whoops I was totally reading the reports wrong earlierBut Clare summed it up nicely in her questionIs the consensus that they were deleted but the wrong revision was winning because of a bug in the tree making it look not deletedTo which the answer is yes +Part of a bulk update all resolutions changed to done please review history to original resolution type +Great +The Task should also use setConfiguration with the jobs configuration +So maybe we should choose the request size as the variable instead +Would anyone want to try this +Ord with no transition cost on the largest segment should be a nice win on its own +First the synchdir does not need to be set in distexample +run +looks good to me +Contribution to Spring +bq +If use true even when I generate everything the module depency shows up as a library dependency +A similiar patch was also applied to andtests were added for both. +I suppose the overlay would not work in that case would it +bq +Thanks Todd! +Ah you should have seen the original result of createSchema doubles were mapped to an intvector type +Bad that my Ruby is so lame and good because Ruby is such a beautiful languageRegardsYanko +Ive yet to be able to recreate this issue +is default +Failed isnt enough exceptions are handy +The database should be created by before running the application +Hi Stefan thank you you are true I am using +layout attribute should be added with values horizontal and vertical +still +Id have to dig it up but some user complained about a similar issue when was put inIf you add create as a public method then all of the Lucene classes have to be changed to call it otherwise if you override it in a user Query it wont be called on that QueryBut anyone with an external Query class that overrode createWeight will not call create and wont work correctly with classes that override it +That would greaty optimize Java VM +Its crucial to either filter these or sample down their interactions +Name Finder should display an error message if there are zero sentence or token annotations in the CAS. +Turned out it was a little more involved +I think I have the patch backlog cleared up now if youd like to submit the patch +If my understanding is correct the main difference is that they are blocking on adding to the call queue whereas what I did in is sending the request back to the client for retries +Thanks +Attaching a patch +Do you plan to use it laterYes since we have to know whether the transferred replica is a RBW +please go ahead +Please let me know where I can help +May need to them for the assembly but will deal with that later +Looks good +Of course you can write +add entries for configuration +I cannot unpack the attached rar +If the Servlet Container doesnt clean up all of the threads before returning the to the thread pool we have the following bad cases content leaks until this thread happens to be reassigned to a request that uses the at which point it might still leak Application assumes that a has been initialized by this request and thus reads stale data The supports hot patching of application content +Check the HTML spec +dhruba borthakur Thanks for your commentCan I ask whether this feature has been completed? I suppose the only thing left are the test plan negative unit tests and performance test rightWhat patch above Shall I used to test it and by the way which version you think this patch shall be merged into +The one appealing feature of annotation is that it keeps information about table layout near the code which reduces the cognitive burden on the developer +All members are injected using Expression objects which should be evaluated on a basis +Id love to see a patch for this effect +Applied patch into trunk with thanks to Cdric. +Committed to and trunk +Clearing fix version +Well see which gives better performance and that too if I everfinish itPRegardsApurv VermaIn either case we are going to use and I propose we have a reusable byte array service that provides you a wrapping a range from that byte array +Remaking will update ASAP +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Guess we will have to try and see +Actually there is one required change before everything is go srcjavaorgapachecassandrautils needs a APL header +I think such method would be usefull +Chrome renders the problem area differently than FF +Yes that it is correct +Thanks very much to Tiago Espinha for this contribution + extends without needing exposure to all the private variables +Its because of the Find action and I have not worked on this yet +So this may make the case for block versioning +well +Here is a patch for restoring stateless property of with Deprecated annotation and deprecated Javadoc tag +TamasI created this patch from the files you supplied and I also cleaned up the code a little +I figured out what I did wrong +Probably end users should be made aware of that since their apps may depend on PE +I am closing JIRA. +Hi ShazronYou can use the code sample from +Closing the issue now that it has been resolved. +What I end up with is all the projects in the archive imported into my workspace +I just factored the test runs out to the nightly jobs +Granted the problem is annoying but the itch needs to be scratched correctly +Set up two queues with both common admins and different sets +This patch will address your comment +Ok it was fixed +This patch still use get on Linux because I havent found a better way to resolve symbol links on Linux which did by get. + +Im not sure what the advantages and disadvantages WRT this but Iknow its definitely a good practice +As far as I understand it there were specific changes made to that improved Tomcats ability to reload deployed applications +Sure Id be happy to +This issue will need a patch after is submmitted +Merged to the branch in revision . +Sometimes the version history is empty on the first print sometimes not +The user would have to click on the to see the child they just added +Hi Balaji a quick suggestion +This application did deploy in AS with of heap versions of Weld should concentrate on not reflecting the same classes repeatedly as is currently done +Thanks Adrian for looking into this and providing the solution as well +Thanks for the post! I lost the email you sent me between work and home +Thanks for such a rapid fix +This behaviour can be observed when the release plugin is configured by parent pom +was already fixed by +Please see linked SOA Platform on your comment on +Patch includes Thawans suggestion and C client changes that were missing from previous couple of patches by mistake +Fixed in svn revision Thanks a ton Amila +Overlapping dependencies are removed +If the duplication becomes a serious problem we can move it into the build. +So it works as designed +This is an extremely interesting feature for LDAP +Sounds good +No problem +Please go to login with degenaro and upload the ssh keys thereor come to asfinfra I am thereThe keys on minotaur is local to that vmrgdsjan Ips +Id love to see that restored +CheersClinton +The problem of XSI type not beinggenerated for soapjms for some tag is a bug +Pete Muir May register the for timeout methods +doesnt run on patches +please attach a runnable test case fixed in CVS please test. +Thanks for pointing it out +You are hitting the JVM limit on the size of an array +Thanks AkashYour patch is in trunk rev +I cannot reproduce it since I fixed +Do you think if we add quotes around each arguments will all the scripts and applications tolerate it? How about having interpretation for delimiters +If you can provide a testcase that doesnt involve J please this issue. +Seam Servlet is now ready +After the fix I was unable to force a deadlock +Whats the status of this issue Rick? Do we have any concrete plans to change the current approach? Else lets simply close this issue for the time being and stick with what we have +Committed to trunk +Now we are saying that if there is a conflict it should still not fail +Riftsaw is just fine +also remove extra tabs in +Is this still an issue or can it be closed +What would you suggest instead of using the class name attribute as an ID +The highest value wins strategy strikes me as essentially undermining the case for Cassandras cell level update resolution strategy +Marking closed. +Improved xml formatting in issue you very much it worked! Im sorry I couldnt figure it out myself Im new to all this stuff +This was fix on trunk a little while ago +Wow that was quick! Thanks a lotI reran my tests and everything works smoothly also ran the latest snapshot against some other java libraries that I had installed on my system +eclipse maven arquillian project to reproduce the Versions AS and +Heres a unit test with the fix +If the source cannot be reset it should throw an Exception + to force to rerun +Its a good start point for extendingrewriting it. +Rebase for trunk +I have updated that wiki page so it links to the user guide +Backported to vGACP +can it just not +Suggested path +A proposed patch +Why not just configure the socket buffer size based on the maximum IO size we want to block on? Say kWe had two issuesone was blocking on the full write which I thinkhope we fixed +Test cases are attached +Attached patch places a note after the first paragraph indicating the JDK requirement for examples +All the more reason to get them correct if were going to do them +any comments on this +I can work on the issue as well in parallelPls +the price of parsing JSON +Can you clarify +Commit set on BRANCH on BRANCH for release +It was cleaner to redo it from scratch +We also set up Hive on an eleven node cluster and performed the benchmark +This also includes reworking configurations and improved support for concurrent access +Fixed along with a test case. +fcecbdfbadbacf +The file descriptor problem seems to be causing another problem the JVM memory info isnt showing up on the dashboard +Given the dynamic nature of wiki content I assume its no longer relevant if it hadnt been fixed. +that when unpacked in a Hadoop tree put them in the right places +Also Indexer is modified to check for negatives in and make them +Not done yet + +Can we open new issue for future commentsThis JIRA has been resolved +I think this code is actually simpler than the original version and importantly more efficientThere is a clear iterator that hides all iterating logic insideIf there is a way to avoid storing references then why storeThe way Nicholas did this is by using standard Java library classes and methods +And youll also be able to get version +Following your review I will be happy to commit it to the camel source +trivial uncontentious change +I have committed a fix for itt + all issues like shown in the link ryantxu sounds swell +They are the actual starter kit which is needed by the start examples which helps launch using configuration files the Jini services +Here just exposing admin command to operate on +The problem to change the formating now is that most of the patches that are still open will fail to apply +Yes that works + committed yesterday by G +Moving this to the next release +Id like to commit it soon +bulk defer of issues to +This assumes that the URL encoding done by may not produce an identical POST data string +Bulk closing for +It seems likely that you could validate it on a single node though +Definitely a XML SchemaLibrariesEndorsed issue +I think perhaps like a Hadoop input format split we can define at the segment level as to where the documents live so that if one is splitting the index as is being implemented with the splitter can be smart +Looks good to me thank you +Hello RuthThe patch has been committed at revision r +I dont feel a schema registry is a good idea +a screenshot would probably explain it better +Thanks Paul +I may push a copy to github so that i can try to get Igor to help +Thanks for the detailed advices +Showcase is fixed. +trunk +HelloI dont understand why Nicolas Lalevee said its not possibleIve just finished a patch solving your issue +Would appreciate some help on writing one +It uses the version of ant packaged within sonar. +Yeah in the interest of time I think download a clean Eclipse and installing is the best choice +verified by Paulex +Not on Blockers list +This patch fixes Race Conditions in and some issues in +FYI other s already have this like solrtikaconf +This patch fixes that +Fixed for by using maxIdle in place of minIdle if minIdle maxIdleIll look into porting this to and next +Thanks for the patch +Most of the calls are implemented as queries against the catalogs using the user transaction +The logs are completely clean which surprised me given my recent results on and full +I can choose other files and reliably reproduce the issue +Fixed at revision +If you agree Ill apply it +this is not going to be changedRead the duplicate rejection chapter on the docs +seam directory +The Delayer in fact uses the Id attribute even inside of a chain +Added a second patch with documentation for aspect and adapter indices +Will finish the architecture side shortly +enables usage of Parascale storage +Yeah I dont think there is any uses cases where you want property values that either start or end with spaces +ThanksD +Source file and line number info has been added to compiled rule bytecode but this has not yet been tested +A lot of the code is in it just needs some generalization to support both cases +In the contained youre using the standard file and extend it by an entry for FCK editor +What do you start with in your source documents and what is the result? +We should stop delivery only after the subscription manager is notified of the lost topic and will not accept any more subscriptions for that topic +That is true would do the job +Thx for finding it +I wanted to somewhat second Stefans comment +I can start there while we work out the other far as the host controller and the process controller maybe we just add the handler there and in domain mode we just have a more global operation +for the explanation Manish +Though trying to traverse through the code to figure out a proper way of fixing this is a bit tedious I think a few things can be alleviated by using the to handle the relativization of the paths +Marking Patch Available as this is ready for review +If any of you is able to apply a patch to the source build it and test it I can provide you a patch soon +I see youre using JDK as well maybe thats the culprit +zoomulti and zooamulti crash for the same reason +This makes them have a good name +Theres actually an additional problem we cache the dashboard repository which is normally good however when we switch from a standalone console to a console we dont update the dashboard repository +Not sure why this wasnt working for him but I couldnt reproduce the problem +The back end applications are for business use no company wants their applications telling them what browsers to use across the company +Here is a cleaned up version +How old is your could you try with a recent of the examplesI also see a bunch of warnings for aspects that arent applied thats not a problem and you can ignore those +This feature was implemented for alpha +Was this after a pull from last night or from an earlier revision +I dont remember when exactly. +How does that work for you +Thanks Ill look into applying a patch and possibly releasing if it becomes an issue +I replaced type with clazz and print out a warning to the user that a binding file can be used to specify a more appropriate name +Swapping out the entire stdlib for would be in my opinion too big a change for +Suresh and Colin thanks for the comments! New patch uploaded to address the comments +Thinking through some potential solutions to this I can go one of two ways +Yes must be compliant +Sijie Ill review more carefully and give you comments +One comment on the patch +Thanks much to both of you for your help in debugging this +Gotcha +Attaching patch by Kevin Conner +Some relevant discussion is at +Would you mind if I do soBy the way are there any reason why does not implement Writable +Youll have to be more specific about it not updating +I can try to add a warning if the file already exists but Im not in favor to change the default file can use the property to define yourself another name for this file +The decision has already been made and documented in +How will this affect the upgrade path for those customers? +Btw I am bit stuck as I lost access to almost all ASF services and waiting for password reset and of course do a whole log of search and replace in the template files +Sorry for the spam but Im still getting used to this process +Mojarra was fixed that bug for release +Err reassigned to Farid the author of the attached patch. +Thanks looks good to me! +Thanks AaronI am using CDH and will email to Thanks for the info. +But this doesnt mean you can unassign yourself +Please see screenshot attached +YuCan you attach patch for trunk so that Hadoop QA can run test suite +which should be pretty soon after we branch +Hi DheerajHow do I compile this on a Windows machine? Do I need Visual StudioThanks! file without Michi yes you can compile this with Visual Studio +A test is messed up in the last patch +Mostly looks good I had some minor comments +this is a summary of the other tasks related to moving off seam +we still need to upgrade to equinox to resolve the remaining issue in runtime +Can you explain more how this is different why this is better than the existing bmtexample +Thanks Johnny +Applied the patch into trunk as it breaks the behavior of old Camel version I dont want to port it back to other branches +This helps when folks connect old client to new servers +Amareshwari do you mind adding some illustrative examples for the HADOOPOPTS? For + Although the patch fixes the issue it is just a workaround because it the GB +classpath this is better handled through JBDS and documentationWe should check the docs and make sure that what is present is sufficient +server is killed well retry a good few times +It will only become public if you change the security level +Hi Andrewdid you make any progress on this? We could help finish it as we need it soon and there are other volunteers willing to help on it +patch with review comments to serviceutiltoolsactionjms packagesstill pending commands +OK pls close this bug +Interesting results +This pull request contains the fix suggested by Schroeer in his commentPlease note that this problem doesnt occur in Firefox anymore +There is not much more that we can do other than log +See for example of such deployer +It has been a year since there was any activity on this +I confirm that things work smoothly over here +Fix formatting and comments other git merge the Lucene secondary index we need to report back the RAM consumed +Please reviewOleg +At the moment I have no idea though what is the cause of the problem +Thanks! +but they wont know about the knob unless they follow the conversation on the Derby lists +Regarding the this is a good idea but its moving code and modifying it in the move at the same time +Marek can you make sure we mention the workaround in the notes +I was able to build and deploy the war to a tomcat +Verified on build Version to set a verified by Alexey +Lenny that would be an improvement and this ticket was posted as a bug so please create another ticket if thats the case in the future because the bug it originally describes never actually existed +Ive tested this connecting to S cache store. +Changed the as suggested by youNow the Refund process should not break and run normally as usual +It is as simple as that +Thanks Todd! +What item should I open to see that problem BTW +The problem appears to be a missing definition in of disable +Committed +I ran RAT tool and found a bunch more xml jsp and script files wo licenses +Changing memory settings only changes how long it takes to OOME +There is also no insert before so that discounts my theory of it inserting and then flushing some modifications +If you enable the servlet class loading model using the attribute then you are responsible for packaging the classes appropriately. +In general Derbys subquery handling could be improved +The grails app attached to this issue should exhibit the Hibernate problem the latest changes in GROOVY branch the sample Grails application from now pass the tests successfully +Konstantin Given that we couldnt reproduce this regularly I think it was in actuality a different thread doing the modification +resolved in GaryCurrent Camel trunks version is if you wants this patch to be applied into the you need merge it from trunk to branchDont worry I will do the merge for you in a minute patch is in camel trunk and branch now +Committed at subversion revision +applied trunk did make a change however +I received the Cant create session factory when the JPQL in one of my annotation was incorrect +Including documentations in javadocs might do the job +I admit it is a pain to wait so long before you can make an update to downstream projects +Ive committed the fix for the NPE +This issue is very old the entire plugin has been completely rewritten since this issue was created +Duplicate of which has slightly more related info +If set to false it will skip looking for the projects in the workspace. +Contains an apache ftp server configured to timeout connections every seconds and a camel route that writes a file every secondsEvery other file write fails and a new connection is made the next time +Furthermore the wizard errors telling you that you first have to create an application +I see myself unable to fix the issue without a test that actually reproduces it +Once the async command has succeeded it reports so and exists the loop +Davanum means that he cant recreate the problemRegardsthe code that ive posted is fully working the problem is that after the execution of the client the file that has been downloaded cannot be removed from the server but it is locked +Hi Lily on my XP machine with I am getting a hang running +MergedThanks! +Ok so whats the status of this issue? Closed? Wont Fix etc? +Patch that does the fix I described and adds more info WARN logsIt probably needs a unit test +If you would like me to implement the full solution please let me know +Ive verified mvn clean install +It could be a spring configuration issue +Reopened for port to per Tim Brobergs request +Not saying it shouldnt be done but I dont think its going to be enough +I think we should create a new issue and fix it there +If we have config enabled by default then will postgres be broken or will it choose slow datanucleus path +So please your Spring version and make sure that your do not have the expose flag set to true +Kathey thanks for looking at the patch +but if a bug is introduced by your program +This could also be added as a Struts plugin +Verified in EAP CR + would leave too much pressure on me +We are good for now +rcommitted tested on Ubuntu expect to see features work w applicationof +Part of a bulk update all resolutions changed to done please review history to original resolution type +My changes are extending the way clientPort is being identified whereas here the patchis identifying the hostname to be used with this port +Hi JuanDo you have a simple testcase which I could use to recreate this problemRegardsSen +attach version against patch still adds srcjavaorgapachecassandradb but assuming you dont commit it from me +But please hold on commiting and afterwards + This bug has been marked as a duplicate of +If I understand correctly it will work globally speaking not only when running Ivy with Ant tasks right? so it could be used to resolve +on describeschemaversion on describering +I like Hoss suggestion better +Hi devs Andreas Here i have attach patch with an integration test for JSON and two isolation tests for and +trunk +I would like to see something like Andrzej is mentioning the metadata section at the endOur use case is specifically the example hi mentions the record countCurrently keep the count in a shadow file ie using a custom +Add some bundlized components at revision changes to trunk At revision +revision on. +Sorry that makes sense +Otherwise our tars wont have the junit jar in them +Confirmed Ludovic Claude executed individual CLA with reference number on December +Id be happy to help out +With no further information I am closing this as already fixed +I am not sureAlso in this user need to be careful making both Scan and Delete object +I think that such data should be located in applicationsparty or in applicationssecurityextPlease could you clarify location of the? ThanksBtw Its good news about new branch I did not knew about it +Generate the same number of preferences for each user as in the testing data +Patch applied at r +We found some more bugs in the existing code and new code that are causing the jobFile to use the incorrect user +but that doesnt mean im out to get you +re b +Each with a single step in it using the corresponding +It was actually applied at r +As others have tested on winvista successfully Im not complaining about this much but it does prevent me from running the test suiteFor now Ill see if theres something peculiar with the machine +The version of the patch Ive posted is reliant on being applied to make any sense +Thanks for the quick turnaround Arun +require specifying the classname as a indexoptionI like that approach too +Can you add them as properties so a schema change is not needed +Question mark added Is converted to lower case understand changed to understanding HTTP changed to web +The same happens when you have multi i though we could at least check if the value is NULL return arbitrary truefalse or at least throw a more meaningful exception stating that pipelining is not supported for this component or something along those its not the exception that bothers me +Technorati ping Enabled for all new Blogs +Yes +Skipping back a month or more results in the forward link set to the current month even though posts exist in later months +As a workaround you can switch queue off by setting false in +For instance is fine for MavenSince it is currently not necessary for Tools to expose its parent pom publicly +Not very DRY indeed the method mentioned in the forum would be very turns out that in the previous patch was a serious bug +Attached the just in case you might need it +Per the JIRA triage meeting on Oct Henry Z +The method is called from this allowed only because they belongs to the same packageThis is blocking when you want to extends and rewrite one of the method wich use get writePage is +We also tried googling for the specific problem but without JBM cannot work with MS SQL +I dont see the point in separating them +I havent had time to explore this yet +Could you please give another try to generate the patch file +Thanks for the detailed report it made discovering the bug quite easy +I like it +Well be upgrading to as soon as we can but this seems like a significant issue for anyone doing range scans does it make sense to backport to +bq +Marking this as a dup of +It does what it is described on the pdf +Great Ive cloned it and will have a look at it tomorrow +For instance now the following will workmvn install pathto up old issues +Revising +Committed the a patch to trunk revision +As there is no test Im considering it resolved until proved otherwise. +trunk +New patch where the method allocates returns the contents +I my last commit should be fixing it. +Hope for instructions especially from my sincere mentor Myrna +After assigning this to me Ronelle told me she thought itmay already be fixedAssigning to her to confirm +on patch +Hi AndrewThanks for quick response + +I tested a couple of queries in ij and it seems to work fine +if I have both jars in hbaselib then stargate tests all pass +Next step fix on behalf of Andre +If I look at sort then we are putting data on MR in form of +Tangent We should migrate to anyway +member variable row was never referenced apparently its reference was removed in +Is it possible that the patch for interferes with this patch? Applying this patch fails for me +Something might be wrong +Verified JBDS Linux +added support for show partitions +Committed patch with revision +Fixed included and +btw +A patch I uploaded is available at +the ability to change for field foo from one document to the next +The problem is due to the metrics that we have to check to verify if configuration is valid +Hopefully the logs will have what was missed +s in the previous comment +That is what the Atom spec means when it describes the link tagIf you were expecting a single document with lots of metadata for all the link tags then Im afraid you will want to review how RSS and Atom actually are supposed to work. +Thank you! +its kinda wonky having list of regionserver coprocessors showing in the UI load +Looks like a buggy connection pool or a buggy Transaction manager implementation to me +Thanks Andreas for giving Axiom developers perspective Good to learn further about the real cause +Those are just the static definitions of the flows what views they contain what stats actions etc +Failover work is basically done aside from some additional unit tests that Id like to add +What is the rate with the original syslog adapter in XD +Will switch gears and work on that a bit instead +Seems reasonable +The v patch left out treatment of the object member function +This task also must include a relaxed schema definition for MSSQL that does not support circular or self referential cascading deletes +Should be fixed during Jive plugins in SSO plugin v +Does somebody know when this bug will be solvedThnx +Its close +I suddenly realized I dont have commit access! +I suggest we change the current geronimo build system to use java instead of java +Parenthesis was wrongly patch should get the caching mecanism up and working +But I wonder about thatChange it if you dont like it +Seth had sent me a stylesheet and xml doc to reproduce this bug +I see your point +I disabled this test for client until this bug is fixed +Looks great +I can confirm this one looks really ugly in my app +Though that code will have to poke into the design doc to handle that warning case +Yes verified that it all looks good +Do you think it would help w the issue you saw Gao Jinchao? If so I can commit +Secondly it shouldnt go down itself on sync disconnected +Thanks for the comments Josh +sebb are you sure this issue is on our side +I was just referring to future translationsThanksG +Ill try to make it test this +Sorry for the failures working on identifying the root cause and a fix +Thanks +Im not sure if it has to do with my changes or the way initial content is indexed +I couldnt find anywhere so I tried +We have some internal API that needs to scan based on a external readPoint +Any chance of adding an explicit spec to ruby lib for this +Claus saw this on OSX too +ok but only this afternoonIts not urgentthank and java filessmall to solve the issue but at the moment only for all the patch just submitted here seems to solve the issue but at the moment only inIf all is good I can apply the same behavior even in all other places +Spoke too soon +fixedPlease +Pushing issues to the next CR +Works fine for me now +Revised patch based on comments from Abe +OK great thank you for the explanation. +Some labels moved at Common levelIf some framework committer can review it I can continue to cleaning the labelsThanks committed in revision +The configuration file used by the enclosed program. +Bug fix in script +A little test project would helpNote however that relative links are looked for locally +Can be reopened if this is to be put on the roadmap. +Everyone seems to think this is good myself included +Acegi still has its own problems with the reauthentication path after timeout +Petr Do you have an ETA by which this issue will be resolved +Ive read through the specification that youve attachedCan you suggest me an entry point to implement the spec? I can see that solving this bugwill help to solve the other linked bugsThanks +Like +That is we raise a too much contention exception only if we have looped longer than the timeout set byWith this change the too much contention exceptions disappear +I was hoping the new feature was getting the tires kicked +Really cool +This should probably be implemented as an optional feature that can be turned on or off by some means +It seems wrong to fix the class name within the Javadoc plugin so at least it should be possible to override the via a user property as well so the user can add it to their the Oracle tool has some bugs so it would be useful to be able to use a rd party version of the tool which does not have those bugs +Ive just kicked Jenkins again +Console +mono is available usrlocalmono +Anyway commited in revision Thanks for your consciousness +So what is exactly the problem does the selector hides some resources? Could you please describe your usecase to be supported +this looks like it would best be a web console pluginIm not against that will have a look later unless someone beats me to it +Thanks Ben +I have planned to submit an ICLA I intended to do so but due to my vacation it was forgotten +actually that would work but then each release would result in commits and that is not what you want +New patch that works with localization +sbt publish +See for further details. +The version information for a file would have to contain the URI prefix in addition to its current contents in order for the incremental crawling to work correctly if the Web Server prefix were to change +Resolving as Fixed +Changing the title of the bug since this really could occur on any version if the plan with the problem is chosen +Fixed +The only edge case I can think of is an application which was coded before we introduced sequences and which created a table with one column an identity column used to simulate a unique id generator for the whole application +never rocked my boat +Im beginning to think something was introduced in the SSL patch that altered the behavior the socketsNow Im getting failures wo this patch too +So if our goal is to exactly mimic Java the reflect API would generate unions with null for nearly every field +I think I was scanning for return this and somehow missed return path +Fixed it here +I couldnt care less what the address was + Olga +Maybe we can streamline the process a bit +Guys great job in moving this forwardI am sold an all the improvements in the patchm going to give it one last and make sure that passes otherwise Ill commit it shortly +This check is just for trowing away an evidently incorrect combination when heap size exceeds the limit for compressed refsThis value should be changed if compressed mode will support more than G heapI guess OS limits checking can be performed by GC before Java heap allocationAnyway if specified heap size exceeds OS limit then memory allocation in GC will fail and GC initialization will return error +Ill take a look ASAP +The embeddercli will check if MHOMEextensions exists and processits contents during reason its not in is that is for the classworlds booter and contains no mavenspecific configuration whatsoever +This issue affects servlet resolution and is fixed in the next target release +The plugin must execute AFTER cobertura to prevent cobertura from having a corrupt classpath +Switched from using to and added overrides for the dispatch methods in order to support encoded responses other than the default. +No problem AleksandarI have removed the step from the repo version the change will be applied at the next stage! closing the issue then +Please try +Nothing changed as far as I can tell +This patch for subtask provides these improvements adds timeouts to all wait for condition loops adjusts config parameters to speed up test execution a lot improves log messages clarifies code and fixes obsolete and incorrect comments + committed to trunk. +No this doesnt affect yahoo +This patch reorganizes the config file to have API properties +I attached the thread dump to this issue +No need for IT +bulk defer of open issues to +added as resolved in application servers JNDI properties were being set by the +Retrying Hudson to pick the patch +Nobody gets paid for working on so please do not be aggressive with us because nobody wanted to handle your requestThank you very much +close off release +Thanks Venkat +If I understand correctly the job configuration file is named after the job id which the unique identifier would be a part of so they would not clash +Duh +The differences are A merge conflict in was resolved +Hope that makes sense please let me know if you need more info +I started to work on a unit test for this issue +Please look at +See especially and for more detailsI guess thats enough for now on this issue so resolving as fixed. +Please review +I would rather recommend using the classpath style which only picks the nearest resource to begin with thats usually the first occurence in the classpath +yes +Applied GEOT patch and tests for DB are ok +I believe that the changes for this issue cause a regression seeRuwanGlenCan you have a look? +and some other classes that extend Projection will be changed to extend +It is true that they can be cleaned up manually but it is a pain +Meciej this is awesome +I would prefere to handle now the whole locking semantic in our own classes to handle this behaviours in a global waySee mechanism is meant to synchronize only JCR API calls so unlike with XA transactions there should never be a need for two threads to use the same lock concurrently +The RM does not read the at all +Should all metadata keys be? If so then we should instead probably implement this with a Properties a using +This is a modified version of sample app faulthandingChanges are in andI dont think that I can look at the code you have put up until you grant a license to the ASF. +Patch file Committed revision . +LucianoI modified all the modules to use Felix rather than in the branch +Go for that Dan +I think it is definitely another manifestation of the same bugAttached is which includes the test case as part of the testsI ran the test case interactively and without thepatch I get the NPE and with the patch I get reasonable resultsSo I think that we can mark as a duplicate of thisbug assuming that we decide to proceed with this patchReviewers please review instead of the v patch +beautiful +Ive confirmed that this is a problem +Covered by. +Im just resolving it + +Here it is +validation kicked in after the route was linked together. +I think I found the problemMy bad! I tested it against my local version of mapreduce in which I had taken out + +So reducing the number of operations in the CRC algorithm may not lead to a better performanceI tried more varieties like Crc but my implementation did not perform well +Thanks again +Hi KurtPlease see my previous postWith we introduce a new property componentRegistry for cxfseproxy you just need inject it when you useAnd yeah document part is what we need improveCheersFreeman +it is trying to get the connection for a reason +Also use the good url wherever guvnor uses it such as homegdesmetprojectsjbossdroolsjbpmguvnorsrctestresourcesorgdroolsguvnorservercontenthandlerm also pushing a modification the the file that changes the scope of the that bothers to deploy the war inside tomcat +Will you also remove the dependency from +perhaps you should look into getting more RAM as any modern app server JVM various frameworks your app +See new test cases output and output +A twist is that the server needs to ignore the initial SASL message that the client has already sent and continue with simple auth +This patch has been committed +I think the page should just be turned into a wiki page +heres the patch to fix MDWnow more tests other than fail +Patch for for trunk +Can you give me access too my id is metskemthanksHarry +so we should decide if we want to include it or release it as a gem or what for +This is fine if one family per region but if ten families per region and k regions on the server could be a bunch of juggling of files by master true +Guys please review +what if we just applied the and reverted the two changes you have made for this issue? Wouldnt that make sure we lose no data +Yes it works with myfaces latest snapshot and with RI. +Thanks Ted +That would be good +I did have a reservation about always guaranteeing a sort order +Mechanism of the queue collection reader was changed +run from the qpid directory +You can manually change column type in DB +Some compiler implementationswill make the pointer go with segment fault when access is attemptedoutside the methodDeclaring a static variable in the class as an empty string will preserve itsusability outside the methodsI will create a modified patch and post it +Agreed +Example patch showing proposed enhancements +If you dont agree assign back to me +Also made some changes to the test case so that the queue subscribe timing is more deterministic and added some more stats output at the end of the test +Patch is ready for review +Unfortunately the guess example in the showcase is throwing an NPE and due to the bytecode manipulation it is difficult to tell where or why +yes works for me +Sometimes those columns are the last ones +Marking closed. +Thanks +Thanks! +If not then probably the classes should be extended rather than creating a thread safe wrapperAs to formatting AFAIK there are no Eclipse downloadsAs for any project keep to the existing formatting in a source file unless otherwise advisedRegarding test cases for different formats the class requires the format as a parameter to the constructorTherefore it is vital that the tests include more that just one format string to ensure that the parameter is handled correctly now and continues to be handled correctly if the class is ever changed +closed +Also I agree with Daves idea of including a method on the interface that allows a developer to return an exit code +I see no other way +With this aproach Roo could to generate HQL queries using specific operators for the DB you are working on +Im not sure that its useless though there should be tests for non distributed mlt so if distrib mlt matches it that seems like its doing something usefulBasically I tried to bang out the simplest thing I could that seemed to make distributed mlt possible I had absolutely no need for it but its a common user request +Im also quite reluctant in adding some syntax just for this mainly for the cons expresses above that is the fact that it confuses quite a bit the we dont do message and it doesnt make it very clear what the restrictions on that are +Not found with latest runSheng fixed many vpc issues +Also installed fine into Fedora x Eclipse Galileo JEE bundle x Open JDK and the above update site zip +PS I disabled this in SUs internal build for trunk because it makes the master UI impossible to use on a larger cluster +supports DOUBLE as primitive typeAlso it supports other primitive types thanSTRING and BYTES for MAP and RECORD complex types +This cannot be avoidedIts a pity the class methods are all staticOne other possibility would be to allow the default to be set by using a System Property this could be but the value could not be changed once set +Thanks +In the case that max of returned entries is each directory listing requires only one RPC callMax of returned entries per getListing RPCDirectory of entriesDirectory of entriesDirectory of entries +but sometimes if i ran all tests with it would happenthe reason this test is not reproducible is that this test uses n seconds as a limitso whether it passes or fails depends upon what your computer is doing at the momentI think we must change it to limit itself by number of docs instead +TOVC should continue to work as it does today +Hi MichaelThanks for pointing the language resolver issue out +Contexts must always be joined if they are alive +Trails runs beautiful like it is suppose toNow I could start making comments about the brazen contempt but I am going to bite my lipIve got commercial stuff pending and praying for your bestYou may want to closing this and opening it up and creating a test case for itIf the capacity to produce a workiing app that shows this for you comes anytime soon to me I may deliver but right now it does not look that goodTo double confirm my findings I bumped back to and Yes there was a minor usage of contrib that I previously added which was not part of our codebase beforeThe above revs appear to be handling things now less a new exception that I will chase that is propably of the same natureSo for now leave this closed +I try to provide a patch the solution is to create a wrapper of the request and response object +Done. +Some clarification fromXSL spec is needed and is expected in the CR version +ok thank you Eugene I have installed it and will have a look at how it works. +patch +build absolutely cannot return an empty string in any other context wo wrecking havoc on token processing +Suggesting patch based on branch +I think we can close this issue +Requires an update to requires a fix in Shindig and an upgrade to +Agree with the proposed way forwarddirk +Sorry but I didnt understand why you think thats not a bugIm using Findbugs in both cases and have the same effort and threshold in both casesIn fact youve also admitted that it did slow down for you with medium and normal thresholdSo IMO there must be some difference how the maven plugin controls mustnt itOtherwise I couldnt explain how that time difference could be +M forgot to update that piece when i updated the Title to be a general cross domain feature request +New patch attached +as I need to add documentation +pleaseignore attached today its the same as privious patch +Removed junk file +is anyone working on this issue +Stefan thanks for quick fix +See at line +Oops typo in the XML made it malformed +On Windows if the testing path is Cbuildtest then this logic yields C which is not an absolute path +Looks like it is fixed while fixing other issues +We build with a Java JAVAHOME variable and maven As for a command I just use c mvn cobertura you run mvn with the option X and attach the output? Please make sure that no passwords are creeping in +And when there arent a default properties file like but only bundles for the specified locales then the dialogs table wont be initialized for this bundle +Looks like resolved with addition of synchronization block. +It is not sure whether the exception is thrown from Xalan or from the trace handler +I then ran it exactly times and it did not reproduce +Attachment has been added with description Patch adding static declarations and removing trace this check has no cost +smile I think its a good compromise too +Hi UweThanks for investigating thisLucene is repeatedly hitting Javas inner and outer boundaries also in this caseI agree that were safe here +And here are the tests themselves +Marking closed. +thanks of making the parameter to look for the certificate configurable in Rampart wouldnt it be smarter to define a standard message context property that would be used by both the servlet based and the NIO HTTP transport to expose the client certificate +Benson fair point and I would certainly not want to add to the startup time +See attachmentThe behaviour of the second case is to be expected because there is no possible breakpoint +this should be already solved with version please try and if it works close no longer is being thrown in version +Martin I tested your patch and it yields the same performance benefit +I think this issue is now resolved +r on rubyspecs now pass +There are no information whatsoever about the MINA version je used java version and from the two exceptions shwon in the trace it seems to be an issue in some initialization in the user codeThus I close the issue +Also whenever we change the interface or any of the associated classes it is an incompatible changeIn our case uses which is a part of +Im closing this issue for the release +I believe a patch would be very welcome +I think you was rightBecause Create empty has the functionality of old wizard it created only wizard in the specified location +is most likely a classpath problem +Thanks for the bug report! +pushed to fixVersion to +Added fix version see Make sure all fixes on branch are applied to branch. +Made this issue major priority rather than minor +Patch looks good to me +Does the path find its way to the subversion repository +Awaiting SMX people to release an osgi bundle of flatpack committed with rWillem do you mind double checking +fixed this in branchtrunk not happy with how error handling is done since main publishing does start but if error in module publishing then that error comes afterwards +r This is easy to fix and IMO a release blocker +Yeah + a change of +For example I am having similar issues with extended ascii characters +I just verified that the AS Beta launches on OSX even with spacesI also verified that it does not work on Windows but that is a different scriptscenario so I opened to capture that. +OkaySo is it official to use these properties Ive found in the codeAre they documented anywhere? +ran on ubuntu on ubuntu +Such things should be configured in the pom of the specific mojo +applied +So well see what their response for the update +Since its in the examples a few brief notes on the initialization of and on nextPoint would help orient readers +Also will give a trunk version once based version is fine for commitThis delete performance is being discussed in different user mail threads I guess +Jeffrey the bits youve pasted from the pom is irrelevant thats just used at build time +While it is true that in the extreme case people could configure a different for each field which will lead to a large increase of disk seeks here we would like to use the for all the standard fields but have a different mechanism for specific fieldsWe would like to store certain fields that requires a different type of data structure than the one currently supported +Actually its not only the config files its also the classes that are still there unused +A comparable should still have a valid equals implementation right? Also from the performance point of view this would be better i would the end this pull request should solve the you +See for details +Bulk closing stale resolved issues +Also marked fixed in to fix fix to fix fix version +I would suggest that default be changed so that if youre choosing to update a model that the user needs to select to have it deleted rather than default +but thanks for letting us know +Attached unit for the test case AlexeyProblem was inAttached patch fixes this in plus adds the to existing. +This makes the clean actually work for test databases which should improve reliability no end +Just trying to start a conversation +I suggest running the integration test and you will see one failing test which shows the bug +No one has been able to reproduce this fix and it has not been reported for a good while +Im confused +Committed revision +Miss configuration +I dont know why Hudson is not picking up this patch +JimmyYes you are correct +That class is using the JSR and defines an additional validate method +I asked about j because badConnection was listed in so some clientserver tests were at one time run with j however inappropriate that may have beenGood point about the exception text +Since the X token profile requires that a wsse is used in ds and also specifies a limited number of mechanisms by which the wsse may referenceinclude an X token the situation where an embedded X certificate is present as a descendant of ds cannot arise +Ive added an explanation as you suggest Thawan +all of the empty plugin name providers ive found in the about dialog have been updated +We should probably only do that for Application messages +Your approach of caching writes locally and then writing it to HDFS might work as long as the writes are in the same HDFS block isnt itBy the way a new feature ability to reopen a HDFS file and append data to it is under the works and might be coming soon seePlease let me know if I can help in any way in getting your server hooked into HDFS +I am looking at this jira entry to see if something rings a bell in the code which might cause this npe +VikasRegards Akash Jain +Also when I skip tests I run into the problem described in +Release audit warning is due to jdiff +I can add a note to the documentation +Closing old issues +Using spaces +must be used together with a proper startRow. +A second consumer will see the new messages but a good message cannot jump a bad message in the queue for a particular consumer +Verified on build +it might be a good idea to have a version of the assert methods or at least for select that takes an array of users so public privileges can be asserted against all users +proposed patch +Committed at subversion revision +When that error happens anything in the console log? Copy and paste into here +make versions tomorrow and commit +Tests passed cleanly for me on +any thoughtsThat is reasonableSeems like the patch has gone stale can you rebaseFile a MR ticket too as a followup tracking MR changesPatch looks mostly good +Extracted out all messages to properties file +Documentation fix! Love it +Also added the class +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Link to JBAS feature request +And your argument is super generalIf you want a real test for monitoring perm gen usage make one but its silly to count on variable ceilings in the wild to be high enough to avoid that kind of bug +I think LDAP Servers provide for storing a hashed password and the algorithm need not be same for all passwords stored in the LDAP server +Could you please provide with a sample code code and configuration in more detail? I am having difficulty putting this together based on your forum +The problem with the code is the way that it uses the Binding URI to look up the target Component and target Service +Thats okay Kristian though I think it would be helpful if you backed out the patch because based on Stans comments I think I want to go back to the previous version for some thingsStan thanks for those additional comments +Extending this to also expose on which disk on a datanode a block resides would enable even better scheduling on a rather than coarse basisI am not sure I understand your motivation +Youre right I was being misled by the comments + + +Once I understand I will be happy to make the changes +Hi James the fix did resolve the issue +Claus Im working on the documentation +Whenever the merge must merge doc stores and at least segment is referencing the currently open shared doc stores IW must flush +Let me know +BTW we are employing a Fedora Linux bit system +FYI +This is only on Windows +Committed new patch to SVN +Thanks for your help! +Username ssswdon +Test fails because xul does not initialize and test begins +Thanks for the patch Tony! +I am still not clear how this approach supports change of Checksum between blocks +No more refactor for testsuite +Any? As of now maps and reduces can manage separate filesarchives themselves +bq +Please verify +I will implement another say orDo you archive any other files besides log files say +I think it would be more correct to generate the files under target and then configure the to include the generated files in the war +I added and in the require bundle section of +I guess its good enough to have a separate with a higher index than the then? In that case Ill close this +A new controller state change request and response is defined and +Regression tests ran cleanly for me +No worries thanks for the review +Thanks +fixed will commit later +Hey RakeshI modified your patch a bit and came up with this new rev +About potential disaster well maybe I should give it a try thank you +The test runs fine locally +In IDEA its not an was fixed in r. +The EWS one is scheduled +Reverted in revision +When I said two nodes going down very close to one another I meant that for this to happen we would need one node to go down in succession that had the correct processes running on them +Need to look at this and figure out if that scenario is actually possible and if so how do we fix itNOTE This scenario was a problem even before rev was submitted so were still better off now than we were before +If we commit this we should have an idea of the impact +the node type has been added to the node type cnd +Am working on a patch which i can upload for verification +Marking the issue as resolved +If Guest traffic type is supported on more than one physical network in the zone the tags have to be defined for all physical networks +A tiny aspect that does what you want at load time +Pretty sure its flaky +See the readme in the source code tree root about how to run daytrader in geronimo we will keep the readme file updated to the latestFor jboss we do not plan to write a detail doc +This appears to work fine +Currently the shuffle is still the slowest step and until that changes thats where the attention isGigabyte blocks could prove useful +Yeah a Linux server running FC +I think the same goes for autocomplete +Fsck is useful when Namenode is misbehaving like not replicating blocks not deleting blocks etc +Hrm +But tcloadBundle has a different semantic +I remove the test from the JBM testsuite for new TX m closing this now +Can I help you if I set up some DEBUGTRACE logging for some packages? is tcplocalhost nothing special configured there +Perfect +This is a unified patch that adds all the source to contribsrchbasePlease note that this is a so while it does compile it is not complete +Where are you doing this? I get a completely fine dialog when I do Run as Run on server in project? jing +If you are certain that you dont need it for a particular use case then you can exclude it using your build system +run mvn javaccjjdoc +Weve spotted this problem before fixing it is tough +with test version make me happier and lgtm +I just committed this +Thanks. +Ill attach what I did for my project +It always returns null when the schemeauthority is missing +The code for PMDCheck is done. +I have the same issue +I merged the patch to. +I opened couple issues around download and upload and I was talking about iPhone not android +didnt see this anymore in the riftsaw with jboss as forward to QA to check on the SOA if it still exists please it +Im tempted to mark this wont fix this is some problem +Thanks Eli +Could you provide a reproducible test case for Strategy on Embeddable +Thanks +Hence moving it to +now we did it in parallel +Just reopen this JIRA item if its not working right. +Thanks for the help with this one Rob +It is a good idea to ask to about this issue so lets see what happen +Christian if you can get the patch for to clean correctly I can start commiting these various changes +This target can be helpful for contributors to verify that new code has a correct formatingIt is a own target that can called by ant checkstyle +The unit test should load the vocabulary and output file as a resource instead of hardcoding it partly +yeah can do +The changes look good +Please tell me where exactly to write +Cancelling patch under discussionIf the issue has changed to fix the broader doesnt problem then the titledescription should be updated. +done +Resolving this as Wont fix based on Fernandos comment +This is an initial patch +I have committed this +VS non reentrant readers why not +Thanks Eric Ive fixed this in SVN now and itll be included in our release today. +Hi Andreas the main cause of this was that the checksums were being created and the metadata updated after each artifact deployment as the consumers were executed +Second attempt at making Jenkins happy +I was trying to file an JIRA +Attach a patch to add an async interface for multi request in java client +Added utility method in to make this easier +I agree +It also used a separate dataDir but thats a completely different thing +However I am attaching the relevant code snippet and stacktrace +Simple patch that honors as an appropriate state for activation +I think you mean the justed linked issue? I just happen to work on that issue +patch to fix the and rename the do you need this to get into +and then it works ok +x directory is prepared and pootle changes analyzed +Thanks applied patch +HeyI was wondering what that actually helps? You can already specify a combination of columns or families with the existing parameter +There is still the potential to revise the concurrent flag to make it batch journal writes. +Its a little more complicated on servlets since we need to calculate the servlet path but its possible +Any chance of a review on this patch +What we used to do was bogus treating the scheme part as a name component +Verified on app Engine. +keeping deprecated stuff around is not a bug +It facilitates checking this fix +Im not sure how useful this is +Defer to along with +This would decouple any contention from the scheduler from calls to this REST API +Hi TimIve made XT Ajax compatible with Servlet Ill commit this change soon I think more people could find it m attaching here a new JDK compatible version of the give it a try and let me know if its ok forSergio B +What I wanted to express is that its not easily doable at the moment which is actually a reason why I did not yet started and let Patrik to handle itJacec +bq +migrated configurations to +Alan assigning to you lets try to review and land all these patches we have in the queue and make a release +BTW none of yarn daemons use jsvc +The latest code in CVS has my serializer contribution +I see +Initially memory may be sufficient to hold the array user sets in +Then when invoking hooks use instanceof to detect whether the class implements +Fixed in r. +Agree with Robert here +and +You can help speeding up the fix by fixing it yourself by cutting ainternal release of and load it thru mavensextensions to see if the problem goes make a quick fix just comment out the place where knowhosts fileis fix this we need to do a comparison between the local copy and the one in sch instance +Marking closed. +Would you please try on WindowsThanksAhh +Thanks Uma! I didnt think I had privs to move but I learned I can move it myselfUpdated patch to add the tests that I forgot to svn add +attached the proposed patch for injecting service properties using a Map +Thats what Im doing now +Koji I added a slightly modified version of your patch +If you guys want something more comprehensive done to these tests Im happy to spend some time on it I feel responsible for setting this course of action into motion anyway +Deferring this +Aniket it would be great to get committed +Patch to update the Eclipse configuration +Ales fixed in and its subclasses to reduce logger creation +If we want to remove the other side from each Task we can do that in separate jira +Do you tell mehow can i use snowdrop to complete integrate +Just extract run mvn compile execjava and it will start an embedded broker using a configuration file similar to your own +That change is in get method +This looks to be fixed in the branch now +This has been implemented waiting for an OK before merging this in +Thoughts about? My gut feeling has a slight preference towards no +But in this last trace for example isCase does this since it is an uncached dynamic callhas someone reading in this thread any idea how to prevent that? I could apply some logic from the primitive optimizations but EMC is disabling that +Instead of the whole CID just the parts before and after the +committed a fix in close of resolved issues. +API seems to be fine need UI action button to acquire UI engineer needs to pick it up +I think have changed the old style of ASF license but have not added one to the files that didnt have one before +Does the new flag cacheBlocks need to be added to the Writable serialization +I agree the test failure is unrelated but would like to post another patch anyways to address comments above +Added table as Appendix in Users Guide +Thanks a lot for your quick review Suresh +The observed object name and observed attribute name are provided in the java explicitly ignores null monitored values and fails to emit the OBSERVEDATTRIBUTETYPEERROR +options should probably be limited to +Debug log +I miss you here +heres a use case for load balancingIs it a point against this or for this +Apologies for the inconvenienceIf youd like to help out check out srcpythonsrc +Pull request +Attached second patch to prevent unnecessary creation of during component shutdownDocumentation has also been updated in the wiki +The work around that he had proposed was not effective. +Bulk closing of issues that are old and have been released or marked invalid. +bq +annotations unless the annotation is present is correct or wrong +submitting to hudson +m create another patch without changing the formatting as it doesnt help to understand whats really changed +The test is very simple If you have a product like which has only units in stock and you order but set allow split to true and then create a picklist does the picklist show or of? Previously it would show in the quantityavailnot avail columns +Community documentation page was my first idea where to look for documentation +this bug is still there +Rising priority to Critical +We put their udf search path first +I hope to fold this patch into that work in the next day or two and commit + +patch reviewed results on the branch overall +Patch looks good +Reassigned to Dave to improve the error message that produces +There can be too many high frequency typos and low frequency correct spellings +Will be in next merge + +There were skipped for fileUpload component +You are welcome +Fixing as per comment also moving some changes +I think that deregistertrue is fine +Another comment akin to what is to be done to get this included +Thanks +Closing all resolved tickets from or older +quickstart +As such of classes is a necessity and the XML instance documents will have to be altered as well +Red Hat Palette Tools to next +Just stating that kafka is not present on maven central right now and it needs to be pushed server separate and client separate with the proper dependencies specified in the pom so that integrating the java client with a java application becomes easy +Nice one Marcel happy my side if you want to close this one. +Ive d on reviewboard with a few minor nits +Ive created a sample project for this +We are already releasing artifact which is consumed by downstream projects +The for example fails +Is this a issue or specific to some flavors of Windows? From Alan Bs comments it sounds like a generic issue with these methods +let me know what you think. +Committed to branch and trunk +This is fixed just needs an is up and happy. +lets give more helpful errors +Test added Davids point is if bigger than use an unsigned short otherwise use an unsigned byte +The currently under review for sync now includes +Let me revert this change first +Logging should be something that the user can enable not something that must be disabled +Now that is fixed I can now see the above stack trace with the latest Seam trunk and AS trunk +Also note that it was secure setup +Sending Committed revision +svc files though +Why not reopen this one +We have the same problem with Hibernate and SybaseIs there any chance that the provided patch will be included in the next release +Again does that sound correct to youIt is easy for me to get lost in section +There wont be any difference its kind of a syntax sugar +Here is a test case for the doesnt emulate the entire auto login cycle but only check the systemOperation flag in it fails with current trunk and passes with the patch +Ah thanks for this catch +In one place the network driver assumes that a will have exactly characters +Is that a safe assumption +and the manual workaround. +also can you add describe partition in the test +On further thought all or almost all of these problems can be solved by adding add and remove methods to Fragment that add and remove to the private fragment list before security filtering +Is this now redundant wrt +We can repurpose this JIRA to have that documented. +I also combined the into the as they should be one and the same +Probably it is right to keep waiting and let the timeout monitor to deal with itWith the timeout out period will be dropped from minutes to minutes which is a little bit helpfulThe issue here is that we cant see the assignment of root is completely failed +rr +test output after hang +There is no Tools release on the calendar nor is there currently a Designer release on the calendar so I dont really have a preference beyond +Are you ready to provide a patch +Removed tab characters +mrepositoryjavaxmailmailCDocumeUSERNAME +Hey all Ive just this using the latest code from the repo and it works great +Here is what Alberto had to sayYou need to release the node I dont think that detaching an element makes it unsearchable +Thanks Rahul! +I am going to try to repro in house and will open a ticket if I can make it happen +Thanks if you give invalid input the API layer sets them to the default value +But as I do not like having empty stuff lying around I rather implement support to cope with missing stuff +However this never happenedId like to see either an option or just the default to be when a name is given in a configuration file Hadoop always does a host name resolution on that entry prior to connection +Attached patch +Should I provide output myselfSure it never hurts to do this +javadoc deprecated tags have been added to and removed from trunk +I dont believe there are changes suggested to the API just possible additionsWe can iterate or improve before +The reason is that if it does not work in osgi then many users will get disappointed +Thanks Regis. +Heres an initial patch showing the life cycle plugin +Please use the users list to ask questions +No startup errors with latest Roo code +I see +Hi SanneI dont think Ill be able to dedicate time to work on it as I dont have a development environment set up for Hibernate Search and I tried to set up one quickly and its a bit cumbersomeI think I detailed the problem quite well so you should be able to fix the problem very quicklyIf you have any question about the problem or if you want me to detail anything feel free to ping meHave a nice day +What other patches are you concerned about in particularSami yes the fuse stuff would then make a great hadoop contrib package +Thanks Aaron +basically it just needs to be docFreq +Yep that fixes it +Ive been trying to formulate a possible race for this for the last hours but I havent been able +As discussed in the mailing lists this problem should not happen with the current codePlease provide a quickstart so we can see what breaks +Ive switched trunk to use to start +I ran all the tests locally as well +at that time excel was not added to the codeI have since added the new mimes but have not had a chance to check it out +This bug is not reproduced anymore with JPDA tests and Eclipse debugger scenarios +With commit adfcacbcdfdabcaedae and eaeabbbdeada this work is complete +If a project was checked out from a release stream the ivy settings would be overridden to point at another repository for the specific release +So same as regarding storing this request in the session +Im going to take this back and solve the problem in the client initializer codegen +Attachign the patch from to this jira +Planning to look at this for marking as critical +Werner as far as I remember I did these tests to the different milestone releases after and each of these showed the same behavior +Sure I will take care of it +Lets first make sure we migrate and can run the server and targets plus all our tests and then worry about the continuous build and how to setup everything for releases +Converted to feature +also has its own delegation tokens so we should consider all of the interfaces +Fixed in svn revision thanksdims +No changes have been made to the skin selectors for train +I forgot to do thatThe tests are actually in the patch for which depends on this patch +So if u get sth for the submission version well its hard to believe that the samething is not working for Final versionPlease do provide me with a test case or atleast a way to this problem in my machine so that I can fix it IF there is a bug +It returns different results forthan forab even though RFC says that the two are equivalentI saw the bug cause an error in live gadgets +Nothing to do with Hibernate +Apologies I couldnt get it on to trunk there was something screwy with the merge +is there a reason why this is limited to pages? can we not have something more generic that can be used in case someone somewhere needs to serialize something elsealso Pedro is right +cleaner code is definitely a to move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Removing that annotation will solve the problemAny updates regarding this issue +AdamThe WAR doesnt have a file therefore its not a bean archive and theres no defined for it +to trunk +Youd have to hack the client a bit which means that still on you get the retry was thinking so close +For those that needlike having slimmer client libraries they can still do that those that feel safer with a fat jar go right ahead +Obviously something that we should do +Regards Paul Elschot +To generate change the PE so that cells are bytes in size instead of the default bytes in size +Testing effect of changes various complex queries were created and run against and versions andthe results and the test descriptions are being attached to this issue +I have a gateway use case where all the workflow run on same topic but uniqueness is handled using experimentID +Grepped through the code for c and didnt find anywhere else that needs this change +At this point the includes are refiltered based on the created for the project +There is no globally shared salt at any point +I dont think you want custom reset times +Len need your to proceed +Again this issue was about shipping a complete distribution not about behavior +Expanded your hack so that its less hacky +Also the configuration mechanism will probably change for probably making this an irrelevant issue +I think we should restrict the changes to just what is needed for the release +And how can I guess the actual enum type for this entityHow can I create an appropriate enum object when loading the object +The fix in looks good thanks! +If use client command set a then use the spring interface incrementthe result seem to be right +Thrift switched its logj dependency to slfj in its like Thrift is going out of their way to make life suck for us this month +If there is still an issue please open a new bug. +Im pretty sure it is hadoop specific +Committed patch almost as is +moving open issues in X DT to X DT as discussed with Bob +Tag library reloading works fine in my tests if it doesnt work please with attached sample app that reproduces the me the problem exists with and SVN head version as of today +Verified that failing tests pass locallyCommitted to trunk andThanks for fixing this Cheolsoo Koji +Per default Axis uses as the port value for the auth scope therefore the port the target server is listening on should not matter +Im not sure why you are currently allowed to enter a type +and are unnecessaryThe changes for and in should be handled in those separate +I am using as my AMF provider +Retargeting to +It is a feature request but approved by PM and GSS. +Tests running successfully and the sql scripts have been carefully converted +Have you run all the unit tests +snapshot on the rat incubator website theres no indication of any official releases and their documentation on how to get started with rat just says to checkout from svn +Thanks +Its not receiving new data so the score doesnt change and the dead host is still rated the best until the FD removes it as an option or enough demote it +FAQ should probably be towards the end As this is specific to Hadoop may be we need not mention as a bullet but may be can include a note on it later in the text +Perhaps we can updatebreak the interface for +As a nit you could remove the whole Known bugsmissing features section since there arent any issues listed +I am going to deals with it +Note +I guess its the same problem with no embedded fonts as in +My vote is to put it into trunk and backport to +Also I tried the Geronimo with JDK which ships the JAXB API those cases pass +I will count the number of clusters and divide by that +Sorry +As soon as connection n I would see the initial data from the server on the screen of session n +Will or do the trick? A far stretch will be utilizing +Taken care of months ago without benefit of JIRA +If Block is chosen it is not very optimal as hosts B C have only bytes with them +I tried to address the questions above Ill take another pass through it tomorrow to see if I missed any of the commentsIll also work on adding a section to compare and contrast the two alternatives that were still debating ie file names with transaction vs file names with sequential numberingThanks for all the feedback + +This should be okay as I dont modify those lists much only on a view change more or m running the test suite right now and will commit the change nce it which other protocols did you encounter this problem too +Youve removed the Context from State and now grabs thecontext from the notification registry +fixes the issue +All operations are anyway protected by the global lock +I dont really like the idea of having a clone method that returns a copy of the structure +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Sure that might end up being where things go on some issues but lets not get there within hours of a large hash out +Trunk and branch +Could you provide information which build you are using under Mac OS X you have downladed JBDS and linked it as plugin in existing eclipse instalation or have you download it already bundled with eclipse +BTW Im happy to see you put effort in improving ll try to move to fop this weekend +in which documents have you noticed the format problem? which other documents do you suspect are affected + ran ok with the patch +Now that I thought about it shouldnt it be fixed upstream via introduction of something like OOZIETMP location and a proper handling of its cleanupcreation phase in upstream strartup scripts? +Or should we standardize such stuff inside of itselfIn any case added comments shall help greatly +For the time being we can live with the memory management mechanisms we have +Reason x +Otherwise +Thanks Maas Vdberg +Overall they look good here is some feedback patch in the code conventions are not followed no space after comma not sure why there is a need to patch +I know but as was said earlierFrom my point of view main thing here is that violation was detected in header of a file which typically contains licenseAnd thus as a we can simply ignore +The point of switching to would be the assumption that code is more stable and performant +No objections here +I suspect deploying multiple distributable webapps in the same AS and then independently hot redeploying is a bit of an edge case more common are just a single webapp or multiple webapps packaged in an ear or full server restarts when applications are upgraded +It can not be included in because that branch is closed except for bug fixes +Check for an empty file list after searching for +transforms every local variable into a field of the implementation +trunk all resolved tickets from or older +Thanks Ning +bulk move of all unresolved issues from to +Carlo wants current patch for integration reworked to use a installer freezing the installer system itself +Anything show up in the Error Log after you try drop the message onto an endpoint +bq +Hang on +Most of the time it seems people can just use different tables rather than do complete isolation +Ah right bummer indeed. +NadyaThanks for the reviewI might have overlooked Debugging VM JIT and Internal profiler files +Fixed a sealing violation which surfaced in when run against insane builds subversion revision +Looks like license header is missing on adding license header to +Im saying that if you have a JAR which contains logj and then you try to get around its access restrictions by putting a class named in the JAR it wont work +Not working on this now +i am using openjdk any known problems with memory meter +Code cleanup needs to be finished +the parser thought it was a URI and treated c as the protocolscheme of the URI +That being said we prioritize company work ahead of outside work in our work queueFair enough this is not an emergency +any update about this issue? It is quite annoying since we cant use the with the embed mes maven +Patch v adds the first Path that triggers in exception message so that I can trace what operation that file goes through in the test output +Patch rebased to trunk +Perhaps dependency got introduced recentlyAnyways installing openssl package on the ubuntu build machine fixed the failure with the patchThanks Ranjit +The same problem exists for the linker options in srcrunConfigure and in samplesrunConfigure +I do add more tests for bulk insert +if you said conf was additive to conf then you would get both +Im waiting for your responses +Thanks Vamsi! +Since test code has since changed I would expect many merge conflicts +thanks for the patch Shaun +Regarding the filenameFilter we need to determine whats the expected behavior is whether its for inclusion or exclusion +This fix is no good for a JMS provider because it uses JMS methods only available since JMS +Same thing happens in Chrome also on OS X +This patch turns into and so propagates upstream +I will take a stab at this +As this is only a its not really that important +I remove the assignment to me because I cannot go on before is fixed +Means if this is a issue it is in both engines or i can only reproduce the issue with a valid have done the same request on a local page and the request header looks as expected cannot see a problem reopen the ticket you have to give me more details what railo is doing wrong or a example including the server part of this testcase +I dont see any reason to use schemaLocation specified on the xsimport +Is it possible to have httpclient not follow a redirect when the redirect is the same as the current URL? This would avoid the problem + delimiited set of bundle names is great +I updated the plugin with your fix +I have not run in pseudo for many months but I will be going back to my home rig very soon as school begins and will try it again +Apologies for the delay +will probably have to set it to something but probably can be large +Thoughts +I agree that its an improvement over the reverted special casing logic +Please let me know if you were able to produce some positive result with gathered writes +Fixed wiki markup Manfred could you please confirm that this displays correctly for you +Attached source code reproducing the problem +I am against applying the patch +please verify branch +Moving these to please move any bugs back that will be worked on before the November release of. +Good to hear! So will there be a new release of any time soon +Id suggest disallowing the removal of the nic if its being used byload balancer etc rather than just blanket removing everything thatmay be using itOne NIC per network for a VM is fine its only currently allowed toadd more than one on the same network for consistency because deployallowed it +for the review NavisPatch committed to trunk. +This is the first version of this patch +patches for the newly implemented methods to find a node in which is greaterless than the given revision . +On closer inspection this issue actually extends to cover two separate tests from the same test suite these are test test +Just wanted to point you at some utility we have already that you might not have known about and that might have helped you composing your addition +Please delete Spring IDE from your Eclipse installation and reinstall +closed. +For the record the thinking behind the varied sync marker is that it makes collisions less likely +Committed to trunk and x +Attaching looks great to me +hi karlIts wrong anywayIn that code you append an url to an other url +I will definately test the same with jBPM and get back to you +this is caused by the username is a registed ingres sql function closing stale resolved issues +Tim sorry for late response looks good thank you +Ive also added which looks for a schema in the metadata and a test for it +We are about to start a project under the name Grumpy which will do static checking of Groovy code +StefanIs this up on your Git repo? Git really isnt that hard Ill be glad to bring it down and see it in real life +In case theres a problem please the issue +Ravi the patch looks goodOne nit do the conversion from to in not in +No one can reproducibly run your benchmark +Ill mark this as resolved if theres no more changes expected +It would be better if only one dialog appeared +Yes indeed think default value for unions field is not valid and should be compile error or warning +Max could you please look at the patch +there shouldnt be any diff in this area +committed to trunk and branch added new future versions in JIRA +Works correctly probably after changes made under +Please report them under new a JIRA +Passing information out of methods via exceptions is always a bit weirdMaybe in the general case but throwing an access denied exception if a security policy fails check seems natural common and unsurprising to mebq +In reply to comment I tried the patch +I am an advocate for making methods using positive language containsAny vs containsNone etc +delete will actually delete on JVM exit right +but it seems the regex used now is a bit floppy too +apply to trunkmodules +Im fine with breaking my existing code if it means this finally gets fixed one way or the other +No one ever this issue so ignoring +Has this truly been confirmed as fixed? Its happening to me with the Samsung Galaxy S as well for Cordova +ok so I can resolve my own problems +heres another technical noteandbq +Incremental builds arestill significantly faster than full builds so I think its importantto have that working reliably +All tests of our test suites pass +WSDL comes Wrapper element is added that will wrap both WOM and the WSDLJ definition so that it willbe available for each and every this the the WSDLJ definition can be taken from this and can be altered with the required binbing and adding the correct port and can be serialised to show the correct WSDL version +Lets say that reusing the timeout defined on Sonar server side should work in most cases +Hudson didnt catch it I guess because of the multiple groups requirement as Nicholas mentioned +java files and the diffs +it make makes sense to do this via a confgurable timeout on the blocktrylock on the transport mutex that is held till failover reconnect ocurrs +Older versions were added manually in the central repository following previous jira issues created by me +yes please create the like theres been a bit of a the podlings name is RCF not RFC. +Thanks Jacques +Now the order may change and this was not taken care ofWe should file a new jira for fixing this +Ive actually been needing this too and was looking at implementing it my selfHowever in is missingOverride in doesnt really override anything it implements it +Youre right this would be less polluting the pom +I made fix in so it can properly read CFF data from font +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I dont think this is necessarily to do with two different points I have issues with only a single point that is miles away not showing up until a radius of at least miles is double checked and it looks like our issue is not solved +Users are managed by the Mailbox implementation +When can be which fixes the inconsistent names for the meta fields +Ive started to discuss +resetting severity from major to normal pending further P open bugs to P pending further was generated with trunk but the output is the same as in zooming in one can see the borders have the same width even though that does not seem to be the case at zoom. +Updated patch with some additionyl codestyling and renaming of tables +for +bq +Spec change Committed revision +Should account be changed to domain there or is this sentence talking about a different feature +Please disregard the attachment it was meant for another bug and hasbeen correctly attached there +arrh I miss the readObject codes +How are you running FOP? From the command line or an embedded Application? If you are running from an embedded Application can you try running FOP from the command line to see if the error persists? If your can be processed from the command line then can you check that your embedded Application can find on the classpath +Can you pl link this with the one in the common project because that JIRA has the performance numbers +Looks good +Default behavior seems better or same for bothidentity and sequencesThe preallocation of single user sequence is the worst case I can think of +the file +This port is required to merge and test +Im not sure +Attaching a new patch incorporating Devarajs offline comments +Good comment yeahMy testing didnt show any compelling reason to increase this value though but at least this way its not hardcoded +This code was not changed for very long time but good hint i will change it to work with homeDir only as well +Awesome! Thanks for fixing this +to fully remove all warnings on Visual has to be applied Konrad! +I have added a thread dump containing the ejb threads which are inside the infinite loop +You can see the diffs by looking under Activity and clicking on the Source tab for +I will investigate the differences from my side +Thanks +Ok i have the line fix +Currently only one in one physical plan so if the execution of a physical plan fails remove the associated output fileHere is the plan +CarlDid tests pass +It is just always reporting coprocessors in +If I launch from Eclipse environment it pass the test + +Merged into the master and branches. +Why is the initialization done in prepare? Just to be a little more lazy than in init +given that I spend hours on this a week +Committed at subversion revision Touches the following filesM javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslang +Actually the above seem to be the same from our Brew build of geronimo the very same build accordingly to David +Attached patch for marshaling the object into output stored in Result interface +Build on Mac should not include linux jsvc binary because the native library and jsvc would not work +Logging a new bug for the test harness adjustment and when that is done Ill see if the test needs further adjustments +In fact the generated EAR is empty +fix the problem that Kan Ogawa mentioned above thanks Kan Ogawa +Commited slightly modified patch in r +I had a stab at recreating the spirit of the patch by short circuiting early but it didnt improve speed in my tests +If they are different then you get the split wsdl things +Actually this isssue might have been fixed +Can this issue be closed? Its really just an action we choose to take +The next version of JAMES will include Derby so I believe that you can count on the presence of JDBC in all future versions + +I presume the master rewrite will incorporate something similar but implemented with a different strategy +Doh! It most likely does +As already mentioned parts of the required information has been added as part of the work on +I found the last of the thread contention issues in the class now all tests pass even on windows. +Can you use the link Attach file in the left pane +committed +How will switching to help +Passed all testcase in hadoop +First page still comes out with a heading but nothing else. +jars in lib to make the attachment you trying to say that doesnt attach a date time converter to input fields are bound to Date properties on backing beans +I am way too tired +I just ran into an RSS problem that may be related +committed +maybe it was only to not forget to close them? +Ill any patch that doesnt have it +Maybe crash happened very late in the shutdown stage when properties are no longer available +Patch looks good +Ive applied the v patches and have successfully offline scrubbed reinserted a number of nodes in my ring without further occurrence of the previous issues + +Thanks Anoop +Having two s in same repo violates client server viewpoint. +Manually applied patch to trunk in SVN rev +Thanks but I was more worried about why my initial moderate mail had no effect despite being accepted by the ASFs MX +I think Im using it in that fst suggester patch actually +All the tests pass +Hopefully this wont break anyones build scripts +The statement is not used at all during the dependency resolution by Felix and the issue is related to the fact that Karaf exports all the JRE packages withversion +Linux version + +Reworked version of the patch based on Stacks comments + +Thanks +Resolved in EAP trunk r and EWP trunk r +this patch contains class for Identity project and for Forums to right release versions +I have to correct myselfThe proper options line is tag generatorClasstGenerator Classinstead of tag generatorClasstGenerator ClassPutting the quotes just around the tag head causes javadoc to abort with Illegal package name Generator Class +Dimsshall we have this bug fixed for the release? was fixed by Glen in r. +The test failure is unrelated +I ran a few other experiments and they were fine too +I just checked in some code that guards against a potential null pointer exception +Verified reproduction on trunk ORDER BY seems to be the issue here Scrollable updatable result sets are not allowed with ORDER BY sothe query gets downgraded to being read only this generates a warning that probably breaks in the protocol somehow +fixed concurrency issue with creating a new entity +CDI integration with has been fully rewritten +Ive just discovered that Emmanuels workaround doesnt work either because that makes always exit with error code even if the build succeedsAlso note that my fix needs to be added to all of the other validation points in this file including the JAVAHOME and MHOME verification checks +youd have to figure out a way to ensure th sdk is on your classpath at. +no worries done +You guys have iet right that you cant know the next row until youare already past the first keyThe problem is that we would have to hold on to the snapshot and kvsetuntil the end of the scanner when it is closed +I must have been drunk when I wrote that line of you for the very detailed bug analysis and fixWill be released in Version. +One of them should contain a class that is also available in the jdk but the jdk class is required to be loaded and the jar class is loaded instead because of our special class loader setup +Putting the weights in a file and making the argument be the file name would make changing them easier +for the same we can create one ftp account for you guys to download and test from their +Sorry Shaun would you be able to clarify which book this is in? There are quite a few and it will save time if you can specify +json output format seems to be ok with this patch +Arun this feature was added in H and was not available in H +incubator grep sschaffsschaffert +Attachment has been added with description works for has been added with description leads to an has been added with description use this in FOP and probably earlier transition resolvedfixed bugs to closedfixed +Could youplease try again with nightly build or later of andreopen the bug report if its still a problem for you +should this issue be marked fixed? Or is there more work? +Try moving the installation to E or even Ejboss for example +Process with failure of timers AKA Jobs +Thanks Nicholas +Scott importing a single component would be useful but that doesnt help the bugzilla side of things were that component would still existOlegs work on the bugzilla side to move a component up to a product is necessary anyway and then the existing JIRA import should work +Committed with the whitespace sorted out +Closing this as invalid +Not an issue now as were using Jive forums +If you find further issues please create a new jira and link it to this one. +Thanks for reviews Todd. +Does it make sense to call this add or something like thatLets call it add since it also adds blocks to the block map +I did update but as you know files are replaced from eap distribution +Verified this issue with latest build and found that delete cleans the lb monitors automatically +Closing issues which have been resolved for some time +Comment crossing againStack and I had been discussing some kind of escaping because the Meta comparator is the only reason why we pass around comparators everywhereWed have to guarantee that the escaped byte sequences will still sorts correctly +Extraneous comments coz of TODO fail the application on the failed transition keep the master in sync with the state machine nm to avoid confusion stopContainer Container containerId hasnt been neither started nor scheduled to start instead can be Container containerId is neither started nor scheduled to start In all of you shouldnt explicitly refer to runtime exception +Patch for this issue is included in related Patch Collection. +Implemented support for Range GET requests in the in Rev +I can only supply code patches for the series because CXF presently wont compile with WSSJ +I would be happy to make these pretty +Looks good to me +See attached patch +He is working hard its am in California true but he is in germany sitting in Stuttgart going to hit the Weihnachtsmarkt soon lets see how the merge goes after several glasses of Gluehwein +Btw I was also wondering if s is a good value or not +need new unit tests +Cleared references are cleared when a future registration or notification occurs as a Thread for this purpose would be overkill +This really doesnt need to be part of a release. +Added a test that should verify this +Some of these systems have very large data volumes and want to minimise indexing time and disk size +Simply copy the labels to and change there until fix is released +As discussed on the user list on Feb +this was done long ago but the jira remained open. +One could set up a filter that restricts by author discarding all comments from Bob +Each method has its corresponding generated which seems to be what is causing the +The patch for jdktools module +Module the authorization parts of the above comments the weaknesses can be seen back to Derby. +Will commit in a bit +I think repeating implements is a good solution +jsfplugins +I found the problem I need to do a ll commit it as soon as +I was thinking simply to return a null for the counters which is what the client API returns for that same situation +Committed to and trunk +I applied the patch in revision as proposed except the changes to the Encoding class which has nothing to do with the LZW filter +Fix for the issue +You must refer to the documentation of your vendors JSSE implementation +Maybe the sitejar goal needs a new parameter absolutize so that it is possible to use relative links in all files and make the sitejar goal update all references to files outside the jar currently being built to absolute paths +Romain please let us know if this addresses your need +Jakob can you please provide a test project that would allow to reproduce this and also clarify what you do in order to run your web app +I agree that this patch should not be muddied by an overhaul of our Clob implementation and that you should log a separate JIRA to track the Clob defects identified by this discussion +If this fixes your problem can you please close this JIRA +If the directory has no Jars then it is put in the class path directly +It is utterly pointless to create issues without submitting simple easily runnable testcases so we can reproduce +Windows Firewall is to be blamed for it. +Sure I will have time tomorrow then it can go to tried without success to reproduce this possible problem with nested forms then I just realise that only the root form write the hidden field on response so the committed fix should be just fine +I think it might be problematic for some languages though especially Dutch +The path should take care of webkit sending too many cookies to Jetty +this could be client caching try refreshing browser and doing it again or at least closing all tabs and then opening it again +can you try again with a fresh checkout please tested and was able to commit to that repo so setting and perms are like that did it +Latest patch in synchrony with PIG trunk +Also will get a later version of too +Easy to understand +Verified in CP CR. +Nik any updates to this +Yeah it used to be that way +Here is a more complete patch +Providing a mechanism to close an instance I think would resolve both this issue as well as this issue +PhilippeYour comments and patch are very much welcome theyve helped me greatly along the way towards making a default implementation +as far as the encoding itself is available in the ICU libraryICUs API requires to use and for inputoutput +Why is this the case? It seems unintuitive to like requiring a Resource annotation for this special case is inconsistent with the expectations of the user +bulk move of all unresolved issues from to +Committed to trunk +Thanks a lot! +The object can be generated for serialization from the data stored in the. + +And still the seek op itself slightly expensive than get op +I also tried escaping the backslash and just a and were printed +What versions of Commons DBCP and Pool are you using and how are your counting connections? Is DBCP itself reporting the excessive number of connectionsAny updates in this issue +The test case currently shows that is not functioning +I would bet that most folks would prefer to use the POI exporter over the basic one +It has been very useful for understanding how to implement alternative Stores incrementally. +The implementation borrows some code from the gbean reference classes +What about building a new or using the decorator pattern to implement throttling +Thanks looks like list items dont have modify dates +There are many cases where it is necessary to iterate over the entries withinthe index directory or a segment subdirectory +Verified in EAP ER +We can just remove itLooks like metrics is in +Since this will be fixed by and Yip is assigned to that one +bulk defer minor issues to +Honestly I must say I also never saw constants being defined in this way before +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Yes I agree thats a good focus +Open synchronized window with skin Galaxy S III +Well to be honest when I wrote it buildnumber didnt have a timestamp +The existence of two logical connection holders for a thread does not matter according to our understanding since the sequence of two logical close calls should still lead to proper state in the endWe couldnt reproduce the use of two physical connections in such a scenario so Im closing this issue once again +They are loaded even when the plugin isnt loaded. +well spotted +Thanks Paul +This issue has not been resolved as indicated in the previous logI have a form with most fields having validaterequired and tdisabledread +Attaching latest code +At this point in the release cycle and we can not add feature requests +Python patch looks +Here is what the looks like in notepad on right out of the said hed take a look at this so passing along +Would be a better name for this new config +It seems this is currently not the case +Andre I tried it with a new connection +Closing resolved issues for already released versions +Committed to and trunk +Removed some TODO items from related to setting storageID +Thanks. +I had some comments in the reviewboard +Webhdfs will fetch a new delegation token if the renewer has failed to renewreacquire the token +Potentially we can make support independent jcrpath predicates +I would be inclined to close as WONTFIX +Profiles and directory listing as referenced in customer email +we can just throw directly instead +Does this need to be mentioned in the Release Notes for EAP considering it has a helpdesk ticket reference +Im not sure what that evens means to be a blocker against a released version from a community perspective or what further we would do from the communitymainline +I think the easiest way to do this would be to not strip off the shared from the filename +Implemented +Diff including all applied and the site has been updated +Ill see if I can knock something together tomorrow morning +Problem was backup was trying to a read a pages from a container before the firstallocation page for it is created with a invalid page number +It would be nice to have the fix in the next EAP CR +Not replicable on Fedora bit KDE +Afair is now already bundles +The way its named now its inaccessible via maven or maven aware tools so why bother putting it into the maven repo +I am not able to run this test either before or after the patch +The user scenario where both and connections are used in the same application would be quite rare I thinkSomeone can reopen if they decide to fix. +I cannot reproduce this +Perhaps most people just use jar dependencies +I have hit this too as will anyone churning out a baisc ROO security app and then switching to framework course it also breaks on vFabric tc Server which incoporates tomcat +My vote would be to leave it around until we are done done and staging is removed + results JDBC calls TCK without query results JDBC callsTCK with changed results remote server +This gives me better chances for bug tracking! +Devaraj K verified and Eric Y reviewed +See +Thank you for this error is logged if the default user group does not exist +I applied the patch and it builds fine +It looks good +Sorry I submitted a wrong named patch +This ticket should probably be titled allow sorting by index value since that is not yet possible and the paging concerns are not valid until it is implemented +It is possible that may be there is some bug in optimizer which thinks that the index used for can be used to avoid the sorting required for +Closed after release. +Fixed for the Clover plugin and lots of others +First of all please update your JDK yours seems to be outdated + Fixed some Paul there are conflicts to the latest trunk +There are no of platforms in the product that I can think of +Yeah nice catch +Attaching first attempt at this work +unread is there to be able to push back the last read line nothing more than this +WillemYeah that was also my initial thought but wanted to know anyway +OK heres an updated patch that propagates NOW for all shard requests +I used the latest hibernate core jar I used the data direct dirver to do my testThe collections never get loadedIf I run the same test against or the collections get loaded +Hi Hadrian thanks for giving this prompt attentionI think it is better to censor the values of the secret fields rather than strip them completely +BTW I have seen numerous misunderstandings with our current boilerplate implementation +I thought that was a regular collection mapping and not inverse +Awesome thanks for pointing that out +Its fixed in the latest build however the older builds are still present in the repository too +The first one is a unit test showing the failure the second one is the fix +As the patch is thats not really possible This makes a lot of changes in core code for something that is a +Thank you for the steps to reproduce the issue it is very helpful +But we want to use the original block object when passing on the request to the next node in the pipeline +Brew build fixes this missing jars +Good suggestion +It is not okay with getstart Japanese characters diff +from the point of view things are simply since the proper wiring will Im find with any decision if that means packages will not be splittwo jars which both include classes from the same package +I am sure there are other traces of pig in there which will remind of pig in the harness +I came across this issue looking for a solution for my current problem with my case Ive been trying to bind values to a domain object +Lowering to critical since theres a workaround +The idea is that the Mylyn Build becomes installable from Central similar to and other on its dependencies it might be included together with the existing mylyn functionallity but lets assume it would be would go to Central click the software tab and search for keywords like jenkins hudson or build and a plugin named Mylyn Build would be shown and then use the normal installation steps to get it it is installed there is a Builds view where users can setup jenkins instances such as the one they get when setting up a jenkins server on openshift +Let me go it +Extraneous wakeup in remove which inverts the locking order +The most simple solution is to set Enable scaling checked by default +patch looks good +Whats in +No objections to close. + are always welcomeWe prefer patch files attached in the ticket though +Bulk update fixversion to +This was probably related to an earlier issue with mule where the DNS had been configured incorrectly after the change in imagine got dropped from the ve resynced the list and you should be fine now +If there are no objections Ill commit this in the next hrs +Test test attempts to creates a blocking condition and then looks for TPEBLOCK error but no matter how big I set the send data I have failed to produced a blocking condition on a windows socket +I did have a few wierd instances where it returned empty jars when I didnt think it should +id like to commit this if were missing features in the new ui ill add them +I could externalize that to get a out of something but is it gonna be easier or more complex for you than a Classloader implementation that does the trick +sorry for the inconvenience +Im just merging the code +Its a warning not a error +Sorry missed that +Its pretty obvious that we have tons of integration tests in addition to our unit tests though +Tasks is probably a throwaway but the rest sound like theyre right on the money +Correct fix version maybe doesnt make sense to set this for Cannot reproduce but I guess it shows what version I believe this works for. +And it will be very complete tests +But thats a drag to have to do by hand when Hibernate could be doing it very straightforwardly +So I close this issue. +Our query syntax is such that a user can combine clauses against named fields with clauses with no field specified +But indeed you can use validate update function to forbid doc insertion +Nag me if you see no activity! +Taking the actual issue a little further youre overriding the field types of some of the methods with the PIXXX variant and that is the actual issue all previous samples did nothing of the sort +I attached a patch setting the chunk at MB since Im also affected by this issueThanks +ACKd since we like watching the silly british kids +Is this going to happen +IT through +ppnyc Ive just pushed a new snapshot build which contains what I believe will fix your problem +Adding release note warning about increased exposure to +New patch with Brackets added +I think I can create a patch to customize some style properties after a template is loadedThe style properties which need to be customized includefolanguagefocountrystylestylestylestyleBesides these properties are there any other properties that need to be customized after a empty document template is loadedAssign to myself +classpath Usersdmitriysrcbin +I think its ready +Until somebody confirms this as a real issue it is minor +Help in resolution is appreciated. +has to fix the streaming job +Can this be marked resolved +Without more info or a reproducible case I have run out of ideas on this one +In summary async at user level can not improve latency +QMF V management messages were being batched by accumulating up to a certain total size of data +The Create Configuration button does not exist any more +uses tablesystemctags to get cluster names +Thanks Rohini for reviewing it! +Thanks Ive incorporated all your comments except one mentioned belowbq + +They have been removed as they were an artifact of +The testing and discussion seem inconclusive on this and it hasnt been followed up on in a good amount of time so I am going to mark it as wont fix. +Correct it is indeed stil hardcoded and not overruled by a value from a name was used instead of the proposed file is missing thats the reason why the email is always jbpmjbpmnoreplyThe file can be provided to illustrate this i dont get your far as I remember it should be possible for people to specify the property in their file +But any query that explodes when you change the index is arguably brokenObjections to this going into +Reminds me of a similar issue in to master. +I have two colleagues with the same problem +in answer to your points No overlays classes end up in classes as per the currently behaviour +After some investigation I think that the problem relates to the class loader hierarchy +Yep in many cases we do a mock for the master the regionservers but not for +Adding Ignore to +solutions +I support this move as well as long as the packaging remains such that the out of the box setup for both languages continues to work without issueWhile were at it we can fix the location of the default container configuration as well +Unwrapped exceptions from remote exception has same message as instead of picking only the first line from the server stack trace +Patch checked in +testAutomaticallynew persistent class jpa name +well same algo isnt pkbdf in too +Thanks Mike! +I dont think we want to update the RPC in mrv just to make the test pass +As my opinion it caused by the mavenjunit framework has load the implementation of before the test case running so it will call ahead of registering all the timeSo the test case can not reproduce in mavenjunit +another possibility was to drop the and check accessibility onin other words let the operate on the and only performcheck for accessibility once the TreeProperty is really accessedIMO we currently perform far to much accessibility checks while accessing items in the hierarchy + +Ill put it back inOut of bounds crashes the job so I dont think the speed of trycatch matters +All that matters is the with the stack trace +Issues regarding UI deadlock and extra delays seem to be fixed +after the discussion on dev list I agree with you that we should fix it in drlvm +OK thanks +Thanks for the review Konstantin I will commit this patch +I dont think we need to add serial to classes its absence shouldnt cause javac warnings +Patch applied with svn revision +Attached patch applied +Why do we need to do this when folks can use the test jar in their projects? It is surely meant for test purposes only right +Looking forward to them +We do not remove outstanding requests at the end of a search +The is suffering from the same +Nothing else jumps out +do we have any free bits in the token header? Can we just use one as a the sign bit? then we could just use a simple vlong. + +Needs to be mentioned in any upgradeCan you please tell us what has changed as we are preparing the release documents now? like nothing needs to be documented as the issue was later fixed in a separate issue +Time windowing for shutdown should be addressed byIt might be reasonable to build a command line tool or script that stops all in a given subdirectory +were backported to as well adding the latter fix version. +Patch applied at revision . +After the latest patch fixing removing inappropriate dynamicBase properties attaching fixed example dynamic field responses +Anyway now getting the same deps as Committed the revised in revision to test the patch for +Some day I will complete it and then the Forrest Eclipse plugin will be independantId recomend that the project should be called both and projects into Eclipse +It could lead to an application working entirely correctly with a native id generator and then developing a bug when switched to a dialect using a native id generatorI suggest that the correct behaviour for replicate is to always try to insert the given identifier and if the database is unable to do that throw an exception +Thanks Benjamin +We already have lots of its time to clean up the duplicate codes +Splitting into sub tasks so its easier to review smaller patches must require all subtasks for it to work +Compiled and tested +I just committed this to trunk and branch +Rewriting it as a unit test would be silly. +Ok nevermind +This is just a misconfigurationIn case Im in mistake please reopen +Performance is not the motivation for this jirabq +Thanks Stack Ted and Matteo for reviews +Any Comments +I made the simple fixIt will be in the temp build +verified by Vasily +What do you mean by clicking times? Is that a double click or confirming confirming again +Can you check a bit further what actually happens in your scenario in that respect? for the time being as it is likely to be a problem +That did the trick +I have similar concern as Lars has +Im also getting this error using the Data Management release to import the attached into Continuum +You dont have to write the tests but you might ask for some advice or help from someone else on it before resolving the issue +Attached patch includes the drop and the test output +Separate state store clean up thread +Can you give an example of the binding configuration you used when you got this error +thanks +I just need an ack from a drools despot eg +bq +ll look into this issue +Ive tested the nightly build this morningOn Linux with Oracle Java and Tomcat it worksThanks for the for confirming +Will investigate the issues on Solarisx as time permits +Ive tried ti reproduce several times for a different builds +Ill keep that in mind for the future +Related to +I dont really understand why do you think this is a major bug in Spring Security? There doesnt seem to be any evidence for that here +this is really very time consuming and makes Maven unusable for productive build environments after it fixed in Maven? + +has been marked as a duplicate of this bug +I think the general opinion was that the extra flexibility wasnt necessary +There is no need to create new tab by name Validate leaves +may be better at executing the Ruby code in srcjruby than MRI is but I think the code in is slightly inefficient + +Looks good to me +Not sure of the implications of this as Ted says +The update to JSF seems to have fixed this issue. +Fixed in close of all resolved issues in preparation for release. +Im sorry that I had other priorities at the time +The only difference in these patches is that the default for and is now Ill probably add one more patch later to implement the change to merge key caches +NET binding using the +The current code resolve at the Definition level +So Im confused +Not sure how a single error could cascade like that after +Currently I cant use Solr because I have indexes with field and for each field a is assigned +Fixed in both branches. +This patch includes parsers beans checkers and test cases for DIT content rule descriptions DIT structure rule descriptions and name form descriptions +When embedded is exposed remotely that means you are trying to redoreplicate the functionality of the server that Teiid already provides through AS in a different host environment +I am thinking of what kind of error message that user is expecting? I would prefer to have a generic error message for this sort of errorThis is because it uses the invalid wsdl op it fails to create the Ode thus return the null Ode MEX and then cause the following NPE. +Tested environmentsJDK AS Server defaultI suggest to deal with this issue once jBPM BPEL will be tested on AS series too +Your particular issue is fixed +Not sure why this is categorized as a doc issue +Can you please verify? Thanks +Thanks for the tip anyway +Meant to set status to resolved not closed doh +Reopened because of respin +how to get from String to URLCommitted to trunk with revision Hopefully the last commit on this issue +These were probably there before the version to trunk and branch +Integrated into Mahout Trunk marking this as Resolved. +Thanks Konrad! of the Windows specific files got missed in this sweep +This patch add a filter to the obtained after the call to the make function +sbt package seems to default to scala +Pull from this morning assuming my updatedbuilddeploy dance is working correctly +Reopening because of integration. +I dont think we should remove that target unless we are sure that its useless to the eclipse communityWe agreed that was the last release which would carry eclipse plugins +and contribhdfsproxy imports contrib committed both the patches +I have submitted a request to have this taken care thanks! +Thanks +Resubmitting patch after +This is unrelated to the code and the test passes on my local box +If we only change here then the reading part might not work +Hi HansWe have completed the functionality of Training Calendar in Employee profile tab and will soon upload the updated patch +Any idea when your pull might get accepted? This happens at the discretion of AS committers +Possibly some corrections will be required due to different environment +Tomcat or Jetty +Fair enoughSadly the fix is deep so its not gonna make it into +Thanks Helen! +failed on +Seems to have been fixed. +Patch was attached +Konstantin created Test test fails +I agree +to applying on +Since the change to use block pool ID is not complete there are a lot of TODOFEDERATION comments identifying further changes that are requiredIn tests and there were instances where protocol helper methods were not used to send requests +As far as vs +Ill try it tomorrow +Dont worry about fixing this for now +Once my patch for is in Ill bang this out as well +Havent tested as the latest svn build is broken +Thanks Yongqiang +That could make maintenance of the repository and right management easier +In confcollectors file you need to put httplocalhost to send data properly to the right port +Yes I pulled and rebuilt and get the same behavior +The original test with fixed expected results and new test messages to prove different scenarios and prove the previous inconsistency +Hopefully this will help nail down why some of the requests are taking so long +Patch isThanks for the patch applied to SECURITY module tests at repo revision Please check that the patch was applied as you expected +Not sure this will happen for so move it +Ok Ill close this duplicate issue now +Output should bebut real output is +thRuth +Closing resolved issues whose fix targets have already been released. +Patch checked in +Please check and resolve the issueRegardsSimranjeet check the revision i mentioned it is fixed there +My understanding is that an API can have different implementations and that an API does NOT know who is implementing it +Both trunk and its now pass when their location contains a space +This is what Jeff found out Found that EAP and used different version yesterday the EAP is using while the AS uses the I think we should go with +Found out the problem is at the class where it can only handle one jdo configuration ever +Wrote unit test case to demonstrate the defect and test surrounding functionality +Thanks. +hdfsorgapachehadoophdfsservernamenodemetricsNew file added for unit test + +Is this ready to be committed +Nice that will do the job +Can you create a test that shows the bug +Im not quite sure what resolution to use +com was taken +If we wanted to distribute compatible files I say lets write a script as doing it manually is brittle +Question if the trackingId is in the token why is it extracted and stored in the? Currently that object tracks info about token idents that isnt in the token ident itself +Again the leak only occurs using the Debug Multithreaded DLL runtimeThe attachment is a zip of the MS VC project source and sample XMLDTD +If the child RS can be removed always assigns the more specific partitioning columns to the parent RS +Ill leave the code as is for now +Closing old issues +Would making the Renewer non static work +Postponed +add a test case to test sorted +This was just a test exercise in seeing if we want to have a expansion or a modal dialog am going to go with an expansion and then have the buttons act as they do in the mockup that you edited +Script to open a region +Closed issues related to Roller release. +The final keyword here has the similar meaning as it does in Java cannot be overridden +This works now +Of you go for the verification +Thank Tomasz for your great work and thank Suresh and Nicholas for your review +Sorry for the delay Lionel been traveling to clients and didnt get much time! I committed the change you requested +I see one has to be able to reproduce a local context thats representative of the global map otherwise the data resulting from the transformation might be differentshifted +I gave my technical rationale for this +Burr ok so this is a pure runtime issue can enable zipping of deployments on the server by double clicking it in the server view and toggle enable use compressed archives me know if that works for you +I believe that this patch should get rid of all the tabs +Should I create it +Ok so accumulates all the Views it can find including but then ignores it since inherits content type texthtml from and that doesnt match for a request with Accept applicationjson +ears at least on each JEE application +Thanks! +This is done. +Heres a patch showing the sort of refactoring I was talking aboutUnfortunately the surgery was more involved than I expected I had forgotten that was an inner class which had to change +and we already had just the same kind of issue within FOP seeanyway i can make a workaround for this in my own logic for time being and wait for new stable FOP release +OK seems not quite done yet +Hi Suresh +If you only use it for a counter then you should use instead as thats from the JDK out of the box +I also declared it synchronzied to be sure we dont see a here +what is local +DNS failures exceptions on every single part of an IO operation are all candidates +thanx! +So if this patch merits a commit I hope the committer will also remove these files from the repo too +You have a patch? Thanks. +the usertransaction filter we we jackrabbit works perfectly with xa under websphere +If I include the CXF jars in my project isnt the Jboss classloader supposed to load my classes first? I will look at that and also try my project in some other containers and see if the behavior is differentI solved my problem another way that I am very happy withI used your Maven plugins to generate WSDL and JS client files into my project resources folder and then those files can be fetched using a js or wsdl URL by the client +I think the doc conversion went very well and can be closed +Exporter methods represent a logical end and dont need to return the archive in question +When were doing analysis we dont know which of those different meanings the user intended so were providing them as all as options +I believe this approach is comparatively rare +Applied patch in revision +Without any resultAnd I duplicated the issue because I wasnt sure that someone reads comments on closed issues +Update to add cglib to pom dependencies +Just committed the patch +It wasnt there just because it hadnt been needed before. +The only file retrieved from svn is which is not installed by JBM +The use case is IMO rather theoretical and an application could still handle the exception caused by a concurrent importI think the important part of this issue is to make sure the repository keeps consistent +Im trying to reproduce this on quickstart but can not yet +Patch replacing the link target +Howard can you please take a look at this as well +See the discussion link above +I will prepare an addendum for the security branch +thanks Joana +If your namenode fails and you bring it back up on another node then Accumulo will not work +If there is no interest in the proposal I can close this. +Ahh +I see it on both and +One way to fix this would be to add the list of parameter types of the client method when writing the invocation since those should match the server implementation but that would make each RPC call bigger +I think we can move the dirs and restructure it appropriately to reflect Maven dir layoutAs to the question about work with srcidl and well move them too when the restructure works take off +Apologies for the duplication +Im not sure I understand the problem do you wish to save the entry as a Draft and then have it flipped to Public at the appropriate timeThe intent is more publish to future or publish to past where by you control which date the entry is associated with +of course +Patch needs to be updated to trunk. +Otherwise it remains at the defaultAlso added MAJORCOMPACTIONPERIOD to +Patch that adds a type for T such that we can pass the compilers generics checks + the patch looks good +Bulk close of resolved wont fix cant reproduce etc +How come this is marked as resolved? Theres still documentation on how to get flume NG to work with s anywhere and its been months since this was initially resolved. +If a test is using say as the key its still possible to use the set methods +Now the sentence is short and easy to parse +Committed the revert in r. +progress on this one Dirk +Thanks Carl I will take a m getting conflicts attempting to apply on latest trunk +ThanksIll fix momentarily +Verified in JBDS Linux JDK +Modified so that only a single dialog is displayed +ok thanks +Commit changes to trunk At revision and snapshot At revision +Other libraries likely provide this support better than we ever could for lucene i think the focus shouldnt have anything to do with data redundancy in particular but just making the storage API in general so that everyones needs are met not just your log file needsTotally agree +Yes Aaron I noticed it and corrected in next commit +Task is in progress donehecked new articles from RF wiki and cookbook +Attached my client that causes the crash +Peter could you create a new patch which only has the changes once? Makes it easier for me to Peter we have added tests for all supported for both input and output +there are N publish request in hub server HUB in queue +I have made the relevant changes to support the new version +No +Can you please have a look atHere the problem is that we need to declare weblogic parser to register prefixes first +interesting which of the files should we listen to when it comes to connection info +Fix use real operand size when spilling regs through stack emulate pushpop instructions for wide types on IA +to automatically verify the validity of the incrementals something more needs to be worked out for example keeping a persistent monotonically increasing sequence number for all the + to qa to stage to new GK release with SLA breach. +Now that work is well underway with AS all previous community releases are +Tried to set everything to but dont help +Here is a small patch that allows reading the compiler plugin configuration from pluginManagement +Patch fix see Bug should never set autoCommit during a global is able to use a non XA with JTA and does setautocommit to false when needed + +Oops original patch didnt even compile +What do I need to do to turn it on? Its been a looong time since I took at look at this +Attaching a patch that provides a generic way of specifying scheduler options +Thanks for you kindly reminding I will update the wiki page today +All you have to do is to set futureOnly property to true +I still can reproduce the issue with broken insertion of qualifiers +Patch as described in original issue comment +I think introducing more complexity by adding weird hierarchies is completely the wrong direction! how is this hierarchy being determined? how should this hierarchy technically work for flat environments like? what about jigsaw? what about servers which unpack and packagesFolks introducing any hierarchy is exactly the same way broken as BDA itself +Im having exactly the same problem right now with Grails and Vaadin and this is weird +Thanks Arpit +resolved in Ive fixed +Attachment has been added with description Patch as described in the bug report +I rebased trunk and created anotherI also broke out the white space fixes into a separate diff +Ive committed this +Even if new replica is assigned to a node that has recently gone down this will be detected during data transfer and a new location will be assignedDont see how it should correlate with the delete policy +I like the orca the best and the orca second +David do you plan to provide a patch for it? We love contributions +Mickael created hudson job it works so I mark this issue as resolved +Actually this is something that has been bugging me as well but I never got around to reportingfixing it +This has been committed. +That seems reasonable to me +page filter to the list +Ive added reference +Now we have it implemented when messages are added to the store so theres no need for this checkThe commit also includes a fix for a cursor problem spotted during the test. +Randomize the initial load of images content +Committed in trunk and flexWARNING Please do not merge revision to flexIll will not merge this to branch as the is protected and this change need to invert th PQ order which would break BW compatibility +The tests pass for me now with the patch +Is that acceptable +Update this is now not happening for me +Final patch for review +dims +patch v named events interfaces more clearly added testcase fixed a bug with unhandled exceptions not being +Pinged Jiri for any additional suggestions again today via +Yeah +Will get this out by was done when the test module was part of in this is a separate project now we can create a tag for this release +commentsmy initial idea was to add tests for the SPI interfaces to the projectwhile it seems feasible to have the tests packed to a separated by defining the goal with the i didnt manage to make usage of that for the tests of a SPI implementation such asunless someone knows how to do that i would therefore suggest to create a new project that similar to the contains below srcmaininstead of srctest and allow to be added as separated dependency to any SPI implementationprojectcomments? objections +what does that exactly means +Please close the issue now +I agree I was talking about where the contents of each module are coming from not necessarily new package namesAre the contents of each module correct based on the current source +DUPLICATE ISSUEThese are not failed on my desktop now +Ive tried to make this patch python compatible but was unable to fully test it during a deployment since paramiko doesnt seem to work correctly with python +Renamed and to and +Again perhaps just a little bit out of scope of the original intent for Thrift +Have a fix in hand for this +The methods attachCopy and attach should be removed as part of this change +That should be all. +The patch for the same +Im glad I was sifting through and found this one +I have verified both on and dev branches and both builds are failing for me. +I came across get +Filed +we just have to set the owner to the outer closure class that would be enough +Manually tested +Just make sure that you use the dialect +The has changed to but not all of the calling classes have changed along with itIve found the problem in Ive had some trouble with subversion so not all files did get commited +fixed the bugs and added some new tests +Patch look good to me +Sometimes +Any input +There must be some difference in configuration for cluster +Marking patch available for Henry so that runs +Please try it out to make sure it is working OK. +When I look at the download page it still contains references to snapshot builds +Successfully tested on SP +Polished a bit +Committed changes to address issue in revision +Updated patch incorporating review feedback also fixed all jsps to encode urls again so that the decoding done by the browser doesnt lead to wrong logFile parameter being passed around +Duplicate of +to exclude UIMA types from and PMD warnings such as or +Convert to to perform comparison +Or is there a way to uncustomize a custom config +Im going to commit this momentarily since the difference between this and the last is just comments +but is there a way to not specify the startend and have that based on the values in the index? start end? In this case it would be nice to specify the gap as round numbers +EAP ER works though +in Karaf is better +I was playing around with other crazy opts passed to the mvndependency task in ANT +If you assign it a Destination you will get the advisory +IIRC you asked if we were to definitely remove from documentation features which had been deprecated but were still supported +Ravi patch looks good except that there is an additional prefix in the path asrc +Removed form name attribute being written +Christian could you please submit your patch for this issue +bulk defer of unresolved bugs to +Thank you for your comment +If so it may be good to create a service in Clerezza that allows to Schedule arbitrary jobs as I dont think it should be the responsibility of or exclusive to CRIS +How badly do you want it +ClausThanks for applying the patch +Thatd be better in more cases +The usual way to achieve this is to configure Springs in however some mechanisms bypass such a listener completely +Patch applied under r + +What Groovys integration mechanisms you are referring +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +The WSDL to xmlbean parsing has been moved to the class +The STS name is also set to a unique one which should not conflict with each other +In the long run it may be better to encode the cluster as part of the groupIdwell in the ideal world all artifacts would have their own original coordinates not the ones derived from unit test for all this? Oh boy +It seems like the and other related format classes need some serious testing done so we may need to just write a full test case to make sure this little fix and others arent introducing new issues +svn commit m agent forwarding The main change in this commit is the switch from mina as the transport mechanism to plain tomcatapr sockets mostly because the mina apr transport does not support the unix domain sockets +Thanks for the patch the package names and tuned the entity class names used by the test case for +Now the warning only appears when the target of the rename is in a groovy file +The test case associated with this tool seems to be failing as well +And each has a full AST below it so a ballpark is around +The test case did not pass on before and after the patch this is only to address the additional error +I like the idea of and I think we should make all the other exceptions extend it to ease error handling on the client but I dont think we should get rid of or they are informative +Reattaching +This seems like the right place to check for it since the compaction down into a single file is what would have caused us to exceed the region max file size +Archive containing a document and all the images used for the to mention that I filled out and sent in a CLA this morning +Thanks Ratha for the contribution +Fixed in revision by keeping count of active readers and clearing the active when all reader and writer locks have been released +Application was created succesfully +Please confirm the current status of this issue and if its incorrect change it as required +So now the question is why my appender didnt get a Layout +After running several tests with I cannot find a memory leak regarding the tests that are failing exclusively in linux +Can you post the command that was run by make check right before the error? Also what version of boost do you have +However we still need to be able to specify host and port in addition to the ID from the command line correct? Or are you leaning towards not being able to specify host and port through the command line tools but only be able to set those programmaticallyOne other possibility is to completely remove host and port from +I cant find the other bloat error +Moves the cli parsing from to +Thanks Tom! I committed this to trunk and. +Can you attach a little example of this such as the template file which refer to the other template +Attachment has been added with description Changes to FOP similar to that of whereby the files are downloaded at compile time +Given these two facts the first execution of the mastercompute runs before preApplication +I will submit another patch to fix zkpythons make file +What kind of route are you trying to open? Does the route editor openI cant see a single occurance of a Fuse class in that stack trace so it might even be add more details how to reproduce +Untested patch +The version is but the are +Like in the readers +Include my test +to trunk +unborderedWith the of the cells in the bordered row are rendered differently +If so this issue can be closed +Patch with fix +This means regular are created for inner types rather than +Whats missing? I know I need to add some stuff about demux and archiving +There have also been requests for cells between time X and time Y +Since for a given id there can be several repos depending on branch for instance its better and less to move to as mirror key rather than repositoryId +Thanks Sumityour patch with some changes is in rev +I am closing this +Confirmed with ER that server starts up on machines with with G of RAM even on the supported. +Is that xercesxml +Ill close this and clone one for fix to allow me to spin rc for +I misread the issue +bulk close of resolved issues in. +Patch applied and committed revision +Committed. +Attaching a patch for clarity +test is bit of an outlier because it uses cgroup hierarchy instead of the testing hierarchy that all the other tests use +Almost there! Can you clean up the indentation to use spaces +HiIs there any update in this issue as we need to get it resolved for a clientThanksDarren +If youre ready to add JNLP component descriptor generation please see the patch I offered in +Patch has been committed. +To make it a little clearer why I see this as bad separation of concerns consider that you could accomplish the same thing by putting the impl in each of the distributions and then having use the impls in the distributions +For Eclipse CRTLA CTRLSHIFT I agree that it sounds like a separate projectBtw Im not interested in of XML but of validating how other programmers did the formatting. +This patch solves the problem +The change was +Thanks Sreekanth! +a lot for the review Todd +Currently SICS UI uses for instrument login dialog and status bar button +We will need to rework the rules that define how dependencies work before we can solve this. +It also include two unit tests +verified that specified stringspaths have been removed from above in ER +Please feel free to reopen if youll still have this issue with build +on the IOD project weve been using the term XCAS for any XML CAS +Setting the +is available now +The whole statistics object is but I put updates to it in a separate thread so that they wouldnt block the current request threadSeems to work for me +By the way Im using openjpa code enhancer during development time not using Spring for the code enhance at deployment time +whats happen +made communication timeout for bulk importing dont need only configurable timeout but need to retry slower +Note this change when commit +closed +I dont expect the patch like to create so many warning though +This zip file contains the source webservice client classes +This definitely seems inconsistant +Should be pretty easy to fix +Upload the path from +I think we are assuming a lot of people use queues and queue +Many name them after the issue and then add a counter in case there is more than one patch per issue +I will not enforce Test dont kill me +Hi Ant I could not replicate this issue seems to be working fine for me +I will try adding pragma headers and such +Thanks Marquis! +We also experience this the new relative link processing done in doesnt work with used the attached to to try the latest this issue no longer seems to be a someone else please confirm this +The patch file is already committed to the branch +This is also addressing an issue Ive seen some time ago while testing Whirr with +New patch addressing Sureshs comments +Removes the fetch lease checker because it is redundant with the socket timeouts +Fix in SVN revision dont believe the patch for this issue fully resolves the problem +RicottaThe fix in has been and that issue is once again resolved sothis issue should also be resolved now +Therefore the current regionserverhlog sequenceid could mess us up +We could use the enforcer plugin in a profile if you think its needed +It is in too btw +The internally provided and classes have setters for the and helpers which will be called on configuration update. +Syncing patch with trunk +Any possible issues with me removing the qpidgentools tree? AFAICT it has no content now +Nice to have feature +Is it a state right after opening file? Do buttons look ok when I select an item by mouse +I am attaching several screenshots +After changing memory parameters I had one GC every several seconds which completely solved it for me +on both ideas tho +Yes and that does occur +Phil it makes perfect sense +Im a little reluctant to lose that +reopend for clarificationare you saying the seam examples have been removed or just the central tutorials +These have a sign encoded in +Im marking this as wont fix although it seems to me most of this task was actually completedBut weve moved on from. +I guess the behind this feature takes care of this? If opening an IW on the source indexes can lead to corruption of the target index is there any way to avoid itI think this patch is ready for commit +ee tests moved to +We could mark it experimental in the beginning if we want to give it a soft launch for users only +attached patch +The Python team here wont be able to look at this in the very near future so I wanted to pass this on +Bulk close of old resolved issues. +I created two +As for as possibilities in other software Microsoft Word supports a side by side viewing I feel like I have seen some adobe products do this but I can not remember which one +Differentiate between the case where a function does not exist and documentation for the function does not exist +in +Need to come up with a test case for it +You can run it by doing a mvn test +Then I set up the web console on broker attach consumer to broker and producer to broker +Thanks Arun! Ill commit this shortly. +A new session will be opened on next operation trying to access the DB +jackrabbittrunkjackrabbit it is then +Your problem might be related to bug which we have already fixed. +Ferdy yes not filtering and normalizing is implemented as well but the patch seems to be missing too +This is most important for the archetype repo but the quickstart repo should also have this +Closing since change appears in latest alpha docs +now that its time to remove the attributes in Myfaces that are not present in the RI Im worried about breaking users applications that may already depend on them? For example I cam see how people would use the onclick attribute of commandLink and be frustrated that it no longer works after its removed +trunk Jeremy you are welcome to test with latest code from trunk to see if it works for you now +Some comments on the patch In the logging should be done after is added as a resource +Thanks for reportingI did a simpler fix by using Random that is easer to understand. +Works great +See the Weblogic documentation to understand how to configure that when you get to Weblogic Try the t protocol and let me know how it works out +Added anchor tags to help navigate to entities within theThose are great improvementsI made a few changes to your patch +gitignore from this patch +My first patch attached changes as detailed in this ready for of the in in patch will mean that the caller thread will execute the +But in the latest patch download button doesnt work +The stub file should store usersabc and we should look for usersabc +In MRI Mutexsleep appears to return the number of seconds slept +Please contact David Carter if you need immediate assistanceNote The information contained in this message may be privileged and confidential and protected from disclosure +patch for seems that your patch is not simply copying codes from but also contains some additional codes +Closing as requested by Viola. +Right the replacement is done in the region server and in this case it gets bypassed +The additions must have a matching entry in sarr of srcsoap +I can then the method Ive created here renaming it to removeArgs Start with getAll +Im not sure I follow what youre saying though +Do you have a sense of when the other one will be done +The fix has been committed to the Beta branch and trunk +Makes a little more complete +Patch attached. +Bulk defer all frontend issues to +Applied the same fix to both in the second patch +The test is for hadoopIs there HADOOP JIRA for this bug +Also changed the logging from TRACE to INFO +classify and set fix close of issues +It makes it so much easier for us to review and comment if patches are created from the top level directoryThank you also for your attention to detail regarding the formatting +Fixed in also for +This file changes the sample code so that the class takes the XSL file as a parameter +got everything here for at least staticuserrepository +Thanks +Just a workaround may not be a good fix +Running tests now +Sorry +However per my experience of using other portals as end customer it seems I can always dragdrop portlets immediately after registration like liferay jboss portal etcWhats your thoughtregardsMichael +Ben thanks for the review +This no longer seems to be an issue in trunk as of +thanks for the report which version of Grails are you using? This sounds familiar and am sure that issue was m using version think its a good thing to allow different javascript libraries to be used in grails but wouldnt it be possible to have the same syntax for attributes regardless of the library used so that a change of library would be transparent regarding the usage of the gsp tags +Just invoke maven cactusLet me know if you need more I could reproduce it I will try to fix it tonight. +AlanThe differences I observe when running on actual csv files are within the margin of error sometimes comes out on top +It wouldnt make things simpler but Id expect either design docs having a property allowing anonymous reads or at least a databases secObj could hold a list of design documents +Ill keep you posted on what I hear back from them +I will use for nowThe failed unit test seems not related to this jira +Closing this issue because it duplicates another. +I would be worried about baking that into a core supported feature of Roo since it would make Roo applications require a license which would defeat some of the purpose of Roo as a toolkit without licensing I didnt see it +Lets experiment before we commit to a new parameter +CDI tools model takes into acaunt both the qualifiers and the type of injection point +I just filed. +BumpThis feature is needed. +should be fixed per other related jiras. +need to change the release release note according to comment in +Thanks to Robert for providing the excellent test app +Please change to if you intend to submit a fix for. +This does not look like a groovy bug to me +DanielYou are correct but what you are proposing is a useless feature +as well as attributes +Ive now added you to the apsite unix group +But it only does not build if you locally updated to openejb +Assume closed as resolved and released +I had no idea that this was considered a bug let along a fixed bugIll set up a test case and see if it indeed still behaves this way. +Chatting w Nitay I thought we could continue hiding ZK from noobs by doing something like adding a new property in named and in it wed list all quorum members and then in background wed write the but Nitay just raised his eyebrow when i suggested thisSo Im with him now +Instead of the partial LGPL licensing would the author willing to license the entire library under the unrar license +Updates servicereference handling of operation selectors per handle the JMS reference binding issues in a separate JIRA when Lukasz has a chance to look into it +This is being caused by the difference in PartitionersI am using Murmur Partitioner on my CF but since I didnt specify it during Table creating in Hive the Hive Input Format defaults to using Random partitioner! My bad I didnt notice it while testing in MapReduce as I set the Partition thereI will be testing it extensively over next few weeks including one deployment in production Will update if there are any more issues +It is not explicitly supported though and neither is any secondary build plugin +applied provided patch +Rev o Reverting all changed made in rev as it resulted in build break +However the will ensure now the outgoing message has the right STOMP command +Update this commit does not seems to be the one breaking and showing Factory errors +Stripped down XSD +The issue would not be caching it but rather knowing when to discard a cached copy +Dont commit this yet +pushed pushed upstream +They are actually the same thing +Thank you for your improvement +Well I may have just thought of that last one +jieryn you are right but at least I volunteered to do something not done in yearsAnyway give me some more time and Ill add testsNotice however that old tests work +Could you clarify this? FurthercreateDocument method I couldnt see this method in or +While looking into this I suspect the problem was introduced in where several dependencies were removed from the advice of the mvn dependencyanalyze reportThe report is correct hama doesnt depend on those item but instead has a transitive dependency on them +Modifying the test case such that the message is prefetched but the transaction is not committed and then a new consumer is attached to the same queue result in no redelivery of the moved message. +I generally agree +This is definitely involved and I would not recommend for +I have upgraded to mustang build and that indeed fixed the problemSince the other test is broken it sounds like this patch is not ready to commit +and Im not sure it makes sense to add mins and maxes for all of them +a release plugin bug +We can consider your work a continuation for itAmine +Thanks. +Looks good to me +That sounds like a plan +Only when it comes to the alert print from there the slashes disappear +reporting schedule is monthly for now. +GlenI should have removed the annotations in my samples to make the point clearer but just overlooked them +Patch should be integrated to branch as well +everything seems to work fine +Sorry still waking upthis morning apparently +workarounds remove xs in branch +This was on a node which was down during upgrade and got revived afterwardsWould it make sense for to refuse connections from running an incompatible version +The previous patches cover the case where the XSLT document itself has XML errorsThis third patch covers the case where the XML document in the request body is has XML errorsPlease also include this patchThanks +Cancelling patch also no agreement yet +Tests pass + we ever rerun this? Wheres the SUCCESS log? +Added necessary glob handler to spring config for oracle or other +Thanks for the patch Jochen please if you have a test for this uploaded as a new attachment + target findbugs warnings suppressed +This should be fixed in as fixed in Seam on Stuarts advice +If you ever have any fixes to the code you so graciously provded please by all means share. +I went back and checked on and and found that while the core embedded problem existed in I do not see the network server exposure closing sessions on shutdown with the branchThis is very interesting and may explain why myrna only sees hangs with her runs and not +Do we really need both and? The name of the generated interface and methods should avoid potential collisions with interfaces and messages perhaps by using +Arnocould you please verify if the fix Ive committed in rev +This problem was solved by Upgrade to of . +But Im leaning toward writing out a table of compressed record lengths at the end of the file +I insist that the SHF properties in should either be fixed or removed +trunk +Im willing to help with the rebase if you need +His main comment was that while my previous update made remote logs stream the local logs were still loaded into memory +For Windows we skip test test +in some cases you dont want to do a full installbuild you just want to test your changes so it wouldnt make sense at that point to do a clean and then a build +maybe jacques you can figure out what happened there if u havent time + +Its a matter of resources +I wonder why I didnt get the new code + for the patch +HiYesterday night in the middle of my reply the JIRA went into maintenance mode +Navis please supply a patch to place on top of this one +I think thats rapidly becoming an outdated idea and not worth the issues it causes +My pleasure to help +neweditsfalse is now supported for more operations +what stuff belongs here vs collections APIGood question +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +If you have a more complete definition and it isnt aggregates then please raise it on the GAE DN plugin issue it possible that the shift to has broken s relationship sugar? introduced some changes which handled unowned relationships on GAEs datastore and these are no longer being applied + +Interestingly a few of the processors did support the cvv response but it wasnt saved anywhere +Thank you now it worksJrn +Im thinking there will be more improvements to these perf tools as part of +The downside is that rename refactorings will take longer for Java classes defined in Groovy projects but the benefit is correctness +You are correctThe concept could be expanded further to give applications a toolbox of layouts like etc +bulk defer of items from to +Added log statements to print the theread number acquiring and releasing lock but log shows all in order +I know of at least a handful of production deployments that have circumvented this with a simple cron job looking at log dirs and doing cleanups outside of the framework and do not this this to be a blockerKonstantin its your call now +So example is wher c like ccase c is BASIC collation c TERRITORY collationcase c is TERRITORY collation c is BASIC collationshould this just behave the same as where c c with different collation types which I believe just throws an error +we can close the issue now +good question can we apply this change with the version installed there +James are you happy with the solution Roshan has provided? If yes we can close this out +Hey Luca is it worth returning? I know that doesnt love null +As far as is concerned an InOut gets created for each requestresponse so creating the instances inside the interceptors wont buy muchIs that behavior changed in the trunk +I will work on this soon +Closing released issues. +Thanks +Verified by examining the generated service control from the following WSDL which contains soapenc for various typestestwebservicewebappwsdlschematypesVerified using svn rev +I was trying to build the source bundle of but when I used mvn sourcejar I gotNo sources in project +im not sure which change and when fixed it but its already fixed in +Resolved thanks for reporting +Unless there are objections I will commit this soon +Changed test the test case now tests for different kind of job movements which are possible Submission and completion of the job alongwith removal from waiting and running queue at correct times +Please check that it was applied as expected +JSP docs cleaned +While this solution works for most of my problems it fails to address some of my more complex issues +if something doesnt fit in it will be made to fitNo offense just facts + this could be a reason of could you please verify that this is fixed in beta? I am unable to reproduce the issue +Lets look at this one once we will have investigations done. +Please help correcting it +Some minor code changes are missing +Currently I return in the patch +I wnated to make some changes to the bookkeeper api so that it easier to use and can be extended to write a streaming interface on the ledgerhandlecomments +Requires a update and theres no stuff in there for it yet. +ASF will be used as a fallback for message body agree +I think really on the Thrift side it should be the slimmest wrapper around direct communication via +Attached is a modification of the along the lines of the current +Done +Hi LukasIt is very strangeWe have people in our group +The reason is the the order of the class path entries in the surefirebooter jar is different +The attached portlet source and portlet descriptor can be used to demonstrate this issue +The widthheight cannot be changed either +Thanks for the runtime classpath +Perhaps this is too much informationI was trying to be as clear as possible about when users may see a difference due to this change +I see no changes to in this patch +The dataset is extremely small rows +Is this ok +since subclasses to add logic for shutting down the db server its not going to be useful for connections +No feels OK to work for feedback +If this occurs I added code to log an error and build dummy data to be returned to the caller +patch applied and tested OK +Done +please see the previous commentColm +Will soon commit this change in Derby codelineDerby currently requires all the character columns in a table to have the same collation as the collation of the schema inwhich the table is getting defined +Thanks Andreas for the timely contribution of the class it really helped me to solve the issue with Hessian large messages +The correct system property is of today there should be no occurence of localhost in the webservice test sourcesresources +That means you need to have a client handler active when shutdown one of the failover server in the cluster +Target version revision +Please open a new JIRA for any pending issues +queue wiki update should be checked + +I understand nowI will try to find some time to play around with getting the AM ID to not have a wrapper at all +v is working here thanks +Please also add junit test for these unit test was added for token services +The default behaviors of the current code have been causing troubles on my very heterogenous cluster for a very long time uniform distribution only really actually makes sense if the same amount of space is available on each drive +Just checking but does your gadget have osapi feature included +To fix the compilation error is sufficient to make generic the +Pity +Dont know if its worth working on the current UI thought maybe its better to handle this when well find the time to recode the UI +Rather than putting one counter into the general class I just used the existing mechanism for reporting extra infoI also adjusted the formatting and style +Youre right +Looking through the code base was the only pipeline component I came across that allowed the element it deals with to be set in its configuration +how much of the seconds is due to the loggingwould it be simpler to change the log message to DEBUG instead of INFO and see how the performance of the regular commons server behaves then +Does something need to be done there? Should I submit a patch that removes all the broken to make ant build in piggybank work on trunkTo get piggybank to build I had to remove! contribpiggybankjavasrctestjavaorgapachepigpiggybankteststorage! contribpiggybankjavasrctestjavaorgapachepigpiggybankteststorage! contribpiggybankjavasrctestjavaorgapachepigpiggybankteststorage! contribpiggybankjavasrctestjavaorgapachepigpiggybanktest! contribpiggybankjavasrctestjavaorgapachepigpiggybanktestevaluationstring! contribpiggybankjavasrctestjavaorgapachepigpiggybanktestevaluationIs this just me is this fixed on other branches +If the error is present Ill check it againregardsonce again its me sorry Thanks for the update + + and unit test cases have succeeded +Heres a new version that throws when the anonymous mechanism fails including a test for this situation +Applied +Phil I believe this is best marked as Major +Please take a lookThanks +To fill in some color we had a customer complaining of data loss after a crash +dea +There was an issue fixed as part of +I did download the source and looked at all the attached files and took the source from there and build the project +Hello Pete Hello TakeshiDid you consider contributing your in extensions to the slfj project for the likely benefit of everyone involved? Best regards +Backport patch for +Thanks for patch v +So we can schedule it to run even every minute alsoIf there is a need for such complex scheduling we can consider that syntax +What about Solaris Mac OS and BSD +Could it be that the target version is set implicitly when explicitly setting the source versionSo how about setting the target explicitly too as we do in the JSP scripting budle +provides access to the in addition to the current event and current user samples reflect using the. +Main methods are good practices +I knew what you meant but wanted to keep my changes isolated +Raghu any progress on the JDBC driver + +The groovyeclipse patch issues have ben groovy patch is not needed as Groovy is nukedThe groovyeclipsetests and features patches applied +People might want to convert to different formats or even use Unmarshallers +Thanks Mikhail! +Revision for on branchRevision for on we need to wait till is released +But it is happening +Does this and resize volume partly address +If I put in place an hadoop jar that is minus it all passes +I just tried it and it will work perfect for what I need +It would be nice to be able to specify two things on the Date fieldType +Yes I would be willing to try and give feedback on a nightly build +Needs separate patches for and because of the directory and package restructuring +All tests pass fine +what maven goals youre runningin hibernate checkout and run mvn clean install +I just checked the Solr tutorial and saw thisThere is also an optimize command that does the same thing as commit in addition to merging all index segments into a single segment making it faster to search and causing any deleted documents to be removed +Committed trunk revision Merged revision +Following algorithm will be used for dominant DTDs collation determination +Updating fix version to reflect that this did not make v GA. +Thank you Kathey for submitting the second patch +The provided patch adds three classes to the test tree of javaflow +I can recall cases where errors in shutdown were important to diagnosing user problems so would prefer we not introduce a blanket assumption like that +hmmmm +No tests are included but I am looking for feedback for correct approach and what areas to test +Verified with JBT. +The result with current is an entry which looks like thisdn snAmosousystemsn Amossn AmosobjectClass personobjectClass topcn Tori AmosExpected result is that snAmos does occur once in the entry not twiceIf someone reopens the issue I will attach my test case on this +Thanks for additional clarification BikasDo you feel we should make some changes to the test? From the perspective there is no need to circulate the PID value back thru the pidFile right +Moving to CR for triage +update release script needs some more updates +Patch attached +Patch with the sample code plus a test caseThe test case failed at first! Ie the returned string was over the specified limit +I would like to see the number decrease before we put such a scheme in place +is deployed on codehaus tagged +It is a very good idea but it also requires changes to and other classes and I didnt want to do it this late in the release cycle +Hi Ioannis +Those are renamed files bizarre that theyre not going through in the patch +It is implemented by searching for the word Release in the full database version string and taking the following number expression +We plan to address this in the short term by making the JVM OMM much sooner than is the default but it is still just a on the underlying problem that unless there is back pressure there is always the possibility for incoming requests to overwhelm the system +Willem whats the status of this +Ive updated the release notes field with information on how to configure the old behavior for users who dont want to switch +Thanks. +They just need to have two extra null parameters removed +Turned this into a documentation issue and lowered severity a bit. +For instead of commenting out huge swaths of the code just fork the class into +Closing all tickets +with current version i am hitting a bug in jpox that is fixed in the latest version +bq +I think the only difference is you need to use patch p instead of p +Could you please add link to EMF bug report if any +Your fixes conflicted with a change of mine and restoring yours I noticed the bug was not fixedfor two reasons in the web module contains serviceStratagy so the lookup always ends up in NULL the same file has service names that are totally diffent from bean ids BUFFER SPEED FILEIve fixed the typo but I guess you have to the rest +Im sure I will learn quite a bit in doing so +The fix for includes this patch +But the jdoql tests did run and pass. +This gets tedious when using different branches for creating patches +Who is the directory owner in your test caseWe are setting owner to a newly created directoryTo my understanding the creator should be the owner under POSIX mode including HDFS +Yeah baby +looks good will commit if the tests +Take it if you want to +This covers accessing the old way via home interfaces locally and remotely versus the new way were you look up the EJB proxy directly +Hey Slava can you take a look at this one +Applied the patch in rev +I forgot to commit some changes to a test class +No I havent fixed my machine yet +The following jar files are in ECLIPSEHOMEpluginsAnd in ECLIPSEHOMEplugins show and describe are ij commands not SQL commands so even if SQL Editor updated its version the commands would still not work outside of ij +AlanI have tried setting terminatesm flag to true i did not get any crash im running load on the machines since yesterday. +This is really a bug so changing its issue type +Maybe you would like to add this feature to forge CR wouldnt you +x r add assume false to test for java are some jenkinstest issues to look into +Committed revision +we should just check if the DN is already deleted +There appears to be some problem with the level in particular +I dont have access to change states +So the user clearly is informed that something targets head but head is not there to target it we should add that as well +In the error message you mentioned the error The content of element application is not complete +I think that centering the whole page could be good +Seems like counters dont work for local job runner +hours +Dist packaging +I am submiting a new patch that addresses Kristians comments +Good catch Ill patch it +Duplicate of. +weird bug though +can we call it a Header instead of a Handle to be consistent with the specDoneThanks Doug +As I said ive tested this with cvs H version which is only a few days newer than H beta and thus no changes in this areaAnd this stuff works without ANY problems hereSo until you can come up with a case that fails on something else than your computer then i wont do anything +also repo parsing now removes all white space and newlineendline using regexp sI dont think we can have URIs with spaces but I could be wrong +Yes you are right the website isnt clear +IMHO this would be nice to add support for newly announced +Tested this with manifest atrribute values of heightwidth set as +We could use Externalizeable there with no XML +I would like to the changes to the public API before I implement the support for symlinks in HDFS +For this we may need more elaborate catalog system that maintains and keeps statistics informationFor them it would be a good starting point to refer other database systems +This would block the code from compiling +start the app +Please update code to work resolve the issues Ive . +Feel free to change it to find if you feel strongly about it though What we should probably do is incorporate this functionality wihtin the crawldb reader instead of having a separate class for it +buildUsersjoaDocumentsworkspacesmyCompanyworkspacebin +Status update all features implemented in the core the following extensions are needed in the admin console that are currently enabled only for users +it is an error condition so should be ok to log info +The patch will need new hadoop sources which have not yet been released on apache so until then the patch can be used against hadoop trunk but since pig build picks released hadoop this would not be seemless. +Awesome Carl +removes one line from each of the two affected files that belonged to another JIRApatchCommitted patch b to trunk with revision +Is this a fax andor TIFF onto which youve put the other text +I havent retested it yet +We never figured out why the performance was bad when using Java threads to handle voluminous output +Unified patch against +is able to extract text in many configurations including non latin character sets like russian japanese chinese etc +Werner Guttmann wrote For this case do you use the extends attribute in the mapping file as well iow do both classes have an entry in the mapping file +Another attempt at a patch for V +AFACIS we talked about providing non breaking API extensions to support multipart handler plugins +hymmm +What servlet container is this problem occuring on +The docbook structure has been checked into CVS under webservices module +Does this happen to only consecutive joins with the same set of the keys? the above does not work for the followingA left outer join B on right outer join C on Consider the following rows for a given value of cA a aB nullC c cSince there is no join no pruning will happen and the following output will be producednull null cnull null cnull null cnull null cwhereas the correct output isnull null cnull null cNote that extra rows will be producedSo I think the patchs approach should be to add more you add explain plans for the new test cases +Hm I am not sure were actually saying the same thing +I guess its not implemented for now +mvn U clean install eclipseeclipse makes the trick. +New patch initing the postings arrays in instead of per term +You have the in the wrong place +Also changing and to pull error from public static constant in +Although this can have problems if the address changesviews change midway during a call +Since it loops the disk fills up very quickly +Other tests are more than welcomeT Jake on last read new version gives compared to with current version of BRAF seems like it does not blow away page cache or I misunderstand something +A patch file for Chinese localization of Accounting +Link to the jacorb upgrade +Then you can annotate the exception and to make that test run you need to provide a property to System or whatever you have +However it shouldnt be too hard to add this configuration option to the Zookeeper discoveryIn general I agree that the connection management of the discovery implementation needs to be enhancedCheers the enclosed patchI found the best way to do this was to restart theThe restart is done when one of the interfaceMonitor gets a event which means it has been able to reconnect to the zookeeper server but it can no longer watch the node because the session has expired +The provided scope is unsatisfactory during development time within the IDE +Verified by Paulex. +interesting idea but what happens if shard changes its role? Wed update this central overseer and hed need to pull the entire cloud state again? I guess I am not quite following this +Removing a stray logging statement for the review and comments BobbyI havent included an exception +Add clover related targets +Is that a shared installationYou might be onto something with the permissions though +I may have been imposing my closed source background where functional specs and implementation specs were always different documents and were writtendiscussedreviewed separately +That is fine + to trunkCommitted revision +Please take a look +Nothing works then +Admin please delete older file +When that is said I have to actually verify that nothing is deleted until the repair is completed +This might have been solved with but we should confirm this with your test webappOne more thing +My question still stands though wont the jar file name changed or not reappear with the original name whenever an upgrade is performed +deployed into production +Hi StanI am closing this issue as it is fixed now +Yes the channel moved as part of the SI work +So would it be better to put the encryption where we write to disk where we already compress? It seems like this might be less invasive and would be more completeIt has downsides if you do lots of spills but it is much more complete +Ive just patched the assembly to also use the release profile all tickets +Bulk close for +I only bring it up now as its a directory layout issue +Green Codehaus LF looks like crap +However perhaps there is another way for to achieve the functionality it is looking for +It seems I deleted the wrong one +r on trunk to be merged into branch +throws an so I need to catch it in or +Cool thanks Andrea +response timeout I close this issue ifwhen readyRegardsHenk Penning + +It would be great to see some activity on this issue +We went through a painful period where we discontinued contrib projects got like Pig Hive etc +Julius were you ever able to reproduce this issue? Are you actively working on it +Since this one is committed lets continue in. +Ok fair enough +should have caused a failure on Hudson +applied thanks. +ant FEFBEAF Dtestcase clean Without patch Total time minute secondsWith patch Total time minutes is a patch with some improvementsI still have not gotten to the root cause of the slowdown +Ill try to find some time after this release to see why those errors are being reported +the for +Theoretically the release could be cut without it +Either the builder is missing from the install or it belongs to a project nature that is missing or disabledENTRY MESSAGE Skipping builder for project +Jason Anything left to do here? Can we close this? Should we file separate bugs on more detailed tasks? +Magnolia and SBS have been updated +Thanks Tony it works fine. +This patch encapsulates the EOF marker used to terminate persisted character streams +Struts keeps looking at the web root instead +I think it is better to add it as an option so only the client will need rampart to compile +test fixing way +have been experiencing this problem too +patch hasnt been tested cuz I stay at home can not connect to the companys vpn +I fixed this and one additional hidden field error in the same method +A page is a leaf nodeAlso review the Graffito project as we are considering implementing a Page Manager with a Graffito backendThere is also a Site Manager portlet which I was planning on working on over the next month +But this would be a great format to have since it could be available in the default install showing geoserver supporting more than one format because we can include the data +Gabriel here is a patch with associated test case +Unfortunately Xerces no longer supplies the class necessary for the FOP commandline application +But if there are ideas or suggestions Id be interested +Closing all resolved tickets from or older +You can close it +all rich classes present in templete and fixies in junit test were made +In particular the one I feel most strongly about is separating out the keys which are used in weblog rendering +Good ol pragmatism? Why sanitize user input and adjust the token when you can hide errors by making core functions more flexible +Feel free to it to me and Ill take care of it +The updated list of library changes is in. + +An example of the header property is in the for but under the properties arent being shown I still have to investigate why. +Connecting using the for to leave my code unchanged Ive run into a strange issue though +Should we check whether is neeeded to checking permission in collecting score nodes +Currently we display the collection size which potentially requires persistence access something rather expensive in light of the main purpose of this better debugging support. +The best approach is to call encodeURI on your URL and it should work with jQuery as well it returns what you expectThere is a doc issue that tells you what is required for the API function +Reto! Not directly in Clerezza but I actually develop a mail fetcher for Stanbol and this is where I see the problem of transform in in OSGI environmentAnd as Clerezza already have a bundle for I think more useable to share this bug fix into +Propose to mark as closed in the next few days as the resolution gets reviewed accepted. +If someone changes their similarity to use a different norm type at than atthen he or she is an idiot +The comments are replete with when a particular field is added +I lost a good time finding that the problem was in the user +I would suggest asking in the Teiid forums as to their schedule for this support +Therefore resolving as out of date. +Just synchronizing the patch with the latest trunk code +We will look at this in the timeframe +Im not sure this is really an issue here +has been released +Done +Ok it seems to happen only on first series of index updates and only if no query was ever run before. +Now we have and +Also do you have any other warnings for example about missing Ast transform classes +I am using Indigo +it +Might also be worth considering migrating your project to svn over is a better alternative all around closing the request. +update trying to rebuild the slave now with updated fbsd done had to build a new vm with etc. +Could you please try again +Its really for reference on this bug for adding +I dont have the Manifest +Patch for helps address getting Traffic Manager working again +With Axis this works fine Fabian +in cabfafbcaaadfdeacda +Why do we need another lock? The caller of process should already have the lock rightYes youre right +The testcase works fine with +release bulk close +When are you planning to release? +New improved and without my modified develop project file included +It is possible to remove multiple white space characters ina single loop +Sorry dude I must have blanked +Attachment contains requested modification +It allows new WSDL extension elements to be introduced in a graceful manner +remember to run your eclipse with jdk too if you want to load jdk compiled m not sure you understood my point I am using Eclipse with JDK +Here is a patch that switches the build and adjusts the plans to use activemq +Thanks +You will need to delete the files under the webroot directory click for the to be redeployed +Build source tarball make dist do build from source tarball step above should be skipped +avoid putting asyc io in boost io queue when startDelivery again +Added test case in the patch and verified results against pg +Compare s existing optional use of Equinox API if provide a backport +bq +I just found in Jansi +I think it would be more useful to have the unchecked faults logged at WARNING. +Enclosed for detail +Im posting the result of the latest test run with the port boundThere seems to be some issues with replication although that is probably normal +forgot to mention the patches are based on tags +It happens occasionally during our automated testing whenwe shut down our server but I havent been able to force it to happenon command +Ive merged the necessary bit of toolsrelease +Have you any suggestion how to use the exception hanlder and not loosing the ability to print filtered stack traces +FWIW from a hive perspective it looks like if the automatic profile activation was removed it would be easier for us +Shouldcould be extended to all other SBS pages too including articles +It didnt break any of the test cases but that doesnt exactly mean that its the right way to fix it +Looks good +I think its due to the problem of G server deployer +Thanks alotTo all Anyone objecting to committing this to a new samplesnotes project +JSF support for the Messages API is already provided by the Seam Faces module +For me it seems like you are adding portlet jars to your app which will be part of the web deployment which thus will collide with the jars provided by the runtime which would result in you getting like this +No this is open but I think I can resolve it with acceptable effort +I tried to execute testsuite with suggested changes but I still have many errors even on trunk without my changes +I attached patchfiles for and +resolved issues for already released versions +Uploading patch +I have not run tests but the build completes posting for feedback about bumping the jersey dep in general +I plan on committing this later today +Hi Jesse +Ive now the page to remove the spaces as before and no change occurs +Let me knowCheers Assigning this to you +FIT and its family are truly innovative testing tools +Even build failed and the same testcases failed due to the same exception +Any user on the same machine could shutdown any network server even if they did not posses any valid authentication credentials +The rest is done in on the clients browser +Jonathan I havent seen any patch uploaded for +The patch needs to be updated to the changes in trunk and then many missing pieces need to be implemented +review merge with as mentioned in the review +Support is included in Logj . +The spec doesnt say anything about for loops and coercions +bulk move of all unresolved issues from to +This problem also occurs with exclusive +First draft of patch +This made it difficult to connect the code and the accompanying explanatory notes +The bug does need removing though +I dont really see that as much of an issue for anything we have right now +The CRS API has the interfaces to do this but currently the CRS objects do not have the data so it is not useful as yet +Always enabled +I can take this one + +With that said I believe that the boot class path is being modified you just cant see it in Java system properties. +At line uri is not necessarily of local fs so other components of url needs to be preserved we shall use url here alsoI believe your test case will pass if you revert these changes and keep other changes +allow as the type to in rev . +the Hive wiki may only say it requires column names but I think the syntax may allow functions in orderbysortby +Maybe there will some kind of prioritiy handling in Jackrabbit someday +All we need to persist is the single master key used to generate these tokensCan you please verify that this is the case and RM restart will continue to work after this change? And file a jira under if it wont +for the patch +Isnt it +rails app has been updated so that the produced http configuration implements this +Task log redirection is done in the class +Christian +Seems hudson didnt fully recover from its long hospital trip +This is old and Gabriel gave no response so closing it. +Preparation for Gora release candidate +My bad I thought wed only change the public +The NPE is now fixed and an error message is reportedI created a new ticket for the check if multiple consumers is supported when manually starting a route. +Have messages passed into components as parameters with defaults coming from the components message catalog +Someone else should take a look at it. +I just created the simplest possible interface plus a default implementation which does not deal with attachments or elements +Also I tried rolling back that last few commits to but they continued to fail +Moving them to clientnegative +John Can you please take a look at this? Thanks +Moving out of +Therefore the field descriptions would be used also in the help systemWhat do you thinkCheersChris +Meanwhile a Simple implementation has become available so the chance that there is real use for a Simple implementation has become very low. +Thats what your WSDL is doing +Perhaps that can be adopted hereIssue One of the unit tests is failing +If failover is successful stop needs to be called to terminate the log shipper thread +BVT java tests +Changed +Scenarios ab and c does not seem valid +Patch Prog however now is more dont think this has been Done isnt there any type of resolution that would better reflect this issues status +Patch is attached that fixes issue +Hiifve upgraded to jsvc and it works nowThanks a lot! + +There was an error first time you accessed the portlet it seemed it was looking for a local LDAP by default +The below examples are what Checkstyle wants to have us doSo does that mean code not in that hideous style will be flagged by Checkstyle? Im confused by the next example you posted which says Checkstyle wont enforce indenting post line wrap +I still am getting some errors and failures but I still seem to have a lot of interaction and dependency between tests +All other tests passed +Can a committer familiar with the compressiondecompression code please review this patch? Thanks +Added missing test case +It could be that previously I was confused about the actual behaviour because of the bogus test. +At this point Im more concerned about getting the format right than optimizing performance +suggestion for discussion +Ive patched my local version and how the Run target from Netbeans works with classpathScope of runtime +Just to clarify the Prefs bundle had assigned +Patch to ignore comments when looking for LDAP referencesAlso includes visit from and police +The problem had to do with the fact that the default charset encoding can vary jvm to jvm and platform to platform +Increased the width of fields in the output. +Id like to get these changes committed + +Ive added the test message and committed a fix for this issue +Do you have a ready made reproducer for this +Marvin do you have a summary of the rationale +So we dont need it to be available all the time + +patch looks good +This Java client would be extremely useful to us and we would like to see it in the next stable release +test and proposed fix +Here is a patch that resolves the issue +Will fix this as duplicate of +The patch moves all common map functionality to +The vote has passed the artifacts have been promoted on the central maven repo. +See the class in +bulk defer of open issues to +If this is then it belongs in the mapred package no? So if we think that might be useful to applications then it should go in the fs package but the code that creates a based on the value of surely belongs in the mapred package +Thanks IgorPatch applied to LUNI module at repo revision rPlease check it was applied as expected +Leave issue open to see if it fixes it +cordovalib when you point it at a local path instead it just copies directly from the local path +EdellThe level of the packaging of the created is pom +And i think they share the same confBut not sure about how compression info for queries like load data local inpath We can focus on correcting the insert overwrite in this task. +This has been fixed marking as resolved +only fail maps and reduces that are currently running +Please close +I just fixed so you will be able to try the used the javadoc plugin version and the problem is still there +Committed revision +new jar verified with the latest bits that it is indeed fixed +Thanks Kim +Agree +Sorry if I was not clear enough but Im not deploying Cassandra inside Glassfish +screen this record since is now released. +Finally the changes to binhadoop did not name the log files correctly the correct log file name should be normally set in bin and used in binhadoop +Reporters of these issues please verify the fixes and help close these issues +Hmm +Thanks Jacques for your commentsYes we are proposing to addIve gone over the discussion on customMethod +Moved to CP for triage after ships. +We can just postpone it if it was not possible to solve it +Thanks Mark! Ive always hacked something like this in to projects for easy running +Mapred is really just part of client in my mind +and committed to contrib area +This patch is to be applied after the first and adds ship estimate functionality for PURCHASEORDER +Just run through QA to make sure + needs done first +I can submit a patch here +I tried annotating the entities generated by JPA but Roo complains that its not an entity even with the Roo entity annotations +Id like to add for the sake of completeness that I ended with the following commandsu RUNASUSER c exec CMDNICE WRAPPERCMD WRAPPERCONF PIDFILE TRUE the m was causing the root accounts environment to be used instead of my continuum user account +Note that the code is executed only if doingREST is set to true +it should have the following passed inXmsm Xmxm XXmWe also need to be able to stopstart the app server correctly from the jbossws file but it currently doesnt work correctly because JMX is secured with a userpassword +Ive been testing this including setups that continuously reolad the cores to check for memoryclassloader leaks +We use exactly the same modifiers as Java does so it is no problem to reuse that intend to make two distributions one which has support for indy the other not +There are a lot of people looking for this we may pull it in +Fixed by changes committed for +This issue needs to be re opened and the second test case tested +Verified by Anton. +We really havent had time to revisit that system to run more tests though +I wonder does the broker in your code have an associated amq sslContext +applied to the future a maintenance patch like this could be applied to especially since will remove this code completely from trunk +But when the session makes at least one lookup toEntity the problem the lookup SessionEntity doesnt keep state Why the link between the Session and the entity keeps the state? Username alvaromota +Verified this issue +I included besides the improved renderers default icons default skin entries skinSelector s an image with the CristiSorry i think youll have to provide any new files that you added as separate files and a patch for the modified files please include also each new file as attachments here and their corresponding paths and ill check it out and apply it afterwards +Yes you can close it +Why use the system temp folder and not the databases? The databases temp folder will be cleaned up on +This update ensures that on shutdown they are removed so should prevent any further problems +Insure Job Process Init and Run Times have end time when Job Process is abandoned +GreatIll see if I can include some code for the detection of the JRE in the Windows installerThe user would then be able to pin the application to the taskbar once it is fully initialized +It should be fixed in trunk nowCould you try it out? +Pretty efficient +Patch with the change +Committed into trunk rev +WDYT +With I fixed a warning that had resulted from my original patch +Sounds like a bug +The problem was just that I couldnt extract the bean definition from directly +Rodney could you submit a pull request via? This will allow for easier review and preserve author attribution + +I like the approach but wondering whether there is any impact on the writing speed +So no point in letting him continue thinking everything is finePatch attached to do that +Want to make one for the branch +Resolved as still to be documented in the Wiki +Patch is ok imho +Closing all issues +FernandoCustomers have deployed client applications throughout their enterprise +Verified the issue still exists in need a bigger cluster to reproduce +Ive lowered it to debug +Attached patch that sets the OPTIONWORKSPACEMANAGEMENTSUPPORTED to true only if no value is provided in the configuration +committed +Committed revision +Uploaded a new patch to address Bikas comments +Thanks for the review Andrew + is already used by itself so I used +This is a significant defect for the AWS API feature +If this is off then priorities of jobs are not used by the scheduler +will require cluster shutdown and restart but thereafter we should be able to upgrade running hbase across major versions +I set the to be minutes to wait for the repair to finish This time however the actually have the old values as well +I agree with you about being careful for the race at startup +Roger I will be glad to help +fnx file outside of the CFS file +for this patch +Attaching a patch siteb that shows that change +Done +triaged for marking high value +I thought of getting rid of Dispatcher too but then I thought that this change might be to bigA minor nitpick Id change the Task subclasses such as or Query to use the super constructor instead of using the constructor with parameter so that the subclasses dont need to be aware of DEFAULTEXECUTORSERVICE at all +If someone doesnt know some complex math problem but sees that they wrote somewhere in it then it is wrongUnlike some I do not pretend to know everything which is exactly why I asked the quesion in the first place +the version of versions after September or so should allow you to upgrade to a new upgrade +Verified +Let me try +A patch to fix the problem and a workaround by closing stale resolved issues +The current names areThese variables are going to be commented out in so by default they will not be configured +The real cause for NPE of store in is that there are bean definitions for with the different bean id which is also not default bean id name based on the classname +The revised fix packages the dependencies as part of the travel sample and avoids the need for environment variables or separate downloadsBy default only the dependent jars that arent already part of the Tuscany SCA binary distribution are packaged with the travel sample binary distribution +I dont know if it currently configurable in Thrift to change from static dynamic but please dont force dynamic dependencies +Patch to fix dependency for hadoop exclusion for hadoop common jar dependency in +Should the need arise it could be reopned. +This problem seems to be present in +Infact this problem would have introduced data loss easily in simple RS going down scenario +This limitation simplifies a lot when reasoning about the order across storefiles and client formats to make it as fast early outs and simple list instead of map as possible +He is trying to create a simple test case for in order to develop a patch for it +There is a problem with Tomcat because it sets the http encoding header to and cocoon serializer does not reset it to correct value +It would also be the start of making interceptors more public something Ive also thought some about but hadnt yet decided if this was a good or useful thing to do +Please find patch attachedThx +The receipt of ANY command from the client during that interval SHOULD suffice to reset the autologout timerSo minutes is the smallest timeout we MUST use +Provide an inclusion facility for ivy files to semantically includemerge the contents of some other ivy file similar as include for the settings files +Since I worked with this code a while ago I go with what I rememberThe relevant layout managers are line and text +Spoke with JD on this +Testing in progress +At the very least that lexical analysis would be needed to determine the overal structure and identify the tokens which are aliases +The now parses the path and appends any queryParameters to the parameterMap +I just reproduced it with the same project upgraded to m misunderstanding something +One micro change Updating s AM memory value as a new default entry was added via recently +Hi RishiThanks for this suggestion +This property hasthe system just scan through all the log records without actually applying them and then writing debug info about them into the +Please no error messages in the console +use svn diff and perhaps the option to make the patch more readable +Proposed patch. +Ah my mistake it is I misread it as +Will not get to for the release +I think our private repos are missing them +If we are not supposed to use beanName property it should be protected from framework user +For the Hibernate ORM integration repos I assume that would be a single repo with a branch for each version +Thanks +It is now created before the project is defined +pending Jenkins +Verified at +The mirror is now available at gitWe can make the contribution folder available as a separate repository or you can move it into the main repository by moving the folder either to branches as a custom branch or to trunk as a subtree of the main Lenya codebase +but Ill be offline for next hours or so +One of the hints to deal with it is to use nested tables +No need for +Unmarking patch available its a year old. +Allowing Hadoop local execution mode will be very beneficial for testing with a subset of data before going into production +It looks like the updating that the Grid does is a shortcut feature but can cause data corruption +Thanks Rajat + +YesNormally the response is set on the IN message if its NOT out capableif OUT capable set response OUTelse set response INBut in the world of WS you can have which is a bit special WS +a good attempt at defining terminology has been done and if theres a question we should address it as it arises. +Luckally I could use the attributes of the ant task in stead of the nested elements as a workaround + +Thanks very much Leticia and Jason +This is related to all documentation related issues to final since they dont represent a change to the codebase +You are welcome thanks! +Changed logic of method to check if the collection or one or more given types are equal to or derived from the primary type or mixins of the Event node +FixedWhats left is testing and tagging +Added OBR descriptor in Rev +even guess that would be fine +Quick update I havent had time to progress these issues because Ive been working at a customer site but I hope to devote some cycles when Im back in Malaysia next week +The thing that it fixed pom projects is nice returning none was a really good idea of the maven devs I changed the patch a little for back then since I didnt know that +I added a patch that were trying for Grails +will do more rigorous manual or are basically the same thing +is this expected +Thanks for the patches Dag +But at this point since weve never ever gotten everyone to agree on all the terminology perhaps Ill just wait for someone to put up a patchAs Mark says good luck +This should do it nowSo you may close it now if it works for you +I have removed that report again before making this patch patch is for +This ended up causing data corruption for our customers + +For example imagine a applicationpage has link to pagepage has link to page and backwards link to page +Im resolving this issue +it is fixed in it wont in its just a sample +Created a patch where protobuf classes are defined +Another approach to this fix would have been us have an initialization stage and make sure that CFS creation and statistics clean up happen on it +Making this explicit for CAS feels like a Good Thing to me + to replace iterator with for point AndrewI had changed this for entire pack in +As of today this is not possible +There is no dependency +Attached is a patch that adds some functionality based around the embedders callback and against the HEAD after the migration to can you please create another patch and this time include only relevant changes and do not reformat any code +I think the first step should be to review the +The summary line misnamed the test as numberformat it was numberingStill fails latest retest +The terminology is certainly different but Im not entirely sure if youre not recreating hedwig hubs +I cant do anything else +I havent kept up with the state of this ticket to be honest +A restart of Eclipse fixed it You can close this Ill reopen it when I have more information +Considering this issue fixed. +Talked with Enis +IMO we should get rid of MB mouse handlers +After all there is a lot of talk about whitespace handling in the spec and my understanding was that this is all dealt with in and so any whitespace left after that is intentional by the user and by preserving it FOP is conforming with the specOn your second point Im afraid I still dont follow completely +If dropped columns are causing problems somehow I would prefer to fix whatever is breaking +Some trackers that join early might get a task which is running on a tracker that has not yet joined +is bundled with the distribution but its really a separateproduct +Thanks +Workaround is to delete the build definition directly in the database +Had toremove the file from CLASSPATH to get a match. +Ive opened an helpdesk ticket +The XML namespace for Spring Security was unfairly favoring standard configurations by being very concise in those cases +I see both slowness on restart and the exception +It was by design we made SOAP as the visible api +This has converged to a solid solution +Thanks David! +Would suggest not setting them though since that will only add to the size of the data being localized per task. +Will add more in subsequent tests +Looks like the frontend is not checking upper bounds on generic instantiation +You are not sure what fixed the problem and you present a spaghetti list of how we start up clusters in our test util to conclude a little extra manipulation is needed +I dont see anything wrong with your markup or the way things are handled by FF or Safari +Stack IMHO it really depends on what the coprocessor wants +please attach errors from the jetty logs when it tried to undeploy the existing there are no errors whatsoever when undeploying or redeploying but only when opening the page in the browserI did some more testing and I found that the problem only happens when markup inheritance is involvedIf desired I can attach a sample application for reproducing the problemAnyway I dont think that merely asking for more information is reason enough to mark the issue as resolved +So a more complete approach would be to grant a connect resolve permission to the ldap host and most likely to the dns server host as well +Aha Im not an encrypting expert but that seems to be the showstopper +In all retries this add a max of seconds +For me its ok using Jdk again for your help. +Ok it will export now +Thanks JasonI put this into trunk and +combined patch for this issue and is further discussed in. +Can you think of an easy way to strip some of the modules from the package for download process +Closing this issue as it was reopened over a year ago and has seen no forward movement since +How is this solved? I want to use it in jbossGunnar +Are you working on a patch +I realize that this conclusion does not help with the memcache issue but I really think there is nothing we can do here +Although this is an unsupported module it is included in Dall builds such as on Hudson +It appears to me that the test runs the compatibility suite in the vacuous configuration where the client and server are both at the current rev level +Considering other Messages might be referencing the same instance the potential repercussions seem too dangerous. +Admittedly I havent looked at the code in detail but wouldnt the same logicapply when setting the base URL programatically +Thanks +branch. +Thanks. +rename scripts to bin add a binpig bash script add a conf folder add add a that allows to configure cluster brief logging etc +Can we change this to configdir instead +New patch fixing a last minute issueIt supersedes adapted the patch to applied and tested it against a local with more Bruno Can you please post the latest patch for review +Please commit latest release so I can continue with the setup +jar thats not an Apache product should be accompanied by a +Place this file insrctestcomopensymphonywebworkviewsjspui content +Looks good Carl go ahead and commit +against any product that has been +Maybe because of the sortingAnyway you can change the limit by setting the system property +Closing issues after Configuration Admin release +So do we want this fix in the branch? Also does anyone see an issue with this proposed fix +bq +TimThanks the problem is not reproduced any more +Retry against hadoopqa +I know that as soon as I rolled it back everything was ok both on my machine and the hudson one and the error was the same as I saw when I applied it originally +If you would like to do it yourself you will need to check out the site directory from svn which is just below the pdfbox root on the same level as trunk +closing +patch with pep clean yeah +Can you define a variable for the Avro version number in and reference it in the rev field in? This would be more consistent with the rest of the file +maven build +Integrated to trunkThanks for the patch RenataThanks for the review Stack +Please have a look +Integrated into trunk and +This issue will be resolved when is +I agree my main argument though is that we have a history of doing it unknowingly +e +This workaround is documented in the classI will commit this shortly +Oh and here is the updated patch with deleted version comment +Thanks Paul +Im attempting to turn off activerecord to see if it will start up +How do other folks feel about thisThe feature is great! We just have to be careful to avoid adding unnecessary to the core scanner classes +method is still stripping leading from the context path to determine the application name causing event handling to fail +Ive updated the attached file +Sybase ASE certification will be targeted for the next following unit tests also fail for the same reason Sybase issues to the next CP +I will go ahead and make the changes +I do not know that I have this issue also came from the lab currently we allow in the guided rule editor to enter in any free form text they were hoping to have a bounding box of available inputs be defined so a end user can be guided +Ill create a unit test or two and attach them as a patch tomorrow +will commit if tests pass +After some more thinking I think there are two problems here +Also adds two unit tests to verify that both situations are being handled as expected +Im reading the patchesI have a small comment about occasional tabs in the the patches +and then open target Eclipse will always say that its out of sync with the file with the release and this is still an issue +So far this is a puzzling bugFirst a workaround if you do this you will be fine do +my mod would be to simply allow pointing at an existing config zip file in hdfs to minimize the job start time +The current memstore does not exhibit the require behavior because it is just a flat sequence of with no understanding of row or between a row +Heres a patch to add the desired constructor +However if you just modify the thread loop like you suggest it might be good enough. +I Context Providers instead of doing the writes out again via the so no need for the CXF classes +JoeAs I said OWB not written for loaders +If you look at you can see that somebody meant for the to be pluggable +Im not seeing that error in my environment +Hey DougThe most recent patch seems to write both the length of the block in number of entries as well as bytes +it logs Finished reading which means its successfully closed the reader so theres nothing in Cassandra blocking the delete +I am using Fairscheduler and submitted jobs in second job got failed with this other jobs passed +Sure Niall please provide us your patch and well review it +A patch that makes these changes against +The only exception I know of is the one I discussed above namely the names of parameter classes used in RPC +It can be committed imho +rev provides initial support for this +filters are added dynamically +Itll need to lose the rest of the analyzer dependency and have more test cases in addition to any other concerns raised in the review +think it would be +I think users expect command line interfaces to be more stable or backward compatible than API +The svn content has no history and now git repo has weird deletion instead of something easy to filter and migrateBefore I create site please restore git and svn repository to contain the right content then I can filter it with property historyYou can assign this issue back to me when that is ready and I can create the right repo +Every time you add a new metric youd have to make sure the time is properly synchronized +Fixed in trunk in revision Thank you for reporting this issue +Please consider the patch +Actually Ill check your doubts about the removed rules first +Hi ThomasGreat contribution thanksI looked at and it uses an MIT License which is compatible with Apache so no problems thereI dont really know enough about Flex to do a good review hopefully someone else here can help out? new version of the Flex connector for wookie is available +Thanks for the patch it will do for you please not hardcode and script names to be able to skip adding plugin jars to the classloader? Maybe accept a command line arg skip insteadI am writing a script which is trying to install several plugins from the local file system and it fails because it cannot delete those plugin jars first. +I agree that looks like an error +We can then load the same terms and measure the comparative write speeds +This may be a bug as it looks like readline may maintain a circular reference +bq +Attaching +Some comments Though it works in most cases it isnt logically correct to expire old token only if a new container comes in or succeeds +Im afraid Im not familiar with this area of Lucene well if I merge two consecutive segments how come I dont merge their doc stores +I was doing some bug fixing yesterday so maybe you got a version before that +Artifacts which where available with CR and are still available with CRbtadminArtifacts which were available with CR but are not in CR were only three blacktie commits between and none of which modified poms +I am so sorry +which version are you using? trunk +You know this whole discussion is telling +Interesting +As far as the original issue goes Ive not been able to repeat it at allFrom the stack trace posted at looks at least for the CPP client side that the client has sent a Message to the broker and is awaiting a response and it looks like there isnt a send set so it just going to wait forever which is part of the flow control mechanism +I imagine that ISO is an alias for ISOFDIS but thought Id ask in case you knew for certain or had a link to the JDK encoding documentation +Do we also want to archive beta documentation since it is being announced to the general public +cant it be done in +potentially a security concern +I tend to be in the Java perspective these days it would have the same problemin other words we cant really control this well +Benefit this patch brings is that its easier understanding what is going on rpcing +They can leave the safety switch set in which case they get the current behavior +What exactly is the problem +NET information +The attached file solves this issue in in SVN +passes Ill now run stress to try and find whether this makes a difference +example patch with test for summary description to reflect broader scope of revision Committed revision Committed revision +Find to and trunk +Needs codetest patch +The modules are the ones Jeremias are talking aboutSo since they were migrated into committers +Cast to is done only when an extension of JDO that is not supported in JPA is needed +If the fix was applied to why is this issue still open? Jeff are you waiting for something before closing it +I just committed this to both trunk and +It sounds right that we should try a different snapshot but according to the log messages you posted it sounds like the problem is that we are not catching +I have no doubt you found a bug but without some kind of project or example or steps to reproduce Im going to reject this one as cannot reproduceIf you do reproduce it please reopen or create new issue with steps to reproduce +patch to add to and trunk +is similar to in the sense that it has an error code per partition and a top level error code just like any other state change request +patrick +I did a quick look through the changes and the tests pass so I am on the change as isAre there any to deprecate the configs from where they reside in the code +However the still would +See if it reduces hangs before closing +More investigation requiredWhen I set to false I found a few tests that were dependent on this being on by default +Currently working on management console for AS +The reporter of the ASF BZ confirmed that this is not reproducible in httpd and thats what were upgrading to in EWS so Im marking this as resolved +Patch attached to fix that +So Im closing this as fixed +push out to +groupid has to be it +Hi Please check the simplest Mapper that I have written it is a work in progress and not tested at all +very good idea lets do it this way +No response from user so. +BrianCould you post an actual patch for this? With a test caseThanks +the patch looks good +Request has been given to BPMC team but will not yet appear in jBPM + +I have just hit this problem +and +I have updated that branch to simply not close the channel for a see there is a pull request for this +SI maps this behaviour to a +Closing issue out as part of JIRA cleanup +Do you have any testscases that are failing +Am I mistaken hereRegards has been held up for long and no one seems to need this feature +HiWillem Thank you for your helpMarshal +Olga please provide class for that case and return issue back to Konstantin +I dont see why it would hit a problem going from bit bit +Anybody willing to start making a build + +Ill look into this issue +Mind adding your test run results in hereQA test result is recorded above +The same applies when storing object Person in embedded mode and trying to read it via REST +Due to previous refactoring of +Closing all issues +Somehow this has slipped through my queue sorry about the delay +What I can see though is the remember me filter correctly detects the remember me cookie and gets the users roles +HiCan you please attach to this bug a short FO file showing the exception? Itmight serve as a testcaseThanksm really sorry but I dont have a short file which may be used toeasily reproduce this problem +Yup works for blocking out the hfiles and +committed in three parts thrift and released issues. + +I have borrowed bits and pieces from other open source projects and have tried to attribute credit to them when I do +create command would not let me create a newproject in an existing directory +Well since this JLS incompatibility is about one compile error LESS I changed JANINO to mimic the behavior of JAVAC and be released as. +If you finish it before we release beta please correct the version +Verified on +Furthermore Ive added a attribute to the entry tag as convenient shortcut alternative to a key element with a ref bean child +I think this may be in the nether region of things that we want to improve but are scared of struts +I run it around times +Verified on both eap and master. +It isnt a full implementation of just enough to enable the correct behavior in CSLMAs for the test failures Ive never been able to get cassandras unit tests to work locally so I always had assumed they were simply ornamental +looks pretty solid +Ted yikes! My patch is on top of hbasetrunk while Jenkins says At revision +Im still not sure what to do here +I tried creating some comments then using the Comments page to approve and them mark them as spam etc +Ill commit this then in a new JIRA move out the common code to util or some place +If Hans can reprodcue the issue with beta then please repoen it. +make autoconfigure script accept hadoopconfdir as argument +This has nothing to dowith fonts or whatever +Ive tried declaring an instance of in the hope I could get at it via the threadlocal but the context is still nullThanksDave +Helpful label +For now you might just want to skip the tests +he fix I committed is final not something intermediary as it deals with the issue appropriatelyNo need for a better approach imo unless someone encounters a problem with the fix itself +When loading the data the user would be responsible for providing data in the expected escaped format +Perfs is almost max with and multiple threads to send test to measure like scenarios is +The issues with class loading and the fallback transformers can be resolved as part of a separate JIRA +But this will probably happen afterthe releaseI suppose the documentation related to this subject is not up to date +Yes +Sorry for the delay I shall be releasing a new version to Maven Central very shortly. +When we have the new unit testing framework in place I will optimize as per Edmonds suggestion +But this was without cleaning up after the previous build +Its not something we need so I didnt look at it when I implemented the patch +Im starting to feel the creep of parsing plugins make their way back into Nutch instead of just jumping over into TikaThe creep so far is just which we were forced to add back because Tika HTML parsing was totally inadequate to our needs +The patch for the existing files +This patch provides the basic default serverview page will be improved upon through the main jira associated with this subtaskAfter applying the patch the following must be donesvn add clientsrcmainwebappviewThanks +The copyright date for the new class need to be Im not sure if this class should be part of the public api should it be part of the published api javadoc or notShould it be inAlso I think the class will get an exception if started with no arguments again goes to if we expect it to be part of the api or not +Applied rest of the changes +Ill clean up the patch and add a test case +Coincidence +Yep this is an open feature request for the autoexport plugin timed exports. +I can contribute an Avro version if that would help + +Your test case is much cleaner than the one in the patchOnly one problem latest build seems to be leaving a couple of stale +Logged In YES useridI committed code to the branch on Oct that no longerhas this problem +can this issue be closed +Sid the patch looks goodMinor nit pls define a constant DEFAULTSCHEDULER or so and set it to CS to be used appropriately +Thanks. + really fixed +When this is fixed the release note should mention that the fix cant resolve past lost space only future updates +Thanks Zheng! +So how do you update the Action message for a new action item? Will Actions with the new enum case be backwards compatible with earlier clients? Or will they throw some kind of exception? If old clients can ignore new enum values then I dont careEdit Fix typo +cancelled and it already works locally was a very small change +Attaching the fix basically moving one line between and +Is this causing problems or are you just concerned that it might +It must be documented that ftpdeploy will work only with maven +Thomas once again can we please drain your brain +Anders Im glad you picked up this issue Here are some other cases There must be a way to keep the plugin strict +Ignore Accumulo patch I uploaded by mistake +In addition in the future if the project needsBTW another one question +Any preferences re naming the new exception +Sorry about the sloppy cut and pastes +Perhaps something like how handles vs specific stuff +Sorry guys but weve finally reverted this change as there were also some side effects on Sonar results +I suppose you want to keep the source of the site in the trunk so that it is easier to include the javadoc as well? I havent thought about that but it makes JulienYep thats what I was thinking +Attaching a revised patch and +Is there an open jira to track the final demise of Mina? Im very keen on this work for similar reasons to Rajith +Horrible! I just say die die +Resolving and rebrewing for +This will be part of the configid changes to make the version usually optional. +We can also add an introduction at the start of the page to explain that the links from the page only work when you have built the Doxygen reference +I have no idea whats happening so I need some help with thatThats it for now +Applied thanks +Of course we can do that +My current patch if anyone wants to review as its a bit long haired in the ClausIll review this +Changed to enable easily using Infinispan cache and cache containers instances +Hi Brian it seems this problem has been introduced with mobileExcluded changes could you please look at it +Doh I really did not see this one +This can result in some completing and some not +AS should probably warn if this occurs but the exception should not be FatalThis is currently blocking us from getting apps deploying on AS +and thus I do believe this is a bug +also committed updated testsuiteThis change really breaks stuff +Nice job Ricky! Way to work the process. +done updated seamgen eap on last GA build +this is actually a new isue should be separate not related to capabilities +Im mainly waiting on a public beta for +I forgot to mentioned that using a public field is not a good coding style +I refactored the code in part ofThanks! That makes sense +If we plan to share the code separate branch for does not make sense either +Thanks has always been only for testing +I quickly showed the app to one of our accessibility engineers today +This is an old bug +I agree +will not be available since thirdparty client will not be able to include the extra data needed for this on the server side +The change has been to branch. +Could you provide a patch +Now adding jobs is deferred +As indicated the clean build is a workaround for any eventual remaining sync applied on did we end up saying this patch should wait or was it applied on both +One uses a Decision object +added link to pull. +I mean really beat it up +added unit tested with and without security +Attaching the console logs from all of the nodes in the configuration file for MR guess the logs were too much it looks like the job was interrupted after hours and it wasnt even finished with inserting the initial data +Thanks Raghu! +It means passing the instance manager to the super class constructor replacing field members access by set and get method calls setting the delegation field correctly computing the manipulation metadata computing the handler set by analyzing the requirement defining overriding rules configuring the handlersClement +If the problem persists please reopen +I think that the problem is a problem of versions + + for the patch +Committed to +I am attaching here Basic structure of and +With the JBDSSOA split in UI this will be an interesting issue to fix +I still didnt get it to work even though the is fixed +It doesnt work +Resolving as duplicate we can start over in . +Fixed as described in SVN. +Its been covered by thanksTim +committed to with the release. +Upgrade download urls from to thanks. +In general it seems like a lot changed and everyone is chasing after it + patch +Any idea about the issue? thanks very muchThere are test failures in systestsjaxws with this patch +Ill run some sanity checks and commit itOne final nitSince is static I guess its better if we replace int buffers with int buffers +Could you add this to branchthanksj +This is because the redistributor is treated just like a consumer so a message group ends up pinned to it and causes messages just to be resent to the fixed this so that grouped messages are no longer redistributed +After fixing its not reproducible on windows either +Then I could understand why it is necessary +thanks Rakesh for help +Attaching patch for branch for review +We seem to ignore from source folders too +the s test suite needs to include something that verifies that you can authenticate this way too +I have the same issue with so forcing it to use a classpath URL in seems to work for the webstart case +I had to force adding the spi package and not the has that package included out of the box. +If you want to continue work on this patch that would be great +Ooops should have left open for Marshall +is the change by updating the cql to UPDATE USING TTL DEFAULTVALUE in class a correct fix +Issue closed with release. +It doesnt include any new unit tests since the code path is covered by all the existing tests and its just an optimization +The deny list suggestion was intended for use in the case where the spam from that account does not stop +All JPA annotations tests pass now. +all of the files localized by the TT are written into systemjobid After job localization is done all files under systemjobiduserfiles are moved to usersjobid to be consumed by the users task and so owned by the user +Do you have a particular in mind +Are the logs you attached for the same +and all hdfsproxy unit tests passed +DEBUG log output +The submission form completed as far as possible for Martin to review and commit along with the signed software grant signed by Patrick Lightbody as managing director of codebase as of revision to be donated along with matching MD checksum and signatures +I dont see any annotation scanning in that classThe javadoc for indicates that if you include an annotation class youll get all the classes annotated at the class level with this annotation +to trunk +Java DSL works fineWe will have to revisit XML as startup ordering play a trick here with Spring Framework. +Please confirm and close this issue +This patch fixes all issues with a much more elegant solution we can just set the f and f values right away +Just a friendly nudge +Sorry I mean Aries JNDI requires Aries Util +Just noticed that JIRA converted n in parentheses to a symbol +Closing issues +fixed increased to row count so it does not over write itself on display. +Since this is expected behavior close this defect +Disabled unit tests for now in Rev +It also provides a utility method to update framework Counters that arent defined in +Could you provide a fix also for that version? Thanks. +Reopen to fix docs +Hi didnt see it until now just forwarded it to aleks +class files the jar containing downloaded sources the jar containing javadoc Maybe the contents of the mclassifier attribute should be appended to the artifact names +Sending assemblySending assemblysrcmainAdding nmrauditAdding nmrauditAdding nmrauditsrcAdding nmrauditsrcmainAdding nmrauditsrcmainjavaAdding nmrauditsrcmainjavaorgAdding nmrauditsrcmainjavaorgapacheAdding nmrauditsrcmainjavaorgapacheservicemixAdding nmrauditsrcmainjavaorgapacheservicemixjbiAdding nmrauditsrcmainjavaorgapacheservicemixjbiauditAdding nmrauditsrcmainjavaorgapacheservicemixjbiauditfileAdding nmrauditsrcmainjavaorgapacheservicemixjbiauditjdbcAdding nmrauditsrcmainjavaorgapacheservicemixjbiauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditfileAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditfileAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditfileAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditjdbcAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditjdbcAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditjdbcAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditluceneAdding nmrauditsrcmainjavaorgapacheservicemixnmrauditAdding nmrauditsrcmainresourcesAdding nmrauditsrctestAdding nmrauditsrctestjavaAdding nmrauditsrctestjavaorgAdding nmrauditsrctestjavaorgapacheAdding nmrauditsrctestjavaorgapacheservicemixAdding nmrauditsrctestjavaorgapacheservicemixjbiAdding nmrauditsrctestjavaorgapacheservicemixjbiauditAdding nmrauditsrctestjavaorgapacheservicemixjbiauditfileAdding nmrauditsrctestjavaorgapacheservicemixjbiauditjdbcAdding nmrauditsrctestjavaorgapacheservicemixnmrAdding nmrauditsrctestjavaorgapacheservicemixnmrauditAdding nmrauditsrctestjavaorgapacheservicemixnmrauditAdding nmrauditsrctestjavaorgapacheservicemixnmrauditfileAdding nmrauditsrctestjavaorgapacheservicemixnmrauditfileAdding nmrauditsrctestjavaorgapacheservicemixnmrauditjdbcAdding nmrauditsrctestjavaorgapacheservicemixnmrauditjdbcAdding nmrauditsrctestresourcesAdding nmrauditsrctestresourcesAdding nmrauditsrctestresourcesSending nmrSending Transmitting file dataCommitted revision +See for further details. +ability to select your pipeline in the http request no change to that functionality +I plan on committing shortly +Can you try latest srcBTW the error sounds like a compiler error so maybe it is a problem with gcc +Even if you goof with the filters you always get compile +I renamed the +When it isnt a property that is being accessed then iBatis does arbitrarily insert the whitespace. +Hopefully it will not happen in Jive Lets test it when will be upgraded to this version and issue if it would be neededThanksLibor +Comment from Raghu and comment from Zheng are not yet incorporatedThe new patch overhauls things a bit to support Partition level lineage and does this in a post execute hook +button +To describe that more fully not deadlock just really long pauses no cpu or harddrive usage by either hdfs processes or solr for a long time threads hanging out in socket waits of some kind it seemedThat is how I first saw the slowdown with the blur fix I was running one of the tests on my mac and it took min instead of seconds +The deadlock is indeed detected and a victim is chosen and woken up +Marking as closed. +Attached image showing blip padding in the Wave In A Box client. +ah point taken +The runaway test process on was caused by a instance started by the integration tests failing to shutdown properly +I agree that backup tool is more involved and is over weight for this purpose +Thank you as wellIs there any chance this could be fixed for maybe? Otherwise is there a scheduled date for a final release +Marcus Holl posted a +Incorporated fixes for those in an updated patch +Thanks Navis +Patch includes unit tests +Francois Can you please include userstatistics table contents and logs +The cost can be extremely reduced in some queries when we can reduce the number of tuples needed to reconstruct +Only schema but Ive added that option anywayAlso switched back to the database owner to connect as sonarproductionowner +Thanks for the hint +Patch sent to me by email +I would prefer to stay with the original assumption of is the only sink in nested plan +Ive put the v in the review boards a few days ago awaiting feedback +I will see if there is a way around sledgehammer approach is to forbid use of at all as target names +Attachment has been added with description Patch to improve font definitely fixes but I havent looked closely at yet +Im running it through tests now +This issues has been resolved as part of changes made to resolve issue. +Again Im sorry on making incomplete dump +I guess it is a problem with the calculation of the width of the single characters or the spacing between them +Oops +revision Thanks Yonik and Mark +If someone could assign it to me that would be great +This was meant only as a preview of what I intend to do +I strongly believe we should notfall into the same trap here +Whats the status of this shall we rebase it and get it in +Minor mods to make work +fixed in revision Alexei please check that it fully resolves the is ok thank you Mikhail +No problem Francesco the wording was more hoping that those who follow will also get the idea nothing against yourselfYou will receive more detailed help from the mailing list rather than here which is Infrastructure type stuff. +So I went ahead and made the change +Gotcha +push to RTF parser doesnt yet support embedded documents so this would be a new featureSee the and classes in inside the components for the place where something like this should be implemented. +Any updates on this +The default is plain as it was before +There are two intentions and unless a sufficiently acceptable value for stale can be found that indicates both then Id say two parameters is probably bestChrisagain While I agree that staleok was never intended to serve as a way to delay index updates I wouldnt be surprised if quite a few people are using it to schedule index updates for traffic times +There are also multiple LICENSE and NOTICE files under the but I was less concerned about those as I cant see those causing any potential technical issues +Patch against added +I could only understand this deviation in behavior if an alternate scheduler service was will be experimenting with the to see if I can reproduce +Yahoo version of the patch +painful but once setup perhaps rewarding +Streaming tests do not pick from srctest +Add a fail fast option that sets e so that we exit immediately onAdded timestamp to this scripts emissionsMinor cleanup of log messages +Ill redo it when I have a chance. +IMO testing that a instance different than the singleton is used and it is the same by both the and the is enough and avoids adding secs to the testcases +Maybe that caused your failures? Anyway the test is still failing now +Ive removed the unused nodetype in rev +In this case the plugin is removed and swapped with a different one otherwise it is left alone +Already compiled and tested it before supplying the patch +could automatically register its dependencies for example +No luck investigating this today +None whatsoever as there was no reason for it to be a subclass of in the first place +hopefully this is not an issue since we just reviewed that +One other thing +Applied with updated to the test suite and the site +One possible workaround before BCEL supports Local Variable type Tablesis to simply remove the table +rolled into dev. +It could also be supported as a mixin later +Assume closed as resolved and released +If the aforementioned simple Groovy is at fault then I would understand that the fault is mineYour thoughts would be appreciated +Any tried this in +Ive just uploaded a new file with accompanying checksum filesKeeping my fingers crossed. +Image added thanks. +Hi sorry I really didnt have much time to work on this latelyBut thats true we still need to figure out how to apply the patch against there are some extra work to do to make it work with +Yes it is +agree this would be really handy when it comes to running an appl on different environments +Anyone got good ideas for this oneThe idea is to have a simple pattern to set in the URI configuration or as a option on theMost need eg to move files into a backup folder with some kind of pattern usually a date pattern +For example SSHA with a random salt per user is preferable over an MD of userpasswordrealm +Creating the right abstractionAPI will require significant changes and we are not attempting to do that right now +See the discussion to the associated security design form topic as jira should not be where discussion happens +Applied the correct version +Im a bit torn but can we see a patch with that option? Also do you really need this in the client side or is it best to just remove the uncaught exception handler completely from that part of the code +It fixes this issue and doesnt seem to break anythingIs there a better way to fix thisThanksdo you have a test case that can reproduce this issue +This should be handled when we rewrite the to add JSR objects +So please ask Rajesh or someone who has the authority to set that +Cannot reproduce with CARGO packager working properly for. +The is CP will be can confirm that is affected by the issue +This doesnt seem to have much to do with the use of +Pls add some documentation in around the skip feature +Adding updated patch to resolve the build problem in SMX features. +chain task validator Eugen +m directory +Just released a new version with the fixes in the +Reopened at Chriss ChrisI will certainly have to work on this any update? updates for this +esb archive +trivial patch to add logging + +I like this though maybe we could omit printing the statusstate key all the time and put that in the help instead +provides useful info for debugging misconfigured servers +Test cases are checked into trunk and +SankaI saw the possibility to set the hostname parameter +Please try again was able to replicate and then fix the issue seen +Looking over this now +Is there a way to get it to run again? Very confused +Closing all resolved tickets from or older +No worries I certainly didnt have to +test to more reliably reproduce CME during range +One of them appears to be a bit of a race but I dont really know for sure +Great! Thanks Sharath for looking into thisR +Hive and projects both asked for their svn and git repos to be added to which they have and are using it why are you introducing yet another code review tool +r +I suppose this can be closed +Awesome thanks stack +This is Sun bug +I didnt see the problem because I didnt run all tests before committing +This was removed some time ago + + changed for HQL nested news support +Youre correct but the stand alone test case I provided you has the same result wheter the hints are set or Ill try +Proposed patch implementing the suggestion in the attached patch applied the output from the Lynx trace gives grep homeyyzthauser texthtmlcharsetHTMIME PICKED UP texthtmlcharsetHTMIME Extended MIME is texthtmlcharsetHTMIME MIME is texthtml converting to wwwpresent texthtmlcharsetHTMIME PICKED UP texthtmlcharsetHTMIME Extended MIME is texthtmlcharsetHTMIME MIME is texthtml converting to wwwpresentNote that there is no longer any duplication +To do this needed to +Edward looks like Yoxos didnt make the proper distro could you try using plain Tools or Developer Studio +One could get Avro working by just passing suitable implementation for which reflection is usefulThen please make this change or remove the Avro code +Committed revision in sandboxJason please verify the patch applied as expected as I had to svn addrm some of the files +batch transition resolvedfixed bugs to closedfixed +Looks like this issues been starving for attentionWhats the status on this patch? Are we waiting for +Because its how we get the tarball and makes it portable to debian packaging and rpmI will go and fix the patch against Romans idea + fruit for +Thanks for updating your patch committed +Is this an expected behavior or there will be a fix for this bugBecause there is no row on that table and this is a single node setup +I agree the solution in comment solves the problem elegantly + +Verified. +Renamed to +Doc patch +on cr for review but assumes it is to go for future version since its a big change +Committed revision +definitions can override the base ones so simply copying from the base definitions to the definitions map is incorrect +Mikeanother thing I just noticed is your new testcase doesnt remove the directory test at the end +So Im now leaning toward moving the current uimaj project pom up one level in the SVN tree and therefore also in the default checkout tree +Dont know whats your opinion +ear and the +If you look at the current file in our svn area youll see the presence of two jdom dependencies one appears to be for tests and the other I assume for runtime configuration +A test case showing the behaviour of javaimport with and without the Rails Objectclasseval definition +Also make sure that you have the on the projects never mind +Fixed revision branches +Added comments to the description indicating which behavior is expected +I should be able to make a reasonable guess as long as the argument lengths are correct + +Attaching patches for and trunk +Thanks +I agree comments are definitely important for this changeI meant to add them before but forgot Thanks for catching thisAttached is a revised patch the only change is improved commentsin andIm intending to run a complete set of tests verify that javadocbuilds cleanly and then submit this patch +bulk close for the release +I quote from docs To stop replication of a database connect to the master database using the stopMastertrue connection URL attribute +There are no distinctions +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I tried to start replication and then shutdown the master database but did not get a hang +but I think its fine for now as it is +Also how is this intermittently not delivered? If its being delivered but for some reason without XA its still being delivered +The generated thrift Java interfaces arent really clear that some ints are actually enums +it related to? Looks like one is a duplicate of the other +It would be nice to be able to cache the anonymous users but not real users +But on Hadoop a from a committer who is not the patch author has always been required for significant patchesGood to know that +Will check in after release of. +I like Colins suggestion of prepending IS to the OS flags in Shell +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I mean MD is MD always so perhaps this we should only be attempting to use the provider path if the JDK itself doesnt support the digest algorithm we want to use +Actually I think you have to do a full build with xsd and csc on the path before stuff works +I got a new seam project added actions did the hibernate stuff basically did everything in the handson lab and everything incrementally deploys absolutely perfect for me +Also this would make the change to unnecessaryGiven that Im making a bug fix I should also add a test case that catches the bug +Too bad that C and dont play nice out of the box but them is the breaksOnly the first two lines of your code are strictly necessary I thinkThanks again. +Ill make a test tomorrow with the snapshot to see it but if it passes all the tests I think it should work +opening to remember to create jira for remaining issues +sintervalperiodg in the new config options to be consistent with the checkpoint options +I have created the project and set it up as much of a match as I can +Patch applied +Definite for +The patch seems good as far as I can see +I am completely open for other options +If you tell me what you want changed I can build a new patch. +completes this issue. +You might use the current trunk by building it on your own or you have to wait until the next release which isnt scheduled yet +Ugh lost track of this one +This issue needs more investigation +What does this do with regards to people who havent compiled the binary? It looked like it was being automatically required somewhere though I might have imagined that +You can simply construct a URI using the and reuse that. +I dont like the changes to though +Not sure if this warrants a new bug but constructor addComposited method still accepting arrays rather than varargs +I hope to commit this patch shortly +Hi maybe we could simplify some things here if we were to refine this issue as its described maybe not all tokenstreams but just someMaybe just tokenstreams considered to be more corestable by solr should belong in? I would think word delimiter filter at least maybe html strip and synonyms too would belong in this category +Juhani I just posted a patch + +Suppose you start with a set of WSDL files and want to modify them +I think its highly unlikely that there will be a at this point +Please use the user forum for help. +The sooner the better if you cannot test it before the beginning of next week it is fine with me +Merged into the master branch. +thisd be great to get into +I am looking into have retested it again and I dont see any test failure +. +The thread should have locked monitors and I am investigating why it doesntIt doesnt really execute wait on the acquires monitors of synchronization it was a separate question because I didnt find any mentioning of it in the spec +Setting it up to conditionally include vendor or site dirs would not be too difficult. +Applied patch to Subversion repository +Here is the test mentioned in the description of the issue +Ive applied the patch and atfirst glance it does indeed seem to fix the problem +Sorry about that +WillemYes this one is missing from the shared library +Suresh any update on this +Perhaps did you forget to svn add the file +The advances property has been changed into a write property +I could see a specific issue that can do IO of PMML into Mahouts but I dont see any thing running natively off of PMML. +been following bookkeeper as a WAL destination for Accumulo too +Frank I think I may have misunderstood the serialization format mentioned in the spec +Well need to spin a Joni release for RC +I dont really see anything that points to a bug here since its a Rack mutex thats apparently misbehaving +Can you please build and test the latest SNAPSHOT version +So why hurry the release? +So in order to use the optimistic locking using Version it requires us to set up a repository factory which provide the jpa detach behavior because if not detach we cannot set the version of the entity as it is managed by jpahibernate within course this detach thing is tide to the concept of transaction which in not the case of other persistence framework such as mongo or what ever +This is something wed really like to go forward with for Tools since it would replace a lot of complex conventions were maintaining to deal with our composite build and its multiple there something we can do to speed up the process +If my machine has GB I might want to allow a MB container to fit in between a GB one and a GB one but might not want to allow MB containersThat doesnt sound right +instead of waiting for the file to be moved to done it may be better to block the move if it hasnt already started and respond to user requests fasterYes I agree that would be good but that would require me to make it so that reading the Configuration file and getting the location of the file are things that are done via the so that the Configuration file could not move out from under other parts of the code +Ivan sounds good to me +Assigning back to Darrin the review is done the book sis unchanged and good to go +In practice most likely reason for such a scenario would be a large number of datanodes go missing +bug title to more accurately reflect the nature of the in trunkShould I apply this to +Im posting it now becauseId like feedback on the interface for compression and decompressionTypical intended usage is present in +That mostly takes care of it +During that process the prefix marker with prefix is also skipped +gets the class only from the configuration irrespective of whether it is old job API or the new +xml files in financialsservicedef for timout and replace with the correct value timeout +Updated on th need to use abstracttrue in your modified binding as youd see if you tried compiling the binding +Can backwards compatibility be maintained in Geronimo by detecting the schema version and processing the plugin XML files accordingly +Committed to trunk +It removes the ipc retries that worked independent of retries and pause +Do you use ST flow acegiWhat happen if you dont set the subject property on the exchange but only perfom the login call in the componentI dont really understand who how when the subject is given to theEJB container +Fixed +Verified this issue +Not sure how we can do this in a cleaner way though +Eric i will try to get some time by this weekend and look in to this +Better loggingWhile profiling also noticed an inordinate amount of time being spent in DistinctInitials bag registration fixed thatThe task that I cited as taking seconds with this patch earlier? It now takes seconds +Committed to SVN with minor tweaksMatt Would you like to be credited as a contributor inThanks. +For example a file sink may emit a message when a file is closed or after X number of bytes is written +and returning null wanted to do this but didnt because itll break existing builds +Well newest version of archetypes are displayed by default +Carlos can you apply the patch. +This is expected debugging output +Will look at your patch as well +I also renamed things per uwes advice s does exactly what it says +Or at least of a true leak thats something we can do anything about +Fixed in! +Tagging for family and +Unit test for array expressions in methods and constructors. +Hi KatheyI synced up till about yesterday as svn is down now and applied the p cleanup patch and ran the with sun and it works OK on my win laptop +Attaching here since FB is broken +It verifies the tombstones are added on cleanup +Theres a bunch of javadoc warnings if I include thrift generated files +Ive just verified that this fix works for me follows +trivial patches for both trunk and revision in trunkCommitted revision in branch +Circuit can make use of a separate jar of +The comments indicate that they have a way to customize the serialization per column in which could be interesting but its +I changed in order to address the same behaviour in the embedded driver also +Hi NickThey are the same in too +As I indicated I have found no way to set the name automatically using requestdispatcher and so I was forced to include servlet in controller +The class cannot throw itself away obviously +Sounds good to me +The other comment was about a typo in writ +Resolving since its been committed +Moving it to Future as there is limited time for +Trouble merging can you rebase and review +has anyone tried using IP addresses instead of host names to see if it helps? I didnt think that was related +The guide was only missing the reworked images from the stream +bat files can be left as is with minimal functionality and we can do proper scripting in the powershell scripts. +The missing path to get it production ready is a uniquefile id and a way to access a file with it as nfs IO operations done with filehandles +They only can request a update +xls and +This is now fixed under SVN revision +fixed at the same problem +The common project needs cleaning up +So mving them in base class will make it incompatible with or +Jopr revAdded discovery of apps to handle scenarios where we want a resource in inventory even if its down to begin with +sorry for the false alarm +I think going ahead with and then looking at this patch will make things easier +Keith could you review this change pleaseThanks look reasonable to me. +Thank you +Can you please let me know whether you have succeeded in having an appealing solution to this issue +Hi Bassam thanks! Could I make the request that you attach the changes in patchdiff format? It makes it easier to see what exactly has changed +All these images were taken either by me or by John Kaplan and we both agree to grant the ASF license for them +Set time stamp to current if its can get jobid from file name which is in the meta data of the chunk +Thanks Owen. +Thanks Olivier for merging this in +Ive also observed dismax blow up if the query starts with more than a single dash +What is the criteria for the new logical plan being stable enough? This ticket kind of scares me as a user of who has had to disable the new logical plan to get my scripts to work properly +So this needs to be fixed +bq +Its a txt file despite its +Problem is I cant run with encoding since its broken without the patch +Old habits die hard +YongqiangCan you take a look? We have been testing it for a month in our test cluster +OK +Committed to the trunk as revision +hashtable lookup of clock in clockPhases fails due to equals method returning an incorrect by r +The useCase variable is set to falseI guess there was a typo above useCase Ted corrected the typo + +Thanks. +However your definition took me by surprise and I cannot see any advantage in it +Would also be good to have a separate JIRA for the scheduler changes +Committed at subversion revision +Havnt gotten that far yet Still just doing quick standard micro benches of each +Attached patch for review +smile +A lot of them contain still incomplete patch contains the mdocifications test cases and updated versions for the dependencies to the latest minor version + of a bulk update all resolutions changed to done please review history to original resolution type +What needs to be done to implement this API? Id be willing to take a stab at it if someone could point me in the right direction +Thats already the way it works +Another issue is that the JVM is continuously optimizing your code when you restart the VM you lose that information and it has to be reconstructed in the next run +I committed the patch +There are extensive completed in all the key classes and there is a full set of automated unit tests +h +Slight correction +So it should be fineFor me I put the generated files underShould I put them under too +The issue requires a running client and service to reproduce so that might take a day or two for me to get to. +Splitting it up in the way I proposed might not be fine grained enough though +These keys are currently required by sonar to reload measures that are not kept in memory cache +Its been fixed now. +Thanks David +Is schema synchronization confused? I cant think of any other reason for continuous flush here +Cannot reproduce and have not heard back from the user +Should I just add JDK check in installer and warn about JDK requirement for AS +Spelling fix checked in +The store keeps track of available files therefore unnecessary bundle copies can now be avoided. +It looks like there is now logic to find entry points from inherited modules so wont immediately return false for these cases +As a result the patch its about the size of its previous incarnation +Id like to get IO out soon +Improved patch with double quoting enabled and changelog entries added +Application to other makes sense +Objections to include that in RC +Updated Eclipse patch that can be applied to existing trunk subprojects +Any Apache committer can edit theses pages want to give it a shot +This is definitely an error condition of sorts +We dont have a reproducible case and it will be a while before we can test on We have just identified an NPE in one of our custom functions and I suspect thats the root cause question about Teiids exception logging The stack trace here appears rooted in and therefore I wouldnt normally expect to see a Caused by +I think this patch should be committed as is +Id prefer Davids suggestion as Im only using Java SE on and think keeping Overrides improves the code readability +That should be it +This should be a reasonable threshold because it will only trigger a writer flush when it is inevitable anyway +on the excellent doc Enis +This may or may not be related to autoload +Apparently there are some missing that we need to work around to maintain a single source +Reproducible on branch no more under active development +Thanks +As Ive already said I dont have Mac so I could not test things properly +Committed revision . +Updated patch about ordermgr component +Once thats sorted we should look at removing this limitation. +Changed directory name from links to link Left the and as they are rather than move them into the new directory +I just committed this +Madhumita i expect this to be resolved with ER +This violates the dynamic design principles of Flume +I just wanted to addThe issue doesnt exist on on which version the URLs protocal is jar +META +I just add a patch for this caseIts add a property archetype to the mojo this value will be used by in resolve in addition to default exclude patterns. +I think though that even though incomplete its helps +Reopen to consider the enum based solution and fix the is implemented in another it was reopened +I have enough experience in programming in Java and therefore these two caught my attentionI am interested in being a part of the Gsco and would do in this project of Test and Fix + +Nicholas could you take a look at this? I think needs to be backed out of or should be pulled into to fix the build +Since this is fixed by using a mapping file and its best to stick to established naming standards and Castor shouldnt try to how names should be handled Ill close this as won issues are open recommending that some types are put into a different package than default so those issues will be tracked there +Actually remove from the does not work either +Ive added check for connection in the outbound adapters with attempt to XMPP connection will make an attempt to reconnect on its own so for inbound adapters there is really nothing needs to be done +I agree with Yoniks thoughts here we really want to limit the effect on the overall API since currently its just for +Thanks for the efforts Ahmed +Thanks Donnchadh +I cherry picked the commit and built a custom for +Yes this is a duplicate of so you may close it +Hi JulieThat bug actually is fixed +This is similar to tapjobjobNamestepName Q Can I tap into the source channel for a streamIf I have the stream s http log wouldnt tapstreams be the data coming right out of the source +Patch +follow up on it +That is the first thing I tried but it didnt help +Actually with that change in place I cannot detect a statistically significant slowdown on my home machine +ok upgrade is now complete heres hoping it helps! +Use beehive beehive as userpass +Can you verify this patch before I push itRegards patch has been tested by msimeonMarked as closed. +Hence giraph is still broken for a large number of profiles due to a patch +We can adhere to the language of the specification and add a message but still throw an exception +Build log +I think it makes sense as there should always be a version where you can use the old and new api as this allows a transition period for the projectsAn alternative solution is to copy the classes from into a package in +can you be more specific about where you saw this error? If its happening in then it needs to be do you have a failing test case we could use as a basis for an integration test? This would make it much easier to be sure we got it fixed +Each Receiver has a distributed pipe and a corresponding Receiver connected to the other routers +Logged In YES useridReopen with more details or an example +DavidIt could be compiler issue +if objects are not detached executed are efficient +When an element was created in presenation its family will be set to by default but it should be presentation +Targeted to +orgapachehadoophbasecodec +please +Closing in favor of which describes changes that address this problem but cut a little deeperI think your onJoin problem is separate from snitch activity and needs a separate ticket. +Its not an extension or a feature but much more an information about the fact that the ACI interceptor is active or not +If this is important to you feel free to create a subclass or +This way if users use hibernate part of the spring code he can at least put hibernate as a dependency inside the user project this is similar with hibernate as therere many ways to provide cache for hibernate +Did not use nightly snapshot since I did not find a +If libexec exists it will be used +Yes +Nagendra will you be providing a patch? If not a better home for the above links may be on some Solr Wiki page +Also attaching a copy of the integrated patch on the Jira +Grant! Game plan to get this committed +The patch does not include a fix for this +Fix in place +Patch to fix this issue +I fixed this after +I am actually not sure whether I like any of these ideas sufficiently +Also adds memtableflushqueuesize configuration directive defaulting to +Dust this one off too +Just reopening to hear what you want added to the release notes? I dont see much new and noteworthy about this +Im also running into problems taking advantage of Enums with and Groovy +New topic has appeared in Latest Alpha Manuals so closing. +Will open new issue if fails again. +I dont have time to chase this. +Also I dont see the benefit of storing largish models in XML +If Avro is released this week as expected then this could be trivially changed to generate JSON instead +Thanks a lot for this enhancement +In I made an exclusionTests pass +Still works fine on Safari and Firefox. +Moving out of +If the timeframe on is further away then we need this repository to be manually created +Closing all resolved tickets from or older +With this patch tests are added that test the policies with for a faulty service +Ill just have to live with the extra overhead for the time beingMike +Its clear that do need random access to get to the portions one is interested in +Ive submitted my ICLA and Im waiting for access to CWIKI once I have access I can publish it + isnt actually a test and is something you run manually so you should see the output all that checking would do is generate more output you would still have to look at to know +I have changed the title according to the document New Features Roadmap Draft + +Please verify in ER verify in ER have looked over this issue closely and am confused +In more than years in geo search related companies I never went through a case when one wanted to sort by distance anythingIt shall be very hard to have a understandable UI for such a listFilter by categories other thing and sort by distance afterwards yes +Thank you for such a deep explanation +As with other chunk issues moving into has been completed as part of the major documentation upgrade. +So the lazy instantiation of my treeis of no use now. +sorry for the delay +I agree that HTML versions look goodNikoleta +I dont know about the future plans of +Jessica what do you mean +Warnings in editor due to ng +Stopping the would unbind the from JNDI and when this happened before the shutdown of the the error in the description of this jira was displayed. +description update +patch for code and junit fixed with minor changes at revision thanks a lot for the enhancement +I see counted in to the build but not +Lets just punt on this for now +There is another issue in JIRA that details what I would eventually like to do in terms of this +Thanks for reviewThis adds three unit tests of master failover +Better fix usage points they are not that many and the fix is relatively easy +I just committed this. +Adjusted code as mentioned not as efficient but should always work. +After no app is running with the tokens and the keepalive expires the tokens are cancelledNote that by default I maintained backwards compat in that tokens for oozie jobs setting the false will never be cancelled +I submitted to track this issue further +I have attached find results in PDF view for Accounting Transaction and Accounting Transaction Entries +So all issues not directly assigned to an AS release are being closed +After turning on logging to debug +will commit afte tests! Thanks Syed S +Committed revision in trunk with updated status from Jason. +I can see a Javascript error dont know yet if this is related to the problem is undefinedhttplocalhostwicketresourceLine a simple workaround so that it works in the moment but I am quit sure that the necessary rework for using jquery etc +The problem seems to be in where it checks if moduleContext is null or not but thenm outside the loop it continues using that even if it was null +Original problem identified is now ve also now commited some changes which do a compatibility check on the javascript code during the import process +This has the added benefit that It works if CDI is disabled +There is one new class in testmodel which I will attach next +Can you please restate what issue you are having? I find it difficult to followwhat you put here +Is assuming that configuration of the search component chain is the same between all shards OK +Not sure about the fix here +When the flush is completed we can set the size of the memcache to the number of bytes submitted as updates during the flush +If you installed using other modes such as +Unfortunately I can not speak for all customers but I do know the customers that I have worked with over the years find it a must to connectors +bit later today Im going to check in a change which should substantially reduce memory consumption for stylesheets which use variables +So all issues not directly assigned to an AS release are being closed +I am also getting this error when running the assembly plugin in maven +the patterned used to date is to rename the old entity create a new entity with the new fields or without previous fields the use or java with delegate to migrate the data +Also the fact that needs to be changed to support the new ledger storage indicates that there is a strong coupling between the two +Hey David would you close this if its working for youThanksRich +Ill check them in +Will then append remoting and the callback clients session id +bulk defer issues to +Thank you I will take care of it. +Though I think HOD should still take responsibility of cleaning up what it createsIt seems therefore like a useful addition +Version is attached it removes getting split size as system m fine with leaving that in for now +The tree is obviously down to the new plugins and the splash working points to a configuration difference +All work fine +Attaching Rebased patch on latest trunkReady for review +Fixed in +Ill keep you updated on these changes +The change is available with revision +Closed. +Merged into the branch. +I just committed the trivial fixIve opened to track the findbugs warnings in. +svn commit +I should also comment that I checked that I got same pictures in from using both OL and +See my updated review issues installingusing the package under ubuntu +Ok I misunderstood the spec +Its value is decremented when it is resubmitted in subsequent attempts +I will work on this JIRA today and make the necessary changes +Closing as difference. +ThanksSuren +the patch looks good to meDaryn has Tucu addressed your concern +That patch looks fine +Added a series of tests for standard fault codesThanks to Fred +First we had the EJB and uber war cases and now the cases +hi Angelathis is my patch for this task +First attempt looks good and passed smoke test +A patch that attempts to fix the problem +I did run tests with the change and IBM before checking in +Cant reproduce +I couldnt see other ISIS annotation specific to HTML is it a good idea to call this one? or may be we can choose something more generic +Now there will be two settings for the two separate functionalitiesPlease provide a quickstart in +To mimick the behaviour of the autocomplete component one must explicitly set selectFirsttrueOnce you set selectFisttrue in metamer the behavior you are describing is achieved for autofilltrue +One area that needs a little more thought is the for actions during the evaluation of a consequence +is already commited no worries on this a new patch after the tabs spaces commit +Thanks DIGY +But probably we need a new bug for that since most of the issues have been resolved and we need a new more targeted one for F +Closed upon release of +Resolving after committing it to +Cheers looking at the stacktrace I posted I see the artifact in question is actually using even tho our parent pom defines +Say you run a public facing proxy in cluster mode where you want to communicate in between on private between cluster peers +You guys are discussing the global config topic on the forums +problem is that type all is not supported we will solve for input +Linking this issue to +The minimum the change the better +ASL? Yes +Candidate have submitted the fix +My concern had been about getting into a state where Preconditions was used for some but not all such checking +I dont know of any reason to keep the LDAP groups +There needs to be a push of state as really mitigates this issue +Any thoughtsWhats odd is if I inject the into my test its populated there just not in my converter +brainfart +Fixed +We need this fix for one of our releases that consumes this branch +Thanks! +Closing this record since is now released. +Would seem to be very brittle and hard to make changes +Fixed the issue now creating the healthchecks on basic zone are working + +Unassigning due to inactivity. +What if the feed contains entries from last month and last year thats an area the journal very likely wont cover anymoreA query is one solution but you might also design the content structure so that it is easy to generate the feed via traversal. +When receiving blockReceived from DN checks the length of the new replica If the new replicas length is greater than the default block length or smaller than the current block length mark the new replica as corrupt If the new replicas length is greater than the current block length set the blocks length to be the new replicas length and mark the existing replicas of the block as corruptI believe that most of logic for has already in branch +I didnt add any snapshot repositories to the pom +We also tried to kill the thread before the app exits but got a crash + at +The properties file not found message was displayed as there is no in the same directory as where lives +Use insteadof +Hmm +Anyone working on this these days? Id love to see this in when a table gets out of sync with meta it really would be great to have a tool +Does this test still fail? If not lets close. +Im investigating the issue +Thanks. +This looks good to me +Beyond that because of we update the +Tagetted for RFE is tracked at +Thanks John! +Thanks for the extra information +So each category is a type of event for this ID +DavidgetCause does not work in +Logged In YES useridit is also in jboss Username cazzius +This is the pdf file which I have testedAdam thank you for your testingCould you test with this file too? +I dont think Ill bother to do the comparison until after Ive tried out the approach though +I will go on for these issues +We do not want code committed thats not yet usefulthat leads to bloat and sets a bad precedent for the project +Patch that does this +Fixed in upstream by adding module that has ibm jdk specific classpaths imported +Will address the unit test later in trunk patch of +I will let you know if I need any more info from you on these seg faults +Thats right +Here is the correct patch +Laura this issue has nothing to do with the documentation like the jira says +which groups are duplicated? Screenshots +CSV output format moved to core excel one to extensions +I think we can do better though +Of course these are features so should only be considered for the musta missed something the JIRA is for like REST calls to provide convenient RO access to some of the data that is already available either in the form of basic HTML webpage or via JMX +thanks Nick! +Thank you Prashant one more time! +The eclipse configuration under which this occurs +Im taking this one +never mind about setting the managed property to false will try to avoid that. +Itd be after current filename which is ts IIRC? Or is it random number +The client write the message to tcpmon at port +Yes this dialog did not read from JDT +Please do some code review if you intend to use it since it is the first ever patch that I submit to the GrailsFabien +Im nervous that ignoring exceptions entirely could eliminate a source of information when things are breaking + works if using quickstart warnings are okay but issue in validation is fixed in CVS HEAD +Thanks for all of the help everyone +Hmm +Cannot reproduce with the sample application against head closing. +Needs to run multiple times to reproduce +fixedCommitted revision +Much more important than how far in the future the cache date is what will the uri of the resource be? If every page will have a unique uri for its combined Javascripts the client will suddenly have to download prototype scriptaculous and tapestry for every page instead of just once for the entire applicationI suppose Im just very paranoid about the possible pitfalls and Ill just have to see what you come up with +Updating my previous patch for the maven plugin to resolve a SLFJ runtime exception Failed to load class +To be confirmed in a future jBPM release +Resolving as duplicate of. +Correct the default rollover file size in the logj XML configuration to match code value of the comment to make it clear that the unconfigured default is MB and not MB is good but I think the update to use a MB configuration setting is a making it a bit too large and it should be left at MB +The source is included in the jars themselves +unit + +Committed +Can you fill out the details of this issue and does it still occur on +Will not implement. +Works well JulienCould you just Add the following text on top of the main sonar preferences page General Sonar Settings I guess its not possible to automatically update the standard Problems view as soon as the default severity level is changed +pushed +This is what it look like +You cant delete a directory that is the current directory of a real problem is that clean fails if it cant delete a root dont know about any other handler problems that this patch doesnt fix but I just cant see whats so bad about my suggestion +The dependency from will never be removed since coverage plugins MUST be based on +Heres the zip for the sample +it is reproducible at least against EAP looks like the m not a Messaging expert but for me it looks like is not conforming to what other message implementations are doing +For issues Ive opened the rest of this to move this relevant any more? I found this by accident while investigating a client question +Does that mean that anybody will be able to connect to HDFS as a superuser via fuseIt seamed like a rather simple move of a command from one class to another +Recommend use of the for that reason. +I dont quite follow the overall structure of spring project subdivision and thought this could be related to the classloading of spring dispatcherservlet thereby fitting into the module called Web +It works in the Ibator build but may have some quirks +Makefile was is the patch that makes to call constructors +Alright attached patch will make the following possibleBuild from Source ant toolsstressbinstressInstalled Binary toolsstressbinstressNOTE executing from source file only works on Unix like systems i dont have a machine to test hence left it untouched +Logged In YES useridI have tried Sachas suggestions and now it works +Btw Jurgen I havent been able to patch master with your patch most probably cos your patch has not been updated to latest code +All tests question on this +Lines contain the changes +Kishore thanks for the logs +This has been fixed as part of +Do you mean to say that i need to create entire again in the didnt get what do you mean by use Spring transaction demarcation for the Job Please could you help me in solving course your Job can use the same Hibernate just like any other component in your application + +Latest patch applied in r +No doubt someone else will raise another JIRA if it hasnt beenZoe +I understand that this will require some changes like adding code to the followers to handle connections to observers forwarding INFORM messages to the observers syncing with observers etc +ui quiet a big oneDI will have a look +About the PR we should check all current avaible skin and we should not move the class have updated PR +Done in +if this is true Ill wait for that to reviewapply the patch +All three findbugs reported during the last run are old bugs not introduced by the patchI fixed findbugs in all of themBut the two in related to Write to static field cannot be fixedThis is done intensionally and the warning should be ignoredThe patch is updated to current trunk +Just go to the method I pointed out and read the doc +I tried running the command directly but it didnt work +Closing as Later. +Got rid of javac warnings and uploaded a new patch + +All Sonar schema tables created +passes +I have submitted patches for it so should be fixed +The setter on works +Im not sure we can fix this in since it will make the existing instance data incompatible with the Daos +I attached a build of the patched version to save you and others the hassle though +Hadrian maybe you can take a look at this when you publish the XSD for the Camel release +Must be something in our pom that changed +Would you agree to put this into? Seems a natural fit to add it there +Are you ready to release WML through theStruts site? We set up a package there but no files have been addedyet +Eclipse has no way to find out what has happened on the other side of connection thats why it reports such error +trunk + it just seemed like a higher priority issue with the scope expanded +It does look a worthwhile feature but if it more like an enhancement then it may not be scheduled for right? Or enhancements are still being taken up for +I have created a JIRA for tracking this addition should it become necessary +Actually theres nothing to fix here +the patch looks good to me +Sorry for being late +Added using the lib classifier artifact +I try the test case above and this issue still exist in current rev to because this still occurs. +Fix confirmed +ClausI cant really take credit the scan framework lends itself nicely to extension +The workaround is to do forrest or forrest run in the directory before test +It might then make the entire package structure a little bit better as well since the stax input source would extend +unzip to luni module +defer to +push out to +I am currently working to migrate all our java code to use slfj since that gives a very simple logging interface that we likeThey also have a package to directly redirect code to call slfj which then in turn can call any supported loggerUnfortunately commons configuration class references the implementation package causing a runtime failureI suggest that the commons logging stuff could be replaced with sljf in the first place and make this much easier and even be the one to make the patch if others want this too to get rid of interproject dependencies +Pull request linked +I dont see that you need to use a handler for this +Unless by working you mean allowing an MDB to infinitely attempt to process a broken message maxing out the container and requiring a restart in which case I stand corrected +bq +bumped this issue to to as duplicate to +Sorry for confusion +I guess it is clear now that your Updater was seeing the event of the not its own refresh +As branch has no longer its own forms block but onlythe external linked one this should fix the error both in the branch and the trunkThanks for your bug report and the patch +In certain deployment data node and region server are not on the same machineThe above would pose performance issue +Pull requests solving the issue +I will work on a general clean up patch for transaction handling and put it up for review soon +Im removing that from the fixaffects JBPAPPGACP switched to Messaging. +I had incorrectly assumed that the cumulative patch now includes all four jiras +Committed revision to trunkThanks very much for your contribution Jianbin! +diff made in modules rev +The babysitting process was root cause +Closing. +Thanks! +Sean Sorry for the late response +The failed to resolve artifact error still encountered during releaseprepare. +Updated site with a link to Mina on the front apache page +Repro and some traces per previous comment +Alternative idea we allow the programmatic configuration API to configure instances so the configuration code can pass in a which is already having its dependencies injected +Attaching which adds to the list of branches mentioned on the Derby website +Patch for design review +Well revisit this in the generation possibly formally declaring such attributes as type string expressing the actual target type integer in the form of other metadata that tooling may to mention There is a defaultTimeout property on since Spring which can be used to specify default values across all transaction attributes and which does accept a placeholder valueJuergen +Not sure what changed to disable this +I assume you are talking about the Description tags in the classes +Patch +Thanks for verifying Edward +My application runs perfectly on the older version and on other app + +One normally copies the database from the master to the slave and then boots the slave database with startSlavetrue +Do you have to have long support or is it just a nice to have +Can online docs for CP be updated +We can reopen if patch is updated and still applicable to current codebase. +it is stated in the docs of hibernate entitymanagerannotations what version they work I used the version CR instead of CR +oh sorry ignore my previous comment +dont think we got it cleared out +However there does not need to be a test in srctestjava +So the solution isnt even just for the async JSF request case seem to be experiencing the same problem with a Wicket app running on Spring +Committed to trunk +posponing to beta +In the old patch I had renamed to +My apologies the earlier patch is missing an import and will cause an compilation error +ScottHere is a patch for the C implementation for you to take a look at +It depends on whether we want to try and have a really clean separation of Impl from interface which is impeded by the share package a somewhat constraint I will admit +what do these changes do +It really is ms base pause time +finally! build test failed is not related to this patch +defer all X issues to + is not the default repo +You would either replace it with a new Event Injection point or a new Qualified event from the which now omits this information +Pls commit + +I think in all these cases we can use at most RECORD compression or NONE +All you have to change is the property which is the base directory where stores its data inIt defaults to hbase +Added delete were made to the tck tests to check results of delete operations +Adding a test case +closing the issue is logged in +v lgtm +Committed revision Thanks Mark! +Cassandra issue has been opened asAny ideas on the longevity issue with this one? I think that as per the rest of our dependencies we will just stick with the manual task if upgrading at interim stagesIm closing this one off for the time being +How should these be installed? Only the Release? Both in separate directories? Change the DLL names to reflect debugrelease as the cpp dlls do +Deisgn document is attached +Max Andersen figured that little trick outAt this point we dont have a way to do the library flip flip automatically but this is a more consistent workaround in the meantime +Geronimo is currently using for JSF support so as a Geronimo committer I can work with you on defining and testing this interface +and the producer will reconnectYou may get exceptions and have to retry your transaction what is expected as documented +Currently the type system is too simplistic to allow for additional facets. +If you just apply the changes it works +We need to decide whether use tuple of tuples vs map for Cassandra map type +this will have to wait until +You can see it in the intermediate format output +fix for after is going to just copy all of lite renderer when it is ready. +closing with release +Forgot to grant committed to Flume codelineThanks Juhani! +duplicates this issue but closing this one as better describes the issue. +No rebuild doesnt help +It is a habit to include a backreference to the origin of the change that has been pointed out not applicable to this project +There is a checkbox to mark when resolving the issue and where the changelog line goes in is different +oh I realized I didnt explicitly say +Of course! How simple and how ignorant on my part +Patch to add output to file option to the scan and config commands in the CLI +Initialization would fail on differences + patch looks good +Any chance of a test case +Hard to see how it was missed +It calls +Layer level security allows to setup a per workspace authentication +The only thing approaching use of this directory is some configuration inWe could either remove logs or add a README in it that tells people how to set up logging to go to that directory instead of stdoutBut that could just as easily go in the README in example as well along with other info on stuff you may want to change to go into production +Looks very much like a String COW bug in join +Of course this brings some nasty reflection mechanics as a side effect difference to your implementation is that the patched plugin doesnt leave the classloading hassles to the +By default we should everything +I added a follow up issue +please note that new servicemethod is always added at the bottom of the file though with exception like logical or related methods are added togetherI have also done this change +I havent worked on the filter dispatcher so Im going to leave the patch out here for someone else more we have a patch Im setting to the next release +Patch including change unit test +save a bit of headaches +The problem is that nsapiredirect is poorly maintained upstreamThe best would be to file ASF bugzilla issue upstreamWe might get faster response to that issue +Apache resolver library found resolver will be usedValidatingTry using a fresh seed siteIs this running forrest direct from the or are you calling the Ant task from elsewhereIt seems that it is not finding the resolver jar from FORRESTHOMEtoolsantlib +See you have to hold an instance variable in the decorator. +Whats the latest status did you get past these build errors +I actually recommend that we do a full config review and just change not only things that are poorly capitalized but also things which have inconsistent naming conventions or that are just confusing +Like other links something like targetblankDoes that help +Conceptually a snapshot is composed of hardlinks with COW semantics for files metadata and last partial block +Checking this in +Moved to wishlist +bravo! lets release it +Sending trunkmainsrcmainjavaorgapacheservicemixkernelmainTransmitting file dataCommitted revision . +On the other hand customers will have a chance to learn something new all the time +Fixed and tested in M through the classpath support in the IO module. +Im more than willing to do but how? you mean rebuild the project from SVN? Should I use mavenArentt there some nightlies +The link works you need to log on to the RH VPN to access it + +Patch for the preferences view provided by Fred +Moved to CR for planning purposes +Please see the fix for +HelloAny update on new release +It still makes sense to me that the constructor with the string argument should be used for xml resources stored in files +v that uses hadoop as default couldnt reproduce failure on MacWill investigate more +This issue is complimentary to +is the query plan for the query that returns row correctly I believe +I assume you are taking this one on +Closing tickets +Ill upload a scenario to explain what the problem is +I dont get why we see different results can think of a way to ensure is always added by the script though +Bulk update to fix version +Approved for SOA CP +Committed seans patch. +and orgapacheivy +Using it would mean correcting package structures the application +It seems related to this issue but I cant see whetherhow issue will solve this one +Assigning to Rob to commentreview +Thanks Alexei! Everything works as expected +Looks like the same problem applies to the record Tomcat has been released recently +Unfortunate that existing Setter unit tests did not catch this +The global catch all for the forums project +on commit +EWSRC TC for RHEL +The INFO level was just a temporary workaround for us +On we accidentally nuked hudson +Correct way to do this would be to first check the sizes of two lists if they are equal iterate through both lists and check equality of items at the same index in two list +Based on patch size you just became the largest contributor to Hadoop +applied +We can not accept patches without this licenseThe general rule with JIRA is not to attach files while creating the ticket +Attaching the patch to fix the issue +Personally I do not understand why a serious project would stay out of securityIts simply the current stance of the project +change reporter +I think this is what you need +What kind of error do you get? That sequence should be a if no can be found since remove will simply return null and close skips the close call when given a null reference +That matches up with my expectation and understanding of the problem and it should mean that the issue is resolve +Verified that the logs from the second job showed up in the UI and that the logs from the first job didnt +Although CR is available and should be used for validation +What use case is this fixing? If most operations complete within time X but a few need time X Y then why not just make the global timeout X Y +resolving this specific issue since it works as be separate issue on having a way to clear the usernamepassword without having to finish the full wizard. +Ive undeployed and redeployed test applications numerous times now and cant reproduce this error +ThanksI have a test that can parse the modified log +You need to use a cpptasks built from its SVN HEAD +Fixed also in branch +Closing stale resolved issues +The bea docs lists a required version +Reopened for Release Notes workflow +Moving all issues to the CR +This patch includes the tests that Pat wrote and the fix for the problemThe fix involves three things The leader creates a that contains a reference to the doing the sync +I have this working but getting it into the build is going to be a pain +so this is a feature request +This seems to be identical to how we currently handle all of our other external dependencies +DavidI can go ahead and fix this and commit if you are comfortable with it +Some cast warnings are from the generated code of javacc which is out of our control +This happens with wars as well +Can you post a log of the error on AIX +Unfortunately I cannot recall what those reasons wereIf the jsonrpc or jabsorb are not being maintained its probably safest to rip them out and replace with something like Jackson +Should we wait indefinitely? The lock is acquired in the async handlers not part of the rpc and there is no generic way for the client to track the status and get the result of master operations + +Resolving as cantrepro +The reason for that is very likely because the region is split +proposed fix +ASF Sat Jul UTC genspring Support schema validation for Java EE Deployment descriptorThanks Rex for the patch! +Generally its more specific than a general contrib grouping thoughAlso something to think about is the possibility of having independent release cycles for some of the modules +A thoughtHow can I determine whether the maven build is using the relative path correctly and there is not a setting elsewhere evaluating targetimages as an absolute path? +But getting errors in second level details +If we want to fine tune our initialization configuration we may want to look at details like how many jobs have we initialized and how many of them wait until they actually are scheduled +Changed to critical since there is already a release with the hadoop class dependency +If more people show interest that the number of partitions to keep should be configurable then we move this to config file +Verified fixed. + patch is no longer required as it has moved to parsing using the modello generated model which accounts for have not applied the in patch +Dont think this is going to reader by the time I want to cutCan revisit for +Ah sorry Ive been very confusing mvn package assemblyassembly fails when using maven +Fixed eclipse added an implementation which uses Spring AOP instead of proxies. +m sorry for the trouble i caused you were right i have forgotten to set the provided scope +IMHO the best way to fix a regression is not to create a new feature and force everyone to adopt it +and in at +In the meantime the only way for us to get real job isolation is to move machines from one cluster to another +Do you have a use case where you would like to pass in your own spans +Prefixing an ID member there with just the name of the next outer element is not sufficient +Fixed in rev +However I think thats a huge can of worms not worth diving into +I think the interesting question will be in the same read as what? Surely Safari cant expect all HTTP reply headers to appear in a single read +Not that i know of +Are you seeing theseIn general can you say up front on attaching a patch whether it passes all tests for you and if it doesnt what the plan is for resolving the failures +Patch to use +Seems like the patch is undoing what did shouldnt be doing thatThe real problem seems to be that is not equipped to run that need extra +Tests for this are passing locally for the part that is independent of +Including editorial changes with content changes makes it hard to understand which is which +So I think we should merge these with your new functionalityIm all for replacing newImage by a document type independent approach like your proposed +Thats not true in this case since the in the async callback is not linked to anything its has no effect on message delivery +Adriano Can you take this issue because I will fly to California on Sunday and have no time next week? Of course I will fill out the evaluation form once it is available online +Unless theres some super stealth sneaks all references to jquery ui have been removed +Looks reasonable to do for the next do you have a few classes handy for us to reproduce the problem +This is a general update to the processing to deal with the latest changes made to Natures and the +This looks good to me +Please verify that it solves your problems +If we dont identify such methods as overriding then the interface calls are not incorrectly optimized to the virtual calls to the methods +I too see the same message Language groovy not supported by this report compiler +Let us know what if anything you can gleen from these + +I think theres something else going on +for the patch +If you can provide a test case that still fails on the current release please reopen. +Updated clustering tests with a new test using for clustering Hibernate SearchThis should hopefully prove that its possible to cluster indexes using masterslave configurations +look like funny +the controller information is the part that is the same for all user reference to the file still needs to be in the i think +All javadoc warnings are and are unrelated to this patch +Thanks Thomas +the issue was moved to the maven war plugin. +Based on real world observation it would appear that the issue is not driven by message frequency but by message numbers +Its brokenWe applied the attached patch to fix it on our clusters +Actually never mind I just had a look at the code and it isnt fixed + for the patch +Fixed in trunk and +Either or is fine +The script creates a table which has M rows and then indexes it +This is where I think your framework is broken but please correct me if Im missing anything +I think I know how to only emitting fields associated with the current record +Spring registers only one trigger there so it has to be Quartz firing twice for some m afraid youll have to look closer at what Quartz is doing there +Cant reproduce issue on latest available build +OK I will move the sysinfo code to somewhere in that was my original plan anyways +the unit tests will not pass without fixing getPos +If youd prefer for me to put this on the other project let me know +sorry for late response missed it somehow +Bug has been marked as a duplicate of this bug +Note I subsequently had to work around bug which meant some source changes to my contribution +Thanks +I like that +targeting beta small bug just needs an easier error message +For the first parameter I prefer to use target instead of trg +No objection only some remarksOne bigger issueThe latest patch defaults to docs in order above performancebut my personal taste is to have performance by defaultAnd some smaller onesOne could still adapt to take the possibilityof docs out of order into accountSome performance tests with prohibited scorers could stillbe needed to find out which of the boolean scorers does betteron them +In this case its a hostipcport right? We should be clearer about that or actually make the argument be an +It worked for me +Attachment has been added with description Patch that this fixes a real bug with the fop maintenance branch in cvs +accidently got should be able to attach now +To me tests are as important as application code and they should show the same level of good design code reuse loose coupling etc +If you use when you click the button youll see an alert with if you use you will see null +In future if you think it is comfortable to allow duplicate security realms we can adjust the page again to fit new needs +Changing the approach a bit in order to avoid the manual process of znode removal +Yes slowestTIP is actually reflective we launch speculative attempts for which have the slowest +you can try after several hours +This patch fixes the bug + for option +Yahor makes no sense to me to have launch configuration complexity for something as simple as launching what for users are just a browser +it didnt cross my mind that this might be the case +Commit of the patch breaks the test case in +I am running regression tests nowThis fixes the problem Kristian identified +Ill try to provide you with a new including the in the +Client A gets new block starts writing +Rajesh commit used wrong bug id +releaseLocks clears the set itself +If we get our data from a map other than the default we specify it with otherMap +Is that what you meant by the logj user library? If you did then I dont understand why that wouldnt solve the problem since you would be using the same class loader as Logj uses to load its own classes +Ive added compatibility tests with Aries Blueprint +Bump! Should we no close +But as you see pivot still take the character from keyboard directly not from the input barSo the bug still stand asFail to input character other than EnglishDo you agree +Rebased with no serious changesRather than adding more features to this particular patch I punted on online index additionremoval the current implementation logs warnings that the index will not be available until after restart at which point the previously described method synchronously performs the indexing +I dont want to embed UUID cfId to file nameI like this idea a lot +port +This requires that the configurations live in the repository +Also thank you for downloading the build from jenkins and helping to test the latest +And all the methods are documented via javadoc +Maybe we should warn when validation is disabled and throw exceptions when validation is enabled then enable folks to disable validation in toolsYes I think we should by default conform to the spec +It may not have when the ticket was opened but at this time it does. +Please confirm that you want them replaced with the versions I retrieved from the backup +Agree +I hope that explains it and takes care of question and +r +I am pretty sure if you make both versions use identical SSL initialization logic and identical both versions should behave consistentlyOleg +Junit test case that demonstrates the is a valid use case for this +Ill double check the fix was pulled in to +Fixed in cvs +Ive also run the benchmark but will comment about this elsewhere where it is relevantIf we can clear up the jdom discrepancy I note above then I am on this one +Several problems here I was using twice so it was passing true as the filter value +Yes lets cache the good dirs +Then the message is sent to a second WAR on the same Tomcat instance +This issue needs more information as at the moment it is not clear what needs to be fixed +This only occurs in the all profile or any profile with Cache as distributed cache. +my mistakeBut part still seems to be a problem at the cant reproduce the problem with +The optimization could be applied to as well as to qf right +Hmm +It is ok to fix the test as the problem is related to timing issue +class files then inspect the classes for mojo impls and then extract annotations to generate the +I can see two options how it could be fixed copy lots of code and forcefully reload model shown in the effective pom add a listener so when model refresh will be completed content of the effective pom will be refreshed the second approach you still wont see the changes immediately but it has an advantage that changes made after effective pom editor is opened would get there without reopening the editor +Follow on issue for finding the last ones filed as +I was out of town last week +First of all I would like to say that this is for the Bundles plugin explicitlySo my suggestion is +Thx a lot DanielI will create the mail symlink later this evening +I believe the sorting earlier referred to the file list sortingIn that case although gives consistent sorting for HDFSs listStatus and such note that Javas File do not provide the same consistency while using getmerge over any +We already have a plain new which is supposed to be faster and I hope has some fix for problem like this oneThe problem is that switching from the actual parser to the new one is not a simple matter +Ok I get the point nowBut I still think we should have specific unit tests that reproduce specific scenarios than using some monstrous tests that happened to stumble on a seed that revealed a bug +The goal I am trying to achieve is to have a single HTTP Binding component act as a gateway for subsequent dispatchingrouting based on message properties +Verified by Oliver. +It looks like we have met this issue in in the link +Thanks for your advice and checking Bryan. +Thanks for looking into this +forogt to mention to list the quota limit and usage use listquota a junit test to run with for quotas +The patch adds more than test cases to as the feature touches all the logical operatorsAll unit tests pass +Why not leave the old distcp as is and add a new command for distcpSure I can add a new command for distcp +in JBDS +fix javadoc +I think Hongs idea seems reasonable +I would like to be able to set the context path with its own parameter +And my test server that has enables works without error +We have tests to test automation in the repository already to start with +Heres a patch Todd +Im afraid When I reserve a room by sending an email means the room is reserved via Calendar +all issues to the CR +Commit above was actually for the CXF jiras mentioned above are fixed in the version were using can you please verify this issue is fixed and mark it as resolved in that case? Thanks +Make sure its Tomcat +The log information displayed in the create bit of confusion in both user developer perspectives +Code committed on and trunk +removed it was for to +The failure in is unrelated +Yes we intend to have hosts authenticate themselves to the domain +Looks good but why only available on which outsiders cant reach +C is extracted in a separate issue except for B which is extracted to a new issue +htmlUntil we settle down the definition of the skinconf I have just revert them +What platform is this? Android right +Marking it as an enhancement to category XML code generation +The idea is to provide a way to opt out from power assertions +On second thought it might be a better idea to change isValid to a protected method so that it can be overridden as needed +Ah never mindThis is some code we took from Cocoon without changesWe will have a look at this and probably rework it somedayThanks Simone +I just noticed today that the problem also happens if you call an method in your parent controller from a subclassed controller +As from the original issue description Phil intended to move the to the general package +IMO the s code requires a cleanup cause it has the same code which uses to create sslcontextIf the fix is acceptable then I propose to create an extended operation to reload the sslcontext with the updated digital certificateSeelmann this patch includes the original patchs contents also +I think I see where your going so if you wanted to do a distributed search and have different parameters passed to different shards +So because the fault is and the interface is the throws Exception will refer to. +this new patch improves the previous one and adds support to color spaces to the +Eric Baldeschwieler wrote why not store the cluster in the data nodeWe can alternatively store namespaceID in every data storage belonging to the clusterMay be this is conceptually cleaner +Regarding the use of avro for urls perhaps avros to leave room for a tcp avro urlI would like to minimize the number of transports as they combinatorially explode the compatibility matrix +Moving to +Contains the patch for the srcmainjava is the patch against the srctestjava the patches have been applied at revision r thanks a lot for this enhancement please verify the problem is fully resolved as you expected +bulk close of all resolved issues. +Here is what I did for working with Enums +I need to modify the toString of the object being mapped onto the check box to work as the get of my property editor should be fixed as part of now +The Launch Tab Provider api just isnt good enough +Attached patch which takes care of old metaroot edit +It takes a while until commits are visible in public CVS though so Im afraid youll have to wait for a couple of hours +This is the project where the bug all modules have different namesseam nameseam in beta +Device has been cleaned +im quite busy right now but trying to send patch for this issue as soon as possible +We now require the quantile file to be created before we run +unless the property is a huge binary i would argue that reading the propertyInfo is not more expensive that having an SPI call to check the existance and the validity of the new Value +Can you send us the JS code that youre using? We already ruled out the splashscreen with the other issue and I see that webcore is the part thats crashing +The original patch has a slight race condition within the time window between testing for the presence and creation of the lock file +The project causing the NPE +Hey OlegI agree and Im not trying to spark any problems here +Tentative classes have been coded +Let me do it right now thenFYI support this feature without the +But the errors above indicate that we are unable to mount a test hierarchy because there is another hierarchy with some of the required subsystems attached +Thanks for pointing it +About the changes in evaluateStep there are not so many differences +Note that the Data Bindable field will be removed from the table +Ill wait for the go ahead from justinandrea before doing so +Thanks Mark the patch is fine +The sample app is set up with support for embedded jetty which is good enough. +Endpoints that create new messages are traditionally so either we have two markers for the ssame responsibility or we should pick one or the other +bq +Verified on latest build commit cccaffdbcdecTest Passed. +Fixed the webapp samples are not running into this exception anymore +Hmmm +Yes I brought back some of the changes that deleted +still consumes the post datathe bug also exist +Beauty +However as the test corpus did not complain of the change and extraction of ordering value from an instance is code in I have committed the change but with some reservation +I was actually indifferent +Thanks! +Maybe going forward we should change all WKS stuff to use gperf or some other perfect hash mechanisms. +Please add info about since what version or camel features are available in the wiki +After Ive spent some time with my improved version Ill contribute it back to the projectThere are a few cool features I came up with to support collaboration +Right the real solution is that we need to slow down writes if we get in that sort of situation +Arturo can you please confirm that the problem you reported is still an issue? If not Id like to mark this issue as resolved as discussed above +But given that this is pretty rare it is better to be safe than sorry I guess +The issue is the same but IMHO has been said as duplicated while its for cases when a single value is something more then just Object +Thanks for raising this Kenny +fixed and tested. +Support implicit and explicit wildcardsAlso we now strip parameters from the mimeType we are looking rev and told it works +Please take a look and let me know what you think +Eliminating broken files already deployed in central its not hard its simply impossible +Confusing terminologies! There is no point in supporting OASIS because its dead +The ticket indicates this is fixed +Ok I had a first glance at the code itselfThere are a number of problems in the code itself notably exception handling and dealing properly with session setup vs +Thanks Alejandro! I have started with your patch and am starting to look with which does not need a +You could do standalone but patch looks like it does the job and its small change. +I think and tried to compact the same region at roughly the same timeOn INFO starting compaction on region contentcedccabfeadfcfc ERROR Compaction failed for region contentcedccabfeadfcfcOn INFO starting compaction on region contentcedccabfeadfcfc INFO compaction completed on region contentcedccabfeadfcfc in the title of this issue to reflect analysis findings +LGTM +Thats fine + counts the number of prepares good starting point for a test +Yes we still fire the online event twice as the type of connection changes but we eat the offline event +Closing this my repro was at error sorry for the noise +I have a fix and am planning to put it into +I would recommend defining your own Factory bean like that registers exactly what you need +I am sure it would be straightforward to convert the current Accumulo tracing code to push to a Zipkin collector too seeing as looks like they borrowed a lot of code from Accumulo tracing code +fixed checked in trunk +You are right this does work perfectly +The issue with the date fields is resolved and several other test that failed now pass +This patch just makes the be in sync with that +Then the second goal reads the list and deploys the files +To me the following are bugso any attempt to add a writable page to the cache for a table should just get an early error rather than fail later at flusho any writable log action to a readonly db would be nice to get a reasonable error rather npe +Leonard no youre not too picky +Claus yeah looks to be test only +Dogan are you still working on the? I am asking because I will need to develop an for a project that I will start working on this coming May + Added a new project to vs solution titled daslite Implementation of and classes Integration with sdo onbuildGraph method is not working correctly Also added other classes of daslite library but not completely implemented classes suggested are implemented and resolved issues. +Depends on adding site locales data to the rendering context +So this week good enough +Here is a first cut at supporting multiple concurrent clusters the image instance sizes zone availability and ganglia +Jason are you working on a patch here +That did not seem to fix it +There is no exception thrown in ER when service C is called from service A +Debbie you going to work on this or what +Committed revision +Assigning this to Rick hehe +Thanks ToddYes I see the merit of adding a weak reference map in the Path class +This value should be set if user want to optimize performance of streaming lob from server to client +Closing stale resolved issues +or we could add flags in to support thispublic static final int INTREVERSE public static final int FLOATREVERSE public static final int STRINGREVERSE +Postponed to when resolved we can postpone it to +Both sides need a warmup +Do you think we have to put a limit somewhere to the boundary length we allow to avoid a very long boundary to make us allocating very big buffersI think it would be much better and cleaner to enforce a max limit on header length +In loadContext has parameter of which is not used +Assigning myself +So the reason this weird error about filenotfoundexception was not because of the tools but because of the old ssh configuration being wrongThis still leaves us with the hardcoding of the keyname but that is a simpler problem to solve +I do not even know how to overrule the order of the repositories that are searched and comes first delivering broken pom files +The continuum zone has an updated account as well. +added a boolean flag to allow the user to decide if this behavior is wanted for a particular flow +I decided to create a new JIRA for that So I can close this one. +thanks a dont like the approach of commenting this line out +Also verified the actual behavior using minicluster and it appears to correct the problem +Ive since fixed that but wanted to go ahead and finish off the rest of the parameters and such before I upload another set of patches +Pavel thanks for reviewing surely usually I clean patches before commitApplied at r +The thing I dont like about the blacklist idea is the following +Thanks! +Instead of binding to revenue a list of doubles or a list of strings I got a list of String ArraysLet me know if you need me to test this differentlyThanksthis is a different thing +hi julianfrom my understanding the SPI impl must only provide amaybe i dont get your pointangela +I am a bit confused +And thats why it needs to be done thru config +So its documented in bugzilla yes I did. +Applied to branch as well +With it definitely occurs for me +Test in trunk +Attached patch uses modifies get to use get which avoids trying to resolve the +then this is duplicate with +As far as I can see maven sbt are really good tools +I committed a much improved version of the code in revision +Patch has been applied +Hi Pavelagree Im gonna make the same updateRegards . +You probably had your very good reasons to do so +Will request some community testing and close in a week or so if no issues are reported. +Here is a patch +Must fix for the next release +I think your workaround will let your app run in a cluster but as you mentioned its not really a good solution +Reworded storage and fsimage file messages +any progress on this bug? I just encountered it using would it be possible to make the same patch for? I saw that the content of in your patch and in is very different +The application on top in this case decides what to do with that information +I have been doing a bit of research today axis seems to be fine however and are both affected and what is in the latest snapshot to the best of my knowledge +Patch coming up +I just ran into the same error while testing with Hive +Yes its in the CVS +For us nested anonymous definitions simply isnt a choice and while this patch is trivial and not best practices it would greatly reduce a lot of cascadetrue noise +patch includes ACL corrections for grant revoke get and AppendPlease review +Manually verified the feature on a single node cluster +Hi Developers or AdminHi Adrian Dick Sorry I ignored another statement which need to be correctedHere I all the two statement should be corrected as belowFrom the generated code out is really a complex array which is a new addressingaccessing method introduced from and be used elsewhereThe complex array is an object not simple memory block so should not been deleted directly by delete +guys for this effort +No longer relevant +Move issues to next release +This feature allows you test your transformations from within the console very handyThis is already implemented and sitting on the workspacetfennelly branch in SVN +Converted into enhancement request +Clicking on the Orders shortcut of any Party produces errorThis patch for this Issue +Bulk close for release +Hi Logan I had looked your patch +I also noticed this +Oops my mistake +So if you have new lines tabs or any other kind of spaces between two tokens it should not affect the trained modelDo white spaces in a sentence annotation affect the trained model +a byte is fairly compact and would let you track some problems if you are packet sampling client just needs to stop incrementing once it is at xff +It stays in this status unless you click Cancel Patch +would vastly prefer an option that doesnt require escapingmaybe a slight preference for over in case we implement SQL arithmetic in the distant future +Right now theyre in the middle of CP releases. +What about development of a simple tool that supports updates through import of EPSG definitions in for example WKT format? This does not address the larger interpoerability issues related to an outdated distribution database however it would be very useful for individual administrators +We already have a builtin class reloadingconfiguration reloading mechanism that could give you some integration ideas +You basically create the impl then you create a class that extends and then register this extension in srcmainresourcesservices file thats it +We are putting our main efford into JAXWSJAXRPC tools issues may handled by community should fix this in wsprovide +For the other mailing lists I received a WELCOME mailDo I miss something here +Can you elaborate on what is not working? Can you provide a patch to the existing tag instead of code for a new tag? All you need to do is modify your copy of the original tag and run cvs diff u on the file to create a patch file +Also added a default switch option that throws a as all of the JCR values should be explicitly handled in the switch casesAll unit and integration tests pass with this change +There are some style complaints that need to be fixed +The problem is actually in the class itself +Do away with the unit tests all togetherAlthough +Fixed on trunk revision . +Scott you know have the karma to assign tickets to yourself in JIRA and edit the wiki pages +Pushed. +Moving out +Synced from there next times you just need to deploy there and ping us to poms were not present so they were not added to I finally added the relocation pom myself for this version +Hi ShawnAll the required bundle configurations are left there with no change +rebased +I have exactly the same problematic +Its an incompatible change that breaks backwards compatibilityFwiw I think you assessment that most people do this is wrong +I think performance issue is yet another issue and I believe we can find the way to speed up +Shantanu do you have encryption enabled +Can you post the test code compressed by a much simpler way such as +So it has to be replaced by another method or removed +them at startup time? The namenode doesnt keep track of what nodes have which blocks why should a storage node keep track any more rigorously within its own state? Id argue that all of that complexity is needless the cost of maintaining a consistent state is way too high for little benefitPlease make it very easy to change the code +this issue to talk about cell style inheritance +This file has the Getting Started instructions to address this issueAdd this file to APOLLOHOMEsrcsitecontentxdocs directory +Not critical so for +Applied regression fix to branch +Ill take a look at to confirmfix it there too +Committed to trunk +The patch is in the SVN repository +Patch for the revision of release. +Attaching doc that covers a range of information on how offline snapshots and recovery work +Im pretty uncertain about the Set howeverI am not sure either but I think it is an idea worth exploring +Patch allows lookups to be configured for +Thanks Nigel +Guys before we look at committing this have you tried the trinidad trunk to see if this is still an issue +Luckily its not necessary in the current replication scheme which no longer relies on snapshots +I missed it. +I think your suggestion is a good one +Hmmm +Ive kept the sources + +Fixed +Did you change the file yourself or use the standard one +Let me know if you see anything interestingoweni took a look at the netstat and nothing jumped out at me but ill show this to a couple other people to see what they thinkim posting an excel spreadsheet with the before after and delta numbers in columns here is the timeline Running job job map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce map reduce Job complete jobJob ended Tue May PDT The job took secondsFantastic! I wasnt aware of that optimization +You want to read this body +HFTP was always designed to work as the HTTP front end for HDFS +ROWID is actually in the database +maybe a minor +GaoYou got this in? or someother version +Super Thanks! +writes reads both? K regions on servers thats it +Attached trunk last so Jenkins should be able to run itThe testDelete method was already there on the test class but not declared at Test so was not running +Waiting for new numbers later this wk +r rChanges which speed up the encoding of qpidmessagingVariant types significantly by usually avoiding the intermediate step of using a +After the exception come the connection is released and everything works well +good jobThe patch looks good for memvn clean install works well +Thanks Todd! +Issue is that once connecting its is blocking another operations +Maybe I was wrong +Thanks again +It still needs to be fixed forTC +Better use synchronized maps to avoid more such issues as class and its inner classes used by multiple threadsSubmitted patch +If you have an offline payment in your store that you want to refund the procedure is to issue a check +Manually tested! +Yes it looks has the same problemSomeone has to improve it. +This is not an issue on right +Updated summary to better reflect what kind of actions are expected on this issue +Moved trash tests to fs HDFS test of trash to a subclass +This fix was put in after Cache was tagged forSo it would be a known issue for +Thanks Kasun for pointing this out. +tweaked the css to narrow the div and fix the overlap appears that the fix has caused problems with the title is not specific to IE it occurs whenever the user overrides the font size +Thanks for looking into it +they built in +I have updated the wiki pages +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +Retrying hadoopqa +bq +Provide user input list and host selection by role type +Ill commit in a day or two +Removed dependency on conf from mapred jsp pages + +Magesh +Attachment has been added with description Minimal no access to Mozilla just realized that perhaps script event handlers should be executed with the on the moment of script handler assignment? It alone would be sufficient to close the current bug and perhaps prevents manyothers like accessing Jython interpreter or any other interpreter and using themto assign event handlers and escape from sandbox +Not thinking straight today forgive me +To start with we may in fact have no indices and then add them later +And I dont know why that choice couldnt be based on what exists in the string if it exists +The current editor is really hard to use when you start quoting messages that have already been quoted once or twice just that makes it interesting to have a mail gateway as soon as possible +I never managed to reproduce this issue in the past because in the tests I made at WHO we always hit with GET requests as opposed to POST ones +Can someone look at faces and see if that needs line ending tweaks it simply looks retarded in vim on my as fixed +Will be removed before AS project is past CR this should be removed before releasing final + +Yes i got it to work as expected using a proper +Distinctly possible +doh yea this is dup of +Committed to issue as it has been released as part of release +we dont need to lock while reading right +If this is easy go for it +Yes great +Also since the test now uses the string client twice and wont add the new test cases unless both occurrences have the same spelling perhaps you should define a constant to ensure their consistency +Committed revision . +I tried to build from source but the parent artifact is missing +Oh right of course +JIRA SOAP Service WSDL states that the environment property can be nillable +The current patch fixes the reported bug +Heres the patch for this problem +Juergen is doing a great job in raising untought scenarios +linkng docbook stuff to +Modified patch which handles the previous mentioned to allow better integration of the and the new authorizor method +The approach taken in the xml input sources was to use a static utils class to help with restarting etc +Else we would be forced to have all table names to be globally unique which would different from database semantics +You didnt remove the +Ok so I see the problem +What is the status on this? The new scripts dont work properly since does not pass the master URL to the slaves +Kathey I can see where you suggestion is coming from when I look at the examples that you provided +In trunk and +I am working on backporting this fix to let me know if anyone sees a problem with this +But since its a last minute addition and not tested thoroughly skipping this feature for release and differing it toLahiru +Patch applied with commit revision cccccbcfcbecbfbThanks +do you have a sample url +This would ensure that the users +Attachment has been added with description Source SVG has been added with description PNG you rasterize a SVG file with the latest SVN version of batik +Madhumita are you planning further investigation on this? I am also having this assigned to Strong so hes aware of the regression from the JIRA notification +But seems that the addEntry doesnt go thru journal just move entries between entry log files wo any flush guaranteeSijie I dont understand what youre trying to say here we do write and flush addEntry requests to the journal +please post the error you think I have narrowed the problem to special characters in password to be the issue when verfiying the dsn even though it successfully creates it the first time. +Assuming it has been fixed given no negative feedback on the fix has been provided +Yes I dont think Diegos solution is correct +Operational questions should be asked on the users mailing list not the bug tracker +For instance the AS build timestamps and builds only once and packages the SAME jar in all server configs which should have the SAME mdsum +Once we flip the meaning in the code we need to preserve that defaultThe problem kicks when we deploy that in production on old configs +I committed the resource checker doc as part of the commit +WDYT? +In fact the error is class not found it can not find the class that is in the jar put as pojo in is a view of my guvnor result of previous enum correct patch is forget the two others now all tests are passing +Permission issue is happening and the reason is similar to the one observed in +Deprecated code removed +Were not going to accept breaking changes just to make maven happy and we wont make any legal commitments or sign any paperwork +One shorthand method is no big deal but if this becomes a pattern that others follow then we could have a lot of shorthand methods creating code bloat where we dont really need it +Der Fix repariert aber NICHT deine kaputten Diagramme! Ich kann es wie gesagt bei mir leider nicht reproduzieren +I reproduce hung up +While doing cogroup we first put tuples from all the relations in a heap then we drain the heap and generate the output tuple as appropriate +Closing tickets +implemented in a in prod now. +Fred! Have you tried this with axis transport? Is the problem solved with the new transport or do we need to keep on looking for a solution for this +Applying just the part from that clones the SC returned to callers fixes itSo applying and after all since weve apparently empirically demonstrated that returning mutable objects from the memtable that caller is supposed to remember not to mutate doesnt work in practice. +Patch was committed twith revision +Thats could you please next time indicate what the deadline is for the next CP so that Im aware of it and I can fix it before the deadline? Galder Ill add dates in next time + +So lets keep it. +v deployed to and BSC now points to +For making it compile with I will resolve it +Please let me know your library of choice +Increasing PLEVELGoing to apply patch with echo pp Etoddgitcommon git show relative PATCHecho srctestbin Looks like this is a git patch +For this corpus I was unable to improve it in this wayBut maybe I made some mistakes in both these cases so anyone feel free to try this themselves +it should be download while building or something like that +it should test most of them +waiting for feedback from mark newton on how to do this isnt possible anymore with the new cms. +All clear the description was a little unclear at first but now I see the problemI will look at itregardsMartin +That seems to have solved it +Thanks Tom! +Just curious what prompted the move from to a custom script +Seems like a simple clean fix +See the files as they list any additional maven repos. +release candidate to be available tomorrow +patch with new unit test that includes verification of execution path in child map taskworked with Chuan and have it tested in both Windows and Linux environment +Committed +It appears what is happening is calls into which calls the +fixed +I was using Eclipse and I did not realise that I had to restart Eclipse before the changes in compliance would take effect even when doing a clean of all my projects +Before trying your own files why dont you use a binding file to specify the custom handler during code generation and stick with the generated sources +This is intellectual noise to me +If you are a table that is arranged by date for instance and also performing you will eventually have a large number of empty tablets +Thank you very muchI am marking the issue as fixed in the trunk +clustersrcmainorgjbosshaframeworkserver to and also checked in only remaining step is to document how to configure the file +Anyone still watching this bug? Im going to close it as Cannot Reproduce since the original device showing this bug doesnt normally run ICS +This is very similar work to what was done in +Is there a patch for? +Oleg and I discussed this one a bit also and we were wondering if there is any way to refactor the adapter so that its actually a polling consumer with a Trigger implementation that is aware of the restricted rate requirements +However I dont think thats a sustainable workaround +Bugger I dont have access to edit this issue and change it from bug to improvement +What is the possibility of you moving to MVEL +We need more of these! I took a look at the trunk patch after reading through this issue +Committed revision +adds emission of the row causing the problem +Thanks Vinod for good suggestions here +Im pretty sure that Android does not support widthheight on DATAURLs +so to check on legal front would it be ok to override the class which has a different license compared to t we provide a fix for logbackI think overlaying the class until its fixed there is fine +nothing came for the thing you mention there are the ant targets that you can use for that purpose +How easy is this to reproduce? Do I just need to start running mapreduce jobs and it starts to show up eventually? Do these need to be long running jobs? You just need to start MR jobs and it starts showing up +Unfortunately we are not maven yet for this test +Perhaps it could also be labeld a task or improvement +On the above Id say leave region creation in even if happens only on table creation +cks! Im sure this can be simplified so I dont have to grapple with fixing the unit test before I can run simplify and this case if you want to +Lets get it fixed in +Implemented serialization side support so that wrapping can be enabled easily by setting to true +Would that work? Are there existing performance regression tests? If you have a preference for a different workload or configuration let me knowre the other config file settings maxloggedlatenciesforconsistencyprediction is possibly useful +The patch looks good +Is it ok to backport? thanks +to trunk thanks for the closed. +Actually rename is one of relational algebraic operation +russian article on Tools and JBDS last realease is done +The Redmine ticket is not settled and Matz explicitly states that should check the rescued object as doesUntil it is settled on MRI I for one am unwilling to change this behavior +Run db to set up table +It was handled previously because the code used executeQuery which always returns a result set even if it is empty +Should we close this generic migration issue and in the future make new issues for each migration step we add? Right now this issue isnt really adding very much +Thanks for looking at the patch KristianCommitted revision Ill leave the issue open until Ive had a chance to run the test where Dyre saw some unexplained behaviour with the concept patch +Running the unit tests has tested mockito and running up our features in Eclipse has tested the others so they do work +I think the current behaviour never results in duplicate identity column values being produced for use in a row +Does this sound correct +Its exactly the lack of consensus we see here thats why I am against having the setter approachIm totally against some deprecationundeprecation loop because we in future releases another settingwants to be liveIt seems the only way we can avoid this is for javadoc to be the only specification as to whether a settingdoes or does not take effect live +Please redirect if you feel it doesnt belong to you +Fixing them resolved the problem +The only drawback of this implementation is that this is no more possible to import some test suite reports generated for instance by an Ant task +Adam can you tell me how these properties are currently set for envers? Is it just another value in the current Hibernate propertiesconfiguration file +Changing ownership of the created +AndreIt seems that is not translated to BrazilianCould your please translateThanks VasiliySorry I forgot to translate this file +Naturally we do expect the bug reporter tests the code after having applied thepatch +Changed bug to reflect that +See also Issue for input +Bhushan You have some changes in which looks like are unrelated to this patch +Im overriding the version in JDK using the endorsed mechanism to ensure that its the correct version and Ive checked to see which version is being picked up at the command line too +logging in as a user with the same roles as you i can see all the screens correctly +So a single ZK instance on the same node as a RSTTDN is probably the worst setup ever + +But for the time being I am resolving this issue. +By default the session replay buffer is limited to either commands or Mb worth of commands whichever comes first +First patch for replace all the head style to head style +Hi Markus thanks for your patchBut when i create a new category DVDR the widget encoder encodes it to DVDR not to DVDBR +I dont have another time source Yet it is better than nothing +Ok well take this as a documentation task to warn people about this prior to getting fixed. +I will adjust code that when doesnt founded there shouldnt be a +Please let me know if this is commtiable +For your convenience I made one combined patch for and +Move mouse in the + for the patch +In most cases the people running the data centers will have complete control over that and application developers will need a way to accommodate +If you dont plan to implement this please let me know and Ill swith to another oauth implementation +Using this the admin would disbale the replication workers but the auditor would continue to see bookies dropping off and marking all the ledgers as underreplicated +Alex port this feature to So no more action is needed. +Ill send it to the list +The code in Hadoop Common is somewhat reusable for this so it shouldnt be too hard to implement +Should we mark this as closed as it was already committed over a year ago? +But there are some methods in that do not acquire a global lock but access blocksMap +Moving it out of this as fixed for +fixed according to the tests +I made some minor tweaks to the part as some code was altered beforeThanks for the contribution! +I totally understand the reason is not a due to the lack of multiple inhetance in Java +Too extra needed code is so tiny that there is really no reason for not doing as far as I see it +Maybe the error could be avoided by using versions in dependency management +After a lot more examination of the code and combinatoric testing I see why it works and is the right fix for this incarnation of the DIHAnd yes unit tests for more of these options are on the way +bq +Would making this constructor private resolve your concerns +Adds better debugging output too +Please copy the page content the tests the ones you mentioned in? Are you sure that is picking upPlease change and in both and +When you use Open Office or Symphony you always see one of the alignment options highlighted +Even if you gracefully handle the exception you cant redisplay and ask for corrections +Is there a library or application that needs to a +batch transition resolvedfixed bugs to closedfixed +But howll I know if it is installed properly +If you were able to provide a patch it would be really great +bulk close of resolved issues. +Yeah agreed +It is be possible to use spring rich without ever using jgoodies we have defined several dependencies as optional to minimize the memory footprint which is required to use spring rich +Are we sure this thrift daemon behavior doesnt cause issues with multiple simultaneous instances? But we could just change the check from a flag to a counting semaphore to achieve the desired behavior in light of this +Im not really concerned about it though as Id rather just keep the bit of the sha and just make it eye +Itd be nice if we could find a more generic solution that fits in with other injected methods + +Multiplexer integration has been enabled for +When record skipping works more generally that method will disappear and the correct jobdata methods will be used +You cannot share the security domain with the is you disable the unified class loader +It is a very serious issue for us as we are approaching our delivery date and we are not able to get any workaround for thisCould you please take this up as a high priority issueThanks and RegardsKhem +I would recommend updating the documentation to reflect the limitation though +After following up with the Hue folks I agree with Daryns point that the server should not set the service +Maybe +There are many things being for +Great thanks Oliver +can you provide some examples of what sort of data might be in each of these tables +Issues resolved in +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The double click demo was excluded from mobile showcase demo +I cant reproduce this either well need a better idea of your projects structure to investigate +Rolling over to reduce workload for pre can be placed in packages now only limitations is controller Prefix still has to be unique +Can somebody will look into this issue +we have the ability to do this with existing commands dont weThe name of rollback snapshot is not returned +Actually this is fixed by +Please reopen if not. +If that doesnt work then need to download and install pyro +for and its test case in committedIncluding TestWeak and TestWork correctionsDIGY +Meant to close up being an issue with QEs ant build script +Removed the initial connection moving its closure showed that only the encessary threads in the pools were started and so it wasnt having the desired effect of taing the pool size out of the equation afterallAdded a processor count based variable delta instead +i had the same issuedeleted in my repository it was downloaded again and worked fineseems i somehow got a corrupt jar in my repository didnt help but mvn clean eclipseclean eclipseeclipse tomcatdeploy worked somehow. +If you stop it you have to restart server to recover +I took a closer look and doesnt seem to really add any new synchronization help on the tasks list +This jira has been closed as part of a review and archiving of the aj project +mark adds the corrupted block into corrupted no matter even if the block is not +Attaching the logs from both the setups to the bug +For I wanted to see prototyping of this and the work that has been done has accomplished that +Implemented all feedback from reviewer +Fantastic Mario! thanks. +Please upload your aar and we can reopen this bug +So similar cleanup code for the should be applied in as well +as long as the concern is addressed +Vincent can you either close this or refactor for the new classThanks +Close Doc issues in release versions. +I realized some of the patchs are not related after sending it +Duplicates then +Ill take a look at that next + unscheduled issues to the next release + +Can you advice on how to get from ? I dont see how it can be done without exposing additional methods in +Index building and updating works the same but is now encapsulated by this API +It was actually backported to from trunk with a forth digit version bump so maybe if someone really needed it they could look at it again +Ive not been able to reproduce in a lab only in production +Ive seen that in the http core the IO for file is handled by the class if i not mistake and tghis class gives one method for read from an and another to write to an +The codebase only has the optimization on +This most likely affects many projects that are being ported from to +Mine is that i can bend Ant to my will with ease +Yes I would prefer adding as well and I think this would make sense as well +Also fixed for add new +Also need to tinker with create in +Not exactly the goal of the title of this ticket but allows for the same result +Great thanks Christian +Need to think a bit more about he fact that we are duplicating status code when no is specified In this case status code currently set in payload and headers +YourHirams help is appreciated +It does exactly what it can +Users can however optionally provide their own MAXMEMPERTASK limit for each task in a job +Do you have a problem with that specific fact? Trying to understand what actually remains here +I got it to work now and if its not really affecting anyone else then I guess its not worth keeping open +Close after release +know at least other Codehaus projects using this ids +introduced a signature change to endFunction in +I found another bug +covers the same problem but for the +Does it happen with master as well? The HTTP Range requests were majorly modified on master +If its a product issue i would suggest to keep this defect open and fix in product +HiWe dont request the Wifi state permission +on the current design +Committed patch to documentation trunk at revision +The testcases jms producer endpoint was not using a pooledConnection factory so when you send larger number of messages through the NMR it eats lots of sockets +Status +Added a unit test to test thread safety +committed thanks +One way forward on this issue might be to split the work into smaller increments +Fixed at r +Should be is empty or something similar +Thanks Lars. +Per Satheeshs comment in May this does not appear to be a I agree its not a bug I could see this as an improvement +This is covered by another JIRA item +Sounds like cqlsh needs to set CL back to ONE when selecting from the trace events +were seeing rather consistent bad results but only after documents and multiple commits and wondering if anyone else is seeing anything +Thanks Sergey. +Hi MaciejTheres all sorts of havoc an unauthorized user can wreak if an unauthorized user has access to the UI +Fixed rev +I needed the table index to be constant so I created a property file that defined all tables and their index +takes a step backis this even a scenario we should be supporting? if the cluster has a schema he should be using autobootstraptrue and we should be looking at ways to make it more difficult to do the wrong thing instead of some of the sharp edges +Disregard the last commentI have now cleaned and rebuilt and now it still locks up +The test case provided used the identity transform and so still does not run +The following tests have failures +The check in line is not needed +noting that we have several centuries of headroom in in bitsWe can definitively do that good idea +Screenshot shows the expected behavior of the test application which works with Tiles jars +This seemed like a better place to put it since the destinations are initially created on the interface +Please confirm they have been applied as expected +Yup as well +Note I ran into an issue when using it with sonar +Conversely it is necessary for every submit to include the extra content? If one wants to disable it there should be a wayCheersNick +Next failed in different builds Avg Avg Avg +Completed in Git ID edecabaacdccfaafca +Nice one to and TRUNKThanks for the patch ChunhuiThanks for the review Stack and Ram. +Or is there a workaround somewhere in the scm or release pluginsI see that has the hidden features is there a implementation then? +there is also announce and private lists Ill assume to delete those too +I keep all those code that closes streams +Constraints forced me to handle it this way + +If multiple threads access a format concurrently it must be synchronized externally. +As all the queries are having the new query in single line will maintain consistency +this king of patches make always me think Should I create a new issue of reopen and overwrite the patchWhat is the suggested way to send this kind of recurrent contributions +If you want to run javac add a reference to the classpath with a dependency in the plugin declaration. +Plus changing begin values from the page using input box leads to java script error +OK I solved the mystery +I cannot replicate this issue +reopening for backport to maintenance branches. +Portlet name is currently used as identity of that child node which results in conflict in case there are two portlets with the same name imported into registry service +webhdfs or webhdfsv +Close for all issues before incubation +Im curious why youre proposing different keywords for use with RANGE than ROWS +on this week +Pls attach your as reference anway +I ran pmap while the client was running to confirm that the block files were in fact getting into the client process +I know some of the other folks whove been around longer could probably provide some more guidance however again my for committing this +This bug has been fixed in Xalan Java +Job done +HI SebastianWe rely on lunatics err +For Designer we implemented the changes necessary to maintain current feature set +That is fine +Whats missing here? Is it still an issue after all? We are using more recent Dojo versions Jeremy is working on the next one as far as I remember. +am attaching a patch so that the implementation can work with Saxon applied under r +clean this JIRA still relevantThe component used by EAP is built using Brew and pulling the sources from anyway so no generated artifacts provided via or Nexus are ll make sure next time I use JDK when building and releasing artifacts +It can also be noted that if the serialization frameworks wanted to reference Hadoop which granted is the of this exercise they could support foreign types that were serializable using Hadoops framework +create new ticket if +Issue occurs in JDK and +this looks good to me will check it in +Brian This label appears in the Project Explorer view but I realize this is something I should do myself there was no reason this issue was assigned to you +Yeah thats a current limitation +fixed +If I remember right security was added for most of the new features in +Ill try the link idea out but I wouldnt get too hung up on having the application run on two many older Rails versions as it changes so rapidly +Ended up skipping to use security updateThe display problems with appear to be resolved in. +Although Teiid has unit tests they were written using embedded +So I thought it would be practical to have one available out of the box to use noregards Dan Aki sure totally agreed re the motivation behind introducing the interceptor +to trunkTo those who follow trunk closely be advise that the script now starts a ZK instance unless HBASEMANAGESZK in is set to false. +If hama pipes is not necessary it seems that would be ok for just skip it +The interop tests fail because of those +Jonathan can you inline the latest on how jruby udfs are supposed to be created? Its a bit confusing to read this ticket as you guys are working out the right interface +Patch attached rebased on trunk +I already asked the Roo team +I didnt debug to demonstrate whether this is a result of a javadoc plugin bug or a site plugin bug +It will also add AS runtime and Seam runtimes for EAP included in JBDS +Im not sure about this functionality just found out that disqus is HUGE does a ton of ajax is the big player though +As a result it seems the state of this issue should not be resolved. +Thanks Alex! +Thanks TimThe fix looks fineBest regards by Andrew. +Changing the machine or updating the JVM to UU solved the problemI think thats a bug in the JVM itself +Resolving report as fixed + has been marked as a duplicate of this bug +the handler itself is in CVS I just need to create the documentation in the reference test casethere is a different JIRA issue for documentation +This problem is resolvedVerified by SeanThanks for your help +This patch adds maven profiles for hadoop and hadoop +not TomThe empty string test still fails namely bc it is expecting instead of +As discribed about this will require protocol amendment +So I went ahead and wrote both of themI guess the best way to talk to you is by email +transport +I just noticed that the same thing is happening with and +seamplugins +Absolutely your code is valuable and Im sure itll be handy at some point +I confirmed this is indeed the behavior on +The issue has been fixed in tranql driver for informix. +The comment still applies to but what the comment doesnt says is that if you use a x as the it expects you have no remaining component +to and trunk +the SVN diff also contains a typo fix +But thats not fast enough +Thanks Nicholas +suggested modification to solve this issue +Maybe documenting would be good at the same time +We need a way to retry the rehash as we cant afford caches in a failed or invalid state +HelloWe are also interested in having an inline way of rendering gadgetsWhat is the status progress on this issue? I see that code a patch has been submitted for review but it has been some time since the last time this issue was touchedBest regardsHasan Ceylan +old issue can be closed for now +Long term fixing the kernel so the OOM killer can not deadlock in this and other ways is on my TODOThe other reason for switching from to if we can is that the kernel code paths where the kernel does not kill the processes itself arepoorly tested and buggyThe application where this was observed triggered the deadlock about time in it triggered the OOM killer and the immediate fix was just to disable the broken application +Reopening to move to release +Patch applied. +I think Its ok as it is now +provides a similar implantation except they opted to incorporate the core algorithm into the patch rather than use a third party library + and others should be timeouting because of newly added shutdown hook + +Is this one done? just checking to see if we can close some targeted issues. +I figured that by disconnecting the nodes earlier from the underlying overlayed state the changes would not be overwritten +the important thing is it got reviewed +committed +What would be great is a that downloaded and builtThe README could also include more info on installing fuseOn Ubuntu I was able to do this withsudo install libfuse If youve never built things on your Ubuntu box then you might also needsudo install make gcc +Somehow thread blocks dont reuse and old handles doesnt released if we store reference to thread on array or many local variables +I renamed the artifact id to be aligned with the other camel artifcats +You can see the diffs in the new image +It seems manage memory more efficiently too +it should be since it is an message although maybe only advisory +src attribute is empty for menuItem more iframes for menuItem also generated for fireFox +can serialize classes even if they dont implement Serializable +My classlib reversion is r + issues to Nick and had to set resolution of all of them to done +After this file deletion was working +I hit an error trying to cluster some documents I added with solr cell unknown field AuthorSeems like it would be nice if we could handle unknown field types gracefully +This fix plus the one for will provide a way for the Eclipse plugin to dynamically find the areas where files can be published +Error code and message invalid stored block lengths +After applying the patch all seems to be working +jmhsieh Shall we close out the Jira? Since Flume is on and the scripts are on the right place now +There is no way I can add a contrib test to this patch if the HAR creation is performed by a new version of that must be committed to MAPREDUCE +I just committed this +applied to trunk closing stale resolved issues +Some consider desertification to be desertization desertization encompassing both natural and causes of extension of deserts +does the total customer benefit of having the right behavior outweigh the inconvenience of forcing the change on customersI need to understand the impact on existing customers +Re weird +The Add operation has been completely reviewed and this problem does not apply to it anymoreWe have to review the modify and rename operation +Addendum that modifies we sure that this didnt introduce any findbugs problems that might be a problem? I thought we had fixed the findbugs finding so we need to make sure that no new bugs were actually introduced or if they were that we consider them ok +well Ive copied the code over from maven integration however I encountered which prohibits this feature fron working now +Attached a patch +The issue in was first raised under +Sorry weve stopped work on the branches +Optimization can wait until failed to find any null check in the mentioned iterator neither in the current trunk version nor in the version at the time the issue was created +This one was completely missing from +Is there a link I should post here to the reviewboard +Please reopen if its not the case +closing this one +Heres a patch for this and updates to tests +implemented TC +This part appears to be working but since the GCU thinks that it originally had no imports the old imports are removed and nothing is added in their solution here would be to add some more error recovery to the parser so that it can still produce a module node for the sample code above +This patch conflicts with the fix for +I agree with Mikhail +The paper seems to imply that yes but Im not sure why +Patched checked into branch. +Marking as a duplicate though Tim has proposed one extra feature a maven archetype that is not in that issue +maybe we need a +At this point future Ajax work will have more specific JIRA issues. +Okay well take a look for +suboptimal +Vlado no I dont see a reason for creating a new issue +Do you have negative ratings +There maybe some servlet container trouble but triggered by change in os x webdav codeHas anyone used jackrabbit webdav from Mac OS X with another servlet container than tomcat or jbosswebIve also some doubt about Chunked with uppercase C +Thanks for the patch Brett +The test itself was removed from the intermittent exclude list +bulk close of issues as part of making release notes. +I would appreciate if you would use the diff and test it for your case +The id of the referenced condition is now saved and the condition will try to get the referenced condition later on. +Im resolving this as wontfix because I dont think the chaining pattern would have enough value +If you have an EJB artifact declared with an ejb dependency it will work out of the book +This dependency should be resolvable in an environment. +Apart from class everything else should be in core +What would that achieveAlso I think the code needs some more comments there are a lot of magic numbers with no explanationAnd some Javadoc to point to documentation on the Snappy format would be useful +Nick not sure if this is limited to but could you review toPlease fix the problem for select component. +Someone with some serious foo will have to look into that one I think +unit tests havent been merged yet but the fix has been +This issue lays the ground work on which all these other issues will depend +Pull requests sent for both and master branches +Start the server will contain the new scanner +source maven file to reproduce the is how i run the test cases but obviously you may do it differently depending on your setup so feel free to modify the test case as you need +rvt +Moved to an officers area. +Follow access +Assuming the vote passesAttached is the patch for the website changes for the release +Why does Webtools need the commonext component? It seems to me that is an inappropriate dependencyIf the goal is to get the framework to run independently then we cant have any components in the applications folder enabled +rebased +Jakob Theres lots of interest Please post the patch even if its a WIP +Connection framework totally rewritten + as r. +Thanks pat. +The fact that the client transmits messages in Sequence message order doesnt imply that the service will receive them in the right order +Hi guysThis is a dependency issue in +Fixed in Rev +Verified in +Scott are you still interested working on this issue +This is fixed in MRs in and trunk via +ScottI think you need to mark your work complete before seeing the Submit Patch buttonAfter you click Submit Patch future attachments would be picked up by Hadoop QA +If I do get time and investigate it Ill reopen with extra details. +Working on it +Should we it instead +Is this blocked on me for something +As you have mentioned just because a connection is used for HTTP does not mean that it cant also be used for HTTP +This is an improvement on the previous dependencies +from my point of view at least but if they have the same blocking behavior then i guess we can live with it for now if not easy to fix +I accidentally found out that inserting rules with an empty LHS for a certain will trigger it to be handled +Localized setup steps in one file that gets called by all scripts +So if you have a type Header defined in your record and it is not found in that same file we look in the includes directory and try to get it from there +Yeah Im still seeing the performance problem +Pull request sent +Thanks for reporting Cedric do you want to try to work on a patch? with an unit test that reproduces the issue and proves the fix works as expected +Ill have to do some digging +There is a bigger issue here pig does not support boolean type anywhere except for conditions like in a filter or split +Do you have the same problem in a environment +I cloned and patched with Jeremy Tans fix which solved the issue +The patch looks good to me +jsp better modeled Im going to provide a way for a system test to reach JMX beans and pull down any information available there +Not sure what bruno does but a test against the deployment version of the current jetty stable works here +You actually want the following two characters t to appear in the html markup +Can you checkout the code and build it to test? If not Ill try and get a snapshot deployed later tonight +can you look at it. +But a misalignment of the labels in relation to the data fields +My in memory implementation uses Vints and Vlongs so they arent that bad for storing the fileoffsets and blockpointers +I just checked and we dont have dogma so we should add that as its own entry. +can you please add +Hi Nick thanks +Needs lots of comments explaining whats going on +Dropping recoverable memory looks good +Duplicate of which has a patch which looks good to me. +I just committed this +We can probably even get you commit permissions on gt so you can put these fixes in yourself +Most likely your ZKnetwork was in a bad state at that moment +In this case we have to do a fallback to non batch execution +It looks like need to update the docs +Finalized is not the responsibility of anymore +for the patch and marking as blocker +This would all need to be testedAll that said theres nothing to prevent you from doing it as you propose in Pig without +I applied only the changes to and it didnt fail +After running more tests it still seems to fail every now and then at about rate +GSG images fixed +Does it also addressed hdfsproxy issues mentioned in +Knut thanks for looking at the stack trace +Will continue over in trying to replicate root of this issue using Haijuns prescription an entry added to the wrong family. +Ah ok I had assumed it was on the schema itselfWell it is a boolean on but theres no setter only a getter and the ctor which is only called from the factory is the only place its set +The patch for Nutch trunk +Heres a patch w tests for the java side of this +suggested patch +Should we consider moving the consistency part of the discussion to +The patch following the patterns from and Avro Mapred +Brill was kind enough to do the initial version but I never had the time to incorporate it + is the new patch +patch applied by Noel now we only have to make some test to see if its resolved so we can mark it +when a follower connects we figure out how far behind the follower is by comparing the lastProposed zxids and taking the difference +Fixed on trunk revision . +Trunk revision +When I set the size up to it works fine +Needs a test case in order to reproduce reopen once a test case is provided. +Uploading a patch that cancels the delegation tokens asynchronously as well +maybe one of the committers will have an opinion on vsThanks a lot for your work Karthick +I did not see the same bug on trunk +on compressing the entire fileThe VERSIONS file should have a entry of the formcodecif the fsimage has been compressed using gzip +Unfortunately that is part of the interface so has a somewhat broader impact +I am still having this problem init seems it was happening because I was not using submitModenone +Once that is completed please this issue +I still need to add console opts for encoding and write these new constructors into the producer and consumer stuff but the way I have the patch now I believe the properties calls have hardcoded defaults to so acutally I do think even the constructors would be safe with this in the case where no character encoding option is chosen +Furthermore I also found a solution to the problem with bean validationSo this issue here may originate from the commit in but has to be solved elsewhere +Can you point me to the class method entry point where I should set the breakpoint for watching what happens during deploy? believe its same issue with +See for more details +The real issue was that when the exception was thrown the Camel File component failed to deleterename the file even when the exception is caught and handled +Im now done streamlining the class +It is a distinct problem. +In the meantime if there is any way you can attach an app that consistently reproduces the problem that would be a big help +Bulk close for the release +I just updated it to also catch them for attributes +It should fix the broken trunk build on Hudson +Now I got the same error with slightly different stacktrace from netbeans from daily build from which comes with Jruby someone have a look at this? I think this is a pretty serious bug +bulk move from M to old and t reproduce on build +Quick n dirty fix +Added managed send for camelroute context runtime configuration to be managed +This is behind the scene work +You can safely ignore patches provided at Batiks Bugzilla and use the one provided here which removes all references to rhinoClassloader +This is a bit of an interesting issue +I suspect this has something to do with the fact that the mahout driver script is actually calling the hadoop executable even if I dont really want to explicitly run on a cluster +Do you have any indication as to why this is Markus? Which plugin are you using to parse your html +Hi Vishal Weve had discussions about Zab not being equivalent to Paxos in the mailing list of and we have a paper at DSN discussing how they differAbout the code I didnt mean to imply that the bug was there because of poor readability +Fixed +A good way to do this might be to depend on the jar task and put the jar on the classpath since the defaults are included in the jar +Will commit onSorry its a bug introduced in this patch so Ill fix here +Please try it yourself with the attached project +In my mind the required annotation means that we expect it to be present not that we expect it to be fully understood +It is that information that is used to seed the LDAP entries for the PMC chair as well as the LDAP entries for the PMC so it is imperative that the Apache User ID parsing in be robust +This jiras headline is about having the client honor the negotiate +Not directly related and I dont know if this is possible +Was this one fixed +Intermixing them makes merging harder which makes for more work for committers + it to Dimitris who was involved with deployment for +i see what you mean we need a better error need to add javassist to the Deployment for some containers +Closing. +if I throw an exception in the method the client keeps retrying until reaches the limit of retries number +both +The insert succeeds at A but failed at B and C +Jason how come your JAR file does include a mapping file? As you are generating sources from an XML Schema instance Castor will create the Descriptor files anyhow +The bug was in the shallow logic where it returns a negative value when dealing with an empty fetch requestCorrected the logic of shallow to return when there is no data in the created from an empty fetch request +The projects and a file showing the exception +Kirk please go ahead and make the changeYongqiang was just trying to get an svn compatible patch from your based against latest trunk from SVN +Closing old resolved issues +Closing resolved issues for already released versions +This will be doable if the fluent query API is making good progress +The attached file implements a fix for this bug and updates the tests +Patch. +Do I miss something to understand the issue +It includes some package poms that have to be hand created to provide geosciml and its dependencies and one unit test that seems to pass intermittently +Gabriel can i get a sanity check on this patch before committing +Committed +Closing resolved issues. +looks like this disappeared my radar +Maybe you have a broken proxy between you are the internet? +Confirming that against the master it works +committed +This patch adds a to to make uses aware of limitations for certain platforms +Forgot to include librecords and librecords +This works fine in the latest code +As to line numbers the only failure Ive seen with sane jars is the assert failure from +I have the same problem for me! +Deepak Bhole has provided a patch that has been incorporeted into the build files and will be present in the resulting build +It is fired for selection changes only +Can you point me to more info on altest svn is that a module in Subversion +Fixed +I do not see why that view should be in the resource perspective in the first place? plus reopening to get proper fix version +It is the thin end of the wedge +As I mentioned earlier I have manually tested this issue and confirmed the corrected behavior with the patch +Ask what category the old categories are changed to +bq +Yes actually both the management server and your manual run of report that the host youre trying to add does not have KVM modules +If we just fire them we could run into cases like you do with the Timer class where you can end up with a large bunching of events so we should allow for optional past due dispatch +Please can you clarify if there is an EAP issue with the code in these folders as we have previously decided not to make cosmetic changes to the structure? Tom Im working on this with Paul one by one +Duplicate of +Relaxing the issue criticality +Moving till next release some unknown importimport +I should add that I started with a clean repository applied the patch +Resolving +Method calls should be able to be repeated for the same method like in the add case really miss such a feature and like to encourage Juergen to think about it +Patch applied see in Apache Commons Digester release + +We really should start enforcing the field names we allow IMO +bq +New image looks good. +Sorry I was confused with the from +that test case seems to do a lot more than what the issue is describing +The patch is still not review the patch +UsingTBH that contributors corner page is useful the only thing I thought it was missing was how to set up the stuff +I dont really see the connection to this issue and what exactly this issue is about +Applied to TRUNK. +Changing affected version to and fixed version to the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +You should add validate to class as abstract +Could you please add them +Thanks Anil your patch is in with rev +This also uses the latest version of Netty in both cases +Im pretty sure this is caused by misconfiguration in your Tiles definitions instead somehow defining an include Thomasas Jrgen already said This does not look like a spring solve your problemTake a look at your must be an tilesinclude tag resulting in this loop while rendering the you want to check the generated servlet class at line to find the incudeBest regardsRafael +However in order to avoid timeout issues it needs to be split in a few smaller +Was that with security enabled or not +Also you will need to give some through as to what provider would be responsible for taking care of this processing +Although on second thought +One consistent way with only minimum options seems to work better for both +Hi MirekIs it a regression or newly found issue? I dont seem to remember the relevant config has been changed since last releaseThanksHoward +Good then Ill proceed with that change +I think if I had applied the patch a long would be stored in ZK insteadYes thats right +I dont see a bug in the current code +Bin Chen please bring this back to a release milestone if you plan to work on it. +I think this patch is ripe for a commit +on also behaves the same way +As LR support mentioned they made the build based on JCR api pls let me know how much JCR code API changed to the latest version? How about JCR? I assume very big changes right +Moving all documentation issues to temporarilly +ChrisThis got marked as a dupe of which I have not looked at yet +It looks like your bridging should bypass iptables though +David Blevins pointed out that we should first try to create the object then get the set of unset properties and if there are any complain about all of them at once +You should first file this againstas they are of the actual integrationIf it really turns out to be CDIWeld issuereopen with more info +This must have changed somehow because we used to exclude them +Thanks Sebb patch applied at repo revision rPlease check this fixes the issue +reopen to set fix version to again with correct fix version +So then is this complete +This was also seen on an problem +cannot implement a workaround it was alreaday the workaround +I did my test before update then merged then did my test +Reopen if this is still a problem +This decision has been the result of countless emails to our IT department complaining about it +Can you provide detailed steps to reproduce this issue? OSGI container list of installed bundles etc +yes thisd be a very nice improvement +A straight average would be misleading +It seems there is no way to easily determine whether to activate a filter for a particular viewer +resolved and no more reports of a problem. +lgtm as well +Becuase the user role account uses the dev mail address he too gets copies of mails for issues affecting him so therefore that is why the mailing list gets copies +Now clustering is being done using +Testing in progress +Its consistent between the and release even if it changed during the milestones and its too complicated forLets revisit it after since theres a simple workaround in my case. +we can have global preferences but each server can override them if they want +the content of its bound persistent workspace as modified by the pending changes in its bound transient storeThis is indeed different from Jackrabbit where the query was always executed against the persisted state of the workspace ignoring transient changes +tested +I think we should be all set at this point + +Talk to Fernando fix will be part of a second build if we create one +Yes Shelly mentioned it during the meeting +I am not sure what definitions in HDFS you referred to +Thanks ChrisYou patch is in rev +Should IMHO go into cocoon ajax just next to the +Ill run some tests today and commit it plus Ill backport this to you create a separate JIRA issue for the other issue you found +Test case that shows how this issue impacts the class +and addressed some issues pointed out in +Maybe JIT fixes not sure +This was reported in March of +Its filtering out responses on all requests but shouldnt be doing it for wsdl request +So far I have not found a if the response container model is not a template model +Pavel The user just need to setup properly accordingly to his expectations +I use groovy all the time to write utility scripts for servers that cannot resolve grapes and I always end up copying dependencies manually +Thanks for adding the table of system routines in the functional spec +Assigning to Brian if theres any other elegant solutionThanks +Verified that the patch was applied and works with a VM that implements the new VMI interface. +Both patches are the same +closed with the release of +It seems like we need some public API in the embedder to loadingverifying feel free to attach patches and please stop referring to the unrelated its this I missed it +Now it is possible to use multiple containers +New TP includes all the bundles I need +Deepal the client waits the reponse via callback +A configuration wont created until you call some remote debug sessionI see what is happening now +I fixed this issue in revision +With we removed that class to remove the duplication of the code since the compress code already existed in ALTER TABLE code path +to trunk +libUsersjoaDocumentsworkspacescompanyworkspacebin +It breaks it in several places +was added to the for +Thanks again Erik for the contribution +Changes worked well. +Things to be done Hook the collectors into the random test +How comes provider works when forkModeonce +I take that your profiling has shown that the performance degradation is due to the changes introduced by this Issue? Can you share any more detailsWeve seen a regression since an snapshot from a few months back +perf fix removed +ps +Ive noticed that Mono threading and collections can sometimes cause some real slowdowns +By correctly catching we might be able to prevent multi boots with in the same jvm using different class loaders +assign Now user cant assign reconciliation to inactive check Now user cant generate deposit slip for inactive Also I am using to show button on screen instead of showing button in screen widgetAwdesh you Awdesh. +s almost sure I may have forgotten something in the patches since Ive done further refactoring in my local working copy but please take to review them and tell me if its thanks +Hi DavidYes you can replace the NS like the blueprint does but it will introduce other issue that there are two versions of JSSE parameters schema one for spring and another for blueprintSo I made some change on the patch to let the JSSE parameters schema share the same NS with and it is easy to update the part +So we should do the other issue first +Committed revision Thanks Noble +Ive put a design doc in SVN let me know if you think anything is missing +dfs ls +I thought that it is same as the one present in SVN +Integrated to trunkThanks for the patch RajeshbabuThanks for the reviews +Neither is a subset of the other +Thoughts? Objections +IMHO I think we should fix as much of the legal issues before putting a release to the IPMC +Only issues that are actually fixed should have a fix release. +En canvi encantat pot voler dir molt feli entusiasmat o b distret embadocat +This issue was marked as LATER so opening it up again now its in JIRA to be more visible +Image that shows the irratic test pattern this causes +Srikanth great addition! Ive been running tests with this for the past few days and its very solid +This could obviously come first +So whats a practical situation in which one would choose to use vhosts on the AS? Only when the AS is directly exposed to the bad ol internet? Does anyone ever do that? +Its only required if you enable security +Im not sure this will not create other I think I will add this issue in the bug tracker of logj +We shall fix this in the next release +Since the voting regarding committing the patches on branches passed I committed the respective patches there as well. +class attribute for notification tag M +downloaded the jar from the link provided then uploaded to central and regenerated the checksums. +Attached patch to resolve compilation error +Reattaching patch +Hi Hope this can be ve added a patch for this +I would say yes based onHowever the final word may be with Stefan +Committed to trunk and. +Hi SimonThanks for the patch +Wait this happens on a CLEAN start there are no locktemplogrunpid files +I dont remember how to do this +According to the tests it is now fixed correctly. +GrahamEmily do you know know what is up here? Looks like we are getting a mess of tiles back and returning them as individual grid coverages rather then stitching them up into a moasic +Also since its such a fringe request well have to be extra careful to add extensive unit testing for this feature since its not likely to get too much exercise in the community +Ive only tested vibration and device +Will wait for and commit if all is good +For I am not sure about that +Has this been used in production by anyone? Has there been any issues +Link supports confirmation facet nowSee +Attached TRACE logs from nodes +A set of classes in the package created by Sreekanth Ramakrishnan that canbe used as a starting point for this support +For example selected text in Foxit Reader and other PDF readers is corrupted in most of the cases +Thanks Otis +When I run the test or our application with it fails +until that time will only perform well when it loads templates via a template attribute rather than having the template as its bodyMaybe it wasnt clear enough for you then you should continue on the user mailing listthis place is not meant to discuss user issuesStill we can maybe leave this ticket open to remember to implement this caching oncethe of the engine is out +The patch looks solid +However you could do the testing in Infinispan since Infinispan already includes around +Thanks Chuan for the reviewI noticed that is not added to the java library path in the downstream Hadoop projects causing some tests to fail with +Denis I added couple of others requirements +German properties translation added to svn +I think the latter is the best default +closing the jira since there is only one thread as per Sureshs comment +Yes has a patch to fix that assertionerror. +Identified another bug type not being displayedNMCONFUSING +should not NPE like that at all +This is a Spock bug +Bear with me while I learnIn any case I have submitted a fix that has a and a export so hopefully Hudson is happier this time +Not applicable to which does not include Cursor switch over from yet this is in +Hi should this bug be backported to? I found the same UI issue in the release +This is really a blocker for us as the problem happens consistently in production +This looks useful +on bens comments +If batches go thru the pipeline st one succeeds and nd one fails before notifier thread wakes up wouldnt it report the first batch also as failedYeah I think you are correct + an issue +Can you see + +That is a possibility +Are you able to work on a fix for this +We should have all parts of GA ready friday next week +see comments +I enabled the HTTP binding tests in revision since is fixed +Having some kind of info object factory in might be needed but Id still prefer to avoid moving the +The error says that it searchesfor +I think it is OK to not try to acquire permits in it could resolve the issue in +I cant publish our bisness application I hope you Alexei Thanks very much for this information +This is the same as. +Project lead changed +No an include tag will not help here Spring IDE cant handle this tag did try creating a config set but I didnt have any luck +I believe all of these issues have been corrected but this should be verified in the next CP +As you suggested via email maybe do a snapshot release and check that you see the pom etc being uploaded +After changing that tomcat groupId it works for me now even for +Is it the scenario we met here +should generate a new nonce for each op +Just mentioning in case someone else hits the same problem when tracing the client +batch transition to closed remaining resolved bugs +That package is installed already in version +Ill look into this issue immediately after the BETA release since it does not seem to require changes to the public APIOleg +The field types are always loaded by Solr so whether they are an inner class or independent does not matter too muchIts not a problem as much as it is inconsistent +jcasey actually mentioned that a that pushed this information into a specal descriptor might not be a bad idea +any sample test project to reproduce the issue +Multipart tutorial would be nice but we canlive without it agree +bulk close for the release +I am using the Sun JDK on x Linux and Im still seeing this problem in the source snapshot of HEAD from +A trivial change such as joining rn lhyzip rn to a the single line lhyzip rn fixes the problemOn the plus side this problem occur when I use the latest trunk with my outstanding JIRA patches applied +Looks cool it even uses the new API +Committed revision +Ill commit tonight +Thats really niceIf it is a proper redirect even other clients than browsers should be able to follow it but that is really not that terribly importantYou could call the file instead so itll just come directly up? Or maybe a actually is good so that one can see the different versions? Probably +Nope I believe this is something different +As a side note the most updated version of also got committed +I agree the behaviour should be selectable +If you believe order of creating of nodes by sequencers doesnt lead to any other issues feel free to close this JIRA you. +Please log in all just requirements as comments to this Jira and I shall address themAlong with the patch i have also upldated a small note that provides some information about using this implementation of +The assumption I made was because there was a brief explanation of the KAR feature in the documentation that this should be in +I think the use of forwardsredirects might be best encapsulated in the kind of strategy object discussed in +should not depends on +Im still running a similar patch locally on trunk because I think the functionality for POST with keys for list views where the view function is in a separate design document is still missing +thanks ted matteo +References obsolete version +Ron can you verify that it works on +Applied to trunk +it includes the changes from Mounir some this is the patch based on revision you should be able to apply it Peter it has been am now waiting on an update from Mounir before I can close the issue +Committed to trunk +Please set your editor to use spaces rather than tabs for indenting since that is our preferred style +Mircea think so +will fix this immediately. +The patch contains analyzers which use this new base +The goal here is to achieve data backward compatibility so to me it seemed more logical that model version knows how to read files from version but model version knows nothing about version +First reported in with less details about resolution +That was a fast reply +Same applies also to in the master. +If this exception is never thrown to any client we do not have to throw it we can have the prepost methods return false and log whatever we need in the coprocessor hooks +Interesting question is whether this should be on by default +easier document +Sorry abt the comment +so as soon we change the compaction algorithm this test is flaky again. +imo the correct behaviour should be in that case to just create the document and return since neweditfalse will be respected in that case +Modified which was not doing the parameter replacement correctly for null values. +defer to +Can it be deleted after you got it +If we delete the +Has the date for been decidedIt so happens that im working with a lot of maven projects these days with very long builds I can totally use this feature +another option would be to maintain the and in zk itself +Will see what others will say +The maven pom is generating an artifact named not +There are lots of ways to drill into a stolen database and these examples dont strike me as worth special mentionThanksRick +Its worth noting that set true in with the intent of making all Hadoop execution environments use this setting as of endJan and that need was in part identified by which also relates to! for this proposed patch which should do the same for HDFS unit tests which indeed typically do not usePlease also open Jiras and submit the corresponding fixes for in COMMON and MAPRED components even though failures due to have not been detected there yet +I did not intend to insult anyone in pointing out this issueAnd YOU are not giving ME free software +Just a unit test for the sample slightly modified version of the which just throws an exception on the th item a causes a rollback of the current chunk +Ill commit it to tez branch shortly +JIRA is not letting me attach files give an error from httpd +No more problems has been reported so it seems that everything is working ok now. +for adding an issue to implement that would be very handy and definitely seems like the right way to give access to aggregation like a count of elements +Ravithe latest patch integrates Nicholas comments reverse comparator andthe is per filesystem instance +Bruce any idea about that +For running the project it is necessary to add the libs and modify some configuration files like and +i dont understand this is already added to the automated synchronized thinking changes has not taked effects. +Ill start some investigation soon +for s patch +I dont konw What are the consequences of doing it +Your welcome +I think its just a matter of this being a little in terms of when to do this named +updated to trunkchanged valid Strategy to JOINEDupdated Unit could it be that your latest patch lacks a few things such as the changes to the class +If the units are radians then toRadians makes sense why shorten it to save typing characters +the jar in tmp will be deleted when the jvm exitsIMO means to delete the file in hdfs rather than delete it in local fs. +Theres no time to fix bugs with an unsupported dialect +Create another sales order for with product +This issue is closed as out of date as per the middleware process rules for aging of bug reportsIf you believe this issues should be fixed please reopen and add reason for reopening +Thanks everyone +I am fine with moving the creation to get like mapred but that still leaves us with the fact that throws IOE +I guess that could work +Maybe we can catch if we dont release it soon +Bump Shall I go ahead and merge this +I suggest that the portals site should provide the consolidated documentation +Closing this issue and pointing to a new port issue +maybe ivy could establish some locking mechanism via lock files or rename the directory after the publish is finished and delete it if publishing fails +Is this only in debug mode +TCK test result listing all test failures caused by the described NPE +outline of actually means draw an outline in the default opacity which is I think we will need to change these as of is the new Style view is deprecated and many points here are outdated +assign to branch in r next Maven release was in r and this was +This might also have been because of Bug which has now been fixedso this report might be just a duplicate. +Resolved in CR. +closing issues associated with existing releases +Were almost there +Would you mind creating a new issue for us to investigate this specifically +kept in session +Seems like it doesnt compile with Axis and fails at runtime with +the attached patch contains a junit test case inside is an outdated document but it describes my requirements a little multiple feasible nodes per page can interfere with the extension code +Do you have any particular examplestest cases in mind where there are problems or is this theoretical +Fixed by version . +it hangs on sybase as what I should have done in to begin with having a calendardate in a composite PK is stupid and has caused all sorts of issues +I have read Jesse planned to drop static state in ognl version +I know there was some debate on this and whether or not we wanted to expose users to posting real contact info to a public hosted app with no way of deleting it +could be Directories where taking snapshots is allowed +Anyone that can look would make me happy +Any errors in log? what happens when you start eclipse with debug anyting relevant printed out? +Can you give use the emailaddress or try to contact him? Sounds like great but I dont know him +Closing as cant reproduce +I was expecting ant +as and FAULT are optional as well so should they not be +Moving remaining bugs to +need more information or sample code to reproduce this problem +so I guess its +John A few questionsWhat exactly are you migrating too? Which files do you think you need apsite karma for? Cheers other info you need from me +Another piece of information in Spec The embeddable class must be on the owning side of such a relation and the relationship must be mapped by a foreign key mapping +I would like to move as fast as possible on to respect to less overhead for dependencies then using a common dependency matrix for the larger ones makes sense +In any plugin loading we may just need a Plugin class name and probably that is itIf the plugin has dependency jars do you expect it to be a part of the jar itself +Jan could you please verify and close this Jira. +Closing out as per comments. +fair enough +I tried a couple of options +Please advise if you think otherwise +add datatype for the patch. +SteveThis looks good +Attachment has been added with description Performans tuning has been added with description has been added with description Correction for wrong a new patch for the color problems with the cells +The pull request is still open but it doesnt appear that the maintainers are inclined to move forward with this enhancement +Otherwise the code looks fine to me +Attached is a first pass patch I created for Synonym support in the docs +With this patch a slave broker in a pure masterslave configuration will not open transports connectors bridges proxies etc +Test for this error by setting the browsers primary language to some language not in the project +Please worked around this byrm fr HOME +If the team agrees Im happy to submit a patch that makes that change instead of updating antbuild this would be a better Graeme +I can contribute for this improvement if nobody is doing it +HiCould you please provide more info eg +Stein given that youve created this patch what problems would you see tryingto move these methods away from Database +has some discussion about it also +bq +Found one more issue in manual testing which made me go back and add automated tests for this feature +I am moving fix version to ER +Ops maybe you need to turn on the IE script error notification to reproduce this +But is open to change the behaviour as followsUpgrading a database from to should automatically change database owner recorded as owner of system schemas in sysschemas from pseudo user DBA to authorizationID of the user attempting upgrade +fell asleep at the wheel +BULK EDIT These issues are open to be picked up +These are not bad names IMO +YES! Computer name differs from domain name on windows systems but James cannot support changing thatI usually use Linux servers but sometimes it is necessary to deploy on windows. +Bulk move of issues into after vote that we merge branch into TRUNK up on list +Going to commit +Marking as invalid per last comment +Look harder it isnt my fault the search doesnt workAnyway you are talking yourself +Basically it seems to be viable and provides improvement for large regions +I can see the case for a htmlwrite tag a beanwrite that defaulted to thebean for the enclosing form tag +The VM validation method was incorrect and has been fixed as wellThanks Arthur for helping test this! +Closed upon release of +I use the Query annotation and I have a query which could return n results but Id like to set the limit to because I only want to receive the first row +What is the import from? Are there any issue attachments we need to worry about +Reviewed by Rob +This has been fixed in both trunk and +But they will not be able to cover any new features you do not yet have in your s what I told you +Correction typo on my part its not I meant the patch is the one that fixes the issue +we do not use our own JAXB API impl and it works well +Im happy to link them to the umbrella issueI dont agree with postponing the resolution of the JIRA to after as complying with a known is the point of using a Web Service bindingWith respect to your comment that the current test case does not fail thats because it does not check that the data flowing on the wire is correct +Based on the comments bumping down to major +Why cant the channelFuture be closed in the finally clause? +I get the same mdsum as you do +Unless you do kind of a lot of extra work you will find yourself writing a lot of unset primitive values to the wire +Logged In YES useridTry using the same column name for foreign and primary keycolumns +For the was added additionaly forward for the +Will check it in once Jenkins blesses +If you finish it before we release beta please correct the version +Earlier you had posted an example where the super call was also made +Currently the test is commented out on all branches + to trunk + +Zhijie Robert didcan one of you test this against a real cluster or just jar? +Uses the canonical uri instead of the uri +Screenshot and model project in missing icon. +In particular we should remove YARNHOMEsharehadoopmapreduce and YARNHOMEsharehadoopmapreducelib from YARNAPPLICATIONCLASSPATH +Integrated into trunk and. +zip kit work under linux as well +log for rat check in +I assume this is what you also got +See comment by Vijay above +The trunk patch applied with some fuzz and does not exist in +Attached patch should fix this +Have you been able to try this patch? Ill bug Ted to check it in and your builds may be healed +We will currently always have ambiguity +Might want to make it clear in the comment then how that case could occur and also clearly define snapshot path and path +I sent you the contact info and legal stuff a week or two ago +INFO Running for rows took ms +The original code where the problem was found was ported from a solution that worked fine using ADF controls +the fix looks right Ive applied it and slightly modified the test to be configured from code rather than from config file +Feel free to start creating that would define the desired BOM for the next release +but i cannot close the bug since im not the reporter. +Please let me know if there is an error in the test steps listed +Committed in the closing all issues that have been in resolved state for months or more without any feedback or update +just committed this +I am gonna place an inside this will maintain a fixed pool of threads for sending +Hi Andreas LehmkhlerThanks for your time and solution +Feature request for pool to work in arbitrary environmentsI agree with Mark that it is not obvious that we should aim to support +Uploading a jar and error messages in a second +Reopening to get rid of Fix Version +The config is unmodified but the conftriggers dir itself is missing +but this change would take many days as every single page and component would have to be touched + for the new patch +The original problem with create account page having errors has also been fixed. +Feeds are handled differently now. +I should note that I also changed the property name to +I ran some test cases and they passed though I have not run the full suite. +what and where are the differences? I think we can only begin to make progress with this when both incorporated issues as above are resolved. +I have been using it on my site for some time now so I know it works but then I do not use annotations in general +Andy please take a look at this an comment if you have a moment +The default of is rather low in order to prevent users from killing their region servers when using large write buffers with a high number of concurrent clientsSince we now have protections against it and aint that low +Ready for review +. +I am assigning this Issue to version although it really doesnt require a change in the code +Upgrading mine to hadoop to test it +and the are a better place to track this. +It also fixes a NPE issue seen in when optimization runs on udfs with no arguments +Let me apply this disabling to trunk +Also is time across the cluster consistent +trunk all tickets +also on for looking into this Gary. +I was just wondering how it would be possible to have a message with no recipients +We are generating hadoop test jar in the target +The Java activation framework java mail and spring context support jars are added the file +Patch seeing an occasional race +This is a problems in the Microsoft C library as well +Running the build shall update the patch soon +Resolving as incompleteNot enough information provided to reproduce this problem and it looks like this is not aXalan issue but an application class loader issuePlease if this is still an issueand if additional information can be given tohelp us reproduce this one. +Alan does the get build for the regular run? If yes I have a unit test that reproduces this +Thats not good +This task is complete now +It just has abstract class I am looking for its implemetation that is Do you know where is that +It should stay failed until the issue is resolvedWe need to have something like FATE in accumulo to queueretry actions taking several steps like splitmergemoveIt is a to keep a history of region state transition +We will be addressing the larger issue here in +sdw and +I have not looked at sqlline code but how much does it depend on actually SQL dialect +Pull it back in if you think different. +do you want to do it? +This getsaround the issuesThanks again Wolf +How critical is this for you? Were in the middle of a MP release so need to determine priorities +This bug went away when the process deployment page was redefined +The suggested patch +Hairong we should also add the image MD digest into +There may very well be a relationship but it is not quantified at the momentWhen the OJB DB implementation is in place it is likely that it will become the working PSML implementation +Great! I dont think anyone is working on this so feel free to do it +This one seems to have become a dupe of +I was just working on updating the images in the doc plugin +Fixed branch vGACP in revision . +Summary patch to sync the compiletest youll need SATJ see another with SVN diff including the intermediate changes +Ill work in this issue too call me a docbook guy +Thanks for quick reaction and response +I dont see a class in this patch +Thanks! +We may need to do more than updating pom thing for the coming up CXF +this can be generated by but cannot be by cxf +Canceling patch +I added those title attributes to the headers with some soothing explanation +since there isnt a way to set the JMXPORT except by using custom scripts i figure most folks just dont do it MAXHEAPSIZE was already coded to be configurable since by default it is not set +This seems to work but it wont deal with a slow master +Richfaces document is already created +Using CSS for layout and positioning would have huge benefits on this project +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +And just as you say theres no need to overdo the TEXTTABLE function as dealing with field length or referring to other rows could also be done after the primary extraction +appl file is seen again I should think +To generate the documentation extract the content and run grails doc on the content root +Minor questiontypo In should modify be called as add? Both add and remove are modifying the entries but modify here means add +Relative instanceDir is already handled relative to directoryIm really sorry for the confusion +It is against revision but since other reviewed paches are pending for that revision i would prefer to diff against a newer version +And well try to address the problem in the next version +Very good Sean! I had exactly the same results even without cutting the final predictions to bound values! I think its a good start. +Closed with the release of. +Thanks! +I dont think we need to rerun tests for this change +Fixed in bpel see bugzilla +The connection is only accepted if theres a thread to handle itYes but not from the normal pool +Committed thanks +Does that suffice +comments +test case +Document describes using the jetspeed realm with tomcat single sign onDocument to be placed in xdocsguidesPatch will follow for xdocsguides to include jetspeed realm guide in the guide index +Can you see if things are now working with the latest autotools commits +Sorry cant give out more info than thatThanks for any help! +Eli a zombie TT causes lease expiration +nasty workaround but its fixed +I mentioned earlier in this JIRA that I was seeing test failures whentesting with addresses on my local machine +I will commit soon. +Committed. +Patch based on the beta code +Integrated to trunkThanks for the patch AnoopThanks for the review Andy +Nevertheless this restriction may constraint too much the kind of traversal you make over a DOM treeIn my example I removed the node +Uploaded patch that fixes both and the issue +What about removing the file webportalsrcmainwebappconforganizationI think this can be removed because profile is using and cluster profile is using two mentioned files and I am not reopening since it is only a minor thing +The search term is retained in the box until you click on the link for a new resource category +Thanks Mayank! +But this means security is not simply a different implementation of Hive security which was one of the goals here +I also added a small comment in the test about why we are skipping the deprecated values +Performance is still an issue +I cant reproduce the issue here using boxes directly connectedtried and. +But I still have a bizarre runtime issue to resolve with the +Which patch +Dont know how to resolve it +tested +This is by design +HiI am working on functionality to select an existing account while converting a leadThanks RegardsShradha ShahiLabsLT at +If not the message will get lost +sh files to launch activemq +Closed. +I had found my mistake it is not a bug +Im sure there is a bug on the use of x class field which scope is also avoiding the bug you noticedBTW well done +Append check for an infinite calls to converter method +Released Beanutils reopened issues should be looked at as a part of and evaluated. +The patch looks good to me +No longer required solved with +Interface documentation for the last few remaining interfacesI also removed the interface since nothing used it as far as I could tell +OK Ive added the new constructors for +I suspect that this has something to do with the fact that this plugin is written in Groovy +What happen if there is no context? Things will keep working because the context is only relevant for the bean container +Do you want me to post some instructions about how to run those tests +The original patch you send has created the fault message Qname using the operation name space +Furthermore by only extending this method we make the class unbalanced regarding functionalityThus a new name is needed +Would you be able to upload a patch? changed files based on you have diffpatch files instead +It happens on DB as well +I believe the root cause of this issue was resolved with we no longer rename files when writing a Lucene indexAre you able to get this issue to happen with Lucene? +Thing is One should be able to use as the Hibernate second level cache even if one dont cluster +Thats funny +fixed at with the release of +The jar containing all the resources is attached +Set the nant property in order to generate help style Chandler from Helpware has given permission to include as part of the download +for commit +the should be done by h EST +Thanks Nellya! This one looks good +Dont have time to get to this in does someone else want to push this forward +Although I dont know why a custom FIF would want a directory. +Actually what would be better is just to use gravatar images +This patch contains the code and tests +Add placeholder but I was not able to locate where to edit ln here +It would need some improvement before it could be integrated into Struts because it does not work with Messages externalized Strings +The Ant bug only appears when you set the id attribute on a task. +Ditto performance regressions +Hi FolksHere is the patch forHTHAnkit Jain +Reopening +Also what is on your classpath +jsfplugins +I got halfway through a fix that extended the remember me classes to put a into the registry +Moved to next milestone release +This is must fix for upgraded env +Patch with the fixesCommitting shortly +I guess the second file is for when is not availableDoes it make sense to extract classes under crypto into their own module +Lets postpone this to since I may not be able to deliver a stable solution before +This config file is used by Spring framework and AMQ components +I looked quickly at and do not see this as a problem with the parser +Ive committed this and merged to. +Does engine check visibilities for accessing XSD models +Scott you said A test with the latest CVS code outputs xC xBC exactly as Joe specified +Updated patch for IPN +Thanks for writing this up so systematically Rick +Marking LATER pending course of action +Whats the purpose of versions if you cant improve the source code and documentation over time? It doesnt matter that version says X version says Y and version says Zit only matters if version s Z is incorrect +Thanks for the explanation Henry +Thank you! Would this then be documented in the docs at some point later? point Ill add warnings about this to the javadoc as well as to the ref much! +I left that line commented out in the attached java file +Could you check the two new screenshots andThank you +What is the file system with which secondary NFS server is formatted with +The main problem is to decide when to save value property and when not +realized today that this is a worse problem than I at first thought +Also I do not think the return type on the current method could be changed because the return type is not part of the signature +Fixed in master thanks Anthony S Baker via github +Resolving as fixed +accidently closed + for the patch +This patch fixes a bug of incorrect handle of pipeline heartbeat in yahoos hadoop security branch +Please try latest svn or Axis issue still exists with Axis against +section added to admin guide +Is a completely new splash expected for JBDS +Thanks Daniel +Ive also signed up with my address . +Could you give it a try? +Navis Since the tests passed can you please commit this yourself? Thanks +CR Whats New is reviewed +The final join should be run as a join ideally +Docs Q I am assuming this will not change what users have to do +Is this really a major issue? Does this cause the site javadocs to mess up? Im not sure what to do about this really +Once I am fully set up I will review and commit this change +i was trying to setup GWT SDK but I cannot find google settings in Preferences it is called Web Toolkit in Preferences sorry +SO this kind of a example Endpoint impl will help folks +AJP shouldnt timeout in the first placeSo it could be another system property but it wont be in the domain model +Cool I just committed this +Whats the line length? The command of operation in the shellFor others yeah I will add the Constant and also the api usage in +Michael do you have any ideas +Okay finally getting back to looking at this +It also looks like you have used the dropins folder for at least one of I am guessing is that there is one or more plugins that is behaving unexpectedly starting up early and forcing JDT to start before we are gives me some room to explore +does not workCan you file a new issue for that? And having an example document as test case that illustrates the problem would be helpful +Applied the patch thanks! Also kudos for finding and fixing those typos in the code and did not apply the change to since that would have made the transport layer which it currently isnt +The wild card for regular expressions is dot star +Ive commented on this in +to and trunk. +Furthermore reusing means I dont have to add a separate beastalthough some modifications may be required +Thanks Jarcec! +The Overview Download link is only for KarafFor Cellar its a specific download link on Cellar DownloadDo you expect that the Download link contain both Karaf and Karaf Cellar +Project with which you can reproduce the problem +Total file size of the devserver zip goes from to because the included libraries are more specific now +No IT +Carstenis Mike right with his assumptionI suppose that something went amiss from your last comment rightregards there have been no further comments Im closing it. +I tried to point out that not returning GML by default for a request is actually not compliant gml should be requested specifically +In my testing the Instagram module works fine with this change +Applied the patch +The simple patch should be good to go then +I think its close we do without the etc and just change freq to long +Upping priorities for jiras were looking forward to actually fix in doing validation of requests is not by any means trivial to solve +After switching the code to use an ajax converter the deferreds served only to remap callback arguments +All inner classes are analyzed properly +Manually tested! +Defaults maxidletime in to instead of and updates the various ipc keys and defaults in Client and Server to use the relevant constants to prevent similar future mistakes +Bulk closing stale resolved issues +Sounds good! I totally agree we need to support Hadoop +In any case it appears that it is jvm dependent working on windows and not on linux +Updated patch appears to build okay on Debian Ubuntu +Hi XiangtanThe geotools part of the patch looks good +If he confirms that copying files from another project would not be considered bad mannered I suggest we just keep the files as they are +skinning is added to the guide +As for requiring I had started out with requiring Collator rather than +assigning away since not going to work on it now +Had a discussion with Ashish and Raghu about thisSince we already have create temporary function the above approach may not give us anything new +for the to trunk. +Fixing +It was Nhat Vo who attached cardemo +The completed docbook xml could then be generated into PDF HTML and other relevant formats +I use and have never seen this error +This changes another reference to in the testing +. +Id vote for assuming that you mean KILLEDUNCLEAN and FAILEDUNCLEAN +The output of the REST method will be a list of Results containing target key and action status for each target +It only use the trick for repair because that is the only place where we open lots of sstableScanner but I suppose we could apply the same to normal compaction and lower slightly the memory consumption there tooAnyway the patch includes a unit test to exercise the new code +Ive tried the way and it works fine each cached parser is garbage collected with the thread its associated to through the a thread local reference +I have modified and made to define nonTx and Tx datasources +I wrote the fix yesterday I will see if I can merge the two together and post an updated patch with tests +Even the STABLE message does not contain the highest transmitted seqno there are only highest received and highest delivered seqnos from each node +As advanced user I personally would like to know why my plugin is not running +Wicket follows the focused element all the time +Patch applied at +fixed in a separate released! +committed fix for this issue +proto file and it just slows things down +Now that work is well underway with AS all previous community releases are +Is the sign allowed anywhere in the of the email address? Is abc valid +If you look at the SSVD algorithm for instanceyou can do the subtraction as part of the original projection very easily +is a problem with the geronimo version i thinkin the top level pom geronimo is referencedIm specifying in my local pom under platformsgeronimo +added a bunch of licence headers in trunk in rev in branch in rev a few more headers in trunk rev in branch rev a few more unaccounted files remain in the web modules +Rejecting since simple test program for replicating not supplied and has not been validated against this if test program can be provided that fails against TRUNK. +Please check it +This is EPP specific +Its OK +Since it may have uncommitted txn in its txnlogInternally we use so I didnt put more effort on server related features +Thanks! +Sorry thats r. +If that process doesnt exist or isnt a ZK process then the lock is an orphan and can be removed +This sounds good to me +All child tickets resolved +We would like to keep as a short release and +Sorry above by this files I mean png product files +bq + +marcus should I just pull this into? We can resolve the bug once its over in and only close it after confirmation +Thanks for making the change bluedavy +Thank you Mukul the fix seems to work +I did see some issues in Findbugs and put a scenario and bug report in the Findbugs +all should be fine. +Robert Yeah I removed some left others but also added asserts so tests should fail is regression +More work but certainly the better solution +Lets commit the patch +It looks like this has already been done for both and +Shouldnt it be possible to use a version thats more easily programmed againstOtherwise patch looks fine and builds the same Constants file over here +The attached patch should be all thats needed +Tried cloning but could not edit this issue after cloning +bq +Resolved in r. +Attached is the patch that will add to Axis Message Context +Yes it works +Utility that converts mongodb collection record into a flat file of to be injected +googleing for this exception gives me hits seem to be an issue for glassfish tomcat and other appservers +I appreciate the good work put in this library and will do everything I can to help improve it +We havent seen any problems while the client is typing as of yet +I have now merged this code from the code line to the codeline so bug is fixed in both versions +Indeed thats something we need to do as soon as the website is updated and the new release is announcedI will update our release guides. +Ok I think Ive gotten my head around what needs to be documented hereFirstly is not in the EAP instead the EAP has an equvilant file refering to jmx users called should tell a user that they should start and stop the server locally by using the script which will ask for a user on the OS +I encountered the same reasons unknown to me running the plexus runtime by +So correcting this mistake by introducing a work around into and using it forever is badI did not exactly understood what was vetoedTo clarify Correcting this mistake by adding translation code into trunk is horrific and as you pointed out well be stuck with it forever +With Mb we changed back to a single module build +I checked Schema and loading Its fine no changes needed +If I attach the file will it be helpful +rarely its slower +Thanks Andrea. +Thanks Amareshwari +This is likely a configuration issueAlso +the feature is in there and now available in the ui right? There is just the caveat of correctYes and warning about dialect should be removed from this feature after will be resolved +Sorry I missed that part +If class A is only available for the groovy loader for example because the file is not compiled then you will have this problem +Any chances of getting a new patch attached that could be used against SVN trunk +The plugin classpath should not affect the build classpath when executing other things +Fixed thanks Rick for applaying the patch +Hibernate does not know how to map a LONGVARCHAR column returned by a native query thats why you have to use register +Thanks for the patch +this record since is now released. +Please do not commit until the fix is reviewed by by Blake +logic to be extracted out into a generic view class that can then be used in a dialog or internal frame etc +Please see my comment on on not supporting symlinks +Thanks for the quick fix +Closed as requested by +It shouldnt with a fresh database +Then consider a plug point for replication policies where the extension code can decide what to do when presented with an integer +Possibly having a very visible warning displayed in the console could work +Not for commit +thanx a million +see logsYou need to configure +Committed fix by Jason Porter in r +I just committed thisI made one change making rather than public +I agree it likely not related to your patch Himanshu +Hey Abiola Aikucould you please inform me if the wsdl you used was something that axis generated with wsdl +Not sure why it isnt for AegisThat said not supporting the minOccurs thing from the aegis config file is definitely a bug +Ill update this tomorrow +Hi i solved this problem in a way lets say a workaround so this Issue can be closed limitation with pattern seem to work like there always must be anempty response or request element and complextype defined for the wrapped Xfire generates the code from WSDL in a proper way +See +Thanks Nick +Here is a test case for this is a README in the tar with directions +Sorry +Fixing inconsistency between code and +Craig do you have a unit test for thisIm not sure I understand how when max size is hit items might get flushed out of order +Closing as incomplete +Indeed I may have been losing sight of the forest for the treesthanks for all your work on layout! transition resolvedfixed bugs to closedfixed +No longer used must be replaced in the test. +Yes I agree that this patch is more likely a workaround but it resolves the issue at the moment without any detected regressionsSo Ive committed the patch and created to track the better patch creation. +instance in eclipse +I have already created a jira for the documentation and I have to do some more modifications to it +Sorry I think I wrote the description wrong shouldnt selectchildnamespace work? Selecting all child nodes in the given namespace +Note that this feature will cause some overhead on the rules processing +That target just substitutes true or false for the SANE literal +Take a look +v patch Committed revision in HEADPlease anyone watching this issue report further issues if they still experience problems with the new functionality introducedThank you for all who have reviewed and provided comments on how this is working for your use cases +I also have doubts about eclipse project references see my comments on which is a related issue +I committed these fixups +The more work to do is to make the report model more consistent +For instance let say in your you inject Bean Autowired Bean +Thanks +The notifications dont specify what changes have been made to Nodes so cant determine what keys have been modified or removed +Crash dump for this to SVN +I tried your suggestion of putting the after the in the and now it works great and intercepts the url correctly +So I think you are perfectly OK to insist upon it indeed I think its an improvement because otherwise somebody might restart the agents process with broken locks and your patch prevents that +Are you interested in our software +Moving this to as discussions are still onI think it is reasonable to assume that all the tasks progress in similar speedSpeculative execution condition should be based on thecondition that a task progress considerably slower than the average +Try it out +Disables region balancing until all regions have been assigned +For more sophisticated handling of default packages IMHO extra functionality is needed +There is a patch there +its certainly fat here are the results on runs +Can you please look into the time out issue I reported on the th + +Mirroring has been set up for us +There is more work to do +Resolving as Fixed though I still need to update the public sitePS +Thanks Robert for taking care +The table then behaves as if it was set to but the last column takes all the space that is left and gets too wide +committed to trunk +If you want to pursue then please take this there. +The win is nice in that something is either or and there are some classes that do not need to redundantly set reduce tasks to when they run on the map side +The release audit warnings are related to html files +Howeverour rebalance algo currently considers only of loaded regions as the loadfor region servers +Fixed +Maybe the only difference is that nothing in the tree is propagated down +I can commit it today +I generally agree but would include Class in the list of interesting and useful things to configure externally +As the user interacts with the applications checks are performed to see if the data being requested is somehow linked to the Organization Context the user logged intoIt seems we have traveled parallel paths on this subject +Please create a new issue and try to find an exampleI also added some more and the translations of Thomas comment from Thanks for the contribution +Youre rightThat is exactly what i meantIm sorry for opening this ticket but i havent found anything about using sql fragments in the tutorials or documentationSo this ticket is resolved and can be closedThanks for your help +Sounds like the solution here is to include both the JMX Console feature and the above archives feature on the webtools site so that they can be installed at the same time +The default is false due to the spec as well as for backwords compatibility +I guess the easiest path is to secure the admin page fully it also gives away information about the this as wont fix as has been retired +However it doesnt work if delete is replaced with +These are the exported files +Subclassing the types in question will take a considerable amount of time though +Java isnt slated for release until next march so at this point its a bit early to worry about test failures +marking as resolved since we cant reproduce QE please provide exact steps plus attach the files slava is asking for? thanks +I was wondering was there another bug filed for the regression introduced by the patch? I did not see one linked to this issue +To set a default for the promoted reference the target should appear on the composite level reference +We know +dyslexic title change +We can and address other issues +I just committed this +The above changes were performed in the featurecleanup branch of the sandbox +I already have an issue open for fixing cleaning upand making those consistentI dont think we should add a generics parameter V to Lookup especially if itself is stillwired to float +Done thanks. +the goes to solrsrcdocumentationcontentxdocsimages this is committed thanks for the contrib +Check the history in the Server class +marked as to be fixed for and an issue for only +Sitemap builder now validates sitemap files against the schema +Phu semms like much workI am back from vacation now and will review this and your other enhancement inthe next few daysThanks in advance +Invalid lines should be either discarded or the exe should fail on such inputs +m +Heres an update +See for a possible explanation of what is going wrong here +svn upU incubationUpdated to revision . +Heres a patch that implements this +See patch for which addresses this issue also +Committed thanks +I have install opensolaris x and have build solipse +I needed to use the underlying logj logging implementation +Looks good to me for now +And I have run mvn test all the UT passed +output for error + + +Wolf explicitly pulled this field out and replaced all references to it withreferences to similarity or idf +Note that this patch have doc impact on CREATE TRIGGER statement syntax which is described in the ref manual +It will be filled by us +We can focus on the methods needed for the Cell interfaceSo Matt you suggest that i give an uploaded patch with the methods added to Cell alone and remove the for now +cannot reprouduce +No reset required. +Taking owbership as unifying the authenticators with the realms is something I am planning to review further for AS +Create Backup I dont think this is going to be supported in JBAS plugindeployDir this needs to be setup as a traitdeployExploded Emanuel can you investigate whether we could add another argument to the +Is there any functionality that we absolutely need now +the comments on this issue by all parties were extremely helpful +Even if we all agree that this can only be a solution Id rather keep it as a safety fuse +Henryyes I am using +When you refer to the details of the wiki I guess you mean the vague hint that besides the its possible to implement custom but which will be exposed by CODI +Closing resolved issues. +By default is invoked only when distros are prepared and schema directory is included into jar distro +I will create some user instructions soon +If the problem is still the same underlying POI perhaps these POI issues should all be handled in this ticket? Or should a new ticket be opened +Everyone else might get los in space. +Worked on typography a bit and attached diff files +Hmmm this feels slightly strange implementing at the indexing levelWhat are the adsdisads vs just lowercasing for indexing and capitalizing at the presentationapplication layer +Cheers one +At least I found a plugin that can be used in a profile to download the dependency +Thank you for the patch. +This was also done in although I suspect by a different method since there appear to be copies of the console parts there +However the point of this JIRA is to collect together all of the known limitations of our small device support so that users can figure out at a glance whats supported and whats not +In this way the code that caused deadlock will be invoked only once and does not require synchronization +Sure Ill try +Please do not reply to this emailIf you are receiving this email it must be because you sent an email to my old email address +This has been committed in trunk +Nonetheless we can still perform the double checking to cut back on writeLock acquisitions +find name svn also returns nothing +Might be already fixed by other recent checkins +At the least +This allows you to create multiple registries with different back ends without causing conflicts and invalid keys +Ah well +The test program I ran was in the project with the s option you can callbq +Specifically it works very often except when there is something strange in the certificate which makes the structure slightly different +Should we be concerned about updating versions of dependencies for reasons in a bugfix release? Its possible that updating one of our dependencies could break a project that depends on Avro +Fixed in trunk and in bigbang +Okay I have reproduced a litte unit the issuetrunk A workaround for you would be to use multiple s if possible to reduce the chance of hitting this issue in CamelYes we have pervasive use of dynamically constructed routing slipYes it seems so +Committed patch from Ron +I press the INTRO before entering the whole title +For some reason SVN refused to recognize this as an updated source fileAlso attached is an updated class should go into common +Brad can you verify that the patch fixes the problem in your environment +I think we can close the issue and go ahead with release +This may simply create way too many outputs and kill the FSDo we believe this is the current behavior or an ask for an extension +Whats the best way to set custom attributes on the Coprocessor? Eg I want to tell the Lucene Coprocessor where to look for a configuration file in HDFSSee and +Pushed upstream. + a test that current trunk will failThe bug is in since it only cares about String but not Path +We can certainly agree that this issue has nothing to do with or MozillaPer Sebbs comment above we have been presented with a license that gives us access to unspecified IP under conditions that we ourselves dont meet +Our priority is not to have our webapps mergeable with custom webappsIn the future we will forsee deployment classloading so custom logic can be added in a deployment. +Great feature! But it seems to be missing a capability I need +and closed. +A fetcher can be removed after the request is made + for the to trunk. +Perfect thxIts working fine +I need my client code to use a real API rather then needing to build the correct String query representations and parsing resultsAre you able to give us some more information on what your client code needs? Is it just being able to instantiate a Shape and then convert it to a queryable String formatbq + and look that weve not seen this report for quite some time now on nor am closing this out as its gone stale. +I had major compacted a column family on all of the nodes then did a simple pig job to read the contents of that CF then I got a lot of minor compactions for that column family +If you include a patch I dont mind how many issues you post +Thanks Rod +The second one illustrates the related but different issue test compile target is not included in the test classpath when using a non standard directoryNext time I will provide exactly one lib patch and one spec patch +Set for now have the and annotations available for tools +Added a basic test case for tld scanning in revision +Hence why the FAQ and the Camel in Action source code has been updated. +If it passed validation it is given to the Generator to generate the and +This is a patch against the current trunk and is intended to be applied on top of the changes represented by your first patch +Let me know if this would be useful +Sorry +Can it be closed? +Geezz +Wouldnt it be nice to roll all of these fundamental concepts in to a single project like SISI agree but the elephant in the room is the LGPL library JTS +Patch addresses making TEZ a only deployment +Patch applied with a slightly different approach that should work also for factories mapping a class with static transition all resolved issue that did not see any further comment in the last month to closed status +So seems we might not entirely agree on the exact naming but that we agree its not really a big deal +The exception message has reference to o instead of o. +Initial trunk patch +Please attached the new unitary test and the related patch +new revision from review board + +Using the hex of the hash I added a authentication mechanism I added the ability for +Im really not sure any change on an should lead to all bundles handled by the listener to be updated +Reduced case is +I had problems with the UUID generated for while communicating with a using the windows communication api +Thanks Ram! +Theres a lot of ceremony required for quick changes +I thought this was a simple edit +Ive committed some changes to SVN to ease implementing a solution for this Added resolveId concept to ivy core +Now that Object Cache is backed by Cache does that also mean that the cache is distributed across different in the same Teiid Server cluster or is this as with +Anyway glad the worked for +So if the dst is a directory the src checksum file is incorrectly moved to +Why would you ever want an entity to be partial? then it is not an entity +resolving as fixed in +Yep I think there is still a problem with the Archiva dav as I couldnt open it via the Network Places whereas I was able to open a dav repository via Network Places +Now Im having problem uploading the jars to this ticket + +Reference +fixed in rev +Has this proposal been raised on the dev list for discussion yet? Icouldnt find itYes a discussion mail thread was initiated in dev mailing list it occurred hours later than Jira ticket update as the proposal mail was stuck in outbox due to issues in our Exchange server +changing affected fix version +Enforcing token auth always on Using irrespective of kerberos Fixed testsOne side effect is that Unmanaged AM is completely broken now even in mode +I can upload a relocation pom pointing to another jar in the repo +In the attached diff I took the quick and dirty approach of using axiomsoapfaulttextsetbasenode and axiomsoapvaluesetbasenode to set the node pointer back to zero after the tree has been freed +There is no guarantee that the read does not complete before the callback is assigned even if it is asyncIt should be considered a requirement to assign your callback before calling read +For example change +Thanks Dag! to commit +Please check +As long as the instance variables are private then how come we cant get rid of the mBytes variable +for the lazyfix +Ill be adding a sysprop override as well as changing the sysprop that exists currently for stacks in a backwards compatible manner +Trunk will follow soon +junit output for tests +Thanks a lot for the reviews everyone +All the code compiles and appears to run correctlySorry I got attachment wrong please use Rick +If it were a problem for File Install it would be a problem for the framework in general I would thinkOut of curiosity how many bundles are we talking about +I like it and it looks like it would scale fairly well with the number of tasks but it does not give you the same information that the swim lanes would do + Could you provide a patch for thisJosephno need to make yourself one because I already had this functionality around see the patch + +Executed mvnDebug X deployThe result is my code is never executed even though maven debug clearly says it is selected for runtime +We are finally getting somewhereAdded the logoFixed the message issueSync the pantone of the main menu with the original siteDid some h style sync with the original websiteAdded some creditsNeed to test alternative sites to the homepageSmall style missmatches and coloring issues +Based upon my work to add support for polymorphism I think that this code really is not used at all and hence shouldcould be removed +With my testbox ran iterations with one test failure +Thus it is not possible to write into columns that are outside the PRIMARY KEY +has two problems maintaining the year is not scalable +Our understanding is that at one point this had been road mapped for the release +The problem persists in Activit +already fixed in M +Thanks Gunther! +But we have three items which data our address we have and not it finds the first loaded item which data address we need replace our successor of by successor of +Looks good to me + +Dave +This is a one line fix +Still have to figure out why one test occasionally takes min on windows though Generally it takes under a minutes but every few runs +Likewise for NOT +And these UDF functions are simple enough so I dont see much difference between testing with constants and with data from table. +Is this still an issue and should this be addressed now or deferred to HV +From the attached log I do see that the same split file was read again and again +So this one +No time to look myself right now. +the patch looks good to me +Committed to trunk and x. +downside of this is that I will have to go through the urls and look at the jars there to remove the junit jar +Lets try to commit it today + +In PSML Great +Configuration problem +The attached patch corrects the entries in + +At this point I think this is better moved to +bulk close of resolved issues for release. +I spent a little time looking into it but traced back to the storage handler and decided I was in too deep since I dont have time at the moment +hopefully the problem goes away +See some discussion following that on the the dev list +That seems to have fixed the problem +committed has documentation issue from patch when reordered cause build to break +This issue is duplicated by this issue is the same asIt was moved to RFPL as it seems that it is Metamer related +Native authentication cant be turned on until credentials are stored for MARY +Heres an updated version of Ivans patch +Verified in +There are some limitations related to lists and tables but for regular text upgrading to FOP should resolve your problemsThanks transition to closed for remaining resolved bugs +The version of Struts was invalid didnt existin Struts +OK +Ported from trunk to branch at subversion revision . +The resides within the lib apparently refers to what is found there +Its done now and weve clarified the correct procedure going forward so its fine. +Ok im clearly losing my mind +The best proof is that in my attempt I used a vanilla jee Eclipse which has no Egit installed +Does this issue also apply to the branch +To have RAT run use the mvn audit profile by issuing mvn P audit followed by whatever other commands you would like to run. +What do you think? Not suggesting we reopen this +It ought to be possible to suppress unwanted anchors +Im seeing the failure too +Hi Jacques Thanks for the suggestion review +If I can configure in a single repository to get stuff from why cant I configure a single repository to put stuff in? +Definitely blocker +I too have been busy on other projects +Agreed. +Hi I would like to contribute to this as I have taken this code and substantially refactored it for and much more of the security features such as remember me anonymous authentication etc present my code is very custom and I would have some trouble to split custom stuff from vanilla portlet security +Feel free to if you still see this issue with JBWS or greater. +Tried the patch +I am happy about your feedback +A Product Store can be defined using the Catalog component or by using the ofbiz setup utility +Looks good to me Wanna make the log message debug +What problem do you mean? +Sorry for inconvenience +It was applied with minor changes in and +on trunk it was already fixed +Slight changes will be needed if we adopt the groupId change for the api jar to be instead of +this was fixed a while. +This is a jdbc helper frame work and I think having support for stored procedures would round it out +I added a test and committed this. +But can you see the pdf? Right now i cant see the pdf it will just show me blank black screen with the Error Unhandled error +Fixed the find bug java doesnt allow the derived class to reference a static variable from the parent class +alternate locations from within the classpath remote http and local file urls all work +Hey Per thanks! Ill try and get back to helping finish up this issue at a minimum Ill get a commit or two in on it today +Ive committed an additional change for setting the ws ctx to null in the target bean after the invocation +Seems your patch introduced various whitespace changes +How feasible is that considering we have a time frame of weeks to complete our projectThanks I believe the feature freeze applies to hadoop core and not contrib right +We cant use newline as terminator since we support multiline is a special case where newline is in string literal +Assigning this one to Kris since he was the one who originally created the patch +Should open in a new tab Would be nice to have it like the other ones where it has the modal dialog that explains how to use it show how to import to Eclipse maybe? What is the icon really +Just need to it to edit the fix version +the main test program +Another possibility is the in Jakarta Commons +Committed +Space was restored! +bqWhy does need to set NM related env varsOriginally set those variable for the AM Container +Committed to trunk after addressing the comments below +If it still doesnt work Ill repost the patch +This issue no longer shows up in tests with SOA ER +I havent seen the failure again after the fix went in +AM calls a startContainer +svn ciSending assembliesfeaturesframeworksrcmainresourcesbinkarafSending assembliesfeaturesframeworksrcmainresourcesbinTransmitting file dataCommitted revision +Need info about Sybase version that support JSP info about servers was added into guide +Invoice search and creation form in Accounts ReceivableRegardsSurya patch completes the implementation of relevant screens in APARThanks to Sumit Pandit and Awdesh Parihar +Patch with unit test demonstrating that structs with nested containers do not break cleaner unit more patch cleanup sorry +Ok I see now +Like I said above for the more scary changes I think its important that we give some config flag to enabledisable the new behavior where we can +Suggested code as a patch errors +Bobby as I alluded in my Hadoop Summit talk PigHive dont have to be limited toIAC you are describing a lot of work which can be done they arent necessarily part of +Ive committed the patch to AS trunkCan we get someone to try this in a cluster env +Thanks for fixing this +Id just like to note that I built it locally and it works with no errors +However minValue has length chars +Consider tracking some metric such as facebook likes count or twitter friends in a social media app +validates schema agreement after CREATEUPDATE on KSCF and CREATEDROP index ALTER TABLE +I think I know what the problem is but Im astounded it didnt crop up while I was building and testing the archetype +Committed to trunk +Hopefully we wont have this problem anymore. +It does not related to your patch +can you investigate the options here +Updated the components associated with the issue since we have two instances +So that I will try to make use that in this case +Also affects. +I guess this is ready to commit thensvn stat srcrefM srcref +is this a regression or +Rebased against trunk and with added support for compression in the write throughput test +A javascript event telling me that tapestry init has finished would be helpful +It only returns true or false +Jordan the dirs were created but we cannot switch to using them until you populate them +is what we are using in ORM as well so it might be a good choice +Struts is model layer neutral +committed +Verified at by Anton and Spark. +There are no codes or anything for us to check +I am sorry but we cannot implement this change + for the patch +yep yep +Nadeem Can you select Apache SIS for your project? I see it is blank now +It seems ready I will commit soon if nobody objects +The uses the data and displays them in the same way +PR for master merged in master +Hi AshwinNo rush Camel is still in months developmentThanks for the explanation of the Netty annotationI suggest you copy the which tests how it works with concurrent requestsresponses +My main problem is that I have two types of dependencies +Closing issue as it was included in version which has been released +A blocking RPC client without any connection pooling etc is not very difficult to write and for administrative purposes that would be sufficient right? If we had such a thing available as a relatively small C or python library available would that solve the issue just as well +Maybe we want to avoid messing with the tests for now +Ive created a feature branch called vmsync on thisYes Ill try to address background job execution issue together with the sync logic as they are closely related but Im not sure if I could address concurrency issue for all orchestration flows will be focusing on laying out a in first stage and select one typical orchestration flow to exam the effect +Sounds good to me +Just out of curiosity have you tried this with Grails +copy templateiso from zone to zone and it is successfulWhen we configure the basic zone management and public ip addresses in different subnet then the routes are configured properlyPlease see the below mangement ips are in and guest ip addresses are in root route nKernel IP routing tableDestination Gateway Genmask Flags Metric Ref Use Iface UGH eth UGH eth UG eth U eth U eth U eth U eth UG ethroot So in real deployment the basic zone management and guest subnets will be in different subnet and routes will get configured properlyIn that case copy templatediso from zone to zone should success +Forgot to mention I also removed the JEXL jar because Hive does not depend on JEXL for a long time +And now that the mochiweb branch has landed this is also fixed in trunk +Or wont help much in any case +needed by applications again now properly recorded against Fix Version as well +so you are saying the is actually not even imported into the jsf project +IMHO that should do the test +I would commit this soon if nobody objects +Just checked the test results and this test has not failed in the past odd runs +I am surprised that this problem is not solved for related workloads at your companyDhruba it used to be your company as wellAnyways as I stated in there are cases datanodes would not come up when disk goes to a bad state +That should encourage the fixing processI updated the highlighter test and added as a test dependency such that this can be run with ant +su camelthen cd clean installeverything is okbtw it should depend on camel version but not the version in the generated Jack explain more about how he saw the problem step by issue is with the archetype version when created with the Meclipse plugin +Slip out of t reproduce that problem any more. +However this variable is getting reset to the full path of the sharedLib folder +WHEW +ksslSPNEGO changes have completely broken the patch +Reopen issue if there are any problems. +testing for hudson +Currently it doesnt have a Fix Version specified so it wont count or be reported on the changelog for is no fix +Strange +This looks like the old code path that had the deadlock not the new one +Voting readonly and failing to do work are not the same thing +Committed +The dependency is marked optional so it should be possible to use Compress in an context without XZ being present. +I did some of the formatting manually so I hope I didnt make any mistakes +Im working on adding the tests for this +Want Mylyn? Heres a single bundle of Mylyn stuff +I am +This is part of making Oak proper bundles also requires removing all MK stuff from the module +Super patch looks great +added config param +The reason is we are exploring different class loading models for the embedded and server to reduce the memory foot print distribution size as well as controlled dependence on the third party software +PaddyI have commited a fix to camel trunk +Reopen to fix site as well +completed at revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. + +selenium test controller +Your patch is in SVN +Committed to trunk +How can a stack trace be expected and desired behavior? It this actually fixed or is it a wont fixHow can I explain to users why there data disapears +In the document it is worth mentioning that since is the classpath hadoop jar is still using the jar +Ive just committed this +Hi Tatuthanks again for your patches +I am able to apply v to trunk using git applyI generate patch using svn diff in cygwin +This patch fix the problem for has been added with description Fixed patch for transition resolvedfixed bugs to closedfixed +Looks okay to me +I dont see any problems +It takes multiple minutes for me on ecffddcccdcceecca as well +I am curious though we have had this solution working without any null for a number of months so either we are doing something funny or there are some more issues with axis that I am unaware of +Marking as Major with Complexity Low +Changing this to throw an exception instead of returning a connection at a later stage should be very easyIm leaving this issue open until junit tests have been added +Sorry for the mistakeregardsCharitha +Added synchronization locks around calls to the registered loggers +Phillip since the dependent issue is fixed in jboss messaging trunk can you build jboss messaging from trunk and use it for testingThe nd EAP build which is scheduled for today might include only jboss messaging s a good idea to continue testing and not get blocked on an official build +Ive just stepped through the deployment process and the problem is that and dev map to the same pathAre you already working on fixing the or might I look into creating a patch +committed +Thats the intended supported usecase. +patch with revision bfdeecebffbaafbadedfaccThanks! +Hadoop with proper support for is now released +Fixes this issue + thanks. +I will commit it shortly +Josef + +I doubt that this is desirable +The logic and explanations here make senseStack Could you have a quick look too? This is tricky business and some more eyes on it would be good +According to my tests this patch doesnt change the behaviour of other failing or passing tests +Help button is absent when call New Project Tools Web JSF only happens if you use the New toolbar icon and select JSF project from here +Its just how gossip works +For example is it a good strategy to run all the tasks for the first job in the queue on a subset of the nodes in the cluster? Maybe it is +Wow great catchThanks alot for the patch! +The implementation currently just takes the first character in the string as the character value +Thoughts on incrementally updated viewsIn a first version a View CF could be defined with a normal column family definition plus a function that transforms a mutation against a base column family into zero or more mutations against the view column familyAt mutation time inserts can immediately be transformed into inserts to the view +Hmm that could be related but Im not seeing any exceptions thrown by the close calls +Moving to +I dont know if it is by design or a bugFor the moment its by design +Any one who wants use this can implement thisI meant this could be a utility +Same changes I was going to make +Looking at I do not see this patch rolled in there +Group Management issue on the right you get similar problem on Tomcat bundle +bq +Error was in the file +Anyone hereIt is abnormal to include dependencies in the manifest that have only a test scope +Its not a problem for the exec program but it could be for some of our tests +Could work do you want to explore that in the spec changes +This patch fixes Cloud Files for users with DFW endpoints returned in the service catalog +Looks allright to me +People seem to have been happy with browse being AND for the past years so why change now +Problem already fixed in. +HiDo you have any news when iPojo will be availableThanks that the new manipulator is released we will plan a release of the core and composite +In this case dialog creates at start new images and then each time you switch to Edited properties tab + test close if verified fixed. +any way we can verify how many keys we are missing translations for +Ddas Can you merge this to the PB branch +Minor update +One patch apllies to trunk the other to the sitePlease tell me if this is ok I would commit it before packaging the RC +I dont care that we get it exactly right the first time just that the API have some standard way to say replicate highly we can improve the implementation later but we should only have to change client applications once +And I did not see that exception happening till I updated to the recent trunk +If we choose a name for the thing we are pulling like image then we can use makeimage pullimage etc +Yes that makes sense +If a user wants to override the logj properties that are bundled with the JAR they only need to make sure that their copy appears first on the CLASSPATH +type +I attached the project that represents the current issue according to the reporting guide +I dont think this should be necessary as the Apache directive rewrite the location URL regardless of the context path at the back end + +I am not sure if this issue is active +Would be useful if there was an option to treat responses like asproposed above +Thanks Nicholas! +defer all issues to +Hello I am interested in working on this project +The question isRestoring this in plexus? or reverting? +I know this are really implementation details but still +Looks good from my end. +Thanks a lot. +Thank you Laura. +This may be closed by now. +Thanks Kathey and Myrna +Ive attached a middle ground patch that follows what Martin recommended an exception explaining the problemI realize we could just leave it since the does give a stack trace that the user could follow but having a more helpful exception wouldnt hurt. +Previous DIFF of Jeremias merged to currentshouldnt we use components for as many as possible of the ODF Toolkit +David exactly right +Is this bug up to date? Should be solved +if the artifact name is en why is the +We need a reviewer for this patch +Please check whether you are using the Xalan contained in JDK +The test jobs where not creating any inputoutput data network traffic +Just to be sure what is the expected result of the decompression of the given value stream +This is just a potential starting point to whoever wants to do the job +I will check codegen of Java backend +Nit looks like test code required format for the latest patch on addressing the above nit +Is this correct or is PREFETCHT better +Need specifics +The include the inlined finall blocks which they should not +Please reviewJuliusCan you think of a better API to deal with the SSL engine configuration? like the full path is the troublemaker +generalized types of map reduce and scan on Array on r +have committed this to trunk and will commit to soon +Making blocker on to branch +Bumping +Ill raise another bug for that. +It separates a new injectionScore API out from the initialScore API +I will be committing this momentarily +Hi AlessioWhat do you want to know +Verified in additional and issue not was found in affects version is +Ashutosh yeah i understand your point of moving the validation from client to metastore server +At first I had block mode disabled but it was failing? The delimiter is case disabled block mode went through my scripts and changed all instances of go to read GO and it works! So close this as not a bug but maybe add that gotcha to the docs and also delimiterType is not documented in the config for all of your hard work on this thing very useful plugin. +With the new proton work and varying cleanup methods already put into place this task seems to be a bit out of date +not for beta not super urgent +is ready for committing +I dont think that this module is usefull anyway +This ee profile switches the build to and excludes the package in the configurationThis works already +We now have a draft website build and published +File type value for directories +Are you using the latest and greatest iBatis release +Verified on RHEL. +But we should check thisAnother place where the checks were omitted and instead AIOOBE is catched is +configure make and off to the races. +Thats what I thought you were trying to do was looking into the explicit option as well saved me some time +Migrated to test +Thrift in particular is important to maintain so we shall keep it in core +You have to add the filter in +JacquesI havent looked at the code but off the top of my head I think we could setup an Ajax request that would return localized dataRegarding using color names in the CSS in other css files I have made a list at the top of the css file that shows the colors used and their hex values +Please check that it was applied as expected +The result could be small enough and be a good alternative to the patch +Does anyone run it against IBM jvm? I did not get any failure for from sun jvm +Assigned to Wesley for definitive answer on this issue +The should load conf before setting default property values that would ensure that the correct config is used +Still get compilation error but its not immediately obvious why +I had to reformat in Eclipse to get it under control +I had a similar problems with spaces some time ago I think Ive forgotten to remove tabs as well Ill give it a in. +See for further details. +Assigning to Arnaud Blandin for investigation +Yes Christian please file a bug +Fixed by passing thru variable from pom ant +i will hit the forums +Yes its fixed and works nicely with the release +I get the same exception +Please apply the patch and push the build +I guess the next release of ODE is pretty unsure isnt it? So this ticket could hold back a new version of SMX no +Try to use modal found the problem when you combine the modalPanel component with the ajstatus and you try to open a modal over modal the navigator blockedI just have to change the modalPanel component for another javascript modal based into a for your help +I have confirmed that this problem is resolved in +The only solution at that point to have the node join the cluster again is to restart it right +Thanks Devaraj K +Still is always better than protected if at all possibleFiled to fix this +patch applied +Patch for +some features also should be documented later as the component has couple of update tasks +The only exception to this is when we get contention for the processor locks +Replacing the requests errors and timeouts counts with Counters and adding a reproduces all the existing stats and gives you the histogram of request times +I know there is a workaround but I need at least a stable nightly build that works in order to use it in my organization +Patch with fix and test to demonstrate issue +Also the chinese docs have references in the following filesWill file another jira for thisbq +Oops looking at some of those other files I guess youre not considering them to be Wrappers +Thanks for Bruno +JBC testsuite ran fine as did the ejb clustering tests from AS +This additional change made the implementation more robust good +Last patch had some bad characters in it from the license header somehow +Here is the patchRegardsAshish AshishIve committed your patch other files in rev +Thanks. +Thanks +This will make GA and should work with any version +I also attach a java code to demonstrate the problem +I cannot reproduce the problem here +This is same as but targets alone +I am using Safari on Windows +fixed this by accessing and Namespace properties of using nice one +Please do not make multiple changes per issue they should be tracked individuallyAlso please do not make formatting changes just because you can no longer stand it especially at this stage when we should only be doing bugs +But since these are all estimates seems like there always should be some cost thatis linearly related to number of rows scanned if there are actually not rows +I think Pavel is right that we need to give users who rely on the warm cache some kind of solutionOne alternative is we could provide a cache converter tool that fills in an cache with position information from the index +Now we have a meaningful warning instead of NPE +supported operations +Bulk close of issues +Eli fixed this with a quick patch in svn commit dbcc +can any one tell me more +Patch for branch +Fabien Marsaud thanks for your explanation and workaround. +Mind filing a new JIRA Luke? Else holler and Ill get it done. +Map support is checked in. +All the MTOM tests inculding the security ones seem to be holding up finethanksdims +Are we willing to break things for our users so soon after +Not for lack of trying though +The key is to have a name +Maven Shade plugin +thanksThe issue Jan points to where you can select multiple dialog windows is still there +please get the older version of the wsdl to see the problemthanks u please try googleSearchs wsdl? i remember seeing similar constructs in that +If false the bind authenticator treats null or blank passwords as authentication failures before attempting to bind +Please speak to the Tomcat project +I agree thats why the change was made in as you dont want to delete all your data after reverse engineering a production database +Executed from outside of sqoop home directory +What benefit is this large set of changes going to give the community +I filed to track this +I rewrote it so that it fits our purposes better and so that we dont have to deal with licencing issues. +I am running tests +But I do appreciate your eg since it isolates the problem and will make it easier for debugging +Looks good +I should add the tests pass with Java too but with a log message indicating that use doesnt work with the Java version in use +Hello GeorgeThe fix looks good thanksBest regards by Andrew. + is generating invalid setter signatures in createAccessors so any code that uses Property fails on openJDKIts generating invalid signature in both but the implementation of the bean Introspector seems to be parsing the signature instead of using the compiled Method object to determine parameter typesTurns out its a typo the raw field signature was used instead of the generated signature +Agree with Hemanth on having two knobs one for completely disabling preemption and one for enabling a dry runRegarding leaving some heartbeats before killing tasks I agree with your arguments and am fine with leaving that for now as the timeouts may indeed turn out to be in the order of minutes like you are saying +Ive added an attribute to the http element which maps to the errorPage property of the default used by. +If we do that the penalty of scanning quickly might be fairly ignorable when you consider the other variations in work load a DN is exposed to anyway +we can close the sessionWe will not actually close the session just delete the log files. +The patch only currently works for field values not for values or for +The solution Im working on is as follows +Proposed approach if a managed bean is declared as an EJB interceptor it may have wrong lifecycle interceptor methods and they should be ignored +Thanks +Thank you Marcus + +checking in +is used only inside a method in which is never used +Scheduling for specific fix version for this. +Laura you changes look good +I think a hotfix should be issued with this fix in it +ok cool so we are on the same page +Moving to This has mostly been addressed but there are a couple dangling issues +Allow setunset coprocessor table attributes from shell +I was able to add a Maven Installation with this ll put this in the backlog for a while in case anyone wants to try to reproduce it on Windows but Id suggest just setting there a situation in which you would NOT want JAVAHOME set on the machine that is running Continuum +logout mabye is a method from a previous developper in my companyAutologin does the Cookies handling +I guess we need to decide if this is how we want the grouping done. +The logic is copied from Hadoop +This will make it easier and clearer +Task couldnt be closed because guides is updated online periodically +I also enabled shallow iteration in the system test +Perhaps this is an upperlower case issueYou created the schema TEST but are trying to compress the schema testtry changing your procedure call to pass TEST rather than test +Im not sure we need to do anything about my standpoint returning null makes more sense than returning an empty string so Im glad they fixed it in +Thats what OL is doing and works great I dont undestand why uDig is doing extra work to get an inferior result. +So I tried to the Web service by deleting the servlet configuration from +Unfortunately I dont think you can post a patch for this sort of thing so youre going to have to find a committer who actually uses svn +svn compatible diff +Oops let me have another go at it +Commit changes to trunk at revision changes only make traditional applications work for due to no is generated no way to known the dependent bundles while starting the application +I am currently reading the source code to Hadoop Im slightly stuck in the gap between what appears to be a version implementation emerging from the works but even so I am very interested in contributing towards the effort suggested by this ticket +HiFor those looking for a solution to do LDAPActive Directory logins that bind as the user and then query the roles using this binding I add here a possible implementation +Not sure thats true +This issue is fixed for now +which will make me feel better about pushing this to the branch unless you can explain why we should make the other changes +As per guillaume if you dont want a workaround please use Handler +We need to change to put the key data into messages sent to the broker +Linking it to as per Aruns reference +corrected +The config stuff is still too specific though +This limits concurrency +We havent heard about any breakages yet does your definition and your look like if anyve reviewed the initialization process and the reason the occurs is that I had an entityManager reference to and a sessionFactory reference to +All javaDocs are now available in two places one under reference docsAREA javadoc and the second under the Javadoc heading. +Thanks +The VM seems to not find collator support for the no locale +But in both of these cases the error code is a request level error code +Committed to trunk and. +Proposed patch for the orgapacheservicemixsoapThis patch sets the wsdlResource field on the object which facilitates reparsing the wsdl when reloadWsdl is called +Maven simply has a bug +Moving out of +There is still problems for as it is a bit hard to find the correct maven style artifact from the reference style location +Ill look into the nd one +Please reschedule this if its not certain to be done in the next week +lets see if I can click the mouse correctly this time +neat are you putting this change through Classpath itll be great to lower the Builder overhead +However Francois has been testing fairly close to HEAD +This is now available in the latest snapshots. +If a long running client wants to delete its temp files some time later it probably should add the temp files to a list and then use the list to delete them later. +Have you tried enabling from shelldisabling from shellIf I wanted to enable hbase lzo where are instructions on how? Can we include in javadoc somewhere +osgi bundles are too +You canshould use JSTL instead +The only existing file modified is the base pom all others are new files +Brandon I cant reproduce this difference eitherIs debug logging turned on? If so try with it offWhat replication strategy snitch are you using? Is the dynamic snitch onThe only piece of code on the insert path that these patches touch is calculate which is cached in anyway so its not immediately clear to me what could cause such a discrepancy in performance for youIll do some more thorough testing and profiling tomorrow +The second step will be to create an instance of portlet that you get as a bea consumer and add that to a portal +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Username doychin +Thanks Hari +ah? Awesome +The issue is still valid though there are inconsistencies even within a single test case +How about outputting the classpath to a file +I tested the fix in the trunk and worked perfectly fine thank you Karl. +This is a sample exhibiting the behavior described in this bug +Fixed with svn +committed +I am assuming Jenkins comes back OK still +One possible design for this is to create either local or Break Continue and Return classes mark a given context as a finally context +Not a functional problem so not blocking release +Only during EJB mode the looks it from +This mechanism works for simultaneous request of the same block +Obviously it should be committed and now if its done but if nobodys moving on it +Im working on it +The TT should use that time to decide when to send the next heart beat +If this condition is evaluated to true an uninstaller will be created otherwise the creation will be skipped +Dont stage those chunks where was renamed to +Is this still a problem in andor +I did this grep r l appendconf charmonizer xargs perl pi e s appendconf appendconfginstead of this grep r l appendconf charmonizer grep v +I havent figured out how to upload it to review boardThanks +Thanks! +Hmm normally on first run JPA will create all the tables defined in the schema +Access URL is httpslocalhosttestajaxtest click on the buttons several times refresh page the counters should now reflect the clicks click on the buttons more times refresh page the counters do not reflect click after first refresh +What is the difference between the twoAs for it can be reproduced iff the original compilation is done with the version of was an in valid bug report I had manually installed the newer version of the groovy this JIRA is a problem +This is pretty hacky though no? fetching metadata should not create topicsthat is like a getter subtly changing values underneith you +Verified with JBT. +Yeah this been bugging me for a while as well +But even in that scenario the default behavior should be to copy all the files for the sake of wroteAlso this is not the best place for the question but is there a particular reason why there is no configuration type existing for jbossI dont believe there is a reason could you please open another bug about it and assign it to meThanksMatt container has support for existing local configurations closing this issue. +looks ok to me +settings may influence the build it is better to remove it from SVN I will do that +I just committed this +Can you put up the full logs somewhere +The is necessary to avoid choosing smallest compactions in cases like and getting screwed later because all files are dissimilar +svn revert coresrcmainscalakafkaserver +This patch also does not affect the acceptancerejection of any other patches you are referring toAre you this patch on any technical grounds +Closing as wont fix +I like LOCALSERIAL over ANY +Ive invested considerable effort some time ago to make cyrillic fonts work +for the patch +This is the XNIO worker is not an Undertow issue but rather an XNIOWF one +Works if text is of a bulk update all resolutions changed to done please review history to original resolution type +Looks like RB is down +A new zero amount replacement order gets created which can be checked from Order List +Bulk closing stale resolved issues +I am now ready to commit the connectors reorganization also once I hear back +Please change the param name to be +So if this isnt closed as wontfix then I would suggest the version be changed to +Hi TimYour mirror will be back in the list soon thanks for your supportGroetenHenk Penning +The new test is now runas part of the in and the old test is removed from derbylang +dead servers is subitem of servers and current tasks is subitem of the tasks section +This issue is introduced by the enhancement of the to support the stax API better +is a tool in project. +Sorry for the confusion +I think the Gentoo guy was using +Based on this fix I also added a integration test case to test the REST functionality of Axis +what about to rename it to? It will be more clear about what it does +I didnt really try to figure out why the test was failing +Thanks Jarkko Nathan I didnt add anything to the change log given the nature of the changes let me know if this appropriate +Moving to +All in all its just a String representation of a integer +Add reference to distribution +I was using Jetspeed in a projet with a lot of hits and I had to maximize the application throughput +This still needs to be tested prior to being committed +new patch only checks if the transaction isActive if necessary +Could you attach the please +An to be thrown for getAt when index is greater than size +However the problem is trivial allocated memory is not issue is invaalis because the SOAP fault exception model was changed with updates to the Exception class hierarcy +Sure I can! Do you have a patch for? Or only trunk so far +Ill investigate +Source files used to reproduce the issue +Thanks! +Closing all resolved issues for. +This issue should only be resolved when the final version is actually referred to the releases Felix Framework in Rev +Fixed in mater the group tab is now looking at the category value in the workitem configuration so for example if in your workitem config you have category Integrationit will place this workitem in the Integration group +combined the fix into one patch for prev version +Lowering to Major +I talked with Raghu and understood his concern is that the number of calls to DNS resolution might impact the performance of network location resolution performance +Committed to trunk +At the client side we could store the method names as a hashmap +I discovered two issues +the example above includes the workaround take this out to see the exception +I am a bit new to JIRA so +Weird +HI devs Seems earlier patch having some errors Therefore i updated previous one with two new patches and for all changes +ie the serialization format is likely to change and the change is not likely to be compatible with its old format +very good ideawe can even minimize it more +trunk +No Problem +Hi I just find this project idea and have interest on it since recently I am coding online CF algorithms on other computing platforms +You can probably resolve this +Enis Id say new issue +Due to this is no longer relevant since the JSF scaffold is not used +Committed the patch +Maybe I dont have permission +Dont add this extra ve testcase file a new jira for this limitation +The configuration is invalid +Modified orgjbosstestclustertest to incorporate new test +But now I see just out +For reference we only use Tika core and parsers on the following two classes +Notably the html ones which are generated in runtimelogshtml +Yup it needs to be fixed +Im still mulling over how best to write tests for this which is a tad difficult on a machineI tested this manually by setting up an HA setup where each NN itself binds to but has actual addresses for the other NN +Your java doc should be probably be included in the hadoop java doc so that it is available on the hadoop website +Now it looks better with the new configuration readreadreadreadreal user sys It would be nice to have a doc describing how to use this feature correctly and its relationship with the previous shortcircuit implementation +If there is a problem please open a new jira and link this one with that. +here is the diff between output and command line is definitely different Forking command line DProgram Filesinseeapplicationsjdkjrebinjava classpath deewu +Can you enable TRACE logging for the and categories and try again? There will be a lot of output so beware you might want to log those categories to a separate file +Should be fixed via new API +I dont expect much changes in this areas +No problem there too +I dont think this case should be treated differently +If the component is needed for a temporary conversation then it should just be created a Startup component is already present in the conversation context when the conversation is begun then its already personal is for making sure components exist in the root of a conversation stack so works for can look at doing a patch if the above sounds reasonable +MarkoIt should support credential delegation however I havent tested it +If you type nothing it succeeds +are not used when looking at this file it seems also that default values are not used dBasedir dIncludes dExcludes d defined but not and Eric could you please try with the version in SVN trunk +But the problem if that the was unable to fetch the status of the job from the Vinod means to say is that the could catch the exception if it happens +I was thinking that in some cases where it is not a hard error caused by syntax that it would be useful to see results of the validation tests +utilsrun work +srcmainjavaorgjbossforgespecjavaeeservletWould they all be moved to javaee directorypackaging? What about the only java file I found with javaee in its name? Should it change as well +Verified in nightly testing. +Thanks Owen! +Thanks +Dtest +fix looks trivial lets include it for the next version +I think we can wrap such information in the exception and throw back to the publisher +Updated patchWhen I get a chance I will build an update site based on this patch that is compatible with meBenson in general it is easier to apply patches that are created from within eclipse and are created with the workspace root +The reason for this behavior is that youre adding the as a module to the EAR +Or we see the bottlenecks because of the proposed workaround +Registering registor when the value is null is probably redundant but the case we are inspecting would happen if the entry already contained some old value. +Over reduction in CPU time used by the +Im going to test this latest patch with the applications I currently haveThanks +EricMy concern is that when a new user checks out mapreduce trunk and does ant binary the compilation should not fail +You dont need it unless you want shell access there the web login can be used to admin the Hudson jobs +Applied to the two branches +Hi Hairong can you post the details of your experiment? What was done and what results did you achieve +reopen for backport. +Patch to add Eiu option to wsdljava so code generation will ignore and log warning instead of an exception when an unexpected element is found while parsing response +Thanks Kev +Patch rebased for latest Trunk +We can commit this since ryan added a separate test for his case +Please close it if starce is the same +I am also tracking the dev releases of +down into would take a bunch of messing regening the site and messing in poms +Actually it is Rampart +In the UI Knowledge Packages are also referred to as simply packages +Yeah this might not be so bad +ok how about you can also partition on the column family? Its the next item in the hierarchy doesnt add too much complexity pretty straightforward +Thus it appears that there is more to your test case than meets the eye +Actually this simply looks like youre not getting the headers +Merged into for. +Imagine if we have hundreds or thousands of nodes usually we have only a few merge coordinators but many participants +Memory graph for running script +I must say I personally prefer the explicit usage of the encoding because I am often on systems using multiple encodings and relaying on the default encoding has often been proofed to be a bad idea +Looks reasonable but is there any chance you can also implement STATUSDEAD and STATUSDEADANDLOCK? Would be really awesome if we could support those cases as well and it probably shouldnt be that terribly difficult +We will want to indicate that this change can possibly break any automated backup solutions people have +You may not want very unbalanced racks nodes +Therefore this can slip too +This issue has been fixed and released as part of release +mThis will mean your artifacts will try once to seed the cache and then never there should be no need for an exclusion feature. +If you set it to Annotated classes must be listed in the validator will complain if there is an mapped enitity but not listed inYou can also change this option after the project was created by going to project properties JPA tab. +I found another leak. +What Im not in favor of is changing the ZKFC to ever cause the standby to become active without some fencing method succeeding +Resolving as Fixed +Worst case do an equivalent of a grep on the windows include directory? Good luck +In any case its not about the parameter but really about the return type! Ive removed it and the standalone TCK now still passesIn most cases the defined checks for Serializable are too strict in +Jouzas is there a release of CGLIB due soon +Makes sense +because is saved to snapshot while renaming itself +Please allow me to attach a patch with the enum +test error unrelated +I think that on the push dialog is not even there and the UI is waiting for it +Push it. +In this particular case the issue arose because bundles were overriding packages provided by the system bundle +All tests passed +Issue Resolved with revision +As far as managing the configs in ZK is concerned what if we invested in creating a dedicated resource for doing that? We already have a zookeeper module after all +Thanks Eranda this looks great! Ill look into committing this soon +Should be easy to fix forAwesome! Thanks! +In we concluded that the right thing to do is simply not expose these to Thrift describekeyspace. +Not applicable +Thanks HariIm actually against a deadline on something else but Ill verify this after the weekend and commit it got to check this properlyIt works fine and the checkpoint is valid but even with patched in it still hangs at the end +If thats the case then Id like to propose that the details of get hidden from the end user +Docs are now updated in trunk and CP branch +does anything need to be backported +Same patch but runs with Increments and Appends that atomically update two column families one with VERSIONS and one with VERSIONS +Verified with beta builds and EAP closing. +You should use javacomp +I wrote that article btw + +I tested this by looking at the output from jstack and could see a bunch of blocked threads relating to the index updating +I wrote a script that keeps accessing the endpoint +But its that what the developer using compress should be doing +Isnt a full URL that is a URL pointing to a file resource a legal base URLwhich will now fail? For example if I set the base URL to the file I amprocessing lets say filehomemm the previous version wouldaccept that and resolve references relative to that while now we are appending a and it will fail +socket write operation in is now done without holding dataQueue lock dataQueue lock and ackQueue lock are merged into one removed the above findbugs warnings +Attach the full stack trace pleaseIf you have a test class to reproduce that would also be good +If you are not using Python you can run +small patch +I can not reproduce this now +Attached patch fixes JDWP agent code to check for JNI exception in the appropriate place and removes affected tests from exclude list +version of patch +As long as neither garbage collection nor such backup mechanisms are automatically run by the repository I think it would be better to handle such coordination on a level above the repository +I ran svn up on a different directory than what I patched and hence got an older version +There is some value in the STOPPED status +In fact this one cannot be reproduced +A list of directories and the target file size may be enough as it would catch directory config issues +pls help me +So just reject this if you do not agree +We need To unify to one Path impl able to accept a Path or String prefixparent context Tests for creating new Paths and obtaining the right context via get +Committed revision Bowing down to public demandChanging the methods to package private to insulate the public API from internal changes. +Thanks +The doesnt work +we should figure out why it aborts +Then lets just drop that for now +Should be easy to implement +I will do this before commit +I attached the patch that shows what I did +done +Ive updated it based on the information youve provided but since I dont use jscookmenu its probably better for you to update it directly when you find errors +Built using it loving it! Thanks again. +YuckAlternative proposal User trying to override the values via the properties results in an errorI dont want to introduce back doors around the ASs socket management layer +it would be easier to manage these types of issues if someone would reviewcommit my maven support +Im seeing this too with in code which is evaled after being received over a connection from MRI +Simply removed from the list of affected have removed the affected and fixed versions it was a bad move +Heidi could you try the attached patch to see if it resolves your issue? Also be sure you have a search category linked to your catalog with all products you want searchable in there +generating certificate on agent needs class +Fixed in r +needs to be fixed before we can start encouraging the feature +I guess the easiest is always better if works correctly +Closing released issues. +the original patch to trunk patch would need approval of the release seems to have broken the HDFS tests such that they no longer log anything we were apparently depending on those for the test classpath +So the package structure goes like adaptor cache crl ocsp pathvalidationBut all the sub packages are for certificate validation +All tests passed without complaint +Nothing new from last as rejected since its not relevant and much easier to see when reviewing the whole doc set jira +Cool! try to generate a SVN patch or tarball is also OK +Replaced Stacks with Red Hat Application Stack as suggested +Sample project attached +The error occurs when theres a null value being passed through a patch for this tests yet included in this +This is a patch with a test for this issueElements inside lists should now be indented in a way similar to statements inside of closures +Thats the only timeout we have everyone else is content to wait forever while the attachment receiver is stuck waiting for a message that will never come +Makes more sense to be in CVS anyway +Thanks for the doc patch for this Kim! Ive committed it along with a clarification to the mention of frameworks in the getting started guide with revision +The only problem happens when while previewing it reaches the end of the stream +If the value is false the RPC client would throw an exception if a token is not a available +This is a key property that needs to be set to really secure a database so should be corrected +Agreed thumbs up for check in +should have closed them earlier. +Please excuse the lousy title I used on this issue due to lack of contributor interest to implement this feature +Ill check these issues out more thoroughly and update when I have clear answers +to trigger Hudson tests +only question is if that is possible +This patch also fixes some streaming javadoc comments that were causing warningsBefore committing you must perform the following svn commandssvn mv srcexamplesorgapachehadoopexamples srctestorgapachehadoopdfssvn mv srcexamplesorgapachehadoopexamples srcexamplesorgapachehadoopexamplesThe patch will then fix up their package and class names +I hesitate to add a global lock to open since it will slowdown mapreduce jobs when all tasks open a file in the beginning +Looks good Devaraj +I belive the bug is not up to myfaces +Please submit a patch with codebase. +Fixed in revision +Chinese japanese needs to be fixedItalian and Catalan lead to a blank pagespanish and portuguese have some corrupted stringsnorwegian german korean russian french for Catalan and Italian this record since is now released. +Moving to +sh scripts and the dpk i you also add the permissions errors you saw +fix committed in r +well this whole thing is a crazy idea I dont anticipate working on it until a lot of other much more important things are doneAs for your specific brand of crazy idea I think log shipping is a well proven and simple method that should really cover the majority of use cases + svn commit Sending productsamplesquickstartsconf productsamplesquickstartswebservicewsproxybasicAdding productsamplesquickstartswebservicewsproxybasic +Small patch on the to use to package so that its +Done +Fixed +The reason behind this is performance issue that when ever user needs a project with the axis web services support and with the webapp console this validation happen and it took some time +However if you must do this you can disable the installation of a security manager by specifying the no command line option when you bring up the Network Server + +Hi LukeIt seems to be a conflict with another pluginI am trying to setup a test project where i can test all plugins that i use one by one with the mail pluginBest regardsJan Schumacher +Thanks Navis +I would propose to mark this bug as fixed leaving it to Dirk or someone else toreopen it if there is an issue with this behaviour +Thanks for posting this EmmanuelIm having trouble applying the patch +Is there anything in the logic of that does not provide? +Id question the quality of an application that had code paths that could result in the same listener being added twice +The check to see if a jar has a +Hi MarkI have the same concern than you I proposed another algorithm +Cant switch assigneereviewer for some reasonanyway and committed Jounis patch thanks Jouni. +Give me a day or so and lets see what comes out in the wash before making any changes +I can see Dans point but I do believe that it will be good to have one central place to look at all the exceptions irrespective of what code caused it +Would you please try this patchBest patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I can send a patch if needed +There were no OOME encountered on this run +my thanks to everyone who contributeditd be cool if someone submitted a documentation patch describing this new functionality +Thanks in advanced +The links to nightly in should be removedDone +A place holder for dependencies will have to be generated this page wont really be needed until is out so its not going to slow down the RC process +Duplicate of. +I think maybe its worth holding off on this one until we consider implementing +Resolving as wont fix since this is a commons validator issue +Sorry for the confusionBasically if the bean was declared inside the service element it didnt work but if I moved it to be a freestanding element and added an id attribute things started working +Patch against r of the branch +Added marking for entrylogger without this it possible that the bookie will start without loading all entries +I disagree with this +And patch +the task is too general more detailed tasks on guide restructuring and update were created +Well that was the case up to now +I did some manual testing of this patch and in doing so discovered a bug which blocks this issueAfter applying the patch on I confirmed that this patch works +RELAY would do what you want +I will contact you directlyif I have any questions +The former I think is an important purge to do for +this is the test cases pom project +fixed findbugs release warnings +Killing! INFO Removed completed task attemptm from trackerhosthostxxxxx INFO Error from attemptm failed to create file benchmarksiodatatestio for attemptm on client because this file is already being created by attemptm on +The double star is interesting a method is easier to implement and doesnt require IDE developers to catch up Im not really sold on either of these d be happy to have some more feedback from others as well +Still actual +bq +Hint the to modify that default +I dont this is related to KVM or any hypervisor in general +s idea would be there a workaround way to do this +To complement what Eric said The idea of the login selector is actually to have one single mechanism to create login forms while the actual login and authentication is done by the actual real authentication handlers +missing a final bracket +It it possible to count the number of connection open? Then we would just have to check that it does not increase after two callsBut may be there is a real issue behind this flakiness even if its not what the test is looking for +Works for Collections and MapsArrays has a small bug in the preventing the feature to work but it should be fixed soonNext step will be to test keys of maps too. +Mike +In theory the changes are not stopper for tomahawk to work with Mojarra even without the change +Stefan Do you mind if I take ownership of this issue? The original complaint appears resolved with a compound operation of two add steps within the compound operation to add the domain definition but there is still a problem regarding the required type to define the module options I really need to discuss this further with Brian as there is a backwards compatibility issue to this as well as an issue relating to validating happening at different points + The build still needs to produce a distributable which can be downloaded from the website +patch committed thanks shravan! +Hey Aidan Can you review +It happens to me too +This is now fixed + would improve this +Thanks for reviewing Ted. +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Cool I will review it tomorrow! Thanks James +fixes some more bugsStill need to fix that modification time is somehow changed in a snapshot +ZIP containing to reproduce problem Derby tracing JDBC since this is a we cannot retry to create the connection +I think this ticket can be closed +The simplest way to do this which is would be to use a static int or some other fingerprint on the integrity of the tables +I just installed it and the problem is solved. +bq +plugin updated in r +The build just finished +Thanks Amareshwari! +Now moving target to GA for documentation of engineering efforts between Beta and GA +Ive changed them all to match the Test +. +I think this is more of a desired policy that needs to be thought about +Thanks for pointing that out yes it should be public see. +We should focus on parallelizing rename operations so that there is no such bottleneck in +So you can upload fresh photos easilyTherefore there should IMHO be just note generated for mobile devices +In most cases patch testcase immediate commit +Worked well with RHEL bit and fc bit with and without m +Please reschedule if necessary +I read the comments in +It now stops the +Resolving. +Committed to trunk and! Thanks Hari! +Fixed in Rev +Such requirements should also be mentioned in the under transport sender configuration +Is there something wrong in my installation or +Thanks Filipe. +Doug you argued against thisI still dont like version numbers in methods but if semantics are changed in a way that cannot be handled otherwise and theres not an obvious better name I dont see an alternative +Ive confirmed that Marios slightly modified fix addresses the problem I was noticing in my own project +You might want to change such as to including or some such there are potentially other directories as well that shouldnt be touched +Please list what I should do to reproduce this? It works for me for any hibernate configuratoin i got so need some more details +Described the additional samples and removed the old ones +Committed revision +OK cool +A detailed testing is in progress +When HA is enabled and the RM is in Standby mode the services among the stateless services should respond to the client that they are not the Active +Ive just committed this to trunk and branch +Yay! + +Added the correct fix with the whole path +If max is not set then the client tries forever to reconnect and this seems to block the clients shutdown +Very simple test to demonstrate effect of IOD and calculateRequires that be set to andShould be set back to final before release +The removeSession call causes the client to send a dispose message to the broker and removing it from the state tracker used in the Failover Transport so when thats followed by the remove of the consumers and producers the state is not validIve attached a patch you can try out +Im this close to having the live tests passing for vbeta with so once I finish that Ill dive through your stuff +What do you think +My current thinking is to use popupspanels that become visible in the appropriate context +But that doesnt make this a critical bug but a RFE +Fix and testcase for your observations with light weight transactions I can only agree +memory that is used by a task attempt should be freed to as much extent as possible before the next task is run +So for now we cannot use Sonar in combination with findbugs in our there any estimation when this bug will be fixed +doesnt use a separate factory class so maybe just for consistency reasons we can have them in +clientjavasolrjsrcorgapachesolrclientsolrjrequestThat should be it unless Ive missed something +anillocalhostJBPAPPmainsrcbin svn ci m server security manager policy to make use of signed jars Adding bin file data +Ok With Linux no problemBut now testing with our own mediators we have too many open files errorWe do not use files in our developmentsDo you have any idea track the cause of this error +Ill take a look at it +Hopefully not this patch +patch for the patch fixing the issue w Sep codec whereby if one field has payloads but the next does not and you reuse the D across those fields then the enum incorrectly thought a payload exists in the nd fieldAlso I added a safety in so test doesnt take too long when you get codec +I named the target not since the resulting jar file is named +Why should a warning be generated +VDB name pattern must be +That wound be great +Patch fixes the javadoc indent issues not appearing in checkstyle +I was trying to limit the scope of this specification to only the features strictly required to implement SASL but this seems like a fine addition to support potentially dramatically improved performanceIll go ahead and make that change as well as the change regarding extra data with the COMPLETE message and post an updated specIf were cool with it Im going to call this protocol spec complete and rework the Java implementation to implement this new spec +Final patch for review. +Luke I do have to say that starting out with Hamlet was confusing for me +Is this still a problem in or the trunk +Jean hold your horses we should only upgrade this when spring framework supports quartz +Dmitriy the bundled hadoop dependencies are only supposed to be used in local mode +So this environment setting seems a good solution. +Its not that I dont want to its just that Im not artistically inclined +If this is holding up a release then just send a notification and release the code +Would that be acceptable to you? Pardon me if you werent volunteering +seems the failed test is caused by meant the failed test was caused by +Here is an initial version of the patch for review +The reason for synchronizing on the object is to prevent the token getting corrupted while its being inspected +Hope that helps +patch for Yahoo dist security branch +Your workaround worked for me + I have added the comment about that to the guide +Attaching a patch for dealing with the this is produced by git but can be applied on an svn checkout viapatch p i please reviewscheduleapply as you see fit +Theres no evidence that this is anything other than a security manager configuration issue. +Traditionally this has meant merging all segments into a single segment as is done in the default merge policy but individaul merge policies may implement optimize in different waysThis actually almost argues for your solution +ho I didnt see that issue before it might be a duplicate ofI also had the same issues about a freezing eclipse +It unfortunately confuses tools like svn blame making it harder to find the origin of a line of code but its better to get it done sooner than later +I dont have any more insight into this issue at the moment but I am certainly interested in getting this resolved +bq +I just committed this. +We can fix them together +not +Accepting for review +applied set fix closed just updating fix version +I dont think this should require super user privilege +The new patch includes several changes +We could just check for first +Well but the error message is still very confusing +we should actually think about pointing the user to this setting in our wizard +seems to be appropriate to add this test +I agree this seems to be more trouble than its worth +Lets discuss +version in them +thanks for cleaning it up! This looks great +New patch addresses all +Provided we find this significantly improves the write performance do you think it would be reasonable to request a Jena release +svn ciSending Transmitting file dataCommitted revision +Committed revision There are no hardcoded paths left in +Hi Werner Have you had a chance to take a look at this? +You mentioned there were million files +waitForevers not working correctly and its being investigated in another JIRA +Asserting that this undocumented policy should apply to this issue until gets even more discussion isnt a gap easily bridged for me +Glad its all working. +Part of a bulk update all resolutions changed to done please review history to original resolution type +should only be used for RPC request and not for requests received on webhdfs interface +pressed enter too soon I meant multiple archives in a single jvm wont work +Also I was short of time in the last weeks +Delivered +It was discussed but I dont think we ever managed to get to creating the new JIRA for using the Smooks config namespaces to figure out what should be on the process graph +Users should instead use a FOR EACH ROW procedure. +Great to see another recipe implemented as part of the release artifact this will make our users lives easier! +That may help clarify some of the questions raised in other commentsThough I dont understand the suggestions fully I would be wary of implicitly linking RM logic with cgroups +Thanks +QE please be trivial to fire an event when the autocomplete clear method is called +Happy to change this to a Lucene issue if that makes sense thoughAnybody know definitively where this should go +attached source code for idea should be able to use the variant without any issues there +Hi AkiEither solution is fine with meColm. +looks good Thanks Bhallamudi Venkata Siva Kamesh +Hi FlavioThanks +Here is the patch for this issueNow it works fine in case of international orders as well +I think that all the data that is being written in the is not necessary +I test ont thethe web service runtime jboss ws +Please email me your private email to I can then give you temporary credentials for svn and you can suck whole project will then send you details on how to get it running locally +Change variable stack reset to just get a new object +in millis but other elements have the quantity pattern +Youre right if we wrap into wed get rid of an extra step +fixed in added back to last night after Torsten had trouble building the frameworkI added it with releasesenabledfalse so it is the same definition that would be inherited from the Apache pom +Read section Client View of Session Beans Written to the EJB Simplified API +I have seen this issue on a new install of OM +Patch to include you can take care of your own patches now +Attaching full from the failed unit test found in derbyderbyallstoreallstoreunitstoreunit revision . +Thanks again for getting to this +Will try it this weekend. +We can add the reconnect to and wait for Datagram support for push it to +However the patch is good for Xalan buildYash samples to utilize the Derby embedded revised sample code requires the derby jars to operate +I have got the same issue using AS and its embedded libraries when my machine is behind our Squid Internet if I use the same WS client classes outside in a simple java program with same WS implementation libraries it works even behind the same proxySo it sounds like there is something specific to +I just committed this +So well +Sure +Is this bug a duplicate of +Ooops sorry +Excellent! Fixed in +This fix should allow the defect to be resolved +Im rejecting this issue but please reopen it if you see it again and is able to provide steps to reproduce +Patch to introduce common interface layer shared by all elaboration on the above Keith and I have been working on this together as discussed on the mailing list previously +The problem is that we want to keep the order of the elements as well if you for example do the call on a +With commit abedfdcaddbabbddc this has been further research needs to be conducted on pointcut matching +I would simply like the ability to do it because I realise that in some cases separation of domain objects from persistency is vital while in other cases its something that requires extra work and will never be usedForgive my ignorance as Im new to Hibernate but is it the projects philosophy that everything is either black or white that solution X is superior to solution Y under all circumstances? That would be unfortunate for me as Im a big fan of analysing the tradeoffs before making a design decision not the other way around +But the issue is that that exception is swallowed +version of Visual Studio does not have the following project types installed or does not support them +Here is a patch which changes addValues addFields behaviour to always copy collection but keeps setValues setFields behaviour to use whatever it is given +This looks mostly good +Updated to r +At least it will pass the proposed test case + +Please include this class in the componentsprofilersrcjavaorgapachejetspeedprofilerrulesimpl directory +I dont think we need to do checkConnect because we use Sockets under the covers and checkConnect is always called on creation of a new Socket +I checked in the fix +attaching a patch with the unit test +Added a null check +Im having the same issue here but Im configuring my endpoints programaticallyHow can I set the xml to use scope prototype from within my codeThanks +Regressions ran OK +fieldName postalCode notNull sizeMin sizeMax add field string class +It appears that the POM is working as expected for this case +I wanted to say Edited title +Afaik theres none so far +Would you be able to make some proposal for every axis of the technical debt and propose some default values +Wrong patch attached first +It appears that Groovy is internally using the synthetic bits by itself +Uploading patch with above mentioned comments considered +Sorry by this change I mean +I currently implemented the following other transaction exception dont bubble anymore but the method still not transactional +Claus Thanks for the feedback all your suggestions have been incorporated +Try on failure to load seems like a very sane thing to do +If there is some speed up then the patch will be worth it +Anyway I have no luck with it and decided to use IBM native stack instead +If there is a problem please open a new jira and link this one with that. +Added equals code to +There was an apparent problem with a casting in some of the test +Hey Hari given that this error is only encountered with an invalid config I think we can close this JIRA as well as +We have discussed several times the issues behind choosing a datatype or a task on dev list and last time we agreed to undeprecate the configure task and make it the default way to load settings again +Essential in JEE or environemnts +Already resolved with +It is tied to device ready event +Thanks KevinPatch applied to test support at repo revision rPlease check it was applied as you expected +same patch but remove a useless log info +I might have missed the merging of child workflow metadata in the +The wizard prompts the user and informs it that mongo is required for mms! is also resolved +Have a look at d to see how I did it does incur a very slight perf hit for all new calls but its barely noticeable +Im not an expert of this area but it looks goodThere is just one little formatting issue one tab in the first code chunk +A couple of screen shots demonstrating the screenlet widgets ability to include form widget navigation in the screenlet title bar and the ability to collapse screenlets +That way we can update as needed without commits +We are really stuck without this bug being fixed +Version of proof of concept Using to get the map file instead of the classloader +I have to disagree with you on this this is working as designed +Unfortunately I have to revert the changes and provide alternate implementations of those listener. +thanks Devaraj +Ive been unable to discover why the b test fails +There was a problem in the stack offset calculation when method parameters included longdouble fixed +ooh v was missing a bunch of code +Renamed issue to reflect shift of focus +filed jira deb pkg installation fails on ubuntu as the groupid is used by gdm +Hi NitinAlena had a look at this issue yesterday +Changes committedOne problem with this change is that custom Fields that override onProcess will need to add this check otherwise there is the potential issue where two Submit buttons in different forms would both be processed +Just to not scare you we dont need glossy v +Peter what was the reason of reopening this one? +Patch committed +Me too lets leave that be +Now when i am trying to use ant in solr after following all other steps i am getting compilation error +the bookie would still fail to start since the local bookie files are not consistent with metadata formatA consistent instance setup you should run local bookie format after metadata format +Please extand pool size +Patch that implement the new behavior but keep it disabledIt can be enabled by a configuration symbol +Thanks Amareshwari! +I made a few changes as I was reviewing the last patch by Rahul and am attaching a new patch with those +Im having the same problem using version Ulrik what was your solutionThanks thereIm having exactly the same exception when unsing version What was the solution? +bq +This patch has a little changeyou can merge this change to your versionBecause our network of laboratories alse failure so I cant verify this +Does the test always fail +I will check this in on if no objections from anyone else +The logs seem to suggest so +The only thing I took note of is that in my file surefire was resolved to while the other surefire modules were resolved to +Its a little tricky to describe the conditions under which this user error will be encountered +Sorry I think broke the diff +Ive integrated it into the kitchen sink archetype +Your patch seems to be good but Ill accept to apply it only when all project types will have this option +My point was merely we have to be careful since FOP is used in so many different ways by so many different people a typical isnt easy to define. +Close this as a dup N +The attached patch includes this addition with test cases +When we add new features we need to make sure the support is integrated into the command line options wonderful! I will try this out monday but tell me did it already work for you and pass a mvn verify for check style etcThis seems like a for commit any opinions +However If we detects order by only use one reduce sampling is not necessary +Right +That is probably fixable thoughI read the issue +Applied s patch thanks! +no validation errors means that the content was goodnot validated for example because an attacker managed to inject a form fieldvalue string that crashes the validator under a implementationphilosophy should abort the process under way +Im going to look into whether we can avoid the requirement the need for this by having the +Patch to issue a SELECT COUNT query to get the size of an uninitialized collection +thank you for cleaning this up! +Greatly appreciate any help with resolving these installation issuesThank you +perhaps that utility could be changed to keep existing line breaks if they are short enough to fit in the line +If the contact is not synched by the time we are ready to return from the save command you can get a false failure +it is weird and I could not see the staging repository either but after restarting nexus it showsnow you should be able to drop it +Checking limit is only used to decide when to probeadjust the average variable row size the flush is called in a while loop until it shouldFlush returns false ie +This patch fixes most of the findbugs issues but also filters out a few of them where changes looked like they may impact the overall functionality of the code +Ill try it +Im fairly certain this is the true cause of and other gossip took longer than RINGDELAY bugs weve had +Marking Patch Available +This doesnt look like a bug more like a classpath issue with your Util class where have you put it? Please move this thread to the drools user mailing list so other can benefit from your if you get it finally working you might be hit by these guys and s still a classpath put your class into classescommycompany a bug +Here is the svn patch sorry had problems with subclipse and eclipse +processing of handler methods init binder methods and model attribute methods is consistent now +And that is not should be such a condition +And this should solve the purpose hereRegardsMridul would even go one step further for marketing it is only interesting if the message was read or not +Looks like it works perfectlyWill try with EclipseTested +Olegs patch +No but I will push the code in the next mins. +resolved with changes better encapsulation. +Ive noticed that is closed so I will not create a duplicate submission thereEvaluating +Hey Vlasta it works great verified on Arquillian Graphene project +But I am still unsure how to interprete the log +I changed the issue type from bug to new feature because the current behavior is by design it is not a bugIf someone wants to include an option to filter the tasks in the task list then they are welcome to submit a patch. +If nobody has objections I would like to commit it +Feel free to reopen if needed +Cleaned up the javadocs +I just committed this +Supplied correct think this has been implemented now is that the case? +Otherwise I need to add additional code to use if I know the stream but directly if I only want identification +Patch fixing fail and kill task metrics +I get you now +Fixed in the trunk +There might be implementations where it makes no sense to even speak of start and end keysI think we have three options here Keep pure as is and use directly when we need to know about the implementation +It just cleans up some jars +Done +rel is not called using pig API directly not using Grunt +Patch applied +Am marking fixed +We wont fix this for the point release too as its really a new feature request +FYI convention adds to even if only +the patch is generated based on +My first thought was that there are no implications on JDO but after after a while I realized that there are some +Alfred in out tests and demos it does +This is fixed in the branch +which I think it should. +Note that the maven archiver did not support the add property so I just upgraded it to the latest and greatest version + Bug has been marked as a duplicate of this bug +I havent started on this yet +Please review you can do that Apurv if you like and have time +Without the fix applied setting its size doesnt change anything the button stays unchanged +Im not a Maven guru so maybe there is a better way how to enforce right version of artifact nevertheless this is used in Hudson jobs and works for license issues sjsxp is licences under both and +Please help is the file containing source code for update profile +Why +I THINK the latest version of the TCK fixes that issue but havent had the time to try it +Thats a good ideaA simple global hash that maps a table name to a row key formatter and then all operations on the table use that formatter unless explicitly given think this should be a table property and refer to a Java class name rather than doing it in ruby +For convenience the diff of your changes on top of mine is shown in the attached +But Im thinking to keep a copy of this file specially for and let the build file picks up the right one +My fault missed it in the plethora of updates due to the byte array modification earlier in the week +Committed to trunk and commit to the proper on +Yes since Spring is supposed to be complete in terms of support it would be great to address this in time for and ready for testing in snapshot +Discussed on mailing list +I have developed a patch to the camel commands that solves this problem +If only jQuery is in use this cant be replicated +Suffix +Fixed in +Not really something I fancy at the moment +An empty collection should mean Absence of permission +I fully agree with Phil +reopening so I can modify fix target to +Hi Davidthank you very much for your feedback +Ah sorry forgot to git rebase before regenerating the patch some of the recently landed patches introduced changes to the testcase output +No intensions to reopen it just wanted to know that only this property name is corrected or there is some code modified to actually tie this property at JT level +If it is the same code patch is enough +It was actually coming from the collectionsclasses +Asanka AI do not think this patch was yet applied to the SVN? Hence this should not be marked resolved until such time dont see why the FIX transport should accept an XML structure with an arbitrary namespace +This should definitely be fixed! Scheduling for think that is probably a good suggestion +Ive updated the documentation to state that limitation +will build a jar from class files if a packages jar cannot be locatedI see lets do it in your order then +HmmIt will work but Im not sure this is a desirable way to do itLet me explain then I let you decide what to doThe default behaviour is that each time you create a document of the object you will have unique IDMy intention was not to change thatIf you set the documentID you practically turn off that behaviour andfrom now on the object will use a static IDIf this object is used over time for some reason as thedocument source it will always use the same IDI dont know if this may lead to any undesirable and unintended effects toother users of +failures are in streaming +sounds good Amitev could you try it out for your usecase and let us know if you see an improvement +I think should probably resolve cannot reproduce unless someone can provide a reproduction or more information +testcase showing that returning a Set results in aVery interesting +The note in the COPYRIGHT file is fine we can either move that to our top level or just include that text in what we have already +Can you guys debug this and tell me a little moreregards this from Blocker as a workaround is availableregards issue my commit for helpregards patch works for me +I just committed this +Also would suggest you emit at least as DEBUG the decision the write came too to help trace id issues +When a failure happens this Note spoils the table layoutand The page has all the RM related information like version etc +Mike you can flesh out Mock as you like many of the methods dont do anything +Are you sure this is not a case of different server and client versions +Thanks Tim +please share what you have and i may dig in that routeying +The web service parameter plugin should be renamed but it is kept for +Do you feel it imposes a lot on the user to ask them topass in an and an? I think most people have no ideawhat those are and shouldnt need to know +If this happened the real node would loose its original data because of theThe fix was to add reset and close methods to the so that it can be consumed mode than once + to update on. +When creating a Palette object make sure you use a that has been constructed with the constructor specifying both display and id expression +But I am definitely connected to the Internet +Any objection to this approachIm not really sure what the Maven implications would be either but clearly the current approach of building the war ourselves using Alfresco materials is certainly a concern +Include path in exception updated test to check +This seems to have solved things for me +Updated summary to better reflect and documentation updated +Bug has been marked as a duplicate of this bug +Why do you think that? The break means the switch statement will exit and return control to the while loop +Would you prefer that we close this or keep it open waiting for the new patch +Note also that patch and patch contain regression test for +We wont be supporting in this release +Please improve the titledescription and provide a quickstartIt is not clear what is happening with the current info +lpetrovi in JBDS GAdpalmer needed by BRMSTherefore will wait until at this point +Same issue +Thanks Colin +Closing as not a bug +This should be fixed ASAP but backward compatibility also be considered + +Would it be possible to upload a failing unit testYou also mention your db is not in the list of approved dbs +I am looking into it +of the user guide Using the +Song titles are longer than place namesbq +What if the node id is for ex anonymized on each restart? I think such behaviour should be configurableAnyway I dont think this ticket should be a blocker +If there isnt anyone who wants to resolve it I want to resolve this issueWhats your opinion +Revision +what should be the i put that will it resolve my issue? You +The negative thing on this solution is that there are many small changes in the nutch code base +Im not sure that there would be much benefit to putting the timestamp in the filter +Thanks Nicholas! +This is code I ended up using not sure about licensing issues for just including it as is into +Ive added a bit of documentation about testing and using the testhandler +Dn is primary +I committed this +Moved test folder from testdelphi to libdelphitest Fixed the problem generating filename with only extension +Why is this assigned to Mark Little +Can you please check both the code paths in the unit test? Looks fine unit test +This patch has the changes from an initial review +test folderOk Ill open another item this evening +Hi RobertPatch looks good to me +Its only used by xenserver hypervisor host +To run hcat tests for hadoop use the following commandant hadoop the shims for webhcat attach +I just committed pat and flavio. +Here is the diff +This issue was reported on an old CVS drop but the issue still exists in svn revision Using a string would be the best since m itself is a string however there are many placesto change in the code +Fixed on and trunk +the ideal way to do this is by releasing hadoop for EC as and have a yum server on SIs the idea that the nodes install the Hadoop RPM from S on startup so you can have a Hadoop AMI that evolves more slowly than the Hadoop release cycle? This is similar to the patch for this issue but using rather than tarballs +This patch is ready to go in +Done by redirecting to the WS Tester instead of Web on JBT Betab +Unscheduling from but leaving in for +The CAS is only contended if the given handler already would be invoked +Could you open a new Jira and maybe have it for? For failing the dfs put should be fine +I think the problem is in +Applied Ravis patch to branch at rev +The patch looks good +At least it is much better than it used to be if it is too complex to provide more info it cans be enough for a first version +Patch for and that separates virtualhosts configuration mechanisms +Change source tree for accommodating maven package structure +could you document that rejectBehaviour member only applies toThanks Keith +Thx +Will need to add this property to Hudson instead +bulk defer of unresolved bugs to +those units can be used for dates to specify a slop interval to be defined in month days hours etcFile contains the changes on the query grammar and the File also contains changes on the parser classes +Fixed findbugs warnings addressed javadoc changed Token type to accommodate Nicholass feedback +Ive added a release if you want to assign it there or get a patch in today +Still have no explanation as to why it works sometimes in some environments and yet not in other cases +Thanks Avner +Looking at the the structure of the sent SOAP message is correct +has been completed +Please use in all the filesHive history filetmprootOKbarintfrom deserializerbarintfrom deserializer +Patch looks good now applied. +if the url is something like httpmyhostshopmyEngineservices then context path shopmyEngine and service path servicesAre you saying the above is not working or you expect context path to be equal to shopmyEngineservices +Documented +Moving to to later release +the patch looks good +Confirmed bug for version and added it as affected +After reading that I begin to think that you dont need this feature at all and that youre mistaken +Thanks guys for you help +I think I will leave it in place right up until we actually release so we can test all the new bundle version etc +Thx +if its a regression its do you happen to have a test project that displays this behavior so we can get it fixed? Im not certain I understand the issue here +Reason being Hive already uses it and were integrating into Hive so unless theres a strong reason not to we should do things the hive way +This error appeared about the same time that I did the test +committed +HTML DHTML JSF prefixes are added to the tables of attributes. +That would work just as wellAndi +The reason format requires a cluster ID is the following When you add new namenodes to the existing cluster they become the part of the federated cluster only if the same cluster ID is used +Yep +John Should we apply this patch +Sure will do +Committed at subversion revision +Vinod I mean that sure it posts to but its an automated email where if Im on a phone and read it I have to navigate through JIRA Email threads are a bit more pervasive and easier for folks to follow along the convo +I loved the M release structure +weeks and laptops later Im back +The second problem is commonly caused by defining a overlappingthe +Is it always going to be loaded again and again when requestedAny caching involved? If yes how do you manage the resources from urls can be expensive if there are too many callsWhat about consistencySay a core is loaded and the version of a resource is x +I have my working with a proxy setting however my EclipseCVS doesnt support pserver with a proxy +Well need to see if this is myfaces specific and the what is happening that the value is not being submitted +They should all now have the present tense displays +done for the change request or the implementation +CREATEDATE the time of the inception of the MG and UPDATEDATE is the timestamp when this MG was updated the last time +Same as last time only with all references to the acronym WSDM hygienically scrubbed and author attribution in place +Wsdl File used to generate the problem is due to not handling the empty complexType elements properly +High level patch i saw +Bug is still present in current trunk +Thanks Rohini +Alex Im not sure I understand what youre saying by on disconnect you mean when client is done with the server and closes the socket? How do you know that? And what if the client was scanning but then never comes back +Attached patch that fixed the issue above +Not to mention because the SQL is in an XML file all that identation is preserved too +And please fix your dependencies + +it would remember the last value enteredWould that be betterFYI you can do this now with all fields by doing Create X Edit Save Back Edit Save +You dont actually have to have such a web server but you might be able to use that as a model +Autoconf installed on libtool expat are now installed +Have not been able to reproduce problem +changed +Unfortunately this causes DELI to throw an exception so no profile information is returned to cocoon +Thanks Trevor! +has been released closing issues. +Thanks +Ive checked the command above in terminal and it works +Patch attached +Looking back changing a runtime or servers ID is probably a bad idea and may need to be revisited +JacopoJust curious what is the relation between a missing file an not being thread safe +To clarify my question Click deploys resources as follows deploy all resources on the get +A NPE will then occurThe attached patch fixes the NPE +attaching theIt is OK +Adding a verification to check if the class is like +This patch is dependent on hence the failure to apply +Session A is on node and node +Previous upload was incorrect +Closing resolved issues. +Does this mean that XCP is not supported right now for S image store? Or is the version of XCP on devcloud too outdated now and missing this stuffTom +This patch doesnt look quite right at the least the removal of the new call cant be right +yes its a known issue +Installing in PREFIXlibcouchdberlanglib would make difficult to detect plugins or are you suggesting PREFIXlibcouchdberlanglibpluginsWhich could be better so we make sure plugins are related to one couchbd version +fixed in revision Nathan please verify it fully resolves the good thank you + for not exposing apis +Tried reproducing using steps above +Sorry this bug slipped through +Duplicate to +And this event is right before kill event +Committed rev +Removed from page Listeners rely on the Courier interface to pick up and deliver Messages +Yes +Bulk close of old resolved issues. +Id like to second this +But this is not how the socket server behaves which works in the favor of this feature +Remove all traces of orgapachemavenit from your local repo +I needed to apply all of these changes to get fb to build and install cleanly and to get scribe to build +I just committed this. +I have resolved this problem +They only change versions of bundles that actually have changed +As a convention Nuvem Maven build only identify classes ending with as test classes hence try to rename your test class +moving back to m going to try to start this one first thing after M +Sure reassigned +With respect to versions is our new release +Let me know if I should not have done that for any reasonI have some few changes on x to submit yet but I was wondering is it necessary to deprecate a class in x if its ONLY going to be removed in? Not sure if I understand how these things work yet +If the javascript method has same body it will only be rendered once per request +Button text is already configurable and the buttons are fully styleableI was thinking of the buttons being replaced with something arbitrarily different or possibly as the of another component +Whats a quickstart? If you can send me a link to instructions I will try and carve out some time. +Verified that it is ok now. +Thanks Mike! +I wonder if storing Web content into isnt a canonical use case +Is this still a problem with the recent changes to the way are handled +appears to be duplicate bugs while others have saw this similar bug was reported earlier as fixed but still able to it as of today fixed in redback the checkbox is always there for the System Administrator role but even if the user deselect it the role wont be unassigned +Is it still an issue? I think does set the Accept header now. +binary and source much appreciated work +The uu version isnt a problem in the current codebase that was added as a part of creating aThe octal escape isnt handled though and needs to be added as a new UnescaperEscaper pair in text translate +With your current approach this use case is impossible to achieve +So we have to use original libs from sources +I committed the part of Lilys patch that changes the backup user to the database owner to work around +Koen can you verify the test and make sure it is on both the SOA branch and HEAD +Hello Would you please try my patch? To apply the patch please +Seems to be fixed I had no issue when following the steps above +yeschecking openssl usability +The way is now to stop using replaced by was this closed? Its not done yetIll take it from here if you dont mind + + pending ve just committed this to trunk and +By default this is uncheckedIf unchecked when the user on a node it opens the ESB editor with that particular component selected +I have a change checked in for this that appears to be working +Trailing spaces ant are only difference from +Numbering is fine now verified in +Hadrian I am adding additional unit test for this in and the patch is buggy +I had assumed the limit was much shorter +I wrote this proposal because I had an absolutely real use case in our application where I had to send special HTTP header from several controllers +I will switch to as I have run into other problems using Hibernate +now gives a couple of warnings about overriding CXXFLAGS +But I am ok to change that if others think it is best to call top level dir piggybank +The main advantage is that this way we do not change the distribution +Closing old resolved issues + +The patch with the new test cases +However creating a link to that page is automatically resolving the blank issueYou can easily mimic the effect by just pasting the in a plain test wiki page +push out to +Id rather see hidden out as it seems like itd have more visual weight appearing only for the view stuff so that people arent wondering how to enable itThe patch seems to work out +Currently both and mapreduce creates raw comparators using comparator class names +Hi ScottDone as suggested at r +Actually may not be a problem because UNICAST should stop retransmissions once P is dead +I confirm Darrens fix works fineThis issue is not yet fixed in hibernate so we have to patch it againIt becomes unpleasant to patch every new versionSo please could you fix this bugThanks +Note that if I remove the data directory there is no more problem as the bundle is not any more in the cache thus no more present +feels the bestone option would be to escape them in the case of a float +note one additional benefit of initialize is that we should be able to wrap up myid file creation as part of this another stumbling block for inexperienced users + +Our experience as a community shows that this is usually not the case and the old behavior is confusing so why not remove this artificial limitation altogether and if users do want to keep each cycle in a separate directory they can do this by specifying different output directories +clustering is on the edge of experimental that this point anyway +Patch that cleans all warnings of save for usage of and and one instance of s non usual constructor cause a job reference is still required there +Thanks +I just committed this +Attached patch for documentation +Currently webhdfs does both spnego and token auth for all operations +Reviewed the changes and they look goodHowever I would get rid of MAPWARNING and only use MAPERROR +I agree that something needs to be added if only to facilitate debugging of the various problems Ive been seeing +Sorry looks like trunk changed again +Thanks +Please verify + +Im not wedded to a search component +This is the fix that work for me I have patched sources of found underAlso mvn test is ok no test I have addedViewing the source I asked me why this part is not implmented using Ibatis +The only trouble is the use of Java specific method which is a very minor problem +The thank you page will now be displayed to any link backed up by Downloader +Ofcourse the latency is affected by the number of regions of the table and the total of regions being used on the server +Additionally adds different types of Crossover and Selection Policies +Here is a patch that should address the issue +The first case removes an element from the objects list the second tries to add the same element to two different objects lists +Will commit in full as soon as RC is out +Its about giving a fair chance for the job to get executed in the order in which it arrived in the queue. +A workaround for you would be to set +Patch committed + +I bow to your l coding skillz and shall never question them again +They currently live on sigma +Thanks Ill review it +Merged for in revision . +Unpack +Otherwise well asking the Erlang VM to store the entirety of the view query in memory and thats not feasibleFootrest looks like its coming along now that it has the ability to use another database as a scratch padI also dont think this blocks the collations +Thanks for looking out +Fix applied to the git branch rev dedaffaddc +Help in resolution is appreciated +Many times the users use hive and hadoop interchangeably to accessthe data and this might be difficult for them +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +GB is already the default in the code +Can you double check that this should also apply to trunk +Tested with jdk works with WCTME and just from the set of derby jars. +This will likely need additional config parameters to configure the list of aux services being used along with their service + +It may not work because of thge limitation of but please raise a new Jira improvement ticket for nested annotations and we will look at it +The problem here was that some global variables were not beforethe bytecode was attempted output + +Closing resolved issues for already released versions +Just as a sanity check before we go too far down this rabbit hole are you sure you need to build the code rather than just using the packages? Are you trying to modify Accumulo or just use it in your application? Sometimes the README can be misleading and imply that you need to build the code when you can really just download the distribution +Im sorry +Looks great +OWB has different code path for specialized producer bean +Andrzej do you think it will be possible to commit soon +Let me clearify myself when I asked about Update Source Folders I know that it is suppose to include the generated sources into the classpath +The above stack trace is trapped from debugging into and trapping the exception thrown since Ant does not print this exception +Does anyone else have experience in or thoughts on defining precision and recall in this context? is arbitrary just pick the top n or top n Id imagine +This is what it should be likeUsing these catalog descriptors svnasfforrestmainwebappresourcesschema svnasfforrestbuildplugins catalogUsers +Opened a JIRA for executing server actions on a separate thread +And that is a little more I simply leveraged the solution that I had available to me +I just committed this patch to trunk +The stylesheet ran fine after that +the problem was in my code the hierarchy parent child references was not set +Ive made a change in SVN trunk to cache the local host nameCould you try again to see if the situation has improved +this code looks brittle +The failure looks unrelated +They must be downloaded and installed before the WTP can be installed +For example Ive set this variable with following valueMAVENOPTSXmsm Xmxm XXmSee java X for more details +Please try this when you get a chance and let me know if it is working for you +The specification has been updated. +mvn P systemvmKelven +Other URI options are documented +Please update the patch to add a test verifying the fix +do you have a suggested workaround +I second that request i would like to have such a feature as well +Meanwhile users can set their umask appropriatelyMy second preference would be only to only change permissions by default with java and higherI think we really need to have tests for the default configuration and tests might have to be customized to different environments which seems like a lot of effort for something that wont be used moving forwardI am sorry I didnt chime in earlier +i will be busy this weekend see if we can squeeze it into can do this but the problem is that the patch is not a valid patch that eclipse acceptsand the Imagebutton is already from sep +I prefer the +Thanks liang xie +Produces a readable representation of columnsupercolumn diff clearly shows being replaced by patch I just submitted only works when applied after previous patch +Second it looks like the patch doesnt apply cleanly to TRUNK anymore +definitely better than + +Tested on ubuntu and fedoraGeorge +Otherwise as observed the data will be inconsistent and broke the replication +Hi GreameI got a chance to debug the problem +for these ones may be recoverLease will return true at the first callI think so +Could the file be placed somewhere that the jobtracker does cleanup +I can see the benefit of this in many ways and the nice part about this is that you dont break backwards compatibility here +Looking more closely at the code and the message Im really not sure that is the answer +Thanks for the patch patch is to fix a remaining issue that should have gone as part of original checkin +Reset assignee so mails go to list +Updated patch after merging changes +Thanks +bumping from readd it if makes it in. +We never created one +I think it is ready +My mistake it is not on by default +For s of source files that one might edit in a day this is a tremendous drag on productivity +Tried extracting and running this test +Anyone want to give it a try +Currently the build fails +Tha patch has been successfully applied. +Since the DB is it is not permitted to create a container for this Hashtable +Dan It appears that Jeff fixed this +I had to implement this in a slightly different manner as we have to allow for loading of descriptors via different classloaders +Ripples all the ways down +bq +The updated war with the upstream groovy jar doesnt fix the issue +Updated patch to apply to current version of public vectorization to branch +The last patch on RB is weeks old +The is not a very friendly construct in the debugger +branch revision trunk revision on. +It works now but icon has and label has +Eclipse is not reporting which file has the error but is intested reporting this against the entire project +Please bear with me a little longerOleg +svn that I used for my other global changes +I know in general the community has avoided such properties but perhaps it is justified in this case since there seems to be no other alternative +running patch through hudson + +See should we keep this for or move out to +If you have any questions or concerns comment on this so we can keep track of care All I have attached +Even if I close all browsers and freshly open a session this index page is not loaded +I confirm this +This could be a general Condition is always true warning +Times out after s now +what is your expected behavior? I think if transaction are trying to accomplish the same thing at the same time one is supposed to fail? We are not taking out write locks but the entire update should be in one transaction +Please test this to make sure no obvious problems were introduced +It parses so you only need to enter the least significant portions +Actually the issue was not that some methods were just not marked for the patch +If you are targeting this to I advocate we invest time in developing generic importer then make use of it for data sources like so that we do not have to do the metadata import once for the Designer VDB and once for Dynamic VDB +Spencer et +Thanks Scott your patch is in rev +Moreover we should not retrieve identifiers from SCM now that we use the email address instead of the name in Git. +If they say no we dont add it +This seems to be fixed in and versions already +on commit +I remember talking to the Twitter guys in the context of about connections that appeared to be alive but made no progress failure detection is also prone to false positives in the sense that a transient network failure shouldnt a streaming operation +war applies to the actual content not stuff that gets generated +Hi cforceDid you resolve the problem? Can I close this issueThanks i dont understand what i shall find there? Can u explain? Tx a lot +has been committed to trunk. +Will you should be able to verify the lucene version with this linkhttplocalhostsolradminIt should be different from thisLucene Specification Version Lucene Implementation Version buschmi +Committed this during the JIRA outage +Added a simple lock timeout mechanism and updated the related TCK tests in revision . +Because of that the heartbeat always fails with the NPE trying to read something that was never set +Slava are you sure AS supports Servlets +I believe the null name is only done in some of the unit tests that arent creating delegation tokens +I have checked all of them +Thanks Jukka it looks good to meIf I understand thing correctly we need to override the get method in too to allow the reference key to be shared in this scenario when using it +both patches are peer reviewedDeciding what to was based on issue severity and we only critical stuff for which there were no +I am also affected by this annoying it be possible to deliver a patch release for the release? +bq +this version should probably be released in the maven repo to keep our builds job working on that! +This isnt a bug as explained in the forum thread so closing this JIRA. +No +Confirmed no additional regressions +I realize that doesnt appear to be the topic du jour but +I fear that I wont be able to make any relevant changes without your help as I find the EMF part hard to get a thorough understanding ofFirst Ill try to update the qad to see I can trigger the issue you mention +The profiling is currently woven through the code using several anonymous classes each time so it isnt possible to just turn off from a +Marking closed. +Sorry for the delay + if a parent version changes rebuild think the priority of this issue should be higher +It was introduced by +Thank you Chris for the review and testing +Let me resolve this issue +Please refer patch in +Should revisit this issue once we do API audit after. +Probably classpath problem in POM +This did work for converting a full listmap but not for individual listmap elements + +Yes this is what I expected thanks +There were several issues for DocLitBare services The webservices for doclitbare testing were not set up correctly in the service control +Good point +Patch checked in +Sounds good +Will upload patch on Monday +With EAP going GA and the bits being made public wouldnt it make sense to update the build from the EAP ALPHA bits to the GA bits and therefore the patch would no longer be needed +No problem on my side +Sachin is testing for confirmationTake care. +There are a few places where it does not meet the guidelines +But only has Java and Java at the moment so its not for sure +Starting with the text in the right upper corner is rendered too +This patch fixes that +I am not familiar with code +fixed in +Trying Hudson failures are related to +Therefore you may test anywhere that richinsert highlight is being used in +Still doing some more integration testing adding unit tests where possible and adding some javadocs +Updated according new any test case would be helpful for the verification +Attaching patch your patch in revision with additional unit tests +fix in svn revision should be marked as closed +please it +Im going to assign this to you so it doesnt get flushed tomorrow +This patch changes the to and +This project seems moved to project and it is very different by the original +To change partitioner you would have to edit right? There was no such change before this restart +Closing issues resolved in released versions +Should it also generate code for array elements? should also reject creation of array whose element type doesnt have default value without initializer +As Jerry said defining a JVM parameter will turn off the use of time for remote service requests +This appears to be a dup of you tested this using the latest release +Its possible that was executed twice by CI and so one of the runs had a blacktie issue +Frankly speaking its also strange for me why Object Helper works may be due additional checks which it contains +Im seeing that the fix is targeted for but still open +It is noteworthy that advertising works with EAP ER on RHEL +Adriaan could you elaborateWhich html elements are missing + +Ive also added a target version of which corresponds to trunk +See comments should be configured to use runtimes +Thanks +Tim thank you I cant reproduce the bug with latest sources +bq +Guys I fully understand how valuable it would be to cover this use case but everything done on javascript side would be a dirty and quick workaround while waiting for +Let me merge it and go ahead +Now typerendering seems to be hanging Im getting rdf but when requesting xhtml I get no response +Deferred to +of the warnings are from rumen which never got built beforeAlso the need to be udpated too as common and hdfs have more then its actually generating +Client side Stub can define on bean definition and can acquire configuration context bean. +Merged into the master branch after reviewing the changes and verifying that all unit and integration tests pass. +Ravi what is the issue you are experiencing? Please give more details +What kind of queries would this break? Just to be clear when I say weight +Verified in to BRMS CP release notes as resolved About dialog box was displaying an error message on some systems instead of the expected content +Because a symlink to a directory is no longer reported as a directory believes it doesnt need to recurse into the directory and the job failsI tracked this behavior change to +Wish there was a way to include the check but without an integration test I dont think we can. +Next patch toward the server +Ill check file settings if I any are missing svn native Ill post to +I understand and the fix is fine my problem was simply that one should also add a short description of the issue that lead to the failure thats all If Steven would have put that into the issue description I would not have complained +in +xsd schemas are installed only on the Jetty assembly and the +This is the wsdl which generates the interface +With no s registered there is no blocker +I seeSince we are just starting to use hadoop archives I didnt care to make my patch backwards compatible +PR adding the stuff to the sitescaffold not included yet since not part of pom modules list so I assume its still not ready for inclusion +Should we set this and the related issues set to resolved +It can be removed from the file and from the libjars directory +In the panel the string is displayed +Resolving it as Fixed. +Thanks! + +This is a waste of time +oops it is not there am out of my mind it is already in beta +committed +Thank you Jing! +Good thinking +doc reader could peek into the file and return either textplain or application +heres the actual patch I plan to as fixed for now +Keith and I have been working on this together as discussed on the mailing list previously +If no longer has the problem why file this bug report? Is this at all Struts related +let me think about this and review the code +Im pretty sure that IVY will have an equivalent of SNAPSHOT in Maven we should probably use whatever that is instead of timestamps and version jar artefacts correctly. +I dont think the patch is correct though the oneway call should just throw the exception and not mess with sending responses or exceptions to the delegates +Any reason for this duplicationWhat role does the different Store objects in the metastore java dir haveThis patch does not make sure that a partition drop doesnt delete the original data +fix for multiple callbacks removes recursion changes approach to breadth not correctly implemented incorporate Brams fixes +bq +Moving to the bucket for consideration +Scott Kryo was using byte code instrumentation and so on to avoid reflection +I cant look at this today but Monday I want to look at my original use case to bypass the admin admin login step and use the sign in as the authorisation for Spring me this should be a configuration option of the Roo Social +All tests still passing against does slfj add any performance benefit to cassandra or improve the logging in a dramatic way? Im just trying to understand the motivation for such a change +Reformat for readability +I would delete it and it but I dont see the option to delete attachments +the usual way to deal with this is by calling clean even if I agree it should be handled nicely by Maven +Thanks again Jevica +updated of course mv should be usedrebased as that the file structure is all over the place thats why this patch was so incredibly painful to doseems i only did the disk space checking in sub directories in the backwards compatible patchi had on algorithm for detecting what kind of file it is in the old backwards compatible patch it iterates over all data directories and figures out which data directory the file is in then it knows the keyspace is the next part of the filename and can check if there are files or directories in that directorybut i agree it is incredibly ugly your modifications make it alot better and yes this would be a good time to do this properly not like me just trying to do a minimal patch that worksregarding keyspaces in file names sure why not guess having a header with this info in the file is out of the question then the only meta data we have is the file name right? A problem could be if we want to do later that would increase the file name length even more +IT added +I will add it +Sunitha can you check the results in your environment and close the issue if things look all right to you? Thanks +I sure can +I made an enhancement on trunk to not keep a consumer active if it has not dispatched messages that need acknowledged but otherwise you need to call commit +Ive fixed a bunch of the validation issues too but didnt get through all of them +Should DMC read +Its fairly high priority +Done in now generates an empty resources which will be picked up by +Thanks Adrian! +Just deploy it and you will see the WARN statements in the log +Im not sure exactly when or what caused it to start failing +File this is the class that the javaWsdl tool is using +I spent a little time getting this patch somewhat updated to the trunk and running various benchmarks with reopen +OK perfect thank you very much bu upgrading JCR +Ok let approach it from a different angle +Btw +PDO doesnt seem to allow variable number of columns per rowbq +The application starts now but there are some major problems with the images referenced into the Web pages +have values such as true in their enumerations as java will be choking on key words +Looks like if you uncomment that line it fixes it! Ill experiment with the workaround until is out +Thanks Bryan! +The use of UNDEFINED would also affect the selection of the default or customUNDEFINED Beans and collection Items might be translated to Proxies that execute a lookup based on some Path Conversion scheme +when patches are ready you need to submit them to make them patch available +Im dont think the current behavior is wrong +Removed methods in in lieu of checks already made in +removes the interface change to +Nothing to do with +Moving out of + problem is what portal itself activates transaction for request +This made the warnings go away +modified srcwsdlorgapacheaxiswsdlwsdlwscppto fix this srcwsdlorgapacheaxiswsdlwsdlwscppliteralto fix this problemPlease notice the mark like mxiong improve to avoid compiling warning +Borrowed code from POIs to extract Shapes instead of Runs +Done will be in EWS +The transaction is never committed +Thanks Sanjay! + +I verified that it fixes the test +Raves overlay file wasnt updated with the original change to Shindig and there are a handful of property changes between and +in this case a coupld of other features need to be implemented the code from that does the project aggregation should be moved into the project builder as an optional function dependency resolution that is triggered by a goal on the main project would also need to resolve dependencies of the collected projects on an aggregation enabled too much risk of destabilising ahead of plugin becomes an aggregator plugin when refactored to handle reactored builds. +as per which doc please +Please ignore the previous commentThis patch makes it possible to build the docs target using JDK Disabled sitemap and skin stylesheet validation in the files +Solution is fine with meAn explination behind the purpose of maxlo as a multiplier would be really nice +Like I said youre trying to copy your whole drive +The popup was correctly moved upwards in firefox but dropped down in IE +In your case if you dont provide an Initialvalue annotationthe parameter is null when it is put back into the pool +Those two parts of the application are different +is best Ive run so far +mrepositoryorgspringframeworkHOME +What is the status of thisMoving to for now +Yes I thought the problem would be solved by enabling File caching in the +To make sure I fully understand the request you would like to change from the current to right +I dont really see how we could be part of the problemThere is no batik code anywhere in the stack traceMy first suspicion would be that you somehow corrupted theclasspath when adding batikSorry I cant be more help +Multiple other images and icons also appear +This file is from the testsfailtestfolderBefore running the test I had cleaned the tests folderI used the following commands to run the testsset CLASSPATH +Integration tests were updated. +It looks like he is taking a good approach by distributing the collapsed Gibbs sampling and then by optimizing the sampling of the Markov chain which easily provides a x scale up improvement +Thanks JohnI put this into trunk and for finding this Nicholas +This should make the job of porting to different databases much much easier if it worksCode as it currently stands is in brancheswarthog +Committed revision in HEAD +Let me know if I can help in any way +Net Sp? It looks like this issue only occurs on versions of prior to thats truly the case should we mark this issue as bug and require users to use SP or should we try to fix it so it works correctly on both +One minor comment +autocomponents only exist during the render phase so they cannot receive callbacks. +We have made exceptions to that upon request +Resolved as per +As I write this eclipse is still generating the details +The database backed vector should be configurable enough such that fields could be selected +What Martha mentioned applies to as well +Implemented with commit revision! Glad you liked itThanks +Patch applied at r thanks for your contribution please verify +In Caolites example lets say that of the records are skipped at what point does that go from being acceptable to being a bad input file? In most environments Ive ever supported having a bad input file would signal a particular error code which would cause a certain page to go out and someone would manually look at the file which would usually start some type of process with the personentity that sent the file to have them resend since its obviously bad +It need to be developed tested used +There has been a flurry of fixes for spins in joni some fixes to joni and some fixes to +Kathey here are some answers +Issue is the overall modularization effort +It looked odd to me when I checked the details of a bundle +I must have been really confused when I opened this issue +actually has a method to deal with this set +Suresh Given the security wont work without it why log an ERROR thats a step more longer to catch? My preference is to keep it automatic as stated on the description as well +Are there any performance metrics available +I understand that the fix would be available as from or in SVN +Fixed documentation problem and fixed problem that might stop Logger from working if interceptfalse +seem to deserve a spot in command line args +Patch applied r +You could try the there a release date yetI configured our nexus to use but I cannot find +batch transition resolvedfixed bugs to closedfixed +Good job Charith +I know that Id like to be able to edit my comments +Granted if you assume that no one would do this stuff then it is okay but if you worry about these sort of things then it is not so great +Heres the rough instructions +we cannot tweak this for because it would change existing behavior and would require code modifications from the users +If not we can bump to +When will this mirror be checkedadded +User sees no problems with a single route they have routes with hundreds of endpoints deployed within +committed the patch in CXF trunk and CXF branch +An unmentioned change is that theyve also made GG an allowed duplication +So this issue can be good to hear that it works for you nowJuergen +These issues have all been resolved and just need to be closed. +As is using the skip parameter it is by design +Where the user does not have permission to access broker +v version of patch which works with the latest trunk mapred mapred with and without Dresolversinternal + +After we should be able to replace this rule by our own and without this limitation +My guess is that floating point intrinsics do rounding in different ways +Ive opened to have this feature configurable on a basisThanks Thomas for starting together these two issues should lead to an ideal solution +It seems that the patch is OKI just commit this. +Bob Is that just a matter of changing the data structure that is at core of Context +Here is the proposed javadoc +Yes I agree +Tagging if can be slotted in if not bump to to issue not Cordova issue. +There are some differences in these test compared to the specification +Tests will be included in next patch +A new patch has been attached with the message arriving before the receive operation problem has been addressed +There has been no activity on this issue for a year. +Thanks Evgeny the patch was applied to SOUND module at rPlease check that the patch was applied as you expected +The suggested solution to make the Bundles class Session instead of Application scoped while a good thought it is not recommended to store large amounts of data within the SessionThe ideal solution is to retain the Application scoped nature of Bundles while also allowing you to retrieve a bundle based on the current users locale +RayWould you mind spending some time updating the user guide with a small section that includes some instructions on how and what you needed to do to get s working happily? It would be really cool if it highlighted some of the gotchas and workaroundsThanksJon +This release contains the fix for +changing it to an invalid dir does raise an error so this is going wrong only when valid paths are supplied and after the early validation +On event I would expect to be calledobserver of on an instance of of on an instance of of on an instance of of on an instance of of is NOT called on an instance of because it is specializedOn event I would expect to be calledobserver of on an instance of of on an instance of of is NOT called because and override it and no instance of exists because specializes event I would expect to be calledobserver of of of is NOT called because is specializedOn event I would expect to be calledobserver of of of is NOT called because is specializedwdyt +Thankfully there is always the HTML versions to fall back on when the issue is like this one +Thanks Arun +Overwriting initial rat report following the most effective way to debug transitive dependencies has been mvn site and look at the that is generatedI built the pomtools plugin from the latest source but it is broken and there doesnt seem to be much development on it so best avoided unless it is developmed futher +I ran the HA tests with this patch applied and they all passed +This way we could easily manage some future possible issues that could rise up if we tend to use some other features of SAPDB SQL engine while providing support for new features specifically in OQLSQL translator the other hand with current set of supported features solution works fine +committed to branch and trunk. +this not an inconsistancy +It is also now possible to specify a wins server in the bean has been updated with and had Abys changes from above rolled in +Will attach the output +I committed the final patched version at revision +No this is a bug +I am trying to backport fix for but this is blocking my testing of it there +this makes sense Id set the default to larger than . +Also in the widget store Add a new widget After adding now again try to add the same widget +At a minimum please upmerge the patch +may be true but is not obvious for a simple user of the framework +Unsetting Fix Version for unassigned issues +Apache Directory Studio version has been released. +Erbi thanks for catching this +I didnt use it before. +Happens for as well +Did anyone verify or not if the patch is the right way to fix this? It seems to fix the path issues Ive been having but Im wondering if Im trading one set of problems for another +Summary of changes +So there is nothing to do +Field naming convention doesnt seem to belong in there +Patch applied +This is looking good +Is there any chance that you inadvertently deleted or modified the version field inSee delete silently ignored if updateLog is enabled but version field does not exist in after release. +Hibernate is instantiating persistent objects just for temporary configuration reasons there not for active use in the application it checks the default value returned by the objects id accessor in order to determine the unsaved value then discards the object +Resolving +Ah thanks for the note about +The map is a new piece of information which is ultimately communicated to Reduce tasksUnit tests pass +It will require someone to work on or sponsor the changes to replace the socketpair usage with something else or make the code work without that facility +I didnt see that Lohit had done a patch for this until after I had finished mine +Committed revision +Disable context menu action if weldscan element already exists +esb archive +pending the have committed this to trunk! Thank you for your contribution Mohammad! +patch is still not available for trunk will work with the author so I can generate the patch for trunk +If devs are ok with this i can provide a patch WDYT +For testing please put the gadget definition xml and the template library xml file in same directory +Sorry about that +Shouldnt have resolved just set the Fix version to AR +groovyc disableASTT +The unit test was committed at the revision r +Verified by Miguel. +Im guessing that in that case the split items need to be trimmed and need to be checked for ? right? +According to the JPA spec the falls back to style execution then which unfortunately doesnt work with traversing result lists please where your transaction boundaries are and that youre initializing your result list within such boundaries only +I understand the portability issue but if you use a JDK with JSSE downloaded from sun my patch will work since it provides the packagesMoreover if you look at the code you will se it is already using the framework +I have attached screen shot for more applied and testedimplemented by Vivek Kumar +It depends on the that is used in this case it could be the +Thanks for the feedback! How does the Hedwig connection factory type get registered in JNDI? this is specified in file in the classpath that specifies this parameter Does the protobuf format need to be in or could it be moved under? I dont see how +s and Gregory S +Fix for test sock data loader unit test +The only extra work that is involved in making this work is batching +Committed to trunk. +its removing the keyspace arg and using setkeyspace for the keyspace state on the clientState +Third patch even better +Didnt realize that the logj filename has been changed to logjclosing the issue +Yes this is actually what I was suggesting as well +Conversely if the inbound message rate is greater than the trigger rate then the channel queue will grow with messages waiting for a poller to pick them up +Code looks good +Because of of after inline and following type propogation we can get more detailed info about copying types +Dejan Thanks for your commentsI agree a production system would probably not kill a broker every seconds +Removed that AnalyzerThe last example is one problem of the explicit UOE Now a user can suddenly get UOE if he uses a query where +I kind of see this problem from both sides +to know +Just keep in mind that Im still profiling this and will post my findings as I go along +Attributes delay and showHistory will be removed in. +New patch attached and posted to RB +get ready for the years anniversary of this bug! Ok seriously what you guys need to get it fixed +I had this feeling as well but I thought it might be too many extra parameters just for the regular expression support +If we delete dir deep in the tree they are just noramal so flow goes in diff path where we are not wiping out on some conditions +Is there anyway to formalize this tracking process to help us or shall we just use a JIRA or two for each releaseThe changes I see +just a thought +Verified in dist built from SVN Thanks for taking the localhost check out of too makes it easier to test from my remote machine +Apologies I just got back from vacation +disabled the type validation as Gary suggested. +FWIW I dont think you sent it in the original zip but its there now +In addition to fetching https websites Ive also run into this issue when trying to send SSL email +Very interesting +This patch updates the for newer Jetty verison + +Daryn Good point +Lets upgrade to the latest ve uploaded the tomcat bundle and the to use Tomcat +I just realized the javadoc doesnt need update its described on the class level +this will become more clear once i actually get a bit more time to put together a full proposal but this is a standard feature of the other major blogging systems out there so its not something we are inventing +Oh! You meant zookeeper server will first process the request and after that will process the request +Storing the indexes in the DB reduces IT complexity for backups database clones transactions and unit tests and although performance may be slightly impacted it seems worth the price in simplifying a bunch of other things +Thanks Yuri the patch was applied to SECURITY module at r +I was more concern in replacing String with But if is even more efficient lets replace both String and with itExcept in case when there is no need no string accumulations of course +Actually I was thinking again about this one +These fixes will now make rc +Now may i ask why not using SLFJ? Might be a good modification since its what hadoop uses +Upgraders from releases should recompile their applications to ensure they will not run into this break at runtime. +I have attached the axis log for this +It is great to have a framework for converting the recovery tests. +Yes whatever the jive guys did resolve the issue +I have sent an email to the cite forum asking for guidance +Both +RuwanThe model does help alleviate the issue but it does not solve it completely +Again Id like to that this feature causes serious performance issues on the JNI calls are very expensive a rouge client for this feature can easily cause severe harm to the and hence the entire clusterHive can use JVM Metrics for the same functionality given that this is already available in JVM Metrics +I will take care +Is this problem still happens? What about testing the server with a newer JVM version? Sun JDK is nowAny clue +verified. + Test Case is updated against modified CA Proposal insertion following Tests are added for the issue +Please also backport to SMXThanks +Committed derby to trunk with revision +When a task attempt is killed failed and the overall job fails still the coutner info is shown When a task attempt is killed failed and the overall job passes still the coutner info is shown +Need to check all columnsTo me it looks like the sorting is done but in blocks as if the UI reads a part of data and sorts it then reads another part and sort is independently and so on +Hopefully the guys that have been pushing the UI forward will lend a hand for further work in this area +Applied patch to branch this bug still happen on FF and IE +And I borrowed it from Apache Directory Study to provide the functionality in fulcrum +Integration test added in revision +srctestorgapachetilesconfig needs to besrctestorgapachetilesconfig +Please allow me to see if I can reproduce it +Im gonna to commit my changes +Looking +Running locally now +Once applied this fix a is received instead of aMore information in user mailing list thread Unexpected withAffected versions and +Though this is a good bug to fix to be sure I dont think this should be considered a blocker and hence Im lowering to critical +I wish that this quirk was better communicated in our processes since I only found it in JIRA after searching +Hello Gabriel can you please review the patch so that the issue can be closed asap thx! +Username alinasz +I need to do some final touches but I dont expect any new failures +It should be true by default +The problem with the current implementation is that the is collecting the lastModified if the etag is not set. +Thanks Navis +Cool! Thanks a lot for taking over this now applies to revision and from my perspective ready to commit! Looks fine to meThanks +Duplicate of which has been resolved +You need to checkout from the parent folder if you want to build without problems or install parent dont want to add the snapshot repo to all poms +Tidying up +True on close examination the help message makes that clear +Ah I see +add the following remove allocation of a in by defaultThe patch is estimated to give around on a large on bit JVM +If there is only one row left would simply clear its text but leave itIs there any way for me to capture a query after I have gone though all this trouble to create it? I mean why not store the parameters in the display URL so I can copy and paste it +I seems to me that you often needs to add the shard number in the value of either the indexBase or indexNameThe redundancy seemed very bad when I wrote the example +Now passes tests had a cast exception in renderer +Though it looks like runningofbiz ? Sl java XXm sodukasddn jar Anyway now I will reload manually and update the fileThanks +Can you attach the docs as a file attachment to this JIRA? Then we can accept that and include in the docs +if it doesnt work all the time i would rather not build it and just educate users that if they do weird things like that they have to call attachdetach themselvesigor +Looks good to me +But I would try your suggestion when I will find time for this +I will mark this issue as duplicate +Changed DFSSUPPORTAPPENDDEFAULT and used it in NN and DN +Now all thats needed is a fix Good to know what the problem is +If youve built the jar files yourself you can try this by running ant clobber and ant insane before ant buildsource buildjars +Just a note a number of our developers are seeing this on both Linux and Windows +By the way i used on the tests +I will review your patch and apply +Upgrading your JDK level is a good idea +PR for master the issue as the PR has not been merged. +It works fine for me +I forgot to mention Patch not yet tested as I first need to boot my Snow Leopard and install SVN there +my mistake the app was configured to use the working server +Im not sure what is special about and that would warrant an extra issue +Hmm great! Well lets it +Hi StephenCould you please ZIP the whole configuration and attach so we can have the same testcaseThank configuration folderadded userpassword testtest runs with jdk started by you run the application server on your local machine or on another oneEverything works fine when I deploy remote against a local instanceThe security only comes into play if the jboss server runs on another machineAll I have done is download unpack changing some ports and creating a user due +failed with a timeout +SergeyIn general a forrest of slightly different ctors is harder to use than a fluid builder +Whether or not I get to this for RC will depend on if the quality of the testng plugin improves +Updating fix version for fix +It handles the detection of an Ajax request correctly but still uses the Spring headers for handling a redirect +Initial patch coming +No since the patch has been rolled back and we need further investigation +Are you using an +I think the biggest issue from my perspective here is that some docs might be very large and once they are loaded the will hold on to potentially way too large buffers +This would require merging this change to the release branch which will have to be requested once the modificatiosn land on trunk +The issue is actually from a duplicate annotation on and the build processes them in an unordered fashion so the outcome of which one was mapped is +settings +The patches in this code can be tested and enabled by placing the mailcap file in the JAVAHOMEjrelib directory but this problem is not fully fixed until the jar build problem is resolved +That make sense +Were you waiting for further commentsreviews +Ok thanks for checking Ill close this issue +updated comment to reflect what it actually +As I wrote in previous comment I think its better to instruct user to drag and drop elements to source part of VPE because it works always +the last reported exception should be fixed in +I will try ASAP +See for more info where Lucene uses SPI to load its codecs and analyzers +agree to move out of +Ill try to fix the first one +Fixed with revision + +Rename patch to I did with course I also checked was well associated withI will tried your process to understand why you have this errorThanks your review +Max asked me to resolve this issue since the most subtusk are done +May be we can remove tagsource for +On a session by session basis you could swap the and a in the session +Per Aarons suggestion mark as part of this JIRA +Actually I dont know if this was the case in but notice that omitting any results in a schema valid +loads the revised file as well +I added some debugging and got the same issue will IE and +I need to keep an eye on the IDE +Hey Venkatesh is aimed at getting Hive and Pig to express their sequence of jobs as an Oozie workflow +It might be doing too much +Upgrade reflected in release notes. +Thank youLeo +The new patch is indeed addressing all the concerns pointed out +it does not connect to the another JBOSS instance in the cluster. +If yes it should return silently as the config would have been dumped in the JT itself +I ll try to pick this up for testing ASAP. +What version of Mac OS did you use for testing +Can we resolve this one and open a new ticket so we dont lose track of the issue? Is that an acceptable solution to that +I am not sure if there is problem if we just remove recentInvalidates +If youre interested please me irbouh at gmail dot. +A couple folks have raised concerns about the PEquinox logic included in +not a bugmaxQuantity means the count of files which will be uploaded from the fileUpload +Can you provide minimal configuration of HTTPD to be able to make such setup locally to test +It would be nice if zk can up updated with the last committed record id in the ledger +In any such case the Spring remains an internal container of each application unit might allow for a bit more flexibility in terms of hierarchies +Breaks out the new indexing framework into its own patch +Thank you Haohui +Thanks Knut for figuring out how to set the sleep and deadlock timeout +The only downside is that shell command application forks another process to run the command which can be avoided in a java app +Found some additional issues and want to add optimizations for Single phase commits for local transactions vs DTC transactions +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Blocked by with latest build from master +Patch applied +I was always a little suspicious of since it doesnt obviously produce anything +All set up you should be good to go +Also seeing this failure in +I fix some issue for attachid on line for +The patch resolves the issue. +Thanks issue is closed now +Will upload the patch again to add tests +As Jonathan said this is a soft copy so its probably contributing a miniscule amount to any perf drop +So root does not exist in this system except in some strange areas +Please have the patch for trunk +no test as this is a change to build file only to allow unit tests to be run +Snjezana I dont have a windows machine +I also updated the documentation of the bundle in the wiki +There was no any file on the Usersjluooutbox directory +I attached a patch to correct this +Deferring to we are currently unable to deploy the latest ODE release on due to + Username jgangemi +Patch with tests and new methods forBufferSetBagCollection applied. +Both seem useful to me +Looks similar to what I wrote but it uses a more data structures +How are you meausuring? How is the test done? Can you share the code +Perhaps this one needs to be explicitly excluded or it needs to be specified with scopeprovided +But maybe notDoes the thrift API only provide a way to create a batch scanner? No way to create a regular scanner +This will no longer be true so people need to be aware +Enter integer value in field QOH minus Min Stock less than +When you say offlined what do you mean? They are offline parents of splits? If so they should probably not be counted but no harm done Id say +The test still failing depends on on Tom Whites can commit after which seems ready to commit +Applied +Patch for +Assuming this is out of date is something still needed? +Sorry should have looked at that first Thanks +File or String for the path +is part of the solution but some rebalancing util is needed to +rev issues +Thanks Deepa for the info +Will likely be resolved by removed from derbyall +Fixed +I have upgraded this issue to blocker as it prevents me from making proposals +Let me know if there is any problems with it +cfg file +lugins are optional dependencies for +Please no enable op +Im looking forward to this feature +The changes have been included no warnings remains and the API changes have been resolved issue for release +Sometimes it can happen that this problem occurs again and therefore i havent found any workarround yet +This will allow us to use the same plugin for the admin console as well as JON +Overall I dont really care one way or the other but it does look like change for the sake of change which we probably should avoid +Im the idea +Sorry +Jira to remove the workaround from the TS. +to add a module globally +Ill have a look at that +Same this happened nd time this setting was enabled +Separate code contributed to the mailing listAlso keep in mind that we can start looking at embedding as well as Derby +In my JBDS Beta there is still Select an alphanumerical name instead of Enter an alphanumerical name +Agree with Folke on this in principal however javadoc seemed to be the most prevalent style in the codebase +Changing the Jboss cache configuration is one of those solutions the only solution so far presented in the very least if Seam as a component container truly considers this an inconvenient set of defaults then lets see the convenient set of defaults added to the Seam documentation. +I read through about half of the doc so far +Jitendra for the reviewI have committed this +Todd mentioned that hes seen edit log corruptions from the log volume filling up +Good point! Never mind. +First run at patch for pulling the Accumulo StoreLoad Func code into Pig +This new feature has been implemented for Apache SMX needs to be backported to Apache SMX branch so that it show up in Fuse ESB srcmainjavaorgapacheservicemixbeansupport srcmainjavaorgapacheservicemixbeansupport srcmainjavaorgapacheservicemixbeansupport file data +Thanks for reportingFell free to give the branch a test spin on your system +Its really nice that all the methods are sensibly and compatibly namedHowever its easy for me to say Ill go with whatever the wider community supports especially those who are more aware of the impacts of a binary incompatible change +So if you move the listener call we run the risk of going too far the other way there may be cases where it is never called at all if there is an related error and the recoevery never gets to detect the error before the step fails +Review and fix for the next release +This would enable starting an aggregate service without requiring that all of its delegates are available +Geronimo must be applied before this patch is applied +update committed to trunkr +patch +Overtaken by a bunch of more pointed jiras. +One question do we need a call to executeBatch like process has? It seems like in local mode this might be important +You can check by using with the variableIf the patch works then I think we can improve it a bit as imports a bunch of dependencies which the client is not concerned of eg scala libs if you are using java +Verified on EAP and EWP +orion tomcat and resin all support didnt have any test cases for its datasource configuration +A matter of coordination this ticket is currently assigned to me but Im going to wait on any until you guys figure out what you want to do +If you could try it out that would be appreciated +Unzip the file that contains the html and graphic files +Nick the component was central which only has one download feature afaik but i made it clearer in the titleAnd I did not assignset fix version for this since it was just an idea for now thus I want it to show up in queries +bq +So you could argue that to be consistent we should do it the same way hereIll create a helper class in the next version and add a comment to each of the relevant classes reminding people to remember to update the relevant helper class when they change the class itself +Would anyone want to try this +Do we have any good way to enabledisable readahead drop cache and sync without restarting namenodeJust thinking when the workload changes the customer has to restart the namenode to take advantage of the performance enhancement +See documentation for more information. +When you are ready please say in your submission comment that the patch is ready for committing and not just reviewing +Doing all XML handling with streamed APIs like SAX and +r +We had an unusable release because of that and it took us roughly day to find out the reason for thisGiven that there is a patch which fixes the problem could someone please apply it +Started the role management portlet using and the Tree table component +So all issues not directly assigned to an AS release are being closed +Moving this back to CR and marking a blocker since this is a performance issueTo Howards comment I do not agree that this should be the default configuration for server profiles like default and standard which do not support clusteringEven for other profiles which do support clustering when used It would be good not to use the since it has an impact on performanceI agree on the configuration part though +IBM jdk seems to load the anonymous class more eagerly prior to controller was registered +Added some specs for this too +With large enough dataset some reducers fail with Error GC overhead limit exceeded +This means we can leave things as they are +I did this by making the init methodpublic +Ill investigate a bit more it seems weird to me that the change on sender gives speedup but on receiver doesnt +Thanks Rick + +Fixed at resolved issues to so we can generate release notes +In particular theregistrationForm was tested for thisTrying to track down the cause it seems after loads thestreams through the plugin the args in the form fields have no data +In such a case we should close all the resultsets no matter if they return rows or not +I started to put together some stuff to replace what was there but the Groovy community werent that interested in contributing and I didnt have the time to do things quickly +same as but threads reqsec +The attached script works fine on the current CVS HEAD +It would be good to print out the random generator seed +Ive also attached a patch for +I have committed this. +for eg Invalid Card +Unit tests pass locally +I dont think we want to define at runtime whether UDF needs to count nulls since it complicates and slows the codeI dont think there is a precedent of AVG counting and so I dont feel we need to support it +this should be possible do you need help configuring it or were you just not aware it could be doneThere are examples on the Wiki for other both I guess +I havent seen the test fail so far +To be reopened whenever officially supports such extraction of the underlying +Add Image heading in table above the user profile image +Andrei Pozolotin I have already attached the patch in this thread +Ill comment again if I figure out the juju to repro +This is abviously wrong what with it being the wagon setup for deploying the site not the projects publiclly addressible location which is simply provided by projectUrl +I didnt run any tests with the patch +We would only do stuff with the followers later in the logic +Well the test is heavily threaded so it could be it only fails if you have enough coresThe failure is also intermittent though often I see it fail after just one iteration +They are on my TODO list +This can work off of some implementation detail stuff that Im doing for so Ill schedule it for AS +Hi Bertrand did you commit your patch in the end? We could close this ticket if so +If the thread is not dormant it will finish what it is doing go to the top of the loop and exit because the stop flag is setSo instead of using synchronized use a +There wont be a release +The issue relates to modification of +Work on this should be in the git repo already +A couple notes belowbq +Maven is looking for this in repositories however I cannot find any repositories that have this version +Its reasonable to define one property before attempting to use it and the ordering in the xml reflects the order of operationsIm resolving this as Wont Fix because even trying to support this would result in pretty major code changes +can not be negative +I just verified the fix by run my test with +Hi Jimmy Patch applied at r +Its running in a jvm w everything under currentlib in the classpath +It seems SRX pre configuration issueCan you please check you SRX configuration security policiesPlease attach the SRX configuration in the bug +Because I think that sometimes we want the method to throw an exception rather than doing a check for an empty cluster + +Modified file svn statusM srcref +in Camel we have annotated mbean classes for JMX management +For easier review I removed all the generated codes from Giris patch +Starting implementation setting the shardMapping parameter to any value in the request will add the field shard to each response document containing the shard as specified in the request +Thanks +Sebastien While ipython is its own binary and does as a quick library reference for a developer +for the patch from my side +Sure +Committer need to execute this script to rename files in the repo +They seem to have changed this in except for the forceResume does NOT work in Weblogic +Regards SOA TP Release Note could I have some clarification of what the resolution of this item +It will not have any effect on my current implementation so feel free to change whatever you like +Bumping up the priority to blocker +Looking at the source code of and your last patch quickly it seems that nillabletrue is actually specified whenever a type is not a primitive one +Ill assume that the goal of this ticket is to make it possible to run mvn on each subproject directly which would be nice but is not utterly essential seems to me +Since the changes are relative large and subtle we should probably revisit this post +Unit tests had to be refactored such that artifact names are unique +Its a trivial golden file update + +test patch passed +Note by default now we are building jars instead of one +dims +I dont think its an issue having windows seems quite useful to me or you would prefer a new tabHelp is refreshed but user may not see it if help is minimized or hidden behind another windowNot sure what you mean from Accounting main with help opened I moved to Invoices no changes in Help windowI will have a deeper look ASAP its appealing +attached patch +Looks like two int is more convenient than one long and also saves the conversion between to milliseconds +select this that out of those. +Some more effective some less +Itd also be nice to be able to do this in a Spring configuration file +bq +I suppose the relevant change is to be found in Springs you please for the sake of isolation try to take Spring s class and use it with Spring +Ill commit shortly +It seems that is only to start the broker and even that is only a convenience for the user +Its probably a config +whats your opinion on patch ? My plus one was related to your last patch +Sigh +The biggest remaining cost is copyField of the various fields into the catchall text field +Renamed to with no changes in functionality +Could you try to create a test case that produces these messages +Just changed to +Do not delete table if corrupt +Here is a patch that should resolve the issue +However if any of them fails them we will discard the container and try again later +I strongly suspect that the test failures are spurious but I just kicked Jenkins again for this patch just to be sure +Done +Patch applied + +After some discussion with Karthik I made changes to remove the remaining findbugs excludes and posted a new patch +The issue is finding the XML files that are loaded from classpath and jar filesBTW instead of providing some of the feedback that motivated that jira as comments on this jira would have been better +pattern to make it clear that these are not generally used types. +nothing to do it is already now being computed internally during compilation +Some of them immediately some of them after a timeout +Attachment has been added with description Header file for mirrors download transition resolvedwontfix to closedwontfix +The main idea of our argument was that the default make sense for the new user I believe that change accomplishes this +Wont fix +I looked at a testdiff I have from one of my largedata runs and it looks like the master update seems to only take care of the exception name change for some of the testcases and not all of them +Closing as Fixed +The issue is rescheduled to as it could not make it for the cutoff +Looks good Dart +Please attach version comments from pavel +I would like to commit this in a few days +Moreover this will break when Sun adds something to the protocol +If found that would become the bound web component +Closing all resolved tickets from or older +Thanks Army and Bryan for reviewing the patch +The length comparison will fail if the value is not decimal +ASL for this patch granted and patch committed to trunk and branch. +fixed +Instead of running mvn compiletest just always use mvn install and youll be fine +Looks like all that needs to be done at this point is unittests +jimi is no longer transition resolvedwontfix to closedwontfix +This part of the code has been quite brittle +Thanks Uma +Moving out of +If Edgar or anyone else can provide some code to reliably reproduce the issue we can reopen the ticket +Do you know if this bug repros on any of the emulators +Implementation of DBSCAN clustering + removed eventsQueue fixed render attr +This isnt quite the same as what osgi gives you but is quite powerful +Note that it does not handle the anything tag +This step fails using source and reports error +PS +Thanks for pointing it out +Thanks for reporting this +Hey Ryan commit this to branch and TRUNK +attach proposed looks fine to me +No news yet +Anyway here you are. +Thank you Bikas! +Since this issue has been closed you might want to watchvote for which replaces it. +Change the new patch for release it contains the s patch and also some other version changes +Any update on this issue Alexey +Id add it for this still relevant? +Christiancan you please add the URL to the wsdl in the comments section of the bugthankst recreate the problem +The patch misses some informations due to some Tortoise limitations it is why I attached a zip file +Unfortunately it has always been that way +Integration tests were updated +This patch only makes it possible to deploy from the console +I think this is it +For now I close this Jira it will be after discussion on the mailing list. +The LON cluster doesnt know anything about the NYC cluster or their youre using the wrong tool RELAY instead of RELAY would work the way you want but RELAY is not actively maintained +to trunk +Issue is resolved once I upgraded the libraries + must be exluded also +Well if the reason is to swithch easily to an alternate disk then the resolution to an absolute path could be problematic +You MUST NOT put ze in ze portlet applicationPlease REMOVE ze and ze portlet vill appear +applied rThanks. +You can now use the help option on commands that have required options without including the required options +This is a good find +Its a I add a or a it works fine +It also changes the reader so that it can handle sync bytes at the end of the file +Were in both and Maven Central nowNext step WORLD DOMINATION +Hi Ive applied at r please verify thanks +I checked in the fix under r +Upgraded along with xerces lets see if there are any issues +As well as well as RHEL with Firefox +If these were to get created they probably fit more into thought that was more focused on giving to the Java Platform aset of advanced mathematical functions rather than rewriting some of the JDKclasses in a more efficient wayMaybe I was misleaded by the presence of the and classesDo you feel I should close this feature request and reopen in +We are usually encoding source and destination hostnames in the name to be on the safe sideAnyway the problem also occurs under heavy load with distinct network connector namesRegards Amineusually that meant that the broker has a lot of things to do +Alright committed +However I hope the patch is minimal and only fixes these inline problems and doesnt just change all the div field stuff to span to fix some possible display problem +I think that the should also be updated? It should use the instead of all the different collectors +I you only want to filter you can use the filtering parameter that was introduced in +IN TXT vspf ip ip ip ip ip ip ip mx a include include all +It has an unit test that randomly triggers appends to file of random size with random replication factor +Fixed for IE +Marking this one as resolved. +When the comes up the and are registerd as with the registry +Btw it too need to have the Apache License text +Thanks. +I know that it wouldnt impact portability as far as requests go +is not exactly thought to build a single script and simply execute those parts of the script you give in +So any external mechanism used for the groups needs to be used in authentication as well +I very much agree with having a property for strict vs lenient parsing +Logged In YES useridthre is th esecond thread dump after new start of other bean +Did this make it into +The JMX spec requires that for a given attribute operations corresponding to read and write are created +Right now we do fair sharing between jobs in the same pool so the jobs in the default queue would share the excess capacity in the cluster equally +Resolved as invalid at reporters request +OK why not let me try +Can I ask you to have a look and comment on whether the patch is any use for your activities +Please close the issue +The problem we see probably is +Patch looks good +I dont fully understand its purpose FOP already supports cyrillic fonts as Ive generated PS using cyrillic fonts before +I verified it basically works but have not run tests +What about Commit by server adapter +The commit was unfortunately tagged as instead of +The test looks right to me +I know I am cutting it close but I have expressed an interest in another Giraph project +Ive made the locationmap view extend the class so now it is just a case of fixing the selection changed event in the change to the navigationView class to properly open files when a project is selected +please Amita I took a scan of the patch file and it appears good +Yes I can not reproduce the failure so Ive applied the patch to DRL VM at rPlease check that the patch was applied as you expected +Fairly simple additions to examplemulticore +Whichever is the cleanest +Running on android browser does not connect but I of course cant see the error that causes the failure +verified on EAP Affects and Fix in Version +I did not rearrange anything this is the baseline Java EE Indigo and Java EE perspective with Tools included +It requires a new build of +Committed fix to branches Branch and master at gitronsigal +Thanks Odi +I have created a new patch to solve last dependencies issues discovered on Apache Karaf todayRemark Im able to deploy a jar file containing an EJB stateless will add a new patch where I have osgified the modules and provide soon a new patch with features file modified to deploy openejb on Apache Karaf +Jieshan A few questions + +Any chance of this getting into? Its a fairly innocuous change +Thanks for fixing this +I Version Version +Actually we should probably keep this one open as we wont implement this for all containers at once so will probably be split into several issues when it comes to implementation +Should it be against SDO instead of SCA +TedThanks for the review +is a blocker which will be fixed for I think it is fine to close this JIRA +ACS feature freeze is Nov th +Page preview +Look at wps it is an extension it has a separate web module same goes for monitoring afaik +Looks good to me +When i create a product add inventory facility and receive inventory the invemntory item says but inventory says +As it makes the contract more explicit this is a fair point +Resolved until I get further input on you are right the newer browsers should work fine when keeping track of the focused element with +Theres a way you can use platform mxbeans to find the number of open but I cant quite track down the right API at the moment +Weve done enough work here +Verified by Paulex. +Ill examine the code some and see if I cant figure it out +All tests succeed with this patch +META +Hi WillemThanks very much for all the work on this one! I will give it a try and let you know how it Dave +The issue has a fairly simple workaround so we wont set it to Blocker +I made some progress in the implementation of profile versioning +I suppose it would be more symmetrical with the job parameters but if no one really needs it perhaps we shouldnt add it +Not sure +like if it had way to register w IW and be notified when segments changed. +I just committed this. +r updated cattrack to a variation of this format +Maybe its normal javacgroovyc behaviour +Thanks. +Moving to +Verified in +At Salesforce were working on a feature to separate old and new data for scenarios where we have to store very many versions for the odd historical query but most queries are for the latest data + +But I havent been able to reproduce such a deadlock to see for myselfThe deadlock in the bug report is caused by the index split problems fixed in so the exact same deadlock wont happen again on trunk or any recent releases +bq +We should alsodifferentiate between request write recoverable exceptions failure readingresponse exceptions in order to provide the users with a finer control over HTTPcommunication +Wouldnt it be better to have it work with the +No we removed any dependency on moddtcl and modwebsh long ago +First priority is to have integration tests +Patch for plugin tools +Is that expected +If everyone is OK I would just cleanup the patch and commit this shortly +Please test +I think we can assert that the value should never go below because that would mean we are trying to remove more containers than we requested which would be an app bug +Thanks for the review Ted Enis +Thanks Mark the patch was applied as expected +A note for Windows users absolute path there is for instance filec +The attribute description above states that so my comment about the description needing to state that was invalid. +Realtime get can be a separate issue altogether +Im still trying to understand how this change could have an impact on exact phrase queries +Move to and yeah Ill assign it to myself +Is a need because in server environment when memory gets low the application degrades and slows a lot +trying to understand the of the Hudson logging for which only fails in Hudson and not my build environment +The API in the model package should be kept using +You must have set to trueThe only case that you are trying to optimize is show tables desc +This patch will allow users to decouple their external file resources from their EAR in a portable and JEE standard way +Resolving this issue because it has been ported to. +Solving it will solve this too +There may be a problem with the way the srctestresources was extracted from the patch +Im willing to implement this if I noticed this was fixed by Paul along with +woops i switched affects and fix version sorry +Updated to address issues with revision Thanks Richard! +Yow that audit is feisty +Also do these changes break the shell surgery tools +attemp starts +String operations turn out to be quite expensive in java +Oops ssenessense +Make distribution retrieve das sample based on dependency on the war file containing the java source +updated summary and description +bulk close of resolved issues. +The gossip period is critical +To get the primary node given a conf and a secondary hostnameget is moved from Namenode to +Thats it +Resolved +good enough for me. +But the other one is remain +Perhaps we could include the sources in directly +I hate nodejs dependencies for exactly this reason +Maybe it should be fixed to work how it really should with a flag that can be turned on using to allow it to run in the old suspect wayComments anyone +The former remains an undocumented commandSeveral issues remain to be resolved +for datetimes theyre generally easier to remember than random stringsAlso do we want the datetime of the jobtracker to be in the job name? Would it make sense to just job ids every day? When scanning for a job it would be easy to scan for job from Wednesday the th rather than job after the restart on Wednesday the th +We could have turned off mavens parallel execution though get is not threadsafe but you can seed it with different start ports so that it becomes thread safeI take a look later at it when Im at home +There is no plan for maintenance builds +extract vs identifyThanks more testers for this issue? Shall we commit it +Thanks Julien for the review! + +This is not a blocking issue the split package messages are only warnings +RoyI would try to schedule time during this weekend to find my demo +There hasnt been much progress on +Another potential situation for duplicationYou have a table A with foreign keys to both tables B and C +Thanks issue is fixed and released as part of release +You just need to be able to sendreceive messages in clustering +Core patch applied in r patch applied in r that was an epic patch thanks! +Are you trying to build a high frequency trading system or a test suite JK +Is there any way that the code being used to interact with FOP could be posted here? It would be exceedingly helpful in finding out whats going on +My patch attached +If anyone has a workaround solution for this please share +If were is some time between the the first one might be already into a store and the nd one into another store so we should be fine +The patch is now applied to trunk and branches +Updated the comments +Splits into and with including the serialized bits from the AM +thnaks for the patch +Move the majority of and blocker issues to the next point release +Added adaptor type +for a further note +should be supported out of the box if it adheres to the JPA +This patch supersedes I had an unsaved change in Eclipse so the previous patch was missing some code +Please help me review and apply it +No concerns about usage in cookbookBased on a quick peek at the web site it is not clear to me that the usage at runtime is from code in Lucy written in Perl +This fixes the problem described above +Hein have you been able to put above patch at some use +Patch changes transport Session to use timeout value specified via JVM settings or This patch enables tests and could you please review this patch +Your test seems to pass anyway +looks to branch thanks for review Eli! +Im not familiar enough to be absolutely sure that the omission is accidental rather than intended +Sample Eclipse application that recreates the issue +You use it just like +Dont thnk this is worth rerolling M for +It was made stranger by the fact that it worked perfectly with but did not work withI have now changed to load it as a top level script and it seems to be fine with +The change broke binary compatibility so cannot be part of a releaseAlso the change was incomplete as not all instances of article numbers etc +Patch attached +So the property should proabably have text in its name +Can someone try this alternate patch +For the other types we also throw a hard error if the incoming value is outside the signed range of the target + +Ill save that for a follow up issue though the solr cell morphlines tests actually upped the number of dependencies tests hit quite a bit at leastThere is also a test speed issue that is not on the critical path on my fast machine that does tests in parallel this adds about minutes to the tests +Thanks! +and most people create a separate directory anyway +It generated a component type of where it should beDoes anyone know why or can anyone help me with this issueThanks in advance for your help +done at rev I found some typos and errors which were correctedThanks Nicolas for your work +Ahmet could you open a separate ticket for that? That sounds like a definite Cassandra bug if are still showing up in after they are dropped whereas this ticket is a cqlsh feature request. +You have to have enough independent thought to realize when the advice doesnt apply to other situations +Patch does the following Adds a test to jdbcapi suite +Despite my error would it be possible to upload the sources too +Set Fix version to +This is a duplicate of +Ill make the update to the +Closing this JIRA as EPP core docs development is stopping for the minor issues please open a JIRA against EPP this JIRA for blocker or critical issues +This patch must be applied after the previous one and fixes a related issue where the rules String is the empty string +Sorry I changed the from java to java for testing and did not change it back for the new build with sourcesAs statet above java does work correctlyWill upload a new build in a few minutes +Bulk close after release of +However I think the redirect makes it easier for filters to differentiate urls +Since these are marked as Ps they are assumed to be blockersAssigning these to Len for QE using CR as the version they should be fixedensured by +Synced to trunk +Sure alternatively we could use this jira to track just the refreshable part of the configuration alongwith etc +this leaves the nominal case simple +Please vote +Reassigned to Alexey +New class Identity missing at the last patch +Canceling patch +For hex values if the number is padded to less than digits it becomes an int if more than digits it becomes a long +what is your workspace path +Perhaps is the more correct place to make this change then +Dain will take a look at this one +Thanks +I created for that +It is very useful indeed though it was used only for jsf config file and has been forgotten since +Thanks Owen! +Additionally the directory names within in shims have been renamed to +BULK EDIT These issues are open to be picked up +Instead of using common list form with implemented a separate list form for accounting component +One additional comment if the config name is changed it should go into both and +Much better +Interpreted as a encoded long this is which is an invalid item count +The beanManager was injected with Inject annotation in my resource class by CDI +bq +Thanks for the experiments and analysis Myrna and Knut +Made not to extend as and will not work with a pre class +If the connection is unmanaged on the other hand the trackers are rechecked before they are needed and hence abort becomes a for any +Closed again now properly recorded against Fix Version as well +How about create a new package for the APIThat seems like a good idea to me +Return null and log or throw a more meaningful exception +Are there current applications that want to write in parallel to multiple local disksClearly MR +Adding patchThanks Arun for your agree with your approach +what has this to do with JavamailI dont get that patch file to follow Trinidad coding stylestandards +I make all js stuff slower and has side effects + Search menu is accessible by ONLY +One more question is this implement support wildcard query with the index table +Verified +but why not just use Spring Security instead of a hacked? +Resolves the conflicts with +I upgraded the ConnectorJ library from version to and the problem described by Ralf quite so perfect +I changed it to provided. +We would be better off prompting the for vs during the installation wizard +Ill try thatIs going to be a pleasure to help the teamtkx +please think the issue is still valid +What is the specific problem were trying to solve here? What would acceptable functional solutions be +Restructured to provide encoding implementations +Emails are now being generated as described +Hi JacquesIve a new patch for the pagination issue in bluelight and droppingcrumbs +I have attached reproducer that shows the problem I described above +Reopening to write release note issue now release note text has been written +I think that Onami community has to be useful for the industry in order to be simplify its adoptionI am personally to explore more phases in Jordan would it be possible to implement various lifecycle handling in separated module +I have the Mx version merged in my local checkout however this causes a few issues that I have yet to look into +I have replicated the position is out of sync after reload +This should allow us to release the beta sooner than that time frame +All fixes and features must first go to trunk if needed +Revised plan modify so that we can get the history node id upfront this will enable the versioning fixup to do the right thing without having too touch even more pieces of code +Thanks Chinna! +kill the region server hosted the table +Denis found more issues +In general I think we should try to nuke these methods like copyXXXX and appendXXXX they are dangerousEither you allocate all the time and are safe or you dont and nuke data that isnt yours +Oleg youre right +Yes thats right +According to the attached JDK bug it was fixed in +old dir nuked. +I think it is unnecessary to create a branch +v removes some unused imports +Cant you just go Component on the class and Autowired it in where you have the val +Mike thanks for the commit +Unit test missing see comments on mailing some point Id like to fix up this page to use a more ajax style a left pane listing the services and a right pane with all the details for the selected service +close this issue. +At least by default +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +I think trying to control thing from the tool is overkill +Patch for +There are still some issues about Selector left such as the interruption etc and some of these issues are under discussion on the mailing list I think they are worthing of separatedThanks a lot +And I still need to fix the OOME that happens randomreading because the blockcache is not getting processed comprehensively +The test case I have could probably never work and its no wonder that none of the AMQ developers have come across this if testing is only done with Geronimo +Are there any other arguments I need besides +If the datanode is really dead the client should not get +Unscheduling due to inactivity +This enhancement will be available in the Xalan Java release +So if the log files are smaller the log recovery should be faster right +Issue is the same with a fresh download of Firefox +Suspect this wont happen in or timeframe +tag couldnt run such a check until the page rendered +only +After discussion we decide for the moment just put the examples to the Messaging download page and add a link to the upcoming AS release notes +For a clear contract it would be better to throw if one of the classloader is null at the parameterized constructor +I like best although its also a bit too curled up +I backported this fix to to meet urgent need there +However since boolean is just a synonym for tinyint in and the JDBC for a boolean column is of type BIT Im switching back to the bit column type. +Committed to trunk and issue Oozie is released +So ideally this should be a preference somewhere but to simplify it lets make it lowercase + +I think for lets just document how to wire up Fedora for torquebox to run on boot +Attached could review the patch +Any ideas for a workaround for this issue? ThanksNo workaround from the top of the head +havent figured out how to make one of those yet +And you may enhance this spring descriptor overtime +They are still on the ftp site as well but it looks like they have turned off directory listings +I can confirm this is fixed with the the test updated to check for this issueJira points to a modified test class but nothing seems to be changed for that classMODIFY openejbtrunkopenejbcontainersrctestjavaorgapacheopenejbconfig +But after some further testing with a patched I was still able to reproduce theSo either these two tickets are not related at all or the proposed patches does not fix all casesIf anyones interested contains an attachment with and data files from when I reproduced the issue. +By the way in flex also queries and Wildcard queries are speed up +I made sure that I changed the embedded database Windows Services name so it is now JON Database and its description is now Database for Operations Network +I think we should leave the ability to use qt enabled by default +I dont recall ever reverting it +Ok that makes sense what I mean is that you can not ask for the same type with different parsersYep seems fine since people can work around it if they need to +Hi Lahiru Attaching the patch with necessary modifications +Monty Hood already sent me links to mockups and also mentioned that you volunteered to help us out a bit which would be very helpful +bq +The customer is right +No problems reported +Here is all the source code of intelligent analyzer for Chinese +Thanks Thomas! +Hi Freddy thanks for noteOf cource widgets will be movable +This will start a new build at each child and that particular build doesnt know about all the siblings so the dependency lookup fails +Added informal parameters to html and to branch + for this issue however your solution is not that universal +It is greate valuabe for me because I have leaved away Open Source Community for too many years + a test case which checks following +Unfortunately Im not seeing any difference with the revised patch +I rebuild the package and restart the and I get always the same problem Thats no solve the problem +I have not change implementations +I think having prominent folding filters without adequate explanation about their pitfalls or usefulness may lead illiterates into a false sense of sufficiencyIf it makes sense to have a filter for TR I think that should be a separate issue +A support person should be able to impersonate the user without forcing the user to reveal their password +I thought this was the case but thanks for confirming. +my was not related to your comment +Closed after release. +Im fairly certain this will be fixed when upgrading to since Java exceptions are no longer wrapped in Ruby exceptions in +If it is a new message then the default English message is used when the new message is not changed in the other locale +rerunning with more minor nitpick +many thankssorry it took so long +ill have to think about it for a while +Temporarily resolved this by changing the threadlocal to an Object array +From looking at the patch it looks to be localized mods to only +It draws every glyph of a given font and creates a png pic of themAny comments and improvements are course we are interested! The easiest way would be a patch +These changes make work on Windows +Currently SSO integration doesnt work with Tomcat bundle. +Your issue sounds like the error isnt being reported at all +At revision I committed an initial version of the rebase contractAt revision I implemented this for the H MicrokernelAt revision I implemented this for the Mongo Microkernel +Hey Brian question about this +Please see stated in this was just a bad artifact creeping into the local repository cache. +This is good for commit +This way it is possible to get the logEnd after the scan is closed and check whether it has been advanced since the last log record was processedFor the verification of prev I guess that the best is to verify this for all files processed during recovery +Seems we should be consistent here +OverallSuccessfully executed smoke test for this patch which covered following things Allocate Deallocate Info List operations withwithout static dfs and tarball using hadoop andSuccessfully executed hadoop script using script option +As for the attached patch Im sorry I havent had time to get to it yet but I committed and Im pretty sure some rebase has to happen +Thanks Knut +Rails appears to generate a tree of methods or code +Now that work is well underway with AS all previous community releases are + +It is in the dev branch and will be published soon +theres still the problem that has not been deployed and the reference to in its pom is still instead of patch with correct versions in here it is the important partIf you run the unit test youll see it fails we havent figured out how to deal with collections yet maybe somebody can figure out whats going on class doesnt register the class descriptor which certainly causes this kind of issue +Move fix target to updated +Starting at add method allow the parent product to be passed down thru to theThis patch also corrects the selection of surveys for an add to cart survey +Added get method to that returns the name of the parent activity that spawned the activity for the monitor +Committed with revision +i tried also with Tiles +The problem is not about the import version range its more about that scrdeployadmin are not deployed at all +Ive attached the patch to JIRA +Attaching a patch which is a first step in converting this test it only handles the following test cases set to false set to true but not set set to true and setThe next test case will need to examine you forget svn add before creating the patch? The one I downloaded was almost empty +Its giving the same +I wasnt looking at the mockup though just unifying the search and dependencies pages so they have the same look +Sarvanan Can u pass on your client code which you used? I am looking into the issue and unable to reproduce itThanks ashitosh I am now looking into this issue on this side as saravanan is no onger working on this +Merged to doc branch at revision +when is fixed this would resolve on of the most important partsFor me the solution sounds ok +I think we need to document it well and put it on visible place +AwesomeTomasz you are some kind of an Antlr deity +Support for using to save data sink to different table by demux parser class annotation +I have attached the new patch tested it and it is working fine in my local environment +Thats exactly what I thought of but I can not give you any hint on how to include that with XML test suite +Thanks for the reportWe do not use Commons IO directly because this would mean an additional dependency just for this single method we use +Yeah I would say good enough for a workaround for now +there should be a stack trace for the batchinsert error toothe range one involves bootstrapping +I can make configuration changes for you as I have recently done this +patch to introduce the rmf commandthe last patch contained junk from other changes in my looks good running it through tests now +Since runs on ever start there should not be a need to write it into etc on master as per Chiradeeps comment. +Fix has been committed +I need to determine if this is sufficient +I tried recreating and hit the same problem with an IBM JDK but when I used an IBM JDK it worked fine +You can run the test case to reproduct the bug and apply the patch to see if it can resolve it +has the same browserversion do you use +Axel have you resolved this on your own? Id like to close this bug if soDid you have any suggestions for doc revisions? Please comment +Namit Please set the fixversion +Here is my and you can see all the mailets that are activated + +How to efficiently go through the auto testing result? Is there always non relevant info in itThx +It can no longer be used as the new API expects the partitioners to extend the new Partitioner class +and i try to select that my web browser hang up does not respond +Im out of town for the following three weeks will be back Monday October th + push to +bq +Initial patch for review +Looked at the dump +I tried to reproduce it without any luck +I would like to start working on based implementation +Thanks for the good work +If we ever did that then some of these might get a new lease on life +At the current time please use FF +An even better idea would be that the installer checks and shows a list of detected partitions on the network one can either join one or create a new one +Click Next +Add Navigation Rule dialog that helps add a is documented. +Commented resources and fixed yes I completely in svn repo +Thanks for the help +btw the compiles for me now as well +manuel is not an email per se +This test has quite some subtle changes in +But the question still stands what are they supposed to be set to +James youve been working this for quite a while now so Im assigning it to you +that is the generic way of exposing the direct methods like get can be exposed as well +Fixed in branch and trunk by removing the static workspace +If not let me knowCheersCould you please refer to a path for what you call Docbook content files? I guess its the HTML generated file? Why PDF +I will be submitting a patch for review soon that embodies this implementation +it looks good flavio just two suggestions i think it is too much information when you start talking about the naming convention for the znodes in zookeeper +Pass to Charles +I committed this +Duplicates +and committed to +The log shown indicates Maven is being used via the Roo shell +Otherwise looks greatExcellent work fixing and thanks Andy for the detailed explanation +So I cant reproduce the way that I check the optimistic locking is a bit different +Not sure if theres a more appropriate way to fix this though +Hi Luciano thanks for pointing out the problem the new patch includes the missing files +You may give the head revision a try +What do you sugggest instead +Lets just iterate on this with actual changesets instead +Sorry +I check and system properties and cancel tests for all Suns JVM earlier than +The Clearcase client is +Quite scary unit I figured out what was going on +JavierThen use the nd java class attached this does exactly that +Thank you Tanping. +of patch was tested before I the code +The new rendering system code which has gone into Roller provides a better chance to catch and respond to errors during the rendering of weblog contentWe now do a better job of catching any exceptions that happen during rendering and give more useful error messages when appropriateWe still need to improve this for the struts based adminauthoring UI though as often exceptions generated from those pages result in a lame error page which is not very user friendly +Trunk version functionality is complete + +degreesAfter some testing I realized that the text doesnt have to be rotated by degrees but the shearing has to be corrected by multiplying the values with +It looks ok to me but Im not an expert +by adding the serial number or by using the SHA hash +Looks good to me we ended up requiring a field in the validation function. +The validator in previous versions of Spring IDE wasnt elaborated enough to care about the different types of bean references +Yes Im using that oneThe findbugs plugin is seemingly just cancelling the report and the Maven build is reported as SUCCESS +Completed refactoring +TypoI meant simpler in my previous comment it will be simpler that way not similar +Made a few more corrections in several of the links +working on a fix for this +Setting port numbers is being discussed in +This was fixed a long time back +Good catch on the missing javadoc +I was just using it to help debug +Thanks for providing the patchI have applied it in Rev +It happens in Pax Logging when Exceptions dont have any message +Fixed in rCopy annotations from the original method to the bridge methods in +Credentials problems resolved patch applied +there a workaround for this issue for those stuck using for the time being +issue already fixed in the trunk +Thanks for the report! patch prevents from skipping nodes if the switch is set +Please can you update the status of this Jira to indicate when it will be ready as we intend to release M on MondayThanks jtsremote QA tests are failing the functionality will not be ready for it work on? +Set Fix version to +Closing after no response to the comment +no problem you can provide a patch with a test caseP +What are the module dependencies for jsoup +Yeah its quite verbose in the log but thats the pain for the gainMaybe I can force the rollback with an exception triggering it +Backported to +This is a significantly shorter example that triggers the described behavior +Integrated to branch. +Did you try changing the settings of the +I believe the checking on for this issue is causing the failures in the junit tests reported in +I dont currently see how to fix this do you have any ideasI could change the ERROR to a TRACE +Im sorry Ronald but your argumentation is not fair at all youve used this property as a workaroundhack to support your need and then you come back to us saying that weve introduced a regressionbug that should be fixed asap +Were now well passed still requires BuildrThe plan is to upgrade to the latest Buildr once is released +I have the relevant pieces of ready to include but I will wait until I know whether this is the appropriate issue or a new issue should be filed +to making this a blocker +The notes youve added are pretty clear so I dont think that comment will add much value even if its reworded +Thanks! +. +No such rich at +Passing by hadoopqa +This is not a functionality but a test issue +Unfortunately an error appears in a complex application +So it has to do something with Byte Order Mark +Add commented out line with GC logging options to HBASEOPTS in good should probably go before it so that the HBASEOPTS part of the addition works no? Otherwise ye thats fine I can do that +Agree config file option seems like a good idea +Thanks Mike +More specifically I have added variants of the methods open add and create that take the buffer size as a parameterCore tests pass fine for me +Add compression iobuffer dont use templates anymore +However how to tell if it does not have transactions if the file is not emptyWhat do you mean by this? If its not empty but has a valid header and an OPINVALID then it is the same as bytewise empty and can be deleted + thanks for fixing this +Pat may be able to confirm this but my understanding is that the xhtml theme will continue to use table elements and the xhtmlcss theme will continue to use pure css +Marking it Patch Available to trigger automatic tests +This sounds like an issue where kvm module is not enabled on the host +I could discover an important bug +Editorial pass over all release notes prior to publication of +Closing as per request of the reporter. +nfsbf +Unification of the resulting models is probably much easier than the unification of the model building process itselfSome of the problems I have seen includea all of our clustering and classification models should be able to accept vectors and produce either a category or a vector of scores for all possible categories +And as we got tree structures elsewhere people may want this to be consistent and in all places where we have thus its much more work +Brian the new patch defined a ESB Project With Java facet preset so the default java facet version is guess its what you need +I always have to remember to +Thanks for the comments Robbie I will look to address them +Fixed the link in svn main site will refresh automatically shortly. +So jani will be joining the admin team and also help in maintenance coordinating with existing admins +Cool thanks +eclipse has a code cleanup tool that can easily do all this automatically +Ive just committed this to the HA branch. +Ive added it to rather than it feels like a better fit there to me srcSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision . +Bulk close for +thank you for contributing the fix! +yep Chris is a committer now. +Add hbase contrib tests to CLASSPATH + +Otherwise I was going to look at running some longish stress tests to check the throughput and latencyThat seems like a way I would do it +Im not able to reproduce on my Galaxy Nexus with the latest Untappd +Wendy will provide some automated test in the near future to prevent any possible future regression +STRSPLIT return a tuple of variable length +That seems to work for me thanks + +If it is handled at the same place as preventDefault then it wont fix this issue +lib +Then i switched to Iimap and then it seems to work +Thanks! This is awesome to have in HiveCould I please ask you to document this feature on the Wiki? Or if you can list all the features it supports Ill happily write something up too +Here are patches for and trunk that resolve the javadoc dying for the patchesI have applied it on and I also fixed all the shared javadoc warning +Hi cassieYou are right but now my question is what about the gadgets which are already using the old style gender like which comes with Sample container and there may be many more +I suggest we do it as part of +All Diffs are found in single file in the root +I agree we should never return null from get and handling the different cases under more than one issue sounds like a good idea to me +Ill stay tuned +Why turn it off in production? Current functionality only allows someone to change the ecommerce theme from the store screen in catalog manager +FYI the issue is that by default the H test setup assumes data in the GEOTOOLS schema so either the test setup needs to create data in that schema or the test setup needs to configure the datastore to use the public or default schema +Dont create a in +Thanks +It was my mistake +forgot +No my unit tests are already executed by Hudson I just give the report to Sonar +Gary I ran the tests in several times and they succeeded +Spring should support got to work it after a little hacking with adding missing JPA implementation methods into and injection of in example classes replaced by declaring as extension to for database access +This issue could be related to +No bug is about only +Looks like we have a work and quick installer now! Thanks for the help guys. +I made some comments in the RB +Would that work too? It seems simpler +The being to eventually allow users to plugin in their own entity that requires which in the long term is correct to do this cleanlyHowever is far more ambitious than this issue and is therefore bound to the +Something we may want to consider for if this catches on is making splittable +Then I could go further +Is this documented +The problem of choosing the wrong package is not worse than in the dynamic import case as there is also no guarantee to get the correct one +Send us a diff of our changes and well analyze them +Added formatting tags +Let me know how it goes +small fixes to increment algorithm +This might be useful for the frequency inverted lists +Does somebody already came up with a good name for this? + doccomment spelling fixes +Dmitry isnt resolved in JPA details for the first time when update its ok +Going forward were focussing purely on the JDBC persistence option with the journal as an optional performance boost + +Im sorry I cant reproduce the issueYou have probably had a connection problemTry to install the connector again + +IMHO if it is reproducible on other environments it must be fixed before since it makes the Studio completely unusableMy +Updated for latest revisions to more time. +Proposed patch +defer all issues to. +HiWould you please try my patchThanksBest regards current Harmonys behaviour is fine +Mostly because by design the dialect is allowed to throw an exception for any method that takes a connection +Verified by Paulex. +Bulk close of resolved wont fix cant reproduce etc +Harumph +So there is no problem when running with local meta store +Its an update query but fails both on insert and update +That is not an oversight it worked you got an idea what is going on here +pushing to to to to the approach of adding a new jdbc importer property use defaulting to allows JDBC drivers to use the specific name as the teiid name the name in source remains the same +Thanks Namit +Yes Arjen tried with disabling auto but no luck this time also here we have two representations with same root node is there any chance of spring caching the first rep +The WSDL shows a plain http address but when you try to invoke the endpoint using that address returns an HTTP response with the https location +Can you also add a negative testcase for skipped jobs +at the new screenshots Feels like one pageversion is missing? the case where a server is already installed +I had a typo in my junit test so it looked like it was passing but it wasnt +Attached links it to Note facility and order help will open to BIRT help in current helps system because they do not have a webhelp implementation +Unit tests and code coverage still packages in question cover functionality provided by the RHQ plugin and Swing based management GUI +Moving to +Closed for +So let me get this straight +Fix for issue is not reproducible on current DRLVM sources +Since the tab size can be different depending on environment many OSS uses space indentation +So do you want another patch or can you just remove the trycatch blocks from the patch? Also I dont know how to mark the tests as fail for Jackrabbit +Note that Im not building full release packages just pushing snapshot jars +I wonder whether there is some problem in your environment that is causing this +This time all code related to this fix should be in. +It fails on the test +I tested it again with the newest versionsWith version during creating the third cache UDP bandwidth is at max +pending sure what happened there Alejandro told it to run +I have attached a for after the problem occurs +xml it adds to the file name +In the class this introduces a compatible change by by replacing params of the type Object with SpanAnd it should stay Object because the counting logic relies on equals and can also work with many types +Let me know if there is anything else you any specific reason why you reopened this issue? Which specific still causes a problem for you? this as fixed until proven +By the same tokenFRED cant set A eitherUploading the spec again fixing some opaque phrases that Kim spotted working on +Another thread dump +Id need some insight from someone into that before I would put it into placeAlternatively feel free to fork and commit another version +Committed as of rev +Version is no longer being maintained +reassigned for review. +Nothing to verify here +Possibly XHRWITHPAYLOAD is leaking in some way +need to do any service class changes +Closing all resolved tickets from or older +This number is likely an as the experiments are in mode on a VMNote that in the design represented by this patch I move the process into the cgroup asynchronously so the latency is not incurred while starting the process +But without this we cannot continue with the patch in which will hopefully ported to Hive as well +It appears that kryo generates binary output which is thus notI had difficulty to serialize Hive plan objects using Jackson which seems good only for +Verified on my vmware setup extract volume is working fine now +Closing this as everything is implemented on the c backend +The u switchinstructs cvs to create a unified patch which afair instructs cvs to take intoaccount new files as well +This is probably the same as you have a full path as your? I have done the test but havent met any problem +Also the improvement on will minimize timeouts but it doesnt help at all with dead nodes which we should discover and avoid +There are numerous exception in core for Border only and even itself is rather complicated +Tests passed for this bug looks fixedClosing this issue. +maybe seconds is too much +Did you have the Roo shell running +So this will affect existing applications and release note is needed +Currently running test on branch +Can you provide a sample application that shows this + +I updated our machines to include +Also Im fairly confident it works in as well +With this fix when an empty element is present soap body and all message headers targeted at the role are signedThanks +Thanks very much Dyre! +The issue will be closed after testing on Windows +bq +I tried that first but for some reason it didnt exclude the jars +Attaching correct for hudson +I updated the title to better reflect the issue +I am attaching the whole scenario and how the client uses what has been described +After this patch go to trunk then we can go ahead to make patch available for which depends on this +Resolving. +includes the suite derbynetmats +Go ahead +Hm I must have missed that part +Someone might want to take a look into the Maven possibilities for Manifests when the requirements for a manifest are certainMarked this issue dependent to the Java component feature +Dan can you take a look at this + on the idea +This should work +took a look at the fix seems ok to me +closing since I dont see any way to implement without crippling array indexing performance by injecting a virtual function call to +Its clearly not cross platform and wont work with the jdbc thin driver both of which make the change unacceptable +There are no implementations of the yet so I am bit hesitant to even throw out a proposed API for that yet but the current one is pretty generic which is both good and bad +You can set to false in and the URL will not use templated preferences. +Bobby I see your point and I agree that keeping them missing is the simplest solution to avoid complicating the issue +looks greatNit excpetion +Branch has been merged to master +Committed patch to the trunk as revision Thank you for the contribution to Derby +Woops thanks Renaud I just committed this +Verified with +I will try some things out and get back to you +I have committed this +application initialization includes ADV this is a good point +Ive a basic implementation will wait until after before committing as it might need a little hardening before we release +This is a good solution to limit the space of tables that is searched in. +Works fine here artifacts resolve out of the Apache SNAPSHOT repo +Initial Committers Marcel Offermans Karl Pauls Christian van Spaandonk Clement Escoffier Felix Meschberger Carsten far as I know all initial committers are existing Apache committers +So you have a pointIn any case give me an example where that behavior is desirable and may revert the change +The commands git push and git reset were being overlooked when a check to see if a git repository was available which caused these commands to be available when they were not supposed to be +INFO Running for rows took ms +This interface is used to cache nonces and also a combination of Timestamp Created Strings with message Signatures +also applied it to +This also seemed to be the thought mirrored on the forums by some of the folks +Im going to work at getting the initial functionality working on this then work with you to integrate it into the settings is now working on branch +They also do what they do as part of a bigger picture of making it easier for dumb users to use the systems +I will trigger the build manually +Im happy to test against SQL Server +I see from your logs physical address wheres mine logged the actual pc hostname +now Im having trouble reproducing it +I just need to change the messaging there and refactor some bits +Fixed the alias generator logic to correctly change the group of the order by item. +It passed on my machine. +One issue though is when you subscribe it stays on the same page and the image is not refreshed +This patch modifies the otherwise the compile will be failed +The option shellcommand can be ignore by setting shellcommand is binbash by patch can let option shellcommand not be mandatory when running distributedShell program +The javadoc tool did not generate any warning messages +NikhilCan you please show example usagethanks new failed due to incorrect port associated in the endpoint URL +My point about the newer services is that I had originally thought that these were going to be part of and not +Patch using +As an interface its fairly simple to manually patch +fixed by +Committed at subversion revision +i hope the user will be able to understand whats happening +The fix will require upgrading the Plexus components +on comments closing. +But that would be for another ticket + +Working closely with eucaools developers to fix this issue +In a larger cluster where many nodes are being used as coordinators each one will have to observe the faulty node performing badly and this could be observed as a global degradation in performance at least until the snitch on all coordinators decided to quit using the node +Committed a test case in r +However I think the requirement is also to support features like streaming +IF people scream for it by default then we can revisit +Integrated patch v to trunk after adjusting the javadoc lineThanks for the patch HimanshuThanks for the review Nicolas +core +Id raise the severity a bit because all the docs both on the website and in the distros are wrong +Pavel field is changed in synchronized block that is your patch breaks access serializability +Committed +Someday people might consider this a miracle +Sandy +Any chance this issue could be assigned for review +Added an ee test since I needed columnprune limitoptimizer +Additionally it adds cometd client id to messages coming in through cometd +Is this problem not present in + this is it duplicate off? I have the same issue without annotations with this is it duplicate? Exists some workaround? Thanks +Id like to see this in if possible for things were working on +log files in the source archive after runningmvn install after a successful build +Joakim thinks that this issue is related to on different projects and it seems to work properly with the latests patches +Fixed. +The changes are isolated to the class and they are very easy to unit testSWAP is the default +Thanks Chetan Mehrotra for the patch +I am all for the new PHP namespacing feature with logphp and leaving the old naming style in series as it isThe pear style of naming must die +However a small amount data may not be replicated to the followers during a failure and will be lost +This seems useful and from the description alone I think I understand what it does but you mention an attached figure that explains it +Is this temp fix available for download? My app is going live in weeks +will be rolled back from branch. +This is a pervasive problem throughout the code +You have list of what needs beyond mighty Jesse +You could close this case +If it introduced compatibility issue how about we moved it to next release +Extracted +Thus when the tried to read in the config coupled together with which currently has this set to false the config was being overriden to be whatever was specified in the default file +for embedding inThink it out loudly +Attachment has been added with description to reproduce the I seem to be having some problems with Bugzilla and attachments +If thats ok with you Bela Ill go ahead and do it +we have been using axis for years and our platformframework is build around axis +Could you please upload your latest patchOh sorry it is already therePatch overall looks good to me +I thought that maybe a better fix for some of these issues is to override the in and tell it that closures are actually functions rather than bracelists +But that class either uses stored in the index or using the analyzer passed to it I think? Are you using +Hence this fix will be excluded in +Can we see a use case for this kind of configuration +For what its worth the following issues for the same functionality on other have been committed so far +Sorry false alarm I was using the wrong dialect in the hibernate properties +You can try by just changing the path to the projects I submitted +Every expression or function that operates on numeric values will return a double +shutdown script does now require an servicejmx url but mbean still writes out a jnp a sync so that the urls match andor can be used by shtudown script as discussed in issue description +prefix for system properties per Willems the latest patch has been appliedWe should add the docs for this new component as well +Test failure seems to be due to +Revision Author ieugenDate Message Issue Migrate James web sites to Apache CMS initial import for James site with Apache CMS we can use to put other svn versioned documentation files dont you use the same template we used in the maven skin? Im not against changing the html theme but mixing things doesnt sound as a good plan this issue is about using Apache CMS not altering our skinlogolayout +Give it a shot! +It uses javaxfacescontext +Please verify it +Should we also add links to popular places that make some of those lists searchable? +Lars This patch does include the generated file which is the only generated file that has changed +war file within tomcat webapps directory. +Update the issue to have Brian Minchau as the reviewer once there is a patch +Do you mind confirming that this fixes your issue? +Added a option for finer grained stats per +There are no changes to which seem to use this mode +As it is Im closing this +beasting uncovered a good seed +Thanks for the patch we applied it +It seems to me like a good for the moment +There is no reason for not adding more complicated subclasses if you like thatJochen +were backported to as well adding the latter fix version. +Committed revision +bq +thanks chris +I also own ejb now +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +generate in was printing the methods only for immediate superinterfaces not transitive superinterfaces +Maybe theres an error in the engine? +I would try to checking out from SVN that revision then patch then updateHope that helpsMatt +The ESB packages that are attached have dependencies on this queue so Im not sure how that happens +Hence this is an environment issue +When I attempt to explain to others I tend to focus on its ability to facilitate mobile apps from one +It is still missing in the Windows Installer +This patch removes many calls removes and fixes murmur +Christian can you please explain what plugin throws up mean? Also can you please provide detailed step by step instructions how to reproduce this issue +Defer all open issues to +Many thanks for this enhancement +Verified by Ilya. +Performance test actually doesnt use jBPM +ran cleanly with +Draft patch that passes down the router and uses it for hash based routers +See on Slive andTodd Which bug +It seems that when applying changes to in revision the modification for application got overlooked +I like the idea did not review yet +Patch applied at +I synched with the latest upstream kitchensink and still notice this problem +Please dont assign time estimates as the form says its Only to be filled in by the assignee before starting work +Ill commit this today +Also need to add disabled labelClass selected attributesdone +Thanks very much. +Verified +I just committed this +Good news +r applies the patch that adds ehcache and the config file to get rid of the warning +The problem is the compaction process may never catch up with the writes if they never let up and eventually it will run out of disk spaceCompaction should be attempted when the write load is less than full capacity +main diffs plus test case +attach a new patch +The addendum I am testing now +Ill close this when there are +BTW using TIDY gives me a good workaround for the test suite but this will be unacceptable for customers who want to use Neko +I wouldnt suggest you drop them to your production env +Closing this record since is now released. +Why isnt the tests jar a hadoop compliant jar that includes junit like any other hadoop app? Is there a reason to treat it special +okTest userdata as POST size k +Can you do two things to help us diagnose this? Try it with the latest posted build? Attach both of your +The PDF looks much nicer now + +The fix for XA HA has been committed to HEAD only currently +I very sincerely hope this is the last Ill ever see of this test. +Pushing this in +I prefer to do this in wicket app and only for wicket app +Hello Bhumiryou are right +I added some tests and committed this. +but when I changed it to configuration is changed to custom not to Seam My point of view is that AS is working only with Seam so when this runtime is selected everything else should be set properly to be ready for seam runtime added on last page of wizard +This will contain documentation on how to use it +Alejandro thats exactly what I plan to do +I think this can be closed +We pay for this bigtime looking at hbase under load with a profiler +i know now why our cluster has problems with heartbeatwe are using ibm jdk with gencon garbage collector +Thanks for noticing this Kathey +Number of cases printing reduced by commenting down to a few key examples +Youre comparing with Right +Its unlikly that this is a bug +After its important for CREATE OR REPLACE VIEW to preserve any existing view partitions +I understand the importance of having a forum topic and I will be sure to adhere to using that avenue for future open ended discussions +Attaching patch +JFYI patch above broke bean lookup logic and removed support for context beans. +Does it make sense to create class that would extend File and override unsupported operations to throw +Thanks Sebb. +test case to make sure does not configure objects is better to add this test code as part of the patch in the common itself +It is actually a duplicate of +Yes that would probably work too but if someone does happen to put in a count that is close to LONGMAX we can still run into this situation so I would prefer to just leave the code as is +Only one nit we can rename the new test from test to test to be consistent with all other test cases +I agree jspx is fine +Fix patch so itll apply +Thanks Jay BalunasSee ya +Also somehow some callbacks need access to the origianl spring integration message to access values from the header like the soap action one +ThanksOk so well discuss this one wednesday call to ensure it fits our needs and if everyone agrees we can switch to this approach next missing I think it could appear to work currently because the current aggregated site may not have been cleaned +now performs this check and Status Code responses no longer result in an exception being thrown. +Jarek is there a timeline you are looking at +It just means that at some point someone changed the group in which the file resides without changing the owner at the same time +Pushing to handling only for the length parameter and added a note in the release notes that negative start indexing is not supported by dbderby. +Avery what do we do in this case? Looked at the log dont know why the build failed the one I did locally went fine +Thanks Jaroslav +The patch replaces the previous file +The v invoke is better +Hope you got my point +yeah +Implemented. +Attach a new patch to fix the problemRex would you kindly help me review this patchThanks a lot +I see this in IDEA too thought it was that IDE only +Maybe RS will go into some crazy state where noone can reach it except client +That is why you see it after fifty invocations and you see it first time when in body of a script +There probably needs to be a discussion how this should be done in the first place +Unregistration fails but the application doesnt have the chance to retry according to the configuration +IOW I would be happy if the old code structure in loadResource is retained when pulled to the new method loadProperty +In that case we have a bug in the plugins or the s in that were not specify a repo list right +really glad to hear this! +Thanks pass over all release notes prior to publication of. +Having alternative break characters was another way to address the issue that ellipsis would provide + +Same issue with +You had clover installed and it worked for you with java? Maybe the comment is old + + isnt a bug but can be improvement +It would be nice to translate full content via some script this may cause some formatting problems though +For that matter we might deliberately make these hints in the implementation so as to avoid having to bump the edit log version number but persistent in the trunk implementationbq +The plural method doesnt sound all that useful for me but if youve got tests and javadoc with it then i dont mind adding it + +Any idea +Can you provide the Requestmap entries that youre using? Theres no indication in the sample app of whats secured and whats the sample app says youre using of the plugin but the bug report says which are you using? did go to the database for each request when using Requestmap but that was fixed in using caching +Thanks for providing the patch +Weve always said that its which kv you will get back if you have two writes at the same timestamp +The same error for class bridge methods is not considered. +If you add it manually everything is good. +I must be too tired or something +Note that although this is likely a Derby issue the IBM SR does not show a +Hi SvanteThanks for taking the time for reviewing and feedback +running tests +This is the unit test +committed externalized strings in in patches look really safe also +Committed to +I opt as well to change the constructor +there are a lot of hunk failed errors +I updated it just now again to handle some bugfixes +thank youcould you pls +and sources +Seems to have worked this time closing again. +Duplicate +I think that makes sense as Admins will not be able to stop Hive CLI users from bypassing this mechanism anyway + Committed to trunk +ThanksCan you rework the patch to NOT depend on catching aAdditionally since the buf is becoming volatile you should use the paradigm for that field +patch applied thanks Ray +Hi I just used standalone and AS deployment of seam project with mvn integration facet enabled does not work +This change was inadvertently removed in revision + you please test with the generated client and send the whole stack trace +If the would store the messages in a internally this would also be solved +Perhaps a key reason that collections isnt generified is that I dont use in my day job +this file has been granted to the teamyou can use it freely +There were quite a few changes in the API as part of the changes but all of the common code should now be in shared base classes as suggested here. +Thanks for the explanation Kristian +When building EAP you must not use your local +Moved default to new closeexception handling method too +Please attached a small program that demonstrates this behaviour +Attaching a diff with additional changes so I can link that through the forums +Sorry about the formatting I didnt realize that the description used wiki markup +The directory listing paragraph is already in production +It sounds like the permanent fix for this is close +What do you think +Possibly implement their recovery interface or a wrapper to it. + +Example of using xml component to invoke an or fix later or not an issue now +Itd probably be pretty hard +All addremovecontains etc have to scan the array to find the match which will be slowThe decorator can be made to implement List and will perform better as it has both a List and Set internallyIs there a case for both types + +We can update Hives Thrift dependency once this releasegoes GA +when we eliminate all the getset and exception object creation this becomesbbsome PEI with an exception edge to +Were currentlydiscussing relaxing the keep conditions for these situations but the spec is alittle ambiguous about this +wont dont think this is a reasonable position +Done +Would it make sense to settle for approximated answers and use +I find it really bizarre that you would be seeing this problem on Windows +Not seeing this in JBT +Will review in a day or two once things on my end settle +Code changes for this jira are already in +The Features mode is the new feature I added in which a user can choose annotations of certain type and further filter them according to a certain feature and feature values +I promise thats my lastOh yeah and co credit would probably be more accurate +Apache machines are loaded +Am I missing something on this one? When I omit the version Im redirected to the Spring Framework home page +Thanks could test this in right +Resolving the issue but a comment on unit tests is still sure the is just a bunch of stuff portted from memory datastore +Search tag elyogragspringclean +The old clients no longer exist resolving issue. +Thanks guys +I still need to write some tests for this issue +What is your unix account id +The patch looks generally good but I dont know enough about it to say for sure +is out the door adjusting on all remaining unresolved issues to by per the first comment this issue is not well described but other tickets do describe the real problem. +Caused by a jQuery UI upgrade and resolved with +I think Im using the embedded in the Seam directory but with an updated that has all the new hibernate code +Revised Summary and Description +It will make some failure of test case +every table calls balanceClusterMany logs as Skipping load balancing because balanced cluster will be printed +I suggest changing the exclude list comments according to attached for exclude list applied at r. +I committed the patch +I think we should punt on this one +Again you are fixing an issue around the same time in multiple branches its exactly the same whether you have one JIRA for all branches or a JIRA per branch just less cumbersomeAt least thats how we handled issues before EAP came when we used to fix a bug in AS trunk and branches back +Given that Im not sure it really improves things Id prefer avoiding that if only for the sake of having less chance to introducing bugs +This appears to be more a problem with tomcat rather than with spring +What is the daily cost for running? With the backlog of patches we have we may be running for a bit +If we start cutting a RC of Camel then I guess we should have the bundles sorted +Able to delete the private gateway after VR reboot +close off release +Resolving as fixed in an unreleased versionComment viewable by all users +Committed revision said above waiting for an explanation as to why the class loader trick should be removed +Im not convinced this is our bug +Lets only handle diff report computation in this jira +Agree with Flavio instead of exiting we should think of solutionClosing the session might be difficult to implement checking if pNode exists after exception seems to be sufficient +having a custom interface that if implemented by the bean would be called to return the actual bean +My point is not just limited to the default value +Stefan please submit separate bug report for this and provide some test projects that would allow to reproduce it +We wont migrate site +Committed changes in revision . +reopen for backportreject label +The problem in of Seam +Patch applied as suggsted +Its non trivial for us to do a complete solution due to the number of hits we get +For information it does not occur on Mac +something failed internally but was allowed to continue to do the dir rename at some pointThis is not desirable +Unsetting the fix version for now until we have a suitable Spring release available to support this upgradeThanks for pointing that out Freeman +The build with these changes has been pushed to Brew and will be live within an hour +Patch attachedBefore applying this change should be discussed +to make this public +if this causes failures when having ejb put into wars which is important for EE then yeah we should get this inA shame really since if they are exploded things like jrebel works much better +because of this issue i ended up to drop spring and use pure active client connector +Ok I will drop this +Why would you make reference to a file in that no longer exists? A properly applied patch by the committer will remove this file +mrepositoryorgmortbayjettyUserstucu +targeting for Beta but there is nothing here explaining what the problem is +We also tried to integrate with Hbase but we are getting the same with need to use the bin script or the crawl script +Looks good +Asking client users to manually pack and unpack data according to validatorclass seems a bit wrong this may even be rather difficult in some languages +Closing as invalid because it appears this is not an problem +I think the underlying issue is a possible false assumption +Patch for the branch to move to Saxon +Thanks a lotBest regards +MS Access also map boolean values to numeric database fields and they also use for false but often they use to represent true +Also we had this nice trick in place to generate the category dinamically from a stacktrace this was quite clever but is generating lots of stacktraces which slowdown the boot process not really AS compliant for the Enum love the idea +Patch incorporating Hongs comments +However if the username or callbackhandler class arent explicitly specified and if the user doesnt specify a value for property then EJB client library implicitly adds the to connection creation options to allow the Local Auth mechanism to silently use the configured on the server side. +Thanks Dave +Linking the memory leak issue when using smooks +Added Hadoop as a dependency to the pom so we properly build test +When a cluster goes from being to we call process to make sure that we look for blocks which were not considered when there was only a single rack that should now be on separate racksThen the check for whether or not a block is on enough racks becomes is the desired replication only OR has there only ever been one rack configured in this cluster +Sortable criteria combobox not combobox now supports a new attribute sort which if present passes the value to gselects value + +Rebased for as v +If its a bug then we can log a point with the ICU team +I have the same problem +It needs a lot of tidying up though which Ill be doing tomorrow +normally we try to have that without any IDE because the IDE may influence the new archive under the same link has NB project structure but supposedly does not need NB ant build should only has source files under Caiss Modelsrc and Caiss Sagesrc You have a static import and on that static import class the compiler tries to get the return type of a method +Coding in text defeats that goal. +Clarifying wrt jndi names EE components would get these default resources in the names the spec recommend and these default resources would get the stuff managed in the threads subsystem using MSC deps. +In particular Those seem like internal packages to me I might be misunderstanding Isnt there a judgment call to be made about what can be exported within a module +Thanks for reviewing. +I have just fired up the example from a trunk checkout thats a day or two old and taken a quick lookOn the analysis page Is it possible to have the information fields always appear in the same order? That order is up to you I just want it to be consistent +Ryan Please dont start on this issue yet Im having some other errors could be related but I have to investigate a bit further before you start processing +final member variables +I did not decribe it more precise I just create Drools project and several mouse click in several places +Removing one more tab from and committing +code committed need to update wiki as well +Could you check that +HiIvanI will try to split this test case into two just as you suggested +I removed the wasted in For +The whole point of the prototyping effort is to identify exactly where code has to be changed to make it work well with native threads +Im doingthe to make sure that my changesoverlays do thetrick +Having the same issues shrinks view logic is reversedEach time i open and close the keyboard the shrunken view shrinks by another factor of the height of the accessory bar +Chapter there is no graphical installer for available now I tested this chapter with from +Juca please verify this fix on the CR docs bundle and close the issue +In SVN +bq + +Marking as rejected because Im not able to reproduce the problem. +Its unfortunate that if you serialize an Avro schema using JSON encoder and schema for schema it wont match the standard JSON version of the schema +Patch created against the release +Example program that can reproduce the patch mentioned in bug description. +I forgot to ask one thing hopefully this is not required to add in cli command list +Attachment has been added with description xml data to be has been added with description stylesheet to be transformed rendered by has been added with description output with missing last transition resolvedfixed bugs to closedfixed +Manually tested +Dave do you still see this error happening +Thanks for this info Brian +It Works fineIt contains junit testSubstitute with Comment loggers! And add guava jar for substitute the references with appropriate implementations +Thanks +Is this the official new maven plugin? Is there any news posted that tells people to use this one instead of the existingold +The job info files are not being removed from the local FS after +JDBC spec hints that the length specified in set should be equal to the length of the stream +Pushing out to R to remind us to revisit this later +Is there anyone opposed to having the patch here committed only to and having the more comprehensive fix of to? The other option is to drop this jira and have also committed to +Leaving this issues unresolved just for the sake of verification after goes in +I applied the patch but then noticed that the unit test isnt being run its got the wrong package on it and so applies outside of srctestcore +ScottPlease can you rebase the patch? After checking in this patch no longer applies +No the getter is not invoked hm +Transitioning to waiting for from pctony via IRCJukka anything we can add to this please +My guess is that this issue is related to the same JDK reopen if the issue still occurs on. +Correction I meant +Attaching incremental patches +The root cause of the issue is while merging schemas the code recursively merges subschemas if a field is a tuple or a bag +I just committed and resolved +rule should be numeric value now I can insert letters etc +Postponed +The patch was made against the latest SVN from ThanksCheers Chris to trunk +Id also suggest to remove Environment Win flag as this issue is not but exists on all platforms +Looks like the command line client cant be used to create KS for now +Have we tried the same VM reset plug pulling with +Is the change to supposed to be part of this diff? There is a possibility that your temporary local variables could collide with field names +Thanks in advance +Patch available +declaring the issue solved as per last comments copypaste bug was corrected in revision +See if patch applies under issues raised by Michael Michaels concerns and runs in my environment +What if I want the default is lib? You will using lib as? Is that correct? What are you thinking about? +OK the attached patch v includes both sets of improvements +Sounds like a missing feature +Marco the patch for failed Im sorry +Edited issue to specify that this should be fixed for the next release candidate + +It needs to be at this level I think to access the above +on checked in +The begin insert complete insert stuff doesnt actually get used at all for in this case the memstoreTS +Yes Kasper +I made some corrections and additions in this version +I guess using it under srctestjava is fine right +Since this issue was added the idea plugin has been completely rewritten and significant changes made to the eclipse plugin +why has this fix not been incorporated into version already +With the changes in the patch this is no longer the case +Id suggest committing to as well just in case we do another release. +Probably the project has not been imported correctly +The CP framework can fail to open a region if required cannot be loaded and initialized +Verified in Beta +The complains and you have to turn it off but thats all that happenedThe type has to be set to pint before you add the document +Stephen apparently the version youve advised works fine! At least those two issues I complained about are gone +See for further details. +I think it should be as used on WFK this issue you created a new archetype with version so what will be the last final version +Hi Eli sorry for the delay for getting back to this issue +To clarify next version being ESB not a more comprehensive contract model in ESB that would support both and would be very useful +Integrated to and trunkThanks for the review Lars +Does it boot ok if you comment out the yamler syck line in all affected libraries? Syck really needs to die + for the change +Please reschedule if necessary +you could have a very large number of small documents +Hi AnneSorry we do not hand out Windows for testing purposes +Let me give a try +some more details attached for analys if that also helps also mande change to use onclose and track statement connection mode but didnt helped much +I know nothing of Drupal but ought to know most about Mahout and recommenders together +Attach a patch and remove one empty line from the old patch +Updated screen shot which addresses Todds comments +Looks like your is what we need +Attached is the patch +Sounds interesting +Well also revisit locations resolution at that let me add that I can do this by forcing deploying to MUST supply a system properties which is working fine becuase if its missing the webapp will failed to startup +I also think it is clearer but Ill add a comment +eh Denny didnt you get a esb icon just last week +Now it works +How about bundling patch here into patch for +Thank you Michael +What I mean is that when we deploy it to our server via WTP mechanism the ear file contains the provided libraries +Furthermore integrated JDTs type filters to filter content assist proposals in the Beans XML editor +I with and without my patch and it passedI suggest we commit this patch to +Another solution is to get all the splits for multiple ranges of the same node in a single request instead of multiple requests and each request per a rangeI am holding off the implementing this feature unless we want it later +issue is still present in +Or is providing an iterator that the external system writes to? A more complete example would help clear this up +Thanks Aaron and Suresh! Marking as resolved fixed since changes have gone in. +Thanks +Committed change to class in to add additional methods to this mock class to match changes in core and updates to core committed and updates to committed changes to core assembly and committed in example the JMS binding now supports native async +The query needs to rely on the analyzer used but Lucene analyzers do not generally implement hashcode equals +I have to agree with Kevin as well on this +Prasad Its possible that people who ran into problems were before were using a version of older than +What do you meanAll you need to do is to create that implements and then adding any Ajax behavior to your component will automatically do what you want +this patch is excellent we will finally get rid of those annoying exceptions +This is simply a coincidence and is statistically more likely to happen under another note I always recommend starting with cacheLevel for jms consumers because this will show far less CPU usage and provide better performance +When you use selectorAware option the message will be sent to the queue only if theres a consumer there that matches the need to document this better + fixes compilation and disables bbp pass in emconf files +Hi BenOn examination similar changes should be made to ands a patch that applies Bens changes to and in addition to +Quick update Ill be sending a pull request for this tomorrow. +Yeah +Source and +If you just commenting to comment that does not help but actually take away from the codes readability +Got the from the model and set in the dispatch requires that we send a and does not put any requirements on the content +I see what you mean now yeah our options in this case are to either not perform the optimization or to push the operator chain above the sample loader and sample its outputs instead of its inputsBut on the flatten thing it occurs to me that we actually shouldnt allow the merge join here as we cant guarantee sorted order after a flatten +Hi I would like to say that at least the container behaviour should be consistent +update patch for humanres component +So assign to Pranav to fix UI first +Closing issue. +We shouldnt however limit ourselves to only two opinions +Hi James can you please attach a patch forIn Eclipse Team Create patch +The class defines several assignment operators but only one copy assignment +Anyone able to test +The default Runtime isnt named as EAP in th e build from +When are packed ints in their own file today +is an issue to enable the jvms jmx mgmt bean so can invoke full gc from outside jvm so we can do Todds suggestion above from the script added byNicolas Want to open new issue to implement your fancyIll resolve this issue as near dupe of +contextzip +Updated title of issue and replaced the patch with a right one +Minor nit add deprecated to isDir doc comment so that it shows up more clearly in generated Javadoc +Fixed issue here after upgrading from to all of my unit tests are failing with this issue +Oh yes youre right +Ill have to look at the Hadoop source code +So no Collecting was in effecttrunk avg number of terms avg seeks best time ms worst time ms avg mspatch avg number of terms avg seeks best time ms worst time ms avg msYou see the effect of the caching +As everything system test related goes to under system test runtime classpath needs to point there as well +But still Im more comfortable w the simplicity of the approach +bulk close for the release +Once I have feedback on previous comment i will close this jira +Added DROP PARTITION to need to show partition after dropping partition to make sure dropping partition succeeded +Ill close this jiraThank youMirek +It was confirmed that the workaround fixed this issue a more proper fix needs to be done as part of an refactoring of the Sentence Detector +Well renamed files were missed from the patch +Lance IMHO we design for of the users not +Ahh I hadnt noticed the road map +If the fetch plan changes discard the statement and regenerate it during the next execution +Yes I am logged in as WIKI is deprecated and EOL +We could assume values and fall back to if we encounter an exception but I worry that were getting a performance hit here for every field +There are a host of maven related issues related to this and its not clear if we are doing something wrong of if its maven +Vladimir and Alexey could someone of you please close this issue +Let me get familiar with your comments in and Ill ping you if I have any doubts or other ideas +I am adding as we speakYour timedqueue looks a little odd though +Even now there is lot of apprehension over how neutral state can be an alternative to various fencing techniques +Are you able to provide a nutchgora patch as Markus suggested? If you are unable to provide this then maybe someone else can get round to it at a later stage +This has played out over time and were fine +Thats why the return value is a List +i wasnt able to reproduce the original reported problem +Thanks +Almost all instanceprocess deletion queries have been +Well I gave it my best shot but no luck +However being a good OSS citizen I reported the bugFeel free to lower the priority as I guess it isnt major +But it sounds like Remko isseeing otherwise +We never have this issue within our company +bq +Schedule for an alternative to the original Lucene which does not make use of rewriting to basic sFixed at revision +was able to reproduce this on but not on using xen andIn i can see the execute in sequence is false for stopvm +Angie this is open for some time can you review this issue +and I agree on the nice to have as well +Hi JulienJust to let you know that we thought about this valuable feature +Will get closure on this for this might be a candidate for a plugin. + +It is against the branchYou know first start the bash script then patch the files +mvn clean install works fine when using Maven +Thanks for review Stack and RamPatch addressing the review comments +Finally Im convinced this is a bug. +I also did not have the ability to debug the server locally +I understand but unless you have a trace indicating that a thread somewhere is stuck in the its hard to see how our code is directly involved +Awesome Anand thanks +thank you Nadya +Now obviously the parser itself could be profiled and it could be determined that under a large number of nodes it starts to suffer +rubuntu smoke and +I know where the problem is +Thanks for the tip +generate the client code with wsdljava tool and use it to access the service +If that work is disconnected from this task then feel free to create a separate subtask for it as appropriate +Other issue dealing with Map +Fix versions +by downloading them +I will say that both at fine for me +Prabhat do you have any update? Its been weeks. +Used wizard New Other +nod Discussing that in a separate thread makes sense. +Having a simple to use Main class in is easier for Camel end users +The applied patch does not increase the total number of javac compiler warnings +Resolved +use groovylangScript are not supposed to run a +So the query means give me all nodes that dont have siblings or that are the first sibling with that nameDo we need to support those queries even if we dont support same name siblings +Any issues regarding this feature should be tracked as new issues. +However when trying to handle the reducers requests to fetch mapoutput files of those map tasks the http server complained file not found +This is a problem but too much ugly work to fix +Added the provided patch +Closed upon release of Hadoop. +or Ill just inspect the latest release first +Oh Im waiting for something like this too +Hum just tried this with loose bbox disabled I get no loose bbox enabled I get the two features you mention +TOKEN via or TOKEN via SCRAM etc +Thanks +See the linked bug +One could in theory specify that to develop a JDK is the minimum requirement +see object keep increasing but never get released per request hence the memory t create conduitObserver for each request we should create conduitobserver per endpoint +Thanks to Martin for pointing out my mistake so quickly +Committed +Dennis patch worked in trunk +Logged In YES useridmarking CLOSED +Patch applied and new tests added. +closed +In his case he wanted cp clientto treat the case where you have the Derby Network Client but not the JCCdriver as a successThis patch behaves that way with plain old cp but not with cp client +In particular getNodes doesnt seem easy to do in context +Didnt seem to fix it for me +It is with script +From it is workinglooks like there is workaround for my fix on somewhere elseBasicly the problem occured because the version was not specified in the pom and maven has defined in the pluginManagment +Embedded dependencies cant be easily changed +We can just leave the MIT headers for the entire component and add the Apache header before the MIT header in each file and then later on remove the MIT headers from the files we can +Tx +This will allow you to build a implementation of the interfaces for unit tests that dont run in +bq +I think the performance impact is minimum +I have tested it countless times on my mobile app +If it is the case I will try to move these two jars by Ryan revision in ER. +exception call rowsnull is the signal to stop looping over that inputsplit so thats not going to work very well in the general case +Just left them out +Test not included since it does not affect normal cases +You may take your time to review the branch +This was linked to to deal with but we will take a different route there +Seems like I got confused by the phrase from the docsIf both items to be compared are convertable to ints a numeric comparison is done otherwise a string comparison is done +Fixed in svn the english bundle has been renamed so that it will always be used as a dont think that renamed the s bundle is the best my point of view it is the contract of the site plugin to find the best bundle +Test shell script modified for output and additional jars for Axiom +Would these languages be defined as part of other ToolsJBDS developer? Anything is possible +I have reversed my decision about a Release Notes item for this issue +Patch for KT Driveradd testcase pass unittests and pepbut it has one problem +Could you verify if the NPE problem was fixed too? I still wasnt able to replicate the NPE previously just the You have access to no repositories +Are you checked this under JSP only +Hi +Itd be nice if the option was to specify a themeLocation instead that was just used for our custom theme header resources and itd still use the extension filter for the resources for the and +Checked in changes to allow c code to compile with clang +We recently resolved a concurrency bug and improved performance of HLL +The shard gets created incorrectly +looks like a packaging bug in which contains beservices took the lazy approach and just unzipped moved the file to the right place and rezipped it +defer to +Just for posterity this is really a dupe of +having said that relates to the same test and can also be removed +cleanup +Could the regression have caused what Anton saw if he was running from the branch +While Im all for a more feature rich rewriteregex engine after benchmarking I dont think its needed to fulfill the needs of this ticket +Suggestions appreciated +This patch fixes this issue and adds one more test +The long lines stopped viewing in jira erroneous call to getText when the argument would be null. +Lets discus in the new issue. +Move the actions to the Search menu +Does that id come back in the POST response? That should be quite doable either way +TIKA has own SPI parser +The new Scrunch version allows client code to determine if a REPL is actually running and includes methods for creating a jar from the code compiled from REPL input +I have since fixed this problem and I believe that it was specific to the PerlAPI and not +Next time please open a thread first on the Nutch user mailings list before opening an issue in JiraThanks +It means that this issue forced us to disable part of another test so when thi sis fixed we should go back and improve that test as well +You are more than welcome to email me if you have any questions about why I implemented something one way or another +by introducing would be beneficial in the long run +Thus removing the web console as being assume that the fix is sufficient closing this issue +I agree a long is enough even it the biggest number is only +Thanks for looking at it + +fixed this issue we just need to add more tests +Memory emulation system tests initial patch +Heres a patch that works for my codeFor some reason its not working for the simple example after I set preservetrue and make the adddelete buttons immediate +Thank you for the unified patch +Thats really bad +Here is the new patch which will fix most of the issues you pointed +Marking resolved as apparently this has been fixed already. +Fix the check that determines whether access times is enabled or not +So you dont need to build it locally +I guess not +Take a peak in the test package +the haszero method guard on Array breaks the codes +In Line of there are double semi colonsSince Tajo has hanging problem now I didnt try unit tests +Thats mandating protobuf RPC +However for filtering and security of requests is much better +bq +I think fix will be to compile testSerde but dont add it in classpath +Thanks +Stefan can you please attach whatever is required to be able to replay your problem +Committed to trunk. +Not setting a fixVersion as this was fixed some time ago +What Lars said +Rollback of blueprint changes were applied to trunk. +This patch works for but it wont be committed since no new features can be added toFor please use the same patch for trunkFor please use the patch in. +Please refer to the linked JIRA for analysis of the issue +Previously the sequencers were defined in the JSON configuration using an array of nested documents +Jun could you please merge v to the branch so that work on is unblocked +Patch for the is only schema for we want to run as part of the tests +Sounds good to me +Update to the latest axiom code +Hence if your component is capable of doing that then we may be able to pull it over soonerWe could then state on the wiki pages for it that at runtime it requires JDKWe will open a discussion on the camel dev list about when we should make the jump and go for JDK as minimum +Attaching and with the remaining changes to the Getting Started manual and an additional change to the build setup to prevent the table descriptions from appearing in the PDF outputM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srcgetstartM A libThe change to the file means that no desc element appears in the PDF output +That would degrade read performance while the repair is going on and lead to a large compaction when it finishes +And in the case described here the deletion of a task would mean a loss of information +Hi Ralf I have noted down the points +much delay heres a new take on this one +Needs additional cleanup +its required to use the termspostings apisThats why I marked this issue blocker +The attached patch adds final modifiers where possibleIt requires a minor change to the test codeThere are still some internal classes that are not yet these have been marked with TODO overlooked another test that needs access to the to rewrite Entities +Marking as resolved with the implementation of +Suggested Dennisquick update could you post this to +People looking to work around this right now should define their interceptor in spring. +Great thanks +Applied patch +Thank you very much for this great contribution Xuefu! I have committed this to trunk. +bq +I was aware of the change in logic but wasnt aware of any documentation of the behavior +Revised the release note to better reflect the changes in this issue and . +A patch for this fix against trunk at revision +Therefore feedback is welcome before I apply it +First stab at this have some test cases too if your interested? I appreciate its not complete as Im only supporting JMS which is fine for our use case any feedback on improvements wold be welcome +to merge into release +targeted for bug fix release +This is pretty standard across projects in the sense it addresses the issue of duplicate creation of resources and the issue of releasing the resources after the last use +It worked for me on your test projects so Ive got them all building in Eclipse +A few lines of change in also incorporaed the few lines of change in also incorporaed the with I moved this to project Hadoop Common since thats where the patch needs to be appliedIn the future please dont set the Reviewed flag unless the patch has been reviewed and approved by someone in the community +I cannot find the module javadoc within the repository homecvspublicon the host you provide me the correct url? +Even if we accept that that is a completely separate issue than the one being discussed +This is a problem on trunk as well +The archive is created on Red Hat Linux using tar command +Let us know +Ran out of time in beta so moving to beta +Just committed a patch that fixes the problem with the test case +I just had to check and again to make sure I wasnt missing anything +Vladimir if you think theres more cases needed here add them to our tests or to rubyspecs. +I believe this issue no longer exists +You can also configure it in MAVENOPTS like MAVENOPTSUTFAs far as I can tell its a problem under the hood +But the same app ran well under jboss and +deploysfine in a running instance of jboss as well +Thanks +Id venture to say lib as a single additional root +Andrea with your recent work on pushing D points through can this issue now be closed? If not I should probably mark it as wont fix as the patches are for the original postgis datastore +Is that sql something that designer is generating or part of a test script? If its the latter then the script just needs updated as it looks like it was never fully converted over from the older mmx style update procedures +Could someone the bug? More patches to comeThis patch fixes some more Javadoc warnings +bq +Sorry for a delayIt seems difficult to meI need some help +The changes in revision for incorporate a preliminary fix +Yes I think so +Attachment has been added with description what I has been added with description main source FO used to generate the PDF sources are generated with this unique source FO uses fox uses fox +quarter of and is unchanged until now +It is close to the one I was talking about above where we have two poms per module +This issue has been fixed and released as part of release +for more clear the list of portlets should be sort by Display Name while adding to a application list after get Applications for render +Distributed EnvironmentsWithin an application marked as distributable all requests that are part of a session must handled by one virtual machine at a timeEnd ExcerptThough not definitve certainly suggests the same session object for all the request threads +Done +As I said before is basically a speaks HTTP protocols its behavior should stay as close as the originalWhat about removing the checks from and create a separate jira to support the user pattern in HDFS? Im not worried about special characters since jersey takes care of all the escaping unescaping automatically +At one point during the R spec process we had debated around allowing a bundle to require itself +Revisit open issue links for details. +I have changed my local copies of and validator to fix this problem and I have also created a seperate file with all the static so that you can include the file instead of having to include the staticJavascript in every formWho do I send this to so that it can be evaluated by other users and possibly added to a next version of struts +Its weird it does that only for java files apparently +Still need more for the GA +I have a feeling well be building off of it for years to comeSolr The Cloud Edition +Ive had little success with the samplestests +If I have a hierarchy like thisTable Animalsid name cat dog parrot cockatooTable Petsid name cat dog cockatoowith Pets being a subclass of Animals +TedAfter there are some more changes +For singleton beans the overhead was negligible to start with +jt is supposed to mapped to mapred +We might just want to build in some round robin action or something +Those functions are only used in tests so correction isnt so critical +Have you tried the standard option to specify the directory into which the generated sources should go? If this is not what you mean you need to explain yourself a little more +Ill create another JIRA for this +Most likely you are running Eclipse using JRE and have JRE as default JVM in your Eclipse? Check Java Installd in Eclipse preferences and make sure you have JDK in there as default one +For this task is using in CXF test code still okay +To fix this problem completely Remoting needs to be used on the client as well as the server +Attach the patch +NET the console looks like this when completedInstalling +trunk Needs to be ported into +Thus the iteration order is and also different with a different JREChanging this to a normal list solves the problem +A small refactoring that you may want to apply on the doesnt solve the problem but I guess scalability problems will be identified when more containers are supported +Comments added in Review Board +Hi Nade please use the and provide the server logs +It renames the class to so youll probably need to do the associated file move manually before applying the patch +Which issue is this a duplicate ofI would like to watch itI would like to vote on it +This is fixed in the Panther JVM +Adding an additional XML configuration option to the Core docs might make it hard to understand +No need for other patchesDIGY +Ive added zoom actions to visual context menu +Upon digging in I discovered that is still in the bin directory and there is no +But FOP is no longer actively maintained and so any bugs there which arent in wont be fixed +The fix to this issue was applied to which was removed after starts to use blueprint to start the brokerCan any one confirm if we still need to track this issue +New patch version attached +Just wanted to share thoughts in which I saw some benefitsPlease let me know what you think +The fact that curl is failing but not hadoop fs hints towards your assumption that is curl +Quartz has been released +You mean the error message column alias name already exists +runs cleanly +My badI forgot to the conflict managerIt still loops forever +branch is safe +Therefore this is no longer an issue in and will not be fixed in . +Closing resolved issues for already released versions +bq +Hope that most users wont hit this issue in the st place but feel free to reopen this issue if it becomes a nuisance well sort something out +will just use +Cumulative patch for +Column is always TRUE or FALSE when used as a filter but may be TRUE FALSE or NULL when retrieved as a expressionAlso it appears that maybe values in a SARG test against stripe metadata can just be consider TRUE for the purpose of that test since you have to look at the stripe in that case +When close the wizard panel using default close button it close all frames including idea main frame +If one does exist it calls update service +Fixed minor issue and added unittest +The patch I used as quick fix +Another of this same change is that the extensions in tasks +The following ALTER TABLE statement worksALTER TABLE ADD COLUMN udomain int +Costin do you still need this in some form +Integrated to and trunkThanks for the reviews Jon and Sergey +is not a public class so a public method in another public class should not return it +This will make it harder to remove cruft +Implemented Junit test case for Payments received from customers for Sales InvoicesThanks to sumit for helpingThanks and RegardsAwdesh Parihar +But its true that outside of the debugger you get C or Java stacks +In that case what name to you prefer for the complement to ? Do you prefer or something else +Im just afraid Ill end up Ivy within my resolver to allow for all the various extra attributes revision patterns etc +Will this affect users of the plugin with regards to the way they configure itIf so then we need some documentation for this in the plugin siteIn version it worked this way but in version it works like this +I agree that the dependency on xalan is very irritating when wanting to use other xsl processors it seems to me looking at the code that the only dependency is on the class and this is only used in This might even be a longstanding error as a quick look at the code suggests that the standard could easily be used insteadXML Security has a dependency on XALAN though so is there any point in moving to in the WSSJ codeI think we can fix this given that support for JDK which doesnt contain has been dropped forMarking this as wont fix +I am newbie here and I would like to get familiar with code +Would have to at least be in different threads and avoid sharing any state beyond trivial stable data classes +I am running test suite on branch +Hi AlejandroWhere should the HDFS apt doc go +Weve introduced some changes there and the bug seems to be resolved +Rob could you try and do a ping on this one to see why noone commented on it yet +Reopening +Closed at user request. +So I added a resource method that does as you describe but with still annotated with +The policy mechanism is within the replicator +I have signed up as backup mentor for this project +Merged into trunk under revision r. +Patch to port java spellchecker stuff to +I cannot see an easy way to fix this. +Can you please take a look? Thanks. +Attaching a slight update topatch adds a note to translators for the new message not to translate Properties or in the message because they are Java clasnames +Thanks for the clarification +Rob might this be a candidate for Spring Modules +Thanks +NiceBut will a core test and its backwards version conflict? Or we use different subdir for backwards +Merged toCommitted revision . +The big introspector refactoring is in Issue +Even if it is not per restart every patch release will require a replacement of the sharelib and all the jobs that were running during that time will fail +closed for release +Patch looks good +But still surefire didnt create that confusion from what I see did by changing its behaviour isnt itI think only a small warning in the doc would prevent many users from being surprised +In the attached example I have eliminated the use of s eachchar function to get the same functionality +I would either move the Equivalent to material higher up or I would reword this paragraph to something like thisThe rowCount is a integer that specifies the number of rows to return +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Routable has been removed anyway now as part of message model persistence +. +That means exception handlers need to be implemented specifically for the context they are used in which isnt nice I think +Patch is provided against trunk +If we fix we really ought to add a buildbot for it so that we can know that it kinda worksJ +Sounds reasonable +team have concurred on the opportunity to release an additional artifact without the hamcrest dependency that is released well upload both artifacts +Event handling feature is somewhat like replacing the missing leg with multiple new ones +Oops I see that there is albility to specifiy autowire mode +Thanks for the effort Carl +Hi SebastianThanks for the patchLast time I used and patch for Nutch indexed all the multivalued meta tags +For example in order totest the OSGI support in Derby we may need to include some of theOpen Source OSGI mock classes that are used for OSGI componenttesting such as +Isnt this the same as +Are you usingI see there is a version in the Central repository but is the latest +We could put the unit test along with other tests under testorgjrubytest +for the patch +We could also rollup task failures in the JT against the TTGiven where we are in the product lifecycle for MR Id vote for +Well what the application server exposes here is a not the directly +Isaac can you help here to make sure we get the right links for the productIve filtered the original full list to only include doc related the action links still works right +The annotation will now look up the interface inheritance hierarchy if a is marked with Inherited +Please take a look at it and let me know if you see any issues or need more infoThanks Dims Im using Axis version not thanksm seeing the same thing with a client generated via wsdljava +Ive committed a similar fix +bug on CR +Have you checked out clean Geronimo sources and apply it? Did it finish properly? Im concerned as others have reported it worked fine +And can you please open yet anotehr new one for changing the directory structure in lib as discussed +Ive triggered a manual export to help it catch up hopefully it will behave from now on +Thus it may be useful for lists of key lengths and value lengthsYou provide some benchmarks showing the advantage of zero compression for random sequences +Thanks MamtaThe patch looks good to me +That interface would be specific to jrubyWhat do you think +Simply disallowing it at all may be too aggressive and cause more confusion +Attached VDB editor add model file +Branch has been merged +Close this JIRA as note that the JIRA is not about add operation but add command +Then itll only write the value to the output and it doesnt complain +If you specify a boot password then Derby generates an encryption key for you +if was old code +Thank Greg for the patch Lars for the review. +Looks like that this issue is fixed by Netty execs threads and metrics threads dont get cleaned up properlyNow theres no orphan processes after successful job termination. +I havent looked at the code but it is possible that tapestry uses this method when posting a form to a zone +Verified at cost two days to get it work before I read these. +Jozef how can this be ambiguous? There is one extension per extension class +Raising the priority +I think we must include this and others in! pom ivy and schema solr fail to patch current trunk +is this the same bug we had about certain settings not being regenerated properly +This seems fixed the race condition issue the nd mapping phase passed wo +Regarding the knob I think taking a number that is the size of the range would be fine +It seems to be due to a change in WS WSDL parsing +Ashutosh If store longdouble types as a varchar instead of storing it as a numeric type we can avoid evolving the schema when we add a type +Have the same problem building is almost certainly the problem yet again +thanks! +Thanks Ning. +Thx for the reviewI went ahead and committed my latest variation on the original patch with revision Leaving open so Knuts comments can be addressed in a patch +Whether I go the route or do something like the suggestion here Fix Version will be M +Marking these as fixed against rather than against. +Upon clicking the ok button the user has the option to cancel in case they are no longer interested or experiencing difficulties +do you think this might be the problemIll set the same time on both of them right now +Shahbaz can you please update the into the repository +Im quite used to commiting to as opposed to infinispans github +Based on latest branchRemark valign does not work with include this very usefull feature in the current trunc +attach my patch +Not solvable by now and the current implementation which relies on cross encoding and decoding seems to work anyway. +I have this almost complete +only assumes http in the createConnection method will use the create method +Do you have benchmark results showing an improvement in actual scan speedWhen I looked into scan performance with oprofile a few months back I found the same as you that a lot of time was spent in these calls +line +New patch which uses the JAR instead of a one +Alexei the patches have been applied properly you can close the issue +Attaching an updated patch for the security module and a patch for the file taken against the classlib trunk +Closed upon release of. +for the patch +As part of the fix for I create a Teiid Category that was a sub category of the db category + looks is addressed in +I noticed it as well +Trunk patch please +I just committed this after working with Hitesh to test on a secure cluster +Hi GuillaumeWriting to the same file in append mode is another use case that I wasnt explicitly thinking about but it is definitely a valid scenarioIn my use case I have a system that expects an incrementing counter embedded in the filename that works across the cluster +Changed component to dfs +patch against fixes off by one bug in substring +Works +This causes an error when creation is invoked +We need the same behavior for all languages whenever possible to have best interoperability between the implementationsUnit Tests should use and implement the same response for each test case thats the idea behind Yes I would like to see a patch for that bug +Committed to closed. +Dominique can you isolate the fix for just this issue from the commit? Otherwise I cant includethis in the Jackrabbit release as the change is IMHO too large for a patch release +As Gary said afaik this feature is in the ODE branch also not sure about if it is finished or not +Looks good +Code unit test case and doc have been checked in and updated. +I used the RPM built from +Ive applied the patches to a fresh copy from svn +if fsRunning is false then do not leave safemode just exit +It would be interesting if they have any useful input to Jena design +Attached is the +patch for approval for maintenance this patch instead +lets see how it evolvesbtw accessing properties independent of the parent is also neededfor the the reference stuff +But could we have some Javadoc and maybe even a test if that makes sense? I should also note that neither of these classes are exposed to and thus that has no impact on this commit +Pete has confirmed that section of the spec should say that if follows the standard +We could leave the old constructor and let it construct a new Configuration object and call the new constructor +Marking all current RESOLUTION LATER issues to be against the Nightly Build +Initial config committed to branch r with the following changes remove warinplace execution remove antrun config for removing struts jars resulting from warinplace add emma plugin config in build and the above fixes to trunk r and r. +Verified. +Hi David the repo looks good +Yes it makes sensesAs part of the Mavenization this would be trivialBecause of that wouldnt make sense to wait till Mavenization is in place for this +The test failures are likely due to will resubmit when the latter is resolved +ready for merge from trunk +JohnI was going to submit a patch to fixupdate these API docs but I noticed you were working on updating the list of available protocols in +It will be in the version released for +Modified to include the case that the list is null +Thank you for the help +It already does when the jobs priority is changed and so is already aware that a job can be in both queues and thus needs to be removed from both +If thats the case I will create a new issue +example of the approach attached as a maven it isnt obvious in the example a application of this approach would not use locks and conditions to wait for the async reply +Can you attach your test case please +Heins patches added +I have set it to false but maybe the tests need to be run twice? Once with and once without block cache +Can you provide a test case +v changes the state machine to use the same logic operations used in +Customization of the javacomp lookup needs to be implemented using or descriptors as described in section of the Java EE specification +The test is incorrect +Eligible for closing. +This will be very useful for the hive users +I think its a general issue with Eclipse or STS +To use it you specifiy sortexternal on the table tag then it bypasses any display tag sorting +I tested the same setup on windows with the same what is your config +Since we are using connection exception are handled in and retried automatically and we can also define other type of exception in with different retry policy if needed +for got to grant the license +For a quick fix you can just drop in a file into your classes directory under or if you do normal builds just make sure its put your own JAR file underclasses orgapachetapestryformThis way you can continue to use the original Tapestry builds without having to patch anything other than including a property file in your own build of your application. +So the limit value will need to be divided up between the splits including the remainders +on above extendingI like the idea of message back to master moaning if HRS cant open a regionserver fails to record split in META there is no other means of master finding daughter regions +If i use a list that is on value stack the same problem is not ve tried to reproduce the issue but unsuccessfully and I fully agree with Lukasz it is definitively a question for the user group unless you can provide something like a test able to reproduce the issueSo Im closing the issue in the meanwhile +So task is to update the subsystem to register with the appropriate flags +Thanks agate +I have started working on this release +But as a first pass generic instructions will suffice perhaps listing which distros theyve been tested on +Meantime Jimmy is figuring prevention of assign while a regionss logs are being replayedThanks E +While not understand all of the patch it looks correct so far +We cannot just trim the string because there are many other existing snippets in other palettes that use the same properties for insertion +Also ran Sort benchmark with on +After scaling back the number of threads of unit tests per host I have not hit it +You need to enter this issue in Bugzilla +Now they just have one type of token that needs tomatch one of the outstanding tokens +The callback when this happens is not executing the code that deletes the lock commitet but now needs dependency on for a constant it exposes +I misread one of your updates +This feature will simplify things for Solr users who are doing a query to get all the fields following by an add when they only want to update a very small number of fields +This change will require an upgrade script and an entry in the release notes +try it once more pleaseregards it works like a charm. +A for every exception would need to be in a separate since you can also chain those one thing Im not sure of is how walking would combine with walking +from use m jars for pars poms and m doesnt accept a distributionManagementstatus in pom files so I can parse file and remove this part from it or do we have an other method that doesnt check distributionManagementstatus +Patch which sets the initial delay to the same value as the patch with test to reproduce the for the patch +Thanks Tim the patch looks fine. +Also added to reduce the time before problematic merges are re still seeing issues in the failover tests but they are pointing more toward Cache issues so Im resolving this one +Using Codehaus its one higher than when using Java It will raise problems if our tests are too implementation specific +Though I do have some recollection that I might have fixed it already in HB +I think I closed this too early +Now tried a th time still got no entered sorry slurped by a mailbox filter. +should be unrelated +This should be completed please m using Eclipse Indigo with Tools plugin and there are only palettes with components. +at both the places +May be you need to do clean turns out that Groovy does not build cleanly with Java +The feature implementation itself seems fine but the tests could use a couple of changesThe actual message recieved should be verified to ensure the correct ordering as the message might be from the correct group but not be the expected message +Attached patch to display wizards in sorted order +Those are just warnings +Are you experiencing a particular issue +It will do a naive merge no matter what a CP might want to do possibly duplicating work +Patch for this issue +Todd just because there are questions about features does not mean the release is being used in production +Closing resolved issues. +Hiis there any progress on this? Any chance that this will be fixed inMoritz +Nice catch Bela thanks! Ive committed a fix for this +patch for review ensures all resources are closed using +Brett can you shed some light on why this issue was reopened? Are the patches solid and is the use case worth complicating the process further for the marginal gains it will provide? Which patches are even active does anyone know + +Julien could you just replace Unresolved issue statuses by Unresolved issues by status? tested +Seems like we have modified some configuration attributes and Studio does not know about itWorking on a fix atm +a static initialization block and a static setter for this both stand out as an to be avoided this will introduce problems in the long run where multiple tenants or the like require different settings perhaps even in tests we need different settingsI dont believe the answer is to continue the but rather to refactor the utils class to accept the settings as part of the interactionHitesh what is the scenario that you need to override the settings for +This patch is ready for review +Brian That could be since Im trying to do this after the fact after doing the release wo maven involved +However cql driver couldnt be started out of the box and I havent found documentation including code examples how to successfully setup a working JDBC connection as of now +Thanks Jacques +Ben let me know what you think and Ill update the patch accordingly +Well needanother fix before closing the issue. +Hi Jeff Hive was using the for the plan from the beginning +Another pass will be needed in order to remove all deprecated methods +Jacob and are updated by r +Yes I think so +I applied you regenerate other patches with new code in scm +unzip the patch to modulessql +Actually I dont see how this solves the original issue of eliminating IO +Cause with optionalfalse and and inheritance strategy we could have joined subclasses with discriminator values +Verified on rBoth the affected functional tests and the reproducer one passed as expectedThanks for the fix. +so is the problem resolved? Did you changeWhat Id like is for everyone who as experienced this bug using or better to report their OS and exact JDK +is a patch to servertrunk which replaces concurrent class usage with backport classes +Attached logs and assigned to this issue is thereits a test case issue Girish can you take this +All changes implemented as described in the task. +Rather than having a static IP for the name node we attempted to use DNS to move the name node from one node to another +You can already use set on so the machinery is already there +Output of this one test attached +The ZIP contains products which are included in the platform +I upgraded to and didnt check that setting +updated to address phunts comments +I was wanting to experiment with these new pubsub controls but for some reason all patches beyond patch set dont seem to do anything when attempting to apply them via tortoise svn +It should be called for clarity +Nice optimization +Heres a clean update to HDFS that enables obtaining the local file a block corresponds to +Added Ted as contributor and assigning +Could you give any detailed information? How you test it? whats result +There seem to be some few more hidden leaks +The controller should not generate a populator method for Embedded fields at all given there will not be a drop down box to be populated in the first place +I did and I was confused myself +good that you spotted it +Heh now I have to read this spec +Thanks for the details Slava +Anything else gets handled via a Parser fetched from the if required much as were doing for container formats like zip +Maybe it could even depend on it +Thanks Michael! +According to section Note HTTP servers are allowed to return responses which are not acceptable according to the accept headers sent in the request +Ok this is quite an uncommon usage so scheduling for passes fine against Git head +Just unbinding the server port isnt enough because that only affects the creation of new Connections +can you create a sample project to reproduce the problem +Committed to x under. +pass over all release notes prior to publication of. +It needs to correctly reproduce the signed request parametersI also removed the code that copied the query parameters from the original query from since those are already copied into the signed request and get treated by the code that reproduces the parameters from it +The manifest version is indeed incorrect in M +Changes look good to me. +Updated a patch to log the blacklist additionsremovals as INFO +Link to duplicate issue +patch using ready for code updated in line with some comments on a similar jira +Thanks Mahadev! +Robert and do not form a linear stream there is no way to reliably consolidate these two into a single file even if you flushThats the way things work in real life why do we suddenly start caring when we dump it to file +It turned out to be caused by advertisement blockers +adding specification patch removes member functions from the interface +According to Simon This is the case for all the widgets that contain images with dynamic width +Yes Max that is correct +Note This is a problem only in Eclipse you want I have fix this +How can you have a file that doesnt compile into a file named? Can you attach a sample project please? +I forget to comment one thing +As suggested marking as Cannot Reproduce +Patch to expose the lifetime bytes compacted +no one use msys to launch maven in windows +Resolving as fixed +I cant figure out how to delete it +I forgot to mention when I create a new Bestellung object then this new object references an existing object of the derived class Privatkunde +It seems it could be useful + +Thanks Patrick +I cannot reproduce this issue on Seam +Solr can create a separate Field instance for the stored value and so can other users of the module +Deployment synchronization issues with original fix +Removing as it can be included as a boolean property in LIKE BETWEEN IN and NULL objects also using NOT is not appropriate for string formation as NOT have to be optional and in between the string +It would be nice if this piece of code could be made a little more generic so that if a JSP container is not available then no attempt is made to wrap the containers assumed the dependency was valid for Seam +If enough of these outputs are queued up waiting to be merged it can cause the reducer to OOM during the shuffle phase +Yes the provided instructions works well to workaround the problemMany thanks. +Thanks +pending jenkins +Unless the idea is that this file will be updated immediately prior to a release in which case this should be documented in the fileThis is mainly to be able to tell one SNAPSHOT from another +Fixed on and I am correct I believe this is a divergence from how services are supposed to interpret filters +Are you saying there is currently no issue then + +I just realized that option is available only for wsprovide wsconsume uses vverbose guysYes we can get the messages from runtime + so the initial strategy were going to go with for this issue is Write a servlet filter to redirect to http or https depending on what is being requested +Includes and a unit test +Have a look at the unit and functional tests of the JAXB project where you can see how things should be used +I believe we could add an extension point that would allow additional pom modifications when converting dependencies depending on the dependency kind +batch transition to closed remaining resolved bugs +Thanks +The being exposed in the wild I want to be extremely careful about +reopened to make sure the comment is not lost in the noise +I agree with your complaints but this issue was more about MTQ queries at all and strange scoring +Thanks again for your input +At least I dont see why theres a meaningful distinction between setting quotas and say chownSetting quota is pure admin method but users could use chown for changing groups although they cannot change owner +Any chance you could us us a proper patch? +let me know if it will fix the issueThanks again for your great workJacopo +Yep once it gets resolved Ill retest it +This also needs to be fixed in trunkWith that change for the patch +OK the problem also occurs directly if building the master using maven ve just stumbled upon the same link and information Thank you very much anyhow! Ill give it a the mapping layer will be quite some work since were using the relevant artefacts through out all the plug inI attached my current progress but right now Im fighting against a class loader clash Im not quite sure where it is coming fromAfter fixing the current problem making the thing backward compatible to will be quite another challenge afterwards +I just want to make sure that the problem is not Windows only related. +This is as you would expect because we asked for a weakly referenced changewatcherIf you click Test which registers for events in a model layer deep you can see that change events fire when you click Change B and Change C +Leaving open because of the second issueGood catch thanks for this new patch +The default test run is too long even with the most time consuming test sets commented out by as they are by default +Oups artifacts from copy and pasting between two projects Sorry +When I made the last two release I used autoconf on centos +Recovery decompresses the message to get the starting offset +Im pretty sure that is part of the JRE spec so theres really no way that it cant be included +Found a bug in the tests attaching fixed are in capacity scheduler and streaming unrelated +whats the best to way avoid this failureMake the test assertion based +It refers to range queries not the equality operator but the issue might be the same +So the bug here was that there were tags that dislayed the label from body content that also contained children +For I think we would still require non of the operations to fail +defer all issues to +having applied with deepaks patch +dont sweat the repro project if this test demonstrates it thats all Ill need +Not sure is it easy for you to try this example against mysql database and let us know if this is only h specific or not +Can we not use the? I think this should be consistent across store implementations we should also have consistent logging within the API for handling the case where mapping file is equal to nullThanks for taking this on. +. +Christian Justin this is up your alley +There is a way of sharing connections with traditional Ado. +Ive modified the action invocation to separate the storing of explicit results returned and normal ones so I think now both cases will work +mmm are you speaking about? yes maybe we remove some +Added the deployment to server install the hsqldb database driver library. +The fix posted today should clear this issue up and likely a other issues that havent been detected or reported yet +Looks fine +It will be very helpful to know when the fix is comingThanks for your help +Ive tried applying your patch to Solr and it is working well so far +The problem doesnt exist on or Solaris +This facility was broken for longest time but I believe it may have been fixed recently +This request is alone the same line as and few others as well as a lengthy discussion on the mailing list +This is a regular Hudson plugin so you could at improving it +I am sorry as they are incomplete +Does this affect? The most common use case for the postOpen hook is performing some initialization upon region open calling this multiple time could be bad +added argument to the on as closed as resolved and released +please apply thanks +The issue is more basic than transactions though the integration code attempts to access these variables outside of a jBPM code in question was intended to do what you suggest return the actual values to the client but the way in which it is doing this is incorrect +This patch appeared to be within inches of being incorporated years ago and this feature is still sorely missing from Hadoop +How much more explicit can I be +This looks good +Also I as part of fixing that +Apologies if it is incorrectly formatted this is the first time I have attempted to submit open source patches for git +Excellent work +I probably wont be able to work on it in the coming month or so +AddedRemoved catalog select when there are no catalogs available +No checks nothing +I do hope this issue resolved on this version +Or rather is not necessary like I originally thought. +Its been documented and the Maven dependencies are correct regarding SLFJ +No rush. +If it doesnt work can you use JOPR and confirm datasource is validAlso I just opened JBDS and did your steps and got results back +Closing all resolved tickets from or older +There are some edge cases with initialization that I am looking at before committing a test and resolving the issue +Committed to trunk with revision +bulk close of all resolved issues. +The Provider preference behaviour is referred to in linked issue +Committed to TRUNK +patch for the plugins for the patches from zpeedy leaving this issue open until all such dependencies are history +I will commit it soon +Attached project doesnt include and JSF in +See for the discussion on mTime aTime semantic +Logged In YES useridI am not accessing the non serializable attribute from am just invoking some other operation on the +My still had set to false +Ill leave you to itI just did a couple of fixes +This feature certainly is helpful as is but it might not always be practical in a distributed system to have a version ordering across updates as its likely to involve a single point of coordination if you need one +This was backported to and releases +This patch makes sure that all other threads are down when +Thanks a lot Mark +Thank Juergen +could the concurrent merge possibly break apps that add a doc and then expect to be able to find it immediately after? I suspect this is not the case just wondering based on some odd new behavior I am seeing +Thanks! +Please let us know if you need any help +I attached the final successfully runs in eclipse except The fixture test was successful with but giving syntax error withI think it is ready for commitYour suggestionscommentscritics are highly appreciated as alwaysThank U +I propose we implement this in the instead +Please find my version of are currently on +Concurrent requests will override the parameters field +Thanks looks good +SMTP and log to context. +Tests now have to run +It uses an original wsdl which contains both wsdlimport and xsdimport +It all depends on what groovy enums really are doesnt it? In java there are several synthesized methods and enum inherits but this does not necessarily apply to groovy enums +I applied to trunk and branch +Im not going to spend time on this +This particular issue is discussed in section of the functional spec attached to the master roles issue +Found error in my test case this problem does NOT occur with passivations disabled +Ive upgraded the wc of build is that the one you meant +The Library configuration has both boxes checked +Alexey please verify +this is your own individual work +Just cleanup. +Or if you get a chance please go ahead with it +Added the check for the null +couldnt be fixed in branch cause we want the db schema to remain the same in minor version tom +For the benefits optimize is not a good name +Snjezana It seems to be your Quick Fix for project examples wizard and your Junit Test +sounds like you have hit on the solution +I gave it a test and as expected IT WORKSIt also spits out debug messages in the eclipse console +Niels can you modify those tests to make them dump the renderedimage to png and the attach the resultsWould like to understand if we are just slightly off with antialiasing or fonts or if the result is fully attached +The only thing is that i had to change the get method in to public +Commons File Transaction has not had a commit in monthsI for sure would not go in production with this stuff that wasnt anyone around to keep supported +If you submit a patch I will apply Ill get into it this is a diff which adds the nullabletimespanhandler along with unit SVN thanks +and thanks! +So youre suggesting that the path would be AXISROOTDIRtoolstcpmonSounds fine with me +TedNice useful backport +This now appears to be caused by inclusion of the selector auth bundle +Its unfortunately a plugin contract change but I think well be ok since the schema shouldnt need changing +This could be related to long lived servers but my understanding of JMS is that temporary queues should by garbage collected when the connection is closed +Will submit for if the results are good locally and then commit after +See the document in trunk +If this continues to be a practical problem we can change from to and let kernel kill processes and make progress which should prevent the entire machine from hanging +Yes Randall +please apply +binnutch index webtable +its sync against branch +If you need this patch please apply it manually to the branch in trinidad. +Kristen is planning to finish it up +Note +let me know your thoughts. +The release note also seems to be talking about branches rather than releases +passed around +Thanks Uwe Ill add rhino to maven builds +Need to assess if this is still relevant before addressing +Attached a patch that addressed Devarajs comments +I confirm upload completed +Note this is not a User Documentation issue and must be changed in the code base +I quite did not understand the passing the cursors to procedures and functions +James is this config not in the struct? If so we should move this out to sometime. +Own your code share the pain of maintenance +Im closing this since there is still no need for this in the framework itself +I will look into why it doesnt work with Ruboto +this must be resolved hence creating new issue with major priority +yes it is +IMO this isnt a bug +I swear it was a pure accident +Fixed with addendum. +apppath doesnt support full url normally since you deploy in the webapp you are do you have a full sample of your issue +I like the ambition! Purely out of curiosity does Yahoo! see this feature as a blocker for the release? It would be nice to do a release before then for certain then +I have committed a fix and polished the logging a bittrunk +Hidid you try putting it in conf +Looks good to what do you thinkI guess we could reschedule that issue for Groovy as well then +Without this scripting udfs do not work in for both relative and absolute path +Moving to as this seems to be a core feature of jbossweb svn module can now be used by the AS build + I am not sure what is triggering eclipse classpath and release audit warningsI think you might have done something wrong +Probably not important short term +Committed to trunk +a string for a numeric field then I get a +To be released in RC as well as next week +The attached simple patch works +committed +missing tools +The getSplits method uses the reader to correct get N lines and perform the splits so getSplits will never return half of a line you can actually configure how many lines you want on each splitYes this is also a valid concern and I took care about it +It makes working on individual features faster and easier +Specifically phoneME runs testcases in the opposite order of other jvms we are running tests on +This is the standalone archive missing previously +META +Yup an oversight +I just forgot to remove the comment after doing the necessary researchYou are correct +I just committed this +Im nearly finished with the docs but Im waiting for the outcome of our latestdiscussion how we want to build blocks in +Patch to clean up the comment +Thanks Igor! Hope it isnt a problematic change for Rhino is itAnyway works for me now and seems to scale very well +data processor class path is not referenced to hadoopjars directory +If it works please close this issue as Wont fix +Sorry about that obviously worked for me last night since I had the file +Well this is strange Went through about of the failed list but they do not appear related to my TT status patch at all + +Also as currently configured the ant script to build without using Maven will fail when it tries to copy the now missing +Thats why I merged into which is located in +It seems that proposal is fine and patch is provided + +Maybe we could add a to the already existing SIMPLE and VERBOSE +It is easy to miss the fact that there is no license on a new file during the review +Furthermore some of the sanity checks and tests conclude that the presence of an edit log file with some txid X in the name therefore necessitates that txid X has actually been written to disk which is not necessarily the caseTo address this we conclude that +Thanks Ruwan for looking into this issue +Thanks +HelloId want to add more API hooks TSHTTPPOSTREMAPHOOK TSHTTPPREREMAPHOOK into state diagram based on Heitors great contributionTSHTTPPREREMAPHOOK is new name of TSHTTPREADREQUESTPREREMAPHOOK since +So I added the flushImmediately method to Flusher that allows you to immediately and safely cause a region to flush +Fixed +Verified by Mark +Patch for branch revision on trunk and on branchx +Patch for attached. +Patch appliedRegards +Right good catch + use getter and setter in to access blocks +The issue is opened to correct fix version +Attaching same simple fix as +this patch removes the automatic copying of JMS message string properties from request context to reply context +iOS can change the volume while playing but I implemented setVolume to match the Android implementation +Committed to trunk. +running applying the patch needs to be added to the directory trunktestlibs The jar needs to be removed from that same directory as well +I think its definitely a good idea to have unit tests for the forge tooling which test for schema validity but not sure this is necessary when users actually use the tooling +X clean install and mvn X clean install from the root directory and both comleted without error +Want to upload a patch here for the server scripts to fix this issue +Hmm i did clean the build on trunk! Anyway the second patch goes well on a clean build and my messed up builds +Minor update to the previous patch to add the new methods to in +For ZK CDI support go to their issue tracker +See is what I committed to branch and +Or maybe the arg parsing was borked in some way from the start +In theory it will create a little less garbage to use instead of an iterator and since addNested is called repeatedly that means less time overall since that garbage needs to be collected +Fixed at r +also added a test +Hmm +Hi PaoloI stumbled upon another issue see second zip fileIm not sure if it has the same cause like the other stuff but it is at least related +Hopefully that will work for you. +So our model ignores them +Perhapssomeone will tackle URI for our release +See +Thanks for adding a test case +Report back if you have the problem with this you dont have it with this configuration do the same test by adding each of the other you want to use individually +Thanks a lot Richard and Claus sorry for the inconvenience of the extra work +Possibly depends on priority in respective of other tasks required for CR +Maybe close since JSTL is the way to go +If the code is compiled with maven the problem is not present +Append a character to the progress printed viaI wish the also used instead of digging into that way wed sure that whatever gets exposed on is also available for the cmdRPC clients +The right patch +I made a slight change to the patch +click on some other composite servicereference +did the rs not recover when the master started +Committed in trunk and nutchgora +svn statM javaengineorgapachederbyimplsqlcatalogM javaengineorgapachederbyimpldbM javaengineorgapachederbyimplstorerawxactM javaengineorgapachederbyimplstorerawxactM javaengineorgapachederbyimplstorerawM javaengineorgapachederbyimplstorerawdataM javaengineorgapachederbyimplstorerawdataM javaengineorgapachederbyiapistoreaccessM javaengineorgapachederbyiapistorerawxactM javaengineorgapachederbyiapistorerawxactM javaengineorgapachederbyiapistorerawM javaengineorgapachederbyiapireferenceM javaengineorgapachederbycatalogM javaengineorgapachederbylocM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTeststestsstoreM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmaster +Java encourages Reuse and thats what Im doing rather than by bolting on JDBC code to make it do what I want when the package includes a class that does the job +I will try to commit a patch later +QA Passed +We dont currently have other uses for these methods +No changes required +Then anyone can easily setup a jenkins to run tests after release. +This means we will generate compliant with the Metadata specification while still understanding generated with previous versions of axis +Hi Owen I have filed sub tasks and initial patches will be attached later +This looks great +Use version. +Thanks Ricardo! + Will not contain new look and feel +Glad to hear thatI created that type of server +If mkdir doesnt fail on existence then p can be a +Done +Will attach updated patch to this ticket +Added it now in rThanks for spotting Ray. +It looks like Alejandro has started on that in +The problem are the embedded true type fonts +Zapped it +Then the first thing you should do is upgrade in case there is a bug in your version +And change the log file path to target directory so it could be cleaned up +So back to method +I would think a sanity validation should be fine on Master if thorough testing is conducted on feature branch +How we missed then +sonal do you have a stacktrace? Im having some difficulty reproducing the bug +Thats the subtle world of bootstrapping I guess +As we are almost ready with our own jSPF implementation I think we should prefer it over the use of an external daemonprocessCan I close this as wont fixPS The contribution is welcome anyway and maybe well include this in a future third party mailet repository if well ever have it +All done with new refactoring support +Someone managed to squeeze in some more variables in Put so the test failed fixed so that test pass +This patch supports both Provider and injection and makes the existing codebase a touch smaller by utilizing to find the correct object to injectproxyThis patch also contains the requisite tests that support the expected ve made a couple of modifications but essentially used your approach +Thanks for the report +It looks like this is the jar file thats produced by building itest +Thanks Karthick + Username letiemble +Double ugh +Thanks Dmitriy! I will make those changes +Previously i remember that it was minimum of the seq id that was used for naming the replayEdits +Sounds good to me +Will take a look +AmirI have a perl script that I use to generate data for pig testing +Thanks Yongqiang +I understand it better now +I have removed all of these dead links +was successful to view properties +Documented in the release notes but more information about this would be appreciated if you can provide a summary Marek Martin +No new test code changes to documentation only +Pulled the request in +so it is a similar issue as the one that was previously found with the +Just realized that the title should be missing getter not missing setter +That makes sense +I think there are a lot of opportunities to clean up the code but shouldnt hold this ticket up +Id be happy to take a look at what you are doing to see if I can help you rework your code to go through the APII know that the guys are aware of this we discussed it previously as I recall +Test case idkey can be used as an example +I cleaned this up earlier today +This better integrates jdiff into our release build process +So one of the updates will fail I think which is ok no? The database end up in the right state rightAnyway to reproduce I can for example update the serviceName concurrently? I suggested is just want to let you could reproduce the exception so just said it in a simple way in real environment the parameters will not exactly same the normal scenario always like thisThere are three hosts and all of those provide a same service but with different endpoint and then they need register their endpoint to juddi to let client to lookup and some time they will register in same time so will update the same service entry in database what we want is all endpoints could be registeredThe behaviour we expected is the server side can let all these register success and when throw these exception the database state is not right its in a stateI always reproduce this by using a script to start three server to register in same time it can be reproduce quicklyBy the way I find that when I save a service juddi just remove old one and save the new one is there a function that can let client do a update? that means if the client save a service with some endpoints the old endpoints of this service would not be removed now I just get the service from juddi then merge it to the new service and then save do I make a wrong understanding of this +You can workaround this by explicitly disabling it in the configbq +Ill close it out if we are all happy with the current resolution +Would that mean a date in the far future would not be persisted or simply not indexed +Jaikiran just passing this your way as you had the final piece for AS I believe this is ready to schedule and resolve +if someone has a large number of unique queries will it not take too long to commit +many filessrcblocksportalsamplesskins +in keeping with the way it is written in the rest of the release notes +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. + +This text seems to relate to the resolution as commented by Danesh on Aug rather than a workaroundNote that this issue has a release notes status of NoneIssue remains resolved as of ER +Attach +MINGW still defines WIN so that part seems to be fine +OK done +Would transactions help you? A speculative task can show up right after the map task decides to commit the transaction and youre in the same place +Heres the status of this JIRA issue No more DTM IDs available error message +Ariel could you please try the latest svn head +No subsequent have the same problem. +Test case showing the failures +In the old agent we had already integrated some support for this but the new agent can first download and then install a deployment package making it much easier by design to support this scenario +Please file a separate jira request for this enhancement. +PR for master merged in master branch +still downloading the jdk if you want to test the ant task it should just be a matter of updating the ant test to point to a that contains the ant tasks +Ive closed as dupe +BraveFoolish I certainly am one +Fixed in current could verify this issue was resolved with nightly build Thanks. +This is blocking compiles for users with clean repos +In my opinion the spec is pretty clear here all is sent to the server none sends nothing and is omitted +So when the metadata for is loaded the jdo implementation will have seen the metadata for and know it implements +Puts off having to spend time on why the Reader is null at particular juncturesShould go into +Code to introduce automatic usage of Commons DBCP for Java SE apps in a and setup when Commons DBCP is on the class pathThe default behavior is autobut users can force usage of DBCP or fail if it not available by setting dbcpor force the old behavior of using the by setting simpleThe setting can be used to provide Commons DBCP arguments like but the will use the following defaults if none are provided +It is difficult to keep track weather the symlink should be removed during the upgrade +just I cant reproduce it +Patch applied at r with testcase +It turns out that skips the URL handling tricks that we do +It works many thanks +Should be assigned to me after fix without resolving +Just to clarify an upload button will remain in the bundle list but without the start level and start checkbox to make it look nicerIf someone wants to the upload the new upload panel is the place to a first step I removed the start and the start level input fields from the bundles listA new reload button reloads the whole listA new button opens a new page for uploading several bundles in one go by using the jqery multifile this issue as it works and no feedback has been received we can open new bugs for further enhancements. +I also reproduced it on RHEL with both openjdk and oracle jdk have found the root cause and I am about to push a root cause is that the esb distro have multiple versions of xercesxalan under endorsed lib +Resolved +otherwise +merged to docvalues and realtime branch. +Maybe we could also detect IE and put a message to suggest users to use Firefox or another modern browser +I suggest exchanging so the search is from the inside out +This has become a problem recently because deprecated the use of the address in July +I once tried repeated optionnal but protoc wouldnt compileSo I think repeated should be good +Tested +So keeping the issue open not only create false expectations with users but also the future direction of the frameworkFurthermore JIRA is used by Click as an issue tracking system not as a wishlist +Another data point if I disable join flattening the result is correct + +thanks for the corrections +According to the comments should not accept using an undefined default servlet +Need to add Security component in issue type to improvement as nothing is broken +I think this would complementimprove performance for as well I will patch this in and test against that build to see what happens +Beyondthat the logs do not show enough informationWhat would be really useful next time would be thread dumps of masterand region serverIf that is really all that is in the logs except for the NSREssomething is wedgedIf the master never receives MSGREPORTCLOSE it is never going toreassign the region +Update hdfs script to add the two missing commands +This is a not a bug I had another problem with starting the task instance that caused the problem +Touches the following fileM javatestingorgapachederbyTestingfunctionTeststestsupgradeTests +To start with here is my user name smarru +I checked +In regards to using for the RT parallel arrays if they grow to become too large then SC could become a predominant cost +Patch v +modulessrcmainjavaorgapacheairavataxbayajythonrunner +My mistake +Hi BecJust shooting this one your way for triageCheers Eva reckon you have time to fit this into the guide too? Let me know if not and Ill get onto it +Thanks +We get all kinds of errors +This issue appears to still exist +Would you please try the patch +Attaching pcap dump of communicationIm not using James server + +So I have put a quick fix in template and the build problem goes +Eli Thomas would it be possible for one of you to take a look and see if its fine? This patch is helping a few of us develop patches from Mac +Translators are not currently updateable in JoprJON so there was no need to add it there. +After this shall I try to tackle the documentation so that we can release an actual +yes please review +Hi Kyungho I am sorry for the late reviewIndeed the issue is present in but not in and trunk +If you can produce a test case there is some chance that we can investigate without that it is impossible to know what is going on +It may be possible to some spring configuration logic but Im not sure how the examples from transfer to the possibility is to rely on module properties when we get a service instance we can inject properties declared in the corresponding +On the other hand Ill drop Tools from the first lines in the text area and call them Version Components +However code correctness is not a sufficient justification to fork the code when there is no real world bug reported as a result of these issues the classes are clearly marked as to lineage and that they are not to be made public +Integrate this test on hibernate apply to as well need to look at the impact +Thats no longer an issue +There are two patches one for geotools one for geoserver +Attaching sources for proxy and test class I behavior you are testing is no different with or without proxiesAs the documentation indicates however Teiid does not provide application transparent failover +The output from the ORB service says Using null which means the parameter isnt set when the ORB is created +This does not sound like something which the component should do but rather the responsibility of the application using the component +yeah i looked further and agree with you no longer an issue. +Attached corrected test case +Emmanuel suggests this wont go out with misunderstood each other this will be part of CR +Fix for the ve applied your patchThanks! +Way more testing esp +This got too annoying so I looked into it +The calendar is correct its just the labels are wrong +Supported in CDI +Heres a quick summary of things +Closing released issues. +Added it to reproduce the above problem +Any ideas how the API could look like +What failures are you seeing +Verified on EAP ER +Should be fixed at revision main parts are in place now +Need disable magicquotesgpc directive in in PROD +sorry to jump in late +First this option would imply that all UI scaffolding support it out of the box and second it would require some sort of new field level Roo annotation which is not available right now +This has become critical if we are to have webservice that do more than just pass OM Elements +Could they be shrunk down from a pt font to a pt font? That change may make the serif body text look more as resolved. +Done +The workaround for Java users is to type the following at the Roo prompt at any time after creating the projectadd dependency groupId artifactId version confirm that this workaround works as expected in my environment +This will also log when a node is updated but the binary didnt change +We wont be able to do testing in this state because data cannot be pulled from existing based clusters +The compiler result is baffling +It is possible this report can be closed if the patches for are committed +Currently it appears Artix relies on the maxRedeliveries setting of the broker to avoid the same message an infinite number of times +This particular argument is not a valid arg for the shutdown method so it is causing an error and preventing clean should be able to specify startup and shutdown runtime arguments independently +Thanks! +coded to run on DFS + for the patch +Perhaps some sort of converter base class would be appropriate to help parse the HTML +In fact Id suggest that we remove the completely from the shutdown hook +Im attaching a preliminary patch +Also improved samples for step execution +What the hell are you waiting for? +In revision I reverted all changes from revision until we have a better solution +So its not a blocker +And is an interface. +So it has +Identical JVM +Moving these two classes to another package updates the imports correctly +Granted we dont really have a specific pathological test case anywhere for it so we wont really know till we get one or users start testing itBob Dionne said he was going to try and get a case that exhibits the behavior tomorrow so we can assert more forcefully that this does fix itAnd futon finished successfully +batch transition resolvedfixed bugs to closedfixed +Committed to branch and trunk +Apologies for the oversight +Im inclined to leave it that way since its the only way of testing in an environment similar to the target environmentThe build file already provides targets for building running the tests etc +lock file +Trivial patch +The patch does not apply because it is for the root build not common +After and perhaps a few followup changes new SASL auth methods will become much easier +Ran on the ubuntu hudson hosts which have older jdk causing compilation failure +patch created from geronimo root on windows xp +These are on two test boxes under trivial load so this is all still very tentative +Oops I stand correctedwe still dont have a logo +Its true Spring Security has a history of using filters but they can make it a but hard to follow the flow if things go beyond a relatively simple things are better off in filters authentication has to take place before the decision is made for example but once things get beyond a certain level Id say a controller is probably more flexible and easier to understand +applied in r +I can share a generic visitor thats somewhat more handy than default visitor pattern in javaEarwin I think we should make a new issue and get something like that implemented in there which is more general than what I just sketched out +Id like to commit it +I just committed this +Hmm that is not so easy to fix +Hum is it really needed to turn the geometry into an explicit envelope +Heres a patch that removes all of these deprecated methods +It is not obvious and has some overhead +The solution +Let me know what you think and of course how it works for you +By default this property would be null and we would just use IOC to populate the flowExecutor +Got it +Resolved for cxf now +BULK EDIT These issues are open to be picked up +just to be clear most of the issues above is probably not limited to OSX +patch disabling runtime enhancement and updating the build time enhancement to use the maven plugin for such +verified in JBT +if it could happen due to wrong configuration the client might be better off verify the configuration the first time it gets it +Updated the patch +I have the same problem with +I saw failures in ee test but theyre not related and will open jiras for them +at the two downloads and are replaced by and +I also pushed the updated parent pom to Nexus. +This works in some +Tentatively scheduling for the window +just remove the changelog cleanup +Only shade the needed geronimo mail bundle isnt enough we need some meta file to +If everything goes well I will check the fix in tomorrow +defer all X issues to +Patrick also thought that Clone would provide an opportunity to mutate the descriptionSee the last paragraph of the description for the interesting bits +until someone adds new stuff w the old label. +I have updated them to handle the +Patch as explained +Fixed will be in CR. +Theres no way to free direct buffers and everytime you initialize a server you allocate some which was happening in every test setup +Tagged for +Attached patch b where I rewrote the instructions for running the ant targetCommitted to trunk with revision +Patch was committed at the revision +This is definitely an improvement +Also the fix r invokes setParameters twice for error urls from get and after it +bulk defer to +Looks like a good fix to me +Function Metamodel retired in properties still need to be accounted for whether on the function metamodel andor the relational metamodel +In that case the position should be set to guarantee the top of the window in view +so is probably currentIs there any way you can get this pushed into Maven? Im not thrilled with dependencies on svn although I guess we could do a direct svn dependency if there was no other way +This fix has already been committed and in the release notes if this is still an issue I will create a new Jira task +Server Source Java and Messages +Should be fixed now! +Resetting the fix version to. +thx +Manually tested +Its not clear to me what would be causing in such a case +Note that while this is not available on the front page of the UI it is available within the information for each job +Hi Rishi Ive submitted the patch to the trunk in r +This wont impact the core API at all +Ive checked the release notes and I wasnt able to find anything that transation policy default changed +I may have the only app that will be broken by the backwards incompatibility but the change seems worth it +jdk used fixed +Attached service code and +I like itAttached a new logo form withoutgreen weed showing in the web page. +Letting the new patch go through Hudson +Some typos in I think +I believe Antonys work on for database names will solve this +ESB on AS ESB has been updated so that these operations use calls instead of remote calls over RMI +I dont know more about restful web or Brian any ideasIt is a little similar with need to create a new wizard and some facet +Rob is this issue really fixed? I dont see any commitments here +there is a work around in the meantime which is to set the correct using alter table +Reviewed implementation was not complete +Id love to help more but I need todo some digging to see where this examples app is headed and I want to lookinto the app +I prefer the string approach as well but id prefer it more if a point release of Axis didnt require me to do a ton of on my existing code. +Assigning to pending discussion +Tim thanks for the patch +Screenshots of the crash +its just summing across the ram of its packed ints structures +ProbablyCan you test with last resources plugin snapshot +Removed some unused code +Meanwhile Ill try to help out on some cli systemvm issues +Pushed into master +I attached the final patch for itRegardsCosmin +Not really at least with the existing heartbeat since that signals the node has restarted and the subsequent actions for that differ from a regular state change +Gave the patch a quick spin the unit test fails on my Linux platform you should try this out on a machine and see what other issues might exist +The problem happens in command line +Whether you have a memory leak or not depends on how fast you or applications and how intense is used by other web applications on the same tomcat +Does Velocity support the interface Iterable inWhere can I get some infos about the new version +Ack sorry should have read the patch first +Looks good to me. +is now out so closing all resolved issues +The composite element has a different name but same namespace +What difference would it make +I use this tag in JSP but somewhere with javascript submit again +This should not affect you normal processing of messages +That worked well enough for my needs +I think there were some languagestemplates that have in as reserved word etc +Ah thats right +Patch attached +I had opened another issue for this but closed as duplicate of this one to how to recreate listed in that jira. +The first time it runs it requires the user to accept the license agreement +The patch is precisely as noted in the prior comment +Im continuing work on this bug +unit Vladimir the updated patch was applied to LUNI module at rPlease check that the patch was applied as you expected +Will commit if tests pass +ETA has passed Murali we need to close the release +simplified patch based on changes in +Could you point to the jira created for this in ASEAP so we can track itThis issue says is fixed should we create a seperate issue for it in JBDS which only is an issue on Windows +This is related to +have to check the this one has been fixed on and trunk +Connected underlying functionality so that no intermediate input stream is needed +Perhaps we could tackle the startup situation now and open another ticket for making sure we dont try to stream incompatible sstables +I take a look over weekend monday! +weve had this discussion on the list +Thanks Francis +can you provide the stack traceand more importantly some sort of a test harness that would start jetty with this clustered session store you are using +I committed this +Please try it out. +Ill delete a couple comments that I ported to the forum +just committed this +Chukwa test cases were divided into categorieschunkinput toolsagentcollectordemuxdatabaseThere used to be test target for each components +The solution I suggest is much more flexible allowing a View to be assembled from a layout template and any number of nested templates andor keys +minor but much more readable and consistent with other parts of tools and especially consistent with majority of the file for names +Arun please consider merging this into +Otherwise make a new test +Maybe its some problem with merge +I just needed your result to reference it for the INFRA team +Hi ToshiroYou have a valid point to some extent +Currently the web console configuration support requires meta type information as from the service to draw the configuration GUI +The changed was committed by Hairong before +Please let me know if any changes are needed +Patch for extending the with the Annotation Ontology Ontology files for the UIMA module +After reinstalling another time it is working +I believe Brett committed a fix for the problem anyway since Hibernate working around it or not its still a bugCould you please copy the link to the Hibernate issue for future referenceThanks! +And they do it because asm needs to be as small and efficient as possibleAnd yes we do jarjar to avoid the jar equivalent of DLL Hell +New patch with findbug warning fixes has been uploadedThanks just committed this +applied in trunk has a policy for values for this policy +We got the new onCompletion DSL where you can do custom routing in eg when its failed onlyAnd thus there you can send the compensating transaction message +Note also that he got different results by changing the order the dependencies were declared in the dependent file which should not IMHO affect the result of the resolutionIn general I think that if a version of a dependency is available in the repository at the time of resolution than that version should be the used in the delivered resolved of the project being resolved +Submitting so precommit build is the change in maven config it should not require new or modified tests +The test failure is because in the current copy constructor of we set the parent field of all the child nodes to the new node +Ill open another jira and fix all those +The Mapping file can be generated automatically +But I really have no idea whats going on here especially since we bundle ANTLR into our lib dir meaning that shouldnt even be involved yet it obviously is +sorry for the noise looks like the the crypto api needs warming up before using problems goes away is a dummy hmac is done in construction +Sorry there is a typo on my comment above I meant the patches for and as what the attachments namesFixed in NIO module at repo revision Please check that this fully resolves the problem +Out of scope no much value in that +It looks pretty good +The main website isnt actually versioned with the release it can be changed any ll put this on my list for tomorrow and copy over what I did at Continuum +After all we are going to ship amajor version upgrade +which would allow people to use such bugfixes whenever available +Thanks TomI agree + +Let me know if you have more questions +Im in favor of keeping this as it isYou can just change the CSS for your own requirements +It should have all of the above fixes and hopefully a fixed jar too +If you have no evidence that there is a bug then you should ask on the user mailing list +I see + is now the default dialect as per and +Issue reopened to correct Release Notes field. +I also like the original graph image I may do a with it +Thanks for the question Tony Ive forgot to mention it explicitly +reopening to track this problem +User code in the AM then adjust the payload to instruct the tasks to downloan multiple partitions +I vote for difference +closing issue. +File DescriptionSpace Contains valid resourceApplication application Launches Jetty hosting a servlet to demo bugExample httplocalhostwadl generates WADL for Space resourcehttplocalhostnospace returns a textplain documenthttplocalhosthasspace returns a blank textxml document with a response patch for issue +Attached patch modifies the file to include how to run the demo using the network client +Mostly autogenerated POM have several typos in your configuration +I have tested and applied the patch provided by Graham +Unfortunately this workaround doesnt work its true that this error did seem to come on at the same time that I renamed one of the config files in the external Spring project but I have taken care that every config set in the application references it under the new name +Many thanks Musachy for your effortsThe code looks good and the results even betterPlease go ahead and apply your patches to and trunkAwesome work +has now been addressed and the updated Server Config Guide will be avaliable with the new CP +While this original was correct at the time it was created I would now prefer that it compacted the first files not just the first Maybe it could abort a if there are already items in +It use Mingliu Chinese data files contains both characters from PUA and non has been added with description Result PDF has been added with description Sample XML file contains both PUA and chineseI have uploaded XML data file XSL template file and result PDF file +This file contains the sample text lines included in the bug reportconsisting of repeating s and s +This patch lets Pig operate on jars and scripts that are hosted within a DFS such as Amazon S +Attached second patch removes another check making sample to fail +Added check for class and if not found assume a JRE or invalid JDK version +Ill see what I can do to refactor this +the rpcs are killed at the top ofhmm +I have attached a patch that fixes these exceptions +It works fine when they are files or parts of file names +By the way I am just a new user of ant looking for a solution +The only downside I can see is that a query might use up more processor resources on each node +First preview of a patch +trunk rev rev Shawn could you please check it works now and close the issue +This also fixes reporting +Thanks for the patchI have marked as deprecatedI updated wiki page with the new option as well. +removed the method changed visibility added final as requested fixed comment +But I also feel that when something comes from a specific schema which in these examples is SYSCSUTIL it seems like it should get collation of that schema +I made the changes as you instructed +so if the bug affected or would affect the correctness of fencing it should be resolved in latest release +Ive eliminated as much duplicate code as possible +Updated unit test to reproduce the extranous mappings to. + +Have another look please. +I was unable to reproduce it on my computerIve done some investigation about this Unable to get the current logfile size with stat error message and it seems its not related to the installer itself but to the Tanuki Wrapper that Apache DS is usingWeve already have issues with such an error message +Please be patient +In particular the no longer finishes +It needs more workId like to get this committed because its something we can use here but I dont have time to correct all of the errors + +It includes hcat server installation script test data generator test driver and initial tests for Pig Hadoop HiveThere are several known test failuresaborts Tests failed due to Rounding issue when benchmarking with RC file +I guess it is not solved then +Merged into SVN +svn diff for the label for the patch +by Denis. +Im guessing that it is placing the faces in some classpath that is visible by the admin consoles resource path +Despite that Ive updated DSW to support the relative endpoint addresses in combination with HTTP contexts which is pretty much the way things work in CXF when say a single is deployed but multiple jaxws or jaxrs endpoints with different relative addresses are addedFor example you can do HTTP Context serviceWS address barthe full address will then become httphostportservicebarSo if you ever hit the issue like this one just differentiate further with relative endpoint addresses +Patch after svn up +So I think we can resolve this issue since it it no longer a problem +Problem is that we presume cellblocks when we return mutations rather than checking what client supports as we do in other locations +The using is as follow +Nothing stopping commit of the initial code drop from my perspective +Going to resolve this issue unless there is further feedback on the correct exception to be thrown when accessing a method on a Session when the Session is good +If we could get rid of the split message and shutdown being passed by wed be able to drop and just rely completely on zk client ping +Agreed? +Other databases have no issue. +When split dns did not find default server use dns handler as backup +I put in a fix +You opened an issue on interning +It contains changes to and for the testsjdbcapijdbcapitoolsThe changes boil down to ensuring that appropriate mechanisms are used to accessing supported files and in the case of blobclobBLOB and lobStreams I ended up copying the supportfiles because they were used by other tests with another encoding +The patches depend on the local resolve patches + +reopening so we can have further discussion about a messageStore default by convention +Dont seem to belong to thiscategory but this was the closest I could find +If Valid annotation supports groups attribute API may have INCONSISTENT behavior because the validated group MUST BE PROPAGATED to its referenced objects annotated with Valid annotation +Please help me in this ASAP +I patch to make and work with +Only this time I get cannot be resolved +Just classifying bugs as Keywords bugs that pertain to specific API calls or advanced usage of Xalan +there any reason for keeping the stream open as long as possible +fixed in subversion repository as of r Concerning the exception it is not possible to check the matrix without trying to decompose it so providing an external check would be a waste as it would already do almost everything +parentheses got deleted while uploading last time +Once is resolved this should work correctly in console and CLIHowever if you want me to hide filesystem deployments we can look into thatReassigning to Brian for his comments +Here is the promised patch +hmm +Also once the Java Bean Creation wizard is finished we should the canvas instead of just plopping the java class on there +patch illustrating most recent comments on saving row position +If we extract these changes somewhere else this patch would lead to a code +In that way you ll be able to removing the eclipse framework comparing the cached stuff +I need to check in again and verify ifwhere its wrong but it was not an oversight it has to do with the way the Colt code does +It seems much less fragile than renaming the Range class but thats just my While Im thinking about it does Python have a range class too? Will we run into the same problem there? +As part of I want to unify the two +Each application should have its own service domain at this point so I dont think there should be a collision between these two apps even with the same namespace +When disk errors occur in HDFS jobs dont fail corrupt blocks get deleted and replaced with good ones so they dont get noticed as much +all set +What should happen if you specify parameters Camel cannot map and whatnot +The code is still not as the various nested classes are notHowever provided that the XML and other static Entities are not updated externally it seems the static methods ought to beThe only issue is whether the data items are published correctly to all threads +Theres also another subtle approach being lost here and in fact a rd option +Got it to work +Moving these to for now +Patch wasnt able to get TC to pass after migrating it to new testing framework +The ant based src tarball hive released are bad and should be completely ignored +Prepend for test resources and pages with application context are now passing the TCK +Merged into the master branch +exposed the setting in. +Does that help with your situation +Ill add null check and make the logsdocs more clear +Resolving. +Van I assume you mean when it is deployed directly not embedded correct +As there are only image libraries in the current code only image is support now in partuza +Fixed on revision . + +will as suggested +this has already been added in so we should be able to merge it across +Makes it hard to use jruby as a thrift client without resorting to native java thrift interface +Which is the expected behaviour +fixed as part of patch +No special reason I guess it was only residual from many upgrades that class had over time +This is intentional +With local DB the network latency is much lower and thus the impact on performance is not that big +Thats bad that there should be such limitations +Because this project is shared by riftsaw jbpm console I will need to find a way to get it configured properly. +Please resolve this issue if it is done and open a new issue if we do the same for EAP +It seems to make a bunch of calls to generate some mapping examples from your current location + +I too think this should filter and not new parser provides proper error messages. +thanks for the review Jimmy. +For some reason it is with not specified as dependencyCommitted in rev +I too am facing the same issue +Patch checked in +that will help us understand the condition of cluster function did an experiment with stable version following the way you setup the cluster and changed the relevant arguments and the results demonstrate that we can use the full cluster mode and connectioncollapsing feature together in version +set email encoding to for text mime new patch additionally sets the mail subject encoding to +I mean you are making it sound as if it is a marginal feature so that the priority should be trivial +Thanks +Ill await others comments to find out how they feel with the current structure +Thanks this will be in as its ok to be included after the RCI am going to backport to also +for the patch +We should backport to +svn directories +This seems to be fixed now +OK +Im not sure this code even exists in the code base anywhere so there isnt something to patch in a traditional sense +But I ll keep it with you if you think its a side effect only +Nevertheless I think I have it working now +This is not a new observation although I cant remember if we discussed it in a jira or not +This is a needed feature +Please change it if you think otherwise +Thanks MikeCommitted revision +These are automatically passed in for JPA EE container managed applications but not SeamFor a workaround could you try setting property to value seambookingsessionfactory +IMO its a bit out of scope for the deploy plugin to be removing things +The general jdbc mode should workSeems to be a proxy error +It doesnt appear to track dependencies between columnpermissions and the underlying columns +Since it is a bimap its possible that there are two cfs with the same ID but different names +Ben can you please have a look at this? dont think that last link was quite complete +Fixes double free errors I forgot to remove calls to delete +Currently it is simply inlined into the source which looks not very professional I think +But in this case the jetty is not a container but a way to transfert the war on the remoteAnd if you dont add this dependancy you force the project to declare into their pom and for maven its a worse practice to let the project to guess the transitive dependancy +The web connector would probably find the links but not the RSS connector +I think we should stick to what we currently do +This patch addresses by patch. +The problem is not really the batch but the followingWe use Equinox +Complete initial patch so that all junit tests cannot reproduce the junit error but from the output given by Emmanuel Venisse I saw that the initial repository where ll add a patch that makes that initial step more robust by forcing add and by always use absolute paths +Just incorporated the changes Keith checked in on the weekend +Is this really necessary? And if it is then why +Nahi I talked to Tom and we agreed the change to is fine for +I have been in contact with Stephane Dolivet who has created the LGPL licensed editor that Bunkai uses +Ok +Hairong can you please review this since you removed as a part of? Thanks + old JDBC datastores are unsupported out of use please reopen if the problem is still valid against the new closing all issues that have been in resolved state for months or more without any feedback or update +I will try HEAD when it comes back in process of setting up HEADI reproduced with updates + +Its also not clear to me that the issue can be fixed in a manner +There will not be a moving to for fixed this at the time he reported it. +I will add some descriptions to illustrate that change the default start level will not effect the server side +Hi VivekI dont think we intentionally drop stderr statement in +Not sure if such a scenario will arise and what problems it could create +Thanks Army those are good suggestionsIm intending to move ahead with final testing cleanup and commit +I know it is not caused by this patch but I think we should handle and the stop event in Application and not container +Lars did your deployment include an file or a file? Before that is you worked around the problem +Yes Im working on it now +Thanks +We just tested the patch and it works as expected +Please note that there are some new for the patch +This issue is closed now +I would love to have a matric type with Pig +bulk close of all resolved issues. +If you can it would still be great if you could provide some test setup + was a nightmare on Ubuntu and I gave up with it +The choice of eap as a value is a mistake which unfortunately we have to live with as the intent of that value is to identify whether the admin console is running on a community release or a product release +looks like well need to provide custom scm url entries to all modules because they dont match pom scm would think it better to move the samples to their correct dir would fix all problems regarding urls + +Your testcase seems to demonstrate that output from the java process that happens after termination of the test itself seems to create trouble and I can reproduce that with or without any threading involved + Username rodburgett +You say it is fixed but you say no in the next sentencePlease resolve with proper fix version if it is done thanks +Ian I applied your patch but I dont see much difference between the first and the second patch are you sure you uploaded the right ones? +Regards a test case +Changes made in for the conflicts due to update +So the sharp decrease problem appears +Attaching a first attempt at implementing thisIm running now +Thanks for explaining MichalLets make a trunk patch so that we can get a run +Neha go ahead and run with it +for the patch +Thilo please review +Even though the API has not changed it would be good to have the title on the page updated so users know the are looking at valid API doc for +Thanks. + +just so much easier +Other test cases can be written in the future for uncovered cases like multiple families using the foundations in this commit +Looked a bit into this +how to join usermailinghow to send my doublts +Here is patch fixing them +Thanks Rakesh will go ahead and commit +is not final. +added to the chemistry group +This is a roll over from previous EAP release +Im a fairly strong believer that adding the Jdbc classes was a mistake in the first place +Thanks Timothy +Added unit tests in runtime +A simple safe fixThis little patch has been available for a week +The OOME happens while copying jar files before compilation actually begins +I am just while I backport to +This is great +The following patch fixes it and enables lazy resolution on Linux x +Sad since the code was fully backward compatible and wasnt causing any performance lost if get wasnt ok +We use NAS drive mapped to the application folders where log files get generatedWe use and havent tried internal loggingPlease suggest next steps for monitoring or debugging. +Hi DagThanks for the revised spec +also for the othersaddadd +This is a language design issue or a Java backend implementation issue +I think deleting the files is a legitimate use of a finalizer +so I think you need find the old code about octmeber the system can work +Committed to trunk. +show the dump always show the dump only when a debug template is outputted do not show the dump at all +Thanks Liyin +Im not sure if theres a way to setup Mongo by default for everyone +What should we do if the parameters fail the sanity checkShould I create a new exception? If I do so this will have a strong impact on several parts of Hadoop +my pref is using a library for this +Found it a file is generated at each build and contains info about runtime configurations to use +please note that there is no the Sling webdav implementation but its in fact is just a thin wrapper around the one exposed byjackrabbit +That said allowing to change the workspace was discussed on the ml and agreed +Hope this helps +Assigning back to Dave for incorporation for next release +Ganglia +but theres more loose ends to tie up to make this a robust solution +for new patch +Thanks +Suppose one client writes the data +Ill test as soon as clustering is unbroken + +batch transition resolvedfixed bugs to closedfixed +It was synchronized there for a race condition that showed up in testing +that analysis is correct + +Hi Dan I fixed all these problems in core last Friday +This is a good catch Simon the file encoding does not get injected when the include is performed +Submit apatchRobert +Its a good simplification +looks like a bug I introduced w the mmap code +The per connection is more general I guess you only have to keep a flag around to avoid redoing the spatial checks every time a connection is created +rPasses smoke and on Ubuntu +findOne delee update +For a datanode that has K blocks it will take only about MB of memory to keep a list of blockids +the task id b +Currently IWs ctor code performs the initialization however Im thinking to move that code to an init method +This is the kind of thing Im talking about +Give these a my previous comment should have their descriptions be Server for Operations Network and Agent for Operations NetworkThe description of the services installed by the attached binaries are Agent for JON and Server for JON +I think this is too big to do in +Trying again the problem was that some java files were checked into SVN with windows endings and the svn was not set on the files +I hesitate to add implementation details or refs to javadocs in the shell output +Teig there is a simple way to bring me some data +Committed +Theres no yet +So if we get an estimate for what it will cost we can pool our resources to get it done in a timely fashion I imagine +Given streamed results there is no such timing as a purely internal cost. +Were auditing which secondary indexes that we really need because it affects write throughput that muchIn other words resolving this ticket would make it so secondary indexes were no longer a compromise wrt write performance + +Lets do it +Fixed with commit +Will apply if good by Andrew +Impacting me too cant load the EclipseSTS plugins due to this error +If you describe in detail what your interaction with this hypothetical API is that would go a long way towards clarifying the need +old issue +However that warning might be indicative of other problems +Brief summary We cannot provide snapshot isolation as a guarantee since doing so across regions is really impossible +I think I should add a test here to verify that +It is the differentiator with the similar command +Im providing a cleaner fix which can tolerate this +Assets are recreated in Flash +ivy folder has these filesI would add AL as a header on these files and resubmit the patch +I think there is an issue with but I havent isolated that yet +Yes Ive looked at this code before +This bug has been marked as a duplicate of +trunk +Id like to do so on Monday or Tuesday of next week so that should give plenty of time for further an in Chriss comments +The patch also changes exception swallowing behaviour to swallow Exception not Throwable +Even the class of the new tests in the patch is called +This problem seems to be resolved now +Perhaps you should try an JDK update and if that fails to help maybe provide a stress test example that demonstrates the problem +Do you think we could close this issue +I hope many of your concerns would be covered thereAt the same time I would like to know if you were using why did you need to use autoconf? The expectation was for the use to use +In response to Olivers question the latest version is on FUSE Source +Verified on docs stage for both EAP and EWP. +still needs to be fixed for server working with clientKathey +Here is the patch for the same +Its on the blocking list so dont worry about it missingHTHPaul Davis +I have subtask some closed jira since they are part of this efffortmostly for trackingIf they want to be I will use them +Ill open tomorrow different subtasks and update the description for this one +I committed this to +perhaps using the implementation details in the manifest +It seems fixed. +Why that particular orderingNo strong reason +bq +Spoke to Mark about reproducing this and he advised that its a dupe of +So its mettter of this resolver to recognize such +I set the assemblyDirectories for all the projects in the file +It includes demo code and an article +See the tests completeness and because you asked +Findbugs and javadoc warnings not related with this patch +I still dont understand this error I mean I have added safety mechanisms to avoid things like these +Thanks Mauro for the quick reply! +Oups +Sorry but I dont have that last line +Thus the smx bundle will probably need to be used +I guess we basically all agree on this report +Verified with JBT. +This was not an issue for will review during next release cycle. +This isue only ocures for Jboss tools I tried Eclipse Galileo and instaled Jboss tools Tools succesfuly from Development Updates but there I have other isues with Jboss Tools +Alexey PavelIf the target is achieved I believe we can close this issue and continue in the next JIRA. +mrepositoryorgapacheaxis and runmvn C clean that will not resolve the issue +The other bug is +Marking this jira resolved as it has a workaround and will be fixed on side +There is little point in supporting old because of the work which has made significant updates to the specifications +Updated patch with contributions from Werner +Obviously building on the committed version is the way to go +Id also like it to implement Closeable as thats generally handyIf consensus was to adopt the guava lib as theres no reason why their couldnt be used but Id still want a under this which would be our base class and have extra methods in there +So it seems that was only fixed on trunk and then i went and fixed it on +Unfortunately the fix didnt make it into the release you will need an SVN drop to verify it +It gets tricky to display the the JHS jobhistoryapp page though +Committed to branch and trunk +They appear in the sample in the proper directory +We are under high pressure to resolve this issue. +LuciaCan you elaborate a little more on your use case? I cant remember having a prompt for password at Eclipse startThanksXavier +And while here note that the doesnt get along so well if you try to have multiple shade executions since the name conflicts +Lets use. +sounds good done the changes were simple enough for me to just redo them after refactoring as such +Would you please try this patchBest can you help me to understand why we should do this +w r +bq +Im not sure what the npn directory and the classes inside it are as there is no documentation +shows the design framework and the checkpointing service is implemented using the class design +Fixed +I blame JSFFaceletsPortal libraries conflict for it and I will need probably more time to solve this +The issue most likely is an environment or a setup error is a build issue +Fixed on trunk +The top level build works minus the following three components are not building and are currently commented out catalog pushpull and there are four current deps that cannot be found using a clean mvn repojenajarojdbcjar xmlrpcxmlrpcjar apachetikajar +i think i fixed this back when i was at JPL and just never closed the issue +Since the other modes are pretty advanced i think we can more or less not advertise them for now until we get the modes sorted out +As a minimum could you at least attach your so i can see the cause of the s the for the integration tests exercising. +How would that shared be exposed to a? Via the or theI was thinking addition to interface or and implementation inContext is meant to hold state scoped to the invocation of a CP chain on a particular hook +So +Just noticed that something similar used to Eclipse in Tools +Testcase im speaking about +Applied a more proper fix to trunk and branchOleg +sh files over to your HADOOPCOMMONHOMEbin directory they should work +It makes more sense to have everything in one place as you suggest +Great if youre thinking of fixing it Martin the and also use parseInt +As a hive user affected by this bug I would have no problem with truncating the variable to KB +I wasnt clear in the description +In other words someone else has also donated an implementation of the Axiomatic Retr +rebased to latest trunk remove unused and setClass in improve new for backward compatibility +Cheers issue can be closed because it is fixed by issue +Agree More investigation is needed as to why compactions cannot keep up under high write load +Fixed in r +Marking these as fixed against rather than against. +Test case files pt +thanks to Dimuthu for his contributions. +Patch applied in SVN rev and to trunk and branch. +Can you let us know if it doesnt work for you +Anyone wants to post the makefileProj changes as wellIve done those and still cant compile or link the debug target +you need to delete buildredserverdist after that all related will be recompiled patch attachedThis sets Xmsm preventing VM initialization errors +I am not sure if I will push this change to be available in STS but it will definitely be available for final +Patch looks good Uwe thanks +So it seems that the only solution may be changing the groupId or the artifactId +I do not think we do not need to add any extra complexity than it already is +that works against TRUNK for +SeanI just finished manually testing the fix +Absolutely no reason why we cant have a singleton instance available in the codebase that everything defers to + +When we complete I can think of a useful enhancement where we can dump if some deprecated keys are being defined in the configuration +Thanks Dave +Not sure about the use case for that but it would allow for it should it be limitation that I think we will have is that classes in the project cant be added to the classpath as this mojo executes in the phase and the projects classes havent been compiled yet. +Also adds a warning to the set javadoc to make its behavior more clear +There seems to be a linking problemgcc worked without problems +OK looks like what was happening was that asynchbase went from using dependency ranges to specific dependency versions +Ok here is my project with maven +Bump! Shoud we close +Verified on EAP ERorgapachehttpcomponentsproject contains expected parent +I did not review but did try out building with the patch and ran a single test +is at the very core of jackrabbit +And from those discussions in the past it seems that the community always recommended to users to close the connection explicitly in each web service all +Maybe you can rebuild natvies on Harmony since there are modified native codes included +In other words replace your tt with t and t and you get the same behavior +Patch applied +Closing +My Regex way of testing for valid XML didnt work in all cases +Fixed for both NHTTP and PT transports + +Sorry I completely missed thisOn IBM JDK the issue is fixed +Updated patch fixing checkstyle issue and changing the data file concatenation test to use the current time as the random seed +Move listFiles to be a method of Util +What is really being requested here is an adaptive more advanced pool management capability create headroom during loaded periods but then quiesce when load abates +Replied on phabricator for the comment +This stops up creating asdocs to include in the release +Verified on. +Hopefully itll work on Hudson too +Wether your patch is the best or not it is a hint in the right direction +Committed to tez branch. +check if any failed test cases are related to the investigating the failures I realized that I introduced a bug with the parent id of workflows when redoing the purge logic for +I would love to see something EMR style api that abstracts away some of the exact details of the cluster being used specifically the mapred jobtrcaker being used +As these are deprecated I removed current OBR implementation actually support those optional flags when using the internal bindex to generate the OBR constraintsThe information about the provided service is actually really important as service constraints are supported by felix obr +That will make things easier to reproduce +This fix does the following +I will try to be more preciseIn fact when I am writting a new job using a file I am able to define a custom jar using the property +I see what youre saying and although its still a slightly confusing mechanism documentation can address this +It uses Platforms Support Case Workflow v scheme which defines some special workflow for Support Patch issue type and common jira workflow for other issue types +Using a in the interceptor handler is definitely wrong +Sounds good to to me like this is already handled by the cdata attribute of the javascript tag +Thanks we should commit this commit Committed revision +It seemed ridiculous to turn longs into strings only to immediately turn them back into longs all over the placeThe patch passes all of the tests for me +Alejandro makes sense +If the issue is that the prompt facet is still visible and possible to click when the showDetail is undisclosed then perhaps the fix needs to be specifically for the s prompt facetno other facets or children +We can certainly discuss whether it makes sense to block always or once in a while but that seems like another discussion +In most cases where maxVersions of a table is set to a small value it can alternatively skip at most maxVersions keys +Oh hey did you set your and stuff that is not welldocd? Does wordcount or pi work on your YARN cluster +Hitting that limit will cause the JVM to exit with the same exception +It throws a in binringmaster if no tarball is specified because we are assuming it will be available in the patched code +The attached contains a couple queues that can be used to demonstrate this issue +That way the executable wont link if the library and headers are from different configurations. +Comma is special in + +Patch was submitted +Previous notes belong to +Thank you +Does this mean there will not be a? Since its listed in JIRA I had assumed there would be +Fix for causes this issue +In which case Ill leave it in +Tested with a Cordova client running in the iOS sim receiving push messages from an running inThanks for the fix. +I can prepare an updated one if you like +Revised patch file with no empty files and the bmpm package renamed to bm +Hm this looks like something I ran across when I replaced the thread pool impl +Since this is a contrib you dont necessarily need to deprecate so you could alo drop them from or even from for reviewing the patch SimonThe patch was applied on revision Adriano the is the dev branch for minor release like +If you find further issues please create a new jira and link it to this one. +I am in favor of replacing the existing implementation with the alternative one based on code +Heres the test case +Sent from my iPhone +We have huge old project +And I am not trying to check if the connection is alive +Btw for those who upgrade from Maven to they see broken links too +Let me fix it ASAP +Closed after release +The module should be seen as an option to developer in case he wants logic in his application +We can do that but then we have to change the Root interface whats exactly what I tried to avoid +Its rather dissapointing that so many people wanted this but I had to go and add the issue myself +Closing. +To avoid the timeout error the iterative method was used for mult as a each row of matrix A matrix B +It works now +However I do think this can be committed to at least have query time join support in x +In the recovering phase currently MR is only looking for history files in staging dir maybe we can also look for that in the doneintermediate dir in case the files are already copied over +The goal of this defect will be to enable all of the tests and make sure it worksThere seems to be a number of problems +The next iteration of the patch does a manual syscall so it works on RHEL now as well though it seems to have actually harmed terasort performance by or so in its current formI fully support someone who has a system doing the experiments with fdatasync though +Hmm +Any comment on why this happensIt still happens in I just ran my testcase in and it seems okMy main application must have another issue one of my bean references is coming out null +Is there any chance you can build the latest source code and test this? Your assistance is much appreciatedThanksGreg +Moving to M as this will probably involve some interesting mockingtrickery plus the existing tests cover most m closing this since all we would be testing is the session handling of the browser and not the functionality of the library +Trying an image instead +Still interested to hear if the issue persists +Hey Colin it looks like the patch you posted is against trunk and not +Can a committer apply this patch? Is there a specific person that oversees c patchesAll issues with the patch have been addressed and the change is rather trivial +Proposed SOA TP Release NoteThe FTP store method now throws an exception instead of returning a null output stream +The Problem still exists in +We should keep this jira open +But this list is populated in compute as default not doing any fancy things +Talked with Namit offline +Not necessarily the correct implementation but maybe a start + +Add new widget gives me following error +The mavenArgs would need to be updated with the P flag for this patch to workThanks +Can you explain in a bit more detail what is required here? Im not sure how to address this +Messy patch I was using on hdfs trunk +didnt mean to go dark for so long +Update Item unit price and quantity +I couldnt find anything in the documentation about this feature being deprecated or no longer recommended for use so I assume it is supposed to be functioningWe cannot switch our project to without this feature workingas we are exclusively using this way for dialogs with lib +updated in revision . +I guess we can postpone this for final and see whether in the meantime bndtool has this problem fixed +annotated shutdown methodand container will handle super classes and initshutdown them in right order +Things to do +Hi FlorianGood to hear that it works for youThe problem is that some of these order dependencies are in bundles that are outside of the CXF domain so Im not entirely sure whether all of these can be solved within the CXF subproject +Hi Laura Patch looks great +No dependency on additional third party jar should be introduced +Heres a patch we applied to our fork to deal with this issueWhat it does is bootstrap new brokers in the same way existing brokers are bootstrapped for new topicsIt includes a test +Submitting a patch +Attaching simple patch that throw an if the user tries to use CAS with the protocol v. +For this kind of issue we need to isolate gadget instance related variable and API +could you review it and provide better user case description of your issue +The repo format of Maven repository is just logjjarsjarsjars original bug was as of last May using Maven. +The test suite has run successfully in jenkins. +I think Arquillian dependency can be upgraded at the same timeMaybe thats already in your plan +We used to go through the FSN directly to avoid the checkOperation checks but now that checkOperation is inside FSN we needed another way around them +ReviewThe dictionary has the case sensitivity flag and implementes multiple copies throughout +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Disregard my method question +Remove whole test if you cant move it overThis is a big change +Sure +ant also failed +bulk defer of featurestaskswishes from to +Ive uploaded a patch file named which does not include system scoped items when writing the classpath +Would someone mind adding me as a contributor so I can assign the JIRA to myself? Thanks +Youre overriding dependencies that are already declared +We should look at getting some of those +Passes didnt run the entire suite +At startup time you know if tx recovery is on or not but you dont know if HRMemcached servers will be in use +We have lots on our plate and I dont know if well really find the time to do this for gt if it ends up in it will be included in geoserver but most probably not in since the change is really big +because of some package changes from H to H we are registering some exceptions by name instead by class thus not having to have branches +Please update your local copy and make the patch again +This is a corner case +Not sure if this should be critical but it kind of is to me +I get intermittent failures despite the patch +New version of patch that also clarifies that bytes are unsigned +Unscheduling from +SMX developers you might want to update the component for this issue to +Can you briefly explain the fix +If they happen to arrive eventually then we could look at doing this. +file attached JustinI was unable to find a suitable util class to factor out the is method therefore I made the method in great Victor +I would guess that the jps executable is not on your environments PATH +I remembered how I got to working after the NPE if you omit the protocol on the connect things work + +with the plans to add Rest services to Spring MVC +However to completely certify it we need to run the entire testsuite with different databas do all the process of release testing as well +The Security Guide does not contain all the information required +issue resolved with Revision + +done in +Since we break a lot of stuff in Conductor with this I suggest this bug as a blocker for release until we figure out how to fix it +You said that you could also cause this problem by using an identity column in the history table instead of a sequence +what do people think of metaIt seems like it could make things easier on clients removing the need for +yes i meant noticed I missed something when merging from the previouse verion +Perhaps we could get more accurate dependencies is used class folders instead of project references but then when navigating across projects in Eclipse it wont be possible to edit the code linked like in all you could use one of the workarounds below disable workspace dependency resolution for project module so it would use the actual jar deployed to the local Maven repository and wont bring test classes and resources make sure that each projectmodule is using its own namespace so even resources are brought in to the classpath it wont affect code excecution in the patch to trunk +This means that it must be included as binaryTough it must be modified so that it can be used by the Cas EditorHow can we solve this +I removed the unwanted methods in the code +Wait Java is faster than C? Something seems wrong here +List of files changed Command New New Statement New New Note Stored Procedures are not supported for named params get applied under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Instrumented class fails on interpreterIt can be incorrect bytecode which passed verifier somehow or bug in interpreter implementationMoving on +beta is just crucial fixes +it doesnt have anything to do with this issue +For this case it was really easy to do it by hand so no worries +authcore Rev +Do you think it should be synchronized on the or on the writable itself? It seems to me it would be the responsibility of the one calling read to ensure synchronization not the writables +It turns out that I forgot to make include all of its dependencies +On the other hand this may allow us to better handle multiple forms as well +to create configuration for the POST Servlet to enable autoCheckout for the duration of the tests +Indeed the classes created in the package werent serialisable +Look okay +Fixed in trunk +Michael changed to Apache Software Foundation added are no longer optionalPatch tested attached +Please try this patch +comments incorporated and tests dont fail +Thats why I considered to remove the timeout thing in the log split managerI would this +Hi AllCan i start working on the implementation of ID and algorithmsAnyone started working on itIf not please do replyThanks +Seems the issue is fixed by the commit +Cant we have some generic solution which supports all the databases +Please reopen and provide additional data if you encounter it again. +Please verify in the CR build and on EAP details in previous comment problems with test JIRA opened for it waiting till patches are commited to repository so Jason can fix this testcase. +I can imagine some sort of join in a custom would be required as in DRM +Thanks for reporting it since you never know! +Attempting to upload the patch again to get the newer version +Martin favoured matching names in the code to those used in the published references but I think its confusing +Is this patch going to be applied on trunk? I searched the code and found that it isnt and comments say that listInstances will be deprecated since ODE trunk patch is tracked by missed that thanks for reply +The previous patch didnt apply because was not committed at that point + to trunk. +Im I missing something? +For the love of Pete +I can certainly add the automatic async view updating in time for if no one else does +push to +Very you also add some plain text as a comment that documents the changes from an point of view? Ill commit the patch after that +Thanks Nicholas The patch is attachedRegardsAbhijit +After adding the name attribte code generation worked perfectly +Alternatively use the same basic workflow with Git +committed +See +Bulk move issues to and +You can create a folder in tomcatwebapps and just drop it in +After switching to a redirect it was simply left behind +I wonder it if is browser or location related +The data source connection manager pool is stopped and a event is not yet notified to the Managed Connection so Teiid sends closedConnection event to the pool but it has no way to cleanup properly as stopService is already called on it +fixed in revision for pdfbox patch for is coming soonCan you please take a look at the commit? I have done some little changes most of them typo licence headers and one improvement inside the objectWith the typo changes some method names changed +In both analyses I still see some duplicate jars +No I havent +Turns out WAS is more strict in its adherence to a specific web container behavior defined by the spec +Im not sure how much that bothers me since were breaking a load of other stuff as well +I suppose this must be fixed with +Please use a new or retry it manually at the testcase +Skimmed the patch +Please check the attachments +Also translates the incoming MR config into a Tez config with per stage configs sets appropriately +The above would be a natural extension of the existing in feature of Forrest +CCE no longer thrown and are less verbose now +Ill make a new task to publish the other document +Provide the patch +Hmmm +I just committed this to trunk +So wed lock even before the commitlog entry +Suneel can you commit this if you think it is good +Closing. +See attachment files +This is working as described in the specification +Something to do with slowing down the clients +Thanks Andrew! Were upgrading our version of Solr and this will make our users very happy +Committed to branch +Happy to agree the default is enabledI believe there is another JIRA somewhere about making different config examples for users depending on their risk tolerance or interest in performance +Sorry missed it earlier and saw it nowbq +Pluto Portal Server +Are there any updates to this +Hi Tonythanks for contributing +Dhruba werent you going to write a test plan +We can add properties for these elements as well and make sure that if these properties are used that we create the corresponding constructs in the human task service. + is done +It is probably important that I add the class I am testing is a controller and Tests class extends +is not is not bundled +NET +Also like I said earlier the new configuration for can be called as they are not overridable by users +This needs to be moved to the iBATIS user list if it is still a problem. +Tested with latest code as of noon Pacific todayStill see the the way I am not testing against sandboxI am testing against cluster installed by we need to have the old sample topology available OOTB as well withthe hostmap provider disabled or removed entirely? That provider really iswhat makes the sandbox topology specific +In constructor should we check for blockId? Earlier we checked for blockIds array +The interface implementation separation is to support the strategy pattern for alternative implementations +Therefore I would appreaciate if you could briefly summarise for the uninitiated Westerner the line breaking rules for CJKThanks again for you interest in and support of FOP +Mark Do you think it a good idea having the documentation in two places in the xml description tag and then elsewhere in the software or up on a wiki? In my experience the two tend to diverge pretty soon after they are madeThat said maybe were missing documentation of the key configurations or better intro doc +Patch attached I ran the tests and they all passed +bq +I think its a very convenient feature and is currently used by several customers that in production +Have you guys considered using a consistent hashing method to choose which server a region belongs to? You would create buckets for each server by hashing serverNameportbucketNum and then hash the start key of each region into the bucketsThere are a few benefits when you add a server it takes an equal load from all existing servers if you remove a server it distributes its regions equally to the remaining servers adding a server does not cause all regions to shuffle like round robin assignment would assignment is nearly random but repeatable so no hot spots when a region splits the front half will stay on the same server but the back half will usually be sent to another serverAnd a few drawbacks each server wouldnt end up with exactly the same number of regions but they would be close if a hot spot does end up developing you cant do anything about it at least not unless it supported a list of manual overrides +I guess it should be some tool on windowsRegards FreemanIts probably due to the tortoise tool on windows as it doesnt allow you to add new files into the patch only changed files +Merge with branch please +As there would be several need change I just append a script file which can do it as one goalJust run the in SMXHOMEexamples fix that and release just after +This has hopefully stopped happening now +Created for the unicode compilation error on RPM systems. +From the description in this ticket I expect another patch here +It might be your local environment like you suspect +Here are the logfile and the resulting graph which shows that should be evictedexcluded +Hi CameronI am working on a similar issue that you were were able to resolve +The pach also touches some other files that did casts that are no longer validnecessary +Thats what I was thinking Ralf +I still think we should also commit Roberts patch +Wow I buggered that up +Applied patch for trunk as Rev +Can you please update it? ThanksFixed conflict with in +Thanks for the feedback +Verified by Leviev. +Further to the property I put it back in and set it to validate instead +I the module with the project depending on the ows model jars as opposed to being a project dependencyNow Ive regenerated the project files and then regenerated the and the model seems to be fine now + +Default version for cookie parsing is and not +bulk close of all resolved issues. +Thats fine +See earlier comments on this +done see for the code changes. +Again I agree +Yes it should be all upper case. +One argument for keeping both would be consistency with the shell which already supports hex notationI think this was the ideabq +Redirecting to the Referer by default breaks the httplocalhost example form +Hi Wangda Thanks for reporting this issue + +A test case is added in r +I just committed this thanks Chris. +At the moment I do not have time to look into it but it will almost certainly occur in since the JAAS code has not been touched +Lets try the next build +no immediate requirement +Pat is essentially the same issue except that the suggested solution is weaker since it doesnt handle the problem of time jumping backwards nor suggest a method to detect the time jumpsThe numerological coincidence on the issue number is quite striking +I am able to redeploy these wars with both set to true and false. + +Attaching patch that makes faceting work with field collapsing +On the left sidebar there should be a section entitled Available workflow actions incl +I just committed this +It will depend what it covers. + + +Thanks Mahadev +Here is the patch for this issue +What do you all think of this simple patch +would sure be nice if you knew your bug was confined to a certain module +Confirmed +deployed is added +Committed to. +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I always do my testing ie I NFS mount my development tree and just use startstop scripts from there +The current searcher already handles this type of search +Nate did you get a chance to test the patch and verify that it does not introduce new issues as you saw with +Out of convenience I have only really looked at for now +pete +Thats not supported +This is fixed on trunk it was caused by custom toyaml method that returned strange values +Reduce some loops and sleep for a short while to make test can hit pools maximum size +with works fine +From what I can tell the while loop isnt going to break until all pending are accounted for either by error or by valid results +But as I told before its problematic because of licensing +when will this be available +Do you want me also commit it to branch? I just dont want this changes to get lost in the transition process wherever it goes +I cant reproduce this on Gingerbread which tells me that this is a weird Android quirk +You need to add a reference to the located in the lib folder of the binary distribution +Havent seen a jira for this go by anyone else seeing this +Perhaps it is because Im using Cygwin +If you generate a Mock object using this version of it adds the following first line to the build randomHash personally dont like the version numbering scheme so I chose to go with this one +Some implementations assign a single shared username and password per schema so the connection will default to a specific schema +Amareshwari more details about your example? From your example i can not reproduce the problem +A more clean refactor +Also simplified teh interface of to comply with passthrough I guess +Reopening to revert changes +Ultimately any EE container has to be able to process all the environment metadata independently of web beans +export LANGbq +oups +for the patch +I am going to close this as Wont Fix and file a new bug for varargs. +Good fix! Theres another case of this bug in +Salut Olivier if you want one I will provide one +good question +configurations are much more common and the problem is again about limiting local access +Thanks for the quick fix +Robbie could you please review an updated version of patchA newly introduced by previous patch was removedAlso connection closure was removed as well +I think a better solution could be client notify the datanode of the possible corruption and let DN to doublecheck and report to NN +I just found same issue when trying to fix seems there is still something to do on this +Thanks Jonathan +I understand the value of the improved testing but is there any tangible benefit to users to converting the +Common has or so warnings and theres a couple dozen spread out around the others +Using a listener is the best decision for such cases +Oh wait I see it already was +Thanks Billy for all the hard work and endless benchmarking so nice to have a block codec that is simple and clean and reuses our packed ints optimizations +The hashCode difference was because the internal to was different +Received an email statingSorry for the late reply +I am not concerned about the code duplication because its only like lines +Backup master becomes primary master successfully +Apart from unit tests and loading did you try killing some region servers +I just committed this. +Fixed the issue in close of issues as part of making release notes. +Not sure whether this affects the branch as well but I wouldnt be surprised if it doesIMO we need two fixes we should fix at the write side to properly canonicalize the path before logging the operation we should fix at the read side to canonicalize as well so if someone ends up with edit logs from a prior version that have this issue they could upgrade to repair the cluster +This issue has not been picked up please review and see if you can help fix the blocker critcal was created before July please review and resolve we are approaching code freeze in days +incubator +As a result of discussion I am reopening this issue and attaching an updated patch that provides unit test reports for each module within the QPID java framework + +Fixed +In any case I am doing now. +The above blog could be mentioned there +Our typical key size is under bytes +If you manage to reproduce this issue with a repeatable test case please reopen this issue. +However other projects prefer a more configuration based approach and prefer to simply wire in the id so I provide both +This enhancement will go into patch is really valuable one +If I understand you correctly you also want to know what statement a body tag belongs you model the relationship between a body tag and a statement? Unidirectional? What way? Bidirectional? What does a statement look likeI still dont understand in what situation meta information about statements would be useful +Reviewed by several folks +Since this is typically around I would be surprised if a small buffer size results in lots of seeks since the OS should perform readaheads in its buffer cacheWhat OS are you using? What do you have set toIf on linux what does iostat x show when things are slow? How about sar n DEV + +It was due to an error in the testcase +It is for some reason the last digit got lost +Marking this uber issue as resolved since all subtasks have been completed or marked as resolved +JasonThis patch just wont retry sending the data when hitting a +Otherwise Ill try my hand at it as soon as time permits +if a am back at the office i will send you the whole transcript output +Unpacking both the unsigned and signed files gives mefind +This happens pretty frequently in testing the tree +Thanks I have fixed all the to and to in to use the english locale. +I was able to do multiple iterations of rolling bounces withoutissue +I know I have been bitten by the retries several times and it is painful to wait that long in an automated test before it finally fails +Hi MikeIs it a fresh installation or an upgrade from a previous release +Just saves one trycatchparse if a value type is defined as Long but only works as a Double +David is right about the constraint language not being expressive enough for this case +This will eliminate concurrency issues and safety issues +Thanks Lahiru for the guththila doesnt write additional namespaces +It might be a better idea to have a separate timeout for reads +Oops +But the resolution that has been given does not work + test case to prove that may corrupt national characters on large strings +Its great for me work with codes coming up this patch +way to go +The schema generation for the databases happens from the ORM during the build +I will try with mysql later today but I just realised perhaps another difference is that the apps you are uploading contain only some parts of a grails have to run grails upgrade before I can use them so maybe an important part is lost in the script that creates this zip or do you create the zip yourself +Thanks! Hemanth for review the script and tto provide the valuable suggestionsI am attaching the script as patch after incorporating all your suggestions +But youre probably right if were not using that now it would be a good thing to look m also all set to fix the missing catch of but I have not been able to reproduce it from the command line it seems like its something specific to embedding +Commited in I didnt noteced that it is not fixed in +Sorry for that and thanks for reporting +comments +If you really wanted a patch to be included earlier than for a release a little bit of your time wrt running the CTF test suite even now might cause wonders to happen +Im not sure if its tested there +Per the directions in the java doc at the top of +First patch was against TRUNK this ones against +Delete all exisiting snapshots of a tableJust want to call out more pointedly that it deletes all of them +To coming from a commercial workflow background audit trail is a very essential part of any workflow based solution +Sorry the problem of spring transaction and session was fixed in the trunk + +Page functionality should be finished +Thanks! FWIW gsingers and jukka were s mentors in the incubator looks like we never transitioned wiki permissions properly on graduation. +It works! thx a lot +closed for. +Looks good to me + +It has that runs fine and with exactly the same code but it crashes. +AndyDo you want Hadoop QA to run the patch through tests +Forgot to mention that this would seem to be a regression of +So all issues not directly assigned to an AS release are being closed +No problem let me known when the webbased solution will be can i find the terms +Ok above is the fix someone plz commit it +Thanks Devaraj +Hard to do this feature in a scalable wayIf wanted we could do something like cassandras UUID to mint that go in a chronological direction if thatd help. +A provided means that the artifact is available in the target environment +I am planning to add some more grantrevoke tests to be run after upgradeHowever I am wondering if we should also test this switch for databases created in itself +Will attach the message. +Initial patch for review +deployconfviewerlist I can only get the changed template discovered if I change it before I start SP for the first time +Fixes indentation and Exception messages according to Averys comments +Removed EAP since they are in maintenance +I unfortunately still experience this +Thank You Andrew for the clarification +Ill spare you more blathering there are good examples in the wiki and even more in the mail I have sent you already +The current is quite big and it takes long time to load +Heres my minor refactored patch +on the patch +bq +Fixed with revision +Probably not a big deal we can just change an existing component in tomahawk nightly like we did for +The ER bits are in the JBDS final build +No need to create a clone I believe +The other unique thing implemented in the Twitter search as described by the shared slides is each posting is a single int +Temp files are removed either by explicit close or via an implicit close issued by the shutdown hooks +No sense in cluttering things up with the extra If when it only executes once in a blue moonIll apply the patch with that change +Hi I have started the work now but have run into a small issue with the element names +bq +Thanks Ted Yu now I get it +I think its still worth doing +This is not a limitation of is fundamental to how the JVM and operating system works +although the isnt necessary as it is already part of wOops missed that one +Defer to future release +do you want to create a review at and add the shindig group to the reviewers +will all start with or Im not sure what you mean by groups +So for the final merge into the reduce and the intermediate merges a counter will be provided +Still researching +Here is the Bug I was asking about +In fact your patch will make Harmony act as the description above +Current version is and I am able to produce this bug. +Ive done a fresh checkout and a new build +Marry to ! +Thanks Hoss and Rob +the same exception in interval between undeployment of ear and datasource must be longer to avoid this exception +Also the user could log the message as WARNING or ERROR using logj etc +Repo was +Using this feature will create zipped files for deployments +I have noexperience with Maven and I do not have it installed so I cannot do this task +If working on the backport for this issue +Please verify whether this solves your problem. +Other comments There is TODO in +So I think it is reasonable decision that to get user out of that since these are pre produced jsps by axis +complete source +Got some unrelated off the head of trunk when running Reran only the and it ran without problems +But such test would pass even with current code +Committed +The title says is that right +Just thinking out loud not sure what the best solution is +Hi Plamen Please feel free to take this up +Logged In YES useridYes this needs to be rewritten +However it will really a great help if you can answer the questions I haveThanks Regards have sent the query to user list +A client is one that uses a socket and selectpoll to determine when it is ready to readwrite fromto the server +I will let the community know once I have something that I think will be of benefit +its fixed and will ship with the console no good in EAP CR console +bq +I forgot to mention that I used Kdd instead of Kdd for so a Ratio of in the results means of the whole Kdd dataset +Either way this isnt the same fix for chujihuiyauns problem has not been applied +Sorry I didnt explicitly state thatIf the repository supports optimistic locking then only one camel instance should be able to recover any one exchange +Thanks +Just committed this +I have replicated the original problem on a Tomcat instance under Java where roller was also running +It is very awkward if wizard finishes normally but puts editor into a broken state +Ill be working in the evening again but if youre faster go ahead and commit it in +Also if I am snippy I am hosed on a project without a fix here + good catch patch looks good +Patch versions of the changes +I think people use varchar in Hive probably because of legacy data from +Example operations where previous value is used to detect if operation worked fine include amongst others put replacereplace remove +This is not supported lucene cannot read lucene indexes +even now +It turns out the Java are raising but because of the way we have our finally blocks set up we never see it +Verified and closed it. +Adds for IO testingAnd a in example that uses this +This is Morphia result in same application just add new action to controller WARN Time for object is ms WARN Time for object is ms WARN Time for object is ms WARN Time for object is ms WARN Time for object is ms WARN Time for object is msThis is full objects with real peresat swear and help deal with the problem +If no vtest need to rewriteI tried other scenario b which is public field with setter and it also failed too +When I run the unit tests on this I see Test failing +In case its useful take a look at which is my attempt to provide a abstract implementation of Channel with threadlocal Transaction objects +This adds a function to get the path of the folder +Repeating from thread on list not needed imo +Im creating a ticket to improve Sonar WS tested! +We can close this for now since the installer is dormant and likely needs to start over. +batch transition to closed remaining resolved bugs +No need to wait for the next release if we create a class with a new name and the desired new functionality +Unit test is available +fixed and issues associated with existing releases +I wonder about JVMTI threading functionality +Im on dont use Concurrent collection for pushpop as the is always thread safe +fixed +This would also hold true for having the view know about the model and the controller know about the view +Bug has been marked as a duplicate of this bug + +fixed please check out the and check your tried build on the issue to Achim Hgen and Michael Kurz for providing patches +Next time i will do it as adviced +Decrease priority mark as improvement since it only affects demo +This should be available thru cmd line and web ui +Adds dummy values to avoid breaking build +Confirmed that the patch in works for methanksdims +Jaroslav can you provide steps to reproduce the issue? Thanks +bq +If committed I assume the target will revert back to Java SE when the infrastructure is in place +For +I will submit a patch that homogenises the undefined tests as another bug +Went to backport it to and it had a merge conflict in the module code +Hi Eric I am not sure why changes have done in preinst file +If anyone sees a reason why it will require api changes that I missed please let me know +The optimizer will be responsible for generating the physical plan which will manage things like join type and ordering +Oh! there are too many attachments now +This needs to be fixed +pending jenkins +fix checked in both branch and reapply to checked in HEAD +Committed and trunk +Renaming bug to better reflect the actual problem +Gavin added this a while needs updating still +V +I verified that it is doing what is expected +Applied branch and trunk +toggle is working well and used by paramchem +Hi Brian I tried to apply your patch against a checkout at on the th March +Test case that fails +fix for the problem +Further bug report conversations of will be done on +This can be done on client side using new API see +I have created and provided default implementations of the interface +After that I will try some test for performanceThanksLeo +This and the fact that a cache can now be associated with a classloader means that we need a marshaller per cache. +A disk seek takes ms so at a rate of events per second youre looking at a lot of seek time if the disk is being used for other stuff than the flume logs +So the impl lets the superclass attempt to find the bean in the container first +Both share this problem +Im glad you like it refactor it as you wish! This patch should make it possible for mahout users to use their own customized similarity implementation for the distributed patch with some largely cosmetic style changes +I hope this is not terribly urgent +Just to update Im preparing a second patch which while it doesnt change the functionality improves some formatting issues as per comments on another patch plus adds a unit test +So unless we add structural subtyping to the language or come up with some way of turning it into nominal subtyping this code cannot be valid +Attaching a release note for this issue +Thanks Jothi and Billy +BPEL User Guide and BIRT Integration User Guide these guides are relevant only for JBT because these plugins are not included into JBDS +AFAICT thats only need for the ability to bind to service ports and possibly this file access +Could you please rebase the patch against the current Git HEAD +This feature been available since build and since build you can enable it from Maven page in project preferences dialog. +This patch was submitted a while ago and the implementation has also changed since then. +The patch is simple and straight forward I am assuming that Jekins is OK with it +Closing issues which have been resolved for some time +Cloudscape did not include a CLOB datatype even though was introduced a year and a half earlier in JDKI dont know why anyone would use LONG VARCHAR now that Derby supports a CLOB datatype +if that flow isnt working then that is an issue but the case of an admin making an account with no password is valid design +can you please attach a working example tnx +Could we rename lastKey to last and then take the has out of that assignment? Since you also added it to the big if statementI guess deserialize is only for unpromoted indexes? Because for promoted ones decoratedKey will never be null +I also added tests for a bean based server and a Proxy based client +Itll be something thats totally within the using special context variable names +yes much better thanks +The patch does not apply against our current Git master +Google catal alto el foc x alto de foc x . +patch committed thank you +adding memory management settings to for compiler this is probably because of sphinx that generates the docs +Thanks for the fix Jeff. + would be nice in certain situations but you have to be careful with overloading a given Quartz Scheduler instance with too many tasks or it starts misfiring +Sure lets remove it for now and add it later if required +more than one of the above problems +I left it as is to make it a straight refactor but so long as people dont mind the slightly breaking change of changing the log category Im fine changing it +It tells you what the problem isCaused by Unable to read +The behavior of all the other queries will remain unchanged +bat is certainly the easiest and safest solution just use addon cleanup for now it wouldnt be too hard to change that to a focused deletion of files later on if that would be preferable +So this may be a more general requirement for how the subsystems handle +This would lead to region server sending RPC requests to master prematurely because master is not really readyThat happens in normal operations so no need to avoid +Its been great to at least get the tests running for this module and long overdue +any reason not to put the convict in markAlive +I have committed this both to and the trunk +when we intend to use Directory for writing +More appropriate is a test if assignable to Closeable then invoke close +Thanks +The tasktrackers would then require to handle priorityfairness among the multiple jobs +Does this fix now cause the graph to be drawn for these metrics +Congratulations Konstantin! +Do you have a code snippet that you can share +It was important to show how this workedFor example perhaps what we should do in Clerezza is ask that all Principals implement another interface so Clerezza can be open to newtypes of logins without those needing to be placed into the core authentication module +bulk move from M to move from M to HTMLCSS CA processors cannot be used to serve style attributes of JSF tags also there is no possibility to propose a CSS stylesheets collected by created by to HTMLCSS CA processors for the correct style proposals calculation +as discussed probably that stays in its own sub projectBut leave this for you as it could evolve in another fridays ant vs +As far as I recall we dont consider CXF a CFX runtime but a runtime which behind the scenes are CXFAnd from what is described above that is what is happening or +Any chance of a patch gaojinchao? Thank you +There are some changes concerning including additional resources to the dmgzip +So when is Grapes needed +There was actually a bigger issue in that the tracking information was being thrown away too soon +Hi Gabrielit is my understanding that codes aide in detecting and correcting data corruption +Trying to understand how it was implemented in the back end +Looking at the code I can explain this problemThe Agent Context ensure all handlers are started before the Default Controller component +Thanks a lot for the contribution Jeff. +Attached patch + +When I am awake from my Tokyo trip I will take care +Not ideal solution though +fixing JIRA metadata. +added attributesshow to func +closing underlying stream of gbzip streams +ok after backporting velocity changes committed revision to xwe could still consider in the future if these jetty tests should use the example schema or their own in +For its not a big deal in all case both write read path needs to use the stale mode +What about adding an extension point at the dispatcher level that can be implemented inside main and registered into the spring context instead? If the interceptor can get the service information it should be able to get the and decide what to do no +Otherwise continue to help me reproduce it so I can fix it +Youre the firstIf the issue becomes more frequent or a workaround too painful we can certainly revisit this and fix the discrepancy +Note that regardless of this issue and patch is incorrect since it doesnt have mapreduce stack trace so that test is bogus for all hadoop version +I hope that doesnt cause problems +Changes look fine +Sure +This work is all done in the file and does not depend on or use the file +Hey Max I have continued to try and rectify this issue +I think users dont hit this because typically its that calls this method and returns theId rather fix both of these classes to throw an exception if you call their getter methods more than once than silently pretending the nd time there were no results +Im not real happy with this It puts data to the right of commands where usually commands are on the far right It shows Module without any more useful data for the children It increases the height of the display we should think about whether to show child modules such as or child modules such as or both if were trying to show everything that will be stopped if the module is stopped the list should show transitive children and the entry for +Since the are quite small I combined them in a single patch instead of opening up a separate jira for each UDF +But as of now I am starting with writing fragment to the I had mentioned above I have just started off I will get back to this issue once I am more it was good that you provided extending the even as a fragment since I could start my work right away +Whether or not the jobtracker can support some stuff should better be decided by the jobtracker +That way we dont have to support this stuff beyond people asking the inevitable questions on the mailing list which they will do anyway regarding this issue +Thank you. +Committed +will get fixed when we releaseupdate M docs. +Filed separate bugs per Arvinds suggestion +The example above shows how to remove additional packages from the stack trace +Given that the subtasks do not break the trunk I plan to start reviewing individual jiras and committing the patches attached to subtasks +My comment re was of course directed at the version in not the new proposed in the patch which I havent reviewed +I do not but it is on my list of things to do +The guide is separated into two guides now +For the workaround should work +I think these are the same issue +Why should do thisIf you add this to then it will become some monitoring tool that is collecting data from the instances to see if they are aliveI dont think this is up to the loadbalancers should handle this themselvesWidoI am not expecting to do any health monitoring which as you mentioned is out side the scope of +Im still working on the scoping of the resolveIm at the moment trying to have a correct behaviour for the task +Since the priority queue doesnt need keep all elments in sorted order it may be cheaper than aOn a separate note it is common that the most impressive optimization for keeping elements is to simply avoid inserting into the priority queue if you know that it will just get kicked out +Cannot reproduce it but fix should be simple +provide details +Previously we were sending back SAS info including something about the default subject +Thanks a lot. +However it gives us the locality when we are running a jobAnd this time will only be added to the job setup time rightI guess my approach was making it right and then looking at the ways we can optimize it rather then trying to hack up a fast solution right from the startDo you have any other ideas that may be worth exploring +After I believe we are at parity though until random read framework gets fully baked we will still internally use this for experimentation? and Not sure what public means in this context? Does that mean we provide support to users for subclassing these classes? Or do they only have access to public methods? this class needs some refactoring since it contains job related fields Id consider it evolving +Does this meet the plans for AS +John since this exception is occurring on a index keys can you remove the index keys on the matview table temporarily to see if problem goes away +I am in the process of uploading this +I believe for this round this is as good as it will get since IE is a hard nut to crack and this in fact might be impossible with the select box +A program could not easily update partially +Appreciate your help +But there are now two new failures I created for that. +I have updated the patch to address Hyuns comments +First of all the tilde has only a specific meaning on unix based systems and it is generally the shell interpolation which expands the to the users home directory +Hi Brandon Fixed in v +The contains all the automated test instrumentation +OK that is good to know +We could add some special treatment for it based on the type declared but right now we dont support it either. +Committed to trunk +As I said before let us first commit this patch +As part of this +Is this a known one? Any JIRA issues open for this? The way it stands it does prevent us from running JBM reliably in the clustered production environment +This could cause hints to expire before delivery if one of the for a mutation has a particularly short +testcase included flag to yes +this is not JBDS specific moving to was released last night which should have a fix for this so should be updating the bundle forge to this +Added Affects Documentation so people know the Programmers Guide and thus the documentation needs to be updated along with this +Patch applied and missing files added +to write release note issue now that release note status has been yes +Ben has pointed out that this bug is pretty catastrophic in that it caused all rendering of the layer and layer group to fail completelyWould it be worth hardening the data reading path to be more during rendering +Mass closing all issues that have been in resolved state for months or more without any feedback or update +This means the plugin is now functionally useless +I think this is the issue mentioned on the mailing list where DWR has switched from cometreverse ajax mode to use piggy backing so clients dont receive messages until they make a request themselves +See +added missing file indicated by patch until issues are addressed +Patches applied thanks. +Spec has been updated so may change in future +Will use the initial approach with a better name for the internal next method +may essentially be a duplicate +Im marking this fixed because Bretts comments seem to indicate that at one point he was able to reproduce the problem +Marking closed. +Those inputs are already in files +Committed revision . +Stack us know if you have issue w thatNo problem +The extension now builds and appears in the web UI by default +Ihave made use of this in all the three places where this is currentlydone so as to hopefully make it easier to maintain this pattern +Can you rebase +And by the way IMHO its a blocking bug and not a new featureMatthias please change the priority and type of issue +Ill take this review over +Thanks a lot for confirmation Greg +Applied patch +The fact that it can be worked around by specifying the plugin goal absolutely as abaove +yes its only for cant tell developers you have to add all manualy two times a day each day. +with JBT. +Patch implementing the above proposal +Fixed by r +i am hoping no additional work is necessary +Do we have any +We use it to change the description dynamically with value thats generated at runtime +Updating the NMWEBAPPADDRESS with NMs webserver http port +theyre both illegal argumentsIn essence I think if we document what are valid arguments and youre a consumer of that API then you need to ensure that the arguments you provide are valid arguments +Just tried the latest snapshot and it seems to be working fine +Assigning to myself till we figure out which area this issue belongs to +I have committed this +to leave as is for now until we have a better option +Fixed +Ill do this cleanup in though since it will be more clear there +I just applied a fix for this +Instead it will log an error and return null to the caller since Configurator must have a core not a. +Graeme any update on this one? How did you produce the problem? Do you have some sample we could have a look at +Thanks David +This is a JDBC requirement that iBATIS must comply withThis issue will be closed unless you confirm that it doesnt work with the JDBC typesCheersClintonImproper configuration. +makes the target stay out of buildtest and only use buildcontribnametest +I use the generators because I do not want to edit the WSDL manually +I ran several hours of load against the nightly build jars with no errors +From trunk build results on Apache and EC Jenkins mapreduce jobs passedIntegrated to as well +All it does is return a path to the so +Improvements will be done in the context of. +Pls provide your patch is for String is used to store pid and ppid +Added the section about accessing the archived emails +Here is a patch which breaks the ICUC materials out of into the separate file and adds the required text to LICENSE and NOTICE +I searched all the files which make up this book and the only instances I could find of capitalisation were correct + +This overcomes the disability of webbrowsers to set headers programaticallySecond uses a header like to override the HTTP methodNote that were already using ujax as a prefix for request parameters interpreted by jax the pattern for the first filter should be similar +Wouldnt removing it break usage from supposedly to definitely unsafe +From stacked password +Got too much going on in my working copy +The new section is titled Upgrading an old database to use SQL standard authorization +This should be fixed now +good plan at the lest I need to know how to tell cobertura to build a new site on each is already fixed +submitting +fixed in rev . +I suspect it would be a better use of our time to actually use annotations and drop the fake javadoc stuff entirely +What do you find weird about that +Applied patch +the is actually a duplicate of but I will hotfix the problem by resetting the system environment variable basedir after embedder execution +The proposed fix is to throw for type unable to be resolved such as the test case +Remove duplicate jar classpath to include Hadoop libraries +As a sanity check Im running derbyall on my linux machine +Looks like this gets fixed if I removenetsfhibernateca the file to etc directory. +No its not in CVS +Eventually the test ought to get run through a sqlscripts junit test +docs have been reviewed and are being pulled in +See +I wonder if we can do something about that as shown above the usual java solution is to have the installer program collect firewall information +I believe this is no longer an issue as I cannot locate this text in the manual anymore. +If you upload these I can check them into the Mahout repository +conf files +I guess that is actually a bug in our grammar and the ANTLR AST build from the rules +I just modified the list of available bundles to be those included in the release and telnetd from since it is. + for the patch +Id still like to get it fixed as our group uses both unixlinux and PC build machines and a common provides a lot of value +Part of a bulk update all resolutions changed to done please review history to original resolution type +I dont know why there is not any Teiid connection profile type +? Shouldnt everything in the entire system technically be hadoop +committed to the trunk. +Done +What would you guys think about using for the binary type +Hi PaulJust wondering if you had any thoughts on this +and are in this patch as well +JamesI am not sure I like the idea of introducing a new concept of transitive context attributes +Feature as described is not needed +See comment above for this patch +I feel so sorry about it +I was able to run successfully by invoking Ant directly with overridden +In that case theres no file to touch no mtime to compare +Central Time +Should work now +In fact doesnt Jiri even confirm that it remains stable after a weekend soak +Initially I thought the results would be perfect given the lack of overlap between latin and Japanese character sets but looking at the only lines in the +That is why all of the requested configurations for new hardware included the purchase of IBMs RAID card for this model +for the patch +Wouldnt it be easier to have a logical that would maintain a list of target datasources? The logical one would choose a target ds from the list and try to get a connection +This patch adds a length property to the map making it behave like an array +It no longer needs any mods to the file +This is not how we classify issues but we cant really rewrite the documentation of the tool we are using +Fields which is where the token class came from originallysaid that at on etime they had a complex system with different types of tokensand lists of tokens + +Committed to +Currently the best documentation for the build is the javadoc +Moving out of +It should be fixed with the latest enhancement on the producers check +Weijin is this something that you can look at +Thanks Wing and Jacob I will try splitting the file to keep it under KB +Karel what about the bom? I cant build that one because it still points to in central can this be changed to use the bom instead +More descriptive subject +Hi IvanIm not very clear about what you mean add the close action in the following catch blocksIt looks to me that the is thrown explicitly and earContext is not returned to the caller so there is no chance for the caller to do the close action +generally the first call startkey will always be empty +Updated patch for trunk +I am not able to reproduce with booking +The UI code already had the proper method to show or hide the box +Got the point JoergD I close this issue since it has no meaning them +I am going to it up on Google code was done just so I could start using it immediately +we are on the same page then +Nor does it yet have bindings for other programming languages +I have just checked the trunk it works for both spellingThanks Maarten! +Theres already another JIRA open about handling errors during state transition +thats a good point +All vts tests listed in this bug report passed on +The regression is fixed but I found a case that is not covered! +I feel with you +I know that this are malformed document but letting the parser crash IMO isnt the best solution +We store maps for return nodes and merge them when needed +How did the region server fail? Was there an exception in the logs +The very same exception appears after on each page when theres no database at all +This case happen in our testIt causes the regionserver cant do flushing cache because step is not executed and s state isclosingfalseclosedfalse fasle +Anyway Ill followup +Please provide me the ISO you used for this testing I would like to reproduce the issue +So this works with the delay but fails without itIs there more to the stacktrace such as something that indicates the vdb is inactive? Even in we did not explicitly wait for a vdb to become active before rejecting a connection +but It is not possible to change the the configuration when ever he wants since you have to alter the source file which means you have to compile it every time +I would like to be able to generate a project with and the following types of service from the Choreography ModelBPM ServiceRules ServiceBean ServiceFor the BPM Service the BPMN process definition should be generated with the Service Tasks being the services + +It is in another quality patch but we may be able to find additional by optimizingselectively caching. +Updated location for int and method +Ok yes can easily reproduce with attached test. +Is this still an issue +In essence I had a query doing an intersect of a circle with points stored in the documents Geo context point vector strategy no other filters or terms +the fix is already in review board +Just out there for initial feedback +Quite often these failures are due to pretty printing of adocument post signature creationThanksCheers Berin +I think there was a bug in that version wherein the spaces between the comma and the word was causing problems try removing all whitespaceThis is fixed in later versions +The distribution would become unnecessarily large and confusingI agree its much better to set up the sample ant build files to accept something like a a or an to pick up the majority of these jars and just add a comment in the readmes to let the user know +One thing I forgot to test running this from the Sonar Clirr plugin +Without the patch any files in progress will be corrupted +Reviewing would be easier with multiple revisions with rbBut posting patches becomes much much harder +Wrong result of the reprojection towards +SWF flash movie demonstrating the problems with case sensitivity in the code completion +See attached file for stack traceI have raised this issue with Oracle as well and will report any of their findings +ok we are hairsplitting +it was a lot of changes since +That would be better +Its just that in the case of a forward slash it gets stuck +Linking to because patch addressed the capitalization problems with XCL +If ignored then Camel will skip aggregating the timed out and continue +Currently speculative on maps often doesnt happen when it should because the map is stuck at I dont think we should be tracking the progress as each record is consumed in the merge sort +Fixed on closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them + +This is the patch that implements Rumen +Other todo enhance to visualize Joram +Attaching a modified unit test that is more automated but still requires an external this test I can reproduce the problem fairly quickly using in the unit also tested against and despite a high number of tests am not able to reproduce the problem +Please check +This prevents missing plugins and also multiple loading of one single plugin +This is java +The acegi branch is broken at the moment so this patch hasnt been tested yet +Marking as incompletePlease reopen if you have new information that indicates something is wrong with. +assigning this to Justin for review on behalf of Niel patch seem to only affect the extension? I am not the maintainer of it so probably not the one to review +Patch ignores primary key field of a Embedded field while inserting +Also theres another slight tweak to the patch to only go to the K allocation when theres a binary of unknown sizeEither way Im fairly certain that while changing the minalloc to a single byte shows that there is a bug its not the proper fix for the bug. +bulk defer of open issues to +Let me explain a bit on how namespaces are arrived at in javawsdl the wsdl definition has a namespace that you can specify using the commandline option tn +to trunk and +All the test cases passed in Hive root directory +Change signature of to return Filter +Will somebody try this +Hi Todd This is nice work +Closed as a temporary workaround has been applied on master and branch. +Here is how it startsURLhttplocalhostexecutioneseventIdcheckhttplocalhostexecutionesURLhttplocalhostexecutionesURLhttplocalhostexecutiones +I am very sorry I get it +Take into account when searching for namespacesPatch applied +In my WSDL I have two ports with different bindings that have different porttypes with different operations +If someone wants to take this on please feel free +Closing the bug based on Abhinandans comment +This is a dup of rightYes with the variation on which branch is the commit target +Hopefully I wont make such blunders any more +Sounds reasonable +But it requires to rewrite many longer investing in EJB CMP enchancements. +Can you do a simple test and collect the logPut just host from a working state to maintenance see if it works and collectcan you provide required this record since is now released. +I had my workspace synched up to a older revision when I created the patchI am attaching which changes check test to test that starting a global transcation when a local transaction is active will give an exception +I really thought I had my environment setup to limit to code +Should work in commit right solution was in using the goal izpack by mistake which used the old Maven plugin which does not inject the installerFile String in pico container +From the fix we create a new cookie whenever we initiate the airavata client instance +As far as Ive seen they are not focusable +Thanks as always Ivan +Ill move checkstyle to the test phase and commit that on this JIRA later today unless there are objections +Do you know if is there a DOC ready that explains that? What is the icon reallyIt seems a lego build block +benhenry any one of you can volunteer for running some benchmarks +Theres no way for the client to know its expected to use a token +I think I can take it if no one has taken it after that +In your Oracle database run the queryselect from VTIMEZONENAMESwhich should give you the list of available time zones +This one has been on my list for a while now but I am still looking for the best way to obtain the order of the properties +Marking as resolved +Ive added more tests more messages more checks against failure to try to prevent this and better diagnose when something goes wrong +This upload application should be making an per thread or using htablepoolResolving as invalid. +Still not there it seems +We just so happened to have cooked one up in Java +Ill get to this today +Here is the list Column name VALUE changed to COLUMNVALUE FRAGMENTPARAMETER RULECRITERION PAMETADATAFIELDS PDMETADATAFIELDS SECURITYCREDENTIAL Change from BIT to BOOLEANINTIn SECURITYCREDENTIAL UPDATEREQUIRED ISENCODED ISENABLED ISEXPIREDIn SSOSITE ALLOWUSERSET REQUIRESCERTIFICATEIn SECURITYPRINCIPAL ISMAPPINGONLY ISENABLEDIn PREFSPROPERTYVALUE BOOLEANVALUE +And nullable constraint for primitives doesnt make sence for me +With Macs Preview the thickness is the same although there is a discontinuity in the border lines between cells +Dont know if that matters in the fair schedulerThis is correct behavior according to the way I was envisioning it +Attachment has been added with description initialization block should beThis problem was fixed in the CVS repository shortly after the releaseThere was a problem with a missing static keyword on an initializer. +This issue also affects Axis version +Fintan I am pretty sure this is taken care of +Patch for Yahoos distribution for branch . +So this is this one that must be used +Adding everywhere option to select another is not realistic IMHO +svn wondering +I got exactly the same issue when ending process +Unit test was actually not failing for me for some reason but since patch made sense added forcing of US locale for and unit tests. +Bulk moved to with JBT. +importType added missing in original upload +bpelc. +OK I see that much theres a root mean and square +Assign EAP issues to Fernando +So if I understand you correctly Vincent this issue is in reality a doxia issue +Nice Cedric + +It references classes in the package which are not part of IBMs assume that has never really been compiled with the IBM JDK because otherwise this would have been am not sure if anybody at will tackle this issue or if it will be my responsibility to clean these references and then feed it back to +So all issues not directly assigned to an AS release are being closed +Kiyan thanks for doing this I had a lot of fun playing with it this evening +ThanksThanks to Filippo for the patch. +The extra src directory befuddles the site plugin Im researching +Ill try and get some perf numbers sometime next week +Just trying to figure out how to do the logging. +Can you provide the source file you used to create the image? Maybe as Photoshop or any other file in advance +Mladen it sounds like the attempt to disable the workers fail +Ben pattern Repository is more suitable for big projects and require less coding and code generation especially with JPA then proposed Data Access implementation can all objects using to build specific queries Criteria pattern + +JMCan you try with e for stop row +Yes there is +I just committed this +Are those needed? If so please attach to this issue and Ill take care of it +I recently switched from Windows to Linux but forgot to move my key +sorry previous comment should have said while we still support +I believe the release notes are correct +Since the server is meant to replace the one it should act the same wayAnyways this change broke backwards compatibility in a point release +Now ADB uses the serialize method to buil the as well +If you can think of any worthwhile refactorings for the new implementation please raise them in a new issue. +As far as I know that is the only way that you are going to get this to scale beyond data that fits in memory and likely the only way to get it to work well even for large data that does fit in memory +The nd variant of the new layoutNew VPE Preference Page screenshot was attached +Fixed for real this time +Tested on does not workIm retesting on +Let me know how I can help +This can be done by maintaining a stack of Bean method invocations not unlike how circular Import usage is currently detected +There is another JIRA issue out there for referral handling in the JNDI provider +I was hoping to get this into +In my opinion this seems +ShalinOn the inside it is very similar to a just customized to be making use of our interface + +Does it really make sense to throw an exception over gracefully falling back to something reasonable? We can of course add a flag to rather trigger the former behavior but I think people prefer the server to do the right thing than to force the client to resubmit the request +You can edit the scripts EAPROOT variable to change this if needed +This is a corrected patch +Committed rAdded the lazy option +This sounds similar to +Yes this is a duplicate please try with trunkWeld +Great although I was concerned when you mentioned the pipeline +Patch integrated to trunk and patch did not change the binary group with the change in the value of +This concept was never changed since the initial implementation +This was incorporated into the ultimate resolution for There is no independent patch or change. +But if you are going to committe a cleaned up version then Ill wait for that one +buildtesttestsfailed is empty after running ant based on patch v +Ill give the online update site a shotThanks +on dropping +Thanks +If this is made mandatory then the presence of might not make much sense +I think this can be resolved now +The Access based factory is now completely gone form version does not have it anymore due to lack of maintainership and lack of complete support for the SQL standard but Ill port the test over to HSQL and on and trunk +Closed based on original reporters request. +Reproducer for this issue +Should I close the issue as fixed +Release audit warning is about lack of license in the which is ok +thanks for contributing +Removed from SVN. +I suppose that this issue duplicates +I dont think we need to modify the heading code iOS is the only one that uses the options object +applied and run the affected testcases on trunk +Please review +This looks weird +Already fixed takes care of it. +I keep getting the bug testing on a win bit system +Actually I think it would be a good idea to have the examples around so customers can benefit from looking at them +That means other hadoop projects or applications could be using them +Ive been able to start servers in a domain without issue. +I can recreate w but not with SVN HEAD therefore the bug has already beenfixed +Any action +Thanks Johannes for the patch +Closing as this was released in +The issue was solved by switching JVM +So either the implementation is wrong or the javadoc needs an updateIn any case I think it would make sense to have a method on the request which returns the originalfirst resource which was targetted by the m not against setting request attributes but what about making this a little bit more user friendly and add as methods to the request object +I ran the tests after V but not V of the patch +I think that should cover this within scopeAlthough there is a lot more to say and do about documentation I am resolving this one after years +Please update trunk to explain how to run command line tools +Attached patch fixes described issue and also corrects DRLVM behaviour to throw instead of when an attempt is made to override final method +Difficult to provide review comments on javadoc so attaching it +The field should have a Party look up that shows only the Person +Thanks closed. +In your case that is nothing +Thanks for taking care of it +No this is done through jcr api +fixes the formatting errors seen by Knut and Bryan +The outlier is what does it get you other than some magic around ensuring that dates are stored as ISO dates? In the end its just a special type of though too right? Not trying to be difficult just trying to understandIn any case Ill focus on for now regardlessCheersChris +Infinispan also exposes JMX hooks and has helpers to achieve what you have done above +This is the crash log for the agent +Ok Ill clarify the help messagebq +Thanks. +patch a lotThis is committed and will be backported to branch +The OOTB layout has not changed in years +Will continue with UI tweaks on Monday +zip using file but not work against httpthe content is identical just the protocol is different +Testing now +All done github should pick this up within an hour or two +HiIve applied the patch +Moving Fix Version from to +Reopening this fix hasnt been ported to neither I see a persistence of the bbox I see a field for caching it so I guess its computed only the first time its requested +Im hoping to balance that out a bit by pushing back on less important issuesSo here we go +Laura I selected on purpose from system table to show that system tables character columns will have collation of UCSBASIC and the character constant T will have collation of which is TERRITORYBASED for this db database and hence the comparison between system table character column and constant character string will fail when the comparison is done inside a user schemaNote that character string constants pickup their collation from the current schema which in my example above is APP schema which is user schemaIf the same query above is executed inside SYS schema then the query wont failij connect jdbcderbycdellaterdbdatacreatetruecollationTERRITORYBASED ij set schema SYS rows selectedij select tablename from where tablenameT TABLENAME rows selectedIn the case above character string constant T picked up the collation of SYS schema and hence the comparison did not fail +On further testing by Vladimir he found that it works perfectly fine in RHEL but fails in RHEL +Is this feature actually even relevant to have on the update site +Please update with manual testing results +Switched to betwixt and format +I was able to reproduce +Verified on. +Patchthanksdims +Patch that refactors the to and then Deprecates the original with a doc link to the newer interface +Then run ant server and you should get errorThis example is scenario bunch of pojos that form tree structure and basic send over the wireSimple as it can beI would be more than glad to hear this is some kind of my mistakeRegardsEnvironmentJDK Windows Ant Cxf are using Spring +seems like a very straight forward implementation and I wonder why none of this logic was ever includedCan someone else please comment on this +Sailaja please provide details as specified by Alena and hope attached logs will help to find the root causeThanks it even the right logs youve uploaded? It says CP bug and it looks like its a duplicate of Management Server stopped responding in a Multinode Management Setup not this bugAlso please upload the logs to ASF Jira +For us it means we can not use Jackrabbit at all +Thanks +Yes +Thanks Kim +Ive already started the release process for Roo so Ill bounce this minor bug to Stefan for the next release +Im not calling this a bug as is the correct dependency +Ah depends on the timing its probably late for AS so Id say the first AS release coming after that +At present the test is not functioning +If curious why we were running joinring false after the node had data it was because we had ran out of space on a node and many had been created but unable to compact +Launch a python shell and type import pyro +We are going to stop using for Nutch +someone who has idea on this issue pls suggest +Remove the commented assert statement +Im able to reproduce this now and Im looking into this +But I dont think there should be any problems with backporting it in case someone wants to +The src subdirectory uses the CPACK variable version values +Will post it tonight +I put in a few improvements +modified filesjavaengineorgapachederbyimpljdbcadded call to in constructorsmoved checking for closed connection before check for valid flag in checkValidity method +If it exists then it tries to describekeyspace to see if it has the definition +We know that a class is an exception if its name ends with Exception which is not accurate +He can see himself among the users +to commit the patch +Closing off bulk of work is done +We use Groovy in a Web Start JNLP app and one of the final build steps is to sign the dependencies so the app can run with special permissions +Web part also ported to the running context +Justin does this ring any bell for you +The rationale behind this task in a bit unclear to do you mean unless required by the user +The version number could then be stored in the configuration +There shouldnt be +The upshot would be that timeoccasion diacopvegada needs to be explicated with examples including +I will verify the fix and get back to youAnd I will surely be cautious while describing the behavior difference infuture bug reports +That defers the resolution of the service until after and are fully constructed theres examples of this approach already in the Tapestry codebase +Can you please commit this when you get a chance +Ill commit this later today if there are no additional comments +The debugger doesnt work at all right now in +Metro always create this instance internally for address resolution +But changing it would break backwards compatibility unless we did something REALLY involved +Try to fix this problem the default behavior is not use this feature + +Attached a Flash Builder project for benchmarking the selectItemdrawItem performance +Please post your additional comments to the new issue. +AlekseyYes we can add this option to default server configuration +Ok Im closing it as fixed since I already did the changes suggested. +It is not a good practice to modify the users configuration instance +Ok +Erik can you elaborate on what compilation error you got? is the problem just that we dont have a deprecated version of inChris your patch has some weird stuff in it in particular a whole lot of previously deprecated Params classes now log messages saying that they are deprecated and to use the version in but thats not where the valid versions of those classes are located +Hi Marucuswe can fix this in +Ive committed the patch. +If only because the markup used by JIRA comment make it unusable as is + for the NPE fix +This is a feature of the Seam New Project for also deploy the datasource descriptors and database jars so the project actually works when deployed +ThanksCommitted for in rev +Can you put them up on RB with some context about what they do +Since Hudson reported success resolving. +Many thanks SebastianOleg +Is there any progress on this? This seems like a low hanging fruit to increase robustness of Lucene if it crashes at a bad time +It still isnt clear looking at both products and costs modules where you assign the relationships between cost and weight and the measures +ill try and motivate them to follow up on this for me +Yeah A sounds good to me +Fixed format of href to Apache the fix at revision +Marking as resolved +Verified on JBT. +Tell me what you progress or decision taken for this sharpening plugin +resolve for M should probably look at the method name post M +until something comes out of incubation apache cant release it +This also impacts + The actual example is good but its just to hard for the users to make a similar one for their applicationsI can confirm that I also gave up using because its just too complicated to extract for that example the required reusable code +I say close it out +Please verify. +Both issues modify the NMS API +Refine the patch for this issue +All tests passDouglas does this help your performance issuesDoes anyone object to this change +Hello let me clarify +However I propose to update the comments in because I think some of them are misleading +Attaching a revised patch with Apache license in new file and map file with just the change +will fix now +Thank you for your contributionPatch applied in revision resolving fixed. +It was initially runtimeSide note Eclipse unfortunately doesnt distinguish between mavens runtime and compile time scopes so people editing in eclipse will have compile time access to +Should we move it or keep it like this +New users cannot report issues +although this unfortunately wont currently help classes like used directly on a because a is no longer used I still think this is worth committing for users ofThoughtsobjections +class files. +The thing is that function could only be used for none start start events start automatically after being deployed +Thanks JasonI put this into trunk and +The complete class is also and deprecated +Going to leave it alone. +This isnt an Apache build issue +It would be nice if we could all come up with a common commons picture onthis topic +So can we close this + +Fixed in again in r since the coming iteration reimplementation will not be sensitive to rehashing +Clover did also not run but it does not use nightly or more than one iteration +ear and adds to deployment unit but since was added to +In discussions about this security aspect of the project we decided that doing both would be best especially since there are possibilities of holes for both filtering input and encoding outputPlease consider that we want the defaults to be as secure as possible and allow ways of doing things in a less restricted way when it is needed +Also if you store plain byte array via library mode or you cannot specify Accept header when later retrieving the data via REST +Verified by Mark +The code changes look copylocal flag in the parallel test +Removed instead is used for the exact same purpose +The patch is dependent on the localization fix of +Triaged July Assigned normal urgency +Its not like a car accident its like filling ones gas tank with creamed corn +What I meant was the manifest included in this patch was generated using Apache Felix +Thanks for the report! +That excludes because of incompat +Well +The xffff separates the property name from the property valueSee +This appears to have been resolved +I talked to the person Ann Marie Rubin who reported this problem in the first place and she saidWe found the problem in the xml that gets generated during the build +Please use that to fix unit tests +I was seeing this yesterday I will look into it +This can be done already with the patterns and though I think +We only had such reports on IBM specific setup of that Sun JVM debug mode or the like? just run it with default parameters it fails +Hi RonThe useJms just for demonstrating that it could be doneYeah the fix for trunk would be come soonCheersFreemanJust wanted to mention that useJmsfalse should be the default as the old JMS transport only used the APIs +Applied to module of the specs tree +Otherwise we must issue new + push to +handles dropping traffic to old members running on the same you had a chance to test whether fixed this issue as well +bulk move from M to ve tryed to reproduce it now on empty xhtml jsp html pages context menus on visual and source parts are its already resolved +Where should this be done? Integrated into the core code or move instance construction behind the SPI +Also we made this change long ago because splits were taking up memory in the JT and were causing +This patch deal with Windows short name and network driver also added some testsI just tested on Windows will try on Linux +fixed +The implementation is not done yet this is just a requst for comments +I have fixed this problem it doesnt care with connector +rightgoing to apply this patch +Patch applied +This would also avoid the overhead of iterating through the properties to exclude properties when components should really not be setting properties on a in the first placeIs it possible for the se to look for cxf properties itself and remove them prior to sending the messageExchange containing the message to the delivery channel? I think your comment makes lots of sense performance wise +bq +Excalibur has been updated in both Cocoon and +As Ram said the data check is also added +Ive run the same tests in windows with a VC build and again had no issues +Closing all tickets +Closed all tickets +Hmmm +However before closing this issue need to do interoperablity with interoperability successfully with WCF +I have not had a chance to test out the +Re sending the on the DMR requests I was reluctant to do that for fear of blocking out a custom client in an edge case +I figured we were applying this fix only to as a stopgap for a proper fix using JNI later +Already can access via +stateless and there should only be set created per EJBI think the real problem would be elsewhere +will commit when tests pass +a patch and failing testcase please +This is the same patch but compatible +You are right the order should not matter +Though even there Im thinking I should verify that the Xinclude happens before it is sent around the networkSummary Id go for the python option +Made most of the necessary code changes to correct this problem locally but cant commit them yet because Derbys functions are limited in the current release to not allow CLOB arguments +Thanks! +The ui has same structure hierarchy as that of the sanboxgsoc +That works for me +A fix was committed but unfortunately it was reverted by r +Same thing for snapshot to archivatrunk seem to have solved the snapshots dependencies are now unit tests for requesting a snapshot artifact from archivas repository with every possible configuration of the snapshot policy to test the snapshots plugin issue +At this point one of the admin commands will get an error and it will exit +Thats fineNicklas are you putting together a pull request for this JIRA? +Of course having the export without an import wouldnt be terribly useful so Id want that also +Thanks. +here we go +Thanks very much for the advice +Paul what should we do about this +Do you think you can fix this issue in the next release +I have to look into this deeper Ill let you know +As feared isms have crept back into these scripts in this case from +Is there any update on when this can be fixed? Many thanks +benonly your patch was applied +I find really confusing having classes in different packages with the same name and I am pretty sure it is the same thing for most know your project my use them in fact I am not saying we should remove them but since they are not used anywhere in the geotools tree I do not think they should stay in core coverage but in a plugin or unsupported dir or something like that +I think its a good idea to the configuration of the test cases +Ive put this in the core package for shortness sakesvn ci m Returning to Lang after it was removed in Lang per the request in +No there isntMigrating to lucene involves significant changes to the jackrabbit internal index readers which extend from lucene +Nevertheless your solution is solution is a good one +Once a file is validated the +Sorry I meant wouldnt and not would +I have attached the font that I am trying to use +recreated version of which I deleted by accident +Im trying to find out what changed between and to determine why +Closing old issues + +Currently this makes it hard to initialize an embedded C process inside a hive unit test +Nice patchHow does this interact with the cache effort +Hi LanceFor this release I would like to support JCC as an option in the demos +My bad regarding javascript issues in attached example +Can you highlight the flaws +I think the best thing to do unfortunately is make an for each snitch allowing them to expose their specifics if they choose toGaaack! Really? Theres only one in use at a time per cluster right and they all provide? Thats all I want is the methods on that interface +Breaks and +Turns out that there is no need to traverse the parent hierarchy +Im working on some comments +If youre not familiar with the regexp format Ill help you outcheers Roland +Attaching a patch for branch committed to branch +But not all the time +Patch creates a common Interface for the two CompressedFile classes to allow access to the compression metadata without casting +Sample VDB have tested with the supplied VDB that works +Did you have implemented this switch for Android only since? Because i didnt have any issue beforeAnd yes normally i have the error with websql on creating table which exists but its not causing the app to crash +As you mentioned we can provide better solutions with new architecture +Just put the jaxb artifacts into the central repo. +It may be clear to you and its clear to me now but not by looking at the code +Are there any more specifics necessary to reproduceI used IE and did the steps listed without any other changes in the portlet +I checked with Cosmin and he is OK moving to whats now trunk +The bottom navigation bar does not show up on Android Content flows out of bounds on both devices landscape mode on iOS is zoomed and content note Mobiles bottom navigation bar is extremely buggy and glitches appear on every page transition or orientation change +Looks like hudson is running tests on unpatched versions +It does NOT fix but rather has a hack in it to protect callers from the it may throw +Documentation patch and images attachedThanks +Building docs could you please add a section on using mvn site to generate the website under docs +Its still not possible to install JBDS on Windows XP +The bug is in the way the broker uses that fieldDeleting the field in the future is a good idea +Patch to handle apparent change in Cygwins for this patch just wanted to let you know that it was helpfulAfter upgrading from Forrest to running Forrest for building Lucene Java docs failed with Target +Logged In YES useridthanks +the info bit is key to investigating this +If we decide to go with the open classes approach this code will be legal +titrating +works fine with jdkBut can we make it work with jdk? because we are using java script engine functionality of in our application +I have the same issue +Kindly tell Oracle that we are tier one client of Oracle +Commit the fix to trunk at +Please add an initParams attribute like the other Tomcat objects as well +The two BPEL processes have been separated out so no longer part of the same deployment +Probably we need to add button Add as opposite to Remove +Cleaned up patch +So thats why I removed this +I didnt look deep into the failure I just noticed the and was speculating wildly sorryYour planned changes sound good and with them I think this should be ggFwiw my dev machine is fedora and I run the tests through eclipse there +ant resultsfindbugs are not related to this patch +Bulk close for the release + +Thats right +jruby is too big +Added code for validation of all keys values of history file by comparing them with actual values obtained from JTAttaching the patch + +Thanks JeremyYour suggested change is trunk at r at r at rIt may seem obvious but I would just have preferred that you would not have omited the setScale partMy apologies the rounding was already present in the divide operation so the additional setScale wasnt necessary. +Talking with Tom White about this on a different JIRA the original idea was to remove OKJAVADOCS entirely +Looks good but would you please make two minor edits +I will also split this into a Common patch and HDFS patch +Are you purposely trying to disable facet results? Why not simply pass in facetfalse parameter +This bug got into release because current sample hadnt showed a problem +Im notstunned there might be problems in such a thing anything we can do about it +The other bug is about creating volume from snapshot There is no NPE and its about operational failureBoth seems to be different +User Guide patch to update the documentation now included in +I believe that the patch I attached is correct +In the approach I took I initialize the hosts whether or not the number of random locations is greater than zero +Please increase the priority of this bug because it really break things and causes loss of data +Please review the patch and let me know your feedback +Im just finishing testing it +Ive committed this and merged to +Anyone have any issues with committing this patch +No functional changes so no tests included +So I think we should leave it that way unless there is overwhelming support for a compressed format +Works great Emmanuel thanks! +There must be some incompatability with the trace debug level making it through logj to hibernate +It is currently listed as a issue in the associated JBAS JIRA +I have even ran my test suite against SPRs OXM and it seems to work fine +closing with release +Thanks Mark +Excellent point +Unscheduling due to inactivity +Youd need to integrate some on the fly config loading and caching but this might be the best way of actually achieving issue forThe for routing the requests include the dataset name before the service nameStandalone a relevant servlet is mounted at each service point declared in the config file including no servlet if that service isnt offered +The escaping is an accident of the xml encoding if the in memory representation contains escape sequences there is a bug in whatever generated them +Jaigak thanks for the patch Ill look into itTypically we mark the jira as Patch Available when you want it reviewed Ive done it for you here +Whats the minimum version of Ruby that must be supported? orWhile trying to work through some of this I noticed that the mongrel dependency will cause an issue +Even though its possible to use the to get a single backslash escaping within does seem a bit inconsistentGiven that the idiom has been adopted from programming languages it would be nice to be consistent with their behavior +Should we remove it or update it +Attached stacktrace +Fixed spelling and grammar errors so that output makes more sense +trunk . + Bug has been marked as a duplicate of this bug +I think the current time headers are probably fine to be honest its more a case of using them in the IDE +I close it please reopen if problem still persists +This should be fixed in +KiranIll try to replace get with as prototype developers recommend +There are other ways to solve this issue +Looks reasonable to me +Committed to trunk. +We had the same issue for the book announcement. +Otherwise we fail the build +I think can be removed from clerezza +My project did not have a global style to center text but I did have a style to center all tds in a scrollable +verified under +do you know why that reinit is being commented out +Will this be a priority? Is this something you think should only be addressed by an addon? If so maybe closing the issue is more helpful +They verify that the init of stacked filtered filesystems correctly sets the conf in all the chained filesystems +If you could make that consistent that would be great +We should figure out why this example works in projects and what we do wrong +Verified by Sian. +There seems to be no apparent fix for these extremely long paths in Windows +I added sections for this to the user guide +Makes also sense to have it outside core +I have the same behaviour with a JSP that does the same thing of your quickstart +If user doesnt provide password installer will show a popup +audit output from I am not sure what the release audit warnings are +Ill do a bit more fixing later tonight +Closing as wont fix. +Had to do a little but it works. +Currently minimum shares in the fair scheduler will go to other queues if they are unused +Im just missing the info on where inside we should point the runtime? We dont want to modify +This only happens with OIO so far +Moving to Ma has made commits into trunk for this and other issues so I am taking this over +In the past Im fairly certain that wed always observe dupes in at least one of those indexes +I just got this errorfatal protocol error bad line length character Remoerror error in sideband demultiplexerTo httpsmgrigorovreposasf aae +has been marked as a duplicate of this bug +reason for the behavior is the same as +Ive changed the color to at +I attached the patch +Switching all issues that have been in resolved state for more than one month without further comments to closed status +All great points +released +Fixes the regex I was +Thanks Petr! +A new comment for mentions that the behavior can be reproduced by opening a file +but since this is a breaking change that would need explicitly called out in we should get pretty wide consensus before moving forward +However some decoder implementations try to resolve invalid byte sequences for example in case of fdral separate byte can be treated as character because it cannot occur separately in a well formed text +The fetching of rowscould have functions that produce side effects the application is expecting to occur +Hi SamiI tried the patch but it didnt work for me +If additional areas are found then this should be handled by individual Jiras +Fixed + is the only has this problem +Reopening this issue and Richard will add more information into the JIRA +Patch for trunk adding the hits field to the s +Regarding the backport for Clouderas distro we will definitely include this in in CDH and its looking like we may backport for CDH in the coming weeks as well. +Need to fix before GA +Also it now deletes the output file after running the test rather than before +revised diagram in Beta documentation have attached the directory listing in the Beta I think Isaac did some work on this without mentioning it here +Oh I see I mentioned that in the original bugreport +This issue is also about specification compliance +Hey Raghu the streaming benchmark is in theres a +Set the same fix version as +PR merged in master and provide steps how to verify this Jira +I actually had a chance to play around with this a bit over the weekend and it certainly suited my purposes of being able to restore in a reasonable timeframe should disaster strike +I think we should get this change in +I assume you want to do it because you want to maintain just one markup file irrespective of the various locale which your application should support +This is all done except for figuring out a way to hint range mutations during a partitionIve had several ideas that would keep all hints in the same CF that wont work for various reasons +Are there any further details available for this issue? Id at least need the maven project theyre using and the validation error theyre experiencing +It looks a lot like +Replacing patch as previous patch had extra unrelated changes +Search for in the attached file for fixes +Fixed changed calls to Widget factory to grant access requests by default +Thanks SkipYour patch is in trunk revision +Will take a look tomorrow +ACK to the patch above solves the problem for me on sinatra feel free to push to back to Ramesh for testing and closing of the patch provided by Michal and works fine +I have run many times on my local pc all passedIs there any other problem +Ive committed this to trunk and +Thanks for the detailed reply Dan +just used console appender not file +Im afraid this one cant be fixed due to limitations of the current Maven archetype +can you please create a common patch and mark this jira as dependent on it +It is a fragile approach +New patch that trims values at parse time +Addressed Teds feedback that includes moving some code into base class + +Committed to trunk thanks for the patch Karthik +is this still an issue? +defer all issues to +Attached patch appears to solve problem +for that +However the path during the controller failover will have performance implications on the controller failover latency +fixed in found that when using JAXB schema generation some XSD features disappear +setting property to! windows should help but after a initial test changing the stty didnt work the same way as on other posix systemsthis is not a issue but rather something that must be implemented in jreadline. +If you find out whats causing this we welcome patches +JIRA upgrade will be probably coming in the next few weeks +If alternate processing is required then please this Issue +Care to make a patch +The previous one had the problem that it also blocked during creating the Filter +Would that work? IMO its less risky than creating a separated module and would solve the problem of having a large pom +Tried this on the latest version of Groovy and the problem seems to have gone away I suspect when we upgraded Commons CLI some time back. +patch of +q file +This is typical log when BPEL process throws exception and everything works fine +I downloaded and added to my projectapp and I see that does what you stated above +This bug has been marked as a duplicate of +bq +The release contains the saxon jars but has no code or config changes over. +I think we could fix the pom as it is broken but we dont know the right versions +what alpha version of Maven are you using +Closing this issue. +If there is a problem please open a new jira and link this one with that. +If you have the BV but not a provider then youll get a warning in the log +TODO is to equip one of the examples with a process diagram so that Erik can test this +There wont be a release +changed to resolved in as it was known to work in release +Actually we will just take care of this inside the Shuffle package itself since the use case seems to be pretty the same usage exists in too +The issue in JIRA is a part of this initiative taken as part of this new umbrella PR +I dont think so +These are not broken on docs stageI changed the Javascript example to highlight as Javascript +Buenas me inscrib en la pagina del INEI para una de sus convocatorias pero ingrese mal mi nombre quiero cambiar esos datos pero no puedo me voy en ayuda y me manda a esta pagina me pueden ayudar +Stefan do you want to look at this issue +I will try to see if I can start it via Spring +Btw to replicate the issue just uncomment line in its line +A more uniform scheme would be to use clientId in both and + ready to revision +I will need to think about it a bit checked another mail client The Bat! also can not receive the same mails from JAMES +CProgram FilesAltovaxmlspyCProgram FilesAltovabinCProgram FilesAltovaxalanjDbinCProgram FilesAltovabinxmlCProgram FilesAltovabinCProgram FilesAltovabinCProgram FilesAltovaxalanjDbinCProgram FilesJavabinPlatform is XP professional and I have run the transformation from the command line usingjava in xsl out Any help would be much appreciated. +This was fixed earlier and can be closed +When we originally created the Simple Web Service wizard it was discussed that it should be available where users will look for it in the wizard categories +As I can work with the standard command line client without problems +TJWS container shares the same logic on provider loading +This patch provides the functionality I am requesting +Which is to say that the moment a column map is constructed with empty or null column names in it we should immediately register a failure +I ran some small perf tests in the original JIRA if you want to take a lookI did see the intended effect of utilizing more disks in case wmore I havent got a chance to test it +how about commands on the txbridge tests +Committed to and trunk +nevermind i see +Comparison with Phoenix type implementation can be done first since Phoenix is actively maintained +The commit should fail so the user doesnt think everything is ok as far as the cache is concerned +Use the adjustTimestamp method instead. +How about BTW replacing Hack with some form of Intercept would be better IMHO +Sorry but I cant provide you specific information with such number of testsTests run Failures Errors Skipped Anyway +Hi DanielThe situation I was thinking at isBundle A registered an SSP x withBundle B wants to override the Renderlet created for SSP x by implementing a typerenderletIf A is restared the service provided by B will disappear and only the typerenderlet provided by will be available ven though the other typerenderlet would have a higher priorityProblems also occur when one registers a service via and duly unregisters it in the deactivate method +Will open a separate Jira to fix test case +just to set version fields +All the others example I looked at were Russian dollIt is clear how this design aspect influences the generated classesThe problem was that I have a very complex schema and I havent possibilities to change it +a proposed patch to prevent throwing introduced default values for the domain and the extend attributes so that they cant be null +Considering this one done +It does not implement +Thanks to Owen for inputs +Update lets make sure we document the change in default +It failed because the lookup of host failed with +We need to fix array library exception usage +Regenerated v after an svn up there were some changes to that I hadnt updated to +It gets an undeclared exceptionwith the cause to be an IOE and the upper layer doesnt know how to handle itThe Set in Invocation is used to decide if a protocol a pb one so should be used +I also changed the test case to reflect my observation but it does not make a difference +Any chance of it fixed soon? Im still working on tomcat so it is a real +Sorry that I am not being clear +Thanks Yip +should be combined to something likeYou can set the CLASSPATH environment variable in the operating system either temporarily permanently or at run time when you start your Java application and the JVM +Triaged for +SvetlanaFixed in archive module in at repo revision Please check that this fully resolves your problem +Heres a corrected the result of a bad merge +Thanks Kim +We wont be able to wedge in anything pluggable in less than weeks and so we wont really have improved anything for users +Pretty easy patchSince the hdfs block locality stuff isnt really used for anything other than metrics just ignoring the case where no hosts are passed in has no adverse effects +commenting the merge operation will do what you want to you even try? Because before writing that down I did write a unit test based on your example it fails with merge and work if merge is commentedIn the book cust is a detached object not an object loaded by the same entity manager +I would suggest no new states introduced unless good case made that it is really needed +Maven wants one version of Spring on the classpath +But I didnt get any errors in axis libraries +So down its level. +It looks like theres time to get in before the next RC so this ticket is rendered obsolete. +Earlier I was considering a but it would be hard to make it work with all query parsers +I wouldnt be super concerned but it is something that we should try to look into +Could you brief me as to how we could get the response other than scanning this stream +Fixed by +Well it seems to be packing ok so its a regexp issue +I rising line when the replicant is suspended and then it drops close to when I unsuspend the replicant +Take care! Wonderful project +No one else has brought this up and this is not a major or critical bug +We may need to analyze all together once and handle them in phases +Jim please feel free to commit itIll work on after this one will get committed +All again passes +This looks great to meHave you had a chance to benchmark it against? Id expect it tobe faster with rare terms and slower with common terms +Seams to be a problem with missing runtime dependencies orgxmlsax Could not initialize class you clean your whole local Maven repository? In my case I had a completely empty repository and only compiled Stanbol with +Sounds good +Fixed as of rNo backport on this one because its a behavior change. +This bug is blocking QE acceptance of CR and has the highest must be documented that the minimum version of JDBC driver is was fixed with the change to modulescoresrcmainresources the jBPM source for +Raised the priority on this one since it may lead to unexpected +CONDSTORE enabling command does not trigger Fetch Responses that contain MODSEQ all the + +You have answered your question it would be nice to have another upload protocol that is easy for firewalls +There are a Jenkins jobs running framework tests continuously in Firefox and Chrome +Hi KatheyThis should definitely solves the problem with the split methodThanks +Sorry forgot to mention that I am using Kaha persistence +I agree any mention of seems weird after understanding that the attribute only controls whether the deployment will be slathered in arquillian juice or not +However the header should most likely not be set if the value is null in the component +However the patch is known to work well if you check out that revision so you may want to test that +Is this still happening with or trunk build? Please let us know +The usual solution is to cache and reuse the +Would also be good to have a release note here since its a new feature coming along with new configs +Will fix this issue in because it might affect the overall API design on the connector side +Thanks Ryan for the extensive documentation on the results of your testing +Actually I am not sure proposed change would be faster +They all have a conf directory with a file and three files have an section concerning the jndi namePort with an value of anyone know what configuration situation could occur where all three files say and yet netstat shows port not in useThis doesnt seem to make any sense to me +Also fixed the facilityId for DROPSHIPMENT set create +Patch added +John that icon is the icon for Seam why should that be put on smooks stuffI just grabbed the same one that is being used in the ESB project wizard +The only information currently sent in STOPFLUSH is the view is not sufficient for this case as the joiners FLUSH viewID is the same as the masters dont see a way to fix this without adding the member list to the STOPFLUSH message which introduces a backwards compatibility issue +We solved it buy adding a null guard around the operation but we might want to consider an upgrade to the quartz component +Or it should be included as a transformation everywhere in the API and that is a lot more than this one place + +the reason product dropdown is empty is because certain groups are not setup properly on you for reporting the issue +I am able to build EAR projects with both version and someone be able to upload a project that displays the error? I tried versions through to of Maven and get the sames am using OSX +Patch to update classpath and enable more tests +Number of Readers is less of an issue since hfile +Please see previous the changes in the Properties View are reflected automatically in the Graph View +JMS transport now allows for a to be injected which can be used to change the message +closing issues associated with existing releases +Closed upon release of +If the defaults dont work in the face of a slow chaos monkey we should change the defaults +Verified for closure by Julie ZhuoClosing on her behalf as she yet has no rights in JIRA. +you can help them to prepare their artifacts for the apache repo and get them uploaded through there they dont need to use mavenFor any questions regarding the apache repo repository at apache dot org +Modified the patch to use the module which is part of the base Python distribution instead of the libxml module which is notOn Cygwin which I use Python is at which doesnt include so the support doesnt include attribute predicates as a result I had to break queries where attribute checks are needed and perform them with code +This should make the new code a bit more future proof to any ASM changes +Oops Attached patch to the wrong issue. +Fixed this in r. +Stack This jira should be good to commit right? We can address the edge cases with +Closing. +I hope this can be done in open source hbase and can be pluggableCan you do your research wo requiring that your researchy code be committed to core +Got it +We can take it in many different directions and I just want to open a discussion to gauge interest +it is optioanl because +I want to but could not find an option to reopen this issue +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +Branchx has been created and the trunk next release version has been set to +what we call our Permanent Link +Im now getting this errorError is not a functionSource File httplocalhostLine There is no longer an error on the tomcat log and the ajax response xml is correct +If not please reopen the issue +I couldnt find that docd though +The approach will be available in M +FF screenshot is attached +I dont really know I have to look at this +Removes the one method from +Ok that makes sense what I mean is that you can not ask for the same type with different parsers +will kill client threads if their queue is ever empty Interrupt handling doesnt seem like a clearer solution for killing client threads what happens when an interrupt in received during a mutation? I dont like the idea of indefinite retries pretending that the cluster is never unavailable sidesteps Hadoops own retry system As mentioned in IRC batchSize does not seem like a good value to hardcode +Its a surprise that the master wouldnt open a new searcher not sure where that would get short circuited would have to dig a little +Thanks a ton +Some methods throw due to missing clareification around a new version of the +Ive fixed this for both RC and have you had a chance to give the latest snapshot a try already? It would be really good to know that the fix actually works for you +Actually the easier solution is to pull down a jar and using it to launch your mapreduce job +consolidated latest merge patch may not be correct and might be missing trunk changes and +Ill file a revised patch +Is it against latest trunk +Lets fix this before it gets reported for the fourth time +this seems to be jaxb related and i have no jaxb experience right now +See +All col names are +Ive verified that the supplied patch works fine for me too +Seems like we read that etcpasswd and rewrites it +I committed the patch because it looks right +Updated patch addressing review comments and the +I will take care of it +Yes there was quite a bit of cleanup to do +I think its a general problem with any url that requires authorization + +Sounds good to me Lars +Yahor is open please apply there as you see fit +Close this since its a duplicate of. +LOL! I didnt type that that is a message from JIRA +See +Thinking more about the problem with the wrappers I came up with the following solution just create a of every wrapper that overrides correctly and choose at creation time whether to use the original wrapper or the +Verify and +Do you know whats happen? How can we exclude these jarsbq +Seem not avalaible in latest also not in trunk SVN +When the trigger is fired the SPS will get recompiled +I will hold off committing this patch until Friday +Yuki can you have a look at what would be required to make compaction able to span multiple target directoriesMy gut is that this is not a small change in which case Id rather mark it Later since its a pretty niche scenario +and is from my previous patches on this issue +Ill see how much work is needed for that +I just committed this +Fix checked in. +Thanks + Username ehillman +We have been using metric in our prodution cluster for a year and its good enough for our rapid disk repairment +The SOAP Request without the price SOAP request with the price service Pascal Im having a hard time seeing whats going on here +Work in the guidelines +Closed Issue +I have another patch for that too in my queue +Thanks +Ill play with it but I assume the test is working with the current mongodb master right? Also I just realized that we are currently dependent on but in the verge of releasing where we cant depend on milestones +After a first look in the code this will be hard do to the fact that you have to know the size of your object +The foo component is installed by default and the attempts in the file to override its scope are ignored +If this proves useful to people it could probably be hosted somewhere else like in commonsOne drawback with this approach is how the FSM is setup +This code doesnt include EJB or Web integration or integration with a work manager +I would suggest that this problem should be solved by changing the CMP plugin to quote all table and column names with whatever the method get on the database meta data returns +The reason I picked is that the incorrect length was caused by the server setting the wrong length +Patch against JSON to support discussion on the user list and suggestion from Sandro Ive reopened this ticket and attached a patch against JSON instead +Heres a patch which addresses the issue +I will look at it in the Spark land thoughI will address your other comments and the patch +Anything returned by doCommitdoRollback is always an. +in the attribute value you expect the attribute value to be stored in one lineIn a manifest filename valueis the same asname v a l u which changes the current way of eliminating all CRs and LFs to merging consecutive CRs and LFs into a single blank +Still awaiting a review and a sold vote +I think its definitely worth moving to and consider the impact fully then +When I get a chance to abstract the problem out into its own little project I will update the incident +added issue to the release as requested by Eddie +Fixed in dabdbdabdfffdeb thanks for the patch! +We should expect best practice for a very common cell ACL is a factoring of it to out a CF or table grant to avoid any IO checking cover for mutations +Closing as it was caused by a slow job run rather than the test itself hanging +name a head n +I will have to undo Jaikiran fixes in the test and try with upstream to verify Stuarts fix for the other problem +AhhI went ahead and implemented a random matrix to make things like this easier. +So you have to add the attrs as classes to the markup +We could certainly have a Preference page that captures the defaults for the page but I kind of like the one page config concept for of what we need +So it pulls in all the dependencies again that we would like to get rid of +Net latest code from has been committed +Thanks Vasily and Mikhail +actually i have already tried with git clone and svn check out both then even manual downloadcan you please send command you tried for checkout +I could not find any option in GUI installer which asks User for log you please provide more information in description or maybe a screenshot where this is configurable from user the fix version on this JIRA was incorrect +So not specifying resource at all and explicitly setting it to null is not the same The former is supported by design the latter really is a programming error IMHO +Ok my original idea to let user to supply the obrPath in the resource resolver construction and if user doesnt supply any we use the default local m repo +I didnt experiment with the performance impact in +thanks +The lookahead in the grammar of my previous patches did not work for some parameter types +Fix was incomplete dynamic scaffolding is broken +Will inspect the patch the apply it to trunk +Commit changes to trunk at revision +Will keep it open for some time to gain more confidence +See is due to a new view being issued but both existing nodes assuming that they are running and therefore dont need to kick off the DIST join sequence +This patch removes the wrap group bys in a subselect rewrite in the parser +In the last test case I submitted all of the schemas are related +Hi Dan Sergeythanks for your commentsI introduced a policy provider that can be configured to assert specified policy assertions +No assumptions about the platform default encoding are necessary at any point +I have no commentsThank you for all your helpAvishAvish YeshurunComverseOffice Mobile +As per discussion this patch removes byte block pool forwarding address rewrites by always allocating bytes at the end of each slice +is set to is there a reason it can not be upgraded to +Thank you +Ill check out and test your latest patch +Unless you have specific questions on this code lets hold off on discussion until I can post this document hoping it gives us a more constructive way to frame the discussion +Verified fixed +Personally Im happy with the current slicereplica terminology in the code and I dont much care if it matches the external doc terminologyWe have agreed and disagreed on all these points in the past and Im sure we will again in the future looking forward to the next dust up but I certainly dont think its worth breaking all those apis to change the names in the code now unless its a couple minor consistency issues +Resubmitting just in case +Committed to trunk +AndyI will look at in more detail +I suspect that you have a different bug +i didnt realize the wrapper existed for a single request only +One is esbcontent the other java classes +Will post progress and comments here +for the patch +Assigning to because I believe Steve wants to get this work in to next release +It may be that the standard did a poor job of integrating functions with CHECK constraints + backport as this corresponds more directly to the published artifact than java packages +Lukas any updates on this one +The release note for covered this issue too. +Fixed at r +Discussed that we will add a further openshift option in the Server Behavior combo Local Remote System Deployment Openshift +This one should be right +Numbers in last comment are in ms +Added seeds to output +Just deployed a new SNAPSHOTCan you try again +If someone wants the new behavior the need to change their configuration +I dont think we need a workaround for this +Retrying just committed this +Ive just committed this +looking forward to the link from the English docs that would be better for Chinese users +Fixed in IIRC +This needs to be called in two code paths when the model is loaded from an and when the model is instantiated with various parameters and artifacts as usually done to create it in the first placeIn the later case a different or no validation is performed that should be moved into the validate and then used in all cases where the model is created to validate it +forrestwhiteboardpluginsresourcesstylesheets + +The needed colorspaces are used in to convert colorvalues from a soucre colorspace coming with the to the sRGB or CIEXYZ for displaying and printing the +Ill check this in tomorrow if nobody has comments +pushed to a new client and pushed to. +It overwrite the previous unanounced release. +Attaching a patch for should anyone need to resolve this issue for an existing systemIt is different than the trunk changes because the Queue was rewritten when moved to the new module during development. +We dont store it as BB anywhere we just convert it to the BB to pass to validate method +I just committed this +Weston thanks for the pointer +Attachment has been added with description rev of to latest PDF transition resolvedfixed bugs to closedfixed +Bulk close for +Closing resolved issues for already released versions +The loop predication bugs in the original Java release have a chance of hitting some of the optimized loops we have in the binary decoder or encoder +patch and shell to apply revision . + could merge the trailing headers into the aggregated +The programming parser of is different than the stuff in +My apologies old version got into the test. +Does the dfs shell pick the user from environment variables or from a conf file +I didnt migrate the artifacts since I didnt see any previous releases. +I have been testing this currently in a sorted CF but this only gives incorrect search results when there are capital small alphabets mixedBut thats the rub isnt it? If we need to slice we have to do twice the searches sort in memory and return the result +There seems to be no easy way to fix the names of the partitions directly in hive + +So in any case is the repository problem only appearing in pax exam or could it also hit karaf in production +Looks like Dag has a promising patch and I am out for a while and dont want to hold up progress +reassigned to myself for further research as I am not reproducing the problem any more at the moment +Inlcuding pack selections and all user input +Bug has been marked as a duplicate of this bug +Thanks! Ive been wondering why we had trouble changing the ports on this. +We assume that there is a single client writing to a ledger +Just having a look at the implementation hierarchy of +During code generation time put use flag in to the +Adding link to related deployer issue basic idea will be for undeploy to wait for a configurable periodof time while some deployer is in deploy for that deployment +Thanks for digging Uwe I just wanted to be sure here +I have to check build helper for this task +these patches are mutually to Dennis Byrne for fixing this. +Hello VivekHere I am uploading the patch for the functionality to create a pick ticket from order view screenSteps for testing the functionality +This is the regression introduced at r as the fix for +Its checked into branches now so it is possible your change and that one are both in place +I updated the wiki and since it has the most complete information I think the Glassfish Installation Guide could be just a link to the wiki page +Also can you comment on the frequency with which youd observe this spurious test failure without the patchThe patch looks good to me +Please commit latest release or a KEYSREADME file to distreleaselucene then I can continue +I found this patch helpful but inadequate +fixed as part of porting the data directories to structure +Closing the bug as I couldnt reproduce it on the latest build +Patch should apply with fuzz +I assume that the test is broken as well +Revised as per comments from Todd and Nicholas +Hi John Please try the latest released version +Let me add the address +The Patch fails when applying +The trunk patch applies cleanly and is from Hadoop patch for patch for timeout to timeout to test and following findbugs! I had forgot to sort based on compressed size +to patch +What I committed. +I guess we still cant reproduce it +Committed the provided patch unchanged +This latest failure looks like +A quick thought Maybe an alternative to sending through the original date in the success method would be to just set the lastmodified date for the resized image to be the same as the original +closed +Its better to use we should encourage use of that utility throughout tests +content +I can provide more information if needed +Should we file another jira for itI was working with patch for which has directory specific fields I saw for the first time an EOF error while loading the fsimage +So then what is wrong with PROVIDED the way it currently functions? The dependency would not be included in the war because it is provided by the container and project C would be able to use the driver at runtime because it is provided in the classpath by the app server right +For that I would use the delayedwrites option of the Erlangs file moduleBut i might be wrong +We should probably look at the site structure and separate out the guide such that svnpubsub can deploy them separately and make it a better experience to maintain the website separately from the guide and vice versa +Thanks Amar! +Rob looks like we have some unguarded codeMaxim does this really require meclipse etc +Heres one with tests +i dont expect we will implement this any time soon if at all ever +Please add a different issue for your feature request. +pinging this issue to see if anyone picks it up again Looks useful +Eclipse IDE metadata changed in relation to this. +Approved for. +Thus moved this and other issues from target release to +Moving this out as this is an specific class and not part of the tried to change the test case to initdestroy on a test method basis and Im not seeing an out of memory exception anymore +Ill take a look at it +Sorry the code is from branch +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +bulk close of all resolved issues. +bulk close of all resolved issues. +Lets see what hudson says +Thanks again! +If you can reproduce this using some version of the unit test please reopen and attach the updated testcase +get returns the correct value +The known issues section still needs clean up +Patch coming soon +O I forgot to mention fixed in +consolidated patch +jsut asking the to load a class first is no option since the system loader may have an jar we dont want to have +Proposed SOA TP Release NoteErroneous trailing blank line from input data file for transformCSVXML has been removed. +If you dont have the tag interface how youll decide which public API records are serializable +I think we can achieve it by suspending readwrite on all the sessions +Can someone please confirm? +Finer grained control of this functionality will be added in a later release +I think we can find a solution Im just not sure what it is right now +Thanks for the contribution +I will look into them and also will try to implement the and will upload a new patch +Ill be testing this with other databases over the next few days +I looked at it I dont think thats your diffs fault +It looks like a good solutionJB why did you envision putting that logic in rather thanThe proxy is loaded by the system bundle so we should avoid adding any dependency to it +Please add any additional comments to a new JIRA issue with a fixin for a future release +Convert to Java standard without breaking unit tests +The unit test case that expose this issue along with the proposed patch can be found in +All help has been removed from the example schema as help has been added to the cli in +A fix has been pushed to and will be in the release +Committed to trunk thanks Ning! +I already have a dedicated commit in +Closing stale resolved issues +To be reopened if the fix is verified to be +spec file is currently correct. +Looks good I had some minor commentsInstead of making property file dependent on system property would it be a good idea to make it a configurable parameter with the default being in the configuration directory + attached patch +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +I could use the test but not the patch +Patch attached with comments incorporated +I had previously had the usernamecredential for the made available to the LM through the LM options only and not via callbacks or stacked passwords +Will commit soon +to and trunk! +Can you reopen it? +How about regression test +Same patch than Bruno but without change on main auth and a few labels in in s a good way to do things for the webtools webapp because it is a nice introductory page that is always there but I disagree with having it in other apps +Yes the big F was a typo +If a search term is being used to filter results then it needs to be displayed in the Search if the search term continues to filter results when I change whats in the View dropdown why does it not continue to filter results when I click on a different resource category +I detect an edge of something in that question but Im going to take it for face value +Empty string not handled correctly in JC so handling it here +I have tested it lightly mainly through unit tests +Thanks Ravi! +Since the bootstrap libraries are hardcoded in placing the dependend jars in JBOSSlib doesnt have an effect +Thus Im opening this issue up to any objections for closingresolving it +Done in trunk +This doesnt seem to be a problem in Xerces +if this is a requirement of the plugin the plugin should declare the the tag requires and the plugin lists junit as a dependency but its not in the root classloader +This functionality is really important lets write Warp test first + +Remove unused variableslabelsmethods in packages +Possibly Use local config? That may help reinforce the distinction between the local box and the remote one +is the jira for that. +Ill later check if it is because of some nonsupported operands or because of a problem with pdfbox +Thats why I think it would still a good time to make this kind of changes they are necessary for rendering and data wrangling in general and we should not be breaking a significant portion of users so it still looks like a good time to make this kind of change +I still have this problem using +For us the problem was fixed with the final version we actual use +I have ran and verified the patch in our cluster a while +I thought a JIRA had been field for that but I wasnt able to find it at the moment +It should make no difference for clients either wayIf you make a proposal patch Ill be happy to review it. +Marking as complete since no outstanding comment +did a clean build verified the test and source are in the proposed directories +I have to learn jQuery I will have a look at your last patch though +I expect the read and write transactions to protect their activity against other suggested patch in the previous comments resolves the problem reported here +I dont think theres any risk +It is also worth noting that even if you could get an asterisk into a phrasequery it would not function as a wildcard +holds a reference to a instance as a primitive type shares the same info in class regardless its possible parameters +How should users edit this if it does not get shown in the editorIf we had a way to toggle rendering of a component without changing the attribute then it would be fine but as long as we dont have that then the current behavior sounds like the best comprimise +Do you think the wrong order of the dependencies in the classpath are the reason for our problem? I just concluded it from the changelist of this plugin. +I would expect the test to fail if the close cant complete cleanly but we are really not propagating up any error in the case some component doesnt shutdown cleanly +There is no tie what so ever to any page +Closing as Wont Fix for now +The callback changes seem like an improvement +Fixed with +Meh + +sample multimodule in r +Basic mode customizeA user can define which pages he want to appear in the basic mode which considered as a shortcut +The rule to check that switch statement ends with the default clause already exists javascript +Im making really heavy weather of this Flavio thanks youre completely right +Committed to master + to trunk +I think we need to do the same I did for prototype +Canceling patch while unit test failures are resolvedIt looks like cacheString and cacheString arent getting broken up as they should be from xargs +only to run the tests and not for the distribution jarsRegarding the failure I havent tested on Windows +Why cant you use the file property +Rocker if this bug is DUPLICATE change status please and add link to original issue. +Hi RickThanks for the help Ill be working on this again come Monday so Ill give it a try +My batch lookes like the followingTOPnet start activemqping n nulnet stop activemqping n nulGOTO TOPI think scott restarts the broker more often than me +Committed for in rev +It may happen thatlines fall on top of each other +I think its ok for wont get fixes as imap and mailbox are now seperate projects +When you compile on it works fine +Tag resolved issues for already released versions +Looks like a dup of +We have a default value that gets used when the user does not set a value +If you have any feedback before the release it would be most appreciated +Now to track down how to detect large files and use a temporary directory instead +Solution This is a matter of timing +Thanks for that analysis Knut and thanks for the extra info Myrna +Youre trying to compare two unrelated things +If we want to discuss a new HV specific feature for we should do so in a new jira. +Description done +Name Zipsort Zip returns gridzipsort Name returns gridzipnamesort Name returns gridzipnamesort Zip returns gridzipnamesort Name returns gridzipsort Zip returns nothingThis way the zip can be sorted and name can be sorted independently as a secondary sort inside of the sorted zips +The exception had this information but was not propagated +thanks Pascal +Done +Looks the same +Thats supposed to be a minus then capital O on the sudo wget line not however this page decided to render it +In this case maybe contribI dunno I think I would be for it if I was willing to work on it and maintain it +IIRC building libhdfs is a manual extra step +Simple fix +Moving to. +Attaching the patch taken from my patch +bq +Alls fineThank you Alexei +Might be a good idea to split things up into a JPA specific part and one that could go into commons similarly to the XML namespace implementation +shows the output of the fix +Another option would be to create a custom JAXB javaType binding in the JAXB custom binding file and specify a parse and print methods as well as setting the has JAXB attribute to true so that the curent could be passed in and retained in the generated Java objectIf desired I could demonstrate either of these approaches using a recent revision of the trunk prior to the conversion to +KarlI tried applying the patch to our custom build of but it didnt apply cleanly over my own patch +Thanks for making these changesThe nodes in and correspond to scheduler queues right? My suggestion was to add Queue to the names to make this correspondence more explicit +Could use a little more information +Attaching patch for trunk +I vote for not flattening them on the basis that they arent collections +Linked which requests to have issue was migrated to +There is a regression in the latest patchIt does not work properly in a environmentIt seems that the values of the counters are not properly serialized in theWe need to add a test and fix the bug before committing the patch +for a year or so so it does become a race to get all the features you want to see in the next year +But false positives here could result in replication storm +can this issue be closed or does it need +I can work with the improvement but it take some time to fix the issue +our custom indexing chain to handle indexed documents and then register a codec to intercept the code path for norms and CSF but this ends up with customization hooks for the same indexerThanks! Simon Was wondering if you had a chance to look at thisThanks I looked at the current code and it seems that everything delegates do and I wonder if you cant just use a for this and extend it in DWPT +NickIve checked in your patch +This patch checks in the version of the release notes which will be packaged in the first release candidate +Actually I forget to upload all files that must be patched +Im experiencing the same problem +Mostly saying this for myself so I dont forget what to do tomorrow +Changes look good to me +Change looks good to me! +Thanks that is great to hear +I have next started looking at how soft upgrade should be handled +Added patch provided by Thierry Guerin +Fix committed +And I strongly believe this is the right way to do second pointI am not saying that this shouldnt be generalized +I think its fairly clear that false is a better default +If we choose not to introduce a dependency on cgexec then we can return to modifying the C code in as the previous version of this patch didthanks Ill ask againCan you please provide a simple writeup? Im confused seeing new interfaces in every new patch +Could be if deemed valuable enough but closing for now +I only realized now that Pat added xdoclet jsp tag to generate tld files so attached is a patch for that use such tag in the new field added + +Do we want to log the exception as a string or do we want to pass in the exception as an argument to the log so we get the full stack trace? In the case of a DNS outage we may not want to fill up the logs with useless messages but if there is a different error happening I would prefer to have more information to be able to debug it +Continue following for more updates. +Adobe lets you catch it +bq +Would it be funny if a root cause analysis of the explosion of a prototype JSF fighter that killed one of our military personel and cost billions was traced back to a coding error in and introduced by yourselves? Yes you are dead right people do kill peopleThe very reason I hound people on these sorts of things is because yes the devil is in the details +KonstantinThanks! We got it working +Thanks Lahiru for the patch +simple also takes sout reference to DB JDBC driver +Andrew should we move this out of? Do you think well get to it in next week or two +As per the general notion that JNDI is part of the Configuration which will get scanned automatically +It might be good to open another Jira Issue to document this changeKathey +metadata +The only issue is if too many mutations queue up you might OOM but this is the same problem we currently we have with the write stageRight +Updated excalibur store package in Cocoon and +Whenever I combine these it never works in a way that is completely understandable +My patch failed to remove the import statement for +Yup this is a bug in the way we load the certificate chain file +This patch also includes changes to the the help text in bin and adds tests of delete and deleteall to binAll tests in bin passed +Fix the example for selection attribute of scrollable +I was confused with comp name ejb name and jndi name in my post please read comp name and jndi name +Hmm +The task changes the content of jar so redistributing it with modified data is too far from my legal comfort zone. +Other issues may require separate ticket. +I tested it against Joans and Marks Python scripts but I dont have junit so I wasnt able to run Curts test +Fix will be in the next nightly snapshot. +since is closed for business +Ill investigate +I first thought it was only for gridmix unit tests but the problem might be more pervasiveCan we check that the values defined in srctest are actually being read and used? Maybe all that mavenization messed things up +Thanks to Michael Kurz for this important patchregardsMartin +I think part of the confusion here is the terminology +What I had wanted to say was like the comments in I am thinking about how handle the package overwrited the class oeverwrited and the overwrited +So Im not sure what apache would prefer to do in this caseAlso more generally would returning the attributes as a list be sufficient or would one need to also return the data from the other two elements? Later I can check what is returned for amazon +I tried full clean and rebuild multiple times but they did not fix the problem +I think ideally we would not weight or score MUSTNOT or clauses at all +Hey guysI think we should mark it as a new feature +Patch for assembling webconsole related buddles into distribution web console is not included in the assembly on purposeI think a work around would be to add the apache snapshot repository to the list of default repositories instead +The attached introduces the new methods and event flag for the swing and render modules +They should make the Eclipse M release +Other thought is that the logo should somehow portray stabilitySome wordsphrases I associate with Mesos core elastic kernel of a dc flexible multi frameworks +Or can eclipse tell that those deps are already provided by the plugin I didnt think it could do that +The checked attribute is for overriding this automatic detection and forcing it to be either checked or not regardless of the Graeme re your comment The checked attribute is for overriding this automatic detection and forcing it to be either checked or not regardless of the value This is really confusing because just submitting a valuexxx makes the box checked + +on the side +What do you think? If there is a reason not to require it we can keep two implementations of +Maybe Sandy also mentioned this problem on another JIRA +Stuff in lib are usually meant for C libraries hence Owen suggested to put jar files in share +Thanks Ari I just committed this. +Is the file from the project actually deployed on the server +I actually did find two more files the readme docs in java and have added them as deletes to my client +Yes this issue doesnt show up with the branch just prior to commit at least it is not obvious +Were trying here to just printout the problematic exception and then move on +I have a need for the exact same functionality and this could save me a lot of time grokkingwritingdebugging etc +Now that I think about it you wont be able to define keyspaces in the yaml in and we dont want to break existing configs in so maybe we should just wait until the cli becomes the One True Way +I think the simplest solution is to remove the lazy copy approach +Prefer doing this in a separate jira though this one is more of a refactorbq +Doh +The changes is more of a feature change and should be done in a separate jira +Resolving this issue for now +I committed some improvements Java DSL builder method just like the DLC so you can create transaction error handler in neat java syntax Added missing builder method for new option in DLC Added TODO for the bug above with the delay Improved unit test for wiki the example is actually bad as the direct component doesnt retry as its backend system hasnt a transaction manager such as JMS +A patch was attached for fixing these two test case failures +SureWillem will start the release beginning of next week +I have a doubt here you say you throw a fault which is a throwable but not an exception? Usually this would mean it inherits from Error instead of ExceptionIm not sure if this is a good idea to catch Errors like OOM or suchCan you provide a bit more details on your use case +It is a slight modification from a quickstart project distributed with Jboss Developer Studio +The tasks in time out one year in the future +Indeed its a regression from an earlier implementation where I used +Apparently they were abolished back in +Can we do a rebuild in mead for EAP even if there is no code change to resolve these differences +This bug differs from +My wire dump looks nothing like yours which has some really strange things most interesting is the packets from port +Issue closed with release. + +Nothingthere are following additional properties +It seems like a more logical choice to meThis patch more or less implements the suggestion made by Archimedes Trajano above +After giving a much deeper thought to this work I came up with the following class hierarchy which would be need in the scenario +codes are committed +After pondering this after the original patch was committed I realized thered be some confusion between field values and fragments +But Im not sure what to do backwards compatibility wise +looks greatNit the line where you pull out the nested exception in could use a comment indicating why this is necessaryAlso wrt your comment that its not kosher for to be using the classes file a MR jira for that +Its even possible for one goal to trigger another goal +Hi Shai here is an updated patch that achieves the same thing now the tests dont create redundant Shai i got myself confused and thought you were trying to cachethis patch is good in case a test has multiple methods +I think the original trunk patch did have some problems +Eclipse product +If you find further issues please create a new jira and link it to this one. +It shoud use toString? it does for declared types with imported beans we have no control though +this seem to change the default for the ant tasks which isnt the original problemThe original problem is that in the IDE it comes as a surprise for some and I agree we should make the default from wihtin eclipse to not export directly or at least add a warning when running it +Looked at the data in detailThe tables should besmbbucket v v vsmbbucket v v vSo the above query should result in rows both the results are bad I did not see the entire results so based on what Joy is saying it does not work in minimr did you use for your query? can only work with +The order of the configured namenodes does make a difference +Well indeed something like that works better +Seems to work fine with Groovy and Spring too +This patch adds all the functionality and unit test +Attaching new Yahoo! Distribution patch +Hi ParimalPaul who sets the requirements is busy now he will answer you when Information on this +Applied to rev +this is a great feature for HDFS HA but it was Fixed in version? I hope it can be released during the first yarn stable release. +Heres a patch for this +Richard if you are not working on this could you please unassign this issue +If the data and S support it we should upgrade +These were present in which was unable to find them +Hi PierreThanks for report but your patch has a problem +The first two issues are more difficult to accommodateIm going to close this one for now please open a new ticket on the reporting of modifying configurations via trafficline. +See. +Matt you wanna take it? committed slight modifications of these patches to trunk a couple of days agoI included in the first line of the commit but its not appearing on the page presumably because of the infra changes +blocks deploying blocks jBPM resolved and verified in IR removed from blocker list +If possible please post a unit test exhibiting the bug that youve pointed out +Unsetting Fix Version on unassigned issues +Thanks +Am I missing something +from me but perhaps ammend the commit message +We can apply the patch once Commons Compress is available +push to +Porting to will need some merge workClaudius your patch wasnt complete and had some design problems which led to a complete rewriting. +Not sure my approach is totally in line with the Tools manual so I need some feedback on itIm clearing the patch available flag and also assigning myself to the issueIf someone wants to add to the content after the first patch is submitted feel free to assign yourself to the issue and bring it forward +If you dont have an interest in pushing this along further than I could invest some effort in a couple weeks or so to provide missing elements that should make it amicable to committers +Fix the failing test case. +Darn sorry +Done +In the future it would be great if you could submit this in smaller more manageable chunks +fieldName houseNumber notNull sizeMin sizeMax add field string class +should be still ok +It seems to be working as it is supposed to even with more complex Accept headers +Please provide a patch file so we can easily review your changes and apply themI did didnt I? It should be attached on the jira issue +If youve got data coming in thats a gigantic decimal and youre allowing that through and then youre also trying to turn decimals into strings +We intentionally dont support namespaces +Bikas my question was more along the lines of what operational state youre looking for here +Attached a patch to add headers to some files that were missing it according to ratI couldnt get the rat plugin to run automatically on the first try so the patch doesnt have any pom changes in it +I just committed this for now +Adding the generate flag to your first call of the source generator should do the trick as this way Castor will be able to resolve imports as a result of importsincludes itself +Any other issues can be opened separately +Applications that desire format flexibility can use the other interface +This looks to be the same issue as reported in +Tail log on master to ensure that the master is not doing anything +Is there any new info with respect to this +Found em pushed them into archive and markmail. +Thanks so far for looking in to it +I think tests should be fixed +this patch should be applicable for latest version +Can we plan to handle this +jmsconcurrentConsumerstransactedhttpNone options is commonly usedjetty JMS component have a lot of options but most of them are for fine grained configuration +Finished cleaning up code +Great catch guys! I also would like to know if you have a reliable way of reproducing it but it seems plausible to me that it can happen +So the folks would be interested in solutions that allow for this +It adds some contrib changes and and javadocs + for patch +applied in master I managed to reproduce the issue and I think we can consider it very minor it happens the very first time you open the generic wizard in that workspace doesnt happen if you opened the wizard from central first doesnt happen again on subsequent tries doesnt happen again after you restart the workspacecan you please open a new JIRA +Thanks for the further discussion on categorizing the plugins please see. +This patch only handled unknown topics and unit tested it for all types of requests +Could you please test this minimal patch +Needed optimization for Netscape +This feature needs to work correctly in the case of concurrent compaction request +This is especially bad because if executing via a Hive Server it will cause the service process to crash +Agreed this isnt a CDI spec issue +Attaching patch to make SBAP mimic the old for workaround If the problem is observed in hcat commandline then instead of running hcat e show databases run hive e show databasesIf using webhcat in secure mode then theres no way to get around it using hcat then in that case the suggested workaround would be to disable authorization or SBAP on to get around this problem +So its no surprise that we dont have a test for that case +It seems like old nabble has been down for quite a while +Youve combined a service factory with a Simple server factory +The source stream is read twice first for seeing whether source stream isis not empty second for streaming it +Michal Can you please clarify was this a bug with your configuration or the stock configuration? Your comment makes me think the former but from the fact that I see a fix version I infer the latter +Right +Also changed to work with new API but is not a part of the patch at the moment seeing some weird issues with the RS after applying the scripts Trying to contact region server for region unknown row unknown but failed after attemptsExceptionsfrom orgapachehadoophbaseclientin getfrom orgapachehadoophbaseclientin nextScanner +I think says you can detect changing tabs and set focus back to the SWF +for the trunk patch +Ill commit it today or tomorrow +Thanks Dan thats a much better suggestionThe only thing is the propOrder doesnt seem to take effect for the test cases which are failingSo with no change test test and test failWith the propOrder specified the same tests still failI know the propOrder does have an effect overall since if I specify a propOrder of name id then those tests plus a whole bunch of additional tests failIll have to dig into this a bit more to see if I can work out whats up +I want to include it into upcoming s the patch that works both for branch and trunk +I like the idea of the progress monitor tracking the errors +If you dont have time someone will get to this eventually but getting it now would definitely be an appreciated contribution +I lowered the logging level to INFO which is the current default value +Yup that seems to be reasonable way to go about it +The tool didnt find any problems in the and javadocs +Regards your suggestion would you still be able to pass a custom partitioner +bq +CP fixed in revision fixed in revision +updated estimates +Create a new page and add the portlet from the categories remote or or Producer WORKS +Hey Aaron +dropping priority as we now use Maven cli version as default so this is only when using embedder build +Were getting errors because the protocol maintains persistent connections between httpd and tomcat and tomcat has limits. +doesnt apply cleanly because of other unsubmitted patches +Thanks for the quick turnaround will save me some time +Thats the whole point of optimistic locking +This is the error when the disk is full +DoneRunning Tests run Failures Errors Skipped Time elapsed secThis mainly happens in which consumes the from the fragment after the host got resolved and pushes that to the merged +Dup of +And committed. + +I cant reproduce on XP maybee i missed some updates I will update and try again +CorrectionIn positive case there will be deployed two beans and negative case there will be one more deployed bean s see if Jaikiran can duplicate this and drill down on the actual cause +Committed revision Excluded file as it was only a change in spacesPhilip its just a little easier in the future if patch files are from the root of the tree not aThis allows committers to applie patches consistently without having to figure out where the root isThanks for the patch +Patch to IDE to work with new mojo +gateway support is covered in separate issues and + +I have a pull request ready to fix it but also have some doubts about it. +It will likely require work from the Mojarra team Since the situation was improved in Im not convinced that this is a critical issuePlease feel free to discuss and debate about what should be done when and if more information comes to light +Since that might be useful in the future it was left in although not currently used +i missed this one +from what I see via the jcr tests this is good be fixed now please reopen if there are problems. +Here this has a rough interface definition on and SCA component interfaces +Patch attached +The behavior of is correct +This way setting the CLASSPATH individually for various scripts should be easy +Closing the issue. +Fixed Polish calendar translation +Further work on it will be covered in new issues. +Reducing priority because it seems to me we didnt have many complaints on this topic +I am not sure what else is needed but it seems like using standard API is the way to go if possible +Once i confirm that this issue is fine in Linux then i will make it as minor +Then I went back to the st page switched it back to remote mode saved and checked the deploy folder option again +It removes the partial file in the destination if rename failsSorry that I did not make my proposal clear +Great to see this is fixed inDoes someone have time to try this on the latest which is still in production a lot of places +To get the exception it is sufficient to deploy the ESB app deploy the process start the process undeploy app deploy app and finally start the process the above steps in the description are not necessary +AFAIK the projects build process relies on the standard plugins +Should I provide a patch for correcting the problemThatd be excellent +Patch committed to trunk +Patch file of code only still working on regenerating test logs +transition resolvedfixed bugs to closedfixed +Todd so Hadoop already has a similar webpage? Also one of the critical issues is adding so we can also generate a list of the config values read and what their default is set to +This could indeed be solved by an interface +If you find further issues please create a new jira and link it to this one. +I am attaching a patch which adds a getter and setter onAdded a getter and setter on to set the isolation on the endpoint +I think these two solutions makes sense code coverage per module +I tried to iterate on Simons patch doesnt modify offsets and emits all of a single term at the same position uses a sliding window +where is snapshot +The tests should be excluded now +Mass closing all issues that have been in resolved state for months or more without any feedback or update +this issue is still in progress since message boundary events are not yet supported +Will flag that for newbies +Thanks +Rob any chance to investigate based on it +I think the Login button on the right side off the application navigation bar is for requesting the Login screen +Then the Import Into field is being empty and the wrong error message Selected logfile already exists +Thanks Navis +CommittedDIGY +I just kicked off it and ill see how it goes tomorrow +In any case Ill look at itI was probably just not paying attention +binding is absolutely useless +Abul why do you need to maintain page markups? Its perfectly ok that only has associated markup +Patches were committed at the revision r +Committed +This generates graph by recursively iterating through dependencies. +Thank you for the submission +since its +Im going to look more into this but just wanted to assign it to myself since Im interested in fixing this now +This is the last drop wed like to make before working directly in Apache +Patched pushed to master +Excellent work tested the new Spring context loader component and it works great +This publishes the app and you are in sync +only for adding version added a reproducer to Metamer httplocalhostmetamerfacescomponentsrichAutocomplete was in +For DB initialization at deployment time in should be changed to +Patch already applied +Resubmitting to see if Jenkins is automatically picking the patch up or not +Ive committed this and merged +Moreover when I run a test it still looks like the LHS of r did not go into the LHS of r +this is integrated on master only +attached tests locally and Billy the reporter claims it fixes the reported issue +Here is the final patch for trunk +Fixed with rThanks +Surefire reports from failed tests +packageshivemetastore +I have looked into this problem but I lack insight on how to fix test appears to be running successfully now +Thanks +After some consultations with Sleepycat I have determined that the cpu hogging Ihad noted was not due to the cache preload but happened when the db environmentwasnt closed properly a consequence of other than normal servlet containershutdown +The real problem is that doesnt throw the expected instead raising an +I put the errors screen secuence +Thanks for the feedback Ashutosh if this is something youd consider adding I can update based on current trunk if things have changedWhile there may be perf gains this is needed to integrate with our compute grid +Id be stoked to see this happen and help out if the licensing issue is now a. +Attached is an extended version of Martijns initial patch +The instanceof is NOT a Runnable instead its got a proxy inside of it that is unwrapped for the Java method call and then rewrapped on the way out +Also add support for db +Either dcommit hasnt finished applying all the commits or it broke +But reality is different. +Lets add Checkstyle to the build before we commit something like this so the changes are maintainable by all +I think Dan can have a look at this issue. +Closing since the mapping is arguable to begin with and there seems to be no interest in contributing a fix. +I also use the Character static methods directly where possible without the indirection +hadoop archive archiveName p tmpsomenonexistdir somedir userknoguchinull echo After the patch hadoop archive archiveName p tmpsomenonexistdir somedir userknoguchiException in archivesnulltrunk knoguchi echo I guess we should also fix the NPE when src doesnt exist +Based on the below text whichcomes directly from the Jakarta project site this is supposed to be themechanism to request feature enhancements +Any update on the junit test case +Felipe unzip works well but jar dont why dont you try to use jar instead of zip? I can do so but itll be by time it worked I suppose you used jar +The patches for viewing will be +Hi IndikaIs calling the setProperty method on cfgCtx sufficient when storing these parameters in the cfgCtx? Or do I have to get the and call addParameter on itThanks IndikaI believe I have fixed this correctly this time on the trunk +One of the problems in appears to be a builddistribution problemFor some reason is missingjgreenedraught develserveralldeploy unzip l grep c Authjgreenedraught develbuildoutputserverdefaultdeploy unzip l grep Auth orgjbossnetaxisserver orgjbossnetaxisserver was code changes in since +I think you will need to rethink your solutionRegardsFred Preston +Adding and to support release +Doron I havent really been able to keep up with the way this issue has evolved or dig into your new patches but to answer your question about the Ord functions yes they are very useful and it active use in Solr +I will look at the tabs and spaces issue +Ill deal with the web archiveeither over the weekend or early next week +I have tested creating PDF files using and Save as PDF plugin in +Hi UmaFor the first method +Excuse me I mean in cases where you have a collection that is populated with a complex object +This is already fixed beta +brian you beat me to it +bq +b Do not remember the events during outage +A reference to the clients datanode descriptor is stored in pendingCreates +Would be a nice one for someone to get started on. +Adjusted the subject to better reflect the situation +If you are using a bindings file everything is fine correct? If you do not use a bindings file you are getting +but i cant figure how it may end up being singular given that i generated it randomly +Release Note The was not updating the address of a message when the initial EPR failed but a subsequent one was successful thus the sending of a reply message would fail +Thanks Shameera Im assuming that the defect will remain open until is release and it is verified on the release? Im not sure what the Axis policy on defects is +Yes +Thanks Knut Anders and Oystein +MR AM recovery should already handle that case today if the staging directory is in place to allow it to launch so we do need to fix that +We should handle it under the xstream databinding +Feel free to reassign it to someone who will provide the testcase +Updated patch makes ssh tunnel work makes changes suggested by comments Craig +need to copy files from esb build and also modify ws allow the schema to be located it should be possible to take the schema folder and the from the and add them to the directly allowing for xmldsig to be term it should either be requested that this is added by or an extensibility mechanism provided to add new schemas to +Hi we had the exact same issue +Your patch only works for trunk and +Moved to please move bugs back to which you will work on in the next weeks +Gives each map copier thread a unique id +Tests pass +Committed to trunk +Does not seem to make it for going to do this costbenefit doesnt make sense +Attached is a modified tutorial example to illustrate the am having exactly the same problem using and EJB setting updatefalse insertfalse does not help since I need to insert new entries is there another work aroundDave +In the example in trunk where there are two mvel rules +Thanks +Thanks Patrick! +Some how related it was mentioned before the patch allows you to override via configuration settings an impl defined as service +Hi nkeywal I think you should make a review for this jira so that we can compare the old and new code +JohannesDo you have the time to commit the changes or should I do it +be careful the project root must be added to the list! +AlsoI am still interested in a real life example where this is actually useful this would really help people understand the motivation +Flavius Please supply a test case that reproduces this behaviour +Im not sure I follow +This patch introduced a bug that prevents the first searcher opened from ever being closedIm looking into a fix +very weird +Better reading code? +I got the latest build +Perhaps the intent of the patch would be easier to follow if I attached the patch that I was actually using Ive corrected my attachmentAs I said the fusedfs binary does not appear to have moved yet in the version used by Bigtop so this really isnt all that useful yet +So I thought they might not be as surplus as they first looked and went ahead and applied the original patch. +I just committed this +reopen for setting proper fix proper fix version +I added the bundleContext lookup to jetty and Rick added it to tomcat +What is +Assigning back to Max +Thanks for the what OS are you using? Because the only part that I can imagine still failing here is that an update of the file is not seen because of timing issues of the OS +One techniquewhich is artificial is to deliberate modify the code for rolllback so that it fails +Contains the correctly formatted canonicalized SOAP response body with the empty namespace declaration and attributes preserved +Additionally required upgrading Jackson to and some changes to make builds work on mac. +Thank you for the patch Harsh. +But actually it may be worth dealing with this at the same time we deal with +IIRC property had something to do with it +Verified Hence closing this bug +Test catch this is a regression introduced inCan you make the testcase more easy to reproduce the problem? I mean if without the change in this diff should get an error or incorrect results when running with that testcase +Can you rebase? Sorry I didnt quite get to this in time +Build packageV +Entered for the issue discovered by Kathey in +Hadoop Tika the community uses the wiki extensivelyAny thoughts here about potential improvements or other Nutch specific documentation we need to add or need to update ASAP before we consider this issue resolved +Babak thank you for working on this issue +Great +Superceded by times needed this again today and in its here +Per I can confirm that we dont need the following classsrcmainjavaorgjbossarquilliancontainerglassfishmanagedIt seems to be a relic of the container implementation +Removed the changes to and changed back to an Activity +ve made a commit which applies the same jars as previous asx streams until we can get this worked out +hi Alexanderthanks for the test case! Ive included it in this updated patchAndrew +in +OK pasted SQL included queriesFirst query is absolutely correct SQL +So introducing a configuration may introduce compatability issueI filed a similar related issue +Yes that works +But same problem at the end +I suppose this affect jboss as to dependent issue +provides a set of guarantees and the auditor complements that set of properties by healing the replica set of ledgers +Got to ask Adrian a question +This behaviour is back when using the release plugin in tandem with +change subject from to +Javadoc now looks good +We changed our configuration and specified a high memory limit +When converting the pig tuple into an hcat record we apply conversion rules on the way out so that our hcat record and hcat schema matchI believe if we follow this approach the schema and records will always match and we can avoid having to keep track of original data types if fields have been converted etc +Right +Thanks Areek I just committed thisI just had to fix one assert in test it didnt hold up under beastingAnd fixed up a few things on backporting to +Sorry I missed the rebase notification in my inbox +Hi JacquesNow if you want you can do the changes in POS for the same improvmentThanks MarcoDone in I just wonder if we should keep reasonComment in since in POS its redundant +Thanks +Im not comfortable trying to change around jitted methods class names this late into and since we have only a handful of IBM users this is probably not worth the risk +Sbastien what is the size overhead of including in installersIs there a way to produce a version of? Or do we have a way to make the inclusion of only if your feature is used by an installerThanks +Woops the new testcase is failing on windows +Integated help documentation now generated via ndoc nant task + +Trivial patch to allow chaining of sets in DAG and Vertex +We should raise it on dev list unless good reason for keeping up +I think current behavior works as design +See how how this is used by the users and enable it either as stated in the first bullet or remove the restriction altogetherIf the current restrictions takes away some use cases so be it +and that will work in any container +The behavior described is consistent with missing commit statements in user code + server is not actively maintained. +Initial design doc for compaction in version of the design document +But then it works everywhere +Ubuntu not +Marking it Blocker so that it makes it to +Will upload rebased patch soon +ve committed this +Patch applied Committed revision +Thanks Heshan for the contribution +Ill take that forward in +The method is where the lookup should occur not where it is currently being of the other tasks we should have if there is time would be to simplify all this integration code +most likely to be pulled in for wtp sr +Ive just resurrected the +Simple fix we were using an method for Hashinspects but since we were in mode encoding was null +Are you sure thats the right test +Do you have a patch expliciting how you want the Groovy console to behave wrt that context class loader +Sorry guys this is an enhancement not a bug +Committed revision . +However a bunch of the vectorization work is in operator code where we have vectorized version of the hive operators +Icould find where this exists myself and will post a query on the User list tosee if anyone else knows +Please verify and close this Jira or provide steps how to verify it +Here is a fix +I will copy this to the branch soon +This patch has a that shows the problem +version with incorrect manifext per this JIRA issue downloaded from unknown location on +Somehow I hadnt noticed that parameter +This patch provides better reporting of the state +I understand but I continue not knowing where is the problem +bq +then I have no clue about how this can happen +Very nice sample +Ill bring this up on the mailing list +More detail would be helpful +Rebased and merged into the master branch. +Also should be released in a few weeks +The base problem is we are closing the same socket connection twice +imho its too late to tweak this +Bumping the priority to critical because we are seeing a very high NN load because of the lease renewal and I would like to get a patch in for this for patch is a bit ugly but it does fix the issue +Hi all is there any wrong with hudson? We have tried several times to make the patch be applied +It might make sense to provide a constant for it somewhere though +I will attach the out put from those tests to this issueMy setup was exactly the same as the one Simon is using +Updated interface in revision Please note that this is only a preliminary interface and will be replaced with the official interface once JCR is final and available +Unless that is explicitly what you needThanks for reporting that +about the manifest it could either pick it up from the jar configuration option or from the classes manifest is already being used + +Updates to that blob should cause both reads and writes since the pages should not fit in the cache +thats why for maven from gae were rewritten +The bug remains open until a better solution is implemented +It is fine if every node acts independently as long as each of them respects the SLA +Until there is a model for all system resources in the capacity scheduler this patch lets users with that kind of problem achieve reasonable behaviorWe are missing the woods for the trees here +Should probably be using internal variables for nanonum nanoden offset +I look forward to mid next week Thanks again +When Irun tests I frequently add a UUID to some base table name so that eachtest is isolated +Output looks identical to the xml source as it should +will +Knut you are right +Theres lots of duplication already in the platform +bq +Gotcha +Although is not a required property it is present in both Suns and IBMs +Chemistry opencmis artifacts are already bundles +There have been no updates in days and there is still a DNS issueWe can always if these are resolved. +Thanks Suren +Committed revision +the banner file +The fixes will be committed to Branch soon after I ripped the extra log off the weve tried the patch and things certainly improved but it subscribers slowly started to fail over time +The tests are implemented with new jiras decribing the test failures +Is there a reason for this +Patch attached +I forgot to include the test recorder xml comparison file in the first patch so here it is by itself in a second patch file +explain the GC problemThe trimming feature of CAB doesnt come into play with the parsing algorithm Ive chosen since that searches explicitly for token start and end rather than splitting at separators and trimmingIve made a minor modification to trigger different exception messages if there is an invalid character after a token or if there are two tokens without the required separatorcheers Roland +Updated patch with support for both option types. +Tests are now integrated. +Thanks for the feedback +Attaching a test program +SeeMaybe this resolves this ticket +This patch worked for me Thanks commited. +duplicate of +Integrated to and trunkThanks for the patch TianyingThanks for the review Jeff +works finemvn clean verify Pintegration Dmrtrue jclouds KEY +An database doesnt add value in production +I like Joshs idea instead to keep the commas out of the copy buffer as it wouldnt clutter configuration for something so small like this tiny bit of presentation stuff +Updated patch some fixes and more tests +Patch looks good +Ralph I think you did some work recently in these classes WRT to SD elements +threads increments per thread columns per increment Each increment uniformly distributed over rows trialsWith new locksWithout new locksso about a difference +As a workaround you can configure project and add maven build step with sonarsonar goal so violations report will be generated after sonar execution +Any chance to get this for? Its blocking which will be nice to have +Attaching snapshot Thanks issue is fixed in and master branchIf we are going to cut another RC for need to pick this changes also else we can document this +trunk depends on released camel +The problem does not reproduced on Windows and Ubuntu clients only SUSE clientException shows that client send all data and do not wait for ending responce from server and closes connection +PR was merged resolvingThanks Tristan +Why did you create this issue in the Spring AMQP issue tracker +Issue is resolved +Please find new patch attached now test failing +wsdls + Patch looks good +It also shows the headersIf you think it would be useful I can add it to SVN +to GCI changes +there is one error in line of isWritable is not defined +Test to reproduce the Configuration you could also try BLOCK rather than this is expected behaviour if you dont give a server enough memory it will OOMThe paging chapter already explains the semantics of so I dont know what you want is a simple misconfiguration on your the server never tries to load more than from the journal +New patch fixes the build and also verified the runs successfully +I think the builder should remain a syntatic shortcut over the and not start adding too much Api to it +The related documentation changes are covered by bug. +Apologies. +ehcacheehcache depends only onehcacheso there will be no duplicates and all artifacts which still use for example ehcacheehcache will get the actual artifactNOTE my latest patches for rubygems code allows to use a newer plugin via configuring it with HOME +Addressed some more comments from Amar +At least Im keeping it a release blocker for +Weve also unsubscribed you from both lists +With such configuration the sibling project targetclasses would contains the sources and during reactor build they would be accessible to for this detailled explanation of your use didnt understood the returnscanner +committed +Hi Cdric yes as per there was apparently some problem at the time wrt Grails +Marking closed. +This patch only makes sense in trunk not inWhat are the exceptions that now are different +Until a Gant is released you will need to edit the file sGROOVYJAREverything should then work +Introduced and a default singleton implementation +If you place the assign element to a wrong place this error what do you want when you put the assign element to a wrong placet allow user to move assign to an incorrect place +It also supports SSL +So I will put them back in or I can no longer use the tool on my machine +This works for me with you have to start manually before running the examples and integration tests. +Editorial pass over all release notes prior to publication of +Hi ColmThanks a lot. +lol no +I shouldnt say that it will fail +My previous commentbq +If a corba is specified and you dont specify a portType then it will generate bindings for all portTypes within the wsdl file +Cancelling so can patch now that findbugs appears to be working if patch passes hudson now that findbugs appears to be working sure why this has been back to me +And handle the project split issue in a jira maybe for HadoopRegarding the second point of a more complicated mapping again we agree that valid use cases for this could exist +I think the current behavior is better than the behavior in the beta +This recently caught me out too +Ill update the patch to include oneThanks for the with tests for cache invalidation after a design doc tests and fixes for module resolution and just took a glance at it +This is probably too late for so maybe we should aim for a solution for where we can make incompatible API changes +I submitted the patch with revision +sh scripts because if one day soneone chooses to change the port or the password key it would have to change them in more than scripts while they could be change in a single place in the +We just released are you going to be needing that as well? Why the need for so many versions of the compiler? it would be easier for infra to manage if just the latest version was on the servers installed as a deb rather than having to maintain separate paths for multiple versions of could we find the deb packages +Fixed by +last change broke the tests +Yes I think thats fine I did not realize it was covered in a separate JIRA thanks +Hi Adrian any update on this issue? Were discussing cutting a Karaf release soon and Id like to know if this issue may be worked on resolved soon otherwise we may push this one out to +Merged into master. +The existing mechanism using pessimistic locking is the way to go +Committed to. +im on the patch on the branch revision . +I could verify that recovery works with parameters +Mamta One question that I have for you is about your comment on CREATE VIEW +PIPE +I am uploading a new version of both patches based on the tag since this was released this month +The pull request needed further work +Committed to trunk and x +Patch w the trivial change +doh for somer eason the PR didnt show up in my browser +Earlier the classes describing extensions have been implemented as internal classes of the Extension class +oups you fixed this and sorry +Alex is your latest patch usable on the release? It is included but I did not chek it +Folks any progress on that appart from reverting to a JVM or +This comes up because I wanted to tokenize strings in a simplistic way to test that the highlighter can handle different tokenizers and it just wasnt working right +Triaged for +Closed after release. +closing +Affects Distribution Distribution Distribution Nate nor I can reproduce this +Thanks OliverI modified the since tags but otherwise committed as issvn ci m Applying the copy of the code to stop cyclic references over to per and Oliver Sauders patch srcSending srcmainjavaorgapachecommonslangbuilderSending srctestjavaorgapachecommonslangbuilderTransmitting file dataCommitted revision . +See proposal in +Attaching a patch that is just like the previous except that theres some minor changes in the location of the returns and tabs to facilitate visual comparison in the outputI will commit this shortly to trunk then port it back to and later +seems to be because of on ubuntu still is likely something specific to how the java is installed. +The HDFS unit tests all pass +please look at the function is added to Configuration tab +I committed the patch to and trunk +And yes theres an effort going on to fix this but time is quite limited for some of us these days +Bulk change of Fix Version for all issues fixed or closed for +So its a bug in because Chrome doesnt follow the ECMA standard? Applied in and trunk. +And even if we made all loaders like it is still possible to use groovy classes without using any classloader provided by Groovy +Uploading a diff for the admin guide to it can stay in synchwith the policy template file +OK +What kind of issues are you faced with? I use the plugin to do lots of web development with no issues +But I have already tried some of the suggestred combinations and none work weird +A great improvement for those Camel users making use of the transactional JMS Thanks +The Checks To Print link will only be visible for of type CHECKRUN +myfaces springs security you tried Spring Security with Default security context implementation of wraps the requests by default so I dont think there is a need to implement custom security context implementation it should work out of the box if the is configured and afaik it is configured by default using security namespace configuration of spring +If everything is OK I request for v to be considered for a commit +Docs look fine +A web service which demonstrates the problem and the invalid WSDL generated by Axis +Update on this issue +triaged for I don think this has happened much +Thanks atm punted on the last one since there isnt an method that maps from response code to string +Closing again +Do you have a sample DTDXML to verify the fix? Thanks +Mathieu I need you patch too +is no longer used in IPC. +Suneel if you have a patch I can take care of it +Sorry for the confusion there +Good catch Konstantin +I can see where in code a NPE is caused but how to reproduce itNote this is the VM failing just because a NPE occurs not much we can do about that since NPEs will occur since we cant guarantee against NPEs The vpe on linux is seperate issue +But I didnt look into it carefullyThanks for cleaning up so much code in this patch it is a big improvement overall +No code will rely on this adding a message behaviourregardsMartin + updates should be created for missing cases but otherwise this may be an easy one to fix +Applied thank you. +Feel free to use my patch as base for your development or simply apply itLars +Id be marginally in favor of or instead but dont know if its worth building a new class forthis single method andor whether we think other methods might get added later + +Thanks +I have no estimate for this +Assume closed as resolved and released +Im not ready to commit myself though +This patch includes themHowever I dont think these are solutions +Defer all unresolved issues from to +The javadocs are incorrect about thisShai pointed that out to me this is why i opened the issue +There is probably not much we can do without more information +getQueue is not passing through the code path change you made in an offline discussion with Hemanth and Vinod +Hmmm +Deferred unless someone has time for it +Best guess patch at revision +Remy the attached jbossweb patch fixes the memory leak +policy +No regression foundadceaabaf defccbdbef + +We need something like this patch for the short term +implemented +Hi Bill +There are lots of JIRA issues to process and very few people here to work on them +Each would know their dependencies and poll them until they receive a yes Im done then they could begin on their merry key problem with this is that Maven assumes that all build artifacts go to the central repository so relying on maven itself wont be a solution +Duplicate of +Posibly because this is just a work around moving to means the it does not collaps with a but with a so it only reduces the changes that this collaps takes placeAnyway because I am not able to retrieve the patch I am going to modify and added to the svn. +Last comment was for Wendy the one provided by Wendy at matts but not yet id tag inside contributor must be removed for next time +Correct release version after merge +Adam if I follow you right youre saying catch the badarg on listtoexistingatom and leave it as binray +bq +attached junit update for initial test run that I executed +Ive updated the patch +Completed At revision +Closing as fixed since all dependencies are resolved and this issue doesnt contain any additional specific tasks +We can also document a workaround namely add as start key suffix and as stop key suffix +patch with Checkstyle issues fixed +I merged the patch to. + where removed from the release notes by request prior to that release +Changed the output file path ruler committed to trunk +Made a follow up patch for Linuxwbuildrsvn ci specpackaging m directory entries dont exist on linux so this spec fails we now check that there is no file entry insteadSending specpackagingTransmitting file dataCommitted revision +HiCan you try and reproduce this with the latest source in the branch of svn? ThanksThere have been a number of mutex fixes since fixed in. +Were going to close this out as were not doing support for anything but Eclipse at this time. +This is a copy of the thread dump +Also I will see if MVN is working fine with the values +in and CVS +Since the problem in the has been fixed Im getting rid of the workarounds I had added to the test cases +Please advise further if you see this problem again. +from ID D rows selectedij select +Here I had kept only the necessary changes +This should be used to set the service address for tokens +Non bugs should not have a fixed version +Ok the implementation is pretty fine but I am not yet fully convinced with the nameIs it clear what a is supposed to do when just looking at the nameDoes anybody have a better suggestion or is this name acceptable +Oops imported the wrong Singleton Bug found and was fast Thanks so much Arne! +We can take a shot at the dependency monitoring seperately this implementation was just attached to give some hint what we are thinking there +trunk Polished the example and fixed CS +Let me fix comment on commit +Unassigning this for now I am not able to work on this right now +Duplicate of +Look for in the commit comment +the example above shows the problem in action for me. +How does one build a compatible version of Pig? Do you have to download the jar yourself? It looked like if you did this would manage the rest properly +Im not able to get this last patch to compile +Based on discussions with the team +now using a streaming checksum +Patch to fail on node dead and exit nodetool with error status code +We are going to simply add query to the interface +As I recall I pretty much took the code from Mikes comment above and pasted it into as he suggested and it worked just fine so I should think youd be able to build a custom version of the Derby code for yourself with such a change without too much work +This is probably the wrong place to place questions +typotypetextxml charset + doesnt support the some camel annotation as the doesIm will update the wiki page with new added tomorrow before I close this issue +Committed to trunk. +But if there is only one big block of sentences it will all end up in one object which does not work for the cross validator +This is very nice +looks good so far +I would prefer to see setup split up into three different functions +Found a solutionMove from entities to ejb moduleCheck the file for detailsAfter double check ejb specs I think this might expose a bug of Geronimo +Fix appears in latest alpha docs. +The only situation this seems to work best for is when you use the new runtime wizard AND check the create server button but it is impossible to tell this entry point from othersA workaround for you is to create your runtime first with no server then go create a server and select your runtime from the combo box +It affects any unix configuration. +Such information should be cached +I do agree however that the configuration code could be more robust and handle invalid configuration +I was adapting AVRO to work inside and already has that infrastructure so it was easier for me to it +The name if it can be trusted again says it all +Ive named the property disableCaching +I like the last version very much +Even a test case would be helpful +This patch is for this fix applicable to also +Have fun withRegards Florian +I agree though that it could be made more clear as to what its function is +Did you replace the ampersand? Does it work nowId like to turn on your project page +Please check it thanks +Sounds like a good way to for now Jacques +It now passes withJCC +Im facing the same problem with February like Oliver Haider +Each class in a hierarchy should be reflected at most once +Added testcase in svn revision profiled for an hour memory usage remained below mb for the jvm in this time +Thanks +IMHO we shouldnt use our deprecated methods in releasesWe should mark them as deprecated when we already using a different way and the method is no more used in our codebase +I attached an extended patch +It could be network jobtracker or something else +just committed this +The models are immutable objects +Some other fix may have fixed it +Im flagging this issue with the gsoc label in order to set it as a real project proposalAs I said ealier one student as already showed interest in this project on our dev ML Keheliya Gallaba +It was put on the mailing list to solicit comments about CSS coding guidelinesI recommend removing the attachment to help avoid confusionThe is a much smaller step in working toward the ideal fileFrom your perspective how should the style sheets be separated +Fixed in rIf self is available use its type +Does the user then need to stop and restart the server for changes to take effect or does the server stopstart automatically? Or is stopstart not necessary +bq +While you are doing the update you could also solve this license information issue +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Committed patch to documentation trunk at revision +it is not good to use this seperator since it will be different on each platform +this appears to be an issue again +Ideally we should change this method signature to return an array to avoid unnecessary object creation +INFO Connector started INFO We are using the agent +I am not sure exactly how this should be done but yeah simplified timedate handling sounds like a good idea +Ill commit a fix for it and port the fixes to the branch. +This is a patch to the whole project except the fail directory +bq +But I need to replace with my own query signer +Smart service providers will put frame busting code in their approval pages to avoid UI overlays +Integrated to Thanks for the patch LiangThanks for the review Lars. +Generics are used and the cant map them as the name of the generic class are changed with each instantiation . +Defer to defer open issues to defer to not essential to address for +for the class keyword maybe something along the lines of transforming to classElement or classAttribute as appropriateStill leaving the option to choose something else via a binding file. +So I think thats the question +Amos I also have a feeling that the same problem will occur when using a transaction between a C xatmi call using RTS and a java one because the same code exists for handling the messagecontrol parameter as it does for handling the messagexid in stompconnect +Tablet servers should refuse to load a tablet that hasbq +The jar file contains the fixed jsp pages that should be extracted into the root directory of the application +The jar file used to be included in lib but it is now included in m not sure why it has been moved but I believe the sar is the correct location for this +bq +I would like to discuss more about this project +Creating a wiki page with requirements examples etc +Since the module in question was mathcollections maybe this is a helpful step anyway +we handle these by yanking the viewport size out and creating a fake APM header which works for a lot of pictures but sometimes produces the wrong scaling +alternate fix committed in r. +The codes only dependency is on for tests during compilationAll the codes headers should be changed to Apache License as well +Added Requires to joda time packages to check for existence of adding a dependency on does work around it with CR +Thanks for taking care of this so quickly! Greatly appreciatedBruce +I have Eclipse configured to have automatic builds +patch look good +bq +Not the case do think an option to not override the local and remote repos with eclipse projects should be an option +I tried same thing on my home computer and got a java stack overflowAt work I am running NT and JDK At home I am running XP PRO and JDK betaI have just downloaded the nightly build If the nightly builds are not the latest then where is a later jar I can use +sample pom for ojsp samples +I added a test in testChroot function because the bug was introduced when the chroot capability was added +Id say it shouldnt cause a problem + +i will commit the patch including the equals and drop them in a separate commit +This one is correct +Thanks Pierre done at r in Erwan! +The text above was an issue submitted byGuillaume +For us continuing with the wrong schema is bad but we want to notify the user of what has happened. +Julien do you think that Im being unreasonable to expect that someone at least determine if the patch is completely objectionable within a month time frame +Completed in +There is an issue somewhere else however since readbuff shouldnt be called with readCredit in the first place +select from where DEPENDENTFINDERViewDEPENDENTID DEPENDENTFINDERPROVIDERID PROVIDERFINDERView providerid points to T table in systables no way to know which fields the view actually usesView providerid points to T table in systables no way to know which fields the view actually usesThe other fix that I am thinking of is to reuse the codes that check if the fields and tables are valid entities for example during a CREATE VIEWoperationYour thoughts? ThanksKarl +If we doesnt support more eclipse then its not reproducible under eclipse onits not reproducible. +applied thanks! +I am starting to wonoder if the fix for this and for other devnull and NUL related bugs is to revert to the we used before +All unit tests now pass on Linux but fails on Windows +dom attribute is fixed for panel is fixed for domform +The location is an additional thing that only comes up when you start using paths that contain a drive spec like C or D or even D which means in whatever is the current folder of D +An internal Red Hat one yes but only after artifacts have been deployed to nexus +is possible Ill check it +Is anyone looking at this bug? I would be very interested in trying a patch fix for this when its available or being notified of a possible workaround as this is a serious issue for our project +This may take a few moments +However if youd prefer it to be that way I will change thatYeah I confess I didnt read your code style guidelines +thats where some of us tend to hang around anyhow +Hi Elvis sorry for the late are you saying that your code correctly extract the command object but doesnt work for other model objectsI just want to understand what actually works and what doesnCheersSergio B. +Ive also attached the patch to produce it. +Attach your notes to the Jira and reassign to me when done +mark a JIRA item as Blocker if theres no test to recreate the issuethanks this from bucket as theres no way to recreate the problem yet +x you pleae fill a new issue for this since the background is completely different +If the last call to returned false calling Current throws an exception +In the TCK distro we have a lib dir which we bundle + +I can run it in my english workspace it works well not too long +I tried your code with a very simple example and I think it would work for the above purpose with a bit more work +to trunk and +Confirmed with hitting the backspace as well +If the patch does not fix the problem for you just reopen the caseWire log will be highly appreciated ProblemWhile the fix did solve the duplicate cookie problem it created another one +You will need to figure that out before we can commit the patch +Thank you +I glanced at psych and it sort of looks like all YAML files should be treated as +I just pulled this out of +Closing was caused by IBM JDK custom configuration in jrelib +This is fixed if provided profiles are set as default for the same language Sonar server stops +Exports file in csv format If we create multi language program app we externalize strings in property this file is in formatuCuEuBuF uuAuuuuuu uuuuAuBuuuuBuDuuu uFuEuBuuuIt will be difficult for translator to translate this +So all issues not directly assigned to an AS release are being closed +To solve this we have to add the synchronized isClosed check on our side which kills performance +I will look into the logging problem! +Attached patch fixes the issue +Found workaround copy the and logkit jars under tomcatcommonliband forrest will coexist with the other Tomcat webappsStill getting an info that Forrest will use the screen to log instead but its not a blocker for this release IMHO +Thanks for the update +I did find some other issues that I logged related to security +If were no incorporating the patches I did for march detection I think we need to do what Andrew suggested and change the order in +I am guessing this is a result of the fix +Thanks for your interest +Specifically the diff to pom is missing +Ok thank you for testingSounds like something else might be broken +to changing the API to use either ping or heartbeat only +Is this fixed or is there any plans to look at it? This has bitten me in the past +Attaching updated patch with the changes +Thanks +The plugin should make the contents of the distribution configurable booter policy file booter script file lib directory contentbut provide plugin should not depend on the version of the booter envisaged booter booter with goal allows the user to dynamically create distributions with the content of the maven to add deploy functionality +That was a very quick fix thanks a lot +is now no longer optional both in the webmvc and aspects modules +Closing this issue out. +This would help me confirm the fix +Can this branch replace the current git provider in trunk +KonstantinIm not sure I understand your objection +Committed revision + adds the users authorization to a sorted set +If you are at all interested in helping with anything please let us know but at this time we are doing everything we can with the people who are active in the community +by lektran so I think that this issue can be closedThanks on behalf of Marco Risaliti +There is no easy fix +Hi NealeI have create to track this issue can we move the discussion there. +Configuration of Tomcat inside is extremely similar to the Tomcat me know about any issues +The object generate is String and is our +Thus there is a need to access the original parser +I see I had no idea that the comments were not added until you touched a button named clowncopterize +can you try with Trinidad +for the patch +Yes i was running some tests in mode +Attached example app patch file took a look at the new patch +modulessrcmainjavaorgapacheairavataxbayajythonlib +At the end of the day it is still a test rather than an test +Without the patch after a couple failed security attempts I am seeing the server crash but I am also seeing the same behavior without the patch +Reviewed all these issues and marking them as closed. +If in the future someone would like to tackle this feel free to open is and attach a patch. +Starting with Hibernate another option is to use Javassist instead of CGLIB since that feature was added specifically for use inside. +Beta will have BPEL. +If this is the result of a file descriptor leak this might be related to what I am seeing in. +What could you do with thatNetworking toolkit route makes reverse resolving by default +I finished my review of the doc no serious issues +Bullen commit access permisson to work on he is one of the few users that makes use of this datastore and is will to create current patch is a bit too small to count as a code review we should try and figure something out for his next change +The fix was not to create response message context for Operations +Ill address this on Saturday morning +The docswsm file references the in src but it is really in One other note about the documentation of it would be good to edit to add so that the doesnt have to add the line +fixed will be in the so we can merge into next fuse release +Can you please investigate and provide a patch? Thanks +Otherwise you would need unsafe casts all the time in the ASM codeThe patch looks correct when I upgraeded the policeman checker of Lucene the only change was to move from implements extends and the ctor param +I am new to the process is that the correct way to report documentation nits? Chris +The issue is fixed +For this reason we should make consumer cluster a more explicit construct +A conspiratorial group of plotters or intriguers Espionage is quite precisely ita cabal of powerful men working secretly Frank +All issues found with the MED editor will be written up separately from here on +This requires changes in to get around the async nature of the launches +No response assuming ok. +For another project I had to add the to the installed JRE under Eclipse +agreed +Not included are elementswithout text content but with child elements +ftl where am I going to write it +Committed at subversion revision +after release. +How many bits do you needAndrew Purtellbq +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +Editorial pass over all release notes prior to publication of +done in to use out for moved from to in so Iwe dont forget to undo changes for the in r +Its usually the message size limit that stops messages from going was set to kbytes which is absurdly low +I thought about opening a separate issue for RT +Also take a look at what can be closed here please +Looks good +See +There are no new tests included since the change just reinstated some methods that were previously removed. +ThanksI agree that leaving out math for now is a good idea +can we bring this commodity to the maintenance +coresrcmainjavaorgapacheanymime +I I did the defualt install which I assume loads the I did the DEMO install and it could not find the IDNow I agree with yoiu that if it is not in the table it wont be foundso what happened in the Default install that the data did not get loaded or was not accessible when the code looked it upJacques says he can not repoduce it +pglocks relation locks mode virtualxid locks mode relation lock mode granted now explicitly handling transactionscommits on the lazy loading temporary session fixed the issue +Ok sent in the CLA request how long does it usually take to get processed +Thanks Vinod! +Youre sending to target not to hint +The linked jira has been created to update the code cleanup +Also verified that hdfs oiv using a fsimage with symlinks look OK +Also the idea of maintaining the last vote intact is to keep the pair used to decide upon the current leader and currently we dont use the zxid field to determine leadership if the notification says LEADING or FOLLOWING +A function that is evaluated by the source RDBMS directly is called a source or pushdown function +yeah +This is a really interesting project Haohui +HansBut is there not a problem with the default value you want to set inI mean if the person logged does not care +Looking at the code I curently cant see how this state could be reachedCan you turn on debug logging and try to reproduce this state? If something goes wrong there should be an exception in the log +Hi VenkatI was very tired I think +What is the command line that you use when you get this errorCan you please put together a small complete Maven project that can be used to reproduce and test this issue +Owen do you think this looks good at this point +XX and XX didnt help +What that means for the route representation will be up for discussioncheers Roland +Version M has been releasedClosing the issue. +Okay so I guess the question is how do you rectify the package structure that is inherit to a java object and the package structure in BRMS +Rob you are certainly right +Yes go ahead and integrate this next +Also removed the min and max allocation mb configurations +I just changed the reference to what is getting downloaded +really strange +Dont be to strict with that +Committing +We changed the process of converting a floating point number to a double in addition we changed how Railo handles Objects before they get converted to a primitve double and this double then get converted to a String to display have now a more direct approach Float objects now get directly converted to Strings +Ill make the corrections required +Not sure +This is how it was done for the yarn directory so did the sameA srcmainresourceswebappsclusterA srcmainresourceswebappscluster +and to the Dev +I cant take that on just now +Im pretty sure the problem Stephen ran into is an error in the bitset allocation inside the collapsing code I ran into the same problem when I customized it for internal use about half a year ago and unfortunately forgot all about the problem until reading Stephens comment just now +Feel free to commit it if you feel strongly about it +there is no risk the HDFS data handling goes over the internet between the nodes +Thread dump during incident. +Committed in trunk for rev +Patch to skip whitespaces +Overall this would be a very useful feature since it is possible to add fencing around cpuio resources in addition to memory usage for MR tasks +This has been fixed with commit aaefcaabcddbec +It may just ignore the annotation if present on an interface +I would say set rollInterval to and let idle close the files for you +Let us use strict mode for thatAdding more and more properties may be more confusing. +has been created to capture this discussion +it makes me think you are talking about when you are in the browser you are clicking on the submit button is that the caseIf it is then this is a very simple question for the user list not a bug +This needs to be explained in the spec +Updated code checked into SVN +Updates and to show cleanup task +verified on +Bill Havanki Solutions Architect Cloudera +Using async JMS on the server side requires a bit more work but might be done along with leveraging CXF continuations +Worked under Axis +But I get what your are saying no large need to include this +It seems like it may be a little weird for the user of the proxy however if they only see the first or the last exception that was thrown +Just adding to the discussion These bugs are related to the optimization issue which introduced a new in the in order to accelerate access to by name +I will see if I can push on this a little more since it is cropping up regularly +You should be able to spot them by locked +That part of the code I have reviewed before I inserted it and it looks good and I think may have also been looked at by Nicolas +Changes made for handling client side +attached +for the change +Hadoop users will have a more straightforward path to Cassandra adoption and vice versa +I also updated the test case to include the the rounding error scenario noted in this bugThis bug is fixed with SVN revision . +The more dependencies a library has the harder it is to integrate into an application and more chances for version conflicts +Very odd resultRandom writes Timer for and cols msRandom reads Timer for and size msRandom writes Timer for and cols msRandom reads Timer for and size msGoing from columns in the family to columns leads to x slowdown +Same to each line end for most db sql to the line ends. +Sorry I didnt think to check for licenses outside of the code +The next version would be which we have just started working on +Moved the mucking around to tmpThanks Hitesh +The epoll bug is not solved with jdkHave you profiled with trunk +Works for me. +Final patch attached. +Moving out of +I forgot to refresh all projects in eclipse so the and the jar were on the classpath +minor comments on and responded to comments +HiThe following is my sub class of to allow for TLS support fot Ldap the TLS negotiation is rather expensive implementation of TLS connection support will be much more efficient once transaction support is implemented +new comment from Robin I had a similar problem and was able to solve it by simply supplying one file inside srcjavamain +Also it seems that your xmlbeans project is generating sources for tests only +Theres a beta out and a full release expected in a couple days +bug is related to JISP persistent store +upload a workable draft patch +Closed the issue assuming its okay like this. +Will add comments from the other one here +Thanks +Fixed ATMs nit mentioned above thanks nice catchWill ping Eli to see if he wants to do another review pass before committing +Unix group created +Then Id like sitecmstrunk +bulk close of resolved issues in. +Hi Jeanne i would prefer the patch added with Grant license to ASF for inclusion in ASF works +As I said do the google search +A with r. +Though the scope is larger +But if you select something else and then reselect the composite it shows up +Committed trunk revision Committed revision close for +Can not reproduce this problem. +Verified with ER this is still an issue there. +to branch thanks for reviewing Eli. +My guess David is that many of these files have been updated by others reformatting code so when Foster laid back down his code formatted as it was in the original version then it looks like hes making massive changes when this was indeed the work of othersThat being said Ill get with the Fost and get this cleaned up so that it can make it in ASAP + +Only issues that are actually fixed should have a fix release. +move it to as its not used by release +How does this work with vs? You push different jars +it should be trivial to fix its an obvious bug introduced in a stable release and now another stable release has gone out the door +An identifier part character is any character for which the method returns true. +But Im happy to hopefully I will not have to it in the future +I havent articulated my intentions +It looks like Jukka had already added this magic back in r in We already had a hdf unit test in r Ive added a hdf one +I think that javastubsjsr sounds fine +It is the changelog that is missing +MVEL pre is currently used by drools. +Verified. +Its not actually a huge problem for me +But where do the and come from +I notice that the INFO output for the individual cluster density is written here +I fixed the javadoc issues and committed. +what kind of error appears +I had to clean project to make it working after that +Sorry for the delay +Add to the list +Would be nice though to get a working version including working test cases +If there are we need to uncompress all messages and with new offsets assigned +not onlybut method get wont be called +After applying the patch please do the following +The patch changes the code that has not changed for a very long time it bypasses using the cached portlet reference every time going to the source for an reference +Working on branches again with this as the previous effort was lagging too far behind changes on trunk + +I tested Axis Pojo deployment without Eclipse plugin and it seems to be working without a problem +You want to be careful there because the effective POM includes not only the effect of inheritance but also the effect of applying any runtime profiles and Im decidedly fuzzy on whether or when roo wants to take those into what conditions does Roo need to know what Maven dependencies are being declared in the POMUnder what conditions does Roo need to know what plugins are specified in the POM for maven build +Yip do you have any interestdesire to port this back to andor +will still use the TTL that is configured in the file after you apply it because the wrapper only stores String Object mappings and is not aware of any type differences +In addition to fixing Knuts nitsthe server side code now uses a threads default test configuration directly and I also added a comment explaining this +It is not recommended to use use this inside the because it will not always point to the component where Ajax request was initiated +here is a first patch +Apart from when unsafe is false you also need to check do this when unsafe is true we are in a TX and the command is conditionalAlso I wonder whether youll need to maintain two ops as well when conditional operations are in used in general +Please reopen if this release note needs to be updated or removed +I will give it a try now +Hi KrishnaPlease set to include a more reasonable Xmx +You should mark the JIRA patch available when you upload a patch so that the Jenkins runs +Patch for is attached +Also if were going to be going like this and pursuing a broader integration is it reasonable to create a branch we can iterate on? Submitting repeated patches is getting painful +However I noted I was unable to find a time where something was actually removed and wasnt at index +Maybe we should up the waiting time a bit or add some retry logic +Can you try with ANT and latest JDK? Also can you mount the share onto some other machine? and try a build there +What was the resolutionI am looking at the same error +Alan youre right I changed to Hibernate due a problem with the dependency +Ill be out of the office from Thursday Sept +Will commit if tests pass +It looks like this bug is the same as in so I am closing this issue as duplicate because there is more evaluation in that issue. +Check in the changed files to CVS +I think routing implies that you get the same message not a copy +Thanks. +pushed +OK I will fix the sample applicationDoes the java mode require the start method to wait or return or does it not matter whether it waits or not +Thanks for providing the patch +Duh sorry above comment is supposed to be on Ill copy copypaste there +yes I do have this package too +Thanks Ivan +Hi AlejandroI just want to clarify your two comments +They are not critical for and are not for typical use cases + for patch +committedOleg +changing to bug so full CDW ack is not required +thanks Mahadev! + has been marked as a duplicate of this bug +We now fully understand this scenario +Thanks a lot for the review Eli +I have tested it and it works ok so closing the bug +For CREATE VIEW is the following statements now accurateTheview owner automatically gains the SELECT privilege on the view +Any progress here +Thanks Enis +Please reopen if that is not the case. +I dont think adopting this patch commits you to any particular approach in the future but in my opinion a sensible policy is always to use the most recent protocol version supported by both peers +Seems like this caused an on Windows +This is the kind of work already done by in + fix +I think Jon has a great point +Incorporating comments from Hemanth Add javadoc for Refactor common code to a new class rather than using static methods in Apache license header for the start time issue which was missed out while refactoring the code +But I think it is a convenient feature to have location of the shown in the field. +Converters are mantained in a map managed by which is usually unique among the Application lifetimeA new instance of is created in s constructor and that single instance is referenced by every thread in the same ApplicationIm not even considering internals since its javadocs states Date formats are not synchronized +My conclusion is that I dont feel strongly about having one way or the other but Im slightly inclined to leaving it to the application +Bruno maybe you care to comment? +Moving out of +Im going to change this from Major to minor and itd probably be easiest if you create a patch for us +will be at same locations as the not the github as Chip replied +Just to clarify the code is by Hans not believe he said in the newsgroup that it might be a good idea to move map and zip etc +I have set it up on a properties file in etc and that seems to be workingBut I was trying to track down where the id and version were coming from when you execute the About commandI tried extending the class but the thing is that the Branding class doesnt have a version op but I guess we can try and define spring to pick that up from the file +To which persona would you want this feature to be available? As a programmer creating the Pipeline Java Action that Oozie is launching? As an Oozie workflow owner assembling multiple Java Actions and therefore not able to access Crunch? The reason I ask is that some of these capabilities are already available in certain situations +I removed a little too much from the first POM +Agreed fixing for +If we think that doing addresses the issue to do with initializing too many jobs simultaneously in the capacity scheduler then I propose that we dont fix this issue at all for now +Again on some systems the delete will fail if the file is in use while on others it will succeed but the other apps will still have a handle to the deleted file and in effect will be writing to devnull +The OASIS java CAA tests and are testing this +For some reason the insert method on Linux for vector is very slow +Verified. +Also I have to take a look at and process Stefans comments here +I see therer some major refactorings going on related to debugLogger etc +Yes I can where I can find changed version +I fixed this bug +Any opinions on this patch +AttachingThe ZIP file includes a few deltas to the statically generated classes and to the test case statically generated classes are different due to a subsequent fix that went into +It is the same error everytime you try to connect it +Applied patch +That makes sense to me as I believe once prepared nothing should interfere with the commit but I say that just from what seems logical not having done any spec or product research on the matter +can this be committed so our tests can pick up fixed code from trunk + +The Difference is for each Node separate report is created. +Just a thought +I believe is a duplicate +Sorry about that +Additionally Ingres use requires an updated CMP from what has been shipping in AS for some time now +trunk Committed revision x Committed revision +See are trying to following the rules called out in the specification +Tom +From what I can see from the diffs the problem is that the sequence of output is not always the same +I just dont understand how the Hadoop community can pass up an elephant with a water fountain on its head +generating empty tokens +Agreed such a test would be useful +shouldnt our editors be using plain eclipse API to access the resources and not s abstraction? +Committed to the current. +If multiple versions of same VDB are deployed connect to one specified with ANY or if there are multiple with ANY then connect to the latest version with to +Heres a patch that fixes this issue +Here is an integration test that demonstrate the issue +This is mostly just check style violations +This DTD file validates the +So you should make the same changes in there as well +Please try version The paging has changed in that version. +Ravi Teja can you please look at the latest code and update this patch? Thanks +Put the new patch in response to Pradeeps comments +Work in progress patch that adds a Cassandra service +This patch implements positional read for andI will later normal read in terms of pread in a separate patch +I think we can close this +Ran ant package true to verify there is no more javadoc generation with flag on +Above was solarisIf I use on x linux problem is presentNow looking deeper +We plan to remove crc files as separate entities replacing them byper block crcs on +distributing tasks according to data requirements +DROPRECREATE doesnt works because name is randomso it will create a new unique index everytime and never drop the exists onethe point is use a fixed index name rather than a random name break compatibility with previous versionplease NOTE in release notesthanks! +patch missed some files +Updated patch fixing this also in and +I reviewed your patchI was expecting a single script that has essentially full control of the logic of how to handle a request and it would unlikely extend to do this +so whether return null from getTerm is right or wrong I dunnoIn fact its probably wrong but at this point its a moot point +Rex had already implemented the feature inThis JIRA is to simply enable the commands for Karaf shellHTH +Assigned the issue to me. +Security class updated and file alsoAll the tests have run successfully on my all tickets +There are a number of other corner cases where the STOPFLUSH membership should not be the same as the the simple fix does not work and fixing it correctly would be much more out the current fix +Could you provide a way to set java mail property +can you two sum it up again please? When does it work as it should and when does it notI get so far it works from the command line but not from the console +I think we should do this for but change this guy to use FST for So if someone has a big dictionary it wont eat up tons of RAM and also enforces immutability +A Unittest to reproduce the bug +All integration tests still pass +The other problem is even if we fix the replay issue its still terribly slow due to excessive throttlingI like the idea of changing from a push to pull mode for hint delivery +thank you for the prompt replies +Fixed location and added text to user the correct. +As the wiki notes This is an API that should only be used if the application is taking complete responsibility for update concurrency replication and sharding +this looks good to me +Siying I get a test failure in +It would make sense in a My Portal sort of application where people go there by default when getting started with and some introductory text and possibly links and such is meaningfulI dont know if this is the best way of doing things but the problem I have with all of this is you are both arguing based on only one thing consistency +Thanks +The code base is pretty solid but it can always benefit from additional analysis in advance of finding bugs +any patch you could provide to improve documentation +There is still a performance impact but it is smaller and tlog sizes are under control +When the journal flushs it creates a new journal mark object and flushes it +Make sure to run at least the two tests when you submit your udpated patchLooking forward to it +All set +Id say wait for just to have less moving on and trunk +Given the code is the same this may not be regression compared to +Yeah it was easy to fix with your instructions and AMQ is planned to be released soon +Chris Want to upload it as a patch +Its just an field for which the should be assuming a default mappedName of javacomp per the ejb core spec References The container must make the interface available either through injection using the Resource annotation or in JNDI under the name javacomp in addition to through the interface + issue out as part of JIRA cleanup +Honestly who creates copies of their repository in tagsbuildsWeve excluded indexing of tagsbuilds as it was overwhelming now fully indexes in minutes instead of hours +Merged into the master and branches +Thanks Paul! +Can you provide a more specific use case or description of the enhancement youd like to see? Its not really clear what youre looking forIf youre looking for an system like in Java Swing that doesnt seem to fit with the RequestResponse model of the web +Also I believe this affect as well so I would suggest fixing it there +Any updates on this? I am blocked from upgrading to +We can merge it back to and if they are to be released +looks good +oh I forgot to mention that the temporary workaround is to disable the continuum support in the IDE and just use the maven support +LGTM +QE would prefer to specify those profiles on the command line +bq +It can go either way depending on to which patch some of the changes go +bq +The failed tests are ok on my machine +Hooked it into also hooked it into for as a way of showing how it works. +Thanks ThomasDo you want me to provide a patch +Ill look into this issue +Pete I guess it is done +Sorry did not check the codes carefully the service is of bare style so should not have a wrapper request class +First waiting for reference count to reach zero wasnt part of the original semantics of obtaining connectionSuppose there were two clients A and B +Rather in our application we need this control in the endPoint levelEg Endpoint connector Endpoint connector etc. +Further check into branch. +another issues with actioncomponent and whats on the stack when you use it +Please andclose the bug afterwards +Specifying a pattern takes precedence over the alternative dateFormat and timeFormat options used to specify datetime styles +Note that I talked to danielsh about rolling up builders but most have sufficient different steps to not warrant doing that part +can we close this issue and move on to refactoring +Good to hear that it works for you thanks for the quick testingJuergen +In spinning up ssvm in zone I dont understand why it will go to look for cache storage in zone? Based on our implementation it should look for cache storage specific to zone which should be there +No these were the last ones I could find everything else was already covered by existing +Wed still some way of passing on form context though +There is no any NN for central. +Yeah I would like a more pluggable interface for the thrift service +NickI dont know the actual status of this but flow was created to catch all the commonalities between the jpdl seam flow and rule flow editors +ve committed this +To get a successful build I had to take the corba module out of the top level and then build it separately afterwardsDoes anyone have any ideas how to fix this more permanently +enabled a response to an HTTP OPTIONS request with no the patch contains a wrong test and it doesnt passes the corrected testissues now contains a version with your patch and the corrected test +should have a final patch by morrow +DanielAny chance you could reduce your application down to the example and attach it to the issue? Basically just keep deleting code and tags until the problem goes away if anything else is deleted +You can read the configuration setting there as well +Good work +bq +From the code I dont see a strong need for adding tests like that since we already have tests for some of the levels +The patch contains come functional and style correcting for previously committed Synth public classes and the addition to internalization property file +This jira is over and years old +Yeah Id like to have this mess cleaned up soon +See for the information Todd +This used to work in richfaces but doesnt work in v +It was released with the fix in +This email explains exactly what you need to do to submit your app +Need to evaluate whether can already be removed +Almost all complete +The vdb name is +My patch shows that adding optional client initiate is simple and future proof +But that is minor as it is debug now and the patch just changes the information logged +Lets add the link tot he wiki page as a comment in the a link to wiki page for good measure +Verified in +there was a jar already in the m repo dont know why it didnt got copied over to m +I figure that some changes will need to be made before being accepted +If I read the code correctly this will cause a new instance to be associated with the servlet. +ChrisOnly committers can commit to the branch +If they have a buggy IBM JRE they get a warning +committed the sleep workaround temporarily +HelloI have tested the update and there some new bugs +I think description is put here by mistake instead of I did not understand header and description of the issue +Heres the patch +The token ops dont send data payloads so the of those operations when the spnego token goes bad is not an issue +Closed +Hi EnricoPlease use based configurations when securing messagesYou can find samples within samplespolicy directoryThanks +I have fixed the issue some parts of the calendar are hidden at r and created for the trivial scroll issue +It means that we cant have more than one import active for single JVM but it could be considered a minor issue because we have a cache system to avoid excessive import operations. +I guess it was due to newly deprecated methods +We just need to move it to the client side and pass it across the API +committed to trunk and branch +bq +I committed this. +As a design guideline this is a guidance not a rule +The in the source of nllogj specifies Required to build the javajmxlibjavajmxlib I will fix the naming of the license files in the next few hours +You do not need to patch to achieve the goal +Thats fineHowever if these documents are going to continue to be published on the Hibernate site the errors should be fixed +Also add some logging to the http an option to set follow redirects on with with algorithm issues +Please fix it in too if possible +eeb on master branch +Committed revision . +Assume the stale BBxxxDEADfoo is stored on node A +to exploring deploying within a CF +closed with no testing Ian says this is a change. +its very strange but i cant help on this error because i cant reproduce it +Theoretically the datanode should be able to send an error code to other parties since the bad state is about out of disk space not network problems +I ran myself and I trust Jarcecs test +There is a patch for this on the parent JIRA +pushed to in Developer StudioVersion Build id Build date +Here is a patch for +Code added to the all pass through attributes similar to th one from Mojarra wrote test that checks attributes behavior for different atributes works the same in Mojarra and danger of that optimization is missed passThrougth flag in generated attributes or attributes implemented by hand they would be missed from renderer +These are not done The title should be Web Map Server Import or something similar +I need to wake up +the patch will fix it +Introduced interface and implementsIntroduced methods to serialize an with or without the XML declaration +It is already working +Your test case revealed a subtile bug originally caused by coppaste +Special thanks goes to Christoph Boden for his help on this issue. +Jacopo please keep that completely on this is certainly not an infra task and we definitely shouldnt bug Joe with itThat said I looked at the patch and it seems fine to me nice feature just go ahead and commit it +Ive added this +For the author tags Ill see if I can put shared project specific settings to ease the process +So Ill take the lessons Ive learned and share them with the community soon +It almost appears that if you close the sink end without writing any data the source end never becomes readable +v attached +xsd files has a schemaLocation element +The result was than when you upload multiple documents it just the first documentFile has this fix +bq +coming soon +Since KEY is null that part of the index is blank +please review removed in Fix Version field by mistake +It is not normal Java practice to document all specific subclasses of an exception that can be thrown +The same decision needs to be made for the current proposal alsoThe only benefit of a view is it seems cleaner +Apply zookeeper client jaas on server + +Setting pollDone is possible which will cause the loop to exit with zero items +All possible missing strings are simplier to add afterwards than start to correct all those letters +I take was going to look at this last year but ran out of time +I reviewed the changes in and related classes +Reattaching for hadoopqa +Do what is the best for you. +So it supports +My suggestion is to fix is asap and not to wait till ver +juergeni agree with you that it is odd and that i should investigate the issue further +No type parameters should be needed +Committed revision +We intends to remove it in the featureOn Pig reading data from one file system and writing it to another this feature is supported since PigRichard +Attaching v of the funcspec incorporating feedback from DanDan it would be great it you had a quick look at the latest rev to verify that we are on the same page here +For +Test case issue is resolved by the fix for +Guide has been updated as outlined in this JIRA +So it depends on your timing if you want to change the bundle or wait +Perhaps you could provide a patchDavid +Thanks for these patches Kim +Thanks Joydeep +I am ok with the null pointer check +See for more details +Eventually that might be true but merging is currently a operation +This patch collapses the two optimizers together +yes Page Section is out of date +I feel that the use of lex Yacc is a too big project now +that way cglib can create the subclass and no one outside your api will see it +change dom element name will have to regenerate code for doc layer so there will be a lot of merge work to do including the existing convenience layer API and serial version IDas to the serial version ID does it need to be auto regenerate by eclipse again or must use the original value +I just committed this +it detects that my script is a pig script and puts a pig icon next to the file name +ColmThank you very much. +There will not be a moving to for due to inactivity +Also the weld reference saysThe first thing that a framework developer is going to look for in the portable extension SPI is a way to inject CDIbeans into objects which are not under the control of CDI +Is there a reason why is not generated if Errai IOC is not included? Can it be a configuration issue + +A minimal test case is provided +move SIGQUIT constant to top of class some whitespace indentation seems wrong for sig Can we refactor some of the various code in to share common code? Lots of very similar methods +Should we check and add useless dependencies to exclusion list every version update time +I am very reluctant to change the current default value as theres still users out there that have not switched to Java and above +It will surely make its way into have request can you please give me the guarantee that the code that you attached can be released under the Apache License +I agree with Jesse its not trivial +also the images you posted were borrowed from gimp lol +james echarles ls lib sort libjames echarles ls lib sort libjames echarles diff lib lib +Is there a way to recover from this once it has happened to a project? +With BSP model we can have checkpoints when computation reaches the barrier synchronization which forms a consistent global state +I applied the nd patch by Ilya +is worst I say actually +Attached files with console output +Here are the two dialogs +Yes +That way the staging repo can be used for testing it +Seeing it herePages and +No interleaving should happen +When the detects global workflow conditions it compiles a new and assigns them those conditions +used to just read head in sync collection but we had to change it to account for branches as you already guessedWouldnt it be easier to limit usage of head to trunk only? Branch commits always come with a branch revisionId so we know where it belongs to +The behavior of the strict verifier will remain the same +Attached is the proposed fixes +Hi TonyI collected almost all ids for Pig committerssmsjulienolganchandecpradeepkththejashashutoshxuefudvryaboyalangatesrdingCan you add them to the right group? have now all been added +That may be something to handle in to tighten things up +I threw some code in just because it was an easy place to experiment from Eclipse I do not think this should be the final resting place for this code +I obviously misunderstood the code +The issue is because of the documentation mode selected in the consoleView windowTo get the console View change the documentation mode to IE as described in the attached screen +However we dont enforce that in broker startupb +Closing this issue as per the reporters comment. +so this makes clientmode not make changes to local storage is there anything else i should be noticing +I checked it +So that eliminates NAKACK from the equationUNICASTFC is better for state transfer but UNICASTFC is much better for smaller messages +Committed to trunk with revision Closing issue. +Thanks a bunch Army and Bryan +Attaching patch a which in addition to adding support for running off a distribution where the test jar is stored in a different directory than the production jars cleans up some comments and renames a method getInstance to newInstanceNote that neither having old distributions with split jars nor having distributions where potentially all the jars are living in different directories are supported +Candidate for SOA Documentation +The type of problem we want to catch is described in +name cloud print shows +And it would be nice if the backoff period is somehow dependent on either namenode load andor cluster size +What is the scenario that exposes this limitation? We dont want to break existing apps just to fix this bugFinally it looks like you are adding another potentially full loop through all of the focusableCandidates +I have created a small patch which integrates aspectj Changed the dependencies to aspectj Fixed the tests Use it already +Attached patch fixes ant tar for me +Not angry at all +Yes the user should still be asked for a password +to have it removed +This issue depends on to make the change events for jcrsystem content appear to listeners +Committed revision +Jonathan do you mean there is no need for array index lookups into the index keys +Changed summary by request +Hi DaveIm not sure I follow the scenario entirely +Ill fix this +I understand this is a negative test +Slava You cannot copy Sun licensedGPL licensed code into our code base if it is more than a few trivial lines that is I still do not understand why we need to use to save these files when eclipse has a perfectly well working save functionallity? +Stuart can you please take a look +Here are the steps I have to do when checking out a simple maven project give it a name there is no default name suggested by the New Java Project wizard specify the jre set the srcmainjava and srctestjava folders as source folders change the output folder I prefer to separate it from targetclasses so the too compilers dont collide use MavenEnable think those steps are worth a wizard +generated once and you simple the generated classes in your repo? That would be an alternativePlease note that Andrew told me that is is not sure that the xslt approach is coming into the final release +It would be helpful to have the SVN references for the updates +However it still adds the login successful message which will display the submitted is fixed in by the addition of the already logged in event +I think I am running into the same problem with version and maven +let me run it past Jenkins +Sudden thought +NRT Reader shared live with getVersion isOptimized get get were all broken. +Committed to and +Didnt observe this bug on my Build Setup +We build the web service within the IDE and exported it as WAR +batch transition to closed remaining resolved bugs +Ill back port it to shortly +Markus Jelsma reuploaded my patch back to JIRA and refused to delete it +fixed on trunk plus a bunch of other cases involving inheritance and it works perfectly nowDo you happen to have any plans to port the fixes to a or would that require to much effort +patch replaces each in a directory with an +I think the problem is that the key in has a space at the end +Files mentioned in the previous raised to cover bug +I changed the version number +whats the problem with keeping it for the experts who have configured their application so that they do have the level of control neededStill unanswered is there a specific bugdesign problem with? I mean if commitWithin are not enabled and the client does explicit commits is there still some bugAnd maybe rollback should fail if or commitWithin are configured or the latter is explicitly used on update requests or in update documents +See my previous comment. +Post upgrade to later releases the conflicting opcode is not usedGiven that a tool that works with the opcodes seems unnecessary since the problem is only in alone +But since we do we can indeed define the grammar and get all the verbs +Yep at first glance that looks similar +But if dir exists too it creates dirdir +So please fill in another jira for this the way I wonder if we should use process goals instead of generate +No real reason to submit this to Spring now so Ill close this +couple of spurious newlinesI intentionally added the new lines to separate the different groups of fields +The following interleaving will lead to the tserver seeing the udpated compaction count but not the iterators +The most important bit to focus on is the quoting. +Yes this lack makes mutexes impossible to use +Pushed to master in ab. +This is blocking us from using +found it patch m resolving this it does not happen in torque +The patch is correct +Brett and I had some back and forth trying to debug this but hit a wall and I havent followed up +interface is now implemented + +Please review and commit +Thus the number of active instances in total can effectively grow unchecked and thus an OOM will occur. +Take a look at class in +We just assign an icon to the server and WTP uses this one for all the bars views dialogs +To make it work now it is possible to use tdataTable preservetrue +Id rather see surrogate kscf names hereI hesitated +Thanks Ramya! +ignoring exceptions you are right +This patch for g jetty and patch for g please help to verify it +This attempts to avoid duplicate efforts +Do you want me to close this issue or do you want to leave it open for the backport? +Check if snappy is loadable before running the unit test +Targeting to +g CLASSESFILE CLASSESdonecat echo i tr sed s +Just to add to this the issue seem to be that AS does not have the JMX features we depend on for adding url a move of M to d it +This patch will resolve the problem +There are several reports about various incompatibilities due to poor support of by some web servers proxies +Fixed in bugs with invalid fixed for release as fixed in. +Yah +I agree +AS Tools will work fine without Project Archives but for zipped deployment project archives is required +Hardy or Emmanuel please answer the questions in my previous comment +So closing this issue now. +By the way the pull request containing a fix for this is available +Attached the patch for the reported issueThanksHasini +Fixed the bags +I know lots of users that do that +Martin was right Ryan that was what I meant +Great catch Eli +so this is a new patch for this jira +guess we cannot merge this patch within the meaningful time frame unfortunately +If all good please do go ahead and commit +The fix looks fine. +That exception stacktrace does seem to indicate a bug +Use an explicit type attribute in +lock on client applciation while consuming another queues messages introduced +Oh looks like you want more than +Inside +I also created a single patch file for all the modules which I uploadedThe following files have been changedCheers reuploaded according to Lewis suggestions +Yea wed really like to move onto asap +Ive deleted the file but left the full one for the minute +Move to +Rescheduling +Needs parameters +Update wiki as issues +How does that apply? Boo is not C and the syntax used to declare extension methods in C doesnt have to look anything like how C does it + +Note that some of these instructions are due to aries being under svnpubsubcms system now so the CMS needs careful migrating +Actually it is a very basic function test and it should work all the time +Committed the a patch with revision +Also much of this stuff has default Eclipse Javadocs in them + Please attach the issue as a file +How about not blocking by default but providing a w or similar to request blocking wait for commands where it makes sense +So CXF works ok on but the hang is occurring onWhat else have changed? jdk jdkNo one has reported issues against on TomcatJetty but may be you can verify independentlyAlso perhaps it is worth just for the sake of trying to narrow down the problem turn the map into a regular? thanks appears that this is some kind of subtle JVM bug +I think you can solve this more elegantly by having a test profile with the extra set of dependencies +You may have tested a previous snapshot already but please one of the upcoming snapshots in order to check this different solution for the same problem +Hi HarshanaThanks a lot +was fixed by +hmmm +I believe Terje had compaction turned off during a bulk import +Just committed to. +This kills the benefit of CDNIf you want to bundle it then download it locally and add it in the bundle as orI guess someone will want to be able to fallback to a local if there is no connection to the external url +Guys i think this needs to be implemented +Invalid since the changed contract for RootTree now explicitly allows this +This might end up being a feature but Im adding it to so we at least explore that option +Only the subject line for the mail notification of Prepare build is not configurable +a useful example where x perms would be useful is mounting hdfs as a fuse module +Thanks +Hopefully it is moved forward now +Ill look into this issue +I found that the first rev failed to build when applied to a clean workspace +So if you can give a xml output of just DFS Summary we can include it in our dash board without putting any load on namenode and we can use the same info for some of our monitoring scripts also for checking how many nodes are livedeadLet me know if it is not clear. +If this solves the immediate problem we can revisit this as more of a design issue later and come up with a replacement +Here are the results +Heyjust checking in you left a comment on the github issue I repliedAny feedbackThanks? +Will have to inspect why is this the case +The issue is fixed in a PR upstream but thats not getting merged +Thanks Rick +Please fix the logging to be human readableCan you give a patch against trunk too +Completing the issue +I use my personal computer for work and had the file template set up toad work boilerplate +When we are to cut a release we will look at all the bugs fixed for Camel and backport the bugs that can be done and make sense +Closing as invalid for reason Syed mentioned + +EAP is using Supported and Tested Integrations +Yes it took me a try to two to understand how the table is setup +Committed revision Sucks to rush a commit when an issue is under discussion +Reimplemented to download sourcesjavadocs of particular artifact on activation of +No I didnt make any changes there +I thought there might be some possibility of working around this restriction byhaving a trigger on A that updates B which then fires a trigger to update A +bulk defer open issues to defer issues to +Regarding to the speed the axiomatic retrieval function should have the same computatlonal complexity as the default function if we could compute the average document length at the indexing time instead of search time +I know that it has since stabilized but I am still nervous about such a large changeUnderstand that it is a big change but if we want to address this issue we need that patch +Feel free to if you can reproduce with the specified or otherwise provide more of I could make it work withWhen will be released +So I am looking for the next versionHave a nice day +Some strange things about this error The contains the following Plexus Archiver Apache Maven mockbuild Application Server JOPR plugin Application Server JOPR plugin Note that the field very clearly has a value +Unnecessary attributes are absent for toolBar and tool +Illl attach jstacks from this run as well +reset for a return to the write followed by commitflush +This patch solves the first problem and demonstrates the second +If you want to test mixed modechnage the mode entry on the Config variableInteresting fact I was performing my tests using the attached Java codeIts main responsibility was to establish the external connection poolsfor the SQL Code +If you still have issues please file with more information against EAP +Q Could a JAAS policy be changed dynamically such that the original code work normally +Any chance of a patch against a trunk refresh? Thank you +WernerIve created for the fix of query element placement and as well assigned to me since the same patch contained this implementation try to provide both patches over the weekend +Interesting that you would close the ticket and mark it as incomplete without communicating your need for a license grant and without giving me the opportunity to provide that license grant or to ammend the patch +You should decouple provisioningdecommissioning from the workflow stuff where possible +Patch applied +All the reported issues against have been resolved returning to release check listCheck list Build tests on Mac Oracle Windows and IBM Java +Suat has applied my patch for this bug +Imagine we have a union of Foo and Bar +Could be closed I guess? + +Thank you Brandon +Committed to the trunk as revision I dont intend to merge this to any of the prior branches at this time soIm marking the issue resolved +Whoever reviews this needs to test that functionality +I downloaded this latest SNAPSHOT and successfully ran the tests without regression +AFAIK they mean at most its harder to get to this gbean from other apps +Should be available in tomorrows you very much Juergen +Descoping items not being worked on for M into Unknown Fix Version for now +The failover mechanism is useful for applications with multiple database nodes when the data is actually replicatedsynchronized by the underlying databaseSo DN will try to connect to a slave automatically when it cant connect with master +optimizerTracing The new tracing is intended to help diagnose performance problems in production +bulk defer of feature items from to +Look like it been fixed in previous SDK +Fixed in SVN revision . +No longer see this +thanks! +This is because when we use a default package the grails system throws a warning message and waits for the user input +When all else fails try to parallelize the scheduler dispatcher +Makes sense not to have the class name in the exception message +servlets delted from SVN trunk +closing. +I forget if you told me +Make it a subclass of +Eric I backported this fix to the branch as wellI assume this issue now is fixed and can be closed or do you have additional pending changes in mind +Antonio your approach looks good +We can run sql script when a datasource or an emf is created +AndreyPlease review the documentation for technical accuracy +I have questions before I commitIf lists were not handled how was the working at all for JSONI cant remember which part of the social api needed a list as a base container it must be there somewhere +Thanks for reviewing the Patch +Fixed I came to the same conclusion during the weekend when I was thinking about what I did last ll fix it by providing multiple configurations +For instance the build currently dies on ARM with the Oracle JVM running targetnativetestlibhdfsthreaded +If you just replicate it in the same way like in trunk it does not make sense +andrzej thanks I like this design +Using custom solves this. +You will open another issue to do the doclib +Hi MaciejBug or not how difficult would it be to add a clean implementation of this functionality to the authority connector? Is it something youd have time for +Reduced priority of non critical issues which have current issues that arent going to make it into to t for me +Hehe maybe not fixed before monday +btw all tests pass with the patch applies +My sense is that the table compression code has a number of weaknesses and we should devote some serious effort to overhauling it +plist file only inside the +Weve been using for a while in serializationOops hadnt though of checking there + +Could you add a check for this as well? Right now Webstart is not working with Pivot as far as I understand +As weight package screen was reverted few days back again introducing weight package only screen with following functionalities Now user can weigh packages and enter dimensions with weight package only screen only when he has gone through Verify Pick process Now pack process and weight package only process are independent of each otherSteps to test Create a sales order should be in approved status +Hey guysThis is a feature that I need implemented in Android as well +The name of this action in the management console is also vague and potentially misleading the reload user data in fact seems to only reload the contents of the user rights file and not as it may appear to suggest the usernames and passwords +Now pending a new release +Hence resulting in exception +I just did a test patch it works patch with thanks to Christian for fixing all these ticketstrunk +Tested by Adrian +updated fixaffect version need to verify in latest build about result +Could the counter name be based on the named output rather than the base filenamePossible +bq + +This list of dependencies is then used to assemble the tar file in Can someone please review and commit this small change +For you John I assume this is for +I tried +Thanks a lot Nicholas. +Attachment has been added with description Problematic file +managed hibernate session availables Create method is IMHO the best place for this +Wow! After Koji figured out the cause of the problem he talked about the same timestamp solution as what Doug suggested +As far as I remember from my ongoing refactoring of s and s there is still some code that is intended to deal with enums that I actually do not know where it is +Upgrade to to pick up the fixes +Rebased on trunkNamit JainThis is not backward compatible functionality so making this as a default behavior seemed to introduce problems IMHOIll change it if you insist +working on some verifications +My current theory is that this is not actually a loop but an exponential explosion +Once this patch is approved and committed Ill clean up and get that committed then open a ticket to add the interop tests back inThanks should have noted the new patch contains an implementation of the file object container format specified in +Verified. +This is already implemented and the test suite seems to be stable +moving into modules is an orthogonal discussion +Test failures are unrelated +Patch has been attached please help to review +Unless somebody tells me otherwise I am assuming we have no intention to do a new release as a result I think we dont need to produce or commit patches to that branch +Modified other classes to make this happen +The second option also runs into fun issues with the possibility of having the same commit show up in multiple emailsThere are also a number of variations on each of the three options +Does anyone have any objections for me to take a look at this +RohitThis patch is against the trunk version can you test itThe patch fixes issues with the view indexing of the links on the keyword results page which is the first obvious problem I could see +I think that we would need to introduce a new option something like map +I think theres a bug in there but Im not sure if its related +We currently the version to the latest released version of so even if the list is empty you get a good default value. +Fix for Trunk Fix for +Some of the commercial Agile planning tools have a distinct backlog queue from which one can dragdrop stories into various the principle of the simplest thing that will work please consider starting with a single Project Backlog that would be used as the holding tank for incoming stories +Himanshu A thread dump when its hung? Might tell us somethingLars Done Im triggering and watching builds +Fixed +Hum +PaulI think your idea is better a simple Geo entity boolean field should be enough +Could you please provide a testapplication to test this patch +I only see the messed up icons under leopard +I have written test cases for the options available in +I like Asparagus +Youre making some innacurate guesses about whats going on Tapestry should not care about the method names +I tried setting the basedir config property and debugging output shows it being set but the report seems to ignore it +decided it is better to have an solution for knows now two modes one to define the Reflector class and and one to do the usual class loading stuff +Updated test case showing how to make this work on a SNAPSHOT +With zone element it not work on +I can only assume javac got irritated by the wrong entries or has an optimization for inner classes that could not kick in +do we want this for alpha? is the forge stuff in here okstable to release +Max could you please review the patch before committing to M +Spec updated rev +Could you triage? Thanks +Its quite minor and should not affect anyone adversely +Velo has done a great job with but its way more than one man can handle +Arguably HDFS should move into a set of packages under and these would go into the namenode +Thanks for comments Nicholas +Does anybody know how to update extpaths stuffs +I dont feel super strongly about this so if this is your preference Arun lets go with that +I suggest to let wait if one replicating table is not present likes the scenario of peer cluster is unavailable +what do you mean by current working directory? Do you mean something in the local OS filesystem or a current directory in the DFS +I think this has been fixed we can set this issue resolvedthanks M +Have you tried beta +Bulk closing stale resolved issues +hopefully Im misunderstanding +Now the NN closes the file and the header sector that has the correct LAYOUT VERSION is flushed to the disk whereas some other pages of the file encountered an error while being flushed +What happens after we log it once and a node tries to connect? we cant fail silently the node will not be able to connect +One way to do this is to use Maven for building the iiop module since this resolution problem is fixed. +I had the problem with Andrews patch also so Im not sure whether this is just me or an arc issue +oops wrong issue +Did you install the File plugin? is different than havent heard from this person since its possible that the plugin wasnt installed +Well the Jasperreport can work well directly with Hive official release now +Patch implementing reusable analyzersSimple filters have been converted to use the new APIComplex filters such as synonym and WFD have not been converted + +This bug has been marked as a duplicate of +the next time Ill create a test for it +Thanks fixed. +Yeah this is fixed in the svn trunk. +does appear to be a duplicate +This has been fixed on trunk +James assigning this for you since you said you had a patch +We should implement this as an option for sure +Ok a dispatching could make sense +Will be available in tonights snapshot feel free to give it an early try +Xidv for markedXid +I have just checked the documentation and now there seems to be no mention of client side load balancing +Working on this issue + +Just realize this is a duplicate of +Issue the failover or stopMaster operation on the master database instead +Your patch was committed in revision +Too much work +These can all be executed relatively easily from the command line using the new command of how to use them needs to be added in particular explaining how to load and unload them dynamically into a running program +Please open another ticket with some concrete examples if necessary +If we decided to go this way Id want to then structure our unit tests so that they check that the separation is maintained +I was thinking a hard error +Also I agree with jgray about leaving out mention of allowing nulls in the constructor +Unfortunately this was a bug in CDI not all session beans are passivation capable so CDI assuming they are is just plain wrong +By the way I hope issue for Japanese PDF file is talked at +I also am using xdoclet as Frederik is too +Closing the issue for now. +I dont think so the patch could raise these failures because the scope is just adds the new property in a xml file +Well use the workaround we foundCXF doesnt pull in xerces +merged back to +You can ignore that null in the log message for now +But it would be great to see a Scala sample project that reproduces this to +Im not sure how this differs from the way formcheckbox works at the moment because I seem to have the same problem with a checkbox +I am running not the unit testsI do not know how Owen derived this is related toI dont see any exceptions at allI do not think we can release our distributed software if it does not run on a single node +The new patch reflects the suggestion +Matrix of operations and what it takes to make each compatible with the +Will post a patch along the lines you suggest +I dont know the exact update version where it was fixed though +It fixes a bug that I introduced on in the patch file I provided yesterdayAdrian +Although Im not sure because changing the file didnt do anything +attach my patch +If the property is set to true then the plugins behaviour remains unchanged +Derived from +Ill try to fix it soon +Thanks +This patch allow the resources to be loaded in a jar as a diff file applied from the root of an old version of the projet I am not sure that it works today as the set of the three modified files OLD and NEW patch refers of course to the old sourceforge version of +Also I removed the cloneable proposal from description because it has been included in the new patch submitted for +The JAF handling code was lifted from. +I am using Chrome Version can also see it on Firefox what are you using + +running udig with a nl en flag will use english as the language +I wonder whether that would have been truncated guess it matters what state the original project files are in +Using this tool we can determine if the namenode is not purging the list of blocks to be invalidated +Or is that no longer a worthwhile objective +Pinaki Agreed +Maybe we can do that for and keep this jira as major for future effortI think this will address the vast majority of cases +Could you please fix only EOF exception problem at the moment +As such one could consider switching to the classes for this patch +Thank you +That sounds fine as long as we define what we mean by. +has been marked as a duplicate of this bug +Will send patch failures are patch to reflect the current trunk +Thanks Kim! Those are very helpful suggestions +Theres nothing anyone directly involved with Solr can do about this its a generic problem with the entire wiki installation that already has a bug filed duplicate +Thanks Paul! Ill get to these soon +This is a more generic problem none of the Master operations goes through replication +Not sure if I can and may mark issues as wont fix here +Can you post a patch? Do it against the trunk please +Details tab had no problems at all. +PR for master merged in master +Instead I changed it to explicitly catch those and rethrow them as which is probably more appropriate than RTE +Bruno and RayRay your patch is in trunk +The line of was commented to avoid a compilation error +Axel Thanksjdb +Its an old post issue but I agree with the issue creatorIt would be very very convenient to be able to make a release from an earlier revision than I only used CVS from mid s to early or so but I think there was a way to checkout a branch +Now the sucker password in must be in encrypted form +How about just disabling hibernate cache for? It seems like itd be fine but I just worry about what would happen when two threads are trying to add a Tag for the run +Before you reject this case you should have a look on the reference implementation +Thrift should handle it automatically by adding a default value to missing parameter +The final patch I am sure will include changes to master files but I thought in the meantime folks can look at the source changes I have made +But rules are not getting loaded while starting sonarI added following value inCDSSWSWSLatestSourceWSWtargetsiteparasoftCDSSWSWSLatestSourceWSWsrcmainjavaCswCdocI am using sonar Thanksrudraemail +My mistake. +Also what does this add beyond hitting the config end point to get the list of other end points? Though config might be gone or named differently +First draft of iptables patch I still need to write tests and do some live tests but I wanted to get this first draft up for comments +Once we have a release candidate well do more testing +This feature should be considered as a Usability enhancement for issue +A flag disabled by default and activated by a web config param is valid or a new mode which is the same + +Got it +The current design requires user to turn on both namespace and schema for Schema Validation +Same request but not same LICENSE file name +Hmmm +Will leave the decision to merge w Branch between Dimitris and Carlo was not ported initially as not indicated in Fix Versions +I checked the code and it does appear that were unregistering +Fixed in i set CONFIG INT The Cache cant response gzip header the gzip takes request with VIA as from proxy and refuse to send any gzipped content +What is about the problems mentioned by Daniel from Mar When implementingIs this not an issue +dropCf was intended to unload a CFS from a table instance as indicated in its comment +Changing this to a tracker issues as I dont believe any changes should be made to the CDI spec to accomodate this +thats what the bug title says +First go +I love inconsistency +But blocks can be in the read cache of the underlying database because a write transactions reads them into its own transaction view of the data +It is burried deep inside our code where we actually create the log +In all honesty Id forgotten that this class was even still in the codebase and I think the correct thing for us to do is to remove it +Albiz! +Mario Ivankovits from they use scp with jsch saidIve never seen this problem and there were no reports like it be that the time after creating the session and requesting a channel is too long idle and thus the server ended itJust a wild guess +And this is only after submitting a st order +The problem with overriding the various xmlrelated classes is that these are used all over and not allcontexts are correctly limiting the scope of the classesthey use +One to download the patch one to test to see if the patch applies and one to do the initial build +If you have any other suggestions please post them +While annoying the issue doesnt affect users that much +bulk defer to defer to +Why add the complexity to a stabilized code line +My preference would be to configure in the +Done. +Perhaps you have not configured the project you add testsrc as a source folderIs testtestsrc also a source folderWhat happens if you remove the first test in the location text boxCan you attach a simple project where this is failing for you +There is an advantage in multiple lockable over the same as being equals so I think the implemementation in your comment is just fine +Ooops yes the group is this is a typo mavensynchomenexusnexusstorage rsyncssh Filippo De Luca +This can subsume +Attachment dont love adding this logic directly into the encoder to deal with complex features i sort of prefer the flag but the patch is simple and we do do checks for complex features in other places in teh parser and encoder so +All tests now pass +Interesting +Updated one attachedWhen build will be successful I will upload the and versions +Its ok to show in this easier to read format +Only +If youre recompiling already Id rather have the improved internal decoupling and force the recompile +I should be able to take it from here +Uma the second option makes sense +Attached fix +I think we need both in case for some reason the znode makes it over the migration +We need to do something about this +applied thanks! +Steve I wonder how things work in Oozie then I can compilerun testcases using and without changing imports +I am using the trunk as of revision for my environment +A patch file for TS that allows the background fill config items to be altered on a per transaction basis +These features will ship with Axis release + +I found two defects in the process and this change did not introduce them if Find window is open when Frame is closed the window remains open and the process doesnt die and Find previous does not match the first occurrence of the pattern if that pattern is at the start of the text area +Can you please provide a rule that causes this error for youThanks like a duplicate of +oops I was wrong that isnt causing the problem still investigating +Thanks Ashutosh for fixing this. +A year ago marked as deprecated so I just think that now could be a good time to throw it away +This should be fully tested in the EDS QA process +code looks good +Of course also Vinicius can help +Removed in keyword bundles from core apart from default en locale +This is the closing +The patch is manually tested +If we think the small duplication of the assertOpen isClosed code in and is worth creating a common superclass we can create a common parent for them called something like Container or somesuch but I dont at this point see the wisdom in getting rid of them +s localhost a jmxrmi u admin p admin gettypeServer make sure that is the same username password combination stored in the server editor inside eclipse +Regardless of whether its a NIF or not I still think we should look at this in terms of modifying a JSON binary as opposed to concatenating two random binaries and crossing our fingers +Patch using the wrapper class added by fix of the issue +weijin is the queue for events coming from other threads? Ive been using the patch for a few days now and it has completely resolved the segfaults +In terms of relevancy and scoring having a real title is not the same as populating one with the filename better to know what is what +Patch looks good +With this change we no longer depend on the ASF Jira instance being available during fresh builds +Ok looks like I should investigate how your scenario plays out with my last batch of changes and possibly revertadjust it +Reporters of these issues please verify the fixes and help close these keys were fixed we would not focus on the ajax file. +I have added a test case to cover when a relative path is used in the +So if and are not merged to then I suggest we instead just remove this test from and still merge this patch there to keep protocols the same between the releases +A testcase has been added to the test and test +updated to apply with latest codeNow that we are into I would like to commit this soon +Okey so this is pure upstream issue then we dont control this dialog +We could do that but since theres a version of this function that strikes me as the way to go +Patch to fix this +See new UI in +but not used is depending on what every session store you makeBut i guess the session store is the one that makes an undo manager +I think we should make firstName and lastName mandatory for service updatePerson +And a good idea is in the WARN message to say that +Thanks ChristianI verified tested and committed your patch +Youre thinking of NEGATIVEINFINITY perhapsThank youYes +is ready to review +Its not the cleanest but Im open to alternativesIve probably missed a few spots in the MR framework but Id like comments on the approach before I go further +They must be appliedconsidered in order as they wont apply out of orderThere is a seventh patch coming containing the various changes and additions of commentsdocumentation +But deployment framework is totally differently and EAR deployer is parsing this you want to further continue this discussion then please start a thread in the related to how deployers locate and parse deployment descriptors belong to the deployers project parsingvalidation can be discussed in XB forums but in this case IMO XML validation is working properly +Dropped and starting over with with as there was a showstopper starting the new release is out! +Lets leave this tentatively as remove decorator until we get closer to releasing and make the change then +Submitted a revised patch for this one removes trailing spaces from adapter params +We could do as you suggest as long as we do it before we call resize? So wed have a conditional where if the size N times the buffer then instead of resusing we go get a new BAOS +Then open one of the entries from the bookmark +Done +Not a ton but still probably worth it +We should most likely catch the and throw a special error to clientI think this is the case where a client thread decides to do a local read which will perform the read in the client thread rather than in the stages +This allows to override with additional config options on the if necessary. +It must be the closest to the Scans stopRow which is less or equal +installed so youre only dealing with the plain Camel versions there +This is needed by and other related jiras test automation +Ive tried to reproduce it with the supplied configuration +Updated fixes the dependency problem as discussed in our mailing list +Will post patch tomorrow +Patch applied thanks Kristen! +patch new files as previously described +We will open additional to deal with specific issues as we go forward. +fixed now +Im closing this issue. +First cut of the completed HTML pages +Suggested patch for this issue +Problem no longer exists with new version of iterators. +See prev comment for fix +Forge just uses the same maven resolution as maven itself + really need some test for this stuff +All hibernate libraries lucene slfj libraries are needed +Ill build master and will check for it +Change still required for topic exchange +See which is currently scheduled for after but before +Or even no valign at all thus should be left to the users can be changed with as a breaking thanks for the patch and support! +The agreement was to have replace it entirely +I see thanks +I didnt feel good about applying the change that continually created new so I addressed the thread competition issue by removing the synchronization altogether the worst that can happen is that multiple threads may write the same value into the map but once the initial few passes of this have taken place there should be no further effects beyond the avoidance of synchronization locks +See also +Jon I want to commit this +It aint notpretty but that appears to be the only option we have that does not require APIchanges +Hi NicolasIts maybe too early but dont we need to clean much things if we go the CSS way? +Problem w caching +lingerMillis +whats the purpose of the new get get apiIs there overlap here with +I sent the ip clearance mail. +This is what I meant why not to make relativeLocations cache also expiring like it is done for composite +Hi Chris +New patch which addresses the review comments +Patch that includes all mentioned improvements but needs cleanup documentation improvements and junits +However we can start with adding a new field and do the optimization later +I agree with Bryan I think its better to study more before refactoringI unassigned my self from this issue for nowThanks. +Its even more complicated than on Oracle because you have to move the LIMIT and OFFSET paramaters at the first position even bevore any parameters used in the WHERE expressions +Is this causing a problem? If youre running a combiner then progress is reported for each record +bulk close of all resolved issues in preparation for release. +Lets integrate it after the code freeze I you dont mind +Thank YouRahul +After all these modifications I get an Illegal Annotation Exception +Attaching the udf CiemiewiczYes UDF takes a bag should work even not in generate statement +Hi NicolasI tested your patch +So this would be? Thats how hints are TTLd currently +Used your dump and loaded it into a encoded database then I issued your request after properly the filter and it works just fine +Can you verify this is still true for nightly builds? There was lots of changes fixes in management for messaging subsystem +we tested using conditional compilation to disable stats collection +As far as I can tell it is working as per the spec +version that has the fixes from these two commit ids +I changed a number of things There was a load libc and other java related upgrades for my OS which i installed I started it with the Xint option and Xmxm and I reduced my cpu overclocking +Needs a load of testing +Perhaps override setTaint in and friends to raise an exception? And perhaps we should modify setTaint to raise an error whenever Java code tries to taint an immediate +Actually since Alan is the reporter hell have to close it +Thanks for this enhancementBest regards GeorgeThe fix looks fine thanksBest regards by Andrew. +I have merged these changes with and attached the patch to +I verified that this is still an issue with +But ifits better to have only one service i see no problem to merge them +Its not fun but it needs to be done +Thanks Michael! +So it must be something we downloaded somewhere but I have no idea where I can tell you. +for Option +Note that the method does not return fields declared by the interfaces that the class or any of its superclasses implement +patch to allow users to run dont think this patch really belongs in the codebase +defer to +See +found solutionencodingStyle in wsdd file should equals blank string +This patch introduces a new file which is just a factory for instantiating query objectsAnd the class is modified to use use the factory to build the final queryThis is backward compatible +Linking to the broken credit link bug which is likely due to the same hope this has been fixed by the new Cocoon in CVS +Test change passes locally +hmm +UI is sending following addCluster call +So how hard would it be to get this in Ant instead of Maven? I dont see Solr converting to Maven anytime soon and I certainly am against build systemsFigured this one out already the new version will solely be based on AntAs for the dependencies the two main dependencies are indeed GWT and +Lots of changes in it anyway so we might as well do this upgrade there +The dependencies should not be in the classpath anymoreWould you mind applying it to a local kernel copy and see if this works better and solves the issue you found +Will close this when we push the site +And Im not sure the caching here is not a +Agree with John +I observed the same error on sun jdk +Whoops +GreatThen this issue is ready to be closed +I should mention that the Connection Shutdown timing is only a conjecture I am currently not able to track down the life cycle of where the bug occurs so I am just guessing that the exception happens at shutdown +please attach console output from ant d doc run on windowsplease do this after a ant compile run +I presume you mean in trunk? I dont see it on my release + +js file you check in Matej resolved the modals not closing for me +Also it would help if you use GEOTOOLSDEVELOPERLOGGING log profile and attach the log file in the future. +This issue is probably outside the scope of this jira though + +Never mind this was a misunderstanding +I think not being able to create a page with a lower case name is an issueOften as in our case of LDAP items and as in is the case with manyprogramming or technical values case sensitivity is a requirementPlease reconsiderjimJim Willeke +Yup I believe there is a plan to support other testing frameworks like +Oops +We are putting our main development effort into JAXWS +Corrected behavior or in two ways When reset it must also reset its parent context When you have multiple chained predicates all but the last one must collect their individual node sets. +If some match user is offered to update other projects as well +Niall I change my position and am fine with Martins suggestion of clientValidation +The default wagon manager couldnt used there due to the difficulty to change it and some bugs but it has better testing +Can you take a look at this? Happens only on devcloud +John thanks for the patch +On and even if I set the soapaction as long as the wsdloperation is set then Ill get the wrong resultFixes committed to the various branches +I will be uploading drafts so I can get early feedback +This issue is closed now +Will try it on the weekend and report then on the fix +Yes my was generated under views +Open it when unsatisfied. +Thanks +I was beating my head against the wall trying to figure out why b would be expensive since it looked fine everywhere +I will try with the new version of WSRP and let you know how it progress +I suggest we reject this JIRA +Ok +I am attaching the patch here +Were trying to increase the available memory in LIBHDFSOPTS well post if this fixes things. +A common spi would be great BTW. +Need Matts approval before scheduling for s add this +Is it possible for this to remove an endpoint that it shouldnt +There are problems with index replacement in windowsWindows does not allow as to delete the index folder because it is being used +Has been fixed there in +They are called before and after. +retrying wont fix anything +Thank you for your contribution Hari +just committed this +verified +Thanks for applying so fast Robert +Lets start discussion more on this to move towards closerThanks Vinay for your interest to work on this in your free timeI will also think some options to night and post if get any good thoughts by then +assign to Daisy review +The problem was that in xwork the parent list on the object was in fact a set and when searching for the default interceptor ref the first hit was returned +Whats a pin +Adding a synonym for as cssrs corrected the warning message I was seeingMar PM get Could not map SRID to CRS No CRS for srid But the behavior is still a problem +So its a latent bug. +That the normal way a bug report works +MOO +I am these methods so that they forward to overloads which dont have the typeName parameter +As it stands right now if Communications Platform is supposed to have its own category we will need to make additional code changes +Can somebody pl comment on how this ties in with HDFS permissionauthorization? There is a small subsection in the doc about this issue but I am unable to understand that part +Thanks Mathiasyes you are right! That was meant here +Marking this out of date +Trunk srcmainjavaorgapachehadoophbaseutil I have attached the patch to this JIRA +Oops Its caused by missing class files in plugins and during build process +Thanks for the patch. +Hi GuysI have created a patch for this issue +Then it flushes the logs to file +New patch with the renaming and removal of Map in favour of a simple SetAgain ready to commit +bq +shortterm fix where as now takes an as argument. +I can handle the other CLI changes +Closing it. +As far as I understand if you patch is applied min and pref will be filled with sizes of the last component in list for given index +Close. +Heres the proper patch +Can one of the committers please review +Erick couldnt reproduce it either +Not very simple to implement unless we have some more contextual information +Can this work still make it to +Patch for this problem +Merged to the branch in revision . +Thanks Chris and Konstantin! Resolving as fixed +Add more changes from the finding of integration tests Resign a task ASAP instead of time out when split returns resigned status Adjust default configuration settings for replay such as rpc time out number of retries and single replay operation time out Add synchronization for logSplitting during master initialization because there might be a case that multiple places split a same wal +Just need to get rid of completely +Review OK +Updated to use the rpcaddress field +The difference now is I am running on Linux and you are trying on Win something you can do to help me +Ive gotten rid of and replaced it with which is an immutable object +Thanks +Can you please add description throughout the comments and assertion messages that the value is since it is not discoverable in other way than look at +I rerna and looked at all javadoc erros and none of them have anything to do with this patchI did run sanity tests +I pulled this back into +The update to the canon for derbynet is okay +It would be interesting for you to try that as well but I think these results are sufficiently indicative to conclude that Ive got it basically cracked +See attached screenshot form IE provide us full page source +Thanks! +I think I can take care of this one +Replication is the failure point +modified the handling to log the situation and properly shut down the component regardless of failure to unregister the service or not +Duplicate of +Tim do you still see this as a problem? +The attributes Map in is always instantiated anew for each project it doesnt seem to get populated correctly from the pom +Its important to remember that the control files are dead stupid and I dont think we should try to them +I meat the two changes to are NOT necessary +tested +This possibly is some sort of a race between JMX Listener and some other resource +resolved. +Can anyone confirm +Odd it works fine for me but I did have to remove the spaces after the header names +Several more names in +This is how is supposed to have been addressed but couldnt in +This will help keep implementation details out of the way +BTW by default should be true +I only know that Hadoop local mode can run on local FS +Ok for Patch v but may need rework whenif the client library brings support for isLoaded state on collections such as Userdomains +i wasnt sure if we could since it is a part of the JDF thing +I can make the changes in Javascript +I removed the fix version because this specific test is more of an update to the release process and will also be automated as part of that effort +Patch has been committed. +Latest list of unit test failures on branchTestcase classTestcase MethodProbable Causetestorder bytestmerge jointestskew jointestPrune columns issue with order bytestskew jointestskew jointest needs to be fixedtest needs to be fixedtestCounters need to fixed hadoop issuetestCounters need to fixed hadoop issuetestCounters need to fixed hadoop issuetestCounters need to fixed hadoop issuetestCounters need to fixed hadoop issuetestCounters need to fixed hadoop issuetestCounters need to fixed hadoop issuetestLocalCounters need to fixed hadoop issue +Can someone please assign this JIRA to me +Closing as such. +For the Spring DSL it would be really painful to keep the old and new DSL +This issues contained patches for two unapplied patches to the branch +Synchronizing on the in assignContainer would prevent this interleaving and avoid allocating and filling a new of priorities each time a container is assigned +Created for the sync issue mentioned before +Thanks Andy! +The best way to do this is probably expose a counter in JMX and nodetool +I havent reproduced this issue for subversion revision either +Verified at under too +File level data is still not kept for more than the current snapshotThis becomes clear when i look at the modified source +I have committed this. +So in the end you dont have to avoid the call because it is inexpensive +i remember you had the problem while we were preparing the docs in Neuchatel +To generational collection in Jitrino is required for Dacapo performance Feng should we keep this JIRA opened? It looks like it is not used since August months already +conf used by the +Sanne found the pom +Oh and it would also be better if the poll could do this in a streaming mode with the aggregator so we dont need to hold X exchanges in memory +I will this versionCheersAshwin +Declared dependency on +EmmanuelDid you have a chance to evaluate my response? Im sure you guys arebusy with the next release but this is only a small change Im happyto send a patch file or my changed versions of and the Hibernate DTD for my proposal +Thanks Manula! +Allowing a separator is a luxury since the effect is the same asomitting the and altogether +Because of that because no one else is currently looking into it and were so far past the initial planned release of Im going to bump fix to the next version +The default constructor sets useful defaults the properties constructor configures the configuration according to the properties +Maybe there are also another solution how to implement operationsBut current problem of CXF is using operation breaks integration with Spring CXF is actually responsible for the integration and it should be fixed here +Any possible rotation value will be automatically taken into account when rendering the page to the given graphics instance. +But still an application needs to call get first +Should we reopen this issue and add a patch or +bq +File Contains my first attempt at adding a Log class to classes in the Main module of Karaf +And Spring Source was not yet giving us resources to hire an author for this kind of stuff +Thus closing as CANNOTREPRODUCE +Closing the issue now that it has been resolved. +Data Tools Connection Profile for +The problem is you shouldnt be using the dtd for this file +Im going to clone this again if it doesnt get reopened +Hi thank you for the explanationStill you might have a problem lets say you have an entity with the two fields annotated as in your description having the values ID and Title of this book and you use an Analyzer which includes lowercasing and removal of digits +in bytes the average length of a chinese word exceeds even english because cjk ideographs are encoded as multibyte sequences in utf +Ok I understand Tony Ill do my best to prepare the package +Adding +This is quite surprising to me since I thought this case was handled +More patches are attached +downloading beta at the moment to check if the problem is there is there and parseablethere are no problems building the project just the project view and editor view showing errors and no libraries +meta +It is also personally embarrassing for me because I have been evangelizing as a superior as well as faithful implementation of RubyI can no longer do that with any confidence +You saying I should have mvn install from the plugin project dir? Doesnt it all get built and deployed to my local repo running? is definitively fixed in the current svn version of the eclipse plugin +I will submit that. +The current patch looks good for the test +I have also seen the shell give very uninformative messages in the past when it can find the class but the class does not implement +The patch attached to includes a fix for this bug +Move it back in if you think differently +Patch is also attached for this solution +Then if a region server dies we process both sets of log files +This was missing in the CXF filters +otherwise why would we have the catch block constructWhats the core reason for this rule of thumb? The only thing I see is the performance argument +Actually we are going to allow regular servlet filters but Im still the Shiro filters for better performancebulk closing issues that have been resolved for more than a year. +This dead code has been removed. +Because this is what they are +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +Investigation under progress +Patch for to use the for consistency with the decoding are we still using default charset after +Two thoughts +The patch is on top of +What about having a callback makes this a problem +Thanks Peter! +When I was doing this the first time I just ran into an issue in the way that we do it +I would like to see what others think as well +We should add a page that described hot to contribute and a link from the main page +Assigning to Scott. +change is some typo mistake +Fixed in SVN +WillemThat fix resolved my problems in the original project +Very much appreciatedYou are right it works as you say +If we are no longer doing a tech preview then this requirement can probably be ignored +Hi UmaThen this issue should also come in hadoopRegards is no different from on this +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The issue will be closed when initial support for other containers is implemented +I just committed this. +Is there some documentation code or a test case I might look atI just committed some small improvements to the javdadocs for this can you look see if its understandable nowAlso have a look at it consumes the bulk API +Setting this property will make the builder use an xsdimport rather than inline the schema +Please reopen if still valid +Ive seen some weird things with linked in drivers going exponential around this size when BIFs and NIFs remain fairly constant +Sorry for taking too long +Unit tests should pass even if native code is not available +As such how about agreeing about a approacha Implement a solution for serialization where an object that has not materialized will be materialized before serializing it +I opened at the time +smileIt can thats the point of this bug! +Raising priority to blocker please can someone provide steps to reproduce +use transactions as they are intended +Even though it kind of makes some sense that the SQL implementation is the base version in practice a database that isnt one of the two supported is the rare beastThat said if you really wanted to do a big proper clean refactoring to put the SQL implementation as the parent I wouldnt object if it were nice and tidy +Please go ahead and commit it on trunk and +Thanks Gunther! +I selected in error. +Woops +Am sure it is related to classnotfoundSent from my iPhone +It should still lock the since it violates the assumption of the method and could easily be broken during later maintenance +I was a little ambiguous +Thanks a lot George +ah now I understand +applied to HEAD branchOleg +Feel free to pop a ticket at project if you feel that we might help with the integration issue somehow +Gabriel as far as I remember the requests are to a generic test servlet which in turn executes them +HiThis sounds like a bug and is probably something that happens in more places +So it does not make sense to make that reservation at all +Replaced with runtime check for the existance of and reflection APIs for method invocation +Tidier not to try this for by testing up front that waiting for +Yuri strange you didnt see it in openSUSE +Looks like Anoop changed or removed his JIRA account +Cedric Guillaume and myself find all dont like this approach +Upgraded to +I had done the latter but you asked for the former unless I misunderstoodcan you clarify what you want +Whats left on this bug? It seems the patch was applied is it ok now +Insufficient info +This has been fixed in the shade plugin trunk just need the plugin release to happen +this is no longer an issue the behavior was changed a while back also verifiable by the attached test +This issue duplicates +But that also happens if the testproject IS definitely the first added module so it acts like the root one +However its easier for me to read the code w standard set terms +If you use JBDS its much cleaner and easier to see what youve installed vs +Added more top of class comments to explain all of the config optionsAddressed a few code commentsPhabrictor seems to have been broken on the move to modules +Thanks Shawn. +Attach another patch addressing comments from Dmitriy and Thejas +Lets do both +is performing a null check on usage id in the constructor +But this should only contain a single method of accessing a remote repository +But this is correct behavior +The is no more refered as a Model class. +Turns out it was not doing the correct calculation for reexported packages +Much nicer than referencing a beanFintan the rest of the examples should be done in this way too even the sticky handler +Punting to post +Tutorial added to the docs +Guothanks so much your patch is in rev + +Also iOS support is being removed in so we can use blocks to help with this +bulk defer to +Thanks Sreekanth! +This has the consequence though that any calls made with CP code to a GA CP or CP server will result in the error above +Merged into the master branch and into the branch. +Same as can you try again with the svn branch +reviewed and committed patch +Xavier could you please verify it by yourself or are there any steps to reproduce it? step to reproduce the issue is about removing an internal jar to reduce the plugin size +All fixed in CVS +Okay +Currently I am having a problem with Java Mail. +Does anyone have any opinions on itLeave it as it is remove rewrite +Chris can you please create simple project that would allow to reproduce this issue? It would save us some time +Is this still an issue in later releases? +I spent a bit of time trying to manipulate the configuration to avoid the timeouts but without successFor now Ive modified the problematic tests in the AS testsuite to add FLUSH to the stack to circumvent the errors +Addition reverted as indicated above +Would anyone want to try thisThanks +Fixed in both +This patch applies to the current with minimal problems just some imports on that it couldnt resolve properly +to reproduce unpack the and run mvnsite within the failure contains a simplified scenario use that the enclosed reveals this issue was fixed in +Thanks Sergey +Is this in the bin folder in the root of the platform repo or in the cordova folder of the project +Reporters of these issues please verify the fixes and help close these found this issue in latest automation runs +Updated patch as core wasnt compiling due to misplaced ant dependency. +This issue conflicts with the patch for +Updated the attachment includes NOTICE and apache headers on top of filesRefactored the package name and added base on the events pluginThe MD sum isbbbedbeabfbd As for the grant who should be the Licensor me or the CEO of Labs +Reviewing the code i see that create in should be called +Done and live. +I hope to attach a patch here shortly based on the code base + +Here is the scenarioLets say the current leader A is at zxid A participant B with zxid joins and gets a message from leader TRUNCB then calculates the length of log up till zxid +Do you have any better idea or concern +Cheersc +Im not sure i can solve THAT issue for now but as long as i can solve the majority of the cases you listed things should be fine +Nicolas how did you configure your project in terms of languages? What did you give for the reactor project? What did you give for the faulty moduleCould you give it a try to put flex in the reactor and nothing in modules to see if it also crashes +Ill take a crack at this and post again when I have something concrete +v fixes some variable name confusion in and deserialization from a direct buffer in replay +In r done for find for for select selectRejected collate +Attached with test for user is now done and tested though the configuration mechanism is going to change +As util should not depend on processor I had to introduce an interface in camel that could be used to abstract from and +Id rename the package to what do you think +If I change column with column in this last assertion then it does not fail +Tested +Of course that doesnt mean it wasnt there but I still think something changed in possibly elsewhere in the code to make it more common +I have resolved this issue as the requested documentation changes have been completed +Verified on JBT +That works around but iFrames are still screwed +This is no longer reproducible on trunk +Patching of dont think thats going to be much of a win checking the filter is the fast part its merging different row versions that is expensive +Sorry Gregory I meant to ask would you like me commit this and then in separate issues work on the outstanding stuff or do you want to update a new patch +Defer all open issues to +Sorry possibly my fault but +Thanks Knut and Myrna for your work on this issue. +Thanks Suresh +This patch fixes the issue and includes a unit test +Added new property for improved container logging +I think I have successfully threaded to the codecs and index package wherever required +Adding a div will fix it +Tinderbox run ran cleanly after commit so resolving this issue. +Id describe it as a performance fix If you enable support for highlighting in the query handler and upload documents where text is extracted the query performance will drop significantly for hierarchy resolution +Anyway if such thing is recieved we should respond with a HTTP Bad Request +If you put a better message I guess I cant see it unless I upgrade my Seam which wont please my Developer Studio +Merged to the branch in revision . +Evan Weaver invited me to throw some ideas herePlease take a look at to see different explorations +Can you add a wiki explaining the user behaviorI mean when the parameters are set what is stored in the metastoreHow does it impact when the data is already skewed due to list bucketing +Thanks. +Assigning it to Yusaku to take a look +New patch with ALTER IDXPROPERTIES stuff backed out +This is identical patch to the one on with only diff being listeners are now instantiated with constructors which takes conf as parameter +Ya i have tested with and its working. +Please verify + +I have assigned this issue to you Dimitri hoping you are the correct person to work on this +Tom can you please take a look at the patch +This should be included in release +passesI want to keep the version number at since I referenced it in to branch and TRUNKThanks for the review and Andrew +This is a fundamental change in the IPC layer +Moved the issue to M since the outbound adapter are set for M +on patch +Patrick who wrote the plugin would be happy to contribute it if he hasnt already +Small patch to solve above issue +I think it should go as is +It works with Sequence files because thats what I needed and the get had to return something +Adam your ICLA has been received +Ah ok thanks for the input +we dont if we dont test for it. +modulesimageiosrcmainnativejpegencoderlinux +Tests pass with the version +Fixed in Rev +While it will better that we could update the codes a bit for eailer integration +I think callback handlers should be in the space +Thanks for spotting this +We could just have with Backup and Checkpointer +This issue is the reason of failure of the following VTSVM tests on +Yeah I was just looking at the wiki device was never specified for run +The patch also corrects incorrect in example configurations for three other connectors +Thanks Navis +Of course multiple databases are allowed and work but often lead to unexpected problemsAdding a nd db in a normal case may double the memory requirements of derby given the cache per database model currently implementedn background threads per active db +The EMF should be relatively easy to bubble up in the UI once things are in place in the core models +Any pointers are appreciated +Modified persistence XML properties and my previous error is gone +Changes +This issue is the same as . +Functions and even the model itself could be targets as well +You are also using IE to view the map +Im not sure which online help you refer to but the same documentation will be posted once is released +Actually I have most of this done in a proto I did earlier +Reopened the issue as the newly introduced stack via doesnt have the patch changes +The patch looks fine could you comment on the bug you saw with a deadlocked shuffle you saw once since I do see youve reverted back to original patch wrt the fetcher threads and stallingWe have not seen the deadlock at all it appears to be a one off issue +Thank you for posting +Hibernate uses some of its own names and prohibits usage of cps setup but that is not essential for doing the update to fix the leak bug AFAIK +Mostly we just try to emulate the official Ruby on Windows as close as possible. +Added Apache license missing made some headway with getting this working +Closing tickets +patch is created based on latest SVN that has patch in +Already in progress +These patches have only one author and they have not been committed yet +closing issues associated with existing releases +ll include this in the SP release. +Attached a quick deck with a summary of the design and semantics of this mechanism +nope nm was testing incorrect branch is legit failure +There is a label See Tools Visual Editor FAQ on all three screenshots +For what its worth Ive attached a patch against that is working for me +Looks good +Before EMC was introduces this problem was probably more hidden +I will send an email for help +There is a dummy VM that gets created during this in VMWARE and gets removed after this operationWe allow to resize the volume even if provided the current size if the volume +Im still not convinced this is a Struts problem +Manually tested! +temporary move set fix version to +Updated the reviewboard to include the +I have a patch for this +This is only for Branch +From the doc Neither nor is +You will get also not get the fields value +If you have time to wait for that release I could fix the Doxia upgrade issue as well. +Its just a little bit more complicated if we start a threadbq +A Pattern insteadBite how? The enum order is part of the spec +Ill commit this momentarily +and the set to Resolved Fixed again to correct resolution which was lost in Bugzilla JIRA conversion +Whats your configuration in the? Have you looked atAfter short look i can see that there the Blobs are freed correctly +Please verify that it was applied as expected +import a project within a workspace manually repeats the same problemif it works fine manually then something is wrong in our import code and we should fix it +Check out in the project if you want to see what I did. +And internally this should lead to create version When the Quality Profile is updated for the first time after a first use this should always lead to create the tag version and changes can start to be tracked +Thanks for the quick review Knut +I believe this problem showed up during Dhrubas tests on k nodes +Well the lock timeout will only abort the transaction if you rollback the transaction yourself? I mean if youre managing the transaction you can catch the exception do something about it and then let the transaction carry on if needed? Infinispan will only rollback the transaction if youre running with autoCommit on which result if no tx are running for each cache operation to start a brand new tx and commitrollback when the cache operation finishes +But its still not right +Oleg which version of OSX and which version of JBDS did you update from and to? to? or to +I think the javadoc is just wrong +Those that do mailing lists might miss the comment in this ticket as the SummaryDescription do not interest themFor clarity please open a new ticket for your completely different request of closing down mailing lists +Please enable that property by default +Ill try the tests and see if we start getting the intermittent failures again +Hi JacquesI have spent some time to find a way to integrate in the best way the help pages that are being developed on Confluence intoAnd this is what I did findThe best would be to be IMO to periodically export the space in HTML put all this content in an help component zip everything and put back on a Confluence pageUsers could download this zip file and install it in theThis should be all a user should do to install an updated helpThe solution in the patch at the moment is not so versatile but I do not know how to do what I have describedIt would be great to have help in doing it +Attached is a sample set of files that implement a bare minimum of functionality for sending mail +Eg if an app s and refCount becomes but doesnt trip the assert because assertions are off we want to still hit the ACE +Closed as per state in Bugzilla. +We have to accommodate both hadoop and hadoop Certain form of shim would be neededTake a look at srcmainjavaorgapachehadoophbaseutil which I introduced in I will send out a poll on dev list for when old metrics should be removed +Has been resolved in AS still see this issue while running the application using glassfish v +Its then very hard to read logs +I was just trying to be helpful to tell you that deserializing using should provide equivalent results +Is this still an issue using +jar +I will create another JIRA to track this issue here + on the patch +It does create commands but a different one for each +The submitted path do not yet eliminated the shuffle phase barrier +I committed this +Oops +Jake So problem with JNA requirement is solved by your v patch? If main goal about direct IO reads is reached maybe we can consider this as resolved and add new task for write Direct IO improvements? this one is getting heavy +Committed! Thanks Alex +Is it a regressionNo it looks pretty ancient but if theres a including this would only make things better +Format your patch using git diff instead of git +Hi Albert This was my fault +I have provided the source code +In Myfaces a new dependency was introduced +Thanks Kim! Both the patch and the html look good +Updated version to +I dont think it does happen every time others seem to have run the smoke test and not seen it at all steve saw it some runs I see it every run on my really fast comp +Updated in release notes. +Dhruba can you please commit the proper patch to so we can close this issue? In the mean time well move it to since many people are running cdhb +heres the correct one that implements extraction for each of the integrity defined in the dialect +You have not provided info to reproduce the problem and I provided you with XSPwhich works +fwiw contains unit tests that will catch these kinds of problems in the future +this patch move as part of the bundle and remove from to the POM for to add the patch with ontonet content patch remove from the dipendences to +bq +The client uses the CONNECT method and headers to interact with a proxy +Sorry this got lost in the stack +I guess so though you probably want to loose the quotes around the XY in the shell triedJAVAHOMEcProgram FilesJava gantand that worked fine so it definitely is a format of the environment variable problem + of to revision +Thanks Thomas +Please use a packet sniffer like tcpdump or Wireshark if and inspect the traffic directly on the wire if you are still having doubts +Thanks Shengsheng and Alex +The intent was to implement what we discussed in ejb last week +Looking forward your proposalThanks. +We need to monitor both builtin tests passed and pdk tests passed +committed thanks +The stack trace of the is similar to the one we saw in +Implementation imported +applied thanks! +Jeff Jencks anyone can you vote this ones version of the patch applies and builds for me Ill repeat my +All javadoc changes to add missing docs in. +Im working on a patch +Ill try to have the patch submitted this weekend +Patch fixing to properly handle the incomingacceptDocs +this patch provide select java source class path and option wizard panels for javawsdl wizard +This still needs unit tests +Ive rebased the patch on the current revision +Added to CHANGES and put in a comment about calls to firstSearcher initialization +Instead of throwing an exception restart the listing seems to be a better option when the startAfter inode cant be found +I dont see this in Nigels new queue up on the new hudson +Thanks for your help +I took the liberty of composing a tos method out of the spaghetti and leaving engineeringValue and floating as the remaining larger methods. +Thanks again Rick +This is after all your task to complete +Actually you are right that is intended to be extended. +Cool thanks +The main argument for not using it is that its one less arbitrary limit to worry about +Bulk close for +MichaelSorry for the delay there was no change with as well +committed +Nick marking this for you can you take a look at it +Seems to be configuration issue +I suspect it has to do with the comparison returning numbers other than just and +Closing Out of Date issue. +The problem is that it would be incompatible with +The changes spans over many of the surefire projects so the patch files base is the top surefire project +This will be fixed in a release after +Attaching a version of the test case that does not seem to fail +please note that this +fixed +Sorry on holiday last week and just catching up +Path applied +adjusted for Hadoop version Matthiass patch applies clean to +Nope it was my bug +Patch applied +I can confirm that this problem exists in standalone mode also +Thats why bug is seen in +It is required to process directories with spaces in the for loops +Actually the have been stripped from the zip too so the tests refer to Java its not an operational project +I have committed this +There is a connection timeout mechanism built into ssh +Could you please outline how you intend to handle ant builds then? And I think I am not the only one that uses antYou also did not address if you still intend to stick with Java for the other modules +Maybe we can also add the files path info as well as the timestamp when it is moved to trash +you commento A separate credentials database is only necessary for systems which already manage multiple Derby databases +Will be done in a later release +I am specifying a runtime arg that I need when starting that works this arg is also passed on shutdown +If you lose the connection between log collation slave and master the slave should log on local disk until the master comes up +Password just sent +Added a test to to verify bracketed comment support in ijRan derbyAll and and saw no errors I could attribute to this patch +There are unfortunately a lot more +The interface only allowed for the receivable qualifiers on the event it never allowed for the actually received qualifiers. +Updating bug title to reflect the actual problem +same patch again for hudson +IONATIVELIBAVAILABLEDEFAULT rather than IONATIVELIBAVAILABLE +AFAIK there is no way to translate this label +This is completely blocking me from moving forward with my project +Looked over Todds patch +Cleaned up version of patch attached +There was a bag with file +Yup this is done +Ive checked with SDK this issue is reproducible there too. +The current revision could be read from the file during startup stored internally and then accessed for display +patch attached +on code changesI think tests should be true unit tests not runs of pig scripts +The actual solution uses dynamic proxy It was a temporary solution +In essence this URI tries to have two ports +pending ones uploaded +lgtm Flavio Ivan for reviewing it +We agreed to continue to live with the WC license and there is no reason why we shouldnt also live with this one +Hi TomI think the default should indeed include trivial nodemetadata for localhost using the assumption that you are on linux or osx and will login with the current username with default ssh keys from the home dir +bq +I can successfully run and I can use this as a basis point to diagnose problems I am having with Configurable and Hibernate. +Thanks +On that basis it should then go ahead and use the params filter to further refine its choice +Hi Eric +I definitely want to explore the issues with state transfer of concurrently starting replicated caches later but that can be part of other JIRA here my priority is to move forward to be able to continue with performance tests +Could you please let me know if depends on any other files in package apart fromThanksCiu +One way or another both will result in an invalid command +One last time +The two classes seem to overlap quite a bit +I copy hintedhandoffmanager from c to c +New patch now available incoporating Aruns feedback +This sounds more like a problem in to me +and as Fred says some more specific requestsinfo would interesting rather than first class +For example when iterating through the list of windows reserved words the error message should indicate that +spring has beans which should be recognized as a Class for an Application +This seems a lot like what big top has been doing too but perhaps not on a nightly basis +I refactored things a bit to avoid code duplication and got rid of the static methodsr +I knowthat some people probably like the current behavior and it may deem it an errorif the code silently ignores a null we can make this a configuration parameter in the binding file or file +The changes are +Now arent you supposed to be on vacation! Go drink a liter or two of beer! +After further testing I have seen the same problem with the unit tests hanging when using the RB runtime and the older jinterface jar +This is not on Axis +If the user presses when the script is running a confirmation box is thrown and if the user chooses Yes the thread running the script is interrupted +Do you want this to remain open Scott? Or is it done +It seems this only affects Java on Mac rawagner just tried it on Ubuntu and it works for him +I want to create this test as part of the +Second comment is to make for the existing classes such as where we are changing the accessvisibility of methods we should do it in trunk first +This is a duplicate of +Not critical imo +If there is any interest in further development here I would be glad to work on it +Should be working now but Im going to do a clean checkout to be sure +We need to work on fixing this for future builds +The patch is simple and it fixes the test +Can you reproduce using ij to both create and reconnect +Can it be resolved +What happens in this case is that the client is still dropped but there is no exception + +Bumping from +Do you have some specific concern that you want such a test to address? This issue doesnt involve the upgrade logic as far as I can tell +Importing is doing much better however it still sometimes has an issue importing a model +would contain fix for allowing mapreduce job to run on hadoop v allows to pass using is the story with? I see it in tests but it is not being run currently +Like Angelo says this should already workJust to explain a bit further the Deployment Admin specification that we use allows you to add support for new file types to deploy +Hope you all like it more development +I get the highest PID and if it is then not the latest I the store question if I do an update does PID get synced +revisions trunk rev +Closing all resolved tickets from or older +It is not fixed in the CP branch +Added +With the modified program I am able to successfully encode and decode +updated to there any objections on getting this committed? Thoughts +Probably I am a touch late on the comment +Worked fineTried running against a real deployed cluster +Attached a patch that increased the test timeout time disable a table only if it is enabled +Committed to +Thanks for the fix to Simon Kitching +pom after I changed the war property from String to I hope this patch helps +I guess we can close it +Putting bind in there makes it clearer but then its different from what HDFS does +and committedOnce again thanks a million for the patchOne last request I really would like to have Giraph in Bigtop distribution +I am also having this issuer also solved by downgrading to +In order for YARN applications to implement their own properly they need to know who is connecting +One possibility I can imagine is that if the NN writes a txn group to the local disk and fsyncs successfully and then fails before writing to the shared storage we could have this scenarioNot unless we had a failover and failback though right? That doesnt seem to be what Hari was describing here +Can anyone think of any real life use case where this change would break existing code +Will get it to shape and attach it here +could be modified to do this as well +I cant think of a scenario where that makes sense really +Tamas This looks like a bug in my implementation +The patch doesnt remove any deprecated methods from contribanalysis neither does it mark the other Analyzers final +Preparing for release +The +Can you test it please and post back with the exact DB version number that you used +There is still a lot to be done +Model project set for m attaching what i think is another example of this problem +Could you please try this patchThanks very much +Todd +I agree with you nevertheless the functionality is here so its expected to be working +I have to kill the java process to get it to terminate +Here is a new patch that increases the timeouts +Not used any more +Please vote! +Im reluctant to apply this change by eye +I just committed this +Can someone show this JIRA issue some love +if the builds indicate not fixed +I have checked for the cases pointed out as part of the comments to this issue +The investigation has been completed and the results are available on the wiki +Ive committed my changes +Andy you can just simply start daemon on some nodes while start daemon on other nodes +a handle notifications in async way which may lead to handle notifications out of order of the events happenedDo you mean the events generated on one same znodebq +Can somebody help? If an admin can grant me Pootle admin rights I can solve this problem on my own +Hi David +What remains is the Spring JMX stuff +If there are other related issues then another bug report should be opened. +it seems that this way its tied as a webapp library +from +It seems likely that the fix to this issue like the one for will affect the messages shown in the Getting Started guide +I just committed this to trunk +Integration tests pass on Amazon EC +Hoewever please note the following the Maven Dependencies container reflects test compile time scope so it is supposed to have dependencies with scope provided content of the Maven Dependencies container is not directly used to calculate runtime classpath for launching tests there is different mechanism in place that uses Mavens dependency for your fast replydoesnt that just boil down to a bug in resolving provided scope for workspace dependencies? As normal repository dependencies seem to be added to test runtime classpath in provided +The strange thing is that the findbugs analysis takes ages to complete compared to +After some thought I think we should better update to use a Queue internallyI am failing to see a use case for accessing the iterators in the chain +Basically what should derby do when it encounters an interruptI think it should always throw some sort of error back to caller and would be nice if the error or the nested error indicated clearly that it is being thrown because of a thread interrupt +Last one is rather hard to avoid +Tour choices are using the extension in XSLT or waiting for XSLT apologies I missed the point about not reporting an error +Need to investigate this some you ever tested a picketlink SP using as the? This is what I am testing against now and setting up an SP and attempting to access it via reproduces this issue +Creates new master files for these testsRan derbyall on Windows XP with Sun JDK and IBM jfoundation VM +svn ci serverSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientAdding serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbserverejbdAdding serversrctestjavaorgapacheopenejbTransmitting file dataCommitted revision . +Can we see a sample of the code xsl and xml files you are trying to run? Did you try running the transformation from the command line +Actually the fix has been done in sandboxakarasulurc. +Even more debugging +I dont remember if we need to do anything more than just commit the changes or will they be picked up automatically? Thanks again please close if we are done +I am completely in favor of using cglib with ASM +Sorry I forgot to mention that the patch is implemented in terms of the web dao objects so the web services have access to the same information as the web page +Could you regenerate it please? Thanks +This solution avoid to Parse any content would could caused any errors in the parsingAny thoughts +If that happens both the admin commands might see that the path doesnt exist and try to create it +socketmultihomehomesis equivalent to socketmultihomehomes +Can Dhruba or Todd can you please review this patch +At the end the problem wasnt the queue itself but this anti pattern was being used by accident as the failover was starting the connection and right after failover and closing it the failover side maybe we should just not resume invocations on close +It does seem like we should be able to detect it +No need to the patch +Recent work in trunk with Proxy Connector concept makes this easier to setup. +The advantage with this is that when you change a format significantly you just have to write a reader for that without modifying any earlier versions +Move to close +Ive attached a patch that should apply to both trunk and the branch +Thank you +Removed test from exclude list with commit r. +Thanks for spotting that Knut +After some investigations decided to not support doHeaders method at all because code becomes too cumbersome +I propose the following elementsslingDocument Root entry slingContainerslingComponentThis is very specific for my use case though I think a root entry node is necessary +Whats your recommendation on where it should go in the docs? +Hmm +Still seems broken in +requires that we validate bean names as unique at startup +Looks like Igor already fixed this one +bq +That reduces the list of things that can fall out of sync +Note that there are two branches and +we get this in as well +It is ok to require for my own code but what about existing librariesfile written by other people that I use? With this requirement it will no longer be safe to reuse code from other people as it is impossible to ensure that the code then adheres to the rule +Ill keep trying but it seems that this exception was just a result of the fixed onesThank you occur of this exception again +Are you referring to what the web UI and agent do currently and that upon resolution of this ticket we will not be doing so +Are your Type fonts accompanied with AFM font files? But it should even work if theres only a PFM +Were using qmail at the moment +Deadlock related issue +I just committed this +Good morning LucasI made some research on my application here and I noticed that the error occurs in another scenario +I was not able to build this on my Linux environment unfortunately but the code in that demo looks good to me and should be committed in a demo folder under the CPP tree +When would this be soClarify may poll and synchronize +a Visual Basic project were seeing it suggest Company as the groupId note the space in the company name needs to be removed +The issue occurs when using forkModepertest aka forkModealways +Its pretty tricky but I think I have the right way of approaching the model now +branch version of this patch +Delete from javatestingorgapachederbyTestingfunctionTeststestslang +To major compact a stripe all L files if any can be compacted and split into stripes then mergecompact all files belonging to the stripe +You may need to build against a clean local repo +It contains a quick outline for some of the accounting documentation to verify the formatting and structure of the Docbook commands +In addition we could make the following enhencements Check if the locking could be simplified parallel job processing it should be possible to process different job types in parallel while jobs of the same type will be queued jobs should be marked to be only runnable on the instance they were created addition a job could signal a retry could remove the need for locking by using the id as the node name and not allowing same name siblings +is also related. +As long as the is scheduled after the scrub +Ive just committed this to trunk andThanks a lot for the review Eli. +We are having a similar discussion for FLUME project on and will likely use the same solution for both of these issuesCan you please join the thread on and share your thoughtsThanks is readwrite and the first issue should be thanks for the prompt +Hi Carloswe would like to replace the graveyard background with retired that sounds less creepyPI tried to modify it with Inkscape but looks like Im too nb and didnt accomplish can you help pleaseMany thanks in advance and happy new year +Reported issues were fixed in Committed revision Thanks for the update +Yes the IN filter has special handling in because IN has a variable number of arguments and we need to prepare the arguments to add them to the appropriateI will investigate how to do code sharing for the string hash function via a Hadoop JIRA + +Part of the code taken from Clerezza +As such its that edition which is on the site not the EAP edition in which I made these work around that for the moment I have attached a PDF version of the book +PR adds ability to and remove multiple transformers from the list at once +However Im against copies of the code Id rather have it move back to examples than copy it. +Heres a better patch with a test included +Ive updated the timeout and time to live for message send and receive to respect transaction timeouts. +The issue happened on linux when ejb was chosen with profile named different than default +if it gets more popular +qquiet Do not print out much when executing +same error for test and test +well not sure to understand your pointusing filtering on icsinternal I was expecting only folders to browseand using filtering on internal I was expecting all the other folder with no merge as it seems to be +This will also require to update commonssolrextras modules +So this jira should rather be filed there +So I am not sure if it is good for +On v Ive tried locally it workedYeah +Thanks Harsh +Tested end to end and works like a charm +thanks guys! that worked for me though +Heres an updated patch which addresses your feedback +Yes it was up since early until very recently +I dont understand what is incorrect about this +Will provide an addendum for sooner +Chris thanks for taking up on this +It would not be an api change. +Would it help if I submit a patch? +We are planning to add this functionality in a couple weeks +There are stil some tools issues but marshalling and jaxrpc are fine +I was proposing that we changefile scheme to point to checksum scheme to point to +No activity on this for a long timeIf anyone feels the itch please help drive this issue to completionIt is not quite clear to me what is actually holding it back but there seems to be uncertainty on how it resolve the issue +The remaining ones are releaseaudit findbugs and +Set the relation to issue leads to hang the reliability test in debug mode +I have reviewed the code and ran a bunch of tests +Deleted until we know who was playing with this jpox renamed to and to jpox +Besides there are applications sharing the same FS object and hence it would break them +Fixed in rev snapshot deployed please give it a spin +Review of +Thanks. +I am not green but gave up due to vacation +Thanks for the report! +Should have been created as a Clustering component improvement +cgen JPA template is checked in to srcmainresourcescayenne +Offline event does work but its firing too many events instead of not firing this event +shipped in +I can immitate the tests failure when the property is switch from default lower to mixed but now I dont have any idea why overriding the property value directly in the test didnt fix that +Your suggestion of keeping all logic related to managing replicas in makes a lot of sense +I think when we complete everything we can together decide what to do +Not enough to kill RC IMO since its been this way so long but should make next RC if one and backport to Id say +As Ive written it looks like doesnt solve the problem +Mapping Diagram has it own Open SourceMapping changes of Hibernate Configuration View have no effect on Diagram action should be changed too +Its definitely a different bug from this one happens even with a single node but only in invalidation mode. +Added template +Mark without the patch in the very clear repos beans unit test still got vm crash +It is in fact the size of just this buffer +I am curious on the performance of the split how long the region can be splitted my understanding the lob family should participate the region split and the region is unavailable during this period +Cancelling while feedback is addressed. +I guess Id just have the users download it +I agree the problem should be solved in a different tool +BTW the reason this wouldnt apply is likely because of +A new module could certainly be done +Cool I guess fixed this and we can thank Anders for it +So I had to make some changes in local mode job submission codes to directly invoke Hadoop instead of going through Hadoop scripts + is not updated with hibernate bits which are in FP is in rebuilding the whole EAP with upgraded hibernate bits to obtain +Completed in commit +Thanks for the update John +That can lead to real trouble +Hope you understand +lib entries are not duplicated into tuscanyboot and tuscanyextension directoriesBTW does the CL hiereacy follow the order web CL Tuscany boot CL Extension? If yes should the libs in boot not dupliacted in extensionTa feature is in for M. +You are welcome to contribute a patch for though How can I get the necessary permissions to commit this change? One needs to become a committer in order to get permissions to commit code to the ASF repository +I bet you are using an older version of +etc should be submitted as a Hama jobOtherwise I think its too incoherent +Best regards George +Bulk defer all frontend issues to +Going to try using instead of to get rid of all the extra dependency stuff +slightly improved good minor nits or should be and if Snappy is missing should be and snappy is not installed on your host Most of the time these options should not be necessary should be something like These options are not required if the snappy libraries and headers are installed in the system lib and includes directories +Im ready to commit this as is +somehow it fixed mistake it is still occurring +srcjavaorgapachecommonscollectionsmap +This was actually a general problem with inheritance of static methods in generic classes. +What you do here is to improve the standard with enhancements +missed out on a compilation problem +Canceling until Chris comments are addressed +I can confirm this bug is still present in Mapped two entities parent having a to child unidirectionalWhen executing a Criteria query both the parent and child are flushed to database but not the join table insertWhen manually flushing before the Criteria query is executed I can see that the record in the join table gets inserted so still a Hibernate flushing issue. +Closing resolved issues for already released versions +Its not by choice believe me +John have you had a chance to look at this +Sorry somehow failed to get these latter notes +test seemed to work fine thanks to Joe and Roy. +Sorry I miss typed the title of this issue +I can file any that I know tomorrow +Marking as patch available +The handling of this is also not right when transitioning from Running to Failed state +But it should be applied to Seam as in advance. +Changed the requires to use the enum intorduced in JSF +This model object makes writing a testcase much easier than testing the logic by hand in the running portal +Will address later in a general review of how we now let out +If I understand itcorrectly the rules are +Should be put back as Fix Version +on patch +Hi Pavel I was able to reproduce the problem above with my sstable file +I just forgot to commit to trunk branch +EvertYes that is the idea +Ill run an experiment to verify that SYSCSUPDATESTATISTICS work as we expect stay tuned +I was not able to replicate the error but Ive attached a patch that updates table header titlenbsp in the jsp pages to title +Is there anything else in the log fileOh and it is usually better if the first time you have a problem you bring it up on the users list rather than a JIRA you tend to get faster responses there +The is created and maintained when the database is reverse engineered + Vladimir the patch was applied to BT at rPlease check that the patch was applied as you expected +I committed this +imo this should not be treated specially +Notice that decodeIdentifier may return null if the identifiers class is unavailable or throw an exception if the format changes +I disagree on this and I actually made it that way willingly +Is it a safe assumption that this only ever occurs when an infinispan cache is involved in a transaction +Attached a patch +Any thoughts +Alexey please verify +Should be modified again before release +Whether a worker is sending out partitions as it builds them doesnt help us because that same worker is also receiving partitions of its own from other workers that read a particular! So this is the real memory danger weve seenSo +So it may not be easy to get rid of it +I think of it more like a graceful recovery from crashbughang +My first potential patch +Applied to TRUNK and branch. +I wont need your patch though I hope you dont mind +Not sure if I was clear but correct the patch to not add the +this problem will not be fixed in the development stream. +ear file can be directly deployed in the serverShould the also deployed to the server directly? Shouldnt this be part of one of the other deployment unitsNot sure if you noticed but the has two top level folders with different capitalisation +Note that there is no change with respect to the order of callback processing and if the Namenode processes them correctly today then I dont see why it wouldnt with the proposed solution +This looks like fixed to me on and can close this bug if someone confirms +I will try to find a way to reliably reproduce the problem and reopen if I that last comment +If you can submit a patch for our tests I would appreciate itHopefully you can add this to your regression suiteI will through +The issue in the discussion seems following the same lineAlso I think hive is still in its early days and it doesnt even have release yet +Roshan you are right +Removed monohtml references from subversion revision +Beta will include the jBPM identity component and add it to the BPEL service archive for compatibility with jPDL processes +Was working with a tricky schema at the time with xsany in it +This complicates the code but left the bulk of the logic in tact +Changed call to so that it uses and upgraded to JSTL released in. +Some content bundles are not added to the applicationBundles list when the server restart after a hard stop because it still presents in the bundle cache +The upgrade merged when egress firewall rules added toI am talking to Murali I will update the bug +Im just trying to present some wild ideas to spur discussion in different directionsbq +Attached is a simple patch against the that corrects the classloader situation +Even if you have issued a shutdown f or shutdown yes and the console commands have been removed this will still allow you to exitThanksScott for the update Scott +I suppose you also want to complie your sources +older issues for Apache since were no longer actively working on these at the moment. +I have committed this +Please check that it was applied as expected +See details in attachment +If is this the way i could take care of this +Bulk close of old resolved issues. +This may or may not be fixed in CR as well +Hi Gary thanks for the patch +I briefly tested building some stuff against the artifacts you have deployed on staging area and everything worked as expected +What happens if you run Checkstyle on it manually? Im trying to establish if it is a Checkstyle bug or a Maven Checkstyle Plugin bug +This generated file will contain all files inside the test directories +The screen renderers are convenient but using them requires a lot of kludge code +Theres no way to get the private part of a key pair afterwards +Thanks +It would change a lot of the internal architecture +Updated patch in reviewboard as well +you could add a feature request for your suggestion of config element for handler +I will submit a patch for this and file a jira for making the Snappy codec exception better +For me the total running time is approximately seconds +i have the same question +Added quickstart to demonstrate the problem +Closing resolved issues. +Patch for version +Please verify. +I wouldnt be a problem for me to add the Derby in to my Maven build +The weird special field for specifying a value is due to how the bindings work for example +If there are no severe structural changes then for now Ill say a new servertype is not necessary +Removed the expected exception for test test because it doesnt happenOnce I fixed those things then on the last test I get an error when it tries to redirect +People doing the headerlicense verificationmodification can just attach modified tarballs to this issue if necessary +Changes look fineGeorge +Make sure it passes locally for youNo worries on findbugs +Is it a good thing to build an type that silently discards most or all input as opposed to complaining or delivering something +new thanks! +Its not a decision come to overnight +I use a sequential instead of a random so no block is added +Why offer the opportunity to set SOLINGER if we disable it prior to close? It only has an effect when closing the socket anyways +I added the plugin to the form tags example application but could not reproduce the issue reported by GregDo you have the modified webapp that you used anywhere I can downloadCheers Greg +Since these changes are related to of duplicate code I had planned to do them as a different patch +The driver will also need to be for raised and resolved. +Some are defined inline in their associated class some are defined outside of it for no rhyme or reason +So marchpentium is a minimum +In the case of heterogeneous list every thrift type must be a child of some int float and bool +was just an ideaId say lets dropOn Deletes all is good to me except seems like a delete at now will have no effect right? You have to know the timestamp to delete +great +Horizontal scrollbar has some minimal width +Fixed the problem with two superfluous title pages in com guides +Thanks sadeep for the input +Thanks for the report +Itd be awesome if you can take up which would help us write tests for JMX stuff outside HDFS +This way if is down we can still authenticate against IT successfully and read in the information from the table. +Christian is there more work here? We want to close down on Camel to cut a release +Merged to doc branch at revision +If you want to investigate this for the next release otherwise close it +Committed to branch and trunk +The only way I see to get this into the log is to programmatically inspect the garbage collectors and log the info ourselves +good job +It adds a new configuration item in gossipperiod +Patch applied at +I think the same should be true for default input as first part sounds reasonable +MINA and sshd have been removedWe keep MINA and MINA available on the dist server +And then providing a scheduler to update the stats on an ongoing basis +Pushing off to release +NET or Mono framework then it seems you could run it with a different appconfig file to force it to run on a later frameworkI think it likely that the target will fail with the same obsolete messages that the mono target encountered +And found has similar problem as well +afaik what does is it filters the overall subscribers based on if their corresponding predicate returns true or not when event is postedif a predicate of subscriber returns true then the methodsubscriber gets invoked else not so in above case with if client A has sent response then predicate returns true only if A is receiver and hence only A gets responsei cant reproduce your exception +Tag above is corrected +For instance I prefer the button in the menu bar in Wais proposition but not the font used for the text in normal buttonsAlso an advantage of Wais proposition is its very condensed +looks great +It looks like the problem here is fixed +Attached test is right here but we must assume that even if we delete some element in the the cursor we have built on top of it should still be able to return the correct next element as it stores the current positionI think the problem has more to do with the way we manipulate the cursors internal pointers than with the way JDBM deal with modificationsI will investigate this aspect anyway +On the other hand if we rollover the file every day then job start and end events for longer jobs or the jobs that start on the day end will be in different log files +We checked the correct implementation and tested again but the problem still exists +Attaching patch that applies after is committedPlease review and provide your comments +Daniel might you want to increase your to ensure youre not verifying before the invocation has happened? +For the default timeout I think we can cut down the connect timeout +jersey has a built in support for a GZIP content encoding filter that just needs to be configured on +But we need to revisit this in the future +Or am I missing something +Committed revision Thank you for your contribution PatrickJarcec +Unless were moving into a approach Im not super excited about this +I cant find where this was fixed +a few thoughts Ways and whether to effectively cascade down a on the first fail +What happens here is that you are confusing socket timeout with connection timeout +This patch makes the version for the generated variable +This patch makes the changes in both and +I just committed this to and upI ran on on windows +Sorry +The issue still exists in +I have standardised all references to Apache to Apache HTTP Server as this is the Apache web servers official title on the Apache projects web page +Lets just take the sample out of in a number of old issues that I raised but which I dont think are still relevant +bq +It will be really good to integrate this with the current RAID +QJM is writing Tx +Applied! Thanks Karan! svn ci containerSending containerSending containersrcmainjavaorgapacheopenejbutilReplacing containersrcmainresourcesAdding containersrcmainresourcesTransmitting file dataCommitted revision . +You can get it from the HRS web UI or by using jstack on the machineThanks +James My question is Is it relevant to a large noof people to have this last patch? Can you achieve your objective by just adding an implementation? I mean will you need to patch DIH to meet your needs +Accidentally created this with Major priority +I was mistaken +Should submit +Of course we dont want to do thatWe would kill the permgen in secondsPlease no +and others like youThanks +Patch applied at r please verify +Oops Henry youre right I did commit the fix to this defect as part of my patch +fixed in trunk +Besides horizontal scrolling should take into account that several first columns can be fixed so that used can scroll the data but relevant column and row titles remains visible all the time. +Can you svn up resolve the conflicts and the patch +If anything they are just a hassle which adds useless time to test runs and needs to get updated everytime we change the plan +Decrease priority +Okie dokie thanks Roshan for the clarifications +Well not really as it happens since that caches the value elsewhere +In v patch use +The second part is how to access the same information through and which is the way we should prefer +Committed +then why are you opening it again? you arent using the version it was fixed in +Second caches are much harder to implement and test than the cache information in HTTP +It s far more important that there are tests +Again Im using as my testcase +Patch with more commentsAlso added a return if reflection was as there is no need to try more +HS is working better than ever before +Bulk move of M to to M because of no updates from +Second version of patch taking care of specific case when the separator is the first character after the string separator +When I did it I revd my copy to +Attaching patch along with updated test caseAfter digging through the code a bit more it would seem that the running list of found elements for encryptionsignature should likely be inspected at any time that additional elements are added to the list to be signed or encrypted +I create a reader bean in spring context and its still not used by my cxfrs endpoint +Isnt it generally easier to instead just check procPIDlimits thoughDoes this also work in context of a secure DN startup? Does the logged ulimit reflect the actual JVMs instead of the wrappers +If needs to have a flexible layout this probably needs to be addressed as a separate issue +Rule A uses Query A and Rule B uses Query B +So I just caught that case where domain isnt found and made it ok for that to be the case when attempting to stop +No test is necessary as athis is a trivial fix +Wed have to merge that logic into the serverport as well if we merged this +And I dont want to let them bubble up to the Servlet Container which would most likely write them to the we could allow for this scenario too +Do you mind working on a refined patch +Test project should also be adjusted to selected server isnt itwhat else +passes +Just raising the priority to appear high in the list I know Brian is working on this +Resolving as Incomplete without more specific description of whatwhere needs to be done +Committed branch and trunk. +When a container is killed before launching the TERMINATED exit code seems fine but the message can be something like Container terminated before launch +My main goal is to make everybody aware of Lucenes changes and to have more documentation on this +The question becomes can Groovy hide the issue or failing that we need to document this as a known issue +We cant do a thing otherwiseAnd next time please mention this BEFORE we create your Jira space! people have started created issues on your new Jira instance and they will get blown away by the import so make your people are aware of this +Final patch incl +Youd read what codec to use before you started pulling in blocks +Version uses String form for the value of NUMINITIALREGS so that the display on Web UI is meaningful +Im not sure what the impact of this is but it may not be acceptableThe alternative is to move it from the avro project to the ipc project +Im not sure what the fix version is though +Replaced with deploy +running through hudson again +afaik this normally works fine on oraclecan you provide a failing test case +This is because of the difficult in extracting the final result when running multiple iterations in combination with extracting will look at clearing this up +Heres a patch that fixes this by throwing +else +verified on Developer Studio +See +Also dont get confused when the bug is fixed +The commit also included the patches for +Thanks +My estimate is that its around a runtime hit but I wanted to put the patch together to test that out +resolved in trunk and on JBDS Alpha respin a +I hope the recent commits and the above comment clarified the design goals +I will look at putting it on trunk +Works fine in still working to get it going with +Does anyone objectI imagine it should be OK to test from the browser as long as you can examine the raw text response before it has been but I am not an expert in programming +QA did not see any problems with the to hear but I have problems with this workaroundMainly with the webdav library do you use webdavlibThe VFS test cases do not pass +Only the response is guaranteed to have an order there is no such restriction for the input which can be a Set or a Queue just as you care about the order then you pass a list and get the list +I have no power of that department and dont know what restriction they put in place but I dont think its the container since that is the Sun supported Glassfish +Talked with Prasad today this should not be supported +I can offer to help more on this matter but I guess for now committing the code and renaming the file so that it doesnt end with spec is a good workaround +This bug has been marked as a duplicate of +Interesting and thanks for all the research +Slip out of the hudson job will test latest glassfish builds only whatever Weld version is integrated gets tested as part of that +Attaching a patch to make your life easier +I also have test systems on Solr and Solr using a smaller subset of production data +Patch containing code changes to Position Commonin and associated test classes +By the way you can work around it by making the update call yourself. +The problem with the API is that it subclasses to pick a counting algorithm +Committed +Bulk closing stale resolved issues +Xiaobo can you please add details on what tests you ran +Thanks again for sharing your solution. +Report with affected tests results + +this patch is against +Were coming up on and it doesnt seem like there ever was a consensus here wrt immutabilityIm also still in favor of removing freeze +Source files list and test files list are now cached +The following will be added to the various jUDDI create database scriptswsdltypesuuidxmlnamespaceuuidxmllocalNameuuidwsdlportuuidprotocolsoapuuidprotocolhttpuuidwsdlcategorizationprotocoluuidwsdlcategorizationtransportuuidwsdladdressuuid scripts have been updated to include the additional outlined int his issue +Create a fix to pad with leading zeros when necessary +Reran tests +issue with optimizedDispatch resolved in +Lets close this issue for the release as Wont Fix. +Ive just committed this to trunk andThanks a lot for the contribution Sandy and thanks to Harsh for the review. +However this bug itself has been fixed by the original commit +Created issue +Patch is almost ready but not quite +Touches the following filesM srcworkingwithderbyM srcadminguideM srcadminguideM srcadminguideM srcadminguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcrefM srcrefM srcrefM srcgetstartIve built the these changes into the html versions of the Working With Derby Guide Admin Guide Developers Guide Reference Guide and Getting Started Guide +I added the factory bean created by Ben to the samples and it worked without issue so Im going to go ahead and close this issue +Will this go into the next release? Should this ticket be updated accordingly? +Im guessing since hive has no notion of a unsigned type we should just use a bigint for unsigned types +Unfortunately Oracle is about the only driver that actually depends on that system +The script needs to confine its self to binsh features +The UI allows for recursive deletes now and you will get a list of the layer that will be removed along with the store +Nodes that arent fully in joining mode just wont show in gossip until they are but you can always check the node itself. +ADD less error prone BTW Nicolas if ever you go the nd way have a look at applicationsaccountingservicedef +Do you mind helping with that +If the task was completed by the first AM and recovered by the second AM the name of the task attempt in the jhist file will look like attemptm but the Recovery Service is trying to recover a task with attempt id attemptm which appears to be the format for attempts that completed successfully with the second AMI need to understand a little bit better how these names are determined and where they are set so I can determine how to fix the issue +patch applied migration docs for this change are really poorIm wasting time and nerves trying to migrate some code +was removed with Lucene +I changed the patch now it seems compilable +The three files being merged srcjavaorgapachehadoopio srcjavaorgapachehadoopio srctestcoreorgapachehadoopioturn out to lack the Apache license header picked up by release audit on a different patchThe next submission is a patch to correct this and insure that the findbugs and javadocs tools are also run over these three files +Name size date would be good for starters +The good news is that code seems to work +fixes this issue +Both the short reproducer test and the corresponding functional one work as expected +Implemented the initial version of an within the new module +My centoss version isremove the setting and install successfullyThank you very much for your help Yusaku Sako +That should be Either class level use should be made legal for the annotation or the parser should actually parse the columnNames attribute at the method I wasnt aware of the which resolves this +Thanks +in +Ramya can you update the patchEnis regarding your comments please file a new Jira +Finally I changed the annotations in the byteman tests as suggested above and added the scripts directory on the scripts are found correctly with these changes both on Linux and Windows +Had uploaded wrong version of patch this is correct +This has been fixed in trunk and a test case added +As we really only need the user name in the repo we could do it with a session contextual component but Seam integration woudl be nice for future +This is being changed by the new Failover you explain why this design was chosen? Unfortunately it requires a complete application layer outage in order to stop our s of producers and consumers trying to reconnect to the servers while they are being restarted +But this now exists and cant be removed so we have to be consistent in them allWe currently have and the new supporting this switch so the old one must also support it +quick fix +Its tested against and I have the patch in a review request for jzb +Not nice but I dont see how to get a good default at the moment. +Forgot to mention that the dialog should accept its content as the body of the component or as a block +Just placed it in service because thougth of it as locking service for the whole server +It is true that the Create Image button should not appear for a running RHEVM instance +Resolving for the time being +related! Thats problem in AJAX JS sources +Hopefully the solution I provided is reasonable? +The removal of fakemetadata is under way +I have applied the patch +If the image checksum at seondary is the same as the one at primary Namenode the secondary neither download the image from the primary nor load the image from disk +Bringing into to take a look +Patch added +You can test by running manually +probably can only do what you said in the first place and use the doubleName attribute if nameid are missing +This change is causing the generated Manifest for wtp projects to use the eclipse project I am not a wtp user I dont know enough about why the jar is named after the eclipse project +I hate to say it but I did not know abou these release notes and they are actually exactly what I was looking for! Given when the user downloads the odfdom ZIP file all they get is the I think it would be very useful to have the README file recommend and point to the release notesCheers back to Svante to commit the README edits once reviewed +Applied Vincents fix. +Roo doesnt support wildcard imports or constants used in what about the disappearing columnNames +Interestingly I get this error when I run the tests also +Source code changed new jar and instead of copying it in its a delete move rename +If the parameter interceptor took a list of exclusions you could achieve the same goal +Groovy doesnt seem to correctly cast into a String when a is assigned to a java beans String property +Works fine for me in Win XP IE and you retest and see if its still a problem +I dont se any issue with adding that +This is a bigger backward compatibility problem but obviously its a right thing to do +If the tests comeback cleanI think patch v passed QA runThanks for your patience Jon +Out of Date +master properties as well +This is a feature request and not a bug +Nils we should leave this until the major updates inAs it comes out the process to add this feature is very complex and requires major refactorings within the configuration and interceptor stack functionality +If thats the case you can already do that by specifying the classifier +good candidate for to get the regression test failures resolved +It needs to be configurable and tests are missing too but I will create a patch from it today +This renames to +Here is a patch adding the usage documentation in mapredtutorial for libjars files and archivesAnd there is a link from tutorial to commandsmanual +rename it to something like get +You need to try with the latest release is old and many issues have been fixed +BULK EDIT Moving the unassigned features and improvements out of +The attached patch updates the rewriteParent method of to use the dependency version supplied by the user +Sent at PM you got that I think we are probably good to apply +Wouldnt it? If so the annotation should take a list of Controller classes for which it does generic general Id like to avoid to setup two one for Views and one for +It means that it is not necessary to add to servletconsumerdeploy to have WSRP WSS working +Shutting down the RS fixes the issue +If we back out the change to add sysinfo to we should take the other approach to fixing suggested by Dan updating the example doc to include in the client classpath not +you make a request using suggest and get an error about it looking for +Thus the font library shipped with your OS seems to have a bug or the font data is corrupted and the library cannot cope with it +committed to trunk +Thanks again +This is the EAR application that contains the Portlet used in the test +I changed the maxAge field to Long +Ran all rpc ipc and token tests in hdfsmryarn +Attached reproducer +I ported from my changes to which I run on dev clusterSince hasnt been committed I didnt include readwrite count +Ive also was getting this is undefined javascript error +Using remote reply queues for remote delegates will simplify message memory planning for a service simplify the documentation and the CDE interface to the deployment descriptor +The issue about selection of an appropriate seed has been raised elsewhere +Min will you address this in this bug forAlternatively I could close this bug and open a new issue for the removal of the flag and creation of a flag to replace it on each s object store instance as John suggestedTom +This seems to be broken in CI +check might work +preparing IP Clearance file for Sling Replication +Thanks +I can work on getting some unit tests done this weekend +So this has NEVER worked to my knowledge +There is support for commenting albeit not exactly what you are requestinglines start with are line commentsyou can also use and as long as those symbols are first on the line eg create a keyspcecreate keyspace foobarnow lets see ituse foobardescribe foobar +Seems like we might want to revisit the process +Mehh +Matei wanted to understand this a bit moreAnother use case that I heard of is the following If I have a CPU bound job I may want to restrict the number of tasks that run on a node so I can use the cores all for myself +Yes there are a couple of minor bugs that people are waiting for fixes for +Hey its been SEVEN months and a half since it has been issued! Is still under development or at least still maintainedWe are now backing to in order to use an official working version because we are on a hurry and dont have time to learn to apply the for your help if you read this +With clients into a cluster sharing the connection is almost faster +Im a bit embarassed to say that when I personally ran into this problem it seemed more pressing and there was a fairly easy solution by using a delegating login module that knows what classloader is required for the login moduletrunk rev branches rev branches rev you give a quick rundown on how to use this? I couldnt figure it out just looking at the diffs +Hmm I think failure is ok if the target already exists even in a scripting environment +Thats the reason why all the replicas were placed in the default rack +Instead of just including crypto blindly when ssl is available we should check for it as well and either send an error message or set a flag and remove clients that require ssl and crypto +defer to +Let me add that rebuild retest and attach the new +Added features file for to remove the filtering in the in the etl example as it causes mvn eclipseeclipse to break +is this possibleAnyway thats why we developed the value +The only way to fix this is wait jdk and NIO +This approach will block tinputUpload at Myfacess +Ill apply this patch and close it as fixed in the understanding that should check for context classloader first and then to class classloader to be friendly with +Someone can explain how this should work correctly +I fixed the issue in revision but I didnt use Simons patch as It was just a workaround + as out of date because of month no activity +Old serialization version UID added. + messages +Similarly we can unconditionally log blocks in the future if benchmarking doesnt show significant impactReasonable +Case sensitivity is a known issue for Windows clients that causes issues so I assume you had two issues The wrong case +Pls regenerate the patch +Could any ASF user please copy the code related to this issure from trunk to branch? +httpshudsonviewviewjob +Addressed all items in Todds last comment +The patch contains the fixes to bugs discovered while enhancing the test +We have to specify a particular target for bamboo specified clean distribution and that failed +I restarted the whole osgi framework used a Firefox with all cookies cleared same results +Thanks for your descriptionDo you need the rest style navigation for store this page as a bookmark +on patchThe added call is already in trunk +See We ran into this need too and it turned out that via the use of D options this is already possible +Test both the service metric config and the service metric config templates. +Committed to trunk. +fixed revision in fact the wip permission hook is the culprit +Please use the latest build +I am updating the class to remove the fileID counter only when a commit is recorded successfully +actually this is caused by the Embeddable class for exampel we have Embeddable which has no Id and no parent class has Id then it belongs to its own this is work in progress +Bulk closing stale resolved issues +Youre welcomeThere is already such a request in JIRA +Hmm yes +Bulk move to issue related to templates edition deep changes +The command line I used to generate the stubs wasjava l c or d adb wv s S unrolled u uri is due to having all attributes and no elements in the complex type +the patch for trunk also contains a fix that I had to put in for a race condition in where not all the regions where assigned when the test started +In our organization operations are now moving all users from one domain to another and this will go on for about a year +Jongwook Im glad this allows you to continue your work +Would appreciate confirmation that it fixed problem for you +Thanks Ravi and Alejandro! +i basically have one linear path of commands for each chain in which each command invariably executes +Thanks Juhani! +Ok thanks! Sorry I cant help with qpsmtpd would be cool to get it in. +Sorry for the empty files this is due to the way IDEA exports Subversion patchesHere is a new one is now exhaustive +Cannot reproduce on release a popup displays Authentication failed and the login screen can be used again +So just tried connecting repeatedly using binclient while the server is getting hit with MPS and everything worked fine no client hangs +Documentation patch attachedIt contains documentation for improvements I provided for Callout mediator through and +Lets open a separate issue for core tests +I think this is working now +of reimplementing from a blank slate I propose we split each Maven module into its own TLP +Work committed and reviewed by Martin +The carbon and bit cocoa versions have problems on Java with SWT +bq +Allow user to issue groovy commands to invoke methods of the beans +The changes have been applied to the and branches +There should be no old deprecated code in the is still in the code base and has been deprecated since should it have been removed for? What other deprecated pieces should have been dropped in but werent +It works with JSF RI it should work with myfaces too but I havent tested it +The problem I see with is that it has really become an internal interface +this isnt actually necessary though since if its taken out of the list of mutated index columns the obsolete columns will only be applied to the main data row and including obsolete columns there is harmlessVery truebq +This patch removes disabling code which prevents us from declaring and invoking functionsprocedures with LOB argumentsI havent run regression tests yet and Im waiting for community feedback on why we dont allow these arguments todayTouches the following filesM javaengineorgapachederbyimplsqlcompileRemoves the code which prevents us from declaring functionsprocedures with long argument typesM javaengineorgapachederbyimplsqlcompileRemoves the code which prevents us from invoking functionsprocedures with long argument typesM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangAdds some simple tests to verify that you can now declare and invoke functions with LOB arguments +Many thanks to Kim for the doc changes and to Myrna for the commits +Probably not final word on underlying issue though +fixed in rev +Not blocking GA examples that needs multiple projects and today is split into multiple examples should be updated when possibleneeded +please open a new JIRA if more work is needed. +Yes my classpath contains here only appears on Linux but not on Windows does not appear immediately but after time +just to change security level +This part is about using SPNEGO in the context of EPP so it wouldnt apply to EAP as the original request has been actionedOther considerations can be dealt with elsewhere. +Ill give a try to manually define a Number class in a model and see if this fixe the problem +So will convert meanshift as well before committing +Martin Change done please help verify thanksRgds now compiles without warnings +Another hint dont use cygwin on windows for this it only works with ve got the same problem on Linux SVNI could fix it by specifiying the new dependencies for the plugin as in the patchCould please someone release a new version with the fix as probably a lot of people are affected by this problem +Bulk move issues to and +This looks great ShayWhat was the use case for subclassing to translate the filter into OBS? Is it a custom filter cache? Makes me nervous because the app really should create reuse this OBS filter usuallyOn the Collector we try to keep our Querys +Havent thought it through entirely but I think this could be fixed something like this add a field last to and rather than using validateIntegrity to set a shortened block length in set the last flag in use the following logic If all of the recovery candidates have corrupt last chunks truncate all of them like were doing now If only some of the recovery candidates have corrupt last chunks dont allow them to participate in issue is somewhat similar to +I have the same problem +fixed with the revision check the existing class names before generating the exception broke build +Marking invalid as per Roberts last commentRobert could you file this with the Felix project +copy of the latest version on the review +The attached patch cleans up offending commented out sections in module and project templates +Patch applied in revision +im curious why you only list the community ids shouldnt this also work for EAP based onesand btw +How about we go ahead and move from mr to hdfs and address additional build changes in another jira +Bulk closing stale resolved issues +Closing tickets +It is very unlikely to be misconfigured +is easy to have a circular reference in mediumbig project +The specs do not define a specific order for executing an event and render parameters atleast if I interpret the API correctly Thanks for taking the time reading this issue Greets Im closing this issue from Springs side then +I dont see that as being this urgent +Another try with new configuration and CSV files +Ralf it looks like we can only add one of the tests as one of the tests makes use of Java generics right +very rare. +I verified on a XS setup +Apologies I meant the approach wont accommodate changes such as multiple steps working on the same file. +May somebody pls try this +Im assuming this is fixed in EMF +Once again thanks for updating it +I suggest adding a new context to the three existing ones and letting the client decide which one to use +theoretically having a variable map and doing substitution before handing off to the parser should work right +change renamed to for consistency +Also this is the message it gives to the UI Property dob must be a valid word on this issueIn grails I could specify a Date field in a Command bean and the MMDDYYYY would populate from the it does not without registering a custom +What rules should be used to decide where to put the ellipsisShould it try to leaveEqual numbers of characters each sideEqual numbers of words each sideorA sentence that still makes some sense? +for releasing it otherwise when a new snapshot is deployed projects that use this plugin might mysteriously change their behaviour for no apparent reason +Therefore I do not spend time on fixing them +Hi if you look at the supported from there Drools is deployed as itself +Please change to if you intend to submit a fix for +Hi AllenFor the jsp that showing local filesystem I will comments on for detail requirementsAnd to simplify the design I agree that we could move finalized blocks only and do not touch existed dirs +The previous patch is not really good +v attached that incorporates Sylvains suggestions +Thanks StefanShould this new be used only with the tests? Or should it also be there for other interop tests +Note sure its a WAR issue +But we could make a clone Tool interface for tests if we wanted to +v attached to encapsulate the logic and not require multiple passes for each mutation replayed +Looks fine +The thing that makes me wonder is that only and cause problems with your patch +However the instructions for running test suites in the old harness on CDCFoundation say that the property should be set explicitly on the command line so the properties are in fact used on that platform +Youre right after restarting the problem went away +Patch corrects that issue +Devaraj Patch looks great but shouldnt we catch Exception instead of Just in case +As you can see below its just not there +its not a bug its a featureI thought we are allowing it inWhat is nodetool doing if not calling force +to branch and trunk thanks guys. +Good point EmmeranIll improve it. +Hopefully someone from JON team adds it to +Yes I actually see the deployed process +I was about to commit this but then something struck me the shared documents must still appear in the same xdocs directory +thanks Ivan. +that seems backward +If you successfully replace the current implementation of the OQL parser by aANTLR generated one this is a good thing +Would it be hard to you to attach a patch that is the same as the last one attached minus any Velocity related changes +It was me who first let it slip through +As such I dont think its gonna beAs a workaround you could always attach files to the ticket +I just committed this thanks Jerome. +If I get your suggestion youre saying that we could reuse most of and focus on generalizing its constructors and the init methods +PeterIts not clear what is the problem +Logged In YES useridBillI am in hell right now with +Committed on trunk as svn +Should check whether is related since this mentioned Batik +I also want to talk to him about why schema generation is always disabled with JTAI completely agree with your point on TM testing with +the patch fails to apply on trunk +Which for anyone who has tried is essentially and therefore to be feared +Not sure this patch is what we want to do for the reasons mentioned above +The missing filter jar and how to get and install it needs to be documented in or an additional in the base of the EWS product +Can it support characters written as uff format in the mapping file +ve committed this! +But if the user has used addProperty to add listelements they would expect the elements from the first matching to be returned as wellOtherwise that sould very good and I think it would cover all relevant cases +The page with the exactly same content should have only one unique URLThe canonical link does not help much since all of these pages are indexed separately anyway since they have different and the pages are not generating redirect. +Debug is checked to view the inevitable stack trace +Since creating a templatevar and default geronimo server instance caused major TCK problems what if we just create multiple instances of var? So youd have GHOMEvar default server instance GHOMEvar next server instance +Committed. +I dont plan to backport further. +After is fine +Jacob the shots looks quite decent thanks! Now whos going to put them on the site +reopen as wording not implemented +Uwe The point here is to let also take care of the filtering logic without doing any extra score computationsFor example that involves changing to not only accept Scorers but also and use these together with the Scorers to skip to the next matching document but only use the Scorers to compute the score valueWhat is the point of adding a score value to Filters when that score value has to be ignored during query search +All external artifacts for portal and projects now can be resolved through default maven central repositoryUsed portals apps and portals bridges projects are next to review if certain dependencies might need updating to more general available versions +The individual cores do basically +Committed to all four branches +Anupam please define the config key in for trunkIve committed the patch to +This is useful in cases where you want to define your preGoal in your subprojects and subprojects dont have any tests defined +Script to create and add unit tests folder to the version still needs hudson right +John reopen if this does not blockers and critical issues are resolved but not verified +We could create an SP to fix the serial and revert the writeExternal change and change them in as well +In your case they should be cloned but what about checking anywaysIf they are modified without cloning they will be rendered once only selection or notIf its not the in place modification of geometries maybe the default feature collection is stateful + +People seem very keen to remove the Callback Server in a future version of Cordova for Android and this will most likely be removed in a future release +Both issues were worked on in parallel +Mark resolved and assign back to Ajit for verification +Fixed in now that is released. +Good work +do you have stopped continuum during a build +related not the same +Convert method rewritten to use regular Roberto! + If you look at you will see a failure inside a trycatch statement +I have not been able to reproduce it so far +Attaching and +No test case added I assume it was an erroneous encoding declaration. +The module represents what we would put in trunk once a CLA is submitted +Patch looks good! I updated patch and moved the test into the test class +Youre talking about variables which are not static +I always prefer to err on the side of cautionNow if someone wants to set their acceptable and minimum factors then its their choice but in the end that doesnt change anything since you can just set a bigger and have the same effect +That being said we probably should see if we can relax that for the actual tutorials as it does not make sense there +This seems like a fine change +Ill get this right eventually +Otherwise we can move the classes to server common but these classes depend on the other classes in RM project +patch added +bq +Closing as INVALID on the basis of a reread of Xaviers comments that the formbean was indeed in request scope and that it works if he moves the bean tosession scope. +Otherwise the patch seems fine +Lets stabilize it and release it +It seems that finally we have a proof of hardware failure +I dont yet see how you can get the membership information Vector docid Canopy Id +We must preserve the old one +To be honest Ive not looked at the actual docs on the wikiConcrete suggestions for changes are very welcome though +I will try to write up a couple more unit tests so we can clearly define expectedactual behavior +is an other solution to connect workspace projects +The parameter in the soap message is still qualifiedIll need a full example then as Ive fixed all the cases Ive been able to reproduce +Regression tests for patch b ran without failures with Java SE on Solaris +I know nothing about Gradle so this is in no way an argument against supporting it +App callbacks are now handled by a separate Executor since some calls need to be made from deep inside a call stack +its the first time I see indexes created with the specific typeAll oracle dumps users threw at me so far have generic names +I need more information about this test +Add a constant instead of creating everytime +Filed for adding the testcase +Ill have a go at adding wrappedunwrapped support based on the code in the AntI have an updated patch based on the latest code for rd time +rebase one place +Ive not heard back so am assuming this is now resolved due to the upgrade of +stabr spaces with a strip of trailing tabs and lines containing nothing but tabs and tests whitespace only are no tests for this as all it is doing is removing tab characters from the source tree +So we should include API diffs in and +Alright just looks like debug mode is the issue not working right with MLT and the latest patch +Thank youBest regards by Richard. + Committed revision resolves the issues mentionedremoving webappdocs moved licenses to correct place cleaned up unnecessary memcached and hadoob jars removed workloadconfig and workloadsbin directories +Moving to I do not think we should go this wayLooking at Mule and SI you need to add a ton of XML namespace wrabble which a lot of people forget to do correct and you get poor errors at start what the real cause isAnd the Spring XML is really good as is +Patch to enable load Scott for the patch +The next step is to finish the Quick Start guide so that they are documented +Thanks Lewis +how about documenting the json +Please check that the patch was applied as you expected +Ive corrected it but I cannot verify the fix at the moment because my dev environment has been destabilized by a process Im still in the middle of +listener element should have maxThreads attribute which is no longer present in the referenced type +There may be some additional improvements pending +The tests listed in the description all pass now so this issue is resolved +Committed +Committed +Will be refacted in Phase III +I want to check if they are being fixed in one of the current pending patches that Vishal has been working on +patch against the gun too quickly +But building that up would be a long way to go +For me left and right did not even workSvante I like the the method name that you propose +java jar +Verified on trunk +Super +After that I will create a patch for this +fixed committed in r +Patch for javaPlease go ahead and commit the code and test casethanksApplied patch and added test casethanksdims +Nicholas are you saying that we should keep the additional fileId in the JSON response without being documented? Do we need it exposed in REST API? If we do we should either document that field +Seems to be fixed with jQuery +Im not against people trying to do it all in the LICENSE file and for tight projects with low levels of dependency that would be best but I dont want k license files +Part of the admin portlets are already Wicketized but not everythingFull completion will be done for release patches for the wicketizing the Service +The plan is to have some feedback before the next release +You could likely take the build simply copy the code for the package over it and then create your own reverse engineered patch +Integrated to trunkThanks for the patch binlijinThanks for the review Stack +Ill have a look at it +Bulk defer all frontend issues to +Attached patch file containing a second version of the API prototype changed to adapt to the new package structure +Will continue this strategy as more issues pop up + duplicatecommit m Duplicate log of HTTP header Committed revision . +As I am new to JIRA I do know how to share this +That single object allocation and the resulting GC is an order of magnitude more costly than the difference between instanceof and The cleanup is nice and if changed to an enum there are more gains a switch on an enum is faster than cascaded ifelse +Sorry the patch attached is the reversed patch + +section is is donetask is done +Revision unit tests and created +Finally got most of the ee running on my environment +I guess only the admins can do that +submitting am not sure whether this is bugThe trash behavior to me is a trash +This bug is still present inThis bug produces huge warning messages in user code and it doesnt seem likeit would be that hard to fix so I really think it ought to be fixedI could try to work up a patch if that would be helpful +Regression test case and the fixIve just placed component installation above other parameters validation +push out to +Definately vote up +Now bookmarks that have special regex characters wont work in the bookmark plugin but at least they will not cause it to blow up. + +push out to +Thanks again +verified fix presence in EAP + in r +Not generating these cached CSS files in the future would be an improvement but could be handled by another issueIm going to upload a patch soon which essentially handles the issue were seeing but none of the performance concerns +Windows XP english locale dutch dateformat set toThanks again for pointing out the date format issues in different locales +thanks in advance +We are good now. +agree with clockfly +Fixed +Andalso we can miss some docs eg we dont collect in advanceShould we instead move the child collector intoIt has access to all the scorers for involved in theparent query and to all child docs for each parent doc collected +I have an ajax filter above the table so without a fix the user sees the table header borders alternatively switching on and off when no records match the filter +I will attach the various back to Doug +John Thats the correct behavior +However it is a cause of concern if it takes longer than thatPlease feel free to suggest alternatebetter approaches +Ive attached a patch file for the changes related to exception propagation and the newHadrianI submitted the ICLA via email last nightBengtIll try to explain the issues you raise when I work on the documentation +It is good to know that I am not going crazy +When I tried out the current Chapter code of Spring Roo in Action with Roo using Hibernate this error returned +Note that this issue depends on Woops right I had skipped that step +Asokan +But the logs show no error in factPlease recheck and upload the right log +Patch applied to Cocoon tried this bug? Can we close it +JPA is implementation agnostic and if we are coding to the JPA standard it should work whether we use Hibernate eclipselink or +Gustavo you think you can contribute some notes in the documentation and javadocs +This is more similar to delayed components being activate by service events on many threads +revisit once it has been fixed in spring contextzip +Thanks LeonYour patch is in trunk at revision Trivial fix I did not backport I will maybe though +I have committed the changes +it is the scenario two in the my second commentA and B are friends so they should be in eachothers rosters +I tested with Maven. +Derby does not support changing the collation of an existing database at this point +Rob I assume this is an WTP issue +Problem is repeatable under +david you need to be a tad more informative than this +Committed to trunk +Do you have some example documents that you could share with us +make sure you run the test suite before committing. +Ask Ilya for attribute name for that bollean attribute currentdefault should be true calendar should additionally set client current time to date if time in pattern +Close after release +But I think the problem is not project specific + +At the Castor BNF attached there is at least a failure in orderClause definition +It uses the same logic than processAt least a x matrix is nothing else than a x matrix without the rd columnJeremiasThanks for figuring that out +Cheers +If you dont mind it would probably be good to provide a test case for this +Ignore me +Moving to +Applied patch to branch in SVN rev +closing issues for released version +I include the smallest example that I can come upwith +For now I have extended the and set the default value of field in its constructor +. +Hi AlexisCould you please switch to users mailing list and describe your problem more detailed? Dont see reasons to continue this discussion here since this issue marked as closed +It will likely need to wait until after the ESB release in September + +Thanks Tsuyoshi OZAWA for your comment +I believe the entire issue can be close because the should be correctly handled as part of the rewrite that happened some time ago. +I was about to fix this but we do actually handle properly + +Just checking +we split on this and the affixes are in the stoplistthis is how the whole analyzer works more examples inthe tests +HiI apologize it was a CVS delay fix is working you very muchMisha. + +Pattern object is a static object instead of being created locally in a function since the regex pattern object is very expensive to create +I presume this is not just for reduces +The descriptive text is being internationalized +Empty is devils work +Lets create a test for this and see what happens +There will not be a moving to for should be fixed as well. +Lets revisit this post as all our need a review. +If this is OK with folks Ill commit the one today +Fixed in rev +I am happy with the patchRobbieCould you please commit it +The test failure is unrelated to this patch as previous patch on Jenkins also show the same failure +Committed r trunk r branchx +We add a shutdown hook in precisely for this type of situation +Eh I meant for the second list item +looks good to me +Hi LeonardoThe centralized configuration you mentioned as another use case is alreadysupported by the Framework +HiDo you have the link for the whole plugins bundleI tried to find on the website but there are too many release and Ihave no idea which one to use +We tried to look at the requests that occurred during those times and didnt see anything in common +We can raise new JIRAs for these +Like selling cars and let the drivers walk with the broken clutch or on Route and leave them in the nearest car service +This was fixed in and is a duplicate of. +applied +Patch looks good to me +build on +Thanks Isuru for the contribution. +So how do you expect it to work? The maven plugin finds the files in the usual locations and adds all the type descriptors it finds that way to the generated descriptor as +but you dont show the static imports in the doc +It usually aborts after documents crawled +Even JSF contains a bug added workaround fixes your issue I tried your test case on ASThere are still some issues with JSF on AS but I will create new issues for that +I concur +Confirmed AWT command line option repaired. +Does DB on zOS not support any DATE columns? What type of error were you getting? Can you provide a stack traceThe reason that SCHEDULEDATE was originally a DATE type is because it doesnt really need much more information than what day it is +A new clippingpath has to be intersected with the old one if present we have to set the clipping path before every to the ve forgot to mention that the issue has nothing to do with my first guess +osgi tests added to both trunk and branches. +Np just trying to clarifyI dont believe that this will be fixed unless someone stumbles into the cause +Reinstated templateWe will likely have a project archival process introduced shortly +Thanks again for the patch +Randalls patch applied in think now that the js tests run with make check something is needed here +People are still using record io +For now we dont have a big issue because of this but we should fix it to make the code extendable. +antcore test passed now +fixed and applied thanks for providing the patch +Thanks to Jamie for the patches. +MaxI have test this issue on my linux machine +Patch updated +This is the only safe way to do it. +Anything else we need to clean up +Browse to httpslocalhostmyapplicationIt should sayServer running +Adding relation s not obvious to me that exectck needs a different JIRA component +Vasily Ive just succeded to reproduce the crash myselfyou need additionally copy jar files from HARMONYJRElibboot to the folder where you compiled test will use these jars as entry data +Theres been zero response to my bug report in +I also want to check that distributionSite and distributionAddress from the POM still get honoured. +CME +It is definitely disappointing when a tool cannot deal with generics correctly +I would prefer to have some concrete code in SVN and I think the stuff from Rida is a good starting point +Can you regenerate it +Reopening this issue to get the following confirmationFS is updated for Database key to the be same across resgionWanted to know about the managementserverkey? How is this being used? Is there a need for it to be the same as well across regions +Undo +The patch works for me +I guess the question for is how to package and distribute such a script +I am going to take a look at this next week to see if I can pick up where Ian left off +I dont think you need to change inside +Verified looks good +Roman +Thank you Simone I also concluded keeping track of the document level was the best solution. +Correct me if am wrong +I take it back +Vitaliy many thanks for your activity on that issue +Furthermore Jacek found in his experiments that encoded scanners were actually faster than scanners on decoded blocks +We could cancel and skip this step for example +This issue makes debugging more painful +Doacan AFAIK this patch is ready for committing no? If so I will go ahead and commit it and close out the issue +The verification will pick up any situation where the old data is somehow mixed with the new data +We also need to address backward compatibility so it can work both in scalar and a distributed cache file mode +Thanks +vote complete passed +This change is already epic +Addendum to previous patch to fix a minor error in Counters format +After rebuilding everything works +Proposed fix to be applied to orgapacheactivemqbrokerregion by SVN revision +Hi InfraWhat is the progress status of this issue? Wed be glad if we can have a git repo the sooner the better +Thanks RonYou mean i need to do some work arround in lognet dll for this +It received same exception as reported here and included previous version supported I use your fix that is jars to resolve this help me in this +The only thing necessary there is the requested data length +This symptom is only seen in versions of Derby or higher +I am working with the Cobertura plugin and this issue still remains +Marking it Cannot reproduce please if anyone disagrees. +This already exists was added two years ago +Its not my favorite option as I was hoping to keep the beans schema out of the security configuration entirely but this is good enough for my purposes +will fix it in the CVS +We can ask development team that please add support code to IRB +Please see my completely object to this rejection +Rebased patch on trunk +Heres a patch that fixes this +Path comitted closing enabled issue is now closed +Sounds like a race condition between the eviction thread and the starting of the file cache store +hql file +Please provide your feedbackRegards think that for the we should require that the table exists +I plan to work on this soon probably next weekend +As there doesnt appear to be anyone else whos bothered about clustering performance with Tapestry weve implemented solution and tested it with our application +Note that this patch depends on +Modifications were chosechosen prepended is valid herePlease verify that it was committed as you expected +The patch which fixes the problem +Not on stream or in the stream +Considered implementing a implementation of extending and adding it to but decided against it as that seemed a little overboardAgreed +Patch is ready for review +Concrete use cases would help a great deal in the design of this API +stack ok +bloomfilters have little to do with scanners +Committed to trunk and branch +High priority for M but this requires a complete rewrite of the content assist infrastructure + and are used AT creating a installation jar with the so called compiler they are not used IN the installation jar +If the layout is set to positioned the result is as expected +Great thanks seems like I made wangc poke around bad file +Thanks Tom for reviewing +It would great if some one familiar with could look into it +I did apply the trivial patch +Issue is now resolved +look at here is a big list to describe picture properties are grouped to objects like Bytecode CP Cache Connections objects will always be available under node Properties +make installMakefile src No such file or directoryMakefile src No such file or directorymake No rule to make target src +on removing those flags personally I find them unnecessary and they complicate the buildAnd I would love to Lucene parallel like Solr now +Unbelievable noone has seen this so noone must be using and JSTL together +DRL rules already supports import so this is a purely GUI issue +and sync with me after youve committed +on SOLR side good point although the SOLR dedup is based on signature only IIRC and does not take the score of a doc into account +Still its only that one thread thats accessing this codestate so by the spec volatile should not be necessary +just committed this +its not a problem of the Evaluators but of the see just tried on this is still unresolved +Running Tests run Failures Errors Skipped Time elapsed secRunning Tests run Failures Errors Skipped Time elapsed sec +Merged to the branch in revision . +comments on is available +doc to xml is ok need to update the support hypervisor versionClose as fixed +firstColumnsecondColumndate instead of firstColumndate secondColumndate +is a regression from and this should be fixed by +See if making Cache field transient solve this issue in CDI test +The issue is not about its efficiency but its behavior which may surprise users who are use to various checksumdigest commands +Gary This has passed through QA on all releases +is this incompatable because the seralizer deserializer is changed +I think that this will confuse users of NATIVE authentication just as it confuses users of BUILTIN authentication +Proposed SOA TP Release Note The JNDI default configuration has been removed from quickstart files +Ill make sure that it ends up in the jdocs. +Hi JorgeWhy would the minimum be MB? You can assign any value for memory that you want and it will be applied like that +OK we can tweak few bits later on if needed +What regression? I dont know of any regression introduced by this patch +The attached shows the memleak + in ER +Proposed patchWDYT +Toby said he thought the missing SCMTracker combo in Safari might be due to a Safari bug +Can you attach a failing project +Can we resolveclose this +SateeshOnce you have that code in place please give me the commits so that I can cherry pick over to for retesting +I think we should close this JIRA +This is also a problem because it prevents classes from being loaded from a dont think this is required for. +Committed revision . +WC Director Tim approved the use of +Do you have a sample that is failing +Oops day numbers were too big in month view +Ive included info from my testing as well as where to find the copied repos and test dumps for your review in the next comment +When I originally wrote the I had that type of functionality but it was removed for the initial commit because of exactly that issue +no problem i was just wondering +I should figure out how to get to create patches without it +So there is no need for this jira +Theres a problem with Internationalization on the Calendar +Shall we go ahead with thisBy the way I have an Apache process question should I even attach patches for such stupidly simple things like this? +What if you have multiple modules using the same nativecomponent? Since the +The main point of this jira is to not wait for updation of metrics +FWIW there are some require perf fixes on master that may help +Got this on win bit as well +Duplicate of. +is resolved +May be you just need to clear off the browser cache +Ive added the xerces lib since were including other required support jars in lib +Jim I couldnt decipher your title let me know if my bestguess translation worked +I wouldnt want a rd occurrence of this +The problem described above was building on windows then running on windowsI just tried running on windows the launcher full jar built on linux +all if you need the functionality now you can always checkout the source apply the patch by Peter Butkovic and put the resulting jar in your local nexus repository +One minor thing you no longer have to check for uf random never returns that char + +It seems to me that the if statement there checking for a clientName should actually be checking whether its the last node in the pipeline +Hi DennisWhat is this client code doing? Is this your program which uses HDFS to talk to datanode and namenode? ve run the patched for a few days and the number of CLOSEWAIT socket to port drop to zero +If there are no objections Ill commit it shortly +The analysis report is not self contained as if I try to drill into the details its trying to connection tohttplocalhostmemory +committed r +If a developer has created a partition on hisher own heshe should be able to decide what should happen with the context entry +is only created in which has the config object +Reverted changes to committed to revision +hai jacques I did as you said and continued and if i click on Create Order i got some error like this The Following Errors OccurredDid not complete the order the following occurredOrder creation failed please notify customer serviceBut i dint get as you said in the description +Never seen this youll have to provide a method to reproduce +It now looks as if SQL Roles will be documented at not +Thanks for the review Andrei. +Does this design satisfy the back compat issue for the response and exceptionsI have to review closer its on my short list +Unzip on demand fromThat seemed like a decent ideabq +Awesome! Thanks Andres. +Current behaviour is what Mamta says +Once was a proxy issue +ThanksAlso when you check this in could you also removesome empty directories +This is required for our action processing + makes no sense for numerical values +Is there more work to be done for this issue or is it completeIf all the work has been completed please close this issue +Yes I understood that part +The test case simulated the error condition and you can say that it is flawed but I didnt write it for amusement purposes but to simulate what we had seen previously +Thanks for reporting +js tests are OK +If there are no objections id like to commit this one +Ive made a few small mods +Jason feel free to take back over if you can suggest a better solution. +The server would periodically check its own status +Patch allowing use of PDFBOX on apologies I was distracted last week and failed to attach the patch file when I filed the JIRA +resolve as out of date because of month no issue apparently didnt solve itself during the last reopen this Stefan +Despite the alpha label there are a lot of people using it and breaking the compat should require a really good reasonCertainly its a shame to have some items in there but theyre very simple serializations of strings +Attached is the new version +Moving to +Ill be away for the next three weeks +This will not be fixed in Only AS and standalone. +Hmm +Could we add this item as a troubleshooting tipIs the issue that the user is not able to log in to Management server UI +reattaching with ASF license changes for JIRA and JIRA as both have some common files +I would like to verify if this issue still exists +Will commit it now +This issue isnt relevant anymore with the move to git +Take back my last comment as Kurt was talking about uddi +to trunk +Any major rework like this should go in as part of the new pluggable admin portlets in +I forgot to actually not seek in the previous patch heres the updated patch +I see here also a problem with declaring the depends relationship via a method in the target +Approved for +I see that has already been changed for Feeds +more to come +Once the release goes out will commit these to the documentation on the website +is the patch plenty to look overThe method access might not be as much slower as I thought might be closer to a couple to than the I first guessed +Mathias can you resubmit your patch using Subversion please +Caches are preloaded at startup and these preload Fqns are loaded regardless of whether it is in an active region or not +Not sure about the word simple Is the implementation too slowmemory intense itself to be used or is it the probability of complex stuff someone could enter? I dont think an ldap filter would give better usability for of usages compared to a simple substring matchingIn this case id just use a substring matching +Verified. +Thanks for the update +It can be passed to the optionsCollection +Committed at subversion revision This patch introduces a new runijscript target and runs each of the smaller load scripts individuallyTouches the following filesM javademotoursdb +Just to make this explicit I reported it as a Windows bug because it doesnt occur in a nix thought I made this clear but can see that I did leave that unsaid +Message level are attached using to the binding heirachy +Yes Utils should be better +please file additional specific bugsenhancement +Agree with Felix +It should be the serializer of your XML parser thats performing these magic conversions +The last patch missed an import statement +Or we could add a config option to disable the check thus enabling longer than chars for where this is supported and the default would be check enabled +Thanks NathanThe no cache patch looks betterApplied to LUNI module at repo revision Please check that the patch was applied as you expected +I could easily imagine someone wanting to perform a rolling restart of in a cluster with clients wherein the whole rolling restart process might take less than minutes +This will also affect proxiesThe question is if some payload is submitted with POST how do we know if the intention is not to actually gzip the outgoing payloadShould we introduce a contextual property so that GZIP can also affect the original outgoing payload +note i can help move these around if needed but I need to know the intent of these jars and then your build needs to be updated to account for future releases +Could cope with something like that? Either way I still dont really see what the problem would be +But there is no JSF code so I dont know why +It seems like a fine way to specify it to me though I think it would make more sense to get opinions from the people who plan to use this feature +What else should we do here? Avery seems like most of it falls on you as I dont have access to Jenkins +We might want to change this +Thanks +I notice this because I am diffing the and code +ouput is also attached +Patch as indocated above. +For example for me Harmony is reasonable in the first case while RI is reasonable in the second and third cases +There is probably the same issue with the other archive output streamsI dont know if its best to change the code in or tell the user to initialize the with a buffered stream +So this is a must have feature +Suddenly dawned on me that this may not be a problem in the main web application and unsurprisingly it wasnt +Thanks for the patch! +Which one it picks up is configurable via environment variable +Draft complete and committed to SVN revision +Thats absolutely right Robert +I have started again with a fresh copy from the svn trunk and I do have a questionThere is a class which should group everything related to counters but then every other TestDriver has also test with countersIs that normal? Where should the additional test for this feature should be added +Its worse than that plexus remove it before calling the mojo +Two things scare me off +The fix version is +Should be available within the next hours +Otherwise could remain. +Thanks. +Im slating this for for the built in solution +I committed this to the branchThe C code should probably make use of the std namespace instead of using global imports but that is a broader more pervasive issue that can be dealt with separately + for this fix I cant imagine when this code was ever correct but i can see how junokepler would be affected more +thx +Rich also add a note about this new option in the release notesThe community have frankly asked from time to time about this improvementSo its about time we listen +Yes the doc should be converted to forrest and put in srcdocssrcdocumentationcontent has been already done please see subtask of this believe that all aspects of this JIRA were addressed by an appropriate subtasks +Are you concerned about details of the implementation or additional demands that this places on the user +Thank You very much for Your positive review and for improving the patchAlso a small correction to Your previous comment it should be unit tests included in her patch as You know Im one of the girls in this years programIt was a great experience to collaborate with You and I Thank You for everything +Note that you can use either of those in combination with a location pattern so both of the following are validclasspathcommyapp +Is this at all related toDo these two need to be coordinated +There are other forms of messaging that can be looked at Java Message Service etc. +Reverted change. +I just committed this +Committed to trunk. +m mvn archetypegenerate localor +This feature is related to the possibility to expose the UNION graph of the Jena TDB store implemented by the with revision +Perhaps ACF does it so that functions such as and work? Adding psqno works for railo but breaks CF +The more I think about it the more I am convinced that cluster administrators typically have to administer many other servers other than Hadoop servers +Not sure how this had the noob tag this sounds like a fairly major undertaking +Each of the steps of the algorithem have been numbered in the comments below and those same numbers are used in the actual algorithm below so it is easier to understand and maintain +with me +However the test still fails for me also +Junit test was added +Attach the +Apache Felix Declarative Services has been released +Patch attached +Sorted out state nodes fixed node leave and added cases for only and opens up same problem for node and so now working on these for the same three cases as state nodes +Though the second approach has a drawbackThe user has the option to configure if he wants the datanode to be formatted when namenode is formattedIf the property is not configured then the behaviour will be in the normal way +Sounds like a pretty comprehensive test plan to meI see that impersonation tests cover the Oozie case scenarios +So I believe it is correct but I wont have a chance to be certain until I am outside the firewall +So is this an issue just when using the meclipse plugin? And just on Windows +I log the message at info level the first time handler called for each threshold type in subsequent calls to the handler messages are logged only if some object has been spilled to disk +Hi Ivan has now the moduleThx for your review from what I have experienced building hbase depending on is tricky in the sense that hadoop source trunk is evolvingI have come to the conclusion that maven build pulls the snapshot jars from the repo but this can give inconsistencies and break the hbase build when you enable I have overcome this by installing hadoop locally with mvn install and after packaging hbase with the maven o option so no dep is fetched from the network and I am sure the deps it depends on are not fetched from somewhere elseA the time I have submitted the patch I was sure everything was fine compilation and even working with a hadoop yarn clusterNow I should retest everything but I have just changed job and country yesterday so I prolly wont be able to do it within the coming days +I think this approach is clean and wont need to be revisited and wont add to the maintainance burden in the future +updated patch to newest trunk to patch against latest trunk +Submitting to verify the trunk version of it +It may make sense to have users either predeclare the variationsAll in all I would like to see Solr have better support for languages in both the schema and the config +Committed in r +This is an important issue for me because I have foreign keys that dont relate to a primary key +Thanks I will check it out +Alex could we extend file to have HTML set of attributes +If they happen to have an MSDN license also and they provide that Amazon charges them lessEven though Apache developers are allowed a free MSDN license we obviously could not distribute it in source code +Please mark all new as evolving Please add all new configurable keys to commented out with instructions for people regarding what they mean param job job configuration its actually a Job not just the config right +It is working for me both from Ant and Gant +I hope could have as many fixes as possible as I need to build something with it badly +I really just forgot about the group Id +If I crank down the logging to debug I see the being logged times for our application +Luc that is precisely my situation +We have to update docs or restore the previous whats your take on this solution a combination of the old and new behavior +So creating Archetypes are not released +I am not sure if its possible have to look at a the patch in more detail but would it be possible for to be a subclass of? Basically it attempts to run all the safe conversions from the parent and then falls back to some additional unsafe conversionsAndrea thoughts in this one +Sorry hit Enter involuntarily +Though not a likely a problem in the THREADID case there is the possibility that if some transaction manager uses extremely long XID that truncating could result in two transactions that are actually distinct being mistaken for the same transaction +The timestamp changes the file size changes etc but hte server is not picking up the new jar as a change +CI has been happy on this issue for a couple of months +Snapshots live in the Codehaus snapshots repository not in the main repository so it is necessary to tell Grab where it may find the dependency by adding the snapshots repository to the lookup list +I changed get and get to use get +Ok now I understand it is more simple than it seems to be +Ive got about junit tests to there anything I can do to speed this along or is there beta code out there someplace that I can use in the interim +Latest changes looks fine Chris +So I did some debugging tonight +There should be tripple opening and closing brackets around the code in seperate lines +I think that passing it through URL all the time is not the best way for that and I think that this should be dealt with similarly to the active language so that once you choose the language it is active until you choose another language +Incremental publishing doesnt work for Java files within a same thing happens when creating a Touch descriptors action restarts the application you start the server in the debug mode the standard Java hot deployment will work correctly +By startup do you mean my or ondeviceready handler? I do various initialization in each of these but neither should run on a reload because then they would reset my application state +The class in JDK is provided by the JDK not. +Modified to add and javadoc links to the Manuals tab +We create triggers dynamically when the master instance is running using SQL statements +So thanks for taking this up +I had to force Maven to use of version to work around the problem +The main work on the referenced is done moving this docs tasks to. +Closing issue. +Attaching a patch with the test case and code fix +Turning off fixes the issue +The clustering produces some funny values for the clusters which I have not yet understood +So now Im handling this from whithin wsclient code +Contrary to my previous comments I can no longer reproduce this problem so closing this issue. +I modified it to release the lock before officially going to sleep and that should prevent this ever happening again +From my experiences with Nutch Bixo I think that URL normalization ultimately needs to be more structured ie first break the URL into pieces then apply rules against the pieces +HiI will provide a sample portlet war and upload it into the liferay community somewhen in the next week +Yeah it seems +If servers are upgraded but not clients then du will not work +The fix was committed at the revision +Tried Dejavu with Japanese and characters didnt come out +Thats fine with me agreed + +v is what Ive committed to the snapshots branch +Updated patch I forgot to add a check if the was already closed +Since there are lots of test case failures in the trunk I would open another JIRA to fix themAt the time I make this comment the testsuite is able to execute so I would close this JIRA nowAny concern pls comment or reopen this oneForrest +If I cant figure it out Ill open a new issue +In the attachment you will find the whole sample as exlipse me the case is no longer present with GA and can be + for the patch +Thanks for the patch Sai! +Some minor comments +Reopening to evaluate patch or move to Java +Please attach a patch before switch to other workaround might be using a Toolbar. +Another colleague reports the same to both code samples is that static methods in a class are being be a clue +This change in generate phase can be found in and +The pattern will select the last IBM tickets as the filter is applied before the sliding window as of DroolsThe new work on the sliding windows is being done under ticket and will be part of Drools +Note remember to remove the Ignore on this test once this bug is fixed +Thanks EliCommitted to trunk and fixed this by merging to +works for me +HA Partition name private String partition private String partition +If you find further issues please create a new jira and link it to this one. +No problem Pete I kinda guessed that this issue would take some time to resolve +All besides for which the standard states we have to return the highest implemented I dont know this would make us even less standard compliant +Did go final yet? dont think it has although must be close +Thanks +Tom Fennelly saidif theyre going to be singletons they probably need that +I think that to meet the practical need that led you to open this that is going to be needed so I would say first just patch the current impl to allow the threshold parameter +I propose to use the aliases as stick to what we state we support in the capabilities +Patch v from review board +Sean please open a new JIRA for the issue. +jobdir needs to be freed before execlp +yeah that alsoCleaned up the sublist callChanged random number generation so that every class gets a different set of random numbers +changed fix versions cleaning up bucket +I output only a slow request in log by setting the threshold in a propertyThe default threshold time is ms +I will commit it once Jenkins s it +Hi Keith can you review please? Thanks +In the meantime Im reopening this issue for tracking +but doesnt work for el wich has been provided threw design options dialog +URL injection is only possible on the Client side error persist with any other type of the field annotated with +Im preparing the release and I will update in Karaf as soon as the artifacts will be on Central +Patch against r including unit tests +In case its not obvious the primary purpose of pings is to make sure that child processes die promptly when their parent dies when their job is terminated etc +I might use the auto schema generation next time. +For one thing they dont allow you to wrap the request and response objects which is essential functionality +If requires native code it requires native code +Resolving. +Todd so do you think the patch is good +Related to the following both the old model and the new service registry based model +The naming utility detects when identifiers returned from dbmd require delimiters +Do you happen to know a perfectus build number that this will be included fromeg I am using now is the fix in that +to merging as well I prefer having an project with a concise name and a precise focus +Yeah +Will commit when tests pass +So thats the bug to fix. +No problem +Patch to add warnings about newness of this package to the javadocs. +Otherwise Its a waste of CPU and network bandwidthYou think we should up the major compaction time default to be a week +So I needed to save the current selection somewhere and I found no better place than the +Added zip of debug version built with Adobe Flex I posted a zip file of a debug version built on my machine +Moved the call for readDefinitions inside the init method ofModified the calling code accordinglyDirk can you test if everything is ok? closed due to the release of Tiles +Is that intentional or should it also reference theThanks Peter +you have a but the API only accepts a String if it would accept a the could be passed in directly. +Do you mean the me extensionIm afraid thats the way me is designed + lgtm +I just dont want repository definitions to be transitive +moves task initialization in the capacity scheduler to a separate thread +But a null envelope raster is certainly not being copied the layer back from arcy via and now its fine! +I am using version +Thanks +I shall commit the changes but have them commented out for our next release +I can see that behavior as well and will work on that. +Earlier Suresh commented that we should support more than MB of preallocation at a time +What is the status of this one? If no ones working on it I can take a stab +I have committed an unit test which shows the issue +rdBest RegardsCameron Goodale +Patch applied in the pure html pages are converted to +Table editor only shows createTimestamp creatorsName modifiersName and modify. +I tried a series jar but that wasnt enough and I dont care to debug class path problems +Added logic to clean and rebuild project containing the generated web service model after the models imports are rebuilt +I reverted my changes +The following comment we received from a customer is relevant to how accurate our platform availability measurements are +submitting a fix for and test you! I will as soon as it will be available in some release +I just committed this. +Yes that would be the options available AFAICT +That avoids expensive copying of the unpacked bundle contents to the location under the GERONIMOHOMErepository directory +SO reopen this jira to patches are created according to the s the complication is passed +Both runtime and workspace is unchangedOnly happens with SRRestarting the server has no effect when the problem occurs +They should be removed +IMO this would be beneficial from a build systempackaging perspective and also if we port it in such a way that we minimize dependencies on other modules other people might be using it for their purposes considering that we will fix several generic issues. +If it is a bug in the mail jar we ship the issue would need to be logged with geronimo. +That does appear to be the issue +If multiple matches are found then the user can choose between them. +Committed revision +Tim removes the searching for new jobs and moves completed jobs to done folder hence keeping the folder clean +Gleb the as default value is applied when inplace is updated with new Class attribute +Any ideas +patch does this +Bringing it up to the CDI expert group is definitely a good idea +Good remarkMaybe added a Langage menu in the page when someone have migrate the Chinese versionWe can choose this way for the existing ofbiz website page with a transalated versionBefore migrate to the Apache CMS we should not have a large number of pages +Am I the only one who cannot build hive with offline option +Can this be integrated as part of the regular tasks done during a release or development cycle +Attach is full output from sonarsonar +In the future someone may want to dynimically load different transports +committed to and ported to and after release +not much point really in my opinion since all of the work is already done by Maven and the suppose I could see an instance where someone would want to create a test deployment that used some resolved libraries but again I dont see that being a very common scenario since the tests more than mostly tend to rely directly on the actual addon under test +I will hopefully soon have time to work on a little sentence detector refactoring +Verified in JBT nightly +push out to +I see the following errors in your attached logchgrp changing group of homeraghuhbuildcontribzebratestdata Operation not permittedSo I believen your tests has encountered disk permission problems +Oh and one last comment and the only one I would consider blockingWe need more documentation in the comments about what is expected how it works I will try to send in a patch tomorrow +Need to update +I dont think using respondto and send would actually add much different behavior +I guess the problem now comes from a bad file URL +the release of components servicemix and archetypes are now completed +Since map isnt sorted sometimes this puts STATUS before TOKENS and immediately notifies for it causing the NPE since TOKENS isnt set yet +Side note When we do this to use as single sign on needs to be removed +See my comment +this is expected behavior if you are using databasesREPEATABLE READ The default isolation level for tables +Thanks DeepakYour slightly modified patch is intrunk r r r r +I cant check this in just now as Im waiting on SF to move this class to as part of +testtemplate was created +This enhancement is applicable only when the runtime environment for Muse is +extends +Fixed in +Glad I didnt doc it +Thanks for your help. +That should be rare in practice since typically if a cluster has no data you just start it up without bootstrapping but still we should protect against handing out tokens that already are in use +Because timers execute asynchronously the underlying timer events should behave more like a fixed rate timer than a fixed delay timer +V removes errant hbase shell test +Issue is marked and actively being discussed but has no assignee assigning to most active committer contributing patchesdiscussion so far to triage wether this canshould be pushed to or not +Thread dump illustrating the problem with the locked threads +Hey Markus I cant replicate the issue with the latest snapshot +Unit tests are also implemented for to test for correct translation from a standard expression to a +But please add a if the page container has a the will need a higher style value to show upto the Javadoc +Well after a lot of testing and debugging I finally figured out the problem the geronimo mail implementation used by Axis is faulty and not able to parse the correctly +Defer to defer open issues to defer to not essential to address for +So the ideas in it can be reused but not the class itself Im afraid +testing when online shows it breaks that so still work to do +The intention was to allow to control whats going on after an exception occurred and to present detailed message in devMode +It does not adhere to the whitelist currently +Thanks Jeff +What about using the default collection editor for this? It will make trails easier to only problem is to know when you want the child element to be removed +Hope is that this backport will elinate current intermittent failures in the nightly tests against as they did for trunk +Better yet if the character is appended after +Frankly I dont really understand how having a Configuration being Serializable is particularly useful and the complexity involved in this seems to be way out of proportion with the benefitIn the case of a that would mean serializing all the configurations that participate and then including them in the serialized composite +upped level to reflect the SOA modified in revision they still have to be applied to the CP will also setup JDK runs of the CI on trunkCP +BTW there are lots of CI tools used at the ASF which DO support maven +I hope someone monitoring this issue will supply a hint +Fix committed into JBPAPP revision +this patch should fix the following issues incorrect source repos in missingincorrect provider missing licenses missingincorrect descriptions etc +If anyone wants to have a performant implementation its reasonable to expect him or her to use the streaming API +If you agree I will contribute a patch for this +So no need for Patch for DIGY +This bug also contains a different Similarity implementation as well as a +Something along the lines ofThe classpath can also be specified directly on the +I will start a discussion about it +To clarify in this jira you intend to +Have you monitored the jvms for gc activity? Are you using CMSincremental gc rather than the default GC setup? Im all for adding balancing but it would be good to rule GCswapIO out as an issue +what a rush +Added link to implementation +This isue should be closed obviously +Hi Dinesh I am having a hard time fiuring out what the latest attachments are +well you can build this quite quickly in the old version and use itIf you will wait for a new release it might take some time as we have to go through the Apache Release process and that can take some time +Hi AlexeyI reinstalled my OS today to XP +We have deployed and tested plus this patch to the affected cluster +Wait a minute +Each site would probably have their own settings and a way to gather metrics +In your case I have no idea what Olaf Merkerts solution to the AES decryption means in terms of licensing + +and will be pulled in as Geronimo uses Glassfish JAXB implementation +No problem thanks for the updateregards Malcolm Edgar +Fixed in trunk +This essentially improves the average similarity of users in a clusterAbout documentation I am caught up with a few urgent issues at work and will need more time +Committed +and heres the test +Please see prev comment +What would an address with a prefixed TCP port number look like +Changing relevant will not do what I am trying to do +sigar is supposed to be loaded in the server +Well still need to ship the backported for hadoops and unless squeezes in soon for early releases of the line as well +on latest patch while still hoping to get another committer to take a look +If anyone else is interested please feel free to pick up the issue +I am uploading a new copy of the S Integration +This was causing the dev mailing list to get x notifications for every update to this issue. +This patch updates all the dependencies that are available in Maven repositories and all tests succeed +Robert any news on this patch? Can we get it applied for Lucene +Actually an update that this is still failing on my home machine so I need to dig a little deeper +Hmm my mistake +i think you should make this change and proceed forward + +Note the latest changes are in andAlso attaching the revised source files +work +Thank you for the review. +have to defer +EJB security alignment will be an ongoing work until the unified meta data comes out and ejb meta data gets aligned with is why this JIRA issue is not resolved yet +Not an issue on or upcoming so wont fix +Fixed in the release. +applied patch please now looks to have been fixed +In all works fineThe test case used was Password is hctrfnot workingeebbecaefworkingEEBBECAEF +Is it better to work off the github or would you prefer I continue to upload tarballs here +There are no test failures on Hudson anymore. +I could not find a way to give separate committer priveleges for just this branch so we would need to manage changes through patches the same way as we do for the trunkThe current state of the branch is the same as of the trunk +it mostly works only under certain circumstances the values arent preserved +Pushing to +You mentioned that the duplicated documents have a number of conflicts +Modified to use a to create the result. +Will close as soon as we have the first version of release notes on the doc server +I think I wasnt clear on that my domain object and the from my example code is not +The objective is to move a subset of file so large that naively copying the whole file is a poor idea +If there was a way to determine the expiry time in advance then it could even display its remaining life to management toolsweb +For the time being I kept this as a separate project as I am still not sure where this code will fit in Castor project package names etc +Hey there GabrielI was thinking what would be the behavior if the WMS request matches a tile perfectly but includes TWO layers? As I understand the situation GWC will not handle merging two tiles together right? So in that case would fall back to regular WMS renderingIf that is the case then I would suggest to have implement this functionality namely to ask GWC for each layers tile and combine those tiles into a single image that will be handed back to the client +let it be a stand alone java beanAm trying to gather information here +One additional note +I was simply responding to your assertion. +Thats interesting +Have a great day take +Regressions ran ok modulo an error in dblook +If this is a bug that was introduced in then we could try waiting it out until the Eclipse team corrects it +Markus sorry I didnt see you already provided a patch when I wrote my previous comment +Now i also made the strange anonymous inner class a named inner class to get rid of the strange boolean holder implemented by an array +Just out of interest how did you go about cleaning up the initial patch? +Im moving all bugs out to and we can move some bugs out to to get some release action going +This issue is duplicate of +The patch has been applied +Also in. +We are planning to commit this soon +Or leave it open to track the upstream issues Eclipse +It could either ignore versions or programatically update the grails project at this time itself or it could invoke setversion to do the job at this time m all ears but I propose +I still work with all projects at once but Im absolutely willing to chage it as I hope that working with the will be fasterupgrading to Java would be great Im doing some crazy stuff in JAXB to traverse from J to m not sure who the mysterious group are that still use J and ANT and arent willing to upgradethe only one point Im aware of that is keeping us on ANT at least partially are the test suites so lets give it a try! I think it is a step into the right Werner I was aware how it functions as I said Im using Maven in all my projects and no ANT +This would be a GREAT feature +One thing I would like to emphasize is that the error is not deterministic with the input data +The mq distro is really for deployment +includes test that has ordrord failing without it +Indeed configuration in the conf file would be enough. +patch after release. +Does it affect as wellThe only issue where I had to touch the linenumbers was but this was part of the releaseSo first of all provide the smallest test asserting the rownumbers +Ive attached a test caseYou should modify the value of the variable childrenCout into the method insertDataIf the value is about the error not appearIf the value is about or the exception describe below is thrown +Note that the attached file implements a slightly modified Evaluator interface where the newContext method takes an addition parameter +Second rev of patch to fix correlated references to outer tables by arguments to tableFunctions invoked inside subqueries +I amafraid this one will have to wait till +Installing help into via a msi installer is a major pain +But SaschaWhat happends to those who are already using the renderTemplate older version without cache parameter? It seems that its deprecated to me +Fixed sun specific import if JMX +was as I reopened it +Hi Todd Hairong Added the additional tests to reproduce this scenario! Patch contains Hairong code changes Todd review comments fixes Added additional testcases for reproducing this scenario +Im experiencing the same problem +No it was reported by users +There are also many people using Spring Security who arent using Spring MVC and we need to be able to continue to support issues can be addressed by using Springs or a similar filter implementation to set up the before the filter chain executes +My question in the original issue description asks whether we need to specify the schemaLocation in our file the test file spring suggests that we still do need this declarationIs the point here that the file specifies a mapping not from namespaces to local files but from schemaLocation to local files? If so why not just establish a file mapping rather than trying to intercept requests to the schemaLocation? +I really dont know what is happening with the contrib and core tests +the patch looks good +All set. +april guides are looked through necessary info is added +The downside to this is that its overly broad given that for Hadoop we only declare as needed +This WSDL demonstrates the problem +Trunk r x r Thanks for reporting this Junaid +After that we talked about my proposed fix here of turning the exception into a warning +Owen isnt the same generally true for hdfs as well? It doesnt implicitly acquire a token +Facelets Support section are updated +Default value of should be true and we shouldnt put it in it should be a hidden parameter so we can remove this in the next release +Updating patch for trunk +And I did fix this issue with an instanceofThanks James for your attention to this issue + +Here is the patch that adds a enabling users to define an endpoint for dynamic keywords floating around the route +I have a and Requires passedto my service constructor +Id like to get the comments updated as per the latest comments on RB +I have been unable to test because I seem to be running some other version of the plugin even though I tried specifying this version via pluginManagement in my pom +I checked in the fix +Servicemix will use CamelFreeman +Is that the only thing holding this issue up? Someone just posted a request for this to the mailing list and Ill see what I can do to help +Lots of people like to compare and measure us based on performance and itd be good to do as well as possible there +Great find +Are you sure the full mvn clean install does passElse you are indeed right that there are not many unit tests CARGO has many more integration tests +Thanks for the nice patch +Hi OlegGenerally speaking I think this is probably the right way to go +No need for IT +Jython engine has been released and deployed +On systems which use lockfiles its common to lockf the lockfile so that an orphaned lockfile can be easily identified by being +If they want to do it separately then they can do that too +Bulk closing all of these resolved issues. +Committed to trunk +If this also applies to please open another issue with a new patch. +ie this doesnt allow for arbitrary mixing and matchingOne can produce a downstream stack definition which will be dependent on whatever version of hdfs and mapreduce packages one is pleased with +Heres the information for Jie username jietang email jietang full name Jie Harsh I havent heard back from him yet +Im adding the patch now before I commit it +Nathan submitted a modified file directly to me with changes to fix this issue in it. +The other did not make sense. +Patch applied with modifications +In keeping with my earlier point re making the counters mechanism as generalpluggable as possible having a generic implementation of the log a modification event stream replay it on read that could eventually be exposed as a primitive on which users could build their own functionality would be a big win +A mismatched client and server will not necessarily cause this problem but the only way to ensure it does not occur is to update remote EJB clients when upgrading to EAP +It seems that this is a only issue +This also ensures we dec the counter as soon as we start procesing the exchange so the counter is fully up to dateWe should expose the JMX attribute in also +with suggested changes to handle proper errors from SSLConnect and to add the full SSL error stack to error message +Ok +qa didnt catch this because weve separated jboss testing from db testing matrix +With more complex projects where plugins are configured in parent poms but actually do anything in child poms the overhead is larger +I checked in the code from the patch for your useI have begun to look at the code in more depth and will be proposing changes to it and also Andrew Stitcher suggested I put this up on the Qpid reviewer site for wider review +Please reopen if disagree. +Other providers palo alto will add the rules +I think it should be pretty easy to come up with an early prototype and attach it here one that actually ignores XPATH handling for the time being +hi Hardy remember we declared the API as experimental while youre using it please consider a in case you need it you can still change the contract b think about if you like it so we could mark it eventually final in some release +I tried to run the test against the latest version in svn it works finePhilip could you please check if still you experience the exception +I think the problem here is that we forgot to ship the headers for cext with our binary dist +I can run secure jobs without this patch we still need to investigate +The other lesson for me is that we need to up the coverage of our tests +But the real number of threshold is debatable +If we do not have the docs we would possible need to remove the component from the distro +Added KiranYour patch is intrunk r r. +One application of this is like in process map reduce on the cluster +Merged this test case addition into releaseSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestslangSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +javadocs checksums added. +Maybe Im not following it but it doesnt look like something many users would be impacted by +Just got confused by the way a diff was being the more general side I do not know whether it is a good idea to remove codefragments like this +Thanks Jonathan +Rolled project revision revision +I committed this test under the aegis of this issue +this works should be correct I think +Looks good +And needed to be tracked separately from this particular JIRA we cant mix different issues in the same patch +I will commit it shortly to +applications usually dont make guarantee as to the order that worker threads are processed +It appears this is not a recent problem +Here is the patch for the same +Can we create a unit test around this? Hooks should print during testing and then we have some coverage here +I will leave it up to you to decide if this issue should be closed orll close this issue for the time being since this is a really special case that isnt generally intended as a proper feature +Patch has been committed. +I should have to corresponding patch for trunk done by tomorrow +hql menu is now in the menu +Because in that case its not a necessity to provide a custom comparator to the constructor +I think that it should be more focused and should be really just exercising the cache here instead of emulating a full run +Fix included in release. +Integrated into trunk +sorry +But I think we should do that in another issue +Isnt the fixing of the thread name possible with the old structure? If so I suggest that you do the refactoring in a separate jira +Sure! Blog post would be cool +Thanks Pavlin committed +Longest time to fix a failure +Hadoop QA hasnt run tests for the patchXingCan you run test suite and post back the result +Dont try to get the message size if its alreayd been delivered to the change of resolved shipped +You have the local revision number available but not the actual changes that is why Ive added the poll operationNow Im investigating a lock on introduced by this patch so Ill upload another one soon that will take this feedback into consideration +No comments patch applied. +Youre right I have just committed the improved javadoc +Jeff is that when converting the configuration +Thanks Anil this is in SVN r. +Thanks Teddy! +Sounds like we should increase the heap space +Post questions on the forum +Im moving this to M but we should be able to address it as soon as M is released since the mapping logic has been refactored for message handling endpoints and this will be closely related +Thanks Abhinav for the updated patchI will take care of this as soon as I will get a chance +Further testing required before we will close the ticket +Hi Dag +Can we have decom and move call the same method to calculate ranges to send to then do the actual sending and the same for bootstrapmove? Right now that move method is pretty hairy and it seems like it would be good to have that logic in a single spot +Upd +META +Yes thats why I linked to it +Sorry for the annoyance but i think its good to figure this stuff out +So for now until the details of the SASL over TCP transport are fleshed out HTTP is probably a better first transport for implementations +Verified on EAP TimSorry for reopening this issue now but I need some info for release notes +Even if its a bit more work +Todd how many RPC responses go beyond K in size? Roughly what would be your guess on what of total RPC calls this is +to commit +Ill try to find time to add the examples +Stefans code wasnt ready to check in and my change was too early +is set inconsistently sometimes with setMode sometimes directly +Can we borrow libraries that will manage the lifecycle for us? Would it make sense implementing atop some lifecycle supporting framework? Would it make sense using say any of the DI containers wiring upThe regionserver and master have similar need of a lifecycle as has been discussed elsewhere +The goal is to reduce and cache the annotation information +here is an starting point with lots of but something like this would be very powerful for users that have their own codecsits really just a start for discussion +Joydeep your patch probably also fix +The RM wont restart the app because it unregistered successfully but the client is never notifiedbq +The performs as expected in Flash and AIR for desktop +Seam objects keep references to source objects in Java model and +What if you decide not to have a setter and simply set the id upfront yourself +Please take a look +Im experimenting with something to assemble full configs from subsystem snippets +Does that satisfy the requirements? dave +sighYoure right is required +solutionremove the sync blocks from the methods +Reattach the patch file with granting the license to ASF + log file when aborted +That would be true if it took into consideration the comment density of what the Override provides +Barring that I think we can just leave this be. +updated web page +And then when compress table does a bait and switch trying to replace all tables and indexes it gets confused +nothing fancy in therepaul +Given the bug mentioned previously SLFJ version and were unusable under JDK but only if a version incompatibility was detected +Sure Oliver +I know I could do better +This Jira is about adding it again +dll +You give it a try +A new implementation that returns the value. +I am attaching a modified repro which executes some statements after setting to false and before the call to set +This patch was also available in away +Thanks Vinay for the patch +Camille It sounds right that we need multiple acks from the follower but it sounds awkward that they are acknowledging the same message +And if there is simple uniform timeout may not be the best way to deprecate an item in the cacheI will vote for simplicity of the code for now +The test case was insrcorgjbosstoolssmooksuibottestjira have attached the patch +since is not already released it would be nice if you can put this in +Gavin that is not our development way +I will commit the patch +Yes we can can u post a patch jason +Im not sure if strict immutability is necessary theres everything in between tooOne can simply say that all changes should be made before first use and after that point its undefinedUnrelated question I assume that this would retain the same flexibility as we have today +bq +Seems buggy to me if I setup a runtime and the IDE doesnt know where that runtime is like it may lead to further issues +This should be resolved by commit sorry +Committed to trunk +after testing it locally for various environments I would consider it fixed for now +Thanks for checking Mikclosing this will follow up with a different issue to put a clear warning on the docsHardy do you think it would be possible to create a better message in the exception? +I added you to the committers list in JIRA see if you can close the JIRA now +Basically the project will need to decide to go ARC only for this change possiblyIll tentatively slate this for this wont make it for which is being released soon +latest patch merged with the trunk +If not then Ill probably just add the field to the entity when I commit itIm satisfied with this implementation +The functionality of the current scripts is no different from +Another set of comments +My bad +think this will just work as expected. +Thanks for pointing it out. +Patch with some test cases and corrections +I dont experience this problem in the new variants of +The patch seems reasonable to me +This is because you are not the PMC chairAs the Wiki says PMC chairs have access to execute this commandPlease ask your PMC chair to run the command +Unit tests are also implemented +Oddly enough though presetting it with fixed width and height or with visible overflow doesnt seem to do the trick +This is an incompatible change inIf we want to support we have to reintroduce the munging code +Ill attach new server and agent server names that are detected need to be unique across the inventory +Committed at subversion revision +That would be very costlySince the maximum time allowable in one iteration is passed to balancer why would using local timer to gauge elapsed time be expensive +But the fix should apply to trunk as well +I unpacked DEV went to the cxf installer folder and ran ant then started the server and ran the deploy for the bpelsecureinvoke and it deployed fineWhen I ran the test it had the signature verification problem which is a known issue Did you do anything different to what I tried +Were tracking the solution to the problem in the wagon in +action now throws an exception is a fault is returned without message content +It would be nice to have this in +the same patch to fix download problems +Ive checked the example and indeed there are no more exceptions +Investigate for if they should be in SOA Van will set the Fixed Version +Also I could be wrong but I thought some didnt allow identical constraint names even if they are on separate tablesIf someone has a valid implementation and docs feel free to request this to be. +We did test secure and setup with those changes +Two more failures on trunk for me +Ah stupid me the classes inherit from Iterator +I think I saw an access point for that +Closing fails to compile on +So I propose to close this ticket and eventually open more precise issues for phpcs integration for +If you need to set those then wed need to think about how to implement that dynamically and safely when a failover transport reconnect uses one of those cluster updated URIs open a new issue for that if its really needed +Need to test for other joinpoint types callersfieldscontructionconstructors also have the opportunity to make sure that per instance lightweight aspects do not get called for need to test PERCLASSJOINPOINT and methods with return types +Issue JimUse the to log errors and check if the output contains overlapped names +Moving all documentation issues to temporarilly +META +this patch contains three files is patch filesrctestresourcesldapSchema is new resource file for testing should place at jndisrctestresourcesldapSchema just call svn add to add file +I had a second thought + +This is not hardcoded +I know that the generation of the metamodel code is not a requirement of the spec but since the can provide an entitys meta data we should be processing itIn the mean time there are a couple of alternatives +Ive altered the logging level for situations to see everything of interest as an error. +The test suites finished successfully except the junit which fails me also on trunkJust for explanation of changes to embedded driverWhen I wrote a test I tried it to run also using the embedded driver and it fails with the XAERPROTO exception +Generic type parameter classes are replaced with type info class hierarchy employing a more OO approachAs a result of the refactoring I think the code is cleaner simplified +Updated patch to latest branch +potential committer has implemented this see EJB wiki for more was already working only wasn some standalone unit tests. +If A has scope provided shouldnt that mean that A and everything that comes with it is provided including B? Its how I would expect it to behave +It would also be interesting to see if the CSS resources to the body solves the problem +An untype String array that accepts a? Once stuff like this goes in that we havent thought about very much has to be supported forever +ill see if i can fix this resolved in +A connection got dropped when the connector was attempting to authenticate +Applied to SVN trunk +the patch of i leave this have never discussed the two of us on the user list about i have never participated on any user list about proposed only a patch for what was seem like a bug for YOU DONT HAVE TO be so unpleasant like proposed a patch and i wanted to be sure you have understand my you think it is not corresponding to the principe of maven ok no you could explain it with peace and tolerance. +If you have test please post here + +I tried Germans patch but it doesnt apply to +Passes all tests locally +ive me afew minutes to verify that +i would also appreciate to see some examples using richfaces especially the backing beans needed to render the exadel live demo lacks of this +clean up op before we enter into optimization step +Added my complete small Rails app to reproduce the thefile pattern to match exactly one file save and restart the scan after it has completed +like +Studio is on Windows and DS is on centos +Closing. +Create a pig job with multiple mapred jobs +Closing issue fixed in released version. +reopen to set proper fix proper fix closingVersion Build id +All the implementations that currently use IPCRPC do not handle this exception gracefullyI also think we should create a jira to ensure this is tested in and does not break +Resolved by all issues +Here block length still would be in NN +ai file there add an SVG version add an HTML version for the latest if the stuff is there Ill just close this as duplicate +This is pushed now Ill close as nobody bleated in the last weeks +Ditto for hadoop profile for use by above script so it can find where to set profilesM srcmainjavaorgapachehadoophbaseutil Remove unwarranted user of slfj +Now granted I was an English Lit major in college so perhaps that comes into play and I take things for granted but to be is clearly future imperfect tense +Got it +This is specific to already got something working Ill commit tonight. +Hari I think we can do this with an optional interface that can be implemented by channels that allow themselves to be polled +How to reproduce this one? +Done along with a fix for +Justin can you elaborate on what issues you are seeing wrt +closing so we only have one issue open for it +It would be great if you could include here some numbers along with the code you used for your benchmark and a description of your setup +to commit +By setting to true the new version of the process was picked up and a new definition instance was created in suggest that the documentation be modified for this option to something along the lines of Set to true this will create new instances of the process definitions found in the confprocesses directory each time the server is started +If REGIONINFO in meta is null write out warning w key content probably daughter split columns? and just continue +Thanks for working hard on this +If we cant manage that though Ill add in some getters +Setting for fix on +and more than an hour later the issue still persistsIf Im doing the workaround mentioned above then everything works fine +Im working on this +Ive switched the cacheUnresolved flag to true by default for Spring only keeping the default value of false for the backported version in Spring +The output I printed was from the original repro posted by JasonIm tweaking some things on my junit test and will attach it soon +Implementation idea add a custom resin property by creating a class +The second one is a hack that populates the dirty cl object by a quit command when leaving the loop so that it can at least close the connection on ctrld +Committed to the trunk +The only catch is that if you invoke the new API from a new client but the server is using the old version you would get a or smt +Please review the patch and tests +Its certainly a hell of a lot easier to remember than VWT or Web Toolkit! It will take me a little bit of time to rename everything but seems well worth the effort +addresses this test failure +As stated in the details of the bug the html is generated as expected but Dojo ignores it there isnt much we can do about it. +With ujax I meant only the type of pattern for special parameters I agree that this has nothing to do with jax in particular +Thanks man +Perfect Core was released. +Patch v makes extend catch +I simplified the endpoint so that it does the same thing but with a lot less code +This is especially bad for tests which are running on an actual cluster but from the source workspace +Thanks +Memory use changes as followsadded before gc MB after MBadded before gc MB after MBadded before gc MB after MBadded before gc MB after MBadded before gc MB after MBadded before gc MB after MBHope this info is of some use +Fixing Affects and Fix in Version +So we can use as soon as it will be released without patching +Added a condition to skip indexed properties +need to be careful exactly how this works as it will be easy to break things with respect to precedences of type coercing +Thanks Mark for pointing this out +Looks like everyones happy now +We might keep both in here for now just for the convenience sake +Any other ideas? Thanks +Attaching patch fixing the issue pointed out by RamyaPatch removes the condition which checks if the TT was already blacklisted for the same reason and ignore further processingAdded test case to verify the error messages were changed even if the reason is same +fop problem with seem to start with the second marker +Thanks for fixing it so fast +See my answer to your answer. +This issue seems to bug a lot of people PS Alan please write up new logs for your findings instead of reusing this one since it is really only about implementing Serializable +I think errors like the unregistering might be disconcerting +Verified. + changed description has the same bad seems this issue is already fixed in r +srcjavaorgapachecommonsconfigurationbeanutilsimport +It seems like from discussion and inaction weve decided to keep this in the acknowledge the incompatiblity between and series and keep this in series allowing for compatibility between and +I have applied the patch to master and +Adds tests and makes some improvements to the readability of the relevant codejavadocs +Fixed the the issue +Fixed in close of all resolved issues prior to release. +Removed the earlier patch because I forgot to set the svn native on the new class +The AMQ team indicate the issue was an invalid usage of transactions +Im attaching my from which shows what is going on during startup +It can take qmail a week or so to figure out whereto send the bounce messages +SO I deleted Cordova file and added Cordova file +Added documentation for nested definitions +You most likely built the war in JDK but are running your server in JDK +patch for struct returned by thrift in describering already has both a list of endpoints and a list of rpcendpoints +Other parsers like do not preserve colors too +Hi Mike your point is true +Work on this new version has already started +closing it +Please change the path that points to those two avpr file in to run it +But that was not a standard way +Review board has been updatedNote this patch includes binary content so once we get a I will have to commit it +I am new to caster and i am facing this problem Can anybody please tell me how to fix this issue? I need to steps to install this fix +Committed to trunk and x +But as we are still at the beginning it may not be good to modify anything on the site +Thanks Bikas and Sid for all of the help on this +Assign to during tidy up prior to the issue now that it has been resolved. +Yes +Probably long obsolete +I think there may also be some overlap between this test and the existing test +IMHO no matter what the reason is its not good idea to add these kind of parameter automatically +Created like any other private list moderated subscription privately archived +Adding it in as Rohini suggested makes sense +Changes look good to meAshutosh and Namit let me know if you want more time to look at this + +Thanks Sid +In this project an ajax navegation incrisis the memory but when the page is reloaded the memory desappear +Regarding skip performance I would say its hard to tellThe code in the JVM might be optimized and perform slightly better even if it is using the same approachHowever in some cases the stream implementation might be capable of skipping an amount of data a lot faster than reading it +If it affects more than one EAP major release please make sure it is cloned so that there is one JIRA per major release +bq +Have you found a solution yet? My case is this when test standalone no exception thrown in web container it threw exception during update but not persist +actually in the patch used the callbackWorker +Im resolving this issue +Is it the artists intention for the upside down logo to be read as the word doopey +When this plugin will hosted be by codehausThanks +Verified on JBDS +Patch generated using SVN diff against the codebase +Well I think we have to choose sides here +Patch applied. +How it is used and it if requires configuration repacking etc +For the unload thats weird + merge with branch +screenshot taken from chinese version of windows after bug was caused by something upstream where the encoding was not specified for the readingwriting of the filesThis patch has been committed to the branch +Attaching a new simpler version of this patchTouches the following filesA javaengineorgapachederbyiapisqlcompileInterface for node Visitors +Its true that if we estimate N correctly then we know that eventually jobs will get to launch tasks before the next preemption interval but even that is not as good a guarantee as saying that as soon as your timeout passes we will kill some tasks and give those slots directly to youI think the scheme I proposed above is the simplest way to achieve this without requiring any sort of estimation of heuristics + to test hudson patch t this be fixed in the branch too +Lars contributed wiki documentation as wellThanks a lot Lars +It doesnt fix anything but adds some additional logging which might be helpful in resolving this +Yea same for +Ill have to build an assertion in to it later +The difference between highest generation stamp seen from the shared edit log and the highest generation stamp seen from any DNI couldnt think of any legitimate use for this +Please verify whether the problem is resolved as you expected +Coprocessors in architecture and implementation are exactly like loadable kernel modules +should be committed after +Seems unnecessary +Please try latest CVS Nightly and let me know if you still see the problem and we can reopen the bugthanksdims +My recommendation is to post this information to the solr users list and see what they make of it +This is one of the major reasons of slowness of printf in X. +Patch includes a test case called which performs some tests on this annotation and on this new feature +Please add locking around the isDone method in the context and updating the done flag +Net clients which were removed from the repository months ago and not the current +Topher +Here are the improvements in the code and functionality for the cost centerWhen we run the functionality and add the amount percentage for different cost centers the problem is that it is working on Ajax Request for each row submitNeed to change this approach as screen flicker on showing and hiding of each amount percentage changeSteps to get right functionality Change the event handler first from jsonservice to +I will move to and start cutting THPF over to it +Actually Ive submitted a pull request that makes that proposed right fix +Closed for release +Hi JiriSo what youd need to do is Define and maintain your model separately from the project youre testing in +My environment seems to render the site a little different +Patch to allocate new Storage as +This patch added more to our test cases which is good +Seems I already had a TODO there to check if Kerberos OID was correct +Another solution to avoid probable differences would be to create the hibernate query that would retrieve all the task instances and allow the super manager as stated before to view perform action on note that this query might be an over kill when the existing tasks grow large and there has to be a mechanism to allow ordering grouping and sorting those tasks to reduce the processing web app blurs the authorization a bit +Thanks Ivan the patch was applied to BTI branch at rPlease check that the patch was applied as you expected +The logic is as followsa If this is the rd or greater recovery attempt round robin around the data nodesb If this is one of st two attempts round robin around the data nodes and pick the one with the minimum heart beat +On Oracle g when deploying an application there is a part of the deployment where you can configure the deployment plan +Though your sample is just a bit broken on nix filesystems Ill fix it up and push upstream. +and failures seem to be unrelated and both pass for me locally +image used for green side of progress bar +We need to ask all current hudson account owners to sign up for infra +patch was tested on winlinux ia and linux could you provide a patch without formatting? Otherwise its hard to understand from diffs what was changed +We should have something in the model that allows us to do this +This jar file was a step release during the JDK era which became standardized for JDK +It can be changed +I use junit +It seems like the best solution here is to allow for the salesforce datasource to allow for specifying its own bus configuration just like the ws logic +it looks fine to fine +I did not quite grasp what you were trying to achieve with your patch +The patch gives as is applied to Windows not On Windows Lets get GroovyVersion JVM exit to terminate the shellType help for command helpType go to execute the statementsgroovy println ABCDbEFGgroovy go on Linux ets get GroovyVersion JVM exit to terminate the shellType help for command helpType go to execute the statementsgroovy println ABCDbEFGgroovy goABCD on some linux but Not work on other linux. +Could you provide stacktrace for the error +Chris Ive added a sidebar in the reference documentation that touches on the differences between the two +expected actual rule is not about having a right curly brace alone on a line but the first character on a line +Very niceIts too bad there arent hdfs tests since it wouldnt be hard to do +Ok I didnt know about the property in Why are you unsure to modify those from the Eclipse plugin? Is it because of actual modifying or because of the solution wont hold for projects +s behavior matches that of MRI. +A few tweaks to get the configurable by +A test case for the changed functionality is also included +The configuration has been changed to support the various attributes of the vm starter command and the starting of the vmconntroller now uses those attributes. +This is very strange +Paulex thank you very much for the fixIt works fine confirmed +Despite the fact I remember seeing at least one other person on tapestry users mailing that had experienced the same exact behavior I now believe this bug report is invalid +Antony why this issue depends on +As it will not break anything else +Attachment has been added with description I didnt realize that I could attach the source image file +Ah whatever +The problem is only half solved +Thanks for the examples Benjamin +If its closely tied to Proguard that project might agree to host the plugin +What about call refresh action when or after saving edited gadget? This will not occur so often and will not have significant impact on performance +JSDT related issues listed in thanks +Closing as out of date. +This ticket confuses me +I have only checked this on windowsNormalc RSCO Tx Dr Clerk Normalc Dec CSTNormal NormalDESC PRICE QTYNormalNormalConfigurable PC T Normal RAM MB Normal HD GB Normal HD GB Normal Normal SALES TAX Normal TOTAL Normal Normal Cash Normal Normal CHANGE Normal Normal Barcode NormalcAC +Hi ChibaHave you been able to reproduce this bug with the compiled classes I providedSince our project cannot continue using javassist as long as this bug remains Im wondering if you have any idea whether or not this can be fixed in the short term +In addition to Todds comment contains the change for which you address in +Set two values in the MDC context the former would be a formatted string that contains all of the information above and the latter would only include primary principalsI am currently leaning towards the first as this would give users the greatest flexibility +One possible solution may be to provide binary distributions for Xalan that do not include XSLTC +the other hand uri validation is overkill and does not really helpinstead just try to unmarshal xml into a bean return true on success false on parser exception +Right sounds think this could work +I agree it makes sense to merge this patch into and since all of the other changes are compatible with respect to the public API I think it further makes sense to merge all three +This fixes things for me +Hi GlennThanks for volunteering to help testing a fixI found that this error is not seen on Derby and earlier so Im marking it as a regression +Your step might have removed file that existed on your setup +for example port +Applied the updates in r +Can improve on it in later patches as get more info +Thanks Vikram +Does it materialize the clob in the union all case or is there something else going on? I am not quite sure where in the code to look but am looking +Hello Vijay attached the file as requestedThanks appears I mangled the patch file while making a simple change sorry +Nice patch Ram +we in Scarab are still using a patched version of so we can provide the feature and might make a new patch if neededThe history of the issue and the comments just say but the Wont Fix status is not emiting good vibrations. +So I suppose my patch isnt really needed +set RN status to not required due to lack of feedback. +We have two clients hitting the same cache of data +However looping through dates until a match is found seems kind of clunky +can make the statically accessible if they wish and use to acquire the in threadsThis means only a single web app can run in the container however +If you know of a json library that is known to work Id be happy to fix the erlang json protocol code +have you created the issues you mention for the criteria to resolve this? please link them to this +The testcase already works in both and +mfojtik is going to send a pull request to fix this. +Since Im only retrieving n out of the set at one time anyway I should be able to iteratively call this request to see all matches +when i remove the from +been there done that +So to do this we should add an export to the cassandra users bashrc file +thanks for changing the method name +Might as well not half ss it +Out of curiosity why do you need multiple src treesI cant really think of the usecase why you would HAVE to have this +Marius are you able to take a look at this +Im using Eclipse versionThanks and Regards this is not the same issue as +So i think if this Apache book store doesnt exist anymore we can close itCheers +Linking to which is for logging operating system solution of logging this information directly in without going through the sysinfo tools code and without any special properties seems preferable +Hence this issue could be closed +I see various examples of other services doing this such as Should I create a separate subversion project as you did for? Or should this live in an existing project +The Filter after a request completes should delete any files in a known temp directory +Can you please provide a simple test case which can reproduce the error? Thanks +Yeah with above I mean trunk +An updated patch that ignores short names +Each server had logs to proces +CopyPaste by mouse clipboard in RTE is another ticketThis ticket is related only to copypaste in Link dialog not in RTE and I guess its fixed nowSo we can close this ticketDan could you please test it? Thanks +Here are my two cents if i may chirp in +Since this JIRA has diverged from current sqoop development might I suggest closing it? We can always reopen or file a new one if we will need to address the interactive command selection again +wars classpath +Added the code for generation of asked constructors in service locator file +Roman would be possible that the created from GH issues have the GH number at the beginning of its subjectThxs +Yes you can close this as a of. +on the patch +r fixes this +We are exhibiting OOM in other areas of our test effort +Changed classes based on version code +Any workaround or possible solution +Its been annotated although commented out as far back as +The other option is to break only if a dependency has no revision and only warn when we cant locate the parent +Im not a sure but I think I had the same problem with Apache Derby and a Date as the type for ordering +the output files just need to be linked somehow from the webapp they reside currently in the appscontinuum ive tried adding a symbolic link from the webapp dir but that dint work somehow +fix to with revision +Checked in code fixes this issue. +It wont play a difference but well spotted +Im not sure theres a compelling use case for translated log levels +There is a cleanup task which cleans up this Map so this Map will not grow infinitely +So then wed have both options using a JMX console and by sending a message to a management queuetopc +This will not affect the new shim layer which is now inOne other thing you may need to be aware of together with package rename the maven coordinates of published hcatalog artifacts will also change in the same way +Ok there you go +Heres the initial patch +Thats it +Matthias I have added you as a watcher to this new issue as I had to address the same error you saw in +However I am going to leave this issue open and reported though because I think it makes sense and were going to need to deal with it +However it is not applicable in my real case which consists of a custom builder custom binder and running an external script using +Cant reproduce this install fuse ESB and create camel su archetype with +Great! Hope it comes out soon +also note that the license of the module shall include the licenses of the libraries if they differ +I use the example of to reproduce copy the project and add a sec delay after the Exception in the client class I replace the remoting with I can start the first exec and after I see the Exception at server side the client sleep in the I start the second client this call produce the same for me it looks like working nevertheless whether the client is still alive +is it enough? +Ill look into seeing what I can find +When we have a packaged schema then we dont need to generate one nor do we need to add stuff to the packaged schema on the fly only applies to unconfigured DII clients +Please go ahead just correcting issue title to be a bit more specific +Attachment has been added with description Hyphens do not show up when preserve +Fix the cart before the horse by creating a discussion thread +Description of patch I judged that should not be thrown in closeReaderclose +Attached +The current patch is sufficient to catch compile errors before commit until the antlib has unit tests the only time someone would actually build it is if they have a particular use for it +I have experience using classifiers for my text mining course and also I have experience using recommendation algorithms available at Mahout +Id prefer to keep it in the same mojo and define the output types html andor xml +but here for refernce +controller automatic name +I am one of Chase Wolfingers and I am trying to apply this patch to move the bug along +Is there any way I can change the stylesheet that Maven uses so I can supress those icons for my project? +Feel free to move it if that makes more sense +bulk close of resolved issues whose fix target is either unspecified or from a prior release. +Patch applied at r please verify + +Fixed in subversion please testIf you have some time for a contribution please submit a with this at least as +Perhaps we should update to not prefix commands with a dash? Whitespace in is to make the if clause consistent with the following one if you dont mind Ill keep it FC uses failover recursively for failback definitely better! Added rebased on the tip of branch post +Sorry i just search how to help +See Integration of lesscss in Tobago +A comparator per table will not be happening anytime soon +Did forrest clean between each theme switchUsing pelt skin +Do you have log snippets that demonstrate the fix working? Thanks Gao +to be specific for key iteration start withfor the hashing look at first it sends a digest message to the nodes this computes a hash and if there is a mismatch then it does a repair step in the catch block +When a contribution like this gets made we need the original author to donate the code giving up copyright +fixed by adding an option to not detach the event handler function calling +Verified fixed +Yeah this would need an update in elementree sax npm re moving in the direction that the platforms is a build artifact and not to be edited by hand +Im sorry I copied it from the wrong workspace +Attaching patch with the fix +Im not sure why Solr is trying to send such large amounts of data to the shards thoughSpecifying ids to be retrieved I imagine +Has anyone tried integrating it into Hibernate sources? Any tests failing +On a clean shutdown each region is closed by theThe steps of closing a region are +Since we already have a util package with a I think it could be put there too and referenced from +If this is still TBD why do we have? Should this be closed as a duplicate? +I appended to the seed and it failed twiceSo i think maybe the test is becoming more going back to +Let me try to get it fixed +We should also bring along our file as well now that we have conf +The test is expecting an error in response to an ack but gets a message instead +This is when only one language is specified the languages dialog does not appear and the given language does not get propagated and set as the default language +There is example code on how to do this in whiteboardhenningjdk and an actual implementation in whiteboardhenningjdksrcjavaorgapachevelocityutilintrospection all resolved issues for Engine release. +Your explanationof how this change fixes the problem seems good to me and your testing resultscertainly show that the patch is effective in removing the problemIm not very familiar with this code so these comments are mostly for my own education +So it can be a race condition +If we find a conflict we could fail the upgrade with a message advising the user that the conflict needs to be fixed firstThanksRick +Thank you! +Sounds good to me +Forgot to grant license so attaching the same files again. +Peter does this bug affect too +Yep good catch +Completed At revision Updated jar target to add in requried files +If at some point a different serialization protocol is used that supports enums the wrapper class can be disposed ofThis patch Creates enums for and that were previously named constants in Provides for conversion of these enums tofrom integer values for sending over the wire Creates a class +Tapernate removal is work in new will continue as part of work +I think load balancer should be temporarily disabled at the beginning of instant schema change +bulk move of all unresolved issues from to + for me but I think it would be a good idea to have someone else lik Pat having a look at this +A will always override it anyway if it supports a custom factory +we could divide up the rest and do at our leisure +If you think this is an STS feature let me know but IMHO it would compromise the idea that Roo does not depend on a specific regardsm happy to advise that STS offers this functionality +Am I missing something +passes +Dup of verified. +fixed in +Nice patch Uwe +Problem is erratic and still appears +This issue was fixed in and works fine on Linux based systems +Yes the first example is similar to my problem +looks good to me +Might this be more simply solved by just adding UPDOWNLEFTRIGHT support to and having it look to its parent container for other components to focus +Is this complete +New PatchAdds more testsFixes a couple bugs that prevented correct resultsFixes the debug path for the single node mltResults are not currently sorted the same way as they are on a single nodeI dont really have a need or use for this so if anyone that does could help with testing that would be great +But as far as i can tell from reading your commit mentioned above your fix looks exactly the same what i did while analyzing this issueRegardsStefan +Nice catch Hitesh +WillemThanks your help with this issue +The version throws exceptions for modification methods + +Im hoping someone can review this patch +This might be a dup of +Reattaching the same patch with license granted with the name change for the friendlier +This is done +Fixed with r +Its a new feature so for just committed this. +for the patch +bq +It seems that I forgot to fix this issue as I said I would so it will ship with +Assigning to Pranav who is going to look into this +Yes the callers of mkdirs will need to check the return value +Note widgets are still disabled +Done +Fixed in +Created a patch to add the jar goal to the package my cents but if these test base classes are going to be consumed outside of the samples module they should be in a module which isnt named samples +And the efforts to make it more generic are awesome +Moving out of. +out log to get Class not found exception on startup +svn commit m Add test scope to junit and ant dependencies api Sending api Sending Transmitting file data +yes I attached my update NewSyb however I would like to delete the old one if it is possible +eclipsekeyring arguments os macosx ws cocoa arch x product keyring Usersburr +dims +Updated to use and included the changes. +There is also a licenses directory but it is not in I see the issue more clearly now that it has officially bit me. +Anyone care to code review +If local build is downloading dependencies from nightly composite it doesnt expect they going to be moved to different location +The exception is thrown only while shrinking the data volume +lib x local +Help welcome. +I dig it +That is consistent with auto create behavior throughout Seam anyway +And the above does work without quotes +Yes DROP TRIGGER is good though +So I should come back to this issue to test it again +bq +If the user applies a manual workaround to visit these nodes and then the visitor automatically visits the nodes as well then the node might be visited twice +the patch to both trunk and branch +Hi Howard isnt this JIRA related to the failures we are seeing now? The ones we talked about this morning +fixed in svn +A first try on this one +But nothing ever resets the iterator so thats only meaninfgul on the first recovery pass +zip and +I have applied it in Rev +If the Delete and next Increment occured the same millionsecond the Increment will not take effect for followed Increment because the Delete will lay over the Increment +This patch should fix it +Reported as resolved by Chris Custine after applying +verified on windows with is still present on with +How about this one? has been integrated in the web site but not yet in the main site. +Similarly for the +When RM comes back it scans the store and deems every app remain in the store as not completed and resubmit the job +The first use case can be done with existing facelets templating facilities pretty easily +It would be better if the syntax can be modified to something like fieldfieldfieldwe should be moving away from delimited param lists not adding more special syntax like comma seperated lists of fields make it harder for people to use special characters in their field names +I am also facing the same issue +Any revisiting of the locking patterns should be done under a different issue. +Their gauges and some of their Google Maps integrations are very powerful lookingHave you tried the walk through onDo you see that much benefit in designing it anew from inside? To utilize an OLAP natively inside would I think require quite an extension to the entity engine +Mattre documentation agreed and thanks for the reminder I just patcheddocuserGuide with more information +Im not sure is idealWe need something to provide a basic test harness functionality such startstop testsuite and such +Committed with revision +If we havent received a response in seconds we can assume we have failed +This can happen if people use super a lot in modules +I will check if that leak happens without it +This is a test EAR application containing the WAR fileAll is packaged as maven applications so that two maven builds are required +This is a recommended practice anyway so I dont see the harm in making it officialIf the group really feels we need to protect Apache releases from releases then we have to decide what to do about +I missed the REST client project when making the changes +If anyone knows how to do it and thinks it would be worthwhile please say so +I think it will be more effective to use something like ORC Parquet rcfile etc +Patch updated +This bug was original brought to my attention as an issue with the entire mapper taking more than the timeout time +Added generic support for wildcards in addition to things like foo also work now +updated. +Sounds good +Extra test case patch now committed. +this is the patch for and +Can you try again our logs are rolling too quickly to spot the problemIve increased the log size and hopefully will be able to catch it next timeCan you CC the email to supportThanks Ben +Thank youfor the comment reviews and the test run KnutYou are correctI just had to copy and paste the case from the case where the locator was not being used +Jason it would be better to hash out ideasfuture directions on the wiki and than to create stubs in needs to refer to an issue to be relevant over time especially ones that take time. +Additional Description of Testing derbyall passed +Since the patch becomes outdated already I am attaching a new one +Ill commit the patch shortly +Im going to hold off on committing this until I hear back from Arya + schedule for +matcher code will be added to the sandbox first +Thus blessed by the power of lazy consensus committed. +Note this now works in trunk for +When region is assigned to the same server this seqnum can changePlease comment +Thanks Stack +Resolving as duplicate of +Closing resolved issues for already released versions +date fieldName ts notNull class +The current patch is still not very clean +Remaining subtask is optional +What we currently have set up with Rob is working fine but its still not pushed at Eclipse +Okay thinking about this issue more in response to a GET Request I believe that JSF typically executes a createView instead of a restore view +miss license headerCurrently is used by andI wonder if a config parameter +call start +After further investigation I believe the behaviour you are describing is the correct behaviourThe element causes a search criteria to be used that searches all network interfaces to identify the interface with that name bound as you dont have a network interface with bound no network interface is found so there is no interface for services to bind toThe when searching only needs to detect that an interface is a loopback interface to be able to select it to bind to this is why the works as it no longer needs to search by exact address +yes put in examples for proof of concept +If I try to ping the server once the hang occurs my client hangs as though the is accepting connections but not designating a thread to pick them up +This is duplicated by +I started to make a response and then thought back and forth on this should probably be out on the dev list? Want to start a thread? +Im not really sure what needs doing here +I should of compared your code with mine +Which applicationJRE version? +Im going to change it to see if I can call nice from the java containerexecutor code as it builds up the command rather then inside the code +Attachment has been added with description example supporting the claim + for the +This issue still exists after I applied the patch in +FYI Im also attaching the file properly encoded in +the problem with this is that application threadlocal might not be set on the thread deserializing the session this is a behavior so it might not always workinteresting little problem +Hi Renat yes I agree +Using it and using it correctly are two different things +DerekI have already commited the changes +encode it back to UTF and copy it into the output buffer and send it to the client +Already did that but still get an error +Thanks for confirming +Right only x family is supported +Im going to revert it +Platform includes wizzards for especialized industries +This is still not fully functional as mentioned with comment +OK well I got my environment to build and the attached patch works wonderfully for me! Hope to see this one get into a build soon +BJIm not having any problems with the Username and Password showing the current login are you sure its not your browser putting them in as part of a saved login feature +Thanks Robert I committed first part +Sorry I did not respond sooner +This is a browser issue use the reported workaround. +Manually tested +Code changes fast around here +Changes committed to and logging to use the Problem Monitor when logging the exceptions +I have run an inhouse test that uses bulkinsert and vti to populate data and the test ran fine +If deadlocks and cache wasnt be interested for months so this one should be opened +Maybe I cant tell for sure until we find the code that has the issue but weve been running for months on multiple clusters and never had this issue whereas this one cluster has it +Took another look and I was confusing the code to do all C static initialization for with the static initializaiton of a single closure itable The code happens to start with a closure itable initialization but it actually does static initialization for a lot of data structures +I went with classname because that was a better match with the other AS subsystems +Issue closed with release. +Thanks George. +I dont think different can calculate different values +Filter array array for created with cvs diff for created with cvs diff of diffs +Noticed that the document still does not capture f option though the copy command does +Committed thanks +New supposition is that an updating client thread got a region before it was moved to retiring queue inside of batchUpdate +Why is this even logged here +Her is the code for can see as well that if you switch the root logging from DEBUG to INFO you dont have a memory leak any more so it should be a dispose method somewhere which is missing +BQ +Compatible with Mantis +Presumably we do this during the normal read path for block checksums but full file hashes are too heavyweight for that +Tomcat is therefore started without any messages logged in Windows Event Log or Tomcat log files +it would be good to have a command on the shell to install the daemon script you start it interactively once and then xou can make it a service +This will help me building the fixregards Malcolm and handler changes checked into SVN + We need to remove the Create Salesforce Function Model action from Designer it is no longer applicable since we include the for salesforce in the model +Hadrian can you get this fixed so we are starting to be ready to cut Camel +But this is a total duplicate of +pushed +Wed also need to add to the list +For example would we need a rtl rawURI parameter as well since we have both uri and rtluri parameters. +as it doesnt support weblogic standalone configuration +All the previous comments before refresh is present and I can delete the file after application is refresh +Can you please merge the changes and the patch +The data for the engine is placed into the templating environment as Java objects +Id be tempted to assert +That means it came from me +Otherwise what are you serializing + +Check with Stefano for the m waiting for Stefano PR but perhaps I misunderstood your comment and you mean that it will be Stefano that will review my PR + change only for LOCALONE +Currently namenode will list dead nodes +Thank you for your patch Arvind +We have a single workflow that gets startet hundred times a day +need to retest this because of H Could you please retest with the next nightly if this works for you now? retest ASAP +Upload a new patch to accommodate the newest trunk change +I guess that the same filures also happen without your patch +I havent had a chance to investigate it and try it +Please clarify what you mean by splash screen? Are you talking about adding an image with the application name that will appear for a few seconds during the install process +Resolved as fixed in revision +If there are default Args named Args start their positions after the defaults +Changed logging of error to a message dialog in the action +Should it be made less critical or renamed to move exception classes to common or closed due to lack of activity and people can file specific if they want to move classes +Open new issue if the new load balancer committed there is unsatisfactory. +see attached patches +Can you please attach a simple two entity Roo script +Would be nice to make a list here and make sure this gets applied everywhere +Have you tried using a custom type handler for this +In any case weve seen similar problems in the past when a second copy of derby classes is loaded this definitely is a problem with the embedded driver and possibly with the network driver +I still have one views content that is oddly smaller and the overall text size in the tabs is smaller so I guess it is expected behavior now to have to adjust these for the new dpi setting +If the client would reconnect he receives a exception with subscription allready in use +integration tests added tobranch merged to masterWido will look at adding RBD resize support if he gets the libvirt bindings and has time but its not a blocker for this feature +Will show next time we roll the website +As you recommended use to let Camel manage the thread pools +Thanks for your attention. +And naturally I cant reproduce +It completely messes up a lot of legal characters in the path such as which osx likes to put in temp file pathsnames +Thanks Allan for the contribution +Thanks Gert! +Closing the issue now that it has been resolved. +Closing resolved issues for already released versions +This would be a more obvious clue for the user that something went wrong and what it was +I dont think we should do bookie recovery on an open ledger +Should this get committed +I believe that I understand why this happens when the object is validated in Validator class in effect the is called and throws exception but currently the which is suppose to add the id to the test context is never invoked +Simple test seems like gc bug so I changed the subject +I think it is a bug and the fix is correct +HelloIve updated Geronimo unit tests automation scripts according to latest BTI requirementsTested on Windows ia and Linux Nina Aleksei the patch was applied to BTI branch at r +Hemanth That still doesnt provide any information about how to reproduce the issue. +Why shouldnt the objects version be used? And why should it be incremented when there was in fact no change? Replicating an object without modifying it is IMHO no reason for incrementing a version +Also their are largely using the normal compiler +all bulk edited issues have in a after release. +Is the new interactive demo on showcase good enough to close this ticket +Fixed in r +In the shell script andor cmd files in the bin directory change JAVAHOME as appropriate for your environment +Dan whats the status of this +This really isnt dependent on the new api +But we should definitely have a closer lookThanks Peter +Manually tested +Carl did have a good point this patch did not fix the case where a class is removedIn my newly uploaded patch when the test class path is constructed it adds the jars from builddistlib before the Ivy jars and excludes all hive jars from IvySo now for the cases Carl mentioned this patch fixes them for ant test +SVN and laterNew current trunk if the above mentioned patch is appliedInput msm mxmOld output msm mxmCurrent output msm mxm ms mxNew output msm mxmInput Xmsm XmxmOld output Xmsm Xmxm ms mxCurrent output Xmsm Xmxm ms mxNew output Xmsm XmxmInput msmOld output msm mxCurrent output msm ms mxNew output msmInput mxmOld output mxm msCurrent output mxm ms mxNew output mxmInput XmsmOld output Xmsm ms mxCurrent output Xmsm ms mxNew output XmsmInput XmxmOld output Xmxm ms mxCurrent output Xmxm ms mxNew output XmxmInput msm Xmsm mxm XmxmOld output msm Xmsm mxm XmxmCurrent output msm Xmsm mxm XmxmNew output msm Xmsm mxm XmxmInput null or Old output Current output New output +It will only help of the issue +will be applied in next out into production +Which version of IBM JDK this is for? IBM JDK SR? Is the patch needed for IBM JDK You should still use instead of introducing a new one in +Heres a patch which addresses the issueNo tests are included since security has to be enabled to test this code path +The fix version was set to should be? Also it was left unresolved +Note that widgets become DISABLED Stop server +I did not understand the problem at first +Does that address your concern Yonik +I was just curious about what works and what does not +The rebranded pdfs have been checked into the soa repo +I overrode both of them to which looked like it fixed things locallyCan you attach another screenshot +For sure the hadoop classpath is not being picked up by hbase when its started +Once you seek to the offset the data transfer time is insignificantPlus given the KB page size used by Linux all IO is already at this granularitybq +Its using the XML stuff from the JDK itselfSo they are already splittedIts only that is in the component which could have had the name as a more friendly nameBut renaming a component is not something that is nice to do +has no such thing as far as I know and thus the only way I can change the context root is by changing the deploy name +I think everything is gone crosses fingers +This does not support searching across index types so all queries must accept index expressions of the same index type +provides all the functionality that provides but using the text interface instead of the html gui +. +Verified on. +It feels a bit like using String as my superclass for a POJO instead of Object +Can you please put the patch on review board as well +But we know some are using other strategies coupled with ant +Pending sync to see if it worked +comparison of fragments +Updated mojo to manifest directory if it does not already exist +Closing resolved issues. +Thats coolbq +Per discussion with Zheng added a new shim method to deal with the incompabitibility +pending Hudson +Here is all that has to be done +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +It may degrade respondto? performance slightly but still leave it x faster than C Ruby versions +The Browser is Hardware Accelerated +Hi fabriceNot really sure this is the re using tomcat for our integrations tests on sonar and never had such you using the same version of tomcat +Hi Rishi FYI I was able to duplicate the problem +upgraded to apache common daemon procrun +checked in +Bumped to beta since beta is released but we need to do this before letting many people in to we have a backup in place which writes a pgdump to a private s bucket every minutes and keeps the last backups +Granted there is a danger of duplicates from other plugin +Im not sure its the better solution though +Patch applied with small changes in both sources for cocoon andsrcblocksscratchpadsamplestransformssrcwebappstylesheetssystemAlso js code was moved to a separated js file srcwebappresourcesscripts css code was moved to a separated js file srcwebappresourcesstylesPlease cross check and close the bug +Is there a github link you can provide so we can see your codeCould you also give us some feedback about what you didnt like and why that caused you to write your own client +Edell + +IMV by making the pluggable we can address these issues +DanielAny chance to get that support in the next CXF releaseThanks +Many thanks to Evanthika for the review +Probably we could do a separate patch here and resolve this first +Updated the EXPORT topics with a paragraph at the top indiciating that the FILENAMEneeds to be a new file +I can reproduce it +Hi Julien yes this was changed and then reverted as per your comments and that the tests failed +Im afraid this patch needs a rebase +Changed post hook name as Ted suggested and passing mini to itThanks Ted and Anoop +Same comment as above +Certainly the attached project works +Brazilian Portuguese Translation for Human Portuguese Translation for Human Lus AlfredoYour patch is in trunk at r +I just updated the file +has been marked as a duplicate of this bug +So the JS file with the wrong licenseshouldnt be there +So the are normal? +I cleared + +Other thing that Im not sure is How to decide on the version of components that doesnt have the version as part of their jar nameSticking the checksums into google works most reliably +Ill take a look at this but it may be tomorrow before I get to it +Patch with applied comments +im wondering if we should do something radical and flexible in this arearight now if we use IP address we can make it hard for clients to talk to hbase since if their knowledge of hbases IP and the one uses are different then bam +The test code was left in place because it might be of use to the HTTP client project in generalThis code submission is authorized by my employer with rights granted to the ASF +The only part left is order of the result +I thought that this was handled already by the activationprepared statement mechanism Ill try to find some time to look into what isnt happening that should +Sounds reasonable +I waited an hour and the instance remained in the state stopping +I guess it will give some headache to maintainers on OS when they will have to deal with xulrunner js +Pull request merged into master branch. +trouble that can be solved by this issue is the follow use case Client loads objects A A and Client loads object A Client changes and commit object A Client changes and commit A A and A in one transaction but get on A because of changes in step then rollback the transaction and only reload recommits the transaction and get a Timestamp mismatch for A and occurs because on processing of transaction Castor changes the timestamps of objects and in rollback dont retrieve the old values of timestamp +Justin comments on the pull request that perhaps it would be better to log the exception than to let it simply be thrown +I am going to leave this open to voting +Its in +Attaching patch has been applied thanks Richard one more patch point +Anders keep it simple This usage is exotic to me. +So for the need of our project we need to patch on and on to make it work + +Thanks +BTW In there was the status verified but I cant find anything similar in Jira. +Committed to too. +This is critical for us +Although I wonder if Hudson will be able to find the Facebook Hadoop jar? Would we need to add some repo information to the to tell Hudson where it can find this jar +archives as jmx move from M to want to keep this open or have we reached a point where the more specific issues can be used +I think this is expected +I did note that it is only the web content folder that has this have a feeling the problem doesnt start straight away +This is by design +This looks good so Im going to close it +On revision Ive successfully tested Windows Linux and +This Jira will drive upgrading Jackrabbit to the latest version not just from a clustering perspective but also from a maintenance standpoint for include bugfixes and enhancements +the guy said that the error disappeared after updating his Java installation to the latest available version +Brad how about that log analysisThe POST is being done in a new session every time while WSDL is retreived in the cached one +If team members availability were really that volatile then I would suspect that the project would have some major issues delivering anything anyway +It is not needed because the statement cache handles its own synchronization +The one after the error that only occurs once and the other during the infinite loop +If I can get one or two good nights in I should be ready +Assume were ok as is +was closed as part of a jira cleanup not wontfix +Update the patch change the type of snapshotDir from String to Path in to make it consistent with the createdelete API +Its probably been like this for ever but lets try to at least understand the issues for +If list size less then we dont show menu item for this listIf list size equals we show menu item but if user selects it we dont show next step we show CDI doesnt work as was agreed +Remove the old test +Add scan which optionally takes a Scan as input +It will probably be a little easier to implement but I dont feel strongly about it +Copying +As per the discussion in the nabble thread the automated build of snapshots should also build the source enables people wanting to work on one plugin the ability to debug and step into the source of dependencies without having to checkout every snapshot dependency +Henning or others any thoughts on this? Objections to committing +Dags comments in suggest that the statement that causes NPE is +Thanks. +Heres my patch for the accounting app is it possible to rename this issue to encompass the whole migrationRegards component with Apache Licence +Thanks Han though I wish the commit message wasnt resolve compilation error as this doesnt fix any compilation error +Thats fine too but is an implementation detail +Adding to for resolution one way or the other +Im curious and this may not be completely relevant but what is the structure of your image files? Are they all children of a single parent or is there more of a structure there +Ive created this jar out of the classes in deployclasses and put it under the lib directory so that both and share will need to produce such jar out of our build system for AS +if you can fix it in mins you dont need a holiday instead you deserve a holidayi added a test and it is checked in for +Thats in now off to add the logic. +as mentioned in the bugzilla your problem is a complex mix of jeetools and servertools +im going to commit these based on Chriss so we can make progress on the rest of but Im still interested in Juns feedback if any +I use the maven Artifactresolver to load the wsdl into the local repository so the updates are handled like they would be in a normal dependency +Changed fix version +Hope that helps get to the bottom of this a bit more + +Applied patches in r +The main files to review areThis patch depends on the following outstanding patches +Hi MaciejWere now getting emails alright which claim that it doesnt know how to invoke mvn +I was just trying to distinguish my comments from my logging +Instead what we can do is penalize hosts based on the last time they replied with a cap of the update interval +Im excited to see progress being made here but I have to agree with Joan the security of this concerns me even with the new youre putting together +So finally you tested the fix on or you are going to put in production? excuse me but which is the file with the correct patchOne more question After applying this fix would be any way for deleting old log files in order to reduce unused space? I am at G and I need to maintain the current state of the brokerthanks +ThanksIm sorry to keep the commit waiting but wanted to be sure about few things before we commit +just committed this +Ive been profiling to try and figure the slow down scanning in that reported up on list and its crazy how much cpu this key making is responsible for in the admittedly warped view the profiler gives you +Resolving this as duplicate as is already opened. +this really is a question and not an issue so Im leaning towards INVALD myself +Im sorry Ive no had time to dig into this one because it interest me a great deal +yes every line in one test run to check variations between runs +I am not sure what the value add here would beWith the lazy startup I have a bias towards eager loading at startup but maybe the compromise would be to eager load when in productionprofile mode and lazy load otherwise +Modifies to use batchmutate +Actually the bean that is produced by Arnes code sample is passivation capable +Can you try this patch? Thanks +Updated patch to latest changes in +Login as portal named prabhat Create portal button atClick on Make default link next to prabhat +trunk Sending trunkmodulesejbsrcmainjavaorgapachegeronimosamplesdaytraderdirectSending trunkmoduleswebsrcmainjavaorgapachegeronimosamplesdaytraderwebSending trunkmoduleswebsrcmainjavaorgapachegeronimosamplesdaytraderwebSending trunkmoduleswebsrcmainwebappAdding trunkmoduleswebsrcmainwebappdbscriptsAdding trunkmoduleswebsrcmainwebappdbscriptsdbAdding trunkmoduleswebsrcmainwebappdbscriptsdbAdding trunkmoduleswebsrcmainwebappdbscriptsderbyAdding trunkmoduleswebsrcmainwebappdbscriptsderbyAdding trunkmoduleswebsrcmainwebappdbscriptsoracleAdding trunkmoduleswebsrcmainwebappdbscriptsoracleTransmitting file dataCommitted revision branchesSending modulesejbsrcmainjavaorgapachegeronimosamplesdaytraderdirectSending moduleswebsrcmainjavaorgapachegeronimosamplesdaytraderwebSending moduleswebsrcmainjavaorgapachegeronimosamplesdaytraderwebSending moduleswebsrcmainwebappAdding moduleswebsrcmainwebappdbscriptsAdding moduleswebsrcmainwebappdbscriptsdbAdding moduleswebsrcmainwebappdbscriptsdbAdding moduleswebsrcmainwebappdbscriptsderbyAdding moduleswebsrcmainwebappdbscriptsderbyAdding moduleswebsrcmainwebappdbscriptsoracleAdding moduleswebsrcmainwebappdbscriptsoracleTransmitting file dataCommitted revision . +Attachment has been added with description example of the table was messed upprinter and viewer test IBM Infoprint IP OK either not valid or not supportedOCE printer OK OK IBM workbench OK OK BTB Afp Browser no image OK ISIS plugin no image no image IBM workbench ver BTB Afp Browser v ISIS Papyrus AFP Viewer Set allows using FORMAT X RGB but my printtest with a IBM Infoprint IP results in the same error ERROR The actual destiantion reported an OCA field code taht is either not valid or not supported +Andrey I still see the crash of from time to time +This problem should be fixed in the release installer +Since the merge of is imminent and this bug is not present in I think its best to target only for this patch +Bytes were not with xff causing for instance uc to come back as uffcCommitted to trunk with revision +The patch has been appliedPlease verify +After all what is the second query parser +What does it mean to set the version number of a? The only way to support this is by hacking the sequence which is what the older implementation does which classes are mostly now deprecated +The developer has their beans and impls and such in a particular place but they want the generated stuff to go elsewhere +As agreed on the dev mailing list some month ago in Karaf the commands nomenclature should bescopefor instancedev +Find the reason for issue not reproducible now reopening the issue +It might also be something worth porting to +The schema and doc has been current logic in will not read any transitions if the handler is found +I thought I had refreshed those but I guess not +Thats an interesting point +Hi Aidan can you review this change please +Reopening issues for to branch +this must have been remove earlier I dont see the method in the or the drl test file. +Committed to trunk revision + fixed the NPE +Heres a version that just contains the bugfix +Setting default visibility for Physical Models in to FALSE for newly added models + thanks +Two items First this bug is marked for trunk as well as security +Joernthe unit tests run fine +Ill take a look +I committed the change +wait +Assign to Tony +Hi NadyaYes the issue is resolved +Applied the patch with thanks to Preben after fix some minor CS errors. +does exactly that +Ill try running that +must have hit the wrong button while adding the resources locally +Please attach an example that reproduces the ZIP folder only contains those files which I have created and are needed for the test +I got the UI test working and made some progress on the IT test +Please take a look at and the code in branch +NamitKevin can you please confirm +not a b with +It sounds like this is an application server bug so Im going to close this for now +Looks like the same or closely related problems +Thanks KevinPatch applied to BEANS module at repo revision rPlease check it was applied as you expected +Attached is a ptch which resolves this issue on trunk +I think if possible what would be good would be to have the few changed files as the Subject rather than the commit message +Sorry previous patch had the leading path wrong +Nkeywal We need another patch on top of +David how can I use this feature +I can confirm this bug using Axis RCThe code generation fails with xmlbeans but not with adb +Remember that this is common code so we need to support all platforms and configurations +Thanks for looking at the patch code Knut +Hi guysI am confused what we want +The root build files logic remains simple just pick up what is in connectorscmisdistlib and use that +Sorry for delay because i tried several times and finally found it was an issue with my SVN GUI tool + +The code generating in the client is +Seems that Axis requires castor +cece master aad +Ive added extra debug that will pop up in the log if this happens again. +Attaching patch that implements +Log warning in if deployment archive and file extension mismatch for known types +Hi Folks Ive wriiten the Ingres db files for inclusion with jUDDI +Its important that we track down the root cause of this and fix it +It should be a simple to be fixed based on email feedback. +Thanks +Javac would choke on this +A bit of uneducated error only happens if I have instances of +The attached patch changes the test to shutdown the engine instead of the database with the teardown method before removing the direcstory +removetoken command could make efforts to redirect hints to new destination in case a hinted target is removed +bq +here is a patch that applies to the trunk +Found during collation is a standalone repro for this issue. +Hello NathanThe fix looks good thanks a lotBest regards. +Wheres the rest of the source +trunk r r +This patch needs reverting will wait another h before doing so please read comments on +Yes does sound reasonable +Thanks for the bug report Im assigning it a fix version of so we can investigate and make sure this issue gets fixed before the next releaseAt a glance it looks like the problem is the path CUsersnashrailsappc Im not sure whats generating this path but the multiple c in it are indicative of probably an issue +Thanks Pat! +I tested the proposed fix by subclassing the applying the patch there and configuring the patched version to be used in instead of the original one +batch transition to closed remaining resolved bugs +Instead of logging in as the KSSL principal we now always log in as the hdfs principal +I am confused +Patch fixing javac warning +Thanks for doing this cleanup Myrna! +I also see potential issues with the connection bound state +Otherwise the connection may stay open +I think it can not be done +How about moving the discussion of Whirr HMS etc to the incubator thread +Nothing else was requiredFor I forget if this is required by wickets logging or but it is neededThe Glassfish libraryJDK To run Wicket with Glassfish you do need to add and to optglassfishlib because of some issue Ive half forgotten with when the logging modules get loaded +How about? I think that makes it clearerAgree that we should move the commands under cacheadmin +Curious why is the resolution status invalid? If work gets committed to a instead of trunk does that mean the issue gets resolved this way? Perhaps it shouldnt be closed until it lands on trunk +Yonik Im not sure? Jans comment really hits home at why I opened this issueI think its worth considering optimize command to throw an exception in Solr for this reason +With this patch Pig runtime no longer passes an to through the bindTo method +Patch for fix needs to be ported to trunk +One of the good design principles in Velocity core has been to not to reserve keywords from the context +Otherwise lets track it through the original issue +Committed to trunk +Think you did this. +So we should be ok to commit thisWe can work on the maven issue separately as part of once this gets committed +As Owen suggested I will just put a log message if a spill happens on the Map side +In both interpreter and compiler we cache the blockBody instance to save allocation +I think the onus is on me to think this out a bit more and make a better effort to understand the internals of the codeThank youJames +Changed some calls in the installer and uninstaller to call instead so that the registered cleanup listeners will be invoked before the installer or uninstaller exits +Changed to decode all instead of only mentioned components work inside accordion now but they dont work inside collapsible panel +Can we also add the number of slots to the UI +Could you give me an exampleExample is pretty synthetic just if something wrong with and version number cannot be extracted from jar file +For instance chromedriver binary is by default referenced by system property +project +Lets discuss the implementation details on the just wanted to point out that your suggested change may break existing code +How public should such an API be? Should the interface be public? Should the implementations? The most conservative approach would be to make it all package private to give more leeway for evolving the update API +ThanksResolved in masterbfbdddebdcfcbbfebaeNow it should raise exception +Hello It is version now and it still does not work +It may be easier to identify issues which we may have ignored so far due to the high noise of the other issues +to three separate properties +Bumping to LATER +If such renderer exists in the project use it. +bulk defer to +Just got a clean build so Im hoping this fix actually worked +I made the UI changes as a temporary fix we should probably consider using Lifts functionality to make the change +Closing +Works fine on and CDH +So it will not try to pick the next element rightCorrect me if i understood your issue wronglyThanks +Fixed in trunk +Control Bus sample was added to the JMX Ill move it as a separate sample since it is not really JMX +Closed after release. +I think not because I dont see any formulas expressed +Was caused by having some code compiled by and some by +The performance difference actually looks fairly large on first glance though so it might be worth working around for a while if possible +reopening +is getting a bit large so maybe a new class would be better +Wire parameter +Mapping to GBK instead of GB according to Windows Control PanelRegional and Language options which is as same as RI do +Recall that for array elements used the type xsdstring for each element but uses the type soapencstring +I have no problem to make enhancements and changes for KarafBut as Karaf is a minor release on the branch it should be consistent with already released +for patch +In the default case it will take atleast min to send a notification +There has been enough testing to show that this is no longer an issue +Created and +to remove from webapp libraries just delete them +could someone add a note here when the bundlesspecs are released? MR trunk needs to be updated to use these before it can be that update we should be good to go JonathanI am working on releasing the bundles +FTR we do not support that because it would lead to inconsistent pagination and result size +Added information on message property to control sequence close along with discussion of JMX for runtime control and monitoring of RM to documentation. +Is it possible that this issue has been fixed somewhere else +We should remove setName +Vasily Your patch is in cvs please help verify and close the bug if it works thanksRgds Im verify cvs version +This patch adds a check to the to make sure that the retrieved Column is not deleted and fixes the bug in the by switching the logic on the isThe patch is for the branch +Trying just committed this +upgraded to RF +Thanks +There are a lot of places in Seam Core tests where we build projects and stop testing to wait for all builds are take no more then minutes on my computer +I have applied the patch then gone a little further and removed the impl method it updated. +Darrinn this should have been fixed for the very last CR build that we did and CR builds going forward +Thanks +NET have anything to make this easier? The can be updated progmatically but there are lot of classes to support that +Can you use a portablestandard compressor for this attachment? is not widely usedTry zip or +The actual value of the string is NONE and to ensure Acegi doesnt spend time on any file requested from the images path of a given request from the servlet context one could add the following line to the filter of the configuration imagesNONE +Steven I can see some other changes to +If this is not the case then yes this can be closed +I am not facing any when importing and building projects +Im attaching files that are modified from my first patch to to this issue as an svk diff and the entire files +If the table doesnt exist we should throw instead of returning falseI propose we reopen this jira widen its scope do the fix that throws the Exception and add some unit tests. +Still works fine for me on Safari Mac Chrome Mac Firefox Mac is it now tested and OK on IE +So I feel better both should be presentYeah we should fix the both ends +In a replica is added to corrupted even if it is not in the blocksMap +Apologies for the long delay in closing this +Done. +I am git clone the codeand checkout branch +Can I assume that you want test +fix findbug and release on looks good to me +Thanks +Not one with each interval +I think we should check for the exception and log a warning that support is unavailable +My patch brings the hunk errorI think the reason I copied the text of the patch in my linux terminal and pasted the text to notepad +This member must be declared as transient because Streams cannot be serialized +Committed thx. +Can pull it back in when gets resolved. +No need for this work +When you finish the quickstarts in the jdf repository you can assign to me again so Ill update the site +Apologies for the first few patchesDont apologize +BULK EDIT These issues are open to be picked up +The is runtime scope but it seems the WTP integration only publishes it if it is compile is included in +There is no update on these items and are in open state if they are not ready for please move them out to Future +Closing all resolved tickets from or older +Sorry to take away some fruit +As per discussion with Ivan here just deleting the inprogress znode may not be safe in the case of an Exception +I couldnt figure a way to remove them using just the parser +moved what looks like changeslargish feature requests to issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +I reorganized the structure of the source directories so that I could use the antrun plugin approach +What is to be done to finish this up for? Ive marked it critical so gets attention +defer all issues to +closed all issues which got released with +committed this to in Tools B +Sherry you need to verify this on a real cluster to be safe before we commit this +However this field has access +Examples +We are only considering JON discovery based on Teiid as a datasource for now +the issue in is one of the errors I encountered during testing the above cases +Can you post the class that is causing the problem so I can take a look at the generics being used? again I have created test to reproduce theI added testSPR to and everything seems fine +patches for trunk and it +Do you have a configured in your Spring MVC configuration? If so the will detect and process multipart requests possibly interfering with +Thanks for pointing this out will get it +Releasing this as a duplicate of the better here +The reason is that the generation of the manifest for the messes the bundle version because the maven plugin transforms it to instead ofThis needs a new RC of jetty +instead of parsing the path for each class in a jar it could cache tokens of the jars pathBut I havent looked into VFS internals so just my p +I agree that the tests under discussion should be and will remove them for now pending the submission of replacement testsBest regards George +Yes it happens when server B that server A makes a WS call to is running with Kerberos enabled +After a very brief look I dont understand how this would result in correct behavior +I suspect Im just treating the symptom with that patch so happy to see a better solution at some point. +New source file thats missing from the patch and is needed for the solution +does run now +if you can post some more information or a sample project Ill reopen +Donald personally I dont like lots of gorp coming out so I expect whatever we choose someone will be unhappyId suggest a vote to gather community consensus and lets make that the default going forward +In other words I wonder if we can have a version that doesnt assume Lucene? Thus I dont want to rush this in +HiI just started to use mantis! Ive been using maven and to automate de build and documentation +Also the example which I am trying to implement doesnt support Spring Integration API as the implementation class in the example has imported the classes from +Committed +Applied now as well. +Why do we needdo itBecause I wrote it and it seemed useful to me so why not? We know how Strings look like at least in their worse case +Verified with commit ebfaddffecedeb in +Sorry broke your patch Jon I applied the Writable and pieces from this patch so I could close two issues and fix +or a bad memory I was talking about the demo serverfor the windows there is a hickup in the core system the virtual file is fragmented extensive use of the virtual file is being used to swap program memory segment large virtual file that most is being usedit is a combination of thread racing and delay in response from the virtual file when doing program segment swapsthe above can be alleviated by using Norton speed disk that put the virtual file on the outside of the disk tracks and defrags it completelyand or have sufficient real memory not to use the virtual fileper the code segment both condition are set in the class +run Nowhere in the above do you install the resource adaptor on the instanceThe easiest way to ensure the RA is correctly installed in is to follow the instructions for installing HQ in from the quick start guideOnce that is done replace localhost with the standalone server ip address +Hi Ted yes I clarified what this operator means with Jacques earlier +Charles are you sure +bq +Thanks! +added full table scan syntax +The bug seems to be fixed in ticket can be closed +This is the source code used to reproduce the problem Ive described +and does a Clean publish fix it? could be the automatic deployer kicked in before the copy were finished +verified on DRLVM classlib revision msvc +Are you saying you could potentially add the JMS one at this stage if one has not been specified in and have Axis process it or is it too lateRegards one Dinesh +For the time being only hot redeployment of the entire web application is guaranteed to work a patch from BEA support for BEA WLS +True +OK to close +This is a effort being done as part of +Absolute paths are just not the way we do things in java +no problems +Right you want to adapt to effective reply time not just the network part +Only seems to work as IP address of the transport connectorname of the current host with or without domain extension with usetrue or usefalse doesnt work +Test case to highlight the problem +Ive created a branch to solve this called bigdecimal +Attaching logs +Thats why ORIGIN is a way out of this +Committed to +The problem with the current system is that people using external systems are linking to the URL of the page title because this is the URL displayed +I updated the for the requested URL so that I could add a group Id +this patch fixes the issue +This issue has been triaged and decided to not blockpreventhold the EAP release +Sorry for responding so late +bulk defer of issues to +having put a seam project together quite quickly and easily I am finding Seam and Seam security to be a bit of a regression. +Closing as duplicate of +Other use cases auto summarization document classification topic analysis text similarity metrics authorship analysis language detection plagiarism detection document clustering reading grade level estimation and so onThere are dozens of NLP text analytic operations that can be performed if we have a simple text extraction API +Sorry forgot to mark this as started writing tests around this and +Attaching VDB file and both datasources +Thanks Trevor +This isssue is fixed +bq +Im also including a minor optimization that switch to when we clone the updates in Commitupdates +I have added some test cases +Now supporting all +doh is not released yet +Therefore none of the classes find the service implementationWhereas in HTTP Server implementation calls to set it value to axisservices +Ive tried to setup a similar scenario +if you have an implicit List for Object one for and one for String then a String will always be assigned to the rd list a to the nd while a Boolean goes into the first +Well we will see + when I try to create project without google plugin it creates sucessfully and runs on AS +The next version of will be replacing this simple implementation with one that can support clustering and therefore should resolve these issues +But if project A depends on B and C it should be possible to parallelize B and C builds +Darn it +Thanks Charles +There is now a which should be listed in the UI when the model is updated in rider +It has been decided that the control constraint validator will perform minimal checking on the property in order to provide maximum flexibility on the value that may be specified for this property +It wouldnt normally go in the container unless the container is actually using Logj +I dont think there is any meclipsemaven specific about this problem the culprit is to be found in the project structure and its eclipse defined dependenciesMartin if you could create a minimal project showing the layout and dependencies and having this bug would help us alot +Adding a portlet to any region results in a in the dashboard page. +I believe the javadoc of is correct and we have an issue with +This should fix the jdoc format issues. +What we need is avery simple peace of software which we can run to see how the index is createdand the queries are generated and processed +Looking at it now +If there were no licensing restrictions what is the harm in having this in the tree +We cant have different configurations based on databases +it runs without +I think we would like to work together here +same patch try tend to think that the patch is ok and that the errors were seeing are the usual flaky stuff but lets try again +Already done on features. +I also wonder if is where the REST API should end up it makes sense from a pure repository perspective but with the HTTP nature of REST it might be just better off if the functionality were located within the WAR an explicitly HTTP based Java artifact +No this is not implemented +Have verified it on the latest build and latest UI now fires the remove with UUID only and things are going fine now +one extra command does not bother me we shouldnt slow all of our jenkins jobs down to make jvm easier +open httplocalhostmetamerfacescomponentsrichAccordion +You mean so you can cherry pick which ones are turned on? Makes sense +Committed revision . +That way an can manually check to see why that file might be missing +OK thanks +has since released the repository exists as well so its not clear which one is ready for prime time. +Thanks Owen! +Turns out I did miss couple of genericization pieces +the same problem seems to exsist for jboss have the same error on Jonas too. +We have thought about writing tests first +Its not clear regarding the context of that manager can get can set and will succeed +As indeed there is noerror if DOM output is not used +Still no responseSen. +I wait one day and my assigned issue gets completed automaticallyD +Attached is an alternate automatic fix for this issue +This is introduced by +Have a look please +Attached v for that also tested that will not return the same numbers even being called at roughly the same time +Will have to think further into this +Roger do you feel like opening another ticket regards adding the field ID to the output? +Now I am building trunk this evening I will give you a wait for a while again +Could you please help verify? tested Rexs fix and it works +If not bounce back to me +We can open issues for more specific things later. +My application uses HEAD in a number ofplaces and having to revisit all my code where HEAD is invoked just to solve aproblem I never encounter seems weird to me +We would like to get this bug fix into the branch +same for trunk and look like something that crept in to the code by some IDE the sun classes are just shaded versions of the original ASF code so this switch should be harmless +I still get this error message when I deploy my EAR project to M +Shouldnt this cause some sort of performance degradationI do not understand what you mean +Thanks AdrianYour patch is in trunk revision +Could you please test it? Sorry for the issues +All quickstarts listed in the description work fine on ASJava and ASJavaexcept webservicewssecurity which fails on both for the same reason quickstarts now work on both ASJava and ASJava as well as under restart did tweak some files explaining the usage of internal wsdl in files as well as files for webserviceproxy quickstarts +Part of the redesign negates this. +Please remove resolved versions if the issue is rejected in any way +committed to trunk thanks for all review. +added to in revision + +Hope a good tripIn the meantime I will check one more time and commit it if everything looks good +This should now work with any precision on the date picker +Added baseEncode Xpath function to synapse +Log into Asset Maint as that userTry to update the DEMOMACHINE fixed asset the operation should complete successfullyUsing the same demo user try to create a new equipment +Thanks +Includes change to just went though your patch and realized that the wile loop will kill the CPU power we had the somewhat similar mechanism sometimes ago +That Hadoop QA output seems a bit confused its a failing build applying +Anyway Ill add the table name to the assertion message +Ill commit this soon unless there are objections +Could be related to useing seda as the queue +But stomp is differentThanks. +Thank you very much for your fast assuming that the ACTHIVARINST table contains the values of variables of a finished process it should NOT contain a removed values of the removed variable should be available in the ACTHIDETAIL table +the patch in light ofNow the contribstreaming build no longer pulls abacus code +To reproduce the behaviour deploy kitchensink application open httplocalhost register an user with a long address table fixVersion to assess the issue again +We see this in as well +I dont think theres really an issue here +Perhaps you guys can implement some type of API that puts the web app back at a certain state based on a hash tag or something +on technical merits +add new file have done some more work on this since my original patch please disregard patches and use the higher numbered ones from this update + +adds the change to to avoid sending the stack trace as part of the message and also adds history saving for the last commands +Logged In YES useridSend it to In YES useridI have looked at the testcase and the problem is that theAnt Project is using a call to load the javaccompiler task +this is a dupe ofI just saw jiri peterka as something that tested on a full install level I was looking for something to run on a component for itself. +The fix was not correct +this requires update to parsing as refactor on webobjectdefinition to move scope property to iobjectdefinition +fwiw here is a patch to use the algorithm from the unicode std for utf in utf sort orderthey claim it is fast because there is no conditional branching + for the patch +Setting it to false should cause full form submission on suggestion you please provide page code +This is indeed what is created when running from the leaf project not what I showed appears the only thing that could have broke with were the changes to and +I do not plan to backport it any further at this point. + code just committed this +And heres the mockup +Closing this issue. +That would certainly be friendlier +John can you also take a look +Assignments go on child issues +well we mark the page as stateful when you call urlfor on it but it seems that the code that fails is ahead of thatim not sure that works because the only real way to know if the page is stateful or not is if someone calls a urlfor on it +Patch looks good +patch and test case. +Thanks btw I had a bit better name bq +The big problem here was not chained exceptions it was more how solrj was dealing with exceptions in some cases it had the code to do something more informative but an earlier exception was short circuiting it. +Attaching test case which reproduces the bug +committed and tested +yes the dev list has recently voted to make that the new to fix in patch in trunk +Created for. +There are maybe other cases like that which should be improvedYour patch is in trunk at r +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Ill look into it +Works wellThings to update rule should be active in default profiles all deprecated rules should be removed from default profiles theres a typo in the code snippet Cimpliant Compliant +I get the expected compiler warnings on Java +XX +Having said that it looks like well need a new test to be added to our suite +You read the size as a varint then read the appropriate number of string bytes +Works fine in fop +Thanks Graeme and Jeff for clarification +Is there any limit on journal file sizeNo. +Fixes tagattribute issues in StephenI fixed the value by hand in revision +includes and part of that provides for setting mojo params from variables now even more usefull when in need to check params values from +Works for me too try bretts suggestion +Ill need more information here before applying the patchThe attached quickstart works perfectly with current +If Fernando et al are OK with it I can do a +Closing all old Resolved issues +Which repository are you guys using? Could the root problem be Nexus related? Would appear that maven does not find the dependency in Nexus +Im actually kind of surprised this is not supported by Jaxb itself Kohsuke Kawaguchis native language is not really or proof +v. +Refactor all packages to +I will leave them Resolved going during test cycle closing. +Max allowed age I Current disk usage +See +move to scratch the package thing +In case this is already sorted out and Ive been ranting in ignorance please do point me in the right direction +Mikhail please verify this +I noticed these kind of problems when working for example with Intellij IDEA on Ubuntu +Let me know when you are ready to use it and I can create a source zip for you +Tried on and seems to work ok except old state value gets An error ocurred +fixed commit in comments +Very nice work as it really improves the look and feel of the GEP deployment plan editors +I have do some hacking but have not gotten what I expected +from the source tree +Haykel I could give you some reasons but I wont because I dont want to sidetrack since the question is out of scope of this issue +Very cool Lance +Here is a fix with specs +Removing Fix Version since dupesinvalidwontfix bugs clutter the release notes. +Just for my information do you want the conditional expression to show or actually the name that you specify? Either is fine and I can add asap +Im marking this WONTFIX +Srimanth let me know if I could help out on migration +This is why I it under a higer priority level +Undo. +AFAIR the current approach was chosen such that the jcrspi tests can be for other SPI implementations +Thanks Arvind for the patch +The embedder proved to be useless in which is why we removed it +UI for queue listeners of an AS following is the UI for the queue listeners of an AS aggregateI used the Wiki markup +committed to masterThanks Cristiano +Download attachment to the lib folder see picture with required libraries +So closing this issue +patch suffix + Patch looks goodCould you run and unit tests if you havent done so +Thansk for posting back closing as INVALID + +Comments and questions are invitedLooks like a good direction to me +Same thing for other be ok now tests needed +Ill hopefully be able to look at that before release but its not looking good +But we need to see each of the keys +This seems reasonable +Srikanth +Could you please add a rat exclude for that directory +I dont know if you like this solution +Great! Lets schedule this to roadmap +This issue states that the does not work in most cases +It is interesting that Jenkins is not started after renewing the patch +Option is now available in sever editor enable in place shared lib in which for a given project being deployed any jars on the build path or referenced projects that are not configured to be packaged as libraries or utility jars inside the application are deployed to the shared library. +Looks like its an established pattern then and not something that needs to be addressed in this issue +added conf to the ignore lists +This class would replace the existing class in the jar file shipped with Geronimo +Closing all resolved tickets from or older +Something along those lines +and I get the same error even if its not show in the same way since Adrian has refactored the way FOP error msg are rendered +This should be in Windows help with the Windows scripts +Normally the exceptions should be handled by cxf interceptors. + Id actually prefer to modify all methods that take JCR paths to treat the referenced item as not foundIf a would be thrown we could react accordingly +absolutely the same situation in my case +my bad there was a problem and the local repository just needed to be flushed +Will provide a status today +NET development +This could be much lightweight than repair except the network IO pressure on the bootstrapping node +I looked at using Toolbase but this really is just a general startup class for other classes who parse their own arguments using Toolbase +In the problem description the CASE expression in the ORDER BY clauseis identical to the CASE expression in each of the SELECT statements inthe child table expressions of the UNIONThe problem also occurs even if the CASE expression in the ORDER BY clauseis unique for example replacing the ORDER BY clause in the description by order by CASE WHEN IS NOT NULL THEN ELSE ENDstill results in the NPE +Tabs are not menus +Ask to hudson +I am not an expert in HDFS but I have read through the changes and they look good to me +The exception handling in Curator isnt that great yet and leaves me wanting +Could be a lot of things +uniting is and recoverTask into a single apicombining the api has problem you wont know that recovery is supported or not until you use the recover task api +It beats not doing it +Hi MirekWhen a node in a cluster is disconnected from DB it is expected that the cluster will block any DB related operations during the retrying to make sure no message is lost or duplicated +ZIP with the three repro was already fixed. +Before if iPad was in landscape mode the orientation of regardless was in portrait mode + +This patch fixes that and has the code for the attached issue so that these failures will cause junit test failures as the should in the future. +Screenshot from Interactive SQL tool +Related method at had already throw and +Perhaps something needs to be done so that such queries can be handled earlier even if it means coding against the servlet API rather than the Tapestry API +I just committed this. +Did you try this with a snapshot build or only with +You would need to either use or Talis RDFJSON both of which have implementations or you would need a plugin for each JSON document +Thanks +Ben can you please add javadocs for those +If I supply the required text changes can someone take care of making the changes? If not can someone point me to where I can learn about forrestI believe Doug does have commit privilegeRegards members maintain the site +Patch checked in +Not a bug +Okay this patch completes the conversion to +The problem is caused by not flushing the blob data after an exception occursAttached patch makes the test pass +patch applied thanks +Attached test caseIm closing this issues as no one has been able to provide a reproducible test case for well over a year now. +A user saw the HDFS exception for their job so they stopped and thought of going back to and tried to start it +can you assign this to the concerned person +The original patch has gotten stale perhaps you can bring it with trunk and add this case as well? And ideally since you were planning to work on it anyway you can add some specs under specjavaintegration for these cases? There are some good examples there and we are in dire need of Java Integration specs +Anyway your patch is harmful and commited in trunk revision Thanks +we need to build for all the objects we want to navigate not sure about the performance cost associated with that as well +bulk defer issues to +Please verify if possible. +My IDE is showing no warnings +I should add that with lob inlining small lobs in a serial mode for even a client will also be fine. +It will return true or false according to the posibility of deletion +this is a big blocker for committing anything because it will only create bugs +Any updates on this issue I do not see anybody assigned to this defect +This has no use in and of itself the others in your recent string of improvements have made sense as standalone changesMaybe create a parent issue where you layout the all the changes you want to make so we can talk about the best overall design? +bq +ok uploaded a better tested version of the patch +fixed verify in JBDS +OK just went over the comments my concerns are essentially the same ones that Todd Suresh and Allen have raised +Since it is only a test class no users will be affected +Konstantin I dont think the work pbing endpoints and their protocols has been done yet +backported to and +No filtering type pom was not necessary for my use case + is a patch against svn head as of today for modulescodegensrcorgapacheaxisschematemplateThis patch fully resolves this issue +I just committed this +Please continue this usage thread on the Maven users list. +Alexey patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Moreover the catch block does not contain anything so the exception is silently ignore which is not relevant because it will result in an unpredictable behavior +So I am going to close this bug as invalid +fixed the javadoc warning +Patch looks good to me +Is there a with the current Hibernate release? Without a solution Hibernates usability is greatly reduced. +Since we compile under JDK we should make a simple unit test throughing an exception maybe from a method call in a templateThanks again for the contribution +would likely present the same problem +Same here canonical options are better imo. +Maybe its not getting applied correctly +BTW if we are not maintains this module IMO we need to remove it from SVN WDYT +Id be happy to hear any other workarounds from the Spring and AOP configuration to implement many thanks for your help and attaching the it wasnt very elegant +It will take about another months to release +Fixed for trunk and +This was mitigated with WicketAdditionally each redirect causes a new temporary session to be created +JBs JACC implementation seems to be a bit of a moving target ATM +Alright I just created please check it +Thanks Bikas! +by my reckoning security is the next blockerLooking forward to seeing your patch +Set a better title +Yeah I was thinking it would be cool if we could wrap a config screen around it to manage all of the config xml files. +I am ok to keep subtasks under HDFS as of now +tests this bug +It works fineRegards +They need a review though to ensure they are conforming to naming conventions and completely convertedWhiteboard plugins are not considered a blocker at this point +Ive committed this +Ok I have updated the components and adapted them to the incubator +So if someone else could work on that it would be really great + fix in rev! +Tentatively targeting for +JBAS is already at jbossxb should suggest the jbossxb version to use for is going to stay at CR unless request new urgent fixes +If youre willing to wait a bit I can spend some time trying to fix it +I indexed a few million docs using cuss cloudsolrserver single docs and bulk add nodes shards replicas and Ive yet to see any update lock up myself with the latest code +For grins and general knowledge Jenkins provides a history of job configurationWhat you have done is better than just reactivating the old build config btw +Ah sorry +Then there is very little error from the expire interval and we dont have to worry about a monitor interval at allpriority queue This was exactly the same thing I was thinking the first time when I wrote +upgrading from current to latest fixes js problems in +It will be checked in the next build +using r which is future proof and in central +I thought you are talking about regular presence stanzas as you didnt include typeprobe in your examples +Also local variables and parameters are not handled +Move it back in if you think differently. +Thanks Ying +Okay lets try again! Two issues I was only giving an incremental diff from an earlier patch and I also hadnt rebased against the latest trunk +Thanks applied in rename of propertiesApply to core +That benchmark really isnt valid for a high number of threads though notice the difference in hitRatioIf you have many threads quickly adding items and only one thread at a time removing items the goes over its target size and thus increases its hitRatio making it artificially fasterThis isnt a concern for its use in Solr though since the generation of a cache value will be much slower than clearing the cache +Uploading the same patch for running via Hudson +I think well have to get in touch with the mapfish guys to see about having the print module get its configuration from as well as implementing the proxy URL option. +Presumably there is another bug open re the console that someone iswas working on +Go ahead and commit +patch files here for posterityAlso this should probably use Accept content negation vs +attributes +So I think Ive figured out what I want to argueTo make another analogy think of this feature as a tool +Thatll do for now. +It was not a problem in how we profiled the environment +It seems this versions is totally broken with the docbook integration as well +So we can mark this as resolvedThanks the issue since the patch is already applied +Now it works for rc thanks +Although this is a bug fix Id prefer review of this patch before I apply it +The patch looks good to me +Related to but the problem here is that the connection fails to move to the connected stateWhen watching the await state call it will return that it failed to connect and hence will spin in the main threadsleep +Build a new context every time? Remove all references to rhinoClassloader +I made the svn propset changes and committed this patch +On the right the gc events are much much more visible as spikes +But thanks a ton for this patch well try to get it in as soon as possible +Ok we will cover scope at the phone call next week and come up with a statement +The test also assumed this to be a valid case +Thanks for the review Daryn +Serialize this across RS massive delay +eg +I dont really think that there is a perfect name for it and I dont think it is worth breaking backwards compatibility just for a name change +bulk close of resolved issues. +See for further details. +Just a note I think R was meant for running with java and generics may broke it +Cool thats short +DanShould this be set to Resolution Done as opposed to Resolution Unresolved +Gating each send on available usage up front may be the simplest approach but that will have lots of side effects like not being aware of flushing to disk when a cache is exhausted +is there a thrift ticket open to fix that +This seems to be an issue with the not working +You know I personally would liketo see retry logic moved from to however for thetime being the patch is a good interim solution is is apparently misspelled I think isRecoverable flag should apply to writing the request as well asreading the response +Without answers cant see any reasons to keep this Jacques but just FYI Im quite capable of handling the issues assigned to me +Also reopened in CPwebserviceconsumer tcunning svn commit Sending file data +Made this a general jira for issues with the seambay bidding +Heres a patch that implements the suggestion above exception is now a list an exception in one faceting command doesnt cause the preceding ones to fail and all shard faceting exceptions are added during aggregation +I have discussed this with Brian and he said that these additional artifacts needs to be added to whitelist +This issue is duplicated withThe patch is in the trunk and branch. +I have increased the memory and ulimit on my VM see if I can get a full test complete +Attached project can reproduce problem +bulk defer to +Actually the problem with the incorrect assignments to choice values will be resolved through +I am wondering why is set with minor priority though. +It is quite useful since I met the same exception in the hicc service log +thanks +Best regards of demos +I dont think this is the same as requesting that those points not be clustered at allI dont see why these distant points affect any other canopies +Instruction to run the test caseA case start an standalone broker start server start as a testB Working case start an standalone broker start server start as a test +What is the current relationship between ? Is a subset of ? Or should the patch be applied first +if i have added axis kernel jar to the same lib directorythen its workingBut I need to get it done by Axis kernel +The solution to this problem should be consistent with the direction the spec takes +These two issues intersect wrt +This proves that the isse is within Hibernate code +Would it be possible to change this to run on Windows? Although disabling the shell script would allow the build to work the result will be different than what you get on a Unix system and will require the folks working on to have to do this work +dunno +the patch looks good +It exhibits very high variance and its results are very much dependent on mapreduces scheduling +Thanks guys! It built fine now +bulk defer of issues to +Tried above mentioned scenario after changing the group of the user but copying is still possible +Moving to unless somebody makes a patch +Regenerated patch again with the trunk +Hi Tim Thanks for the quick response +Welcome Dave Vronay as a new Contributor +TIA Wolfgang +Migration in progress +cant reproduce this just editing the file +Rendering a JSF view from Spring MVC does not support posting back. +If the consumers fail to start they will back off and restart according to the process defined in +Now that work is well underway with AS all previous community releases are +This is not yet an ideal patch as static variables is bad +newest fuse dfs which should perofrm better on reads +There are no automated performance tests +Fixed in the CVS +The associated ticket isCopied the classes from Commons VFS to start package and removed the Accumulo versions +So I would say would be helpful but we would need someone to really spend the time investigating any bug reported as quick fixes might make things even more complicated. +I have been having this problem too but most of the time its fine +Ok this patch is updated to the current head +I dont commit to branch now but we may when we feel a need + and translations to spanish at revision stuff Gabriel Appreciate the work +Patch containing modifications to class and unittest +If everybody created a JIRA any time he has a question about the JCR it would be a chaos and JIRA would be unusable +User must make sure to use either XML Spring Integration namespaces or to use the XML namespace. +Backported to branch. +This is the right thing to do since the purpose of the cache is to record the last decision made + in the header file add better documentation about the possible return values of hdfsRead +one last unread field +practically optimal with performance point of view during editing complex custom templates for vpe and makes very smooth and prevent from intensive update exceptions +One note for the query above input format is which is not friendly to this kind of query +if we used the rather than +Resolving as fixed for with the component being in place +A file which triggers the error condition +This patch addresses the issueTwo additional XML templates are created to produce proper pom files for pigunit and pigsmoke artifacts +We can copy all the sstables which are present in the replica and keep a View of sstables copied +an incremental patch to enable running of the query as the enduser +Thanks +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Attached update with new ASF headers allas you have probably noticed Daniel Goodwin has been so kind to attach to this issue the relicensed version of his workSo I think that all the license issues are resolved now and that we can continue his workRay would you consider to recreate and submit a new patch against these new filesDaniel many thanks +Issue was fixed in latest SVN head +Close and reconsider in light of jBPM +These issues have been marked as resolved for some time now with no recent activity so I am moving them to the closed state. +No followup from original poster and a related issue was fixed so marking this fixed as well +DimsApart from not using schema with spaces in file path I cant think of a workaroundAm curious what is the design consideration for space AND commaRavi +The Double may be converted by throwing away the fractions and clipping to without throwing an exception +Theres nothing more to improve thereIf you have ideas for improving this further could you add them to the description of this JIRA? + patch looks good +Jenkins already has IVY only some changes needed for build scripts to cleanup correctly +Updated root by applied part of in in SVN Rev +wrong +Closing issues before next release. +So impact is not as horrible +Update the guest OS installation in the same wiki page +I added the required event handlers I overlooked the Log messagesThanks for pointing it outThanks to Jenkins to see if the I created is working properly or not +Just refreshed and test passed +The third time was the charmAll unit tests now pass and I have comitted this patchThanks for contributing Stefan. +Nope the one on the map itself is what i had in mind +I think youre making the speed scalability mistakeNo Im simply acknowledging that theres no such thing as infinite scalability and if this scales to the machine counts people actually deploy on then its silly to do something more complex +Opening the sample ruleflow file generated by the wizard seems to be fine +So I think there is only a little overlap withThe balancer did not have storage information so that I added get +The only reason we use YUI is because of the compatible licenseSince Nicedit is compatibly licensed and much smaller than YUI on this change +Please crosscheck and close this issue then +We might look at moving it into the merges directory to be less magic. +the individual will need to rescan all the artifacts to populate the database correctly. +Andithanks for reporting the problemIf you have not set database transaction isolation level on Oracle system level the default level for Oracle is +Closing based on the availability of +Looks good to me +Thanks RickI havent included the change to include the new property in since core shouldnt depend on contrib code. +I thinkIs the only benefit in this approach that you dont need to manage a sidecar taxonomy index +Updated Apply Template section and Smooks Configuration testing using Smooks Run Configuration section +Hyperlinking Java types also stops working +OK + +This sync is only needed for sends but it impacts acks so it needs a revisit +Unfortunately no +I will double check +Verified on +Any chance that you can submit a patch +Currently we parse JRUBYOPTS env +Ok whole patch committed with changes +Same behaviour on +End user configured endpoint incorrectly and uses invalid HL messages +Attached patches for and trunkIncreased the roll time to minutes from minutesAs Konstantin suggested we could improve the count by updating the count when a block moves from pendingReplications to neededReplications and when a file is abandoned +Continuum has recently addressed this so we can leverage the work done there +collections issue was fixed in but that issue does not have a fix version can you please update that issue and then I think this one can be cancelled +Bulk update to fix version +If there is a significant degradation then it will delay our adoption of until its optimizedI think a good performance test is to create a namespace of M paths +Thank you for the feedback +Just need to commit to repo +Bulk change of resolved shipped +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Need patch and fix for use incorporates to link the upstream bug +bulk move of all unresolved issues from to +Note that there is some instability in Derby which affects the script +Will split that out +Added serialversionid all client side classes all their references to other classes through their interfaces. +didnt we have some other issue that showed that bundling directly where the wrong thing to do +There could be two possibilities +Let me know if you would be able to take up this issue and try to find a way to automate this or if its not possible as anticipated become the vmware ova systemvm appliance maintainer and assign the ticket to yourselfCheers +in revision could you let us know if this will be included in +Good point Peter although I dont think that is a cause for concern +This was reviewed and committed by stack back in. +If Derby can catch a bad boot password or encryption key before trying to decrypt the conglomerates then Derby can raise an intelligible error messageIf the database was created with a boot password then Derby expects to find an encrypted attribute in +So this is not required +the bug is that it has test classes in the compile classpath which could cause other problems too +cleaned up less noisy version on trunk updated syntax for shared dir in multicore config +Accidentally committed to branch but this seems safe enough for +some tests are still problematic at least on windows +Ill avoid the use of this and avoid syntax modification of that file. +simply writes any image of a pdf to the disk +Please note the additional features as well +In the case when we know there will be lots of results we dont want to just show results with no constraints unless the user explicitly submits the parameters form with no constraints to request it +we decided this would get put into m attaching a patch for the proposed API + +You really dont need change at all +So it is not distributed thread for mapsone thread for reduce +I agree they dont add a lot more to the basic underlying class andgiven that your atLeast concept is effectively a unionor concept there areother boolean logic ways in which people might want to be handling such thingsSorry to be knocking this back after a year of silence +Yes it is just an entry in the log I will insert check for plugin +I cant find the way to do virtual guests in +All addressed +The user must set the relevant system property to remove newlines from generated xml which isnt ideal but this might get fixed in a subsequent version of. +method introduced at r thanks for requesting +Other than my reservations about the whole business fine with me. +Today I am at a customer so I cannot spend to much timeWeb projects are no problem at all +If a configuration detail view is opened from the configuration list a POST request is done to display the details +Upload a new patch to address the review comments +I see! If you couldnt make it how can I make itBut the idea is that you could have some kind of object as output instead of a string like for example an array of tokens plus maybe some more informationIt would avoid to parse again the string output and somehow the text as needed to have a snippet that we could provide as output directly +I will commit it shortly +Red Hat Middleware is what I was told to put extra sentence will magically disappear in the next build +Yes two different exceptions are also ok but it might cause some backward compatibility issues if the changes are applied to the existing method +It may be hard to judge with so little context +what dose open a copied file from another project meanwhat type of file? any case I think this is a eclipse WTP core issue not JBT +This appears to be due to firewall issues. +Attachment dd has been added with description A unified patch against HEAD +Everithing seems OK now +Do you have a patch for this? The site sources are in srcsiteapt within Tika trunk +While this is an installer only issue it ends up in a broken picketlink distribution +I have no experience in Openmeetings but I am familiar with Java and I worked a little with RedI have some questions about the task +I just returned from some time off and am in the process of updating the patch to the latest in branchx +I changed it. +They will be expanded to the Teiids view of data sources which in turn may be JCA or JDBC data sources or something else. +Just needs to tie loose ends and a LOT more testing! +Ill let you know if it failsRegarding NTLM I just wanted to restate that yes some people use the domain in the opentext username and some dont +We may improve this in future and after we adopted osgi extender model for geronimo configuration it would be directly supporttedFor now lets just doc it +Ive just committed this to. +What do you say commit on or wait for the release +Please submit a patch +Even if the filter selects most of the rows select from T where +Reflection is used since it is possible to use Tiles under a portlet environment thus using instead ofAbout the upgrade why dont you try at least on your machine with Jetty? If the bug is still there we could fix the problemWhat I dont want to do is writing a workaround in Tiles for a bug in Jetty +Ah okay +Logged In YES useridCan you show the original error that failed the deploymentor at least describe what the problem was +So we dont need to keep and in source code of each plugins +is about the type of the column +Doug would be nice to have rpcsend as well +If it is too come to under the apache license I will need to see about a couple of classes that were provided to me for an EPL project +No luck with the move to Linux +Test cases are modified to match the expected behavior +Ill commit this shortly +In the meantime could someone with higher privileges in this issue tracker add to the affected versionsUpdate The problem appears when using asynchronous loggers +I have reproduced the issue on my would need to introduce an option to tell to use a background thread to create the hdfs file system you would be able to start the bundle but on first access to the hdfs producer it will not have any either we should add some timeout value to wait until connection could be given fail asap block until hadoop fails leave it as is +This was merged into the codebase back in April so marking as resolved +When running a Hello World application through the debugger I see abcxyzFrom what I understand of what Ive seen the Request parameters are separate from the query parameters +thanks for the report +Thank you Harsh + am closing the issue +Click on prescriptiveACI +Assigning to you bob for closing and feedback +We wouldnt want to hardcode the normalization +Please try my patch +It contains an older version of thrift that is no longer available in +I find this information the key purpose of Release Notes +The relevant section of the affected region servers log +I am not sure why this is invalid +If the UUID of the is equal the UUID of the SNN we cant download the fsimage of +Patch for yahoo dist security branch +bq +here is the patch for the a patch for branch neededI dont see srcjavatestorgapachezookeeperserverquorum under branch + on revision +I was told that this code path is used when replicated documents with attachments +Submit a patch for +It wont let me add static stuff to the annotation class so upto any complaints Doug has with naming I think it is good to go +Upload a patch for the documentation for +Ping +Patch looks good +Seems awfully complicated +In addition would be possible to store this info in a properties file in the JAR instead calling a script to generate a classOf course the implementation of the interface is module dependent +Whats the Traffic Server bug here? I think that for this to be actionable we would need more details on what the problem is and how to reproduce it +Revision Constructor call chain +Method to check output regardless of gshell text markup +Thank you for your patch! Now Im still on a business trip +Any query where there is an extra will cause the HQL parser to treat that as the end of a statementan exampleSELECT FROM Table t WHERE ORDER BY the ORDER BY will not be parsed an will be dropped from the query with no error +This is a bug in the s already fixed for +If ext is not supported should I buy an external hard drive and make it ext? Also great to hear that I can switch to non AIO mode thanks +Thats actually a weakness in the JDKs class or in the JAR readers usage of that class +Could you provide a patch that works for you and that is safe for browsers that dont support it? Cheers +Perhaps we can use this to solve the classpath woes for our friends? I like it because it doesnt keep us in the business of custom classpath parsing and set manipulations on strings passed from shell scriptsSounds good to me +But not sure if it can be done cleanly +StackPlease review once +Can work around this after inspecting the code it is not really an issue anywayThe documentation recommends setting the delete rule to something different from the default no action and states that nullify does not do anything if there is no reverse relationship +Bulk closing stale resolved issues +I hope this is what you ment Gurkan +I went ahead and changed the initial size of each cache segment from to +Fixed the URL +so tried to keep it in the app layerbut obviously would be more efficient to implement in the sequencefile code itself +there you go is this something that is fixable in the branch for the release we are using of jbpm or +I think we are good to go ahead with a perf comparison +Lets try to stay focused just on what this change does making listStatus and globStatus resolve symlinks +Until those are committed I would not recommend relying on symlinks +See Beetle and +Also ran on and got the same results +I would open a support ticket as well if you have a deadline. +Actually it would be nice to have an offical glossary for terms like this +Thus executing the web jsf setup without options more than once leaves any configuration as is +Thanks for the review. +Committed revision +JKS keystore works fine on the and IBMAs stated in an earlier posting we cannot include the Bouncy Castle in Geronimo due to IP Licensing issues found +its real was mistake on my sideGrails use ivy cache and dont update from new local maven repoAll works great +Im not having classcast +More widenings +patch for for branches and +committed to in JBT +I really hope the issue can be fixed t a patched version be release with this fix before? It is a total break to. +Committed revision +srcmainjavaorgcodehausmojohibernateexporter protected String outputFileChange String to File that should fix it +In my case i had a typo in a file on one node shardashard was specified for one of three cores +So all that this patch does differently is try to define a broader concept of query context rather than the narrow fix Simon proposes +Ive found comments of other user having the same problem with Toplink and Tomcatclasses is the correct location for web applications perhaps hibernate could give a hint about that +Oh in the stack trace the root exception is +Ill update the patch +The messages tab is now added to the process properties +If there is space in the same volume then the rename occur within the same volume +I think should work not +Patch for including bundle in servicemix a ton for this patch UlhasSending assemblySending assemblysrcmainTransmitting file dataCommitted revision +Not sure what scm this ticket is for it doesnt say +FindList form to find employees in system +Testing now +Included in. +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +RE Tests the patch includes a whole bunch of new tests for all the new functionality its in so hudson doesnt see it +new versions with mentions of yahoo proprietary tools script for RPM for check status +patch looks seems to be messed up at the moment I cant see the test results but Im sure there no issues with the contrib tests +I do not see how that can be done unless you create multiple endpoints which will clutter the applicationI might be wrong in my assessment and will look at this further and thanks for the feedback +How about trying this out with the ER release? Youll need to start up two instances of EAP on two separate interfaces let them cluster together and then shut down restart one of the interfaces as above +Just expose add the extra code arg +Fixing this issue correctly would mean toadd a whole lot of checks to the URI class +Javadoc of says this Column names used as input to getter methods are case insensitive +Will copy the book over to branch just before cut next +When clustering development is complete then you will be able to deploy this +Yup feel free to up the priority to be sure we fix it for the next releases +So this may be the culpritSo please try this as a work aroundIn the broker configuration set schedulerSupportfalse and see if the files clean up +Reporters of these issues please verify the fixes and help close these this with latest build +So good news this stack looks to be working out well for me +These are the Eclipse project meta files +Thanks Nicolas looks likes one is new in trunk +Initial contribution of +A patch would be more appreciated Shi YusenPlease note that smoothfeather does not exist anymore and have been replaced by bizznesstime +At the moment this wont work in automated installation +Since it is a perfectly legal thing to have a null text node it seems to me that DOMSAX should be able to handle perfectly legal DOM documents +One question I have is if this diverges enough from the current project that it warrants a new subproject +Changes were back ported to and committed for release +Now we should discuss whether there are important use cases for the health script where it would be problematic to make it call lostTT any cases where admins want to use the health script to prevent new scheduling but not fail the stuff thats currently runningThe point of the was to quickly determine node faults +They could either open up sources downloaded from maven or go directly to the instance at Apache +Thanks JM +Could this be expanded to allow injecting the entire sql query? In some cases the names of the tables involved may not conform to an inhouse policy +suggested good +A similar issue was fixed in beta +added license header +Closing all issues that were resolved for. +Sort is not mentioned because it is handled in the +Ill keep an eye out +Good to know we have backward compatibilityBTW John all these items can be done in a transaction if needed +Now that work is underway with AS all previous community releases are +BTW if you want to see for certain if the fixes were made I gave the revision its included in the source is open +Marking this as Invalid In case there are other users I will be happy to provide a revised patchThanks to all for the reviews and suggestions. +Just did a to confirm that his has indeed been fixed +In our installation with TB hbase balancing iteration is about minutes minutes took by block move and minutes by exclude blocks list preparation +that is bad +QE please any update +Enis you up for finishing off the change or you want us to take it from here +Yes it continues to page within the range +When removing an endpoint the gossiper should notify MS to destroy any conn pools and remove the timeout tracking +Is this still an issue now that is solved +Martin thanks for the report +Thoughts +Mea Culpa! You are of course right Evgeny and thanks for calling me out! After my post I am embarrassed that it does come across as demanding +Jira is not a support forum +You need it because after recompilation the information of being cacheable would be lost else? what happens if you disable caching for the string based version by default +Verified with JBT. +What I applied to +cotains patch text for existing filesBoth of these files are for expansion or application from the tuscanyjava directory +Marcus had introduced a check on the hypervisor type that the resize property is only valid for Xen serverKVM and not for OVM +Its just lines of xml +Just curious maybe howard will also be usecase +Thanks Adrien +I should be able to overcome that with a custom that would be used within respectivly right? Or maybe you have some better idea +Assigning to Rich for reviewThanks with revision +She is also willing to iterate on the ideas or do fine tuning so feel free to suggest improvements and changes +I got this working by setting JAVAHOME in etc +embeddedejb +It seems with this field cache approach and the recent on trunk that Lucene has a robust and coherent answer to performing efficient sorting and range filtering for float double short int and long values perhaps its time to enhance Document +Great job Ravi! Thanks Vinay for helping with the testing +The write rate at OOME is about kk a second +Ill nail down whats going on and file another bug if necessary +Im not sure why this is filed under menu enhancements +I have updated the description +Ive fixed the bugs +Please prototype the example basically I need to see when a custom parameter is reported back Im assuming it has to be done after the user has authorized the client but the question is does it have to reported back as part of oob response or after the client requests the access tokenIt doesnt matter which step +Closing this issue. +Chris How do I reproduce? iirc you just deployed the app +Hey Pavol I cannot recreate this issue can you try this with your JDK tests? Thanks +This would be a nice win for new output formats +If you can fix that we can the test that confirms text as well. +Brock will probably know I think he added that feature +Im opening a separate jira to track the capitalization issue that David mentioned in the pull request. +Please check that it was applied as expected +Marking bugs with invalid fixed for release as fixed in. +Log a warning if Solr is running in cloud mode and is not registered at getbq +After is fixed this task will be just to update Remoting to a stable version with the with Remoting +I suppose youre right +Ive checked mvn verify +Not much because it was there before but rather because its the better position when you need to enlarge the lookup +I believe this is fixed in the current code which will be released as +While we sort things out with the Hive community +can we change the exception message for invalid accessPlease ignore this it is being handled in itself + +The property constants are now all defined in the and the default settings are in the and +Partial review on phabricator +It might not even be the component thats of interest to the application +all deprecated are removed +It is a test case that failed too +Its a WIP because AFAIUI only default price works so Price Rules and Promo cant be following Davids David JacquesIm the initial reported of this issue and Im take a look at this to understand if the last improvements done by David cover also those requirementsThanks David JacquesI have tested last trunk revision that includes the improvements done by David and it seems to me a better and easy way compare to what I have proposed here in this task some years agoSo I could close this issue has wont be fixA small difference I have noticed is that normally in site that sell products with VAT taxes when you add products to cart the taxes are normally included or showed immediately and not like now that they are calculated when you select the shipment destination addressNormally the taxes are calculated basically from the shipment origin once selected the productThis can be considered as another improvement can be done laterThanksMarco +cannot completely solve the issue because it only takes into the account the amount of time the logs have to be retained and not the disk usage because of the cleanup mechanism itself is not guaranteed even in terms of timeWe should have a concrete mechanism to limit the amount of disk logs +Weve been running a similar syncnotify logger at Cloudant for a long time now without issues +should be closed as fixed. +In addition the capabilities of an instance are distributed in a cluster with a property provider providing the property +This might involve serializing the to the on DFS and using the same for authorizing further requests +Alexey can you confirm if this should be excluded or if the serial should be fixed +I will verify in CR +Thanks +put it in i a different category +They never are + exploring this I found that a large part of the performance issues stemmed from a lot of polymorphic call sites + +Ive reproduced the error with r +Thanks for spotting already working my fault. +What isnt working when building there seems to be issues here +Closing resolved issues. +Transitioning to closed. +For this test the client reported the failure but the job actually restarted on a different host and finished successfully +Perhaps someone else? +compareTo use intead of +Maybe this is related to the fact that I already had the plugin enabled for the projects then updated the plugin and restarted Eclipse +Let me know how it goes +Added a to identify references to remove +Thus it complains that the directories dont exist and wont build the projects due to classpath errors +Thanks for the discussion Todd and Nicholas. +This would serve a a no nonsense way to get data into a CF +Unit tested but not integration tested +ERROR Not jar seems wrongjar tf in JAR file format +I think projects bundling artifacts should take the NL files at face valueIf an included dependency has items missing from its NOTICE or LICENSE files that is for the relevant project to sort out and address in a subsequent releaseAt which point the current project needs to review the NL as beforeI dont believe the bundling project should be responsible for checking further +and will receive a patch today in SVN +But this may be needed so that the client is sure that the table is createdI have heard of a use case where every day a table is created and the same is disable and dropped on the next day before the next creation of tableSo i thought our apis should enable them to know if a table is created properly or not +If one is found it is set to be submitted at the onchange eventAs in your second page the component is on a form the textfield component on the same form will receive an empty string on the form submit parameters +Can you check if the commands have been registered in the registry using the osgils a command and check the log for any error +Attached patch that updates the reference guide for the new implementation +Is this option implementation independent +Has been fixed now since r +set proper fix on +Can this issue be checked against ER? I believe this has already been fixed +Here is the patch as file +Jonathan Any update on this bug? Given that it was supposed to be solved as part of is it now closeable? +m file in the right place and parseable? is the project buildable even when the the errors show upDoes IDE restart help +Workaround the user can create the adapter from the server this is ugly the user is not even shown an error message +This makes sure the values will be available for the resolveconstvalue call +added logic to calculate the concrete state on an executed project after a forked lifecycle completes so real values are available to the bridging mojo that triggered the fork +if I unapply the aforementioned patch that added the synchronized keyword to the writeXML then would the deadlock not happen? I dont care about dumping the configuration at runtime for the moment but if I can get my jobs to run then that would free up a lot of other internal tasks me and my team are doing +Committed to branch +So this means that responsive portlets will work for both Responsive and Default skins all skins we ship with +So all issues not directly assigned to an AS release are being closed +Move the fix to +Since they all have guaranteed capacity of the timers start + +Can you provide us with the pdf +This fix along with the could be applied as part of issue to change Release Notes flag. +Checked in enhanced for the api project +it is close to complete though +This is fix is provided as two separate patches +only onswitch not really critical to have according to onexpand and oncollapse presence +The final conversion happens in getIntall as a result of the inserted +In that case I think we should do what you see in in a shell show the choices +Resolving as duplicate of Ive been testing lease recovery on a cluster using I can kill an hbase RS then do a failover and the master is able to properly recover the file without losing edits. +Thilina just a reminder that the mtom guide still has a databinding section that isnt correct +We will check existence +Too many background moving pieces +Feel free to file new Jira issues to get things improved +Ill try to scare up an expert to document them +My comment should point out that inconsistency only +Moving to +I think i will commit the patch as it is +You may as well get it now + +Added the description on hotkey method for Build Project Archives action. +Virtual columns were added to the functions that as far as I see are only being called with the pruner expression that have been compacted to only contain partition columns +Is the Initial Committer List the same as the list of pmc members +Ive finished the task using the Path algorithm described in the Pregel paperWe have now a working example that can be submittedObviously the coding period hadnt started yet so I will focus on my Dijkstra Please checkout the new example and tell me what you think great +Note to future selves dont bother trying to grep for strerror or perror in Erlang sources +egx +Completed At revision +HiI think this isnt an ERB bug but e bug of window console because alsoirb a print is really related to the windows specific console parameters and not to bugs with invalid fixed for release as fixed in. +Or if this is a bug grails should copy resources for guess I should have saved my comment for this window instead of the other window +Kinda cool in the project docs not so cool in the platform that site I think is the latest +New patch to make JD happy after the Mavenization of +revision in build. +Thanks for reviewing Suresh and ChrisI have filed and split the patch as suggested +Another approach is to implement escaping using X pattern as a solution and restrict allowed characters set starting from jackrabbit +It works with the version from SVNThanks +It would be awesome if this were fixed +I agree that using the actions bundle instead would be better however a change like that should go into not the stable branch +Could one of you review this change and submit Prasad +I just committed this. +Verified the above issues in CP RC build +Cannot fathom how this could be missed in +Yes of course Claus +If someone could point me to the relevant code for this I can probably create a patch for it +This would be really handy +for the changesI think we can comment h +Commit changes to trunk At revision +Does Hudson agree +The supplied patch will support what I need to do for my projects and allow different levels of checkstyle alerts to fail the build +Committed a fix in txmanager trunk rev and branches rev this jira be closed since they are fixed in the txmanager code +uploaded new patch +That said it would change this into a feature request from the pov of the platform + +I hadnt noticed it +Attaching and with changes as followsM M M srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcgetstartM srcgetstartM srcgetstartM srcgetstartM srctoolsM srcadminguideM srcadminguideM srcadminguideM srcrefM srcrefM srcrefM srcrefI created a property for the minimum JDK and used it where appropriate +Grid any ideas +Moving out to that work is underway with AS all previous community releases are +This code is only in trunk updating the fix version accordingly. +top on the hosts shows all CPU time used by the JVM running the regionserverAttached as is a snapshot of what jprofiler sees of a process started as +permissions formatLets use octal +New translations +Let me know what the requirements are as far as unit testing +svn wo wildcards work +This patch remodels the layer groups creation workflow to make it impossible to save an empty layer group moerover it makes it so that catalog loading skips over empty layer group thus sanitizing data directories that have been corrupted by an empty layer group alreadyPlease looks good +Defer to C hosted +using a setting is causing the issue. +Rebased and merged into the master branch. +This issue was assigned to me by mistake +Bill private variable is not initialized before is called in the next line +changes is acceptable +r removes the workaround I added to to prevent the CCE +What about this issue? Any chance to get this into next release +sorry have not had time to apply after the super speedy mmigration was successfully merged to trunk! +cant be using the same connection concurrently AFAIK +Thanks Kihwal! +Restarting did not add the index to the table +But im not sure it is something we want in the iBatis framework +No need for IT +checked in. +The jars and are already included in the source and binary distributions +I have made a few minor changes to the class and added a class as a convienience class +There was only one occurence of a usage of more than elements which was +this is a factor +that would be for mere humans it is necessary to have code formatted to understand it +As such it would appear that sharing across threads is not actually the JNDI lookup does not return a distinct copy of this class on each lookup then this is probably a bug that I need to raise with IBM +MilindYes +Nate has your testing progressed far enough that you think this should be committed +Finally I found the attachment guess I was a little slow that morning +All mirror sites are not updatedDoes there has command to trigger mirror sites update +Anita What error are you getting on attempt to apply +Yes the test failed on a trunk build +thanks Adrian +Ah this is nice +Please for the inputs ThomasI am unable to cancel my previous patch +I believe the fix was all right maybe its something else +That other data is not needed until save operation that happens when user SAVES the form +I think we have optionsFirst is to put is as TRACE or DEBUG the nd one is to display it only if the last one has been displayed more than x seconds ago or if the last value was x lower than the current one +Cool +Setting the divisor that way shouldnt likely work on trunk anymore since Lucene dropped it +lib sounds like one more thing to add to the migration script +Howard If you look at the test times youll see that did not happen in my tests +Applied patch. +The only thing we changed in the assembly itself is the encoding of the in the informationalVersion attribute of the me know if there is anything we can do to help the investigation +JinchaoI find trunks logic of carryingMeta and carryingRoot have also the problem you mentioned aboveIn trunks logic of carryingMeta and carryingRoot before master finishing Initialization carryingMeta must be false because of empty logic is very complex +too as that would move us away from the preferred programming model. +The scan included xx VSVS debugrelwithdebinfo variations +Code style is as maven code styleMax Interesting point with the package names forked code will likely need a different package name to avoid just that issue +Look at my next comment to get more information regarding this +Can you please delete all Spring IDE artifacts from your plugins and features directory and reinstall the nightly build +The default for WRITEXMLDECLARATION falseSee the for for details +What it help to provide a concise example of issue with all the files needed to build it +Not many folks implement their own filters of those not many implement Filter directly there are workarounds and for Dave it no longer matters +I have tested installing the CXF bundles in Framework version and can confirm that class space resolution now works flawlessly +If you find some time before the release feel free to reschedule this to +Hi Karthik SureshI think everyone wants this to go in right? Both you guys gave this a earlierI think if we have wider points to make about compatibility we should do it in a mailing list thread or another JIRA +docs? +I cannot think of a situation where you would not want to check this box +A node could manually mark it self down before starting a drain and clear the mark but not startup gossip again after it was drained +I reschedule for t as this becomes relevant for the outdated issues +The includes files are now expected to be in the srcdocbookprojectname directory with names akin to +Leaving issue open +I will reassign it to him when I figure out how to add him +Im working on getting this demo running with ode now successfully deployed as an bundle +Same problem for me using the classic command line mvn releaseprepare +for commit to branch and trunk +just mean just as config in yep +Fixed and added testcase + +we need documentations for both ledger manager and hedwig metadata manager so other one know how to provide different implementation +I imagine the overhead isnt all that great so what you are suggesting may have merit +I am glad you opened it up as i now can safely remove Put in a separate section within transformer under from my local todo list +heres a test that fails without the patch passes with it +Without any steps to reproduce the first error just indicates the ssh key or usernamepassword was not approved and the second error could be a lot of things but if you are changing IPs while running then I guess a lot of things can suggestion Dont change IPs and expect everything to continue to work you can reproduce this by getting a new IP then please do let us know and outline the steps to reproduce as otherwise I would suggest we reject this as cannot reproduce and assume this is errors caused by out of the ordinary network setupchanges or a misconfiguration of permissions +Will work on diff report separately +Andy do you think this is a bug in Roo or is this something that is best addressed as part of? Its just Id like to close this issue if theres no changes required within Roo +Patch with excluded tests +No Im thinking about blueprint bundles for examples which are started asynchronously in a separate thread after the bundle has moved into an ACTIVE state +ArvindI agree that we could cache resulting parsed objects +But I think that should have been outlined clearly in the early stage of development of a featureSee above +It seems like everything is fine in version +This is what I did in my patch +SorryRunning test +Patch attached +the problem has something to do with this new index stuff that replication uses this thing always gets in my way +Have you tried a recent checkout from Tika trunk +Thank you very muchBest regards by Richard. +I can figure out how to do this on the backend enough +It seems GC is doing this +version and owner +Heres the list of failures I see +Assigning to JF for investigation +All the jars that used by devcloud are copied from your local source directoryYou can check above jars on your local source directory after mvn P deps +In other words we have to make sure that we account for a handler that turns around and posts a new event +The same doesnt appear in the layer to shape export also +Attached patch implementing the comments +Can you rename the DFSDATANODEDOMAINSOCKETPATHKEY constant to DFSDOMAINSOCKETPATHKEY as well for consistency? Otherwise looks good +Problem was reproducable with the we had hereAfter moving from to the problem is gone crawling is fine +Hi KoushikI have updated the bug with my comments +Sample jstack output taken after the build has failed +I thought Gianny had put in some support for this but couldnt find a note on the topic +Oh sorry my bad for the confusion +Attaching a small test case +Can this be scheduled for EAP +Verified with JBDS +Do not use this on with large rows or with high writeread ratios +Id like to see cms included in all the CMS headers for consistency its in a random sampling of the headers nowactivemqutilProperties definitely needs to be exported for the DLL as its used in the CMS interface +The version of this patch is required for s unit tests to pass consistently +I created two advanced zones and separately enabled both in quick succession +Added the REST web service apiuserproperties the queries in Java ws client the ruby API for widgets JulienThanks for this info documentation updated. +Eclipse not wakes up even after some time +Joseph do you have a test that exercises this +see like there are a many different ways to achieve the same thing +Git Info of the Tree is as given belowcommit eaeabdfabbbatree fbdaceefbfccedcaddparent fefbcccadaaccd +v. +for this new contribution +The delayed validation is not an issue +So unless theres a reason why we must authorize per connection rather than per request it would be easier to authorize requests and would better compartmentalize the code +Can you confirm whether is building in Continuum in addition to being resolved from a remote repositoryAre there any other details you can provide about your project and infrastructure to help us reproduce the problemI need to understand how the local repo is getting populated in order to figure out how it should be cleaned up +was already fixed in in a different issue +Removing fixin I dont think this is a must for If it implemented in time fixin can be set when the issue gets marked resolved +Closing this issue. +It still has some nocommits mainly related to and carnality but I this its fine as a start +The URI should be removed or changed to the Attic URI +If no one has any issues with it Ill assign this to myself +fix merged to +cant comment on this no info available on what this means and WFK doesnt exist yet +Committed addendum to and trunk. +Committed revision . +In add S as secondary storage is also having scope of REGION we always treat S as secondary storage +There is a problem with which introduced a bunch of duplicated code including the component annotation pulling in all the layouts instead of just using the id and letting the factory take care of it +to trunk thanks Stack! +Its probably worth creating subtasks for the remaining fixes as they are all fairly large and unrelated to each other +Duplicates. +So this is a good changeAs for Avalon interfaces Id expect to move away from all such things when we JAMES +Again wasting hours of my time on the bootstrapdeployment +Congrats Werner I wasnt able to convince Bruce to includethe new CGI library to implement lazy loading +change to buildpackage scripts in r +Succesfully created application on development env and embedded mysql cartridge to it +then the person downloading would be responsible to do the now will close since I put in a solution in the commentswhen I get the final solution will submit BJI put a link from FAQ from Resolved to Closed since we do not use the Resolved issue status +seems that s dependencies are hided by s dependencies during resolving +the pathch contains also the implementations of and +Im happy to add these as new date formats +My tendency would be to do it just after a branch +AmosCan you add a comment to this issue whenever it happens so we can get a feel for how severe it +Reassign to a release that you feel comfortable resolving this issue in or leave as is +To match your example yes +Yes we would need the listed issues to be implemented before we can even start working on implementing vertice operators +Expected or byte longyou are passing a value to something that expects a thats my folly +Is there any reason you cannot add catalog entries for these in the IDE to enable auto completion +Hi George If there is a misunderstanding about the location of code please feel free to move it to another directoryDIGY +Canceling patch to upmerge and disable user checking if security is disabled +Maybe someone has opinion on this +Is this Issue a Blocker or Critical priority for EAP GA? If yes +This patch was committed in SVN and SVN +Ive seen went out so is imminent +In the backport I forgot to change this in the poms for the bundles +Lets get a test tests pass +Pushing to is a use case that will be enabled with the completion. +Verified and working fine +arent in the default maven plugin classpath +Evaluate usage of rather than in XSLT Too difficult at the moment + +Thanks Julie +Moving to EWS well have the jar and properties file at a shared location for tomcat and has filed a docs JIRA to document the usage with tomcat +We were using a version that was about weeks old FYI. +yeah that current code would break NTLM +In that event they would need to author their own implementations of +This patch implemented support to allow using key to specify user name pattern such that numerical usernames can be supported +Patch looks fine +Patches for this issue should be applied after patches for otherwise it may break other pending JDWP patches +After an update to the latest master the deployment with the also works for there is still this issue with the Maven Deploy plugin +and edit connection +I tested some scenarios and thing seem to be working +If you specify the number of parameters you specify the arity of the ctor +Does support special editing of Ivy settings filesOr you mean that having a pattern with organization in the settings file makes the completion in the editor of the not work +The patch is ready for review +test for this issue +As you mentioned if upgrade will require full stop of the cluster so trying to preserve old schema is meaningless because we dont have to provide migration at all. +How would this with a separate deps package work wouldnt one have to check out multiple projects and make sure they are in the correct relative locations +Looks good! Sounds reasonable and I like the idea +Following Igors advice backed out Desugarer change and fixed at r +Comment these users are all flagged as committers +The Preferences page should have a tab bar menu with two items Main and Visual Theme +Bulk close for release +Anyway you can get a stack trace +If you want to propose changes for which could solve this please supply a patch +But TBH I dont think its going to change much in +Thanks Kihwal you are right +How about naming the MIME Type parameter simply as It is already documented to be only used for usePuttrue so this is looks cleaner and more without sacrificing any thingWDYT +can we increase priority of this one + for giving this a Ill look into this. +The new issue for is over at previously mentioned is not an appropriate solution here +The case statements with magic number ints are a little hard to explainIf you dont have time to do the above I can rev the on all accounts +Currently it is only used to store booleans not as a pointerThis test passed for me for times and still continues to work +you can apply it and testRegards +patch will boot and communicate there is not sufficient permission to for +Jonathan as Yoav notes you are defining a method +smile the diff tells me but I look at the title of this issue and its confusing half of the code isnt related to this issues scope +patch for to branch. +patch attached I will also commit the binary sample documents if we decide this should go in +Release Notes set to Not Required as information to include in the RN field was not provided in enough time to allow it to be included +The same bug still you for your help +Help in resolution is appreciated. +crashed there is with headroom calculations when changes Im wondering if that was the cause here +addresses the problem in general. +We need to add a test case to verify Generate BARE java interface from a WSDL Declare a reference with the BARE java interface talking the Web to at revision +bq +This test case exercises the NPE +I created for the issue with the embedded AMQ broker +It does indeed exit if it cant bind to the fs +Thanks! +The workflow you describe strikes me as unusual and Im not certain we should allow the disabling of security features + +Hvae a and let me know if Ihave missed anythingHuge thanksCheers Berin +The attached patch has a lot of unrelated changes I believe but considering the fix is the one line change in then +actually clean OR closeopen projects both seem to work if there is a just one project but it definitely takes a clean if you have multiple related projects that changed +lol yes I forgot the issue +Merged also to branchesSendingSending Transmitting file dataCommitted revision +After looking at this further and a bit of experimentation it appears that the only thing thats really necessary or at least a nicety is a set method for +Please reopen with more details if needed. +Can you come up with a test case to demonstrate this? That would really help us fix it +Check if still present in with a test then resolve +Using Maven actually makes things worse +Thanks Thomas and Knut +the credentials section has been moved up +Can you provide an example showing a failure +type f name +Thanks for report SumitAny other volunters? Leon I saw you voted did you review orand test +Whoops I am runningMy bad +Ill have to look at it closer I didnt write the stuff so I need to spend some time getting familiar with all the internals but at first glance the sleep makes me very suspicious +Its a line change +Please find attached the thread dump for the hunged testsuite +How long is it taking up too much heap space and what is too much heap space and what is the problem of taking up too much heap space? Is a probable Exception the problem? I would rather not be able to read any information and catch an if i have to process an email that has a larger headers than i can handle than not being able to read any information and catch an caused by an if i have to handle a message that contains a header bigger than any arbitrarily chosen size that must be smaller than the possible size to take any effect +Closing. +If so we should document in as setting values in mm +I think its safe to backport the whole fix to as long as we remember to bump the version number on the branch +thanks as r in branch. +select as c from t group by order by doesnt workI can see ql now doesnt allow but is important for the project Im working on +I do follow your way of thinking and I do agree with you that you have the equals and hasCode methods to override when needed +I am applying this patch but while it will eliminate some NPEs which are not specified in the contract it will not make Trinidad compatible with browsers other then Firefox and IE +Ok thanks +I will finish work for today I am exhausted like the enums +Required unless the relationship is unidirectional +I have incorporated all of those +I have also scraped my source code for both Case Attachmentplace in acfexpandpathplace in railo Michael I do appreciate where you are coming from and my last comment looks more snarky in words than my jovial intent +I did find some lines with exclude but couldnt make much sense of it +Updated the patch to be similar to what Hadoop and doAlso added the default to the client as well +If there is an EJB module that was previously installed before the patch was applied then you need to reinstall it +We should definitely upgrade I also think we should upgrade branch do you agree +The provided patch does not work and no new patch is available +Could you attach your patch to the issue with a flag for approval of inclusion in Nutch? So we can test it more and include if all goes well +will do thanks! Ill let you know what happens +Please jump with comments and improvements +correction of formulationwith +guvnorinfo folders as and isDerived so this sorts of files and folder will not be compiled and not displayed in Eclipse client tools +I have not dealt with this and I dont yet know how to do so nicelyA couple tests needed to be changed to use a because on rebooting the writer no existing index was found in theStill mostly an exploratory ideasolutionIll post an ugly patch in a while but headed out for a bit now +I answered yes as suggestedwhich kills attached file for the best information on thecase where I tried to change a field name and it got an infinite loop +This is an excellent change and makes complete sense that it will resolve conflicts faster than the first way +Support of commited to CVS +Particularly in EC where IO performance is variable itd be nice to bypass disk IO whenever possible +Its probably not the me stuff that is causing the problem but maybe something else has been updated in the +KeithWhen you say it was fixed some time ago Are you saying it was fixed somewhere else in another branch or your local codebut not comitted to the main branchBecause the main branch did not build yesterdayI suppose the code is merged into trunck now check the trunk +Let me know if this works for others than me +Its not a proper patch +This is a duplicate of which has been fixed in and will be backported to. +Backported in r r +I have encountered this issue with yet another WSDL +This was affecting our benchmarks because this was causing taskNow benchmarks look really good +Talked to Bowen offline the error I was running into is a different issue and is being address by +Attached patch to fix verified and committed +It looks like the fix just didnt make it in time for ER Ill verify with ER when its ready. +Catching as seen in and handling them is okay +As is it can never be closed because well never add a serial version ID to every class in the product and well never know if the ones weve missed will cause problems +Thanks +Reporting all known deadlock occurrences with as much detail as possible would help but in the end someone needs to bite the bullet and start rethinking the locking behaviour withinSome people have workarounds that explicitly synchronize all the versioning operations of concurrent sessions +Ralf hmm I didnt deliberately modify any characters +There is already a JIRA for this issue +The lines look bright where the legends colors are more subdued +eclipseplugins +Do we know what the internal width here is? If this happens all the time I think this is more than optional +But you could stop and start it which would be nice to be able to do. +Hence and structures will remain non synchronized for communicating data structuresPlease correct my understanding if I understood it wrong also provide some pointers so that I can work on the issue +Looking at should it live in packageRightbq +Otherwise unchanged from aBoth derbyall and passed with patch b +Maybe we should have QE give it a try too? +bq +Until a long term solution is found would it be possible to get something in the documentation warning people away from? It was featured rather prominently in Eds talk so people may mistakingly believe that is the preferred method to create dynamic indexes +But having no choice Ill give up on defaults of course +d reopened this and had been leaving it open as a way to track in the JBAS project +The patch removes the custom code from and makes it source like the other scripts +Fixed in line with the comment posted +your screenshot does not say anything besides socket is closed +Yes Im sure the content doesnt contain the semicolon character Ive already tried get byte array from the file and store to other file it also didnt contain the characterWere using and defined in Chromattic side to store this data Ill check in Chromattic soon and confirm if the problem is Chromattic projectAnyway please help us if your team can its really magical if the character is not given to the JCR it is enough to investigate at JCR side +and you have to ask SM for the list of online servers to call find +Even so people not using and just using git directly would ignore the settingsRe Unix endings yeah I was suggesting moving to Unix endings +on the snapshot dated th mayAll of my consumers use queue browsers and sequence numbers to either subscribe from if they are new or x if they have been restartedIs there anything you can suggest that will help me get you information on whats happening to my instance? I can probably throw together a test case but as far as i can tell its random +Thanks +Scheduled for along with +Still need to verify in the original environment where this was reported +As the previous logic have issues with concurrency when multiple threads invoke the marshal unmarshal method +What I wrote to last patch applies here too +r +Hi ErichI agree with workaround one but what I did today got me up and running although it did make me downloaded the nightly Spring build and replaced the lognet assembly with the version +Might be a good time to run an automated link checker on the entire Tuscany site to assure validity +Screenshot of application in revision +Even if we should redevelop part of our I think we should create a sample application and check all of important things +releases +I think thats a good change to make but this patch is against the current state of affairs +Thanks for the help. +You can also delete the Ontology Manager Store from the release branch +Actually I wonder if these parameters should not rather be plugin parameters instead of rules parametersIn the stylecop conf files they are analyzers parameters and so they are applied to a bunch of rulesTell me what do you thing +Introduction section of the docbook. +No problem +Ted Its not a runtime requirement that client or ensemble be +Not sure we need an index for entryDeleted +Thanks Stan you will make Max very happy +I read the docs build from on branch issues I find +I just committed this +When a datanode restarts it moves all the blocks from datadirtmp directory to their final location +Since this is asynchronous we can also update the leader and isr path with the new controllers epoch on receiving a successful leader and isr response +That was a whatever its private decision +Do you have any objections +Ack no I didnt think to look in the jetty socket factory +Broadened the scope of the issue description +Committed in revision + +I have added two tests cases already they are in +I verified this problem in it works as the description I posted on Apr +Many cases has been handled based on the experiencestesting and also running in production from last monthsThat is also has State machine implementation as you proposedIf you have some free go through once and if you find that is reasonable we can take some code from there as wellAlso i can help in preparing some part of the patchesThankss GreatCompletely Agreed with you for completing manual failover firstOk lets continue the discussions on design parallely whenever we find the time +Whatever you did it fixed this bug +I have a book that lists the SQLSTATE codes they use if you cant find that online +The applied patch does not increase the total number of javac compiler warnings +Please verify if possible +close the issue +I link this issue to which addressed the analog issue for table level privileges +That code looks fine but it is still missing something +I updated the patch to make use of +updated to HB to update to latest Hibernate CVS to get the EJBQL changes that were done +PK of e attach an ear with sources that has the setting disabled i can deploy the we cant reproduce this + looks good +is part of +This actually is not a minor this value truly global? It cant be set in our specific calls to be higherif its truly global then providing a link and even better a direct way to increase it directly in the flow of our wizards would be better +we should pursue this and figure out whats going on since we need these tools to perform reasonably well +Woops Robert you are right +if there is acceptcharset or multipart standalone +Applied in and +DirkDo not mess with the headers +Is there a reason? I have talked to contegix who host Compass and they say that rsync should work +That is not Forrests jobWhen your graphics expert has the replacement logoif it is still named then all thatis needed is to replace the image atnutchsiteforrestsrcdocumentationresourcesimagesthen the siteIf it has a different filename then declare the newfilename in nutchsiteforrestsrcdocumentationat the elementIf your team is not able to tweak the logo imageto add the trademark symbol +As I had previously used this adapter the tables were already in the database +This isnt the case with Gradle as far as Im aware so anything other than a maven solution leaves us having to update IDE dependencies and build script dependencies independently +Of course it is +It throws exceptionsCheersBob +Otherwise I will. +At the end of the month individual values may be collapsed into a single column +You cant control the directory that way +No more features in central or extras site +I have committed this +What do you thinkThanks! rishi +fixed in RT +Is this done +Yea I guess step is the real security issue +Updated the patch file to remove the prototype of from the header file and the comment of it ever existing in the +Since theres only a single instance of the plugin in the agent it doesnt seem clear to me without serious intervention that there is a way to support this +As an aside please make sure not to use tabs for indention +Does it really migrated +fccbbfafcda Reduce noise in Kafka server logs due to reviewed by Jun Rao +Raghu and Hairong yesterday raised a few relevant issues the superuser name is not currently known on the client and until it is we can get false negatives in permission checks the goal of false is for admins to be able to set examine and alter permissions before they are enforced so that a filesystem may be upgraded and returned to service before permissions are completely configured +I think it would still be useful to have a test that uses with data in HDFS so perhaps the patch could be changed to do that +Playing w Jackson making it understand our HTD and HCD would take a bunch of ugly work +For proxying we need to make sure that JBR http gateway and handle these in a consistent manner +this was an initial attempt at a loading status indicator +Added the flags to the newest snapshot codebase +Then the page reload and checked the snapshot status +I think the patch is ready to be committed +Attaching a patch which adds an additional test to jdbcapi +I agree it would be nice to not always include the method if it wasnt necessary but it should probably be done on the branch to preserve backwards compatibility +Stay tuned +Havng read through the above I am still unclear on what the best workaround is or if a fix is in progress +It would be better if you posted the code snippet for which doesnt work +Better error messages would be appreciated if possible +moving fix version to since we might just put in gwt support and refactor this kinda logic out of the actions which is where it is right patch for this in rev +RickThanks for putting these in +er just to be clear it will compile under Windows when we merge. +Probably it duplicates the logic in the engine with the intention of having two separate implementations to better tests the functioning of the engine behaviorI believe it is used +Certificate installed. +Let me think about the side effects of patch v +Yes this combination will not work as things are +If that sounds good Ill open a separate JIRA for it +Applied qmxs version +This is the transformed WSDL with the missing namespace +There was also source control cruft in there from some other source control system that I had to removeIm not going to have a chance to look at this in any detail until at least this weekend so if you could check out the branch and let me know if it seems correct and seems to build for you that would be great +I was saying this because I think the use of is discouraged and in JS and JM there is quite an extensive use of it +Provide a patch pleaseThanks +Is the test appropriate or still need another +The patch is incomplete The skip operation still uses the +This constitute that the writes are made k at a time thus the buffers suggested here should be no more than k thus cant possibly lead to the memory leak. + to trunk +Will be committed together to ensure build does not break +The cron job will be kicked off twice a day from this point on with the first update at AM Arizona TimeThanks +At least now kernel tests dont crash any more +Attaching the zip file for the Test +I changed to and the imbalance lessened +Reattaching to get another run +Does that mean we should be forcing people to use a by throwing an exception when none is configured? +comment +The class is now the instantiation of +json and cannot receive a banner headline +Currently there are rework underway mainly relating to resource monitor etc +I think the best solution might be to allow multiple views or have a tree with multiple server roots +please assign to me thanks +See the attached +This patch seems to fix the problem +I spent some time trying to fix the test today but no luck so far +Moving issues that arent going to make it into to update yes +on first loop +Test cases the to deliver callbacks on a separate thread from the timer task thread just in case the callback handler hijacks the thread on the callback +Rerunning with latest +Created interface for Transfer Rate request to follow pattern of Concurrent Login patch +Ning in case you are hitting the problem Paul was tracking down with ls can you run his test script on your dev box to test it? My dev box was affected so I found another one which worked +Im not sure what to do with this as I can get it to fail and tabbedpannel produces an output almost identical to the one suggested +Alternatively you can implement directly possibly using BCs classes and parse the name fully +Pressuring vendors to use a standard is a good thing +Incorporated comments from Raghu +So it might be best to wait a while before submitting an updated patch +bq +JacquesI wouldnt say that the is deprecated there are probably situation where putting the forms in the context by declaring them in the screen definition wouldnt work well +Hi are very welcome upload your patch and Ill review and see this class but need an idea how to configure it +I believe in JBC would blow up if you didnt specify a default region and used nodes outside of your defined eviction regions +For listIsos and listTemplates command current search logic will change with recent storage framework update so this should be revisited after that is done +Denis I will be very appreciate if you find a time to check all this stuff +One thing we need to be very careful about it is features that need the same bytecode manipulation +You may need to make accommodations on your specific machine to allow all tests to run +Just to make sure I will try to change it and then run the test suite tomorrow +Im facing this problem tooCore Tomcat Spring ve tried last suggestion but I dont obtain any in branch and trunk please note if you want to use this please keep available in the classpath otherwise the default property will be to beat a dead horse but now that I understand the problem I like the way this was working before the bug fix +This is because of the different structure of the master to the indices indices have a handle on their record manager so the sync operation on an index commits the record manager +Upload third patch and measured result in this trialDifference in the patch is that new patch take nullbyte into consideration when calculating size of memory to be allocatedThe size of nullbyte is subtracted from memory to be allocated as bufferRegretfullythere were no difference in measured result between before and after the patch +to and trunk +Committed to two branches and trunk +I replaced the provided libtool with usrbinlibtool + +Are we closing this issue or moving to Beta +Could you better explained your need? Christian do you have Javascript errors +Excellent cant wait to give it a whirl +stack Ok +Could anyone review this as soon as possible +so if we cant do that we just need to locate and read the manifest +Not sure it is worth backporting tough +Rethrowing no cipher suites in commonI see the above times during renegotiation attemptsWhat is odd is that the cipher suite lists on each side are the same whether keyalias is used or notI am attaching log files for both cases +this be interacting negatively with? If the exceptions here really are innocuous we might want to supply some defaults for and which manage to exclude these from consideration when calculating if the shuffle exception threshold has been reached +It will not be easy but I really hope someone will be able to pick it up where you left off +Apparently theres no debug info compiled in so I cant even look at the variables to see what the deal is +Assigning to Justin he developed support for per workspace +Could you please encode all queries in unit tests to use double for meLets create method to do that +Thanks Mamta for looking at the patch +Patch applied. +v attached to generalize fix to other +ij runs a system shutdown on exit so simply exiting the script like in the repro doesnt actually cause an unorderly shutdown +GC count reduced to GC cycles during the test +Then the decision of where to put it or can be left until later. +Committed support for casts in r +It was a throw in feature +thanks jeremy +also can you edit my initial ticket to remove my domain name +demo to adhere to new demo kar design +I just committed this +But apparently this is not possible in jetspeed unless the user creates a new portal pageThe data I see in the DB seems to support thisI login as user A and I have a portlet preference for one of my report portlets called report +Thanks Mathieu for your work on this. +about need to have SYSREQUIREDPERM system table +Ive also created a patch to address this issue +If anyone feels this change requires a full version bump I will do it but I think it may not be necessary in this specific case +The problem is reproducible on my machineThe timeout for the session is set to sec in the +Patch applies to both trunk and +The developer setup is superseded by the general UIMA developer setup instructions +Heythis was certainly not noise! We really appreciate those kind of reports +Rolled back changes that created a file r +Note that the location of has changed as a consequence of the implementation of the consolidated trust store +You would have to write a small amount of custom code into Archiva but it isnt exposed as a user level feature +We have two choices suggested by Yonik update the wiki with the missing analysis components or think about switching strategies to pointing at generated +Marked deprecated because this will not work together with button controls +Now that work is well underway with AS all previous community releases are +Please test if it works for you and solves your problem +Can one integration test be added to the patch +No you are right this just went stale + at +Strong can you take a look at this test failure +Thanks just waiting for your patch now +i like that idea but i cant follow that because i am not doing this at my endi dont want user to especially set my implementationi got another idea though i dont know if its good or not +the same for hibernate guide +Additionally we need to build javadoc and source jars +Any suggestions +Its the simplest solution and Autobase is due for a rewrite to use the updated Liquibase engine anyway. +In it always reproduces for me +Yeah i think soBtw Ive started writing the jaxb spec and will also need the jaxb spec so currently they are in servicemix but it makes sense to move those to geronimo along with the other ones at some point. +Thanks Hyunsik and MinI am excited to work with all of you on Tajo +Triaged July Assigned normal urgency +If a line of code is involved in several different duplications this line is going to be counted several times +Closing all resolved issues with a status. +Thanks Yongqiang! +But yes this removes a checkAs I dont want to decide this by myself I guess it is better to ask the listfor a correct handling and so a community decisionJrg +I did check for regression on the other Jira but thats not the case the test example attached to that one runs smooth +Unfortunately conflicts with this patch +Meaning that if the original variable were not declared as a variable but rather a gettersetter to begin with I am guessing +Consider merging to? Seems like a change and will help with merge conflicts +The console requires Axis for the UDDI client code +Some additional information contains in we can trap that or is that purely caused by native code +I guess by threads that loop indefinitely thats what he means too +Deferred until dependent issue is outdated did a little deeper investigation and it looks like is actually recognizing the configuration but it fails to write the correct endpoint address in the WSDL +Fixed with commit rLink now points to network bridge page for networked consumersIf you can check out the next nightly distribution and verify that would be great. +Indeed this is an option +This is the initial drop of the Portlet Bridge for Portlet +Even though its less convenient than if you want to use and you can still use and by definition you wont have too many of these for assuming and by default have a look or track +bq I also think that this may remove the need to deprecate persistence its a lot more maintainable nowI sure hope so it was pretty ugly for a long timeBut the deprecation also has another goal moving all the core definitions out to discovery mode rather than this current hybrid +He owns the mailing list +If you have not resolved the problem please let +But how do you apply that to the first composite model with the sparse fields +I just committed this. +If there are some remaining warnings I would appreciate it greatly if you could correct them yourself so that we can close these issuesIf there are still tons of warnings that you dont feel like correcting I will do what I can to fix them but hopefully this is the endThanksJared +I will modify the TC accordingly +Fixed on trunk revision +Simplified the boolean expressions +Compare the raw value from the original Configuration with the raw value from the cloned Configuration +Im marking this as fixed for +Making a feature the xsd was wrong here +But afaik there is a you can use in gradle as well as from the command line +Anyway the test now runs for over minutes Ive attached the patch to make timeouts shorter at least it takes about half a minute now +The documentation looks good +Clean patch for this issue +Moving all issues to the CR +While its ok to aggregate warnings we should switch to stopping on the first error we find +It includes EL code completion validation refactoring etc +Hi there The copy lib samples in srcxercesc is notsynchronized with the change made during the implementation ofsaneinclude and it is not consistent and some how misleadingthough it is no longer part of the build processRgds +open daughters on the parents regionserver +I think in this situation the correct action would be to halt early +We have been able to implement the file attached to this ticket +I wish I could mark Abandoned on this +Not done yet +Hi Cheolsoo find attached a revised patch as per our previous discussion +The milestone is now available in maven central +Corrected patch is attached +License this for Jeffs contribution +So FOP dont respect encoding in embedded fonts +bq Why arent we also changing ROOT and the first META? Im afraid this will come back to bite us when we actually try to fix meta splittingI opened a new issue to bring these over too +I did not realize that XSLT core functionality could and is implemented through the ICU LibrariesThanks +the same fix should be made to the getHeaders method +Thanks TimApplied as expected. +Thanks for the patch Clint +Reopened so fix instead ofOr should I rather open a new issue +actually this join query is wrong +The Activiti Modeler is not shipped with Activiti anymore +Trivial patch that fixes the links +not many databases handle that high number of in +This issue is now occurring in Geronimo Aman I could successfully send messages from a standalone client as well as from a was no exception while closing the session or connection +I filed to fix dead links in HDFS document and I attached a patchI found no dead link in YARN documents and just a few dead links in MAPREDUCE documents which are missing Go Back. +Its not usual to specify the umask in octal +Java was not my issue as I am using C and CPP +Ill toss this into +Half of us forget an issue after JIRAs +There is code in that was related to collecting statistics for MVC controller invocations that I did a while ago for the conference +which jboss are you running this against +I think the instructions need updating +Preferences Groovy Editor Templates now allows you to tweak groovy editor templates +The interface for does not define +Actually lets see if we can squeeze this into you going to get to this in time for +So the large files will impact the performance due to the transmission of files to all nodes +Adding default proxyWizards in central in master and beta +Committed revision to close for +Something likehttphostportsetlogleveldebugwill push that server into logging the class at the debug level +Vieri thanks for the patches +Moved to upon release of +Is it possible for me to disable it without commenting it out from the pom fileOr better yet just stop it from when I run the quit commandMy main problem with it is It performs a commit without me asking for one +In this patch Im just adding test scenarios to cover this specific JIRA +On second thought just allowing any decimal is fine to start with +Add Deprecate annotation to remove warning introduced from the javadoc deprecate tag + thanks Alessandro +addressed an issue with the unbind case not correctly being propagated +just committed this +atomicity it is outside the initial scope of this JIRA +To Xerces the SAX filter acts like any other SAX handler +How about startContainer passes the app token that will be used to later authorize stopContainerget +Closing this because the basic work has been committed and there doesnt seem to be any more activity with it. +fixed in as i wrote in the issue description in the schemacolumns table it is a class data type in the create statement is timestamp as it is and should be in the jdbc context. +Good point Yonik +error handling had been improved and this should have been fixed please it if still on build Passed +It shouldnt break older versions +Otherwise +Oops +Setting this issue as depending on This is because I do not feel comfortable committing this code in trunk it is too complicated thereInstead I would like simply work it against the flex branch where we can make it flex patch after and updated the patch because of my last commitYours looks good my change was only adding the method param and removing the access to the noew private tenum +Reopening because it is not working correctly +I suspect that the problem is related with the name of the directories +Replaced it with a generic class from how the analysis time is aggregated and reported for JMX +btw closing a is a so you could keep it simple and make it volatile +Thanks for the feedbackCelemens I in fact considered the name But in the end I went for Engine because it is the main driver in the Sling Engine of the Sling Engine module +I would say yes to delete cause it wasnt tested at all after much refactoring work happened on the code baseIt is better not to include it +Or If it hurts when you do that dont do that null is not a valid XML characterId recommend marking this as INVALID or knocking severity down to Enhancement since if we were going to do it in Xerces the right solution would be a separate DOM implementation for use in untrustworthy contexts +Try to reopen to see if i can change this the affects version to be since thats what affects +The +What OS are your machines using? All the same +Attaching and +Is this something that canwill be resolvedAlso what happens to the existing repos? Will they stay around as forever? We are wondering if we can go back and look at that if we find something missing in Git at a future dateHow is the External git repo lookingThanksAny update on thisThanks Ive been traveling a good deal this weekendweek +Ill make the change in the patch for. +Ok so Im NOT using the Vista special just the regular win version of SWTFYII also tried running with the IBM VME and Harmony class libraries and Eclipse runs ok with that combination. +Im no longer able to reproduce this +Youre right Brian and there is already another ticket on that subject +is now at Roshan does this mean groovy will behave now +Patch for YS +I could fix this in trunk yes +I think I will keep insults to myself +It seems it has been used in many Java codes and scripts +As the GC starts to run the producer on route is reclaimed and a consumer is leaked +Thanks Eric +looks good to me Is somebody volunteering a patch? If not Ill create one +Marking this as an enhancement for +can add refresh for a refresh hyperlink to trunk yes please reopen if required. + updated +Btw which version are you using +Final patch for review +Unfortunately it gives the same error +It may be possible to generalize the code to handle arbitrary recursive structures but Im not yet familiar enough with the codebase to attack that problem +Detection by type wont get us far here I suppose since those trigger beans will almost always be of standard cronsimple trigger type across all schedulers + +reopening to change the component to Module API +At least as far as i can see +adds a basic aggregate as list strategy class and test thats frankly not intuitive to aggregate using a constant expression and all into a list +Yay! anyway +The workaround is to just use distinct source names +What we need to do is move to but thats probably not going to happen for a while. +no more features making it in to block GA +class files go in a jar +Item assumes a certain naming convention that the majority of apps probably do not follow +A couple of small changes are needed to make it work +is some how not loading the class +Youre right will try to fix +Request and response for the volume attachment issue is already fixed in the Abiquo fork but needs some cleanup before I can actually send the patchIll try to have it merged for though! +When running derbyall this test failed when I added code to throw an exception when ended up with the harness configderbyallderbynetclientmatsderbynetSeeI didnt run the compatibility test so it may be dependent on the harness code in still +Thanks a lot for the review +Please look at this Torsten +I never got a response back but I believe this fix to be the correct one so Im committing it +Attachment has been added with description has been added with description AFM encoding has been added with description patch in r. +The problem in TM implementation which consumes a lot of memory +bq +Since this is the only way to get the new revision I think we should merge the headers here +Or maybe this feature should latch onto minversion being so that we dont introduce additional parameter +IMHO the easiest solution is to change the internal representation of the float to +Except is located under lib all other missing files like are located under usrlib +Why would this work under and not? Additionally if you examine my latest sample youd see that with you get an that describes not only indexed write and read methods but regular write and read methods +The ab prefixes are due to the patch which is why we are using p instead of p +How does this block +Closing the issue as version has been released now. +and solved +Nah might as well do it in this one +Sun vendor version +added constants added handling for IntegerShortLong and intshortlongcleaned it upextended to self The attached diffs do not account for DTD nor HBM did link this issue to +Please verify and close issue +Attaching files in agree +Peter what runtime and runtime version are you running against +Fix error of in in rPlan to enable it after enhancing can with parameter list +Bumped project version to hibernate to +the conversions so far have been very straight forward +attachment included to help recreate in recent code +Here again it does not matter if if etc +Patch with added javascript code to and files in master and slave webui css and js files need to be placed in mesosbinwebuistatic while all the png files need to be placed in mesosbinwebuistaticimages for the data tables to be sortable +Even wait X seconds between bootstrap operations is not a prereq I am comfortable withUnless the above is incorrect I think we should bite the bullet and fix it right +Im not here for support the docs are saying it is part of cordova and not a separate pluginThat is why I asked if it was a documentation error or if it was accidentally removed +insert +Hi Benjamin Sorry I missed your memory settings in the new +Is this an improvement or a bug? The summary sounds kind of buggish + +Turns out the test webapp Id gotten from a customer had a copy of jgroups expanded in classes! With this gone the problem went away. +string describing the SERVER the long run it would be great if we would have a Details linksection that describes all the used versions +Patch includes implementation of string matchingThis implementation uses the optimization to compute partial matches of and adds a prefix of size so that the first character is contained in the same number of as a middle character +distcp must turn it off before submiting the job +or doesnt matterThanks Gary. +after searching more deeply one way to confirm the issue is to check please try to use or of they should not cause the problembut or cause the problemone workaround is to change the plugin versionanother is to avoid dependencies far as I can tell this is a consequence of then a duplicate of if youre still having an issue when removing MPIR dependencies report please open another Jira issue with a test project to reproduce the problem +Nice refactoring as wellIs there any advantage of sending out the masterslave counts in the response as well +I would prefer it either DEBUG or a better error message +Is binary compatible with? It just seems really weird to me that uses aAnd yes I think should also depend against when using a profile to maintain consistency across our other active branches +So the file still has the old name +The patch makes getKind synchronized +Definitively not empty +Task Define ISAPI Virtual Directory Step Missing +I would suggest to make this configurable +Ah great +First thing you should do is create some unit tests otherwise any fix can break at any time without warning +Sorry for the delay +is the unit test code that tests the with different zip files having different number of entries +Initial patch to replace the old file +Would you please try this patch? Thanks very much +Thanks James. +But you do create anobject of that class +Good point about running different client and server code +Verified by Mark. +Theres a deeper problem with Lucid that needs fixing first see +And there is a class named many also need change given that NULL is renamed to NONE +I think this is failing because of timing out at least in recent builds Ive seen +Then wed need documentation added to describing how to publish maven artifacts and to convince those that produce Hadoop releases that it is worth their while to generate and publish them +Does signing from Eclipse work? Where did you get those commands from +nodeid is now mandatory and the instanceid has been removed as nodeid is unique. +Looks good to me. +Groovy bug +Ill be scouring our logs for the debug message you mention +With the commit Im about to make a large part of this is solved +Keep it open Id say +I think it probably makes sense to just add a long get function and let the JNI layer take care of the details +Closed as not reproducible +what about we keep the test as it is and set it runs only on oracleand we add a new test as your purposed to run on others +Assigning all documentation related issues to final since they dont represent a change to the to be already fixed +Sohil Wont it be a good idea to use new hibintegration libary with this as well? There is not much change that would need to be done +sheduling for next rev + is the same issue as +Thanks Wei! +This project has retired from Apache and is not supported by the Apache Foundation anymoreThe current source code is located at zetacomponentsYou can get support on the new mailinglist available at zeta +bq +The fix fixes this +BTW you need to svn cp srcjavaorgapacheavrofile srcjavaorgapacheavrofile before applying this patch +I think it makes sense to port it to +Using either the with the or using with Named were both can be considered solved +Fixed on cellar trunk revision Fixed on revision +Maybe a path issue? that forward slash isnt very +Wow + +This issue has not been picked up please review and see if you can help fix the blocker critcal was created before July please review and resolve we are approaching code freeze in days +just committed this +I am experiencing this same issue and I can help with testing +I excluded the logging change since will take care of it +All test failures are fixed +I added a fix to enable the connector without any index defined in we have to discuss if we need the check for the index specified in the configuration or not +Thanks for taking the time to run these experiments BogdanAt least now we know the problem is not caused by or and we know that it must have been introduced some time after and beforeOne thing I know was changed in that area of the code in that period is that started using +Otherwise it will be a surprise to them if they find out some region is not assigned and not shown in the web UI +I dont know why this was created twiceSee meant see. +ve committed this +For now I think Ill just apply your patch and hopefully one day it wont be a problem as I refactor my way through the Ruby libraries +Added get accessor in +I have verified this +I have also attached the updated regression tests +not so goodHibernate pleaaaase we want a! Would really like a bugfix +The above suggests there will be a jca ldap and text connectorconnection +reattaching the file after renaming it to so I can flag the grant to ASF button + +The addresses example is actually correct with the assumed bounds it doesnt count the network and broadcast address as usable +Looks fine to me +all files and directories will be owned by the super user and super groupThat seems fine +fixed in to set fix versions +However the latest patch needs rebase +I think it looks better in caps thoughTested more on Windows today yeah I just want to be sure about js and css +Thanks Vinod! The new patch addresses the commentsbq +Is this progressing somewhat + +do you have it around +Is it possible for you to move up to a newer version of Xalan +on all branches +Let me try coming up with a new patch +In r removed required changes have been done closing it for now. +Ill file this separately +But anyway this is standard Tapestry behaviour +This update to the features assembly descriptors poms will move the demos folder currently being packaged in SMXHOME to SMXHOMEexamples +The Supplementary channels are listed for and the sun JDK is specified in there as well +just thought anything more was a bit verbose +Fixed in both old parser and new parser +Was wondering what the status of this was as we are using crowd and I would like to connect via this plugin inside of jumping through hoops with the acegi current version of the plugin will work with Crowd however I am in the middle of updating the plugin to use the client libraries +will track individual tasks underneath +No new tests are needed since this is a simple refactoring +I would need to look again since last time I checked the maven plugins were not doing things well compared to the ant support IMO +For projects once the CDI facet has been installed if you manually disabled CDI support it wont be reenabled automatically even after project update +I think diffs are ok then +Currently all IO in all tokenizers is broken and buggy and does not correctly handle special cases around their bufferingThe only one that is correct is but at what cost? It has so much complexity because of this Reader issueWe should stop pretending like we can really stream docs with ReaderWe should stop pretending like GB documents or something exist where we cant just analyze the whole doc at once and make things simpleAnd then we can fix the lucene tokenizers to be correct +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Issue makes sense patch looks good to me. +Applied patches in r +Its not out of the question but we moved to this approach when the graceful one didnt work as reliably +I think weve got enough information in this thread that we should probably all hash it out in front of a whiteboardDid the above hashing take place +Output goes straight to STDERR +fixed in the schema +I committed this with revision +Rescheduling to since they havent been worked in a while +I wonder if we can configure this something like the console so that we can have a generic war and then tomcat and jetty specific plugins for easy installation +Wrong +I am not building the console I am using the one that comes and is configured with GA SP new JDBC driver upgrading JDBC drivers in seems to be hard I matched the driver versions in jBPM with those in EAP +What would be good would be a simple example project showing what you find currently missing in the for the Maven project +See comments for details +Reporters of these issues please verify the fixes and help close these test cases passed in last run closing this defect +I also ran into this bug +One point of samplers is to save time so adding function calls is silly +It is legacy now +Im a little concerned that this ticket is coming in so close to the release +That way on shared zookeepers all the content can be dropped into giraph or whatever +Ive divided the task into two pieces delaying the application storing operation and adopting event model for the application submission +to fix Field +Lets leave it in for now +No I have not demonstrated that the cost string concatenation in debug existing log messages is insignificant but neither has anyone else to my knowledge shown that this cost is significant +The challenge with this kind of change is that we break the strict DAG of the execution plan +I am not sure what happend to hudson + for the patchI have a patch if you want it dont even need to ask about patches they are always welcome! its like asking hey do you want a icy cold beer there is only one answer yes of course! +We have no test exercising currently +Could it be added to regression tests? +It will change current behaviour of the framework and users can be badly surprised +Theres already tracking and explaining this +Or perhaps it is being corrupted during the Ant +A small stylesheet that strips out the original indentation with the xslstripspace element so that it can be replace with alternate indentation has been provided +Revision is a commit with a fix to a javadoc typoparameter targetTD should have been target. +Committed to +Hi JohnDims just applied your patch to +I think you are agreeing here tooNext Patch I will put the CP hook also +the descriptors created are completely functional and only differ in +Reassigning to webapp category since this bug only seems to affect the live Cocoon webapp not +I rewrote the method to do all of this in one loop over the message so it should perform better than a string split and a separate processing step for the other characters +Do you have a Linux build of your that I could download and try without having to rebuild the whole tree myself from source? Thanks +batch transition to closed remaining resolved bugs +The problem is a name clash between the url parameter locale and the Locale object that is kept in the service contextI changed the url parameter name to newLocale +Mirceas busy with the transaction recovery stuff so Ill look into this tomorrow +So the fix for this exact issue is to specify sourceenvBody and wrap the XSLT output into a envBody element which will replace the original body back again +We have to have different deserialization per znodeBasically if its a leaf under replicationrs we print a long +Steve I understand this is probably a bad database design +I believe they were killed as part of the shift to the new UI +Please regenerate the patch +Uploaded a new patch to fix the findbugs warning +Since you found a solution Im knocking this down from Blocker to Enhancement I agree that a hardwired limit is undesirable but it sounds like were working correctly within that limit +I have some vague ideas that at the same time would allow us to plug in different compression algos but dont hold your breath +This is a big maintainability issue for me and would like to see this bumped up higher than minor +To me it sound like this is then not sufficient but is at least a step further +Hence closing the bug. +I think these inconsistencies arise because the EAP build has its own logic of packaging config files using the Dsoa switch and BEGINPROD ENDPROD vs the messaging login config change was incorporated by other two might be related to you please give me download links to the zips of the EAP? I think this was caused because I was trying to build from the tags and ignored some differences that I thought are related to test build issue with jmx and web console sharing the same security settings has been discussed before since the installer asks only for one usernamepassword pair it wouldnt matter if the security domain is different +read and write writes balancing properly? i think childrens from split can land in one regionserver? and what with increaments this is special case because this type of writes doesnt trigger splits so I did in is to rebalance newly split regions onto other region servers +I dont see any problem with Alessios approach +The new attribute is called +Lets see what mdobozy says about Firefox perhaps +For now this patch is just fixing the incomplete cleanupIf you think its useful Id be happy to go through the tests and make it more consistent so that a temp directory under target is used with cleanup of temporary filesdirectories only being done as part of the clean target +The patch looks correct I see no issues +Fixing log level +Nice +You can commit this seems trivial to meBut please let this open we should take care of the things I mentioned before +There was some done on this front in the past some sort of Ivy based deps generation for +fixes. +new patchrework the loopsrename copy to cloneit doesnt use did you try rather than +Revision +Thanks for the feedbackSo I took the liberty and committed the patch in Rev +The patch is intended to be applied AFTER the has already been applied for brevity +The code cant return the full path since it is different for each keyspace but the directory name under snapshots should suffice +dhruba In your setup are you going to be setting to MB? depending on the production cluster response sizes and patterns we may have to set the resp buffer size to MB +Thanks Ravi +It equals CSS Class Dialog +Svn history shows that gabriel was working hard in the code base in september and probably introduced the issue at that time +for ve committed this +YARN uses a protocol buffer to do this +Thanks a lot Eli for the review +thx +Thanks GrantI will begin step in a few hours +is just a glorified and has exactly no performance advantage to a correctly presized +It wont download the artifact unless the user removes it from a local repositoryWe could add artifacts to the me indexer but this indexer requires the projects context that doesnt exist when debugging +I am reviewing this patch +Maybe we can put it in docsexamples +Update in kit release notes +But this worked out just fine +could you provide set and set for the two newly added settings +we should warn the users upgrading to +I think not but this task is about to rename packages +Resolving as invalid. +Sorry I mean issue error happens when two methods with the same name and argument types are defined for the same interface +when +Applied +fixed chars error in precommit buildfailed test case not related to this javadoc warning but its not related to this patch +Tom you made me curious +We might need to add the generated vector aggregate code to source to avoid eclipse reporting errors +Right now there are version and latest two versions of in and everything else is in module. +Switching all issues that have been in resolved state for more than one month without further comments to closed status +See Ramkrishnas question onIn this was disabledIt is enabled in the code was committed in even tho both and have the known issue that its disabledSee Todds email titled Two requests from a grumpy old man if we start changing fix versions like that its gonna be a nightmare for those maintaining those branches +Thats why we have lost the fields State and Name +Why are you changing EPPs producer configuration if its being used as a Consumer to Liferays producer? Also why are you mentioning a LR consumer if the scenario is LR producer consumer? Please clarify +Hi thank youI tried to verified it with telnet as well and it is able to connect to the address on the port without problemI do my tests with and without zone id passed as parameter of script but it has no effect +Thanks for the help +When I invoke any tooltip and press F as focus tooltip of Filesets is shown +David can you please look into this? Thanks +Cleanup of unclosed issues. +This might even be according to the spec though this particular case is not covered as far as I remember +Thats great thanks Lewis +is definitely allowed and is allowed but there are restrictions on its use and well need to refactor to properly understand how a may be used for path parameterization +Again the failure mode is for classes in +The latest patch seems to have resolved the problem. +Verified OK now +In my locale IBM vme returns GB Drlvm returns whereas RI return GBK +Therefore we cant use this key combination but what about the following proposal? F always opens the corresponding element in the xml flow definition file CTRLF navigates to referenced elements like actions bean actions and subflows WDYT +Almost there +If so pls tell me other patches are needed to apply this patch castor needs to be up to date +that is I think only for traveling persons really critical +Then update the provisioning modules nodestatus routines to use this instead. +Thanks Ankit! +I think we should reopen but leave it without a version to be fixed in +Its an and shouldnt be applied hereThat being said the patch for the fix for should be applied but it should not be replaced in or +I am now using custom runners on of my test classes to pull this off +Patch for to add the objects +will remove the findbug issue +Hi VinodThanks a lot for reviewingRegards forgot that Jenkins wont pick up the same patch again +Fixed the broken code after Checkstyle update was done +Hi Alan only the attributes that JPA marks as +resolving as WONTFIX since theres nothing we can do about it in +Done +I opened. +Started going through this document +Also research possibilities to create Carbon bundles +Kenney please file a new bug. +I deprecated globPath in this patch and I changed the semantics of globStatus which returns null if the user supplied path does not exist and returns an empty array if the input path has a glob but no path matches it +With this should no longer be an issue. +But we got to know the problems faced by various bugs in Maven as its in its early stages +Any chance this can be fixed for +Hey Catherine my GIMP is weak +Closing as wont fix +I think that was the case with this change and Ill flip back the ordering +issue not reproducible so resolving this as fixPlease open seperate issue if the new issue still occur +I bet that this bubbles up a bunch of complaints complaints that are legit but that we just rode over clueless in the past +The admin has to copy pighivehcatsqoopdistcp jars to the sharelib whether it is local filesystem of Oozie or hdfs +Patches both and database setup on init succeeds and database tables are present +Works as expected +Nice to have them verifiedI think Ill have to submit a patch to the to support an additional method for identifying artifacts that are part of the current project +Perhaps there is an issue elsewhere in my configuration +Apparently Eclipse is parsing the path variable and uses the first JRE it locates in the path rather than using the JAVAHOME env var to determine which JRE to use +So replacing the jar didnt work in my this change be done in the UI +Introducing the thread fixed test is +what is the data bind you usethe problem here is you have a seperate element with the nameas an alternative you can use p option to specify a defferent package to fault and web services specific classes +prefixing json +This is not a problem for our current release because depended on +Also moved from coresrctestresources to sharelibpigsrctestresources so that it is available in classpath +The issue you point out is certainly legitimate but its a compatibility issue to fix it now +I was wrong here +Is the correct +I have no idea about that one +Documented as a known issue in EAP release notes +keeping open to know +Not finished yet +We already have and +I uploaded a patch that modify in pdfbox and in preflight +Hi RegisIt is OK for me +Attached the patch that adds the subj featureChanged the root stats facet lst name to fFieldfField so that we can have multiple parameters for the same fField +Should be fixed with newClose this issue. +See for details. +Integration test added +Simone Bordet does not work on this project. +Fixed with r in WicketWorks fine already in Wicket. +Attaching the patch with the fix +Other services like RMI and JMX have already been refactored in Spring XML configuration. +Patch applied for the first issue +Hello both of youI slept over it last night and came to agree with Adam +I also question whether this needs to be public +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +Release completed +It seems like only parametrized object with collection type T causes the problem other parametrizations work +added localizer for messages visita takes a list of graph edges traversed while walking the graph in DFS list is initially empty always contains the edges walked to reach currently visited nodeb the cycle is calculated for back and forward edgesThe cycle always exists for back edges part of path back edgeThe cycle might exist for forward edges Reasons for forward edgesi The graph is disconnected or ii Threres a cycle +I vote for using exceptions internal to the to keep internal interfaces clean and elegant +Moving to is the patch files for s patch has been applied and checked a gadget from page to page works a bit too well +bq +run speed has increased + +I think it should be safe to avoid sending the stop replica request if controlled shutdownhas not completely moved leadership of partitions off the shutting down broker +Thanks Alexander the patch was applied to BTI branch at rPlease check that the patch was applied as you expected. +resolved under after release +Thanks for fixing some of the bugs that got introduced with the implicit txn patch +Ciao Paolo the aim is not to publish Nutch artifacts but mainly to manage the dependencies +Open Server view +Though if this code is such that we dont have any contention over it this issue could be resolved before the other two weve got outstanding +I dont think there is a need to mark it package private +Thanks Viji Balas +Moved the bug description to the Description field +This is the path +We have support for doing this already by using the in combination with the Tagish Login Modules +It is a bit of a big patch but it does solely address this issue +If you have additional tasks scheduled for this GA release you can add them there +Im thinking we must have gotten all these by now so Im closing it pending contrary advice +. +Sure if that is easier for you +Im now trying to localize the problem and provide another standalone test +its probably better to move it to mainAnyways the plain is like you have suggested +The victim is trying to get a shared lock on the row +patch fixing the issue for of curiosity what are you using that identifiercode evaluation for +and maybe the next step is to even work without component and page specific classes +Give me filename and line number where i need to change code with solution +better to shout loudly and not act at all vs come up in a potentially broken state to fix strange state in has been released +Hi DeepikaWere you able to repair your indexes? Has the problem happened again? Are you as Kathey wondered using a version of Derby? Without more information it will be hard to make progress on this issue +Fix in along with other +The idea is that the dashboard display can be trimmed down to display a small set of useful and relevant information +I would say thanks for the attention and go ahead and close it. +It is a class that instantiate a channel in a thread three times +Im pretty confident that the test failures are unrelated but the findbugs warning seems legitimate +yeah seems that way +This will only effect users trying to start using hsql +Will follow up with an explanation +Maybe later you can reuse the code for making messages persistent +The tested patch to fix a warning suggested by Mark +It looks like we didnt use sessions right +See the modified class are bound by the Java compatibility requirement for base module and therefore may not use of Java features +Please try to uninstall meclipse features and then install them from scratch +patch dropped the range queryfilter changes so we can refactor those separately +I would try regenerating the patch against the latest trunk and see if you get better results +This may have been fixed by. +Nathan could you please take it I have no env set up now +Ill try and help out as I can. +Also if the startstopContainer were to complete around the same time as when the timer kicks inPossible fix would be to synchronize in the main task on the when we dont care about interrupts and always synchronize the on itself +Territory based collation is supported by both the character sets too but there is no way to specify the use of that collation at this point +Verified that the configure throws an error on the new one +In general the exit code is set correctly +Correct proper fix has not been implemented yet +I will indeed create a new issue for client driver once the changes for this Jira are committedIll try to commit within the next day or so if there are no other comments +Velocity does not provide access to private or protectedmethods and classes intentionally +But we can do that in a separate improvement issue +Done thanks to update +thanks +You should be able to reproduce the NPE with the attached sample project + +So if you are sure there is no concurrency problem here then we should be fine +Just got an Eclipse crash +Hi Lewis I am using just atm it works with that not setup for the other datastores that will probably be required for running the GORA unit tests +Independent of the suffix mapping problem why we need a standard way to expose the current instance? +The default options are in place for what is done most of the time +bit map indexes are also mentioned as a solution +svn patch with latest zhCN translations +So far all the work has gone into actually building a and using that to build consumers of the the describes +should be thrownIm not sure if this feature worth the effort +its now possible to dotest somedbdesigntestrewriteor test somedbdesigntestrewrite +It fails randomly when it does so the transactions become locked and the stack traces exemplify the output. +am diffing the and AS messaging schema today for all the persistent configuration elements I have seen a few minor differences +And since isnt even included in this tree I think its fine to relax the restrictions as suggested +demo. +Ok the class name will stay mandatory +Ill take this since the pending test recorder test infrastructure changes will actually add the entire jsf to the test distribution +Verified with all runtimes in JBDS B. +You only need to change configuration of Defines the default XML parser to be used by castor The parser must implement file +What Yoav meant was that this mustve been fixed in the recent revisions +We are looking into it +aCBE +IO thread registered for notify +as gives us an extra hook to do some extra Groovy coercing of like things. + +This is necessary to allow access the of the wrapped Analyzers without making TSC public +patch for for +May well be a related change but didnt see anything immediately obvious +Thank you PatrickIm sorry that I missed it in the manual +This appears to be orthogonal to the other question unless the persisting iterator objects are retaining access to a DOMDTM which is in turn retaining access to the DOM +Removed duplicated +Now changing to Fixed. +kicks in when Bad Things Happen lifetime numbers are just noise when it comes to diagnosing the immediate problem they have +Deferring to +This bug has been marked as a duplicate of +Tests pass now +The actual fix is in srcmainjavaorgapacheactivemqbrokerregion other files contain improvements +Theslave will have to unserialize the byte buffers to get the logrecords but that should be fairly easyThe byte buffer strategy requires far less objects since eachbuffer can contain many log records +pending additional few line changes in SLA test case to avoid timing to trunk and +delete the renamed directorybecause is saved to snapshot while renaming itself +Let me know if this sounds good +After discusssion with the remoting team and other interested parties our requirements will eventually be catered for by the new Remoting Remoting will not be released in time for JBM therefore we will use MINA directly and swap it out later for JBR need to make sure that MINA satisfies our requirements for HTTP HTTPS SSL scope of this task should be extended to analsye the current wire format and alter as client and server side abstraction should be created allowing messages to be sent from either client or server and handlers regisstered to receive messages synchronous and one way sends should be supported +Code was added in that serializes access to the compilation stage +Ive committed this +added +But I digress +has been added to and is now +It seems silly but please attach that patch yet again with the appropriate checkbox set +My vote would be to overwrite and warn +Second patch enables running all tests from MavenThe tests in fails ATM this issue should be looked in toSending orbAdding orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoSending orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoSending orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoAdding orbsrctestjavaorgapacheyokoSending orbsrctestjavaorgapacheyokoprocessmanagerSending orbsrctestjavatestpiSending orbsrctestjavatestpiSending orbsrctestjavatesttypesTransmitting file dataCommitted revision +I think this could be called right after Deployer line +Those that occur in both the overlapping ones +Since there currently exists a problme where the network bridges fail when temp destinations are used across them the default setting should be to diable temp destination bridging untill we get that fixed +Looks good just a few comments All the default goals in the files now call jarinstall instead of jar +Lets wait as long as possible however +This wizard is on its way +I just tested that second scenario without but a different PU on each page directly injected into the pageand that works fineThe code you provided still should not make any difference +issues +Ok cool +Nobody objected so far marking as resolved +Please help how i can get around with this issue +There are many instances where we make store a reference to optparams and later on set values on it +But to clarify youre saying if you specify either value for Android it always snaps pictures back at the user +Added documentation to our committed +Doesnt matter will fix the failing case regardless +Ive made you a contributor and have assigned you this issue +Sorry createPatch maps patches to the maven modules not the root directoy +patch fixes it +There are a few potentials but have asked the team to check. +OK +Test Case is added due to verify the issue +Actually the subject of this issue should be a little bit clarifiedIOreopen with File with FileCREATFileEXCL mode fails +duplicate of +Some state is built but the state is not guaranteed to be propagated to all threads at read timeIm sure Ive left over other classesHardy lets chat together to see how to solve this class of problems +The localedefinitionMap is completely cleared when the refresh is neededI suppose that the correct approach is to check if refresh is needed before loading a new definition file +Rgds Gary just wondering is this patch okay for you +Patch looks good to me +Kristian noticed that some files were missing from our source distributionsAfter applying this patch the source distributions include some more filestoolslntoolslntoolslntoolslnREADMEtoolsreleasejirasoaptoolsreleasejirasoaptoolsreleasejirasoapsrctoolsreleasejirasoapsrcmaintoolsreleasejirasoapsrcmainjavatoolsreleasejirasoapsrcmainjavaorgtoolsreleasejirasoapsrcmainjavaorgapachetoolsreleasejirasoapsrcmainjavaorgapachederbyBuildtoolsreleasejirasoapsrcmainjavaorgapachederbyBuildjirasoaptoolsreleasejirasoapsrcmainjavaorgapachederbyBuildjirasoaptoolsreleasejirasoapsrcmainjavaorgapachederbyBuildjirasoaptoolsreleasejirasoapsrcmainjavaorgapachederbyBuildjirasoaptoolsreleasejirasoapsrcmainwsdltoolsreleasejirasoapsrcmainwsdltoolsreleasenoticestoolsreleasenoticestoolsreleasenoticestoolsreleasenoticestoolsreleasenoticestoolsreleasenoticestoolsreleasenoticestoolsreleasenoticestoolsreleasetemplatestoolsreleasetemplatestoolsreleasetemplatesTouches the following fileM toolsreleaseCommitted to trunk at subversion revision Ported to branch at subversion revision Ported to branch at subversion revision +Bump. +Should be fixed for as well +Vikas + +defer all X issues to +I modified the to log exception at DEBUG instead of WARN +Changed priority to minor and assigned to is that an application can take care of that itself by adding a decorating property to events perhaps using an Interface or abstract class +I was not aware that these system prorperties are controlled by the start script +Kind of obvious ones +The patch to fix the problems +har to be an archive only if it has index file in it +Any chance you can release +Ive removed the synch +Assuming fixed this is very old. +Whats more its sporadic it doesnt happen every execution +What do you mean what wasnt right? My comment was just regarding a performance improvement otherwise this should be fine! +Like I said I think you need to extend +We have servlet code available if the Shindig community wants this feature +Struts strictly adheres to the HTML and XHTML standards and will not support code in any manner +My proposal is move tests which depends from Xulrunner into plugin +It will come out with Axis +I have added protection in the class for null plotType parameter because of which the null pointer exception occurs but I dont think that with normal run of the program this problem should occurDo you create a separate instance of the class somewhere? Is the correct parameter passed +This should also be resolved by The code stored RBD volumes as QCOW in the database thus confusing the Agent +I just committed this. +Ive been staring at the SELECT statement for a while and found that Imwholly confused by what its supposed to mean + +Closing old issues +Created and added a patch +MaxHere is a build driven by an ant script +File caching based on the size of the Attachment is done +to see the integrated under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +OK + for the patch +Sent it to dev as I should? +Core committers please review and comment or commit +bq +composite etc? Looking in the src for the I dont understand why it needs to have the call if the defines the Tuscany servlet filter the shoul dget started automatically +Patch was updated to apply over +We work around this in our avail checks +Mass closing all issues that have been in resolved state for more than a month without further in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +Im trying to get an intersection of the special handling for the A normal delete also triggers the same bug thoughdeletionPolicy sharing could have something to do with this given that it doesnt seem like Solrs original deletion policy was meant to be shared by multiple writers +Your option is the same as right? Though perhaps a subportion of it since its only checksumming the transfer +Depending on which file you select the current renders invalid json +batch transition to closed remaining resolved bugs +Apply your patch +Reindexing mule this usually fixes it +s a cool workaround +Keeping a resevoir to release around oome is common technique +I think its ok to skip importing the bugs from the OSAF bugzilla +Assign to Ratislav to provide steps to reproduce +else conditions to the controller workflow significantly reduces clarity and focus proposal demonstrates that the various success view types would be managed by the framework not the application and that they are mutually exclusive +Ok +Ive finished refactoring the lower layers viz +Theres a json library that the rest of uses which might well suffice but someone needs to take the time and do it +helper JMX in case the user wants to query the active CLs this can go to the tools folderwe dont need to commit it to the core +Utility projects were never expected to have further child projects +This patch resolved my problem when reading somewhat big contents through using thriftReading the description and referring to my experience all the contents that raised was bigger than bytesI hope this patch to be applied ASAP + +we may not able to fix this +We use though so not sure we will be successful +bryan check this and commit +With LSR directory creation will be handled by the output format and will be store function OlgaYou marked this bug as fixed but from your comments it seems like insteadits a duplicate or child task of some other JIRA having to do with LSR +The simpler and more standard we can make all path handling the better +hi Sebb thanks for looking into thisbq +Marking it as Future in case it is followed up on. +Committed revision . +But this depends on what type of replication policy one can come up with and I would like to leave this for a future time +The special case canonized wrong results for the network runM javatestingorgapachederbyTestingfunctionTeststestsjdbcapi +Just wondering cause we are so close to the release +BTW the test suite does not run with this patch due to missing the permissions +Thank you for adding the doc page! Very much appreciated! close for +I just committed this thanks Terence. +HiCan you confirm if this is on the master branch which revision and which command you ranThanksDave +Local test passed Submitting +thanks for reporting jing +Yes it may actually be that ticket then +Patch good +Reopening in an effort to close +Thanks Ill get to processing the pull request very soon. +Constructs a defaults in the init method which is used for getting default value specified in for only count collate and extendedResults parameters +In your implementation wed have to have a different spellchecker component handler defined which is fine in most cases at least for me +Note that Ive closed the issue as I committed the patch but more work is needed +Same pull request as +Seems I caused some confusion about if I mentioned TC in the first post ornot +Your patch added a new class but its not used +Thank you for your suggestion +Will be solved as part of related issues +onbeforedomupdate event doesnt work also +Commit the changes to trunk At revision and At revision Thanks for the patch viola! +bq +I just have small comments on had a cursory at this patch it is great work +Rob +Please try the patch on +Since the hadoopjar in CHUKWAHOME is package and supplied by chukwa at build time it should always be compatible +Done +But without making this patch much complex if want to go ahead for committing I have no objection as this will be covered anyway in +Were just making that decision for themAlso does this change in behavior have any impact on the normal callpath from a performance perspective? Pinaki is hinting that it will affect performance +Attaching patch with proper license grant +and are different patches +Brian the effective pom editor is using internal model that is refreshed upon editor save so it takes some time to complete and if effective pom viewer is opened before model is refreshed youll see content before the changes +fixed +The selectors work for the binary +When the example tarball was renamed the package target was broken +Its the normal mode of operation and if wed fix that behaviour we wouldnt even be able to compile the Groovy source base anymore +Although Commons Validator has a resource attribute in its dtd Struts is currently ignoring itI just posted something on about resolving that bug with one option to add a boolean resource attribute to +At this point thats all I know +Touches these filesM javatestingorgapachederbyTestingfunctionTestsharnessD javatestingorgapachederbyTestingfunctionTestsmasterjdkWith this submission we are down to the wisconsin noise plus some diffs in sysinfo and sysinfowithProperties +The contains the class you can then either instantiate an instance and call the additional method I propose or rely on reflection to call find +the fix also goes into eap should documented as fix on reopened to change Release Notes text +Figured out a way to help here +In revision I added a test to reproduce the issue +Added a note to the documentation +Although the benchmarks demonstrate that we do better with threading I thought it is good to discuss the issue before we conclude on the benefits of this optimization +Not sure I agree that we need the extra qualifiers on the request side but we have plenty of room for the labels +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Slava doesnt Helios have the new servlet version +Attached sample RTF file to reproduce the think this isnt RTF specific as I can get the same empty title with a plain text fileMy hunch is that its an issue with how the Tika App is rendering the xhtml sax events +If someone can describe how to reproduce this please reopen. + +war is ther ein ER but there is a deployment issue +Maybe someone can recheck the mode semantics +It appears that isnt actually correct +Can you please verify +for delaying this +I attach the patch to view what has been doneTo summarize if the temp and temp are not null and that the user provides a name the sender endpoint generates a file using the prefix and the suffix +Thanks +There is no dependency here +I committed the patch to trunk and +I think that we can stick with this single issue but the problem has expanded in scope +girar +Hmm Im wondering if simply not incrementing the digest wouldnt be the better and simpler solution if a joiner gets and then gets again it will simply discard it +committed thanks +FYI opened to revert this JIRA. +Moving the default deploy location into the workspace could be done in a LATER release but im not willing to muck around that far just yet +The documentation for Attributes classes in this package is not generated because these classes are and the documentation for private members is not generated by you please tell how you arrived at this link? Was another part of documentation linking to this URL or you manually came up with the URL as you expected it there +This issue is fixed and released as part of release +I should also note that this is without doing grails war +I have an app that currently stores data on HDFS but I would love to make it run on zookeeper without changing my application code +Hi ChrisAny thoughts on this issue? Do you see anything in the jstack traces attached that indicates whether this is something which a customer could hit in an application and therefore should be fixedRegards have seen a lot of this in the component tests over the years so I am not too concerned to add this sleep if it helps +HenriI have the rsync location setup with your public key however were not Maven users ourselves so is there some easy way to go from the we have to the POM and structure you need to rsync +What do you think? We could use respondto and send plausibly +is a subset of this +End users should set it on the correct type such as a such as the fromThe thread children is now also configured with the inherited error handler as suggested by BelaThanks Bela for reporting and digging into this issue +Fixed CCE +What do others think +Working fine now +Well it looks like Maven doesnt even support a packaging type of so Im not sure why the Nexus indexer is even looking for that packaging type +Simon +Given the result of the test so far I am going to commit the image tar ball and kick off the tests with the new patch +Or even comment out ldap in +The goal is to generate a pom using only Ant and the jar that contains the maven ant tasks +Yes that sounds like an SCR issue is it using the correct classloader +Have done the rest wireTap had also onetrunk +Committed at subversion revision +Added fixed updated xml in properties Final is released +Thanks Nick! Patch committed +Bulk close for release +It work both for common patch works on top of version AJ and distro to was v +Does the latter part of this sentence lack a subject +So when symlinking etcsecuritycacerts to etcsslcertsjavacacerts the SSL exception is gone +can we use that +Sorry for the trouble +There is a mismatch between the metadata and the as the latter is returning an integer in the place of string column +Applied to trunk +so both the enums behave in a very similar way +Any volunteers +Dont worry thats not the the datasource doesnt support PC directly it does take part in the PC transaction +fixed format based on Virags to issue as it has been released as part of release +Changed target version from to since just came out and were a few days short of code freeze +Tested on secure cluster too +language features +NOTE The workaround of downgrading to is not really an option +Simons suggestion was just to throw an instead of a which is probably ok and definitely easier +patch looks goodThanks Jothi +The unit tests all passAs this is my first submitted fix to Geronimo suggestions for improvement are welcomeThanksDavid Farb +have you measured what made the biggest impact on performance changes to Metadata orchanges to IO indid not have time yet I would quess that IO changes make most signifigant part +Mario do you have project that would allow to reproduce this issueIve committed fix for the NPE but still not sure what is the actual cause +I have attached all the required logs while reporting this issue +This is not just for no I dont see why locus should refer to any of these sites in their associate sites +Instead I added a initial size instead +bulk move of all unresolved issues from to +An alternitive is to set the system id along with the input streamreader and the parser can check for duplicates for you +Please remember to attach a complete test case that can be used to reproduce the problem described by a bug report +Why force the user to pretend to be maven when we can do it for them +The attached patch removes the unused parameters from the signatures and updates the callers accordingly +To try to get some debug info I ran Eclipse using the attached file with the options shown in +This issue is currently being fix inWill close it once patch gets commited in trunk +Thanks. +Ill go ahead and commit this +to trunk +One new problem that popped up since velocity moved is the fact that the ant build erroneously builds a WAR with and in it +You can check the RM web UI to see if there are any nodes registered +I cannot add much new information cause there are still no for the two Tests and the one for only differs in some line numbers from the one already uploadedAs a consequence Ill build with true true from now on and wont bother you with issues anymore at least as long as Im running on this old and slow machine +I am closing the issue +Thanks! Hope it gets rolled into EAP quickly +messaggi tradotti messaggi non tradotti +Hi YeThanks for submitting the patch + has been marked as a duplicate of this bug +That sounds really odd +note that only the second instance of that line is incorrect and needs fixed +Thank you for your comment Jacopo however this required segmentation has nothing to do with glaccounts this segmentation is at the transaction entry level even lower than that because the transaction entry can be further split and spread over several different segments with this implementation +Its pretty small but the amount of work Id have to do to reclaim that performance is pretty small too so Im apt to do it +Attached affects version log outputs using SNAPSHOTS from +svn ciAdding assembliessrcmaindemosdeployerAdding assembliessrcmaindemosdeployerkarAdding assembliessrcmaindemosdeployerkarAdding demosdeployerAdding demosdeployerkarAdding demosdeployerkarNOTICEAdding demosdeployerkarAdding demosdeployerkarsrcAdding demosdeployerkarsrcmainAdding demosdeployerkarsrcmainresourcesAdding demosdeployerkarsrcmainresourcesAdding demosdeployerSending demosTransmitting file dataCommitted revision . +Im not sure about the use think given the current approach the use case for such script would be synchronize with remote repositories and automatically adjust IU versions +In other word stress n F will do insertion on the same row +The current Hive Parser is doing good forThere are something missing from standardSQL I think we could add them one by one extending the existing parser and maybe also add some for any missing functionality +you releasing the plugin on the dev update or the release update site? +Also if the user clicks on Reconnect button there should be the Server Dialog in case the reconnection is not successful +For servlet endpoints we are able to pick the meta data from the property +Removed reference to userClass attribute +Thewrite still reports a a release on the pSerializer instead of doing a delete on it seems to have solved the leakIs this correctThen we do not have an issue with writeXerces developersplease note this +In the little time I spent on it I was thinking of taking exactly that approach +reduces the memory allowed to a smaller value +Stefan I actually think that trimming by default is not +is a heavy object that is reference counted so there may be searches going on etc +You can set default part as source and reproduce it +Will do +Please add unit tests to the patch demonstrating the issue +Brandon can you please split this into two separate patches +If uses the static method to get the object instead of accessing the structure directly and goes a bit further in getting messages and passing back the cause exception instead of the rollback error exception +Your second part looks good +It certainly does seem more reliable than in +right that works for the kerberos impersonation but not the case it uses the common scratch directory +Will put up a patch for new package on +As the Jenkins builds are still very flaky I have asked for Buildbot jobs to be setup see +the probability that youre going to be adding a node in a small time window of running with ROW set to off seems much smaller than the probability that you have to +same as Jon +So it is not a good way +When FOP is rendering SVG if the font isnt found it may have to resort to stroking the text which is much slower than just using fonts to draw the glyphs +In addition there are other places that we use reflection to do things conditionally and a conditionally loaded jar would be niceSo Im for +Commited +Fixed at revision close for the release +jpeg and it does notWhy would the consumer do this check? It already knows the file type in your example aboveIm not saying this functionality wouldnt be needed just trying to understand the need here +jslint is being removed not relevant for CRGA +If it proves to be an issue we could consider a based workaround for with no guarantee of future support once the proper config element is introduced +Committed to the trunk and branch +The patch should be relatively trivial but like I said I have no idea if there is other important stuff going on there or not + + +patch committed to trunk +The next build from hudson should be considered upper than the released one. +John can you have a look at this +hiwe think this works now have you taken the latest code in the past day or so to check that it still broken? Adrian made some changes and it look sgood for us +WernerCould you update the patch against the regarding the new +Dup of +New version of that patch that uses ACTIVEMQCONF and adds ACTIVEMQDATA as env vars +Is this still in progressI am willing to contribute resources to help push this through +Same patch for trunk +wink wink nudge nudge +bulk defer of featurestaskswishes from to +If it is virtual and some of its variant is out of stock or all variants are out of stock then it shows the proper message regarding the unavailable variants +Can be done for primitive xhtml templates are added +Thanks Mayank! +comitted the code in r +Krista please take a look +I already fixed it +Is it a final release +The fix in makes this issue irrelevant. +Hi allService Client is responsible for the headers added and we are not enforcing that +Look after a filedirectory called in the downloaded zip +Why not check if the directory exists insteadYou could run into a case where the dir exists but the file is not there +AndreaI do not think adding an extra dependency is justified in this case +Could you please review the patch and apply it +Please check +This was incorrect + updated just committed this +removed extra hidden field +The moderators +It would appear that I nailed this one was happily loading old page information when the wiki was restarted from the serialization store +Which released version has it included? I have tried and it appears to still be occurring there. +I guess one problem is that in order to generate a good hash you would have to look at all the elements in the vector which is problematic as vectors can be very hugeI think it is generally not a good idea to use objects as keys in hashmaps that have a very costly hash function +If you are using the driver please report the issue to. +Not sure how to go about tackling these kind of issues +Committed in r +Its specific so will fail +hlink and hbutton are the ones related to outcomeTarget +Ok now I can assign this to Alessio back to +BTW your engine does not need to be SOAP for a service to be exported +It still doesnt include Dmitriys unit tests but the other patch from him does +Ill try with a larger model +Patch for as per the behaviour is implemented in the RI +I understand that my change broke the test library for the ultra strict environment but it seems that if any merging back to needs to be done it should be done from +Actually its not a big deal and I once thought there would be some changes after review so I didnt update it in timeSorry for that and Ill always keep the patch newest. +Has no tests in it +Looks like this bug was introduced years ago and youre the first to notice +Here is a patch +Maybe thats what you meant +If one really needs the Debug CRT one should also have Visual Studio loaded on the machine to be used making install of any MS file unnecessary. +m is the correct place +Still support and guidance would be appreciated +Testing if this helps +Please generate code using u and g optionsIn this situation there are type hierarchies +Thanks for the bug report! I should be able to fix this momentarily +The patch should be regarded as no more than a signpost pointing towards a proper fix +some newerish unicode logic in packunpack? Will look briefly +Thanks Vinod! +All I can do is to schedule it for and see if I can fix it +as discussed this moves to timeframe +Patch including recent do you some have time to review this patch +I changed the bean tree to demoTree as a +I reopened this issue and backed out the patch because it broke securerpc +The only other alternative I can think of is to keep track of imported files and associate them with their corresponding targets +All version ranges that affect EAP have been removed the remaining ones can be cleaned up after GA. +The Sling class launching the framework ensures the framework beginning start level property is set to to not fully start the framework automatically but give the framework listener installed by the give the option to continue startup or to restartImplemented in Rev +For the general scenarios seems no improvement even be worse it needs more test to verify +Do you mean log it after a block is written by datanodes or just allocated +Thanks for the patch +Ryan it dont look like a patch it dont look like a diff +I just committed a cleanup fix +ThanksHPP +I will also fix the itermittent testtwitter failure reported by Hudson +Nice catch on that race condition Todd +we need to fix properly to make sure the user can only enter proper date values for a user +Actually the issue is resolved +Attached is a simple patch the add the to for this example +I wasnt able to find a C driverIve seen a branch named v on the projects github siteIs my understanding correct that as of yet the v C driver has not been released? +Thanks for pointing this out a patch for this + +Regular expressions issues I currently throw an internal error +Added sample implementation to pool +I say defer until +It should work now +I recommend that we follow through with the alternative suggested in the source code comment sort by weight and divide evenly +The receive loop in expects that any type of connection will be long lived but streaming connections only send a single header +Ill try to pass my patches through dosunix but Im not sure if this is enoughBest regards Bartek +Sun JDK is not supported for EAP have tried to reproduce on Windows bit SP and it works fine all tests passed for seampay have extended the delay between the first and second payment to get rid of a failure in some combinations of fast HW and jvm with EAP +Providing validation on the other hand is very much a part of what Validator doesWith regard to clientValidation not describing the kind of validation well can you suggest other kinds of validation that someone might think of when they see that name rather than and get confused? If you want to get pedantic about the name then scriptable doesnt do it either because that name omits the scripting language and also omits whether the scripting will happen on the client or the server +I do not get this error when connecting to a windows based Filezilla ftp server +Can you post a sample of this bug so that we can figure out where things are slowing down for youAlso you may be able to mitigate this with a splashscreen +method only accepts a +what is your binding style? document or rpcNote that multipart is only allowed for rpc if you have not declared the part to be used in the binding +Can you point me to a test that do a conversion fron a POJO with jaxb annotations into json +remove is for removing config properties +for I meant can you see an jar in the server? does it look like a fully working +I think this was targeted against +Under normal load they are instantiated as expected and accessing their properties works +All tests run +Can somebody from INFRA please take a look at this issue were facing +Compiling should always fail if there was something wrong on parsing this condition. +However the token is actually treated as and the policy is ignored fails as a resultColm +Moving out issue not being worked on though would love to have it. +confirmed fix. +It works as intended. +Ill apply a modified version of this patch tonight +Weve changed some of the communication parameters +Dan We really need this featurefix as soon as possible +If we filter directly in we either do it on the leader or on the observer +I was able to fix it by changing the order in which I delete entities and by removing some sI cant get the problem reproduced in any simple setting I though may be the reasonIm sorry +Latest version of signed jar from build +Any commentstnx +Votes please +call httpxxxxand see the memory for Object name and runhttplocalhostYou need to hit the page twice to see the result of +have committed your change at revision +still investigating if this affects trunk or not code change significantly +Why is the console using WS endpoints? Why is it not using the local transport +Roger thanks for the very nice codeIt is now in contrib +PaddyThe fix should have been committed just recentlyBut the fix is part of a big refactor of the underlying file transport in Camel that the FTP now sits on top off +Can you subclass it Nick and make? It is ugly that we have to reget the offset for want of this class you are trying to add here +Works well Evgeny could you just change the following thing in the description Deeply nested if statements are hard to read +If you are in GORAHOME then you can generate a patch simply by doing svn diff or if you are in git then please do git diff The patch seems to combine numerous things which are not useful at this point +Committed revision + Michael +If youre referring to ndary indexes then thats indeed but otherwise thats a duplicate of +How come +You have constant offset from constant base which is an absolute address the two should have been combined before going to BURS +To me it looks like this is no longer a problem +Hi Matthias Im also interested in the portlet project +No response from the person who reported this yet so am trying againIn the mean time I have seen myself in my own experiments that an output filter can be called an extra time after the filter has already been closed because of the mandatory flush that is occuring +Something like a build path in Java Projects would be nice to have in projects as well +Attaching v that I think adds the intended method +When you create a new project using the create script those files are copied as the default wwwIf you are just modifying your existing xcode project then this shouldnt happen and its worth that your xcode scheme is set to your app and not. +Theres a problem with some characters not being matched properly +Dont think they are needed +Closing it +Feel free to rearrange things to your liking and make corrections that youdeem necessary +Reopened in light of additional information provided by D +Also can you rename HDFS Design in the table to be HDFS Achitecture since that is what the title of the doc is when you click the link + issues solved in released versions +Change Server State Detectors Startup Poller to Web Poller +Thanks +just a of the wall idea temporarily increase the replication factor for this block so that it is available in more racks thus reducing the network cost and also avoiding +receiver listner is not starting up hence no way to get the is fixed now +Thats what I see when I have an empty index +just committed this. +i mean if you remove this line and keep the line set true +I have an addendum for the last two warnings +Has this been voted on andor committed to? Since has already been released Im moving this to for now +I come across this problem should complete those creating files when encoutered +Yes it still occurs with +Snapshot of my toolkit to build XALAN and web sitesSee the file The toolkit is created for Windows platformsSteven J +Hi HenryIm in vacation right now so need to hurry for this reviewI can confirm you that we will need to fix all other stores to set the conf I havent found a better way to fix this for nowRegards +r makes the necessary infrastructure changes to allow us to seamlessly override the database driver during testing. +Is usrlocalbin in your path and is inside thatIn point of fact is the way autoconf expects you to bypass the path so if my suggestion proves correct this isnt a bug +Ill try to peek at the implementation again but I think it handles this okShould this go in or ? I figured Id put it in either way +Submitting file for hadoop QA +for pushing the addendum + fix is now a synchronized method +Thanks for raising this! Ive included a corresponding note in the that was really fast thank you very muchKarsten +Sent the patch for review +Also add tests with one BP and with two BP +When i have tested the behaviour of the two implementation was not the same +Ive noticed a lot of patches going into the branches recently +Applied in fafdf. +Leaving in +This is the patch +Thanks for the pointer to though I do believe this custom binding functionality should be made available too in webflow +Oh yes of course should have seen that +That sounds strange maybe there is a problem with the class path duplicated jar files or somthing else +Hi BogdanYou patch has better chances for consideration if you include unit tests +Maybe could just have this and throw away type +This patch adds comments to flag more properties which are undocumented +So I have put combo boxes in place of the check boxesIf you really have issues with the project specific settings Jason please open another Jira issue. +Deploy the war file to check the response +So as far as I can tell one global prefix for all tables sufficesGreat to see that this will come in +The directory removals will be done with svn rm instead of applying this patch +But I dont recall what it is +Please double check this class +koji I took the issue until the code grant is due etc +But even there the connection pool is only able to cache per Connection +Ive looked through the code and I see only one solution here someone must detect system locale and set it up into system propertyI believe this someone should be class library +Changing resolution state to Partially Compleated +Committed to trunk. +I will take a look at why this is still breaking in your case +This is not a minor anymore its a huge usability issue for users trying out ESB and should be fixed ASAPUpgrading to Blocker so it does not get lost +Okay Ill create a new ticket for it seems a shame not to use it if we have the code there. +Jason do we have an eclipse formatting config file +Cleaning up old issues +But try to reproduce in a new project to rule things out +JAXB is a good example of this +I also got rid of a dangling participle inPlease let me know what further changes would be helpful +One thing missed could move to common as well +Moving to in xwork trunk r +Closure is not used at all in the UI so this can be removed unless it is used on another component I am unaware of. +I think the idea is totally reasonable and hope we can come up with something +In my opinion Bounced and Read both the status should be handled in a single entity +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +Thanks Dan. +now I am adding back distributed profile in main in the interest of making progress on this jiraI have created to track the issue where webapp child also defines the same profile +I have tried it out +Yes I have tried with AUTOSCROLL and I have the same problem. +Swati I am assigning it to you since I am assuming you plan to work on it for +passes all tests for me a quick look through v +If you think Damiens solution is acceptable then Id say glink and gform do not need controlleractionid attributes at all since you can just cram the values in the url attribute as the questions +i dont know what is in context of report in a different issue. +Attached screenshot showing the added priority is fixed in version +I think we should make it per project + +The broken test are fixed +Railo error from my stacktrace indicate that this happens in a rest call but your example does not show a rest call can you please post the stacktrace of your example? with ehcache right have you tested with other cache types +I can try the current dev build want me to check in thatexample declare field does support the minus sign to remove a member but since you started with just running would not have upgraded your release +changed to false not true +Contributing patch to fix described networkTTL deficiency +But providing more documentation and examples on how to configure Eclipse projects that support would be an excellent idea +Patch seems fine +collectionAlso did not quite get the last comment about the workaround using Resource annotation +this could be a s not a problem lets think this difference between Runnable and Callable is whether or not there is a return value +I upgraded to Git and enabled the script. +Now this connector is based only on a regular expression matcher +Closing this issue because the requested change is not generally applicable. +I did some further research on this This problem is not related to tomcat it happens also on jetty +On it looks like the JMS correlation ID is but its the wrong ID so SI cant correlate the request and the reply. +Let me know if Ive got it wrong +Eventually the namenode will mark it as dead after lack of heartbeatsAre you actually testing trunk +I ran this test in isolation with and without the patch and confirmed that it works as expectedIm going to commit this momentarily +The latest version of binds config schema to ESB +the code changes at and at are irrelevant because they delete only commented out code +As you come across such cases please create new Minor severity for them. +Forget about this +Please change this jiras title to something more relevant to your actual issue +Would you mind modifying to return null instead of an empty list and then rerun the MBM with the code for the left in and also commented out to see what the difference is +Tara if you can locate your preferred interface change we will be happy to accommodate you +But I guess hbck can run into those kind of cases so patch is good +not blocking M +Second attempt at a patch that applies correctly +the sample ejb jar to recreate this problem +So you are right this method can and should safely be removedThanks for the review so far + +Since I cant reproduce the problem Im not sure how useful it would beTo be clear the executable is doing a transform and Ive let it run for an hour without a dump +the new patch looks good +I reformatted the entire file to get it to comply with current coding standards +I think this is not a struts issue this is normal behavior when using ajax +Thanks Jacky thats a great solution +Closing all resolved tickets from or older +Never mind +I set the priority to Major and the fix version to +Plus UNICAST might soon replace UNICAST so this issue is not that important any longer +Designer sets the name and name in source to but the logic is looking for getUpdated +Changed the fixed in field to CP +In fact if there had been a timeout in that log there would be a log message saying that the response was short and I cant find such a message +repositorygeneramajars please post test report as asked by have already made an attachment with the test report +I was thinking that this could be deferred to an activator vs +But its a delegator that caches modules full of all the methods it used to eval every time +A patch for +This is in testsystemauto and + This bug has been marked as a duplicate of +Thanks looks great Felix therefore closing this bug. +So why is this not a bug in this case? Please explain +Thanks Jacques will look forward for reply from you guys +So the employee in the role can only view partiesparty groups who are in the role of Supplier AND only those Suppliers who are members of the Organization Context the user logged intoHaving something along those lines built into would be great but I see where each deployment would have its own set of rules +The alternative would be to checksum data on the Datanodes with byteschksum and validate the new checksums against the existing dataDo we keep the configuration variable or do we kill it? Do we simply copy existing checksum data or do we itI dont think simply copying checksum data is enough since the checksums can themselves be corrupt +I reverted this +Ya I started to think this a HDFS issue +Needs testing +OK This only seems to affect deserialisation on the server +Hello Matthiasfixed the bug in and using your testcase +I did this for the base UIMA and for the top level of the sandbox project +Note that because the archiving process involves several filesystem operations there can be errors if its run concurrently on the same partition +The assertion can then be done in the success callback of sendBatch a first stab I assumed that it is sufficient to clear out the pendingCallbacks list and reset the instances state as in sendBatch +To be fixed in both think that the initial problem has been fixed in both and versionsTests has been implemented and they are now passingCan it be confirmed StefanThanks a lot for the report and the tests +The primary issue is that were changing the API in a fundamental way +I just committed this patch +So does fixed this one +I can redo + +The format was chosen as a defacto standard format for consistency so no matter what locale the user has selected data entry will always be the same +Still I am looking forward to a general solution +This uses the deployment and config mechanism provided by google +I would be tempted to sand it down a bit Since the new relational operator could be something other than you might want to call it newRelop rather than newEquals +Attached patch adds some comments and fixes +Will commit if tests pass +The confdir is then simply setup by using config like it is done for the rest of the hadoop +Thanks +Please try this fix and unit test +At the moment there is no way to disable the setting of the wsa header because the clients always generate a message id if one isnt explicitly specified +I think it is necessary to debug geoserver +Are the examples still available for JBT just not JBDS? They still should be available in JBT +If a user stops a bundle in the running system the system should not automatically start it +is config file +Here is my review comments +Thanks for the patchI was not able to apply it to the latest code in the trunk as the trunk has changed significantly since when this patch was developed but I have checked in a that does the equivalent of what your patch was doingAlso some lines in the patch were wrapped around column I had to hand edit and fix it to make the patch tool consume itRequires on the implementation class are added to the assembly Implementation model objectRequires on service interfaces are added to the corresponding service model objectRequires on callback interfaces are added to the corresponding callback model objectRequires on operations are turned into Intents containing a corresponding Operation objectI ported the test case you had contributed to the latest code in trunk and the test passes even though Im not sure if I adapted the test logic to the new policy model structure correctly +has been released closing issues. +I am also working on writing a junit test which will check if the readme files get created at the end of database creation +Patch v removes commented out code +I added test cases for +Closed. +The hashmap the values in the hashmap and the usage +I think this change verges on too radical to be backported +Ah okay yeah that could be a good idea I had the hunt that you wanted some osgi support in +Thanks Gary +I like your solution +A new tag for is needed + +Still I think that adding some advice to the sitewiki on how to profile Jackrabbit would be usefulRegards. +Yes I too have schemas defined with this problem +I did find a lot of records + +I agree with the proposed new feature but the patch needs to be improved with an update to the documentation in src if possible a reformating with curly braces on the next line so as to match our prefered code style +what about using the or instead of xpp +Patch for +This can be closed +committed +I agree +bq +I imagine this is the problem hudson was complaining about +Less likely +This indicates that even if users share writeaccess to the same SESSION schema it will be safe sincethey see their own session local tables only +Now those properties show up in and can be edited in the properties view +bq +NET the latter fails +Good catchAnother solution might be to use the following patch it changes to use a to hold its Transaction facades rather than just having a single global one which should mean that multiple threads never get the same Transaction facade + fix compilation problems +applied in revision thanks for the patchAlexander please check that it was done as patch was applied as expected! Thank You +Closing this as the related ticket is closed. +Attaching a new patch leaving those behind +The above error is due to the d ins interpolationYes sorry for that typo +rebasedi also note that if adding a token to keyslice does end up being the best way to solve really large numbers of keys it can be done by adding an optional field to the struct wo breaking backwards compatibility +Im thinking that the problem may be that we have some sort of wrapping situation in the DTM but Im not sure +Ah! is really relevant to this issue I forgot to duplicate the patching task in case of local version is will test and submit another patch +Any tips on what plugin was run from where when was specified on the command line and no sorry for spam +This was already implemented in. +I uploaded the same patch twice +Tested code from trunk +This patch fixes the issue for me +hbase is an apache project this jar is provided here as a convenience +I think we should try to find a way where we cache the resolved resources to avoid the additional latency introduced by the actual resolution +Hi MadhavAll feature versions of your RAD installation look OK to me +Thats what needs +please test against current trunk as this might already be fixed in +HarmlessIll fold this into x when I merge into x r after release. +The SDK feature has been removed and now a proper sources feature is available +Let me know if you think this should now be would be really keen to see this change reflected in improved Groovy support +I guess this is really what we want so that we can pass Ruby options as normal +Also my scrollable is the entire body of the modal panel and therefore isnt surrounded by any divs or centers +xml typo related to chmod under fs action in the example in the chmod xml closing element typo +I corrected the test and will commit it to see if there are still any problems +There is definitely room for improvement with respect to concurrency +Should I try again with a more recent version +Complete with rev ddccafafffbeaaaccc +Many thatks to all those commented here and helped +AES encryption test runs without errors on Solaris + this patch looks good to me +the ldemangle additions from +This issue is fixed and released as part of release +yes Romain +Attaching simple patch +committedOleg +It is not based on the patch although I will be looking at the patch to include any missing functionality I didnt think of +Please and if the memory leak continues. +Otherwise everything looks great +admin attempts to impersonate user but the impersonation of user failsDoes the attempt to exit the previous impersonation in step happen before or after the second impersonation is completed? If the implicit exit happens before the actual nd impersonation then wont this scenario mean that attempting the nd impersonation but then having a failure after the implicit exit will boot you out of the st impersonation as wellIm not sure this is really a problem but perhaps an alternate solution would be to do the nd impersonation and if it succeeds then overwrite the st impersonation +Can it be scheduled for fix in next release please +Please verify and close +So for now I just changed a logic of test cases +A mistake I dont have to close the inputStream from parameter +The component is operationnal +JanThat wont be doable until we make the btree balance itself during writes +This could also be an opportunity to port Surround to the new query parser in Lucene +one more commentthis patch only works for portlets using the dame trinidad version +I have a few tests that require embedded jboss ready to go +Thats fine +I would throw an exception just so that we can hear about those architectures nobody has ever heard of + Extract text OK in but badly in space between char almost OK in but worst in OK inDont extract anything in +Thanks +Lets start from this delay decommissioning until the NN gets out of the safemode +can still make the UI clearer and explain how it works there but that would be separate issue. +Attaching latest patch with changed incorporated from Vinods comments Added a new method in the Configuration class to set float values +All related tests still pass +Hi GaryA couple of points +Checked when jobs reach still the job and task numbers are displayed correctly without any erros +For now though w multiple people working actively on it well need to do it locally +Thanks to him +We explain it in our trainings and it takes a while +Please test review the fixOleg +The fix is simple and use deleteTable to delete a table in clean up phase +Please attach source files to reproduce this RF the issue is of a bulk update all resolutions changed to done please review history to original resolution type +updates were committed under +Good idea Tim lets use after every X attempt instead +has done this. +So any help on getting that to act like a real plugin would be greatAre you still toying with this? I can take it over and get this rolling if youd like +Ok cool I did not look into the code if configLocation was optional +All the tests pass +I created based on Nicks comment +I need to port it to master yet + removed from for and Extras sites +In chapter its a bit hard to follow what to do + + +Dims said to attach it here +Migrating bugs marked as resolved before and tagged as version as fixed in RC +Ill try to find the new version fixes in order to implement in version +I missed it +what if project has multiple poms +Attaching smaller patch incorporating Miloszs comments. +Ive just committed it to the trunk and +Hi Christianbut why did you change the component to render its children? For rendering attributes you would not have had to do thisregards it was a dumb question from my side you are absolutely right +Feel free to backport it to it is a couple of extra classes but easy +in the of a panelGrid +I havent had the copious free time I expected to work on this +x +Ive reproduced the previous scenerio and the m version works wellThis is my first time to take part in open source projectThanks you guys +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Aggree on this db change and will help to solve +Modify installer so that it executes system tar which is always present on Solaris and include provided build which can be gunziped before +Thanks Paul! +Latest version merged with trunk and containing the patch as well +then we should adjust the spec to set this prefix up in the prior to calling the bridge +As in my test execution the only test that failed at Hudson was which is not related to this patch and is already broken in trunk +There also needs to be a way for mapreduce daemons to list parameters that may not be overridden by jobs +Yeah +Patch testing v + +I doubt we will run into too many people bumping up against each other and staging is only marginally slower than doing a local build anyway +patch that fixes the duplication and adds oracle jira was last patch from? Can please you link that jira to this jira +Logged In YES useridAh yes +Note that I didnt include the nested tag example simply because I dont think it is a supported syntax +This was committed marking fixed +for the patch +first problem is that is really misleading +Experimentally weve determined this provides a good performance boost without too high an increase in buffer cache usageI also took the liberty of adding documentation for the other fadvise parameters to in this patch +The patch does not contain any author tags +You can delete a local gadget using the trashcan icon to the right of the appropriate gadget in the gadget list +Especially if clicking outside it already closes it +Please use this one +QE please +I just submitted a patch that will provide an additional argument sharelib to to upload sharelib into hdfs and an update flag following sharelib for updating sharelib +I already finished it! +Do you mean signout support for the plugin or the idp? The former is quite simple and could be done in a patch release. +Maybe a test case would allow everybody to understand the original problem +Line Service call manipulates global Debug settings +Please update and run your tests again +There is not our bug +New patch with changes detailed above. +sorry create before I finsh it please see this issue in No you try again with current SVN headthat sounds much like the problem which has already been solvedregards of +I would agree with using QR if we dont let the algorithm be chosenFrom what I learned least squares is one of the cases where numerical stability is really needed and LU fails +Additionally I have the second phase of this up at +Committed into mastereaaefeaaefabcfa +Patch that makes usecase behave identical +since context is now lost on default fallback urls +These lists are explicitly created to be used on this resultset only and cant possibly be modified by anything +Mind if I take this on +Ive changed this for Spring right away does not define a group anymore for ease of using custom please give this a try with one of the next snapshots +Im sorry the previous version was working this version the problem added a test scope on dependency between second and first project +It should go in to EAP CP +Sorry for delay +So if you are able to reproduce the bug I am sure the spring team would appreciate a unit test. +Closing this isnt likely to be done for +Youre right this feature of the enforcer fits my needs +Looks like we dont have strict configuration in Infinispan Server +It could be that a conversation scoped bean is being abused +must also be modified to buffer request objects so that they can be written preceded by their size permitting to determine when each request has fully arrived +improved job Sergey +Maybe this has to do with the fact that target is also aliased as +We already have gotchas in the codebase like where the user needs to know that they shouldnt change VertexEdge objects retrieved +Simple KV pair with separator should do +No problem +Could anyone help +Add patch that ensures the perftests all run with a prefetch +Thanks Hari! +Ill change generated sources directory to target and see if it a lot +So it shows that the namenode handles lease correctly +BTW just for the record you can add attachments to instead of doing a cut and paste into the comment buffer +I see Dougs point about it being perhaps easier to modify an abstract base rather than an interface +uses a custom CF serializer that is +This issue is duplicate to I think +A is thrown if there is an inconsistency +I am willing to provide some modest level of support from the OGR side of things to a developer interesting in building this it could be done as an OGR specific set of Java classes independent of and then I could include it with regularOGR distributions and would just need to implement a layertype on the API +So Im sending a new patch named herewithThanks SaliyaThe attached patch breaks axis please fix it and attach a new patchThanks patch on the issue +I get the error on the th iterator for process instances and on the th iteration for process instances +unfortunately the from run is too large to get attached +This issue should be closed +Are there any other tests you did that failed +Tested +Revision of resourceschematoolesb should fixthe original problem +Am I missing anything +I polished it a bitCan you test the code on trunk to see if that speedup your uber JAR +This change affects only the +I did a fresh checkout from master and it works of courseBut I meant that is not contained in the download from sourceforge +pull request filed +Patch checking those entries in of all solrlucene WARFor I assert that the contents are identical in the main distribution and in the WAR +It seems that option should work fine +Actually getting a minute to think about your second case is there a reason the value returned is kept in the tx context and not just directly written to the data container? We arent updating the value with this and it would be fully consistent +Will be posting a preliminary design sketch this week for comments +The YARN portion of that patch will be attached here +In any case it should have been deprecated and then removed in a later why negate the possibility of reusing existing code? It breaks nothing +But the error just wont give upJAVAHOME CJavaMHOME CWorkplace have found the root causeI added one line in set MAVENOPTSXmxm XXm +Could somebody with access to the Jenkins host take a look and adjust ANTHOME or the ant installation there on host hadoopHere is a snippet from the build homeshudsontoolsantlatestbinant Dversion homeshudsontoolseclipselatest homeshudsontoolsfindbugslatest homeshudsontoolsforrestlatest true true clean configure tar findbugsnightly homeshudsontoolsantlatestbinant not build has been modified to point to the new tools. +I am using gradle +Did the namenode use all of GB when it started? Thanks for your Nicholas I would hate to put in a reference to an Inode object from anywhere in the fs subsystem +If problems arise we should it +patch a faster way to uppercase the letters a through z +Vinod maybe it is not time yet to split the documentation IMO +Temporarily reopening to update release note info. +Yup right you are +Need to fix this so both insert paths can echo writes to the bootstrapping nodes for now uses it looks ok +Simple maven reproducer attached +Closing old tickets +Add to the list of repos in Review Board +As long as you dont use you dont have to link to +A test case to go with would be just great +This one still needs some further investigation but downgrading it to minor as its definitely not a major issue +Once the content of the file is correct it can be converted to a mapping class to avoid parsing during the execution of the test suite +bq +Mike thanks a lot for taking care of the rebasing of the patch +I tried to upload the jaxb api into the Sonatype repo today but I was told the artifact was up loaded already by other guyAnd the Sonatype people I touched cant find the JIRA of this artifactGary did you have chance to get touch with the Sonatype peopleBTW I just checked the it looks like the JAXB is released recently +Thanks Mike +Fixed backported to and branches +Makes sense +Where are your jars contained? Do you have duplicates? Also I dont know if you have Apache CXF installed as well +One more note compiling xwork using the source distro was a PITA +committed to trunk branch and branch +Renderer factory created +job is using my branch +Added more dependent jars +thanks +How about using some information from tpstats to make this adaptive +Raised to critical because we cannot start the QA until this is resolved +Ok actually refresh button on dashboard or somewhere IMHO only confuse users as it refreshes cache behind the sceneWhat about add this button to App Registry or call request on backgroundOr add preview window with refresh cache button +This is the wrong place to fix it +changed summary to clarify are right Emmanuel +Patch to add junit tests +Heres proposed fix +However I am not sure if the needed version of Scala is supported etc +but there was some reason +status to resolved to svante to issue as already integrated to the repositorychangeset adccfuser DeveloperweihuaWang Reviewersvanteschubertdate Tue Jun summary bug Convenience API for Version attribute. +Looking forward for the remainder patches + +Minimal test case that manifest this Activator is in the bundle itself but the generated manifest reads and the framework is not able to resolve this package +No response from user. +Committed addition to the feature +Removing from target versions because this change is dependent on code that only exists in right now. +reduce num of bookies to in reduce number of entries to in you figure out why the test cases were not failing? Before this change the disabling of compaction in test had no affect +Im using now and experiencing this error with my placeholder as the version number +Whats the situation in your site +Its really very fast and easy to +youThis issue yesterday afternoon has been resolved that causes this error hadoop version is not compatible with the process of compiling forget to specify in the hadoop version. +If this is likely to get accepted Ill add testing +Yes I am they are line by line compared against the common converter codeThe part that is lighter is removing the non string conversions and checksThe was only set up to convert Strings so we were only ever calling with a stringThe other nice thing is we will not be creating so many temporary objects during conversion looks like a wrapper was creating each time beforeOther notes I am done my patch Sorted out where the the extra converters were coming from +is modified as a demo of the use case +With an allOnes facility the first reason disappears almost completely and query times can be also be guarded in other ways for example by checking for document frequency and then trying to fall back on digramsSo the message is please keep this in +committed +Patch applied please verify and close. +I am also having this problem +Fixed again in and. +Implementation with serialization check and SPI for message serialization +Thanks I am satisfied +I hope to be able to provide the content in the next few days +Bulk closing stale resolved issues +Closed after releasing +will result in this errorERROR XRE DERBY SQL error SQLCODE SQLSTATE XRE SQLERRMC Replication master has already been booted for database db +This would allows us to be more scalable as no of bad records could potentially be high +orgapachecommonsjexl +for the patch please change the Ganglia rrd cached base directory in the Misc tab to Ganglia rrdcached base directory +It also was not coordinated with reading the jhist file or many other things. +closed as of release. +attaching a new patch which removes the unnecessary imports +Pat Any reason this is not resolved +I saw free Aries tools has been added as dependency +If at first you dont succeed try try again +Yeah AFAIC its a bug +It also fails appropriately when I remove the null check change to +I finally got around to writing a proper fix for this +because the spec actually doesnt require the client secret to be sent in the authorization request +Yeah its just addl fields to existing attributes but a unit test is always a good idea +We can weed out any issues we want to push to a later release from this set once weve done the consolidation +Appreciate your help +We still need to work on +I did not change the asymmetry for chooseTarget which takes a list of because this list is actually generated from the and represents the datanodes where the block currently resides +Maybe not but you probably want something in combo even if it points to the default processor +I just committed this. +What tends to happen with hive and features is the iterative style produces a final product not exactly aligned with our initial spec +Should not be a clean shutdown +componentType side files is this since original problem is fixed and we can work on and as new issues +Wrong patch +This is sweet +For this to work the needs an additional String field to store the wsaTo address +In order to run the test client you need to run +I have no problem understanding the logic in when I look at table in the JDBC spec but the table is for setObject +The problem with using a BLOB is a are designed to store large volumes of data in the order of and are hence stored outside the row +Need to test under performance tests for specific performance differences +Having it directly on the reader meant that foreach would dispose the reader after enumerating which was not always appropriate +and Win do not have fsutil +It seems this is dependent on how do we proceed to get resolution? What is the normal maven proceedure to do thisIf I can assist in any way let me know +i want to review the fix for this issue. +see assafs comment +However the result is exactly the same when mounting with our Jackrabbit +Duplicate of +With a little more intelligence I have a complete solution +Thanks +if you swap between booleanBoolean in Activity do you still observe the bad behaviour +Dup of according to Vasu +closed +Does the merger require two different counters? Isnt it sufficient to pass an optional counter for each record the merger emits? The caller should have enough context to know whether the segments are coming tofrom disk or memory +This should be fixed in the next release +Upated exception ignore patch and some converted tests +Tested and works fine for detection of duplicates +and what you ask is generally the usual idea is to reduce the number of database hits +Fixed +After a clean install on Ubuntu I am also experiencing no PREFIXetcdefault actionIm attaching both the and sudo make install output for reference +Yes I think we can keep feature as is and add specs bundles to appropriate features which need them so that feature could be still used by downstream projects +I fixed all of them +Yeah spotted this as well +Thanks for bringing this up +I pushed a commit for this yesterday evening see commit dfcdfebdaffdfffadffaVerified that it works just fine on my hypervisors +Yes it is a issue. +Ive reprod the correct behavior in my regular Chrome browserit works fine now + looks good +Fixes being made for will resolve this issue +Just committed this. +lgtm +they have the buttons on the UI to do things like Submit Patch Start Progress Workflow Resolve Issue New are missing these functionsSo it appears that the read past EOF error is causing to be only partially created and left in an incomplete state +updated the patch with the check +I do not expect it to be accepted for M but as Juergen says it makes life very difficult for those using and so I am forced to push it up for request +I dont see how the logic is broken in readSubscription past the fact that theres a possible NPE which my patch fixes +It cant be that only I and Christian are using aspectj with spring and +Mobile browsers can replace the old element with the new one in the PPR response +Just realized what was wrong +Targeting to as has now become a release with mostly fixes and minor improvements +the STOPPED process restarts with a new whereas this proposal resumes with the same execution +status for RCsame as for forclient differs +Checkout fails +In it works fine even withautoDeploytrue +Ive committed this and merged to +I committed to the sandbox. +log of news thanksAnyway I think classlib gurus will investigate this issue later it can be a hidden issue in classloader +Verified on Windows +I have attached the patch a very simple one +This is against the hit me up if you want one against a different release +pending hudson on the latest patch didnt apply due to conflict in +Im trying to use to validate the JTNN syntax +Well see if the community want a more fine grained approach to addresses other use cases +If you can attach the JWS that caused this error I can verify that its fixed with the patch on my local system +I am mostly in favor of just doing something at this point +are you sure that patch belongs on this bug? was very small isolated clean and an easy read +Will do +The only solution is to disable the managed entity interceptor +It can really add value to our test cases +This jira has been closed as part of a review and archiving of the aj project +could you verify if you have? Thanks. +This fixes the +Marking closed. +Todd and others whats the plan on this one +The form should be a from +You might need to run grails compile before running again to trigger dependency resolution +Waiting on hadoopqa +committed to always providing excellent text thanks +I will post updated the patch in a moment +bq +Patch applied. +No that hasnt changed we still want the raw URL +filed forAshutosh new jira will be useful +David you are definitely on the right track +Includes example of building this waySending modulesassemblySending modulesassemblysrcplanSending modulessrcjavaorgapachegeronimojeedeploymentSending modulessrcschemaTransmitting file dataCommitted revision . +Does not work in +Oops meant Spring Security Kerberos Extension M +Nick and Mike thanks for this patch well done! Im ready to commit it but I have points Nick I have to remove the authors name from +Note that an interesting issue deriving from making this list is that both and use the sessiondirectory option at the moment +I got the mail I just didnt process it immediately +What you propose as the common case then appears to be uncommon in practice +If we do it for some that it can be misleading for users and again the functionality might be overlooked +Hi KimYes STRONGPASSWORDSUBSTITUTESECURITY should be added to the list and as part of the other already listedThanks and good catch +Patch refreshed for latest trunk +Im just brainstorming here but it is highly likely that it will be a separate project from at least in the beginning +Now noteInfo has removed from the url +However Lukes patch is slower with Kerberos than mine +Patch on top of changes. +Check regionserver stopper during version create didnt throw when node exist +Workflow and are also done +Here is the revised comment gleaned from the comments in this issue and my understanding of it +So we should check all the mapped folders if the validating file belongs to them +OlegWait a little while until we get more comments from other developers +should be added release +passed Im running the rest of the tests now +Related bug fixed in r. +I see that the commit is made to the ASF but I still see the same issue even with the latest buildHence reopening the bug +that would be a small change internally but the effects might be very big on the outside +I just committed this +I dont see much of a diff between modify and postModify event +Thanks for the very rapid fix +I guess the problem you are referring is the one mentioned in see Not sure I agree +I tested this application and it works fine for me using SVN HEAD +Do you have any further explanation of this? Im looking into the interface at the moment and have some ideas but dont want to goalstasks that are already ongoing +No idea Ill add this to my list of is an old old bug can anyone reproduce it +That said I do not think this can happen in this scenario as the TM is never crashed in your scenario +I am able to reproduced this with +Applied patch closely following suggested code fragments +For example if you deploy a queue with name exampletestsmyqueuever you must specify its jndiName attributeSo to restore this behavior we simply ignore those destinations with a attribute defined in the forward slash checking code +Need to change to +there a reason the std namespace is imported into compilercppsrcgenerateAlso could you possibly update the fixtures data to reflect how the generated code will look? I know weve had a regression or two recently because the generated structs we were using in tests werent the same as whats actually being generated at this point +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +Documentation and format is there +I dont have a simple test case right now but Ill try to create one soon +Patch applied under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Fix checked in under change to. +Verified. +We will let you know the outcome +Im attaching the current file that I used in my scenario to get the problem for historical reference +this one looks quite interesting to me but one comment in here the problem is that this patch will only work with eclipse active but weve no way right now to map features to active osgi impl shouldnt we add some way here first before we apply that patch +patch to fix the that is attached to this issue fixes the ability to now set the feature without throwing the however when an undeclared element is encountered during validation an ERROR level validation exception is being thrown rather than a WARNING +Added the descriptor file in Rev +BTW I am using Windows VPN client for remote access vpn +The documentation issues are interrelated +Happy to see this picked up again but I think addresses the fruit here. +By default the scratch dir can be based on date etc +Teddy when do you think you can update the patch +Unfortunately i do not have a list of all supported Function Setsof my IP or any other printer +See +But it Im guessing you have it applied anywaySounds like we should raise a new JIRA about the deadlock issue so I raised since it doesnt look like its related to this particular patch +push patch and close this issue. +Patch for changes committed in revision +You can download a snapshot from the maven repo you just need to change the version +Thanks Clint I applied this. +fixed locally +After applying this patch modal dialogs no longer work for me on IE +fixed in forgot about that thanks! +the most easy solution for this kind of problem is to remove the offending jar delivered by Groovy and located in the lib directory of the distribution +For smaller size clusters the current policy might continue to prove the best +I think we should do this +Even if there is a maximum limit would I not see a capping of the results as opposed to a drastic drop +If the threads receiving from the channel are managed by Spring it would be possible to make sure that they are done +Defer all open issues to +an EAP with existing deployments +Im starting to think that behaviour introduced by patch might make a sensible default +This patch will work only with the patches of above Please review the patch and provide your comments +Can we have the master set watches on parts of the client is currently looking at in the viewer and then push the data to the client as events get triggered? HTTP is though so I suppose you always need to end up having some poll period so this is probably all moot +Im adding some clearer messages for the future and will check whats wrong with the concurrency +We can use the UDF template to implement CASE and IF +not a bug +Tony I know +This will hopefully allow the main Continuum build to completeDeleting srctestjavaorgapachetuscanyscaimplementationdataAdding srctestjavaorgapachetuscanyscaimplementationdataTransmitting file dataCommitted revision +This happened only once but if it can happen at all I am appreciate this is a hard bug to track down but it most certainly appeared and was very are the instructions or scripts I should invoke when having an odd issue to get as much information as possible +why fixed on trunk and not? +Ensure that there is no tablecfqualifier level permissions are stored in ACL in pre +The error here says that the input file does not exist +The output files look good but as Raghu said can you add a test with lots of unions +I am going to try on ubuntu +process standard input +As a future improvement we can add a feature where users can configure an optional sequence to handle this messages +Must have been a bug on my +Its been years since its creation and Nutch has made it just fine wo these logging improvements +Confirmed this problem on my VM too +to and to trunk +Postponing to allow more time to reproduce the issue +Stefan what should we do about this? Maybe we should bring it up in the dev mailing list +fixed in +Updated patch with a new testing validating the common changes +bulk close of resolved issues in. +I would imagine changing projcet facets might also since that can change classpath containers require a full build etc +update Now user cant assign to inactive Financial Account +Change mapreduce parallelalgorithms and sequentialalgorithms public static final String DEFAULTHDFSSERVER hdfslocalhost For HBASE public static final String DEFAULTHBASESERVER localhostThese are read from the hadoop conf and hbase configuraiton file +I think the issue happens only when newSearcher passes an? I wired temporarily to always pass that and it looks like its always failing +The provided patch resolves this issue +Thanks Jarcec! +Marking as resolved since no additional problems appear to have arisen +Reporters of these issues please verify the fixes and help close these this bug as create volume from snapshot is succesful +Looks great +I merged it with the actual revision and combine them to the new patch +I think it is a good sign if the project uses the software it produces itselfSince it was not accepted by the Apache foundation we did not further maintain this component and yes indeed it can be removed nowRegardsHans +Other than that great approach I really like it +Patch applied with thanks! +The best way is to drop the verification framework and to reimplement the validation rules with wtps validation framework that we have for seam and cdiWe have rules for rules for rules for +bulk close of all resolved issues prior to release. +Im testing things now for any side effects +Franco ok I might look at it but I have some higher priorities a patch would be very welcomeTsuy I think this could be avoided by having the bundle with a higher runlevel otherwise this bundle is very busy during shutdown as it removes whenever a bundle is the critical issue issue is that each time I start the platform x is generatedStarting x s doesnt matter if I shutdown correctly or not +QA testing for about a month didnt find an issue with the above workaround +Good idea nkeyway! Back to you +Except for that I dont know what else might be required +nothing to do with Roo nor primefaces +Thus this appender can not be found which causes a +Logged In YES useridIts ii and obviously this needs to be fixed +I totally agree and I opened for this +Fixed in subversion repository as of rThe same problem was also present in the add methodsThanks for reporting the issue and providing the fix + Obviously Id like to see the new code integrated with + +Id really like to see it fixed in would be great if someone experiencing this problem could run the debug code attached +java jar plugintarget a Directory of ejbtrunkaserverdefaultlib AM AM AM AM AM AM AM AM +Deploy packages are excluded too +I am not sure whether the issue is with File System settings +Thanks for persisting with this one +I agree with Keith the concept of coprocessors has been defined in many ways by different groups +This bug has been marked as a duplicate of +We need to do this +Can you please rebase +Thanks Ivar looks like just a delay as the latest update site now has the fix. +Verified on JBT. + to such a test +advanced proposition to use this dialog as a way to fill properties value for default +Merged the change to the branch and committed as revision Ill ask the JIRA admin to add a new tag to the Fix Versions and set the fix version then. +I upgraded to the latest trunk version and the persistence flag is now showing up in the stacktrace +Thanks for the patch +Just fix it. +Right now the only ways deletion is detected is by getting the version string for the document that no longer existsAlso FWIW I make free copies of in Action available to all committers upon request +please split the refactoring into a separate patchWhich refactoring? I dont think I did anything that wasnt necessary in order to cap the number of available buckets +If a token is available why not use it? Under what scenario do you envision a client for any external auth requesting a token and then not wanting to use it? If a cluster not using tokens wants to talk to a cluster requiring tokens then doesnt it have to send the token regardless of the local config +As I said three years ago the intent appears to be providing protection against OOM for unframed connections but since its languished broken this long and we presumably have way more modern framed connection these days than otherwise I suggest just ripping it out +Cathy this patch is for the current trunk not branch +If you are merging you should simplky replace the old BW branch by the new one I recently created for trunk +Is there an unit test dependency on this change +I am working on hbase support but it will wait after to go in +Seems that the time reported in the XML file is wrong +EAP contains Changed to so we dont have to update with every micro release +Cleaned up unused variables indentation extra file +Clarified the link relationship. +Some scripts need to be made executable before the tests within contribowl can be run +What exactly happens if you dont include instanceDir? Does it just use or does it use the name of the core as instanceDir? If its the name of the core then IMHO the inferred value should be explicitly persisted on RENAMESWAPThis will definitely be a lot better with core discovery +It sounds like there might also be a custom errorChannel in the other module +Thanks Dag thats a more helpful thing to emphasize +Yes I agree you should return the collective value when you ask for the superType +So the in Seam which used this attributemap to do its work doesnt work +just brower httplocalhosttestftl +Any commentssuggestionsThankssuresht +no changes in beta branch the change is in trunk only +What do you think +It was apparently removed when the charset was introduced. +The fact that other distributions have is irrelevanthi suresh I have to agree with allen on this one +It would have been useful +However I dont think we supply this database with Derby any more +I agree that it is a security hole to be able to send a URL that will bypass the check +I finished packaging for Mageia Cauldron and was dismayed to find out that it no longer handles the documents on my homepage properly +As the descriptions says we need the file name in the exception think of a simple way to pass the filename to the constructor of the +If the order of the flush is such that any C instance is flushed after A is flushed that C instance will have valid object id +This means we are not showing the total capacity for cluster right +And yes thats the right place to check the revision number +I just figured out myself that my above question is actually irrelevant to your problem +Committed Brendan Long for your patch +Updated patch with a few more test cases +Try instead +Please reopen if you have a patch that does the package moves and ensures that all tests pass. +Feel free to revert otherwiseThanks Stack for review +The problem is that the pom is not even parsed which prevents resolving of other dependencies necessary for test +update set to Harmony jre home +Will run the tests locally and check it in +The current way of storing hints using pointers just doesnt translate over to ranges +triaged for +If we throw an exception log the violation with an detailed message we can give the user a hint how he can disable the check +Reported issue is fixed +Ive granted you despot developer ll create confluence jira despot privs granted to can you please create a user in confluence preferably bernhardttom so theres less hassles when jira and confluence get hooked onto our ldap have created a bernhardttom account in have assigned you privileges in far as I am aware the only outstanding work is the list creation +Try requeuing to see if thatll get an entry in the queue +Thanks +Patch is not attd to this ticket +The framework should do no interpretation or quoting +Thanks Steve orig hardcoded that link and I made it relative then realized it didnt work cause didnt go out in dist just thought that was on purpose so went back to hardcoded +Added compliance document to docsUpdated sitemap to handle it along with a stylesheet to convert to forrest documentNeeds adjusting to get the colours in the table cells and dtd validation +Not really without a good amount of work +As far as I know our plugins does not list these dependencies explicitly thus im surprised that is happening +The use case is subsequent observers wrapping scanners created by those earlier in the chainAlso we should deprecate and remove the older more limited hooks now that we have a superset interface that admits more possibilities +this is a limitation of entitymanager when there are spaces in the dir removed all defult dirs that have spaces to reduce because we have another issue in jbossas about the exact same thing. +Committed as r +you RickCommitted patch to documentation trunk at revision +You can see the stacktrace using such a URL httplocalhosttesttesta original nested params are httplocalhosttesttesthttpAFFauthhttpAFFsreg the problem simply results in the usage of nested param names +errant release +the error message is not clear We should fix this in have submitted a patch in which fixes the error message problemThere is a new test in the patch such that it starts up a cluster with one user account and run distcp with another user account +Ill finish it Ill address Teds comments this week +The patch is a bit hard to work with +Moving out of +So therere problems here +User Error +Thats an interesting idea Yonik +Simple patch that adds new methods to do the checks +Max maybe you mixed it up with from Palette? We do insert namespaces when you a tag from Palette +Consequentially these issues can be linked and this one can be added to the fix list of. +We should document explicitly that the check done is valid MX record for the domain and not simply resolvable domain nameNot all resolvablevalid domains also have valid mx records configuredThey are different checks +Brian you were right about resources plugin +It worked but the large compaction queues were a symptom of a problem not the problem itself +as I understand the request is about reloading console configuration on java changes that would result in severe performance now marking as wont fix +This is a duplicate of +does work though +I need to have a better look at it. +Attach screenshots with steps for reproducing +Maybe towards the eow +Maybe a filter like interface for the descriptor enumeration therebq +String contains both single and double quote marksThat last ones painful +This seems a better abstraction then asking everybody to skip the files starting with +However it has far better unit test coverage than the rest of the cassandra codebase and more importantly has been deployed successfully in production at scale +I agree with you about the fact that REMOVEDTOKEN and STATUSLEFT are the only states we need to worry about expiring +Im closing it. +Committed to TRUNK +yesafter i use JAVAOPTSJAVAOPTS i do not catch the CCEthx +New issues require new jiras please do not hijack jiras for unrelated are not meant for discussion or questions please use irc forums mailing lists instead +Uses a different committer for the targeted unit tests that always return true for need +Invisibly failing over to the lib in theparent could cause the user major confusion + +still broken +Each terms enum is unique to a reader +Stuart has already patched more at the code I think this should actually be resolved. +Upgrading this to critical because the effect of this is that with rules at multiple scales get their geometries written out multiple times +Attachment has been added with description attachment includes a file some test images a and PDF and PS output of the test file +Even adding a +I mostly wanted to set the granularity of as returning classes with the appropriate implementation +Tested up on cluster +Verified upgrade in ER build +Patch test in +I see google widgets choices +Heres a patch which refactors and out of +This patch allows to be +teiidresourcesruntimedeployers +Committed to trunk and. +License we developed this code for a private client initially but then decided to give something back to the Spring community +These two failed tests pass in my laptop +addresses the same problem. +A way to run query asynchronously so that query progress can be monitored without waiting them to complete +Have a look at it if everything is fine and hopefully I did not break your usecases +really means entities association wo lifecycleParent really means pointer back to the entity ownerThere is an open bug for Parent AFAIRthe unidirectional issue is fixed and will be fully tested through another issue for betaSo this is half fixed half rejected. +PR handles model changes and binding UI + +tnx for input +Thanks for looking at the patch Narayanan +Looks good to me +I think we should add other segment events such as flush open clone close etc +Just upgrade to and you should be good +Which hadoop did you pass to the Giraph build for this run +Thanks for the patch Erik. +This may help a bit if there are rules about the precisionscale of a result based an arithmetic operation between values of known precisionscale +I didnt dig in enough to see that it was only Delete markers that made it through +I have the same problem use and best +Yep I was seconds ahead of your post so I didnt see it +The request response Via headers and Server header can be set via the variablesCONFIG STRING CONFIG STRING CONFIG STRING ATSYou can change the version info etcGeorge +Generally we want to limit the time code as its a right royal pain but no reason not to consider minor changes +I close this as it was integrated +Patch for the improvements +Fixed in this patch +So equals overhead is not negligible compared to the actual logging which is counts field updating in this caseBut if we think that the overhead incurred by equals is negligible on the overall performance we can use equals +Hi can you kindly point me where I can get example and see if it has the problem +Nah feel free to take it Ill be busy w the salt stuff +Well there is another related problem here that constructor proposals are not supported in +EricI understood the goal of indexing +The patch is ugly +Clearing patch available flag a new patch is expected later +Patch checked in + added excludeParams property to to take a list of regexes of parameters to be applied however it was against so it wont make it into Struts since it depends on the +bq +It is said to be a recoverable exception but is that true even across restarts of the ZK ensemble? If not is there a way to tell the difference between a network hickup and an ensemble restart +Manually tested +The queues are deleted when the connection is cannedHowever when the connection is lost the queue is deleted and after failover a new queue is createdHowever if we specify as the reliability mode then the above is not the correct behaviourHowever it was decided not to fix this for the current client as it requires more substantial changes to the code baseOne big issue is that The destination object holds the subscription queue +r to fix the latest problem +On a related note what do you think about using apply as the naming convention for any methods that need to actually modify and passback a new query and encode for those methods which just contribute to parts of a query being built up +Missing words +Is there a possibility of anything else happening here something like a disk failureNo the disk is working fine +Patch for by making visitTree not call setup before invoking the during a ppr visit +This is an improvement that might be considered in future implementation but is not function critical +Recommend you try a SNAPSHOT build of and see if you issue is resolved +Didnt put any hooks in for nD coverages since I have no idea what the interfaces will look like but it should be quite easy to update things when they come online. +Sanne I do not think so +The very first picture that I posted to this JIRA is from a workflow XML thats over lines and certainly takes a handful of seconds to generate and is quite a bit in size +Hi +Attach +Id bet in the JSF engine +I would kindly ask Renato to upload this patch when he has the time +Committed revision +It will be good if you can let us know the exact problem that you are facing here +Attaching patch and containing first drafts of three new topics In the Reference Manual these topics will appear in the section Setting attributes for the database connection URL +If anyone has an alternative proposal for how to resolve this then please with the suggested approach + +I know the correct TM config appeared to be used but I cant understand why node would rollback propagated from node as the orphan detection shouldnt touch them +If we fix for it would be diverging behavior +When it works it will just print the word out When it doesnt work you will see the word out and an key is that has a eclipse project dependency on +I pulled this into too +Youve probably set NUTCHJAVAHOME then +Are we talking about the same toppicI am not talking about the of your plugin documentation and do not talk about any +dup of bad Gavin must have hit the create button too many times. +Patch with unit test demonstrating the bug +for taking a look +Indika Please update the documentation pages and them mark this as resolved +Also the tests added in are substantially similar to the test added here +weird +remember they all claim to implement porter one passes all snowball tests +No test because theres no way to squeeze in a barrier +why not make a wrapper transport that just hangs on to a copy of everythingI only want to log write operations so my plan was to extend the Processor and peek at which operation is being called +Justin Ive applied the attached patch at rev in order to be able of editing coverages but require your review so please take a look at it when youre back and feel free to revert if needed and tell me if theres something wrong with resolved while waiting for jdeolives only issue with the patch is an aesthetic one +glad you got it working! + +Robert i opened to track the issue of moving into the correct package and committed your suggested changes under that issuei know for a fact that part of your patch didnt make it in the characters kept getting corrupted in one of hte files and i couldnt figure out an obvious reasonsolution so i just committed what there was at that point so wed at least have all the files in the right places to move forward from thereplease take a look at the current state of the files and let us konw what still needs done to make them good +So I had the idea to embed the web console that ships h directly into the geoserver web app +You could then do a simple recursive file diff between the instance installed as per quickstart guide instructions and a installed with RA via your script +When I downloaded the Diff file it only had files listed in itsrcadminguidesrcadminguidesrcdevguideAre these the only ones that you changed +appcasted files should of course be build artifacts and should be accessible from an URL that is also compatible with rubygems and! +Nick can you just I havent missed something to add a new maven feature to our build setup +There are two variables one called JAVAPREFIX and one called PYPREFIX exactly for this purpose +And aspectj is bloated and asm is inflexible and requires a lot of boilerplate There are pros and cons of every choice I dont deny it +Next Actions Put this information in the Graphics and Screenshots section of the Docs Guide +And since the syntax differs anyway for IF NOT EXISTS and IF XY forcing a high level API to use the rest of INSERT or UPDATE syntax doesnt feel like it makes their lives much harderBut Im open to making them fully symmetrical down the road if it turns out Im wrong +Its not a problem +Took a bit of searching through repos to find it +Verified +Hello I am out of office on vacation and will return on Wed Jan ndFor any urgent issues please contact my manager +We loose the jboss esb tModel because jUDDI during startup does not find the root publisher in the DB and runs the whole registration process again +Made the changes +What about this one +I just checked on linux and it works fine +does not makeas much sense +prefixes added to both standalone and server level directories now created for both in top level directory includes the minor version like standalone I would suggest that within the +I stumbled across it since is trying to create the missing locallog directories and ensure the permissions are properly set and the URI used in is for a path that typically cannot be created +AFS impls all end with fsId generally prefer FsFs is better +DEBUG Compaction requested for region test because regionserver INFO Starting compaction on region test DEBUG Skipped compaction of file compaction size of actions Skipped files size INFO compaction completed on region test in sec +Thanks +Ignore updating accesstime in get if the namenode is in safemode +Committed revision +this is a duplicate of would also notice that the VE polygons are grossly simplified why is it so +I will find out how much performance I can gain this way if any +Excel on the Mac puts a comma in +Here is the patch once again in a single file with relative paths +But it required running servers on multiple ports +Current shapefile reader can load all of the attached shapefiles without closing all issues that have been in resolved state for months or more without any feedback or update +Could someone please supply a release note? Thanks +Hey Ivan Great thing to do to remove the execution of elections from the zookeeper thread +The new version of the xsd file was not published when the site for was deployed +There remain numerous references in the build and testing trees as well but they do not appear in the javadoc so did not fix those. +Here is a patch to fix parameters parsing to get them better in report +To clarify nothing bad happens +Fred could you take a loook? Is it related to maven intagration +MartijnThanks a lot it works +Committed to trunkThanks Eli +pushed to master and also externalized the string +Another patch needed if applying this to the branch +Thanks +Thanks ElliottHow so? Wont all be well on the next go around of the balancer +Priorty was inadvertently changed to Blocker +Bulk close for +This is a duplicate of. +Duplicate of Make sure is included if client behaviors are rendered +or If this is a base type this element is not presentBut since it does not then this would be interpreted to mean this xselement +Let a test run all morning here with the above changes and Clauss fix no and memory usage remained in a stable pattern so this should be all set + +You may give it a try +I mean that when is created it is known that would breakIt seems that your patch is not generated from trunk since the line numbers do not match +thanks to Evgeny the fix make it to release +It is not incompatible. +I just committed this +Closing ticket +Automatically fixed by +yes but I got another patch that I think is better let me spend some time writing a post to the dev list +This is not within the timeframe of so I will move it to future +Marking closed. +check English and component functionality +The UT failed is a issue as previous Jenkins build with the same failure +If it is to be included in please let me know and Ill convert it from a Known Issue to a Resolved Issue +VERIFIED +Fixed +usrbin +Also I dont think its because of the file size +Ive applied this patchNiall +Guys please look at the last comment in the original issue +Although that bug is fixed the fix doesnt really help us when running the upgrade tests since they still test against the old buggy versions +for the patch +Expecting test failure of to be fixed after this patch +It wont be clear to users what these files are and howif they are relatedI do think that having this as a contrib would be fine thoughcontribfatjarcontribfatjarconfcontribfatjarsrcjavamainorgapachewhich can contain your mainClasses etc +Marking as optional since this is only really needed to convey support for additional convert functions +I hadnt noticed it myself. +One of the problems here is that is a terrible api +getClient and putClient no longer synchronize +How it is doing with another language did you test + +Might mark this as fixed for M then +Andrew Im attaching a debian side of things plus some trivial modifications to your RPM codePlease update your test code to run against a real cluster and I think I can then submit this to our testing machinery to see how it faresOnce again thanks a million for helping us add this useful project to Bigtop +Thanks for the clarification Sebastian what you said sounds be compliant with my proposal so Ill updated according to Commons conventions +Reopening for handling in Struts +Could be useful verifying whatever fix we have here + changes made in revision to buildersesb Commit message Add changes apply the SOA authentication schema to the copy the versions of ESB configuration to the reopening to update release note info +Going to mark this one as closed. +I will cut the branch from master on that day +Attaching a patch for +The actual class and its supporting infrastructure might need to move to some commons package either in Spring or this issue is related to +Great turn around on fixing this +I attached a file that contains target view of new Web UI +And maybe answer our questions as well +HowardIf you need more logs or try new jars just let me know +They can be set independently +I forgot to mention that this is in project mentioned in the comment it is already resolved +When is version gonna be out +I will retrieve the logs requested asap +The source package is the important artefact +As implemented we could change the contract to say anything instantiated through the that implements will be informed Alternatively we could add and to the list of valid componentsHoss do you have thoughts on this +Please report it feature comes from WTP structured XML editor +Make its own maven project +Committed to trunk +had some Path manipulation logic that incorrectly mingled forward slashes and back slashes + sounds like a great idea +Neither are currently in SB I Know there was an issue opened for the archetype but I havent seen a hellow world sample job added +Verified +Heres a version of this that still needs debugging as it fails tests +Moreover the point of the Avro RPC support in Common is to support Avro RPC in HDFS +Thought Id share that with anyone else who wants to try it too +patched files are +is fine as is +Builldr just passes the argument to javac javac decides what to do with it from this error looks like javac refuses to load so its not even looking at the argument +The issue was investigated and it was desided to use common description field for this purpose nowl default value are come together with attribute description +Yeah using CMAKESYSTEMNAME to determine what platform flags to use seems totally reasonable to me +It might be worth to use the xsd from the zip file thoughNevertheless the problem still exists +I believe that publishing data to a public repository constitutes redistribution +This creates a choke point for later policy enforcement +bulk move iteration added javadocs removed all nocommits renamed getReader to get used a instance instead of a string in all get variants so we can use a cached instance instead of looking it up for each invocationI think its ready +Implement so we can suspend our transaction when Hibernate wants to perform non transactional m afraid this is simply not supported with The whole point of that option in combination with a is to always expose the same Connection to client code within the same your particular case why arent you simply using? That should properly support Connections for id generation and also expose the transactional Connection to JDBC access code like does + +It will be useful to learn a bit more about the testing framework +What I found was that numberguess does eventually get deployed and will work properly if you wait about minutes after step d +Andrew +I opened this by mistake +Would you mid the scenario with that version Dean +Are there any workarounds? Getting nervous because this is a show stopper for our app +i havent found differences in wording so i presume the change is done to limit length to characters +is fixed. +Looking at I can see that it was done in shell bundle not in the fileinstall oneSo that was the bad design idea that timeShould I move the issue to fileinstall component +The patch process still seems broken Giri could you please investigatePradeep since the patch has been sufficiently tested and passes outside of hudson please go ahead and commit it +This same configuration works on EAP +SELECT FROM statements so that we dont run into a later in the compilation +Works fine with deefdccafcdb +Please could you give me some hints? Willem I am sending you the patch against the branch including tests please could you navigate me to eclipse formater so I can easily format the next patch for the trunk? Best regardsAndrej +Sample project which demonstrates the problem +moving to M +Just a debug message +In this case a solution may be to detect and decommission slow data nodesWhen decommissioning datanodes and tablet servers would need to ensure the whole cluster is not decommissionedAnother source of slow reads I have seen is at the network level +Jasmin test source close the bug since it is fixed as a part of bug +Thanks Elliot. +looks good to me +Assume closed as resolved and released +Due to not use the handler we just register in the reusing handler will work well and not fall into the infinite loopSo this bug may be difficult to reproduce in junit environment +We took the general decision that the activiti engine does not do any security management access control +Please disregard my previous comment it is wrong not looged exception exists but in different situation +Thats a good use case and I agree with everything you say +Yes I was planning to putting this in branch and trunk +I have started work on this +Thanks +Please no intrusive message it as an completion proposal +I can do that if you would like +Ivan I agreed with you +Perhaps it is a JDK difference +Added to the ujax module in Rev +This is a BLOW OFF by Apache +i guess fixed means we just have to set maxallowedpacket variable +thanks for catching that +Hmmm +Tested the example and it crashes with a few other issues first +yes i am in the process of that +I believe it would only need to be done in the manifest parser similar to how the bundle capability is currently added +OK +Ive recently been looking into and really like it +If she fixed the TCK error in naming the parameter typeclass the next failure is this one +Also fixed in all tickets +I believe the checkin for broke this test +Added test showing that BPP registered via definitions are not applied +Im using which I believe is based on Eclipse + +fieldName onlineAuthor type +if the schema is a mismatch then flag it as an error +I added a detailed description of parameters to bookkeeperConfig +Its wrong so dont do it! If you dont know how to avoid it then you are certainly messing with that you shouldnt be messing with +Why is this not compatible with ArchivaNick +Equal tuples will get increasing rank numbers +I think the relative path is enough +Next version of GAE has built in unit testing support +so this is not a bug! this is working as designed +I dont know who should make a long as it got a from a commiter its good to goSo any commiter should feel free to commit it since Edward gave it a Note also that it will still need a few patches since I dont see any service script +Thanks. +Attaching the new patch with hemanths comments +Thanks Bob! I think we should commit this +Had attached commit message rather than patch +Patch containing notes to add to collation key filteranalyzer classes javadocs +Thanks for your work on this Kim +Fix applied in revision included examples should now build if mvn is run for the examples directory for example the basic directory +Done in r +Patch is based on trunk +Maybe this help +No harm in lowering the priority reassigning or just letting it lie fallow +Flusher goes to work but any subsequent gets will not see what was in snapshot until flusher finishes +The was a bug in my own test code +committed some template changing issues see still some minor quirks and still ok to add methods to though I would like to keep them to a minium and allow usersexporters to add these and own a bit easier +I also created a second version of the Aggregator example that uses the offset function +This is my sugestion for the +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Ok Arjen thanks +Ok +Hi Lars i have small doubt regarding Root and +However I have a couple of concerns about this patch I dont think that generated files should be checked into the source tree +First take on using filestamps in lieu of md of crcs for the filecache Ill continue testing this +I just committed this +It sounds like we are telling users that we are performing a design timeoptimization and not even providing the alt code pattern +This is handled by other JIRA issues. +Does reinsert a failed job back into the job +trunk Committed revision branchx Committed revision +This version tries to do a better job of reporting more information before exiting after tripping across a corrupt record +Attached patch +are we going to postpone CP until June +Updated to work with latest patch +Uploading patch to terminate unmarshalling for the first error or fatal error +We have discussed modification of this pull request thats why Im rejecting that +File used when mysql persistence for JBM doesnt work +javax +What other tests should be added? After checkin file issue for lock timeout if column is not accessed by JDBC +This will be partially addressed by new Best Practices topic in the documentation and the improved Click Examples. +With I havent had this problemThe same way i fixed this with a new project as I did with a old project +Just to clarify When Robert said if you ever decide youve done enough andwould like your name added submit a patch I believe he means provide a patchto the file which contains the author information +This is basically a clunkier implementation of right? Should we close it as duplicate +labels +I like the changes but appears to be a real issue where the test is not finding the needed token identifier +I am using trunk revision fixed in trunk revision Part of jira issueAshish Vijaywargiya + +Its now only difference is +I only review as Pranay tested successfully +Globbing is not internal to HDFS its purely client side used by for any file system +please test how things work +Fixed the pkg names and moved into srcTesting is at the moment +Thanks! +Fixed that now by using theIncorporated the rest of the review comments too +Goals to run on project import Goals to run when updating project configuration ProjectBuild Automatically is user settings file contains nothing special only a http proxy definition +Hmm +That means for the common case users are going to have to wrap calls to this in a try +MVC does not exist +After much debugging Ive discovered that this is solved by the fix to +Derbyall ran cleanly against jar files and against the classtree on jdks and +info is added to +I will not address webpoller in this jira +Anything other than Major aka never implement me +Thanks for your time Tooru. +Yes of in trunk +Thank you for your interest +could you give me as attachment the whole logmvn e Xdashboard +Maybe something is now broken +I can into the same problem +So I wonder why you think they are not? Thanks for any clarification + +Patch with all of the graphic files committed with revision The files are +We may want to push down into Lucene eventually +no more exceptions and I am seeing HH being disregarded for the node we removed now +So I think the issue can be closed +For now used the same property names as in +Bill sorry it took me a while to get to this +for the latest patch +Thanks Tadeusz! +We could add a neutral abstract class that has those commonly used methods and both can be extended from it +You could also verify that externalsorting occurred by watching the sorters behavior in the debugger I suppose +We can forget about coverage tools imho and thus about pyramid. +Tests still pending +but thats what my testing points to +Thanks Awdesh your changes are in trunk at rAshish Vijaywargiya +See Global Channel Interceptors Configuration looks a bit confusing would like to simplify it interceptors can now be applied as described in the last comment +Clearly its not from the. +Jason can you make this change? At least incorporate Pauls patch +No worries I was thinking that you are waiting for a review of your patch +Rajika what IDE are you using thats complaining about the Override? Have you changed the default java version to be within the IDE? Are you able to successfully compile run the generated code? +I appreciate the compromise +So there is no compatability problem but the scenario of partial enabling may not work +Attached and updates +Let me try to understand +With in also this is good +This ticket will be automatically fixed by fixing as soon as is out i will test it +Committed to trunk +Couldnt we test it here Its synchronous to avoid getting the process mailbox flooded in the unlikely case theres a high rate of updates to the same ddocAnd the updater shouldnt be crashed +Merge with branch this pelase +I cant remember what the current one was but it can be sr or if its currently my gmail one it can stay as that whichever is easier +You are right I committed a partial patch that changes encoding to +released! +Patch looks good +Committed in r +for the patch +Yes! For some reason with my torquebox configuration even compatible code isnt functioning unless I specify anyway +I applied the Lars Trieloffs patch with a slightly modified failure message indicating where to look for the problem in Rev +The whole order items view is very cluttered with lots of information already and Im not convinced adding this makes it any easier for users other than developers +Checked in to trunk and +JayIve committed the fix into module +See other comment +Checking why this is the case +I was originally thinking it was in use exporters which would be ones that actually have a wire to them but a resolved exporter is simply one that has been previously resolved +I will update the patch +Thanks for helping us out at Felix +does not sound secure to newbies but Im not sure what a better name would be +Closing this issue because the root problem is gone +Also should backward compatibility be maintained just for the previous version or for all old versions? If its the latter the code may become harder and harder to maintain going forward +If there is a better solution to this I would be interested to see it +I think that means that the IP issues have been settled and that the felix jars are distributed under the Apache license +Actually its not in +The second is the of +Please let me know if this causes any is available in rev . +It will keep track all the time instead of only at the time of checkpointing +part of release. +Maybe its useful though +As part of modularization review need to look at new code paths versus old code paths for various features +I agree with Daryn we should figure out a way to selectively open the +Only than the input CAS would be returned to the client +Command objects will be implemented differently +Patch attached +Another important point that must be addressed is the package structure of existent tests +rev correctly identifies across server restarts +If resolving the permission resolves the test can you please close this? If not please reopen with the script errors +Since falcon does not support for the same cluster name to be defined in both source and target definition of the feed I had to create two cluster with same end points +Thanks! Perhaps its a good time to rename this to or something since it does work similar to that launches managed AMs +Lukas is this a doc dev or QE task? It is not clear to me +My observation is that this problem occurs when downloading maven dependencies takes longer +So yes we will work on that for response +We cant change the job apis to add one to force kill the job for +added as fix version +this tool can also take one or more regular expression patterns for region server FQDN for user easily useI use this tool on our internal operation so I think that other people may have the identical Andrew Purtell I put this comment on the wrong ticket move it to and thanks for your reminding +Probably we can change how the unit test does verifications later on so that we dont need the special countertracker +Need more info such as operating system etc +Ok thats plausible +I filed to follow up on fixing it for both Linux and WindowsI ran the new patch locally and confirmed that it works +Thanks to Uma +Without this fix back porting of fault injection framework followed by reusing it from Common will be quite painful +Yes I agree your proposal would work correctly. +Proxy settings can be accessed via from bundle +Logged In YES useridSorry I didt see that there was a version of jbosson which I cant reproduce this problem anymore +Uncomment drools related code in class changes reverted to previous +no code problem. +forget about my patch +why is there any difference in behavior here +Ill take a look at it soon +The mind boggles + +buttons in flat grey seems to be okSo long Jacques i tried its working in both opera Firefox properly i dont know whats the reason at your side? AnkitI suppose you tried with the last trunk version +Oops missed in the previous patch +Thanks Rick! For now I think Ill just go for the patch as it is +Im using the old style portgroups but still I dont know if it has introduced some changes in the APII was thinking that if you have the time it would be great to meet on some IRC channel or whatever so we can narrow down this problem? I dont think I can get a vsphere for testing but Ill see what I can do about itThank youXavier +Created to continue work on it. +Other option is to create a file and document designcode in place +Id like to stay the default behaivor for use sendsync for backward compabilityYou can set synchronousfalse attribute for the cxf bc consumer endpoint to use send +The above branch targets however it can be backported toIf anyone is interested in testing I would refer you to Stu Hood for advice on setting up the node setup scripts +bulk close of all resolved issues as part of closing items. +Committed +Logged In YES useridCan you attach the WSDL file +Ill try to get some time to see why this is so +Note Patch includes test case as well +Please have a additional patch has been applied +Right in that sense its the same problem as +I will send a message to mailing list just to inform peopleThanks guys +I dont have a good way of distinguishing that situation from the manual saveNamespace where we dont want to recover EDITSNEWI dont know that we can actually fix this with the current set of files and determinable state +The udp stack and stack now use the UDP config. +I have used workaround provided by you +My plan is to create a new mechanism on the Java broker for the which will convert the hash value read from the file in to hex and use that as the password +By the way the code above does not work +Input send as initial messagesThis sounds clear how do we split the data? And how does this going to be send to the tasks as initial messagesCoordination between client and BSPNot a part of the IO System is it? Sounds like a more general synchronization between the and ourInput has random read featureThis is great but a lot of work? Additional feature +Are there any other ecj related compile issuesId rather fix all related ones in a single JIRA issepatch +Please provide this issues Cause Consequence Fix and Result information as is usual +Diff and patch attached +How likely is that? If we see a problem here then we could never utilize event bubbling. +A checkbox has been added to control this +Looks like the unit test fix didnt make it to branch +We shall add more functionality based on feedback +Thanks. +for the patch +This isnt really a bug its more of an announcement +The patch has been committed to trunk and. + and will be patched together +rev rewrites the jetty integration to use the Registration interfaces and an info tree +Hide it from the node tree in +looks related to which is fix in project that shows that this issue was fixed in version. +bq +didnt get used +Patch applied +I checked out the branch +was named in case you cant grok it +Okay it was the slash in the promote attribute that was what was different in our I fixed this with a combination of changes to Deployment and which wasnt taking the slash into account like verified the fix using Jiris attached test case +Seems like a problem in our calculations in itself +Committed to and trunk +Commit +So cant fix that for Beta +So dont worry about thatThe attached tar file does not appear to contain the source for the implementation +I looked into this beforehand especially the seda endpointIt was a bit confusing as the seda endpoint is kinda like an internal JMS queue so the uri notations is kinda like a queue nameSo the queue name was the unique but the parameters didnt matterBut in Camel land the endpoints differ here is whyLets imagine we have this endpoint already registered sedafooAnd you want to lookup this endpoint sedafooconcurrentConsumersAs the code is now the two endpoints do not match and you create a nd endpoint +Hi AntonioThanks for this good job! We are currently setting up the project in order to host all Sonar plugins +I will update the patch +The multichanges plugin has had some difficulty bc adding multiple porject intelligence can be more challenging then it seems at first +was created to fix an aspect found when working on +Ive been researching this issue and believe that I can put forward the following upon there are two outstanding issues regarding this Improvement request JBI container should be able to initialize all first and then start them +Which one did you meanYour report is not a bug but a request for extension so I modified thecomponent accordinglyRegardsVincent + +This look like a duplicate of to me +seems the patch is ready to go +Also this is being showed because I added a query show tables +Ive only recently got involved in the but Ill look into it +The now has code to do the actual web service invocation tests +One example is aggregating all KVs for a row into one and write that value +I guess this issue must have been resolved in many WAS in market +IIRC there is something about this in the core model +You query the whole collection unless you pass the param distribfalse +There are dozens of ways to do this adding an additional mapping switch is not one +I think I had it left over in my model from when I was experimenting with the AOP mapping +is the same asJust add it to make it more clear +Can you open a reviewboard I had a few inline comments I wanted to make +Hey Nick yes can you pull a mirror onto? Thanks +Other than the coprocessor acl table is not special to the rest of hbase so not sure how we can handle that in the least hacky way +ok I workarounded the issue as described +Verified OK in Automated does not work with assertions enabled +Incrementing priority to major +There are no getters in that case +It seems I implemented it the other way around +Changes based on comments +ThnxI had to use the followingaftablecontent trhover afcolumn aftablecontent trhover afcolumn aftablecontent trhover afcolumnaftablecontent trhover afcolumn Tricky for sure +Thanks Slavik +Had to modify parts of the new streaming code to pass on the type of streaming operation performed in order to incorporate the clean operation +The last time the progress was changed +Unfortunately when it fails it blocks the C tests from running so I dont know what this patch is doing to the C portion of the build +We also have tests that Ill submit separately + +Try a method called get or set +I would like to make sure that we get the numbers with this recent patch +Assigning all documentation related issues to final since they dont represent a change to the codebase +We will close the issue after that +closing issues resolved for final +The server view shows that I have two servers configured +We pan on moving to maven maybe this problem will be resolved with the upgradeNot in Blockers list moving to next believe that this has to do with buildmagic not passing on the properties when ant is used +The dependee will now do a complete restart avoiding this migration +HiYes we are using now JDK I think since it version is working with JDK +If data is still available the connection is closed because going processing again would be a little more complex +there is still an ongoing discussion about did see no further comment on the list since then +version is given by version +Verified by Oliver. +Changes look goodPS there is one javac warning +I think we should use the same for calendar reported some time ago by Chris Howe at sorry for duplication jiraI think you right its the same problem I recreate my patcch to correct both issuethx thanks Nicolas +Even UCS uses the ASCII code for this. +cardinality statistics only tracks the average number of duplicate values per value in an index +The convention seems to be taht that if patch is small or a doc or site change go ahead and commit otherwise its good to get someone else to your patch its someone to share the blame with should your patch break the world +is the fix +If we identify any we will reopen it +I believe the first version addresses my problem and allows me to modify change its behavior via properties +That being said I do think we should probably change to match ANTS +works fine. +It applies there +Fixed typo in comments +Site looks good now IMO the refs to xsddtd should be added as soon as we have them see. +Removed the startstopcreatedestroy operations from the ESB Statistics level +The specific test case I think is test +In my rush to complete the expressions feature I forgot to code in a check for empty strings in groupsexcludedGroups +In general I see two different ways to solve the problem +The issue affects also Opera browser +the output log is stdoutstderr from before logj is initialized andor libraries that dont use logj +I think the miscommunication here is that Greg is using the command line mvn eclipseeclipse which makes the +Perhaps we could have an Excel response writer that could create a spreadsheet file +Unfortunately I have seen plans take many minutes to compile depending on how complex and some customers often the plans at application startup set the cache size such that the plan is expected to be in cache +I have just commited the generic array converter also created a JIRA encompassing a whole load of Converter refactorings +Purged +theres something else weve missed then. +Im not going to try to do anything about this at least for the first cut +dtd +The proposed fix would probably raise security issues +I looked at the code and see now what Ate was referring to +Im going to create on this issue for the work I still have left to do +Does this make sense? I can certainly document it though +Moving to TBD EAP +Otherwise +Default uses IEEE R Decimal formatCommitted revision jexltrunkjexltrunksrcmainjavaorgapachecommonsjexljexltrunksrcsitexdocjexltrunksrctestjavaorgapachecommonsjexl +Parsing should be a bit faster now +You just need to set two properties and +Otherwise there would have been circular dependencies with util referring to classes in other packages and other packages referring to the in util +the versions are in sync now. +Some points to note are If rn option is used anywhere instead of nr it does not work +Do you have an idea of whether youll commit the patch or fix this in a different way? Thanks for looking at this issue +Sadly I dont have time to focus on this for a while +Downloaded installed and verified the new is present and contains the JBDS Teiid Designer doc per Barrys comments. +A datanode has been moved to a new cluster or is rebooted after a long downtime +This should make a big difference. +New functional test apps are welcomed aim to have many different functional tests that test a range of things from installing plugins to scaffolding and so to me that your integration of my change was a bit fast as you use grails checkouted sources not the release +Fixed with rev Thanks for the patch Delos +Added more units tests for the client side plugin +Are you saying that simply creating and closing a in multiple threads will do itAnd im a bit troubled on how setting the callback will actually affect future instantiation +Test cases are modified and Ignore annotations are removed +Most of those decisions need to be made dynamically based on ram availability and growthwhat I was saying is that the user should provide its intend so the codec can optimize +For more general service description info I suggest using the admin servlet. + rocks +This is the number of ports used in +Short of writing a tag handler Im not sure how facelets can perform the correct type conversion since there are no gettersetter methods indicating expected attribute type +Ive marked the configuration properties as experimental so they can be removed if this turns out to be a bad idea. +I already expressed my views hereId prefer making things more consistent always requiring def or a type but not both at the same time +verified by Anton +If we can afford the weight might be a decent fit in here +I actually didnt mean to set the fix version to +We have several different QA CI jobs using as well as a personal one on +Its fine to close this as invalid +Its intentional ve also tried adding an per your suggestion elsewhere but due to a typo in the file thats not working +Is any chance that this will be integrated? Or exist any other solution how to get unique incremental number from GIT for tagging build number +we released wo it +This is the first checkin that changed the default skin +Patch applied many thanks +Patch combining v and Stacks test +Interesting discussion +A script named scrunch was added to the project that when run launches this modified Scala REPL +Verified javadoc is fixed +Lets break this down into smaller tasksBasically the large messages should be stored on a local disk before considering compression +I agree that the methods classes should be in the messagesparts +also optimizes a lot of queries mapjoin followed by groupbyThis is great + Username rodburgett +I just wanted to use LIMIT and searched results in ROWNUMBER of derby before +Should I try harder? Please note the mass indexing problem is solved and the PR is ready to be integrated +It means that the method dont return what what it should and should be renamed to estimate but as we would to the same for all of the rows in the cache I dont see how it would harm LRU if we do the real calculationIf you are getting hot n keys it will be making hot keys even more hotter +Also the problem still exists on exception is raised because methods dont support command object or as method arguments +Marking all current RESOLUTION LATER issues to be against the Nightly Build +no longer needed after the install plugin update +And also after apply the patch there is also no error +I believe the synchronized keyword takes care of that but waitnotify are admittedly subtle +Nevertheless I put the priority down to minor since this is not critical nor breaking the user experience in any way +Both generate messages +fails though so not committing yet +Seemed not so think this optimization will become more useful if it also considers the limit in query since in most cases queries is accompanied by limit +I will wait a day or two if somebody has objections +The dialog works for me apart from the redraw on close issue that i can reproduce +Do we have an update on this? If this is still an issue Im willing to take a look +Yup + Program to reproduce dual boot from two classloaders in LinuxChange DERBYLIBDIR to derby directory and run without derby jars in our classpathjava cp +I will come up with a solution for that soon +I think Its gonna be too tricky to do that with the current state of things +Pondering +Are you planning to continue working on this +Just create a forrest or forrest and alterdelete as requiredIm waiting for a moderator to approve my request to join the dev list +Ive only tried them on Windows server +I just committed this +Is there no cross platform way to achieve this that doesnt involve directly checking the string +This issue is reported against a version which is no longer maintainedwhich means that it will not receive a fix +All regular and integration tests now pass +I didnt say I was interested +Ive committed v to the snapshots branch +Can you please create a patch that makes the change I described above and submit it for review? Thanks +Fix included in release. +A build option XMLMACOSLCPTRADITIONAL is available to revert to the previous behavior. +Do they need to to the new lists? I thought that would be transparent to them +Ive encountered some brew issues over the last couple of days which delayed this guide being pushed back upPlease revisit the and look for the guide under the EAP groupSorry for the inconvenienceCheers review the MC user guide from stage not the attached version +updated description to be more descriptive +Thanks +Thanks Boris +Shall I raise another jira for thatThanksVikram +Also can we make these two parameters configurable +I had modified version of this file locally +Sending srcmainjavaorgapacheservicemixcommonxbeanSending srctestjavaorgapacheservicemixcommonxbeanTransmitting file dataCommitted revision +contains the bug fix +The whole idea is that this string searching logic will be a generic one and mtom caching thing will be benefited from it +for this release forget the synonyms just have +Put an error message in +Can you confirm you tested with Sonar RC or RC +The style sheet in version has some changes related to positioningIt is possible that your application is still using version +closing issue +I just committed to and trunkthanks Ted. +I just committed this +You think we can fix this? Infinispan was released already so I dont think we can change the binary format anymore +I think I will apply anyway shortly the code is cleaner looking +could you provide a patch +stack if the class is in hbase core the users could run different versions of hbase without upgrading +Treat unscheduled issues as Native SOAP stack is not supporting JAXWS any more +Ok +Thanks! There seem to be some WFS sources that still casues problems +Instead for an externally managed transaction such as Springs youre apparently supposed to set to in which case the transaction activity check will delegate to the JTA subsystem only and not insist on a local active flag within you confirm that you were using in the setup above? Does the behavior change when you switch to? from documenting this in Beans javadoc Im also adding a set method similar to what we have in the Hibernate variant +Tests are added and functional with no visible errors +Seems like this is a regression since +The instances of initialization still shows up as our limitation +Maybe its then more a WeldCDI spec issue to provide a way to skip the processing of any given extension as I initially thoughtI think the easiest way to provide a switch would be only to provide a java D parameter as its also easy to bootstrap programmatically +Other than that the patch looks good +So when a programmer works with Oracle database he wont spend any time to make sure same columns names have same case +You are right that the XSD file should be versioned I will update this in the TRUNK ASAP +GlebPlease review the issue +Let me reopen +Im sorry about your deadline but it is unlikely that this is an iBATIS issuePlease dont use JIRA as a support mechanism use the mailing lists +Without a fix on this one has to add redundant exclusions etc +Committed revision +changing operation to getGreeting results in expected behaviour +But anyway glad that it works! +Unreleased does not mean that is not out in the wild it think it lived in the sandbox for three years +This is not a Clover issue it is a JavaAWT issue +Incorporated Amars offline comments and uploaded the new patch +for the patch +Thanks works +Since this is a Datanode configuration problem I am closing this issueI will create a new issue for improving Datanode handling. +The exceptions like the reported seems gone but as Tina pointed out the failed to start for some reason we cannot tell the specific from the console output and +Just submitted the dev list. +Ok I think this is all fine +fixed in trunk. + can be used for a backing meltdown +fixed in svn r by applying jukkas patch +Heres my patch for this +Love to use this for testing out Whirr services likeDavid What has to be done so vbox processes also work under Linux +Irrespective of what is going to be decided here one thing is clear +update to trunk +Attaching a patch which +I even moved km +Set and patch to remove unused classes brought to our attention by Kiran Chitturi +I cant reproduce it by turning my network on and off or by turning on and off +Now packages are +Assigning to myself as Ill be handling this kind of thing for M + +Both approaches have legitimate pros and cons +It seems to me that we should have added the flag +Great! Thanks for the pointer I saw only in there and somehow missed the latest version +Actually after a bit of review this issue isnt a regression per se but rather an intentional breakage of backward compatibility +The remainder of the functionality offered to Castor will be integrated afterthe release has taken place +I have edited the stack overflow code which claimed that it was fault of formform without model attribute but that wasnt the real issueThe problem which it caused was not because the first form didnt had modelattribute I have edited the first form from stackoverflow which is the same result +went in on Jan th +All it does is translate apostrophes to two apostrophes which is not worth the method unless more is implemented +Please sorry for not verified sample. +This is not so +OK I didnt know that its the exact same string used here as in the other jira +Tim what would be the recommended size of a paging fileWould be sufficient to work with Bridges? I guess a better question would be what is the default forI will build and test this over the next few days and would like to know in case the docs have not been updated! +Open IE with google on the first tab Log into on second tab Close tab while applications loadIE goes berserk and opens ifpc call is gone now and I havent been able to repro this +I will start to work on the merge now +I will separate the job and setup more tickets for different target I will take this issue as a key to cleanup all the related issues + +Hi ColmOk will do soIn which namespace should the custom GCM suites beMarcHi MarcOn second thoughts lets leave the for the moment +Can anyone from the Spring Security team comment on this issue +It works fine +Looks like the same reason caused +adding this committed in revision . +Solved by reverting to +on is available on both places +Did the tests help? Is there any other way I can assist +Committed revision . +update will keep the region in transition + +I think we need to put in before committingAlso I dont know ANT well enough but is there anyway to have it automatically do the contrib packages without having to explicitly specify it? Just one less thing to remember to add when adding new contrib packages +We need to consolidate the attempt retries discussion along with what is being proposed at +dcaveneywould you please add a simple xml input document that has the problem you describe to this bug report and its corresponding outputThanksBrian Brian +Its interesting though whether the commented assert in still fires on full patch +Thats probably excessive +bq +The errors are still the only errors present +I was thinking a codec could implement the Decoder or AND +The attentive reader will have noticed that it should be tar xzvf in my previous comment +Patch applied at +Now the test works for hadoop and +This does not actually seem to be an issue +I just committed this +Really should be moved to the installer project +That said if there is anything we can do to improve the situation were going to roll it into and +These tests capture well the correct behaviour +It seems like a trivial change to not check for disk error +I dont understand scenario? testcase +Attaching screenshots of the warnings I get with a vanillaIve never touched lint settings +Jayapal I marked the issue as In Review +I also changed Fetcher and Fetcher to use instead ofI tested this with a small url crawl and then used and they all work successfully +Thanks Mike for the explanation +This should be handled the same way as in other plugins +Brett will you push the fix before release? +i will no add any WARN logging output +Fixed for Burlap in this fix as an umbrella +I didnt manage to reproduce exactly the same behavior but on my side all measures on rules at project level are deleted on the last snapshot +Thx +Thanks AlexeyPatch applied to LUNI and SECURITY modules at repo revision rPlease check that the patch was applied as you expected +Since there were no further comments Im marking this as Wont Fix for. +junit test confirming header delivery of JMS properties +bulk defer of open issues to +JIRA cleanup +New packages will be created under as follows entity containing etc activerecord containing etc identifier containing Classes remaining in root package are +The close is often omitted for error conditions causing slow leaks +I added some checks for the cases you pointed out and resolving the issue now. +Andreas is right it is an application developer responsibility however we can clean up the template in the demo to ensure that an exception is not raised when default values are entered +There are couple missing dependencies in various poms and most transports jars referring to unavailable versions +What if I came into your house and painted in large letters if things were better for me this is where I would like to add a sunroom if I ever have enough money on one of your walls +Punting issues from RC to final +Are you sure it makes sense to share the code for getting the current user? It would just be replacing the call be a wrapper to it and replacing a trycatch with an ifelse +I plan to this integration based on +This patch uses the name get +Thanks Aaron! +Purpose of hidden in It is used to select only the fields required for our target service from the input map +patch to exclude targetresources from source in SVN. +Does this typically with high or low loads? We just want to know what to look for. + on this +bulk defer of open issues to +We should document the change so people should be aware about security implicationsThe spec doesnt allows it +Looks like there are still tab characters in the new patch +Can we mark this jira resolved? +Closed per. +Implemented is a pass for x copy of arguments for callva +Closing issues which have been released. +Any takers +Youre right so I reopen this bugAnyway have you found a real case where this problem appears just to explain why it is good to fix this problem +Verified fix is okay +Please be more precise and tell me what you want or need +There is no real way to fix this without changing how the installer is put together +Are these two issues duplicates +I need to look into what is going on there +This was fixed whilst fixing other issues +I could not find any license information on the project site nor in the code itself +Thank you for your contributions Andrew! +Im happy to give the related tests a look if you have a quick cheat sheet for what to check out and how to build as it relates to this issue +Thanks for the quick fix Thiru +In the trunk we currently carry all clusterIds in the replication path and we could optimize this later when there is a need +Please verify and close this Jira in case its done. +sample connecting to As always comments and suggestions are welcome +Mylar has been mentioned as a possible way to do you are still eager to do this without an layer I suggest that we at least look at something light that will get us in the right direction +I had a look at Hadoops implementation and this did the trick +Backs up both the localStorage and databases and fixes +attach my patch through hudson +You have just replaced the mocked objects with a real instance of Token +Please let me know if this also fixes and mark that issue accordingly +See for background +One line fix took care of it failing locally for me +Roland do you wish to bring the other comments you had on dev over here and integrate them into the final patch? If so that would be excellent +Jut for completeness I make use of SEAM project that contains all source code and that you can use to build +I have encountered the same issue in the version of hibernate distributed with +Please feel free to if you hit this on . +we dont host anything on +This will show that the path being passed in is truncated at the first spaceFor exampleCurrent directory is CProgram FilesJBOSSHOMECProgram FilesJBOSSMODULEPATHCProgram FilesmodulesFirst call tocall JBOSSMODULEPATHorgjbossremotingmainecho pushd The echo will print outCProgram will start but no modules are added to the classpath +this is a document i wrote about this in +Distribution Management is my mistake The code is from Ant do you have Plexus style I can applyWill make another attemp the code comes form Ant does it really make sense not to use it directlyIf not possible I wouldnt change the package names or formatting so future diffs of both codebases are easier +I attached my little projectPlease just run the with the current configuration to reproduce the issueHope this is helpful +This was done in conjunction with +David Id like to get this rolling +rev trunk rev +Hi DavidFirstly my apologies for the delayed reply +Thanks Konstantin! +These tools are tied in pretty deeply to the format +Integration tests were updated. +Thanks +Im using mvn help on the child POM and seeing the behavior I expect where the junit dependency in the child overrides the scope of what is stated in the parents depMan element +Implementing this will take a lot of effort +Tested +Changes to distribute wrapper to distribute wrapper thanks for patches +This sounds a lot like +The patch did not include a new file +Change the syntax from START to LOADSubmitting + +Please vote on if this is a feature you consider important +Thanks anyway for the patch +Ive took a look at the test plan and I think its overall complete +Is it better to provide two patches one for refactoring and one for locale supporting please? By this way the change will be more clear to read in SVN repository +Prior to conducting the update I verified that the error still occurs and I left the partially edited file in the work space +db files are still getting generated +removing imports that are no longer used +Attached v patch same as before but fully rebased to trunk +We should probably open a separate bug for changes to the loading process such as using get and so on +I wasnt aware of it +Close +Then whould we not closeo +No problem if anybody wants to jump in but i wont be able to work on this for the foreseeable future +is there a JIRA for this? Service Contract Definition on page describes this feature +After checking there are some changes in two files are omitted when applying the patch to branch and they are consolesrcmainwebappviewrealmwizard consolesrcmainresourcesTalked with Ivan offline and he will apply the patch again +Thanks a lot for your timeIve been busy lately but I plan to go over your other patches soon +applied and built and passed in my environment +I have not been using that system too often lately so it is available for a bit if desiredThx have DB left installed on my Gentoo box at home +Moving issues that arent going to make it into to already got the first phase working in my +As description says it is for NN also raise a new jira for hdfs +In r Added a German flag to switch to German in the IN example +The Issue still stands +lets verify this against the latest have had reports of this before but not a reproducable situation +Hi NirmalIt sounds like you have some ideas about some possible fixesWhy dont you try writing a potential fix and attach it to this issue as a patch proposalwhich will give us something more specific to discussAlso when you write your fix it would be great if you can also Run the current regression tests with your fix in place and report the results Include as part of your patch some additional regression test changes whichverify that this issue is fixedIf you need help enhancing the tests just let us know what problems you encounter +It works after you add the certificate to the JRE CA keytool import alias toto keystore hometheuteToolsJavajrelibsecuritycacerts file mycertificateWhere mycertificate is the one created for SSLWould deserve some documentation +Showing the available containers as in that case makes it clear why more containers arent being allocated even though memory appears to be availableI agree though that users would rather think in terms of straightforward resources such as memory CPU etc +Your latest version will work much better by leaving the decision to the custom binding impl how to handle paths +Im curious Martin what is your reason for not using? +Without example code this behaviour can not be reproduced +can pick this +I can no longer reproduce this issue and I suspect this was caused by. +Ok Ill wait +Will try to reproduce the problem with IBM JDK tomorrow morning +Unable to recreate on current codebase marking as fixed. +Marking this issue as Closed as the attached patch works to allow ssh to an Openstack instanceQuestion for Marios + + +Below the full stack traceThanks in advance +Updated +After rebase more deps started to cause that so I need to go through the whole list again and eliminate themThis takes few hoursSeems we will need some automation for this +Something even as simple as a radio button directly underneath the theme selection drop down on the backend that says Allow user to change theme? Yes or No +I came upon the build when trying to see which Namespace test failedIf long lines were posted back to that would be confusingI am in the middle of reviewing +Robert The problem is that the script used to build changes came from the but the source TXT file from the current working trunk revision Committed revision ok I see +fixed and extended unit teststhis attachement replaces all prev attached files +So I think it does belong in just like get etcI will check up on adding this to added a test in and suppressed the unavoidable findbugs warning +Ugh +I think I can deal with itI use the to encode paramsit worksthe detail see Ford for all the patches youve sent in recently great contributions! are welcomeits my honor +No tests added its a trivial typo fix +Fixed a minor RST rendering issue on commit +Yes we can try to release a +Does it sound good to you +Since this is a change in thus affecting only new generated code this change doesnt break backwards compatibility. +Sorry wrong issue +Closing wont fix +Impact of Cache configuration on identity management of EPP when number of users is and roles are +Noble Ive got your point +done +Working with can exclude Part type from restore and avoid this error +Does Eclipse not use the same JDK you are using for building from the command line? I am confused on which part would be Eclipse related? Are you sure there is no cachedreused jar file somewhere? Have you tried created a new Eclipse project to avoid any stale datainformation +Can we have confirmation that fixes are in place for +to Dennis suggestion +bq +The last path with tests +anyone needs this code send me a note ndefreitas at iso dot com +I agree with what you said for benchmarks vs real use cases but for messages I dont see how it can get worse than since one cant generate messages at higher rate than there Maybe memory is leaking elsewhere one thing you can try is set number of messages and open requests to something really low it will be extremely slow but just to see whether it will crash +It would be better if it didnt have to do the table name to id lookup each time and that would be avoided if the app had its own cache based on name +Our Credentials and Identity which are scoped at SESSION actually do preserve state across restarts +ReportedI will revisit this JIRA later once is fixed +I committed the patch to trunk and +Thats great Han Ill have a lookI can be a mentor for this +I will commit this soon unless there are objections +Im the original poster +Martin DirichsCould you provide a test case that reproduce theThanks +If a user needs to read a file without verifying its checksum the user could use the raw file system handle instead +Thanks Jon +It sounds like this would meet your specific needs correctThe other possibility would be adding a general tab that appears on every job regardless of output connection which forces specific metadata attributes to specific values +So I dont see a problem here +Blocked access to via etchosts before and after the change to visually inspect +Heres my opinionFor and NLST should not show hidden files unless a is presentFor MLSD should ignore hidden files +Yes I can reproduce that using the JSF RI and the tomahawk candidate release. +Do we want to create a test wherein we create a and submit to a map red cluster with multiple directories +Latest patch adds option to deal with the EC issue +Would be glad to find what others think as well +Awesome + +Tried the latest snapshot and behavior is different +I updated the patch to address Alejandros comment and also added a comment clarifying why the merge callback occurs outside of the lock and after inProgress is cleared per a side discussion with Arun +Was able to do a merge on +initialization is added because it looks strange otherwise +Overnight test completed OK with that patch +defer all issues to +Just to clear up the confusionNO I havent implemented what Jacob got into JSF +Agreed will add the unit testsThanksMayank +This patch is made from configs directory +The path normalization in follows RFC so the collapse of is part of every Paths construction prior to its interaction with a +That said I had to do something similar in configurators and I did this by adding wrappers to the get +The +I have reverted the patch committed to +A combo box with a dynamic list of options for example would only need that model attribute set for the create and update forms +So Vm does not get started in the clusterNow in theory for advance zone VM should be able to move out of pod +Question What are the rounding off rules involved hereExample Splitter thickness width splitRatio +negative feedback so I suppose I can be closing this outregardsMartin +updated YOM and tests now pass +The first file is in SVN rev +Added new patch with UTC date format for the start and end dates build trigger type and Build Id URL there a better way to get trigger type? I note the does it in what looks like a more elegant way. +first scratchlets fix pagerank as well in this issue +I agree I used it in a previous project +Fixed as part of close resolved issues +I have added to the repository and this should be good enough for what it matters for us +What do you thinkCheersPaul Foxworthy +Thanks for implementing this Ive hooked it up to geronimo and it seems to be working great! +Retry +By default we can keep this prop to be null so that encryption is disabledregards aki +No one opposed so I relaxed URL parameter parsing as of dont find it very relaxing to accept a request that is invalid just because we think we might know what the user wanted +Verified. +The exists because of the following reasons +It sounds like we should coordinate work +Subversion allow us to coordinate smoothly on most things +The API checks for a runtime of a certain ID +So on thread we have the iPOJO bundle that is stopping so it deactivate all known bundlesOn the other thread we have fileInstall that decide to undeploy a bundleQuestion do you remove a bundle from the deploi directory of fileinstall during the shutdown sequenceOr is this FI that during its stop process perform an asynchronous uninstall +sorry Im wrong +Heres a patch with a few operations to start with +Tried locally and passed +What would a fix for this issue look like? Since this vulnerability seems to depend on the attacker having full control over the JVM process and how its booted I would assume that even if we make Derby prefer classes on to classes on the user classpath the attacker could just make sure the Derby version in the user classpath was one with the old behaviour +and configure did add the correct lssl flag to LIBS for the cpp Makefile for me on Ubuntu +Our build is bound to ubuntu because of some dependencies for running the tests +Namit I filed two followup tickets and +Retargeting to to match parent bug +But we still could benefit from a caching mechanism more localized to the callers to allow them to cache an immediate reference to a method. +I did misunderstand part of your explanation +lets get it in there + has more background to this issue +Please test and let me know if the updated pom attached solves your callout issue +Also gracefully handles duplicate primary key error +It looks good +Made a minor change in the service description to remain consistent with other services description +Temporarily assign to yourself if you backport and then reassign to Mamta before closing +A little unrelated question to the patch how will the packaging work around this this +Fixed addReplica logic to use get instead of getLog +Fixed in revision +Yes this issue is described in +Fixed +Yes it was correct for Roo version +Classes have been removed from trunk +Assuming that this is no longer a problem for you +Is there a way to just pass in the hadoop root? Can it only run this code if HADOOPHOME is not already set? Btw it was nice to meet you yesterday at Amazon +According to the BOM the version of Resteasy is not +Youre data is from an old version of JIRA So we first need to get that upto our current version +Given the feedback I think that the best course of action is to make this interface conform to the current rest interface +The latest diff seems to be missing something +This is the branding plugin that is responsible for adding a logo using the Footer extension +Changed from to +I want to understand how this issue relates to Ive tried your solution but Ive run into some issues +Merging to revision on branch +Also when you start working on this could you change the state of the issue by clicking on Start Progress +We need to be sure that were not tying embedded Solr to containerHTTP classes +You are right +Kris are you aware of this +In general we use which keeps track of if it has been already logged and contains an error codeMany places in the code catch unexpected exceptions and wrap them in aWhat would you have inherit from +after findbugs again attached patch after fixed findbugs +On lab +Created a new issue for sorting applications in wizard +may be i will do it for my Jspresso pluginIt works fine in debug mode +If the bean is in the extension array when the policy engine starts up and need the extensions it would ask the locator for all the and this would then find them +One more Improvement in Find Order screen as per suggested by Ashish Vijaywargiya now Previous and Next link show properly ie Previous link LHS and Next link Ashish +I seem to recall someone on the list attempting toimplement this a loooooooong time ago but never hearing from them again +We got his blessing to post it herePatch is against a recent +Ok i will create another +these requests should also be able to be tied to queues +I just committed this +I get fbadbbbddfdbcf on OS X using Codec +improve status reporting throughout all plugins +Needed for testing +Hello RuthI have removed the unnecessary check about the unresolved IP at revision Best regardsRichard +That shouldnt be a problem as AFAIUI most logging services provide bindings for both those +was created to show error message +With this patch matches the MRI year conversion logic +Patch committed to branch +temporarily removed to get M out the door. +I do not know how to edit this after submit this issue +Add a new constructor to Improve error type in Keep the original Throws Exception in parse code in Check HTTP return code in +Is used for me +I got the latest code and tried to build locally but I am getting failures in the +Inserted casts in forloop foreach ateach codegen to avoid problem +SergeCongratulations! I wish happiness and joy to you and your family +I know of that option after the little discussion on the maillinglist +I recommended this morning but I confirmed that is just fineso I recommend the latest version for Cassandra as well as. +As now we are throwing instead of +Thanks Ravi +Now that it seems to work do I still need to take a look at it +jmstopic handled in handled in +One unscientific test showed performance hit with the flag set on a PC +I dont see any better ways at the moment though +Fair enough +Trying this I produced a region with k flush files to compact +Also has guards against an update going in with same timestamp as the update its overwriting the case Ming outlines at tail of +Perhaps ditch the second Logger and just log through base logger? Id rather have these messages ON by default Hadoop should be easy to debug and troubleshoot out of the box +Ah good catch +Thanks Henry. +So for that its either that we need to use the configuration file for cassandra thats already there based on the static initialization of the vars in or we have to create a mechanism to configure itDoes that make more sense +I have written a test case and will push it to my github fork +If not Ill make one +Tests now pass using latest trunkNo BUS errors on Solaris +When actually appliying this patch I get back a instead of the +I guess well need to roll this back +Attaching which revives the target that adds to the compile classpathIve verified that it compiles cleanly both with and without pointing to various including the failing case with jlib defined and jlib not definedThe patch also adds some type parameters to to silence compiler warnings and it removes some reflection magic thats only needed to make work on Java and CDCRunning regression tests +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +The patch is already in trunk +This patch is for trunk +I also think we should provide this and had proposed it when we started to create the osgi core jar +Moving out of +The end user will still consider this a roo bug regardless +Moving this out as it affects only examples +bq +I changed the ubuntu version too +Is it ok to do it for +You should now be able to override the behaviour of as a result of so I dont think any additional changes are necessary. +Diff file for to reproduce I reproduced it. +There is a but is used only inCauses that this issue +We should do a cleanup and check that all Entrypoints are actually used +Ill dig +This has been resolved in the EAP release notes +Can you give us a complete that we can use to test this +Thanks Arpit +With your patch this situation shouldnt happenWhat do you think +Thanks! +to check on the health of a consumer you would need to look at both the max lag and min fetch rate across all fetchers +Set Fix version to +Kool unfortunaly I have no WLS enviroment hereonly WAS Greets set to fixed +Thanks to Jakob Korherr for provide this patch +This should be good now +bq +It seems that all that is needed for this is for publish to throw an exception +Im attaching the solution Ive provided to my projects +Now to reduce script to simplest form for you +This issue should be under GUI component +the patch fileI only know diff patch commands used make patch diff Naur patch patch p i patch p i have created a patch file for this issue which works for mesee the attachment for detailsand Sean sorry for the late response +Updated the test to sleep for a duration which is twice longer than the s timeout. +Ill leave this issue open until I can investigate +i change to rebase on master +Updated patch +fixed using nd approach you recommended +Sample project +Attaching screen shot which displays mapreduce counter values properly +Ive found in my system package that prvides also libxul packages +So I guess its a problem in the shim layer then +Sharad can you have a look at this? Thanks +This is resolved for CONTINUUM just not Confluence +No joining a terminated thread will not produce an error +it will be a local variable which is not visible to subprocesses +At least it does here +This has been running successfully for in my clusters +d For sync test wait for ms for each message to be received null is returned if it times out without receiving any messageI think a reliable way to receive the message is polling in a while loop until received message +Latest patch from review board +First attempt at a patch +Committed with revision +Patch against rev +Steps to reproduce like Hans changed the decorator name of editcontactmech from to in r which broke this view because the ordermgr does not have a +Oh I have conflicts! Which documents are theyTheyll have to write a view to find out and that view would also give them the number proposed hereDiscuss +comments addressed and ran tests +is also not a unique identifier for the contents of an array so if youre using it this way your code +No problem +This means that classes exxisting only as +Move the fix to +If someone does submit a patch that reduces the overhead of having many connections without complicating the programming model or tanking performance I assume it would be acceptable rightIf someone feels they can achieve these aims Id encourage them to sign up implement something +I can reproduce in +The exceptions were put into because some of them depend upon the protobuf classes in and any exceptions that can possibly make it into the client need to be accessible in +I didnt realize you could have two dependencies like that on the same code +Hi BryanI tried with another command prompt it passed in both occasions mentioned above +Thanks for verifying that Peter. +unless of course the standard also mentions the behavior seen in Derby +I tested this patch with and I see that correct values of start time and finish time are logged to history file with this patch +Fixed thanks +Tested and it is fixed +Not good +Attaching the first patch for this test conversion +Kevin can you take care of running tests and committing itIvan on a side note for your future patches I will suggest to avoid whitespace changes that makes it harder while reading patch to see where exactly important changes are +Finally got a chance to look into this +The title of the section is Prepare NFS Shares +The patch has been appliedPlease verify +Firstly is there any way from GWT to tell it to stop sending this info? If not an alternative could be to simulate a failed authentication that way the browser will believe it no longer has valid credentials to present +See the Velocity Tools sub project for more info. +Guys please remove all the attachments except things get messy +Snapshots are now being published +Consider that pipe has a different meaning that slash +libPIGHOMEbin +I have linked the issue here +releasepatching file servercan you please provide the steps to reproduce thisSince this is minor well probably fix this in the next release. +Thanks for pointing out the process +This is the final stage indeed +PigComplex fails and a number of Hive tests +Let me know if would be helpful and if so how I can share it here +Typically under Windows failures occur because the file is open but the file should not remain open +Version of framework used was taken from trunk after fix +Simply saying words dont make things true For running the clover target we set the perm size to m quick fix that bug! Oh wait thats a stupid thing to say +Updates complete +regards agree we can improve this behavior but however it was designed how it works now +Do you see any obstacles integrating Hibernate Validator into Apache CXF as a bean validation provider? The project is quite mature and stableHopefully the implementations are interchangeable so Apache could use later onThank youBest Regards Andriy Redko +why would the FD be null? if it shouldnt be lets use an assert thereDirectory FD could be null when its the first time we see directory +I just saw a demo of the METRO Config Service interop with the +Sorry forgot to mention the version +This adds the comments and more javadocs for the Maven plugin classes +in JSF then you need an own anyway +I checked that the bug is still present inCould you please put this issue onto your radarscreen. +Good catch thanks +Dont include a space between the option and the arg +It seems like you have been toggling back and forth between two different master files that both creates problems +Closing all resolved tickets from or older +Client is external to the cluster +Integrated to and trunkThanks for the patch PrakashThanks for the review Stack Jimmy and Chinna. +Simple patch one line +Thanks for the patch +Heres a patch that addresses this +I know its similar but in the reason for having multiple delegates was that there actually were multiple delegates the code now doesnt make much sense and should be fixed there too. +Does velocity help with this +Proposed patch looking forward to it thats pretty powerfulPatch looking good to me after a very quick glance only +Can I get one last sanity check before going ahead with it + now advises +Controls can be added through the factory bean for command groups +Just committed a fix for both issues you mentioned +bq +I have committed this +Ive had a look to see which code is trying to access a completed transaction but still without success +Happy to contribute +We can only binary deploy it +TYPE was because it was the original solution that I had found while googling +No further comments resolving. +Thanks Virag +There is likely to be plenty of information about feed status and process completion status much of which is well structured with clear schema +Go Jason go! +Is queuingservicexsreporting a JMS endpoint +Can the discussion be done through ML +The it patch seems to have some duplicate content +strange can you give a sample project causing thisI dont understand how you get these commas since they are points in google java api and you will see the web links contain a comma +Its time to start thinking about and I intend for this to be part of that release +Maybe create a bug in WTP side +yes Patrick reasonable tips a patch for this is attached that because I didnt make the patch based on the latest svn chunk version? Should I make a new patch based on it +sysctl agrep The same value as you suggestion +Hope this one is ok +Applied in r +I cannot run on any patch +Any reason not to commit it +Verified the issue on latest buildAble to add the shared nwHence closing the issue. +Essentially the evolution of is making our job harder than necessary there +svn ci m Applying my second patch from adding the Caverphone algorithm srcAdding srcjavaorgapachecommonscodeclanguageAdding srctestorgapachecommonscodeclanguageTransmitting file dataCommitted revision Closing as fixed because I dont think theres any reason to add the Caverphone algorithm. +What svn rev is deployed on the machine? Are you sure the code was builtIf you cant find a deployment issue then I see no way to fix this problem +Disregard my previous comment +Samindas API refactoring included the specified managers +ah right updated the title thanks for reporting of +For this to work youd have to define the Codehaus snapshot repository in your project POM +It is intended to be used in a group all script where all records passed to it will be added to the filter +Ben the current patch needs a storage directory to retrieve the edits +I think this is as close as were going to get to consensus so Ill close this ticket now. +If you get a minute can you let me know if the general approach is what you had in mind +I also searched all of the plugins no return either +original WSDL WSDL from Axis patchthanksdims +I would guess though that someone changed and forgot to apply those changes to the +Can you reproduce on? Weve applied lots of bug fixes since this is now resolved +This issue has been fixed and released as part of release +Additional investigation shows that the issue hapens when special chars are sent over String objectIn our case a certificate is generated and the certificate string is sent to the web serviceAxis fails to create valid XML and sent it over +hrmm +unfortunately logging is a bit misconfigured on these nodes +bulk close of resolved issues for release. +And then we just need to give users the url to use to play with time we will look to better integrations get it as part of the UI etc +It appears that depends on from drlvm +You could try that with a minimal direct Portlet implementation bypassing Spring MVC +of course +We will need modifications to build files to generate javadoc and source jars for each of the artifacts separatelybq +This will ensure that only master creates the base node and no other component can create itIs this change fine? I have tested it in my testing environment +bq +Remaining +Is that okay with youRegards Geurt +In minutes it is using m in mins it is using m +sure i chose bad wording +Please backport to SMX branch if possible +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian this issue. +Its using Netty which was released just today +I am fine with synchronization configuration +In the end deleting the lines and some surrounding ones and pasting them back was the only way to get them to have temporarily reopened this issue just in case there is anything that might be related to the usage in this case andor any caching you may be doing that would have contributed to this problem if this is simply an EclipseWTP bug please feel free to close it again +Since this will be a new feature it will not go out on the patch branch. +Closing issue +Theres now a Synchronize with Sonar server option in the context menu when in server mode which works as expected +So Im a little confused by the Patch generally looks good but its good to have clear logical naming +As far as I can tell the existing server does not indicate a dependency on the security moduleI will move the classes to the security module +Moving out of +This issue should be resolved by for should resolve this issue as well. +This patch removes all from Metastore directory +This does work to the best of my knowledge The relevant code is hidden in s parse method which gets called by parse +Committed revision to trunk. +Thanks a lot for the contribution liang +the dirty NSRE exception in logs generates queries on regular basis. +It is based on the trunk of this morningIt has as its base +The patch in fixes it though. +The same with the project case if volume belongs to project the template will be owned by project too +examine calls are jvm thread dumpwhile the first thread is sleeping the second one tries to obtain a lock in why +I agree it is important and Ill create a new issue with this subject as it applies to all the distributions +And activemqquery to show some stats based on a queryThough they all print very verboseWhat we need is the table overview that you get from the web consolePS You can then use the watch command and see live updates +This has been fixed in but Id like to apply it to the patch on the branch and it works OK +Fix unknown endianness when building Python on Solaris +Ive the rule application logic in a functionaloo style rather than following the PHP approachAll tests now pass +Coho copies the js jar and xml folder from the example folder that gets generated when you run +This patch has dependency on other patch which covers the suspend and resume process functionality +My does not show the pages anymore when using instead ofSo far as i am concerned a fix is not needed anymoreRegards think it is solved byThanks for updating us! +Hi Joachim it looks like the class changed a bit and I cant apply this patch as you please have a look at it again and resubmit against SVN Trunk +Will continue with review as soon as find some time +change theSo new prepost hook should implements the interface +Introduced domain and application class loader into and moved all get calls to get +Note that the code makes use of generics but it can be easily adjusted to work under Java +not sure whether it would affect selection when two deps are at the same level in the transitive hierarchy but it seems need to examine the commits related to this fix to see. +Here is a patch +Here you are +One in particular x Subject Area which appears to be where the face detection routine chooses for the AEAF target +patch submitted to to support this. +bq +any reason not to proceed with that +Marking the issue as resolved. +Would you like to submit a patch +Also when I allocated more memory to the JVM and switched to use parallel gc the issue seemed to happen lot less frequently +We had some issues in doing this +This is from our XML entity resolver which we use in xhtml validator +Otherwise I dont think we need to discuss the proposed patch which isnt the one we used. +Please explain. +Thanks Dhruba for the awesome work Thanks Ted and others for all the good reviews +giventhe fact that we make use of jukkas all over the place inoak i have the impression that we will add minor modifications hereor there +So can i commit this patch +Patch applied thanks +I need to think over the overall eviction threading issue +classes there is no good reason not to use the tcl. +documented as per pretty please user request +For removing STREAM stage Attached patch seems against and contains diff from different issue though +The last patch should include all unit tests as well +thanks +Hi ChrisWorked like a charm +v draft has a few minor typosstyle fixesIll try to come up with some better naming for the real eligible stuff +fixed in +Sorry had this in my queue for some time +I didnt hear anything after my last +Thanks +Ive changed the design for the next generation +Thanks for the patch! I was already in the middle of fixing this issue when you sent it in +But publishing via WTP is what I need when actually developing should be able to see it with the previously attached test projects +I solved my issue +As a reference in Win Harmony creates for every configuration at the beginning whetherever they would be used or not +I have made a request to hudson dev group +defer remaining issues to +Committed the document source to trunk as well. +It was worse indeedId have separated the commits by myself but did not quickly spot the part Thanks for your help +Working extended implementation of +Removing from the affectfix versions since it cannot be reliably reproduced there +Submitting updated patch +here it is slightly modified with a unit test +Based on this introduce a new global setting to control this +I have not tried to build the rpm on Mac port +Is there any timescale for this being made into a release version? +David event is called onselectionchange + +Thanks Bryan I think the latest version looks good +Otherwise we risk alienating users that cant figure out how to pass in files to the Woden Ant task. +Its not a nightly build its built using a continuous integration system so the majority of the time what you download is exactly the latest code +Probably srcplugin is a better place +Its affecting production so I went with what worked and was easy +Unfortunately Im getting some null pointer exceptions running unit tests with this patch +The steps Ive tried are +If you need this for your app patch the Struts code +Is it a seam testsuite? Shouldnt it be filled in JBSEAM then +Contains the mapping for the Castor mapping XSD fileTo execute the test case I ran the following under JSDKjava cp sorry for not mentioning this but the regression test suite does not finish without any problems at the moment +I do not agree with calling them legacy though +This is a legacy cost +Thanks I was preparing a use case anyway but I havent had the time to test it +ChunhuiNice work +Geronimo rev openejb scattered across random directories + +Done +Ill have a look at the hopefully until tomorrow and file the issue +thanksHow I wish there were an svn patch that would add files for meYonik +During Acquire is under going other requests of same topic will queued only in topic op queuer in until the topic is acquired +In JSF spec section it specifies that Java EE runtime should pass the instance to the JSF runtime by setting an attribute named in the servlet context + +It is confusing if the Flags says Required and the explanation says + +Works fine in rev +Yep it appears when I mouse over it + +moving to +We are currently using Hibernate ga with without this issue +Yes it is already possible but as you noted it is inefficient and can likely be optimized by an implementation +Update the fixes for this the patch for this JIRA. +Committed to the trunk and branch. +Its a totally unacceptable way to blow up +Simple patch which makes sure metastore does a login first before attempting to create dir +The commit you mention was reflected in that mirror within seconds of reaching svn. +Vineet is livereload integration installed +I find that this bug only happens when you have an empty table +Add more Overall totals +bq +Ill look into this issue +John has fixed this on the branch. +looks good +I am aware there are lots of other things to move +Were a little behind apologies +looking into this +Upmerged patch to and trunkSame patch applies to all +For the record I was able to get it to build with on Solaris +So you might have to educate me herePbq +except is really bad +Let me know if there is anything else hit us up on IRC if you need any help with anything as wellThanks +Corresponding mysql dump +However I did realize that there was perhaps some confusion regarding whether the resolver should be keeping track of in use vs resolved packages +sighAfter calling getrangeslices with a limited to N rows how do you make another call to get rows starting from N +I like Roshans thought of a queue +Moved to +Address javadoc warning issue +Fixed namespace to Removed widgets and comments +However thats not unique to this jira thats true for ant test in the hbase directory in general +When the reply message comes back the client invokes the callback asynchronously using the arguments passed on the reply message and also releases the waiting caller thread to that it returns from the forward call +Anyone has something to say about thisThanks in advance. +I recommend removing them from the trunk. +I would like to see this feature add too is it on the roadmap +Probably this should be really tackled in the itself and not just the +I have a patch available for this however I am not able to assign the issue to myself as I probably dont have permissions +Added test in r to confirm. +The IT test work with Maven through but I can find a few projects where I think I am recreating the issue +They all came back +You can change any text label using Administrationlanguage editorIll try to replace the text with shorter label can you propose one +Will be available in release +makes sense +Prepping ve been away for a while but I agree with Steve that is a better solution with less HQL plans createdIm glad to see this out in the release. +Another one! Well since your country produced Cowboy Bebop I guess I can get on board here as some kind of minor repayment. +Thanks Daniel +I seem to be missing the class +You probably want to update the docs Programmatic configuration while working on to discuss implication of wrong is a temporary fix for the CHECKNULLABILITY issue for merge events is a improvement for persist save and merge events. +Ciao Christiancan you please provide a test image as well as the style that youa re using in case you are using one +Issue is still present in +Forge cannot handle repositories with p layout +Thanks for the thoughtful review +On the other hand if the entry is needed every entry is included +Committed to trunk +I added you as contributor to hbase and assigned you the issue +I wonder why none of our current tests hit this bug +Nicolas AnneNicolas your patch is in trunk at r +Excluding findings is going to be the norm the vast majority of the time +I just committed this +If there is anything I can do to help please tell meAFAIK I dont have much karma yet but this need to be changed anyhow +If I force build the projects right after attaching the SECONDBUILDQUEUE to the default schedule again the two build queues are utilized + +Attaching the piece of code which configures SSL on the server and the client using CXF Java +Added support for testing both Token start or end offset Added javadoc comments for new just needs a copyright header otherwise patch looks good +Those are some pretty crazy screen snaps what does the right answer look like? ie what are you trying to draw? It looks to me like what was wanted was a series of polygon symbolizers on buffered content? but it is hard for me to see the intention +Move it back in if you think differently +Please find updated patch with explanation for +Backported to branch revision d like to this issueI think that full urls as tokens is not a good default for because i dont think users ever searchon full URLS +Lets decide how we want to fix this on the forums thread rather than confusing end userswith discussions about things that were never done +The prototype will be up on github and well move to a kitted solution in. +You can search for TODO JSF in the code to find where code has to get added +I totally agree that a more generic solution is required +checked into develop branch +Go through issue thoroughly youll notice it is related to and in there is a comment leading to a blog entry with a solutionworkaround and you can apply that workaround without Spring being changed issue is a duplicate of and is duplicate of +I know about and Ive looked at and others as well but they are not like this. +java files and awesome +I have run the tests with the patch +Can you point to an example of where you implemented this or a commit hash please? +Since you can reproduce this problem on your machine using a unit test would it be possible for you to apply the patch and rerun this unit test on ur machine? If it fails can you pl attach the output of the unit test? Really appreciate your help in this regard +Well the consumer is a portlet which allows uploads of files by clients +bq +See I also see fail in strange ways on a pretty regular basis +This is a simple example with a cluster that has map slots +However the zone is up and I cannot ping it from my networkAs for access you appear to already have an account on this box and you are already in the sudoers list for full root accessuseradd d exporthomepctony m c Tony Stevenson pctonyUX useradd ERROR pctony is already in use +Ive been running this privately for some time without problems +This is one of the pitfalls with action chaining +Yes but you have to do that with the current method as well +I just committed this to trunk and +Resolving this issue +Do you know if it happens with older versions as wellI also adjusted a few fields in the reportTo me it looks like the heap usage is a lot higher for Case B which you describe as steady memory usage +Sorry +I added the file into this patch I must have missed it when I moved it from needed to update and rebuild the sling apiall integration tests about the swallowing that should be changed to just throw exI can submit another patch for that if youd likeAs for wrapping the calls I can do that easily but I didnt want to have to keep a wrapper up to date +Ive attached a patch that fixes things in my limited test +Thanks Rob +trunkCommitted revision +Typically a user goes to earlier pages much more frequently +With a little less words that was my thought when writing The only project which could inherit is the executionProject all modules should extend +The fact that a workaround exists would to me not mean that theres no interest in fixing the bug I reported which still exists regardless of the workaround +The solution above should also enable the to work with Sharepoint and clams based Active Directory authentication as Sharepoint has claims based enabled by default +Not sure which build to kick though +But one need String formatting as well to properly format objects like dates. +do we close it +Another genserver timeout likely because the database file is being deleted making the system slowerI think its safe to add infinity timeouts to all couchdb genserver callsJan do you think you can test the attached patch +Attaching a patch for Betwixt for mvdb to review that fixes the gump the incubated Morph as the more mature bifurcation from returned to the fold +chris please do open a new jira +If we simply wait until containers have completed and stop the NM immediately afterwards then we can cause fetch failures leading to a relaunch of map tasks that ran on that node +Very similar to offpeak hours tracking should they be unified? See also Calling is after every KV seems excessive can it be improved? Esp +The failing test cases are annotated with applied under revision +already fixed in add a reference to the domain mode test that verifiesdocuments this functionality +Both of these dont seem relevant to the world Im still working on +now which is current pathtrunk +o +And even if you dont modify the modeler yourself wouldnt you still want to test changes made by Signavio +Im going to use this issue to be the master issue and then openlink other issues +As deprecates the option file in favor of files we can remove file option using this jira. +I mean maintain record of what objects are dirty and not what are +OK +Yes one of our processes have a wait activity but it works as expectedIs there any limitation for bpelwait activity for processesWe have only processes what are other restrictions for these processesThere is only one receive restriction described on the documentation site +The test fails without the patch and succeeds with it +This explanations sounds reasonable +clone is a better name than the API method from copy to clone in Rev +At this point weve achieved of what we need to do for JME support +Workaround available for Release +Yes but that isnt where the failing is happening +Making it configurable so everybody could tweak it you never know what HW or file system one usesYes we could think about adding static getterssetters for read and write +What others think +I can actually bump aside an ongoing RIT +dims +Thank you Chris. +Passes mvn happens if the vertex value is legal? Probably a bit cleaner to create implementation that does the right thing here +Worse yet is that only have a limited support for encryption in itskrb implementation +attaching a new patch that merges with +Plugins can also use the conventions of navigation plugin already and it will just work +Fixed in r +Note The cluster this happened on didnt have fixes +Nearly three years old now and no response to the comment since several months please as soon as someone sees the need again. +Id be glad to offer some pointers on Cactus in a dev list thread if you need themI am generally opposed to introducing changes to the branch that are not strictly necessary but the changes here seem to be localized so with a test Id be OK with them +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. + +Good to know it worked! +Eddie wrote Isnt this really just a duplicate of bug ? howevercomments on that issue indicate that it was not about the same thingI still think the ability to have optionally be instead of only would be nice and I dont seehow it would harm anything +Static code analysis could help us elimite potential vulnerable code in the future +I see you chose Dataresourcehow do you see linking this to the source of the data +Implementation used native sql query in class and did not take use of configurable database schema +Example test case for. +I need to move the assembly stuff out to a new module +Cool + applies against todays svn head +Also requires that dnaxmltext in repository XML be changed to dnaxmlText + +patch but w prefix stripped +Though this one looks different may be related +Patch files for codebase as of PM on. +Once I commit the changes I will be closing this issue as it doesnt seem to be an error in the but in Flume itself +Xmsm Xmxm XXmnthuge Xemserver propfile to show the versionsvn r +Reopening +We are not able to figure out the request that is causing this issue +Does anybody know where to checkout the realtime branch? I am very interested in it! Thanks! +James +Just make sure that changes down in HCM areWould be sweet if didnt have to do all testing spinning up a minicluster +This patch includes the test case originally supplied and a possible fix +Ondrej could you or someone else in QE work on updating it to the new testsuite? Separate the various JPA impls by package +Will be applied after code freeze +From a high level format returns boolean +While doing this would certainly prevent anyone from wanting to bikeshed the names in the future it doesnt seem any less fragile nor worth the complexity of migrating the system table schema +Actually the specs formal def of operators needs a bit of tuning too +SCR has been released +Looking forward to +For the record both versions of the patch are meant for inclusion under the Apache License +If I understand this change correctly it seems that you have replaced a full rebinding of all expressions with a more limited rebinding of just the result columns +The other examples that I believe need to be converted areblogdvdstoreglassfishhibernatehibernate icefacesissuesjpaportaltodoThe conversion will be made much simpler once is done so it may be best to postpone the rest of these examples until then +Something like thisIts a dir using first char +They do indicate a problem with the SQL statement +I just committed this +That being said I think theres more work here beyond the merge +Only reviewed but its obviously looking good +Thoughts +Lets see if it helps +Worth a shot to apply this and see if the problems go away +Updated patch +Fixed by change pathinfo parsing in per above comment. +You used to basically create metrics for a resource and then enabledisable them in two separate steps +Alena Anything remaining on this ticket can this be resolved +thanks for the heads up +Please test and verify that this works +I have prepared a version which should deal with different classloaders that can be loaded from +closing since fixes are in released code +Patrick Hunt commented on that atomic writes of the snapshot and log files is not necessary +The attached patch fixes this for both branch and trunk +Patch applied +Please ignoreclose this issue +upgraded to spring as is already released months. +can we open a new JIRA for thisSure let me close this and open a new ticket +Some articles have not been found to the RF cookbook are added to RF guides where its necessary. +rhq rev added a new operation to support calltime datajopr rev adds the plugin that the server is collecting metrics for the remote API and that those metrics are available on the monitoring tab in the UI +Thank you for the clarification +The first one would just be a disaster waiting to happen +By default the advisories got out in JSON in keeping with the supplied patch +Nicolas thanks for the script but I still get SCRIPT No such interface supported error when I try to order a column in a richdataTable +Changed title amended description +Attached source and test case +Do we really want to open up all paths under? Or perhaps just +Can I take this patch to trunk +is happening in Railo as wellwindows apache for the input and sorry for the delay +Anyway this issue is not the place for this discussion +You did it exactly as I would expect it +In I have added more cases to the network error filtering logic based on reviewing logs from tests of rapid NN failovers +it would be great to have a unit test reproducing this +Ran tests and everything passed +I merged this patch to to be picked up for +Is there a wiki page on how to use this? I need to implement an index with nested docs and an example scheme and query would be awesome +committed w HHOM on normal especially with multicore compaction one could argue that there is a good reason to want to adjust thread priorities for compaction this time to actually care about CPU usage balancing rather than IO. +After having discussed the issue above with Brian Chan from Liferay this behavior also appears on BEACheers dont think Liferay breaks the spec +to the end of the message +Attaching patch fixing the issue +I believe you +Were still seeing issues here reopening +Fix by from what I can tell +Does anyone know a better way than the following stepssvn rm sitesvn commitsvn mv targetsite sitesvn commitAnyway its done +Hi Stack +For this issue a fix has been applied in subversion +Closing since changes have appeared in Latest Alpha Manuals. +bulk defer of unresolved bugs to +Any hosting service offering hosting should be running their JEE app server in the server VM +yikes +I have been traveling a lot and havent had a chance to look at this again +patch is similar +Im not worried about access so GC and such isnt germane +We need that fixed anyway +md filesIm planning on committing this patch soon +I just have one question would it make sense to call put a second time for the same key and verify that the second call overwrite the existing value for the key +Whether this is due to the same issue as the ipad is undetermined +Im also going to blog about this effort +I will resolve the last of these this then Ive manually synchronised this user into Confluence +There may be more things we can do but they concern me because theyre very different from normal Java integration behavior +test for conversion to results as results as bar test to mention the best case benchmark was seconds on my test machine +M provides some form of resource filtering +Bulk assign all unscheduled issues with priority major or less to target +Relying on JDT weaving introduces other costs and we have to weigh up those costs against the cost of having to copy code +A Maven Archetype would be nice too +Thanks Harish! +Thanks for confirming +Im little bit confused with last commentDoes it mean everything is fine? And log files contain to much noise which confused usIf yes than customers would be confused too and it would affect GSS explaining several time that its expected behavior +OK then I suggest you update the site SVN files andBy the time that is done hopefully the site will be deployed and someone with apsite karma can update the site +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues of an aside to this specific issue but in relation to all of these massive number of issues Im not sure what I think about all this work if its not going to jive with + Have you set to true? Can you try this with +I dont know if there is a way to tell Eclipse that those modifications are not to be taken into account for another build cause they are generated by the build itself +But Im using a workaround as followsThese works well +On happy to do that as well +I would also add that we need to do interop testing with Axis and +Its The in toolsjava should be upgraded +If this is not reproducible in your side can you please see if the attached logs provides any insight to this issueIf not I can try to reproduce this issue again +However when using state saving there are problems when restoring the view after a container restart so these should be destroyed I think +It will be propagated out with the detailed cause of Exception +Committed to and trunk +Feb +This feature needs more discussion in the dev group with regard to consistency of the concept and API +Made Chris a contributor assigning to him +Ill add the patch over here very soon +This code was once all clean but its been slightly crufted through normal evolution and bug fixes +shutdownNow will force the operations to finish but this patch doesnt use that +One way to do this is to expose whatever Client sees +I could add an assertion for this +Vamsi the patch looks great +I just committed this. +We dont currently use SYSCOLUMNS to describe the parameters and return types of functions and procedures +It has worked in and is now broken again in +Satisfied. +Yes thank you investigation Andrey but the sample was fixed and extended by me in last days and it works now as expectedThe statescapitals was bad sample to sort. +This is patch for branch +Eugene do you mind attaching a nd patch with the compression patch +Use it to install hornetq to EAP and then unzip to JBOSSHOMEserver +Vinod what is status on this do I need to do something here +This will be addressed in. +Thanks +Ups sorry +We should investigate what we might get for free when Teiid and are deployed in AS and then perhaps add to this view whatever is lacking +The version appeared to be with the latest spec +We must keep in mind that if we go for setting these properties in JBT we should ensure we dont set them in some other for JBT and JBDS +you chose to implement your changes in a form that I will not commit I think I was pretty clear about that +Most probably log is just a misleading nameOleg +Committed patch as svn resolving +New patch changing over to returning an +Will file another bug for a separate issue that already existed that was uncovered by adding specs for this to master ffc fe +no worries +do a google search on jboss tools installation and jboss tools updatesite and try and make sure any hits on the first page is as simple as it gets +Lists are evilP. +The disad to getting rid of is it would make it much harder for someone to easily improve the look or functionality of the admin pages +Very occasionaly Acrobat Reader shows error with text similar to Can not remove aceArial font +Im generally comfortable with the purpose of the patch and with the approach youve taken toimplement it +I dont think i created many Writables in Mahoutso yes it looks like a moot pointok L +Returning always just for decrement purpose does not seem intutive +Is there a deadline for this version +Sort of like coprocessors +The patch looks fine +Please reopen when there is more information +It implements the three in the way +Thanks for your comments Bryan! Yes I agree with you proposal that the API could be adjusted hereI am currently working with an experiment to extend that API method with the desired tablecolumn number instead of the string +What about this Ram? Its about four lines +doesnt use its model at all the actual form value is handled by the recorder component +days and no bug m sure that Murphy will take its share on releasebut for now I plan to release as stable tomorrow +for patch. +OTOH Maybe this would cost too much as the session wouldneed to keep a reference on the collection +I dont know what you expect Roo to do or be changed to deal with bad data +Let me ask him what he thinks of thisWhats this change mean? private static final int MAXQUEUESIZEPERHANDLER private static final int MAXQUEUESIZEPERHANDLER has to be in HRS? Weve been doing work to break up the massive classesHow about a unit test +trunk +Steps to reproduce +really needed protection because it is a public static method that is part of the Java API +These updates such as typos or adding indexesare unrelated to the issue but needed to be corrected substantive change was made to the file review these files for this issue +We should either add or remove +I dont use ptpd so I dont feel strongly about including it in the recommendsChanging it to ntp works for me +ResolvedAlso committed in +BTW I also updated the karaf feature of camel and cxf now they just have the bundle dependency of +Many thanks for the patch + +For now I will create a separate project and fix the mentioned problems +Which revision fixed this +close off release +The patch doesnt appear to include any new or modified testsTheres no good way to regression test this since its a very rare race +but as Ashish points out we may not need to expose it +Any steps to replicate? It looks like its trying to start hte local instance actually with an incomplete launch configuration +See Jan th Thread Message History And Reply initiated by Colin items not being worked on afaik out of M Fix agree that the functionality would be nice however I think we would want a a clearer set of requirements to work from +Is just for Tika +This bug is now present in EAP on +This is unfortunatly a spec issue which I cannot resolve multiple workarounds are described as comments +looks like a better patch as well. +Perhaps its possible to implement a new Lucene API to do it more efficiently +bq + +We need to at least remove the old confusing doc and document the Maven plugin and potentially the command line tool as well +Ive started this work +I think it might be added as a step without changing any other code +Per Kihwals first comment we havent seen the error in but this is the same code so we suspect the problem must be there +or location be a http address +dececadebdc Allow SSL certificates to be selected +Well spotted! I observed another gain of about on my machine with about documents using your patch +committed +Attaching which describes the messages I will touch +Sagara big thanks for you to test and apply the patchI just created and will work on it. +shows all changes +You should never assemble a URI as a string +Sandesha works fine with Axis but not with Axis RC +What do you think about trying full width layout rather then the fixed pxThat was already mentioned this week on the will try to use relative units so that it will hopefully scale on your screen perhaps we have to restrict the for the really big screens +Unmarking High Value Fix as this query does not work with other databases and may not be SQL Standard +Very helpful +I propose to do it in phases +Lets avoid confusion +eh do you have gtk installed? what java are you running it with +Our general recommendation is to refactor your service beans such that no circular references emerge in the first is unfortunately a known limitation of setter injection in combination with AOP proxying You cant resolve the circular dependency with AOP proxies on both ends in this case +Hi RobertIve committed some changes for enabling keys on Mac +Still works fine for me on Safari Mac Chrome Mac Firefox like this works now on all browsers tested. +Yeswont fix. +Unfortunately some proxies unconditionally decode F to before forwarding the URL so we also have some special cased logic to things like design doc ids and what not +Yes Ill get on it +Refactors some code out that waits for RIT to be cleared out as a method +No additional info provided closing this bug +Since the Visual Source Safe plugin was showing the same conflict if you contact me at I can get you a copy of to try out +It a fixes the lint problems I saw on phabricator b Adds a new line to the files this patch adds +Javadocs out of sync in both the and +Attached patch that sets the content type to application for both and +Next time I will follow the steps +I asked around and it turns out that there is no simple proportion which does not depend on all smaller proportions +It is using News as the but the two in the page for the appropriate heading are NA and ProjectNews +Does fail to correctly process data still remaining in the input buffer? My concern is if we remove the check altogether will no longer be able to handle situations when a connection is terminated forcibly by closing the socket without proper SSL handshaking which is not that uncommonCould you please test a slightly different patch? Please let me know if that fixes the problem for you +Thank you so much +Grouped with or duplicate of +Fixed and I added a test case +If you want to change to noCFS for docvalues you need to subclass a codec and give it a name since those settings are final now +I also updated the documentation to reflect the changes +This is the example exported as an Archive i have installed areSubclipse Tools installed from update siteEclipse is have more info in the picture you a lot +Sent +In we use HSQLDB thats all +can send a full sample with poms content +I simply copied your configuration example for the filter chain and +As can be seen the differences seem due to nature of shuffle +The improvements to test cases is really aseparate issue. +In your docbook example the doesnt match that pattern as defined in yourWhen Ivy doesnt find an file it will use a default one defining single jar artifact for your moduleI hope this gives you a hint on how to solve your problem if not please ask any further question regarding this on the mailing list where youll probably get a quicker responseMaarten +A new patch addressed the review commentsWill put a few into followup including the stat update +Thanks for your patch +bulk close of resolved issues. +I pulled this into as well. +Patch has been applied +Like you said there is no other implementation that provides these features +If the contents are empty they are not saved +Specifying does remove the bug +Are you thinking about setting a particular status code in the response headerI dont know enough about these protocols to judge one way or other +Nick provide me with a draft of what youd like to see and Ill make sure itgets expanded and committed +I forgot to commit yesterday +If we would not have removed compression in the file format would have been identicala great example of why major release shouldnt be just removal of deprecations +This sounds like a issue rather than a Wildfly one especially given that it happens with both Web and Undertow +Simplified the heck out of the CLASSPATH construction by relying on HADOOPCOMMONHOME HADOOPHDFSHOME and HADOOPYARNHOME on the +Also there are some whitespace issues in javadoc and in test file +Provides minimalistic error messages +Patch for the trunk +I agree that we can probably drop the argument now but lets wait to hear from Konst +commited +Nevertheless has been tweaked to ignore inappropriate connectiondirectives in order to make it more robust when dealing with proxyserversThe patch attached will also make report a waring when an invalidconnection close directive is encountered in connection established response +I do not see a reason why methods should handle retries in one way and another one in a special way +Why not List? Set seemed natural to me +worked as advertised +Thanks. +it works +Sendmail seems to apply a more intelligent algorithm it read the mail and use every line as an header until it find an invalid line +Attaching which removes the break statement from the loop so that sets the types of all untyped parameters from context and adds a test case that verifies that the NPE no longer happensAll regression tests passed with the patch +I elected not to move at this point since its such a small interface +You shouldnt use setset without providing unique ordering yourself +We dont have Ant build scripts for any of the current samples nowyet so Im going to close this + +I guess youre right +Hallo therethis bug is not fixed in the M version +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +There is always a certain time lag involved depending on OS JRE TCPIP stack implementation Are you sure the port stays allocated for the entire JVM lifetimell look into it in the coming daysI committed a slightly different version of the patch +I also noticed the doc diff but Im not very comfortable to update the spec directly just based on the java doc without reading the spec update itself +released on Beta format +Fixed in r +Aries is using a synchronous bundle listener to react to bundle events and it is trying to grab an internal lock which is already held by Aries extender trying to register a service on a different threadThe framework is forced to hold some locks when calling out to user code due to synchronous listeners and especially in this case since this is on the STOPPING bundle event and still has work to do +thank for the quick fix +Gianny thanks for looking at this +Fails our CI builds also +See for get +ALR assigning to you +The answer may or may not pertain to why this is presently undocumented +Why is it problem if methods are public +Can someone please commit this +Not a bug because was specified incorrect url so its just render url and show border +All looks good resolving. +All aspects are simply a matter of setting up files before hand and copying them to the directory to be included into the DMG source dirsrcmainosxDSStore builddir +Uploading a patch for +meaning we can close this one or not? +Fixed +part of release. +However this needs to be solved sensibly to avoid duplicating the tedious blocks with instanceof checks already used in the repository +Also affects the can you make sure this get fixed as part or your profileplugins work in +Done +I stated that above +I added a note to issue to cover it. +See in the archive +Risk of incompatibility too so I can close it the way it should +Later in the week yesSo far I have been a helpless watching all this activity whileincapable of responding or helping to now +SorryDaniel Kulp only said that a workaround for this inside juddi might be possible but was unspecific about how to do itBtw thanks for your help so far with the issues I had previously with juddi and scout +Thanks for the update +Doing so would lose access to the process reference of the actual vm etc +I raisedBe careful with this patch it means that wont be able to deploy in KarafI propose to wait the fix on before applying the patch +Ramya did you use the patch I attached +Hongtu I think the behaviour is right +The issue with this one is that someone might want to define an extension that did allow this +You are attempting to get the attributes value and the bean is offline so there is no target to access the framework will try and return the offline value which needs to be persistentThis is separate from the first issue which does appear to be a problem with the lifecycle +ops struct has been dropped form many structs +I talked to one of our build gurus here +I can confirm the workaround but it doesnt work on a Mac +I dont care about making it +Log file showing a second cache view installation starting on a different thread before we have finished invalidating the keys for the first cache view +duplicated by. +Can this be moved to +This issue is now fixed +I have seen the property files so that I have checked the tmp and datatmp folders several times with tens of code execution but as shown in the error message and df h jobcache folder is generated at tmp so that it has the less storage error +Should I change the for this or should I create a new scheduler +This also affects Embedded +I think a simpler alternative is to override the deploy configuration for the to skip deployment +Please post your additional comments to the new issue. +The last patchs test case data may not have been sufficient to test out the changes +Please verify that this fully resolves your issue +Hi Matthiasi hope this is the right wayregardsMarkus +I think the import looks good to me +Nice commit message +Helps if I upload the right static test version that dislikes I didnt look at PMD although that is more down to my personal dislike of code lint tools that do source code analysisshrugs It was a fun distraction is marked and actively being discussed but has no assignee assigning to most active committer contributing patchesdiscussion so far to triage wether this canshould be pushed to or not +It is not a public APIOk +Attachment has been added with description This is the test file +Weve replaced with JSP wiki due to incompatibilities and extraneous features in that would inhibit usability in the Forge +I guess the downside of which would be that we might actually want to know what the units are in order to interpret the value properly instead of assuming pixels +command object binding logic is in Ive done this fix on Grails I have no idea if itll be applicable to previous releases +tests are passing for just committed this. +as well as the binary diffs on the +Ive always wondered why this support wasnt there +If you find further issues please create a new jira and link it to this one. +Import HTML project from Central and run it on AS +I think this is ready to be committed unless someone objectsCould you rebase eveything to trunk? I cant seem to apply to my local branch +using may be a good idea on system that have many disks or have disk name changing issuebut that will not prevent you from destroy the root disk we should confirm that the disk is not mounted before disk init +We should also open a ticket to add asserts in various places where weve shied away from error checking due to performance concernsNote that well need to fix an assertion failure currently triggered by before committing the ea addition +needs to be included in the usage string also +Dave please let me know if this is sufficient for your use case +For example a user may want to a snitch that always sends reads to predictable places +I applied this patch and updated the to use it +Fixed in nightly build +mysql on additional info here? Otherwise we will close this ticket within the next days +Do we want to remove as well? It contains asynchttpclient which I doubt we need in the world +Reassigning all XJava debugger issues +fix has been added to to add the revision numbertrunk r branch r +Hi SbastienI completely understand your objection to using mean and standardDeviation +This does not alter the directory and allows subsequent maven builds +Thanks Arvind! +Updated the patch to set in this value is parsed in the following order +I have added the subject of the dev discussion happened earlier wrt to this feature +We could create a new rename participant which will listen to corresponding events and ds resource if its path has been changedNot sure we should do it in Seam component +If it is the added code will go on to check whether the container is at KILLING +I am attaching a new version of the patch which contains the licenses within a data structure to facilitate their manipulationPlease review so I can commit it +will not compile on Attempt op implemented and tested Signed comparisons fixedtested on Linux Windows ia for isZero and toLong ops +Added mock JNDI objects to and corrected an editor problem in +patch looks good +And as far as I remember methods returning lists and arrays are generally expected to return empty listsarrays rather than null so that each implementation that returns null for array is potentially unsafe. +I am going to attach a java core I took while it was looping +Mostly looks good +Trace log for I meant is that the issue is probably the resolver is not configured properly +detail in svn is possible to receive duplicates when the messageOrdercursor reverts back to the start when there are no new messages +And potentially the correction could be done on the node update event before it makes to the scheduler impl thus transparent to the scheduler impl +If this is acceptable Ill do the producer and consumer in the same way +We should not expose these deployments through the consoles or block the users from meddling with them through the consoles +Hi Kevin Any update on this +Here is a patch for your perusal +Should fix also this issue +Couple small changesclose and in replace a bug where where stored in a tree instead ofIf there are no objections Id like to commit this shortly +Jozef can you have a look at thisThe issue title is misleading the exceptions are not suppressed but rather it appears that the methods are not run at all which makes me think it may be a problem with Weld interceptor bindingsIm not really sure where the issue might be +Now all of them use +We will work more on this issue as our knowledge of the core grows and allows us to dig deeper +dont have anymore code taken from. +Bulk close of old resolved issues. +This is JBOSS without any deployments +The skip counts should be accurate +And there is nothing new for M anyway so we can safely reject this +Ill think about this a little more and get back +Marking as resolved since has been fixed. +Thank you for clarifying +I cant get any HVM to actually start via jclouds +handlefinally +Please check +If exactly such IP is found we can use that ID +Ive committed tims javadoc improvements +Thats why Im removing the version for now +I added a condition target to check for istrue on the property the actual echo target depends on this being true +The build works as expected. +the find button is not submitting the page at allthis bug is related to the jQuery mergerthanksrohit +It may be instructive to first find out it the attached sample works on EAP. +please with rev +I have version. +Reviewed by Robbie +Finally added for Spring as a public method to be called on +In the future if lucenes developers make the reader more realtime so it sees more changes as they happen at the writer should be able to handle itQuote from the user using Insertion speed while we cant really explain this we are able to insert k records per second at a steady rate over time with RA while we can only do k at a descending rate with normal Solr + +Doesnt the filter just fix encodingWhere the issue is where you want to read off streambut Welds cid check already eats it +New patches add a second assert to the test and fix some formattingreadability issues +I need to be able to show the status of the ping handler on all my servers +Please look in history for what the actual resolution was. +A quick look at the patch looks OK to me + now handover completion to the Exchange so it works out of the box for FTPFile components etc +I think we need to fix this because it will affect performance +All the problems stated here have been addressed there +All Done. +patch attached +The attachment contains the fix for the issue +Two questions Fix for is set to and but commit is only in trunk Cant this be fixed in a better way than catching +Minor change to make spacing consistent and made a method package private rather than public +Ill be resolving all of them +Updated patch based on comments +I see this hang of time on my builds starting last week +However I am not trying to pass the buck +The properties that have historically been known to mess things up if not properly configured +We do not in fact need the encryption so feel free to turn that off +Mina versionCommitted revision + +But looking further the answer seems to be that the server dont need the trustStore part +And new collectors likeI havent see an issue regarding post faceting yet +Lars patch w some null constructors removed and some fixes in tests to deal better w Result that has no kvs +Fix by r in trunk +In general the patch looks ok other than that you threw in removing a bunch of files that as far as I can tell is unrelated +For now that is done without a proper pom that defines the correct dependencies but it is still better than nothing +Thanks a lot +Heres a patch for this +QA Closing. +Since we already have Perl bindings for this we could implement a configuration DSL +Wanted to get this in before the cutoff +Tomcat was designed to do this it has a service administration system for running as a service and all you have to do is install tomcat and copy the +is not super active and I have been pretty much the only person actively involved +problem is already solved +Ok so were done +The patch was merged into +Left some comments on RB +Closing issues resolved in released versions +changeonly reduceremove call +Fixed in revision this to seems so Im leaving this to. +Attached the throws always line from the of several methods +This is puzzling +So the remember me for two weeks on this computer would become keep remembering me for an additional two weeks every time I connect +Sorry missed that handy parameterset boolean +I think its a reasonable riskAs for the UUID being stored as bytes from the numbers +From Nick It appears that the image is missing in the markup +Good point +And add test on both memoryStore and +However I seriously doubt this is a container issue +Committed at subversion revision . +max Policy entry is applicable to both +Hopefully I didnt break anything while doing the merge +draft patchNext step is to use Davids sample code for verification. +dayjob and tons of travel kept me from dedicating much time to this +Right +this is related to fix events Lisa to make sure that we get two events when we have a comment and a state change +Latest version splits class affinity map into a separate class and holds separate affinity lists for source and event classes +Only if we cant estimate row count in CQL rows patch is attached which doesnt have estimated cql rows per token range which is not calculated easily +Theres no reproducible test case and no other reports of seeing this +I have looked at the patch but Im hoping Joe will also have another look and give his cents +Java script error was coming due partial access of internet to the +Hiya very much for this you are spot on with your analysis +Now hit the same thing with JMS way to diagnoze this issue is to wire up a debugger make an exception breakpoint on and check from call stack which class does it choke into +The simple answer is deploying the kitchensink quickstart on doesnt exhibit this behavior +Thanks for the IT +If you put it as jtstarget then even + Modified SM related configurations in to Updated to support extracting and configuring Changed is attachedPlease apply the patch +If not I can make the code reflective +What I really want is for them to disappear +see +Removing fix version since this issue hasnt gotten much attention and doesnt appear to be a priority for anyone +There could possibly be a limit to how long a module would be allowed to have that status after which it is either promoted or dumped back into unsupported +This also has had a lot of work done already for RC but I think its not seamless yet for embedders to launch a new instance portably across and +Resolving it +This patch cumulatively addresses the snappy issue Joeys NIT and JoeyDougs suggestion for a test case +is not always defined +This issue is a duplicate of bad +to deleting unnecessary code and classes +If you hit enter after that the process continues +Checked this one in as its been a couple days of the build being broken +nice so whats the benefit of the table using ajax? returning to the previous state would be great I think but am not sure what well be losing +The current impl works reasonably well it simply picks the first connection +Yes these are features that are supported in the container and only affect gadgets that need something these provide like pubsub open views. + +Fix applied +But my implementation still holds and optimizes the computation of divide and plus operation over sparse vectors +This will require upgrade to Sonar thereany news on thisThanksthosor +Some thoughts on the text transformationsThe upper case transformation looks good on main navigation and the application titleThe lower case transformation on the application navigation looks bad +I will have a try maybe still give it a free portand add more comments +Thanks! +David Is this patch ready for review? If so can you please create a review request on and change the status of this ticket to patch available? Thanks +Im guessing its a quirk of the file system to leave that file in place prior to a flush of some kind + +Attached an addendum to fix the jdk test issue. +You can see that when the nodes kid and others are expanded but the sign is not displayed in the tree ui +NoUltimately it would be cool to be able to expose an externally managed field as if it were and thus any code that uses could use it without changing its code +I noticed that I forgot to add some test related file +I am not really aware of any objective benchmark in this regard +Retaking issue as is on hold +in +Andreas it might be that my change was a bit too big +Changing the priority on this to Major since its actually a significant problem +But you said the problem still persists even if you removed the initialization in the static block and pushed it directly before the code which uses the converters +Thank you Guo your last patch is in rev +As it is now the attribute serves no function so its removal is backwards compatible +Patches would have to be made upstream as well as convince that benefitI would rather separate this issue into another ticket so the issue of renaming projects to reflect their TLP status could be tackled by looking for an easy ticket to start with. +The easiest solution would be reordering the fields in general +The job is scheduled to be executed immediately +Thanks Tom for the infoI am not overlaying the Common and HDFS directories +James Cheyenne said to direct this to you +Are there any disadvantages of enabling this option by default +bq +decimals sounds like a lot but I guess it is necessary to cover the full set of significant digits in a java Double right +Certainly no need to hold up for it +has a diferent docs system and I cannot find where the path needs tobe applied +resetting P open bugs to P pending further does not happen anymore with and trunk +Its simple patch so Im giving my here. +i think this is not a acegi bug but a ehcache problemis not part of the objectgraph that is put in cache +Add this file to deploy dir before startup to duplicate make Beta +Thanks +we keep archetypecreate forbackward compatibility + committed a while back. +An improved but still not perfect update is to jez about fixing it is still unanswered +We should always be outputting valid and usable stuff. +Anirban Id really like to have a discussion on where this code should go +do you mind if we hold this for post since its a feature? I dont want to risk any late breaking regressionsThis will give more time for a more complete solution as it seems youve indicated this is an interim idea of this patch is to prepare for a more complete solution with as limited as possible changes on the existing patch simply converts a static util class to a plexus componentIt also provides support for that will be required by ALL maven users that will use archiva as a maven understand you want to release a final ASAP but the curent code base is fully broken for maven users +The were fixed within issue +I dont think the original patch fixes the issue I still cannot get ints to round trip properly in the compact protocol +Jean pointed out that Cliff states in the mail that he will be making a formal announcement of the new requirement to committers and that the likelihood of a Sep +Thanks Dave. +Reverted in Fedora and Linux in general is a requirement +Note for some reason bit windows can launch JVM +Leaving this open till the full Hudson run completes. +It should be fixed nowPlease the issue if the problem is still there +I see youve put tomcat in the environment section of this bug +It does not test but is sufficient for the desired interface testing +Instead of using Eclipses CapabilitiesActivities framework I believe Max Brian and I have agreed to use the visibility and enablement criteria available through a menu extension definition found in the +WCS works this and WCS needs a fix +Further investigation showed that the changes for are the causeChecked out revision which works OKChecked out revision then which shows this errorSo jaikiran its your turn +We will revise all in our hand +Not required as the new lookup is very quick +Care to turn your code above into a patch for easy application +LGTM to trunk and +Hey Ken thanks for the important to let me understand it well the default head renderer does ignore facets +For example after running tests look at the class generated in buildtestsrcorgapacheavrotestI agree that this is +screen shot with additional check box for unique scale patch implementing it +to on GORA +I am sorry I just got to this commit and it is out of date with the changes +Juraj could we work on incorporating the Components Objects into Graphene +A fix is in CVS +Here is the patch for enhancements in screenThanks to Jacopo Rishi and Sumit +But we dont want to rely on another dictionary that along with dictionaries for Japanese tokenizer +same occurs if initModel fails when running subsystem teststhis patch does not address this it is not as problematic as original issue but is quite annoying for subsystem writers +I think it would be awash either wayI have a slight preference for using the new znode all over where possible +Marking as regressionThe results I see with the repro are not consistent with what the user reported +Thanks Bernard this help will be really appreciate +Logo combining left and right just added another version combining the left logo with the toolbox icon +should be and it shouldnt be in the revision history for this version of the document either FIXED +I still feel like this is the Wrong Solution given that logj datacenter properties and snitch properties all still load from classpathWhat about an option for bin to override CASSANDRACONF variable +Patch applied to branch only +Since globbing is this requires two round trips one to list files and one to list their blocks but that would still be a huge improvement over RPC +ESB and Smooks as the primary source of friction in this area? Ill chat with Tom F and maybe we can come up with something to help +Mass close all issues in resolved state that have not received further comments in the last month + +Pushing the latest patch through think in this case supporting the backwards computability is prohibitive +A lead dev on our team noticed another issue with inline gadets link tags inside inline gadget CDATA blocks are not injected into the HEAD of the document +Just a minor comment on formatting +Any objections +I was looking at doing this and I noticed that currently All delegates to which only runs the jdbc suite if vm +The user now gets the contents of s begin pageExecuted repro stepsAlso there is a bvt at +I still had some trouble getting Eclipse to recognize the XML so I doubt they are all removed +Martha what eap tag was taken for ZIP ERI havent checked all rhels but on Rhel tag and are signed +Because what is shown +Note must be installed first +should be extensible to become archiver for pig data format +MRI handles this by trying to read whatevers in the buffer and if that fails it sets the fptr to nonblocking and does another read presumably getting whatever datas available on the stream without blocking itIn our case we can certainly do this when its an interruptible channel but all other cases will have to just return nothing +I have downloaded the and the two tar files areinside opt on dev for Sparc and dev for xSome should install to work with CR +So force users to reset their passwords on login and then store in and validate from first before Salesforce. +Right +Is this still an issue do you think Tim +always fail on my test machine due to stack overflow +His patch I believe confirmed that the sstable was written corrupted +I will ask Alex to provide details about what comes before and after this status + has been committed so once Pig uses Hadoop this issue can be closed +Apply to +Hi MaricaSorry for the late reply +Committed to trunk +If one just wants to allow shorthand method calling surely one could just add wrapper methods with the new namesThe old names could be left for backwards compatibility. +Thank you for your contribution. +Couldnt an approved order also end up being cancelledmodifiedreturned? It would be nice if a requirement was flagged as automaticsystem generated so that we could remove them if inventory becomes available +DhrubaSureshGiven that we need to a new API to that sounds like new a JIRA to me +refresh patch against the trunk +TonyThat would be +I think I have a encoding of the XML was modified to on my Mac +Many of these queries pass for me +Done in trunk at revision I will backport in releases after in r r r +Big thx Mark! +The limits default to off so it shouldnt be an annoyance to ops because they would be the one that define itPlease help me understand the more desirable implementation +And that is a setup that does not exist as of yet +I will use it for some kind of port forwardingreverse proxy to open the socket on another host. +This bug is still open because it still does not integrate with Spring Securitys Authentication object. +update for vAdded forgotten file +Will commit tomorrow unless objection +Sorry for the confusion please close this thanks +This is now supported in integrated into the ESB codebase +but I will do it on since there is nothing broken on redistribution policies or statistics +Sure. +This seems to work for a while then stops being effective +Attached is the amended patch removing that edit +Defer to +Karel I suppose there is no need for object encapsulation of request headersDo you want to contribute some code around authorization to Ajocado utilitiesI can imagine to have encodeCredentials there +What exactly does it govern? It seems like it for buffering rows pulled from reducer iterator in memory +Thanks for bringing this to our attention +is a new merge request but we cant get the regioninfo of merging regions it seems user send a wrong requestBut i felt like while doing completemerging if due to some reason we dont get regioninfo and we retry why not the same thing happen here in CREATEMERGE too +In other words object oriented techniques of polymorphism and such can be wonderful but sometimes a direct approach is better and more flexible +All loggers are now copied in so this should all work as required +I plan to add a separate else block for line and provide another patch +Chetan Nadgire I just try ant in trunk as well as but all success did see the failure +Will have to move some arguments. +Matthiaswhat would be the next step here? Frank +Why dont we just incorporate it as part of the build first? Get it in there as is into the buildAs ideas come up we can track them separately in JIRA and people can volunteer to fix themA lot of people use this today and it is being left in the dustBill +Testing for valid versus invalid scenarios is hard there since potentially EJB session beans could live next to instances of the same classes in some scenarios +I think the NN can continuously try to factor in fixing failed while scheduling the next checkpoint +Its in the trunk. +a proper solution to this will look something like thisfeedback messages are stored in components metadata with the componentsadd new support for persistent feedback messages these are not cleaned up after renderform components report their errors as persistentform components clean their reported errors once validatedso form component workflow looks like thisbefore validate clear all reported errorsvalidate reports persistent messagesonvalid clear all reported errorsonerror do nothing messages stay until next validation or above solution is a little too much for because we want to get it out asap so scheduling for + for the to trunk +However meaning of external was specified in scope attribute and the missing attribute seems to be format attribute +which leads me to wonder why Im having an issue with the TLD location under that I didnt have in +Verified that this worksFixed in revision +Method! +Patched in version +Here is my first pass on top of Joels work +It shouldnt have even gotten that far without it on the classpath +Updated VPE preference for clarificationIgnore Visual Editor Warnings does not at all make me think about this it to Inform if in a project without settings needed for working Visual EditorAnd please put the full preference name of this setting so we can set this flag when running unittests +Well oun bash launcher does try to be a bit more clever than native launcher and recognizes some of the typical ruby options and knows when they allow optional white space +Argh +This is only donewhen the union iterator is not a part of another iterator and when the unioniterator contains one or more reverse iterators +I attached a patch for removing the code where I found the bug +As far as I know me does not support EAR and does not do packaging so I am not sure what you do and what exactly happens +closing with release +The current version of the plugin already checks that the version in is in sync with the version in the POMUsing the goal you can make sure that a valid date has been entered in the file +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +infomark +What are the users going to do when they cant act themselves due to the OOM state +fixed in current SVN +Hi ShobanPlease describe the problem a little more in detail on Oozie user mailing list user before opening a JIRA which is for identified problems +So far it looks like less moving parts and cleaner overall +I am sure you are not suggesting that we drop the use of tools like PB and Avro completely at all layers +Writing a section in the user guide describing how to build and deploy the jBPM web console will be tracked in a separate issue +Ive checked in a basic fix in rev The problem was that was not dealing with the case where it had insufficient spaceThis is a bit of a blunt solution but at least its simpleBill please test and see how it works for you. +In it wont +Im not a fan of the being a different font and spread out from the word Seam +The close can fail on statement reset or on physical statement close +Its interesting that were failing on maven plugins that arent available in Central that seem to be referenced from JDOM and JAXEN +I dont think we have better approach for now like more compact jar for the KDC. +This issue has been solved in. +If you dont think that approach will work lets work something else out +Its about a stable Writable for exceptions that doesnt rely on the far end to have every exception in the Java chain +Thanks Jie! +and +This patch modify ctor interfaces an instance of the generic type is added +I do not believe that this invalid end location is possible for method names +the document persists in the view on server even though server reports the document as deleted +Ive updated the CXF site exporter to process this and output the same script tags as it would for Confluence +If you want to review it please do so +Another approver can approve this and Ill check it in +Much appreciated! +I want to store the start date the end date and the final grade as extra attributes for each row in the studentcourse table +Were still leaking +I think this would be good and now is the time to make such a change but what do others think +The patch moves the classes into the exported package and leaves a note in the Usermanager pom for when its dependencies are upgradedI have not updated the exported version of the package +Someone else questioned my choice of RSS spec last night +Patch to correct issue +I think one extension property on FO is enough +Thanks for the review SidUpdated patch addresses Sids comments Moves to gridmix Removes unnecessary code changes from findbugs warnings are the same from earlier and are being addressed in +Thanks again for the idea colin +Somehow it seems that in may already have the fix in it +I checked all the use of this method in srctest and feel that the change will not cause any problem +The email blizzard is an interesting point I suppose the digests are helpful +using git to handle things like this but this needs to be done carefully so that the entire community is included in the process +Try Jenkins +Improved patch that handles namespaces and includes tests for those as well +Set to crawlspace you will start getting reasonable numbersWhat is the cluster size +Ive been given push rights for jffi so the changes are there now +Joel could you take a quick look +My worry is that doing the latter would overwhelm and hide configurations folks might want to actually tweek +Youre right of course +Documented the issue in the users guide. +Is that correctIf so when in this scenario was node last updated? There is a check in the code to prevent unnecessary remote loads that checks to see if the remote event was within seconds of the local update then ignore it +If can be implemented we dont use the concurrent skip list anymore +Global description of changesNew class extracted from classnew class extracted from class it uses a callback for credential verificationRefactored to use callbacks instead of subclasses renamed it to because it is no longer an abstract classRemoved all private subclasses of from this functionality is now included in anonymous inner classes created on the fly as callbacks for theCreated new package +Since youre suggesting to raise the log level to ERROR So it could be sufficient for you to simply see this as an error log entry? What else would you do if you could implement a custom exception handler? Do you have code that needs to be informed of the beginning and end of a recovery phase or needs to check for a recovery phase? we have support for a standard JMS through the exceptionListener bean property and are invoking it for listener exceptions including listener setup failures already + +Trejkaz thanks for working with usYour patch has been applied many thanks! +Seems to be fix somewhere between and. +Update the patch to address Nicholass comments and fix the test failure +Thanks folks I installed and yes it works here tooIm impressed by the quick response and bugfix. +However it might have been an indirect fix +This patch doesnt break any of those and as it fixes Juns build after I fix the findbugs +DenisI created a simple unit test Im not sure if it is testing exactly what you want and Ill attach it +Our project will continue to get build failed notices to commits until this is resolved +Im unclear why you need to do this in the first place + to trunk! Thank you Navis! +The proposed fix works perfectly and Im building some pretty complex queries +But good work so far +attach my patch +Modularize the Kryo object creation +Could you please upload a summary of the tests failing and the logs as well +Im then as long as we if a test util ever starts relying on a hadoop compat +thread pools are now definied sepearately and timeout works +They seem to be sufficiently separated +It really is quite a bit of work to downgrade this to lots of generics but it also depends on enumSo if you dont want in contribanalyzers I vote for simply removing it from trunk now and reintroducing it in the trunk +New patch with both fix and testcase included based on branch revision on and trunk +At some point Forrest will have to move to only supporting +The hint rework looks good +Are spaces allowed or not +Fixed in both branches. +Your patch was applied in revision The servlet API require Java to compile +bq +Rather than anyone have to wait for me to get the patch file created I thought it best to get the source uploaded so people can start using itThanks Peter +If you dont agree please reply otherwise I will close the issueRegards tool +I agree we should do the first part +We could but unless its a heap more work than we thought I remember liking the ideas we sketched in Philly about field set meta data +Committed to and trunk Thanks +Shouldnt we add some tests of the generated Python code? We might just add tests in the python tree that refer to +Ill commit this soon unless someone objects +updated example file please +CameronI sent you your wiki password privatelyIf you havent got itor got it and stillcannot log in please follow up here +I would like some one else to review it since I do not have a lot of windows scripting experience +Attaching +Any chance one of you can verify +The integration test passed by the way +But have discover another problem as it in +Now its closer to of the time instead of the original +He didnt like the non scalable queries +We dont want code in core thats not both widely usable and actually usedSince is not a replacement for and does not support reading up to the latest sync while writing we probably cannot use it for +Sorry for picking up the patch late +Besides we can borrow generously from some of the projects that did some of it already +So changing the behaviour makes sense in this case +This will make the NN exit safemode Dhrubas feedback on the review board +I am working on it atm so should have some completion by the end of the week +Right before the branch +The object is created by teh itself and set as a request +I have thr different computers that I pull code from the SVNI do not change code in this projectsI move the code to new projects to work onIn the svn trunk projectall of them match as of all of them have the same errorso why can you repoduce this? +ManfredI believe there is an incompaibility in the behvior between and which the attached sample from the Suns blueprint catalog illustrates +Will then loop back and refactor as necessary +Can someone suggest a way to test this +Because I made a full install with Maven and Java in Ubuntu without a problem +It divide windows into multiple work region +History of parallel added more details to the audit log including the parallel GW Parallel Gatewy Start Event End Event Intermediate Event. +The file should now stay in sync without a still need to deal with the unexpected loss of an agent that is still configured as enabled +Changed references to to in the connection code +We can unify and refactor down the road +Moved Pi Songs additional improvement requests to +As you can see that exception is inside of not Curator +Still do not have a chance to work on it +Imagine Threads context classloader loads a different version of Hibernate Validator +This is somwhat strange can you give a hint which goals do you run +Ive updated the validators home page as well as the Bean Validation TCK wiki pageThe reference documentations need clarification still +Refactored dispatching in Rev The list of services to dispatch the event to is created synchronously with the actual delete call Event delivery is scheduled synchronouslyThus are scheduled immediately before the configuration updatedelete tasks +One with a and here everything works fine +This issue is a duplicate of which has an updated patch available that merges to current trunk +I agree the extra column that may beadded by ORDER BY is very similar to the GROUP BY case +Old and from only one client +here you can see that the initial vertical position of the is to high + +Sample certificate +Ah sorry about that +Its a good practice to preview your patches and make sure that each change in the patch is required by the issue the patch purports to fix +On win I get about ms improvement in the configurationOn I dont see any effect or it is minimal +trunk +or +If it is a pain for you to run the tests just let me know and i will be happy to run them +we sync from svnrootspringframeworkreposrepothe poms for have svn to native so they are converted to unix end of line while the checksums are done against the DOS fileseiher the property is removed or the binary property is set or something like that needs to be done in their side for the next have converted these files to DOS format so the checksums should be fine nowunixdos converting file +this means a duplication of code for him possibly in multiple languages which is no funTrue but thats the case for a api as well as a binary one +For further review in the development period +I this the currently committed code fix this stuff in the expected wayAll tests still passIf you have external code depending on the previous behaviour please provide a testcase so that we can discuss the need of body to be unfolded in the parser and not by +I meant to say clusterFilter +Damn nasty problem +Yes I just finished the just attach the is that someone is in charge of my patch +After running the test onceapparently theres some data stoted with HSQL +If you find further issues please create a new jira and link it to this one. +Are you usingHow much data do you have on disk? How many sstables +When jobs finish all running tasks should be killed via +Refactored to inner class to ensure order of dependency of instantiation of components is preserved +Its definitely complicated so opening a new jira to track this. +direct Blueprint XML deployerits really stable +can not reproduce +in Public if no marking of what follows +Once your grows the usual method would be to interactively rebase your changes into a single commit +Is this going to support transactions +Implementing seems like the best place to start +removed the action event annotation on the method causing the VM bring up works fine without any issues in the latest build +We definitely have to keep this issue in our minds +I am running a full build in both profiles now +I changed move to create the parent directory before moving a file +Is this still relevant? Invalidation is in place and works fine without such a notification +Updated patch that passes ant precommit because it uses +Running tests for branch patch +you also shouldnt need the filesystem tool you can get this from in code to address all issues identified. +Please see Thanks +I think adding a message attribute sounds like a good in r + server is not actively maintained. +Also added in some documentation for it. +This looks like a good change and it applies cleanly for me +Verified. +After discussing with Venkat I understood that the datatransformation interceptor will not be inserted in the invocation chain if the target interface is not marked as remotable +Issue fixedReplaced linear successor check with eventual successor check that is recursive +Can you check thatAlso do you have the env variable AXISCPPDEPLOY set? If so then do you have an file in that directory? And what does it contain +Please attach your stylesheet and xml to the bug report +im going to need you to come in on sunday because its saturday and I dont know the name yetyea +I do a lot of stuff with terms and those terms will get pushed out of this LRU cache very quickly +mvn dependency is not producing the classpath in the same order as mvn dependency with has been to cannot reproduce. +Committed +Since then Ive implemented a different approach to solving my problem which does not require a Stateless Session +When this happend in Sakai I noticed that changing the from to in both Sling and Sakai caused the same problem which is bizarre since that appears to be completely unrelated. + +Test finishes correctly with +Hello HansonNow that I took a closer look at I can see that CODECMULTIPLEOF is about input length +The wikisearch code to me is more of a prototype application than a part of Accumulo +In general for many components that should be used carefully as for example all popup ones will be influenced by this also +Ill reproduce it myself tonight and hopefully give you some example code you can use when debugging it +Alexei could you please also check and exceptions seems to me the same as the one for this bug +will introduce some refactoring that will affect this patch +It should be able to share most of this logic with the I think +get that is accessible to all classes in the same package and the applies overrides during a restart by calling the method appl +Also modify the jUDDI configuration in +Bit set copying is ok only if the mappings between languages and bit positions are the same which they are not in general +But the class still does not call close on the in +Javadoc options are I cant include some classes and exclude others from the same packageThat saidDavid is having the classes in Javadoc a problem? any chance we could get rid of split packages +jar to libendorsed +s serialization routine is straightforward a C indicating bytesize followed by bytes with a sanity check to ensure valid +I separated the ownership part from to this patch here to make sure we have an agreement on the changes for the topic ownership +Never mind +Dont bother with useSasl and just use useWrap +Great! So it means that will be fixed soon! I will be glad to see it happen +Patch for server mode +If you dont want to resolve it for me give me the workaround at least +More likely the ant build will be removed. +But the stack trace should help further improve code and I can try to update my Eclipse android to perhaps reproduce problem +Ok I will close the issue +Thanks Sergey! +Initial support committed to trunkbut Testcase in missing +It has the eclipse templates on the same lines as hadoop +Hi HansIve looked a bit more and the error I had was because in the demo data the clients have no currency defined +Hi Maximillian can you try again with trunk r or newer? We rewrote the replicator and improved its memory utilization +Data directory of the node that wont rejoin zookeeper ensemble +I agree this should be the final fix in this area +This is not in the patch but I thought Id bring it up since were getting our hands dirty +yes big +Output from running mvn X Dpackagingbundle install at directory +We can defer it if you like +The easiest way how to see whole thread and reply window is click right on Reply button and open it in new window +Please test against CVS +I didnt try to build it but having a look at the goal and implementation make me think that this change is definitely worth a things where they belong is always a good thing +Behavious is now as stated by the spec. +Unscheduling from be already better solved by which would allow us to close this issue +Updated test patch +I have also tried to be optimal since this is a high touch segment of codeI will add unit tests for shortly but would like to see whether this code satisfies the original intent in the issue +Just trying to work out a sensible way to do it +Committed +Also take a look at what can be closed here please +This isnt something like a Tomcat server +Is there any chance we could get a high level design deckdoc that illustrates these processes currently and what looks like after we move to this proposed mechanism? Also what operations would eventually get ported to this mechanism? I think discussion and an example at the designrpc comms level would help a whole lot by grounding this conversion in reality and not require diving into the code +Sorry we cant rename it for compatibility reasons +As i replied in the maillist i think VM snapshot is generally and volume snapshot which is for backup purpose is more likely to be scheduled +Thx! +Then you can add yours back and do an svn diff to get the next patch +Ive initiated the transfer with dotster +This time wo prefix +Its gonna be tricky to reproduce this in a sensible test case +Thanks Erik! +JIRA project admin must assign my id to the project +Please also be sure to make this change +Closing stale resolved issues +by the way if and are available then and should also be available +Committed this. +Fixes to hide the title bar if there is no title and no icon +Committed to trunk +I tried the sulotion you suggested by creating definitions it worksThanks good to know that it works for you with the individual filtersClosing this issue then +cvsignore files because usually these files should not be included in packages +My patch against version which circumvent thread local variables from not being garbage collected on application redeploy +Dougs patch for fixes that by implicitly making all methods think a warning should be issued in that case +Also occurs when output limiting is set as output last every sec which is our preferable timing changed to msec for testing purposes +The pending request is to have it also applied to +Fixed in SVN +Apurv I have not looked deeply into code yet I shall answer after I know the differences +Please try this patch +When a exception occurs during view rendering like a or we dont catch the exception in and clear flash +bq +That mitigates the issue if the permission check is the first check applied to an inode resolved path +There are still references to tomcat but always like You can even run this on tomcat so it is clear that only some things can run thereI think it is correct closing. +The last comment was tested using Java It is my understanding that in Java synchronization has become even cheaper although object creation is cheaper as well although of the instantiation time of is in the init code not the object creationI know there has been a lot of discussion of the problems with +Is this still an issue in? +Syncing patch with trunk +Patch committed to trunk +as I had mentioned previously I am able to view the content when connected through SVN but there again I wasnt able to perform any publishing activities +This PDF renders just fine in the trunk version about to be released as +For now I think well just have to not support WCCP for until someone really cares about it +Closed after release. +Patch checked into trunk branch and branch +What can happen here is that the first message sent is dispatched and processed before the second message even makes it onto the Queue and so it appears to be out of order but in fact its not +I will expand it eventually +whatever copy you end up putting in the formal docs I can post there too +Andrea +I havent had the chance to test this thoroughly but this fixes the case and passes all tests +true skips all compile test run test create +I added the ASM jars to jtsktrunktools in revision +Heres a new patch that removes it and tidies up instead +bq +All the test failures were caused by Too many open files +bq +Attached patch is an example to demostrate Resource annotation for simple environment entries +I would think that this all or nothing granularity could be a problem for some applications +I made some changes with revision on Sep for this test but I think new fixtures have been added that require a similar treatment +And maybe someday when jQuery would slow down we would include your plugin in Struts trunk +I know for a fact theres code out there that depends on getting an array tuple for hasheach elements +Changes done as per review by Gary and Stack +I get this exception even with allow effects fixes the NPE in question +Yes should be fixed but the bug description says that there is the same problem for class +The patch implements both of these enhancements +Heres the fix +I just made you a contributor can you try again +Approved by Andre +Initial version of document was described +It mixes the best ideas in this thread and uses nothing but FTL +j test has been passing +Investigation is still on whether this patch actually solves the problem or not +That doesnt mean it is a bad idea only that it is visible to all FS clients were going to have to make sure it is consistent across all filesystems write tests for it etc +Thanks for look at it +bulk close of resolved issues for release. +Strike that reverse it +Done! Thanks. +Thank goodness for compilers +I would prefer to keep the actual rereplication logic in the client package as it does reach into internals which we would have to make public otherwise +Have a peek to see if you can make it break or if there is something missing. +Adam points out that the key would need to include the concept of the application +incorporated amareshwaris comment +A patch file is attached +The TTL iterator should be updated with the property value when it changes or at +ocell de corral o au de corral +Fixed in +Im pretty sure thats whats happening but its not clear why +Oh OK could be then that there are only architectural m intending to use the one I built in the larger contextof the RPM Creation and Quality Control Automation effort I have proposed on the mailing list but Im not saying that anyone else has to use could be that both produce the same end result and then everyone has a of the reasons Im sticking with the EMF model generation is that it generates the baseline for the eclipse plugin that goes along with pomspec mapping so thatwe have a leg up on the corresponding eclipse work as Ill publish all this soon so that we have a clearer context for discussing various approaches and corresponding user and developer use case Ole +Keith wall has posted a probable cause on +In I am able to add a second build definition with Typeant to a project that was added as a maven project and has the original default maven build definition. +I do grant ASF license to patches +Id had a similar problem with AMQ reported in which I had fixed +Yes there should be another JIRA ticket reminding us that references to should be stricken and replaced with using the constant elements +This may be a good thing +I added a testcase for my particular case and fixed the grouping if a term gets to trunk in rbackported to in close after release of +Check your connection pool configuration +IT added +Enis Soztutar do you have any bandwidth? Since youve shown interested earlier I am up for it +This patch shouldnt be committed until consensus is reached there +Patch committed +trunk +However mapbuilder cannot access the data on either the postgis or oracle database directly +attached patch filters out the mbean from but the rest w configured are +Seems a bit odd to remove just the return and not cleanup some of the other odd bits in the javadoc but its fine wmeThanks Thomas! +Binglin are you planning to contribute the c library based on this work to Apache Hadoop +Bunches of the MR tests seem to be failing like this too +When I switched it to this error went awayHTH. +name +bulk defer of unresolved bugs to +Nasty little bug that does indeed appear to go back to +Enabling zone will create sys vms on the host and it will be marked as Shared +I havent really thought of a great use for min +Hi AngeloWhy did you created it how and when are you using it? m using this attribute because in my application all the screens are linked to a field in the context +This is caused by the ejb proxy optimization issue described in the linked to issue +LICENSE file updated in so removing Graduation version from AffectsFix version +Ah nice +Installation generates in target folder +why dont we just use the result from the first mvn call +added handling of SSH keyboard interactive authentication method prevent SSH session from looping while sending bad passwordsNote the patch still contains the binary icons which seems to prevent patch from being able to apply it +The behaviour is caused by thenature of Basic authentication preemptive authentication can be done if andonly if credentials for the null realm exist +Improved throughput in comparison to Hadoop RPC for data transfer +I remember having bizarre issues with the behaviour of this method in IBMThe actual implementation of is also so there is always possibility to run into bugs with current implementation also +I added it back so this test should pass now +I swear I reported and fixed this problem previously but here goes another patchCommitters can you please review +I also think this a Roo related issue as the Entity code is generated by Roo +Patch looks good to me +the project is finished by now maybe ill try again for the next release of the project since were not really happy with what we have today +Actually what is being stored in the session there is the implicit model Map which happens to contain stale session attributes as well +In all cases roo was running in a dont know whats different this time but basically I cant reproduce the problem any longer +Sorry I couldnt get back on this we couldnt spend much more time on this but in the end our analysis pointed in the same direction you suspected so I agree with your gut feelingI am not using atm but suspect the situation has improved since CXF is now the official WS stack in? +The trick to bringing this bug out is to make the window as small as possible but not so small that scrollbars show up +also the varloglibvirt can be useful since it will sometimes spit out why the storage pool failed to create +Closing issues for +Aaron are you happy with this patch? Should I commit +It changes to check for a servers root url to verify that the server is running +has been marked as a duplicate of this bug +I dont thinks I need a test for fixing this typo +Please test it and close this issue if it fixes your problem or reopen it if not respectively. +those for +This cannot could be circumvented easily +Thanks Sameer +That requires reliable locking of course +Even starting as root resulted in the same failure +I think it would also be useful to make the test report that the failed +A couple minor issues I noticed Under New Features it says This release addresss a data corruption issue +Needed to move the retry of the checkpoint out of the synchronized block +Theres a workaround if you can use spring integration +If not this can be deleted +First up an element for the be revisited forAny reason this was dropped fromThanks +I also want to talk this over with James but it wont make much progress until M +Cool beans +Verified with EAP reopened to change Release Notes. +List created +So I would prefer it as it is +Sounds good +patch available in branch good! into the dev branch commit afacdefedeededdThanks Aimee for the feedback! +x +The HSQLDB picks up the extranous space and refuses to register such relationships +Passing in system property requires either setting ANTOPTS environment variable or using XXXXX which either requires changing environment variable or typing extra command to change from one groovy home to another +Attached a patch that would redirect user to a confirmation page when deleting a purge configuration +Marek would probably be the best person for that +stack NP thanks for taking care of it on your end so quickly +I dont have time to investigate it now but its a real misteryThanks its a possibility although for this one I was hoping we could get away with having a DB on disk its so much more useful for demos +We together can make Spring the best Programming Platform. +We may split it to be scheduler specific but do we really see either scheduler not using it in the foreseeable future? Perhaps we are causing more grief than good by splitting them +I guess this can be closed since the whole thing was caused by using the documented but disfunctional command line options +Interesting problem and good to know for the future +Why do we need to defer the notification? Since theyre immediately moved out of the current directory into the to dir they are as good as deleted even if we crash and restart +thats the solution +I need two things if this is going to make it into the docs Is this content agreed upon as appropriatecorrect? Where should it be included in the docs +Im fine with turning on by default though it doesnt seem as critical to me as to you +I think this was resolved against please retest against that version +Created for the problems in +should be fixed in as it has cleaned up url generation +Closing issue since this is already covered in the implementation of. +Resolved +I cant reproduce this problem in both GEP and GEP +Is it possible to update the Xerces JAR to include the release +looking into this +RM and RM have set to rm rm respectively and logical name set to for both +Both UI and API can do a check in register template to enforce to true in case of SSwift as secondary storage provider +Please verify if possible +no storage is the right place I just meant dont put it into Pig builtins +Could you please svn up and then regenerate your patch +Ooops left a VIM temporary file in that last archive +If we are collecting this data to be able to output a warning it would be good to also keep metrics for each disk +Fixed by adding methods to and setting them as part of field annotation processing. +Also please turn the example you provide above into a unit testThank you a new patch with tests +Forgot historian class and +I agree with that has been released this is one of our top priority issues +just playing around with the idea of using map projections in the logo +thanks +bq +Thanks +Ill commit shortly +Config file for wireframe compile file for sparkskins compile file for sparkdmv compile files for Mobiletheme compile files for mobilecomponents compile files for halo compile files for charts experimental and compile files for automation projects compile CarolIve finished converting the compile targets for experimental charts and all automation projectsPlease take a look at the patches when you have the chance +Uses instead of +thanks a lot +Please check that it was applied as expected +Hmmm +after the case change +What I applied to and to +f on and rce on master branchesIf you have a chance would you try latest one + committed a few changes but they havent shown up at after minutes +I dont see why that cant happenJim +The concrete change I would make is to move the serializations to contrib modules so that they can be more easily updated than core +Please this issue. +Thanks Andrea +will no longer override string values with of +This will go into when that is released. +Update patch with cleanupInclude changes of R to be declared asprivate stable +I tried the nightly build with my and it runs wellSo i marked the bug as CLOSED. +Will the fix be added to the next release +Added patch for trunk +For this case it makes absolutely no sense to add a description field to the various instead of to the main entity neither from a pragmatic point of view for the design nor the actual details of what is intended to be implemented +Why do you need them +ok rolled back the cleanup change +works well with big indexes now +It download the JBDS install meclipse follow the steps as generate web service the web service wizard will choose a dynamic web project from workspace while if no dynamic web project it will create a new generate web service client it does not work because the wtp web service framework only support dynamic web project +Now that is no longer used in jdk I will back out the patch for +From what I can see the change to is unnecessary and without a further change to the code would not be that seems necessary is the first part enabling within +I am able to run the CTF test suite fine once +Downgrading to minor priority and undefined fix version until there is actual demand for it +Please use policy based approach +I agree +Bob Pecors suggestion helped switch the java home in maven preferences from the JDK home to the JRE home +It is also the only case where the synchronization takes effect at the moment +Applied the patch in revision + by SVN revision and +If you find further issues please create a new jira and link it to this one. +I think the concensus at this point is that APT is probably not the right choice especially now that theres a GUI available for editing Xdocs +Anything for more performance! +i think thats all +It was already either committed or rolled back by the time the afterCompletion method is invokedI agree +Moved to SAVARA v as pisoa monitor will be used for v release but only as a debugintegration test support tool +pull looks the same? I am closing that one +Added some exception handling in to better explain a when looking up objects out of JNDI +Added more tests +Some more thought needs to be put into possible interfaces to accomplish this especially for streams +after r reback test seem more jdk safe +patch looks good It might be a good idea to scan quickly all the places where the getDT api is called to verify null value is expected there +We wanted to make a setup that looked like to wiki meaning x contains appliation +OK I will revert +Hi AamirWeve got james on freenodeI will comment your application today and will send you time availabilities to talk aboutTks Eric +Is this something we should consider for? Since Grant refactored all the stuff this is a pretty simple extensionThe only real issue i see is how to include the noggit library? Since noggit is in apache labs it can not have a release there +Forgot to add the darned file to svn. +Heres a pdf of my proposal +has a lot of paths +Can this be moved to +A separate configuration file is required +ClintIm ok with being only a patch unless you want to use it in +The string passed to the fingerprint method is in turn passed through to +Closed +Well you got my attention nowI wasnt aware of this donation +is the patch for +anyone got the link for the release planif its too late doing the export sounds like the feasible thing to do but what happens when we move to then +An interesting situationAlthough it sounds natural to locate it in the statement spec my guess is that this is best exposed from the since it really is a parm +Ive had to hack ibm jars hibernate etc +I doubt the team will patch the code for this problem since it works in +on the rest of the patchEdit Forgot a not in the nd sentence +If you find further issues please create a new jira and link it to this one. +pdf docs from different scanner vendors +it will parse your number there as well as hex formats and other things like Infinity and I think wont throw exception if the value ends with d or falternatively there is also in ICU +Not sure how Ruby Gems plays into this but upgrading to did NOT fix this problem +I think we should revert this until we can do a detailed review for security +Looks good to me too +Unfortunately this history was lost when during the LOB work too years later obviously looked very convenient for this engine work but we should probably file a bug to take references to out of the engine +Im not seeing the problem after all +Alternative implementation of that has the mechanism for attributes that are indexed +Code added to correct the way that intents specified as annotations in the Java implementation are copied in the component service model +Hi It seems thatstorefile only includes datablock size +Fixed trunk r x r +Agree +Why do that instead of letting the users who want to enable the feature choose to turn that option ON +Waiting for the bug fix of the underlying project issue +As I am building applications that run on top of the status quo is a big win for me +I dont think we should backport those fixes to the branch nor do we have the in place there to test them +I will start work on the backport patch after this gets committed to trunk +weird why mozilla should not be able sart up with a different locale +Maybe we could do that in the dtests +the second patch also removes a silly equal check in third one nukes doShare in as well since those codes are in misc and doSharefalse sometimes makes sense Im not sure whether this is the right wayIf were going to totally omit doShare maybe we need some javadocs in so that this sharing increases fst size is not forgotten +In my experience tends to be the more common way to model this on oracle +It seems that SVG file is saved after the refresh of the designer and for this reason the bounding box it is not the minimum bounding box +There is no fundamental difference between the test case that you have and the main class that i am runningI will look at the test case and see if i can change it to make it failThanks +I didnt carry this forward as a fixed field with but it could be added and honored as a field +the bug is actually that we are passing it up to the upper layer which does not know anything about the follower thread +Im saying this because there a particular workload involving multiple sstables which shouldnt be all sorted and if assertion is not firing it could mean that something is broken or assertions are disabled I will double check that tomorrow +Yes the testcase did not failed prior to my refactoring because when it was not successful after waitforcomplete it just said OKThis problem is there since ages +Modified the patch a little. +I will just run this and start thinking on integrating this with +Patch could you supply a patch as an attachment for this and I can apply it +It would still expose the presence of nodes and properties thoughRight he fact that there is a conflict gives that information away already +Per the soap specThe faultcode MUST be present in a SOAP Fault element and the faultcode value MUST be a qualified name Thus the soap fault message is invalid +htaccess file +I just committed this to. +Updated patch which addresses sureshs feedback +It was used a mix of and + +This feature is for speeding up BOOLEAN queries. +My apologies +Has an alias from to mentors been created? I havent seen anything in your commits +This patch is needed to make Hbase work correctly on Hadoop +I dont think its relevant to take into account the number of inlinks +While is definitely the intent the attached is sufficient to work around the bugNo unit tests though this highlights a gap in coverage +In the previous discussion you suggestedYes you are right +Ok documentation updated. +target path is the path where the file will go in the jar +Example application that replicates the bug + +The environment variable ACTIVEMQHOST can be set to the name or IP address of the machine running +Thats really weird I am sure that these Tests have been running successfully but after looking into it they are definitively wrong +Ive changed the two BDB scripts to avoid use of readlink +Ill remove this as a blocker for now this issue is out of the way for the release +Committed revision + +Will try and get this stabilized in the next +Bulk close for the release +After digging the tests I found the tests keep the out message null to no send the response back +I believe you get two rows because the results have been partitionedinto two groups as you specified in the GROUP BYWithin each group you are getting only one distinct value of fieldbut there are two groups total +may be this should be for asf infra +Oops I take back the committed part Im getting test failures +I quickly looked up jdk sources and it seems that doesnt override equals so doing equals check on lists would result in reference equality test which would be incorrect +I think the effect is pretty minor disk is cheap and class loading only happens once +please set your indentation to spaces for Java code I keep forgetting to correct it in patches of yours I apply +Added transporter samples to doc +This works for but not for +Feel free to move this to CR if it cannot be fixed in time for CR cheers that an address can only have double colon per spec +I think I can change this +Patch committed in r +I just finished my first round of reviewto be fully completed next I am done with my review +If I use an analyzer for indexing the word dogs it is stemmed to dog +Thanks for the review Andrew +HansIt could be usefull for users but as Carlos said we need ASF license +xml should be good s only really placeholders that can become a problem here +I applied +It is basically just your patch moved into the existing if statement +Im also not sure what the logback behavior is if it comes up with a bunch of files +David quotes the SAX documentation which says that startend events must be reported +How are the rest of the migrations looking? We are ready to test the other git repos when you let us know +Enis is coming to officePlease give him some more time +Will include that in the classs docIve thought a little more about the and Im not sure its necessary as there are already constants in for the special exit codes and there are only two ABORTEDCONTAINEREXITSTATUS and DISKFAILED +The capability set used by the service registry is indexed by objectClass so any filter that evaluates against with will automatically hit this indexIt isnt smart enough to do query rewriting though so you are best to put indexed attributes toward the front of your query but in this particular example the objectClass is added as the first clause in a conjunction so it will hit the index right away +More will be merrier +Reopening this one as the other bug is more critical. +If a constaint is violated in a plain text box the warning is a dojo widget exclamation mark +WDYT +I wonder if there is any way to take your example and transform it in a Generic Class that way you could create a Serializable wrapper for any propper subclassobject without having to duplicate code for every class you want to serialize +Hello and Happy New YearNow its working fine with snapshotDoes the version is on track for January thThanks a lot for your to hear that it works for you nowYes Spring is on track for a release in late January +Would a flag append version suffice? The things is that Im imagining the possibility for the Eclipse plugin to go look in the current workspace and find all existing projects thats in the workspace and try to find the ones it can make links to +Ups sorry made ignore my previous is a solved this problem by rebuild my beans I am not using annotation in this case +r changes so that it remembers the JNI address and will only invoke it if that address is unique +Nicholas try running the same touchz command over and over again quickly +Committed to and trunk +Patch removed a few trailing spaces +The parsing problem occurs when the character is followed by n because the value doesnt allow inside +please have a look at with a nightly build +It would be a good idea to let the user specify a partition on a bunch of files as well +I accidentally uploaded older version of the junit test with the previous patch +Explicit changes now applied to and +The reasons can for not implementing it can be found in my previous comments +Peter release is in progress +subversion version +patch for this this case I suggest following RI and do not throw exception for one directional textFrom this point first test is correct it show difference in the behavior for Harmony vs +Good on you DD +Well what is the the filename that I should generate? For GML and shape I used the first feature type name in the query list should I do the same here +years ago or so I once suggested to not overflow but to enlarge the number like some other languages do +I ran the tests before commit and realized this needs a smidge more work +Its crunch time for project delivery and Ive had no spare time at home +Might want to take a look ssee if you can repro this +Is there some added benefit to declaring methods in a final class as final +The patch removed is the static one +this saves me that effort +I dont understand why extending WTP wizard is considered simpler when the only thing we need to create is a new file optionally update gotta be easier than trying to extend wtp I do not think facets are required for any of this beyond java and war since it is not generating some WS platform specific +So I will switch to wsdlArtifactI do not think a dependency is strictly necessary +Logged In YES userid did not null the association between the and the so youare likely using classes from the previous deployment +The definition was changed because it was causing major warnings with gcc due to the negative numbers being incompatible with the type used to represent the scope +this has not been done AFAIK +Filed JIRA +I think that this is a serious security hole +Pinging for review would like to see this feature integrated soon as we keep running into large aggregated logs and have no good way to view a portion of them without downloading the entire log +It basically reproduces the writeread test from testavrodata but does it times and times itIn this case the reference counting isnt as much of a bottleneck +httpd restarted as per Maven Guide +One line Patch that addresses this you try this patch on a james instance serving real mail traffic +Calling it a string when it may or may not be is holding languages that do have real string support hostage to those that dontIt is a string regardless of whether or not it is or Unicode +The correct XML is making it to the client though +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and + +I also added an it profile to each test so that the tests are only executed when the it profile is enabled +Im hoping this will minimize changes to existing test code and also minimize code duplicationWhat do you think +As discussed on IRC this patch creates a with all permissions +Im really stuck because I cannot generate the java files for my web there a workaround +MattCould you at least give a summary list of what you did to make it work? That would help me figure out if this can go in or laterThanksDid you find more issues? are further issues regarding use of in +It might be nice to remove this dependency +Any tries +So this should be fixed now +Detail of the updates +Patch apply thanksI just deploy a new snapshot with this fix please can you test it in order to be able to close the issue. +I need to put the ASL into so that it will automagically do that +Ah +it prints abd now +Done with new UI +Release Note not required as this only affects internal Red Hat testing +Ill experiment with a few font families and submit another patch +fix added on Review Board +Currently only links in src and href attributes will get rewrittenThis patch affects module onlyThis patch adds dependency on module +Good enough for this to change fixed for +NethanelOne more thing i forgot to you change the latest axis jar alone or anyother jars like saaj and along with it +Instead I thought that it passes on both versions +Ted please test this and see if the error goes away if so please mark as resolved +I agree with the change to +maybe the exception could be improved especially it should not give the BUG! message since such an error should maybe expected +Recent logging are an example +Defer to +Thanks for your contribution Trevor +Wouldnt providing b and c be the best combination? But maybe parameterise the ensure read by a logical timestamp +Great +That is a major messll go with Stuarts proposal unless anyone objects +Ive made it work now so you have to split up the classpath again like I did above +This patch includes a new Test Framework Javadoc link from the Solr websitesCommitting shortly + + +Hi RobertI dont quite understand your approach because we need to provide thepath of the profiler libraries to the instead of thetasks +New mockups with changes incorporated + for patch +Its harmless though +Rebased the patch with Teds commentsDeprecated all the implemented apis +It would be helpful to have the release note for to understand the scope of the behavior changes +Without these methods the scheduler would have to scan the list of jobs in the every time it needs to make a scheduling decision to see whether it has changed from last time +The HTML files are included for review +I sent the master and region server logs at DEBUG level to stack at his gmail address +Looks ok to me +Hi Thanks for the bug report +If its necessary I can put the blank lines in again although I feel this is rather petty but heyAs for the bigger picture I agree that this has gone from what could be a biggish improvement to a rather minor one but thats what Doug wanted so as not to introduce backward compatibility issues +We can reopen it if we decide to backport it +The get methods in now throw an when the values which are configured by the user are invalid +Updated patch for supporting useOrigin attribute previous version failed on having configured a root has been somehow fixed with since now useOrigin is set on the cache itself and a cache with useOrigintrue can be used with cachefileset +Pretty much all the multi war issues revolve around the statics in and Lifecycle +Unfortunately I wasnt been able to validate it is working as expected because I cant compile gridmix on my system as it aborts on some generics complication issuesAccording to a multiple confirmations from other ppl +Thx. +To you need to do the followingA Compile B Run +consider for GA I had to push it off to a release due to time constraints. +compiled problem occurs in also +You might consider changing tofrom to to or from to cut down on all those slashes but its fine with or without such a change +Please report any issues you encounter +In addition cpasrcmainresources +indeed we are definetely not going to build this into the core abstract behavior +Without the changes to Requestor setting sendfalse now fails with these changes is passes +Using node packaged with webworks does not cause an error +Closing jBPM issues resolved in a released version and reported by me. +When I save I can see connection working message +I think I can fix it +Committed to trunk +Could someone please review the patch so I could move onto sorting out the Cursor stuff? in the way the input buffer gets filled with data in order to avoid unnecessary arraycopy and eliminated Cursor related classes +Planning to use testUtility to allocate ZK cluster instead of explicitly allocating ZK and clusters +Maybe Im overlooking something but I dont see any place where the HANDLE ptrdifft data types are interacting with the and membersI think the changes related to the and members are great but Im not really sure what was accomplished with the HANDLE ptrdifft translation +Committed the test services in Rev +and derby tests passed +Brian told me that there arent any differences during the text extraction test +the point is not running but starting from a fresh checkout trunk and at end do a Eclipse Clean project to rebuild inside Eclipse using the +is buffering data from the codec but is checking the codecs position after each record returned +fixed in the trunk and in in JBDS Linux to set proper fix to issues that were set to +Input needed for redesigning portlet. +I marked this a blocker because it is what is causing our rigs to fail +Thanks +But I am confused +Duncan yes Ajax style fetching is another option +Anthony I see no design docs or ML discussion +Attached patch that should be trivial to review +Ok I will remove the shading at update on when will be released +Laxman seems reasonable +Looks like some issues were handled but quite a few issues are still open +I see that the change is only to serializing this class + +dont really know what happened here +It has been fixed by Revision in ESB CP Branch +everything seems clear and simple +Please go ahead with your proposed changes +Please review not happening any more +If you start putting commands back on the queue you could have weird createdelete command retry reorderingI also have not switched to requiring or respecting a replication factor I was thinking perhaps specifying nothing or would give you what you have now? An infinite rep factor? And we would enforce a lower rep factor if requested? For now I still require that you pass a collection template and new nodes are created on the nodes that host the collection templateIm not sure how replication factor would be enforced though? The Oveerseer just periodically prunes and adds given what it sees and what the rep factor is? Is that how failures should be handled? Dont readd to the queue just let the periodic job attempt to fix things later? What if someone starts a new node with a new replicas pre configured in? The Overseer periodic job would simply remove them shortly thereafter if the rep factor was not high enoughOne issue with pruning at the moment is that unloading a core will not remove its data dir +Hi DavidWorks great thanks for your helpRegardsScott +Thanks a bunch Stefan but next time please submit a diff instead of the file itself +Or am I missing something +I just noticed also the to method should be explicitly tested going to work on it and attach a new on trunk at merged to branchx on fails in x because the constructor prefix parameter cannot be issue with in branchx +Please always attach such a big documentsdont paste half of mega Kb into a bug description +Ive checked the alternate wkt form and it seems to be working fine +Ive copied everything off. +But with it works withRafal +Well in fact the wizard should only ask about adding as it got dependency on junit +As such the dummy file is good enough +Ive looked over Robs proposed fix and it looks good to me +Moving to CP +Reopen if you find it not tobe the case with the lastest from SVN or in the next release +For me all tests except for work ok +Attaching the log file in case you are not able to access rhts +Likewise have the AM also run with the debug log level if possible +Where could I install the bpel editor? Please give me a linkIn the update page there are a old link for Developer Studio for Web and SOA Development when I try to install it it say Developer Studio for Web and SOA Development will be ignored because a newer version is alreadyBut I dont have bpel editor in my jbds +Checked in a number of optimizations. +Im not really sure eliminating this instance will eliminate the Parser +Thanks Ram +Please check that performance is ok +Is that documented somewhere? We should set up a Hudson job to run them +Why would there be anything other than byte array in the memcache? In case someone was silly enough to have stuck some arbitrary object in the memcache used by throwing seems perfectly reasonableOleg +Thanks +That is why we think on a networkConnector shall come to rescue but it does not have any effect at all so far +Mladen Im moving this out to but you obviously have the option to land this earlier on trunk if you like +Adding wrapper around so that I dont have to maintain new map for app +Hey Karthik +committed +So we might remove that no? +These blockers and critical issues are resolved but not verified +The actual distance is calculated somewhere else in the flow I havent looked into where that is but I believe it changed from to due to the refactoring and this is one reason it was no longer stored in the WPVW for each clustered point +You are right i didnt notice tikas ability to parse pdf files +Im using maven when a child project has a profile activated by a missing file the profile is always active when run from the parent folder because the file is never created in the parent folder. +Even if it did any operation requiring the block index would have to load the entire block index because it does a binary search +This happens with EAP as well +a failover or just general network problemsRPC retry should be done in RPC level +xmldocument xmlexists and additional support for xmlelementxmlforest would be good +Trivial patch +Patch applied +This currently affects the following tests chartsSWC m going to close this issue as resolved unless someone can reproduce it +I hope we can iterate on this quickly and commit fast since its blockingI propose to commit this by Sun night PST if I dont see further comments we can always file jiras +Nit remove the commented out code below Set RPC engine to protobuf RPC engine since you set this via the command line option +Thanks Craig the initial run through looks good now +Thanks! +However doing so prevents us running the full test suit +Lets open a JIRA to escape a more complete list of characters +some eap docs are public some are private but they are moving enterprise books to public repositories have enquired as to whether they require any particular structure before moving populating that the docs referred to above are the docbookXML translation source files not the final switchedconsole code depends on which version of jbpm we end up withneed to merge into at some suitable juncture +Committed to trunk. +new config using vm broker url on fixes branch ready for +This issue is very old +Committed a slightly modified version of Davids last patch in rOther than minor javadocformatting changes to make checkstyle happy I also made the shared source of randomness pluggableI am not happy with the static randomGenerator attached to though I understand and support the need to ensure reproducibility for some applications +checked in +Im going to commit this shortly +will post a prelim patch on this soon on review board +The new usage of a is actually configurable via a new builder property +Shell script to provide the and via maven repo +Can consider this later +I want to share with this list another note I wrote to the European Legal Network in response to a question there One of the key stated rationales for such as the ASFs is to allow the Foundation to relicense if necessary without polling all the individual copyright holdersLets please use terms VERY carefully +Committed to +Slightly confused I have checked that the patches are included in my current build of the trunkHowever I still get the file not fount error tmpdata as decribed in the SubtaskHave I got the versions right +Its available both on JIRA and phabricator +patch to raise an exception instead of ignoring replicationfactor option in NTS +Otherwise patches with javac warnings may get in without being noticed +How would you want to configure the scope programmatically? What scenarios do you have in mind to use thisIn any case this sounds very much like an improvement not a defect +I just checked it and cant find a problem +for it +Seems some part of does the trick +Updated the patch with some early review comments and added a testPlease have a look at at your convenience +Ive updated the class to provide a find static method which is just like get except it returns null rather than throwing if no default constructor is found +Cool Ill have a look tonight and will let you know the result +I examined prior branches to see where the changes were made and its only in trunkwhich is why I didnt see it until checking out trunk to make the changes +Im testing the patch now +Assume closed as resolved and released +Closing issues +Patch containing fixes to implicitly spawn streaming operations within the OS default shell to handle scripts more uniformly across processes +Fixed the wiretab all issues +any solution for thisSTHKS think we can not use any variables like dot in AxisHowever it make sense to make minOccurs considering your initial two points +All tests passed on x including weekly and nightly +If your Mysql db is using can you use the +That issue has now been addressed. +The problem seems w false +Hi JaromirWe are not able to browse the queues if the messages are near to When we click a queue with more than messages in it we end up in java heap space errorThanksSubathra +Im using the version posted by Martin Zeltner and have run into no issues +This patch removes optional fields from constructors altogether +it would be nice to extend the parser to accept infix but thats a different issue for another day +But izpack overrided my class at the time of packs are installing +Need to add instructions on setting up logj to output to syslog port and have flume listen for it +See how hard to do +We are not replacing anything +Also I had missed one synchronization case which the test found +Ok +Ive updated the patch to do this +Part of a bulk update all resolutions changed to done please review history to original resolution type +bulk defer minortrivial issues to +Closing issues for M +It is a very simple implementation that is currently +The patch addresses the problem +Central already requires the AS Project Examples Seam Runtime +Committed to +I dont have a vmware environment handy can you tell us the mount pattern for devices on vmware +Instead of throwing from sqoop method I catch the exception and throw as suggested by Arun +Sure as you see ant lib +The question when to set a timestamp into BU when no immediate flush? I think like in transactional case where BU is not flushed immediately thetimestamp is set during insertion of the BU into transactional log samecould be in the case when BU is inserted into cache +if you are using coordinated versioning then the parent is always released with the child and this isnt a you arent you should leave the parent version as the released version until you need the snapshot +Moreover by default the option Only NEW violations should be tested +Btw on my HD widescreen monitor I can crank up the font size to without any scrollbars +Tested with SNAPSHOT in maven repo and it seems to be working finethanks +Im not sure what your environment was but I did once see a super slowmgmt server two days ago and it was actually running out of memory +ALR clarified and I confirmed that remote would work if mappedName as not specified +This is a lot more complex than I think you realizeIll admit that the documentation on this class is sparseat best and ignores some of the most important issuesPatches are always welcome +to commit! Thanks Doron +it should be issue +So all issues not directly assigned to an AS release are being closed +The issue isnt really fixed +Looks like this patch is for trunk +Fixed in r w testcase. +This is working for me and I see it in your demo +Looks like the group cache bombs on restart because of a bad group member +Ran ant test Dtestcase +forgot to close this earlier. +The is merely a holder of flow definition metadata and actual flow data +Patch for trunk uploaded +Closing with committed unit test in Rev +It also includes APPC however that part seems to be a little buggy right now +I mean EWP +But we had to change that when it come to real module like security and RM they wanted not to throw that exception +This is the last file +I do not believe this restriction would be acceptable +It looks like its still not working +I will still look if theres more info +Also if you try for instance to search a party group containing a string or a letter and hit enter the lookup surprinsigly attach a new patch WIP not intended for inclusion more for myself actually +is attached to set it to empty byte array for keep it null for. +Thanks for raising this +on the docs +I improved the code by not locking globally and use a blocking queue for scenarioPlease reopen if you think there is a place for further improvements +Sure thing +Hearing no answer to how to continue optimization with a single mapcontext Im closing this +Normal load balancing will balance overall cluster load over time +Hi Alan it isnt an Hibernate issue +provide steps to reproduce and we can reopen the issue and take a see this in Grails with our app +Please close this bug if satisfied +The output from WSDLC including the jar versions used +this issue is always happened no matter what I add a parameter name or not +LarsGo ahead +Ill take a rest until thing quieten down +Thanks for the patch +SBS provides margin by default. +Emmanuel what exactly is the status of thisWe need this resolved before releasing +Probably current is the best we can hope for +Im moving this out since theres no viable patch for this for +If you leave them in the program does not run at all +Attaching committed patch as I had also cleaned up a couple of whitespace and one relevant indentation fix while committing. +Max +I think we want users to be able to define a default CL +Did you ask the Red Hat documentation team what they use? No I didnt +HelloWould you please have a try? Thanks a lot +I allways end up on line with a check to a parentItem that is invalid as there is no parentItem around +It seems that the examples are now prepared to be used with RHDS but this is just my guess +Have it all done including the changes to the class names +Unit tests have been updated to cover these cases +So if the original problem is fixed as you say then we should close this one +See if same is true for Hudson +Thanks Dave for the backport. +Commited the fix for this +SVN is not setup on my VM so please review this update to README and commit it +This is the patch for DIH version included in Solr distribution +Hi Chrisunderstand your view point +Is all the required work for this in complete? +consider for as soon as it does not impact other class where user can define floatleftstylefloatright moved to default class. +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +but oracle is not yet +The more I think about it having as the was all around a dumb idea +Seems related to +I will have a lookThnaks Andun for easy solution! +These operations will be fairly common in the optimizer +as the workaround exists and we have to wait for underlying issue fix I am moving to the next version underlying JBIDE issue was resolved. +Resolved in r +The second which attempts to verify that override doesnt change the existing behavior fails +To add schema Add schema file to incubatordirectoryevetrunksrcschema Modify incubatordirectoryevetrunk with schema and any deps +good i think it is better because with only the jar if people create their own ejb packages they cannot forget the enterprise issues on which I wont work in the near term +Closed all resolved issues related to our projects. +Patch file including test cases attached +So the Jira is about adding support for the system property +Cant reproduce the with Firefox and for Linux for Windows +Eventually I will have to use more from camel directly though +Anyway there is still an error when wanting to create a PDF file with your charts +The best way to build it is to really run ant war from command line +Anyone feel free to reopen if there is serious intent to continue preferably with a patch +A new patch will also help get a clean test run on Jenkins +Attached patch fixing some more nocommits and renaming +Thanks for the patch Yuan. +. +The primary issue with a solution is zookeeper +In the meantime you can edit the existing links to refer to instead of to get them to workPlease close this bug if satisfied +Wrapping content of the scrollabledatatable column in richpanel works fine for meUsing ajoutputPanel doesnt workTested with Richfaces Final +Should we add it there instead +I notice that you mentioned this issue in your stackoverflow post +I apologize for confusing the issue +It may not be every last one but its most of them +FWIW +Japanese translation have done +Do we want to hold off breaking it up to and just deal with the missing field? Or do we want to take the time to break this thing up now +Aurelian The patches look pretty good but applying the togo patch causes to raise a +We should have a cleanup +I have two small comments and a request The original reporter already added a test case for the bug toI think the new test cases are quite good and should be retained but I think weshould also enable the existing test case +ldapsrcmainjavaorgapachedirectorysharedldapcodecsearchcontrolspersistentSearch +Bryan Can we close this +I am not in favor of adding more property files as it gets confusing for the users and difficult to write instructions +Did you mean that we need to upgrade Spring Batch +This is expected and not an indication of a problem here as far as I can the fundamental question remains Why does the transactionally created ConnectionSession pair with synchronization on not automatically participate in the ongoing transaction as well? It would be great if you could debug this a bit further analyzing the Session thats passed into your etc +Changed affectsVersion to because will stay LukaszI have worked on +Version have installed fedora and under this os the issue reproducible for me will look into it +fix applied to svn thanks for the report. +I therfore resolvethe issue fixed. +Resolving as fixed +Two changed imports only +If another issue arisesis confirmed re EAP on AIX a new issue will be verified +bq +See +Yep I already caught it and have fixed the branch and trunk +m still getting an error in STS GWT M when trying to add the one side of relationship + +push out to +Can you keep us posted on it +wrong reference was deleted +Hey guys we are ready to work on this as Forge is in CR and heading towards FinalPlease let me know how to proceedThanks think that is a very good idea +Since it is specific to the catalog +NB +So apparently findtype is not enough as it leaves all the SQL NULLs bound as CHARS +Yes patch doesnt depend on this issue +Dont use +This patch applies to both trunk and the branch +Security constraint in contains tgmixedoverride and servlet is mapped on tgmixedoverridesrvTest is failing because access on non confidential transport is possible and IMO it should not beThe truth is that if one changes security constraint to tgmixedoverridesrv test passesI cannot see any part of spec +Your approach is simpler so Im on that too +any objectionsThanks for cleaning all this up James +Sorry for my mistakeI do not need this ticket + +Confirmed that it was an xbean issues and have developed the fix +to change fix version +this issue was solved M can be installed now. +Ah! I had not made the connection between the BSP and the type attribute so the error was in my understanding +Yes but having split individual nodes data to more subranges we can make negative impact of not well balanced sstables practically negligible +proposed patch for the jackrabbit API as well as a patch for implementing the changes in oak +Manually tested +I could not figure out what I was missing +confirmed fix +Andor check out the further discussions should be taken to the forums because JIRA is not a good substitute for a discussion forum +Figured it out +This has been fixed in can you please try the SNAPSHOT to confirmthanks +I agree with you about dependencies and as you said we can move the classes later if needed +sketch of MVC abstract portlet and of existing MVC that were discussed Interceptors the code to support class lookup implementation +I checked this patch on Windows with +So the RPC version bump looks like itll happen soon +So the finalizer for will close which will close the associated channel and will just be responsible for tidying up that map in Ruby +sorry about that. +OK thanks for the analysis +I tried reproing but I aint too good at waiting around so missed repeatedly +Please review +closed all issues which got released with +And no I am not a search the duplicate issue yourself +We will test this in a test environment +Thanks for your review Chris I worked on a new patch I hope it solves theproblem elegantly enough +Is the checksum type also stored on NN? If so we can prevent the concact of such files +How would you like the API +to set fix version +Do you mind taking a look at adding that to the patch +Thanks Fil Ill take a look +on fix forMaybe there is a way to do without references we already mark the parent it has daughters? Maybe mark the daughters with their parent and list references +It was an incompatible change and users were expected to update their code to be as they migrate from toThe migration has been slow and we have finally decided to support symlinks in +Create web service using the wsdl file +Marking as resolved since the test cases now pass. +Can any one confirm this solution +Cleaning up the patch to get rid of would probably be a good idea +Ive checked that eclipse startup is now fixed +For applying the patch you have to first do a svn renamemodulesqueryparsersrcjavaorgapachelucenequeryparserflexiblestandardprocessors modulesqueryparsersrcjavaorgapachelucenequeryparserflexiblestandardprocessorsAfter doing this the patch applies +AFAICT this is resolved in. +SVN s great that you were able to create a patch +Fixed with build of +Ive discussed these changes with Li Xu and he is going to review +Note that the example itself is not a scenario I choose to use the Spring interface as the bean interface because it is readily available +still validates the task directories + for skipping the tests on Java for +As indicated above the third of the feature list has been implemented alreadythrough bug +Ill see what I can do but feel free to step in and help me get this right +This keeps coming up +Thanks Jukka! +Patch fixed +LeeThe message The given paths +The changes look good +I tested this with a little java program that instanced a and calls set +Thank you Rich. +Duh +Got it + for the to trunk. +I have no idea how this approach would compare to a Thrift service +I am thinking that the namenode has to spend considerable horsepower to keep forwarding block reports and blocksReceived to +It certainly doesnt matter to me how it happens +for fixing the test Jonathan +Reattaching for hadoopQA +Ive worked around the problem by changing the name of the directory in question to something that isnt a substring so I cannot justify spending the time to do that Im afraid +I also made it so we exit with code if successfully wrote compressed file and! if anything else +This reverses an optimization made inCan you please provide benchmark output demonstrating the performance improvement? The output from before and after this change should be sufficient +added also test cases for profiling classes and a convenient callback classIn the profiling should now profile creation of creation of creation of Action execution of invocation of invocation of interceptors invocation of action invocation of invocation of Result +no test code available +Hey CodyI have committed the fixcould you please check that the selection works for you again +Will also adding something similar to the postgresql convertfrom function and possibly a binary type +Hi SiI think I already have something that does this +to be clear if we did start doing this the nightly site would still need to be there for old builds to work AFAIK thus snapshot shouldbe an additional place possibly just a symbolic link for nightly +Ive got gc logging enabled +If I make a Maven package for XOM can Maven projects like jaxen still use it +The patch I provided for doesnt fix this issue as the ruby arguments are left as is but a small change to the batch and the understanding that these arguments require quoting would get this example working +Lets investigate some solution and try to make it injected in +Ill update a patch to log that we need to set related value to enable ha feature +Closing issues +Ill look at this issue +Thank you Ivan +One trend Ive been randomly introducing here and there is the use of in the file +QuestionAs far as I know HDFS has implementation that helps protect against changes during execution +batch transition to closed remaining resolved bugs +We always appreciate help from the community in spotting bugs and we always do the best we can to deliver the fixes in a reasonable amount of time +However when the readonly attribute is set their values will be submittedWould be of huge help if this would be mentioned in the documentation too +Shard leader is marked with prop leadertrue in its cloudstate propsAlso some of the tests are improved so that they are less likely to fail due to timing issues +Done +I dont think its a reason good enough to abandon the idea fixing the issue would be better IMOSo I think the fix should make sure that eviction of M is only done and flagged in the test root configuration only with no influence to the runtime root conf +There is only one issue at stake does not work out of the box when deployed to GAE in no small part because of odd GAE security model dealing with Reflection +The tag was added for +Also why does the file still define CORDOVALIB if its no longer used +Its up to your clients to ensure that the names of the temp destinations that they create are unique using a UUID for instance +I dont agree that build should depend on the order of dependencies for now I roll this change need to enforce rules by tuning the comparators passed to the not internal like I am wrong it really behavior of the resolver when the system of inequasions has moren one solution final selection is completely in the hands of the optimization function optimization function uses to orders inside order of used to create optimization function defines which of otherwise equivalent solutions is selected this GA ordering is defined by a Maps implementation used by the solver s behavior changed from Java to Java thus tests failure Bens solution to use which has predictable behavior between Java and Java seems to fit the billI Bens fix and will be running to make sure nothing else Ben +We resolved this issue by making the sessions persistant on the load balancer. +This patch contains the fix I suggested +If adding of additional property to invoke procedure needed then please reopen and target to another fix version +Reclassified to RFC as there is no columnClasses attribute at richdataGrid +Im running Jboss Jbossws native +Also I have included modified golden files to make the log to phy compiler tests pass +Nick Thank you for reportingIll add a new sample to Metamer to reproduce this issue +After reading that byte I get the local valueAlso this happens when the DELETIONMASK is set on a record +Thanks for the clarification that this is intended to be used internally only +Okay got it in now +This issue is closed now +Investigating +Lets get this into as well +Barry anything you know about? put on CR for review assuming you might have info about it +install finder class +Thanks for the fix +Make default similary configurable maybe make sense Such as We can use instead of just through modifying configure rather than writing other custom schemasimilarityfactory +does mess w scanners +It is to merely note the scope of the problem +The patch looks good and fairly simple too +The code needs to be changed to use the post hedwig api +Id prefer if we tried to keep our maintenance branch changes confined to bugfixes +YCSB wont work I use a ruby script now nut that take some time to create a excessive number of columns +But just to be clear this is not an issue of the client side its the message broker that has the problem +This happens in thecontext of passing headers not the message body content +I only started to use Maven after I had trouble with in Maven gone through this myself my feeling is that if Maven can output a message about going away in it can save users quite a bit of headache +In that context what does it mean to disable allocation of containers on a node which sounds like blacklisting the node +Only one nit now with the change some shouldRetry methods no longer throw exception thus we can remove throw Exception from their definition +If you can produce some smaller app that I can build on linux or Mac I can look into it right away +This issue is fixed and released as part of release +But in this case I get exception because example does not expect nil message +What do you think +Good catch Uma +This is going to be related to our batcmd scripts and were pretty poor batcmd scripters +This was due to not having any PKI certs in the system and then attempting to we have doca link to how to add a PKI cert +Committed revision . +Any opinions +I also improved documentation +Committed the Maven POM in revision to get people to test it out +Or we can strip out the schema string and put it in a metadata sectionThat would require the parsing of schema string to figure out that the schema matches across the files before merging no? I agree that there is a benefit of making the files self contained +for the patch +Both and use to pool codecs +How about keeping the calculations in bytes but allowing ppl to use GBMBKB suffixes in the config file +Can you give me the test case that when you commented out it worked fine +itemStyle and itemClass attributes were fixed and tested on dev +When mvn site was executed on the latest checkout of the javadoc plugin the source URLS are already updated +wrong priority +Im getting the same error I cant find this plugin in any repository +Closing for now as I cannot reproduce it +SQL statements in was modified to run only forIve also changed the default to use mysql +ThanksAngelo +Yes this issue is related to Java DSL +I think it is much more important to get the pread performance up than add a second datapath +Which is as I would expect +The idea is to the property only if user did not specify the property value +storage live migration is expected to fail if source and target hosts are in different datacentersBTW wondering how could clusters from datacenters allowed in single cloudstack zone? May be setup issue or a bug in zone to dc mapping feature please raise a bug in such a caseIs this upgraded setup or fresh installPlease provide commit id full management server logs +I have committed this +Temporarily reopening to update release note info +Less critical than what I thought because the default for most is hrs +I think its under trails +Yes obviously the profile should be activated at that time + Forge JBDS commands Clear the screen clear Create a new project in the current directory This means less typing +I think that it might be interchanged easily as the adaptation involved interface implementationAnd I think that has been extracted from library as I needed this functionality very quicklyRegardsPavel +Its definitely somethig that should be included in Ruta sometime but I think it wont be straightforward to add it +all seam demos are checked +Erik I cant get your patch to apply cleanly to solr patch p i patching file solrcoresrctestorgapachesolrhandlercomponentpatching file solrcoresrcjavaorgapachesolrhandlercomponentpatching file solrcoresrcjavaorgapachesolrhandlercomponentpatching file solrcoresrcjavaorgapachesolrhandlercomponentHunk FAILED at +Im including modifications to samplecontainer in this patch so people see how to use the code but I wont check those in +Perhaps the best approach is to treat single core as multicore with only one core? me some time to understand sorry YonikThere was a hole between finding a core and acquiring itI hope I got it right this time +lets add a note to QE to validate upgrade with various types of config edits +diff tested on mac os and centos bit + +could you list themSounds like something that should wait for IMO +Give me a chance to notice any other issues +It seems that all RPC api with String parameters have this problem since it does not prevent the client putting a large string into the parameter value +For those who want to try out the CA portlet on codebase +Seems like a reasonable feature to have with a slight caveat that the retry limit should be bounded by the limit configured on the RM +Ive been testing quite a bit and havent been able to recreate the issue and Id rather go forward with an official build rather than my hacked version +Committed to trunk. +. +exception namespace was not a good idea +Resolving. +Moving integration test related issues out of any particular release bucket +Not sure I understand +I am having similar issue where tomcat does not shut down properly because of activemq threads +Until this bug is resolved the work of getting certified to run on Railo is pretty much stopped. +Because of that we migrated to and kept alive +The spec will be revised after this syntax change is checked in +When the XML file attached here was built using axiom and serialized using axiomnodetostring the output xml is not equal to the input XML due to the fix for this issue +Closing all resolved issues due to a successful release +Nevertheless is above seems to be a reasonable defaultNote that here deny means OGNL will not automagically update the value referred to by this parameter +Am I missing something? eric see updates at a reasonable rate IE soon after each flush or every k bytes or so with less than a seconds latencyThats a new requirement that Id not heard before +As far as I can tell the problem appears to be with references to global variables +Regardless I just wanted to upload the latest patch before I move on to other work +And the code is slightly different +It should be mentioned that crashes even after the is applied +JuanThanks for wrapping things up +Thanks again +Applied +Some of the scenarios I wasnt sure if what I got was to be expected and one scenario might be a real problemScenariosItem Specify an with includePaths id property on a no Indexed on the OKItem Specify an with includePaths id property on a of no Indexed on the no on FAILItem Specify an with includePaths id property on a of no Indexed on the with depthx on OKItem Specify an with includePaths id property on a of no Indexed on the with includePaths with an id property on FAILItem Specify an with includePaths id property on a of Indexed on the with includePaths with an id property on FAILThis last one is a problem because it basically means that if an Entity is part of a includePaths and is also a an Indexed Entity itself it cannot includePaths with id properties itself Item Specify an with includePaths id property on a of Indexed on the with includePaths with a normal property on OKItem Specify an with includePaths id property on a of no Indexed on the with includePaths with a normal property on OK +we should go the assumption of file job +Done but also need to check if we can dynamically limit the length of textentry in inputText fields +Thanks Tim verified +I agree with your statements +Heres a version with more tests and few bugs fixed +Related to and +This issue was migrated to +NET dependent +Assigned to Sarah +Attachment has been added with description The test case +Im out of the office on vacation until Tuesday the th of January with very little access toFor technical questions or concerns please write to happy holidaysAndras Szerdahelyi +Thanks + + +Ive found the absence of this feature to be a real inconvenience when mapping to legacy schemas however Im starting to wonder if it is ever going to get fixed its now more than two years since the issue was raised +Thanks Shanyu for the contribution! +Cool the patch is still very +May be we can close this issue now what do you thinkNo one regrets the old logo +The release you are running is a snapshot +It would be so nice is someone who actually understands Japanese could confirm +yea but that API isnt til +We may let user applications to specify the compression codec +I think we can close this +DennisThanks for the feedback will be updating the release note +This included creating a new class this bug to the groovy project since it seems to be coming from the compiler not greclipseI have confirmed the bug on and Paul has confirmed it on +If we check META for the first region and its not there cant we just stop and say the table doesnt exist +Were spinning off the build problems into a separate JIRA but Im unable to duplicate the issue youre seeing with the custom source folder +But still wed like to know for which use case the this lock mode is intended +have fuse dfs rename things into Trash +Meta action failed due to method that was fixed in it is in trunk but not in yet it should be backported to there or in separate JIRA +Any progress please +Eliminated descriptor and the generator generator manager to simplify and to assure that the generator names are global also across generator types +I didnt trace through the code far enough. +Because if other operations have this module engaged they will require its special operations to be available on the service +We have to consider your thought here +Wrong +Just tried c using Oracle JDK u on +sorry this is not possible for Version we can add it to our future wishlistOr you contribute to it +The applied patch does not increase the total number of release audit warnings +Not sure about those changes since they may not meet requirements +No +Moving the taskdefs out of our common build file fixed the problem. +Assume closed as resolved and released +Please let me know if there are better ways of submitting a reviewThanks +Any update Brett +Sample couldnt reproduce this +Thanks +Edited description and summary upon submitters request +Jndi source is changing fast these days +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I ran the tests that failed in the previous run locally and they all passed for meChanged the names of the backported Quantile classes so that ifwhen a new hadoop version is revd the names wont conflict +This is already taken into account by Maven itself +I think that it may be an aritifact of me using gasp clone +What exactly was the issue +Though Seam is currently in maintanence +Default is to run against the Derby Client so youll have to start the server on first +How was this resolved? I dont see a patch and there isnt a screen shot or screen cast of what the new message looks likeDenny can you please add a patch andor screen shot to this JIRA so we know what was changed? Otherwise QE is going to be scratching their heads wondering what they need to test +Patch attached +And Im not really willing to support multiple launch configurations +The connector jar was not in the classpath so i added it and apparantly got passed the CNF exception however I now am hitting a Loader constraint violation because the class is getting loaded twice +At the bare minimum we should have some basic tests where we have multiple cache managers sharing a channel and see if we can operate against these cache managers and deal with lifecycle events of cache managers in the correct way +The fix to read the epmd stdout and stderr streams seems to be working OK with no build hangs so far when using the current RB release of ErlangOTP and the current version of jinterfaceI have therefore committed code to read the epmd stdout and stderr streams everywhere that epmd is launched upgraded the jinterface pom dependency in and reversed the previous commits r and rFixed under revisions r and r in the branch and revisions r and r in the trunk. +I think its OK to stay in contrib since these are functions but I agree with the other itemsBill can you take a look +Got a new laptop and and forgot my password +Dog is barking +there is already no guarantee a ec instance will boot having one hang up during config adds for one more thing to debug +Should be verified in gatein master and reopened if still valid +in memory usage in this scenario I dont see anyway to significantly decrease the memory usage +feedback in think it is only important that users have the functionality +We are no longer supporting the python contrib +can you make a minimal test project to reproduce this +Committed revision +Applied proposed patch in Rev +schedule for fix in release +If we commit this change well use the svn mv command above +Closing +Am I wrong about that? Anyway theres more to do here regardless will get to it and post something ASAP +Is file an absolute path? If I remember correctly distcp cannot handle relative paths +I also double checked that the new test case fails if the fix is removed +Just have filed to do the refactoringI really dont understand why introduce code that is due to refactoring on delivery +It seems srcjava has been moved to srcmainjavaThe srctest has NOT been moved to srctestjava howeverCurrently except for java mail all the test sources is in srctest +Also modified to do the same + +comment +Is this correct? +So this is not a regression at all +It seems like we might as well always ask for a persistent connection +Closing stale resolved issues +Hi WillemFor the spring orm optional import problem we neednt hack spring orm jar to add Just useosgirefresh spring orm bunde id after you install hibernate bundle jar so that the optional import could be resolved now as the needed package is already thereBest RegardsFreeman +If the user is explicitly setting a bundle name instead of using the defaults that bundle definitely should be there +Thanx Edwin for itOtis what about contributionsanalyzersnl + +Brian I consider the schema public API and hence raised the level to blocking +This patch also passes jdbcI will commit this patch shortly if there are no further comments +Here is the patch that removes the unnecessary whitespace +This highlights another more important issue which is the fact that even if we set startstop priorities for components within Infinispan these might be rendered void within a managed environment so any assumptions we make in core about these priorities are only partial which can lead to funnyodd bugs like this +Meantime Ill open an issue to look into how this help stuff is being interpreted by IRB +Thats true +I was actually reluctant to invoke the plugin discovery dialog from editor because it would create the impression that every plugin must modify and apparently it does +Im still working on updating the jUDDI docs themselves which will contain more details. +We should really create a new JIRA for the additional aspect noted and relate it to this JIRA so that people can properly track what was actually fixed when +Here is one +Attached is a new version that is a hybrid implementation +The workaround with LDLIBRARYPATH works good on RHEL +Thanks ClebertClosing jira. +Ive heard of font size issues when upgrading +I dont know whether this session expiration has anything to do with if not please tell me and Ill file a new bug +Another Keychain integration would be nice on X comment +Im changing the title and description of this issue to reflect the new scope +bq +OK closing +Need to create a bundle out of this jar for working with you use HL in IPF could you once you got the time upgrade to use the latest HAPI all resolved tickets from or older +I forgot about my question about exceptions +Dennis could you please regenerate the patch with the current trunk +Richard please update once this is the s option in the test case so that the JNDI lookup which twiddle uses to create an will be made using the AS instance of the testsuite and not located via JNDI autodiscovery +bulk close of resolved issues in. +In our case debug is for developers of WinkINFO intended for developers that use a library +You provide nothing that any of the devs allow to reproduce the problem +Most of the work to port the build to Apache is finished now smaller issues which I guess will come up will be fixed in new and problem specific jira issues +Again as a submission to a new module +Option Parse for as and parse as as is fairly new it wont affect a lot of usThere arent a lot good options here Im not sure how I feel about this one since its definitely a hack but only appending the number to the DC if might be the least painful for existing users +Changed my mind +NOTE This patch also includes the fixes for +Now that work is well underway with AS all previous community releases are +the project file cannot be loaded +But maybe we should update the documentation to make clear what is meant by empty +which is a jmeter test that calls the test page tag to provoke the leak +Windowssvn +if hash is same then namenode and datanode have the same set of blocks +started defaulting to the eclipse compiler with and janino dependency was made provided in not sure about jsr as that is still a dependency in the module if we are no longer building this module that could explain +I have checked it multiple times +Not found this issue in latest automation results +it works. +Patch appliedYour code regardless of the size adds a valuable feature +Now I set one and remove it afterwards +You shouldnt have to know a thing about auto commit or have any config for it to work fully with commitWithin +The published site just needs to be updated. +In the meantime can you answer Alejandros question wrt we we need to keep the SSL HTTP configurationWe dont + +I have figured out that fallback converter which I coded up is causing this issue in all versions starting from +Fix is in CVS + +Lets see how that goes +If there are no objections I will commit this in the next hours +I committed the patch to trunk +Ideally I would like to see one consistent Metadata API defined in the Connector API +JSF form validation in Seam two xml code blocks without the opening lt bracket hform +Committed +Unfortunately as far as I can see in the current design there is no way to specify a guaranteed dispatching order to the synchronized listeners +note The idea for the unit tests was to check the resources were generated correctly does not necessarily have to check that they compile to be able to validate that. +Yeah +I think we need a try catch around setting the watch on the next guy in line he may have been cut down between seeing he was next and setting the watch we probably want to check if we are the leader again if an exception is thrown setting the watchTried making a stress test that could catch this but tough window to hit +Thanks for posting. +Again the listStatus is a superset of this +Committed to trunk +Thanks +Cleaning them up as I go even if this patch is not accepted I would like to see the javadocs updated +It should throw an error but there is no error thrown + the log level is already configurable from the admin pages I thought that had never worked +I see the problem here I forgot to include a signal of the condition variable +Attachment has been added with description test fo that illustrates the text and line height m going to attach a revised version of the improvement patch it fixes the calculation errors better handling of preserved linefeedIm going to attach some test files tooRegards Luca +Ideally we could include something like a reference to the keystore gbean +I verified using Carlins DRT and tested against SVN +Im afraid this issue still exists in +just in case we remove the later this issue needs a test to ensure this really works +Its unfortunate that you werent able to verify this bug before beta was released +Thanks to Nikolay Petrov for supplying this patchregardsMartin +So all issues not directly assigned to an AS release are being closed + +This same behavior also occurs with replacetoken +After the behavior has changed +For that project I will make small fix +patch commited +Its your own choice +Ive checked the previous headers from the jira emails and it appears to me that they were addressed directly to +A depends on B +SunilInteresting points here I thinkSome QsHow often do you have to face in practiceI do not get your other tools points care to explain with an example of sorts + and are built as wellMarking as resolved +Marking as closed. +This class has set of boolean flags which control point at which transaction should fail +Thanks Yi Xiao for the patcheNote Test suites run successfully on win xp sp +Attach quickstart +I agree put a message at the bottom of the output which says Now that you watched tokens scroll by you probably want nodetool status +Hope that is not too complicated +Questions should be raised in the FOP user forum +Attaching patch that adds support of reading blocking arraysmaps to pythonPlease review and provide your comments +Any update on this being merged? The impact to an hbase regionserver is pretty substantial when this happens +Its the XML parser thats complaining about your input XMLBTW you might want to reuse the instance instead of recreating it each time if you want to profit from image caching and such +Pull request merged into master fcceAndrea suggests waiting to see how everything goes on master for a while before updating as resolved on if we backport we can just edit the jira and add the new there any reason not to backport it to now? I dont think there have been any issues with since the upgrade a month ago. +Unless someone can reproduce this with or later I believe we can call this fixed +committed +In general from a development perspective it is better to throw exceptions + +Committed to and trunk so resolving. +Trackcontribute for support +it happens on Ubuntu I can put in a fix I will do this in the next few course if Daniel didnt use spaces in paths we wouldnt have a problem The command line scripts thing is a different problem though almost certainly a quoting andor backslash issue in startGroovy +It would be nice to have a small test case though +for renaming the file to target it for deletion +Bulk closing stale resolved issues +Woops had it turned on by asked for a test to show it working +Bryan thanks for the test case +The old solution isnt complete so I cannot apply it +I can try and take a look at this later today tonightregarding bq +I spoke with Pat Hunt today about this issue and it sounds like a good starting point to contribute +Please close issueThank you. +Thanks Virag for the review. +I dont really think we should make this public +You made size the determinant +Hmmh actually reading description more carefully perhaps this is not the reason +Attached patches use geronimo dependencies instead of an inflated +FYI We see the same thing happen when our rails app sends mail as well +With Ravinders patch the fixture passes correctly +What do I need to compile and install to get this issue fixed? Only the plugin? +I added a filter to my mail inbox +Should not have been any progress been made on this +Tools update site has it under the Maven category +I can make the change back to once all the hooks are provided to seemlessly use in the Struts base classesFor example only accepts in its method signatureI think we need to go through the Struts code and deprecate and provide alternate method signaturesIf you want i can lock away some time to do this +I dont have the ability to use SVN and therefore get access to +fixed in trunk. +Should the number of attempts and sleep time be configurableMaybe we should start with reasonable defaults in code and make these configurable if we discover situations where they dont work +libhomeedwardworkspacebin +Fix released in. +This version of Derby has had a lot of changes with interrupts so I am trying to see if those might be contributing +James are you willing to work on making the proposed changes +Doesnt matter to me +I believe it was fixed or they arent having this issue anymore +This issue is fixed in my tests +It also affects +for Uwes fix +I have the TCP or UDP problem resolved +Assign to me if you want me to look into this +And once again this time with the correct file +I know weve avoided it in the past but its a possible fix for the current issue if were going to have limits on the number of os processes +Committed to trunk r and branchx rThanks Robert +is screenshot showing is result of selecting +Built +Fixed in and trunk +Closing this issue +We should not mark an injection point with type Conversation as ambiguous if its project has weld jars in the classpath +The fix version is wrong +the rest of the code is tracked under all SOC issues +If the directory attribute was of type Resource or URL that would be fine but as it just a File it should not be used like a URL +fyi some additional feedback on +Heres a summary of the use cases as Fred and I understand them +or soIn either case this behavior has to be changed and unified in both Failsafe and Surefire to follow includesexcludes +Ive reassigned to myself and Im resolving as duplicate +Thanks for taking the time +Always glad to help if things are so easy +ms que necessitat ganes de dont think it is idiomatic to attempt to gloss plorera with an English noun +We can do something different but I dont know what should be done in that case +Method selectors are gone. +also the cloverclover goal is now failing +containing a simple Outlook Message Parser implementation using POINOTE The test relies on the test message in the previously attached has been checked in on +To a lesser degree +Hi RajithThis is actually a deadlock session lock times out after some time thoughSadly Qpid is full of such fundamental issue +Committed +It should all depend on how you configure the recursing parser you attach to the parse context +After some detailed package analysis most of the codebase packages have been restructured to make the code more modular and reduce circular dependencies +svn rev verify +Hi JacquesI will need to spend some time looking at how the help system works to be able to answer fully + +properties is needed so I modified the test codes to use existing resource file test +but what prevents a leader to repeatedly trying to become leader and to repeatedly fail +This looks good + +But this particular issue seems to be fixed in the trunk +Added a testcase to test the same +You missed it +Add of savingrestoring autocomplete attribute in components require little refactoring uses autocomplete property from ALONG with autoComplete map attribute uses OWN autoComplete property used in mapThis should be unified to use only propertyRegards Zdenek +B +I do see the problem reported but I dont see that tests reproduce it +can you please attach the files +New version of the patch with an extraneous comment removed from the pom file +Whats the general feeling on the JDBC PM should it be included in Jackrabbit coreI would be in favor of including it as the default persistence manager +My screen first text field is with autocomplete +Attaching the Tools maven build log +Reneno your latest commit with patched iterator implementation does not fix this issueSo it seems not to be an iterator related bug +No apologies uploaded two that I failed to upload previously +Thanks Eoghan +Im trying to modify my patch and will upload it soonHowever I dont quite understand your saying expose this and not just metrics +Have also added a unit test for the static method calls +Remember that in some cases a is not enough +project w overwritefalse which is goodIf there are no objections I will commit it tomorrow +The user provided getter is in place correctly +Will let you know if I get stuck +As for implementationI think adding additional columns to existing queries is fine and perhaps preferable where possible because it avoids duplication of queries and should not impact existing applicationsI think changing the results returned for existing columns is probably not OKI think adding new queries is fine too if needed +date fieldName visitDate type notNull pastfield reference fieldName pet type +Oki +Breaking the JDBC URL across two lines is fine but it would probably be good to add a comment that the lines are one continuous URL +If you get errors other than a permgen space problem or you cannot resolve the permgen error by adjusting the maxpermgen then well revisit +So I just left that comment myself +However we need to check if this method is used correctly in we can probably work around it. +I think the selector keys have no hashCode implementation so this effectively uses the object memory address which should be plenty fast +Would be a better place for this setI have no major objectionHowever maybe we want a way to know if something bad happened on a region and a coprocessor on it went away? One could comb logs but that is hardly a convenient way to get online state +Keeping it open for trunk +Am I hearing we should completely migrate to ivy for its dependencies now rather than making a small change to where Hive publishes its internal dependencies +applied thanks! +No to mention fixing bugs you couldnt reproduce +Ill also look to update the slave +This is due to the architecture of SBS and has been an known problem for a long time now +However this could be done in and we could try and include this in Thanks +srctestjavaorgapachewssecuritystaxtestsaml +bumping to since we have no time left for this issue in was fixed I cannot reproduce it +I will apply the patch tomorrow +Regards ThilinaI have just found where is the problem +New patch that fixes the merge issues +To allow Seam to remove their POPIMAP +Reassigning credit +Committed to branch +Nothing particularly but as a patch is available the flag should be setIt allows to filter on issues with patches which should be applied firstI will commit this patch asap +what will we pass it? The location of a table in hdfs? This table could be under or elsewhere +Another option is to consider external coprocessor hosts +Please provide it as soon as possible +Please verify the problem is fully resolved as you expected +Hopefully this will fix the failure in seen on Linux +batch update of issues with fix instead of the correct +in a cygwin xterm I had to invoke groovysh with the terminalnone option +We dont support serving things as applicationxhtmlxml for many reasons and support for this certainly isnt planned for the line let alone +Also if you compile both target files you will also get the correct dir output +What is your release plugin version +just glance real quick at testQueries and you see its a BQ with one MUST and two MUSTNOTS no parsing by the brain required +Let me make it critical +get the plugin to appendmaven resource entries with any bnd file specified properties +dump is a cfml based functionnamed arguments are not supported for java based build in functions +Using listeners for logging seems to be a good answer for the described concern unless anybody objects? +committed thanks! + was uncovered by tablets moving around frequently during case not optimal +Further directives are written directly to the also an updated unit testcase reflecting the actual requirements +I agree with your reading of the C standard +Having said that I also expect some problems with other database engines but leave them to separate issues. +To reproduce try to switch services from the left nav +Thanks for the patch Dhruba +instead of would do the trick +Why are we using to try to accomplish this instead of changing the code directly? That also seems like the wrong thing to do to me +Removing the OS check from the Windows commands +This is the conclusion that was reached for +Trying to start stop the bridge based on the foreign connections state would just lead to out of order message delivery on the Queue +Updating the patch to use temporary files when creating key and token files +Id be going for the documented fix at this stage +Thanks! +Classes be fixed in thanks. +New version of the patch +My suspections are the same it shouldnt be hard to fix +Not fixing something for a beta release that is no longer live +So if I understand you correctly the incoming Query +Cheers +Martijn can you scp the tagged parent files into the Maven repo in the right place to fix this + +needs fixing before anyone on Windows can look to solve the problem. +No bug fix involved +applied in revision Nathan please check that it works as good thanks +It appears that is not handling attributes that are objects in the sessionWhy is that so? Feature? Bug +Attached Patch is from branch to fix this and other problems +Ill let Jenkins do its thing before closing this out +Verified on EAP ERRN text is fine changes are present in and +Quick Dirty patch for Invalid Sate of transacted in trunk. +Ill check that it also happens with the static library tonightAlso this usage scenario is quite realistic + +Thanks BorisPatch applied at SECURITY module at repo revision rPlease check that it was applied as you expected +I can confirm that its fixed +Would you mind to flag the patch as Apache licensed +Attached diff file with +Luke Ive added you to the group in JIRA so that you can assign issues to yourself +Thats ok for me +Resolved the comments issued on Check User Availability functionality in EcommerceProblem occured while adding somebody taking care of this issue +In addition to Juliens comment it also needs header and javadoc +When this patch merge into our cluster the memory usage of tasktracker is more stable +Thus Id like to see a testcase to see what is happening +Committed. +I have committed this +Modified some failing patch containing the latest licence and ve reversed the latest commit on this until we have a fix which doesnt break the blog sample +Unexperienced with Maven we could not determine how to do this +copy these files into dependslibsbuild with ant x +Committed revision . +So it is like described for both HQ and CXF unzip the respective +Will push in a +We will remove those bits preventing to work and move that issue to a later version then +patch looks good +applied to trunk only as there are merge conflicts for +It doesnt occur when creating the triggers on the APP schema +Our goal is to run w the minimum allowed by the jvm +No its fine as is +Get this thing to target INPUTSUPERSTEP and make larger reads possible to capitalize on HDFS block sizes and I will be the biggest fan of this one! The metrics on these jobs I ran with the patch still indicate INPUTSUPERSTEP is chief culprit for memory problems during a run +TESTS derbyall test suite passed on Windows XPJDK with sane classes directory and insane jars +Specially the version is not neededIf your application uses H you can still add it there +So Random may be good enoughThat is different issue though it could be a work around if we dont fix the real issue +We need to make a distinction between and options +That way they are easier for someone to resolve them later +while integrating with i created an action which ports s Lucene query to an action +The scheme gets correctly stripped in so that we end up with the correct absolute path +Nathaniel This should be fixed as of +Please reopen if there are updates +Removed update from on because the partition state machine will read ZK when issuing leader and isr requests +Patch committed to and trunk. +The combined version for Jenkins to test +I can detect your username through the system property +look good no comments. +Column +Change looks great ChuanTwo minor comments +I tested on an instance thats due for an upgrade +The length passed includes the bytes that encode the prefixed length so well need to account for that +Hi Junping I have covered this in +Im now going to spend a bit more time looking at dataSource and dining so that I can understand exactly why the former is being initialised early enough for the latter to be took less time than I thought it key here is which is found via component scanning before +This solves the issues brought up such asundesirable synchronization +What I am proposing is we recognize this and implement two types of triggers +John we will do exactly that deprecate the support in favor of support +Clarified docs. +But now we know where the code is we can look into it +maybe I did something +if we fail over to the SBN does it continue to checkpointIf the only active NN comes back it will start in standby mode and then do checkpoints against the other one +This makes it a simple deterministic computation +s complaint was address by this patch as well +it is attached now +Im installing now +I added an EMS implementation as wellI modified the and to create objects. +Would you mind confirming +Only documents with the same collapse field value that appear more then twice will be omitted from the result +After review historical commits in master this issue appears with this commitBAD COMMITcbecedfaaceeecbfde Support wildcard membership typeOKacadafbdebccaffdbe Remove only what was added in and and reduce warnings on interface has been modifieddropped between and and this can be related with the root cause of the issue. +Pros cons etc +Revisit dependent issues are Tests run Failures Errors Skipped Time elapsed sec +Thats an interesting one however I think we should hold fast to the rule that since is opaque +The new documentation as written only requires that the quota violation be logged +Patch is for trunkQuestion is if it seems ok to backport to +please show it to us. +Any workaround +I dont yet know whether its something special about or whether any malformed file or any XML document at all would be able to exposethe bug +Sorry I messed up in the description field +Fixed in r was to blame as it didnt convert stack via convert protocol. +Seems unimportant to me to force case sensitivity +When the cache is restarted the dependencies are rewired including the you provide a test case to replicate this +I will test next time I have access to solaris +Thanks matteo and ted +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Attachment has been added with description The proposed P open bugs to P pending further review +Looks ok to me without running the server were you having any issues when you use this +the number of redirect hops is often quite small the memory overhead of maintaining an extra is not worth the reduce string equality testing +I followed these instructions navigating through to and then manually entering the URL httplocalhostmemberID in the browser address bar however it successfully displayed the page without raising any exception +Yes Ming thanks for reporting and providing test case +I understand the intention is not to instantiate objects on the web UI but may be we should rather consider making most of the methods fields of static instead of constructing the object +I cant quite recall why it did not apply +that would work in this case but theres plenty where the object might not be called user +This seems fairly reasonable +You need to run mvn clean install on toplevel directory though theres a couple of compile errors due to no longer being thrown +It can ease the transition for users doing expert stuff w Lucene todayI think it only helps if you counted on the reader having every doc id in itIt needs more than the current patch I think we cant rely on people being able to plant a Searcher on the Weight +Their seems to be a bug and we should fix it +Actually it seems to be caused by and and I overlooked inconsistent data +Closing for feel free up open again as necessary. +Nulls can be handled with a for objects with more than fields otherwise a bitmask on an int +Does the BPEL component drag in a different version of Xerces or Xalan in its dependency chain +There are always bundles export same packages and you need take care of how to use it +fails but not +Sorry for the delay I got caught up in another project +Confirmed and exclude lists m not sure why Gurkan is asking only for one test method to be excluded +This patch allows hadoop logging works with +Please t reproduce +Thanks a lot for the feedback +please discuss issues on the forum before opening +Thanks for addressing this +Committed r Please close the Jira +Integration tests added +trivial test patch demonstrating the error +Last comment of Micchael should be sufficient description but just ping me if you need something log store view onto the top level participants the No Release Fix Version +This might mean that we need to be able to set a different proxy server for this connection +Petercould you please upload a sample for us to recreate this issuethanksdims +Deletion only helps until the next try on updateVersionsEclipse Sonar Eclipse Sonar Are there any workarounds available +Making sure all Ps are at least marked Critical to make them standout compared to P and Ps +Maybe some JS errors would show that could help +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Postponed +Had originally taken them as I thought the SWT widget key mappings would work +Attached the process I used for producing this issue +Keith from the gist it appears that this utility just reads the text in from the given resource with the only functionality being stripping out comments +How do you suggest actually retrieving all of the documents between two endpoints based on ordering +Is there any chance this fix could be backported to a release +Sure I can do that +looks good now your server admin is wiseI added your site is in the list it will be in the reportin an hour or soRegardsHenk cool thank you for quick reply and help to fix it Henk +Issue may already be resolved may be invalid or may never be fixed +Yes of the time when I was reproducing it +So I think this is how it should workIn the undertow subsystem we define some config for global session managers +Id say this should be in improvement to log against Maven itself as its maven that passed the config to our At revision The fix removes all new lines from the parsed XML +The sample should work now +I agree with Carl regarding the parser lets move it to ANTLR +Thanks for the clarification Felipe! After I get done banging my head on my desk Ill update my pom. +If this goes in as is as part of I will clean things up so that the thread is not started until aggregation is ready to startI have manually verified that this works on a cluster with the patch +Implicit join across a association is not +Can someone look into this trivial fix +Is this notcovered by any test case +The described workaround does not always solve the problem +However looking at code around bug I suspect other problems with formatter save action prefs +Thanks for reporting this +Looks good to me +Providing test case and screenshots +Seems like it would be a useful feature to be able to configure this impl +and it might even be that thisisnt a huge issue for most default scenariosbut i do think that we should be aware of the weaknesses of our approachand the consequences this might have in cases where data consistencywas really important +lets if needed when testing RC +I have created to track the discussion on the deleteselect with ORs on multiple columns +Moving unresolved from M to M in preparation for M has a fixed time out of Minutes but the JIRA calls for a configurable timeout value +Otherwise the patches that are given are insufficient. +updated priority this is an occasional fail on QUORUM read after a QUORUM write when RF so its pretty serious +Thanks. +H is a single jar full scale DB server +Its not a bug its a featureA feature which was missing in so I didnt realize Now Im happy againo +It has some advantages There is no longer a REGISTRY of Configurations memory management is better +groovy folder on my system +However this will be a major thrust of as we move towards +Assume closed as resolved and released +for having a default set of configs to start with want to take a crack at it +bulk close of resolved issues. +In order to do this I have to let create temp file but it should be OK since close renames file + +Committed trunk revision Merged revision Merged branch revision You can fix this in you local installation by using the latest stable branch if you cant wait for +toolbar action group to be visible in the Java Java Perspective and Hibernate Console perspectives only +The entrylog doesnt prewrite so this is safe +Tested by compiling and running mvn package Pdist Dtar false both sucessfully +things are working fine after I did required changes in my setup. +Verified in ER +The compiler check is invoked when running the tests through the ant junit targets + for getting rid of writing temp file +NikolayI have a couple Maven issues to clean up first before a release +applies on branches +It appears to me was fixed in +Please reopen the Jira if you see this issue again +It works if view selection is carefully monitored but it is kind of tricky +shows that both libstdc +Nice! It seems to be fixed in version for me! Glad it works for you +What information does the number of spills provide? Whats important for performance is how many records are hitting local disk as intermediate data +The error still appears to me +I must agree with Emmanuel +Would be great if you could turn this into a test case to allow easy reproduction of the issue +I am going to run the contrib benchmark tests on this patch vs trunk to see if there is a difference in performance +project key KALUMETfirst set of project admins jbonofre correct component. +Attaching the patch to fix the issue +Will be in be fixed with the RC drop if not please verify by marking CLOSED +I dont plan on doing more work on this users can droprecreate triggers as a workaround to benefit from. +updated +One typical wrong case at the moment is with Oracle which supports pk defined on if the view has the pk the datastore assumes its updatable if not not updatable +Yep thats it +I will take a look what else needs to be changed the bottom line is that if you navigate to the object using the model API then everything seems to work with the patch but when you want to execute an HQL then it is failing and again is looking for the column in the base patch that will fix the HQL to SQL generation and it will include also the fixes from a patch for Hibernate Core s not fixed inKrasimir please attach patched jar +If you can attach your work so far someone may take up from there +The attached patch adds support for unique and in the process for any other eventual aggregate function that has multivalued return +I tried it and it is true +Hi Jeff Thanks for the patch I have already applied it in the trunk +These can be unzipped below root of offline maven repo +The source can then determine how many of the values to batch and control the atomic nature of the insert an autowrap transaction in detect mode will not be needed +see the attache patchWDYT +The library project holds a library of useful JSPvm templatesmacros and other examples for using in Bridges portlets useful velocity macros JSPvm templates that can be used in portlet applications example spring resources example commons validator resource examples of how to use entities in buildsI thought it would be helpful to the user getting startedI could use some documentation indeedDo you want me write up a xdoc in this project itself or are we putting all Bridges xdocs under one roof +Looks like this one has gone stale archiving it. +I do think that JMX should be tried first though +my patch to grant license to ASF for why dont we just do rowCount for now? Well deal with the configurable stuff later +From my point of view looks good. +Why was this assigned away +Need to upgrade Archivas Redback to the latest in trunk or to when released +Sorry the typo in my original request wont of helpedI am requesting that the header be included in the digests for Tomcats users and dev listsAs far as I can tell from the ezmlm manual this requires creatingediting the DIRdigheaders file and specifying the header in this file +show in web browser fixed +Thanks for the quick review +Please review and provide commentssuggestionsLarsRamSorry for delay +Attaching SQL authorization examples for patch +Thanks for logging this issue Dag + MESSAGE data formate doesnt suppose to touch or change the stream contentsIts just reuse the CXF transport +But I have not been able to induce the exception +If no anymore came in I would close it today and cut a release candidate for today +Now the strange thing is that REMOVING the flush in actually gets more output from the test +The duplicates are even not excluded when I declare them in the within for fixing! +Makes sense +For that will be required instead of forbiddenLooks like there are no security or privacy concerns about it anymorecheers Roland +Ill be able to provide a clearer estimate towards the end of the day +yes solr +Andrew your proposal looks good as the standard case +Did you looked into the logs? Ive added some status outputs it is possible that the hosts were just writing the results to disk this takes a while +ant user forgetsdoesnt know about +Hi GianmarcoI think we can stick on which is the major use case +Attached patch simply adds an argument to the Eigencuts algorithm to specify the desired rank of the eigendecompositionStill working with the algorithms author to draw up a method of calculating the desired rank +Now I see it is a completely different file +I will review this and apply the patch autotools the libaxisengine and libaxishttpcommon libraries cannot be built on due to the circular dependency between these libraries +So although it not desireable to go into an infinite loop at any time it is unnecessary to ever Eagerly load a field in a primary key object when that field is actually the owner of the primary key object +Andrey once qe answers the question please add this to be reviewed for +Also introducing counter per namespace is very awesome +Then on failure the parent knows that skipping that range will skip the offending record +Includes Dan Harveys option patches with a couple of further changes +Finally got to close this case using a manually pasted case is completed except for the classloading task that will be rescheduled for the submitterscould you please provide more info as to why the classloading patches in case are neededPlease add any information on that particular case this one is now closed. +I think we should WONTFIX this one +Done in r +A duplicate of. +Second version of the patch that inserts the home mapper in the list before so it is asked later +Added Forum t need redesign just remove the method declaring the dependencies from Remoting class +So what is the status of this? Does this work for emails fired by quartz or only a problem with MDB +Yay bikeshed. +So Id recommend putting in under contrib unless there is an objection +bulk defer to defer to +pushed to cr +err I mean robert to decide what to do here +Can someone merge it to and trunkThanksBlair + +If they arent then then they are not called and thus the proxy infrastructure equality will apply +Was I successful at that? I dont know +It works the other way round +Thanks for the fix Ted! Ive committed this to trunk and. +Assume closed as resolved and released +I cant reproduce this on the Nexus on Master +does a when storing the page at the end of the request and it gets the new session id from there +As I also commented under we cant just change the way the form tag behaves in this regard but there are other improvements we can make to accommodate the request +HairongI am unable to reproduce the problem building either or trunk +running with a HDFS instance creating and deleting a single file very quickly +We should try to resolve this for is in SVN +we upgrade to javamail could you see if it make any difference with +The document seems to be broke +The problem is that Grails has decided not to load the other plugins before it tries to install the new version of the Hibernate plugin +is not +Here is a you Wojtek as I said I have reproduced it and going to investigate it closer +Resolved in +This is essentially a duplicate of and has the same issues as the original version of that patch +I solved this issue +Alexey please verify +I didnt have the problem on my Ubuntu system but I did see a similar problem some months ago on Ubuntu on my coworkers machine + +I will post a comment as soon as everything comes together +Thanks +I see mentions of CMS +Need to checkHere is the updated patch to check for null for cmd before using it +Is that possible +Ill upload a patch soon +Fixed in revision +Previous we just flushed the oldest region though the oldest WAL could have more than just the oldest regions edits in itThis patch isnt enough though because flushing gets held up for long periods of timeOne such reason is hdfs running slow so flush takes a long time +No more work is planned on CDL support if this is considered critical then please. +Aslak you seem to be partial to the eyes that are in and +ftl screen in it +Indeed as expected even with numOwners duplicate hash ids are given to different nodes +Did this work with Wicket or this is completely new code in your application +Is there anyway to pull the values from the ivy config file +Fixed in r +I meant you should open one +This is crucial because as Ive pointed out in various posts one key design goal of LCF is to make the connector developer provide the UI for their connector and your proposal may well force a violation of that principle unless you have something clever up your sleeveThere are still a number of points in your document we have discussed in the past which remain but whose controversy goes unacknowledged +Yes the test failure is unrelated +did you send them to janstey +Committed branch and trunk +Im not sure how we want to track this one +Harsh Im confused +Attaching a new copy that includesonly Vaibhavs work +ok info please +Typically should complete very quickly since it just writes a few data structures in ZK. +If you are able to reproduce the problem with a recent version of Interpretive please reopen the bug report with a simple complete test case. +diff with news for integration of this patch into an official releaseIt works like a charm into my company since many +Nicholas I will try your patch in a couple of hours +Is there a reason Im missing +Just as in other issue waiting on access +Also there were a few ifdefs that checked for freebsd but not darwin when the underlying darwin and freebsd functions were the same +Fixed in SVN head +I will look into it +I didnt see this issue +hard to confirm the fix when we cant recreate this easily +Thanks Sanjay +Which DB are you using Vincent Derby +Any news on this subjectI am having the same be fixed in +Thanks Andrei +Here is a patch in case anyone wants to attempt to build the dashboard +The pdf file was downloaded successful on all clients +cant map file errnocollect ld returned exit statusError generating assembly code for datamv optbinpkgdata name xercescmessages mode static d +you already created ive created for it ive created for itdone for this ticket +XML file is zipped +This might work +I wont do anything on this one until you clear it w Legal and assign it back to me. +please merge +So HT eg +So just need patch for now +There is still an edge case that allows a stale messageAck to get to the broker which is the root of the problem in both Java and +Committed +Attached patch to increment maven artifacts from to to allow a clean build to be performed +Are you talking WFS or WMS? The has been designed to be tolerant to bad data +This test now runs on weme with the resolution +I prefer to let David commit it if its OK for him as he has already commited when you get a chance could you complete by applying these patches? Thanks +Oh is that so then do not worry I will commit the changesThanksDeepal ThanksDeepalThe highest tower is built one brick at a time +The test contains a normal case that typically succeeds because there are few delays in +This patch is against branch so expect it to fail qa botIve added a number of tests to try and verify that the rollover will both work and that subsequently any restarts of various quorum members will return consistent results +Sorry for the late reply +We can fix this in a later version +Implemented in current Remoting APICore +release audit +If Karls patch achieves this then I think we could use it as the basis for doing the evaluation +Fixed WSDL endpoints to handle soap ports correctly + By the way Ive found monitoring await in Ganglia useful and have seen over ms average waits on a busy large cluster having both the host await and a similar dfsmetric would be useful for debugging +Reporters of these issues please verify the fixes and help close these issues +As you know all of us work whenever we have some free time +Patch v looks good +I think the problem is as simple as an incorrect ordering of statements +that said it could be good to have the FD in addition to the other so that if a node goes down for a while that doesnt have much traffic we dont lose the first attempted message once its back up unnecessarily +Even if there is no clean way to ask for it the scema supports what they want so we shouldnt errorI considerd breifly what it would take to make TVC support overrides for the various options the trickiness comes from hte fact that it doesnt iterate over the list of fields itself + +It ensures we can have it working when using and other containersWe need unit tests for as well as can play tricks on youAnd can you provide the patch in smaller pieces +settings that were safe to be renamed +Committed to. +fyi the typical way to deal with this is increasing the memtable flush this what you were thinking of? The patch adds a new which can read columns from multiple in parallell +Please report this issue to WTP project +Many thanks. +I am out of office on a business trip for the next couple of days andwill be slower than usual in responding to emails +Im getting the error randomlyThanksSupun +This issue is part of positive failure. +The other one is for connection closed and in this case we throw exception to stop further processing of repair +OK now all instances of +applied +Tom do you think the sparklines would be useful? I am somewhat reluctant to add a dependency +on of a bulk update all resolutions changed to done please review history to original resolution type +Closing as wont fix. +default invalidate and break its more obvious looking at it what the point is and unnecessary invalidate calls will be harmless +Otherwise we should clearly state that should work but we dont support it +JBPAPPGACP switched to SP CP +might have a common maxim could we get an status up date on this one +HiAlthough TPTP behavior strangely after further investigation GEP still need to add extension points from TPTP to provide this function +This might require a bit of rearchitecting on the client thoughAlso a test for this would be nice +Zhihong yeah I broke this out to a separate issue because I was fixing other problems wanted test to pass but didnt yet spend the effort to figure out the rest of the problemsSince I got your confirmation and since this is a minor test update Im just going to commit +Please address this +I have just tried the wiki and I could edit the page +Added the link to the dev forum but will not merge into main trunk until after MR is released +Fixing both comments and throwing different exceptions for different scenarios in +Trivial patch attached +The simple versions were committed as hadoop issues now umbrellaed by +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content ClebertIve attached file +Committed. +Guys thanks for your comments +I will let you know if I need some help to figure out anything +Thank you very much +we should decide on how were going to maintain gdataAre many folks using Lucenes gdata server? If no one is using it and its not maintained then it should probably be dropped from trunk +ok +Switching to prompt for all input because the annotation doesnt compile with the current SDK installed +Related to Ive done this for now +Heres the patch I used to fix it for that case I hadnt noticed that it was also broken for the and possibly others +Apply using patch updated patch that actually falls back to the thread context class loader. +To elaborate on this a little +Thank you so much! +Hi allcould we close this issue as it seems it is now implemented with lookup screensCheers +Diff for proposed patch +as I understand it is used for testing +I dont see a problem with that +Hit issue links and then the trash can icon +If anything the might have to be modified a bitBasically almost all of the work has been put in to make this happen +Committed w additional comments +Since its a temporary path will remove in the end of that function we can use a relatively path +Commit is aedefefabdabacebffdaa +Some tests are failing though so thats why Ive yet to commit it +LGTM +Check for additional details +Helloany patch for current versionI started learning erlang yesterday for that purpose +Patch that errs on the side of printing nothing rather than print dataI went through where we print out pbs and removed or shortened the output +Note that jenkins d this on the combined patch put up in +Somewhat more subtle than Rs comment it would not be possible to create a file in a directory with zero quota as creation requires that a block of quota is available +There is an arglist needed for the alternate colorspace so that it cant be optional +Closing all issues created in and before which are marked resolved +In fact its not a bug when width of source code is greater than windowI still fix the issue in drilldown pages +Changing your axis instance will involve just the modification of the environment variableRegards there are no further comments on this issue I guess the reasoning Senaka has pointed out is accepted +shift to next GA issues on to CR since GA shouldnt have issues +Hi ClausIm sorry for it +Tomek Issue has not been resolved yet +Im going to attach a new patch for trunk in a moment +Should be Set it to FALSE to enable this new functionality +Fix in r although a similar fix is likely necessary for sysCalls +Im syncing on the buffer +Committed to trunk and branch the new error messages +Also keep in mind that we have a context menu in the browser which already provided all needed functions +Maybe at least a log message that makes it clear what has happened and suggesting that the state needs to be manually updated to allow a restart? The point is that there is no way for the framework to tell the difference between a running job and one that was terminated without prejudice so there has to be an external intervention to get back to a restartable state +The lucene related jars had a dev suffix previously and now the dev was removed and hence my eclipse had jar resolution problems while building + released +made the cache class package private after moving it to common +Lets wait for that patch to land and then triage +After commit the file is opened to write backsubstitute the keywords this is where it fails +Patch applied on trunk and in Developer StudioVersion Build id Build date +There are no longer any customizations around attachments +Changes are stored in the wrapper +Attached CND file with nodes definitions +or at least shouldnt be our problem +Comprehensive reports lead to prompt bug fixes. +But the message of is more descriptive. +The upstream JIRA has been fixed in ejbclient +We could add web hint but then if you arent on the Internet you cant validateAs to schema validation on boot we validate as part of our parsing code and once again during the management conversion doing xsd validation on top of that is slow and doesnt buy much +Im not shure that its a right decision +metadata +Rebased against trunk and passes +You can commit after verifying that tests checks are passing +ViolaI try to define the properties in the pom that actually uses it +Thanks +no feedback given in time for b +or maybe not +I have the cocoon release from november with +looks like this can be committed +The Xalan PMC needs to decide on the fix priority +Marking PA for tests +If no one makes comments I will mark it as cannot reproduce for the next no one said something I mark this bug as not m reopening the issue since the problem is still there +mike patch looks good but are we sure we are not accessing the live SI somewhere down the path in unsynced context +I agree with others that an incubator seems like a better place for it +Hey Michal could you please create a pull request for this +Having it done well before a May code freeze should be no problem +Ihanged the class to use a for the set of attributes using the attribute name as the key +Patch that adds the exportedtrue to the +It works as a controller for validation so apply it across composite components does not cause any problem at all because at the end the validation rules are on the managed beans +Ill run it through Hudson +In fact using transaction synchronization mechanism caused memory leaks on Weblogic +One property missing in this +Ill create an Alfresco Jira for the trailing dot problem as this appears to be and not related to the client library. +fixed in rev issue is not resolved yet the statement still fails with Line Incorrect syntax near XLOCKin MS +Bulk closing stale resolved issues +Conceptually comparator is in the logic of Tuple +The I posted previously doesnt crash but does leak memory +Request is for adding support for setting request request in +Sure I can merge the two patches together and submit one here +Harsh Mind if I take this over +In times when it does we should upgrade the client server side jars consistentlyHanding different versions in production could be tricky hard to test so we should try discourage this where we canHowever being able to check up front whether or not a client or server are compatible sound like a very good thing +I attached to the wrong bug +Im not planning to them further for now so Im marking the issue as resolved +llyas patch worked +I take it this domain is now safely renewed and can now be transferred to the ASF as requested? If so are we ready to tryUpayavira +version updated to in Beta branch and in in JBDS Linux +However must be set to false for connection sharing to work +Link to JBXB issue +The jbpm documentation for SOA does not define a default value for but IMHO it should be too +The does not expose +code is in place but needs a unit test to complete it + has a different behaviour +bulk move from M to like it is implemented in +If it was going right into then I would have reviewed it better +this patch prints the bad service name rather than npe +Im sure my app will be fine without it +Did you mean xsdhexBinary and xsdbaseBinary have a common value space but +Web UI is required to username selection for openID account creation +Im running it in standalone mode so I dont think itS related to HDFSShould we open another JIRA for that and keep this one for the scripts updateWhat will we need to extract from the faulty server? I will start right now to get it fail to stop +So out of support by the time the bug was reported and fixed +Marking patch available since there is one +Please add your changes in diff format to this issue +The total time to balance is halved +Doh there was a couple of mistakes in the last patch uploaded new version +Marked and as abstract in order to communicate the of r +This may not have been updated in the scripts yet +So will be easy to compareI will make those todaySorry was busy with meeting yday +Ashwin just to be sure its MR +Committed to trunk +This implies some changes to the interfaceI will work on that tonight +Committed to master as bac +My intuition is to call it BTW I got some calls yesterday without messages left +This listener then sets the server as StoppedIm affraid theres nothing we can do for nowPerhaps latest Eclipse dependencies will fix this issueIn the meantime Ill resolve this issue as Wont fix +So in general I personally do not think the behavior of browsers should beused as the most common denominator + is very coolBut Im not sure if you could render the stream fast enough as XML in a live systemWould it make sense to have some kind of parallel logging process that only dumps an XML file containing say seconds history prior to the broker dying or receiving some kind of control signalJ +Please feel free to close this issue +Seems rather unintuitive that auto commit settings would apply to commitWithin +Thus its hard to ask OWB for the configuration and better let the user manually add this feature +Finally I realized that the problem resided within our configuration +doneBut usually are able to setup jenkins jobs +This is much better than before thank you +All issues has to be filed against from now +Id like to see snakeyaml adopt this patch +as there are too many folks dependent on itthanks am having problems with the Amazon Google Word Sense and Paypal WSDLs +I dont think this is a bug. +Ive added another check for that and hopefully this is fixed for good now +Ive tried a bunch of other examples all of which propagate constraints correctly +will commit if tests pass +I am doing a simpler check in +Rebased and merged into the master branch and into the branch. +Next time it would be nice if you could send a patch instead offull source +Admittedly the section about qualifiers sort of suggested referring to bean names using Qualifier there +I believe this is fixed now new snaps deploying shortly. +committed +The Groovy method will unwrap the Exception for you automatically +Hey guys thanks a bunch for working on this! It will be great to get Brix up to the latest version of Wicket +Here we go +I would do the later +I wonder why that is different code from Ctrl N wizard +why is the groupId +Ive made those changes +By mistake I didnt upload the patch with implementation +Fixed documentation formatting wording and online help +I see +I reopened here before checking the other one +Could someone please merge it in the source tree +The change Gary suggested does seem to fix the problem +See. +should be renamed to something more appropriate +Also improved unit tests to verify behavior +Ludovic mind creating a completely new Jira issue and attach the patch to it +I can put failed log here if you think it is helpful +Note that the problems in seem to be fixed in trunk but interestingly it is not backported to +Looks good +This patch is working fine for me +are updated +Christian +This is where the current Environmental Variables button is located however I have been discussing having it on the next screenThe button will most likely be placed on the screen following this one as a nested group which will appear as the Advanced buttons groups does on the observed view +Thats what I was looking for +Ive added the code to check for the and the code to check for the existence of Bean Validation +Since Xalan doesnt support the new IDOM you cannot have used the IDOM for the transformation +So Im closing this issue +I added those new collection converters +sh files under scripts and check that we can actually publish bundles on Nexus staging +Hidden fields have no visual representation so they cant be the target of focus +Patch that calls delete if value is null +When I start next node after previous is up and running everything works okI also get strange issues when starting multiple nodes concurrently with indexing turned off +EAR project case is fixed +Where do you place the property files you need for your mailetHow do you load them +There isnt much to be done before releasing +Sorry for earlier comments +Committed thanks for review Cos. +bq Commandline version of importexport can be enhanced to accept job tracker parameter as well for this scenarioI thought you could do this already? IIRC ExportImport subclass hadoop Tool so you can pass +please merge this to trunk as well +Ive not done enough testing as yet but it may be that the common factor in failures is theI still have no time to work on this issue though +Implement conditions +Canceling the patch as it no longer applies to trunk. +bulk defer to defer to +Lets discuss if any of the ones I implemented wont be needed on a regular basis +Ok this is absolutely a total hack job but I wanted to see if it would work taking the package tweaking it slightly to make it and installing it directly in Mahout to makeIt and the associated tests compile but Ive run into a problem of failing tests the cause of which seems to be that it wont write files to HDFS etc +Ill fix thisRe Thanks for finding this +fixed +I dont think there is an easy way to put unit test +And then if I am solve it that way I will have the patch reviewed first and do the changes only on +So I think you forgot to test the category before +Patch has been committed +bq +kk I should be able to do this for without any issues +Im waiting for so long time and nothing happens at this thought that we are the only ones that has such problems with the new Spring DM are still running at the old version of Spring and Spring DM because the solution with the start order of the bundles doesnt work in our situation +I have created a plugin which depends on a jar in a heavyweight application server which has a number of hard coded class path entries in the jars manifest +Hi SandroI started out making the changes in trunk and merged them to branch +mvn package assemblyassembly still fails with the same checkstyle errorCan you tell me which maven version and which jdk version you are using? Ill try with the same versions then +We need those as well +Since maven plugins cannot be executed in this mode anyway perhaps there is no need to support the proper attachments +This would not be needed any more with this patchI did not test the patch but I thought about it this night and I think it is a great patchSome more thought tough it would result in attributeNames containing UTF or ISO +Ivanov suggests inI think its better to use for storing registered because of performance reasonsOn the other hand I guess there wont be many items registered thus making performance gains very low +Seem to be ignoring the new directory for the most part +Note that one session is shared across all endpointsservlets started within the application so altering these settings will affect them all +There was an issue with the cache that I put into the valueOf method so Ive removed it +Ok perfect! I will look into this today still +However the initial unzipping process could be improved. +In any case I think its best to layer the lazy reading on top of the interface instead of below it +v clarifies the error message about the possible use of removetoken force if someone tries to removetoken a node twice +Also I am really worried about the code structure in +Please send questions to dev save tickets for bug reports or feature enhancements +Not easily testable and easily reasonable so no tests +perhaps we could keep the wiki page to give people an opportunity to go add more papers and have a site page for fundamental pages +Looks good +Add the problematic filename text content to one of the filename +Could we make the header contribution smart enough so it could inline the external CSS references for IE? Or create a smart wicket tag which would inline the css if the client is IE otherwise would use the external CSS +Ill let you know if I find any more inference problems +Sounds like this has been resolved +Thanks Robert thanks Chuan Liu +the developer should have idea about xpath to use this +Good +It looks like you may not have your connection configured properly +This patch touches the following filesM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguide patch and the correponding html and pdf output +Committed second patch with revision +Lets wait until its released +Forgotten to mention this +should pass tests +Ill investigate the problem and provide the fixed stream class and expanded unit test hereRegards overflow error is fixed +Hopefully someone will either reopen this and mark that as a dup or just start work on that oneSee Issue you submit a simple test case for it +The is prepared for create a URL and the is passing a value without a protocolI change the to pass a correct value for the +Right now it is a separate list because that is what the project requestedIt requires a vote of the project committers to change that setup +The are pretty big as well +Looks pretty good! Unless someone beats me to it or has concerns about this patch Id like to transition from my earlier mentor role to a committer and take care of committing this +Looks good +resolved in r +We will probably need to implement a fix where when not disclosed only the visible facets are manually processed +Do we think this failure is meaningful or not +I have focused lowering the complexity of enablement of demossamples and thus placed logic into one place. +I have the same problem and looking for a solution Ive found this issue +Ill buy you a beer if anyone notices +Are there any special characteristics of your environment which might influence the triggering here? colleague reproduced it on windows machine +We want one slave to handle deltas and one slave to handle full imports +Also I think someone else should have a look at the issue because I currently have no idea how to fix this +I attached a tar file containing the following README dependencies +I did the test after all nodes got the upgrade +Im going to bring down the server and comment out the Schedule for now +closing +Postponed to the next version +Todd i have addressed all your feedback when you get a chance if you can review the patches that would be great +This issue is hit consistently +Patch for trunk +Go ahead and commit. +Great! Any chance of knowing when it is in a buildrelease on the maven site +Is this hitting you on a real code? Weve known about this as a theoretical problem for a while but hadnt seen it matter in practice +So Ive added these buttons +If you find further issues please create a new jira and link it to this one. +Closed all tickets +bulk close of resolved issues from releases prior to +environments like RHEL Windows +Thanks Jacob +think one of the keys in importing this is that we setup a versioning structure such that we can copy the guide to a new version and then edit that version for a specific release + +Refactored Gregorys patch to be compliant to new cache interface and improved it in ways improved synchronizationlocking added support for capacity limit in addition to the already supported ttl limitThe new implementations arent tested yet in any way but I like Gregory and Werner to review whats available now and add their comment suggestions and so on +If you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +Do you at least have the vdb +Here is a patch that exposes the instance in and makes Solrs specialize accumulation when a instance extends +Cristian s idea above of a +I was having the same problem as Tom +All the rules are now consolidated in the and the tests now pass. +Then I call mvn o site and it fails +Not intended for inclusion with svn srcs +Thanks ArunYour patch is in trunk at r +Ill take a crack at the test case today +I think its fine to just make sure it run against one version and gives desired behavior +It would be great if you could give a more recent snapshot a try as well or higher +It should at least be considered +Attachment has been added with description Sample file showing the issue +closing with the release of +Done +Thanks Alan +But as there are people who care about maxDoc additional bytes in seem to fight against this idea so I would leave this openI would like to fix the issue any comments on the new and implementationclass structure +yes and Im using Spring it havent been fixed there either +Ill be looking into this today +Pull parser implementation in srcguththila and parser lib implementation using Guththila in srcxmlguththila I did not put the build stuff in the CVS +Were having consistency issues elsewhere so do you think it could be because of the lack of sync +Or if we have to do it relative to some kind of SOLRINSTALLDIR or something +So there must be something wrong with the command line args for the run as groovy script launcher +However it may not be very useful since we cannot catch interface as it no longer extends Throwable +The issue remains but we are expecting some genuine code +While making this change issues I found in if confirmed should be addressed in separate jiras get grabs readLock but checks for safemode? commit has no check for permissions against the path +Committed to trunk and branchx. +By using GETPOST correctly you mean that the form submission should be POST? or that we need to have two examples for this +Im not one to be hung up on names but I dont understand why the first thing you are proposing is to rename everything +Please give it a try if theres anything further to refine let us know this one was fast marked as fixed before I could upload the mentioned filesP +However I dont touch the root cause increasing the perm sizeSee newly attached +So please close this issueI am still having trouble with consumption but given it is a weekend I am busy with other stuff +Thanks a lot Wendy! +This looks good to me +Tomcat has good production support at Yahoo whereas Jetty doesntCan you elaborate on what good production support meansIt seems reasonable to have an ant target that packages the hdfsproxy as a +Hi Jacqueshere is the Pie Label Percentage update you requested +Thanks Andreas +The following comments assume the current file per indextable allocation implementation +The above fix I commented was tested against and worked for me just that the issue may be more related to general spring injections life cycle when they are done dynamically within a method + +Maybe this is OK but we need to step back and decide if this is what we want +Problem is solved I just committed an updated tutorial including working source code +Verified by Vladimir. +Added test case that failed before current round of changes +No word from anyone on how this is coming +Well provide this feature when a new dedicated snapshot status will be available to tag such deleted snapshots +Bulk closing stale resolved issues +I guess the hard part is the suggestions for a test are welcome or Ill ask you to review it +Update srccorecontextresourcesschemadocbookwith support for Docbook Someone please create folder and copy Docbook to it +Fixed +Created for +This will cause this case as well as most other rd party integrations to delegate to raw URL handling +See relengbuildersproduct line better in. +Push it in Lewis! Ill fix whatever breaks here revision Would like to say a big thanks to everyone for keeping me right on this one +So probably the correct thing to do in this case is to just remove them form the GUI Web Service and any RPC calls that include them when asking for a partial job +the problem is that that value should not be stored as plain text value in a file that is readable for everyone that can write new File +Ive looked at the code now +is committed in trunk r +Please test with latest commit bcfdadacefecfbabd +See +Now I cannot reproduce the problem +Please verify +Seelmanns recommendation is valid and IMO very important to consider from recovery pov +Well see if that really did solve the problem +done +Fixing the m release process so this does not happen would be highly desirablerwxrwx myrnavl db Apr rwx myrnavl apcvs Apr rwxrwx myrnavl db Apr rwxrwx myrnavl db Apr same for although those have fuzzylogic as owner +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Thanks Andrew for looking at the patch +Thanks. +Would be part of table schema +Razim should be Radim +Keith? Ready for some more to be added +Emmanuel it seems this is fixed in trunk +Manually change the buildsitemanuals to reflect the necessary change +This is by the way exactly what does so perhaps the cleanest way to fix it is to create a wrapper around +The approach it takes is one to change a set of tags but there is no logic or consistency behind what tags are in the set +Closing all tickets +Attach the patch +That way they do not interfere at all with normal method processing but are there when required by groovy +Reference Guvnor Perspective Connection Wizard Repositories View Resource History View Guvnor Preferences Guvnor Repository Connection Preferences Local Guvnor Repository Resource Decoration +r on trunk and added some regression tests for rubyspec failures remain +I believe this is no longer valid. +QE please fine in +Excellent +as per phone meeting thanks Bryan +Removing fix for status +If you find further issues please create a new jira and link it to this one. +But we might also include access as an option so that if folks do have the Enum in hand they can use it to get counter values noTrue +Attached is a patch for this issue for the branch +The fix is simple Ill attach a fix test in this patch +Have you tried specifying IP address in URL instead of host name +I figured that if some sort of web gate is in front of the TB server is must feed TB with the header when requesting +within a singe read there are retries for each blockI have no problem if we want to do it per read +To be effectively used as an embedded ldap applications should be able to pick and chose the optional dependencies +Yes add it to the main solution +Resolved wizards were removed +if it is regression from then yes it should be fixed in too +Or you reuse the old class loader +In Java your file name must be the same as the name of the public class it contains +Arun will this feature have a design doc and test plan +issues related to this will be resolved in though +That way Ill be doing even less work just applying your patches without rewriting their subjects Thanks +we need to make sure that its going to be easy to fully reproduce customers install by using +That way will always the empty string and if ever the string is composed of two double quotes then the result would be +I have some comments on the functional spec +Thanks a lot for the review Todd +This seems like something well need to think through a bit more but probably as a separate issue +Ill start on this and then Frederik can take over to work it out in more depth +Tom I agree with Grant and Ill assume youll update the patchAs for that Ill just commit it as is once the patch is fully ready and then Ill rename classes in a separate commit +Ive added our outstanding subtasks to the ticket +These are a little out of date and the there is a incompatible with current trunk +for replacing with +will commit once the tests succeed +I strongly dislike adding sleep statements to the test +I believe this happened when a cluster was being forcibly shutdown in the middle of log splitting on the master +build absolutely cannot return an empty string in any other context wo wrecking havoc on token processingDaryn after Jitendras explanation are you still +Coming right up +batch transition to closed for remaining resolved bugs +Simple test that take that backMy test was incorrect +Details on the invoker issue +Closing verified in JBT L +This would be very disruptive +We are a search server first not a lib +Ill make that change and add the reference at the same timeIs it specifically the INDEX for which you are looking +I cannot access ibatis trunkTherefore I applied the patch toAnd my test cases work wellThanks! +There are only three and one of them is already committed +Currently decided to give the patch for this JIRA along with was covered in +Lets move this to as there is some API changes needed +Perdona Linda per lentrada s solitari ria +is a little more into the available JDK process for spinning up and waiting on child processes sounds very reasonable +Thanks Nicholas +So there is a correlation on the limit threshold and max +Pushing this back an iteration because and will definitely keep me busy this iteration +Hi AreekMaybe I wasnt clear with how we can avoid the user to put in the type of the fieldName required for sort when using I have uploaded a patch where providing type is not required +Posting a little early in the process of running full test good +Reassign to a release that you feel comfortable resolving this issue in or leave as is +targetprotoorgapachehadoopyarnproto +Hi Carlo this issue has been around for quite some while and still is messing up QA testing +not sure how you want to handle it +Finally as I said on Friday Ill dig into Robs code asap +testCounters looks like a unit test and only emits log messages +Seems to be only in master +Feels more like a dtest than a unit test to me +The change is really minimal in this case +These changes resulted to normal reaction for in my environment +is this issue resolved now +We no longer support the branches +Logged In YES useridYou get the same failure if you simply in an EJBs classpath +Thanks Suresh +The hadoop tar build spends most of its time building docsWith my fix for in place it takes me just under a minute to build the docstarget +Just to upload a full can confirm this is fixed in the that was just released +Thanks LukaszIll try this patch today +This has been done for EAP +Plus branch is dead. +You are correct +I ran my local test again and caught it too +The Criteria API is largely considered deprecatedlegacy so we wont be adding new features to it. +Definitely not otherwise there is no way to shut down if clients stay s easy to test run cassandra and in other terminal session connect to it using CLI and try Cassandra server without closing CLI +Fixed by fixing +Lets track it there +in the long text of VM options find false and change it to trueAfter this the console works fine but looks weird due to the ansi control characters getting printed +I would absolutely love this +Adding scripts to bin is not responsible for hanging tests. +said +Effective Java is primarily aimed at library code +I think I get it now in we have made two fields dont care and we update the epoch value +Small patch +Is it possible to translate a hierarchy of to at once +Please test if that is working for youCheers +Retargeting to fix in since has been shortened and made more focused +bq +Closed upon release of. +All it takes is to associate the code format with the project in Eclipse and apply it +Adds in a section for highlighting +Reopening issue since the patch is not gcc on cygwin doesnt defines the WIN macroSo I suppose that we can replace everywhere MSCVER to WIN where we suppose to use Microsoft libc +Attachments the Eclipse project used to reproduce this problem java dump after the deadlock occured java console of the test program after the deadlock cannot reproduce this problem with Sun JDK +In current ESB runtime definition framework require the user specify a location where the ESB runtime installed the location can point to the following three types of ESB runtime distribution +this does not seem to be related to the update to as Im getting the test error on the current plugin code from svn which still uses pmd and the tests worked on my machine prior to being looks like its something wrong with but Ill look at it today and try to see whats ve submitted a fix to that corrects the problem and the update to should now work at least it works on my machine +This is about changing the design of and and its too big of a change for right +If this approach is acceptable probably should add this to the interface +That test fails because does not handle concurrent conflicting commit detection correctly anymore with the changes +It looks like this was resolved in although we do not know exactly which change resolved the issue. +This is fixed in SVN for trunk and for branches +fixedWould you please try it again? append your whole workspace the serverclient codeThanks very fault I thought that the enclosing element set the default namespace +I have test it with EAP and it works for me +Thanks for the review Stack. +Option Proof of due to rewrite of serializer part of this particular problem should be solved +Which history are we talking about here? previously hit commands using arrow keys or any thing else +Is that another permission +However looking at the code we must make sure values in filter attribute value assertions must be normalized +dont have time to try it but it looks like changing to on lines and of cameltrunkcomponentssrcmainjavaorgapachecamelcomponentjdbcwould do the trick +Assign to release vehicle +HTMLs table syntax does not apply to doc book +incorporates improvements in the parascale unit tests +Have you tried doing it from the command line? Sometimes the complains if it was not created with svnFrom root directory svn diff This will create a patch file you should be able to upload without problem +Do you think it can make the release? Thats be great if soAnd Maurice thanks again for your fantastic support to resolve this +I wonder if we should rename that to add to fit the convention +no sure of component setting services as a starting point for failure to start error +Fixed this one +Note that if were streaming wo first writing temporary file locally there is less reason to keep all outbound streaming but Streaming contexts currently assume there can only be one outstanding stream per source node +Lars beat me to it but I also want to clarify that the osgiheaders command that Lars mentioned accomplishes what you want for listing exports of a bundle +Ive also included some documentation on the Bank Reconciliation processThanks SharanYour patch is in trunk at r +The parent POM can mandate that are created for all child projects +Please could you check that they have been applied as you expect +I was hoping others would express an opinion on this discussion but oh wellThe first and most important distinction I want to make is that the data model services and even user interface elements are NOT hierarchical in nature they are more of a graph because that is how business processes and data are in the real world +Duplicate of +Fixed in trunk and thanks to Thomas all resolvedfixed issues of already released versions of Roller. +Siarhei If that is still an issue would you raise a separate JIRA issue specifically for that and describe the details thereThanks. +The discussion on this is going on in the mailing list +I will open a JIRA and attach a patch +Thanks everyone for the reviews and helpful comments and the attention to detail. +Do you do something similar at the HDFS level +Yes +Having that information in the exceptions isnt very useful so I omitted them +Thanks Tom for the docs on APIs +Verified that change propagated to site +I just committed this to trunk and +But then what is the point of back porting the JBOSSGA with the fix as mentioned in Bruno Haleblians commentBruno did not backport the fix to +Committed to branch. +But that is not so interesting to implement as its actually similiar to directly programming on message producers which is already implemented +the libs are in libhttpunit and the test in in the source folder src is still a problem with java script +It requires that all cluster users have accounts on all slave nodes in the cluster +Version has been releasedMoving all related issues to the next version. +If you marked this as blocker because it has fixVersion then please move it out to the next release +The group has administrative access julien is the only member so far +It has clustering features +Added a that bootstraps the +Placing the tests somewhere else is weirdYou cant do this it will create a circular dependency between core and +Hi Wei Nan has this been done already +we must remove blueprints setting for shell otherwize while running shell will be into empty object +Please post in the forums if this is still causing you problems +I tested on +Closing per Bryants approval to do so. +In MSDN documentation I found that before calling any GDI functions it is necessary to call function +fix with from primitive to primitive after removing boxing +We need this one working since is pretty popular +Another Jira MAPREDUCE deals with similar issue where map output lost is tolerated but it has the benefit of not wasting resources in the node during the whole rolling restart process +I took a cursory look at the patch without applying it and as far as I could tell everything looked all right +But I fear I have no time now +For unexpected zk exception we should abort the master instead of logging errorIt does +Committed in r on trunk +Add classifiers for each hadoop versions to help with multiple hadoop version compatibility +Change color to Red and submit query +All junit tests pass fine +The Sonar User mailing list must be used instead +The patches have been applied as expected +ClausAs a friendly suggestion please refrain from qualifying other peoples work +Hi SimonAs mentioned in comments for we support suppression only for all violations +In any case at this point that looks like the only plausible explanation to me +I agreedbut the existence of ledger doesnt mean the metadata is formatted +Thanks for picking this up Chris +is a solution using a char array +I have been able to solve the problem by implementing a +ftl file if the file exists? Or direct me to a link where it explains how? Remember that I could be using the same +Returning back to the original single write scenario +Ill take another look and at least add the comment if I cant figure it out +Marked as later +This is not a defect +Is that in fact a new directory that needs to be added or am I misunderstanding something +ChangesCreate a mvn profile called visualize to generate graphviz state graph for andAdd maven tips to add two missing source files + +Ill confirm and cancel this defect if my hunch is correct +I laughed so much about this patch! Only is missing in it +This patch contains both comments added to the various test cases and changing asserts which use literals like to use GB instead so that the units are not as easily confused by think documentation in code is enough for explaining GB I dont think this patch is required may be if we just add documentation is enough +Description updated for example project updated for Seam issue due to recent +Totally unmaintainable +New Version + has been marked as a duplicate of this bug +I can force an early load of Struct which should help it +Allow me to clarify my positionI was prepared to accept this for +Please if still a problem. +Now that work is well underway with AS all previous community releases are +average docfreq across all terms even maybe the or so is sufficientThat could work +A new task can be created if more work is needed +Its not the temp file generation that failed its the directory creation +Hello Colm +The patch for the existing files +Please reopen if it does notCheers Kurt i am aware that these properties can be set in +So this patch is more or less preventing us from deploying speculative execution as the default for our clusters +Identity equality may result in multiple entries for a given schema but the cache should still work correctly +I sure would appreciate your review once I have a patch +put the screenshot on the wrong jira not sure how to remove it +I think should be resolved first for that. +I wanted to know one thing whether we have rights to change the code under package? Anyway this question unrelated to this JIRA but curious to know the fact +In the new mapreduce we still have to introduce the new input format as we did here + seems reasonable to me +As discussed by email the parent class already handles the session opening and the exceptions +The only reason this test was passing was because of the bug in + compilation target does not echo anything about the number offiles printed Upon inserting a bug and setting proceedfalse the compilationproceeds instead of failingI have fixed this in addition to a small change documented belowa Removing the quiet optionb Setting failonerror depending on the value of the proceed variablec Removing the echo task since if the exec task fails we would want the user to know the reason for failure which otherwise wouldnt be printed if echo occurred after the exec taskd I am also removing the part of the build which checks os and setsexecutable to ant or and am instead using vmlauncher attributeof the exec task and am setting it to falsethanx Narayanan +This Patch packs management server feature set to stratos committed in faedddbefebaa +I have studied description of +raising priority +We now support this kind of implementation + for as it involves quite a number of changes +The patch has passed all build tests +is probably spending most of its CPU doing its work inefficiently mostly in generateSentence and Textencode +I think that the problem is more critical see key doesnt work with many other components and in addition in different periods of time VPE chooses different components to be deleted +What updates are you referring to +Give me a bit more time to defeat myselfbq +Need to add the new dependency to libcore directory +ok hours sounds good +Downgrading from blocker since this seems to be the consensus here +Also how come you externalised the variable +So since you are decreasing the heap this makes sense +Spun thanks for the patch +Thanks Sharad +Bug has been marked as a duplicate of this bug +Simple code example attached Jody +Sending srcmainjavaorgapacheservicemixhttpendpointsSending srcmainjavaorgapacheservicemixhttpprocessorsTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixhttpendpointsSending srcmainjavaorgapacheservicemixhttpprocessorsTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixhttpprocessorsTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixhttpendpointsSending srcmainjavaorgapacheservicemixhttpprocessorsTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixhttpendpointsSending srcmainjavaorgapacheservicemixhttpprocessorsTransmitting file dataCommitted revision . +Can a committer please take a look +Attachment has been added with description Simons Simonthanks for your additional commentsOn your first point regarding trailing spaces just to be clear are you saying there shouldnt be trailing spaces according to the specification or is this your opinion? Either way you may be right I just want to understand this problem from a specification point of view +and no release is in sightI tend to disagree with the above assessmentNow that Gary has nailed I would say the shape of is getting clearer in terms of featuresWe just need to tackle the outstanding blockers and critical bugs +All these segfaults today are really killing the post election buzzThe error seems to be in the APRs global pool destroy not sure what is going on in there I know it creates its own mutex to try and protect the pool from corruption +I analyzed the use of threadlocal and its clearly not needed anymore as flow control credits are dispatched OOB and INTERNAL and are thus using their own thread pool and arent getting queued up behind other messages in UNICAST therefore removed it will also remove it from a release youll need to create your own JAR for now as has issue and I dont want to release it with only a single issue +I am on linux bit debian I have created new copy of eclipse and new workspace and install builded version of CM eclipse plugin into it and all is okBut your listingosgi ss cayenneFramework is launched +Confirmed this issue no longer occurs. +Bumping to CR since this wont be done for to so we consider this for the release or a future bumped to for now +Because of that reason I changed in that way that no modifications of existing Hibernate code is necessary +take it this wont make for +The issue is reproducible starting from threads that makes production use of Jackrabbit impossible +This isnt going to happen +In VFS we need the first version where nul will be returned in case of listFiles didnt return something which indicates that this was afile and not a directory and thus this file had null children. +Earlier patch doesnt have the code change for boolean variable to keep track of whether the weights are fixed or variableAttaching a new patch with that changePlease review and provide your comments +Both test plans look good and complimentary +Note that we shouldnt be able to beat LZMA because we neither compress values nor do we compress the overhead +Ive checked that the proxy configuration is correct too +Sorry I missed that one it is greatbtw +Class that implements but delegates to a +looks good +A conversion of NL to CR NL was done twice resulting in NL being converted to CR CR NL +Hi BenThe output from development mode is +Once we confirm that it behaves as expecteddesired we will make it the default behaviorWhen I make that change you will need to test that hostname verification passes for Hive JDBCDoes that make sense +Looking forward + +the embedder doesnt seem to use the file by default +Hardy I would like assign this to myself if you have not start it +OK I opened +Okay +Hmm I think the logging option is better if people provide the right configuration you have what you need and even more +the workspace argument is not relevant IMO as in a givenjcr workspace the versionable item is clearly identifiable and its perfectly legal to have different permissionsin different workspacessince the current situation where access to the version store can only be turned onoff in globo poses a security problem i have the impression binding version store permissions to the permissions of the versionable nodewould be favorable +Trunk version of the Lars change to the test to and to trunk +Please add some explanatory usecase or description so as we know what we should verify +fwiw Carl actually implemented a reworking of the test execution to work around the binhadoop from the tarball issue and thats part of this patchThis doesnt change anything about building shims rather it changes how the classpath for building the shims is assembled +I wonder if we should refactor the view engines use of external OS processes so that view generation can intermingle with other types of requests +That also means that there arent any bug fixes being planned +Further we cant add a dependency +I think it does a good job of explaining the problem and the solution +so that portion should be ok +Gleb when you understand what and where to write reassign this task to our juniors with explanation +I think I need to fix the small hole here +Thanks Mithun +But even in that case I think an exception should be thrown rather than accept an empty field name and silently provide zero search resultsI am still in favor of text as the hardwired default +But do you want tomerge the files? Or should I close the bug +but If you dont provide us persistent on the cloud +StefanI have created a small unit test for and I doesnt notice any incompatibility in with this regexp +I had an offline discussion with Karam about this issue as mentioned in his comment this indeed happens for a short duration between heartbeats when the task slots are actually getting assigned for a particular job +I have amended the commit +interop tests in whitemesaround +And its very big +Can you also look into it as jobtracker will get fixed by this patchRegards SenthilFor introducing the guava dependency is an incompatible change checkout which should work as wellThanksEli +created for the titleProperty of screen +Excluding resources is not useful +I attached patch with test case +Thanks for reply ChrisI tried with property but still the heartbettime is going to update every min +Actually my last comment was wrong +this applies cleanly but fails to build against current SVN +We should change the description to reflect this +Attaching the patch used to partition the hash tables +what do you mean by JSSE issue with JDK +What I have so far +Initial patch +Need to generate a version for but the one uploaded is ready to be reviewed for me +Other impls should be verified. +dslr when you start using as that will allow the IDE to correctly recognize and work with your rules file +I just committed this to trunk and branch +Ive applied your patch into SVN trunkThanks a lot for your contribution! +Please post a patch if you want it to be fixedThanks +Verified on EAP ERorgapachehttpcomponentsproject present in maven repo apachecommons is present too +Sergey im curious how are compactions of the stripes being scheduledqueued? Does a region still make a single compaction request and the compactor picks a single stripe? Or can multiple stripes be in the compaction queue at once? Given that regions could be allowed to grow much larger with stripe compaction enabled it would probably be good to allow multiple stripes to compact in parallel +Partition keys are also useful metadata information for users browsing views In a strict mode where a predicate on a partition key is required for any partitioned table a user has to know what column the predicate should be applied which then will be translated to base tables partition columns +Given the narrow scope of these classes I might be leaning this way +Hum +Ok how about this then +I agree +This patch does not implement checksum on HDFS files right? Do you plan to generate for HDFS files too? For HDFS does it make sense to create a checksum from the blk +Works with tests provided. +Verified in ER +studioeclipseplugins +for the patch +Steven J +This is a system not a user forum. +just keep in mind that SNORQL is now a webjar +this is a duplicate of which weve taken care of + check existence before cygpath +I have scripts I used for previous version to create scripts so I can build the project has moved so the group name may need to change from version one that uses +Its a problem with the Modify operation adding times a value make the growing up to Mb +Username djencks +Thanks Wing Yew! pendings Jenkins +the problem by copying and to the endorsed directory +that way round it actually makes sense +uploaded trunk patch for review and upload a patch for +I did some testing around replication nodes on same lan heavy replicationheavy indexing and did not see any sockets in CLOSEWAIT state after running it for about hour +Yonik may have a final word on + Username stephanenicoll +I have attached a video showing strange behaviour after disabling a node using status manager +The rest of the stacktrace is as in the attachmentCaused by cannot be cast to happens with Dispatcher +I just committed this +bulk edit of issues that for some reason were automatically assigned to me +Matt can you see if you can turn this into a failing system test +This may not be the clearest place to check this though +Thanks. +Added refactored patch file +I will report back in a few days time on how the application holds for your review and help +Ping +If you want to take a whack at an fallocate patch I can run some benchmarks +Thats what odd +Step is necessary because we have a different tokenization pipeline in C +However you said you need something so that we can finish the setup +Added another patch + +Fixed in r . +Moving to the issue to its version was updated to +If we dont pass these parameter values to the reader it will ignore the user choices. +You dont need to have a failed mapper but can insert the same row twice +However it obviously goes very much with this issue if you do a little back reading you dont want to have to reopen a reader each time for realtime get you want to pull from the transaction log when you can instead +allowing arbitrary file reads from the server would be a security risk +Main focus is now on RS which runs inside Switchyard so DSP no longer required. +Closing resolved issues for already released versions +Just committed this +Incorrect Hex values in the caused tests to fail in patch in r +I dont think it to be possible to change interfaces at this stage +We also tried DBCC shrink commands +Build a new from +Attached patch +As far as I understand the testcase from Clovis this is currently not the case as only cache is changed but not the database +Thanks! +Closing committed as svn . +It is generaly public but the module with the impl is privateThe slfj javadocs come together wed have to do a rpm fr on the orgslfjimpl but that requires recompiling slfj +bq +attaching proposed patchThis should solve this particular occurrence of the issueI am afraid this problem might be a bit spread all over the place a is used +And adding some dummy messages doesnt work here because the wsdl file is generated by javawsdl with extraclasses set +Bump should we get this into now +Ill create a separate issue for that +then I expect well hit this +why on earth would it match anything you have not even touched the Apache install +Seems fine now lets reopen if we see it again + +If we had defined at the host level this might wellsolve the problem +Marking as resolved +Let me know if you continue to have permissions problems when committing. +Graham I seem to recall some discussion about this on the dev list + +We can diff the output of the intellij formatting with the output of the eclipse formatting as a test +I dont have access to the original Adobe data right now. +Could you create a separate issue for this +Does that fact belong in documentation for the Reflect APIYes +If there is no need to keep different sets of annotations apart for different code generation targets then it is perhaps way simpler to just add the necessary to Property and Association and add keywords to the concrete xtext syntax +But until that comes I am thinking this is a nice side effect to have +Attaching the patch with the test case +Otherwise it would be silly for it to rely on old files lying in a workspace holderGert V is the one who have access to the box is now all resolved tickets from or older +Updating the latest patchThanks discussed We will do the changes as part ofThanks +This change is just kind of +OK thenanyway seems worth mentioning in the docs +Moving a fix for this into since thats when we can really iron out replication issues +FWIW you think my comments are angry well its more frustration that I have to keep repeating myself +BenjaminIve pushed a new set of changes to github with the error code refactored as you request +Now were doing that everything works as expectedo +Example app and sourcecode +Please notice that you also have a error in your example +I welcome feedback +actually not so harmful can +It is nontrivial to estimate which change is tolerable and which not +Patch looks good +This patch will add the missing features of codegen tools which are available in the command line toolSending srcmainjavaorgapacheaxistoolcodegenSending srcmainjavaorgapacheaxistoolcodegeneclipseSending srcmainjavaorgapacheaxistoolcodegeneclipseuiAdding srcmainjavaorgapacheaxistoolcodegeneclipseuiTransmitting file dataCommitted revision +Diff file and original chinese simplified xml file as diff file might be correct +pushed +DestroyingI wonder that there exists problem around startingshutting down in the test suites +The logic was to create a new pool only if it didnt exist +You missed to close the single quote on the exception message + to trunk. +In your humble opinion should this patch be applied? Afaiac +Ive moved the sorting so that it happens in itself and it now covers a lot more of the edge cases +As you guys pointed out the current patch still needs a lot of improvement and measurement in performance +just a you send me the page please +See linked jira +This will happen as a natural consequence of getting the conjugate gradient stuff into trunk +It was a bad choice in implementation but easily path involving is a full walk of the tree and unfortunately we picked instead of +Will try to find outIs this issue the same as +Also how easy is this to reproduce? I added some more robust tests to the replication test at one point to try and tickle this but had no initial luck +Heh this is becoming quite a novel +Each row key is a different index and the columns within that index are composed of different composite types +Testing turned up interesting issues like needing to tickle lease renewal between lease expiration and while waiting on log split to complete but complete testing not possible at this stage because resultant cluster churn with new servers coming on line and loads being rebalanced provokes the outstanding issues in +Good idea Anil I need to do this for our internal people in any case +Closing now for impending release +And I never saw the article I really just read it in the posts on the dev list how it was recommended that plugin developers use for addressing this +Screen shot of current pool stats in +Looking at the code a bit I think part of the problem here is that Idle timeout seems to be considered even when there are references to the which would go against the stated purpose of which is to close out Connections that have sat unused +The problem is that JDT assumes that the file name and path is also the name of class name +Great idea thanks for the patch +Dunno so referred to insteadThe other aspect of this issue is that the navigation needs improvement in general +yes I tried ur sample and it runs smoothly sry I cant reproduce it eitherthx anyway +Sorry about that I committed the changes only from within the server subdirectory +Id like to see a of the mailet configuration +I can make a new one if you give detailed instructions +Please check commit notifications +But thats not a debate related to this issue in particular so lets leave that discussing to some other venue. +Nicholas patch on did not cleanly apply on +This JIRA should really be closed by now +We need this for i think as per the changes in +because of all changes to the backend the natural speed in Search has improved in the mass indexing area and following the advice in the book and reference docs will get you acceptable timingsThere are still lots of optimizations to do but I am beginning to think that most API changes I had in mind are not needed anymoreIm currently evaluating the option of adding more Lucene backends could be selected by configuration and offer slightly different strategies to improve performancebut sharing most code with the current one in trunk +I hope I didnt miss any +on this patch for being technically correct but it actually does not help +Implemented changes discussed with JVS +Need to create generic patches for both tomcat and dependencies dont seem correct to me +ok i may have after all I am a little confused about the description at firstI will try the script in and see if it solves the problemthanks Stack! try deletinghbase deleteall +Also attempting to adjust values on object this in postPersist throws a runtime exception which essentially states that the state of the transaction is marked as rollbackLet us know if you need a stack trace for that scenario +Im just trying to understand what the reason is for creating a list and looping through that list when you could access the map that you already haveThanks and I hope you dont take my question the wrong way +Of course it wont be in the logLet me look at the +Would a introduction of errorKeys besides errorKey introduce too muchhastle? I dont see a reason one form field would not be abe to handle more thanone error type? +There hasnt been any objections to my comment about this issue being a duplicate of so Im setting resolution to Duplicate. +Sorry I overlooked your patch +sounds reasonable checked the code should be easy to fix +If communication to the jobtracker is subsequently restored while the job is still in progress the jobtracker can easily mark the lost and found tasks as succeeded +Package untagged revert to avoid accidents if the prod team has to fix something in and rebuild +At some point it could overload users with options +This is not my case but there were some other SQL queries which I wished HQL would support too +Im tickets +bq +This requires changing the agent protocol syntax so its a feature at least +MAXPRIORTY is and is set to square when the longitude first hint is set Access based one is set to Hsql based on is set to Generic threaded one is set to So it seems that setting to anything above would do the trick but still its funny how the priority layout makes most of the range between and unusable +Where are they +Hairong is uploading a new patch merged with the latest trunk and a small bug fix +Right patch this time +I only ever saw it under one extremely specific of circumstances that I dont even recall +push to +I set and I can deploy the bean from an exploded EAR from EARejb +If they are not specified entire envelope will be used as the source and the targetNew attribute called endpointKey is introducedSerialization test is also included in the attached patch +So was this actually a bug or notCan we close it +I think its safe to attempt an update of serde prop irrespective of whether the user passes in the tablePropIf the table or partition record is using the Haivvreo and has a property defined then I agree that its probably safe to update this field even if the user didnt specify it +from jboss souce do I have to use if I want to do some thing similar on the am generating the server dynamically through code and the requirement is to validate it against the whole bunch of xsds +Zip file containing test classes and test output log +Committed to trunk +Thanks +I just noticed that you can set the launch configuration booleanAttribute key to false in the launch configurationin order to prevent Eclipse from appending its own environment to theenvironment of the launched process +r xr trunkThanks for your patch Michael +One more thing we may have to take care is in the purge logic can you please check thisAny reason why it has not been marked as submit patch for to run +three sounds very specific to certain widgets +submitting patch +For now though we should remove our KI on this topic or at least provide better guidance on a reasonable size setting. +bq On test +Once again thanks for the prompt response AND the information! It makes sense you have a forward index of grid cells to documents no geo comparisons required at run time +I wanted to check whether there was some code hidden in core Java somewhere that did this or whether we were forced to do it by hand as per patch +Must be that test framework does not initialize same way as the real thing +Patch included except for the is method +We have to go back to version +Merged into the branch and into the master branch +t get confirmation as to the necessity of the DOCUMENTUM environment variable so I went ahead and made the change as if it was still required. +thrift specification as long as it is valid haskellBut then again I dont know what the other code generators does +The object returned by snapshot doesnt necessarily need to be threadsafe +I dont understand this issue isnt scm update the only way we know if there has been a change in the remote repositoryOr do you mean when there are two build definitions using the same working directory where doing scm up once will only trigger one of the build definitions? If so the resolution to this issue is IMO to give each build definition its own working directory +This version is then mixed with version from other dependencies +Otherwise I will do it myself in the near future and submit a patch +Im just thinking that a workaround would consist of setting focus to a with display that is actually hidden behind a without display +Heres all the work done for the clone able solution +Attaching a test reproducing the fixes the Julien +Plus anything with a suffix on it I expect mileagekilometreage to vary and we work out problems as they arise +In order to make this work in there would need to be a considerable amount of work done +Verified in ER +Added the patch to the for your work +A version event is prepended to all history filesHistory viewer and History Parser have been cleaned up and duplication of code in the jsp files and has been removedHistory files are named username +Ted can you take a look at this +Wouldnt it be better if implements iterator interface directly if there should be only iterator expected on +bamconsole needs to be pruned +It imports a WSDL that causes Ode to throw a +Patch attached +I think we should address this for +The latest patch looks good to me +Could you also please post the rule which is being injected the contents of the script file if you are using a annotation or the annotation field values if you are using a annotationAndrew that doesnt seem to tally with the error message possible trigger for rule Throw in on nd write in class Byteman says it is injecting code into a class called +Need to run tests to make sure nothing is affected by this change +the priority as this needs to get into +Easy to repro by reverting the fix +Thanks +We still have to test on a Xoom +Duplicate see this is really a clirr issue +And third this kind of thinking is the same that is drove people away from JEE Lets give the API a lot of power for advanced stuff but make it hard to do the common simple stuff +This may have some side effects I have not caught yet +API stays the same and if sum is available you also get it in the +But the script has a variable defined for widerowcnt which implies otherwise +patch otherwise hudson will not be able to detect the file type +The list FS nodes were not properly mapped to the feature structure that the list FS was a feature of during deserialization and hence when the list FS was modfied the encompassing FS was not serialized during delta XMI serialization +Thanks Chris for addressing this +The method was implemented +It could definitely unity a fair bit of code +Minor fix to Job tracker it passed only client protocol class when registering the rpc service + for the patch +You will find out what I am saying +There may be one further one based on and that will be it +The Annotation Editor selected a block of text beging at the index of lowest start index and ending at the index of the highest end indexI assume you would like to have single text selections for multiple annotation text areas right +looks the same respect issue Parent page should be modified through a +I have two more tasks to this issue +Ruud thanks again for your precious help. +thanks Roman Ive tested the patch and everything is working as expectedWhile reviewing the patch Ive seen a few things it should be fixed in the scripts but that is another JIRA +Looks good to me. +Retry just in to and trunk +Normally a method call from inside the closure should not affect the closure +File isnt part of the default studio files though so it shouldnt matter. +The build problem is being caused by poms containing workaround for a bug that was fixedarchivaarchivaBoth should remove the excludedGroups parameters since they are basically workarounds for a bug that has been fixed +There are several things about this patch I dont like yet +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I added this check to The rfc also report this text while discussing the ending boundaryNOTE TO IMPLEMENTORS Boundary string comparisons must compare theboundary value with the beginning of each candidate line +Hi BryanAll problem solved +Thank you for applying the patch DrewIve tested this and it works. +Also to use it Tomcat must be installed and Tomcat is used for +Agree to close it. +It can return null to indicate it is unset but it cannot return a NULL object nor does it get a default to pass in traversal calls +The main beauty of maven is that it just works for most cases without having to tell anyone how to get it to work +Running tests nowKathey +Fixing this is more than a little tricky +Fixed in r for trunk +While we are cleaning up and redoing this section of the code it will make things clearer to get rid of the code pertaining to ancient layout versions +Added Justin as a watcher +This is fixed in trunk as of r Ill hold off on fixing it in the branch until has been released +I vote to undo the patch +Closing after release of Jackrabbit Server bundle +This could result in deprecating our Function metamodel and ultimately retiring itWell be rearchitecting the model extension framework in so wed maybe just have to create a standard Function Model extension thats always applied to a relational model +Thanks for creating this ticket +The portlet definition in the file will always be changes are committed to I need to create a new update site for you? +Thanks ill verify that +Do others agree +Latest patch seems good +But Teds right this should be considered a bug +The above patch allows a to accept an outputMode +This looks really uglyI would prefer to configure that on the and leave it out of the URLAnd there is also a destinationResolver you can use to lookup in JNDI etc +Patch applied to trunk I hooked up the new params to wfs code as well +This is actually a problem in the tags in general +Thanks for your analysis +Hans can you tell me what youre doing in the Axis service to generate the fault +Hope the new version addresses the issue Spring Form Tag generates a null id when not specified and duplicates the name attribute regardless as +Maybe they changed this behaviorAnother undocumented change +All passed +After discussing the problem with Rob Davies in the IRC channel Im reopening this issue to rollback the synchronization for the readCheckwriteCheck +Logged In YES useridTo run the test case and the client stuff will have to be in theCLASSPATH then typejava youll get the usual output stuff +Apply afterRender judgment on whether its really OK to cut off checking the queries after +Hold off on this patch unfortunately +It only has to be a sentence or two but it is confusing as it is +I thought the parser and formatter would be created per message and pick up the new line parserformatter on that occasion +incase if this test needs any special permissions +Also before we add it to the code base we will need to assign copyright to geotools PMC +Thanks you very much for that information +Rado does that ring a bell? I have a vague feeling it wasnt ever part of any actual release though just something someone could hack in somehow +The HTTP server updates now mean that at least one side of the connection described in the original issue will not close the connection is accidentally established +heh on fixing them all +Max I suppose you ment Sergey Vasilyev? Not Sergey Smirnov +Since the constructors at class is already in use we can not remove them +I changed character to +Alexey please verify +Committed backport to +removed localhost link in section Service Endpoints expanded section Usage Scenarios +Cheers For Your Best Help +remove from cloudstack put it in some websit +Thx Tony +what weve been trying to get you to understand the difference between session expiry and destroying the session +Attachment has been added with description sample to the cant see the problem in the test FO you attached +If isnt being set then lets just go back to using digheaders +Glad to see this bug + +I guess that is an issue for OASIS +What do people thing about the attached patch? It tries to always return valid transfer syntax and not to destroy information at the cost of not always returning a float typePatch looks good +As the form says its Only to be filled in by the assignee before starting work +Bug has been marked as a duplicate of this bug +Rebase the patch +Youll have to tell me how to switch to XML +Test to reproduce the problem +Can you rebase +I saw this was committed +Done +Fixed in +Does this preserve the intent of? My read is that if two threads require the same file at the same moment one of them is going to think its a circular require +I am guessing this may result in changes in query plan in other +Excellent suggestion +Its too bad the sysinfo tools complains about missing DB client jar file and driver class when testing for the network client +via the assembly plugin +I also updated the and files in the schema directory to add ingres these changes successfully starts and runs for me with one exception +patch attached to fix +Committed a bunch of work for this +Our processes are very short processes and we dont need hydrationdehydration feature for it +What if a job gets unlucky and its th attempt gets killed because it happened to run on a node where a rouge task of some other job +If I activate only a I got the issueIf I activate only the Producer I dont get the issueThe only way to recover Ive found is to delete the store directory +Reopening as Later is not a good part of JIRA workflow +Assigning to Dan and setting fix version to to Vemund since that is who contributed the patch. +This looks done +I still have to verify it but I think Ruby Iconv behaves different in mac and linux +Closing resolved issues. +By eliminating step carrot could also be fed by term vectors with crazy Highlighters +Committed +Just tried but Maven cannot find itHas it been published +Can you please add a banner to Apache License header +I will post an update shortly +Let me move this issue out of +And it runs well +Third times a charm +Still cant confirm that it works but well look into it +The point was that nasty Bug that I just reported +Is it possible to visualize EVENT relationships between Components through information added in annotationThis is a useful task which Id prefer to be done in a separate JIRA issueAgain thanks for doing this +how do i give signal to the framework that my first tasks were accomplished so that i can start the next one and do you mean by spawning A and B launching both tasks at the same time? +Both patches are the same FYI +Working for me too with EAP CR and WFK I need to send Pull Requests to Missing artifactjar this seems connected to workspace specificsWorkaround create a new workspaceCorrupted workspace provided to Fred for research +Create a jira in Tools and assign it to fbricon then let him know about it in irc +Sorry I should have mentioned this issue when I committed my recent change +The same should be true now theyve moved to the type system +Closing all resolved tickets from or older +transfer The return value boolean is not being used +Closing resolvedwont fix issues. +It only tries to assign or unassign based on ENABLINGDISABLING but does not set the state of the table +Thanks Boris +I was reviewing the component code and wanted my approach to be clear as to whether we were doing a load of system properties using a location system +Please give some advices and Ill see if I canfix the issue thank you +VerifiedThanks a lot! +Its about to pass Futon tests as well +It does not support storing complex datatypes as +Lars thanks a lot for the reworkGuys I could be of more help as soon we get this into the repoIm looking forward to seeing this component on Camel! we have to thank you for donating the code to us RegardsLars +Feel free to change the throw to anything you prefer that will be easily recognized during testing +Ill attach the patch +Yes it always false in the inializtion phase +Otherwise I would like to close this Issue +At present this link is broken +It would be nice if an example could be added so people can see how to use it in practice +I just committed this +Alexey please consider if we need it for +Good catch +Err + +If we are getting new connections each and every time then this issue is not getting new connections each time in a while loop is a performance there are many JDBC calls in our code it will be a very huge change if we are planning to change it in such a way that it takes new connection every hibernate has a property as afterstatement do we have any other setting in JDBC to indicate that every time a new connection has to be created so that we can overcome this issue for time +Excellent patch I might actually be able to read my own code again Thanks Tigran! +Can you be more specific about what is not correct +try should not hardwire any algorithm but rather use the preferred algorithm of the filesystems involved +Thanks Julian and Deepa for verifying the patch +Any chance we could have this asap? futher help on reproducing this problem so marking as a FUTURE for now +I think we can close this I think this is now fixed. +This all seems pretty involved +I check and I see problem with start two database Failed to start database matrixhudson +Patch copies the from +My point is that using the full URL by default is in contradiction with the HTTP specs. +Fixed in revision +The problem most like comes up because the pickList is missing a ConverterIll update the forum thread with a suggestion to try nesting a converter in the pickList component +Updated the patch with a with no values set but explains what all the user can AshishUpdated the patchI think we also might need to revisit these package files +Or to do it just for auth to make the test not cause that error in the logs +I think adding encodings adds more complexity than its worth but I dont care that much +We should have one anyway for both c and java +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +We need to work out the process for getting this information from the individual projects and putting it into the release notes +Working as designed +I need to see what actually our data contains? I will do some research on that +I will do everything possible to stop this +I think that its not the real problem since we can share our opinions and then if necessary make a vote +I think doing it manually makes sense as we can control when the resolution happens +I have never had this problem in on Windows or Linux but I am still running version or on those platforms + +Thanks for the review SureshNote not needed in branch. +This patch also adds an hdfs temp path to the test harness so we dont have to worry about whether the bloom filters are already existing in HDFS +Please Andreas send an email to the Sonar Development mailing list for such issue +Lets make the new behavior official I can also think of scenarios where this behavior is useful. +Trev to create the JBAPP JIRA and +I wasnt really intending the source code I submitted to be a patch per se +The client should use Basic Authentication +Given that I am not sure if the jira or the title makes sense +If the RM is not having any support for scenarios then we can leave that out for later when such support does arise +Never tried though and dont really know what code changes would be needed besides perhaps instructing the datastore factory to use direct connect somehow +I think we should redo the parser and switch to the porcelain option finally git blame pBut that will take some time +Ive long been meaning to add find but havent had the cycles +Doh missed that one +You are proposing alternatives yes but to the detriment of the benefits which the system is intended to bring +Im going to commit this in a moment and create the JIRA you asked for +I saw this with IE and +OK that makes sense Simon +But how can the implementation detect which version of the struct was passed? A common technique is to have a version or size member at the front which has this information but that might be too ugly for usAnother completely different alternative is to do the socket thing and have separate open and accept calls with the ability to set options via other calls between the two +I will start looking at this sometime next patch for code review +Thats what I have noticed too +Embedded doesnt support JDK automatically yet +I didnt get a chance to test it today just uploading it so that it can be reviewed +Its ready for review you rebase to latest trunk please +I notice you fixed this bug as blocker so I guess this issue may be urgent for your applicaiton +expand +Max says to assign this to you +This is mostly to avoid when iterating over values returned from such methods +Here is the draft of the design proposal for the shared cache +Addendum for debugging +Should I retest on the most current snapshot or should we just close this out +deprecated the method for and removed in trunk +A junit test case that can reproduce the usecase changes that make it work included in and batch reduced to in java code message to kick. +Committed the fix to the release branch. +Would be good if we had an example showing how to configure JAAS CXF in the ESB for example something like the test case attached to ESB Security Guide is a tutorial style introduction to security in ESB +Otherwise I wont bother +associates page locks to a single thread +Its been a while since I looked at Lucene but their sandbox approach was in the back of my mind +fixedmanually checked with are reporting beharior in release +ThanksIt may be worthwhile and not too hard to try and do something that works well enough for the common case with one liners +Why do you assert on locking heartbeats instead of sychrnonizing on it +Yeah that test is failing because the user defined path srcjava does not exist +Forget about this patch the one attached to also solves this issue by different the problem is also in it forces tiling in the image while removing the alpha bandWill open an issue in in closing all resolved issue that have not gotten any more feedback or comment in the last month +All other log messages in the go to +Begs the question do we even need anymore +This issue may slip to next depending on when release is +Scheduling for fixed in SVN +I just added an formqualified to the addressDetail element within the complex type def for and I dont get compilation errors any more +The release note looks good to me +on either v orAny opinions? v makes life easy for do provides for a more general API in the future that we may or may not use +OK +So once the master sees let it change the state so that znode version changes and let RS work on itStackLarsWhat do you think of the suggestion way could be make the RS create the znode in SPLITTING state +Max could you please reviewupdateapply patch cant run queries outside an app container and then assume it can grab the containers transaction configuration will have to be changed to work in that scenario. +Gregory I think is only a workaround +a while back. +In that case if I add a Map to a namespace you wont be able to iterate over the keys and find out ones which have a +No problems no errorsI have svn compatible patch files if anyone needs them +I tried this with a couple of hundred regions +Tested patch with existing cluster good for initial review +Ive investigated the problem on with jars +thanks Ivan +The error makes no sense +And line +Patch committed in. +Is it right for you if I create a version file where I will put the changes made in the component and add it to the zipremark in this case we will avoid to pollute too much this jira ticket +The code will be committed after is released +This sounds reasonable though Id have a comment to add to what you describe as the current behaviour for runaway scripts +Changed the Graph API to improve the methods of getting blocks of children including adding a new technique for getting the n children that follow a sibling +I am attaching for those who are interested +Ill resolve this once is in +This belongs on the users list +According to the log tunneled connectiongets closed after each requestresponse exchange +Well the other issue is quite explicit remove the jars so it wasnt obvious that was what you were asking about +Perhaps we could at least get agreement that this is how it ought to be and I can change the RI and any test accordingly +Ive pushed the change out to all users +I have tested on Ubuntu and it is fixed +I think Ive found the problem + +adding a release note +In every case original issue was resolved +Might end up just closing all of these as Out of date +Thanks Gabriel right you are +I mean if you use tomcat for example and listening on it is not necessary to setyour ports on itHope it works for you +Please verify +Testing Debian now + dont think using of template for configuring components is a good idea +If we really want to have support for any ant path syntax Id rather go with an antpath attribute or something like thatBesides this everything sounds great +I just kicked off hudson again to reverify the failing tests +Add a new directory with qltestexcludedClasses which contains classes that should not be in the test classpath +and are both passing on my machine locally +Alan Yup doing that right to trunk andThanks! +As the tutorial is currently part of the this build it needs this restriction +But there were no Operation Timed out were found in the logs during this bug +At least the stack trace you show me there isnt corresponding to the current line numbers of the trunkBut then as I was redoing some test to check it was actually working on the trunk I got other kind of issues like the some strange unresolved dependencies +I think we are done with nightly builds. +The idea behind the patch is to introduce an type to handle cases where the URL shouldnt be interpreted and basically change into an abstract super class for both this new type and which used to be the old + Check listPrepare in kit and online release notesUpdate dependency table entries +Update the sample code in the file +and derbynetclientmats passed +Im still not happy with the fact that the field contains http when I open the link editor +Hi teamCould you please let me know if there is a plan to add GSS API based Kerberos Authentication to TCPSTOMP connectors anytime soon +The failing tests should pass after changes +Hi thanks for tackling this! you beat me to getting to the testsYes we should remove this stuff +Integrated to TRUNKThanks for the patch JimmyThanks for the review Lars and Stack. +This issue appears to be fixed in the branch +Lets finish the integration before release and end of this year + +scratch that comment +So I think its due to the host name mappingeg +The equals method should be independent of the backend codegenerator +see also see also Bug get better error messages at least provide some background info as per +What do you think +You only need to send me a note and I can go ahead +In addition it breaks compatibility in the DB within releases +Galder what kind of thread information would you be looking for? Getting a jstack every minute or so +They are not public as we have not made the protocol itself public we may do that at some point and at that time we would make them public +I didnt port the hbase code in my latest patch +Attaching and with table descriptions for the URL attribute and Derby limitations topics +Is this correct or did I get you wrong? After all Im no native speaker +Check out bugs and +Ok I have it reproduced +Deferring to NMR to to +NET Ill try more tomorrowSecond Ive attached a first draft to the interop guide +trunk on Camel in rev Cancelling future tasks is now done using a running boolean instead of a keeping a big array list with the future references +Are you saying that you would instantiate an objectof the same type as the return value of the getter for the property and thencall the setter using that object as the argument? What if the return type isan interfaceabstract class +Can you please the patch from latest sandbox area? SureshI have attached conflict resolved patchThanks Shameera +I think that the time has come to eliminate this compatibility code +next time +Good morning Suresh yes I have signed and emailed icla +Resubmiting the patch by removing samples folderThis patch changes the doc target and calls it as docs Thanks Nigel for mentioning that +Unless I get objections Im going to zap the lot tomorrow +Ive just removed it here for now +xsl and that worked for me but it sounds like others need recursion so that wont work +See my comments above +Grid I would not worry about it for now +Lars not yet but trunk is evolving such that we can move code sections implemented as reflection into version specific implementations guarded by Maven profiles + +I agree that it is the easiest and securest way to get rid of that problem +The MEASUREMENTDATAIDTIME index no longer exists +Resolving per Harshs comment +More especially trace wil be added soon +Resolved in acf +Thats not what Id expect +I think the current snapshot has it +It works now +stop +So which jsp pages are you referring to + patch looks good +Jacek thanks for the report +Raising priority to Major because were hitting this a lot more now that we allow rolling upgrades and a corps internal Hadoop release cycle may not necessarily be correlated to the release cycle +Thanks for your feedback +Tried to import some data with jsonsstable to compare sizes but seems like jsonsstable is not entirely working yet +Am I missing something? I see no patch available for this issue. +Please also remove and which are unused as well +Is this something we want to keep in core +Since this just reads from the database and doesnt actually persist anything I didnt have to go to the extra effort of using ordinals or ids in the enums +So what has changed is that all messages are internationalized and all exceptions contain a valid SQL State +What is the error on your classpath tab +I have committed this +Ive attached the code as I have it so far +Do you have a PDF file with these glyph names so that we can verify that the integration works +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +As there will be no more new features on the jbpm tooling and as per Max Andersens comments it is not a bug but rather a new feature I will resolve the issue as Wont Fix +There is another simple optimization that might be considered as the sourcepaths are looking for a given file once one is found we can break out of the loop since there isnt a good reason to look across the rest of the sourcepaths +committed +Merged to doc branch at revision . +Reorganized +Committed trunk and +From what I see on the dev list this change did not make any problemsJust remove the prerequisites from other than the parent +But I strongly suspect it is already fixed in TRUNK can you give it a try +My bad I was wrong about the filtering arguments I went through only part of the code. +YuckAh argh +BTW why thrift does not use a well established unit testing framework +Carlo changed the version and left the nexus version be used + +One other data point Ive been able to gather since filing this bug is that we tried a deploy with Maven and specifically downgraded to and we still had this issue +is an updated version of the testsThe tests are not yet enabled inCommitted patch b to trunk with revision +All patches have been applied and the resulting build succeeded tonight +Is it not the same issue as in +Use class for setting select fixed revision There is padding in default class +Thanks +on the effort Sheesh +attached miss you here +Also we should point to the file where it is actually needed to be configured +Okay so that makes an even stronger point to align these because currently Im seeing places where stereotypes are not honored +So all issues not directly assigned to an AS release are being closed +with the latest development build I got this errorUnhandled event loop exceptionXPCOM error believe this is out of date now that we use xulrunner please check JBT M or older. +NG datastore does not suffer from the reported issue anymore +Put me as the owner so I wont oversee this any longer and review it ASAP +Petras thanks for spotting this +Fix for this issue +sure it was with? cause I remember having fixed that a while ago +Do you have more information about thatSee the Escape analysis section in httpdeveloperworksjavalibrary +Reopen because it seems that there is still some memory leak as seen be ukasz +I have run all of the unit tests against a build containing this patch without any problems +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Another to adjust is sizes of caches +How worried should I beAre you keeping your promise to ship the patch in some form withIf you are really worried about the browsers Id be willing to investigate your irony is that s broken negotiation is causing us to have issues with patch is a complete sollution ready to use +This issue needs futher investigation but i will change the status to Resolved for now +Now i am trying to make it work with the latest trunk +addressing this limitation is required for I have an alternative implementation strategy which allows an easier backed implementation +I wouldnt expect you to write tests for example as a programmer +Now we know that it cant really work that way +Ancient bug report against Crimson +The AS supports the JPA spec will try to update the to the xml namespace spec +emptyData seems more clear in what is happening in the produceRequest object Vs clear +We have geohash already tiling is a different hope you guys can make it to my session at at which Ill describe my geohash prefix filtering technique +I think the issue was that the migration tool was trying to process all the libs in the EAR in addition to the application code? Im not sure +We now atempt to bind in a loop up to seconds before failingfix in trunk rev +Documented as a known issue in the EAP and release notes +Thanks Dag! Well directory and file name can mean both absolute and relative so I hope this new patch works all right +So all issues not directly assigned to an AS release are being closed +Thanks for the detailed review Thomas + are now invoked during the processing of EAR to discover persistence units located in the EAR library directory +The first which tests override enabled passes +I know what you mean +Can I ask what the use case is? The token endpoint is supposed to be the only handler for a token request +I think its better to move this issue on GA +Added dependencies and wrote POM file +I decided to start it back up +This patch includes the previous +Okay this patch fixes remaining threading issue inand three tests that were expecting CFS segments and werent getting onesdue to flush now respecting no and no default of your test looks good +Closing for now if its a continued problem please reopen +Thanks Jon and Dave for the review +In Criteria tests we had used a different scheme where a single EMF instance is used for all test cases that are using the same set of domain entities +This fix appears to have been merged to the branch in time to get onto the next release train so I am marking this as fixed in +Can anyone guide me how to install the? I has downloaded it but dont know how to installI use cygwin for window and my jdk isThanks very much +Ill second Rolands analysis +slfj seems a significant increase in complexity +Is it possible for you to check if that helps you +Did someone else than Anil was able to use WTP editor in Eclipse and view documentation? JacquesIts working for me Im using Eclipse and it started working after I removed +I guess providing some common adapters will satisfy of the needs of course its more xml to dont know yet +please find the patch of improvementI put some check to handle the error coming on current implementationThanks Awdesh Parihar +Converted to a separate issue rather than a of as I think is close to complete +I applied your docs many thanks! +note that id Type is a valid approach +Do you have anymore specific information about your object structure? What I really need to know is the exact return type from both of those methods as well as what the values returned in this instance actually were +Now I will have to read the spec like a we can leave this open I do find the WCS Request Builder very helpful. +Not really a concern as that constructor is only used in unit tests +It hasnt happened for me for a while +This is how DN handles assigning default generation number initially +This was fixed by +This problem seems to be related to the that was used to redeploy the siteGenerating the site in trunk for Fluido Skin works fine and has the correct footer +Used attached pom +The steps would be Unroll package Run bin startThen you would log into the app where it would ask the relevant setup questions such as entering the license key backend database mail server initial users great example of this process can be seen with the JIRA and Confluence installers from the meantime though we are still planning on creating a text installer for Windows that has the JRE and Postgres binaries included +Trivial patch but lets have it verified +The Target Versions field is the one generally used to indicate what version you want the patch to go into. +moving to beta +I currently have the snapshot surefire installed in my local repository +were you moving nodes around at all +no problem +Ignore my last +Not making a positive claim just sounding caution +Offhand I think that logic should have worked +Works if streaming gets out commented in test +Removed Branch as its a dead branch +Uploading repro script +the developer will have to do this manually by copy code from Wicket the developer will duplicate the details once where the original header item is created and second time where the replacement header item is declared +Perhaps the solution is to always depend on the contenttype bundle +Patch applies from inside of langjava is that the new intended behavior for patches? If so we should update the How to Contribute wiki page +If it causes too much of a lookup performance hit the Builder could just build in order too +Gary My gut feeling for this issue is that seems we have the version number issue again? like using int but not long +I removed the offending assertions in +bq +After many attempts +I suggest populating svn with the current version +Needs reviewing and wont make it into issue is not going to be part of Nutch which is likely to be released very soon +Thanks +snapshotbinstartvnc +Where does the output show? In the test output? Good stuff +What is it about this problem that leads you to believe it is a Struts issueI have a monitor class that listens to events on the network +I will leave it up to the reporter to close the issues + +Sigh I guess you are rightHowever i just checked the filename is around characters not including the jobname so please get rid of the truncation too in +My bad was setting even for cases +duplicated issue +The thing is that those archetypes are not in the correct version +If you dont want to have trouble dont use plugin apache snapshots you using mvn can you try with upgrading to or +Renamed to now extends +I cannot reproduce it +It may actually be best to close this as WONTFIX because BOP has been removed from +This will go in along with the related protobuf changes +Patch to change filename to from thanks. +Hi DavidCould you provide a Pull Request with this change? We can then make it available in Forge which is due in August RegardsGeorge upstream Thank you! +Sorry I am not sure what issue you have reproduced this one or can reproduce in JBDS B but it has been fixed of these issues have been fixed in JBT master with the latest jpa +Valid point if you ask me +Holger if you believe this patch solves the issue please close it + resolvedSo now as I understand it we will read splits concurrently in each worker task and the ZK split list will only be read once from ZK and then propagated directly to all the workers? Was there a speed increase in the split reading phaseAlso what did you mean about compute threads this patch seems to deal with the input split phase only +Files to reproduce the could reproduce it with release +How did they set the isolation levelI cant seem to reply on the forum even though Im logged they attach a small test program that shows the bug +Should they be read only for Edit action? Like it is now +I might have jumped the gun there with my previous comment +Looks like a duplicate no +Please check if this is still the case with the EAP reopen Fix ER if still present +I updated the cache wiki page with the new endpoint options. +Im definitely not arguing that it wouldnt be useful only that the scope is large enough that its not an task +I broke my own install oops. +Mike fixed the missing merges! Thanks. +If and when you do come across a scenario that would benefit from the async call I think it would be more efficient to use the approach in in whether or not to make available to a pageIf youre itching to use dojo for the checkout screens I would think the wizzard widget would be a much cleaner and user friendly route to takeDo you think it would be possible for you to outline what actual issue each file that is changed is in regards to? This would make it a lot faster to get through +works for bothPlease commit the patch to both and +Koen That sounds good +Mattdoes this issue still exist with newer version of Castor? In other words will the very same test case fail against Castor? +Sorry but thats too much information in one go +What do you think of +Shouldnt the invalidation take place on all threads in all instances of before returning from an write operation in? It seems as if just setting the cache to empty would allow a period of time where write methods had returned but not all threads have seen the mutation yet because they are still holding onto the old clone of TM +This fix did not make it into the beta branchIt would be nice to apply it for the beta release +It also is more pedantic about interpreting the results of a call to parseColumn +Committed to trunk +op is only acquired in initiate and do which are only run in the context of the application thread so this deadlock cannot happen as described +We cant check the status of bundles immediately after the call to refreshPackages returns because refreshing of bundles happens in a separate thread +Agreed lets just remove the custom implementation and maintain the binary compatibilityThanks +does not include a overwrite test forAlso for I got errors in input join and join input is ok in my test +Im on having this on by default FID that do not conform to the expected format wont return any valid result in the luckiest case and will break the pk column extractor in the worst +This bug is most probably the reason why the option true is still ignored in the brokers file +what is the ZK node in question? Also how do you test this best on local machine +With the sub tasks being done I still see stack overflows when running the integration tests with NITPARAMCREATEABSOLUTEURI false +If a program or the OS requests additional memory the OS will happily and instantly give up the lowest priority cache data to satisfy the memory requestYour Solr admin page seems to be locked up while trying to load the dashboard so I cant see the actual numbers +This appears to already be possible just not documented +Thanks for reporting the bug +Reads much better +So they dont need any javascriptAlthough I think it is very stupid not to change the classes when the user checks or unchecks a option I will delete the javascript part of my patch and submit another one +Reopening to make the described improvement that ensures consistent field numbers +They were very industrious +Oleg I just picked up the changes to +The test is still incomplete due to Mikhail and Stepan +committed +Unfortunately I cannot dedicate myself to such an endeavor +Add service and implementation bundle Add service login methods to and Add implementations of new the temporarily removed changes to the Jackrabbit Server bundle in Rev +bq +That shouldnt collide with existing URIPrefixes but I put the value into the restricted prefixesuris just in case +Tested using aforementioned projects and setup. +If this doesnt handle what you were thinking let me know and Ill. +A Maven project should never use files outside of the projects directory +Suppress warning +For the deadlock case please open a separate issue as thats something different and its difficult to keep track of work when so many things are lumped into one issue +It is probably fixed with the current snapshot of the SCM project but theres still no good feedback. +Oops +GavinIs there a a design issue with providing the with the? Again I think the interceptor should be allowed access to all that information just as the hasI dont see how I can use the to my advantage +Now makes sense +The code compiles and all junits pass with Java you for your contribution +Ronald Id just like to get an idea on how to proceed with this? Iow do you have an opinion with regards to what I have said? Or can I just go ahead as Id like it? +However in this case almost all the mappers request the same block at the same time so their retries will be also at about the same time +This work will be dependent on the Locator to LOB mapping that is part of +I would call the new invoiceType Payroll instead of Employee Salary to cover hourly and salaried staff +Do we still need to get this fancy now that we have +All tests pass. +The problem here looks like the end location of the testFoo method in the JDT model is incorrect +these could undoubtedly use more cleanup but they do exist and more or less work. +ant producer this will consume messages at a time +Set time Estimate and Severity +Updated release note +Joakim is fixed can this also be closed +Ive stumbled into this problem with the release and just started to explore my options +If there is no redirect to another page then there is no need to bind the session right +New unitary test +As per Manik this shouldnt be an issue anymore given the locking fixes in this release +Injected parameters of producer methods observer methods and constructors should also be validated +Thanks guys. +Sven i see that the ticket is still open is that intentional? If this is the final decision on this matter im going to ask dashorst to add this one to wicketstuff htmlvalidators ignore known wicket bugs filter +Committed into trunk rev +along with all of the other testers. +We can only do this if we add experimental to the random access detetion method +Ok pretty sure are sorted with builtins last. +Marking all current RESOLUTION LATER issues to be against the Nightly Build +See +The patch looks good +This issue doesnt show up as a feature request because subtasks dont allow it +Heres a slightly modified version +Fixed with rev am closing the issue +My updates are related to the issue but they didnt solved it +Ok I will revert my change of this morning and we can work on this when we get an answer from the jsf open listAnd the problem may be in Primefaces but I guess it is due to Mojarra and behaving in a different way for the same code +According to that youll need at least version of the assembly plugin to preserve the permissions properly. +Dan +the other disturbing bit about this code is that it is not outputting any info about the actual being thrown which will make it rather hard to debug other consumer issues +There should beDescriptionEdit +Can we apply this to trunk yet? Chris +I will do so in the future +Then for example we wouldnt have to worry about Point or all of that +I guess somebody complained about address being an IP instead of hostname but for us this change breaks our reliance of it being an IP instead of hostname +Great work and thanks for your fast responsesEnjoy your holidayI added an NPE guard there +investigate with Nexus indexer to take care is karma on repos +did not mean to assign this +good catch +Were not incorrectly failing to find the Literal constant +annotations? The Maven CDI Configurator adds the CDI facet if the project has the war packaging and includes the dependency +Clearly in Seam it is being when the same class file is loaded but two different classloaders how can you tell it came from the same file location? Given the abstract at which the Seam deployment scanner works Im not sure it is simple to figure this out +Apologies for not cleaning out this dependency before. +Committed to and trunk +If there is another please tell me about it. +Its doing pretty well with the small number of tests Ive been trying and I think its at a point where some of you might want to take a look at it +To and at leastIf therere no objections Ill do thatFrom your analysis it looks like k would be the sweet spot but since you ran it in production with an k window lets commit this +The simple solution is to just not shade bookkeeper in the bkjm jar +It seemed like synapse community is too lazy to response or doesnt care to find a solution +Thank you +This patch simply checks is before writing to the log +Please review the amended text and confirm whether or not its for the inconvenience to yourself +Move the fix to +I need more time for this on review and we want to do a M soon so pushing this out for MRC +No it should not be +Fixed in SVN +There is another expected unit test failure which is only on MSSQL certified are oracle i g mssql postgresql mysql +What to do for the case where a shop has chosen other than supervisord to monitor their processes? I suppose we could let them do the convertion from supervise to god etcI dont think we should worry about it for the first pass +to enabled Flex remoteing that sets this up for you? Of course itll also then need to update it if anything changes in the core +Commit changes to support absolute URL form in the annotation. +This is slated for +But you can see that this does not happen after hours +Ill do and +Have you tried to delete your browser cacheChromeBut I expected that the messages appear automatically so maybe I was just too impatient +HiI deploy all required snapshot artifacts so the build should work now with +For large files Id want to consider writing the view in ruby though which has a C mdsha moduleSven Its relevant to that ticket but I think youre touching on a much bigger subject there that of revisions for individual attachments which is an issue in its own right and beyond the scope of this one +Closing resolved Release +One of the regions of table B is actively written to and is split +Thanks Ted +That v patch lgtm on principle from reading the patch but itll need to be rebased +This doesnt contain any transitive dependency +Winzip logExtracting file CUsersAdministratorSourcefusecdctargetdist to CUsersAdministratorSourcefusecdctargetdistUse Path yes Overlay Files noExtracting bin bin bin bin bin bin bin datatmp deploy etcactivemq etc etcclustersstatsdefault etc etc etcdefault etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc etc examplescbr examplescbr examplescbrsrcmainresourcesblueprint examplescbrsrctestdata examplescbrsrctestdata examplescbrsrctestdata examplescbrsrctestdata examplescbrsrctestdata exampleseip exampleseip exampleseipsrcmainjavaorgfusesourceexampleseip exampleseipsrcmainresourcesblueprint exampleseipsrctestdata exampleseipsrctestjava exampleserrors exampleserrors exampleserrorssrcmainjavaorgfusesourceexampleserrors exampleserrorssrcmainjavaorgfusesourceexampleserrors exampleserrorssrcmainresourcesblueprint exampleserrorssrctestdata exampleserrorssrctestdata exampleserrorssrctestdata exampleserrorssrctestdata exampleserrorssrctestdata exampleserrorssrctestjavaorgfusesourceexampleserrors exampleserrorssrctestresources examples examples examplesrest examplesrest examplesrestsrcmainjavaorgfusesourceexamplescxfjaxrsclient examplesrestsrcmainjavaorgfusesourceexamplescxfjaxrs examplesrestsrcmainjavaorgfusesourceexamplescxfjaxrs examplesrestsrcmainjavaorgfusesourceexamplescxfjaxrs examplesrestsrcmainjavaorgfusesourceexamplescxfjaxrs examplesrestsrcmainresourcesorgfusesourceexamplescxfjaxrsclient examplesrestsrcmainresourcesorgfusesourceexamplescxfjaxrsclient examplesrestsrcmainresourcesblueprint examples examples examplessrcmainjavaorgfusesourceexamplescxfjaxrssecurityclient examplessrcmainjavaorgfusesourceexamplescxfjaxrssecurity examplessrcmainjavaorgfusesourceexamplescxfjaxrssecurity examplessrcmainjavaorgfusesourceexamplescxfjaxrssecurity examplessrcmainjavaorgfusesourceexamplescxfjaxrssecurity examplessrcmainresourcesorgfusesourceexamplescxfjaxrssecurityclient examplessrcmainresourcesorgfusesourceexamplescxfjaxrssecurityclient examplessrcmainresourcesblueprint examples examples examples examplessrcmainjavaorgfusesourceexamplescxfjaxwssecurityclient examplessrcmainjavaorgfusesourceexamplescxfjaxwssecurityclient examplessrcmainjavaorgfusesourceexamplescxfjaxwssecurityclient examplessrcmainjavaorgfusesourceexamplescxfjaxwssecurity examplessrcmainjavaorgfusesourceexamplescxfjaxwssecurity examplessrcmainresourcesblueprint examplessoap examplessoap examplessoap examplessoapsrcmainjavaorgfusesourceexamplescxfjaxwsclient examplessoapsrcmainjavaorgfusesourceexamplescxfjaxws examplessoapsrcmainjavaorgfusesourceexamplescxfjaxws examplessoapsrcmainresourcesorgfusesourceexamplescxfjaxwsclient examplessoapsrcmainresourcesblueprint fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofilescamel fabricimportfabricconfigsversionsprofilescloud fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofilescxf fabricimportfabricconfigsversionsprofilesdefault fabricimportfabricconfigsversionsprofilesdefault fabricimportfabricconfigsversionsprofilesdosgi fabricimportfabricconfigsversionsprofilesdosgi fabricimportfabricconfigsversionsprofilesesb fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofilesfabric fabricimportfabricconfigsversionsprofileshadoop fabricimportfabricconfigsversionsprofileshadoop fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofileskaraf fabricimportfabricconfigsversionsprofilesmq fabricimportfabricconfigsversionsprofilesmq fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofiles fabricimportfabricconfigsversionsprofilesnmr fabricimportfabricconfigsversionsprofilesnmr libbin libendorsed libendorsed libendorsed libendorsed libendorsed libendorsed libendorsed libendorsed lib libext lib lib lib lib licenses licenses licenses licenses licenses licenses licenses licenses licenses licenses licenses licenses systemcomweiglewilczek +What do you propose then? I think this would only be used if the plugin will be instantiatedused anyway so why bother about the instance being created when getPlugin is called +Closed as +closed after releasing version +I think that the database was not updated for some reason I dont know why +I think this is a new version of a previous bug where ! is getting reset to nil somewhere during exception handling +to clarify I think that some level of attribution is somewhat essential in this community +unprotected returns boolean to indicate the existing rename target was overwritten + seems to have a bug that requires cli +As reported please use insteadAs of version you can even put multiline JVM arguments in your pom file rendering the XML much more readable. +Burt since you volunteered to look into the problem Im assigning to youSince there is a workaround feel free to downgrade to a minor as Wont Fix since theres not much that can be done on the Grails side +All committed +My did show some errors but I havent had the time to investigate wether its related to your patch or not as of now +Since seralizes verifier and credential these tests need to be fixed for the review Brandon! Update the patch to address your commentsbq +Thank you for that. +These issues have been marked as resolved for some time now with no recent activity so I am moving them to the closed state. +Added a test in the +Updated patch which removes the additional cached locations from PBPB translators and does a precondition check +Maybe? Im liking less and less the namespace name probably would make more sense with the way things are named at the moment +Logged In YES useridThis issue was fixed for the release. +On X I can reproduce this over and over +Attached patch adds support for injecting controller entries and menu itemsIt works well but the code is pretty rough +Alright the first two schemas given as part of the issue description seem to produce compilable code with the correct code initialization in place +Ok I was blind +We just need to watch those new children +I will fix the spacing problem when we get a consensus on other problems +The workaround in the JSR layer can now be removed +RobbieI changed Test which extends now and uploaded a new patchCould you please commit it +FolksIf no one vetoes this contribution Ill proceed with integration of the patch in a few days +Verified on in trunk +New text has been created. +How and where you change the http port? By default Geronimo uses for http request listening +bq +I disagree with your assessment of the dangers seen in +In case of LDAP clients should talk to the LDAP server directly +I really really really want to see this in trunk soon despite the limitations +I tried to reproduce this with the bit F same Java and resolution but couldnt +Fixed in the SVN +Ive removed the cache so the pattern will compile each time it is retrieved however Ive added error checking to the editor to ensure that the users are aware whether their pattern compiles or not +Will somebody try the patchThanks +This code should not give the array to the outer world in this particular case unless its documented that you must not use modify the you pass to to anywhere else +Itll be a configuration option +Does it mean that has to be enabled for Harmony VM launcherIt is the first module that is loaded into the memory first +But I agree with both of you its a premature optimization I will try with first the way I use SSTSUW is not completely appropriate I have way better result doing one key after the other +Added file Regression test for the issue +Im infact looking for how to set it per server instance implicitly so that people dont need to set it explicitly internally +The Example as a zip deploying I only see the message Test no Hello World +change getput to putreplace for the issue described here +After that the classes should have been generated and the generated sources folder should have been added to the Eclipse project +Otherwise it looks correct +The new patch refects a fix to the new bug +Currently the binary does not kill the process group instead kills only the process id +Ill test it out +I suppose you did +This is part of a feature we are explicitly moving from tech preview to fully supported +First try at validating values length +Sounds good +One way to handle literals is to place them with the code they represent this would allow visibility using instruction pointer relative addressing. +Is that using the script? Wont the the namenode mark the datanode as dead only after mins? Should my test be thatCheersRavi. +Have you tried changingfixing the SUID for +Attached updated versions of the JCR API jar and related POM file +Bumped java stack size to in in +has been prepended to the relative paths has been introduced to replace the string in the output +Looks like we both worked on this in parallel and with slightly different approaches +Problem is and does not happen every time +I tried the approach that you described but had a difficult time to find out if the actual test method failed or if its execution was successful +assuming tests failed due to +Closing the resolved issue. +necessary? I just excluded in the dependencies in +Verified. +Thomas please be just a bit patient we know about release as we participated in it +The IS components are defined by the composites which are defined by the individual component tech leads. +added +Can you post and update of the tooling I can tryAlso are you running your tests on windowsThanks tried the latest IDE installer from your link above and I can now see the option for the IONA Fuse ESB Server +Please review +This may have to be slipped if we dont get the WB release done in time +I applied the patch to trunk since all the tests pass +Will commit +same issue here +It makes sense to include the delete topic feature in the beta release since most people might create test topics that would require scratch in How topics are deleted +Scheduling to a proper release as to not bump from Beta to Beta +I am not sure about the different callbacks but I thought I would leave it in the patch for nowParent Child and Scorer became generic argumentssimon +thank you Gert +Ok +Im using templates for compilers because the type matching is stricter +Making the roles configurable is not required +The problem is that I cannot change the CusrsapJEJCjeeclusterserverbinextwebserviceslibcontaining the package as that is part of the environment and loadedby the classloader of the SAP Enterprise PortalHow can I determine which version of the is used and which version has to be usedIs there a way to still use the older API in Axis and stay compatible with the older saaj version? +hey Ed trybin props config topic topicintwo the file is missing the property but if you use the you should be good to are correct wrong parameters +Thanks Ankit for reporting +Patch applied +The first file is the pom and xmlbeans config +I just added our the release audit to the patch process +It will add a bunch of deps necessary for the whole to run properly +This isnt an argument its a discussion and a mighty useful one at thatIts helped to flush out design decisions and guidelines and get them out in the open for all to benefit from +it may have been fixed in the meantime or may have been some issues with my when I was working against versioning +would it be possible to apply another resolution logic for +So the solving of a problem was to add a support to interfaces of a proxy object +If that is the case just let me know + +Do you have a test or a template file which illustrates this problem +I forgot to include the to git +new patch including unit test +If bag fills again all of its content are spilled to disk in entirety again and so on +If youre running with the service engine it will even try to the service when it detects a deadlock +jar it appears that other nodes get some sort of message but the +Hi KevinI have a very minor request for your patch +Jrme is right +Reopening so we can resolve it for LATER rather than WONTFIX +it doesnt have to work if you mount a page outside the crypto mapper then why should urls generated back to that page should be cryptedperhaps what we need is what sven suggested something that exposes a bookmarkable url and redirects to a crypted version when its hit +The description of the jira does not represent what really happens but Ill update as soon as Ive got to the bottom of it +for sure definitely something I wanted to do +done +I ran the failing test with the patch it passed now +Fixed in trunk +As it turns out I can. +I just figured out a way to implement the strategy one we could do some work on the to let it contain the url parameters hold the reference of to access these parameters when it is processing the message +Attached proposed patch to alter to parse iso formatted dates as wellI believe this is less error prone and makes implementing the easier +Since it is a medium test it is ok +Tested it locally and got passed +No worries Micah were all shaking off some of the holiday cobwebs +RodrigoI beginned so I continue +Can you wireshark it or similar and see what is in the payload coming backRequest for more information ignored after one month +I should probably run that this afternoon before we make a final decision + +BTW We will be using the for maven related activities +No problems Ill make the change and commit +You want to give it a tryRon +I am using Castor on Tomcat and JDK +I used dosunix but Hudson still cannot apply patchThe problem is that the committed code uses CRLF and your patch uses only LF +BTW this issue is supposed to due with global filters but not ssl as stated in the title +Given the perf number improvements we see can we consider up the priority +Attaching three build logs +Therefore cannot provide this token +Committed to trunk +Paul this ties in with the EJB issue +Itshard to immediately determine how thatd work given the slot concept whichseems to be an ord or value per reader thats directly comparable? Is there anexample of mixing multiple comparators for a given field +Kevin can you refreshSome patches went in between and this patch is not applying cleanly +We do have integration tests that useThey may not cover every part but if its completly broken it should show up. +unixdos converting file +The use case reported here is another good example of one in which the client would prefer silent disposal close in a finally block after the badness has been encountered on use +By using setvbuf on the c side and on the java side I was able to change the data transfer to larger chunks especially important when submitting many pairs with small valuesstrace shows now larger reads and writesI attach to mention the name of the patch new version of patch to qa review +Rev would you be able to post a small test case to rproduce the issue? Are you using Axis? Thanks +Uploaded test page and panels for reproducing these mustve fixed it somewhere along the way now works even fine on opera +OK not an easy one so Ill do it for the next release +Correct content +Committed revision +Happy testing and please take the changed package names into account when updating your application +Bingo! I tried the and builds neither worked and they both had goofy manifests +so given these guidelines then it seems like we should remove the release option and set the arg to something since this is the fourth incarnation to support CMS +tested! +testAutomaticallynew persistent class jpa name +We shouldnt allow interruptions in service just because Windows doesnt allow the unlinkreplace semantics +Yeah this patch just lets the exception bubble up +If you dont have one what I usually tell users is to assume the db they are accessing is corrupt but accessible +please find the patch attached to resolve this issue +Moving unresolved from M to M in preparation for M items not being worked on afaik out of M Fix Version +i think we should include the into avoid the lzo loaed twice +Ive cross checked everything fine closed the bug +For fossg +Ok +Yes happy to commit your final patch +A sql driving query input source has also been added that handles composite keys +It would be nice if we could control the interface the stressd is bound to and also had a way to cancel a running action +Thanks Zhijie! +See for the shutdown hang +Applied latest version of the patch based on Pavel O +Thanks Jacques would you mind commiting this in +This is very close to what I was thinking about Im looking forward to the performance tests +So please can the appropriate svnignore properties be added to the lib directoryAbsolutely +patch that only do the copy paste on lastest trunk +But when setting it to false I no longer got the NPE. +Ive added an initial patch +Sure thats great +Unit and nosetests for testindexscan test failures after rebase +Even if last were true themethod would go into an infinite loop at the end +Consider something along these lines for the branch as well +There was a bug in the patch before sorry +The same index in with compressed stored fields is a little lres than GB +I can give you examples later for this +I will resolve this and a new bug can be filed if anything else needs to be fixed +Resolving as invalid see the parent issue. +Added test case to to test this behavior + +I mean +Thanks for the advice +right +I know how to fix it but it will require a lot of refactoring of Suns code +Now that work is well underway with AS all previous community releases are +Can someone help to take a look? many thanks +fixed issue in fix concurrent modification issue in enable for sijie for providing a fix +My first attempt to set a breakpoint on line of was not successful +Adds the ability +I agree with Sergey the tag should contain a relative path instead of an absolute in repository was tested for relative path and implemented resolving system variables. +The same fix will be committed to trunk +we are ready to create a patch for apache camel and build files created and tested +What ever datanode has is the master copy +closing this as its checked in! +to backporting to as well +But this affects the storage engine below schemaI think its reasonable to make the change offline especially if an upgrade tool is provided to make it simple +change the shutdown order so there are no inflight transactions in the persistence adapter when it is by SVN revision +Screenshot of my profiling test application +If decorators are configured assume the one configured in the table prop is for html and use the others with the appropriate +Cannot reproduce this problem +is likely within a few weeks +Ill need to do some more research +I will make this available in a few hours through this issue +Resolving as Wont Fix based on Julians comments +The should probably be registered when the repository instance is registered as a service Probably should overwrite the registerService and unregisterService methods to registerunregister the service appropriately +Stack Checked your patch and this is exactly what I was trying to fight with +Thinking about it I guess its more an issue ofshould Boos builtin functions reset the enumerator once they are done iterating over them or not +Sounds good in theory Michael +im using and i have sorted the filters in according to the faq +With SVN Commit addresses the mapping with a when mapping is annotated is two independent mappings at both ends of the relationship +One idea that came up was to emit a script that could replay the requests that had been made using curl +Maybe the compaction requests could carry some reference to an optional callback method + +Now that work is well underway with AS all previous community releases are +esbplugins +Submitting +Thanks! +Were now inspecting the version property instead of the id for deciding whether to start versioning or not +Thank you Knut for tracking this issue down and fixing it and for the test! I think I have become too accustomed to ignoring the noise in replication tests and didnt make the connection with when this showed up a couple months later on zOS +PS thanks to Alan Matthias and Cheolsoo for the eyes +IT added +xml parser is now setup to be namespace aware which seems to resolve this issue +Ill install the trunk version and have a play +Ive seen multiple new test failures that I believe were introduced by the patch +javadoc javadoc warnings +Added patch Replaces incorrect tests with the new one +I dont think there is an Eclipse launcher for semantic search +Attach the patch that contains the changes for the singleton implementation +What we have now is a port of the code that works onParallel block copy is at the top of the TODO list though +Perhaps separate the after invocation provider into two an after invocation authorization manager and an other after invocation provider +The default will likely remain as caching forbidden but I assume that would not be a problem for you if you can turn on support for caching when desired +Is there a place where proposals are collected? Maybe a wiki page +This doesnt happen unless you run the git setup ideas how I can get this to stop happening? The only way I have been able to so far is to comment out the git addon from the roo build pom file +We can add that script to to be ready for V removal in +Adding patch +Oh and all the underlying work has been done it should be very easy to implement this now +Attaching the initial solution for the problemWill be attaching more patch with unit the new patch with unit test caseApproach in is changed to make sure that listeners are notified in the event of any state change of Jobs +Well fix bugs but there will be no new features added to the codebaseAs you have noticed the URI class is a mess +Just this warning message is annoying +extras directory looks ok now thanks! +Proposed patches for +aggregator for input super aggregator when loading edges at input super patch. +a fix for this issue was committed to the roller trunk at revision +If this parameter is present group memberships are collected in a node structure below NMEMBERS instead of the default multi valued property PMEMBERS +As I understand it the time zone is strictly a memory artifact in Java and unless the database has a specific type timestamp with time zone then the time zone information is lost when you store the value in the databaseAs far as I can find with a quick search DB and dont offer a timestamp with time zone but Oracle and do +I unexecluded the test and applied your patch at +I was having issues trying to apply the patch to and would like to confirm which code stream is this patch intended to? Also it would be good to provide a test case to avoid similar failures in this area going forward +This patch replaces the RPCs call stack with the one relevant to the client +Yeah thatd be great Elliott +By the way I dont know if you can get the from HEM +Id like to toss my vote in for this d personally like to set up a few archetypes with IDE configuration files too +contains the latest merge from trunk to branch +Test statusTests run Failures Errors Skipped +Hey guys I ran into this on probably because the patch contained in the file was never committed! Why would this issue be closed +shazron Thank you for looking into this so thoroughly really appreciate it +The key can be either CLOUDSTACK or ACS at your discretion and I am happy to serve as the initial project lead in Jira +But as a matter of fact we have an open internal ticket on this +and allowing suppression of that warning using a System property or something +I suppose you could split out where those patches were applied to as well +Probably wont have more time to look at it tonight but maybe later +Adam if possible please reproduce putting all your classes into srcmodel and disabling Seam debug mode +Will the patch be applied then or rejected? Ive made the name change as requested let me know if I should change anything else +I just saw my dream come trueso I propose it +Ive committed this +If we take the aliasing approach we should probably fill that public static final A MODULE thing with aliased instance +updated and attached that includes all used and attached that includes all used couple of questions Why do you have the live and backup on the same node? If that node crashes the backup will crash too and will be useless Does the problem occur if you run the tests outside eclipse? I dont recommend running test clients in eclipse if you set breakpoints then that will cause pings not to be sent and the system will fail over +Attaching a patch which does fixes similar to the ones introduced in and and it depends on these jiras +Sorry merged to doc branch at revision +So do BETA and BETA as you prefer Ill test them along the way +Its colors dark blue and green on a light grey or white shirt +The unit tests work now as expected +If we cannot have a single generic version than we should perhaps create two webonly versions one for tomcat and one for jetty +Given your findings I think it would be a good idea to leave this for now and wait until the AS trunk settles down +Will commit this in the next days if there are no objections +PrashanthAnother thing +Good idea Aneesh +pushed pushed upstream +This patch removes the debug parts of the release configuration +If you dont have time to add this Ill try and get it done +applies cleanly unit test pass +The patch is perfect +DavidI applied these change you made to Dialect etcbut still got these errorswould you please recheck this patch +The related vote on prc passed and I committed the change as proposed in revision I included Jims comment by adding the following text in revision The criteria for including including companies to such pages should be clear and applied equally to all companies +I am looking at MR which is as of now the biggest app where in for all medium to large jobs we are creating threads inside the AM to launch them +Opps +Great thanks +Sorry its actually Section fixed. +If true compaction is wanted +Thx +Odd +To have to update a gsp by hand once you have custom code in it should be unnecessary +They appear in English on other locale machines +Thanks for detailing this issue +I reverted this patchWhy? What broke +A CF can be defined to be indexed either by name or by timestamp +Thanks +Will make sure we get on the PHP side too +There have been no complaints since this patch was applied +Why are you askingOK I will double check the setParameters usage. +Please report to GWT user list. +I wonder if this was a transient bug that was fixed after M? Can you still reproduce it if so how? +trunk +A security domain element now accepts a attribute +use a to homepage instead to keep the page in your example stateless +Close issue in preparation for release. +we selected the red button to remove the style +The user may want results in sorted orderIll add that capability after we work out the data structures for reading and writing +That timestamp defines a realionshipWith KEEPDELETEDCELLS enabled for CF this allows for nice queries as well as idempotent operations +Okey dok I created to track it +Added two field with name delegate and comments in service definition and using them at the time of updating task +Patch for this bug was submitted as part of another patch for +We dont want to madate the use of on everyone due to the endorsing stuff +ok you can provide a patch for it +Same happened to me in the source due to personal experience in the source and +Verified out of date issue. +This would not happen with a String array but it would greatly reduce the functionality +Most Hash objects have string keys so the serialization routines in keys as an optimization +Look at the fix in. +Logged In YES useridYep I definitely do not want you to copy BEA on this I would like to have is to have some +Corrected null aggregation issues when docs contain null values for fields pivoting on +The patch you supplied for this in was for beta +bulk close of all resolved issues. +This issue reads very similar to something I am looking to solve for Remoting JMX for long running JMX invocations one thing I am considering there is enhancing the protocol to allow the client to check with the server that a request is still being processed +No worries it is one less thing to check +Give me a chance to create a test case for this and Ill post it. +Tom Hudsons results on seem to be OK +Seems to me that it should be fast enough that we can run a few thousand with no noticeable test performance degradationAs for the code I believe Scott wrote it with some guidance from some C source but its definitely not ripped from another project I googled far and wide for a good implementation in Java and didnt find any that were this fast +As a general rule I think applications should not expect to be able to hold on to pointers to objects passed to them but should expect to be able to hold on to pointers returned to them +Although Im still questioning configurable collectors this issue certainly makes sense on its ownCommitted x trunk +The patch does not need any unit tests this is only configuration change +Theyre using make and calling the main method to compile +This is the right way to do it +If this is still a problem for you please reopen this bug report and supply a complete XSLT stylesheet and XML input document needed to demonstrate the problem +Changed lowercase break to Break +committed +This is the patch that was accepted and that Im committing +like ability to expand the node or expand the node together with all its parents programmatically + +But this behaviour is new to +the error is only possible if you havent installed all mevenide modules that are necessary or your install became corrupt error shows some error in the maven embedder not sure what the problem is make sure you install all mevenide modules +The patch is against SVN trunk +Jason for now I am going to resolve this jira as invalid since you have not posted additional details +After more testing the posted patch doesnt work as intended as it removes the transitive dependencies of deps +I havent updated to yet but I didnt see anything in the changelog which might have been relevant +Might be an issue with the timing around the installation of the stdio redirection handlers. +Assigning outside and into this one to next release +This model project set may be used in reproducing the bug +Was missing entries in +We could change the error message to requires a positive odd initial value and then remove the extra check for +Short term fix is to ensure that this exception isnt thrown +Attached patch a which adds the method assertCommitted to trunk with revision +Help in resolution is appreciated +The following is only a stop gap but may be better than nothing but may be easy and at least plug some holes +I did forget to open up the firewall for port so yeah works now +I think is related to this one +Sorry about that the gemspec was part of an old setup to build on Github which of course doesnt work anymore +Persisting indexing errors after the Lanczos verification job +And it appears the patch is now committed to the codeline so all is good in my book +I put this into trunk and but it does not appear to be a clean merge to +It looks like this is simply using to separate out common functionality +Yes the initial committers list is the list of PMC members +Ive opened for this. +Comment if a not equal classes conditionNeither form makes grammatical sense +Any further tweaks that might arise would be distinct Jira issues. +Please provide an update on your issue as a comment +Closing this issue as this is a guest os specific issue. +Build it on a clean main trunk and it seems to work good for and. +Give er a go! +avro is also used only by contribpiggybank and contribzebra that should also be moved out of +Im getting the same exact compile issue +the in could be used for the same issue. +It was probably an oversight +They were there as a convenience to avoid requiring users to download them but legalpolicy requirements are likely to force us to stop this practice +I will add it and also check other parts +All seems good. +Leaving the jgroups in the transport works fine +Should be fixed can you please try latest trunk +In my view it should never produce them +We have been waiting for somone to provide a fix or sponsor working hours to solve the the above expression works for values greater than one but for those lower than one I have doubts it may be better to add DBLEPSILON in that case +Added an id attribute to the fileupload input verified it in respin +that was the wrong patch +I corrected the Copyright line and notice on the Copyright page in the Japanese translation of the Getting Started guide in the trunk revision +The DRDA type checking does not need to be in forHowever I plan to do some more testing of soft upgrade before I close this +Will have a look +Agreed dependencies of the final distro need to be manually checked but this should be getting done regardless +Even the empty JNDI name is error is fixed on Teiid side we should also log a JIRA on Designer side as to why it sent a blank JNDI name +Could you please test with the latest trunk +Thanks Bobby! +nicecommitted +I can see if theres already a SOLR ticket and if not create a patch and attach it but I cant do that until this evening + +If the target does not exist we start screaming Im not sure this is a correct behavior +with no children of tag a it will be ignored +limit up to mb give it another Done. +Sending serversrcmainjavaorgapacheopenejbclientAdding serversrctestjavaorgapacheopenejbclientTransmitting file dataCommitted revision . +There is a in release for some reason its not merged into the stream +I updated in r so all doneIll double check then delete those contracts Thorsten said to delete later todayThose to delete are obsolete contracts +Mithun Can you please re upload the patch and also make sure you grant license to ASF for inclusion? Mahadev +This can lead to and hanging tests +Attached fix and test patches + thanks for finding and fixing this one +Ok +Sent notification to dev list +What version are you using? Your line numbers dont match works in CR +I downloaded again and test is successfully completed on my clusterLets commit this patch first and later improve the code! What do you think +Any progress on this? Id really like this feature for a distributed application environment +Is there a workaround available? Would more information be helpful +We made initial changes to fix the issue and in process of initial validation + branch build breaking as the test fixes for and were using hdfs protocol related changes that have not merged into +fixed in my original intention was to add these files to the source tree for metro so we could have the metada files for all the possible configurations +FixedI really hope were done now with this one +Caches the inThis patch gives on Clovertown and on Tigerton for MT +addressed nicholass comments +Sometimes we still have a disconnection related to the unmatched ack +Can we make this simpler by not supporting manual failoverYes +We could deploy to our own repo if its not available anywhere else +The synchronization protocol for standard maps should include both reading and writing +These blockers and critical issues are resolved but not verified +We should patch and trunk +Looks better in the code verses a dummy variable +Closing resolved issues for already released versions +the line in the logj file could be commented out with a line above saying uncomment to turn off warnings the javadocs of configuration should mention this as Steve Steve can you please take a look at this change? warning is also output in shellThe latest patch looks? +If you see that the threshold is not getting checked during volume attach reopen +This is more about a separation of concern System properties are global properties +This jira covers how the replicas are load balanced not how the shardslice requests are managed +Client A retries submtting job and is returned Bs jobDid I miss something or could this happen? An easy fix is to make more unique + +attached source +Reopen if there are any more problems. +In general Quorum based readwrite requires that Qw Qr Q +Verified that the expansion and contraction states now appear to be updating in IE by following the repro steps. +That should be fixed with t abuse the ticketing system +It looks like the replay was actually from the last checkpoint itself +Given Im parsing the on the fly to look for a place to add a charset to an existing header or to add a header if its missing +personal problem that should be dealt with on a basisI think its more of a project level problem we have to decide how we want to handle logging and stick with it I think +system can you review +Votes please! And please bear in mind that thiswill include four new JAR files in libtests +Attaching +Another ideaWe could add the bug number to the branch name so for example use the branch name branchbug +It would seem that this may also apply to is that correctFurther this does not only apply to the FP version the EAP but all versions of the EAP so I will update the version information above to reflect this +Fixed in nightly build. +Ill see if I can reproduce and post a screenshot. +how can i do it? please let me know the steps what should i do? Thank uemail hardsujan I wanted to tell u what I understand about requirements +If you dont want to make geronimo tag you can modify daytrader root pom and change geronimoVersion property to that let you avoid build geronimo code and make daytrader build work directlyI need to replace tag by in geronimodaytraderbranches only orin all the files in all the subfolders +Is there an angle in Bigcouchcloudant that makes this much more important aside from sheer size ofWhat does DROP TABLE or whatever require in an RDBMS? or zfs destroy etc? Any bulk store has this riskfeature +We may just need to write the bytes to disk if we find this +tests now pass still need to look at adding typed SEI have a working implementation in my tree will commit it with a fix for in final testrun I ran was on Windows XP within cygwin +I add two patches adds interface Focussable and changes and +Will attach the test log next and will be looking at the failures +Im still not convinced about hashing +I will run using the trunk test against just to make sure it is not a test change that is popping the issue +Yes Ill try this now +thanks! +Could you please take a look? Thanks +Regression tests passedCommitted patch a to trunk with revision Resolving issue I dont expect more work on this. +There is a bug in and you would need to patch hadoop core with +the patch has been merged to master +I dont think those are used for anything else I will wait until the release of to commit +I am having this same problem and have gotten it to work by making the aggregator pom the parent to the parent module but would like to know if there is some way to work around that so that the modules do not have to have the aggregator in their parent hierarchy. +Been having the same issues +Let us know when youre ready to convert the live site over to the cms +Scheduling it for since it is +Code committed but this Jira Left unresolved until it is included in a to CP release notes as support for Japanese and Chinese has been added to the BRMS Web User Interface +I have already had use of it when debugging replicationCommitted as revision +UTF orWhat is the borwserWhat is the default languagePaul let me clarify why I think browser should have absolutely no saying in the a resulting encodingNowadays all browser should be able to handle any encoding just fine as long as it is state in HTML page header +I think we need the email address to allow tickets to be raised and commented on via emailSo for outgoing mail would we be expected to use SMTP through port on the VM or use the official apache SMTP server +Please help to reviewThanks +Looks good +could take a long time before it finishes and we could get very big regions! So now with the client handling that we should have a better storySome nitpicks some classes have unused imports after this patch like HCM and I also wonder why you use inside as we never used that class anywhere else +But since Id like to use the provided operation classes I did it all in one huge dump nowPlease be gentle with this patch +Please let me know? NOTEI am using and have it deployed on webshpereThanks in advancePratik +Besides end and set its mostly finishedThere is also an integration test which uses staxsax bridge for unmarshalling +Pull it back in if you think different. +But shouldnt we? Ie you interrupt so that the thread stops sleeping and throws an exceptionSomething somewhere higher up must catch the I think +In that case we should probably avoid relying on it +Problem is registration to common web server it should be done only for default hostserver +great +Thanks! +committed patch to trunk changes should be merged tomsvn commitSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestsmasterjdkSending javatestingorgapachederbyTestingfunctionTeststestslangSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +Triaged for +About the lines fix can you provide a test that shows the bug + we shoulnt have different subclasses of users might reasonabyl except to get the same subclass back but an implemenatation is only required to store the uri reference itself +Good point will use Javas rename in case the target does not exist +I assume this is to enable special consumers that can fetch from any replica and to be able to tell external consumers from followers +HiI backported it to branchIt works fine for my failure and I verified all tests were passedWould greatly appreciated if you merge this intoThanks patch includes test case changes plus changes attached by Hiroki Tateno +The patch +User can configure the thread pool profile by setting the maximum to make sure scheduled thread pool doesnt eat up all the memory when the request load is the user sets the is to be false which means the delay invocation will be synchronized the thread which call the throttler will be blocked and it will send another feed back to the caller by slowing down the request +I think we should do something for too though +Maybe adding an Attachments section to the users guideAnyway if using solves the problem it seems to be a bug in or any of the classes related to itRodrigo this problem also affects to Axis on Tomcat with Sun JRE on Windows XP for documenting this issue Ravi +Ah it appears my patch generated with git does not have the binary changes within it +Agreed one main difficulty with updating ZK from bookies is that no bookie alone has all the information about a ledger and bookies dont communicate with regular readwrite protocol +Is it plan? I am not sure +Committed fix to branches Branch and master at gitronsigal +After further discussion with Pete it seems that the CDI spec isnt absolutely clear about valid methods for enabling alternatives +Using the prefix cl collector parameters +The inputFormat for mapred is but the mapreduce inputFormat is Im am not sure of the history here or if it should impact the trevnis inputFormat naming convention +In the case of Accumulo the table scan returns nothing because it was filtered on a nonexistant column +Putting the fix in sounds logical to me +I will update design doc with addressing your comments soon +The intention is to have it default by CR watch for that +If we do roll an rc for other reasons it will be incorporated +Given that BSF no longer includes any engines I think this issue is no longer a problem +codes look goodminor comment it would be better to use allowedChars to construct the help message +The revised patch looks ok to me +I can sort the commands into buildbot for the initial setup just tell me what you normally run to perform a buildOnce I have a basic config in place and working youll have access to it to tweakadd to it etc no problem +If somebody finds more welcome to use this JIRA +After getting feedback from Knut Anders and reconsidering in addition to the limited time left before the release candidate is produced Im strongly thinking about closing this as Wont fixThe points Knut Anders raised about code maintenancecomplexity and compatibility are very valid and good in my opinionFurther just using the simple release mechanism is still an improvement over the current situationUnless someone strongly disagrees I will close this issue and the subtask as Wont fix +Im not entirely sure why you would get different results +This is either a bug in Grails or a bug in Tomcat +now looks elegant +This causes a downward spiral that takes quite some time to recover due to the filled up response queues +Yes +Good catch BenjaminDaniel not sure of the status of snapshots but if we would provide an updated snapshot release could you test it +Fix findbugs warning and terminal async immediately if error happenAnd I also change the order of radix tree which greatly reduces inner ledger id node +Deleted documents are supposed be able to contain meta information about who deleted them etc because they replicate +Otherwise it uses the default encoding +Forgot to mention a quick updateWork on filter pushdown on primary key range scans have been completed via Note that filters are still not pushed for keys +Patrick any luck digging this up? And can I assume that you provided customreadObjectwriteObject methods on correct +But and still catch Throwable and eat it up +Obviously if the suggested fix doesnt work feel free to unassign it +Gary +I couldnt have just one map because the key lookup of the computed value should be the original class name uncapitalized +Because in client side what you are doing is sending but not expecting a resultAnyway we should not have those invalid read if a user done a wrong configuration as well +since its a ratio then it needs to be to actually do something right? comments should reflect that probably should be the default off +The meaning of is pretty much clear now its what you need to start restart a job but doesnt contain the exit status and execution dates etc. +Incorporated Konstantins review comments +The series no longer requires specific suffixes. +More like joyful amusement its been broken for so long +Hopefully by the end of the week Ill be able to check into sandboxprototypesqlparser some code that uses optiq and a dummy schema and converts a very simple statement select from table into a logical plan +Not even two way communication? Id like to be able to push events to the clients is that possible? Has anyone done it +Verified in SVN +Due to my local env setup I was not testing what was in the patch +Patch w fix plus some cleanup in +Thanks Gautam! + was a similar issue where the ping error prints but the test passes +As a follow up the economic tests should clean themselves up since that is the real issue here +Got your point Andrew +Patch for issue +Which should not affect the hib catalog afaik since xstream only applies to rest and those properties should be output the eh xml representations +The issue is not reproducible on current DRLVM sources +Well lighttpd would serve files out of the file system which means the OS will effectively cache it in RAM +I typod the comment in that commit but it captures the idea +Here is the patch to resolve this issue +If you could could my patch to handle shards be applied before you reformat so I dont have to piece it together again and resubmit +I will need to rewrite the plugin again from scratchThe internals of Lucene and Solr has changed drastically since this was first implementedI will keep you informed within the next weeks +Ok +Need to make sure this patch works after +this is for hadoop +I think just count in normal scene will be enoughit will suit for most of usif we need count something with bussiness logicwe can write one ourselves +gtr invalid backslash escape at end of stringecho i tr sed s +Toni do you remember who did the regex DSL things for the guided editor? Was it Joe White +Kim sorry for jumping on this one but easier to explain the issue in a new pacth Btw I also removed the comment about this pertaining to system properties only the permission needed is unrelated to thatDoes it look ok +Attached is a patch +And my disclaimer I didnt design this Im just interpreting the code +the rename of the and its impl was for aries jmx to pick it up correctly +The closeSession may do a massive amount of deletes so potentially we would need to go through and figure out everything to delete and send a cversion for eachive been looking and i think we can simplify things +The distinctions reflect how the server lifecycle is controlledDOMAINSERVER separate process controlled by the SERVER separate process controlled by the end user typically via a launch script but not always +java file in it and can be other one is I am looking for workarounds at the moment since I would need to move I blow away my current jboss setup and try again +Can u please explain a bit about the issue +I should upload a patch to shortly. +I just pushedTake a look and improve if you see somethingThen Ill merge it to master +We do print a lot of output to stdoutstderr for many tests and if I make redirect false the tests pass +This will test that if anybody tries to upgrade an old version then the upgrade will fail +META +This patch depends on a SNAPSHOT of +So we need to be able to set the maven version for each project +This happened after I applied the v patch +What do you think? Tom youve raised some concerns about adding the ability to setup arbitrary environment variables +i was using the party mycommunications email from unknown parties form Hansthis is fixed in rev +If you finish it before we release beta please correct the version +OK Ive closed the issue for now +instancesstatic is now removed. +The test case doesnt build using Boo trunk +Could the difference be that on Windows you have the option of having the hardware detect when the cable is unplugged and the network interface is completely torn down whereas on a Unix box this might leave the network interface configured but the socket now gets the EOF after the TCP stack detects the timeoutJust a thought +testcases were removed as they seem to be excessive +Also as I said preset aspects such as Transactional is another case where the problem Ive been claiming for a month now is that adding a controller interface in order to overcome this issue IS NOT THE PROPER SOLUTION +And provided yet another patch +Is that correct +Its dependencies are already on the central repository +Ill go and scurry back into my corner now +How about we resolve the operations in the datanodes so that the namenode dont have to add and delete the same block multiple times +Thanks +You have changed a significant aspect of the Jira so this should be a new discussion on the +only fixed the of the profile activation that you observe with SR vs SR +You are right Brian done long time ago! Thanks for all the help with this issue I miss working with you +I tested with and couldnt get this to work +Ive updated the patch to use the same parameter names as which will be inI successfully ran and unit tests +As the issue is fixing after restarting do you think is there any chance of not loading the class at the time marshallar needed it +Tom looks great +The tablets key is encrypted with the table key +ll commit this one in the next few days unless there are objections or improvements +change to integer representation based on days since epoch pull in lib as dependency support date literal syntax limit the amount of type compatibility of date with other types Hive server jdbc support Support additional classes Additional is the same as the that I had posted to +Clearly a cosmetic issue an equals sign has been added to delimit the key and value +Just one question +Spam or mistake either way there is nothing to do. +processing the partitions that do not exists yet and add them +Im confident the javadoc warnings are unrelated since this patch doesnt update any javadocs +Ive noticed a similar possibly related issue +I wonder how this got introduced! Would you be able to contribute a unit test and perhaps a fix +Is it possible that the wrong file is attached to JIRA +Still its not hard to add up all items sizes which should bevery close to requests size +We should be deploying with +either java sources or aspects will get compiled with wrong sourcetarget level +Thanks David +whats the reason +Patch applied at +Think about it it doesnt make sense to specify an executable if you are not going to fork a new process. +Thanks Robert looks coolAnyone know what the policy on the stop word list being BSD license is? I assume its compatible with Apache? Whats our BSD license policy? I dont see anything definitive on a quick mailing list search +bq +It appears to be a tomcat issue +Documentation is pending +Committed the patch to trunk and +Going to update the page right away +So this patch adds one +Id be tempted to assert +EmmanuelI will investigate my environment as you suggested +Code is implemented and it calls into the +hi Ivan i have updated the patch with your comments +I think this is an issue that should be given high priority since it makes useless +Could that work? If the client is going to send to Datanodes then we can have each split be a filename or a list of filenames this will work fine if we dont want to validate data +Plus from Maven side that means the reactor will do much more work to resolve module dependency thus slowing down the buildFinally I advice against merging into one this complicates significantly troubleshooting +Thanks Adrien this looks great! Ill commit soon +I am also adding a test case for the concurrent split alter scenario and I am validating the same +The attached test case should demonstrate the problem +Is your screencap from a installation +I think the approach of including the job history file name in the URL since the beginning will cause more headaches since the job history file name includes some things that are unparseable by humans + link was broken before javadocs were just never generated because it only had package visibilityI think in this case the link just has to be qualified as or or whatever. +Thanks Ivan +Extracts into subdir nowAdds which is a simple test for a table with millions of columns in each row +Jeoff you are correct in having extended +fewer issues on the this seems to relate to a problem in the lexerI created another ticket to cover that properly +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsernodes +Since previous patch generated exception update the code and attach the new for the inconvenienceI think the wizard can be launched with this timeThanks very much for your help +Already fixed. +Zhihongs suggestion is to merge the changes you have into the updates in +In previous WSDL versions it worked fine but they added declarationsThe problems are the constructor parameters hashcode and equals I thinkBTW everything will compile properly but the moment you try to run the code the exception is generated +Revised patch with coding style +Suresh thanks for posting your notes +did not looked bak once changed updateRemoved comment along with this patchIn trunk waitFor is throwing do you think I can file a small trivial bug and change for consistency in code +Looks great! I will run the tests tomorrow and if everything is fine I will commit the patch +I will get on the Artifactory listforums and try to find out +Lease recovery should always be default otherwise we will allow data loss +Perhaps Im persuing the wrong method? If not Id hope that we could come up with a simpler solution +Its fixed by issue still occurs for me +At the end of a superstep partitions with not active vertices and no messages get dumped to disk and substituted with those on disk that have active vertices and messages if there are +bq +Changes will be visible when the next version of the site is deployed +Thanks Andrew for resolving as duplicate. +Applied +sounds like a good idea +Closing now as the last update was days ago I assume that this has slipped through someones net +ibiblio was brought in only to serve zookeeper dependencies +I wonder if this patch is needed for +HiI have tried the fix with a Oracle Data Base and its not workingHave this been highlighted earlier and has a fix nowKind RegardsCharith +Thanks Jason we will fix it +Sorry for closing ladsHimanshu is messing w passing a local flag so we run the mapreduce job all in context +Are you sure? In my opinion it should work build should be checkedmodified +Deploy and the modified to the default configuration of Jboss +yes it should be direct to HTML eventually +Im having a hard time reproducing this problem in any small context +Coincidentally I think the patch to fixes the problem +change reporter +Wouldnt this be a perfect candidate for M? This should include Iterationrepeating issues according to Brians blog +Thanks for the patch. +Confirmed +Thus there should be no problem with implementing the getNode change save cycle for importing jcrroot with REPLACEEXISTING as specified inYou are right in that the REPLACEEXISTING is defined to remove the entire replaced subtree includin in this case jcrsystem but I think the same logic as we use to prevent modification of protected properties and child nodes in normal imports applies here as well +patch applied +Hbase uses so for backporting we can go with or if it is binary compatible +What exactly still needs to be done on the side? Can we close this now +for which removes the and related files +To give some background about replication the reason that REPLICATIONSCOPE is an integer and not a boolean is that it is meant to be used to as a way to encode routing information but this has not been implemented yet +xml descriptor files generated by our build +Heres new combined patchCommon build was successful and all artifacts were pushed correctly +Its missing a letter e between m and tThat is why it worked in your test case +Dup of +That is to say only page name instead of the full name with category is needed inAnd if we need to support user based navigation customization it does require configuration in to make the installed plugin be accessible to a particular user +The hierarchy is arbitrarily ignored why does exist when there is already an established hierarchy? What is the difference between +fixed in r +stupid copy paste in windhoos +rename connector to connection in above +The rest treat it like a struct with optional members +Its cleaner to not have it in the main namespace +Reviewed the patch looks good and matches what I started over the weekend +Gary Im assigning to you because I know you have quite a few ideas around this +If enabled for client the userid would be passed as is to the embedded driver and be handled like it is in the embedded driver +The question is not whether the elements of depth are included but whether theyre each stored in a distinct column +Not very object oriented +although it breaks ofGerman Analyzer. +Everything OK +Updated patch which brings this issue up to speed as of Dogacans comments +Resolved on both and trunk. +is it the best name +Patch applied on trunk and all issues that have been in resolved state for more than one month without further comments to closed status +Sorry for not providing a patch but I have a not quite stable of the installed +Committed to both and trunk. +Looking at this now +Done. +Added to HEAD +in an XSD documentation and java code are all generated from one source so should be a +In order for getting this to work I had to change the implementation of a bit as well +Trying to find where your unpacked app was in previous geronimo versions was a challengeIm not sure this feature needs to be enabled by default it might be enough to supply an explicit alternate location when you want to do this +st Screen shot screen shot +Added the file for testing this change +The way Jackrabbit handled this case is IMO not acceptable +I think we can move those tests to their final location whenever its time to move the code into the actual SDK +In the meantime it is possible to use the hex editor to load a +So thats why I didnt bother describing the scenario +Dims Yes that will solve the problem and it will be the easiest and more appropriate way to fix the problem +Time based bounding would also be a lot more complex to implement as it would require a garbage collecting thread going over each message checking it against the bound and if so deleting itThe nice thing about the number based message bound is that it does the cleanup just in time and it allows administrators to reason about how much storage their system will take without having to worry about tuning garbage collection +What if some user uses the node runs a package manager to update things or uses scp to get things off the server? the fs cache is likely to get screwedI found out memcached supports domain sockets Im now working on implementing this around domain sockets +This is a platform quickstart we need to deal with it through SOA +Uploading a new patch +ChrisWe currently dont do this with any Producer in Kafka + +The more rigorous the rules for RB end up the less they should be for patches submitted after review so that committers dont have to exert the same effort twice +Verified +Having all things compiled reference +fix in rev can you fix file permissions on apache snapshot repo? I cant deploy a new. +User authentication must be enabled before the Network Server installs a security manager +Compiles cleanly +This issue has been resolved for over a year with no further movement +The projectartifact that requires is jbpm thought our products were not supposed to include things for Jetty +Thanks +closed. +in name loses status as node set may be related +Ive attached my modified POM if it helps close thisddDavid Trustin +Ive checked in a patch in the s not the rewrite I was talking about above +Sounds good +not going to make no one was interested in hacking on it yet +sounds like a resinapache problem +I am yet to take the builder code to fix this +This is a simple copypaste error +Theres actually some advantages to not breaking with the first error as there are operations that just mess up one feature and not all of them +looks fine with patches +Whats the status of this? Can we get someone to review and commit +Spring Security has a class called which translates these events into events. +My bad forgot to resolve but that has been committed a few days ago. +Proposal is to add an openSearchertruefalse option for hard autocommit config in Future enhancements could include openSearcher flags in the actual commit command and also on the URL for update commands + to trunk +SishenI dont believe it matters from my perspective as Im always dealing with byte arraysStackIll encode the row cell name and cell valueI have Michaels updated json jar and have incorporated the base changes into the json serialization +Ok +Now it does. +The segment dumper will serve as a plugin for Nutch to dump files from but I am still not clear that which it should be mount toThe dumper requires a file which contains the mapping from mime types to file extensions and it also requires a third party library. +This issue was about the message decorator on the message flow which I dont see on the diagram shown with the latest version of the toolIll attach the pdf page to show what I meant as well as the bpmn model. +When set to this will limit the number of pending exchanges when sent async +Stack thanks for review and commit it. +Patch available for review +Fixed +batch transition to closed remaining resolved bugs +Applied Sergeys patch in r +I can set the execute bit when the scripts are committed +Reviewing this Ive had problems last year with bulk IO and these filters +fixed in cvs +Konst asked that the patch for apply against trunk instead of depending on this one so I moved the test from that jira to this jira and also made the tests waits indefinitely for the required failures instead of just a computed amount of time which Dhruba suggested in +I think I will take a crack at this Ill bet an interceptor and maybe a tag could easily accomplish this +The issues turned out to be a good thing +Ive just reverted the trunk commit but youll still probably see some Hudson integration comments here +Temporarily put a copy of Javassist into the Formos maven repository. +Works fine as I can see +It is done with strings within the methods so no fileaccess is involved +I wrote the core logic to many years ago and its been patched a numerous times forvarious issues so the code is in definate need of refactoring but thatll needto wait until we have a more thorough suite of ll review your patch as soon as I can +It might need escalating to Atlassian +GET is the right semantic for a query since it doesnt change the resource +Change the metrics keys to follow the conventionnumspeculativemaps speculativemapsnumspecualtivereduces speculativereducesnumdatalocalmaps datalocalmapsnumracklocalmaps racklocalmaps +When that code was migrated into it actually caused a mem leak on if I recall correctly because of a runtime bugMaybe the right thing to do is to make this optional behavior because if you have searching threads you probably dont want this data to be cached and want to release it as soon as possible +havent helped thetone of the conversation either +Testing this with the most recent nightly build of fixesthe problem +As i said Opening a using the Configuration object would decide between local fs or hdfs based on the execution context Yeap the Sequential algorithms use this principle you mentioned it determines which file system it should choose according to parameter hdfs is given or not in training and prediction procedures +Well first of all if you are using dynamic method invocation you dont need wildcards in that way in the action mapping +This patch is against. +Ok reflection looks ok +Yet Im not totally set to option neither +ok I think I was mistaken +The patch resulted in compilation errorHadoop QA has stopped functioningPlease revert the patch fix the compilation error run the whole test suite and publish test result + +He is interested in doing a interop with Axis for MTOM like what they did with at Java oneNow Im working on delivering a MTOM enebled current Axis build to him +code samples to QE for review + +Patch lgtm w this caveat +Sorry I downloaded your test project and couldnt reproduce the bug +Coda metrics represents the industry standard and is used by hundreds of projects +The problem only exists when using any activation context and additional parameters +We just want it easier to set the default height of widgets in the properties file so it can be changed without rolling the application +Wondering if that is the important difference +Trunk looks good to go +Thanks Bryant +to trunk Thanks! +I sense theres a documentation update coming out of this issue +You should not overload setters on your bean +I guess the issue is then online schema updates are disabled by default now on account of flakiness observed during integration tests so a flip without disabling the table isnt currently possible +There are some redundant imports in files you modify +Patch v uses to sync log splitting with test +Added a pointer to this issue to in case these details are needed +Typo Section The second paragraph has two periods at the end +This patch injects javascript +added warning +OkayI have done another check +So from the last exception it appears that CXF cannot find a class Since the same service deploys fine in native I suppose there is some issue with the way CXF processes the web service generated by EBWS +Updated on the two earlier patches for this ticketThis patch enables the last log file to be recognized +I dont know enough about it to determine if this could be fixed on its side instead of applying the fix proposed here +Under Linux with JDK to develop makes sense to me as it allows someone using Castor under JDK to debug a problem they encounter +Thanks Philip! +Ive updated my github repo with an implementation of thread dumps + Submitted test revision +Run mvn f help +Its not a navigation but a menu on the breadcrumb navigation +PS I was thinking how awesome your contributions were about a year ago they are still alive and kicking today! +I am also voting for this issue to generate the pure java code +I am looking forward to using it instead of jconsole +Found a solution which works for me upgrading to +In the future we will try to follow the process +Yes that would be ideal +Yeah I was thinking the people might be matching filenames based on a regex or something +sigh +Can you please verify? Thanks! +Should our unit tests do the same in some run? Absolutely +In a terminal I do the followingelihusmails telnet localhost TryingConnected to localhostEscape character is hellohelloConnection closed by foreign hostAs soon as I type in hello I get the Connection closed by foreign host +do you think this patch can also work for Series? If not is it easy to port to? Please let me know your suggestions +Does that sounds ok Fabio +Show all future iterations up to the page size +Awesome +Brian it didnt work neither in nor in +The reason for that turned out to be a typo in the Tomcat file +tab settings changed +Ive create the task for +AmitPlease create another JIRA and rebase your patch for trunkThanks a lot +Do the two combine together to make the whole thing atomic overallIf they dont I guess the worst case scenario is that two callers would get the same return value while the value would still be incremented by two +Freddy Sonar updated to log errors when a review command fails Rule title added to the ticket summaryAnd documentation updated on the Wiki +Tucu do we have an test that checks that it can read files big enough to require chunked transfer encoding +move reporter from Steve to defer minortrivial issues to +Would you like to wait with these and similar patches? Im guessing that Ranas new code covers most if not all of the current code base which would make most other patches cause conflicts +Username diggermaus +To be reopened if there any remaining Im glad it wasnt just me +I assume that we would have the abstraction here so that folks can try out different systems ldap zookeeper etc +thanks Tony +Ran Sort benchmark on big cluster +for Stephen Smiths patches +Currently this is the cause of a blocker for a major release at my company +retriggereing the patch build +Still havent take into account some of Aruns comments though +We can revisit at a later time when work continues on this. +The change was deliberate but has unintended +Yes this was observed on a trunk +Instead there is a separate popup in the case of in memory raster readingThanks for the comments +if text of errors is present it must be updated +I am not sure the spec has been updatedWill go over the spec and patch in tomorrow so that I can gain better understanding of these two +SLD will be attached for reference +This is a very useful improvement to the javadoc generation and I want to make sure that it does not drop off the radar +JDBC tests run cleanly +Sorry yeah +Heh +Such methods permit changing the format of parameters renaming parameters etc +Please add new issues if there are less broad more specific issues. +Can you think of something besides putting in a check in that blows up with a cleaner exception +Not sure exactly what you mean about positioning an example would help +Generating interest is a very very important task +Not sure what happen with local index +The numbers captured reflects teh failures properly +bq +Thanks Zhijie! +The idea is to replace the existingcurrent directory or file with the content of the specified snapshot +fixed by ve got a few more changes on this one so I dont close it completely right s a nice of this fix is that theres no more encoding issues on GAE. +Michael thanks for reporting the issue and providing the test +I need to further change the join type to regular join to make my idea work +handlers is enough since there are not that many to datanodes +The local repository was moved over to labs earlier today. +Hey RebeccaI see why you want to associate it with the other JNDI information but not sure the chapter fits in Part III Clustering Guide +I can look at offering a extension to Tika Server to do with supporting multiparts +Perhaps we can improve the error message +What I wrote before was plainly wrong +Kalle what i see is only your unwillingnest to solve things +I know and supporting it +This error not there in the trunk +If you deploy a service and theres a typo in one of the dependencies you get Error Unable to distribute Unable to create configuration for deployment load of testfoocar failed Unable to resolve dependency asdfasdfjarThen if you fix the dependency and deploy again you get Error Unable to distribute Configuration already exists testfoocarMeanwhile undeploy says Error foo does not appear to be a the name of a module available on the selected server + +We usually reserve comments for code that is unusual or hard to understand and private constructors are understood by most Java programmers +Cleaning before doing the distribution build helps. +Hi AndyGood catch! I agree with all your changesRegards d suggest after looking at the data as I interpret it that the expectations are wrong + +Just checked it works fine now thanks for this. + +This is checked in since rev Atompub TCK was with mvn install see results below T E S T SRunning Tests run Failures Errors Skipped Time elapsed secRunning Tests run Failures Errors Skipped Time elapsed secRunning Tests run Failures Errors Skipped Time elapsed secResultsTests run Failures Errors Skipped +Sorry should have mentioned I ran it as a unit test and a system test against and +Verified onWill close when it will be fixed also in JBDS version. +Were implementing a WSG +Downgraded to minor because SDB is mostly maintenance. +Closing the issueMore work needs to be done to support running the tests with jars but that belongs under a separate issue. +Should note also that I didnt add any tests because this was a fix for existing functionality and frankly Im not exactly sure how to test it +Would it be a big problem for you to put those in srcmainresources and use the standard resource filtering mechanism +Applied the patch and resolve the issue. +It seems that as far as this bug is concerned is no worse than +I am hoping with that approach these kind of problems wont show up +Right +So what to do about numbers convert them into nulls +Any objections +And we think a complete implementation should send requests such as refreshNodesrefreshTopologyrefresh to both the active and standby NN. +Not just VR +I included more statistics in the report number of minimally replicated blocks which is useful for checking condition +I would definitely like to see this applied to TRUNK as it would simplify my custom themes a lotPS It appears my recent patch introduced some TAB characters into this file sorry +The Model and +controls are enabled by default so it oblivious result of custom policy in the customers domain +Is Apache on Guvnors classpathIIRC adding to your Application Servers lib folder or uploading as an asset to Guvnor should both work +at the moment and dir poms will only fetch jars +When executing configure from outside c configure sets the INSTALLPROGRAM variable in the to the absolute path location of +In my case I was trying with the queue prefix which should have been a good hint +hey youre right +Plamen could you please post the exception here +With the patch applied the surround tests have no more stdout +The topology can change after startupFinished of course +VR started successfully after installing ssh client +So this is unfortunately for nowWe discussed to add instead what do you thinkSorry I hope that is understandableCheers to provide some closure we got over this issue by using mybatis sql inteceptors +configure yes make +Im going to close this issue after p goes in +I will continue tracking the memory usage to get some more insight +Actually I wouldnt fuzz so much +and a cvs diff to update time a diff from the repository root +no longer relevant content has been removed +Closing resolved issues. +This is introduced after committing the +Standard and Web Configuration Removed list until we can get more details +ok I will add it today or tomorrow +Have created to track the new patch +The complete stack trace +If you add several components to the ART all will first disappear and then one after the otherThe way it is done now with prependJavascriptappendJavascript isnt very useful IMHO +Yep I think that just after the commit call the maybeMerge should called. +I have been building grails ok with a Xmx of M +Even with this patchAs additional info +Multi app version of +We dont call updateReaders in anymore +Martin you still can deploy the web project as a part of EARSo its not a bug +I did not take the DEBUG mode because it is quite easy to feed a from outside +Implementation for a saving initial components state is too complex and will be done at next release +bulk close of all resolved issues. +Hi GertIll investigate how to fix it from activemq sideRegards get resolved on Activemqlist jmxlocal still works of course though +Attached to this Jira is the results of iostat xd sda and mpstat +Pretty sure this one too is I can when the UI patch is integrated into the core +If you plan to remove the config once the bug is fixed wed probably be better off just reverting it +for the patch +fixed with svn revision ThanksIll leave this ticket open until I know whether I can enable the test +This patch fixes both problems +Simple modifications suggested by findbugs +It looks good to me +Dear JiraI tried to use the example orcherstration with jsr but it is not working +yup its a seamgen is a must fix for because it is so easy to do and it causes breakage of applications for demonstrations +Unless theres a patch forthcoming in the next day or so Im moving this to since its a minor issue +Thanks Anil that did the trickApplied +WasunCould you please explain what its working for you meansMy understanding is that if you dont create a product with Id with httplocalhostecommerceproducts you get the list of the products in the category +Is the original description here still an accurate guide to the approach taken +to separate authorization from other things that might occur after invocation +I strongly encourage to handle this in wicket +This patch does +IN MX +Im missing this issue in the release notes +In accordance with what you said regarding adding more code and files so we dont break other function I added this extra method in this file +How do we do this wo breaking compatibility during the upgrade is my question +Remove Patch AvailablePatch was tested and reverted because it never worked +Just discussed with Jason well come back to this post title of the JIRA +OhOur comments were conflictedI seeAnyway it seems that I need to work on current implementation ofI will try + +r +The patch was tested with automatic class unloading implementation +is that wrong? its definitely simplewarnings will likely be missed when the thing is running from hudson +It is a duplicate of which was deleted by mistake by me +Was this how it was used? or was it given as an argument to some GUI for running cpm things? If the latter which driver was used +I originally committed these files in error so they were not in the original evolved version of the patch was committedSending trunkapplicationssrcjavaorgapachegeronimoconsoledatabasemanagerwizardSending trunkapplicationssrcjavaorgapachegeronimoconsoledatabasemanagerwizardSending trunkapplicationssrcwebappviewrealmwizardSending trunkapplicationsdemosrcwebappDeleting trunkapplicationsdemosrcwebappSending trunkapplicationsjmxdebugsrcwebappSending trunkapplicationsmagicGballsrcplanSending trunkassembliesSending trunkassembliesSending trunkassembliessrcvarconfigSending trunkassembliesSending trunkassembliessrcvarconfigSending trunkconfigsactivemqsrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsclientSending trunkconfigsclientsrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsdirectorySending trunkconfigsdirectorysrcplanAdding trunkconfigsAdding trunkconfigsAdding trunkconfigsAdding trunkconfigsAdding trunkconfigsAdding trunkconfigsAdding trunkconfigssrcSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanDeleting trunkconfigstempSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsjettySending trunkconfigsjettysrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigstomcatSending trunkconfigstomcatsrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkconfigsSending trunkconfigssrcplanSending trunkmodulesassemblySending trunkmodulesassemblySending trunkmodulesassemblysrcizpackSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcplanSending trunkmodulesassemblysrcvarconfigSending trunkmodulesassemblysrcvarconfigSending trunkmodulesassemblysrcvarconfigSending trunkmodulesassemblysrcvarconfigSending trunkmodulesassemblysrcvarconfigSending trunkmodulescommonsrcjavaorgapachegeronimocommonpropertyeditorAdding trunkmodulessrcconfSending trunkmoduleskernelsrcjavaorgapachegeronimogbeanruntimeSending trunkmodulessrcjavaorgapachegeronimodeploymentserviceSending trunkmodulessrctestorgapachegeronimodeploymentserviceSending trunkmodulessystemsrcjavaorgapachegeronimosystemrepositorySending trunkmodulessystemsrcjavaorgapachegeronimosystemrepositorySending trunkpluginsSending trunkpluginssrcjavaorgapachegeronimopluginassemblySending trunkpluginssrcjavaorgapachegeronimopluginassemblySending trunkpluginssrcjavaorgapachegeronimopluginassemblySending trunkpluginsSending trunkpluginsSending trunkpluginsSending trunkpluginssrcjavaorgapachegeronimopluginpackagingTransmitting file dataCommitted revision . +Closing resolved issues. +Otherwise all the callers of them need to be modified making the patch even bigger +Reopening in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +Im starting to implement this and schedule to +The reason for this is that an application can set the to an application specific string and it does not have to involve the +Dupe of +Addresses stacks comments about multip ops against that were not synchronized + +I have a solution for ready that would solve this issue too +Does this test fail consistently? I ask because we have been seeing random tests fail during the build process +this patch removes the code which wasnt useful and dependent on +The proposed change is to perform deletion incrementally +If you take that approach please make sure that the error is propagated into the logsThanks +for the patch +This should be ok +We will evaluate traffic for another month or two and close the list when appropriate +Im trying this on Mac right now give me a minute +But this just wont happen the way the plugin is designed today +Thanks a lot Ivan for your opinion +It succeed +This is something we will look at when opening the design of to think Ill be able to offer that alongside the ability to load objects from the second level cache +Prachi it is done now +im marking as resolved since it fixed the only time i was able to replicate it +Just waiting to see what their status is +There are a lot of places where those configs get serialized and its surprisingly hard to change things on the fly the nodes generally assume things are static for the life of the process +Today there isnt any interface separate from the implementation and hence the dependency on the implementation itself +Fixed in r +I opened that bug without realizing that Brian had already reported this problem +BTW trunk builds fine for me if I apply this for the align attributeant version java version Eclipse +replacing the sort logic with a custom algorithms such as or query to +Moving the overall task to CR because I still need to do the envers tutorial +I think we should reopen it because it sums up documentation for all the functionality +Here is the patch for this issue +Ok to increase warning count +Good idea Ill add those two checks +Fixed in r +Thanks. +look good +I still think logging all exceptions is useful +Looks like used to cd to HADOOPHOME and libhdfs test uses a custom +Alex Which site should I log to +In Cargo land we have changed some coding conditions from Cactus were using column widths were no longer using the theXXX parameter notationMost of the code in Cargo is aligned to this except the deployment +i have written integration tests for annotation based rest support services xml based rest support in Axis +Yes this would need to happen afterIsnt the workaround a bit involved? For folks using EDG asking them to write their own startup code within AS internals is +Of course this should be handled at application level but at least there should be a way to know that the data it got is not whats meant to be +Thank you for the clarification +Writing release note and attaching new and noteworthy label so it is included into the release notes via the release note script +There is goodness in what you have created here and we need to see what we can migrate into SJMS while keeping it performantThanksScott HaraldCan you share the example code you were using for the integration +I opened for the missing offset bug +Minor and never reported in the wild bumping to the indeterminate future +the repairment to official website and thebluesky group an use the patch Qin Jian +This patch makes the creation go through the standard utils to make sure it gets configured +No by we I meant the entire community +However this is how is defined in Lucene so its something that we have to change in Lucene too +The remaining things are harder to fix as they mostly references Methods not classes +so what is it you like to actually have +So i stopped Zookeeper removed the data directory and restarted Zookeeper and the Solr nodes +Feel free to create a new issue for that +Jaigak if you take this project to the incubator I would be interested in being part of the project. +I tried making the change suggested in the patch but it caused numerous unittests to fail +Thank youI summarize we did following changes +I was trying to generate this file from our Axiss databinding we use to write the output +Im gonna spend a little time on trying to make a patch to resolve this + +for the fix +This patch will make that the current tool that converts the m repositoryto m work better +great thanks +to +Clean up all javadoc as much as possible in preparation for full release +There was just a mix between tabs and spaces in the patch which I corrected to stick to our coding conventions. +There is another public data set that breaks the old code but it is about MB in size +I uploaded the second patch after removing some commented out codes +Fixed r r URL to be the main Release Notes page instead of the specific release in EAP CP CR and EAP CP CR. +Im using debug logs and println outputs to check whether the beforeInsert event was triggered or not +You mean the datanode rightIt would be worth considering adding these as flags an HDFS fadviselike API on that changes the behavior +We might actually be able to contrib modules if we had a versioned coprocessor API they could hook intoSimilarly to this right now on the read side of things once we get the seekreseek optimizations in place we are going to add a new method to the filter interface so that any filter can pass seeking hints in +Only noticed very minor improvement but didnt try many cases +Nick this issue was fixed in JBDS for ESB why do you want to reopen it +Bulk close for +Only foreign key column names I have not been able to find a way to eliminate +Here are some more refactored tests +HTML included for review +But because and Maven differ on some points its ok for me to choose either one as long as its an explicit choice and not some workaround to make it easy to set the bundle symbolicname while the does that correctly by default no mather which groupIdartifactId scheme we useConsidering your comments I would like to suggest to use the artifactId you suggested and to remove the instruction from the bnd config since the s default value is equal to the artifactId youre going to use +think the problem is that you are only in and not in groupBtw Ill check the space and mark resolvedclose this issueThank you very much for the fast activation +So outer scope variable are visible within inner contained scopes other then the obvious global +postponed to to since we ran out of time for Beta and it wont hurt anybody +Both the messages size and the processing time are available within that +Please try to upgrade and see if this resolves your case its not please get more precise about what not working means +Definitely a bug +So they would constantly be struggling trying to correctly maintain that +I reported which has been marked as a duplicate of this bug still exists for you please test sitestage for your next release +Thanks Ben +The reliable transport is only intended for TCP based protocols ive reclassified this as a feature enhancement to add a kind of reliable connection for we know if this is still the case in +Thank you for tracking this down! I did not realize field number assinment was volatile across +At least we come to the same agreement +Hi AlexeiThe setter of the Interfaces property clears everything that was set before it was called so the order in which properties are set shouldnt influence this using new method. +I think this issue is worth while fixing before release +Theres no problem adding this to our vm startup command in the config +Instead it will get clarified in that such things are forbidden otherwise behaviour results. +one to do the offlining close and delete with another to plug the hole +rvs this patch includes my fix to generate the keytab only on nodes where httpfs is enabled +This is interesting +Then the release plugin user can turn it off for branches heshe does not want to push and can push at their leisure fact I find not pushing to be much more in line with the idea of git to Play local until confident push only when ready +fixed the problem for me +And if we add it as a default component that is always there then stuff like debugging output etc will list this separate componentIm not sure I see a benefit to this being a user visible component +That phrase should be used in the Bigtop documentation +At the moment it is difficult figuring howwhere to insert +Committed patch to documentation trunk at revision +Architecture document for the the doc into the patch so that it can avoid build failures +I havent worked on this +fixed in r. +So it could be simpler than Mappercontext in a bunch of ways also. +Committed as r +Thanks Tom! +ThanksHeath +Some comments Include ASF header in all files +Output from a failed run with configuration +Thanks Mahadev! +I have pulled the change to branch in c which came from fc on master +Will go through the code and your comments and update the code +And with caching disabled it is a horrible bottleneck since it synchronizes while it is reading from disk which happens on every request +Wizards for file and folder should be standard now they look different +Adam Can you please put in the guest os of the vm which are failing because of this? Are all the guest os failing or some of them? +The CPP client doesnt have a class to do that yet +Thanks Amareshwari! +Hi WernerIll grab the Cygwin port of subversion and will try to release the svn version of these diff files within theAlex +The code in has been recoded with reflection and moved into +In case of the IBM JDK I have been testing with the path ends with classic +Anyway I test the problem with Hypersonic provide you a backup of petsclinic where the issue exist +Introduced by in regressionArea +Iphostname +After installing I had two additional update sites in JBDS pointing to +Only possibility is to add the old libraries to the option pluginpath +Whoops relationships connecting TBLSSDS and IDXSSDS got lost will attach another diagram which fixes that +I was able to run everything successfully the only change I made was to the class to add the setter back you are not using JSF I recommend you update to this and give that a go +Looks like youre having fun I hope this is going to be a long stream of many small commits +Maybe someone with a Linux system and the Eclipse IDE could helpJens +Even if it was unusable why couldnt one + from me +and through the weekend +Committed revision . +Thanks for the notification on this +Ram I think it is because I didnt delete the assign znode +Please close the ticket +Bulk change of shipped and resolved. +How is this different from as per Arun I am separating out issues which are causing this failure +Bulk closing stale resolved issues +This applies cleanly against the Trunk version of SVN +Patch applied to +Geoserver sending a kb response instead of k at the expense of a calculation is worth it in my eyesmy c +We only need a +It will be even harder to understand the dependencies between different configuration files if people dont see and together in the same directory +Attached is a patch for maven because maven uses a different version +I attached a log file +We will need to do something better in the future but for now I think this is good enough +OK no worries +I will try the patch tonight +FYI theres an easiermore efficient fix +Just as a note that last set of patches to the files isnt reallysuitable for use against the svn trunk youd want to turn them intopatches against the files +Hi will try to get more information on the wiring being null do not have an easy bundle deployment yet which I could give as an exampleNevertheless since it may occur IMO different behavior depending on log levels is a bad thing and should be avoided in any case especially uncatched runtime exceptions +But and use +The patch looks good +addSynchronization during beforeCompletionIs there a really good reason for making asynch calls? If so then this raises issues for the asynchronous addSynchronization client as well as the implementation of beforeCompletion +The entry and element declarations in were out of order + +The proposal here describes a way to share an object within some kind of scope scope using an explicit reference with very simple functionality +As for a review of the XSD One could try to create unit tests thata use a couple of interesting example documents as input which we know are validb push these documents through an XML parser in validating mode which employs the the sample documents pass the validation we have some assurance that the XSD captures the intended structure. +It looks like decom is broken +Let me revert +Thanks for reporting this issueThis sounds like the same problem asCan you please check whether the latest SNAPSHOT solves your issue? Thanks +is the scheduled release for dropping JDK I am not sure I like the idea of binding to uuid +Uploading a new patch with Dmitriys feedback incorporated +I did not getcould u please attach a stack tracethanks am maked this as fixed if you still have the problem please reopen the issue +You can do the same if you want to add more detail to the body of the response +What it should updates then? Belive that should only load its data +If we did awaywith that object this could be a problmem +Lets keep the Jira for tightly scoped comments about the issue at hand and use the mailing list for more general discussion +I will try to have a look. + +Build is still failing +Just makes it so we do the ZK transition and remove from online regions even if +Ive just updated with the a propose naming for artifacts +I would try to finish the review these days and update to you. + is happy +probable design would consist of implementations that implement the various R client protocols Servers that understand the R protocols which forward requests on to Some way to use R service discovery to set up these clients and serversIt is possible that it is prohibitively complex to implement the entire protocol at which point wed have to have further discussions +The Attachment contains a test case with ENABLELIGHTWEIGHTDIALOGStrue in the and partialSubmittrue in the launcher button for a test dialog to reproduce the errorIf partialSubmit is set to false it worksAlso If ENABLELIGHTWEIGHTDIALOGS is set to false a dialog launch with partialSubmittrue can be done only once +Actually Lohits patch is mostly unrelated to the bugs description and is about getting counters from the job client +Shouldnt it show +bq +The static deployables are because they require modifications of the containers configuration you could detail your use case on the list so that we can think of how to fulfill it to the best? to change the status to wont +Username mrfredsmoothie +yetanother versionstore scenario +DTP have that option for their database connectionsMarking as since this looks a lot like a bug to users IMO +I have clarified this with an explicit check in which now throws a descriptive error +Add CSV export to the result view +Committed at subversion revision +Thanks everyone for the review of the patchesIll submit a patch for shortly +Very much appreciated +These dont seem to berelated to the patchI am going to commit this patchThanks for the patch Ole +Using a new map per request for the response context is exactly what did so that should work +For both the failed test cases the log message is Timeout occurred +Right + +Silly mistakeP +This screen actually incorrectly lists my services as not working +New label added to in r by Anil because it is part of framework. +This gives rise to the following conditions and the proposed behavior of the system for each Both MAXMEM and MAXMEMPERTASK are set to invalid values This will result in the current MapReduce system behavior +Sounds like a good approach and I will give it a try tomorrow +Nor is this patch meant to be a substitute for regular backups of the filesystem image +Yes this is incompatibility +Id like to get this into so reopen to make it easy to into is a prereq for an automatic should pull in r as well +Patch is OK +Have you something particular in mind +I wonder if this is not a JVM bug +Now it seems to workCommitted in Rev +We are currently doing final testing on release so I will check out the patch once the release is finished +works but lacks test cases javadoc +Sounds good to me +I cannot reproduce that that with the current latest snapshot from the builds +this issue is already fixed which build are you using +I could do that but right now it depends on Spring not sure I like that in xworkAnyway I will update the wiki with some information on the new element +were you having file system problems on +Thanks Daryn +This really isnt a bug per say but more of something that I just have not had time to implement at this point +are they logged somewhere +I just stumbled upon more occurrances of this problem +And merged to the branch in revision . +I think that rather than creating it it would be fine to just print a nice error message if the directory does not exist +Thanks Devaraj and Owen +All tests pass +Why wouldnt strip default to false? + +when I look at the zip and run buildr there it claims the compile taks uses javac classpath homekristian +This patches to keep the last key seen and throw an exception if an earlier key is seen subsequently +I have committedSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaSending javadrdaorgapachederbyimpldrdaAdding javadrdaorgapachederbyimpldrdaSending javaengineorgapachederbyiapireferenceAdding javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestsmasterAdding javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestsmasterAdding javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTestssuitesAdding javatestingorgapachederbyTestingfunctionTeststestsderbynetAdding javatestingorgapachederbyTestingfunctionTeststestsderbynetAdding javatestingorgapachederbyTestingfunctionTeststestsderbynetAdding javatestingorgapachederbyTestingfunctionTeststestsderbynetAdding javatestingorgapachederbyTestingfunctionTeststestsderbynetSending javatestingorgapachederbyTestingfunctionTeststestsderbynetTransmitting file dataCommitted revision +is the same table as +How about moving all imageedit related classes to a new package? If it is not easy to do how about moving the methods you want to use to a new class +See SVN rev . +Now that ihave some time i m still stuck on how to get the schema out of the source class +Ill see if I can drum up some relevant HTML source too +The ideas is having something to grep +I wonder if this is one of the first applications ported to YARN. +Pull request coming shortly +This can be closed nowAs you have committed the patch recently in rev RegardsAshish Vijaywargiya +So I propose that we resolve by doc fixes for +I was able to repro this after running the test back to back for days on my box +So now everything is included from the other file but everything can also be overridden in the current file including the event list blocks though those will override the entire event list from the other file which is good because it is the only real way to remove an event from one of the lists +I want to refactor both ONI and OCINEI for readability and clarity and such gotchas before is out +Hi thereIf the id doesnt change we might run into a problem with locator registration +Scott let me know if something is unclear +unable to replicate this locally will investigate unable to replicate this locally +Tests all pass although I dont yet have a test for this particular case +On x two tests failed with server errors during shard splitting +Seems like it might be +for the patch +I do have copious logs for the time of the outage on the order of megabytes +Canceling patch until we sort out all raised issues +This is ready for review +Failed test seems unrelated +The turn around on this was impressive +Even though the ZKFC is currently only used by HA these classes are all in Common and thus the descriptions should go in instead of +AFAICS your suggested patch effectivly turns off encoding +I know we had this issue with Catalog listeners as well and in the end came up with a special exception that would allow listeners to legitimately throw exceptions but in other cases like NPE the exceptions are caught and execution continues +Can we deploy this on a live cluster +I did not think of equals +What are the changesIs it required? Does it work better with KeplerAny reason to update +That patch was committed to the trunk on but its easy for that commit comment to get lost in the rest of the commentary on this issue +Do we have more crypto related stuff or in other words would this be a or going into lang or codec +A link will be enough +After some testing the booking application seems to function as expected when the session times out but unfortunately my application still has the issue even with the latest from CVS +It works fine on the latest snapshot in the trunks already fixed at the latest snapshot. +Incorrect patch uploaded previously +Unit tests are in place to continue verification of this correction. +Fixed in trunk and branchx +Please let me know if you have any problem while running the for the patch and changes +jar to be deployedlibtest +Hi Markus the attached patch also includes your boilerpipe stuff I am reverting those parts on the patch and trying it out right now +Ok Ill put something together +Done for client side +Thanks Peter and Neil. +A quick fix is attached passed twice with it +This would do +The noSnapshots rules is deprecated and replaced by the rule +Ok this is getting annoying +Issue may already be resolved may be invalid or may never be fixed +My best hope is that theres going to be another release of or later which Hadoop group would more or less assert as a stable at which point Mahout could switch dependencies +I think we should rename it to Common tools for jBPM +Hi AndreasIve done some more testing and have come to the following conclusions When you mvn install something into the local repo it will always be a SNAPSHOT version When you mvn deploy something it will depend on the settings of the repo and which version of Maven you use +I created new issues for setting up hudson builds and for resolving the remaining issues with the to a limitation in maven the retro modules have to be built individually instead of as an aggregated project +Please see that is the option we are going for for have some tests which were disabled because of this issue and still are failing +Upgraded to Camel releaseRevision . +this should have been fixed already +Filter +Closed upon release of. +Applied branch and trunk. +After this everything compiles and eclipse is ready to go +In order to subtract two Integers you have to create a third +David what needs to be exposed byif needed we can add that +This makes Castor easier to use since mavenwill download the jar youPaul Spencer +I used to compress the image +Good point +attached patch based off committed! +Review the file for detailed instructions +never mind this may have been a misunderstanding +Fixed as part of. +Closing streams in finally would have a bad properties +We dont like to use Dojo just to handle the onload stuff only +has been just released +The orignal fo has been added with description My fo generated rtf has been added with description The generated rtf uploaded patch does work for me +Also when we get this listing on the wiki we should link to it from this jira so it is clear +See HL where all the MSH fields is now added to the exchange in message so routing is much easierCheck out the HL component page for a routing sample with this new feature +Fixed on and trunk +On the other hand it says do this to enable JNDI storage +Now rows is running way slower more realistic +Ive created +The same for all Converters and ListersexportShema produces an app that will create the db schema at deploy timefar less java output some more xml outputdate types supported removed from unfortunately the jira notification system were down so didnt notice this before +Verified +investigated this further using test irunning single placeXNTHREADSXSTATICTHREADScompiling dynamic libs on final executable on failure with GC ok without GC valgrind reports no errorusing same dynamic libs from final executable on ok with close of all resolved issues as part of closing items. +We should add options to control if the backlog tracer should drain on dump truefalse whether to drain the queue when dumping retain period N period to retain only the messages for the last N period eg last minutes +Thats parsed from the URL and getting the title from the document could be difficultAs far as the community and howto tabs go I noticed those and fixed them a few hours ago +Globbing means give me any data that matches the globe +and then itll start rising again + +Reported as working in so I can edit the fixed in attribute +closing Dojo related tickets +But if append is false I have to use a classifier so why is classifier deprecated? As soon as classifier is removed then my assembly will break +Removed config name from closed as resolved and released +Not sure why yet +There is no need to sync on the leader to access allPartitions +I just ran the steps and didnt get an outdated project could someone else try it i think maybe this is the previous comment still m closing this task as Ive not been able to reproduce it. +So all issues not directly assigned to an AS release are being closed +Trace problem with the last patch so further trace to several classes to help problem diagnosisAlso in the method removed the duplicate setting of the a trace point ni the removed the unused import +Heres a patch which adds headers to all java and idl files +Will look into it +Trying to figure out the reason +then youve got a bug to fix +Great! I will keep contributing the best way I can really glad to help. +Did you guys get an? I see that there was a build failure and then a successful build and as per log it should have sent mail to please add yourself to or if you are a member of then you should have got that as well +For the new runs use a build which includes commit beffdfaacbfebd +Is there a conflict between using plugins and declaring dependencies in? Or between plugins and declaring other plugins in? +Yknow what I think is the best way to handle this just have an API that you can call from an action method to validate the model + +In addition unfortunately to date weve not managed to implement tests yet so this is another reason to further support your patch and its backing for integration into the codebase +Are you proposing we issue a single compaction submission when streaming is done instead +Full crash report RemyI do not know +id leave that to copypastebest text editorformat sql +I get and the collection that it inherits from is not orderedAny news on this issue +Committed for in rev Thanks Jul. +Please try it and report back if the issue still exists +User did not understand optimistic locking very well +That seems to work as then I can open up the file and the graphics show up +But basically it looks like I cannot replicate your problem +So perhaps if we identify some reasonable use of this method and we make a release then this could be revived there in the branch only +The second patch implements the comments above +I personally have no strong feelings about this particular problem anybody else +Ramgopal are you still seing this issue? If so could you please provide additional details about your environment and steps to reproduce +it should give you some respite to hear that axis has passed the standalone tck for saaj and jaxrpc AND jee tck for jonas and geronimo +I was using Hadoop but I deleted that and installed HadoopI know this isnt the right forum for Hadoop but if you have any suggestions it would be appreciated +So far the following API have moved fromimport import import import toimport import import import +When I say +eddie can you take a look at this or pass it to somebody who can +Fixes an issue with applied under svn revision . +nothing in Violations viewWell where can I see log for this to see if something is wrong +At this point Im more you have discovered my plan to keep you out of core! Actually Id like to keep my own code out of core as well +Do we wanna update as well +I think this bug is not minor +Just updated it and committed. +Ok patch committed to trunk +My pleasure and two questions +duplicate. +Trivial fix +err by language constructors i meant language constructs +Removed that disclaimer revision Any discrepancies will still need to be documented in the release notes +Looks much better thanks for adding the test case too +bq +Would you like to submit a patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Applied the patch provided for the +will commit if the tests +for a Daryn I share many of the same questions +I will look further into the possible causes of this error and report back with my findings +Thanks! +thanks Namit +What types of objects were you getting which were unable to be cast? Also what line was throwing the exception? Is there a sample PDF which we can use to duplicate the error and use for tests +The beans wizard is now complete as it is though the drag and drop to anything else than the editor will be very difficult as drag and drop is not cleanly handled by the Eclipse platform +It seems we had this problem sinceIll pull it in if it gets done in time otherwise itll be in the next point release +Please reviewThank you +This patch is ready for integration +Table suggest Ajax utilizes dojo heavilybut it is not the inputsuggestdojo does not have anything similar in its core lib +This is invalid defect +Hi FredericDoes this issue still exists? I have checked with the notify sample with MEP specified and unspecified +part of release. +attach my patch +I have no reason to believe it wont work on other platforms +It seems Jenkins didnt kick off a +Stress testing has been performed andit looks like the release fixes the problem +Could you post the patch as an attachment to this jira and click on the Submit Patch button to start the build + +Fixed in r. +After that I try to modify other file with symbols and Ive get dialogChange to encoding does not change anything the behavior remains the same. +i thought it might be considered all the same problem but it doesnt matterif we change the test as Adriano suggests it fixes Korean problem +Also please confirm this is still broken in +The logging package has been moved to the common project. +Attaching unit test + +The test files are actually checked into the source as resources +applied patch and tested locally +So it isnt an issue and current approach is much faster +OK thanks + The VDB deployment is fine Can connect to Teiid and VDB via using Teiid Final jdbc you can connect from then what assertion that you make that the issue in the Teiid? Did your debugging reveal anything? I believe this may be due to some eclipse classloading +Checked in a thread local storage fix to r which makes a big difference for +Making the datasource not participate in JTA transactions is not an acceptable work around +port of protocol test as found in the Python tree +This has been fixed for a while. +JaquesIndeed I used Idea Interllij and did not sustain myself from pressing the shortcut to reindent everything +Actually does exactly what I want that is the increment goal to be specificSorry did not see that this was a duplicate of before now. +This would let us provide a registry +I take your point about the exporter +But if you like feel free +Actually on its been observed that tasks using an invalid token will pound on the NN every second on one cluster this happened for a monthThe JT immediately issues a token renewal and then uses a timer for future renewals +Agreed +Fix for Hadoop I was too hasty with the third patch +I am spending a lot of time on creating an assembly but that is not really the part where maven shines +Claus however brings up a good point +The disadvantage of this strategy is that it creates a wholesale import of unspecified environment variables to your project which can effectively reduce build reproducibility +We cant support the past forever and Im loath to change the behaviour for AIX users who have more compilers +but in we could change it include some large collectors for these people and the whole thing would make sense +And one more thingThe commands should be run at the root of the checkout and not in the studio. +Which version did this happen on +Arun is looking at it +Fixed in the issue now that it has been resolved. +attached Design Ashutosh +What you need to do is to find the reason why the parent node has been removed +Simple +I have added you as a contributor to the project +I changed from the subversive to the subclipse plugin which seems to use a different patch root by default +Closing resolved issues for already released versions +Heres an interdiff of the v and v for reference +Lets close this as since was reverted. +Attachment General XSDany XSDfloat XSDpositiveInteger XSDstring nillableI have attached the log file tooAnyway Ill see what was wrongThanksChinthana +Please fix that and upload a new one +The tests in other branches define the order of execution +Does it work when invoked from a regular application server thread +For more security I would suggest you impose another directory called +Dave can you please fix the owner of that file in the RHEL tomcat rpm build? could you take a look at this please +Im going to add a warning about not using this for and then change to feature request to support the alternative pooling +Carlos I dont really know this mojo +I think logging is ok as is +I had to leave out some stuff that I worked on at the end because it was incompleteIm unassigning this one so someone else can see what is left to be done +Feel free to select the skin and figure out what changes are required to fix it +Thanks Alexey +This is a fast form of assigning ids to records but arent continuous +and delete the +It does not support other HTTP verbs as well such as OPTIONS +Also not directly related to this ticket but can be parametrized just likedidnt know this +I suggest creating a separate issue for the outer functionality as this behaviour is now enabled by default +Please attach a sample project +HiWhat do u mean by returning the image to the user? Dont you try to draw the image and have a look at the mapAnyways my work is to draw the map I am almost done if somebody could hep me it would be really +I would verify that each segment of the path is a workspace project refresh check the next segment + the fix is committed as part of another JIRA. +Thats great news on all fronts! Just one correction +Regardless of change of documentation however I dont think it should be possible to actually trigger a scenario like this in the first placeThe system should protect the user from thatI also noticed that in this case we have RF +Fixed in r. +yesinvalid +Full test run is still running +I suppose you are using Maven +Commited as adbdffbdbIll marked this issue as fixed +The integration test shown in section of the user guide extends not +At least I think we should add description about this problems to VPE FAQ +Your promptness is REALLY REALLY appreciated +I cant see a situation where having to add the same field twice with different flags is good from a usability standpoint +with Wendys comment +Odi All I am saying is that might be a better place for proxysettings as is getting a little too messy IMO +could you rebase +Merged fix to and committed revision +looks good +Tested with and the same issue remains. +Tamb tenim dogma i doctrina +Why didnt it fail when ant core tests was run? It should have AFAIK +Since it is unusual for an MEP to be transacted it seems quite inconvenient and to be forced to explicitly set this flag to false +The patch has been appliedPlease verify +I can see the value of consistency +Ill wait to see if I can get someone from Yahoo to look at it too +Make Description easier to read with proper formatting +Unfortunately the simple webapp with just this code behaves precisely the same wayIn testing with as the backend I experienced no issues +We could add the new API to the package +This heartbeat log was helpful at least to me in the past while debugging production issues + will address this too +yes happy to look at this upstream +faceting is now slower for the term field and slower for the term fieldOne way to regain more performance is to implement some kind of stateful iterator over the values in the field cache entry instead of looking up by ord each time +this works as design because of the different of Geronimo and lifecycle. +Since the release candidate has not yet been cut this would seem to have no impact on others +not a bug user didnt realise that addPackage calls fire at the end. +As to RPMSOURCEDIR I was just copying how install +I wonder if SOAP is supposed to support circular referenceAs far as serialization supporting circular reference there are a few models out thereFor example keep a list of already serialized objects and do not serialize again +Do we know the upstream bug ? I dont see any links or references to it +Ill post a patch soon +Yeah Ahmed make senseThanks Ahmed +Log attached war attached +I have also created for difference but I dont know where is the entry for it +Bug has been marked as a duplicate of this bug +If we have a custom widget to install to jira we should open an ORG jira issue +Implementation should contain mechanism for limitation of loaded records into memory and this mechanism avoids OOM exceptionJournal file is quite big and it cannot be attached to this JIRA +Passes tests +here it is +Applied in r. +oops still working on validation removed this from branch and tag +The root cause is G does not shutdown as before +Assigning to Shelly for approval +Chunhuis patch for TRUNK with minor renaming +as far as i know nobody is working on this requirement has been addressed by. +We have nested associations for which field updates are causing a load of all related entities +A quick grep through the source says this is only supported in the and +defer to. +bulk defer to t fix believe it will result in less clear code than doing nothing +Not found this issue with latest runs +Patch to allow apps to make extra reservations slowly over time +Checked in patch and verified fix thanks! +almost a year after the issue was first reported +My understandingexperience is if you have an Authorization you should also have a capture if done within business daysso my first concern is why you not getting a capturehow have you determine the capture failsI suggest you put some in the to see what it is returningthen work you way upBTW which gateway driver are you using +So it really means releasing the plugins releasing the parent updating the poms to the newly released parentits doable but not right now because isnt quite ready to be released. +The package name from the IDL and the generated java code cant be different at the moment +Id rather not have a back and forth in jira +Do you want to batch insert by default? Or present users with a choice +This really helps clean up our dependencies and makes it easier to develop. +Marking as closed. +I have a suspicious feeling that changing to use the shapes API will not fix the scaling issues based on work I did back when this class was created +Moving integration test related issues out of any particular release bucket +Closing issue fixed in released version. +Doesnt do what you need +Part of a bulk update all resolutions changed to done please review history to original resolution type +Pig still hasnt changed since +I rewrote the whole class and split some code into new classes based on the final sourceBut there was another potential point of failure In srctransportaxis there are operator overloadings to read from and write to a channel +xsd fileLine should readCopy the newly deployed blueprint +The only explicit supporting import of schema is through EBWS and it does not sound as if this is what you are referring this work previously +Good morning +I found the reason why there was no installIn our setup the build extension for archetypes was not defined in the parent of the multi module build but in every individual child +What operations do you perform before you the web applicationIt works for me when I do a login read a couple of items and logout again +This patch fix the problem +This patch removes the jdk variable from the javadoc targetsWith this patch I have successfully built the javadoc in the following environments +Use cores update a Slice to inactive state and check on another core for the update to show up +Done added launch button in job instance Javascipt to set job parameters on click on previous value in job instances view +Bec is already working on this assigning to herI think youre already done on this Bec but please let Jimmy verify the text when done +so you created an app that imported and created an EMF to use in your own Activator +Instead of reusing PermissionVERSIONMNGMT I opted for adding a special permission for this PermissionREMOVEVERSION in order to avoid confusion with other uses of the PermissionVERSIONMNGMT +Only the http target is filled out for now +The Vhost fixes fixed it +only after this was committed did I start seeing it +for the patch with committed this to the Federation branch. +Otherwise we can add a new fictitious component GROUPED ISSUES and assign this component to this type of issue +Irrespective of RM restart Oozie needs to set as for the launcher job and for the underlying MR jobs +java files we have +If you do it in parallel that might only hurt the performance since the disk will do seeks all the times +Matt fixed this by normalizing the Strings to remove platform specific line ending issues +Would that work +The directories are created as part of the make install command +But how can I get this as anbinary distribution +AaronThis particular case that triggered the patch submission is for Microsoft SQL Server +Thanks for taking it care of it +OK Claus so what do you think about creating +Seems as a simple patch so Im giving my here +I tagged the new specs for this case on the master branch ddcc +Kurt if this does not fix the issue please reopen. +note this also include the references from properties in the version storage to nodes the version storage +Infra would appear to have completed all of the necessary work just a few other items remain for us to tidy up. +Updated patch with Teds think this read lock will be needed for SPLIT +Opean eartarget +Ok I understand thanks for your quick responseIs there already a way to the shortcuts through any or will it be available in further builds later on +David are you okay with us taking on the documentation for Logging as part of Solder? Well include a chapter dedicated to the understanding is that Solder will be the window through which we expose Logging to the community +Pavel go ahead and see if this would be a quick fix but if not lets table it for now +Fix merged to trunk. +all is fine! Although I dont like the of I am fine with the fix +Note the version of Saxon I tested gave the same result as XSLTC +It also has some wonky behavior around committing changes that I cant seem to figure +Open up the Preferences Groovy Compiler page +Maybe should be called or something similar? Anyway outside the scope of this issue but worth mentioning +Does that mean we should create separate classes for the mapred and mapreduce with static methods that match those +Made most fields final and removed the setters of these fieldsIll commit this in the coming day or so +It applies to anything that extends +They have tablet definitions startstop compaction events and mutations for multiple tablets + +bq +So now is having the exact same algorithm as +Paragraph is working I only need to document libraries it requires to also did some changes to its design according to the new CSS and removed hardcoded loginpassword +Committed the a patch to trunk revision I havent found any other places where we generate code for row templates so Im resolving the issue +Will the Resource annotation be supported as well +During upgrade namenode will first wait for current requirements to be met +Having that configurable somewhere would be nice +I dont see any difference in performance with and without the patch +Reattaching the file whitespace issue is resolved. +Support like this can come from core frameworks I think +Aaron Thanks for the fix I have added your mirror to the rotation and it should appear soon +I dont actually understand whats supposed to happen here +Ive also added the test case sent by Jari and a new ant target to run tests just compatible with ruby to build with the lastest master changes +The presentation of the info is bound up w the composing of the info +Closed after release. +Deleted +Also it may result in a smaller size taxonomy index. +See. +So I checked in a fixrI verified that the form posting all looks reasonable +You can use the FTP wagon or switch to MavenBrett +These tests were running pretty close to the timeouts in my environment even on Mac +Not sure why debug calls wrapped in if would cause performance hits in an enviroment with debug disabled? I assume in your production environment you dont run with debug turned on +Im for reopening +Thanks Owen! +grep and sed will do +Check in the izpacklib +git diff ll be happy to after you respond to my other comments above +Please reopen the bug if it persists. +Most likely Flash Builder not Flex SDK issue. +Verified at r +Revision . +Favoring high IDF terms or moreaccurately high scoring character position spans and favoring fragments with high term diversity are not mutually exclusive +Moving out to until we have a contributor to work on this +Sorry I forgot about thathtmlunit comes with a newer version of a common +Please correct me if Im wrong. +small bit of js required for ie +Attach the crash log please for iOS +installable plugin war with admin console Export Plugin till we verify that this is fixed on + is the sandbox issue where the proof of concept ideas go +This is also an issue for the header especially when used with + intentionally has lots of tests in it to cut down on our spinning up cluster instancesPatch looks good to me +This actually has to be fixed in Redback but we need to leave it open here to avoid duplicates. +This is needed to avoid object marshallingunmarshalling if there is class loading issue. +Address for passed +I think this is a good point +The rule above really has rules into it +Which browsers do you use where this problem happens please +Add hbase runtime dependencies to +Yes typically you make a field persistent and then bind the parameter to the persisted field. +I think we should just turn this off entirely +So can you please describe exactly what to do here +Thanks for the patch Peter! +ops +Fix all the issues not just this one Darran see for more information +fix where clock is constructed for a local Blocking for mutate methods reflect updates shandleadd +I cannot reproduce these and still not received other info so or create new issue if more info comes available. +Per instructions from PM RHEL is no longer supported as a dev environment for running JBTJBDS but is still supported as a server runtime environment for AS EAP building JBTJBDS on RHEL is a WONTFIX +We want to set a custom header response decorator based on the one uses but as I said the runs first +Set blocking for because I believe this will break the database file format +I should have seen the patch! I have taken this up as a project +Patch against trunk +Committed to and +Here you go +This looks very good however Leonardo has to take a look at it since he is currently the one who does all the core releases +small one liner removing the revision . +Beware that the line number where we change the token type to text may be different in yours my version includes changes for that Supun wants to review before they are applied +thanks + +and yes has to invoke the custom onselect handler when the user presses enteror call it onautocomplete or whatever the name isnt important just has to be documented +So controlling these parameters one can control the extent of recovery +Are you able to run unit testsI guess I will try with new unit tests that just got checked in and see what sorry can you make one more patch agains the latest tree +JAXRPC tools issues may be fixed by community contribution +for internal central repository cleanupI am required to do daily release +Kristianpls patch or provide details about the bookCheers! for the inconveniencehere is the site updated +Hi MaciejWere you able to add a FORRESTHOME declaration to your script in order to get this to work +As before I this assuming it does not break the build + resolved issues +Thanks +The stopmaster patch had a conflict with v +This kind of information is better in a tuning guide +Attached as +Get rid of recursion +Thanks VladimirMikhailPatch applied to LUNI module at repo revision rPlease check that the patch was applied as you expected +The application code can go into directorystandalone somewhere and we can have it use some of Nick Faizs work with the standalone build +For users who are writing in order the performance should be identical today +Fixed in +This still includes the generated thrift classes +rename issue +It looks like this may have been due to a botched upgrade from to revision +Sync up with latest trunk in v patch +No progress has been made +Can you test again? I know that is kind of try and error is very unpleasant but as I said I cant reproduce the error on my end any more +We dont have a javassist change yet that is safe to bring in +for the patch pending successful Jenkins run for the latest version +I fixed the issue this morning by changing the scalatest version to +I committed a patch for this +To use filter the will need to subclass to +We could introduce a class that encapsulated all the configuration options in one class that could only be set while the is being configured either in a constructor or before after +Thank you Andrew very much appreciated +Alternatively and preferably they should be final +This patch adds stability and intended audience labels to classes and interfaces that I assume will be used directly by users +i consider this to be part of which should address the modified more flexiable namespace handling +Unit test for lot of things were changed +The way it is used it can be easily replaced with a single method that takes a varags Object and returns the array +Thats why Im surprised that you still have do some tests again +If you import it then should no longer keep its copy of the code +Sorry Im a little bit busy at the monent but it is not forgotten + +Nope we dont have to Warp Alpha brought critical issue which our tests suffers fromI wouldnt upgrade RF unless itll be critically needed +It seems to me that this one is coming out from s during batch callsor flush +Asokan +What is for +Per default does not check for presence of HTTP HEAD response body +no longer using column TS as this will make things backward incompatible old values have higher TS and cannot be overwritten +in is missing this jira description +I certainly cant see the downside to rearranging the query but its a little disturbing that were seeing different behavior +See +Those dont suffer from the same memory usage problems as they are applied one row at a time when the rows are indexed. +enough Im linking the associated issues and closing the issue thenCheers +This one has been sitting in Patch Available queue for a while +Log file +Attaching patch to Switch to JNDI multiplexed Cache region factory Provide an explicit cache region prefix +Its s +My bad not correctly used searchYes i start studio with debug +I committed this +Issues are set to go to the issues ML for the component I setup +Useful when you want to play with row cache sizes for individual +bogus issue now. +This patch incorporates the first two comments +I should clarify that bonding works today as long as the IP of bond is where IPC comes from as well eg if a DN is configured only to listen on the IP of the bond +Fixed on was told the fix should be in the yet unreleased but downloadable RC +on both a tiny tiny update +Already present in master +It would be nice if Ilya could doublecheck the article +Closing this issue as Invalid for that reasonJuergen +now contains the constants for notifications and is completely deprecated goodwhy was job priority completely disabled? replaced by the new threadpriority inalthough ive not fully though through the concept of job priorities it looked useful for me first if a new job is added to the queues which should run faster than other jobs perhaps from the same topic and even running in the same queue +addressed NPE in +Sure that would be crazy +Closing +Stupid question did you register the as a stream in the step +on EAP with removed signing data +I dont +Seam Ear web project deploys to server wo +the link for Getting Started would become href +Add it to your and Cluster journal +Ive just committed this to trunk and. +Metadata map now checked on trunk +Basically this means better performance at the database layerAlso our DBA would like to map the JBPMDrools persistence table name to different names in our environment +I will check the rest of the code to see what can I do +Alexey I see differences between spec and RI behavior and consider what situation is unclear +Minor patch updates for x +As you mention we have to introduce a semantic change so we should do it rightBTW make sure to test for and +Closing it. +Problem is that only our editors and views would now about it thats might be a good feature request for eclipse but not for +Committed to branch also on Stacks up on IRC. +got that one again on clean installation of was not JBT but Developer id date +bq +Ive seen it before too but also cannot reliably reproduce it now +Nick we already have a separate issue about raising an error in case of JAR Id say we should just resolve this one +Carol Frampton added normalisation of the paths using in r. +Committed to branch thanks for reviewing +something wrong with formatting in changes for +thanks for the contact address thatll do fine +Carl can you take a look +Hi Jacques it should work with trunk although the error message given to the user should probably be differentA Product Store has not been defined for this ecommerce site +Prior to this we were working with a partial set +Maybe Ethan fixed it during his other s possible that its no longer an issue +The previous patch was getting quite hard to apply so I updated to the head of trunk and regenerated the patch +added as well + unscheduled issues to the next release +So we might end up with the IDE resolving the version differently than the cmd maven being used later for building +Unfortunately this is not easy to test in a runtime workspace +No its difficult to say what the cause of this is +Fix released in. +Personally I would rather see more rather than fewer functional names as they make it much easier to find things +Here is how things broke +That doesnt make sense to me because it will render HDFS unusable might as well put the cluster into safemode and shut it down +Sorry projects can not be renamed not without exporting the entire Jira instance as XML editing the XML and then reimorting the XML file which wont work anyway as our data is too large to do thatI would suggest in your case to use the key AAPOLLO perhaps or other projects use something like AMQAPOLLO etc +GraemeThis will work although there should be an easier perhaps automated way either thru plugin or something like that where user does not need to make the changes in to run using different grails plugin has mechanism where the default setting can be changed when glassfish or some other server plugin is installed? Otherwise it would be kind of for ayone who wants to use something other than the default jetty server +Fixed the MB mistake +closing as dup of +whether we do an RC or not after this I dont knowTechnically if we dont want to its not required +Do you know whether LIMIT means return all rows or return no rows? If return no rows we would need to issue LIMIT bigconstant instead +I think this patch is a regression +So heres a little addendum that fixed things for me +i dont know why i use a german comment thereD Jacqueshere is the patch without special Charcaters +The comment just says that it is the metaserver port +I just committed this to trunk and +Thanks Jarcecs contribution and Abhijeets review +Elliott Want to say more on why the jitter is needed +What is your kvm config? Can you debug this test commit fixes the dependency where we assume credentials to by rootpassword +Since nobody commented or reopened the issue Im now closing it. +Fixed in Joram that sounds great +Any comments +Reschedule if these happen to get fixed sooner +I guess the Single File Code Generator for Visual Studio doesnt get invoked on a Rebuild All which is odd +Patch w last round of fixes +merged +The spec file should default to using the in conf +The instance has been restarted several times to free up the file descriptors without changing the design documents +That makes contributing extremly difficult as I can not start from a base where all tests are green +youre right sorry +Ill also research what we can do for standard JMX in the Spring seems there is not much we can do about this in Spring for the time being +This patch passes a rpcTimout parameter to RPCgetProxy method +We will produce a patch to fix also +bq +Thank you for the Review Ashutosh and thank you for the patch Navis! +Metadata requests are cheap +Moving to were trying to narrow the scope of +The implementation files exist in the camera plugin and theyll stay thereThe issue concerns removing the geolocation code out of and into the exif module +Git status output +Thinking about it more however I think this is probably a good change as well +The patch needs to be the patch as invalid as it no more exists after is commited +Looks like a good compromise for now +This looks great to me passes my testsOne improvement Id like to see is for to not serialize the call twice +Attached patch fixes thisplease review and applyDumindu +Thanks for the fast reply Carlos! Sorry! silly me I had it written down on paper correctly then failed to commit it properly +The patch is known an affects only two linesRegardsm proposing to change this to WONTFIX +An active detailed test has been running that verifies this +If the exception were to accept a collection of that would be nice tooWould another alternative be to extend the interface? Maybe that has custom lifecycle methods to rid Exception throwing? +To deal with that might need to separate that part out a little bit so as to be able to construct a really lightweight comparator that is constructed each time but wraps an object that has the main sorting configuration done only when needed per the current approach +Set to blocking as the changes for this issue affect the JSON passed to patch One minor nit we might want to add hbaseooziesqooppigzookeeperhive to the rpm list +Forgot to address point +If so we should pass the servicename and portname to wsgen when they exists in +thanks +The patch applied +We usually create new bundles when the version used is different +JD if you look at the above posted logs you will see that it was failing with We are missing some rowsIm guessing it was +of whether or not all this functionality makes it into the next release the modifications made to and should be added to the release +At some point you would evaluate the performance vs another node in your cluster look at your performance and make the decision +Therefore there is no test for commands +safe than sorry +Oh yeah sorry. +will probably use Maven to generate the manifest. +hopefully fixed in codebase for netbeans crossed as I still cannot reproduce the given error message that starts with Compilation failure. +would be interesting +This ensures that there is a change in the znode versionNow only if the version changes let the RS carry on with its further steps of SPLITTING +fixed with svn revision +Otherwise there will be a serializationdeseriablization problem +The documentation is already updated in +Have tested and will now map the connector port for each of the port configurations when using binding manager +Latest patch is rebased against the commit for +I cant reproduce this condition either with or the latest version or Roo are you using +Test to reproduce the issue described in the description +I am marking this fixed +that would work if this were a filter +for applied on trunk and in Developer StudioVersion Build id Build date +A workaround is to copy the plugins andor views to your own reproduce with latest SVN head as of rev +looks like we are not using the latest WTP with the patch +Forgot to start the workflow +Fix a bug regarding ignoreCase in the attached patch +Thank you for the input +Do we really need this new option? How about change for only append +I disagree +Its not commented at all so use at your own riskFor du in the above chart I used the accumulo shell du command +I am more or less convinced that we can stick with prepared statements +does not require loading of entity classes of course +However I cant see that this can be considered as any kind of real release artifact as it isnt self contained and Im not even sure that it is buildable by itself +bq +eg kafka start +I dont mind removing it too +The pool can set the schema once per connection it creates whereas here you have to set it each time a connection is +can you please post the whole stack trace? AND are u using Axis? its probably fixed by now +Configuration changes made to vhosts +there was a Smooks patch release run off last weekend +Do the same patch on your old AMQ JAR and it should work with the newer Camel releases +Or that deprecation has to be read as in the final solr just use and ignore the warning but remember that in the next version the things will change +Adamam I wrong or do you have already recently fixed this issue +Applies against r +Also caught a problem which Ive fixedRewrote it to use a static constant and test whether adding a would cause overflow +The best way to write tests for them will be as part of a system test thats using the jar as a bundle in a real container +Well I think Spring is a much sought after framework which has a lot of appeal for many users +I think it might have been related to synonym fields tokenized using +The fix by kafka is there so I guess the patch is not needed +Thanks Nathan! I committed this to trunk and. +However it appears that the API is starting to pull away with what would easily be reverse compatible +commits seems to be what weve settled upon +The receiver doesnt seem to restart or even detect the disconnect which could be a bug in or in the CMS code +Is this not the exact same issue as? That issue is fixed but youll either need to try an incremental build or wait for the next release to get the fix locally +closing as issues are fixed in +I know what this isI introduced the change that caused this. +Why cant we wrap up the discussion here? Im currently on a design which makes the snapshots in fact not in the case of concat +I have fixed it on trunk. +RobertI agree +That isnt a valid soap fault message +Changes has all the types required for the protocols has service definition request and responses required for implementing +Verified in JBDS to write release note issue now that release note status has been set +I think this should be closed. +This one is against the current head revision I will post a different one for +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Here is deletall implementation +Refactored the code in response to comment one from above +Do you mean the provider endpoint maybe +Attaching and with modifications to the DROP VIEW topic +passes under both embedded and +will try that in another ticket +This makes it possible to run an in memory instrumentation with Emma and have your code coverage tested +Please confirm whether or not it works using findservice with findservicecategorybagkeyedreference rather than findservice work around the issue omit the findtmodel portion of the query then process the result set via code +Fixed since revision the wrong issue +HTML output file included in the zip file for review +i cant check it in myself +Works for if the is set by upgrading +Fixed in commit daefceaadbfbb +Splitting the query is not a viable option +not just the asset serviceAt least it is better in with friendly +Arun Ping? Comments on my above questions? Should we really raise the defaults +Tested +I opened it after the fact +Added a test case +Referenced be decompressed into examplesjms directory of beta distribution +patch to fix the looks sane making menu faster it also unblocks the ui when showing the url works for me +Thanks +Need to investigate on two fronts fix to comply with the contract should the contract be altered to ignored nonexisting properties +There is no rule that says you cant encode A to to avoid URL encoding ugliness when used on binary data + +Should we make this a separate method in so that other languages can also add this kind of functionality? I would not make it abstract since some languages may not be able to do this +Can be closed +If you consider this is equivalent to the Seam way you can close this issue +So when will the netcdf make it in? Andor can this code be written against a generic interface? In we try to only use the from to not write against any particular data format +reassigned for review callback approach has been implemented although fisheye only show the older commits that used setters +Still the mep printed is not correctWhere is set the MEP for the server +I wonder why the code in was insufficient +Sorry Gagan I didnt get a chance to review things this evening +I was seeing the same thing until that commit +Proposed patch I think its the simplest solution nuke the confusing ctors +The rest is mainly testsThanks Avner +Yes I use a function for the other fixes I was going to add one for this edge case too it just wasnt very high on my list as its ugly code BTW I was reading up on lisp advice apparently you can append to an existing lisp function +Thats by design +is this the ruby equivalent of +Rev +Compile error in renderxmlI did manage to fix all problems +One thing makes no sense to me is why we are not seeing any produce output events firing even though the session clearly has write interest and readiness flags set +language checking is Docs Build checked +All the tests get passed with this Axis versionthanksthilina +Ill open a separate jira to change back some of the logging. +If you get any trace please provide i will take a lookI dont think there will be traces aspectj support is totally broken in trunk and it hasnt been fixed yet +Attachment has been added with description Patch has been added with description List of updatednew will take this on since Andreas has mysteriously vanished +The identity deadlock happened minutes after starting +I suspect there will need to be some optimization to get some of the resolution methods work as lookups and to clean up trailing spaces and other things +Fixing the findbugs warning +It is probably my fault +Closing this issue. +Current behavior ignores any property that has an invalid name and that is what is being exhibited here +A select record type function +Brian I am quite willing to include this patch with final release but at the moment I am not able to replay the problem even with the test case you provided +In any case I wont get around to implementing until the end of next week probably +I dont think its trivial as the might go into a recursion +Testing a backport of related patch as first step now running iterations of +Personally may feeling is that the version stuff should be hidden from the user +Weve decided to work around this +As part of this work I think that we should remove from source control +Seems odd that the initial size of the buffer should be so important so long as it is long enough and is plenty long enough for localhostportIf it really is more efficient to get the exact length why not save the port string temporarily and allocate the exact buffer length? Is that faster +committed to trunk +Now that we send strings for the method names instead of ids this jira is now possible +bq +Committed to the trunk and branch +In this patch I dont use +Any updates on the new skin? I think it still havent been implemented on the current BZ even if the preference skin is pointing to +Set correct fix version +it facilitates creating a separate CG pass for IPF CG +classes +Im looking forward to so that highlighting can be more efficient and doesnt need to duplicate the query interpretation logic anymore +Added in runs fine with this change +I think the problems was in the get method +I will test the patch and commit +patch to add a test case to cover this believe the original fix for this issue is actually more correct than the modified fix although the modified fix might be considered a reasonable M workaround for +It works fine with release +The code is not committed +The client will stream data to the datanodes directly possibly in chunks of K memory buffers +Changed to forward the package target down into the srccontrib dir rather than run the packaging of contribs itself +Maybe you will have better luck +So upgrading the Hibernate components to the versions below wont break the compatibilityCore ANN from ANN +Alan if the server is the then it obviously is a HTTP server +Logged In YES useridguess that this is a case for are basing our JAXR implementation on UDDI +Apparently Oracle doesnt follow the JDBC spec closely there +There are cases when the sessions entity managers are accessed outside the context +backported patch to +Im going to commit this +A DN will successfully start with a failed volume as long as its configured to tolerate a failed volume A DN will fail to start if more than the number of tolerated volumes are failedThis is the expected behavior with this patch +Not sure what is going on there +scheduled for +It required me to hook into in a way that was not planned before but well there is a clean way now and the reload test from before works now +If have to pass m explicitly to get iaw +So you are making WMS requests using the tiledtrue parameterYesGET geoserverwmsWIDTHHEIGHTLAYERSbfpSTYLESSRSEPSGAFORMATimageFpngTILEDtrueFORMATOPTIONSantialiasAnoneTILESORIGINCSERVICEWMSVERSIONREQUESTEXCEPTIONSapplicationBBOX HTTP I have turned off memory usage and that probably fix the problem for me but not for others +HI Hasan Really appreciate the help +The patch for the unit test class +Actually it only fails on stack! Looking into it +rename PREALLOCATIONLENGTH to MINPREALLOCATIONLENGTH to reflect the fact that we can now more than MB at a time +Will commit when tests pass +was this happening on windows +The Jaxr implementation in will use a Saaj based implementation +But as far as I understand from Sergeys blog continuations provide a solution for something jms does in any case +its a great help to all +and have not the pb. +Falcon fixes up through bddfafbfdadeedaAnd SDK fix acdacdafecfeafd +That was actually the reason for creating a separate interface in the first place +This has been fixed time ago by with the layer grups +The embedded assembly codes of getting base frame address is dangerous and have poor portability so I remove them from the sources +Im also getting this error +Ill send you some links via private emailinsteadFor what its worth if we could decouple the issue of the headerfrom the issue of copyright notification +I asked Eric about this he was thinking about indexing the logj debug logs in parallel from each machine in the cluster. +Should be relatively easy to special case to just not load BF partition summary in batch mode +Simply wrapping sasl in protobuf headers would be sufficient to address our near future needs +If you really want to see it you could do a simple flush or hack the code a bit etc +Usually they work fine +made a few changes Please review if it makes sense I will verify in a real cluster +Then the is processed just the same as regular jobsAny comment on this approach? Will work on unit tests if it looks good +The code to run the terabyte sort competition +Please review it +UIMA has problems parsing type description files if Xalan is on the classpath +I would also like to simplify locking in the cache by getting rid of the for the partition +Let me know if there is anything else I can do +Wont fix +This approach sounds good +Ive been looking with a debugger The problem is that needs the which isnt being set in the the was having the set then it should work just fine the logic looks I couldnt find where the is being created or where the set would be called on it so I dont have a patch for you +Perhaps other improvements could be made but Id prefer to limit this issue to that improvement if possible +I did a little read performance at the beginning seems not much impact +New Version Rename and friends to as that is what those essentially are +It seems to be fixed in the master branch +stack Todd whats the jira number for showing all configs in hbase contextThanks a lot +There might be some build issues with it too +Implemented as part of. +This will need to be done before is released +I believe the best way to coordinate this is to have someone take each component and make all of the necessary cahnges to that componentThe drawback to the way this got started is that different contributors are overlapping each others work +Although it does set me back on the issues I have just created but thats okay I can remake those patches +If Affects Version and Component are not set the issue will be rejected +potentially you try to quote that private column namelike private +Looks like a bug will take a look +Thanks Julien I made your change +it would be nice also to support is. +bulk defer of bugsminor issues from to +We can afford to live with the finalizer. +Im attaching the XML dump of those +Ive asked Richard on the mailing list if he could provide thisAfter that we can turn off all the infrastructural items. +If there are no problems I can do a GA is causing regression in TCK +I dont want to write foobar all the time +Can we say the is final? I dont think anyone disagrees when I say the feature would be definitely useful +If Im following this discussion correctly I think we should consider choosing a common API that is more similar to the Java API or at least includes the slightly higher level functionality that it presents +Work around was to add the project as a shell script project and then continuum can deal with the checkout and you will have to specify the path to maven with the lifecycle states you want to build +I only could reproduce this error when trying to register after a initialaborted anon checkout +I think only the bin file was missing from the main repoMay I suggest that the infra repo is made until its time to delete it +I tried reducing it to Mb but still freezes +You could however select +Committed revision Thanks Bernd +svn revision +The repeater is showing the contents of a that always only has one item in it +Having a good UI is a plus +Hello Manojyou wont have any error message but you will not get information of column family +I didnt think that the RM supported blacklisting +I think this is resolved with the new config admin and file install versions that do the persistence themselves. +A user will have to explicitly ask for in order to use them +Hi Andriy I agree it is not a high priority issue +Please fix it +Will backport to +Investigate how to reproduce the i cannot reproduce so we will resolve it +Release +Oh so looking at the code a bit more it also appears that if an error does happen the thread that writes to the log dies and the event queue then grows without bound so technically there is a memory leak involved here too +Juergen +Konstantin tested on his Windows box and it worked +Cool +For KVM its not possible to live migrate between incompatible unless you are on full software virtualization in which case the module would never have loaded +Ideally these would both use the same code to build the classpath +Throw and catch unsupported exceptions by user manager and connection. +Attachments with todays date contain changes required to implement +If not you receive messages over and over until you run out of RAM and the test dies +Isaac can you document the above +The default behavior should check whether the codec is splitable like other input formats do. +attached patch to add CDATA section parsing +Closed with fix field unset +HiThis is the working patch without slinginclude tag +Patch applied as expected thanks! +We save bytes when typeXMITRSP and bytes when typeMSG! +Oh this supposed to be my morning issue +You can see that even just restarting RS can cause the issue since I disabled all CM actions and actions and region move in actions +Is that a from you as well RakeshI was also wondering if we want to commit this to the branch too +Currently the constructor serves this purpose and the Builder API supports the case where defaults are desiredPerhaps we could better document this? The generated constructor might include javadoc cautioning that no default values are set and that the builder should be used if theyre desired + for the patch +Thanks Ning +Ithink there should be no more dependencies +Attaching test case update and proposed fix +Patch applied in trunk and Beta in to fix title forgot to close after reopen +Oops there is another big issue you need a watcher if you want to get the info on session expiry +Closing this issue due to lack of activity +It is after the parsing and currently we do not annotate logical plan with line number +Unscheduling due to inactivity +Thanks JohnCommitted to trunkAlso asked Arun if he wanted this in he said yesCommitted to v. +Its not even possible to check whether an explicit environment has already been set because will always be set. +But the above handling seems cleaner +I agree that this would be unnecessary complexity +To execute test suite for +Ill try to find a way to embed this into Print dialog I dont know has Eclipse print preview functionality should research how is possible to implement GMF has a required functionality +Nice Mike! Definitely what needs to be done and very cool +Both of those files arent actually used anywhere +patch as work in progress of my experimentWill look it later but have some other issues to work on to make it intoAlso I want more feedback of a more generic solution as how the exchange is currently being routed +Applied proposed patch +Octavian we are not looking to drop we are just trying understanding what the real world situation is +I get it working using but therefore I had to change the character codingin Mozilla after every request to +bq +However our samples jobs use two base classes that extend the functionality of springs dependency injection tests +If thevalue is zero then exiting online archived log files files are not deleted +rev rev brrev brrev trunk +In Rev +Test harnesses are not cleaning up after themselves too many files in tmpNow resolved +This was already done +Just saw that the MD checksum on m repo is different to the dist folders +It is pretty clear that v is the best so far +Attached example code shows how the deadlock is triggered +batch transition resolvedfixed bugs to closedfixed +tests that partition setting works as intended +It works if either using or disabling filtering for the fileset in questionRobin can it be that you use for the plugins and that your tries on Windows simply used a different version of the plugin than on Unix +looks good to me Ill check it in +Please use instead. +HelloWould you please try this patch? Thank youBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Thanks Koji! +Hi KimI didnt change the release id in the version of the release notes yet +Im still getting the Keyboard interactive required supplied password is ignored prompt +should be fixed as well on branch +in other words Tapestrys instance would be present but would never be invoked because your custom instance would be the one visible to the rest of the application during the render. +Slight improvement to patch which is simpler than the prior solution and fixes the reported issue with the first patch +It seams no one is working here at the moment +Committed revision . +may provide a better way of implementing this +Where server state wasnt saved prior to shutdown +One concern I have What if the file is actually gone for some reason? In that case it seems wed never stop retryingIf you go up in the file youll see that after we looked everywhere I currently dont have a good solution for files that are missing completely +The most critical ones for me are the ones in the procedure to recover a ledger +I dont have the ability to do that. +Pavol what you see is a combination of several problems the memory option Xmxm is not correct if you use Xmxm Forge starts correctly because of the wrong option the VM halts with an error message that is not appropriately shown to the user this needs to be fixed but is not critical IMO +This sounds like the approach that will alleviate this problem to a certain extentOption Option but we can do this as a separate JIRA +Promise! +iow feel free to critisize +Lets see if we can change the to only auto wire by name instead +what is the work around exactly? OVERWRITE +Punting issues from RC to final +an IT no idea how to create an IT +ad +However additional burden on the system admininstaller is to rename the automatically created to after the installation is finished for a node +Patch applied +Thanks the default assignee has been changed but existing bugs still use the original jakarta assigneePlease could the assignees for existing bug be changedThats not something I can do as far as I can tell +We had to fork with this fix and we think that because many are using the sharedlib feature this patch should be appliedIt works with both or lib scenarios +Yes youre both right! I misread my own notes on the bugIm writing tests for a real fix now +The source build now only contains source code as it should and there is a separate archive with just the binary dependencies +one more file new patch to test the build +I must also document the properties syntax to override default ll try to do it tomorrow and Ill close this issue +Thanks for patch Tudor and review Michael +It works fine +I did try mending the patch myself but looks like there has been a lot of refactoring so Ill leave it up to you +That has been done. +Is there an easy way to write test case for this +Iterative patch Javadocs changes to HashWhat about the API signatures? looks still they will expect default value right which will be unused now in definition +Im working on a new compaction strategy that would use a manifest +Please try the updated patch +Override in a few methods of could be added +Committed the patch after reverifying it works in beta +Patch of work so far +We should have a clear message when two nodes are started with the same have tested this and this is working having two on the same cluster would show a clear message on the logs. +Is this acceptableProbably the best way to go is to make it configurable but isnt in yet +this patch takes care of issues with speed +Note that the preparser is only able to obtain referred graphs but cannot completely generate parsed objects +AXISPRINTFUINTFORMATSPECIFIER +Looks like we dont have this fix have documented this in the release notes +Marking it as resolved. +Yes thats a solutionBut still the RI is doing something not entirely good hereregards you think that the solution presented by Mathias will be integrated in the source code +I had enough time to go get a soda before this was fixed. +The only change to the patch is depending on htrace version as opposed to +Thanks a lot Jacques. +Since the PR shows up as an issue Im resolving this one. + for the to trunk. +So every time an unresolvable class was found even if the compiler had just tried to resolve it another resolve attempt would occur +Tried restarting eclipse with just the editor containing that file open +Updated patch with the following changes Changed the dag submit to accept a configuration for the AM +Is there some way to hide the annotation from Jackson without modifying the entity class files +One benefit for patch v is that it allows different components to make decisions based on whether JVM is for master or region serverWe can revisit this topic when more config parameters are set in this regard +Also resolved by downgrading jackrabbit to +There wouldnt be much point in checking if a storefile has the next row or not +you cannot decide to create based on a +a minimal not configurable implementation but should be enough to allow reverting problem is that google bing etc do not use an accept header +are written as a byte integer or as a variable vInt and whatdrove the decision behind that +In that case theres no point in providing skip logic in the Processor +I think that was what we planned on the meeting +Thanks Roshan +I created upstream and updated it with a patch +FWIW I dislike using the word token as it might be mistakenly associated with text analysisGreat point +The goal of course is to build community around it though and eventually graduate it from labs to the incubator once you have a community +I need to check how far that is already integrated with our rollout plans though otherwise it might not be that useful +This issue was fixed months ago and I need this fix very urgently MarcelThanks for the reminder +Yes that is not needed anymore and will fail on float splitters until the bug is fixed +Sample project that works. +The changes are not yet verified +Well review it once again when is more stable +Hi we have a load test setup under qa which fires requests concurrently and I have not seen this error +Thanks for the quick turnaround! +In general I think we may want to traverse the graph only once hereI think you are right I make it twice only because both json and gv need this logic and I extract it into a separate method +Still the patchs smaller than your patch +Well whack them one by one +Bug has been marked as a duplicate of this bug + for meThanks +No JVM crash is observed +assuming issue is now resolved +Updated final patch for new link analysis framework +Simple test exhibiting described for the reportThis is a problem with inner bean name conflicts in the object cache in some scenarios +do you need anything else +Added test hudson after adding the suggested test AnkurWe see this kind of issue with tests from time to time +Was this already completed +If it is strictly an internal API then the argument about it never being used makes sense of course +there just isnt a test for itthat seems like important enough behavior that we wouldnt want to break it to fix this bugEven if matching on overlapping subspans wasnt an intentional feature of the fact that it currently works and the documentation is silent on the issue suggests to me that it should remain supported +OK I see that there is configuration to be done prior to packaging to the newly generated project and untaring it to the deploy projectWe are getting thereIll see you on list + +Saving to Lucene indexes is a different use case +All tests ran cleanly on the second try +Hi GianfrancoI believe the JBI spec requires that message exchange semantics be identical for both sync and async use cases +Will this also break the ring persistence we added for +This is not an MNGECLIPSE problem but a general Maven problem and is a result of the xdoclet plugin extending the antrun plugin which I will have to look into +Any implementation that tries to do a check previous to the insert is likely to suffer from this issue which is one ofthe reasons I still think it would be better to always to do the check after the insert and not try to do any duplicate prevention locking schemes that are likely to present unintuitive concurrency issues to the application +Ill open another issue for changes +Ive committed this and merged to +If we fix then this is probably unneeded +NOLIMIT has been removed +I got distracted by some other issues but should be able to focus on this again +Looks marking this issue as having a patch does a committer fancy applying it +multiple threadsYes it would be really nice to know if anybody has run into this kind of issue while developing S applications +The same things was under inplaceInput Page for me +Yes I mean the unit tests will not fail if the setting is not done +So we did not prefer and empty region to split +AwesomeHow come I dont get that kind of response for commercial stuffThanks to bypass Enumerable logic for Rangetoa when both ends are fixnums +I pushed a change that defaults src to file and will change it to only if the xhr succeedsHowever if you say that js execution actually halts this likely wont fix the problem +I would like to discuss here a little bit +to commit +Ive not yet found how this can possibly be +And it still doesnt unload classes after redeploy +I did some tests with larger data set and the results are consistent with what we saw before +Keep forgetting that the code is in MR. +Yeah I thought of that but +Thanks for the commentsWhat version of designer did you use to create the first diagramDid you use relevant upgrade scripts in activititrunkmodulessrcmainresourcesorgactivitidbupgradeAbout priority maybe it should not be minor but I think it could not also be one of the only two critical issues +I would recommend to debug your code an see were the fault is created and how the setDetail method in it is invoked in each case +The code under the srcit directory is meant to be run by the +Im probably being dense but I dont understand what resolves means in this context +What is the problem if this is not fixedIts generally considered good programming practice to remove unused fields from the code +I am unable to replicate at the code the only way an NPE is given at is if model is null but on there is an explicit null check +Im happy to trade some performance for code that its easier for others to maintain providing that this tradeoff is understood +If you have requests with an average latency of milliseconds for example it would take seconds to execute them with a single rpc thread +test case attached +Ship as a utility for resolving a jar path based on a class name based on the available classpath +Btw could you use svn diff instead of diff to generate the patch? For new files I think you could just bundle them into a zip file and comment here about where they should be placed +Patch adds a check for sequential seqnos and also a unit test that shows that we were failing to maintain this property before with empty hflushesRunning unit test suite now +It would be good to know how this relates to the Pool effort +I figured out +Thanks for fixing that howler Ole! +I typically do the build with JDK but try my best to keep the code such that it continues to work on +Marking blocker so it doesnt fall off the list. +Also you can see how getting peer ID is ugly today we need a new API to find peer id from the task id provided +I found a bug is not thread safe but is used in a MT context +Also if we do that if the znode is changed in the assign flow we still go with the deletion and will abort the master if deletion failsJust my thought +But anyway +I certainty think a JSON based approach will offer a lot more flexibility than the current as methods +println method is not actually defined in Writer so the println that is added to Object class by Groovy is invoked +Im not sure but committing this may mess up Todds work on merging Fetcher and Fetcher +It could collect statistics record them for later user us and pass speculative decisions up to the JT +Dont commit this yet +any ideas what to test here? i dont see a way how to test captcha or validation +Mark +Verified with the release notes in staging. +Closing all resolved tickets from or older + +was a little autoboxing issue +There are some known issues with reverse DNS resolution that we are working on for +The metadata is completely defined by the serialization plugin +sxc open fine in myopenoffice on macosxIll attach the generated files if you want to compare with your versions +bq +Updated patch which simply removes the zero cores checkthrow this mirrors current trunk code +My monitor resolution is at x in a monitor so i really couldnt ll have a fix for it in a while +Implemented so fr are integers big ints floats and doubles +Do you think the rather then a should be the default? I think we want add a configuration parameter that leaves the weak reference the default unless the parameter is explicitly set +Im having the excat same problem with I meant version +on commit as well +Sorry the minor change I made in the patch that changed to is wrong +As far as I can tell all the raised items have been addressed in the commits +Document the rearranging hereWell do a clirrjardiff comparison and decide on renaming later +Here is an issue for the documentation noted in my earlier comment today ysteins use case is supported only if a rewritten chunk is exactly the same length as the chunk it is replacing +If you are able to come up with an example that reproduces the issue please attach and reopen +Corrections improvements are welcomeI just committed the revised Javadoc of the Header Cookie and classes +I have the same problem on my MBP +Forgot in the previous patch to remove the unnecessary call togetheadervalue I noticed afterwards that we have couchhttpdheadervalue therefore avoiding changing one line +Closing issue due to irrelevancy WRT current status with descriptor +Your design looks good otherwise +We may be able to revert this now that Ron has fixed Remoting +That is a known issue in +NOTABUG +Im looking for a volunteer to help increase the quality and test coverage on this before we commit to trunk + +So are we ok on the name +ChristianYes I worked my last patch against on lucene +Use JDBC instead. +I hope that before Xmas ill have time to fix +This system test has been instrumental in testing and I see how this can be useful in testing more Kafka bug fixes and features +Basically although it might be compacter writing a process in which conditionalguarded transitions leave nodes is bad practice youre trying to do too much in one node and it wont be immediately clear to anyone looking at your process later what is being done +So we have to make that change +need to change uima in addons projects +Its not very likely that this is going to happen properly the code is simply too hard to completely verify through unit tests +Usually only committers resolve or close an issue +Duplicated by +There are also cells with encoded characters +I guess they should be made available +bq +Should be fixed in r +David do you want to take a stub at it +I see that your CLA arrived +The order by has been removed +This is an untested patch +we will try the new version +Successfully Pack successfully updated with this new feature no need for. +Im doing a unix find command on the entire dir and I cant even find a file by this name +This gives us a good start +Fixed tabspace issue I noticed when applying to patch branch +This assertassert is there for when in the future the currentversion changes so that more cases are added to this method and thus this assert should remainNot sure we want static imports used in cassandraimport static patch looks ok other than that +Resolved while upgrading to jline. +a for this patch +we are limited in terms of what YUI calendar understand please do not use the shortcut formats. +I did not test constructor +Verified in IR +Suppose that you have avg ids per transaction +Labour Party has its own entry. + +The patch looks fine to meJust to check that Ive understood correctly removing the wont affect the performance negatively because the underlying stream gets its data from a byte array so buffering the data in another byte array on top of that wont improve the speed at which data can be read +Found a similar issue instead it stays in PENDINGCLOSE state for the state flow is different zknode was created by RS instead of masterSo the general scenario +Nice find patch to and though the JIRA said it was merged into this was one of those that was lost when became and was recreatedRan across it on and noticed it wasnt fixed so I pulled it into + +The restart job was failing because the step configuration had not set savetrue +The patch overrides the superclass find method and avoids an NPE by checking for a null delegate during for branch and trunk +Matt +Each HibernateJBC combination needs to be configured +Any other ideas +Marc Schoenefeld Red Hat Security Response encrypt the database password thing is there for passing silly security audits +I think that the NFS solution looks like a very good idea +After that this patch will be review and commited I think that also the issue can be closedI start my work from that issueThanks new patch permit to use the paging onoff in the searching of products and after that this patch will be applied it will be possible to export all the products linked to a category to Google Product Search without surfing all the pagesThanks patch allow to export all the products to google product searchThanksm closing this issue now that the basic stuff is in svn. +It contains more test cases covering various combinations of modifications as well as previously untested modificationsI also extended the with an additional method that allows the merge to distinguish between an externally addedremoved node and a locally moved nodeMartijn can you please test the patch and report back if it works for you? thanks +fixed by r +A good host is a machineport where zookeeper is running or will run in the future +I believe is descendent of +The remaining failing tests are in. +I think its not possible to satisify everyone of the time +see above re lesser of evils +Not super concerned about this but on first skim it looks like it could clean that loops logic up a bitAlso Im wondering if there should be some sort of throttling on how quickly the scan for databases to compact runs +Bug still appearing in andWhat is the status on this issue as it is already open for a long time +fixed in r +Here there are one for corrupted ones and the other for archived pathIn between there are lot of deletes and exists call +Would it be possible to make a plan for fixing this +Hello Davythis problem might be due to Ant not to looks like it Ants bug which is fixed in Ant can download Ant from to check whether this solves your it with ANT but unfortunately same problem +That should fix the problemOleg +table but rather using the methods to put and get regions in and out of the map +This seems to be fixed in +I am not convinced either +Ive manually checked the files are not used by any process after iteration is done +It is quiet possible that is related +This appears to be my day for batting +Brian please reassign as you see fit Im not sure who is the right person + is a blocker if you use axis or higher +I will work on the unittest for this one soonLet me know if there are obvious problems with this approach +It is the external WSDLWeb Service that it was waiting for +Ive attached the new patch to the JIRA +I was leaning towards as wellThere is one issue I am worried about if we use czxid as the index into the bitmap it will keep growing over time as the recently created nodes will have a the latest zxid as the value +Yes this issue has been marked as resolved meaning the fix has been committed to SVN but has not yet been retested verified as being fixed in trunk +to your suggested layout Steven +the latest patch looks good to me +My assumption is that the properties were added for specific reasons but without corresponding comments its hard to tell +So changed the generation code to the one I used +bq +This should work +It seems that Micheal was packaging the jar inside this jar and using it in the project +So for I marked this one as optional for me +It has an expanded Javadoc like Enis was asking for +Gary what you think? It adds new dynamic metrics per Region +Nop! I guess thats itThanks for the information these version conflicts drive me crazy +Patch itself looks good +if you do I think that this uses only one argument +a group of letters will have only the first filled with the fill color +It implements the ignoreadd capability rather than the as described in the initial descriptionAny and all comments are appreciated +This sounds messy and probably is but the ASF certainly will not publish code of other people who dont allow it to do +and my production application too testcase with changes on attached +subclasses one or two compiler classes but I couldnt find a difference that would explain this t that make it a issue then? I mean is it not possible that team looks at it first and if it really turns out to be a groovy issue then move it back or split as the case may be? Just curious +We are going to support OID macros Im just saying that users should not use themRahhh +For use plugin must run before this plugin +The patch also adds transient identifier to attributes of and making the logical plan serializable +The download page needs to be updated a bit +The patch has been appliedPlease verify +Also I am not sure if the clean in mvn clean test is required it deletes the +Patch file +If any changes are needed I can still make them on the trunk and if a second RC is needed they can get into the releaseCommitted patch to documentation trunk at revision + +Fixed at revision +Thanks to Dyre and Kristian for contributing data. +the list +I found that if I create a closure for focusLost and remove the the errors go away +Thanks Harsh Do I need to add parameters in and restartservice namenode +I was hoping I could get by just deleting the old keys and writing the new ones +Would be good to figure out what is going on +Ah sorry I can still reproduce the NPE +Patch against which implements a value for MRU cache setting for freemarker +The URL for xircles is not shown on any mailing list m subscribed to user and developer but it doesnt show up in xircles +proposed patch which validates the return values of all pthread calls and only sets the priority iff the policy is set to SCHEDFIFO or SCHEDRR +Should be verified in gatein master and reopened if necessary +Good to hearJuergen +added in an which is just a writable wrapper of +ok i will take a deeper look +well revised the code and you are right +This doesnt seem to be an issue in trunk for this bit of code as the custom no longer delegates to the parent thus no stream is opened +There can be a good suggestions from othersI will explain how we were addressing similar stuff in older versionsbefore rolling the logs we were having size check if verification logmax entries check +I will log a separated JIRA if there isnt one already +Does the benefit come because the are sequential or because they are monotonic? If the latter then would utcrandom be as good +The same shutdown problem appears in the following tests +Reviewed +I sure will after some more local testing. +Same error hereubuntu tomcat java geoserver pyramid plugin gave all privileges to the pyramid foldersGS was able to create folder and move tiles in it but thats itWhen I click the save button to create the pyramid store pointing to the pyramid root folder it showsCould not list layers for this store an error occurred retrieving them This plugin accepts only File URL and String pointing to a fileHavent tried with tomcat yet. +Attaching + In case of Registered user selected state list is showing on page load which was entered by user while creating profile in both the checkout process and manage it was showing all the states of all the countries but now after removing states screens and changes in both the JSits dispalying selected statelistSpecial Thanks To Rishi SolankiOnima Onima RishiPatch is in rev +Shouldnt the patch be against? Or is this only relevant to +Thank you for the patch Roshan! I have committed this to trunk and! +But it has some flaws +defer all X issues to +I still see a reduced amount of heap available and a lot of remote connection threads +Will be included in the next release +hi Michaelis there a workaround? I have a large app that I want to switch to Railo but this is a major thing for us as we use stored procedures a lotthis happens with stored procedures that return a single result set +DohShould be fixed for and +The addition of nullptr to the language does not mean is no longer a valid null pointer constant +So I would agree to get rid of this warningcheers Edell. +Why did line of change from o to o +Merged to doc branch at revision Will file another issue if it makes sense to remove the Parameters to get topic and make a more general statement about catalog parameters in the topic +You wrote that some parts of the code are automatically generated +whenever is ready I will migrate it. +the Queue browser queries the actual server queue for the messages whereeas receive acts only on the client theres a window between the message arriving on the server queue and reaching the client where it will be visible for browsing but not for depends how you define available +svn ci Sending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigSending containersrcmainjavaorgapacheopenejbconfigrulesSending containersrcmainjavaorgapacheopenejbconfigrulesSending containersrcmainresourcesorgapacheopenejbconfigrulesTransmitting file dataCommitted revision . +Isnt it there? The comment says Use of ORDER BY and FETCH FIRST in a subquery +Could I ask you to attach complete stylesheet and input documents necessary to reproduce this problem +I am running the new tests I have uncovered a new problem when playing around with this +Attached a new version that incorporates Toms feedback +Did you change how is handled on the HRS? Does the HRS still run the safe mode thread for slowly ramping up compactionssplits on startup +bq +Committed patch to trunk recommending using the Ant propertyfile option to the trunk with revision +But for our upcoming release were stuck with the current construction code alone +Removed the useless bin directory +It looks like Mono does not support framework unless you configure it with yes +I like the idea of keeping the Struts deps as minimal as possible +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the codebase +Closing this issue with the documentation fixes provided by Uri +this shouldnt be specified in rackspace and overriding it to nothing is a bit awkward +Thanks RaviNo I did not actually build a package and run the daemons +Hence assigning it to her +Did we ditch this idea? should we close up shop +There seem to be many places to put something like this +Ill get that change in place +Thanks ImranFixed in LUNI module at repo revision rPlease check it was fixed as you expected +Poop +Although master reading ZK to make decisions may introduce additional delay especially if master just watches the nodes and builds up internal state instead of querying entire state periodicallyRegardless does having different coprocessors on the regions of the same table make sense? If user has logic relying on coprocessors for some data querying or correctness it seems dangerous +Attaching patch that implements most of what was proposed +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be in +since we should not have any Seam references in Perspective. +Looking at YARN and might benefit from similar changes +See to infra and svnpubsub +Minor tweaks for a and version +Pretty simple toString for so that the ui gets the numbers again +Thats because I think that only updates AFTER the evaluation and the evaluation never ends so it never updates the user interface +Bulk move issues to and check and returning an error message +check the router iptables rules for acl rules +Closed due to the release of Tiles. +Is there any reason for thisI think its just not done yet I havnt looked at the merging since I started playing with the I think most of this stuff becomes moot with though this will turn more into an API overhaul than an reopen time saver +committed with the same fix for the nd occurrence of the RSRR in +It is just for safety if you see what I mean +Your patch looks like the same design as the original one experimented with on this JIRA which doesnt take into account security etcI still think the design I outlined using over UNIX domain sockets is the right approach +Fixed in trunk behavior now changed as discussed in the issuenote in VS Microsoft is automatically added if there is company name registered is blank. +the issue will enhance usability so Id like to get this in CR it is rather simple to implement +Source code for the microbenchmarks +I created a patch which I just uploaded that changes the index and welcomepages into jsps +Just out of curiosityI could probably look in the documentation but I wonder what you need all those additional options for? Is it for fine grained control of the mappingAnd in which situations would you use that over the default convention over configuration +Could you please attach screenshot how looks opened in XML editor in your JBT installation? And fix description how to verify this Jira I dont understand it +The more compelling you make the use cases the more likely that becomes of course! to Juergen whos already been in communication with Werner on this topic + resolved issues. +srcorgjbosstoolscommonxml if used doesnt caches the schemas +More general solution implemented at. +Daniel I agree with that a warning would be very useful especially fornewcomers to Velocity +Closing +See also for an older issue with a similar though more generic use case. +Will investigate further +on a branch for this and well land it back onto +We should add a check before we parse +Should we just set the headers and call it good +May be another JIRA +David you should commit your patch for this now +Was even a year ago +Fixed minor issue in Rev +Could be a race condition here where a background merge gets done in between commits causing there to be fewer then segmentsI was thinking about just taking out this assertion and using the rest of the test +Sounds a bit dangerous as users dont know what the default values would be then. +Help in resolution is appreciated. +No I never saw it again +I updated now it uses the comment which is not escaped by might still be around but seeing no additional commits fixing the test thus punting to +Attached patch that cleans up the file system connector page +remove the outputUnigram option from and configure both a and a to operate simultaneously over the same inputs +Committed r on branchx. +Fixed by rev +Unfortunately the cordova prepare command overwrites the manifest file so we need to work on CLIMaybe some section in applications www +The previous patch does not handle deprecations for mapreduce and hdfs +If the code appears to be the identical issue Ill include it in the updated patch +Doc needed for AS but will not hold the release +Must have been a local db thing +redundant with compression due to insightful mark edouard comments +Thanks Tim +lowering down minor since no one but me has witnessed this +Hi Pierreany progressCheers +Suresh deleting the version in the code could cause backwards compatibility issues +If the connection is lost we should actually and kill the current node so that no other partition is fetched +Ive done better I was thinking about how to do this without locking the solution to AS +The sharedLib attribute must exist and point at that location or you will see that the classloader loads the jars but the classes in those jars are not found when the config or schema tries to use them +Its fine for me if you close it +Ill assign to him so he can assign it back to me when the review is done +Thanks for the patch JeroenVivek can you please review and apply when valid +We have no influence over what they include in their versions +Reopening to add new accessibility task +Now we no longer run the formatting if no parameters are passed +Good choice when total data size is not that large but there are large number of generated +If we knew that the other tasks were going to be IO intensive we could them on this node but we dont +If someone can try it and inform us itll be cool +for the patch +Ill use the same approach as for the other open issue +And that is why you can use the which have all these included +We also need to document this and ensure people who set this to false do it with the understanding of the risks +Above test case unified and relative to srcbugs +is one of the most popular DBMS. +The list was complete +Poking this with a stick today +Will be fixing this properly once this is fixed on Sandesha +Please seeThe best fix is probably disabling the tests in local mode +Ah +Looks good to me + lenientValidation lenient should have possibility to tune it think having a lenientValidation property is excellent as we can introduce this new behaviour without effecting the existing behaviour +Verified on build +OK the solution here is to mark your script folder as a source folder and set its output folder to be something like devnull +Why not simply use float without the raw? This one normalizedWe used the raw methods for speed reasons as we assumed that sorting makes no sense and cannot usefully select those values +I will test your changes +just left it runningI have restarted it and turned on DEBUG for spoolmanager popserver smtpserver and mailstoreHopefully well have more info the next time it diesAnything else I should check +The way to not have to maintain this rack information is to simply define all nodes to be in the same rack +Dan mentioned this in an earlier comment +This is relatively common effect so Id say we should include it in the main project. +Moving to as it didnt make Release +Please post the JSP code for the dataTable component +Thanks for the contribution. +Arrr Sh +just committed this +Also cleaned up some strings in the class +I hate subcomponents in jira why doesnt it allow updating fixversion based on parent one +I agree with your suggestion +dont think there is anything further I need to do and the fix was in +Forrests own documentation has broken links not forrest seed +Modified save operation to create temporary file in state folder rather system temporary folder since that could be on another drive and prevent the subsequent rename from succeeding +Yes Ive verified now that this is no longer occurring now that the Maven repos are available again + +Gump is using s trunk when buildingtesting RAT so I know it wont hurt to upgradeThis means Java is now required at build time but I dont expect that to be an issueFixed with svn revision means Java at buildtest time while means Java at runtime +addressed flavios review and the findbug issue + +As far as I understand this problem its more about a problem with table aliases then self reference relation +resize it also affects the Details panel +Problem The spec javadoc is clear about this +Closing +Please apply this patch after applying patch for even this one seems to be working for me +I was somehow the patch +I will recheck +What could be the reason for such conflict as to clone a bug +Can you atach a sample to reproduce +I implemented a new perform phase to check whether dry run was set to true when preparing release +All three XTS AS integration tests fail but only the WSAT one has any server output +another reason why we should use an OSGI container also for standalone repositories this way we would find such issues early +was working on some other project and I could not test it +Re String marshalling +The issue will be closed +another issue want to be able to pass the outputDirectory and possibly some resources to the report and the option of not having a sink so that apidocs and other reports that generate their own HTML can do may also be multipage reports later which will in effect be similar +Please vote and comment if this is a feature youd like to see +updated to current trunkLet me know if it works without problems now in the doc values branch Simon +As I already said implementing quick fixes for JDT error markers doesnt make sense in my opinion as it could very well be caused by dependencies not being declared in theRegarding automatic download of the missing repos via the runtime detection mechanism we need to be able to access these repos via a public url +This has gone stale +Correct me if Im wrong but arent these AST transforms and wont any compiled version of these classes have the transforms applied so wouldnt we get the correct information from the class file anywayIll get a test case together for you +Ill just add an appropriate catchwarning and treat this as effectively the same as supplying a jndi name to a missing resource. +I dont see the difference because mirroring would fail anyway +You can change it and try again +If you believe the issue is not resolved please reopen or open a new issue +Thanks Amareshwari! +Also if I pass securitygroupsenabled false for the createZone API explicitly still the API response showcases it as true +The existing data were already persisted in HDFS earlier +Paul what scripts do you actually refer to? Can you give further details? Currently the is referenced by the testsuite the manifest and the ws tools said Native uses the jar provided by the AS +Thanks I will be waiting for the file +Jochen By mistake docs have not been updated in the patch +A first shot is implemented in revision manually tested with the above scenarioAlso tested startup with config B overriding A successfulNeed more automated tests leaving this issue open. +This button is only visible if you select one or more rows in that table +Just as an update Ive continued working on this today and will try to get a patch up this week +bq +So a value of should never match the position +This should be fixed in r +Thanks Enis! +nrm changes from were committed +For now I just need a way for a user to select a time zoneIm not trying to pressure anyone Im just trying to make a point that some of this is needed to fix other things in the Workeffort components calendar feature +I suspect there is work to be done at the DUP level +Many thanks for contributing it JonHow certain are you about stability of interface though? Once beta is released we will no longer be able to change it +Even if this feature did not exist and a malicious user got hold of the administrative account they could do just as much damage +Content from xhomenicolasjavassistisnt copied to mavenjavassistIve added versions and and see nothing on central +The isTransient flag was added to and serializability checking done in impl annotation processor as of svn revision +This way we have both top weights and top terms for any vector +was rewritten for +This is a good one +The question is what we can do about this +Thanks Jerome! +patch looks good +Ive also seen this issue +Allow full qualified custom Platform names +This context is active in all structured text editors +Another option that might be simpler to take async network IO handling that is written specifically for RPC out and enhancemodify it so that it works for RPC Data transfers clients disk IO etc +the session is null in cause this issue tracked on is the root cause of jms session nullthe bridge sample works when apply the fix for +Anything we can do to reduce that should lead to speedupsWe should run some experiments +I see this is now scheduled for I have an update on the likelihood of this happening given it did not make HQ bad jira +duplicates this and it is more recentA thousand pardons for filing this twice. +This magic currently exists for all the support that went in +Opened to handle and to handle +verification of block placement by fsck +Gonna push this into but patch will likely happen +For backwards compatibility this would probably have to be explicitly specified for example through a corresponding flag on Juergen you certainly didnt irritate me you just surprised meI have a question about the fix that you made does this mean the query method will just return if Commons Collections is not possible? That worries me because removing the Commons Collections JAR could potentially break an existing application in a very subtle way if the application was depending on the to be Im worried about introducing a dependency on Commons Collections +Verified fix by running lang after the commit +Thanks for the quick fix Alan +And if it has a purpose then all the failover properties should be exposed and passed through the multicast discovery again uri +I dont know anything else to help you work on this but keep it in the back of your mind that it could still be a problem +Charles please close if you are satisfied +code looks good +by Richard. + +I just committed this. +I was able to reproduce the problem out of runs by restricting the test to a single CPU withtaskset x java Dframework jdbcapiI will play with sleep values to see what is appropriate to let the finalizer catch up + to trunk and branch +now uses DOUBLE PRECISION FLOAT and REAL respectively for these types as these are available at least in Oracle and newer. +can you give more details here +Thanks for the feedback Neale +The fix is in SVN +OK got it +We can always add this later if need be +I ran into problems backporting this + +Ah +Updated patch resolves a conflict with a recent commit +I dont think failure are due this patch +Thanks! +Close out issues. +I must ping Sean to see if he can fix it on his side +Frankly the right side is already stuff full of content +Added test class +Aint nothing can do about this issue given paucity of info +Looking forward to using the new Macintosh transcoder +Not completely fixed still drops formatting during pagination as of release +Authorization example deploys just fine +The admin and config models are part of core +Using and add unittest +The codebase we are using creates a new proxy for every message passing through the pipeline and they are not invocation aspect appears to only be relevant for the trunk version of wise and hence ESB +Looks like has been closed again so considering this done +Can you also take a look at GWT VEWB and the Android stuff? Android tooling GWT Visual Nick +Sorry must have left that comment on the wrong JIRA last week +Do you have a problem with the initial proposal or with Thejass suggestion +Are we waiting on anything to commit this patch? +Works for me +Thanks +Sure +Alexs test needs little changes like start mini cluster and stop minicluster +Yes I checked + to say The above snippet is how it is implemented in current cforms trunk +Also any opinion on to support? +Here is a patch which makes sure that the M Extension for Eclipse never writes an illegal settings m still looking why it happens but after a restart of Eclipse the error was gone +A typo corrected +We can then derive an optimized JCROak observation setup from thatb Optimize how threads are used to deliver JCROak events this is a separate concern and might be totally independent of a +So all issues not directly assigned to an AS release are being closed +Verified in JBDS Build id +Bug has been marked as a duplicate of this bug +we unrolled revision the problem introduced by this revision was that the requestcycle kept reusing old pageparameters in the urlfor call + +Finished Applications showing and +It is able to find inherited methods just fine +Putting all status into a single object sounds like a good approach +Committed revision +Move out of basically use the first nics MAC of VR to determine following nics MAC in such situation. +It was a wrong decision to revision merging to for +I just committed this +Peter does this only occur in Spock specs or also in regular Groovy code +Forgot to mention the patch is against trunk +So the message should directly discuss that use case and provide as much context as possible +Tested with ASM with zipped deployment enabled and plain new Seam project +The second patch replaces the first the firstMartin +That is a huge patch +on patch v + +Douglas +An object can be null +In i have tried handling all these scenarios but currently it was suggested that we can wait for +Thanks for verifying +sorry yes this is not pageflow +doesnt update strategyoptions for KSs that already have +Blocking has just checked in a fix please test if everything works as it should +There are still a handful of small tasks that must be done before I can cut an RC release including proper documentation testing and ensuring that all dependencies are brought +If it turns out it is Ill update the issue in case anyone else runs into it and then close it out +svn folders Im using the default excludes and hope it will exclude svn cvs and other folders. +This version is not available in the public repositories so we use dependency management to select xmlenc the assembly will attempt to resolve version causing the build to fail +Yahor told me the same thing via IRCIt was working like that before this refactoring and without much trouble +The patch looks good +The backend work for this is mostly done +in paths +Well after looking at a bit more Im not so sure what the correct solution is +This causes the last transaction to be rolled back after the connection will be closed even if the was called beforeLesson learned Better do not use the same names for arguments and class members +The priority of this task should be Major at least +I will look at and see if it should take a +Postponed +The Central GWT archetype uses this issue has been fixed in errai PR fixes this issue for JBT Beta and JBT CR by upgrading the gwt archetype to can check this by replacing the with version of errai in the projects +Im not sure you are meant to use L for a serial version uid but then its only to shut the compiler up +To add a bit of historyMost of these fails never would have been fails when most of this code was written +Mass closing all resolved issues that have been closed for more than one month without further comments +v LGTM +I get it now +Daryn Bikas have I addressed all of your concerns in the latest patch +Hispanish translation asks only for addresscheers you! +and contain code for different consumers of the startup progress information +Forgot to include Konstantins javac warning agree this looks better +Thanks SiqiI applied your patch then I verified that the EJB Server page was properly displayedMy tested environment is Windows XP IE Firefox the patch to trunk At revision thanks for the patch! +Its ironic that a fully optimized index is the worst thing you could do when searching segments concurrentlyBut I still dont understand why is not merging the little segments from +sorry +Thanks patch + +Should this issue been closed +Correct +Defer all open issues to +We rather explicitly allow for extension through subclassing like youve been actually think is a real lack of really expect to be able to add in in your templates whether you are in web MVC dev or just generating a bunch of. +I will take a look at this for +JeffWill this patch break CLI for sites not using ? If yes then I donot see how it could be accepted +Looking through it everything seems to be good +Like you I looked but could not find a +Started DFS by running usrlocalbin rollbackWhen you upgrade from v to v you do it by running upgrade on v +Note that this can be worked around by ensuring the bundle is started before any bundles that attempt to use +reopen to set duplicate value of +Fixed in Mike +Patch adding the exception incorporating Aruns comment +regenerated patch to check return value of mkdirs just committed this +From Ralph ApelPlease dont create a package +all issues not touched in a month to +If need be I can switch it to but Id much prefer switching all the to I am a strong advocate of declaring plugindependency versions in the first place they show up in a tree of you control +bq According to these guarantees we would not be able to make regions of a crashed server available until after their log replay completes correct + hintThe test in the past has been a canary in the mine for memory issues with Derby and Network Server so it would be good to determine if this is a regression +Manually tested well done +Maybe its me but I cant see any difference in signature between the two snippets other than the Objects have FQNs and its all on one line +So to make changes generally not permitted you will run in admin mode and all other times you will run in regular user mode +But Cygwin isnt just for the scripts since it is also required within the code for a few commands +Think of the situation when you have a project A checked in to scm having a dependency to a snapshot of another project B +Hi NikolaDont worry about the maven stuff I can try help with that its mostly to get a unit test running that boots up an embedded loads it with the necessary test data then runs the test case +I oped against the wrong project +Now its overlapping communication and computationSetup should be ignored +Attached a patch we deployed to production that fixed the issue +But thats not the only use case +Fixed in trunk +If we leave it where it is we have to add to the class path which doesnt make sense +Thanks. +and there doesnt seem to be an edit button either +Patch looks good +In revision is served as system +for +Thanks MikhailPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +In this last version I added a Rays suggestion take advantage of properties file cache for the default blocked view propertyI think it will be the last version and commited in trunk revision +Can we force a scan during addition of a repositorySounds like a good ideabq +I am able to verify your change for works +it seems there still is one issue remainingwhen i pass some parameter which is not in the alias list all values are set to null +I think it is ok to just have the method that takes a boolean parameter +Note this only happens with the paging collector +Shouldnt the classes be in? Note the final s in plugins +It wasis defined in but we dont load that we either should modify the stockMRI version of and load from there or we should just implement the method in the java code and be done with it +Configured the new attribute ssefalse in when using Tomcat As described above this now requires version when using Tomcat The patch is now no longer needed and thus removed. +Generally I would be more than happy to do this +Im trying to program a functional and put the source code in the Wiki so other people can use it +fix will address this problem too +Binary block compressed file +Yes thats right +Great thanks for looking into that +It looks like Im still at the latest commit on master so unless I have some time to track this down Ill just wait until a newer version to verify its working +It disables all JVMTI events when it calls compilation +I cant easily adapt my application to to see if the problem occurs and dont have time to do a full testregimen any time soon I wouldnt let this be a showstopper +Rajesh there is no web service specific error +So I guess this bug can be closed as wontfix as Stefano impliedIf trunk isnt using javamail either u might want to upgrade it thereThanks +New patch fixing issues found from Zhengs code is the output of the unit tests +This issue is already fixed in +And that makes a nice thing to have once again +yes sorry +I see now it uses each one independently +Adding an Appender that does that would be fairly simpleOn the other hand if you want them added to the logging methods then I would say that the Message interface is already sufficient for that +Lucene doesnt cache postings but does cache metadata for the term though I cant see how that could lead to this exceptionThis could also be a hardware issue? Do you see the problem on more than one machine +translating them to Should perhaps be + +This is for sure based on a +Number was done in +Please check if gets calledDo you use the JSF impl or Suns JSF RIIn case of JSF you can check if gets called and ifa the is in the phaseListeners listb it reach the Happy debugging +This jira is still open yet there are commitsAlso Im tracking and might be related to this commit +I wouldnt mind more detailCan our LRU be resizedMemstore upper bound can vary but there are interesting effects like if its too big flushing can take so long the memstore fills before we get around to flushing it again so we blockNit minutes seems like too coarse a granularityGood stuff Enis +MarkDo you have an example process in which the fork activity failsThe fork tests in our test suite arent failing +BAT syntax was weaker and we couldnt face debugging it +testcase for input +works after add the dependency +Why do not we wait for me to post my version and lets see how it goes from there +The addresses are on the Gant home +Is this actually a restriction of making controllers singletons if they werent singletons this wouldnt be a problemThis workflow isnt one we want to code for because it isnt a valid scenario however it is a scenerio we are unable to for the time that web UI frameworks such as Spring Web Flow provide a conversation concept that isolates conversation state in more specific ways not just holding one conversation per controller and +Hi Erwani think i removed the hint because that the fact that it is is not obviously but if you like to keep it +Rajith thank you for your comments I understand now better what is going on with the timeout +The current version allowed making members final and whatever we see in perf the current Collectors work w local refs rather than derefing +Its not a big deal the message just needs to be dialed down to debug +JeremiasOK I will treat easy on the percentages +Can you apply it from your side +Added EBCDIC support to trunk backported to branch +Untrusted certificates are tracked in +I think thats better +Zheng would be able to give you better answers hereFor the error that you are getting I think you have to end up using I think but Zheng can confirm +Ok this issue also fixes the double include of and externalizes the oamSubmit functionsLeonardo and Jakob could you please have a look at the patch +Caused a CMEDont need to iterate over entrySet in get +It seems that resetting the property variable to null before invoking the antpathconvert solves the problem +fixed in master +This issue has a very negative impact on the usability of Developer Studio by default it deploys exploded archives for an to see changes they need to restart the server +ok deleted the class and also had to remove its package from the manifestonce i did that i was able to build and test and everything seems to work properly now +Im preparing a patch for an improved which should also resolvethis problem +I put in some more debugging statements intoand +May I ask why a class marked with exposes and which are marked ? sure +Patch file +But this was done when we didnt have the process store or a better way to access a filesystem resource at runtime so that should actually be refactored at some pointAnyway let me know if your experience is different and well find another way to cache those resourcesAnd thanks again for your patch its most welcome +Attaching patch by changing the log statements to debug level +scratch that last commentI had copying the new jar so the test used the old the new build properly in place it all seems to work great now +Uploading logs reproducing this on my setup with master and but wasnt able to +My concern would be if the params are added to the Jena execution context then there may be some security issues +Not directly involved with this issue but something you can do easily when working on this issue +thanks again +Applied. +Changes committed in revision +It also fixes the indentation of the second list +I agree that leading or trailing whitespace can be confusing and dots are odd but I think forbidding or allowing that should be done at a higher level than kafka we should allow anything we can tolerate +This is goodSo perhaps the thing to fix is that the namenode should not automatically create an empty filesystem if it is misconfigured +Thanks. +Fixes for both are attached +So basically the event was not usable for users once the entity got detachedBecause of that I would be in favour to even make enabled by default but thats a policy decision which is not up to me to decide +A related topic here is what happens when an INSERT is done to a table with an index on it +The support will be included in M and then we will work out the equivalent for M +After some investigation this will take a considerable amount of work +Let me know if we really need to add a unit test +Superceded by new CMS site. +We might as well keep it simple for now and extend those as and when required +It looks like the customer recently asked for some sampledemo code on using XA transactions for the JMS endpoints +Current executeStatement implementation in blocks until the operation is complete +Ive done a pull request to get our fork of that into EAP +My gut says its more confusing to put updateseq and lastseq in the db info +Same for for +Many scanners only return false when there are no more current to return +JacopoSorry about the fat fingers +Service assembly with source +Raised for this issue +The stream get closed immediately after been used +Youre right +Anyone +updated above list as well +Instead we use our own interceptor in to generate the eviction pojo event +It is not meant for inclusion +to the patch +Eric please review +Unable to reproduce this in Tomcat either +You forgot the +Closing +Alternative route as discussed in the mailing lists make protobuf required for the maven build +I will ask around for more guidance +I remove the from the first patchI ran it using eclipse and it tookhope this solves the thanks for you effortsfrom now on will submit in parts so we dont run into thisdont want to make more work for the tireless committersD +Patch that warns rather than throws +Import failed +added Clearspace by Libor themes by Rysiekreorganized by Jozef +Apply as you please +Let me know what your concerns are and I can try and alleviateaddress themIs there a devstaging branch this can be applied to to run it over all the unit tests in the tree? I admit I only run the tests and even those take about half an hour +bq +An additional check of the is is required to make sure there is a parameter underneath +Actually I can reproduce its fixed unless that you have built your our jars from subversion trunk the problem is solvedOr just wait for Annotation +Please try it out +That helped +As per demand here is the patchPlease note that it has not throughly been testeed by myself +At some point it will always be possible to get a nsreIf thats the case the compaction management code should handle that scenario more gracefully +confirmed to be fixed +IT added +Accepted +I pinged Paul Ferraro on IRC + +BillTests do not pass on either or trunk with this patchsome failuresant test Dtestcaseant test Dtestcasetestant test Dtestcasesome x failuresant test Dtestcaseant test DtestcaseIm not sure whether these are test problems or more serious +This patch fixes some synchronization blocks on and I believe it doesnt require a test since it is not coming from the observation of bug during a run +I am still figuring out the root cause +Das Encoding der XML Datei auf setzen +applied thank you +Patch for the change proposed in this issue +In the old code a variable was necessary sincethe trigger was referenced before creation +I will try to take a lookat it +Hi Devinthanks for integrating the patch +Most of the data in the world is implicit +So we wont have any backward compatibility issues as well? This means table create is still left to global admin +Verified the patch with on using a trivial patch +Yes +feel free to reject it +If you would follow that convention then there would be no name clash +A valuable addition to this issue would be to be able to add a feature by which one could add exclusions to the rulesets for pluralise and singularise +Pull it back in if you think different +Patch applied with many thanksI guess Ill leave this bug open until the proper solution is applied +n endto poasses in mri and fails in m not sure whats happening here but all these examples give me EXACTLY the same result in MRI and latest Toms test case +Released in For nightly docs Near each component its necessary to tell a user when it was or will released +Would you please try it +Test failures are unrelated to the patch +Ready once the discussion is resolved +I am suggesting we have warn on comments +All the ones you mentioned are not already part of the Admin API +I have polished the javadoc to explain the behavior of the doStop method. +But that doesnt help in finding the errorBest behaviour would be to report correct line in file where evaled string was defined +Changed the constructors for to more closely match what was available before +bq +Will be considered for a to method to. +Hi MikeI think that the other failures may be effects of the failure in test +Updated title to better reflect issue. +The fomarker is still broken inThe problem is that the ts in the parent object is notinitialised +It served its purpose. +Spring lives up to its advertising as a pleasure to use +Please update summarydescription to reflect exact change +s classifer field is now deprecated. +ok this solved my issueMaybe you can put a hint beside the NPE that the producer should be started before usingthx +Sorry for the inconvenience. +I am not able to apply the latest patch to branch +Initial plugin rev worked once +pushed +To be consistent the keyword user should better be changed to uid +Mike How can i set the flags for transaction logging +It doesnt do anything when deployed but it compilesVerified in dist build from SVN . +is another issue where long columns may not get space reclamation issue +It is easier to deduce from the code what type of exception can be expectedOh man I just changed it no worries patch coming soon +applied to trunk under revision +Committed to x in revision +Thanks +bq +However the problem I reported was triggered by the DS problem in conjunction with a problem in my code which caused a large number of connections to be used simultaneously +Patch attached to enable operator in statements on objects using to release in branch +The default signer should probably fetch the private key from a file that is configured via +pmdcheck is done +Hadrian has committed some parts of this to the trunk as ofChristian I do think you should do a fresh start to see how we are with the tangles and create new patches +should be fixed in error is gone when using HV together with Hibernate s the right combination HV together with Hibernate +Local test passed +When that patch comes in the javadoc would make more sense +See steps below to reproduce Login into the portal Go to Browse Resources and select Enter Tomcat in the Search The list of Tomcat servers is displayed Choose Servers from the View Instead of the list of Servers being displayed an empty list is shown +The is NOT correctly printed to standard out when usingfork mode +Wrong selection range is the effect but not the causeI suppose key handler does smth wrong +I still can see failing due to assertion failure on Windows on branch +Will raise other Jira for this + for having this be a profile +ResolvedThe application is created +Skimmed the patch +Diwaker Would you be able to run the tests along with Eugenes patch on? If not please let me know +The app config would include the beans name +Thus it is bzr that really presents the first big hurdle for compilation +Most sections were sections touched changed tense from future to present +Interesting +Fixes both and paths for patch +We have to restart Archiva periodically because if this +I wonder if shutting down the slave can provide a workaround +in case its necessary +But I dont see why opening a Geronimo plan with the Geronimo editor should result in an NPE + +It would be great if you could post a patch to fix it +See comments in SVN and on The customer is facing an issue thats directly related to the way concurrent paging will happenThis fix should be included in the current release +Here is the patch with the fixes and unit test +Thanks +No luck so far unfortunately +They should be consistent and all have a preview +Here i see the storageId is still with fresh data +What sticks out looking at you are creating a bundle top and then attempt to uninstall a bundle with a BSN of bundle +Perfect task for a contributor +to Pavel for antlr help +Thanks Alexey everything works +should been dropped from the trunk +The patch does the followinghas a mapping from long to acls in the datatree each datanode has a reference to acls +Maybe others know +Edits to existing objects are not saved cause is +thanks +For any other type of licensed work please contact the Legal PMC +Thank you for the tip +Committed to trunk +Sorry for the confusion +I had put them their because I didnt like that one of them came from and the other fromA smarter client will do things like make sure the protocol semantics are being automatically followed +added a second kind of disposable stat those with a single column and unique key +Note that the test also requires the currently in +Closing. +Im not a good writer so anyone who has the inclination can feel free to improve it +I believe this can be marked resolved since the fix was committed to the trunk last week +If you are willing and have the time to restructure this patch well need it by late October to get it in version +Tom can you confirm this is still happening +Do we actually want a temp +Is this still an issue using +I dont see the change in the CVS +Sorry for coming in lateI think doing statics is a mistake we should avoid in MR +Im not sure how to address these +avro +etc +Now maven and maven builds are independent of each other +Ill believe you that is fine I dont know why the versions acted differently. +Trying to figure out what to tell users that may have hit this +Lets get a feature branch going +starting with If you are using this bridge version from +One additional note This task will append a x to the +I have one ready but before its done need an init script +I dont think adding information about specific sql states is a good ideaIts not very frequently where error messages needs modification +I see the same behavior and I would urge the maintainers to come to an agreement on the fix +This allows the userdev to handle it a custom applied with svn revision +lib +At this moment Im in the user group without the permission to assign this issue to myself +It is specific to snapshots which were created before the log files and db dumps are in bug happens after failure to create a volume from the same snapshot +It would be a lot simpler for us if we dont use the sonic NS service we now use who provide a far better NS service than we had when this ticket was opened +Could you live with adding them +What about providing both strategy at the same time +If possible we could only reload the class if it has changed and only if in development which shouldnt cause too many problems with the statefulness of per my discussion on IRC with bbrowning the fix for this issue will likely not be tied to rails for the record. +Those tests passed +Long ago in I carefully figured out what changes you did and patched my Solr trunk +Its working fine here in the UK as well now but was definitely behaving badly on the day I reported it. +Will that be enough +I really dont want to bug you wicket guys but leaving the compression on will make wicket unusable on SAP JEE Server +Increasing timeout probably is fine +Is this still a problem for youI have tried to reproduce but I cannot +This is believed to be fixed +I found that I needed to modify not only the table subquery topic but also the Boolean expressions topic where EXISTS is documented as a Boolean operatorAttaching and with these changesM srcrefM srcref +If such beans gets exposed through service locators or through remoting endpoints or through service references then they will qualify as public at the application architecture level +Hehe +This bug has been marked as a duplicate of +committed to and trunk thanks for the review! +Really needs this +Will migrate to Arnauds solution when Maven is out or when we move to Maven +Can you try it with popupReference to a popup outside the sheet +I think the intention was to add the new tomark set to the existing set +A property can point to the customers installation +This patch would have to get merged with the recently committed one from +yeah +Woops thought I had gotten them all +The patch for the API and TCK sources is complete and checked in +An illustrative patch which achieves this +Thanks guys for reviewing +Backed out for now +attached patch addresses issue please to master commit hash cdcedeedafbfcdefac +on patch +I will remove it +Ill let Simon N decide on this +Ok I looked further the project was added to the silly incubator permissons scheme Ive changed that to the better default permission scheme +Would you please assign this issue to yourself? some comments on phabricator +It is true that an output can provide a workaround but thats only a workaround not a fix +Because in both sides of the communication they should have the same code which include the comparator code +Mailing lists are in placeApologies for the test messages +ah I meant the defaults for the working directory etc which end up under is probably not a big deal to do in we should tie it in to the configuration review which is another issue somewhere. +Ive gotten confirmation that my ICLA has been received by the ASF +This problem reproducible when we closse VPE on VisualSource Tab +is also done is going to take some graphic design skills I think to make interesting images +Will do +Reverted +nm looks like you added it to the server adapter feature which is already contained by the +Each Set of parentheses should contain dot star date and then there should be dot star dollar at the endIn any case none of this would account for the fact that select part col col and select produce different results +It should be described in tooltip Func +Thanks Ashwin and Claus. +If there is something important to do then its important that it completeIs this not true +Thats fine +I didnt observe any negative impact by this patch +As long as youre using Forrest version or higher you can just create a symlink to Java and name it Java and all works well +Will report back +I think we would have to add the complete AS feature as a dependency in the central feature +Verified on +This patch implements the JSON encoder and decoder for C +Closing. +Docs are missing too but Ill add them until final release of +We can definitely do something with the components to jazz the pagination up a bit +Linked to +from Product as +Does the MODEOPTIONS table exist in the Oracle database +ATM In future please open a separate YARN jira +Added formatting +No because it wasnt fixed in right fix version is for the branch +May want to remove some of the logging messages but I think they are handy to have aroundIf someone could take a look at this and see if anything needs to be changed +Unfortunately a regression was introduced recently +I think that this is an errormeta description is a standard html tag for search is used almost on every page +Moving the file to trash and open requests arrived at NN simultaneously +Ill try to create a patch +This only happens on multi project builds where the artifact is in the same reactor and the compile phase is used +All things done +bulk close for +These files can be deployed and undeployed by running ant and ant + +In the meantime I would appreciate your feedbackThis patch tries to set the classpath properties if they are not already setIf the following library properties are set they may influence how we set the corresponding classpath propertiesjlibjlibThis tool behaves as follows If the classpath properties are set nothing happens and we simply exit +Further information +IT is online +Must have leaked in from my version of the scripts +Same problem here when trying checkout change checkin in one transaction +Would appreciate any pointers about where I should be looking thoughIm posting this patch in case the store fix turns out to be too trickyTouches the following filesM javaengineorgapachederbyimplsqlexecuteThe fixM javatestingorgapachederbyTestingfunctionTeststestslangRegression test case +all except for dependency to Seam still needs Interceptors and to like I can probably also get rid of those dependencies but that is not so critical +It worked it took about mins to export +It was pushed into master only. +While I agree with your analysis of this code this isnt the only cleanup code +I would also like this feature at work +Assigning it to you as you said you wanted to look +bq +Whats the size of your messagesHave you tried on a SNAPSHOT +OK you convinced me +It may have been added after truncate code was added +The last metrics are almost all moved over +To future travelers this means that each file needs a discrete dataDir property in +This might involve some code changes but without taking a closer look its hard to sayDoes this modularization need a vote on the mailing list? Just to see if you want to go ahead with it or not +Reopening as it is failing in trunk +Thanks for the patchWe should get it into the trunk +Fix one line chars +If size based weight is turned on and an app has demand I think giving it fair share is the correct thing to do +It cant go in until goes in +Just having a single config constructor makes sense but currently itll be pretty awkward to use +Mostly mechanics +This feature could also include +I will commit this and update the wiki +yep that was it I resent using my addr and theyre being delivered now +It is the only way to be sure the understands your intent that I dont think adding a splitPath method would make things better +These versions supported a byte max length for +What we trade for that is the universal expectation that stopping a Cassandra node never triggers a flushIs the very small group of operators who run with writes dont have to run nodetool drain such a compelling win that we should change a fundamental behavior of Cassandra making it less predictable in order to obtain it? My answer is no +for the patch +Needs additional research +Patch changes cleanup support for spot price unit testsI have done some manual testing on and everything seems to be working fine +I didntsee that you had already taken it +Modified the Sling API project as well as the scriptingjavascript and the microsling project +David C do you have any commentsobjections to the fix proposed by David B +Im pleased you like it +As mentioned in we can have a flag for this +This is configured statically in code to +Starting work on this issue now +I have the changes made and tried here on Ubuntu Solaris Mac OS X and Cygwin no problems that I could find +The guideline we used for the first round of tooltips was characters maximum +Yes we can close this +to patch this time with the correct name +The class has been changed since this defect was written so those changes might have addressed the problem Paul was seeing +Does that workI take that back +This issue is not pertinent +Tweaked the instructions in the poms a bit. +will need to update later +Thanks +hope it does not blow up anything +Sorry +Let me check +Unless someone objects that is where the build target will expect to find the summary +Email sent to Arno Unkrig and Eugene Kuleshov +Patch Fixed broken links in one file +Also in test case test case didnt wait until VR is stopped then check the state of VR +This is a of using to read the +check to make sure that a doc directory is there and is up to dateGreat idea Chris I will try to get that will do something like check to see if the date of the docs is within one day or so of the release +Seems that I cant make this happen again and theres a solution +The file will need to be checked in +I ran into this issues as well + +We need to create a Forge group in there +This problem actually presents itself in the test bucket that IBM has been running trying to work through the performance issues of and puts the runtime behind the competitive implementations in the scenarios above +Committed on trunk +Working up patches for and trunk +Moving patch to file your patch +Fixed r in r in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Anyone else done any testing? How to move forward to have this patch reviewed and submitted +Required changes in Jackrabbit changes in the Jackrabbit API +From what I can see there should only be one change the persistence service +This looks like my badWe used to have a problem with this and I turned validate on and my build broke so I placed this JIRAI seemed to have tried it with an old versionMy latest code with RC works with validate onMany thanks +Default styles can be used to test +Committed +If youre going to unify clustering and classification then some classes will disappear right? Perhaps only core writables should be pushed outThis is a poster child for using a few weakly typed data structures instead of many strongly typed structures +To be clear the exception is only thrown if an actionListener binding is used +Updated the file again with another column showing todays status +I glad to see you are picking these up Senaka +Ill continue work on this +If its not b or higher I cant without the version number its impossible to know +This test has broken the build again +Tom have you checked in some things for this JIRA? Just thought I saw something about this a few days ago +No more work will be done on this. +It seems like we have way too many namespaces for SOAP in particularRegarding having different transport ids youre saying that if the endpointinfo is created from wsdl it has a different transportid than if it is created from javaThis was fixed a long time ago +It might be possible to simulate this through. +Until then feel free to lower the priority and Ill get around to it one of these days. +Sorry its been a while since I use expect. +Fix +Yes I hope youll at least handle the relocation but I fear this wont be enough to be able not to have a and a spring jar in the same classpath +Changes from review and added some extra ACL intro ok and committed +Reopening investigation of an apparent memory leak has shown the previous changes result in leaking the sessionsconnections in the management layer by causing duplicate registrations to occur +I have created to track the +Thanks Jim! +Only matters if you are deploying as a War rather than an need to make sure that you have in the EAR lib directory +Done committed into trunk. +so this cant be related really +Thanks KelvinPatch applied to LUNI module at repo revision rYou patch also included a modification to that appears somewhat unrelated +Duplicate of. +I guess I keep this os around for a while for testing then +I have done a new instance in to enable the start of since there are two composite files i felt thatz the way to specify which one to start instance of +Thanksgeir +This approach is similar to what LinuxUNIX supports +Patch rebased for +Fixed the unit test and uploaded a new patch +Ill mail you and Dimuthu so we can compare the commands you are both running +I didnt see any errors when I tried this out do I have to enable it somehow +So artifactId resolves to the current project and extension is not set hence it resolves to null +I would like to have a few more constructors added to the classes if nobody objects +Weve had some problems with notifications which are fixed now probably too late for this +AFAIK there is no published API to do this +Perhaps we should change those consistency checks to assert calls so that users can avoid doing them in production since they are being done in synchronized blocks. +fetch row dereference the statement force garbage collection try to drop the table with the prepared statementI t hink if the is not closed on garbage collection you will not be able to drop the tableI have not tried this but just a thought +I tried to execute TS against EAP ER and test failures are present too +This is exactly what I had done if the onclick attribute of selecthad worked +Is there any reason to use pool name as the artifactId? I propose we use pool name as artifactId if there is no in the pool otherwise we just use the pool name portion after as the artifactId +I think it should be removed from as well and only be in the Dictionary classIf we have duplicate entries it does not matter because the dictionary is a set which contains all strings mentioned in the dictionary file if there is one twice it will still only be one in the set +Patch corrects the adds unit test forI ran streaming command with cacheArchive without passing symlink name +we have reduced the colors used for cfdump to colors means now a monitor with only colors will display the dumps right +ps shows the qemu launched sitting thereusrbinqemu drive filebuildappliancesifedoraJEOStmpcacheoffifvirtio global off nodefconfig nodefaults nographic m device serial stdio chardev socketpathtmplibguestfsidchannel device virtserialportchardevchannelname netdev useridusernetnet devicenetdevusernet kernel tmp +Ill change that +bq +Straight forward failed contrib test is due toThe failed core test is clearly unrelated as this patch is a documentation patch +The correct issue is +patch for based on stdcxx +Hi Mathias Filipe already backported the fix to the branch +Upgrade from to seems to be causing several problems with session and prototype scopes +I had empty the folder tranlog and restarted the server +Address space separation of primary namenode and backup namenode thus better fault toleranceThe namenode when invoked with the backupmode command line option functions as the backup namenode +Please review what I have written Groovy has undergone a major update from version to +IE is not supported by Wicket anymore. +Moving to since G depends on this fixDavid J +This patch uses the DOM to get the values to return +Is this interpretation really needing support in Hibernate? If JPA doesnt forbid it it certainly doesnt allow it either +i have changed some files to get a compile without errors +So if we cant get that working with Rich Faces then lets include instead +fix javadoc ignore the proceeding comment it relates to a different JIRA +do you have any ideas to solve that problem +should be doing less retries than ms but the graph shows it doing more +Adding methods to public interfaces is evil! +Thank you! +added to docs +Thanks Stefan! +Good thanks +This is an improvement Apos +bq +I clicked back on design from source view and the GUI was rendererd as screenshot tryI wonder if we need a special logic in Fuse IDE to render doTry +That looks good to meThe part that irritated me was that it only sets them when unset +As a committer Paul and myself can request for any infrastructure help +Thanks +ok have a functional solution to this in place nowEffective Roles renders a list of effectively assigned rolesAssigned Roles renders checkboxes letting you remove an assigned roleAvailable Roles renders checkboxes letting you assign a role +For now I am just targeting a decent fixed size while the bulk of the grunt work is being done +Set this to resolved +attaching new patch to fix went ahead and committed the code changeSending snapinstallerTransmitting file dataCommitted revision +Passed all test except in in hadoop +rtc item real problem was that what the user supplied in the source folder field in the New X Class wizard didnt correspond to a classpath entry for the project +So is this actually causing an issue with something? That console needs some serious revisiting anyway so if its not actually causing an issue then perhaps we can push this +An alternate option is to have a set of that depend on both hadoop and different are marked as exclusive and user can install them +It is wired into and will be available in JON when our integration there is done so it is available remotely +Hmm I have hit a bug accidentally! Shouldnt it have terminated the nested block instead of waiting for more input when there is nothing to +Did you resize the popup window or did ut open with these dimensions? +Thanks to Jacopo and Scott for discussing requirements thoroughly and providing helpful implementation tips while implementing this feature + +How are these supposed to get cleared? Is the threadsafe isolation making it so that these are not getting cleared +Ok +Lets take security out of this issue though Lance +Thanks for adding the tests +Slava does look for resources by path lelated to the root folder of the project or +Maybe add a that does the casting instead of Configurable? Still what you have patched is better than what we have already +PatrickClass is not in the patchSimon +I know you are all nice guys +Patch for trunkThe two places pointed out by Jon above are already fixed in trunk +Hi Guys according to me this issue is not a bug but a request for improvement +I just committed this +I was not able to reproduce the issue but I still think there is something strange in the behaviour +Unit test reasonable +Thanks once again excellent Spring support +Might be easier transitioning that when we have the wiki +Instead provides a method set for setting checksum type max and checksum size +has been marked as a duplicate of this bug + should provide capability to transform each row before writing into target hbase instanceThis is useful to reconstruct the scenario of without disrupting normal operations to the table which was broken but later created +Ill need someone to commit it +Youre right in that it does not need to be a JDT project +The minor back compat issue around it is not worth the slow default +patch fixing this project +relative parent directories +as soon as possible will probably be an independent hibernate tools release during or after eclipsecon +So based on type this is what I thought is best +Migrated additional test cases +It provides a and for +The first release candidate was cancelled due to a discovered snapshot dependency in kit +Looks good no issues with the import +I would still prefer logtail most developers will use this often and an option to logdisplay will be overlooked im afraid +If the only mappers left are in the penalty box then somethings probably wrongCall the home stretch percentage +We talked about this in Vienna and it is still on the agenda BUT it is not just about making its about ensuring proper semanticsLike if we change the default of String to something else should this also affect you when setting parameters on a query because here is also in effect? And no it should not affect querying Thus we need to ensure we keep the overriding as isolated as possible mainly to the Configuration and constructionIll hand this over to Michael as he were so nice to volunteer +Also should the refactoring ask the user if the local variable should be typed? Or should it always use the inferred type +Vinay before I commit what name would you like listed in the changelog? Just Contributed by Vinay or would you like a last name as well +Protected by fixes will now make rc +is now a separate interface. +See the plugin as an example +Nope we dont use SSL +Theres precedent for this in Derby and +We can when needed +Running the full test suite for sanity +Hope it can be better maintainedNo function change from previous patch +If there is a problem please open a new jira and link this one with that. +i will update the document +I did not change the file since it did not seem neccesary +What about posting a minimal test caseAlso note that was ported while was at version and not much fixeschanges have been done after thatHave you tried? is workingThanks a lot DIGY. +I would disagree with the last one there might very well be a number of scripts out there which expect the to come out in a certain format +Is that ok +I suppose the answer is that we have to bind the application context into JNDI using an +Longest time seen msUncontroversial change going to commit soon. +Other circumstances could be When httpejb request takes more than X ms then generated a thread dump +If there is any further discussion regarding how may be more flexible or used with other web frameworks I recommend we discuss it further in the forum andor open new issues to track those items individually. +Likewise for the soap thing +I have created quick fix that just resolves expression for attribute password and that unmasked password is then used insideProper fix should be as Brian suggested using +If that fails then we should not let this compile +Are you using? If yes pl +mmm solution looks ok to me +DN reports those blocks as valid to NN +I have not looked into the failure +It would be really great if someone could double check the changes and comment +In my zeal to comment I didnt that I read the documentation for which claims to to whatever conversion is necessary to convert from the native database format to ASCII +As part of this we should add a test that tests input topics with two different partition counts +Im gonna commit it +Committed to trunk +Okay I see what youre saying now +The latest patch looks good to me +Attachment Not able to read FOP pdf has been added with description Not able to read produces Adobe PDF +If disabled rename should throw an error to user +Currently the bucket with the Solr configuration tarball is not destroyed when the cluster is taken down +I just committed this. +it will be much better if you could supply a self contained unit test for us with drl and test method and facts which will allow us to get round to this one much quicker +to +The fire is needed for using Maps asAnd name is needed for using Maps as ListenersSo in this cases they are requiredAswell if you have multiple registered and their Events host the same event method you need to change the fireMethod namesHaving and not in mind I would say that the options should be presentI dont mind about the class level annotation I think it is a matter of style or preferrence but would stick to the field level +Close issue in preparation for release. +Hopefully that work will resolve this issue +I dug in and it seemed that state was being spread across two different locationsTo fix this I moved the flushRequested boolean into the class +So thanks forthe confirmation +I just found the after the split doesnt work as expected so I commit a quick fix and update the test case for it +Then assign this back to the QE +It would be great to call HIVE from grunt +Closing EAP CR +I just committed this +Maybe it would be easier if you just do the command line arg parsing yourself +So the issue is not that the SBN throws +Can you please test this using Castor? There is a fix in for losing the timestamp precision after being converted to a +Let me try to cancel and the same patch +The servers view is designed to be placed in teh same pane as properties and most users will not benefit from this change +When doing a test with I was expecting to get this message when the property was not provided but in fact this is when the property is not provided +Committed revision +Thanks +The issue reported here is very likely to happen in setups that use secondary namenode even inbq +You ought to have waited until the ICLA was filed before committing this patchThanks faxed it this morning +We also had a very simple recovery mechanism we immediately kill the process that got disconnected and the shell script that launched us will relaunch +Could you please obtain and retry your case with that +Will let tests run overnight +And until aether is actually included this issue remains unresolved. +I will move some of the other as well and post a patch +it was overkill +Attaching sample headers +Ugh +Like Oran said it does not appear to be fixed +I think the blockIdoffsetbytes can provide a accurate records for each IO operationsThus maybe it will be used for some replay tools and useful for IO pattern analyzing however I have not got there yet +I love the new test case +Hi Knut If the test is stuck how do you print out the stack trace for the waiting thread? I think we can use the same tactic to be the fix for +Remove the unused free +Thanks +At least a test case is required to prove or disprove +I also added this for the ANTHOME variable as well hopefully this is a sane thing to have will push out shortly. +to my question Ive already read it but this info is not really answer for my question +Hi Mark Patch committed in revision + +It would be good to have these menu items reflect the state of the checkboxes on the Maven Classpath Container config page changing their text between Enable and Disable and also have them only affect their respective properties +for simplicity something like DfiltergroupIdartifactId Both must be included in groupId and artifactIdSeems more simple than the ga +Im wondering what the default should be Spring or SpringAt present the default is still Spring +Tom does this work when running tests from Maven? Asking because if I recall correctly Maven heavily uses classloaders and if they are used for the project classpath most likely the JVM wont have the project classpathYes it does +Wont Fix is going to get removed +Should be fixed in the next release +The commit message for r has the wrong jira number +in cvs now +fonts has been added with description FOP configuration has been added with description console has been added with description output PDF file +Everything looks goodThe patch can be applied to branch branch and trunk +not for the user + is not committed and the issues arent really related so this is not resolved +I am misreading the output of the +I modified the tests slightly to handle the exceptional exception cases using the convention preferred by the project +Patches are welcome to improve FOP +Per my first comment my initial thought was to ignore expired tokens when writing the image +Initial implementation of a Tika document parser that does not its results and does not look for links +Thanks. +verified on Teiid JBDS +To be clear it would be great if we can simplify the code in but I am ok if that is hard to achieveBTW were you able to test this on a real setup +I got confused since is used in quorum but this part of the code uses Socket and +its a pain isnt iti was thinking recently about how we could have a better and extensible test set for the +Seeing Weblogic screens is only meaningful to those who use it which is probably not the majority of people on this list +Patch attached +Will submit a zip with a new directory and files +Do you have a theme of ajax for the form and submit button +This is no longer an issue. +Lets assume that a blockfile deletion request is queud up +bq +Hey Kihwal +paid special attention to PDF plz look through all the links in the JSF Tools Tutorial and Reference guide Struts Tools Tutorial and Reference guide as the guides were moved from the Visual Web Tools guide there may be links still refers to this guide and doesnt work now +This was fixed ages ago +The design doc patch have been posted for weeks already and there has been no feedback +I do not use the linked project feature +Here is a patch that adds the test and fixes the bugs +Thanks Matt Todd and Cos! My first patch into Hadoop +The requested updates to commands are implemented +Why are you running a MapReduce job with input from a directory that has not finished being copied? MR was not designed to run on data that is changing underneath it +to that just committed this +Omitting it is better. +Hum thanks for the information Ill check the m then going to build another test to assess how well we match the old +for the default queueIf you need a list of all queues you could either scan or add a property that lists queue namesThis might not be optimal but it would be functional +maven works differently than maven? Maven resolve the dipendency without problems with the last alpha +Is this what you meant for decoupling the cache from the transaction manager +veerified +Its clearly wrong +On every write this flag is tested and when it is true and the row key changes I close all currently open This gets slightly more complicated due to the fact that we only close the but dont open new ones here because they may not be needed +has been marked as a duplicate of this bug +Tested after Admin Console login screen had been up and was able to log in successfully. +And I guess we will also need to introduce an inconsistent exception like MDB is object but has some bean types and qualifiers +Hudson doesnt seem to have kicked off a QA Build +I see no record of that occurring so Ive cleared that flag +dupe of this test case all operations are in a transaction. +Axis C server response did not contain but the Java clients could understand the response irrespective of content typeHowever going by the HTTP spec there should be some validation based on +Eclipse nicely reformatted that code for me and I will see if I can fix that over the coming weekend +Hi Martincould you move the and from lib to +patch applied +Revision +Richard i just realized that i misunderstood you you meant that the committed patch is trying to kill two birds with one stone of course not mineSorry just ignore my remark +Committed to trunk +perhaps the best solution here is just to add a bit of custom code that formats the requested string directlyDone in revision +Added patch for deploy tool as well as console to prnt all the module names stopped started uninstalledMarking patch available unassigning so that someone can review and committed as revision in trunk. +I assume we can close this one dont we? +This api isnt for general use only from administrative tools +then +It fails if I try to set the env variable from wi the build. +changed +Please look in history for what the actual resolution was. +This is EPL code +patch file for fix to fix from trunk to branch with svn +The current released version of this spec jar is +Ok Im going to take a look at this bug Ill let you know of any progress +Please look in history for what the actual resolution was. +if the initialadaptor loading is not working just remove it otherwise just leave it default varlogmessages streaming +I have tested this basically by deploying a postgres datasource on a and didnt find any SQL look ok to me also since I have a lot of experience with the patch was applied +Thanks! +Not all packages are being exported. +wesley i can port it to the archetypes just ping me when you get around to fix it +The view handler requires that you have Tiles Setup in your Web XML +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Clark I am going to fix the problem that Michael described but it is different from the one you are seeing +See Mikes comment +Fixed in Rev +This is for the same reason as I cloned the issue to handle the anonymous inner class case and will now resolve this one +Ill try to merge it tomorrow evening +If can take multiple partitions the stats has to be gathered according to the input partition column values +You are right +I understand how that can be very frustrating +Moving it out to +identical tasks for different usersWhat I would need in my case for example would be to define a task in definition and during runtime create instances of the task each assigned to a specific user +Fixed on trunk +Ive been seeing this test deadlock as well on my hudson +fixed +Other tests are of course good as well +Needs to be supported in many thanks for the excellent installation guide that saved me a lot of time +Thanks +for previous version not for commit +Patch attached +from my original comment +Just remove it from the enterprise features but make sure its listed as external projectWe do want people to experiment with it and contribute feedback and code we just need to be clear about it being a subproject and not main part of Karaf. +Which unfortunately has the same name as another file in another package I may change itBut besides that this patch correlates with a note I just sent to the dev list about a lockup Im seeing +Okay is now committed so we have a better API for shutting down thread pools +the exception is very clear the problem is a version conflict for +So we are not ruling out that there is not going to another system like model in the future in Teiid runtimeMy personal preference is not write logic to compensate for users lack knowledge +Wed also be better off if folks would resist the temptation to back port anything and everything +The goal of is not to redesign but to refactor code +This works as expected on Android if it is possible then it should be disabled for what about Jurajs proposal +If the build order is dependent on the previous build state of a project Id like to suggest resetting the build state of newly released projects so the build order wont get messed up +However sadly the backwards tests do not agree with you +Closed. +Did you miss it? +The way it is the pool size grows to the configured and stays there +I dont object to using the first patch then +That said the caching module isnt actually a proxyits just part of a client +Q was it sensible to have it in the remove method as well? added a little bit of javadoc to the class removed extra space in the separator line + +Thanks Jothi! +Thats a good point Im not sure why I didnt just override get! It seems like that should work fine + +Also if this check does not catch anything and bad are created anyway we will know that the source of the corruption is exogenousI imagine that it would also help in testing prospective Cassandra releases +One more question +Opened to address the ticket cache concern +Also I should add that in my use case there might be several files of MB MB and I really only one download a single file by timestamp which is built into the file name +However I see a couple of problems cfstats shouldnt allow arbitrary args specifying an invalid CF works but just gives you the KS summary there is no way to examine a secondary index such as since secondaries also have dot you still working on this Dave +Patch for the first testset in the CORBA testsuite +is a good idea +However I think using grep is okay currently +In that case will be able to require bits per value for all blocks that contain a single unique valueBut I agree would likely better in general and remove one level of indirection +Sorry I am a bit dyslexic the command was indeed jdpa and it resulted in the message I pasted it is a copy from terminal and it does not have any jdpa optionI found on the net a that fixes it I am attaching both the original as the one I am using now that works +Trying hudson. +clear +Sandy apps are running submitting when this occurs +The only challenge is to compile a list of useful and rules +No Guava on branch so could not use the Precondition util +Caching partial blocks There is no problem with a DN caching only the hot parts of a block and still declaring to the NN that the block is cached in ram +Attached patch +I dont know version OS its a third party sftp server where we need upload the we would have deleteFile in FTP not in SFTP + +for an updated patch addressing minor comments in RB +Added transitionExecuting hook to +i think from release artifacts perspective this would make a lot of sense you would unzip and see core analyzers grouping demoSo people wouldnt be confused about where to go find stuff +once this error occurs the route cannot be used any more +You can specify use it via some new constructors that take a bool indicating if you want to use the stemmer instead of the default stemmerThis wont break compatibility with users who want to use the old default DIN stemmer but enables anyone who wants to use the other +duplicated with +Testing with Done inserting in secondsThere are rows in usersThere are rows in messagedusersDone querying in secondsRetrieved rows totalThere are rows in usersThere are rows in messagedusersDone querying in secondsRetrieved rows totalGiven that it takes nearly minutes for rows and that the performanceis definitely getting worse as the number of rows in the result grows its a bithard to predict how long it would take to run rows +I just checked Searcher and forgot to check +And it boils down to a JMS type flagI also added typed metadata properties in the protobuf messages but these can be used to filter messages even without using JMS +Created to factor out conf validations +Updated patch +We could do what we did with and create a proprietary interface its not like clients use a Connection for anything much in Spring AMQP so we could keep this basically internal +The log you produced was that by running Maven via the command line or from inside an IDE +So I dont think anything needs to go to the branch at this point +Might be the limitation of WTP or P so downlevel the severity +Thanks Klass! +I fear the story continues +I created for the JAXB version of the problem +Thanks for the head start on these Reference Manual fixes Anurag +Pardon my impatience Enis +You should have searched JIRA first +Please close if it works for you. +when attaching a file +I think has a limit of n seconds before forceful shutdown is executed +Since are deprecated I rewrote the test to use Facelets and JSF. +Any objections to that approach +Adds fix arg to hbck +Hi FrancisAny ETA for? I need to retest after that +Much of is written specifically so it does not saturate a single servers resources +Id like to review thechanges before doing my patch to the trunk versionIf you have a use case in mind Id be glad to hear itIm hoping to make some progress this week so stay tuned +We have an additional changeset for that which fixes the tests +r committed the fix +Im leaning toward limiting expression support to a limited number of attributes with metadata in the resource description to indicate which +Hi KurtSorry this issue has been open so long +Patch applied thank you you also have a look back at to see if some of its enhancements could be added? +Notice there should be a period at the end to finish the grammar +Sounds good +At least itmakes sense for me. +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +I added a few lines to the test to make sure the options were still intact when returned from +Strangely I noticed a considerable slow down on +Perhaps they are stored in thesession? Once I figure that out I can begin to work on a patch +Propose that we fix this for its an easy enough update +I thought we just passed the URL down to the JMX connector do you know if this is a problem with the deployer or a problem with MXJ +Pi Yeah I see that +Moved fromRavi quothbq +While doing +Exactly what I needed thanks for your help! +They cant be availability until I reinstall proj +Not paying much attention but I think that ejson has an issue here +you could also try to get a recent Xalan versioncopy the into FOPs lib directory and use again +This adds to the Derby rdf descriptor +This problen is there in the tests below as well +Added a css reference to the bootstrap css file on the mobile template +I think the code to push down bloom filters would have to be written separately for each engine +Closed upon release of Hadoop. +Early patch while I continue testing it further +Incorporated Devarajs comments +Alex you are compiling and running using JDK onlyAnd then the problem is AFAIR JDK ships with in the core and thus the endorsed mechanism is needed to upgrade toAlex can you try the other examples to see if they also need an dependency on jaxb? running cameldot +Or did I miss something? +Feel free to assign yourself +Hi MirekWhen you are ready with the logs Ill take a look +Thanks for the review and commit Jason! Ill keep an eye out for the +I think it makes more sense to have a directory with empty output files than a missing one if there was no input dataIn terms of the fix to the immediate problem I think that all changes to the job state should be set to the listeners +PDF filecontains an analysis of the relationship between the minCount cutoff set in seqsparse and maximum cardinality of or centroids +One wrinkle I forgot about made executeInternal +No from measure filter have been removed +Just in case you want to know +Sweet +Using the right patch name format this time +generates multiple fault when necessary +Leaving this jira unassigned and open +Sorry should be Component closed as resolved and released +Fix attached as a patch +Id prefer to get rid of the check altogether though I dont think its necessary +Here is how the tutorial looks with this patch +We actually implement things around CD now. +output of users step to reproduce attached + to + +We are going to file a bug on the JRE saying that behavior for classes that dont exist is inconsistent based on the platform and the name +Attached is a eclipse project with the my test classesFor debugging I suggest to set a breakpoint in line of file Vamsi for this excellent observation +Mass closing all issues that have been in resolved state for months or more without any feedback or update +this cannot reasonably be completed until we have an offline mode +I will look at using the with I dont have access to RSA I have no other option to postpone the ticket to. +Thank you for the reply + +See next screenshot +supported added in. +We do and we could +Perhaps you are using an older version +If I understand this issue correctly well see the problem regardless of Derby version +Sorry release notes are generated from jira +I think what Karam is pointing out is that Current might mean whats currently being used whereas what we actually mean is whats currently guaranteed + +classpath if you dont enable meclipse compatibility +To create a patch first get a svn update and then do you changes +Attached patch should resolve this +Regarding the first error its really a god catch +Youd have to apply the patch on first +The JSON output format is still the same as proposed above and the depth of included subtrees is set based on the addressed resource instead of the included nodes themselves +Closing issue +bq +This change worries me I think it should not be part of this patch and instead separated out +DavidSome reply textSome more reply text David +Added thanks +The idea is to have a JCR that starts faster up to now on my local machine when I launch the JCR unit tests the JCR needs seconds to start +Unless there are objections I will leave this bug open while I round out the testing +fixed some comments +Nothing +Bill are you using maven to build? Im on and getting errors downloading artifacts +Sorry for the delay but we just finished up the Felix framework release +caused this issue. +Marking backport reject. +Use to execute them +Thanks again! +I think that should be another JIRA +Wow it seems that great minds think alike I had actually done a ms ns environment edge for experimentation +Patch addresses comments +Have Grails go down to the embedded properties and generate those during scaffolding +I wouldnt be able to do anything beyond noting that it doesnt work so I dont really see the point +thanks for clearing this up +distance decayThe agent will break into the house with a probability which is directly proportional to the profit value of the house indirectly proportional to the probability that the recognizes thethis is not coded this comes out as an emergent property +After switching to RC some code stopped working throwing no mapping found for HTTP request + +Hi AlessioYes cxf does implement the required class but in a different package to the one used by Native as are the other types provided by CXF +iPOJO rocks +I did not mention that I have Message Consumers also listening on the same queue +It has been long time ago and it has not been released yet +The original description is not quite valid since workaround for JVMTI was removed +Jelmer are you already doing this in your app? How did you implement itI see ways to do this but you have not provided a patch for eitherId be happy to commit your patch and associated test case when you upload them +I verified the changes on branchx and it seems to be good +Grants last patch for created a which isnt visible outside of its classbq +Its a lot better than waiting for a failure we know will happen and then trying to deal with that. +This is absolutely by design +If we have the same bundle configured from somewhere in the and also we drop it into the load directory that bundle and all of the dependent bundles are refreshed transitively +The patch as it stands still has issues with the JNDI tests +Well see if we can get it fixed soon +OK I asked around and was told to file a chore +Switching all issues that have been in resolved state for more than one month without further comments to closed status +So the SQL would still be portable also for DELETE and UPDATE +Just because its documented does not excuse poor decisions +Does the system need a mechanism to addremove audit loggers? When a failed logger is fixed do we need a way to refresh the audit logger so it is picked up by the Namenode again? When you have multiple audit loggers is there a need to keep them in sync or out of sync audit loggers okay? Alternatively should we consider a separate daemon that runs off of the audit log written to disk and updates other syncs instead of doing it inline in the namenode code +lgtm +for the patch +When that is done the host while stopping and then reverts back to green +If the values in the editor doesnt fix it then that means it is likely some type of server error as opposed to us accidentally passing a null username +The initial version +See for further details. +MRI test suite is something you have easy access to +So instead of sending it back to the sftp producer just write it to a regular file or print to log or then test with stepwisefalsetrue +bq +Patched applied +committed the patch long time ago and the issue remained open for benchmark results +I believe that alone would not cause the destabilize the runtimeexception but rather could cause a security or marshalling exception +When I looked at the method I didnt see that it actually returned an Object selectable by the constants +Thanks Prasad +Yes add a comment in addBlock and in the code that generates a quota violation exception that the quota check failed because the quota is smaller than block size replicas +I still think this is a bugQuote The clean up of default egress rules is not needed for VR when we destroy the networkFor other network providers it has to be implemented in the shutdown network of the Firewall Provider resource base +Did you try with Commons Math +Can we call a truce on this? It makes it difficult to contribute patches +So far I cant reproduce the issue neither with nor with +Owen I have made the comment change that you suggested and uploaded +Is there still interest in this? I talked to a few people the other day at a meetup who were interested in this and I wouldnt mind putting something together +I think it might just be a deployment issue in the maven repo +This breaks down my ability to test my business logic unless I can mock JSF contexts +Thanks. +Reopen to document backport to +The above tickets have been resolved +code in has been pushed to which extends +How were you creating the project if you werent using the create tool? Also did it work after using it +Hibernate is very different from Hibernate +Uploaded patch addressing Todds ve uploaded a new patch which reduces the number of objects as suggested + plugin source file for the sort of is a merge point but with a properties file plus it has specific attributes on the plugin for the standard hibernate properties +int stands for integration +The failure is unrelated see +going to rename jbds launcher to jbdevstudio as we discussed with launcher to jbdevstudio and added icons for windoes and mac platformsfixed installer panel responsible for update of isues related to new p enabled product build are moved to on JBDS +Im moving all bugs out to and we can move some bugs out to to get some release action going +Ended up bumping this up to so we could use their guillaume to the contributors group +Not with TCK +Hi could you give an example please? Do you talk about environment variables that are not interpreted +bulk fixing the version info for and all affected issues have in +support is also operational +Yes I agree on the other classpath issues +i could build this feature with headless pde build also +The other thing I thought about was only changing the to make the checkbox protected and then only override the constructor and render methods in the case +Thanks Andy +It appears from Apache Flex not from user code +I have to be honest I didnt run the profiler to compare the difference +seems to be dupe of + nativesamples be fixed now +The test is currently marked as Ignore +This draft patch is meant for design review I would like to save the following improvements for the v patch depending on which way we decide to go +patch applied see rthanks once again for contributing Bruno! +It seems you compiled against hadoop? Can you do that with hadoop? Otherwise it wont compile at my workspace +Then the state store can use it as the lookup key instead of its own key +We dont need driver +Sorry Joseph what I actually meant to do was mark this as LATER +The reader gets the Codec through and that has access to all the meta dataI think I have everything I need to implement a patchMy first attempt will use the following parameters to read and write with a Hadoop codec when not running mapreducereflectionCodectrueWhen running mapreduce and going through the only the following parameters will be neededtrueSo with this we get two things +There was a COBOL program that contained years of business logic that they did not what to re write so they made a COBOL stored procedure in DB on the mainframe that would call the program to retreive the data +Assigning to Paul to make the necessary changes to pull that new release into of points Due to splitting Hibernate out into modules we will need to account for Hibernate features actually utilized within the AS +If you have a visual studio sln file you should be able to run sonar without too much trouble +If we have to Id prefer explicit arch detection along with the version and have the decision based on those +However since Derby doesnt store the timezone offset timestamps in the hour before the switch from DST to standard time and in the hour after the switch are ambiguous +Thanks for the fast turn around! +Of note is that we dont use either the JDK or OS profile activation anywhere in our build +im still running the full test sweet but posting here for other folks to review early +Since there has been some performance work tooYou didnt say how much faster would be fast enough +Verified on ER Affects and Fix in Version +Ive attached a patch for the next portion of this issue creating a separate classpath for the iut enhancer and the jdori enhancer +I am not sure that the plugin should address this use case +Theres no hierarchy which leads to the tight coupling problem just observed +This one gurads against the quurom ports +Thanks Tucu for reviewing. +However I am not ready to review a patch file of k considering that the actually required changes are most likely not so big +Subdeployment classloaders now have classpath elements that are already in the parentweeded out +unresolved in maven +bq +This was an unfortunate regression in caused by a change in exception propagation +I tried adding a +Fix adds a model validation error to models with the extension +Patch gets the correct reference to from the key which is always of type +Created a for the editorwizard page see +Patch applied +Ill keep this umbrella ticket not for details but just to show how all of the individual tickets fit together +Looks good! Please move over to the asf repo +Thanks +This seems to be an issue with the AS M build +In terms of review time the first patch looked simpler +Under this scenario there is no way to notify the application that an error occurred +So the type parameter is not allowed prohibited banned to be there +Please look for the xapi commands being run around the time the stale socket files are createdCan you also confirm the out of etcxensource file whether it is openvswitch or bridge +Shaz I can take a look at this +Oops I had missed a git add in the previous upload +this issue has been fixed in re right +So How to create a patch to this changeThanks AhsanI see what you mean the svn diff format is rather bulky and even though yourchange is very small the diff is quite largeAs I understand your suggestion we can approach this by breaking down theproject into N Change the import statements for the first set of files +Alright +For example in a production environment it will be assumed that flag will be off +out format represents NULL as N +In fact its something I was sure we would have to do somedayYet modifying seems a little bit too much to me +I very much like the broader approach as I have another use case in mind for which this kind of postresponse creator can be handyIn my application there is a validating postprocessor that inspects the modifications done within an operation +Attached client files main program used to trigger bug Files generated by uri httplocalhostaxisservicesServerwsdl +Cool I have add a patch for this on my todo list +please remember that seam is fully ok with +This is the source as it currently stands +I think this might be a timing issue if you schedule an event for the current time in libprocess but Im not sure +Just a heads up that is slated for Wed +Ingres calculates the size of the target character field based on the maximum possible digits from the source data typeThe cast needs to trim the result or use a varchar target attaching a patch +No further backporting beyond is needed because this regression does not exist in those codelines +Verified at +I reopened because of some event handlers not working +that an administrator could just make accounts in a secure environment and then have people go back to their desk and login and setup their own password +eba file with the updated contents +It would be great if we could just use the one the leader is queuing already +Hi RenThanks for your patch but I disagree with this suggestion +Hey JukkaWell if we ever consider doing documents they exhibit the same behavior so might be good to generalize itCheers work thanks! I committed the latest patch in revision I guess we can mark this as resolved now +But it doesnt occur every cant fix this without seriously hacking into windows +Im assigning to nick I would like to get this looked asap. +It appears to render and perform actions correctly in and Safari +But the friendly message may never be reached because the condition that would cause the exception to be thrown would also cause an NPE before the friendly error could be reached +The import ordering is handled automatically by Eclipse Ill look and see if theres any way to configure it +Should be better now +Hello TonyWould it be possible to map desruisseaux to the SIS group +I see some changes related to writing Map in +changed subject of the for attached + +The only issue is that at a switch or rack failure is slightly more likely to take all replicas of a block offline until the switch comes back up +Now the smooks tools only support Smooks I will develop a new tools +add code overlaps with +the code has been checked inThanks to Gordon for that +It is on the list of jiras to backport +It thus might make sense to disallow calling of all builtin methods and make someone wrap them in a actual scripted Python function to call any of them whether they have arguments or not. +Filter out private +Fixing this now +Example Thread access access change state and passivate for cluster replicationThread now using SFSB that has no better approach to me would be to use something the entity interceptor in front of the interceptor to serialize access to the SFSB +Sorry but we dont believe this is a bug in Camel +Is this going to be safe for highly concurrent environments +Selection criteria the simplest framework that will fit into jrubys ant buildI am a tycho convert where tests are easy to run +Brian Geffon Yes used for events coming from other threads +Added documentation to srcsitexdoc patch +I wish I articulated this visionThe approach youve taken here surprises me because it wasnt what I expected +Can we use that instead +We need both patches +I committed this +r on trunk +We also need to split this patch into two +Ross Im glad you found a workaround +It works but goes against modularization +Ill consider adding common customization methods on the interface as well temporarily storing the settings and transparently applying them to each created underneath +Replaced the following filenames in the EAP Installation Guide with with CXF Installation section to EWP Installation GuideAlso moved the note about installing the platform as a service on to the section about installing the platform as a service in both EAP and EWP Installation Guides +I believe this has been fixed in Axis +Committed revision for the lucene wildcardregex testsI will look at the solr problems under this locale now they probably need to be merged to x also +I do think that should be good for trunk too +Please use See for details +Recent builds of the component are not using index directory when updating index and instead create temporary index on disk to reduce used memory. +If I have problems in the meantime with String converters I will look at an eye on Web Flow which adds Java based flows with a programming model that is similar to Spring MVC annotated controllers. +Is there a suggested workaround? +Also when I do brunch w s it does the following but at httplocalhost I dont see much it just says loadingbrunch w s Apr info application started on httplocalhost Apr info compiled in ms +Eg might be due to that but never got a chance to confirm that +Ive had success using this patch in getting declarative transactions working in ariestrader +thanks +to do something like this +Lets change the default to locking +OK Ive applied the modified patch +Anakin sounds good just because it is an apache project +First of all thanks for reporting the issue Steven +Making the necessary changes to enable the use of locators for Clob +A unit test is also provided in the branch listed above +sorry I dont think the patch can prevent this crash and need to be reverted +Do you still see a need for coercion overrides +This is due to the fact that Portal is a non exploded sar containing a war file +Patch provided by Jonathan have applied Jonathans patch +esb archive modewith standard sequence ant deploy ant run ant undeploy delete paragraph to run standalone mode? see meant ant deploy ant runtest ant is no run +Run core tests manually and they passed +Still looking at the actual failures as not able to reproduce on both trunkThomas can you confirm that you have JAVAHOME set in your environment? The regression tests have a tendency to fail if this var is not set in your environment +I just committed this +Is that the case or is it always broken +Do you actually have an issue or did you just notice these in the logs +Update nethyperichqproductservlet to reflect the new name +defer to +has some very similar points +Ive attached the and java class file +No other plans currently +Wierd this was already fixed in Hibernate module but not in Hibernate +Ill work on this once I have committed the patches on +These take an URL and return a normalized one +wasadded in to address the lack of synchronization in the old client andprovide the core for a http based applicationThe problem is that we have different client classes that although they dopretty much the same thing their interfaces are quite different due to thesingle session nature of and the multiple connection poolHere are proposed requirements for this feature +Fix for descrbed issue trivial patch to +Attaching patch and regression test +ini files it doesnt come from anywhere else +I updated to +I see several other things that need to be in the main +Theres a good chance the code will be included in the next Derby release +Basically any action that requires a Designer model project will need a open model project check and if the the resulting New Project Wizard is cancelled the dialogwizard should not be launchedNote that any wizards that require model projects that are launched from Eclipses New or Import wizard framework cannot be tweaked to the wizard +fixed in trunk and new for fixing this +this is the complete folder structure and implementation formI did not know how to make it into a patchyou will see that the widgets and controllers for each application is seperatethis allows those to be moved into the appropiate applications if it is decided that that part should be with the applicationNote this is a work in progress I should finish sometime this weekendi submitted it so others may comment andor help +You can either lay this class on top of your existing openjpa jar or you can preprend it to your classpath + Please dont create any but ticket before having first discussed the issue on the Sonar user mailing list This is not a Sonar bug but a bug on the Sonar plugin so this bug should be logged on the sonarplugins project +Would you comment on that? Thanks. +the write throughput does the graph reflect compaction in? Are the regions hosted in one machine? It reinforces my argument of not using or as it is +patch wrapping in and removing not needed from alternate colorspace is used instead of rethrowing the exception +Previous comment clarification preliminary +Patch They have included a patch for the test Ive not yet applied it +Please attach a zip file of a complete Maven project to this +Preparing for release +I understand what does but Im not sure if there is any other use case than the test case? I want to refactor and as static inner classes of the test or alternatively introduce some sort of more generic in the utillity package +has been marked as a duplicate of this bug +Use the runtime it uses the groovy stub generator +Patch looks good to me with comment about how we handle +This java class demonstrates the problem +Take it as a proof of concept of how well the test data can perform +to the latest patch +At first glance of the jpa sample I got a there it should be caused by bean scope naming context is not built successfullyI opened another JIRA for the jsf bean scanning +I plan to commit the patch early next week if there are no comments +Commited with both the subdirectory change and the rename to Thanks again Bertrand. +This issue can be closed as its fixed in through other commits because now the testcase requested closing. +Closing resolved issues. +Also reproducible on windows +Got it +Added missing license header. +patch uploadedWas able to repro the problem prior to doing the fix on a local setup using a client which load lots of rows +should return it out and log extreme countsLet me try Todds suggestion next +But I didnt have time to look deeper and thought youd probably identify the issue more quickly +Then we can have serverextension serverextensiongooglewave +Definitely a bug +Thanks Chris looks good! +IIRC needs to unpacking the native library for loading requirement +Will try this week. +This appears to get the dispatcher working on Windows without breaking Linux +Here is the updated the patch +Wont fix +The processing of idle beans can probably be made a bit more efficient from what it is now +Thanks Colin +Requirements Each job has a unique name github repo branch SCM check schedule +Reason being the thread priority is a tricky thing to do and unfortunately this feature is not pluggable +Your issue is not related to SDN but to Neojs query language Cypher please try if it works if you use Neoj as please put in some effort to format your issues in a readable way otherwise no one will be able to handle them + patch is ready and will be added after verification +Verified in Developer StudioVersion Build id Build date to add release after having added release release issue after correcting release to set Release Notes main issue will be documented in the Release Notes instead of all +If else has already started on this I would be happy to give this a shot +Abstract statement not happening specific statements possible +This code satisfies make check on LinuxIt also compiles with gcc on Solaris +Unzip +Weird thing is that DOS wont print the correct values for and argument variables unless you put them in the beggining of the script +Yes well need to perform housekeeping on expire +Current trunk also affected +If you find further issues please create a new jira and link it to this one. +OK ill do it tomorrow morning +to and trunk +The test failure do not seem to be caused by this patch as it was failing in previous build as well. +Maybe Flash Builder got confused someway the last time I tried thank for the help anyway. +yes I took the M samples made them compile and made the tradeJob work so the list should be fairly complete from users perspective +Every time +Sounds reasonable +Probably it is a different glibc version +Removed namespace table create permission will be addressed in a separate patch +I cant reproduce this in a test +In the sequential design this means it will roll from editsN to editsN even if editsN is empty the NN then downloads the empty editsN and everything works as expected +Thanks for this documentation guys! FYI it took me more than hours to read it end to end +This patch also fixed a problem that and forward to the wrong context +Important contents like tables or grids +The reason we wanted that feature was to reduce the initial startup time and make sure that the printers are not instantiated when the plugin is not openedAt the moment the printers services are retrieved not immediately but on the first attempt to render the Configuration Status plugin +They work now +I think this patch what we want +which removes the content during template parse + +I just made a minor change to it and replaced the array of Headers with our new shiny +Stefan what would expect as a user? The thing is that theres already a remote repo called openshift attached to your project +Otherwise put just the newly added files in a +oh and the fact that it works with is that the check is missing in that version altogether +bq +I have downloaded and checked with and I was unable to reproduce this bug +This jira is about doing a reseek when you finish one column before moving to the nextI think we should close this as resolved +I dont think that there is any wiggle room in the Standard but I also think that the correctness problem with identity columns is not one that most users are going to care about +More changes under srctest caused by fixing DATATRANFERVERSION to DATATRANSFERVERSION +You wrote once you had some debate with some eclipse developers about it +Shall be more vigilant in future +was my mistake +Nevermind it looks like may solve this issue +The patch to revert the changes remove and add redirection from to +It works fine on we should reserve this feature have run your test with and the latest code and the test fails in the same way for all +Fixed on revision +If it is not easy to override parameters at project then the number of parameters that can be set in the general configuration are indeed limited +Pete what classes are associated with this issue +Actually every Tuple generates a and push it on a stack which is sent by MINA +All other operations do not require spnego they just need a token +The changes would still have to be backported +The reason we order by partitioner is so that we can give results back in that order which is necessary to do pagination across multiple nodes +RickLooks like a doc problem is the desired maximum number of elements +Devaraj I think its better to open a new jira +Rah I read the wrong entry sorry +This is the right one. +Changed to just to confirm +at sitesrcdocumentationcontentxdocs +That seems confusing and unnecessary since there is no submodule relationship between parent and children. +Would it be worth adding the testcase to the samples directory once the bug is fixed +Go to the Sonar interface and click on the module name The login page is in such case the module doesnt appear in the Sonar dashboards so how do you access it? As a rule please dont create a JIRA ticket before having first discussed your issue on the Sonar User mailing list +Ken A couple comments on the API proposalI think given the fact that we have pnsessiont and pnlinkt elsewhere in the API the pnsslsessiont and pnssllinkt names are somewhat unfortunate +We do have tables with wide rows and do have tables with ttls but dont have tables with wide rows and ttls together +No Binary should be removed from the store if it is still used +Resolving the issue as Later we can add this test when required. +Basically I must know your configuration to be able to reproduce this caseWhat I said is base on my source code understanding so example can be handfulRegards devMode is set to true throw will looks like a place for some refactoring +I think from now on you should be able to assign tickets +Yes this is the current situation of any jar +Ive confirmed with Erik that the stale locks are not really stale locks the problem is just that transactions sometimes take a lot longer than expectedIve created for one of these problems when state transfer can cause a lot of transactions to time out. +Both bugs involveprocessing of columns in the ORDER BY clause but are probablyno more related than that +good to me appliedMike +Yes I am sure is used its the explicit dependency set in try thismvn dependencytreeThis should print the dependent Ive tested my patch on latest sources checked out from Subversion and test passes +release bulk close +Out of date +If cversion counts the number of created children we can always learn the number of deleted children by subtracting the number of current children from cversion no? I was also wondering if there is any use case youre aware of in which it needs to have both counted +While not perfect I think it is ok for so I simply applied it +Files listed by Paul has been updated to reflect the new ASF copyright +Heres a change to that makes use of the +Looking at this POM it doesnt state any dependencies +Thanks for looking. +Sending srcmainjavaorgapacheservicemixeipSending srcmainjavaorgapacheservicemixeippatternsSending srcmainjavaorgapacheservicemixeippatternsSending srcmainjavaorgapacheservicemixeippatternsSending srcmainjavaorgapacheservicemixeipsupportSending srcmainjavaorgapacheservicemixeipsupportSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestjavaorgapacheservicemixeipSending srctestresourcesorgapacheservicemixeipTransmitting file dataCommitted revision +It still shows the OOME in my environment and it runs a lot faster since it uses a much smaller test table +So the API actually is consistent. +The only case I have found the usage of useful is for source code generation +Closing this one as duplicate of is duplicate of +Its not mandatory for me to retain the file names thats true +Now lets say the primary never heart beated or was lost +Right now most thrift implementations cannot talk to my Java server and that is broken +So yeah either way use CheersChris +The IBM Developer Cloud driver has complete functionality and tests +because irfames javascript cant change size of iframe in parent document as they are on different domains +looks good committed file revision +In I dont see standard as a common lib but it still works. +So it was a problem only when the slave broker system clock is ahead of the Database server but not behind. +Great job James but the next time remember to assign the issue to yourself +Hey AlejandroAny updates on when the code for HWS will be available to the open source communityThanksBy then of this week well posted an updated specWe expect posting the code next week still in the process of ironing out documentation examples and distro +Thank you +RAT is clean +It just makes the temporary empty file to qualified +Broker now has some messages left in the Queue that arent being dispatched +We are relying on QUORUM and on the the fixed TIMESTAMP for the superuser +JacquesI would prefer keeping HTML out of UI labels +on the fix + +And you didnt give any reasons for the mergeI think there is enough motivation for the feature posted in the jira +Ill create another jira for fixing the root issue in +Fixed +For us the is just OK and I do not see the need for any will be a good idea to add this kind of behavior for different +attaching patch with squashed commits +I think this can be committed +We could be using a running number for all we care +how to import the full jar to one maven projectmaybe add one POM task for it +Verified on. +is a replacement for this issue +that looks like an error then +Committed +Tried to register at their website too and got runtime exception from their Jira. +I also added a test that should fail without this patch +i had it on my radar but will not have time to do it for +has been added at some places since debug is off +OK Ran rat and got it down to files and +Thanks Yi Xiao for the patch +There appears to be no SAX compatible parser in the classpath +should be fixed in revIds +yes jodatime is on a maven repo I have it pulling via ivy into my local pig trunk +Hi Maximcan you provide a svn patch including a test and the fix against trunkRegards I cannot just fixed it locally and have no patch file or test case +The javadoc warnings are unrelated to this patch +Would like to get verification from a few others that tests pass before committing +Thank you for the patch! I have committed this to trunk and! +well maybe the speaker in me is the problem here +Fixed in +bq +but IMHO its good enough +will test and commit +The problem reported by this issue was fixed in revision +Do you want to check it in once you have commit rights Jeff? That way we do justice to you for finding out the location in the code +Basically a remote connection to a instance with jconsole from Sun Java works great +I have updated the patch +Im going to look into next +But that schoouldnt be a problemNext features to implement Window close when clicking outside the lookup Background fading to focus the lookupHopefully next week i find a little bit time to handle this issus +I can take a look the issue since this seems related to both are setting environment variable for child tasks +I was able to compile with IKVM +I just Googled joc descales with language set to Catalan and got back three web pages +patch has a problem i will attach a new one +At least changing this behavior would help things like supervisord +Implemented fix +My bad +This makes outdatedAnd possibly many other hibernate related issues +I will take a look +This latter sample is updated to set the application scdl for its parent +Great to see all the work on this one thanks everyone and esp Sergey +Yeah I think youre right +Should I make this a separate issue in JIRA? my failing test patch from and created new issue +to a common sample +Thank you for the have made some corrections to it so it should work under various environments now +Closing. +DejanCan you update the fixed version for this issue +Then youd get the cleanup done and the next patch kicked off right away +It should rather be according to since its a new featureenhancement instead of a simple bug fix +The patch also fixes most of JVMTI code to not invoke Java in the implementation of TI functions +could you add it + +I should point out that the error goes away when the servlet API is added as a provided dependency in the project POM +Possibly final version put on staging with working javascripts +I am attaching an updated patch that contains everything included by the previous patch plus integration tests using karafThe integration tests do the followingi Test that whirr properly installs inii Test that all whirr serivces properly installiii Test that all whirr service properly register the required to the service registryiv properly bind to the servicesIt also fixesimproves a lot of stuff from the previous patch like typos etc +check the screen shot +After working with a number of open source projects this has to be the fastest bugfix that Ive ever seen. +Some tasks not mentioned such as moving the website etc +All I want is a check if the string to be converted is the empty string to return null and not When a text box is not filled the browser sends the empty string be fixed in CVS by +Regards because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +use closure +Marking as wont fix because there was no objection +Thank you Oystein for sorting out the complexities of this issue +Closing as suggested +to be created +I think thats a fine default and shouldnt create too much hassle for the users of the site for a default selection +I think I was to quick on this one +I would love to see beta released have been waiting for this for a while +Well move into cookbook. +I suppose this issue is no longer a blocker then +ChrisWilliam can you bring this up to date with the current spatial parser? Also see other comments about the update processor +Issue is resolved +Or we could change the methods to do what we want them to right away and make new methods like get that are immediately deprecated but give people a way to change their code minimally +There was a when I tried to create the page and that seemed to precipitate the rest of the oddities. +CPU load +Please feel free to commit it next time +When you have multiple Spring within a single webapp and each of these servlets have their own app contexts you might want to share beans between them +Attached a and committed +rc looks greatSorry about this +reopened for fixVersion +It is difficult not only with JDBC component for SQL component too +batch transition resolvedfixed bugs to closedfixed +patch looks goodneed release note +My apologies. +For information it links on project dashboard if maven job sonar homepage if freestyle job +Looks like its just some date parsing error related to YAML +you install and resolve one of the bundles install the other one +terminate block recovery nicely when the block is already committed + Please see for the discussion of JDBC vs JDBC compliance and the question of whether compliance is a moving target +Hint If you need OR capability pushdown in the meantime try using IN criteria instead since IN pushdown is supported and the result will be equivalent +Updated the to be more explicit. +Run the on the remaining sstablesThe downside with this strategy is that during compaction newer sstables could be mixed with older sstables and the resultant compacted sstable gets marked with a max timestamp of the newer sstable +Closing this as something to be considered later since the direction has slowed down. +Moving all open items for XDT to XDT since has been released +I had a few bugs in the implementation first was the edge case with contexts Andrew mentioned the second was due to the fact that numeric values being used as keys in maps were not being quoted properly +I think Im going to do this manually for to be sure that I understand everything involved first +Is this not relevant to trunk +Otherwise looking forward to you finish your other project. +So does this mean everytime we need to reboot the router after upgradeAlso why the creation of with the same router succeeds? And why dont we see this error before upgrade +Counting is inefficient and we dont want to encourage bad habits +Very minor change +and in our code there isnt a reference to getCould you better explain? Is it something related to our eclipse pluginThanks +So to fix this we will have to make changes to the compiler +A better mechanism for looking up a unique possibly namespaced class is needed +bulk assign all unscheduled issues to next release +some minor changes to the minor changes to the text of the problem description +This sounds like a JSF impl issue +It breaks +Could you be compiling and runing with two different versions of the jar +Verified on Tools +Reopened issue to correct fix version. +Other tests fail in TIW +Moved to upon release of +Hi DavidWe are using i could find this information for confirming that i am using this patches +Sorry bout that +When I run on the es locale I see as a new Spanish engine message and and as deleted +I wasnt aware of that ticket +Reason is these classes were available in default config in EAP so to meet the purpose of adding them back they shouldnt be limited to allproduction configs in EAP ll leave this open in case anyone wants to discuss further +Hi MarekIf you can try the same steps with going through the Execution History Tab instead of from the Instances more execution history to see if you can see the same issueDue to the I am actually thinking of removing the execution history from the Instances More +This mustve been fixed with the recent set of robustness improvements +If we receive the donation under an Apache Software Grant then we get enough rights to distribute it under our own license +CDI setup +Approved for SOA CP +Running a profiler on the namenode showed that we are spending a lot of our time in getListing and +is now fixed + new PMD rules have been added ll create a dedicated IT to detect regressions on PMD rules +I can commit this from my client so we dont have to worry about the diff not quite picking up deletes correctly +Once you get past the point in your dev process that you cant use you need to do manual migrations or use a migration library or plugin. +Thanks David and thanks for reviews Stack! +This was fixed in Struts +Properties and modify it to work with custom encoding +This is supported on Linux as well as windows +I would like to get a with this in it out soon +The fix should be easy and lowno risk +I do not believe the problem is with my changes I am debugging further to find out why the test fails only with these releases +and must throw on invalid pathsHmm +please fix this in branch and head thanks +What is not demonstrated in my attachment is that the IDE works fine if the config file contains only app context files in its subdirs +message given the guest user has Global Observer role so I was only able to verify that as fixed +Fixed svn rev fixed using local beehive build synced to SVN MVerified that this bug has been fixed and that the code compiles +For a DB to pass the Java EE TCK a feature list has to be is actually HSQLDB Pointbase RDMSOS Sybase and Ive seen some literature on how to get a lock through so the only remaining issue is Sybase +Each method belongs to none enhanced classes and they are struts actions and holding parameter values +Thanks +we continue to evolve the site but in its essence this has been done. +BTW get should check read permission for the entire subtrees of the beginning and ending snapshots +Also should be noted that current behaviour is exactly the same as behaviour of PMD rule for Java whereas we want to behave as Checkstyle rule for Java +This is the corresponding original byte code anewarray class javalangObject dup iconst iload bipush ificmpne ldc String goto ldc String aastore astore aload ldc String aload this is related +if it wont be ready it should be updated to when it would be ready +move limits into good! Could you add a quick test that tries to serialize a token ident with a field that exceeds the limits +Moreover the problem is relevant for hashtags and links as wellThe reason for this problem is that after the message is parsed it is stored as XML +Im not sure exactly what Anils application is doing but looking at the Java code it is not creating a constraint on the table at all +Thanks +Maybe I dont know how configure for that +It passed then failed +Attached patch that has validation for the notifier page +The problem is at server start the actual configuration is used to determine what is deployed +Not for commit. +Currently there is nothing interesting in those +Ran the failed test locally passes for me +This is needed for the product documentation +Not sure if I understand this entirely but as I read this thread it is in some connection modes possible to shut down the engine underneath the network server in other modes it is not possible? If so it seems wrong that it should be possible to take down the engine when the server is running +I committed Jesses debug patch to trunk +The aborts are caused by the compiler bug described in +ok +I tested with the project and with our unit test in our product +So lets mark it as resolved +Sorry for trouble +The intent was not to create more syntax outside the SQL Standard +To start with I will look at various subclasses of Optimizable to see if we rely on the statistics in those subclasses +I have applied the patch to current svn and patch is committed and is available in current svn trunk. +To move the listeners field from to this issue must be I guess since this need further discussion well move it into +Makes sense +Hi JacquesI have updated the lasted code I changed I created a without cursor may be its not the best way to fix this problem but it can work now if there is any problems please let me know +Closing as duplicate. +Aleksey is the problem you mentioned is caused by this patch? If yes I can fix it and submit new patch ASAP +And even +Looks fine. +For the current this would mean there should also be a implementation which directly hooks into the internals +And this also contains the wiring among the pages +Strange that this works I checked it with the unit test and it hungI can offer just a vm cluster too +Ive applied your patch and already have the skip thingy running +could you rename and upload the images? Perhaps that would be the simple work around +In the second case it does not run runuser for performance reason +I am actively looking at this +This feature would be a great which might remove lots of rework and perform searches faster +Move to +In WAN andor when the network is busy the round trip latency can easily be in hundreds of msbq +Hi RalphThank you indeed +This code is rather susceptible to be improved and to be used as a basis for richer examples +I want it to work with the taz demo openlayers only so we can show off the capability +slip to next to AS was unable to reproduce it however looking at the trace led me to believe I understand the project is associated with a runtime to get the proper jars +Thats a known issue leaving open until we can update for a version that isnt affected. +The old admin gives results thoughBTW this whole effort is a makeover Im glad youve taken it on +looks good to me +Patch for this issue +Done +labels +This looks ready to commit +Bela will decide +Perhaps we need something like and for +Sandy +so we need to guard against thatIs there any generic Linux OS mechanism that would make the running containers die when the NM dies +Thanks +Hi Ben +like any fix in binder is ever trivial +Have you noticed anything dodgy with this +Please give the snapshot a try and let me know +Im not positive how to setup the environment to test the if branch +In case of a reference it has to be uncompressed +Added to the help info on property +Additionally Im worried that making this architecture change will tie our hands in the future if we want to make any calls from the NN into the Plugin using hooks like inI think we had talked offline and decided that as long as the old Thrift server remains in tact your use case wont be disturbed for now +As of you should be unless this was previously documented as a API then it cannot be removed in the codebase +What is more this value is likely to need to change as the cluster fills up with data and as data is deleted off of the cluster +Initial triage for release suggesting postponing to move back to if you disagree +The class added just to pass a proper context to use counters in if you know the better way to do this please let me know +Updated patch has the suggested changes +Because nobody is interested and I see no effort for a long long time +static inner classes as necessary for concatentation of strings and not String concatentation as the latter is immutable +In some far futre release deprecate the option and clean up the old codeI dont know how intertangled the two implementations are +Patch using only checked exceptions as described +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +This is good one step forward +I am aware of the redundancy it was painful to add it I think refactoring that would be a good next step +We should add an extra check and skip the scan if not needed +A partial workaround could be by adding on to fix for +Prefixed bean names +moving this to Beta since we already are code frozen for M and by far enough to do for M +I just wanted to make sure this wasnt supposed to include some actual tests as well +There is a bug in handling of rows and column views of matrix views +Warnings should be viewed as a hint that you might want to rethink your code not as a mandate to revise itIn any case if you believe you have a solution I imagine that a patch would be welcome +this was resolved with stub scripts closing +This sounds like a flame +Which tools are used in Apache Mahout? Apache Mahout is provide suitable for supervised learning or unsupervised? Which preprocessing methods tools in Apache Mahout +These tests used to pass before the project X parser wasreplaced +Attached test works in the same way as on RI now +Im not saying this is totally correct just saying why it behaves like this +Thanks Shawn. +I will try to do my best but i dont promise anything +I guess that the group navigation has been cached and didnt reload after changing navigation we also had some bugs relating to caching data +It sounds as though I should see another provider under the? Or am I misunderstanding your first para +Patch to implement this new Feature +svn ci m refactor to use for IP hostsSending srcmainjavaorgapachecommonsvalidatorTransmitting file dataCommitted revision +I have fixed the NPE in +Added optional wording implement elevation for end attribute tests and toString +Nice work Gabriel +Verified in now both JDWP tests and Eclipse debugger works fine +needs to be updated to read comma separated since thats whats being done in +I forgot to add the enum class name in the default value I used in the example +Patch attached that should resolve the issue +Would changing this have an effect on compatibility? If you wanted to support this change over a rolling restart or anything like that it would probably be rather complicated or impractical +I could go either way +Apply this after the patch for +Closing all resolved tickets from or older +Yes the solution would be sufficient for my issue +If it is ok I can commit this today +Heres the patch per my last post +An option to switch it off is fine we can do this against an inlined version as might be the biggest concern with Olivers pull request actually It simply uses Objenesis when found on the classpath +I have added the style to + +in +Still should be easy enough to work around a null value in that case as well as you +well the Red Hat one should be the one that works best so need to figure out why its failing in this case +in the patch maintains metastore client in tss which makes it thread safe +Thanks for the review Stack! +Groundworke Groundwork Monitor for RHX +Ran out of time to get this in pull request. +Can you please provide more info which version of are you using etc +All of the code we have written that consumes python thrift already requires +I am horrible at naming +But decided agaisnt implment it from scratch since I thought it will the wheel +Do you suggest to let base class implement this interface? But you still need to construct and initialize the processors even if they are wrapped in the interface thus my suggestion for a client side version of the factory +We might also consider replacing the with an injected executor which might of course bring a whole new set of problems +This is cruft built up over time +It should just skip encoding data if data is NoneI will work on pushing a fix to trunk tomorrow +I dont think that it is a problem to ask to the user to add a dependency on its project to use jimi even if this jar must be installed manually on its repository +You patch only moves the synchronized block out of the try block +Adding the jdt dependency to config jasper +For legal reasons I have to ask you to post a diff with the licensed to Apache flag turned on for me to work with this +OK +Thanks Jeff +In my opinion the parser should obey only to the encoding seen in the XML declaration as it has no control on whatever envelope was used to transmit the XML +So for example you could have a project crm in contiuum and give the team crmteam from ldap the necessary rights in continuum by populating the authorization group two level aproach to authorization is a very useful thing in big companies. +The method that I am looking into is srcwsdlwsdlwscppliteral and srcwsdlwsdlwsinfo +Modified tests as mentioned in the comment above +Please see my comments on RB +Where is your libpthread installed at? Is it in your library config +When this patch gets applied please upload some snapshot date build into maven repository +unable to reproduce withEDG Alpha SNAPSHOT revinfinispan last commit baeedceaadbcedff Mon Jul AS last commit eebafaeaeddfadb Mon Jul will reopen if reappears +update patch to reflect comments from +I would prefer that Apache s default behavior remain although we could possibly enable support with an property +Does the paging manager already contain this informationIf not how will you compute it +yes checksums that are reasonably strong do you need SHAMD or CRC enough +Im unable to replicate this problem +The implication as also documented above is that nothing that relies on an ID works anymore +Fixes JIRA fixed on the branch in revision and trunk in revision +Includes unit tests with Dumbster used for testing of sending SMTP messages +Will try to see if that fixes it and reply soon +binhbase +Did not make the reviewing old +I agree that it should be replaced withGeronimos for licensing issues +WTP will package your dependent jar into lib as it should! and as maven doesWTP will also do the hot deployment of your changed classes during debugging so no need of copying classes +Committed not in so we have time to tweak it if necessary +Thus when Iogging on BK there wont be any storage directory available when reading from bookkeeper so nothing to read +However with thenew patch it seems to be with the same timeout as forstart of replicationMinor Some has been added to previously empty lines +It looks like it is related to IIS web gardensIn our live environment we do have worker processes in IIS hence the three log files +change int to the sizet +with latency going way upIm up for more exploration if you fellas are +It would be really nice if we could provide a URL to redirect the users to an External mechanism to be able to create and account or change their password when using LDAP etc. +If i were a regular hadoop user i would prefer the issue that deprecates anything to be listed in Incompatible changes +I have implemented what I described in my last commentI am not very sure about the correctness of the implementation regarding the wizard creation it doesnt seems to be well supported by the JDT API +See for this issue opened against. +What stacktrace do you get with +META +at version vector contextcomments appreciated +sounds better +didnt pass rawOffset value to ICU and subsequent call to returned old value +One more fix +Sorry the patch name is wrong its name is but its just a typo its indeed a patch for this applied the patch +Attaching consolidated patch of all changes previously attached to so that we can get a Jenkins run here +A year ago I wrote a custom listener similar to to do this environment independent log level control +Page Second paragraph should be +Im guessing they check for null after the resource lookup now which is why they dont return alt and title if the message resources are set to nulltrue +What I am saying Martin is that if your intended behavior is that if your threshold is bytes then anything smaller than that will be retained in memory and anything larger will be written to disk then there is a bugbecause files of size are written to disk even if the threshold is set to MichaelThanks for your insights +Keith could you review this change pleaseThanks reasonable to me. +It is +Try running Warbler with MRI vs +Similarly I had to do it in a lot of places in other jsps too +Code review requested +Closing all resolved tickets from or older +Added new plugin on users profile page that shows the posting history +It cant just be ia so I choose IAW as defaultOf course we can make IAN as default if you think thats better +I expect that when we add tests to ratchet up the code coverage that we will end up with some tests +I can totally see a badly behaving client opening a lot of scanners enabled and then were stuck with a lot of garbage in memory +I have applied it in Rev +I just committed this to trunk and +seems pretty similar +I updated all the scm urls to not have incubator in them but it didnt seem to pan out so I just deleted all the projects and them +Probably PRC should have it. +If I switch back to there are no signs of it +Work around is to remove all comments from the fileThe shipped example contains the Apache license headerRemoving this will not cause a problem but will ensure you do not corrupt the password file when editing is done via the MC +Patch looks good +Test setting column from unique constraint to null able should pass +I cant reproduce this issue with either or the latest snapshot code +Thanks Ed +Can you point to a website that this patch worked to pass the form auth at? I need to verify that it is working for me but cant at the momentThanks in advance MaxIm sorry but I dont really use the patch anymore so I wouldnt be able to tell you +For me its not working too +Patch to correct problem +I think the particular example below is fixed with latest patch from +Thank youWill this fix be included in the next version of M +I think Ive finished this already but I have to check Ive fixed this already just tested it again +I am also not sure if it is really a critical problem +Closing the issue now that it has been resolved. +But I guess it is good that it did get fixed I just wish I knew how +Later was introduced in the file xalancInclude a template doing the job regardless the type. +Hi JacquesAttaching patch to fix following issues in ebayStore component +I have installed Railo on windows machines and all do the same thing +A lot of people were using the older patch +Feel free to post the rest of it here just in case theres something we missed. +I dont think I need a different API as Infinispan should be able to figure out what it optimal to do according to state and configurationMaybe a boolean could be added to the cacheentry to track if the value is already present in the store or not +Heres a patch for trunk which addresses the issue +Here is an implementation of what you suggested on the mailing list +Thanks Sharad and Arun +Attaching +We are looking why in this case it not works +Working program with fixes for +Added Darrin Misons patch to the style bundle +Basically commited but to activate the fix I need Hibernate core till the next core there a way to define the type of the index +Thanks Brett for the patch and Siddharth for picking up the testing and driving it to commit +Jay can you provide the content to be added +The NN should release the lock when it finalizes its current segment +As mentioned in making the rpc listen only on one interface fixed the problem. +Hence closing the bug. +Basically weve asked doug to create an approach for showing a simply example of launching from the command line +Hopefully that will happen today +Mircea how about a plugin for the new framework that uses different types of string generators as defined in the benchmark config? that is doable +for fix +I will merge the buddy to hadoop code and upload the patch v late +Thanks for the info on the ZK stuff. +Marking closed. +Resolving +do understand your use case but I still disagree +Sorry Uwe the policy for allocation Hudson accounts requires that you are on an Incubator project or on a graduated PMCDiscussions welcomed at buildsPlease this issue if I your status +oops sry I thought I marked this as a feature not a bugYou are Mandrikov can someone please respond to this? Our normal builds take with parallel builds and end up taking m on Sonar buildsI understand that the Sonar builds will naturally take longer but thats a huge lose and a potential reason for people to not use Sonar +Use of new logical operator plus bug fixes +Including Ashutoshs suggestion to avoid +Just for the record +Could you try again? Youll need to delete the cached zip in CUsersWeipeng +ok i am sorry i didnt realize what you are sayingtaking stuff from solr and putting in lucene without fixing solr to use it from lucene is terrible we should not do thisi think we want to remove not create duplication +Oh my godIts so strange +looks good will test and commit +Willem should this new feature be scheduled to so we can start feature freeze and close in on a release +Minor change to allow the configuration to have more than one in any sequencer. +We definitely need to do some changes once annotations which appear in other places can be created or if we should choose to support such annotations now +It is a CSS bug. +Great work Filipe +Simple awk line that dumps to a variable +Attachment Advice for has been added with description fo has been added with description image has been added with description image has been added with description image has been added with description image have attached the fo and images file for your reference +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira +Patch applied as discussed thanks matePlease cross check the fix and close +simple benchmark python beforeWrite Read afterWrite Read +to separate mapreduce and job classes +Assign to during tidy up prior to resolved issues. +Also you can take a look at the package structure of jXLS +uDig works fine when I drag the WCS links and WMS links to it +YangYes i meant using EXTENDEDMETADATA as a string +Maybe fix for can be unnecessary but the same fix for doesnt work +Could we add a validateConfiguration method to +This is fully workable solution that depends on Im asking because it is currently not marked as Patch available + +All Geronimo instances will likely have to have a second repository just for local deploying +The release plugin uses this to stage releases +I have not rich experience on hbase do you think it is often required by users to specify other arguments +to close since we have no repro +I just committed this with a few changes added a few more checks made it so that so unit tests will not pass if checkstyle fails since checkstyle is LGPL we cannot bundle it +Looks like also bincqlsh doesnt have svnexecutable set +I committed this to trunk and +Work has been committed but a test is required to validate this functionality +Yes this is clearly a leak +bq +Could you please provide some info regarding your implementation +My thinking is that the Java API would initially only getset values + +Im fine with either approachWe still have to look at thread count usage with the this though and do some general checks +Barry helped me walk thru on my mac and was unable to replicate issue +That should work +Status? Can someone try this out +Thank you Lohit. +This is a duplicate showing up in the dependency yes I can do that the attached zip has my and file +Agreed + +It also would allow us to use tokens when security is off +I just rebased and didnt look into thoroughly +I need to email opts to have the set this up +Ive just committed this +My svn co had some messed up state +Opened to get attention for this. +And this makes the decision whether to bind the session not so easy +Worth looking at for +The physical Portal Site component site view implementation has been tested extensively with the new and appears to be functioning as intended. +Could you please give a try with a latest version of Ant Task which is for the moment +Aggree the latest patch in covers this issue I suggest to resolve this issue as duplicate +Strange +support will be replaced with +verified onsubversion revision on work properly. +Sounds good to me +Applied the volatile patch in revision thanks! +Actually I am wrong in interpreter mode the catch location is incorrect +Then the reader wouldnt have to flip back and forth between this class and the +No response assuming ok. +Mark go on with the RC! Thanks a lot Uwe! I really appreciate the speed with which you have been addressing these RC issuesIts taken a while to get this release out but we have barley wasted a moment in cranking along +Also added a menu item to the Teiid View that enablesdisaables data preview +I created a patched remoting jar for the hotfix since it was broken in the remoting jar that shipped with CP. +This issue has been resolved for over a year with no further movement +Bill did you also test the other patches +Nkeywal +Marking as out of date since the old xml connector logic no longer exists +Not sure what we can do about this +Tom no really I think they are separated issues +They are easier to merge and easier to see the changes +Havent seen this for a while +binhadoop namenode upgrade +implementing write operations correctly would be quite tricky since they depend on jdbc commit +Yep youre right +Thanks ArunYour patch is in trunk rev +I havent done much profiling so far +Its written now and currently in default serialization the field names are also written +Sorry +Thereby across application attempt response we can lock on specific attemptI think this is important as we may loose container than what were requested +This screen is in component +thought the naming scheme and using the spring standard registry was common behaviour +OK to push to master +looks good +Should we not close this issue +Kiran can you confirm that you were able to reproduce this behavior in and that its fixed in trunk plus provide some information on how we can obtain such a duplication of createThanks +The tests are still dummy tests but the actual builtin works +fixed in r. +I assigned this issue to you +Will commit unless there are objections +It isnt that it is possible to try to workaround these issues in the Flex SDK codeAs grows up it will hopefully be the case that you can use Cordova with Flex to create a mobile app without these tradeoffs +after release. +Patch with Hibernate bootstrap +Maybe you have more luck applying this one +Thanks for raising the issue and your testing of the solution. +In particular because it acts as a pool using core threads and an unbounded queue adjustments to maximum have no useful effectWeve wanted bounded queues in the past and we definitely still use growable pools in places so thats another reason to keep both +Moving to CP to look more closely at this matter +Thanks a lot for translations +Thereforethe that serves META will be considered more loaded thanothers +A few more applied many thanksThere are certainly many more fields which can be made final because in the first versions of Configuration the concept of final fields was not taken into account +I dont get it there is no node with token +merged in MODCLUSTERGACP +Thanks +This patch doesnt add authors to readme nor to the website since I think they arent necessary when everything has been developed within the ASF +Actually both jobs use the same source tree +I am far from being an expert on this area of the code +Unsetting Fix Version for unassigned issues +Michael B +Thank you for the patch. +Definitely ok with considering disabling the truncate test until it gets fixed but I wish we werent claiming this work done when its left us in a broken state as far as the build is concerned +Builds locally +Documentation is coming soon on the wiki +Trunk cannot pass this test +Patch to make match in honouring JAVA +HiI have the files in there as part of the output parsing tests to ensure that we dont accidentally try and parse the logs +Thanks Harsh! +Will post back +I view returning null without any indication of what the failure was as an undefined behavior +This is just one possible solution to the problem of the long test duration +Without any feedbacks we cant help you +Changing ctors to private was confirmed by running nunit test which is marked as testing mentioned by Digy in original email Andy +Again +The snapshot driver JNDI lookups from a JSE client arent working so I couldnt test the snapshot driver +Yes +There are certainly others reasons as I removed all and devhgofbiz in the file but still get conflict in +Remove the full build and promoted the bare build to be the default build in Rev +conf doesnt exist the corresponding changes are stale +Sounds good +That is correct +Example if Interceptor is entered it will list any class with that in the name +HiFirst of all thank you for patch +has been marked as a duplicate of this bug +Committed at subversion revision M javatestingorgapachederbyTestingjunit Added a header comment exlaining how to configure your environment so that the XML tests runM javatestingorgapachederbyTestingjunit Fixed a javadoc warning while I was in there +Maybe you wanna take this on too +canceling patch as one of the tests is failing +I tried build of beta and it now works +If we know that name prefix resolution causes this amount of pain shouldnt we consider adding addressing in +so you dont accidently release before this is fixed +par I can deploy it via the httplocalhost web application UI just cant deploy from JBDS +yes you can mix xml and annotations provided that a given class hierarchy uses the same there a status on this issue +Workaround is inReplace with attached file in issue in +Ill open new ones with patches in the future +Currently we run a thread on the namenode but that thread could be run anywhere its just convenient to run it on the namenode +I should be able to commit these mods and build test passed on winxp and lin RHEL gcc +If necessary we could check for the presence of a default constructor and only actually use Objenesis if we cant find one staying with regular CGLIB construction otherwise +Tested +Attaching patch +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +No reason at all it must be a dirty default +Marking resolved becaused upstream issue it now resolved +Attached patch fixes all the interfaces +Perfect the tests are validating both logsstatic case as well as the pages +Also I have it set up in one of my project to go to mysql but I didnt make that change in the sample! Im a dumbass +I made a directory named which is located on root directory +Given this it seems like only other way is to offer this via the component and not +Any update on this? Seems like it should be fixed for +bulk defer to + +BTW shouldnt the abstract internal methods be protected rather than public? Theyre not very internal if theyre public +a patch with the dutch that sorry looking at wrong naming convensions +It allows for people writing API classes to perform a check to see if its successful +Linking the patch that added the problematic code +This helps just in case the original owner is also working on backporting it to +Master startup should be quickThe MAXIMUM FLOW solver is a good fit for Locality Checker +It sure seems as if the JVM has picked an exotic class loader to put it in but I cant find any evidence +Proposed new license files +seems not a fundamental solution +Thanks for fixing this. +Ha BrunoJust forgot to say that you forgot some tabs +This would not only allow the provider of the progress monitor to know whetherwhat problems were encountered but it also obviates the need for additional parameters to all the places that already accept +I am still getting LF endings on windows +Id be on the mailing list and not in JIRA at this point except that xircles is broken and I cant subscribe +So I removed methods duplicates +First shot at this +adds documentation to the reference guide +Since users may skip versions when they upgrade it may be just as relevant in future releases +It ends up uses surefire +Logged In YES useridI have reverted the and branches to the release to address this issue +It also failed in both my linux and windows machines +stands for Table and stands for Table m stands for Table m +Thanks Gunther for contributing! +We could generate the static docs as part of the build process and have the monitor page dynamically generate the docs +there any plans to address this issue in the branch of code +I have now attached the file to this spreadsheet summarises the split between generic and test +But we werent always consistent in specifying the index filename and instead of raising an error Coordinate would just return something bogus +There are other complications in testing this as well since Suse aliases sh to bash in sh mode so you need to use dash to get raw sh support +I can see how this works for the WAR version as we can provide a SQL script that you can run against the DB using any DB utility +testfix all test revision Thanks Mark +Olga sorry for reply you late + +Wont be done for +Can we change the title of the issue to NPE when install target is missing file and pom type is JAR? +BUT Where I CAN DO THAT +I could not find any other explanation except that at some point some converter was using the wrong time zoneIn the end Ive fixed the problem setting a default time zone before the execution of the testAfter a quick search on the web this seems a good thing to do anywayBy the way before Ive changed it this tests used to fail sporadically with the ehcache tooThere was an obscure comment about it in the code with no explanation given maybe you know something about it +I think this code has an infinite loop coded inside if you push a buffer back into the stack and then pop it back you will be looking at the same buffer for everI will try to code something equivalent +Which is official way to contribute my fix? +We used it elsewhere +Will add it to description +I think we should add the shutdown hook but add an option to turn that feature off in case you dont want it +Maybe you can add this for all the queries Invalid for delete by query because +on the latest patch +Sorry I meant branch +Since aggregators are usually registered in preApplication the master cannot use the aggregators to send a value to the workers +where is this happening +I tested this on Linux and it worked OK but it also didnt have the problem described with with spaces in the name +I think we can upgrade our pom to the most recent version then +I think there is a typo on line should be Regarding Bryans comment on renaming the property +Hi Aidan can you review this change please thanks +With this patch wouldnt you still get as many notifications as watches you had set for the znodes you have manipulated in your batchIm still interested in understanding if the performance difference you claim matters or of it can be fixed some other way +Test randomly fail often +I tried applying it but the patch failed +Committed patch a to trunk with revision Ill wait a few more days with patch a hoping that someone will have a look at it +Jason I see where youre coming from with bad web citizen but it refers only to a specific usage of as a traditional web server +Also it is missing functions or procedures is form of a function +Pushing this in +Either the setter or the would be a clean and simple place to create this yet Hibernate wont let me because of this silly exceptionI now have to code a very ugly workaround where I manually track stuff and build this Set after the first to its setter +It works for me can you confirm this is not a build issue +I have a and version but neither are anything to do with the original arjuna interop consists of a war file containing the interop tests to execute and the proxy for logging all traffic +The bug is invalid +This error is caused by the reason that we choose a file to generate codes +Adds support for parsing should go in trunkmodulespacksrctestresourcesorgapacheharmonypacktests location next to the code + and into a single Option type +This is the error that Xerces was failing to detect +It seems these two bugs will be solved with the suggested replacement as wellASF Bugzilla ASF Bugzilla except the unix regex must to extended with an e link type +Hmm the injected entitymanager is this simply could the class be made Serializable? Currently it is not +New FeaturesYou asked for input on this subject Central Cheat sheets for Start from a sample project wizards are now automatically opened when projects are created and the cheat sheets can be subsequently accessed via the Show In option in project context menusSee the attached screenshot for an illustration on what this looks likeand Server Tools Additionally you can choose to either stop or undeploy anapplication from the server via the administrative consoleIm not sure what your question is here +I verified the tests failed before the fix and passed subsequent to the fix +our discussion this is a fix way false should be setup as defalt property for Console Configuration +Done +Patch for branchAs Enis suggested private class is introduced to align ordinals for classes across major releases +since it wasnt the root cause failing +You are right that it must be owned by rootThe error message really is completely correct Must be owned by root and not writable by group or other +This step wasnt happening with certain skipped keys meaning they couldnt be reduced properly. +Khaled has committed the code in the patch so resolving this issue as fixed. +I suggest simply using static imports +I just committed +everything youve provided looks fine +Patch makes sensePlease run through security test suite and let us know the result +Where would be a good place to drop in some sort of +Nice one Trevor +Error handling has been taken care of in a seperate issue a red banner is displayed for a short period of time at the top of the screen to indicate a server communication error +I wasnt happy with my conversion to because of the lost info +Resolved as Wont Fix. +I agree the term deploy is not intuitive +testing on Windows I get rowss with hsha instead of with sync +If desired I can help with that +Patches so far have been applied to the trunk after branched on Aug so I updated the fix version to +Patch to fix the layout errors in the docsAlso removed all tabs and replaced them with correctly indented spaces +Flat structure is still slow so Ill keep investigating that +I would say that of course weighting more highly linux and solaris should be done over results on macs because while I love my mac Ive yet to see a production cluster running on Pros +But this is not necessary for further problems +In case the reused container got new configuration but it did not receive logger from container thus replacing with +This should give us more info the next time the test fails +Thank you for splitting and showing kindness toward the reviewer for the patch +I think this was against the master branch so not so relevant +Just a quick note I probably hit the rsync server more than two times today and yesterday as I was doing the initial checkout of the repository as well as testing the cron scripts to make sure they executed properly +So another only needs a few times more connections to run out of memoryEdit Most of these are writing new blocks +But I am not sure of the main title +remove period from javadoc use instead of to mention jira for other places where we need to use read range checking +interface is a bit confusingLets wait for clarification from the first +This is because NN may get shutdown any time +This is Directory Studio version has been released. +Documentation is updated in. +In My opinion the transactions that cannot recover should be just quit +Im a little confused because the code which combines the list from get with that of get has been around a while +Go for it +It depend on Ubuntu smoke kernel +Ive just committed this. +dberindei does concurrent writes so invalidation failures are legit +I will take care of this. +the wsdl file needed to reproduce the stub code generated by release will be end of this month +There is already junit test case testing theCould you apply the patch +I was thinking of the general failure scenario but perhaps we can ignore it because both parent and sub shard leaders are on the same JVM +trunk +How about making this a core component? +Segment starting with logical offset on broker does not have continuous logical offsets +I have revised the text of tree component + +This issue has been resolved for over a year with no further movement +So the discussion at one point was to ditch the daos and just have maps in a repository and put it into a test module weve been discussing +looks good will test and commit +When the dependency from to Foobar is recognized that is when is not possibly two issues herea Why does not record a proper dependency on Foobar b When it does why does the compile of run without is ready to go but too serious to put into it will be in a dev build straight after ships +and then it makes sense for Split to give the ordered Tuple return functionality +ll take a look at this one +Ive created issue and Ill try to do fix and release as soon as possible. +Feel free to reopen this if you find any problem. +As for the is there any chance to make it fully dynamic? As in avoid any need forextra setting in the configuration file +Those indices match of the nine literals in the query and should be the preferred way to answer +Moving to want to work on it but I have a busy work schedule ahead of me +Subtask completed new JBN property is added to config +v. +Currently you cant define a mirror that needs login info either because it seems to ignore the server section for something defined as a mirror +Making a fair bit of progress here +In addition to the patch I had to change some of the tests since they were also +It probably is more safe to have a spring feature excluding orm and tx and an extra one for the rest +Please find the attached file we have an attribute something like sorttrue in future release so that we need not have to do any work around and decide the sorting pattern at the tag level +please add the following to the doc as a NoteUnlike Xenserver and KVM dose not have the capability to mange the memory allocation to dynamicallyCloudstack sets the min and max values of memory that a vm can use +toolslibDid not work JVM could not find toolslibThe iterations could take forever so I aborted and just copied all files from toolslib to lib folder +We might be able to further simplify it by using a base class as Doug suggestedI really dont like the extra rpc for any method that has a pathname as a parameter regardless of whether or not a link is in the path +Do we need an additional request field to indicate we want itThe current logic already returns the AMMR token only for unmanaged the AMRM token only when in ACCEPTED state Typically you monitor the state of your launch AM until gets to ACCEPTED they you proceed to register thus effectively the AMRM token will be returned only once +Thanks for fixing this Jeff I really appreciate it! +Here it is +I have made a patch Please review +I think we should wait until we got started withIf you got the documentation ready then you can attach the text to this JIRA in the mean time +merged into master +This issue is closed +Thanks Richard for the example +So I close for suggest adding a error message reach eof unexpectedly +Was not feeling well last nightVerification on hadoop passed. +bq +ok + +This error is still present in Wicket +maybe something elseThe above omissions can easily be but lets see the review of the existing changes first +Oh sorry not needed Ill reopen this one +The issue is with when response has validation errors not you want I can create a simple ws for this +Good catch +Can we convert this into a wish issue and keep it openI dont like closing issues just because noone is handling them +Were planning on a repair tool for region unassigned patch available +rename it to omxmlsec +Thanks Rick +Things can be passed as bytes in Pig by passing them as bytearrays +John good catch +I verified that the tests still pass after updating the patches for and +Are they in svn somewhere +Thanks Suresh! +Yes client module should be always loaded as RESOVED state should not have gbeans started +Looks like all of them has as Copyright holder in the LICENSE +Trivial patch need to be some discussions with Herv here are a summary apt sink produces ASCII document xml sinks produce documentsWhat to do with others text sinks like twiki or confluence? I proposed to escape all characters similar to APT. +Bulk close of old resolved issues. +Committed and to subversion as revision +However I cant see that this functionality has been implementedi did not either and that is why I thought that you had thought about stop +Before we go on this can we please make sure to do a little benchmarking? Java is pretty good at optimizing for repeated strings so there may not be much performance to gain here +committed thanks Ram + +Commands sending is done via a Writer connected to the old socket and thats why I get a Connection reset in IMAP +I forgot to say that the failed test is not related to this +Hi Per Ive added you to the JIRA Solr contributor group which enables you to be assigned to JIRA issues +Im trying to find the code position that causes this problem +Looks great Kim! KnutCommitted patch to documentation trunk at revision +it remains an the cluster password is not set there is no way the cluster will work +Specifically the cxfseendpoint is simply not generating a new class in of these servicesQuite baffling +I will try collect more data and update the Issue +This should be fixed if you are using an EAP build after th July see rev +Just committed a fix. +d DEBUG and b works as desired +edge with Nailgun grep s. +Modifications for bin are not tested and might be incorrect +Im trying to figure out the whole picture +Ill close this issue if I dont see any argument in a day +for each method call +Also there are no errors regarding JCR or content initialization in the server log. +I tried it withNo chance ti deloy +If i reproduce this issue again +Now the test works for me and I hope the EGA scenario will work as well +It would be nice to provide an official solution for this in any you provide an official solution for any time soon +we have evidence that it actually helps in some cases + AWe would probably have to open the Configure Maven Repositories wizard instead of See how to setup +to trunk +There already exist lots of solutions for service management we assume that the user is already using something like puppet daemontools supervisord cron etc to make sure the daemon restarts eventuallyI did not find a reference to an external monitoring tool in the HA design docs +Moved to +axis log but not a blocker +The class using modified compiler are you using? Suns javacIt would be helpful if you upload beforeafter the bytecode transformation by you +Closing issues resolved in released versions +I think we can currently achieve this only like the old highlighter does am I right? Maybe we can make this pluggable and have different implementations +I just havent found time to properly investigate yet +looks invalid +I believe the additional code Ive added to has enough functionality to solve most use cases you present +Patch which includes the demo code +It is something in the nodemanager tests +Closing all resolved tickets from or older +I need somebody who knows what they are doing to point out what I am doing wrong +Getting annoyed and walking away doesnt help anyone +Initially was added only tests for bundle installation +rob This is just a simple extension to the Bundles page which shows a summary of the status of the bundles installed in the system in terms of how many bundles are installed how many of them are resovled and how many are started +Created for this +Trunk doesnt have this NPE problemWhy not? From a cursory examination of the code it looks like trunk should have the exact same issue +I propose the following spec change in chapter on page Add the sentence Property names are +The only limitation we have here is manpower +hyp files areserialized objects which are serialized during build processafaiu the exception means hyp file cannot be deserialized into object due to version inconsistency +Thanks Sijie. +Moving these to back to for now +If you havent removed the corrupt files you need to do so +In a new format data directory the raw data directory path is supplied to which bungles the conversionPatch in testing +I didnt see a question on the maven users list about this so I posted one +I guess my network connectivity and EC performance would do the difference +Or did you spot a more obvious problem in the +New patch w above feedback incorporated. +Even in that case theres no warning in the wizard +Im not sure if it is possibleThe wsdlxsds are provided by one of our customersAlthough Im on vacation currently I am going to discuss this issuewith my colleagues tomorrowRegards Karl +However in code first approach generating an rpc WSDL is not used much +Good idea +I will report back when I know more +Because of that I had to change the test to not look for nanosec granularity after the merge of into codeline +Looks goodRe Well probably need to redesign metadata handling at some point to better reflect various different types and sources of metadata +It seems like with jars and security manager on we are able to dual boot the database which would be a bad bug +htmlRight +ExcellentIndeed Im using the HTTP transport without Jetty hence the NPE as Willem explained all resolved tickets from or older +Is there any overlaps with? I submitted a patch to modify the to have support for nestedPath extending the existing code used elsewhere assuring consistency +Obviously taking the latter approach would be preferable from the perspective of the OODT project but from their point of view there are pros and cons to each optionCan you enumerate the pros and cons? Heres my general rule of thumb if something is experimental or requires changing a bunch of interfaces or adding code then first develop in locally in a project repo field it experiment with it decide if and how it can be made general based on the above experiencebq +tested with +Its working with JBDS +Feel free to post your suggestionscriticismquestions +Effectively our circular reference resolution mechanism is a best effort kind of thing it tries to get by with passing in references to beans that are not fully initialized yet +We need to handle monospace as an escape for other stuff +Honestly Ive seen a lot of creative CL usage but this beats them all +Fixed and Verified +Good suggestions David Ill make them today +The caches could be kept there instead of whose instances may be shared by multiple threads +Resolved and merged to the this is still not working in or trunk +i added the class for this refactored the elements of script to variables +Bulk closing stale resolved issues +FYI I have previously made use of the class where I have needed to catch stray signals not sure if that may be one option +No test case was submitted by user +Closing issues resolved in released versions +Be careful because some changes like exclude is a windows only solution +add as affected version +I dont think this is a bug +Please attach a sample values werent handled correct in the constructor +With this patch tests still pass for me on Ubuntu +With a small amount of work we could abstract away the needs of and such that you only needed to implement one interface to add support for a new cache +What good is a completed transaction associated with a thread? Is there any chance the jta spec will in fact change to specify this behavior +It does specifically happen in the continuous case when there is a batch delivered after a cancel has taken effect +I agree its a good thing to have +snoopdave needs to learn to send from his apache address or request to change to his gmail acct. +O +I dont see what OS you are on +They actually do launch separate processes even though the NM and RM are running the same process the AM and tasks are not +Rename all sunri references to metro +Bulk closing stale resolved needs to be skipped also +Added Useful for the edges option in and probably other uses in futurePlease feel free to look and comment +bq +Sizes of trend icons are now x and x pixels +Andy do you have any update on this issue? Its a pain that we cant use this Groovy feature because Groovy Eclipse cant handle it +patch for this applied +expand allcollapse all feature is added to XSL bugs were fixed +grails directory below your home directory and then try to reproduce the problem let me know how that goes +Or perhaps a separate knob to turn it on and leave leading off +Still an issue under +is related in the sense that the demand has been for clientside spring validation +We dont get these chances often +Looks good +Yes the mgmtnode was unable to ping the server I was trying to image +test I used to for spotting this! Fixed for +Actually no the sockets wont get closed the way it is right now +Maybe I am missing the concern that Brian mentioned in his comment beforeJacques I like idea to do the reverse by using Drill for Gora is interesting too +More later +Attaching patch to fix this issue +This sample is a pure client with a combined standalone domain and node and there is no need to start a jetty or tomcat container and open an HTTP port for the domain and node to communicate +Committed +Do we still need k in the alterStatement rule considering that the ties can be broken by k lookahead in alter +Ok thank you for verifying this. +there is already an fix in place if youre working in the Fields Tab +Applied to branch and trunk +The nature of this attribute as described in the JVMS is somewhat vague so Ill have to dig deeper into the two class files +that deletion was replicated to the same database on server +Done. +Findbugs is now running against are using the default configuration which for the moment is what we want +fixed in revision +something like fill properties values and further use it when drop tag into the view but do not show the dialog every time after click or even try to drop tag on the viewI propose something like thisdouble click on Palette item show this dialog with possibility to edit values by default for the tagone click select item with possibility to drop it into VPE like WPE does and possibility to scroll and select other tagDont show this dialog again this option also suitable but it should work for ALL TAGS not just for particular selected one. +We should make sure the final are okay +I just committed this to trunk and +Hi Greg could you also provide unit tests for your fix? Thanks +It wont fix such a source attachment and users could have an invalid attachment if dont use the Source Lookup plugin no matter whether we add that dialog or not even though all have me +Im assuming that wasnt intentional +This is an improved version of the xmlbeans maven plugin +Pom now deployed. +I need to investigate a bit more +All unit tests succeed with this patch +Any additional info I could provide to help diagnose the issueAlso would be ported toThank you for the input +Amareshwari sorry for the confusion +Thanks again +Completed as part of Release Notes make explicit reference to this issue and went live on customer portal on Jan +You can configure a longer timeout using options on the configuration page of the CMS site +Pinaki Great point but I want to keep this JIRA focused on properly designing the infrastructure for post creation callbacks and moving existing dependent codeI opened as a defect for us to be more intelligent about when we can eagerly load metadata +Yes fsynced blocks may not be in complete state +Solved by + +If there are none then the installer should failConfiguring the script is a workaround IMHO +Committed the change to in trunk with revision also made a slight modification to to say can be in the top of the source treeIll wait a while to see if anyone has trouble with this or objects to backporting this after all +In such a scenario the wicket jars are installed as bundles in the container and are shared by all the deployed web appsI faced this problem recently and the only workaround I found was to embed the different webapps in a single bundle and assign a different wicket filter to each +Uploading a second version of this patch we shows that there are actuallytwo intermediate states we could encounter before we reach the end state in step +True the updated RM will not schedule more work on them +Hi Brett Could you please confirm it works under? I just my previous comment that it was broken under for me both of us were mistaken or possibly something has changedI am happy to have this closed if solved in a final! to so we can then rename to +One comment +Where can I find the patch? could not find a patch either +About unit tests you should not have to play with velocity at all to add a new test +Ill make the change to the thrift file though +I have now committed your fixes in Rev +Lets close this issue and create follow up issues +This would likely help out flush out a real world impl that matches your use case +Shall we rename it to something like notYou are right that not is much better +Ok I was able to reproduce it +Let me know if you think otherwise as per Apache pig guidelines +Thanks +It simply collects all packages with ep in name to yum remove commandFor RHEL it should be something like this rpm e +Thanks! +Also there is no password sync between Xircles and you should be able to reset your blog password via MT +It seems like the web site doesnt provide any intuitive information on how exactly it is integrated with other frameworksI also dont have much idea on what classes should I add the tag +I am feeling this patch is only a temporary workaround for some issues in threading subsystem +OK I fixed forget the hijacking +For example on the grid you see about threads at the most intensive processing moment in the reduce phase on each of the datanodes +Hi HiroakiThank you for the patch but its already fixed in version +However I have replaced the ledger identifier input with a object for both addComplete and readComplete +Path to provide join that this is a common use case although its one that I usually use a Cogroup for to get a Collection of both items +Filing a separate ticketFiled +I propose we disable this ulimit thing for hadoop on cygwin +Marking as fixed since the patch has been a unit test is though still something that should happen. + to thiswill be quite awesome to have this +Please close if satisfied +Deprecation note for enum +I dont think that would be that bad for instance I believe we already ignore testcases when they dont exist without failure and if qfileregex doesnt match anything we dont complain either +bulk close of all resolved issues in preparation for release. +Your patch looks good +The project is attachedNOTE The models folder in the attached file is empty +Or was that replaces in favour of using unique indexes at some point +Need to confirm whether Aries dependency can be removed +Apply same methods as recommended in apply into the AS Common +Its very strange that the rule of create counter is triggered here +I voted against the change nobody seemed to be interested in keeping the old version so it got changed +Actually it cant throw Youre right +Does you service name match the interface name? From your component definition it seems that there might be some mismatch there +So this issue is not seen in kvm +I agree that using the foundTask flag is a better idea +Patch applied as expectedThanks Mark +Looking at this sorry for lack of response not much activity in this component but as I answered your previous mail I felt I should look over at itI dont see an obvious homepage link +Yep the thread contention is on the critical path and the map search is on a background thread +Attaching documentation patch for this jira +new patch addressed Ivans comments +The reason is simple the code is very brittle with strings flying around in various places not even symbolic constants for them are used +Fixed in subversion repository as of rThanks for having done all the initial implementation work +or will the vote determine the golden master PNG file to be used going forward +Leave the setting and simply add the All to + to trunk +Release +Thanks Karthiknew patch attached add comment and threads check +Indeed starting with rr there is something very badly broken in the code +Here is the patch for this +Please add a unit test for this +Updating a deployed WAR or EAR with an invalid update reverts back to the originally deployed application +NET System nor Mono define such a means every library dealing with complex arithmetics has to define its own complex type +Excluding the test from running in the framework +To be honest i dont like it either it is just something that we should be able to do without needing to add loc in each service definitionThe patch i put in our server is even more big and i certainly do not see myself having to do the same on a bazillion other servicesThe ugliness in the patch shows that there is something lacking in the framework to handle those cases +charset +This doesnt seem like something that should be a built in feature of Solr +Just it will create zero size files and modifiedcreated time as current timeUpdated the appropriate message +My mailets and the util class they have been tuned for a while in a local deployment +looks good thank you GeorgeIll raise JIRA for serialization completement later when if you dont mind +Currently each produce request blocks on a response which means that a client cant send the next request until it has received a response from the current request +Thanks for the note lets put these still in +Currently I am waiting feedback from both Eric and YashNow as a secondary effect to working on the SQL Extensions and talking with some of the other Apache contributors at +These would be similar to methods for manipulating locality groups and iterator settings which are also per table configs +patch available would you pls try it thanks a lot +Also if you could deliver the fix as a patch file then thatd be great +Note will not work with if both and are installed on the box +Updating Fix For to Unknown on issues not targeted for +I do not need to ask if you have assigned the corresponding beans xml configuration file to the web flow definition +Not reallyBut we are using jdk we had problems with generating the classes +I would say this is related to but not duplicates since assumes usrucbwhoami on Solaris +OK so the folders above use the same root so your workaround wouldnt workAdding another folder to the root of the project makes the references deploy to libSo is it a JBT or an upstream issue +Equinox is available +Thanks Arun! +This implementation is incomplete. +Thanks for the review! Ive committed this +is being called by +Many of the issues has been resolved +We can move the current state of implementation to trunk work on this have been done +The issue can be reproduced if we have a partitioned table with average size being in some range so that get in the returns a position not in the joinContext +Great so are you proposing to solve? Cause that should be able to handle thisAbhi +What about additional docs and instructions for toolings where to find templates etc +What forIt uses less virtual memorybq +I am looking into this +I wish Seam PDF was documented better cause there no information in the docs regarding how this can be done +For now lets revert +Ill take care of this in a bit +Ill see what I can do to unhack it in some that pretty much forces us to this filter and push some of the features down the stack anyway +Not required if maintaining current situation with family +This is not a bug in the plugin +Is there any difference in how tried to create the dir and your manual creation? Like the user running NM and user who manually created the dir? Can you reproduce this? If we can find out exactly why NM couldnt create it automatically then we can do something about it +It is up to our downstream users to select which actual logging implementation they want to use with Jackrabbit +This problem still exist for versionAndreas thanks for you workaround it solves a problem +etc and other type of eviction settings for that particular regionAgain please explain in more detail what is it that youre unable to do right now +So this is a bug +Based on JMSs look this is an issue in the logging and isnt a real issue. +Anything further we can look at in a separate issue for +Passing back to Jeremiah until these tests are reworked +Ill take a look at the patch +okey butthere is no such code in the file +Ah! The real problem is that I was specifying an id of simply Id which conflicts with +It contains two test configurations with cyclic dependencies and one with the target set as a targetName and one with the target put to the advisors chain +Added support for cond operator in the projection pipeline statement to handle expressions in the select statement +Works well when starting with a Java IDE Juno edition +Please reopen ifyoure still having problems. +crc files to be accepting any files +It seems SVN and GIT patches are incompatible I applied changes to and just in case attached SVN patch file +Fixed in JPOX CVS +The patch has been appliedPlease verify +is in and that has no CDI dependency implementation so it shouldnt rely on itEither we CDInize or we remove the annotation +Feel free to if you later find any other related problem +Thanks for pointing that out Aaron! Im terribly sorry for the silly errorNo biggie +Works for me! Pull out the stuff fix the bug and add in the graceful degradation and Im happy to see it pushed + revision Thanks Sangjin! +Rebased and merged into the master branch. +I fixed failing assertion at +As I understand this is different in the v branch and the v branch +Hmm I noticed one difference with the artifacts I published +I attached a new one with squashed commmits just tested applying it to a freshly checked out geoserver clone and it worked for me +Ive just run your two files through my local working copy and Icant reproduce the problem +Rerunning it through Hudson so I can commit it after its blessings +However it would be nice to go back and useWhy? What are the benefits +Bulk move issues to and +Sample buildbot configuration with change sources schedulers and build factories for and +Seems we agree then that this is not a regression but indeed related to our redefined behavior for method parameter validation +Note When you upgrade from version to the URL for will changeHello Brian thats a very good point that definitely takes the edge off myconcerns I should have realised this myself too much coding +had the same issue +Comments orsuggestions about the patch are welcome +Hi BenoitThis is very strange as a would appear to indicate that you have a miscompiled package +Sateesh I remember this bug was fixed by you +Thanks JonathanOpening a separate jira to add display names for counters which dont have them shuffle and a few job counters. +If you have problems please either open a thread on the forum with some description or fill in a new jira with clear reproduction info +looks good to me could you remove the commented out imports at the top of the patch and ill commit it +Hi Willem +Does this handle the degenerate case of jobs? It sounds ridiculous but I recall previous for this situation since occasionally people have a cron job that periodically processes a given directory +This is working as of at the very least so resolving this. +Page title and Wizard title are setPage description may be hidden by validation header is now Select an alphanumerical name and a type for the application to create +Axis correctly produce the code for rpc style. +bug in the +push out to +assuming that this is a duplicate docs on shards issue since we havent seen it elsewhere. +It is a Modeler and users may not care about DB connectivity at all +Thanks Min +Can this be included in the release +Marc I recently committed +Closing on Jeremys behalf +There are many situations where a select will not initially have a value and the required attribute is needed to trigger an error if the user does not supply oneId look through the source code and try to help with a fix but I did that with another bug last November and the bug is still marked as Open +The selenium test Im news +You reference warns that It may confuse the server however this should not be a problem since the http header parsing will simply fail and the socket closed +To implement this feature we would require a worker thread that is synchronized with the appender +Could you take a look over my changes +AshutoshIve solve the problem by injecting a new into because this phenomenon is only appear in mysql backend metastoreIts not graceful but it works +Alexei could you please verify that this bug is no longer reproducible +If there is a release it will certainly include the fix but I am not sure ifwhen someone will choose to volunteer to make that releaseAs opposed to the trunk build I think it is perfectly reasonable to build a with rolled up fixes and use it in production +I fixed a couple cases that I found but without real patch I wont waste to much time combing the code for these +Here is the file I referred to in this feature request +The space keys cannot easily be changed once set +Attaching it to the document works +Anchors are now detected but less outlinks are found! Anyone has a good suggestion on where to fetch our outlinks with the anchors from +I do have some ideas for further functionality in the code but getting these operations in place will be very valuable +seems this has already been done by made this a blocker + +I experience a lot of latency with firefox since postAbout this ticket I really think we should change the http status and make something to track file deletion +im having related classloading issues while implementing drag and drop as my action beans need to import from the ajaxjsf jarimport import so probably kind of similarly the event objects should be loaded by the ejb classloader +Overall runtime was mins +resizement done +If youre seeing a hang but not a memory leak then this sounds like a different issue +Which version of hadoop are you using? Probably your Hadoop bundles another version +Runs fine for me in and command line +Thanks JeremyYour patch is in trunk rev +I think thats nice to know for users as well so I wrote some lines for the Mini Suite Guide +I have reviewed the patch and approve it +Hello Jeremias Thanks for your suggestion +Yes the above comment of Mar PM is correct +I dont think thats what we want +Add Jira Component Plugin +Could you add the patch +Thats when I had a look at what apache rampart does +I didnt dig deeply into the code but Im sure thats what was happening +You can code these variations with Velocity +Please make sure that Affects Version and Component is correctly set so the issue can be triaged +At the time using Gora it happened consistently +Well revisit this in the context of our general expression language support in Spring to comment that EL now supports the navigation operator as used in groovyshoppingcart +If we require separate builds for and its one more thing that can cause confusion for new usersAs I understand it from Dmitriy for this to work we just need to stop packing the Hadoop JAR into the pig JAR +If you have a suggestion on what should be done then we will gladly discuss this and create a GEP together +So is this the decision on this Sam? No offense but do you have the authority to make this decision +If any feedback please feel free to share +bq +IT added +It needs the apache license headers attached to all the class files though +It looks like an intermittent issue and I was not able to consistantly reproduce it +We dont set the titles of any msgbox merely is not a GUI application +Checked High Value FixBumped urgency to UrgentThis seems like something we should fix for +Since this is highly dependent on the TZ info provided with the individual Java version all I can say that I currently run JDK on Linux +No just a dumb coding error +Revision . +Can we get this on +Just moved over from mapreduce +This NPE only occour if the is set to TRUE and the related does not support the element of the related mapping is will fail also with a NPE in that case the line numbers are slightly different +I think its necessary because the method read could thrown an exception an leave the stream openAlso why is the rules object initialized to +Also will attach a test like the Hadoop QA bot attempted to apply the +Thanks +A patch to do achieve what I am asking the patch supplied to resolve this issue please be applied? +rd party to the ASF or rd party to this product +I havent tested this in a while but if you edit the logj file Cassandra should it after a few seconds and enact the changes +I tripped across one puzzling sentence but it wasnt something that you changed In the second longer paragraph of the BUILTIN bullet we recommend LDAP and schemes as alternatives to BUILTIN but we dont recommend NATIVE +Any further feedback or should we close as unreproducible? Thanks +mrepositoryorgapachemavensurefire CDOCUMEeewuLOCALSTempsurefiretmp CDOCUMEeewuLOCALSTempsurefiretmpForking command line X C DProgram Filesinseeapplicationsjdkjrebinjava classpath deewu +Remarkable difference from previous patch The error messagewas modified +Not sure what the issue might be +could you attach to this ticket an xml report where this parsing issue occurs +The previous patch for Yahoo! internal distribution seems to be incompatible with the changes made in +Please if not working. +If you finish it before we release beta please correct the version +Would you agree? Is it also a need to have translations for? the patch successfullyNo translations for this version is out of maintenance +pending JenkinsNo worries + chapters are plz make changes according to my remarks +Nice idea +applied to I think in the code should be reword to use a Compoent rather than to with rThe code could be as Juergen mentioned in the previous comment in a different ticket + +Im going to revert this due to +In this case by instantiating a new one every time it does not remember this +JacquesI just upgraded the Derby driver +is correct patch to apply +Code is now merged and commited into trunk +Applied at rPlease verify if it works as you expected +I will work next on adding regression tests and upgrade tests +do you want me to update the patchYes +So the dependencies for this will change a bitWhile I would love to have these features available I wanted to ask your opinion of the likelihood of these features being included into MCF given that as soon as you make it work with Jetty it will have to be changed when you upgrade to a newer version of Jetty +Reuploading patch with a min timeout on this test so well get logs if it reproduces on Hudson +My opinion is that it is ok to do this only during NN startup time +Forcing the type to INTEGER in the changes anything +What test are you referring to? I updated Hadooptrunk so theyd pass with this changebq +ah excellent +For now on the fly XML generation should work. +Looking forward for more contributions from you Ben +Hi Martin I submitted a modified version of that is first looking for an available port before running testsI reverted the too in order to include again that test caseCould you please run tests and confirm me that all is workingRegardsAndrea +Andrew please confirm that this resolves your issue too +Can you explain why you think the Text plugin have to be added to the? Do you mean because of the contract? Then we have to add as well the pdf et al +Verified with and No problem with +fixed that +Just upload the new patch with same filename and JIRA will automatically gray the old one out but its still visible +Works for me tooHappy to it if you still have a problem +bulk defer of unresolved bugs to +I also removed the check to make sure that the import path ends with jcrroot +We do not need a compression script for Windows because users can download a compressed build from Sourceforge +Im back up for air though so Ill take a look at the backlog later today or tomorrow. +In addition there are some refactoring and changes for ServerAgentStore version compatibility +Not very likely but the code gets more compact as side effectPatch ready for review +xml files to some degree +I would have thought it should have returned only for zero and zero for all values greater then zero +Then we can use the fileName in the exception message +Knut thanks for reviewing the last patch +removing fix flag +Hi Johan Resolved is good enough +are only a week or two away. +Please dont use Jira for such questions repport to the user listJira is a tracker for confirmed issues that where allready repoorted on user ll add an it test for this case in the plugin and try to find a global fix +Whats granularity? If this checksum based on each event does it affect the whole efficency in a high frequency environment +Just as a note the workaround was integrated so the problem shouldnt be present anymore post Beta +Updated doc svn rev diffs show content is present and simply needs to be pushed to the server during the next release +Mark Little saisOASIS is dead +Sorry +Thanks Kihwal +Assign to Damien per discussion +Attach again +Going to try with your new code +So do not need to contain attribution +bq +Apply the attached patch inside ROOTsamplesclient applied in revision +Id be happy to add this in but Im unsure if the old documentation is still valid could someone verify thisIt states that load balancer rules need to be created for ports and +Also need logging for activation spec and admin object construction +The uploaded patch contains the patch of since it is required for the design +As mentioned before I wasnt sure if this is a good idea to remove the line +I fixed the include warnings you get with Ant +For the first issue I created which is resolved in rev +going through old jira issues assigned to me is this something we still care about? or can it be closed as wont fix +I filed this issue and I agree that its an enhancement but I dont think it should just be closed outRelocation patterns are very difficult to debug properly a little more sanity checking would make Shade a lot more useable +Hey I was curios and Ive tried myself you are right the targetPath needs to stay thereIll apply your patch as is thanks. +Ok i will do it. +Added the patch for my test +Fixed please verify +It is caused by incorrectly sizing the datanode JVM heap unless one shows a leak in +Ive been running the client application via the ant run command +Ive attached the new which should replace the old in the lib directory +Tug told me hes doing it +At the moment I can not find the mail thread about this discussion +They snuck in via inheritance and I thought it a nice to have +HiI think adding a keyword is ok as long as documentation is updated to make it clear what is the difference between the plain OTHERWISE and OTHERWISE NULLABLE +Above Jenkins build failure didnt provide any hints so just suspect something wrong with Jenkins job +The DI for this diagram is now generated correctly +I committed this to and trunk +It could be useful to filter out certain projects such as if was written in and youre just going to rewrite and toss that existing code later on +Do you want to upload the logs for both scenarios? at present we upload logs only when application finishes +Its been applied to branch +Its likely attributable to something in the usage +If there are newspecific issues with JSON please open a new JIRA record. +Can we mark s resolved +This is a plan worth pursuing +Thanks for the patch Kim +There fixed my. +No repro +May be some gaps are there which we have not yet foundI should accept one thing is I did not try to change the in RIT and was seeing to use the only +arent injectableDavid can you clarify what you want +Sounds Good +So then heres a new theory The test produces flushes +just committed this. +ARAT should be flagging these +Patch applied at +here is an updated patch I think we are close here so a review would be helpfulAnyone Yonik Hoss +Please test against working correctly against. +First basic version of the gallio widget has been commited +I see you removed existing test in old test test +generate true should do the trick +I think that this is actually bug selectdeselect werent being written back properly +Could you please give it a try and provide the messages being printedThanks +local disk +This is already covered by issue +So I will commit this patch to both and trunk +Is the DTP jira known so that it can be tracked and when this jira can proceed? +I have added an overloaded method to to allow passing in a +Compat is important but not so match as unimplemented features and crash bugs +New patch addressing Elis feedback except forbq +Thanks you are right I dont see any filters coming in the way of test url after it is authenticated from WAS can my application know about the authentication details in this case as forwarded urls are not getting intercepted by none of the security filtersHow should I configure in this case +My only comment is you may want to delete out my debugging calls in the test +SiCould you provide a link and a more detailed problem description please +When Java timeout expires connect method throws but when OS timeout expires we getHere is output from my test connect timed outTimeout specified sec time elapsed sec Connection timed out connectTimeout specified sec time elapsed sec +In the meantime Ill keep a lookout for initialization issues. +Thanks Sergey +to Jons comment +Added mojo runOrder setting and documentation in r +of the time I need to see the dep tree in order to resolve version conflicts +This issue exists in version as installing the latest flash version +Hey RaviThanks for this improvementPersonally I think Please configure another method +I will contact Weblogic to see if they can provide me any will keep you a lot for your help for the time being assuming that this needs to be addressed within WLS through a fix or through corresponding WLS be reopened if theres anything we can do within Spring in line with how BEA recommend to solve +If Mauricio adds this capability let us kjow +Once question do you see a problem with calling get and doing the filtering for every socket creation? Thanks wasnt happy with reading the keystoretrusstore files and doing the cipher suites filtering for each socket creation so I ended up creating two patches Does the filtering the down side that the filtering is done for each socket thats created +how does moving the removal out of the for loop fix the problem +Im going to look into a patch for the Dialect and get it posted +I still think that we could have the conditional actions as an addition to the API +Could you please give these a try and see if this is the only source of a leak or whether there are also other issuesThese should be placed within with C CR +I have looked through the patch but I dont understand what you mean by this approach can be harmful when portlet URL returns path with the context prefix +all issues assigned to +If you dont use the Spring +Issues resolved in +User should not build controlsWeb before runningNot sure the reason for having user build coreWeb in package +The existing jar minus the code generator does worries +Since this issue was created the internals have changed so that is no longer used +Please let me know if my description is +The same issue applied for shared flow declarative validation messages that used message keys +Should I file a separate issue for that? permission write This property is used by Derby to prevent the accidental boot of the database by two class loaders +QE doc reviews are completed. +because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +gotta get the parser up to snuff specificly what we need to do is to improve the API to allow retrieval of generic parameters similar to the new reflective API methods in +I am going to rerty this following Brians steps +Lets do this in as well +Description clarifies this would typically be a preliminary step in job rather than standalone job +What is the distinction between Tested and External? Tested are hosted by us? We should indicate that should not say Tested by Red Hat +The subsequent request header from the browser can get pretty big and overflow jettys buffer +srcjavaorgapachecommonsconfigurationimport +It seems from a debugging session that the two last join queries Army posted both fail due to failure to classify some of the as noopsIf a PRN is classified a noop the RCL nodes are maked as redundant and during code generation we walk further down the to lay out the code at a lower level +The basic idea behind is to perform single sstable compaction if its droppable tombstone ratio is above thresholdThis works because single sstable compaction drops tombstones and lowers droppable tombstone ratio which prevents recursive compaction on that sstableBut there is a situation which compaction does not drop tombstones +Thanks +Thanks Sandakith +Closing this +svn revision +need to handle the possible test names problem is when the test src file is called tests and not eunit the module name in the xml wont contain the tests in its name +Found a solution to this problem +That is the pointAlejandro how about starting this in github if you are so worried about incubator? Frankly I dont see incubator as a big deal though +It works fineIt is ready for review +Unfortunately there doesnt seem to be a better solution +I wish I could track down some logos for and +Lets try to get this feature in for. +batch transition resolvedfixed bugs to closedfixed +Patch to create better error output in case sourceRef andor targetRef are not valid within +Ok +Fixed but there are still some questions assigning to Lev to reviewQuestion is this the right fixIf ex is null ! are printed to while the rest of the messages are directed to +I think that may be part of the solution but in looking at the other libraries I think we need to just enhance the protocol classes to support writebinaryreadbinary methods and implement usage of those methods +Thanks for the solutionTo be tested and commited by a commiter of +It is not a common requirement and adding a bean manually only requires a single line of configuration +Comments from Denis in IRC suggest that old TPs inclusion of TPTP and were in error and DTP being including is a good thing +we are using hadoopThis simulator is developed majorly for stressing JT and NN +This can potentially lead to deadlock on systems with small network buffers such as seen hereBut by default this flag is not set +Just so Im not being confusing that checkout process IS NOT in thats a custom one that I did back in the day for a customer +However copy is not in scope is it +Basically most of the methods will have to take as an argument +Hmm +Included change missed from previous merge to set local +Forgot to mention yes I did a clean +Original patch as unified patch +This patch doesnt modify other RPC callers +bq +I initially tried to do upgrade to spec when switching over to bundler and reverted it due to the build servers not being able to run that gem +with static variable or a cloneYou mean create a clone method +How about this little patch to avoid creation of if not needed +Fixed with respect to the current semantics of interface properties in r +Patch applied in rev +perhaps a shared memory design or loadable module design +I read that but thought what I was doing should have been okThanks for being on top of all thisJohn +When is searched in query everything and a match was found in a package name the search results should know that it was a package that was found +bq +here is a patch as a start +Hi AndrusIm sorry for the extremely late answer +closed with the release of +You are compiling the file into the wfs resources + +Both PEAR files are modified to ensure the PEAR Merger test does not fail in the Apache UIMA environment +I tried to display the complete path of the source column but it was too uglyIf anyone has great ideas for the UI let me know +Could possible just make the command fail if it would create tablet with over a certain number of files +Any more info on this? I dont see it in the repo as of yet +Changed the proto field also +I have tested it with the attached to +Note that despite the same input data the results are different depending on how the data is added to the message +Patch to decode get parameters as they are parsed +For more information see the comments there +Thanks to Arun and Giri for helping me with this +Can you attach a test case that allows us to replay the problem at hand +with true and false +Works correctly on Windows Vista +If it could fit in with Wookie development roadmap I will be happy to know that my work is useful +Its not a bug its a featureSuperclass field access NYI means not yet implemented so its a documentedJoking aside I will try to implement the missing feature soon +The URL given here now s should we close this issue assuming that the reporter never followed up for a wiki account +Ill post it later +or rather the +lgtm +are missing the bean class +Well thats just frustrating +on solr +What version of Maven are you using +Slightly more conservative patch to reset temp in catch blocks +This may not be an issue for you just double check your baseline code to see if you notice Windows EOL characters +Attachment has been added with description Patch to revision of branch Temp +I think it should fail to start +About Timer failureI think this stack trace doesnt help in any respect +Its a while since this got updated +I just committed! Thanks namit! +Thanks for pointing out Mahadev. +By default the auto progress is false which means that the timer would also be offIt is a change from the existing behavior is that acceptable +Thank You for answerYes with force work post installation and formatting to ext but image with ext not workIm provide log from boxgrinder and from amazon system log +You probably have exported PIGPARTITIONER but is using the default partitioner +Investigate +We have discovered a workaround too but I have been too lazy to we did is we put a classpath in front of the path to +Paul sorry I really meant to include that in the original bug +No plans to apply the TCK changes across to. +adds a public getter for a new field called creation as well. +It will require some work but I think it is worth +Patch attached +Port Configuration in eclipse plugin and settings wont synchronize with each have to keep it consistent manually +for the to trunk +In this situation we have control over the session lifetime ourselves and may or may not use a transaction inside that lifetime +skip that I misread the issue +This site is for Juno DEPENDENCIES not Tools built on Juno +Do you have any error in the logsHave you tried to launch a resolve before viewing the report? relies on some report in Ivys cache maybe it got wiped out +Abhijit the test still seems to be failing +I recreated the patch against todays code baseThere were some conflicts +The bundle URL field apparently has a comma attached to the +I think the previous comment I made about API changes should be seriously considered +can we also get away from checking in generated doc +thanks +I am seeing a case where if I configure different targets all of the Puts are getting routed to the first +I looked at what it would take to move from to SNAPSHOT and there were quite a few changes +Corrected the package name +This has been taken care by Sebastians fix for +Our application works fine on and but we would very much like to have it working on Geronimo +Thanks Brock! +what are the features of your project? any generated sources? +If its too much of a problem to smkdir mkdir p Just replacing mkdir with mkdir p doesnt fix things because Hadoop doesnt support the p option so if you just add p everywhere your programs break on old releases +I should be able to provide the code as well as the refactored that leverages this code by the end of the week +One fix is to modify the to ensure that Enum classes are also loaded by the normal classloaderA similar fix is there in +Thank you Jing Zhao +and probably wont scale up very nice for many layers +Right I used a duplicated dot to do the escaping +I just uploaded a fix to the main Android repository +will post you with my findings on by hrs actual value of prefetchSize will depend on the size of your messages +Worse that default value is also set as a range constraint in the db so if I have a bigger than bit numerical DB type I cant use Range +At some point we must have decided somewhere that including the test case with the bugs in place was a bad I check it in with the bugs enabled +Ive tried both trunk and revision +But I thought the stream wrapped by the QP stream in my fix is actually reading from the wrapping the root stream? Shouldnt that work +It got added mistakenly +If not objections I will checkin this patch +Updating the Seam example to the newest version resolved the problem +The issue is fixed in rev +I added detailled reporting with the application being deleted and the of the total work thats done so farI am now also checking job cancelling +Addtional patch that I hope really solves the problem +Not sure if you want the same behaviour for am just pointing out things which look a little bit odd to me so please feel free to ignore my suggestions +Changed all places where this is invokedI rand which invokes this piece of code to make sure is loaded and recover the is patch with changes incorporating comments from Stack +Patch in trunk should fix this issue +I think that we should avoid adding a new directory in src +Weve removed the employeeWS sample from the v distributionMoving this to TBD +I have committed it +trunk Committed revision branchx Committed revision +Works and committed to trunk +Lets rename one of them to something else +Greg In general interpreted methods should run a bit faster than MRI and if they get a chance to compile down to bytecode they should always run faster +Result as expected due to the smoothing parameter +Doh! Just remembered that in your comments you envisage another few patches being applied for this issue +Havea look at the competitors like Axis +No integration test +Ooops not a major issue +I think that is a reasonable behavior +I havent had time to fully digest it yet +When the data source is created with no credentials user name is assumed as anonyous with password as null +The server timezone is not relevant to what is shown in the UI +Ashutosh I left some comments on phabricator +But why is that necesary? You have all the hooks from +compiler writers often optimize once its in the standard +Here is a simpler test which fires the same assertion +filling up the block cache I think this is the biggest cost we are paying for online serve +Also dont forget about which is also the same bug. + since the site was updated this issue is fixed +We actually need the in the values too +Patch merged with latest trunk +has a unit test reproducing the same issue for +I cant guarantee for certain this is fixed but I put a null check at that line and I would suspect that specific error wont be itself +Axis maven build picking up the new Woden jar files OK +now it makes much sense to me +Trunk patch is to follow +I searched but did not find how to set Restart on Fiscal Year as invoice sequence method +Tools does not validate Java Script +only upgrade in trunk and add a warning in pointing to this Bugzilla issue +Simple patch implementing above approach +I just forgot to resolve it +Kelvin is investigating but it appears that reads and writes at ALL are not blocking appropriately on the server side hell open a separate ticket for this tomorrow +This is a really interesting problem could be a nice project for Googles Summer of Codedominekif youre not moving gigabytes of data its unlikely that you need to worry about one or two buffer copiesif an application does move gigabytes of data and you have some other use for the CPU while youre moving data you definitely dont want that data flowing through the file cache because you are incurring a buffer copy and the data volumes are too large to get any benefit from the file cache +Committed javadoc fixes +As I said your error is fixed +Im under the impression that the MAXSUBST is intended to prevent recursive substitutions of a single config variable and didnt consider the case when multiple replacements might all happen in a single value of a config var. +applies after in this caseJust to show the correction better +Here is patch which addresses most of Phus concerns +Tested with latest build fromAll the test cases below passedScenario NFS as secondary storeZone XenserverZone KVMRegister Template with All Zone Register xen template xenRegister Template with All Zone Register kvm template kvmDeploy Vm in Zone using kvm Template list during Vm deployment should not show kvmDeploy Vm in Zone using xen Template list during Vm deployment should not show show xenScenario NFS as secondary storeIn Zone add kvm clusterDeploy Vm in Zone Template list during Vm deployment should should show xen and kvm +Thanks Ted for your test caseThus we need to handle in the class +Although its a great workaround I think we still should fix what ant eclipse puts in for the project name +Verified. +Hi The cluster node went down due to journal overflow +The Tomcat option worked greatThanks. +We now do Java reflection which looks nicer +A more meta and general question is whether or not we want to record not likely to happen enhancements in JIRA +Dropping from content as it does not seem likely that this can be solved in timescale +to the patch +Thanks LilyI have now also fixed the css problem for the deployed docsClosing issue. +It is probably not worth the complexity of trying to determine which is which and as was already mentioned in this thread this line of reasoning arrives at the need for something like a annotation which for reasons aesthetic and otherwise is not desirable. +I think all information are in Steps to Reproduce the issue here is that in FRESH workspace and with NO Archives view opened there is no chance to create nested archive via archives explorer +fixed in rev +Good find it should never return +this an old feature documented somewhere or is this new and needs to be documented +For we can though +I committed to trunk with revision and built the to confirm the changeThanks for the patch Jazarine +HiIn fact I ran mvn clean deploy and then mvn sonarsonar and it works +Latest patch +Thanks for the patch DD +This new patch should account for it +Yes Brett thats exactly the problem +Is this fixing a specific issue Donald or are you just looking at getting current +Hey Todd thanks for the update +If I use delete the CRC file should be gone +Committed wo review though a fairly substantial patch lots touched but its mostly just fixing failing tests and remove of SC and SCM. +Attached patch is a work in progressTests pass except for some order by related tests in the JCR TCK +In my opinion there is no problem with that +Ive committed this with some minor changes +Commited test case of this issue +What is really needed is a server side interceptor that can be used to build a better failure response +calculate Not sure if calculate is the right word here +Otherwise we will break the plugin for version +Is there any way to avoid messing up cache hit ratio with these optimizations? Maybe pass a flag to getBlock saying we should not update the metricsAnother suggestion from Kannan maybe we should move metric updates and other extra logic outside of the lock and only leave cache lookup and HDFS IO inside the lock +Since these are not timevarying int or timevaying rate +Added call choreography stencil +This means however that the constant that names the subdirectory name should be owned by and not by +Thnaks MridulYour patch is in trunk r and r +I got this errornls +I tried this with the latest built from SVN and +Before doing that though I would recommend running the following tests standaloneIn addition I would recommend running the upgrade tests because they test the metadata in old versions of Derby and you will want to make sure that they dont break when you change the metadata inThanksI am attaching patch which is intended to implement the spec of this issueAdditionally this patch is addressing as well both are so difficult to separateI ran the full regression test suite successfullyThanks +Fixed title +Thanks Ferdy for the commentary +I modified the callback support for streaming so that the code should wait for all streams to finish before confirming +Hmm my second patch didnt the id from the vector is removed I believe it will affect all clustering algorithms +the suggested change has potential to break existing applications that rely on the current behavior +Deferred to unless we have a patch resolved issues for already released versions +Thanks for the patch +Heres a patch that doesnt require loop status and passes the same testsComments +All the tables in this section has number only +Fixed for now any further changes will be improvements so resolving this issue. +To get an answer for that it will be the best to ask Mark Struberg +I also think that we need to manage documents by release version +You can support multiple backends for dynamic classloading jar groovy closure etc +This is currently a contrib in PIG +Do we have this yet I remember reading something about adding it in the past +Im trying to avoid what happened on the hadoop ML over the weekend re got over most of the above API changesFor Filters I extend them from +were not developing Deltacloud any longer +Attached patch display message No installations available instead displaying the list prevent NPE +ok tweaks are committed +Did you check the receipt for configuration information? I just checked the patch and its in there +I dont have Fedora +Is this desirable behavior? It would be easy to abort on first error if thats the consensus and I could take some tedious out +I will run some more integration tests on it tomorrow for now it looks promising +Fill the second column with the table that you want +bq + unscheduled issues to the next release +but it will never handle when we have where clause in the mapping fileIf is mapped using subselect it works properly without any when its mapped to the table and having the where clause it gives the Foriegn constraint voilation while second It doesnt consider the Where clause +We were just debating that +Attach patch that moves from ipc package to unit test locally and it passes +Since the project examples feature is different for JBT and JBDS we have to set Central differently +Please include into release +The reason attribute can now be a code resolvable through the s. +If the reason the issue isnt being addressed is the need for a patch let me know and Ill finish adding the feature and generate one +Maven Creator Replaced the hardcode file and applied to r Maven now builds in Windows XP Maven Java +But for Xmx and Xms there can only be one such switch so having these be built from a directory seems unnecessary +I propose changing toThis is the installer for EAPSimply run ant from the dir you got afterextracting the in your EAP JBOSSHOMEIf this is acceptable well close this need a confirmation that the text is acceptable by early next week or we will have to return this to TBD EAP +I am sorry for asking to remove the changes in the mean while I will reduce the priority +As it comes down to distribution of counts vs +Closed after release. +Its upgraded to today +Theres some javascript running that page that was thrown together for the demo and needs to be replacedfixed +The model jar +Ill get this patch applied soon +The last comment about not having a manger seems not globally true I see one in jboss and the manager shows that the war is distributable +So this scenario should not happen +Note that the attributes called a and b are not displayed +Fixed in CVS now. +JSVCHOME is configured but JSVCHOMEjsvc is not executableNo tests are included since security and root access must be available to test this +Agree +We could still come up with a way to use the test suite in our Hudson CI build and individually by each developer but we probably cant keep the documents in svn and we definitely cant release them as a part of +Thanks for the thorough review Daryn! Updated the patch to address all but one of the concerns +When would already ever return true? If it never does i suggest we remove this check +undeploy helloworld +Thanks Marco. +However there is a long logicalclock in as well +Using the payload proxy would only be effective if we had a mechanism to specify the legacy locations for the pipeline and as there is no easy way to determine these values from the pipeline this would require further ll take a quick look into this +Jason can reproduce but Ales cant +we didnt need it Cheers at r thanks Sascha +Enums can now be defined as attributes with the standard API +Permaineexcuse me but you didnt catch my point +Min is working on eventusage stuff +But if you feel strongly the other way for seamless backward compatibility I understand +I think this needs to be finalized based on your agreement +For testing on actual clusters Ive done this by shutting down the active NN then just rsyncing the storage dir to the standby then starting the standbyYour idea of skipping is one solution for +Perhaps the fix has introduced this regressionColm +Increases of MB from MB to MB is the way I was quantifying was percentage of entire java heap memory +bq Would it be ok to have this on the branch as well? Done +defaultparallel setting in Pig will map to in hadoop +May be I can help in that case +I was wrong about reading the split file multiple times +The test was looking for interrupted status after the second callPatch v should make it passtest passes locally +Increasing priority level +This was fixed in the later version of per last comment the reported issue seems to not exist any longerClosing this now. +Problem is not fixed in! Line +Is that because happens elsewhere in parent? so doesnt show in a particular regionThere is look at a region that was created with a table +The SYSCSBULKINSERT procedure is in fact even worse since it doesnt quote the identifiers at all +Closed upon release of version +Patch for G which is a slightly modified version of the and does not remove the changes made in G due to the needing the ORB class from the Yoko spec jarI verified the minimal Tomcat assembly starts and the deploy and workI verified the JEE Tomcat assembly passes the Samples and EJB buckets with the same number of tests passing as before the patch was applied +resolving this issue since it is fixed in Axis if you see the problem in WSAS please report that issue in the corresponding project jira +Larsh committed to. +We just need to handle the special case where all apps in a queue have weight and make it so that this does not result in an infinite loop in the computeShares method +It seems that that there needs to be another way to handle properly getting the correct servlet context in a portletservlet independent way +Yes please for +I wouldnt veto the patch for it but it is certainly not fond of itOnce Abacus is moved into the streaming can stop including abacus related stuff + +bq +Have a single component provide both local and remotable service +You trade storage vs IO +Unit test demonstrating this error +If the race doesnt exist it would be possible to simply lock using the inprogress znode +Voting should be locked on a comment basis nowPlease verify and close the ticket if youre happy with it +Fixed performance problem by removing usage of for normalprocessing flow +mailing to user list instead +RTC defect all remaining issues so I can release in JIRA +An entry in retry cache will be retained for a configurable period of time +The local baclup tree is searched first for data before any is sent to other nodes so the stale data will be used +It shuts the connection down slightly too aggressively causing an exception on the remote side before setVersion gets calledI can see that the initiating side could get pissed that the target closes the socket uncleanly what I dont get is how a sleep could make a difference +EHWA works for me on both Linuxes ia and +Jyotsna Rathore thank you for reporting this error please remember forms and ftl are not the sameHowever you problem has been fixed +Can you post a snippet on how you use the template engine? +And require allows you to plug in whatever description you need to make the error message meaningful +Its not the problem that the ids are generated after the is calledWe had the same problem with duplicate ids +To preserve backward compatibility if BV is not in the classpath we keep the not null checking as it is todayIf BV is present in the classpath we disable Hibernate Cores checking +Heiko +I checked this into trunk and. +t +I created for the broader issue +It appears the lease handling needs a bit of improvement +Attached +Ryan Regards key size was small in this case and not tweakable +Id really like to see this patch incorporated sooner rather than later +I think one test with CXF as client and one with CXF as server should help +Rather an instance of is returned +Ditto on javadoc fixes +Done +Note my last comment was meant for Stefano not for the issue reporter. +I thinkthe US Post office is pretty accurate. +Looks okay builds work tests succeed +Based mostly on +Using more efficient Monitors and such to synchronize access will make the threading infrastructure much more scalable +Mass closing all issues that have been in resolved state for months or more without any feedback or update +PMD might require something else +I tested it at least using in example directory on Mac osx and Rhel +Great idea +It would be risky if the thread was killed but I dont think the JVM kills any thread it just waits until its done then shuts down the VM +Thanks! +Would you be willing to try again +Thank you Frank very helpful will take these under advisement when I can schedule this in +Im not sure its as we upgraded this morning +patch added +Idea for rsync probe a dummy rsync module which has logging turned off. +The registration is based on reflection and does not have to match the property name of the bean +so i would devideClient Interface stuffService Provider Interface +These ideas went into the work. +Numeric case insensitive and case sensitive are required +timed out error occurred but tests are passedi just committed this. +Thanks Pavolthe screenshots have been updated and the text around them updated to mention the greying out of the versioning checkbox if the server doesnt support the feature +Martin I have created related issues for you please vote on it if you would like it to be implemented in upcoming versions as they are documented here will be renamed in scope of. +Using the patch I submited before and this new template you can have a webstart for the various architectures and with support for various architectures and d like to do something about this issue Could you attach here a example project about this featureIMHO Doing some logic in the template is not a very good idea it would be more cool to have a real special configuration for the nativelib support DYT +good to me applied to and HEADMike +But it looks like in this very case this is I do not have a concrete idea how to go about this right now at least I think this is what the problem is +Execute the vdb and this queryselect SNAME sumweight from +well spotted +Reopen for to myself temporarily for port to +See for further details. +Hi Roman Do you still think something needs to be addressed here? or we can close this bug? +I think is probably fine +Ill make it +Ill create a patch at the beggining of the next week +Please close this issue its not an issue in in forms is currently not really supported by see commented by Geoffrey this is not a bug in the masterdetail implementation. +If energy is going to poured in here the other thing we should consider is a database mode so that users can PUT and POST but not GET +You would still have to check if the record is the first one +Any updates on this +Remove the link to or make the href an absolute ref and the problem goes awayProblem in +I was very happywith your patch and was only logging my changes to your original patch +should reimplement the default behavior in the new methods? Anybody who inherits would get it +Reopen if not +I am seeing this with IE as well with Solr nightly build as of August . +never heard back from Paul so resolving as incomplete +is the new patch for trunk +and for fixed in Rev +We also like to close this one but do not understand it can yo elaborateRegards there is no more mypage and lot of work has been done just close for nowwill open a new one if I find the same with my portal. +Please reopen if the issue is not actually resolved or file new bugs if there are additional problems. +Hi StackActually there is another scenario another process test took the port between the time we killed and the time we try to take the port +Filed for a v backport +See andIs there something beyond this that you are requesting +Wrong commit message on that the change is in source code of svn ER tagIt was difficult for me profile the deployment speed so I checked it just in codeWhether its not sufficient please return the status back +bulk close of all resolved issues. +base decoding of attachmentsoh gosh back to +We had to revert the remoting upgrade in upstream so this will likely not be fixed for AS +A few comments It is a bit hard to see that its the Opera house +With s acquisition of the nature of AMQP vs JMS the rapid growth of s user base in my opinion this is as least as important as JMS and probably more for enhancing adoption of Roo +CR is posted and future enhancements or defects should be logged as a separate Jira. +Otherwise please direct all comments to the existing! code formatting tags +Mind you it is not the archetypes that pose the problem +Ive run the test in a directory where it was previously failing times and so far its been passing consistently +We are embedding in the bundle so we can get to these classes +Could you try again the trunk version and tell if it fails with a clean message +If the user wants to use something outside of that list they have to quote it +Perhaps its locked againMy colleague and I have added links to the file days ago but we still dont appear on the public siteI checked out the infrastructuresitetrunkpeople locally and ran the ant command and was able to successfully generate the site. +a lot for the several reviews Todd +Parameters setting is working for both primitive and reference types. +I am wrong +The same issues applies to the Stage plugin seeIt might be possible to add some class path entryies to the plugin on execution to work around this? + Added cookbook and tutorials into the book page +It is reproducible every single time + take a look at looks really interesting but Im agree with AndreasWhy having a bndtools integration in EIK? It could be cool +Should be fixed now in CVS +Also checked the studiofeature plugin +And this both in and +Im going to commit +Committed patch a to trunk with revision +Ah +I have to retract my claim that there is a memory leak with Solaris +I can take that +or eap windows m usingRegards is caused by the fact that uses and our does not anticipate thatI fixed it locally but cannot commit it because of SVN authorization issues +What do you meanContent assist validationIt should work +Replace with Semaphore to prevent and allow during startstop to prevent deadlock +here is the patch make cluster happy and it passed my testing +Applied +add more volatile to variable that can be changed through JMX Add and and store the message in the same +Patch commited in rev +if you just want to create one big patch and we can mark your other JIRAs as duplicates feel free to do that to make it easier +Possible improvements could be to either add a println or turn the labels into comments +This patch does authentication only +Well I dont see Wicket anytime soon +The patch will become quite big but I think the resulting code will be pretty clean +Users will not be forced to change their implementation if we add a function to +I had the same confusion when looking for the appropriate API +Maybe we should just talk about upgrade +Have you tried to install it with normal java and then run it with Java +It would seem that you always need to check all files in case of moreover when we look at individual scanners we are not processing deletes correctly +I will try to find some time at the weekend to do that +Thanks Nicholas youre right those two tests are failing for different reasons +abstractimplement or the current childextend with id it is the functionality is which most important +One thing we might want to consider is what copyright attribution policy we are planning to have for contributors +So this one should be better +Taking this off the list of things to fix in +I guess this cant completely be considered solely a packaging change since the stress script itself changed but I guess Ill leave that up to the committers +All issues has to be filed against from now +So if were not careful this issue will work against since fixes are in released code +Making a target of as it would be nice to showcase this new integration +Updated trunk patch +I do not even see a CXF build scheduled. +Works in head revision +It should be committed after C analysysC overloading of operators should NOT have a return +It would be good to mention the bug number in the comments where you talk about the bug so future readers can refer to this conversation +im just debugging it to find the solution +Yeah jon I was about to suggest this +Run without security manager the test passes with jars +For example removing it from aggregator annotation implies that the plugin is only able to build the aggregator pom in a build +Duplicating the data there may not be a good idea as there is a lot of data to be maintained +The reason why image files arent actually deleted is to prevent problems if a user accidentally deletes an image +Any chance you could provide the URL to that post here +I think we should exclude this from as were getting very close to the code freezeWith the current approach tests are failing and I agree with Koji that new functionality like this can and should be added as a new +If this constructor is called will start Jetty as well as Grizzly +Perhaps the offset is usually Also the above fix didnt actually fix the in my tests +Fixed trailing whitespaces and failure +We simply cannot put everything in the response +We already made some code changes to reuse the connection for sending data Created class which has the JNDI specific configuration Created class which holds all data associated to an open connection to so that all resources can be closed appropriately Adapted the to check if the connection was not open and to reopen a connection on failing to connect to +I think this is what the lads over in originally suggested and I thought it wrong I still do +summary of JIRA to reflect what the patch does +Done +Fixed dependency versions of said bundles in Rev +I think the problem has been resolved but I am leaving the issue open for now as the solution has some overlap with at least one other open issue at the moment and I want to make sure they are all working before any of them are closed +go for it +Carl when you supply patches please make sure they can be applied with patch p +actually this can be to big somebody up as +Is the merge prevented only if flatten is present or for any udf +i think both soultions are correct but personally i prefer the patch i submitted earlier since it would be the most simple and consistent solution with the existing implementation +What went in differs from the latest patch here only in that some javadoc has been further modifiedcleaned up to attempt to make the intents of the new classes clearer +Are you doing anything special with FOP? Maybe you should show us your class +If code is creating a background Thread for its own purposes then it needs to ensure that it does not hold on to an application class loader +Patch for attaching the patch +Im still having this problem with +In the next bundled release we can look at better ways to do this. +Here is one suggested patch +Sorry I forgot to mention that the service wrapper i used is in the ll attach the binaries tomorrow when hudson builds are done +CharlesI think the fix for Case only works in CamelPlease feel free to try it outI think you just need to update camel version to in the pom maybe you need do some modification in you pom if you are using some other version of CXFBTW I think CAMEL should work with the fresh release CXF +But since I cant find a specific record of why we dennied you site I have allowed you accessNote however that your mirror site is not following our recommended httpd configuration +Good to hear +Ive started converting some tests to select table names from the test method name instead +files +Attached reduced resolution tif +I cant think of a case where the s responsibility to transmit data extends past the closing of a session so a warning seems most appropriate +so they changed the interface and in order to upgrade we have to handle the exception +Looks like both passed +stackOf course It work but i think this formal command should be providedGenerally A query file used for construct schema table and column references in the auto installation batch program +Validated that lowercaseOperators default is true in and trunk +LGTM +Alan would love your final eyes on this +Probably this is obsolete +Thanks +Id really like to upgrade to a newer version + +Fixed +I also commented out a few other isolated lines that were stopping the conversion +I fixed some issues around concurrency and also some stylistic issues around naming +For another day +Ive committed this and merged to. +Triaged July assigned normal urgency. +Unfortunately I do not recall exactly which version I was on at that point but I believe it was some iteration of +Hi RichardDo you have something specific inI just tried your without it and didnt observe the problem you reportedRegards Richardsorry for my delay in replying +Please commit +Oops +One of the major trusts of virtualization is to fully encapsulate the virtual environment +It incorporates the from and the from +It adds Ant targets that run the junit All suite and create a reportI could probably eliminate the need for the test policy change by dropping the showoutput tag in which is fine since theres no output it setup this way anyway +Thus I need to assign the above constant directly +Reopen Andrew if you see hole in my reasoning. +And forgot one step +Patch for apply the same as trunk on the branch +An means you have misconfigured either your app or Crowd so you want the application to ll see about catching and throwing a Fatal exception +r +Updated to fix the very valid findbug warnings +In that case CLI etc shouldnt even have any changes +Applying this patch to will provide to +Do you have chunked header? That should work if its thereCould you share sample with headersRegardsArnaud +Also the patch is a bit premature I dont have where to apply it are you willing to keep it in sync with wicket trunk so it can be applied once is branched +JDOM seems less affected by the problem. +All sample jobs can be run from command line or JMX since configuration was split into and + to trunk. +Patch upgrading UAX tokenizers to Unicode +I will correct that sometime todayDont laugh at my crappy scripts youll hurt my feelings +Implementation details we need to know list of repositories where fetch artifacts from +I have changed the logic to throw a when this is encountered +We done lots of work in this area and this should not be an issue anymore! +If you make the nd select and get you have actually just reserved not +It also includes the name of the database or document that youre viewingIve made a new svn diff patch that has your new changes included +Yes I noticed +assigning it to to investigate +OK Igor that make sense to me +For a particular time the collector is running or not is achieved with the help of ps ef grep collector wc l if its value is less than then no collecor is running +So really all we need to do is create the maven assembly +Well is running on my LAN so do you think these reverse lookups will yield any fruitful resultsI tried making GET request to with Apache HTTP Client and received response successfullyI am wondering is there any code in the Jena library which does something like thisGiven serviceURI httpsomehostnamesparql extract the first component between http and first and treat it as hostname? Such thing can lead to exception like thisCaused byBecause somehostname can contain port number +Seems I have to be more careful while expressing myself as there seems to be a bit of politics while choosing the wordingWith My personal vote I did mean My personal opinion but according to the link you did provide it does really mean votingThough still a question I did also put some funny comments into just for the sake of a relaxed atmosphere as well as good mood +I am sorry for wrong spellingthe orginal words should be why dont +Will open new issue for. +Thanks Ted for coming along w the cleanup +great my first patch worked fine I dont know why I changed it +See changeset ecdbccdcdcaaffbeede. +Will be more careful next time +I moved this issue up to release since itll be implemented as part of adding +Just checked in a change to move to +Try that and see if that fixes your issue +Can someone review it and help with findbugs issueThanks +Moving to as is tried this but it does not seem to work with auxilary jars +Closing this JIRA +Yep we tested it as well looks good +Perhaps numFetchers should be renamed to numSegments +At least put some tests for the lookup of equinox launcher jar the patch and add neccessary changes keep it backward compatible with eclipse should work with eclipse by copy plugins into +I saw the messages twice and assumed that was the same issue reported here +add was adding the node to the blocks map even if the block is an replica of the last block in an file +Fixed on trunk +But I think we can at the very least incorporate this into a feature requirement for MSC +Todd Instead could this be an attribute of the scan so we dont have to duplicate the number of method entry pointsI agree that sounds like an elegant approach +Thank you +Patch applied +The distance between the titles and links seems to be dependent on the width of my browser window so it is not too serious a problem +Also please mark the Affects Versions as for related activities +Any objections on this? Would like to revision +Yes they are illegal +Let me know if not attach the patches you have and leave it with me +The patch is attachedPassed build test build on WinThe regression test from will cover the issue +I just went through and added the permissions for all of them +Some comments The change in to do with wasRunning is problematic +As indicated above. +Thank you +See if these changes fix Hudsons passed +At revision +I have applied the patch to produce a but the existing is a and should be seen as consistent with the declared. +I wonder a whether this is still an issue and b where the attachements mentioned above have disappeared to +What do you think of Marcos dirty solution to solve this issue? Does it have drawbacks that I should be aware ofI think as Philip Gust that xstream should work as ignoring fields that are missing from the class and Marcos solution seems to do exactly that +Applied patches to Branch and trunk +Kurt has started a discussion around some for this feature +These features only work properly if documents belonging to a group reside in the same shard +Any updates on this +I have exactly the same feature request +Been a while since I filed is possible that Netbeans IDE was open as an editor at that time but I dont use any of its convenience functions +I think were going about this wrong because we should be figuring out why the implementation isnt working rather than just making everybody use the reference implementation +Change committed again Transmitting file data +Problem was that the preview jobs were starting before the and were locking the model resources +I made to make sure this not necessary going to +Its not necessarily the that would have to create it they could do so in java +Verified in its ready to be closed. +To clarify the bugs I fixed in the existing compression tests The generated test file starts with a line number +that wont actually be a problem for me since all of the documents in the group should have the same facet counts +If you are using KDE what is its version +Tripped over this again but didnt even seem to relate to META +Please take a look +I fixed it in version +retargeting from SVN head to so they dont get overlooked in release planning +Also includes improvements from +Tracked this to line ERROR An unknown transport called https exists +Editorial pass over all release notes prior to publication of. + for the patch +We seem to have some patch confusion at this point +Heres fixed patch +Sebastian I had allocationSizedefault but the sequence had an increment of +To actually accomplish this I belive the +same applies to the attributes on query engine +We should probably move the discussion to the user mailing list +Well +I think I know whats happening here its a lack of synchronization on the classloader creating interface impls and so two threads might generate an impl at the same time and step on one another +I use version +Thank you for copying the dtds back with revision . +Changed the serverworkitem and requestworkitem to log processing exceptions with only the key information but the full stack trace at the detail level. +I dont see the error he is seeing and Andrews code compiles fine for if the code is compiling for you guys I think we can ignore Matts latest for this bug I believe its now fixed +bit and see if that works +sounds good to me +I do have a fix for this locally though +Resolving issue +Definitely ifyou can reproduce the deadlock report in a new jira we are off topic in this JIRA +I confirm that Paolo will work as a committer on qixweb +CAS Editor Documentation patch for the docbook projectNote This patch does not include the images +Im calling this task complete for tooIll investigate these CI environment issues offline +new patch +Thanks for the and the nice breakdown Knut AndersThis information is very valuable it would be neat if we could figure out someway to incorporate it into say our release notes since as you say this isgood news and backed up by solid evidenceIts great to see that all the hard work that you and others on the team havebeen putting into performance is paying off +Im marking it closed not a problem. +problem as described above is still present in When replacing the everything works fineCan some please replace the as described beforeYou can reproduce this error very simple + +I just committed this +I suggest you either try to get the JRE standardized or use another renderer if possible +I have documents that can have the same score but when collapsing is performed this only takes the first received document +The second set uses the first set +Doc is already updated for +What remains I think before this issue can be resolved is for the wiki to be updated +The patch break async calls +Each framework can have arbitrary final state like say ABORTEDToday we call as finalState may be we should name it betterI saw that in when an application finishes you used the column for printing the when the application finishes +This class does not exist in could somebody confirm whether this is a problem there as well +Pushed the new PR to provide steps how to verify this Jira or close it +doesnt call while holding a lock on the +I will open a new issue for possible fix of streaming LOB +Luis patch has made it in +classpath m not sure there is anything that can be done on the UIMA side +rev after release +Is that kind of configuration supposed to work +True dumb mistake +let me know when you need help on it. +It would avoid exporting conf for every tools +Actually LIKE and RLIKE etc are also functions but I agree its an existing issue that they are parsed and called differently in existing code +Thanks Jothi +about for each table extracted all foreign keys for all tables I dont see that in the code it always ask for exported keys for a specifc table +This is in keeping with the original intent whereby doesnt provide any default logging and you can use or Servlet error pages to handle your opened a similar isse +I will need to look closer into that but suggestions are welcome +BVT failure so marked it as Prasanna should this be resolved? I see commit from Prasanna to fix is working on KVM but is failing +Jstl component has been removed +Cant reproduce it too now +So one byte is serialized as two chars between. +Maybe the issue was a restarted server with a different IP but same name? In that case the master would still think the server has all these regions when in fact it doesnt? +what database are you running this onis it case sensitive? Then try and use uppercase in the reveng xml let me know if that works +There might be more work to be done to make the test more stable to be part of regression test +The documentation says that any jar in that directory should be picked up so they would have to be taking advantage of this bug in order to be surprised +Hi DarrylI sent the email out +Fixed the test failure unrelated +I see no reason to introduce special machinery to model this clause of the TRUNCATE standard +Start from the original +Allows the tests to pass on all jvms Keeps a single bumpversion target for releaseThis version of the patch differs from the previous version as followso the explicit printing of the version information from the metadata callso Eliminates any need to change the odbcmetadata canonso Adds the new metadata canons to the list managed by the bumplastdigit release target in toolsreleaseI have cleanly run derbyall against this patch +Having unforeseen issues with stored procedure handling +Im really hesitant about pushing JSON through a string +Added optional option as packetSize which just overrides the configuration value when merging +the eclipse project showing the FreemanI was a bit puzzled by your comment but got the confirmation +It is not apparent to me why you need to do this +Youre saying that you changed the file in your local repo and then you expect it to update again from the repo +Thanks +That was a joke dont spend time on that +I forgot to split them up sorry about that +Closing per reviewboard as fixedship it +Sorry misread that. +Resolved for on reopen to be able to target correct fix version. +Just a guessIll attach the failing WAR so you can dig deeper +So I suspect you would see the largest performance increase from native maps when Accumulo is empty +If an is thrown then it is a bug yes but because it is a missing +Fix failing unit test on of the major problems is fixed nowUnit test for the SFTP part would be nice but it requires some mocking as we do not have a SFTP java server at the moment. +In JDBC you can pool both + +There is no loss of precision +I couldnt reproduce it lately please anyone comment if you face that not reproduce anymore someone must have fixed issue is still present +actually if maxrank we set returnentirecachetrue maxrank +The patch patches srcmainjavaorgapacheuimaconceptMappersupportdictionaryResource but there is no such file in SVN +Hey Uma Im surprised that the RAT exclusion should be placed in the instead of the +thanks for yourCould you please explain how is generated? What tool is used in order to create the file? The fonts? Any specific configuration etc +Deal +I think it would be quite demoralising to submit patches that clean up warnings while other patches were being applied that inadvertantly increased the warning countRegarding the manual intervention we can mitigate this to some extent by using the annotation +HiI have a same issue when Im trying to map Domain Grails Class trough Hibernate XML + the patch looks good to me +It might make sense to open a subtask for the doc change if the behavior isnt changed to allow unfreeze from a separate connection +issue traced to several problems stale data in the local repository cache refers to a snapshot which is already purged from the server repository updatePolicy is set to always but stale data still shows up there is a SNAPSHOT jar and a couple of timestamps in the repository +Committed to trunk +Attaching a proposal doc and code for the first iteration +At this point it is clear and specific version of API requires a specific version of IMPL so the manifest entry is correct +I think this bug existed since was commited +Libor please have a look at it thanks +so use with care the patch for as well +The fix for +OK no worries +what behavior does this fix? just gossip trying to reach the old node +fixed and correct version is deployed onto +sorry +This has been addressed by +thanks close for Solr +Committed revision . +delaying till next release +Frequency distribution is not simple measure depends heavily on corpus composition size +bat files for that environment +Screenshots and other reproduce on Windows XP with +Simple enough fix +deploying are both appearing in the deployments dir way before the progress dialog finishesThis is not a bug +Blocks +Array sounds good to me too +Im not familiar with Spring but the of make me doubt whether it is really opening a new session per requestThis class is a concrete expression of the Open Session in View pattern which is a pattern that allows for the lazy loading of associations in web views despite the original transactions already being completed +Nathan Ive actually removed these as JSP functions +Postponing to the next as Frank explain changing default scenario could cause major problems +This isnt a goal for core Cassandra +the condition is different because that is what it did before +The findbugs warning is not new to this patch +SO do you really think a problem like this Where the hibernate gurus like you guys cant solve can be solved by us with in this short timeDont take it personal +Thanks a lot for the reviewcomments Owen +Pat mentioned to me offline that he tested via running a Hadoop tarball and verifying that it rolls when the log is at the configurable amount and that it keeps the configurable of files ie mb and which seem like good defaultsJenkins wants the patches to be split per project so youll need a MR and HDFS jira for those changes +As per +are used to attached to which does the buffering already +From some profiling I did it looks like the is spending of the CPU time on UTF and it could save about of it by using Text +The CLASSPATH in your shell is unchanged +Closing released issues. +Yes this also occurred to me when trying to adapt the workarounds attached here to work in the more general case +Override get to return primary key nonclusteredSound reasonable +Caching resolved resources is a very good idea for a new issue Having said that there may be such an issue somewhere in all those XML issues available +I am Python developer do you need this to be handled in Python +From a design perspective there are many different ways to address this +and again +The patch looks right to me even if it is not losing data in practice +The patch submitted was invalidI changed many code that could have optimized the spoolmanager but the change proposed in the issue is not applicable to further improve the spooling we should change the spoolmanager interface +Would you like to submit a patch +Also the Add and Remove Projects +unless your keen to look at them yourselfcheers +Yes my question is the connection waits indefinitely and does not connect to the vdb even when it is active +Thanks Prashant +But I believe we see the issue in nightly runs against SANEfalse so it is not specific to stuffwe do in ASSERT logic +You could add them to the module we provided or in your applicationGenerally if you look into the source code of Hibernate Search specifically in the you will find some great examples to run automated tests which include packaging dependencies and deployment on AS using Arquillian and the custom module referred above +Its here for ptests to pick it up and run +So really it seems like this issue could be closed +to trunk +Im now running into this same issue with a jBPM project +I will try to explain a bit +Patch attached +Thanks for the latest patch +Closing +Ill work on that +Yes! The good news this morning is that Hairong now has a good understandingof the issue +Yes I will change this behavior apart from that it seems this JIRA is complete +Christopher is this for an ASF Project? If so my preference remains a tightly scoped exception +Tests seems become very slows when page use ajax +Anthony What did you ever do about exposing the information to technologies +A locktimeout exception should throw a rollback exception and not a +Im thinking of enabling via Dboostrapcollection that will do everything necessary including copying local files to zk +There was a Struts announcement on about the GA releaseof +hi gabriel +The kernel did the correct thingWhat errno did you get? Easiest to just paste a terminal session showing the commands and error outputbq +Ill have a look at the factory bean and probably file a ticket in their bug tracker to make the implementation less invasive +So the only things that we have instead is the metadata +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +we need to test this patch by removing lib the newly added jars to the eclipse classpath +Should make it more but you never know +coretargettestpiruntest +you are correct again there is a internally which needs to be notified of the issues that have not moved in a year +What is the reason for it to return no sequence? Is there no possible valid sequence? If so the tagdict might just be incorrectOr is there a possible sequence but beam search doesnt find it +Attached is a patch for the server only in +Could we create a new ticket or sub ticket with the renaming work on the non stream that Martin mentions so we wont forget it + +ok whatever fix you think is appropriate +Hi ChrisIm sure you have lots on your plate but I would like to have this wrapped up for our next board report which is due on the th of June +Probably the first thing to try is to upgrade to Axis or and check if the issue still occurs +I will open a new issue to deal with requirements. +has no dependencies +upgrade to wagon done. +HiIn my opinion this is definitely an error because I am just reading an file via SAX and piping it straight out to another file +Hmm this sounds very courious +The latter +Shouldnt override skip to just bump its posCan we plugin various schemes into +Added +But Im dead in the water when it comes to browser supportIf it matters the version of Railo I installed was the all in one installer that included TomcatDo you think I might have different luck if I used a different Servlet Engine Application server +and +META +Moovida do you think you can also check this? Im not sure if there are any other ways of changing the edit preferences + sure the patch isnt right but Im submitting to initiate the conversation on the right implementation +Just tried seems to work +With the patches I have provided this issue is solvedAnyway I would like to hear if anyone sees problems in making aware of blockcontext URL protocol +Does anyone have an issue with committing this +Having process discussions on a jira will just get them missed. +There was discussion thread long back based on which I created a sample and its there on my blog as well +When I applied the patch it lead to a few problems +Wanna make a patch +so the bug remains as a reminder +I am having difficulty uploading and the because they are each greater than MB in size +Assigning to pipeline for investigation then fixclose +cd kafka +This patch addresses the points Karam has raised as part of his review +bulk defer of open issues to +It provide an option Not redeploy JSP files in server instance +Ill have to analyze this first +Jira will be closed once verified on as well +Checked in patch v thanks for the reviews +Can you please rebase? Thanks. +bulk assign all unscheduled issues to next release +I know we had a long debate about this a few months ago +Thanks Dominic +I have been able to reproduce this problem only for ObserversThe reason seems to be related with the different handling in +This meant that the tree state on the server was not updated but on the client everything appeared to run normallyI fixed this by putthing the stuff into the response to prevent IE from caching the results +Anyone else have opinions on if this should be committed +There are many paths that do not involve clients which still serialize objects that need would just merge your test result into the existing +Committed revision +Someone can use the artifact if they dont want more work to do but if they dont want that and choose the artifact then the work is on them to provide that themselves +In that case the VM is usually toast anyway so it really doesnt i would see two options explicitly ignore the OOME or only handle +i can understand that this is confusing so we have to change we have to remove the list or improve it mean give out the message that the key is null is the right way +and both can have multiple columns and also a required field for row +I identified the build problem and fixed it in my environment +In fact Im not sure whether problem comes from JCR or not I only see that the file content is stored not correct with special files as attachment +Its not generic but it may work for your use case +Would you please help to try it +I killed root on loaded cluster and I then killed server with meta so clients would have to go back to find new root location +I have proposed a patch that theoretically should solve this issue for both and +Matthew this sounds great +Event handling may it be application wide or between specific component instances is basically just about routing messages from senders to receivers +Indeed is part of +Resolved in fixes +ah and one more thing +The whole issue seems to be in +Looks great please make the sort relevant select drop downs use sorts now. +NoSo this would then mean version renamed to so all and every issues are now moved along with it +Then examine calls at the front for staleness and discard calls that are too old + a patch that adds cursor default to Andrea! +I have NO idea what is going onI think in Jimmy is ripping apart AM which would hopefully cover thisShould we attempt this at all on or just let it go? Please let me know +Attaching file +Then if the result of investigation indicates a bug you can file an issue in Jira +Right now we use standard JDK http client if we must use something else to allow doing this in an isolated manner we could work on goal is to avoid requiring to use the binary protocol which requires specific versions of remoting example on how this would affect the client +Dmitry ok I found your agreement +No we dont have to teach how to set PYTHONPATH thats explained in a gazillion places already on the Net +These catalog parameter could be made unnecessary by having a default catalog name related to the sitemap +I figured that this patch makes a lot of code obsolet so I removed all the foreach stuff and the Function interface +Second attempt trying to get each POM to explicitly provide their SCMdeveloperConnection attribute +This patch file also contains a patch for the file controlssrcruntimeorgapachebeehivecontrolsruntimegeneratorresolves the NPE issue +For example there wont be a anymore and we rely on Hibernates own JTA support for JTA synchronization now +Same problem here +Fixed the link on the page +But if there is a strong case for paging Im fine with it +I hope it does not include incompatible signature changes +This means that after I added messages I can wait forever for delivery +javadocs for and related records +How much time do we have left +Its very helpfulI noticed that one needs additionally to set environment to use proper settings +Central is more or less what JBT brings on top of individual components thats why I think it is a nice place to host a few lines of JBT branding +think its a bad idea +Thanks for the patch +Will with Mikhail +You wrote I am assuming topics that talk about default should already be in the manuals and are NOT new +The code relates to planting compare and exchange bytes which is only used in and therefore not by bloat hsqldb pmd and jess +In a way it does seem like there is a bug in Android JVMs bytecode verifier I dont think it should care about missing class +Any ideas +When I saved a local copy of the schema +So from my perspective its fine to include +Hi DanYou can just copy the following keystoressystestssrctestjavaorgapachecxfsystestwswssccertssystestssrctestjavaorgapachecxfsystestwswssccertstortwssecuritysrctestjavacxfColm +Contributed by Suresh Srinivas +Unfortunately I have no idea how to fix the checker Mike your turn g +In a later issue I intend to deprecate the constructors on various implementations that currently take a parser and replace them with constructors that take a parser and a writer but that can be done after goes in +In fact I plan on experimenting putting native maps in their own module in +Moving it out of branch +bq +This should take care of the problem OlegLooks pretty good to me +If we had different timeouts for writes than reads then it might be nice to use say of the timeout for the normal write and reserve for the hint phaseDifferent rw timeouts is being added in +maybe we should move the check into getKeyspace? theres a bunch of callers that dont call vCF afterwards +But its just my opinion But also a marker exception would do +Some strange conditions are met into this cluster which causes a random corruption to happen +Patch applied with thanks to Chris +Should be available in a day or two just testing the building +Fixed as part of released +cant simply swap in for the dest with any arguments +Not sure if you can force it to do so with luck +Automation of such a solution assumes we will integrate Grunt build into Maven buildWe can make running of those tests part of release process so far. +Thats only done on the Preview VDB sideNow the keeps track of its own Preview objects which are prefixed with UUID +David do you mind posting your Camel route +Ignored test for this in esbsmokesrctestjavaorgfusesourceesbitestspaxexamkaraf thanks for fixing the connector issue +Jobs hit by locking exceptions are retried immediately whereas jobs hit by other exceptions are not retried even once due to an oversight +Hi Arpit I did the work for what was the issue exactlyAnything I can do to help +my preferred is the first one you added today let me ask the some team members are not much convinced is it possible for you to propose also a few more options exploring different designsPS this doesnt mean the current ones are rejected in the end it may be the one with more votes +Thanks for the repro project will have a look. +Still going through the patch +Unfortunately that means the next release of the plugin will be and above only +Fixed number conversion problem +In if a datanode in the oipleine dies it is very likely that all the threads in that Datanode become inactive so it will not matter whether the thread or the packer responder thread exited first +I dont agree that we should just swallow the exception +attaching files from metadata failure. +will it if we have new string trim featureannotation +Ceki thank you for telling meI looked at the readmes for and to see if I should update from but didnt see any compelling reason +I committed changes to only +Editing of these books have been completed and will be included in the round of listed. +Hi KonstantinThanks for your commentThe question is whether every applicationframework should do that or the web container should sanitize the headersWicket now also replaces n and r with as Tomcat does but not only for the value but also for the header names +this problem appears again in build +The supplied sample project works if I update to. +Looking at the logs for I remembered that in their case the thread was able to read a valid id from the connection with a Nagios server +JDBC driver connectivity class +Ive committed the fix. +In my opinion we have two options we can to use the LOB annotation or to create a Stream annotation +Remove unrelated changes in other files +The DEFAULTTYPE is always meant to be a shorthand for the log type so is correct +The root cause for this issue is how branch commits are later merged back and made visible to other readers +This would just be logged on the NN side +What you describe is typically a horrible way to design software +Lets commit this to unblock QE folks while I work to stabilize + +Write me if youd like +Ah that would explain some strange behaviour that I have seen in a project using insideUndeploydeploy does not work unless I restart the server +Resolving as. +Ive just committed this to trunk and +This issue has been resolved for over a year with no further movement +fixed by tracking all operation objects in a special. +Hi Live Migration operation being timed out observed only on a multinode mgmt setup +Thanks again Christian! Such an awesome addition +Workaround jar for unique timestamp versioning a lot AndreasIm still playing around with the code you sent +How can we make this work if we spill on average once per maps? Or times for every maps? Right now the case requires spills collations and fetches per reducer +Thanks +It looks like this patch fixes only the NPE but not the reconfiguration issue as Ralph had alluded toThat still remains a gotcha as Logj should not silently ignore configuration as it does +Could a CSS expert have a look? +and rev +Just managed to get it fail with TRACE on TCPPING +Straight fwd patch +The looping through all the resources needs to be eliminated by offloading to the db +I tried cygwin on an XP machine and it worked for me too +This was fixed upstream and will be included in the next rebase +Small patch to this region picker +like a bigger refactor than we bargained for hereI think initial patch is ok +It looks like there are collisions with AWT synchronizer +I renamed the properties as per your suggestion +I think you did not read my commentsee my comment above +I dont see why users need to be aware of this and whywhen they should force it +Create a consumer using beas WSDL +PrachiThat is a separate issue +I am also seeing this issue +Could someone please have again a look on it +It will be included in the next components release. +There is java docs missing frommap same comment applicable for other public method +I prefer having a few occassional false negatives than having to wait for a long time for every build. +Done +It depends on the changes in which havent shown up in trunk yet Ill wait until trunk is updated then submit a clean patch +only oozie should be INFO hadoopmortbay +new patch moving into core and adds additional to I think this is ready +Fixed thanks +Thanks Eric! +If the values differ it will attempt to pull down the remote artifact +retriggering patch sessions timeout on last submission unrelated to change +Screenshots are attached +MingFeel free to provide a patch +I think this bug may be closed +Committed to trunk +Patch v fixes the javadoc the remaining javadoc on to trunk +checked in the case +It might be an interesting hack to move the map sort into the couchjs PID instead of the writer PID and refactor the btree to accept a sorted list to grease the wheels on the write side +I just committed this +Edited the char long lines rest is same +Committed to trunk with revision I dont expect there will be more work on this issue will close in a few days +The same thing already happens for bundles specified in the features files being handled so Im not sure why this little addition would be such a big problem +Could you do a ant clean test to check if you see this again +Doh actually I dont know we dont have any lib folders anymore in the repo +Ill take care of the branch +OK thanks +Current patch is unlikely to introduce any of those +I am having the same problem and look forward to this bug being fixed +This is not working for me +Im not sure the best way to fix this +In my case contains only comments so there is nothing to compile +Dan could you please try and get it in or at least a patch for saw this error in our node cluster in which the network connectivity to one of nodes is not reliable and has high latency +So theres probably a piece of the module puzzle that I have missed and it might be due to how the JMX module is currently started Im not sure since I didnt write most of this code and I would most likely need to spend quite some time with a debugger to understand all the details +Using it may cause a performance hit but its arguably better than occasionally seeing what may be the clock going backwards +If the original request did not have a host header set manually you would not need to change it in the first placeI suspect you are trying to fix things in the wrong place +Stefan I just filed a bug in Spring Core +Thanks John! +and other missing there is one called usequerystring as well +Everybody should update the code and raise the version constant to +Now vjoin only uses live docs +I just tested this a few times and it seems to work +I think that ball and are close enough as to be indistinguishable for practical purposes +As above were pretty happy with how JNA is working for us. +Ill commit this momentarily +Thanks for looking at this for me Stefan +Base any patch against trunk +Just added the required description to the rules +Vivek please review if this issue is also solved on desktop or not need to think over the future of dojo based desktop engine +This issue is fixed now in +The block file of blkB on DN shows that the block size is +So my suggestion is that we do the version number change in trunk and redo it when we have released +Third times a charm +Guys do you think its time to create the branch for series +Ideally this patch should not break the backward compatibility because it just adds a command +Feel free to use the simpler diff format +We can keep this blocker open but a temporarily comment out the failing test as you suggestedCould you provide the temp patch +The diagnostics mcast address and the udp mcast address must be different +Updating Fix For to Unknown on issues not targeted for issue out as part of JIRA cleanup +This patch fixes the test +Thanks for +they return false +BTW feel free to steal this from me if you are planning to work on it +Closing issue it has been resolved for a long time and nobody reopened it. +Matt I will close as cannot reproduce if I dont hear back can be reopened this if it is still an issue but at this point I cannot reproduce it and I have done multiple reverse engineer commands without an issue +should this new node fail loudly or just log and keep going +I guess one could call it a design flaw that nobody thought about what might happen if the statistics were slightly out of date +If they start with file then the exclusions will relate to absolute path +I assume you want this to return before any exception is thrown +Unfortunately this bug will exist until we have native launchers for each platform + testCaseMoved just validation of failing test to so this test can run in daily smoketest added keyword Pri to Low since this only happens in edge condition when user supplies that ignores fatalError +I had the fix on a local branch and forgot to push it +Thanks +they shouldnt be there +applied patch +As a result the main class becomes smaller and easier to understandPerformance is faster with this patch mostly due to the caching of the default values of subproperties +has been released closing issues. +a category somewhere in between unsupported and formally supported meant to flag that we need to exercise more care +untar the attached file in hadoop home dir +Fixed a bug in a client API that was causing a Ping message to be sent immediately after Process timeout +This looks like a duplicate of +has been marked as a duplicate of this bug +bulk close of all resolved issues. +Well do that and get back to you +Address above review comments in v patch +Assuming they want to continue with the old behavior would a resolution at this point be as simple as take down the server chmod the and add true to the file +the l and f thingy +My apologies if this was just a thread safety issue the entire time +Thanks sadhu +I had seen the same failures which you reported when I ran derbyall in the new trunk but now I realize those failures had a different diff +Rename the C fields to c and the test will pass +And on the unit test front there are now several tests in that show the version update behavior +Nigel built most of the releases I think + +Since the JIRA is up and down most of the time today we could not add the comments hereSo I am for committing this and opening a separate JIRA for above comment +Im not really sure the best way to do it yet +I attached patch +Fix included in will not backport to . +Should have been fixed in but got overlooked +table false doest work as well +is a revised version of the sysinfo patch proposal incorporating comments from reviewers +The contains the exclusion configuration to prevent unnecessary dependencies being brought into the build +I like the new title better. +Ill remove it also +some thoughts get rid of compactions during startupshutdown +Added the test case to trunk +Attaching a patch with the test case +I did the reversion because I got on the changes and I also wanted to have the version that I was served up when browsing match the content that was on the screenI wasnt paying enough attention when I first edited the page to know if the defects were there already or were some conseqence of my making changes. +Patch file has been fixed +Please review and commit +Running run Failures Errors Skipped Time elapsed as testsuite was broken hiding a couple of regressions +Don thanks for testing +rUbuntu smoke kernel +Fix package task for for Committed revision Committed revision +Hans +please adjust the priority to indicate that the Jira should be considered for CR +Kristian Rosenvold At least basic support would be really nice so thank you +Sorry Remi no idea how to do that +It looks good for me. + +Merged into the master and branches. +As with Lucenes this version is implemented as a Foldersubclass using the decorator pattern +Patch includes fix test case and modifications to the file +the reason for the exception was I used instead of Asset it would be nice if injecting an asset which is not defined in the jwcpage definition will result in an error similar of defining an asset which does not exist +Need to make sure fix doesnt break AS test suite before closing this one +I agree with adding a stored content field but I dont think we should add any more copyFieldsOne of the biggest out of the box experience items that people make their decision based on is performance so we shouldnt make the example schemaconfig slower +Excellent +If there is exception with that too we will just let the existing code take care of it the way it does todayAlso another outstanding issue was what if user has provided user name and password and protocol level fails then should we still try protocol level with no user password and passwordWould appreciate any feedback +Only services I see there belong to +Fixed +Whether it will be part of or not really depends on if we are going to RC as or closing all issues that have been in resolved state for more than a month without further comments +Tom have you had a change to add the taskdef for +Committed thanks Paul J +bq +This means user will need to edit that file +Minor one pls specify the unit for value +This can be something like the G download which enlist all the version on one single pageThanks we need to keep this JIRA open for all the GEP releases? +Can you list the specific changes that you are proposing? Asking as some of the configs that are common to all schedulers are termed RM configs soBTW I did think of fixing all config names but felt it was too late +I verified the test on Mac and Windows +There are three tests in the package and +It is set to For installations that are seeing this problem maybe they can change to a smaller number say or so +fixed +I misunderstood I thought this was Confluence site +Suresh your proposed code looks good and simple +Whilewe dont have a solution I think we should disable the automatic update feature +Was thinking it not critical enough issue to go out on branch +Oh Jason renamed the Mojos so the patch will be a bit hard to apply +allows setting an which is responsible for creating the readers and writers for reading and writing properties files +Works fine +Sure thats was committed to branch +It doesnt work for me +Can you double check that this is still an issue? Otherwise Im going to close it as a cant duplicate bug + has been marked as a duplicate of this bug +Thanks +tested on +Changed hash code implementationI am not sure I follow what if if it a Text Writable ObjectWont it go to default case and error and String are both STRING primitive typeThe relationship between Text and String are the same as that between and Integer they will have and which both implements +So I suspect that the problem is with the deployer +This is the fix I made to +Thanks. +I think that would work in theory but given weve had a lot of production testing of the patch as is Im a little nervous to make that change at this point and lose some of that confidence +This should save a trivial amount of time at startup when processing config files. +Assigning back +registration authentication acl dashboard recent activityaudit tags license management Maybe these should be each separate issues since its just too much code for an issue to implement allFull module support in Click would be the killer feature indeedWith the help of the above modules a new Click application could be ready in a few minutes +Next time I will run the tests from within Castor first before did some debugging myself and you are right the tests still fail however for a different reason +Reviewing will post comments shortly +testcase verifies that the primary group of the user submitting the job is used as pool name +and it is mentioned at as one of the things to be doneCould you contribute this code as a contrib +Fundamentally ASF votes on source being released not a tag or binary but the source +See the attachment Source Code Java Doc TestsWorks fine at my place although I already noticed a small typo in the comments +Else it is the rudest thing I saw +Having show create table XXXX will be very useful +I would suggest looking at and for examples of using +Transitioning to waiting for folks we are blocked on this for our first TLP release +K +Cant repro either now +VERIFIED +Testcases are passing +Very good +Other than that it looks good +Marking closed. +Other than the schema everything else matches dist +Moreover the security is correctly applied +What I meant was that it can be implemented as a replicated cross which seems to solve the problems of implicit dependency and using distributed cache +Bulk change issues +Can you add a unit test that illustrates the problem +It is important to create the list of all the dependent manifest creation is so slow I will not commit this patchI think a good compromise would be to rewrite the target to only make manifests either when a system property has been set from the command line or when generating release artifacts +This patch implements the assertions above +I thought that hedwig client already handled similar issue well +Closing all ResolvedFixed issues lacking a Fixed Version +Tests will only run with profile otherwise Maven is only compiling the Java code of the tests. +trunkCommitted revision + for the patch +dimsPS I can run ant clean under +Moved executeStatement and bindFields method to +Ill do another pass and post some docs on over thereThanks Jonathan for doing this +Do you think we should get rid of that? agree lets keep the simple case simple and try to create the queue if it is possible but in a consistent manner +Mass closing all issues that have been in resolved state for more than a month without further comments +Please that youre using the correct Commons Collections version if necessary using the exact from your Quartz distribution +Ive patched the code in a different way +Revision of namespace proposal with additional optional class association in +I still think this should be removed or moved to the MTQ query itself then a setting on the queryparser could set it or a user could set it +This seems to be a very good feature +There wont be a release +Looks much more promising +Unfortunately trace contains this warning as well +I see that the code in which calls was added as part of the changes +I uploaded a new version of the patch +Cos thanks for your suggestion +So I added a second source lookup that runs after the Java source lookup that can find Java or groovy files +I am excited to try it again +I was hoping for an embedded Jetty for example +Something to do with the Hadoop version on the jenkins machinesTed might know the details +This occurs for example when in a narrow typesetting width only a single word or a part of it fits in a lineI am working towards applying this patch despite theseregressions for these reasons This patch is a good piece of work and a step forward for FOPs layout +here abeebfbafebebdeaThank you for your contribution Hari! +OK thanks Koji +But I thought today that not all clients may understand such a soap fault +The problem with my approach is that I duplicate their byte array as a String before its garbage collectable which will aggravate the memory requirements +was not reestablishing upper join groups which caused later planning issues since we didnt detect appropriately where criteria was coming from +Improved Thrift Actionscript compiler and version of the thrift compiler for windows which generates the new style AS interface client and processor +Closing issues for beta . +testsfunctionalorgjbosscachereplicated +For some reason Apache saw the previous patch as a binary file +Verified by Aleksei. +AnnopI open a new issue and upload a simple patch could you tak a view +Was just a test that I didnt change for the new ordering +Ah I see it now +Good find! Ive taken the liberty of updating the patch to work against trunkNot sure if Im allowed to when Ive uploaded my own patch but from me +but doing nothing is even more dangerousThen we get complains that wicket doesnt set sometimes a value without reporting that it couldntno i dont want to kill the exception +Updated patch with simple test had updated with wrong patch +I reverted the change +Move the fix to +I took out the exception but theres still some INFO logging +The wasnt getting picked up fromSimple patch +bq +Fixed +Bulk moved to +Im attaching a screenshot of my terminals in either of them I cannot read current working area +So if we want to tidy this up we need to alter Section to require the client proxy to only delegate business method invocations not all method calls +bulk move of all unresolved issues from to +see also +no traps recieved on both snmp servers consoles +one question why open is put in statistics section? this settings would be setting right +after to trunk! +contains info about the solaris machine that doesnt crash and the one that you probably wont see eth on solaris the most common is hme +Since no modifications could be made while rehashing the invalidation issue cannot happen since it relied on being in use and this is only used during rehashing +working on this +Updating Fix For to Unknown on issues not targeted for updated to reflect scope widening from only path being unable to use heartbeats to all protocols paths currently being unable to support good to me +Rescheduling to +The board recommendation dating back to about author tags is highly contested +Okey doke Ill sit and wait and bide my time +region hyperlink show regions correctly but it doesnt open corresponding resources anyway +More performant too than doing two passes +Stan can you clarify your acceptance criteria above regarding make the greeting text smaller +Status +I bet this is something hive hive combine input format is going +Thanks Uma and Ivan. +Verified that derbynetclientmats skipped jdbc with JVM and ran it successfully with JVM +Im working on a fix +issue is ack before add reference is complete subsequent add would leave a dangling reference +HeyJust started to look into it now +Thank you +I have an idea about what might be causing this +Yes I see website isnt updated +Shelly +Fixes compilation issues for test class srctestorgapachenutchprotocol and temporarily make sure that the expected values correspond to what is returned by Tika +Closed upon release of Hadoop. +Confluence detects if you havent made any changes to the content and wont do anything if you just open a page and press saveThis means to keep the docs current you must periodically open every single page make a change just to force an update save undo the change save againAnd because there is no way to know what pages a snippet appears on or which snippets have changed this must be done on every page with a snippet on a regular basis +Also please be careful not to lose the Deprecated markers that James introduced in and watch out for javadoc warnings following the code move +All the files compressed together +Ive uncovered the problem and have discussed the fix with Michal FojtikBecause does not provide the nic attribute in the VM listingwe need to query each vm when getting status for a single VMThis Jira is encountered when one of the instances in the list isdestroyed between the time the list of all is generated and theactually queryThe solution is to ignore Entity not found error in httpgetby using a beginrescue and skipping the instance that failsto fetchA longer term solution will be to request to fix theexpand nic attbibute +But he is not sure about the older versions +Im not sure why the temp file AAA +What about the non ordered ones? executed before or afterDefault order is then you can order as you like or ? +Peer review completed before release +I could successfully run random writer with hdfs +Hey Aaron +marking as blocker the qunit jQuery license and copyright should be fixed +So should be all set +First run with mvn clean install and then run with mvn webstartjnlp true X +Hi can I provide you with any help to get his resolved? I am one of the Castor committers so feel free to throw any questions towards myself +Instead you should use the injected to create and send the exchange +Fixed an issue in the Scan benchmark +Almost four years passed +These new options are distinct from it would be reasonable to have them set to true and useip to false and vice versa +The test of runs ok currently +This patch fixes the concurrency problems with the previous patch which were disclosed by the experiments described onThe fix is to go back to updating SYSSEQUENCES in a subtransaction of the users execution transaction +This patch does not work correctly if the add command happens before the file is available on disk +al you have the solution right +This is with earlier discussion on this issue as per Kishans comments. +There should be a way to balance the reducer utilization and small job latency thoughts +Unfortunately infra is not at its best today so Ill try to create a snapshot tomorrow. +Eclipse didnt like the trailing +Hasan please see Hoss and my comments above. +Ive just committed this to trunk. +Ups we should fix the typo in this error messageThat explains your issue the tag dictionary denies all sequences beam search could advance therefore it can only return null which is causing the NPE you seeWhy isnt cross validation not checking the dictionary does it not support it +Committed patch to resolve defect +Gary any idea of whether or not this is an issue with just the mvn jetty plugin? I have not had the opportunity to test yet on a full jetty server? I should have specified in my previous message that Im running the jetty maven plugin +All the issues should be fixed +I expected to be coincident with other functional components like hselect hselect richorderingList +This is a signal to us that the keyboard has popped up +I cant seem to make you a contributor and assign this jira to you +Please unzip +see it seems were good to go on this one what do you think +into there is not hashCode and equals anymore. +Assigning to Nick fora was not? QE passed it inb What about? Should we also add more to the site to avoid that problemc need a PR for updating content on too as that site must mirror whats on so that JBDS Central can pull from +Wouldnt I just copy and paste what Ive given above into the patch file +Committed revision . +John thanks for the tip it works +Hmm maybe I dont understand the context but shouldnt a pool expand to have one instance per calling thread? I wouldnt expect any hard limit to be imposed +Try entering info or ls to see if you get that other commands still have no effect I would like to get more info such as OS so that I can try to reproduce the specific environment +Patch to fix this If I didnt get what you said +In an email discussion of this particular feature request I suggested that the solution for this issue be made generic enough to be used with all the of the types available in the package +So the th attempt gets killed will not fail the job +But there are the sites which have improperly encoded URL links in Japanbq +Yeah current behavior should be preserved +No new test cases added as the existing checkPermission test cases in already exercise all of the hasPermission code +using URL encoding and decoding of filenamesThen is it an incompatible change? Does it work with the previously created har files +Did you plan to reintroduce the feature +I hope I havent broken the world +I like the function specification that you completed +Sure +test case +Added patch using constant strings rather than enums +Ill commit this +LGTM committed +bq +If this is the case the problem is not that it doesnt work but rather that it doesnt fail sooner after recognizing that you are trying to deploy a car file which has already had most of the deployment actions done to it +Ill have this reviewed and committed by the end of the week being in the middle of a relocation process slowed me down a bit more then usualIts high on my list though since it seems to affect everyone and makes the trunk unstableInitial review is good but I need to run it through a few test cases before committing feature list sorted +Thanks Thomas for your helpCould you test this issue +Latest workspace and a jstack as blocker to keep attention on this +I missed these binaries as part of the original checkin +I agree with Jacopo +The mappings are a pretty large amount of data to carry in memory +Well be going to the filesystem for every store file on the regionserver to find its size +But in we expect shortname to be the lookup key for the delegation token. +Some of the above appears to be a logging issue but it still seems that the first node correctly updates the DRM replicants but the second node doesnt see that +Verified upgrade latest build MS ASF GAhost KVM ASF GA +Is vnode enable? Its not recommended to run hadooppig job when vnode is enabled +Leaving you with a very short rd line +Ill convert to automake sometime soonitll give me a chance to better learn those tools +There wont be a release +New SNAPSHOT deployed +With these changes Ive been unable to reproduce the issue +markw bin Welcome to cassandra CLIType help or ? for help +bq +In the meantime to work around this problem I would suggest you do split up yourdocument into page sequences and try increasing your JVM memory space with theXmx vm option flag +Initial implementation rev +This can be improved +Reviewing the migrated tests I only found some duplicate test methods that I removed together with some test cases that where disabled for all database engines and didnt made much sense either. +Applied addendum to branch. +So we would have the same kind of configuration with tolerances set at integratorsolver level and maxEval and function pointer passed at integratesolve method callSince we are just in the phase we change interfaces this would be a good time to add this parameterDoes this seems reasonable +Im concerned we might have a causing these spurious values Ill start digging +Thanks Andreas +What font are you using +Transitioning to waiting for what does waiting for Infra mean? You closed as Wont Fix +Update was made to correctly handle relative and fully qualified bootstrap. +thanksagain +Is this patch necessary after the recent fix toAlso Im uncomfortable with putting the does SOURCEPARAM need to be added logic here +scratch that +Geronimo is compliant +Maybe this change could be made forOh when I saw the CI tables tablets were not evenly balanced I did not notice that all tablets were evenly balanced +This looks like a bug in the browser +for the one in the middle +I must say I dont feel a huge imperative to fix this since it obviously is broken on the rd party server +is this still an issue +But what happens if you add ? to end of your new line? Might be and therefore useful figuring why kill of pid failed +Here again going with the philosophy of not changing supplied manifests we should be taking dependency information for the embedded library from its manifest rather than scanning it. +I also had to modify putNext to expect a piq schema without the implicit wrapping tuple +Moving back to for now +Access the verify certificate link to verify that the certificate is downloaded and installedSummary of the scenario CA is setup a certificate request is submitted through web browser and issued certificate is downloaded into the web browser +All unit tests pass so marking as resolved +The IC is looked up from the thread local not from the transaction table +What do you think +It still cant magically fix severely corrupted files like the exhibits here though +Thanks Nicholas for the review again +Introduced a new class +It would also be nice if you could specify a different repair rate for and repairs +I have this situation in my project +Indexes and Constraints +Jason arent you the NRT specialist here +I committed the patch +for the change +Should be ok with Metro too being a fix in JAXB. +Evgeny Ive discussed this issue with Simon and not make any new metrics available in the alert service +But I would prefer not to change it and keep it in this way not to break anything +Please close this issue if you are satisfied with the result +If there is no value selected there should be no reason to submit +Check this will not help attach at list here +Changed to call on nested Number class in +Is that the ideaEDIT Actually that would be invalid with this patch so the question stands +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen + +Also the tasktracker and jobtracker commands dont make sense in mrv +Patch applied thanks +No more comments after months and Im pretty sure the issue is fixed as per the above mentioned fixes. +I do feel bad about cancelling this patchBigger picture To be honest I really dont see this improving things that much +resole the previous findbug error +Seems this issue is still present in Jboss EAP final. +even though its instanceof +The patch is incomplete it only support a natural id in basic properties +Nice one +The httplocalhost application is always using the in the regardless of the value set in which appears to be ignored +Patch attached +The sorts documents in the index according to Ids or Payloads each of these is a diffrent implementation of the Sorter interface +in the latest patch new features? bugfixes +I think a space in the location of GROOVYHOME can also cause the problem +clean b installed removed this error +NET has been added to the trunk +v +Gary WDYT + and added +failure is unrelated and I filed to track it +Fixed the patch to work with the trunk +Offense was not my intent +for we can also make it Comparable and have a custom equals this way it can absorb the corresponding methods from the class +Those would also be valuable and reveal nothing about your data +Jay created as subtask +Needs to be in its own assembly under security +I kept what was there as is but moved the logging of the request outside to kafkaapi +Hi CarstenCompiling Installing the Apache Felix Inventory and Apache Felix Web Management Console the menu switched back to a flat action list as shown in the attached snapshotI am closing that issue as it is already fixed +cditests +I dealt with exactly this same frustration on the test results about a month ago +I believe there is a problem calling the stop method twice although after cleaning up the code my issue remains +Reassigned to Burr to get additional info assumed temporary problem and definitely not specific to hibernate since its just a perspective switch that is causing it +So everthing came back toNote about the last commit in trunk I have merged lazy loading of the proxstream +BTW is it OK that we can see times the erase method with the same parameters in the call stack +Bringing into +Flavio +Then again the new call on the new Scanner it wont throw any way +I watched your wiki pagesBtw Metrics uses and of Hadoop +Please let me know if we have any disagreement here +I actually changed this to LATER a while ago but it seems it made it back into as wellThis is imho just a minor issue that can be fixed later +Keith suggestion sounds good enough +Marking this as incompatible since any applications that had dependency on old jackson version now break and would required change to upgrade +My apologies for the lack of class comment in the annotation +Ill attach the logs from the failing tests +This bug has been marked as a duplicate of +So I reviewed the code again and for some reason I must have been blind before but it looks like does indeed implement the Val interface +If an application then tries to allocate a block for this file the block allocation will fail +I have attached which is a modification of the original blueprint sample +We can always merge other users post import +FS impls out of common is a good ideaBut moving all of them to a single module is not a good idea +It seems that this bug is still present in as well +Ive added a check if parent is null which should fix the NPE. +The line is there in every elease and has been for years now as far as I know +This is safer faster and easier to reason aboutLet us assume that some languages have practical difficulty with this kind data structures +With the patch it does but it seems to be being passed an invalid har URI +Looks like it have something to do with Seam? You can try to deploy the unzipped to AS deploy have tried it with a simple EAR and WAR file and it works fine +I missed the signes in the stack trace. +I deliberately programmed Scenario behavior as generated by areas die when you navigate away in the Project Tree view and raise an exception during undo +The second case should be if a patch is only fixing a private distribution but not Apache Hadoop trunkbranches +Thats what I read from the initial issue anyway +So its a matter of usability versus not logging warnings +I digged a little bit deeper and found that this output is from hibernate not from we upgrade to hb next release we will have to this problem with hb +Could you please provide a patch for this? Thanks +A patch to fix the issue +on patch because test cases under srctestorgapachehadoophbasemapred no longer for the just committed this +this affects too +Hi lars Thanks for the patch I will review and apply it on a few nodes in our cluster to monitor how garbage collection is affected in a high volume environment +is the one that changed this behavior +Im still getting used to thrift +Hey Suresh possible you didnt regenerate the patch with these latest changes? I dont see any mention of in the latest patch +Ill attach two text files with the plugins names and one excel files with the comparison +bulk move of all unresolved issues from to +Thanks LeoPatch applied to LUNI module at repo revision rPlease check that the patch was applied as you expectedAll current LUNI tests pass ok please send along a regression test when you have one +Lets try this out on releases +doc looks good for me +The test appears to be running fine on now and also on Java without Xalan in the classpath +Well also need a timer if the fork locks up to count as a failure +Used for a months or two and seems to work well. +bq +and even then it seems unlikelyIs it possible for you to provide the actual application? I could probably fix this in minutes if I could reproduce it locally +the bootdelegation property between the system and custom file are already aggregated +Do we know why this has reappeared +Any documentation patches are always very welcome +Yes that is the problem we are now trying to figure out how to put xerces and bcel in the endorsed libs so that they will not conflict with JRE versions +first phase complete Dierk can you make sure there are open issues for all the above points you feel would be valuable guess I have now opened issues for all respective topics that deserve an own issue +I am not sure if its an ordering issue +Start the polling thread and jetty thread by using Interface which will be implemented by the Application Server +I have a project will aliases so Ill test and apply Peters patchThis is a rather strange issue though +This is not an error in +Is youre preference to just match based upon a prefix string? Are you also requesting the ability to support continuation recordsTEXTTABLE has support for quoted values and the use of an escape character +There is a big improvement comparing with the old one +Or to be precise I see the failure when I run +Dont know why it created twice. +if xxxfoo is absolute we can no longer assume xxx is absoluteIs any reason or benefit of making X relative +We are dealing with special Overseer considerations and I dont really see the strategy for that as a consumerA standard consumer of the queue API does not necessary need to detect this kind of zk connection flap Its kind of Overseer specific I think +Lets close old ticket that are not in demand. +for the patch +I would really love to be able to specify a different thread per Connector or + +Or we can disable the two failing tests +Im agree that we have serious miscommunication between GC and JITJIT does not expect that GC moves objects during the enumerationIvan does objects during the enumeration process +Will try and get a test app for chance of a test app? Be nice to have this fix in the next release +just committed this to trunk and +Since organize imports is also broken I think it is critical that this bug is fixed +Do you see it +When the time threshold has reached force the writeTo function to returnAn alternative approach would be to make the s Acceptor smarter than doing to assign requests so that incoming new requests would be assign to processors that has shorter response queue +Thanks Arnoud +This patch gives Idea about where I am going with this +I think he made the changes to for the drop +The patch also adds the property to the Infinispan and Cache connectors and the simple model in the JPA connector +Initial patch for review +org is broken for JDK provided by AppleTo compile it on your own you need to configure likeCFLAGSm arch x +Verified by by Denis. +Hmm good point +Could you please try and capture the request and responce using tcp monitor and send it +Patch against Henry in the future please set the affects version and target versions fields when filing a JIRA and only set the fix version field when the JIRA is resolved + to commit this but looks like stack already fixed this in r +No problem Adrian I will do it tonight +I think the real sticky point here is in providing functionality on the component that isnt exposed via annotations or any of our other process control mechanisms +Description mentions renaming a ProviderID Channel as an example so that maybe it does not exhaust the matter +Fixed the code to read correctly. +Attaching and committing it at subversion revision +Is there a login required for this repoWhat groupIds will be coming from this repo +We might also want to check whether sstable is sorted and there is no overlap +Inverting lucene documents is not done with the help of a thread pool +verified by Anton +This patch addresses the race in the test and cleans up the unused Conf parameters +adjusted subject its not SPI tests that are failing but tests against the JCR API +just went into + +The one in was put there by the patch you sent +I am resolving this issue again because its already been committed. +is this issue due to some jar conflict +Closing resolved issues. +Done. +Safari console screenshot with websocket connection errors +Bulk closing stale resolved issues +AdrianYour patch is in rev +Done feel free to release +Seeing s intermittently through the day today +Added new scripting property to Field +Ill see if I can figure out whyOne other note became because it no longer has any test methods of its own +rebased and new patch attached +What I dont understand is how you expect that to be altered +Thanks +ll bump it up to ish tomorrow and see if that fixes it and get back to you +Attached new dump from latest SNAPSHOT +One of our customers has expressed the interest in this JIRA issue being implemented for the ASEAP +Thus our assumption that they never occur on Oak paths is wrongSee notes in the source with respect to this ticket imply that identifier paths should be handled in the path mapper +Attachment has been added with description simple one line patch to fix Adrianthanks for reporting the problem +I still think we should remove the alias command +I cant grant permissions any moreIt has been escalated to Atlassian +Thanks for raising this KenJuergen +Manually tested! + for the patch +We could do the same in the associate factory objects to property paths and let the bean wrapper call them when an object in the path is null and need to be do you thinkCheersSergio B +Ive made almost no test but the following should workjava conffile pathto ivyfile pathto publish theresolvertowhichpublish revision if we could have a little more docco on the command line that would be much appreciatedThanks further info on what is needed I set this issue to resolved +Let me know if their is room for improvements +With projects you are forced to publish the classes jar +call it hflush? its the new styleI did not try your patch +Commited the bundle replication module to trunk and merged to branch +either per destination or per connection +The problem is that was uploaded later so that will be tested +I also ran just in case and got not failuresIve committed the patch and resolved the issue +Updated patch +Thanks Raghu! +Hi DamienI have patched you v version and run the mvn package it pass the unit test +This sounds like a good programming practice which could be applied to the whole code base do you have any plans to do this +This forced the use of a regexp where the unstripped quoting was valid +Native procedures for JDBC LDAP Salesforce are added +If you find further issues please create a new jira and link it to this one. +verified by Dmitry +Ill look into both ways but if anyone has input it would be welcome +As Nieraj discovered this appears to be a regression from when we rebased on Eclipse +I was not thinking of new release notes for +I am referring to the patch posted on rd feb +The patch does not contain any author tags +I think we are good for GA. +Please verify. +We dont have a working patch yet so this is rescheduled for focusing on cluster and partitioning late +This patch should fix the issue +I also marked IWCset as expert is my first review round +You can do this by adding a file called to your tomcatbin directory with the contentsJAVAOPTSXmxmexport JAVAOPTSTry that and let us know what happens +We have fixed the decoder to support TCP last year but we are likely to have left some bugsIll give it a some unknown reason the KDC server receives a PDU starting with bytes which contain the PDU sizex x x x xA x x x x x xA xAre we missing something in the protocol +Is there any news please +Fixed as described +This plan looks fine to me +Might as well do that under a separate issueAgreed! Progress not perfection +Since Namenode does not assign it can not work around like the the current patch +It should be CR +In trunk this situation doesnt arise +The attached Java program demonstrates that multiple copies of rows are returned by getrangeslice +I think that its ok for XSS but I did not check yet XSRF and Session Hijacking +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I have added this dependency to the BIRT feature +STS and our plan to improve that story +In our tests it only occurs with HQL +I tried a number of different options like forcing it to but none of them seem to alter the results we get back from Java for this +If so the source directory could be HARed and the result files then using the current RAID +If someone can take a quick look at this patch I will highly appreciate that +This patch will add FTPS support to the current trunk + +What do you mean by Please make a more robust grammar if you implement this improvement +I expect I will be finished within a couple of hours +Using the management console to pop the top message or connecting a consumer that doesnt use selectors will allow the async process to terminate +No update for ages assumed slipped to next please set the appropriate fix version + to set correct version. +according to API guideline api has to return the only lower case api name +Can you pl update it +Thanks MaciejYeah you are right +patch added +Minor change in Uilabel +push out to +Im thinking the right thing to do is mark it affects then move the for the issue out to and make a note of the partial fix in the release notes +is a valid representation for in US while in most European countries it would be +So all issues not directly assigned to an AS release are being closed +Tested the patch attached to this jira by adding an assert false to and made sure failed as expected +Thats a bit of a mouthfulWe could call the package authoritymapping since the net result is to map to authorities and use attribute rather than string +Added experimental to the Javadocs and I probably need to add more +And if they do use Spring classes as remote service arguments I would strongly recommend using the same Spring version on both said I will add appropriate serial where necessary +Patrick thanks for the feedback +Ill rather keep this issue +Robert unfortunately here we are signing jars in a folder while the jarsigner plugin only capable of doing project artifact and attached artifacts the last time I checked +Expect a fix later today +Do you want to verify any test cases or should I just close it? concurs. +Warning was due to +If this would create a problem for the jobtracker asit will not be as with the running tasks status as wed needthen running tasks status updates will need to be somehow piggybacked on thelight weight heartbeatsI think that once we agree on whats done in each heartbeat type we canthen decide on what to call each type of heartbeatsKhaled +This was the result of the s replacing this with the containing class name +Also youve added an import for for a single line in the code which just copies an input stream to a string I get some time Ill update the patch to address those but if you get to it before me Ill apply the patch and get this into the first release +a quickstart project to demonstrate the the in a Form itself solved the issue +I dont think we need to retag this +The trouble is that theCONNECT method takes its parameters from the client ignoring those set at themethod and host level +A copy of the prior attachment but modified for WS that is bundled with AS +Adam +Deboosting expanded synonyms is quite important I think +Close after release +Can we run commands in setup before execution of the binary +Im all for committing it but I wrote it +please provide a quickstart or testcase which which shows the expected behavior +And when they are complete it closes them or releases the references +No Im using +Embarrassed +Assume closed as resolved and released +Reopened because reproduced on taskTrackers on HDP +The updated patch works fine no memory leakage also now its significantly faster +Sure that sounds reasonable +committed with changes noted above. +latest patch looks fine +Also I see some can we remove them +Example for earlier versions not to be committed. +CraigI know you are busy so just looking for an update +Although it kinda looks like eye glasses +applied to closing all issues that have been in resolved state for months or more without any feedback or update +BTW Is there any reason NOT to use commons for this +Patch makes sense +bq +I think I just pulled before and was still looking at a trunk containing the original version of the patch +Rob I upgraded to and all is well +In the finally block all which might have been thrown are logged and then programmatically swallowed +I just checked something in for to log the term and other info +Should be some time this week +For now we handle this by hand copy an existing profile updating the name apply changes then make it default export checkstyle pmd and findbugs rules and commit the XML file to source control +Unscheduling this issue until a correct patch is older issues for Apache since were no longer actively working on these at the moment. +I will try again shortly +My unit test for this webservice is passing but my actual webserviceis still failing with the same errorI dont know if I can simply it down enough for someone else to duplicate though as I cant figure out which part is actually causing the problem +GavinYes it works fine today cool +As a developer I would expect the new behavior as in the patch and it seems unlikely you would know about this functionality unless you looked at the default factories +Namit What is the compilation error? I applied to a fresh checkout of trunk and did not run into any problems +I think thats fine +Thanks KnutCommitted patch to documentation trunk at revision +there any chance that youre using the from multiplethreads? Try the wire logging to reveal what thinks isbeing sent and received rather than just what your packet sniffer gets +Closing the resolved issue. +Ted Fair enough +The patch has the changes proposed to solve this issue +There is no need of new testsIve committed this! +See +Again many thanks +Ive just committed this +I have a patch that allows some races will revisit this after I figure out the EE deadlock. +Patch fixing the more typo fixes for +We also need to update the Chunker sample descriptor +Targeting for release since is released +Version of patch that will go against Thanks for the feedback Owen + + +I have committed this +I used Searchable which is for JDK or lower +FYI you cant put binary files in the patch when binary files are included the only thing svntortoiseSVN do is list the file with its properties in the patch but you still need to provide the binary file separately. +For example its crazy to use this class to back the corereader closed listeners +But there are still other javadoc warnings left +Hi I support this RFE infact I was about to file the sameOne Addition Let the Display the right after a successfull deployment +BarryThis has been fixed for +Heads up sent to the mailing list +In order to run the sample cases you have to add to the classpath +Is this something we need to push to upstream Tomcat +Thanks for patch v +Manually test updated. +Follow formatting style of rest of sections with main section being the first part the necessary changes in r closing the ticket +Thanks Philippe for the detailed report +Basically Im suggesting that the invoker take the bits of metadata related to the child elements of from the acknowledgeMode messageSelector destinationType subscriptionDurability property values for use of the legacy with the syntax +Related to +Id be happy to do the research +This is related to the issue +commit has to be rolled back the applied fix the right behavior is the one reported in the Service Provider documentation and not the proposed one thanks Joerg for in release +a sample project to reproduce the issue just run mvn site +Separate test source folders are fairly common in OSS projects too this is true even in Apache projects +No new tests required +tiny patch +Dominic FritzCould you attach you servers log file to this incident +Committed in revision introduced the class in to make it easier for people to submit new test cases that they could easily run both within and outside the main Jackrabbit test suite +made some minor updates to comments and committed. +Also could you describle the problem your change is supposed to fix? Do we need to use chinese language to see the problemAlso I wonder what other commiters think about fixing widthes in pixels like this +working on it this weekend +patch committed to and branches. +This solves the problem mentioned in +There must be a bug in the leader election code then so Ill check the code +Ill be trickling in thechanges as I can determine that they are safe +I will attach the code that has been created as part of our Tuesday session to this task within the next few hours +There wont be a release +Hi Uwe now are you ok with this being aCan we discuss where to put it maybe contribanalyzers under tr +is already overdue it was scheduled for release last week +same problem with composite you post stack trace +Adjusting priority +The difference between sessionCreated and sessionOpened is that sessionCreated is not filtered by filter chains so that it is always called in the same thread with which the selector thread runs in + +I was able to reproduce and found that were not working with defined? checks thus constmissing was being incorrectly triggered. +minor correction in the doc removed in the command ending with file +Pulling a ready prepared feature branch is a lot more simple for me that fiddling with cherry s keep things simple for the testers +That line should not be executed for the POJO deployment since EJB context does not am looking at the changes in util but it does not seem to affect anything +Perhaps its time to start a thread on dev on any outstanding issues that need to be resolved before we branch +do not buffer results of processing and document processing will be reapeated after estimate of day although the change looks almost trivial the implementation will need to change significantly +thats been failing for a while its not this patch causing it +Thanks Amar! +I am curious why the XML DOM model cannot be used as the representation of complex structure of components with some extensions for functionality specifyc for TapestryIf the reality is that Tapestry framework doesnt allow structuring of components as it stands now then the framework should be recognized as unstructured some kind of system of independent components without parentchild relationship which IMHO from object oriented point of view makes it less pure from the architecture point of viewThanks +Hi Bobthe original reporter was hujn and it is aYou seem to be reporting a different bug that the generated output differs +Xiabo I had missed the information that already ported this to +I agree with you that the upgrade requirementswould necessitate special handling +patch that breaks out common code into that is installed in the same directory as the actualy executable scriptsproblems is that we should probably not install in usrbin etc and if we drop it into say usrsharecassandra we would just have to try the various locations in the tool scripts again the same code we just removedso id probably close this as WONTFIX unless we want to do a real overhaul and perhaps have a wrapper script to all the small toolssomething like sstablelevelreset aa bb +Verified in SOA ER bpelsecureinvoke quickstart is running fine now. +The error was only there because of some broken deployment +The blocksize is KB +Here is patch for cleanup of this file +I havent added the fallback capability discussed but it shouldnt be too hard to addIve also fixing failing tests and reduced the number of deprecation warnings I cant get rid of all of them until the deprecated interfaces are removed +Adds a service to to allow us to get UPS services with just a zip code +Im up for disabling since it fails frequent enough +Please retest I have set an rate now for the micro to maybe that was missing +Thats what fails as key bit is having BUDDYBACKUPADEAD +Wizards that create new JSP CSS JS files should be replaced with the WTP ones due to the fact that WTP wizards provide more functionality and convenience +Could you provide the Maven log file you get when analysing this projects? Please send you result to the user Sonar mailing list +Did we ever get a test case for this? Can this be proven in the lab environment +patch with various small fixes and unit tests +Great to have some tests in the tree +new tests added since this is an annotation and javadoc changeThanks Todd for the reviewI have committed this. +Reopening and rescheduling for looks like this got fixed on trunk it was just originally scheduled for and seems to have gotten shuffled around. +Oracle support when under a transaction uses a plsql block to call execute immediate in an autonomous transaction to create the temp table +I think all available patches has been committed. +Removed test that no longer applies +Left some comments on phabricator +I can make a priority to clean up my pending code and post another review particularly if someone else is chomping to pick up the FATEPlease advise +SDK build successfullycheckintests successminirun components success +The encodes data identical to the existing implementation +for considering something like this +It adds BINARY usage for CS queries Select and EJBQL when ci is trueAlso I figured out that not only may have case insensitive databases and Derby have such dbs too +Add a new method get in +Hi Marek thanks for your testing +I agree with Mete an exponential backoff will make it worseWhat I was thinking about recently is something like a database scheduler which coordinates concurrent writes +Please stay tuned to this issue andor for the release announcement +Eclipse seems to have a bit of a problem processing the patch +This patch removes the boolean parameter which would be always set to true otherwise +Would speed them up a bit +any preference on names? contribanalyzersanalyzers and contribanalyzerssmartcn? once we figure that out i can create updated set of svn moves patch for this +The failed core test is also failed in before test and trunk nightly test +This patch fixes encryption to store a digest computed from the padded password rather than the unpadded password +The time os hour is not configurable + +I added my version of you are right +Its not a bug anyway you just need to be aware if imap vs imaps +will simply use the token from the attempt +This makes it easier for to know when it can operate on and when it cannot +I checked the change log for majorminor fixes but couldnt find any references +The later uses commons http client with a multithreaded connection manager +Okay Ive got some decent logging in place another hard one to parse though +I think the final output is not too hard if the key ranges do not overlap +Please review +sorry stopped reading description after or issues and went straight to the patch +I think I may have fixed it +So for me that would make inclusion problematicFor this looks awesome +to trunk +Mahouts Lucene version has been updated to and the analyzer instantiation code has already been updated. +fixed in and trunk in by r +Hi GiovanniIve recently added support for schemas into connector so Ill take a look on this issue as well +It is also a use case that is better addressed for the page author through an attribute than through tab index +Can you please provide the exact instructions +reintroduced and deprecated it +Ive just encountered this while searching for the type of issues we discussed in the list and decided to fix everything at once +well the idea is that the atom links are what provide the navigation through the rest api +nothing to look at +If you have a large file and a slow link it fails +Should it say schemas here too +note this issue has some relation to dont think its related to the mentioned issue and i think that its better to get the libraries downloaded rather that searching and sucking in the versioning dilemma and i see that the request is duplicated so the users think they need it +the lists have been setup a long time they are just unused. +Upgrading to tycho did not fix this +And now when both oOo and has great support for SVG one would expect all ODF resources to at least not crashIm currently on holiday so Im not much use I cant help you reproduce the problem +Its more of object pool where each key has limited number of equal objectsYes +I will run full regression tests nowTouches the following fileM javadrdaorgapachederbyimpldrda +Use for now as we rant into TCK failures with AMQ +This patch eliminates one of the lint diagnostics +And I dont have to look at the docs to know I can use the result of the query these ways. + thanks +Generally its the persistence providers task to log SQL statements +We are all volunteers +In this case the cluster center is ok as are the other identical rep points +Test it with the latest build +Yes +NET MSSQLLinux Windows Windows XP +There was never a resolution posted as to what needed to be it just something that Im doing wrong? you could join us on the Castor IRC channel over at +I had the same problem with balancer on my dev cluster and this is the solution to fix it +Is this necessary? serial are not meant to be human readable +Thanks KristianCommitted revision . +In what way is the username and password being ignored +All major upgrades of are incompatible deletion or renaming of fields change to tree structure etc +done +Thanks AntonioEric +Attached patch +Full diff for creating the new module +PR for master merged in master +Why does the patch comment out in pom? Should it just be removedI am not familiar with JAXB +Cos Apache Hadoop already does not build on the target platformsSo staying at the current version is not an option either +Code added tests temporarily in the cunningly named tests folderAttempts have been made to keep it binary compatible. +committed + +Im looking at them now. +Assigning to Laura as she is the new lead for EAP docs +Is it impossible to include as +introduced by in rUsing the nametargetNamespace attributes in does not help +I have updated the code for this but I dont know how to make a patch for it separate from as it is dependent on and changes the same files as +I would really appreciate if this could be committed by someone so that I can stop worrying about keeping up with revisions +David you are a genius! I had the same problem and your approach works. +This seems silly but we may need a contributor agreement from you +But in HA world it shouldI think we are discussing this in + +However before making any sales and or discounts you need the default gl accounts mapped +So this is very redundant as you can see +Can you please tell me in which release of will the fix for this bug be includedThank you +The clickable area is cm x cm which is the minimum dimension for clickable components on touch devicesI have no trouble clicking the expansion icon on my Android phone. +The sets the encoding of the request which is for the body and in this case the parameters are from the query string +If anything here is a bug than that unhandled cases are silently accepted +I think it would be better to investigate why proxyProvider is not null +Grant i did read what you wrote Im not jumping to conclusionsbq +I do not seem to find at least in precommit builds failures of this test +thanks Andrew +I meant to put this in as a blocker but I dont appear to have the privileges to change the bug type +So kindly release this in as revision . +I had great problems getting my application to run on Karaf due to dependencies to different versions of Camel +Hi TolaDo you have more info on this? proposed fix for the behaviour +Hi JeanI agree and I am correcting the patch for the numbering +The reason +Thanks for writing the content for the release note field though that helps me determine if it needs to appear or not to set proper fix in Developer StudioVersion Build id Build date +I was working on the latest svn on a linux desktop +Thanks issue is fixed and released as part of release +Fixed Jopr r +rUbuntu used and +Serge thanks for the patch +This is for me a sign that it is probably no call site caching bug +Any chances for reviewing it +I will need to rebase now +The call to persistBlocks writes only the path and blocks +on committing the attached patch +Wed like all developers to do the same so your feedback is highly welcome +Sothe attribute namespace is not common for all the derived classes of i think that theres no point of adding an API to create Namespace declaration for the class +We need to find whatever is eating the memory in the first place +HiThis patch includes improvementsfixes to autoscaler rules implementationPlease review and apply the patchThanks applied with commit revision bfcdfbbebThanks +Unfortunately I am concerned it doesnt mesh with customers requirementsIll catch you on IRC and we can discuss furtherMatt +Will this issue preserve the current behaviour? I think the current verbose format is useful but it would be very useful to be able to list all contents recursively as well as described in the issue description +added but renamed config element to m really anxious for this fix +making sure the context was initialized before any custom init and +Fixed by append this variables to component state + +Thx! We will update patch and the test +Committed removal of test from trunk and +Any opinions +Two small comments The table in the Derby properties topic says that the property has scope S D which I read as or +I was being an idiot +Each issue should be resolved in a separate JIRA +problem was with the URL given in the +The property could be declated on the commandline as well as in my +There were subtasks identified aboveHow many were addressed by the WIP patchOther than all are addressed +Ive updated the documentation to be clearer about howwhen the headers are set +So all issues not directly assigned to an AS release are being closed +The current implementation can be always cast into the MarshallerUnmarshaller interfaces +Currently having a problem uploading new patches +Make platform repository service mandatory and move the noop implmentation for this service to a separate bundle +confirm that the problem is solved by upgrading to httpclient Thanks a lot for the supportThis ticket can be closed +Committed to trunk +A processor is REQUIRED to report an error if anoperation marked Oneway has a return value declares any checked exceptions or has anyINOUT or OUT resolved issues for already released versions +Why wrap in for the token methodsThis ends up used by methods which throw onlybq +As an administrator of a cluster I find it easier to set a time limit for a rpc conection to bail out if it is not receiving response data continuouslyI am not sure if all administrators want this +Example of a chaining interceptor that can use this +I wanted to get workspace all set up first adapter is only piece left and is blocking you +Full and diffed version of the are provided based off the latest CVSAfter integrating this defect should be considered resolved +. +Which tool did you use to generate the patchI would be able to apply it if in unified format which is somehow standard +For example the release notes point to the release note page and points to. +This bug was closed as duplicate which it wasnt really because the one it duplicated was fixed and closed without fixing this one +Hi Sebastianit seems that the implemented feature of using multiple does not have any issues on the performance of the crawlerWith a standard network configuration all traffic has to use the configured default gateway which in this case is the bottleneck for network trafficA configuration for advanced routing tables seems to be no real solutionBest practice for avoiding the network bottleneck seems to be your suggested solution with the bonding interface +Can you confirm that this issue only occurs when you use a model trained by the UIMA Sentence Detector TrainerSo you do the following +Ill submit a patch soon +Should be fixed. +This is a difficult one to address +The patch to fix the bug +Duplicate with jira is fixed in build after verificationso close it. +This rev the context id for +There was also the open question as to whether this is really just an IO encoding issue +batch transition resolvedfixed bugs to closedfixed +Please verify whether the problem is resolved as you expected +Can you make it minutes and see if the issue goes away? AFAICT the only way this NPE can occur is if a node where the session isnt active locally expires the session removing all data for it from JBC and then another replication message arrives without the metadata +also I think we can call it lda in the option list +What do you mean by raid API +the company logo is an issue organisation URL is separate +and committed. +how about posting it there? +Comments and thoughts appreciatedCheersAshwin +Im going crazy +committed to patch broke Linux on all architectures because newly added symbols were not added to export file +Can you elaborate on how youre using the mode and nailgun server under RVMI have shown a couple of clear workarounds +Hi MattCan you remember what version of Chrome this presented inCould be useful for release notesWhat version of EPP did you observe this inCheers no longer reproduce this issue +There is still some more work to be done with this new library +has moved since the patch is uploaded +We shall not use such internal hacks which are even not working for minor revision changes +pending patch removed some more code that is not necessary +Eran and Bohdan seem to be MIAId volunteer to finish this +minutes? Maybe less +Thanks a lot Andrea +Jira messed up the text of the initial description formating I edited the issue to try to avoid that problem +But after updating I now get a classcastexception in get +We have just upgraded to Jira so that may be a factorThanks Mark +I just checked in a change to the CVS source tree that will unwrap the and you will see the real exceptionThe bottom line is if you are see in incompatible class change you need to something +This would be a really cool feature +I noticed there is a good error message when trying toaccess an unanvailable form field +Done +contrib test failure are not related to this patch +enhance goal adds the information from the POM into the annotation +Do you like them now +Linking my issue to is in SVN. +hmm the PR was not merged +I reproduced the entire install procedure and I never got the if this was an error every JBM user would have been reporting this as this is such a basic use case +Happens with Tomcat Tinidad +Committed at subversion revision This patch should fix certain situations in which you see the following You build the codeline from scratch +rev thanks Mike +Looking forward for your findind. +I have tracked this down to the following +line to run the testant updated for current revision +The issue has been fixed in git master but the job uses the tag when creating quickstarts +I just noticed that you already have signed the Contributor Agreement +I added comments up on RB Gregory +This looks related to Kojis patch for alternatively make these files standard for all yarn apps so that the container executor can read info from them +Currently running using memoryAtatched the file counts the number of mapper tasks the JT scheduled at every minute time pointThe first spike corresponding to the first gridmix runThe second spike corresponding to the first gridmix runThe third spike corresponding to the first gridmix runThe tail corresponding to the first gridmix run +See above. +Thanks so far Suresh +Found the one I was thinking of +any chance getting this patch applied to the trunk? do you need a testcase before applying it +The only thing is that in the group is named Copy ofDC but were the real thing +will be marked closed after test is basic test. +Issue was created on suggestion of Andrei Shakirin regarding post on CXF Users multiple WS calls to the same URL as different user from same VM. + Username janb +Reverted trunk rev +What happens to the pid file? How is it named? Has the possibility of and on the same host been considered +thanks +you have mail +Apache HTTP Server version updated from to in the EWS release notes. +active item has inactive icon and vice +Any idea why thats the case? Any possibility to supply a matching version +I think this will help the reader relate to the Login Module easier +The patch includes a new Id class for and the appropriate hooks so that if will be usedI chose to move the SQLDATE constant from to its superclass +Attached patch with ve just pushed the ignored testcases +Resolved as part of +Thanks for the help on thisBTW I am using a custom its just that the code in my handleResponse needs to take into account whether the connection may be kept alive to decide whether to create a new connection to do authentication on +The idea is that the code that generates them will run again on the next deploy +I didnt go through the whole review but I will do that once I hear your thoughts on the comment +I have one comment +Knut points out a portability problem with the Postgres syntax +Sorry but I dont see how this has anything directly to do with genericsNor does this enhancement seem to me to be a +Existing tests are good enough for correctness and sanity coverage and no new tests are required +Im also having this problem is there any workaroundThank to work around I run a patched version of Pig that doesnt raise an error on duplicate imports +On a heavily loaded server it could take many seconds to completely flush everythingIf you want to ensure your data is to disk use full commits +bq +Im going to work on it soon +After some experiment add a pointer to consumer in can solve the problem and past the python test +Im not anticipating any problems though +This bug has been marked as a duplicate of +This will be really nice to have +I have a patch almost done that Ill past soonish +I can understand Java is a requirement for running the application server +I dont know of anyone using etcdefaultcassandra in such a way that this would be a problem +I still dont fully understand remap deletes so its commented out +When do you think you can take a look at it confirm it is an issueThanksPeter +Your plan sounds reasonable to me MahadevI have reviewed this patch some time back and Im happy to review it again if necessary +Closing out issue this relates to the old ruby client which was removed from the repo and not the newer ruby bindings for the C messaging API. +Rich could you please give these changes a reviewAlso the svn diff patch does not provide deleteddirectory information +built seam jee have provided property in and it now works to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Seam GA contens +They test xslprocessinginstruction +i will not veto this patch but im not going to it +The failing unit test on Hudson is a Chukwa test +In my case I send k messages in a transacted session so I would expect to see k message on the remaining node but I dont I see less then k +Is this an issue at job level progress only? Are task progress percentages for all tasks under ? If a particular tasks progress is shown as please provide the info of which phase that task is in when you see progress +Have a new patch that handles deleted docs but realized that returning is not needed +Committed to the trunk as revision After we get a bit of experience with this change in the trunk Ill investigate merging it back to +There was a bug in where I wasnt handling the caches correctly specifically in get and doThis is also smaller and updated for I created new issue and added some remarks to your last message too. +Veronika please deal with this +Keith if you happen to have such an implementation available and that simply needs a bit of adjustment I think that would be great +Flume has a lot of new features compared to which is months old at this point +Several test show errors that are not logged as intentional +Updated the description field as suggested by Mike +Found also seems unlikely because the would in theory use the same config and are able to talk to the NN but this is the most likely culprit +Ive commented out the controllers entries in rev +Retested with revision +already donethis one and seems to be fixed but still existsthanks greetingsResolved by upgrade to +I have reworked this code based on comments received on my earlier submissions and am uploading a patch titled TrackerItemApr +caching added to branch. +With it was not working and thats the version used in the provided setup +Just wondering if it is worth trying to upgrade to version or if we have discovered a different bug +Two issues come to mind + to trunk. +We have a customer who sends us XML queries via email and we respond in XML using email +Screenshot of you have a Fuse runtime its probably the one causing the Fred removed Fuse runtime as well +HiThank you I will try to look at it today if possible and give feedback +Yes I think so +Thanks Simon + +patch looks good +Also I added +AntoineLook forward to working with you +Verified +for the patch +I am going to see what I can do to get a test for this +Removed the possiblity to specify the delimiter +Personally Id prefer to use something like here +When that happens in Commons we just post to the issue with a link to the JIRA issue and either giving them an update or asking them to use the JIRA issue instead. +Latest java specs say that although its not neccessary to initialise a collection it is always preferrable to initialise it +This is mostly fixed by r and r +since theres no response and no votes and the length of time this release has been out there I think we should stick with the behaviour of +My concern is if a fix introduces other issues unintentionally nobody would look at it +related issue +Did not make it into scheduled for this still wanted in +And when I restored the change the test works +I had also updated a few places to use more appropriate apis +I guess we should explicitly mention the option in the reference doc as +If wed get a similar amount of growing pains for Id happily sign for staying with + +They all make sense +When I create new issue which fix version should I set? Max or Denis explain to me please +It obviously does first require an mvn clean install of the jclouds core deps but I think thats an easy one to add to the instructions +So a check is still required on every write if we need to open a new As we later need to know which files belong together to a region I save them using the current task attempt id and a counter to guarantee their uniqueness The current tests all run with my changes which is a good signThe second part is the loading of those files which seems to be more complicated and which could use some comments +Thanks Amareshwari! +Patch including generated code +It would also be useful to see what the public api javadoc would look like with this solution +This is a two step problem having a replace event go out having project notice and make use of can you assign this to jesse when you are sure your replace event is sent +However the wrapper and fault bean objects are included in the application but they arent being picked up for some reason +Max did you mean AltShiftS keys +Roman can you verify that the fix works in the next fuse build? ThanksDhiraj +Workaround implemented in revision is much better now but I guess its still slow for large diffs +This is a very important feature for us too +We then choose to allocate a block to one of these buckets of volumes randomly with a configurable frequency and within one of these two buckets we allocate blocks on a round robin basisThis scheme allows administrators to control both their threshold for what they consider balanced disks and how much theyre willing to impact overall concurrent write throughput to the node vs +Sample file that shows the problem +Changed update delay applied in unit test so that it verifies a change from the default value +If you are setting explicit timestamps you are explicitly telling that it should withhold judgement about versions because you know what happens logically before and after in your system +All tests pass on Windows Linux and Solaris +Thanks! +I will trybut if I want to keep cxf default style and remove the minOccursminOccurs in webmethods causes interop problem with the +Marked as a duplicate of. +Then I will have a look about the dtd fileThe good thing is if you have in future xsds following the same style the generation works out of the box +Im trying to follow the code but it would help to explain the format of the expected and actual strings and how they are generated +Ask me any questions that you need +Inside the Continuum software not cheddar +corrected my spelling errors +Click More Actions menu and select Link to associate this JIRA withYou can temporarily click Cancel Patch and wait for to be integrated +Committed revision +Patch applied in revision thanks for the bug report and patch! +It doesnt appear anymore in the overview of a forum +Need to rebase on top of +the trim is a good suggestion that ill keep in mind for jPDL +Patch file to expose the patch which adds the settings to the code +I trying to use ATS as a reliable and high performance transformation proxy +The latest patch looks fine +Patch applied +Attaching thread dump of process when it is not possible to shutdown it cleanly +The log should contain the user name who issued the refresh request +Thanks Andrus +I think keeping two different interfaces for will lead to confusion in the long run +Uploaded SSL +good thanks. +beta R Standard SP Sun JDK had to use the ERRORLEVEL like Brian mention above on Nov AM. +Im not planning on fixing this any time soon since the code doesnt cause the problem to arise +Attaching patch a running regression tests +Johnathans outcome after his review add missing apache license header to the KMP matcher +Uwe thanks for the reminder +However for this maybe we have to revisit table ownership rights +Closing per previous comment. +Thanks! +Having to pass the command name to user written tags would be unfortunate but Ive stumbled across too many restrictions in the way tag files work to be would actually prefer to be able to set the command name on the form tags even in the jsp file so maybe the formwrapper tag isnt such a bad idea +Have you tried? is no longer the ones lacking patience you can remove the discriminator column and recreate it +Not a problem working as designed +Whatever rule we choose is going to be hidden by the GUI +Please add this to the +I downloaded and ran the scriptpatch +I have one question though whether it is too aggressiveYou do indeed have a point I never thought of before +applied with addiitonal reference to this jira issue with thiswhen an admin creates a user with an empty password when the user first now logs in with the empty password then is forced to change it the Existing Password box cant be empty atm +The screenshot is old +Can you please try to create a small sample project because I cant reproduce this issue with the information given +. +A few mirror update processes were stuck recently which probably explains the outdatedness +What do you mean by save a lot of time? Could you elaborate? +Currently available impl +Can you get us the manifests of the bundles that you are resolving +If Angsa isnt found either is used +This should be solved in verify and reopen if you are still having issues. +This patch sets a very high timeout only for the test and I also decreased the scanner timeout for the other test back to where it was +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I dont think we can use the same ID as the ZK node +Talked this over with Daniel and we modified the patch slightly +from my side +That means it looks at the bytecode +For this is fixed and working +I apologize for the incorrect info +unlock keystore availability successfullystep +create to track +Verified on EAP +Added the other versions to which this issue applies + +Thanks AdrianYou patch is in revision +I just committed this +you dont use the same pwd as you use for the rest of the ASF infrastructure +bulk close of resolved issues. +I guess there was no interest in applying this patch therefore this issues should be closed +I will send an email to the PMC members to see if there are any objections to accepting it. +Sorry it took me so long +Most callers simply care if there were files found or not +Most people will only read a small fraction of the docs if anyI dont care strongly here but I think one configuration flexible to hold most common configs with little overhead would be preferableThat said I moved it out for now +Talked with Tom and am moving this back to is the one who assigned the release Mark didnt request it +It currently handles portlets only +The missing patch +Typos corrected +Here is the patch +Ill commit the fix if the tests pass +have you been able to? This bug seems to have gotten a bit drawn out so perhaps we should file a new bug for the new breakage that resulted after your set of patches +Arvin and I exchanged emails about how the current code strips annotations from blocks and thus may break subsequent passes +This patch performs configuration conversion as discussedI tested it with new old and mixed configurations +Patch with nice indentation +Does anyone have cycles to quickly bootstrap this and see if they can replicate this issue +adds another test addschanges documentation for removes the and does the required changes for the removalRegarding the test I changed it to also add an id to the test table to be able to select specific ClobsRerunning and derbyall no errors so farPatch ready for review +Applied patch with thanks to Johan +missing licenses to all filesPlease review and apply the patchThanks +Robert could you take a look at the commit if youre happy with it could you close this bug +You also cant know the number of keys or values for other reduces without a lot of extra traffic from the +Hopefully with the next synch everything should be fixed. +Im not a big fan of the goto in there either but perhaps thats just me +To run against Derby embedded use java setNullTo run against Derby Client use java setNull clientTo run against JCC use java setNull patch +Such default might be less performant than storing in HTTP session but at least it will be correct +I filed to track that idea +A separate jira should be created if the patch does not apply to older release and requires substantial work + Looks good to me +Thanks Terence. +It could be nice to have a review of this part of the code +I think we now have the foundation in the subsite infrastructyre to get this one implemented as an Administrative feature +I had to also add in some staging directory cleanup to another test to make it also pass +This patch fixes the issue +Thanks to Girish for the is the patch file for updated documentation for API level changes +Do you have a timeline for and when it is possible to get a patched EAP version ofThanks to me +this patch fix the issueProblemWhen axis receiving a message it sets the property to avoid addressing header in the response if request does not have them +Use insteadofDont know which one is better +HiIs it any workaround for this problemMy project in Russian language and izpack installer fails to load if in encoding but I need for Russian words in it + Bug has been marked as a duplicate of this bug +When I try to build the PDF I get an error but the HTML builds +Now that you mention it I think I ran into a similar problem while using the hadoop jar command +I confirm this is fixed in trunk +Can make it happen but would need to wait until at least because it will break backward compatibility for those relying on the public static String +Also the default value is since a default of would imply you were using cosine distance +Yes this has even been done for Spring already +Any suggestions +A patch is attached herewith that upgrades xmlbeans toAll the tests were successful after the change as well +The node is deemed to be unhealthy only if the script successfully finishes and spits output in ERRORPATTERN +KT Ucloud API has one bug +They generally shouldnt change every time though +Thanks Rob for pointing this out +Fixed the author tag +You misunderstood me i know what intrinsics are +Patch up for review on rb +I did some more experimenting and I think this is acceptableThank you for your patience +To run this just go to the child directory and run mvn +Also did some monkey tests on the code and did not find anything strange +Test cases LRE and LRE added but excluded from automation until thisis found to work +I also remove copying the bytes which doesnt make sense here really +the gain I was looking for in code clarity may be surpassed by the slowness of the apisI took some small bits of improvement over toclosing this as not a problem +Looks ok +This seems to be the only way I could find to do this other than implementing this component as a polling consumer let me know though if thatd be preferred +Its untested +Otherwise it creates a new endpoint reference +Looking at the patch for reviewcommit +I just committed this. +I solver this by creating a that I could att a pnael content to +Still have a long way to go +I just tested this fix for the first time and it doesnt resolve the problem +Otherwise you could create an HDFS file as lock +You have to run and then +I guess the questions have been answered +Hi Avery Yes it does should have indicated that in last comment +Ill ask Bill to take a look +It doesnt help that makes this harder than it has to be but at a certain point thats what we get for using a language that was designed to be used in a browser +This may get into M perhaps M +Thanks very much KristiaanCommitted patch to docs trunk at revision Merged to branch at revision +I was not aware that the hibernate jars are not includedOnly JPA for hibernate is in modulesorgjbossasjpahibernate and not the hibernate jars under modulesorgjbosshibernate +For the boolean values we may want to only set a value in the generated XML if the value! the default value +does sound like a bug changes in? files should not trigger a a better solution for would be to do an scm update after a release +This is a duplicate of created by mistake +Its already been used in some other test code and popen is deprecated in current python versions +The propositions made in this ticket are already splitted in several issues +Andre do we know the reasona fix for thisIn any case does not sound like a big enough bug for at this stage so moving to is due to the fact that the sample I include was compiled against an older SDK +Fixed in r +No problem +Ive just link this JIRA to and posted my previous commentBTW to see these changes you may use All bookmark to see all modifications instead of default Comments +The serialization of datetimes must incorporate timezone +It is completely desirable for me to ask if any blocks in my directory are missingI agree that move and delete both should require write permission and move should be to some location that is in the users home dir +fixed in the sense of recovery is attempted +I will take a look by EOD today +Most tests are now using FLEAll tests pass for me +The byte array issue had already been solved under +Heres a patch which addresses the issue +This applies for listerners interceptor and lifecycleUse a separate session connected to the same connection if you need a session to do your validation +It now contains two characters which will when posted again be mapped to four characters then characters etc +Should be good to go passes mvn verify etc +feature was removed recently but the abstract test case still looks to load this feature causing all tests to fail +Safari thenTested with Mac OSX an iPad +Committed revision . +Most people dont need since was introduced and I wouldnt recommend using it unless you have a really good reason not to use step scope +Will you please verify? How on earth did you spot this +Replicated even on new workspace +It is a of company so JPOX doesnt use that ORM metadata for the classes required +patch Feel free to reformat anything I submit to PEP conventions but I need to follow the conventions in place in my organization I think Im mostly compliant minus naming +submitting patch +This only occurs when the alias is named the same as the column +statements +I am using the embedded derby which comes along with the Sonar +I have applied and committed patch thanks Svan! +Verified. +Obviously I meant for each vertex in the above Im saying is that the concept of output format is relative to whatever produces the output in this case Giraph +Is Bernard right here? try the current code base and see if I have fixed the problem if so I will take this fix to the multipolygon encodings +hi as Burr said the available in adminconsole is cut off I am sure that will be do you find any other console needs horizontally scroll bar? Ryan the issue is not whether we should add a horizontal scroll bar the issue is that the narrow display requires scroll bars while at the same time we have a large part of the browser display that is not used at not use the full width of the display instead of making people scroll horizontally? Thx +This patch includes the check for the parent types fieldInfo for the simple content mentioned by Werner above +Update in Requests +The goal of this issue is contained within which has a much better solution for being free of break stop and return via control objects +in fact it was not working for all descriptors excepted on trunk +PR for maintenance +Is it KVM +And there are many sites out there +Resolving this issue as cannot reproduce. + codeline has its own URI implementation which has been the single largest source of issues bugs +Nope +Ultimately I would expect that at the tuples would flow out of a single root node at the topThis visualization is an improvement over what you can do with xml documents in Firefox today The visualization looks cleanerit doesnt have the distracting xml tags mixed in with it Probably this approach will work in a broader range of browsers besides FirefoxI was wondering about how hard a different visualization would be Instead of a structure which looks like a folder tree how about something which looks more like a tree? Thats what Derby query plans are today +Jobs are fine now +These projects will get their server and ESB runtime from your workspace preference +Any reason why this code could not be implemented in the read function +New attributes attachTo and attachTiming were added +flock should be locking exclusively thats the implication from the API +Id like to use it to link an installation to the Wiki +For row wise median which I think the majority of use cases would want to do we will have to precompute it with one more pass +Please post the FO so we can run it through FOP ourselves to debug thisBTW the producer field in the document properties seems to indicate this PDF wasnt produced with FOP at all +Am I on the right track +For the operations of that transaction to be seen by anyone connecting to the previous slave the failover command must be issued on the slave +I thought that the parsing method would be build not build +here and this is really a featureCould anybody please clear up on this point thanks so much +CorrectionOur fix actually moved add to the top of the loop of processRequest method in +Committed revision Thanks Ted! +invalid +refactor the implementation of decision forest +If its read only we should definitely give some error if someone writes to it. +Some more query results will probably change so not submitting yet I will run overnight and update +I am providing a patch for this issue +I think the first version was not nearly paranoid enough +The reasons for equals and hashCode in the interface are there to call out the notes for the two methods +It turns out that the test that causes the seemingly infinite loop actually does generate that much data due to what looks like a bug in glibc +after making a noob issue to add a test verifying deferred a patch with the typo fixed +ACS feature freeze is Nov th +Thanks again for hardening this component +Committed Barteks patch in revision +Patch that tests fixes the three scenarios +Do we normally specify the exact SQL error state in our docs though +Will look into it +Im of course aware of your effort but it is easy to forget when it is not marked explicitly in the jira +Now I think it would make sense to make the default plugins configurable Maybe in some cases you dont want some of the default plugins or you want to easily replace it with your own versionNow the questions is if you want to do this by code or by configuration +I investigated the combination of Zookeeper Solr and againThis is good news +Would you mind taking care of it? +Its possible restrict tests to specific submodules by modifying the list of submodules contained in the macro defined in the file. +Back up now. +This time it ocurred with the targetNamespace attribute in place +Thanks for the patch JonathanWould suggest you file an issue on the protobuf toString issue +New y version of the patch +Does this really belong in the assembly plugin? There is now a NSIS plugin in the Mojo sandbox and I assume plugins could be written to wrap the other tools as well. +Is that now the candidate fix +I think something happened with the encoding of the files +Renaming something like that is usually not an option because of compatibility +Kevin could you comment on this? Looking at it also list instead of for ESB jar making it impossible for both support users and tooling to identify what runtime is actually being used here +Any outstanding connectionconsumerproducer is replayed by the and this is visible with TRACE. +Ive downloaded the version in the SVN repository and tested it but identified two problems +During normal scenarios the logging would be omitted which should have minimal to no impact on the execution time +m not clear on this one +Otherwise it will fail +Fixed by resending the first message in the retransmission task if the connection is not marked as established +They are still in the jBPM CVS +Is there something better than that we should be using instead +Youre right +How do I use it as a client? It seems like I would want access to the DOT file wo having to actually kick off a run +Marker handling has been improved a bit as a side effectFixed now in CVS maintenance branch +Lets make sure they are approximately evenly distributed +just a dump to show where I am trying to goI havent done anything to WhitespaceLetter yet but you can see the changes to +Restore HADOOPCOMMONHOME and environment based control +Anyway I made no change to any database before trying to add that user so everything should be original to the installation +The casting could be one line I can remove in about places I got my idea off of Spring because Spring can types and I find that kind of stuff appealing +rev passes all tests. +As a consequence this would trigger a exception +Thanks again Knut! Committed patch to documentation trunk at revision +Can we close this issue now as fixed in TRUNK +Once we find a good release we stick with it for quite a while +Got the last piece fixed so it works in osgi as well +I have refactored the older class to not to extend from ie +To expand on that a little the replicator could have avoided this bug if it only ever saved checkpoints when it had consumed a full MVCC snapshot of the database and verified that everything transferred correctly +Alexey I dont know what you mean so probably no I didnt install Forge projects +From my experience the same is true for large xml files? being xmlxsd +Changed bug category +Hi denesh ayyaThank you +This bug is in the wicket in the release version +However when I enlist in I get error enlisted previously +Heres a pass at it +server startSqoop home directory homeusergitsqoopdisttarget +Wrong patch +I have attached a patch to that fixes this +So when embedder will provide hook for this we will have to implement artifact resolver in our plugin +Cache flushes will happen when the cache fills or when the region is closed +Im also using the amq +This issue is part of positive failure. +version is managed to be the same as ofWith that modifications all logging from cassandra and all its dependencies will go to slfj API +Committers can you please review +This service will have two require parameters returnId and return +Thanks Rushin! +In the configuration for the new BPM Processor action you have a list for commands it should include and the new was resolved +Will reference this Jira once we get to fixing this in trunk. +Move demosassap to contrib +Verified in trunk brunch +Ill keep trying until it gets through in one piece +We are currently offering a not really supported maven build for a maven driven site and we might update to maven exclusively for site building and reportsBut the official code builds will not be done with maven as long as I can it +removed as a target +I verified this fix using conf test copy and it passed. +workaround one I run cd CATALINAHOMEbin before starting you can see neither of them started and runs just fine so it has to be some change you introduced with +bq +slip to next version +exercises the two methods +You would have to build and install the JBMAR and update the AS POM accordingly though or just hack it into the modules directory and edit the corresponding at your option +For the test failures I created +This is actually misbehavior on the side of there are supposed to accept any values for setValue simply returning it from getValue if they cant do anything about it +Alex you noted that its resolved in the new devcloud image +Perhaps the Seam debug page can add some JMX controls for hitting Logger API +Thanks very much for letting us know +Do we have more details on what exactly consumes that much memory? Is it specific to some queries or is it for any kind of query +Currently on snapshots +In addition I wonder whether you could attach a small XM sample document that I could use to improve the quality iof Castors CTF test suite +Does this class work with Spring ? Thanks +I tested this with the WSDL and the options provided in and the return type of the generated get method is +Thanksthat works is there any way I can change that value on the command line ratherthan editing the properties file +Currently the patch is usable but is more a workaround than a real feature +thanks so much. +Committed to trunk +the issue should be resolved now testing is s great thank you +I think Id prefer just converting that code to use a CQL query or two all it cares about is is this kscf pair valid which isnt a big deal to fixSide note looks like some code duplication in and client be nice to clean that up +I think the textile filter allows raw HTML as well this is required for our front page which is a bit too complicated for textile markupAs Ralph asked is there a way to see if this actually works +No special account setup is needed for Express. +Problem solved! Thank you so much for helping me out with this issueRegards Olivier +So some of the failures may be caused by the class library being confused and using the encoding inconsistentlySeems like you also ran out of perm gen space +This is a fairly major bug and will probably require a new release for Click fairly quicklyregards Malcolm checked in will be available in release +Ideally this would be a +Debug level seems reasonable for that log +Feel free to reopen this if you can provide a consistent service +Proposed changes attached as as revision +Thats specifically the point of the FORCEMANUAL flag +Once the dbo guys will move the code to a public repository Ill let you know +i have the same bug on hibernate using db vr +Log a warning if jython version is not shall solve it by pulling the right +The global sync I referred to is the normal POSIX one that flushes all pending data to disk +Would like to remove from Xalan distribution +I will follow as mentioned by you and I will get back to you +The second patch contains all the comments +optinoal scope instead +I cant find an old enough version of the Thrift compiler to regenerate after updating the IDL so Im just going to pull up to todays Thrift trunk r and commit the result and the removal of as a single change set +Latest patch applied to rest have added the annotations in the interface to integrate REST features in itI would like to get feedback about the REST URI scheme in this patch and in the discussion in the dev list +Tony a geotools patch should be attached to a geotools jira +I created class that includes the class name of the exception that was the cause +The builds correctly with ant +Can you make this patch compatible with branch? it seems like you use old directory structure driversjava +AM test separately passes got the integration test for this running against the real cluster +As we talked about this long before Itd like make only concentrate on converting into and move context partition integrity checks into or implementations +Another example that link area calculation is stillseriously broken in +test case was removed. +When should one approach be used and when should the other approach be used? If the intent is to keep the approaches then your first point makes sense and adding an additional would surely help +I tested this and this worked fine for me without jce jar +Yes +Nicholas I guess we are stale here and it is matter of personal preference +I think passes Strings if a code cant be found +After thinking about it for awhile it does seem to make a lot of sense +Yes I believe this was fixed the proposal only has items remaining for under its tasks. +PR merged +Tests result in a few hours +Assigning back to Gary pending Paul yes please. +I think it would be great to put a note in the LTW setup guide in the manual to warn people about the issue with Tomcat LTW. +Assuming that it is should we do it for or +Yes all Path instances returned from a method should be +Fixed in +See also. +That is on purpose +The JMS one is no longer required this should not be added +We will see more use cases laterh +bulk close of all resolved issues. +If I will be able to fully reproduce it in a test case I will reopen the issue. +Totally forgot to close this + committed to trunk and branch. +Should case sensitivity or insensitivity be dominant +Dave please investigate +xsl +Yes they should +This is fixed and will be release with within the next few days +I dont think the calling method will get you far +Thanks for the updateddescription and clarifications + +Ive not checked it +Could perhaps be nice to have an option to remove the sld as well +Rebased the patch against latest directory layout +The returned list is in the order local replica replicas and replicas +Hadoopuser no problem +Based on my understanding the Cloudera ZK install will put a working in etczookeeperconf +Changes in dont influence on deadlock +If this reduces the of reads required for a worker to find an unclaimed split on the first round of iterations the clumping problem should be solved and the ZK writes that begin to pile up at scale will not slow down the reads so much that so many workers never make their way through the whole list +Ideally this detach area could use some refactoring +Thanks Amareshwari +OK thanks +Patch is applied at +For a sane classes run I got the followingtimes for the loadfinished creating tablesTable done inserting data num rows Table inserting data executed in ms Table done inserting data num rows Table inserting data executed in ms Table done inserting data num rows Table inserting data executed in ms so a total of around minutes to load million million million rows +isnt foolproof +For example instead of imagepng we could use something like imagepngI should take a look at the RFC to see how standard that would appreciated already do aliases in a sensible way +patch looks goodI will commit this later today +Colm can you look into? +Duplicate issue +I will try the snapshot +I think the server should adapt to whatever namespace is used by the client as with other parts of CXFSemi fixed a long time ago by providing some ability to configure the namespaces to use +Checked into bugfixing branch in revision +This is the reason I think its not a suitable solution for managing memory. +Suggested patch applied to and trunk thus covering and trunk series +Patch removes isDirty and compares the epoch from +Heres a patch for +I just checked in the branch and saw that there is no operation there which probably explain why it was not a problem in have no objection to make more tolerant to exceptions while scanning plugins +principal without knowing that theyve already done so for a different service +Before we do so I want to have the discussion in the so everybody is on the same page and to avoid issues now or in the future +Also this has to be implemented for PT transport as well +What server do you need to check? Can you please list them Pete? the mentioned examples have a set of readmes and which are the builds to booking glassfish and ocjjpa glassfish jboss jboss tomcat tomcat weblogic webspherehibernate glassfish jboss jboss tomcat tomcat weblogic websphereversions for the server is in the readmesDone are jpajboss hibernatejboss hibernatetomcat hibernate jpatomcat jpa hibernate cant get transaction handling to work correctly on jpa jpajboss hibernatejboss jpa on websphere except for the fact that aj js scripts arent loaded +Attach the corrected biggest change was to remove references to the production config +Interesting +Not sure if that will cause us problems later +So I believe my solution as outlined above would be the best in this case which allows us to keep the existing behavior +Turns out there was a test that already had the invalid message so the repo is in corewebcoretagsform that useful tag error messages are given when a form tag contains a badincorrect method value. +Could you please provide a patched binary? This is a blocker for me on a major project +Thanks +related but not the same +I think the right thing to do in that case is to just use a cast +meansounds like the classloader is not isolated correctly +Resolved +Can you provide this as a SVN patch or simply attach your s the same as APT +Please reopen if needed +I thin we should keep ntp as the default so change the line toRecommends ntp chrony raring ptpd doesnt provide +Stan the reason for not using a weak reference is that the structure is cleaned and removed on undeploy prior to removal of the class loader hence I dont see how it would cause a memory leakPlus as I said on an to Pete I ran the tests with and without and the amount of time that it took to reproduce the out of memory error was the same which took me to the conclusion that was not contributing to the memory leak +I set the reporter field to Julien since he knows pretty well this part of the code so he cans help +. +batch transition resolvedfixed bugs to closedfixed +I have updated the connector design approaches document and patch against trunk +And I can confirm the bug is gone +By the way I was also wrong about all lists being sets +I to hate this problem on the iDevices however its not limited to Rave +I have put NEW comments to indicate the places where the change needs to be made +are now agnostic of Transformers NobleThis patch has the following changes on top of Nobles patch Added close method which will be called at the end of import +I guess it would need to go into cql +and committed to trunk. +Is this ready for a commit to trunk? Unfortunately no Allen +Steps to reproduce? Looks like maybe trying to create the same network twice +It will have to be synchronized +IMHO it is ready to be committed +bq +Extended the JCR event model with a custom NODESEQUENCED event +What if the site is suppose to work without +can you please test against trunk and verify that this was the same fixed resolving this ticket closing +Starting from version ajinclude component is updated on every ajax request if ajaxRendered not set explicitly to a workaround add ajaxRenderedfalse to ajinclude +Otherwise if you go back and forth between options tab and other tabs you will get more than one internal notes and shipping notes at last +Committed revision Thanks Cdrik! +I think you can go as you want +steve that code looks fine +Sounds like a good starting point +For this reason I think we should be looking at a search request control which makes the search request return immediately with a search response control showing the number of entries that would be returned for the search +The javadoc for close has some strange formattingindentation +When choose xml javaxsdwsdl input type the input configuration panel will show Specify sample data in the Input Data sectionWhen choose no input it will say Input type should be specified +Thanks +IMO its no use in restricting ADMIN from READWRITEThough ADMIN can allow someone to grant themselves READ andor WRITE permission ADMIN should be a bit that allows DDL and table management ops nothing to do with READ or WRITE permission +on close +forgot to set fix fix version +Attached patch in the event we want to apply to the. +I absolutely agree the most important thing is that Spring explicitly throws an exception when the expected behaviour cant be assured +It might not be worth the extra encodingdecoding time though +cpp shouldI dont follow this rule +Instead of latch it now takes the lock before making the call so the notify cant be called until wait releases the lock +I will research and see if I can duplicate this bug +Howard +switched to in +This is almost certainly a classloading issue with +jar files +What do you mean exactly by there is no concept of ngroups or group facets? Does that include that there will be no possibility to return the number of groups like the request parameter currently doesWill it still be possible to decide if the faceting is done beforeafter collapsing similar to +tests seem ok +Mark as patch available +This is not new we have the same type of behaviour when we switch languages. +From ircjohano i know we have that file in some projects at work will see if i can figure out what depends on itjohano seems its used by emailing etc and a friend says its now part of the jdk anyway +Got it working today and tried to update but it still complains +It needs a more sophisticated protocol stopping threads during the creation of an instance +xsd because the schema folder is at the same level as the that import the schemasNow at least the editor recognizes the message types but Im getting a strange partnerlink mismatch in the receive activity +v has assert rather than throwing UAE +change committed at r in has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Thanks RickSome notes version should be version Rev Date I didnt think +runs all the tests in and reports how many milliseconds each fixture took +This patch synchronizes access to the while sending out the +the stddev varies on read from +Proposed patch version Is there something obvious that Im missing? Any other feedback or comments on the detailsId like to nail down the exact wording of the proposed policy change before I call the vote on general +Would you like to have the wiki editor turned on for textbox fields +Try again with the code I committed +While looking at this area of the code I was wondering about this patch wouldnt it permit the mirror issue of enqueue and not advancing the offset since the interrupt could occur just before the fetch offset update? So the new fetcher may fetch the same offset again +This is seen particularly for those nodemanagers which had a few containers in execution +Introduced a that examines the contents of the process archive and invokes either the jPDL parser chain or the BPEL parser chain as appropriate +Ill try to come up with a test case by today +and Qdox is allowed to push the bar beyond what Sun would dictate for Java +Does this work because you have moved the splitting after other DFS ops so it will stall there until out of safe mode? If so thats right +Any alternative that does not have this problem but solves the would also be great +Patch includes new tests for and to test correctness over daylight savings time changes +new patch looks good to me! makes sense to execute the aggressive only if rest is on! in +Logged In YES useridfixed in and HEAD +The same Exception is thrown when using the get option in readseg +I think we should check if there is already a before setting it +from me after those other changes then +patch has been applied by Nandika +I will create a patch against contribs svn trunk. +When a user resets password from the login page using Reset Password an email is sent to the customer with the new +Let me know if theres anything I can do to get this out sooner +I see that this patch was already committed to +For some reason it doesnt work to build with jonas or jboss +Plus we already have metrics integration for lots of things that havent been yetSo Id prefer to keep exposing data via metrics and then use a JMX context to also publish to JMX +bq +Thanks Craig +A warning will be printedIm not sure about this one +The current revision that I see is +Any concerns on this issue +Committed to trunk +In this case waiting will happen at the right time +Whoops I wasnt very is the message the compiler is generatingBCE Cannot convert to +above the scratch version was fast so it would be worthwhile looking at I think +bq +The original project was too big so youll have to copy all of the Axis jar files into the lib folder and then add them to the project after you import it +bulk move of all unresolved issues from to +If this patch doesnt go into SVN perhaps the wiki should be updated to show it +In we had a separate java class which did the compression and did its own exception throwing in case of errors +Here is the plugin revised according to the comments made by Doug Cutting +Closing all ResolvedFixed issues lacking a Fixed Version +It would be good if our system can have ability expressing data layout instead of how data is accessed by tasks so that the system will be more flexible than restricted access modes definedCan you give an example please +bulk defer of unresolved bugs to +The getProperty interface method returns an Object to ensure that if in the future more complex additions are required they can all be added via property keys without necessitating future api changes +Sorry Srikanth for the confusion +On the other hand you still depend on the API I guess and the other two dont make the methods visible like requested here +What remains for this issue is to complete the calculations from the representative points +Im opening a new ticket for the test. +I think what youre asking for is some way to configure Enunciate to say generate multiple and then somehow pass configuration describing which resources you want included in which WADL +Replies create virtually no extra load and it makes the clients much happier +you still arent ever removing any stopwords but using this solely to speed up phrase queries by forming bigrams around the common termsIsnt for that case +The most common cause of this is a PDF that uses subset fonts +Putting this off until later as the performance hit is pretty substantial and I think there might be a solution that gives us both performance and correctness +Full stacktrace Please use the forum for support Please do not set the fix version this is for project team members only +I do not have the right to close it! +As far I can see I use version +Tested. +they were purely for testing the quickstarts for the last Community GA +Patch committed to trunk +If you find further issues please create a new jira and link it to this one. +If you only but not sync it with then it is not obvious what to do now that have to be supported what if the earlier interceptor does only update but not? What I mean is that can try and sync ACCEPTCONTENTTYPE itself to make sure sees the correct values but it can not guarantee it will do the sync direction correctlyIf you do the header updates before is used and not using filters then please also update a bit more work yes but it will be the correct way to do itDo you agreethanks +Im not sure whether replacing the interface with directly using Tika is a good idea +Thanks Jeremy +Read load wont affect its ability to complete however. +in SVN +Yeah I think we may need to update something on the spec here +I will upload needed jars to your original exogtn case +Thank you very muchIris IrisCould you please create your patch against the latest trunk code its more easy for me to review and apply the patch? Also you can use something likesvn diff to generate the patch its useful when your patch include multiple files and you can only append one patch file hereThanksFreeman +Ari this is not compatible with existing checkpoint file +from me +Gavin it was successful +I didnt do anything specialIts only like in that its triggered by using the +new patch rebased on still has a conflict +Initial patch this bug as wont fix this was a patch uploaded to put up code for growl a tool we developed to import data on to hcat around the timeframe and our code has significantly evolved since then +Write some unit testsThanks the last one of the test failures are not related to my patchfixed one assertion failure due to unit setting +We can then start implementing the reports mentioned above +The patch is applied together with the reproducing issue patch to the latest branch and it is working correctly now +patch removed unneeded getuser overload +for having this for versionIn fact I was trying to make a port and test and as per that raise a new issue for porting +Do you mean the contractThat is working fine for me when I added it to the +after release +The attached patch fixes not only the incorrect caption but also does some cleanups on the code and adds some documentation to it +Verified in JBDS Build id +That makes one think +Ill make changes lads +in a similar fashion to what does with? Is it to avoid doing too much work upon reopen to compute things like byte widths for numericsActually it hadnt occurred to me to wrap like we did for I guess because a DVF is so much easier to write than a PF +Fixed in metamodel branch +A workaround has been committed for now but this needs to be reviewed so that a more appropriate approach is found +It is worth having a look into the rules defined by Axis C project in has been fixed to a great extent during the integration effort. +I could only find examples in ui with editor textbutton using label Edit +These days people are reporting it as a bug +Hello I wonder what the situation with this issue is currently +Fixed in the trunkYou were missing which was causing procrun to crash +Inside abdera they are creating a class which extends and use this internalSerialize method in itRegards Fixed. +Should we branch? Do we justtake the plunge and say that any new versions have to be JDK compliant? Doesthe trunk continue on a path and we just creaet branches thatsuppor the legacy +We will not fix this. +silly jira giving me the wrong bugs to edit +Thanks. +err correction times before it goes to the default then once more before it goes to the specified +I just committed the patch for this to trunk +Applied on master and branches +Yes this should probably look like rsync +Migrating to a different release would require lots of additional testing to avoid breaking stuff due to subtle break +ah I had generated from what had been committed so thats why it was there +Do you have particular views on concrete changes to this end + +Should make some scripts a lot less verbose +If your tests show that this works in distributed mode then its probably good to go +Would stored procedures be implemented in a TSQL like language or java code? In the near or mid term is this viable +Attached patch to fix the problems +I allow myself to rename the title. +Ran tests and examples against Hibernate +Attachment has been added with description The generated has been added with description The PDF obtained without a soft is a fully version of the custom generatoreditor that is used by our customer +The patched build process will determine which common buildfile to use before import check whether should be executedThis is done using the presets found in When creating distribution the will be copied to distdir the will be copied to distdir the in distdir is edited to standalone setupPlease review apply inWe decided to just keep a copy of in the CVS module itself so that should take care of everything +Vinod thanks a million for attaching the patch! Im building it right now in Bigtop and hope to run validation tests tomorrow! Will keep you posted on a JIRA +The itests and qtest target require the profiles as is desirable +Correct you can override this by setting the table replication for theMETADATA table after you init +ok now this is moved to patch for the way if no one objects I plan to commit this shortly so that we can improve things directly by patching the trunk +These can be invoked independently +But at a minimum well not accept an implementation that has not been documented and discussed on dev. +Can you please check if it works okay now? Can you also please give an example of what exactly happens without this patch? We can then add a test case +The above fix takes care of all session bean types +removed LICENSE NOTICE and files from contrib copied text from file to file at ambari project root +If enabled cluster mode ts crash +Please consider only the patch +Triplets are a maximum of bytes therefore Attribute Value Triplets can legally hold a maximum of bytes including the header +I think it will confuse users a lot if +duplicate of +I might require cleaning Java cache from the one comment we are considering to remove the dependency on on before gets into a final version. +The methods of instance in network driver use the instance to send all the DRDA commands throught the socket +Marking as Wont fix for nowFeel free to reopen Himanshu if you think you get to this. +Thanks Nicholas I forgot was not in +Changed target to etc + issue patch ready for inclusion before branch creation +It works when I use Resource instead of Autowire +See published patent application +Thanks guys +All arguments beyond target IP address and port are rolled in to the struct +Can I also ask Nicks question +There is a grain of truth in your proposal at least with respect to the java and org prefixes as you do use the Java standard libraries and the Apache Commons math library in Native annotations and thus these prefixes may leak into the generated code +If it is really such a peculiar requirement I guess it would be best if I followed your advice and created a custom implementation of the cache classes +Clean up OS detection to fix Mac OSX test runs No more bash c +I suggest that we add another section next to Tomcat one instead of replacing the Tomcat section +I would expect there to be some kind of log warning if this took place however within the last minutes of the crawl before the bad state was seen +This patch has been tested at user sites and is believed stable +Thanks ChiragChanges are in trunk at revision Ashish Vijaywargiya +Thanks for the patch Laura +Since no one is stepping forward to fix and failure of this unit test prevents contrib from running in test builds I am making this temporary disablement permanent until is fixed. +This is a problem because the next time the model is generated these changes will be wiped +Inheritance and provided classes have unresolved references and I cant reproduce the issue using different entities +Fixed hang +Hope you will find a solution at the machine level +Hi Myrna +Patch with this nit and similar addressed +if the dependencies are not the same you might have different jars at runtime than at compile way Im managing the dependencies right now is to use the maven assemblyassembly goal with the descriptorId set to This creates a single jar that contains all dependencies combined into one jar +Documentation patches welcome +Ok Lars +bulk defer of unresolved bugs to +Thanks for a hint +Thank you for your contributionIve applied your patch into SVN trunk. +Why remove With federation there was only one suffix possible to key +Reason ForbiddenPowered by Jetty +Closed after release +Good! +int he process of fleshing out unit tests to confirm +just found chapter idea is definitely worth thinking about I just wanted to make sure weve got the basics understood and tested as the scenario seemed a bit more complex than the average use case +Output from jstack for and test number looks like a spurious error since the configuration matrix for is showing all step where you run the client sets the number of coordinators to createrollback +jar +I removed an old reference to legacy entities committed to SVN revision +Applied thanks for the patch! +One thing is the change will make config incompatible +Reopen if still a problem etc. +Already code presents under namesystem lock +Will write it up +However as Keith initially argued and as PR experience has shown the decision state is an extremely elegant and usefull construct so in this case it pays off even though its a bit of an exception +There wont be a release +In and nodes config DB is bottleneck of the demo application +So I think well have to go without a test for this one +Unless you are running into random jar related issues like staleness +Ready to go +Provided patch for trunk as per the above comments +You have the flexibility to calculate median atleast on a single server +If you see the error please reopen it with WSDL and client code +Have you run all the tests? There is a test which should diff after your changes Also add describe and describe extended for the new function that you have patch rebases a testcase now all tests pass +I was off a bit even a non graceful shutdown should not cause this if you are not indexing when you shutdown at worst nodes should sync not replicateIn my testing I could easily replicate this though replication recoveries when it should be a syncYonik recently committed a fix to this on trunk +The RAM reader implements all methods except norms +RightOr you can just test this manually +Just checked OS X and these versions seem to be installed alreadyIm not sure I understand the changes in the patch +toolTip text in the box comes out of the box +Ill upgrade my libs to include the fix for and start poking around tomorrow I started looking at it today while you were very kindly looking at but other things got in the way +Data nodes should be banned from reporting the same block copies multiple timesif they are intentionally or unintentionally started to serve the same data storageThat is why data nodes need to register and need to keep a persistent storageID +Thanks Bryan! Yes these are good improvement suggestionsAs for the whitespace changes around line I am aware of thoseI was vacillating back and forth on that in order to make the diff look at readable as possible but in the end I also prefer to keeps the diffs minimal so Ill reverse that change +Committed latest x patch to branchx +And it would be nice if you also merge the regression tests +Please review the changes and let me know your feedback +Thanks +Access url of httplocalhostdsAccess url of annotation httplocalhostds is supposed to only suport servlet +LucasNope + patch add support for update request field typeStreaming is only implelemented in XML Codec +Fixed the reaper thread in revision fix is in the for JTA and JTS so that we are certain we do not inherit anything else from the wrong context +Skin aliasing for backwards compatability +Patch file + is a test case that can be deployed in Fuse ESB has been merged to fuse camel branches +The XML based version is an alternative file format +This file contains Type fonts and displays as blank in to improve provessing of +Hoss thanks for the modifications +I hear you and my apologies for the same +Alan It wont affect Zebra and +I added the case number from salesforce +Checked the Maven depends yes it is differentWe are using the and +Great to have and it completes the schema picture +I missed before +One more cumulative patch +Issue resolved by applying the patch. +IT implemented the redirect +JimThe agents module got left out of the patch sorry about that +I should probably run a once it goes in +Lucene +Ive attached a fix for this problem checks to see if the exit code is and if so does not report the service stop to the windows service controller at all +unzip then with mvn compile jettyrun you can start the webservice on httplocalhostcxfbugservicesWSDL at httplocalhostcxfbugservicesdummy? you for your reply + +Hi RobertYes I can connect to Sonar server via https from Eclipse Sonar Plugin but there is a trickThis Sonar server too old version so button test connection doesnt work but I can ignore this button and browse content from server for example via Sonar Server Navigator +Is someone else planning on improving the implementation I provided so that itaddresses the concerns outlined by Dmitri? I am willing to help if requiredRegarding Dmitris comments +So the simple thing to do here is to add an filter query but that seems wrong since why even go through all that work when we know the exact results that should be returned? Of course that assumes that all the docs are actually in the index +This is an issue that we will track in the EAP PM team as we need to provide GSS with a supportablebuildable product +As long as thats the case we have no choice here. +Just tried it with CR and Im still seeing the NPE +This seems to be at least partially a duplicate of you clarify what functionality you are desiring about and beyond that in and provide an example of what you expect to put in the pom and the resultant entries in the spec file +This points the Netbeans project at the new location of the Felix classes + is the correct version of the jar available on the Ganymede site +Which store do you mean? And what updates in particularThe Derby store should already be updated +Dont confuse Jenkins logging with Maven logging! Maven cant know that your password was passed by Jenkins and even as masked password for Maven it is just another pair provided by someone or something +This may be related to how checks for the completion of modules by scanning the log file and could go away once the output will be. +It cant be booted using any of the three involved boot passwords after were done +Why does need to be accessed from a Configuration? should be dependent on not +In the meantime Id appreciate it if youpointed me at the source tree for your buildsPlease go ahead and make the site liveColm. +Did you check this scenario is valid before submitting this jira issueI am not sure it comes from Portal +Committed with revision +have written my paper but tried to fix this bugAttached a diffWorks for me maybe not the best solution maybe the fix which is right your turn the supplied patch that will be available in the nightly build +We should probably document how to deal with spaces this is a common question. +Katheys suggestion of simply removing the sentence that was only true for GENERATED ALWAYS columns seems quite reasonable to meAttached is a proposed patch to accomplish this and the resulting HTML page +I just committed this +throw +Please discuss such questions first on the mailing lists +Added a program that reproduces the issueTo answer the previous question what Id like to be able to do is return the error to the user and when the next request comes in be able to still use the connection +has been marked as a duplicate of this bug +Hi SureshI Just noticed this issue was fixed sorry for response later +Commit at subversion revision + This bug has been marked as a duplicate of +I was testing with a patch that had only one tab so I didnt noticeSomething like checkstyle probably makes sense at some point but right now we have so many violations that it would be a big change +There are plenty of examples in the project +Xavier could you explain me how to generate this file +to trunk and +Good find John +Updated patch with missing files +Thank you! +Really closing it after long time +I had the same problem +It worked when I switched to sane jars +But that is beyond the business of this very JSR +If its gone would you please resolve otherwise fix +I will check supporting it is possibleFurthermore according to Javas naming convention Foo is a static nested class in Foo which is a nested class in +This just looks like a difference between INSTALL in trunk versus the one in the branch +Nope it just errors in this case now +I do another test where instead of use and components i used only components with only threads for pool and i see that requests are enqueue correctly and handled yet if there are few threads availableWhile with this doesnt work +It worked +Only line has works +It may be fixed in CVS but can you confirm this bug in the released versionI dont want to pull CVS from the head and use it in a production project if you can issue a patch or are implementing a patch release that would be appreciated +Previous patch has some undesirable +And whil ewe wait may we please have highres so that the versions are a little more readable +basically because we dont include rubygems into jruby jar file its included into though +We are using annotations not files +Hi Peter I was wondering if I could get an email address for you as I need to discuss a few things about the mail adapter +implemented +Moderator changed +Closed upon release of Hadoop +Builds and passes all tests locally +for the bug as originally defined +We do some configuration of the maven settings internally in Forge as well +I created a synthetic test which creates empty threads which just reproduce themselves and deep recursion threads which eat up to Kb of stack in native code +for your remaining bugs to. +Im just adverse to unnecessaryegregious use of endls and secretly wished that I could get away with it +Nobody has spoken up in its favour +lgtm +Committed revision Thanks for the review Stack +Oliver this has been done for awhile I I commented but I dont think you are set up to get notifications +If you created unit and integration tests for the same domain class I think they would have the same exact fully qualified a bug +To fix this we need to restart the node completelyFor example if the masters connection to throws it loses its ephemeral address node in and everyone will think the master has died +a test in annotations is enough +It will help for those developers who clone the git repo and git add +According to Maxim who has discussed this point with Stu at the summit this is still fine +Ill try to think of somethingI personally think the name is OK +I added a test for the new option +If it is a limitation of current FOP? Thanks for any advise about itJackie from Taiwan +Defer to all open issues to need to have explicit for each optimizationanalysis we intended to forward port +While rebooting JIRA daily solves the short term issue of JIRA it ends any chance of fixing the problem + log with multiple client deregistration you see if your still having the same issues with? The queue browsing has been improved tremendously since this is all working now in let us know if you can still reproduce on and we can same problem ini meet it +The build with this commit was successful so Im marking this as resolved +Also according to me experience the merge is required in order to get the update to work when there are file changes in starteam server +This test depends on a specific implementation off how a repository handles node type constraints +Anyway I dont see any link to a managed bean class in VPE +Thanks +Done in trunk +Instead the Search box just has Resource Name in it +The attached patch just removes the line which sets create to alwaysThe default value for create is never +Sonar log file shows that it cannot initialize the data directory of the embedded Derby that the cause is another missing folder which was not extracted by crappy it is +Since the seam dispatcher needs the web application to be available for its lifecycle to work the timer event fails +Havent had time to check the validator appthough +In fact I felt it for the past couple of years +Please update the changelogNevermind +Attaching an file that can be included in the API docs +settings and project +I can commit the change for trunk but notEdit it gets even better +Sure I can look into it +The remote repository with the parent pom +Discussed with Nick and probably we could place the schema to and make redirect from in order to have pretty URL +I wonder does this test also fail with Advanced +What I can try to do is to support the subsequent retrieval of the referenced resource files same as it is already done for the referenced meantime try to use a single WADL with multiple roots this should give you the best experience as far as the user seeing the complete WADL and you seeing multiple resource root classes generated is thanks! It is what I am doing right now and it works well but the file starts to be big +Close out issues. +Thanks for the review Vladimir Ive modified the patch to avoid the creation of the object in each call and I also added more tests to check the behaviour +Assuming the new datetime library comes out with Java and that Lang is dependent well revisit all JIRA issues then. +Its only beans at the conversation context that then make things ugly +A soap endpoint in CXF can handle both SOAP a lot for your reply Daniel and FreemanI have a question for thisIf we send a soap message to a soap only endpoint means the endpoint only has soap binding info +Would be really nice to have a relatively fast smoke test group vs +Net bindings for the C messaging API. +Also fixed in mostly parallel sister class issue also applies to +This has been implemented as part of SWS +Running the tests showed no trouble with the latest patch +Any code contribution are welcome. +If find anything wrong with the latest changes just go ahead and reopen the issue +Thanks Jason I just committed a fix for this that changed it to milliseconds. +Moving for now +Perhaps this is the start of a rules perspective +Whats the status on this? Has there been a component upgrade? Should I backport +Differences from trunk patch +IN TXT vDMARC preject ruamailtod rufmailtod +Dont know for sure if this is the right fix but one way of solving this is by maintaining a sorted order of queue just like waiting jobs queue +Hi Carlo Im not sure who on the EJB team handles documentation but I dont really touch the EJB docs at the moment +James see and +WIP patch incorporating the comments +did not introduce the problem mentioned in this issue +bq +used to be a lot more now its just there for deferred log flushing +Attachment has been added with description patch + Instead of treating as a noop it would be better to throw an with the as its cause +For maybe we just need to replace the whole +Attaching screenshot with broken layout on richdataTable filtering controls in table screenshot with broken layout on richextended filtering controlls in table header +But after editing all future ones shoudl work right +Its the very first collapsiblePanel on collapsiblePanel page of the example warPlease see discussion thread on users with Subject Tomahawk on also have the same problem running on Glassfish V B using the Tomahawk nightly build of the examples from +This is great it will be really helpful to be able to properly run all of the tests against Hadoop thanksCan you also create a JIRA if one doesnt already exist to switch to a released version of Hadoop when its available +Applied for hibernate and forge +replication failing with hadoop and distcp +Is currently being used to hold styles for vector features in so it is to be kept +Is this issue related to? I fixed on bramley +Reopened to set the Fix Versions field to JDO beta. +Here it is +Hi Jacques if you want I can provide two different patches to solve the bug also into the releaseBy the way I have seen that I have provide a wrong patch with no relative path to the ofbiz directory and so I will upload a new patch to solve this problemThanks a lotYes of course patches would be Jacquesthis is the patch for in improvement ofThis patch can be applied to trunk only and not to the release because its not a bug fixThanksremember to close the issue and if this issue will be closed because they are duplicate of the same issueI have move the improvements on into the new issue +Do you have any suggestions as to how I could get tests? I had enough surprises when I ran the tests in English that Im reluctant to just plow ahead +actually each executeQuery call creates a new revision as that method calls the get method Ill update the title to reflect this +war file and deployed on tomcat +The only option to run the test project with EAP seems to be to use JDK and add true to the JVM args while running testng suiteWould id be possible to not to enforce JDK for the test project and create special testng launch configuration which adds this parameter? This way we would get experience with EAP Thanks Alexey I am fine with this solution +yes all the more reason for us to change the seam wizards to behave more nicely +RobIn order for this to have the correct visibility I would think creation of subtask jiras are in order contained under this parent +Ivan asked me references enough to make a blog post and I guess docs too +Then we need to initialize rpc server in RS with the hostname recieved from master after checkin right? Otherwise we will have this issueThe regionserver just takes the name and uses it in subsequent communication w the master it does not change where it is bound based of the name the master gave itAre you suggesting that regionserver only set up an rpcserver after it has gotten name from master? What if this disagrees w what the operator told us use in the configurationIsnt what we have here a setup problem we have regionserver on localhost and master on an ip? Can you have regionserver bind to same ip +What platform does this occur on +Got the pluggins to compiletest with both maven and maven but when you exend your buildlifecycle and use the pluggins plexus blows up with java language exceptions +I just committed this thanks Sourygna. +Revision . +Please let me know how I can see itThank youBest regardsJames +Id like to see either the name changed to be more specific to its use case or keep the name and have the templates restored +with JBT. +Thus the XMLSCHEMA fix really isnt neededI also added a quick system test to make sure the import is generated from a real service +Confirmed that it fails on Ubuntu Server bit with Erlang RB as well +Changed Target Version to upon release of +would be good to get this in on the front end of the next iteration +Closing +As Ive never seen enums used in this way in a Grails app this way before would it be possible to include an example of one of your Domain classes that uses enums like this +Youre right my bad +addHints is called from the constructor at this point in time delegate is null +The point I am making is that should work for all applications not just +Thanks for merging +will do it +pushed to master pushed to in JBDS B +Shouldnt we be honoring even if it happens to bePerhaps this should be fixed by removing the +First cut at this seems to work as desired +Thanks Jukka! I didnt consider the possibility of such a limitation and so didnt run the Jackrabbit test +I suppose the biggest thing we need out of this bug is a better error report if we cant find an appropriate JAVAHOME +When was this fixed? +therefore idecided to move that duplicated code into one single location without major refactoring but just added +Is this issue now fully subsumed by +Committed revision +Sometimes the +I think we are mixing unit tests and functional tests here +So Id suggest implementing a different that implements this +But then the Guest and were created successfully without any issue +I agree with your preferences +Thanks for asking for the review too it is good work and I am glad we could get it in +Sorry I should have been clearer in my initial description +Youve definitely found a bug +Added the test to trunk srctestjavaorgapacheactivemqbugsrevision It works as expected all messages are received +Implementation notes and are both hard coded to use artifact name for packaging +not removing parent region unless the daughter region existed +Tests pass locally for me +therefore naturally fixes this ticket +This problem is easy to rectify if an error message is produced however linuxprochandler a small C program used to trap KILL signals for java tasks started up by was masking the error +If anybody else isreviewing it please let us know +Im currentlyusing WindoooooofIm definitively encourage some way to point that out both options have their advantages and disadvantages +I can add more choices perhaps. +Im closing this as wont fix because it didnt generate much interest on the mailing list and a clean implementation would seem to require major changes to et al. +After the project split this problem needs two separated patches one for Common and the other one for HDFS +Also let me know if anyone has a better name than NULLABLEThanks +Fix along with I am badly missing the auto generation of transient fields in scaffoldingWould it be a problem to get it back maybe by respecting a displaytrue for the property in the constraints block +Wrap the authentication result and identity attributes into the common token +ERROR Not URL above works for me + built wo maven rebuild when maven is +Unfortunately my brain cells start firing low after am EST +click it and it shows up in the browser +This is fixed for +The issue of sorting the output of the various compressed archives can be handled separately +AzeezThanks for the patch +Since there is still a lot that needs to be done for the release I am just going to update thirdparty with a tagged build from remoting +Let me fix +I have no other way I can change one of the other sql server JDBC driver JTDS so that you can set the socket timeout +the fact that no one has been able to confirm whether or not these properties work as describedHasnt Dag been describing how these properties do work +iml +There are a few jiras about exluding some projects from using project references +How should we go about suppressing SUN jars etc? Also isnt that a larger problem of wanting to suppress them from both m and m repositories not just the m repoI can start on a blacklist component of some sort to suppress conversion but it still will not handle the m case +This patches the method that writes to a file which as far as I can tell isnt used at all +Done closing. +The patch resolves that issue +Anh Vu No i changed the fix version because you reopened it so I guess that is because the bug still exists and thus need to be for another JCR version + +Currently that exception is only used by clone and importTable in which they are inconsistently wrapped in and respectively +There are a couple of things here +Perhaps there is some way to builda custom version of the classes in that jar such that they dump a stack trace whenthey are accessed? That way we could see what code is accessing those methodsand figure out who is calling it +I will add some detail to the javadocs +User code does partitioning and so Tez cannot understand or make changes to it +FFI has been renamed to FFI to disambiguate it from standard in commit eecdebdfceffeb +I couldnt find where the problem was though +Once i reset my properties everything appears to work fine in and above +Im rather +Just the WSDL +This should be back to working and the example is not contained in the contribclustering plus I the downloads directory +Is this ok this way now +That way we dont have to anticipate all these cases in the plexus definitely agree about making all this testable +Once has been committed this issue becomes easy to fix +We might even just want to change the docs to match the code +Hi JacquesI did set Patch Available! You can see it in the history from Tuesday +I could not figure out how to do it with hints. +Minor upgrades in backwards compatibility layer and javadocsWill commit this now +Ill have to check but I thought we were locking +Will the Component and other annotations with classpath scanning be supported +but the rule thats blocking it should have nothing to do with http traffic +What was the problem? I have the same issue? After adding dialect to my file I get Dialect class not found class checked the class is in the supplied by Grails and I get teh same error if I add my own to the lib +Hi Alan what should be the new way? Thanks +Might release without this +does it still work with revenue as expected with the latest +I think this was when it was sleeping +Sorry about that +Is this issue solved? Since the filling of the memory is the always case in Sort like jobs is it preferable not to close the input stream and maintain it also to avoid the unnecessary reopen? This will not lead to job failure as reopen and will occur but the hosts will be regarded as slow hosts and put into the penaltyBox thus slowing down the whole progress +I think it will warrant its own page +Yep well spotted its really a nightmare to implement this kind of checks the AST is slightly different for every possible syntax +The next key classes to review would be the Mappers and Reducers the rest is detail which we can easily collaborate on to get rightIts OK to jump straight to the implementation if you feel confident about the structure of the computation but it could well also be useful to review your detailed design just what are the what are inputs and outputs +The attached patch is improved over the last one in the following ways Removes duplication of code by wiring in on the other two phase boot code in Shuts down the database after the first boot if authentication fails which means you will be able to retry startSlave with other credentialsAnd regarding two file I have no idea why there are two of them +Also affects +The design of key distribution and key management for support crypto codec +its just a little changeIs this already been fixed in some version +But what can I do get this working +Thawan please regenerate the patch use with git otherwise our automated testing isnt able to apply the patch +You have a miter that should be moved into the header +Please somebody apply these somebody add these filesheader to rampartincludespurce to rampartsrc add +Hudson do not respond committed patch anyway +Seems worth it to fix this in the branch too +But dont see any change when I try have the text file resurrect the osgi config on a different instance +A locationmap should be able to be supplied subsequently without restarting Forrest +Are you trying to run all MR jobs in parallel? Because I worry about conditions where the tests are making some assumption about unique access to the filesystem or mini clusters +I think perhaps the first one but dont have a great idea +I have my asbestos suit on so fire away +Reopened issue to the latest patch of CSS changes to the bluelight theme +This was resolved in ER +voters please move your votes to +Most of this comes down to how to handle the old jdbc modules Justin do you have any thoughts on what you would like to see happenSuggestion close all existing jdbc db oracle postgis issues remove transitioning all issues to jdbc leave the db oracle postgis components available to document the new datastoresI dont really know where the issues for postgis datastore being reported currently +Committed patch revision +rob please set fix version for this +Daryn By not Im referring to the fact that external filesystemsGood point but the new apis in or the new you suggested will not be implemented by the external file systems and there is no default impl that works except for not implemented exception +This is on a branch of various build changes that I was working on but I stopped because Ruby was giving me trouble +Oops this got moved to common this morning +with a bullet +Were you using a custom implemention to get the +Sorry if I left something out +I created a patch for it just ignore the Exception that is thrown while scanning a jar fileNot sure whether we need to do it or there is a better solutionBy the way Jrgen Weber does your war contains the? if does it should be the same reasonThanks for any comment +Attaching a new rev of the patchThis rev aborts the build if you are using the IBM Java compiler AND you have explicitly set EITHER jlib or +What Robert says in Comment is mostly correct +And which DB are you using +Sandrogreat! thanks for looking into this yes I will setup osgi test project and let you knowAndrei +Patch from patch from RB +So it should be doing it +Ok Guys I dont haev enough time to do the full mavenized site yet + contructs the link with the wrong GET params viz +Agreed with Tsuyoshi that we should make these upper case +Was even able to fix it without a change in the ReaderStore interface +Unlike the managed request scope in this would go away after a single actionrender request pair +It seems bindData is completely broken as soon as there are associations collections +Obviously we have a dont seem to have the rights either +Ok Kris its clear but this bug is quite critical to work in production +closing since fixes are in released code +We happily take patches Im inclined but I think Oleg has a point about having time to do itThat said this is something Ive been thinking about and I think the rough approach David outlined is probably right +Yes we played with it and it seems to work fine. +Ill update in the evening +Please check that the bug is fixed now +I have been experimenting with splitting my own javascript files to load some bits and the rest later +you will need to fix your since it has snapshotbest if you register with Sonatype to deploy your artifact resolution should be wont fix +I have downloaded the same attachment from a JIRA issue multiple times and gotten different results either truncation or Ive even seen a OK in the middle of a diff +dont need data nodes for this test use set rather than array +RMI using the or SOA includes in the directory +Patch attached +Now there is no way to do mapjoin for this multiway join as the same query fails with this error after changes to +Trying to fix javadoc +These are local artifacts from the current build +The patch looks good to me +Fixed in trunk +Is this still an issue? I was unable to reproduce it on or trunk. +Im sorry for any inconvenienceTake care file openejbexamplespatching file openejbexamplessrcmainjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplessrcmainjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplessrcmainjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplessrcmainjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplessrcmainjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplessrcmainjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplessrctestjavaorgapacheopenejbexamplesinjectionpatching file openejbexamplesHunk FAILED at +I believe this is done. +Its pure Ethernet nowFor ER all elasticityresilience tests were run on hyperion again with EthernetWhat happens on hyperion is perfectly valid now +The motivation behind this is that the arrival ofentries in isnt that easy to predict since the work iscarried out in a background taskAlso some cleanups in Utilities o fixed some o fixed formatting o added method sleep o simplified repeatChar +Let me know if you receive mails +This can be solved with +For trunk we need to figure out what regressed for pure Ruby mysql driver +this is just not the behavior i expected +to come +AndreasTim do you have a recommendation on how exceptions could be detected by client software that doesnt break anything +For some reason the crlf line endings are not picked up in the Mead build so I will try another approach +Raise the severity. +Hi just wanted to ask whether there are any news or plans regarding this issueThank youCheersBen +Hi Svante The previous patch does not work for some cases +Confirmed this is working well and has allowed me to dispose of some nasty workarounds +He says he is running branch but also lists as to fix for +If youre interested in shell access to a Fedora Core machine with GCC and Valgrind installed let me know +Tests pass for me with SVN sources and Ant on Windows XP SPpatches with Sun Java Sun Java ll try it out on a few other platforms +However here is another patch that will avoid compile of any c code for binary target unless specified Dcompile +we will have to rerun pmd for when its getting closer +I noticed that after this patch although recovered the repaired directories they were not actually used +Attached a single patch that removesupdates all stuff +If it is I have to ask why as any effort to provide a consistent and as of externalizable entity resolution layer is not be used by the xmbean layer +We could add an interceptor to the server and catch the first createsession call but i am not sure how the RA would register the interceptor or get a handle on the server +in a sense the goal was always just intended for converting a netbeans IDE install directory to maven repo +just committed this. +Im trying to research if there are specific properties I have to enable for these enumerations to work with nmtokens and what happens if I make a list of them +I guess it makes sense to be able to add multiple cache directives for the same path +I had similar comments as Colin posted +I get this test failure about once a day +Jay appreciates my notes practically always +I will be investigating this +Where are we at with this one? Does it still need mailing list discussion +Yeah I can move the files if you let me know which files you want to be moved +Now the issue is observed in the branch as well from today +Isnt that a lot of users? Wouldnt it mean theyd simply ignore queue and then job actually work fine as documented +Changing status to Patch available so that this issue will appear on our review queue +There were a lot of conflicts with the previous patch because it was implemented before was committed +Option B sounds better but perhaps trickier to get just right +Yes for publishing in JMX +Thanks for testing balancerGood finding +whoops sorry different issue probably +Committed to JBESBGACP with revision +ERROR XJ Java exception ERROR XJ Java exception DES key length must be bits This is interesting because we say the default key length is +Sounds good to me +In my engine I also expect the user has already called declareObject +Manually tested! +Yes resolved +test fails for me on with this patch and passes without on the same clone +Are we going to go for this for +Create new repository tenant +Ive changed the config and will talk to QA to see if we can try measure the impact +Shaz any thoughts on this topic? Only thing I can think of is providing the standard swizzling util functions as a part of cordova to aid in the process +Might help us get what we want +Here is the that causes the problem +Isnt that array wholly made and manipulated on the stack +This issue not occuring in ode anymore. +To be more specific creating keyspaces like create keyspace keyspace or create keyspace keyspace or create keyspace family would fail and raise syntax error error +Thanks Pavel Alexey and Gregory +Chances of can be further reduced but again not eliminated by all places that iterate childResources making a copy first +v removes some dross included by misstack +META +From best I can tell and have never applied to the code path +The patch looks good to me +Bulk close for Solr +i am also facing same issue with the ASF master am able to successfully deploy a Setup with System running on them +Attaching patch merging with lastest trunk +Thank you +The view directory is not in the classpath so the tests would break I guess +There is one test that fails but it is unrelatedShould the patch still be submitted +Not a bug in X due to removal of values. +Thanks Eric for the fixed patch + Attached Patch that matches v that is up on review board +The problem is exactly that the models are stored in a tree and they do cannot be detached from root instead from leafs because leafs are independent +No problem Tammo +It seems that grantwas actually demonstrating the presence of the bug but we didntrecognize it before because the error message didnt obviously showthe problem +In addition I noticed this and some other conditional logging checks were logging only if logging was disabled + +That might help explain some of the other random encoding issues Im seeing under jRuby +My feelings are that this could turn into a pretty huge task as Im not exactly sure what this would this entailAre we to implement this moving forward prior to release of trunk both +This toll currently is invoked by binhadoop jar fileA fileB +features descriptor is not in the NMR distribution system repoBut its not required as no feature is in the bootFeaturesThe features will ship the features descriptor. +This patch should take care of it +I will update +Fixed in trunk as noted it was a regression from my recent rewiring of the string case of split +I pulled the descriptors code to see what I could do about this but got lost in the code generation so Im not going to be able to send a patch for this one in a hurry +The context object was holding its own lock when calling back into application code +I think you have it right +Chatting with Jon this patch needs more +I wasnt able to repro in +Verified the final set of instructions in +Fixed the problem Now the prepare soap envelope method would take SOAP version set into account when creating the SOAP envelope +Committed to and trunk so it is in place for and we can see how it does under duress. +Thoughts +Im pounding the code with threads in a tight loop +Today Im going to directly ask it to the owner of jenkins +could be replaced withThe reason is the request and response could in the future change to include additional optional fields +Verified test case passes and manually verified on a cluster with a fail job that the diagnostics appear on the RM page for the failed app +Anything that starts to product multiple artifacts with a subset of a source directory might not integrate well with other features +On the flip side we cant make changes without disrupting existing applications +Many of these expressions are pretty generic and different systems are likely to have similar semantics for these +I try my best to keep code organized +If you feel the rest of the patch corrupts the consistency with the other connectors you can just close the ticket +Release Notes Text is correct +Bugzilla changes have been done +Agree it will be possible is lock is set on the region in the file which we wont be reading +I think this is fixed now it was failing due to I didnt check the default one +xcf files tooM productpluginsM srcconfigresourcesimagesM srcconfigresourcesimagesM relengartM relengartM relengart +for the to Max for review +There is not any comment that tells us why we should read them +This bug had already fixed in +I will open a separate JIRA for pattern matching +We should rethink this in the light of we deprecated the and have the new CRUD support I resolve this as wont fix +Can you confirm the IP address of? has address +Most of the code was borrowed from the test modules of theThis code has been tested although its definitely not for production use at this point and probably overlooks several axis features. +i dont know that being wrong X of the time is that much better than Xwhat if we name it something more generic? columnFamilyparams or columnpath +As saw discussion in todd said HAMA use so asking here +This issue was discussed in HDFS Design Lounge +John while you are at this please change the to use +revised patch looks good do commit. +It is so much cleaner and easier to read IMO +Resolving as a dumb idea +Merged PR +Moving to will not be a moving to for all of this to +Now it should be fixed + +Patch which addresses the issue +Probably we can skip the call to when we reopen since we are not really reusing the file container object for another container we are just reopening it +See the new annotated test clients as examples +Example web app shown working with sun reference impl +I committed what I tested +I was also able to generate eclipse files using the eclipse plugin +I can confirm this works thanks Thomas +Merged onto the master and branches +The unicode escapes are merely by virtue of the fact that they caused an error previously +I am closing this as we cannot reproduce it if you have an example or steps on how to reproduce it feel free to. +Ill commit the patch immediately +I think I wasnt right about removing check box +Duplicated so close +This should be the correct one now +M javaclientorgapachederbyclientamThe getLocator method that returns the locator value associated with this LOBThe isLocator method that indicates whether this LOB is locator enabled or not +in +But either way works +The is from Tyrex +Changing issue title to focus on the error cascade since the original typechecking question is pretty much cleared up modulo the fix to the type of the closure in the methods +rehashes itself +Is that true +This is not using svnpubsub the static files get generated and put into a folder called public in the thanks! +created xwwwonly in svn but not in effectvhost entrydns entryqmail rcpthosts and virtualdomains +Ah probably because you use Instance not direct injection +Saxon does the conversion automatically provided thatxslstylesheet has version +mosgi only compiles with +Sure Ill merge it with the latest version and create a new pull requestThanks +Im attaching a patch file which fixes the problem but in a manner that is probably not desirable +resolved in branchsoft links are added for linux and macosx in installation root folder +Trying hudson +Seems pretty useful to me +I fixed the same issue in the CSV as well +Usually snapshot will be null but if a snapshot happens during an outstanding scan of a sudden the snapshot will now have values in it and well start satisfying memcache next out of snapshot +design completed +Test KristianAlthough Id love to get more involved at some point I dont quite have the time to learn Derbys test framework and adapt the test case +I can probably do that +Had a hectic last week +Please convince me that this use case is spec compliant +Could we go for a deeper level of integration? Instead of storing the data twice as Cassandra row Lucene document use the row as the document Source Of Truth and just let Lucene handle the indexesYes sure but still requires constructing the full row before writing it to the index since the client may be updating field but indexes are on field and field +I am loath to put a kb flag on I frequently have todo merges on these files +Patch checked in +Ive created a simple test application +Which really wasnt possible before Java annotations were introduce I believe +Updating the release notes fields +If you find no real difference between changing the signature vs +Once we restructure to not put logs appends between memstore puts we are literally talking about the speed of adding a few dozen entries in an array +Can you elaborate on the change to shutdownBroker in? Ithink we need to include shutting down brokers because the previous shutdownattempt may have been incomplete due to no other brokers in ISR for somepartition which would have prevented leader movement +I have committed the attached patch for +I would prefer to address this with a consistent overview rather than altering log levels ad please comment here with log levels you would like to see changed so that we can get a good idea of what is useful and whats not +I wonder though why the behavior is different when using the example like the example logging I have above +probably need to try on a different VM what version you using +We also need Perl tests. +Modified line to correct the quotation issue +Michael +Need more thinking before applying this patch +Marking this issuse as fixed since all unit tests except now pass failure will be tracked in +Thats the error you get later because the management model is now corrupted +Please check that the bug is fixed +Clean up old issue was deferred in planning. +Apparently that code is part of dead dirs and we still need to implement this for YARN sorry for my ignorance. +We agree + +As i know you cant send object of different types inside single map. +Currently is using a different version so we should try to get these in sync +making this a blocker as the change introduced by in the filesystem client requires a WEBHDFS JSON field +I just skimmed over the bigtable paper and could not find an explication about what they do in this case +We need to provide strong usage examples and point to the right classes in the new API +I can add in such logic to the test class to check the job state consistency +The soruce url can contain variables which will be resolved each time the input module is used +I only asked Wensheng to seperate the changes +Thanks +A few preliminary requirements as I see it +not yeah +Test for the detailed report BernardI was able to reproduce it with your sample +Thanks In that case Ill wait until after the overhaul to revisit this +ported this back as far as +Thanks Jervis! +I have used this patch against a certain version of hadoop security and have had success running queries +sound like unrelated issues please file separate reports with their own patches +Looks good +AlanWhether to introduce support for JQL or not deserves its own JIRA issue +Companies which offer free licenses for open source projects often request that you put a logo and link to their site on the projects site +I have all tests passing in intellij in windows +Ryan do you know where it is being set to DEBUG? A grep for in the source tree or the built tree return nothing +This issue applies to codebase prior +Check out both as the landing page and then as potential internal pages +Flash demo movie is on. +I see similar errors reported in where Parameter directory is requiredDirectory containting property filesis reported +Also can you provide javadoc for it? Any reason not to use the more standard M type variable? Some Javadoc for the class would be nice here too +I took the liberty of defining couple of new properties so that it would continue to work with Hadoop scripts as normal if above properties are not explicitly set +on linux needed to add packagespackage list becomes sudo install g gcc bison flex gperf patch ant junit junit epm subversion autoconf rpm I to conclude that you just want installedHaving an upto date package list is nice but it hides from me what you actually want installed +The screen will be located inAccountingGeneral LedgerAccountingReportsBalanceCan you please provide more information about this issue + does this now fall under what youve been doing +The network is full in a short time when do checkpoint leading to Jobtracker access Namenode to get relevant file data to fail in job initialization phase +I have verified that converting the spaces to hex and adding a backslash allows the ldap authentication to work +The rule of thumb that we worked from was the Source was and interacted with external resources whereas Item concentrated on the objects +I hope the m war plugin is not so flaky +To generate all config files is used +This isnt a real big show stopper in my mind +Interestingly enough +Added with a fix to the to ensure that it stores only client information. +I just committed this +Im not an Ant expert anymore obviously so I can guide you but you probably know best how to fit it into the existing ant build +Same here +Issue resolvedThanks Andrew. + +Heiko has confirmed that remoting is good to go for jbossws +This section you worked on is about Functions +I dont think this is exactly a bug that we can justify changing in +Needs clarification in the docs +committed this. +Yeah it is sort of like faceting but I didnt want to get too close to it otherwise were just duplicating effort +Jacques Now I think its ok +Problematic collisions areCollision class class class class class class class class m not sure why there were problems all that dup does is add an attachment to an attachment list the ordering should not matter +file causing the unable to determine structure of projectSince this is a project single is not enough to investigate this problem and complete tree of required +Edit More detailed explanation You are proposing what sounds like a HDFS approach for rearchitecting the REST interface +Thanks +It would be great if you can have a quick look before a commit itThe solution is simpleIf the session is timed out i check in the request handler weather it is an ajax request or not +Changes look good not sure on the tests will have to think on that +Can we do that +Then we should manage our own Utils for String Class +This would be great as a community contribution IMO +I changed the pom but did not realize we had versions in all submodulesHow about this patch +Tests with Axis Java server are successful +Running all of the functional test takes to hours +Otherwise its a bit confusing to look at +So to answer your question why File and URL have no encoding property Because they are references to things that may or may not exist and that may or not be text data in a certain encoding +In order to enable it interested customers must replace the mapping resource orgjbpmjob with orgjbpmjob in their Hibernate configuration in addition to creating the column +Attached standalone test that demonstrates the problem +Attaching MSApi logsdb on master stable branch passed +Good ideaAdjusted entry settings +Not just ignore them +It is up to the user to choose whether to use cocoon or http +I will make the changes you recommend +bsearch on an array will be much moreThe Builder pattern seems overwrought here +The patch previous in this jira +Patch should stop region server from hanging even it it doesnt stop the failed because it was configured with too few retries +Latest version fixes missing file +I just improved itNow they are separated. +Concerning the compilation sequence I guess that in my example the production code should be compile first and then the test code as dependencies are quite simple +The philosophy of Maven is that the project should be completely described by the pom +Unfortunately I ran the tests again so the original log is not available anymore and the problem didnt occur +You can get HV to work in if you dont patch it to JPA level and rename the jar that resides in the modules directory +but if needed its an option +I have added status handling if no handlerroute is found +Fixed in trunk revision . +Thanks again +It looks perfect go ahead and commit it in Skimming thru the patch on a Highlevel +Marking closed. +Second part includes +We have added templates to visual part for milestone +As to the performance the current patch at has a bitwise linear search to do this +ppr cases do not work but about of the page works with ppr which browser did failWe might have a small javascript issue here which might be triggered on a certain browser version +I have also seen this bug in the wild +It wont overwrite an existing definition so in the case of declarations the first match for a method will always be used. +Thanks +Can you attach the ear to this jira +Simple patch that makes this test as a unit testThe code to start and stop has been refactored into separate methods in patch that tests this functionality as a unit tests passed on my local patch has a timing issue +LGTM +Considering your last statement sorry for the noise +This does not appear to prevent the dependencies from being downloaded. +On the other hand if you know that recalculation is not part of the upgrade process and you know that you can just kick off recalculations you would do so for the test runs you care aboutI dont feel strongly foragainst either options though. +x +Could you provide an example of Surfire XML report available in the target directory? Martin without your Surfire XML report we cant reproduce your issue so Im closing this ticket with the resolution Incomplete +Remove files copied from yammer metrics project + +Nice patch provided +Just to be specific can you post exactjdk versions of the jvms you are running the tests with when you post your results +Fixes checked in should be working in next buildDone +I committed this +Improve the performance of the query that determines the for each weblog +Ray can you take the pom tweak Rajeev describes once Ive got working again? Or remind me where and how to do it +I added another patch +Maybe leave it for a specializer codeOK I agree lets not do this one for nowNew patch looks good Ill review it some more and then wait a few days and commit +Ive added some hack to pickup these generated classes so they are visible when running tests from proper solution would need to update Mavens model to allow to register and introspect folders for generated classes and will requite to update xmlbeans plugin to use such api +Part of a bulk update all resolutions changed to done please review history to original resolution type +Add fix for archiva layout and unit test forSelenium webapp test needs to be fixed and webdav +I have created +years passed since this issue was openedAny changesStatus Resolved ResolvedResolution Rejected Rejected? Why? This is very important functionality +Chris can you put all these Tika TLP issues as under one main issue thanks +My first example was hills while the second was spikes where I agree that the or something similar works well +Adding test case +Problem is also among the default output format not being the correct one that if a client uses it to make a request an exception is on trunk and +What should I do to get the patch please let me know the details +Thanks Nigel! +Committed to trunk and x +Patch applied at +All of these issues have to do with text extraction specifically and should be in that section +Can someone look into this issue +Yes sorry its under +Thanks Denis +Hi im working on it right now i already have the resolved but i cant find the refreshInterval value how can i get it +and respectively provide versions for applications that require concurrent access to statistical aggregates by multiple threads +Serialization time time to encoder each event +I just commited +This is ruby stuff I am not familiar with +Just applied the patch on the trunk +Andrzej are you still working on the carmel policy? I see conf isnt yet implemented and I cant seem to get it to prune anything with just a default threshold +Closed +I made this patch to be applicable to the trunk +Committed as r +Cant find an Assign to me button +The attached presentation describes the Authentication Point Support for Wink +Also compiling with groovyc and then running the class using java with groovy and asm on the classpath works fine +Tests applied the patch refactored the class to be an inner class of the testPlease verify the patch was applied as expected +Any reason you think this must be included into the base API and not kept as a util method +on making a separate dist and example for each contrib folder +Closing based on last comment. +And Im seeing loads of this Shutdown of DFS in my passing so doesnt seem like it always breaks tests but for some reason its now causing breakage in +The tricky thing is and this may be what you are seeing when you having profiles that add executions to plugins wired in the main section of the POM in this case the order of execution is determined by the position of the plugin in the main section of the POM +Very tricky +Yes looked at it earlier + +Committed all JBDS GA doc sources to the branch +really unstable yet such a critical part of our apps and precious screenThis is all being tested in Phonegap Xcode on an Xcode updated project +reopen to set fix version to again with correct fix version +Good work! on the patch +Really this was fixed in but were accepting Roberts observations as proof its really gone. +BerinYes youre right +Could you give it a check over in terms of whether the examples still work to expectation etc +committed fbaeaddbddcdbbefefd +close it. +I believe this version of WAS is shipped with IBM JV SR which has this problem +please confirm fixes are correctI dont think there is much we can do about the bpmorchestration quickstart references at this point +It only works up to though beyond that its large pointers again +I guess the main concern here is that concat is specifically a operation a new base class with concat featured in it is feasibleAnd sure it works but only because it would get the job done +What happens if the path is too long? Hopefully theres some decent error +defer issues to +Chris you are correct +Note that this patch is against I can create another patch against andor other tags if desired by reviewer +Latest patch makes those changes +Patch looks good +This isnt ideal but is safe since Hadoops calls setConf before it calls run +Created static in Bookie to generate from conf +Well create a new bug if necessaryThanks a lotve decided to manually apply the patch from dated to current SnapshotIf this patch wont become official at least there should be a new class to handle the thread safe issue +When this is option is used the will work as delegating fashion instead of for the specified packages +Personally I would delete this but I dont know what parts of your implementations are grounding on it +Ie the level byte is only needed when the slice isnt full yetBut if you move the level byte to the start does that mean its never +I have checked a patch and found one more situation when an doesnt Use the same project from description Open page Make an on page should be FAILED isnt opened +You are right that public sounds correct +Per IRC we should also clean up misleading comments in w comment fix +refers to xslattributes attribute namespace not the namespace declarations +Projects could even include a configuration in a special config branch that only exists on that server for the purpose of defining what artifacts are published where and do not equate the practice of making useful tools generally available with centralization surely its obvious by now that policies geared towards centralization for its own sake have not worked particularly well at least not in the last years +It would be nice to be able to use the other JMX implementations but unfornately somebody introduceda dependency on the the mbean registry which should not be there +So I assume that more concrete enhancement requests would help you to improve the +How come? Note that all the talk about session flush modes ect isnt realy much to do with the patch the same situation would happen if the hibernate interceptor were added declarativly in the bean definition file +Patch applied +And none of the arithmetic or other logical operators work +Looks like we are past the point dims +patch lgtm +It appears for example the C implementation of schema resolution is partialId prefer to refer to that as a bug than as an established implementation level. +But that exception is removed from the sig of add but is also added to the sig of removeShouldnt we keep it or remove it from all methods +Fixed +Ilooked at using the Xerces URI class but hit issues with questionable windowsfile +If there were setters in place then every instance that I can see of the replacement objects wouldnt need to care about the concrete class at all which is more in line with what the interfaces are supposed to abstract out +Hi Adam it seems this is not a problem any more +will it has implication of generated source where we add addition dir to source rootIn the case of phase are we able set excludeinclude as well +Consolidated typos correctionsIncludes the differences of the previous patch +That will kill your performance in a high load as you require +i dont remember writing it that way but we should change that to create directory after the transaction has been committed +I then decided to simplify my approach and did pretty much what Uwe has suggested all is working now +The fact that the Datanode is shutdown message appeared in the log a few milliseconds after is meaningless the DN shuts down its bits in a sequence so it may have already dropped out the cluster before the message was printed. +Meant to mark this Patch Available before but forgot +I had no idea about fetching binary data sorry +I face the same problem +Appears to be caused by ODE synchronization issue but only occurs under very specific circumstances +There are no errors in the log while I can see the commitlog replay is accomplished +I guess should be them deprecated we update docs somewhere to indicate how to get debugging information +While looking at this jira noticed custom partitioner being dropped when run with multi query optimization +A is created per WAR deployment +Please provide an NVR where youre currently seeing this issue arise. +I may have to deferr this to because the generic type handling needs to be completely rewritten and various need to change which would break backward compatibility +Adding patch for +this was added mainly to have a test for doctermsindexs termenumso i think srctest is the correct place +However Ill leave this issue open because we will need yet another update after is resolved +Merged the change of camel back to trunk +Looks like we have forked patches +Thanks +that was added because of Oracle client installation +Well at least the fix works as far as fixing the system typeDoes the code work if you use a simpler file name +I havent had a chance to look into thatpriorities had shifted +Shows the changes I made to get past the initial configuration checks +My changes were to move the weak map logic to a separate collection class outside of +I assume this is working for now closing. +That way the old version is still available for posterity +Whenever I was called to fix a Maven setup it was because people setting it up initially werent thinking in Maven +makes certain assumptions about its use +I think this functionality might still need a little bit more error handlingGenerally it looks good and it is definitely useful if you need to rebuild an index and you want users to allow to use the old wile the new one is not yet donebest for feedback Rupert +I plan to commit this patch on Monday unless I hear there are issues +Please check that the bug is fixed now +I can confirm this one using latest release of maven and karaf branch on my archlinux s the same issue that we had before +Still nothing to figure out the tmp file name +Try them with and without storing the field stored fields have to be copied during merges. +Is it possible to add a +In fact this is not critical as the password is correctly the success message is update bug priority as I dont have the rights to do it +Defer to defer open issues to defer to not essential to address for +I think that a sufficient amount of testcases have been generated for this XSDJava process +Ive added an overwriteMode attribute to the ivyretrieve task to configure when the destination files should be overwritten + load and after refresh callbacks of openjpa were each resulting in a post load call resulting in the duplicates +Sending componentsSending examplesSending Transmitting file dataCommitted revision . +This is a tricky change I think but shoudnt it be more like thisqueryNorm takes the Weight rather than sumSquared and then you can call sumSquared on the weight or skip it saving the few calcs and isolating the work to the queryNorm implThe patch is unfinished though its an issue that queryNorm throws I dont think it should and i removed it here but thats not back compat +For central why cant central feature just depend on AS tools and get all the right dependencies +I will look to see exactly which conditions this happens in +Maybe instantiate but I prefer the current nameHave renamed some of the other methods some as suggested others with slightly clearer names and have added some comments to make the tests easier to understandAlso added documentation to stating that derived classes must include a String only constructor +I really appreciate your submission happy to work out the kinks +Thanks for the patches James and for the reviews Elliott! +Anyway will list out the probs that i faced with one and znodes +Kostis +Tested with the version. +i guess a impl for the that can expose position information thatd be greatI would imagine that for a single segment were still very capable of knowing the position of termI have code that depends on this functionality i would like to be able to try it all out against lucenetrunk as it moves forward +I send messagesA priority xB priority yC priority zI expect to receive messagesC B Abut instead I receivedSomething like that +Im hoping Jeff or another tomcat expert will comment on thisWhen we get it implemented for both containers we can get the xml into the generic schema +Uma do you mind if I take this over to finish up your patch? I was planning on working on which is closely related +I can see the argument on both sides +Ill try to come up with a class loader that is not subclass of and see if the error can be seen +trunk already contains the fix it was applied soon after when was resolved +Well try to make sure that includes Rake again +Add support for URL in upstream Aslak any plans to implement also variant +does jvm tuning out of the by enabling Concurrent GC Sweep in the For Sequential Versions we can enable it from the Shell ScriptFor Hadoop jobs to get the benefit it has to be put in or in conf closed +Type aliases are now case insensitive +This patch includes definition of entity for associating with +John yeah i use with the latest trunk +The issue isnt reproducible in JBDS +If you want the same package names go ahead +Hmmm +In my opinion the reason isnt valid +Perhaps javadoc should be tweaked +Tyronne to get back to me the week commencing Febrauary with info +trunkCommitted revision +This is a feature of the Graphiti framework but we should able to solve this +If you find further issues please create a new jira and link it to this one. +It is necessary to tell userdeveloper that the called method is not available on startup and it is being called on that time by some code +Ok because we are going to go with JSON and not use more or less unchanged I think I will take the core of it but I will probably rename it so there is no confusion with the Tomcat one +thank you +And try with latest release Camel +Aplologies if i have misleaded u but i am using eap +No complaints here +I saw it thanks +i dont have that log but it should be easy to reproduce +Committed to trunk +Code reviewed +agree and will create wiki page for that +Thank you Simon +Which version of Hive are you using +It would be great if you could reproduce the problem with the wire context logging turned on and attach the log to this ticketPlease your application against the latest SVN snapshot to verify the fixOleg +Method added thanks +I need to bind it to the lifecycle disable inclusion as a connector when it makes sense +classpath loading +I forgot to commit a change with r done at r +Hey Bob any additional information about this one +My bad changed and uploaded +The suggested support for existing Jackrabbit should be a good starting point. +I have attached a test case that breaks for the given scenario +I take this to mean that does not provide the ability to iterate in even an approximation of least likely to be evicted next order +Can you run the samples on your machine? Thanks +Good catch +Can this be in earlier release +need provide the container value in the sample page +Moving out of + +This is not an issue if you only have Pets but say you have thousands? Not only do queries for single Pets take an age to load but your house smells funny +We may be dealing with multiple bugs or bugs masquerading as other problems +I applied the change in on master which just comments out the newer ivar table logic +This works as designed +I uploaded the patch as a starting point +I just committed this to trunk and branch +Congratulations on issue +Dear Mr +Thus the name reflects it by no following the convention for the rest of the target +Hmm I dont know how to adapt your advice from Intellij to ve set a General Project and converted it to a Maven Project +Attached is the patch that fixes this + is marked as a of but in reality this is a separate bugThe existing comments in and repro conttinue to apply for this bug +Closed upon release of Hadoop. +Yaayyyyy +Please commit the patchThanks. +All right Ive committed the obvious and hopefully change to CVS +bq +So people had to care for those headers or the resetting of those headers in their custom subclasses +because is an avalon context value and the semantics differ +Also does bjoin needs any specific types of field configs in +The second patch implements the comments above +please check into php next to java +Both trunk and +Do we need to resubmit +Humm +In the meantime need to get Alpha released so pushing to we can just add and the test suite? was fixed by adding AUTOCOMMITOFF to the URL registered with the datasourceThis change caused failures in +Thank you. +Im not in a good position to really test in a environment so if you could change the class to extend and convert to convert I believe that should do it +Closed all tickets +Please review that big patch +I do think this is the right sort of place for a JSONP callback to be handled because irrespective of the implementation used to render the view its something I would expect to be handled in the same way. +This is very bad as it can have dramatic security implications +We are interested in implementing the Connect specifications in the Apache Amber project so we can contribute it back to the community +previous ran on +It has to walk the inheritance tree +After is branched we can come back to Guice and start the process slowly as splitting this patch got too hard. +Im guessing that you would get an error if you specified an encryption key that was inconsistent with the specified encryption key length +Since this field really applies to the weblog and About me doesnt really work when you are talking about a group weblog how about a compromise +This is fixed by now +It definitely looks like a step in the right direction +Moved to beta +Thanks! Done. +Damn acronyms Can you please elaborate on what does PIA mean? Thanks! +The root Entity makes its way into the index but not the sub entities +thinks it can only write GB of data to the data directory but that is not true +Would you be able to come up with a patch +The effect was that the SF reader knowing nothing about compression couldnt read compressed entriesThis v patch fixes the test for when WAL compression is enabled +Does this test fail consistently for you +I have updated the to successfully find the configuration and removed all old references +All the connections I mentined above are ESTABLISHEDWorkaround with minIdle also helps +It looks like we will need to provide two versions of the file +Please try to use assertEquals assertNull etc +For clarification this issue with only uppercase propagation type doesnt affect because was fixed after release and the bug is only in branch not in tag therefore I am removing Affects version s true this bug does not affectThe bug that actually does is this one which eventually made me try to use a snapshot of Seam and discovering the bug introduced in the way I guess this fix is also welcome in the branch +Pushing all issues from to +As an update Ive been working on updating this patch and finally have all my tests succeeding but Im now told that the KVCell changes are being rolled back now in +I promise I left all the fun ones still to be converted +Looking at this issue now +Yes I agree it doesnt look good Ill try to follow your suggestions you please close this issue if you find a better way to do that +Hi Cristian I would like to see a example usage of those annotations +Sample test app +Thanks Fernando I will rework the Jira subject to clarify that this applies to the XTS component and reassign it to my colleague Paul who leads the XTS aspects of after working the subject as per Fernandos comment this is looking like a duplicate ofI wrote description of the jira not clear enough +and sorry for the mistake in naming the patch files +Thanks! +Regression tests are included in the patch too +David still looking after the printing module +Already fixed would have been a better resolution but alas there is no such option +I cannot reproduce this problem using your example when I make a standalone test case out of it +Will it be given more memory if I increase the maven opts? What do you set yours to +Thanks +To raise a JIRA for feature you seem not to share that opinion +All other clauses are logically evaluated before the SELECT clause and thus cannot use the output column names defined by the SELECT clause +Subsystem test has been fixed and enhanced so both and versions of the XSD are covered. +bq +That will make it easier for me to change to the official release once it gets released +I just commited this +this patch includes konstantins changes adds a test and makes some minor changes to fsshell +If the Position named window is based on a POJO event class then that class must implement Serializable +Attribute names were updated +Im ok with all the comments I will take them into account for commit +LGTMNit would prefer avoiding allocating unnecessary in trim when the common case is not to need to trim +I have correct the signature + running to trunk + TrevorLink works now. +just applied the patch on ubuntu and it worked the same there too +Closing released issues. +Thank you so much for your prompt response Ivan I really appreciate your helpI have already maxed out the RAM on the machine it seems very strange to me that adding a whole other GB of RAM did not fix the issue already +There are lots of methods classes and interfaces in the Java platform that are deprecated and none of them can be deleted because it might break some existing application +Id prefer to see it as an external function rather than a method of the structure so that the structure code doesnt become dependent on any one transport implementation +I switched to the latest WTP and cant reproduce this issue anymoreSo the bug seems to be fixed +Now think of the same situation over cartesian product of sets of languages frameworks and DBALs! And what if the DBAL isFor now Yii DBAL works with Cassandra near perfect thanks to language similarity except for some bugs i do fileSo increasing similarity of CQL to SQL will be for good of everyone less reworking in all DBALs and frameworks more argument points to use CassandraPlease the bug even with far milestone! +I am moving this back a release as it involves invasive changes to the current codebase +One of these will be created for all that do not themselves contain a or +When i use hbase hdfsuserclustermbeoutput MBE instead of hbase mbeoutput MBE the issue is resolvedMy problem was solved using absolute addressing with port number instead of relative addressingI think that the difference between providing address with or without port number is an issue and has to be resolved on next releases +Investigate further +If there are no plans to upgrade I will do some code review tomorrow and reconsider the route +It could probably be done while uploading the image from SNNAs I pointed out in an image on disk or in transmission might get corrupt +Attached is the patch for the new +Committed atsubversion revision Currently the release instructions say that you should create a cgi script for the new release by copying the cgi script from an old release +Username will always come from the preventing this issue. +Smiley is looking into revamping how this class should work in and as part of that all the existing methods would go away but some new non static state methods would be added for a new type of hook to access the current while processing documents. +A patch is attached +rebased +It force the secondary supervisor to restart so it should also restart all services attached but I may be wrong about thatThe reason I use a daemon section here is to allows a plugin to be loaded as a supervisor or as a worker +Its possible I was just mislead by the fact that the compiler accepted it +Im unable to login with the password test I tried to reset it but I never got a reset email can you update the email for this user to eli and send me a reset link +Added the config files that I use for each node +After I added the check the code for the if and the else was identical +Thanks Ryan thats exactly what I was thinking of +MD and SHA are fixed bad PGP signature is still in place +I learned many things from your code but it needed changes +Tested with Chrome Safari and Firefox +I will up the period make it two days at leastBut was thinking too that we should mark files that have been major compacted write the fact into the files metadata or into the file name and before running another check the column descriptor to see if TTL is forever if it is do not run another major compaction if only one file to compact +But replacing the fop jar with the fop jar still doesnt get us past this problem +And I will try to add test cases for this +ill checkThanks for reportingHave a good dayhere is a patch for the reported issue + execute them from the shellexecute them using hql not shell +Created the test directory as requested +the patch looks good to me +Closing the Bug as Verified based on the screenshot information +log remote address as part of handling for the patch Sriram! Few very minor comments +Well Ive done it the other way round the maven plugin doesnt autoset the configuration home if remote +Changedso that start and shutdown are only called one time and reload just refreshes the sources and drivers +Patch + +If NN exits is torn down and any subsequent checks by tests will race +Closing this for now +testing that right now +I did a wireshark on it and didnt see it trying to hit the web site for the schema so it looks like its getting the schema internally +Even in a world with it is reasonable to expect tasks to be constrained differently from tasktrackers and jobtrackersLong term we will likely run tasktrackers as root and setuid to the submitting user and at that point will suffice +Few more tests for the comments Mark +ant f produces the exception +I think it is possible to relate to stomp and hornetq + on simple you guys think it is worth doing above? It would save byte per KV +This is implemented +In real life the timer can be first undeployed before server shutdown and this is actually what the README file says to do +dupe of +gitignore is relevant yet useful Heres new oneI will run commit the patch one more. +Ill keep my eye on it but Ive never seen any tangible symptoms from this so for now its just a curiousity +I have MB files that cant be handled when giving GB of RAM it required GB to process +Finally the change is made visible in the freemarker template +patch files generated by the diff command +All done +Not a Windows problem +Those are indeed not appropriate anymore with the new for pointing this outJuergen +I cannot reproduce it +Thanks for the patch +so my machine I am testing on is osx +This issue was resolved in release of beanutils +So if a user or group is not found locally throw an exceptionBTW what Linux does if a remote directory is mounted to a local machine but some accounts used in the remote directory does not exist locally? We probably should do the same +Linking the issues because is the documentation counterpart of this feature +Jan please update the list of attributes that are still not covered and assign to Jiri +Also I dont think this needs to be merged to is that correct +Also Larry closingthis as WONTFIX simply means it will not be properly resolved untila genuine use case appears and someone reopens the ticket +Since the blob being used wasnt a it would get duplicated also in the source repository and thus the duplication problem didnt show up for the backup +This file adds Java security to cover +Im working on it + +Hi Eva can you please verify if this has been included and resolve? Thanks +The only plugin installed in eclipse by default which seems to allow downloadign the runtime is tomcat and tomcat is using an extension point which is deprecated +It would likely be one step out of sync with a working config unless we were also to adopt as a test platformA Setup Page on the Wiki is a possible compromise since it can be maintained without requiring synchronization with the releasesAnother question for ChristianHave you had any issues with the use of other paths within the exploded? Like uploading? Or updating the blacklist file? Or have you specified an alternate upload directory outside the +Only fixing on trunk as is a bit different in trunk all tickets +Would be great if you could verify +Thanks for finding that potential NPE Knut +I think deploy tool should throw an error when trying to deploy to an existing configurationredeploy command should be used so user is sure what hes doingmoreover undeploy and deploy should do it toobut should never redeploy an application IMHODario Andrade +Tomas can you share your definitionIt should work without needing that to be explicitly defined +Can see anything that might cause a problem +Filing against for now +I just committed this +This doesnt work well in most tooling environments like where forge is Forge XML parser is very capable IMO +This issue is superseded by. +Applied +Only change is that ant prints a descriptive error to user if does not exist in top level lib directory +Vitali please set a specific Fix Versions +Hi HansI added a fix to the trunk which may have fixed this +I could go either way but technically it is a backcompat break and it would be easier to just leave itIll just push it off for nowIm just trying to resolvepush issues didnt mean to single you out for any specific work +This is Disabling XML reporting might help +Thank you for the patch Subramanian + +We just have to look at everything put the priority on making life best for users while balancing somewhat with devmaintenance headaches and come to a consensus easy +Yes I miss to. +v + +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +If you find further issues please create a new jira and link it to this one. +This is a bit out of the scope of this issue but will be useful for taking bug reports with the VTI +When the namenode replication engine triggers of a block log it +this would allow an componet to load its setup into the db without programmingI also think that this should be the basic code should be in the frameworkcommonand the component specific should be in each components seed data to include the screens each component would have a ECA and SECA for setup as neededeach component would have a webappsetupany thoughtsnote the migration would also use this setup controllercopy of email on dev list +Thanks Namit! +Id like to get it in for let me swing around and see what works remaining +works and its consistent with the name of the analyzerIf i svn move the files in my local and submit a patch will it ensure that history is preserved? I am not an svn expert +The problem is hidden when there is exactly one delimiter character after each valueThe patch removes the completely and uses the length of the string to adjust offsets for fields with a string value +will be done on needed basisA subtask will be opened for that. +I think someone needs to do svn propset eol +This is completed though I need to check with Lisa to go over what requirements has when logging in a user +Reopened the wrong issue +Recommend same changes be made to EAP +When we upgrade to a newer version of a dependent project we take the heat and do the necessary changesWe also dont necessarily need to depend on the latest of the projects we are using but on the latest stable as much as possible +I think we should discuss the requirement for populating contributors +AgreeAndriy. +Atomic operations on Linux using APR appeared to be using mutexes instead of assembler +Out of curiosity did you also tried saaj impl +Thanks for the quick fix! users on X were having lots of trouble with the because of path problems +I will take a look at soonThanksve committed this and merged to +Saving porting to trunk until the XTS tests are moved into the quickstarts project. +As Garry pointed out added all necessary comment directly in the configuration file +I think theres a possible bug in the equals method it only takes account of the file field and ignores the user and fileName fields +Closed scanner in a finally Applied patch to r Merge into trunkThanks for the patch Dave. +Also added groovy validation script for +patch includes upgrade to as is not completely synced +Considering that youd have to implement an EL function with the hack for JSF I think a component provides a much cleaner solutionIts also easiersafer to install a new component than it is to install a new resolver +Does this issue still make sense? What are the main differences between the patch in and the one in +Remember to run Java with a heap of less than MB +Hm seems to be a problem reading the LDAP schema +To that effect I propose that these modifications be done via exclusively and where necessary things be parameterized via environment variables +We need a new version of the demo that runs on AS +Perhaps its just me that use ngrams for something completly different than what everybody else does but here we go Adding the feature as suggested by this patch is according to me to fix symptoms from bad use of character ngramsBOL EOL whitespace and punctuation are all valid parts of character ngrams than can increase precisionrecall quite a bit +Would make sense once u can load to an existing table +Ive been concentrating on getting into production so I havent tried this out only seen the screenshot and read the discussionTwo questions a Is this suitable for inclusion in? b Can the XML display code be made smart enough to deal with xinclude? My file is mostly made up of xincludes +The failures in are because of this +Nevermind +Checked this into trunk and +Curious why this would be closed as deferred instead of left open and set to a future release +Sending trunkjbiclusterconfigsrcmainresourcesblueprintSending trunkjbiclusterengineSending trunkjbiclusterrequestorSending trunkjbiclusterrequestorsrcmainjavaorgapacheservicemixjbiclusterrequestorSending trunkTransmitting file dataCommitted revision Sending trunkassemblysrcmaindistributiontextetcSending trunkTransmitting file dataCommitted revision . +Create New Product Content if we select additional image works but no for Image Additional View X Size TypeWith the first case your additional image file is automatically scaled to smallmediumlargedetail size related to the But the second case with Image Additional View X Size Type the user interface Edit product content isnt adapted to upload an image fileSo we have to improve this case + +This isnt an error in the documentation but it might be more explicit if stop is true an unrecognized token stops the parsing and the remaining tokens are added as extra arguments +Hi GuillaumeFirst of all thank you for the patch +Ill post the announcement on the plugin list when ready +On a more philosophical level it is a bit curious to me that if we have the strings abcde fghi that we are fine with b being at position and not at position but ab needs to be at position +But NO need to hurry +I moved this up to and Ill take a look at it today +Commited to and trunk. +Closing issue out as part of JIRA cleanup +It occurs with the EAP version as well +it does a pretty good job explaining the root issue hereThanks how I should go about making my application immune to these problemsDont share Entities across threads that are associated with an active context +What do you think? it seems were back to where we started from +Thanks +maxSizeminSize deprecated only for numeric properties +Committed to sqoop +patch which also change the URL to point to latest trunk packages +class IAE for call site duplicate class missing constructor +I did basic scratch testing +Documentation on their use is in the javadocs of the newfileThese enhancements badly need testing +That is why we need to know them to reproduce the issue +Now that work is well underway with AS all previous community releases are +There are some backports here of new test code in etc required for the tests to run +updated +The patch is correctWe should revisit this issue for JDO to see if it makes sense for the checks to be added if the implementation is or if another flag would be suitableIf there are no other comments Ill check in the patch +Ive attached a test case for this issue to +Closing as there has been no feedback from the reporter for significant time +Deprecating might be best since is going to releases soon and has the fully maintained version of the application +Hi Prakash would you like to review this one +Can you reproduce on +Now I have replaced with +Okay +When connecting to an existing database authenticate using the credentials stored in its SYSUSERS table +NP! +Verified pls close just needed a as requested. +The person requesting the transfer can log into their account at this registrar and view the status of domain transfer orders +Just checking no strong opinion either way +I am content that at least a legit problem was uncovered in the process +No interpretation +I think that since this extension is for tests then it should by default allow only classes reloading +upgrading to I noticed the exec sink was taking much longer to cat a named pipe +Just testing small mod for this will commit back in committed will close in few days if no feedback +Also please upgrade to the most recent Hibernate release +Really appreciate itI think the pach process had some problem in applying and merging this patch with trunk +Closing this as I believe is a superset and I already applied it to svn +Have manually tested this patch +If not Monday should be alright +Im not clear if I need to do anything with this JIRA for EAP +bulk defer to defer to +For security reasons and others Hadoop should not add another networking stack +I clearly see the weird behaviour inthe logs though +First implementation of holding foreign apply this patch you have to have the castor up to date +Thanks +I wrote a using the +This can plausibly be related to timeout handling but there is no way of telling that for sure without seeing a log of the session that exhibits the problemOleg +Patch is too big to include at this stage safelyAfaics it never respected the temp deploy folder anyway so its not even a regression +Then we can hopefully give you the go ahead to make the git repos rw +Thanks for the clarification James +Lets just add reads to for now +Further resolution of this issue to be handled in +So the question is whether it makes sense to have to have this convenience method even if we cannot guarantee consistency? +This was working on you confirm that you have replaces the etc with the etc? That is is required for it to work on Karaf +but why not keep it simpleI guess people dont put together the context of hostname and the page content +not an issue with the latest version of the plugin. +Now that work is well underway with AS all previous community releases are +The patch doesnt appear to include any new or modified test +committed +Still it would be nice to get these javadoc warnings fixed! +Finally got some numbers to share from a running testmax topN Selector reduce output records happenThe above looks fine +Committed to trunk and +Please let us know when you determine the exact cause +It would help those of us not familiar with MIDP to get some explanation of whatmakes not work today andor some possible suggestions for howto fix that +officially the code freeze is today +From a design perspective on the Screen Widget I have concerns about this change so please no one commit it until there is time to discuss itAdrian has expressed his opinion on it which is great and even in the form of a patch + to add the right Fix Version +Id like to commit this soon unless there are any objections +Still not sure why Jenkins isnt failing patches because of thisThe changes are fine now +Committed possible solution for this issue in revisions and +What are your concerns when it comes to removing deleted in a minor compaction they are still going to be removed eventually and there is currently now way to undo your delete to get them back so the way I see it they are just a burden for the system +Think how many robots are out there and think how such behaviour would impact human users visiting the overloaded server +will do +make sure our ID numbers are large so we dont conflict with Hyperics numbers as they add more +something similar applies to modulesintegration the issue title and environment field of the issue to make it clear that the issue is specific to Java . +The idea of having it on sounds good to +Without it there are no headers generated +I dont agree on committing it +The default implementation of the new session handling strategy which is responsible for migrating the attributes has an additional list of attribute names which it will retain even if migrate is set to false +I also have created the keystore and imported it into keystore +I committed a patch for this +Thanks Gaojinchao +it looks like we could have an interface in called get which returns a used by this ledger manager to do garbage collecting +tests are still running now +After updating to the latest native build I can no longer reproduce the seg think its safe to consider it resolved +Since the element is optional in both schemas the new schema could be modified slightly so that its element has a different name +Moving to CR +There is a properties file that is created as part of the perform process but the descriptor should be produced always and should be the authority for performing the release +It does look good to me but Id also like for at least one or two more core devs to look at it +Assume closed as resolved and released +Thanks Chuan seems like a simple fix +The the unit tests for the patchTo demonstrate the original bug use this unit test +Please add more information a bit more of the code of that appears to be failing and the code that contributes your filter into the service configurationWhat youve shown me should would but its clear from the stack trace that the code is not returning and is dropping into the standard page rendering logic +I think its the best compromise between developer easy of use and addressing security concerns for a default M Stark wrote For whatever reason our long standing use of unsecured consoles is now being reported as a security hole +See in for an example of the pattern thats supported +Hi TomWe currently have a cycle configuration per user so we load that with theuser name +Verified on JBDS +It works by this fix +to the new package needs to be resolved first +Grahame in your tester did you mix jgroups and +Added license and notice details for EXSLT stylesheets in rat mainwebappresourcesstylesheetsexslt +So the is definitely at least part of whats causing the movement of Filter in this caseSo I need to fix up and and I need to fix my test +Thanks for buildr I use it every day +We need to provide a new for do you want to to write a patch fordo you have test script with continuum +If you are able to supply ant scripts for which reproduce all the existing maven functionality I am sure there will be a great deal of interest +Thank you Drew for commiting +This doesnt seem to make sense +Hi Vincenzothanks for the detailed description you didnt anything wrong +We are exploring our options here and I was wondering if it is valid to upgrade to but leave all other jars at +Thanks Bart +It also seems too late in the release cycle to try and abstract away from a disk index +yes that is correct +So if the IOE that we capture is fatal then when we try again we will get the fatal exception again and the job will fail +Patch v corrects spelling and wraps long line +I am not sure if this is the semantics that we want to have for situation where data doesnt match with the schema provided +It is not optional to implement because core functionality depends on it +I set that issue as critical +If I can get some additional info on this one so I can reproduce it Ill try and get it into the final build but without more info on reproduction steps Im kind of stuck + +Perhaps something in the +I got a reply from Azariah Jeyakumar indicating that it was possible to use the Servlet filtering capabilities to automatically correct the charset indicated in the POST request +A schema name must be supplied if the user has access to other schemas of the database otherwise the thema might be empty +subversion history shows this was applied on Sep th by aborley +Thanks for the feedback DanI have implemented changes based on Dans feedback and am now resolving this issue +It worked fine +I would also patch it for the future in either case +I also had to substitute with cdvjk in the +o files in plugin runtime context so that nativelink can figure outhow to link them document this issue +First problem the Rules for the tag handler have a namechange in their instance vars it is now changed from Instance to INSTANCESecond problem orgapachemyfacessharedimplutil does not exist under that name anymore needs to be fixed +Im not sure what you really can do about this in the framework +In the attached archive theres also a Ruby wrapper for the class +Pushed to Grails Repo. +Thanks +Will try to fix +After this error i am not able to run any workflow +Thanks Tim for your comments +Ill post more when I hear back from IBM +Ill link this issue accordingly +We should provide an overridable directory locationThat will imply modifying the classpath specified in the generated script +I think this is no more a problem. +I thought the problem is that something is not recompiled +We will do minimal manual testing with IE FF and skip rest of browsers +I will refactor the JCR component in order to use bodyheaders and submit another patch. +Ill try to take a look in the next few days at my hotel +It implies cascaded DROP SCHEMA +Attached a trivial patch to the POM for to update the version +Im closing this issue as Not A Problem because there is a working solution. +It should instead be something like a trailing time window or better a trailing ratio of successes vs exceptions +The number of returned blocks is limited by the total size +I think this makes sense and the patch looks good +Can you please give a link to the JIRA about class if such does exist +I can remove the release from the defect and you can try this out in or trunkRegardsKiran +Shouldnt this issue be resolved +Committed to TRUNK +includes this patch +From todays IRC discussion +Tests pass and committed to trunk +Will fix whatever failures in subsequent patch updates +Richard It works! I close the issue +Not a bug of of a bulk update all resolutions changed to done please review history to original resolution type +ramWhat doubt do you have about my patch vI update the test case to verify MVCC for scanners after majorCompaction +Otherwise I run out of memory +Pack file which triggers this error in my workspace +could someone using ROP actively take over this issue +bridge supports the latest seam security changes after seam +And what you are saying is exactly what I mean +The others are working sometimes but mostly failing I suspect they might all be related to File list sort order but I have only confirmed it with one of them so far as mentioned aboveThere are a few options as far as I can see commit the udf and open new ticket for the other three tests or try to fix all the tests in this ticket +If the issue has been fixed I think they should be removed from the exclude list. +Ill have to dig a little more + is fixing multiple issues +AnkitYour patch is in trunk at r at r +It includes documentation for afnavigationPanebuttons afnavigationPanechoice and afnavigationPanelist +Keeping here as a back reference if this weird issue is causing more problems. +Well there were also warnings from other build +In other words the XMLCTF now runs fine again +Sandeshas in the same state +I just did it because to me it made the code easier to readpriv is called twice so it seemed better to have one method than duplicate the code +Thanks Charles! +Patch applied as provided +If your collectors are heavily loaded it makes sense to pull this functionality onto the HICC server +Ive created an example project +by the way can you also provide a an url with which i can reproduce? I also didnt come up with a reproducable url for +Some guys leveraged some geotools code for Informix +Thanks Sijie Ivan for the reply +I changed the default recommendation to RELEASE in case restrict does not find a valid version +I can live with BTQ printing match although it actually just hurt a certain docIs it somewhat strange for BQ to count a negative score contributer for its min condition +was solved by recent changes to the code base +Not getting fixed for +Proposed patch to +user +Does provide this functionality +I just committed this +Example maven and eclipse project for reproducing the example project with source code +Yeah ok imprecise word choice +Hitesh there are that you added in you intending to fix themThanks Vinod for verifying +How could I change it by myself after creating an issueThanks dont have enough rights to change a jira after its creation only developers m not sure if this will help but have you tried using? See and how it uses that class for non square see what you mean now +Verified atttached repro demonstrates the problem. +It looked like the tests were testing that code so I removed those as well +RI and vms were added +Please verify if this is fixed and close it or bounce it back to me if not +Change applied in r Thanks +Currently all classes in B have all As classes and dependencies available at runtime guaranteed by virtue of being in the same WAR but they are not available at compile time +agreed +This is the upstream directory structure of tomcat it contains serverlib and commonlib since the zip is close to the community release theres nothing we can do about it unless it is changed upstream +Ive attached the project +assigning to Denis so he sees I forget what reasons we have now to run this job when jenkins can accumulate the results you recall +This patch should support the specific generic and reflection Avro types +I will update or enter jiras as needed +o Could you describe what tests you ran to verify that the jar files are correctly and that the corresonding regression tests continue to test what they used to testThanks! KnutIve already tried altering the main heap space as well as the permgen space in various different configurations up to M of normal heap as well as an equal amount of permgen now Ive retried it after uninstalling openjdk then uninstalling then reinstalling so unless its the OS then it could only be a very unusual bugIve tried junit in case now Ill try your precise memory settings Id guess Im looking at reinstalling the OS now though funny as I did so very recently progress is relatively slow given the time takesDBest regards patch available as there are outstanding questions on the patch. +It has a lot more functionality +The only UDAF that I know of that does this is count +committed to and trunk +Going to fix +Thanks Ashutosh for updating golden files +In the pom file they look similar but they are not treated the same way during the generation of the for the comment so i have to report this to the springsource guys that they fix their Manifest Parser. +For access inside of the server the expected mechanism of accessing mbeans is to use the directly +Authorize is testing authorization system before +If we do svn co ant test instead of svn co ant jar ant test test introduced in the patch fails because test checks for which is built only in jar target not in test target +Yes +Just noticed +Please let me know if there are any other style issues +Doh! Updated patch +I also upgraded my project to grail +Fixed the new test to run correctly every time +I will probably file a new JIRA to think more deeply about what the interface should look like +Could you please try to reinstall in a fresh directory and make absolutely sure that you have +Thanks to Hazem for provide us this patchfixed at revision +This patch contains the fix on this issue +Thanks now Ive got it especially the point that thread functions indeed may be residing not in a making the whole system much more flexibleI am all for committing this patch +The libraries in the repository should not contain the version numbers in the filenames so I corrected fact I see the same mistake is made in the version but I didnt correct this there since Im not sure where else this is already used +This patch works thanks +It has the same finalizer behavior it just logs where the that have to be be cleaned up by finalizer were allocated. +Also in okayThe site does validate okay with WC Validator. +They contain the old docbook xml docs +fedora w Chrome andor Firefox +Ok +Solved r +Just want to check to see if anything has been done on this issue before its anniversary comes and goes +ill do this on fore the on trunk since revision +Applied the patch as I was committing another one to get all tests to pass +Example original version of this has some issues I made a few changes and will upload the patch later this week +should in fact read The motivation for is that the specification does NOT allow a +Also should not be static since it isnt in +This is not specific to optimitic locking though same with pess we be doing this? Still debatable as per forum discussion +You check out build and run +This patch fixes the issue +Hopefully your patch will be applied soon +Thanks for the detailed explanation of the problem and a workaround +Id like some feedback on it since it adds a catalog reference and a query to the catalog on each rendering of a freemarker template +Thank you! +Next would be a question Which way is the cheapest to transfer that information from to application levelPlease note these considerations are purely from transport layer perspective +So it can test soft upgrade from various release is working +Later as we gain more experience with this feature we should on monitoring by default so as not to waste cluster resources because of programming errors +Ive just forked the project if I make progress Ill send a pull request +Fixed in +Good to know from my tests on web site contents all seems Ok Ill ask even to other developers if something is missing and post an update here if necessary +DoneYou can login using the same key details as do ssh reinhard and use the same password as on but still cant login +Noel your problem is similar to and not related to this issue. +Verified on the user provided sample code +It would be a lot more useful if you presented that as a diff from the existing mapping +It works also with BaptisteThanks for reporting this issue +I havent seen this lately in the nightlies +Committed fix to branch +Sorry the proper support of the completion of import is too much work for me +comments and attached latest patch +Must be merged manually on the this issue on the failing in case of activatesuspend +Test case problem is reproducible on the stock multicore example configure multicores to replicate after startup and commit launch multicore example index some example docs ask for commits httplocalhostsolrcorereplicationcommandcommits reload core httplocalhostsolradmincoresactionRELOADcorecore ask for commits again httplocalhostsolrcorereplicationcommandcommits observe no commits being returned after reload +I mean how should we fix a bug that we can not reproduce +What should happen is that if they are idle for a while the resources get freed and the connection goes into a dormant state but for the revised Solr connector I dont think that is happening +This should be resolved by the fix made for +Currently its just for legal reasons we need a record of submission that is in our systems github is really not +Will commit momentarily +Each bundle imports the transaction manager but due to the missing unwrapping I ended up with multiple entity managers opened in the same transaction and only one of those getting flushed +NMR distribution file is already goodGuillaume investigates how to fix the issue in Karaf in a better way +We need more information +Please check and let me know + +bq +committed v delta +Addressed the issues as desc +Resources source classpath entries now point to proper targetclasses pr target output folders +Btw is the name case significant here? If so we wewant it to be? Thanks +Check runtime list in preferences +It will be resolved the mistake with this patch +Thanks Ill take a look +patch has been applied +I have attached a patch that fixes this issue + +Cdrik Please feel free to provide a diff patch +Are you starting with an clean workspace +Please reopen if you still see the the issue now that it has been resolved. +As in the docs youll need to set the for the queue to allow a user to go over +I also refactored the exception handling so that you could write a handler for any thrown exception +Ive just committed this to trunk andThanks a lot for the contribution Ahmed +create keyspace testcassandrause testcassandracreate column family test with columntype Super and comparator create column family test with columntype Standard and comparator +might help solve some part of this. +pending ve committed this and merged to +i think ted is right this is about how we detect time +This initial text is just used for a bit of parsing and to find out what the supertype etc is +The onlysolution I see is wrapping the Access Denied exception into a but I dont know if this does really buy you anything +Probably the simplest change is to have a seperate list of urls of top level directories that contain unpackaged suffixless deployments +On the next tier you have things like DBCP Configuration that depend upon other Commons +Moved to we wont release a think this was fixed last year when the operations were reviewed for speeding them has been releaseClosing all related resolved issues. + +Ill go ahead and make the change to our jira source so it gets picked up in the next deploy +the fix works as expected +Verified the integration in latest code working as expected. +on ve committed this +Verified on. +Model is also affected +Aligning multithreading configuration through Hints as on trunk as revision as revision closing all issues in resolved state that have not been reopened nor commented over in the last month +Wait Im looking at the examples right now and Im seeing no line breaks or spaces involved with the properties in the value tags +This looks good! Thanks Tony +And then wrap those policies with a custom policy so we get the callback from the JDK when the task is rejectedThe discard would possible not be needed as I would assume the thread pool will reject it asap when you try to submit it +We can ask the upstream jython developers to produce a jython minimal jar alongside which doesnt include the extra classes and get this deployed to the central Maven repo +Targeting the issue to to get more visibility for the next release +yesits not a bugwhere can i get the information regarding this. +The solution is to simply resend the messages in the until the is empty +Thanks Michal! + for making Hadoop the default build profile for +Thanks for testing! just to set the resolution to fixed +Add to its own Condition column that is NOT shared with +I am posting the issues to the dev list +I somehow didnt read that the test fails on +Will test it am not able to replicate this in trunk neither with the attached project nor with a new project of similar structure +xsl is not lost +I just committed this. +There was a javadoc warning +this is +spaces to indent the codeIn addition to Jons suggestion for fixing up old files we should also allow white space changes to a limited extend in any patch +JNA supports as of +Possible fix attachedNo yet +Will commit when tests pass +It means that the Apache Commons http client is attempting to retry a request +Committed to main with revision Intentionally not closing as I need to commit this to the CP branch as well but holding off until I check with Kevin that it ok to do so. +User reported that with troubleshooting it appears to be related to container not Jackson. +Zookeeper is available and claims to be backward compatible +Stack your patch for this broke trunk +pushing out Im not following what the real problem is here but it doesnt sound like its something that will be fixed by R release +JIRA is not a support forumPlease subscribe to the Commons user list and ask there + to +So you can reuse the message object in your vertex compute method +x if possible +Dale brought up the point that not including the directory path breaks tab completion +I am scheduling this bug for and +There is an issue in which the navigation state from the url was only being set on a render url and not on an action or resource url +We just ran across this issue on another project today +Thanks. +Setting to resolution state before reopening to change affects. +part of release. +Upgrading patch to allow to run the same jar code for running Slive +Protobuf marshaller prototype using output file and matching sent protobuf class name to equivalent java class name +Volume migration is not supported on KVM yet +Verified with JBDS +Needs my fixes to XSD importing to workI am still extending the code but let me know if you want to add it to the samples +So I found a solution online thatshowed me how to use the +The change to makes the exit from jruby a little more quiet d tells cmd that it should Disable execution of commands from registry +Can you say whether your application might be experiencing thread interruptsThanks issues here all have the feel of the database getting broken sometime in past +A spec failure and not a real app error +KEYS restored from last version and moved to the project under docs +Any reason why weve made these classes final anyway? I can remove the finalkeyword from these classes but itd be nice to know if there was a real reasonfor them being final +If it becomes reproducible then I will try to fix it again +Thus the compiler tests if it is a class at compile time and if it is not then the compiler assumes it is some kind of property +ve created the equivalent patch for +Thanks! +One more task for late night action hero upgrade to Maven to get the release candidate going. +Using removes need for several +Strangely the quartz jar IS on the projects classpath. +Yes I understand that and it still doesnt change the issue hereThe paper here deals with a data set with no ratings picking any item as test data is as good as the next +Nice style youve got there for writing docs +I think were good +This GSG example uses the example in the EDS kit +Based on the exhaustive information I provided by now I restored the priority to Critical as that priority is in line with the nature of this problem +I for one only have access to a handfulofBill +applied. +We have a design for RM restart that can be completed post beta +Will be completed after he joins the team fulltime +If a new message get enqueue in the queue the old message get dispatched and the new one stuck in queue +I really hate to see exceptions used for control flow but I can see your need +I do think that even for solr you need be able to configure and before startup +I understand the where an item may be composed of several elements I have also ran into it at clients and the samples jobs illustrate it one way as well +Ping can we have a PMC reply please +I am on the patch but I do have a question did the patch fix your issue as the unit tests wont test this format +Hi MichealYea The fix seems very straight forward +When the is passed by the client we are not closing the same +Many thanks for the patch +ok missed the partBut anyhow that code is completely valid and I dont get what the error is from your descriptiona runnable testcase is the only way to prove that this is a bug in hibernate +We dont have any details as yet but it may be something we need to investigate at some point +Note that the plugin already adds web properties to GWT services such as request and wasnt watching this for a while but looks like implementation went quite wrong like i just add security features which present in controllers as well just like many other features which are added to controllers in third party it quack like a duck and call it a goose. +For which use cases does the user need to modify how the scoring is calculated? Id like to get some concrete examples in the doc for each feature to grasp them more easily +Closing old bug reports with no updates the last you feel this close was unwarranted then please reopen or create a new one with link to the old and we will revisit +Could someone give this issue a feedback? Its a pretty old issue by now and we have no idea if something has been done to address this performace issue +is to the trunk and branchThe trunkSending modulessrcmainschemaTransmitting file dataCommitted revision The branchSending modulessrcmainschemaTransmitting file dataCommitted revision +fixed +After creating a patch I found it was already posted hereThe attached patch by Dimitri solves the issueKind regardsSebastian Dimitri your patch is in trunk Rev +One file modification required by issues so the patch is combined +Which methods exactly are you talking about +Not high priority to get fixed. +use to slice a hundreth off a region count rows in the th and then make estimate on table row think we should be storing statistics about in the new meta blocks +It is not necessarily the case that an order and an ivnoice must be associated with one and only one billing account +Hi SumitThanks for your inputBut we cant removed status valid change for ITEMCOMPLETED to ITEMAPPROVED as this is valid status change +For we only support what you call a persistent clientspec +Logged In YES useridIm seeing that there request username is not found in thedb prior to the ejb exception No matchingusername found in you put together a sample ear that demonstrates thelogin action here +Got it +Unless someone would like additional changes Ill plan to push it in tomorrow +My is bowditchchris +It seems that the VM hangs loading some native library +I just committed this to trunk and +Thanks for pointing that out +The just takes a desired size and gives you exactly that manyMerging them makes sense from the point of view of a user but there is little in common between the implementations +Closing all issues created in and before which are marked resolved +Forgot to mention that the cannot be updated by the user +My only questions about the would bea How would a user trip across thisb What would the user seeThanks for including a test case for this exception +I dont know why is in libtest +Applied +Here is a subset of my data +Yeah we used to have some circular dependencies that meant we really did need the laziness +Refactored patch +The disabled links doesnt look right +Thanks Rodrigo I will commit this patch +In a different jira +You document what you consider to be acceptable values and throw an exception when the client supplies something else +This was regression due to dependency exclusions not being added as a collection. +Ok thanks I will add myself and try it out +Property accepts the testPreprocessor configuration string and uses the same format as explicit configuration +The same patch is applicable also for trinidad +it should have extended and forwarded the ejbgenerics values to the +The annotation processor does generate the meta classes with JDK despite logging a warning the warning is just rather confusing to end usersNow that Java is officially at end of life should not complain about Java sources and as Piotr said I suppose its enough to simply upgrade the +Looks good +Ok +Remove msgs that show on the console a list of active threads when the service is dumping active threads to to continue on CPC error handling to continue on error +I think we should use a different versioning scheme than date stamps though +Couldnt this lead to a quite chaotic situations with some images having thumbnails and other not depending only on whether a user with write permissions happened to request a page containing themRather than working with on Id rather suggest to implemenet a service to generate mising alternative representations for existing resources if needed with some more expressive way to specify which resources need which alternative representations +Thanks for the correct line numbers! My last comment I deleted was wrongIn the current stable branch there is an update related to that maybe take the current stable branch +axis now supports multi dimentional arrays. +I cant figure out how to delete the old one +Sushanth thanks for informing +Closing as. +AndyThe class used the property names directly instead of the constants from the above classThe new class which sets the read and write buffer is not using property names either and they are not configurable as wellI have some unfinished work for the java client configuration and will fix the above as part of it +Allen the solution is bullet proof and +Im against quick commits like this +It has a similar test for clob where it extracts two reader and compares them + +Now since the does insert a special propertyquote What File Install does though is adding a special property to the newly created configuration which allows the configuration to be related back to the cfg file +Reintroduced missing method and enhanced in Rev +I ran manually +Filed Im still looking into other ways of doing it +About not needing a set of already seen bindings This would only be true if all variables in the Select clause were also part of the Order By +Stack and I talked a lot about it heres what we came up with +Thanks for reporting this and submitting the test case. +Anything in MUST run in the main thread + +I have opened to track my changes +Does anybody know how to fix it +Hi Sebastien Im sorry but we had a discussion with Evgeny and we decided to not create the git repository +The response message is complete bogus which makes me wonder if something is happening on the server side that is causing the bogus response +We would need to have some knowledge of the fact that it is a numeric field and what precision it is +Ready for review. +found this issue in amrita HTML template library +It happens with all of us. +The spec does not enforce that so there is no way for HEM to have access to the Session beans +Closing due to lack of response from reporter. +If I recall correctly we try to unwrap the JDK proxy and look at the target class but we cant do that with keeping the params in sync and please report this sounds like a case we should document +Sorry +First I need to understand the scope of the changes +Updated patch run through patch +Verified in L +These classes solve most if not all issues related to older xxx classes Any arbitrary action class can handle several events +We have known and easy to use workarounds +assigned to Rob since this is a server adapter issue +If you use maven under windows you must patch because it doesnt set the error closed by am using Maven on Windows XP but my does appear to contain the logic to exit with the same exit code as the Maven java you point me to the patch you mentionedThanks think if you have upgraded your maven you shouldnt have the not please reopen the issue. +Will follow up once I have reviewed your code. +Will add a test and test applied to master and +Thanks +could we fix this +Fixed in r +Committed to trunk and. +Looks like the commit number is r where was fixed also. +I think my brain has started to take Fridays as a holidayI meant it reduce the of the output PDF by cached pages and their resources if the same page is embedded multiple times within a PDF. +I hope I have not screwed up this time +I have committed +Submit a Certificate Request through web brower using Request Certificate link +All done here now. +I see this failure for other patches too. +Precisely +It will be part of but that is blocked by this. +Another improvement would be to accept a Comparator as an optional second argumentI wrote this quite awhile ago and I dont have time to do the fixes anymore so you can do whatever you want with this +Thanks. +Im going to commit this now +But I am not saying that it wont be in a if we can get an implementation it will be welcomed to be integrated. +Seems related to datagrid interactions + tested +NathanWe also need to fix the jbossws file in etcjbossws that starts up the web needs to +Hi we dont use more modern casts for portability reasons +Thanks Nicholas +Attaching patches for review +I fixed them all. +java files in the grails project +resolves nesting problem with nested collapse tagsReported space issues seems not to be reproducible +Would be nice! I try to get trindad running with +Its still a bit light and may need to be updated in the future though +We can then load this into the test instance for you +To me this makes it easier to manage keytabs and principals +I do not have a way to test on the Mac version of Safari +Verified on ER +Ok so if I explicitly add a classpath entry it should be added to the launch configuration? If so I am not seeing this +One question from an MR nb in MR these configs were very error prone since different jobs took up wildly variant amounts of RAM +Ill update this issue but we may need to push this one off to when Cursors make their debutPostpone yet again +Your patch partially undoes the modifications introduced to solve and will cause a regression +This looks good and passes test on +This should fix the guys were right thanks very much for reporting this +OK Ill try it as is though probably name it something other than newstyle as two years from now when someone wants to add a third style they will not be so impressed +is just too static in its current state +Oleg +Im working on adding a test for this case to the end to end tests +Additionally as I said the number of lines where this happens is very limited +I fixed to fit the current directory structure +Im at so it may slip a day or two +In your nbm project you should have only the dependencies on other projects that your project needs for compilation that are explicitly used by the code in the project +Ok Ill bring the discussion about the default +I rebased it successfully the other night but held off committing until I tested it a bit more +This one is pretty smalluncontroversial and is now committed +May beClears the time portion in the self parameter and return his +Closing out old issue. +Close all resolved issues for Engine release. +Aslak hows this for the bald spot +The directory under webappssonardeploygwt contains lots of stuff see the attached let let me know if you are looking for some specific stuff +Attached a patch for the velocity artifact processor that uses weak references for its internal caches +johnathanmeehans solution fixed my problem +Hence Priority is updated from Blocker to CriticalThanks +Thanks Knut I agree +Nice the OTP validation issue has been solved lets move the code into an official AG repo at demo was implemented and moved to the official AG repo under the name +Joan dont close an issue until the code has been committedThis looks useful its always bugged me that the list wasnt sorted +Thanks YakovDo you think we should consider this fix for? It seems very safe and straight forward and we have time for simple fixes like this I think +This has a couple of other positive side effects for the overall bean container will be available in tonights M snapshot +Fixed with patch by Ken Glidden that merges a single diacritic text chunk into the previous text chunk if they overlap +Defer to all open issues to all unresolved issues from to + +This issue has been fixed and released as part of release +Verified on ER +We can probably close this issue for now +Mass closing all issues that have been in resolved state for months or more without any feedback or update + combine those methods to delete +Initially I did in bsh because I am working on Asset maintenance application in which users need only the list of facilities of type Warehouse +Thus I close the issue +Thanks for that functionalityIn the attachment Search Analytics there is a broken link This shows how to add the functionality yourselfWhere can I find that oneAlso the link to power point earlier in that pdf is broken +Alexey how to reproduce this? Just creating the EAR app and run it in JBDS does not seem to cause this error +We need to reimplement something on the lines of which we used to provide with our earlier xlib releases and ll be working on this along with the autoconf distribution making work that is currently going on +Lets also rename DATANODEID to DATANODEUUID in +I could upload the patch if no one else is doing +This bug has been fixed in Xalan Java +I forgot about this JIRA and opened a new one +It just dawned upon me that this limitation of Within finding could be remedied right now without waiting for the faster approach in +fixed at for this also take into account the deprecated tag in and include the associated comments +bq +Resolved by +Close all resolved issues for Engine release. +The txn log only contains accepted txns +Unassigned so someone else can pick it up +Im wondering if users are going to interpret CPU to be a hardware core and Im not sure a vcore will map to a hardware core in the typical case + +Regarding and and new issue vs adding to this one do whatever is going to be easiest to assist you +Updated patch which includes symlinks for YARN modules +running patch through hudson +Hi Oleg can u also verify that the current resolves the restart for youWe aim to have complete by the end of July. +AmilaI think you are the only one who can handle this +I close this issue and open +The is identical to except Goal changed to altcompile A parameter suffix added to allow an easy definition of the file endings to compile Parameter compile renamed to sources and made editable Parameter outputDirectory made is identical to except Goal changed to altjar Parameter outputDirectory renamed to sourceDirectory made editable and given no build section from a POM which shows how to use the new think these should possibly be a separate plugin but am not sure +On a hunch I suspected something might not be getting cleaned up properly from the transaction for the DROP PROCEDURE statement +In my location it works compiling manually +Closed upon release of to. +I think this patch is awesome +Attaching a patch for with some bug fixes +enough information to resolve this issue. +Sounds like good improvements you suggest +Dont try to get for fragments Use the implementation bundle context instead of the s bundle context to track services Downgrade Core Dependency to after release +Push And Commit was removed since was extended to offer this +No idea if that patch was applied or not but all I know after checking out on Jul is that api build fails +Base conversion to junit is complete and the old test has been removed. +New patch addresses the issue +This would allow users to hook these up to kerberos credentials also if they want to +Of course it means that the http buses are deployed in the servers web container so you dont configure the host and port +add patch for patch for except v with minor +I closed this because I didnt think I would be able to resolve the image problems this out that I could +The default Similarity implementation is named because itwill replicate Java Lucenes scoring model +This patch adds some members to the group of elements which contain the src attribute as well as removing all whitespaces and non visible characters such as tab n etc from extracted microdata +Do you have time to also get this change into the release +Ive added some scenario tests which use the skip and take operators to control the result set +The old container has a couple of problems in this regard +requested Robert +The JDT preferences are confusing +our goal is to have a standard wire protocol for Avro RPC so that clients and services in different languages and from different institutions can interact +bugthanksve done some more testing using the generated code as you did and that did not reproduce the problemThe generated stub has code that correctly serializes and then parses the xmlThat explains why you see the correct request responseSo in order to reproduce the problem that I am seeing you need to use manually coded StubPlease use the attached srcThanksSathija +Coverage has been fully disabled from this job +Its clear that the existing hierarchy scheme would be insufficient to handle the case you mention since currently name resolution in Spring only progresses upwards to the root therefore if bean X has visibility to your dataSource then any bean Y which has visibility to X can also see the dataSource +I encountered this because I wanted to test in a cluster and I wanted the to be viewfsIm going to commit this momentarily +Yes i have replaced the jar +Other clients wouldnt remove the new connection handed out to the first client +v here includes cqlversion as well as cql and cql and fixes the most glaring problems with cqlsh use with cql +Will open a new Jira when I can identify what is going on here. +Currently the grammar accepts conjunctions AND and OR +GUI is added to editaddremove import chapters language is the screens and the content a bit are updated. +is replaced by and and these factories can be used by tomcat. +Fix available on r +Storm is a stream processing system +Thanks Craigps +on +Cant reproduce on mac +Finally get back to this +Yongqiang can you take a look? +I will follow the other issue +so yeah we can comment out the valgrind test for now +Could you also have a look atWill do. +You have my vote for itAhmed +This jira has been closed as part of a review and archiving of the aj project +It is required if you intend the patch to be committed into the project +Marking closed. +see issue comments +but it was a configuration issueSo this was not a bug indeed. +Changed from blocker to to camel in trunk +agree with robert can we try to survive without dependencies? What is the reason to have this stuff on github its your projects anyway right? also the spatialj notice file is a copy of the commons langs the reason to have spatialj outside of Lucene +I need a osx to fix thisrps please provide some patches to fix is fixed in master update on this issue +Super new patch looks good +Fixed as suggested by James. +The library builds cleanly now +Thanks for finding these bugs so quickly Byron! Good testingHeres a patch that should fix the issue +Pulled the changes back to and noted about the lack of this property before apache hadoop +Ok some more fun +I think it would be good to modify this improvement descriptionSeems like this could be a simple as removing the word union from the descriptiondoes that sound reasonable to you? Or do you want an entirely new description? Something like Subquery materialization via hash join or Support subquery materialization by allowing the optimizer to cost and generate hash joins with subqueriesAny preference +Postponed to. +Performed a search and replace on PROFILES +yes we have a couple of quickstarts for those +Im by no means a SPNEGO expert so perhaps this will be educational +Thanks for creating the issue +The are the netbeans ant tasks that create netbeans modules +code should first be aligned to be similar to what we do in to be asynchronous and leverage the Camel async routing engine +It does +MapUnmap based and committed in r + for the patch +Anyway this looks fine +bq +It is only exported via to the test plugin +Hopefully it will be fixed this time +Works for me from a clean local repository +This might be due to the fact that there are multiple JAXB runtimes on the classpath +thanks! committed to trunk and +Because as we have found here already it is allowed to have multiple line number attributes for the same line of code +I think there might be more leaks when using maps and unions +passes derbyall now modulo the wisconsin noise +do you envisage the latter part of this to be a lot of work? I am happy to put in work to the Maven stuff if it would be helpful +Roo trunk will not work worth with of GWT and you must download the source code and build both to have it continue to work now. +OK thats fine +Turning this off by default will benefit most of our usersI do not support this idea +I have committed this to and trunk +Includes fixes to all mvn site errors +Im going extract to be an interface with default implementation and then will change all the code to use instead of +Another point It would be a big chanellge to provide component transaction support from Tuscany from my point of view transaction is the most important matter for a enterprise software to go into reality +If that applies to SNAPSHOT builds then that is fine with me +Attaching +Changed the a bit for a better look for the patch + +Another thing Ive noticed is that is still using logic +We could move this into the but its very difficult to know what is and is not a model class without more information than the would have by itself which is why Im thinking it might need to live outside of that. +Lars are you referring? It seems it is not going to make it into wdyt? BTW I am all in favor of since it will make zk the absolute authority about the state of the regions and simplify splitmerge +The issue is with the which is treating a correlated reference list as having correlated references +patch ll revisit laterAnd if someone in the Scala community step up and help work on +Ill a new patch after running the tests +I posted one possible patch to the Google Group +Just a note whatever we do here we should also do to +It looks to me as if all these other components are not created in the default and Im a bit reluctant to propose changes that I really cant test at all +In my examples the response is always textxml charsetAny ideaBest regardsTobias +Add test tosrcsrctestjavaorgapacheactivemq has been fixed and is available in problem exists in snapshot of Mar this is a slightly different circumstance with a different test case +Setting this to resolved to not block the release. +It also contains the changes from my patch for because the tests are failing on my machine without those changesTo resolve this issue I have reused the fromTIA for review +it would be nice to make getBuckets suck less in +Fixing Fix version field +This patch makes create an absolute path name for HADOOPHOME +Im going to give this another shot +I would prefer to simply remove the methods open isDir in Namenode and +Not a blocker for SOA fix into. +Updated the patch based on Francoiss comment +Once divergences are made it will not be possible to simply compare two branches of the java code in order to identify the changes that need to be made to the code +Ill try the new nightly on Monday and let you know if I find anything else. +Somehow that got dragged off into the weeds about voting and other silly things +It seems in version before couchdb will parse String to boolean but now its more strict +It worked nicely to break the lines but failed to preserve the spaces at the beginning of each lineAny suggestions? I appreciate it +I will take this one on as part of the Cocoon +This way if we arent querying that dc is penalized but within a reasonable limit such that if the local reads get bad well exceed the threshold and go if we have to +Triaged for +Committed in rev +This implementation is likely to change soon as more things are added +I can fix this but there are more problems with the news blog than this and you cant tell at this point because you cannot view any of the old posts +Its easy to forget how it works under the covers +As far as the priority goes I really dont even rememberselecting it +Because of the fact that is called at the time you call the setter rather than at test run time its important to call setwithConfiguration before calling setwith or setwith otherwise the custom config parameters wont get used +Ant has been successfully run on this patch +Updating fix version since this needs to be fixed on trunk as well +Added missing comments +Why cant the recordReader get the rows in a range through thrift? This can be done by first calling getkeyrange followed by a bunch of gets one for each row +started and hung but many tests failed before this occurred mostly related to network server bind errors +Can this be released as? SpringJPA has been since RC due to and now that that is fixed it would be nice to upgrade to a non release candidate +So I believe I misinterpreted how options are passed +I would recommend to refactor the webtests to avoid the duplications in testReloading. +Lets kill this off for now +Is there a possibility of anything else happening here something like a disk failure +If it can be done easily then I think we should do it +Im pretty sure we can get rid of the isReplaced flag now +that isnt true +Jarek can you log a Rules JIRA to reference this +name is only added when the user is supposed to name something +and combination as suggested is being tested +We experience a very similar problem in +branch has been updated to use +A day lost should have been phrased as a day spent learning other things I wasnt expecting +I get the same behavior on the command line +The available features in the framework which can be used at all for limiting user logs are s limit via Log cleanup via is not usable in the current format because of its limitations If this is used showing the userlogs is not possible until tasks finish or fail +datadata +Thanks Vinod +It is scheduled for eventual release on May the cut as of looks good and does not seem to surface any lock contentions any for the prompt turn around with this issue. +ThanksWould mind also adding to documentationcontentxdocs to document the availability of this wildcard option +Applied patch now works +I see how that could cause a +Felix has been released so I believe this issue can now be closed is that right +The fix seems simple and straightforward +Grid Im guesing this wont be resolved for this release +Effective Roles vs +I have been eagerly awaiting for a fix and definitely have a need as sean schofield described above to have a version of the radio button that you could include in your own table +I agree that this is a very conservative time +Its should be already fixed at the master +Not a big deal with timeout and friends so well nail this on M. +Thanks Jimmy +I really dont think this is something we need to deal with maybe we should mark this as wont fix +The issue appears to be that the generated file is empty and does not sync with the source editor +Though on some systems with there is no making it impossible to have a include in It sounds like a reasonable concern to avoid including directly +. +Indeed it was a typo. +Similar changes should be made to the async HTTP service as wellThe since tag on looks wrong +No idea why there was configs before when you cant use more than one scheduler anyway +flavio can you please upload a patch for branch +This SQL statements against jira DB should helpupdate issuestatus set sequencesequenceupdate issuestatus set sequence where pnameQA In Progress +I think the right steps areCreate a new property with default value Change default value of to Add a unit testCan anyone tell me where to find the initialization of these properties +AllI updated select radio and checkbox today so that setItems takes Object rather than String +This OS stores all filenames in a decomposed format +Are there any plans to release a? If yes can this fix be included +attaching a first version patchAggregation support can be turned on by wrapping the lucene index in a and proving a definitionThis should work the same way for solrThe patch adds support for named child nodes only +durr test ve also added tests to rubysecs for this. +I simply took the proxy class from the simple sample included with Muse and modified it to output the xml request using different ways of using an array for an operation parameter +This is the patch for adding the Ant targets of generating and cleaning Eclipse project description files + I noticed this too while using ideaidea the top level projects all have the same name +Whatever we do somebody will be whining about it Yup +Since there might be users of we keep it in docs examplesdocsexamplesexamples moved out of because it needs to be deployed seperately now that the registers the +Thanks Zhiyong and Raghu. +They seem fairly straight forward for what you are doing +Yea I think it can be removed assuming it still works we should probably keep it around for one release in case anyone has scripting against it +Jeff When that patch is committed we can look at putting a snapshot of Avro trunk on ours +Will commit + +Could you confirm? +For default JIT setting probably does not matter but for wierd JIT testingsituations maybe +A general purpose text decoder that uses regular expresions +Ill try make the time shortly. +Already filed to address this +hope to fold this into this operation +Thanks for the patch Jiawei and your reviews are much appreciated Ravi +Gregory here is a missed change for the previous patch +An initial patch for for everyones review. +Default should probably be permutation is OK +this compiles and tests from parent pom +SLFJ is needed by Thrift so we cant exclude it or limit the scope to testThe build works again but I havent yet run the tests. +Thanks for the comments Brandon! So Ive renamed to and move to the nfs package +The big one I see is a potential deadlock in which seems odd because there appears to be no other synchronization anywhere else in that class but I have not dug that deeply into its parent class which probably has some synchronization in it +Shawn is this also applicable for Tomcat? I have set the max post size to MB this +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +This should probably be updated as well +Patch applied to first error is a warning that no custom hazelcast config is provided and that hazelcast fallbacks to the default settingsI think the second has to do with shutting down the hazelcast bundle +Now it is working dont know why I get this error +DB is the new label for the DB option +Mr Dixon if the above suggestion actually fixes your problem please close this bug report +I confirm that the fix works! I dont have anymore the timeount on latch error messages +Im still having trouble reproducing it under Mac OS X +At this point two things can happen Block received comes from datanode indicating finalization of block +Changed string to so it now shows up correctly as sca syclusteredtrue syloadBalance +it just uses nothing to do with dependency should just be removed. +The main exception used on the XML side is so I think we can get away with moving the to a shared package with minimal the branch well just leave it as is but the main branch this is an acceptable move +With we have to use non shaded jar which can be generated directly by just clean install build optionsSo we will need to keep bk jar explicitely +I was going to look at it again today but looks like Ill get to it around tomorrow? Will keep this posted +and autoconf v +Thanks for your patience +and passed +Move all the util methods in util package and currently there are three util class which areCommit the changes to trunk At revision . +But if the or is used then only family and qualifier are matched the value is not considered +Attached management server log file and cloud db +POC code attached +Okay now I am confused +and is now at version so it is in line with current everything with no issues +IMHO Karaf should go the extra mile and support commands on style as well in +Creators prerogative to close Im planning on something else. +I think only for master should be enough +What is the point that you are trying to make Ted Yu +Fixed in revision +Updated the to recognize login regardless of case +I feel our Poison message may serve well the purpose if we combine it with a few new concepts +I dont think facelets has something to do about this i would like to help you out with this but for now i cant reproduce it a sample app would definetly help out +This should fix it +more details to latest patch is against the latest code base +I have fixed the GDAL module setting the default tile size value on both and trunk +I tested this by Running mvn P localTests Dtest test successfully Starting a master and a couple of backup masters checking zkdump killing the master checking zkdump restarting the master checking zkdump +Sorry additional question is this documented somewhereIt seems like some tests now have to be run via testdriver and some directly on real cluster +To me the current solutoin is fine. +Can you tell if the dependency managers that are experiencing the are optional after the reference that causes the activation +That could be used but this is also a bug in the since it assumes that the build tag is present in the POM + +Let me know if you want words on the homepage mentioning its atticnessUpayavira +fuse rev where I deleted the component from the fuse repo +Currently the alias for a is enumerate +It almost seems that everything targeted to or and is a MAPREDCUE JIRA is mrv +Verified +Itcould well be that such relations will not be ignored during updates despitebeing declared as will have to revisit my test cases but with my current job this will not beeasy to achieve +Verified by Mark. +Unscheduling due to inactivity +For a long running client once failover has occurred the client knows which NN to talk to +cant seems to commit into apache svn will try later +It was working by fluke because the build file happened to be in the current directory and was being. +So tis +had no direct dependency to so I ve added that on both master and the branch +However Im unclear what to do about +Thanks for taking a look at the patch BryanCommitted revision . +Ive already tested it using the Ive attached +Ill we try them out with my current SPI implementation as soon as I get around to it +Confirmed If mail arribe from root and send this untouched to it sets the bitThis is the default in Seems that +Sounds like the best thing is to have cqlversionVER along with cql and cql shortcuts +Thinking longer term could the nested maps be converted to simple concurrent structures? Is access through a key necessary? Could iteration with type checks suffice +Id be interested to seewhether theres any remaining data types that are not being covered right from that does anybody have new references to the mailing list threadsdiscussed above as both links seem broken +Therefore a fileset cant be constructed if one goes outside of that directory branch with relative directory constructs in cache +zip of test case +Heres a patch against the tag +The reason why this patch does not contain any test is because it is a trivial change to scheduling info string in Job status +Just as simple check the same code wont work with hcommandLink in pure JSF env. +Im happy with this for now. +Those messages are expected more or lessIs there a functional problem andor a stack trace +Using direct buffers for RAR and CRAR may also help avoid heap fragmentation +in jbpm we will focus on reviewing the node types for workflow patterns support and making them persistence and extending jpdl with the proper xml to addess all of the tom +Xen VM snapshot is a new feature +Where is it coming from? Its probably my mistake from but I would like to correct it hereI dont see the param which file r u referring to +Just plugged the jars into my application and I seem to be back in +Updated this JIRA so that it is specific to the dependencies and opened to cover the dependencies +Committed to and +in the directory jbossFinalmodulesorghibernatesearchenginemainthere is no +Feel free to reopen if required +I patched that one single line and Ive been using Weinre for the afternoon without any other issues +any more comments? else want to check this? Otherwise will commit tomorrow to trunk +Thank you very much for the quick work +I was reading that differently +Ill file a separate Jira entry to enhance the testWhen I ran derbyAll I found there is another problem too with the test that tracefile goes to a different subdirectory based on whether the test is run in a suite or independentlyI will fix that too +is officially released +Just hit any action that is not mapped +Thank you for the and branch patches Heath Kesler +Mostly Pig works with s +Thanks for your patches Paul +Wo your patch we are broke +The thing Ive done thats saved me when the namenode dies is simply to have a cron entry that rsyncs the namenodes files to another node every few minutes +Essentially the related functionality was inlined into each strategy +committed to tez branch +Also the source tab now updates in whenever a change is made on one of the other tabs. +Done +It should open mapping file if anything +Module migratedDep list pruned +Looks good +In my opinion this is a bug +Ok found it cactus was reseting the test to a wrong value + after release +It doesnt really do all that much useful +The lineto be deleted by that part of the patch is not dead code at allI did not apply the fix to because that dead codeactually exposed a bug +I meant file system cache +Verified resources are now delivered with the correct +BULK EDIT These issues are open to be picked up +I forgot to say that i also changed the way in which the rpc service name is a static in a class means that i cannot have two services in a server using the same class because the two services would register the rpc handler with the same name and rpc calls would end to the wrong changed them to use the service name as the rpc name so im sure it will be unique +But maybe next week +The problem is the enabled attribute is persistent and updates the configuration on the slave HC while the enable and disable ops are and only affect a running server +It is wicket that for some reason is considering the session stale +It simply adds the context to the fault +It always failed immediately for me when attempting to downloadWhere is the? I could not find it +This code appears to work correctly in all my tests to recognize general decimal and scientific notation for input +The fix may not be the best fix but it works to address the problem at hand +Then can be mutated in the same way and metric naming can be handled by static helpers onlyOk i see +Many thanks for this enhancement +This should work however that the annotations have to be on the interface and cannot be on the implementing class +Minor build problem needs to be added to also needs to be added to Addition of the Close method to the and interfaces in this patch breaks the corresponding MSMQ Session Connection and implementation classes. +HiTested out Nightly build of March th works perfectly fine for my use caseTHANX a lot OlegBest regardsThis is not an error but merely an info that the server behaves a little funnyYou may either choose to disable INFO messages or activate handshake as IIS seems bent on using it +So the user can simply add these kars to the deploy dir and does not have to pollute the system repo. +Submitting Patch to trigger Hadoop warning are not related to this patch +Ping +Fair pointWe could also go in the other direction reach out to Jon for a grant to port to Apache as Sergey said and then carry it forward maintained in tree +First I have to give credit to Ciemo for the idea he mentioned it a while back on the ticket +Very slight tweak to the release note +Big thanks to everyone helping to provide this cool frameworkAnyway please close this issue. +Then I will reassign the ticket to Hitoshi in order to add a Japanese translation of the error message +Here is a first take on this one with very preliminary testing doneI know Devaraj has some but yet +Attached patch a which implements an alternative test parsing the client connection trace fileIts a bit more extensive than the existing test but may not add any extra valueUploaded for reference as it may come in useful for debugging purposes +As the comments said for one thing we have also from past discussions on the it would be a breaking change from a Java perspective that wed return a value which may not be a boolean which would break the existing contract in place. +Have not seen this issue in awhile +Can you please create RB for it? Also would you like to change the jira title to reflect it is for +The version of this package does not have repolib so is not in the MEAD maven does not even have a POMAs a result any builds from the JBPAPP branch are broken including the Hudson ones +That will be helpful to make a good test subset for Guththilathanks did any fail that dont fail with the xerces parser +ThanksThe wiki would be a good place to lay out how to run this stuff +So ironically now that we are ready to roll it out to our larger audience we realize the magnitude of this bug makes it impossible +I believe is thrown when corruption happens sometimes which is why it was catching +I ran derbyall and ran derbynetclientmats with to verify the issue was fixed +It would be great if Single Sign Out is supported in next Fediz release +Someone there may knowsomething about the bleeding edge changes they are doingJust out of curiosity can you point your program at any OTHER ftp site and getit to work +The workaround as it is clearly a fault of Maven as a search shows this is not the first project to encounter it is to set the proxy settings on the command line of the build +From netbeans it looks like there may be an issue in netbeans itself in adding Ruby platforms +Here are the new patches and integration tests. +Might be an issue with the JRE used +So any one can show us the script etc +trunk all tickets +Is it possible to have replacement of maven variables in the given xml file when installer file is generated +from Maven Indexes view and from Windows Preferences Maven Installation Reindex local repository +The testcase isnt using delimited identifiers so itd be a bit misleading to put it in that package +Making blocker so it gets consideration +Shows the error shown in the editor +and gotBUILD SUCCESSFULTotal time minutes secondsThat is the correct way of building the AS cd build and +Review and improvment of step of BNF reduction of of +And the put this tar file to distributed cache +we arrive on the th we wont need them until the st which is when the conference starts +Bhavani can you verify this fix? +I see lots of code checkin in olivers git branch +There is some indetation issues in +push out to +The changes that were put in for are causing some issues +Done. +I just committed this. +No sorry we dont add users to AnalysisPlease let us know what you need doing and well take care of applying itThanks +I havent checked our clusters to see if we see the warning message but the jetty thread is actually there +setQuota and set is also allowed +This patch allows one to call the cn functions with the canonicalization algorithm +All screens are retaken with resolution +Thank you. +A complete minimal program that demonstrates this problem +Have you taken a look at the in? Ill pull your branch in and take a closer look now that is almost wrapped up +I dont like this change because the value of this feature vs +Fixed that findbugs and release audit warning +Jirka S +I dont think statics are advised in coprocessor and thus providing an example using them is setting the wrong expectationsDo you think the current way I have it is bad +We have brought this to the attention of the Maven developers before +patch applied and modified test included thanks +Committed. +I see +Hope that they will add this kind of flexibility on meclipse side but in the meantime do we know when fixes will be included in maven corethx Frederic +I assume you mean something like this patch +Thanks Arvind +Updated issue to use the more recent version of the Apache parent POM. +Initial proposal is to keep it dead simple Keep a static map of keys in the Configuration class that maps the deprecated key to a set of new keys +I found another bug testing multi values query parameters get wasnt append more than one query parameter to the query string +Thanks Alexandre. +but we will never really be able to do logging nicley until geotools smartens up and uses. +committed to trunk. +Could it be that some of these extensions have a which caused you to submit this change? Can you provide an example of the problem youre trying to solve here so I can compare the generated PS code? Ive seen your test case but Id like to see the effect in PS code +Thanks for following through with this Deepa +Landed in r. +I am in mind +Hi Chris hopefully this is along the lines of what you were am also working on WIKI pages for and over the next few days +Todd I would like to see the system proceed with launching the RPC handler threads and starting to process Initial Block Reports +AdrianThanks for the prompt response well take a look +That file is generated when you run a nant build of the libs +Thanks for picking that up and fixing. +I tried everythingTo run the prototypes you just need to changed the connection in the config files +Please see my earlier comment on discussion related to API and wire protocol stability that we had months ago +Updated patch +The fallback resource bundle mechanism that is part of webwork should satisfy the requirements of this issue. +I was talking about rules names descriptions and sample codes +The right thing to do is to wrap all the commands run via ssh in a bash l to force a login shell to be used +The REST server registry is now the responsibility of the project +After installation of Tomcat login is working +putting on for now but without samplesteps to reproduce ill reject it as incomplete shortly. +Similarly detailed statements would need to be made for each Externalizeable noSerialized objects in minor releases will work +Lars is dead on here compactframed doesnt work for php thus its a non starter for us +Thing is somebody on the outside doesnt want us to see the classes so we shouldnt try to get them from some place else +will follow workI did something in java for yahoo interfaceit was so could access secure webpages that dealt with order status in the storewould like move it over to this if possible +The main problem in driver reload is to handle multiple loading +Im wondering if maybe some of these things can be rolled into a list rather than separate tickets for every individual tweak +Ill check out repo and contribute solution proposalI can help you with maintenance of if you need another pair of hands +Here is a first draft of some new permission code for content +you dont have to put a wsdl to folderCan you please send your +A mention in the methods and the class level doc on CUE +Closing per the release. +As for your thoughts on the redeployment the script you mention would be fantastic and wed greatly appreciate it if you could help us out here +Attachment has been added with description FO to reproduce GuillaumeIt appears that this bug actually is a duplicate of +Ok think thats all fixed now as well +rename overview to index goal links dont work I dont think the example of pmd in the build section is that relevant examples links dont ll update the docs per Bretts comments +You could for instance remove Product unlisted from the and add a button or next to the for users who indeed did not find their product in the list +Shared environment classloader and all that stuff +Closing this issue. +It also has an attempt to remove the jasper dependency which might not work and is not essential +I explained the reason in bluesky list +Hi Mano Davidit would be very appreciated if you could provide testing sample minimized to show your case since we are not able to reproduce it. +services sample file Is it possible that Spring can provide a similar solution for Converters auto registration? +Many thanks for the patch! Ive committed it rIll leave this issue open and have a look at part s for the quick reactionConcerning part Unfortunately detecting Windows versus other seems to be not that easy +patches is added to CQL mode it would be ready to be committed +Its tricky to figure out because most people seem to use solutions so I thought perhaps I had stumbled into a less than common corner case of Spring dont use any DAO and have no entities representing Users or Roles in the database +Hi Steven can you please submit a quickstart that reproduces the problem you have? I dont mind fixing it but i need to have a testcase +Schedule so I dont forget about this forever +I wouldnt want to do anything under the covers to promote this bad practice +Thanks Rick +the shortcut was only implemented for describe not for dump +Hi KristianThat is a good point +I am still amazedwondering why that the JMX doesnt offer such a security scheme out of the box? Or does itBTW I like the way you implemented with the proxy and for checking what the operation is performed is a read only or not +Committed to trunk +Also updated the unit tests and javadoc to reflect the new modus operandi +Patch for implementation Cocoon that adds possibility of getting connections for +At least lets stick a static member somewhere to make it trivial to do that so I dont have to write the class myself +Thanks for reporting +RobIm proceeding with the missing close protocols at the level of the Solr index and the Solr analyzer and I hope that you and Hoss can agree on what comes next after thatbenson +The ejb reference is cleared was also cleared during passivation +remember to use +Hi David Thanks for the patch +Closing issues before next release. +The getters and setters for UTF and ASCII strings need to be added too +last bug fix for! +We should not need userspace checking as this was a buildfile nd code error +to trunk and +Ok new patch for trunk + +Now youre complaining about the size of the splits not being even +Difficult it is not always an +We still need to define the behavior of shifting by negative numbers in the manual +Currently the delegate injection is broken for all decorators +It seems like we could avoid losing a lot of work if we had such as mechanism in place +I suppose we could add a flag to switch on automatic registration +replacing with the released version make things fast again +Ive been developing the based on these changes and everything is working +Will upload a new patch soon +Closing it. +Ill look at it ASAP +if you dont already intend this it would be nice to extend this so that if no p is specified on a arguments line and socketserver is then they also assign themselves a random port +Create issue in Eclipse bugzilla about that +to and Trunk branches +I support a whitespace policy as defined by Cheolsoo above No tabs no trailing whitespace spaces +We are deprecating for M and by default will use so explicit configuration will not be necessary to make use of Document or any other annotations +Its fixed isnt it? +better patch +Verified +This looks good to meHas anyone tested to see if this fixes +Im sorry Julien but I have not had a chance to try to debug this +It looks like is obsolete now as it tests Request and Module scopes only +The fix works for me. +Verified resources were moved +The rationale is that its not just about Cacheable its about any put operations that happen within a transaction +The section numbers will likely change the next time the book is built +So in this section of the code I dont mind attaching jobs that are not intimately connected to the core code +For example using an inlined service manager the bean property or argument shall be a object +Heres a version with the other half the test. +Please check that the patch was applied as expected +Regarding theThere should not be a element in element only in JoinWe are missing the definitions in join element key and value for Index and Unique +Fixes appear in Latest Alpha Manuals. +batch transition to closed remaining resolved bugs +His pull request is buggy Android device shows the desktop version +Since it is not generally viral and provisioned by the user if they accept LGPL pieces it works out fineSee for example aprhttpd consumption libiconv available as either BSD or LGPL and not required for the operation of the software +I committed the changes and Im closing this issuehalley +Mac users can comment out the dependency from their POM +patch against trunk rev +Patch v from review board +Ill take a look at it today +Its just svn diff the one included to the comment looks really bad +Not fixed on branch since does not use JNA +Any developers who want to implement it can apply the patches to their local copies +v patch Since REST and Thrift are not authenticating for HDFS access only secure RPC additionally check if authentication is set to kerberos +I may think too much +In some cases it might not be so easy thoughThat exception is telling you in cloud mode this feature is not currently support perhaps we should tweak it to be more explicit about thatSince Im not familiar with browse or what it tries to do internally its not high on my personal list which is quite long +rerestarting is solved by view id. +I mean that is the default behavior where the colon is not used +Committed at subversion revision +Enis do you think this can be committed now? +It took forever to attach a patch so Im creating a new comment for the small patch that will fix the template +Limit will reorder data unless the statement right before limit is order by +Libs in lib are used only in hadoop driver during job submission but not in mapreduce is a major regression if its true +the json tools are for debugging not migration +Ill take a look over the next few days before going into hiding for the rest of the time Robbiesorry for the late answer I was on the vacation +bulk defer to +Perhaps theres a bug when a record has no other fields besides a +Thanks for the reference I wasnt aware of that +ditto +From Slava We have preference Tools Web Editors Always use Tools editors with Open option which is used by Web Projects View +Thanks Javier should be fixed in r +will not automatically quote names +Patch basically follows replacetoken but instead takes a host id and looks up the tokens +Ive fixed the maven configuration to not include dependencies now Im not sure what to do about the binaries +Hadoop was just released. +only logging it during compactionAgreed that its the best option +Verified fixed at rev +looks good to me so long as this doesnt introduce a race condition for frequently writtenI have an aversion to inside the DB its not a nice thing if ops people restore a +waiting for some fixes in the felix properties before I can now committed an impl that directly changes the pax config file and includes a fixed Properties class +Applied branch and trunk. +I will not be changing behavior of identity columns +I guess we will have to live with SPNEGO in its present form for now +This issue is duplicated with +Beware that sometimes this breaks being able to cut releases when upgrading pluginsAnd the compiler plugin or was it causes the components to be compiled x times so it was much slower in the build process +I have a prototype of this undergoing cluster benchmarking +Let me put it up on review board too because would appreciate some review +Ill try to attach a patch ASAPI already did most of it yesterday night but it was a bit too late to be able to trust myself Ill run some tests and if all goes well Ill post it later tonight + +either an exception is thrown or the queue is transparently created +able to start vsphere instance in aeolusvmwareToolsfrontendIP Address State running +The patch is part of either way is ok we can include it in this patch since it is small enough +This is fixed +But since youve this now Im happy Thanks guess the benefit of the new behaviour is a fast failure with a more meaningful exception than before +I just committed this +Mr +I start to create new Wizards for new API Hadoop CodeLater one can select what version should be used +Brien Voorhees can you please attach an image that demonstrates this issue? It should be fairly easy to resolve +Buffering was added as part ofAdded tests. +Hi ArtemThe testcase alformantioned passed on r on windows xp sp could you please verify that? Thanks +closing since we did not see any objections to the approach +Ive changed those to use getInstance method instead +Jason Ok +In JDK a list of providers are loaded and sorted according to the system property but get will still only return one hostname +This doesnt sounds critical for Beta does ituser can just press Frefresh to have it discover the change right +Or perhaps increase to a higher ratio if you dont need that much physical memory +Looking into itRegardsRal +The change looks good to me +This has not shown up as an issue since resolved issues. +changes to code to test for empty myContainer up frontas suggested plus removes some unused imports +SOA now includes jBPM +Fixed in nightly build +That version works for me too +with further details if still a problem with current version of Bcel +Im working on a small test for this +I was just curious why you were enabling and then injecting a +As this check can have a big performance impact this should be configurable at the level of the resolver settings andor in the ant task +Please create a new issue and attach your test case +Ill be sure to update the wiki soon JanThanks for your guys help. +Ill have a go at a rewrite +Thanks for taking the time to check it out +Attaching patch against trunkI think the patch does not track moverename properly so I suggest to svn copymovesrcmainjavaorgapachewicketextensionsajaxmarkuphtmlformuploadtosrcmainjavaorgapachewicketprotocolhttpservletbefore applying patch +Missing patch +is better describing possible implementation we need +One thing though the current docs also includes API documentation which is not generated by make +Thank you Bertrand. +Verified on JBT +Ive just committed this +The JAAS login takes place inside the mbean method which then calls the ejb +Vladimir reassigning to you for clarification +For example there wont be a anymore and we rely on Hibernates own JTA support for JTA synchronization now +So it seems the connection to the database was lost +Not critical moved to M +Hi Nick +Resolving this issue as it is fixed inThe fix for other versions have been reverted and the same has been taken care in +Ive run the tests repeatedly and wasnt able to reproduce the problem +I will extend it later +PR withdrawnNeeds a more robust solution because amqp jms etc generate their own log messages +Its probably the same for though it should be fixed tooThe other way may be having something like and we test against that +I use this all the time +Yes this sounds like a good idea +But I dont want to require that the name spaces are distinct +This situation currently causes lots of issues since to the resolver system bundle packages look completely unconstrained since they lack all uses constraintsI am leaning in favor of adding this metadata +I dont have a failing test but the working assumption is that including the full list in an updatedocs call generates spurious conflicts +One problem I see with this is that in a normal SWF config the bean is not explicitly defined in the application context so Im not sure how we would export it via JMX +the patch was committed to title to make clear that this has already been committed to trunk and only maintenance remains +I guess well need a branch. +Done + +This object only holds all the detail about the status of the method which was executed and also the response entites from executed method +Thanks Sushanth +Four patch for the new style to TransactionsTax Authorities forms +So we have removed it for in +He was able to get a little bit farther but his solution seems to have some concurrency issues +It needs some improvements when write namespace +bq +this looks good to me if you have a chance could you turn the reproduction into the integration test format +Attached a patch for review containing what Ive got so far +Committed branch and trunk +true if the Type represented by the c parameter and the current Type represent classes and the class represented by the current Type derives from the class represented by c otherwise false +the precedence looks right to me +Added failling test m running this at the cmd line +Wei Yan can you please put an additional patchBTW I pulled the earlier patch into also as other patches in arent cleanly applying +attachedit fixes cpu burning when not specifying a sleep timerThere is still an issue when cancelling a tail f command you have some character already on the next shell prompt which is invisible but prevents follow up commands to succeed +Im not quite sure how to do that yet +Checked release notes +Then if you want to take advantage of this you just delete your xsl stylesheet to merge an input with a custom +NET call into the +Great thanks again for your help + for the patch +updating title to include ref to AMQ Donald I will prepare a doc for the JMS portlet soon +Verified fixed in ER build +If thats what winds up being done Id like a pointer to an existing example so I can stay consistentYeah basically like picking a spell check implementation or any of the other subplugins we have within the various s in SolrAs for my involvement here Ive got no time in the near future to contribute an implementation of my idea but should be fairly straightforward to leverage the ideas of how other Solr get their implementation details via named plugins +Here is one use caseRendering dropdowns with same set of values in each table row +is unfortunately hard to address +KannanI think you were referring to +Closing all resolved tickets from or older +Any update on this one +Draft complete ready for tech by Ilya. +We need unit tests for also +import removed. +Harrythe problem might be found in pwebjspwikiwhere you use an absolute path +would it be possible to use transitive dependencies rather than listing every transitive dependency? This will effect the generated pom file +The migration script is missing but you can start testing +Patch adds support for adding itself to the and also exchanged the in for the class +expects to have this done within a day or two +HiIm using and JDK on Win XP +Closed as a result of. +More infoThe problem seems to occur only when the class name is using +Is anybody working on this +Ben can we please take discussions to the issue to track what needs to be done not to determine how to do issues should be separate tasks with this task closed when we are happy it is complete +Should be easy to add a default constructor for thatStill a would be a better solution probably +What is this actually supposed to do? Any resource can be used as context already +Does this only happen for multimodule projects? Are you able to reproduce the issue for projects +Findbugs was alerting that rit wasnt being used +Fixed in master commit dbcf +Please dont leave code inThanks +On trunk can this utilize +We decided on the dev list not to fix this so Im closing this JIRA. +Max plz review fine using same method as in for lookup +services authenticate between each other via kerberos as practical usecase? Or most people would just do trust for Kerberos to accept users from corporate LDAP and thats it? +Sorry my second comment was not a reaction to yours it sounds real sarcastic this way and that was not the intention +Looks goodIll close this issue when the docs are out for QE review +If you still find issues with the latest Axis trunk please reopen this issue and attach the service to reproduce this issueThanksIsuru +Attachment has been added with description PDF has been added with description FileOK if anyone else has a similar experience I had a bit of trouble parsing this report at first but it comes down toThe example examplesfoadvanced that is part of the distributiona has a reference to a TIFF file that does not exist leading to a SEVERE error during renderingb contains a barcode in embedded SVG that is not FO leads to errors and wrong output. +Can you indicate to me what to do in the descriptor regarding filtering +Please verify the fix in the latest are included rev +My apologies for taking your time +From your error message it looks as you may have only updated the lucene folder and not solr +Much simpler implementation which piggybacks on and +Bulk closing stale resolved issues +Thanks for the patch Lars Francke. +Patch committed to trunk and branch +Manually tested +Thanks Raghu! +Approved for +Hi Andriy thanks for checking it +Can look closer later but seems good to me Jesse +Yes but we cannot test this on CI as we cantshouldnt adapt the archetypes just to fit CI +Thanks for the patch +We should also add recovery count +Oops forgot to add item to actual list +Applied patch in version +I have committed a patch to fix this issue +The batch size seems like the case we care most about +Maybe its a good idea to consult jgroups guy Ill try the priority to a proper value +Still not sure I like the whole feature but if we have to do it in this is a reasonable approach +Lets move to on Jenkins +Fixed in revision Thanks Fix version to +Is there any developer willing to comment on the issue and explain whether or not this will be fixed +I have tested on master branch by using jtest tool it seems not happen again +Thanks Aaron +I dont think this is an enhancement request +hope this time nothing would break Svetlana ill look into your port patches soon +Then in after most of the data is passed in on the command line if we can also migrate as many of the tests over to not need to be explicitly enabled that would be great +Todd As Arun had mentioned in his email to general we have already fixed this in branch +I also need to change my current organization info because I will resign from there +Does this problem only occur when the server is bound to a loopback device +For a normal scan wed delegate the locking to the enclosingThat way there is no repeating lockunlock during frequent calls to nextseeketcIt would also force delay updating the readers until its safe to do so without any complicating reference counting logic +As its pretty simple so Im bypassing review board and giving here + will commit after tests pass +Thats good newsHowever requests are not supposed to be long here they are simple search with a DN as key +Was wondering if any updated status on this patch +Thanks AmareshwariThat totally make sense! I will upload the patch soon +LGTM overall +Im going to call this fixed since the bulk of the work is done and file other bugs for the remaining missing bits. +Please attach a gdb trace if possible + +Thanks +Tests look good for this patch +Attached the screenshot after applying patch +Thanks Dag! Im glad you caught trace it is not mentioned in the Network client tracing topic in the Admin Guide only in the parent topic so I missed it +Seam was upgraded to Drools agree that we can see it in compile time but what is the point of that exercise? We may discover as an obstacle and that requires to do something in Seam +We have had to do that in a number of cases +Either that or I am seeing things differently from you +The soa plugin pack contains exactly the same bits that have been tested +Michael can you please reupload the file because it is lacking the right to use it in Apache and dont forget to check the correct box +will write on SOAP api irrespective of whether the recd message is SOAP or +Thats why onall winboxes you can just type ipconfig instead of So this change was not required +Thanks Pavol +Patch increases the max heap for hedwig server testing +So let me rephrase that was it working on? Or both +a patch that lets you pause will go ahead and commit this +Attach to fix it. +move to for now +When creating the adminobject Connectorbinding the identifier will be assigned the deployed name and the name will be that the name assigned in the Designer. +I tried your latest patch with the latest trunk code and Im not able to reproduce the problem +Count me in as well +There is an that is being shut down after the server causing the server to go into TIMEWAIT on this port +As I mentioned in my previous ticketsYou might be working on the directions given by Si Chenthankswith best regardsVikrant +Added order by to the queries to get a deterministic output +Fixed in +I guess this can go into trunk and give it enough time to settle before the release +In reply to comment This change will be available in the next nightly snapshot as well as in the upcoming Spring release +Bulk move of M to to M because of no updates from +Thank you +There it works bug was fixed after Tools were released + +rFix to detect circular reference and avoid infinite loop. +batch transition to closed for remaining resolved bugs +ok done i changed the Dominguez SarmientoI just ran this test locally again and do not see the test failure would you please provide more info +I see no way how to differentiate an expired session than a completely new user +Closing as requested by Petr. +My preference would be to describe runtime dependencies only and be done with it OlegOK lets leave it with the runtime dependenciescheers Roland +This is something possibly influencing the MOP thus I set a new fix version +Let us do this after the branch assuming we have a branch pretty conflicts +Thanks AllenJitendra please provide a unit test then provide patch for trunk and get to runThanks +In my discussion with Jing and Nicholas we had considered the following Allow snapshot of as a special case and that is the only snapshottable directory with in which another snapshottable directory is allowed +I plan on doing this within the next few hours unless any objections come throughIf anyone wants to send the message to the two lists go ahead and let me know here otherwise Ill do it myself later +Im closing this since now you can explicitly set the intancescriteria into the and components +Thanks a bunch for the patch! +But this is not a problem database Web server or anything else +Thanks Kevin + +Hi Eric do you have who can take this on as Jamie is not available +Also can confirm that its back up to speed if I set JAVAMEMXmxm +for Raghus proposals for this issue +The data field phone was bound to the data model property phone +When service to cluster is rejected the DN shuts downIts unclear what the correct results are here +Patch looks good +Hi AdrianI like the getDocument function but the getNode would just result in more typing +patch resolved issues. +Thanks Dan AkiYeah the patch is useful for the client side I will apply it soonFreeman +The repository descriptor QUERYXPATHDOCORDER still returns true +Avoid caching instances +Change some of the mapreduce classes to expect instead of Configurations so that I can get the System directory using the method +Snjezana thank youWe will try it +on patch +Test case added and issue fixed +The specification does not need to be changed + +Removed methods inadvertently put in lucene +That should be fairly easy to add +I havent tested the current IO classes with fragments +I was able to deploy struts application on without with actions mapped by convention plugin annotations +Sorry for the noise. +I am not adamant about it and if someone where to change mycode because they feel differently I wouldnt stop them. +If theres any interest in seeing my implementation let me know and Ill put it up on Github +Uwe had ideas for helping in that regard and documentation can probably still be improved based on future user interactionsIm not as dead set against it as I was but I still dont think Im for the change myself +is the intended behaviour yes +Looking at this issue Im pretty sure I introduced this bug with +bump to +Probably adding srcmainresourceswsdl and srctestmainresourceswsdl as well does it make senseAlso Is it ok mark the fix version to +The user can get sessions metadata via list +We should make a for this rather than requiring a completely separate module +Will leave it for now +Especially trying to apply the patch to the latest version was problematic cause that one changed very much compared to the version the patch is based on +The second with common commonb and the clientThe service tracker prints a Adding service or Remove service string +fixed at revisions on trunk and revisions on +Fixed in r. +I hope these patches can be accepted into +Also based on the forum postsrequests I am upgrading it to need to decide if we will use Spring for IMAP but without this feature it is a not an option +doesnt apply by request +iBatis is not active maintained anymoreSo we gotta have a component for mybatis which is the new and active maintained project +Patch submitted by Martijn DashorstSending guidesSending guidesTransmitting file dataCommitted revision . +Thankyou Matus. +Yes I did +So I think everything looks good and thank you for the quick response +As a matter of fact speculative execution was enabled before launching distcpRunning distcp with speculative execution turned on is definitely not supported +Please cross check and reopen the issue if needed +For the long run I think this is the right thing to do +Hari is concerned with balancing between multiple consumers while I am concerned with the exclusive property +This patch is an improvement over what exists in terms of performanceI am in two minds regarding having an interface esp +No problem +Sorry for all the problems +Ill nag the list +Maybe have to come back to the handling of stateful services for can plug in two different loadbalance strategies or or you can plugin your own load balance strategy +I fixed these in the attached patch +Next time we could also tcpdump traffic for socket between DN and DN until DN gets Connection timeout out error +Per concerns raised in jobdata metadata map setters now set a flag indicating that they were used +Tests all pass now +g as used in wfloat +Commited to svn with revision and +Patch for adding wrapped patch looks good +It basically splits webapp resources off into filtered and unfiltered ones leaving the war plugin to copy the unfiltered resources and using the resources plugin to copy the filtered resources with a cunning +I would backport this change to again only if knob was included +Done +Made obsolete by. +but would like to think on the lines you suggest for future +Committed revision Thanks Simon! + +Thanks Eli! And thanks to Owen and Ravi for code review +Configured to add it to Apache snapshot repo by default +are you working with lobs +This looks like a duplicate of so can you provide the debug logs of the above command to verify that you were actually using plugin version and not something older +Im always looking out for hints +I will commit it after running tests +Attached the Spring and Hibernate libs +Patch for Netty version change is against Hadoop Common code so I filed +probably that will be can try to reproduce it as Sutter wrote I wont have a chance to until next week +I would love to use an rsync mechansim +for the patch. +patch is good for too. +trunk closing stale resolved issues +this is a simple webapp archetype based applicationi have it all my pom please attach your Eclipse configuration details from Help About Eclipse SDK Configuration Details +Steve applied my changes to fix this issue +Changes are not related to the failed test +Thanks for reporting them Sebb +edit I did more testing with XSS and I was not able to exploit it +Is this still a problem with Resin +So we need to discuss it before this commit +Step should not be necessary +Reopening issue because we might need to downgrade back to +Why would we need the GEF site too if its already on Galileo? And why would I never see this problem when installing from Eclipse Classic which contains even less than the JEE bundle +Currently the dynamic client API and simple generated stubs are supported +Will commit if tests pass +Hi KimYes I think that if we forbid this syntax then we should close as invalid +But I am not sure the documentation need say that +I dont see an explicit statement of what meaning is attached to a which doesnt have a port spec +OK I ask a different way is it worth it to request those sourcecodes generation +This is now working and the createarchetype is an example of this working. +Attaching the first rev of a release note for this issue +Thus the solution of importing all the namespaces in the catalog would bridge the lack of communicationHope that clarifies the solution +some stuff has changed in hthe method you refer to is depracted and moved toThere is a new better way to define what the sql query returnsSee in how it is done +im thinking of starting with adding a tracker similar to how you described in the blog +I also fixed but havent touched the test cases yet +Ive the version of the plugin thats the problem! But I couldnt be able to find out where the is stored I tried the apache and the codehaus snapshot repositories without success +Thanks guys. +Patch attached +Fix in place +The agent and ajax code in TRINIDAD predates IE so Im not surprised it doesnt work +Thanks +Its a rare case that the interpolated model is needed +A document does not currently exist but I can write something up that would be good +Aha then that is in fact a bug Ill look into it this weekend +Same for ORACLE fixed +Some sort of flag in the request is needed to trigger model attributes +The use of Control is actually a legacy from our days and can be removed +This patch includes followingAdd getAnd getChange schemaTree from public to protectedAdd getImplement getBug fixingA typo in comment retrieves retrievedAdd java documentModify Java documentEliminate unnecessary spaces +That stitching has to be kind of fancy because there is a lot of polymorphism that goes onThat was a big point of the +Added enum support with a possibility to specify multiple values +Deferred this issue to be broken down into smaller more gradual refactoring issues to evolve the container into a stable state for V +How about adds a simple searchHandler attribute with truefalse? something like that sounds reasonable +Sample seems to have the same issue +Thanks Rob! Please confirm and close +And if no caching is enabled? One might not want caching for entities but still benefit from adding a proper cache library +Will edit the summary now to include it +defer to +At revision in branchesAt revision in trunk o Enables using DB portlet for databases with connection pools defined +Sorry +I have added second jar to the project everything is ok +Migrating is too much work for now +A new patch addressing your comment +Im just pointing that out +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content issue is out of date and will not be fixed +sorry the addresses arehttp load balancernode AJP I added to +About the suggested the is a meant to serialize the received messages so it has to accept an despite the current small difference between an and an it is cleaner to use the first also keeping in mind future evolutions of such interface +The error is just because the triggerfunctions need to be rebuilt +If we use pooling in Struts it should probably come from classes. +long and last server selection is done at sec +Did you make the code change I suggested two comments ago? Ive done it on my clone of your project and everything works fine should I send a pull request +It has the details of the test including the GET request +OTOH I would have been used a also +And the and should have their versions updated by the daos as well +Talluri are you still facing this issue +The serialization issue for the was fixed +is committed to as well +Upon looking at the sources it occurs to me that this could be caused by a mismatch between the public key they used to encrypt and the private key I am using to decrypt Ill make sure those are a match first +Verified with JBT. +Ive closed itThanks! +Fixed on the release branch and the trunk +One more time trying to get to run +Patch for points and +Added comments to the final commit. +srcmainjavaorgapachehadoophbaseregionserver dont think that complexity is needed +Maybe exclude could be handled by two options and exclude where each line of the txt file and each comma separated value in the exclude option are compatible patterns +I was able to close the staging repo now +Without it this patch will not work +on v +This file reformats into a patch so that can be applied to the new commons subdirectoryNote jms jmxtools and jta are not yet fully integrated since they are Sun jars that have to be downloaded separately +You will need to change your local impl due to my signature changes to the get not heard anything so Ill assume it works fine and will be committing the changes great Steve +I havent had the time to look into it +added back to svn head at revision are still big problems with this patch +I just committed this. +then changed it to be a cache archive so that the classes and lib directories could be added to the classpath +Problems include signing issues syncing permissions and most importantly very bad documentation +Verified with such seam runtimes snapshotSeam runtimes AS EAP this is great news! +I ended up having to implement my own id generator class basically a hybrid of and +I just dont want to go down the road of copying spaces without attempting to truly solve the underlying problem which is that we need versioned stored in Git and the ability for those versions to be displayed dynamically in the viewer +Released in. +Support for semicolon delimiterClosed by cfcdfdceabdcc +When the expiration period is reached the adaptor does abort gracefully +A pull request would be more than welcome +Ive now seen very rarely that with this approach the multipart parser can steal some data of a subsequent request in the connection which is simply silently discard afterThe solution here is to use the multipart requests and keep track of how much was already received before each data function call +The logical operator +Its probably easy enough with JNI to grab this system propertyor Statically link into libhadoop at build time +I think we need to try and expand this case so as to create a test case that fails for am guessing that the above code fragment is embedded in a function method or closure +on patch version +However I dont think group is something special if we employ access control against other attributes like role they should be important too +Patch applied +At least the error message This page contains both secure and nonsecure items +You are correct mlock is an implementation detail not relevant to the +Please verify +Created a temp jar file including tld files and added it to a classpath. +Committed modified patch +Fixed in . +I am attaching the client program which is having error finally i have to implement x enabled security for this as the web service is x compilant +I have removed all the comments regarding mobile showcase +No additional unit test is needed since it just changes variables to be private final static +Instead you should upgrade to a more recentrelease in which this is fixed. +Patch updated +bq +All the attributes in the are Strings +Fixing this myself +looks good to me +has been marked as a duplicate of this bug +asynchronous delete takes placeSo the NN would get contradicting information +Thanks Luke! +closing it as fixed. +Attach a same patch for branch +I noticed it when searching for the correct style to disable the button +Once this is stabilized we start to consider of requests as part of the server load +Changed the test to use returns method from +I dont think this is a blocker for +Hi Sergey Thank you for looking at this +It extends the add filejar functionality to download remote files +Moving this to the next release +So we will not support this. +If someone feels strongly about this they can feel free to run with the idea +Already fixed in SOA docs need to make sure it is fixed in ours +Wont be included in though so moving to fix unknown +Put the more specific ones first then the wildcards +My machine has only x packages installed +Examples are usually viewed by people who are just learning about MINA and I think if our examples become too complex people will lose sight of the MINA portions of the example +that makes sense +will enable use to use it in smartcn too +thanks +When the job becomes stuck the reduces are still running +batch transition resolvedfixed bugs to closedfixed +You should join the users mailing list and discuss your ideas there before writing a application. +Is this specific to vmware +After the patch it was pretty fast but like I said I didnt evaluate any correctness just yet +make it non blocking for next release +Ill post back if I hear anything +Attaching a patch with the changes to the interface and implementation +I am not sure how realistic it is here but it would be great to standardize more +Whether this will be implemented or not depends on whether going forward manual addition will be supported +Rebased and merged into the branch and into the master branch. + +Hi Laura should you look this please because related to Authoring Guide? But we are constrained by Confluence and Scroll export plugin features here +My bad +InWhen optimistic transaction is used In the commit in line secondaryObj is expected to be persistent non transactional so it is not expected to be cleared and jdo should not be called +To take real advantage of this we need project scoped repos +allThanks for a discussion Ill benchmark approach maybe its a solution indeedOne thing still is not clear for me how the batching factor could improve gets performance? The get request is synchronous isnt it? So in mapper I issue get to obtain value from large column and wait for it to be ready +It was laying below the pathIt also explain why the install started out by downloading all the found it now but why did this happenThe only thing i can think of is SP for Windows XP which i installed two days ago +Xalans is not thread safe +done +Stephen can you please describe why you are trying to load such a big SKOS thesaurus into Ontonet? Do you plan to do reasoning over the hierarchy +Is this still a problem? I know a lot of things have changed in the last year + +Ive started to work on that and it seems to work pretty well with maven repo +They are aware that this is going to go away in next major version but it will be some time before their applications will be migrated to the newThanks for having a look +Setting a target version instead of a fix version +Comment to reproduce the bug by j haleyJust to clarify from within this project pull up +I fixed this bug at +I concur that is should be move out +We can get the out result by specifying aggregation strategy for multicast processor +bulk close of resolved issues. +fixed in r +This time I clicked Grant license +Maybe this could be done similarly to maven handling which we dont support yet but is marked as a bug example of Ant I gave above also comes from the maven repository +It is not clear that Hadoop considers the record io a stable interface to be used outside of HadoopAll public features in Hadoops core are subject to the same goals +Its not a bug its a new feature as the component did not support that option before +Thanks for the offer +Yes that might be a useful alternative +Adding support for Comments +Here is what I think the test should do Setup creates a new instance commit the trsnaction and stores the oid in an instance variable +Good catch Hong! The patch looks good +We can easily produce class files that work against Java and Java as long as we do not utilize new Java features in Accumulo +We do regex matching if and only if the column name contains special characters +This is a duplicate of of. +Ive committed this +Which as you notified has been rectified now +There is an improvement now I can see the shortcut label when I on two apps to be removed +Ok this is part III have found a way to reproduce the error in eclipse +wont fix this in release +Apart from trying out things like the webserver in Java one thing which will help solving the GB limit with the current jetty version in Hadoop is to have Chunked for the HTTP responses if the size of the file happens to be more than GB +Whats happening rightnow is that multiple SUCCEEDED events can make it into the +The timeout is canceled when the oldest reply is received +Delete link works without any problem +I am willing to take the bait as I have a definite vested interest in contribstreaming and contribindex and I can whip up some stuff after you commit +Here is the +To generate the Makefile then add the following option to the CXXFLAGS variable with the using gcc did fix the problem. +I want to parameterize the location. +But it works on my ubuntu boxLet me take a look +Thanks Luca +I might have pastedsomething into the summary field when I submitted or something I have noidea +Given the gains for I think this is an OK tradeoff +It is not meaningful to show only the first entries of the dictionary +Using the from creates a mix of WS native versions +classpath and +To create the bundle which using hibernate using mvn clean install or mvn clean install PhibernateTo create the bundle which using openjpa using mvn clean install Popenjpa +Reattaching was missing file before +checked locally +Thanks a lot for the review Todd +Updated unicode from +If its not rendered then it wont trigger an actionId recommend using tsaveState on your backing list or making the list session orI think theres also wiki pages on this topic if you dont want to use saveState +Maven isnt going to support import scope so its Wont fix for this branch +Applied andThe remaining patches dont fit anymore mostly because Andrews patches from the last two weeks already changed a lot of Collections into ListsCould you check what is still missing now? Greg and Andrew closing. +Moving back to for now. +Anyway Ill let this task open maybe in the future with more time available we could try itbest regards is the benefit in rewriting all of that stuffWe do get new bugs new code to maintain etcWas there a thread where this was discussed +no regressions +At least not when creating a datasource +Fixed +It has now been and a half years and this issue has yet to be assigned to a developer +Roo will auto generate the case and the user can then easily refactor to start customizing +Its probably best to localize the assertion messages just due to our international developer team +It was buffered +Attached example for earlier version not to be committed. +One liner change in to include confdir variable +see the attachment for my patchset +Fix is committed. +After switching to the exception was gone and my was able to marshallunmarshall my know the exact cause but it might a jdk has been the bane of remote unit testing of session beans for me and predates java +The latest version of the patch also removes the strings from thevarious locale files +EduardoShould all be good now it was an issue with the CIGood job +This reverts the identifier factory to the same as datanucleus +If this is true that is a big bug +Patch committed. +I am attempting to migrate a live cluster to and Im hitting this issue +thanks tom! +This code also handles correctly the casesfor where the new publisher code has introduced problems as described aboveand elsewhere +You can work on cool. +. +I just committed a fix for this +So whats not working limit or limitRender +What do other filesystems that have a Trash feature do in this case +I was under the impression that seqencoded could be configured to encode several data types simultaneously such as body and lines from the News example +Behaviour of hosts should match in this regard maintenance mode should imply the host goes into maintenance mode. +Have to update the documentswiki to instruct how to install and run +No other changes were made +Sorry wrong resolution +Ill all system and drop definition from the method names + adjusted added in rThe intercept is the first element of the beta coefficients array +This issue refers to the BACK button on the for fixed this. +Minor comment if you use the same name for the patch file JIRA will automatically gray out the older one thus making it easier to spot which patches are out of date and which one is the fresh one to look atThanks for the patch this sounds useful +Fixed this issue added new macro to updated wsm samples s to use this macro +Its not there anymore +patch attached +Should this be in or +The fix has been committed. +I have committed this change +have attached a correction patch +But if YARN common lib changes I definitely want to run RM and NM tests +issues fixed +From Adrian Brock Adrian comment Will this also be integrated with the installerand the Network for ease of installationchoice +You can find more information in the log. +Thank you now I can reproduce it too. +this should result in Failure +Ahh right I remember now asking you to raise another bugThanks I will take a look before we release m sorry Ive made a mistake +Worth noting that the above overhead is in the neighborhood of +This framework spins up an accumulo instance to run test against +Should serverxxxdeploymanagement not be affected by the installer? Yes the file is the one instead of the one my r r r in CR +Updated patch i added a test to try to test all indexsearcher methods for thisit currently fails +btw the patch is ready to review +I believe there should be a version bump to whatever version this fix is backported to +Editorial pass over all release notes prior to publication of. +I think its also better if we fix some remaining issues in and make it public rather than return null hereIll look at this thing and see if i can get it in shape +Some changes had been made to class sinceadd describesplitsex providing improved split size estimatepatch by Piotr Kolaczkowski reviewed by jbellis for attached the patch +Something for the wiki +Also should this be backwards compatible against different lucene versions assuming that the indexing scheme in lucene is somewhat consistent across versions + +Isnt this a separate issue? This ones a wish for an Ant task +Here is a second patch that includes the jst upgrade from toIn addition I +Per instructions from Andrew I checked out the demo directory and updated the file to add the proper paths +Perhaps it even does what you want +Not for commit here +Mass closing all resolved issues that have been closed for more than one month without further comments +Ryota had run all test cases locally and they passedCommitted to trunk +As discussed on IRC sure its probably wrong but it did work for versions up to wicket and could be considered a regression. +does the command run under the same user as continuum? do you need a cvs login first +FO files crashing FOP +bulk defer of featurestaskswishes from to +Verified at +Yeah +this never got applied to trunk +Part of a bulk update all resolutions changed to done please review history to original resolution type +Once that happens well see how itfairs against our more extensive Kodo test suite +Creating new +The patch is small enough that I hope having it around will help discussion +The patch doesnt compile due to missing renaming inOtherwise the patch lgtmnit Maybe in we could keep the blockFor value as a field instead of the table which would avoid having to care about null checks +Okay Paging over an mapping is done in memory but in our case it is a mapping +Committed revision Thanks Noble +Hi SanjayAre you started working on this? If not can i take this issue? Because in our cluster we already implemented this mechanism +However if anyone has the time to start a plugins project on sourceforge and make this one available there under a suitable license I would be more than happy to make the code availableIve been meaning to do it for a long time but Im afraid it is pretty low priroty +I will research and assess the amount of work needed to be done here and time it would take and then also take into consideration that this book will not be part of JBDS from version onwardsThanks I may interject thats a strange notion +I tweaked grailssrccommonsorgcodehausgroovygrailscommonsspring reran the app and now the resources are found +Can you check this with? Seems like it might be related to checked it with Maven and it still does not work mvn site fails but mvn compile site runs OK +I think you the wrong commit +Could someone please try this patch? at r +This is what makes APP the DBA +Moving to TBD EAP issue has been triaged and decided to not blockpreventhold the EAP release +committed bens patch to revision in trunk and revision in branchthanks ben austin and chris. +Selenium tests have been addedupdate to verify the attributes described in Attributes missing in myFaces The Selenium tests run successfully using Suns RI against a browser +I think we should just suppress the warning +Thank you the patch is applied correctly +Reviewed and updated version fixed small typos and spelling mistake and used the English state for the Web Flow msgs as status is not the best translation +foobar project that replicates the only give a blind guess today but bets are it has to do with bridge methods again +We should consider a module but the axis module might have to adapt +Please include any source needed to reproduce the problem +On the same host when the app runs we are running cssecHowever I will try to gather more information with strace etc +for making the behavior of supporting concatenated gzip files as default behavior +Once restart is implemented this jira should not be a problem as theres no notion of relaunching a new AM in restart the old AM will just spin and resync with RM after RM restarts +ivy directories below your home directory and report back to let me know if that has any impactThanks +thanks for the hint john +Shane will look +Added fix versions of and given that this was identified as a for lets open new for the other as improvements and then close this one +This implementation is currently not used +Need to write up some docs and a release of +It avoids the iteration by maintaining a for the keyed by +However the error has not +Actually join and path implementations are fine but is wrong it returns false when it should return true +Package documented in EAP beta release notes +Corrected a spelling error in the description +I think its enough just to disable this +Hamlet since the AST controls are still not correctly getting populated most of the times and you are working on it I am moving the issue to your name +The unit test was updated in revision +Sorry I the code is the index for hdrrokenstrs which in this case is Viaso is valid +I believe we should probably have a similar implementation for Puppet +Sure no to generate release update with commit version when completed +This was fixed by removing the port rolling in cluster in. +Patch to kick off compactions after changing strategies +This shouldnt hold up the release +Theres nothing accidental about it +Fixed broken link of blog Add git mirrors information Changed header can update more after releaseIm committing now +Pull request processed. +Heres a patch that eliminates the extra calls to get as much as possible + +So this cant be filed as a bugOn the other hand a lot of people are successfully using with +Thanks I could reproduce it now +Do you need that for tomcatrun or for the exec war? or for both +Without integration test however I think that unit test is enough in this case so I close ticket. +Its possible build number was reused by Jenkins hence one of the runs may have had a failure but we do not have the log for it is a jenkins issue +Looked at the test case also +Ok i am closing this as wont fix then. +So Ive done so far is add some checks for nullity to fail earlier should the same problem happen again +Hence closing this one. +Julie Please verify +to trunk +one application having visiblity in another applications channels +Updating parent +cleaning up. +The clean up of default egress rules is not needed for VR when we destroy the networkFor other network providers it has to be implemented in the shutdown network of the Firewall Provider resource baseI think the rule ID case is specific to palo alto +my only problem was with the error caused by the jdk and libraries jaxb +classpath committed into the repository too +I have a patch for it +Thejas Had a look at the patch and found that the ANTHOME path is removed +keep some int constants since javah cannot understand enum values +We have a catalog with tens of thousands of layers reloading them all is too expensive the reset would only kill the store caches which contain tens of elements much thing would be to accept a layer name under the reload endpoint +Added new interfaces in r +Fixing will resolve this issue +Theres something else going wrong that mysteriously is linked with having multiple virtual processors +Thanks Namit +Shouldnt be too hard studying what goes on there to figure how to get a little test in +fieldName ordering type min field string class +Committed to and trunk +It will make jaxb model with high coherence +On the other hand it must have done something useful at one timeI also think that only existed in JRE and perhaps only in Suns version of +The patch also enables empty intrangesets +Confirmed it lookes like a jbossxb issue +This is the first Ive heard about the need for an email server but the localhost on the labs box will relay mail through our production SMTP gateways just fine +I prefer they create mirrors locally to Fisheye + for the patch +Maybe I am wrong +For the first pass just is supported +And is there something you want to do on the AMQ sideIf not we should resolve this ticket +Committed revision . +Done! +This is a separate binary but might as well become part of AS specific implementation of that SPI which includes the EJB integration has been moved to the jboss code base +Whatever minimal dependency can provide that Im fine with using +Besides finding help text more efficiently we need to be able to associate the help text with the label on the screen instead of the detyped name +Patch to fix the bug +Fixed rev . +SMX already uses CXF +betterIt is acceptable from my thanks for checking this! Although Im not done yet +So hopefully well be able to get this work integrated before the dependency scopes are set and before the configsassemblies are converted to M +Thank you for the report +I didnt want to actually define it as using as the API is supposed to be script engine some extent I agree with the earlier comment that symbols are similar to enums +bq +No progress since last move and there is weeks until the Beta release is do +Thanks for the comments Mike +I know its a pain maybe I should submit this as a patch to selenium? Im attaching an updated jar which includes jquery based on +By the way Parsers dont kill people +relnofollow implemeneted in version of plugin +Can someone please take a look at the code +This change moves the methods to the interface is s public API +The appropriate place to change the host is in the server editor NOT the launch config +Josef can you verify this plesaseSee my comment above +Committed revision are some more tests that fail with only in contribanalyzers +Also the webUI can query the namenode to list the identities of the Secondary Namenodes +fix affects version field +The new version makes the DROP SEQUENCE action clear the sequence cache +Why not use the same mechanism as the pid file + +This patch updates the documentation profile assembly for profile criterion and database scripts to insert the profile rule and a new user with the profileThe new user is geo it is a generic user for demonstrating the future geospatial features of jA client application for selecting a geographic location will follow +What do you think should be shown other than ? I think correctly identifies an error and its not a bad thing to show +reduced issues that arent going to make it into to +Also beware that the code is not up to date +Even if the can authenticate users they still need to know whether a user has the right to access certain data blocks +Also the assembly plugin attaches a secondary artifact +I am curious to know whether the previous fix fixes the problem mentioned in the description or not +Specifically for this last is false while for the earlier records its true and by the comments in the code its not legal to add record after interposed ones +This is by designBefore a message can successfully be createdreceived by Camel all error handling is component specific +That being said I went back a number of versions before the refactored anonymous checkout was presented and the gift card did not work in that case either AND paypal was not an option out of the box +Thanks Arthur for your valuable comments +Todds patch doesnt seem to require much polish to me +It expects an xml payloadI have fixed this issue by doing little chnage to service client code so that now I can doechorest mGET httplocalhostaxisservicesechoechoStringtextHello World without creating a om payload by the client application +invalid is not a proper resolution for an existing bug we just cant fix. +I could go back farther if that makes sense but the chances of another Derby release for and below seem very small +Verified by Svetlana +Patch that include an extra info tracing level that must be enabled to log the message +classpath file +has been resolved so if there is a test run available pls let me know +Would it be easier if I asked Terence to attach this version to the wiki page as well? it d be really confusing if people could get two different jars for the same version of gunit so Id really discourage the version on the wiki isnt marked as version +There are two binary distributions available Whats the main differences with them or add info on which flavor to choose and why +Thanks +Im not really sure what the problem is that the assertion ve figured that the problem appears when reloading of project can be multiple reasons for that +Unfortunately the rules for Noun conjugation is much more complex in Norwegian than English and there are many irregularities +patch looks good +Having the special jruby but not special jrake is a consistent consistency question +At one point we looked at moving entirely to but its about x slower than using Double +I like it and I think it could be easily scaled for mobile too +Sorry Gavin I only noticed the assignee setting once Id already done the update +The attached patch has reduced the import time from about hours to minutes in one project +I Believe this was fixed as part of but I need to verify +Matt FYI +Yeah Ive definitely seen this as well +Havent looked at the code but the first thing Id try is using a fullabsolute path to your dictionary file +theNice +But the same was working before now it got broken +Or someone can propose and instigate an alternate process. +I shall find a fixI ran into some issues when I used the class +As you mention in your previous message a viable solution to this problem is not readily apparent +note implementing this as a freemarker macro would be cumbersome but it would you deal with this in hibernate tools then dan +Camel can now auto lookup the so the PROPAGATIONREQUIRED spring bean is not needed anymore if you want to use the default settingstrunk the JMS component as isLast piece off code committed all tickets +Both logging and authentication are done in other places besides from HTTP servers +Enforcing that claim might depending on what license is required when it is enforced +Part of parent as part of +Done in trunk +The patch is applied +Preparation for Gora release candidate +Otherwise youd often be tuning in the dark +Input focus has been lost after Up button of a bulk update all resolutions changed to done please review history to original resolution type +Fixed on trunk r +We have corporate counsel we need to use them before publicly publishing legal policy. + +The testcase launches diff r and hangs when the output buffer is full +Another option may be to downgrade the log statement to a warn rather than an error +Our intention is to track the Google Guava version that the AS team is using +Now say we are going to use unamanged AM for debugging +Closed after release. +Committed to trunk at revision We may want one more test cases to cover these annotationsInterceptorNote that interceptorinterceptorBinding is optional according to the spec +Omitting JAVAOPTS in those situations would give the wrong resultsI suppose we should search the JAVAOPTS for this modifiers and call the java version with those params only +page files but you need enable this feature first manuallyIn the Eclipse go WindowPreferences +OK in CR. +WernerI think its more important to fix existing bugs than to introduce new functionality in RC +The compiler including the command line and maven plugin supports the processing of annotated Groovy classes just like Java classes without any additional configuration +I just committed this +That is the purpose of the staging site to show you the resultsof the builds before they are made live + +The delta is nothing but the difference between the last call to read the delta so this metric is not entirely useful +but that attribute must be defined +AFAIK there was severe reluctance to accept that the pool was being used in such a scenario and there was in my opinion no interest in any fixes +In other words building that should not be mandatorySamisa +If the IDL says binary then you should write it through with no encoding +But it is not easy for it refers to and Task which then refer to many othersSo i think this would need a refactory of packing +Minor unused import added in import imports no longer needed in Otherwise on the patch +Hope this helpsSorry next time I will post the complete stack trace +Id love to see the package reorganized so that this kind of thing would be easier +The container Maven Dependencies references non existing library CUsersBurr +Will close if no response in a month. +SSL should be a childThe model has changed then +Patch with comments incorporated +latest patches applied in revision Dmitry Alexey please check that it was applied as works fine for meThanks. +I added javadoc saying this could be one of the reasons for an USE +srcmainjavaorgapachehadoophbaseutil +The best well be able to do is shuffle things so that the operates earlier and works against the rather than the Request +The quickstart publishaswebserviceinonly is not in CP or CR +Hi Ben +If there are no objections I will commit the a patch tomorrow +bq +For sure it helps with the above situation +Peter is also a Groovy committer +Ive noticed the same thing on our internal build servers just hadnt got round to reporting it yet +For we dont get it from maven because hadoop has not uploaded their jar file to maven +low recordCount +If it is now going to use the already used numbering for product then there is no migration required +components inside the enclosure are visible but not rendered +If necessary uncomment the line and restart etclibvirtd +I think this should be represented as a policy in UDDI V. +patch was commited to revision +We added as a dependency to our project and the missing came with either that dependency or one of the transitive dependencies from the module +verified in Metamer httplocalhostmetamerfacescomponentsrich +When the stack trace is printed because the initialises its super with a null descriptor toString throws a and the useful diagnostic information is outcome stack trace contains useful information including the name of the type and the reason for the outcome trashed stack trace and with no message +jar is under +We failed to retest what looked like a harmless change +Committed to the branch to fix the test case +The reason I add this is because which inherits +If we want to get some real projects using this before releasing should we toss it into samples first +AlanThanks for your help +The example signature is notvalid according to the xmldsig schema +Looks like Windows does not like this oneBUILD FAILEDCJenkinsworkspace The following error occurred while executing this lineCJenkinsworkspacelucene The following error occurred while executing this lineCJenkinsworkspacelucene The following error occurred while executing this lineCJenkinsworkspacelucene Unable to delete file CJenkinsworkspacelucenelib is creating a lot of noise +Updated patch to fix the flakiness +Cant we omit the when the commit is not for a branch? mongoSync should already contain the head revisionBtw why doesnt just read head in the sync collection? Is it because head may also point to the head of a branch +Thanks for the KUDOS Im glad you got the issue resolved. +The intent is to consider the outputDirectory not as the directory where the classes will be put but as the root for the main and source output directories which will be kept separate +We only need to write the final one to disk +Usually running fixes the problem inthese cases. +Thanks +Some time ago have opened a bug to track this +Committed to branch thanks +Hi MartijnThanks for your help +So yes fixing those files would be a worthwhile effort +Works now +I suggest we just fix the broken tag and let this one sit for the day when we decide to remove projects for depending on snapshot etc +I just think having as part of examples was handy as the audience is different from test +I tried to set true at the following places Inherited Project specific Command lineNone of them solved the issue +Thanks Ryan +reqsec reqsec reqsec reqsec reqsec CPU +Uma great point about Trash not extending Configured +Thought this approach would be unsatisfactory from a mathematical point of view +Initialize ugi earlier +Should clarify version is as of this weekend +Committed to trunk +IE is the only browser having the issue. +OK +Passes on SUNStill fails on DRLVM because DRLVM reports owned monitors in the opposite order relative to SUN +In general the policy of ignoring all blacklisted nodes after the threshold is reached seems incorrect +Hi sorry for not respondingfixing Issue in mpparser +Hi RobertSure no problem +What do you think + to trunk +Attached a diff of the testsuite module +If you are using in Spring you cannot use add repositoryService or executionService in the application will cause a cycle reference in will append the on the top of you invoked the get it will look up in the but the first environment is it will go back to application context and find out a bean named repositoryService declared in the application context but this bean is not yet will continue invoke the to create this bean and go back to jBPM then there is a endless you want to use in Spring you should remove repsitoryService and executionService from processEngine +One of the step in modifying a table is to reassign all regions +Otherwise we exclude classes from publishedapi and derbydocs and we use the javadoc tool under JAVAHOMEbinThis patch touches the following filesM A toolsjavadocA toolsjavadocM toolsjavadocM toolsjavadoc +This would have to be updated to trunk and then benchmarked before commit +Oops +bulk defer of unresolved bugs to +I vote to move to anyway if its in Contrib +The CONFFILE and MODULEDIR macros are currently set on the compilers options with different values for qpidd and libqpidclient +I did some changes to the code and get this working +Same here +Revision +Thus if you do not first make Test known to GSE by running it GSE and the compiler will be unable to resolve the Test and Test dependency in Test +closing cannot reproduce +Need to figure out how etc is being created +As discussed in irc +Wow +bq +Since this has been fixed in the trunk I think we can mark this as resolved. +This makes total sense to me +This allows a secure client to work in a heterogenous security environmentAs I said if this is the design the patch is okay +Please change to if you intend to submit a fix for. +Other than that +Someone could come along and implement your feature request but no one is under any obligation to write your software for you Since its been several months then you probably need to implement it yourself +per my earlier comment patch to latest trunk +The attribute is removed from the attributes of the component +Committed to trunk +Saxon XT and Xalan have no problem +Thanks Amareshwari +TBH Id rather hold off on this until we have a concrete use case for it +Here is corrected patch +I was a bit surprised about the magnitude of this myselfI can try switching one of my cores off and measure this again if the issues is memory fencing we should not see any improvement with one core only +Im not sold on any of the bugfixes being release drivers and the JIRA version only has a few items that are there because the code was deleted in and a fix was also not readily identifiableDepending on the snapshot build would provide very useful feedback or you could take the source and fix the one bug affecting you in a local version. +Triaged for +Just remove the testcase for +This logic would also be applicable to staging tables in xml planning +Should we do this for +I guess that would require some more surgery to get that considering that is the default in s hostname + +Two questions why was the merge strategy overridden? Whats in the patch is basically the default minus a few cases +Verified by Spark. +moving to problem is that Servlet version needs project facet of version which is missing in the used version of WTPShould we try to create available version of to work with servlet version and +In generalmore config more confusion for user to use itmore config more confusion for developer to maintain code +Looks good +np closing the ticket +Once deployed the distribution URL is no longer of any use anyway +Marking critical so we dont forget about it +leftover from is a duplicate of the other +committing shortly +for the positive feedback Mamta and Rick +patch file for here is an additional line in the stack trace that mentions specificallyINFO Cannot serialize session attribute for session EBEDCBAEEAE +Thanks +Verified +As i said i have added the Simple language goodnes support for operation and key parametersHave a FUN +Moving this to M +Its impossible because for now Hudson Sonar Plugin uses internal classes from Hudson to invoke Maven so if you can run maven from Hudson then Sonar Plugin will work for you +This looks like its working properly now +Marcus simple JIRA for u +I dont know what the th field default could mean +Looks fine +This will definitely take up all bandwidth until the release and possibly you want to help out with this you are very welcome +Thus this is something that may already be fixedCannot reproduce with supported versions of CXF and request for testcase unanswered for over months. +The regression tests ran cleanly for me last Friday except for a known instability in the stress multi tests +Committed some JPA updates in revision +The problem is not with the database server the server DOES the value peeking when it is doing the prepare +The onclick should call javascript code +updated code against latest API refactoring the whole build pass +Could You please Your tests in Your environment with fallowing patch applied +What device is this happening onAlso remove the code thats trying to set the orientation thats probably whats causing your app to crash +But now Im hitting some problems while our bucket attempts to enhance the various classes specifically the class +The output data may be deleted anytime when it is no longer needed +See for ongoing discussion +trunk +Have you done interop testing with +actually flavios suggestion can go in quite well +Tucu Im a little disappointed +A fix is in SVN +This is really a bigger in issue where a set of threadsafe components are not logging to the originating thread but for the time being this issue has been resolved for netbeans +So I dont need to change my source file headers from their current Home of Professional Open Source +actions added +A java db would allow us to get to a point where you download the thing and launch it that would really get this project going +Committed revision +patch build +If it doesnt please post a new test +properties file +That is why I reply on to verify if the events have been received or notAnother way to test is to put a sleep in begin method and make it long enough until the event is receivedCurrently is a simple and reliable way of telling whether the events are received or not +Could be +Making the project easy to import in eclipse is not done by only adding the me stuff but only by cleaning up the entire pom structure + +It eliminates methods that test how many times the task has been run and ensures some task to be run using a loopWould you pls check it? Thanks +so i have said that rd part plugin will load some fields in +Thanks Paul! +If the names or meaning of binmahouts and options is stable vs experimental this is very useful for other people to know +Michal one final question I see driver threads in each client +Looks good to go +This is same as what the spec says +Any updates on this case +WDYTI fact I dont know the purpose of this feature +for directory creation passed the umask value as +packages file inside eclipse and edit it directly OR if you edit it outside eclipse and then refresh the workspace projectThis is a design decision +Note that this is not a true profiling issue but rather a consequence of using the security constraintspreferences instead of a profiler approach +Awaiting patch from user no access to HPUX on my end +Committed +This is not a bug the solution for the problem is updated in the bug comments. +Can the destQueue not be made Transient? Being able to monitor the Queue level via JMX is a pretty important feature. +You can use an Any component to simulate this I believe +Ill leave the exception mapping in my mapping layer for now and well have to worry about fixing either or both or neither products later +This diff breaks long running conversation restoration during JSF Ajax requests +just uploaded a new patch +However I dont know if its a good idea to allow cycle to be created when injecting aliases on an ancestor +You cannot get location paths from WAR files unless they are exploded +Good luck with the release +So to me this conflates the two models completely and also requires extensive rework of the documentation +I attached +In general it looks good to me +I had thought about eviction +bq +Closing again as fixed in . +Priority changed to minor as suggested by Bastiaan in his mail +Committed revision to Committed revision to trunk. +Ill update the docs once the patch has been applied to SVNBabak Thanks! Its easy to contribute when you guys are so proactive about giving feedback +Hibernate is much better at not needing JNDI. +Sounds reasonable to me +I have opened new task to track the other issue +There are smart ways to lookup the job tracker history based on the directory structure ensuring that you do not bring the system down +Are you using Camel on the client side as wellIf for instance the Exchange is sent over the wire as then it requires that you have the same +Attaching +Verified +In that case we have no option but to fail +I think this is an invalid bug +Thanks for the fix Ashutosh. +Drop partition get because of this designDrop partition is a base functionwhy this design exist if a base function can not work +Patch Prasad +defer all issues to +But we already have issue +No worries Im still in the thick of it plenty of little strings to wrap up just wanted to make sure I was on the right track +thanks +There are no memory leak fixes committed to github after was released +If were either positive that this is strictly an illegal namespace or we decide that Xalan is never going to support this kind of namespace then we should just close this as INVALID or WONTFIX +This works for me +Patch attached Thanks Pranay Ravindra PranayA different version of your patch is in rev +Add columns Type State +Verified by Sian +verified that the given commits are present in +To avoid splitting arguments on spaces set IFS to newline at the beginning ofthe script like so export IFSnPlease try this and report your findings +alternatively we may move out of this container and treat it independently +As that turned out not to work I tried hard to understand why and to bring it up including a new installation applying the war target +log file and missed theall thanks for can confirm that this fixes ZK launch +We have a lot of that logic in place but most of Array works with our internal class directly rather than the class where we have the encoding logic +This can save the store testing every product that is returnedThis way they can poot it back on shelves or return it to the manufacturerzoly +Not sure what there is to track as the issues was related to an incorrect documentation example in the Hibernate Community documentation +So all issues not directly assigned to an AS release are being closed +Im not sure I understand this issue correctly +passes with all its existing Normalization tests +Not sure how much work was done to also support www only +Change SQLLINEOPTS to this to use embedded modeSQLLINEOPTSu jdbchive d n saOptions such as this can also be overridden on the command line when invoking +Is this the way of starting httpd that would trigger such problemstmphudsonhttpdredhatewssbinapachectl k restart d tmphudsonhttpdredhatewsetc cant find xmemalign in the build log +lets use and for the interceptor configuration names +To understand this point consider when the response actually gets committedIt happens when the response is flushed or when the response buffer fills up +check revision to apply this patch i getthe chunk size did not match the number of added removed linesany ideas? Did you make this patch differently then before +I think we should break this up if it is a story or do you envision subtasks? A feature that would work just fine is to support just the uploading of job configurations that dont require any additional elements in the classpath this would cover a wide range of jobs type for spring hadoopDeployment of custom would fall under the epic for that +I dont see the null check for keyvalue being the beneficial part of the optimization +What if i alternatively return Integer max value rather then +Since this is quite a large patch and touches some files that are changed frequently +Hope fully this should be fine +wrong bug +Nice start SimonBesides tests we should probably make the max a parameter tooOne issue that rears its ugly head here unfortunately is that the Surround parser does not analyze any of the terms thus stemming at indexing time will cause confusionproblems in trying to use this parser by most folks Id presume +Hm didnt understand u +Can you please also provide a sample +Did youtest the tests against the RI +Found the problem booter should include the still have the problem +No this lock only protects concurrent execution of twoIts the same as synchronized method except that the latter produces bogus findbugs warning that will have to be suppressed on class level +Attached a proposed patchThe patch is pretty big as its the result of about a dozen refactoring steps required to replace all references to +we should look for uses of the artifact and at least warn if there are anyI think the functionaly is sufficiently useful to add anyway. +The issue is in the creation of the prefix variable where an additional dot is added +Id made a conversion of types rather than changing parameter types of functionsIf you still not convinced and prefer changing parameter types then change them consistently please + +It might be trivial to implement a JNI library but I dont think its going to be trivial to build architecture dependent code at release time field issues debug etc for N platformsIm also open to alternatives but Id need to be convinced that the cure wasnt worse than the disease +patch for testsAdds test defs to and creates to xwork srcjavaBasically if no class is given to an action definition delay giving it the default class from the package until that action is actually asked for +this patch changes the Boolean returns to boolean +Removed test +This should also allow you to add the facet to an project +the application context path should be prefixed to the path given by the image attributeI could not find a note about this in the spec +Credit of course goes to Dick +committed +I have the same issue +This method can be implemented by the persistence provider to support Java EE by simply delegating to +This is a first version of the patch that still has some debugging messages +If we find it is not in other JDKs we can disable the test on those environments +patching the log at line CarstenLooks like there is much more than you said in this patch isnit + is a xslt template that will scan for sections that are missing an id attribute +Fixed in trunk +I dont think that well have adequate discussion from all parts of the community by just having it on the issue +Actually theres a few compilation problems but some of these issues are known and dealt with as part of existing issues +documentation has been fixed +urgg just ran a test +We have downloaded the latest version of RF CR and it still not working. +Maybe you can give it a spin to make sure the services work okCheers where are we on this one +Nice thanks for the patch +Feel free to provide a patch with testsIt would be great if Matthew P +Please let me know if I was doing something wrong +This could be very slow especially if there are big andor slow update sites +Yes +I understand that security remains enabled in the added unit test since a previous test sets it to kerberos +Is it possible to commit it to upstreamThanks on EAP ER +and as well as rev +We can build a small framework to allow passing parameters between plan operators +Bulk Close for all issues before incubation +Shawn I have been reading though your comments +Did you try with the xsino like in mine exampleThe doesnt define a target namespace thats why you cannot put the elements inside a namespace +Patch committed to both trunk and branch. +This patch corrects the logic to respect the queue timeout +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +now we generate composite keys in the recommended way +Another good news the class doesnt contain anymore code related to fop so the very last issue is the classapplicationscontentsrcorgofbizcontentemail +The first part is completedev general users mods added to above lists as requested +Moving our MXfrom OSUOSL to UL seems like a step in the wrong direction and since the current MX has no problems coping with the current traffic load this issue is IMO not a top priority +New patch is coming the current one doesnt work +That way the actual cache data structure could be reduced and the memory could be reclaimed +Yes this is not optimal + the patch looks good +then no need to touch interpolator any more the format strategies will be applied automaticlly +is not needed see the comments on and + does the jdbc test suite catch this at all +Fixed. +Quickstarts helloworldejbws and helloworldejbwswss still cannot be compiled in ll split those out to a new issue as they are nothing to do with ESB if they were written properly then they would have been fixed by this +This bug can now track the fact that a compressed message not containing a results in no errors logged or returned. +Please check the patch thanks +You can imagine also automatically creating issues in bug tracking system +Ah ok I see it now it was from a recent patch +Incorporated oneoff jar from +Somehow the compiler is recompiling these sources and ignoring the classes on the class path +With this option the definitions with include a namespace for encoding whose prefix will be used in the place where the undefined name space is used +We can look at the table status in the UI every seconds but that will use some cpu resources +The Shuffle was substantially rewritten for by which also did a big refactor +Thanks Lars for jumping in and Ed does what Lars suggests work? If so I need to doc it +Here is my version of the plugin not yet made in the hibernate versionMuch neater now reports can be included mostly automatically. +If you want me to assign you the issue just say and Ill add you as a contributor +With our Teiid Designer release weve removed our use of UML feature and added a Server s our current list of dependent features +Is anybody working on this issue +revision trunk revision on JBDS +This issue can be closed after applying the patch +Would love to see this addressed in earlier rather than later. +And it has always worked in that way +Looks good to me. + +Track down the closures andor +Since the fix worked in the code Im developing closing ticket since I believe it is complete. +See the simple README in this zip for instructions +I need more help +Again once this problem happens a number of thingsare likely to break and the real problem is not the insert but what allowed the allocate of the free pages to finish without the space ondisk being forced +Tom sorry for the delay. +Needs some testing and probably a few tweaks before to branch +Pending resolution +I have modified the patch as per request +Yes this version also works without schema defined +And as Ravi pointed out it appears it has been observed previouslyThe failure in could be explained though +The thorough job would be to check their distribution packages +Polish not a bug + for patch +Sending jbideployersrcmainjavaorgapacheservicemixjbideployerimplSending jbideployersrcmainjavaorgapacheservicemixjbideployerimplAdding jbideployersrcmainjavaorgapacheservicemixjbideployerimplSending jbideployersrcmainjavaorgapacheservicemixjbideployerimplSending jbiitestsSending jbiitestssrctestjavaorgapacheservicemixjbiAdding jbiitestssrctestjavaorgapacheservicemixjbiSending jbiitestssrctestresourcesorgapacheservicemixSending jbimanagementsrcmainjavaorgapacheservicemixjbimanagementSending jbimanagementsrcmainjavaorgapacheservicemixjbimanagementSending jbimanagementsrcmainjavaorgapacheservicemixjbimanagementSending jbimanagementsrcmainjavaorgapacheservicemixjbimanagementSending jbimanagementsrcmainjavaorgapacheservicemixjbimanagementSending jbimanagementsrcmainjavaorgapacheservicemixjbimanagementSending jbimanagementsrcmainresourcesspringSending Transmitting file dataCommitted revision . +Moving this JIRA to ER +This way a site admin can prevent users from taking actions that perturb region assignment if desiredAgreed on the possible affects on the global state of the cluster for some table level operations + if add a small sleep in wait +Extended the testcase to pass a file in and validate its permissions because distributed cache files are set permissions during localization itself testcase with distributed cache does not go through the code path of setExecutableAll contrib tests passed +Would help to work out a solution +The patch posted on Nov fixes the issueTo verify this I ran a hadoop cluster of nodes with both and + +It looked to me like it was from python wikis etc +bulk defer of issues to +We should enable intermediate compression using LZO +v. +For me it only happens when I clear out all the rules and then add am not sure what is causing it perhaps there is a bug in the partitioning +We may simply pass received blocks and deleted blocks in separated arraysI think this is a bad idea +I will update the documentation to clarify this point +The issue appears to be at least in part due to the user implementing a Proxy that does not implement HTTP Challenge Response correctly though I have yet to get a detailed HTTP trace for debugging +Was done +No problem +Patch that uses +Thanks +In regards to the last admin removal this will be covered in the dedicated issue +Cool +Im confused about why a client listening to continuous changes should care whether or not updateseq changes without emitting a document modification + +The testing plan does seem to be localized to only phase seems to only cover integration testingIm not so sure about having further schedule +Implemented as described above. +Hello Would you please try my patchThanksBest regardsAndrew +Fixed in svn trunkCould you give it a trythanks! in +Committed core part in revision this is the rest of it minus anything trickyi will leave the hard stuff to uwe +If I know that my server has a limit of concurrent connection set to then I can never have more then concurrent downloads or uploads right? If so than shouldnt the be tuned up to have no more then active threads for this case +An internal private class has been created within called +Additionally torque ignores the exit statusIf this is acceptable to all we can do the same as wellHemanthI could certainly merge this in with Im busy dealing with svn merge issues right now and dont want this held up +Test case for you explain what youre using and this package for at all? is sort of weird not used by the rest of the code base except for a few constants not well documented and overall just not very clear on what it is or why it exists at all +Lacking any more info on this I resolved it as proposed earlier +I think nobody is currently working on porting these to +fixedCheers SaschaCommited in trunk at r +Final patch for review +Thanks. +Need to fix +I will add it to the root README. +Attached the corrected images +deprecated the cache thread to avoid asynchronous cache eviction +It will be possible to create unique constraint over nullable column in upgraded data base +Finally with Ivy we do not pretend to be backward compatible at API level +But it seems is only used for test +as if DEFAULT is set +Please phase the above line so that it is easier to understandIt works the same way as is there anything in particular you think needs more explanation +BTW were using our own builds of both the ICU and Xerces source +Thanks very nice +Possible this is an rvm bug but I figured Id file here anyway +Its fixed +Besides this patch includes a unit test to check if default values and settings are correct +Len Jervis Im trying to understand this issue a bit better so I can help out Denny on +Commited in revision +This sure looks like a case of if timestamps are equal all bets are off to me +Previous patch is missing newly created will NOT work +WolfgangSounds good to me but I cant find any license information neither as header nor somewhere on the website +changed Affects Versions to and Fix Versions to Jespers suggestion added a conditional macro named jdk skip jdbc ant target if under java +Thank you for your time +However since the changes can have a widespread impact I recommend widespread testing of this feature AND of Portal in general +himake an effortI wish you succeed and enjoy in itwith best regards! +Thank you Chris +is not a core concern for me personally +Correct you should not expect this check the EJB spec +The real request is for an option for daily updates instead of a steady stream of individual emails. +I have svnI just tried it and cant reproduce it either +The only way out of this would be if we can validate such requests at the time of the allocate RPC itself and throw an invalidrequest exception +Just a test for a search filter. +I think Flume needs at least the tests jar to use the minicluster +Yes however keep in mind that it is very +Not reproducible now. +Thanks for the report the and the hint toward solving this problemFix committed in revision +This patch fixes the issue +releaseWill correct the Fix Versions field when +I ran into this too thanks for addressing the problem +Im afraid not +Yes I do remember that I had similar problems using JDK +It needs to be done when the patch queue is small because otherwise itll probably hit a conflict +Why is this jar part of public API javadoc +Same as the other report this is unlikely to be fixed since the new UI completely replaces the current expected behaviour would be to have gray areas in the bar that do report disabled datastores +It has a parent of the which does have a and a configurationDue to this a new is created and it is reconfigured with the configuration from the propertyYou can actually see all of this if you enable statusdebug in the tomcatOh and the is also theIm going to have to test with but my suspicion is that it doesnt make sense to choose the current class loader if it has no configuration and it has a parent that does have a configuration especially if that is the +A reviewer should probably apply it themself before committing it +fixed a few more compile some of the collector stuff for the sequential case +Closing as duplicate of. +I believe Ive incorporated your changes minus the use of srcdir since Im not sure how to reconcile that with XALANCROOT +Reopening +Yay! You rock Arjen! Thats perfect! I cant wait to try it out when it becomes available. +Im submitting a patch to this issue +This patch includes the shutdown command +What exactly do you RunDebug +and passed +I am about to commit this +and intentionally cause to test how the are handled +patch which adds predictor support to added the proposed patch in revision +The decodeIdentifier method is only reliable for the issuing services server +For instance if the source instance contains a tuned file the cloned instance should contain the same thing +Only just now I noticed that true is actually the default for which makes it even weirder that configuring the default requires extra configuration thats not needed when you just rely on the default +So unless we have a JDK distribution of then we cannot distribute EJB with it +When the nestedinner rowkey evaluation completes it erases the state of the outer rowkey +A +The outputs in open source JDK are also correct +Could you please try this patch? Seanpatch applied at r w minor modification on testcase please verify if it fixed as you expected +It seems like the user should be in control of if these docs are added sorted first regardless of what the regular sort is +So any synced data should be available to any reader if they were available to a reader before dfs was down +on patch for patch as well. +Really treat the symptom this time see what else fails +Implementation from +Rebased and merged into the master branch. +Emmanuel Demey fixed this a few months ago. +Rebased and merged the new tests into the branch and them into the master branch +Fixed since the creation of all Web Fragments and reuse of +The users list is the appropriate place to ask such questions. +The routine of Guvnor tools installation on Eclipse is also this is specific for JBT +So that should also fix right +If you visit the url in the original description you can get the background into the schema itself +It will enable me to have control over the source +However probably Igor has some ideas. +This sounds awesome how do you use it +Scanners removed in +Hi Yaron currently the page map is removed from session respecting the least recently used sequence +If ever +It is possible to enhance this patch to create a new data strcture called rack at the very beginning +Please post configuration questions to the mailing list +Can you please provide a patch +Will fix in +This means it will try and include the libs directory for every single module which causes the stackoverflow once its been added to the project multiple times +where all backslash have been removed? That would surely be a problem in that case +The fix is for both client and server side +It used to fail on my randomly in the past but now its more frequent hopefully it will help identify the cause +bulk defer of unresolved bugs to +Namit I added some comments on phabricator +Todds original patch rebased against +Attachments estimated time h +Ive checked it didnt find any code listings + Bug has been marked as a duplicate of this bug +fixed per correct my earlier two comments +Thanks for the patchThe reason why we needed to add the max memory of M was because this Ant target exited with an on Windows for some reason +teeing up for next release +This means that when its asked for its object type it has no idea and therefore returns Object thereby causing the match to explains why breaks component scanning has happened too early +patch be committed +Moving out of +So are you finding the payloads not at the right term if the terms match ieterm term termYou might get the second terms offsets for the first term and vice versa +currently people can migrate their apps from to in about a day or two +Updated patch changing classifier from h to h based on Alejandros review patch with comment explaining move based on Juliens comments in reviewboardJulien Need a from you for the h to h change +eq is the same query the jboss plugin uses if this also crashes will be much easier to debug from the command line +Removed all the variables and added commentsAttaching a fixdo evaluate and send comoments +This is the DB look database is based on open source data so I have no issues with uploading it +However I think we still need some more work to this patch before we apply it +This is a patch which resolves the issue by including the header for attachments when converting a document to a stream as well as including the length of the headers in the computed length of the stream +See to reproduce and fix +We should think of further improvements +I tried different revisions from the repository including the latest checked the source for ql etc +Modified patch with more comments + is the loader pair for have you tried that +Marking these as fixed against rather than against. +Thank you Nicholas +You could never place an odd number of consecutive backslashes into a slashy string +How large is your file +If this still interests anyone please reopen and post a patch. +Much appreciated by the community +I have a patch similar to this one based on +Id prefer to see this committed soon so we can take a look at the overall architecture and see what parts we want to move into a module +Ill create more issues for them. +Fixed via symlinks +Both jetty and tomcat DO work fine for some definition of fine +Not sure if the old chart data is being released properly every time you render it +Since Oracle stopped making new releases of the JVM public and the old version not up to date wrt security patches Ive stopped using +Ruth I think I can understand your patch for the but I think its not complete for example after applying the patch of seems reasonable now but the assertion from looks like same case but without the while loop the only significant difference between the two cases is the first one waits for ms while the second one waits for ms IMO they are almost same unreliable +The code appears to try to search dynamic imports +It significantly affects existing code for no use case benefit that is immediately apparent +Hi commitersIs there any one of you interested by this patchThanks at rThanks Pierre +Correct? If so this wasnt completely clear from the if I just run mvn releaseprepare the release plugin doesnt prompt me for the release and development versions +Added support for separating integration and unit tests in the main pom +This is a change from current behaviour but Id argue its a bugfix rather than an incompatibility and Id be very surprised if it creates any problems +This can be resolved with a rudimentary fix of changing Line to not use the absolute file location from access +The patch now includes the missed changes +Christian can you tell me why this isnt set by default? And what the implications would be if we were to set it by default +am i doing something wrong +The test holds events and starts separate thread that waits for events during default timeout +I dove into the code and found the exception birth intoWe can get the type of error with but extracting the actual data need the exception text which is not so pretty +I dont see any comments about failed tests and I dont see any jiras filed about that +You dont give FOP a chance to breakanywhere +I had successfully tested my patch on my computer +is now but just reverting to Struts fixes the problem +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved. +that jira should fix the Exception that we saw on our cluster AM Kannan this is the master failover for client? AM Nicolas Spiegelberg its decoupling the from the master AM Nicolas Spiegelberg is the one that has master failover problems but a client only uses it when disabling tables creating new tables etc AM Kannan i was under the impression that was the more critical one +I have not really worked with this low level protocol code much so it is great to have such a great tutorial included with the patch +SFTP support code has been added to the URI changes you made and checked in your workspace +Isnt it +Committed +Thanks in advance +Fang can you review the patch i can apply if it looks with any other fix we need everything on master +Hi no currently the build system only producings shared libraries and not static libraries +Please reopen if you can find the issue against the new closing all issues that have been in resolved state for months or more without any feedback or update +Thanks for the info Vinod +Ive attached an example patch to that uses the version of transcode that takes a locally allocated buffer +Carl can you update the patch to pull from delphi as well please +and deployed new snapshot version +Actually the series will have SCR its just not there now +we publish the cxf jar to apache maven repo mannulaly so cannt guarantee when the latest update can be publishedYou need build from the lastest trunk if you are eager to try this featureBtw can you check the timestamp you are using? you can find it in of repositoryorgapachecxf for some reason cxf soapbinding namespace is not as same as it defined in I will fix it very is supported for getting to this so quickly +Later patches have meanwhile inlined the scorer queue and removed the offending javadocs. +Can you give it a shot andsee if it all runs ok +Dynamic deployment and hotswappability Its cool +i would strongly recommend not to do this but keep the original node type definitions as they are but just deprecate thembtw i find it a bit inconvenient if the docu section is updated before the code is modified +As to the rest +What about having a single custom instead that would be used during the injection and would take precedence when using the? If the default fetchschedule is used then the custom value would be used obviously +wrt Master restarts jgray and I were discussing it will be a scheme similar to zkUNASSIGNED but in a different location +TRACE Information used for extended debugging suitable for developerscontributors DEBUG Information used for basic debugging suitable for end users +No progress DonI made a svn update but nothing changed from the last update +OK then I will use that rather better solution +I was waiting for Francis to see whether he has any more comments +no news no test caseI suppose this is an old problem already fixed in +Maybe one could be removed +Fixed invalid return code checking +Thanks Mayank +done via profiles in parent POM. +s actually a dupe of to confirmThis issue requires the option to configure the JMS backend processor to participate in a XA transaction? Should I create a seperate JMS back end processor which is configured for this purpose and have a option +Yeah I like that better too +In form +Cos again can you please use to discuss the specifics of the UI? Thanks +Submitting correct patch +So you need to reconfigure your server to ports that are available +Partially yesBut please keep it opened and think if there are more possibilities to reduce memory and CPU footprintThanks +Ill defer dealing with that mess once is out +And thanks for adding the good comments in the codeOne question though what is the actual symptom of this bug? If I am understanding it correctly the symptom is that the optimizer may pointlessly continue to investigate a possible query plan which it should already be able to reject as being too expensiveDoes that mean that the only symptom here is that the optimizer may waste some resourcesI guess the thing Im still struggling with is the bit of the change which adds reload true at about line +More details required then OS browser typeversion etc +I really appreciate that +I had trouble applying the patch as uploaded +It works for me +A single pass test of most basic RAM reader functionality Working deletes via sequence Were only copying the spine of the byte block buffers and the int pool buffers per reader +You are responsible for jsp and seamapp test applications +I visited Order Manager and click on Order Entry +Ill look into this issue +My understanding is this is resolved this the new devcloud image +Yes it maxes one CPU +Closing all issues that Ive reported and are resolved. +and +Ideally they should have two different config files one for bringing up the cluster +Thanks Mark +It only fails with +I have verified that the fix is included into the CP codebase +Dynamic partitions error logs and storage formats are linked to the Hive docs and various Hive docs are linked to each other and to the docsAny further changes can be considered improvements +Is there an example of something similar that will work with arbitrary and +If we decide to later include additional information with an staying consistent with this approach would require us to add additional methods to the callback handler which would break API compatibility +I plan to commit this tomorrow if I dont hear backKathey +So Ill be sending a pull req in a sec +Posting for reference +As previously mentioned no tests are included since doing so would require Kerberos to be enabled +Have you tried updating to +The example usage in the docs for remoteLink shows the params attribute in action +Reopening for review +Move back and assign as necessary +Found out the issue is actually happening on all database based stores but it is apparently a visible problem only in Oracle because of the slowness of that specific query issued by the JDBC driver when we try to get the structure of the SQL view result +So it should not go into GROOVYHOMElib but given as command line argument to the cp option for the groovy command +If this is not the case then if someone can give directions we can do the needful +Yes the values are screenshot for the issue. +Although replication is a viable disaster recovery solution backup and recovery can also be very useful for users who dont need a live standby cluster and for situations where programmer or operator errors occur + +Both the previously uploaded and this patch apply to both trunk and branches + +we may not need it right now but its important to be future proof someone might access fields in there eventuallyThere are also some char lines +Since this driver can work with both DBand Informix we need to check url protocol first to determinethe correct +not for M moving to above was done with JBDS H blockin M Dima you left no comment on what you actually did for this issue? could you updatefollow up based on Jirkas comments? field was mapped properlyYour case is possible I have to recheck that +Thanks Hairong +Something that isnt going to address this issue but is worth knowing is the ability to see all request mappings in the STS view +Ive just committed thisThanks a lot for the contribution Jim! +PedroIt +Reopened for some more of cleaning +Looking forward to working on it +Hi BobI have restored the +This change is part of so not submitting separate patch +The stuff was first put in place many years ago +If this coding convention is not accept by upstream I will have to change our internal coding convention as well because all pendingfuture patch has this style +Thanks alot for looking into it +Im not sure if you need a new set of poms +or when i use some alias the right value is changed but as it seems all other values are set to nullset both values by passing zipsets both values correctly then passing zipcodechanges the second value correctly and sets first value to nulland finally passing zombieBindsets both values to null +AteThanks for the work around we are currently working on a fix that will prevent people from using that work your argument on updating the examples I fully agree and will spend some time on it before the next major why would you prevent people from using referencetrue? Its perfectly valid to use it +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +If aggregation is being done over timestamps the actual timestamp shouldnt matter ever be interpreted +Please confirm they have been applied as expected +Bumping to while we gauge the level of community interest in seeing this component ported +Looking a bit deeper Im not sure how feasible it is to remove these errors itd be a bit of a hack +The trunk patch contains reordering of imports which kind of distracts from the goal for this JIRAOtherwise patch looks good +This is an invalid wsdl +Tests added and based on jsftest I call openOn action from page for jarecho its has opened page in xml editor + clean up use of multiple releases of dependencies by different modules some of the dependencies are several meg and we use several diff releases derby xerces saxon cleaning this up would reduce the size and speed up builds from an empty repo +Archival file is not there +done +Why are we removing HBASEPIDDIR +Thanks to everyone for their patience. +Which is of course beyond this issues scope +Sounds good +Can we please collect information about artifacts that are currently not in maven central in this issue +The workaround that Marek posted workSort of +VERIFIED. +Nologer relevent as we dont do the tomcat deep integration now +fixed verify +The issue is not producing PDF its producing plain text output +Shane what do you think could be related to +Needs porting to trunk +In addition I already tested their send and barrier sync featuresHowever I do not post the unit test class because it need a zookeeper server and I cannot afford to consider the unit test with the zookeeper server provided by the hudson +linking issues so we ensure they are considered in move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues there any way to advocate for this feature more +I think youre right but only with the new aggregation framework +I will apply and see by myself +It wasnt clear to me what should happen on theslave when the master is stopped +theres a comment relating to gzip support NOTE Features for gzip files of Rubyzlib are written from scratch and using undocumented feature of zlib negative wbits +LarsIm going to post the reviewboard tonight +The only caveat is that dynamic fields which are added by the transformers must be in the correct case as defined in revision +Ok I guess this needs to be cleared up by the spec guys first +The shared layout is using native permission mechanism of container not param +So group commit made a significant improvement for continuous batch writes. +t is the in that testIt could be changed so that most of setUp and tearDown code is moved directly in the test so that further tests with different could be written in the same class +This bug appears to be caused by not coping with the single point when calculating for reporting youre right +If we just rename it to an invalid name +No one knows why but everyone uses the javac fixed but documented in the on how to do the javac HACK +Patch that shows the problem using Apache which removes the need for a manually configured SSH server +Patch added +not sure if this is a sql or store issue so marking both components +The help is there in +Checking it will store a cookie in your browser allowing you to automatically log in +In the multitable scenario if we assume different tables have different schemes does that mean the application mapper implementation need to take care of input from different tables +I was thinking initially that deprecated methods would be removed on version Im not sure anymore +OK this patch works and passes the test +Tests executed with a db on did not failed so far +Bulk moved to +Fixed by r. +It probably cannot handle custom font encodings +That resulted in creating less buckets than it would have previously and the preloading up to certain size in the JDBC cache stores dont take into account the number of entries in a bucket +Description updated +However they did not change the core issue here +This wont happen again! +Good old vanilla Roo +So all issues not directly assigned to an AS release are being closed +postponed since theres no time left for this to get fixed in +Last test was +Sorry +Below that the startup sequence is divided into phases loading fsimage loading edits saving a checkpoint and safe mode +Is it worth to do it? suggestions are welcome +What does preserving full source path specifically mean +Whoa! Didnt know that the anchor tag had the same properties as Location +Agreed +I added an unit test in instead of the jpa test you provided +Can you try to set the flag as we arent in a position to release Seam that soon +Sure I can try put a little project together +Adjusted which shows the you were right +I do feel ashamed now yet very thankful for quick and helpful. +When interleaving skip data between docidfreq blocks the performance loss on queries still seems to be unavoidable +This is complete +II agree completely that this portion of the client code should be changed to retry for a large finite amount of time +I think its a bug in the dmr +revision . + +For the other examples contents under the node are not viewable in the GUI as explained in the tutorial meant that it is not updated instantaneously +In your screenshot the XML Configuration node has zero children +I totally agree with Christian +Nomen ist omen? +You refer to a JIRA issue that was marked fixed in June +another parameter replication if the user sends a value of can we reject the request at namenode rather than after the s fix this in +Thanks I applied your patch +Now the various clustered services need to use it +to the search path as well +Verified in +But there appears to be some problem that comes up intermittently when the interface throws clause does not match the implementation +Id be more inclined to try on users the approach first as it is the most visible +fixed in +however in the meantime ive left the company where this problem occurred and i dont have access to their codebase any more +Ive failed to reproduce this issue on Windows on Harmony rCould you okease verify +The attached patch is created after merging the patch of +Reviewed by Ilya +Appart for HSQLDB and non who incorrectly implement isolation levelWe should raise the higher Write cannot be set by a user in the current code +spanish Alex Mayorga Adamefrench Emmanuel to translatesrcmainresourceslocalizationsrcmainresourcesorgapachemavencontinuumaction +Fixed in +The JIRA for the compilation problem in May be it is not a Guvnor problem but a compiler problem +JesseThanks for your replyI think is a online DB +I guess it gets broken just recently and the compiler tests has not updated to the recent rev yet +Resolved +Of course the example is heavily simplifying the scenario +Jacques that email link describes what needs to be fixed NOT why it should stay that way +Yes +Would anyone want to try this +You can resolve this ticket when you are happy with the docu +It has now been fixed. +This patch does not address that issue +Need to come back around and verify that this is still an issue +At least now I believe that some PARAM are really necessary for us to know what we are doing and notRegards BillYes I think you are correct +Clients can monitor RS liveness via ZK and respond quickly via watches +Currently the only thing that can map to actual service addresses is the proxy providerExactly IMO it shouldnt be the proxy provider rather a different entity that proxy provider usesI agree these issues should be discussed in a different jira +What kinda rates are we talking here Erik +parsing and encoding +File a bug against JIRA itself too +Closing. +Hey OlivierDo you think that kind of issues is now fixed with the new filtering implementation +was added into libgen on EAP ER now correctly generates the entities +Ill revert this change until I have chance to look into the issue +Could you please deploy Seam project too +Creating a separate structure to manage hierarchical directories +Had to disable some testing because of for the useful and good patch! I did a quick review of the code and have couple of comments Is there any need to keep the am still? Since there doesnt seem to be any users of this may be this class could be dropped? What would happen if a client tries to send a SET STATEMENTTIMEOUT statement to a server? There doesnt seem to be any mechanism to prevent unpleasant events +I think its because the new closes the stream once an exception occurs +It inherits skinning from base menu components. +when you start whateverThere is a default implementation that does what was previously done a couple of simple optimizations +Move out into its own package private class +I think things might have gotten messed upActually since you are an Apache committer already and javaflow is still in sandbox you could also just get svn write accessPlease let me know +Agreed that this is ugly but if we do not publish the artifacts downstream projects cannot easily consume hbase jars build with in their profiles +it also will be problematic when we do federated +Hudson returned a for contrib tests though I cannot find any failing contrib test from the console output +We will add this as a step for rpm will not run into this issue +Thanks for clarifyUntil the bugfix is committed Ill use the inputText elementGreets Leonardothanks for committing the bugfix for this ticket +Nevertheless SLFJ reduces the impact of this dependence to levels +As of now its a push model where every change in the schedulers state results into an info string which gets pushed to all the jobs +bq +and a similar fix needs to be done for branch as well +If someone could test this fix on Windows that would be great +we need to fix the first by creating a by. +Sorry I didnt have that much time to work on this feature +Attached dump +add a test to for the new feature when there is no jms binding specified axis reports that no JMS connection factories are configured +Will be part of ER handoff +Cool +Fixed +Fixed +closing all issues that have been in resolved state for more than one month without further comments +Hopefully phabricatorarc will get stable enough soon so we can start using it to apply patches +correctly reported the number back to me +Hi guys this Jira is not duplicated with +As as workarounds keep us posted +This patch is to current UDIG trunk and has working coverage and WMS rendering +Will port it along with other patches +I ran all existing unit and tests as well as the tests provided by Arun +Can be a post install step executed by the admin +I reported it because I thought it might be related to the CSS issue in the release +Let me see what QA saysTed Stack Thanks for the reviewTed Yes the check in FFT can be simplified +offset should be the try block offset the inlined finally block offset another finally block and offset the normal finally block +Is the Git repo out of sync with the SVN repo +This test didnt work with the original version of Hibernate annotations and works with the patched be applied to v +Hello JuergenExcellent thank you for the quick reply +Yes this is needed +Reattching the patch to trigger hadoopqa +updating DT is added to such that can send separate update completion notification not to confused with storingremoving completion notifications +for the patch +on the patch +The Central plugin doesnt depend on this pluginI have committed the fix to the trunk +The valid options to me seem to be Disable these tests somehow such that theyre only run when some specific flag is passed to ant +I did and I couldnt +Does anyone actually utter the expression uncork a bottle? I cant imagine saying to my wife after leaving a bottle of Rioja standing in the kitchen to warm up Shall I uncork it yet? Sounds bookish +Fixed in cvs now it displays +Many thanks for this very good to trunk and branches +I believe this issue is fixed +If a proper response is given it assumes the server is up +bulk defer of featurestaskswishes from to +TX Resource Manager means any JDBC Driver that supports XAYou cant make assumptions about the order a Resource Manager would give back to you +This appears to be an issue for all xxx SQL States +xml in our package to be sure were using the code defaults +Remember the last values in case of refresh and reload in same session +its great to see all this progress hereI had one suggestion I felt this way about Version too +My rationale is that if a user is trying to set the throttling threshold then we should not proceed when the input value is not correct +I dont think this need to be done within itself +Im not sure that this is necessarily a problem with the migrate command although I suppose it could be a bit more tolerant +The patch is untested since I cant compile the code seems to depend on libraries not present in lib directory +Again Brian asked me to enter this as a bug +maybe just call it optimize or faster or align +I have to rewrite my code +This work has been done on the branch and is awaiting pull request processing. +The current scanner implementation expects clients to restart scans if there is a REST server failure in the midst +Thanks Antonio Perez Rupert for the support in identifying a suitable solution for the errorPlease review and apply the patch stanboltrunkentityhubindexing +Is this YOUR project or OUR projectOff course you can reply on my comment that it was a stupid place to put it and that I should add it somewhere else but thats all +and committed! +If any pop up during TCK testing we can reopen. +I think we should fix this as uwe mentioned +agree with Aaron let us please keep the discussions here focussed on one issue +Please do not checkin legal advice or policy into the website without having it reviewed on first +Commit +Wouldnt it be better to invest in a more configurable and extensible retry mechanism for maps? If a task fails we should have hooks that permit cleanup of any data before retry andor to move final results into place on success +Here is the patch that contains my suggested changes +This would be a nice one to have +I actually removed the topScannernull check from the above and the single file scanner was fasterThis optimization would benefit both and implementations right +hiverc would be of value in case of hive server +I dont think well fix this for Ill mark it as +Can you look into the case and see whether we should fix it or simply not run it on WindowsNote also the tests are from the Shards project oncvs dpserveranonymousvarcvsshards checkout rubytestIdeally we would want to submit patches back to that project and update our copy +Hi Otis yes for a serious whole web crawl I need to change this reg ex firstIt only hangs with some random urls that for example comes from link farms the crawler runs into +I just tweaked the Sonar code to use the and an updated file for Checkstyle +Doesnt work in new rootless trunk +You effectively have to do a switch statement on every element which makes for unwieldy code +Had copypasted the title but it didnt include select which may have confused people +For some reason the didnt mergerenamed in lwassemblies branchDeleting configsAdding configsAdding configsAdding configsAdding configsAdding configsAdding configsAdding configssrcCommitted revision +So you want the parent to be visible to TeamSynchronize operation etc +I have verified the build locally +It would be really great if there was support for creating arbitrary bounding boxes that do work over the th meridianShould this be changed from a bug to a feature request to that end +Are you using the same version of and? The versions must match +Can you resolve it +It should be the application itself to decide how to deal with the change in a compatible way +Fixed in revision +If you find out anything further please create a new enhancement requestJuergen +New patch is attached +I have cleared all cookies and the warning stopped occurringThe question is why the cookie was not replaced if it was invalidAnd another one why dont you provide clear message about the cause of the problem instead of that ugly exceptionKind are two good questions and ones that we could address in the form of adding them as new features to the codebaseThanks so much for verifying that the cleared cookies solved this problem that confirms my assumptions in my first comment above +There are only entries in the addon list +Please update this once resolved +What we really need is some way to preserve the page context not tweak +Fixed as part of change for as done since was resolved as part of +Broker log file +Missing path separator in classpath +See the attachment +Gives a pretty detailed demonstration of how to use the client mode repository includes a sample application which runs in this is complete enough to get merged into the main webflow codebase and to reinclude the missing client mode repository functionality +Closing as the signatures look correct in the current code. +Once again this one affects any repo not just maven repo usersSo Im hoping that artifact signature checking becomes an option in maven day +Yes I was trying to use the font support feature +dont really like the names fixed vs fetchInterval thats confusing and unclear +Patch is in trunk and branchThank you for your contribution CheolsooJarcec +Observe the screenshot +A patch for the test cases would be appreciated +Id like to volunteer for this issue after my patch for which involves the same part of the codebase +It works because the malformed part is after theimportant part and we have no need of the malformed part. +Ive added a also depends upon the intermediate being before the new deployer based on sort to use transitive additional name compare to uncomparable one is a cycle check at the end of it all +Thank you very much! +Ah right I should have guessed there was another issue based on the fact that weve discussed it in the past +This fixed the issue with the imports. +Committed revision on branch +After having restored the code which handles this case it seems to work wellStill having some issues in other parts of the code so I cant commit atm +Hi Is this issue already resolved +do you have a time estimate when the will be released +Trying against hudsonqa +dialog will appear +Do we need another state? It seems like the problem is we default to hibernate which as a dead state wont show +jstack of one of the other nodes we found in this this log appears to be a capture from stdout and not the file since there are no class names or line numbers which would make things slightly easierThe one thing this log indicates is gossip is not the problem + +Hi RobbieIve addressed your review comments and the patchcheers applied. +Aswath your reported errors amongst others have been committed in rev and thanks for reportingRegards note that communication is now a general coomunicationEvent application +The false setting is more performant as it avoids calls into the +Patch that fixes the problem and adds test case for the issue +It triggers the same exception than solving works like a charm +Erik I believe that there are many solutions to version compatible problem +Hey LeonidYes a patch would be greatThanks motivation for this fix is the following +Good work Saliya with a one minor comment do not do additional formatting in the javadoc to make it look aligned in the code which will result in messy java docs +can you create a review board for this one +The meta data field is still called Is Disabled +Do you have any scripts that can reproduce +Thanks +For example shouldnt have a static open method attached to it +Ill look a bit more into this and decide how to proceed +Linking in a new feature request for a real not matches operator +And that thread is written entirely in user code +Site is the difinitive source for info on struts and I feel that it is critical for new adopters that example links not be broken +does indeed use JDK thats why Spring is still compatible with it! All of the Java support there is optional Spring will automatically adapt to JDK when running on +I should supply a sample file only +WolfgangNo worries Im glad you were able to find the issue +I suspect other users will encounter the same problemThe argument above is that the current implementation has been deployed some where so the current implementation is what we have to commitCan we figure out something to bring along these legacy folks Francis? They will have to restart anyways going to the new stuff? Can we treat this problem apart from how ns goes into trunkbq +Thanks for reviewing the changes +Builds local +Done at +Thanks for your analysis and I agree with your conclusionBased on advice in asfinfra I added caching for pages and for images that seems to have activated ATS who in fact did not cache beforefyi I have asked to have my accesssudo right to this vm removed for other reasonsRgdsjan I. +committed can you follow that example now of just loading a constant filename from the classapth +Resuming addition revision +Verified in JBT +Thanks Trevor! +This is an implementing patchThe patch removes allocation hythread t to the VM sideBesides it fixes cleans up and enables for LinuxIPF cunit test suiteThe patch passed build test on platforms +netstat ano shown NO any connection to the client +rev used in geronimo +Im really concerned about including the in hadoop +This is not a problem for the implementation I guess but it does make the new interfaces effectively useles until implementations for all the dependant parts are created +Harsh could you please try thisThanksYang +Committers can you please review +they handle it at the application layer by having s reconnecting to T in rA and establishing a new subscription in rB +Here is the diff file showing differences to the file and the Broker and Queue classes files +In stead I will have two abstract methods readDataread in Checker and two abstract writeDatawrite in Summer +Projected to Metamer test suite. +Roland did you ever get any more information about this +screenshot is correct focus is in the right panel and there is selection in right panel btw +Not going to support testng in V. +We dont use the regionserverhlog sequenceid +Anything needs my to do please let me does something have to change in our documentation to explain how to make this work or does it just need to be release noted? Please explain in detail +Both and are failing +Yup sorry thought its the hung runner issue +I get the errors that youll see in the next few screenshots +I still havent been able to reproduce this bugDavids patch is probably just fine as a fix if we want to get this into +So then is there a chance of this happening? Or are we just going to have to expose the current one used by all projects outside the firewall +If you are still having problems let me know +This report looks like it should be reported to jQuery project insteadCan you attach a sample response that takes that long time? Id like to reproduce this because it looks like a problem in your installation of IE +It is shame that project like Hibernate have so much annoying and not resolved bugs +This patch also fixes arguments array memory allocation for method with arguments +While were at it we should update all the c docs to provide such detail not just init +Rein yup applyed r +Flippo Looks like I will get to this tomorrow +Awesome +Thanks this was a good bug to find! +SJC now builds on Spring Build. +So theres still some internal friction to be removedJames lets run this through the profiler tomorrow to see whether we can pinpoint that friction +Is there anything else to consider +The jndi nameis the binding of the home interface that the client islooking up +There are a handful of patch series that developed under the ant build and have been reviewed in various states +Commited a pull request that may help with this +Patch applied. +I will create another jira to track it separately +Attached patch adds block to JDWP agent threads execution +This has been fixed in SVN but the site need to be deployed +So all issues not directly assigned to an AS release are being closed +Added some links to related issues +Would you mind rebasing your patch against current trunk and reuploading to review board? Jarcec this was probably one of my first patches in old days +If a test fails normally by exception then things work correctly +It gets assigned to nobody +copy the two jars under the project of srctestresources +It does not have to be called it could also be called or whatever +Verified. +Move it back in if you think differently +I dont think this really needs to be fixed +Waiting for a to add this and to clarify if it needs to happen in a Beta respin or can wait for BetaCR +The addendum will do for now +See also +At this point the JT refuses incoming connections +Could you regenerate it please +Could someone crisply state the compatibility implications of what is currently checked in? I think that describes potential issues +Thanks. +This whole problem doesnt make a lot of sense to me +If there are any improvements to the java code in the second patch attach the diff after taking a svn updateAnyway great piece of work +Applied in r +Closing issue +Thanks for the review ted I was typing too fast +This looks very similar to the problems which Dag has encountered working with authentication and shutdownencryptionupgrade on +There is a unique parsing code for both cases this code considers the time information as optional +Small bug in using default vs config +Jacques Please find the attached ScreenshotIn this screenshot you can see the Association not found Product Id Product Id To Association Type Id From Date message under Edit Product Association screenlet +When Ive needed this functionality it was always the code requesting the value that knew how to create the value if it is absent +Is there a reason to include it also in this release notes? has never affected any release +I think we need to do the same +Thanks a lot for the patch +Please give a more concrete example of what fails +Change the visibility of and back to private +If we can figure that +This is now the only HDFS doc to in XDOC formatShould we file a separate JIRA to convert it? Or just revert this one and redo it as APT +Sorry about the GIT slip up +how long till a lucene user complains about bom markers in their stoplistsWe can still use ant to keep the files in sync automatically from the lucene copies +But need to confirm thatDevaraj I dont follow +If these were all commited then you would see inqueue outqueue which is correct +It is good to confirm that though +mrepositoryjavaxenterprise +Pavol how many of those attributes not working are regressions? Only oncomplete +All should be good now + +I agree with you about the initial configuration for Zookeeper +Fix will be included in you for reporting and providing such a good test case. +Thanks for confirming. +Thanks Quinn +The later case is often called entity disambiguation or entity resolution +This works for cobertura includes the oro dependency that the original was need to put the asm ncss and cobertura jars in your repo for it to work +After all we did test against Essentials Hibernate and +This is very important because many developers use Spring or EJBS with a couple of methods implicitly using transaction around which calls ksessionjBPM API in some moment so keeping it thread safety for these scenarios even with performance implication is better than let the whole context in in a non reliable state +OK Ive got this test up and running theres value in it and Im just going to it based on the that I see in the original test file from the patch +Added Export support +Setting the destination and properties is perfectly acceptable and is normal for +you want us setting the context class loader in Ant in general or for a certain task? Because if it is for Ant in general wouldnt this be an issue for Ant then +So the issue is that it contains duplicates files under subdir directory +When the block was created the initial pipeline has datanodes and +Please reopen if you are still seeing problems. +in my view optional is the best approach. +has been resolved since december so I think it is ok to close +patch for review only +The unix stat uses b for blocks and s for size in bytes +I still keep getting the same error +stephanCould you explain a little bit more the context of what you are doingHowever just FYI if you dont put an estimate on a task entering time WILL change the current estimate to match +An attempt at updating the BOSH pluginRemoved usage of Jetty APIs and replace them with Servlet APIs instead +It seems so AndreaD +Almost there but two questions Is bounds the desired word? We also use bbox in other places the draft spect doesnt cover After a quick discussion with crschmidt and mloskot on geojson I was convinced that it would be best to use a polygon wih points to represent the bounding box +The custom reject message would point them to a wiki page where we would say something likeIf you were looking for the legacy marketing mailing list please send to If you were looking for the legacy security mailing list please send toEtc +Would be nice to get this one in for +Hi DIGYthanks for the tip +After furter testing I can confirm that the problem was linked with a specific version of Firebug installed as a global extension on Firefox +As long as RM is restarted and ensured that it has the same state as it is before restart +Great +So I suggest we go ahead with this and open up another ticket which will address the problem of all null arguments for udf +The usual recommendation for source and javadoc artifacts is to use the same name with a different type +as blocker due to described issues in fetcher +I agree that the behavior should be to throw a deserialization error +The sure way of ensuring this can also be to mount hadoop on ramfs +It was setting generate to true in It should be set to false in order to run the test case correctly +We are getting rid of zkclient in next release +If you look at the servicedef for ups it implements calchttpslocalhostwebtoolscontrolavailableServicesselservicenamecalcIn this service the required fields are correctly identifiedThis isnt just a display bug if you try using the service ups and leave out the required fields such as shipping there will be bugs and data errors +which doesnt seem to be ideal +Maven and WTP and PDE took last patch from Scott refactored it to be more integrated with builder and committed changes to are some issues with rebuilding dependent projects on pom changes or when project that contain dependency is closed or deleted +A few minor nits Typo contets in comments of +If I dont get to the type validation Ill create a new JIRA for it +defer all issues to +Committed +Committed to HA branch thanks for reviews. +bah wish had more time +If you look at the unit tests it expects the above case +It fixes it by implementing a which delegates to several underlying +I cant reproduce it either today +xml actually putting the files into that packagedirectory as well +passes tests locally +In current code once the action is triggered an event is broadcast and relevant listener would update the Composer resolved now +Just to be clear the trunk as well as our branch builds WORK perfectly when we run them locally on our boxes and the build problems described above happen only on the +I just committed this +xargs ls r amandalay amandalay +If youre packaging youre gonna configure it properly anyway +Info about command line arguments is added to the guide +This might give the false impression that the node is stalled +A cluster of Solr servers behind a loadbalancer with caching functionality +So this option is in fact documented in the lein plugin README which is pointed to in the main documentation +There is no indication that anyone is actively working on this issue so removing from the fixVersion assigning to James to assess the existing patch for reconsideration +Bug fixed according to recommendation +I saw that too pretty easy to recreate +Whats happening here is that the Inject does not get called again when the cache is base configuration population should really be happening in an Start element +Fix merged from committed revision . +Basically the tests are modified to check for elements without all looks good to me +This issue is now closed +Can you show what SQL is being executed and why you think Derby is not behaving correctly +maven users wont use the zip and can use the original maven reference +Still working on getting the tests running for this +Open jBPM context before parsing in both deploy task and upload servlet. +security and PAM always show up with the correct servlet and servlet mappings +Thank you for the patch OleI had two unrelated failures while running these testsSending javatestingorgapachederbyTestingfunctionTeststestsreplicationTestsTransmitting file dataCommitted revision Will merge it to after running with +Attaching a slightly different approach with cleaner logic +Adrian thanks for jumping in +Maybe its a configuration issue but I dont have anything special in my or +The shiftoperation was not correctly incrementing decrementing array indexes +I have reviewed the path I am basically OK with itThe only thing that bothers me is the design decision to use a static inner class for +committed with revision +And some of the other testing and build scripts call out and reference specific versions of Derby +Tomaz uploaded a new patch for the release can you review it +The second problem can be easily fixed +Patch to add Dimension data to the request sent to UPS for shipping Estimate +Reopen to mark affects version for possible backport +I decide to ask data nodes to register after a name node restarts +Great catch Todd and thanks for providing the test case +Running tests +With a direct constructor there would be the problem of missing data type informationSorry what does that mean +So you could do something like Just an idea but is there some way to extend or to provide these kinds of overrides +patch removing duplicated logic and test looks OK. +Linking to which represents the quickstarts which demonstrate the capabilities outlined as P in the PRD webserviceproxybasic webserviceproxyrouted webserviceproxyversioning +So a batch scanner thread would fetch a batch and throw the entire batch on the queue +Applied and fixed thanks +Screenshot attached +Perhaps the another compilers may has the similar feature +sigh more incompatible changes +Fix should be in tonight +Checkstyle requires Java so we would need a retrotranslated version that works with Java first +Setting up layers that automatically reproject by default is one idea +Dismissed as this was not a bug +Committed revision +Its way too involved to go into the current release +Just wrapping it in a small genserver ought to be enough because its not important if config changes are serialized I think +I apologize but this may take a while as this appears to be something targeted at Eclipse +Deadlocks in +This is already covered by and i would rather not mix this into the test case which states in the that the last segment of the path and the name are being compared +Added a new to maintenaince +Do you want to contribute a patch for this +this was actually fixed in +Use injected Gauge to override getValue with buffered message size +For some sample classes that do not seem to manage crs information at all look at in axis order issues to thats it just tried the test passes now +I checked the DB part closing all issues that have been in resolved state for months or more without any feedback or update +I think I am the last person touched this code +In the use case of we want to configure with different values of bytes per checksum. +The compound properties are shifted so that both base and compound can be stored into a single int +however isnt there a risk to have weird where some fonts are scaled and others dontWould you be able to attach some screenshots of installers to demonstrate that +Logged In YES useridThis still happens as of version +Failover seems very solid so far when exception is not allowed out +Hold tight or you can also patch and build source on your own +Attachment implements a unit test that shows the bug as well as the fix in should be considered for +Whats the possibility to get this into Sling ? I have run into the need for this several times on a current project +Maybe what I should do is make the authenticated url stuff only be used by token operations +somehow I would like to fix this laterbut dont let me stand in the way here +Set correct fix version +Then when I quickly moved it all the way left the leftmost portion of this screenshot is much less garbled +I plan on working first and I will update if this task starts to slip +Please check my patch +Did anyone had a chance to review the updated patch? Thanks +The Tabs must be there in the original file as well as my eclipse project wide settings are set to use spaces +This definitely seems like theres a missing maybe? Could there be a problem with the feature +Any comment please let me know + +Sounds like something that would be a benefit +this is by design +Committed to trunk and. +I have created a SFW ticket to see if we can get this done in the framework rather than using a workaround in Roo +Ill turn it into an enhancement able to change issue type so Ill have to start again +As is it is intimidating +Can u please upload your sample service as wellthanksdims +IMHO we should not only think about today but also tomorrow +We have long term problems with Permgen space if we are running our test suite that repeatedly deploysundeploys ESB applications +Should not be a big deal and is in linewith existing design +Patches replaced by rebased versions +bulk defer of unresolved bugs to +Attributes policeman sees no problems in the attributes +But came to know from RUBYDESCRIPTION it is upto. +Ive checked the code and a variable init miss at the beginning of the function send in ezcThe expectedLength must be set to false +However Ben has been doing alot of work recently and it feels like we are getting close to being ready for Validator +Can we get this retested against the CP branch +I tested it with an artificial patch that had changes in various modules and verified that it ran the correct tests +Thanks for the feedback just two more questions Is it fix or workaround? Is it expected by the nature order of things that there will be deadlocks on the database level? Should be there deadlocks? connection closed probably a remoting issue could you elaborate more on this? You mean Remoting or what kind of remoting? This was never a problem with remote JNDI resolve the issue so I can close it tests are no longer failing +In Id use the package space for a similar purpose +Sync patch with recently trunk branch. +maybe +need to tweak comment +Is there anything else to it +The previous rev of the patch raised errors in the regression tests because some canonized message text has changedThe revised patch is identical to the previous rev except for the addition of a couple more corrected canonsM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterCommitted at subversion revision +Hi AndrzejI wrote this code and got it working some time ago but then had to abandonit for a more important bug +Attached patch with fix +groupId is now +That is was I was looking for +Thats an accident + +Or can it? Over unique ids and then sort on id would be broke right? Okay it would be kind of nuts to try and sort on that many unique terms but in the futureEDITActually one seg would need but you know what i meanEDITOkay I guess my argument with JIRA cleared up youd have to have the second segment or later with over terms +Patch to use a so the Iterator can be counted on to iterate over the contents in a specific order +Step is to get an actual definition in the javadoc for what it is trying to compute +Cant we determine the size from the collection +So it will be in It will remain removed in the and +lets see if it is any better under the new test harness +The Seam phase listener shouldnt be adding messages directly +Likewise Im not aware of a concept of a realm +Im pretty comfortable switching the representation for lets make a separate ticket to move off heap for +In that bug I mention an additional test case that I created for that tests the interaction of and the static property would like you to look at it since I proposed a possible change to to resolve that issue. +I think this is ready to go in +Excuse me but its not enough information to resolve the bug without yourassistance + error log mode you are using? full ejb or directI cannot reproduce your problem but I encounter another problem in full ejb modeIf possible update your code with the latest branch +Estimated completion date is build has been released to all modules under head. +sbt clean package +Ill wait half an hour or so and retrigger this +Thanks Wendy! +There is one more problem on permission Suppose the owners of file and file are foo and bar respectively +Im tabling this issue until we have a little more time to look at the impact on the descriptor and how the installers and archivers really line up beyond being superficially dont want this to hold up the release of other important fixesfeatures for now +Sorry I cannot find a proper link to download the Ive attached the patch +I will let Amos answer conclusively but I think it is going to be a bakeoff between JNI and socket so it should only be of short lived value during this dev cycle +Indeed we return unsigned byte values from String and IO methods that return bytes so not being able to stuff a signed Fixnum in the range +If you trace through the load method it catches Throwable from the and returns null +I think this is more of a feature request and not a bug caused by not using a thread pool +Ive built several times too on FC +Its been updated and the resulting in the are consistent with the jacc spec +Basically no volunteer to test itHowever it has been fixed in too if you care to upgrade +With a title of Limitations of revoke privileges on tables and columnsI need your help in determining where to put this new topic +Scott there is something wrong with the patch +Thanks for the patch but we try not to rely on system properties like is a component registered with the runtime that implements the interface which is our preferred way of retrieving a Stax parser +A file with whole stacktrace as as cleanup +contains new files to be added +If your data is largely constant and this creates a problem in your application as a workaround you can provide an alternative Percentile implementation to using set +Can you block the effective pom with a message saying that its updating? +binmahout kmeans i +Just one minor comment +This is not intentional and is automatically created by the buildBut these jars should not be inside to begin with and should be specified in the pom +In your case I think the behaved correctly if job and job were failed without being submitted due to one or more depending job fail +Verified in ER +With the text of the token being the default but other unchecked options for raw bytes positions offsets etc +I brought up this example a while ago +This should now be fixed +Thanks Brandon! +Cos I just have tried +It has been updated quite a bit since the last time I had the problem +Make log and high in replica atomic long +switched all connection related actions to the command framework +Makes sense +Have you an idea +Spring is not released yet +Obliterate them +Jrgen we have talked about this a few days ago at in Copenhagen +The code comments also indicate why the change was necessary which is greatIts a pretty small change it makes sense to me and derbayll passed +I believe this was resolved so closing it. +Can you try this with the latest plugin snapshot? I think I fixed it while working on some other stuff +revert some changes +I am not sure what is being suggested here +revert the change +The check exception now display though it is limited in its use +Remove legacy site document. +ASF Security dont evaluate vulnerabilities they just act as a central point of contact passing on to the appropriate PMC to evaluate +But datanodes network locations only when script based mapping is used +I have fixed the NPE upstream can you try another run with Jaikirans changes +A point of concern is that I am now using the osgi icon and I am unsure of the legal issues regarding it +Ok I will talk to Stuart as it will be me that will probably be fixing undertow subsystem. +And the garbage collection background process results in the stores holding onto no longer used BINARY values a bit longer than they might normally do that +This screen was fixed in rev +Marking as dupe of for now please reopen if still an issue +Keeping platform neutrality would be beneficial so that the usecases continue to work for non Java AMtasks or on Windows + Please remove the invalid dependecy in following DD +Seems to be already fixed +gem packaging support into optional plugins +we have to check what the difference is between your case and the railo admin the railo admin used cfadmin as well for this task +Test case added in ConstructsPoint in you probably want to just grab and apply locally + +I just committed this +dev seems to have fired up again but it looks like theres a number of people that have been running a release since the rel is years old and looks like it might have a few issues +Thanks! That seems to have done the trick +Less parallelization per RS +HiIf you switched to mysql then you should configure Sonar inside Hudson and specify info about MysqlOr I cant understand your issue +Using Thrift library. +I added it just now +This allows for leading wildcards in the if set is set to true +I put the same change to HDFS and had a unit test thereFor both Kannan and I tested manuallyBTW why doesnt simply use Hadoop RPC +once this is addressed +I have a few questionsSection what is a Source Window SpecSection I cant see any way in an MR framework to guarantee order stability without adding an order clause +Attached patch from Review +Sorry I should have been more explicit +zlib for C +Can anyone suggest a solution so that I can get around this temporarily without building activemq? running in Karaf +Between this and we regressed on Kerberos it takes an extra round trip +Its not clear from the release notes if message boundary events are supported or not +added log messages will do cleanup in a separate jira the patch failed the unit tests +Test passed at least once but may be unstable when parameters were more aggressive Ive seen some put failures with no useful logging to explain them +I will also open another JIRA issue to clean up duplicate entries for +I think this is fixed with the changes for +Looks goodThanks good to me too +Linking the patch that added the problematic code +Most if not all JSF core annotations now can cope with annotation changes +Thanks Michal for the issue and the patchI will correct this problem +Included is patch against SVN trunk revision It exports suggested attributes +applied thanks edwin +Marking as a duplicate +I think it is important we are consistent on this it would be quite confusing we have different rules for the different API sets +Please confirm the current status of this issue and if its incorrect change it as required +Thanks Stack +This has been fixed in in. +I plan this for +And eclipse says generated wsdl as an invalid wsdl so what do you guys think about this +Sorry for any inconvenience this might have caused +bq +Can you explain what you want to achieve with your component +Patch for patch seems to be applicable both for trunk and for branch +Im still seeing the issue on beta +would be better than Id say +A patch to with the test case and drl exhibiting the problem +This isnt even a dm Server issue its fundamental +Stop trying to fix it +One of the big upfront filtering jobs seems to currently the most for triggering a No space left errorAs noted above we have about TB free +TCK tests TBD +Unassigning issues on which I wont work in the near term. +Should we not add them? Should we add an option to the wizard to add them? What are you looking for here? Whats the easiest way to kick off a build via code when the wizard finishes? Can definitely whack something in there +I noticed that has the fix version set to when in fact this fix was placed in the branch and thus should be present in the release of embjopr +The patch for these changes is ready to go +Thanks for working on this Vinay +So looks like I was hit again by which is not fixed. +just committed this and ran its unit test passed! +Lukasz how do you beautify the code part +Thanks +Timo when I originally concieved of the enhancement I was working onthe code base +Id rather recommend implementing a custom then since the impl that I sketched wouldnt really be complete without a filterinterceptor that sets the attribute +Need to update Credits with libraries and their license +On a side note the change in the class name cited in the previous note is here to stay. +Double check you arent trying to return a null object from the producer method +The problem is gone because slashdot turns back rss +It should clearly show that was not set +well that just means that there is no appropriate test You mean that the tests just happen to work with new and old format by coincidence? I just really am in the assumption that a query is done on one index at the time and wether this index is in the old or new format does not matter +Why was this resolved? The linked pull request has not been merged. +Thanks Junping! +Followed steps mentioned in the bugCant reproduce the issue in latest master +IT added +Some version relatedtests now pass but a lot are still failingWhat works now is the creation of an initial version history +Ant tests passed on my box +If that is the case it is probably fixed on master +Basically the approach is to avoid throwing and swallowing in +Building the rpm works using this patch for building a SNAPSHOT rpmbq +Patch is not workingpatching file coresrcmainjavaorgapacheshirorealmldapHunk FAILED at +negatda +The comment on this revision states Fixed bug when using extends of a class from an included mappingAbove is the attached diff for the fix +What is the current status of this issue? Is this feature still needed +Change context root as described above +It failsPlease look at these failures +Grouped listing based on deployment type as it is in admin console would be avesome +Thanks Jacob. +Fixed by the patch in fixed by the patch for +Could you pls offer which link you used to download? +Trying hudson. +I was working on and when I patched that back to I just brought miniacc along for the ride. +I ran the two tests locally with patch and they passed mt Dtest mt Dtest v addresses Anoops and tests passed for patch +Works ok for Tomcat Looks like Jetty has issues with that +I have modified Sanselan to read compression type images and am willing to contribute these changes +Now I can see the problem + +That is each web app will have its own session ids associated with it internally inside tree cache +updated version of simple benchmark program that uses complex input data rather than a real input +Untar the file and run warble to create the war +The difference between the SSD and spinning media here is noteworthySSDk more writes but only k fewer reads totalSpinningk more writes but k fewer readsThe best way I can think to explain this is that possibly the SSD test was CPU bound while the spinning media test was IO bound +Layered pdf has well layers as we have learned to see in various GIS programs +Forgot +Although this issue can be fixedcorrect on its own it really needs JTAJTS and XA Resource support in the connectors to get the transactions to exhibit transactional behavior in most environments +Updated patch +Its ok nowThanks a lot! +bq +mozconfig file which i have used fot trying to build firefox +address Ashutoshs to trigger the precommit test +Had to apply manually because I did not had the previous patch applied already on +Committed +In those cases setting a null value in Configuration will result in NPE +If you run through a real stylesheet Xalan will normalize namespace useIf youre using the default identity transform Xalan bypasses much of its normal processing +has the partition state information like leader isr replicas for a list of partitions +Fantastic work guys! This is why is amazing +Crunch and others that use Avro internally have been very very happy with it +I prefer to place the jar even if it is a dev jar in an official apache repositoryI tried the above link and it was ok +Thought you wouldwant to messed up the last message +Created a simple PR for for master is on Jenkins +Again this is a moot point since the Cache requirements dont seem to need singleton status +That makes total sense + +Hence removed action from and moved them to and +Thats probably why I only found instances of the issue +I had made the modification and posted the changes also +It seems better to track conversion to java in sep +Mostly it is changing from to +Set select for the tried that solution already but the issue still appears. +JazarineOn a quick look at your patch I see a diff for nbproject which is not part of derby codeSo make sure you submit patch that only contains changes to files which are part of derby codebase +I sorry for that +but I agree that this should be for +Committed +It seems that this bug is so Ill try to attach an FO that reproduces the issue on all platforms +Im looking at using Spring Data in a new project with the latest Mongo and its drivers +In my experience few services are written thoughtfully enough to contain minOccurs or maxOccurs +Then if anyone accesses those portions of the object the wrapper can reload the +is more needed than just having two open editors one with the xml template the the other with VPE +Unassigning so a committer can grab it +The customer also saw this failure error which seems to be in a different place +I just committed this +patch looks good +Ive found some code to replace the dependency so that this wont be a problem anymore +We can relax this restriction later if we come to understand that it really is harmless +trunk updated wiki as well +Can you try uploading it again +Closing since we now have verification that it fixes the problem +This issue has been fixed and released as part of release +Right that is correct +That is a cool tradeoff to be able to make +see previous comment +These pages look good to me +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +pushed to master and marked as +By default this value is set to false and the tool will ignore errors while processing resources and continue without failing +Snapshot deployed +And therefor the select method will elect it for compactionThis patch is to validate that if there is only a single empty file we dont elect it for compactionIf there is more than one file even if they are all empty they still can be selected since its better to have just a single one at the end instead of multiple empty filesPatch for trunk coming in a second +Commit ffcde +pushing off remaining issues in order to release version in JIRA and create release notes +Clean all projects build run works class is found and classpath is ok +Though please search the existing bugs for an existing bugreport +bulk close of all resolved issues. +Will remember next time. +applied patch +Ive tried both and my custom domain and I get the error in both cases +That way I need only look at ONE thing when hacking up xml files the xml files +Ok yes this actually makes sense +Example you! +To make things more interesting theres also no heartbeat mechanism to test if a connection is alive or not so you are left with using some timeout to determine if the connection is dead which can lead to needless failover for clients that are simply inactive +The ec installer is also affected +Manually tested! +Please +encountering errors attempting to apply this patch will try to piece it together manually this patch created from modified code of the current site trunk? Im observing a strange mix of correctly identified changed lines in the diff file and other lines we are different but not marked for changes +Either way let this issue know once something is there so I can continue with the setup +Fixing Affects and Fix in Version +I have not got a version handy to test with. +So if random bytes fill the remainder of the file we degenerate to what we do today and its essentially random which one gets used for recovery +Not sure this is related to ORM as I have just installed a new server with Railo with virtual hosts and got it straight away after I had set up the cache on each and then restarted Railo +I should not upload them +Thanks Paulex It looks fine to meBest regardsAndrew +I have use of an alternate bootstrap partition jar working in the sandbox branchI still dont know how to test if a normal apacheds server will start with these changes or if a classpath adjustment is needed +This has already been fixed +Yes Ive this configuration and the dll shpws up in my compilation target directoryI activate the MDA and the stack trace is equalI try a long shot add this line before zip null and all gows rightIts very strangeThanks + +Please check it once again when you find time +Thanks for review Luke +it seems come from the jms flowwith just seda one i dont have any pb +Thanks for the patch Lars Francke. +Actually I believe jdeolive fixed this one the day before he left for vacation +Ben + +Updated patch with changes suggested on reviewboard +I couldnt get it compile because it is complaining that the namespaces like Newtonsoft and lognet are not found +Jason I have a subclass of that overrides the getFile method as described but Im not sure how to register it to be used +Attaching the patch +Ivan we use the parent directory of the symlink since the link needs to be resolvable at creation +The bug is not reproducible with attached page without additional styles. +I think it will decrease the performance for write +The server on receiving the acknowledgement will then remove the message from the DB before returning to the client +I just tested and succeeded in editing the actual ticket + on this +Committed revision in trunk and sitePlease provide feedback on where we still need to improve here +Please consider the following a patch for +If the client might be using an alternate Avro schema they can specify it using the OUTPUTSCHEMAKEYIs this likely to come up in practice or can we get rid of it +I think its just a problem of wrong messageTo prove it Ive added a couple of unit tests showing what is happeningIve also moved the message of domain correctly set into the setDomain method so that it is displayed only when the operation we should add a warning message when setting an incorrect domain for more clarity +Ill commit this +Assigning to myself +Done +This patch contains a diff that includes ONLY the single change necessary to fix this bug +I will check the right way to add the gbean +This is still valid for Confluence docs +like I said that was intentional and I still dont understand what does properties are not available mean +In addition it would be great if we can find a good usecase for this feature otherwise we do not need to fix this for release +Should have been +Sure +Just one comment the roles and responsibilities of the mentors are duplicated in the Roles and Responsibilities document and the Incubation Policy +I was wrong in my previous comment the test issue is still present +VDB driver and data source character string that is returned with a binary collation will be returned as bytes from getObject +Scratch that +Reopened to set the Fix Versions field to JDO beta. +Please give the exact version of DB and of the driver DB Driver IBM Data Server Driver for JDBC and SQLJ enable the debug log level and provide the resulting log file +Patch integrated with few enhancement to prevent socket from closing on response +I made the topN decision just like +Added quickstart +Also if it can be replaced with an off the shelf library we can do that later without mucking anyone upThanks for your work and patience with this +BXML is primarily a shortcut for creating Java object hierarchies +I am currently working on a project that is using Informix and in urgent need of this issue to be resolvedSergio Moredata an IIUG director a major contributor to the Informix DB will gladly supportassist on any issue regarding Informixplease contact sergio for assistance or any other issues questionsthere is an open bug report for exactly the same issue TX Alex +Bulk close resolved issues +I then hit F in my browser just to refresh it and after I logged back in I get back to the Control Tab screen and now the errors no longer show up +Fix applied was able to prevent the abstract base class from implementing this allows to act correctly with instanceof checks +It would be nice if somebody can apply it +Good point Christophe +After updating through svn rev I have verified that the fix is applied and the behavior is correct. +to trunk +The user reporting this was unable to reopen this issue so Im doing so on his behalf +We currently work hard to move Apache Logging into a better light these days and of course a certificate would help here + since I need to test that the artifact conforms to maven central rules +we need check latency when updateLatency +I just committed this. +When I run ant test in contribzebra I get failures + +Since its not changing any other state information besides adding statements its perfectly safe for your singleton +TestedFull unit tests during compilation +I also had to add explicit dependencies to the core API in various modules +OK great! Ill have an updated patch shortly +Is the union above a union of string and something other than null +The comment could use an update then +Ill take a look +apparently we have a lot issues with VPE in x in Tests +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +See. +I have the same problem +semantically its a creation not an edit so I think we are good +Sorry for the but all the patching reverting and diffing is not yet so smooth in my workflow +Hey Mark not yet appreciate the understanding and we will get to it shortly +Besides its a class library issue not a frontend one +For example we could say the attribute must be set to a session which has certain access rights +Allan were you planning to continue working on this +Shouldnt the scaconfidentiality on the service be ignored since it is exclusive with an intent that appears lower in the structural hierarchy? This is how Tuscany is handling this today and why you get an error reported when is excluded instead of confidentiality + + +Thanks +Did you actually try the glassfishx container with Glassfish ? Does it work +Thats from. +This patch against current master fixes the heartbeat Ive seen these guidelines +This feature has been moved to the version by agreement of the development team +Release note cleared +I closed this one one year ago because according to me if the sort order of a filter really matters this filter should be updated to make this sort the default one +doesnt have the same issue since RScloseRegion already throws in this case. +The user showing it had used mvn build first and saw his target in the UI right clicked the and saw Mark as DeployableHe had no idea that he could actually drag the project to the server instead and get incremental publish etcThis suggestion is to provide some feedback to the user that he is actually marking something as deployable in a project that already in itself is deployable and if he is sure that is what we want to do +Merged to doc branch at revision +There could even be a small tprototype library that does this contributing for you so all you have to do as a user to maintain the previous behavior is include the extra dependency +Duplicate of +Anyway its nice to see such good implementations and to see this finally implemented +As we know that all the methods in the are synchronized and in Geronimo we are trying to map the configuration lifecycle and bundlewabeba lifecycle by bundle listener or something else +Add current rev of doc to jira +It just seems easier to hide the platform differences in the options rather than adding new apisAlthough Im not sure how developers will feel at having to provide all the necessary orientation popoverOptions at the start of the call +Ive added the suggested method to and it is now called from instead of the version taking the method name +We can commit this +It is and not +See is itself the hub of a whole bunch of other issues and Id hate to have to go fish for the issues depending on it and tying those to this one instead +Any discussion would be around how best to resolve it and frankly I dont care where that discussion happens +Otherwise separate jobs may mingle their logical plan maps +Committed to branch +The current test is not verifying the obviously failing case yet +See +I tested this manually on a small cluster and verified that works +I dont really care about specifying locale in constructor +Please verify in ER are still some issues in the Getting Started Guide Chapter should be +If icus people solve that it is they issue I create corresponding bug +I assume any role with insufficient privs +I think so but Jan seems most familiar with the issue +Thanks +Generally when you talk about inheritance there are a couple techniques lookup or copy down +The condition to print the message is now times the sleep period +BTW The properties generation configure is parent antrun plugin section +If the bug is critical for you please try to fix it yourself and provide a patchOtherwise please attach a sample of the data a small sample enough to reproduce the issue and be patientOr you can look into commercial support options if you want a fix fast and funding the work is not a problem for you +Patch from Ramkrishna +Tim why is this a Major severity + actually only deals with supporting key based authentication but not does provide ssh agent support +Carl Thanks for the reviewI will go through your comments on but I dont think I will be able to get to it today +I wasnt able to generate a patch using svn diff after rming the file + +This behavior should be logically impossible by I am seeing cases where different session bean instances are being injected into prototype beans where the session beans should be the same instance +I wont be able to work on get working on this soon so making this unassigned so that this can get picked up by someone in meantime +lets keep it safe as it is for now +The broker will now allow up to receipt responses to queue up before it stops processing addition frames from the connection +For now I am moving this out to for getting it done right +Dont guess clock type from thrift input read from +The States interface inherits implements in several classesFor we dont duplicate code when just merging and whereas this would be the case for States +A duplicate of. +Lets design how to modularize such omnipotent bot and forge the interfaces +Issue is the description on how to verify +What are you movingwhat plugins does not work +Fixed in SIXPT branch +Its anyway not on critical path usage of Grab in scripts loaded via GSE +I had to apply it by hand so there may be some minor formatting changes buteverything worksEric Pugh +If the Java classes are not available we will try to fallback to the Booster Pack stuff. +bulk fixing the version info for and all affected issues have in s go ahead and make this a blocker +We should complete the documentation review before releasing +groupId cant be that should be am the admin of the project + +What I was saying was that it worked on CVS HEAD before I changed anything +Any concern in changing the existing latency numbers with the await metrics? or should we just do this as a separate metric +Verified in jbt build. +Changing test class to an enum class results in linkage error in jdk where it works in jdk +As for the bundle version it basically is not important what version we set but I would start with version just to signal this is something new +Closing I found out this is related to a row with millions of columns +Doyou have any other ideas how this problem might be reproduced? have this bug too +You could have added the dependencies thatd make your life could +Added to +Its actually s a patch that should hopefully fix things up wrt get +Cant reproduce this anymore with the new embedder committed to trunk +MED have been changed to include a schema and year +Thanks for the additional info. +You must provide a simple easily runnable test case +I do appreciate any clues or hints people like yourself and Filipe can provide because this probably is a core couchDB issue and I have no idea what correlation there is between infinity timeouts and the session HTTP interface not recognizing newly created user accounts +Verified on SVN Revision . +Check s javadoc for the details +See default one +The file +Is there something about the workflow that is required but not documented in the readme file +Hi SergeyThanks for the feedbackMay I ask you which patch did you use when testing? There are two one by me and one by MikePlease comment directly inIm not sure about the behavior +Im not sure I see how the NPE is a result of the lack of the wsdlLocation +i like to join your project +Committed +Now that work is well underway with AS all previous community releases are +So more accurately that commit broke things +api change moved to point Child Abstract class ctors are now synched between the and codebases. +Thank you very muchIs supported multiple params also with XML mapping +OK that makes sense +This is no longer an issue +Any update on this +Ideally wed have tests that check that the example app works where we could codify expectations from the example app to make sure things like this dont break as componentsinternals change +Hi ChristianI was busy this week +see previous comment +Thanks Uri +Thanks. +To track further progress see +Created +Committed to branch and trunk this should help fixing a lot of problems new users were getting and provide faster instantiations +Username jsimone +Thanks +The Maven source plugin to bundle all sources to the JAR was in the wrong place now a JAR with the sources is generatedThere is no within the sources JAR as the name of the JAR already states the version +created to address the comment issue +Ive tested with J it does fix the test case copy +One is to implement the way the server does ie using something equivalent to on the client side +Im using this class in all my SpringAcegi projects +Thanks! +And weve already got preprocessor macros that we can check to see whether each particular codec is present +Attachment has been added with description test XSLFO file +First I tried and then with latest nightlbuild please find attached screen shot let me know if you have any you do a screenshot that shows the project in the project explorer next to this screen +Sorry wont help if commit logsstable writers arent covered as well +However for the time being a remains the central hook to use for such +Let me add it to this patch +Note that these exceptions are catched and logged in +This is sort of but I am fairly certain Equinox does it the same way since I have talk with the guys there about itHowever I think I have committed a refinement to the algorithm with checks to see if the requester has no wire but still has access to the class +To clarify Page +Planned for early morning Dec st before RC release in afternoon +I will file an issue on the PDF after someone commits the +Attachment has been added with description copying the LVTT entries from method to! Applied! +Scenario aCreated the namespace with version with mutiple FS operations +And yes my experience is its something most users run into somehow The same goes for joins think rows joined with rows +What I really doubt is whether my code meets all requirements of the CONTINUUM project and whether there arent any compatibility problems since Ive been working with +Didnt run since it just modifies a native build fileIve committed this to thanks Andrew +Patches committed as svn revisions and +Thanks Scott. +I agree that installing all bundles first and starting them all next is bestIdeally the scenario with the webconsole would be +Im fairly sure that one is the random map ordering only because I never saw it locally until Robert turned on the random map ordering for java now I see all the time +However we are currently very limited in our resources and we have to prioritize +What do you understand under graceful handling? What do you expect +Not unless its on by default in the I restarted JBDS and picked a different table and now it is working +Filed so well catch this in the future +on lowering default maxWe should probably gather up all the information on required memory and put a separate section in the docs for it +Im not sure that we want to handle those this way though death may be a more appropriate response +Actually I just reviewed the pull request and you dont need to do thatYouve put the code in the createargstoparams method and this should already do the trick +Luckily you did the part which i dont have much unerstanding +Yes is more appropriate +C isnt Java so if you close something and it deletes pointers that would be needed for a later operation on a child object then there no way to use that object any longer +Just a nit please generate the patch against zookeepertrunk for uploading the patches +fantasitc! Thanks for the quick turnaround. +In the new format we continue to support this method +vitaly has done the seperation so code completion is no longer dependent on VPE? nice +What does this latch on to and how does it work +In general we should convert it to a implementation +Ill close this as duplicate of as we seem to have broad overlap in scope. +here is the patch to add jline m still unsure if we can use LGPL binaries in ASF code +So the way we have reasoned about it is that all the deletes except delete removes all the values that match after the specified timestamp if no timestamp is sepcified that is set to now +Do you mean desertification +Lets do this via a toggle to allow people who upgrade from pig to pig to have a consistent view into their data +Issue can be closed +the attempt thing is really needed only in the first part of the test +Patch is applied. +Hi I am using and modcluster still does not picking up the port offset? It only identifies what is set in the +User can use the new Publisher Web Interface to post an activity with a particular tag then use this new Demo Streams web interface to see the stream returned based on that tag +Current list of files +Looks like this is inactive now? +Grr +Thanks Pradeep +If you can create a screenshot for that then I can include it in the ref guide and close the issue +We need to find out which so that I can plan how to best fix this +Kevin can you come up with some unit tests for this or better yet some specs we can add to Rubyspec? I committed my change in r +The Apache Aries project is working on implementing RFC JPA support and should be the venue used to implement this support at Apache. +I think your database size method looks good but take out the commented old code +That actually had a as well +I dont think this is a bug anymore but I would like someone else to confirm andclose it out +Fixed at revision +I suggest that we change this jira or continue the discussion if you believe it should be part of the Corpus Refiner instead of the Corpus Server +It has been fixed in version for reporting +Attaching a version of the patch for branch you want to enable count updates during fsimage fseditlogs merge +I will commit it shortly +it cannot be included as long as it locks up eclipse fully when you have duplicate DLLs installed +Fixed on trunk r. +Fixed +fixed in But I would really suggest using properly formed JSON + This bug has been marked as a duplicate of +Invocations work as expected now +Indeed Marek the screenshot of the violation with codesample inside the code viewer +Could you elaborate on what you mean by an opaque structure? I do agree that exposing more info about the client cert would be useful but as you mentioned previously this is a starting point and therell be chance to extend it further if needed +Renamed the following testcasessrcmainorgjbosstestejbnointerfaceunit srcmainorgjbosstestejbnointerfaceunitmainorgjbosstestejbjbossxsdunit srcmainorgjbosstestejbjbossxsdunitmainorgjbosstestejbjbpappunit srcmainorgjbosstestejbjbpappunitmainorgjbosstestejbwardeploymentunit srcmainorgjbosstestejbwardeploymentunit in r. +Ive checked and am seeing this issue when everything runs on a single instance +Patch looks good +social add example +Right Weld is behaving as the spec intends +Once Julien applies Ill respin a RC +My Bad I had this bound to a normal and not the extension +Existing code does not allow any concurrent modification to input directoriesI agree that allowing concurrent modification could be an option but lets do it in a different jira +It is against the trunk +Again this works ok in windows and seems to have a problem when running on linux + is for master and requests merged +This appears to be because the output file hasnt yet been created before the editor is opened + +I experimented with this approach and it appeared working quite well + +Did you see this in? Just want to know if this is an issue in trunk +Then the client can populate a temp table with the results +Is your a dependency of your application? Or are you actually creating a gbean in your application deployment planUnderstand the general nature of your request +Strange that weve not encountered it before but it does look like a valid scenario +We will begin working on this and target for next CP release +not sure if its the best approach or not +bq +I just committed this +Jan Willem correctly mentions that due to a flaw we currently dont stop affected bundles at all but attempt to update them while running +This failure prevents updating testexternalsrubytestproc to latest +reported this issue on the SM codebase +We should create an issue call out the offending classes and add test cases that demonstrate the issue so that we can eventually add them into our unit tests to prove we are meeting our contracts +Make that a partial fix +what exactly was the beaviour? did you get a popup dialog or entered something in the output window +Sorry for the confusion +I have verified that the problem does still exist if you correct those errors +If there was ever anything confusing or missing from the doc youd often find it in the comments below +Committed a with revision + reference this is +Thats an excellent idea +bq +I committed this +All patches submitted and committed. +Easy to test start a PE insert with clients on a single machine +The fixes are in SVN +Committed to TRUNK and branch. +on shipping with multiple example configsI think that is settled + +Closing as stats are now upward for earlyaccess and downward for https +We just have to extend the existing framework with an additional safeguard to restart failed worker threadsOleg +I need the link which is not changing through releasing new versions to add into Seam ref guideI dont know why component is set to Build but Max is managing components and the issue attributes so ask him +So all issues not directly assigned to an AS release are being closed +I use my own mvn bash script which sets temporary JAVAHOME and LCALL env variables so that Wicket works with non EN locales as well +However I dont think weve addressed the problem actually raised in this issue yet so Im going to leave it open +appreciate if we can commit before someone changes the code again +Ok we tested this after field failures resulted in the TT going down +relativebasetrue sets basedir to the directory that each +I actually missed when this went in +We need to check for as well a failed commit can easily turn out to be a plain rollback if you look at the exception carefully +r to fix more issues +Attach another one +I could verify increased memory requirements but no memory leak +Thanks Kathey for the helping me review the change and submit the changesThe attach file is to attempt to fix the insert logic to avoid primary key exception and continue using delete cascade and improve the insert logic for with new schema +When the server constructs the response to a metadata request it will try and build the iFrame URL and part of the iFrame URL is the view to render +There is already an unload operation for modules in there is a bug as described in issue +Thanks karthiga +staging site available at +Ive replicated this within a test case and am investigating what is going wrong +Just saying that you may need to turn on debugging to find itIt works for me locally so I kicked the build again +Hi AkiUnderstood and It doesnt matterJust ensure we follow same practice when we work on same projectBtw youre welcome to take this issueBest Regards Freemanthanks +It will only generate a SOAP bindingYou cant invoke a service without an HTTP bindingThanksIsuru +We can still have a portal name that contains upper and lower case letters with out easiest way to fix this issue is to not let mysql use VARCHAR to setup the table we need to use something else which is case sensitive +So that I had created it as separate files +Thanks +If you are fine with using the engines objects then the simplest hook would be to allow you to provide an implementation for which simply takes the resolved command expects that youll throw an appropriate validation exception on an illegal access +Martin thanks for reporting and with the details you doHave you been running with your patch in production for a while? And did that fix it for you +you can actually edit the area just need to click if a workaround that makes the editor always grab focus on key or mouse click in the editor area. +But just my opinion +Whats your view +bulk defer to defer to. +works with gif jpeg but not png +those expected results were incorrect in either setting +A job should be able to override the run the job method with extra cleanups +I cant see or edit the field on Android in the emulator emulator or Galaxy Nexus runningIm not even sure whether this is a problem with Cordova properly saving the birthday value or a problem with retrieving it from the database because I cant convince anything to show me when the birthday is setGiven how much of a citizen birthdays appear to be on Android Im lowering this to Minor and unassigning it from myself +Im not sure why it is not working for you +Did they? I dont know I havent looked. +Ryan thoughts on that +was used when the thread dump was taken +Simply ignoring such characters is IMO worse than trying to escape them +resolved in r +Patch that need to be applied in addition to the original one to fix problems with that caused failing tests +I come from a world where there are no tests for code and this piece of code is particularly hard to test +Added to +An automated regression case has been created +I have just committed this to trunk and +Also added quite a few test cases that useAll unit and integration tests pass with these changesSee for details +I agree +As requested by users I updated to populate the system repo +not just the default +Can we remerge this branch back into the trunk so we can avoid the me issues around unmapped plugins and start using the new packaging of plugins that avoids unpacking them and maybe get some of these other problems shaken out +Mass close all issues in resolved state that have not received further comments in the last month +Attaching patch v +Regarding Joshs bug it is a bug in inferring the type of the indexBoth bugs have simpleBowen iterate with an indexFoo and then assign to BazJosh specify the type of the index explicitlyIm leaving this bug open and continue to work on it but there is a simple +Just confirmed the demo reacts to changes in both the url and the posted xml +new patch which reintroduces the functionality +Only or individually +Thiru any idea on the best way forward here +I just committed this. +I hacked this patch to test parsing with I still cannot get it to fail +My bad +In this version the connection is not closed while the write lock is held and a timeout is used while waiting for the connection to be closed +here is an initial patch +Will wait for validation result from Roman before closing this JIRA +Researching best option to correct +Here my proposal is if we enable persistBlocks feature addBlock call itself persisting the blocks in NN +has not been a problem since +Looks good +Submitting patch for review +Select Developer Studio from the sidebar menu +There were times when we had multiple implementations of a service interface that needed to be autowired in these cases autowire by name suited best + the heck add exclude list dont see a way to specify at the CF level which seems like a common use case +This looks good to me +Issue was resolved several months ago so closing. +will see can Christian become a committer +With the release of Ive removed all earlier releases that have javadoc issues so patching is no longer needed +This compaction strategy is useful for time series data +The error message is the same it seems this is a duplicate of +The POM should have been rejected in the ftpserver build and will fail the build in Maven +Waiting now for a new report +and theres always the option to not change any code but just javadoc all the behavior as a fix providing stuff in contrib or elsewhere that works correctlylet me know what you think +Just to confirm the fixed type doesnt store a length byte so there is no extra overheadThats correct +Logged In YES useridanother diff file +Philany updates on this? What will be the fix version +good catch there was a code change after the DDL scripts were created and thus they do not match on some ids +I figured the comments crossed paths +Ill update as you suggested in a few wonder why the ITs didnt catch this when I ran them? Surely I would have seen this same behavior unless the pom files were change after the fact +I dont think itd take much time really and I think the code would benefit from it +Add stanbol layout to ontonet and reengineer web component content are empty +A first attempt at a patch for this issue +Thanks all for the good comments +Attached the patch +This was fixed in revision . +Thanks to Kihwal Lee we got a workaround as followingAdding the following in HADOOPCONFDIR in only client side get rid of two extra INFO loggingWARNconsoleWARNconsole +Fixed will show up in the release notes +I verified this. +These changes were committed into SVN trunk with r and r and into the branch with r and r +Merged into master. +However I was very disappointed with the Java libraries for openid +Oh one other thing lets not use xdoc going forward +I think we should just add a bit more randomization to the tests before committing +Backend project consolidation will follow soon. +Fixed applied to +deferred +Duh +Can you provide a diff version +There is no easy way to make it work +Seems reasonable to me +Ill look into API more and seek a better way +As Adam says discussion is about as far as things have gone +I reproduced on my side as well +exceptions are not considered for treatment +stole should be a better name for but Im out of inspiration +I still went the derived class way because I got hung up trying to work with the registration in +Weve discussed that the dist we ship will have a proper developer oriented standalone setup out of the box +Heres the fix for the +What version did we add the column limit code in +Affects Array +In my first post I used the option fileSet +Here is my revert of this on trunk and +war +Done +I committed this. +sorry I thought you were referring to p install you are just comparing downloads then x more direct installs is not surprising and the numbers verify users actually download it +Breaking compatibility screws users regardless of distro +I wasnt aware that GAE uses the rather than the annotations thanks for pointing that out +This problem remains in +So my question is if the portal thread and the filter threads are different how can you expect to workAm i missing something here +Attached simple patch to fix +The google guys already liked the spec so that wasnt a blocker but itd be good to cover all previous use cases +has been marked as a duplicate of this bug +Fixed in rev patch that fixes the logging statements + +We need to put the iteration in a synchronized block +Im not totally happy with these class names yet +Fixed in trunk +Refactored the code a little bit and created a new class for detectionThe code relies on already available information the filewritten to the bundle private data of the system bundleImplemented in RevisionHow about calling the referring to a regular startup just STARTUP instead of RESTART +The replication is my complete couchdb log +Based on that if other issue didnt do that bootstrap I can certainly see the issues being the same +Seems like something we should fix for +Had some problem running the new script but am a willing apmail guinea pig if youd like me to try I used your emails as the moderator Brett thanks for your help +Thanks for the quick review Knut +This behaviour is defined in The annotation is used on a method that needs to be executed after dependency injection is done to perform any initialization +In case you add a new test case Ill be interested in having a look at your code for another issue Im working on +I am currently preparing a patch +And once again sorry for submitting this so late in the release process +Committed +bulk move from M to an issue link to +That is the point +I think that you forgot to enter the required Site Name and thats the reason for the errorCan you try again and confirm before I close this issue please? was using the svn trunk code the productstore ID for bjfreemanteststore is but the code passed bjfreemanteststore instead of please check again +These blockers and critical issues are resolved but not verified +Great nice to hear it. +Fot the example of luis this means key and x only exist inside the for loop +The attached implements a change to attach all the namespaces in scope on an element node being detached to the element so that the detached node and all its children will still have valid namespace references even if the former parent tree is freed +Not surprisingly I see the same problem on linux +Its definitely a bit hacky but it works and it seems the easiest way to allow us to add new fields +Reopening +Committed revision +Should be available in tonights M snapshot! great! Thanks. +This could be related to this still valid +all issues not touched in a month to an attempt some time ago to pass sortMissing on to the currency value long field but unsuccessful +it seems more appropriate to throw a than to return null +Is this a runtime error? Closing this one for now +On my side I need an interface to represent any implementation that takes a set of servers and verifies if that set is a quorum or not +This issue was corrected closing. +Defer all unresolved issues from to +doesnt look like it will make to as there is no patch for consideration in +Note the Installing section in Those will be the instructions that will be in the installation guide +Should be easy to backport +This probably makes itless efficient when writing large files because reallocating a single bufferas the file grows is less efficient than adding new small buffers +The failure is the result of the new test which still needs some cleanup +Lemme see if anything breaks I think that could cause a test to fail since we compared the seg infos though +Looks good +Have founded the problem +Theres some ripple on effects in and the root licensenotice in the source tree +The culprit was a silly usage in of a manifest constant defined in have moved that constant to +I have fixed point and by creating new zip artifact variants of cppunit and expat and uploading these to nexus +Because of that I said one could use a tool to enforce optimization or the new can also do the copy old to new index +FYI in general no gratuitous changes to the formatting +I also checked locally +This feature has been implemented without directly using this patch report +You can also use that to convert it to a I implemented that for I think it is for the best. +Im on backporting thisReasons In my opinion is feature complete at this time +Defer all unresolved issues from to +That Simple header list doesnt quite seem to correspond to the lists from the WC spec but okI agree that a method for customizing allowable headers would be good +It shows your dedication to the quality of Derby which I and I suspect the community in general really appreciates +user adds any notes or attach any files for report +Thanks +Committed to MATH branch +The version for testing will be updated to by +here is a proposed fix with parts master remove the assignment plan when we offline regions! regionserver add a new sync block to prevent threads from closing a region at the same time +Yes +But how to verify it here instead of verifying it with the plugin +Which is only to say that I dont think excludedocs is responsible for the issue that you opened Marek +Patch allowing to give options to the SSH variant was applied in r Thanks! +Good point about the two halves workaround +I dont know if we should fix this for since generated advisors are not really supported there anyway +Tests passed cleanly for me +Could we instead try to openextend the current code base so that we are more free to extend it on our sideCan you list what you exactly need andor provide us with a patch +The patch also adds Maven as a requirement in ll commit this soon if there are no objections to this +Now you showed me that I should check out and it works the first time I tried +Oscar theres a few issues with some of your changes +I will make sure it is mentioned in wiki +We might need to tweak the query code a lot to allow for this type of mechanism to work properly and Im wondering if the benefits outweigh the costI was considering bootstrapping in some way a property index on top of that hidden node state but looking at the it also works with s so wed still have the issue of extracting the from a Tree +The attached patch adds getset methods to and uses them to control the encoding used in zip streams +Just ran another test with redelivery attempts +Patch applied in r +I dont suppose this fixes tooIt doesnt no +Agree +Will commit when it passes tests +I meant JBT M of course +If the and are fully compatible and let Tika run fine no problem I will give a try +I also moved Spatialjs which is tightly related into the same destinationIll commit shortly to both branches. +Looks like all the elements are covered +Cleaned it all up +Ok wrote tests found actual bug in the code that manages the linked list. +This should then be fixed in +While that solution may fix the bug it doesnt solve the issue +I think we should at least fix up the current samples if not include both now +Confirming that this works for me +Uses the new in the namespace to handle the entries as well as cleaned up general code revolving around the entries field +thanks. +I am having difficulty reproducing this +The address binding error is caused by the datanode rpc server port which uses the default port number instead of the free port +Looking at it +Fixed and pushed as to npm. +I reviewed the changes Robbie made in the light of +I thought not to break somethingWe could create it always in the short form +Scheduling for since this is a Blocker +The idea is to instead specify a maximum number of index terms to cache in memory +Initial implementation doneRunning Tests run Failures Errors Skipped Time elapsed sec +Please if you have some ideas +I have no access to the you change the doc accordingly to the change then +Switching all issues that have been in resolved state for more than one month without further comments to closed status +above carrying stateful filters across regions it cant be done easily in because we can get a at any time +To display the button we need to know if the process has a form linked to d need this accessible in the task view as well in order to determine if the Complete button on the task list will open a form or actually complete the task +Personally I think it is useless +What is the readerThe service that presently sorts the logs +I am no longer getting this after using the jars from +I cant reproduce on a different device running +added that makes sense thanks for clarifying +I think the problems arewere mostly caused by the use of fixed port numbers +If theres any protocol that cannot be implemented on top of Netty please report it as a bug +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be +seem like is not in the patch +We should implement another data handler for Hive. +What version of CXF are you using? If its not a recent version please upgrade +Thanks Satish! +I think we have a couple other cases in the Hadoop tests that have this problem as well +Thanks +These were very complex cases to describe here and I was able to find another ways to resolve my problem which was related only to the static which Tapestry generated for the HTML elements +I am marking as a blocker for this bug as we would not be able to verify the fix for the patch if this one gets fixed +Im checking for alternative ways to solve this +This is occurring on index reproduce +bat files and the has shell scripts for Linux or Mac OSX +Im sorry for the delay in replying +Hopefully the tech writers can field this oneIt would be appealing if the documents shared a common file of parameters +Bulk close for Solr +Id like to add the steps on how I got this NPEGiven Sonar IDE plugin is activated for a project in Eclipse In the Sonar perspective The Sonar Measures view is openedThen I opened a file in EclipseExpected result The measures are loaded in the viewActual result I received the NPE mentionned above and the measures are not loaded in the view +I think we already have those too +Fixed in revision . +The closest the old CMP implementation had to optimistic locking was the Instance per Transaction Policy +Ah looks like this had already been fixed in head +Try different sqljdbc mappings +Thanks for the patch Jim +Just thinking out loud havent read the design preview email on the dev list Does the method need to be in the Cache? Couldnt the method be in a particular component which gets injected +Ill assign this issue to a release and in the meantime forge ahead with release process. +Patch is applied at +Implemented with for the EPSG code and for additional codes. +Deleted the subject exsd file +The issue is caused by a conflict lib version of Xstream between the one in Geronimo and the one in Hudson +Thanks a lot Eli for updating the patchEven though we are running the tests with append off better to handle it +Thx for the the apachedsdocsusers is out of date in two respects +If the ordering in the child POM is respected the bug is fixed +can we just change the wrapper name class instead of chanhing the c code +The specific input boxes for several properties were not being correctly set to null if the unset box was checked +It can then be accessed in a browser at httplocalhostNote that the standalone EL implementation requires Java due to the use of generics in the base EL resolvers of the reference implementation +And youve enabled the native query server +HelloI tested this patch and for now it works +the Apache snapshot repo is listed in the poms but it shouldnt be we should never download LuceneSolr snapshots as part of the build +I have reread and it seems that your patch can solve itI just realized that you may have patched the source you got from the distribution archives not from svn didnt youI would be great if you could checkout the trunk of recreate your patch from there and attach it to +Minimal sample same exception there +The Job Tracker also gets an ERROR log message which is incorrect and should be removedOwen Could you give some more details on this? I didnt get which ERROR message you are mentioning here +There was a reason I ended up keeping numerics as literals in the SQL +I have run junit All on this patch and there were no failures +Everything gets packed together in an ear file +Automatic GL posting service triggered when inventory is produced by a work effort +Do you have a newer patch or hack? Whats the error +This would be really helpful for the poor developpers who are often deploying artifacts with wrong version name and do not have the rights to delete them from +bq +If so we can not commit this patch into trunk right +To get the running i had to copy my importing XSDs to the homeworkAxis folder +Ideally I think this would be done with an rather than a but that doesnt seem achievable with the interfaces as they stand +Why arent you using the reverse DNS style +Added similar issue for the new Messaging +If you do periodic checkpoints using secondary dont you check the correctness of the namespace image by that +Committed to trunk. +The issue is caused by rpm reorganization in ACS +Still need to undo parameter setting in Multi so can let go of +Finally fixed in H and H this fix should also have taken the default catalog into account to be complete right +Maybe add the copyright headers too? +committed +experience so scrap a more relevant example is whether anyone ever thought to combine the war and cargo plugins I doubt it +Thanks for the patch Jonathan! +Youre right Ive attached the requestresponse captured through Fiddler +OK could we please replace the Middleware banner as the event took place already on th April +Ill commit to trunk if there are no objections could you to the release branch SteveYes Ill merge into branchx and lucenesolr +Reopening the ticket +I hope our tests account for that +turned into relation instead of subtask +changing to to match the Push release offerings +Will we retain the mirroring functionality and just identify relationships by the masters URL then? That seems like a pretty good way to do things except for one lingering problem authentication +This was a decision made many months ago to attempt to facilitate the sharing of server objects that reference runtimes by name rather than some obscure ID +So I think we need to write login module if not there one already +When mouse cursor is moved over Visual part selection in Visual part appears immediatelySee attached screencast +thanks for the reviews Todd let me see if i can generate the trunk patch again +this is different to did you determine whether this was an issue with the shell script or not +Patch + is a bug +If I just get the Enum type then itll how can guess that Invalid have to be represented byDo you hard code the binding in the itself +This issue looks very similar to +Thanks Travis! +Indeed! Removes a whole lot of paramaters from the function +Triaged Dec as a blocker +Removed unused methods from the module in commit all unnecessary casts in Git ID for the s URL in in Git commit commit fceacbdeddffaa fixes double nesting of generated HTML PDF fixes typos in simple addon docs adds useful tip to release instructions in +can somebody just the custom sharding tests to a new class +How is the full locality more than the hostname +Im not sure what this buys us it basically starts the micros component at a random point in time +has been marked as a duplicate of this bug +Agree that H is not ideal for the task +patch is a patch to resolve the issue and an update to the test that also tests and fixes the negation operator for the media type as part of +This can be moved back to a specific release if its going to be available +Deferred to s one option +Could you also post a version against Yuki +bug brought forward to visibility +Its worth remembering that once you go to JNI you are more at risk from memory leaks pointer problems race conditions and other CC coding issues that can affect programs +Ill provide details in the next comment +I also understand the preflex problem and I totally agree with the solution here +the patch breaks compatibility with existing encoded property reproduce backwards compat breakage see until there is a good solution found. +how large the matrix iswhat is the size of the blocks +I initially named this runner junitcore testrunner and I think the name should be changed back from junitjunitcore since this provider is really about invoking tests through the junitcore api +Oh I see +Hope its ok to assign this to you Justin +This is a patch of tld file that includes the html scripting events +with minor style modifications +Ben in my case I can wait for a solution to be included in some upcoming Seam release +This is being worked on in Cargo land +the first part says its a second part says to leave open as that is a jbpm issue +bq +One more point to note is that when both apps are deployed together they work correctly because of the Domain picking the right service +Patches for trunk that deprecates the misleading property and provides a more relevant new one +Updated patch passing sourceID and vertexValue to the edge +Perhaps Tapestry should change to render all attributes in single quotes +This is because from an SQL perspective the WHERE clause would follow a SELECT clause and so will confuse user if the second format is introduced +Luke +The overhead is that we need to maintain the code and also have sufficient expertise to tackle issues raised over user group +This is fantastic stuff! Thanks so much for the patch Ill let you know if I have any issues applying the patch +Thank you Yuki. +I mirrored what scans already do I put the key extent in the thread name +Heres a patch that should work but the analyzer will have to be the same instance for the fuzzy equals hashcode +Thanks for the report and test case! +Committed to trunk and +maybe this can be solved using a flagujaxmoveFlagsreplacewhich will enable this patch including the check for replace +Dropping priority to minor assuming Martin meant that was what he originally intended. + m still able to reproduce this on correctly in you provide the link for this build? This is not the binary I see where I have been instructed to get builds +what do you mean it isnt on the branch +Is vmbuild SolarisIll get an instance of Hudson setup on vmbuild so we can build up our consistency across platforms until the big iron shows up at Contegix +More over the message does not talk about conversion error but about a mandatory value from an another validator Or to support both date format according to the date format given by the backing Any message at the user screen instead of having a silent error +Fixed in SVN check in +So I submit this one for build a release +if you want to exportimport users or acls and it does not work right away and give you some weird errors like jcrprincipalName is mandatory then you pretty lost +Bikeshedding suggestion what about describetokenmapBesides that it looks good +Does the newest version fix this problem? If so what was the problem? We prefer not to update unless there is a specific fix +well be it bug or not it was accepted before so I think it is not good to have that changed now for +has a script in place for automatically mirroring our repositories. +Please reopen if solution is not since we now can use component to do this task to revisit the solution we ll leave it like this +duplicate to +Lukas can you please assess +We should try to contact the author of and ask for LGPL license +No progress has been made +At least then Id have control over having it do ornot do initializationLogj does not provide a default configuration +Thanks +fixed at r and created which is related but this one ended up just not respecting its Serializable nature +Net results in one of the few ways Ive encountered to completely defeat the CRL garbage collector +Im abandoning the sub tasks since there are changes in a given class across the subtasks and finding it hard to maintain separate file sets +Attaching a patch whichs adds a reset of the current rolein reset and added a test case toI also added a pointer fromRunning regressions +I propose that we fix this in +Tested manually with Java Java and Java. +bq + +Attached file contains lib sandboxsrc sandboxtest samplespetclinic derived from petclinic in Spring main distributionI will update petclinic sample to include Cache JMX managementRob could you kindely commit this in cvs +is using allow true +The code was for a project at my work and Ive been taken off that project over a year ago +will if you can try out the patch ive attached to and let me know if that solves the problem you were trying to describe it would be much appreciated +Very good job Sergey! +I will reopen if I discover problems +Committed new test to repository with revision closing. +You dont want other people to post to right? +its a part of baremetal donation that has been in I forgot merging UI part when I merged baremetal in ACS +I tried to use czech phrase and it works but czech letters are not shown even if I set up the exisitng font +Pushing this inOrthogonal issue The related params dont seem to be following the naming pattern of other configs +If need a particular feature or bugfix to be included in you are welcome to get involved and contribute +I cannot reproduce this situation +New patch +Removes heapSize from IncrementFixes broken test test broken by this m not too familiar with the RPC area of the code but I thought everything looked good conceptually from the perspective +For standard development work the cost of building and deployment can be significant when aggregated with no benefit + is new in +Ill backport to x first +I am attaching a patch which moves recording to trigger action spss dependency on trigger table from create trigger constant action to +Basically I think the decision of whether an enum or union is valid should only affect the enum or union itself +haha typo fixed +I think this is low enough impact it could go into the branch too +Attachment has been added with description input crash does indeed happen in but it is not present in trunk +Resolving. +I appologize for that +Using +Now it works thanks! but thats just weird because I used the same connection URL at connecting to the database right after it was created and it worked! How is that possible +Another thing I have noticed with this issue is that it is more pronounced with large number of map processes +Was committed to trunk a while back +there are many possible reasons for this not working +So marge or marge de maniobra could be good translations for leeway +The existing list looks as followsTemporary tables cannot be specified in the following statements ALTER TABLE CREATE TRIGGER CREATE VIEW LOCK RENAME +I am currently working on a testcase that uses everything required to reproduce this and I am going to attach it soon +Please attach a Roo script to this to confirm can you remove the annotation value pair and see if the error still occurs +Currently it backs up homebdelacretaz homedaisy homeconfigI saw no value in backing up any of homecdemosBackup currently runs once a week I can increase that when you start using the zone again +They look so similar that I got confused with. +However I would be in favor of an exception that provides more information to the developer using Apache CXF. +sorry +As far as I know Grails creates a +I just committed the patch. +Stack thanks a lot for your testsThis bug is very weird +If there is a problem please open a new jira and link this one with that. +My concern now is what happens if my master goes down +Everything else looks good +Fix is pushed. +that follows the root path completely +Note that even when things are made to work again on Windows after this change it will still be an incompatible change since programs that created Paths with backslashes wont work the same +not sure its related to the install plugin +Im not sure why +I think it is more important that does not depend on +Issue is fixed +Lars and Stack Thanks for the reviewTed Thanks for integrating this +Could you change the test to wait a bit longer after bounce +Here is a move issues to and + +Smooks editor is no longer being developed +neat idea could you give us some usage examples for thisCheers revise. +I recommend for now you use the selectKey element to grab the value from the sequence +Contents of this patch areM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyiapireferenceM javadrdaorgapachederbyimpldrdaM javadrdaorgapachederbyimpldrdaM javadrdaorgapachederbyimpldrdaM javadrdaorgapachederbyimpldrdaM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangD javatestingorgapachederbyTestingfunctionTeststestsjunitTestslangD javatestingorgapachederbyTestingfunctionTeststestsjunitTestslangD javatestingorgapachederbyTestingfunctionTeststestsjunitTestslangD javatestingorgapachederbyTestingfunctionTeststestsjunitTestslangM javatestingorgapachederbyTestingfunctionTeststestsjunitTestscompatibilityM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterjdkM javatestingorgapachederbyTestingfunctionTestsmasterjdkM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestssuitesM javatestingM javaclientorgapachederbyclientnetM javaclientorgapachederbyclientnetM javaclientorgapachederbyclientamM javaclientorgapachederbyclientamM javaclientorgapachederbyclientamM javaclientorgapachederbyclientamM javaclientorgapachederbyclientamM javaclientorgapachederbyclientam think there are some files left after the v patch +I will work on migrating it to codeline as well +Hi Sorry for the slow response! I just took a look over the patch good catches +Integration tests were updated. +Appears to be long defunct +in the log +You should be able to add it to with little changes +Ill commit this tonight or tomorrow +The fix is for it not to do thatNote Ive also done some minor cleanups in this file simply putting back the uses of Java generic types that I was required to take out when I originally checked the code in because of the need at that time to support JavaI think this fix should go in since it is a serious bug with a pretty small safe fix +Running through tests now +I will review all the other classes to close this for good. +Its safe to remove the delay but since its only required once in the clusters lifetime I think its okay to leave it +Thanks Ying +Just a status update as I head to bed +Are there any dissenters? If anyone that has the power to commit this doing so it would be great +It seems to be resolved under Geoserver its not yet a stable version. +Removing the following seems to fix the issue remove modulessaxon project remove reference to saxon project from remove saxon download ant script from remove saxon download ant script from samplesClean build from branchClean build from source distribution generated from the branchNeed to do some testing with samples that have saxon dependency just to double check +Could you test the patch which was submitted on Dec for me please? + +But under this scheme its diffs that are worst +closing +I wanted to simulate throwing a on a proxy call but Mockito cant mock a proxy object as far as I can tellSo all those changes are just so I can setup my own RPC Invoker to run which just randomly throws +Its a part of the HTML spec though so thats why were going with it +Pending update of trinidad sandbox +just remove the branding of smooks for should remove it from seam and hibernate too +The value is stored as a local variable and then used to pass vhost information off to jinterface calls +Would anyone want to try this +hbmx +I agree with Benjamins comment +Applied the supplied patchAdded some extra handling to make sure streams are always closed. +I will do some debugging on my system. +With the JMS is called irrespective of whether the connection was automatically reconnected or even if it automatically reconnected ok youd still be then closing the connection and recreating it +Stupid mvn does not resolve pom +Therefore something like true should be available. +Thanks for confirming Manukranth. +Thanks Mahadev! +classpath file for forums updated +this diff should catch most changes thus far and includes speculatively +Running test suite to check particular components. +fixed +After updating to the latest trunk one test in here is not passing with my patch +Logged In YES useridI have added when upgrade to write lock failed +Patch attached +We might see a FARSDFARS notice or some other text that refers to the statutes and regulations that set out the policy about copyrights that you stated +This batch is backward compatible +Thanks Yanbo and Radim for the patch Alejandro Tom and Andy for the reviews +Ahh I didnt realize that was what the checking was for +If you have questions on how to go about writing a patch or Jelly questions please send them to the mailing list and Ill be happy to assist +The patch looks quite simple +I tested version built on July +class file under targetclasses +You can fail producers block them or spool to disk +Here I could see one conflicts say when a new bookie joins with actual hostname as and port +In high screen resolution I am afraid that screen width is enough to show every things raw formatting issues has been polished +Could someone create working TC to prove the theory? However we will still need some mechanism to timeout poisoned consumer to move prefetch messages to others +So the CAS Editor cant highlight single annotations with the annotation style settings? This would be a bigger difference between CEV and CAS Editor and its somewhat essential for some components and uses cases of the +It appears that the problem is some kind of jar conflict between the dependencies of jBPM and the dependencies of Seam +That makes it a bug +attached diffs for pluginpostgis src and test directories +Guarding it in the header means that clients get a compiler time error instead of a link time error +Sorry for not reading the debug carefully can be closed. +I just committed to all branches and trunk +should be fixed in the current svn branch +Sending srcmainjavaorgcodehausmojowebstart srcsiteapt file data +tld files are present under projectTag Libraries +Mockito allows you to work with the class directly but it seems that there is still advantage to working on an interface +Rescheduling for CR. +Rackapace uses Ubuntu still so both pathways are tested. +Attached patch fixes the issueBefore the patch the randomized fault test failed about of the time +wont be release until next week so until then Ive patched the tarballs with this fix. +but works on Jetty +Fixed in masterafc anddcd. +Ive just committed this to trunk and based on Bobbys and my own testing +com site should be fine as long as is it updated +bulk defer to +This test was failing on trunk because never ended up getting applied there +Do you file a bug on Maven Jira +update info for runtimes as well +Fine I will provide one +got the same patch failures on a similar environment as Hardy except git version is CuneytCan you try tag or masterThanks tried master and it worked +The issue is that configs dont set the correct permissions +I think the expectation here is wrong +Duh +Let me commit Jon +was there any more progress made on this issue I am similarly affectedThank you +Can you elaborate a bit more on what kind of you are trying to enable +The test failures in streaming do not seem to be related to distcp at all +Any particular reason you didnt apply this patch yourself +For plugins we bundle those includes ought to be available anyways in the build tree no? The point here is to make them available for rd party developers no +Even Spring webflow has not been released yet we know what we will do when Spring WF goes out +After updating the configs you can run this command as the hdfs super user hadoop dfsadmin refreshNodesThe same will need to be done for tasktrackers config on the namenode needs to be the correct one before your run the above command +Arun in the last version of the patch Patrick moved from to +The attached patch should apply cleanly against SVN trunk +i get the same error if I run it from the child project +Security schema moved to module +It should work now no need another patch +Since user can create their own I dont think its important featureMoreover once separated from Graph job it will make able to specify desired number of tasks user want +With patch from applied and maxallowepacket set to M everything work as expected and error doesnt occur anymore +I committed the patch +I dont like it +Patch +Ive just committed this +You saving a couple k a socket but give yourself many hard to debug problems +Im working on creating fix for this issue +Im confused by default cxfapacheorgutilnulltrace functionCan you show me your orgapachecxftrace functionIts not simple function +Closing question that should have been posted to the list. +Logged In YES useridI am currently dealing with the same problem +Do you have a draft notenot really I still need to figure out what this change here exactly added +Yarn clients could have their own own property which they would set the to for the RM clients they create +Just one comment there are some simple examples for flash scope in It could be good to try them with the patch you provide +Fired the change event from the setValue in respin +For sure this is an expected behavior when launching Sonar but are you sure that the Sonar server doesnt automatically recover when Sonar has been launched successfully for a while +Can you run the test +However the point is still valid although is slightly different Im able to read bytes after seek +Thanks looks good although its not. +Can you please attach the relevant files one by one as it seems that the archive is corrupt +Work in progress +Have you tried using the latest NMS SVN codeHave you tried using a brokerIs there some other step to reproduce that you havent documented in this issueI see above you are referencing code that would be invoked if a Failover occurred are you stopping the broker at some point during this test and restarting +This saves the round trip to disk and also since the order of input to mappers usually is not important the second mapred job could start once the reducer of the first job starts to output +YAGNI. +Fix Versions includes as this will probably be a small and safe fix +Itlooks more like you added this so you can work around the problem in yourenvironment +If we make the and the Processor itself more not depending onBut the update processor should have access to +Move to +At the development time of the menu we decided against the possiblility to have commandItems at top levelSo this is currently not possible +This should not a SMB join this is due toCan you apply the patch for and check if this works +Actually I dont know what I was referring to as upon checking there is no such thing as a display name in the spec +But ivy fails because of the bad checksum +Im theorizing that since these were real projects with many classes and web pages the longer processing time was what made the errors show more consistently +The format command prints a usage message explaining the usage so the issue is not as big as it is made out +Thats really the only way I can imagine this slipped through cracks +The reason for the files to be in the project is that they should not be just about the adapterThis file should also always be created no matter if users chooses to click create and setup server adapter +It would be better if the caches were loaded only and correctly invalidated on cluster changes +Committed revision Thanks for the catch Akshay + will remove Continuum from Xircles soonishly +In a clay HTML template all markup that is not bound to a component becomes a verbatim component +We can it if the problem comes back +Its tricky to test it automatically +wouldnt that be a good thing to have +Lets plan it for some future version. +Im not going to even try just see if it works +thanks +Ill follow up on the linked issue +Im attaching and with changes to two topics I fixed the table entry for in the Derby properties topic to say Static and I made the lock compatibility topic almost like its previous version except for replacing and with Yes and No +I will open a separate JIRA issue for the copyright statements and Cloudscape references in the demos as it is a separate issue from the demo package names +Marking as resolved +Attaching a patch that adds the reverted change back to the parser grammarThis lets the tests pass +Adding mainsrcmainjavaorgapacheservicemixkernelmainSending mainsrcmainjavaorgapacheservicemixkernelmainAdding mainsrcmainjavaorgapacheservicemixkernelmainSending mainsrcmainjavaorgapacheservicemixkernelmainSending mainsrcmainjavaorgapacheservicemixkernelmainAdding mainsrcmainjavaorgapacheservicemixkernelmainTransmitting file dataCommitted revision Thanks a lot for this awesome patch JamieWould you mind taking a few minutes and add some documentation about that in the Kernels Users Guide +Specifically Im interested in the example generator use in which seems to absent from your patches +Andrewusing JSE andor JEE as examples is not valid since they are specs and they are implemented by applications in a completely different hierarchy +Its the coolest thing Ive ever seen! Thanks very much for getting this done for the GEP release. +Triaged for +Considering that EAP wasnt released yet is it OK to run the regressioncompatibility tests on CR +Correct me if Im wrong but this would also require code distribution and restarting of +Patch for trunk +Test applied confirmed fix applied and confirmedThanks Mark and Fredrik + this is a good idea + +Changes look fine to me +Thats a fair point + +Renaming to reflect abandonment of gzip compatibility as a goal. +Demoted unless there is an argument that this is a serious regression from earlier +The xml partial response +Fixed in trunk r. +It looks like were missing the jdb flag in the native launcher which rvm installs by default +commit eadaecdcdbfceedeMaster commit since is now released +Im afraid this has gone stale again could you regenerate it please +Current code howerver increments offset by default regardlesswhether the character is pushed into the bufferIf thats the case then there are more places that needs to be fixed. +Adds a config option to to prevent including inbound anchor text in the index +forgot the version +I thank a simple way to do startups is using a round ribbon assignment of regionsJust a thought +I attached the picture +The help message for text is still missing +I checked the URI and its fixed in the trunk already +Im going to give it some testing to try and see if this gets us any speed up +Since we let users dedicate up to of the heap for the block cache and Memstores we should fully use that space by default +However the test suite may run on JDK or as well the latter is why we ship the in our said It should not matter whether is on the classpath as long as youre compiling against JDK simply since the JDKs own JMX version in the system classpath should always take precedence +It should show preferences for current selected UI element as editor or view +please attach the this issue as incomplete because the reported didnt provide the requested information. +This is actually a bug and not just an improvementWithout this patch the Maven build will attempt to create Javadoc based on the version of the artifacts it finds in the repository and then goes on to build new versions +I think you can merge this check with failedkilled job +Excellent committed this to. +and updated the reference in the launchpad bundle list in Rev +One of the reasons I was motivated to change the javabin format is because i went and actually looked at source code to solr clients in other languages and saw that they implemented the protocol wrong no surprise because Modified is really a crazy java thing +The last test in test leaves one entry in the table +migration of to private was left off the request but I assume it should be done anyway +This bug has been marked as a duplicate of +But in this case there should be directory containing not compiled source code +without eJOPR no need garbage collecting for a long time just minor CPU spikes +Isnt this handled in +Would that workThanksYogeshSure absolutelyColm. +for the fix to the group freeing problem + +I opened +review for the patch I gonna review it +Submit patch +but that is a problem for some classes +Of course I can create smaller patchesIm sure you know that this will not reduce the actual amount of changes to reviewBut if you rather have files instead of one thats fine with me +Yeah +Hi CarlosIs there something else thats wrongRegards you dont own the domain so groupId could be lets go for mavensynchomeusersoououaibskymavenmreporsyncsshOuaibskyouaibsky +Umbrella task to investigate all the HDFS API changes we put into +It was merely made +I always assumed in this case its all or nothing in that people wouldnt be checking out leaf node projects and trying to build them +Stefano can you take a look at it +If you used the latest version then we have a blocker for our Beta which should get released in a few daysIf it turns out that your test tools are simply outdated then Id close this as +Committed to and trunkThanks Hao Zhong +However if you authenticate on node and then move to node you must hit a page within the same app that you originally authenticated on in node before accessing any other app on node +Hi FlavioAt first pass this seems to indicate that we cant replace a failed zookeeper server by a new one but that statement is probably way too strong +I removed all afm files and all cmap files which are part of the +Hi Vasanth thank you very much for your input! Allowing multiple answers is certainly a good idea +However I must admit that the current implementation probably covers of the real world use cases +Approved for. +I wonder if a more thorough unit test would be enabled if we refactor the logic of deciding if commit is required into a separate method and directly test that method with all types of tasks +Attached the effective pom with this +Thanks for this +Fixed in r +any chance you could work out the patch for on as well? I got a zillion conflicts on svn merge Thanks +Proposed fix +If were doing this Id much rather see PBKDF than bcrypt +OK so this is apparently the HTC Evo D +the fix without turning lazy mode on +Committed revision +The namenode start is doing some work to set up directories after the daemon has started so it needs to stay there although it should be made idempotent +We basically have to resort to having a local fork of +How would such a Groovy specific file class be better than the encoding property java uses to set the file default encoding +Bryan Since the has gone through various changes to some extend would you pleaseapply yourpatch to against the latest code base? thanksRgds from Kevin King applied as this patch does not work on the latest sourceCould you verify please. +it would be a great help if Thorsten could try to compile his project using the release to see if the error is still there +Changed the log message +not in tests in DFAtransitionS and and also have many concatenationsSolving them alone doesnt fix it +This is an old task but I think its relevant to what we were discussing last week +If people switch logging through SLFJ it wont work Clicking multiple lines without making a selection leaves hanging there any way to give users the slfj level options and have those translated automatically behind the scenes into the correct levels for the framework thats actually in use +is a wrapper of real Zookeeper instance +Changed implementation to use state variable rather than value of the CL +Added new templates that change the sucker password in deploymessaging +Please clarify which one will be used and if we should remove the other oneAs for the second question not sure what the current policyprocedure is so Ill leave this to others +Sorry if I wasted anyones time +Thanks for working on this this is really coolI committed the first patch please pull from subversion and merge any changes you have if you could the ticket open for now just keep attaching patches if you have more work +Can you check if it looks good +We dont have enough horizontal space in the navigation bar to have a separate Documentation category +rev domain has a resource adapter installed +Closed after release. +Thanks Nick for the patch +bq +Its machine generated +So to keep serialization compatibility this inner class is needed +Experiment shows this not to be the case so I feel its ok to defer this +I committed this. +However I dont believe the failures Im seeing are due to these changes but reflect my inability to get the testing server configured properly +Attaching an updated patch to account for that + I think the user should view the app master UI through an HTTP proxy running as part of the RM +On Oct we decided that Kevin would spend sometime investigating this +It will use an lcc to update the data dictionary directly +the bundle uses impl packages from the bundle +I was just poking around the code so I did a really quick java doc pass in a different patch +Moving Sybase issues to the next CP +But it worksBasically remove all after a region is placed +Im leaving the issue open because we still need to update the documentation +Is this ready for commit to +However Its rare that maps and sets play a key role in mapreduce and if you really need them to be sorted on the wire right now you can just use and implementations yourself +There was a message on the mailinglist to ignore the information on the websiteuse cvs module and specify the r option and useinstead directly as a moduleThat is working for me now. +Marking bugs with invalid fixed for release as fixed in. +So there is no need to runs just move up to target +this patch depends on +A change should be done at spec level +A null gc that never collects space is actually perfectly legal under the java VM specification and has actually been used by IBM in the past for certain operational environments +Hi Thanks for the patch applied at r please at r thanks. +This is more conservative here and I think the correct thing to do otherwise we can easily end up with phrases way bigger than the fragment char size +Sorry evidently it took me a bit longer to check my and write my comment than it did for you to find a way around your problem! Im glad you were able to come up with an approach that works for you +In other words while his opinion is respected thats all it is +See Juergens comments in as they may be related to the undesired effect you were seeing +Attached patch may fix it +If youre in the US or Canada you dont want metadata pointing you at European or Asian mirrors +Thanks for the review Cos +was added in version see for details +Was this a stealth add somewhere +I would close this as wont fix +este tipo de issue esta incompleto de momento no se puede hacer nada a abrir la pero su tarea principal esta cerrada lo permitira? +I was going to remove the name altogether +I was just about to report this +Have you tried? Once Id got the test case running it failed with a NPE but after switching the security jars to it runs without any could you make sure you are running against an up to date version and clarify the situation +hashmap based on by apache are in release +Although all of OUR objects are Serializable our production logs are complaining about the not being Serializable +The question is whether this is a big deal +conf dir in the solr home dir would be handy too +IMO is mature enough to use now but also interested in Wicket +I vote that this be implemented as a struts plugin +Here I have attached patch for export search result of Accounting Transaction and Accounting Transaction Entries in the form of CSV and PDF viewThanks Rishi and Sumit for review this patchBrajesh patch for issue +bulk defer of open issues to +Works well but all the new metrics should be defined like that the linestocover metric should be well but youve forgotten the newblockerviolation metric +Fixed. +Reopening this commit has generics violations in +Ok thanks for pointing that out Heinrich +Rebuild the projectIf you open any JSP in JSP Editor then SSE model will be created by the editor and incremental validation will work +I was able build whole metamer outside corporate network with clear orgrichfaces folder in MAVENREPO. +What version of Lucene do you have in your lib directory? Try svn up from the root of Solr trunk +Since the module mantainer is gone and Im one of the developers Ill fix this straight away +Attached file with thread dump +Id actually be in favor of changing the dqp stop call to be shutdownNow instead that should help fix the first two cases +I will have a look and fix the flakiness in another jira +Setting the backlog queue length is simple but different operating systems may its limit on the max length it allows +for I would rather leave alone but I can be convinced m fine with that +attached a patch which does the looks like this modification didnt made it into source +The reasonable explanation for master going last is that it looks at aggregators and decides what happens next +Thanks for the comments Jun +Its a criteria query due to JPA not SELECT +There were some errors to resolve applying the patch to and +BTW Ive reproduced this bug using both the and just encountered the issue +Same error if you deploy all the by default featuresbundles provided out of the box in the Fuse ESB release +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +sigh It does always buffer it and will block the launched process if it fills +Moving once again to +We dont override the init +My opinion is that all of the things you have mentioned can be implemented quite nicely on the current data model with perhaps a small extension here and there and that redoing this model and the resulting work it would cause would not be justified by the things it might make easier. +of records in the trailer along with the checksum +If we use sessionStorage the data will only be saved for a given browser session +Attachement is the patch for class +Right now I dont think we can tackle this problem +test jar file was missing from the earlier patch +this looks good now +Thanks for your contribution ShoLets wrap this up with +This is not a bug +Thanks Kim +Also isnt expressing a long value as L not really typecasting but just how a long value is specifiedOr am I missing somethingThanksLaukik +When Demux see a row with new tag coming it will know that rows with tags which are less than this incoming tag can be processedTaking the example in the description with this fix inputs of JOIN will be ordered by the tag +Ok I applied Stuarts patch +SOA is based on EAP which is why we are now seeing filter did not exist in AS as the diff which David has added shows +Sorry all I can point to at the moment is the sling in bundle in action +Try setting that to the FQDN +Ive fixed it on without noticing there was an open bug +reopen as user run mvn e X and paste the maven output pleaseAnd the best is to provide a sample which reproduce the issue +First crack at this +Latest processing sketch that produces rounded corners +Maybe we should entertain the idea of creating a separate that could contain project files and preferences +Interesting idea +This has been fixed but Im not closing this until I know what release it is going in +We say that tests fail what tests are these? Is there a file path associated with them + reopen so they can be assigned to +Attaching to fix broken links between certain docs and both xml and html conventions files links to new file were added in the file should be visible via the website soon +Steve yep I did push a snapshot +Incorporated comments from Ashish +v addresses the lack of constructor logging and makes the instance singleton private per request +It seems takes care of this +Thanks for pointing me in the right directionMike +The current spam rate is pagesday +Git commit fbaffdeefc. +The proposed in enhancements will be addressed in a release of Struts after +I committed this to Trunk and not branch because the code is now really different +I dont agree with OSX comments while also is one bad piece of we find something better we can always use thattill then this is the fix. +Ill open another jira for the c tests that we should have tests that run against actual java serverResubmitting this patch +The API Doc of says makes no provisions to check whether the same credentials have been tried already +Confirmed works fine on Firefox and Chrome +You just need to do more concurrent testing +I tested with and on OS X +in process +Nothing to ack +Redo code into where possible +It used to be that it would fail in a manner now I get a nice message telling me my permissions are wrong +currently we have three running +I was cognizant of the bug you fixed in sorry I missed it +wsdl and +Committed to branch +Increasing the iterations by times should serve to demonstrate the issueFor the sake of exploration after I patched I tried running the test both with and without the TTL in the update statement an saw no appreciable difference in the number of failures +I would do this using a timer rather than a absolute number of tuples due to the vagaries of how long processing might take +No this should not be closed this is still this need work? Is this something briang needs to look at +Thats why this integration with Spring should work for both cases I beleive +Can you run the last command from your output? In your case Cworkspacescordovaclihelloworldplatformsandroidcordovarun emulator and post your output +All branches and tags arent there at allCould you please kill this git repo and it with stdlayout option? Thanks +Also the following error does not lead to an abortLookup Error SQL Server Database Error String or binary data would be truncated +One thing Im wondering about is if will implement +Committed to trunk and. +It looks like it is resolved and I cannot close ve given you the necessary jira karma +Problem cannot be tested without connecting to real vCloud +This issue has been raised ing issue as part of JIRA cleanup +Im not going to update them all at once but I will update each instance as I come across them +I had a problem similar to the one described by Grzegorz Oldzki +Patch applied with a few modifications handling of src and url attributesis identical and use as suggestedRewriting cocoon as http requests was a workaround to useprobably because did not existed at that time. +fix merged into Mattjust confirmed that the plan is to with release in the next esb release +Session state is logged to console so you can see two sessions being created +Namit certainly but I dont have a set up ready to go for that scale so it will take some timeIn the mean time it would be great to get some feedback on the patch to see if the approach looks reasonable +This is really funnyThis has been done by LUKEI was checking what happens in the index in the meanwhile in LUKE and I didnt realize that this app not only READS the index but also WRITES it eg +The references above provide some links to good approximation algorithms for the distributions +We send the result back before adding them to the results buffer which means that the lob reference may not yet be populated +It will take me sometime to go through the code and review it +pretty straightforward +Please reopen if not the case. +Merged. +SCR has been released +I think most committers dont review patches unless they are marked patch available +Revision Job i will teste again later +I had the same error and Im using preview as ve just uploaded a minor new version of Spring IDEs support +But forgot to update the docsDone +Add a note that iOS compiled apps when deployed on iOS devices automatically disallow the mic permission +But it seems like this fix is not applied to secpolicybuilders +A patch to include our own functions directory as alluded to in the Whirr FAQ +Since the project was merged into the project i added the database qa profile to the projects Furthermore i modified the mysql scripts and the mysql specific sql in repository and task mapping files +which means overcommit is working +We did all the needed steps and sent the notitication email +If there are compelling reasons this should be considered for please provide more details +Simply adding a test for the status of the abort tag here seems to work +think you nailed it +is in +Work in progress as sorting and filtering is needed +Youre saying that globbing foo would return the status of foo now but it doesnt +I dont know what new points I could offer here that werent already covered above +So it is unnecessary to update this if that value is not changing +Pierre if you could verify the fix and then close this issue it would be great +Why would you use the both plugins in a project? In any way if you really need to use both plugins you can use the fully qualified name groupIdplugin +Patch supplied +This is a candidate for SOA is a blocker for SOA +Looks good to me +Finalization turn off flag was changed from off to offPatch contains changes to support new flag +After requesting a vote via the dev mailing list weve determined that were not going to worry about supporting the JDK interfaces +Dhruba What do you know about this? In particular is this related to the infinite loop in that you recently marked as resolvedDoes the lease monitor inner loop leave a faulty record at the top of the sorted leases map that is repeatedly tested +Tests passThanks for the work everyone. +Please attach one +Was attributes were ignoredBecame Border support introduced +Closing all resolved tickets from or older +Hi will somebody try this patch +I just have not gotten around to hacking a codeline or test run to do it since +For cases that uses we could get these attributes and can decide on the that needs to be added to the hfiles while bulk loading +Having an UNBOUNDED resource in Resources can be handy even though having a local copy in should be enough for this JIRA +As mentioned I cant reproduce this with the attached example +Also I can update or open a ticket in POI +All tests passedCommitted fdaaffdacbffdfdafbbaf on masterThanks Adam good work +some point to a very old ASF Wiki +Shame this one got through thanks for testing! +I think you forgot to include the new files and no +Oh but it is committed to trunk as well. +They become these nasty swiss army knives that have to support every which feature and function people have to throw at them +the fix for also address this issue +It will say tmptest already exist +Point taken here +Thank you Jing +Also fixed the code to always clean up snapshot on primary even in case of exception happening in backing up stage. +Accidentally just attached the delta this is the full patch +sorry I meant TASKKILLACTION is sent per task +Isnt this fixed in +We may want to inherit some of this in spec maps too or structure it better +Easy to add our own +Attachment has been added with description More extensive patchThe PDF result after applying the second patch shows the effects the resulting layout will keep words together if they fit and for the other words the characters are distributed nicely over as few lines as possible +I dont see any reason this cant go in to and trunk +I have checked in an initial build for aop which reproduces the jdk build + fixed in r +Sir Yes Sir + +Why does the port need to be changed +So Ive changed the condition such that the check is always skipped when security is disabled regardless of presence of the native libsUpdated the to indicate that the security check is only done when security is enabled +On the other hand the standalone server produces the same loop reported in this JIRA +Cheers either a fully qualified classname of an implementation or a shorthand name for provided implementations is common across all modules in Search +This is arguably the least secure of Derbys authentication schemes +The latest version of dumbster fixed this but apparently wasnt released +Andrew please check that this fully resolves your issue +But since they are not you can have these operations fail in playback due to concurrency issues +Issue reopened to change Release Notes field. +Attached is a patch that adds test cases that test collection of runtime statistics for SELECT UPDATE INSERT DELETE VALUES and CALL when a prepared statement +This patch only fixes the response for the fetch request +Account was created and communicated with Steven personally +All the other metrics seem to work fine with +I have just preferred to keep instead of just +I have also removed the subscribers from the users lists +Care to submit a patch +getEnum and getPattern seem not to trim the string before usage as well +Are we able to leverage any existing implementations ideas +All of the proposed fixes for this issue will create other issues when DTP makes its fix +Mark thanks for your reply on my question +Is it worth adding this complexity to the code for a speed up? Id vote no +Issue closed with release. +Corrupts the meta file header on one replica +I config the ip list on znode just want to support to set the ipList in runtime +Could you clarify what the attemptMatch is actually trying to match against when the exception is thrownIm kind of reluctant to add code which treats an exception as a as Im not sure whether it could occur in other situations and cause a security constraint to be silently bypassed +Thomas +Moving all unassigned bugs out to +We dont branch off until development divergesI pasted the link to the commit previously +Based on Brians info this sounds like a issue Im tentatively closing the JIRA +applied thanks +bq +as available software to install and user doesnt know that he has to uncheck this check box to get really latest version of Maven Integration for WTP +Also well keep UGI as an internal detail so that we do not need to change HDFS for now +did you try this with the final +Closing all tickets +HiFor the stuff you require usersync avz delete pathtomirrorYou may have to add an umask as described in theYou dont have to send us a request if you want to set up a private mirrorof apache just do it +Thanks for the patch Andr! +Yes surely we can simply repackage the +The test failed on a run before the patch and passed with the patch. +Any function can veto the update +Here they are +adding a new parameter to the client will make the API incompatiblePoint takenI am fine with latest patch +Understood and I definitely can see how this adds new capability +And adding filters for other states +on v patch +bq +remove iterations and cutoff params etc +I am perfectly fine with that +I noticed that the binding lives inside the cpp dirI would assume this is temporary and that it will eventually be moved out to a suitable location +Thanks for looking at this Alexey we need jruby to run ruby shell tests we need to be able to run maven build and then use its jars to run jruby tests +oops my bad actually javadoc is missing besides why so many classifiers? can you provide javadoc and sources for them as well +Im mapping this to the generic bugfor now reopen if the problem is not the input file size but aninfinite loop +I committed this in +Please let me know if anything else needs to change +assign to Daisy review +Not sure thats bulletproofThat is why I mentioned that if we have checksum per column it will work as a protection from wrong decompression on the block level and spares us additional read and check isnt it +bq +Guys its my very poor questionId like to this issueHow can i open again +hopefully restores timings so can pass again. +I was just looking for related issues and I came across which talks about immutable Schemas and caching of hash codes +The thing I didnt understand in the examples I looked at was that the request was epsg but it said it only worked right on the d earth +Unfortunately I have very limited experience working with windowing toolkits to be able to help +The patch also look s good for me +Thank youDevin Hi Daisy Could one of your team quickly review this tiny patch and push it to the master! PS No test possible as a parser that causes an exception is required to create a regression test for this +Ive had a need for this in the past as well +Move to which will focus more on supporting the switchyardjBPM execution environments +Same issue with A property needs to be specified in the pom fileThe value of this property should be the location of the root directory of the analysed solution. +I tested and the fix works for me +Revision +If you use that the test will pass +Version I applied to branch +simon i think ive almost got it ready +This allows for example multiple factories to be registered for the same instance of one of which is selected at configuration time +I have a patch ready as per the above discussion +This patch is out of date +A good starting point will be to use Objects and its model +This makes the patch hard to validate when testing with IBM Java +Good to see you guys are already working on this problem +Closing resolved Release +Fixing a couple of minor RST rendering issues on commit +running from command line would be a good idea thanks +But sure would prefer to have this fixed +thanks for the testcommitted in revision Anton please confirm it works +Ive been bitten by this too +Fixing Bug also fixed this behaviour +Issue is fixed Test is added for the issue +I just filed to address that +Its now a year later and this jira is still open +Also date format is not thread safe moved it inside the method just to be safe +I was unable to get Maven to properly deal with an artifact of type as an dependency of a project +So we can close this issue as everything works just fine? +Will commit soon +In other wards if the mandatory sequence has to mediate only the requests then an in mediator is required to put inside the mandatory sequence +Rev +Patch attached +Kurt Im sorry I no longer have the experimental changes that I was working with in +All issues ear marked for have been resolvedTesting branch with a few different for build issues +I am attaching a patch that fixes the tests +Iterating over stored files is not efficient +svn diff java link to ci you are absolutely right +Draft ready for publishing +Will review parser at so this issue can be closed as implementation is finished +There is no big time delay when creating Java EE Web Project via Central. +OK thanks for the review guys! Ill commit soon +Replaced to sparse matrix +Updated patch doesnt change the renew and cancel API in +Committed to master. + change was done against in case it makes sense to link the issues +This means we do not need the batch versions I talk of above +Assigning back to you just so this doesnt get lost +failing has been fixed since then +I will leave it in resolved be targeted for future +Committed the patch using revision I agree with your observation about the error message +Ive just committed this +Will this fix be applied in branch? +I am confused +Accumulo uses zookeeper for this sort of coordination but what about zookeeper itself? If multiple zookeeper processes are started with the same data directory does zookeeper do something to prevent one process from running +Aggregator is scheduled for ease of use in Camel all resolved tickets from or older +changes went in as revision in trunkBackport went as revision in codelinesBackport went as revision in codelinesBackport went as revision in codelinesBackport went as revision in codelinesBackport went as revision in codelines +No is not impacted +I hope you will like this +Add patch that exposes the Clear Queue method and uses it in the performance tests to clean up when an exception occurs +might be used by applications and changing its format may be hard to do compatibly +But made row edits atomic +It would be interesting to use a JDK and see if that resolves the problem as well +This is another problem we have with predicates and nested predicates inparticular +Wont fixed. +ABCD ABCD +There was a lot of work done on this issue recently +Screenshot from WLSPersistence tab looking for a way to set a specific version of and not allow WL to use the default +But when I move all remaining conversion methods from +try extending from my exception to extend doesnt change anything i have the same error +This is a list of the required jar files used to run the example +I meant to get the opinions of the core team about this patch since its a syntactic change and then commit it myself +Also need to revisit change in and make sure it is not caused by this issue +Its applied +Create a common test case for addresses Native resolved issues for already released versions +I tested the patch with my code and it works like a charmAnders +The validation has been tested further validation was necessary besides the one mentioned in the title of this task. +Wont fix this for as its no longer maintained. +I hope that changes brought in revision are fine. +verified code changes in ER source. +The factory could be useful at the receiving end +Bulk move issues to and. +old smooks editor was cancled +Test failure seems to be connected to some sort of configuration issuesReview of the patch will be highly appreciated +Please try the upcoming M release and let me know if this is still a problem +BULK EDIT These issues are open to be picked up +I cant get mvn releaseperform to work even if I do mvn clean beforehand +Thanks for your inputs Andy and MattI considered these points +i find that this one is the one that makes more sense so why not just change how our default one works +committed in +I forgot to change the scripts +Bug has been marked as a duplicate of this bug +Confirmed fixed on win +Duplicate of of +Hi Stack I think this problem is seen regardless of the size of the table but I have a table thats +This implements functionality for PLT +Works in trunk code with the embedded setup +It is against apache trunk as of now +User did not provide any testcase +Checked in to branch with thanks to Ton Swieb for the patch +test reproducible on build +Would be good if you could test this out. +When we discussed this on the list we decided that we should avoid the problem by using only spaces. +It seems somewhat ambiguous as to when you have to update the NOTICE file as well the Kafka guys updated NOTICE too when they included SBT but the LEGAL JIRA indicates that it isnt required +It is merely us taking advantage of rights granted to us by the Software GrantHTH +adds geoxacml to the community +Done in commits +working branch trunk latest commit +Thinking about this as a fix +Committed to trunk. +I chatted with Arun off line a bit about this and he pointed out to me that the are marked as Evolving I should read the patch more closely next time +There cannot be an object without VT it can be always copied IMHO +bulk close of all resolved issues prior to release +Duplicate of create a quickstart against that reproduces the problem and attach it to this ticket +trunk Improved SSL with thanks to Gareth all resolved tickets from or older +I have opened to track this as a bug that we can target for. +wouldnt it be better use adaptable approach from eclipse +Initial patch for review +Could we add the max lag in one place in and? We can pass in the proper metrics name +removed double query parameter was encoded twice +Hi Andrew +Apparently Ivy this by default and the only way to control it is via filter however I cant see a way to check something is optional via the filter and it will require more researchAny for the moment you can work around this by defining an exclude or disabling transitive I understand whats happening now +Thanks for the details +Committed code to trunk and. +the support databases are out of date today +Grid theres very little difference between the two screens +Should have this done for quoted printable encoding +Would this be a better way to deal with balancingfailoverYes that makes sense to me and will be a deterministic solution regardless of the underlying problembq +Because the problem can be reproduced on I do not believe this is a regression introduced by work on the trunk +Thanks Raymie! +Ill let you know when I have something integrated tested and documented +Reporters of these issues please verify the fixes and help close these issue using the latest closing. +Therefore we have customized the JSP page +in settings repositories are defined in profiles but profiles are not supported yet in Maven Ant Tasksfor the moment you have to define your repository in your Ant buildfiles as a for active by default profiles is a first step to full profile support working on it. +This task is now complete. +Ive refined that exception accordingly turning it into a more clearly worded message with a suggested +I dont think it will be very useful having a single maximum amount for all products +The patch to fix the problemOT the code can probably be generalized and moved up one level into +Made a minor change to the patch for not duplication the throws clause +You do need to be logged in to edit pagesAs for being I do not see any information there being +It hasnt caused anything imminent +But we leave it deprecated so people know it will disappear soon +Weve been doing efficient updating of selected mutable fields now for a long time and filed for a patent on the method +While the change was correct there was no need as object equality is suitable in this case +Adapted version tag to be able to remove status of this issue or fixed or no response maybe can be closed +This patch fixes the issue for flat project layoutsAll tests still run successfullyAll I did was to remove a line in the current code +have expplicit strings like casesensitivemultiline Yeah +Dont forget to grant the license to the ASF +for example imagine we have poorly connected data centers and well connected data centers +Postponed toCurrent status is as Falko stated examples are in codebase but not yet executable +Much simpler does the same thing +The parameters are given as a list of separated by the maxLength and minLength specify the max and min allowed number of characters that is allowed in the description attribute has no use other than presenting a way of adding explanatory text to the segments for better truncatableSegments is used to set the truncatable attribute to all segments for a specific namespace when importing. +class encoded for me. +We will not be addressing this piecemeal +Furthermore go to the user forum for questions +We might consider deprecating If the implementation is passed the Configuration cant this property continue to be used by +Please +Ive updated the examples included in CR thanks Rostislav +I need to confirm first which packages depend on setarch to make sure Ill have them installed +One thing that needs to be done here is that the counters needs to be updated with the long values +Good catch +Its not perfect but it seems to work ok +Daniel Harsh Ill work the other way and fix a random other pig jira +Yes my mistake +May be filter console configurations basics on the db connection url +Thanks for documenting your progress so thoroughly and keeping everyone in the loopKeep up the great work Rossrishi +Ive committed this to trunk +in addition the constructor itself is not created that can take the content +Ping! Any news on this Xavier? Could you check my previous comment and let me know what you think? VinceI agree with your suggestion + fixed in r +I just think its important to have in the timeline before we officially declare it as done +Can we have a snapshot branch to speedup the developmentcurrently to test snapshot you need to apply patches from different jiras and after a while they cannot be applied cleanly and it require some time to been able to run itsince neither me or Jesse has rights to commit do you have any idea how to do that? someone volunteer to merge the patches in the branch? any other suggestions +NET is available inIts use is demonstrated by. +Currently the patch only adds the stats in block but I think they should also be added in close +Eugene can you provide a patch +No tests are included since this is a documentation change +Addressed Scotts comments +I finally applied this patch in Rev +Frombq +What happens when you go to connections? I am wondering if it is an efficiency or a latency issue +Moved to unreleased fix issues that arent going to make it into to the grails war script right now doesnt include junit jars or unnecessary dependencies hence why its redundant +I could report it to Hive folks and ask to change idl but it is unlikely tey will be inclined to break their protocol compatibility to satisfy C folksPlus I do not see a reason to throw exception when situation can be perfectly handled +To run test execute +Combine alterrenamecore and alterpartitioncore togetherTest compatibility issues +Added a help and h option for help Modified the usage messages to incorporate Ravis comments +looks +What did you mean by your comment as in should the framework allow for this? I think it should we need to know if the operation succeeded or not regardless of how the tx context is handled. +Fixed in Branch and trunk +Ill try to get around to addressingthis in the next couple of days +This looks like the same problem as before threads three and two are both using the same +When do you experience theThank you for the patch. +to h +Sorry DonaldI have verified the fix under the following environments Apache HTTP Server Apache Geronimo Server that configured the gbean only startup to Tomcat +The overlay mechanism provides an extra specific featureHence as overlay as currently defined and support through is not proposalHere is how the code should look like in my opinionsrcmainjavaorgapachenamingresources copypaste of code gets garbled up will add code is the implementation for the specification resources mechanism +Want or not want +In particular I dont like the dependency of the route planner on the connection manager +i think its readyour ICU is currently really out of date and upgrading it allows us to delete a bunch of custom code +No new javadoc warnings are introduced in the patch +This means that new entries will be placed in the localStorage for every job page I visit and those entires will never be deletedI see two ways to fix this +Well be putting out a next week that includes this fix. +Also cleaned up some unused properties and added isAlive method to the connection object to actually check if the connection is alive to be used by management tools. +The change makes mutex recursive +Not sure whether this is intentional or not +Can you provide a more detailed description of what was changed +Bhavanis fix is checked in +I ran all these tests on WindowsI am not aware that we dont have a javaw process for Unix flavours in the community build +took quite some time sorry +Committed to the trunk +and are NEW classesPlease adviseThanksAfter checking out via SVN I verified everything looked good +I realise that this does not mean all the renderers will override the same methods or if they have copy the code from the super class +Is there a bug opened for that? I verified that the correct website was returned and the returned page did not end with a newline +The source package works fine for meWhich software are you using to unpack the source tarball? I use gnu tar and it has no problems with the file namesAlternatively check that your download is not corrupted +Thanks +Committed. +pushed +Updated the patch fixed whitespace added first round of tests added key member to result rows to be consistent with other viewsFound open issues doesnt accept via POST doesnt send method not allowed for invalid methods I suggest modelling it after alldocs where the HTTP handlers are really short and use a helper function to do the heavy liftingBack to Adam +Committed to the trunk +All things works perfectly. +I have a test that shows it works +IMO Filtering should not automatically add escape characters it takes control away from a user +Uploaded new patches containing small bug fix in era code +added unit test for updating metadata added unit test sample artifacts applied the patchThanks Gwen +How about and in the path? According to the spec they both need to escape right? See for a complete patch +is in the patch want me to clean it and explain what it does in a paragraph? is missing and it is affected by initialize privatization +The change looks good to me +The hudson seems to have problems of occasionally dropping connections +OK +We need to schedule a release for branch for this +Bruno are the integration tests working for you? Ive also tried to run them today using the AMI packaged by Canonical with Ubuntu LTS and I had to restart daemons by hand +Its true +I would think that as long as its not in the SCM its OK +Patch updated +If we talk about watermarks we should talk about watermark annotations +Attached repro script +Patch looks fine to me +Fixing Affects and Fix in Version +fileWe have gotChanges for version Changes for version butChanges since version +My only comment is that the test patch for includes the comment Harmony crashes with Central Directory Entry not found which it doesnt anymoreI dont think I should close this bug because I didnt raise it +looks great didnt find anything Andy didnt catch +This could potentially overwrite locally installed versions erasing that just need to be more careful when downloading metadata files and somehow verify that no errors occurred during transfer before we replace any existing file in the local repository +got it +The pull request adds a test which shows how to accomplish this +pushed +having a private class is much neater now why didnt I think of that? You can close this oneThis issue is now resolved. +Thanks for the fix +wrong patch can I replace it +The soap stuff should be removed so the soap transport can kick in and do the stuff it needs to doFreeman already fixed this. +I first just added this as a class to ease testsing for but it seems that this could be useful for more than just testing +Closing invalid because this was a JVM issue +Thanks a lot Im looking forward to the next releaseAnd by the way how could you tell that the names a fake? +Id rather see a separate slow RPC log implemented has a function like this and it has been helpful for troubleshooting places where high latency spikes happen +The patch looks good +The users still have a start and stop container API like the one that currently exists along with callbacks when the operation completesI would be in favor of starting to use and let it do the work for us +Patch to allow for sets to be used within php to javapythonetc +The patch doesnt contain the binary +It turns out its not the same bug +Issuse is reproducible on latest +I committed a fix for this. +Yes this would be the idea +After changing MAXINLINEGROWTHFACTORPROF to +The PASSWORD type property has been explicitly set as defined in the Sonar properties +Good point Ill open a separate one +Bob added my DSA key and everything started working. +Please try latest Axis and report back with a complete testthanks you for your answer +Darn didnt mean to miss the java one though I knowingly skipped some of the jspxml +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Updated patch the current trunk +I agree that exit codes can be helpful but in the middle of debugging an issue while looking at logs it is often much simpler to have the text to look at +Hi Devinplease review this minor patchThanks SvanteThe patch you submitted is for + +It shouldnt barf on trying to remove a runtime service that was never installed. +Issue above fixed within the SIXPT branch +Ah ok thats a fancy setup you got there +The threads will be still blockedAttaching a patch for the fix +Resolving as fixed +You should be able to build FOP with an locale +I was just writing in html based on the html document available at the project siteWhere can I obtain the xml file? Im still at early stage and it wont be too much of a problemIm going to use the Japanese screenshots and modify the codes a little bit again during the process because some pages didnt look too pretty with Japanese messages +Ive just committed this +If you want Ill be happy to try out any patch files on my system we can do that offline +michael yes the only issue +Patches work better when new files are git added first +In practice I think this means changing the selection mechanism for the loggingsystem to use +The relativePath currently only points to the root of the parent project +Theres actually several things Ive thought would make sense in butI only bothered to file a bug for this because the Iterators already thereseems a shame not to allow one to use it for arbitrary ranges +the patch I submitted adds jsp pages to view the loaded sourcesinkdecorator for a flume master or node +Patch applied +Thanks. +Search behaviourSome of our projects have a lot of dependencies and graphs tend to get wide so Its hard to find some of the websphere jars if you dont remember exact nameI agree +Another more realistic example zoom or position relative to header +I think this fix will impact both projects +Wesley has there been any progress on this issue? Is the Fix version still correct? Thanks +As for the JSP stuff seems to be related to the map preview which were redoing anyways to use openlayers +Updated start page with new +I think at the moment we dont use batch inserts we have no problem here You can close it +a +Screenshots from eclipse while debugging this +Running through Hudson with the hope that the last patch is acceptable +The NPE is caused by the clone method of as you know extends the object which has lots of objectsIf the instance is started the cloned will not start againIt looks like calling the clone method is not a idea way to resolve such issue +With this fix the sync elimiation is done also +bq +Patch for branch attached. +optimized usage of the added dll creation for icuj added a build for added version numbering for pdfbox libs and executables +Do you think that this will always happenThis notification is sent through the leader election port so I guess that it is difficult to make assumptions on when it reaches its destination with respect with when the Follower connects to the Leader +In terms of code what you would try to do is the Java blacktie library srcmainjavaorgjbossnarayanablacktiejatmibrokerxatmimdb srcmainjavaorgjbossnarayanablacktiejatmibrokerxatmi +Sending trunksrcmainjavaorgapacheservicemixhttpendpointsTransmitting file dataCommitted revision . +Ill try use fragments to provide different content for patch over branch fixes the WinIE bugs in Welcome seem to work fine in trunk lets get it on is looking at this +on the patch +Thanks Costin for your ll do so at the end of this update Bernd +applied Kenneys patch. +Committed revision +Resolving as fixed +but met another implies changes in Hibernate organization services +I will have to see if it still applies for +bq +Ive verified it and it appears to be working correctlyThanks! +But synapse has its own non blocking HTTP transport for handing requests +In particular it looks like attribute restriction is stronger than node restriction even though its distance is greater +This is very arcane feature and will not be supported by. +Ralf can you help me understand this better? Granted I did read the issue but still need some help. +Heres an updateI considered making configurable the number of readswrites before reallocing but avoided doing so because above a certain number it didnt seem to have any performance impact +Moved and to vmcore from open at revision . +What is the fix for this issue +Are you going to update the topic to add the security mechanism info +There still was an old version of the fallback. +It uses standard Maven directory structure and a file +Yes I did +Applied to branch an trunk. +It would be better if you uploaded it again so theres a clear legal trailIll delete the other one once youve done that +Functionality already exists. +jta did a final update of the incubator status page on revision and there is no further need to update it. +My intent was to say that I would always useinterpret the feature in this way considering the fact that we process both json and xml +Treat unscheduled issues as run Failures Errors Skipped Time elapsed secRunning run Failures Errors Skipped Time elapsed resolved issues for already released versions +In fact restarting client doesnt help relieve the stuck message +is different maven artifactBTW in previous version of Seam was there with jboss and eclipse signature in parallelAnd I agree with Fernando this definitely shouldnt be marked as Critical +Thank you very much for the patch! Ive verified that it fixes our problem +Ooops my fault +resubmit for for hudson already did that removed the modifier +Problem no present in RF +It is possible that the test itself is flawed +All tests pass +settings for such as database and usernametablespace and path of pgbulkload are made configurable via system property +Thanks for the detailed notes on this issue they helped me find it easily +Todd thanks for the input +Any approx date for +Patch provided +Patch integrated to trunk. + +Within this patch changes to +I havent checked the tutorial yet but did some really minor edits on the Web Services docLooking good thanks for working on these +Hehe i had just downloaded yesterday +I agree the frequency information is very useful but Im not sure the exact frequency number at just is really that useful for spelling correction assuming a normal zipfian distributionusing the frequency as a basic guide typo or common or uncommon etc might be the best use for it +Chris points out that we need to reset state in in the reset call + +We need to take another look this feature might be important to with the Multiplexer or a shared Multiplexer has been deprecated +This feature is available from Kosmos so you can close the issue. +Very simple patch to fix the issue +Its up to dev how to define +Corrected the logic to look for null values and not record a value for the column name in the JSON response +The attached zip contains two patches for the two options I had discussed +However the encodings seem popular +I definitively cant find a bug in Netty As usual you are right +Hence closing this issue. +Thanks guys I feel better + +Closing all resolved tickets from or older +The tradeoff is to select a most frequent happening failure case and try to tolerate it gracefully +Thanks for helping me to understand this +it works in general if you have full admin rights of course I never complained about this +please attach a junit test case +Bulk move fixing the version info for and all affected issues have in all issues not touched in a month to move issues to and +can you try and reproduce with additional logging TRACE level logging forand also check the client logs for any warn level messages +Can you check if its working ok for you now +Magesh Im going to need a lot more for this one +Attaching a new version of the patch which does what you suggestedThe and ran fine with this +Nested get and set getting and setting attributes of private variables and such +Closing with the release. +Vikas +Do get the error when you start httpd or when you startstop a node +bq +Attribute objectClass is deprecatedline Problem +We now have multiple mechanisms to suppress a test failure. +a version that compiles and passes we have a version for branch Ryan? There are a bunch of failures all in +Reopen to correct the affects the affects the affects field +We will probably need to write new ones +and allowing them to go on methods as well as on classes +Watch out the compilation of the proto takes quite some timeAlso can you please paste the error you got with the version? Id like to get it added to the wiki page +It depends on Use case +ah ok thanks for clarifying that Martin +Committed patch to resolve as closed. +We are working on greclipse now +Revision +bq +Phase II Try and wrap with refactored code +Those are configured to use my main SMTP server at +However this issue needs to remain open until we verify that we can build cleanly on Open JDK after the are fixed thereTouches the following filesM javabuildorgapachederbyM +Patch dated Mar looks good to me +Could you check that the maven property is not set +Can someone else please verify that this is fixed? Please report any issues here and we can reopen this bug. +SPR is difficult to implement due to architectural limitations +Not sure what is up with the other problem so im leaving this bug opened +I checked it and the test Jakob provide works correctly +Thank you so much. +No response from reporter +Fixed use of dangling stack pointers +Patch attached +What ought to happen here? Remove the option +Looks like the linking was the slowest step though which is quite more info compiling without g and linking against libs compiled O g that I had previously prepared cuts the binary size in half and somewhat reduces the compile time to about s but appears by no means to be a p lucasxbinxc d timing c commandlineonly examplesConstructsArray examplesxlibharness p g c Ilucasx Ilucasxpgascommonworkinclude Ilucasxgen Itiming ansi pedantic Wall Wextra timing o timing p g c Ilucasx Ilucasxpgascommonworkinclude Ilucasxgen Itiming ansi pedantic Wall Wextra timingharness o timingharness p g timing timingharness Llucasxpgascommonworklib Llucasx lupcrtssockets ldl lm lpthread o l bdlucas None Dec to +there any issue with this patch +Ill work something out +I ask you to add a bulgarian translation for the new poperty on the Services pageYou may resolve the issue fixed when doneThanks alot. +This should be fixed with now holds onto the term and passes it when opening new during reopen +it was actually covered by tests but I never run them on windows and they passed on linux +Posting a patch that outlines the approach RM starts a service that directly uses for leader election +Right before closing the dialog in makes the Finish button enabled +Hmm Im assuming Jens you mean more the necessary revision tree tracking than the actual HTTP calls +Works better that way as it pointed out one of my fields that I was using the wrong type on +Notifications setup. +Its pretty simple patch so Im giving my hereJarcec +Thanks +rev Support added for sure searching on will help maybe for readCan you elaborate on the ditto for testDitto someone complains because tests exists will move to t be supported if is supported +There were no existing tests for this +It would be a nice optimization in fact! Im still experimenting on this code and cleaning adding javadocs Ill patch this and provide a complete patch once Im done ok +The failure can be a little odd as it can result in empty properties at the consumer or marshal errorsThe solution is to copy the message before making modifications +New proxies have been integrated +Heres an actual sample can you please attach your file as well as a sample XML document I could use for unmarshalling +My workaround was using woodstoxLibrary which doesnt have that +I agreed +I have been trying to read ZAB and PAXOS before I make some comments on some of the epoch stuffAt first glance it seems to me that some of these operations need to be atomic +Using Maven its as if you exploded the WAR pulled all its jars into your project and then added your own code too +back to using commons have the patch merged +We thought about client side stat cache too +This is a patch to fix the modulessrcjavaorgapachegeronimodeploymentserviceTransmitting file dataCommitted revision Thanks for finding and fixing this! +See the forum thread for why this property cannot be true on a have this working in deployments that have the value EAP branch r the interleaving prop still gets reset to false by the Profile Service after I set it to true via a component update +Problem starting service portaldeployerAdapter cannot handle itThese errors are similar to what I had seen when I deployed in jboss +If you could please check is excluded on many configurations because of this bug +Committed to trunk +Section specifies the message body optional section says that The presence of a in a request is signaled by the inclusion of a or header field in the requests + add a requirement the CLI needs to support auto complete like xe command line +re referencing beyond we already need to do something different since the threadlocal CC will be lost during serialization +Those would be more bits that TSjenkins hash +Can you please make it to fit the existing structure +Ill commit it soon unless someone objects +Ning Can you please take a look at this? Thanks +I commented out the ldap module and it still gives problem +Recreated the diff and stat files from the src directory +Patch looks fine to me +Here is the slightly revised patch +I would like to add the directory docs into svn for adding more documents in future +Feedback and thoughts in this regard would be appreciated. +Verified the patch again +You have tripped across one of the speedbumps at the intersection of two awkward facts The ANSIISO GRANTREVOKE language does not distinguish the rolename and username spaces +I used it to figure when the web server switch to chunked transfer encoding when the size GB but not GB + lines fixed with this can either you or Yiqiao please update the errata list when we get them back? Thanks +Yes we could have the import list but in my sample it does not guarantee the exact class to be throw +DateToFromCcKeywordsIf I have read the ezmlm correctly these should be the same as the default headers for waiting for digest to arrive to also addedSubjectThese should be ok now let me know if there are any further issues +username is elyograg im a new committer so permission appropriately +is missing a licenseYep +I think we can use the existing to specify this unless you are suggesting to change that +For example orchestration did not have any namespace entry +The patch looks good +I have attached a new patch that incorporates Micheals quasi distributed patch so you dont have to patch twice +Just added some thoughts about the internal changes that will be required to support the described features +So the namenode removes the file before the client reportsits completionLease duration is not configurable so you cannot control thatBut you can retry everything starting from file creation when youreceive that exceptionIs it true that your writes take longer than a minute +Common annotations is another jira issues. +Ah yes NFS would definitely explain this +savingpersisting are deprecated +That could be part of the problem as well +The error occurred when an was then wired to the packages since the delegation was being sent directly to the exporting modules content instead of through its wires +Hibernate debug log when s search DAO method is invoked showing the output from querying Lucene via Hibernate Search to the point where ORM fetches the database records +We then emit the exact positions which have matched which is a type of highlighting I guess +If you finish it before we release beta please correct the version +I also added explicit testing to make sure the instance dir for collections api created cores is correct +Ill have to back that change out and instead do what you suggested in +FYI I did a quick search of the DRDA spec doc and was not able to find neither DRDA nor so my guess is that the naming is indeed internal derby names +Note the patch doesnt appear in Subversion Commits section +The change in looks fine to me +Ill make it use again +Thank you very much! I have been looking at the changes +Attaching and from the failed run +Johnny sure thing for now just update it to the point where it can handle is git I have tested it with a few services in opensuse. +Thanks Elliot +Using the Criteria BuilderFor more information about the Expression Builder see Using the Expression Builder +Im not sure what Hand over to QA means on this particular JIRA but Im clicking it since its the only option to indicate Im done creating the branch +Attached is a patch that adds the balancer smoke test +Thanks Yongqiang +warns the user to provide a and +Fixed in latest CVSthanksdims +bulk close all resolved issues in preparation for and releases. +Please test trunk +Committed alternative to the fix Composite component attribute handling wasnt working properly changing behavior of fix partial state saving suscribe addremove components view listener after relocate components. +Attached a patch to that potentially solves the problem by putting all custom EL resolvers at the front of the chain +How is Solrs support of function queries downgraded if they dont have the Solr package name? I can appreciate the need to be able to make changes easily but with function queries becoming so widely used in Solr Solr users and probably Lucene users I think having a stable interface is a good thing +I consider this a very important improvement to jcrspi +The setup of properties of a custom bean can be created in and then used in +The base of the component will be the component html select multiple +It is likely the same issue +far as I could infer from Wickets getBean does seem to get used there which would mean that alreadyCreates was already populated properly? Enlighten me how you create those bean instances thenIve marked beans as created when using s apply and configureBean in any case +The proposed solution is only about what configuration properties are sent by a client when submitting a job instead sending all just to send what has been explicitly set and the rest to be resolved using the default values set in the JT +Would be great if we get a comparison in your cluster +I guess I just followed that +Intellectually the notion of a mode for DFS client is appealing +even though now we use the one file locally the ide plugins still create a reference to the timestamped I see the issue +Now that work is well underway with AS all previous community releases are +Fixed in only +I committed this to trunk and +Ilya please check gae still exists with Maven resolved +Would it be OK to use Javas instead of Commons Langs? Just seems more standard. +So if both of them are alive the camel router dynamically assigns a host to me +The component have to disable the default event only within the area its attached to +can you send a screenshot +Its solved by which is available in maven RC +Can you test it with a project path that contains spaces please +Thanks for testing Beta +It was exposed as a dep from some our libsThat said since customers are using it and its already shipped I dont have a problem with including this +Ill keep you informed +EJB embedded is no longer supported +The solution you describe seems reasonable +Patch looks good +Resolving it as Cannot Reproduce +This documentation issue is tied to Derby +BrunoWith your scenario I reproduced the issue in French as well +Not sure exactly how it would integrate but sounds like a great ideaI still think theres value in this jira which could be implemented in an hour +logs PIGHOMEbin +Hi BryanAccording to me patch is ready to commitIf Mamta agree with us you can take a look and can commit the patchThanksEranda +Maybe we could exclude this module from graduationIf we can not delete it then as far as I known this file is basically a dump of spatialrefsys table +to keep this option nuked otherwise it limits acceptDocs to liveDocs +A simple test on trunk shows otherwise +The above proposal says that if a datanode in the pipeline dies the client detects this condition but does not request a new datanode as a replacement +I assume this is because web project view opens thing differently? Shouldnt it do the same +Patch applied as shown above +Verified upgrade in ER build +fixes the problem in trunk and + +thanks in advancedhruba +Backported the changes into codeline using revision +Committed to trunk and +IT +Aleksey could you please explain your point +Thanks Steve you beat me on extrapolating See my later comment +Fixed in trunk +We still run into the issue of Too many open files when running +I am going to skip merging this change to since the port is not straightforward +Verifiedy by Leo. +As you said this is triggered by the javadoc plugin so its not a bug in the site plugin or did I misunderstand something +Brian thanks for you the first suggestion the delay is increased to minutes it was minutes does it works for youfor the second one a is better though when you run the test it will tell you what you need I will add the undeployment I really did you can see the After method +Known issue with Java Try adding trueto work well +Thank you +Hi StepanThe fix looks good +Please send a full description of your project on the mailing list coverage value is +done +My IANAL understanding is that source and binary distros had to retain the Oracle license +There are only very limited reasons to use an inkreleaseassert in particular if it looks like we could be returning the wrong content to a user +A lot has happened since the initial March th submission and subsequent updates +Thanks Jinho +Unless there specific reasons Id be on changing thisIm on Teds suggested API change though +Could you please just attach to this ticket the error generated on server side just after having called the cloud page? As a rule Michael would be really better to first start a thread of discussions on the Sonar mailing list before creating a JIRA ticket +Done by Henk +This adds a denial of service vulnerability to Solr +Applied to branch as well +has been raised to solve other assign related issues that comes from SSH and joincluster +I am seeing the same behavior on a complicated object in my the gives no useful information that would even allow me to identify what part of my object is responsible for the infinite recursion +You are right it would make sense then that services can be versioned too +Looks good Luca +You need to make sure that hosts are up before invoking creating volume command +Same hereJBDS do you know if GWT examples were updated recently +Copy these libraries into lib before executing the test +The spec does not show responses in XML for tokens they document Json only so Im not sure if supporting XML is really you for looking into this +for urgent issuerelated to my jobemail eugeneyumobile no RegardsHao +The connectortype settings should have been coming from the embedded and therefore using the jdbc connector jar but instead were using the values defined in the +See attached FileIs this supposed to work with install? If not you should mention it somewhere +I am afraid the configurator is necessary for. +The patch is just fixing the +here is what I will try to do then Implement the configuration you suggested +Attaching patch for first part allowing setting of keyAlias attributeWill work on adding the initParams support and back port to and later next week as Ill be offline for a few days on paternity leave +Attaching a sample test lets see if this helps +Dont use any other Hibernate Core version! clip Thats true +extension point refactored were already updated +I simply cannot afford to wait min every time I need to save a file or do some simple operations +Bulk close resolved issues. +What happens on the assignment is that the is converted into a long +Felipe +Attached for consideration +Resumbmitting the patch as previous patch did not trigger the hudsonThis patch includes the test Test Failures are not related to the changes so just so that build will get the patch addressing Umas commentPlease review the patch and provide your comments +No tests required +Now process is retiredNow testcase does step submits a continue request to process From the debug log I found a exceptionTestcase does steps submit a request to active process created in step +other dependent code are changed to use the new invokers package +The recommended version of the supplied is ignored in this situation +Retargeting to +applied +Can you please try this Johan +Heres one with a simple test case which verifies that the pipeline is actually working +Pages Specification Arithmetic Operators states that the result will be a Long or Double +I see no reason why we cant close this as WONTFIXfor being a premature request that lacks a current use case +FYI I am working on this +This is more like an internal release well release when we need to do so +mkdir on existing dir and file works the same as linux +Click on the link in JIRA the change appears there +Thanks Siddharth! +For revision The directory structure isnt correct the resources folder is missing and is still below the src tree so the plugin doesnt workCan anyone fix this pleaseIt also doesnt show up on forrest Can anyone add it there please? +Im attaching a patch that will create the dll following this convention +review Details of Usage edit the description of the selection attribute edit Skin parameters redefinitions classes to the Definition of Custom Style Classes Description of the Component Details of Usage according to dev project +do instead of the +This prototype works similarly as the compiler switching did before with a couple of changes so that things seem smoother +Yes the scheme should also be defined in the ide plugin or is there a reason not do to it there? I mean it is an extension point which is provided by the Ide Plugin +I already solved this problem by bypassing the VFS directory listing using the underlying itself +Hum thats an interesting ideaI was thinking that if we go beyond just ORM and Infinispan as a source we would need to somehow rehydrate projections and as such this issue would be usefulBut your idea is to consider OGM and its system instead +will use instance of and will use another instance of +Contains incorrect values section +If you cant clear out the schema then you will need to create the VERSION column manually +Now to rebuild the maven artifact +Correctness cannot be guaranteed for custom impls because they might forget to increment itSame with any subclasses of toobq +Do these results actually imply that batching of maps could be beneficial? Thoughts +IMHO these two things dont belong together and it was maybe a mistake from the beginning to let the user give the chrootPath inside of the connectStringAfter this patch Id like to prepare the next three commits Provide a Path class This will also affect the because Id like it to return a Path instance instead of a String Chroot is an attribute of It is possible that only the object knows about the chroot and doesnt need to have any knowledge about it +I found a cause for this problem +Doh youre right the compiler tests dont seem to use the generated classes for their fixtures +Patch adding serialization to +The only thing I dont like with the two phase approach is that it populate the cache with positions +Dupe of dupe of. +Hmmmm +Possibly this is related to search because the problem goes away if I disable search in the problem is which creates a using the default. +where do you accept patches? here or git +may argue that RFC should take precedence in this particular case but Iagree that is most likely to be expected by the users as a defaultHTTP content encodingAnyways the patch is wrong in using default system encoding when no charsetexplicitly set +I dont understandThe method read cannot return null so I dont see how everything can be set to null even if the properties are not found +Cool +Will address Stacks comments and upload a new patch +Simple example project and the deployable to reproduce the Dan mentioned the catalog dtd is not published at the OASIS web site +This is the patch to restore the class +Ah you are right +I wonder if this issue can be resolved +batch transition resolvedfixed bugs to closedfixed +Thanks Todd for the commentsI have fixed the comments please check +The IP address is same to Gateway +Anything that cleans up project poms should be welcomed and this is no exception +Lets close this since is fixed. +Not sure what I was smoking thereBig Hammer latency reset is fine for now but I would like to have more graceful aging out of latencies so if you have latencies like +Try like the logging change but will this break anyone with a cluster spread across domains +you create the mapping file for tackling down this situation +I see no urgency in delivering it if there are some more bug fixes to put into it as we have a workaround for this present bug but in any case it would be the good move to plan a soonCan you coordinate the effort with and Chirstine StefanThanks +Please add something similiar to the remoteTagging flag or determine the SVN version and apply this completely agree to Daniel +the only oddity is the name of the test now tests the highlighterThats actually not true just looked at the test again test is the same as it was it just used to default to off so for the portion that expected it to be off I now explicitly turn it off thats the only change in test +I still didnt understand this till I saw noels comment so too rewrote the subject +Ignore the previous comment +Incorrect operation property from captured in because my perception of Provider services doent seem to be the one intented by +This looks like a duplicate of which never got applied since it lacked a test case and we somehow lost track of it +When you say small performance hit +Will commit +Committing now +ll try and get a review of this done today +I will try upping the level to debug and rerunning +so no generally nothing is being back ported to the branchWell I guess the essence of my question is if it is ok that I keep providing patches relative to lucenesolr? At least for this issue and the one +Nice patch everybody! +I will start with M and test it with all the successive patches until I find the one which breaks the test +Unless someone can give me a use case that absolutely requires receiving chunked attachments Id vote to remove them and use a straight up +I just tested this fix on a single node cluster so as to enabledisable mapred debug script by setting it to null +Raising to blocker because I think this has to be fixed before +Verified closing. +Thanks +I thought that because the issue was old and no comment where added since the creation it could be fixed by the update done on the Ive tried to reproduce the error my I couldnt reproduce if you want to help me to reproduce the error Ill be glad to help you out +OT it would be nice if these were repackaged to go in a package +If I deploy a web application with the security manager enabled and it has no filter then no exception occurs when the app is redeployed +new configs but I dont see a point on doing that because both these configs would generally carry same values and so creating new ones would just add to the whole nest of already existing configs +Ive left some comments on the PR +Just to make this more accurate its not Inject itself that causes problems but rather Inject on a class with explicit scope definition +Its ok Pierreyour contribution is Committed revision ThanksRegards we keep opened +a more generic set of interface and base classes called Generic +this has been fixed in Id recommend to upgrade to the latest version that is at this pointsimon +Test failures are due to not related to this patch +Im interested in the use case though +This would involve more work but it is certainly something to look would need to split the parsing of the from the creation of the lifecycle so that we can derive the dependencies when the esb deployment was created +Ill change other tests to use the new method in a later patchWith the patch I ran the tests with Sun j and j foundation vm +Additional work for files moved to. +I have created for this. +Thank you Torsten +Kick from Admin area also doesnt workSession gets removed from connections list but video window remains and both audio and video gets revision . +next build time is not easy because depends on the previous build +I should be able to fix the bug now +I am tending to handle this by only keeping track of how behaves +Please if you have a test that reproduces this. +Karl devised a way to do what I proposed with some pom magic so I am resolving this as fixed +nscd generally includes great support for controlling the size ttl negative ttl etc for the cache +Still awaiting confirmation of fix from reporter +We should look to create a patched oozie version as or some such thing +There is not much reason to make things a lot more complicated inside wicket core for a tiny benefit in a very rare use case +A new configuration file and a CSV file with the actual data in it +Problem ensuing from source files inconsistency +Go for it. + +since this patch squashes findbugs warnings +I will indeed the reference bitmaps +I committed this patch to and +Xalan never implemented it and will not implement it +dont close the issue these patches only fix formatting but do not fill up the gaps in documentation missing comments are still ample +I tried to shutdown both hqA and hqM and start them again +Could we attach the weights to the queue to minimize the work done in the scheduling loop? In the innermost loop you should probably break the weighted loop if the thread popped is ever null The changes in are not necessary for this patchThanks Jeremy +It is still putting the jar files under lib with the old names +That would make it much easier to track which changes are being included in which releaseAlso please avoid putting multiple independent changes in a single issue or patch +Diff from the root +Another thought is that we are currently not performing a notification when the deployment fails +Hmm +The confusion is this error is inPut file based and we need to supply a error case based inputWe dont need a job to perform a unit test of the fix +Fixed with another issue +Update to use the mojo site and fix mojo checkstyle it appears I have commit access and didnt even know itI have committed this into sandbox +Im probably calling the and methods redundantly +This shows how the bean will be generated after the patch is applied +to myself setting target milestone for release +Many thanks to Milinda for the patch. +Checked the patch it works +Is this going to be changedThanksIll go with JDK enums +This bug is a complete blocker for those that cannot change their database schemas to get rid of composite keys +If you wouldnt mind that would be a great help +attached error logs +Here is the patch which will align the list form properly with pagination option +I applied the patch and committed to trunk. +Sorry I cant send you the JAR since it is part of a licensed works +Thanks Elliott for reviewing it. +that making the private method public makes the deadlock go away is an extremely strange to hear! +Im in the process of some of this in core and will roll this capability into those changes so Im assinging this to myself +Going with what Sean said. +trunk too now +Im using the latest SNAP but I have the same with you using or something +We do not yet have support for the serde that generically serializes and deserializes maps and lists +Ill pull this intoOn a side note I just realized this sink does not have user docs +Added link to forum thread asking for more input on this feature +Confirmed Massimiliano Dess has registered the Community Committer with confirmation number +There are references to a protected port in that document which wont be accessible from the outsideThe live server is fixed Im attaching the new output +I think the NOTICE and LICENSE files should be present in SVN at the of each tree +This appears to be an issue of file corruption +It still requires translation for French and in rev +NOW is the perfect time attempt to fix this since you are so early in the release cycle +Where is the page source heldAlso AFAICT from the Solr docs the search options may depend on the schema where is that defined +might fix this instead +Use the forum next time +Im bringing implementing this forward as several people did mention that we should do this even though the spec isnt finalized there is stll enough to work with +Thanks Mike +Rebased the patch +Nobody is arguing with that but this is not a part of this issueI like Dhrubas proposal better +Thanks Ivan a lot for finding the bug and making patch. +This patch fixes the problem +is used as the basis for the EJB as a part of the refactoring flavia future references the refactoring is on task +rescheduling on hudson +We cant fix broken Android Webviews currently +patch that includes the changes for to and +being a GIT usercontributor since I can tell you that GIT is smart enough to not copy files over +Fixed in HEAD. +and are versions for and trunk respectivelyThe attached script can be used as an audit tool to avoid such problems in future +But since I had to make these changes anyway I thought of submitting them +Be careful with java version you use to launch and loads different locales for some cases +HiIm experiencing this issue +This is a duplicate of Sorry for the inconvenience +New progress I reinstall the management host computer system and then reinstall the management serverAdd instance has no problem now Thanks +And where do you propose adding the backforward buttons +This code has all been updated with the correct highlightingThanks for your hard work Misty +Not really but I think I can do like this I back port it onto on my local checkout test it against some random test data I have eventually provide some WKT from the wild and add unit test based on it if all goes well I can commit or tell you its ok to on the branch as of revision +Committed to andThanks Eric and Devaraj +The trace does indeed look quite similar +Small maven project that shows the problem +Another one would be a superclass method signature that accepts all potentially needed arguments with the superclass variant itself not making full use of it +Thanks for review Jimmy +Cant recreate +I assume you mean either CXF or +while Im pro putting the logging config file in for easy use Im not so sure about wiring it up out of the box +I will probably stop the experimental patch at this point unless someonewants an update as it has grown way too large and start breaking it up into smaller patches for commit now that I feel more confident that the approach yields working code +A related issue Where do the chukwa jars get put +This patch includes the JSON fix patch from Jens as well as other fixes and updates +thanks for the fast in fixing this +Closing as Wont fix because we decided to not support Portlets in Wicket +Please review +All CI passes. +Deferring to the next release since the discussion in the forums didnt come to a issues to next release +Should be fine +For tablet servers we put an ip address in zookeeper and metadata table I think look into why something different is done for the master +Can you provide steps to reproduce +the evils were strong with either choice and we struggled and eventually went with adding abstract method today for the reasons Shai enumerated +These are things that are worth taking on up front because it increases user engagement and adoption and reduces the burden of supporteducation on core committers in the long term +we can open new issues for any differences +how is this different from +I have the same problem +file version Added XML normalization and proper validation of transformed results within the now validates the actual generated against expected ones +Is there any other obvious issue we need to address to scale beyond the current limit +Someone I work with was also having trouble because they could not use Ruby with Puppet any more +Fix attached +Thanks +Run the attached application and go to httplocalhostinheritanceReduxownerObjectshow to see the exception +The problem is that the original Groovy Plugin would automatically add to any Groovy Project regardless of whether the JAR or were provided in the project or a dependent made it impossible to upgrade the version of Groovy for a given project independent of the compiler being used +So perhaps something like format is wrong for Configuration invalid app path +Verified in and +Applying this improvement for the sequence mediator +I think that having an annotation processor is not a bad idea as much of the time the users wont have to worry about itThis being said i think we got something hereThis implementation fit better if say we dont want a dependency on spring for some hypothetical reason and it is a simple but effective way. +that is true all examples are affected +that was embarassing +I think that until youve investigated this project in more detail it wont be easy to tell whether by itself it is meaty enough to occupy you for the entire duration of the program +Changing fix version to +The patch looks good! Even thought I have one request instead of commenting the old code could you remove it +Attached current changelog for reference +but its an easy one +This is a duplicate of the duplicate issue close this one as well. +Hence closing the defect. +This is a corrected patch +I know the limitations you had to work with +Ok proceeding with changing udp stacks to include FAALL instead of manual test with FDALL all passed +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Components chapter was CSS attribute is corrected and replaced with roleXML + +Im not sure if this should be discussed as another JIRA or can it be sorted out while working on this issue itself +I use on to get a singleton object from different threads +Suggest the attached diffAdds Throwable to the thrown by FilePersistentObjectRepository +I can implement a quick hack that would only work for d s etc +I attached the JSP and action class +Yes I agree we should let users choose I was thinking perhaps even for their final output +Uploaded a refresh patch for which passes +its a new feature that changes stuff directly where we are the weakest eg +Fixed by SVN revision +Why do you need to remove a component child in the renderer +AnalysisAs stated in the description this task will only make sure that the latest version of a resource processor is always sent to a target if it currently has artifacts installed that need this processor +Ah good thank you for letting me know +committed someone please let me know how I update my Cassandra with this patch +Feel free to use it as a basis for your work +I did not do any code beautifications yet +The second attachment replaces the first +Unfortunately we dont know if the value that is being matched is going to be null or not +I think thats what its complaining about +Or am I misunderstanding your comment +Commit aadadbccfccefda That commit refers to this issue by mistake +would need their OK before applying +Methods to enabledisable using JMX have already been created +I also moved the copying of the +I dont understand the performance impact but a little concerned that we are penalizing a lot of users that will not see the problem +Ive zipped the files to avoid any corruption of the character it you will find files is the file generate using the default plataform encoding +checked in fix +Thanks release note so as to not confuse folks about whether is done. +Darren could you please have a look at this? just notice that I screwed up on the javadoc params for checkSelected +Done including inserting example of configuration in +I did its linked in already +Change the log level form info to debug +Id do it as a separate patch +JPA annotation based domain class that register annotation based domain applied to core +BTW has always had a dependency on guava not just for Predicate but also for other classes. +We are not able to reproduce the problem and got no furhter feedback +The list in is not complete i think the generate scripts are also affected +Changed the axiom version to +Dont worry about it +was found during this was found during this deleted +But whats the disadvantages? I would recommend you to start this as a +I will mark it as resolved shortlyIm keeping opened just for reporting while I finish the on EAP ERTest passes on +Committed into trunk thanks. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Thanks for getting back to me and feel free to reopen if you see this again + catch +but what do you want to export it to? The only reason you would have those funky characters is because they cant be represented by standard asciiansii. +I have deleted the extra cygwin script remerging the function definition in the main script +But we have one case saying REGIONPLANALREADYINUSE to differnetiate null and the special case +If it is a fairly simple patch I think we could only benefit by having a version of this tool +And yeah i also doubt that anyone was really specifically watching for +Seems to be working with the provided helper classes +Lacking a champion for this i dont think we can plan on this for +I just found during testing performance of bookkeeper +Attached server log. +The delegation should be based on the configuration on the +Note not talking about deploy time +Per Teds request i am attaching a conspectus of our previous discussion of Teds suggested mods to Tropps stochastic svd +Thanks for the good catch +Sending srcmainjavaorgapacheservicemixeipsupportAdding srctestjavaorgapacheservicemixeipTransmitting file dataCommitted revision +One more solution is to creare a DRLVM options documentation page at provide the link as part of help output from X option +Will that help +It was closed when I commited first and after my roll back I forgot to reopen it +Which function are you using to test +I cant say I totally understand the socket issue but since the new approach doesnt use sockets it should be ok +Dominik this should be fixed in Beta build +Consider opening a new issue for follow up work. +Thanks Jing +Looking at the code this would happen if theres an ntfile node at contentpageimage that does not have a jcrcontent child nodeWhats the output of curl httplocalhostPORTcontentpage before you do your POST +fixed in trunkSending javaengineorgapachederbyimplstoreaccessbtreeSending javaengineorgapachederbyimplstoreaccessheapSending javaengineorgapachederbyimplstoreaccessheapTransmitting file dataCommitted revision +work done and completed. +revision on. +So just depends on whether I like the implementation But its not on our roadmap so if you want it in youll have to write it +So adding this fallback will probably break this right +Timeout exceptions on MVCC on trunk +It is the same thins as with the bscanner before the upgrade is finished block information might not be correct from the new software point of viewIf does not scan the directories datanode wont know what to upgrade +I will be travelling in the next couple of days but give me some time after that and I will try an think there is a design that solves all the issues we agree on Explicit methods Generic logger interface Named of them are conflicting so it is just a matter of doing it +instead of Red Hat Middleware LLC +JacopoThanks for the review and here is a replacement patch that addresses your comments strange it wasnt failing completely when testing but could be Derby +Unless theres still interest in this Im going to close this as WONTFIX +These blockers and critical issues are resolved but not verified +Marking Patch Available to have Hive QA run on it +Possible fix in r. +committed to CVS HEADOleg +Glad it helps +removed the old definition for alterpartition from metastore interface and added a new one along with the implementationlot of code is generated so take a look at the specific files listed below to trunk only +Next thing to do bind the javadoc to the right field +Works out of the box for splitteraggregate pairs but open to suggestion for better ways or more scenarios. +So setting fix version to +Should apply cleanly to SVN HEAD. +does less throughput after minutes +Thanks Namit +free to implement your own retry handler +This feature is already implemented in +I think that we should have the security in the framework rather than the scheduler +Maybe it should be unchecked by default? Whats the best way to submit documentation patches +Committed to branch +Id like to resolve this issue as the major funcionality required by this issue is fixed and to open a minor issue with the interop namespace problemWhat do you all think +The JRE in use is Suns versionThe comments on December above indicate that the problem was fixed +The patch looks good to me +The most simple solution was to exclude from scratchpads dependencies and it was exactly what I have done in rCould someone check if trunk works now? Vadim +SVN revision +Needs more testing +Any other things you are planning to do as a part of this jira + +GB blocks have a lot of issues +This patch doesnt contains some caching features +In the end there will be things in Groovy code that cannot be expressed by a Java oriented API +I guess we can close this one now +has been marked as a duplicate of this bug +It might be better to change the instructions to include a specific ant command +Additionally unit test scripts will follow +Yes makes sense +Spring source files that install web service beans +That said synchronizing undoes benefit of using in first place but I dont see an alternative +Perhaps we can just document the switches for turning on that debug info +done in svn +Typo Page Bullet number The term java should be capitalized +I dont think this is just a convention +Yes the thrift jar is a dependency the cql jar is not the patch adds both +And having getterssetters on the so you can plugin which one you want to useAnd let auto discover them in the spring bean registry if any provided so it can be automatic set just by declaring a bean in spring I have to change something in code or the modification will be transparent not matter if deployment is done on osgi server or not +This is a new feature so belongs to rightWith a superficial check these ideas seem good to me +PR attached +I apologize if i seem a bit eager but this is a major issue that I am sure many users would like to see fixed as soon as possible +Oops at least one line in my previous comment is suffering from bad editing +Pull it back in if you think different. +I think this is fixed now from the patch and was fixed in +Committed as svn +Ive removed them and the build works but can you the attached diff? +bq +Looks good Committed revision Committed revision . +It seems better for all involved if its allowed to stay public with suitable disclaimers and less stringent back compat +Rohit Thanks for reporting issue +If you can please submit the documentation you recommend to the wiki. +I would like the opportunity to research a general tab though before this goes into trunk +Closing for Ryan + +We are unlikely to implement this anytime soon +Now I get a shell on my windows xp have an open issue to doc this requirement +Hi DarranAs you mentioned in your description that this ticket is specific to FORM authentication I went ahead and created a ticket specific to supporting fallback to BASIC authentication specifically +the same problem does exist on master +When performing the same transformations with saxon they display correctly +Fixed +patch from committed +Attaching updated patch +Do you know when you got this exception +I just changed the fix version to because the fix could potentially be included in that release if we can get to it. +Franois I do confirm that Sonar will cover this use screenshot of our brainstorming. +By default Derby does not return unused space to the operating system +Essentially take and convert from Ant to quickstart will act as a template for anyone trying to develop their own new +Patch applied for this issue +IMO this patch is slightly better as it minimizes synchronization +in the version is used and a issue for upgrading to is here now if using maven one could override the aspectj version in the plugin config. +Was this bug fixed for? If so please close this +Logged In YES useridYes my workaround with the dummy driver works +rd party jars are copied to cluster at cluster submit and all workflows reference these commons jars +Please document the next element and indicate how this will be used for circular link etc +Id like to close this as invalid since it seems like the two were configured with an overlapping image dir which is not the correct configuration +No worries +Even without restart if for some reason during MR app master shutdown if unregister from RM fails then the app master will continue and delete staging dir etc +Did the required changesSurya KusumakarKirti DhakadHarsha ChadharOnima the required changes in service checkHarsha ChadharSurya KusumakarOnima KasliwalKirti Dhakad + +Reopen because the author attached a quickstart +Closed after release. +doesnt have the builder stuff in +in a slightly different way than the suggested patch but to the same effect +Verified in ER +marking it for if someone can run some performance numbers with this we can include it in release +Hey Grid +Switching all issues that have been in resolved state for more than one month without further comments to closed status +this looks good +Tested with Quokka and Kowari +The file was updated to change the datasource for messaging to otherwise it is the stock file installed by jbossas +Specification tab renamed to Registry and include both svc and meta data under Registry +believe commas are valid characters in binding keys +thx. +The task doesnt appear to have a particular urgency given its only applicable to the very specific situation of X via an intermediate connector +thanks for verifying! +If product is a domain object created with Roo then there should be another interface and its implementation created with name and etc +Committed revision stuff uwe I will fix now. +Security has been added recently and there are many existing features which do not work very well with it +This duplication of the expression implementationhas always bothered me +Bringing into but going to wait on commit +The problem is that i am not able to make the auto login functionality in +Committed. +updated in CSV updated to +collisions were bound to occur +Thinking about this some more my last comment is probably not correct +Patch attached +Tested on +I think this is now fixed but because I dont have the plugin I emulated the problemby changing some constants and then stepping through the code +Failure could be due to disk failure data ode or region server daemon failures +The test queueA deploymentDid you update your? for the tips Ill have a look +I just committed this + Im an idiot + Missed w this fix +classpath and +Reassign to a release that you feel comfortable resolving this issue in or leave as is +The spring mvc actually is not backed by a stackId thence need to keep its archetype GAV coordinates +Hmmm right fixes the memory leak but breaks cumulative stats +So its not looking good right now +Is the permission readA new permission requirement for for sysinfoI had actually always thought this was required but if it is a new requirement for there might be issues especially now that we dump sysinfo to +I did not see efficient way to determine is curson in the annotation but seems this is only one exception case so we can check it by simple string scan and resolve the issue +Resolving +I do agree that the training data file is too large and I will trim it down before resubmittingI will also get a better understanding of both tokenization and feature generation before resubmitting the patchOne thing I am unsure about is how to break up html file for the name finder +I had a quick look and is failing because the apt parser and servlet bundles are not installedGotta run now but that will be easy to fix +BTW It looks to me with Lucenes default norms that there are only about or so document lengths +What attributes on ivybuildnumber did you specify to get a wrong matcher +Drew has indicated that this is an OGNL issue. +The main scenario where it makes sense is when the task has an ability to checkpoint its work before getting preempted +The current parse filter stuff works fine +Marking this issue as invalid +On the other hand it would be certainly better to have it the way the community envisions it before committing but that requires more than comments or answers +Maybe it could be changed to minor as it only affects hibernate users +Fascinating Good to hear that the API even passes this test! Thanks for testing! +Hi Andrei can you turn on debug logging this should show the exception when trying to read the think this is fixed now +There are a few releases after this one so some of the information might have changed from the original because of edits. +Confirmed that the test now passes ok with as used in current builds. + +Improving precision in timeouts will require more advanced scheduling than is currently available +What we need is MESSAGEADDEDDATE which wil be set only once during the time when message is initially added to the group and than remain constant so we can do ORDER BY about instead of setting the UPDATEDATE on all rows to the group update date use the UPDATEDATE of the newest row as the group update date and then the UPDATEDATE would be the natural order and rows would remain in FIFO still doesnt help with databases where you might have more than one row in a millisecond but folks could use a queue channel for testing +Given the state of Glassfish and CDI integration its probably a bug in their integration but the question is if it is a known bug already +And my point is that I believe on consideration of the alternatives it is expected Groovy behaviour and if it wasnt a pure Groovy test case is required to remove other confusing elements from the equation +Well despite my comment on the dev list Id already opened a JIRA so I suppose I should close itCommitted r +Hm +We need this fix very urgently so Im wondering if its possible to get a of the SDK +set fix to Jiang as hes working on cleaning up the detector contender logic. +actually im wrong had to fix my graphs it is actually client connections and it isnt related to this patch as it happens without it +Committed to trunk and pulled this into +Someone else said something about level playing field and sticking to thrift only and this is where I dont entirely agree +But the basic is being done the states traversed are displayed in green and those not traversed are displayed in +Fixed +on. +Please test and close this issue if it works for you +The get value was modifying the incoming principal name to strip space between the DN elements and this broke matching the alias derived from the DN with a keystore entry +JMX is planned for a rewriteoverhaul in Camel . +Is there any development for this issue +Hi CottI want to see how do you use the componentMaybe a simple test case can help me dig the issue +Switching should be a good idea too though Im not sure whether its really needed anymore modern containers can look into JAR filesThird one looks really really strange +Heres a patch against trunk +This is the way to prefer +I already mentioned this above its the race condition since DN is being treated like a singleton +Fixed as of +Solved with parent ticket. +While readability is important Solrs goal is not to make a product that a CS grad can read its too build a blazing fast search server +hopefully +I applied this patch and tested it out +If this is still unresolved I like to work on it +Thanks +I was looking at cached target folders from previous local buildSorry about that +The following fix was made to trunk in revision which fixes this issue using String description Cest une enorme foret enchante o vivais Merlin et toute la clique String description Cest une enorme foret enchantee ou vivais Merlin et toute la cliqueIll go ahead and check this in as a fix +verified and you are ever so right thanks for your patients +Does it just need that change +I dont think it is related path length more likely that some users have different install directory +does a deleteLogs at the end which should fail on Windows if we had a descriptor leak through the normal path of operations. +Added a that adds two tests that try to replicate this error using simpler code than what actually uses but both tests succeed +I just committed this. +Logged In YES useridTrouble uploading file +Please apply the attached patch by doing the following +I just committed this +good idea I will rewrite the style patch +unassigning as I wont get to this any time vote a big on getting this feature ready for prime time +Containers may now implement this in their own fetchers +Another might be to have the server support a few different comparison methods strcmp Avro etc +Reopening for +The test case added in this patch passes with the changes in +VigaSince releasing weve upgraded the AS we embed from to +Here is a patch for reducing redundant voluminous output while retrying to connect +I used svn mv to create the patch actually +OK +Rosta what is the issue here? not working? creating tmp instead of? Or does it have to do something with the auth dir +I made the change in the patch +I shouldve mentioned I havent been following metrics closely so any help from you Luke is greatly appreciated +Good catch +Committed +Is there a WS Jira logged for this +Ive have checked in Safari and theres no issue with Apache Flex +OK should run through tests cleanly +Thanks Nick applied in trunk +added a JIRA to track fixing the tests under cygwin hudson to run with most current trunk +The name of file is ok with the changes in the config of plugin thanksSorry about the dependencies too I just forgot to comment them for this sample you dont need them +Thanks for pointing this out Steven +Sounds like your not opposed to that +I tried to add dependency to comsunxmlbindmain and it works +This needs to be rebased on +Also in case we think this is really needed we need to add forrest docs for this as well. +This does not appear to be enough information to reproduce the problem you are having +Camel uses and uses the non deprecated APICamel is stuck on +just a tricky place is that if failure happens between and +Have updated the website with the link +Fixed with revision +In fact I have put XERCESCROOTDGNUxerces +Instead was bootstrapping a new node in my case +bulk close for the release +Id much rather only require developers have one build system installed +We need installation instructions for both RHDS and Tools in the GSG +The discussion in this issue is way stale none of whats discussion pertains any more +I think the Clay changes has us pointed down the right path tooWell you are doing great +Forgot to add the svn statusM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyiapiservicesloaderM javaengineorgapachederbyiapireferenceM javaengineorgapachederbylocM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmaster revision +All done +HiI have the same problem at my current project +I am running testcases once again from my end and then will submit here +This patch doesnt wire the get api in the producerconsumer +CDI TCK runner was removed in branch +uncommited stuff slipped in one commit +I can compile our project with the java ant task as described in the wiki with +Central uses N updatesites they should be Im talking about installation tests which can be automated +Passes tests locally +The jvm has known bug with interrupts +I seem to get it of the time +No problem +For example try to create a theme for this file +found rsync entryjvanzylwwwreporsyncsshHenri Yandellbayardsince apache projects are starting using I think we should remove this kind rsync entriesthen we can fix problems in I can solve this by fixing the metadata in the Apache rsync repository +Trivial patch attached +The current works ok +Mahout is now quite old +i look deeper +hard coded text in jsp files are extracted to external resource bundles and translation are added +Youre right Strong tests are changed Ive updated links in the description +for the change +Only change needed was setting the protected rootdirectory property +The attached patch fixes the problem by adding the leading zeros to the hex value if the original hex value had less then bytes +Lets see what we can do there +I guess the first thing to check is if there is a resync or a shutdown action. +Great that will probably save me some time working this out +Christian you are welcome to add your test case to source code for example to help ensure the test keep passing and we wont break it in the future +bq +Added via merge of project and product documentation. +Jake can you test with the latest +There is one small problem though that Sreekanth and I noticed while reviewing the tests +The main reason for this is that we dont have another way to make sure that do not run out of memory +Im thinking that if someone passes in a startIndex that is greater than the endIndex that we should throw an +Added integration test check that all nodes join the cluster +Ted +Committed revision the latest patch +Seems like an interesting case +Alg pot +Then I changed again to and the reappears +For the latter operations the documentation must be adapted to also list the possible NOT FOUND status +We might do the resource filtering without all these moves dont you think? Now our demos code looks exactly like archetypes +This information will be very useful when we build monitoring dashboard to show how many edits from each source cluster +Brokers can fail receiving the message for multiple reasons and being unavailable transiently is just one of them +Seems like this script w default property files etc + +this is the patch +patch added perhaps there are more examples required before closing issue +Or extend the Dialect and those lines. +nd version with changes after last build error +whaouh that helpsYour extract use which is obviously wrong for what you are reporting +Yes either it is not supported by the JVM or ASM doesnt produce a correct line number table in that situation +grails and +It would be great if we could support both but I would want us to concentrate on the swim lanes first +It might be good enough +As an optimization +Heres few advantages of over tests parametrization parallel execution of tests dynamic tests grouping and lists combinations inheritance isnt needed +I committed this. +Thanks for the patch David. +We allow the queries flow through during rebuild phase +Its related to font height and it seems to be an issue with that specific font? At least I could reproduce the issue with your style and shapefile but could not s also true that now Im testing on Ubuntu while previously I was on Windows so +Sorry Devaraj Ive just committed you may have to rebase this +bq +there may even be a limit that limits a task duration to some periodIn all these would make message expiry deterministic for a destination +although being integrating in since theearly beginning the SOA strategy does notmention and the immense integrationpotentials of CORBA +Alex thanks for the reportThe transaction has commited the exception is caused after the commit point +Applied patch for provided by GavinThx Gavin +Want to work on that next +To make the change transparent I would move the core of without the to the commonsosgi bundle and let a with extends the from commonsosgi in the in revision committed in revision +There is not going to be any quick silver bullet fix for this issue +BrettPlease see attached sample for the revisionsnew features requested +We are generating maven artifacts only for convenience reasons but I think nobody ever tested this as we are building lucene using ant and only generate pom files and so onI published the versions but as far as I know there were always some warnings in the plugin during generatingTo be sure if it only affects Does it also fail with or or any other version +Yup but if you look at the log message I posted in this jiras description it says it got a times +I will go ahead and commit this now +Of course the wtp web service should disable the web service client generation when start from a java class +On similar note would it be helpful to log return status? One could then easily query for all failed renames or similar operations +But now +OK looks good +no I was just a bit confused +Linking this issue as it should tie in flawlessly with +Can we close this +This is a duplicate of +Given the initial reaction the priority has been dropped to Minor +This is usually due to not being able to download the dependencies since weve not yet done a real release of ESB fabric in a public repo that the can use to download be fixed in a couple of days hopefully +I managed to reproduce this heres the full jstack output +This looks good Brett +The repeater will lookup the corresponding context and execute the child bindings +Commented out code in a few places with +Can you verify that this is also a problem with regular pojo? Id think getObject in nodeModify will run into the same a separate node I am switching to event listening +Editorial pass over content. +Sometimes when there is a convenient element like simple XML entities can be used to achieve this effect but I have an application where I need a number of end states defined and there is no container element that I could put around these so that I could include them as an entity +merged new functionality with current patch +Fabrice Ill havent got the chance to check that yet +How it works under the hood and what the API looks like is where the debate normally focuses +A suspected and excluded B +Let me know if you have some idea +I have tested on my local setup the above tests seem to pass except and is not related to this patch + +I will convert this JIRA to common project +Hi Bob I agree with you on both points. +So all issues not directly assigned to an AS release are being closed +Was the box doing lots of? Was it swapping out +Closing this ticket. +Yes but its just going to say +Fixed as part of +Thanks a lot for the review Todd +It is working fine +Resolved via Stacks plugin +Now test fails on serialized field +This interceptor will be inserted when either Async repl is enabled or Async marshalling is enabled or Lazy deserialization is enabledSince this interceptor will go before we can get rid of the code to check for isMarshallable in constructor and hence we come up with a cleaner solution for handling cases where we need to check whether the stuff is actually marshallable +What to do with them and with corresponding translation flags +Alright will have a new patch soon + +bulk close of all resolved issues. +Natalya do you have any estimations or experimental data on this improvements? Regarding item +Yep having a third column of subprojects definitely makes sense to me too +Do what Mr +Leaving open for +When you post a patch that is ready for Jenkins also hit submit patch +I dont see any list bucketing heretesttable is a normal table not stored asAre you saying this problem exists in hive trunk +Instead no facets are returnedI cant reproduce this +rev +First the encoding is not yet linked to the engine nor to the launching +The setting to use with newer surefire version is reuseForklast time I checked test methods and categories were buggy but may be it has been fixed +Could we have this in +HiThank you for your help +The problem also shows up when Solr is run on Tomcat +Jonathan thanks for the patch +Also a full stacktrace of that Exception if you have it would be appreciated +Now the dialog shows jar files and allows to choose jar entries but code in Email supposes that the attachment path targets a physical file rather than a resource accessible by the class loader +Youre welcome ThomasWell be releasing Groovy tomorrow so youll be able to use that fix soon +Ive also made an attempt at a script that will generate the keystore and truststore with random passwords then print out the properties you need to add to +Maybe of the time fails saying it is unable to locate the hibernate plugin +Thx +Thanks RunpingThanks Doug for the final touches! +there is a problem with the first patch it would not patch correctly on linux +I know this issue is very old but is there any chance you could attach a test case showing the failure youre talking about +But found that with new runs more test failures came up because the order of methods seem to be random with jdk with different runs +bug id +If you renamemove a Writable impl youre in trouble indeed +Do you know if that change will make it still be compatible with the older version of Derby +Thanks Cheolsoo +Please stay tuned +btw +final design sized on a printer file with each jbug shirt as a different layerCheers! +WIP should fix it + in JBDS h Fedora +environment +I was thinking we could also do the same by creating multiple objects too +I couldnt find any documentation of the METHOD object and it is not listed on the ACL web page +Raymond Any chance you can write some unitit tests for this new code? Currently it has none and if you look at the mdep plugin it is currently very thoroughly tested. +I am currently working on this +Changed public method comments to javadocs The reason for renaming to is simply because I assumed it was a typo since the other class is called +Close all resolved issues for Engine release. +created components assigned open issues. +Tests will be rewritten to follow this guideline. +No response assuming ok. +Converted a few more +Fixed +Ah youre right I was being dumbIll make the changes and post a new patch soon +I will try to push a patch up +However the rationale behind the rules is that we prefer to produce and release XML artifacts that have a uniform appearance regardless of the tools used to produce them +failed contrib unit tests is not related to this s not obvious to me that thats the right default but I dont see any harm in it +Im still creating an IT for this will attach it to this jira issue once completed +Please could someone close this issue? Its two months old and I cannot think of any problem the patch could cause. +sounds good to me especially since itll apply to some merics in both RS and masterMaybe just add a comment line to the file that we ship as well with a short description +I think Wicket is great and very much appreciate all of the effort the Wicket team has put in to it +The language codes should be in lower case en fr and ruIs it possible to choose where on the page the language selector is placed in the same way that you can for the versionFor browser support is it the Javascript that you are worried about +Jopr rIs there a repository from which or can be downloaded? not through the RHQ Server UI? Thanks +Hi againIm not having success using Dsyncpublishall + +Thanks for the quick review. +Had a quick minute to review and apply so I went ahead +This logic should just be part of the main listener though +See comments +It works with a changeset in the +HelloWould you please try the patch? Thanks a lotBest regards Richard Patch committed in revision +After using the constant as mentioned by you in the action prefix is working for us. +Hence actually invoking it should be a last resort +Ill take a look +Ha sorry indeed only a warning was committed you may have a look at. +Ill fix this shortly +Archetype parameters are no longer set to the advanced page meaning we cant click on finish early +When I explicitely configure the runtime environment as then the issue is gone +The issue is not resolved until the patches are commited to the Geotools trunk +The updated connector framework for works with all parts of that are supported in s REST API +Thanks Todd +Now I understand the problem a little better +Also happy to make any changes to the content if requested +Yes problem occurs in remote with domain based url +Meta generation is long process and thats why It looks like it is described in issue +Please find the integration test from the attached patch +Sending jaassrcmainjavaorgapacheservicemixkerneljaasmodulesSending jaassrcmainjavaorgapacheservicemixkerneljaasmodulesSending jaassrcmainjavaorgapacheservicemixkerneljaasmodulesosgiSending jaassrcmainjavaorgapacheservicemixkerneljaasmodulesosgiTransmitting file dataCommitted revision . +Was this not fixed in? I dont see it in the change Log +Any chance of a backport +Why not just have a method new? Or get if its supposed to return the same one each timeAnd alsopage The code only ever retrieves it in order to call get why not just have get or newconverterLocator How about just getConverter? +Test done in new +rsync seems to be prohibitedWhen I try to rsync I get an error message telling me rsync not allowed from unknown +This is fixed in you can use any multitermquery in spans +Had to resolve some conflicts so make sure you pull before making any more changesThanks for the patchAaron +s comment +Feel free to reassign this to a specific release but only if you have the time to do test cases mentioned above have now been enhanced to include the following Web Services tests +is ready for review +the resource code will deal with how to delete snapshot from primary storage +Please use constant defined in for and for the default as well +This question should be addressed to IT team because we cannot have build without box to build on +The log url that points to the NM web can be made available as early as the container is launched +I cannot reproduce the issue with the attached application that uses the YUI plugin +It makes the dependencies clearer between and the Storage which is nice when trying to a in isolation +Username hunterhillegas +This is possible using +This bug has been marked as a duplicate of +updated patch with corrected imports. +issues before next release. +Id like to have bundled features when we add this portlet +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Also if you look at section +Datanode should be connected to both Active and Standby +assign to Daisy review +this only occurs with clean +My personal thinking is that the Red is too harsh Is this the only place that the CSS changes show up? Do you have other screenshots of other areas of EclipseJBDS? I have also asked James Cobb to lend a hand here +Thanks for the patch Russ! +UPDATE if I select the canvas after step then go back to the composite the display is correct again +I tried and +See my comment +STS uses Spring IDE internally so both share the same started to track new features in the STS JIRA project to have a single place for new requirements +thanks +Thanks for the patch +I know this might be tough to do but can you please write a unit test that will exhibit the described behavior +We now have cordova plugin rm f to force uninstallation +the other issue of tests with multiple servers is described in which will probably get fixed in the next release +Checked in patch. +Actually direct is the recommended best practice for handling the and of Sessions +Your internal builds must have asserts enabled +A class foo loaded from class loader A and a class foo loaded fromclass loader B are not the same class even if the bytecodes happen to be identicalAttempts to do assignments or casts between them will fail. +Here are patches fro trunk and m not on a computer where I can try this out but this looks simple and shouldnt hold upAll unit tests still pass Espen +Dave can you please get back to me on that subject +We promise to be back up and running soon +I deleted the original attached files because they contain company informations that Im not allowed to disclose +Its an open source project and easily extensible +bulk close of resolved issues in. +The file is a component of! +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Changes made to project docs +adding link to OAK as discussed during the oak meeting +The new driver works in the same way as the old we can skip reset test in the Quokka workflow to save time preliminary test shows that the driver on the side is unreliable when the reset test is for reliability reason we will not change the current implementation of the driver. +the identifier becomes a protobuf or is encryptedHadoop is committed to compatibility which is possible only if the client doesnt assume it can decode a token identifier +as much as I would like to minimize API changes for it seems that this might be worth the break +See to latest trunk fixed test hang +are replaced by versions of these +in what way? should a file uploaded with a timestamp be converted back or do you want the correct snapshot version generated when a file is uploaded with a version of SNAPSHOT +Since were eager to get a fix for this in soon we could address that in a followup JIRA +Ill also work out the order in which the patches will need to go in +Thanks Daryn! +Thanks Marco and Jacopo +Closing it makes ME to look for the dependency in +Could you convert your testcase to a standalone patch file to that the testcase running from the Hibernate Envers suiteThat would certainly help +Proposed solutionModify batch mutate deletes to use the bulk delete method that accepts a list of delete +This can be fixed in timeframe since STI is slowly going away more and more I dont know if it would be reasonable or useful to fix it on branch +New patch +Committed at subversion revision +For use this use +so wouldnt you attach the behavior to the textfield then? since that is what receives the enter key pressed event +It still fails when I ran it on my box without this patch. +Quick start +I dont think we should make anything more flexible than necessary until things like the producer api are sorted outI can wait for the producer API +One exception is the two Vectors store and store in +Needs to be rolled back and a new change if any need to be signed off by at least Brian Fitz and Max +There is another proposal on this topic in. +Patch for will be checked in CP test during EAP ER test cycle. +The basic memory leak problems has been fixed + +I am not missing what you all are saying +Resolving JIRA. +Committed rClaus thank you for the hint +Updated patch agains trunkTests dont pass after applying the found that a similar piece of code already exists in the trunk written by mkleint in revId +Alexey please verify + on the fix for negative severity + +patch pushed as commit eeccfccbaaeabebff +I tried wiring at the end of the jdbc suite but I get the following error under embedded and network Diff file jdbcjdbc Start jdbcjdbc add main Exception in thread main Test Failed +Dont know about +I improved thecomment a bit relative to the first version since thetempate policy file will go into the manual too +minor changes mainly out the attachments for +You could apply this patch to if you like +Please close as you please +xmlen URL invalid +Now I see job has been changed to use the bundled version calling instead of ant +Create still can not saved the form data +Patch file for for removing system. +Suggestions +Tests are running locally on my machine +Marking bugs with invalid fixed for release as fixed in. +Is this still an issue +Ill address on trunk before considering +I the preferences in the Groovy colours dialog so I clicked each colour to pop up the colour picker and just hit okay and it has reset it +Well report Sun on this but any other idea would be greately you for your support +Will obviously require a remangling of upsert but its kinda dirty anyways +The faults were all named the same +Resolving inferred types is not the job of Code Completion Ill make a new issue for that. +This issue was going to be a show stopper and I couldnt be happier that the problem is on my end +Looks fine +Thanks for resolving this one Uma. +Regards patch. +We no longer have a creating flag +I committed a fix. +all issues to the CR +I made a patchThis patch has an additional test code +also varargd +Updated +Applies to these source files assume Paul forgot to close the issue or wanted to wait for Andy to make another comment +I am even getting customers asking for compliance even though I am sure they dont know what it means +adam Run ant and it will create the file in lib +The update counts from each source participated in the update are added and cumulative sum is returned as the final updated row count to the user +That makes sense to me +I have filtered the changes and committed only changes related to the problem described on the JIRA +I think the easiest way would be to make the INSTANCE field in the public this causes no problems in an environment as the whole class is privateWithin a junit test one can simple set an own through the static field without knowing any implemetation details of the adapter manager implementation +This test now covers either allowed user and a dissallowed user I dont think wed get any additional coverage from a test with an allowed user and a disallowed user +A patch would be appreciated +The configuration is kept in a Serializable form its usage is limited to read only operationsI dont consider this as a public API yet need to discuss that with only with Hibernate core core update shouldnt be needed for this one +Thanks for the patchtrunk +I attach a patch file +Might I not want to limit the size of any process as well as the size of the process treeThis is more secure when it works since it keeps me from even creating the large object +XML is the default +Applied the patch and updated the wiki page with thanks to ChrisAs this is a new feature of Camel I didnt merge the patch into Camel branch. +Joaquin has put up a web server so we can see the problemThe module lists justin and gabriel as maintainers perhaps we should remove that from the +Dhruba I dont think our discussion is productive here +It makes the release notes much clearer and historical eploration much easier +There is an assumption Only one JSF component may be present inside a facet tag if more arepresent only the first one is rendered and the other ones areignored +The PDF maintains the sequence of the screenshots properly to illustrate the overall is the generated +I meant the wrapped table is not exposed outside package rather it is used internally or for testing purposes +Patch is attached +The unamanaged currently monitors the AM for exit and can delete the token file after the AM has exited +Help review thanks +For me it gets the job doneJust look for the javassist string in the code and youll see my workaround +Do you need to do the same w the features plugins and site in order to include the cdi hibernate and portlet featuresplugins +DB Schema Handling works as expected +Waiting for the new Jiras +My file however is truncated at bytes occording to ls l. +actually we also do this in Ill factor it out share itbq +However we have to find a solution that solves all three issues +Thanks Bill for reviewing itI committed it to trunk. + +The patch creates a shower of problems in the regression tests +Sometimes it just works +Our policy is that an existing API must continue to exist for one release before the new API is the default and the deprecated are removedDeprecated methods should include an deprecated flag in javadoc and a Deprecated annotation +In addition a migration process must be written because of +Ive committed this to trunk and branch +Also can details of related tasks for any changes to the client be providedlinked please? +Moving it to later since this is not a blocker +In addition to the recipe error These patches fix a grammatical error +Martin did you ever understand what was going on? Its still not fixed in my project either nor +Do you have organize imports on save enabled? Even with that turned on for me I still could not reproduce +Id say keep the header grey as it matches the console +Can you try with version version +If you add mixedtrue to complexType getSequence returns an object +Thanks Mario I was thinking about it over the weekend and I decided the change was not relevant +Thanks for checking this out Kris +Hi LincolnI tried the latest snapshot today and it seems that whatever youve done to fix my problem is goneGot anotherone anyway but Ill discuss this on the list or in the forge chat later onBest regards! Sorry I couldnt reproduce this +Thank you for pointing this +worked like a charmIve just committed this +Fixed a bug where case sensitivity had a mix match +Reassigning it to Gary since the issue is related to realized this was for the outbound case current PR is for the inbound adapters. +Attaching patches +I did not run even for Ive just configured +Can you verify these comments from them +Publishing resolvers should be overwriting but Im not convinced they are Im concerned some resolution is occuring against the repo on an ibiblio basis +Plugged in Greenhouse S information +Although I cant figure out where it isBasically MVEL does not currently supporting instantiating objects in a null path automatically +So the solution I think would be to force portrait mode for again +From the email list Graeme hasnt yet implemented direct PDF generation his current workaround is to use a PDF print generator and print the HTML file to it +That causes a potential dataloss in both approaches +It seems like much of this could be done with a VIEW +we are not going to do this +Please use Ptar to activate Pdist and build tarball +I think since we are saving thiswith function information with name we may not be facing any issues +On trunk I can see the filesrcmainjavaorgapachehadoophdfsserverjournalservice +Basically the same as trunk and golden files +The next apache release of Hive will definitely contain this fix. +Thanks. +. +Still valid for build increasing priority as this is blocking quickstarts installation flow +I committed this. +Otherwise mark them as transient +Support for the or findQualifier when searching with findbusiness findservice findbinding and findtModel using a has been implemented +for the patch +. +The master was performing a +Theres some odd race condition going on somewhere +I can confirm that with PL this problem doesnt existER will contain PL +The only change I would make is in the name of the new property I think a full name would be more readable +Theres a first thing that should be easily the adding of config files to a should just check that the added resource has an +You should have written on the jira directly instead of going through a private channel with Pat +Youre right its still broken +Thank you very much Ive built on Cygwin and this is fixed +Logged In YES useridThe check was added to EJB verifier +Adrien +Are you going to do this? Do you want me to? +I have merged this to. +Currently my first serial port devttyS returns false from eof? when there are bytes to be received and blocks if there are none +so the selector basically means foo bar +that idea may have some merit +Came up empty asking on IRC if anyone testing was seeing this +Sounds reasonable +Tags files updated in trunk branch. +Could you please confirm whether this problem is still applicable with Grails +Dunno if youve fixed this one already rings a +The order of execution is different when we try running from eclipse and when we run mvn tests +It is roughly +Filter logic needs to be revisited though +I ran and found no problems +sstables are compacted producing sstable and sstable has no tombstone +Apologies I meant to post to close for +will commit if tests pass +Please if the problem still occurs. +FYI We likely fixed the equivalent bug in +Committed revision +Should the admin object instances be tied to a specific instance of resource adapter? Id say not since they arent actually hooked up to one in particular they are just javabeans +Minor correction to issue is fixed and released as part of release +patch update +Modifications to code level comments in the patch thx Karthiga +I was using the default option for separate meta log property +What I have in mind is to have a look at the code in the dependencylist mojo to try to reuserecopy it +roflyou call it Wiki Spaces too +WayneThanks for chipping in on you had the Groovy code working and the Gant code does not then Gant has a problem and it needs investigating +When I add this error happen +Will commit if tests pass +Even if it doesnt exist over everything theres no reason to continue existing bad habits +Ok +OkkeThanks for your interest +We have tried to integrate the most recent patch into our install +Is anyone seeing this who can share a project with me that shows the failure? It might require more than one project in the workspace and is due to an interplay between state preserved for each +As already mentioned I use Eclipse +Anything left to do here? The tests are passing +One more thing to consider if we continue to maintain mock is user experience +Thanks for the fix Izaak. +bffa on branchAs far as I tested using a given code the bug was fixed +This looks good +the error is because a schema cannot contain two global components with the same name +Im going to leave this open a couple days for comment I know Eric was not fond of the idea over in the original CQL syntax ticket so I dont want to rush this through while hes traveling +Specifically I am inheriting from an abstract class +Please attach an FO file that demonstrates the problem +And a switch added to shut off the buffering or shut off the closing of the connection +I will send you the location offline +Thats an excellent workaround I confess I feel a little foolish for having missed it. +This is still an issue in version at least for. +That shouldnt be to hard to fix so i will look into it in the next daysThanks for reporting +I forgot to mention that the reason why I asked is because the job spent minutes failing on that linkI was just trying to find a way to clean up this runNo worries +Here is my first pass +My company may have to stop using this feature for this reason +Transmitting file data +Addition of headers values to the Exchange based upon the when the Message is pulled from Rabbit +Thanks Prabhat +Whats your cluster like +I think you should let users decide if a new issue must be created or not +Hi Gavin Ive received a reply from cheapdomains with the auth code and have forwarded it to the Apache address as you instructed +Bulk close after release of +Thanks GavinThats what I thougt also +If you collapse the screenlet and then use one of the pagination controls the screenlet will expand itself +Committed revision Those tests are currently disabled as some of them sometimes hang and other still old resolved issues +Koen found Beta NN which explain all so no need for info now +Attached a prototype of +Bugfix for issue Thawan found with watch resets on read requests in one session racing a write request affecting that watch in another session +Created file is created as remote not local I think that is the problem +Attaching new file with Patch format as requested by Please take a look at new patch file +The ablility to auto join collections works as expected in and and It doesnt work with versions and with +The only trouble is that it forces me to have one map instance holding defaultSorted attribute values for every autosorted dataTable instead of single one holding currently selected sorted columns of all tablesActually my project is in early stages so Im sure that this issue will be fixed before its releaseThank you very much for fast responseRegardsAlexander +We have upgraded to the codebase +I attached it +Committed revision . +Closing issues. +If the information is contained in the descriptor then UIMA may bark if the AE descriptor gets incompatibly with other AE descriptors +i got same info after i install rollerwho can explain this ques +Attaching patch for the think you forgot to attach the patch +Marking this as a blocker for Beta minimally taking the commit to update the error message +Ive checked in a fix that synchronizes puts to the +Merged for in revision . +Chris Love said on twitter he is working on a REST example for that is a good start. +Thanks Tom for reviewing +Also changes of strings +What scope did maven use in this caseI leave this as an excersice to you mostly because I dont like stating the behavior which others might then pick up as being by design while its really an undefined gray zone. +Thanks Jeremy +a patch to enable the web service compile withBecause an bug of uddi itself +Its some cleanup pig sorely needed +So yes if xhtml validator is causing these then disable for M +It appears to be some kind of problem in the preview resource loadingPriority minor seems appropriate +Issue has been resolved and a test has been added + +Andrea wrote giving back reports is not especially amusing +Working on determing the correct fix +LGTM +Pending resolution +back to the forums +Is this one done +See first DEBUG line in the log file attached +Changed to return STATUSOK instead of STATUSNULL +If only Ant tasks were as simple Thank you closing as per previous commentsThanks Oliver +Not that easy I believe does most of the job but its only used for tests not sure it actually covers all the situations a real URL can present + +And I think its deviousdangerous w that lock factory since that lock factory lets you simply erase the lock out from under another open IW +Putting Cygwins path into the PATH environment variable when a program has a Cygwin and an native version +checked it in +Thanks Zheng +Im going to assign this to Dain as he is more familiar with logging and would know if this is a valid thing to do +I will change the sessionTable loop to do try catch one session at a time rather than the entire loop +In the case the number of threads will decrease when users stop submitting new jobs +Why does the the rails server need to be bounced when you kill +meant a noticeSent from my tablet and thus brief +Please close this issue or leave it open if you think there is value to the Cassandra project andor communityBest regardsJoe N. +Closing. +Actually I use View but this problem appear and in this table too +Those go on the annotation on the impl +Committed attached bug and added entry to srcetcCHANLGELOG +Committed v of this patch +The current state is painful with extremely long builds +Please provide the browser and OS you used +added to all in in CRJAVAOPTSXmsm Xmxm XXm XXm true true true +We need to dig a bit deeper and figure out the right magic for popen subprocesses before finalReverted daf in ce + crashing silently crashing testsuites php for trying! Im confused now is this a phpunit bug rather than a zetac bug? +and AFAIK there is no support for HDFS yet +I also want to achieve parallelism by having a thread work on the combiners during the expensive sync operation +ChrisYou got me to thinkingOkay if the CSSHTML guidelines state that tables should not be used for layout then how do we layout the Find Party Parameters portion of that page? I kept the table layout in the patch but i HAD thought about alternatives before submitting the patchUsing nested would be a nightmare and there is no guarantee that they will line up as neatly as a tableI considered using a UL but that just traded one set of elements for another +If theres some coding standard in terms of using text I should be aware of Id love to hear it +Please try the latest snapshot or later whether this is fixed for you +Hes our Web ServicesSOAP You are correct Teiid creates a separate directory in the deploy folder +thanks +Fixed in rev +patch applied please verifythanks +Committed this to +The errors from the child processes for me are Element was inserted second time in +bulk defer of unresolved bugs to +I agree this should be fixed soon in the form widget +Back to inNormalized naming for the static as ENGINE in the tests that use oneAdded back to Long conversion where needed in +We switched away from this some time ago +I did an ant clean before tarring this upcd ant build war +I tried to run ur service and test cases but I could not since I have latest code +Actually this was done for a client request in which they have metadata documents they want to serve up via geoserver itself hence the handling of localhost as a reflective back reference +The samples are now working +I just ran this on a cluster which had a large disabled table and it kind of freaked out made unassigned ZK nodes for every region even though the region should be offline +Works for me and Passos now +Otherwise theres no tests and nobodys using it so I dont know what to tell a community member in the future when it doesnt work for them +I just started evaluating jBPM engine for a process oriented service executed and consumer through best why now is if I was using xml for process definition jBPM bootstrap on startup using already database deployed process definition it run just when configuration is asked this bootstrap take some time to me make sense both bootstrap on startup together many other I will have to create a bootstrap component for some setup and consequently it will bootstrap jBPM asking for +Ill post some details later +Should have made clear Using feature to execute the same job definition with different names is just one way of causing the problem to manifest but is not really key to the problem +With Daniels Classpath patch mechanism we should be able to enable portable native sync quite easily +Our implementation is based on RFC +I like yours too +trace log test for better m not exactly sure what you are telling me but if you are saying that it is failing on the separately packaged resolver in my sandbox then thats possible +I think now that is fixed we should be able to get through these tests without hanging +Attach a patch to address the review comments +sorry should have closed it +thanks! +It is into Rampart PolicyI will provide a patch todayRegardsDobri +Just committed a change that allows retrieval of multi valued properties +rebase strip out a small cleanup patch that got mixed in +Gotcha! Instead of just adding a new stream iText first inserts a new stream at the beginning +the delimiter support was added +If you find further issues please create a new jira and link it to this one. +Id prefer to keep this behaviour rather than fix it +Lets try againAttached is a much simpler implementation that works with Unrealized File objects No recursion +committed +Still have the same problem with +I will upload a zip that worksthankss the zip with a that works fine with latest nightly of branchthanks good that its working now but Now the pom is polluted with lots of dependencies which are dependencies of the plugin not my artifact +See +It would involve some rework no doubt but perhaps that might be appropriate for Groovy at some point +If I understand well this patch will be never added to version +This issue is independent on class unloading design +Please verify them +Heres a proper patch to retry +Support the in JEE +Is the practice to add Javadoc comments for all private methods +The patch looks goodI looked at class and Im wondering whether it should be abstract +Records can add a name to any type with no serialized overhead +I think we have to write a new maven plugin to generate the password +I added the test cases and ran the following command and I noticed the doesnt runant Xlint Xmaxwarns clean jar Im thinking that because doesnt extend also no other classes callSo the question is Is there a design reason why is not running when running unit test? Is it ok if I make run during unit tests +In the jQuery branch we replaced all prototype dojo based JS code with jQuery onceSo you have to make a separate patch for the current trunk and versionSorry for thatCheerswith the jQuery branch being merged into the trunk +This bumps the last digit of the release id after generating the beta candidate +It actually passes on my machine with from svn trunk with code in description +The attached patch passed TCK on derby +Troves licensing wont be changing +Its much more likely to be compaction related than a memory leak +I think it all works nowrev and . +It may be unrelated but do you have a speculative execution turned on +Giving up for tonight +Define method to call +I guess we should add similar parameters to make it configurable as the chmod command is currently in and. +I just want to say we can not use which is key point to async invocation any more before we find a good solution in CXF. +Looks similar but I think its different +Subsequent dont see the messages left on the queues and dont fail. +So currently a fairly narrow use case +Added description of how to change Axis to fix the problem +Since this is really a pool issue it makes more sense to track progress there rather than here. +Closing all resolved issues due to a successful release +while uploading a file Jackrabbit trying to create the temp file +increased the number of threads in AM launch to and then also saw the same bug in the MR app master taskCleaner so increased that to +Closing as invalidThis is distribution specific +A and are mutually exclusive + unit test failure +Adjust formatting +yes I think the should not be disabled the should have access to its by all well known common the could become a property such as showleft right or something +Ill put a follow on for a later milestone +bq +we can just name itI think in the long run we should do that and rename the taskdef and rename the macrodef +Will do Didnt really mean to format stuff I wasnt writing + fixes this issue for hard upgraded databases but not for soft upgraded dbs +I have applied the patch Patch +Updating version fields to reflect state +If we do change the enforcement to the table level though cluster admins can still opt to not give admin rights on the table level thus they can prevent possibly disruptive operations +Should I commit alsoOr can we close this issue +I can reproduce this if I pass start in the parametersSo the underlying cause may be an invalid parameter and we should report better error messages +Im not completely following what you are afterCustom naming strategy is possible to set on the console configuration todayNot following what you are after beyond that +Its in my patch +if debug is enabled we construct the as above and log the failure to allocate block type messages at WARN level if debug is not enabled then we log a message that says something like failure to allocate block +You need to encode the redirect URL correctly on the server side +Complete log trace relating to my PUT request +Most of the time we actually do not want to skip WAL we just want a deferred flushbq +Then maybe lets not things until we have some profiling data showing there is a bottleneck processing those annotations + introduce for feature this will be in karaf release which is coming soonOnce karaf is out Ill patch camel to leverage this and ensure bundles in camel features can start early then normal customer router bundles +If not reopen to say why +Changed Priority to Major and Urgency to Medium +Never mind. +It seems that is used incorrectly in most places +The deprecation cleanup is already done for the following modules remaining deprecated classesmethods in the modules need to wait for Geotools before they can be removed as of our deprecate wait for one release cycle delete policy +Fixed and checked into develop branch Scroller and tests sparkgumbo tests are failing these need to be fixed or the fix reverted +However this is a first step towards fixing the merge problem since we can now almost exclusively operate on bytes rather than ints floats doubles etc +I wonder if we really should be providing these aggregate implementations? The difference between a header and a footer seems to be entirely dependent upon business logic +Verified +alg file which +Its possible to switch back to the previous build behavior by setting the OAKINTEGRATIONTESTING environment variable in your environment and we can always move the TCK back to the default build if we start having trouble with broken builds because of this. +You are going to want to change max also otherwise you hit the same issues in the per user checksI dont really want to just remove max from the webservices as that is incompatible for anyone using version +Be careful about the live extpaths file it will stop you from publishing successfully until you add the configured dirs to the production site. +fix all unit test failures +thus version versionhistory specific object if the request points to such a resourceb instead of add i would create a method that simply builds such an +but I think its worth the trouble since its an easy improvementill commit in a few days if no one objects +moved because of no update +Please check +And that the problem has already been fixed in trunk +I am therefore resolving this issue so that further improvements can now be discussed in +This is an svn diff describing the fix for bug +Repro control +Sorry but the patch did some organization of imports +Wed like to eliminate both of these issues so that were not caught flat footed once someone at decides they no longer want to keep the lights on +changed in is a backwards incompatible change that requires a discussion and vote please revert your changes until we vote on this +Speaking with Amarsri we found the testcase may need fixing +done +Luis Adriano as the queryparser experts maybe you could create a patch for this issue and for? +Do I close the issue or do you? I dont know how +But it may need volunter time for further testing +although it might be a tad trickier to convert to the new API anything with tests is easierin other words i have the existing cjktokenizer converted but whose to say I did it right +Pushing to to to from the release +Checked. +works quiet well +But for now +Embarrasing Marking for master this fails the same in both and mode compiled or interpreted +Dylan Is there something we should be doing differently in to avoid the need for that workaround +Ran a UT against the latest source and saw that the issue exists +Since the problem is reading the key cache deleting it is an easy workaround +Based on Adrians comment I guess we can consider this issue to be closedWe also get the any tests working hence there are no serious problems with the any have tested and seems to me any is working fine +fixed the bugs described above +bulk close of wont fix unreproduced and duplicate issues +is no second call because the nd call returns true I am following up on this in HDFS +Committed. +The bindings directory under qpidcpp is intended to hold bindings to C interfaces +One of the many drawbacks of the current approach is that the data and parts of the transport are in the model in the SOAP and HTTP cases and more of the transport gets pulled up into the model as time passes due to customer requests +The applied patch does not increase the total number of release audit warnings +If I remove the syspropertyset definition from the call to Cocoon in it wll works as beforeI dont yet understand why it happens so Im still investigating +Bringing this up to fix in V +How does apply this patch +Then based on this I will proceed with committing the patch to Lucene +Ill open another ticket for replacing those structs with autogenerated ones +A sample test case was added +This was fixed in +DSStore +I will see about adding the option +Timur trunk does not run on Java so I think you were probably still running +Can you try this out with which is using a new postgis datastorePatches welcomed too. + ssh mvnrsync connect to host port Connection yes + attached +Correction I have this problem with the current SNAPSHOT as well +The only workaround I can think of is injecting the container instance into the object factory and then do a lazy init of the desired properties inside the object factory when its methods are called +I tested the in a local cluster +Those run several hundred a year +I would like one review cycle on it before I check it in +Yes this is no more in the latest build. +This is the class that I created to go along with the +is is We didnt find above two classes in any jar files which are present in lib folder +It must be required for the user to specify this +taking another stab at your approach is betterBut I think there is a problem in serialization order +Results of tests were not as clear as I would like but went with buffered streams with size of . +Fixed +pom file change included +This looks ready toThanks everyone thats fantastic +creates a transaction on a node A of RAC crashes this transaction and then tries to recover it from a node B of RAC oracleraclb +Please have a look at the proposed patch +defer all X issues to +This document also had automatic page numbering in the footer but that doesnt get parsed +Hi Jocheni apologize for being verbally authoritative when i urged everybody to use the +No longer going to do this +The file you attached has two properties one set to and one set to +The initial bug was fixed in so I keep the issue as redesign remainder. +bq +Hi ToddI will post a patch for trunk shortly +Commons Transaction seems much more useful if I can synchronize it with a database transaction I have to admit +XMLs not yet accepted +And its been around since February and its still not fixed +Right now Im leaning towards conflict resolutionBasically all updates are written out and conflict resolution is handled at read time +Resolved thanks for the patch! I added also test case to be sure! +Yeah the patch looks ok +Unless we are interested in pursuing a greater level of compatibility with a specific database implementation this is not necessary +useful commands for testing archive +reschedule to releaseThis has worked for a while since we save the imported schemas at parse time and then return them directly +The configuration file is case sensitive +Why does the RM side fix involve the client modifications +Ive gone ahead with the initial code drop having heard no reservations +I found the issue this one dupes +One possibility is that it could be maintained on the hbase side so that running hbase in a manner similar to the current hbase classpath can return only the deps that would have otherwise been added by addThat sounds plausible +Will commit this in a bit +Thanks +I dont think there is a real use case for multiple suggesters Erick +Only the latest is left +Just used a finalize method to do this check. +Unit + +Fixed a bug that was causing to fail +seems this is not implemented yet isnt it +Its an xml literals like in description of this ticket +This patch replaces the previous one and implements building JARS for all contribution +commited to and I used to think I have to use foreach task to enumerate all the jars +We selected the second option +Thanks. +Closing for now +Yes Ive tried this with FOP as well and it still has the same issue not sure why the last call to the image resource is being left open and not closed +Pel que fa a desconfiana falta de confiana o inseguretat potser cal especificar que s en un mateix. +Is that correct +Patch with source and testsPatch applied +Actually I started working on this yesterday evening and I think Im almost done +Sorry for the delay patch applied at r with a little modifications to make the for windows compile +Comment on the patch +Pushed to in JBDS B +Thanks for the patch Roshan! +I think the API should pass in the max range it supports and then bestFit could just do a min on the two as the last step +Thus after fprems the next fld gets a stackoverflow re optimistic that putting in an ffree to balance the fp stack operations in the implementation of FPREM will take care of the problem +The only possibility to actually find out where the error is located in the jsp is the exception stack trace if line number information is available +Theyre not safe to be shared across didnt build with JDK but I verified that the JDK compiled code successfully ran against JDK wont be supporting JDK compilation only JDK you know when the will be released +I dont know if this fact could help but I realized that this error occurs most often in miniI appreciate any help thanks in advance +Unchecked exceptions are not safe in the current codebase and cause a myriad of problems with object locking and other problems once they occur all of those exceptions should be caught handled and propagated +No need to mark it fixed in. +They have only a release in github + +Ive also included the wrong PDF the one that shows the unaltered IPD where the text flows out of theNote that the coloured backgrounds dont follow the block exactly but this is a separate issue +Thanks for review Jon and Ted. +I have verified trunk on my project +Additionally if this is done then there wont be a need to fix the issue with the Data Source Info Name not being updated when a different connection profile is set +It is assumed that that user has the administrative privileges +Null bytes should be ok to pass through since they wont be confused with the special fieldrecord delimiters +To complete Id have to make investment in translating znode content into a class deprecated by patch over in +In particular some of the new functions from want to live here +If on JDK or earlier the methodsapproximate the JDK behaviorAdrian Does having print also print out all of the nested exceptionstake care of some of your objections? It seems to make the debugging quite abit easier because you can see the whole exception trace all the way back to theroot cause +Is that correctShould this issue be closed invalid and a new one opened for the documentation improvements +Did you perhaps recently change your jdk version and this is what is causing the issue +Pat I think the core issue here is that these methods like using are all +So the problem is as you noticed that there is really no way to start the server and having it load a broken sstable which means there is no way to run scrub on it +what are you looking for beyond the fix version being set to? I believe Snjezana is working on this but ill let her say if there anything new on it +It probably needs a separate jiraCreated a jira at +Thanks Kevin for your attention to detailPatch applied to BEANS module at repo revision rPlease verify it was applied as you expected. +It changes the lists of actions so that and cant run in parallel with Region servers being killed +xml files in the docuserGuide folder +Fixed +Cloning the element is the solution I had in mindWhat version of ant are you using? ant test works fine for me withColm +This is the XSD I used +Will commit the path for in due time +This is probably more correct but what is the use case? tracks and verifies block checksums for lzop streams but without being initialized by enums to which only has access it cannot be distinguished from +I was going to modify the default to check if the was an instance of and if so get streams mark the beginning of the stream and then wrap that in a custom Binary implementation +Would it not apply to any OSX installers regardless of the OS level on which its being installed +It will still be a little bit while I dig myself out of my hole and begin directly committing fixes to the repoUntil then Ive attached a patch here that should work on codebases as old as June that provides support for larger disks via the BLKGETSIZE ioctl on linux +In theory if UDF still use deprecated s they might see the same issue +But we obviously need to fix it +Patch to call init in the helper is still broken +The patch look fine for me +I can not post the stack traces I saw because I was piping the output to less and the tserver crashed +Converted to groovy into trunk rev +for review +All of my tests were successful with patch version +It works fine with header. +A possible solution could be executing the problematic code in a privileged block this instead of giving access permission to the system graph. +Your patch effectively changes the semantics of addFilter to that of add in the current trunkaddFilter only filter user facing urls which are defined in +If you stop this consumer then the other one takes over +Its in the trunk as well with the otherFell free to work on the doc side of things +Thanks for the patch Gary +pdb and +This is an idea that I first mentioned in +Thanks Dhruba! +There is no update on these items and are in open state if they are not ready for please move them out to Future +In addition to this I am making the stats for all the jobs available +I came across this when an deployed as part of the EAR created an iBATIS from a transaction manager configured to use JTA with +Secondly we wrote the axis integration in such a way so that it could nicely with a spring web application context loaded by the context listener much in the same way that Springs own does and in the same way that the integration works + +Patch file is significantly out of sync with project +add new folderSeems to be a minor hiccup but still needs to be fixed +Just take a look at the way the information is passed around to see what I Have what build? the regular expression parsing? Im not confortable with the idea of introducing it in the current class but feel free to propose some examplesimplementations +I thought you meant the freeMemory method of RuntimeYes freeMemory which you mean does be a perf hot spot because most might happen in this method +Maybe either Nitay or Brian can take a look +Since we dont want to evaluate that logic again on destruction the bean factory should keep track of the dependencies arising at creation time +We may need to release CXF which will have some Fix of the SF related issue will head to release it at the end of this week +Andy My understanding is that this happened on a local serverAs Recovery retried even though the server was already stopped +Do you think database upgrade will be called firstAccording to this patch it seems only is changed +adding to trunk +I built and verified that startup progress metrics only showed up in namenode and not datanodebq +This is excelent newsIs there any chance to take advantage of fixed Jelly in Maven stream +Ive committed this to trunk and +Hi MichaelCurrent version does not contain that bug +Nils thx for the explanationIf we replaceString includeParams nullto String includeParams GETwould it affect the portal portion of webwork? I think that would fix this issue +Patch applied +It seems like you have to hack a solution to detect the state of the last buttonlink clicked +this problem was fixed in +Please make updated patches available in a separate jira when your testing completes +Gonna commit this to trunk but hold off on version until a resolution to RC timeframe is reached +I remember youve submitted that one +but seems negligible for a few really long lived objects +This breaks the site generation when adding Checkstyle as a report +Content assist and code select speed would be positively affected by this change particularly for large files +Thanks for your reviewIve just committed +See for further details. +Additionally this would certainly be of secondary importance to getting our first release out of the door +Mauro can you please test with the recent build and comment if you still have this issue +Thanks +First draft of the Thrift with SASL protocol specification +This is likely out of date +I think were done then +This patch clears the stage during teardown +Why are you reporting this as a bug? Can you describe how this affects your application? What exactly do you mean by uses a bad +But dont you think it seems too simpleI have other XML constructs in this XML feed that match the structure in addition to title there are rights subtitle and summary +Was this finished? +Thanks everybody. +easy at with. +What is unreliable? The original use case for isset was to determine which fields were present after reading a structure off the wire +I would recommend rebasing this against trunk +Workaround is to use a parameter instead of a literal +Correct +it looks like s tests are timing outfailing with this patch perhaps due to the genstamp issue mentioned above +Closing resolved issues. +fixed in +Seeing and its parent Im more inclined to close this as a dupe commit he patch as and then use as the parent issue +Tested with by using maven it was able to resolve dependency properlyThere might be a hidden problem that Jersey version are different between and Chukwa which can cause conflict in the class path in IBM jdkWhat version and vendor of Java was used for this to occur +Any chance of getting this patch committed? Seems like a sane workaround for malformed dependencies. +yes please apologise for the wrong info unfortunately it does not provide help for the target JVM check so the latest proposed patch is still the better solution I have have a look at leatest patch I propose the thread dump service as a Felix instance would it make sense +I have been distracted by a number of other things but I should get back to is shortly +Especially once we introduce features like etc which really need a browser tab separation to work +Thanks Mahadev! +for the patch +Going behind the API would be the faster way to load hbase +Any reason this is not on by default? Does it alter results in any meaningful way? +Is there a way to turn this off for compile and just run checkstyle during a specific run? This would mean that it would be up to the submitted and committer to verify correctness exactly as is required currently with rat +support should be done by now +Merged into +Did you do that intentionally or by mistakeThanks revision please check and discuss improvements if needed +If so the implementation was always a hack in MR +I just asked the incubator guys regarding publishing the informations regarding the DOAP file infos or is listed there this is a subproject of myfacesalso i added the doap plugin to our reporting but is down currently +Im looking into it +A workaround can be using sstablejson and jsonsstable based data porting +Adding allocationSize to your annotations is fine if you have one or two but a maintenance pain if you have manyIf the hilo generator is going to be forced on developers it would be nice to at least be able to alter the allocationSize in the config so that theres only ONE place to maintain this value +Argh the patch has gone stale since went in +Pom version and references to artifacts updated +I have committed this +Thanks +We need to put the security in place before doing this. +tar file needed for someone on a unix with rpmbuild installed to create their own conf +Hopefully I get this done this week since I will be on vacation next week +Thanks for the patch N. +Thanks for the patch NikolayApplied to REGEX module at repo revision rRichard please check that this fully resolves your problem + is trying to compile all list of known issues +Please use one implemtation as a template to implement your own Microsoft Visual C Toolkit +I really like the gradient style +Hi SimonThanks for your help! A patch that corrects these hunk failures has beensubmitted asKris +It looks fine conceptually but still could use some more work on the code level +May I can test and then commit +Got some time to look into extracting the common code between actions +This patch incurs an unapproved license errorIt seems that the license is added to and files or these files are excluded from the license checking +For regionservers with large heaps GB it might not make sense to have a restriction where the memstore size has to be a certain percentage of the heap size +From the header block of QUESTION should we do the same for get? TODO we should enable a way to disable conversation propagation by URL author Dan Allen +Only for those who actually dont remember to close them +Aha! Thanks I missed that attribute +The problem will either go away by itself or Ill get it fixedCould your patch wait a few days please? Thanks +this is caused by a typo in your junit +patch applied to and closing all issues that have been in resolved state for months or more without any feedback or update +Added a vote to this issue +As a temp fix Id like to remove validation from class Text constructors +Mark you might want to review from a performance perspective and run whatever testing you ran previously +In any event your points are well taken and I will make the changes you recommendThanks and RegardsAshwin +Ive committed this +Please provide us info about Improve xmlxmljavaxml template file contentsdocument editing GUI for testcases creation +Bug has been marked as a duplicate of this bug +Will open individual new reports +Patch with review comments from Sharad incorporated +This is a bug now but I dont know how to change the issue type +Have you tried it out on a cluster? See any performance differenceWill get back to you on this after testing +Which I think is a very nice side effect of your work with us. +Would you please take a try? Thanks a lot +Unfortunatelytheres not a file in my +Now we have a place to put Maven task tests +Thanks Shakeh! +Also avoid locking the object itself since it will block internal async dispatcher +Heres a more polished version + the patch looks good to me +Totally! And guess what? You have been elected to do itP That way you will have full permission and I can investigate the issue you were having further +Fix also in release in release +As was pointed out in the email threads related to this the submitters test cases are incorrectThis bug shoul dbe closed as invalid +This is no longer an issue since only one console is notes verified they dont mention this jira. +Since the bug reporting system is still out of commission I have been sending these on to Hiram as I find them +And kind of on that topic are my tests allowed to for example shut down hdfs and restart it +This issue is the same nature as + +ChunhuiThanks a lot +If no one objects I will close it in a few days +Producer performance does not measure latency directly but we expose a jmx bean that measures latency and request rate +Closing as we have the ACQ and the patch has been applied. +If anyone has a need to use core in a Java environment they are supposed to compile the sources themselves +Closed per user comments. +Not sure what broke it +For what is the use case for that? I wonder if store array is better than string cuz if arg contains space string cannot capture. +Patch file attached. +I couldnt even repair that with dosunix program +Attaching patch to fix +Yea in this case I agree get is better +Having a common set of operators will be a huge win +Will advise +Running in fork mode seems to fix this issue however there are occasionally concurrentmodificationexceptions in +In the attached diff file I developed a relatively clean solution similar to what is done elsewhere in Axis +Now that I wrote the tests I realise that maybe the behaviour of the is incorrect as it labels a of a Facet that does not exist in the taxonomyThe behaviour of the and the Standard differFor my use case it is very helpful if all the return a with the same label as the request but I can imagine that this is not desired +Yes I understand that part I mean what put that SNAPSHOT build there +I think we need to investigate what happened hereIn its +How can we delete the whole row if they are the last ones? We cannot because it has a race between check and delete +LG +ear +and since there is already code for that it is not the place of problemIn fact we have a Runnable here so groovy first creates and instance of this class and then executed the run method + +I gave it a go +Thanks a lot to DarynWill merge it to as well +Right now it is not clear if it is part of xerces or xalan +Any other information requireCheers has been added with description Sample FO file to generate PDF has been added with description Sample FO file to generate has been added with description Result PDFSorry I have upload wrong files before +Still pretty confused +Create a controller eg web mvc controller class +Hi friendsI have this same problem with version JBDS the for java types dontworkInterestingly the Type Proposal is enable by default +I committed an initial version based on the CODI +Thus some of the changes in your patch no longer apply +any hints are welcome of course +Ok I am on Maven so I had to upgrade the plugin version +Moving to +It has to be in maven core somewhere +Is this working for you +Yes it is fixed in. +Ive attached the console output from the failure +Modified file svn commit Sending trunksrctoolsAlso merged to. +So that sucks +Tamas Monos looks like I will have to downgrade too +I think the rest of the work is not related directly to bootstrap integration +To my mind its the read rather than the lock thats the problem +The only file which was modified was the for the service control drts all others are new addtions +It wont matter where a table is split +I thought we want in expose the Assembly objects created by mdo to plugins configuration +As far as is concerned this seems like a more targeted fix as Peter mentioned +As I mentioned before this is very necessary for alibrary like Castor +Hi Hairong I did not get what change your are suggesting here +Addressed the comment and committed to. +I havent tested it at all +Waiting for the changes to the docs to resolve this issue +Publishing the first should fix this +Thank you for the detailed documentation and the Spring Integration project. +I think should be something like and include the final resource usage stats +I fixed the related issue ok but this one I still cant see whats up +on inclusion from me simply because I dont generally see the point in adding inverses of existing methods its just API clutter +There is a artifact when the server is built with and a when the server is built with jdk +Looks like the debug build problem is in incorrect timeout settings Ivan reported it could Eclipse dependent +The not seems to be very general better move to piggybank +Cursed Internets! +includes a fix for those a couple extra tests the synthetic test data and a few less tabs +Niklas Any preference on the package name or the existing packages that the classes should go into +That should help a bitIs the mocking necessary? Adding the a mocking package as a test dependency is fine but perhaps not yet needed +Having read this issue I can see that whilst the committed fix solves the concerns of this issue it does not fix the so Im going to unduplicate the two issues. +i was not aware we had so many guidelines very interesting to see +Addressed Iyappans comments +Does it need to be +For example polymorhpic methods should be added that override the base classes methods to say something is not supported or violates the new logic in the generated descriptors should be extended such that only the proper are returned so that any necessary validation can occur for the restriction of the m assigning this to Arnaud as he is the author of the original restriction code + +Attached is an updated patch for this issue +Hm maybe not an issue the new epoch tracking tracking code seemed to handle this ok +Are you missing any specific information from the original stacktraces just a good practice to have traces up to the place that initially caused exception. +Ive got the code you sent me and I will go through it soonMark That is an issue that should probably be cleaned up +Shut down one node and insert some rows say to the column family with CL +for failed it seems not a patch problem +One of the new fields in could be attempt status which can be used to determine what the status the attempt was +Since it is a existing problem this need not be a blocker for +I had not attached the results forall the tests +If it was fixed can we close this ticket +If this change requires a change to the javadocs its basically changing behaviorShould we maybe reprecate the render method and implement another one alongside it with these new semantics? I cant really see how the existing semantics make any sense in the first place but this is supposed to be a stable API. +Due to a small typo +Thanks Cameron. +Did you guys change the required URL parameters structure? I am hitting the following URL httplocalhostsolrselectqpizzaspellchecktruedefault and I am getting a nullpointer exception +Will try to look further +It just contains changes +Did you get a thread dump to see what this infinite loop is really about? Its really odd that it works with two nodes but not three +Patch details The idea is to stream small chunks in bags to the Initial version of the algebraic function in the combiner +have a good weekend +Roland different causes and may need different handlingOf you course you can handle it different also when in the hierarchy if youwant and can handle it at allIt is common practice in the Java API to have any IO related exception underexamplesThese have nothing to do with any IO problems as you mentionedAdding additional hierarchies must be avoided because it leads to a rethrow pollutionof the application codecheersjens +Thanks. +Ok Ill add it to the heartbeat as an option +The conflictingpatch contained a new test which exposed a difference in SUR behaviorin the client relative to embedded +Most of the code has been tested at scale +I verified this is fixed as of JBAS r +Limiting issues for b at this point to anything in trunk for as fixed for b +I think a focus on the constant score rewrite makes more sense and constant score rewrite will be available for all of the queries in the next release +Hudson +Any issue found where we do not align w spec is a bad bad bug and needs fixing +I see no reason why not to commit it now as well +Put something into README about +Most of the unit tests passed and I have verified this patch with my loader which had the mentioned issue +complete +And got the experiment code up and running +first done with r +I dont have any plans to take the beforeafter demos down +Thanks Zo for reporting and Tommaso for fixing +Thanks for putting me on the right track Andrea +The funny thing is that every public method that interacts with internals is guarded by a lock +Whats up with our JVM scheduler such that a thread can get no time for periods that are X the sleep time +here is the correct putting back the from this fixed another test failure +Could you give us some more information so we can try to get a reproducible case for this issue? Can you provide the schema or database and the sql statement you are trying to execute +As I recall RPC calls used to have a timeout that combined with the retry proxy would spiral load on servers +Can you please test against the current RC build +Committed as discussed. +With fixed the commit command really is retried indefinitely so this issue is also fixed. +Wont fix? Right +It would be useful to know which of these is happeningAlso if you can provide any log output at the time when a replication died that might be helpful in diagnosing the issue as well +Well I will take a look at this +This latest version just introduces more comments and more methods have been marked protectedRyan I guess that if you dont like this version you should just commit yours which is in any case a step forward from the current trunk + Bug has been marked as a duplicate of this bug +Updated the source tab to be a source viewer and made it for now +Thats why I have add a bug as a remainder for Axis to make this cpptasks available +Since STMTAUDITOPTIONMAP is a system table it shouldnt be considered during exportPlease be sure to have the following row into your SYNCOPE +Thanks Regis +I uploaded the wrong patch before here is a complete https implementation +Renaming JIRA to signal nature of changes +Im not sure wether on whether we should combine the two languages into a single or keep them separate though +I am attaching the WSDL files used to generate the skeleton C code that contained the reported errors +If set to file it implies deployed on all nodes +Added release note text +Is this Issue a Blocker or Critical priority for EAP GA? If yes +Please close this bug if satisfied +checkout generated eclipse files + for structured xml is updated. +This should help further debug potential connection issues +Just one point is that I might need the job id along with the task id +Set as a response of a inonCommandIt seems the response inside this last is never getting cleaned up +i tested it again against trunk and it worked +Applying. +Here is the patch +Changed the length of the textvalue to +The issue was not with the sort order but with the parsing of s with asterisks In the case where a ended with an asterisk didnt parse well the prefix and instead of removing only the asterisk it also removed the last letter. +Verified in +Ran a check with CD for each TLD this is very slowThis showed that FTP sites only exhibit problems with the subversion directory other directories seem to be OKIts not clear why HTTP mirrors have problems with other directories +Heres a slightly simpler version +Great discussion and some fundamental requirements + LGTM thanks Chris +Bulk close of old resolved issues. +Pig can convert a byte array to an integer and then to a string as long as the byte array is convertible to an integer +Resolved for a while +I just tried the patch on top of trunk and it still applies and works +push to +Additionally over time we can the mammoth to look differently as the team likesSo the mammoth is great as a logo +From where is it picking that version number ? +Fixed with revision +Fixed in current solution to this makes a copy of the entire receivers list for every message that is sent through the is likely to have an adverse effect on performance especially when there are many better solution should be found to the race condition that does not involve cloning the list +The given patch for solves the bug for me +This does not make Struts configuration easier if you ask me Since you do not insist on implementing this I would rather not +Dependency on required was fixed in beta +Oops +In both tests the mean values produced are within EPSILON of each other but the std values vary by +Proposed patch to resolve the issue described above problem described was fixed by good to me. +again +Otherwise if the restarts between submit retries then the same could end up being used by a different client and a submission would be ignoredIn the meantime wed probably get of the benefit of this patch by simply not retrying submits +Did you not see the comment from GavinThe excalibur unix group is still needed to allow the site and dist headers to be tidied +If we put in Ill make sure this addition gets in as well +This patch reduces the visibility of to package visibilityTouches the following fileM javaengineorgapachederbyimplsqlcompile +Is the existing siteYou will either need to migrate to the CMS or at least svnpubsub both ofwhich require an svn url for us to base that on +The log can be found in logs. +It will be fixed with next release. +Why move this case to is it not a valid caseIm also not following your second line of comment can I use it in Java or can not +Template Guide will also needs to be updated +Im speculating but we recently committed a change for +This occurred withDo you have a sample reproducing it please +committed components that needs to implement changes. +can you give me an example maybe if i understand the problem ill can try to fix it +Closing this as duplicate of +I just remembered one of the reasons why i didnt do this the last time i looked at it i dont think has any good unit tests in the Solr code base there might be some tests that use the to trigger function queries but they arent really portable +These are removed in this patch +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +exception is now provided +It does work for a simple test case but still confused +It is currently invoked only in +The new contexts are and +I think the fix for should take care of this bug +Ive moved the priority on this one since jstl is not a jar we are currently using inside Spring +I added entries for Lucene and Solr used for the default and added param for phraseLimit in the new constructor javadoc +Can you attached your generated class +Im trying to improve it using Axiom but there are few issues to figure out +This seems to be a problem with the Maven Eclipse plugin +Actually committed this change quite a while ago not sure of the SVN + +I couldnt reproduce the error however I added an extra check during the request factory instantiation so that it will return the default implementation if a null property is found +On some workings sets having this configured to might be better +This isnt always possible though so I think it should still be fixed in AMQ +Yeah exactly +Stopped using DMLC and wrote my own using the jms message listener interfaceprobably a bug in springs is it closed? Not A Problem +I updated my ee test using this storageTests done Ran the new ee test case on cluster +See related issue +Test case attached please follow instructions in enclosed +Thanks Adriana slightly different patch is in rev +Alright I think this patch looks good +Small correction the option was only introduced in which I just released a few minutes ago +A very preliminary version of a proposal that supports archival and migration of files across clusters +If the components plugins dont depend on anything that changed between Kepler and Luna and therefore theres no fork needed then theres no reason to prevent installation into both Luna and Kepler +Thanks PaulIt is much appreciated +Updated regression tests also checked inSending trunksrcmainjavaorgapachepdfboxutilSending trunksrcmainjavaorgapachepdfboxutilSending trunktestinputSending trunktestinputSending trunktestinputSending trunktestinputTransmitting file dataCommitted revision . +We could get this fairly cheaply by keeping a weakref hashmap of seen warnings and logging the message every time we see a new warning +i have run that +Not sure if this will make it +Thanks +problems when I get the object inserted before. +As a consequence by default any state held as global data in the existing module in memory will be lost +Checking the sample that was sent to me and will update the jira turns out that services work fine you just cant them since the flow wont be active when the controller is instantiated +Checking it in +Can I assume that is bundled only with appendNo +I dont understand the intention of the mutlisend change +So failover during commitrollback will always fail with aThere are other failure cases too +See attached patch +this is in master and has just been pushed to maintenance as well +or somesuch +and is last state that shouldnt be updateIt seems no a race condition +verified in +This is okay anyway +studioeclipseplugins +Also note will simplify such that lot of requirements for stats collection will no longer be present +Again this is desired expected behavior +I can try to add a quick fix that would remove prolog would convert an encoded file into an encodeddoes this make sense for you +Thanks for the commit! +I tried the first command ant Dversion logfile and it seems succeeded +Hi AvishNo problems thanks for the feedbackI think I will need a basic test project +Can you provide more details to describe problem please? is strange but if somebody other can add attachment to this issue and you can add attachments to other issues then this seems as some onetime operational problem which is gone now +mrepository to be sure the build can download everything it needs +Yep I had a wide smile when the relative folder expression worked like a charmJust fits my needs for file based backup of inbound messages +The snappuller script has a shell command to determine the latest snapshot on the master +please apply to branchtrunk +So this problem must fix the problem he opened the issue for +configuration similar behaviour as with the project builder when an artifact is not located in the local repository the build goes into a loop that consumes cpu +This makes it hard to verify that the refactor itself it ok regardless of the new functionallity +Not so with and later +Contains one test case for SMX startstop +Please create a new issue on JIRA and attach your source as diff against the svn trunk version of +Additionally it is not the same tests that are always failing +QUESTION How can I the data for columns C I while specifying the use of the same value for column B? The best I can think of is to add the value for B to every row of the CSV file prior to importing. +The patch worked for me +This is not a problem +In server side service method parameters and return type are used to identify DOM documents wrapped in incoming XML messages and also to serialize it back +Patch file as requested +Looks good with a follow up jira to handle the event +patch looks good to me +then it was reopened and setted like resolved +properties files +How do you want to solve the problem of a tool supporting multiple runtimes at the same time and offering code completion and functionality to view sources of classes if you dont include these sources? Also this is really only a matter of what you want to offer your users +Closing issue assuming problem has gone away. +seems to work ok though +Can anyone state if this is the correct behaviour for cascaded WMS stores +Renames the counter Reduce Input Records to Reduce Input Groups since that what it counts +If in the future someone would like to tackle this feel free to open is and attach a patch. +The recorder should have the uncast valueAs for the issue itself +patch Id like a bit more testing before we get this one in +Ive upgraded to and Spring final the above problem no longer occurs I tried to create a simple test case for the regression suite but could not reproduce it outside of our full application +fixed in r by adding an class with the constants taken from +The problem with write is that put returns something and the other functions do not an inner class is required because of the different return types +Yes absolutely reads like +Based on Jessica here is the UI logic for determining whether to display snapshot option or not for KVM +uploading new logs and db on master stable branchpassed +NOTE I changed one outputField to an inputField in the sortHeader sample! Maybe you can change the sample that way in SVN +If you can change to using the Digester API directly I can commit a fix for this quicker than for an xmlrules problem +In the patch I have added a buffer in transfering please check it thanks +As for Ph patches I am updating manually because I had some issues with local branches for this issue +I can pick this up as Im already messing with the Camel forge support to add Java DSL + +Definitely different jira items no disagreement +That way when we fix it then we know by running the unit tests that its fixed +The old can be removed when the associated plugins have been dealt with +My patch also removes which is not needed +Closed after release. +I pointed them at configurationminimal on the branch +That wont work +In my opinion a jobs queuing time equals launchTime submitTime The startTime is ambiguous because it may refer to the startTime of the AM or the startTime of the first map task +Great Mattiasthanks for the quick feedback! +java file in build dir +Leaving open for a test. + +Attach patch for this problemWe reset the event only when all waiter is notified +Closing issue. +slip to next version +Can we just use the existing is logic? v it works except your syntax doesnt work +etc +Currently it is just proposal +The hftp servlet should use the same user as the http browser since they are roughly the same functionality in different formatsThis is a good idea +hmmm +The scenario you describe seems plausible Prakash +It takes a minute or two longer but it works +This is a nice solution. +Trivial patch to modify default jvm startup options +Marked as fixed for +All tests pass but theres still some cleanup work to do couple of places where I cut corners +If we could get the auto paging through native protocol for CQL then we can easily implement the CQL record reader +I dont object to adding tests +Committed +If it goes well I will post it tomorrow +Fix checked can close this +But in that case doesnt redirect me anywhere +Thanks David but Im looking for a clean solution without tweaking server side. +Let meknow if it needs further clarification. +Any way to get it to you +So please close this issue +I think you need to take the nocamel stuff into account +You are using Eclipse +I believe that this functionality is implemented in the transform goal +Basically what it does is open the META table stick in a bunch of rows that dont have inforegioninfo in them and wait for the master to clean them up +I dont see any other setting to control this in the attribute type definition +Thanks Hoss. +We should catch +Resolving as fixed +I didnt set the search type which is substring by default +MikhailSee perhaps + Username muruga +Are you able to attach it to the JIRA? Otherwise emailing it to me would be was never happy with that change to +to fix path thanks! +current patch references CDH +If the object is going to be garbage collected I think It is quite safe to rollback the global transaction associated with the connection +The main issue is to not reset the on a refreshNormally a revision directory is named something like and after an update a new revision directory is created called +Going to commit this +So in some cases attribute namespaces are not properly for next siblings if a sibling had a binding +Why was a released package replaced with a faulty one? Everyone using buildout and mrdeveloper projects build will now fail when using thrift +If its to support dictionary stemming with wordlistloader then it really needs to be one file not two files +Yeah I agree thanks for pointing that I will ping the JAXB guys because this prefix is automatically generatedAnyway i think the behavior could be enhanced to avoid throwing an exception when parsing such an xml +v +Please supply a test case which reproduces the problem +Attached is the patch which changes the block allocation as suggested by everyone +What do you think ScottAs Scott says Id prefer to avoid adding extra functionality into Spring Security which is already provided by the CAS client +We dont upload snapshots to need to attach the pom for jmf as its not yet in in jmf is URL updated +Not a Bug Its wtp specific behavior +Thanks for the to so this gets addressed in the next release +So it turns out that this issue does not have to do with the order that statements were closed +So are you proposing to add NULL to the CQL language or add some more parsing smarts to cqlsh +the patch looks good +Thanks +I agree with most of them except forComment concerns about user identity +This can happen when a map is left open for a while +Fallback on english bundle does not work +Thanks Colin. +Attached service archive +All of the behavior is in other classes that use the models +svn ciSending srcmainwebappindexcommunitydownloadTransmitting file dataCommitted revision +Closing +I missed + +Assign to Daisy review +Attached is a patch that applies on top ofVerified that is set to the correct value based on the service being started +That way writeXml wont write defaultsite properties +Adam I think you are right +as rejected. +Youll hammer the region servers with individual get requests. +Forgot to change for the change in revision Ill give a patch now to rename to +We should remove unnecessary information from cartridge definition +css class which can be overriden by of a bulk update all resolutions changed to done please review history to original resolution type +This patch does the proposed optimization +In fact there is a third problem one related to the +adding note on classpath needs for JMS source in user doc +I like to contribute and fix this issue as a final year student in Computer Science and Engineering Department in University of Moratuwa Sri Lanka +due by Cache has been upgraded to +So this test require DNS resolving do you think this is an issue for hadoop unit test +Thank you Jing +I bulk changed all open issues from to unassignedIMHO this will solve your issue +Resolve outdated cleanup the associated FIXME +Attached a rebased a patch that addresses Alejandros a rebased patch +bq +In the long run I think we should move the Value stuff to as well this would allow to operate on Name +But such problem existAny help +We are basically guys working on this stuff +componentControl should be of a bulk update all resolutions changed to done please review history to original resolution type +This should fix the issue +It looks like it may be related to the pisoa code +Please feel free to reopen at any time +I am going to have to check into this much of the reflection done happens in OGNL or other places +If the user does not fill in both the first name and last name the creation failsso to me that is a bug +This patch also removes this test from exclude list +Hmm +With we need one client per container because of token +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Normally the error disappears if I change the script file and save it thus starting the builders again +I will try that +Uwe both your ideas are great +Why would this be better? The current design is to add them as frequently as possible without significantly impacting file size +Can you post instructions on how to run the test manually? Then we can do another jira to make it more automatic +Im getting the same error when trying to use my own java agent jar internally this jar sets up and runs SUN JMX a quick note to confirm that by adding the following arguments I managed to start with aspectJ load time weaving agent javaagentpathTolib pathTostandalonelogThe other point as mentionned is that I also added to the boot paththanks Stuart and David for the this to you David to figure David +checking now +We can close this SPR now as again thank you very much for yourm glad that it works for you now Lou! And supporting Hibernate Session synchronization in a remote transaction scenario too is a nice improvement after allJuergen +If I hear no objections Ill commit the patch there are many threads making few calls each to this method we are opening a memory hole +My point is that unless you know a script is really a class it wont make are agreeing here really the question is only one of presentation of information to the user +I think it is good will try to apply it + has been retired so wont fix +In trunk pending change to X +is committed. +Needs some sort of test case to reproduce +Thanks for the recent attachments Ill see what it can be +I just ran this file on the trunk and the extra spaces no longer exist in version +in that case shouldnt the dfsclient enforce it tooRE unit tests Craig and I added a bunch of them to just hudson doesnt know +Changing the output of an existing command will definitely break people +This is fixed for by Uthaiyashankar +I refactored the code to include the fork parameter +Thats bad enough that I think a deployment error is warranted +Besides I dont see much difference if we use null instead of + +Duping this jira +not a blocker +For the class I think its not correct to remove it from the sources its better to add in the java build path the from frameworkbaselib and the errors will be disappearThanks Yin and MarcoChanges are in trunk rev +New JSP File VPE wizard screenshot is attached +Manually added +Thank Dave +And if upgrade by itself doesnt work then try this jar in! that jar wont work but I have a version on my machine that does work +Ah indeed this is even easier than I thought +In the case of quoted metachars the quoting is never removed +Is there anything different about the lifecycle of these beans as opposed to a regular spring bean +Given the small number of cluster coming with a simple naming scheme should not be hard +should be closed +Too fast This just happened because I disabled the Ebay component on which data the data was relyingSorryOn the other hand I still dont quite understand how the two product categories can have the same names +Release first +This patch is not final as i wanted to test some conditions with this +The issue will be satisfied when there is clear consensus from the parties involved + its an ordering issue the locker gets the default statements +The patch looks good +I applied the patch and made the commit +I dont know if this is the reason for the connections but it will be good to rule that outDo we know if the database goes on boot or if it turns after being booted +This looks like the original intended behavior +Fix looks good to me Kathey can you have a look at it too +This website is about Enterprise products not Community projects. +Is there any documentation? how to install on ubuntu? Very little documentation and examples the php examples of the search is for php? the creation of the index is there any support with mysql? +It makes archiva not very reliable +Can anyone comment on that +Could you provide a failing test case? hive script +The issue is the Host header always remain the localhost +In the second phase this may leak and the objects properties are initialized +the issue should be fixed in the latestthis is fixed in the latest snapshotBut I have all the time the same problem that my test classes are not found But this test class is compiled under targetI use a maven project pomIf I call chronosjmetergui I did not see the classes in theHas anybody a working pom file with Tests and +Thanks. +interfaces lead to more testable designsPerhaps but a abstract class should be just as testableThe bottom line is that if were serious about providing to our users and we expect our to evolve then we should not use interfaces for anything with more than a single method simple method +setting Jira fix version +Thank you for looking at this stack! You are right about the +I have uploaded a new +Logged In YES useridYes we can define one datasource per database logon +Each mappercombiner generates its own independent set of canopies and so there would be no common canopyIds to use in the reducer +Verified on EAP to set release notes not required +But I do not think my patch changes any replication behavior because is not thrown in case of block replication +Fixed with the rewrote of +this was the original patch which was agreed upon not to use +into trunk +Tony any results by now +They should insert timestamps directions and portsinto the snoop +Right now it seems theres at least three ways that a TTL value can get set or accidentally be set maybe four +it happens if I run +Community members prefer it +Try setting actionMatchtrue in your navrender invocation +it looks like were losing the interesting detail on the core problem +The node will still be somewhat operating might hold connections open but not respond well +Please verify then close this issue. +Thanks for the patch. +not yet i might be able to do it today thoughthis happened in a production cluster so i will have to try to reproduce somewhere OOM during HH must have been caused by something else most likely background compaction +it is is the patch +Added label attribute to validated fields now validation messages uses filed label instead of element ID +Already released so should be available within few hours +Hope the marriage patch goes well +Its what was desired with fieldErrors also get copied when copyErrors is set to true? My initial guess is no but I could be mistaken +I compared that with release dates for Spring and Spring and did a few more git diffs to confirm my suspicions that the method wasnt there in but was added in +this have moved to a latest build and I dont see the issue now +So I would suggest to move this ticket to +patch failed +Hi MaryThis is an excellently scoped for Visual Studio support +Here is a first version of the connector +Can you svn add the tests +Move onshowonhide events invocation so that forum code of a bulk update all resolutions changed to done please review history to original resolution type +thanksSupun the patch with revision +Also s files should be regened +When h is provided to something derived from it will print out both the arguments and the options +This issue is closed now +So adding this kind of attributes to the loggers properties is bound to create race conditions Depending on which appender is first configured the logger will be differently configuredI would take logging level and additivity completely out of the appender service registration picture +A little javadoc will be great +I looped times in branch and didnt see test failureLooks like the failure is unique to trunk +We were wondering can we have the ability to add comments in the file an ability to use lines that start with which get ignored? That way when we have a couple dozen machines listed in file we can put comments in there as to why they are out of rotation +We are no longer supporting the python contrib +Fair point +the start to a patch some tests still failI figure we can get everything cleaned up for postings and then if we feel like it later add stuff for the other parts of the codec API +One of the server implementations does not currently support MTOM +this would adversely affect the reliability of the system since a follower acting as an observer would count as a failed follower even though it is up +So I think this bug was introduced in the branch +As soon as the last test run is done I will attach the patch +Its a tricky backport so its taking a while to test and bake in +Circus will let Hadoop testers and users run one or more tests on a Hadoop context +Git generated patch +looked at it and here is my comments we should not have a separate plugin for this it should be a patch against the current plugins +Maybe it also has to do with the dependencyManagement section somehow +Really? It was putting the msg under the key instead of the property for meAt least thats what I think I meant +It looks like svn merge has an option to ignore spaces x b which will hopefully make merges less painful and anyone wanting a more complete annotation could look at the branches +All tests pass +Ok Ive reproduced this +Fixed in r +Then it proceeded to delete B +Ill probably apply the patch in a day or so if I have no feedback +add chmod ax HOME +latest profile dump has pointed out a issue as causing some memory growthso feel free to drop down priorityhowever it seems like using the bytepool for the stored fields would be good it would be good to make the pool sharedIt still bugs me that yourkit is claiming DW was using MB when youve got a MB ram buffer +If any problems arise please open new bugs. +Please review Thanks! is already fixed in CVS +That means it wont show up until one has logged in and got a token from the identity providerI dont think that is a bad thing +JSON does serve as a less verbose format but the efficiency of the XML parsing library should make that difference negligible +Done. +It seems to be ok a thank you for verifying it. +But its a good temporary measure +Did you guys get a chance to try setting the field to null to see if thats the culprit somehow? Id be surprised but then again Im surprised that the contents of the method could possibly be accounting for of anything +Committed revision +Could you rebase the patch against the recent revision +Thanks Anders +splitting this issue in to the two separate issues that is being discussed +Willem isnt this fixed merged to FUSE trunk now +I know that we have a switch to bind the EMF into JNDI and Im assuming that there is some way to do the same with a SF instance +There is a simple example +The main work seems to be in the CEP and drools flow P says that the deployment mechanism has changed so we may need to go M a try and see what you think +The problem that Ive observed in the past is that if the field is too wide you lose the separator between the fields making it impossible to figure out what is going wrong +need to set proper fix to issues that were set to +Actually a long is quite enough to store all the information in the ISO date string subset used by JCR +Clearing fix version +Ship it +All of our dependencies are marked as to make it simple to upgrade them and as long as the prefix of the jar filename is the same finds the javadocs and sources +Yeah thats precisely the kind of hack I had in mind +This is fixed in the SVN +HiLet me try on thisHow can I reproduce this bug +Ill give a look into it. +Will revisit solution in Beta timeframe. +I have moved all header files of wsdl and appropriate header files in core +Patch looks good +attaching for use in committed to branch +Yes you might be getting speedups without crashes but you are not complying with Java is only meant to be used when the opt compiler is recursively invoked from to generate a version of a method without synchronization operations because analysis has determined that at a particular call site a virtual method call is invoked on a thread local object +But one big advantage we can have by a single log file is the of various events occurring in a scheduler events happening in the JT +Perhaps should we continue the discussion on the dev mailing list +Updated to use for providing values for rpcrpcrpcrpc +Just fired off an to one of the guys to see if he can see anything +Attached updated fix patch and a test patchAlso attached a make patch that removes tests from makeThis patch probably should not be applied permanently as three tests there fail independently of this fix +So if there are no tests no clover db file will be generated and thus no report will get generated either +Hi AmilaI just tested the same thing using the latest Axis from SNAPSHOT the problem is still present +tokenstream still says token is so we dont forget Marks last comment +Ill commit this tomorrow if there are no objections +files + +What I mean we need to generally get away from connector semantics Teiid still have many references to Added CLI command to read the translator extension properties Added get method to the Admin API Added get method in interface +Quick Start with Server +Issue is probably related to derived templates only +As discussed +In we see the JVM args are +Hey man this is great for actually having a place where to keep track of thisBut what are you proposing to be done here? To link open JIRAS to this one? or just to enumerate what needs to be done +Plan sounds perfectly good to me +yep was waiting whether you had any other comments +Just to clarify youre just wanting Wink to include whatever encoding it used during serialization as the header in the response right +Thank you very much! +The notion of current object in the work is simply a convenient term I believe +go go go Olivier +In case it is an issue I think we should create a osgi spec bundle at servicemix or felix that contains the correct sources build rules so it works +Getting rid of talk about an Input Format for now +The wiki has been updated and weve decided to create new for new artifacts so that third party use will work without Maven magic. +Did you actually take a look at the attached? Are the import and export statements in there correct? What is in the XML above is just the BND instructions to create the exact list by introspectionIf you think it would be better to explicitly list everything when creating the manifest please state what you would like to see in there +May be some new subtasks will appear +Downloaded this email in Lotus Notes account and opened +The attribute only applies to plugin configuration +Claus I wont be able to work on this for the next weeks +Sounds like a good idea can you create a separate feature request? The reason is that itll be easier to track this whether this can get in depends on various factors +No further comments patch applied. +Actually I meant to keep this issue open rather than resolving it there are a couple more minor changes left to make +is escaped by a backslash to ensure it is not treated as the character class +Thanks for helping outI copied the jar from your website to my plugins directory +Yes Galder this is still relevant because the new ST code still uses CACHEMODELOCAL flag for the put commands during apply we can remove CACHEMODELOCAL flag and add SKIPINDEXING to avoid the side effect in Query module? Not sure if this is a legal move +Closing this as fixed as there was a test written +to cutover to +yes the issue to the site. +Chunking is fixed with this and other fixes. + emt issues fixed improved runtime info check for calls allow managedunmanaged conversion in check routine for static fields mptrsthis fix is complete and does not require any previous diffs to be applied +Also added a log message to track when this occurs +The tolerances provided are about as close as I can get to the true values unless we want to sample more +To summarize lets do this Rename to without changing any functionality +Ooh one exception we need to swap in for the hadoop build putting us in a sort of weird position of using an older version of avro for a newer version of hadoop +If everything looks right Im going to add the same functionality for all actions. +With the new cleaner interfaces it should require a lot fewer hacks to implement +lgtm +Reopening. +The version is just about the same but hasnt been fully testedId be glad to have your feedback on thisRegardsCedric is a duplicate +This issue is caused by for details +This patch removes sortingfilter by title except for task id +The stats cleanup would be submitted at the end of the stage while a CFS was being initialized +I agree if we can call the batch file we should have access to the wsconsume code itself to figure this out reflectively shouldnt we +Close all resolved issues for older releases. +This is a newer version of the pageflow used to repro the problem +See my last comment +patch in classpath +Besides the of the directory structures I added a couple files for to create the maven plugins for the source code generation bits +Still to commit that? +add a Select tables to the Seam Generate Entities that allows you to use DTP table browser based on the connection information in the selected console configuration +been a long time ago +bq +also removed in at trunk +Put to the table over and over again +So I asked Power flasher people if they could give you help to solve that problem +Supporting it in Maven is definitely not trivial and it would probably cause far more troubles than it solves +Moving to as I dont think Ill be able to do it within this m moving all bugs out to and we can move some bugs out to to get some release action going +Nailing down execution order will limit our ability to optimize later +Ive committed the first patches then. +Created a test using the provided classes injections are there at deployment in RC +Thanks Amareshwari! +Yeah Im aware of that +I have gone through the jBPM codebase and there are actually two issues to be fixed the invalid check in the fixing to check the as well as checking the current fix these issues the attached patch must be change to adds a protected method which correctly identifies whether a transaction is being managed or not +I agree +Not including a unit test +This change breaks the encapsulation of the Config class +I revised my investigation and In fact the rscontains rows +Yes it reduces the readability of the code do we have any better solution out there +Has the change for this been pulled in so that we can resolve this in the next release? +fixed in branch by adding installed into m or m and will there be a new jbds build for this +My goal with this jira was to protect ourselves from future needs where a copy constructor is not sufficient +It seems that we can have a single protocol but AHS can have separate implementation of it +but then would docview be a better option +imho +what you have left of it +RTC task move of all unresolved issues from to +Open the window widget component test +Applied Andy Millers patch +Thanks! I will incorporate this patch and other community provided fixes into a new minor release. +Im going to try to setup a test case for this so I can iteratively run the code generation tool to see if I can recreate the failureThanksWilliam +If you finish it before we release beta please correct the version +Verified on. +Thanks. +Yes. +we will fix it right for you not going to fix this inIts a bugIts not a complicated fixTheres approximately zero chance anyone is using the feature in given its not particularly useful s no handy for people on +Vitali if we want to fix this issue properly we should implement support of JSP custom tags in CSS +Ah I see the email +I have committed this +The patch looks fine to me +I believe there may be a common issue among these +Hi JacquesThanks for committing ebay component patch and pointing this +There is no unique location for jsvc which bundled in Hadoop +Yes fix to did also fix the failures +I think whats important is how long the lock is hold instead of how many places in the code the lock is usedAs to the I agree with you +Javadoc doesnt really strike me as a good way to provide documentation for users +classifier added if unknown one. +Agreed ideal will be to compute and have the intersection of the nodes in the split information +So every time we checkin into pigtesteeharness will trigger a hive build. +Mark Sure thing Ill hold off +PKCS is the keystore standard too and it isnt Sun dependentYes many application servers have JKS as their default keystore type +Hi This is a bug of enum serialization +Thanks for all the explanations. +So this is more an utility method than a service methodBut if you think that its worth having it in the service we can add it back this is not really utility method but it is a question of abstractionThe get method takes a bundle and an optional name and maps this a user name +Done +In Teiid server there is a option to add overridden translator by specifying a XML fragment just like Designer does in the dont currently expose the translator term in our Importers though we are matching up a translator name with a given importer type +And it seems the diagram gets created when you open the generated jpdl file so I think were ok +You end up w a single index that has N commits one for each pctg you need to test +Apache committers only have commit access to the metadata but not Gumps code +Assigned to release as this is a critical regression that has recently been introduced +ClaudioVertex input superstep is a blocking operation when sending the vertices to the destination partition owners +closed due to superceding issue +I dont think so thats client API stuff not +Thanks Tiago for reviving this test +I applied your patch +Fixed in the trunk +if user settings is not specified text field should show the default value +Rather a followed by an n + +The reason is that sometimes you communicate with customer FTP servers or servers that give special responses depending on files transferred +This is apparently fixed + +The problem seems to originate in an invalid count of closing +The service should throw an exception or return some error code and the filter should do the redirec +Still this problem? It shouldnt happen in CPSeems it still includes GACP Pavel could you please provide the packaged used for EWP? I dont know where to get it so to check +This patch fixes first points +I have prepared a new patch for branch of I review it again. +plans this for +Attach the very simple patch +The casting did not feel like a good fix but for things that inherit from Configurable you cant change the signature and the solutions in other hadoop ecosystem configuration objects were fairly different so I didnt really know that I wanted to make those decisions for everybody or try to plow through a personal solution before finding out if it would meet approval +Some questions +The Subversion commits for this issue clearly show Ive committed to the CP branch but I dont want this left out if Ive inadvertantly done the wrong I need to get this change in future branches as well but that probably is beyond the scope of this JIRA +Hi ShankarCan you please let us know how this bug could be? Is this still around? and is it a rampart specific issue that crops up within the engineOnly found that the hash leaks memory if we free the instance that holds the reference to the hash which may be the case in an inappropriate exit in a scenario +The ones on the active pages are definitely +The new patch makes retry on all types including and +and perhaps more in the time frame +GreaaatTested +A Patch is welcomed +the SF can be put through the regular Hibernate core is now a way to put the in the JNDI +Can you please run your app and watch the following object in the profiler? BTW what is your consumer app memory settings +good to go! nice job +I discovered this while testing on Windows where file locking is enforced more strictly +So your application workflow should send them to the credit approval page as part of the application workflow without relying on a security exception to drive the decision +Im having problems with interceptor stacks used and probably it has to do with cancel issue as well as validation +Its been so long that I dont really remember +A second pass to more closely follow Suns conventionsLots of errors on Override methods in that dont have Javadocs although this is not mandatoryCant say if more cycles would really bring improvements not really thrilled by the effortgain ratioUnrelated added methods in introspection to retrieve all method names of a class so it will be possible to implement in the future added toString in Arithmetic for completeness +Documented as a known issue for EAPThe modcluster manager modules status page is not updated upon worker node failover +bulk defer of open issues to +Lets get it in or +sorry my mistake +Closing resolved issues. +Im going to commit this momentarily +Please define some time +Theres a very minor fix adding type parameters to the to avoid a compiler warning. +I have attached another patch +So the patch was bad because the toString will cause and if it passes that it wont return null. +Thanks for taking care of this +Pruned dependencies revision +Wether or not that is understood is not the pointThe fact that Axis does not support this makes Axis unusable for me and many others out there +Fixed existing tests +Hopefully I can get an answer soon! Whilst you may see most people using CDI standalone what we are actually specifying here is how CDI works in Java EE and above and this what we must focus on +merged the pull request into master it sill affects branch so im leaving this issue open for for the patch +The fix will be backported to RF if it makes sense to do so +I think nulls shouldnt be cached at the locationmap module level and should probably be done at a lower level so that it can be done more granularly +I have learned since back thenWhat do you suggest regarding setting valuesRegards use whats in the context +Which branch did you use to create the patch +batch transition to closed remaining resolved bugs +this patch adds the ability to convert annotations to geronimos equivalent +Done +More rules need to be implemented including the one youre reporting in this issueDo you want be to disable the current validation until it is fully compliant with the Spec? Itll need some devs so we could schedule it for the release but not before +At least Id strongly suggest to add a word of explanation to the docs +And the reason we have to types is so we can have as much information in one single as possible +Put it on but we really should have that on patch were initially done by Jeppe Cramon and Michael Fredsbo Thanks +You would be completely right! Great catchContinue to started iteration story and its tasks are addedContinue to a not started iteration story and its tasks are +in release +The problems with the third option is that it will not work when we extend HDFS to support multiple appenders in this case recoverBlock should not happen on the client side but on the side so that a single recovery action is performed for all appenders +Any updates on this? Were anxious to see the numbers after the Pig scripts have been optimized +We can then call in two modes and pass the list of tests that want it to executed in the multi threaded mode +Fix attached as a patch +Makes dev work slightly more than change the data dirs and boot is allThis is the embarassment I get for making a ticket as I walk out the door. +the issue was fixed for Hibernate Tools for Eclipse to fix the issue in Hibernate Tools Core is necessary to apply the patch +Added log line on startup so we know what mode the operator selected +So to see alt attribute we have to perform manual could you add please +FYI this issue depends onOnce is resolved the problem described in this issue will likely no longer apply however we are keeping this issue open for the time being +Seemingly the order I used on the command wasnt exactly that that maven used +Patch looks good to me Rini +It seems like we should have a parameter to limit this result set + patch as requested +Hi VinayIt might speed up the process if youre able to provide a test case that reproduces this problem +Ben I suspect all of these are obsolete but please take a look and keep openretarget any of them that are still useful to youthanksdave +In the same vain I hit an identical issue with tables existence. +And encountered this errorAll I did was follows +Maybe the best alternative would be to look for the JEE defined binding and if that fails revert to some factory defined global binding +use to store state +What does maven related files are different mean exactly? And whatever that is what is the differencePlease open a separate JIRA for that +Any other idea +has almost all methods synchronized +Im not surprised maven central rejected we should also have source and docs etc +I see +I used it only in one place and not for storing something in the database so i dont knowBut if other people need it for that they can improve it +at the moment this is set in see bug shouldnt be tied to any release yet there hasnt been any decision yet to ship the with and it doesnt look like itll be ready to ship is it also set in Windows somewhere +See +Thanks for the update Aurelian the new patch looks much better +Byte based +This kind of persons is difficult to find in course Im a great example of this person! And I not want to offend the italian person +Cant write a good IT yet due to +It might also be worth mentioning that as Im stuck behind a proxy server and havent yet figured out how to supply ANT with the configuration necessary to get out to the internet Ive manually placed the following files in the +that way we can just pull DV from any field and uninvert if needed +it seems that with not wotks in both cases with and without recursion mvn N clean install filehomelucaWorkegovtrashregressionjavadocsite mvn clean package filehomelucaWorkegovtrashregressionjavadocsitetestcase lukasmy fault +Im looking at this now +The will enable instead of the +Patch checked into trunk and branch +osgi sub package +I did mean to delete that line +Were in a multithreaded environment so this quick fix isnt going to hold up for longVFS is so valuable to us that if neccessary we may go down the road of implementing a proper fix ourselves +Positioning doesnt imply insertion as an existing row can be freely repositioned +I have added a check which stops a field which is named owner in a proxied object being mirrored on the client in commit efeddfedaddab +The COM interface is no longer supported. +Fixed on the trunk thanks for the contribution Rajika +As discussed the release vote continues and I will cut a new RC for Auth Core +It seems that problem is in transformationsProblem is not with font or color +Yes Solution E does sound the most elegant +patch with +well done Emmanuel! +The delete option is not passing the affinity group id to the API callhttplocalhostclientapicommanddeleteresponsejsonsessionkeyD affinity group right after creation works fine for me +Move it back if this situation changes. +Thanks Michael. +caches the image when initialized from certain sources and this would be more in line a crash of the threadCould you share any memory profiling data youve collected +I have done some cleanup for empty strings empty nodesets and entity files not found +Ray FYI your test still fails with master +Can you provide us with some examples +CAT in Installer is not a real language +patch attached +Re I got a little ahead and assumed they would release AIR as the same time as FP it should be there any day nowRe yes the funny little box is relevant and yes its confusing Ive asked for Adobe to change itfix it but nothing was done +A revert to an older version might actually do the trick as once upon a long time used to map its onto int +Here you go Claus +It required for rewrite Clone method per Nicholas assertions in as they are not needed given the nature of the operations on x and yThe as operator is applied which will return null if the parameters are null or they dont implement the collections after the as operator +So weve not adjusted the plugin source to the newer eclipse versionsplugin creation mechanisms eitherYou can add the manually to the classpath in eclipse when developing applications embedding derby +Added javadocs to the factory bean +It was only encountered with Derby DB other databases are blocked by + +I know there are arguments for using it +intialized my array object to how much size i needed in form class then my problem to all. +How are you going to use these decorators with Spring? Oleg +Looks good +However for syntax errors which happen before AST tree is built the file name is absent which should be also part of error reporting +There are multiple issues with trying to adjust the revision ids after the fact +Now it should work as expected. +Re replying keeping in the thread +Param name queueSizehandler is better to be queueThe default value of queue and numReader should not be right +Upgraded to in rev +I was wondering if someone could repackage it as a zip or jar file +Yeah I try but are some other culprits to deal with trying to repackage the lib in a good maven manner +Thanks issue is closed now +This was fixed in +Need to test on iOS as well as on actual iPads +m fixed things +Yes do you think thats the wrong place +Ill tried but cant get the snapshot jar +Again if you really think classifier should be a regular attribute open a new feature JIRA and see if others vote for it. +Alex we should just call the same codelogic as the touch button in these specific cases +It gets us one step closer to using a more common plugin but it doesnt mess up the existing strategy until things are properly tested. +It just allows you to make a Float into a Java float in all cases which makes sense to me +release once issues have been resolved and we have a branches patch available +I am good thanks! Will let you know by tonight my time +The proper tokens will not be created in the case of international applied all tests still pass and committed +what about the case where the node leaving the replica set is deadGood point +Why not just apply that throttling code to trash as wellIf this is really a problem that wouldnt fix it when an application deletes a bunch of files +Here is a patch that at least fixes the problem for now + +Sorry forgot to mention the patch is against applied please check. +pending Hudson +Adding java changes to the just committed this +Ryan you are probably right lets resolve this as wont fix +I tried with it works for me +How do I add others +But youll get another workaround with which is going to be fixed with Sonar +Albert we are closing on our final Release Candidate before going GA with version so if you have a moment please let us know if the suggestion I made in the previous comment handles yourMark +i am not a jmx expert but my gut feeling tells me that there could be potential issues with different deploymentsenvironmentsetcJMX is included as a standard part of Java and higher +I started a fix for this but realized it wasnt as straightforward as I thought +fixed in branch rfixed in trunk you explain more about what the problem was and how to reproduce it andor test the fix? +How you propose we add it in? Itd be put in place instead of the randomizing balancing? Or should it be a configuration where we load different balancing algoritms +Patch attached +Committed to +Also why are the tests all commented out +The only potential benefit of keeping the dropdown when there is no way to change it is that the current Locale is being displayed +WW should be helpful in overcoming the limitations of the technologyAs far as generic and selfcontained is concerned I pointed out that this code could be rolled into the existing params interceptor which would obviate the need for a separate interceptor +For I added CMD scripts for beeline hiveserver and metatool +fbricon requests this additional me feature on the depends on newer builds of to Fred to verify fixAlso when you say require newer I need a specific version to mirror onto and then push into the M target platform. +To be committed to trunk and? Correct me if am wrong +Linking to since this affects using ranges +This was done in a later issue +Are you still having trouble here? I looked at your snapshot and I couldnt find anything obvious +Erlend I think the changes look good except we are missing the Client XA tests +Thank you Tony +I will close this issue as not a problem +If having plugin in the name is something you find to be important perhaps it should be called Plugin Stats or PluginsStatistics or something else that makes it clear +In that the JVM is explicitly killed if it is found to be idle +integration test rules +over a year later Any answers to Andreas questions yet +I replaced it with latest hibernate from troublemaker entry is which starts with while usually all file entries in the root of jar do not have it as leading symbol +like +Attaching stacktrace files +This causes false negatives on searches its safe to downsize the indexing bitset but not upsize as there is already some information loss involved +Cant reproduce the indentation againDont know what was wrong the last time +It was just a question about how to work things and its going to involve some combo of modifying the plugins configurations using objects +Also fixed class cast error caused by Oracle returning instead of expected Double +The first thing to try if you suspect corruption is nodetool scrub +missed the variable rename comment +I think its reasonable for the first iteration to do something simple like superuser only or user quotas then we layer on the complexities of pools priorities minmaxshare and failure cases afterwards +This is intendedThe second behavior described above is really problematicIf a UDF breaks because it returns a schema of more than one field it should be changed to return one field of type tupleOnce fixed it works in all versions of PigThis is only removing an unsafe use of outputSchema in favor of the existing correct use +Can we PLEASE get a list of all tablescolumns that will change +Support for Tomcat and connectors for embedded Tomcat will be in to this addition was the metric templates had to be changed slightly +Adding a patch that mentions capacities can be floats +The FAQ contains an entry for this as well +Had to port it to as many of our users have not migrated to yet and so we need to work with Hadoop +Because if reduplicative initialization occur it will throw Error +Im not so sure someone NEW to CFML would have been able to see the same issue +I am thinking of writing the unit tests in Java so that when we do unify the scripts later we have a good set of unit tests to test for regressions +Ive also checked out the source code for this mojo however I must admit mojo development is completely new to me But Im ready and willing to lend a hand if need be +on patch +Other specific control characters between and might also count as strange +On the inspection of the code It is better if we can provide to ops pointers to assign the two callback functions pointerscallbackopsoncompletecallbackopsonerror +You dont assign it when one of the struts developers decides to take a look at this issue heshe will assign it to himherself +I think its a little bit confusing because you could not distinguish between a property and an entry then +Committed revision +even you update from to the proxy certificate is created with lifetime of +I am sorry I havent been following recent discussions closely +So it looks like youve introduced an infinite recursion with the patch +Ive just committed this to trunk andThanks a lot for the contribution liang xie. +jar files in there +The patch is looking good +Fair enough +For now the JDT preferences will be respected +The stack trace should not be printed and it should not say You have encountered an unknown error running Maven +Brett Trygve EmmanuelI took Trygves suggestion and I grabbed the surefire code from codehaus svn +AIUI James includes SPF mailet and fast fail +Nick can you please close the bug onceyouve verified that it is working as expected +By the way the original PATCH file no longer runs without problems I had to manually apply some of the changes +build has the failure +Check installpuppet function for a cross platform way of installing Ruby +Patch looks good +They pass locally +I ported some RF app to portal environment +Im interested in this feature +Juergen can you build manually with your settings +Created +Patch applied +bulk defer of featurestaskswishes from to +code looks good +Is there a RELEASE version for the module? +Im interested in getting this patch into a release and upgrading to that release in Hadoop core to enable some functionality thats currently broken on Windows +The new target injectfaults doesnt follow the convention of separating words with a hyphenThis target wont be called by anyone directly except a very few developers working on faults and aspects +Bulk move are you still using this +Value of is helpful as a guard value +Under it seems the PATH is used to find the executable but I still get the odd effect of the editor opening in the current dir not taking the into accountIm executing a bat file +Committed to trunk and. +Great Ill try it out when I can get the +conf should be first to override in all case the embedded files in jar +Ive tested version PR and the problem is solved. +Is project progressing now? +One of the goals of load balancer is to minimize the number of region movements so that scanner timeout is minimizedThe current approach to balancing young regions aligns better with this goalFor random approach we may need to constantly move some regions to achieve balanced load +I checked out any building the site with I could reproduce the problem using current its fixed. +So if we have all job history files in a single tree then wed want the directories in that tree to be world readable but the log files to be owned and readable by the jobs submitterTo achieve this would need super user privileges on HDFS to do chown +BunoQuiproco in Im near ready to commit your contribution on datafile I was speaking about wouldnt we put this sort of setting in the database like we do with the other content settings +Is it possible for the optimizer to be able to find a plan for a table such that the optimizer estimated row count and optimizer estimated cost can be ? Maybe that is a valid situation I dont know enough about the optimizer yet to answer that question but I though those statistics were worth bringing it up for discussion +public and private method +Until more feedback is received on this issue Im marking it as Wont Fix +The changes I made to build with defintely break the build +contains a test case using annotations +I hope you are fine with it + +changing component to as Wont Fix based on objections above and the fact that since the number of instances is already dramatically reduced making the size of individual instances much less of an issue. +PHP client for updating your solr index with information stored in your database or manually +I have already fixed this a few days ago as part of my work for +Overrides the earlier one +This fix needs to be in place so the code can resolve parent poms in order to propery handle managed has been fixed +The fixes were manifold As described any methods that access frame or scope fields are now considered to be both frame and +So Ill probably patch it for my own use +My fix for other issue however broke the entity bean unit test +Yes it isMy purpose is more to add extended information on services for instance embed and display javadoc properties etc +Patch with test that fails same as the reported errorNone of the changes here should be committed just showing the error +Testing out all my changes now +Romain showed me the latest Hue at the Hadoop Summit +Andrew and I have been on some other urgent work recently and havent had time to investigate +I did not know about at all have linked it to this issue +The failed test plugins are resulting from class loading issues andor dependencies +Bulk closing for +Do you want to make default to GAE and comment out the hypersonic line or vice versa? +Per a mi engatussar o encaterinar +open a different ticket if you need a furth extension please +Thanks for this hintOf course that also requires us to run an http server +It is impossible to make it work with maven because of the need of netComponents. +A module has a tight coupling with currently module +With compression its fine to set the option globally and each node can start using compression when it gets the schema update +jquery +Sorry +Im working on that right now but its taking a while as I figure out how all the stuff is wired together +verified. +Unscheduling due to inactivity +Work in a lot Jon! The latest patch looks good Will commit it once Jenkins gives its blessings +setting successfulAttempt while parsing jobHistory for a job +is a huge cache for a default GB heap +NET desktop apps for years without problems so I suspect that you are right that it might be Mono specific +Sorry for duplicate issue +Ooops didnt mean to close this! +We generate audit trails for security events +Much appreciated! +yes Manfred you are rightI fixed it and added the new testcaseAttaching the new version of the patch +Having a constant regeneration of the would be great for VCS tracking +This seems like the best way to fix this that I can think of as well +Marking as patch available +Updated changes +I dont follow this ruleDo you mind if the patches move member functions to +Instead add the convenience methods to the subtypes directly and then add the support for these to the builder +Tested on branches andHowever Im not certain at of my patch especially on arrays fields. +By the way Im guessing the interesting terms that the query does return when it returns any are based on the documents contained in that shard only instead of the documents contained in the whole collection +Im not sure how this should be reflected on the page +Theres also some older perf testing in the parent appears to be the wrong patch with recent rb feedback from latest patch no major changes +The unit test and patch test have been successful +Please verify and close +Looking at it now +Thanks Robbie. +Could you please provide it +VPN is policy based and static routes cannot be implemented +Except for +Right wed need to add ACL support to list and show to accomplish the original intent but that was my reason for doubting the feature we can just roll this functionality into the existing feature setClosing this because it probably wont happen +When I use the SAP DB JDBC driver instead of Castor writes to the sql statement so the error doesnt I think you should anyway use another method of comparison with floating point values because there could be a similar problem with other databases and drivers +I think we should have only one trie field else we should clearly document why using int is betterdifferent than sintpint +Also you missed the slippers +Slip this feature request +A few questions If in the we populated the key rather than the value then one would not even need to specify by default would simply partition and sort Avro data +Fixed in xwork trunk there any plans to fix this in line? +But I was little concerned about enforcing this rule at setup +Have you tested this already to prove that there is a gain in time using your approach? Also did you find some optimal parameters like amount of threads so that some sensible default values could be set +Since does not support appending data after reopening Writer so the data is being overridden over thereThis overwriting issue is present only in the sequential version of clusterData +If you have other apps that also use the sequence either adjust them or dont use a pooled configuration +Closing as wont fix. +stackYeah had a blankI would prefer a get that returns a and another that returns aBut I think we should settle on a certain design before further commenting a particular one +Although fixing this is probably correct + +Testcase validates both andPlease review and provide your comments +Ill add a dependency on hsqldb to Sqoops and roll this out as another patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Bumping +Hi DhrubaI just looked at the unit tests and it doesnt seem like we have this test that you mentionedBut you are right! In this scenario the file also gets +Yes this would be a very nice thing to fix +Sample applications are now in samples directory. +Adding a dependency for pom parsing would not be a good thing Now that our default public repo is maven one it would almost be a mandatory dependency which we have always manage to avoid +Do you think the boolean is really required or should it not be in the same way than in the fileThe value implies whole content is required +I will review and commit it after upcoming next release of is made +And i cant disclose the build script because of confidentiality +See last one workaround I have found is to deploy a facade to the on the server this I found I needed to compile with even though it will end up running on I use the client jars from and download and use the jar from jbossremoting to replace the version that comes with this seems to work ok on client server there a simpler way to do this? +I will commit this patch in some time +Dennis do you have a pull request? +Didnt this get implemented +Its obviously not just about our Quartz support classes +get all available fonts of the runtime environmentWow! Is that going to be +Added javadocumentation to brief how to use APT document +I assume Marshall is working on that +bulk close of all resolved issues. +Same problem +I think its an old issue +Thanks AnkitAs this is an improvement I did not backport in releasesYour patch is in trunk at revision +Submitting the patch +Can we make the test independent of timezone +The attached patch fixes the first failure mentioned in the description by dropping the schema created by test upon completionCommitted revision +a lot Robert for reviewing this +This is one but I have also found that if you have consumers for the same queue on different machines for load balancing purposes that only consumer seems to pull down the messages +Hi Werner any updates on this issueCan you please at least artifacts with maven so that we can continue our integration with castor +On similar note the static fields containing processor and provider should be renamed to writer and reader respectively +works on dev +Is it worth considering supporting tailing logsWorth considering but as I understand from Ivan and Jitendra BK doesnt support this functionality yet +The import worked fine not sure if there are any constraints related to export using teradata connectors + Jonathan the patch works for me +Just fixed an issue on and master that the was not being copied to host the fix could fix this as well +So all issues not directly assigned to an AS release are being closed +Its fixed +write sync enabled +in the javadoc at the end of the first sentence +For some reason I did not get this issue thanks for reporting +If the flush comes from flush we will wait since that doesnt hold up the other regions +These are the steps I took to do so +Maybe you could try installing JBDS instead of? +Do not we not have to check for tempWriter null during the first attempt to flush the log +Thanks! Author joesteinDate Thu Jun New Revision LogAdded releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka releasekafka copied unchanged from r devkafka for delay complete. +The first round change Page to use a float representing fetchInterval in seconds +Will continue investigations later this weekend +In order for this to be efficient we should only make a copy when necessary +There is no warning about the same resource being located in different paths +Add for all the packages in +Has this been fixed? has been released as of today +One minor comment it looks like a TODO autogenerated comment was left in the code +I ran into the same problem as in the OP +Changes committed to SVN. +Im resolving this one as Wont Fix for. +My guess is that my original test source was not cooperating at the time. +mvn clean install Dmodule trueRegistry API can be built by following command +Im trying to find a reliable workaround so that the travel sample can be released on Tuscany SCA + for the patch +It has borthered me a bit during debugging that the routes that is outputted nicely with toString in the debug inspector gets very long and a bit confusing because of the DLC adding the pipeline and what notI was wondering if there could be somekind of feature to surpress this from showing? Or is it to dangerous to not show it all? Maybe we need to discuss this on the dev forum instead of JIRA +Well Permaine told me to move sharetomcatextras to shareextrasThere they areI suppose she needs to do the same thing for rhel +I have posted it here to get a feedback on the interface additions and the PB definition + +patch was against trunk as of th Jan +Im my opinion its just too many changes just because we dont require JNA but with hybrid approach we dont have to do any of that work +And its good you have documented that not everything is possible with the groovy render and that some DSL is not supported +In my mind points are a team measure but Ill gladly add it to the metrics page +Added additional patch to undo accidental reverts for Ambari Web +If you have time it would be great if you could create test projects using for example Ant builder and report issue to Eclipse a workaround you can try to specify additional Maven goals in Maven preferences on your project +Fixing it now +Did you seriously ask thatOf course +Do not close the htable +every occurence must be replacedThanks and RegardsParas but no test case no response +Another dummy patch +Merged into and in master +Done. +resolved SebastinI am very new to mobile development and using your GPS Code modifications I need some sample code or samples in java script to call your apis for getting latitude and longitude of an Image read from a photo album or camera library +Patch available +Just to make sure there is no confusion +we should cancel the issue +Fixed having faulty get parameters +Closing issue as the version of the bundleplugin which contains the fix is now released. +That definetly helps +It causes one of only remaining Core TCK issues +Now that work is well underway with AS all previous community releases are +Carlos concern about some refactoring has been solved in seems to be some overlap with code in other places in SVN though +The explicit CL is only needed when adding a class from a String arg. +The specific problem Im seeing here is that the user can launch cassandra itself but doesnt have enough memory to launch afterwards and there isnt an intuitive way to change s heap +The file the original mail had patch was created to keep a copy of the research document but there is no further action required. +Can you help? +Bump +Having a look +Also this is not related to the MVC scaffolding in any way +Pull it back in if you think different +Are you using or? Barring both of the above then we would know it is definitely in the implementation itselfLooking at the code nothing jumps out as particularly expensive but as you say for small files if the open is expensive then we may need to do something about that +Updates for exception mappers integration test for unauthorized close for +Perhaps we could maintain a list of plugins for those interested thus decoupling maintenance of core flume from more specific components +Im not going to this Ill file a new bug and link here. +does that tie this class to CMS GC? What if one uses G for example +Andrew sounds like a good idea +updated a test +ill get back to you by midweek with a better name +there is no headerSo thatll be why ffmpeg and friends cant play it +Could you provide some unit tests? Thanks +I just was wondering if it was a stale issue given the last time a comment was made +doesnt utilize coprocessor the test failures were unrelatedIntegrated to branch and TRUNKThanks for the patch JiaThanks for the review Andy. +This is because is capable of opening fresh repeatedly which obviously cannot be achieved if you pass an open to an I agree thats a work around +A BThanks for looking into this + adds a vs solution that my patch for uses +I dont think is needed as was moved under ASF umbrella some time ago +uses to read the applied in revision thanksThe new configuration format is much less permissive than Properties though I have committed it as it improves functionality but well need to fix using either a filtering or by making the more lenient. +Added patch that solves problem to some degree on Linux and demonstrates problem on +Attached are the bundles that I for looking into this fixes it thanks Felix config properties false +Also we have similar test casePlease check that the issue was resolved as you expected +Ill commit to trunk +Had to modify so that it knows that doesnt need to override final methods +final splash was updated in we can close this issue +that makes sense lets roll that back and if have a jboss icon already in the plugin add that in and otherwise just reopen and we will have that as bug that is not a showstopper +Patch Ill commit shortly +Weve noticed that even though these are bulk operations they dont appear to be from an API point of view +Yet another reason to use dev list for discussions +Committed revision +It is going to be released Monday next week now as far as I can currently estimate +Sorry didnt see that you had already responded when I posted the patch +It doesnt actually create two durables subs it creates two subscribers on the same durable subscription +Great timingDIve spend some time to restructure the models using exposing methods through interfaces +closing the case +The fix seems to work +Hey Juraj thats greatI have commented on so you can proceed with configuring tests in Jenkins. +This work well with how most of the times users add functionality Simply copy an existing plugin and change it to suit their custom needs +It is indeed in the hands on the SMX side and if this issue has been fixed then it would simply let Camel users write the same route which will work in SMX if you need to integrate Camel and Mule you can create a componentFor example we have a component already for integration with SI. +Fixed under revision to during tidy up prior to resolved issues. +A test project to reproduce the issue +It allows to send notifications to Reporter Voters and Watchers also now +By design the storage file is an indicator that the old layout is present +I have some notes for this problem Only happen on local server maybe acceptances using an Apache in front Only happen with Platform not product so that maybe relates to configuration or extended service only used in platform And only happen with a special file as attachment others are not used +I think this is still worth looking atWithout vnodes With vnodes So were slower in the simple case +. +This could change from moment to moment +Similar to improvement request which asks for adder injection based on the requirement of registering event listeners which often are registeredinjected in beans using addListener methods +Assigning to J to spread the load +Is there a small number of bits you actually want from the gump cvs module or just all of it and youll be clearing out the empty dirs later onHop on asfinfra if you want +However providing more implementations which answers common use cases could be a really interesting way to solve your issue +Please consider reopening this issue. +Both using binkaraf and the service wrapper +failure to add a cluster for some reason should not call for of zone from vmware datacenter +brl files in remote Guvnor Repository to accept the changes from Rule editor so the file is not any more +No reason to depend on it I just use it for ll remove it Real Soon Now + +Patch for the +snapshotonifvirtiocacheunsafeInterestingly the command is qemu rather than the usual usrbin + +You can of course very easily just use a normal listener and have it perform whatever cancel logic you need still though right? At the very least the various submit buttonslinks should probably throw a runtime exception indicating that you cant have a submit type of cancel and a listeneraction bound as it will never be called +Remote API reenabled +I have no idea what the markers mean +Investigating further +no real progress since several weeks on this issue but I think this is enough right now to address most of the needs concerning the attributes extensibility +It adds custom syntax namly the operator that binds the wrong way in that it is tighter than +test it on the system and it is working +This will be benefecial for cases where the partitions can fit in the ramfs on the reduces +nice workaround! to trunk revision Committed revision +in the previous patch I have forgot the new contructor for minilang which is used when calling minilang script from screen action section +This looks like a bug in antlr +Hi Jerry +Yes confirming that the change fixes this issue for me and also shows the constraint when its a range +Latest patch runs security unit tests in default profile +you can easily create patch files for example by using subversion svn diff problem have been repairedand this is the new align tags have been changed +Who knows what weird and wonderful apps Pivot might lead toObviously it is valid for no component to have the focus but I wouldnt expect the keypress handling of a container to appear to stop working merely because it is displaying read only data in Labels +Heres the latest rework +Thanks Uma +no access to external services except in startstop +After upgrading to SR even after readding the duplicate httpclient back to everything works for me +That makes sense +TimPaulex thanks bug is not reproducible with latest sources +This has nothing to do with mvn I can reproduce with guess is that its scanning the whole drive which is the reason why its hanging +Thanks for taking this issue Jeromy +Forgot to mention suppose you could run a gossip protocol to advertise the block report to address the network throughput issue but that seems more complicated than providing high speed access between the NN and backup nodes +This is a problem of AS adapter +Hi ImeshI committed the patch +Thanks for raising this +Thanks Anshum +Committed w below message +The cause of the problem is just that the path of was wrongly specified in the original patchPlease check if it works properly +version of the patch modeled after trunk and branch +Logged In YES useridOk +I guess we dont follow symbolic links we use rsync with Lrtivz I removed the symlink and copied the artifacts +bulk defer to +If the last node simply disappears before reporting a checksum error up the current pipeline recovery mechanism can overlook the corruption in written dataSince this truncation discards potentially corrupt portion of block we do not need any explicit checksum on checksum errorAnother new feature added to the latest patch is to terminate hdfs client when pipeline recovery is attempted for more than times while writing the same data packet +Look at Sateeshs comments he suspects VM might be corrupted +The problem is that AFTER successful upload the uploaded files are not listed as they should be +Blimey I meant a read with X and Y obviously since Z dies +Expect it to take a few commits to get a reasonable coverage done +What Im trying to understand is what is it that youre trying to gain out of having a different cache instance for each query region? Is it setting different eviction settings for each query region? Exactly Clader +This appears to be problematic on AndroidI committed a patch to work around this issue +push out to +It might not work for an older version +Ive started looking into implementing this +Here the demo app which illustrates this issueThe issue can be reproduced when you run the included test classComment the test method out and you will see the test will runUncomment this method and the test will hung as described in the issue +This does not actually result in multiple bean definitions however it does result in multiple events being fired which can cause problems for portable extensions. +This isnt blocking though it is something we should fix in for sure +Could this be the cause +Looks good +I just found which looks like it will do the strtok that you need +verified that the two tests passed in QA build Integrated to and trunkThanks for the reviews Lars and Stack +It covers the modern poolbundle for Derby and H and also test the persistence managers +I think we can close this issue because everything needs once the EJB support is removed +FixedTo verify create an AS server or EAP server and open the editor +This is in the JBPAPP branch although Im going to leave the issue open for now pending feedback from some folks in the community who are testing with it +just committed this +well have to do this for all releases of this is a recurring issue that will effectively never go away Ill mark it as a need to keep this open for eternity. +I am interested on why OOM happened +Closing this as wont fix for now +So Im for this but Id rather commit the entire patch over on since it also fixes this for reflect and for mapreduce +Seems to be fixed with new libsThanks with EAP ER no more occurence of this exception closing +Is it still an issue and the path is relevant +This one kinda grew into a bigger ball of crap than we expected so its hard to know when we can resolve it +Internally during the deployment of the persistence unit definition we are adding a incorrectly adding a dependency on the +Keith Babo is now the project lead for ESB he should be the person driving this from our side +Version works as expected +slip to is now fixed +Jasons patch is good request merge to branch +you need this patch to make things work as it used to work +Verified in dist built from SVN +Im going to commit this momentarily +The spec in question passes for me on Windows and Linux +this patch adds some tests to chriss patch +I checked in an initial implementation in the hi svn last night +and are not referenced by any of the Roller +bq +All the java code however is packaged in an EJB jar +I hope wont take as long as to be released +if you want a tabbed panel that works differently simply roll your own it is trivial +Patch improved this is ready to apply though yall might to rename an PaulI thought now some time about this but I dont like to go down the route again +This issue is related to and think the solution is likely to come from those two issues rather than in changing the implementations of all the existing file readers +Adds rowcacheprovider to +Example pdfs Something has changed between ghostscript vs in the generation of the pdf and the embedded fonttext information which pdfbox does not currently handle + is disabled in master but I just had the same failure in last comment did not make it +Robert this patch looks good! Much cleaner now version is not appropriate hereThanks for updating +I was thinking of fixing this if you agree with me + for custom tag definiton was added to files and files which packed in jar file too +a contract with VM do not call monexit if SOE happens in synchronized patch looks ok to me +You will only get transactions when calling through the transactional is the interface and its implementation class is which is mapped as how do i invoke these objects via transactional proxy as i stated earlier i just refered a sample application on spring and modeled the same +Another fix soonAnd I was going to mention that I added support in the mock directory stuff for injecting deterministic IO errors +Also excuse my rudeness +an issue with sort order when an empty sort param is provided reported in +was owned by mevenide great in the old world not so great with apache needing to write to those files. +As CXF has a security fix we also need to update camel branch to use this version +I tried it and the patch is ok +This is the output from Geoserver version +I have added Ignore to Ioannis +So that the browser knows upfronthow many bytes it will get +Could I convince you to switch this to a patch form where it replaces all our current invoker stuff +attached sample project +Committed to trunk and. +The fix changes accessnode to allow single execution for procedures +We build locally but need to keepRunning for our app to work +what the hell? I mean I can reproduce this bug +Committed revision Thanks Steven and Uwe +Running all test in can launch single testcases without forcing maven forkMode add confstandalone to your launch configuration like it is shown on the first screen shot +They may be inlined by a JIT ifwhen it decides to compile and optimize the method some time after the method has been used repeatedly +Contains and +I think this broke the build for me with failing tests +I commit the patch for the exception in array allocation case in r +Patch using git in branch patch using git diff +The main point is to be able to use svnpubsub anyway +Patch is also attached to the bug now +Were you able to reproduce with? If not lets close this bug +Unordered is never the same as ordered +The previous Hadoop QA run stumbled over Please resubmit patch for QA +Then we could have a check for problem characters like whitespaceetc and quote escape the whole option only when we see those +Lars No problem +There may be more elegant solutions however +The dumbster SNAPSHOT is in the m snapshot repository at apache +Im running it on Windows Server R +Morphing this into what it really became removing and the JSR JAR from the Beehive distribution. +This seems to work for me now can we close this +Thanks for the updates Sid! And for the reviews Daryn and Jason + +Getting away from Kerberized SSL makes lots of things simpler most significantly not having to switch users in the NN and NN since the spnego filter handles that itself +As a consequence classes loaded before the bootstrap phase of the Spring wont get transformed Servlet classes being a good example for per Juergens comment. +is the absolute latest cut +Have you got a Unit Test for thisAlso did this fail with and do we need the fix for particularly if it affects SFSB passivation and would be needed for AS +if for caching producers caches endpointsSo we need both +Attached is my patch for this issue +The name match is required by the SCA spec +This is valuable when creating clients Id especially like the ability to return the message as JSON rather than XML +If hadoop does its own TTL code underneath that then you probably get the minimum TTL of and the JVM +I think this constructor is dangerous i dont want it on every tokenizer +Unfortunately its global at the moment + +I am not sure ignoring exceptions is the right way to do the test +The patch is great! Theres a couple of the spots which need to be slightly tuned up +This bug has been marked as a duplicate of +what happens if you just try clusterdir? withouth the quotes +If one changes the version manually then it breaks +As Nadir Amra requested I didnt change the file +Currently Its not possible to drag palette element over visual part of VPEChecking with +For buttons it seems to work without +Run this to see the missing attr this issue +For my c I dont see the harm in allowing the wrapper to remainany efficiency issues could be noted in the javadoc +Shouldnt this be Enterprise Application Platform? Or does it refer generally to EPReopening +Ive corrected it locally and will include the correction in the next commitpatch +Committed +When implementing this please keep the dashboard visually distinct from the control panel +If ever we would have to the wrapper the build counter would be increased thus giving something likeOf course the export version of the API packages is fixed at and will not change as long as we wrap the version of the API +Nothing more +The old numbers of Bugzilla numbers +The fix is to allocate the array only on demand when a get call is made +Also the patch is complete except for the pages I mentioned but those could be added after merging in +Fixed +Thanks for this quick answer +Reopening this one. +Ive already forwarded an email to Rob and Guillaume asking for their advise +A CRM run clears the bit +The patch applies with some fuzz +to the community Sherry +I have tested it and for me this issue is still unresolved +Thanks MarkThe issue is fixed and should be closed +thanks +Temporarily reopening to fix release note. +Afterwards it will use the RBD cloning feature to deploy new Instances +Thanks Amar +I think this may be a duplicate of which is resolved in +Now two ant targets for stress tests just jboss all transports including jboss and raw +Sorry guys + now runs on its own +What you can also do for a low priority task is to give a good direction and leave until someone will do itGo ahead on Hadoop side and please dont forget to keep implementation genericHere I will leave a simple implementation that will give an idea what I expect as a guideline +Please check questions in comment need to track all which must be verified +Hopefully will get it done tonight or tomorrow on the train +Do you consider adding back the Refresh button also to the selection of existing application in the application configuration pagethumbnail +Is this data from that got upgraded +Make sense to me +Maven somehow cannot compile the previous patch +in rev Thanks Erwan. +Assertion fail throwing ISE are just best effort +The changes are self contained in Upgrade script only +For instance as time goes on my will do more supporting migrating more and more old index formatsconfig settings to the latest indexing formatsettingsAt quick glance looks like it applies to writing new content not reading existing indexesDoesnt seem quite like that would do the trick here +Marking closed. +Created to track this +This would imply the error message comes from the default DTD validating SAX parser used by +Hudson CI just builds the packages fully automated maintenance is simple no requirement to move or copy folders before preparing a patch or commit changes similar to the for maven many projects are doing the same provide ready to use packages and having a debian folder within root nobody would like to create and maintain patches within the Debian project as long es we are within incubatorAs you see within that issue people would like to have Debian packages for Thrift +Its possible it might be a windows paths specific issue which I will investigate further +talked with Zheng offline it might be easier to specify a semantic error instead for compatiblity +Geronimo changes Committed revision . +Thomas can you take a look please +onsourcelistchanged and ontargetlistchanged +If you get around to writing code well take care of that during a review +This xml used to handled fine in Axis framework +Igorplease verify that this issue is fixed inIlene put a comment back in Feb that some fixes about the patch refered to went into bugzilla which became in the transfer to JIRA +Look at the please look at the in +Our projects do have empty source directories and it wasnt a problem until this upgrade +Please this issue when more information on how to reproduce the exception is available +Looks like maybe both addRoles and addGroups were missing definitions + few problems spotted on build +Applied in both master and on Jenkins +Clicked at entirely the wrong monent and assigned to myself accidentally +Here a quickstart to track the bug +Currently the supports returns false the fix for this is to make it return true? Can more explanation be provided for the previous comment +Ill take a look at thisMeanwhile can you please add a unit test for the same to ensure we dont break these in future? It could be a simple mockito based test too +Reopen if you find other examples you want renamed as well +mavennbm plugin seems to have problems with the creating netbeans artifacts but thats a different story +mohanzl this latest crash is with +Nicholas gethomedir is required as you dont know where the FS impl creates the home dir +This has mainly two reasons it simplifies the implementation and is in anticipation of JR where potentially either a session refresh is required or a new session should be used +Karl in your stack trace you can see a class named This class exists only in Maybe the cause of your problem is that you have an old version of somewhere in your classpath +I assigned them to beta release so that we can discuss them in context of the next will require that we split the command executor in versions a required and a requiresNewthe requiresNew will have to be used in and in the +namely youve got libc linked to a file that doesnt present all of libcs functions +I can try on another machine too +import with no network connectivity +I marked and classes deprecated to be removed in turned out to be useful for enforcing a maximum limit on content streams classes deleted all issues fixed previously after a brief review of each. +committed this +Would it make sense to add these fixtures to the existing since I assume they are testing cursors +Closing bug. +As long as it fits with the existing coding style go ahead with the commit +Actually Method Returns an which represents a subset of the elements in the source Method Creates a shallow copy of a range of elements in the source List + +Needs to be ported to port for +How is backward compatibility handled with? +I will say that I only tested in IE +Looks good to meThis is not on a hot code path right? The hottest I found wasEven volatiles create memory barriers in many cases so we have to be careful +To clarify force works as a pom configuration element but not as a command line parameter +It was my local problem +files committedJust need some testing before being able to close this ticket +Then the file will pick up required jars from an environment variable or property set by the user +If the logging of this information could be moved out to the main log then this would facilitate the information being picked up by Chukwa + and are broken it appears that regular block yielding does work like it should so this only appears specific to Proccall +on v +Add minute timeouts to all tests in above list of classes +I have a user who has meclipse installed and cannot perform an upgrade +In the heterogeneous situation you should just capture the array as type heterogeneous and then encode the schema information with each element in the array +Your fix works for us +Assuming all test cases pass + +Better bug categorization APIprogrammatic problems versus stylesheet executionoutput creation problems +This seems to have been fixed along with the changes for at revision . +Same as as fixed by Graeme fixed this in see +I would not be surprised if object overhead doubled or tripled the memory used +Will report back if there are any problems +see reviewboard as totally agreeIn general what do we think needs to be done before this can be committed? Would be very helpful +Verified in SVN branch for +We have reviewed the new Git repo and found everything in order +and boom +In this patch the blocks will stay in the list but will not be scheduled for replication because no new rack would be found to allocate sufficient racks +It looks like I might have missed the first candidate but in case theres a second rc ill see what I can do to get the test classes compiled in +Thank you +Pointer to this is somehow corrupted during method execution and incorrect value is passed to monitor exit then we leave the method +Cool +If we agree that this is the sort of architecture we seek should we add it now +I tested by pulling down source and applying patch to my local copy +Oof good catch Jing +Not quite yet but soon +I wonder about that too +Bulk close for release +I updated the havent created an example for this because I think its pretty straightforward to make this work from the docs I couldnt include this into any existing example so a whole new example would be needed. +unless they are really called of should be fixed now +I have added them in the branch as well the the trunkKurt +Easiest is to shut down and startup again in a soft upgrade db but bug would also happen if you managed to access enough tables to age out the created table from the container cache +Retargeting to and +Will fix this in +The console is slightly crippled because of these two issues. +What this means is anyone who implements the interface and does not extend will receive a class cast exception +Index fetch failed +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Just now Ive moved stuff around and Ive got all kinds of elements double or even triple userTasks sequenceFlows their shapes +Hi JurajThank you for you efforts I am so full of hope this will be fixed +Heres a patch +Corrected broken link settings for documentation generation +I have the same problem +Actually if Ive got in progress this must be as well +With this change the stopped providing attributes If neither of the above mentioned special attributes were requested +Hi Koen could you let us know exactly the keyvalue pair you want set in eclipse preferencesThen assign it to Denis who did similar setup for Drools +That is the only reason so if youd prefer I can remove the blocks that error out if its not master or regionserver +The port works like the endpoint using to block for responsesCheckstyle build passes +Integrated to and trunkThanks for the patch FrancisThanks for the reviews Lars and Matteo +What I applied for trunk +The latest version of the plugin only includes + +to any requirement on puppet +we are still working on should work now please does indeed Works both in mvn deploy and in UI loginThanks +Thanks for the update Mateo +Something like weve done for with tracking document scores and maxScore +The bug was a PEBKAC I wanted to install hidden packages and listed them rather early in the package list +This is a conflict with ajqueue being embedded on the same template +Capture in physical plan +can now be auto generated using Plan Creator wizard for most of the web applications +Im going to commit this momentarily +Patch adds logging and return empty results to avoid null pointer from unimplemented applied thanks! +Thanks Claus +Thanks Russell and Jeffrey +This should make the wizard usable now +NET project files too under the existing Jira for that issue +This file is created by admins and put in the cluster nodes +Committed thanks all. +Ambiguity resolved as there was a surplus alternation that had not been removed +The bug title should have read does NOT use +The deployment completes successfully if I deploy manually from the admin console. +The fix for this was included in the patch for +Maybe this is still the problemVeraCan you try to determine if the guard page placement is causing problems +Is there anything else needed +Depends on me know if theres still more to purge here +They are from another example +Lets wait for his comments +But it appears name is still not right need to rename again +fixed error in in r +is there a way to determine whether we need to instantiate the class without having to check in every execute call as the current PR does? +However I have run repair which I would expect to clear up the tombstones no? I still observe the same problem after repair +Leave it open +Moving out of +lets reuse the same issue +Net I think is that theres no syntax that preserves key case and works in both ACF and Railo +Now all missing data has added +The solution has to be at the handle level and not the resolveConflict level because otherwise it could try replace both failed bookie with the same new bookie +Reopening this issue in order to make clear that it is still not fully solved because of relevant issue see comments +Patch for trunk incl +Mark as wont fix because its not a bug. +I think I have a plan here that can avoid the hell of making MS +Closing as invalid +HadoopCheckin and PigCheckin which passed before all fail +I have developed and tested a fix for this against the Tag resulting in the following file touchesM srctestjavaorgdroolsdecisiontableparser srcmainjavaorgdroolsdecisiontableparser srctestjavaorgdroolstemplatemodel srcmainjavaorgdroolstemplatemodel attached diff +Resolved awhile back. +See also Remove reporting config from parent could also use which produces reports like the one attached +I can confirm this with trunk +Thanks for investigating +This takes seconds and slows any other searches that may be going on down +linking this issue to Allow to configure DB persistence managers through JDNIthe described issue could be solved by using storing db connection details in the jndi registry +to clarify the subject does not get rendered after you put send +Verified by Alexey. +and b +The only reason we dont keep the bindings and message in the same journal is because of linked lists +Demo message +Does the s successfully load the classAt this point my preference would be that we close this as WONT FIX since users can easily create their own custom implementation and register it for their own peculiar environments and Im not sure whether this is a valid scenario or a workaround for a Welogic bug or deployment error by the user +Vote passed and the release is out in maven central. +As soon as HOWL is available through maven we can apply this patch +ssh directory is shared thus permitting ssh access between nodes +In the next couple of days Ill port my work to the trunk and attach the patches to this ticket +s this record since is now released. +sadly the CDI TCK has a lot of tests which mix standalone and EJB features +Oh well +Thanks for the test case. +Thanks describing the current uploads to JIRAI have generalized the capability for creating XSLT stylesheet parameters into the rev of +Oops + +renamed itemProvider attribute and static statistics closed as resolved and released +But please note the logs were captured when no pollers were configured against the queue. +The previous patch seemed to be corrupted +I dont feel that people are having trouble editing build files now +In fact the code seems functionally the same just put in different places +What this case tested on the RI or +This patch does the following +So it is fixed +This is the corresponding to the spec +Im attaching version of the patch to rebase it against recent trunk changes +Now Ill have a look at the converted tests and Ill try to come up with more test cases +deals with the that is not payload aware +This issue is prompted by a fella loading hbase with UUID keys would likely spread writes nice and evenly across the cluster requiring the flushing of many regions to go from to +Fixed in revision +Shane is this completed nowI have some test failures that I think relate to this +Please resubmit with further information and steps to reproduce if you are still experiencing this issue +A committer will review and provide you feedback or commit if deemed fit for the project +textbox that shows up +Please verify +Removed trace statements +I have merged this to. +Suresh thanks for linking to I missed that issue +Here is an initial shot at the web site +Mono I know that the Avro project includes a nant script to compile and run their C tests under monoIve concatenated the two patches and attached +maybe for services but not controllers that it wont reload them anyway currently so this is not much of a loss +This is committed to trunk +Isyour Ent basically a measure of doc deltas? Im confused exactlywhat it is Because I would think if you take +Patch that moves read to and reverts changes to We shouldnt add layers of patches for discrete issues +This is mostly deletions +We tested with broker +No response assuming ok. +Preparing for the fact the patch will be quite a bit different +However there is a fairly easy workaround for this problem +Fixed in SVN +But does the patch applies cleanly? I assume it is yet wanted to make sure +Im not in a position where I can try this with trunk but running ant clean test in trunksolr passed with the patch applied +While you could certainly argue that should know how to reshape its MIME structure to handle a plaintext email you could also make the same argument that should know how to handle HTML text without attachments in the same way +Looking we dont have this issue in trunk +Thanks Nathan! Yes your approach is better but id like to improve it a bit +Suppose we have three replicas and have arrived at the beginning of step +Waiting for rsync to resolveclose this issue + is a patch that was committed to trunk to refactor out of +got a lot of work in logj recently +Milosz Frank or Tim if there are additional changes that need to be made please the issueThanks for the patches and the time you spend debugging this problem Tim +please change LealAgenda has been publishedptBR Agenda j foi publicadaHousekeeping rules for delegatesptBR Regras de Conduta para is coming to Brazil +Does it means that we cannot use the same event on tabPanel and itsHowever onheaderclick is not an equivalent to onenter since a tab can be selected other than a click. +ESB project test With always false attitude only the ESB container gets added With always true attitude both ESB container and default container get addedSo this is a legitimate concern +yes API +What implementation are you using +Im fixing this in with Later instead of Fixed since its not fixed yet. +Thanks +Patch looks good to me +Still not sure the best way to fix this +This is how currently works +keystore files being in the branch as well as trunk +To the next release +Reopened new improving JacquesMy apologies for the mishap +Using the and Grab things start but I am stymied because Oracle have only released libraries for Linux and I am throughout +Attaching UI from the latest patch Made changes according to Hongs comment +Updated config on eos and restarted httpd +Sure let me know when youve done enough testing and if nobody objects Ill commit +However since the Spring branch is in maintenance mode already Im afraid it wont be backported +In fact most involve shutdown an area we just have not cared a great deal about in Solr in the past +Im going to go ahead and commit this in a day or two unless someone objects +maven jar completed appliedthanksdims +Cool Ill take an initial look and either leave a comment here or a link directly to a gist +name grep feature xargs n grep Hi codejbosstools +Ill update the patch and post shortly +Isolated the change to in patch +Thanks so Ill continue with this issue in the next days +Please verify and commit it +Pushed in bab thanks! If you can think of a simple way to add a test for this please feel free to submit another patch! +Resolving +Another fixed bug that was left as new. +Please validate and close or reopen this issue +A technique Todd has used so far for is to have one of the NN mounts on mdadm in faulty mode w injected read and write failures +Discussed this with Will and this need to be fixed for +That patch would allow all existing s to continue to work +There will not be a separate +Current will go through the class path to search the but in the OSGI environment we need to use other API to do this searchingWe can provides two kinds of from camel +You should spell check before you check in +Thats a feature not a bug +For EE environments we may have to come up with implementation specific providers +Some test cases is not suitable for windows I have add a new test case in the unix +Hence closing the issue. +Shall login store the credential for jaspic and basicformdigestclientcert auth type respectively? +As discussed on IRC not a platform issue any more. +Probably should close this issueI took a look at the offending routine and it was not obvious where the null pointer was coming from +is the eclipse project the repository that I exported right after this problem happened +In the multithreaded test files weresometimes left open which is hard for me to debug +Alex I will fix the test cases by using the right class and will prefix with BURL for the destination that uses that syntax +patch at Oleg +Only missing part is to have tests that can verify we are not getting worse +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +WIP patch +I dont have a Jenkins job as I was running everything on my VM cluster however there should be latest BT trunk builds against Hadoop in the usual place +The test case to reproduce the issue +Working for me too with mvn test DtestRunning Tests run Failures Errors Skipped Time elapsed secStack the similar work is captured in for trunk +uh +We need keys for this file for which we need to materialize an intermediate collection and I think when we do so then crunch creates intermediate output files which are only +Ok so well wait for PR should be relatively simple +Fixed in r +So I let them there +Updated patch fixes two findbugs warnings +Patch excluded from dependencies +Should this be broken down into two sub tasks One for the commiting thread on the leader and one for the fetcher thread on the follower? Lets not worry about that for now +The and have been made available via attributes on the EJB mbean +new reformatted tests pass checked into SVN +Regarding replacing I think that its good to have a uniform page with information about each project +patch to fix the to pass a test the patch +I think the default should be to include the version of junit that came with the version of groovy they are using +Attached patch adds explanations on how current scoring function is derived from VSMI hope it is not to colorful and that it indeed clarifies rather than confusesI marked it for because if others like it I think it would be good to include it but if there is no concenzus on this modification lets move it to a future release +great! thanks. +Is there a fix for this? It seems to fail frequently of late +Attaching the patch +project and put AFTER a Project Clean and you could see that only java files are processed in target fix to trunk +Others if you want to update this for trunk it would be a big help +So I guess that problem was due to some kind of refresh problem +Please try the latest snapshot and see if its still an issue +The locality was intentionally reduced to test the socket caching +You can use wsdllocation parameter to override the default wsdl location dont know if its sufficient in your case +having a plugin or gui tool is not really equivalent +How do the Apache Zookeeper service writes data to a znode as to which all Java files are responsible for it +Run xmllint every so often to ensure you havent damaged +The counter is called blockSafe +has been marked as a duplicate of this bug +The template is missing from this patch +James BurtonDid you have to provide mapping file or just swapping the jars fixed your problem +Adding left some comments on reviewboard +Fixed as suggested +So for now were going to stick with you needing to put them in the native folder thats already there and try to figure out what the problem is +However I derived that class from rather than as I though this error did not have a lot to do with the MIME spec +The staging area has been updated to revision +Tomcat heap dump of have moved cleanup of into clean and almost all problem are gone but there is still one instance held by thread used to init Struts filter +Maybe they have been introduced to be able to reset the iterator but this is now already supported via the interface +Bumped max heap up to M since this worked for Charlie on bit system +Attaching the code changes that fixes the issue +Yes your infinite loop is caused by delegating to the filter in the authentication handler + +Works on S +Ill therefore continue with both site and dist migrations +Yes excellent ideaI see a few questions do we start the record number at or ? do we start counting with the header or after the header + +alex could you look at this one and verify +Im going to tackle it asap though +LOB Streaming +Recent performance runs say that Robins fix has brought performance back up again +The outcome is that we have an extra project with dependencies on both tools and plugin where nearly all the testing goes on +Please save me from thats not a very acitivityCheers for SWF support for inheritance via a parent attribute on the flow tag +I know Daniel has GMF experience +If possible could you try it on a OS? I dont have any Mac OS at hand From the error message seems the exception is caused by missing server runtime or facet in WTP +OK heres the very big patch +I can take it out +Thanks for filing this Thomas +KatheyI do not mind the change you have proposed I just want to make clear that what the target really needs is to compile against the Java SE libraries +Hey it looks like I committed this r with a stub commit message +so if a starting repository detetcs an existing +Should be assigned to me after fix without resolving +I will do some minor improvements on the java code and then test on a real cluster but Im interested by a feedback on the script +Alexey please verify +looks good to me +first that read just works with different lengthsI think you should combine the incomplete block and full block tests into one with somehow randomized block lengths so that your files had all different lengthsIt would be interesting also to test concatenation of files with different replication factors +Ill add some logic to use sudo as a fallback in case of failure +I think the problem is with the wsdlj + +I have converted it to a documentation issue. +Yes +response +Patch applies to the CP of branch completed +I am tempted to close this bug as solved any thoughts? +reenabled the set projection in the catalog on revision +Im not quite sure how useful itd be Im curious about what use cases this would be leveraged +push out to up +Second issue has also already been fixed in trunk in two days +The subentry class has been moved +accepting for added a test case attachment +I cannot see any reason as to why this should not be implemented this way +Yes we have rewrites for all the SMX examples in a similar style just polishing off the rest at the moment +Also the wasnt available +once more +On the other hand I would think this would be used heavily by others +The reason Im using popen in the first place is because I couldnt figure out how to wire file descriptors up to posixspawn +I do like and would think it more of an opportunity for an add on system it layerSent from my iPhone +closing resolved issue +Changed to with proposed changes +We were not able to trace down the source of the bug but this fix seems to give correct numbers +It is better to keep it open for reminder +I almost spent a day trying to figure why it was hanging and just gave up +Could you point that out +My aim is to have a view framework by which one is able to build html pages with Java code and maybe to have a component oriented framework build upon that +Proposed fix for the Hasan Could you create the patch without the header info and from Shindig trunkI am having problem applying the patch + +As described by Artem this header must be specified as a Long in milliseconds since GMTYou can use a to parse the date and convert it to the appropriate long value +And btw CDI facet version actually is not important +Is there any other branch where this should goErik you set the fix version to and in order for that to happen this has to be committed on the lucenesolr branch no +Same reasoning applies here as with Validators eventually we should prefix theregards patch needs to be reverted +I havent figured out yet if anything is lost from the DB +Draft text for Resolved Issues section of Release Notes statesPreviously the could only handle text elements and not attributes +Cheers +Also it doesnt seem like the output is integrated with the ant testreport target +Ian Im not sure I understand what needs changing here could you be more specific +java files. +Icons are applied +Please reviewIf I hear no complaints Ill check in the patch in a few daysOleg +I think the stupid Token caching in is maybe broken +I have same is a work in progressIf you want to test the SNAPSHOT you can build it by yourself from sources +bq +thank youklarissa +Hi CarstenI think that when an order can be linked to a MRF people in charge of this order must know production status +Several files have been removed +The first step will be the interfaces for classifiers in general +Further discussion in +trunkCommitted revision +bq +RS log entries are misleading +Thanks Anthony +This is just in preparation to that. +The other UIMA SNAPSHOT dependencies however will need to be published as versions since the maven central repository rejects with SNAPSHOT dependencies +The bestway to deal with this is have the correct memory manager deal with the memory itallocates +If a single region was locked I guess it could cause a pause for all of them but it would then would go into a wait state like dominos one after the other +Hi! Our preliminary tests show that injection into superclass works just fine +jdench java jar DERBYHOMElib ijij version ij show connectionsNo connections availableij exit jdench And is emptyBut I would have expected that since it was my understanding that autobooting only applied to the Network ServerAnd the really weird thing is that if I then start the Network Server again with set in my file the databases dont autoboot! Exactly the opposite behavior from yours! When I remove the setting they autoboot the same as beforeVery strange +Fixing this is also complicated + This bug has been marked as a duplicate of +I am using the latest with JDK +see related for how resolved. +I need to get ahold of a device where I can debug this +testcase for input +Do not forget to check that there are no errors for old workspace and provide test case for QA. +Marcin Ive already told you to NOT create JIRA ticket on your own you MUST come and discuss your issues on the user mailing list before +Todd you are right +Would you please try this patch? Thanks very much +Fixed with commits +There can occur some problems but those will be fixed in new issues. +The drawback is that we have to wait +So the guide is no longer a draft version +Resolved this one. +Declaring variables return values or parameters of type is not allowed +These issues have been marked as resolved for some time now with no recent activity so I am moving them to the closed state. +this is work in progress ivy patchApply this patch and execute the script which removes the jar files that are resolved by ivy +I am not sure if method also needs updating to support timestamp snapshot version. +Now its fixed but there are a lot of jobs in the queue waiting for publishing +Logged In YES useridI will confirm but I THINK I have enough sessions +I believe that somewhere there is a or something similar but I have no time to check for that yet maybe in few hours I will find some time +reupload of the patch +Cxf could offer an API function to create a frontend for a bean by inspecting it +Otherwise the functionality is unchanged +create a new Roo Projectgwt setupmanually create a file called under the gwtrequest deletes the file automatically +I think it worth to move this discussion to dev list to get more opinions. +you cant reRender comboboxes modal panels suggestions and other popup components parents after they shown +Take a look at the tests and the FAQ it depends whether the hashCodeequals are proxied or not +Our network traffic onto the JMS server is down significantly from what it was on the trunk version from over a month ago +So imagine that a logfile becomes corrupt be it truncation or full of junk +Yes sure +However now it works except rebuilt CC cant be expanded +Go to the user forum for general help and consider commercial support for personal assistance + +Unfortunately init does throw if the codec lookup fails +Patch that needs to be applied to devstudio trunk and branch if these are to be enabled +Ill commit this one unless there are objections +All prefs should restore to their default values now +I moved the posted patch to +We should either close this JIRA or move it to the transforms module +But limiting the total number of locations for a very job may not be useful because a large job is likely to touch some blocks on every host +consultyour tool vendor for further help. +Actually the bug is annoying because it causes regular QA build failures so I will put the fix on the branch too +Attaching a patch for the second part of this grant execute privilege to system routines that by default have execute privilege to public when a new database is createdChanges are Adds a method grant to which is called when doing a full upgrade +Processing the queue with multiple threads still seemed scary if you were to create a lot of collections at once so it seems just safer to use a different queueIm still somewhat unsure about handing failures though for the moment Im simply adding the job back onto the queue this gets complicated quickly though +Looks good +Fixed +This patch addresses the ticket please Sean +Closing issue was rejected. +Being verified at +ill fix it +I cannot reproduce this +Also I think its a tricky change for the branch +When we delete a row of our view we call a SP which may works on different tables and delete different entries but at the end we only delete one row of the viewSince we are obliged to use the row count parameter as the first argument of our SP and since this parameter has the good value why dont use itFranois J +Same here +We use it everywhere +I think the next step would be for someone to provide the necessary implementation and test with Gemini Blueprint +Please delete +And theres no need to write a bunch of new code for this because you can simply call the existing checking code during the dryrun +What do you think +This allows us to use the exact same code for parsing in both places +This has zero value from a User since its internal and the only value I can see are getting rid of some warnings but the effort involved to do that properly doesnt seem worth the effort so closing as issue deals with best practices coding Java +Closing. +Hi +What do others think +to everything David just said we are in agreement +Patch attached +NetIMAP connection pooling with Commons Pool rightAnd the reason why timeout did not work is there were paths which were not covered timeout block if I understood correctly +So I just remove the sleep on IOE based on your patch +Attached a patch with simple fixFix is straight forward just need to add exists watcher again onTest wise multiple worker test will block most of the times with this issue +Are you sure you mean this plugin and dont mix with mvn eclipseeclipse thing +If one of us knows them in person or is familiar with the code that might make things easier + +Dont expect me to read your mind to figure this stuff out +defer all issues to +Has been resolved for more than months +This should no longer be a problem +Ive felt this pain +Okay those are both committed to branch +Most of the time the master and share that same file so it should cover a lot more cases than the current code does +Please review and let me know what you think indeed +But changes in implementation are Created new service ups instead of using ups ups of this service is always Shop because this service gets online estimates for all the possible shipping method of UPS which are shown in Order detail page +Your argument makes sense +so close it. +See inside first. +One concern here we consume Drone Configuration module right now so we need to make sure it is stable enough +In this patch I made the lengthsquared instance variable in to transient +This is hardcoding it to homejboss +Please push the patch when reviewedThanks and push this patchThanks +Please post your additional comments to the new issue. +Hi James! Thanks for such a quick response! I m attaching the snapshot displaying for the asked i am using and bundled with +Juergen +I was looking in the wrong place +Thanks Jonathan +news Happy +Please close this issue +I do not understand the whole problem +Resolving based on reporter feedback and since I also fixed the same problem for mantisrb specs +Could you please try the patch and let me know if it fixes the issue in your env +closed after releasing version +Image that does NOT give the has been added with description that does give the has been added with description I couldnt verify this properly with the latest source code I noticed that the exception is not thrown any more for image by the transcoder after cloning +I believe that the proposed patch is incremental improvement +What do people think is this a worthy addition +So for dynamic partitioning we have some extra costs added for DFS namenode read +Closing all old Resolved issues +patch looks good +So it must be the ESB adding it +See revision for the changes. +bq +Can you provide some details? If there is a stack trace that would be helpful so that I can track it down faster +Unable to reproduce the bug on the latest Build +The issue does NOT exist with +Ignore the unable to render embedded object message this should bechanged all entries of binbash to usrbinenv bash as per Erics request new patch +Well reopen until we have a test case that proves it +SO someone please integrate it in workspace +Here the patch following the official instruction to submit a patch +On the other hand there is some low hanging fruit that could decrease the size of the by another to bytes per column +Lets chat online +committed +Verified by Nathan. + unscheduled issues to the next release +To the extent that the command syntax can align with any exposed JMX capabilities would help +This appaers to be an issue in the stax environment not in local problem does appear locally when using the JNDI or JDBC configuration files +bulk close of all resolved issues in preparation for release. +Closing resolved issues. +Rado have you seen this issue more than once? If not I think we should chalk it up to gremlins +bq +Obviously we need to fixed the test instead +This isnt in is it +This is the xml that reproduces the error +Cant fix since there isnt any extension facilities for it +see +This patch removes the forced for the suggestion and patch. +Resolving fornow as REMIND so that we will remember to mention it in the release notes +Both unix and windows versions of startGroovy shell scripts seem to handle cp classpath only if they are argument +This section of code is particular enough that it might be easier to reason about a patch even a partial one +Update comments +Fixed lineLengths issueThere is no unit test class for +I think you are right about the scalebar +Artifactjbpmpom is not in any repository because this is the upcoming jBPM version +This seems like a new binding prefix which is something that can be added today without changing the core framework. +You do have to make sure the domain name is. +This is really needed to round out soap support +Checked in working sample into sandbox +Ok I have more info on this ExceptionError happens under heavy load when the logging level is changed +button in that dialog the message I see is Encountered an error while parsing the selected WSDL +Upgraded to to. +Please correct Fix Version to one without x like depending on where it was as a duplicate of to gather discussions in one place +Please verify. +I have found the problem and think I have a fix +Please submit the patchesthanksthe wssj patch is done against the branch and the rampart patch is against branch since this is the last release and the version we are currently using +PR added a new script added scripts for each database that includes the table and insert along with a standalone insert only file. +Backported to and branches. +Thanks qxo for the patch! +I am looking for some suggestions +iirc from the meeting and email thread this issue is different from what we originally understood +Assigning to me +Do you perhaps know a way to have these props added automatically by eclipseIll fix this this evening +I guess a better way needs to be devised +Thanks for reporting thisI think I fixed what you were talking about but if I missed something just add an example test case and I can make sure it passes +You should be able to get those needed files from the mavenivy repository +Thanks for writing a nice test +Otherwise it would be out of date constantly +Users report that it comes up with error reading invalid header field error in a maven build and at of this there wont be any clients that use the broken copy they have all manually replaced the JAR with the good one from the normal download there really no way of replacing the existing bundle? I thought maven should be pretty advanced by now +fixed the issue with revision +note the attached patch +New script with start and stop commands +I do have a stronger preference for IF being the last clauseIm perfectly fine with that +I am hoping to write my unit tests in mockito after learning it are there some examples of using mockito within hbase that I could use as starting pointsNo +class file for imports +Yes it is just that there are other prioritiesThanks anyway +A bundle with no symbolic name which can only be an R or earlier bundle must not provide an capability +Ill make it on Monday +fixed in branch not in trunk have also replicated this bug on weblogic and +Let me see if I can get a similar export for intellij to contribute back +Since no objections I am removing directory for now +Attaching a patch that uses a pragma to disable the ssl deprecation warningsBut there are still loads of other warnings +Open in the project issue tracker and get the developers to fix their broken systems as I am doing now +We have WS now +Do we want this behaviour? The other problem with this approach is that if nfs dir bounces even once standby will go into safemode and this will happen silently without alerts +Let me know if it is heading in the right direction +If the required perspective is already active we need just to switch the Welcome Screen to standby mode +you need to actually exercise the snapshot codeYes +Their usage reporting can be disabled usingusagereportingenabledfalsein the file +Sure every log output +My understanding is that I dont +Though at both times the only one who knows about the trouble is the health checker and not the rest of the worldwhy is why your management tools need to be HA toys themselves need to be able to ask the apps for their health may need to be able to do test jobs to probe system health may need the ability to react to failure according to the infrastructure in which HDFS is running and your policy +I still dont understand +I dont understand how one Delegator object per data source is a performance issue +Ive rerun it so you can see results soon at the same place. +Hope to get some free cycles before end of this week +This is rather inelegant and should be remedied. +Known issues for installed on Indigo Indigo SR are +But that will require change in the storage framework to allow such post operations +Should we make a patch for as well for this bug? Can I simply replace the files changed and make a build for? +I am aware that translation is not complete but question is whether properties files should be synchronized or number of translated phrases may differ in various languages +As parent issue has already been resolved with +The patch actually worked I just did misinterpret the data +Best regards what is the issue exactly? You want a comment in +Dave can you confirm on excalibur +Closing the resolved issueLahiru +I need the steps to reproduce or a pointer to the responsible code +will clarify that this is a cyclic injection not allowed +trying hudson again +I will attempt to fix more today +The tests fail on Hudson +I think we also need to check for any +cfb should be reverted +Thanks for the patch Thierry. +This morning I talked with Brian we think I should add a validation for this issue +plugins Additional Spring AOP Support Feature all aop +The patch has been applied as expected +bulk close of resolved issues. +now it also tries to update the settings file without reordering the xml implemented. +If that isnt true +Ok so this appears to be fixed in rev Is this the branch from which will be cut from +It would also be good to be able to do this from the process instance selected in the execution history +Anyone mind looking over the patch so farIt seems to solve most problems I encountered with encodings in RTF files +Have you checked that theres free space on the disk +Remaining issues can be dealt with in separate tickets as people encounter them. +Added an implementation of this along with some tests in trunk +Im planning to create a filter operator with candidates as a child of the original filter op and mark the original filter op for deletion +Ah rightDoesnt the static block mean that stages will be populated at the latest before the getStage method call? Which would imply that we asked for a stage for a message type that doesnt have one which is hard to diagnose wo the assertion +I dont expect you will get a majority for this +I dont object to leaving the code in place +perform will be removed in a future release and well change this behavior then +Fix pulled into master +Reopening +This is the same asThree files need to be removed from Hadoop Common +I moved it to as a result +Mike how did you generate this patch? I tried to apply it to my workspace and it would not apply cleanly +This is ready to be committed +Moving to some code in the sandbox to see the alternatives the new walker is probably the best alternative to go forward with with a saving on time over the previous implementation +I committed the patch to and +Plz remove +backported committed patches to in rev committed the last patch to trunk and backported to xthanks Simon! hard work but I expect there is moreHopefully this module will grow into something much bigger. +Always escaping seems reasonable +You can run it with mvn installCould you investigate what is causing the issue in your environment or if its not a container specific problem could you provide a standalone test case to demonstrate this issueThanksRegards aki +transparent from the data and index format so this end up having no interaction whatsoever with compressionSo this isnt quite true due to the columntuple index being unaware of blocks and the current scheme has to deal with alignment mismatches tuples which fall onto block boundaries involve reading two blocks worth of data in order to decompress one tuple +Fixed in. +I will look at it soon +Modified the +Unlinking from +Hi Alejandro as above that you said is there any possibility for this patch to commit into trunk? If yes when this patch could be commited +Sorry my fault I m a newbie in this processI guess this depends on solving first and find an alternative to a site while wed like to use git for the actual source code +Resolving as fixed +I had this vision of people at some point wanting this check to be configurable per but I think its better to start with a constantChanged +exclude for x +released issues +Hi MaciejIve discussed the topic with my colleague and we agree that jBPM is close enough to the desired BPM behavior when we use the correct type of handlerThanks again for your help and especially for your upvote on the other ticket which has indeed triggered a responseBest regardsNils +The QA subtask isnt yet closed +Same jars and working on my laptop +Removed the requirement of policy change after usingdoPrivileged method in +Looks goodThere are lots of tests that use these need to be ported to use the new interface and moved into the mvn tree +So if you ask for an official decision nobody can give it to you +Fixed in SVN trunk +I tried it out the maven license plugin and it looks quite nice +This issue is fixed and released as part of release +This is essentially the textual equivalent of the visual instance display so maybe the visual instance display could be enhanced to enable the user to switch between graphical and table view +Hopefully his test case addition will help us nail this one down +I cannot see why this is a duplicate of +Reopen to close as duplicate +Fixed for Web EE Richfaces HTML projectsGWT has a dependency resolution error if the profile is activated so I did not activate it by default +hard to reproduce it has most likely been a concurrency issue in which case it is solved already so I close the released! +to trunk +Definitely doable because the instance would need to change for things such as SWT based Griffon apps +Look at the method copy for added a few comments in the code that mentions where this bug occurs +Added a test to first replicate the problem and then after it was fixed to verify that the problem was indeed correctedThis can be merged into the master branch and into the branch for inclusion into +CosYes you are correct +Changed names to encodedecode returning bytes written from encode +Hadrian it is confusing where to put the constant keysThat is why the xxxConstants classes was introduced +Preparation for Gora release candidate +Well my now doesnt show this behavior so Im closing this one +Thanks Owen! +for internal +FYI tool doesnt support +Thus in certain situations a deploy POM may be used rather than building the artifact for each environment +Comments Can you add a description why this problem happens and how you are fixing it? Should we create a bug for access and modification time mismatch between hftp access and direct hdfs access? for the patch +Sqoop has been removed from closing this issue +But if an alternative is activated it is the only available result regarding this condition so it seems ve now aligned the behaviour with Weld +Mostly fixed +etc. +I think I can provide a patch for this +Thanks Alex. +closing this issue. +Im going to close this in favour of better error reporting and the archetypes for test cases both already under reopen if you disagree. +Initial patch +As hive is using atomic transactions to perform individual operations its causing issue during replication of the database +You got it Lars +just a reminder tomorrow i would like to commit this if no one objectsthis will move the contribcollation components to core and later we should consider deprecating the alternatives that are not scalablethis will move the contribcollation ICU based components to contribiCU and this is where I want to bring the unicode supportThanks +Im already in the process of testing a fix to retain complete compatibility +I got the same error message then that Niall got when he ran with Maven +This patch converts non predicates into corresponding hbase filters and adds them to scan objects to save network IO +sounds like a good idea +is used with Synapse though is more common at this point +Is this is a bug masquerading as a feature request? +This solves at least some circumstances where the master and +Manually tested +txs Bommel for contributing this feature +link to +patch that moves the logic that filters propertydefinitions and nodetypes when build tablescolumns metadata which was incorrectly placed in the rest client +Closed as Cannot Reproduce because we cant find where this problem comes from. +will have it. +I had not checked in all my changes +The problem is with Maven proxy settings +Sure it can stuff! +directly configuring MavenThis night I hope to finish the integration test implementation to customize the Jetty runner +With a server which included ESB the same error occurs the serverruntime is not causing this +Patch also adds a division break in to stop indenting the docsets menu +You should be able to use this jettison snapshot by overriding Enunciates in your own pom now +Ive the same problem! +before you askapply this and then there should be no question about resolving the bug +Still I dont know what conditions are required to error for occur +Allen this is not for hostname resolution +Attaching the stack trace for the monitor deadlock from a stress test which was seen with Cloudscape release and as a result the synchornization code was commented +check it out. +Thanks Leo! +Alexei please could you send an updated patch to the moved? Its hard to see what you changed since every line is replacedThanks +I cant install the groovy ve attached the error I get and the contents of the error log +Good catch about entityInterceptor Ive added that to for Spring with respect to event listeners Hibernate changed its setup approach quite radically and the documentation on it is very sparse +Ah! yes +Ah you deleted the old one and replaced it +Ill take this one +applied thanks +Otherwise well probably end up with merging Sylvains solution for +Thanks a lot +By commenting the various instructions you can see the issues described in the bug report +One planning technique when supporting multiple disparate customer groups with a single team is the create multiple Backlogs and then allocate a of the available resources per iteration to each group and plan stories from each backlog too have created a dummy iteration in order to do this +Issue resolved and committed. +patch against branch +We have previously agreed that this bug would be closed and FAQ with a link to proper IBM JDK bug ID would be created +We should think about how key fits into the schema +This is expected if you redeploy the entity bean and perform a finder call from the session bean that has an explicit reference to the previous version of the entity interface +But I found the earlier patch only solved the case for pattern y +Im now unable to reproduce this in CVS since the upgrade to +version of trunk patch which closes the created in Solrs +I think that the versioning iterator returns at most N keys is the more important contract for it to uphold +This is an upstream issue and I am not sure if it will be fixed in Tools +Attaching the refactored stax input patches contain improved test coverage for stax package sample job output validation +Verified on EAP release note status to Not required +ChunhuiIs this some how related with +I am ok with this change how about you +Actually Sijie had already spotted this problem +Updated the patch and added a docvalues based implementationThings to do Add implementation that uses +has been added with description fo and pdf to illustrate the attachment twice but cant see here +The only exception to the rule is the package which is necessary to bind some maven plugins into netbeans. +Anything we can cut out and still have something functional +I will close the issue as Ondrej asked me. +Streaming code does not set the value of Patch that applies to trunk as well as branch +I think I have a better idea of how this feature is implemented on the trunk +This problem is by design + +Weve got a great use for this feature +Is there anything I can be doing better on my side? I have been running tests on my side iwhtout running into these issues but Im clearly doing something is what I applied to +one more thing to fix bootstrap should use get and rename when write is complete instead of get which is +but if you like the music idea I can change it a bit for video presentations +As a matter of fact Im working with VS +The client needs the servers NEGOTIATE to correctly instantiate its SASL client +It causes problems +Section describes line break normalization and specificallynotes that the correct behavior can conveniently be produced by normalizing allline breaks to xA on input before parsing +Lowered Priority to Minor. +HAH +I like your idea for the wtpversion +Would be good to use some dependency injection tools to fix these kinds of problems all over the place +test case in zip again we do not guarantee the order in which form components are validated +At the moment I dont see a problems with merging if there are only a few lines to catch exceptions +Will have a fix momentarily +This may be an error in your code +Attaching source changes needed to publish on the Derby website +srcjavaorgapachecommonslangbuilder +Is there a plan to generate a more meaningful error? +Introducing package is sufficient improvement over what we had previous +So you need to set up a FACILITY for your supplier as well as a POSTAL ADDRESS +Well after more investigation it definitely appears to be a bug in trunk +Wed still need to specially handle the overall deleteupdate to do the individual deletesupdates. +Thanks Stack for your help +Initially a JCR session has to use that mapping unless the client decides to the namespace locally in its session +That should be the normal way of dealing with thisApart from that having an extra unique constraint is a good idea +Should this be resolved wont fix then +Resolved in trunk rev +The is pluggable +Yeah thats right +Thanks Matt Ive updated the +Isnt is a generic property not HDFS specific? I think this should be in common +Checksumming a length is just weird +I think goal of this jira is to improve performance of joins in case one of the table is big enough that it cant be read in each mappers memory but may be amenable to other performance tricks +Geotool trunkCompleted At revision trunkCompleted At revision Geotool branchCompleted At revision branchCompleted At revision closing all issues that have been in resolved state for more than one month without further comments +Either is also copied to the libdirectory or the manifest file of puts +It would be nice to give users CRCC earlyThe current formation +if necessary additional dictionaries should be plugged in +I manged to get it to work using the attached patchThanksDanushka +With the last accessedseen pattern therell be no problem with an expiry time of just a few seconds which would in most cases allow the garbage collector to be much more aggressive than with the minute time proposed hereI can see the advantage but this would leave the door open for some bogus polling client that keeps some very old revision alive which Id like to avoid. +Not at all keen on or +Trick is in which of loaders method +Nm found it under srcjavaorgapachehadoopmapredlib +It can just use better wording +What is the current release plan for ODEIn other words when this patch will be a part of ODE stable release +Not much left to finish it +Matt in the comments of the Ranchero bug report you reference your RSS feed which is the old Roller RSS feed +These need to go in provide a patch for +There is also a fix in this patch which is not in the previous patch to fix the juddi sql create srcipt + tested and Released +I have been unable to find the time to look at this and I dont expect this to change in the next few days +Looks like the title of this ticket was a bit misleading because when the JAVAHOME env variable is wrongly defined we have the same issue +then it manually tries to complete the compaction by moving the files from the tmp and write +Here is the patch that will first convert the parameters and then perform the entity field type check +consolidate these into. +Solr accepts XML not XHTML +This does not affect these machines +This is more a question than a bug +From what I understand this problem can be solved using assembly binding redirection mechanism? Am I wrongRegards close this issue +This patch contains the new class +this is either an ASWTP server issue or something is missing in runtime servers should definitely not cause or any error for that automatic detections should not call the server but actually what it is EAP or similar if that is the case that is separate bug +Please verify. +What would you like to see in the way of a test case to reproduce this issue +check TODO in submit in section dealing with tovmhostinuse +I wouldnt like to get too far to realize it shouldnt be like thatSo what this patch does +Still tracing +Please accept the original patch or find a better solution +Applied to branch. +Ive run various test for Request Transfer and seems good however there was one case when I got an error flushing the looks like compaction happened before the file transfer could be completed on a request +Konstantin said A unit test would be hard to write for the case since there are no valid ways to reproduce the conditionyou could a functional test with a vmwarexen image though it would take a lot of work +Reopening this issue + +For example I might want to have a more accurate setting for data outside a users home directory +The parent module which module which does not I update to isLastfalse all rows except the last one in snapshot tables +patch file with iterator an offline discussion with Keith I added some utility methods to copy key parts +sujacould you describe more about your issueare you writingreading same ledger? or lots of different ledgers? If you are writingreading different ledgers each ledger would have a ledger index file in bookie server +just deployed +Without the fix the unit test will fail showing that the block gets lost +it would be good too not run the tests for javac on jdk o only compile jsr on jdk +Refactored using SVD recommender +Removes the define command from +Why is it not possible to simply detect whether a is already valid for jbossws +patch to be checked in at trunkworkingclasslibsupportsrctestjavatestsresourcesmanifest +Patch causes checkin tests and mobile tests to fail reverted until solution can be found +The could possibly be refactored to use File instead as you say it might not be as useful +Thanks a lot +Please open other taskfeatures for adding additional diagnostic information +Patch for +I concur with Andrea +I receive this back there was an unexpected issue processing your Codehaus Operations have been notified of the problem feel free to raise an issue in JIRA quoting reference additional information you can provide will assist us I saw the logs come in and Im looking at them are subscribed on a couple of addresses +Hey Eli if you have already configured Eclipse to follow our conventions why dont you send me your +Closing this issue as incomplete because the reporter didnt provide any evidence that this is a problem in Axis +End of story +Tests ran cleanly for me +When using rollSize and data backeting since no more data will be written into the old file the file may never be closed until the system is down or there are too many open files which is not the desiredexpected behavior and thats why so many people complained about this bug +I am not sure I agree that unwrapping in a Helper is such an ugly choiceAny way this could be moved over to translators when we move over to protobuf +What do you mean when you say rpc doesnt work anymoreThe server gets stuck somewhere +Suggest resolving issue as Later + has to have a specific format YYYYMMDD HHMMSS which is characters long including the terminating null +I will create another issue and a patch. +would be appreciated that you can run tests against these two dbs to see if you still can see the issue +This fixes spacing issues +BTW the flex branch fixes this always checks the cacheCan we fix this for trunk too? But I think this issue talks about trunk +In the clone should we clone one more token for the standby as well for ha configurationsNot sure I follow what you mean here +Since nobody commented or reopened the issue Im now closing it. +This shows the hang is in the middle of deployment +Also is this a problem related to virtual nodes specifically? Or the consistent hash impl in general +I think this issue has been addressed with the latest code + +Verified +For instance OOTB orums and Content in frontend are far less demonstrative that would be this posibility and they are there +Ive added an implementation to seampay as a test +For completeness this was tested with TB and tbconsole +so when the is executing the previous persist ops might be still +Id rather not force this breaking change on users so late in a release cycle +tarball containing a maven project with source code and unit tests for AFOR AFOR FOR PFOR Non Compulsive Simple a basic tool for debugging codecsIt includes also the snapshot dependencies that are necessary to compile the code and run the tests +It appears to me that the isnt getting initialized correctly after a restart and therefore doesnt include the necessary properties +It no longer does this via an iterator it does it via the free method on the memory cache itself +I had to take care of another projectThe problem is that I cannot even get to see the wire +If we are getting hung up on the concept of a Theme then maybe what we are really asking for is a skin support where skins are user specific and themes are still global +Modified to detect the running layout +Will try again tomorrow morning +The +Why would the content length be set by the underlying page? Shouldnt the user just not set it if they know the page will be further decorated +This has been fixed in +It would be really nice if we can include the change I suggest to and because we are very close to cutting the releases +Thanks Regards +This task has been tested. +Ive attached the patch +Needs the eventfd fix prior to officially releasing this alpha ami +Fixed in r. +I heard from the user that this issue was not due to a Tuscany problem + +Seems that removing is easier path +Im very sorry StevePlease revert srcApacheQpidInterop and apply this instead +ryan can you comment on what those yellow lights are trying to tell me +Thanks Juhani! +Im not sure what has to do with json +After upgrading to i dont had this issue again +This might fix refreshing for browsers that dont support iframe preservation on refresh +My implementation actually takes a in the constructor so you dont have to add all the constructors +Heres the fix +Thanks for support! +You may suggest this for adding into plugin that runs from the command line +I found a smallspeedup from using that approach on the client side and a performanceloss using it on the server side +This is a very simple code change so I dont think any new test cases are needed +settings +Only the work is being tracked in the STS this make senses exactly what I suspected but its a good thing you confirmed itThanksAlex +Hey Brunook i will handel itregardsi changed a few things +errorContent attribute defines the url of an alternative resource but not the alternative content as a text +Couldnt reproduce it on new workspace with the latest build +Committed earlier by Vikram. +So if he is root he has to be responsible if he wants to remove a SYSTEM node +in svn revision thanksdims +Temporarily reopening to update release note info. +This patch moves connecting logging in and setting up the comparator into a new method +Feel free to attack this one as well now that you are on top of JCE certificates and the likes +Ill attempt to post a screenshot to show you what I am seeing +Maybe we shouldnt throw at all if the writer cant be retrieved +I moved this JIRA to Common since the patch only affects the common ZKFC components +Im thinking of the cause of this can be since this function relies on a lot of concurrent force +Thanks Stack for the review +Ive added a general metricimperial switch to the udig ui preference page updated the distance tool and scalebar mapgraphic to use these as revision I did some coding on the related bug for automatic detections of units for the scalebar I need to check if this has to be done for the distance tool also +I just ran a simple helloworld action with no DB stuff set up and gem and got pretty much identical performance +I suggest you try running the test code again +Yes but we can now remove the NOTE +Enis are you planning to release something soon? otherwise I can work on that +It could be a Windows thing +Thanks Alexey the patch was applied to SWING module at r +Changing to Improvement +Thanks Nicolas! +Ive downloaded the and tried it out in a test application +If we need more specific support we should create a new JIRA. +Ah sorry about that +The code of the is not thread safe and therefore it cannot be trusted in multithreaded applicationsThanks for the fix +You have to use ant based build system as the GNU make builds are not supported +Normally it would require a separate issue but as this is really minor and closely related I applied the fix in r. +Applied in two patches with second being ff +I committed attached patch +Added service wrapper and activemq library to buildsdk script +Attributes for limit tasks to specific target OS are automatically implied to windows on wanting copy a file using Setup API +i think that querycore would be wrong +Moved to as awaiting fix in +This was caused by the abstract check not working for beans +Thoughts +Do not add this property to since we dont want users messing with it +Nick that is my concern +Committed with some tweaks. +Originally I marked this as minor thinking that the mismatch will get the IDE or compiler to complain +In your case since usr already takes GB it far exceeds the reserved limit G +The people in favor of keeping them are generally Solr users or developers +Trivial patch attached to correct the usage to what I think it should be +Hi AntThis issue seems to be a duplicate of +So cant you add a column to the input file + +any justification why possible future extensions are only applicable to fields but not nested fields +Also shuts down the main thread and tries to handle interrupts better +Same issue with exact same code as on your stackoverflow post +Robert the patch is for Mercurial and Subversion as well could you adapt the components +Does this mean that the might leak a few socket descriptors every time an fsck is invokedThis is probably the case +Moving to +Rakesh et alCan I have a or other concerns with this patch +Wow many thanks for your help with this NahiOn your two points Marshal format incompatibility Ok +Will take a look Ted and update here +I backported the a patch to with revision and uploaded docs based on that revision to the web site +I have the same errorClasspath entry will not be exported or published +When you have time could you update your git branchMany thanks +Jake are you up to date? I fixed a bunch of things related to +Committed in rev +patch did not apply onIve put that part on committedThanks! +Vinod and XiequimingThe combined patch looks good +I wont block the change but I personally disagree with it for compatibility reasons +The service definitely doesnt work with AxisAll the other were false alarms +need to fix +Could you help to post comment about how to reproduce this? JCA resource adapter will have the import statement in the generated code +Fixed in . +I thinks this problem has been fixed +Applied patch in rThanks Mark +Everything looked good +its been around for over a year last update was months ago +Can we use Trash if its enabled? I think this is a good idea +Thanks! +For example an ELECTEDSTART event will always correspond to exactly one later ELECTEDCOMPLETE event etc +I am not sure if we have to fix it but i think We can show the hibernate state as a State Initializing in ring will that make sense +Did you manage to get a windows integration running +to commitThanks for your work on this Kim +Added testcase to test for the patch +If you want to work on this please get with me so we can talk +The change that fixes the problem. +Hi Mark Thx for later found that some dependent jar files needed to be in the compile time library +One problem is that this is shipped in so mark as incompatibleI didnt look at unfortunately but that sets the path to be relative to the binary for YARN is that an issue tooI think the fundamental point of this and is easy relocation of the conf by the installers do we have solution for that +Apache logo allCreated a simple logJ logo +Putting in queries also exhibits the same symptomsTherefore this is the same as +And why an user would have more than one session +User runs pig like thisjava For that user can refer to in these two packages without specify the package name +Now when using remove cordova plugin menu item the are no longer listed under plugins folderBUT removed are still listed in file and so present in Platform Properties pane of Cordova Configuration Editor +It was fixed +Yes can you make sure that is in the CLASSPATH +If you could send a pull request that would be easier +The new with the delegate classloader some more discussion we have decided that we are not going to replace the javassist proxy classes with weld ones +Thanks for the Ill commit this momentarily +They will open an issue on the same problem +Duplicate ofWas this fixed in? This seems like a same issue of what I hit in that already had in. +So Ive made them by hand +Committed to branch and to trunk +A couple places were we are using array list on the left side +Thank you for the review. +Yes but arent we still left with FUNC in the description +See also where Charles hint that setting a ref in XML is also not to me to remind me that my cleanup wont lose this fix in the merge +annoying that to write the specification for this feature be verified in gatein master and reopened if necessary +Btw for future reference we need to ensure that the code has right defaults for all configs +Summary moved expected failure cases to own test methods added test methods for null arguments added test rule to verify that correct exceptions are thrownI still dont like that some methods test more than one constructor +Internally how we manage it is not so important i feel +The patch was applied +I wonder whether we are just exchanging one edge case for another +This got already fixed in rev back in october during some code cleanup +Rainer and GabrielYou are understanding problems of ordinary developer +Why else should we support itMakes sense to me +Validation with DTD says the document is valid now but at the same time the current XSD schema makes the same document fail validation +debug as well as adding an explanation +new patch cleaning up the api a little more according to roberts comments +which returns null +Dont know how to validate this ticket Simon do you have any idea +Merged to with revision . +Javadoc plus some refactorings and of the javadoc +I will test today +Discussed offline with Michael and we decided to go with the first optionFixed in revision . +Looking good so farMany thanks Ill let you know if I run into further CDI issues +May be Bruce can give me a password to verify if the anoncvs is really causing the problem +It is highly likely we would KI this +Or another alpha would do for me +So the problem is not the nul points by themselves but trather why the algorithm fails to see the boundary should remain closed hereThere is probably a numerical problem here +Ill go through and follow my steps again and give a more detailed account +We upgraded to hibernate and now our application breaks at several places where we relied on the oldoriginal behaviour +I have created to address it + +Thanks + and could be used for the found that the feature was in the tools earlier but was lost when we did multiple hibernate version support +Closed. +msec transferTo +The Generics policeman will visit them and will hopefully help fixing +Why is this tagged iOS while the description is about Android +this probably a duplicate of saw in the release notes and incorrectly assumed the multi module problems were isnt much in the plugin to do about this +in every modules fixes it +Thanks Todd +And I stand by my previous assertion that there is NO explicit restriction on the content of algorithm attribute in terms of MUST MUST NOT REQUIRED SHALL SHALL NOT requirementsOleg +Hi Pete nice to see you commenting on re absolutely right about an example but this is hard to isolate from the complex code as it doesnt reveal itself in simple using the patch we made we dont get the exception so this should be a proof of the fact thatsomehow the flush method is doing some potential modification of the additions didnt happen in version and ll see if we manage to isolate the situation but that seems Seam shouldnt make changes of the map in this block or make a copy of it as in the attached are your thoughts +I have removed all occurrences related to Tomcat configuration +Committed revision . + check point file should write to CHUKWALOGDIR +fix typo in title +The file deletion is already asynchronous in a sense after which uses for deletion +This is bad parctise and should not be done in query rewrite +This patch is merely the test that must pass in order to verify that the bug has been fixed +Youre right its a JTB I dont even know what is JTB I reported here +Jiaqi could you review what is missing +We do not have a schedule yet for the release +Thank you Clebert for update. +This is the next patch under this umbrella issue +Also Artifact Type is now a combo list +Thanks Patrick! +Sent new docs to Emmanuel +Im looking into whether I can fix it. +Similar situation happens in transactional mode +Then I can replace +doh sorry +Closing. +attribute as the preferred way for specifying xpointer fragments +added doc outlining usability and instructions on how to use is DOC file for better readability +Closing all issues created in and before which are marked resolved +simple eclipse move +I think people are better off implementing their own custom password validators +Oops I did not intend to close confusion with another issue +It didnt solve our problem though maby we are doing somthing wrongThe generation of the qtags is still done with ant +In my run of full test suite all tests passed except for and in hcatalog dir probably due to only happens in case if theres an actual failure underneath +Since and will not survive I am closing this as resolved +Unit test that demonstrates bug by failing an assertion resulting from the duplicateunexpected subscription propagation +Code wise it should be sync with now +I can test with a snapshot but I wont be able to release with a snapshot due to the Eclipse IP Policies +It seems we should remove this from our site this is definitely not true +And a bootstrapping coordinator does not imply this right? +Has anyone seen this recently or can we close it +Attaching a patch that makes use to handle the output from the subprocessOnly tested with so farPatch ready for review +Assuming that the kind absence of feedback stands for an approval that the most recent fix is working as expected +Here the patch for branch revision Committed trunk revision Committed branch revision Thanks Mike for the help! +Had to remove that archetype as it contains no camel context file and it was broken too +I dont see anything Spring could be doing wrong the problem still occurs for you and youve the action class implementations already then please try to reproduce this against a current Spring release and submit a corresponding test +But dont you think that rather than going with a synchronized we should go with by extending this way as the concurrent usage increase we should be able to get good performance +Can be reproduced in Metamer deploy Metamer and open httplocalhostmetamerfacescomponentsaj set disabledtrueWorks fine in +What sort of backwards compatibility are we talking about here? Other mojos extending these mojosIf we copy the classes then when would we remove the deprecated ones? We dont want to have two copies to keep in sync +We might still want to use value but lets make sure we do that with he future in mind +several fixes made now update site appears to work except the Cas Editor plugin doesnt seem to be Editor plugin working now. +Thanks Shai +Hi BrianSo far I have applied all but a handful of the updates that you have attached to my workspace +I tried by putting all the required WSDLs XSDs in to the same folder +If anything I will make sure that the distribution has this in place for. +This may well be it +Please commit when you are ready +It doesnt much matter which system they run in as long as the tests get run. +So this identifiycation is just a dublicate entryCheersSascha +Hoss Man generated the factory for the filter and committed it in r so I think we are good here +I no longer use +This ticket is months old +Marking this resolved will be in CR +LGTM +Kannan Yes +In four runs they pass every time for me now +This feature isnt necessary! Im going to close this +They seem to be implementing quite a few features in CF +Can you elaborate on general rule of OGNL usage with base interfaces? The new version of OGNL compiles every single ognl expression to executable bytecode +The concepts of RPC and REST seem too far apart +Fixed with revision +We upgraded to the lastest and this is now working + +Just wondering whether youve been able to make any progress on this +for the patch +The workaround is to merely reload the page +Done +What do we gain by just having that hash there if its not used? The hash is interesting should we turn it on by default +Is there any way I can help to move this fix to? Thanks +Wont do length as if you dont specify lengths for fields in some databases like Postgres the length can be and this has caused problems +Is there a way to enhance an existing unit test to catch this situation +So I am adding it back in to match the documentation +However weve lost a replica so the data is underreplicated and we dont know it +Fixed with revision +This patch adds a class which captures the metrics in a moving time window +please file an upstream issue for EAP changes in paths +Assigning to Ignacio who is working on this as part of the Google Summer of Code internship program +Nothing more I was just anticipating any comments +Also updated which used to set a number value for the socket test +Ill rename all metrics to proper names +Were getting close to a release and introducing these changes would be too disruptive at this point in time +I wont have time to get it into which is why I marked it as +Update to latest trunk which had a conflict + pretty good +In trunk we have a number of nice features which I would like to bring to your attention +Added entry +It is impossible to persist a transient parent with a transient defaultChild if both sides of the association areAdrian your issue is unrelated as it does not involve +Teiid is definitely missing +This adds some asserts to the test cases added in the previous submission +Any progress on thisWhen can we expect timer clustering workingPlease give some light +especially because these info servers are exactly same as others running in hbase + You are correct Yes Sure Ill fix actions would you dictate I take for and ? The reason I put it as a smaller font underneath the composite is because it is not a critical error and is not important enough that it be placed in the title area section +Since you are past the problem I assume that this issue can be closed as well? +Attachment has been added with description correction +Thanks +Is this something we want fixed for +When you say We moved the same execuatble to place where we have xercesc library without this ICU option +valueOf methods due to unnecessary syncronization removing +So as long as there are no breaking changes I am content to see this stuff inI wonder if validatingParser rather than schemaParser would be more descriptiveI guess we still need to work on a validating parser that reads the schema specification from the root node rather than being given it separately +Why does Zoie even retain readers? Why not keep only the current one +Is this the same issue as described in? +Verify fixed as of svn rev. +This is a bug indeed +There is no easy way to fix this in interpreter as it needs a severe redesign in its code +The bootdelegation issue was a red herringThank you for all the help and sorry for the misleading questions. +Patch committed to trunk. +Hi JacopoWas just reading your comment by chance +New version of release note +Example eclipse template I dont know how to make this template overridemy address author lta hrefadriangtAdrian Brockltagt +Use on branch +Nevermind I found the missing file in original patch +The change you suggest would resolve the particular use case I have +Exposes so needs additional test to validate functionality on all protocol versions +These are the two methods I included in a description earlier +bulk close for +Can we do the same check for both source and target +i backported changes thus far to stable close for +with find by a combination of buidl changes and mysterious fixes to jbas +Rails requires or later but is the last version that can successfully install gems on SolarisWould love to see this resolved as we are totally stuck +Marking High Value fix +A sample wsdl a sample ant file that will build a stub from the wsdl and the source code ve left out the contents of the lib dir it just contained the jars from the for you using? There was a bug in which did this +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian +bq +Logged In YES useridthread dump is attached +Do we have Windows jenkins slaves yetNot yet +You must set this properties httpYOURCASSERVERNAMEYOURCASSERVERPORTYOURAPPNAMElogoutservicehttpYOURAPPSERVERNAMEYOURAPPSERVERPORT httpYOURCASSERVERNAMEYOURCASSERVERPORTYOURAPPNAME httpYOURCASSERVERNAMEYOURCASSERVERPORT caskeytestgrails +Thank you again. +Sounds good to me +skips sending entries if the znode exists +OK +Looks like this is required to for signature to pass verification +I believe this issue can be closed +OK looks like the does not override defineClassTrying with loadClass instead +Thank you for the second patch its been appliedDid you declare any roles in Jetty +corrected the distinct check to make sure it is across just the columns used for that join. +Page up through a file +Its different from what I posted on RB because some other unit tests needed to be changed and only figured it out when running the full test suite +Proposed SOA TP Release NoteError message for Notifier is not longer misleading if exception is thrown in Notifier class constructor +This solves the issue of an erroneous link is create to a is still an order issue with Maven +I think Hairong mentions that already has a fix for this one +Probably its better to define pools as a +Hi guysAny updates here? Would be great to have this patch committedThanksm in the middle of updating this patch and plan to post the result in the next day +What is the groupid based upon? Its hosted on sourceforge project but all packages on code say +No review necessary. +This is especially true since two test cases failed on my dual cpu windows machine while running the tests under Ant but ran fine under EclipseHowever now that the tests are passing locally I believe the changes are doing the right thing but may require some parameter tweaks +again throwing the same +The class has been modified to correctly read and write modified that encode the length as a signed integer which removes limitations previously set on the length of properties in the Messages to encoded bytes +Ill try and test the patch this week +But probably we should do better in this case than just throwing a anyway +The problem is only seen if points to Foundation librariesCommitted revision When this test is enabled as part of a suite we should make sure its not run on Java ME + +jar files from the plugins directory +Works now for any protocol as long as the target service is Stateless +Patch to fix the issue +Keeping it out of MR is a good thing but again deployment has an extra step +This must bring bring all these together +This hopefully addresses the coding convention related comment +The installer thinks for some reason that this schema is the current one and just does not dbsetup upgrade at should probably use an increasing number as db version instead of LATEST +Done +BLOCKCACHE true for info in metaandBLOCKCACHE false for to be working killed my ability to import until its side effects get resolved I put some data and flushed via shell then confirmed that the region server was using blocks of block cache where the region was storedso +I see no reason to store or use the ever anywhere +New patch for this issue attached +For instance logout does not exist in +Thanks again Dennis! +Its easier to just change the code which generates test data +Actually talks about the problem im seeing with bookie failures and publish +I plan to commit shortly +Hi Robert could you confirm that you dont have any kind of reverse proxy mechanism in front of Sonar web server? Could you also attach to this ticket the xml configuration file that you can generate from the Backuprestore service +Do you need direct io on writes to avoid polluting the cache or is it like our row cache and only caches on reads +What about homecvs and homecvsmetrics +Let me do some more testing especially from the UI and I can let you know for certain but this certainly looks very very promisingIll give you a final response by the end of today on the status +At this point it might be more risky to revert the changes as Ive completed regression and things look good +Could someone validate that it fixes the problem from behind a proxy as patched +Depending on when that string is returned to me it may or may not be in the CR tag +Committed at subversion revision +OK anay targetsgoals in the are being displayed +If I wereyou Id just remove the propertyId mark this issue as invalid but Id like a second opinion from the team +Verified on +Bumping up fix to version +fixed in shipped in +properties file +I think the meaning you give is rather obscure +Adding the progress callbacks for download would be great! If you can it would be great to get them in before the next release +No +The patch works as expected +Will test and commit +The easiest way is to attach a complete test project that shows the reported issue +This patch adds a Tab for Children Tree +I intend to put up a new patch for this by the end of the week +Create a close to release do be including the cheatsheets and other this shouldnt depend on meclipse presence will that work +Hello FredDoes this means the problem is solvedThank you for your answerJos Jos FerreiroEPFL Communication Systems engineerThink little goals and expect little achievements +This is simply what the Java IP are returning for us +nobody cares +Besides it would allow isolating the execution of the event handlers +Most likely what you suggested lack of imagination when I was thinking about it. +You might want to try with the latest as well +Please point one of the CSP instances to the DB on janeway +Can we set up a repeating job for this failure or is it not worth it +I was trying to solicit some opinions on what the best approach should be +Patch coming +So the problem start as the entry presents in ongoingcreates +That sounds like a plan +It is not necessary to combine them together since they have different return object +Please review and commit +Patches applied under svn Luciano for applying the patches The search queries below should work finealbumvegasalbumvegasalbumnamedsc +Ok guys here is the first cut of this patch turned out to be bigger than I thought +the import from oracle appears to never returnAs the reference warns the default importer settings crawl all available metadata +Hi devs You can find two patches with this which is implementation code for First approach and which is a json sample for this implementation +In both cases the editor gains focusIf checked on a node freezes it in the view +Im not sure if hibernate jira matches this category or not +A version that actually works +This JIRA project is for the Enterprise Application Platform product and as such it is limited to fixes requested as a part of the product development and support life cycleTo file issues for the community Application Server project please create an issue in the JBAS or AS JIRA projectsPlease open a support case referencing this JIRA ID if you need to discuss this issue further as it relates to our products +Unfortunately patch integration failed +Ive raised this to Blocker as I dont think its acceptable to enter the CR release phase without resolving it first +Ill work on getting this to work for arrays and maps since those are the most common cases +Thinking about it I come to the conclusion that using getResource is really wrong +Building all contrib modules works for me resolving ticket as Cant reproduce +Thanks Britt +This is the formatting suggested by David last have a couple of +Fixed as a part of +Im finishing the integration +changes pass locally and patch compiles properly +Attached the patch addressing the comments +If you keep references to class objects in static fields the VM isnt capable of deducing that this is a circular reference +proposed one off fix solves remoting issue but not XA reports modified patch fixed problem no data loss on test systemwe will roll the one off into the next has been superceded by. +Sadly Im fairly confident that I no longer have the implementation that exacerbated this condition +I didnt know the exact mechanics of so this is better than my solution which included the +by Denis. +I am not sure about the case when the code is working fine but you are getting in logs +Can you verify if this is still an issue with the M respin +all. +I abandoned because people didnt seem to see any value in itand I changed the way I use hadoop on ec around the same time +I think all those functions must use the trick of applicationjson or better I suggest to use a well defined and documented custom headerI can make a list of all those special functions. +Might make for weird corner cases +and couldnt find itTo answer your question stupidity mostly. +Thanks for spotting this issue. +Apparently the code was already wired to handle initiating a session with nothing to transfer but a if prevented that code path +But if you post the problems we will certainly try to help you outbest regard remove two of the attachments +I applied the relevant parts of the patch plus assorted fixes to check that reprUrl! null rev +And thus it can lead to consume more memory if your fill in more messages than it can drain +Invalid setting of the property +Would be nice if that one is released first before releasing the next +That way you could have a listner for sayany security classes a listner for any transactions for DB etcEach would be independantThats only a suggestionWere are going into a Release phase for but with our new releasescheme there is no need to sweat bullets +How does this relates to maven and theThis sounds like a GWT issue +Second patch for the new style to the Lookup forms +We dont care about keeping an accurate count only that once its done with that block its higher than it was before. +Yes I think we would need some additional Bridge methods to accomodate a better algo than the one I initially described +The attached fixes this problem +The test failures are unrelated to this patch +Closing +updated HTML docs +Im fixing itThanks +standelone package DOES NOT COMPILE actually +Ive just tried Dincontainer and the test failed as well +Totally agree with vychtrle +This error happens on the server side as well +Thanks Mark the patch was applied as expected. +Upmerging to the that can both start and stop the job history server +did not exist when the function first appeared +I prepare patch for some tests +the version numbers are numerous +It still think NTLM is a legal minefield and it is not a very good idea to have a picnic in the middle of it +Also add more comprehensive tests +Marking as Cannot Reproduce Bug +can I close it? +So constant caching appears to only be responsible for one of the two problems + writes anything that is necessary to recreate the object +Is this out of the release +My guess is NO + +Rob I assume this is not just for eap but for other servers too +See the last comment for enviroment details +But I quickly checked the man page for unlink on Linux and realized that delete of a file can fail when write permissions are denied on the parent dir the file is being used by some other process the file doesnt exist anymore or the file is on a filesystemThe current solution is enough for the st and th cases +The only problem I see with this is that we currently use to generate the Maven repo zip +So if someone try to do it future it is advantages to keep it like thisthanksAmila +I dont agree that we should never cache an empty list that would have pretty annoying performance impactIve committed a variation of Adrians patch in rev in the trunk and in the branch well there I got a conflict. +A test case and a suggested fix attached +Stein I am just trying to integrate the patch provided by you with thepolymorphism preview as offered last week +Cf my note on Performing the rebinding block only for actually also solvesthe issue as well as a variant query which still fails even after the fix for +Any way this gets in soon in? Its a serious bug with and OOC graph and the attached patch fixes it +Net +Ok heres a new take at it +But that could be because we have untangled that messy piece of code +It would be nice having the general one what its not clear to me is how binding parsed elements to target graphvertexedge it would require factories maybe +Patch forTested with Suns JavaUses a special factory for cloning Readers some implementation use reflection to gain access to private fields in order to reduce the need to read and copy a Readers content +initial patch that I manually tested +Superceded by +Thanks Laura +And that will change in future WFK technical reason why it should stay as is that Tools still doesnt support Seam and Arquillian Test generation +One thing I would say is maybe update the README to say if you already have the antmavenforrest source tars and you put them in usrsrc it wont download them again +I doubt that the use of NIO will automatically result in any performance improvements based on my experience with classic IO and NIO HTTP transports in +Closing all old Resolved issues +Thanks for the changes +looks +Marking bugs with invalid fixed for release as fixed in. +OKCan you say specifically which branches you mean +Do your processes contains async stuff? Ie +Im in favor of the method starting with join for users to easily find it +That makes perfect sense +dibs I assigned this to me and should be able to crank it out +Migration can be done via backuprestore process +Use echo CD +If you are using as a front end Aegis has no control or influence over these values and in fact you cant set themIf you are using the simple front end then I suppose that this might be expected to workWhats your situation +Thanks for clarification Ethan +Thanks for your patch Tomer! Your approach certainly seems correct to meThe file Im attaching has your fix to and alsoa testcase similar to your example that fails before and passesafter applying the patchIm planning to commit this in a day or so +Theres a small issue on the analysis page +This chain creates a cookie by default +Modified test cases to catch the actual failure points in the code +Another option could be to make the source folders stuff inside the builder itself +The issue behind the rev patch also highlights another issue with the clock skew concern in general +I updated the fix version isnt a bug actually but rather a limitation of That configurer is simply only designed to override literal values and isnt capable of handling complex objects or bean references at all +Objs +I have to correct my statement above +I havent evaluated the impact of a synchronous shutdown of application context but I feel it is currently the right way to go until Felix changes its locking mechanism. +The at r moduloPlease verify it fixed the build. +None of this is necessary for Mikes improvements to be used of course +why must parent be merged with target platform buildThose two are separate constructs where parent depends on one or more target platforms +In other words whether the tested file exists or not doesnt matter +Should have tested before making change +The documentation has been supplemented with the attribute listing anddescriptions from the DTD in addition to the the other infomation requested +Yes I do think we need a separate feature for this +All set +bq +Thanks John and DarioCould you also provide tests for the xmlrpc modules +The reverse problem Turn off Link With Editor toggle in Model Explorer Make sure NOTHING is selected in model explorer Open a Transformation Diagram Select Source Table Select Preview in Diagram Toolbar Data is returned Select Preview in Model Explorer No data is fix to is really the same as which has been fixed by removing the action from the toolbar. +this is the patch implementing devarajs coment about host resolution +The writer is already protected by a in +Another way is to rely upon what Ivan has proposed here. +Looking into this for +The model object added itself to but update logicchange in this called infinite update loop +All other changes remain in the file +To start with we will imitate what JHS does throw up list of all apps and do the filtering client side +One option is to add a event which has a veto method which may be used to nuke another extension +If a user is permitted to create a file while another user is +Hi RomanNew release Hama will compatible with Hadoop and we plan to focus more on compatibility with Hadoop in this yearWeve already implemented yarn module for Hadoop but not tested enough on fully distributed mode yet +It is pretty trivial I agree but saves a bit of time for me anyway +Though this is an HDFS JIRA the fix is actually in Common +xml +According to the spec all properties files have to be encoded +I used an unchecked exception but we might also reasonably extend for this I dont have a strong feeling one way or the other +I will validate that it resolves the issue on Monday Mar any update on this +Symbolic links are essential +passes for me locally +The test is failing because the value that it is testing against uses Windows file separators rather than platform specific ones +This property defines a list of variables that need to be extracted from the and set into jBPM context for the particular process instance +cool jobthanx a lotMittie +Not a bug in the parserJust happens that the error reporting by the parser is not accurate +That should be solved now! +committed v +Would you please help to try it +Reviews appreciated +AwesomeOne final remark is that imvho its a bit icky that changes to the list passed in effect the writing +In principle yesIn relation to my latest contribution to this may be removed in future when Queries and Filters get the same! But for easier understanding the term of should be kept +edit memstore +could you attach the generated and a log with running mvn with e +However that is just my preference you can keep it if you disagreeLatest patch looks good otherwise +Also removed redundant readme +The is a start of an implementation of this though more work is required to get it working working using the reliable protocol +Hey Erwannow we are ready for comitt +We just need to create a simple example to duplicate your problem +I just want it to go away because I did not say anywhere that I want to use it +Patch for incorporated inThe patch is also applied in Hudson builds +I think it would be fine though to add a releaseNote to this issue if you think others may hit the same problem but I wouldnt hold up the release for it. + +pull request +Canceling old patch I will submit a rebased one +This is due to CSS in browser and is not fixable in jboss portal +if svn of thrift works for you keep using it +Basically it just jars the classes folder substituted by meclipse instead of jar from the local Maven repo +The problem is that QDOX doesnt support +I think this is tricky since we only want to seek for the first region and not the latter ones but i still think this is doable by comparing with the current regions startKey and the Scans startRow +Also attaching the html output for these changesThis patch does the following Adds a page describing the Basic security policy +Its the foundation for a bake off between different log formats for example vs +Changing a runtime name also changes its ID +It is also possible to cache Class into clss for KVector types +Im hoping that everyone will assemble their files using the features packaging so the dependencies will be coming from poms anyway +the order isnt all that important as the dependency information is used but in a build that might be a far as I can tell the Member Projects are always listed in alphabetical order +There you go +An impl of chained filter using is available in Lucene misc for inspirationAFAIR I had to write my own but looked at this one during the times +The issue was not flatten being set +hoping for a working next command without having passed debugOr at least an extra warning this is known to be broken for stepping through please pass debug or what notThanks +I will revert the changes. +This sounds like a duplicate of bug . +Minor formatting request +Issue closed with release. +Its all set now +Hello HCE DevelopersWould it be possible to post the most recent stable version of HCE for download? It would be even better if you could continuously push your HCE code changes to +It makes sense to have u option to feature +i originally set it to just reset to and updated and deployed + +Thanks Ahmed +The JSP compiler now looks for all +The semantics ofsplit are wildly different in the C and Java Java it maps to which takes a REGEX as its argument +was but its not its main purpose +straightforward hardening of teardown logic +with Unit Tests ensure these cases are covered determine if bug is present. +Today we have x that many hosts and the load on the nfs server had gone up by x so this setting was no longer sufficient +BTW our repo is not cleaned at night +I am using version sourceKindly suggest what I need to do? +The patch should also include a fix to +Can you open a separate jira for integration with sql explorer submit a patch and provide steps to set up hive with that application? Given that the Jdbc interface is so big we probably need to implement a small subset of the interface to enable several applications +FYI I tested in Websphere and Websphere by default appears to decode the uri using +fixed thanks +By changing those two tests failed +at least long enough to cover common transient failures +closing +From what I understand the approach is to dedicate a disk per journal daemon +Ill let someone else close this my suggestion is to leave it open until the JIRA system bug is fixed. +Verified on stage. +patch +Patch applied in revision It could even be slightly simplified +Unless its somehow related to the issuesI dont know the first thing about CXF so I certainly cant help with that +Moved THBASE and TIHBASE all to srccontribtransactional +Is this correct +The representation that strings are read into is the only remaining incompatibility +Patch which is working for us includes the first some additional information after reading javadoc and the spec for the resource adapter this is the normal behaviour +Thanks again thats working fineTried with FF and Opera +No classes were made public +Hans the Dutch files seems identical to the English one +Attaching the patch file and html for this update +but the class doesnt exists in the repository +What do people think about changing this in? for doing that +Approved for. +This situation can be tracked inside the template parser and thats the way the fix was implemented +The auditor also does not provide any functionality in the critical path of client calls +Thanks for the update Daryn +If we want to do something more it will be better to track it in a fresh issue +The current implementation causes each configuration to be completely ignored if the system property is set +Here is v +In this case mutators should throw an exception +for previous version not for commit +If we change it we should change code and the doc together +Added validation check for target project resolved Release +I have tested almost all for and I didnt notice anything wrongIm traveling now and I wont be able to do anything until Monday so either cancel or improve +I resolve this issue +Which version of Seam +Duplicate of close of all resolved issues prior to release. +This patch omits the changes to contrib which were unnecessary +Sorry but I cant find this issue fixed in +Your class instantiates an with only sample pointsCould you test whether the problem persists when increasing the number of points +If you change the uirepeat to an ajrepeat then it works fine +Committed consolidated tracker in Rev +Warning users that you dont plan on promising back compat with experimental warnings seems like a good idea to me +Not sure why the message gets copied twice when it passes through the connection and consumer +Maven s API makes it possible to pass in the remote repositories the attached patch works with Maven +So will we be able to use Compress without XZ after this patch is applied? This is unnecessarily bloating my Android project which does not need XZ support but Compress crashes when the is not bundled +Thanks for reporting Gavin will look into that tomorrow morning if I survive +Ok at least I found reason of this problem +I had tried adding the data source to as suggested above but the data source was deployed after the modules in +This is a known IE bug +The is to dereference graphs resource are not graphs +test script and this is not specific to +Note that I had some trouble getting the specs to run in OS X Java +Thanks JianMarking this incompatible for alpha release users. +Removed legacy code for running from ant source build structure +maybe this will be fixed by a future ejb release +Im bypassing review board as this is very simple patch +but shouldnt prevent installation should it +As for my proposed alternate strategy for resolving timestamp collisions +Check out to see +its a +We will have a look at it as we are currently reviewing this part of the code +style nits missing space after in s constructor definition and its usage please reorder the param in the javadoc to match the order of argumentsotherwise whitespace issues +Please go ahead and configure the website +srcjavaorgapachesolrhandler +Well at least we now know what the problem is! Thank you for all the advice and knowledge to track it down +I think it is because in we only invoke some of the rules +This is fixed on trunk as +Done thanks for the Thrift help Stu. +HiI am also seeing this problem on a Jira instance +I just committed +Only issue is that now on this article you have different versions of picketlink +Done at rev I had to modify a lot from the original patch and put all labels at the plural formAnyway thanks Vikas! +If you have anything to start with I would be glad if you could share it with me to make my own tests +Fixed for the NPE +In order to connect to Apache DS successfullywe should create a new directory instead of a default ouusersousystemougroupsousystemand put user info there +Reopening this issue while I investigate some test failures +And I see symlinks created on branch and also trunk +Thanks a lot for the contribution Jon. +Thanks! +Why not use the default compile scope for this +new patch attached +The test that is failing was passing the partial job in to compare to webservices output of full job +This patch makes it so the vast majority of fields on are no longer used directly +If my theory is right Ill either get a no header error or a header with updates if we write one at the top of the file +Same package with different implementations cause this problem +And add sleeping code in opening to enlarge the proability of this problem +This patch removed the nulling out of the application connection during cleanup from a lock timeout or lock deadlock +Added Rolando as a watcher maybe he can comment further +srcorgjbosscache +More information Im using bin as downloaded from the website +bq +This is stored as a cookie +What credentials does that URL expect for access? +We should have a test which shows the problem you m being pretty strict about making sure every bug is reproducible in a test otherwise we risk getting regressions down the line +Assigning to you for PR approval +Close issue +I am suggesting we promote these classes and the PB definitions from HDFS into common and then use them in YARNDoes that make sense +Resolving as invalid for now. +I think the automated creation of the would create more problems than it solves +I applied it compiled docs and made sure that the links were all working fine +I cant find any official documentation on the convention but it looks like excel CSV format does not support null +Yes you are probably right +fixed size you could see some perf hits +This is a regression in Derby +didnt realize there were conflicts but its fixed now +I have this bug also on the simple project from the Getting Started Guide +Some relevant ones are placed into the file thus influencing the build as well +Committed at subversion revision +Mircea if this is really the cause of are you gonna revert the fix for it +This is fixed for jbds +I wonder where JBM takes that information from +BTW I set the status to critical because this is holding up our WCF interop testing +We shall clearly document it +Does this duplicate +Sanjay I completely agree +Committed to trunk. +I will put this into my +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Ive changed the simple sample to show off the new formbuilder which requires a formlayout +possible patch is Anurag! +Note below the text Create tree with HTML way and use selector tree is not displayed as in the case of html screen you installed dojo first using grails ? This sounds like an issue related to dojo and not Grails itself +is customized with an RBBI file for latin that only breaks on whitespace +The query that cause the error will be helpful than generally looking around where we missed it and it might be that there is a udf which doesnt do reporting +checks response code for checks null for unexpected response code sets for write operations moves the new test toI tested it manually +fix applied to x branch and trunk +no longer exists +I got something working but dont want to commit it for a bug fix to doesnt use +Finally lets remove all the tab for the sake of code readability +wouldnt it be more useful to have a file name where the log is written to? whatever you preferangela + +We avoid the OOM just by modifying the startup scripts if JAVAOPTS is not all issues that have been in resolved state for more than one month without further comments to closed status +Any ideas? Bad db setup? Not registering portlet apps with the portal +enabling XO let the app run of course just thought to mention that in case it might be related to the original issue Ill let you know if I get the crash again +Well I tested this on and I get a lot more errors than I did on my F box +Should we output the statistics via JMX as well +Decommission uses the callback forUpdated with the change to remove bootstrap sources from hopefully one more update +The simplest solution is to ildasm the dll sign it and then use ilasm to recreate it +Removing keys shall not affect the balancing of the tree as it can be easilydone by making the boolean end at the leaf as false +Is it possible to reuse more of the code by refactoring into static methods etc? Similarly much of the setup code is duplicated from +Yes the CLI copies the files to the correct spot +Tested on +Thanks to Ricardo Veguilla for the patchFix applied along with a unit test and removal of an unnecessary Routebuilder import inCommitted to the Camel trunk as revision rCheersAshwin +Fixed by adding a resource listener that handles the project event +I ran times on and they passed +ERIt is sure possible to lock Ok button but previous req was not to be so restrictive warn if there is some inconsistency but let finish this wizard +Also toString didnt agree with the comments or implementation +lgtm +Thanks for getting back to us +has been marked as a duplicate of this bug +My cents +Closing tickets +This is great Mike +This is a classpath transition to closed remaining resolved bugs +Hey Ahmed +It does not avoid the problem of time jumping on the node where the master instance is running +Hi ElspethCan I roll in the other changes youve still got outstanding while youre doing the diff task? Or will that clash with the rebase youre currently working on +I have done that change already +Ill close this jiraThank you problem Mirek +Well a commit has removed storing scm usernamepassword in the s why this is we can fix it for the initial checkout but this will failed if the user choose to not use scm credential +But the main problem +Yes downloading the beta solves the problem +We cant avoid having the affect the accumulo processes so I think we should watch it too +Cursor holdability can now be set like socursorHoldabilityCLOSECURSORSATCOMMIT HOLDCURSORSOVER can be set like solocaltrue looks perfect good job. +thanks back to Ben so its easier for us to spot his contribution +Can you tell me which transporttransport classes youre using to I can work out where its missingThanks to for now +Agreed +My concerns are mainly this is a bad design choice to let the clients send a unique id to a server which has serious implications if the id is not unique +I made a few other changes so Ill use the patch Im working with +Applied update to remove the erroneous space +I am getting A decorator cannot have observer methods but at least one was found on seems that this was fixed within +Trivial +Added a spec tooSending CHANGELOGSending libbuildrjavaSending specjavaTransmitting file dataCommitted revision . +I dont have that project to hand now if it will help I can attached a list of the files etc from the project later today. +Also fixed some javadoc warnings +for maintenance purpose what is the version of d js +Ah I misread what you wrote +Patch to fix scanner metrics propagation when not exhausting all records in a to put write in finally v looks good to me +Basically Iwill just add a new class called say and move all the code in the current into this new class and make just returninstances of this class with the difference encryption properties +The format string is angry about s for the usergroup fields +This is going to require some discussion as its really an incompatible change now that is about to be tagged +Please dont mind my patch name is as it is a typo +The build process will change the connect role to admin +Updated patch incl +Alexey please verify +That would very easily explain this behavior because the session would be accessed concurrently from multiple threads +Im not sure +The local registry is also hidden by default behind the properties view +Verified on revision . +A set of spellcheck files specifying existing Chukwa options +I dont like this feature anymore its too invasive +I think that by installing BPEL feature also AS feature should be installed. +Moving feature tasks to LATER and unassigning them +This extra thread is not required +And then how do you uninstall them? Or are you thinking once youve installed a NEW version of Cordova you can never create a new project with an OLD version but projects created from an OLD template will continue to workIn any case we should consider this use case for the command line tools especially if those tools can mitigate this problem to some extent +I will try out the patch Guillaume and feedback asap although the regression is old Im sure we can agree this is the wrong behavior and should be what I needed to test my patchAnd indeed that fixes it. +these options belong in the and Atomic annotations not in an Property annotation +Makes the following changesM javatestingorgapachederbyTestingfunctionTeststestsderbynet Wires the test into the derbynet suiteA javatestingorgapachederbyTestingfunctionTeststestsderbynet The test itself +base on common ui colorsselection mechanisms when select the label figure the backgourd color change to be blue and font color to be white +Merged into the master branch. +on closing this issue as invalid +I think that adding this really small feature can be very useful +Fixed on trunk revision +More specifically the buildwebapps directory contains files that define the sitemap that jetty uses +attach an example that reproduces the believe this is a duplicate of which is resolved in RC +Defer all open issues to +but I can also see that many people dont like that we removed the source maybe we should bring the source tab back but keep the rest +jira issues that have not moved in a year +Ah right Ive noticed that flashing annoying indeed +If it is new with this patch what was the reason for adding it +a protocol? +Logging output providesJBOSSHOME URL fileoptjemsjbossasBootstrap JBOSSHOMEserverdefaultconfCommon Base JBOSSHOMEcommonCommon Library JBOSSHOMEcommonlibServer Name defaultServer Base JBOSSHOMEserverServer Library JBOSSHOMEserverdefaultlibServer Config JBOSSHOMEserverdefaultconfServer Home JBOSSHOMEserverdefaultServer Data JBOSSHOMEserverdefaultdataServer Log JBOSSHOMEserverdefaultlogServer Temp JBOSSHOMEserverdefaulttmpNotes the method getPath has been renamed to get to better describe function after modifications +I am uploading a patch with very minor changes to trigger the build again + for a generic job empty destination folder parameter so its reusable and it wont erase stuff to tested using COPY mode worked great +How can I find out that my service failed because of this issue? Is there a message logged when a client is added to the list +The original issue had to do with status checking the commits that made it into addressed that issue +Fix javadoc looks like someone confused confused this for when doing a subversion commit of the latter +Missing dependency maybe? Is this pluginfeature installable from the JBDS Nightly update site I wonder +We dont currently have any capacity for a either a VM or jail +Put it in launchpadtesting but not sure if that is really the right place +Thanks Jason +looks good +Compile and runtime dependencies on all Grails classes need to be removed from the builder +The prioritizing by longest match was already implemented but not working because of two bugs +It wasnt until we tried upgrading to that the issue occurredIBM released an update to a few months ago +I really dont think this is cargo related +Closed after release. +for Oliver here +All the refactoring I wanted to do is now completeWhat is missing is Possibly breaking up the url into protocol host port and path +Sounds good +Arvind sounds good +Im okay with this featureId rather fix the bug before we introduce the feature especially if get becomes the default since it masks the bug +Just for the record Ive applied the relevant parts of this to the trunk too +Switching all issues that have been in resolved state for more than one month without further comments to closed status +So I need to know from everyone which plugins have ACTUALLY changed since we released JBT and I can increment those yet if people keep their plugins and features current I can simply grab those values anduse them in the build +I met also this inconsistent behavior. +Actually we did start with the basic URI +I ran the patched on my machine and it worked correctly it also only ran the Hive tests once and didnt print the warning message +Is this going to actually get done in the next two weeks or should it move to +second build log youll notice that while the metadata is fetched again very few actual pluginsfeatures are this time +Want to have a look +Feel free to reopen it if you still have some concern +This should probably be marked as a Minor issue but it doesnt appear like I have rights to do that +OK +My bad +fixed in revision Please thanks +What have now is better than before anyway +Of course it doesnt actually do anything because the execution magic only happens if the class is alone in the file +Verified with CR +Patch available for patch +This is functionality I hope to get to soon in geoscript +keystore and +Roshan yes Ive moved the bug Jochen is referring to into a different issue. +ill take it anyway and well see where we end up +We didnt think mmap could be the cause but nevertheless we tried that once on a smaller cluster and got a lot of memory consumption again after which it got killedI can see if i can run one or two of the nodes with NIOFS but let the other run with mmap +Unfortunatley this is an upstream issue and cannot be fixed +For reduce tasks in the shuffle phase we look at SHUFFLEBYTES counter and so on +I should definitely give it a try and maybe even develop a plugin to ease ivy integration even more +Thanks for providing a test ll write a similar regression test case for this soon. +How does the table definition look like in the db? From the log it is doing everything right the CMP field id is of type +I believe this is fixed with the latest CVS code +can you please go throgh the patch and apply that +see +In cases where you dont have unit tests for it its also easy to forget completely +Committing shortly +Can you give me a little bit more information pleaseThanksSascha +Although we cant guarantee that an exception thrown in will be propagated we can at least guarantee that will NOT be thrown and that destroy is called twice +Who knows maybe well decide to create another examples submodule at some pointTODO decide whether pom structure of wikisearch needs to be modified or notIt builds as is so this can be investigated at a later time +Could you please take a few minutes to provide this text and I will have it included in the Release Notes for +Please go ahead mark this as resolved +Thanks for the commit! Ill try to submit patch updates to keep in sync with any C API changes. +Attached nd patch version which fixes all the above review comments +Replaced code with metavalue specific code. +some pseudo code that adds some entropy to the default baseIdOne random byte may be sufficient or the random could be munged with the hashCode system IdentityThe simplest and most reliable is to use a few extra bytes I think +Good suggestion Dag thanks +running + Thats a bug +Hi MatthiasI have tried the suggested fix +I couldnt reproduce the issue with include +bq +Not sure if there is a separate JIRA for is also an issue with PDFRTF generation with multiple WARS in as EAR +Half of the transaction has been commited in my final example +I have attached an updated version that also includes a bunch of error handling improvements +About keeping the response queues out of the cluster how can we achieve this? How will our application know to which HQ instance to respond to +Thanks Brandon +The same problem is present in FORWARDTOCOORDs +We detail the entire algorithm in the docs +Forgot to do svn add for two new test classes +Attached patch that changes the exception thrown when an authenticated user attempts to login to a workspace for which the user is not authorized +create seems to be the locus of the problem here +It depends on the patch attached to MPARTIFACT +Go ahead and merge it and we will do a release on Monday +Not suggesting this as a complete solution to either problem but maybe it would make sense to allow to mean an infinite number of retries +This is most likely the same as +while writing this ticket I thought about this +Still connection doesnt do anything with these methods currently so the above description still stands +Even though there is a performance penalty this actually is more of the intended meaning of the original patch +This is caused by and being dependent on different Camel versions with bringing in one version of and servicemix bringing in another ofLets await for to come out with the latestgreatest Camel and then switch to older issues for Apache since were no longer actively working on these at the moment. +Apply in modulesdocumentationxdocsaxisversiondir +I checked in a fix under priority to minor as a short term fix has been provided +Unless there are objections Ill commit this soon +Attaching patch again +If you have some ideas in this regard please file an issue +And in my opinion a simply is not anThat being said it dont think its a major issue +This fix doesnt work in Firefox all popups are rendered on the left bottomside of the screen and they dont disappear anymore +Ive made some more changes on the branch +There may be others that I dont know about +after release. + +Daryncould i close this issue if your patches will include this problem +Ok robertthanks for the feedback Ill ping themRegardsMatthias +Well test fails on interpreter on multi CPU system but works on MT systems +It does not however fix the case where it should be happy if theres a symlink that points at a file where lstat should succeed with no error +Ill do that +this looks like its left overIve never hit it since Ive not written a perl server +Thank you very much Charles! +Temporarily reopening to update release note info. +Removed unused imports and reformatted the lines to charsPreliminary runs of terasort benchmark on a smaller cluster did not show any performance degradation +This issue actually supercedes + for patch +What is our policy on thisThere will probably be one findbugs error the same one exists in the I based this off I dont think it will cause any harm +Hi NickOk no problem I will fix itThanks +Why should we do that if its not needed +thanks Bhallamudi. +So if someone else want to pick it up feel free to do it +I implemented the condition attribute on the shortcut element +Strange Ill take a look laterTell me googles IMAP servers address and port and Ill try +Patch is a revert of and then the wrapping of the tool mentioned description +When this is committed should we close this issue saying that form mode is now supported and open new ones to deal with the additional attributes that people find are missing +As long as you figure out the algorithm and blindly implement your own version you are not violating LGPL + +heres the fix +We could define one aggregated DTD and still reuse most if not all of the processing logic +The underlying bug that caused the failures is fixed in +Modified ant task to bootclasspath the right jar files instead of the not a pretty fix +Guessing fixed in roughly +Thanks George +I wasnt able to reproduce the failure with the patchCommitted to the branch with revision +If we get IT tests for these before is final we can take a may have been fixed in trunk i remember having seen a least since was fixed you can use the Ive attached the first time and the error is still present in and +Done for trunk +On Java we try to use the built in tool instead of forking javac +So far the can only add cartridges to an existing application +Im not sure whether its needed or not +This patch implements using of the transform contained in frcUsing of the rendering hints information contained in frc can not be done because of unimplemented Fractionalmetrics hint and implementation of Antialiasing hint +Running the nodes in different clusters is not the standard usage of +sorry for the delay +New version of the patch based on Davids most recent patchThe changes Removed the junit dependency from all Lucene that depend on the +it might be different because you used and the jenkins build is using mvn clean install issue is increatethe first line results in null in case of the abstract method of the partial bean t an issue in with still present with Weld andThe stacktrace is the same as in the second comment +This should be fixed by. +Issue resolved by throwing a in when no container has been found. +issues for released versions +Need to port the plugin to NutchIs this related toLooking at the it certainly looks the same but with a very different implementation. +Does the patch in work for you? +file +applied +Updated merge patch to include fix for +I take it this is on windows? Ive been running the for trunk for days straight +has been marked as a duplicate of this bug +Whats the best procedure JIRA wise on this? Reopen this issue or start a new issue? If we reopen this issue how is the bug fix tracked in changes +Lets see how and develop and make a call based on that +This test failure is fixed with a combination of patches Ive submitted on other issues and +At best it should be I am not sure if internally selects the system defaults +Do you plan on removing the support or is there something wrong with my implementation +Reuploading same patch because HEAD didnt same patch because HEAD didnt compile same patch because there was some kind of weird connection problem +Owen and I were just discussing a few weeks ago if we should always attempt a kerberos connection if the user has kerberos credentials +There is already a collapsable API URL section at the top of the interface +That is correct the patch depends on which in turn depends on +Itd be very easy to introduce an interruptible io into the backing store checkpoint or metadata updates and its almost impossible to test +The buffer does not contain compressed data +now that the patch I requested from abdera is in can remove the casting code +So do we want to do this +Nick Thanks for the explanation +For reference requires to load Rake so right now were about overheadWhat Im trying to do is take the common case and speed it up so Im measuring the time it takes to load enough of Buildr that it can actually load a typical buildfile and create all the various tasks you would expect to have and might run next +Thanks Boris should already be fixedplease test from svn headregards the calendar is rendered now but when I trying to choose any date or change the month I getting an errorDuplicate id calendar +The Jira default of No Component can be used to facilitate other requestsissues as one would expect the volume here to be low in comparison +The default connections showed correctly up in JBDS Beta +Ill try nowIf you mean the mechanism of defining global exceptions Ive of course defined to be forwarded to my error page but THIS exception is still not handled and thrown directly to the default handler which carefully prints it to the shocked testerI must say that in that part of application where this exception appears I do not even use file upload +Cleaning up code an removing double escaping of +Make as Incomplete +Is there any reason why get isnt synchronized in tooDuring compilation of I encountered another problem Why doesnt the ant build script download the neccessary xerces jars +Andy I am not sure I understand what you mean by crude workaround +BUT I dont get your assertion that then in other slices the query should compare with IS NULL +Attached log from running the test +Thanks patch in +Do you mean that it was already fixed before I opened the jira? If so where can I see the fix +Im willing to bet that he has a copy of Xerces that doesnt support in libendorsed +Exceptions area Object val in +This is effectively a utility library which is sitting in the YARN code base to help application developers +Uwe the index format did not change in +Still occurs in version tested it myself and it seems like the DSML export is broken +Only the test that runs in container gets the dataset injected +Can anyone shed light on why introduced this clone in the first place? Seems like a totally braindead thing for performance +I am reassinging this to Bilgin in order to track his contributions +And it should be generic to any repository not just messages in a particular spool and spool state. +Sorry for not replying about the version of the nbm plugin we have migrated to a platform app project instead and we are not having the problem anymore. +This seed reproduces for me actually so either i have hardware problems or its a real bug +After looked into it I found out it spent quite some time on splitTest +attaching a fix its much nicer if you could attach a patch file as its easier to apply to the different branches of code also I cant do anything until you resubmit your patch and select to grant the code to the ASF license +The new code pass all the tests on my box +Different stacktrace but same sort of problem +Now that work is well underway with AS all previous community releases are +Are you saying the source needs to be compiled to webclasses? Why cant you use war packaging with a custom directory configuration? The correct answer is not to hack the PMD plugin but rather try to get the war plugin to support your directory structure because your project is not a POM project it is a war project +I can imagine that being an performance improvement in some cases but in a case were many classes are created this will cause the creation of many many class entries +I confirmed that on gcc libwebhdfs does not compile without this patch and does compile with itTo test this I also had to install and to get this to compile at all neither of which is documented in +Will be great if You try this via Eucaools and show I check today documentation too and find this but work in amazonPlease approve my patch in attached patch for this +. +A service receiving a binary CAS replies with a CAS but if while processing the CAS it sends it to a remote delegate via binary CAS it needs to send the full CAS +Verified on latest build +Patch that contains changed detailed in issue description +Yes it will be just for Struts project just Struts should have access +Its not possible to include these changes in the GA moving to +After I deleted earlier this evening the component I noticed the still was using itI started cleaning up for this and then noticed some other APA component still were not completely clean from eitherOne thing let to another and I ended up cleaning up all the APA components including stripping out other no longer needed dependencies JSP tag lib uri cleanup etcAs a result Ive got now most if not all the APA components clean not just for Ill take this issue back to commit my changes and thereafter resolve this issue +I managed to reproduce the problem on one of our windows machines and found that the same tests fail +Updates to the module to take the new api into account +So lets keep itI think its ready to applies the same changes to backwards LTC +Some sample code to experiment with task timeouts A sleeping iterator and MR driver +Heres a patch for jclouds +Please use the users mailing list. +Hi Pieterit looks fine now thanks your mirror will beadded to the list in an hour or soGroetenHenk Penning +Usually youd want to separate the two getting the info first and then calling a method to presentformatpublish the collected data +I didnt realise that you had to rebuild each time you do an SVN update +which seems very weird +During building demonstration code everything was working fine +This reuses the container configuration for the WAR file +This seems a good idea but maybe another choice has been made before? If nobody see a problem I will follow Christophs suggestion +Updated patch +Committed to trunk. +Thanks for pointing it out +Not sure why we removed that except maybe that it would almost always be trueHow about we make the check smarter and compact if we delivered over half the hints +The logs told otherwise but because the data directory permissions were not setup correctly the database was initialized into a halfway usable state and left +Commit the patch to trunk At revision Thanks Han Hong error disappeared so close this jira. +Now extensions using the should work in +Thanks Navis! +I think the best thing is to bump this to because any fix we make is likely to break something and we should have plenty of time to flush it out +Behaviour is the same that in case with +The threshold can be a follow on +I will take a look and add it to the in rev RC +WernerBack on version the only problem was static initializer +Also added setting the default back in as that seemed like it was the original intention +Denis dont worry about the jsdt stuff +I committed the patch to trunk andThank you Arpit! Thanks to Junping for doing quick validation +also involve a series of to the responsible datanodes to determine disk idsKeep in mind that this should be one RPC per DN rather than one RPC per block +This is still very rough +Nobody has looked at H in any depth yet +A given server endpoint has all the assertions in all the alternatives added to it +any news? +The underlying issue might have to do with sessionId rewriting +Ha thats funny ok time to keep digging and see what I can figureHaving some time to work on this today are there particular unit tests that cause the OOM to happen maybe I can start with where the problem happens and work backwards from there the unit tests on trunk seem to be compiling and running fine on my mac +Will commit this one unless further comments +Not sure what I was seeeing there +Pls review +If we dont set default language setting users create a document without language setting +Im not talking about the features file Im talking about a Jira named Upgrade to Jetty It looks very similar +Attaching a template file for the release summary +Thanks. +Looks like this was fixed in Java +No permission to change my own ticket? lame +Closing released issues. +git diff it is not an incompatible API change since the old constructor already exists and calls the new one +Agree +Actually verified in testing closing now. +Code has been added to +Ill change to topic then +It sounds better I now have failures instead of Buildrinvoke should not force compilation if dependencies older than compiledBuildr should read passed specs from result yamlBuildr should read result yaml to obtain the list of failed specsBuildrinstall should artifact when from is newerRakeTask package should create package in target directoryThese are the new onesBuildrinvoke should not force compilation if dependencies older than compiledBuildrinstall should artifact when from is newerRakeTask package should create package in target ve never had any of those fail on my xp box +i tried wicket examples with the latest and everything seems to be working fine +Efficiently and reliably detecting the presence of a previous value is crucial also for not just for this perf improvement +indicated in the April JIRA triage meeting that he wanted to the patch +Also putting it in a global file also I dont thing is lean and clean solution since an appif not a module I consider it not to be part of the heartof the application server +Operations such as configuring appenders or setting logging levels cannot be performed with SLFJ +Further investigation shows the following facts +Dynamic typing is far easier +Did anyone try to use single machine but not with local mode but with nutch acting like one node? Maybe this is workaround till bug is fixedI need to recrawl about k urls and Ill report my timingCould you also try the patch on and reports if it helps for you +made changes to the esb config file and it redeployed +Submitting for failure is due to just committed this. +bq +Snjeza youd still need to inject a server or look one up somehow so the defaultHandler wont be enough by the current selection doesnt contain a valid server you can choose the first valid can also add a parameter to the command if you know what server should be used. +Bug is still in sure when it happened but i can see this is fixed in the master branch now so closing this one +It looks like all tests of the CTF suite pass without any problems + +I was using +Also included in this issue is a design discussion of what we would like the distributed versions of X to look like structurally +This is motivated by the same problem as +Hi thanks for the bug report +This is the result of missing quoting of the variables in the sendsoap script +settings for the magic project should be fine +This is actually just a symptom of there is a lot more that just that are leaking +Since the change really make sense it does not need another review +Committed patch as is. +ThanksI think in terms of Mahout I think my question is whether it supports REST interface or not +data +I dont like that this is actually syncing on this just for the swap purpose +Possible something got broken in the HA merge +Incorporated jbellis comments +Verified by Svetlana +When I looked at in Oxygen it accessed the DTD at and reported a validation error for each use of lazyextra +Thanks for taking this up +you cant do it with +OK Ive reversed the linkage in Hudson and fixed the bootstrap profilesNow we just need to get everything to build wo test failures to verify +The part for +Ie when I put one message on the queue it shows depth of when I put on depth is has depth etc +Committed to trunk with revision More patches will follow +Bug fixed in FOP +Please consider that when reviewing patches it can be hard to distinguish fixes from reorgs +Ryan thank you for your advice! It was helpful and I finally added a ClusterIn my case traffic labels werePublic vSwitchvmwaresvsGuest vSwitchvmwaresvsManagement Network vSwitchStorage vSwitchI have renamed it tovSwitchvSwitchvSwitchvSwitchand cluster was added +so doing Regions seems oddThe only reason on Regions matters now is because the constraint system is not expressive enough to allow the writing of constraints on and friends that actually enforce the desired semantics +Duplicate of JDBM component has been removed from HiveNo need to optimize this any more. +And we also need the docs etc +Im changing this back to a bugThe way test works is not expected behavior from anyone who has actually done any real shell coding +Also added a state variable since shutdown could possibly be called by different threads +Still You should have started a discussion on the forum +Committed at subversion revision JIRA was not reachable when I committed this patch +thanks a lot +The only thing I noticed was that the were missing the in the will close when they are verified through the sync +Transactions should never apply to the entire request processing but rather always happen in the middle tier or the web controller +For this reason most wont allow you to deploy the same artifact twice by default +I wouldnt go that far as to suggest we use maven to build C +I intend to do New format for the channel with sync marker support Mirror the events on write +Moving to + +attached first draft alpha patch for discussion allows redirect in render response with partial requestresponse +If it is not please reassign it to me +So Download to CSV will work well +Maybe Jeff could take a look + +Bulk close resolved issues +Yes but you want the feature so isnt it up to you to provide a patch +Seem only regex based move ant task could do the trick +T E S T SRunning Tests run Failures Errors Skipped Time elapsed secResultsTests run Failures Errors Skipped patchv for hadoop any if conditions is meet we will set available false and cancel the mergeSo elseifs is enough +Deferred until. +Eclipse Project with test case +unfortunately CND is not something included in the spec and its Jackrabbit specific +Thanks +Looks good +filed under the proper think this issue can be closed as sample container is working now +Postponed until it is more clear what extensibility is required +Is that right? Why I do these in my xenserver then the cloudstack stop and remove the vms there any details that I ignore +Checking the timestamps of the xsd and xsdconfig files versus the timestamps of the generated source and class files is an obvious step towards dealing with stale sources +ah ok +There is number of bugs already fixed in POM editor and you can see path to some artifact if you select in the list of resolved artifacts +Closing all resolved issues due to a successful release +Pablo remind me to buy you a beer next time I see you +The trigger for the bug seems to be autoSubmittrue +Default templates cannot be deleted When no specific configuration has been made for a given component the permission resolver will fallback to the that will reference either the default permission template or the migrated permission template for projects +I will use tika trunk till I can come up with something regarding the link extraction which I consider stableThen after this a release would come very handy +In case you feel like setting up jenkins nightly builds please contact us +Any plans to fix this? It doesnt affect any existing code and its only four lines of code +Perhaps you could give me a better example? +I pushed a modified version of Hiros change with a type check for in masterefdNahi I agree we should look into if its a good fit for +The workaround simply eagerly fetches the association +Sounds like you changed your clusters partitioner +resetting to original owner +haha no worries wasnt sure if it was lazy or just not having done it before so wanted to make sure +jacob added to jaxen group too +Looks good +In the meantime please deploy the attached in expanded form +It turns out that srcmainant has a dependency on and its not trivial to get rid of it +Must be some difference between my test environment and the environment invoked from the Makefile +Ill see if I can fix those tests shortly +Sounds very reasonable to me as long as theres no performance implication +This is probably why proxy classes are loaded in their place +Thanks Mamta! As for the change variable since this is old code Im ok with keeping it your callOr we can be more intelligent about dropping the trigger by doing so only if the trigger column positions have changed +Please work with Lee on this thanks +Ill mark that one as a duplicate of this +Patch suggestion Im not sure it will deal with the update and delete cases though +I need to say I do not have a clue +I really think this was just a simple oversight +I am fixing it shortly +Good luck is tracking it down +Separate Boxes We might do well to rework the shipping box code to handle the situation with multiple packages and have each carrier estimation tool only ever work on estimating one box at a time +makes sense +From a brief glance through the source the order would appear to be methods +Also fyi for some reason eclipse doesnt like your latest patch keeps claiming its not valid +It has been successfully tested here in Hamburg +Thanks IanI have created some themes with my own and for implementing my thinking but Im afraid that the ui template will be changed in future release of so I submited this issue and hope this feature can be added to +I suppose we could make the argument that this can tide us over but it will only take a couple of minutes to add getFields as wellI think Simon may argue that this will work in more cases by default where external queries would have to implement the getFeilds method +RichardCould you please upload a wsdlthanks RichardAs I can see the problem is fixed in current SVN can you please check against nightly builds +can easily be provided by UDF not adding function to core language +There are some unit test failures +Fixed in svn revision seems to be an ordering issue with this fix as per the new unit test +This issue is really about gracefully handling or similarWith as an blocker no problem pushing this to or beyond +One for smooks soap slsb soap +Thanks Tucu I have modified the patch per your suggestion to retry for a max +looks good +bq +Committed to as well +I havent tested with versions earlier than that but theoretically there is no reason it wouldnt apply to all versions of Pig starting with +runs fine +committed. +Its only when creating my own if I use a built in bean like the resource bundle entries for these are pulled out of the Spring resource bundle I have just fine +removed +There should be oneand preferably only oneobvious way to do it +I created for that +Hows your testing going apurtell +first patch was made modified patch for +Thanks for your contribution +himself +Ive also experienced this problem +heres an incomplete patch just looking at found a few wierd things looking aroundI think fixing all these sugar is going to be pretty major and a lot of workIn general I think we can accomplish this issue in a way later sure it will mean we have deprecations but I think this is better than rushing stuff in +This way I am not sure where to place the files + +Should the name Heap Memory really used be changed to something more clear like Heap Memory Active +likely worth documenting +That was why I failed to reproduce the bugBug reproduces in windows cmd consolePatch for fixes the bugThus this issue may be marked as duplication +no input FO file or output PDF file is transition resolvedinvalid to closedinvalid +author +will conflict with this patch I added two more methods to the in that patch +The variable that holds the result of the type flag was not being reset on each invocation of the is fixed in +cglib is now required only for and reverse db generation +could be simplified using mockito Javadoc could be more descriptive for incorporated the minor comments +Id still usecleanup to close the stream and drop the comments though +sorry for inconveniencesworks fine +I also brought up a cluster submitted jobs randomly and killed some task attempts and did not come across this string in jobtracker log +So this looks correct to me + committed to trunk and branch. +Thanks David I took your patch and Im adding tests right now will upload a new patch soon +Ive been kicking this around w ddas and serge and they suggested that +Do you still get these stack traces in the logs with WF Alpha +looks good Ilya please in revision . +and then we would only have two methods copiedacross and rather than across the current mix +should have made it into the common jar by now +Thanks so much Lily this is great and gives me plenty to get started with +tentative to address the is already ongoing on trunk +Reporters of these issues please verify the fixes and help close these +Patch for this problem is attached +Now I feel like a total idiot +BTW the javadoc Ant task outputs many errors on missing filesnames but that something for another issue +The only other option would be to explore other UDDI client frameworks that would not need Axis or could leverage our current version of Axis +WDYT +This patch adds the Eclipse files to the ignore patch doesnt workCan you make a new oneregards this has been done already I guess this bug can be per Jurgen Lusts suggestion. +No my post was focused on priorities +I see the findbugs will be handled by +Now the environment is being used by others +committed to trunk +Applied in r and. +I have added output to log +A test is added that only tests basic operations using AvroDefault runtime behavior of HDFS is not altered by this patch +Ive linked the two is issues together in eclipse bugzilla so that I dont lose it again +The names are the same as those used in the Solr connector. +The archetypes in servicemix will need to be updated when we do a releaseRegardsJonas +It was released with the fix in +If this is done I am not sure why users should be told namespace limit is X worseMy rationale repeating what I said earlier is machines are becoming available with more RAM +Checked and corrected user guide for and filter packages +the codebase will now compile with +GoodSomehow the patch went stale +Tony I have problems to apply both patches with Adding both into a repository and creating a new one and import this did not work for meCould you take a look into it and create a single patch based on both aboveThanks in advanceSvanteJust minor problems we are close to a release of the patchStill some problems here please enable your IDE warningPS You might think it is a little too restrict but when shall we fix those problems in the futureThe good part when you think I am too strict is you may turn the same argument now on me +Ah i hadnt refreshed the page before i commented +Should we add a shutdown hook to try to make a clean shutdown or not worry about it +bq +Opened for the crash. +this should NEVER be impl in should do one single thing and that well +User suggested patch resolved issues for already released versions +So for the suggestionIs there any chance this code can come up in Hadoop because should ideally be fixed in +I am using RVM though +Thanks Alexei +Well so far using isnt working as when I try to use the enhancer on classes in my validation tests Im getting unresolved class references for items included in the tests jar +It would send edits from the to peers to be applied as if some local HRS is receiving local batchupdates +Presumably you have a file in the same package in both your srcmainjava and srcmaingroovy source folders +Im going to create it. +The code is working correctly +Updated statistic bug in asm helpers +Hi KrisThank you for your replyI understand it now. +Or it could create a bundle per module rather a single bundle +Conclusion is you are retrieving a pointer from malloc and you need to manipulate the pointer like you are in C +on the comment from Martin Ahrer. +revision addresses last comment +I have prototyped this for both INTERFACE and CLASS rules +Committed to sqoop +Thoughts +Is there any way to recover the data? The only thing that I observed was that the file was about GB +After I disconnected LAN News and Blogs has warning icon and has content but Project examples are empty after Refresh button for Project Examples is chosenI have fixed this issue and committed the fix in the trunk +Sorry forgot to mark this as started +If there is any valid replica then we use that and truncaterollback the other files to the sync boundary +Binary codec reads complete stream into string on marshalling due to protocol limitations +Patch introduces them again +patch looks good + yes let me update +Im reading old code closely +Since it requires a special environment and will be disabled in the normal test runs anyway +The examples in the documentation in do not include the key attribute +Please go ahead with further development +fix has been committed to head and will be built with the latest snapshot +Heres a rough outline of how cluster shutdown should work +Updated patch +No problem Im not one for prose +Updated in ESB codebase will be in next merge +In particular is it possible to configure jetty LTW without the use of a javaagent? Such a configuration is possible with Tomcat via the +Its only done for trunk +If not then it not being called seems indicate were missing some test cases which would be calling itCan you please clarify the reason for removing this API +Please let me know what you think +I verified them against the branch and it worked fineThis fix will avoid the manual step of adding the two files to the Ui and Doc pluginsCan one of the committers review and look into committing this fix +Unless Im missing something I dont see anything that says defaults to would have thought that in the case of an undefined any request method could map to it not just GET +David I upgraded your JIRA account so you can assign tickets to yourself +Here is a complete test case using httpbinding +We are running into this same problem in Sonar +Ill hold off until some resolution is reached +Miust be fixed on the next to we wont release a has been releasedMoving all related issues to the next version +Ideally though such state should be preventable with or any type of scheduler for that matter +Are you using a recent version from trunkSee that should fix the issue of a always using the default Tika configuration +Fix the bug by eliminating directory if out is specified +Removed this patch to replace it with another patch that tests its functionality +fixed with +defaults increased need to investigate why the increase is needed +Manually tested +Resolved with revision of builderseap message Use memory settings from EAP as default settings +Here the expectation is that injected exception should be propagated to blockThat will not happen with current case ran the test several times +We were also compelled to make the patch for this +Do you want me to check them in under this jira +Are there more references to this kind of error + +thread is a daemon thread it will be ended automatic +JimmyI did not raise any issue to fix that +Inject full recipient details instead of only +Done Im concerned by the day after etc cause it imply something much more complex like day comparison month comparison year comparisone etcCurrently this is only ms comparisonSo if you think its a real use case then open a new case and lets start to think about it. + +but why it is a beta version +So that would have a performance impact too which probably is bigger than a pure lock +Yessir! +If so we still need to add other processor to deal with these message body objectAnd for most camel component there is no requirement that message body should be serializable if we setup the processors by default it will be useless for most of case +Hi TorstenThanks for the patchJust need revise it a little bit as this example use xmlbeans as databinding so the exception here is different with it inBest RegardsFreeman +Attached project demonstrating thats an impressive turnaround! I just ran with last nights snapshot and this works as expected + someday +I dont think those null checks are likely to ever fire however since if the Thread goes away its dead +Need to use the isAcceptable callback just as a way to analyze the response data and then use need to control when the RPC returns +Test now passes locally with v +Hi Marcowhats the status? I think you said that you would come up with another suggestion +Thats fine but then we need to come up with a different name for the class like about +The DSL support folder is missing on the build for GroovyA work around is to copy the plugindsldsupport folder from the plugin to the +The usually fails on my local machine unless I raise the length of the sleep times in test +For example if someone was using metrics filtering then their filtering configuration under the old context would stop working +Yury I linked this issue with since from the user perspective they are related to each other +Set patch available +Any chance I can get the code snippet where this is occurring? If not can you answer some questions +In the email thread I suggested backporting the fix to at least eliminate that from the picture +would be more precise +There is another approach that will fix somehow issues and It is difficult to imagine that we have more than G of executable code at runtimeSo the problem can be solved in another wayIf it is possible VM can support and guarantee that all executable code stubs managed code other are located in the region having G diameterSo even for heaps larger than G we will not meet situations like and +yeah quite strange +Per a mi insistncia i repetici corresponen b amb el sentit central de dint. +were backported to as well adding the latter fix version. +I agreeTODO revert and do as Matt suggested +Roland please take a look and let me know if you find anything disagreeable checked in +Also it does violate the apparent convention that the callER knows the type of the object +Thanks for the comments +Thanks Gregory +Committed and deployed +Agreed to apply that thought to webflow state as well +Other thing Id like to add is to actually be able to pass in an to a mutation for the row so we could completely reuse the bytes everywhere +Not so in this case +cannot do much without additional input +I just dont like having empty parent packages +I will check +Instead it will just use the Collator object for comparison +Hadoop is alpha so it is OK if certain things dont work without this patch nothing would workRohini agree on the TTL for this that is why Ive opened at the same time Ive created this JIRA +What is the compelling benefit of Jetty over Jetty to? Why not just revert to Jetty for +Im confused +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +Peter do you have the file handyI dont have access to the codebase anymore because Ive since changed jobs +Committed revision +on patch v +Analog of page which should be removed +If these tests run in the same JVM then you need to be sure that interceptors trigger correctly +Stein given that has been committed successfully feel free to continue with your original plans +Ive been wondering about this issue +contains a fix to the +committed you +I checked this into trunk. +This one is about support referencefile inside the Spring DM IO bundle +The tooltip descriptions for the Ambari smoke testHDFS user seem incorrect +confirmed works with bincassandra false committed +But anyway less jars in classpath is right solution. +I believe this issue is caused by the test client creating a large number of connections to the server in a very small time +But it might be useful for some other co processor guys +Committed to trunk and branch +Please provide class lets rename and and of a bulk update all resolutions changed to done please review history to original resolution type +Propose a fix on the mailing list +We will need the fix from when we upgrade our EAP base CP to CP +Trying that +woopsThanks for the patch +Just one point is empty in all other languages but English because I removed the redundancy anyway forget it its an exception +I just want to check that I am using the right versionsThanks again and great work so far any reason why this is not yet officially incorporated +Please verify. +I think that if you truly want to accept all kinds of XML and if mode does not accept kinds of cases you have yes a new mode would make if mode does work it can be considered mostly a both xml declaration and doctype declaration are optional and since comments and are ok outside root element it would seem like mode handles all kinds of xml with just one exception it is not ok to have unbalanced trees +Moved from to just after merge of old branch into TRUNK +Hadoop was just released. +Problem was caused by an incorrect handling of static methods in the generated fixed unit test you for reporting and please let me know if you have any other problem +Good stuff +maybe it is not a bug because it seems to have been fix +Thank you for your work Junping I will review and merge the patches to +Sorry for any wasted time. +page to test retested it and it works fine +while i still cant explain how a private branch commit could possibly become a HEAD ive added code in svn r that throws an exception should the same error occur againthe callstack of that exception should hopefully help investigating the root cause +This came from just running a modified lang suite in which we dont run any tests after the failing test +This works for us +please commit to trunk + +I committed the patchThanks! +We also have that problem with our Archiva proxy connector was configured to not use the network proxy but now two day later the stack trace shows that the proxy is used for this connector +Right +I like option +Im going to remove the designation and downgrade the issue to Major from Blocking +After this is documented this should be moved to beta to actually get fixed +Need to resolve the Fix versions section +Committed +The callers subject is properly propogated to the ejb tier +It still works in my machine +AlexIts fine +Unfortunately it does not influence the tests result it still fails in the same way +In my case with this new update the file seems to be equally valid against xsd validator in STS +we need a mechanism +The above could still happen of course but from what I know of it is at a minimum less likely. +This patch uses that capability to remove the timing window that Tom foundThanks the patchThanksChamikara +solved my responsetime problem but introduced a memory issues in other kinds of searches +Fixed at revision +RPC calls from NN to backup will need to have a timeout +In the the connect run by the logs as expected but no logging from the main thread for the JBC channel connect appears +Maps Batch Algorithm is faster than the sequentiel one that I am currently using +patch and script attached would you pls try it? thanks a lot +Weve not been able to reproduce the problem +Will do it once svn is up +I assumed this to be fixable as well but didnt have the time yet to dig deeper into this +Verified +The task is jboss tools guides already contain proper screens and info. +Patches attached +I had have told my friends that jBPM could be released in June for many times +Patch integrated to trunk. +If the constant contails a decimal +Updated test case self contained now so we dont need the broker xml config file +DanielI havent been that closely involved with the introduction of the Model interface so hopefully youll let me off if Im wrongThe interface method doesnt have any methods that have a Model argument so Im wondering if that is the problem +Im okay with any of these alternatives Add a big huge warning to NEWS that this corner case is not supported to but an alternative solution for tagging will be supported eventually +The same test can be used for testing as well +This avoids such side effects one and for less general advice is Dont use autowiring by type for core beans that many other beans depend will be refining the behavior here but this will mean throwing exceptions if an overuse of has been unfortunately cant reliably predict whether will work properly +Added a factory to make adding newer hashesh easier but left the Legacy alone but it will be fairly trivial and more cleaner if we want to refactor a little more +no change in maintainer contact email address +Grab MRI C ruby and do the same thing +SijieNP +Simple patch +Im still unable to reproduce but assuming theres a bug I would like to have a patch asap then this Jira will take care of the migrationSo can we remove the dependency for +Verified in JBDS H +There are two main reasons for that we still dont know what is the reasonable default value the optimal rpc queue size may be different for different applications +The Smooks web site only has documentation for version and which have differences from the version +Capture the generated wsdl +Remove the section please +No I dont need an arbitrary object +Thanks Prashant. +Applied. +Attaching a patch the gets the basic feature working +Looks good to me. +Please add these these to separate Jira issues +for the patch + hereQuestion for you should we remove the use of in general as part of this issue and move to JDK logging? So far Jukka and I are both for JDK logging +Thus resolving fixed +So there is no unit test included +First cut of implementation +Applied +I just wanted to add that Im passing the serializer as part of the initialized method +the IDE should somehow check if javadocssources are available remotely +OK I committed the Release Candidate +Max is right jtidy fixes some xml problems +Heres the ive exported directly from ads +As the error message says your domain class must implement Serializable for composite +I run AS with Hibernate Annotations all day long +Posting it so people can try it +Could use the +Should this be backported to +OK +Just the spam assassin one to go +Thoughts + I usually follow the rule that anything that can go into the +This more closely matches the scenario where the end user is responsible for making sure he or she has all the required dependencies for the parser in question +Though i am not sure why it didnt work particularly in chrome i put those elements in a separate div and this helped +Fixed in revision +Enjoy! And thank you for early adopting the release +Adding patch with ASF license +thanks to Pavitra Subramaniam for the patch +Patch applied thanks! +Especially since in X Char is not coercible to Int +Ugly but how often is this actually going to happen? The general reason for any including is that it indicates a code bug that should be fixed not an occasional condition that should be tested for +Attached patch +Tests still need to Check for Be able to resolve by mappedName Check that can be avoided by specifying first stab working in MC Integration and RI +Complete with commit acaebaadcec. +this also resolve David Richmonds issue with Windows per his comment on Sep or is Windows better off simply not using the dll +Ive being discussing this with guys asking them to provide better integration points no luck yetThey have some valid concerns +Aaron for the test case you can check perhaps +Its been tested on Geronimo and +Absolutely nothing would have to have actually changed in order to reuse an oldIs there a simple way to identify if this is true? Or at least to identity if the commitoptimize resulted in absolutely no change? reliably +Hi Andrew I applied the patch and it seems to work fine for me +this is by design so that it matches inheritence +Since this requires a major rework of how we do things Im moving to is already fixed +if using a interface a key of the map +Thanks Sushanth +Any idee how to get crash dump from this one +From this it seems clear that Tuscanys encoding of this message with xsitype is undesirable though it is not incorrect because receivers MUST be able to handle an xsitype if present +No idea what spring does or does not do so I unfortunately cannot help thereAs for the Hibernate side of things I think that in what you are doing is exactly what you would need to do +Are you sure youve got the context node right? Could you provide a smaple XML file and source code that demonstrates the problem +Diff vs +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Applied fix to these issues in Rev +Both of those JVMs support the server option +Mass transition all resolved issue that did not see any further comment in the last month to closed status +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +move resolved issues to so we can generate release notes + +I dont see a problem with the fix now what could possibly be a problem with this fixThe problem is that after I apply it mail sending still does not work no exceptions I just get a message Cant send mail +We will have a staging rule shortly to validate that they are present and correct for this reasonSo is your previous message on the CLIENT ERROR The server has not found anything matching the request URI url issue resolved +I can roll this in but should the name be instead of as provided in Aleksanders patch fileWhere would the go +I think properties should have names and looking at the current schema that would seem to be the case +hmm eclipse we can cutrepackage springIDE to not force this down our throat +Any idea on how to fix this +The original patch is good but the code has changed locations since it was created +see a cluster +Jason i also think that this plugin needs some work on its logging MOJO commits i ll be glad to add more patches one with usage example for the new param and one for improving i agree with you this plugin should get released the sooner irelevant of any more patches + Reflection is not allowed messages seem to be thrown in several places in the Tag libs +design +If you have a little test program a la that would help +Do you want to mark it as suchor would you like me to do itThanksAhad +Done +User should catch this excetion and performs +Shawn +The Software Grant from prof +you can switch the state to WONTFIX if you likeregards per the JIRA meeting April +Fixed in r +I initially tested with the wsdl fixed the issue for wsdl +I added all the same tests that exist in Java +Screenshot in Opera is only in discussions in articles it works +for trunk and into trunk and +I did that for at least the preparecommit calls. +It has some calls otherwise it looks goodCan you regenerate the patch I will commit if if the tests println +Kristian the extends your IT with a forked version +Comment out the optionexactarchThen install the packagesyum install tomcat httpd do not know the yum so much to be able to say s rather a question for someone else I on my RHEL testing machines exactarch does not have any effect for EWS installation +Ron + +My workaround is to flatten the schemas but thats very ugly as far as schema reuse goes +final +The latest patch looks fine to me +Patch adding the required javadoc comment +I think it is helpful as only one regionserver will print this others will fail +It doesnt produce the right result for +You agree functionality that is gained by making this async call can be accomplished without it +The possibility to create custom distributions add complete different requirements to Karaf than Tomcat and those vars in combination with the setevn files allow a very neat option to configure your app as you like +I included an example doc to test it out withCheersChris +The basic idea of XTS demo is running a reservation of some restauranttheatretaxi seats using both and The test uses arquillian and drone extension +The patch command is a and is not in or master issue should be reproductible with any command that flag the server as + is quite closely related to this issue +Have not tested actual indexing to Solr +Relabelling this JIRA as a Profile Service issue +I gotta polish the patch a bit but it looks good +So this might be simpler now +etc +in case you are using OM you dont need Office headless need to set usrlibif this doesnt help make symbolic linkln nfs usrlib opt oddearlier i try to set the path usrlib but it does not worknow it works fine +Please provide a simple test for your issue that I can reproduceThanks +in Paris we decided that if there is a line int i it should not fail even if there is a class of the name i +I see abortions and failures in ee +Fixed +Sorry forgot to escapeexcludeTables SOMEPREFIX SOMEPOSTFIX SOMEPARTOFTABLENAME +The problem is really that these high end servers is really slow +add user to assets added to init SQL +The Origin must follow the uri spec which contains the scheme imo OPTIONS works everywhere and the xhr clienst wont be able to override if most of the time I will add some doc +Thanks Pranay for reviewing the patch +idState Bundleosgi Is not good +Both Emma and Cobertura has the same basic issue they rely on Shutdown hooks initialized via static code blocks with out any extension model to override +Based on a quick test it looks like this is a regression caused by +Being worked on by Lester Lesters patch in Cargo but Geroinmo support is not working yet and there are open issues +I found this issue by coincidence and I decided to rework one of them to fit into the examples implementation Im sending a patch with my sort example a corresponding unit testManpreet I see that you started working on this issue a while ago I dont want to interfere with your work +In this way I cant see a problem either +FedericIs this issue occurs after applying the suggested patch! +Ill try it today under Windows +If not more care is going to have to be taken to avoid the raceetc +It is but this was not part of my priorities and since nobody provided a patch for it it did not seems to be in users priority either. +Ahad I can understand the problem in but the should use the same principal for both client and server +There will be too much changes I want to avoid the impact of the change to the releaseEmmanuelDo we have a converter from RFC Generalized Time syntax to in +Slice sampling requires two functions the function that maps a user to an X value and a function that maps an X value to a Y value +If not please reopen this issue. +Assigned Normal urgency +Yet the same error messages apear in the file +But I think the problem here is that the MOP methods are added in ACG +Thanks guys! +At the moment its just in the package and has the name which seems too vague to meHow about calling it? Or putting it in a of which makes it clear what context this exception is used in +Im guessing the separate config file is used to support runtime reloading of the tiering strategies? Putting it in a separate config means that users wont get confused into thinking that other properties in are reloadable +closing all resolved issues for previous releases +bq +So you could doexample or example designexampleNow it doesnt work well when vhosts rewrite url like the last test showVhost rule testdbdesignsomedocrewritealias ruleexample Vhost first rewrite the path to testdbdesignsomedocrewriteexample This patch doesnt match aliases then it will be ignored +But you may intersted in and at most in CVS please test + +Cleaning up old if necessary +test case relative to codegensrctest +If you could just rename this issue to describe the problem and provide some steps to reproduce Im sure youll have more luck getting a response +Hi DagYour reasoning sounds good to me and it looks like the patch does what you described in the comment on +It seems that this solutions is the one that is being recommended but does anyone know it this is the case and if so is there any estimate as to when it would be included or released as a bug fix on the update site +Since you mention the extraClasspath I wonder whether this is related to some unfinished feature branch I started a while back in. +It is now working +BXML calls the initialize method on the Bindable interface +Can add that as a patch if yall think thats an appropriate way to solve this thing +This way we might have a group of nodes that is always underloaded and moving together any time a node goes above the load threshold +qualifiers Default Any and no other qualifiers which really means when the dependency will be satisfied by the containers injection point +Meant to say patch and test must be applied first +Can we set to local to trigger master and regionserver in the one JVM? +guessing urls is a losing battle right? in order to make something public in rackspace you have to use their cdn system which makes oddly named urlsBiggest problem with public urls are +Kishan please let me know if your changes have been merged as well +on the patchWe should probably commit this patch to an branch +Good work for getting a complex plugin to work not so complexly +not happening in +add layer +When there are issues with libwebcore this usually indicates issues with the Android itself and not Cordova +su continuum resolved this. +Anything hadoop will have on the cplib would be on the classpath +Im seeing this on linux now too albeit intermittentlyUnfortunately I cant get the cobertura code coverage stuff to work at all with setI havent seen this on ant so that seems to be a solution for me anyway +Add an optional updateOffsetstrue param with the default set to false +This is now fixed in CVS +Works for me rewrites the tag. +Hmm sorry about that Stephan + go for while you are on it I think we should rename to +It now resolves whatever is passed as a first step in hindsight this appears to be a more correct thing to do as wellExisting tests cover the scenario +thanks Garvin for your contributionsAnyway I think it would be better to have a patch based on the latest SVN version which is a lot has changed in that version +Navigate to a page starting a web flow +Hence closing this bug +wsdl and generated javascriptNothing in CXF is designed to support the ancient SOAP encoding stuff +is there are process to review my code and how should I commit my changes because I dont have shell account How can I have account over there. +Humm Could you please explain more precisly your idea +In any case Ill the PDF in the concerning it was my mistake +Patch applied with thanks +I think that these questions need to be answered before we move forward with more code developmentI disagree +We can set the default values of the parameters to the proxy config in the file +So the fundamental design or the kernel of the system is supposed to be totally agnostic of the use cases and we let the users plug in the implemenations in javaJS etc so that they can do what they actually want +So basically the Axis guys have taken the other would be nice to document this on your website because I guess a lot of people are using dates in their applications. +Applied work committed +Please reviewFor the time being I added a of called which can be used to signal cache operation failures +Im using hadoop and want to apply this patch +Why not include this functionality in the existing +Weve already done a lot of work with so marking as resolved. +The script will kill the jetty processes when its done so if you run the script with a debugger on the jetties comment out that line in the script +Thanks for looking at the patches while work is still in progress +Cleaned up version of the previous patch removing extraneous whitespace and line ending changes +The initial input we got back was that matching aa was the expected result +This approach makes creates tight coupling with top level entity and makes it necessary to use custom strategy +Committed r Thanks Sijie. +Possible solution see patch file +etcWe will have a not only the component but as well the interface of its parent +This was a leftover of trying to the resolverOops this was actually still saving effort for the scheduled repair resolve +Yes but the MCAOP still needs to work in other environments where we dont controlthe system the system property will sort out thirdpartyuser code thats harder to ve created a linked task to fix +Bulk move of issues into after vote that we merge branch into TRUNK up on list +java file for it is +Core support is available in the release. +Reviewed with Javier former library removed +It turns out that the default VM args specified in the debug configuration pane were not quoted and I have a space in my eclipse path +The script is clearly too bulky for the VM +I cant reproduce the issue +We can easily step into the file if we need to +Am I wrong with this assumptionOther then that I will dig into this and will try to solve this asap +I have attached MEJB for review +Sounds like we are good +But they still have that option. +pending Jenkins +finally constructYes this is a good improvement over the testsChange the priority to minor since this helps when the unit test fails with an but the unit test is not supposed to fail +Rejected based on comments. +closed for +Done +which is more complex there +This bug exists in close as well +Ill try to take a look for +Omar instead of reopening we prefer creation of a new issue so could you please do this with Affects Version +Also need to see if the changes need to be applied to any other nodes hash joins seem to be taken care of +bq +Thanks Szehon! +thanks for your help. +I just committed this to trunk and +adding patch to return reconciler info to describekeyspace rval map +patch +I propose we close it if we here nothing prior to the release +This will then allow Hudson to build the whole thing and testdemo. +This is reasonable although it is a pain to have another step to perform during the release process to hide and then unhide the messagesPerhaps there is some way we can modify the log level automatically based on whether we are performing a release versus a local build +Where do you think it should belong? Maybe we can just add a reference to it there +updated with currentThe ops per sec in create increases from to in my machine +I dont like how tricky it is +I think this is fine to move that out no need really to keep the IW lock since we already have the commit lock in our hands +I should have enough karma to take care of steps +I think we make too many requests of +That doesnt explain why the build fails with version but not with version when both of them depend on + +In the end the indices were not the cause for the hangups in DBSybase. +Bulk close of old resolved issues. +Could you please check if the user has the permission to write to the local disk in both scenarios +jt ajt bjt bRandomly kill and restart the B nodes at one point A will hang on the solution that works even without FLUSHI would keep resending a request to members from which I havent yet received a response and which have not been suspected until I get a response a timeout occurs or a new view without that member is to +Thank you very much for the feedbackI updated the descriptions and portlet infos based on the comments +What do you exactly mean by attempt to reference the file object? And at what point in time does that occur? Within the same HTTP request or deferred in another piece of code or another thread +Until then it makes sense to have this work like the rest of Solr does +Christian submitted a patch against our crypt routines that we are committing and will have up on the ftp site momentarilyLet us know if theres anything at all we can do to facilitate this +inbound requests to Solr is still handled fully by the containerIf inbound requests are the only concern lets drop all the intra node stuff that requires the Solr code changes and call this done + close for the release +The header Authorization Negotiateis not send but I would expect this to be the case +Fixed +Could we even avoid inlining the and jarsYes I will create a new issue with patch. +So I will commit my change on trunk +Fix for the issue +Unit tests pass integration testing requiredManagement server updates underwayDeployment guide not yet started +test case relative to codegen +On server type setups we should be OK as entropy is also gathered by network and disk activity +I have check it with the command validate and the file is valid if I remove name I have an errorMy idea is that you need to add on master too +So who did the integration and patch of Tomcat? Whoever it was might quickly know why the engineOptions stuff is failing +The confirmation number is +Making this test compoent for now +The patch also introduces checks on the command length before it is executed and surfaces a nice error message if the length exceeds the limit +Hi SankaIf we are going to do this then do that for I agree with your suggestion that should try to read the Policy by means of a actual URL that it refers if the registry doesnt contain it +But since this is true why not go all the way and improve with these features rather than creating a contrib module that has some code duplication and will have to be maintained separately +I just checked in a fix and added a testMikes test succeeds for me with the fixThanks Yonik +Do you have a reference to a document that describes this restriction? Ivelooked through versions and of Microsofts RTF specification but havefound nothing +I cant choose the ejb project so I choose show all projects next +Work is now being done on the branch in newer issue takes care of general refactoring with test coverage improving as a result +just a thought +Can you hold fire on this until Ive completed the Big Renaming pleaseIm changing the files as part of that change and XML is a real pain to merge. +In addition to applying the patch youll need to copy srctestresourcesvfscontextjar to srctestresourcesvfscontextdirwithpluses +Thanks +I am assuming you have mistaken as +What I applied to trunk +From my point of view those dependencies information cant be defined in the plugin but only in the Update Center as we need to be able to specify dependencies between versions of the plugins +Eugene could you ask himher to send the tomcat log files? I cannot reproduce the problem +active is ONLY used when no P are used +This issue is solved in +Verified in JBDS H Linux +Thanks for the gradle debug help +The looseness with the API is probably the fault of the different Sun JCE implementations +Excellent +Marking as fixedLohit Anything you see missing from our metrics when you look at trunk? If you see anything youd like please open a new issue +Now there is requirement to add also and local resources +In the other refactoring only the was left as it sets the context classloader which is not useful outside osgi +please help to review thanksit costs more than one hour seems to me its more difficult to write doc than coding +These blocsk were part of writes by clients and the protocol will ensure that the replicas of the block will be maintained +Renamed and its provider to BM +Thus I approve +The purpose of the test case is to verify Automatic Accounting transactions it follows the following process Create purchase order for a n number of Product X approve it +All subtasks are closed +Thank you Todd and Benoy. +Thanks for taking a looks guys +Looks good +Ok I think I got what he was saying +patch with git option +You have a workaround now so please stop reopening this issue +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +thanks did this patch actually make it into the new RC hadrian just released +If you really want to bundle these libraries in your sar mark them as provided then +Patch adds target to build listener jar +Updated issue subject since apparently that is a hurdle here +Thank you for pointing out the issues +I believe these two are related +Oh and where possible it would be nice to have some unit testsServlet related stuff is probably too hard but things like should be easy +fix javadoc warning about params +validated. +Feel free to close the issue if you issues as per last comment as it seem like problem is somewhere in Maven SCM code +Would this be helpful +If you find further issues please create a new jira and link it to this one. +Under high workload they are sometimes different + +I dont much like doing that but it is a pretty handy tool +Use a local variable to keep track of changes to the metrics and apply the changes all at once +Another one to think about for Seam integration +Should probably throw an or add a guard for the possible NPE +Adds repo to grab put this in ql because that is where the hbase stuff isAs to jug my latest patch has another approach +Which I wont do +Dup of with the release of. +Chris A new optimization for pasing params to templates was introduced in +I think that issue with rd party tools need to be handled separately +That is why my number requirement is that everything stays the same as before and that is confirmed in my test suite +fixed in r +OK I committed this to trunkSince its a biggish change Ill hold off on to for now +Attached complete error you post the agent error log also for the same + +In updated patch I removed exclude from and updated Alex I think that and should be reenabled on the profile +OK I am seeing it in our clusters too and I have also verified the configuration +Would this be customary when you create the database +class guess should be marked as reexported it actually is marked as reexported in do you need that get fixed in branch? If you do assign it back to me please +These methods should be fully implemented and there should be junit test for most of these as wellWhat part of the code makes you think those operations are not supported +I know there are limits to what kind of promotional stuff we can do as an incubator project +Are you going to clean main request from other controller files after that? Bilginnormally when you extend a controller you can override the requested in the destination controllerSo for example if there is a request main in and in another you can define again the same request main with a different definitionAnyway I see from the artifact info logs gives me an error on because it was not present a requestview main and so I have add a default one but it can be removed form the patch I have uploaded +I assume that the use case for working with the CMS is intended for folks who dont have so much toolcraft available +Actually this patch isnt complete +But I did file whichcould be expanded to cover this I think +Moving it out of in the meantime +Thank you sir +Committed the change +I need to rework the test cases to test this and put something in the config file to enabledisable +For all other cases doing this leads immediately to unnecessary failuresIm not sure about the blocks considered lost on volumes bit but if that implies that the blocks become unavailable then I think the approach is too + +Maybe include the path of the in the log message Writing record of this node as the most recent active +See next exception for detailsERROR XJ Java exception Invalid key for AES ERROR XJ Java exception Key length must be between and bits ij I think the key length is so the error message is mysterious +Gavin when you get a chance can you set thrift up for reviewboard have setup Thrift in RB +The zone appears to be missing an svn clientThe doc tells me to expect svn in opt which other zones I can access do have but the new one lacks +But I can add more if you need +Thats true but for this we would need a reference from the class to the +Sending trunkassembliessrcvarconfigSending trunkassembliessrcvarconfigSending trunkassembliessrcvarconfigTransmitting file dataCommitted revision . +Fixed in revision Mark please verify it resolves the good +The problem was that was stripping out the port number if the Protocols default port number is the same as the port number of the target revision . +Changed the build file to generate a jar file for the tests as well +Logged In YES useridRemove target stuff from +No response assuming ok. +The patch which adds the new feature described to tjust committed your stuff could you also provide an example for thewould be greatregards and saving state for new properties was missing now its added +dfs command syntax is the same as hive +doesnt +hmmm sorry about that +It looks like were unboxing too much in the case which has been a common challenge for mode block argument passing +I am planning to submit a patch for this issue in the near futureThanks Ron Ill take a look at the attributions in their LICENSE files +I added a basic to create our current siteThe pom also include all the dependencies needed to build and run testsbuild reports for but it should not be used as the official way to build it because it currently depend on third party libraries from ibiblio +This is one of the problems with trying to be compatible with MRI +against trunk +Closing +Trying to understand why we disabled this a bit better +Yes it is just the collections API +I hope someday itll be in Hibernate core source code but for now you can grab it from this same page +has been marked as a duplicate of this bug +issue. +Could you please take a look +Fixed checked in and will be available in release Thanks for raising this issue and the good patch +This was resolved in the CP release notes +Dominik can you see the error you got with Jan Jamrich? He did the QA for Portlet Bridge and I think you sit in the same office +Youre right its probably not a big problem +I dont know what is the best strategy for this particular exception handling but logging debug is not good when you get an unexpected throwable +Targeting to and +Is because of deprecated methods in +The code now makes sure that we only rewrite the href value if we need to. +The other owner nodes think the call was therefore started by the primary node and thus will invalidate all known requestors for the given key +It is just not committedI have clients using it in production and one has M page views a year and so far no problemshttplocalhostsolrselectshardslocalhostsolrlocalhostsolrindenttrueqfacettrueprice +This was done as part of the pluggable belief +Patch attached +Go to the forum +to provide more complete information about log entries +Good point shouldnt the client retry unless it received aDo we want it retry a timeout? For example in the above pathological case we timed out because we wanted server to open k regions and it was taking longer than rpctimeout +Was the username and password the same from installation to installation +I got some errors on the write side but then I verified that those errors were present before this patch as well +Exclude and from transitive dependencies +Not a difference in working and not working example is just the camel version in +That has no cost on people who have implemented the old interface +The label on the link being The Castor docs command line tools and examples implies to me that the javadoc is included +Id be happy to contribute to this +I do not have time to look into this now but by a brief comparison in the code I have a hunch this is happening only if the namespace associated to the workspace is not the same as the namespace set in the datastore +In the near future wewill be moving to a newer version of maven +Anyway they were already uploaded +If this looks OK to you Mark Ill commit +I committed the fix to trunk only for now in SVN revision +Manually integration tests +Being verified at +I would love to see this kind of functionality. +The stat for message rate is actually meant for monitoring data rates from producers + resolved issues for already released versions +I think the menu should go vertical instead of horizontal to avoid horizontal scrollbar if the menu is too many +The error reporting should be improved though which is way I set the target to +Looking at the bug report that was supposed to fix and looking at the proposed patch its obvious that there was a simple error applying the patch that resulted in the extra return statement. +e +This appears to be the case +You should upgrade to +Thanks for your help Otto! +committing +srctestjavaorgapachehadoopyarnserverresourcemanager +Closing issue fixed in release. +Thanks for the suggestions Deepa +committed +My plan was to begin the implementation work immediately after the semester exams which have only just finishedAlthough I havent been able to produce any work I can share with the community I am now ready to begin with the coding and would be very much obliged if youd be willing to act as a mentor to me during the implementation processFurthermore since Google Summer of Code is starting I would like to submit a proposal for implementing sequence generators for Derby and would like to have your feedbackThanksSuran +I guess that now we only have to add prepare the XSD for the new operation +The patch looks good +Hence the patch doesnt apply +The scope on my maven dependencies doesnt seem to be working to well with the plugin +Whats the getting this fixed? Is there a dependency that were waiting on getting fixed that we dont have control over or is it as simple as nobody having taken a look at it yet +Have thought about it and we cant really do it without making the code more messy than it really needs to be +Special thanks for providing tests as well +Since we dont have exteneded test yet it is worth investigate to see whether we can make this case as a extened test +Added UI work for adding host exceptions. +Ive noticed that for clustering tests there are used some other jgroups configuration +That being said I think this happens on any device based on what I see here +Closing this record since is now released. +Thanks Teddy! +All of the performance changes made for this issue have made it into the tag and will be part of the upcoming release +This sounds like a good idea +What it comes down to is that we need special handling for +The latest version of EAP isnt present in the latest CR build FP was updated +patch to fix two refcounting leaks +to work with your patchHonnestly all this part of the code must be heavily reviewed its a serious crapThanks for the report and the proposed solution Im currently investigating it +You are right +Will let others bring the issue up on the dev list if considered a release blocker +I am not +show as being fixed in +I imagine you can just drop the old library back in +This also helps against those few times there may be a test error due osgi service timeout failure after sec. +It looks like one of the reasons of the size increase of is its number of classes +In this issue was committed changes for suspending was suspended for use case of RDBMS restore workspace if using of repository is necessary to avoid inconsistency DB because in during RDBMS restoring is removing constraints on DB +One addendum is that I have temporarily put the code to format edits directories in formatOccurred +Hi Daniel Can we target this patch for Pig? test pass +Sending trunkSending trunkDeleting trunksrcmainjavaorgapachesshdSending trunksrcmainjavaorgapachesshdSending trunksrcmainjavaorgapachesshdagentAdding trunksrcmainjavaorgapachesshdagentSending trunksrcmainjavaorgapachesshdagentSending trunksrcmainjavaorgapachesshdagentAdding trunksrcmainjavaorgapachesshdagentAdding trunksrcmainjavaorgapachesshdagentAdding trunksrcmainjavaorgapachesshdagentSending trunksrcmainjavaorgapachesshdclientauthDeleting trunksrcmainjavaorgapachesshdclientchannelSending trunksrcmainjavaorgapachesshdserverSending trunksrcmainjavaorgapachesshdserverchannelSending trunksrcmainjavaorgapachesshdservercommandDeleting trunksrcmainjavaorgapachesshdserversessionSending trunksrcmainjavaorgapachesshdserversessionSending trunksrctestjavaorgapachesshdTransmitting file dataCommitted revision . +Ive modified so it no longer eats the exception +committed this +still passes so I think correctness is fineI can run a real performance test if people want to see that too but it seems strictly better to me +Nicholas the next version of the patch has these changes +Its not clear to me how often that will happen but its there +Will do a quick and check this in +There is an identical problem in the quartz example as it is a +Has this been fixed somehow and made available as part of another version +So long as the broker is capable of detecting dead connections on the server is all we really need though I agree that itd be nice if we just did the right thing checked the other connection is still alive first before returning an +Please check out the html filesIn addition to these files there are two other files that I wasnt sure if they should be updatedIll add them to the JIRA issue +Working as expected now. +Updated patch to account for Nates comment +Probably would be best since it has no external dependenciesAlso we dont need do we since records currently implement Writable? However many would prefer it if records didnt implement Writable +In order to consider further we need some affected environments to be identified which will use the legacy Oid +I also added some unit tests that revealed that the algorithm for computing the diff between the bags was flawed +All unit tests still pass +All operations like listing table retrieving table descriptor are verified from shell and Java client with master switching +Thanks Alexei +Otherwise I consider done +A is now thrown instead of a +This sounds like a good solution +Hasnt been talked about for a long time. +FixedAdded a set of tag names which are allowed to not have closing tags +I had to jump through some hacky hoops to trigger the race and being a race it doesnt happen all the time +Fixed in bdortchattrs branch awaiting merge to trunk. +Alrighty then Ill wait to see the updated trunk +Im interested in how you decide values for data and rackOTOH it seems we could use the notion of missed opportunities to schedule rather of datarack +Thanks for the updated script Tars Joris +Thanks +Yeah a new pom formatting plugin would probably be useful and a good place for this goal to go +Many thanksNEntity classes are used by the protocol test cases but unfortunately there are no unit tests specifically for those classes +ok made it clear the history part of the URL before in all these to which is in line with BRMS +Then nothing would be recommended +I am working on it today +Min thanks for trying this patchFor the first bug my patch does not transfer over the wire +Im not sure about this but it seems theres a server side task that regenerates these +implementationPlease let me know what you thinkcheers Roland +cookies are lost after an redirect tester dont set the cookies on request after an redirect response test case showing the first impression is that we should setup the requestresponse calling setup before redirect +Im currently running all the common tests now +I will make the change in both and and get some tests for it +Ive downgraded this Jira to Minor +The error message issue struck me as even if the underlying behavior is by design +Im afraid this is a known limitation at this point +Initial testing seems to suggest it works I will confirm and add tests for it +Could you verify that it works and the issue if there is any issues +I created a fresh install dir for +thanks for catching this out +never mind my idea it breaks way too many things +I also noted that in the wiki but have forgotten to set thisSorry for the inconvenience. +I have a scanner implementation that extends +Also more diagnostics are included when this stack trace is printed out you please update again and ifwhen you see this exception again paste it here and I will look at it +A new Spring resource reloader handling urls and custom cocoon protocols has been added. +This looks like a good change +reopen if it still valid +Assume closed as resolved and released +I changed the genral issue title to only because it is the only builder I know that is doing that +Updated the demo to include the Apache license in all the source code files +I like idea of simplifying but at same time am afraid to touch and am more inclined to bump the hfile version and start writing new hfiles w new encoders keeping around the old encoding classes for reading legacy hfilesOk on the vints +Hi SwarnimWould you mind taking a look at? Specifically now provides a Struct for defining compound keys with support for composition using both and separation +Upper scroller has its own form. +bq +This is actually a blocker +The Integer object does arrive safely in the hash but must be getting converted to a Fixnum and back to a Long on the way out +Setting to future pending a patch +Fixed on the X branch in SVN revision . +Minimal change to try again +Patch applied. +I think I found a way to do ll warn you when it will be done +fix help message +Closing issue as version has been released on. +Not for commit to any of the apache svn branches +I see there are still some TODO items left in this issue +solve my problem +Can you send a patch? I will gladly add to the cocoon trunk +I cant reproduce this either with the test not reproduce this +Could we move this to? Its only really a concern when programmers dont follow the spec +Verified in Affects and Fix in Version +Weve got over session beans and tracking down the exact dependencies that doesnt like takes a long problem is that doesnt work when you have EJB references. +This also appears to be an issue when extending a base styleI simply created a new file and altered the general using japanCherry as the baseSkin and Im hitting this same stack trace +The advantage of site project is that no compilation is required +Read Flick photo seems that you have made some changes to the model objects and have not sent those as part of the patch +Nick explained it and Im going to apply PR +This is basically the only outstanding thing now +Updated patch to fix javadoc and test failures +I definitely agree Chris the interesting part is how that manifests itself in terms of implementation which is where I am digging in at the moment +Only the JDBC constant still needs to be inThe attached patch eliminates the interface by using constants directly and moving the remaining constants over toThe patch also replaces most constants in with their corresponding constants in or Types +The problem is not universal +Please verify that everything is fine. +is not only a french char +It wont make the unavailable all connection factories in are of typeYes subclass means inheritanceLets discuss this on to differentiate connections and +Ive merged changes from the current trunk +We know use Nettys directly which is located in by Scott and Stuart thanks! +If you really think there is some bug in Hibernate you must submit an extremely simple easily runnable test case +You need to specify option l to road the hyphenation patterns first before you try to use them with option f or option h +Dedup for example relies on dates as long stored in Solr for it to work +Currently this plugin generates an xdoc file describing the goals and parameters of the plugin + +Mass closing all issues that have been in resolved state for months or more without any feedback or update +The latest patch in is ready to be committed +Kiranis this a different issue than originally logged? Can you close this one and log a new one +To fix it we need switch all to. +Thus Id rather see a solution like this pushed down to I dont think thats possible but maybe you have a solution +This patch also needs to be applied to +ThomasCould please check whether this issue is resolved? And if it is mark it so? Thanks +OK I see what you are sayingI did test it with extra trace logging added to see what line items were being checked and it was still checking items to cancel but I will have another look +Updates are blocked and region level flushes are forced until size of all memstores in a region server hitsMemstore will be flushed to disk if size of the memstore exceeds this number of bytes +Forgot to attach the patch +Yeah works for me so far was able to generate some wsdl from SEI +Thanks Bhallamudi! +Fix for this issue +This has been merged to the and trunk. +Thanks for the review Bruno +changed to build works. +I have determined this is bc the view refreshes after the nature is changed but does not refresh after archives is enabled +Committed at subversion More specifically this patch backs out the following patches The main patch which ported the work on sequences to identity columns +Now there is a new module in SVN called web +This is already fixed in +The reason the error appeared as an was because an exception was being thrown while a monitor was held +Awesome thanks commits break the unit tests +Sorry our comments crossed in the mail +Attached patch +However this change can be delayed by not allowing an index segment to grow beyond or and by only implementing the long docids for multiple index segmentsSo as long as it is ok to assume that an index segment can have MAXINT docs at most we could use an int interface hereDo Nutch andor Solr already have long docids implemented on multiple index readerswriters or segmentsThe other border is the max size of a document field +Evan any more info on this one? Otherwise Ill close it for now since I cant reproduce it +applied in rev +So only thing missing that I can see is adding a link to the newly made somewhere +I missed one file new file in previous commit here is the svn commit for itmsvn commitAdding javatestingorgapachederbyTestingfunctionTeststestsderbynetTransmitting file dataCommitted revision +Please confirm they have been applied as expected +Let me fix it +oops you are right +It looks like you have outdated sources commented code which enumerates mfailedClasses is added in your patch +If we are relying on the submitted value to be stored in the state that is probably not a good idea +Ive fixed your website we missed that one in the ll take care of your cvs to svn migration in the next day +BTW what is the use case of +Carl I didnt see your comments before committingCan you file a jira for the same? +Marking closed. +Patch from attached +Thanks Owen +OK Ill defer to your judgment on this +on! Is anything more needed to commit +Closing. +Or are you saying that your connection never succeeds even after the vdb becomes active +Im not seeing any of these files committed +I remember this was already mentioned by somebody else +Good on you +Good work! Just one little glitch +Im also entitled to mine +Im sorry for inconvenience +well using from doesnt make sense to meits the very purpose of the Open configuration that it doesnt do any sort of permission evaluationa given permission provider should only be used in a context where it makes sense +This version of will plug hole if it cannot find a region in the fs w the appropriate span +Hi Justin I do something very similar to this in my own code though not with JDBC +can you provide the vdb +The sizes of some columns are also altered +Allowing a remote search engine to pull data very quickly from Solr so it can filter local search results +I could now write a patch that as Jun suggested does not commit offsets if autoCommit has been turned off +Andrea moving back your issues to +If youdbe willing to put together a patch for contrib we could include it +it be possible to fix it also in branch +Hello GeorgeThe fix looks goodMany thanksBest regards by Andrew. +The FUSE Forge is an open environment for the creation of integration projects and is designed to help you make the most out of FUSE questions? View our FAQ +I committed changes to the annotation names to to to to to have the example +Thanks +That would keep full backwards compatibility +will look into later +mrepository +grails simpleapp +Uploading a new patch +Thanks John +bulk fixing the version info for and all affected issues have in are all done related issue open to look at future improvements to making the UI work better when no are availablenothing left to do here. +we cannot ship a production version of a framework that depends on a release candidate of another +Testing it on mac right now +HelloWould you please try my patchThanksBest regards AndrewPatch applied to NIO module at repo revision rPlease check that the patch was applied as you expected +Ill handle it would prefer to wait for one of the but yes would like to have this in soon too +It doesnt work for a button created with a foreground image either +This is good to have since it can cause some race conditions during the client pings +And what is your wish? That it works in all three cases? fun is nearly impossible we would have to make a snapshot of the currently active local variables and then give that as some kind of object to evaluate +hmm +Ill throw some code to not use this on ICS for now but Im going to have to close this as Not Reproducible +binhama missed in the commit +Committed revision Thanks Tim +Just wanted to see if the logic to build workspace specific could be factored out + +Ive verified mvn clean install +Hopefully you will see this fixed in the next ER build +I think it should be possible to write a patch to which would handleall standard components at one timeAll components using acting as a consumer use the send method and all exchangesprocessed by a component use the processExchangemethod to dispatch the exchanges +Thanks for adding the +Sorry for overlooking this issue will look into doing this tomorrow your original request of having per build log files for days should be +Some thoughts on storing block in memory +This patch addresses your comments except for +Thanks for your prompt response. +Attaching a possible ideally we should use the same in the done logic +The previous patch didnt cover nested structures in map values +Commit to trunk +FixedClosing +lets go with reason this request is stalled? We should have sufficient resourcesto process it now I believe +attached latest patch from as move issues to and +API bug +Applied patch as do not forget to close this bug once youve ascertained that therequested functionality has been integrated successfully +Can you put ut the stack trace of the failure +Bulk close of issues +Created a new implementation which maintains a process class loader cache +Ive opened a new ticket to capture that enhancement +Rescheduling for as no time to investigate currently +Yes you right +instead of copy paste from GPFSYeah I mainly wanted to start the conversion and then clean up the details +attaching patch for the fix +Test send the large message and receives and outputs the size +was opened and fixed for the test failure +Superseded by new docs +RI holds up to AFAIK +Set fp per the JIRA meeting July . +That is indeed a work of art +Thank youAs soon as we get v voted on and released Ill commit your contribution +Forgot to mention I tested from a databaseinajar and a databaseinajar to after this fix and works as expected +Thanks for the patch Alex +Committed thanks +I have committed this +If that doesnt change anything please try Lucene and tell us how bug your index is and what fields it consists of +I just committed this +doesnt this depend on the version of Drools used +But I am not an CXF expert +Are we ok with that +Looks reasonable I like the moving up of yarn and its separation from mapreduceAgree +Committed into trunk with into branch with into branch with into branch with r +v patch lgtm +Thanks a lot Nicholas for the reviewUpdated the patch with addressing your comments and also removed one unused import from previous patch +log files for the computation of the st bit of pi +None of the issues we still depends on have been marked as solved rescheduling +Bulk close after release of +ok Ill try to look at it before the release +bulk defer of open issues to +Should take care of the synchronization here? Else it would be a fix that is needed in hadoop common +Verified in JBDSVersion Build id Build date +This was fixed some time ago in the platform scripts for Android +Close all resolved issues for Engine release. + thanks +Marking this as resolved +We can remove it then later and an additional view in the sandbox wont be a problem for now +This is a setting though not sure you had something else in mind +If he does not respond Illtry to provide a valid patch myself +KevinCan you please provide a little bit more detail about this issue for the release notes +Schedule this for. +We couldnt start it from the bookkeeper script along with a bookie because that would mean would have to be dependent on which would be dependent onI agree that the code should be kept as independent of as possible though +Use provided of a bulk update all resolutions changed to done please review history to original resolution type +Assign Devin to review +Ive attached an updated version of the stress test code as wellAny help highly welcome! Synchronization is always a pain +Attaching the patch +Committed patch to documentation trunk at revision +Marking bugs with invalid fixed for release as fixed in. +Are we sure of that? At least Im sure Ive create a CF with compositeType with the cli at some point though I suppose it may be broken since then +The patch comprises required error message improvements a new util class for accessing runtime information about running jvm and corresponding test class +Is there a plan to commit this patch and port to types branch +Patch looks good Billy +I committed it +Committed to to be ported to also right as part of this issue +JacquesYoure the man with the planThis essentially copies the POS so if you can tell me how to connect to your machine I would love to give it a try +Thanks Vinod and Daryn and Jason for the reviews. +Committed +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. + +Should not we close this Jira and continue discussion within this one me this Jira is resolved its possible to Run Cordova project on Android Emulator we found some environment where it is not running we should create new Jiras to separate independent issues +Dennis GA is tentatively scheduled to be released +However wouldnt GSE need to conform to the TC Naming Context Rules? In other words I dont think this is a Tomcat issue? Why was it changed to include both and path names in vs +This does not appear to be a duplicate of to me +You could easily EJB transactions with CDIa yes IF we like to provide such a mechanism in the CDI spec itself then we definitely should reuse the annotations from the EJB spec as far as possibleb there are already working portable Extensions which emulate EJB behaviour via CDI interceptors +Closed after release. +committed to trunk +I this on March th +We dont upload snapshots to those JARs are built from a CVS version I added the date to the version number +batch transition to closed remaining resolved bugs +I think we changed that behavior on can add a configuration of the signout in that would allow you to configure it to let you to the current page +Does this affect JBDS jbt Yes it does +Hi PaulI tried it again with Teiid and the problem is still there +Yeah btw those Mina guys gotta get a grip together and release mina +Modified client API to use the new Timeout logic that uses single timer to time requests +I dont have a Mac to test this unfortunately +increase default sync period and wait for last sync to finish before submitting another +Patch appliedThanks! +Verified. +if is for the portal shouldnt it be then +Created. +fixes by closing this day February th as +bq +BTW displayName is an indexed attribute. +We could extract the default database into but I dont want to bloat the properties fileThis patch adds a roller property independent of the geronimoVersion property in the pom and makes the child poms refer to them +The problem is that contains an Object reference and as far as I can determine that object can be anything at all +Can you use both repos at the same time +Just for info you can tell to ignore JS errors just like FF does +Hi againI might have done a poor job describing what I thing needed to be tested forSo I reopen the issue and attaches a client server policy fileCould you point out an example testcase where the client expects the server to fail? And I will try to add one for this issue +I think its reasonably easy to use and very flexible +Im personally fine saying that it is recommended to upgrade to before if that allow smoothersafer upgrade +Will take another in hope that crashes dont occur with them +Maybe i will have to revert on SVN revision by revision to find out what wrong +SFSB do not survive a server shutdown so no I dont consider a shutdown to equate to passivation +Thanks. +Typically plugins sit in the sandbox because none of us are actively developing but active community development will get the thing promoted as well +To avoid compilation issues the classes are loaded through reflection +Works great +Thoughts +See +brian i committed your patch +How about using the first bits of long value as index of array and the last bits as index of element in arrayThen you have an array of arrays that can be with a long. +I see that now +The jar has already been removed and merged with the solrj jar +Should be an easy fix and could help with very confusing errors that people might is critical but potentially easy to fix +backport is not working +I checkout all of as one project +This restructuring should fix as well +Log file +The alignment will depend on screen size but to help this I used a centered alignment and also placed the button and underneath the collections +Revision same for +It kinda makes sense +I suspect youre using an older version of Axis which uses WSSJ which had different rules based around how Signature Elements are added to the security headerColm +I did not try running the change against the other backends the existing unit test seem to hang or fail +and +This is definitely trivial. +even without the above +Still failing tests +Deletion works fine now +Revision +I did tested this issue but definitely not in locales other then just english and moreover the jira issue does not describe this clearly +Applying the amendment +In this case having a system property control the transport sender is a bad idea and will break the multitenancy of the app server +I verified the tests on Mac and Windows +From here I can use reflection to remove the +This issue is fixed and released as part of release +please close this issue +I extended the DTP wizard and page classes to enable us to pass the connection profile info through to the page and set the driver drop down to the correct also change the behavior of the action such that it uses the profile if it exists creates the profile and moves to the DTP perspective if all of the required info is present and the profile does not exist or brings up the New Connection Profile dialog is some information is lacking and the profile does not also required a further qualification of the profile name +You can close it as not a defect or I can do it if you prefer +No word of easiest way to correct the bug would be to offer only bundles for download that work together +However I do have a question what semantic are you trying to achieve here? It is purely to eliminate the exception you receive when updating an OBR installed bundle? If so then I guess the current patch achieves this +I didnt read closely enough +I think making the semantics for using the is the most important part of the patch as well as this discussion +Even if its used in the template itself +Verified on JBT. +I have committed this +I am fine if additionally we include the core jar as well if we do not do this already +Yes it is applied correctly +I wasnt able to reproduce your error + +Third even having a created with a custom URL it would be impossible to force to USE that custom stacksmanager since currently uses the default urls from +For example consider Client writing to The network link between DN and DN in the pipeline is severed +I was referring to the example not the article one +I am not saying that this is the best fix but at least it fixes itResponsetimes for similar tests on same dataset after the patch +applied patch with the new it is quite useful to see annotation which have a differenttype than the current edit mode +Im now going to use this as a framework to test out some of the work arounds +Removed the use ofThe handles the management of the output dirs od tasks in uniformly for speculative execution +Looks much better Nahi +OK its done +I have looked at the HTML file and the change looks good + +EugeneI have disabled and uninstalled the after restarting Eclipse with clean the problem persists +is really old +The patch is not huge but affects many classes so a slower review of its correctness is highly recommended +Sorry did not mean to make you just checking for completenessThis patch doesnt have any new logic for splitting though +Please check that it was applied as expected +the ok method did not have this type of patch that I have attached will first determine if the status of the maven is not error prior to calling the ok +Hmmm +So I think we can close the issue since nobody can reproduce this NPE anymore +Id love to see a holistic improvement to how errors are handled and returned by Solr +Would that also work +simple clear code cant hurt to have this +This patch should be good for and trunk +FYI there is a duplicate issue at +Adds the pluggable serializer from as part of +Nice find + for the patch +The new class exposes also some boolean methods which are used within the class in order to apply the same rules there to detect a field name +The tests fail for both identity types +Merged into the master branch. +So committing this before server side RPC and full client impl +is required to make pass with the patch +have also rerun all the tests +Ben Ben H +I guess its not easy to get it reviewedcommitted if I dont include itHeres the patch +One can then archive the resulting project and move it to environments that have no network access and be able to build successfully with the o optionIve tested this successfully with Maven and MavenFirst runmvn lib clean installNow the following command will work just finemvn lib o clean recursive descent patch as promised. +Your python script can only contain functions but no top level code +Umm there arent any segment merger extensions that are loaded I believe? I think you need to contribute a plugin here can you contribute an implementation +Sames the case with part of the Grep example +I will work on the policy changes that might be required for because currently it can only read properties starting with derby +The bundles a which made our application log to output to target +If you add to the link element everything almost works the way it should +If you have any link let me know the docbook related things on groupId on the project no one seems to know who made this artefact +It also leaves the page interceptor handling for subclasses since I didnt want to add an API for dealing with both application and request page interceptor +no need to document only affects test. +How do I miss this stuff +Thats correct I just realized the API doesnt contain a higher level method to decompress an archive to a directory +It closes the loop on referencing a controller action from anywhere a redirect from controller to controller link from a view to controller submit form from view to controller +Would it be possible to factor them out in a shared helper method? Assuming everything else is ok it must be decided if we want to keep the special behavior for getBytes and getStringIm fine with it either way +How about assuming file names ending with +Must there be something configured for this to workI did not find anything in the documentation regarding this +addon info command has been adjusted with commit adafffdbebfefddaa +Otherwise you will be stuck with +Hi gavinThanks for your adviceI have fixed point you list abovePlease take your time to check againThanks +Anyway I will keep in mind to build and run tests with JDK for future Jiras I will be working on. +The general point is CDI has Extensions but you can rewrite CDI stuff only +JSTL doesnt the version all is hope you can consider this De La this for the time being since the situation is still the same for Spring +If youre able to supply a sample please reopen the bug report. +An Eclipse plugin providing a wizard would be great +committed to trunk +Resolved by uploading the following files to the public Roo repository +Yes thats the one +It looks like Javassist was build for Java +Fixed in which will be included in ER +Any chance something else might have caused a regression here? In the old code which retried over the restart wouldnt it end up just hitting a and then failing at that point when the NN was restarted? Given that the NN usually takes seconds to leave safemode after starting any retrying clients would probably hit that and fail anyway no +in any case the issue has to be fixed in jboss asweldmojarra +New screenshot with selecting organization context near Language bar +Hi Brian Yes its the classloader leak described in +W Maven +I mean similarity +Just needs to be documented +Enable verbose exceptions get a service excepton that has a full stack trace in it and open a new bug report about it including your style and possibly some data sample to reproduce the closing all resolved issue that have not gotten any more feedback or comment in the last month +Thanks Amareshwari making it just committed this. +You have to be really really careful when using strings for these kinds of comparisons since every operation will result in a new String being created and then destroyed which is why the implementation works on a char array instead of trying to substring +The of your resources will have reference parameters that make them unique +Thanks for your help! +Those two files need svn ps svn simple usage document to get started on creating an associated wiki for the sweet functionality Nicolas +Can I just clarify one last point please? Do you want us to slave and? Or just one of these? Im about to named to allow slaving from our list of NSs Once this is working we can drop hyperreal and sonic +Work started by milind +I didnt realize how much code you actually put together to do this +Were removing in. +Ok fixed this one in a more generic way +One more thing since we are adding avro library lets add some unit tests as well +Uploading a new patch that fixes the issue +This patch only modifies javadoc descriptions +Thanks Jingguo! +cannt process +Yew Poon next time when your patch is ready for review please make sure to put JIRA in a Patch available state. +You beat me Ted +So marking issue as reopened to make sure we dont lose track of it +It doesnt really have anything to do with security per se. +I just committed this patch minus the debug printf. +Closing since the change has appeared in the Latest Alpha Manuals. +Fixed for +Reopened on Aaron Peelers advise to ensure it is included in ver is another issue that I didnt notice before but the ontology developed by Karuna has a statement assigning copyright to himself +Looks like this is duplicatedresolved by +Sorry I hadnt tested the patch after refreshing it it wasnt ready for review +If you dont do that but try to set it in your custom it will fail because the extends which has already injected the valuesConfiguring the is also a delicate issue because not all values are replaced at the same time +No need for extra factory methods the createReParser and createReWriter methods are easy enough to override +doesnt look to or set visibility. +A lot more indirection is required in Solr and a schema is needed for pretty much everything +Any suggestions on how to improve this? Maybe the tab should just say Mailing lists +I didnt select I left everything default only runtime path was selected manually via file manager via Browse button +Thanks for committing this Chris. +Is this a new issue in +Actually you need to be very careful with licensing +postponed got presentation from eclipsecon that can be used as input +This updates the STATUS file to record the release of +Looking at the codes it seems the error message was changed since this issue was filed so thats why we encountered different error messages +Please bring it back to the forum in a different thread thenIt also depends on how the other side is mapped AND if your data is really connected that way +next week +please dont put patches in but attach them to the main issue +So the task isnt fully complete +Forgot to mention all tests pass including test +I committed this +Sorry had intended to this that the content that has been added upstream is have added this row to the table in question. +Yes I understand what you are trying to do we must have missed the +working through the whole document to make references to server profiles and JBOSSHOME consistent +this is can not reproduce this issue +Yes we could use a fix for this +Thanks for the patch +The prefix used in the patch is not +Correct me if am wrong Chunhui +I choose the keyword overwrite to indicate user want to overwrite the fileThe following is the implementation details +Unsecured no longer WARN +Quad predicate cannot be null is unrelatedThe test case for this report can be reduced to a line test case +I believe Ali has done several changes to the build so it is now possible to build with MavenHowever there are possibly some things to verify before closing this issue +q file here +I guess i should monkey patch it in? Do you think this patch could helpIn regards to timeout i am doing something like that but its not quite as simple as i am pooling imap connections using apache commons poolI have tried timeout in several areas and it does not seem to kick in when imap is hanging in this situationThanks again +A new bug report will be opened for the remainder of the discussed functionality +Cannot reproduce error attributes are inserted correctly +I dont think that the extra call to the resolver will change that +NB to be able to use Jasypt I added and bundles in the startup +Closing this off but dont consider this set in stone yet +CommittedThere is same problem with mapping of float for and PosgresBoth drivers maping it to DOUBLE JDBC typeThinking about to follow it in our Cayenne mapping +Yes Gary +Closing RC issues. +This should be working +I reverted back to my original manual test via a chat swing GUI and was able to reproduce in +Tomorrow I hope to solve this last issue on integration tests +Also added more unit testI think this is ready for someone else to evaluate if anyone has the time desire +Moved to be a subissue of since it is committed there and resolving +Theres still a few things being tweaked in the but that should be a separate issue +MubarakArvindi am going to look into this and revert with more details +If there are any other issues then please document them in so that they can be fixed for this is being fixed for remove the patch please +Ah okay +however I doubt it will be included in the svnthat was my only pointBTW I have and eclipse style interface that I ported over the covers most all that ofbiz doesNever expect to have it included in the SVN either +Non singleton endpoints could potentially be created many many times eg +We will warn people to migrate away from DBCP. +Yeah the point is that we havent switched to AMQ yet +I chose a default of retries +No need for tagging recordsWe dont want to support compatibility of disk serialization of some records forAnd as Sid says Records are all YARN private +That would be nice to have +Changes look good to me +bin +Thanks Steven +patchDaniel Key is to set properties through an instance of as this is the only way to apply deprecation logic +I had the exact same problem with the script which is why Im now using the Grails SVN head script +Linking to the patch that removed the deprecated methods +Looks like I abandoned it on the last length +Did you raise the H issue against Hibernate +Please apply +This will hopefully stop the tinderbox and the nightlies to fail due to failures in the various updatelocks testsWhether the absdiff mechanism will be kept or not is up for discussion this commit is primarily to try to stabilize the regression testsCommitted to trunk with revision +The root cause is at the bottom the GUID serialversionuid mismatches +Does this include all the archiva modules from + +Everything works fine now it may take quite some time before we can say it is fixed +Marked as since I was not able to reproduce it and the console has been rewritten +Still diving in +The import job in the end of the wizard now also reports individual this issue is fixed now correct +if you deploy them youll see that the HTML source does not appear to be correct. +Does this make sense +Without that part you wont get something to execute +The neighbor region would have new delimiting keySorry Im not following Ted +good points v previous comment I wrote that estimatedKeys should be greater than +Feel free to reclassify +Thanks Enis! + +Have you tried that +So its ok with me to remove the extras +I havent used so I hadnt come across that feature +Assigning to Snjezana so shell see Maxs question +Even though my ivy dependancies were pointing to wicket it would see tomcat has older versions deployedIt works now after cleaning out stuff. +I think it will fail horribly at start because of several reasons I could have foreseen given my little derby experienceI would also appreciate a lot code review with someone with deep knowledge about interfaces and how should they behaveThat this sounds like it is more than a day away from ready to go in +must be constant so its not possible to guess the separator here. +As for we should use package access where possibleIf not use gettersetter which is documented for internal use only +I dont really see any way that Spring Security can restore the states of beans that are stored in the session and have been destroyed +and also the juddiDB looks good now +Committed to TRUNK +aeeeeeaebafdfdae method has been supportedbut the response HTTP headers do NOT contain because get the correct content length is too expensive +Fixed by +ClausI think you need to specify the version range ofIf you doesnt specify it the bundle plugin will pickup the version range according to of the bundle which is managed by the pom dependencies +Maybe Skip as well might be interested +I was skipping stylecheck sorry +Hi Aleksander which version of do you use. +The migration tool is working well. +You missed two entries which were still atfeaturesfeaturesfixed in trunk. +Please close this issueThanks +Adding the ability to specify multiple modules using lib property in the file as well as ability to support JAR based deployments as modules +Because I was confused about the realm of isolation and have edited the ticket to remove any reference of isolation because I was mistaken +Im attaching and with your suggested modification to the NULLIF topicNot to open a can of worms here but I think that in SQL NULLIF really is a function rather than an expression +from me for merging into the trunk +Thats why the channel factory is specified as optional in the persistence config file +Unless we have control on size of input lets abstain from creating +Looks ok to me +I guess then we can use the old fashion classes as well +deploy The schema tool now only contains the mbean template and is no longer a complete XML file +Mind taking a look K? See there are no objections I commit this change and close both jira and as fixed +Remember when we were at and Jay was mentioning all of the injectable objects in richfaces for headers +Committed +Could we somehow create two different capabilitiesAdd Richfaces for Richfaces for JSFFacelets to work around thisIf not then lets us update the capabilities to work like it should for jsffacelets +redfishktcAre you sure that you are getting the same error with mvn sitesite or mvn sitestageIf so can you please attach your POM and post your stack trace here +The only thing would be to upload a with the right we delete the as bad and put this up as then +Murali is looking into it +But Im now wondering if we should keep just on to avoid any confusion that could exist mixing a veto with deactivateable +This seems to be related to the fact that the QA machines use NFS to share filespaces so that testnodes attempt to install the product multiple times in the same location +No its not supposed to happen nor can I ve traced through the call heirarchy several times to make sure there arent any more places where a builder gets added capriciously +The index logic is looking up index records as if they are children of the tables +I never received an update notification for that reason +And if block is simpler on the implementation side then lets stick with it +ReopeningBtw +thanks Ivan. +They do some ENV setting and they delegate to the Tomcat startup scripts +I found the cause of the problem +implement solution Elis you need to regenerate the diff this one nukes and modifies +I just committed this as a part of. +Thanks for the fat patch that everyone else was afraid to do Gregory +You will need to build the site your self or see the source for will need to split your script into multiple executions with the first with block mode enable +Note may need to synchronize a critical code path. +Seems to be already fixed +Ok so I guess this is resolved and we just need to make a documentation note about including the CSS file +I hope everything is ok since this is my first contribution to the Jira system +Attached the patch and an html copy of the updated topicPlease review +Moving this issue to. +Thanks +I am running tests nowTwo types of problems are addressed in and Mutable Timestamps are copied +And technically this is not a bug but an improvement since the explicitly behavior was to in fact ignore the lob data. +Have you seen something like this beforeThe clean version works perfectly +Ill see if I can address these problems +Assigning back to me while Benson checks some local config issues +This triggered a slew of other problems which at least for DFO seem to be cleaned up +addresses closing streams in a finally block in loadPartition and elsewhere +Geronimo parts implemented rev +HiIsnt the a standard Pax Runner feature? Or do we use some specialty that Runner cant build itself +Assumed fixed +here is a junit testcase +It does NOT set up inheritance in Javascript but it does arrange for the derived class to have all the inherited attributesCan you build from svn and try it? If not Ill make a snapshot after I code some tests that improve my belief level +Updated to updated patch on Fedora and RHEL +Ill start this issue +So i remove this issue from the pacth +Changed it to Disk error during leadership change +I tested it using the unit tests with Oracle the average time is and with PGSQL other remark I dont see any indexes that cover the field NORDERNUM so it could be needed to have a well chosen index composed of the NORDERNUM +chatted with the customer who wanted this issue the most powerset pipeline and the lads said that they want out more than they want this feature and that they can live with this issue being punted to +I have no idea why yet +and its provider and provider impl were renamed to its provider and provider impl respectively and moved to a new activerecord etc moved to package identifier moved to package methods from and moved to and doc updates in Git ID bdfffcfdeaecfbeaa +It seems to workThanks a lot for this workaround +I did not fully understand the intended changes there but one of our reasons for using useOrigin is that with that flag set to true all artifacts names keep as they are when being used from the fileset +br revtrunk revThanks Siqi for the patchRex +Applied and Comitted +Write access within the same session should also be synchronizedBut read access is more important +Just an ideaHow about creating a branch of OSMF for with the fixes we require and call it something like OSMF for Flex developers. +Do you think that this will cause the rpmbuild to work in case is executed from the cygwin terminalI installed maven and got till the point of some error when executing the +patch for into this oneaded more tests make group aggregation with no skew as the defaultEncountered some bugs while doing so for fixed themCan you review again +I dont think we need to do anything anymore with existing instances +I found in have uploaded an updated set of files that has been created using the latest update from the svn +Use Ctrl for the reasonCtrlA also works may be it was fixedIMHO can be closed +Im in favor of driving the type of target source off the object definition +This is the expected behaviour +Given all the weirdness on the cluster the last few days lets wait on doing any thing with these until monday and I will run many more jobs over the weekend to make absolutely certain whatever was going on is internal and not a problem with the patches for sure +I agree with Owen on this +Bulk close for Solr +Thank you +I was thinking it would be better to the constant part of ACTIVESTATE and STANDBYSTATE rather than introduce a new enum but I think given that is in common it makes sense to do so +Create abstract classes for the interfaces +Setting to future pending a patch +Hi SamindaYes +Ive just added this functionality and all and unit tests pass with it in place since the interpolation refactor. +I have not come across the different odering problem but I will try with other jvms on other platforms and see how the test goes +from me +Good catch thanks! Patch applied in revision +Mark any suggestion for this use case? +But I assume I missed this case +Will this work right if I set scanner caching to and then deal with mb rows? In that case every response will be g and it would always block and never make any progress right +Will get around to this later today on trunk and branches. +the application can run before you attach the debugger and it can go on running after your detach the debugger +have you been able to verify this against the latest version +If we decide to release a release then it will be in there +We can just document it on that documentation JIRA +arbitrarily schedule all unscheduled CUDA items for X +Without refactor to Server managerfinal versionmaybe can seemed not so small change +bq +Not for commit here. +I have set this up now as adding the KEYS file for you was sufficient +Andrew have you got any unit tests to go along with this code? Ive started using it and would like to create some unit tests but it would be great if some already exist +that also happened on by box with an italian locale +Rebasing the patch +Thanks for the patch Sylvain +Let me know if this fixes the issue so I can close it +They can be configured through Preferences Groovy Editor Templates +Any other options or considerations +Committed to TRUNK and branch. +Push to a later fix version +bq +Fixed Emitter with r +It looks like they were added when the messages were translated in +Marking this as resolved +for rv option for rv option +Krasimir I assume you mean Helios and and not GalileoWe havent targeted Helios yet and trying to lock down on JBT GA which targets Galileo but if you got patches that make it run on both Helios and Galileo we would gladly apply them +This looks like a nice addition to Hadoop +Hi Michael feel free to reopen this ticket if you still face this issue with Sonar . +Weve fixed only situation then try to edit plain text after special symbol +HiI am using the latest version and it seems that the schema validation fix is not in +We should check if this is now possible using WFS. +Hope you understand my concern Please do let me know if you have any more question +Reopening issue until we get the changes to sorted outRegards BillSeems that your fix imposes no threats on the existing implementation +I guess this is something related to DST +they are going to be put off by the messages a month about Derby internals +Mikel yes it is technically possible but not necessary the best thing to do +Change looks fine for branch to me +is back in the gameAll projects build OK this morning. +AFAIK nobody is actively working on thisIf someone wants to take a shot at this then a good starting point would be to review the changes associated with +I dont know anything about how deals w service or maintenance packs but I know they have them because the place I know that uses it has them and has applied them +You could simplify it slightly more by not synchronizing on the keyLock since thats already held by the caller +Having a single node responsible for each shard fixes this problem while still allowing rudimentary failure tolerance but has the fragility problems described above +In fact expression by itself on the recipient could create a confusion with channel attribute as some might think that the expression is responsible to select a released +Bulk close for Solr +Well technically I dont think it is fixed but since it is related to threading maybe the timing has changed or some other locking behavior has changed and that impacted the behavior +this should be handled by the client +Red color isnt really important it was just a suggestion to be more homogeneous +Hi JoergI didnt change any artifacts I simply moved them from old codehaus repo to NexusBut xstream is not a good groupId so for new releases you should use And you can only deploy artifacts into nexus repo +on backport to +it looks really nice to me thanks Jukka +But what about segments fetched with and without this new feature and propertybq +Can this be closed now that has been resolved +is created without a namespace as it should be +We should change the tag to only trim for the comparison to though +Can you please make it a new issue though? This one should really be only about the serializability +NOTE was not failing due to this issue +Huh? Who exactly is product management +Forgot to mention that the xhtml template would have to be updated as well since it parses the simple checkbox template +Cos could you please commit this + +Integration tests added +If there is indeed still a problem here lets try to pin it down it time +I was getting about minutes with the settings we have in the build and about less with increased Makes one think I guess Great find Roman! +These blockers and critical issues are resolved but not verified +Have you tried? I had a similar issue with assemblyattached and it seems fixed in +ve just committed the last patch +TODOreduce reporting times +Since I understand the value of the FACC codes to your users I recommend allowing both mechanisms and using a configuration item to determine which will be the way the capitalization and underscore replacements were solely to facilitate the use of SQL to access them +This will fix the real problem +Can you please upload a new one +Originally I didnt want to mess up the style of the code by adding some comments to it This patch replaces the first one and has sets of block comments relating to the relationship between the combo and the underlying JRE array list +This bug was already fixed in Ode +Hi Lenthere are some tabs in your patch +See for further details. +The party poster +Other classes already call MapThis is my first patchafter modification i run to ensure modification is correct +To force resolving artifact from your local Maven repository you should uncheck resolve workspace artifacts setting in the Maven launch configuration dialog +See the related issue for more info. +mingjian Since you are looking the distributed code now maybe youd be up for having a go at? Or at least posting a patch that youve tried for Prakash andor I to review? Thanks +I meant the lib directoryI will ask my question in one of the sites you saidThanks for thatregardsCarsten +addition to the fix put in place in the batch script with the above patch has resolves this by removing the batch script completely and relying on the Eclipse RCP executable to directly launch the MC +output is identical under no fork once and pertest now +So were okay there +CommittedSending javaclientorgapachederbyclientnetTransmitting file dataCommitted revision +I cant think of an approach that is cleaner +Sure Ill commit the patches now +Many other systems have stats collectors that are used to augment existing monitoring systems and troubleshootoptimize application behavior I cant think of a reason Cassandra should be any different. +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Verify the patch on G snapshot and commit it to branch at rev +This can be closed for now. +starting now +These methods were kept for backwards compatibility reasons and will be removed in a future release +jsftoolstutorial are checked +And r for +Not an expert in this area +As you showed in the patch the issue was with the exception cases the shutdown was not happening because it wasnt being started +I am attaching the patch here please review once again and It would be great if we can test this +Thanks +Patch available I will check on IRC for someone to review +Committed this to trunk and +Also imported the unit tag this isnt +You have to go into the weblogic pool configuraiton advanced options and disable the connection closing when the connection has been unwrapped +Changing to fixed +stack thanks for your review and clearing the findbugsI was avoiding these changes as these are unrelated to this JIRAram thanks for reviewing the patch +The could automatically add this to the when are registered with it +Leaving this JIRA issue open until then +Not including snapshots is a process problem not a technical one +Sorry I missed the reply from Sandy to post some thoughts on this vs +Tested on IE Fireforx Safari Chrome +Ok half way there +While I am not against useful configurations its already hard to configure YARN +consolidating logj after release. +I am not willing to wait months for issue which can be fixed without minutes. +bulk defer of featurestaskswishes from to +Fixed +verified on ESB Tools +Close this dialog and press refresh in your guess is involved +We do not intend to support clients or servers that are not in Hadoops source treeSo do we intend to permit implementations of the scheduler that are not in Hadoops source tree? If so then we should not use interfaces +Im not sure what the ramifications are to this +Thanks PaulIm reading your code now +Heres an updated patch +This adds site validation to to trunk +What do you thinkAliaksandr +My current list of files that can be ignoredbsp +To reproduce this ran sleep job thrice +It worked well for me +Why not use the clean plugin +I bet youre right +The ant command did not work when running in srccontribraid but it worked when running it in +Attaching a revised patch to address those errorsTouches two additional test canonsM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterCommitted at subversion revision +I did not include unit testing on this but it should probably be added +Avro already uses the Hadoop elsewhere which are intended to be stable for some time +Thanks +and +This tool implementation could be used by Tuscany instead of directly using Axis tool +please review +Looks like Firefox is not sending the cookie upon a form submit +I have just retested the just to be sureAnd I have rebuild the current trunk in order to check the +It is the way the language is designed +I just want to do discovery +Removed extraneous signs in both trunk and rc branch. +Switch from hsql to mysqlRegards it for now if it has problem feel free to reopen or create another one given this issue has a lot of comments here. +rename to +In the strict mode presence of a content body in response to HTTP HEAD request will cause an to be thrownI was also concerned about possible side effects on HTTPS connections +excellent confirmed +Now everything works even the with SEANNOCADDATA using the +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Im working on getting tracing into hdfs and when that happens Ill have full install instructions +But basically hbase needs to know in some way how the timeouts are setup underneathThanks +Ok the nonsense I have above doesnt workHeres the deal +I will remove the formula from scoring and just make it link for now until we can figure something else out +Chemin absoluhomesitespublichtmlcgi +I am attacking the core of the problem as it will solve more issues later +Verified at +Simple application to reproduce the problem +Fixed by +Tested to work on my local install +At the time was created derived entities were not part of the JPA standard though +Thats true Ill refactor the patch to remove the lock unlock and just force grabbing the lock in the stop method +In my case it is not a database connection but an RMI object reference. +Do we agree on the first patches in the meantime +only affects test Release Notes entry is not required as this is relevant only to internal Red Hat testing +Two minor nits then +David can you update the jar to the latest version in the webapp? I have already update the upload string on my side +The design will not be changed. +Patch looks goodFiling against +Thanks Pat! +Adding version to KV +Is there an ETA for this fixI notice the installer has been stuck at CR for some there plans for a GA release +tests included +For now well go against the Devel docs recommendation and add it asa buildrequires item to +I think we will probably have a whole etc directory for each +Hi SureshInterfaces looks very good +This can be deployed via the user home java home or in the classpath in exactly the same way as the file +bq +With this command was renamed into deleteall and the rmr was deprecated in . +We clean up the packages before a new deployment for testing +Same happens to RC +Marking as Standard Library issue to be reviewed with standard library updates in the future +Modified to spill multiple times +To clarify it goes beyond security concerns though those are certainly important +The kinit login in spawn is not tracked by these metrics which is fine because this thread is spawned only at clients and purpose of these metrics is to track logins on the servers +And most of them in this patch are from new lines +Please note that this should be done for both and CP releases +The test case comes from the user +can this be done as part of this patch Kay Kay +Can you add a test +Thats not to say the update processor proposed here wouldnt be useful though +Im able to apply the patch with p to the tip of trunk +All DRTBVT tests available in the test distribution passed +All changes are simple renaming or comment updates +Suresh can you include this patch +Mark this issue as resolved +Do we need to do the same here +I missed the first part +Attachinga patch for remote execute using Jsch +Can you update the curator site xdocs to point to the wiki and remove the static page? Then we can close this out. +Last updated items on the docspace page have DONE in bold to identify them +weird +Slava please dont resolve this issue until the validator is not fixed +Patch looks good +patch solving the issue as described is applied thanks to Ernst Fastl for the patch +What Pete SaidThe duplicate class part is key it is not a duplicate m quite sure it was never intended for Seam to break normal classloading semantics +Seems that I hadnt installed the plugin correctly +Anyway when I have a simpler example why this functionality is needed I will post a comment to. + of a bulk update all resolutions changed to done please review history to original resolution type +Could you update all Sonar widget provided with the following rules Start each property description by an uppercase letter Insert whitespace between each wordSo filter becomes Filter pageSize becomes Page size display becomes Display filter description +There is still the method Randombytes unimplemented and some specs are still pending or so in total but the patch already solves more than specs +is a class from the BV apispec +You should be able to reduce such risk by downloadingchanging resources under different names and then renaming them +Duplicate of +Thanks AlexeiLooks now all necessary classes from print and beans have serial +I ended up needing to make a few changes to your patch nothing to major indexOf is JDK specific so I ended up writing another way to perform this replacement which will work with I actually ended up overriding Dialectapply in the instead since both use the same approach of and because extends closing stale resolved issues +Punting to as not critical now that the release profile is back in I hadnt heard that its back in +I tried to follow David Jenckss suggestion to validate schema while parsing the xml so that we can just parse once when unmarshal to object +Just find the initial configuration style is used in other places if it conflicts with those places I am OK to use the old way +This seems to do the anyone review +Youre concerned that youll lose the freedom to add functionality thatd benefit Solr only +Patch for reviewSince the change is very small and just changed to use suggested replace apiThanksUma +The patch I uploaded does request the postal code from the operator and uses the available infrastructure to send the postal code to +We typically run at full as we are not made of moneyNo you arent made of money but you are +Changing affected version to and fixed version to the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Patch v integrated to trunkThanks for the patch ChunhuiThanks for the review Stack and Ram +Fixed in Trunk +Hi Michael Uwe pinged me about the problem with async background threads +Fix is in CVS +See for detailsBy all means consider it for if it falls in line with your other goalsHowever in the interest of release schedule I will most likely be implementing the approach you outlined previously obviating the need for this behavior in +Committed +is that I already use service layer and repository in my project +I have commented out the test case a lucianos request +Can you provide some useful information on how to reproduce this +Although more testing needs to be done I believe that the functionality described by this spec has now been implemented. +Actually GROUP BY AC is not valid SQL +Is our deferred flush ASYNC described above +Patch applied and now in SVN thanks. + to trunk. +New tests position and reveal this bug for forward and reverse axes +Ive added another JIRA which fixes the tests which hadnt really been run since we opensourced hedwig +Please find a path to fill the table with groupId and artifactId read from the pom file +so is it possible that you create a large numbers of threads? can you use and make threads number configurableThe maximum number of threads created will only be equal to the number of configured volumes +Package private is better though +Both the issues are fixed +Thanks for the contribution Jim! + +I like the aliveSince attribute too as that lets you recognise when a service restarted +If we just want to prevent the as reported on we can just check that the build definition has a schedule +Hi MarekThanks very much for the clarification +Chris the version in trunk is not usable +Please feel free to do whatever you like with the sample project nice that its useful +This appears to be the case in as well +Check all of the listeners in the +Hey Utkarsh can you submit a patch for this instead of a jar file +Also get some performance numbers for goodwill commit if the tests +Patch looks great +Added applied +I would make an interface thats implemented by and would pass it in to the adaptor at initialization time +workaround problem click again on the header separator and release + patch look good for +It would also be nice if the interfaces used by a lot of these internal components were cleaned up remove the extra Operations classes that the interfaces extended +added for this case so I think that this issue can be closed +I expected them to be in the main text +Reassigned for review +patch is applied and running on ll not still close the issue until made stronger testing and backporting it to resolved Saul could test it its in trunk +No reply so lets assume its fixed in +I have updated the goal +I will review your patch +Sorry it was Cocoon +The test works for me now +Hi Jacques I think the patch probably need a little more rework +apologies the version used for the test was +incremental with this option aspectj reads from standard input +Fixed by Git commit fdefcddbcbdca +JanNeither patch applies to trunk as of todayCan you attach patch for trunk and name it accordingly? was conflicting commit to patch for +Naive fix would be just to synchronize and check that the class is not yet defined +Fixed with it on the latest build using an isolated network with LB provided by NS firewall services by Juniper SRX and rest by VR +I agree that it may be preferable for it to render a button element but we cannot change the default rendering behaviour as it will probably break the appearance or behaviour of some existing applicationsregards Malcolm have you tested button with different browsers? If I recall button is quite problematic with IE +This is very nice +Thanks for the clarification +Make it throw an exception. +Let me fire up an instance to try to reproduce the not working and fix it +This is another step to isolateencapsulate avalon specific stuff +Posted review over on rb G +Fix test syntax docMoved the Ternary operator tests to the class +Nicolas Sweet +Thats why its marked limited private +benefits from it nicely when writing +The initial working patchfiles +This will invoke interrupt on all threads which is too aggressive +The last shot shows how the Preferences window appears as a vertical bar +Hi AndreThanks for helping follow up this issue +Patch applied +feel free to commit hotfix to trunkThank you Jonathan for the review! +Issue closed with release. +go for it i have no time for continuum atm +I see what your issue is here +Option is unacceptable it would make all file based coverages very may be a way but wouldnt you give up connection pooling if we keep on recreating readersOption externalize whatever state you need in some object that is created on a per read call a quick fix in geotools and using +Removing these rules from the session causes the not to be executed +Later. +Hi SebastienDid you have any time to look into our comments furtherPlease provide us with steps on how to replicate your issue +Geronimo doesnt use Commons Logging anymore so this is the same modification as above except with the slfj logger change appliedFrom to this file hasnt changed so it should be fine to apply itExplanation a full +hence I am attaching the patch +Oh right were on right now +Done though I fixed a typo and made the argument optional +should have the same problem +Here is patch for fixing this issueThanks RegardsRohit Rohit Done at rAshish Ashish for committing the patchRohit Jain +zip is not the way how to install the solaris packageIts a helper if one doesnt have root access but has write permissions to opt +Select all but not one can be done with two clicks with the All link hence it is useful. +Thursday is November +Not sure why this was assigned to Rich +This works as designedI created another ticket for adding a new url rewrite plugin functionality. +Closing all ResolvedFixed issues lacking a Fixed Version + made abstract +It should but then its an API changeThis is all going in right? I thought we made API changes across major versionsbq +Patch is Simple indeed +Right? +Closing xml and console issues. +Code used during export must be improved not to remove prefix with letter at fourth position but only with number here +Commited new handler which can be configured on SP side +Needs to use the general deploy mechanism +I have amended the directory structure and added the necessary properties and ignore properties for obtaining Spring right now the project build with some errors using ant only but packages using ant package successfully +Can someone review the patch apply it +Yeah but the API consistency is a good thing +Ive also attached two more files LDAP debug output when authorization works LDAP debug output in the blueprint configuration +first raw patch Ill work more on this in weeks +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Solving as rejected see Timothy comment above. +Linking toChatting w Tom White should probably be or made of References +just realized that the patches from and probably wont work independently of each other and Im not sure how to fix it correctly +Here is the new patch please reviewNote that support for different writing modes has been completely removed +It is intended only for the person to whom it is addressed +Out of curiosity when would the result Object be null? Also casting toInteger seems to depend on your JDBC driver +Could you please attach the current version of your test codeIll try to reproduce the problem next week outside of Eclipsecheers Roland +patch that use the property to set the map reduce the configuration key to set the map reduce name for import to and to branch +Merged the second into the master branch +war file demonstrating the bug and psml portal page containing the for use in cant seem to reproduce this +Ok I see no problem adding a logout event as we already have a bunch of other security events anyway +tests included +Tiny patch +Attaching version of Sureshs patch which applies cleanly to ran compressTable it passes havent run other tests +Got some illegal state transitions too in master log +Good catch David +Is using the unified invoker with the http remoting transport +time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity time remaining limit capacity frame size +As users you can participate and influence the design and implementation at all phases +Do I read this correctly as the config option couchhttpdauthanonymousdesigndoc will expose any design document in a protected database to anonymous usersThis seems a tad coarse to me +Please review +Added as part of the work on and. +As usual it was one of those contract with strict deadline you negotiate in one week have two others to build the result +Log showing the issue still exists in Im not seeing the on my machine could you also take a stack dump while this is happening and attach it here? +BTW when using make sure you properly escape it in the unix scripts since its interpreted as a command separator in sh +Hadrian have you checked in the M release jar of that this package is missingI did a local build on apache camel trunk and the files contains that said converter package +Committed thanks for the review! +Attached quickstart have workaround for this but it is +Would you consider editing the in the xdocs to explain this feature? I think it would be very helpful look in docsxdocs. +Problem is patched patched version is installed on the server +jar plugins +There are default parameters loaded from somewhere which get picked up by that +This patch has been verified against repository version +The timestamps all were long on both nodes after upgrading one node +Reopened as Cassandra configurations in ivy are not complete +It looks like these findbugs warnings are the same that I try to fix at +With a simple spec it wont be difficult to fix +Note that we need to run some tests to verify that this is what Derby actually does +I think that the incubator is a fine next stop for efforts that have already developed some traction +So I did it on all the branches except that +Feel free to reopen if youll identify my configuration too old +Fixed in SVN +Patch for this issue over branch. +Ive updated a patch for this issue +Addendum I used a JSP with the JSP textfield tag +Also seeing the same result with the notifications quickstart +master branchesThis fix makes it possible to deploy the RAHowever in the case it is deployed in AS standalone it will not be able to locate its TM or recovery managerIf the user requires to set the transactionTimeout on the MDBs activation or use autorecovery he must use AS instead +in order to back these changes out +Test files +I just committed this +The file in the zip file was for +The maven website just needs to be updated +jar with nodeps classifier meets this requirement +I suppose it might be possible to fix the datanode side but that would still mean that I would have to set to for the namenode +is defined to execute all Transition transitional actions which would infer that the same state would be current for all of those sections. +Please try this patch +The framework can automatically restore the values to their actual class before they are passed to the reduce function +thanks pat. +Post your observations here +Let Hadoop QA tell us the result +since project names usually correspond to paths it seems more has been a contested issue should not have been included in source control without getting a resolution first +so we the patch then +I have a fix for this in my local workspace but it is unlikely to make it into the release +Should have been marked duplicate not fixed +The updates have been verified in the ER build +Applied the patch to r +Great your regex seems to be even more strict than my suggestionYou do not use and which seems to be better than using them for everything except the separator like I didSo thumbs up for the patch from me. +This patch intended to demonstrate how what constants on OS X should look like +As Ive said in list discussions Im wary of people trusting this functionality too much for security and I dont want us to support this as a security feature as we already have security elsewherePutting this in a separate bundle helps make that distinction and if that exposes a need for additional extension points in the resource resolver that might be a good thingMike do you see a way of moving this to a separate bundle +Apparently there are more changes that need to be made when Claim Space is used +fixed in in release +Lets forget this feature for the +I finally know the build details on Hudson +Latest dev build should include the potential fix for this can you try it out +thanks Ivan +Anita +Fixed in rev I think rev stops the scroll timer prematurely +Still it should be fixed +Although a little wasteful I think that this is a valid change to make the API more consistent +the thing i dont get is if you create the znode again on the leader and then delete the znode it should disappear from all servers +If there are still errors that need fixing please open new issues. +The database creation time should also be shorter but I havent obtained any numbers on this yet +patch for only the s branch as trunk refers to those properties in the defaults +I hope the way I fixed this issue wont come up again +As Jonathan said this should not be automatically done for all domain classes but only if you wish +moving to +Assigning to Hasan whoul would like to suggest another implementation + +Onima Kasliwal Kirti Dhakad Surya Kusumakar Harsha Chadhar +Is it OK to have same text in both columns or we can have no text in second column if is Healthy +Im going to up the partition size allocated for tmp to see if that helps +duplicate of you are requesting specific language files then several have already been contributed +This little patch would resolve a descriptor before trying to read it ensuring that it is available in the local repository before reading it +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Also historical notifications have been sentWe still investigate why it happened +Fixed on both branches and unit tests updated to test for worker thread priority +Theres an issue though +This pushes the content down a little further so not to overlap the header +File with all the needed helper methods and name mapping dictionaries +Thanks Rick +The init without the partitioner parameter is unchanged and does not set anything specific related to the partitioner class +A new patch updated for trunk +Ill clean it up if you want it here or before I commit it +Would you mind packing the proposed changes into a patch and attach it to the jira making it easier for us to review and apply your changeThank you very much and kind regards would suggest to enhance the batch to also search for HKLMSoftwareJava Development Kit in case Java Runtime Environment keys do not existI just installed a JDK on a bare windows and had the above configuration +bq +Please increase the priority of this bug thanks +Im hoping this will be fixed on the next branch +I will not have a chance to look at this issue at least for now +I committed a fix on trunk r +Same approach as being done in to support cygwin +Reopening to move to as was issues after move to in +Id like to know why do you remove some of the because they are hibernate dependencies not used directly by this project +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +currently what we do is if any change happens in the element collection we send to the server +Per Keith Turners comments in would it make more sense to host this submodule on github +Nicholas does it seem OK to youAlso might it be reasonable to mark just is +An exception during flush processing in now throws a +Theres no obvious reason this is Trinidad doesnt affect houtputLink +But does already have image for it it is just need to be set +I disagree about allowing separate symbols to control each operator individually +If its alright with now depending on than Ill move all the code back into changed code into +DEBUG Rejected bean name no URL paths really +it just silently ignores the block +Committed revision in trunk +Just seeing the patch I also got the same doubt as Ram mentioned +If you enforce the max during reading you wont be able to process this data +Need to make sure all the work is done before user gets control back +Applied to branch +Thats aweesome +The reprojected poultry layer +I should probably add that all existing usage of the class still functions exactly the same way +Its not a problem with the RPM +Ive been working on some changes on friday to update the metro endpoints as well so I will create a patch with those changes as well +This probably related to the recent changes +Ill do it myself later if nobody beats me to it +rejecting the feature until some reasoning is provided why the existent functionality is not convenience would then apply to every management aspect of Wildfly which would result in a lot of deployment processors that do nothing more than replicate the management ops of the subsystems +Maven can be another new tool for the developer and hence nothing should be assumed +its ok +We decided to include it in although I removed it for +I reviewed and committed this patchmsvn commitSending javaengineorgapachederbyiapireferenceSending javaengineorgapachederbyimplservicesmonitorSending javaengineorgapachederbyimplstorerawSending javaengineorgapachederbylocSending javatestingorgapachederbyTestingfunctionTeststestsstoreBaSending javatestingorgapachederbyTestingfunctionTeststestsstoreOnSending javatestingorgapachederbyTestingfunctionTeststestsstoreOnSending javatestingorgapachederbyTestingfunctionTeststestsstoreReSending javatestingorgapachederbyTestingfunctionTeststestsstoreReSending javatestingorgapachederbyTestingfunctionTeststestsstorebaSending javatestingorgapachederbyTestingfunctionTeststestsstorebaSending javatestingorgapachederbyTestingfunctionTeststestsstoreenSending javatestingorgapachederbyTestingfunctionTeststestsstoreloSending javatestingorgapachederbyTestingfunctionTeststestsstoreonSending javatestingorgapachederbyTestingfunctionTeststestsstoreonSending javatestingorgapachederbyTestingfunctionTeststestsstoreroSending javatestingorgapachederbyTestingfunctionTeststestsstoreroSending javatestingorgapachederbyTestingfunctionTeststestsstoretestsSending javatestingorgapachederbyTestingfunctionTestsutilTransmitting file dataCommitted revision +The cleanup of the other files is intended +Hi Damien the link you sent refers to writes +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Its very ll go ahead with my work using this +Nothing relevant for NN +Appears to be fixed in v cvs but targetting next release as the rewriteshouldnt exhibit the same failureRob +But one query wants the data to be sorted by ts instead +Thanks Arun +commented out incomplete stuff for now for +A bit more screenshots +Please review! The same file should work both with trunk and branch +If it is sorted on another column using the db is not easy +So youre saying we have to redundantly post on EJB dev forum that were doing a task just so adrian brock doesnt yell at us. +Just noticed that nothings been done with this JIRA lately +I suggest adding a new origin argument to +But since now my commit is there in asfmaster you wont be able to reproduce it in asfmaster +However I did try changing the timeoutwait +Please change to if you intend to submit a fix for. +just an idea Ill try to bring up an Version should at least display all available dataMark would you mind to provide a Screenshot of both and your +Closed after release. +seems like this problem was introduced by +override +Found another solution +I think our issues belong to the same category so a generic approach works best +Please let me know in case you need anything else +I think I need backward compatible escapingunescaping methods in the Text util and the provided test case should also work +Vote with the release. +Unless Im misunderstanding the workaround this suggestion does NOT constitutea useful resolution to the problemPlease provide a direct programatic mechanism to add References to binary fileswithout reading them into memory in their entirety +Its been blocking work on for several weeks now and everyone seems to agree that is crucial for the next releaseDo you have some specific criticisms of this patch or is it just that youd like to go further with the changes to make it more like? I dont see how committing this patch now precludes further work on the API +patch with improvements mentioned in the patch has been applied please verify +jdbcsrctestorgapachehadoophivejdbc shows using hives jdbc driver in embedded mode +That revision is across the entire Apache repository +Russ had verified this +forgot to svn add +A start at making for a more fluid design +for that we have in where you can have custom borders around component. +Note that these are intentionally not targeted at because they dont belong in the release notes +But maybe we can add some custom code that simplify this stuff removing dependency based on the code that provides on apache harmony +Do we want a test case of handling two har files at onceOther than that +Ive committed this and merged to +Any plan on fixing that on the branch as well? Having the fix in means it wont actually work before the first stable release of +Could you close it please thx! +The problem here is similar to where an instance variable is modified by concurrent threadsId say s Attach method is evil and s mwriter should be readonly but that would break too manyOther than with I dont intend to keep the instance member and add synchronization but will rather remove the instance member completelyI dont think the stack would really solve the problem maybe make it surface less often. +So Im resolving the issueThanksSupun. +small nits compute Can we update the below comment it still says in +also adds a few distance metrics for clustering +yeah can get hairy pretty quickly +lots of load and the CPU was running at all the timeRegards Noel +yes those dependencies would need to get looked at and ensured whatever API they use would be available in both versions of hibernate tools + +Will we able to claim fully supports CMSynergy with out resolving this issue +Ive committed this for trunk in rev +we are bundling as is +fieldName contact type +Are the others supported also? I can check +I would say its up to the container to handle such attacks dont you think so? Unless theres something I dont understand which is very possibleMaybe a definition problem in this case it is not a attack only a jvm crash or a denial of Tomcat point is how to know that the server is again alive? If cargo is not in debug mode we dont know that the JVM with you for the container +Anytime! +Ok I understood this now +I think it is okay to match the Java type with the protobuf type +chunhui Could you respond to Stacks last comment? I would like to get out soon +Would it be possible to display be default the Lines of code metric? Indeed otherwise the widget doesnt display anything as long as a metric hasnt been defined in the edit mode +Im on Markos line hereIf you are to compare with stemmers consider that these creates unique tokens that does not interfere with semantic meaningsWith the accent filter running the Swedish word kn through the filter would create kon + +Reuploading the Darynyou teach me so many things +Fixed it so an error will get thrown if the state tracker attempt to associate resources with closed out Transactions or Sessions. +I will have my friend play around with the Orca image some more based on everyones comments +Charlies patch made minitest work but the underlying problem reduces in this bug still exists so I am reopening +The names should have instead of charactersThe following patch fixes it +The semantic for this is complicated but in pattern matching I think your reasoning is correct + has been marked as a duplicate of this bug +I was having a bunch of instabilities trying to breakpoint and step through using the debugger and other necessary work reached crisis levels +Ok as the added test project says it works lets say it has been fixed in Ivy probably with. +It doesnt cause the whole app to fail if the connection doesnt work +Attaching patch for presume it also affects right +Changing to Minor its just a cosmetic glitch +this would be a good thing to include in the next release if we can +Thanks +pushed to passing. +We are now able to switch compilers but a restart is required +Hopefully that makes sense +Is it possible to fix this also in the branch? I am afraid that I cant migrate to JSF in the next it is possible can you send me a patch for or a hint where to look that I can patch it myselfThanks a to Alexander the custom state manager must be a subclass of +See isTemporary +a build issue +The XML is written in the default locale of the driver which is the systems default for the +attach that forum thread URL to this jira otherwise A +Also to note its better to rely on the limits file than the ulimit command run by the launching user the formers more truthful in face of PAM config oddities +Thanks Jarek I applied the patchIssue can be is in work is done. +appears you have the purple going light dark from top bottom as I intended but the green and blue have the dark color at the top +The patch resolves the following issues +This allows us to resize readwrite pools without restarting +This is superseded by. +I am looking at it I got a junit test case to reproduce it +dist puts binary release artifacts in the dist directory that test runs unit tests and clean removes files generated by the other commandsIf ant is replaced with some other build tool then the should be updated to invoke the new tool rather than ant +yeah right +This change actually shouldnt impact the Sling feature only the standalone version +Committed. +This is interesting +I can speak from experience that this is perspective we take when trying to explain all these reports and their relevance to others +I already implemented this +Attached a little addition to you patch which fixes help message +If santurio is not available then one can use the JDK provider +add unsafeRead in to avoid closewrite on it +Sory Carlos but I dont see the relation with issue raised here is when a parent project use a plugin without specifying dependencies the submodules can not add dependencies to this plusgin +bulk defer of issues to +it +So can you please explain a little bit more in your comments on submitting a patch? In general what procedures should we take to fix issues on axisthanksXiaofei +We are currently stuck with this issue? when will be releasedThanksBy the end of the monthColm +the patch on both trunk and branch +This action is only available if the user cause trigger exclusion is enabled +IP bonding where you can a single virtual interface over two real network interfaces if one is pulled the other takes over +But the existing code handled more complex situations so I figured Id preserve that feature +I believe this one is simpler to fix +Are you compiling with debug symbols? Also are you using the Sun VM +I have to mention that I have used the css file and Img folder that clover usually generates +In these cases tries its best to ONLY drop the data for these unavailable regions instead of slowing down the overall multiput throughput which may cause more data dropping since the client fails to buffer any more incoming data across all the regionsAlso there is no function if the put is finally failed after N attempts +So the error returned by the compiler is correct +Then how to best to parallalize things like the initialization loop +The implementation has reverted to the use of the partitions rpc support +Updates from Sebastien and further memory tweaks to reduce GC thrashing in AB idea that may further relieve the cpu bound behavior a little bit is to use single precision arithmetics +Minor polish of a debugging code inside be a good idea to contact aruniyer who maintains the FISH implementationIt is essentially built upon pig split and custom loader +bulk defer minortrivial issues to +Though the collation appears to also be a big factor +For my opinion design have to be simple but impressive +thanks +But since we are limiting retries I think it is safe to let the renewer retry as you suggested +Fixed as proposedsvn revision Thank you for reporting this issue. +Would be nice if different parts of the namespace didnt require synchronization because they happened to share files with the same name + +the I think this should be in +If the interface had been an abstract class we would have had more flexibility to deprecate and add +is the updates here +two profiles inDo you really think that CLI param order is something unstable? Usually the builds are run from some CI tool or at least a script anywayPlus this would apply only when the profiles actually overlap +The failures are fixed with the latest jbossts update. +A couple of things to note +Verified on r. +exceptionstacktracetestcase +Id be happy to help on integrating Apache as a caching providerAntonio do you already have a ComponentService I could leverage for that +Please review and comment +No new tests are needed since this triggered by a unit test +for the review Eli +The context class loader should probably be the default in the enterprise configuration +fixed rev . +This code is part of a document management app +Suresh thanks for taking a look of my patch! In navigablemap! null check not requiredYou are right +into an Java byte is definitely problematicPerhaps the right way to do this would be to convert values in the range +Sounds to me that it was implemented similar to isDefined but it should be similar to we should leave the existing empty function untouched and change the behavior of change the description of empty dp something likeThis function is deprecated use instead the function isempty +Ive raised as a more general but related point +Thanks +As cant be reproduced in Hudson closing. +Committed to trunk +Perhaps its an eclipse bugIm on +Hi MalcolmShould we add the method copyFrom copyTo as suggested by ToreJust checking if this approach might be better with upcoming container conceptregards think this is a different problem than the container concept but I havent given this much thoughtTore has concerns about the tight coupling between and +Ok Aaron but if my understanding is correct if you want to cover the second use case this is for instance to compute the code coverage of module by unit tests contained in module +Apparently this can be an issue due to transaction timeouts depending on how the TM implements timeout reaping +prevent the issue from arising againI know this looks like a simple fix but we really need integration tests to apply patches or quality will just for the comment and the explanationIf no one else jumps on it before Ill try to provide an updated patch but I cant commit to a date yet +I havent included a patch although I can generate one easilyAs a matter of course javadoc references that arent included in an import shouldnt be qualified +Close one more time with the correct list of affected versions. +Also v hfiles use the same synchronizing logic for meta blocks so we might have to fix that as well +looks like should work with +Any other changes should be handled through new issues. +We can balance the load by having multiple partitions per broker +Committed to trunk and. +Set to to avoid this problem. +please also commit to +I moved the changes from to since the code has been refactored +Investigating the reason I figure out that your framework user can provide his own setting a custom exception mapper on context ignoring your wrapper +Ill get back here when it is clearified +So for now I recommend remove all not actively supported features and start out slick +Solr and were already upgraded to before this bug was opened +I am not sure I understand why moving jempbox code in pdfbox +This should now be fixed in its not really fixed now +Hi Davidyour patch is in release branch at rev +From this information we will write the final release notes for all issues +Feel free to review +Thanks will do +Btw I was using trunk r +I got some idea from your change so I create a bundle which could be used to replace existing JMX connector so that you could not hack Karaf Management bundle hope it will help thank you for the suggestion I am not sure what is your intention here +Has anybody the same problem it also happens in +Do you know who I talk to about that +Its fixed +Ah well okay go aheadYou can add a similar test as suggested by Armin too then +No I dont think you can assume that +I dont have a listening framework on so Im not sure Ill be able to fix this on +This will be fixed by refactoring the Maven project builder +But it returns immediately and I dont have any feedback about startupThat is clearly not what I wantMy workaround in the startup script is thisI exeecute jsvc check and exit if there are immediate errors such as commandline errorsThen I enter a polling loop +bq +In branch getRegion only get region from meta table? It is any problem when root region server crashed? we reassign the root region +It might be nice to bypass the sleep if nothing is going to be streamed +As is not uncommon in XML documents the use case is extended information that can provide for enhanced experience if the information is understood by the client +They are in the attached patch see what you think +As you say perhaps with bundle extenders we wont need to add any anywayI doubt it would be a solution but would it also work to ask people to explicitly export packages in the app that are used for gbeans +Perhaps something in your setup? I use Linux and Maven +Any progress? +I am attaching patch for conversion of to junit test +Pushed on I am going to close this one. +yes it does retrieve from the resource need to check on trunks but i am not aware of amy changed related to this +Looked into that project but I am wondering would it be better that we have the doc along with the idm project as these two have the same release cycle instead of a separate doc +Do you mean for this to be a part of the spec or is it just a optimizationThanks thoughts on a possible change to the file format to add the size of the blockI have a real world use case now for both concatenation and codec conversion +field number class +Thanks Milind. +Fixed it on as well +I agree Ive spent too much time on this issue already! Ill with at comment describing which tests fail with which version of htmlunit +Joes workaround fixed the problem not sure of the impact of this but so far all our tests still run +examplesportalcurrent compilation time is s +Committed to trunk +The policy to choose the big table should take the average size of the partition into account +Use the generated id based from the alias name and tag position in a page seems to be the best solution in this case. +We will be doing some additional testing but we might rollback so that we dont leave trunk in an unstable state +Ivan can you please specify what version of meclipse you are using and what Eclipse version it is running on? Also it would be a huge help if you could attach test projects or Eclipse workspace that would allow us to reproduce this +The latest patch fixed a MVCC related problem +A file cppsrc is added to contain windows version settings for all projcects or for individual projects +You could get around it by turning off required validation in the JSF controls and instead using Web Flows validation +Though we removed dependency of Job Parameters in Job Instance on the model level it is logically still coupled as we need to pass in the execution parameters everytime we need to find a job just cannot consider it to be clean up +Yes this is a valid enhancement request and I dont mean to brush it off as unimportant but there needs to be some fundamental changes to the way is producedYou give the example of the Java compliance level +What is the status on this topicHas someone found a workaroundCan I help somehow +Patch applied +Currently I have this resolved in IE needs testing and Chrome is broken +This issue has been fixed and released as part of release +When I had been cleaning up issues after migration from bugzilla to jira I asked Keith if this bug can be closed +The other warning for inconsistent synchronization of is due to existing code I dont see this warning with on my machine +Final version of the Tomcat setup templates +Its his patch +The only thing I see is that I have a couple of improvements for the imagemosaic plugin on my machine that I have not merged yet maybe be them solving the issue? I dont think so +I just want to get something committed today to fix the failing tests +My Conclusion of the test is Multiple connections makes sense only while talking over Multiple DCs +it is extremely slow which sort of makes the whole windows thing a bit moot +on latest +The above would imply that datanodeimpalaothers would be running as yarn containers so that they can be targets for delegation +One is a wtp model though deprecated +This does not have a unit test yet +Hi Paulex Your third patch has been committed in revision +Patch implementing changes I proposed +Is there any update on this before we finally change this property +I dont think this is related to issue it seems like the leak might be ocurring in a core component of pivot the transition that uses the uses one decorator for all its work it appears to me that the resources used by the are not being released by the vm as I explained in the description of the bug +im not talking about our ant build im talking about that when we add new features to the tools we should have a way to expose them both in the eclipse plugins and the ant toolset +Thanks for pointing this out! Ive added an eventListeners bean property to accepting listener typeobject m not seeing this in CVS yet +Sorry hit the wrong button! +Thanks very much +Nit remove the period ending the count javadoc +It succeeded +This looks like its coming from the JVM its definitely not from +Along with the JAXB model changes that took place before this release +Ok couldnt resist poking at it the is fine I that the file contains the right contents +You cant have a role defined in and not JAAS and still give permission can youThe case is the typical scenario for defining a role with no mappings +Jody and Andrea should I do anything further to get the ball rolling so to speak +Solution agreed with Jaikiran and Stuart was to simply ignore the static members wrt injection +Im not really proficient in garbage collection so I prefer to mostly keep out of for the time being +Applied the patch with revision +Here is a patch to fix the problem +We also need to propagate exclusive etc +Burying it in About means fewer people will find it +Oracle Sybase and the Derby network server +I kept the alphabetical order as it seems the most reasonable to maintain +The timeout is unrelated +Verified +The patch havent been pushed toAnd my suggestion is only the reproduce steps not the workflow steps +Thanks for the patch Arup and for going the extra mile to improve the tests +For instance Im benchmarking whether we can avoid what allocating the arrays for storing the similarity results so often +Thanks Hemanth! +Actually I dont know if the proposed solution is really a very stable one since the number of IDs and the speed the ID Generator is out of date heavily depends on the loadsize of the system +Committed to trunk with revision and merged it to the branch with revision Unless the ongoing discussion brings out the need for more changes I dont expect more patches to go in for this issue +Confirmed at r +Ok Ill try to fix that diff problem and let you knowIn the meantime given that my implementation is a replacement to the current one may you try to completely replace current source files with the ones contained into the V zip fileThanks +To be clear this patch improves reads by for single nodes only but by when a single nodeNo the patch has a measured improvement on reads +Sources compile fine and the functionality looks good too +If we know what they are maybe we can come up with a solution. +Resolved by styleClass isnt applyed correctly width attribute stopped to work +Ive tried to discover why this was reverted +But the patch should be just fine +The issue is reproduced with GA even with configuration set as provided +Thanks Simone four your work on this next step in order to improve the XSLT transformer. + +Is anyone looking at this +If the used a to wrap the before adding it to the chain then would destroy be called +All the information is in the anyway +would be a cleaner way of dealing with situations like these but in the meantime the patch should be OK +Will investigate +This adds javadoc to our public JDBC classes in the client +This also applies to s yacc I suppose +I havent found a way to work around this yet +This behavior is so strange because a navigation event should indicate a halt of execution so anything that happens after that point is really just undefined +Unless you propose a better fixThanks +Both are quite a hack +Closing resolved issues for already released versions +I hit the same bug withMojarra is a major show stopper as I dont see an easy workaround +is hitting this same failure +ok +Removing fix version of for those that are not in branch. +I have tested this patch on few cases like hdfs format and upgrade etc +Attaching an example of the bug triggered with a quickstart project +AllenI want to clarify that this is a feature branch in which the code is being developed +It seems to me that it is desired if this method would search all the imported included schemas +Im further looking into other areas for improvement +This will help with some newbie errors where the DFS appears to be running but in fact nothing works right because there are no +The only thing missing is the ability to issue an alter to add an arbitrary index +Seems like a good idea +Sorry issues were mine +toolssoa should suffice IMO +Hello do you think this problem still exists in the code +Do you need a jar file to try out if it works nowRegards RobMy plan was to commit the changes to the branch after you have tested them Regards this fixes the problem that I was seeingThanksRob +for instance. +Latest patch added test case +Passing along to Michael +Attached the patch to update the testsuite as well as create new files +It should be assertNull there +If new rows appear in the meantime the action will be applied to the wrong itemIf the were based on the model objects an action on model ID will be applied to model ID regardless of what other model objects are in the table when the form is eventually processed +Patch to address the issue +I believe the basic problem with defining the datasource through the admin UI is the same as described in +Im still being open minded about it but so far in my experience about of the time people want this it means that they just arent understandingusing hivemind to its fullest potential +When the node heartbeats back in the RM will direct it to reboot and the NM will simply shutdownOnce is addressed the NM will restart from the reboot directive and the node will recover +The test case analysis will have to remain within the uploaded bytecode to avoid false positives. +Reopening +Added license header to please make sure it still works with Flash + +will be removed in favor of see. +Will look into ll test whenever its needed again +Hi CyrillePlease see the attached patch it shows the code I applied to the trunkCan you give me a favor please and verify on your end that with this code you can see the counters for JAXRS operationsAt the moment I feel this minor update is the least intrusive way to go forward +Did all of the tests fail? Ill check it out tomorrow +Yeah Ive been working on it periodically as I work on Locationmaps +The code looks like the Boo samples Rodrigo writes +It is created just because publishing cliet needs itBut when you look at publising client code it does not use it either +Attached latest patch which have the following changes +I removed the unused import +closing then. +All right +Without appends in hadoop the edits to the ROOT or META could be lost and so the master would not look for the log of the server that was killedThis is not a good way to do this +With only exception that we shouldnt overuse this facility +Juergen added you as developer for the Infinispan project and from now on can be assigned to you +Please let us know if it isnt and we can reopen this issuethanksdims +Lets see how that bears out. +Ill commit this soon unless someone objects + patch looks good to me +Splash in question +We had some backward compatibility issues at Facebook but we will solve them in a different way +What about if the profile theme attribute is null for a particular user then it will use the default theme? This will not impact existing users in the point is if the theme is deleted afterwards the users who selected the deleted theme before should use the default theme +Sounds OK +The triggered action seems to expect a row which just contains the columns necessary to execute the action +Two of them I see are Use Maven instead of ANT to build because Maven is so much better +Closing this as a issue. +Thanks +Ill see if I can correct +This is now fixed +Question becomes how long do we need to keep it around if its a matter of support +Must be a bit specific issue +While there may be a bug here that can be determined with further investigation I would recommend that for the requestresponse pattern that you use a topic for the response and use to collect the response +called from am using the component and the component +Changing order at fetch and index time to this is cumbersome +HiI think you should discuss your configuration on mailing list +Thanks Jonathan +Sorry about that +I reverted the changes in the and configurations +Ill go searching at resteasy now. +Unrelated to this change though +Similarly for all should start from JT could have a better name is scheduler property +The JAXP Expert Group discussed this issue at its conference call today and decided that xsloutput and the ies methods only have an effect on true serialization to a +Now when I use Batik beta it doesnt work anymore +basically conflicted with this patch couldnt help it +Lets see what happens on next build +If it is still a problem would it be possible for you to attach a cut down app that demonstrates the problem +checked into and trunk +Is this intended to replace +Thanks Im resolving this jira since the attached plugin has fixed this ll close it once Ive verified it in ER +There are essentially two approaches to address the problem +We need to improve the configurability of Planet +rnewson Im confused +I must have been frustrated when I created the ticket. +Ill add another comment if I can track down any more information on the issue +Hi loannisCan we do this in we will need your help here +Created to figure out the exact semantics of auxpath per Pat +To keep the old behaviour in tact of using the class loader added truefalseso if user supplies this property in the URL post delegation class loader will be used in loading the embedded +Closing this JIRA as a. +Please confirm they have been applied as expected +Closing the resolved issue. +stacktrace was before the job the rest are from after it duringafter it abortedStacktrace and stacktrace are from solrIm fairly certain I have the correct PID +Closing the bug as I couldnt reproduce it on the latest build +Heres a patch adding the CDI facetnature for projects depending on Seam The JSF facet is also added for projects depending on +I do now config my deployables in the deployer section and no more in the container config section so deployables will now no more be deployed on startup of conatiner we do need to deploy deployables added to the configuration on startup +I would guess this has something to do with the method being a fast operation +patch applied +Note this patch contains new files which will need to be svn added +Andy GaryDo you want to take a look at the latest patch +We will add the code to preserve state in the next version of Cordova but this will most likely not resolve this issue since it lies outside of Cordova +Is NOT present +Though there are couple jiras opened for better integeration with WTP enabled projects. +In SVN renames property as +DONE +The response came back from WC +If you see any related issues please the issue and Ill take another look. +Out of Date +Thanks James for initial patch and testing +And please add tests for this preference too +I tested on my cluster patch looks goodIve just committed this +Hi Regis Patch applied at r +Thanks Dag +Shifting to Beta +The unit test probably ran in a different order for youbq +Thanks Vinod +JBDS shows the correct number in the splash +cant you update the server? you have to change only the adb generated code +Things you need to do for many DSL scenarios +I have removed the properties relating to keyword extraction and changed the keyword extraction logic +Wrapping seems like it might be handled as an independent issue especially if is a priority in the near term +I was able to sort out some other problems and got the the point where I could check this changeIn order to get srctimeput to compile I had to define the daylight varaible at the end of the RWSTDNOTIMEZONE block as no such variable exists in the global scope +It does not accept inserts without arguments so I had to copy the class and insert the keyword DEFAULT as described and it worked +Changed on trunk renamed to Scripts and wrappers updated to reflect the change. +Thanks Harsh soon we will have found all the genereated and numberators of the hive world +how many pages diagrams etc When will draft material be ready +Now it just stays dumb till the progress reaches the threshold valuesSorry about that logj jar was missing on my classpath +fixed in trunk as of r +Patch for bw branchI will commit soon +Of course +Here the rl text must be presented in boxes in the proper order +In this dependency was required +for classes annotated with and not annotated with +Full correct patch uploaded +I think we should go ahead and commit this to the branch and any polish we want to add we can do in +Class prepare functionality improved variant of the same gcv functionality +Give me some pointers to proceed and I will try +I just committed this +Sigh +Do we want the Vector to be a Vector or a CollectionThe patch looks too simple to be true +Can not set the port as parameter check for to allow patch based on the latest trunk +This issue can be closed +Assign to during tidy up prior to +If you want me to sign up as a use case I am happy toThis is exactly the kind of concrete I was asking for not some made up which may make sense + for files use part file status for dir use +Test case that fails with current River and passes with the proposed timeout solution applied +I just committed this +I think you need to choose another one +Not sure if itll be a solo bug after we decide on a strategy here +Yes the issue is resolved +We reboot jira daily so since that GC feature was alreadydisabled in the live config and we have since rebooted it is now turned off +Ill ask infro to chgrp the OJB files make the wiki and to close the mailing lists Im just making sure there are no objections to this in the PMCThat leaves only the announcement to be done I think +Could you take a look? Thank you in advance +To add +Sometimes upgrades only address the slave nodes so theres no sense having to change the NN +Applied the patch +Committed! Phew! +readLink is only ever called by get so the check is done there +And I may file another JIRA to address the rest of this +Patch changed as per review comments +This was fixed somewhere along the way +Only warnings to go JDK standard to implconvert package after the client validation change +It seems that right now jenkins will fail to compile any new patches without this patch thus it would be great if this patch can land in trunk as soon as possible +It contains the info for RM to allocate container +It would be good to get this commited soon so that the build gets up broken + +am quite surprised that all it takes is a! Looks as if wefinally have a good architecture for connection handling Good work guys +Just go ahead and hack it up in +Is there anything Ive missed or need to do before this patch can be reviewed +it will require yet another api change +Fixed in nightly build +Logged In YES useridIs this bug really fixedIm getting the same error usingWindows XP SPEclipse Username hansd +Otherwise were done here + +Imagine that you have users and you can split them into different profiles and you want each profile to see a different variant of the pages +linked to related discussion about package do we need to preserve history from git? please do preserve it! It would be a shame to lose that useful information +I applied it +Im ignoring the links stuff but not advancing the parser past the value of that property so its not getting to the properties +The issue was not resolved by the attempted fix due to complications with xhtml and namespaces +Steve +This isnt a good idea to do on a sweeping level per Joshua Blochs Effectie Java Item it breaks the principals of encapsulation and what that means for maintainability over time in a framework +No Release Notes entry is required as this issue has already had an entry prepared +Good catch +looks OK to me no tests for any scripts in trunk ant manual verification that the trailing spaces are missing in the patched source. +Patch checked into trunk and branch closed with release. +It may be that the extra processing done in some of them wont really benefit until we get readers out of the way of the write path or get rid of the couchrefcounter serialization point that could be fragmenting our group commitsLots of testing needs to be done +I can come back around later and do a more thorough job but this isnt the only analyzer that needs some documentation improvementsThanks for the attitude! committed this javadoc improvment thanks robert! +If the throttler parameter is null throttling is disallowed +Create an alias entry Best Fruits references entry favorite +No vacation wasted in fact no day wasted night time is when I do my best work +This PFB LITS method is a good quick fix solution in advance of a more general study and refactoring of your business objects as Juegen suggests +icuj updated to version in both and branchPlease cross check and close the bug +Some doubt in my mind about the validity of some of the tests but still working through those +However I see that no is in xs LMP and defaults to which already changes behavior so I think we can apply this change to x as well +If this really comes from the addAll method Im shocked that the JDK compatibility suites are not capturing this works indeed fine with jrockit and the sun jdk but I believe they share more of the same implementation code unlike the IBM one +committed to and trunk +This is a duplicate of. +I hope it was correct to say that the error is likely to be rather thanThanks for any comments +Sorry other Ryan meant to assign it to myself +while deploying it clams that there is no endpointit seems it doesnt affect eg +I was out of station this weekend so could not address umas comment in time +This patch improves per record read and deserialization time by in my tests +causing the command line to ve forked the into an and packaged it into the changelog plugin to make it quick to test the changes + +Thanks Lukas for the contribution. +I can also use the avro tool to view the data files in the json format +Patch is in trunk rev +fixed on the changes made on branch back to HEAD +Attaching simple patch +Was there performance comparison between checksum for vs +Thread two then adds the same document with id and it succeeds +So to proceed we will have to commit a new wps model and change the internal +fix as you see fit Lars +The failure might be due to other test failuresThe other failed tests and the s are not related to this +Freddy I dont see why this ticket relates to upgrade of findbugs +Ill submit for the QA bot after is committed +Setting to future pending a patch +Have you tried reproducing it? I have been hit by it until I moved to exclusively +Applied the patch into trunk branches +Looking forward to it +Need to reopen +If we extended the editor to include functionality like splitting and routing then it might make sense +Also applies to Bundle and Filesystem Resource Provider can now be resolved again after fixing and testing this. +While it was broken before its a small enough fix that we should try to get it in +You can only delete or create a key +If found use it +Youre welcomeIm closing this issue. +This is done for documents but it would be nice to have for other resources too +does not utilize the Digester so these errors are not logged there +the waiting plans lock should not be held while processing a synch query + +A fix has been provided. +Update so that it works with the in release. +Any change in the like adding or removing a pack requires us to manually update need to fully automate generation of the when izpack compiler is run +Thanks Roman this looks great! Could you elaborate on how you use this feature? Also is it basically equivalent toIt would be good to add a unit test for the new logic in +If you moved the assembly as its own module then you cannot use moduleSets and would have to use dependency sets which youll have to specifiy every single plugin and feature in the assemblys pom as a dependency as well redefining which subset of those you want in the assembly descriptor +Sample project for showing compile hint in checked the tld and and it is ok +Latest iteration after addressing comments +Proposed SOA TP Release NoteESB helloworld no longer fails with Tibco EMS as the JMS provider +Patch added +Performance is not a real issue here because is slow anyway +Thanks +Thanks Lily I will commit first thing Monday if your account does not come through by then +Note that the existing build compiled the wicket classes directly to wicket +Dont know if Ill find the time to backport tonight so feel free to do it if you have spare cycles +You might have better like on debianubuntu you probably need to install a missing erlang package +crash isqlhavent finished looking at the codeChanged some deprecated functions to pass control to newer equivalents when calledModified how some error cases are being detected +However that was commented out on the new zone serverI will fix the other cronjobs to use full pathnames +In Addition to the problems above there seems to be no way to escape a plain trying the obvious results in +Patches and documentation have been updated +Ive update the build and this started working correctly +Japanease messeage resource file an patch files stc +I dont like this +Repeating the footnote on every page creates all sorts of problems what do we do if a footnote doesnt fit on the current page and has to be deferred to the next one? The next page will end up having two identical footnotes +Ah sorry +Paul havent checked javadoc but can you make sure that differences in defineConfiguration semantics between AS and Infinispan are documented ASs defineConfiguration implementation if theyre not already? +Changed to M and set the conflicting issue to the new ticket +Otherwise it had a failure +is also referenced from +So fell free to work on more patches +It was an old patch and looks like there were conflicts +Status update Tair? +i forgot that i already added this +If we could separate them that would be ideal into two projects possibly contained within a master project for ease of builds +So i take back this ticket sorry for botheringbq +with rs only selected rows would be lockedtx db c select from ck where id with rr use and keep update locks success U lock on the tabletx db c select from ck where id with rr use and keep update locks U lock wait for table since table already has a U lock from tx and U locks are not compatible with an U lock so waiting for tx finish +is currently only required for security and further is not yet a stable release of ZK +Thanks Florian applied to for no new version + +Id like to second that thank you +I prefer this indentation to the other since it distinct the fun from other level +I already applied the latest patch to the branch in r +Great evaluation PaulWhat wed need is a unit test to reproduceI will be working on this issue starting around May thIf anyone wants to commit a fix before that date feel free to do so +looks like its already been fixed. +Committed revision +A message containing this exception enables the client to cleanup to prevent a hang especially if a timer is not used on Process call +Post migration Jeffs script will need to be run to fix the copyright checkbox +There is which talks about defining the engine interfaces for use in network server +even if we cant merge the code and make it autodetecd dates using the word range to describe the type of faceting seems more clear then number +There are a could of things that are confusing here +This will cause MS XSLT processors to fail in firefox the document will transform but imported types are not broken down and the imported source for each XSD is empty +since this wasnt fixed +Thanks Steve! +Adding the ability to counts at the end of a transaction if a threshold of log is met +Ill switch to + I was thinking exactly the same thing +Havent had a chance to try on a cluster +You will get a notification message that there was an issue with saving your process model but you will not loose your current design you are working on. +outstanding issue +Well if thats teh case why not link to this page from a more prominentplace like +The default value of an EL expression is this expression itself +I believe the issues mentioned here are fixed as of rev +Ill have to rebase and to get a log message Ill see if I can find some time in the next few days +No need for IT +Emmanuel is this the right way to do from a MINA povBut another question is that what happens to an already existing SSL connection? wouldnt it fail because of the new certificate? to reload the SSL context automatically when the certificate of uidadminousystem gets updatedNo clue at the moment how to do this in an efficient way certainly any check put up in the interceptor would be a over kill IMHO +Changes committed to the branch +Everything else looks good to me +In region load is reported to master through heartbeat +Thanks again for the suggestion +I thought all that was needed was to produce data of the correct size approaching K +Work committed to trunk with revision you commit this to the CP branch too? Change the Fix Field if you do +Yes but it happens on the mapper +Maybe we should throw by default and provide some config to do something else such as provide a a file path for temp files? This could be in tmp if you chose or map reduce could default to its temp directory where it is storing everything elsePerformance is clearly not an issue if this is truly an exceptional case +This fixes the crash on fast cache scan canceling +Implemented this support in Rev +Of course I would still like the message to be modified to make it look like an intermediate step in resolution process +tagging for review for otherwise we should reassign fix new fix version to +Removed redundant test resolved issues for already released versions +Closed after release. +I am having issues getting kerberos and hdfs to play nice in a environment so Im going to throw it up here if there is someone who can test it in this condition faster than I can +New entries have been added to the ant build script to download the files at compilation and place them into the build directory +It was using when it should be have fixed this as well as a few issues affecting the GWT s ability to create proxies for entities using inheritence in commit you James very much appreciatedsimon +Topic is treated as a experiment ID in out current implementation +Sarah are you planning to add some test coverage +Using an oracle DB on killing the master broker I see the slave acquiring the lock and the proxy connector and message producer fail over to the slave but the consumer connected to the proxy stops receiving the oracle jdbc tested using SNAPSHOT from Octafter a bit of investigation I discovered that problem is not in the reconnection logic at the proxy broker but in the unmarshalling of messages at the consumer after the turns out that turning off the wireformat caching on consumer side by using an url liketcplocalhostfalseresolves the you test it out on your side and see if customer is satisfied with this workaroundCheersDejan +Verified in JBDS +Any more clarity here on what the right way to fix this is +Thanks for that I specified a custom but with grails plugins everything gets compiled regardless of whether you are using them or not +I also feel this should be part of ping in after. +Does it work for you if you just upgrade the Spring jars to but keep at? just upgrade to but keep Spring atI suppose the issue is caused by but please verify this by testing against selective upgrades there! think youre right caused by +Just try example above it says it all +I dont see anything wrong in the code you sent and this is functionality that is tested and working here so I dont know if you can attach a test case that reproduces the issue we can take a look +Well Ive no problem with the idea of replacing my original fix with something that is better but I DO have a problem with leaving the trunk in a state thats broken +HelloSince this issue is kind of old is it still planned to be implementedManitra +I am thinking of closing this issue +Any ideas +Integrated to TRUNKThanks for the patch JieshanThanks for the review Ramkrishna. +I forgot to commit my changes on +Applied patch for txn part in SVN revision is not completely merged into trunk +No issues encountered so far +A patch is attached +Implemented and now in trunk +the current suggestion is to set up a rsync server on vmbuild pointed at the snapshot deployment repositories and let each project do as they please wrt rsyncing to update on what we need to do here +I would lean on the binary side for safety reasons unwanted modifications are less likely to occurNow that the arrays are cloned I dont see how modifications can occur at run timeSource modifications are of course easier +As long as the thread is alive it references web application classloader +There is no real control on CM updates in place and I think there are some other issues as well +I will assign this to Karaf for now +I have a subinterface of that +Not a issue and needs language design +Regarding Matts comments I moved the code back to the script block and preformed the logic to test if the page status is PENDING thereThis fix makes it so shared pages which are not yet accepted by the user does not get rendered in the background. +I have created a patch for this very minor issue +If Steven would have put that into the issue description I would not have complainedYou know Uwe I assumed that you would be asleep by now rather than ing JIRA in the middle of the nightBut seriously I agree I should have provided some rationale for this change +Thanks Keheliya for the patchIve just applied it. +Targeted to and bumped up the priority +Indeed all looks good now thanks! +Ive just committed this to trunk +As long as its marked executable when its checked in thats fineWe also need to make sure and update the doc on using thisI agree with Travis that this isnt an ideal solution but I cant think of a better one at the momentSo all that said Im on this patch +Good news is that most of these patches are small +Will let this bake in for a Uwe Erick +It looks like Path may be something to look at too but Path is a scarier change so I am happy to check this one in +Please try this oneNow I can select and index subset of all metadata +Closing as a dupe. +Kelvin added a generation parameter to the write methods which fit nicely into the builder pattern + fixed fixed r +there is somebody watching this problem +Committed +If I can help in any way just ping me +I added for Source and Sink as they shall go in separate packages +The VM could be stuck in expunge state for some reason and if you dont show it then you will not notice its stuck +yeah +Patch for +Or at least have a FAQ entry pointing at the right thing to do with gcj if thats the problem +Here is the Nicolas +As soon as that decision is made you can finish this jiraThankssanjayIf this patch looks good I will provide a script that can covert patches for current trunk to the new trunk +in a first step i dropped the idea of using SAX events +I dont consider it something we NEED to do for the next release however +it still hide behind rich it will not cover the parent drop but still under the accordion if the width of accordion is not big want to make it upside of other suggest setting +Please reverse this change when the problem is fixed +In case client is upgraded but server is not my impression is that extra param passed on by client is automatically dropped by rpc before making a call on server side +We dont have a valid use case for fixing this. +Whats the status of this? Is there something more to do +Im open to other means of determining exactly what string was used to instantiate the +Looks like Aaron is still seeing this issue +It can remain on trunk for right now as we investigate how to handle it +Ive successfully gotten the includesexcludes working just fine but this doesnt enable all of the features in testng that I would personally like to use +Party creationhttpslocalhostpartymgrcontrol and in trunk at rev fixed in trunk at r at r. +Also possible is a change to the framework so as to have a param that allows us to reset confs between calls +OK thats a case where we get the node type name from the caller and where we need to support both remapped prefixes and expanded namesIll get to the missing optimization in the name mapper for this case this afternoon +If nobody objects Ill do so in a day or two +I had fixed the issue locally +patch with better documentation of the test case +But this approach is better than what we have now so go ahead with it until we come up with something better +The implementation is mainly completed at the moment +Why would a task from an already running job not be able to find? Why is the task tracker removing content from the cache of a running job? If the content movedis different shouldnt the job tracker be able to reschedule tasks onto a task tracker that has a copy? Why cant the task tracker copy the dcache from another task tracker that does have a copyThat said Im not convinced that in the majority of cases that vs +Please verify and resolve +Apologies for not providing the path in the earlier have provided a new updated patchThanksAbhi +The attached patch finally fixed it for me +Other API methods should work as expected +Im certainly open to being overridden on this I would love to have this but since the vote is already halfway in this may have to wait for +Fixed again unusual version number for special batik version. +On the other hand when value lengths are getting longer they vary a lot +I would suggest putting more attention to +foreach and ateach are gone in favor of for +for regards Malcolm Edgar +Attaching initial run of tests agains product jars This failed miserably with lock timeout and cascade errors due to cleanup etc +Plus AS javadoc isnt really created often +one more questionis it possible to generate the chinese PDF file in version which can be viewed by arcobatMany more informationcan I use your FOP to generate the and then directly forward to client browser for printing +And because theyre gigantic they cant easily be split into multiple should generally affect only C compilation since its in the root of the script +We can throw a lot of traffic to our mirrors and I dont think your bandwidth is adequate +instead of using the deprecated methods this looks a better fix +Is there a reason why you do not use inBecause its not an? Youre right the Streams and the current Streams are very close but the have different access patterns average sizes etc +Please the! +An exporter with a lower bundle ID is preferred over a bundle with ahigher IDIn my experiments so far these seem to be followed +Per discussion on Ill leave this issue closed and commit this fix under instead. +Could you please let me know going forward I have to use github issues tracking or this oneThank you +handles loading saving upgrading rollback etc of image files +Remove the service and then reinstall it remove install +If you need this patch please apply it manually to the branch in trinidad. + in to the different file please apply the patch to enhanceddrlvmtrunkbuildmaketoo +I cant reproduce the problem on my local copy +for the fix +Granularity Example Hibernate Tooling versus the actual umpteen hibernate Seam Tooling versus the umpteen seam etc +I believe this is fixed +If the list is dead the list is dead +I will attach my latest version of the User Guide +To be consistent should the convenience attribute for the help facet be helpText instead of tip +but its practical +My suggestion is to start with the hive support and see what subset we can support in CQLSo how does Hive currently handle our example here +All done +When we discussed this before we got sidetracked by looking for a more elegant fall back +Do you want me to provide a patch for the first or the lattera simple mvn install wont and does not do +bq +duplicate +Hi not a docs issueId put the sample files into the distribution archive +We can live with that for the time beingTo be clear my previous suggestion was not to use addContent at all +I dont think this warrants blocker priority though +Aside from the issue of PCRE vs +inc are data input files which although it may be creative to have come up with do not constitute creative works by themselves +Any thoughts on how to handle that +I apparently read your first comment not the edited one +I think this can now be marked as fixed +The raster is public domain thus it should be no problem if you use parts of it for the bug fix +Ive switched on Precedence Bulk and turned email processing back on which should result in a flood of notifications from the mail queue +Once the initial datarun is done then a nodetool compact fixes the initial corruption my second run is now about x further than its been before and no corruptionbad WHERE IN results have occurred +ill back out simons patch and try to suggest some modificationsto fix the build but also improve reproducibility etc +please feel free to review +My edits attached +If I run with a patch that is simply a reversion of this commit I dont see the crash +The caveat here is with redistribution which we have turned of by default the reason we dont redistribute is because it breaks message ordering semantics as they would if a live server took over the responsibilities of a crashed live server +I see resources are relivered by in any case +If we allow that extension to be used at all in the C code then we risk breaking compatibility +Hi JohnThank you for reviewing the patch +suggested by Jarcec added a more descriptive commentThanks Jarcec +I think it would be good to resolve this before landing on trunkIf we cant score based on positions it seems to me the api is not fully baked +Patch commited +Will you fix this in or should we move it to +Please comment and vote on +Bulk closing stale resolved issues +I had forgotten to commit. +The implementation already instantiates a separate object per Test so Im not sure that is really relevant here +If you have a moment would it be possible to capture another diff against trunk of exactly what you want +So far Ive fixed two tests +Looks like this needs a minor update to get it to work on Mac OSX +Still need a fix for XSLTC. +This comes up frequently in application work +Let me know if you want me to put the hadoop jars in TRUNK Ryan +Sorry about that I only tested with ant and it seems that there has been a change which prints this warning if you reference a property prior to its definition in the build fileThanks for applying the patch! sure think +In the at line explicitly checksfor an external datasource config + situation in CXF version prior to the servlet parameter can be enabled to avoid the unexpected for the quick response and solution +Updated patch with a unit test which verifies translation of reduce side parameters +This unnecessarily bloats the plan size +The value part in this case is actually one of the types so you actually have to deserialize that and then retrieve the value of the cache entry +To my knowledge theres nothing preventing the user from deploying an webapp into Mule +forgot to push my expandingcollapsing nodes are OK now however IMHO the nodes should be bigger since it takes me a lot of time to hit something correctly even on tablets +I need this badlly to continue what I started efficiently +to unit tests +too bad +The latest patch is included inside the latest patch I attached to this ticket +Hi Knut my only reason for the upgrade suite regression test request was to show that the behavior has indeed changed from older release to is the next potential release vehicle this patch could ship with? It would be great if we could pick up a new version of Derby with this problem solved in time for s April release +This would seem more robust than something specified only at file creation time and more managable in the long term if data nodes membership changes over time +I can verify that this issue is caused due to the +Check if remoting blocks until the response is issues resolved in released versions +Is it likely that we can get this completed for +To reproduce the bug perform the following steps in ij with Derby or laterij version ij connect jdbcderbydbcreatetrueij connect jdbcderbydbshutdowntrueERROR Database db shutdownij connect jdbcderbydbstartSlavetrueERROR XJ Failed to start database db with class loader be see the next exception for detailsERROR XRE Could not establish a connection to the peer of the replicated database db on address localhostij The startSlave command fails after one second +unzip the patch to luni module +So instead of always dropping the non matched connection it might be wiser to keep it unless the pool is full or something +This is trunk there are many changes since that do affect users embedding FOP in their apps and this is not one of them +The sample code attached worksSee previous comment for more details +a new patch against testing lucky patch +For one the patch doesnt seem to make much sense it sends an OPSTATUSTIMEOUT to the server even when the server may not have completed sending the data but doesnt actually close the socket on the client side +These must be adjustedThe JDK version is Oracle Update Bit +during start up and the recently added SSH assigmentI was thinking more on the normal assign flow +There are a couple problems I see right now with events as they stand +Sorry about that +Committed to trunk r +This may well lead to confusion on the users part and annoyance for Ops +Were a bit busy getting stuff into production will get back to you with an example when we got a bit more breathing room +Wondered if anyone had any comments on patch? I will work on committing this patch by Friday evening if there are no comments +Applied proposed patch +how to correctly clean things? I tried deleting andor removing everything inside but it had some side effects which I didnt realize in the first place and cost me some time +Please see the dump in the attached file +Closing for long inactivity +This includes all the fixes mentioned in this bug up to seems were dropping the ball on this a bit +Set fix the full blown audit to beta +I just committed this. +Closing again. +Anything after the will not be applied rightDo I need to create one default processor chain with my custom processor before the and the real one used by the update handler with my custom processor after the +If anyone has a problem reopen issue +Again repeating what Chris asked for earlier can we have a testcase to reproduce the issue +Closing this issue +Then restart Eclipse +This replaces the last patch also sets failonerror on the two java tasks surely this is the intention +But the error is happening during rollback execution +Actually not in libext its in the WAS lib directory So Im not sure I understand why wed need to add it again as a dependency of the Webapp +It is not occuring every time it appeared in the logs during long run +Regardless to adding the include as a defensive measure anyway. +All jar files in libext are included on the classpath +the no loop attribute is meant to stop a rule from causing it self to fire again +I will wait for Felix to comment since he is the expert here +Bulk update component to Hibernate. +approximately when will be resolved +Hi LarsCan you recreate the patch against the latest code? Some files in your patch are out of dateThanksit has been quite some time since I did this patchLet me have a look at home if I still have the sources available +Patched to resolve issue +Whoops got my JIRA numbers confused +A very simple patch to remove reference to checked in as revision +Patch looks good to me +Done. +Are you still seeing this issue? Im wondering whether it may have been related to which resulted in a similar revision count +Please test the backward compatible between the old server new client and new server old client +Ive just added a junit test and a fix for this + changes and loading +also adds the following features and improvements over by default but allows destructive edit with suffix option +We should advertise the fact that MessagingTasks are retired and maybe spit out a warning if we see one +As part of adding the and options it will be a good idea to get rid of host and port usage in the logj appender +However the result is really strange in this case +Unless I hear objections in the next couple of days Im going to go ahead and commit this on the trunk branch +Thanks Sishen for fixing something I broke +or am I confusing twocompletely unrelated things +Could you confirm +Is this the additional problem you sawThe locking looks correct to me +I much prefer it to juggling multiple chainsI still worry about the clone doc issue and update procs between distrib and run if we do decide to not let procs live there we should probably hard fail on itLatest patch looks good to me lets commit and iterate on trunk +yep thats what I am going to do in the beginning of next week +Patch integrated to trunk +No I wouldnt want to have that situation as well. +to +Anyway I found out that if changed the signatures of my converters to the ones in it worked again +C can at least return a void for such data structures Let us explore how we can elegantly support this in C +Fix applied to CVS +Then a second query to each shard is sent which requestes the solr documents by its +I confirm that I grant the rights to use these materials to ApacheOne minor noteI am not sure that This means that an operand in the code may be reached at a point where it is yet undefined +Attached testcase +Hi KnutYou raise a good point here +Slavas fix actually works very well for WTP projects +I shall update the plugin and test the same +Are you able to attach a reproducible example? That would help us a lot +Is it safe to simply delete them +Lowered to Minor as it happens infrequently. +Fixed at revision . +Thats been happening a lot recently +This is what token filters should do +I have not seen epoll taking time at least on that testbed +I believe this is fixed in through my work with +As per the last comment we need to test the new IBM JDK package on bit architecture to confirm whether the suggested workaround is valid +classpath as source paths for eclipse +A clean design is fairly involved +Issue is solved thanks +Today I would argue to move them out and let them evolve independently of Buildr +it looks like tag library loading from lib is broken by patch +So configuring the is not behaving as expected +issue closed +fix patch to looks good but the tests do neither compile nore run on versions of +Its a refactoring and passes unit tests +Having a single page to enable user input will destroy the UI symmetry and design since we dont support the widgets as of now where the user can provide input directly on a single page +Releasing your plugin has zero again for the confusion +The requested attributes are too specific and of limited use and do not justify putting them in the main schema +Hi Olivier ErwanWhat is the status on this subject? +so what is the issue againIm sorry but you havent shown anything that points to any hibernate issue here +have you set truefalse +Eclipse reports some additional things wrt overriding synchronized methods without proper synchronization put putAll removeWe could also just simply state in the javadoc that the class is not +oopsstrange things happenGeorge how do you think is it worth reviving the old implementation to compare both? or is it just obvious for you that the new one is more promisingBTW it is hard to me to find where value profile is used could anyone give a hint +Im unclear on what max actually means +Due to the asynchronous nature of the job task and task attempt state machines it is possible for tasks and task attempts to complete even though the job overall has decided to fail for other reasons +Keep them coming. +I have left this method commented out for the moment since it is failing +If the image already exists in the pool the reference for theimage is used instead of the newly created string +So starting new instances can be done via the currently running AM instead of starting new AMs +Committed to branch thanks for reviewing Eli +Reopening +Ive been investigating and although I havent cracked it yet I did try upgrading httpclient +It does not indicate the target already exists +Good luck! Leo +Looks like we are overall on the same page + for the patch +snap should be out now please verify this fixes the problem plz +Any idea if this issue has been catered to in axisThanks for your help +Simple yet working implementation is attached +New patch which includes detection +same problem happens with m using JDK it going to be fixed? is unusable because of this bug. +patch looks patch in svn rthanks +or Mockito +Moreover the introduction of Remoting should eliminate any further need for the multiplex transport +This is a little bit different than the javax +Thanks +I think that the best way is to file an issue with Sonatype for MVNCENTRAL ask them to remove the existing netcdf jar and then republish the minimal version +Im not noticing the on the HTC One X +One approach is get javacc release source code apply specific patch to it +Would it be worth renaming the classes in the engine package so that they are different to the +Can you attach your SOAP message +This is the reason why Owen and I continue to insist that task limits are a poor substitute for modelling resource usage and that a resource model is a for this featureThis feature works well for clusters with single users but not in shared clustersbq +To reproduce drop the attached file in AStrunk default of Date +Since the exchange method provides the functionality this discussion is slightly academic but as a user of restTemplate it seems counter intuitive and to me it feels wrong to evangelize to ignore response codes for put or delete +Thanks for the notification Mic! Im moving the assemblies in the release profile to let Jenkins work then Ill fix the build +a typical windauze file leaking which need a windauze restart +Transitioning to closed. +Attaching new patch with testcase addedPlease review and provide your comments +HelloI was really interested in that option till I understood elegant work around is making the singleton autowire a which has a prototype saying that this shouldnt be implemented but maybe priority is not as important + for trying out Yahoos analytics +Paul whats the scenario? Wouldnt it be easier to rename the dir at the end of the build and integration test process +So were closing it +That fixed the issue +But still there are a lot of things have to be done like write a single filter and write a single class to implement to get the attributes +What a fabulous file system Snapshots clones are very addictive. +This is a correct behaviour for wssj +IMO that is the best way to do it +Yet before I get too excieted about yeah something with a wider scope than would work in this case +If you find further issues please create a new jira and link it to this one. +The x y coordinates are now sent but the source attribute is not which means the commandButtons action is not calleda commandButton whose action is calledpageContentsourcepageContentAidJspa commandButton with image whose action is not calledpageContentpageContentpageContentIm afraid I could figure out how to add the sourcepageContentAidJsp back in +That patch deals with insert only +I just want to state that I am still experiencing this issue +what is its usage +Applied on trunk +Bumping issues to except those that involve with Struts or . + +how about adding something like a userProperty or userString attribute to column override +With appends in place we want exactly the behavior that when a client dies we dont delete a pending file so that it or another client can resume where it left off +Zhijie what is your setting for? That is the configs that should reject allocations that arent satisfiableIn either case there should be infinite assignments can you investigate? Thanks +Because I didnt want to cast too I went with the other solution firstIts not totally obvious but the passed to add is of type and as far as I can see all of Springs implementations also implement +some screeshot? versionIm using FF and and cant see such behavior +Additional patch applied. +When I use the I can change the column type to INTEGER as you specified in your comment +Thanks Rick I was wondering about that bullet list +fair enough but could it deploy at least note that its also installing it locally +Id love for the syntax to supporthod allocate nodecount clusterdir mydirSupporting wordsphrases instead of single charaters would really help out new folks to hod as well as making scripts utilizing hod much more supportable +VladimirCould you please file another bug for the above Java exception crash? Thank you very much +I find it unbelievable that an object mapping framework is not compatible with itself between versions +An alternative would be give a Future back which would be immediately satisfied in the case of a synchronous call and would eventually become satisfied in the event of an async call +Change column header can be done by click Alter View on List Tickets view page then change the label and save the changes +Verified by Tony. +I just need to take care of the mode for displaying the tooltip for a checkbox +Bochun have you done any evaluation as to in what situations this additional knob helps +This is a patch for adding support for the MFMT command + Simplified the code +With the path cache whys this true? It seems that so long as youre accessing a relatively bounded number of blocks the paths will all be cached and you wont need to unless a block moves etcHow do you assure relatively bounded number of blocks? Not sure what the usecase here is +I would really appreciate if this feature comes in the next release +Indeed we are also creating a Directory with create false and then calling with create true to recreate our index periodically during maintenanceNormally we use ANT to run our tool which deletes the directory and creates it but we recently just started running our tool directly which would cause this issue +Hence it is not strictly necessary to make forceFlush thread safe +I regret every agreeing to do it and I bet everyone else does s probably fairly easy to tweak it to add a config option to turn on the fix +Remove jar file from project class path +HiI added the output messages displayed on the terminal where SERVCIEMIXhas been startedRegardsYV +Thanks Arron for reporting this and for the detailed and concise test case +Ill produce a patch for +Any objections to changing the maps from to String? String itself is immutable and a subset of the other class +The skin upgrade is now done in Cocoon +Failure to get something out of JNDI is possible when the type provided in config does not exactly match the expected class a JNDI name exception will not be thrown but the result will be patch addresses this by adding it to the list of failures that throw an exception of type +Adding productsamplesquickstartswebservicewsproxyversioningAdding productsamplesquickstartswebservicewsproxyversioning +Sorry I meant the Release Notes the info is in thereCan you please let us know where the Release Notes is locatedThanksPermaine +This patch probably fixes the problem I am unable to test it +contains a new error message hopefullymore helpful for this circumstance and also adds an additionaltest case to show that ORDER BY expressions are still validon simple SELECT statements just not on SELECT +passes against both hadoop and hadoop Will integrate if there is no objection +Hadoop is not an ASF release you need to take it up with Cloudera via their support channels both paid and mailing list +Mass transition all resolved issue that did not see any further comment in the last month to closed status +as the reviewer +Thanks Anoop but Im not a boss and were having a discussion here +CTF test suite completes without any problems +Committed to trunk +here is one stack trace +Regarding the patches heres what he said I confirmed with my sys admin that both the machines are identical and are on the same OS patch +Ill loop it and see whats up +As this class is in package I do think it should have javadoc stating its purposeWorking on that as well +Ok I think I misinterpreted the spec here +to trunk and to to change to in one place in the failed test + +With M Xmx eclipse takes upthe M from start +We may need to make max a container property used by as well as our prefetch strategy and take it into account in the later +Patch file attached +bq +Assigning this bug to myself +Owen added a for +fixed merge conflict with fresh changes on Robertplease use patch for both and trunk + +already had a patch attached to it +Patch looks good +for trunkCommitted revision +Ive included a patch to sort out this problem +Fixed +board writable is a good ideaI verified that the Git mirror is up to date +Altered remaining matches to use lm consistent styling as asked for by ThorstenHope its right. +All looks nice and cozy +If you use xmlns to put your elements into adefault namespace you MUST reference that namespace in yourstylesheets +Change the name from to and renamed all the related variable names correspondingly +This attachment includes all source and configuration files needed to build a war file to demonstrate the problem defined by +metadata +Transitioning to waiting for ping +I think that was for the backwards compatible layer correct? This seems to be a little different as I have not found a public document describing the Storage handler +No real justification but there is a separate change request for support of image folders +proposed can you take a look at Lukas patch and see what you think +The generated classes are not part of the published interface +a Rather than use LDLIBRARYPATH would it be better to set a runtime link path that used ORIGINb what happens if root is part of the super group +But I am sure my patch does not add more bugs its too simple Hope it will be applied +I am fairly confident of the second patchThanks Nigel +Works like a charm with of a bulk update all resolutions changed to done please review history to original resolution type +patch for new query operator that allows to do whats described in my previous message +My regression test run was clean so I decided to commit this fix so we cansee if it improves the behavior of the test or notKnut thanks for the suggestion about nanoTime +Or a term may be assoicated with few documents whose are successive thus they are compressed much better than a term with one doc whose ID is Mbytes is also something you can measure +It is always easier to start with something that works and then improve on itMeanwhile I will be trying to reproduce the problem with a test case that does not involve SSL and external resourcesHeres the test case reproducing the problem using the s testing framework +Hi Can anyone tell whether this is an existing issue and is slated to be fixed in a later version? Do let us know if there is any work around for this +I will commit this shortly +for real this time +ill fix the patch with review comments for a lter release +That files arent included in java build path and i cant use publish features from WTP + resolved issues for already released versions +As explained above this resulted in the XML file being and the activities not showing up hence there was no way to recover +As Ive seen this is fixed in trunk. +This issue should have CRITICAL priority +the patch as QA did not ran any test files due to cmake error +classpath for rlease Committed revision +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Additional patch that allows overriding of the guess with the CESU system property +Some features were implemented already in trunk +This is no longer a problem in never linux and rcp versions +Dont call the jsp directly otherwise struts filter is not triggered +And setting the child node to ntbase makes the restriction LEAST restrictive +for this patch +Would throwing a wrapped make more sense? Will let the application code decide how to handle +TedCan you help me post it to RB? I have problems posting the samethanksSubbu +Sumit Thanks for help. +I thought there would be very few datanodes hitting such a race condition +Patch +Patch for Solr attached +changed the db info gbean isXA flag to isSpecific and if it is show the properties +In revision I committed a temporary for this problem +Yea all same field shows this behavoiur +This also made us not entirely intentionally it must be said go from using to unique snapshots of Geronimo +Megan Glad to be of service +after changing our test running environment to deal with the new property all these errors have now gone away in the weme runs +Ill post an update soon +Thanks Jonathan It looks great and i think it will be more useable this way +Funny thing Jukka I will talk to Cedric Beust about it because it happens only if it is part of testNG test I just run it in main method instead of Test and it works. +CoolIll check it today +Added a cache for connections and recycling of the connections throughout the code +This issue can not be reproduced on +New attached zip contains patch and html output that adds synonyms to the list of items that cannot be created for temporary tablesPlease let me know if it looks good now +set the isolation level on entire connection object i need read uncommited isolation level on query level so will it be possible to add such support in dielect itselfThanks in advance +Also we can remove take if there are no plans to use it somewhere else +It works for me too +implemented +Thanks you Jing +After some bug fixes in broken portal build by Anil it deploys without problems. +A fix in as attached +Hi LeoVerified Thanks! +Attached repro +Very unfortunate +Hi JochenI just checked out CVS head compiled and run the test several times in both machines and I was not able to reproduce the problem also +well I am resolved to not use any singly escaped apostrophe since of course I cant just stop our project waiting forbut this might come to bite me at any time and who knows what other bug fixes have been waiting in the code for a yearso youre saying that people should just do a local build and just depend on a SNAPSHOT build +The patches brings an overview of opportunities to the account screenRegards we going to do something with this +Right now there may be a few extended requests specific to Eve that we may want to enable like Sync Shutdown Enable DB Diagnostics Enable New DB Partion Disable New DB Partionet +I will gladly create a new bug report if needed +The patch looks good to me +We should see if this can make it into +First Draft just a quick hack +Turns out everything works just fine when I use the constructor +Could you please download a patched jar and confirm the fixNow it seems to work +Current stable version of xdoclet doesnt support true and I use supported true but it is not enough +When will be Jackrabbit released? Maybe we can switch to that version +John Im on committing this as well so please go ahead +To me that makes them easy to understand for the user maximises separation of concerns helps confirm completeness sets project priorities and makes a good set of HINT easy to reason for not starting out that way I thought dropping in working code to a working Roo project was the simple option +So once I commit the changes after a test run I will resolve this issue +Verified on +Is the hand hiding a genuine test failure? May not be a problem but I see in the logs fails with exit code in the test run +Ive recheck the test on release and found handle leak as was reported +Attaching patch to address code comments +Adds volatileMakes the builder function return a codec rather than setting it +OK that sounds like a good plan Ill remove from my area before committing! I plan to commit in a day or two +Verified in. +Postponing because of dependency on unresolved Metro to simplify release plan this is now linked to the umbrella task +The original parameterizing change worked by copying files from the src tree into corresponding directories in the tempsource tree replacing the TOKEN variables during the copy +can you please close the issue this is not a problem there might be a virus scanner or your explorer still holding handles to these files +Congratulations on a really neat and useful product! Much easier to use and more powerful than JAXB etc +since we collided in midair +Im not sure about the reason for this yet. +However it appears to be fixed in +Still present in IR use helloworld quickstart duplicate action and both invocation via gateway and ESB name duplicates are now added along side service categoryname in CR +Thanks Raymond +Please test I guess it should close the issue +It simply deploys +JIRA assigned it to Alan Ive reassigned it to me but am eager to come up with a plan to share the loadIt seems as though python is looking like the winner +AlexeyCan you put the suggested fix into a patch fileIf the patch for latest branch doesnt apply cleanly to TRUNK separate patches should be attached +Hi AndunI have invesigated this bug and found that the problem occurs on the on eventBut I not found easy solution +Can you try that +Closed upon release of. +This makes no sense at all +so I think it must be related to the system font size wIll look into that +I assume you mean that Nutch implementation of it is old it is indeed! This issue is about completely rewriting Nutch plugin to the most recent version of the +SVN didnt like me moving files around so I had to do it the idea is that is where we can create test cases for things we know are broken and plan to fix +I feel it is worth investigating and getting all the answers to the unknowns +it actually happened fully Alex Tkachman then requested its reintroduction for groovy and provided a patch doing so +Committed thanks! +it s going to be faster as all future releases go automatically to problems appears to be corrected now as I am seeing daily rsyncs in the server logs + +With dynamic I mean that there is no set target for this +I looked at the example thoroughly +Passing along to Zach +Closing all issues marked as resolved as of I logged a list of them in release notes. +Patch to improve the packaging of cmislib +Looks integrated. +I can upload these also if you wish. +Assigned to rel +btw the approach works ve had relative success with setting up another lifecycle +This I believe might be fixed by ensuring the deployed bundles are started after all the configuration bundles are started via configuring the right start level +Anyway testcase works now +Thanks +I was afraid that might be the case +I created a schema for Sybase and I run some job on sybase now +This approach used here seems the cleanest by far +could you please try this patch +Thank Dmitriy for contributing. +I tested the following analyzer +btwi use resin as container + for reporting and providing test case +modifying outline to consolidate info supplementing each Platform Guide +Changed name from Lucene Standard and general cleanup based on feedback +Ive looked into this bug and couldnt reproduce it on Windows IA on Harmony rVasiliy could you please verify +Additional fixes for dynamically loading provider implementation based upon external configuration. +Verified the change from staged for EWP +Assigning to you Fred because this issue is related the article you have written +Anyway Ill try to reproduce this one +so I did not modify the local variable hiding the instance variableAnyway I think you are right +Patch proposal eattached +Heres the source code for the test case +I multiple time not a single block +go for it +But this means child RS can be removedreplaced before accessed by graph walkerIve ran about twenty tests before submitting +I separate into three function unit +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The test and domain class as a file +So the driver does not support the standard call +I wasnt aware of that +Also checked in the file and added it to the generated gem. +Thanks Ramkumar for the feedback +This issue should be closed. +Committing +There is a sample web app in shiros source distribution that shows this working +Setting Fix Version to future for issues without a set fix version +Looks good +to trunk as rWill backport ot are you going to do the backport to +I should have mentioned that this occurs when using a +Parth You could verify it once more and if you still see the problem please do the ticketThanks +properties files +ah if the responseTimes is only internal structure for a read operation +Looks like this was not taken up +Is there any other way we could support this +I mailed a comment back but it is not showing up fast enoughWe have internally been discussing the addition of a Coprocessor framework API for reading and writing streams fromto the region data directory in HDFS +trunk +Will upload a new patch with additional tests for timestamp and Boolean types +Yes i can verify thateditabout Tika stripping away lang attributes i can understand that detection from attribs wont work anymore but what about identification? That should not rely on attribs +Feel free to. +If you are actively working on this feature and planning to merge it into master by the feature freee date of for please change the status to In Progress +Just committed a fix for this +Luke we are running against time +I am going to commit the patch soon +Actually I thought we had a patch for this already +Its not quite working +Ron Can you attach your data generator code in case someone later has a similar issue and want to check their work or if an issue w this patch and they need to see how it was that the data generator found no issue w your patch above? If you cant the above testimony works for me and Ill apply your patch +So the priority of getting handling into Teiid is fairly low. +This will break right? Better to deprecate and add a new method +That means it cant be the reason why the mailbox tests are crashing +I have added integration +But if this is a injected type then the same problems will surface +Rebased patch to current trunk +where do you mention how this helps withbq +but I consider that as an improvement request since its a change to the set of features of ivy +Pull it back in if you think different. +Resolved as part of since didnt get +This feels like a bug +Work around committed in r. +There are actually a lot of useful things that could be done with an XSDFor starters we could have validated the the XML i wanted to introduce as part of wouldnt work automatically +and also the files options and files to reproduce the generated javadoc in command line ie javadoc options catched it +That fixed problem although I have to add two tag libraries jstl and standard jars +Thanks Sreekanth and Jothi. +I have been running with this patch since I created it and I can assure you that everything works perfectly +Ok for me of course +Fixed in r +Changed the error message to be a warning message instead +tnx for attached +This has the advantage that if there is a failure on one fragment you can still verify the others independently you can skip tests individually etc +Should this be assigned to Keith +Sam can you rewrite the test for trunk +This patch reverts the first version of this patch and then it makes it so that if you do tar tf we no longer have the +Thank you Sanjay +As far as I can tell needs to be updated along with ivyAlso is references in in the buildJar target when a single jar is createdIm happy to submit a patch for this if that sounds like a reasonable approach +You will not get text if you use +Validation is something completely different and is but an unmatched end tag is not a validation errorHowever the only reason for the IOTAbort trap seems to be a bug in your application +Ill try it to on newest development branch where we reworked the internals and dependencyManagement was added to Aether resolution which should fix your issue +Still got one testcase failure +Fixed in r +issues +includes additional debug information for investigating these problem. +I have attached the wsdl that i used +Updated role list to introduces construct +Thanks Dan +I dont think anymore its such a corner case to not return the reusable tokenFrom a users perspective the solution and your solution should not be distinguishable +Thanks Owen! +If the user wants to be sure which jar is deployed there is still the possibility to drop it in the deploy folder +I did another build clean and build tests and ran it again to the same effect +Such resources are usually not so they shouldnt be used by multiple threads in parallel +to formally defining some kind of with a contract of write out write back in same order +Great +Andy this is another change needed when switching to hornetq can you please update the script in the JBPAPP branch to include this changeThanksPermaine +There is a javascript error clearSelection being null +There are two kinds of JIRA Question and Task +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Hi Benjaminwe can not call it because the new class I added isnt a its a parser for a connectStringI dont know what the you suggest should look like +I really am not familiar with what you want for this jira Can you tell me what you mean by a vanilla ee project? Is a simple Dynamic Web Project with an and a servlet count? Or what +Willem I am working on this there is some changes needed on trunk the patch doesnt work with +Latency of creating a snapshot be small +Yes happened to me tooIt looks to me like Xalan decides to do a stdcopy of nothing and the version of stl shipped with Visual Studio raises an assertion? Anyway I believe the attached patch fixes the issue +Sorry for the inconvenience +ForrestRex please verify there are no additional JAWS failures and then close this JIRA +Its back +looking into it +But I didnt change anything! Does that mean Jon fixed it w his hbck commitJon Let me ask Mikhail why he went to retries +Approved for. +defer all X issues to +Committed to +This makes the disappear even though you can still see them in the +You could of course explicitly add one catch the connection change notification from the watcher and then cause the loop to fail on the next run +The patch which fixes passing float parameters to native methods +I plan to commit in a day or two. +from me +Have you seen the discussion at? Would be interested in your opinion there +moving this to +I agree that specifying allowed fields is recommendable +This patch includes flushing in a background thread +heres the patch with the unrelated bug fixed innow all tests pass +Thats actually the way I wanted to do it too +Is part of. +This seems to work fine +Ah good glad you at least found the code snippet at the bottom of the page in the admin UI +unlike has a nice link to pattern matching in SQL but were not using like mismatches seems awkward and doesnt have quite the right meaning in English +It dupes this +For now I assume all the necessary infrastructure is in place. +Im ready to commit my changes and start hudson testing +The class was designed to be used similarly to String but for bytes instead of chars +I think we should update the existing ibiblio resolver to have this behaviour instead of creating a new one +If this bears out with more testing then Ill find a way to incorporate this into our test suite. +The following tests failedindexcompressionindexbitmapcompression + +I think maybe a separate field should be usedWe discussed this before on the mailing list a while back and I mentioned my concerns + +If not using you need to add the dependency to you project as thats where the persistence adapter is located +Take the issue over to the user mailing list for assistanceYou probably just need more JVM memory +OK first item do you want this to be transparent the API does all the failure handling and restart or do you want it explicit +Does it make sense to add to plugins directoryAnyways if plugin is not added to it should display dialog explaining why +Adding custom schemaTo assist creating custom schema for the embedded server I created etcschemaCreate your schema in etcschemasrcmainschemathen run to get the maven plugin to generate the java source files in etcschematargetIt isnt that elegant but does appear to work +It round robins dispatching messages between consumers even the slow consumers +Second consumer should see the message but will not till another message is dispatched to the the customer describe the current bad behavior +I think this is not a bugWe can control the filehandles with the pools and the GC suggestion makes thetomcat GC operation much faster +I havent used and Id prefer to keep things you wish to internally use whatever it is please provide a convenience method to obtain the current or default all I need to know is what attributes of a bean are required and possibly what the types are + +Done +Invalid this is working correctly in this record since is now released. +I believe you dont even need to create the HSA all you want is to check if something is there or not right +Many thanks patch applied +It would be very nice if developers attached here compiled and files identical to those shipped with but with corrected we would be able to safely patch existing instances to make them usable for remote calls. +Ported from trunk to branch at subversion revision . +Also were talking about OFBIZ here right not some custom modificatinos youve made +I think we need to get the old behaviour working in Form as we need to support existing applicationsregardsIf we the setColumn method we need to document that this method serves as a hint to parent Containers and could be ignoredFor example Form would act upon this value but would ignore it +It will be nice if we can help user somehow to add JSF capabilities in this wizard if the project doesnt have it +Closed with released. +Thank you CheyenneI proposed the logo to the Neuchatel guys and they are all happy but with a minor changeWhen we discussed this with James we talked about keeping the hat on unfortunately that detail didnt make it to this JiraCan you please update the logo so that its Neuchtel instead of Neuchatel? And then well all be happyThanks againThomas +Alright so to make it vmware esxi compatible instead of using we can use ovftool +Since nobody commented or reopened the issue Im now closing it. +Attached log does NOT have error you referred to in the description of this issue +Please reopen if the problempersists and you can provide more details. +Im going to try and come up with a workaround today +We definitely need to address them as we go along +The root issue is that the sorting for grouped queries comes from the which is not modified by the +and if you could give a small line about how we can verify that it actually worked that would be great too +added code tags to code sample to make it easier to look is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +When you resolve an issue please remove the associated in the test resolved issues for already released versions +Yes isPrivate and owner properties are not using anywhere +named +But I have a warning message as above +Here is the encoRedirect Servlet API javadocEncodes the specified URL for use in the sendRedirect method or if encoding is not needed returns the URL unchanged +Added new graphics to +TriquiI appreciate your help researching and debugging have just committed some changes that should resolve the issue +Attached proposed fix actual delete when closed +They said we can use it but should not alter it +bulk close of all resolved issues. +is a node which did not move +Im going to leave this open until tomorrow in case others want to take a look and comment before I commit it +That additional state is not something to worry about from the users point of view anyway as its handled completely +Tree walker dispatcher and processor code to make it more amenable to expression creationetcWill update with a more detailed description of the changes after lunch +Ok adbacaeeccaffcddcf and updated jffi to +if this is a problem with the EAP this bug report belongs in the JBPAPP project against the EAP CP release not in here +On top of there being a lot of methods marked public that should be internal or protected internal I think this might be easier than we originally thought +this one if you feel strongly about this one which would be an enhancement rather than a bug. +submitted a patch for azure policy in explains at a high level why we need a separate policy in azure +Thanks +tnx for input +Yeah onlyThe timeout in is already set to sec +wheres it getting the eclipse stuff from anyhow +The bug still shows Open status and unresolved +To the derby developers that box indicates that a query succeeds but silently does the wrong thing +Hi John that is very bizarre +Except that the scm URL does not only have the URL but also the path +done +The patch needs some cleanup to +Meeting with Jared Morgan from docs team at end of Feb +Not the place for this +Just to be safe I have attached a release note +Profiling identified field look up as one of the larger hotspots in the code +Does this mean that there was only a single authentication process that had this issue +Wont be able to take care of this one before +Assigned to Brad. +Also upgraded to +I am able to reproduce this issue even on simple JSF Kick Start project without using meclipse functionality + +First swing at this probably not ready for submission yet +Not sure when this was fixed +Adding to for the doc issue +Maybe it would be possible to see the full tag when the mouse hovers over it +Im sure you havent been the only one running into thisJuergen +the soap action should have automatically set +Ive just committed this +Sorry it doesnt work +Let me know if you have questions +fix version +a st step I updated docbook to version at revision +Therefore its already impossible for in MR and RM to have the same method name +Its been hardto track down and Im dealing with a JVM bug at the same timeso I havent had time to write a test case yet +What are you trying to scrape? It may be better to work on a proper XML output rather than try and deal with HTML that is is free to change from release to release +Alan could you please review it +great +I actually made this a feature of the core Hibernate transactions so Ill just add some tests there +Perhaps the rigth thing would be to abstract the reporting mecanism from the actual test framework thats been used +Youre welcome! Archetypes will include this PR +I cant reproduce this error based on the information you have given +I have also seen this error occuring on some of our applications delivered to doesnt appear to be restricted to tables displaying large amounts of data +Another comment akin to what is to be done to get this included +I was on a trip down to NYCThe patch is definitely starting to look much better but there are a couple things +This feature would be very useful to us too +The original issue is fixed I think but please reopen if Ive misunderstood or there is a related issue +Actually closed this myself +I havent no idea why it happend +Hi All Whether to provide admin command to start the bookie in mode? Id like to know the opinion from others also +If you reopen please reassign appropriately +One other interesting tidbit +Working on it I have a patch +The Link Pull Request option has moved under the Workflow drop down +Please continue the discussion there rather than duplicating the issue +So we havent broken our contract that exactly one client will see a successful CASBut what if the value committed to A has a higher timestamp than the one committed to B and C? Then when the partition heals repair will treat As as the correct value and the write to B and C will be lost +site was done +But it needs a minor update because of has a minor compilation issue has movedAlso I am suspecting that can be simplified a lot with mock objects +is in the JBPAPP done for CR +Ive done further investigation and found the root cause of this issue +Unless along the way it becomes easy to fix that while Im in the vicinity +NPE in war bugfixmore optimizations +Its not because the program dont crash that there is no bug in anysituation an infinite loop is a bug and should not happendThe infinite loop detector is a good improvement but dont fixe the bug and giveno information about the bug sourceYou have a better view than me to see if this bug have to be fixe +Im not aware of anyone working on such a thing so go ahead and create your own +Its off by default but specifying true gives you autodetection of proxy interfaces provided that proxy has not been set to true as the support for prototypes in Im not sure whether this use case is typical enough to justify adding full prototype support to +Review OK +I just committed this +bq +We could copy the serialization code from Xerces or Xalan to get existing and tested code for this while avoiding the full dependency +There are many unused imports +However the class does not get indexed automatically I assume because of this depth issue you describe +So it might make sense to wait and see before we change +Hmm the dummyHeader is actually written out into output in prepareEncoding +Reviewed the changes look good +In the past we tried not to have hadoop specific options in pig +I added a comment about this in +Please start such question on the users mailing list first +So what i understand from it is that you would like to store an email template in a text file +Removed reference to actionSet +This is looking great +Can someone confirm if the solution is correct +Add this ability for system functions but theres no udf metadata to use varargs from the designer. +pending javadocsfindbug fix +The available releases should correspond with the download pageAt present the download page includes UIMA Ruta though it is not clear why that is still present it seems to be the only old version that is specifically mentioned on the page +Thanks Ashutosh +As Harsh pointed out there are no tests for the commands +There is a test case test that verifies the contract +If you can live with milestone releases go with it +You can even remove the files it claims to be working on and it continues to sit +Things seem to look good for me but I was having difficulty reproducing anywayAndrus can you see if the changes make a difference for you please +This looks like a duplicate of which we closed as WONTFIX + +I tried your test cases with trunk and they work as expected +Here is a patch that should fix it +My fixes currently available in rc release +On my way home yesterday I remembered that it wont work on it does work on about patchNow its exactly the same logic as for +Im burned out on it and nobody else is stepping up to help reviewWe can fix with the current data format for +Which JIT is itAlso making the methods final may help the JIT figure out that the object allocation may be eliminated +This sounds like a different splitting algorithm altogether not a modification of the existing one +i assume you do actually have nodeswhat is the full exception from the cli when you run it with debug +internally reviewed by dhruba +Added Aperture detection to closed. +Fix tested thanks to you Martin! +One change to make is that were introducing just one version change per major release so sticking with VERSION here is fine +Hudson reported success closing. +The result of this effort is a quickstart example to demonstrate the integration +If you look at the above line code to assert the response from the faultEndpoint youll find the requester actually can get the response result from faultTo endpoint +A change solution might be to set this var which is in to a smaller time interval +In a couple of instances keeping the size of this jar down is more important +Those messages are correctly captured in the files from the tests and I dont know where these come from +I dont recall our solution at the moment +Might fit here but thinking about just doing in a follow up +bq +Hope this could help +This patch uses is as suggested by Jimmy +Here we asked the user to change the apache server port to +If the object was not in the cache it was downloaded upon the first request +In fact since we are going to read the first elements from an array better to check if there are at least elementsNot sure how we can improve generic errors handling +button +thanks +backport to revision +I will add tooltip option if thats OK +Bulk move issues to and +Latest version of wssj is +has been marked as a duplicate of this bug +Hi Carl Was this committed to trunk? Im not seeing it there +It blocks the WS TCK +The documentation explains ho to retrieve the different maven classpaths in an ant script but it does not explains how to add optional tasks to the classpath that the ant instance launched by the uses +is a good starting point for could implement ifdef ifndef define and undef as then it would be nice to include them in or Extensions? +Am I wrongPlease give me an answer +personal reasons +Dmitriy can you test it +It seems the new code in trunk partly fixed this issue +This is the case and that solution most likely can be platform agnostic +this prevents the managers of different transaction to deadlock +The Style Map Editor dialog appeared without any error +My suggestion is we ship the features which supports Karaf as a verified feature and leave the features of Karaf as an experiment release +ant version upgraded toCould you please test againhttpclient is a full blown uri util class +HiJust few thoughts to begin discussionIm thinking to the locks using ephemeral sequence znode to give fair chance to all the like below +Where does that limitation come from? We were looking at a server ID +Igor fixed this in by removing the libiconv dependency. +Thanks! + +Thanks Bruno your patch is in rev +Seems to have died on the vine. +Committed to trunk +This is related to the fix of +After you sort against a mix of fields the performance is worse forever +NehaThats correct +windows doesnt do well with url shortcutsSo instead i rely on a few tricks to make them launch +I decided to fix it myself see commit message for proof and python script that did it for me and that it does not break anything +Charles This is only in so it would only go on master. +We do not need the Nagios stuff at all +i am using jdk code to agree with the Javadoc param comment with revision This should take care of the Jenkins warning posted to +I ran the grant revoke tests and they ran fine +I think it is caused by running with an older version of an HDFS jar so I am going to upload the new patch and see if that fixes it +Misty assigning this to you + patch +I have run my test for a week with the SVN and NIO disabled and the error did not occur +this looks awesome +GIF +A formal nitpick Cayenne is all Java now +I get a successful build +It has already been fixed upstream so it will be fixed in EAP ve changed the fix version to EAP and indicated that it affects all since EAP CP FP +also has the same issue +it could be something timing related for example if you have fast disks and I have slow disks events might be firing in an order such that yours always fails and mine only sometimes does +hi for some mad reason the code has started to behave itselfi removed all cordova and phonegap folder i cud find backed up my www folderredownload latest ios cordova from github recreated my project pasted in my old www code and it started workingi think i might of had a faulty cordova install but its working now +Thank you for this option to switch between and default is +trunk +This can be reproduced with should be solved by the solution outlined in doc to proposed solution +Closing as such but if it comes up again feel free to reopen +Thanks for the patch Izaak. +No feedback from Joe +new table style with light blue header +Christian Ill let you know if I have questions +You will find a template project and instructions +Thanks for the reviews +if the TCK being discussed is not restricted by an NDA then I dont know why we need to limit any discussions about it +I find it an obvious enough bug that causes big problems when binding to port that we should put it in all branches +passed +Im thinking of using range as supercolumn and storing sessionid startcompletion times as subcolumns that will give us possibility to log individually each of the repairs made in the session +trunk Karl can you try with latest fix where I have improved the pre to send noop and reconnect in case of problems +could be get enumerate or do you think +especially i threw out the file +These sorts or error messages are normal when the connections are broken between client and server. +create an EJB jira feature request to track down the EJB version of this feature +. +In fact I am trying to figure out if the current state of threading component allows to actually implement the clean shutdown of the agent +I updated the Spring IDE to the issue was fixed +I dont think you are initializing the variable xmpWriter in that code snippet +flexible artifact filter in the core classloader falls under the heading of customizing a maven instance which is a topic were trying to designimplement for +Touches the following filesM srctoolsM srcworkingwithderbyM srctuningM srcadminguideM srcdevguideM srcrefM srcgetstart +At least for logj this is wrong since the conversion of Objects to Strings is controlled by the logj configuration +Perhaps it is a mozillafirefox bug thou I searched and could not find one +note sources are now hosted here sonatypeSo you can play with a pull patch applied cleanly to the git clone that I created and all tests still passed +Most of the driver is in. +Depress this tab you will see the billing account record with available account balance field +The applied patch does not increase the total number of javac compiler warnings +We should have some for the retrycaches name here +Ill check on the Tika Jira if one doesnt exist then I will open accordingly +It has been tested on several platforms and has proved to solve this issue +Itd be cool to have a more integrated and general solution to serve as many users as possibleDo you like to work out an improved patchThanks for contributing to Vysper +Changing the descriptiontitle. +The root cause for the problem between and your own custom site is only the same in that there is no match in the locationmapIf you are having problems with the locationmap in a customised site then please discuss that issue specifically as it is probably not a problem with Forrest but with your customisations +I can replicate this when printing a PDF generated by using Adobe reader version +Verified the fix on a real cluster and it fixes the problem as expected +Both Harmony and RI stops throwing the NPE after installUI call. +Attaching +Oh Im with you nowYup its an issue +The issue comes up from the use of Obsolete in the WSDL for message operation and other names +upgraded mono to +Functionally completedThe user is displayed with an option after they have chosen to install that allows them to choose whether or not to create dedicated hornetq backups ie run with Dbackuptrue or not +Yes MarcoThanks for the reminder o +And oddly enough I can not replicate this today +With jvms each generating distinct looking sql for the jpql our buffer caches were being overrun +Of course this isnt necessary for those app servers that come distributed with those libraries which is why it wasnt done in this case +Could you verify it fixes your problem? Ive sent you my email but hopefully its not needed anymore to send me your model +Eli this will occur on clusters when the AM resource ask is greater than a single slot and it requires multiple scheduling cycles before a large free slot is available to launch the AM +Adding my name to the Confluence Maven Plugin Matrix to indicate that the plugin is being developed +It just copies the preprocessed data into a folder that is different from the one containing the generated codeWith the arguments above taken into consideration I agree that its the best option +Really appreciate that + +debugging information was being printed even when debug false +Do you have an application which reproduces this +Translated DITA revision . +I attached the wireshark log +commited at in test and close +Is there a follow up JIRA to address Robert concerns on archives and corrupt files +You could of course merge these two tasks under one JIRA if you wish. +Now tests are parsing with the C services. +OK thanks Simon +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Also we can open another Jira for +Completed +Closing since fixes appeared in docs some time ago. + +Andre I assume you meant this were for M +Its better not having to muck around with the configuration + +Any ideasWhen I try it on the command prompt like thisZmvn pass Dargumentspass releaseclean releaseprepare releaseperformIt still prompts for the password. +awaiting clarification of one sentence +Heres a simple patch that follows the pattern in the other core admin methods +Please verify. +I think this null message couldappear in relation with onlyI will open a bug report when I succeed toreproduce it +So if you specify a jta transaction suspension should work out of the box of any application server now +Assigning minor since no one seems to care for this its been over a year +fixed in r +The problem isnt in the java code its in the componentorderwidgetordermgrscreen +Fix Richard +When combined with created a major problem with the stability of our systems +Testing with ITTBLL with CM right now +Hi I am unable to find out the documentation for customizable error handling +I used latest EDG to reproduce this IMO this is a problem of Infinispan itself though +yes I think we need the option of using just one thread +What do you think about thatOne thing I requgnized is that you commented testonetomany +All core and contrib tests passed +This issue was fixed in the tag +I would certainly never block a patch based on little issues like these but I do believe that addressing them would improve the patch +Closing RC issues. +for the patch +Summit a patch file for putting the into newly generated object +Well I did not think that farSo you propose to leave it the way it is am I right +As explained in above comments this issue is present into this is probably an issue with our refactored +It looks like something changed between running on AS Beta and CR +Yes I was able to reproduce the issue with break points +Do you have a distribution build +Fixed in the nightly build. +Attachment has been added with description Patch fixing the transition to closed remaining resolved bugs +That seems odd +It seems hasty to remove functionality in favor of a replacement if that replacement isnt going to be shipped with mainline Hive +Heres a documentation update for +Eugene Update Source Folders seems to have done the trick +Thanks for testing it +I have a set of war files but I want to create a single doc for them +At least thats how its specified in RFC +HariPl reach out to Kiran if setup info is required +Fixed for as well +Bulk close for Solr +But what of my repeatedly asked and unanswered questionWhat will become of the static use of Enums in Queries? Will we forever have to dynamically bind what should be a static part of the query or will a new solution emerge beforeWhat are your plans for the future? +g files ok but is having issues with and +Some users or roles may have more rights and so will see more nodes +Julien how about this one? Id like to have this one in before some other commits mess up my diffs as happenend with +good idea this way one can all the generated source with a single target +Here is different test cases path file Park your comment is so usefull to merpad and lpad has more issue then my thinkingI was fixed your issue and repatched +Added whiteboard implementation now +The more enterprise users use Felix the more often they will encounter this problem of not finding some classes +reverted the change +For instance what if you requested elapsed milliseconds since years ago? The method in this patch would execute a loop for every millisecond in a ten year spanI think Ive come up with a good strategy for calculating elapsed time that combines some looping with additional logic +Applied to and to trunk +Submitting the patch +No the fix I implemented is purely a workaround in +Attached patch b and committed it to trunk with revision I expect that most of the work for this issue has been done with patch b but some adjustments must be made when the drop database functionality has been addedTo ensure things get done remove from +addressed Once gets implemented this config should go in the server config dir +Thanks Viknes for the patch on searching with date +Hi ClausThank you for having applied the patch +You are right +I got a small problem with generated test class but rest is ok +Confirmed with SVN r and both Chromium and Google Chrome dev channel on Arch +I will look at permanentRoots in the the record I am deploying by copying a single EAR into the deploy dir +Hi Magesh great to see that youve started to look into this issue +r should fix it +I will run but dont think it actually exercises this code so a visual check would be good +Provided there are no unresolved dependencies amongst the set of bundles to be installed this should be more stable than have to devise an artificial order making everything more complicated +on this patch patch to include the mail recipients and the detail message of the in the log message +The reflect API has a bug that doesnt set the value of partition correctly +Updated api date in test fixtures +Jacques DimitriI found the error in patch and committed in trunk rev Thanks both of you! +Ill report back after Ive had a chance to do a test run in a federated deployment +Wouldnt this leave forceSync to true all the timenot all the time +Is this how you resolved this issue? Thanks! +You are mixing in features from two different types and are not really going to be able to recoverIf you do need to mix and match different sorts of Geometry please consider using +We need this section to be very strong and send the message that although the old approach is supported the new style configuration file is the better solution +support returning a I added +a target to generate licenses would be nice for convenience too +I will revert the patch to +On my test system I see about cpu time in that method which is removed with various versions of the patch +defer all issues to +Well as a general rule the Struts project doesnt add fixes for specific broken containers but if there is something more we can do please at least let us know +project file then File Import +In any case should stay tuned to their new algorithm +Thanks Aaron we have started seeing this in +So why not rename and remove the interface definition entirely +I will prepare a patch with the full fixKathey +I checked the error in more detail mostly to check that its not a regression from something in It isnt +They operate over our RPC mechanism +I meant agent ie in my example above + +It looks like SIS is not even off the ground yetIn my opinion an API is not really an API until or more projects use it anywaysWhy not make JTS pluggable as a separate module into Lucene? Then people can download JTS and add it into the Spatial solution by modifying a config file likeI would love to get this committed and done doneBill + +Only question why not have continue to use and? ie why break out the arguments to File and transaction? It seems like passing the classes allows us more room to add extra data about the stuff to be purged without having to change the interface +Alex looks good to me +Hey Sam +wouldnt catching exceptionsthrowables at least show us the other components on the page +I think this is done +This seems to be working for multiple maps after recent patches and +There were no tests on the previous codeI think lack of unit test is not a gating item for this JIRA +Very much so +Did I miss something? Also a comment on the auth method explaining the purpose of throwing this exception very briefly would be helpful + +tests included +This list should have the all rpms from the CR compose +RayeesNeed details about the failure +These do solve my immediate usage but I still think it would be cool to have on the master homepage some buttonslinks to do it with click and maybe a confirm +The same sql run correctly on versions before +Please check bug to see if this is a duplicate of that bug +i dont think it really affects the runtime since we are really just setting it to something reasonably arbitrary anyway when the leadershipfollower logic runs it will be written correctly +without my clientbase version of clean committed Camilles final patch to trunk and +commit master issues that are still marked as resolved but were part of the release. +In my case any exception is thrown even with Core so I get a Transaction aborted nested exception is Transaction marked for rollback. +If there is anything else I can do let me +The problem persists +Some of the stuff is not implemented Testabletrue Section following methods of the must be equivalent to the methods of the of similar name get set flushBuffer resetBuffer reset get isCommitted getgetWriter encodeURL and encodeUrl +verified that macosx installers includes probuct zips built for mac osx and fixed execution permission set during installation for all supported on mac and linux in +Any other issues I should also consider +fixed issues with patch applying +We still need to determine how this will be done going forward +To clarify the spec doesnt mandate the name of the WSRP service in the WSDL but its considered as a standard as its in the WSDL provided by the OASIS WSRP Committee +I just noticed that has not been updated so our latest screenshots were taken with the old overflowparam not with the fix +I will be working to integrate this when I am back in CA. +He recommended that I open a documentation enhancement bug to that effect in order to get it linked in the page +To the extent that there are still examples in the section I believe the syntax is valid and not overly obscure +They are being run with javaagent but before that aopc performs an instrumentation on the test classes +Thanks Dave +However I think it should work as it is +Fixed with your support to get it +i downloaf war filei put source file to webcontent root directory +Maybe Im just not understanding +I made all appropriate changes from this patch in r +We ran inso this issue more frequently and bumped up the max memory for the compile phase. +Problem is with remote jackrabbit +still revision +patch applied in r thanks Charitha! +Upgraded to is tickets +Closing as Wont Fix in favour of using regex. +This is also related to which exposes a missing requirement from the Web Tools server adapter wizard to our site +Any idea when the releases? The problem fixed here is a blocker for Apache release +One minor point is to comment in the test case for the Fake objects that we are supporting only speculative map reduce at this point +Committed. +Wow havent seen this code in years! The code was taking the button labels from the options parameter and everything is in arguments now +Oh Kristof just one thing if you want please post a followup on the old mailing list post +I would not want to have to use a special jar loader in my application just to use UIMA +All of the test failures are unrelated and known issues see +We could always have the default depend on whether the product is virtual or not +The rest of the changes look right from a +JPOX CVS does now implement implicit variables nightly builds dated or later and the test would pass except for the fact that it has some errors +Would be really nice if those could be used to properly setup these tables for authenticationOne callback onAs for the second issue I think we could change the way that keyspacetable are created and allow altering the systemauth keyspace +Ive been through all the comments here and cant get it working perfectly +Cool! It could be good to have something like this with myfaces core +According to David it has been decided to keep this at module level +Heres a first shot at making this change +the purpose of in its current state seems to only exist to satisfy backwards compatibility for code that uses Lucene at a pre era nothing more +That os certainly possible +Just an update on this bug +messages batching upYou didnt provide a concrete case when this happens +If we check for then its and if we add then its length is so th is the first target value for lookupFor TrunkThe indexing is wrong for paam as i test it for httpslocalhostordermgrcontrolrequest In then it get the partyId form so I corrected the indexingIMO it is a bug so we have to fix it and the patch will fix the problemPlease let me know it the fix is wrong and there is some improvement needed in patch +So I suppose it would become a critical defect for Java as well if we did that +Class and icon attributes verified in +I have come across the problem again and this time was able to track it down +I believe we still have some problems with so as Henry says itwould be helpful if you can attach your test case +Duplicating the dependency just as a typepom made the appassembler include the needed libraries +Thank you for your comment Jian +For testing stuffs +For us it turned out to be a DNS problem in this case primary DNS server down the repository server was resolving network addresses using its secondary DNS server +Assigning to +New patch that relies on properties instead of environment variables +Confirmed this with a little unit test will look into it + +Manually tested +GA the problem still exists +Automated running in nightly +Looking at the actual tar file I notice that it has the factory classes placed in solr directories rather than in the lucene directories as factories are normally organizedBy all means proceed with producing a normal patch that shows the final organization of this new analysis packageSome other issues +This should be fixed nowNOTE is deprecated and should not be used +fix as suggested by Vlado THANKSAnd set now default to for not having too small rendering on osxwindows +Committed to trunk. +somethings has changed in this area +Please open a new one for if this still applicableBTW I think this could be valuable but we will INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +for the patch +Implement delegaion according with specification +If he could gzip after the crash and send that along it might help determine the problem +New context menu showing Add of declarations +It can be done in a +The fix was to include the package in the set of packages installed into the chroot +unzip the patch to LUNI module +Fixed trivial race conditions +being fixed should at least change this from being a blocker to something less +Just updated the INSTALL file to describe this option and list the usual suspects that might be available depending on which OS X youre running. +I think the solution is to add extra checks in the recovery code to make sure that recovered data falls out of the retry logic state machineJon +I was able to remedy this by using the GUI and reopening and resaving all my workspaces. +bq Or are you suggesting to build Pig against and run againstNo that will not work at all +Even stranger if I remove the invallid manually it doesnt get regenerated and I cant see any place in the tree where it is generated +second part of Matteos comment once is implemented we can group by table +I also need to test the effect on large batch writes to make sure it does not have a negative impact +Linked the issue to the new created one +This problems is very important for us Any workarround +the latest complete patch? Running tests on that now +Ive must been tired yersterday since I could sware I tested this in both Ruby +Andras can you please attach a new patch that includes a SVN UP on your side +Thanks Thomas! +Committed to branch and trunk with above comments. +Im sure there are solutions to the problem if you change algebra ops the pq is a naive solutions that would work on top of positive outputs +wal version and compression type + +I am not sure about it as I still need to figure out the nomenclature of relevant cluster Ids +AdrianFor QAre you sure only these labels changed since? If so Ill download the trunk and tryFor QPerhaps you should change the files encoding to in eclipse +added some assertions to the unit stuffI guess the problem could still happen if folks were running multiple against the same database though in that case a real database would be used that properly implements locks etc +Hmm +My last comment was sent before seeing yours FYI +Possibly the overhead of three is more than using an object for every value and this implementation is much clearer and less than using three arrays with synchronized valuesAnd yes we can of course avoid throwing NPE where it shouldnt be one +LenIm not sure what old behavior you are talking aboutThe Runtime UI has been made according to Maxs mockup attached in +Needed to copy some and classes +This should be fixed by r +Uploaded a new patch which adds a comment about the fix and a test case that fails without the fix +In the near future wewill be moving to a newer version of maven +Yes indeed there was a typo fortunately this error only arises if the request failsI will fix this issue for the next release and will recheck for other typos +Agreed that this would be very useful +i think it is better to make the error reporting clear then to try to make an administrative problem a coding problem +Anton I fixed the dirty status when activatingdeactivating cell editor and added test for it +Reassign to a release that you feel comfortable resolving this issue in or leave as is +We do inherit everything else essentially all the specific bean can of course reconsider what to inherit there + +The patch is fine and tests are still passing +But dont take that for granted +daryn Id like to separate out and changes into a separate jiraThat sounds good since there is not code share as you state +How about we do it in a new issue +Committed to trunk and branches. +Fixing some white space issues in the new patch +I will prepare new patch with some test cases also +OK this is done patch applied and committed in trunk respin should include this fix +closed after releasing version +It didnt return the hotel list +for some reason today it does seems that the jars are inside the I misread the maven docs and provided the as the argument instead of the bundled +thanks for the patch though. +batch update of issues with fix instead of the correct +No action required Strong +This really enabled me to spend no more than minutes until the first execution of your test case + files go into axisjavaxdocs +After thinking about this I think youre supposed to use a more advanced container than jetty for this +Tests for this should probably go into since theres obviously a gap right nowI think we should merge this into but want to run it by the other folks first +This would force use to create some defaults for each of the standard analyzers but we would also have to allow users to easily provide their own classesAt this point I like the first option better but I really would like to hear other peoples opinions +So include schema table and indexname arguments but for first step only support null forindexname which would mean drop all stats for a tableNext step might be to support dropping just those stats on a particular index +Our tooling uses the shutdown command and this returns a value on success and otherwise +Dhruba yes that sounds right to meA further clarification should subdirectories be included with empty block lists or elided? My hunch is to eliminate them so that every returned is for a plain fileno directories +I think it is more or less stable now +on the surface to be similar to which was fixed in +The ability to order the repeating select is now implemented +We have the same problem with other configurations +Duplicate of + to get this in +Large streams are a likely cause +Local test passedsubmitting +OK I checked this in because its more of a code than a patch to a stable system and the top level ant still succeeds +To resolve this issue we need to find a way to a built configuration object in order to scan for loaded action configs and build some more +In particular the confusion of task trackers trying to use the same local directories +The close call indeed waits for seconds but if the close executes cleanly then there is no reason to wait for the whole seconds +Ill commit this soon unless there are objections +All that said the parts of the patch I was able to review look ok +Please be informed that I am no longer accessible through this emailaddress +after applying your patch MRI cant run testfile cleanly either +If its in recovery it wont serve search requests until recovery is finished so that could also contribute to the no servers hosting shard issueLet me know how it goes and if you can pin point any problems +The top level accumulate CE supports multiple functions +the best solution here is to add anexport goal with the same properties as does not have support for the export command at the moment +We could derive memory required per task by benchmarking the memory usage and then updating it whenever undergoes major changes +They may already work but need to have the references removed +and able to see that recurse not goes deeper second level +If not should probably go in a release or LATER +Looks fine +smile +It appears to still be asynchronously publishing +Apache Axis wscompile do not generate primitives also but still they allowed this one line of code to comform to JSR +You can find my proposition of a solution in attachement. +Yup the break should be unnecessary now that the infinities and s have been fixed +I still strongly feel that this behaviour should be enabled by default but it is also necessary to give us the out of turning it off +This patch is Aruns latest patch with SETMEMORYLIMIT increased to GB in K in the it to to run m worried about setting ulimit m too high Ill bump it down to M Ive tested it to ensure works on both Linux Solaris +This patch adds Configuration as an argument to +Hi Reymond patch looks good to me hence Raymond for your patch +Should go to. +Reopening to fix for to at revision +Thanksvery interesting +You can reproduce this defect with arrays input whose length is larger than and there are several elements in it +I imagine they are not lost as things are but they also wont come out in HTML anything like the author intended +Unless there are objections Ill commit this to trunk soon +JayWould it be possible to look at the source of the java app you guys have or open source it somehow? It would be great to see how you guys are pulling out the stats and might save lots of work in doing the web consolethanks +could it be allowed on the controller class also indicating that all methods inherit it? +Please offer this feature soon + +Fixed in trunk +Max can we commit it for CR +Hi The fix is fine +Lars keep it in mind doing +Not sure why this issue was assigned to me +Thats awesome will check it outI just tried adding the hudson address to the allowed senders on +Set version to for all issues since more true than no fix versionAdjust the fix version to match actual aerogear plan +Yep +This is fixed with the given patch +I am no longer working at Ingres Corp +Revised doc taking Solr config out see the analyserztokenizers page on the Wiki +Committed to trunk and +Now that work is well underway with AS all previous community releases are +Committed for in +The wsdl attached is the one from Sun Star Office and CXF fails to do anything with it +Closing +Concentrating on EAP issues and will then port over the changes +Thanks Kurosaka +The patch suppose include a web config paramby default false to enabledisable it +Closed upon release of. +I havent tested this yet but I suspect this is not an issue with in but rather a problem that the Faces Response has been made by the time the observer is called +we have changed the namespace in validators +Closing as nothing to QE or doc here. +needs at least the encoded region name added +Unassigning from Brian since this is being debated on the dev list now +I love refactoring projects +This has been fixed in r +This is caused by the fact that the default host is internally stored as null host +thanks doug +I dont see how Geronimo could be causing this error +It never was since it writes to the output file in a way +At the moment anything submited using will go through a URL decode process +move out of the project and rerun +This version adds elements for the new SQL State classes that have been added as part of the client internationalization work +the only issue with it is that you cant have a default value so custom deployments would break until they were corrected +Verified in r +The profile uses hadoop and hbase +Im seeing this as well +It works ok for now but isnt something that should be declared stable at the moment +However I also think that the ant tasks could be asked nicely to push to +Darran will create a jbpapp JIRA and link together with this one +thanks again Uwe +on trying the patch +last pieces committed in rev +It has the associated users +Regex thing removed patch optimizes string splittingIve run all tests in local mode +Fixed checkstyle warnings +and I see your point about how would work alongside so Ill accept your argument against it. +To keep things simple I hardcoded the name of the file to be downloaded in +Thanks Tony. +Assigning to Rick and setting to +Hi DennisThank you for your fast reply +These changes do improve performance and reduce the memory and time required to work with Paths +And also filter and finnaly to have a headers with total number of files and the current file numberIn summary sorting filtering headers with index and total number of documentation needed for file component idempotent options DONE headers with current index and total DONE important camel changes DONE file filter options DONE file sorter options DONE out of box file sorters for by name by timestamp etc +made sure certain defaults are set. +Verified in jbt build. +This can be switched to false to avoid marking the global transaction just because a partipating transaction failed +Ive just submitted a few jpg files with screenshoots and jsp code to prove my prevoius reminderI am using WebworkTiles and Dispaly tag export works perfect in case ofhttpie and httphttpsnetscapebut doesnt in case of httpsie It looks like it is like a mixxed problem between using tiles case if there is no tiles and https it works +management serverSame problem seen in the branch also +intern can have mem disadvantages +It was reproduced when showtrue of a bulk update all resolutions changed to done please review history to original resolution type +The name descriptors might be good finally yes once more to your approach towards best practice +fix patch branch this is going to a branch trunk Jenkins cannot run on it +We cannot use as its implemented in C +patch for appliedThanks a lot for the contribution! +This is resolved in master +I agree +Alexander the problem you are referring to should be already fixed in the CVS version +Hi Scott Our product cluster met a similar problem about job +If the returning disk is cleaned up and reinitialized separately that should be enough I guess +This is a wont fix in and is fixed in +Closed after release. +bq +Can you please provide a bit more information +committed the patches and will close the issue +Ok then how about utilsssl utilssslocsp and utilssslcrl +As to your other requirement exceptions are already stored in conversation scope when they occur see +I just committed this to security branch and security branch thanks Arpit and Giri +The latter can happen if addObject is invoked via the public API when there are instances checked out +Better still what if we didnt need the user to provide this information have the job launcher inspect the ranges and configure the job accordingly +unfortunately I have no scan from +But itsnot possible to test on the filename! There are so many possible bad ones +The issue is not due to addNetwork api as such I am able to add multiple shared networks without this exceptionCan you reconfirm +Please could you verify that it has been applied as expected +Are you planning on finishing this for or should I take it on +Unsetting Fix Version on unassigned issues +Experiencing this Issue as well +Related issue if you create a new table and region mover had an old list new regions get assigned to the region we are trying to gracefully remove regions from +Diff of is fixed with SVN revision +Thanks +Thoughts +In fact testing went against the built broker +Delete will succeed in any case +committed for d like to use this feature from Flare via +This wasnt a problem before +Could you provide the relevant SQL map fileAlso be sure the specify the JDBC type for nullable collumns +They will appear in the lib directory of the bin distro only +I dont see a need to replicate this information in two places +resolved in r some more validation of memory usage and fix issue with storequeuecursor and browse effecting memory usage. +Please verify the fix in the latest. +reopening due issue in EMBJOPR and Seam looks like tag library loading from web application lib is broken by patch +Filters are actually really needed even for basic scenariossince real browsers may request pages and other resources in parallel we cant decide what request to enrich +in documentation now looks like a joke I tried both HQL and native sql as simple as possible and all the time I getIs here any progress? Any plans +Jrme ok to close this with Wont Fix +Hey Rob +Given that Thrift is OSS clients can always choose to recompile the library with their own custom key if the provided key isnt up to their standards +Max new patch attached please m this handles the new defaults and old ones better than before +No response indeed +RB +Unlinking from think the only time the order would matter is in +YukiI noticed that already asserts that the row is in the range being validated so we must be doing a range check somewhere +Hi Dmitriy Im seeing a weird error when pig tries to spill +The fact that your debug log shows two separate instantiations of indicates to me that something is wrong with your configuration +Id happily give an example of fix it myself if only I could find itMoved to and left open for have figured out the issue here but cant figure out a quick fix +This is very helpful Knut I will work on implementing these changes +Ill commit this code for now and refer this to later +Looks good now and all tests pass +It could but that would make upgrading an existing file system harder one would have to compress the ids before upgrading then keep track of the file id +And again this is the time to break things than later +But the only way to properly render an FO inline right now is by enclosing it inside a +Added a new constructor that takes an additional string as a parameter to be appended to the error message +Wikipedia is pretty much the definition of untrusted source for me +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Sorry for the duplicate +The use case is clear. +trivial patch +Back to running w Xmxg +Im talking about timeouts not recent +Question is if what is in that issue is the only cause of this happening but I think safe to go with that assumption until we prove otherwise +committed. +So it works as designed as the default group config distributed map has been populated with node valuesThe first pull the config and after push thats why when node starts on default group he gets the config from the distributed set +I believe most of the users are looking at the wikimanual we need to document there +Im not sure how far that goes at first I was against adding in the decision state since there is overlap with other concepts of the system especially the action state +yeah couldnt it use instanceoffall back? how much slower is the obvious dumb and slow approach +Old one caused a Javadoc warning +Bug fix for filtering coming up +Remy we generally like to stay within the Community color pallet but once weve decided between the two if youd like me to explore some more colors that would fit the Community look and feel Id be happy to +The last remaining files with the old ASL header have been fixed in rev +will close this off as has nothing really to do with the IDE +Yes but in a separate build pleaseOne with Galieleo in the name +Applied the patch in revision +Indeed for this charset concatenation and string to byte transformation are not commutative +Also changes some calls to take ? extends Collection rather than List so that I can use guavas +Im assuming that people wont switch between schedulers very often and so incompatibilities between these wont be a problem. +Currently Nagios is used to monitor the liveness of Namenode we should use more direct and reliable mechanism to monitor the Namenode +Cheers I missed Gavins comment But it works now thanks! +Fix the javadoc warnings of +If there is a problem please open a new jira and link this one with that. +Ive just committed this. +Is this in progress +I would suggest that there is an alternative that an application can implement +Implemented +Once I have those working I will do some profiling and see what else looks like hot spots +OK I tried with WL and Spring and it works like you and Richard Watt said +log and +Thanks! Ill update the patch shortly to include a test for this +Hmm thats what I get for doing this from memory +Such change may break some couchapps aroundAnyway I need some days to figure where is the problem and if this change is appropriate +Without no go Ill commit it this week end. +Any update on the fix to this and if this will be slated for EAPThanks +Sorry +This problem was fixed in release and were up to now +This keeps the existing use cases and test cases intact but works around the timeoutretry problem +Have to investigate how to do it more lazily or make these containers persistent +Patch to allow storing output to cassandra so long as the output matches what you would get from loading either a CF or SCF +Workaround works for me too +Exactly that was precisely my thinking +Committed to in developer mode is failing for me and it seems to trace back to this patch +iBATIS source changes for SQLJ support. +Ive tired this with the latest mvel and cannot reproduce the error +RudiThanks for that tip to zero the GMT offset +The same process with a join gateway works the pull request to master +Would it be any better to distribute a second artefact with all the test classesYes we should do that anyway +Ive got a few Gb to play with on my desktop Thanks for your efforts +Hmm Ill unzip and see if I can reproduce +Im willing to work on a patch if thats helpful the asynchronous routing engine is already there and even has support for using asynchronous producers +The is straightforward +See my comment on where the meta data can pass useful information +Dennis Yes already I realised that and editted my comment a short while agoBoth version and year are for individual filesBut if everone is happy with the following maybe we can move on This file is a component of Apache Foo which is licensed to You under the Apache License +This must be a mistake +By discuss I presume you mean dev +Whats the status of this? Is it complete enough to be marked as resolved for +We now reset the readOnly flag if we hit an Exception during initialisation +We can always reopen this issue or create a new one for the other attribute update issues +in practice i wonder how much it helps lucene at runtime maybe for things like SORT at least it would improve runtime performance by some small amount +Is there a reason not to use Arquillian for in container testing rather than starting the server through the maven plugin +MarcoI tried to apply you file +what do you recommend +Also includes test cases for these +Fix Version should be because all the others have been released already +Event with xmlbeans generated for the schema elements in the wsdl the return type is +There is nothing we can do about itThere is only one option left though +Just use since that ticket had the original we really want to add joda time to the class path and ship this jar out with every job just because its better? You have to remember that every jar we ship with hive had to be put on the distributed cache and if anyone else has a joda time in use by a udf they can run into conflicts +But this cant replace completely because stems arabic text in a way which has a huge effect on retrieval quality for Arabic language textWhat about separating from morphological processing +Thanks. +Applied +This is resolved by and also thought we should have a way to specify which version of serializer one wants on a specific indexThats another reason to move to the approach well be able to introduce some form of selection for which service implementation among multiple options in a generic formStill until we have just one version Id keep it simple when well implement the multiple options logging the specific implversion used by each client will be natural. +If youre using combine it with middle tier transactions rather than let the filter flush at request is intended behavior just enforcing stricter compliance to the usage pattern than before +Need more info to determine if it is or is not going to be done +class instance instead +Then there should be no confusion +The rake file has been changed to download sql scripts from integer or to create scripts through regex +Fixing the spelling +If this property is set to true the sonar maven plugin doent overwrite the PMD and Checkstyle configuration with the Sonar Maven profile +revision Thank you for your contribution Sean! patch seems to be perfect and it has externalized the server configuration +Fixed! +BrunoThank you for taking the time to review the patch +The patch actually lets you annotate by method not just by class +This does not appear to happen in the equivalent HQLAttached is my version with test case +With the Site Plugin reusing it in its current form we would for instance end up with plugin realms for build and report plugins getting crippled and hard to debug class loading errors during reactor m going to look into refactoring the plugin manager +I think I am not seeing what actually gets stored in the btree in this case I dontsee the value of storing anything other than the result of the function in the btree and then keying on thatvalue +I dont know what version of webflow is deployed for the samples on the spring web site if it isnt I can try to test the samples for +or import by name which does not refer to a local variable does have this potential problem +File locking is directly impacted by this as well the implementation of the locking filters is should be much cleaner after this refactoring +Is it what you mind or can we adapt existed code +Edit version and components +Im not a fan of fiddling around in sources until they magically work +Thanks Eli! +Reclassifying as this is NOT a what I am saying + +This issue looks because SLES suffers from it while RHEL does not +Marking as WONTFIX. +Bulk closing Wont fix +It seems to me that userdata should be updated on all that support the feature +Related to believe that this issue is fixed in HEAD and so when is released this will not be a problem +Thanks for the case +Git Info of the Tree is as given belowcommit eaeabdfabbbatree fbdaceefbfccedcaddparent fefbcccadaaccd +You return OK for a query to dbMy feeling is that access to a dropbox db is not useful +On addAll if you try to add a number of elements the first succeeds but some other succeeding element fails then addAll will end up returning true although not all elements have been written +Not to repeat myself but you can have in memory sorting with the new code as well just pump up the MAXMEMORYSORT value and you dont even have to reconfigure the store to switch between a memory and disk strategy +This only solve case a of the original postFor case b we need to use nbsp elements instead of space characters as required by XML +Vlado this is referencing Beta artifacts +Added Teds comments and fixed +Can we use this directly in our java project or we have to create java counterpart project of this +Hi Rickthanks for the patch! Is the attached version the correct one? Im asking because it just minimizes the footprint of although in your comment youre talking about +Revert my mistake. +I run into this issue because my internal repository is of layout type legacy +The test patch is attached +Im working on this one for now +FrankIn both QA and in prod I was able to log in as fmerenda change my password log out and log back in. +Would you please help to try it +Investigating +Looks good to me too +Updated patch for the current SVN want to apply this patch? Or do you think it needs some modification +hmm +I upgraded to version and the problem went away +Committed revision to trunk +So all issues not directly assigned to an AS release are being closed +I wish Id been tuned in to this one it slipped by me at the time +The issue is specific to dynamic sql in a trigger and does not affect procedures +So youll have a set property and when you use this the query will appear exactly as its given + +That makes sense +committed added to NEWS +This mockup includes both workflow and task information and condition information in the same file +Committed revision +Brian if youre able to please resolve this issue +Ne exception appears +HeyDeltaspike is planed to release version soon probably this or next week +Now that work is well underway with AS all previous community releases are +New users who set up single node ZK installs will hit this +OK thanks +ChunhuiSo this happens when master restart and SSH goes in parallel only +Hi Ketan I dont think theres a specific reason however one thing missing from your patch is a reproducing the issue +dependency to must be updated to the patched version. +I just committed a fix for this to trunk +However this feature is easy to do in day +Good point! Moving the issue to BGBUILD +It is now merged into trunk. +If we cache both the native bounds and the latlong box then calculating the bounds from the cached featureType should be able to work for the native bounds as well +Ive made the changes +Please contact my manager Olga Natkovich for any urgent issuesThanksPradeep +jvm metrics information or similar metric information which might be of help to administrator to get an idea about whats happening on complete cluster +I can also report that this issue does not manifest when I use the version of the Spring IDE Core +Thanks +Yes think of region as a +Committed to branch +I think the test should expect a separate directory structure +Using the interpreter I see the hang George reports above but that to my understanding is unrelated to class unloadingAlekseyGeorge Could you please try with the above options? Thanks for retrying these with CU +New patch better error message in NPE change to and also its method names +Should be ok now +Just make sure that the jar files are removed from lib +Correct name for the attribute and facet is iconLeaf +All the jsf TCK cases now passed while we still need to improve the efficiency of the jsf plugin +Will try your patch next week and then let you know the resultThanks allWe have tested the issue against Release Branch and it worked well now the flex components could call ofbiz services directly very cool! thanks +v implements this schema +Patch applied to CVS HEAD +I still would like to do the very last bit of refactoring on but this can certainly be done after the transport code has been spun off into a separate moduleCheersOleg +looks good +Resubmitting patch this needs of course +I think in this case we should test if lib or include is not given +Thanks. +Overall it seems like NTS needs some documentation love with the amount of use it gets but maybe thats for a blog post or at least another ticket +bulk close of all resolved issues. +Thank you Vinay for the patch! +See has in +More over it should build with every commit so it should check for new code at least every minutes +Thank u +Screenshot attached +I will commit this to contrib if no one objects +my bad wrong web news on a fix for this bugWere in urgent need of a fix as well. +this behaviour is deprecated in the artifact plugin so wont be added the latest artifact plugin a version including SNAPSHOT automatically has the timestamping behaviour. +Verified by Oliver. +I committed to branch +upgrading is actually upgrading KVM agent running on host +Sorry Harry I was feeling bored +Resolved but fixes didnt make it into so will have to wait for +That is if he doesnt get to it first +The RPC client will use if and only if theres a token in the UGI +we can always fix the typo later on +Since they both are MO moved the update to occur directly on the RSC thus the state events on DC go correctly as expected and NPE is avoided and fully configured RSC gets injected now. +Heres the patch +fix plus test to prevent it +We need to translate the continuum user details to acegi user it can be used in in instead of the configured in memory database and the login page should authenticate through it when accessing httplocalhostcontinuumsecure of implement acegi interface in Id prefer you create a sub class of out of +I spent a little time investigating what would be needed to get negotiate working based on Ryans attachments but it ended up being more complicated than I wanted to tackle at the moment +For a method that can no longer be supported one might leave the version alone and generate a runtime error to permit old clients to still operate in the majority of cases +It is necessary debugging the lexical analysis +For example in the following uriInfo always seems to be null know when Seam will integrate properly +I doubt thats the right place to do any xhtml validation only xhtml literals should be integrated in xhtml output without escaping if the output is not xhtml then invalid xml isnt a problem +Brunos fix a test +Peter fix has removed the dialog box and replaced it with messages in status bar as discussed here +Dont mount and use a +Same type of problems apply to and respectively +Thanks Mark +I will add an IS NULL expression to the otherwise branch so null values can be caught instead of being discardedPlease let me know if anyone has objection with the change +private int app why +simple patch +Here are some of the problems w FC that Id love to see fixed here hereThis issue could be a foundation for many of those things +Otherwise you could get extremely erratic behavior with multiple. +Does this need to be here + Weve seen due to and +I do have a hidden agenda here though as my main motivating factor is to gather some empirical data to be able to optimize the HTTP transport layer +Im not even sure if it makes sense +I have tested the patch and found some minor changes +So please be way more specific with +applicable to which JEE spec? I dont see this as mandatory so please provide more details reference sample +This update restarts the servlet controler and there the block seems to happen +Ive fixed the immediate problem in by checking to see if the exception is a handled exception +Where possible values aresuiteclass Martin whats the status of this +I dont see how it would be improved. +Bump to next for +If it should be included please mark as a blocker +Bulk change resolved issues before JEXL. +Feel free to close this issue and I will create a new one if I am able to figure out why we are getting this behavior and IF it is rooted in the framework +Part of a bulk update all resolutions changed to done please review history to original resolution type +All tests also passed +I see them! Very cool +Bug has been marked as a duplicate of this bug +Fixed with the latest build hence closing the defect. +Thanks for the review. +New samples simpler than the last part of the issue seems to be changed with new releases and now inner elements have namespace prefixes but they are not recognised throwing Cannot resolve namespace prefix nsxd +Thanks Ashish! +As part of I am moving all the information to job statusI am going to work on soon but if that gets late we can push the quick fix in +This is tagged affects but patch only applies to trunk for me +The issue is script is accidentally registered twice in some cases +Can someone please add more details for this request so I can start looking at it +in to add SP in Fix Version and mark for Release Notes +For indices with alot of terms this can add up to alot of wasted ram +Sure by all means go ahead Ill assign the issue to you +Isnt getting the terms in order more important then being able to look them up in a map +Actively working on it +well it was worth to give it a try +Note that the string may be the service name such as finger or +Its not clear to me whether this is a bug in or +Let me know if I can help +I will change the test +Please test. +Thanks so much +I am debugging the code to see what happens at the execution time of trigger firing +Freddy its fixed with great pleasure o +Based on our discussion and new design can you update FS so that i can file a doc bug for the same +This was an issue on the test only +provides the fix for the multiple pattern issue is fixed and can be closed. +Thanks good to know I wasnt nuts getting the failures when building trunk until +Im sending Nuri Besen a quick patched file but need to look into the issues surrounding this a bit more +The exception in the static initializer should at a minimum be logged exceptions should rarely be ignored like this +I think the behaviour is ok but something needs to be done about flush interval +patch looks good +I think George had started work on this but cant remember which bug it was on +Thx for the patch +Should fix JIRAs and +I had this same issue some time ago while making some tests with ssl enabled didnt fixed it however +There are some for but I didnt see any samples in JWS outside of the +I can test building the without this project and see if the test feature can then be installed +Added the patch to fix this +Im a part of the web team +I used Mockito for mocking as I ended up having to mock and Worker +But when reading up on the information provided I am not really coming to the same conclusion +Thats wrong our current policy for SF projects is that groupId should be equal to the SF project id +At least for Ubuntu Lucid Ubuntu manages cacertificates by package and +Committed patch to resolve as closed. +Let me also ask is client acting effectively as the end user? If it is the same entity then may be it is the client credentials grant which has to be used +ve created a simple project to demonstrate the use of the +Since this problem only occurs when installing via the installer we need to fix it in the installer +Is there a snapshot version that includes this fixThanks proper solution requires to lift the parameter siteDirectory from the plugin configuration into the POM +Dont think Robin ever did this +Alex could you please hava a look at ant Dtest test all other tests commented out succeeds +Ill be out of town until next Wednesday so attaching prelim patch for review and hopefully will be ready to go when I get backPatch Creates a new config option which will take either an octal or symbolic umask value +This change removes the tests reliance on the state of the local disk +So if your module is at the ear root relative and absolute mean the same thing but if your module is at foobar then +. +Im unsatisfied calling this resolved +This is the HDFS part of +Fixed the ordering issue +Here is the missing image +Great thanksIm using Ubuntu +Issue persists even after upgrade +did we tack on the to the end though? if it was us the code responsible needs to be more careful about jsessionid in the does the same +I fixed it +bq +I really dont expect any further changes to this file though. +Since Hot Rod does not support transactions it does not make sense to have repeteable read as isolation level +Fixed in r. +This test expects an exception to occur when readwriting a closed socket +may be retrieved internally through an instance of the +Related I think we need to either not use QP internally or add a only mode that skips and auth +Ive been using the Python interactive shell pycassa as a mock +It is important if the test is generating a manifest using the Harmony classlib correctCorrect +Sync thread is notified as well to flush file buffers +Can you do it again with a +that would be much simpler +IMO we should continue to do chunked checksums maybe a CRC for every KB in the block +Please forgive +Would you like to remove those as well? If so I think it would be ok to change the summary of this issue to be something more general like Update demos to remove references to and make those changes under the same issue +David if you do the same in Java isnt it then also very slow? Also you include startup time in your time command usage +Adding the mapping for JBDS to Developer Support resolved the issue +failure is unrelated +Only addresses for which messages have been dropped will appear in the Map this is not a bug +Ive done some initial testing and so far thread count after failover remains stable so it could be that I need something closer to your use case to reproduce +If any improvement required then please let me know +Id love to help fix this if anyone can provide me some direction +I bet there are are bug fixed in it +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +writing extension and then renamed after successful far as the local copy I am changing that as +Thanks RobertNo theres no need to try even older versions +Here is a patch doing the same +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +See patch for fix +Hey Jan your sample works fine after changes I made +Contrary to my initial fears Im anticipating that this will actually help with and probably even help increase IO throughput in the data directory in a dramatic way +Yes also think we need to add a special type of function to send changes to server +Interestingly if the Java module is included in the client then the marshalled module is unmarshalled as a Module and the sample client code works +That was fast work +The patch for tests excluding attached +The Wicket framework uses rows of divs and then indents each row depending on the current level in the tree +Environment with may come across same issue we need to add it in FAQ. +Changed to allow proceeding in the wizard when only warnings are present +Again a nice to have +Does that sound too awkward +It would be a matter of having a new scaffold provider like the default Faces scaffolding but with separate templates apart from the dependency installers +I think this is fixed on trunk at least the buildbots are happy +the POC is committed to support will be a new feature for JBDS we will not document everything it can do in the release notes instead this will occur in a documentation guide and the new component will be mentioned in the overview and new components sections of the release notes +Or its sth bigger than perspective +OH Thanks for clarification +Is this fix going for +Will try to use your changes to implement thisChamikara +patch is against but applies cleanly to as well +This has nothing to do with FOP +In general I think the memory issue is a real and very serious problem that isnot isolated to a pathological test case +Added html docs removed println in test +To be clear this bug also affects but we wont backport this +Check the airline sample to see how +Adds the missing PRDSCUSTREGISTER to thepatch does not include PRDSCUSTREGISTERas does not my patch that does the same thing +All others can be addressed at a later point and should not block the release +And the volume ID? I always use zero the lapi documentation statesvolumeID the volume ID of the parent object +Maybe we should handle this the same way we need to handle Asynchronous? To add the explicit boundary saying what parts and mechanisms get their own +Fixed in revision +Do you all concur +Yep +After a discussion with John this problem implies that there is an error in the validator +bulk close of all resolved issues. +It adds to the langscriptlang tag +It was a simple line fix +Thomas +This one are in builtins at least according to the patch so they need to be in docs +doesnt work if there are indirect dependencies +One day when all the checksums are valid. +The cleanup runs once a week. +I wont have time to investigate for +The JVM developer that analyzed the dumps seems convinced that the JVM was doing the right thing and there must have been some unknown problem with the state of the machine which remains a mysteryClosing Invalid +Alex it would be great if you could provide a patch given you can verify and also are on the cusp of the issue +having fun. +Is there any particular reason why you felt the need to open a new ticket instead of continuing what was being discussed +Obviously we could rely on the shared version of Dojo provided by the Geronimo server as Paul has already mentioned +Some of the unit tests which use are disabled +I just grabbed the latest code from Subversion and everything is working just fine now +Minor changes to to spell ij consistently inURL consistently in upper case and to remove the strayIm intending to commit this patch to the trunk in the next few days +See comment above +If everythings ok then well close this! +Does it apply for you +The client receives the CNFE but doesnt end up closing the channel +Wow this is cool! Thanks for sharing +Fixed on M aim for more comprehensive solution to setting socket options on trunk +Return code of would be used in many other cases for example if were looking for all versions of a single column once we scan past that column in the current storefile we return to jump to the next one +Jean The list and scan and even put works fineAs Andrew suggesteed i think the relocation of the exception package has caused this +We also need to change the test accordingly +then everything would be great if it was possible to configure the httpmethod on the or the query params would always be added +fixes this issue for me thanks +fixed the issue in class +hod contrib was removed + +Attaching a am a patch from the oozie root directory +Wow +Here we have a massive failure +Im sorry I forgot that in modeler cancel activity on default is always sets in true +I believe the current thought is that the HTML tags since they perform functionality specfic to S can still be used where functionality is provided by JSTL the JSTL tags are preferred if for no other reason than the tag libs havent gotten much love lately +Youre welcome +I cannot reproduce the problem for any kind of project that has it will be of help +Verified on +plan the fix for ve begin to work on this issue and for the moment Ive just introduce the interface and refactor existing code to use it +After investigation I discovered that it is in fact the JDBC SQL logging package that causes the problem +This issue is observed againI have attached the required logs +wasnt being worked on when we ran into our way I think that were getting off track here +Thanks DennisDocumented as a fixed issue in the EAP and release notes +This probably requires modifying and possibly adding utility methods elsewhere +Refactored a little bit +That is true sadly one of our clients uses IBM product to send soap messages into our web service and it put carriage return before the first boundary. +deploy and get httplocalhostjbtestdo to VFS +I asked the Tika guys about their release lets wait a couple days more +This is to address the case where the archive is overwritten in between two reads from the same process +Although I agree that the current format for statistics is not very friendly to humans especially when hardware performance counters or Immix collectors add their totals there is one thing that makes it IMHO superior to the proposed format +Replace hrefurl or srcurl with httphost url +project file for you automatically +Im ashamed of not seeing it! +will commit shortly +Run As +Here is the patch ot fix this task +If in the future someone would like to tackle this feel free to open is and attach a patch. +Sorry I meant. +reopen for clarificationWhat does patched feature meanHow is it namedversionedrepackaged? How does future updates work for users +Can you run on the patch +add a new cli deploy thanks +So basically this patch delays the decision to check disk space as opposed to Todds proposal +That is the expected behavior whenusing option D +VivekCould you please zip up your schemas and wsdl and upload it to this JIRA issuethanks WSDL and test case file +Harder than I thought +is supposedly fixed in which is already being used in the main branchIve been testing it on this +I ran and it came back clean +So boolean foo true +No time left in to get this in +has been removed from EAP +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Thanks PhilippeYour patch is in trunk at r at rI was notsable to merge wih +slip all minor and below +This issue belongs to hdfs thats why patch could not apply +In addition as we have no spring experience on our team its and that the is the correct solution to use +After that we will upgrade to Seam GWT in the process working out how to do it while preserving GWT compatability +Trying to assign an issue to me +If you say so i was informed by Jboss respresenative inAntwerp +Ill go ahead and close. +Will debug through it further +How could you not and avoid the exception? I think the exception is OK +I could save to save the Starting process +Eric can I take this issue +Now I have a way around that now +Here we have to note the anyone can encrypt the message since it requires the public key and the possible problem for the intruders is to find the private key to sign +MD class and test class to as Ive asked Alan to confirm licensing of this file +It would be nice to hack up a test case +Yes if I change field name it you in advance your notifications about +Therefore Im going to creating two modules one for tests with automatically enabled handler and one for test with automatically disabled its going to be all right. +I think if you continued the test runs you would see the cache size top out at +This patch is only for +New test cases will be match through Interesting that a match pattern ending in childx or childx as inmatch is just as good as x or x +I will look at how you did the interpolation and try to follow your exampleI see how what you did is useful in the build infrastructure +Ok so you can change the comment if you want in the ask Remy to remove completely the NIO connectors from the next release that goes into AS +A patch is Stepan +Changed command line in Hudson to bind to a specific interface instead of localhost +Allowing the inTransformer to be called without serialization may prove useful and is necessary forAs a Transformer the inTransformer should distinguish between interim transforms that translate only elements from the current catalogs and final transforms that resolve every in elementattribute and remove the in namespace +I have set up an XP guest os under Virtual Box on my laptop +We got only following test failed which are failing for us normaly +My patch was consideration shortage as shown in the point +Im not absolutely sure if this bug is somehow related to transition to closed remaining resolved bugs +I take the comment about Package Explorer back +The patch fixes a problem by replacing destCM usage with dstCM +Hmm +I mean it doesnt contain any entry but the following headerTemporary index for unavailable sources and javadocsMon Feb CET As the file is created then written its not a problem of readwrite access +They can be I presume by stopping and restarting ofbiz which is usually not a preferred solution in a production environment +Section syntax rule appears to be the text that that Jack observedSo perhaps we should resolve this issue as invalid +you Martin +The panel will be built on the fly when report is loaded. +Unfortunately Dominik didnt specify which version he was suggesting fixes for EAP or result is that eg +So do you mean move the code around handling the connections that is done by the destoryService into the stopService or am I way off I add a call to destoryService at the end stopDelivery after the stopService call everything appears to work as excepted +Perhaps just hack on Configuration to make sure it is thread safe +By introducing the wereshielding users from its location yet well still have a base exception to useon the JDO side +Or not +I updated the review request but not the patch here +Same for procedures +Tested with on. +Patch for upgrading to and upgrading Hadoop test to and all tests pass +Ill find a Vista machine and see if I can figure it out +The take patch works fine for me +Pull wsdl using thewsdl URL +Updated Title +Theres more work to be done set the maxAmount for all payment methods verify all use cases work and make sure the amounts get invoiced correctly across payment methods +Xerces figures out the encoding by looking at the first bytes of the file +Further it would honor failure zones such as racks yes this part of the code has to be correct for data integrity +seam +Its not just the packaging entry as its explained in the linked issue +do assumes the cluster should not start even if just one of the dirs has a corrupted log shouldnt the cluster start as long as only one of the in progress logs was truncatedNit it has no transactions should be indentedOtherwise looks great +Fixing in trunk revision Should we also do this for the release +Gregory? +That makes sense +Committed this to trunk and along with +docsupdatednothing is required to add to the guide +Anyways I just wanted to let you know it was not forgotten about +My suggestion was in accordance with your descriptionbq +Thanks +So I removed the static block and all the static variables +Attached patch file applied in commit befabeaedbebdfbeeThanks! +runtime config does not support adding plugins +Another option is I could just use that classloader as well +Thanks for the patch! +You could vote on the issue. +Clearly it is rather inefficient +Structure of installation changed there is no eclipse folder anymore everything is in studio folder +I now tried the same on a windows vista maven roo downloaded and installed all software fresh from the run the provided script and copied the Converter modified the run jettyrun added a client added an account still got the same error. +Note that theres nothing in the EE spec that says that the java JNDI context should be writable by the userThe mechanism specified in the description allows things which inflexibly require JNDI binding to be effectively faked out so that it can still work at the cost of some perf +API itself may need to improved but this is an active area of development and dev list can help with API improvements while GSOC project focuses on plugins +Therefore the result of the script should be consistent when same parameters are passed to the script +Sorry leaky abstractionsThe best approach here is to define an interface well outside of the component package and have the component implement the interface +couch file +If hodring section contains syslog address then hodring and hodcleanup log is going to syslog +Added patch for branch vGACP +What is the plan for this issue? Open Unresolved Fix version abandoned +In addition seems to be doing a lot more than what needs of itWhat I would like to see is separate patches one dealing with what and need and another to work on the optimization +Revision +Note just to repeat in the CR build the installer successfully installs EDS into a standalone servers default profile +On testing a bit further I can see that the problem isnt an issue with the proxy at all its a problem with the fragment identifier screws up the JQM page view system +I have to run on my local patched version for my +I strongly suspect there was a reason to switch the order I know I have had to move things around in that area previously to get some of my changes to work +you are right +Has native maven support no silly imports +The other piece is to have a mechanism on the classloader to force close the +Committed across all branches +You win +dfsadmin report is too painful because its difficult to show checkpoint in dfsadmin report without adding unnecessary Ill implement and metrics instead +Removed the comment reported by Kathey + +Ive added the clean word and rerun the job two times +Thanks Vinay and Jing +The function now use the defaultAE and return resulting Jcas from analysis +Thanks Tsuyoshi +Youre right Kathey I overlooked this +I just committed this +Note its in CR not M! +Image of scrollable in VPE +Ill keep you posted with my progress +Closing issue. +added an example into showcase that simulate this bugAdding srcmainjavaorgapachestrutsshowcaseajaxSending srcmainresourcesSending srcmainwebappAdding srcmainwebappajaxtabbedpanelSending srcmainwebappajaxtabbedpanelAdding srcmainwebappajaxtabbedpanelnodecorateAdding srcmainwebappajaxtabbedpanelnodecorateAdding srcmainwebappajaxtabbedpanelnodecorateAdding srcmainwebappajaxtabbedpanelnodecorateAdding srcmainwebappajaxtabbedpanelnodecorateAdding srcmainwebappajaxtabbedpanelnodecorate +DavidSounds reasonable +Appologies ill update the test case +Also attached is the test itself to show how it is used +Please close this one off when you have time Dan you +Please verify +But returns a valid object on a path +Paul very good progress +Also upgrading of a major component is not done in at request of our customer advisory use a newer Xalan the customer should consider upgrading to the EAP +then it is safe to change ledger +Thanks Prasanth! +We havent heard from him in over a year +Please reopen if you can still reproduce the bug +Also it seems to me that perhaps umask should use the api from exposing itself as an or at least linking to that in the javadoc so that one knows how to use that valueIts a good idea to return umask as +In the case we do create a new class it might be better to have it in the util package +I fixed my copypaste fault with version +Will commit shortly +Sounds like a different issue. +PS I think if we like this approach i can include throttling too +I think its ready to commit +getContext should ALWAYS return a +Uploaded patch for patch for trunk +Feel like giving it a shot +I am not sure why Ive verified the classpath +But if the caller is coming from Java and we have a method we could then flag that a interrupt is being requested +Additional patch per review comment +The job of the password callback is to supply a password to the validator for the given user so why would it need to know the password that has been receivedColm. +We might need more tests depending onwhat the investigation reveals. +moving to did you say you had a fix for this + +Thanks Jeffrey! +on the revised patch +There is also no test for successful territory based database boot but detecting no Collator support when a SQL is issued which requires Collator support +bq +fixedthanksdims +any update? I have come to this question again and again nowthanks +Phillip is this resolved for this CP? If not please move this to the next CP +It sounds like you havent read my first comment in this issue recently +hmm Its may be a good way +Ive marked a sentence as. +At this point Teiid does not have a Class +bq +So that would explain why I cannot find the code for the GWT Shell it is not provided by this plugin? I was under the impression that it looked so different because it was brewed by this project +Although I havent been able to come up with a unit test to easily reproduce this issue my organization has been running with this patch in production for several weeks now and we havent seen this deadlock occur since deploying the patch +See for instance inner class of +Thank you Lily for looking at this issue +in release release +thanks for tracking this one down +And a trunk version +Just applets otherwise + +to moving this to +Thanks Micah both for this one and for +This works due to other reasons but this patch fixes thisc It will improve on performance +Committed to trunk. +It would lead the casual observer to think that the size reported by after the flush is the number of bytes flushed from the cache +In first cut some tests hang +This test illustrates the XOR split issue +Have you tried it out Moaz? Do you know if it breaks compatibility with? Thanks +I have just downloaded and extracted the binaries +I agree that we should try to do that as much as we can +Once I get that successfully done I will mark the jira as patch available +Please review this +Thanks for reporting this. +Im running the regression tests now and will commit if they pass +I guess it gets resolved with version +Hi Alejandro are you done with your third comment under about the help message +Hi Adrian DickI want to correct some statement in my Description like belowThe complex array is an object not simple memory block so should not been deleted directly by delete +Fixed in r. +I missed in this patch the file file which i forget in the patch +I had not properly uploaded the release file +Works for me +Using the old parser the Example works problems occurs only with the new parser +I fixed some problems to avoid an in revision including a patch proposed by Eric Leleu see for detailsEmbedded subsets only contains a subset of glyphs but nevertheless their may still pretend to consist of all glyphs +didnt remove the ROOTWEBAPPLICATIONCONTEXTATTRIBUTE from the before when shutting down since it expected the to be released in any case +Could you please make it againstthe trunk? Thanks +The example supplied by the reporter should be sufficient to test the patch +I wasnt able to reproduce the problem using the fragment you provided +It wasnt really an issue until we made the change to page in messages for browsers instead of limiting them to a single page +Sorry for the status change mess this is the first time I submit a patch with JIRA and it is the end of the day here in Switzerland +Needed to add optional parameter discussed in the dev list will change the parameter to misclassified. +Just after writing this I noted I had! Im going to have to get used to intellij not showing me these ll close the bug +Matt do you think you could integrate this into a test in the main source tree? Ive tested using your sample and its fine now +Im not sure this is a regression +This is information about how to add another tab to the management console +Tommys response is the same position that Lucas took in our discussion +Let me know if that presents a problem or if you cant get round to it I might have some time over ChristmasI think maybe you were under the impression that this had already been done and were looking for it to be in the code already +Im not sure if that should be an option or if its OK to change the default +This may be a tallimpossible order +IMO Don Browns comment makes sense if the the first module is exclusively usedas a base for other modules +Please check that it was applied as you expected +Ill go ahead and close this one +The problem was a bug in jetty where the JSR filter was hiding the type of thedispatchThis has been resolved in Jetty CVS with an explicit pass of the dispatch typeA new jetty release will be filtering through maven in the next day or two +Patch +Good question +If I do this then users would receive with link parameter languageSOMEID and this SOMEID could mismatch with their languageid which they have in their profile settings +Any update on this for applying against trunk +There was a minor issue in the formatting of license header that caused checkstyle failure changed that in committed to trunk and branchThanks for the contribution Daniel +Its not +In the hour the agents have been running theyve gone from a maximum of open deleted files at once to +root were missing on client side. +Let me know if you think otherwise as per Apache pig guidelines +Sorry I missed asf I was told was the range of hosts +This was working correctly in but after uzpgrading to the ranges seem to be wrongWill add a testcase soon +Im happy to do the workAs to keeping the code clean at present I find the code quite messy as the main code needs to know the internal details of the +In order to walk across all keys on a single node we would need to bound the range by start token last key seen end do you have an opinion on this one? Id like to revive. +Thanks +Thanks for the report Lee +Dont use subfolders please +All those checked looked to be updatedIf a particular file is found to be out of date will JIRA that file. +Created a patch file containing initial versions of timezone handling classes and arquillian terms of formatting thats probably true as I have not had a chance to run the code through the formatter once Dan created it for Seam +Can you also log a JIRA against the ESB documentation so this gets fixed +The primary reason being that the physical limitation of one data center is insufficient to host the entire data set +Applied in r +I have noticed I have this problem only on forms of list type +According to you what is missing? I have already used successfully the plugins on WPF projects +The test passed for me on Windows +Advantage of improving the interface is not just avoiding one copy but could eventually support better handling of direct buffers +Thanks for the help +Please comment either way so we can make some progress here +For MRR the third stage cannot start until the first stage the done +Included at the required places and done improvementsAdded Employee Leave tabCreate Entity and crud operations by Parimal +Are you still using M +Pushed to branch and kicked a build even though I didnt have permission because I felt this was fairly critical for future testing with the next batch of EAP builds +where can i get them +Good comments Julien thanksPlease see the comments for why we keep mbytes in basically its public and changing could break things +The problem now is that Cascading is not yet Hadoop mapreduce compatible so we cant use some of the new fancy things that we are using to deal with mapreduce in +Attach a patch which addresses the deadlock in the following way +Omitting it was an oversightYes +Will leave it open for now hopefully Mike will have some ipv nodes for us dont think is being pushed to +HiIn attachment you will find a patch that fixes this issueBe careful a recent commit on the project makes unstable the validation of PDF with or fontThis issue will remain open until all my test set succeeds with font programBREric +Patch committed site deployed +corelibmanagedcompile +Lists added +That looks like Solr syntax +Got from Daniel in RB +Technical details Allow user to specify layout and change +Seen all the way back to +We need to make a choice on this Do we directly provide features? Do we wrap them in a require way to provide our own grain and label in UI but let people install other versions from external sites? Do we wrap them im an include way to provide our own grain and label in UI but force people to use the version we decide for each feature and force them to update features only via provisioning +This kind of passwords is not related at all to JAAS +I cant reproduce Drop have asked QA to do this on development version of JBIDE they cant reproduce its too +ill push for a stable release of jaxen +Made changes inAdded in implsrctestresourcesorgapachemyfacesconfigimpldigister +There is a link to a Geomajas class that illustrates our current approach based on +There is a basic test at +We are bringing forward to fix the maven plugin resolution bug which means yes many issues will be pushed back to this issue still scheduled to be fixed in +verified in ER +I went with Thrift because we havent finished the import process yet +I like these changes a lot +closing issues for released versions +But there is jar in EAP zip +attached a patch with license box also this to Federation Branch. +So as soon as we deployed our software we had lots of requests which finally crashed the vm +Fred would a simple mvn install of the plugin dependency not work too +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Currently confirmation check will check only namespace dirs +I believe Ive fixed the buffering problems with very few changes to the compressor and decompressor +It just makes a huge Test class which is undesirable +If an externalrequest is needed just use http in attachments src +Project Clean +Andrew Thanks for committing the patch +Committed to trunk +Thanks Ashwin! I committed this to trunk and. +Committed revision forThanks Gabriele for your patience +Hi Bart the problem occours also during cluster initializationSee the attached logBR LucaI suspect you have configured a validation timeout on the datasource itself in the dbcp settings +Colm your patch looks very fine! for close for +re running the script I fixed this script so if you update to head for that file youll get the fix and it wont mess up other parts of the website if you run it +Indeed whats the status on thisSpring seems to massivly jump over to Gradle with all their projects so its very legitime to allow Spring Roo this jump as well +Tims comment was correctly applied to and but was mistakenly applied to this JIRA +Use the select generator if you need this +Fix is to have master recognize the RS restart before the zk node expires and handle it right away +Cannot building problem is fixed but the eba file is not built into the car file another JIRA to track it. +Fixed tests for revision on trunk and on branchx + +the changes in seem to change semantics it used to just WARN if permissions couldnt be set but now it throws an exception +Dan thanks for the report ive commited some more changes but havent got the time to test that i will test after Devoxx which happens next week +Looking looking at the code Im not sure why would be null +Ill make an issue cross link it from here +Committed dont worryMaybe you can find the time to review the contribution +When I stared at it last I was contemplating writing a huge monolithic thing to setup a view and do all that crazy stuff +The patch looks fine but the error handling does not look quite consistent with new error handling code +Here is an updated patch with all the alterations necessary for the new rmi layouts and the addition of annotation +I committed the patch after checking it against the CT +Are you specifying encoding on your xsloutput elementI do not recommend using a Writer in the constructor +Thanks! +Thanks it works as expected +and the will compliant about it cant servlet when starting the endpoint +This way we wouldnt need to worry about bootstrapping scripts or other such detailsIn my view the main use case for this standalone package is to make it easy for people to test Tika by throwing different types of documents to it and observing what comes out + +Please shift it +No disruption is occurring in this run due to +It was fixed and we didnt have to change anythingI dont think we should change the project name because of one buggy implementation +From a quick scan of the mmx logic I believe that we did allow this and it has been part of the SQL spec for quite a while so we should support it on the server side as well. +I looked over the patch and I dont see any problems +While on Linux the default provider is it could work correctly +Sorry my mistake jboss uses old libs +Done will look at the +Will my Groovy beans be wrapped in a transactionThanks think you have misunderstood Juergen comment +In order for Lucene to use something of Solrs it has to essentially fork the code +patch applied at with. +Btw I add these log message via the attached patch +Damn you too? Its been in there for eons and Ive always had a suspicion that no one uses it +I understand Im coming a little late but I actually remembered that for counters weve introduced the F option that allows to specify how many different keys to use +And to be honest that class is so convoluted now that it has to handle both simple and features that its becoming unmaintainable +with the attached jgroups configuration I get a startup excerption +Also coordinators doing RR should only include nodes in the readrepairdatacenters set +the latest patch looks pretty good to me +It shouldnt have fixed anything +BB Contacts API has been addedLorin how do JIRA issues normally get closed? I assume we wait for to be released. +Fixed selection in tables +Anyways Im happy to review and commit this if someone is willing to put up the patches +Hi Chris +In most organisations the admins of a system like continuum will not have direct write support to the ldap server anyway +The patch can apply to other interface easily but wasnt trying to relax in general +The performance speed up with this strategy is remarkable +Spec has been updated to conform to Java long before revision +should be separate as originally. +It is only fixed in community for the moment so I keeping open until that can be verified. +Thats why I used separate signatures for those methodsbq +I think that and validation should work in aconsistent fashion +This patch modifies to display the right currency for an aggregated productto display price with TVA into LAST PRODUCTS screen if configured in store +Im reopening the issue for reopened due to high users is verified in provided workaround +for this patch in its current form +Same Lucene default +The old one is obsolete due to fix of patterns +I found another error in index updating its loosing locks when using and lucene as backend +Joachim is the +Perfect Ill definitely take a look and start with the documentation in following days +I saw one question on the list asking for the doc plugin for the release am searching for an answer for that +You can see for reference +Ill post an updated patch soon +I think this is ready to go and will likely commit later today or tomorrow +Please reopen it if you have objections +Do you think that would be possible +Attached is the XML that Ive used +Alright I think this issue is as close to consensus as possible so Im going to get started +Ted yep +Im going to apply this patch +PR for master merged in master branch +Santigo Is this even relevant? If theres no legal requirement or ASF requirement for having committers in the pom Id prefer to just have the simple text file +I agree but only works on actual Iterators not like an iterator but throwing an exception +I had to change it to dgrails to make it work +Cool +Added output of projecthelp able to reproduce the problem on a separate Windows XP machine +Rick if you have planned to do this yourself please feel free to reassign the issue +Thats definitely odd +We would have to override writereadObject or writereadExternal and serial would not be used +Then we can ensure that the codebase is not duplicated +Im closing this issue for the time being since our preferred solution for modularization is with individual per bundle linked together via service references +Test failure seems to be hudsons quirks +I have uploaded a new patch where i check if the dir HADOOPHOMEsharehadoop exists based on the result i set the hadoop +on Honey Badger +in tests now run successfully +Sebastian thats a reasonable approach although it wont cover everything +Added confdir flag for RPM build for startup script to locate CHUKWACONFDIR +Exception looks like logs in as dulpicate of +Minor quibble set compression type I dont like the hardcoded list of supported options there as it seems like something which can be easily missed should we add more compression types +Pentahoi Pentaho Reporting Pack for RHX + updated and Patch attached +The initial patch I submitted only fixed the addition of field errors +Thanks for review +This could result in some long names perhaps but its predictable and I think its a strategy that many people would use in lieu of a binding file +Posting updated patch for +Never mind this issue is invalid +For what its worth I have tested the bank transaction client by using it as a benchmark to test the JDBC client side statement cacheNo problems discovered. +In the latter case Geronimo should offer a plan snippet containing the changed parametersSee also +it would be scanned again so it would hit block to print the error messagethe patch did two things it would check the log file existed or not before proceeding scanning the file +After change tar tvf +Ok Ill suggest a solution and post it in the design forum +fwiw we found the to be pretty much debilitating for write throughput +Nit binsh should be binbash Nuking the scala reditribution would be a plus but not a blocker IMO +JacquesFor Summary b why dont you practice on the Example component? For Summary c are you referring to or? +Fixed in +It makes the serde more consistent with other serdes +Send me a private email to carlos at apacheorg +proposed interface changes for the shouldnt need to bother with the tracker or status +server creation never ran runtime detection its always been like this. +is a subclass of +I will try the Classpath CVS head to see if this is fixed +Please provide the project version does not match project name isnt correct +The serialized predicate is stored into the jobconf and the sarg api provides a simplified and shared mechanism for record readers to implement predicate push down based on the serialized predicate +Committed r to switch from HSQL to Derby in is also embedded usage of DB in testsAnd as well in testsMaybe they are using HSQLDB as well +See for further details. +In either case I think you do need to handle the null for the ID +But regardless this sounds like a YARN issue so moving the JIRA to that project +We need to somehow capture the actual output from the forked process that executes the release and display that theres some noise in the patch looks like some inadvertent whitespace changes +find a user that exists in user that users username and email in the databaselog in as that user with the users email will create a new user in the databasenow log out and try to log in again as the same user using the email address should not let me know and we will work with a legacy user +May i hav esome more information on this. +Thanks for explanation! +Otherwise should be closed +JMS on AS is currently not possible due to should be fixed in the next week or two from my that is why the push demo needed to be optional +Will add a note to the reference that the invoke procedure may need to have restricted access. +Im going to commit this s the patch I just committed to +Wouldnt be confusing to users because when we package a binary release we only include the scripts not the +in this case it seems that the th or th retry succeeded but reading took nearly hourHemanth could you please provide related datanode logs to see what was really happened there +fixed in svn r +finally we use mina for all our sockets! +So is unchanged +Patch applied and committed not added to the launchpad bundles reopening. +Is cloning theMETADATA table even allowed +If you then get the same exception you wont have to work through my sources nor do I need to shop out a working example +just replace t with in hql for context code complete evaluation it works correct id proposed in context +Its just that using is more consistent with the rest of the code in +Thanks Ed +Upgrade to Prototype while I was working everything over. +Id go with for now +But can you reproduce the problem regardless of any unit test cases +Instantiating the validators with really isnt intuitive and makes the code hard to understand +unfortunately the component got archived we could import the latest version to make modifications or if it only affects a subset of components they could be moved directly to Spring +This issue has been committed for a while. +The problem is that currently the exceptions declared in the interface are not actually thrown to the client and the test code has been written to rely on the behavior of the wrapperWe could Change the tests to not require the exception name in the error string +Added to fixVersion per and svn log. +if there were no separation between reader and writer and small segments were merged on an open why not just write out the result so it wouldnt have to be done again? Now move merge on an open to merge on the close and thats what currently does +Marking as resolved since the TCK unit tests pass. +I see I just read to the bottom of the Wiki page I spoke too soon +You know I just tried to improve it by applying some of the best practices about web design I have come across so far +I had to cut the previous one short to work on getting a release out that is compatible with Grails +For me it works if +I like having the sstable package +approach is better and should be preferred when possible +If this would be a more desirable option let me know and Ill throw up a new patch that works that way +Here is the technical ins and outs of the issueThe Faces Lifecycle object is split into two faces execute and render +Moving to fix +FM in this case its not clear to me why you have an API and IMPL bundle at all +bq +I believe this is a dupe of. +There is also a inAnd I think there is another place in the model in you may need to add it +Do you fancy a try at the ssh client to sshd server gsh command? Ping me if you have any questions +If you have both pointing to the same name dir you will definitely have a problem +No longer necessary. +I dont know how to combine patches and I already committed that one locally +While I was at it I exposed the closeRegion that takes a and a HRI +GregoryShould the Fix Version beThanks +Also remove some unused imports +Thanks for the patch Rajeshbabu +Ill test it as soon as I can but itll probably have to wait until after the weekend +Attached patch for this issue +Thanks for letting us know Tim. +Its actually for each shard in the table +We needed to check if there was a before calling it +Have you managed to update and recently to see if it still exhibits the problem + +Thanks issue is closed now +Updated patch on the basis of version according to Jacks comment +Sounds like its coming from packaging +Cool +bulk defer of open issues to +It was already set to that is a permission thing we dont let users set the m key because of the logic that goes into setting it up in the repository and users invariably dont understand the ramifications of settings things +should we consider adding another field to track the information you want to fix the bug introduced by +This used to be resolved with a memory map flag obviously we do not provide a ui for that anymore +Thanks for including this +I should really stop working as invalid per feedback +We may have to revisit this if we get to support of multiple workspaces +r does Alfresco CMIS Documentum and Dropbox connectors +pathtosomeservice +i know its a stretch +I was looking into the latest that we have built in brew it contains the jaxb built on +If too slow we can come back around later with +Apply this fix to both trunk and branches +thanks for the addfortunately theres a patch available for +Unit tests pass for me just fine +I just committed this to trunk and +Hah were duelling patches +On a side note we need to refresh the branch because a lot of compatibility breaking PB changes have happened since the branch was created +also These identifiers are holding objects back to the Parser somehow +v patch uses a more stable method to check for metrics +The command line were passing to wsconsume via the tooling is as follows c t e v k s Cworkspaceshwcsrc o Cworkspaceshwcbuildclasses fileCworkspaceswsdl +Verified the Deployment section was indeed removed. +was deprecated but is still supported +How one can enableuse it? Please share +I merged this change to. +Pursuant to some suggestions from Ben and Mahadev Im attaching a modified patch that gives a diff between the files we started with and the files we ended up with +Removed in the JDK branch as of revision . +Heartbeats would then be very cheap since it would just be a state update on a few tasks and a dispatch of already queued work + Committed +Does seem to be rebuilding a new configuration and each Configuration is making a new hash table +Can you please confirm if there is no change in this hcatalog code +Patch is against the git repository btw +committed thanks Krishna Kumar! +bq Konstantin if youre trying to kick a new patch build for this you now longer move it to Open and back to Patch Available +btw passes with the patch I posted +If test it is too heavy and you said it should be for integration test +another way is also to simply have that generated as a resource which you simply compile and add to the classpath at runtime +Adding here for a more focussed discussion +The resulting pdf after fixing this just merged the resources of each page but left The document wide resources which are shared with all pages alone +Either way it is unrelated to this bug +bq +Besides the jar Ive attached two test classes +We are experiencing this problem on production systems running of Jackrabbit so would also appreciate a backport of the fix into the stream +And at least it should be possible to show this fail in the my opinion the most important thing is that the user gets notified that he misconfigured logphp or has a typo in his application because whatever we do it will not behave like he intended +what does it log at debug levelim guessing something like Not safe to delete commit log +Will you commit it +Committed to master branch +I assume I will need to modify to load the data obtained via the query expression after the conglomerate is created +Removing the M option appears to work for static mode but fails for the dllshared modeI dont know enough about the icu package or what pkgdata does but Ill hunt around a bit moremake Making all in srcMaking all in xercescutilICUresourcesoptbinpkgdata name xercescmessages mode dll d +There doesnt appear to be a problem in Struts itself +The encapsulation of this projects objects is a bit different than that in so the patch may need to be adjusted +Cherry picked into. +We need unit tests for the new functionality in +OTOH that might make it easier to make mistakes in compatibility +Ill attach new patch shortly +has its own problems which I will comment on separately later +I wonder why this still isnt added to trunk +This fix is then confirmed to work in large environments +ebedbecdb +here is a simple that I the tests actually fail with this patch +It is somewhat like asking the Linux Kernel project to also maintain a PMML library they could probably do it but is it the right group of people to properly maintain it and does it even fit in with the projects goals? For Hadoop proper I would say no it does not fit with out goals but it sounds like it would fit perfectly with Mahouts goals +Lovely +to TRUNKThanks for reporting this critical issue Todd and SebastianThanks for the patch SubbuThanks for the review Michael. +Thanks to looks like its run both sets of tests nowWhy comment out the user test +Corrected the fix version +svn moved +It results as no error but keyspace wont be created with any placement strategy +Many thanks for the good discussion! After digesting these comments for a few days Ive come to the conclusion that I should commit the current +xml wc l find +Is there any reason we cant have default to true? Otherwise its going to get forgotten a lot +Thank you! Y I agree about both parsers failing to grab the annotations +My bad +One possibility is that we use FFI to build a bridge to the Growl API through C. +I missed the code in RFC +same as in a UDIG problem it is occuring in geotools. +You said injection is awkward because special support is required in the parserquery logic to recognise and cater for the tokens that appear in the same position +Maybe we can just call it in SVN +Here is a first stab at recreating our internal tool with nice upgrades like updates for file and directory contents +Have updated test case to check for this condition as well +just sees one tokenstream for the field and doesnt know about this +Removed unused stop and unused default for nn port also moved all System +I suggest to close this one +After checking with Anthony the dhcp entries wont be cleared until the same macid used by another VM in the same network at that time the entry would be replaced with new entry for the new VM +Committed to trunk and branches +Lets wait untill Jan to make this decision +Applied the patch with revision +and unmaintainable +Hey Peter I understand your sentiment here and you are correct but there is only one of me for all of JBDS documentation +Optimistically targetting this back to +But even then is an issue when used by for testing by downstream projects correct +Fixing the fix version +I am not totally sure why the different behavior for this patch +See related issue +With the patch it responds to commands immediately +Thus if someone specifies a final parameter in it is effectively a constant +Reattaching the trunk patch to get it through hudson again +Please review it +Looks good thanks! Can you also attach the file you used +Your definition is invalid so Im closing this as invalid +Thanks ArtemNellyaPatch applied to LOGGING module at repo revision rSlightly modified to include the ASF block comment in the new fileNellya Please can you send us a completed ACQ so that we can continue to use your contribution to logging +You can go on and close the issue sorry for my previous Petar. +That syncing thread was just there in the background messing things up. +Im not sure we can handle this for but lets try anyway +The reason that joining and grouping are different modules is that these are different functionalities +Refactoring the package structure has been in the back of my mind for some time now +Thanks +However vnodes are complex it makes rehashing and rebalancing a challenge +worked fine with this test but failed for another related test +Bulk change to CR to next stream +And you setup a LOG which you never use as well in the same async processor is a bit wrongSee the unit test from You only need to do the complex async stuff in case the policy must do some after work +patch to coasterlucenesolr svn commit m some cleanupSending srcjavaorgapachesolrupdateTransmitting file dataCommitted revision +It would be nice to get these common files shared a little more closely but Ill patch ours +So perhaps the subclass could call the super class after it checks if the conversion is indeed unsafe +I think it is up to the developer of a particular feature to decide whether to use the same JIRA or different +Revision +This is source files +I replaced the with an Exception and it now compilesCan you do the same and check this out for me +Good catches above +security lang auth crypto and rmi packages were addedexclude list statePlz move it to buildtests category +We may even want to consider creating a custom project type for Sling projects +with Teds comments +Regarding Blocker or not +I like but with the p and d gray like +exe is compiled against the Release +Have we reached a consensus on this? I believe Mark is correct regarding the file becoming obsolete once solr moves web its to +then we would have the complete width for every result there is no need to change between a and if its larger than + Also changed the date format to display the date +Maybe this is something I could contribute but it is not yet generic to Jackrabbit and currently has dependencies t our product +USRSSSBPWD will not work with ibm jvms +Test failure is unrelatedIve committed this +Starting with were supporting M but we also have changed our groupId to the M standard +Its model class for JDO and lets keep it simple +would be particularly useful for me +I just want to confirm that it does seem like the above fix works +Please check them in +Im going to assume that when you bring up a node cluster with your setup and these data directories you see the bad ephemeral nodes correct? If so can you try doing it with the latest jar and see if it still happens +Thanks for the comment Karthik +Can you verify and let me know to mark this issue as resolved + maps the string h to the hibernate dialect jdbc driver and default connection string +Jasonits not just for pilot projects +Its not a problem per se just add some work on our side +Verified on SVN Revision . +Heres what I think will work for globs +Both had the same problem +Perhaps it is a good idea to include for current content storage somewhere in the roadmap if the jcr approach proves to be more flexible and easier to use +I will reassign +See my comment on this folder is not writable to prevent accidental releases going to the old repo where Central wont find them anymore +Bulk defer all frontend issues to +Fixed with rev eedfefebbfaae. +Trying to see whether the failures from different Hudson runs will be consistent +Change looks good but better to define this problem away via code sharing right? Also shouldnt the compat test have caught this +I rather think if we have a single phrase that is greater than the frag we should just not highlight that passage at all +reopening so I can modify fix target to +Thanks for the patch Izaak +No functional changes +Reviewed by Robbie +This has been fixed by commitCommit hashecddcfbecebecdeacApart from the UUID we need not log any other details for in error messagesContained in branches masterContained in no tag +Forgot to run verify goal fixes few style warnings +Should an error be +There is a readme in these projects that describes how to run these samplesCan somebody review this and commit please +It may be the case that this is not in need of fixing at this time +The labels for these checkboxes read Recurse into subfolders and Include new files in patch +I did not apply the patch to the file and no changes were made to the branch only the trunk +Were sticking with only +Is there some associated change that needs to be made +Error when removing encrypted content inno the plain content is not removedThat means document is encrypted but the plain text is still in the documentAt leas a major bug +Im guessing that this is mostly a typechecker constraint error +I have confirmed this does not affect preview btw +I think its more useful than trapful if not just as demo codeIn this case maybe contribAnd if we were careful tests a bit could be +So theres no test that covers that the REST API actually does the intended operation? Thought there was a test for that +For some reason the Stax implementation isnt getting picked up when the main class is in a jar +Open in Eclipse attach libraries create user and db execute +Initial patch +Thanks for testing the patch +Fixed in in revision +Found and patched in as well +so exactly how does the skinny war impact serialization? the war doesnt contain the dependencies which is ok +Thanks for taking a look DevarajCan another MR committer review this as well? If its still required that is +Please review and provide feedback +Unsure if this the correct approach as I know very little about the referencing module internals +Currently we have to find how to manage could you elaborate cause its running on my computer without any error for the moment +Fixing a bug that crept into the previous patch to resolve the the patch againJavadoc warning and failures are unrelated to this agree we should have read and readFully as two different methods and callers should call readFully if they want to read all the data +The JSF servlet should not be run in this case + moved testjar testshell to srctestjava but missed updating +Thanks Hari! +Comments in +you dont just reproject a polygon representing a bbox and assume its the bbox for the reprojected geometrytschaub a bbox is special and its an optional propertytschaub if clients want it they know theyre getting a bbox not another arbitrary that spec will be used and Ill add one more assumption up for discussionIf the request specifies to include bounds the bbox of the featurecollection will be calculated based on the returned features +and prevent them from saving it as a new runtime until they correct As for the error not being reset I can also fix that hereAnd as for the wording of the error message Im not crazy about the wording that you propose but I also cannot think of anything better! Selected runtime supports ESB runtime version +Im investigating +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content Services +Theyd like to remove it and put a instead +done +Showed the list moderators how to achieve this removed problem email from mailing user is commenting on the mailing list and the commits list is commenting on the tickets +Ok so assuming client side is identical in all tests it must be something on the server side +fixed in branchSending srcjavacomopensymphonyxworkconfigprovidersTransmitting file dataCommitted revision still need to fix it in xwork head but I dont have the code for it +Moved back to has been releasedMoving all related issues to the next version +When saving the view for the session its a onetime display per alter view +gets the callers class loader and invokes its library location methods +Right +the help of the implementation I have create a patches to support this in HibernateThis is the first of patch files + +release bulk close +Committed +Resolving as dup. +This can hekp all the components to share the same objectbq +Fixed on trunk and the branch +Probably these should instead be different properties +Closing. +Youre right +this patch will just make it is safe to remove the whole cluster config block from if you like and dont drive the newbie crazy when they copy the to another system where there is no eth but bondingD +I think we can close it. +Listener destroy hooked up in rev +Thats why I went with a system property for defining what file to look in +I have attached a patch which fixes the errors as mentioned above +Thanks to Mathias Broekelmann +Patch with the changes described in the previous comment +I just submitted a patch to which sometimes treats incoming strings as double +Closing. +I resolved this issue as duplicate +No need to manage the case where users create tenants with the same name this usecase is hard to manage and makes no sense since tenants are rarely created +Bulk close for release +Configuration files that contain protocols marked as experimental or unsupported do not have schema reference +Sounds like that +added +To be clear the specified directory should be local to the +Actually it keeps going at RF but at RF another exception is probably being raised that kills it +bulk defer of unresolved bugs to +it used jobidtipidtaskid in for temporary storage to avoid any possible conflicts among task outputs +See the Wiki for more info feel free to contribute +unit functionality of a signature algorithm is implemented by provider +Doesnt this already work? Ive used it this way +Can you try that +The basic connector implementation is there and committed to SVN +Sorry Marcus I had no time to look at this but I will by the end of this week +Running full suite of unit tests now +Closing this record since is now released. + +If I remove the properties from the constructor the qualifier still works for my datasource +THIS IS ONLY ON WEBSPHERE all the other servers did not have this problem +Thanks Todd +To complete this we need to write the integration test using the Thrift interface since CDH doesnt package the REST interface +I can fix it but due to other dayjob responsibilities I wont be able to address it for a few days +Deployment side + on this from me Julien +Thanks for clarifying +Thanks Nuwan for the contribution. +I also think the decision was taken in that its going to work on the principle that if its a Map its always treated as a Map and there isnt a good reason for catering for anything else +OK +Thanks for taking a look Julien +bulk defer to is not an XDT issue and has to do with the compiler s ability to provide more or less accurate in the presence of syntax errors +These failures are unrelated to the patch +I think between and weve covered this pretty well. +Sorry Im not seeing any being created +Is there any chance you can try and see if this problem persists? The changes from to are enormous and the branch is the base for our next stable release +Do you think there is a chance to have a official maintenance release? Or should I wait for? + for the patch +CVS comment for fixCatch all exceptions thrown when processing stylesheet and send them to the error listener as a fatel error +I agree +Thanks for confirmation Roger! +Whether the generated code is correct or incorrect is definitely an issue for them but this bug report is mostly about it being different which its not. +Verified +There seems be enough community support for this feature +Nice work Robert! Im on this patchOne concern is the guessability of but if users are really concerned about that they can always generate their ownAnother interesting algorithm +Please feel free to reopen if you still see this problem with new builds and provide instructions how we can reproduce +In order to insure that a sequence number is not given outtwice a synchonous commit is done for each allocation where it was asynchronous in +I can upload the website again for you if you want but let us wait for the guys to reply +Oh my configuration fault +It adds a new binary sstablescrub that takes as argument a keyspace name and column family name and scrub the relevant sstables and does this without breaking if the sstable have some out of order keys +Mubarak Our comments crossed +Looks like this issue is open for a long time +Moved to +What I described was v +This is the SVG I used +The suggested solution doesnt take into account the fact that all changes to the JAVAOPTS environment variable should be handled within the file +You create there two new folders background and icons and put the images there +The docs should probably be updated to not that if using address literals for specifying a preferredMaster hostport combination then rfc should be followed +Sounds good to me! Thanks Werner +Please go ahead! +Could you clarify how relates to the truncation of stack traces? Are you using a variable to store the stack trace? And is this truncated? Or do you simply want to increase the column size in the database so that larger strings arent truncated +Im also very much opposed to a release forcing this sort of change +My mistake assigned back to mistake assigned back to me +The solution is based on the work in. +When I began this ticket I was under the impression that that was a sure thing +Thank you Alessandro. +This file crashes FOP with the second foinline +I ran a test to check that an Avro data file produced using Python with the Snappy codec could be read from Java and vice versaIve also attached as an example of a Avro data file which might be worth checking into sharetestdata +Thanks HyunsikI just upload the patch for the issue + +But we have run the same manual tests with the patch and the errors disappeared +I will wait a few days for comments and then mark the issue as resolved +When doing as above by disabling JMX for new routes then the seda thread pool is although still enlisted in JMX and thus continues to does the customer use seda endpoints in their application or did you just use them when creating this test application + +Looks good to me +checkstyle file created with corporate checkstyle configurationCheckstyle project nature and build command added to + +Your patch introduces what I believe to be correct behavior so Ill it +Did you determine if this was an HDFS bug +We should get the project example plugins into the build good added value +Please check that the bug is fixed for you +It seems that there is some progress on itThe exception is not thrown anymore with that +Thanks Brent +Probably just a matter of adding nls somewhere in the grammar right? +Lessens my work +if SYSLOG is not configured print the date and pid prefix +In addition to this I propose making the minimum size configurable and also disabling the array growthshrinkage an option for users with big name space in their cluster +? and it reports a new test hangingHanging test Running +bq +Max VictorSnjezana is this fix important and correctYes this is important and correct fix for and sure about this certain issue +push to +Chromattic APT need the project do not use the APT at compile time +Fixed by changing instances of into in Git ID deferred write capability to jspx files +Thanks Corinne! +First of all passivity is very important +Closing all resolved tickets from or older +I will post a patch and a review once I can fix that issue +It would be ideal to have hadoop jars decoupled from +We havent received any information from the originator of this defect about the defects nature so Im going to return it +Hey MarkSorry for the accidental spike +Martin if youd like to verify and close Ill leave it up to you +So no usage possible with commercial software +Thanks Hemanth! +Thanks +source cut f sort grep jar wc lNow to investigate how to use it +My code works similar to how I would build the Criteria but is not typesafe is harder to read and is more +The issue is documented and if it arises can be fixed manually +I still have failing which makes me think is still not reliable +Batik import from BZ complete +Patch +Please verify and close +ha +I am not able to reproduce this issue +Youll just have to live with the duplication +Dupe of +This is caused by a wrong path separator in the genpom script that restricts it to POSIX platforms +I have attached the two other failure reports + +Closing all resolved tickets from or older +Backported to Thanks Hoss! +The wording of do whatever dirty work is needed is from the dawn of time if the PMC wants to relax their responsibility they will need to vote. +OlegI talk to our admin people +I applied to branch +PDF with the has been fixed in the head branch +This issue has been fixed and released as part of release + Adds a test case to that demonstrates the problem Provides a solution in by introducing QUOTESKIP Simplifies the method The patch provided will solve the specific case of within string literals it does not handle cases like this SELECT foodooxxx FROM DUAL ESCAPE as this would require real SQL parsing. +It doesnt fit very well into our template but should get the message across +The new patch is for only +The command should be mvndependencies +Martin tested it on and it worked fine +I was using accidentally +I will review the patch soon +its very critical. +Hence changed to the Fixed state. +Note that I have repackaged the bindingservice as a exploded Scott. +Much appreciated guys! +Thank you for the patch Xuefu! I have committed it to trunk +i have tested with gzip compression as well it is still working as expectedtest without a webserver in add more info and answer questions before reopen ticket +Could you link the new Jira issue to this one so I can track it +Do you have an URL to that spec? I see posts from usually knowledgeable people saying that having first in a jar file is not required by the spec so having that would be good +With Marks comments I updated patch withThe part of the message saying To change fileType if want output compressed is a little unclear +to use get and expand the support to cover all databases +The new address will be shown as selected in drop down for Address in the same area and the shipping destination address will be replaced with this new address in Contact Information area on the same page +I just committed this. +maybe use other images or no images at all +This log information will be displayed for each package folder the class is traversingRemove the log information from package folder level put it in more high level if the necessary classes are not found in the whole package hierarchy +Hi Mike Im going to close this ticket which duplicates + +Ill try it out and add the results to. +But itll be better if we can support this at least for a limited amount of time +Resubmitting is ready for are no tests included because it is just a configuration optionPatch is ready for review. +Do you have an example how to use it +Appears fixed in and +In any case there is nothing special that Spring is doing about this +As you stated this does not fix the complete issue outlined above which is why I do not submit this as a resolution to the above issue and made a TODO to support a more thorough implementation +Added some descriptive test about jcr install integration testing to the file of the launchpadtesting module in Rev +What Im after here is something on record as to whether we expect the anecdotal evidence of faster indexing by increasing batch size to be affected by streaming making changing the batch size question less relevantI suppose with JoelsMarks changes to much of the question goes away anyway and this JIRA will then primarily affect the leader forwarding the updates to the followersOf course all bets are off if a user directly uses SUSS +Anyway the reason is a problem in our build script that we have to fix +The options shown in bold are unclickable +Sorry! The new patch I attached contains the right version +It will launch several threads in parallel +Thanks for the details and the solution. +this task since I think we need to make a decision about the various options that have been discussed for generifying query +Select the shape layer in the layer view +These changes are now merged in +Yeah try with relative paths until I get comitted +Its up to you +This patch corrects the issue by separating the invalid cardinality check from the header scanningThe headers are now scanned properties appropriately allocated and then the invalid cardinality thrownFingers crossed this is appropriate and the patch format is correct etc +This probably fixes also +the is the patch to use to add servlet to the jee spec projectthis patch contains the and files as well as a verbatim copy of the xsd from the servlet spec pdf +So well generate source plugins for these bundles only and well manage manually the source feature it will be easier to maintain a source feature manually rather than relying on the plugin and trying to introduce exclusions in it +I also try to avoid chaining whenever possible so I dont find it a particularly useful addition and is just more behavior that might impact overall usability for a plugin that is supposed to reduce complexity. +Attached andIt seems our current implementation is renaming the duplicate element so the first one always hides the second one so I think we are complying to the spec +Any opinion? Thanks +Maciej This is a good point +The changes are in trunk and codeline +Please keep in mind that Rotislavs dependency on artifact only is the way people is expected to set a dependency on the client +As part of this I had to call an instance of +Ithink we need to make and Fields more flexible +Fixing this would require a redesign of eviction in the branch +So who can we talk to about + +Use asp per my JIRA account +Hi JoseI thought the patch fixed the bug +hi Mike that means Flume will execute HDFS flushfsync at each Sink transaction +bq +As you can see from the output the size of the pool is increased in steps of +I believe that this is a dupe of another issue +It is an easy fix either wayThis patch adds three generated files to the directory in the thrift package +The is not supposed to open with the graphical process designer +It is true that administrators are hard to tell which servlets should be excluded in the filter mapping + +meta file +Tweaked code readability +I will check the tags quickly but for the trunk it should be save to use it without the problem +Thoughts +Bulk close for release +Solution implemented in trunk +improve log message +Attach away and Ill go from there +You are good to go now +OK +I see a branch for CXF fixes and the fix was applied to trunk as well as the branch for CXF +Commit the translation patch for rev thanks +or if someone wants to update this to work go for it +It just means the build needs to get updated so that each modules javac output has its own directory so that a proper jar for each can get created +We want to get rid of all of these cases where exceptions are swallowed good catch +Its almost like what Id expect to see from version negotiation but its a little too slow to finish +Also in your opinion has what youve done already handled the basic problem +Can you please let me know if the license is compatible and whether I can proceed on thisCheersAshwin + for local inserts might be worth it now that abm is the default and we went even futher with +Could not reproduce +Sounds like a great location for it +During replaying the scorer is only set once +But after doing so geronimo would not start the wab automatically without the osgi persistence files +Shouldnt it be the case that if a serious exception is thrown the process should die +Just attached a file containing the fixed target already to be fixed in trunkProbably Tom has taken this into his refactoring. +unit tests passed +In the lastest patch there is one more problem +Fixed in trunk for all sites +canceling patch to fix hudson patch again +Thanks Gilles Thiru! +Synapse will be using the latest rampart release +Patch submitted +Its not very clear what form such an outbound adapter would take +Also added update pkey sql to reference example for file manager dynschema +I tried to specify changingtrue for dependency as well as checkmodifiedtrue changingPattern +META +Junit patch committed Revision Thanks John +Ill try and review the code as soon as possible +The masters REST load would add the master fielding HTTP redirects a minor imposition +Strange should work fine +The test doesnt show regression +I agree with Gordons proposal +Fixed +I am getting the same issue when trying to build the commons module on my local machine +makes sense to meCommitted revision +Is it good now or should I wait + you try setting showfalse +Its not that its hard coded its that it is a relative path +Im going to work on this +This is moot as keys was removed in. +So certainly the new results are wrong and needs investigation +Made strategy pluggable allowing plugging in a for consistency with Spring MVC +SureTo clarify a bit first when reducers put mappers in the penalty box theyre saying that this mapper is being slow so it should prioritize receiving from other mappers instead +closed all issues which got released with +I have already posted some comments on the merge vote +So as long as we go through that we should be safeIf we make the assumption that classloaders never pick up new classes thats true +I tried to attach the images but in my tests they were not appearing +Will close the jira once the system test change is done +Fixed. +There would be another method for identifying what nodes are currently compacting +Ok thanksAny chance that the PMD team resolve this bug soonShould I open an issue there + +the attached is a fix to the attached is a patch to the patches in r +I have added this code for all Java projects +see +This patch improves the failures reporting +bulk close of resolved issues for release. +Fixed in Drools the change only on Jan so wed need a later build of Drools to pick this up +See for an example +Thanks for implementing this Luke. +Dart has just told me you can right click on the error to fix it +Good thing we do testing +No need for IT +I will start iterating next week I guess to mike if you feel like it go ahead start iterating +This file is a dummy subclass of the abstract class to be tested +Closing on +Patch you need to get the whole project from svn and then build it with maven it will work +Attached patch to add a primary key to this table +Could you please generate a new patch +Please try this patch thank you +The list simply cant be cached as the icon might change with every request and the logic for displaying the readunread icon needs to execute on every request +If you say no to you handle commits +jQuery has thousands of plugins that i can use them easily without much javascript cant leave it +patch appliedThanks +So you okay with the approach Lars +So why not just and +Not enough requirements gathered +Ok. +Issued fixed in Rev you would not require jars in classpath for non usecases as provided scope dependencies are not for the clarification! +LucThank you for all clarificationsIve read both agreements and decided to sign ICLA +Bumped up priority to Blocker +Thanks again to you and Bryan Ill plan to commit the second patch tomorrow +Sorry for the noise. +Kieth asked me to append these from a problem we had which relates to the configuration just ran into a problem using Castor under jboss where there is asecond file insid the file +So i think we should keep this the same +Hi AndrewThanks for reviewing the code +Uploading patches for both MR and MR +Hope that this will be included in some of the next Ibatis Releases. +Attachment has been added with description Extra information requested in Comment size of the attached file is too large to undertake a fix please whittle it down to a minimal FO file that demonstrates the problem in minimal form without redundant P open bugs to P pending further review +Committed in r +Can you change that? Maybe you need to reattach the fileThen its ready for inclusion in the documentation we will ship with +Also the should be initialized just like within the +Ahh Im not running java so that might be the difference +It is meant to save people time not cause them hours of frustration tracking down a known problem in a jar file +To see teh effect use a with an explicit groupName or extend the postprocessor or the to dynamically determine a group name +How to solve this just wanted to note I have the same issue appearing in Cordova with a Galaxy Nexus on AndroidI commented on the linked Android issue and tried to report it to Googles Mobile Help Forums as wellIn the meantime people not using Cordova seem to be working around this issue by trying to just retrieve the last video file they can find in the expected video directory or something +is my suggested solution to the problem +I also had found a need for the +Heres a modified patch that demonstrates the approach +I send you a new update for forums +Gavin created the requested redirects +Yep implementation is wrong +Need some more investigation though +The class is responsible for constructing the object +Dave I have of the code for the and handler but just dont see how to test it +This appears to be a bug in IBMs and notin XSLTC +Looks like there is no Start Document or End Document event when is read via Axiom +Plugin prepared for out into production +Editorial pass over all release notes prior to publication of +Closing tickets +Missed the +The most of us working on hbase are trying to make it an hardcore production worthy platform +Fixed with r +I have the exclusion file configured in the in my +Hum +updated Fixed For field +First stab at this +Will commit shortly + +Looks like we have too but with the docstring above the maps +Fix applied in bcdfacdbaaafdb on master +Is it empty +Thanks Amar! +I create distinct String objects +Please give me priority of migrating these jobs from EAP to AS with +Roshan no problem! By mapping I mean that the algorithm you describe above requires a queue data structure to be persisted in order to maintain ordering state +It is good to have the test +Could you correct that +Present in cant reproduce it at BRMS I suppose it is I do reproduce the issue by Drools standalone which downloaded from site but the issue seems not occurring at BRMS anymore +workaround Vitali noticed you were not watching this issue so I guess you havent seen Justins comment +It looks like the same issue +Edwin wasnt a committer there until we collapsed the permissions in December so we dropped the ball on this one +I dont think we gain much by an XSD +In what scenarios do you see adding and removing cookies as being an concern +Moved patch to +Screenshot of the Source Folder Selection dialog that is opened when clicking the Browse button near Source folder in the New Java Class wizard +properties header never workedin then lets not keep it in Configuration as well +It uses a to find the LRU item and passes unit testsHi Michael Can you test its performance +I just created bug for the tracking of the LDAP persistence discussion of this topic should be moved to that bug report +Diff for changing input type for enable lzo +EJB team wont be looking at those for Beta +Basically in the end the requires a with sorted keys and null values +Custom resolvers are coming to which is in active development which will address this concern +Also seeing this failure in +Else please feel free to it will not be fixed +Comitted. +On second thought in a workload where rows are infrequently updated it could be a win to use put +and now contain type check fix as well +Im a newby +we will deploy the metadata together with the app +Applied with modifications has the patch contained option to include for! +you wanna take this one and commit it Jan? +Thanks +Added a patch that seems to fix the parsing issue +Whether the infrastructure will support the contents of these semantics will determine am Currently testing +Ive made a smaller change as I didnt want to introduce a new property on the base class in a point releaseI understand +dll +Actually there are more annotations that I thought that dont define a default definitely a bug on their side +Ill try and move the blocking tickets forwardThanks for being on top of this Anupam +Patch generated using git should be svn compatiblepatch p i +I tested via running all the example programs and all ran fine +Ive just committed this +Updated patch now that is in +Still debugging + +Im not sure whether I understand it right that problem with is probably fixed +If thats the case than it would be better to make this behavior optional +Verified thanks a lot! +We currently use majority quorums No the example is fine +Thanks again +Can you update these patches? Thanks +FixedNow single mouse click on site menu to change the state of the right panel is enabled +Can you include the schema of the table you are trying to serve +Looks like were really exposed in several ways +mrepositoryorgspringmodules and into this directory I unzipped the source files so that I have the following for example +In my case I already have the plain file so this requires me to gzip it and then have it gunzipped by tdbloader + +Attaching a PDF of the document +Linked to this because relevantTutorial submited in this Issue +Attached patch premigrates the old snapshot dir in tarball if version is found +The heterogenous data structure is not intuitive code and performs poorly compared to the revised one as put in here +This is a left over from an initial attempt to refactor the classes +We can start without the DSL and then add that a bit later +I think I agree that we cant fully fix this absent API changes that we should not make at this late date +Heres the relevant part of syslog for a node where Tomcat is killed by the OS +For those people their are configuration filesIf someone seconds the I like LDAP idea I will start working on it and submit a patch for review +this should really be discussed on the users mailing went thru a dedicated QA effort and we solved many bugs so it should really be better than any in all for changing the type name the user interface does not provide any way to change it becuse you should not change it +Just commit it this is a much better check than what is there nowAnd hoist the constant See and for more details +Second version of everything +It surely be a minimal configuration but give it a shot +The Schemas Location is not encoded correctly Found this after disabling the order validationCurrent Conclusions Plan The parser appears to go into an infinite loop and although this appears to be a result of the multiple polygons this may have to do with aquiring the document schema +Any update? Moving out to based on Larss suggestion +But of course we can leave this issue open to indicate that were still discussing this +This is a very old jira +so until the day this patch will not work +No only this one +This seems like a problem of implementations not having native AC support +Mass closing all issues that have been in resolved state for months or more without any feedback or update +OK sounds good +Once I did that my appender worked as expected when used from the appender java project +Heres the same patch relative to the Ivy root folder since it wasnt clear what file it was forChange commentAdd support for POM properties groupId and artifactId +Changing global OS settings seems like an odd thing for an RPM to do +I just committed this +Also since we have the table disabled these operations are chainable enabling the admin to chose preferred behavior if we have a restore that doesnt take a snapshot +may be providing a stack trace here could help +Logged In YES useridUnfortunately I dont think theres anything you can do in to fix this +However if Struts provided a way to pass information between two actions it would be kept serverside and hiddenform attributes would not be necessary to get the data over the gapWhile I understand this is not a bug with the existing framework I feel it would be a useful enhancement to enableencourage the use of serverside only parameters within the construct of the frameworkI have changed this to Improvement Minor but I would be happy to hear any reason this cantshouldnt be done to keep the form data serverside +Since there isnt a unique namespace node for every namespace its impossible implement this properly +I havent thought through the ramifications about allowing this but right now I cant see why lifting this restriction would cause any problems internally the select list is rewritten to a SELECT TRUE FROM +I walked a friend through the same steps I did and he wasnt able to reproduce +Your right Glen I had to pull the containing VM out of ro mode earlier on which was the cause of most builds failingOlivier has highlighted a new issue we found between jenkins and sonar but at least the builds will run again +xml mapping files to the console configurationWell sure that is doable but how many users are expecting to use thatnormally your list those OR you have automatic scanningAdd mapping files is deliberately not exposed here. +Not until the split shard patch goes in +The build process is proving to be a little more complicated because the driver was developed using the unixODBC framework which cannot be committed into the apache repo but Ill see if I can include it as an attachment to this JIRA +I am not going to let this sit since its just more of the same and there were no interesting decisions +META +Resolved on Simo Ive address the issues that you raised in the comments. +This is great work! I will commit it as soon as I verify all the unit tests still pass +components like that can be rendered before the dataview and they need just the size. +Mel + +This would mean that they are actually compatible and need not extend +When does the error happen? When creating the temp file? I dont have access to Windows so I cant reproduce a workaround I could change the code so that if the temp file cannot be created then the plugin still tries to run like it did before the security manager config was added. +Whats that mean? Looks like is same as we currently have +targetplugins +Actually there is a discussion going on right now which bugs should be ported to the branch for an upcoming against branch +Patch committed to branch +I opened and to track userguide benchmark issues +RonI was not really following the efforts of implementing Continuations for jms +We have since done some tidying up in that space it might pay to rework and apply now +This got fixed as part of when we changed it to disallow unresolvable type variables +theres also a completely new locking mechanism. +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +Patch looks just committed this +I meant disabling the feature +in is missing this jira +In other words if you have special needs they will have to be expressed in one way or the lets first see what you are really trying to do +Although it doesnt seem like it issues do get triaged there and escalated if need be +bq +Ive just committed this. +Here the patch so only extends Assert without importing extra crap +I understand the purpose of the second change +Thoughts +Thanks for the response +The input document is not relevant in this case. +oh ok so Ive missed a change I remember something like no rows in acl means everyone has accessso I think is good with global adminIll test it and send the patch with the checkThanks +Its simply a matter of using the tool against the version of tomhawk +EricPatches are totally welcome on this +Why is this a bug? is reporting exactly what it says on the code in the resource adaptor has created a session but not closed it +Always good to have you around WolfgangThere was a bug when using the top level accumulate CE with a single function +differs mach from previous version and it is not supported now +Never seen this error before +Made this critical its critical regards our write performance +This bug is caused by the correction of bug will remove the synchronized keyword on the problematic method as the side effects are desastrousAs a consequence should In the meantime maybe people should use the workaround described in ie add the synchronized keyword for all methods annotated with Unwrap +Backward compatibility for config and is not guaranteed until +Richard I added a couple comments on reviewboard +attached +However the unit tests dont seem to test any hibernate so from what i can tell it wont try to the entities +for the patch + to dennis the method header comments +I sent new PR with display separated by Disabled user and user +Sorry Dominik David did not realize that you will need to verify this before you close it +Works ok under dont get the mozilla Bug association here +bulk move of all unresolved issues from to +Maybe this could be changedimplemented in the futureNext time Ill try the mailing list +its not used as a thirdparty dependency +prior to generating release notes +Patch applied. +Thanks Jakob. +Yes these are useful features that should be +Except for all of the other eclipse embedded tools where the me integration is not available +Im curious if I can run the test right now after updating GF with is. +The close and the prompt to save workflow is good for now. +Jiri Im trying to repro this so we can get it taken care of +Also standard commandButton should be working please create small demo example reproducing the problem and attach it to the issue +All the subtasks were fixed +I committed this to trunk +if someone does such a thing i feel the maven artifacts are unreleasable +Thanks Ivan +I cannot think of anything else that should be done with it +You cant ask the network topology about its rack count as that will change as it resolves more addresses +The orginal request came from a time when was pollable and the use case I was trying to support then is probably covered by existing features subsequently added +If the progress tracking API is missing anything will become apparent when people start using it +Fixed in subversion repository as of rThanks for the report and for the patch. +Please let me know +Ill close off this issue and we can raise an issue specifically about defaults since we should really think about defaults for and perhaps at the same time +Adding another dump with thread closing the and then joining thread which is inread +Patch integrated to trunk. +I have some idea what was causing it and will dig into that now +They can be shut down but they cant abort themselves gracefully +sig files are generated inadditionto the +After running a number of tests on Mac Linux Windows Ive noticed no problems without +Added unit tests to validate it +Seems like another issue should be filed to implement different sampling so that we dont this one too much +Ill fix smoke to check for this +I am missing the libraries so all the TEST libraries cannot be built +Hi Knostantin this should be the bug in as wellPlease check if its worth to put in that version as well +Jakob I dont see the relationship between this patch and comments +I change it to improvement had a look at that issue recently and made a trivial change to the grammar to allow that syntax but it created various grammar ambiguities so I didnt pursue +Thank you Torsten +Needs thorough testing I dont want this fix for a small edge case to destabilize the code for the main use cases +The JVM and implementations are not good enough +rev to generating release notes +I will take a look at the implementation and commit tonight if it looks good +En les eles geminades preferiblement susa el punt que al teclat espanyol est damunt el nmero no s si a langls tamb hi s +The same type of analogue can be drawn between and Kerberos +Please use instead of +Alls fine when Im in GERONIMOHOME directory and although the FNFEs thrown the applications deployed +I would guess the is thrown because youre passing the wrong argument type +Unfortunately we really have a mismatch here as Felix wrote Component is the marker only classes with this annotation should be processed however when the super classes of a component are processed its not checked whether the class have a Component annotation and Service Property and Reference are always processed +Assigning to headius in case I dont get to it +Thanks for the review Tom +In most situations the header will be that of a received message and therefore have a CAB +This causes a stack overflow error since the client proxy delegates the call again to the outer decorator +Patch contains desired changes +Does the per topic bean only count messages produced? Would it make sense to have a and a bean separatelyOr maybe Im missing why only having is useful +Is it too late to hide it again? Is this really an API we want to support? If there are individual methods that we wish to expose perhaps they can instead be exposed through or +ant passed on my local machine for this Y! patch +Reopened per Michael Lipps request +For this why dont you use just StringindexOf +The reason is that Messaging recommends you not set both the scheduled delivery property and the group ID +Caches configurable count of latest version +Im attaching a diff and jar of the combined modifications for and this issue since there are many files modified in both and I did not succeed in producing a diff for the changes alone +I encountered this issue as well +Workarounds Use the DOM filter +Set the Derby lock timeout internally to infinite +Im sorry ignore this +Actually function types cannot have constraints on them they can only have guards +Thanks for reporting the problem +But then will not be interoperable with +Thanks joes much appreciated +which is probably better than wiping out all the metadata associated to same named tables in different have resurrect the patches and Im working on integrating them on the code +logs the size for each message set +Sorry Luke +Bulk close for +In we use specific classes and we check javamail results for specific instances +Those methods have the ugly and these others are more specifically named +Duplicate of for creating dupe. +I will have a look this weekend +Test and log file +Set JAVAHOME to the directory of your local JDK to avoid this messagechmod cannot access +Thanks Todd +Added check of iterating a Scan too. +New interface to find all groups find users in group delete group rename group create create add user to group remove user from group +Note I moved sorting code back to where it was before +Mailing lists dont actually exist +Mark This issue was fixed in the trunk already +It is still source level integration +This issue is closed now +From RyanRegarding I found another case today that could cause the error to happen +your patch is in rev +I suppose I could use an instance check as I did on the but keeping them in separate maps kept the intention very clear +Please help me review and apply it +Patch integrated to trunk. +Added documentation for the JCR connector to the Reference Guide added the new JCR connector project to the parent POM and committed all of the code into trunk +Wont make it to but should be in place for the above commit provides as script stub +unittests +Remaining issues with HSQLDB have been resolved so Im closing this ticketr +After some more discussions with Stack heres a fifth solution that also includes some of his comments A Map is kept to keep the scope for each family contained in the WAL edit +We report the hashcode of the current instance in cases where we cant get a real pid +However onclick on input Button is supported +Maybe Wicket Examples site uses Wicket now +Seems like the issue may be because we are opening the FIS and not closing it +I thought I would start with this one exception that I really needed now and then let it sit a few weeks to make sure there are not better implementation ideas out there +As Lucene is coming I backported to branch in revision after release. +The bridge is marked started in JMX but obviously no messages will be passed on when the there is no receiver thread listening +grails test +I got results +we have rows with around columns and around k row size +Oh nice diagram by the +This skin has been committed +You only ran the tests on a few +HI have a hsql jar in my local repository +Mikhail will be uploading the patch separately +I think we are overthinking this +Committed the fix in revision +lmeadows lmeadowscould someone patch the svn access file so that Larry can commitTIA Ted +I would think given an index on column name something like filter rows by eq foo would work but while I see get called set never does +Ill try to recreate and then fix it. +As this functionality not available OOTB we recently give this to our client by writing the same code in custom componentThanks for your concernRishi RishiYes please do you Jacques for green signal shortly new patch will be available +Looks good to me +Indeed thanks for pointing this out +we need concrete examples of how urls were before how they broke in +Attachment has been added with description has been added with description fo has been added with description font has been added with description metric in FOP and probably earlier transition resolvedfixed bugs to closedfixed +We use something like this at Salesforce to track certain slow SQL queries back to the part of the application that issued themFor this wed log some module identifier here +It is in my companys repository which is declared in the as release snapshot plugin repositoryCan I anything? X does not give me a hint of the detector being loaded +revupdate resource inventory links to go to appropriate containing group page with correct params +Dmitri I miss verification steps here +Hi RichardCan you check if its the right log? Because I cannot find anything regarding vpn in the logYou can tail f varlogcloudmanagement to see the log in real time if there is a error in UI it would be show in log as well +will recreate the problem with latest AS master source +what change did you make to the model to enable this +We need two and +I thought this had to do with connections connected to other connections +this doesnt address the skew case +The code is hosted at gitnoahdesu hacephas always please test and provide feedback +Committed this new code +In fact I encountered the issue precisely because my app deploys its own copy of +context and the one in +Committed to as well +The unit tests ran just fine for me even before this change +keeping as ID of the migration to identify apply orderI dont see why youd want to complicate things by doing this +As is there is no context for what and why for a Bean Service Camel Service Gateway Binding of an SCA primer to start and how Switchyard implements that standard would be great +Hi JoeDo you still want to follow this issue up? If so we will need a testcase thatdemonstrates the issue as Joe requestedRegards Simon +Standalone html with perf deserialization process was found under profiling to be heavily impacted by the creation of Text nodes +Marking as closed. +closed all issues which got released with +Which is what I suggested it was on October +heres a patch with the test case added to in revision . +Here is a zipped archive which contains logs for both brokers concerning this issue +Here is the test included as a patch +Which then raises the question How difficult to create and maintain long term +Bump +Let me close this one and work on. +I dont see what that will achieve +Leads to +It shouldnt suppose anything about jsf or web containers +Thanks John! + +The code at present is very rough and needs a bit of work and more testing before I would consider it for a patch +I believe we only have it as a dependency anyway the container provides the actual libraries +Hoss The concurrency issue in can indeed be solved by making the reload method synchonized +In the first place it is a bug of Add JSF Capability wizard +The problem in CXF was between a plugin dependency that was in the modules list whose parent added the dependency to the plugin +okey connection profile to have a jar driver is apparently done via runtime but this does not actually connect to the running database in the server afaik +We could eliminate possible conflicts by appending clients machine name as well +No response assuming ok. +When the task is done an event is sent using the gateway to the global +is out again +The link is correct for beta cql in which is states it is at the top +Closing all resolved tickets from or older +Was this found by a script Brett +We check the plugin feature and update site configuration files into CVS somewhere +Finally I have run the full suite against both sets of jars + +It will help the UI development drastically especially that now this tool supports +Was this issue fixedRegardsVyas status says Open probably means its not fixed +Please dont add any estimates +By implementing this feature via special paths I worry that user code will begin to directly use these paths instead of only allowing the client to use the featureCurrently I plan to keep this as private and unstable feature and not make it part of the public API +It can even be private it will still say you write a setter method for it and make the field it looks like the only way to set the field is via the setter and that its logic cannot be because of the annotation can find the field and set it via reflection which is such a case Id rather keep the field public because it clearly is part of the s why we probably need to exclude public fields with annotations +I posted a message on this to the mailing on October thThis is not in the way of an RC for. +This is by design +Looks good to me commit it. +I think we can add new maven artifact java and add the jar inside the project +Wouldnt it come down to one method call per input byte +Applied with revision +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +Attaching patch which changes title to name and adds two more books to the json patch which renames title to name and adds two more for this to be my second commit to Lucene What do you think +Thanks for the example it has been helpful. +As mentioned cannot be reproduced and moreover I believe that this could be caused by examining process instance object that could be out of date since with persistence it delivers disconnected snapshot of the data and at the end of the process when you try to get process instance from the session it should return nullIf there is a reproducer please reopen. +Situation is hopefully improved with +These system procedures can be called through ij +Regression tests added +javadoc +news? I am tempted to close this JIRA as I cannot replicate but would ideally like to hear your views before doing so. +published +What does it support for internal tablesCan we get a little more details on what needs to be done here +The latest patch looks good to me will check it in +Ported from trunk to branch at subversion revision . +Yes Projection Search works fine +And none of the browsers I used to test can render it +is a patch with the required fix for nhttp transport +Removing old and old +The cryptic name is on purpose so it wont conflict with normal code +No I did not +Ive made the Consumer use an for the closed flag which reduces the contention for the lock +Mike this isnt complete yet +im sorry i didnt have a test env to retry this week +Moreover now hadoop still do not have that feature which may cost a very long time to implement +According to good coding practice all user AWTSWING code which perform painting component modification etc +I wonder if its better to or to stop here +I committed this to trunk branches and +Thanks +Do we need an UPGRADE doc file in the trunk and release explaining where it is +thank you for your inputdid not plan on more at this point just developing the structureas far as committer I think there are a fewthat would support this effortand I was planning to ask to become a committer finallyI think having access to the branch is enough for nowbut you notice I am reading commits and will working into doing reviews +Patch committed in r. +Easy fix will update accordingly. +This option is indeed missing in the current version of the sonar hudson plugin and a ticket exist to add the functionality has been done already and it will go to the next release of the the time being if you cannot wait you can create a yourself and use the false +Youre right not to go crazy with refactoring yetI think updatelocaldocs needs the privileged db handle you should just take the localdocsbtree from the db it returns and put that into the original handle passed into initsecurity +not blocking this few times but I did not find any notable difference on my box +. +No response assuming ok. + +What I meant by is what the Reference Guide just calls an Expression +Updated patch for +But in case of block deletion due to file deletion etc +Pluggability does provide an option for with I feel that adding a field to hrpc header mentioned in my previous comment allows Hadoop RPC library to be used by other projects with a different serialization I understood this to be one of Dougs motivation +Thanks for your help +It seems to require a very small increment over Vijays recent fixes to resolve this +Im going to close it unless someone is positive they can replicate it +but I think it wont matter if we get the updated guide out with its new examples +Apache httpds moudlescache may be a good pointer for this +Using the patched from in jboss allows the tests to pass +May be similar options for the cli +bq +git diff patch +I tried it using both and the latest and I dont see any scattered whitespace +I cannot reproduce this with latest EDG alpha build +Believed to be fixed in definitely fixed in and later. +and fails on my machine even when the patch for is not test also fails without this patch +I would like to commit this change without generating patch +Here it goesTwo vm properties are added andInitially I had an idea of adding one more switch which includes turning them both on but I couldnt think of the right name for such a property +Quite possibly +Most likely the users policy file wont either +New patch with no javadoc fixes +Thats odd we havent upgraded since this was reported wonder what changed +We talked about overridingdisabling page parameters this issue is related to that +Logged In YES useridThe behavior is actually correct. +I have seem similar errors when the input is a mapfile but i am using +I misinterpreted the comment in to indicate that this might be the root cause of the failure +it is missing a change without which the exception still happens +In the new api the loop is handled in user code and therefore cant require the extra call to update progress +Downside the application has to implement ordering callback would pass the message to the app +I dont get MSWord Viewer to reset the top margin to after it has once been set to a value +Id like to see if there are any endpoint resolution errors reported +ok +Committed in a number of steps over the last two weeks +Wont be fixed +Ive removed it +Id like to commit this in the next few days unless there are any objections +I ran a new test to collect more information +So far so goodWe also provide a runnable Jackrabbit Standalone jar binary that combines Jackrabbit and all the dependencies to a single jar file that can be run with no other dependencies than a JRE +It is even more important to have in since the module support needs more links processed through mappings +for the patch +havedefined it consists of all the queue related information required for scheduling decisionSome refactoring has been done in terms of moving out some of the inner classes from esp the andI ll be uploading patch with testcases +The fix will not go into EAP CP +Adding these parameters to the logj properties file will allow the to automatically do some cleanup on the log file +Therefor it would be good if you could show as how you would do that in plain jdbc java +is rebased to trunk addresses Ashutosh review comments +I think we could add a new preference page called SharedCommon Bookmarks child of the LDAP Browser preference page where the user would be able to addeditremove these bookmarksIn the Browser view we could add another sort preference to group shared bookmarks together and appear before or after bookmarks. +Id not want to say without trying something out and measuring what we can get +Thanks Sharad! +reassign it to Gregory as a VM guru +Submitted to trunk +Wed need to see how this component would fit in the distributed flowJoel +I was just chasing a bug similar to before and concluded was the problem +But is that really a priorityAs with all optimizations the performance improvement should be measured and demonstrated to be significant +If you wanted to to go to the bother of modifying Activity you could change that topic to tell people to check their edited program against +Theres so many listed above +Attaching First attempt to reduce build timeNow Ivyresolve gets called only once for all configurationsThis is slower the first time but much faster afterwardsIt cuts my build time by halfComments +Andrew do you think we get a first version ready in time for Spring IDE due end of November +When is that scheduled to be released? Id like to use and above to benefit from the port change detecting mechanism +The same problem occurs withmodifying Summary so that this shows up in the Patch queueThis also applies to CVS HEAD as far as I can see +Could you possibly maybe amend the test case so that a r is not used +It probably wont have a measurable impact +Bulk close of old resolved issues. +can render any TYPE so thats why I have proposed +closed for final release. +All existing unit tests pass +It would really be needed in our case It is not always possible to sign jars at build time +Applied thanks Thomas. +Also it would be good to make service name more specific something like upload or upload +fixed this issue +and so on +But there are a lot of changes that have to happen to get there. +Okay +Your analysis is wrong this has nothing to do with the retry you are saying is that the connection is successfully by the reconnect logic but that this connection cannot be trusted +Still might want to add flushcompact after replay logic from v tho +hopefully we can get to this piece of work in the near future. +It used what should have been a path argument to decide if newlines should be added +Thanks Brian for your contribution +Failed test has lot of errors +Hi Naokithanks for raising this could you please verify the fix checked inthanks Malcolm MalcolmThanks for your quick work! +Included test case to check the report of such blocks +Was there a reason for this or was it just an oversight +Just a proof of concept dont ding me over file locations and suchIt turned out to be better to have separate annotations for each of the possible labels IDE integration was much better and the values show up more clearly in the Javadoc and IDE integreated Javadoc viewer +Ive just committed this to trunk + to trunk +Thanks to all for the evaluation +I was able to start a cluster wo issues and run a basic jobs +Can someone take a look at the most recent patch so it can commit if it looks good? Thanks +greatCurrently crawling arround sites with their outlinks +move fix version to from from to as there is no patch versions as it probably will not get fixed for +IS an bundle all ready +This should be resolved by but a short term solution would be to use Override annotation indeed +These data structures are updated by the and queried by the +As of this is still true +is the same as a except a comment is added to the classBryan what you describe is what I had in mind as wellUnless I have misunderstood you the difference between my current patch and your suggestion is that my patch removes the two unused interfaces whereas youre considering to keep themMy thought was that since the two interfaces will be empty and there is no other code needing them it was better to get rid of them + +bq +Well resolve the umbrella after we get the x s and merge into trunk +Do you see this? Thanks +Such a keep alive could specify a timeout value so that the external process could control the definition of failure according to how often it will send but that might be an unnecessary complication +Will do +Im having the same problem with to set fix is definitely fixed in the release +It is working now +If you are starting from an XML schema Id need the schema +I think it does make sense for me +The parse assumes the unique key is a long +Version of svn is because of the problem doing a release using newer versions +Marking bugs with invalid fixed for release as fixed in. +In current implementation spring application context is started per each test but it is very time consuming operation in common spring based application +Heres the result Failed Errors Execution Time h m s Failed Errors Execution Time h m s Failed Errors Execution Time h m s Failed Errors Execution Time h m s Failed Errors Execution Time h m s Failed Errors Preparation Time h m sExecution Time h m s Failed Errors Preparation Time h m sExecution Time h m s +Check my +This patch includes review comments and removes hidden files in directory listing on jobtracker end while generating splits +Not basename basedir is the one which is replaced by +Do you have an ETA for this fix +They all seem to be related to versioning +I move this issues to the jobtrackerConsidering two possible improvement +Mikhailthanks for your patch +The current bash script which implements the interop tests starts the server as a background processand redirects stdout to a file to capture the port number of the server +Try paging through the user list +v includes test I failed to update +It looks like the parser does not break in this situation but merely assumes that the contents inside the is merely a string +I have successfully emailed it to other developers though and could email it to you +Its broken in through both the executable and the interactive console but works fine in both in +We currently dont store or restore the original Principal as part of the handling of the + +Then could take instance into account and enable users to prevent an unwanted instance from running +I retested this last week but was waiting from one of my dev leads to regression test their stream +A good example is using profile to set classifiers and dependencyManagement for different JEE containers + +Im reworking those tests to do just thatThose tests are attached now +if the server is gone with messages how we are supposed to get messages from I said earlier +committed the patch into trunk and branch +Its not really possible because it needs to implement interfaces that come from hadoop +buildivylibcommongriddevbin +Resolving. +The second link is the descendants link and the CMIS specification defines that only this link supports deleteTree +Adding the source code that is used to replicate the problem +The use flag and the underlying are only designed to work within a Spring transaction manager +Why dont we sipmly add support for a requirement to that you must supply on the constructor +trigger a BR + +Im using to wire my dependency to action classes +Interesting it is probably reading Xlint as an XML namespaceLemme play around with this +Added to the affected versions though probably all previous ones are affected tooHowever I wouldnt consider this to be fixed in since priority is to push out a graduation release +Are you using jboss embedded with your Tomcat installation +Reopen it which is a valid issue +Reopening to address Ferdys concerns +Made sure we dont call close on a null dir +In order to backport these things what is the branch for hibernate core for these modules +you must have a classifer and if you dont like the default one just change itOr what bothers you is the fact that the jar in lib is named +The allLeaders cache stores the controllers leader decision for all existing partitions in zookeeper +Though it might be a good idea to contact the author for clarification of the licensing terms +Patch testing again +If this would be possible to be supported it would improve usability +I looked for hdfsxxx +That will be useful for programmers who want to understand the code +It then proceeds to try running the jar file itself as a Ruby script +Also the policy check incorrectly comes after the enqueue on disk +Actually I saw it once but it was due to delay before the UI was loaded +The libgomakefile and the test +I think erroneously removed support for that property +I was working it for another ticket anyways +and that we should follow this up +The consensus in the mailing list is to include the plugin into core. +As it turns out a similar problem exists for some builds on +Lucas you are the main opponent there so what do you think in relation to this one? Can we do a deal +Its almost as if flushes succeeded or am still researching but it seems that simultaneous flushes are interfering with one another somehow +pushed +The code in and related statement class looks like they add FOR UPDATE to all select queries in dbLocked mode but to confirm we should take a look at the queries executed against database +When I apply this patch the last line of solrconf seems to get cut offSorry my bad +FYI +How many projects in your instanceHow many build resultsbuild frequencyWe have few instances that run since one month with more than projects and we dont have OOME and it dont get slower over time +I havent looked at the jackrabbit jcr commons stuff yet but I think the cnd parser etc +Concerning the interface has been changed to provide conversions for data types +Committed fix to the compiler in rThe test program now properly raises a runtime NPE with both Native X and Managed X +bulk close of resolved issues. +It is not there by default in can be pulled from the fuse repo or from an distro +I just wanted to point out that it would be nice to somehow keep this flexibility once the standard way of configuring this is via XML +I havent tested it yet will do after checking the unit testsscratchDir is the directory from where all file sinks write tothe movetask copies to the actual locationIf you look at the task log there is the default which cleans the job I am not sure what job is used for it is not used for the final output +Please verify whether the problem is resolved as you expected +I cant make a lot of sense of this issue +The patch was updated +Associated with +Etc +is used mostly for search purposes +It seems like this issue is a duplicate of +so no action needed to answer my own question +Im not convinced were quite back where we were so Im moving this onto as a reminder to look at it +For is there any need for? I dont see it used anywhere and it might make the sample even more lean +Hi +to add another workaround +Running the gwt script now just sets a bare bones GWT project. +patch for review +If there are specific cases where access is required or properties need to be overridden then they should be raised on a basis +I was waiting for JBT to be released +Patch for +Thanks +Checked this into trunk and +Il check it on winXP only +Added a small fix for CDATA sections but not a full fix for this bug +even deep in some backend codeThat really scares me a bit +This is for things like an undeploy when the application you are trying to doesnt exist. +Oh sorry I didnt see the patch +This is now fixed and available in following branches +Assigning to Arnaud +Could you eventually detail in new tickets the two last metrics you were thinking about? +We are also intermittently receiving this issue +Lang please check that the patch is applied as expected +I think +move to next release +From SEO point of view it will be very important to add canonical link say crawlers which URL is the right one. +That solution worked perfectly! Thank you +Ill keep this bug open until Ive put together a back port patch for too +It may not be possible today if none of the methods are actually forwarded to the server +as patch reviewer per JIRA meeting June + +Please let me know if you want any more info that could help you debugThanks calling this fixed weve been running tests for a long time and precompilation is a part of these tests. +Same bug as +you should do an svn diff from the top level directory +We should add a note in in and branch as this is an API changeSomething like Deprecated constant +Is it right that is the responsible for remote timeout in the configuration connection factory +check this and verify that its not a setup issue +By latest do you mean +If there is something we are doing wrong I am sure a nudge from Day will push us in the right direction +Please try to update your code and check again +Attachment has been added with description tiff images to reproduce P open bugs to P pending further review +Could it be related to creating the +I have conveyed to Chris that this should not make into the branch +Sending srcsiteaptexamples srcsiteapt file data +With more than a generous bout of help from Amareshwari and Devaraj attaching a patch that modifies the original one to not rely on the getters for the context attributes +For example one might layer a on a local filesystem +Hello thanks for commentsAs result I must recreate DOM tree sometimes for release allocated but not in use memory +A fetcher corresponds to the fetching from partition of a topic +In my proposed circular dependency scheme when theres a problem responding to the event synchronously we try again asynchronously but I think that the default needs to be to complete the response before returningIf we were to do this I think we could use the +Entries from docsdbTables were reduced to a single line as entries were failing +Do we have to mark this as an incompatible change and bump the protocol version number +how to handle source features is separate from this issue though +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This issue may be completed as part of although we may want a smaller scope for this example to be easily consumed +This is the same issue as as except for the version numbers. +Looking at the patch I think I might misunderstood something in the build method for every category the right policy is checked but the build itself is per so why cant the policy be the same for each CLP? If one wishes to get different policies etc I think it would be logical to separate them to different clps and this check should not be performed over each category + +This issue was resolved with svn commit r +Hi Ken I cant see why Adobe would have any issues with this wsdl +It might be worth including an afp extension that can generate BCOCA and perhapallow the user to specify the presentation space +The resulting order will be valid and the optimum insert orderI might actually sit down and do that provided that the hibernate guys would be willing to include that change +Maybe later. +So I added some plumbing into the mix and it works +Updated patch and documentInstead of using reflection in the Encryption facade to avoid compilation failures if is not available now sources that reference that package are conditionally included in the phase if a new crypto profile is selected via Dcrypto +I have replicated the very basic issue at +Keeping it the same has the slight benefit of making it easier for the minority of people who defined their own string fields and purposely wanted norms +Hey guysNot sure where we are on this but it just hit me +Thanks Greg patch has been compatibility should be think that was resolved and never marked as. +I am marking this bug as duplicate of because that issue willcover these aspects + It would be nice to have a guide on pom inheritance what is inheritance and when to use it +Thank you sir +Click on Billing Accounts tab httpslocalhostaccountingcontrol +Fixed +Discussed with Sharad about this +will it be OK to include the solr parts in this ticketSeems the best way to do it to me as well +unaesthetic question build seems to contain old source you +It works rightly if design mode are not turned text is selected correctly after its been saving the page selecting text in the source does not influence the visual the conclusion is in general it works but then breaks +Thanks for the feedback Marek +I think these two tests should be deprecated +Wed love to get involved into things that potentially break usThis is a very serious issue we have to verify whats going on and to decide whether this has to be fixed before we go final with review could it be clis that renames the keys so they dissapear +It is a flag to show the someone wahts to continue working but they dont have anything to do right now + we try to keep close to Java semantics +I had held off committing my update to since I seem to be hitting what I think is a bug in that version that makes my jobs fail +Let me take that back the issue still exists in +Thanks Hong +Ill commit this part shortly but leave this issue open until we are happy with Ill look at the wiki page and see if theres anything that comes to mind good idea +Will clean up test case and attach it to bug +Hiwhen you use you should configure your console using system properties +Committed patch to the trunk as revision +Preference I think +Can you add something to the GEP docs describing how this can be usedSounds like another good feature to show off in a tutorial. +diggingin the darkGrz in the Cocoon core lib dir +It should be added to the next release but the poi limitations I mentioned still hold +Please let me know if you have any comments or need more time to review + +Okay now this makes more sense +resetting severity from major to normal pending further P open bugs to P pending further review +Thanks Steve for all your support +mvn archetypecreate ideatest +Thanks +it would be great if you provide a patch for the trinidad support module +The problem with Portlet filters is that they may be called even if you dont have a faces request so for most usecases the configurators are still the most efficient way to go +If we want to update these guides to be based on the would we need to also create a new archetype for that or remove that reference are the current guides somehow tied to JDF or can we do anything we want +Im using perforce client Thanx +Thanks for finding thisI cant yet see how this can happen yet it does with the protocol you provideIm trying to create a minimal example + +can you try new ui +Added a check just in case +committed patch in close of all resolved issues in preparation for release. +This may only be needed for epoll I am not sure and it may be bad for I am not sure +richdragListener +Thanks for going over the patch +for this Id go for moving to the latest versions that didnt cause too much grief +Opened to track this. +Dispatcher will not leave the whiteboard till after +Fix the issues raised by Ted +Any way of making that happen? Guess Id like to see a message when it starts and a message when it ends +One of the reasons why we get bookies available through zookeeper is exactly to have the ability to remove and add servers dynamicallyAs for how to fix it I was thinking that we could set a watch on the bookies znode before starting the bookie and blocking until we get the notification +Closing all resolved tickets from or older +Resolved +I have kept tracedebug values in not able to get the details of this issue +Perfect thanks +Sounds your just to be more clear my concern was say we have peers A and B and the newly joining B has some old data from A in a cacheloader +Looks good might be steep but it shouldnt do more than a few iterations +Once Ive something Ill create a new issue where we can discuss those changes there. +Tested trunk works. +This shouldnt add any new +IT now passes on CI +I have created a another issue for incorrect value of +Thanks Himanshu +Sure thing +jar related files under +There was some discussion about providing feedback to the scheduler about what resources each container is actually using so that it can potentially over commit +So it may turn out to be interested research but dont be discouraged if the amount of work to make it feasible to commit into the mainline is too much to be worth it +Initial set of mockups created attached +ejbinit trigger the probject to build which I dont really like +Change how the locking is done +If the annotator were remotely deployed this could result in a significant improvement in performance +patch looks good except for the point that Pat mentioned above that its moving the test toward logj than using slfj +This first patch only provides the infrastructure +committed thanks Siying! +Oh man every time I should wait seconds longer before sending content I have lokked into the class and saw also the transitionMap and that stuff must be changed too +new patch that removes accidental junks in +more info needed should be fixed now showing all successful builds +Why? You should be able to specify dataDir in more than one core +My patch changes quite a bit of this logic and it just adds to the merge problem +Removing fix versions usually the committer sets these at commit time +Otherwise a bunch of method signatures needed to be changed and the goal was a quicksmall fix +It seems to me that the build fails randomly and I cannot find out why +This is now back on the branch +Hints are handled independently by each node +Ill let you all know how it goes +I had a look at this yesterday and it seems possible that the entries can use a hostName override using the new system property or whatever name is decoided upon +Committed to trunk and +Note that this demo project has been exported from Eclipse and has been emptied of all the +Trying compile issue +In my rule I had an event which has a reference to another object +Pull request rejected by Brian +spaces are trimmed + doesnt allow it too +Again if we just want to address the fs scalability concern we can be done with it quickly but looking at and this might be a good a time as any to bite the bullet +Quick glance suggests it needs unit tests +should be splashscreen not I prototyped this today I came to the conclusion that we may not want to do this +Thank You +I will test this and get back to youmany thanks OlegI will attach the patch to this JIRA Issue in udiff format +Havent run into this before but Im sure we can easily enhance the plugin to support this +I will not be able to debug the offline machine until late next week +Is this with the EAP server in RHDS +That seems fine to me +There was a bug that threw away the explanatory error message you should have received and I fixed that +Changed Target Version to upon release of +Looking at binbuild +If registration of wrapper is implemented that bridge wont be needed anymore and should be removed +Migrating bugs marked as resolved before and tagged as version as fixed in RC + +Attached new synchronized with latest Java most important methods from into to support current Spring Extension projects +Merged to branch at revision +It doesnt contain the detailed modification plan though just the algebra +So now i have access to the document on Google Drive +It failed when compiling the AS +Roger the patch to me looks good but note that trying to apply it gives to me an error in theProbably you need to update to latest trunk and then commit for me its all goodNote have you enough grants to run a build on JenkisIf not you should ask to Todd its useful to haveBye +Robert Morse what version of ToolsJBDS did you see this +I have ran the job which used to timeout because of the deadlock a few times and it is running Ok now +Moving to this done with the integration tests that were put together? +As i believe username and password as part of connection parameters doesnt make much sense with Cassandra +Thanks AshutoshI will rebase the patch on trunk and upload patches trunk and later today +Patch committed to the trunk +Attaching patch unit test included +It appears that Groovy tries to run only the first class defined +it you cant reproduce the failure Ill and post the log +Bobby please let me know if youd like me to commit them there too +This was fixed in trunk +Dmytro If this change fix the problem with Windows please add it as a note to the doc related to the webdav contains notes about possible problems with webdav on Windows and Mac OS. +Closing along with +Ill nail it down to. +Andrey Here are some review comments on the suggested fixes +bq +Fixed +Fusheng feel free to attach the design doc if there is nothing confidential in it and Shouyan approves +update the zones file on cd etcnamedbzones svn up +bq +You have to know the exact name of your image though +libs +since we see accosional problems in i backported it in revision . +I guess it is a matter of opinion +The only primary difference here is whether the server starts with deploy to the server area or deploy to metadata +These html documents show the gain in performance obtained by applying the patch +Ill look into it +Set the broker URL on the output channel during initialization of the listener that handles requests +It is slightly modified to allow specifying of an authorization id for the grantor +Adding attachment with Grant License +Thanks Pat! +bq +Confirmed fixed on windows closing. +is first added to contrib and then moved to core later +If were going to remove the c requirements then this patch should fix the comments too +closing self reported issues resolved on versions already released +Thanks ChackaravarthyCongratulations on your first patch to be committed +Comments on RB +May somebody pls try this +If you can check and close this issue I will appreciate +LukaszI agree this should be solved with the +Committed at subversion revision . +There is a bogusly attached on the which I need to remove and +I have uploaded which has the changes suggested in your review +Please push it Steffen +Fix checked in +Closed upon release of. +I hope no one will mindMichael. +The same session can be found logged with wireshark inFurther the contains the relevant code snippets we use +pushed aeffcaefbfadbaae +Build a single osgi bundle containing both api and impl classesThe osgi bundle hides most impl classes +Thanks for addressing my comments Olav +Resolving. +I added a +As I said earlier +Unfortunately versions were just released so this will have to wait until but I can deploy snapshot versions right away +move method back to its original locationI moved the two static methods of to the beginning of the class definition +Removes include +On the other hand this is a very common case I think +Created for docs note that because of time constraints this review must be complete and returned to the docs team by Fri Dec +Then you will get the error + JVM? OS +Bulk close for release +more tests for covering different kind of options combinations +Ant project with some fixes +forgot to turn off markup so the above ended up looking pretty bad +This has been can you check and close this issue +From Dicks email Made the change on the weekend someone has to test in Tomcat to see if theproblem still exists +Paul can you +Our web application accesses the process definitions and forms from guvnor +Hi Aslakas I mentioned above migration of blogs is next step of SBS I work on it and I dont think it will take much. +OK Clint +The patch should also add an option for unit testing to the buildit script +Also there could be some corner cases that might trip up the current approach +This includes seam documentation and readme files +Mistake in previous comment the unit test is +On a Mac the first and the third seem not to work. +I also left the client connected all night with no problems albeit we have a m not sure about idle clients when running the example it sometimes times out on the server maybe the reconnect code could do with improving but its quite hard to test this sort of stuff +Done. +Patch is committed to both trunk and branch +Assigning it to me since Alejandro is going to be offline for a while + +Saxon bundles only wraps saxonhe +Forgot to mention that the exception I got without this modification wasStream ended unexpectedly +In the Javadoc can you mention I did a little coding on this +Do you want to commit it to and trunk by yourself? Or by me +I see the same behaviour of creating as in +bulk defer of open issues to +I have created drafts for and as per the visiblity we have at the moment +Both Spring Beans and exist in the EBR +The bug I linked to was also a tough one to track down so no worries if it takes a while to get back +However it was not my primary concern so maybe Ill be more lucky today +Here is a applied for build +I added a patch off of to revert +correct the quickstart name TransformXMLPOJO quick starts transformXMLPOJO quickstart quick start called samplesquickstartstransformXMLPOJO quickstart called transformXMLPOJO The quick start runs quickstart This is configuration is designed to transform a DVD Store message into Java Objects +Then you wouldnt need to use the tool here +I agree with TedIf it was already failing in previous releases I dont think this should sunk this current RC +I recommend applying Transactional at the method level but I agree that this is only really feasible if you dont have any rollback rules or have a way to share your rollback we get this backported to the branch +However Ale what exactly are you doing with that protected method? Are you overriding its use in? Or are you calling it from outside in some way? Would be great if you could post your custom version of it then I could suggest an appropriate solution +Hi JayaprakashAre you able to answer the questions which Mike proposed last June? Without more information it will be hard to make any progress on this issue +Documentation task completed in all resolved tickets from or older + +Can you provide a patch for LuceneSolr trunk that transforms all those classes to static inner ones? This would be nice to have but I see no real need for doing that now so I am a little bit reluctant to do it +I tried adding a maven project and didnt get the error +resubmit patch after fixing the assertion failure. +Fall through seems to work as this accessor is picked up and it does in fact solve the as of not reproduce +I have temporarily commented outtest since it hangs the test suite until this task is sure to uncommented it before closing this also fails because of have commented out until this task is failure on is another case +Just for information passing this additional property to the mvn executable does the workaroundcpdtruepmdtrue +DOH +I am not exactly sure why do we need that +This isnt fixed yet +patch showing what the backport looks likeits unfortunate there isnt more code reuse between Slow FST but that that can be refactored another daythe patch includes some additional test permutations to ensure the param checking works right regardless of which delegate is pickedim still running full tests precommit +Im confused +Sorry guys Ive been pretty busy work but I have been watching this +OWLAPI is now in Maven CentralYou will need to change the artifacts to pull in the desired parts +Please verify and resolve +I was getting a mismatch between trunk and and was confused for a long time as to why the patch would not apply on Jenkins +its been a few months resolving this +This is another version of in r +It requires a JEE server to replicate since a is required +Or do you only use an to generates instances of +I missed adding the file during the splitting of the protocol interfacesUpdated patch attached +Is this one of these? +verified on updsite +Issue is fixed +Just the servlet api +an old FAR jar may or may not work but if it doesnt it can fail in very subtly ways +Its changing several files in several maven modules and all paths are starting with relative path to the module with src +I started at added generics there and worked outwards +thats the one Dart +Since this is an improvement I dont see the need for a test +converted contrib will commit this in a bit if nobody revision . +Ill provide a patch later today +M introduces dependencies on and as they are needed by Axis to generate the WSDL at runtime +i get wcs reference back + +From Section of RFC The Modify Operation cannot be used to remove from an entry any of its distinguished values those values which form the entrys relative distinguished name +I believe this issue to be resolved with the commit +Close Doc issues in release versions. +I will open a new issue for adding build scripts and sources for our test assets which are jar files +How do we handle this? Do you plan to include Bobs? Then I can wait and rewrite my test afterwards +being able to know what other itds introduced would be extremely usefulyet i believe it could be a real challenge to implement as one would then have to deal with addon priorities or some similar this expected to be resolved for releasewhat is its current status +This doesnt look like a bug and therefore shouldnt be marked as such +Assigning to Ales as this should be done in the context of the osgi bundle work +patch looks good +Hence we check for condition first if there is already one of thread pass through then rest of the reads dont need to go through the same subroutine +I can not understand why does not original issue was closed therefore I was more surprised that you still could add a new open subtask +However if the cleanup and the mavenization change are intertwined I dont think that makes sense +All other are gone +Neither initially nor when you hit the play buttonI have run ant and +Have you tried this using the Servlet Protocol +runs cleanAny proposal solution is welcome +So should I go ahead and check this in as is if tests pass? If so should it go back to? We arent really set up for performance regression tests so I am not really sure what testing we could add +I believe that there is a way to limit the export to a subset of classes in a package +I think supporting via a separate JIRA could be good feature +While still not really a full verification test it does at least verify that the GLS impl provided does better than OLS for models with error structure conforming to its covariance matrix +Actually tomcatrun is not an internal it is just an issue of providing a maven option via the tomcat plugin. +However if I removed the explicit transaction the issue never happen at fix for apparently also fixes this problem +Yes I will do it this weekend +section it says the schema will be created earlier than it would have before is this true from the summary it seems it is created at the same time as previously +Committed. +I hit the same bug today +Please move the bkjournal package to contribIn principle I have no problem with this +looks good to me +Closing all resolved tickets from or older +Heres the latest patch on top of Hadoop looks about right. +The information was added to the guide +I think its actually the right way to fix it +im thinking of setting this to wontfix since there is so much that has changed in sh since this was +However I am not sure how your suggestion will work with dynamic maps where the values are computed based on the key provided to Mapget method +You can obviously take that conditional out +Per the JIRA triage meeting on February I agreed to give the official reviewId like this to be Kevin C +well for this commit its an improvement +implemented +Find attached an that might be interesting to some people +I succeeded to reproduce the in the branch and trunk. +Is there anything I can do about it? I think I cannot commit the patch and hence resolve the issue +Maybe we can implement entries in terms of glob when within a jar +is this still happening because we have improved the permgen cache can confirm issue using using Resin on a Win development machine +These issues are partially addressed in +Should include our log categoriesThe Current Structure information is not useful to users doesnt teach them out to do anything +This is to optimize existing code +This is required to avoid FS contention when running mini mr clusters in parallel +Other quickstarts havent been updated to the latest WFK builds +sounds good to me let me know if there is anything I can do to assist. +Please confirm the current status of this issue and if its incorrect change it as required +IT updated. +If you use logj or commons log set this there +Ive got a feeling this isnt going to get fixed for since Ive been nagging Tom about it for quite a while +You may also want to consider supporting IN and BETWEEN clauses because particular may efficiently implement these special forms +Yea I dont see the problem any more since the file is not deleted +Given that still can not be contacted why not close this issue as resolved +Also the version in the root changes were inadvertently in revision +JSP stuff is handled by jetty so no conversion is necessary +Defer all unresolved issues from to +Theyre pretty simple but theystill put in the +Stan the database you created does not have the collationTERRITORYBASED attribute and that is why you couldnt reproduce the problem +Now it is more abstract IMHOIll send documentation for this part together with documentation for statistics module +Observe link popup dialog +On the other hand you are of course right with the object creation taking time that is not really needed +Merged to with revision . +The problem though is that most of packages for CENTOS and alike are real old in the hood of Groovy where the latest is or even later +The sending of the version number in the service contexts should probably be removed for Yoko +Testcase for bug attached a testcase containing both minimal source code and a mapping to reproduce this noticed that it only appears for collectionsarrays for single types Castor gets along fine without a setter. +Anyone? Is throwing an exception bad if theres a? Should we just WARN +This issue was happening as my local set up was connecting to IBM with IBM libraries +I think the answer to that is It Depends +I know that were loosing portability but in the real world when you set a DB provider you dont change it in the near future just because you need to optimize your code in order to that your app +Note that the time taken for each invocation of sort is proportional to the number of keys collected for each partition which itself is a small fraction of the time spent in that phase the number of invocations of sort is roughly proportional to the number of reducers or more specifically the distribution of output keys effected from the partitioner from a given map within the limit +How to use module +sending to you +You might want to send an email to the dev list to see if someone can give you a good use case +The issues are now closed. +This mechanism could also be used for session scoped temp tables internal mat views with lobs or to allow lobs to be open after a statementresultset being closed. +Only needed to remove tests from. +Suresh I still dont see what is wrong with the patch I providedKonstantine this is deeper than a patch issue +This is on purpose to prevent any regression for +After much thought Id like to on the approach in this patch as an emergency fix to and but leaving the discussion of lease close semantics and block length inconsistency handling to a later timeComments to the patch lines in the merged why descriptors is not declared in the else block +Attached youll find an extract from the log file after using the Synapsesnapshot buildIt seems that now the same problem occurs when sending a messagewhereas with the releases the problem only occured when sending messages +Needed to rebase but only really changed moved component to with a package private that handles its share of the parsing and returns the remainderI also confirmed that this set fixes. +Waiting for hudson results +Mixed usage of hostnames and IP addresses will cause problems even though the same namenode is being referenced +Those are simple string properties that you can declare and then use in templates within the archetype +Do you have an update to this record? The asynch appender and any other issues you are having should be opened as separate issues +It looks like SEAM tries to start managed transaction and fails to do so +The changes look good but I thought that there were more tables that needed to be updatedDid you check all of the Derby docsAlso you need to recreate the diff file +Move the fix to +rebased trunk. +I can delete the new code entirely and the test still passes +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsermenu +This will be resolved by +I just was thinking about having the Photoshop files or Gimp files which are used to export as png for the softwareIt would be easy to create new icons based on such templates +SharanIf you want me to put your eBook in the Learning section of the wiki books page please give me the elements I need SharanDone with the help of your email please check. +Added the redirect for jdo to the incubator +type id link did not poit to +Mircea as a quick fix you may try the component approach with the binding attribute and try setting the values of the in your value change listener +I just committed +I send you the last update for italian +Applied appliedSending CHANGELOGSending libbuildrcoreSending speccoreTransmitting file dataCommitted revision I didnt see this issue in other places +Bulk move issues to and +Fixed in trunk. +oops quick typo fix by I meant +reopening and resubmitting the patch to hudson bcoz the patch didnt go well with hudson as the status is set to resolved +If you want a list and know the data well enough to avoid delimiter issues its easy to convert the array to oneThe valueArray function currently has no documentation +Providing the project on such widget is mandatory so when adding a project widget to a global dashboard the Edit mode should be automatically displayed +feel free to take a crack at this if youd likeEddie +nkeywal Where is v? I see v attached here +This class is no longer used and should be removed from the source tree +NOTE We are not accepting any further QE feedback unless they are issues that would result in data loss +Consider the scenario where sets of shards index the same information for redundancy purposes +So we need a better way to examine the entire contents of the while loop to know whether it can be optimized to a simple loop with no exception handling +It is suppressed by default +Hm the archetype command line is working for me +Patch v make the barrier local for better readability +Do you want to close this editor and reopen the file +Ive not applied any code style changes from your patch since it was already applied by someone else +The text you are seeing is the underlying GroovyJVM type name for an array type +Nice contribution Apurv +But since I was already going to be digging around in that class I was going to expose whatever else seemed useful +Seems trunk has o but has o +Chris got time to take a look again +sorry should be lower timestamp will reasonable +Otherwise a delivered hint can undo a deletion if we gc it away too fastSo this would only work safely if commitlog mode is sync AND hh is disabled which doesnt seem like a common configurationOr am I missing something obvious +Theres also the workaround of creating a reader with the correct encoding and then use eachLine on that reader instead of on the file directly +bulk defer of issues to +Prepping +New patch for Y branch not to be committed +I think the try block should be removed from your test +I was looking at the and not the logs +No problem +Sorry for any inconveneince caused +So may be will chk the reason for failure once again +Would it be possible to add a comment that also explains this into the for and for v? this to Rich for review +In regards to the moves I dont believe there are any true mvs +Yes not your issue +Because of statement +This version looks better but Id really love it if you added a testcase to +These changes look fine to me +This will give us sometime to really see if the RS has not got the call from Master or the RS was really slow in processing the close callWhat do you say Gao? TedPls suggest +Fixed tabs with spaces +Hi Ryan Since you have the environment do you mind testing vIt is not a final patch I have to verify the accuracy of the estimate though +Verified at you provide example how to trigger onclear on server sideI use and cant trigger onclear in bean on server side +This stuff is horrid w its h and h versions of everything +How do I make sure the new classe files are included in cassandra runtimeAttached is the complete log for startupThanksWenjun +Are the JSP macros still slated for rc? Has an implementation been chosen or does work still need to be done +Im not insinuating that Abator do any more work than simply translate someones columnOverride into a List of Models and apply a select reference to it +This probably is true because df is already in the path +I dont see test guarantee this sequence of operations +patch committed to trunk +Bulk close for +if we need a means to disambiguate task attempts from different JT restarts we could use the JT startup time instead of the counter +we expose partitioner so that enterprising clients can route directly to the right nodes for local reads rather than incurring the extra hop +r on trunk. +The is applied at +But feel free to do it or add this to another JIRA it there is one for the conversion effort in trunk +Well thats embarrassing +That would definitely be a bug +Fix included in release. +Matthias what would you need from me to assess the problem in detail? +It was a good one so its included in the patch +So the listiteratescroll is to force the query generation to behave as it would in the different cases you would be using this hql +When the tests were written I think we didnt know exactly what we were going to test and made extra classes persistent in order to have some flexibility +Fixed within with JBT. +Exception message Problems occurred when invoking code from +binschematool is not executable and also invokes the now renamed class with those changes. +Ive changed super type to simple Annotation now so the problem is fixed for me +I am still in the process how this interface should be finalized +Christian are you able to provide a self contained test that shows the problem? Using my test classes I was not able to reproduce it so probably something more going on here +I expect this issue has been resolved for +Marking this as duplicate +Its about the of a plugin which is a good thing to report +Increased priority to Major as it is a Major PITA when debugging configurations +If there were a Junit test for each class that were Serializable that would shed light immediately on object relationships that break the Serializable contract outside of the context of what app server is being used or how +Committed to trunkCommitted revision . +Isnt the above a blocker for this JIRA in terms of passing trunk buildNope +It seems to fit on the default wizard window size on my local machine +I was just going on the JIRA subject we somehow figure out the message id header is a general issue of serializing headers is a complex one particularly when folks add things like JMS or backed channels to their flows +For some reason it gives the wrong response and its hard to see why given that it works for other methods and is single threaded +Both look good to me +Defaults to with commit fcafbfcddacd +Thanks Tahir! +CXF doesnt work well with NTLM but the Commons HTTP client does +I meant the tables as in sort the results of the list command +Scenario I am adding Derby version prerequisites for our product +Hi GuysThis is my first contribution to sonarPlease find my patch to fix this bugThe fix what Ive done is that Ill take the unique list of the userroles and usergroups +Do you use maven for dependency management or you have setup classpath manually +in what kind of occasion would a distance value of be correct? only the source vertex has a value of when it starts +I can provide a patch that would move most of hadoop dependencies in their respective profiles and will define the correct hadoop deps in the profileEric sounds good +I think the retention policy of is set to RUNTIME in order to enable automated detection of concurrency related bugs using some kind runtime analysis tools +G +Code changes look good +Yes a builder would also work and would allow the class to be truly immutableHowever it would mean bigger changes to the APIA simpler fix would be to add a flag to reject further changesNot immutable but since code uses volatileEither fix would be fine by me + +We dont have any tests for we should add one +Applied in svn revision to add the right Fix Version +This patch no longer applies. +I cant depend on it to behave the same in future versions +Thanks for reporting that and sorry for any inconvenience +My mistake Translator is supposed to be the local interface of the SFSB +Also includes Mac platform specific handling and handling extra cc attribute that can not be set with properties +Closed upon release of. +The comment TODO Remove this after a fix for WARN This wont work for compressed! should be removed from +This should not be hard to change +Otherwise we should just fix +Can you please rebase? ThanksAlso with LTC gone is this still required? +Thanks for the patch Jody +Without ability to reproduce the problem there is not much more suggest unfortunately +Patch applied with many thanks to Clark +We dont widely practise carpeting in Catalan homes because we never carpet bathrooms +for the patchFor Apache DS you might want to look at startKdc profile in +DITA source files translated and diff file for HalleyIm hoping to keep most if not all language specific build targets out of the top level and keep as much as possible in the files +Here is the patch for this task +I dont know why the sources arent included so I wont be able to fix it this time around +Most of this is done in the programmers guide but I have improved the spec as well +Siying Thanks for making the change +Here is the draft comment for the Release NotesUsing the Full Publish option in the view to republish a project doesnot work around this issue remove the project from the server and then add it again +Hi Mark sorry for the late response Ive been away for awhile +Igor the commented code been replaced with protected method overwriting one from the super class +When the namenode is restarted it will have this last block as and when NN tries to release the lease on this file it will try to recover the block and will never succeed because the block is not present on the datanodesHowever it seems that it is the same case now when namenode is not restarted the existence of this block in memory and absence of it on the datanodes will lead to the same problemOr another case that is similar is when client calls hflush and then Namenode client all datanodes that are receiving the new block crashPlease correct me if I am wrong on this oneAll in all it seems that if the namenode crashes it may lead to the client dying and so the probability of this happening might be higher than my first example of what might happen today +This is in Section additionally this section mentions that RMI is the default transfer + +Sanity check Have you tried the most recent code? Scott put in a fix for a case where we were forgetting to push and pop RTF contexts which as far as I can tell should eliminate the risk of those getting out of synch with reality +Nope +Forgot to review this patch +The change to start mini hbase cluster was only inI am creating a new patch to fix the related unit test failures in as this patch has merge issues in +Proposed patch for +still one release to waitarggh +It sounds like currently classes will just silently start to behave differently +The point of the defect is that this class attribute og the caption tag is NOT renderedWhen is version due for release? We need this fix in my project +If no keys are generated an emptyresult set will be returned +I think this does need to be fixed but its quite simple to doWhen running on Equinox the resolver creates a capability for the system bundle with symbolicname +Reassigned issues to Nick and had to set resolution of all of them to done + +Thanks again Sandy! +Can we get this patch in? We have a usecase where some partitions were created with the wrong names +New JSP File WTP wizard screenshot is attached +Hey I just wanted to with this and see if I could maybe get some feedback +The only difference in Spring is that we revised the handling of which affects the GET case for Burlap There should be no nasty stacktrace showing in the server log anymore it seems that Grails has a special registered there which is where this stacktrace comes from +Furthermore making them ordered identically has a nonzero cost so I dont think we should do itI think we should close this ticket wont fix or we should rename it to focus on changing the test itself to be +Other than that its efficient on thrift server +Is there a defined dependency for Hives modules and if so does this break that? If so the other option would be to move this to the contrib package but to me contrib is a dirty word and Id like to avoid that +One way I see to make it faster would be to convert it to a state machine problem +! +to +Hi attaching patch for this issue and for patch also includes fix for the NPE that occurs when user tries to disable all known content consumers in the repository scanning page +if it s just an ordering issuebut Im attaching one here anyway +Ill check this in +It is probably beyond the scope of this jira +TrashCurrentmachinejob + you provide an integration test it may get more attentionif you provide an integration test and a fix it will get more added integration +Roger or I can see about getting node tests setup within Mendonca Thanks for pushing forward on getting this merged in +Thanks for the commit and cleanup +and please if the implementation stays the same update the documentation to mention that returnObject is null when the method has thrown an exception so that programmers know they have this test to add to each and every of their +Thanks for looking at this improvement request Carl! +The basic approach should still be the same its just that the class structure has been pruned a bit +retest and hasnt happened again +Every time new PE object is getting created and that is the reason +The work space opened without the stack overflow and the partially edited file was visible +test passes Theres an incRef issue still +How does the fact that we are basically deprecating affect this issue? +I think we should throw exception instead of setting startAfter to empty when the inode is not foundFor the new patch it may be better rename to or simply use the existing +Patch from applied see notes over there +Tom can you provide more infoThe does unregister when it is undeployed +This comment came from the original perf tuning section of the admin guide. +a Please dont just closed issues at will +If you get same exception please reopen this with relevant details +will look after this issue + +but do you want to create something at the top level that contains all of those and rebundles them? This is usually how the assembly plugin works for think we need to support this but after more careful consideration of how the module relationship equates to the dependency relationship usually used to build these +As my discussion with Jonathan we want to make the Binary Memtable Flusher thread configurable and we wouldnt want to have it under the same path as cfStore since that should be a different tunable +Having cores in the cluster we are normally using only Thanks for your help +HiI am attaching an updated patch that simplifies the code and adds the new property as a named constant in the class +LarsTedMaryann has come up with a patch for where there is no retry on +Can somebody kindly point me to some instructions to do so + +It does break BC +Attached is the output of running mvn X package using the version range +Truncation of trailing blanks for length less Clobs will be added as part of +Can you apply the overlays to the targets +Following this patch the values are submitted successfully however the checkboxes are always rendered unchecked both from submitted and selected values +Fixed on master and should be fixed +Rooted at trunkqpid this patch adds a detailed description of the C Messaging +The committed patch fixes the issue. +Manik mentioned there was a command line tool but its no longer in the distro +I suppose we could add an implicit one that doesnt change the window frame but that seems expensive +Hmm I dont understand what is doing? Can you describe its purpose at a high level? Somehow its failing to merge those small segmentsAnd offhand I dont know what changed between and that would cause to stop merging small segmentsMaybe turn on s infoStream and watch which merges are being selected +Testcase Fix +extends +Rich Can you please do this as part of the attachment work you are already working onthanks is unlikely that this problem will be exposed by a user +The reason is that the Avro handshake must be completed before any other communication can proceed +I think we want this in as well +needing your opinion on far as I can tell the algorithm is not flawed but I am not fully understanding your use case + +Same as the one provided in Stefan patch the Fetcher set the to protocolNot sure it is the right way it could be better that the set the to protocols +Most of the changes are in +Although you can manually set up a separate bin directory for Eclipse it rather defeats the point of having the plugin +Patch applies to both trunk and without trouble +Joram can you create an issue with Signavio for this and follow up +Latest patch containing fixes for the comments raised by Isabel during code review +I mean in your application code you wont see any generics +Can this be put in to review +Ill post a patch containing the tests +IMHO we shouldnt introduce this dependency now +are you still experiencing problems w the RC +Can someone please check if this was fixed in as Nicks investigations indicate +Your comments are very sound and I will look into your suggestions and report backCheers +Processing of deletes requires +Yeah wed need to exportimport memstoreTS as well +All the disagreement in the world is fine but you have to play in the sandbox +I figured out the problem with +Nicolas I suggest that you fill a new issue if none of your objects is a instance +Tested and closed. +Thanks +Its a decision made by PM +My plan is to first add basic failoverpassivation support for conversation session scope beans +Might be related to +bulk close of resolved issues. +All the lack of hashCode and equals causes is for the credentials to fail to match the cache keyed by the principal name +jar files needed we should remember to update itI remember reading in a forum a user read that was optional +David hasnt been around for a while so Im reassigning this back to the list +Thanks Stepan +I received the approval for the CLA from the University and attach it +BorisI committed your patch +Yes it could be I will take that patch and see if the problem goes away thank you +So to avoid duplication of code we could just use them here +Not sure why I cant reproduce this issue anymore +Sorry. +I think those couple of weeks are long passed by now +test ok bug fixed +No I still dont understand how explicitly updating a bundle helps when the dependencies are missingI am not asking for thatI was just giving an example for a situation where a bundle was installed initially by and later the service went away +xml files into a subpackage +Its from earlier +I dont think we should add backwards iteration to etc +new +Sorry bugzilla fooled me and I posted the above to the wrong bug +It could use a lot of optimization and theres probably cosmetic cleanup to do +Jeff thanks for the contributionAlso marking the jira as fixed. +What do the Solr people think about? I keep it open but I think it makes things to complicated +Is there a way to indicate in the only to build this jar if the minimum hadoop version requirement is met? A lot of the unit tests for this storage handler currently depend on a local instance that the developers on my team all have available +This patch supersedes the previous patch +I committed this to trunk and branches and +I have not seen a sample file +NET CF platforms +This hasnt mattered until recent changes with excludingincluding curl +Robert Newson Great to see that it is fixed in +Yep good catch +The added paragraph looks clear to me +bq +Steveyes agree with you that it is not wise to have multiple files in the way the current JPA spec defines it +patch with configuration of to get it +Wrote all values Status enum can take and fixed spelling +Again with the same cell stimestamp +But have you tried the second option I recommended? Please ignore the fact that target folder is created it is a Maven thing and you shouldnt be worried about that +If both are set it pretty much boils down to the code that checks for one flag before the otherCurrently the mgmt server doesnt have a way to specify mutual exclusivity between global flags +Do we need in addNested at all? eliminates it +As posted to were not going to get to this for +JBDS H verified on Linux and Windows Vista . +Currently treats these connection failures as real errors and will try different datanode +hmm not sure that is technically legal or ever has been +This is almost identical to a previous issue +Im going to suggest that we rename contribextraction to be contribtika and that we just roll all of these things under one area that way we dont have to muck with libraries etcHeck it might even make sense at this point to just move it into core +Vote passed on dev assigned to me as I wanted to get started soon on the tooling stuff to get javac javah and keytool in shape to put in an upcoming snapshot +Thanks for all the help +Verified in dist built from SVN . +Heres a new patch based on your feedback +simple there are no objections Ill commit it +Now that work is well underway with AS all previous community releases are +Wrong usage of constraints caused this not a bug. +Last discussion around this feature is to provide a implementation that is coded for a Hibernate store such that user can configure any store for metadata updates +For SVN trunk +This cleverness will probably have nearly negligible effect but it will guarantee the invariant more precisely than if you create a new centroid from the centroid if it has too much weight to fit on any of the equally near existing centroids +Currently we do not support such scenario I have hidden the abilities to uninstall a geronimo configuration bundle in osgi portlet +David I know it can be frustrating to not get any feedback upfront +Tested! +However thats not a option for me because we have WEB modules that will be deployed in one EAR and having DS in each of those feels a bit of a wasteRegards the and in lib seems to work and that I can live with +Its not a Tomcat its a Glassfish UR +Final patch of refactoring for review +Should I or will youThere is already an issue +But it is fine if you do not want this in the branch no problems with me +from the mailing listrevno is either id or parents or tipinventory is either manifest or less tests fail +Can someone change this to blocking there documentation available as to how and why the Spring Hibernate integration changed between Hibernate and Hibernate From the comments of Good news There is a working prototype already and we decided to promote Hibernate support to the Spring feature set now +The project return or any other information you IDE integrators are now capturing and we have to return those strategies to the default artifact resolvers +Is it OK for us to include these files in Jackrabbit as described in the Apache license +If these scripts shouldnt be executed during a release they should be moved to a profile +And whether we need to make changes within our code depends on how is addressed +Yahoo distribution patch for the issue. +HOWEVER I am up for doing a release in weeks with this being the primary thing scheduled for it +Thanks Hitesh! +The steps given for the st node are as follows +Version Released +I am including them because they are consistent with the rest of the in this patch and because Id like to start a discussion about the best way to include these +Tracing back through the dependencies will give you the service that is not starting up. +Then Id like to create a new class to act exactly like what Text does now +Is there an easy way to do that if is an abstract class +patch for trunk +Ive just committed this +And if I miss a case I get aninexplicably performance penalty +What you thinking? Yes hbase users need stability in and in +Doesnt ever save clear the for +Sounds good but the annotations are required for the XML unmarshalling +Thank you! +Pavel is this planned to be addressed for EAP ? can you comment +Opened for dead server stuff. +My previous comment can be a bit confusing because I forgot to mention something +confirmed fix + +You could also place a small sleep after the consumer create so that both messages arrive in the consumers prefetch buffer and then see that they are dequeued in the correct order +on JBT. +Simple code and new +Options were added to to support multiple job resources +I am wondering how best to associate an with given the presence of the existing interface +Thanks for providing this +A patch has been applied to the branch at revision +while we start on the console then everything seems to working when we start to start through ssh than the service start but the ssh never returnany tipp +Thanks issue is closed now +Thanks to Thejas and Jitendras reply +The No excludes message is quite confusing +Tab completion is difficult +Still you might generate name clashes but the standard case of a date property containing an underscore is unescaping is not worth the effort I think +Positive feedback received +A previous patch with the same code changes passed all tests except upgrade tests +Confirm that the updates are all applied successfully when recovery is completeOther comment on the V patch up on RB +Patch applied +I have no idea what we can do for users +bethard maps to apache ID case it wasnt clear yes the rest of the PPMC concursthanksping +I think is a much stronger solution to resolve annotations instead of trying to match strings +for the patch +Which excerpts dont scan easily right now? Googles KSs Lucenes H or HLucene H +cause I collect full hierarchy I should know relation between classes +The resulting binary aspects just depend on so no compiler dependency there +will stay at and will not take as originally agreed +I cant reproduce the problem using SunOracle JDK +Closed with the release of. +Looks like REST fixed the problem +Here is a modification +Still some warnings to do in the other CSS files and this one may need some tweaking afterwards but this fixes all the errors and warnings coming from +This may not be a major issue so feel free to adjust this down +I guess this is now doable using the new +It looks like this has been fixed for some time so closing the issue. +efdfabcbdcebcbddc +mule implements its own reconnect strategy by using the mul retryPolicies construct +New version of Mikhail for reviewing the patch +Patch for trunk and a doubt on my patch nowWe should track the skipped regions? or the included regions in the st run +Just some mod in the way the expressions are iterated because the code should be compatiblefixed at revision +Maybe Im being overly picky here but we are using angular on our project as well now and weve been adding an update method to our resources to handle this exact case and it hasnt been too cumbersome +Merged to the branch in revision Merged to the branch in revision . +See javadoc in for tests to check that old and new style proxy creation both work and that old and new style serialization also both work. +Thx +Sorry missed this +Andy I had created a perf guide for EPP but could not get a chance to create one for EPP +Secondly this ticket was created years back +The documentation should be in the branch with the source code so that the documentation is versioned with the source code no matter what VCS is used NB revision number only have meaning in CVCS because of the central nature of the VCS +to generating release notes +The NPE test. +Fix patch +A funny aside the continue lines essentially act like goto statements Yeah thats ugly +Can you please check if these files were present in the last CP ZIP +I was wrong to file this request +minor tweak for to +Added link to thread +This happens using the descriptor +Incidentally this still happens in version +quota check for the case where files are in the same with the latest trunk +shows that archetypegenerate actually does workDocumentation updated tests OK with version +Thanks Dmitriy for bumping this up +The does not exit when there is a Disconnect +Looks good +Why should a column family get not touch all of the filesbq +war to this issue last year. +Creating the Timer only if the timeout interval is +the is putting double quotes around each classpath entry +If the user is belong to the pool before could heshe read messages which are in mailbox after heshe leave the pool? +Ill have a look +The new patch combines indentation changes with code changes can that be separated +These test coudl really use any VFS backend and did not need to use HDFS +That helps to reduce the time of running tests +After a restart I can now confirm that is copied you both for the comments! +I checked the latest version and now it seems ok +Shravan could you take a look +This patch version has all correct build modifications in please +no Grab at patch looks elegant to me +Committed patch to documentation trunk at revision +and passed on my box +We had a discussion on this Jira during JBM meetings and decided this should get fixed on the upcoming release +I suspect is not to blame but rather Spring itself +id in sideEffects should beid +Another patch I think fixing the nd issue by doing a custom flushinside prepareCommit that clones incRefs the flushed inside a sync block so that we get a consistent commitI also fixed a test deadlock caused by the fix for the first issue have to handle the case where the DWPT is null because close is inprocessI also evild up the by randomizing everything andfixed RIW to sometimes pull an NRT reader after doing a commit just tomix that upThe test seems to pass now I think its ready to commit +Testing +Ill definitely get to your bug when Im back from +At least as long as you dont use the original package name +I factored out touchIndex cleanup and copyFiles to static utilities onMaybe instead of the static utilities we can have an abstract IRH with two impls and +Thats great +Users that actively use the current branch will be able to use these features fasterOn topic I think the plugable backends is a great idea +Marius is there any progress on this issue? We should update archetype to use WFK at this moment +Dprojectsvn commit message Quick fix for file consumer not freeing memorySending srcmainjavaorgapachecamelcomponentfileTransmitting file dataCommitted revision +We can keep that as a separate discussion probably better had on the JiraI will look into having a bridge available for the get together +The three patches look fine Im testing them on a cluster and Ill commit after +should be fixed +Its useful +They can also vote on Jira issues in case they want to encourage implementation of specific things considering they are dozens of suggested things order of implementation can be influenced by users indicating kinds of things they want to see done +v works out on a cluster for integrate patch v in hours if there is no objection + +I can explain why this is the case +my bad +Ooops looks like this bug is also present in as well +comments on is available on both attachment and D +A merge operation asked a tablet to split on tablet server X the split operation was waiting for aMETADATA tablet to load +we are now deleting the ipalias when last vm form the subnet is removed +I tweaked the javadoc a bit to better reflect its usage lets break out an interface if its ever more than a single shared function +Also if any existingfuture Hadoop deployments use disk striping will this feature work +I will move the test dirsfiles to zzzperms this is now excluded in friendsuntil the mino eosaurora sync is ok +A patch that fixes the problem +I will take a look at the issue on IIS based on the modified logic and fix this issue +if you fix the spacing +Yarn rpc metrics should be consistent with old style underscores +The same problem exists on projects that have been newly imported into the has been fixed and committed into CVS +The inability to build a distribution is one of the issues identified in the validation checklist for the Hyperic code drop +Thanks +And Im not sure whether there is specified order in the JEE specifications +Ah I do see the pull requests +Hey Allen fwiw that attachment is not the patch fix for this ticket +Specification updated +If it does fine but having an emergency exit would be a good idea +There in certainly some progress here +This way anybody could use it without disturbing existing nbpproject already in place +Thanks HadrianThat is good news and answers my question +Im on this issue + +Tests must be added laterM javatoolsorgapachederbyimpltoolssysinfoAdds privileged blocks around system callsM javadrdaorgapachederbydrdaSets if it isnt already set +test design patch and tests +I used to run all unit tests on the test machine I used and everything was fine then +This one should be correct +Thanks all for the suggestions and thoughtsPresently in our code base we are configuring the readTimeout in the channel bookie +this should fix to work well no exceptions yet +I was stucked in this bug and almost forced me to rewrite my code +updating the patch +Thanks Sergey can you please comment here with the jira number of the required CXF changes? That might be of use when we will need to port this fix to EAP branches +However of you can find a way to do something about it a pull request would be welcome +That seems opposite of what it should be +FreemanI dont think so +Im attaching the patch for trunk +ok private list created +seems like we obviously need to do the pushed Martin Traversos fixes for recursive Hash and Array in f and removed the related tags. +Can use findObjects api to achieve mass removal +After doing some more investigation it turns out that there is a bug here +Try with and see if you still have this problem +You probably invoke set somewhere in the code of one of your action handlers while your configuration indicates that you are using user managed transactions +Are end users going to know which jar the are in? Users want to extracting a jar just to get the conf out of it to read the description of the settingAs for what hadoop does +Committing +Ill try to make the fix available on monday PST interception semantic has been fully restored in custom Spring AOP patch +This is a JRE bug +Thank youAll fine +etc +DoneIf Nutch is not releasing soon feel free to commit your last release. +I can imagine this happening inside of builders and the now lets just ignore this problem but in the future we will be adding some sort of project file that can contain some kinds of overrides for seem to be passing but I will not commit until post release +The problem is I currently do not see an elegant solution +For the second questionI think we could introducing a new function of forced splitlog in + +If you run mvn package on the attached project you will see a warning from logback that is on the classpath twice +Uploaded a patch that uses tee to send the standard out both to standard out and the stdout file +Hi Sean Owen is hadoop supported +resolved +The command has an optional backup flag which takes care of backing up the project before replacing existing this means Roo will not automatically update tagx files upon version upgrades +This expression will need a bit of at which point this jira will be addressed too. +is the related bug so can someone close that bug as well along with this oneRegards bug as per martin and ashus comments. +Thanks for the review +It seems that asyncfalse works on firefox and ie but does not work on chrome +Assigning to Nick since all mentioned components now are built separately as CBI files updated in SVN for and including their source feature templates +I dont understand where you believe the bug is? Castor doesnt escape any characters the Serializer does and if you are not using the Serializer then no escaping is taking place +batch transition to closed remaining resolved bugs +x +Done by implementing JSF of a bulk update all resolutions changed to done please review history to original resolution type +Lots of people run releases wo the native libs +Is it possible to include correct version of archieve to the binary distribution +for new patch javac to check the ASF Grant +works in both and added that class to the class and register all these jars and still it dont goregister mepigbuildivylibPigregister mepigbuildivylibPigregister mepigbuildivylibPigregister mepigbuildivylibPigve tried adding slfj to and what not but Im stumped and am gonna kick this back at you having tried and failed to use it for taking a look +Cannot reproduce please if needed +fix verified on Windows XP jdk +works +This should be generated by the build in the target directory +thanks this record since is now released. +xxxxxxxxxxxxxxxx starting delivery msg to remote xxxxxxxxxxxxxxxx delivery success +resovled by using slfj +Thanks issue is fixed and released as part of release +Hence all this looks like a failure to update model to correct source folder +Applies only to trunk +Seems like the most useful information we can get would be if the NN is in safe mode or not because the JT can then go into a conservative mode wrt task failures etc +Thanks Lars +The other repo is one where we manage ofbizext access for those in the community so that explains the confusion +If we get caught up in another it could be weeks before we get it all settled out again +Page expiry is a big issue for users who are logged in invisibly by cookie + +All the combination of network services and network service providers that are possible in isolated networks are possible with Shared network as wellThis feature only enables services in the shared network in the advanced zone only +I doubt I have the authority to reopen this ticket but I hope this comment gets emailed to relevant parties +Dont think we have to take the further step of handling custom Number subtypes that could be looked at when the scenario comes up +Applied. +In one case the generated code is incorrect and anothere I am still trying to characterize +Usually the file storage should occur in a service not a controller correct? +Fix attached +Attaching patch contributed by Hiroki Tateno +Suggestion add a CSS reset at the beginning of +this is example from VS at least it is possible to insert this into left part of tags in the bottom place +The problem is that cache needs replacement mechanism which in Java world only done reliably by Guava and +You need to use a different like +Will give it another run +However it does not throw an exception so Hibernate cannot tell that the insert failed +Ive relaxed the assertion in our accepting a null target object as far as possible +And the patch makes sense to me +This also caused which I have is done in branch +Caused by changes in +Im not clear on whether there are additional steps here is it done? In any event looks like something we should consider to be finished before ships +Also reading the sections on the wikiIm having a little challenge when it comes to understanding but that will probably change soonJust wanted to let you know +I am looking for the the place to plug thereMy next patch will be to the trunk +Perhaps you could be a little more helpful here do you currently have a dist folder anywhere? If so where? etc +resolving issue in trunk +bpmn with both service tasks and a timer +I will try to check in the code for this task this weekend or next Monday the latest +In the clean up step the isolation level of the connection instance should be set back to the previous one +jdk looks ok +Thank you for the test project Nils +apply Thanks a lot! Best regards Andrew +Im downloading Windows Server R and and during the next week Ill try to take a look at this issueIn this way I can start to use my MSDN accountDIll let you know more soon +Moving to the next cycle +Attaching Andrews patch fromI am will commit it once Jenkins says okay + +The else part may actually paint over the earlier if conditions which could have been a problem in my case +The NPE is reproducible on and on all and releases +r in svn fixes the issue so that deployables declared with relative paths are deployed correctly instead of throwing a as we need to fix the issue title and description +John As far as I know is the only release on with a bogus md fileI raised this issue on a couple months ago and got no response +Also while working on this I discovered fixed a bug that prevented STRONG references from ever being located via this operand +Our profile is with the first patch +Ok got it thanks +bq +I have tested it with and it works with all of them +Rather no but you can try its a dependency so grab and used it instead of it how OGNL is designed? The issue with the space being scarce whilst the application is running is inevitable +Id trust the runtime compiler to do the right thing and inline the setter method +Main file +Not sure what side effects this fix is going to have so Im leaving this open for a bit +Build files have been updated +Hi Roger so now is it safe to mark this as resolved +Fixed the doSomething javado typo in r +that reads zk znodes and produces an URLIs that a new function +Was a transposition error in the test on my part +Sorry in rushing to commit I forgot to add author during the commit. +Wont exitWorker kill the fully? Like not just the task but the RS will actually stop serving log splitting +checked this against and its not failing +Enable monitoring capabilities for the new one +If possible remove this file once this issue is closed +Uh ugly I guess we should just catch an exception and display an error message in place of the feature type list +Just added a dependency on bug as clearly this problem would be resolved without any work if we committed bug +Still learning patchingjira +Dont we also need to adjust typetospecargs in order to include the limits for nested structuresAlso it is unfortunate that adding the limits broke compatibility with the old version of the extension module +The messy part is to recover the page instance after the exception is caught +I am not sure if you understand me correctly +See for a clarification to our Class Reference +Can you please provide some examples of the broken projects that in your opinion should work for update sources action +I made mcomponentInstances private again good idea and committed in rev +Its about merging there is a new and an old object for the same persistence unit name +I did try clean too but that didnt help +Please excuse my earlier panic attack +Looks good +Mergey provided in textAfter testing his solution I got the same working result with this patch provided +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Because of this the resulting app cannot be run +The attached patch moves the changes to that were being tracked in over to this JIRA +heres my benchmark is out the door adjusting on all remaining unresolved issues to by call lazy consensus that we keep delayedcommits on by default +Is this ready for consumption yet? From where can I grab these resources +doh +Ive been using your patch for a while now and it looks to work fineDo you think you will commit it +Not a Web or Wiki bug +Ill change the level to WARN and look into checkDirs +Values are trimmed on both sides thus in your test case the two keys are both associated with an empty String as expectedI do not see a problem with this. +Im sorry I mean no Ive got no specifil null issues Ive got the same issue of the other JIRA. +This can result in quite a bit of trash in the plugin when deployed into an application and thus adds to the SVN trash of that s all kinds of crap in there including and Needs to EXCLUDE all +Thoughts +Interesting what is the license if there is no license? Can we have a dependency to something without licenseBest would be to ask the author to state that this tool is under Apache License and we are fine +My guess is your indexed property is not setup correctly +Committed to trunk. +I had to let the memory test run all night before it would identify a problem of course by that time I had reformatted my drives and everything else +Thanks atm. +I thought it was pretty much done +Figured out a reasonable way to resolve the problems including addition of settability of the property. +bq +Linking the m to this +while can now be replaced by? there is also and there is the possibility to use and in +May somebody pls try this patch? Thanks +However this is still arbitrary and not very scientific +Is it really necessary to copy the parameters to another object? +Uploaded the zip attached by Barry to the repository + +Last patch was missing a file +You get the idea +to check if the error still occurs. +In Ant you would have seeBUILD filepath Error while +I already fixed this a couple weeks ago +Verified on +Changed the logger and committed patch to trunk. +An upside will be that I can add support for these queries to the Span highlighter +The potential problem is that when job transition to SUCCEEDED state some job success metrics stuff has already been triggered +Hope someone review and commit it +I think the above description is a bit confusing as Salesforce depends upon and not XMLSOAP +We now have a new datastore based on bindings that avoids the pesky requirement of manually building JNI bindings on each and every platform +requested test case sorry for the delay +Looks like the minicluster adds some users for each datanode +Trying a new one +Ok i saw the answer on the list +Attaching patch file containing test but no fix yet +Besides that maven just does not work that way +Committed to trunk and +I also had a look at the RI sources and there is no coercion in either +Not sure if these two methods would work in windows +in this could be achieved using ajsupports in inputs which will enable buttons on successful validation +Fixed and shortcut changed to CtrlAltQ to avoid conflit with Show view + will commit after tests +While I have stepped through the code with a debugger so I am confident what is going on more testing is needed +It timed out after minutes but spent a lot of its time waiting on to shut down +Either rejected or a dupe +My rest server came up with this change +Oracle model with DB source info +Attached patch that +Applied thanks +It seems to be our snag +Deferred issues go to the TBD EAP bin so we can keep track of them for future releases +I am using Camel and still having this problemThe HTTP headers are all in lower case +Otherwise its pretty much the same as the original patch +SeeI verified with this fix that the test no longer hangs waiting for user input +Drat worth a shotIts something to think about +Im for this too +Matt I think you submitted the patch +comment in as part of the release +This patch fixes the installer license statementsThe patch should be applied to both branches and trunkThanks +This bad boy is incremental but has literally no downside +MaxCould you please provide more details +It warns about using fixed and then complainssomewhat cryptically about missing widths and dies +Related are and not simply delete the block files that are not in genstamp format? This does not depend on upgrade at all +Thoughts +i didnt know that filei have created a patch that sets the last mimetypewhat do you think about that +I guess we need to update excalibur store +They would much rather limit the set of ports and I tend to agree with them +Rescheduling +Committed revision on branch. +Closing. +Thanks for trying to solve this +I could imagine an eventual lucenesolr subproject which would be ideal but until then its going to be in the wilderness somewhere +Thanks for the explanation +Agree that scanners should be passed along +Regardless this is no longer an issue with the version of the plugin. + resolved issues. +Log file from when crash in +have been able to reproduce and have a fix +Stated that the GAE Issue is still opened and Roo does not resolve the problem does anyone knows a workaroundis it possible to downgrade Tiles versionwhat should we avoided to useto get roo mvc project to run on GAE? I am afraid downgrading Tiles would not solve the problem at hand +Should now be fixed in the CVS +well Ive solved the problem locally by adding a symbolic link from our publichtml directory to distrepository +The most important ones are the above mentioned plus one or two for reading data too +I will check this issue. +Now the AM is getting an exception from the RM and so not shutting down properly +Added the new test and put test case into under review because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Place update ngp file +Ill roll back these changes and instead enhance the documentation a bit in this area and possibly add a helper class. +Please link the new Jira as a prerequisite to this Jira +closing issue as this is about documentation and usage and is the behaviour of Hibernate beyond ourJust hit on this rock i understand why it is caused and the wont fix but i would like to make questionLets suppose that books and authors are instead of and when i m going to add a new book i from the list of authors I would have to iterate over all the authors instantiate them and add to each of them the new bookthanks in advance +callbackWorker is never used and should be removed +If I can get access to the fields TFV in the URP I can just run through the profanities checking for each one in the TFV +Ive got an error in trunk version all is fine but on it is hanging +I dont have time right now to review the whole patch in detail but after a quick glance it seems like a very good job +bq +I split of the handling of the constructor argument into its pathpart and its query part so that the encoding is handled seperately +I am unsure how we can improve the error message here since theres no difference between this and the fact of mixing dynamic code with type checked code using some methods annotated with and some others not +David can you look at this issue? It looks like Piotr has posted an updated patch that should satisfy your requests +Now need to improve test scenarios and fix any eventual bug +got rid of old docs will leave other docs in wiki and set them up over time +Please confirm the issue exists with a relational DB ie try hibernate and hsqldb +I think we need some helps here +schemaA is not nullI confirmed that I am not getting the with the new patch but bugs and difficulties in configuring the job in my own code are preventing me from confirming that it works +But Ill need to research how this problem could have gotten past the tests regardless +I just went to look into this +AndreyThis sounds good +Logged In YES useridThanks for the analysis! Accepted +The fix looks correct +Adding it now +The above comment touches item time to service breach +Passing through sorry +However it doesnt always seem to work at least not for packaging type pom +AJAX closes toolTip doesnt work correct in AJAX mode +Adobe reader and Foxit reader can also show it correctlyI think the document is generated by creator it maybe include some new PDF features +Committed revision +The render screen function is recursive in nature and each render step will log every error +to trunk branches +Duplicate with +deploy +this way the score will be equal to the boostthen users could choose which one they want to use +Just some missing builder warningsENTRY MESSAGE Skipping builder for project +I believe however that we can issue a redirect and then have the filter set up the appropriate environment for us +Have a look at javaengineorgapachederbyimpljdbcandjavaengineorgapachederbyiapitypesto get a feel for how some of the RSMD methods are implemented for embedded Derby Have a look at javaclientorgapachederbyclientamto get a feel for how some of the RSMD methods are implemented for clientserver DerbyThen Id suggest having a close look through the spec and ensuring thatthe existing test programs provide reasonable test coverage for the API calls of interest +Maybe that is the problemI wrote my own component to display the list but I had to surround the body with hcolumn tags to generate the correct ids +for as well +Im working to build an amended copy of the book ASAP +The reference to the is kept in object instance +Yes absolutely +Attached patch v that changed the interface +To compare array contents use the new Groovy was just ahead of its time! And now has been fixed to make it like the trend setter +Fix invalid assign from assignation of this issue to the author of. +Fixed in svn revision OK +Added method stubs in instances do not fully work yet +The component works now with the programmatically approach. +This is more than enough for amazon rackspace +Updating the scheduling information after the scheduler has obtained tasks to run on +This issue seems to be resolved +Javas penchant for encoding all and sundry totally breaks this as it and +This way we have the ports configuration next to all the other configurationsIf you are looking for a more central configuration like one setting per all instances in targets well then we need to rethink a bit +bq +Stuart I forgot to ask the following is it really safe to use with Felix? I saw that this option is unstable and calls loadClass unsynchronized + fixes committed to trunk +My client would like to use from PIG instead of a fixed version I gave them in a jar +Compile dependency package dependency seems reasonable +sorry for the delay +Planning to add the following note under the section Working with System Virtual MachinesYou can configure the parameter to create a random system VM password for higher security +Bug has been marked as a duplicate of this bug +I just committed this. +Corrected the title +Thanks Abhijit! +Instead I changed the attribute to scopeprototype +revs + +To export the profile from one instance to an other you should use backup not permalinks +Zheng I have two questions How does one manually go through the metastore tables? What should be done to optimize the code so it executes faster +Re +Older versions of did not support Im using looked at it you just need to skip the database name in yor JDBC URLjdbcmysqllocalhost instead of jdbcmysqllocalhostclients +But patch spreads accesse among other modules such as loaded via dlopen +Needed a rebase +As I went to implement this I realized that in addition to other classes like and would need to treat the session as a as well +For example deploying to the local file system filehomeuserstuff +from me +Please have a look and make corrections you deem necessaryFolks As far as I am concerned that completes my mission +The attached testcase should pass if proxyuser settings workUsing the MR bridge API as Ive writtenverify the testcase with +Pull request soon +Connectors must be in that loading the connector class from whatever library directory is specified in the file is adequate to locate all resources for the connector +The next planned release is no other are planned +and with freemarker you should get a better message +Lets continue further discussion there for the more general case of tracking node instances across move operations +defer all X issues to +Attached patch to process per BDA +There are too many different problems described in the details of this jira regarding loadstyle +lgtm +Thanks for the fix! +Could we allow concurrency between compactions? I assume thats what makes things tricky +I have also tried it with but no the files has correct permissions +Sounds interesting +we can see in this settings page from the release the right list of plugins displayed +Then fixing it in trunk isnt enoughWill I commit the fix to the branch +fixing for now to let QA verify on JBTNo problems foundIf problem still persist reopen this Jira. +are using springs and ultimately Hibernate to go to the db in this scenario +ScottIs the issue mentioned above during start and stop of Jboss server has been yes can you please tell me the solution to resolve the issue +Consider following caseStandby receives a block received message and doesnt find the block in its map +Slava can we be responsible for this long running task? Im worried about the Tools Model Update job +I get this on as well +SNI support in Java and Java looks like a complete bloody mess +This occurred because the URL was used in the name of a temp fileThe URL is no longer used as part of the filename and deployment is successful. +Awesome Gregory. +Attaching a unified patch as for the precommit build to runRegards it seems that and were applied around a few hours back I created the patch +Forgot to close as duplicate please track with the other issues +You can install it with the following command gem install rspec + +Looks good +This is very simple to do +Rayees can check if this is actually fixed and can close is a unit test run by default during the build process +Could you confirm that you want to use the default Derby database as no DB configuration is defined in your file? Could you attach to this ticket the logs file of your Sonar web server? confirm I use the default Derby database on my Sonar Web I think my is correct +I believe the is the only iterator that currently has this issue +Moved to PREREAL to avoid interferring with for forum thread for reason for ve reverted the change since it causes multiple tck regression +At any rate I see no way around doing some work on in order to get the behavior you wantIn your particular case though you could also consider using Mimej instead of which should give you a full control over the process of MIME content generation +My main problem with that is leakage +Digging into the code also helps you understand how Camel works so you get some education as well +Hi Mark XavierI understand where you are coming from +As Todd says it is usually indicative of something that needs to be refactored or fixed or is simplyTests that use too much are extremely brittle often failing in test code B after after changes in seemingly unrelated code A in ways that require deep knowledge of B +These are files carried over from provious releases and not used anywhere in +pushed and +Not sure what to do about yet +I just committed this +Trivial patch +Added. +The code is in sandbox in the +That would be useful indeed +Feature added but needs ll pick for me this task +Hi Frdric Having the same issues with a Flex project +This case is not ever emitted by javac and as such almost all avoid compiling it in bytecode leaving the code to is unknown whether this affects the C compiler but this is a critical performance issue for since there are now two different JVM versions that we know downgrade optimization for these cases +I cannot see any reason +Luke have you made any progress on this? I think a script to copy over the files an event that changes the junit report dir only if there are any files there would be cool +The insane result is similar to which is around ss vs on clean trunk +from the picture I can tell that when going from base line to base line the numbers use pixels as height the text only +So I wasnt sure which way to get rid of the issue +Can you describe this more accurately as a step by step script +I will ask Elena to test this patch more thoroughly +Farah tried the latest versions of the server repository and quickstarts and it workedI downloaded the Maven repository and the EAP server and still saw the issueWhen I downloaded the quickstarts and tested it worked fineThis is resolved +Please verify and close +Strong data typing would prevent applying quantity precision to a money and money precision to a quantity. +And there seems to be an issue with the fact that a is not generated but referenced from another class +By including in both the panel and the page via head Wicket will not be able to filter duplicates therefore will be included twice in the output +should be fixed nowSNAPSHOT deployed. +I dont think you can assign issues to people without a certain access level +from me +Please feel free to not fix it +Thanks +Agreed thats an interesting observation +sounds like a simple and good solution +This suggests this may not be a Derby bug +No no I was testing for the bug on the release branch Im more interested in its bugs at the moment +I guess that this will be automatically fixed when the sparse vector implementation is deleted. +Just add several annotations and you turn your layer class into controller that can be exposed directly clients that send data over HTTP +No no I understand its usage and how it hooks up in meta classing I was trying to find out who is providing it on the users classpath I didnt find it anywhere in code +This looks good would you mind attaching a patch for + has a shortcoming in that is is looking for an exact match +Fixed in r +I will see what I can do butthis is a bit tricky with the current architecture +Good point +The default socket timeout can be overridden at the HTTP method level +If you want to control the timeouts for the RMI connections then this can be done via system properties +Thanks Eduardo for the patch +The patch attached remains consistent with that along the lines of the most minimally invasive behavior +Either ofthem should be changed as saxproject does not comments anything about defaultvalue of this feature +attempts are made simultaneously one each during deployment of CPVM and SSVM +But I havent tested this with latest Equinox version. +Also adds a warning to users +Jira is for bugs found in Struts not for getting help on debugging your application +WS guide is checked +Updated script that copies the server POM and JAR files as well +Do you have a machine where you can install EAP and try and reproduce this problem +Thanks Michael! +Makes sense +Oh I see why people ATTACH things +I get the same CNFE with this patch appliedSlightly odd to have it change my cwd out from under me too +Ive only verified this for a chroot of depth however +The same could be used for the remarkWouldnt it be better if we allow changing the dialog size factor with a simple gettersetter pairThat way the programmer isnt required to subclass +I see no reason why this shouldnt be possible so changing the default phase and will do the same for the mojo +Closing it as +Shall we add an incompatible flag for this jiraBecause adding a new block type is not backward compatible +Bulk close for +No actually any functional changeVerification performed only mvn compile because of patch nature +We shouldnt strip access to but I think we can reach a middle ground where the can define whether it wants a reference? Bean setters anybody? The same goes for any Schemas +Well that sucksAny idea whats going onIf its going to be a lot of work to make with then well just say you have to upgrade the whole cluster at once +I dont want to rush the review but I agree its worth getting intoRussell regarding the jodatime jar are those published in a maven repository? If so you can change piggybanks to pull it via ivy +For the test case I previously added a testcase that covers the cases when using and default delimiters +See where the idea is to introduce a class to make it trivial to link to context relative images +The update chain we defined is enforced with singleton thus the is shared among concurrent threads +I would rather see it done correctly all the way instead of just shutting down the node managerI agree +Attached patch +Given that the problems due to very small block is seen very few times my preference is not to add a minimum size +Add a test case in patchv +I agree with Kathey that we should not violate SQL standard for CREATE because similar violation exists in Derby for INSERT +Do you mean decrement number when a RUNNING attempt failskilled after the task is in SUCCEEDED state? Normally they would trigger their own FailedKilled transition that takes care of the decrement +sort the result and then with another hacked script print out cells that do not have a corresponding delete record +Does touching any of portals descriptors not trigger a restart +Attaching new patch in which I tried to take care of Dans generally looks good but its good to have clear logical naming +Rick will doc the danger in the first AOP chapter. +Patch is actually in +If you have an indexed property you may also have a array property but thats all +Will show next time I pub site +already provides the way to do this with its cache keysI dont think we should move things like fieldcaches back to it. +Yes you are right +Leave as it is +on Bada sdk the problem still supports bada +Without that it would be pretty confusing to a new Solr user who starts off with seems to me that we should use precisionStep for tlongtdouble fields just like we are doing for tdate +AFAIK this is possible using a custom but we dont provide any kind of user management for now +Seems like you are calling the getText method of the Binary object +I also would like to point out that property paths are not officially supported by Bean Validation for validateValue +While this appears to be the way to go it introduces too many steps for me to prepare the dumps thats to be imported +Id be happy to take your patches if you wanted to rework them to be improvements on the existing code versus standalone code +Am I the only one who can close the ticket +I think we can refer its experience +It appears there is a difference with the way handles composite subtypes for Readers versus the way it handles them for content negotiation +Any chance I can get you to attach a small sample that reproduces this? Im having troubles with my code editor in getting that character string into a program +Thanks RegisSome minor comments on the update the keys array should probably be of type the comment on line is out of dateRegardless of what approach Harmony uses from my investigation Im confident that maintaining indices to save iterations isnt appropriate for Android +It is caused by missing in libgen directory and there is also bad filename for +will be implemented post +In future Ill take care +This would work because we know that and both implement the classic interface +is it really trying to load the class literal instead ofcan you set the breakpoint and see what the actual requested class name is +The reason they have failed though is that i made findbugs errors a failing condition +What is more I may need to play some odd games to make it so that a DB output committer even works so that I can commitroll back the results something that the current DB output format does not implement +In the US at least it can be handy to represent stooges +And it should work for any gateway not just Exclusive +Santiago Ive applied your patch at +If I am using svn then it is ok to have the and also for it is ok to have the +Permaine are you waiting on something from someone here? Please specify and assign the JIRA to the person that should provide the fix please +as per the previous comment by Thilina +It is not uncommon that we have had to let it run for a while to reproduce the errorOne of the problems Ive had here was that the logs were not very useful +Stack Is it now or +HS is strongly recommended over HS and in fact HS is considered deprecated +As an experiment I changed the contract to return Object instead of interfaces and ran the tck which gave the same results as beforeWithout the changes I made to extend with once we had an instance successfully returned by it would fail the cast +Therefore the should be thrown out of CTF test suite +Can someone help review and provide comments over the patch +Here is the I find inside the dist bin folder after running ant in the spring module +all that baggage came from people wanted to customize every little aspect of whatever little markup the tabbed panel generates +the +Thanks for the validation Todd +META? The RS right +With Spring IDE M we shipped support for the requested feature +Oh Andrews super flakey ec hudson server is AKA +verified +A lenght response happened in the past when the response was really small though for the life of me I cannot find the relevant jira but maybe you can try a nightly build of and see if the response is still emptyAbout the second there is that Z character which is invalid and apparently cannot be encoded +The five projects are completely new and not yet linked to the svn +Closed all resolved issues related to our projects. +as the other Applier are used client side from the client moduleSo I guess is legitimate for them to pass any parameter they wantAt that point is the server part that must ensure is compliant with the specsSorry again and WDYT +Can you give a little more guidance as to the location of the fix was it in +Declare depenencies we were using anyways +Thanks Daniel +Sure I will rebase again and upload attach the patch for test run +I did some my investigation on this topic last week and found testing data streaming pipeline in hdfs requires fi +Need to decide if we want to do this or not +It looks like this may not have been done for is right Ive did not add any ports as I never noticed that this failed +What bits on the site did you want to add +Ill commit it once I figure out how to create an automated functional test that will ensure that memory stays reasonable +If no objections Ill commit to TRUNK +Ive cleaned up the code in the class loader provider a little bit in Revision The former version did not logcatch all exceptions and might result in strange results however I think that this does not fix the problem +I have the same locally so its likely my patch +The patch is finally ready for review appreciate the input +Here is what Im about to commit +Changing this issue from Bug to Task +compareColumns +Punting to +First attempt at implementing Sprint generic spectral clustering +The spring groupId should be removed moved to meant servletapi and in favor to to other sun about rhino js shouldnt it be js +Reducing the priority as this can be solved by importing all wsdls in the parent wsdl +Now sort order is actually mentioned by Data type assignments and comparison sorting and ordering! very much RickCommitted patch to documentation trunk at revision +I have done some research on this but I am pretty sure youll have a better solution +Patch for +Will need to reconsider +That is good enough +Not sure why this is in the database +With Lucene the problem appears when the current position is after the seeked term +Also Adds a test to Converts Shares readwrite file methods across local testsWill add in the MR side once MR has merged +This one didnt apply cleanly +same patch generated wit signoff applied +yeah the patch i uploaded was mostly to get me out of the woods +This should prevent any problems going back after soft upgrade and keep the spses in sync with the version moving forward without the need for a new stored procedureThe only impact I can imagine is that it may take slightly longer to boot the database on version change +Surely we have to check that once more in detail but for the moment I can confirm that there seem to be no crypto in stanbol +Output can be enabled by passing true or toggle the debug switch in theCurrently output will be prepended by DEBUG A trivial thing but I guess there is no println without arguments so you would have to pass an empty string to get a newlineThe code for println can be found in and the enabling of the debugmode can be found in I believeI dont think this feature has been used much +I will move forward to get to the point where using Puppet scripts for a new service or for rewriting an existing one is a viable optionI guess the interface should be similar to the jclouds statements with scripts stored as resources +Ill start testingPaul can you also take a look +That is looking kinda sweet ncanis. +Are they really necessaryI mean theyre only calling standard android commands anywaySome of them do not work because they require the AVD to be named defaultI am all for automating the painful task of creating new cordova android projects +thanks very much +Update this class also chokes on legal that specify the port number likehttplocalhost +bq +The goal of this Jira is to merely reinstate deleted methods + +Thanks SteveIll create a another issue for making visible also to users for closed releases +But I dont know about Anakias support for +Failing tests should be more seen as a warning and we could adapt the limits +A lua context can be used by multiple client requests with lock protected and wont be blocked by I suggest TSLUAMAXSTATECOUNT to be or more +I cannot fix it now due to time constraints and the lack of an obvious solution +Thanks to Ted Dunning and Patrick would think that these sorts of commands would be better in the configuration fileAlso how would you specify multiple subnets? Can this handle IPV and IPV +The parts in has already been there in my original do NOT apply this TCK in the Apache production repo since it will fail currently +Ill shoot an email off to the for nlogj about jmxtools +I applied the patch + fix for compile issue updates to test case to use datetime type +Closing this issue +Kabir I believe this is done +We have seen this happen with release artifacts too +Thanks a lotBest regardsRichard +Im having difficulty reproducing the bug on a smaller scale +My proposal would be to leave optimize in Solr but make it a +In r I reverted the part I wrongly introduced in r +The Fulcrum Intake component is about to be released soon +Block compressed data clearly needs to be uncompressedIn a lot of situations this makes block compression undesirable +Ill apply the patch +Thanks. +All validators are registered in extension point +agreed. +Fixed and committed on trunk with revision +I tried today again with build and iam seeing this issue in this build +I will commit this patch today in some time +Fix for this the version that undeploys all apps and stops all resource adapters +It seems as if the default machine doesnt match correctly for certain characters but I have no idea why so far +I added chunked transfer encoding to the list of ibrowse options and the problem went away +From this information we will write the final release notes for all issues +I dont want to discourage you from getting involved but can you please start bouncing some of these to the users mailing list first +Denny Ill wait for Max to chime in on this one to clarify what he means +We already do this for DB and Sybase +I added that parameter to allow validation support to be optional +I am unsetting the patch available flag as the submittter has indicated that we should not apply this patch and wait for another patch +duplicate of +we can worry about this should probably just deprecate these super combo tags +Sorry for the bad code formatting of my last comments +A lot of projects have asked derby to work on this with no results yet AFAICT +A possible correction for bug you able to test if the problem exists with maven I agree that this issue is really annoying but for now there is no version release scheduled +Andy Can you try this file +are too verbose in Robert +Yes +Feedback was received long ago +the one for +Asla does functionality works now as expected? It should because problems with sbs notification was fixedThanksLibor +Pete wanted the name shortened +I didnt think it would fail the test I was wrong +Affects versions is wrong +The name in your example is id the label is maddog +the vice versa case linux client windows cluster is not handled by this patch +Updated patch test backported +Hi JamesCould give a few exmples of what you mean +Would suggest revert until we hear back about Varuns effort especially since the man was the one to open the issue and was the person who figured out root cause +Added if condition to check estimated +Moved from to just after merge of old branch into TRUNK +Tab issue in the last patch +Lets just say that in some data stores previously we stored data types as they were instead of converting everything to BytesAs you said we saw the light and now everything is persisted down into Bytes within Gora +The second suggestion also had processing errors when servicemix was startedI believe you had the connectivity to IBM MQ working in your previous posts +Feature request to support the use case in the issue description +except for bug fixes +Freddy I have feeling that this is a duplicate of and all this already was fixed +And your patience with me being during the last few weeks. +Assigned to Sarah M +You could put the length limit parameter into the class if youre so inclined +Its an infrequent race condition but its definitely happening often enough to be a critical problem for meI am using a with +is no longer used by +There is something wrong in my patch +WID wants the directory structure to be just right and unlike Maven its not flexible on that at all +included typo +That issue was fixed in the release +The message sender sent messages the first message was sent before seconds then the message receiver can receive it correctly +I am with Daryn here +I mistakenly added this to the form directory +Thanks Aniruddhathe patch looks good +I think I just fixed this in r +Thanks Nicholas +Ill carry on trying to recreate with extra debug if you get time a run with debug enabled on the tm would be recreated this found the issue and fixed it +It is not meant to signal conditionality or optionality of some construct that is included in the messages +Im not sure that was the problem in I really think the issue there was the race on doing a round of voting in lockstep +has been marked as a duplicate of this bug +That should then resolve all missing me know if that helps +I think that I confused you by my comment +Within you always have to be careful about instantiating classes using reflection for this very reason +I verified the fix locally and checked it into +The issue seems to be gone given the patch of is applied +On tomorrow were going to release that should fix remaining issues you might be dealing withOnce well release U can port this upgrade to your branch and see if theres something still missing to be fixed + +Fixed logic to only return the default value if the itself was not found +If you want to create a new subtask on the branch to make it less generic Im happy to review and commit it. +I left in the log fileI didnt get the broken pipes on the VM but I still go the same behavior in terms of the directories being made but no +Closing all tickets +Has the issue been identified or can the error be duplicated +The issue should be fixed in Javassist and the issues have been resolved +That will retard testing of Axis by early adopters +Ok thanks. +You were indeed correct +Cleanup is definitively needed in there +For I made changes to log it and then close the client socket and resume the loop of waiting for the next client request +And I am able to login with the admin I havent used the runtime configuration so Im out of ideas +Showcasing SELECT statementusing static service API approach +Other than that it is expected to resolve some of the issue that have been mentioned as optional tasks that depends on your capacity and time availability +I have attached a small project that demonstrates the problem +I have disk space and bandwidth as does the majority of the modern world now +Committed r +removed and replaced with and which areindependent of the old harness. +New patch for trunkI also did a doc build it was successful no problems +I have attached a patch that fixes this +AwesomeThanks Ning and Carl! +These two Command Messages get a free ride everything else is blocked until the user is logged in +just committed this +OK maybe the failure was a of the other failures +bq +rulezsorry but which version do use +Patch fixes the problem my tests were seeing +Thanks +Oh I forgot to mention I want the min and max heap settings the same to reduce the overhead associated with early garbage collection and heap growth +see comments +I will move it to srctest +Cloning this but cant reassign fix version to this is broken in as well would really like a fix in if you running a snapshot from January ? Does the problem persist with a more recent build +Do you mind adding info about this field type? +Heres a test case to help keep an eye on it +Thanks for the feedback +Commited to be merged into branch +org awestruct sites is not mobile friendly +Yes mostly at startup. +And since Woodstox has lots of existing functionality there is a good chance that some combination of existing features allows working around issues +I think this bug is related with jira +Reopening the bug as the Usage server still doesnt start +Thanks Kim +still unsure of exactly what this one means +Fixed findbugs error +not a BIG problem new patch to +Marking as this feature needs to work when we ship +I do need to serialize maps with Integer keys but extending the reflection to do that seems fair +Hi VivekAre you planning to the patch +I actually need to reload a store +please review +This should be fixed as part of the new editing model +Thanks for your patience +Currently the version is always passed to that function so the default never gets used but I guess we should update that to default to instead +If the option is specified and stdout is the destination the code also prints a warning and continue +This is not a problem with the Add operation +Feel free to post more details and we can reopen if necessary. +Closing this issue as per the comments from Kishan. +May as well answer this +Fixed the sorting issue with groups and inside groups when a sorting value is null +Are you using a Text editor instead of a groovy editor and does your project have a groovy nature? Also do you see any exceptions that might be related to this in the error log +I just wonder why you did not put it yourself under your Confluence profile any reasons? This question because its less work for us to link and possibly move them inConfluence +It is possible that actually solves this problem +thanks ystein for the review +Ok can you then please provide a few alternatives for the banner with different designs +There are a few objects which are supposed to span a single actionrender +Denny can you and Jervis look into this issue to see if its still a problem +This wont effect Trinidad or Commons +Can I ask that a be thrown instead of an ErrnoEACCES please? Its confusing to report that privileges are wrong when actually its an implementation detailId offer a patch but Im afraid I cant identify where the specific exception is being thrown for this case +It would be nice if there is a way Spring can dynamically change the log levels of the underlying logging implementation based on the environment the application is running +The type of the value is still date so any clients trying to access the value will still get an errorThis helps with repository level operations that need to be able to read bundles without actually using the valuesSo just having old invalid date properties in the repository will not interfere with operations any moreDealing with the errors that come from reading those specific dates is still to be handled by the client +Please provide your opinion on this +Bard I see our comments overlapped +Its difficult to reproduce on another machine +I didnt pursue it furtherI ran the k docs benchmark and the grammar change has not noticeably affected the results +The current documentation phrasing is meant to point out that if entity Address is embedded in entity Company then select a from Address a will never return the Companys embedded address +Hi allI am working on this issue and will provide a patch shortly +replace the test implementation in with a regular scanner iterator +thanks +Adding the EDUMP switch reports that the error is in during an execution trying to read the character content of an +Ill post it on successful test + r +Thanks for the review Stack +As we make progress moving the plotting code in the same direction Ill definitely keep what youve said in mindAndrew +Also another important point missing is the Hibernate Event System +This property is recommended for features with welcome pages that show better in the context of the features views and editors +has a lot of new features and my hunch is that there are going to be some bugs that shake out over the next few months +Committed to trunk. +Released. +Will commit on mondayBharathThis caused a NPE when building the system tests on a couple a my machines +If it is duplicated and the duplicated on is Closed I think you should close this ticket too. +Typo correction formattingnaming fixes and consistent null handling in the other classes in +Im pulling this into the release where Im reviewing this code anyway +I did the effort to make a new workspace and import the process definition I posted +Hmm +a project for preproduction this issue and a patch +Issue with japanese characters is caused by missing japanese fonts on the machine +Where does one get +yeah I missed that I thought this was just used for stuff +META +and should also work +Nope +The original code used which did not serve my needsThe TODO refers to way certificate look ups +patch applied thanks! +My thinking that reflection mode only would be used and not byte code enhancement as described by Bob +Will merge to and push out releases for both branchesbulk closing issues that have been resolved for more than a year. +Fixed some wrong datasource plugin replacements +Should only one break be created by the lexer? In my applications I solved this on the but I am open to any suggestions and improvements +Makes message friendlier +Yes Hari let me take a stab +Patch to fix the XML comparison utility +committed in trunk and branch for branch revision for trunk revision on. +Thanks Divesh Pranay for the contribution Committed at rAshish Vijaywargiya +The next message that actually gets delivered is then the next one and the assertion is is to subclass with a new class that allows all excet currently executing task to be cleared without shutting down the +The class contains two tests one tests an object with a field annotated with and the other test a similar object with the same field annotated with +I had a quick look at the tests in derbynetmats and derbynet seems like a good candidate since the problem appears when using prepared statements with the +With deltas how do we handle the case where a namenode request to remove a block is lostActually I dont know what happens in such a case now +I only added a comment +Should this bug be closed or should the target of be moved +I have added this constraint to +As soon as I have them ready Ill post them +Support for configuration inheritance including dynamic form beans has been added in the current development version +Removes ERROR logging from when output path is null +It produces an empty byte buffer as a message delimiter +Work completed +Patch for for trunk +closing duplicate jira has been already verified +So indirectly when we claim Sqoop would run on Windows we know what hadoop distro would it be working with +Ah got it +Now the navaigation to a low level leaf is much easier than with the previous site since the navigation tree on the left can be browsed entirely without opening a new page click on the arrow on the left of the page title and it will show its children and so on so that you can open a low level page from the top level one +Patch updates flume configuration for DFO reliability level by making a sub sink tha twill never exit with an exception and just backoff and retry periodically +Merged to branch at revision +This issue is no longer valid +Sending srcmainjavaorgapachesshdserverSending srcmainjavaorgapachesshdserverSending srcmainjavaorgapachesshdserverauthSending srctestjavaorgapachesshdutilTransmitting file dataCommitted revision +General error handling in server adapter needs to be strengthened +Prasad I left some comments on reviewboard +Add test cases +Numeric positioning seems a bit brittle unless the standard components have a locked position +An element collection can be in an entity or an embeddable objectJPA query support for element collection within an entity is checked in under revision rbut element collection withiin an embeddable object is not available until we first add the support for navigation thru embeddable object a new functionality in Spec +should reflect the oldest version of Hadoop the patch has been applied to +As mentioned was trivial to for the suggestion +Ok I made it into a stand alone plugin +I will try again with Java and see if the behavior changes +Ok done its in commitsJust testing again locally to make certain all is ok. + +Thanks for the patch. +it should be confirmed if has the same problem or if it is new in confirmed that this is a bug for as well so I am adding as fix just fixed the method selection bug in and now your example works as expected +The patch has been created and appliedPlease verify. +see what you guys think +I havent able to touch and modify the patchI was busy on other things is it ok if you can continue on thisThanks! +Thank you again for your patience on this one +xtrunk +Attached a new patch +Using the canSign parameter to false with the unsign to true then we can double check that no jar will ever be with and +It may take several days +More description for property as relating to running in distributed mode +I have written test and implementation and have created a pull request from githubIts my first PR so please have a look +It is unable to recognize elements from from the super classThe error message is Unexpected subelement my +Works for me on iPadiOS I dont have a Nexus Android to test on + +If spring is included in the application server it would make it much harder to upgrade spring and could cause problems for users that are bundling older spring versions +Can you clarify for the EAP release notes +codes look just committed this. +I know there are a few others that you logged around fault processing +We not longer support running ant from any of the subproject directories you have to run it from the root directory instead +Just in case Evgeny do you have any idea how I can test this ticket? m really sorry Evgeny but I dont see any difference with or without activating this parameter +Further adjustments in SVN revision . +I may introduce them in my fst reimplementation and see what comes out. +I believe a good designed generic event handling solution should work for both use cases +Have also had to use the with deps distro before due to corporate network s a nice to have +Hi Dagwhat does the above comment mean? Has this issue been triaged +Adjusted to recent changes +I have not tested it with continuum +Since Maven does not support profiles in a parent POM this improvement is pretty hard to do +Using Annotations is a nice compromise between using Interfaces and XML configurations +r on trunk +Linked issue primarily in terms of client timeout and cleanup handling + +Felling a bit ashamed because in my snadbox I had the main clas sof the jar without dependencies and tested dependencies only when running single class +This is the autocomplete feature for the lookups +Ill run the test suites again today and see what happens +Please take a look attope level and for thesame +Since the constant score is taken from the query boost idf issuescan be dealt with elsewhere +Roo compiles OK. +Same bug with xwork No problem wit xwork for reporting patch commited to trunk +Provision from somewhere else lacks the seal +I didnt see any place where that could go wrong by changing the perl substitute command to do so and the compiler would find such possible comment errors anyway +Txbq + was not so easy as I expectedCache key should contain locale which is taken from Component so when no Component is given cache will be populated with string from first requested locale and will return it also for all other locales + +This looks like a reasonable start +records failures during initialization of tasks +Closing the ticket. +This will likely rise to be a blocker soon +One note I know that for example contains the configuration but IMHO it should be centralize in one for SOAchanged it to use the and files therefore added steps to build process to exclude the modeshape users and and roles files and remove the policy from the and added a step to modeshape installation to insert the policy to Teiidadded a step during build process to remove the from the and added a setup to teiid installation to add the policy to Didnt change Teiid to use the because these users represent the JDBC access to EDS +Patch committed in revision please could you verify that it has been applied as expected +This is the primary reason that the log file explodes with messagesTo fix it I introduced a special exception called which tells the rendering function that the particular error was already caught and logged so it can just pass it up the chain +Now prints out that it was unable to determine iterator type +Initial patch for backporting snapshot to +For example queries involving sort by and order by can become largely indistinguishable in complicated queries in particular if the sort by only requires one reducer from the point of the taskoperator tree +But that property currently defaults to so that it doesnt have any effect +Fixed please test it for me +This only happens initially though so it is unlikely to a problem at least not for long. +Thanks Narayanan +The reason for this is that generating classes in a build requires use of the plugin project but the plugin project relies on the tools project so we couldnt add a test scope dependency of the tools project on the plugin one as there would then be circular dependencies +It is ready for review +Pull requests sent +This patch synchronized all HA nameNodes display names on UI side including +st run JBDS splashscreen nd run Eclipse splashscreen +Rob EXIN is dead +Patch +However your solution doesnt look bad at all +Well see how it goes +Our source geotiffs are in espg and we are to in openlayers very similar to the thread you referenced but changing to nearest didnt seem to do the trick +For some unknown reason this seemed to make a big difference allowing or more clients to run with invocations +Rename the properties of User managed bean there +We talked this through on IRC and demonstrated that running the jruby command does not use the current shell +Scheduling for We need to verify whether this problem exists with the codebase +Luke Id be happy to maintain the captchaExcellent +bindings fileThis is actually resolved in some stuff I did yesterday to simplify some code. +Moreover Spring already parsed the xml after all thats its job Any chance the method will be restored? +I just committed this +Hi AnuragI have a question about in general +I will take a look at this issue when and resolved +The use case for this feature is that it may be convenient to manage multiple shards in an embedded instance and do a distributed search across them +May be we could correct it and apply the patchThanks +At least what does the field mapping for such a field look like? +Thanks for the patchI slightly modified the facelet documentation part as no viewhandler needed in vanilla work guys +Applied in r +If I switch to another editor then the views are deactivated again +I just updated the wiki page of. +I just compiled our patched version with Suns java and it compiles just fine +Thanks Rongrong +These fixes were delivered along with and +those exceptions happen because all the nodes try to create the tables and indexes at the same time +As mentioned in the description the jmock tests do not work and therefore have been annotated with Ignore +This patch obsoletes +Thats how I run against. +Thanks Mike! +Attaching the patch file +We set the basepath to and yet all links with basepath in it start with a +bulk close of resolved issues as part of JIRA housekeeping for release +Would like to see the patches soon could be useful to test the book examples +Grumble. +Closing old issues +I had written up a big thing about how I dont see why its too big of an if but then I ran in the simulator and concluded that you might actually not be able to compute the new options at call time if say you have something scroll in the background while the popover is up +You can use the libs in the upcoming release as well +Good question and the answer is that it should also be removed from Branch I verified and it was already in EAP as well and should be removed from both versions of the removed +of columns do not match the no +will not backport +Fixed +Thats not enough to have a checksum yet +Are HTTP sessions are matched between threads in the client app and servlet HTTP sessionsThat is if client threads walk the same sequence of mock responses will the Apache Commons simple and connection managers vector the client threads to the right sessions? If not the clients will get a random sequence of responses and some of the sessions will run out of responsesDo I have to turn on something in the Servlet +If you open in VPE attached and youll get the following styles +Lets see what performance gain we get +How do we deal with thatOn the other hand maybe a small refactoring would give us a Main somewhere where it could be an option if you want to take it on +Verified by Spark. +Calling discard on an object just because it fails validation is not a suitable solution +The changes look good to me and I think this patch is ready for commit +I will get some clearification +All unit and integration tests pass +It is nice to have a patch but without the changes to the tests the tests fail +I attached it to the ticket separately +Fixed in SVN r. +The Screens affected areapplicationscommonextwidgetofbizsetupapplicationscommonextwidgetofbizsetupapplicationspartywidgetpartymgrNo occurrences of sessionAttributes where found in Forms +Ill get on in a bit and maybe we can discuss further there +Thoughts +Im with you on finding the culprit but I believe as long as we RPM the jar this will happen +Looks good to me Alexandre +Of the two options the one with the best schemehostport match should be taken +Thanks. +Ok then just for trunk Thanks Ben for the heads upit has become a habit to always just ask for approval for branch and trunk we are pretty keen when it comes to backporting functionality but the GML core changes for required for GML support were pretty big and we are not quite brave enough yet +We can keep these around and tackle them when its needed in the correct way +Done behavior of the plugin has now been properly documented. +Maybe its an idea to start with a smaller subset of the data and see if the problem still occurs? Cant think of anything better than that for the moment +In that case we can start this NNRM as activeService instead of commonService +The proposed syntax actually seems somewhat confusing to into select bid as from Summary select price from what I think I am trying to convery isinsert in select bid +The namespace is not an issue in this case the just came from some strangeness in my environment when I originally ran the test +Please test it for me thanks. +New sourceEncoding parameter and updated to honour source encodings +Updated patch +I was experimenting with the list and the variant I liked most was the increase of the leaders font to and adding a title Project Lead to the link +Thanks +This issue is resolved in EAPEWP GA +I reattach and check the Apache grant flag +Thank you for reporting this I fixed in with r +Maybe having unlinked files would be a way to keep them separate +Moving to as not in scopeprogress for Fix For to Unknown on issues not targeted for this I believe its fixed the Derby store in this time frame was using non transactional connections to the db and as a results its behaviour was basically indeterminate. +Hi ShawnThe response wrapper IS needed because mode so the argument passed in can be modified and send backBTW the WSDL is provided by the WS Stack or simply in our jar +Also list of duplicate points is added to the error message +Committed this to trunk and +Does anybody has an idea how to test this bug +Updated patch to apply to revision +I have ported along with new tests in to branchMike you can consider applying the fix to other branches or close this bug +Looks like a failed transaction is not recovering properly +Great thanks! Nevertheless but maybe as a feature now it could be helpful that such configuration was available in the administration web gui +I committed this +Burr did it make any difference for you on restart of your machine +This idea sounds good to me could you provide a patch +I have a client layer that can do that so it works but it is complicated maintaining state is messy and the vast majority of requests are first page requests so in practice we almost never use that feature but instead do exactly as is implemented here and request the full document count from each shard +Therefore I want to have an option that I can control this behaviour +Regarding issueI think this issue is no longer valid after taking a closer look at the specs +Are you sure that supports the Account Usability Control? Do you know of any other controls that suffer from the same problem ie that are search result specific +I think this is wrong as the sender expresses that he expects to get the correlation id back that he set on the request message +From this information we will write the final release notes for all issues + +Please try this patch thank you +It comes with a very basic unit test which borrows some code directly from some of the Hive RC File Tests +Could be reasoned from the circumstance that my project name! my webapp name + +Moving issues that arent going to make it into to integrated the searchable plugins search feature into the scaffolding templates +Resolve above problems for ADB +I just committed this to both TRUNK and +Can you clarify? I dont exactly understand what you mean +Ive updated the HDFS pre commit job to build the common native bits firstThe reason we didnt catch this here is that the Pre Commit job doesnt pick up on the changes from the uploaded patch I filed for that +Created for the same +So I guess I now understand more about GWT roo but this behavior seems like a bug if folks do not want to follow the +A log file with a lot of user actions would contain a ton of these structures each with only one field set +I attached a draft about it and also add some logging messages to inform users that there are more than two requests are processingbut I dont think the fix is best +To give you an idea an implementation Ive tested this on the patch took a ms request downto msThis may seem overkill to most folks but is critical for my needsFrom load testing I havent noticed any significant over head in creating new creation seem light as well its only processing the request that appears to be the mainwork horseIll make some changes now that is added as I dont need the overrides +This fixes a comment in the earlier to run hudson again +That may be why someone said every field should be optional in PB for flexiblity +Its quite easy to forget adding public to an inner class declaration +Performance is not terribly important since the right solution if performance is the problem is to ditch Microsoft AccessWhile it would be slightly nice to have support for paged queries with the adapter Id settle for simply disabling it +Thanks Deepa +It looks good +I only have one PPC machine and it runs OSX do we have to make it so I have no means of PPC testing +Yes please attache a Vetoable ll try and write up a few test cases for this before I commit it today or tomorrow +Output of might not be syntactically correct +Can you just look at the config I posted in the message earlier +Just an option +Imean when talking to the real IIS works just fine butthe patched version fails to authenticate +I will do a run on linux with the same setting and if I see no issues will go ahead and commit +HiIf the intention is to merge default containers configuration with the selected container then i think your changes are absolutely correct! Anybody to validate +First version first release is +I made a change to the for the plugin and now keybindings are being applied properlyIt looks like a change I made a couple of weeks ago inadvertently caused keybinding conflicts between the Java formattingindenting commands and the Groovy variants of the same +Agree +For trunk its correct Jacques We can start indexing with and then its works fine +Mid March may be when this can be done +Im attaching a second patch and with that correction + +XIRCLES forgot my password too had to reset it as well +Really it passed on my platform +Try taking those out and see if you still are having difficulties +Ran tests against the branch +Running the benchmarks but they are tied in with integration tests so they will take hours or so +Once is resolved well be able to fix this issue here +Can you please confirm + +Could you provide code sample if problem still here? Maybe I miss something. +It is fixed on the trunk with a lot of refactoring done around test directory set up +Marking as fixed per Stuarts advice +for for the first one +There is no difference between these project examples in JBDS and JBDS +Thanks Camille! +The same SPI is used as uses so I need to see if its a problem there too and if so why we havent heard about it +The of setup and cleanup tasks can be and +Its just philosophical question +be sure check the box to license your patches to the ASF via ASL otherwise we cant incorporate them +Fixed in. +Implemented in all resolvedfixed issues of already released versions of Roller. +If we want to leave it like it is the patch is not neccessary at all +Thanks for the prompt answer +I hadnt realized so much had gone into the python based shell +Bulk closing for +Im a bit confused +I think Ive got a scenario for the giving the NPEIf the database holds for hours JPA will finally give the but in the meantime the socket has and the protocols sessions is expired as all it attached fields instances giving the NPEWe probably should better log the exceptions in the JPA layer +Verified with JBT +needs to be committed or rolled into the patch for this bug +committed w change from sort to be this merged to +If you can force gc to run you dont need the log +Committed revision +to +Mike has already applied the patches to branchUpdated patch applied as rWorking on trunk updates now but seeing test failures in the new +Chris do not make the patch available for non trunk patches +This will very much look like a deadlock +Installed JON binaries from CSP and issue is resolved +This is most likely an issue with your Your controllers seem to be designed for +Applied fix srcmainjavaorgcodehausmojojspc file data +Any thoughts on how to hide there +Should it not be up to the user to decide whether she wants less codeduplication vs +I have the same issue and created issue with the log file to reproduce attached. +Yahor please could you verify and close this Jira +Sounds like we simply arent able to identify the war as an EJB module hoping the rest of the code which uses the proper finder will work +Too bad there isnt a +I think the approach you proposed is a good one +Our target platform currently contains these on which you can depend +Ping +I have been unable to reproduce this issue and have not heard back from the submitter that it is still an problem. +is similar to it there are no outlinks obtain by getOutlinks in Domcontentutils then the outlink extractor is used +More test improvements +Thanks +Thanks Cameron remember to update +should be in off release +This is not really a problem very niche usage scenario. +Either way we mark it as provided +Verified in ER. +They have appropriate dependencies to make sure they get deployed in the right order +Done as described +All components define their bundle lists now. +We cannot support any interface beyond Comparable for primitives in both backends without significant work +No in some time. + +The exporter new syntax patch +It gets harder to maintain these documentsNevertheless if you want to take a crack at improving fixing it I will help you get it in +That patch only partially solves the issue after all the discussion and analysis +it be possible to backport this to Cassandra +Of course the + +I think the currently creates a new socket for each block it fetches +Even a performance boost to XML parsing is not likely to give any visible changes to end user performance +Buta the bits built in the workspace the bits on the nightly site the bits on the composite site b Tycho will have JUST finished building the bits so why wouldnt it check +Assigning to Ivo who needs to discuss with Jonathan when and how to do this issue as it is covered by and. +IMO this is something to figure out in the stage plugin +Hi LauraThere is an underlying Remoting invoker object connected to a particular AS which maintains a pool of sockets +I did not manage to remove the indentation of the first element yet +Release closing all resolved issues. +The complete patch is available now +Where is the current code base hosted? Not in CVS any more right? Can you give me the right direction then I try to have a look into it +Attaching the patch against trunk +Which version +Uri Is this something youd like to take? If so please assign it and perhaps move to an appropriate fix version +HiCan the Management class be placed in instead ofI am looking at this purely from module dependency point of view +Bartolomiej Sorry for not being clear at first +But the warning on the top is displayed always +Thanks Jacques +Regarding your last question yes I think that could use the JDBC implementation even when running under an execution environment and virtual machine coming from a Java distribution +Also the wiki should be updated with this new implemented need to add the wiki all issues +now there is a script that manages running of the script and subsequent remote shutdown +I will make the changes +Done +The linked to blog marks performance with collapsing as times slower than withoutPersonally I dont think this issue is ready for trunk +I had trouble reading old compressed using the new code with similar kinds of problems +change and some reformatting is committed to the repo +I have made some changes in upstream that may have fixed this +Works fine with of a bulk update all resolutions changed to done please review history to original resolution type +All patches were applied closing +Patch committed to trunk. +Will somebody try this +Committed revision which might fix this issueCan you please try with Cargo +Ive just committed this +So Im closing this issue +Moreover in general IMHO user name queue name application name and application type should be regulated as well +The hashmap hash code implementation has changed which leads to a different internal sequence of values within the map +For daytrader doc pls refer to the readme files in the source code root +Patch removing Apache snapshot repo from the grandparent trunk r branchx think the problem is now fold With your commit you disabled snapshot artifact download for maven users outside Your commit did not change the POM file in so maven still downloads the internet when resolving artifacts +Otherwise please open a thread in forum have community comment on it +not a bug in http client +I agree with the comments this is not a bug +So I assume this is only for projects +Ive got one more draft here with the smallest of tweaks javadoc spelling errors and one perhaps one or two other tiny things stuff I just would toss out rather than merge but are you doing anything here right now Hoss? I think not at the moment so if thats the case Ill put up one more patch before you grab the conch back +Updated patch to address Sids comment about the exceptionKaram pointed out a typo in the constant +Trivial change approved for +Best +jar we can only add items to it for instance how we added Incremental Publish and Full PublishThey are three actions two actions set Image Descriptor in theirs constructors +Thats correct +Ill have a look +proposed in +Its only when you install everything that it breaks down +Initial implementation committed in r +Or I should have not updated the test on MRI repoYoure changing master repo so fixing by simple refactoring is good for that +Not sure if this is actually a general mvn into tweaking the generated artifact type to see if this makes any difference we have to be careful not to break existing builds thoughIf theres no safecompatible fix then Ill close this as WONTFIX because the safest workaround is to use mvn install or mvn clean install +thanks. +Id also like to request some method to list all the folders currently being scanned +Heath any update on this? +We are planning to add a construct for accessing the mapped URI which will make this change possible +content type could be set for facelets only +Done the descriptions are a bit more consistent! +I thought we agreed on bumping version numbers so all these issues with builds for jbpm and jbpm being different would go away +Details at comments +Fixed for +Otherwise on the fix from me +I still havent seen any good analysis on the size penalty that goes along with the performance increaseI agree +patch for makes juddi jars into bundlles so we can consume them +This is by design +We want to work towards being able to persist in a database but if we have to do it in files what is the location youd want it to be stored in +Im currently investigating if my code is somehow causing this or if this really is a bug in the maven import process +Strike the comment about the target looks like the home directory was set incorrectly +This is still earlydays I need to do a and add more tests +We can try to set the following parameter to make the checks even more stringent for idle connectionstesttrue This is one param that can make checks for idle connections we should try setting in to true +The current trunk compile and startup fine with JDK just tried with version. +Set ignorefalse as a at +In particular is needed to produce the files +The patch suggested by him seems to solved it +If we need to move to bean I think it should be a separate JIRA. +Closing this issue. +Just wanted to adjust the affected versions! +Patch +Patch with test added +Its a reported bug in the release plugin but I dont remember the JIRA key off the top of my head +Should be resolved by +Do we have to have this method set? Can we not pass in the data on Construction of and remove this method? Make the immutable too? Otherwise patch looks good to me +The one possible solution to this feature request +in this particular integration test we know the number of rows +protected field and its setter is protected there should be no injection +It is okay with the tigris skin +It most likely will happen on Apache release as well +Im told that does what is needed +please attached files +Not sure what you mean about we are currently on and are not planning to upgrade +But if it is a territory based database then Store might use the Collator object if it needs to database recoveryI think we can check if the Collator object can be created or not for the given locale irrespective of whether it is territory based db or UCSBASIC db +Hi PaulApologies for the delay in replying to you +A test case and a suggested fix attached +Is it not better to validate the project name and tell the user if its bad +OK it seems like things are working in distributed mode now +Committed to trunk +Who can give me some suggestion +i am using Firefox +I wouldnt go into merging or overlaying etc +Attached wrong trunk patch +Good job +bq +Looks good +region this code will actually search at least the first row of the second region too +The permissions object I inserted in the last patch now has three integer fields +A table could have of widely varying sizes which could make map reduces run poorly unless is splittable +to change the naming is no problem +Otherwise it compiled successfully for meCarlos if everything is okay then close this issue +Could I delete it in M +Did pass based on your patch Jon +I think this looks OKAny comments +I think in the case above it cannot cause the page to break because there is no simple way to fix it +It took some time to actually eliminate Rails from the picture and to narrow down the issue to and the JIT is turned off the above issue does not am wondering how others running enterprise applications on Rails combination have not encountered this issue is in a very critical path for us +I think weve had enough builds to call this a success +Bastian feel free to resolve theseThanks! +the consumer stop consuming is due to systemUsage config also heap needs to bump up otherwise it will get JVM GC overhead limit exceededI am currently observing heap +Restarting the server seems to fix the problem +Patch that just changes the line as shown in the jiras description +This patch cleans up an asymmetry in the generation of Derbys grammars +And your last instruction is quite clear +So no tests +Glad to hear this works! Ill commit it soon unless someone objects +I think ejbexceptionexample would have duplicated the original problem if it still existed. +test has been written +Thanks for the known issue resolutions its the thing that kept getting pushed down on my Notes updated +and Layer are ready to go +Please incorporate the anyone going to start working on this bug +What changed +Can we work on this issue as soon as bigbang is over +Your includes pattern is wrong +cs including the files aboveSharpen seems to have a problem with assert so I ended up commenting out all asserts just to get going +Igor could you summarize +Please fix ASAP +Thanks Werner +Looks good to me +you providing a patch +Malden is this something that we would like to add to upstream jsvc or just patching our build +Just suspecting it +batch transition resolvedfixed bugs to closedfixed +Thank you +There is a slight change in behavior in that HADOOPCONFDIR is no longer guaranteed to be first but we can get the same effect by putting that at the front of the value for HADOOPCLASSPATH so it seems fineI can volunteer to test the patch on Windows +its more the random mix of upperlower case just use lower case and or to seperate words if they are really needed +Here are the originals +First attempt at the patch +The generator partitions by host and has a limit on the number of hosts per queue so i would expect each segment to contain slightly less records than topN but certainly not more than topNWhat should we exactly expect +Ill test them all with xerces and let you know +I think that thedocumentation should include an overview of what this feature is how it can help you and how it works reference material on the new system procedures reference material on the new system catalogs a number of examples of how to use the feature +Mea culpaI suspect the problem with the shiftadd algorithm is in the carry as Douglas cautions in his paper +On shutdown the cluster can stall very briefly a matter of or seconds at most in our case +Generally such should be avoided +Running thro we remove the member variable jobMetrics itself +Fixed by Dave in r +I dont have a group called users +Ironically it was caused by a debug log call where the log related fields had not yet been initialized +Applied changes to branch +Thanks Zhijie. +Peter noted that there is an existing change request to clarify this issue +After that it is getting skipped till the latest buildThe failure you are saying should not be because of this patch +So is your issue that the framework should not attempt to start Bundle againPerhaps so but your original issue description didnt says something completely different +I am able to ssh to wopw +Same here I met this one randomly in the past I didnt see it in a while +Preliminary check of the code in shows that for foinlines is only taken into account when adding the areas not yet when the is generated hence why the algorithm is tricked into believing that that break is OKI already have the beginnings of a patch ready here which fixes the issue by adding the necessary boxes to the but its still too rough to apply +Except changing to package private rest of the changes are done +How do I get the corrected source I have not found the fix in any of the downloadable archives +True the solution was simple but maybe more people could benefit from an extended that allows adding a service parameter to the is of course only required when you need a service parameter with the domain copied from the request url otherwise you can simply include the parameter in the default itself. +Does it also happen when you close the session? If not then this supports my assumption that the error is caused by a combination of repeated cycles of not shutting down jackrabbit and a bug in the index recovery code which is executed when jackrabbit is started up again +addressed kons review comments +Platform is a Mac laptopNumber of the tests including node and nodes cluster did not show any or any other exceptionsFull repro steps and init init init scripts stripped to the functionality of stopping cluster are attached +By the way seems like this affects tooAll subclasses are affected unfortunately +there are three different types of delete markers family column version the delete can have a timestamp which determines at which time that delete takes effectThe shell currently places column delete markers hence those markers affect all versions of the column prior to the markers timestampWe can enhance the shell to indicate the delete marker type but the current default behavior should not change +Now exposes a normal execute method +trunkCommitted revision +I did not know thatI have added the description about to and committedAlso updates the resolution of this bug to FIXED. +The recently added target still produces a number of warningsThese were until recently hidden because there was no javadoc generation for private membersclasses +Removed obsolete files. +I used as a template for creating a testI ran the test suites without failures except failing and which fails to me also on unmodified sources checked out from svnPlease review the patch +SOLUTION Use this allows anyone to ship properties files for a given language without code changes Only internationalize the most important warnings and error messages +Okay Ill take a look now +Hoss how can they use this in combination with an thenWith the current patch this works +Checking this in was unintentional +Ben Nhams hack works! +Aegis or AegisSimple? I think youre going to have to attach a buildable test caseWarning the simple front end may not support all of those annotations +I see now that I should probably be using ASM to get that information from the class instead of reflection +So I need to disable exception usage in the library but I need to leave the exception specs for at least one platform +Maybe it should be marked patch available and we can see if the patch still applies +In your patch can you get the region state and make sure it is OFFLINE instead of checking if it is not assigned +I already have this implemented on a private branch +revision from this morningI used eclipse and teamcreate patchthe only thing I did after the patch was created was replace the tabs with spaces and remove the path infodoes it need to be format +If you need to locate a web service test project to test the offline piece you can use the one from the testsuite testsuite +Once more with licence granted to ASF +Fixed javadoc comments +yes its the case of multiple build definition +Need to change documentation to reflect that and also remove the unicodeescape IJ setting + +Simple testcase fix +Also mentioned rm r for the trunk patch +It also includes a bugfix concerning incorrect behaviour of properties not set before as you will find it in the former s boolean properties +Applied on for all the help everybody. +wars they just fail upon server restart +The directory structure for trunk and is the same but is different +The is gone. +At first glance it looks good to me +Hmmm +Cant the classpath for be modified to point to serveralllib +very unlikely to happen in time frame +Fixing +Thanks Erick! Will try to test +has been closed there are still a number of teststhat need to be written before the jms inbound rar is notconsidered experimental +Documentation committed as attached above. + +Is this a fault of the exception not propagating back or our server side code not spitting the errors back +Patch looks good +Thanks Adrien +Neverthless I guess its dangerous to just rmr on testdata +How is it possible that the output stops at or +I would say it doesnt matter that BSF is a bit older +What is that used for +So I just checked in a pretty complete first take on resolving this issue +defer all issues to +and +Since the older Hadoop project has been split into how should I make the new patchShould I make two patches or merge them to make one +And I think it means you found a bug somewhere with too +I will resolve this ticket as a dup. +btw this bundle is needed by to work with minimum package and have been tested inhouse by our you modify the jar provided by saxonwe always deploy the original ones modified versions would need to go in your continue with version where we will not insert notices files to the +DF still does this so Ill file another JIRA to improve thisIve run the failed tests on my end but would of course appreciate any further testing +I was thinking any combination of comma and space would always be used to delimit by default +Adding back in those tags should be fine but they need to by pulled out of the properties definition in hbases profile and just have them reside in the general properties section in hbase +Done +Ive committed some changes that have resolved the issue for me +Closing stale resolved issues +Fixed on trunk revision +But surely M and Beta are end of life since we have Beta and Beta +I have fixed this in the same the patch +Sorry for the prompt thanks! +Thanks! +This works +on addendum +it is somewhat documented in the goal itself i would rather not add any more details until further feedback from user +studioeclipseplugins +Please commit the patch to alsoCredit for this patch should go to both me and Ganeshan Iyer +Anyone has some idea howto fix this +So hopefully each row is smaller than bytes +This requires jndi lookup entries for javacompValidator and javacomp +Fixed the nits Ill commit this once Hudson comes back OK +We need this for integrating Sentry with Hive +what purpose does the integration serve +Best regards GeorgeApplied to make at repo revision Please check that the patch was applied as you expected +Would be awesome to have this in was fixed in and current Bigtop version is so this fix makes sense at this pointPatch is attached +Did you already run mvn test? If yes Ill go ahead and push it +Awaiting THE results from THE usersIll report as soon as I have information +Can I take it that yougve managed to resolve your problem? And I hope that you are aware that it is possible by the means of a binding file to specify that classes created for complexTypes should have a Type suffix +But currently our clusters are running withMay be it may take some more time for me to bring out the results +These changes should be reflected on the live site momentarily. +Attached patch permitting EOL in Slashy strings +Looks like the problem is rooted at at lines ff +bq +Furthermore we should push down the handling for this into the and not put it in +This bug has been marked as a duplicate of +For queue dependent methods it passes the queue name and for other it simply makes a method call +Can we add more details to this please +the doesnt seem to be working Im getting the old good +Patch fails to apply +Incorporates the plugin concept from Pinakis patch attached to +Oke Lutz thanks for the feedback +I verified the fix on both trunk and +Updated Tomcat Jetty code to consider attached fragment bundles when looking up resources +Hope this helps +Best regards fine thank you George +It might worth adding a test case for this kind of template issues to JBPM unit integration test suite +However with the new one I downloaded today the problem seems gone +bq +It seems the compute derivative method is called for the appropriate times only the output is not correctly used to set the interpolator valuesIf you think this is indeed only caused by the optionality of dense output or it is no longer relevant if output is removed then removing the option would indeed fix this issueIt would probably be better to create a separate issue for the removal of output +This should be good to go in assuming that the s stand +This the PDF file and the Java sceenshot from Java but let me ask you two questions Are you still working on fix this problem? is this problem was fixed in other versionThanksMahmoud +Looks good to me all the references to abdera I could find appear to be included in the patch +Thanks Cristian. +QA verified +When the palette changes will be ported to trunk the issue will be solved there thanks Simone +users dont know why they should use Singleton they already have and they see it as redundant +Makes sense +Fix merged into trunk under r. +Closed upon release of +done on head and branch +Logged against CR but also in GA which is not in the JIRA list can you review this +Eddie if you agree please close this issue +I think we ran into the same issue any work around or config tweak to avoid running into this? Thanks +Updated patch to current trunk +This is happening when attaching test +However for now I think we can table it and just deal with the problem in the workflow manager +I was merely inducing upon your premises +Assume closed as resolved and released +The before and after models with the associated schemas +It would be great to have an example explore the alternative ways to denote the same concepts +Now it seems to be easy to localize to any other language +This problem is fixed by Paulex +What to do with this bug? Mark it as a duplicate of some other bug given thatweve been talking about these issues in various other places +The version in which we see the problem is There are no enums in our code because Google App Engine causes a to be thrown when Groovy code uses enums +I dont think the archiver should try to interpret what the user wanted to do +Also please provide suggestion that in which data file we can include the demo data provided above +Lobprivate String xmlTo solve the problem I removed the Lob and assume by default Basic annotation +The failure propagation model nkeywal described above does not work well because the definition of failures is different +This is a very long lasting problem +Id also argue for keeping the name Metadata and just replace the body of the class with implementation this way we could avoid changing the API in so many places for compatibility we could just bump the version number in Metadata +Have you done any testing to ensure that it works +This will result in running the integration tests for different Maven versions with different Maven versions multiple times +are not included but linkedSee +Therefore changed the component to reflect that. +Scott I looked at the client side code and tried replicating the CNFE on the client side +Feel free to reopen the ticket anyway +Patch applied +Could you try that +I am not really satisfied with the patch +Attaching branch version of the patch +Changing the title and environment to qualify the also want to fix this for a long time +I can open a new ticket but figured this was the better place since the RPM needs these paths correct too +I have attached a one line diff that resolves the issue successfully patch is against SVN Mathieu! +This is needed so that user can modify a bundle that has been unresolved by an earlier refreshPackage +svc files and enforces the behavior in the new templates so everything else will be fixed once they migrate +will be fixed in a follow up issue. +It should be fairly easy to implement unless anyone has objections +Thanks! +The does not get moved from api to impl +can u regenerate the patch on trunk +Thanks Chris! +Checked in loadfuncv patch +Ashutosh have three testing sorry it was the distribute by followed by sort by +Thanks +If someone knows of a way to do it even one that is let me know +So I am for applying this changeBut hey if we now receive bug reports of this quality this means that our code base has become pretty stable +Could you please provide more info on the issue +Anyway this will no more exist if goes inThanks Hauhui Mai and Jing +I have attached a patch as The output of svn stat q is attached as To recap the problem in SQL Authorization mode when a new column is added to a table the rows in SYSCOLPERMS for the table in question were not getting updated to incorporate the new column +what does this exactly mean sysinfo API only returns correct information for the first jar file on the classpathThanksAhsan +Before I go digging has anyone tried or beta +I guess I should have commented on that earlier +Lets focus on setting up the arquillian infrastructure for testing and keep the test cases and coverage to track publishing the to maven central +Marking for +Thanks to Stu for pointing this out on IRC +mike did we hang on since your last commit? can we close this +There is a similar problem in the meclipses WTP configurator +Ralf can you please briefly explain what the changes are to and? License chnage +Patch applied on trunk and in JBT +The more hands off the migration the better because then its harder for it to get screwed up +Should we have hooks around reader and writer creation in general +Should I delay for this? Or just put Matteos fix there +How would I find outThanksAndi +Merged to doc branch at revision Im going to submit a patch for the functions and for Knuts recommendation about functionName +Heh just have to wait long enoughSeriously thanks Demai +A new JIRA will be created for the new design and implementation. +At this point I have no further leads +a good question! the problem is that the variable is used for something i didnt want to break the old behavior! Could you check if it it is still needed? maybe even deprecate it +Currently is used for selling goods and is used for purchasing goods I dont like the idea of blurring the lines between the two unnecessarily +SGTM I will try to stick with the first scenario populating data on the CF creation +The patch looks good to me if no objection I will first try to apply it to branch +It would be appropriate to file a separate Jira issue for this +Thank you again +Resolving as invalid since path and query parameters are different and no feedback was given to as what the actual problem was. +will work on this during as Fixed +As mentioned above we do include the namespaces +This issue is being split off from which is focussed on puppetry + removes the code related to caching of connect bytesIm running tests for good measureIf someone think this is an important feature please create a new Jira for enabling it and close this one as invalidI will wait for a little while before I commit the patchthanks +Tests now pass using sharedLiblib now works because it skips loading the duplicates +This happened a couple of days ago change for in xdoclet that was also made for another issue is also not yet showing up in the saw the changes are made in SVN and the timestamps of the files have been updated in the central repository but the file contents havent the sync process broken +Is it ok to combine it? Because only then the process and SSH problem can be solved totally +Initial patch for review +It appears to be working from the console output +Attached is the quickstart with my ajax filterable patch +Thanks for the heads up +That code is written to make sure that for +We actually are not advertising any of our vendor parameters either +Thanks for the patch Suraj. +Ive been using which as mentioned beforewith large OR queries is somewhat inexact in its precision + +I looped trunk on that file some times and got all of my memory back at the end +Thanks Suresh for the reviewThis fixes compilation problems +Same issue. + talked to Owen this class is better removed directly +Havent tried it yet but the Windows changed looked goodThanksMary +So the default should have been false in that caseStill since we have not removed emit yet so it would be good to just pass the value provided by the user for cluster classificationThe current build is not stable I will commit it once the build becomes stable +ShaneI tried to create a toy application in Seam CVS to reproduce the bug I reported with but Ive been unsuccessful in doing so +Moving to ll have to recheck this and update this issue accordingly +Lopez dacord amb tot el que diu Josep M +the issue is still thereIt has been so long I am sure there is a fix or workaround but I havent fully investigated it +Periodic file checking will need another improvement. +any progress on reporting this to eclipse max +and +I think its a simpler way to go at least to only have one class +What if Tapestry or your own code depends on a library that doesnt specify the encoding but uses platform default +Fixed with. +HI Knut +Agreed on all points. +Test case is updated because is now expected +I added a specific unit test for this and cannot reproduce this perhaps you can provide a unit test to demonstrate your issue +That way the caller can choose how to handle the condition rather than having the whole program dump +I agree with Arun +The patch is for the trunk but if necessary I can backport it to the x branch +I have the same requirement and am using stumbled across the class and see that Field has a processor instance which is provided by the instance in the see that there is a but will have to experiment with it to get it working because I cant find any documentation or examples on how it would be used +passed locally on my +I explicitly called waitMsec there so all the progress calls were in one place and you could see where and how often they were needed +I think this was fixed by and we can close this issueAmareshwari could you please confirm +I could also add a property to the template check which by default is true but you can set to false so that you dont have to override the method + +Better use something like the +TranslationA runtime error you want to start the debug modeLine Error Function generated HTMLJavascript error only occurs with IE +I think it would be more useful to use the jobid instead of the start time in the API +webops updated the page +Thanks Doug for reviewing it. +You must escape it by either quoting the string or escaping the semicolon with a backslashCan you try oozie jobs filter useruserstatusKILLED +This should be resolved now that components are loaded inmap after and have been loaded byloadPlease if youre still running into that issue +Im really more interested in seeing the best solution than getting my own code committed +NET calls that are not supported in Mono? I dont see the feasibility of supporting compile time switches as we move forwardAre there any Mono users of +Tim are the following classes just reformatted or are there any functionalchanges in themM M +this was already supported but it is now the default +If a thread throws an uncaught exception and dies will it be in the state TERMINATED? patch for review +New version attached +that formats it wiki style +The main index page will not disappear if there is no content it just wont show you anything other than an empty directory listing +Probably better to just close as wont fix since portability isnt a concern for the project +Better have granularity with the separation of concerns while allowing the different component to cooperate seamlessly +Id like all range issues in git master and test the patches in which also fix too +My guess on the old harness is thats broken but I dont know how +Design document +here is the patch that fixes problem that Paul found +I committed the patch +Interestingly removing the decoration from causes precisely zero tests to fail +That is unless youre in a JVM +I propose +Im checking now the code around thisWhy do we need itbq +This reproduced the issue +Instead of the servlet base class I would prefer to use a filter here but other than that I agree +This issue appears to affect Spring too +Olivier thank you for your honest it possible to assign it to the release instead +batch transition resolvedinvalid to closedinvalid +I assume this is the same issue with Prime file uploads +bq +sblocker addition stuff like this can also be added in later versions as it does not change API of it just adds an additional ctor +The issue was solved by. +Joncould you add at least a valid example just something as simple as select where the should be the variable so we can write a proper link to some reference might help as well +The solution was to read the whole lob into memory +I just committed this. + into trunk and +Namit i created a new jira maybe there has some tasksnow i finish the work that the table is not partition tablenext i will work for the partition table +Im actually not sure whether this happens currently +bq +Struts requires and subsequently I believe +The problem shapefile is attached and projection is WGS UTM Zone placed on a normal map you can see that your projection is telling uDig that your points are located somewhere down off the face of the earth +Yes I think more than likely this problem was related to several bugs which have been fixed but not checked in yet +If there are more problems please file new JIRA issues to handle those. +It looks perfect to me +If you really need to update your published module metadata you still can use ivypublish again with overwritetrue +Well I have though about it but it fails the following scenario +UpdatedMoved the check to so it is reused by andIm not sure whether the same check should be used for tooPlease try again +Looks good +label and title are rather synonymous +fix for the issue with junits log +ThanksWill be part of the release or can the fixes be backported to +I had put update site into local SVN that requires authentication and eclipse asked me loginpassword +Plus in that case wed feel confident deleting in its entirety so that other people on it are prompted to update and there is no need for current code workarounds simply for let us know your preference and well support thanks know there are examples in Roo code but please try to understand us we have to be productive upgrading our for for and for causes we expend time in work that doesnt add value to rd party we are selling the idea that projects havent any Roo dependency but our clients dont agree because they cannot have features provided by rd party addons when they migrate to to and +What you have is probably fine. +Hi Christian JacopoNot sure if adding en is the better way to go sounds like a camouflage to me +you be interested in contributing with code since you already done all the research +With the actual release I get the same effect even without srconly! No is generated which causes likeEclipse to complain +Seems to be fixed +the interrupted exception is a tricky bit +Ill close this ticket for now as it is not a specific Roo issue as it is a combination of and Glassfish +Do these requirements cover the test failure just mentioned +patch v counter names dont need to be configurable +Patch commited as revision Please check if the issue is fixed there and feed back so that we can close the issue +Ok going to check the test again +Initial proposed fix is there some reason why the stack is not empty at this point? Doesnt that indicate something is being pushed on the stack but not popped off +For whatever reason on my system running the build never resulted in a GC that required reaping that above mentioned weakref while on other systems it did sometimes. +I made those changes in does not apply would you please rebase to trunk +Glad to hear it was only that minor bug +Sorry for poluting this mailing list +Is this issue still open or can it be closed +However we still need a test start and end date the entire idea is that we use only use preferences prior to the current time to make a recommendation +Please verify that the head version is working for +have to look closer at why we do the and then sometimes do for a position +An example of the bug +Thank you Daniel for the review! +You would get errors in the logs but not in the UI +Patch with fix plus new option to limit the bundling to some level of dependencies plus a test for an old issue tomake sure it doesnt come has now been released to the maven repository +Jirka Peterka just tried it on Ubuntu and it behaves the same way when the server is started using +The only way that something can be put extra in is if the original files we copied from had those artififacts that was the case for Seam and I fixed that in the seamgen in seam and for the Seam CR and GA so if your seam is one of these older versions then that explains it +The refactoring is too much to do in one step +If they are balanced in this way then assignments are made in a strictly round robin fashion +Bela I can confirm that thread disappears when the bug is I did not test with CVS head nor yet +Resolving. +Sorry Alina +I can see the light overriding readFields is sufficient +Thanks for improving this +A client can just ask any random tablet server to do the operation +It should not be called period +I didnt know about the trimstring setting +Closing issue code is in release. +a network of brokers with a camel route on both ends of the netwerkThanks for taking a stab at this we really appreciate all the help were getting here +Does that make more sense +I currently have a class to overide my foreign key names but its a little clunky would be better to have this at a Hibernate level. +Thanks that seems to have fixed it +Can you add the tests in the patch? It becomes easier to can you refresh and the patch? It does not apply cleanly and is therefore not possible to actually compiletest and understand +its not actually working properly +Thats why I deprecate the whole class but leave its code in +Reader events should be in a different issue as reader pool is moving out of IW soon? All tests pass +I assume its Linux specific issue and its blocking JBT installation through Install new software functionality +Otherwise there is only a trivial improvement for the patch please addretain the javadoc for the existing constructor in +Right now we dont have enough information about the issue +java files are different +Looking into this before RC release today +BTW having JRUBYOPTSrubygems did not cause a nailgun slowdown approximately weeks ago when I first tried nailgun +That is pretty much independent of the classifier +Switched to use creating +It will help users implement LDAP proxy very easily +Harmless patch it is but Id like to understand what it is fixing +I will fix +I left off +I guess this one slipped in since them +Thanks + +Unit test should make sure that behave as expected regarding state transition and pending operationscallbacksFor this Id add a watcher and make sure it gets signaled right? What other checks should the test do +Implememtation for Confluence should be nearly same as for JIRA +As I explained increasing sstable size helps but does not solve the problem were supposed to be supporting up to of data in +Wont the rpm be overlayed +Is there any update on this? There has been no activity for about a week and this seems fairly critical to fix +Duplicate of +Committed revision in Nutchgora branchI would ask if Enis could now do a final check and now close +In order to get the fix that Tom Jordahl describes +The error shown in the description is the result of being moved into the src directory so the solution is to modify frameworkssrccontribmesos to change the location ofAfter that there seems to be a problem with some protobufs being renamed which is fixed by changing lines and in frameworkssrccontribmesossrcjavaorgapachehadoopmapred b replacing with and toWith those changes Hadoop will compile and I can get a namenode and datanode running but the job tracker returns the error that it cant find class Scheduler and Im not sure how to fix this problem +Profiled to reduce allocations cutting out about ms in a caliper benchmark +So its using long name to get the renewer token which violates the mapping rule in authtolocal property in hadoop conf +Konstantin This was one of those situations where the changes for this JIRA had to be applied at the same time as its parent so wasnt possible to have Hudson check this patch either before or after was committed so I did it manually +It may be still pointing at the old svn repo +Reopening since this bug has the more accurate analysis and solution over bug +A patch for adding the features and property to the code and a new class that was added to enable the features +That needs to be feature does work with the JAR Plugin though but it requires version of that plugin +Olegtake your time +Also added the description of this issue +What are the steps to reproduce this? If I receive an offline payment from the order screen the payment isnt applied to anything automatically +With this patch map side limit operators are vectorized +There possibly is some hcatalog overlap but the lean and mean rational is good and I love the support +In theory should be called after the view is rendered or in the worst case when the view state is written at the end of hform +This passes consistently for me when run from Eclipse +Seems at some point Xerces uses a for a Node instead of a String so in our instanceof tests miss it +Without password substitution an invalid user name correctly complains that the credentials are badij version ij connect jdbcderbylocalhostdbuseradminpasswordadminpasswordsecurityMechanismij connect jdbcderbylocalhostdbusernonexistentuserpasswordadminpasswordsecurityMechanismERROR DERBY SQL error SQLCODE SQLSTATE SQLERRMC Connection authentication failure occurred +Will add today +I was just wondering why would you want to do it in a series of hooks rather then doing all of it in one hook because it seems list must all always be complete and hooks in that list arent going to be optional +ack first version worked on but broke on trunk +This is important for PDE projects +Changed this dependency to provided +packageshivemetastore +Tell me if I did an error by doing this copy +Unfortunately there is no way to avoid this situation +Heiko is this information provided on the server side or we need to add it in the first place +Resolced +According requirements RHEL is not supported +Overlooked this while creating closing this issue as the other one has the diffs already +What happens if you run the Java test case I provided using your same configuration? You will probably have to make some changes to address the URI and user pass for your broker setup +Does this build include the fragments from? The plugins directory should contain or +I think it would make sense to make a +Reviewed by Keith +Fixed in +There wont be a release +The session should be from the connect this is why it is not implemented +eric The format required to add the missing information to DEPENDENCIES has a very high level of replication and is hard to maintain +Reassigned for the latest changes in does it really make sense for these interfaces to extend +The problem appears to be in the Gradle Tar task and this does not affect the Zip would be possible to release Gant as a Zip only distribution that would be correct out of the box +OK I will work on a patch that quietly drops properties where compliant is defined by the statement The set method accepts values of class Boolean Byte Short Integer Long Float Double and String +new style classes +In when we serialize for an old version and cant find an oldId itll probably be better to use a more user friendly message like Cannot send column family X to Y as its version is +tests +Updated patch again +Thanks for the patch Michael applied in trunk. +This serves a similar usecase but its way more powerful +fixed and tested in +Now that work is well underway with AS all previous community releases are +The fact that it works fine with sitemesh in the showcase tells me that we are ok for now +If I dont hear from you in the next week I will close this ticket as unable to alan My project has gone ahead too much not sure if the backup can help now +After giving some thoughts into it I am not sure what shall be done as examples in demo! Any ideas +I could not reproduce it on the current source tree +Or do you think it is not needed +Raised for the improvement +Looks like a couple of the tests can occasionally fail due to timing issues when run on a slow aix or solaris box +still I think thats a valid use in r and documented on wiki. +Duplicate of +This patch is needed to make Hbase work correctly on Hadoop +Do you plan another JIRA for actually fixing the default port number +Can we merge this patch before it gets overwritten with some other patch +Grid do you have any ideas on this one +This runs fine on +IIRC the property is a Map and the base is Foo +I do not plan to test the database itself for a mature database has already been taken many tests +Its such a minor JIRA I didnt think it worth wasting your cycles which is why I am leaving this message here as opposed to email +drda protocol test to which ensures that the error of wrong RDBNAM field sent in PKGNAMCSN is handled correctly +Close this JIRA since all its have been resolved +Set Fix version to +Attaching quickstart showing the problem +When we tell couchserver to restart it has absolutely no effect on the other daemons +project reopening to update release note status. +An example of this needs to be added to the HAN reference documentation that also explains why there is no other way to do it +This issue is just about pageflows which arent persistent. +Carl Rather than debate it theoretically or compare it to Hadoop which has a LOT more complexity in their build I propose that we have Travis make a Maven build file for the combined Hive and systems +Patch coming in a new jira soon. +Thanks +Im looking forward to see this capability in synapse +This problem arises because of the way the features are encoded to not permit to be a valid update for and the fragile dependency of jbpm and jbpm on their feature +Fixed + HAMT data structure used by the allows for efficient diffing in this caseWell that might be the case but the doesnt +The test method runs without problems against the current trunk +So this files werent compiled +This can be done can we have a Juddi PMC member confirm the change to wiki markup +Can you attach the exact tmpdocProvhtml file that produces the textplain output +Minimal unit tests exist. +An updated patch which includes a fix for speculative reduces +Finally fixed +Vertex Name is the user visible means of interacting with a Vertex +For monitoring network connection an instance of network can be added to the metrics system beforehand and a sink can be implemented for receiving related statistics such as source and dest value etc +Thanks again +Also the tests run fine with latest surefire plugin. +Closing all ResolvedFixed issues lacking a Fixed Version +When this get fixed the PRECISION field of should also be changed to match +I just committed this +Reopening for release note +Everything should be normal again +Attached log file and the service see my comments above +Forgot to add that I also tried this with and without dont get an exception when that is set to false +Fixed in r +Remove the request object from +Tested Giris patch and was able to successfully build native hdfs +Thanks I see the change +Please let me know if you need more description of my patchIMHO maxActive should be a hard limit of number of pooled objectsBecause the upper limit is necessary for system resources like database connection limitit cant be allowed to overflowRegards +This issue related mostly to my unfamiliarity with our testing suite +Merged to +Please provide it as soon as possible +The patch is committed into the apache repository I will close this issue when the code is synced to Fuse +Heres the code again as an attachment +I dont think it makes any difference but you can do a git pull I changed the classpath to include the groovy libs rather than explicitly including in the referenced libs +Patch addressing the issue +Ive incorrectly checked the license +Of course I will show you the config once Ive made sure it all works fine with the change +I will be committing this patch now to both trunk and committed to trunk and branch +Started doing bitmaps but it made code unreadable +And we could add support for CGLIB proxies and you could create proxies for classes as well +I missed out one more point +Alright closing the JIRA because it is resolved in +Client who has started screen sharing is not seeing itAll other clients in the same room are able to view the screen being shared +Move the fix to +Please look at the attached patch and see if this is the right approach and if anything should be modified +Which release will this fix be in +Hi Tejas +Bulk closing stale resolved issues +Watcher is assumed to initially be in disconnected state +Failure started in and still present in trunk as of rev +Reopen for backport +Great idea +So in that sense I am tending to agree with HemanthVinod +bq +Works for me too +Fixed under r +The scenario is that libxml adds an additional node when you have whitespace between two nodes +Mono is free to install +Here is the patch to fix this issue +Comparing the contents of these zips I see this +Because its important we solve such issues +two logo designs to potentially replace the current logo +Patch committed +Ill keep investigating +Not sure what the problem is +I committed this to trunk and the branch +Wrong output of German umlaut u with dots +Probably in. +Attachment has been added with description Patch to trunk to add support for static attribute +I committed the patch to and for +Fixed catch selection logic +It still does not work with version and. +This is not a feature then it works as expected +Logged In YES useridThe fix proposed is simple use protocol name as additionalargument localhost localhost httpsThe protocol is the last argument because i do not want tobroke currently existing customer scripts +Please help to verify thx +Ive applied it in r. +Marc as you reworked the internals recently do you know whats happening for Jeff +I believe that this issue is not isolated to collections +Then you need no extra configuration or symlinks to make them availableIf that is not an option or if you want vanity or other customization you always need to configure such mappings somewhere +Complete XML Schema that reproduces the bug +I agree have not seen any other format do this +It also updates Avro to which propagates the correct error string +Added selectunselect all buttons to toolbar to save space +and added heartbeat testing code walkerLauncher method +I think you need to check to see whether there are actually any locks to be acquired and if not skip check +We are working on +Thanks for looking intofixing the issue +Are there any steps here Xavier +Confirmed that the following commit introduced the regression regression marking it as fix for +You can extract these resources in the Karaf branding plugin and delete its own copy of +Looking into this +Check patch through Hudson one more time +Backported to with revision +What do you think of having the token validation and transmission as a separate JIRA? is the meetup +patch updated with revision Darrin applied toThe basic eclipse tooling is spotty at best for formatting what editor or pretty printer are you using? +Verified on EAP Affects and Fix in Version +the trigger listener is being removed from the after the first is fired +Sample Image +If the last used is less than the valid then the beans need to be refreshed so all values could be reset at that timeThe easiest solution I think is to allow the tables to be created +Changes are checked in w new test case and successful junit +QE please can reproduce this error with showcase on both tablets and mobile phonesThe issue is noticeable only when selecting the suggestion for the first time +This problems symptom is not same as the one reported in +wed have to buffer events for a certain time +Patch for trunk +The bug has been fixed there and we are working to get a release out as soon as we can +Other solutions either reduce usability or require massive complexity +has been marked as a duplicate of this bug +Duplicates +User restored to and. +That will be equivalent to the job history web page and could atleast give a user some useful information +See posted again under +Per field would be hard though I agree +Run validation twice or more over on the same JRE +Ive configured both the httpclient and lightweight wagons in the and am still seeing this issue +Thats one of the other problems with this definition +By default I think you get whichever version is supported by the JRE you are using when generating the Java interface +could you use to achieve what the issue opener wants to do +Id highly appreciate a Unit test or at least an example that demonstrates what problem is fixed by your patch +Has there been any traction on this issue? Im just wondering if the should be included in the serverdeploy +I fixed and deploy aI updated Karaf branch to use this SNAPSHOT and perform tests revision +The test needs to be explicit by setting the transaction type to datastore regardless of the default +Today I updated to and sonar and it was not working correctlyWhen looking into it I investigated that the problem was within the application context as the following redirect happens httpserversonarbatchbootstrapindex httpserverdeploybootstrapAfter changing the batchbootstrapcontroller file it was almost ok but I was missing the following Access to index should not require authenticationskipbeforefiltercheckauthenticationonly indexAs this is not yet covered in the version that I downloaded +Then later opened it again and switched to theme view all the elements would be reset and youd have to reconstruct you theme +Factored out so we have one clear place to change from false to true in +The result is runINFO Required EM Launcher CXC RABERROR Unable to locate a logging implementation using problem isnt that it cant find your configuration file +I have been thinking of making this configurable I opened +If a was allowed to leak out of the persistence layer into client code it would represent a break with the layered architecture promoted in a JEE is not to say it does not happen Spring Remoting introduces a couple new runtime exceptions and restrictions on the equals method beyond pure layer boundaries +attaching thread test cases will pass with this patch to trunk and +This issue may or may not be related to the issue reported by of which I just submitted test cases and a patch for +Still if we add something to the XML flow builder we need to keep the in line. +But not sure we need to worthy of it. +I removed it and will create a new one with hopefully the dependencies fixed +You may want to considering fixing it up too or file another jira +So the above behaviour is the normal behaviour but stack traces are printed if the v or d options are given to +patch to add functionality and unit tests +The link works for sure on this one. +If you want the behavior that David describes where the object is empty if the field is missing then you should use a struct instead of a union IMO +Waiting for that one we feel like this is an acceptable solution to have links up to +Since it will only add additional logging information when the database is booted read only due to an unwritable lock file and wont prevent the boot +In addition to my comment on I also want to point out that prior to this feature the servlet did keep a session open past the request +A breaking API change such as this is not something that I would have expected in a minor release +Working on a new patch +Keystore generator now generates AES bit keys +Tests were added for replicated in and tests were added for local in as others have pointed outThese however do not test mapreduce itself and will need to be added into the package +Isnt the SQL generated by hibernate +Applied the patch +API support for this? I would assume not since this is Avro but I thought I would check +Chandan has an upgrade setup using ESX which does not have this issue +Assigning to Trustin since this is related to the OSGI work he did +I can reproduce this issue only on win x +Does this mean the issue is fixed? What version of railo can I expect the fix to be in +That way I would have the full string to build the original exception message +Unassigning since I have not got back to this in a while +Thank you Dennis for this clarification +It may be worth implementing these as seperate MR jobs for higher efficiency The is some code replication between some of the jobs +arcconfig +Why not commit for now and use it to deploy it to the apache m repoThanksVincent +us es +Closing bug as per Eugenes commentthanksdims + for patch +Transitive dependencies in contribvertica and contribsqoop were pulling in so I put explicit dependencies on in their files to sync them toNote that the original motivation for this issue is that Sqoop depends on but seems to be the only component of Hadoop that does so +Awesome! Thanks so much for quickly taking care of these items Ted +Thanks for the quick feedback Ralph! Why is having the password to the keystore in a plaintext file a problem? File access permissions could be used to restrict access to only the Flume user like a private SSH key is protected +That could be done later though +This patches so that can provide columnTypes for path and score columns +and now only close streams that they explicitly open +KB Time msI seems that it doesnt really matter if the buffer size is KB KBor KB +bulk close of resolved issues. +Any luck creating a test that recreates the issues you have encountered +The problem is that my IA assembly skills are nil +Latest patch looks good +Please verify. +And once I move raise to another line it starts to report proper line numbers +Whoever creates deletes +Currently switching between tabs of dont reload all data from storage but we remain this purpose for performance because reloading make displaying slowly and data changing is not often in this portlet +v initializes replicatingNodes to avoid +I have a comment here +Thanks! +Im marking it as rejected just because this cant be a task by its own +The spec prohibits that the Conversation gets touched by Requests in parallel anyway +There are also exceptions thrown relating to local storage as mentioned in this after the issue in is fixed +Cool thanks Nick +Use an Interceptor instead if you wish to add security to results +The directories missing patch for review. +If it is less than that we have the same problem where shutdown will hang +Uses skip list to sort entries before adding them to entry log file to improve ledger read performance +Committed revision Thanks Kay! +lgtm as a lot for the reviews Eli and Todd +Unless i am missing something this has been implemented? Please if i am incorrect +Would be very useful addition indeed +This has been fixed and and works properly +I would like to figure out how to expose the filename in the API +Sorry for bothering +What could be the consistent solution regarding this +xmland +In fact +Rebased for trunk +Figured out the issue. +reopen this I dont think this is fixed in master at least maybe you pushed the fix to somewhere else +Sorry in the delay in responding +Does that mean that you can make this logo? IF yes thats cool newsFYI the is similar to the one of for +more case of excessive +Looks good +I can certainly add a setter method to private classes in are really an implementation detail for that class which are directly tied to it +Empty project which reproduces the problem doesnt matter which goal just type maven in the root it case it helps this is the output of env +I propose using instead of as it is already extensively used +The test failure is caused by it +I think we should apply the patch to the branch as well +The unmarshalling layer should expect to get these pieces of information Class javaType java type of the root element xmlName root element name xmlType name of the complex type definition in schema Does the unmarshalling layer depend on the complex type being available in schemaI assume you need a pointer to the xsd +Patch seems ok to me +Ill also see if I can reproduce it +so once patch provided could we directly repacle them with the patch files +Original issue was fixed though +Patch is renaming to +I can confirm that same happens in Liferay too in Tomcat there any way to manually remove items from session +Sorry missed that one earlier +Smaller blocks dont matter +So I cant reproduce it and can just wait for it to happen +Contains improvements listed +bq +When I changed DEFAULTSIMILARITY to the worst change was a difference of +But blocker no. +Apart from that there are some whitespace changes which can be removed +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Its by design +Users need to be able to drop their WSDL and XSDs somewhere as well +Marc and I will do the +If it is something that you dont want to put into a sustaining release at this time then please give me some feedback possibly with a depending on how adamant you are about it so I can address those issues perhaps by fixing it just in instead +I also removed some lines that are now useless with the changes introduced to the parser +Yes plus moving all the fix into native code +Yup will post one soon didnt have enough time on hands +Running all tests now +Hopefully +I will add support for this +And thinking about it more I dont see any problem with B accidentally discarding useful state +I agree that we shouldnt create a datetime type since the implementations could end up being very different and complex amongst all the languages +Deleted the most recent trunk and x alternate patches +If no DARCH is specified it will now automatically determine which libraries to build based on the users environmentFor example make DARCH will now only build the bit library even when the user is on a bit machine + space is a different option not MAVENOPTSXX do you mean whith MAVENHOMEbinmaven file still has MAVENOPTS Xmxm Xmsmcall this before running mavenset MAVENOPTS Xmxm Xmsm XXm +Also all our build scripts over the past years ran Checkstyle without giving it these dependencies +I find it quite interesting because It forced me to create a custom Message class as well e as customAny comments would be very apperciated +changed some of the semantics of the so Ill be investigating more +I also prefer the naming you suggest +Good catch Barry Ive missed that +It looks OK +Moving issues from affecting to affecting or before deleting version. +Is there anything else needed from my side? +The test now fails without the core changes and succeeds with +Added v with updated other than that patch looks good +Seems like this was satisfied a few months back +Hmmm +The following files are changedM javaengineorgapachederbyimplstorerawM javaengineorgapachederbyimplstorerawdataM javaengineorgapachederbyiapireferenceM javaengineorgapachederbylocIf this fix is applied the database will boot in mode if there is not free disk space in the database directory +Exit the Namenode when the thread encounters a +We could use something like obrlocal and reuse the existing URL handler but Im not actuallly sure its worth it given the current api. +Therefore the composite bundle must have a valid framework into which these bundles can be installed +Closing. +What should be done to close this ticket? In my opinion the pojos way is the best if used with spring +destroy method was provided as part of issue +Thanks Dan +Till then it can float in backlog +Attached a patch that changes the close behavior to not attempt to close the underlying connection + +Patch doesnt include binary files +So I am not entirely surprised by this bug +Ill redo my testing to make sure nothing else is lost but I think this is it +has author name inserted byOtherwise looks good to me +A second is defined in the of +Nitay this worked for me on trunk thanks! just committed this +This will not work as the relevant classes dont have the jdo related hooks compiled into them if Cocoon was compiled without the jdo jars in the cocoonliblocal dir +This target can launch an instance of from a previous version then launch an instance from the current version and confirm they interoperate target assumes that the jar files for the previous version can be found in directory outputinterop +Min This is similar to a bug you fixed +TestedWe may update the version to latest GA if its available before we release the plugin +Yes I know that but im pretty sure they dont use specific index to resolve either and it seem to still work if you have it on maven classpath +Can you reproduce Ryan +displaying the old logo and specifying version +I didnt create dedicated channels for the different tests +I was under the impression that escaping a second time was not an issue +Its my fault Sorry! +There will some be some work to move around some classes methods to achieve a decent separation between the client common server before we venture down the road of separate artifacts +See +Fixed in trunkOleg +Got the same error on Wicket +I also tested on my machine for hours from shells concurrently without any issue +Will an older client be able to make sense of data written by a newer one one w this extra salt in it +Changes to use the patch as it has a bug +Patch for for Ted! Ill fix these +I fixed this in revision by reintroducing the fixed list of known font metrics that I removed during assuming it would be better to allow also other font metrics to be installed by client applications +Recently we ran across a jobclient that failed with an OOM error once we updated the cluster to +This could also be used to specify war jar to resolve some of the cases described in this this wont solve situations where you have a tree like Bruces example where you have warwarjar and you really only want jar files because the war will also be included in the resulting probably needs to be fixed in mavens dependency management so that the tree is built first and then the artifacts are filtered. +bq +Patch fixed a cp error +At least it compiles +Still to do Categoriize and Recode functions +The offline imageedit viewers may be misleading +patch move the collection class and minor changes +Dupes this since there is another defect for same issue +If this is addressed in it will probably need to be done as a simple config option for enabled system functions +Define property and use connection url jdbcderbymemoryI would like to try this patch +This problem came up +Updated Patch for +Committed revision . +I have a transfer request in progress +Also preliminary tests using reentrant locks as a solution for this show that the performance in some cases is then times slower +attaching fixed patch +I suspect processing distributed suggestions would be more robust too but thats just a SWAGIm not insisting on this mind you but I was just looking at this and it seems like theres no downside and potential upside so I wanted to suggest it +This could be helpful to mitigate the problem of too +Jar files on the classpath are +I took the testcase from Sergeys comment changed it to catch as stated in the summary for this JIRA added several new testcases and finally added all this to +I had a similar requirement once I implemented a Collector +I fixed this some days ago but wasnt aware that its in bugzilla +When this thread will be ran calling the first java method will set a breakpoint on the first bytecode if this mehodYou can add the above paragraph as a comment in place of the code which youve added +If the service is already active before the Declarative Services bundle is started this service event is never received by the DS bundle and thus the configuration support instance is not created +So I removed Map implementation from FYI when I added a new class named I refered to class in I did some basic tests and it worked as it did +I tested it and it is workingOne thing though Id like to change your form so that amount applied is the amount applied to each invoice from this billing account which would be possible after is committed +It would be nice that we could use same way to handle the id and key attribute values +Done +Unresolved +oh sorry Flavio +I found that it is needed to make server side implementationcompatible with both of JDBC and before that +It doesnt change the api at all so makes sense +I also think that we need to make this configurable in all plugins +committed please feel free to set it up on Jenkins now +Resolved fixed +Created the class as a common result set metadata provider that can be used by a mocked and changed to also use this class +Thanks! But this oddness needs to be documented so Im going to reopen it as a specpguide issue +When I delete the predefined serverthe graphs related to the server are all deleted +Suppose we want to store foo +I think the logo should be very simple and representative of the function of + +most of headers have been changed +voice XML schemas now compile properly +The reported found the problem and it seems it was not a issue. + +Meanwhile there is JAXB for telling Juergen Ill update it soon. +Especially because this is technically an incompatible change +Also would there any be high availability supportfeature for namenode when hadoop is deployed on cloud? I am sorry if that is naive to askNikhil +batch transition resolvedfixed bugs to closedfixed +Binary version of the failing portlet +looks good +Hmmm +Closing old resolved issues +Gary +Everything is well since r. +Feel free to commit it its easier to tweak the details in svnEsteban The idea is to keep these tests free of functionality so that we could run the benchmark on various different JCR implementationsMarcel AgreedBTW why is the explicit concurrentconcurrent dependency needed +Please check that it was applied as expected +Hi allI have not received any comment regarding the implementation of budget functioanlityThis is to inform you all that I have started the implementation of the budget functionality +I will be attaching the statistics pathces to +I got distracted away from this task a few months ago +I have also added a test in the latest patch +Done in trunk +Note this only happens in the browser mentioned when running Wicket in development mode +Is anyone from Yahoo working on this? Its pretty straightforward to add a config option +I would suggest modifying first then we can simply consolidate all the changes into the script +thanks Vladimir +I just have a couple of comments My spell checker is complaining about a couple of things so it would be good to fix those +Ok +Filter is different because you are providing one boolean condition +Here is a response from OracleHi Steve I just wanted to close the loop on the bug you filled about not being available with the client JVM in u +Ok now we are getting close there is a bug in the ie on certain files if the filelength of a stream is not set in the content header i once stumbled over it in pdfs I will push this bug forward to mario he will look into the content length issue in the resource loader +Is that right? Does that mean that we will be writing metrics to a file on start up when this patch is applied +I have moved the class to package +The implementation is straight forward by decoding the both the cid and the contentId url and then comparing them +Patch for trunk +I tried to reproduce the problem by guessing what the tables here would look like with my own data +as far as i can tell this is fixed on the trunk and tested by +can you take this oneA patch file should be extracted with eclipse that shows contains the diffs between and the current head +New patch because of merge from +I wasnt aware of that and given that it apparently hasnt been a problem Id assume it wont be for for the feedback +Felix youre right I was too early with my conclusion +patch updated +Found a problem with readString during testing of encodings containing AND sequences +Working on the fix +Not good when youre sending alerts! Sergio does your fork handle the bug? Right now thats the last known issue Im concerned aboutThanksJimEdited wrong bug +This case is now unchanged for almost a year +Brian suggested a better way of solving it +i tend to be overly careful with such removals +I sent you an email with a screen shot +But I believe you have a point here +. +It seems to build fine. +If I startup a node wo specifying the collection it would just be a part of the live nodes and no shard and collection should be assigned to it +Adding a new DNS Record Type DNS Record Type of GEO for into trunk thanks +Here is what I applied to +Fixed in rev +Is there a JIRA entry for the Win port yet +Do we have any numbers regarding the improvements? Like its faster faster +Assumed fixed since we are not seeing compile errors anymore. +Thanks Keith +Heres a patch applies on top of andBeen in heavy testing for a week or so should be stable +testcase fixed +Ill see what I can do +having one static lib depending on another static lib is not a normal operation however having one static lib dependending on another inczip of another static lib make sense +As far as maintenance concerns Ive usually found testable code is much easier to maintain that code + patch does fix the circular reference but exposes other problems in the fat client +Id be interested in hearing more about the signature youd need for this to fit your needs +Without this check the defaults were ALL true +Relying on CSF comingin probably wont help this the case if it doesnt make it intothe release +The behavior we saw with release is that if a client in the middle of writing data gets a SIGTERM then all the data is lost +Should I remove the file and put all of the jar files from the in the nightly build into the pluginslib directory instead? Or should I replace all the in the with the same files from the lib directory in the fileThis may seem a silly question I apologize if that is the caseThank you in advanceRegards John I believe that simply replacing the would not solve the issue since other axis dependencies for example like axion +Heres a patch which deprecates the misspelled method and delegates to a properly spelled oneI would like to commit this fix the dependent projects then commit another patch to common to remove the misspelled method in a day or two +bug brought forward to part of the project OSM tile layers are being looked at however this bug report looks to access the java api directly which would be a great new feature. +Bruce +Is there a reason to not use the communityRegarding the enforcer plugin we use versions of all the Maven plugins +i think the brief description not that much useful without the arguments + +Includes the changes to following classes corrected some indentation in update method of +It is really impossible for me to guess whats going on just looking at that stack trace +SteveentrySet calls can be replaced by entryArray calls as suggested in patch for for example +And something should be mentioned or done for Windows which doesnt do Cyrus SASL but regular Windows users for auth +This is committed to CVS already and should be available in tonights snapshot +Why cant it access the config? I am unsure how the testing changes would affect this +Did anybody try to run things? If so are your experiencing the sameproblemssymptoms as I do? Any idea what to do about those +It is necessary update BNF and Exampeles closing all issues that have been in resolved state for more than one month without further comments +OK Ill close this one for now but we have open in the next version so we can figure out a better solution for handling widget version dependencies +Why mix Java versions here? What do you mean by crash here? What program leads to crash? Logs leading to the crash +The attachment contains a testcase for prooving the bug +is it possible for you to come up with a contrived spec that doesnt use a MB jar? We already have enough binary cruft in our will happily apply your patch and a test patch then +Thanks! +in light of +I like that idea +patch looks good +Fixed according to Maxs suggestionsThe Requirements and Project ready pages have been changed +What I saied before applies if you already have downloaded the plugins as zip and have unzipped it into your eclipse directory +Consider closing as there is a workaround for very long state transfers increase retrytimeout in FLUSH +Can you be more specific please +Can you attach the relevant code and point us to the classes which make up the flow +this is not a bug +Hence jobs are now sorted in priority order everywhereIn other words the two identified issues are resolved if both the above are fixed +We use a pom project to establish a build and create a WAR using the +and test was run onm MYSQL +I think it just does not make sense to add this +Simple patch + +Im not comfortable with using XSLT to infer behavior in XSLT +If you can do it yourself please reassign the ticketIs there some approach we should take similar to the way that datanodes to the namenode? That might be even better than the FQDN +I think previous patch has prefixes included for paths like a and b that might be the reason for not applying +i believe this is already the case asindicated by Len in jbpm project distribution should keep its convenience of deploying through the designer and the the only thing that should be done for this issue is make sure that after the SOA platform team switches to the new jbpm branch that they still keep the exclusion of the servlet in their build of the soa platform +The RAT run on the files currently in git +Closing. +with the right version +Committing a patch to validate that the input file path exists before trying to read from the file +From the most important fixes are to make pluggable for as we prepare to merge +Rolling over to reduce workload for pre only works with untyped injectiondef only works in the grails console +This limitation was a side effect of the indexation strategy for resources which stops at the rd character of the resource nameResources will now be indexed up to their first character which should enable the bulk deletion of letters long projects +Oh yeah sorry I forgot to post an update about this +And it works normally when I execute a single table select +Can you verify if this problem was introduced with? It may be regression +Closing issue as it has been released as part of release +mesos shanti incubator Aug +It is not totally clear to me why setting for the sourcepath and make the build work on all the platforms +committing shortlyJust made a minor wording improvement by removing the redundant asa significant to and trunk +Also jaxen should be added to the so that the version can be managed +Thanks for the patch Lev! +Arne is this one fixed? +But I ask to add this parametrs into CA +Committed to the trunk as revision . +Any idea why the tree view doesnt work for the version of the ESB schema +It is not part of the sourcecode it is created by the running of Junit testsDoes the fail directory go away if you run ant clean? I think it shouldbut maybe it doesntIf the fail directory doesnt go away when you run ant clean you canjust remove itBut then we should probably log a separate issue to improve ant cleanso that it cleans up this test directoryIn the meantime Ill look at your latest patch because it sounds likeits getting close to what we wantthanksbryan +bq +Whats a good way of making a unit test case to replicate this issue +Nicely spotted it looks like a duplicate and Ive tested out thelatest script off trunk and it works well +OK I tried to create a unit test for this but I couldnt find any environment variables that are guaranteed to exist on both Windows and unix based systems and java doesnt seem to have a setenv to set an environment variable +Nope +To my knowledge blueprint works like a service tracker that one doesnt stop a bundle either depending services are stopped +Ur patch import classes from package +Created folder for +This fixes it +Plan to commit this to both trunk and x branch in a couple of JanLooks great! to committing to trunk and back portingJust one personal nitpick if were going to add Javadocs to the methods can we add full javadocs +Bulk move is the status on this proposal? It would be a great feature and very beneficial to my current project! Do you have a full solution on this and witch version do you think this feature will be released in +why was this reopened again? didnt we fix these for the version set +Well I am quite confused now +The servers name is a name of servers installation +Removed from as per rRemoved from trunk as per package should not be deleted completly only the deprecated classes in it should be removed +We should commit before I revisit this patch again +I saw the same problem where the callback filter still iterates all methods in the proxied instance with each instantiation +MR uses it as well adding both and to its container classpaths +See if you are interested in the catching Throwable issue. +if you want to run the app on WAS then remove the compiled classes from the war +The fix is in btw +Committed to trunk and backported to +Lars could you please also provide a patch for branch +IMHO Hadoop should take care of publishing and transmitting the right exceptions making it possible to applications to use the transmitted exception as is +d on master branch + Notes marked Not Required as this affects internal Red Hat testing only +Can someone take a look and commit the changes for me? Let me know if there are any questions or concerns +closing as we already shipped ORM with the fix in we should back this change out until we get a test case capable of reproducing the problem +Basic implementation is complete +Yonik youre right about the mindf parameter duplication heres the correct example URLhttplocalhostsolrselectstylesheetqapacheqtstandardmlttruemanucatflid Ryan for spotting the fl param problem Ill attach a revised patch which fixes itBefore that the following request caused an NPE it works nowhttplocalhostsolrselectstylesheetqapacheqtstandardmlttruemanucat without having checked exactly how its implemented I think queries should work irrelevant of whether fields are stored or not as its based on whats indexed +It was much harder messier to do via string parsing suffixes as wed have to add a String based setter then do string parsing whereas adding different setters was simpler and lead to a cleaner can actually support things like limit Mb now that xbean supports a per property +Doh! Sorry just attached the right +I will work on a out of the box example to include in seam including some documentation expect an additional patch soon +Verified on revision +Should be duplicated qnameget not solution patch files formodulessrcorgapacheaxisschematemplateandmodulessrcorgapacheaxisschematemplateThese patches simply check for an element named qname when printing the property accessors and if it exists dont print the getter only the setter +Great +Modify our classloader so it can work with packed car files +I unset the fix version for now. +Log of both can you look at this + +When we found out brokers consumers are stuck it is already late because msg are already dispatched to the broker +So when run as a file in either or these test cases both work +and you start to want to sobte em va entrar una plorera de pensar en el mn que estem deixant als nostres thinking of the world +I have also removed formatting mistakes I had madeI ran derbyall and Junit suites and fixed all the failures I encountered +This factors the common code into its own target +also examples format should be consistent with generator +Will attach the updated patch +I really disagree with adding fake data as I said I think updateDocument should throw an exception here if the fieldinfos disagree +The patch addresses the filename issue +Integrated to TRUNKThanks for the patch Mahadev and Gregory +Thank you very much Jacques for commit and for commentThanks Sumit Pandit +Hi AshutoshSorry about that +This would mean that no stripping of special chars is required +I agree from your point of view +Closing old resolved issues +So it is fine that this does not work. +Your example contains notEqal instead of notEqual +I jumped the gun on the last patch +Attaching add additional test comment removing the check since now we check the permission directly from another user connection +Is this a problem? I dont think there is any escape from this +Implemented this support in Rev +Why is it crashing +I dont see where anyone is saying that using is the problem +Denis is able reproduce it too +Anyways I attached what my patch would be its unobtrusive +CheersClinton +Qunhuan Mei +Theres at least one issue and that is that the Voldemort client libraries are not in any Maven repositories +For ex +Resolved in trunk +I just linked a JIRA where an upgrade to in EAP is requested +Committed revision Ive filed for the improvements needed in order to bulk scans with holdable cursorsMarking this issue as resolved. +Minigems bug +Hi Nutch NewbieI will take a look at this today and take an action to prepare a patch +bulk close of all resolved issues as part of closing items. +One thing do you really need a second sleep here or could you do with some small number of milliseconds? Id think a ms sleep should be sufficient to always fail without the bug fix so I dont see any reason to have a test +The best way to go about improving this is to look into improving the wrapper API or alternatively the arq API to wrap the error +Resolving +Theres a page on the wiki for Anomaly Detection Framework I can make that say visualization too and the Swimlanes is already in there so Ill continue adding information about new pieces there +Its revision. +commit covering backport of all trunk issues to goraamazon branch committed revision in goraamazon branch +Please attach full thread dump if you find something stuck +Trunk is really simplified I will rather hack on it I am not attached toThanks for replies +Patch checked in +I committed the patch +That could be taken care of in an ant target I guessUnless I remember incorrectly this will also entail sticking Axis in the distro since the web app that gets generated exposes a web service via a +I just closed this issue +In regards to Dmitriys comment on Jython I agree that you probably dont want to start bundling jar for every scripting language but python based are specifically called out a new feature in the documentation and they dont work as advertised until you manually go out and put Jython on your classpath +youll only set it from +This was included in the release notes. +If someone wants to test this patch first have have to apply the patch fromThis issue depends on +We probably also want to give the operator the chance to add disks just so that we can bring the bookie back to its original state in the case we replace the diskbq +Tom White also has mentioned this bug in +I assume that at the point the getter is added for the property we accidentally replace the getField method +If implemented the naming convention would be more in line with the other properties but I think still it would just return +applied to revision Alexander please verify it was done as comment fixed in please take a fine +Is it really used for anything else than phone number extension? +Committed class that can be used to profile for bisocket memory leaks. +Attach a patch that allows socket timeout to be passed to getMetadata requests on the consumer side +Normally Java processes dont create other Java processes +I will submitt a patch in a week or so +Thank you Boris. +Why cant it be doneI have seen apps for ios with custom labelsWould really need this functionality +Both the integration code and the giranking as a dependency to avoid code clutter +For EAP CP simply adding cglib to JAVAOPTS bin did not change the byte code provider to cglib +Attaching a patch for +Simple test program to attempt to reproduce this problem +Ignore the patch +For example the could check which outputs already exist and not generate input splits for those +Strange that I do not see this problem as long as I build against +Moving to +IOW the failures due to spring CNFE are caused by the lack of the new spring jars in +Can you give an example of the subtle issues youre referring to? The advantage of HTTP is that weve already tested that code path and it supports things like checksumming etc +Committed +This patch adds a comment in the documentation about the problem tells the user to make sure to use exact quorum addresses and adds a new script to copy all tables description +Odd +Attached is some example source code +and there it is +If there is a problem please open a new jira and link this one with that. +AgreeI always use hadoops local mode for debug it is more similar to hadoops map reduce mode than pigs own local modeI agree with you that adding pigs own local mode will increase the complexity of pig and cost of maintenanceI think few people will use pigs local mode in production users only use it for debugging +Thanks for looking at it JD. +it should not shut down until explicitly destroyed +Great then do we have a winner +Closing all resolved tickets from or older +Hi SteveDo you think what Lukas is seeing here could be related to how j parses XML to HTML? I wonder if it is affecting those block elements that dont have an ID attribute assignedIf this is something to do with the styles feel free to flick it back to JDOCBOOKSTYLEThanks again mate +not a blocker +The next time you run into this can you please capture the RM logs and upload them here? Tx +So this is really not a bug +Patch for patch to attach patch for patch with the same changes for +We dont have a quick test case but we seem to be able to get it to happen within a day or two so I will try to gather the logging output you have requested +bulk defer minor issues to +Sorry to be troublesome Im just trying to think through the implications of the change +I should uses new MR API before merging into mumak +I guess this is the reason why it would be better to use constraint names correct? So hopefully this problem will dissappear inWhat remains is cleaning up the database +added test case as junit test and fixed other bug in related to handle xml composite components. +Done for marshaller. +Kind of scary that didnt catch this +Lets call this problem the snapshot replay problem and move discussion to a new issue +Note also the difference between running with the fix and without the fix for the ASCIIlatin Clob +Please remove the version numbers in the jar names in the legacy repository. +Otherwise its pretty close to being ready for a release if it can be hardened a bit +It appears all our test bundles import hiding this issue + +hive thrift service attached patch Look at the tests for usage +In we needed to handle the case where an was stored rather than a +You are in the group yes +Not sure if this is a dupe or just problem should be reproducible with the project in problem is that for some reason the Javadoc plugin is run before the other projects are built therefore as the associated jars are not present in the ocal +Sorry about that Gregory. +Reviewed latest patch +It looks like there is stray jar closing as this is working now in Seam fine +Rafael Jeferson Pezzuto Damaceno Please can you share our code for for version +More people watch the main list then the issues list and seems like something we should fix +This has already been upgraded for the FUSE ESB release +reopen to remember to look into this fix which looks like it will create an regression for existing XP SPJBDS id date EXECUTE Create Seam War web projectEXECUTE New Seam Input Seam component name Conversation press New Seam Conversation wizard generates java file with wrong class name +Closing this nothing I can do really in GRECLIPSE +Demo project with problematic bean in this is a the service interface is not known +Verified by manual check +Ive just committed this to trunk +Jagane can you provide the patch for this against the current trunk? I also think if we can do a simple version increment without doing any major rewrites that would be more likely to get into + +I will take a stab at part deux of the patch +thank you and Adrian +The password is no longer stored in XML so this problem does not exists in the branch +I believe these changes are sometimes causing a deployed application to disappear from the server view +one is to simply disallow overriding the methods you mentioned +The file being corrupt directoryIn order to reproduce this issue just start with this in varlocal answer the above queriesWere you trying to restart the activeMQ server after a crash? If so what were the conditions it crashed underWhat was the heap space allocated to activeMQ? Were there any persistent messages that were delivered when you tried to start activeMQ? If so please give the number of pending messages +Just ran a unit test using embedded with deployment that included an with a persistence unit that includes +Committed for Version +So hopefully that will happen in the next month or so +It may involve hacking the maven dependenciesPlease reopen if you have more precise informations +Can you attach the configuration file +Thanks! I guess thats good enough for now +Possible patch for you are rightThis is a very unpleasant side effect indeed +So it does not fit in well +Works for me in the latest trunk perhaps you have some old class files laying around somewhere? Thats normally the case with something like +actually this should be version renamed methods vote vote out get +I remember seeing mention of it on the list but must have missed the button when attaching my files! As for the API command +Galder if youre going to do this for RC make sure you are completely clear on and that your fix deals with and tests those issues +We need an calculation which is Cheap DeterministicI couldnt think of any except relying on like all efficient histogram generation algorithms are probabilistic so my optimization idea wouldnt work +Alright Grzegorz I tried Shared Library and it worked fine with CXFI would like to share though that I didnt place in shared library reference +I downloaded from Oracle launched Eclipse with that JDK and starting server no longer segfaults +Thanks Ning +Not pretty. +See +Done +I think our use of Subversion ensures that will not provide useful output anyway +Change to set on its RPC server instance instead of statically +It should be possible to wrap UGI and UUGI with something that uses reflection to determine what platform variant is below +Part of a bulk update all resolutions changed to done please review history to original resolution type +I wonder why it didnt work +Still do we have a test case that uses long strings +testcase based on your input outputs not in Railo and of my test test on win as results on Fedora with the jetty with bit JRE. +The workaround for was to increase memory allocated to maven to perform buildA better fix will come later once gets actually fixed and released +Fixed in Git ID cdeadeecaeaabe +Thanks Jacques! +Thanks Matthias +reopening to write release note now that release note text has been written +Is a test against an option? If so what template should I follow +This isnt a priority right now +Im not sure we should be doing anything differently +Had to trigger update for the boundary event after adding it to the diagram +GRAILSHOME needs to be externally set to allow choosing between multiple at the moment I start grails with a special to define and +I got the fb comment box up +everything looks good to me now +Please let me know if any more info is required +I think its the wrong thing to do +What groupid do you suggest +So with the way hadoop metrics works its not really possible to remove a metric +Patch was applied to CVS +I couldnt it because some internal stuff is used +Errors were caused by wrong namespaces and source folders in the plugin manifests +Thanks Zhenxiao +What mailing list for the notifications + + +Fixed in transition resolvedfixed bugs to closedfixed +Great! I will test this patch on patch contains some bugsIll upload a new patch after fixing the bug. +Would like to lower priority +Thanks Owen. +submitting patch and praying +Thanks +Does this happen consistently? The test passes fine when I run it on the jbosgi branch mvn false Dtestosgi testRunning Tests run Failures Errors Skipped Time elapsed sec +Here is a follow up patch for this issue +We won t end up with only files + +I see that it tries to connect to httplocalhostmapbuilderproxy instead of t that proxy an old one? How come it tries to use this one +Regression tests passed cleanly for me +Hi FrancoisDerbyall had one error when I applied this patch test under +My laptop is a HP Pavilion with Intel Core CPU and G mem +My argument is the locations of a file should be a attribute shared by all file types +Hadoop QA doesnt seem to have picked up the latest version +Box on the left has disabled +This got fixed in the +The http component doesnt support the ssl as the http component does can you double check your patchThanks for clarifying A and B for me +Weird were seeing this on +Test failures are not related to this jira +Looks good +Look under the history tab +Not a bug +You really shouldnt use anything earlier +The fix is in the CVS and should be available in nightly builds from tonight on. +This results in a few additional test failures which is expected +committed +Both the attached services can be invoked any number of timesOn the soapAction related issue if you want to use the soapAction you have to set the action in the annotation +Still needs another run through and review of the MR code as I havent looked at that in depth yetAll that being said this is getting really close +Hi allThis problem is only with maven +on the code change +We have some performance comparisons we should include that information on the performance page at least by the time this is released +The testcases modified were followingThe and tests were already modified before this patch +Simplified the patch +nginx is a reverse proxy not a servlet container surely this is a misconfiguration in morph +Bulk close for the release +Duplicate of +Still it would be nice to have the option to disable the hostname check on certificates. +We can certainly do that but Im trying to think where this practically matters + +Its fixed in all applicable releases. +Assigning to Tatu to close when is released +Does they differ +Committed to sqoop. +I try to launch Geronimo with Harmony +Henry Can you please review this patch +This is the exact same fix as what we used inThanks Sandy +In this light what Mark suggested seems good to me as well +Hi Uma thanks for the explanation +This is the previous full patch minus the XML file +looks like am really tired +The problem is that the presence of this adapter class messes up the way that both logging environments determine the name of the calling class and method +Closing old issues +Patch file for in JDBC patch for aloneOne patch for aloneOne full patch to apply on geotools root maven project +Additionally it would be interesting to know why you would not change the behavior +The current implementation is bound to +I appled this fix on fbb on master and bfd on branch +Thanks Laxman! Sorry for my late response +Changed default regexp to avoid warnings thanks for reporting +In some cases we do see differences +I was copying the Catalan analyser and using seemed like The Thing To Do +Its existing manifest will be ignoredIm ok with the generic approach just make sure to close the and check for header only +Why check the only when the last event is SUCCEEDED +Lets backport this to the branch +Patch is attached +What exactly would you like to see documented there Iwein? I would argue that referencing to higher paths from the classpath is not really recommendable in the first place +good couple of minor things The information will be ta +no jcr strings within the +Wed have to add a workaround like the one given on move the links to a garbage location to clean up on restartThis is pretty low priority for me but Id be glad to point someone interested in the right direction. +OK I took the path of less resistance and made sure seqnos are sent as monotonically increasing numbers so that we only need to check whether a seqno is smaller or equals than the last received seqno +Bug has been marked as a duplicate of this bug +Maybe you just have some garbage there +Also I find it extreme for Maven to add lifecycle phases within a point release +What modifications need to be done when migrationg from WFK to WFK with respect to GAV refactoring in WFK maven repository the from have been moved to +Part of a bulk update all resolutions changed to done please review history to original resolution type +sorry I didnt see that brett moved the issue +We could use the same protocol as used to canonicalize into to canonicalize strings to alternative is to remove cluster names altogether if the user can guarantee that well never see traffic from a different cluster we could drop shipping the cluster name in the transport header +I have fixed this issue in the fix it in the trunk as soon as CR is in the trunk +Damn phone replied to that previous email +I agree that that synchronization is unnecessary +Added code fixed now +c +Should we provide the XX setting in so that the value is a easier to change or b dynamically set to not be on the root disk +Daniel For the time being i have attached a patched wfs jar which will solve part of the problem and I think your specific case +The ones that were removed were not needed +so please create these two locations prior to integrating thingsassembliessrcvarsharedlibassembliessrcvarsharedclassesOnce again this patch was created on windows xp and should be applied from geronimo root +Attachment has been added with description PDF with against latest fop has been added with description PDF with against latest fop that new versions do not change anything. +This is quite a scary patchIm reading through it to try and figure out what its doing +Note to self check with Darran before starting this since Ill probably conflict with disabled for standalone servers in mode +The canvas is pretty clear +bq +Hani When you download the snapshot +fix in is in so this one should be marked as fixed? +This would be the next possible step but I would like to resolve for +Dion made a comment on the dev list at the same time I created this issue +The goal of this issue is to introduce a limititation onto the number of cooccurrences per item to make the runtime of and linear to the size of the input and not dependent on the maximum number of preferences per itemThis should be OK to do because you dont really learn anything new about an item after seeing a certain number of preferences and thus it should be sufficient to look at a fixed number of them at maximum per up the patch +Now if we change width attribute listWidth wont be changed in VPE +I tried committing between each ADD CONSTRAINT statement but this did not seem to have any effect +Proposed solution +Cleaned up a lot of code + +Minor edit to previous version of this patch +httplocalhostsampleCXFtokensxyzbytes sample project +Attaching the patch after doing svn up +Fixed in svn revision +This problem will go away once we move to gtxml and geotools trunk +The svn stat q is attached as +QA Verified. +Does it actually matter if TRACE is enabledI think this is the same as +Im just getting the expected output +Thanks Aaron! +Committed in trunk in rev +Neither I had changed the delete +We tried using the eventQueue requestDelay ignore as well as the Seam conversation timeout and Synchronized with a timeout +ensure this fix gets in the next fuse merged +Fixing this will be complicated though as all the server is based on the assertion that names cant collide +Though change is a very simple character change a patch might motivate someone to commit the change +issue resolved. +Russe Can you add something like the attached hereI have committed this to the open sources doc +No need for a patch I see it in the thread +ThomasI ve been working on the plugin source code for a while before going on holidays for the last weeks and I could made some changes on the source to replace the parent version from RELEASE to the last release resolved wasnt sure if those changes were alright so I didnt post them as I ve never collaborate on an open source project yet I dont know how should I share my changes +Patch +Issue has verified on both failure to load sstables wo upgrade using this tool and upgrade and load upgraded sstables +All updates are limited to convert to the plan like col to col and are carefully checked individually + +The method sig has the headers marked as OUT and not INOUTIf you modify the generated code to change them to INOUT the headers do appear in the outgoing soap message +Here is what I committed +manual cluster a patch details Add a new field finalState in and in +All applicable patches have been committed to trunk +I agree I dont think theres significant value in trying to constrain the source repository +The numbers at the bottom represent the average response time divided by the number of clients +You can set the socket option on the jetty component that can set these options +Cant you do this simply by plugging in your own Similarity class in Schema +But the Google collections jar is over k which would make +Changed fix version to EAPEWP +Yago Im guessing only because of percentiles and such +resolving because we need another release to take the patch and publish it since it is working in the apache repository fine +LGTM +Hive will be adding support JDBC over HTTP support +Thats fairly reasonable and in some of my cases this would work alright +That prefix describes what the data sources are in a positive way and it would allow us to preserve simple for data sources that implement less than required by the full JDBC specification +Fixed as part of. +Image view is completed +Verified by Ian +released closing. +Approved +Steve you have made me a very happy man Ive been procrastinating making that change myself +Hi TomIve checked the failing log and it looks like a memory issue +I always had the oracle stuff in lib + for the patch +Or or +I assume that you dont need SSL to work when the client comes from other IP addressesNo we only need ssl for these two addresses at the moment +limit changed to MB give it another GavinThanks a lot +Committed to. +This patch enables the user to select a Dynamic Web Project in the workspace and validates accordingly +The main issue has been fixed within have had to move some classes to the project examples plugin because it also uses the install action +Look at the CVS to understand how the preferences work. +Hi Clebert thank you for your explanation +Is there a specification API change associated with this +I think this issue should be closed and all efforts should go to. +Please apply this patch +Is this worthy? We can also have tests to ensure no daemons regress away from not using ToolGOP parsers in their main +Because the change could lead to incompatibility it will be slated for I got the point about whether its actually accessed or not but it seems like this is just some theoretical issue thats probably never going to happen in practice which is why I asked for the actual scenario or where this is causing you an issueWould seem strange to me to get to the view rendering stage rendering something conditional on the messages being there but still want the messages preserved after the next whole cycle of request processing has completed +Ill investigate it with our sysadmin and let you know. +Thanks Shai! +th + agree with David on all points above +Sorry for the spam +I just added the header to everything in rblib and rbext +Seems that JSON support is easy enough to add at the same time +Fails when there are spaces in the path +this has changed in so probably works or reports better +greatI will change the status and test that asapThanks you were right +We would have to redo the solr however and modify it with more data in mind +batch transition resolvedfixed bugs to closedfixed +I will take a look +There would be no change to the persistence manager! +Hi Sebastian may I ask question? I digged some old post and found that the best result should be RMSE do you know the parameters being used +Thanks for backporting Harsh. +You can see that this is the problem by placing a comment in front of in and everything will work +that solution requires an extra roundtripIf its just a potential performance problem I would rather defer it until it does turn out to be really a problemI see a bigger potential performance gain if would avoid calling branch merge for the default case +Could you please attach a sample deployment that allows us to reproduce what you are seeing +I think we need to create a new Map add the two to it and return the new Map +The docs you updated Musachy dont mention that overriding one parameter on a single interceptor nullifies any other parameters set on that interceptor +now uses a static method on to test for equality. +Yelp is a gnome help browser and it is the only package that requires on SLES +If we make such a fundamental change to the schema that Avros automatic versioning cannot handle it then we could change the version string to be or something +We should have waited to fix this before releasing ALPHA +Which exceptions will destroy a session +when I navigate to my uimaFIT project open the in the CDE go to the type systems tab and press add I can see a list of all XML files reachable from the that project via the classpath +In the case of an we are not calling next again +I did get the download ok eventually but not repeatably +On the other hand if someone wants to use some part of API I think they shouldnt have to resolve dependencies manually the first time they run on a real cluster +I will change the code so Lists instances are returned to the view so we can migrate to Spring +am resolving this as Wont fix since we cant do much for RAD since we no longer support on Eclipse +created let me know once there is a release or KEYS file in each release area so I can continue with the setup +Can you attach an example project? A patch would be even better +I think this is a good idea +For that matter I doubt that even JRE has any control over it +Im not sure how much attention its going to get here +Half a decade ago. +Validator send labels keys to groovy tmpl instead of send labels values +See class in the same packageI added that class more for demonstration purposes and wasnt aware that it would cause such a fuss since it does not do any crypto by itself but merely calls existing JRE codeAnyway I wouldnt have a problem if we removed this class instead as I have stated earlier +Any objections for me doing so? were no objections so Im closing this issue +Im on jsf spec section for information about this issue +Moved manifest generation to the agent +I spoke just half of my mind there as alwaysAs you may notice any parameter for the scaffold command will have to be considered by all scaffold providers and it may not make sense for some of them +Ive marked that asTIA be better if would be committed first +It does not cause any data lose except for some harmless deletion warnings at datanodes +new patch merged with for the quick reply Bennie +RESOLVED Duplicate Issue +heres the result of running re run this install findbugs and run ant homeandydevtools findbugs +I think we can avoid this once we can do a release containing this fix +Thanks +I had assumed that commit tests are a subset of smoke tests are a subset of are a subset of test +Uploaded another new patch with a minor change +Weve also added workaround for iterator chain bug to check if the problem can still be seen on latest ajaxjsf richfaces snapshots +Anyway just wanted to let you know I can reproduce it +I think key value pairs are too generic and can get very confusing to make sense of +Jeceira is under Apache license only Exo is GPL +To comply with Release Criteria stating that no issues with Critical or Blocker priority setting can be open for the release this is changed to priority Major +targetbin +findbugs discovered problems are not included because findbugs crash after finding about problems +So probably build and deploy needs some changes so that classpath includes HADOOPMAPREDHOMElib RajitDistcp command will run as Mapreduce job +The patch for bug submitted by Manfred Klug resolves this issue +Ah you mean THE OTHER +we should implement IMPORTUUIDCOLLISIONREPLACEEXISTING for jcrroot properly rather than skipping jcrroot +read the docswe need to load the object to handle all lifecycle events etc. +Dump of file File Type DLL Image has the following dependencies Summary +Otherwise if its a situation I can modify it to use regular inline css +XLSX is worse than XLS as the xml processing into objects takes lots of memorySome files are worse than others depends on the kinds of things in them +your project having a dependency on testng should automatically pull in junit as a transitive dependency +Works for me on Windows compiled from master +need to review a couple of classloader we need to restore compatibility with version retain the fix above see if we can improve the classloading situation for things like so people dont need to create their own classloadersthe original bugfix for was need a test case to verify that this is an issue +fwiw I checked hbases and it just synchronizes completely get put contains etc +Ill reopen until we fix this +where is the quickstartis the component path you see there really not on the page? what is wrong with that path +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Ales has committed his VFS aware scanner +removed commented code +There wont be a release +Hi Cos that is correct in I created a new JMX bean to provide the same information. +Item is tracked by +It could go in I think since its +As already mentioned in comments this is because maven expect to find to when compiling java code +I will rewrite the check +Truncate xy to integer +Verified against documents generated from SVN +moving out for now since an improvement + +Thanks KarthickHave been away this last week but will give it a go as soon as possibleThanks again for all your work. +updated in branch and please cross check and close the bug +Protected is just like private however only subclasses can access protected methods and variables or in the same package +Im attaching the eclipse project Ive been testing +These scriptsshould run as against a fresh test migration +Could you please attach this? I attached the fileThx my guessassumption is this is failing because JSF actually does not get registered since no jsf jar is on classpath +The issue is that the ends of the channel may be configured differently because of parameters that are only relevant at one end +With infinite loops I mean cyclic aliasing +Does the declaration order affect matching I cant think of why? If not there should be no issues and using the declaration order to resolve ambaguities in matching would be very useful +Thanks! transition resolvedfixed bugs to closedfixed +Since this has no own implementation but relies on the itself and content extensions there is also not a direct hook to activate it anywayIm reverting to disabling the filter by default so it doesnt produce confusing results in the default Project Explorer but it will remain an option for the user to enable it in either the Activiti Explorer or the default Project Explorer. +Any additional comments re the second patch or can this be committed +See + +Steps to run offline install or more JBDS versions into some folder eg +There is something wierd happening with the with the latest patch will take a look and upload another +A niftily removes the need for collection +as a base component to the +Ive also tried linkcheck and the bug is still reproduce with maven please upgrade maven +Is this tool useful for printing and understanding indexes +Thanks a lot vinod! +On further thought the no time travel guarantee is actually a bit tricky to implement in the face of failure +To save us debate I went ahead and tossed the test case into as well +Lets try again at the real size +Youd only need the board resolution if the PMC is being dissolved +Lets see with Jonathan +Just make sure your spring context file just imports properly and doesnt try and redefine a bunch of the CXF internal beans +Sorry I thought Id been clearThe reason it wasnt working is because the pyramidplugin was not bug is that rather than tell me the plugin is not installed gives me various forms of gibberish. +I am on it +However Im unable to fix the redraw issues. +Not sure if we need to change the wizard namesids or anything else but it seems to work fine +But its browsers so Id double check that its still conditional +Reproduced with HQEAPGABUILD running under EAP but so far I have not been able to workout any pattern behind the failure I did not see the problem with my first test runs but after than it occurred seemingly at random +Should try to include this as a sample app for pending to review for Sample Applications +It was originally as I had only considered circumventing aggresive relase due to open and Iterators + +Thanks for the patch +What issues do you see when you try to apply it? Im new to the patching game so Im not ruling out a basic error on my part +So it was working in yes we need to investigate at least +We do due diligence and not step on anyones toes in terms of library compliance and playing nice with globals in JS etc +This is the last sequence number of a replication target graphed over hoursIt shows that restarting replication gives a speed boost and that after a while the speed diminishes +How strange +Translation needs doing still. +Ive added multiple lines parser and console history +Please create the smaller you feel are necessary and I will pick as and when I can +project file +its one of the few strictly required configurations so if it isnt defined theres a much bigger problem +Having been out of town for the past week I am just catching up so sorry for not chiming in earlierTo be frank I dont see the point of making exec source parse shell commands +Can you try with a more recent version of CXF? is long since unsupported +Done! Thanks +Instead of building a complex audit data management system I suggest making a log tap that sends audit trace to syslog either local or remote +My sample will generate a type create an instance of that type then display the contents of the instance in both XML and from the data object directly +bulk move in after release. +Heres a patch that defaults NUTCHLOGDIR to HADOOPLOGDIR and NUTCHLOGFILE to HADOOPLOGFILE +Sorry Rob +It is a grand idea for you guys to work together +We could even improve our current code +I remember having a good reasonI remember the problem was that there were some assumptions baked into the gwt jar that did not work for us +Ran with part +Looks like a missed failure mode +Lars how about getting into +I have changed the code + +I will handle this one too +My guess is that in Java the constructors are ordered differently resulting in us choosing the public Stringintint version instead +patch applied +I guess for what you are trying to achieve it makes more sense for system properties to have precedence +I will sync with latest changes and submit a patch shortly +trunk Thanks for the patch +fyi I eventually ran out of memory so there is a leak somewhere +Sorry again and thanks for taking the time uploading your work +Can you redo a patch where you attach the movedrenamed files as plain +thanks dude +Should be okay as the difference is not dramatic but seems to remain within the relatively narrow interval +I cannot recreate this under any OS Maven version combo +Reporters of these issues please verify the fixes and help close these an issue anymore +If clearingwiping cache contents was external to all of this none of this would matter would it +It makes debugging unnecessarily hard when your root entities dont reflect what sub type they really are. +Fixes compile errors when you use Java javac +Forwarding to a JSP page as you discovered is one way around this issue as long as the JSP page does not attempt to access request parameters directly +After some manual testing Ive verified that this greatly improves the problem but does not completely solve it so Im still stuck on this +This jira has been closed as part of a review and archiving of the aj project +Modified the with patch +Verified +After rendering the input of component a will show the value a instead of empty +Ill then find the time to have a serious look at it and give some feedback +ok +How about adding a timeout to the monitoring method such that it throws an exception if the desired end state is not reached within that time +I just added you as a wiki contributor Joe +From this we will discuss next actions +may introduce new models for the management of content items within Contenthub +I believe that this applies only to injected not to those created by the user explicitly via +Thanks Holly! +At best Id like to move the two private methods to both commits are fine Id close this ticket as fixed. +i just test in the setupRedirect of if it is an ajax requestif it is then the intercept url will be that of the ajax request page instead of the ajax url +Steven yeah i know theoretically why this is here but it is completely stupidDeprecated is a gigantic mistake requiring me to write it twice once lowercase in the javadocs once uppercase as an annotation +Thanks for the review +I have applied the patch but havent tested it yet will do that later. +Took care of Toms suggestions +Thus why I said suspected above +Affects the codebuild so needs to be in Beta else will not make it +These are probably all obsolete +I accidentally uploaded a patch I created before running the tests where I forgot to escape the character +Ill take this on thanks Andrew +The Security implementation from an IBM JDK is a little bit different from a Oracle JDK +for patch +can someone tell me why the concept of a core is inA core is a solr instance +Maybe this was my fault and Ill test it one more timeSo far scheduling for +one more minor annoyance when doing grails upgrade on a plugin the Upgrade script generatesa bunch of files that the grails Script does not generate +This is in fact something I didnt look into +Updated patch +Actually I toiled over this for a while + +Fixed! Thanks Benjamin +IMHO this is caused by all SMX JMX having different names when running inside +So that makes more sense nowCreated for this since its not the same problem that Marcus is talking about in the description here. +Hi AndreaI would also be interested in the codeThank you +Will file issue to reenable after it gets some lovingChatting w Matteo he suggests that there may legitimately be a window when no files in the store on initial creation +If the core doesnt load you cant use the new editing code +Links are fixed many thanks for the report +We have been using for sometime and are really liking it +Santosh are you still working on thisOtherwise I can take it over +The from the patch wraps the parsed query in a +Will take this discussion onto the mailing list +does anyone know who jfc isAlso Jan patches are a lot more effective at getting things fixed than commands are +That would be a long term solution +It turns out that this bug can manifest itself as data corruptions +I keep this issue open until somebody confirms that the fix is working for Solr +Anyone is intressted in that anyway +Need to review first how we get this information consistently out of Remoting and the HTTP server +Do we throw an exception or return null? As this is not done anywhere else there is not really anything else to be consistent withReturning null would be very uninformative as there could be various reasons that has happenedThrowing an exception would make most sense as at runtime the developers can react to the error on their half for not checking that a doesnt already exists using getSo for now I will throw a with a message describing the problem +Splitting out into a jar of its own is exactly what we did with WSDLJ +Are there functional changes since the patch was generated that will alter the implementation +Here is a patch which allows a configurable no +they are back +Documentation is produced as part of build example make install for httpd creates the documentation as final documentation for windows also has DOS line endings which makes it more familiar to the windows would leave that to our customers to decide weather they wish the single download package ormultiple packages that would require the careful drop in inside the installation tree to have workingand browsable documentation +It may be that it should be changed but it must be tested +Thanks RobI put this into trunk and +I just committed this +Bump can the reporters of this issue try out the patch +I ran NN on my a single node cluster on my laptop with trunk and with my patchThe average opssec on trunk over runs was opssWith my patch the average was opssDetails +Also in order to enable executing these comands from code SSH needs to enabled on the ASA appliances and the credentials needs to be stored in Cloudstack databaseFor now the steps to reset ASA to factory settings needs to be documented and if there is an issue while reusing the ASA appliance in a new isolated guest network only then should this workaroud be applied +Can this be fixed in +Fixed by +Whoops yeah sorry +I attached a new patch +Fixed in investigate with Reimers for investigation +To add another piece to the puzzle such as HDFS would break this and diminish the value +I see now thanks +Deployables +I believe so again for your attention. +Verified on EAP CR +Fixed the check to use as suggested by Pi +For toInc we may not change it at every step so if theres a missing arrow that means nothing should be changedThanx for checking outOne thing that Im not sure about is it better to adjust current and this functionality to it or add new +makes sense forgot that those API were new in +The problem was that if an empty directory happens to be in the end loadImage does not know that it needs to continue to load the empty directory since it has already loaded all its files +Moving from JBRULES to GUVNOR +This behavior exists in the current code so its not a new problem introduced in this patch +Barry weve recently merged a number of fixes to the DDL parser +The change isnt obvious to me can you help be understand it? Please describe whats being tested to determine if the code can return without any change or exception +Without that I cant check it in because you havent granted Apache rights to the code +Can you work on top of? To not hit performance issues can be changed to remember only say applications by default +Whatever work I produce I will attach a SVN patch to this ticket for review +This is not a simple upgrade +Currently itll just return one the oldest but flushing the oldest might not be enough to let go of the oldest WAL +Please note that a similar ticket already exists +I dont believe this affects developers it is a change which reflects a preference for using setter methods in the jBPM configuration rather than direct injection of the field values. +I doubt too many people even have that many custom algs out there but thats not a scenioro I want to help and try to perpetuateSo its not like Im sitting here saying this is a huge deal but I think it should def be considered a bit +This looks quite good indeedAny idea about a release date of the milestone? A rough guess would be enough. +Thanks DeepeshI have committed a slightly modified version of your patch and it is in trunk rev +I dont expect it to be much but itd be nice to know +Attaching the patch +bq +After all the ability to configure and disable this property is not harmful but adds more flexibility +More environment infor plugin version repository is automatically reindexed and index is also updated when artifact are being downloaded. +Ive been waiting until you returned from planet +LarsAre you working on this issue? to work on this for +Committed this to trunk and along with +Please if its still wanted. +Well put that to work thank you for submitting the issue +So the question would be what is a reasonable thing to store where none is specified Null? Empty string? Unknown +Versions are cheap and is much easier to manage modules with the same version even if theres no code change +HiWhat is the reason for this exception? and how can I reproduce itthanks is an issue that should not happen +Do we still want to do this +committed +Chris Im for your v patch thanks for the helpHey Nicholasbq +This is simpler than using +How so +Targeting target versions +So all issues not directly assigned to an AS release are being closed +Currently on vacation but will look at it when I get back. +Your test in the previous comment runs without problems against current trunk code and the branchThanks for the clarification + Looks good to me +Manually done. +I made a comment above but the committed patch still had them in there +About code duplication You can in the hidden ctor of the immutable class make a clone of the builder and keep it somewhere private final inside the instance +Do you want to commit your patch? If it is not important for you anymore pleaseclose the bug +Minor fix to the latest patch to put zhyzlib back on the list of libs in the awtgl makefile on unix +In the case of a feature request please discuss it with the Tapestry developer community on the dev mailing list first. +The current comments state that it can be used with BUILTIN authentication but now thats only partially true +Looks like one of our mentors will have to do this as you suggest +Build and tests run smoothly therefore I have pushed the reviewed fix to the development buildAdjusted the contributer of the patch manully back to David Regards as fixed +ie +If we add support for batch I think we should leave it up to the user if he likes to execute things in a batch +what version of Tools is this on +by object I mean object type TABLEVIEW a full example could look likeALTER FOREIGN TABLE foo ADD cardinality ALTER COLUMN bar ADD updatable true DROP somethingSince the COLUMN keyword is optional we could just leave that out initially and also not expect a keyword for parameter. +The problem is it currently doesnt +branchesJBESBCP +The creep so far is just which we were forced to add back because Tika HTML parsing was totally inadequate to our needs +fixed in i hope +Heres the patch for +Im not yet convinced the proposed behavior is the correct one +I am attaching a patch for this to get some feedback before committing it +Why was the fix moved to? Will be always a developer only version? If will be released for non developer users then please move the fix version back to +the file is pdf with bar codethe files and were generated by transparencyopacity wasnt taken into amount when creating the colormodel for an indexed colorspaceIve added a fix with version . +This could be a problem in MR on YARN so it would be OK to look at this separately +If you change the version to you should get past the problem +Simple patch removing the option and improving how shapes are handled in +I just need to fix the package structure to match +Vinod The patch has passed the test twice +Committed revision Thanks Kasun for the contribution +Ive updated the patch to implement the requirements +Merging the constraint violation for many different mutations may be a little funky too +Never mind +ready for version w better replacement of tests and updated logging +The problem is that the code assums to combine only in the case where there is only one which is the only input to an aggregate function +No need of idempotent for any of the methods +Great WilliamIs it possible to provide these tests as a patch would be great to include them in the source +HelloWould you please try this patch? Thanks a lotBest regards any data to show the difference before and after the modification +I have no idea about the detailBut i want to stress that it was only suspended in OS whose default language is ChineseAnd if the this line is delete it works well again +Reassigning to +fix on r +running tests +The problem was not reproduceable a was thrown not aObviously this bug was fixed some time between build and Im not going to change code in as the solution is to upgrade to a more recent JDK on Windows fix is to upgrade to a more recent JDK when using Windows +Wow! Graeme that was even faster than JuergenoSouthampton is better than finding a tarantula crawling over your porridge at breakfast but not by about the rubbishy format that I punted these typos to you +Hi SaschaYes great that you found that too +I agree with BurrJust to be clear Right now the warning is displayed but the project can still be created +But that would be a useful next improvementIn exporting a UDT column it is serialized first to a byte array +Thanks Martin! +I presume this bug can be closed since it was closely related to that is fixed now. +Why not work towards a long term solution instead of applying a to the current problem at hand? Step back from the tree and see the forest +You could add them to but due to readability issues people are VERY unlikely to ever use it nested annotations are hard enough to read at the best of times and there is no need when you can put more than annotation against a class +Nothing +I just committed this +when is it safe to make one? Youll have to iterate on it to create one +This patch modifies the vcproj files for the SDO runtime and test projects to provide two additional build configurations that use stdcxxThere is currently no documentation to explain the settings required or how to use these projects see +But it does show the danger of custom classesThere are also a number of nasty issues related to wrapping of +Im resolving this as Invalid based on Alexanders commentNote that your comment both users have same node id in cluster configuration in sounds like a major misconfiguration +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Where can i download a new version of jasper +The ZK expire test was in the wrong place after the change +Currently we only have characters so if people want to set up different levels of delimiters they are limited +I cannot extend with specific code as this will create cyclic dependencies +Well spotted regarding Exceptions +Trygve could you see if this also fixes the problems you are having +This patch adds hsqldb to the dependencies of +Instead a destination notifies the name node that a block is copied and a hint that a replica at the source should be removed if the block is over replicated +Chris setting this in will work for RMNM itselfHowever the containers dont inherit NMs environment thus well need to explicitly do it +Robbie could you please review the attached patch and apply if suitableThanks comments actioned please review +Thanks to Greg and Rainer for deep discussion and patch +Thanks SteveThis is nearly identical in intent to +Missed a conflict with +Ive revisited the bug fix again and I think it was still not working +In this part of the code clearly is not requiredIf you have a better way to implement Im happy to hear itbq +Btw please run dosunix on modules module after committing the patch +Uploaded Patch +Thanks for the patch +If not reopen +I think we follow the style convention of having braces for even if they are one line forgot to add the set of decommissionedTrackers is incorrect it needs to be done outside of the for loop +Huh +I dont think we can use anything else than INFO +Verified on EAP +This was actually already fixed by someone in HB. +Seems removing fast and then enabling FORCE worked +This would involve someparser changes +when are the nightlys generated anyway +I added a check that it only does a refresh when the event uuid matches the view asset uuid +I have a similar issue despite removing the incantation in the context +I understand there can be issues with displaying internal Jenkins status on a public site but since the public jenkins instance includes tests results its all good for me +I split part of the patch to +Okay good point +I can take a look at the original patch and see why it blows up on the AS codeThanks take a look at the above and let me know what you think +Reopend as a reminder +And now all of them should pass with RI +My main question is has anybody used Enumerations already + +Okay Ill wontfix given that we have for refactoring client utils. +Im unable to get parent proxy to work either +DustinIs there a patchthanksI should have a patch put together pretty soon +Ciao Simonedoes this mean that for now its impossible to have transparent background areas for RGB rastersSo in order to have background transparency we have to convert rasters first to RGBA? this means that the mosaic by default does not turn anything into transparent +sorry I meant GMT +BillI havent tried the XX yet +is an earlier issue which was incorrectly closed as duplicate of which was wrong IMHO +This looks easy enough to fix without it anyway +The mail I sent to the groupI also had this problem with test case once in a whileSome of the MTOM test cases are also failing occationally due to this connection reset problem +I use the following command line params wsdl httpdcdevserviceswsdl o +Define the request metrics in a separate object and remove those from the Request object +Also please provide from your configuration what you have set to the parameter to? any updates on this +it is an sybase issue please report this to sybase +Committed revision +More details on which was closed as a dup of this bug +ok +I verified if I have write rights on the server +Resolving +This is not a bug try to add the attribute immediatetrue in your cancelButton and then the values wont get updated more commentIf we remove immediatetrue converts data in the inputTexts and if there are errors cancel button doesnt work again +Ok then can you provide your CQL table and index definitions? And maybe some excerpt of the hector code used for insertion? Are you using COMPACT tables on the CQL side? And are you using CQL for the insertion with hector? Im afraid well need a lot more details on what you do to be able to help you since youre the only one to be able to reproduce this on so far +Rebase to lastest trunk previous tests should be fixed. +I still havent had a cahnce to look at the new one in but i wnat to clarify something you said +For exampleset PATHJAVAHOMEbinPATHexport PATHJAVAHOMEbinPATHThe parts about setting PATH should come after instructing the user to add DERBYHOME to the pathIt is important for the success of your installation because the PATH variable enables your operating system to find the scripts from any directoryTo check that the JAVAHOME and DERBYHOME variables have been set correctly and that DERBYHOMEbin has been added to the PATH environment variable correctly in a command window type the followingsysinfoIf the environment variables are set correctly Derbys sysinfo command will print information about your JVM and the version of Derby you have installed to the console +So +Ben can you take care of this for me? Just go through the Messaging documentation and removecomment out information relating to and as above +Jira should not be used for this please use one of the user mentioned by DennisIm just trying to prevent that this will become a discussion board. +Right beside their corresponding runtime +I just realized that the current discussion diverted quite a bit from the original subjectI will create a separeate issue about the usage of head and in this issue we can focus on readmergeNodes in +Its able to get most view names and all the textMy hunch is that the file is record based with eeef fbf being the type marker +Patch updated to latest svn +I also removed the Am I Eligible? section and will provide a further patch that adds it to a more suitable page +When determines the accessors to bean properties it finds this method and ignores get and set so further writing to this property is disabledOne possible solution is to delete is method since I dont find any usages and use and use methods are very similar +I ran this change by Sunitha and she says it looks good so once the Manual is sent to Jean again by Jeff please check and close this bug if everything looks good +One more thing we should consider do we have to add a trigger in the configuration to enforce the check or a trigger to disable the string checking +Or handshake at the beginning of a write we already have the DN send back a +I change to in the patch Ok +Bulk close resolved issues +this issue is outdated since it was reported against the POC implementation for an Express server adapter +I cant reproduce it + was perhaps intended to show short string or numeric values +removed formatting +patch available in both phabricator and am not sure if this indeed is the root cause because this is deterministic cause and the test cause would have always failed not randomly +Furthermore could you please try to run your RADEclipse with a Sun JDK just to test if that is a JDK issue +ThanksThanks problem. +I havent started to work on it go for it +That has two paths +attach the most recent version of the patch to this ticket +I have merged this to. + +Merged into develop. +Is this issue still validCan we close itJacopo +But I want your confirmation on that as well as lightnings +Im going to narrow down the fixing commit but it looks like this JIRA can be closed +My recommendation would be to update your SVN client and see if this issue remains +This is now fixed +Ill have a look at this for +could you send me ? +v +Please update the accordingly +Is it really hard to add a config change that requires absolutely no updating and is back compat +Only I saw a discrepance in Eclipse java project code as compared to the +Related to +Im planning on at least one more release before +Hi Jens Im sorry but were not going to investigate this issue on a Jenkins version which is almost years old +We also have this issue not only in Continuum but in commandline Maven execute during the build to download common resources from CVS +Please change to if you intend to submit a fix for. +in these screenshots describes the behavior of the bug First request to node that content the core works fine +Adding ticket URL. +Or if we work concurrently perhaps youd like to review my patches so it doesnt clash with your designsThanksJoseph thanks for the heads up JoeIll be resuming work then in this section then for your reference so i dont clash with your work you can refer to and for updatesJoseph as closed +Ok thank you for looking into it +Hmm but this is a bit strange IMO +Verified on JBDS. +Go to the Table Edit page from the toolbar +Ive uploaded the second patch that additionally adds ASF license to some missing files +HiThanks for looking in to this problem +There should only be one place on the internet for the project blog at any one time +Now it may not deploy correctly because the structure is different but you can do all the fancy Eclipse things like get autocompletion for JSF component tags and EL expressions and even edit the JSF and Seam +I still see lot of files in modules and applications using old header with the copyright in it +Thanks! +This should avoid any such potential conflicts with other versions of CGLIB that may be around somewhere in the server classpathSpring Framework GA will be released this week so its just around the corner +Please check it +the EXPLAIN output looks wonky +Any progress on that front +Jonathans patch would solve all the issues Im facing currently including reading Python serialized objects from Java and talking to Twisted servers from Python clients +The same problem is observed in build above analysis makes sense +The only way you could get this exception is with the teiid api in your classpath multiple times +Updated patch attach +As for Im leaving it open to be closed as Invalid at your discretion +LarsThanks for your commentsMy consideration +Could you please post some further details on the user forum linking this issueThank m using Tools for development and it looks like Tools didnt realize that my directory changed +adoes not work for the URL httplocalhost servlet is not called when both filters are used saying Not Found in browserThe same projects works fine in +I dont like it but dont know the API compatibility rules well enough to know if it is required or is ok to change the APIWill as is this patch does not let you dynamically change the core +Ok maybe we can clean that up in the future +I was able to create new requestSelect hbase for RepositoryEnter for Base DirectoryLeave Bugs field blankEnter hbase to Groups field +I see no reason why we cant have other versions similar to the way other projects do it inI see two issues that have been discussed and would like some feedback on +The fix for fixed this one too +this feels like the right approach lets get moving +Will check with Jonathan +Attached is a modified class file with necessary modifications in the parse method and a few additional set methods to handle additional object types returned from Custom Document Properties that allow for all MS Office Custom Document Properties to be set and returned in the Metadata +Committed patch to documentation trunk at revision +Dont disabled the Save button when the user try to save the file pop up a warning message dialog tell the user that the file is not option is correct or is there any other options for it +I dont see the introduced state either +OM is for XML only +The plugin now uses BROWSERCOMPATIBLITY Cookie specification which should be flexible enough to accomodate most common cookie misformatting +use dependencyresolve to see whats going on +Please provide a link to the SVN commit that fixes this issue +Basically it scans all replicas of a block for at most times +I believe the problem is due to something in +Done +Tested to work +In fact in such case the resource kee should be encoded by Sonar +on the patch +fixed in svn rev +Could this be minimally deployed to andor released? I suppose theres not a lot of demand for it but as JB made the effort to get it resolved it would be great if it was accessible +Do you have an affects version for this issue? Description says it started when a for started but it affects so Im confused? believe for the issue that was fixed here it originated in and was present up through. +Also when starting a Hadoop cluster the Amazon keys are part of the log file +Adding the parsed jars to the conf classpath +Thats not acceptableBut your use cases are exactly why I said I would like to allow an SFEMF config variableAnd btw what you suggest is certain verbose but verbose! messy +Actually it is really a duplicate of the underlying phenomena described on a default modifiable setting for a session was not implemented for +r partially mitigated this problem +So Zubair you wasnt picked +Patch that restores debug logging to the mojos + +Can you prove this? b +The CTF seems to have five problems all of them in the same area +No more need due to +Thanks! Explicitly supressing namespace stuff does the trick +I ported patch and created a test that is similar to the ruby script mentioned above +This has a few nice properties primarily that MatrixVector are not quite so coupled to Hadoop +Yes it is broken +attached closing stale resolved issues +the patch and updated version of +Thats as designed +That is a bigger issue we will have to fix down the road +works for now. +Were going to implement real inlining for trivial methods real soon now so this workitem is obsolete. +Documenting as supported for iOS android for now +Would this be better commented out so it can optionally be enabled +We are about to ship of the plugin so that s more than welcome and has been other fix +The included test passes in but fails on trunk +FIXED in head pls OK. +HiIm sorry but I dont understand what your comment aims at +the dvd example project does not seem to follow the standard layout so things will fail +I just can reopen it Somebody else has to close it +May have to wait for Maven or maven is out is there any updates? +Thanks for reporting Ill try to get to it asapKarl did you manage to look into this? Im willing to test any patches if needed +This patch uses the from all paths sets the class loader in most casesIt also adds the addedJars to HIVEADDEDJARS so that can add it into the for both the in the conf and the thread context +I dont have time atm to submit a patch +Let QA run once again +Yes I saw that too +But set set fixed value and afterward get doesnt use and getter for backing bean property is not called +to and trunk. +Sorry for my uneduceted guess but maybe you could do that based on of extension functionIf it returns a retain current behavior with recreating nodes in separate document but if it returns or hows that just trust its content nodes to be owned by processed documentAnd also I doubt that would be much slower considering cited core creates new Document and possibly for it in the process. +Modifies file to be servlet independent +Also the reason some code is duplicated between and is because the code might probably go away in the next release or so +Hugo can you check on this issue? We are trying to close criticalblockers on +Hudson has emma plugin +on committing the patch +Applied the latest patch +These links will work once they have finished that setup +Same fix should be applied toOoh Ill fix that +Fixed in trunk +Two minor things sourceDirectory and test look like default values anyway? there is a newer version of jsonic in maven repositories dont know if this matters at all +As of the Software Grand is already on file +This is fixed +Its not fixed yet though but Ill comment that on +Reviewed by Alan +vGroovy Version JVM yes its related to the trailing JAVAHOMEcProgram FilesJavabin +Looking at it right now +Same log messages as tb posted +We agreed that displacement is not the way to add spaces around symbols in fills +It takes a lot of persistence and attention to detail to localize for another languageChc mngPaul Jacques PaulPlease reviseMerge java application raises issues so after using tool for merging we have to Diff by SVN and Action by HAND +It doesnt remove the currently used new goals so they can for some time before news are removed +Thanks for the dirsWe are just from the incubatorIs it OK if Ill upload our first release candidate to the dev in about a week +I thought this method is the best out of both considering the fact that cartridge implementations will differ a lot between carbon and cartridgesWDYT +Yes is true but It is the same that property binding +As recipes are an important of we might as well have a library of primitives that can serve at least as examples for application developers +Our whole delivery process has been designed with this feature in mind +This seemed to work for me +For the EWP version there are still references to we need this guide only if we ship the associated tutorial along with the product +Ok we will wait on Chiba for Maven believe that Javassist is under both the MPL AND the LGPL +SSH action is back +Im not so clued up on the Avro Java API and therefore it would really help if you were able to either point me to a diff of the changes within your fork? Did you have any problems with the upgrade? Thanks very much in advance +Hi KonstantinYou may be correct that the new code use more memory however I beg you wont argue that the new code is simpler than the existing codeI will think about how to reduce the memory usage +Fair enough +Please verify pojo and pojo as examples of usage of that will be hard +If these tests accidentally worked before this might have stopped it +Drat +This was a regression introduced earlier today or yesterday +I believe we have to do it this way so as not to have any licensing finished +Ill try to find the steps to reproduce + must set the system property true if they want equals method to compare only +But this shouldnt require everybody to inherit the dependency into their apps +Committed the suggested fix +You probably need to run javadoc and grep for in the error messages. + to +Added into our customized build tested locally and seems OK so try it on installed into in STAGGING approved by Mark will be rolled into PROD +Could you please be more specific what we need to do exactly to reproduce this error? Would you be able to attach an analyzer that exhibits the problem? Also what makes you think this is an incompatibility with Java does the test run fine with Java ? Thanks +We have to have a resolution to this to switch to SLFJ +Patch porting to new javadoc warning +Please create a new issue +Thanks +Please close this issue if it works for you +Looks like the build is failing for the same problem even before this commit +Yes cumulative is best +That being said this application appears to have serious performance issues and that may cause random issues on devices to to the lack of memory +Dependency on lib used to be there but now these jars are pulled by ivyIm not sure if this is the best place for this discussion +Closing it as duplicate of +Continued as +The Ant task now takes configuration options for all these. +Looks like lot of blocker issues start showing up because ofEnis Great info +Both these solutions work +thank you you think about showing time for executionfetchink +Instead of introducing new synchronization points wee should totally rely on the existing locks +I think Todds right +This is by design +added facade class to insulate Nutch from underlying mime type detector Tika cleaned up static refs to in Content upgrade to +So you can grep the log files if you +Thanks Dmitry +Can I clarify that this is not a request to change the replicate endpoint but to modify the Futon UI which should take far less time and not break any compatibility +Is it a good idea? Or is there any other solution? This problem is very critical to us +What do you think +Marlon you are correct but I think this might be a bigger issue that deserves its own ticketGiven that all the subtasks are complete and there are additional profile tickets I think we can call this ticket done +This is a general tip to help progress +Applied patch r +The above trace is for hadoop +a related note is there a checkstyle definition for this code style? Id like to configure my builds to fail if the Sonar codestyle is violated +Committed to. +Next time I work on hive and cause same problem I will try to provide more details. +Also there will be a little bit more documentation about the peer nature of the dispatchers +Cheers. +Although its odd that only xerces differs from its xercesImpl May take up to hours for the fix to reach central repo +jstplugins +I also tested this on a cluster. +Leif Is this issue plugin component +I zeroed out the FORGEDMUAOUTLOOK rulefor our servers +Heh Ill try to keep renaming to a minimum now This patch is the remaining diff to my local copy +What if there is a collision +tks +I didnt review the code just now but I recall last time I looked that blocks are cut off before overflowing the block size rather than the first KV after they overflow +Weston will you handle this now +If there are a small number of identity hash maps in the bootimage then we could make them work by having the bootimage writer traverse their keys and build up a sidetable of objects in identity hash maps right before it starts writing objects to the bootimage +For instance an object inspector should have a type info instance in order to correctly interpret the data and enforce limit +Moving to Mina hence closing this issueThanks. +Do you have any of these failures that include a line number for where the null pointer error happened +You can customize the behavior governing the addition and removal of deployment scanners to a server through the Server editor +Turns out that it only triggers when a partial message is received so we only saw it when the messages were very bigFor example if the message is bytes the bytes will arrive to the c client at the same timesize will read the size out of the buffer and since the whole message is already in the buffer willcall message directly instead of calling into asio to read more from the buffer +If the problem still persists you should try and make the test case maven runnable so it easier to work with +I added a dbzos option to the trunk and branch have beem updated with new dbzos option + +Looks like one of us cleaned it up then I vaguely recall doing it +I dont think that such link useful for vpe now +I still need to do javadocs. +Thanks for the editing. +No I dont want copyright on those +Im still not comfortable with thisIf our goal is to throw out the maximum possible amount of obsolete data we should perform getPosition across the boardBut if our goal is to be minimally impactful with compaction then we shouldnt do it at all and rely instead on the timestamp check +Thanks for the reactivity Christian! that you can work around this issue in by setting the SQLMODE parameter to IGNORESPACE +Oh +Thanks Im marking this issue as resolved +Taking this for now iocanel please let me know if you want to do any work on this +setupdbdbserversrccomcloudupgradedaoHave anyone tested it +See if same tests fail +The above do not fix the problem +This is not good at all modifiable classes are jgdjwqgdkjq +The doc issue is +What does adding comma separated fields buy us since there is only one field status we are using +Getting them to pass on Mac OS X is not a requirement +At thatpoint it isnt about binary files because youd want it to work for textfiles also +JPOX metadata parsing does not support validation using XML Schema only via DTD +jaikiran pai Yup already tried it and I put it down as the Workaround when I filed the report +Martin looking into this I noticed that there are two extra functions that are widely used by projection classes and that would throw and exception when the coords are out of range +Hi Jun Ive attached a new patch +I have reduced the s sleep to +John can you double check my changes please +Actually there is one test query having a result clause another having a result class and this test query has both a result clause and a result class +This is fine thanksIll check it in with the appropriate explanation +Beware AFAIRC Felix doesnt implement the full core stack +Fixed internally and will be committed in branch +Closing. +Gabrielle raised the issue that tri sounds too much like it means three +But Ill most likely have missed some thingsI believe this can contribute to some new dynamic on this project +Added a new device object with some properties that can be used to conditionally render markup in a jsp. +But otherwise the changes look cool and have been applied in r +In fact I know I didnt do it right bc I accidentally reviewed my own thing but still +Im the author and certainly willing to donate itThe component only uses the cassandra thrift API and does not run in process with cassandra +In addition to the changes Kelvin has already committed for this issue the attached patch will complete the changes needed to rectify this issue +Fixed in rFrom now on transient fields will be serialized if the class has readObject method +lets get this part in first +Allen apologies for missing the point at first +Agreed with Deepak I didnt go through the code till but it looks more correct if only rejected item status should update +Unless I am mistaken the test condition is correct +EugenyI can not reproduce the error with the attached test case and the latest branch code +I got a couple diffs when running query tests +I just committed this. +From there materialization will work as expected +Carlo shares your concerns lets bring this to a forum and all discuss +trunk closing stale resolved issues +the files being deleted will be moved to trash if possible +This is a bug where you want to fix it or not silently removing the port is abug +Okay so lets do this +At time of use the logging system can be not yet initialized +See stale resolved issues +yep both reflect and kml methods of receive a but in those two reflector calls the bbox wont be available +Are you sure you dont have an old xmlschema on the classpath +It passes for me in ant with ant Dtestcase test or with java classpath libbuildtestclasses straight from the as you can see in my command I have in my and includes the runner which youre using +Could you provide test case of your application for us so that we trace down the issueThanks for your cooperation +Improve usability of Console for +Updated with an unrelated config change in the last again now that fix now works for root for just committed this +eine Frage drngt sich mir abr natrlich auf dein Patch kann sich im Moment ja nur auf Entity test cases beziehen da ja natrlich noch nicht committed ist +Unknown versions have a hash +I see the light now! The patch wouldnt apply to so I made changes manuallyCommitted +Im going to and redorebase this patch given when has evolved from +Looks Good To Me +please verify when the changes appear on websiteas for the outdented text that is not fixable i agree we should enclose it in containers for proper rendering +Please verify the fix in the latest +Youll be a much happier camper if you make sure that you do not share managed Entities across threads +Agreed +the text extraction of the pdf reference before loading seconds processing seconds after loading seconds processing secondsThanks for the contribution +Ive been launching this through but in this particular case I dont think that it matters because its the part thats failing +done +Reopening as the list of open ports in the EAP Security Guide is incomplete +Excellent suggestion! Working on and documented +Hadoop depends on a number of utilities so for now using Cygwin is a must and since Cygwin comes with shell I see little benefit in rewriting the scripts. +completed need to test in QA before +thanks selckin +Attached a patch fixxing the issue with setting of queue attributes on queue creation with web management console and restCould you please review and commit the patch attached to the JIRA +msec create MB file +This is not going to be inGili if youre still interested in these can I get you to update the patches with your changes +Thanks +I want to commit the patch if there is no objection +checked in the fix for into the branch +That fix was to use InError instead of the Stream versions since that gives us direct access to the Java streams without calling back into an IO object in the same runtime +Great to hear Yonik sorry I couldnt have given you a working example of the issue but glad you were able to duplicate it +Hi Dasarath aiyaI think it would be better if you can review apply this patch since you are the expert of KandulaThanks same patch as before but against ASF SVN Revision It now also includes a How to run the Holiday Sample section in the WSBA GuideIn short build the site using maven site at the project level +This was indirectly fixed by another change that queues a repaint callback rather than repainting immediately. +AddpanelMenu class was changed to for panelBar should be updated +Assigning to Jeanne to fix by to Trinidads trunk andwill need to backport to and checked into +The file installed by that rpm contains old JVMOPTS definitions +src only contains the parts of Spring that are compiled with target whereas tigersrc gets compiled with target +Were you able to determine if the VM is starting? The error saying that itwont start doesnt tell you of the VM is actually started just that itcannot be logged into via ssh +client h localhost u tesb p tesbIn my case connection was successful but I got that warning message in log fileThe same result is without p option but entering password on prompt + +The setting was loaded by hibernate but javassist was still used +Ben +Will commit it today +adds a property which defaults to k so after processing every k records mapagg will check the reduction rate to see if it should be disabled +Half line fix +Woops patch was against yesterdays common tree didnt apply +hereby grants the right to freely use the information supplied in this file in the creation of products supporting the Unicode StandardApache code doesnt have any such field of use restriction if you will indulge my tying this to field of useMy question is whether a restriction like this is enough to disqualify the license from being in category A +What happensI dont think any of these scenarios make sense +Thanks in advance for trying it out Ben +bug on +Looks good! +In the code your are looking for is actually implemented +Resolved +I have a suggestion to handle this +You can only test it by putting on s +Fabric uses its own jaas realm and default userpass is adminadmin +We can may be warn the user but continue to run +Thanks +Reviewed and approved for +Itd be nice if Solr had a for that +This is not possible currently as uses the Restlet APIWhat you refer to is part of Restlet API which has not been released yet +Then we will be able to reset timer on +I just committed this +Issue is fixed +We can go ahead and commit these html files and open a separate issue for revision Opened for the other issue of local and error messages. +Give me hours to deploy this fix to my production environment and get back to you via this issue +Patch applied thanks Noel. +only have a junit library in dependency tree of my projectLooks like you use really old version of Sonar +sandboxuimafit module names still use the mixed case. +This was caused by a couple bugs in the plugin +This is going to have similar problems where the limits will conflict with type layouts +patch applied +I committed this. +Can these be reworked such that isnt being replaced by DRLVMs kernel JAR? This causes duplicate versions of to be in the classpath since theres one in the concurrent module +Docs on how that works +Committed in branch x in rev + +Closing old resolved issues +I attach a minimal test case that i think is what you meant +Committed on New Revision Thanks Saminda its great that we have this improvement on directly importing the plugin project in to eclipse workbench for developers for directly to startup developing with out manual classpath and other setting altering +And yeah encrypted data should compress very very badly by design +Thanks for bringing this up Adriano +Setting to master as a confirmed issue +At a minimum there should be a way of preventing this from happening since they are not being explicitly referenced in the file +Ill be honest and say that I didnt review the RPM work hardly at all because I assumed from the title that it would not affect tarballs or the dev environment +No test case and no news for a long time dismissed as can not reproduce. + +Matches of property names do not respect the namespace URI The class extends from and adds the path as a fixed value the path should rather be retrieved from a hierarchy manager +Closing issues +Command line testing for runtimeinfo sysinfo etc can be added to this file when convertedI have a question about portability +Czech localization files had the same content repeated twice in them which is fixed by this patch. +Attachment has been added with description Testcase xml file no sensible checks has been added with description The area tree generated from the item I think I know whats happening +patch applied. +The following entry needs to be added to the svnignore property for srcjavalib +Good luck with this +bq +mrepositorynetsfkosmosfskfsUserstucu +shouldnt this patch go to as it contains the documentation? +Therefore could you please review the patchThankful +New integration test which sends a with attachments to an Echo service and receives back the response and validates itWith this patch the SAAJ implementation for AXIS is complete except that SAAJ TCK compliance has to be achieved +When using JavaMR with the sharelib can be a approach as a place of truth for the users can directly pick and put into their libs +Thanks Alex! +All looks good +And this bug is happening for all renderers +Ive added one but it is not reproducing the problem because clickLink is not delivering my page parameters +standardSQL can not support the writting in historicalbecause there is a big difference in some operators +Hi JaehongPlease see which Stack is just committing + +eg two instances of the server might be started by mistake and both would be getting data from the network +Its not needed really + +This patch should solve this issueThe timeout is given to the prepared statementHowever it still does not behave as expected it seems that the statement does not throw a on execution even when violating the +The requested site is either unavailable or cannot be found +KI is the only option I do not think this will ever get fixed in AS line the user needs to grab the required patch from the Remoting and apply to their installation to avoid the issue. +Commits are applied in M branchFor displaying the queue depth in bytesKB a new JIRA is raise +Bear in mind that you cannot copy what Sun has done but you can phrase in your own words what each of the various elements areIm moving this to wishlist +Its not used by solr but its on by default for luceneYepbq +The application deploys fine in Tomcat so things are looking good for Eclipse WTP +It is a minor issue +Thanks Keith for the explanations +It would be better to assign multiple jiras to the individual components who have this problem +However in my case it doesnt make much sense to assign a version number or tag to a property named +In case anyones curious is the minimum version we can use because s build depends on the skipexisting attribute of the get task which was added in Ant +thanks to an actual classworlds version this is working on is now replaced by a custom construction +This makes it very important for anything in a webapp that registers itself with an object in the system classloader to unregister when the webapp is unloaded +With vnodes it is all the more important because now the bootstrap is dependent on many nodes and the chances of failure are much higher +Kind of hard to tell with being so new and all +hasnt been released yet +Proposed addition to Category X sectionThere may be situations in which works under these licenses may be distributed in an Apache project +Thanks for taking a look Alejandro +LGTM +I dont think the minimum allocation logic is enough to justify a round trip to the RM +For HDFS I am creating a separate jira ticktet +By using false as the default value for this compatibility is maintained +Patch looks good except for the extra javac warnings +Please verify it works now +Ill have a look at this again sure I can reproduce it either but it does not matter really since the new logging proposal is going forward +Its kind of an battle to keep up with the changes from the developers +Since we no longer use Jackrabbit this issue is no longer applicable +Thanks Amita +I believe this is caused by this change in Groovy +Yes but the problem is an architectural one we would have to push the base class for stylesheet nodes down into the subsystem and that would be ugly +By not touching IO at allI am able to get same improvement changing the trunk when comparing to nightly builds asI reported before on branchThis is good because we dont need to change file formats at all +Hi FlavioI have attached the patch that also includes tests +There are instruction in that issue to try a SNAPSHOT version of AMQ that has the fix +This should now be ok +This should be implemented to be backwards compatible with standard Java API but also allow use with constructs +Add an option q to the count command that prints the quota of a directory +The IT corresponding to this issue now passes on Windows so Im assuming this is a problem on OSX right now. +This isnt quite how Ant and Gant work +I fixed was especially bad on windows as the bootstrap there failed cant we set this thing up so that we can mark bugs fixed +Hum according to you already choose solution so why choosing solution for exoregistry only? Cant you apply the same logic on this node at ECMS or PLF level +In real cluster deployment it will have conf package for each role +Does that make things clear +Im currently looking into the code but so far the only thing that is jumping out to me is that there are two IO Streams opened on the resource but never closed +Should Property be Properties and be +then once the functionality is removed we can change the factory to fail on init if it sees the option is still set in the I will work on a patch later that logs a warning if you try to use a configuration option for anything but Unicode for these two analyzers +Sorting the list plus adding a second check for an exact match did the trick. +James could you share more information how you built your corpus? Im thinking I should give this a shot using Wikipedia content +See +for the patch +Not sure if this is subsequent edits though +Pushing to support for exception handling as outlined in the previous comment +Most of them should be fixed now but I just checked and I think I missed a couple in the first round +The assumption is that this only happens at the tail of the log and is rare +Sorry for any confusion +I dont have any good ideas for a fix at this point +looks like its still looking for features at the old group IDkarafroot featuresaddurl mvnxml that change but currently blocked on +Attaching which removes the flunking of the server boot when authentication is not enabled +Thats my understanding at least +Indeed it is a refresh issue +This is the quickstart to reproduce the bug +Why is the manifest corrupted? It seems ok to me +Wasnt sure whether it got committed +as it turns out we do want this now as expected +Thanks for good suggestion +This is working for me thanks +Postponed +Merged and thanks to everyone when and how I can download the fix +I just committed this +removed shade pluginIm not sure this is best practice to build multi modules but OK to me +Handle xc ID xcd Process ID xcxfc libapraprosthreadcurrentxcxbf LIBEAYERRremovestatexxaf verifierxfxca ntdllxxc ntdllxdxef kernelxfxaff verifierxfxcf msvcrendthreadexxxacc verifierxxef kernelxHandle xf ID xbdc Process ID xcxfc libapraprosthreadcurrentxcxbf LIBEAYERRremovestatexxaf verifierxfxca ntdllxxc ntdllxdxef kernelxfxaff verifierxfxcf msvcrendthreadexxxacc verifierxxef kernel! version for I am using the attached application to verify the issue +Our environment is pretty good for causing IPC issues occasionally +I therefore unified code first +Oops I apologize +I will commit this today unless I hear any objections +manually change the import using +Also verified that the same fix is present in +At the moment does not support LRU +Same thing for me with from Sren for tip it works better without the new feature. +Looks OK to me +You might want to report this to HPIf you can give me an exact preprocessor constant and value Ill put an if +Thanks for the example really helped! +This ticket may end up encompassing all of and more +so it will be done just once and can be repeated till it will be successfulWDYT +Ok what i will do is create a new project and send it to you so you can verify it +There were problems with the creds in the jobConf +Thanks Niall +Attached the patches with all properties impelemented and fully tested Build Tests other words please go ahead and upload these patches to the Spring JIRA. +If its there I think M uses it +I propose this for and +updating patch with protocol changes with pbimpl changes +Thus this improvement needs documentation changes which will be covered under the sub taskI am glad if someone can review the patch and commit if no special commentsThanks +Under debugger OPT works OK too +No spaces in my repo path homesalvador +Cool! Please update assignees for issues you will be working on +I am about to upload another patch which if I did everything correctly last night will only have some updates to javadocsI got those failures on Windows with +ZIP with a sample project +I asked if that was possible in the IRC channel last week and the short answer was not really +ah right thanks +If the user sets a field to a value I wouldnt expect the runtime to then ignore the request despite the presence or absence of s +Keep in mind that the Struts tags with JSTL counterparts are not being actively enhanced + +Please set the Affects field +As such it does not seems to be a priority for the existing CXF community +This can be closed. +I would like to fix first and thenI think that will be solved by removing static references to the not found classes +Added group captuing with regular expressions +I am sorry that we havent been able to test the patch yet +I named the option +Patch looks fine to meBut should be this part of itself? Im fine with it anyways +The one thing which I and Owen felt should be left as is is to have the same file for writing the commands for launching tasks and debug scripts +Added fix for the problem found in the file. +Thanks James +Closing this issue as its not really part of the security framework at all instead bordering more on the problem domain and support services that help write code +Everything is working fine for me tooReally nice work JacopoThanks for your precious timeRegardsAshish Vijaywargiya +It is an eclipse project +Detects invalid root directory paths +This still seems to be a problem with my account +I committed the patch to the trunk as revision Should we merge this fix back to? If so is there anything tricky in the merge +I recommend to add this code into s action +The second user kept waiting till first user finishedIf you are using then I believe Sybase has the final say on acquiring to hear that has fix for +still need to do confluence +I think something changed between the version of ESB you have and what was released +When I saidIt is failing a test that is completely unrelated to this patch but it doesnt fail any unit testPlease readTRUNK is failing a test that is completely unrelated to this patch but MY PATCH doesnt fail any unit test ON MY LOCAL EXECUTIONS +The problem is due to Esper sharing views between statements for efficient processing and the decision to share the view is made incorrectly in this service pack jar to be posted shortly +It is relying on is which is currently always returning true even when broker and client are within the same VM +Verified. +This arguably became irrelevant through the introduction of the new test context framework in Spring +This is not a small change and may introduce bugs +This issue is an old feature request that was never implemented +will be unimaginably huge Im not sure its a good idea to load the Confiiguration if its not already loadedI think it would be better to hyperlink the module ID in the current display and introduce a new screen with more module details and move this kind of stuff to that screen + +Reformatted description +I will also do some small refactoring to make things easier to understand +Not caused by Firebug +impl and xjc should be under +Access to differential periods for example last version +it is not necessary to have such a feature as it is already available via order in lets continue the discussion in the related thread in forum. +Closing all resolved tickets from or older +Im not sure what is forThank you very muchNaoki +Ok my database was just initialized +Consequently this bug affects +I like the way the command documentation looks in the link above Karaf has a lot commands now so thorough documentation on their usage would be very beneficial +Hello TimThe fix works fine +Linking to upstream Cordova issue +Ok yeah thats the issue +Will do it in a different patch +Bulk close for since this issue has already been recorded in the CHANGES for and you should not attempt to and under the label of this issue open a new bug and link them so once the new fix is applied it can be coherently recorded and there wont be any confusion about when the various pieces of the problem were resolved +fixed in trunk and fyi when i simply deleted the file there was an ugly not found error on startup as if there is a hard coded dependency to load it think there may be the need for some additional code change to the war for this +If protocol buffer has some bug we can postpone the upgrade of protocol buffer +attach the patch +This patch is applied under dir stestsampleTest +Committed revision in trunkCommitted revision in nutchgora branch + +Raw output from the compiler +Seems like this is a Forrest errorbehaviour as I checked the site sources and the mail addresses are well written +keepAlive +I have not seen much value out of those tests +Almost all of these calls could be replaced by calls to Configuration object +How many partitions are there in the topic? If there is only partition all IO threads will need to synchronize on the same log during append +We just believe there should be better way to address a core of the issue +It compiles but quite a few tests fail +new patch with a very first and simple test is added +Attaching a third rev of +Im not sure when a localhost ping might fail +This is just a thought that why the envers cannot audit the data for many relationship without an intersection table? the hibernate is managed to persist the data to the underlying tables without a ternary table then why the envers is forcing for a intersection table +Yes I initially typed it in +Selecting a bunch of text and hitting I lets you get the test formatted so you can read it and work with it +Thanks +Added a new test which uses the binary token file +thank you edwardBTW should we need a test case for +Yeah really +Interesting it worths a look +Ive just fixed s singleton initialization to properly allow for circular reference detection instead of throwing an +Many thanks. +Heres the log file for that machine +In addition to the earlier described changes this version containsthe following new changesA javaclientorgapachederbyclientam For each prepared statement the field is is set to false to prevent committing the current user transaction when called in auto commit modeM javaclientorgapachederbyclientam Change the visibility of member is to package private +sorry I must have mislabeled the issue I only checked and trunk +Thanks a lot David +I found another similar problem and so have written a patch for which addresses my two issues +for the patch +is a bare minimum +The problem also occurs in the SNAPSHOT of Grails which Ive build yesterday +The issue is in SEQUENCER need to investigate why it drops the STOPFLUSH broadcast +GIF loading is now done through +For example a printer which wants to do pretty printing might want maintain the depth it is currently running at +When aligning both Gallio and on bit everything works fineThanks for your issue was not related to the plugin but rather on a Gallio version mismatch. +Where is your? is it in ANTHOMElib? or just un your project lib directoryI think groovys jar needs to be in the lib directory of your Ant installation +The hard coded DOCTYPE was necessary because of the screen widget boundary comments +Updated w null handling failing silent as it doesnt appear that anything canneeds to be done if a mapping is missing +Take out all the code and comment relate to DB JDBC driver +Hi thereI have created a new patch that will get rid of the additional Database Table and Column bean classesThe codegen now uses the empire and classes insteadThe patch contains the following changes modifications deletions package partly moved to additions Used to create a empty and to populate that Used by the velocity templates to get strings like accessor name mutator name table class name record class name +Committed revision to trunk +Ive added a delete call to the which should hopefully fix the constant redeployment of the xml file +User name ajooemail address +Fwiw pass +Thanks Lewis + +Thanks it works fine. +The reported blocks has reached the threshold of total blocks +Fixed in commit ade. +Im waiting for your show them like they are shown today these are the ones I want to have a dashed line These are like +Looks good +Local test passed +Ok ready for submission if it is acceptable +Alessio this is also the reason why you cant get the inner GWC to work the stack trace is the same +Although being a third party library i guess it forgos those requirements +the test incorrectly thought the last term would be an indexed termI also added some missing asserts so the next time this happens the exceptions wont look so scary + is coming from +Changgeng something went wrong during the release itself but we dont know what exactly + last comment +The second part is to optimize the web service calls so that they reduce the load on the system +Appoligize for cloning same issue. +There is no way to select a Build Environment to be used during the release. +If I remove just +Thanks for your collaboration KoenOk ill do it but first explain to me what do you mean by provide a patch? All i can do and what in fact im doing is to change the code from repository it dont seem to look like providing a patch? Piotr +Subsequent enhancements will be dealt in other tickets +Very soon I will check in a fix for TRIM and then I will check other functions implemented in that code +checked a first version connection wizard is missing the rest is to TRUNK +Thanks Ivan for pointing me to SPI interface +Addendum looks good to me +please apply the new NL patch +I created a pull request +Rebased and merged into the master branch. +Moved stack trace methods from sample jar helper into default helper and added extra methods for checking the name of the trigger method and its callers +But there are a plethora of old in bugs in the Livelink Authority UI that we are tripping over +I suppose sometimes those handy IDE fixorganize imports feature can get you into Juergen for all your help +Raising level since depends on a snapshot on that is not longer in maven repo +patch is rebased against trunk +The only thing I found was a format string bug +Puh really ugly + +Close Doc issues in release versions. +Using +Thanks Jason + test and passed +GytisIve updated the sanity check script +It is HTML compliant across all browsers +Summary +For to work correctly in karaf itests is required +The assumption was that the copy would always work to an empty page but for the same reason as described above the copy of the row to an empty pagemay not work if the new page is actually a reclaimed page which in the past has used up record ids +M released +Renamed to make it more clear +If there are other statements that the should supportcould you tell me what kind of non valid statement can come to the +Test suite got stuck in There was no other test failure thus far +Patch looks fine +Ill wait for its resolution before making further proposals +New patch attached unused imports are removed from +SUSE differences are preserved through the fileI have ran an RPM build and verified the init script has the proper CHKCONFIG settingsThis patch should be ready for commit +have added a test case to enshrine this situation decision. +No problems at all Andreas! I need to add some information about this to the User guide too +test problem still present in the codebase. +petasorts merge was huge +This was committed as a part of. +Im going to take a crack at this +Makes sense to me +Please and looks OK to me +Catching up on Closing my reported issues +Also Im fairly certain Callbacks will need parameters +Verified with JBDS its based on Eclipse platform build version +Let me whip up a patch for that +Added unit test for empty inputs +patch for Hadoop QA +I just tried it locally and it built see +Enclose the arg in quotesIve attached a test case showing examples of the above two methods +Ill create a JBM Jira for this issue and fix it re out of time +Merged into the master branch. +It is not unqualifiedEither way if xsany namespaceother gives an error how is xsany supposed to be used +Please attach sample XML XSLT and Java code if applicable that illustrates the problemThanks provide more information on this problem or let us know if we can close the bug +Am just in the process of setting them up on another box just to make sure +More PropertiesTagged PDFYesFast Web View No No SecurityIll try to remove the content without repairing the file +Maybe the old methods need to come back and be deprecated with replacements the new methods that are currently broken? Does that work +This made me think there was some redundancy there after all so I removed them made Abstract InOut interceptors ready to use in cases where a service object can be injected so think will cover a common WSRS case +bulk move of all unresolved issues from to +Actually I think strtok changes the original string so the above diff does not represent a solution +Attached patch from Paulo Poiati +Could you please update components and version fields for this jira. +Ive taken over the routes for activities in the provider +Added patch addressed Stacks review comments removed the move region change to slowDeterministic so the test isnt flaky will address it in different issues +is in the for thanks for pinging me on this Mark +The patch seems to prevent the predicate from being pushed if any of the under the query with the fetch clause is ordered +It just takes a while until the problem occurs +The comments regarding may indeed be an issue but I dont think it is the cause of the error +I just run another test tonight +Good catch +patch implementing all of the proposed steps for trunk +Verified in info about the Web XML Editor +Ill post a link when it is ready +Saxon jar was incorrect in test casemvn test in. +The broke HCM map will work as it did +Reopened issue due to problems experienced by some applications such as TracWell need to the code committed to make sure that behaves properly as a object +Patch attached +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian Minchau +Presumably it would still be on record +Ok In that case can you just go ahead and create the JIRA and Ill do the legwork subsequently to add the appropriate developers? Thanks +Breaks shouldnt be too hard to implement and were always happy if someone helps in the development +This is also documented in the user x kb about MB of data +Cool +Deleting build results needs to be a process so that Its possible to keep the data around for statistics while cleaning up the project view +As an aside this enhancement also helps standard Roo Shell users because errors from Felix now display in red and normal messages in green. +Thanks for the quick reply +This issue is partially done in r for the dependencies stored in MHOME probably add all maven artifacts in the root pom +I will go ahead and work on a new patch soon +Francesco Your podling data has been loaded into the main repository and is ready for use +I have to switch it to UDP for testing at home because my router doesnt support TCPShould totally not be in the patch +It may be added to the distro at a later date +PS This is causing a failing test in the Grails to Graeme and Peter a regression may have been introduced in annotation support between rev and rev seems that the files given are not enough to reproduce the problem +This is working nice on a FF on OSXAre you sure there are no other things in your page and you have a standard jspwiki deployedEventually could you show the rendered html of the tabbed section? the attached page is no tabbedSection but a tabbedAccordionAre you sure this is the right page? me in the meantime I had tested with the tabbed Accordion but this also showed the same problemHere the correct html page with tabbedSectionSorry +i just happened to take a look at that the locks some time agoand found myself struggling with the that turned out to be irrelevant for theproblem i was trying to solve +added in thanks! +While ALLOW FILTERING queries shouldnt be your apps I think there are plenty of creative use cases besides playing with data in cqlsh +create second table Kitten with column catname +Thanks Daryn for the backport. +It is not the best place to get help with this problem you are having +I thought we agreed that it is a regression in GFWicket cannot fix the web container it is running in neither the OS nor the hardware +That suggests that the problem is within how the adaptation to handle is set up +So if there are more than one namespace etc is created +What if we add a combobox with versions to group headersIm not sure if the GIF palette allows to do it +Done +Reviewed tested and committed Kens patch. +Looks good +tests that are not wanted or required made it into the applied patch +A simple proposal is to do what is done in +Sounds like its a random issue + Username djencks +We have used gant as a test case and it is currently compiling with all tests passing +It is a feature request to specify message converters on a basis +pull request sent +I think thats possibly the easiest for now +Then the reduce tasks will read the new output of the maptask but reduce tasks that have read the data from the failed tasktracker have read the old output +Should be closed since the fix is in release +My only fear is that we still might have missed some but in view of the complexity of this I cant imagine what is being overlooked +Hmmm +Thanks for looking it over Bikas +Have tried reproduced this issue with the release as well as the latest HEAD +That was my intension this issue can now be closed. +Im going to commit this +I need to be able to output subsets of an objects attributes they are used in soap messages +I think were good for now with this release +Closing +Latest patch does not have labeling of tests since I wanted to wait and see what dev thinks about the latest testing stuff +partitionwisefileformat was for and others passed in local +Does it affect jvm I have to use to run sonar analysis with maven +why you use under windows? Do you try to use cygwin +Is it an implementation of a standards based SQL or JDBC feature that can only be implemented in the server and delivered in +Assuming there are n stages the parameters r can be used to override number of reducers at that stage +This is just the source patch +Also submitted project proposal for this +Generated from a checkout of the tag. +Still the same issue even if I build the plugin myself from trunk +The queries and methods have been restored to the code base because ESB still uses a number of them +That file works perfectly fine in Notepad and Vi +you dont need to setup a dummy proxy only guess I have is the netbeans plugin isnt released +Please consider implementing it in or something +Committed to +Mark I am running with snapshots from Oct +I know Rich F +Please post a patch here when you feel good about it +This is a second supported format in Dozer since versionI can update documentation page as well if you could give me access to the Wiki +Hey JoeThanks for the work around seems to work well for me also +I agree +Will renew on shortly but looks like findbugs reports need some work locally first doing that +I can probably look at that tomorrow or Thursdaypete +Replace jboss dependencies with individual spec dependencies +The decorator removes the header if it uncompresses things will go back to verify the impact on +Minor change to Dans patch +Strong can comment on whether ER is an appropriate Fix In release +Now i can say i am very happyThank you guysOne more thing +Thanks for the review Avery! Committed. +This kind of issue is typically caused by an outdated version of WSDLJ in the runtime +This may help you diagnose +Yes the document is by Adobe Distiller. +When I add the url http to the urls file and start nutch fetcher says Unable to resolve host +problem staement is correct but in the rule definition i have used a bound variable for the list which could have cuased some confusion I am very sorry about thatIn THEN block i am able to see other attributes of t but not the collections +You say it happened with your update to Spring right? Were there any other changesupdates that might have affected this? Can you give me some way to reproduce it locally +Are you sure hibernate is going to like it +DavidUpdated the release notes +All fixed thanks Bruce! +For eg +If there once was its now gone +Could you anyway please review if the patch contains all required changesBefore committing there is one more thing to look at as the regression test suite shows more failures then usual +With this file the bug can be reproducedThe correct output should be Integer thanks +through the web UI +These files bring the admin scripting fully current +The pull request was already merged with master branch therefore I am resolving this issue. +New were introduced to in r but their implementation only exist for native X +During regression there was discussion on the mailing list for this issue and it was changed back +This way on startup we will be sure that the data in the last valid index entry is not corrupted and we can use it to quickly locate the recovery offset in the log file +That context should hold enough information to allow the Dir impl to make decisions like buffer sizes and avoiding buffer cache etc +None of these changes have helped this +As far as AS is concerned AOP should never be taking over logging it only happens in certain error situations +This patch applies ONLY to +Fixed null pointer bug that sometimes occurred when navigating to msgDepth graph thanks to Sergey Zhemzhitsky for that! Fixed null pointer bug in Authenticator which caused a fairly unfriendly failure if a username was supplied that wasnt in the thanks to Bruno Matos for that! Default to anonymous rather than guest if no username is supplied in Connection URL this explicitly sets saslmechsANONYMOUS in the Java for this case +I think leaving the workaround disabled for the tests is the right thing this way it fails on a broken system which will help users identify that they need to enable the workaroundAlso the description seems appropriate to me this isnt about POSIX compliance its just a bug in a library function that will probably be fixed +I think in my last patch I missed updating the class from the core package +Patch v reverts the change from major version could introduce more bugs have you tested a few commands? Have you tried creating the same table twiceI removed patch vI agree with reverting +This seems to be fixed now with the latest nightly builds of +new testing not planning on backporting this. +Attachment has been added with description A minimal file to reproduce the is fixed in rev +Patch looks clean to me + +Try the latest snapshot + +The source jsp causing the generated file by the bind tag is not supposed to be nested within another bind tag thats where the variable conflicts occur +and will be kept as separate properties since they can refer to different locations in a single build +Moving issue to. +As far as I can tell the processor is behaving as required by the XSLT Recommendation +Some of the Antlr parser objects can consume a substantial amount of heap +How could I have assumed that there is something Spring does not have a solution forI will have a look at if it matches my requirements and reportJrg +committed the fix +Ill add my review notes shortlyIm going to suggest a with Symantec for early December +I believe Seleniums wait is used to handle this situation +Unassign issues assigned to Nate +Logs are already long gone + +It is also worth mentioning that this same issue while not a functional problem could have been causing inaccurate backtraces for lots of people using +In the I create a new Hashtable and add the entries each time circumventing the issue +It does not require the manual jconsole verification step +The associated thanks. +Attached an updated version of the JCR API jarIll update the build to use this jar in a moment +The current jetty profile only works for +Previously there were links both to other topics in the manual and to subsections of this topic which seemed confusing so I got rid of all of them +Even if we can invoke this operation as a batch it will still be slow since the zookeeper readswrites will happen serially + +Hi Jean Do you have any plan about this improvement? One bug of geronimo has a very close relation with this one we need to fix it in time appreciate if u can give me an accurate target date or your current progress +Diff for to allow per servlet config file diffthanksdims +That can happen Closing due to an invalid configuration +And secondly there is already code in there that should only fetch the columns from the object of schema with the same name as the current user in which version of are you using? or trunk? If youre on you will be glad to hear that this problem is almost certainly fixed in trunk +Please check Dworkspacebinding +So the validation status isnt being resetIs this an issue with the new code that was added or is it an issue with the way createForm is used and implements the equality between the fields +The Stop is not cleaning up resources +As soon as Stefan gets our new version which incorporates our patches we will have a Stefan please flip it back for us when we have a tag for that +The change has been made to both and closed as resolved and released + am still reluctant to commit this patch +Not sure if we could expand this jira to handle all of these cases +I will look into that in another issue +Then the designcontract of the method is wrong because unescaping and then escaping should produce the origin string +So the same fixes should apply there as well +We need to Implement to hold the list of sessions and then use it to resolve +Christianpostgresql also returns always independant whether thedatabase stores the data in or LATINSo I changed it accordingly +Will have to revisit this. +Link to in revision +to trunk +The mapper is changed to send each of its outputs to a different reducer depending upon the number deployed in the iteration +In the API there is no concept of a working directory +A few commentsThe indentation in is a little funkyIf were going to put a finally block in s run method we should probably at least log whatever gets thrown I hate silent catchesIf you have concerns about the code quality in the test I think youd better just go ahead fix it +I suspect solving this gets us closer to solving the other issue +How about we rename the core one to +This patch has a few minor changes to the proposal in the jira description +JBM paging is extremely fast compared to since it doesnt use a JBM has producer flow control so +We wait before returning +FWIW restoring the deleted call to bufferedSend corrected the symptoms I had +I think so and Thomas +I guess I goofed on picking the name for maven artifactWoodie Im looking into this today +Sorry about that my bad +Thank to Weld integration as there are no standard hooks in Servlet to notify of when the login page is rendered +bq +Committed to trunk and +Patch applied in trunk and issue does not need to be in the release notes +It is possible that there can be a race between an App when submitted and killed such that you can get a kill when its just newAlso fixed the test failure introduced in +Investigating +Feel free to switch reschedule if you feel appropriate +Committed revision thanks Javier +In addition a new configuration option should be introduced so this can be set via the config file rather than a D vote for which defaults to true +The PDF and HTML Books are covered then because the license text appears as a header or footer appearing in the visible documentation +This issue was solved by patch for +We now support describe functionname and describe extendedIn udfdivider you need to add that and add some annotation in the implementation class to show documentation of the +dont expect to look at this until post at earliest +I have seen particular behaviors of concern +blew the hex subtraction conversion to decimal +I should be in r +I have a copy of what I believe is the baseline from the maven site and still see issues with this when obtaining online WSDL from my application +Stu SylvanLet me try to fix this by using to serialize directly to disk and creating new suggested by Sylvain above +to support the old version of CXF +Fixed in branch and trunk. +I change to the version +I checked the commit in case you verify issues +Still need a patch for trunk. +Is there a set of tests? Happy to pick up any formal testing required and work through getting this patch complying with any additional anything else is needed please let me know and Ill get on to up the good workTake care m superseeding this with a new jIRA which will have a diff for to support Ingres properly rather than just this bit. +This change should resolve my concerns about the running container leaks but there are some points that need to be addressed with respect to logging in cleanup If an application with many containers simply unregisters and exits expecting the RM to clean up the mess or the application simply crashes were going to log a lot of messages for all those containers +Thanks Raghu! +Yeah it has been solved I am going to close it +Thanks for looking at this Bryan +SDB cleaned of RDB +Verified at +is implemented based on for stuff +Agree there are no known reasons for aggregation as of now. +This should allow things to be less problematic when a JVM adds a language and it seems a lot less likely that the JVM would remove a languageFixed in r. +If youre sure it works go ahead and commit it +That will for sure shutdown the cluster +This got to Stefano fixing +Working patch that includes a unit test for debugging this issue +what does that mean +Committed to trunk and +the patch to address the implementation onI will add versions of the tests some time in the next week +What address should we use infrastructure root or something else? Can it be a problem if the address is a mailing list +Any chance the bugs could be consolidated? +Closing this issue +Here is the correct list +On Jimmys comment but otherwise LGTM +BJI just realize that this is open for and fixed only trunk could you check please +This might be related to but it has been raised for module +Just suggesting the appropriate compiler directives for stricter static messages would save a lot of grief +Removes the checksum validation from validates this patch +Option however is still a valid solution to this problem +Such just resolve or push out or reassign this one in tandem w +There should be as well a Plugin Dependencies and a Maven Dependencies +I have updated the manual and feature archive to include the latest this jira as can create new jira issues for possible bugs and new features related to this integration +I will look at that later this week +cool can the name of the audition table be customized +This should be deployed directly to apache repo how can it be built +As source code of Camel which deal with message partis is just like as Camel so this test will also failed on camel. +Im not sure which feature you are talking ofThe patch provided by me alters the Karaf features XML no use of the features XML here +Hi Heath Ive tested the patch update on my local build and Im quiet happy to include this as is +usageevent ADD COLUMN virtualsize bigint unsignedALTER TABLE cloudusage +Can you add the rendered html code please +Of course DRLV porting has to be done more carefully and correctly +Can you possibly patch as well and commit both? Good catch +Suggested patch that makes localhost and DFS the default +I believe bruno and I felt i was better to separate out the pointcut objectname filtering and support only objectname filtering +I would recommend that you vote on this issue and try to get others to vote on it +The new readerwriter implementations introduced by this fix required the jar to be added to +Any other better solution +The common thing of these issues is we want to make sure that masterslave are always in sync regardless of which one starts firstPlease take a look at the patch I provide +Thanks a lot for this enhancement +Closing this record since is now released. +Long running servers or large XML data will cause +Latest patch unit tests +Its package private +New patch to fix unit tests. +Attaching a patch that speedups the removing and also fixes the issue +but Ill live with asm +This works for me now +Yep the script will stay the same to affect only started this be resolved again now or is more work being done +And so incremental builds will work appropriately +Attaching patchVretry only once if as Stacks it be a flag that gets set when we are retrying because of a? The flag called retry or some such then stays set while we retry +traced down the issueIt was another entity that I had created as a view of Book +invoke get a certain java Object as response +thanks bens updates to forrest docs +just did not show up in the diff +Currently this property works by starting all apps as not runnable and then marking some as runnable in the update thread +Please use this patch for heap size extending from mx to mx +Should this be doneWho decides this change in functionalityIf its determined that this be done should patch also include change to docs +The user will have to set the snippets count but this can be documented +Fixed under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This information is still missing in distributionresconf it would be great if you could update that too. +Hi folksDoes anyone have an opinion on this issue? Id really like to be able to do this as it will allow me to parameterize my views before they are shown rather than having to jump through all sorts of hoops to get access to my pagesviewsforms after they have been created and shown +Nothing that I can do here +Does the same issue exist in trunk +How can I get this fix if I want it now? When would the next release be? I mean the one having this fix +Ted can you generate a patch? If not I will generate one as I definitely one this going into +Im going to optimistically mark this as resolved in +Hopefully this can be used as part of the in the future +The latest patch had a conflict with which handled the corresponding issue on the client +needs to be ported +Unfortunately what I wrote regarding the fact that in the Catalog the shapefile is added without capital letters involve that if you close uDig and restart the shapefile is not found +Im not trying to detract from what youre saying though it would be nice to have something to help out with making sure there is enough space for +samplesquickstartshelloworldhibernateactionlibext + adds beeline to in +Now the test are passing on my machine check the new patch and let me know if you do still have some problemsThanks patch that should also take care of the cascade deletes +sorry your description of the issue is not clear for me +We believe this support is huge for mahout users +Possibly in a couple weeks however +Was that comment meant for some other module maybe? +It is not immediately clear to me that this report is a duplicate +Fix applied +Are there any plans to use the V version? I am trying to pass a regular expression such as +to the PM patch +This will be handled inside that issue +I want to commit this +I added tests to to prove works as expected too +Attach initial patch still missing factory methods to create this stuff directly but at least now setter methods are available +So still use as provider that corrupt the test +what is that +It replaces the previous one +Otherwise please provide details of how to reproduce the issue +probably yes but I cant really hardcode that lib directory in the sar module since anyone can change that directory and its not part of spec as lib is for the war +Not reproduced now on issues to Nick and had to set resolution of all of them to done +Included with it are unit tests to validate the results +The Android parser should be calling the platform checkreqs script. +I can confirm this problem +I added an implementation for an axial shading to be used in combination with the sh operator in revision +m configured with your credentials +bq +Rebased the patch to the latest version +In the past spilling was passive it was done when the JVM informed us that memory was getting low +All unit tests pass including and wps +I think that support for a plugin for this should land in the trend for less buttons on Android it makes sense for this to be handled for a fix +There is a tar parser included in Apache Ant +First time Ive ever used bugzilla +Patch Available to get Hive QA run +Verified on +reopen need investigate more detail +they only contain sites +James without looking into the history of the files what were the changes for this one? Something that could be quickly? +But if you set the session timeout higher the garbage collection does not work number of instances in the JVM is increasing +Please look in history for what the actual resolution was. +Resolving as invalid +Hi AdamI got no feedback on this one so I set it to resolved +With this patch I managed to configure connecting to NN and BN +Youre free to close the issue since I have no right to do that +Thanks. +Since Ruchith mentioned that this is working am closing this +Please verify +Changing issue type to bugIf you dont agree feel free to change it back Mamta +Of course to be able to aggregate coverage information at any level all uncoveredlines uncoveredconditions linestocover and conditionstocover metric must be fed +No longer all ResolvedFixed issues lacking a Fixed Version +Let me know what you prefer +Theyd get a warning for every single request which is certainly not feasible in a production a warning about not setting allowed fields per request is not something I considered and is totally unfeasible +Completed At revision Completed At revision +On Oracle for example if you want the optimizer to select from a partition that will only be optimized as such if the partition value is a literal +does seem to me to behave differently to cases like Zip which have multiple and it is different to cases like Maybe we should check for an on the context use that if given and if not cache a default one on the parser and supply that through to the mail content handler +This should do +Attached a patch file that can be applied to the source to build a new JAR file +Removes the class +Is it possible that you are unintentionally mixing in the same cluster some nodes that are INVALIDATIONSYNC and DISTSYNC +Also need to make sure column names that are TEIID keywords are enclosed in double quotes +went smoothly. +The input split size is configurable through some parameters +The question was not about being backported or not as it is clearly stated in your previous comment +However I get tracebacks when ever I start tomcat saying that my has no mention of deleteQuery +I am not disagreeing with Oleg +Issue resolved by applying the patch +Merged to in revision +It may just be me but I often refer to this source code to figure out what locks are compatible with what +right fixing patch makes attempt ignores attemptsaved at FinalSaving also +FolksI evaluated various design options and decided against developing a abstraction layer for session filters in favor of implementing a simpler SSL session decorator +Elliott You are a good man +The code is now fixed in rev +Im struggling to extract a better testcase Ill update once Ive got something +I should have the appropriate rights for this +If there are any other changes coming we should try to batch these together at the same time to avoid having to update clients too often +CSV support Raw Solr Command supportcommitWithin overwrite supportBoost for JSON support +the new patch looks good to me +Can be considered as fixed since both axis and synapse works with the existing behaviour of the serverSide attribute. +removed +on leaving the in the installer +Any chance that after a year and a half this bug will get a second look? +When I refresh a second time it worksSorry for this confusing description but that is how it is +I have included the bug number to the patch name +Glad this could help resolve the Eclipse updates issueI can now run JNLP on at least JNLP workload using Harmony and the no security versionof with slight tweak to get to work with Windows XP +While it is useful we need to study the environment in which the code is run and a it is good to have better understanding of a system to make relevant changes +Bruce as I am observing the same symptoms and I am working on mySQL as wellcan you please confirm whether you are getting the same errotrsfailures +Ill also add a test case to the coreWeb app +I have a better fix now +Strange to see those files as I dont use Eclipse but dont appear in my project and target dir was empty + +Please note I moved your entry in contributors one line up above the find all link. +Verified in CP My understanding is that it was fixed in this release +Thanks Ning +Good +Agreed. +I did not mean the content of the and classThe and should be in the same folder +NiallProbably its better to put this discussion in the dev list the problem is wider than I thought +Updates models for operation +Neha fixed what you suggestedJun +If we remove the wrong input field this wouldnt be noticed at allIs it possible to use a custom validation to highlight the field on a wrong inputMany thanks course proper validation for the date format would be the way to go +Quite a bit of work happened to support chinese and japanes chars since assuming this has been fixed in the when I say assuming it has been fixed I really mean we have tests reading chinese chars out of a dbf files in our regular build +Thanks for patch Mikhail +However I would not mark the issue as resolved until I write a complete test to test it case added +Is this a duplicate of +Sorry for the late comment +Should be good to go and you have project admin access so you can in theory fix almost everything you need too +Also Chriss MDalgorithm uses a nice little he wroteStill my SHA impl is far from ready so lets see what ants got in it +defaulted it to you could file this as an issue at? +Thus preference would be to get patch to iOS Codova and Android Cordova error occurs when invalid credentials are supplied to a implementing basic authentication +I think I was looking at this the wrong way this patch looks rightI was able to use this code with s flat file deserializer based record reader and read and wrote thrift records just fine +for latest rev of part +Why +theres and local but no and no cdt so Im resolving this. +This test passes on both jrockit as well as Sun cruisecontrol I see that the test passes for the + +Thanks for checking Andrea I had sorted out what was meantI am going to fix this issue with a this week as part of getting him set up with a build environment etc +We just need to add more features +Setting for and but fully expecting it to not be totally fixed until +Jesse the release is ready nowdo you have a patch to include before the release +Roger that +Fixed by by + benchmark patch attached +bad for annotated node matching occurs even if the type does not matchthe type should be matched firstif match then go on with the this to avoid unneeded extraction of the +Or would you prefer to have core be pluggable in this respect +Bumping to M as this affects the SOA split not JBDS obsoleted by +submitting +This request seems to be duplicate of +Fixed on master in be also backported to branch +It seams we can cover this with one parser but I do not know yet if other problems exist +It is a custom translator but unfortunately we lost the log containing the query so we cant reproduce to get the plan +Will be fixed in Neoj itself so I close this issue +Merged into dev branch commit de + issue in favor of. +The only way I can think of doing this is +Again please refrain from solutions in the description +The modalWindow patch no longer attempts to modify the request target for iframe modals that are working with a lost have added a quickstart which demonstrates the problems addressed by this of errors fixed by this be fixed in Wicket wit jQuery? It would be a pleasure if we can mix page und panel modal windows. +But after talking with Adam about the compaction swap I realized that it is necessary for when someone is reading a view during that swap +Thanks for the comments and reviewing Konstantin +defer all issues to +Its very interesting work + is an error in my behalf +I didnt implement iterators +Here is an updated patch that only wraps the methods that throw a in a try +Thanks Ted +Why is this not being raised in the jBPM JIRA +updated the new unit test +The if? condition is currently there for a reason +I think its something thats very attractive about Trinidad and before using features is something that we consider to moving to Trinidad especially for +Ok go ahead and add your solution +At the time where I moved mysql to x you commented it positively +Thanks for the patch +Just to be clear do you mean to say that the authz files will now always contain a reference to each unix group and each committee groupThe unix group has to be present for TLP sites but the pmc group might not be needed if there is no SVN tree +Heres an updated patch +I think you could still write a unit test here just instantiate a lock it and then try to do the same again +See +Patch includes fix and test case +attaching is the one to archive also includes a testcase for +See for further details on this +Please confirm if it was +Leave caching of such queries to +Also mind if I assign this to myself +Got it +In my opinion clarify in the Javadoc the underlying implementation dependencies is the best solution now +I do see the concept of Required Interface but that says any unimplemented methods can throw +I suppose Darran will take care of the release once he is done +Im resolving the issue +The may be included in the WFK repo +Is this because the access to jobs is not synchronized? Wouldnt this result in losing a heartbeat +I did a small bugfix on the new support in my adapted file +I might in fact be able to manually do the remapping for the existing mirror +bq +Its just irritating for our customers +Closed following Denniss comment +I just committed this +Added +Fixed with r. +A slight enhancement to your suggestion might be to always skip the statement cache when optimizer tracing is on +vcproj files so it is a standalone patch that replaces the previous one +Patch attached +jar files too +this is indeed working now so closing the issue +We can control content but cannot get full control over ordering +Hope thats ok since we didnt create the website before +Looks like the fix for will be included in the version +Yongqiang is redoing that changes now +Clearly there is a mistake in that documentation because the Parent entity is displayed twice +When specifyVlantrue dont automatically send specifytrue +Thanks for working on this Binglin! Will check the patch soon +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsernodes +Thanks for reporting this Erich +On the bright side I can reproduce your original problemFollowing your steps with archetypecreate and specifying site and pir plugins in the created pom I get exactly the same stacktrace as you posted above +there is a new way to check if a bundle revision is resolved and we werent checking for null +Are you explicitly setting either of these environment variables? Or is the +reporting user said that their filesystem had been corrupted so closing this as invalid. +Or can I see a sample from somewhere? I dont know anything about +Thanks for the patch +Migration is finished dependency is removedNot resolving this issue yet while occasional problems are expected +Hi Sanchit As you can see there was a port in use error Could not listen on port on host Address already in use so the registry did not get started +But I sure dont meant to propose a complex workarounds quite the contrary and right now I think this is neither really complex nor a workaround as this addresses real problems +Changed Target Version to upon release of +apparently successful +Thanks for the trivial patch Ill see about building my local binaries with it + +The patch I posted is against I am working on a patch for trunk too +Committed revision . +Honestly I think that this shouldnt be high priority because Android apps arent supposed to have a menu button if theyre done right and changes to the menu button can break the Back Button +Fixed Steps to Reproduce to fit Description +Looks good +the actual path can be specified in +please + Following my last comment we should not consider this issue as bug but as an optional feature request +Manually tested! +How did you call in such a way that the implementation found in couldnt be located +Nit are we planning to fix get to not use produceRequest as part of this patch or is that a separate jira +This is an OK approach but much more labor intensive than a regression test +Ive committed this to thanks Anupam! +I handle corecompilerfusion issues +I think that we have to start out assuming that Derby implements the relevant rules in Section +Applied from in SVN revisions and to branch +Term info looks wrong due to wrong div box sizes +Pin +I tested with the XNIO jar that you posted and the problem appears to be resolved +Marking as resolved since the two were merged some time ago. +Thanks! promised to look into this but I havent had time to do it +StackWhat about we define the safemode for master like namenode where master is aborting and initializingIn the safemode we maybe could also refuse which simplify solution of +So this patch just broke a bunch of folks and should be reverted +I have investigated further and it seems that this is harder to reproduce than my description suggests +This persists when dialog is closed and opened again +The aim of is prevent the confliction of jBPM directly use as dependencies and it will conflict with from tomcat +Can you reduce the timeout for +PaulWhat is it exactly you are doing that triggers the exception? Can you put together a test case to reproduce the bugOleg + cheersstefan +Some of these issues should be already fixed but some not +Hi Suja can you please update the affects version field appropriately +The filter change did not seem to work +This would make it possible for the users to define this at a central place whether they want the label to reflect the enabled state or not +Schubert can you post numbers +Better do it right +That seems to do the trick actually +But is this script planned to be run often? Thought it would be just once every few months to check which accounts are unused +I did more experimentation in added a comment that shows the three files SCM status varies based on the scenario and that the jpg file is never checked out automatically +However the patched pig is super slow on clusterReason is Im using the Text directly as the replacement of +Now I see that you described problem which should be fixed +However I noticed you have added a new JAR dependency that contains an IBM JAR +This patch contains a Whirr properties file that will instantiate a small cluster for testing a shell script that will configure an EC machine with the appropriate packages checkout Pig apply a patch and run the tests a README on how to use the precedingWhirr sold for me +v +Test succeeded on my local machine +Thus no need to change anything on our side. +Rob you have an interesting point but requests are cancellable not forwardsIf you no longer want the request to be cancelled use a redirect per comment to start a new request +Can be verified after verifying +Manually tested! +Not sure how long that will take this is because of the Maven infrastructure +bulk defer of featurestaskswishes from to +Thought running through my head I hope I have a way to turn this off because it does more harm than good +untested patch +package Class patch file can you take a look at the patch and see if they work for you +Thats okay + +Please try to enlarge the modal window and check if the ? icon goes back to the same line +Does the current read aheadprefetch cache use a LRU rule? An optional LRU shared cache with a configurable hard limit would be optimal. +Its old and not +It may not +Rosolve as invalid +Have you tried with the release +The test succeeds +After upgrading I noticed an issue as detailed in +Hoss right Im saying i dont use that scriptI think we should still lay out the new instructions on the wiki for people who dont want scripts svn committing for them thats all +at least that is what it looks like and with you saying it suddenly starts working again its a tricky thing to hunt downLet us know if it happens again +Now Ive fixed can you retest this? It should be fixed +spent way too much of my time debugging that code to want to see it live +bq +Number of users will be a function of input size only +now the urnxxx decode fails as well +We cannot entirely switch to brat because all the training logic we have is build against the Corpus Server +For the last block if it is under construction obtains the block length from one of the datanode +I just dont know that well get it done in this release +If the fileExist option is set to Override then the file will be truncated and if set to append the file will remain unchangedI appreciate any feedback for my first patch +It doesnt appear to be adding new listeners that are persisting until the next poller is run +Added to release plan +Can you please add ASF license headers to source files when you submit? Are these your original work and submitted here under the ASF license +verified that it is ok now +Patches posted here are obsolete +I dont see any indication nor remember we ever had root of be root of tools +If we want to prevent the method call if no fields should be loaded there should be a I think no? +Im intentionally not adding a constructor to at this time +Since this really is changing behavior between trunk and x Ill put up a separate patch for x I hope todayOne thing this patch does is completely get rid of the fallback string in its made available for tests by being put in a test file +If anyone has better wording feel free to improve it +OkDo you have an issue id for the Enum thing? My opinion on that without knowing the details is that Enums are not much different from classes and therefore enums from different classloaders shouldnt be considered equals +Fine here a from and a successful build on Hudson resolving issue as fixed +Is there a way to add testcase for it +Going forward things like metrics will show up in right place but for now itUpdated our site so we have updated committers list and noted on home page attendance at nosql and hadoop summit. +Committed with regression testsNow code select will try first to match against a method of the exact name then a field of the exact name +Sorry didnt see your comment earlier we do if we enable in our +Thanks for patch v +Problem restoring conversations between Requests give out a new instance of on get +Lots of schema work will go into according to the revised roadmap +Adrian thanks for verifying +Dont have time to look at it yet today but probably can tomorrow +Thanks AdrianYour patch is in trunk rev +OK good to hear that it works for you nowJuergen +How did you do it +Thanks for your feedback +The fix involves using the Use class to determine the encoding style if the message context is missing +Woa! has been stable so it shouldnt be a total surprise +Going to have to close this ticket based on these last tests +Thanks Eli +include all logical relational operators +the reference handling involving the verison storage does not work correctly too +Currently it ensures that replicas are placed on at least two racks if there are multiple racksThere is a reasonable concern that we should improve our block placement distributing replicas on at least racksThis would be beneficial for jar and config files which increases the likelihood of finding the initial files on a local rack for tasksThe patch contains a method that verifies the number of racks the block is actually replicated to vs any required number of racksThe method can be used in fsck once the improved replication policy is implemented +cloned to +Let me do a version of Teds patch where I allow clearing of perm gen and set a higher max size on the child jvm +yes they should I am using it and have not seen any issuesThe Screen uses ftl for rendering table +I sure the steps include selecting a Web project and I do not select WTP option +Reviewed and committed patch . +Looks good test and +For instance if we flush a segment the DW doesnt push it up to the IW but once DW returns IW proceses the event an pulls the new flushed segment in +An approach which required that user code must live inside jar files in the database would run into the following limitation on the Security Manager right now you cant the Java permissions granted to code which comes out of jar files in the database +The proper behavior would be to remove them and add the appropriate regular imports +Source Test cases are okay +snjezana if you find a way to make it portable do let me now assigning to me since I believe its something broken in EAP not our problem +That is desired anyways as there is no need to just relocalize everything because of RM resyncThat is the way it should be +Yes it is still reproducible. +When was released? Is it related to the ant unit testing features introduced in Ant? +We should probably close this task when we deliver and create a new task to track this sort of work for the next delivery +We have been doing this manually by creating linked folders to the overlay projects and editing the wst config file but this is awkward at best when done by hand especially if there are several different projects that use overlays +What AS do you use Tomcat or +This issue is a duplicate of +everything is coupled to everything else and all tested against standard code lists etc +I am not able to say right now whether this will be also enough for larger test suits like metamer has but for now it works for showcase +Move version to dependencyManagement section +If you need someone will have to do a merge down first. +yea hopefully the contributor will collect all html attributes +The NPE is in the same method where you added the null check for tabledescriptor and your check might fix + +The bug was found in the documentation in following places +The current one is fine but its a bit sparse and not very modern lookingOne thing to keep in mind is that the JSP layer is being slowly ripped out and replaced with +or simply delegate the call to revision delegated evaluation to close for the release +Hi Girogio +Grant Date MXMLC Compiler and Tools +Havent looked at the code yet trying to understand the scenarioSo in other words if an application gets submitted to the RM before any NM registered the application will be stuck in pending state +Maybe it should be configurable? I dont think theres a way to give a warning but I havent looked too closely +Mina versionCommitted revision +Check it out and make sure all is as it should beRegardsThe patch has been applied and I moved the gram files to the project +There are a lot of util functions such as info killchildprocesses that we use in all our scripts +I ran all the tests some days ago recently I ran the tests that failed on Hive to trunk +Please see the forum reference +I just committed this +All it takes is form the root in Intellij +Fixed on trunk +Patch is nice +By the way you mentioned attributes +renaming INBOX now move the messages +Special cased and to do the following Print a warn message not a stack trace at ERROR level Not increment the failed request metric since that is meant to track a brokers failure to serve requests for a partition for which it is the +HTTP uses canonical rn so where is the mistranslation occurring? Is it just the tests that are broken or is the code problematic +suppose to match only +Usually when I do this I declare the cluster outside of the try block but initialize it within in which case the null check is desirable +Thanks for that assigning myself so that I remember to solve it before the release +Closing bug. +list in moduleCoretypeServer list in moduleJMStypeServerCan be list moved under JMS or are you blocked by implementation details +I believe this issue should have been long closed as it was committed to Trunk and last year +I think the problem is obvious and example is not needed +New Steinapologies if I misled you but I got confused myself with the code base +This is in package +but it supports them in element level +Specifically libhdfs is now integrated with fixed to automatically copy the executables to tmp on the remoted filesystem and some minor enhancements to the Makefilethanks just committed this +Committed to branch and trunk +netuitestwebappsurlTemplatesurlTemplatesformatter netuitestwebappsurlTemplatesurlTemplatesformatter +Integrated to and trunkThanks for the review +This has been fixed by. +Still reproducible on Spring CGLIB +These issues have all been resolved and just need to be closed. +So your workaround will not be of any help to me +After disabling all the metrics on that resource the agent hasnt crashed +This issue should be closed +Also still need to change CDH code + +Could you please commit this updated patch too NicholasSure +Id like to know folks view of extension points is it preferred to offer them in anticipation or withhold them until a need is expressedWe could either make the factories and all the classes private or make them all protected +I understand the advantages of your source lookup mechanism over mes really but Ive never seen any bug report about mismatch sources in me +is an internal impl and can be changed to work correctly with the existing API +Hi Julien The solrindex commmand and crawl script are work fine after setting in +Let me know if you have any questions +Probably fixed in close of resolved issues. +Did some code refactor for previous initializing work in some constructor calls +Ok I think the best solution would be just to have two profiles myfaces and mojarra and to push the entire jetty part into the main build area +Thanks Prasad! +hmm got a link to the docs that says it is recommended to put stuff in the EAR? That sounds horribly wrong since stuff in an EAR is not supposed to be available to anyone +So you get the benefit of a remote dependency without requiring packaging changes +Patch to add closing for spotting thisJuergen +Try navigating into the connector +just create separate issues no need for subtasks since they are not really related nor dependent on each other +The exact mechanism of how this is done can be debated but I think its a great idea that we could selectively relax constraint checks for trusted code insideagreed +a superset of this bug has been resolved. +Hi Nicolas Should I close this issue now? +Hence a app will never need this +Initial set of fixes to the service check script +DimsI dont think this is related to the JMS transport implementation +Hi ManjulaI took a quick look at your patch +Works for me thanks! Ive merged the fix to the branch in r. +to solve the issue as recommended by Jeroenplease take care of thisthanks the new version of the patch from AshuThanksvenkat + +I think this is a misunderstanding +Hi Micha I have this issue also I have tried also the new task gateway and I get the same version did this fix go intoI tried the latest bleeding edge and get this as well as me know if youd like me to setup a simple coldbox app and instructions +The output of Map tasks of a job should be accessed only by Reduce tasks of the same job +However we should expect that list to be typically in the range going as high as +Next time if you could keep the patch root the same. +This should be in +Whats the impact of this patchDoes it change json output for rpc? If it does change the dates returned will the JS code deal with them appropriately +After spending some more time poking around it looks similar to and not fixed by +In that JIRA a node is trying to establish a new view sending a cache topology control view but the node is stopping +It needs more works to refactor the codes now +IMO this should result in behavior but I will check with the TX team for their views +I just committed this +Although Id not prefer using facescontext to access request params in this case I couldnt find out where the extensions filter uses this method but it makes sense for a null check if its accessed when faces is not initialized yetAlso the file has has inconsistent newlines problems that prevent my commit Ive removed the svn property and reapply it after fixing this newline issue. +Attaching new patch that makes distcp to compare file sizes and checksums of sourceDir with stageDir also in addition to comparing with targetDir when update is given in the first try alsoPlease review and provide your comments +Committing v to trunk within the next mins +Closing as requested by Andreas. +creating the bean manually with the property allow false and a remember me ref in the client allowed the request to the login screen to go through but the page never finishes rendering although that may be my app more than the filter +Any suggestions for Sqoop? I would like to integrate it as it will bring quite a lot on our plate +Alexeyestimation sounds greatthats probably because most is compiled by JET or is it including JET code? BTW we also should count bytecode mapping and inline mapping +Ill after deleting one doc from the base index +Openmeetings is in a stage that has almost all we need +Personally I am NOT a fan of referencing external static methods this way +I just downloaded the latest nightly build on seam and I still am experiencing the problem with the artist initial listing displays the style correctly but when you type a letter like c in the filter on artist name field the listing is redisplayed with no different than +The reason for the original bug is that the logic that determines if an alias used is to be mapped to a relation is not checking the input schema for the nested operation +But Willem you would like to do you code in one placeSo if you need to count the number of aggregated methods or do some custom changes on all aggregated exception you can do all that in the same methodIf we skip the very first exchange then you cannot do that +How can I do this +I wonder if the AS logging configuration is the cause +Burr has some really nice diagrams that we should try to incorporate +Occurred on Linux +I think this bug should get fixed for if possible +Thanks Dan for the suggestions +Im tempted to starting dealing with this issue fairly soon +Congrats +Thanks Jing! +If you can provide some Java code that reproduces this we can try to sort it outI posted it as another report transition resolvedfixed bugs to closedfixed +Latest patch attached +Hi Alexiosthanks for your patches +Thanks for the patches +Pushing problem off to next version +Basic migration into test repository done +Is that the official unofficial extra components repository like for Apache Wicket? If yes and theres no point of commiting it into Camels official repository then there it should be +In a fully app you could then configure this thingy on your if you really need to have vendor interfaces exposed without any vendor as works as designed for now based on Juergens earlier comments. +Ive fixed the problem in trunk +This issue will be implemented in. +Rebased without changes applies atop andOur internal deployment of the UUID based counters is blocked on good read performance for compressed wide rows so our next step will be to integrate this patch with +Currently allocation to different files requires no coordination between threads this feature would requiresome sort of coordination for every allocate file creation file drop and compress table operation +Fixes space issues +Attaching a second rev of the patch which hopefully clarifies the point you raised +Could someone with the necessary permissions please close it? Thanks! +was a nightly build issue +But I think its silly to disallow that just because in some cases with huge documents and a giant max it can be slowbq +Part of a bulk update all resolutions changed to done please review history to original resolution type +Some static methods in are using old api get uses and Jobconf +bq +Currently no +Can you provide more details about your device + +batch transition resolvedfixed bugs to closedfixed +I will do that +a rather complete model of the language is probably any case if you want to contribute I can give hints on +FYI the Hudson integration comment above is incorrect as it is for another JIRA for which Ive used the wrong ID number in the commit. +I will leave it to Anurag whether to inline the code that figures out which message orto create a utility routine somewhere +It was a bug in Redhat thread management. +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +Quoting JVMOPTS is typically best practice but quoting it makes it fail when JVMOPTS is undefined +Fixing now as well as adressing other comments +I will do the modifications in in the following days but I would really appreciate if you could make the modifications in the eclipse projects +On second thought +The results are inAt a high level this patch does several things move from wsmexternal just to external so that netui can access it too add a class that has static methods for encodingdecoding +Who should bein the continuum unix groupCan I pull that data from +Committed to trunk in revision +committed to revision to trunk and committed to revision in thanks chris. +For each client RPC layer creates a thread for connectioncommunication etc +Further if such athread creates more than one reader at a time only the most recentlycreated reader available for possible reuseI may be missing something but it seems to make more sense to share instances across threads much the same way a databaseconnection pool shares database connections across threads +To solve this issuse need to apply this patch and updated in both server side and client change looks good can you change the comment to remove your name from it +Had a few problems compiling the testsuite but figured it shouldnt matter +Any reviewscomments are welcome +Ill work more on this in the next weeks however the current branch version is not completed as soon as Ive a first draft implementation working I can merge it on trunk +Help in resolution is appreciated. +Thus this issue is a duplicate of. +railo provide this attached emails as mails are not lost they are available as attachemts and because this mails are attachement this is the only right way add the to body column makes no can look into it if you can help with a example with cfmail sending a mail of this kind and cfpop that collects the extra information in the FBL message headers they have the followingAOL multipartreport boundarythen the report message at the boundary marker has messageHotmail messagerfc +There is no need to add more metadata to partition +Right +The mirrors are now up to date again so resolving as fixed. +I added a interface so users can supply their own tree model +Correct the patch format +These lines seem to be related to the issue +Hi Brett I am looking for a away to get all transitive dependencies belong to a plugins dependencies section +The version in the footer is the version of the console right? I think Rich wants the version of EAP to appear somewhere in the console similar to the welcome screen +I did manage to get the EJB spec changed so it can be implemented portably +A simple Handler now extends +Rajika may be you can review this and see if the solution is still applicable? I think the problem still exists in the trunk +Done in trunk at revision +The problems that might occur are all solvable but I want to take enough time to think about the possible consequences before starting to codeIf you want you can open a feature request for it or I will create one sometime +However the example supplied isnt enough for Aegis +Ah cant edit issue +I think youll find that policy adopted in much of the code base. +Ram no problem +Those entries are no requirements and thus removing them has no to add a check to not display the red x for those items. +The target for the branch creation is Jan +Patch +Startup safe mode is not affected +If none else looks into it I will take a look next weekN +After better understanding of how datasources are createdmanaged Im no longer certain this is a good idea especially in domain mode +That contract does not say that an order property value has to be specified for all such Ordered beans +Ive attached a patch for two reasons +This should be fixed now with a rework of how functions its much simpler and shouldnt cause this anymore. +Can you do a review and make sure it all looks OKMoving to as there are some complicated issues surrounding this patch that need to be thought throughColm. +Brett can you clarify a little? I hadnt realized that the nested objects dont need to have getters and setters +And if this has a potetial of impacting users then we should perhaps not change it at all +Thanks Michael. +bq +The work manager spec proposed for javaee apparently does specify the behavior you want for use of the executor +closing this issue and will create a new issue when kabir is done with the new classloader implementation. +address some additional test failures. +Thanks so much Gavin! It works for me +It ranges from ms to ms sometimes +I created a mini webapp with both a Java annotated file and the groovy library in the classpath +So I call this one resolved. +Please see this issue for a way to go about this +trunk . +I came to know if we provide regexComparator then can workI will prepare patch with that and provide +Fixed in latest development code +Thanks +Hi for the RPM part asked RCM to remove the javadocs for TS next build will remove the river part we dont ship the javadocs RPM +So on GA +Instead it prints a warning in the log +Merged to the branch in revision . +Process button label is changed showing that form has submitted despite validation error +Use RUBYOPTrubygems as a workaround +I fixed that in revision by using the correct default image typeThanks for the report +If we do so lets be very clear about the direction of things Is bigtop moving towards using groovy as a core element of the core package build and deploying? As of now groovy is only used for and and itest +that is possible however still leaves the client guessing with regard to operations successHow? We reject on the way in before any operation has happened so client knows its attempt failedbq +I think that did can close this a lot for all the effort +Committed to and trunk +Previous patch missed updating the help text for rmr to include skipTrash option +When the even contains exception the runtime will kill the system. +Functional test for file movement from local to remote machine +On the Samsung devices I have once you take a picture it gets added to the gallery then when we run the compression on the returned picture and store it in the content database that too is added to the gallery +Ryans idea sounds like the simplest way to get something good enough to me +I see that the fix version has been updated but Ive already committed the change +Something really generic like perhaps? We could gomore specific and have an +File that results in negative values +I want to continue the other testcases after this patch applied +closing issues for released versions +srcjavaorgapachecommonslangtext +class A implements X Y? Then Alternative class B implements X would only be used for Inject X x but not for Inject Y y injection points +Need to tweak this patch a little bit more +I fear this will make life hard for users as well as QA environments +Ubuntu r +Attaching fix +Thanks +Would need a reproducible test case +I have commited changes to verify that it works as expected +If we only put in newlines between records all of the needs are met using a standard layout +What do you mean under other NTLM authentication termBTW my first name is Konstantin +However I left the method disabled as the cause of hanging is not known yet +An attached document with screenshots for those who do not want to run the example application +to David the data directory is a better place for +Closing this +have you set scope cascading in Railo AdminSettingsScope to strict +Just need to find some time to write the code and give it a benchmark +This way it is easy for others to undo things they dont like which is much less annoying than to untangle such changes from real changes +Seems like we should make the util code common +I also wrote not several timesDO NOT do ssh to use your browserrgdsjan has SSH access and has started with orthpasswdOnce orthpasswd opiekey is successful will be added to suoers and the issue closed. +A somewhat similar case occured in that issue contains pointers to relevant discussions on our list +Will be fixed as soon as we will have merged the current MINA branch into trunk +I am moving it to m weary about having this in before because of the changes in place +Should probably add a few javadoc documentation lines into explaining why it needs to be cloneable since subclassers need to bewareThe class is made final to make sure it cannot be extended +Thats a very simple change and addresses this ticket +Okay forget I spoke I cant see any easy way to get a reference back to the form from javascript since the form name is dynamically generated in the widget codeIn light of that I see now problem with using freemarker widgets are great but they should get in the way of additional functionality +Teds suggestions are good +good overall +What does this script srctestresources exactly do? Looks like it just reads the arguments to this file and prints a value to the console +Hi Kathey +Bulk close for Solr I was hoping to find a way to disable this vulnerability in the +in all cases the PDF is correctly generated. +This task has been completedAttached the bug report the in lib to ANTHOME the modified to jms Changes in are there in using cvs diffant finbugswill produce to be closed after applying the changes to the repository +Here is a slightly updated version that contains a list of all the releases and a reference to the development team +but you should probably removehide the http http from the configuration file +Agreed with Len to document this and close. +Does this mean just the commit test target or all of the tests? Since were relying on existing tests to verify this refactor itd be good to include the and results for the whole test suite +It would be nice if one more person would look at general approach +Hi Kevinwhen i download the test case the Japanese characters become to unrecognizable maybe I have different locale setting with you +I confirm it closed +Marking as Duplicate. +Agreed that this should be done at the coordinator node +Thanks Mahadev for taking a look into the patchI updated the patch with for the aggregate metrics +This has not fixed the issue +With r ASL license headers have been added to most files under importcore +And this IMO applies to all tests used in this manner +Uploaded the wrong patchfile +It should changed as. +same problem with Grails is obviously version problem +Will open a new issue if new warnings appear. +Is there a way I could use the local installation of ant without the root having to update the ant installation +rpm installed before beingable to be added to Cloudstack setupI was trying to test if the generic build exhibits CPVM SSVM problem +Good catch Ted I must ask the same question myself +for and getRegarding Of course I do not object but Im not sure if it would be worth the effortBase encoded bodies can be quite large but I think is usually used for relatively small text bodies +We have a lot queries that are like that because of +This was updated +I dont see that this will be a big perf issue as we already cache resolutions for interceptors and decorators. +Dependent on +To stop this occuring the JNDI definition can be configured to time out if this occurs +Committed to the branch +Why is it marked fixed? Can we the issue? Or should I create a new one? +Juergen You are very right +Also I included the source +Can we remove synchronized from noDupsOK Ill remove it +Its also available still if we want to refactor the script into an improved balancer +Moved to the next version because code has applicable to the code of currently serving for reading configuration parameters. +I want to commit this to trunk and will do so +seeabove why +I think we still have an IT in place that deliberately grabs multiple artifacts that might need to be revised to ensure they are from different groups +The point about consistency is taken +Done as outlined above +Derby logo with a theme +I did a test commit an everything worked fine for me +BrianIs this still a problem +But they are depending on this change from trunkThanks a lot Uma +I will submit an initial proposal so we can start discussing further on the subject in the dev list +Thanks Pranay for your contributionYour changes are in trunk rev +Im going to try to address all of this in the same patch +I am not really sure whether I like the idea of using the INFO log level or not as this might create too much noise for said that what would you think about adding a new property to that allowed you to switch onoff the logging of the SELECT statements based upon your needs +I tried uploading to http and still redirecting to http so no protocol change +This work will listen to the specified tcp and ssl ports on every interface. +Received and incorporated review feedback from Adrian Trenaman +Added specs to the distribution both and rpmbuild One ugly catch to this is proper dependencies tracking is not performed for cppmanagementgen +Version has been released and addresses this issue. +I think well call this one fixed +Stoppng and starting activemq was not sufficient to bring things back +I think what Howard needs is a pointer as to how we expose objects to the container to do this + resolved issues for already released versions +Thank you +The runtime behavior of can be customized in different ways HTTP parameters being one among several +One quick workaround could be to support reading content from a file pathtofilewithexamplestableIf the content of the Examples starts with a it would be understood to be inlined else it would be contained in the file specified by the ideal as it becomes less visible from scenario text but it would get around the issue of the parsing +all bulk edited issues have in a +this needs a closer review since the original reporter had issues with SVN and the patch only covers CVS +VERIFIED. +I guess it could be done via separate JIRA +I want to go ahead and put in the additional changes to along with your patch after I finish reviewing it +poking out of some clouds with the infinity rings circling the peaks +Are we going to introduce this stuff in? I thought was all about generification +I moved the common code to a new superclass so no method of needs to be overwritten in +Seems problem has been fixed on Metamer side +should be added as they are identified +Thanks Prashant! +The app just uses it +i meant the files and to be identical +Thanks Colin. +please check again the formula in the merge part +However Im in agreement with this improvement and fix +Thanks Joy +We need a set of login modules to support certificate based authentication +simple test patch which qualifies the way we cache XSLT style sheets by including the bundle URI in the key was applied at revision . +As far as I can tell it is only the classloader viewer that doesnt workThe dependency JNDI and JMX viewers seem to be fineI have not tested the LDAP viewer thoughI will take a look at the classloader viewer to see what the problem is +Remove the groovy classpath container and things should work fine +It should result in the EPR being removed from the SI cache and the invocation being retried with a subsequent EPR +Trivial fix for y +I think this was kept private originally out of concern that the API might not be stable +They should coerce to since that is what they represent +Yes I removed the Location header as you said in your last mail in the threadThe bug must have crept in when I removed the lineServes me right for not testing it again +The big issue is what do you gain by running TT and DN per node except on testing where you want to give the master nodes more of a workload +Ill see if I need to take you up on itJust ping my name in the issue +All involved doing more work hear and less there or changing this classpath insteasd of putting a file into X folder +So we should not worry about it lets use a clear to understand nameRegarding symlinks yes checked does not implement symlinks was augmented to have such info not to have new class for +Recreated patch to let it work with latest trunk patch with thanks to Piotr. +I did think of returning void when the service is healthy +trialV fixed Ram comment +Punting issues from RC to final +I also changed which was committed in to use junitFor comment the constructor is used in Reflection so can not be removed +Patch lgtm but we have to wait for a jackrabbit release as we dont have snapshot deps to external you make the default constant of type boolean and then use boolValueCONSTANT in your property Julian for your patch +It is not intended that you access the Session in any way during a flush +In Apache Derby anybody who went through a quick signup process could own issues even if they werent committersAnyway I was looking at doing a first draft of this but couldnt really figure out where to put it in the Wiki +This is potentially new issueWhile we havent implemented anything like or some interface to configure invoices I guess all suggested variations are fine because I think that invoice form and the exact text and layout of items will be customized by each user +but do care of the old file if it is not working as expected +We have been using metric in our prodution cluster for a year and its good enough for our rapid disk repairment +For most common hadoop auth work load distributed containerstasks you dont need to guess its the delegation token with scramIn a mixed security env including not just within the cluster but across different clusters with different versions or during rolling upgrades when supported mechanisms are upgraded it will always be a guess +The drawback of having such global checkpoint would be if involved processes in computation increase rolling back to a consistent global state is an overhead +The patch file should be bytes long +Looking to rectify that +I will have physical access to the computers only Friday evening + RandyThanks a lot for your patch we really need this! As I was saying on it would be really nice if we could keep it compatible to the already existing js +r +Of course mongodb does not provide direct cyclic dependency when querying but it does not forbid the cyclic dependency using dbref +At first glance the patch seems ok to me +I used the version Sonar and this issue occurred +Marking patch available though there are some questions about the attached patch +Note that the metrics necessary for this approach could also be used by the load balancer. +I think a new page called srcdoc should be created and we should collect all information that is Java specific and movecopy it there +Ive removed my toolbar from the you tell me where their toolbar is so that I can optimize it and remove the lengthy initializationThanks +Pushed to beta only not +I ran derbynetclientmats on ibmlinux with no new failuresI ran this test test in both derbynet and derbynetclient frameworks with ibmibmibmibm and jdkjdkjdkjdk OK +Can you take a look at the testcase and correct it if possiblethx is a simple test case that correctly demonstrates this bug occurring +Ill look at the changelist for but just moved to for same reason to no effect +Sorry for that. +Hi Andy what work is required to get HTML pages supported hereFor my own web app I have forked Fuseki but ideally would like to make it available as a webapp from my Tomcat web server +Patch to fix the bug for Hive trunk +Ive fixed this now and the state of the repository should be the same as before the tests runAs changes to aliases and mappings are processed asynchronously Ive added some sleeps in the test which result in a longer test run but for now its better to have a longer running time than failing tests +Patch committed to committed to trunk +Committed at subversion revision +However the problem is one of timing +I think that if we verify that with this manifest the package loads in an environment were done for nowShip it +Thoughts +resolving wontfix +John here is a patch that solves the uses of deprecated in unit testsJerome +Sorry I didnt mention this before but I think it would also be clearer to replace task tracker status with task completion status +Yes I also see couple of places fix is get misused +This is the latest work on this issue +Ill give it another go when I have something to read from HDFS rather than S +The problem has been fixed and the task can be closed +The fix is straightforward but I still have to writetest cases +to trunk and +Hi Leandrothis is either an installation or configuration problem but no bug of the softwarePlease use the mailing list to discuss your problemsThanks SebastianThanks for your answerCan you tell me wich config i have to changeBecause the log dont say nothingRegards can have multiple issues depending on your integration code +What is the URL for the post +XML deployments +I tried with both and and both produce same ve been able to reproduce this issue using curl to add documents but using the provided in the Solr example or using a solrj client Im unable to reproduce this issue having tried batches up to documents +Negative +Assigned to PHPCLIENT component +Since the inner class holds a class name of the outer class the outer class needs to be loaded first or else loading the class wont be possible +Tomcat only looks at the first element if present +Cannot see the bundle on central +They are now more general and work with local variables break continue and return +I have an implementation for both andIn I have deprecated onSubmit and added on and on in an extension interfaceIn I have moved onSubmit to on and added onIt works the two methods seem intuitive so that is nice +If the content is included within the configuration it may need to be the XML format required by import +The ESB issue has been moved out of our release for the platform and back into the project release +I confirmed that the test passes on both Mac and Windows for both trunk andOnce again there was a minor formatting difference between trunk and +The problem here is that integrationtests are one big git repo where there are no separation between if its testing core or should seems very wrong to me we are putting soatests into a site which contains tests for core which will have things that are directly incompatible with old think there are some lack of separationstructure missing sounds like the stuff just shuold be on its own site and then composited into coretests and soatests keeping them t that solve the problem and keep things cleanly seperate and possibly to version based on the actual release cycles +For this jira the client will still return paths +Patch looks more than good I am going to apply it thanks +This is committable if the internal use of Spring is tolerable and then I could proceed to add the supply a context file to to customize the databinding bean +Even if there are multiple relationships should they not each have a different tableI know this can easily be fixed by the Grails user by overriding the mapping but the naming is very counterintuitive and therefore will most likely always be adjusted +is a Linux platform which I describe above +If this exists my code +A user that decides to override get would know that we will freely break it as we see fit +rev handle os file path separator. +Sorry I should have been more explicit when mentioning the TOKENNONE string +This is not a issue it is basically to allow a administrator to allow updating datanode jars without updating namenode jars. +I just committed this +if the issue is not persisting for you shall we close it +I have committed this +Moved to release beta +Sending srcmainjavaorgapachesshdclientchannelSending srcmainjavaorgapachesshdclientchannelAdding srcmainjavaorgapachesshdcommonAdding srcmainjavaorgapachesshdcommonutilSending srcmainjavaorgapachesshdserverSending srcmainjavaorgapachesshdserverchannelSending srctestjavaorgapachesshdSending srctestjavaorgapachesshdAdding srctestjavaorgapachesshdutilAdding srctestresourcesorgAdding srctestresourcesorgapacheAdding srctestresourcesorgapachesshdAdding srctestresourcesorgapachesshdclientAdding srctestresourcesorgapachesshdclientchannelAdding srctestresourcesorgapachesshdclientchannelAdding srctestresourcesorgapachesshdclientchannelTransmitting file dataCommitted revision . +The different managers would all try to store the same sessions which might have weird side effects +I applied the patch at revision taking the first two bullet points into consideration +I cannot install IE hereThe source code of the demo application is too big to make a full code review +to have committed to trunk +The patch uploaded has few things to be considered The test cases are passing most of the times +Thanks Antoine thatll be one more beer Ill owe youSending CHANGELOGSending libbuildrcoreSending speccoreTransmitting file dataCommitted revision . +Btw this issue Ive already seen and reported its +Which are the other pom upgrades +Heres an early version of this patch +could you post screenshot here +committed +bq +duplicates +Needs some more investigation +Also it should be more spring friendly configurable using settersgetters style so we have IDE completion in spring xml files +Nadya the changes are on the website already +Im now digging into how link state is managed after the connection succeeds. +It seems that the php version of shindig doesnt return opensocailappurl +Jeff you will need to merge this with any outstanding changes you have for the build file +The code in executes at run time not compile time +Committed at subversion revision +Thanks very much for your help your workaround is effective +Shows distribution of data file Marc +it is implied by the primary node typeSame for jcrcontent leaving this optional allows for empty documents +Attaching the patch +Can you rebase it to reflect the mavenization +Any chance this could be revived? The IDL still doesnt support types +I just committed this +HiPlease import the project into EclipseYou could generate the eclipse project with the maven command mvn eclipseeclipseNext run the test calledORrun the command mvn test +If I want to trust the machine I need TLS which is completely orthogonal to this ticket +the and docs were also updated +I was unable to test however and wanted to just get this up +AndyThank you for your comment +Tested service and on both Ubuntu and Red Hat +Thank you so much for this contribution to the projectOleg +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +BTW this bug is easily caught if the DTD for sql elements would require a name attribute +What do you think is better keep my which maps the related fields or write a routine in java which determine the statesHave a good day Regards version is in Trunk thanks everyone +OK I removed it from the in +You overlooked that two different artifacts are in use and +The BRMS team should be able to provide that content and ideally just update the upstream docs accordingly +Can you please update your comment with key cache contains xy keys to the results on the current trunk but like you did for After this patch results? Btw did you drop the page cache using sync echo procsysvmdropcaches before running After time patch tests +KrisThis might be related to some of the work lucaz did +Also the app is running on Solaris in both instances we are seeing the problem +Defer issue to release +Rafal to the best of my knowledge in maven java source files are first compiled to class files using regular java compiler then they are weaved with aspectj compiler +As its already resolved theres no further work for Kevin to do +Thats what I effectively do in Smartfrog where the configuration also tells the runtime what services to deploy on startup +patch added +I dont really have time to work on it but using format would be nearly ideal for index blocks +Moving this issue to later version as we currently still miss some features +Was the revision done against updated trunk +also is reserved name in eclipse pde I suggest instead +This issue should be fixed now +Milind According to Sanjay the reasoning for this second implementation is lack of wire compatibility in native RPCI said +Where might the info on this decommissioned node be being storedAfter we store dead gossip states for days so that any other nodes that were down at the time of removal can know later not to repopulate the ring with the removed node but this isnt persisted anywhere so since you did a full ring restart the only candidate left is the persisted endpoints though all nodes should have removed it from there after the decommissionremovetokenbq +This way if someone is already serializing this class it wont break their code if they dont upgrade both the client and server at the same time +Since these routines are builtin functions that dont need privilege checking need to bypass creating objects for them +Im closing this as cannot reproduce +How will you handle that when you create the slave database? It is probably OK to say that in the first version it is not possible store the log separately from the database +the Employment Tab +skipTests and model test are probably to blame once verified you might need to raise an AS issue I will added into job to create an empty jar after clean which allows the AS build to AS issue +Please use the user for questionsRegards + +I can have a look at make sense now we are migrating to junit so closing +Would you consider simply moving to wholesale and removing all the facade stuff completely? And by you I guess I mean me since its my dumb idea +Assigning to Brian until we can assign this to Yohei +Corrected fixaffected versions +Committed at subversion revision +Version supports Scalatest and Scalacheck +are published to apache site +Attaching log with lock errors +LGTM committed. +Hadnt heard of the disklog module before but it looks interesting +So its possibly a Cocoon thing after allBut what about applying patch so that we get better error messages +I have reviewed Brians patch and believe it correctly resolves the problem +Testing needs to wait forCompleted At revision +As was announced I want to ask again if this will be implemented in the future +Bryan I have just checked and it looks good now +and setters for timestamp number of versions and lockid and you reduce overloads to constructors and settersThis is my opinion too +The interface sorts the codes as numbers +The attached project runs to error when started from the rootfolder +well I think it is false and only strings like true y or yes should eval to true? Doesnt that mean you will have to check the String befroe hand? Waht do you do if is a Integer? +Updated fuse to latest release which fixes these tests along with some other bug fixes +Ill take a look at that example +If security is needed this is a custom extension to these components and can best be developed separately from the Activiti project. +Thanks for the suggestion Andres. +Im marking this as fixed in since I believe some time earlier Bill added his new which handles these sticky synchronization issues much better +This is because RPC serialization does not allow null parametersI rerun all the tests +Committed POM file for the new Maven project file structure. +but i do it again +Have modified the shortcut to have the correct path +Thanks Lefty. +Just updated relnotes by committing to trunk +Though dont take that I dont agree as a hold up to finishing +updates apr links in +This triggers a startLoad +Vladimir have we got to the bottom of this? Or are we reinstating flush in for now +Fixed again + +Bulk close for what happened to these optimizations to regarding the direct copying from via its? Cant find a trace of it anywhere in the release or in the trunk +This is against the assembly plugin not against site +Thank you very much for spotting that Deepa! Backed out with revision +According to error I may assume that you choose TPTP Integration in JBT installation but do not put TPTP into Eclipse folder +bulk move of all unresolved issues from to +Next up part making sure blockcache gets cleared promplty +i can confirm this though i neither can give more info +Opinions anyone? This seems to be sensible +This should not happen +Attaching a patch +I really need to come back to this and simplify it more as I fear Ive made things more complex +Setting the expiry of the cookie to years in Rev +yes but I need to know which one should stay first +We currently require java so clearly this bug isnt present anymore. +Changing the Fix Version since this is definitely not must fix for since it is a patch against the branch of jdktoolsJim please dont set the Fix Version field unless the big must be fixed in that version +Currently uses the clients bundle classloader and not the proxied interface classloader so you can run into issues when adding the proxying on top +Daniel please reopen the issue if there is still failures +Hey Ben can you please change the followingFind and replaceentity name +INFO Running for rows took ms +This ensures the contract which has the same lifecycle requirement as the +How much work would it be to add just one more bandaid for the stream source thing in comparison +If we can get the fix for this done ASAP I can make the switch and try and sign off on +I wasnt able to get to it tonight but Ill have a look tomorrow +this is almost the same like isnt +Though if we can support it with the new impl fine but I dont think I would go out of the way for it myselfbq +this patch fixes this issue +So either a slow reader or a slow writer might cause a read failure +Some things were not clear so Id like to denote that this issue should fire a code change +The simplest thing may be to just have each node be able to return the factory for itself and then ask that factory to create the clone +which means storing the timestamp in a file each time a target is called + +Looks like this must be a feature of my set up I am not getting true or false coloured in Java may well not be a bug at all +Also I have added constructors to that take respectively so that if this is not the desired behaviour then people can define their own +The error is happening even after Ive removed all Xml +Best regards George +Hi Eli What testing was done with this change before it was committedThere was tremendous testing went into testing these patches +I think that this makes sense +Im open to feedback on this because it is a hack and I was actually surprised when the unit tests all passed +srcmainjavaorgapachehadoopyarnclient +thanks for you answer i test it and I keep you works correctly with the kindly natureThanks +Sorry I didnt mean to set the fix version +Ive also already tried use dev update site without any difference +Now that work is underway with AS all previous community releases are +Fixes attached for and +Thanks Shai +She also found that you do so much work in bug newest patch +I see your point +I merged this to after talking to Matt +This would make it more easy to show providers or images which are faulty in this regard +Done. +Verified fixed it. +Unit tests have passed patch is ready for reviewSummary of changes The hadoop wordcount example modified to use a stopwords file +Later they are replaced by the real gbean datas +Scalebar and other map graphics require interaction and need to be in the Layers view for our current s keep the Layer view clear and on target and start up a legend view for when the user switches from definition the layers in the map to communication what the map is up earlier idea was to fold all the map graphics and background layers into a folder so they can be placed out of the way unless needed. +I am pretty sure that the database cleanupcode reports junit errors when it retries but cant remove files +Retargeted the issue to since is being released very shortly. +Is a string of the form filemyDirmyFile really a relative URI? Im not sure it is the hostname isnt mentioned at the very least +Attaching patch supporting multicast discovery of file appenders and socket appender from trunk revision +First of all I dont see why the renderHead method should be called by pagedetach +The test needs to be rerun with a new Hadoop build +Chandan I am reoslving this issue please file individual issues for remaining properties It is hard to track this way when multiple folks have to fix this blockers and critical issues are resolved but not verified +For Add a option and enable it by defaultFor Make hsynchflush behavior independent of whether is enabled ie you can turn append off and hsynchflush still workNote this does not add a option to This would be useful to people who want to disable both append and sync eg to make a couple paths cheaper +Fair point +log the error if no file list would there not be a filelist? Any idea what the underlying error is +cmake seems not working on jenkinsThis is +Thanks for identifying the problem Ted +number of cpu cores for APR +Thanks Dan for catching this +Im getting closer but I still get an error Failed to start database largedb see the next exception for detailsERROR XBCXC Encryption algorithm DESCBC does not exist +No longer applicable +Other pages are still under construction and they are not linked to the home page although the home page contains links for themThese links and the contents for the other pages will be submitted in the near future +should probably have fix version in preparation of cloning another issue to handle any further method is now available to make basic escaping easier +Can we look at and maybe first +Ive fixed the new regressions in e but Im not confident that the new Delegator is perfect just yet +The actual problem occurs in a background thread initiated by service invoked by a controller + +Moving to next release for now +I will investigate further +What should be changed? A timeout of minutes sounds long +Added patchesPatch applied +Thank you for your help +Id like to resolve this issue as Invalid meaning that it is not a bug against Xalan +Output is expected +TimItd be easier to fix this if you supplied a patch against +Fixed +Ill try with Maven and post the results +libcassandrabin +MAY be processed more than once if they are embedded in more than class +Yes that would be nice +Dns configured website up +bulk defer to close of resolved issues. +The list will be split and helper methods will enter the code elements +lets see if it makes sense to wait for it +Any ideas why? I am running repo using attached configuration files +This patch fixes the problem +I think the build issue occurs after updating to this change with the prior tomcat version still in downloads from an old build +But with replacement just if we run the same command again we can continue from the place where its failed +Please check back in a week or soRegardsShri +closed +Thanks Han Hong Fang for the patch +It would be best if could avoid any dependency on Spring at all +I added a check disallowing not before is using ne reads better and there is no eq operator to link to the not +Final as is. +only target +Thanks Eugene +Your suggestion of is good too +ignore the last comment its referring to have you reviewed and agree this can be committed? Chris gave . +Go the forum. +and a first code contribution from me I cant see the harm in including it +Honestly I do not know +I think it could possibly be a doc error before which then makes the master files correct in and onward +Tests passed cleanly for me on +For you its the wrong spring or the wrong hibernateThanks +How does item relate to and? Item is very specific and it very small patch it confined to just only one function +Or we could do something seemingly ugly but probably very effective and safe Try running ls d devnull with numactl and if that fails assume it is because numactl isnt working +Test too it fails without code changes and passes withJust committed this to trunk and +Currently Im busy with dayjob maybe can find some time in the late evening. +Thanks Prasad! +If IOE is throw during the loop just log it and continue the loop as soon as possible +Done +It seems to me that this problem no longer exists with the axis transport +I think someone else should look this also but if you think its ready then go ahead and rebase against trunk update the patch against RB and attach the latest patch to the JIRA +Good thinking Andrea you are a champion +That file is already removed on trunk I filed for that +If you find further issues please create a new jira and link it to this one. +Vikas I was working on making the page more intuitive +And how is JPA related? If fixes it what was the related Jira issue +Verified at +It could happen that the file could be removed by the retention policy just as we are trying to remove it +fails +I will be reviewing this patch +Three different configuration files were added for the test case +their files before generating a patch +do gpg passphrase passphrase output armor i donein the file? Or is this done by build updates +Does anyone know why were patching a file instead of just regenerating itDoesnt the maven ant plugin generate a build file with the correct javac task attributes +Im testing on server Erlang RB libcurl Browser is Chrome running on Ubuntu +Pete is this a blocker from your opinion +Looks correct now +For sites running and earlier download the acegi jar independently and replace the jar in the lib dir +Provide integration this record since is now released. +sounds very reasonable go for it +The locks patch actually included concurrent access tests see the +Ok I did a small improvement in Axis to ignore the schema of types +I went back into my code and I thought there were places where Spring Security filters could throw exceptions that Id like to handle with my but I couldnt find any +It only happens on my LinuxGTK box on my Mac there is no problem +But I think this will just recurse hit thetrigger limit and throw an exceptionAt some point in the future it will be possible to have users have requested before triggers that can modifythe NEW values there is some outstanding issue for allowing procedures in trigger statements which could openup a lot of possibilitiesMaybe there are other ways to get on an UPDATEIf no are possible then the proposed behaviour is ok though the assumption should be clearly stated in the spec +no should just be a matter of setting the right preferences +Having said that Im open to that approach +Rod BurgettAre you still interested in getting this fixed for groovy or can this issue be closed +We shouldnt ever be writing files to these search paths +We have the same problem +Committed +Also added spec to rubyspec project +The logs would suggest it is being retried eventually ending up in the DLQ +Its all about mashups and random useful libraries +ve committed this +most of the event propagation is done +Ive seen this in other cases that I have not yet received permission to release +The only thing left now is documentation which is as simple as adding a paragraph to the appropriate +You mean any component to have a validation step of some sortThanks for taking this on +See my last comment +Otherwise we can postpone it till release +Just an idea as some problems were reported sometime ago OGNL was built against given version of Javassist and using with the one provided by was giving some strange errors +Does it really have to be within the same transaction +Property view doesnt show properties for tags +Attaching very large patch +AndreyShouldnt cpu system be used to check for thread safety issues? +Then we use this project to generate a web serive the wsdl porttype name is helloworld too +That sounds like it should work and would be easier to code +Its writing a bunch of dataThrift takes mThats it +Whats up with all the changes inI think some comments in the code would help its readability like the fact that we are expecting or a value equal or greater to +Confirmed this affects the branch as well. +Reassigning to to Len as to put on Lens radar +The rest of the patch consists of new tests +Is this available as a packaged plugin? If so how does one go about installing it? Thanks +Committed to trunk +This is just what my spreadsheets do declare patterns without constraints before a pattern with constraint + +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I believe this has been fixed. +Patch is not applying cleanly +If we remove the jobID corresponding tables on different pages will share a single state +Moving out of +Thanks to Andreas and Oystein for pointing out another possibilty that when the statement objects are GCd the activations associated with them are not closed which could be causing the leakI shall work further on this and have assigned this one to myself +I talked with Clebert and theyll look at cutting a or release for AS to bring in this fix +Right I didnt mean all would be woken up but as long as thread is awoken there is extra work being done +Provide more informations if the last comment from Noel is not the solution. +Absolutely +Check +It would mirror the KILLABORT KILLWAIT states that exist in the job kill pathSpeaking of the KILLWAIT state if were using a timeout for FAILWAIT it should be applied to KILLWAIT as well + pending Jenkins +If its in a daemon the NN will notice the RPC server crashed and may initiate a clean shutdown +DaveNice I like it +So the IBM JDK needs to be updatedWorkarounds described aboveMadhumita can you try please +Thanks for the patch Bilung +Fixed usage screen location +Attached an initial patch for computing the total group countCurrently it is implemented as a separate collector +Things have gotton vastly more powerful but also more confusing +There is no way of knowing what the corresponding source file is from the perspective of the compiler +If the model is Basic and false the tests will both pass +If you add a unit test w a MB file I can test on windows +Will investigate +So all issues not directly assigned to an AS release are being closed +What is the distance metric that has been usedBasically I have a use case I have a set of million urls which have an inherent hierarchical relationship and a set of on them +Close after release +GBdatabase with documents +Eric we shouldnt use the directly we need to go through theAlso the readLock is imp +This fix to this issue is incorporated with the Console upgrade in EAP console +This was done inIt will be part of version thanks +In trunk this is fixed. +Never mind I have verified that is in and +far as I know Im not a Confluence Administrator only a space admin for the Apache Hive space +Thanks a lot! +is used only for the first columns code examples Facets richtree attribute is hidden converters are common messages attribute is Richfaces resource dependency maven plugin mentioning in dev dragSupport bypassUpdates attribute is removed +This is add another complexity to this case. +Add a field and set it in setConf +by thrown a if any of the activate deactivate modified or attributes is declared in a DS descriptor +thanks Simon +Id like to take over this thanks. +So it looks like this bug has a workaround. +One question that seems worth asking is is it right that the offset returned on bzip data cannot be properly bound to? If possible it would seem better to correct this issue in the loader that is dealing with bzip data then forcing operators up in the pipeline to handle the fact that not all data formats will properly honor the bind position +Flag this issue as resolved waiting for confirmation that it works for +Note that this resembles bug which is marked as FIXED +If you have time please verify that it does improve the performance for you. +I am working on an improved patch that will be based on the discussions here and I will submit it to this Jira issue for review when it is readyWe will need new subclasses to add support for Dojo check the dev mailing list for Discussion Screen Widget XML format to support rd party rendering libraries +I agree that we should hide the full functionality of the throttler until we understand its implications better if we go with the throttling approach +While having a separate PID directory per configuration is not a problem per se it seems unnecessary. +The now appends extra path information when available +has a custom opscodeSince it is not in Apache we could simply reject the patch +Basically despite the jar being in the right place when the jsps are accessed the container did not pick up the jar because the remained unchanged +The fixfor is not decided yet and it cant be in the past as we havent invented a way to go back in the past yet +That message is expected +Then we dont have to fiddle with the userctx and we do less duplicate work after rewriting. +Attempt to resize shows that it feels it must be flush with right side +On a cluster one can manually delete the corrupt replica which should trigger replication +Jim can you attach your datasource configuration here please +Effective only on the linux system +A patch for +Internally this seems fine but exposing this as a servlet seems like it will encourage the case where runtime configuration diverges from whats in the XML configuration files on diskCould we instead have the servletAPI be more similar to the existing refresh functions where it reloads the configuration off disk and applies a diff? Im imagining the servlet would show you a preview of the diff to be applied along with a list of warnings for configuration variables that have been changed on disk but are not reconfigurable +This works quite well because most tasksphase list are sorted by name +Cool good to hear! Can you change the fix version then +Updated packaging to now include all of Util as it was getting ridiculous trying to get the exact list +Applying it for all dependencies of a plugin would be desastrous and make very brittle builds because an upgrade of a dependency of my project could break a plugin used in teh lifecycle. +I will take a look at these thingsI will need to regenerate again anyway as other udfs have been added since this oneAs to so we are assuming encoded is not an issue hereNamit I just got my apache id! so when we get to a status I would like to do the commit on this one maybe I can contact you offline and we can walk through the steps and any specific things you guys to before commit +Committed to JBESBGACP with revision . +Oops +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I suspect this fixes the issue +Thanks for the patch roman and thanks for the review anoop +Ive fixed the issue with version +Ill try and upload being set to and what is your umask +Please provide a quickstart + +patch file with thePlease review the attached patch and let me if the fix requires additional workThanks on behalf of Dilli via baaeafcbdfcd +The XSLT spec does not say that we haveto output elements with any particular prefix +for Martin can you upload a psds +The CDI Facet version is now infered from CDI dependencies found in the project +This seems to happen only with via our +Marking this as resolved +So have we reached a conclusion on this one +This is a simple patch to correctly receive a message ID from an message +But using sessionStorage would mean that each time we opened up a new session we would have to the settings +User is fine with the workaround +Hope this helps +If equal names are causing problems GWC should at least output an ERROR entry upon startup saying something about this requirement thats broken +But it is different from because the pending Delete is created at a different point in the code in the timeoutMonitor and not when the task actually fails +close in a few days. +I do not see these failure +reworking this to use Seam +also your sample page for the web application looks nice can you post it as a patchIll leave the issue open in Jira to track the documentationtests lack +Attached with changes in newly added stack +Set HBASECONFDIR in +Thanks Vladimir the patch was applied to BTI branch at r +Hi All is for the currency Symbol of Korean the symbol character between ri and icus cldr is different +From what I know about there is no copy to kernel space and the file system buffer of a modern OS is hardly to beat performance wise +Thanks Todd for notifying me +oh sorry here it and the unit test +Patch applied at r +When running my test case extending the test failed in CONNONE jdbcderbywombat by CONNECTION jdbcderbywombat +Looks like Emmanuel wants this issue more than me +Reverting fix causing regressions with remote jmx fix to prevent infinite looping in rpc calls +We can continue to discuss +One more thingas I was going through the example I noticed that we can get rid of and let Springs IOC populate the flowExecutor property of the +This is definetely one of the most awaited features for Spring IDE with upcoming release +implemented activities on top of the jbpm scheduler issues up to jBPM BPEL +Logo is based on idea that logj is a java library which can create log filesas well as log to console +Thanks for taking a look +and you dont even need two nodeshttplocalhostsolrselectqshardslocalhostsolrfls rare it seems to depend on the order of the fl parametershttplocalhostsolrselectqshardslocalhostsolrflidflcatflpriceshows only the idhttplocalhostsolrselectqshardslocalhostsolrflcatflidflpriceshows id and cat andhttplocalhostsolrselectqshardslocalhostsolrflpriceflcatflidshows price and idIm attaching a patch that demonstrates the failure with a test case +Parsing into object would be problematic if the UUID is used as a column nameWhy is that? Does PDO require only primitive or string column names +It wont because you cant do a simple binary search for a range its really a problem of intersection rather than matching and comparators alone dont solve problem of give me all the intersecting ranges for this set without having to compare every range for intersectionNearly every interval intersection algorithm depends on tree traversal and while many of the existing collections are based on binary or trees they dont expose the methods necessary for traversal only the comparator is exposed used to build the tree and the models only expose either iterate over everything or search for the thing I want +It passes on my local box +Can you try copying these tests srcmatrixjavaorghibernatetestannotationscascademulticirclenonjpato the tests and enable auditing and let me know if that reproduces this issueThanks is reproducible in about weeks ago I discovered this bug in my companys architecture and have spent much of that time coming up with a viable repeatable test case + + push to +Is JAXB marshalled + is the original report +xhtml is that because the project does not include that or it is just not implemented yetAnd why is it opening when you select the final result cooking +Attachment has been added with description Patch to clone the SVG document inside +Normally they are just accepted as an array of key value pairs but I think it would look odd in the configuration to have a list of elements under the layout without the containerYou can look at the Appenders plugin to see an example of a Plugin that is nothing but a container of an array +Although unlikely to matter this seemed like the clean thing to do +Thanks Claus! Ive created a new Confluence page for this data format and added a link to the new page in the data format list +provided dependency has already been added to the sample webapp +Commented out the test instead at rev +It seems that youre right theres a NPE when disassociating groupsIll take a deeper look at it +Added patch for the last changes Several functions were moved to header Variable declaration is moved to the place of usage Code format is corrected +Updated to fix the problem +And this JIRA should be moved to YARN +Assigning to Tools component. +Seems that the export code is not compatible with external paging +Its definitely for QE +What should be the expected output if either hostname or rackname is null? We should make sure that Rumens output is consistent and should match the previous version. +Moved issues from to during initial planning process for release +to mark this as INVALID since has been reverted. +The attached patch works just fineI also checked the site generation of it works fine tooSince it is changing not committed directly +Rado also trace logging on for the +Modified implementation so that DRM is now injected into and +It appears ICMP is blocked can you either please allow it or give me access so I can do it myself? CheersTony +I think we need a place to capture usability issues with the site. +Hi RickWe are using tomcat and handled by tomcatOur tests are done from jMeter doing a lot of request to tomcat where each request creates a new user and uses this new user for the operational part of the query to the serverDuring two of these we get the null pointer exception +There needs to be more explanation that this is a generic feature for different types of superusers +Will get to your other next week +Voila just happened again! Cookies were sent +This will happen automatically since IW implements it but I thought to point it out since the change to TPC will also affect the facetsAlso lets not hold up this issue +We are not ignoring case for prefix +The only annoying part is Pig will first complain it cannot get tracking url +make is make to make abundantly clear +your fix does work but I fear it is ignoring a problem about how the project dirsfile dirs are handled which we need to keep an eye updated svn with a fix that tries to solve some of this but we need to see if it is the right fix +I should be back online on +Closed as its not a bug +Reporting corrupt blocks during the upgrade gives us some early warning and the blocks could be deleted right after the upgrade +HelloWould you please try this patch? Thanks a lotBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +README DoneFrom now on then just commit to your release dir. +In addition the JDBC driver needs tosupport JDBC for JDK +Attached patch file for discovery and indexing schedule +I have planned to add example of files compressed with the C version to Lucenes svn repo and add tests that check that our impl generate the same compressed bytes +It was inresponse to a four plus year old ticket that wanted a way to programmaticallyadd parameters to a forward +Closing Issue +Unless its an important exception improvements cant go in after Beta. +Perhaps I will file a feature request on that and leave this bug tracking the problem I just discovered above +Hi MichelActually you can configure not only in for eg +Attachment has been added with description patch file for transition resolvedfixed bugs to closedfixed +Posted this on as wellThis is wrecking havoc with our setup as well +from what i remember we have backwards compatibility issues introduced by the property type of privileges has changed from STRING to NAME between and +When I patch against svn head all unit tests pass +or really apparentlyOK Ill stop correcting myself +bulk move of all unresolved issues from to +Delete unnecessary or incorrrect license or copyright information in the databinding and modules +Viet I might suggest a different approach to saving the information in the database +This patch is for yahoo s and uses an external kinit R to renew the tickets +Im going to take you up on that but probably not until Friday morning +Thank you so much Kathey +Is this possible +Fixed in. +Pls help review +I think HTTP response splitting is not an issue here but the findbugs still report warningsI will suppress it and update the patch +hmm +Need this in into branch +A new method has been added to the class to handle processing of the genFeature during code gen time +The patch is simple enough +Thanks + +fixed some bugs in the reporting framework that might have hidden this bug then +I thought my title would be sufficient Sureshs comment clarifies what I missed +Sorry Felix no hurry +I understand there was a mistake made while introducing these ops to +A test for this is in netui bvt and also has been passing. +V looks good to me +Addresses Aaron and Todds comments + +Im resolving it asan invalid bug notenough information supplied to make a good guess here. +I would also like to modify the first paragraph of this page shown bellowWelcome to Derby! We think youll find Derby to be an easy to use and fully functional RDBMS +trunk Wiki documentation already updated +I have fixed it and it will take now h until Maven Central gets updated again +Brian Can you take a look at this please +I agree that the is an error that should be reported +See also Confluence wiki for Struts on infrastructure list +I dont think smooks feature should be branded since we end up having too many buttons on that about page +Thanks Paul! +This should be fixed now Id like to see it tested on Windows before closing it +Messy combinationandroid already branched committed only on master so fixed in cli no branches committed to master after the tag so fixed in js no branch yet committed only on master so fixed in coho no branches committed to master no tags so fixed in +r should fix this +For more information see the PDF spec section Document Outline +but it doesnt need to since it just tests the behavior when client is closed +Andy lets go with your patch +We need to distinguish between JDK and JDK library class can you go thru one more time to see if this makes it easier for the user per our discussion? If you are sastified with it please go ahead and close it +I am thinking of the following can call submitapplication and send RECOVER event to instead of START +Updated patch making the changes to fix the unit tests +The code being used for resolution and validation is the same for both managed beans and session beans +If you are testing by querying against you may well see invalid responses +I already made the tag for the +Zhihong updated the patch with modification to the test case +patch moves memory barriers from vmcoresrcthreadlinux and vmcoresrcthreadwin to PORTAfter applying the patch these two folders can be removed with their contents +Attaching v patchCan you please explain your point of view? Why do you prefer that is not really a Condition to +as i saw any changes on preference page clear preveiw should press Refresh to see Hide selection bar ger the same me this bug not reproducible for now and looks like its duplicate for +after a quick look at the contents of the zip file it seems that all the interfaces and util classes have been moved to the respective subpackages are there any other substantial changes? Could you outline them to get a better understanding +We will create a new tasklet to fulfil this will discuss it in another scope + can be fixed with part of the supplied you are right +Thanks for the review Jon. +Committed revision Thank god. +Employee can also check status of their training requests assignmentsNeed of Trainings in HR Training and development of professionals fills up the skill gaps and further improve the levels of proficiency +security which extends +It can very easily be replaced with the request handlers support for default request parametersI will commit in a day or two +Class that can be used to simulate the commented error +Updated the description +I have noticed that sometimes the index creation dies on an OOM error +it fixes this one +Anil can you confirm if there are updates need to these +It works on MSQL and +I will post a patch later tonightAttached is a patch that includes a fix plus a test case for this bug +branch is deleted +There will be no need to make an edit anywhereThanksAcco +Alright its past so lets get rollin again +Marking patch available +Using GA I get an error that functions are not supported. +The attached patch implements the desired functionality +Please reopen if you feel it needs further explanationinvestigation. +Its a very convenient way to declare an itemactionbutton because once created contribution item could be used in eclipse menus popup menus toolbars +Thank you Tracy for your good work +This is the possible patch to fix this bug +It might be nice if rrefcreatefunctionstatement and rrefcreateprocedurestatement followed this pattern tooThanks Rick +The workaround I am currently using is to set pushChangesfalse and then manually git push git push tags between the releaseprepare and releaseperform steps +Teiid only supports update queries in the batch mode +there is no explicit support yet +Ben is this something that you guys could look into +Agreed +In JBDS I see this message in consoleEET Eclipse is running in a JRE but a JDK is required Some Maven plugins may not work when importing projects or updating source folders +But thats OK I dont understand lots of thingsBut it looks like Jean and Andrew has taken care of everything to get Halleys translation up and running once the ICLA comes in +So that would be backwards compatibleBut if we go this route Id like to have an update method with the full performance enhancements as well +Hell or high water. +does any one have any objections to it being commitedNOTE there are some negative performance impacts to as a result of this patch +You can then use the SAX stream to drive your own if thats what you want to do. +Took a bit of weaving too to make this work w the security changes +I will have a look to see whether the patch can be applied against SVN trunk first of all +Noel +Exactly +There are currently no restrictions on characters in ACF fields and Id be loathe to start adding them just to support one persons idea of the proper URL over anothersAs for bypassing I dont think thats a good idea either +on patch +to trunk +That will cause epel packages to be not availableHowever is part of repoPuppet does not need to be installed separately though it should not cause an issue +Some updates to insure that negative values are parsed +See for progress on providing in JBDS s CentralUsers who want a newer version of should uninstall the feature from JBDS then install it directly from Springsource. +Back port from patch in +I have committed the changes into trunk with revision and commit comments as follows +the patch and addressed Arpits comment +I think this has to be a wont fix for the reason given it would break usage of the existing password encoder implementations. +Can you try against a snapshot when one is made +Fixed and not a bug +Thats a specific which uses this +Committed trunk too +Can anyone estimate the amount of work +Doug also mentions on that thread that the metrics and fs packages might also be moved to Jakarta Commons +Its possible that theres some logic in the deployment scanner thats making it impossible to pick up the changes after an incremental update +Much of these things are related and could be done in a generic agree with Julien as its the most flexible solution although it may be a bit more more to set up for simple extraction +Do we need fstat and the Stat type at all now? I think not we might as well remove itActually just to keep this closer to the security branch why dont we do that part in a separate JIRA +This issue just needs one line to be changed +Install OS +As discussed the document needs to be updated to reflect the current behavior that only are listed in the list +Thanks very much Claus for looking at this +Trunk has not yet been tested +Closing old issues +Adam want to turn that into a patchI can see how this could be useful when dealing with the Wild Wild and Web +Since it poses no dangers Im including it inThanks Owen! +No new tests added +Working fine with disk closing the issue +Ithink it lacks some and UDDI but it works +The full solution will be based on an interface +Use of the multiplexer is currently optional and not enabled by default while identification of issues is ongoing +For months Ive wanted to change ls to behave just like unix ls but held back for fear of an incompatibility backlash + set SeanCan you try assigning the issue to yourself nowIf it works I think were in good shapeCheersChris +And I guess Yoniks right about keeping the old name even though its a bad oneI dont think its fair to call it a bad one it is just no longer the most appropriate name +Added to jBPM ve tried it but the solution didnt work +My current plan is to make the changes as soon as t is released +Two more things about this patch +Is this ticket really planned for +Moderator has this rights and can transfer it to any thank youCan you please help me out in port settings also +sorry this fell off my radar +Also well need to know what version of Jira you were using +Patch that fixes this issue +Closing for release +Ah thats how you do that +on latest +Its not necessarily means it should be part of YARN framework itselfBased on my understanding I think we should create new project and move forward I am very much willing to contribute to that +What about a core that is down from running +I verified that this works with our most recent EAP build +also it should account for all simple values if we implement this and not only for numbersI will look into it +Marking for +since we already remove other methods from the i guess i would rather remove this as well +This is due to the fact that we buffer all content before sending it over the wire which is necessary for setting the correct header +Can I build this myself +I also noticed that the and fields were incorrectly being sent as firstName and I was able to successfully send invitation requests whether or not the invitations were being set with the field names or the hyphenated field names +Sending tcksrcconfSending tcksrcconfSending Sending srcconfSending srcconfTransmitting file dataCommitted revision . +The test runs inside the debugger +Added some javadoc comments + +Persisting fetched data to disk could unnecessarily put load on the disk +So it will collate if you actually specify collate as an option +hint by user has been proven not very useful +These are all great points Greg +but doing that breaks the API contract and youre back to square one +Fixed author javadocs +Any chance you open a new ticket and attach the exception you see +I can confirm that the test passes on my machine after updating. +Clear patch available while I work on revising the patch +but I would like to see us move away from internal monitoring and making all the data available to a better external monitoring system probably using JMX +The release is in the final stretch of testing and preparation right now so I think we should wait until a maintenance branch is created where we can add this enhancement +Fixed test issues so derbyall now runs with no errors +The patch fixes the test and removes it from exclude lists for Windowslinux iaThe correct behavior is to require a large stack so VM throws according to specThe maximal number of iterations is set to which significantly exceeds the real number for today and guarantees the test passThe test should also check that is not thrown too earlyThe minimal number of iterations in the test was chosen from the current DRLVM behavior and is set to The test passes on Windows and SUSE linux ia +Once the latest patch has been committed this issue can be closed +SELECT should be INSERT INTO +agreed. +Built jar files on trunk and they no longer have the service extension. +Indeed as the Time Machine dashboard is a set of History Table widgets +It works for me + for patch +you actually need two consistency levels here one to do the read the other to do the write +But my instinct is to say no as this issue is about aggregating error information thats currently getting dumped to the log and handling it one just one place +Ok thanks for the feedback +Based on the work thats already gone on in I dont think this bug affects that code anymore for example there is no anymore. +Can you change it to be more intelligent +I have nothing to do with the underlying WSDL generation were calling into the functionality in the runtime to generate the so this has to be in the WSDL generation code +Added an that cleans up the hanging lock +Thanks Owen. +Hi Stanislav Petrlooks good thanks for the fixes your mirrors will be added to thelist in an hour or soRegardsHenk Penning +can we just cache splittableMakes complete sense +Added fix to master and since is now released +I checked in a fix into trunk for under r +Sounds good +comment got lost +I think we need to solve this +Let me do a clean build +Simple doc patch +The upgrade will hopefully also fix the issue with multiple instances some configured for auth and some not which Joe described above +Closed upon release of Hadoop. +I will try the build and let you know. +That code was assuming the plugin jar URLS were file but when running inside the they were jndi +Please try or provide a test project. +Hello AndreasThis issue was actually tested with AXIOM and we are not sure if its going to be reproduced again in because were still migrating and this is one of the patches we need to merge +can you create a failing test case for us? using our latest milestone release? or from your own build of trunk if you can do it +Thanks +In turn the distance map is not filled so you get a null value which is then auto boxed into a double value on sorting +Great! So now I think this issue is completely fixed and should be closed +Bump for review +Justin Ross pushed the patchset upstream which fixes this problem. +Targeting revision +The listener is +Ilya why not add appropriate links back to the forum +one rarely comes up in my searches as my searches typically are bugs I ll be looking at is today and trying to come up with a nice patch for wtp + +The reason was that expressions inside of for loops are aparantly also represented as a type of node in the AST +This patch also makes org +Cannot reproduce this bug in already added a testcase to verify it +To be analyzed +I need to figure out whyI do have one correction on Issue +for author information for dependency on lognet +Since the changes will be difficult to merge and the code is not stable enougth I propose to put it in a sandbox and to perform selected changes on the trunk +but what I am talking about is Two way message routing of asyncronous messages irrespective of the means to establish the connectionWith all due respect I have been prototyping and reading the code +has so many major bugs you cant use any API without your app crashing out +Thanks +Nit Cannot seek before beginning of file block might be better written as Cannot seek to a negative offset given that the outputstream is per file not per block +Hi DanThanks for pointing this out +Converting a namespace from to HA will require a restart +Therefore I think that this functionality is not applicable to majority of Nutch users and I vote on including it in Nutch +removed fixed version as no one has taken ownership to look into it. +But I still cant reproduce it +That makes sense. +I got the idea from the comments now. +were backported to as well adding the latter fix version. +Committed to branch +Builds locally +Hey JustinIm trying to work out where to fit this information into the Server Configuration Guide is this part of the JAXR Configuration +for application to branch +As indicated +Thanks Jitendra for the careful reviewsI have committed this. +The reason for this seems a connection reset error +Bug has been marked as a duplicate of this bug +old date issues. +Vinod Thanks again for looking into thisMy custom AM is not trapping any system signal and I dont have setThe thing is the exact same code works with Ubuntu but not or +Patch updated with the lines to do that too +Thanks again for your patch! +Finally the Bugzilla reports says something about NFS things to test would be using a native partitionand moving the tests to different hardware the usual Isolate and eliminate strategy for locating the cause of a problem +At this level we are working with other advanced developers and can expect them to do some of their own leg work +and if all else is equal same API etc +B should release lock on R and retry later otherwise R would not exist by the time A completes the splitting +We keep the old option around and mark it as deprecated + push to +Please find attached my two attempts to fix this issue +This is the normal process for closing down of mailing lists +Smooks editor is no longer being developed +Although there have been some changes to the core API since was committed this problem still persists in its current form +Fixed +Per PM meeting on closing out will not be considered for +Plese verify +seems we have the same opinion so I wonder what does comment means? The javadoc misleading you or the embedDocument implementation is not right +set affects and fix version to +What do you thinkGiven that you are fixing the native build I assume we would like to enable native compile by default on too after that +Want to commit it to TRUNK then Andrew? to TRUNK +See the linked issue for more details +Thanks for the patch +weil eben noch nicht committed ist bin auch am berlegen was da die beste Vorgehensweise wre +Expected that the file doesnt go to Trash and gets deleted directlyThe skipTrash option will delete the file immediately +we can fix in and +If I create a seam project targeted to AS then fails +I agree the presence of multiple in a html select tag makes it automatically a box in plain HTML +Should work alright now +At least the user will see the results of his manual add +I guess all these should be somewhere in the tarballzipball +This is the fix I need to get this working without browserserver you think a clear call via JMX would be sufficient? is asking for something like that +Now they will need to think where do I get this low score fromSorry but If that is that hard to figure out then I dont see how they have any business writing a Lucene application to begin with +Id like to make some more comparisons on what is being created for sql +Also this was a user requestThanks +This way the controller epoc associated with leader is only updated when its truly needed +All unit tests within the patch now work +There is no longer any +Passed tests + has full trunk LuceneSolr that build and pass all tests +Neither Nick nor I can claim the title of infra volunteer as of yetNick and I have discussed the matter and weve agreed that Ill put myselfforward as the Git volunteer for LucyIs that in conjunction with past service enough to get meinfra suckerHHHHHH volunteer privileges? +After making background it doesnt send anymore and even home page is not loaded +Fixed in the version +Version Final is released +The right fix might be to not call s writeTo in a loop in s append API since the guarantee of a blocking channel would not allow it to return without writing the entire message set or throwing an error +Missed a commit +Attaching demo as a patch +Yes it sounds perfectly reasonable should have been done from the beginning but for some reason the function types on Dialect where only used by HQL not by Criteria if that had been the case the problem wouldnt be here. +The issue is still present +Yes but then they have CORBA that is supported in EAP all and production profiles but not default +Change from Reader to witch is more easy to use and common + +This has always been illegal for comments +I have now split it up into two separate assignments +In theory should never happen as the code should only be executed for numeric and so should never invoke DNSThe intention of reporting the exception was to detect if there is a bug which allows DNS to be unintentionally invokedSo unless you have strong objections I propose to log a warning and return the original host name +It is still has an eary access status +When I went into the code I was figuring it might smartly only apply the increment gap if the offset changed from one value to another but it doesnt even do that and so the current algorithm strikes me as a bit arbitrary +should IMHO not have public visibility itsan implementation detailI will look at the patch once more and come back to you +Including in the same patch because both relate to null handling +I just committed this. +Patches applied with thanks RobI made a minor change to the patch so that the code still works on Mono I used the class instead of to avoid compile errors in and see the TODO notes in those files in case you can think of a cleaner solutionCheers +Still waiting for patch to be committedThanks to trunk. +A big message about it however is far from sneaky. +Thanks this is what I was looking for +Thank you +All tests are passing + +Does it take care of this particular problem tooShould we close this one +Toby I think youve done all this +emmanuel can you verify that this should be fixed now +set fixVersion to +Hi Mikhail The Jitrino fails with any application and even I run nothing such as DRLVMBINjava XX server The DRLVM reports the stack traceI found that the problem is reproduced only when i pass is worked +attaching proposed patch plus unit for providing the patch +As for new tests it would be great to test references to some feeds as well as some tests using the collection. +Will raise as seperate issue. +The fact that cross does not work is a duplicate of +Changes look good. + is resolved +I think we should go ahead and throw it in the mix for beta +The new implementation of should solve this issue test cases anyone +Just compile and run the test code +Assuming included and documenting in beta release notes +Can we mark this issue as resolved? +Ok I will attach an example app tomorrow +That means that we should either include the jeeyul Chrome theme bundles into Target Platform and JBDS or change CSS to use another +Will be available in the next snapshot build which should be available in about minutes +So are these dependencies needed +Only thing that looks a bit weird is throwing wfs specific Hints in the main class might be good to roll out a new hints class specific for wfs etc +I made a start but that was quite a while ago nowHavent had an opportunity to get back into it yet though it is on my list +I committed this patch to +Looks good now. +Thanks +For the moment I think I can just put it in the solrj dir as well +For example another idea Ithought of was to change the Derby code so that the query plans areemitted as XML and then to use XSLT or a similar tool to transform theraw query plan XML data into a more readable HTML web page thatcan be displayed by a browser +Change priority to minor now that there is an interim fix in for the release +The spring issue seems to be the same thing and therefore should also be fixed. +Not sure this test case was valid + works for both trunk and +updated patch of what I committed to trunk that removes this customization of Drools Flow palette on new workspaces +Common four years passed since JPA spec was released and we still have spec behaviour this is a show stopper +adds a session id to tree requestsresponses and should only be applied toShould be ready for review now + +I realize that there is no conditionality here but correspondence in the results is very nice +Obviously moving all of opensymphony isnt the intent but this creates a bunch of extra confusion +update i believe i narrowed the problem the problem is with the fetchselect on element it could happen inside a list and inside set +Thank you for considering this request! I second the original poster of this request An impersonation feature is vital for the maintenance of applications whose users are remote and unsophisticated +I think JDBC connection is not and should not be used like that +Can this be applied and a new release made? Is there something I can help withprovide +The updates have been verified in the ER upgrade in ER build +Thanks +I generated the patch from my clone of the git mirror +Ideally I would have a jar file from the patch that I drop into the solrlib and make the needed changes to the config filesSo Im real interested in getting something stable so im watching the above mentioned linksDavid Smiley would you be able to amend your book Apache Solr ESS which mentions to include how to implement this patch? or do I need to get brave and build the source +Committed revision +Trivial patch attached +Thanks for your patch Robert I have committed it in revision +When a producer sends a message to a queue using stomp and a consumer is not present the message is added to the queue +Only comment I have is the brings in and as compile in JBT +Also I changed to used not +Applied patch and added test case for implementationssvn revision +Thanks Nicholas +Would you please also attach the DDL? I need to this and I want to use your exact use case along with my own +The v patch should address your suggestions +The bug is inThe patch is to check if the total available space is and throw an exception with appropriate message +bulk defer minor issues to +as r +Particularly on step can you define filtering +Thanks Uwe I will incorporate some of this wording into the javadocs and upload a new patchlater if we want we can discuss changing how the nio works perhaps it is best to break the back compat but i am just trying to avoid doing this within the same issue as a performance improvement that is trying to preserve all back compat +In my opinion evaluation from left to right and not being greedy makes sense +Based on the recommendations of the comments this will be closed without a fix +I tested with a Kerberos configuration and I did kinit and kdestroyDid you kdestroykinit as a different user? Did you confirm that the new ticket cache is used after the second kinit +Thanks +Suraj If you have some time available to validate it on your env it would be great +Woops +images would return even is the image was on the server at that specific location. +Not sure who else should look at this but I have run out of things to try +Ok Ill have to check because the last version in cvs before I made the changes was totally broken +Why not simply extend the class? This would remove use of deprecated old Framework completely +This is the patch to allow clients to create a channel for subsystem +I was just trying some things out +Bug reports are not discussion threads. +I will leave this open so someone can address the actual issue and the raid contrib once its working +Robert The permission bits on a symbolic link are redundant +Attaching the patch updated for +Thanks +you can only avoid serialization if both ends are inVM Seems like you are producing to a topic over vm transport but you have more than one subscription over tcp and they are contending for access when they want to serialize +Fixed in the rNext build will not have that sentence +fix confirmed. +Fixed in revisions and +Thanks Ravi and Sreekanth! +Can you please help me understand what patch I should be reviewing +Heres a patch which adds the comment per your suggestionIll commit this momentarily based on your since its just a comment change +Updated patch incorporating feedback +Ive applied a mix of both patch some spices from mePlease check with the trunk version +Thank Mukund! +Many tests failed in Hudson but most of them are related to and but not this issue +Harmony seems to use repository layout +fixedthanksdims +I need this fix for branchx +Updated patch includes both static and methods for loadstore keeps as it isChris are you OK with this version +Ill prepare the patches +JacopoMaybe we could commit the Java version and leave this issue open +This pull request adds Validator support for XML and Java validators +Okay looks like we agree on the code to generate +as said before this should be addressed in the in trunk +Looking at your it would be interesting to figure out what a first fit algorithm would produce +Mass closing all issues that have been in resolved state for more than one month without further comments +broke so now the fix is to disable the test? Things should pass tests before we commit them to trunk +It is in and trunk. +Its stored in +The role does not have to be added to a namenode the node just needs Hadoop configurationAbout the role my idea was to create a cluster with the Mahout jar on tasktracker nodes so you could run a Mahout job from a Java process that has compile dependencies on Mahout without having to build a job jar that contains Mahout and its dependencies +I disagree without the Starting a new period message its very difficult to diagnose the block pool scanner in production scenariosLet me know what production scenario you have in mind +I think I see the problem +I modified the to periodically scan the lru list for expired beans +Ill leave this open until is resolved +eclipseplugins +Please see my comments in +Please dont change EAP AS on this +I dont think that is a good idea +There was some lag between when a consumer got created and when we were ready to start receiving message which caused some messages to get dropped +on the patch from me +camilleflavio I am also a little hesitant to add this to +Its not just an issue of what ends up in the war since we also publish individual artifacts poms +See +Updated Welcome page to link to examples +I think some signatures of methodes must be changed to avoid double code +bulk close of resolved issues. +indicate which partsof the query shouldshould not be flattened and explain why +I think it might be possible to get this working +Id like to understand why the truncation is not working but since we dont actually guarantee correctness in such scenarios I dont think it should block the release +thanks Ivan +from me for v in and trunkThen we file a separate issue for trunk to deal with +The changed will be committed when the code freeze will be stopped +Christoph can you remove the rest of the references to backup and checkpoint node? Ill manually test the patch since its a bit of a pain without JenkinsThanks +Trivial addition of a method +This appears to be related to the new inner class support we introduced in RC +Fixed by commenting the code in and made minor changes to +The patch proposes that if auth is not specified or authnone the transport will default to +That interface implies more functionality that is needed +Downgrading to critical after the discussion with Stack +I tried to change it but Im unable to do it +Closed with the release of. +Can you post the stack trace so I can see exactly what went wrong? If you have a sample app that demonstrates the problem that would be even better +why should plugin load classes from project? what is use case +Ugh last one was wrong patch +I thought this was unassigned +Doc implementation should be a subclass of also. +For trunk it was fixed in rRequire to fix it in Released Branch +heres the patch from Chetans zipped source codeIve put it under integrations +Hi PradeepI applied your patch with revision bcfafaeecddabdaeThanks! +Maybe due to the change to Groovy but seems like an easy enough fix +need for pom +fixed in what was the WSIS implementation? has been fixed +I am canceling this patch because I am not convinced that the current patch does error recovery +Then I removed it +So I think this is actually improving performance because of some other effect like better cache locality by operating in larger chunks +My understanding was that a very large ml would create less and only really really important ones which is what I want +That is already existing code only +Couldnt you just override the checkPath implementation in? It defaults to passing to the lower level but clearly you dont want it to +Added initial approach explaining parent hits +refactory the code as per review comments +Rafi can you confirm this is OK please +lgtm to and trunkThanks for the reviews +Why did you choose to return sometimes but sometimesNo reason at all +It should have been +Yes we need to fix this on Windows too +I can get around the thread limit by specifying Xssk and +Maybe you forgot to activate the maven profile from the command line? Examplemvn +package names and too much other stuff you do at onceI think the can inherit from directly +Regarding the backoff strategy it assumes there is no bug in hbase of course only a temporary problem on the cluster and gives a substantial grace period for recovery by default +This patch made minor change on unit test +There are some better ways of doing this but we should probably wait until the bridge is cleaned upI tested this with a simple guess number portlet which utilized the base renderkit on Pluto +About Given final block not properly paddedI thought this was about session handling or something and ignored it since it seemed not about Myfaces or GAE +snapshot dir +Could we use mockito for this? It should let you do line mock creation + Username jboynes +Yeah I think we should +I have not added test cases for this case +Thanks for fixing! +Would be helpful if you could create a test to demonstrate your usage and show the problem +In fetches still generate loose segments which i just might want to merge for simplicity +These defaults are hard to choose as there will always be someone that the default is not right forI would argue we should only change the default in a major new release and never backport the change to an already released version +Yeah Im in there working on the anonymous checkout anyways Ill see what I can do with this FTL +committed to branch and merged to trunk. +done +I hope to have this fixed later todaythanks revision . +If you find further issues please create a new jira and link it to this one. +Geronimo sample template +I did quick scan of the code and I dont see an in the consumer endpoint +On bucket as host its probably safe to say the s hostname isnt likely to change any time soon +tests passed will commit +Missing the description of the properties in the trunk templateOther than that tested the script on a deployed cluster and able to generate the config with the new values +If I were you Id start with what is there and try to improve it +trunk +apachectl instead of apachectl +The I did was in reference to the previous messageCffile actiontouch is a neat trick +Check out this updated patch file +Great it would be good if we could add an unit test to tests to have an actual unit test running in osgi environment +Rebased to trunk. +This patch adds a unit test for saveNamespace concurrent with edit ops +I the TCP layer instead of waiting like a lemon should timeout when it senses the client has buggered off into outer space +note has a bug in b but its fixed in svncvs +I can test this week +Will look into other issues +I then clicked around + +Copying changes in codebase from to the arow I noticed by nulling it that CMS GC was cleaning items up faster in eden spaceWithout it Full GC kicked in more frequently +Is there anything further to be done with this issue? + +Updated the UDDI integration documentation + +Is that OKI am so sorry but I am not so friendly with License matter +Dialogs for publishing and unpublishing still need to be created +I just committed this +You were so quick in resolving this issue that I didnt have enough time to send a patchJust to do a little I applied the modifications to my setup and the messages now seem to have the correct after sendThanks for the fix +reproduced this in after a rolling restart of the m also seeing this running +Many thanks. +Marking closed. +Committed revision +I ran into the same problem and this seems to fix it +Scheduling to the next release for an for providing the test case +Discussion about removing Extras site from JBDS moved to BYOE questions moved to. +Makes senseLooking forward to the patch +Last patch was wrong +Thanks for providing the patch Kazu +I wonder if anyone is still maintaining it +here is the patch that add the config for the ping timeout +Ill update too +This is really wrapped into the TODO application especially once we get security and IDM integrated +Second the issue is fixed in Axis RC please chcek with that and confirmThanks reopen if the fix given in is broken +As background for the motivation In some production environments we have hundreds of job launches every few mins +Stack Attached the masters log +This is not a Groovy code problem it is a Java code problem +Why is audit log important here? For example rename does not do auditlog when you overwrite the existing file in rename +On second thought Im not sure if Im allowed to simply recommit this file to the apache repository since its not +I am able to reproduce the problem and I am getting closer to a fix +the patch looks good +Where can I find the release? Or how can I fix the issue on an existing installationThanks for your reply +Thanks Haohui and Fengdong! Ive committed this to trunk and. +lets discuss +the file never gets found by maven and +patch to resolve the the wrongly attached patch fileThe patch file was meant for. +I have updated the testcase +Thanks for the heroic effort through all my +I just attached this patch to show the integration of Delete class hierarchy with the existing Regards before creating patches you should always update to SVN head +The attached patch allows as the first letter of a class method or field +Basically if you are removing a leader from the isr then just set the leader field to and remove it from the isr +I reverted your previous patch then applied this one +For users that require encryption in transit certificates are typically well understood and tooling exists to generate and provision +Thanks Virag for the review. +Odd +Updated to and upgraded to critical +As agreed well keep this item opened till the official EAP packages becomes available +This was fixed a while back. +Ben please hold off on the commit +I think the patch is worthwhile +Why not include zest in in first place to avoid this cycle between two projectsAnother note s looks like wrong name for me since there is only one target platform for inside it +Else if this is only unit test code could consider applying before the RC as all I have attached a patch for Test Cases please apply itThanks Lots Indika +ErwanConsider it as asked Thanks +Second version of which supersedes the previous mif module to trunk removed the old mapinfo module. +In particular you mention you are using so seeing how you set that up would be very helpfulIn all cases our code should translate all Xerces instances into +Both should have stabilzed after so is a more realistic as out of date because of month no activity +Thanks for that excellent suggestion Rick! After all this book is the main place where the syntax is used +Looks great + has this does not have this bug +Committing this update to the unit test +That is it is incorrect for the actual record that was seems like somehow one entry in uniqueNames in gets lost +Logged In YES useridI have created a graphicalcommand line installer with that does this exact comes with the CD Subscription though +This module was written in +is abuggy script that should be fixed nowSent from my phone please pardon the typos and brevity +updating issue summary and description to be more explicit about what is move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Up to now this module has a known shotcomingThe table auto creation does not work + +Im asserting that the effective pom is closer than the lifecycle version definition as it applies to this issue +Attached is the same patch except applied to the source for those who are using this and seeing the deadlockThanksAmit +Bulk change issues closed by JEXL. +Also perhaps you can try with the branch to see if the problem still exists +Hmmm I had trouble finding beforeBut it looks like it was used mainly for adding a timeout +Hey Brian Thanks for sorting this out +I will try to get that changed out and will comment back once I test with +Will commit it soon +IMO it would make the code more clear removing themI agree leaving the other methods untouched because they dont hurt or confuse in any way and as you point out might help in future +I think it would be possible to seperate it into several pages. +If youre happy with it working fine would it be OK to commitmerge itAs a cleaner workaround to the dumpload solution I previously used its possible to use filtered replication with a filter checking for! this creates a fresh replica database with a docdelcount on and +the imported code looks untested for simple constellations and there is only a discussticket which is open +Could you do thisthanks all open issues to +I checked again the solution proposed and now Im sure this is the way to go +Committed +Not that I really know what I have done but Ive committed your patchIf there is something broken in the maven build say for the next one and a half Ill ask you to fix it Ill promise Thanks for the patch +SET replication trigger was deprecated as of. +I will attach help update as a patch tonight +done at r by Hans +Attached is a text file showing the files I have under the Apache folder and their hierarchyGathered usingDir s x b Best RegardsEyal SafranProject Manager ltd +Root is getting created from a Length hlog +Autowired and Inject allow for this behavior for components but thats as far as well go. +The attachement contains a test that I used to test it +Clear the codes modified by can you take a look at this +I also updated the a bit and included the script that I used to analyze the job history +QA verified +No itd for services are created +Attached YCCK test stuff thanks Damjan! +patch committed in trunk rev +Hi Unicothanks for reporting this +Probably this is not an error just wrong naming +Closed as a duplicate of +Ill follow this up. +In the meantime can you provide a basic test case or some sample inside the CXF distribution that I can test against the CXF fixId like to know whether this is related to or notCheers +Attached the patch for this issue I did test this patch but it aint easy to test it as it requires the installation of a fake security provider and therefore no test is included +Usage has been removed from the broker +as a result of this is now fairly trivial +Layout api completely refactored from the ground upCreated a model object that handles the layout and movement and orginaztion +We just installed beta and polygon boundaries that intersect with tile boundaries are not being rendered +So this looks like it may be an ant bug and not a xalan bug +commit directoryRegards the should just be replaced with a of the path of the directory that is trying to be imported +I missed the update on this feature. +Can you point me where the jaxb spec says? Should the in setter be used in unmarshal phrase and in getter should be mapped in marshal phrase +Is it possible for you to attach a short test configuration file which needs aspecial encoding? This would help us to verify if a fix worksThanks +has been marked as a duplicate of this bug +Its a mozilla bug +Close Doc issues in release versions. +did what I said not what I meant +Justin you did work on parsing surfaces and curves that happens to be made of straight componetns lately no? How is this one affected +Closing resolved issues. +Attched the patch the patch. +Fix still needs to backported to to should be in the next nightly +Is there anything left on this issue +will take care of it +No changes were needed for hadoop test cases +TPNOBLOCK on a tpcall may raise TPETIME as TPNOBLOCK only applies to the send portion of the call I think +It appears it has been fixed. +Ive just upgraded the Spring distribution to Struts and included such a +I browsed through the code and saw that for the Axiom message the are added +Can you show a concrete case so I can reproduce the problem +I dont think any collection is supported in M +Closing this as review has been completed. +try to reopen and close it again +This issue has been fixed and released as part of release +Im set now +When user navigates through folder using CD and encounters a project folder lets trigger project command so it automatically displays info +bq +This is a blocker issue I cannot find any acceptable workaround to deploy modules into right places AND to generate right links in the parent site +Moving to +Thanks Raphael for spotting this +It works with some classes but seems to die processing other classes withException in thread main Illegal constant pool index for method name in class file Probably unrelated to this bug though as I have not yet got to putting annotations on that class +tests for mixed newlines expectationsNo need to have this in the repository while we have no agreement on how line delimiters should be treated +it seems like will work fine but I haven confirm that +As far as I can tell Objectfbea is the BEANINFOCACHE used by the Introspector internally to sync the cache access +Please apply file +Mithun Uma +Fixed on revision . +This allow the queing of all actions without loss of time. +is there some way i could help you with this issue by trying something out +of course it is check again. +is long out and is resolved so this ought to be as well. +Per PM triage meeting will not fix +If this Class object represents a primitive type this method returns true if the specified Class parameter is exactly this Class object otherwise it returns false +Not sure if you are waiting on me to resolve the issue or is there a process you follow. +Also a few In messages were no longer needed and were removed and the builtin node types from were removed +For wide row table type we could actually insert values in a second which seems not acceptable +The configurator is a way around all the blasted example files +The syslog does not have any error reportsI think it is possible that the problem is due to a software bug +Maybe we should add appends in the branch since it is the one that is supposed to provide more reliability +Ok so I have updated to flatten the svn commit messages into the single list +Validator registers local versions of Commons Validator DTD as of Validator nightly build . +push out to was fixed back at JPL and i just never closed it +Reopen to backport to so that can be more easily backported. +We are good to go again +drools esb jbpm seam and eventually resteasy the latter two actually existing outside the directory +AMQ runs fine with. +These Policy settings do also go into the and into the One example of a is also attached +I committed the patch +I confirmed this is working w r. +Thanks +This would only update the server version nothing more +Thanks G. +CATALINAHOMEbin is not in the PATH +for the patch + +Agree! +committed +sample installed but not started +Thanks! +index analysis would smart detect whether this input was already parsed or notWell this should be possible with the pluggable parsers that this field supports +If this is actually causing an error please the bug. +Stale. +The longer these kinds of bugs stay in there the harder it may become to deploy fixes once deployed code relies on them + +Thanks KevinI plan to apply patch in two steps first only apply the fix the next format code in +Patch and explanation en route +The application must check the zoom level +May you pls upload the file you use so that I can investigate it further? Thanks a lot +A new JIRA will be opened if new content is required to update the section. +Thanks Dart +Les opinions contenues dans ce courriel et ses ventuelles pices jointes sont celles de lmetteur +and doesnt make interruption mare than ms it wiil be work but if user make some rest in typing for example of ms the update job will starts and he should wait second to enter a next group of symbols and wait wait again +Need a setup to investigate this issue +There is currently no predefined condition for OSX +Commited to CVS +In this particular case maven was run with the rf option mvn clean install rfMaven version JDK version x or x on OSX +Source code for test +Im on the patchStack Wanna have a look too +This will be fixed by the patch to +Thanks Kim +This should only affect native builds with visual studio and not mingw cross compilesThis sucks for people using the native compilers but they are probably used to pain by now. +Committed branch and trunk. +People have been using this for years with no general problem at all +So close this JIRAWe can create another JIRA for the GEP updating problem if need. +I hope I did it in right wayAlso thanks Arun for reporting as you can see I have committed a different fix as suggested by Adrian in trunk rev +Closing the issue despite is not create JIRA with Naming. + to trunk +Describing point location of and +This is no more there in the current derby trunk and has been replaced by +This gives the caller of the client the chance to their ZK endpoint list and attempt to recover their session on a new set of hosts +Sample option +By default the extract action will be take effect and a context value could be configured in the to turn off this functione +But if you want to use the code from github the branch is where the changes are the master branch doesnt have them just tried switching back to searchCriteria and it worked fine as well +Diff of for the inconvenience this is a temporary situation caused by releases being in progressWere currently in the process of releasing components +Olivier this is ok please close it +Using heartbeat was also my first approach but it just is not as performant to send a newline every ms just to keep a while loop running +fixed in modulessrcjavaorgopenejbdeploymentM rev M rev seems to be similar to someone should look at what was done for that when looking at this one +Thansk here is a patch for the ajax submit form update area functionChgeers Jacques jep there can be direct calls from ftls which i have missed +Not sure +Committed to trunk +Is there a similar log like the one attached that we can get from a successful run +the given issue has been i have a problem with username +In +this patch is not think datanode need store active NN and standby NN you thinkt the client get the response then go back to the new to get the new block key? ie you might see this warning in the logs somewhere but the client should not actually fail to read the block +Performed a regression test on the news dataset the output produced by the original code and the output produced by the new code is now identical +Fixed in the trunk under r. +Please do also run the test suite on the latest patch before committing +Here is a very trivial patch to apply to jbossas that marks the weld servlet as async supported +attaching two log files with hibernate debug enabled for AllI have an even easier way of +It looks like the patches are needed so that the jsl files use the value at least +Patch applied many thanks! +Committed at subversion revision +This bug is very nonspecific and there are other bugs tracking the missing behaviors in spawn exec and popen +I think we should continue to use SNAPSHOT for the trunk and branch to a JDO release and change the version in the branch to +Of course I meant +There was some recent discussion within osgi of defining a standard URL handler to allow a directory of files to be loaded as a bundle +removed hard coded group in a few test patch actually does not depend upon the the latest patch in though that patch came after this patch +rand the c and java tests +For context we ran into this on our CI server where core was not included as a test runtime dependency +Not sure why he did this +I realized that this may not be an issue for people who arent using Struts model driven concept or arent managing objects in session memory +I deleted several applications here and it never happened to me +The patch triggers the miscreant behaviour by making a WFS request first before a WFS request +Amomg other characters no dashes are allowed +Please close this issue as duplicate +ecedcbcbaaefcacfdcb +I know it is taking a long time sorry for that +Cookies are being set with quoted Paths +Committed to trunk and. +Testing through hudson +Also please start with a fresh install of STS and try again +Thanks a lot for this enhancement please verify this issue is fixed as you expectedBest regardsRichard +All tests ran cleanly +One typo set setNice catch! will fix it in the next patch +Sorry for uploading so many patches but I currently dont have the possibility to run the test suite elsewhere +In I had removed the documentation on and ulimits specifically +Im overloaded at the moment +The issue appears to have been resolved but not marked as such. +The cause of the issue is a wrong assumption in the SO optimization algorithm in JIT +Done +I accidentally created two JIRA issues while JIRA was on the fritz +Alan Im presuming this is a problem in +And Tools does not work with Connection Profiles +Updated the and showcase readme files according to feedback +Resolving for Daryoush because he doesnt have full Jira rights yet +OK shouldnt have any package or import statements in DT that is uploaded however I now strip out the package statement but leave the imports in if they are found I think that should work fine please confirm +Im not sure the performance gain justifies our shortcut anymore with modern computers +Making the change to this API was a hard decision since although it is not deemed to be public it is being used by a number of rd parties +The pdf is valid on linux and invalid on windows +I committed patch with revision Myrna I dont have any more work left on this now. +this is jpa project so Console Configuration should look like are not necessary. +Generating these arent wrong if the user is targeting the old jboss portlet is there a place where the user chooses the targeted portlet versionsystem +committed to trunk and backported to close after release of +Why use docblock comments in JavaI just borrowed the comment block from Hadoop dont know why thats the standard but thats what does in my emacsbq +This patch is for the trunk +Looks like it needs some more work moving to move opened for the trunk port to eliminate confusing in JIRA the work has been done for a while. +Attaching another patch +Removed the invocation context from the response context if present to protect against this kind of issueAdditionally removed the code for logging the response context to avoid revealing potentially sensitive information. +Created for that change will upload a patch shortly +Those after are fine from this perspective +In part it is a naming convention in part it is a contract Pre hooks allow introduction of preprocessing and also important override of default behavior and associated of base processing or any additional coprocessors +Can you create a testcaseI just tested and it does look like the clientCreated is being called correctly when the bean is created +svn commit r in strutsstruts branchesSTRUTSBRANCHcore trunk good to me. +finderName find finder class +On the con side it involves reopening a log for append which isnt an operation we currently do anywhere elseMind if we open this as another JIRA since it doesnt currently cause any issues that I can think of +Yes if you assume that record readers have the necessary logic to be able to skip bad records on the very first try it would work + +Bulk close for Solr +Must be related to user environment created a new user and cannot reproduce the error for the new user. +There is already a unit test that tests append +Hi hairong sorry for not getting back to you earlier +A project group admin should not be allowed to change it +When are you fix this bug? Thanks +New patch I added also the support for the qt parameter +ThanksCommitted to trunk svn txn +That analogy is forced +Why do we have so many for Hibernate? Forge currently uses the one from teh EE BOM +I believe we can make this changes to resolvethis issue +Sorry I didnt get to a review +To fix this bug copy the parse method from the to the +Thanks Viola +I have attached a sample application that demonstrates that using custom implementation from crypto and the legacy package works +If you want to pick it up again sometime later you can always reopen it again. +This will have to wait till +If the classes have been then the dependencies added by the enhancer need to be added to the bundle manifest +The specific problem I am having is testing the recovery of messages +I agree with Todds comments +but there could be some missing changes to that class +Commit the patch in revision +This is the table with no BLOB columnupdate t set status where id But the following update will fail with OOM +We checked against with the same result as aboveThe same thing happens if we may Versioned an Entity and use a collection rather than a do you think youll have time +I just committed this +Verified on EAP on EAP using +Marking LATER pending course of action +The tests depend on the v of the patch attached to which addresses a few bugs that were found to be existing in thecurrent Clob locator implementation +There are more ways to create a build file for Unix platforms which were discussed in the past days on the mailing list but this simple Makefile has some advantages over the other methods most importantly its portability + +Is there a new version of bnd available? If so when could this be available in the +I have inserted the first stub of classes and interfaces to resolve the implementation should be completed for the TCP partThere are many parts of the code that need to be discussed thats just a stub +It would be really nice is xstream could be set to easily ignore fields like this +Another way of solving this particular issue would be if segment readers knew which reader they came from +this should no longer impact the stuff but Ill verify that the other bit has been fixed +Not a current CTS issue but likely to have been one +in JBT +Thank you Etienne and Daniel! +I wonder do you want to change the max in the after the repository and workspace was created? It seems the file is stored in the database in your case so I guess you would have to change it in the database as well +bulk defer of features from to +Adding to we need to verfiy that this works or clsoe this unassigned issues over to the pool where they can be worked on for the in all probability patch +It seems to work +The James Old IO server boss thread is continually locking up due to logj +The xpath isnt finding any nodes +Ill check with Remy and see what he thinks +Not sure what happen the first timeI added the woden group ID to the WS target +How exactly were those logs generated from an embedded Maven like IDE? This doesnt seem to be a plain Maven CLI invocation +After talk with Tim I think we will refactor current code as follows Make all v plugins independent +I dont see this as violating anything +Go for it! Oversight is all it was +This was reviewed by Vijaya Mamidi +This should be checked in together with +Should allow that for the other contentsources as well +said doesnt throw exception at all +Revision fixed in the. +Still it is now the file generator and not jx +Is it a regression? I would be surprised if that bug was there for long time +But it would be nice if you can confirm that your stuff is getting properly into the sources once target folder is set properly +bulk defer to updated patch with changes to. +Why? Is it ok +We should change the demo app to keep the query results in the session context +Keith whats the best way to obtain the for the parent of a? If theres an easy answer I think I can provide a patch within a fewe minutes +This properties file is processed by maven when building and I can then read the version number from the properties +Isnt this one outdated or else done +avail metric control implemented in the build +That seems fine +Instead I added a TODO to highlight it needing a decision later. +Potential patch to read the xml directly not using +Closing + for the to trunk. +No problem looking into code its very interesting work +The fix includes followingThe secure conversation versions are selected based on the security policy version + removes modifies and uses Apache Felix to automate manifest information in the JDO API jar +bq +Committed to trunk +I dont know if either of these has been written up into a ticket yetSo is necessary but not sufficient for doing what you are look for +Therefore I wanted to confirm this with you before deleting anything out of the JBDS SVN tree which you might be using for the Product build +What would people think about using a system procedure instead of a property +Marking as duplicate of and closing +Bulk move issues to and +did you tried to measure standard compaction strategy to see if MB LCS brings improvementsDid you look at the graphs? I dont see how a control of STCS would tell us anything new. +Hi KatheyI think I can complete by thenthank for assigned normal urgency and noted that a repro is attached +Verified and closing issue. +requirement is that we dont log any actual edits to the new edit log until its been recorded in all of the storage directories +These are the changes to the Developers Guide +User Guide section Should there not be a reference to the Installation Guide for those getting started with a fresh Roller server +I stand corrected I wasnt able to duplicate it off trunk either +SSVM comes up successfully and centos default template is downloaded to secondary storage during ssvm template sync +Whats wrong with a internal variable with default of k? How do we know k is the best for all platforms though its a pretty good value +Sorry I have not had time to help on this one +Canceling to sort out comments +I have created the static instance for +Use the to get things like the Eiffel Tower +the null table is a problem in the subquery but these are not extent awareSubqueries are not extent aware +i added steps on website +Patch applied with thanks. +Mentors can do this so I did it +Ok this is partially resolved +This is the place where all the features should be documentedLater someone could add examples to wikis or to the showcase demo web appBut the javadoc should IMHO always be updated with all the features WWXW have. +Lars Yes +These changes make work on Windows +We would need to preserve messages and some other things but nothing to difficult to preserve +There is now sources JAR files +Will do +Thanks Isuru for the contribution. +Patch to add support to to handle JAXB objects +I think this should come post it shouldnt block etc +Ive reproduced something similar wo external file fieldIt looks like if you are aliasing to a field that exists in the schema its picking up multiValued from thereFor example flabcid correctly yields single values while flcatid yields the value in an array +It is possible that the problem has been caused by me as I had to switch between a tag and head revision to be able to apply the patch at all +Let me know which of these you feel is the most appropriate location and Ill remove the otherI also added the same note to the Known Issues section of the Release Notes +HiAndrea thanks for your answer +What I am initially looking for is a set of ws deployments that cover common ws processing use cases +now has a function to report an error based on a file path +To JeffCould you try it again? maybe I have resolved the problems. +So no need to worry +The fix is adding the classpathdep extension point and updating +Jar documentation spawn of to separate issue Released Closing +An alternative approach would indeed be to rollback any transaction which is marked for rollback before handling the exception itself but this is a major resdesign of exception a workaround I suggest you check if the transaction is makred rollback in your exceptions component and if it is do the rollback and start a new transaction +There is no reason to expect that Sun will timely fix this defect in the relevant Java platforms that JDO implementations support +I reviewed the treat function and do not see the connection in structure to the cast functionality +Im waiting for community feedback +these usecases are deferred to the Beta is because these usecases require adding new functionality to the module layer +Sorry I was trying to get to the wiki but its been a busy week +Did a pass through the website and came up wthese edits +It may take us a week or two to get it all tied up but shouldnt be a problemJon +Turns out all the were defaulting to STRUCTURAL +Looks like this is a windows platform specific issue +To deserialize an enum constant reads the constant name from the stream the deserialized constant is then obtained by calling the method passing the constants enum type along with the received constant name as argumentsall enum types have a fixed serial of L +as far as I remember you have a bridge between the servers for the Output QueueSo this is a rejection +Sorry about that I committed broken test +the patch looks good to me +I am not even sure what this test is supposed to verify +As per my previous comment the spec does cover events for these two methods. +The performance degradation is caused by the change on the use asyncSend flag in the from true to false +Apologies for the inconvenience +Is there any planned dead line for a new FOP release +ready now +Right patch this time +I tried having it check for and bailing out but that didnt seem to fix it +I dont know too much this OS for making ports anyway thanks for your offerWill compilate from source +It looks like Jenkins is having some issues with where the files are so dont worry about the Jenkins resultsHave you brought this up on a cluster? Just as a sanity test that the slfj changes are not hurting anything +Hi guyshere are a few more issues we only have migrated the But we missed the andThey are all similar to our POC Tree and i hope that is not to much work to migrate them i asked Ankit if he can support me with a few lines of codeCheers folders above have been deleted at revision +I dont want to move this out any further +If the spec says that Native is right we should challenge Metro +When the persistence adapter starts then it starts the broker pluigns and transportnetwork connectors +Sample xml file with queues issue was found in CR build of EAP no when you undeploy a JMS Queue it will only remove the JMS Definition +HiThank you for your checkingI think my patch will be committed if the modification on can be acceptedRegards +New benchmarks have shown these changes to cause lock contention +on patch +Once final is released we will pick up the corresponding fix. +Hence the already locked storage error was correct behavior +No +for all tests for the patch locallyI just committed this to. +With rev Im seeing some app client stuff work in the tck +The main problem is that are being driven by transitions rather than transitionsThis also turns out to be the cause of +Actually I dont think we any other choice to move to a new syntax +I believe it is very easy +I want the classpath to be stable no matter what happens it should not be the user wants to refresh his jars then he should explicitly ask for that IMO +Checked into develop confirme +Added to the Fixed Issues Hibernate Group in Release Notes CP +SiI am attaching a basic patch for you to review +HiJaxb only support List as collections so use List insteadOr you can use to adapter the data type during marshalunmarshal take a look at javafirstjaxws example shipped with kit where we demonstrate how to use +This bug is closed by revision . +So until standard shell tools support access to GB zip archives I see little motivation for Hadoop to support this +This renders unusable for me +For some reason the files hadnt be attached ve attached them removed the in xsds and it works +I suggest you change the host name to a standardformat or use the IP address insteadThe question is if our URI parser is too strict +I notice that the fix is commented out +The index connector has been removed. +TijsThis is a valid defect +licenseFixed +Niall I think that method should pass the exception back and I cant rely on the method returning null to determine an error +In short this fix really covers a small part of the issues you will experience if you set the to true specifically just one you found in a very specific scenario in general I would discourage to use the flag in production unless you really know what will go into the session +Passed please see for details +Im not mad on BOUNDLESS how about FOREVER? Or could we use camel case rather than pure caps for these constants +Version updates from comments +Sorry the previous comment was meant to go onto but somehow ended up on this one instead +Looks fine! thank You +Committed to trunk with revision Awaiting results from automatic tests before closing +Sample app with test to test this out +AFAIK this removes all unnecessary imports that are not in generated files +Here is a quick patch that fixes ofbizapplicationsorderwebappordermgractionsorder how does one produce this error you are reporting? Can you give a more detailed example +it is true it cannot be converted into valid but its perfectly reasonable to use code units for internal processing like this I am not attempting to write this data into the index or anythingI think data from for merging or writing to is completely different from data being used to searchI know you want an elegant encapsulation of both but I think its a broken designI dont just make this up to be annoying i have applications that will break because of this +When I teach courses people are always confused by shard replica slice replication factor leader master slaveSo perhaps we should formally say Shard replica when we talk about replicasslaves +Luke you want to take quick look? I will commit this by the end of the day +great +Client applications can continue to stream their writes when there is a machine failure +what about add a to do centralized management of the counters +Test failure seems to be due toNow its revertedAshish could you please reattach the patch for clean QA report +Does this bug still exist? Please react for now postponing to version subtasks revision . +To be able to resolve an individual artifact from the command line without first creating a dummy pom +Or maybe a combination of exclusion and inclusion patterns will do the trick better than the somehow stack of +The only real changes I made was to add translation routines for the operators I added +I limited the trimming to null bytes in revision +Attachment is functionally the same as Attachment +We are going to disable it by default in the next release +I meant to say as an example of cut and paste code that no longer works +Juergen +Thanks a lot for the explanation Karthik +I do not know its location in ok got it Thanks +Reassigning to Kurt to handle the project seem to be ok here now with these changes +Unless there are specific problems that where not addressed within the original JIRA I see no reason for keeping them in this state when most of them where fixed and resolved almost a year ago. +Lowering severity as this only happens with custom providers +bq +The other change is that it does not use the FOR UPDATE syntax that is not available for all dbmsI have not included any tests for this new feature but have tested this successfully using the following dbms +jar. +The parsing is part of the marshalling layer +size splitTypo above should be +Then I think this patch should do what it is intended to do +Im concerned about possibly leaking sensitive information via logging of the attribute value or reporting it on an error page +I didnt quite understand how to reproduce a crash +The solution seems to require gwt which is used in trunk now +Most likely related to we should fix this asap or if there is one +if we can get it built with VS and both then it will be ideal +I am also looking for a way to use Kafka client +Yes you can close this one +attach patches for changing pdkey and not changing it +Now the report indeed is generated with the first run but it doesnt show all classes and those which are shown got all a of which is issue was solved in version please try that version +However the beauty of it is that can simply move those annotation classes to core whenever we see fit without breaking binary compatibility with previous versions of +I just committed this +I have added an extra warning to that section of the book to clearly state that you cannot have different databases for different components. +In the screenshot editor is open only to display the identity of the tree we already have this feature for projects that have jsfnature and strutsnature +Please set Affects Version +Patch looks reasonable to me +Have a patch ready +What would it mean to make YARN framework agnosticCan we work on a proposal for a set of conventions for how a single hadoop component lays out its parts and how it wires in other components? Something like Alejandros old tools proposal would make a lot more sense than just exposing a mass of env varsAgain this would be easier to understand if motivated by someReal world examples of how users lives would be made easier by thisE typing on glass +after this patch goes to trunk it would be great that any new patches start to use the trait +This way we dont have to wait for the delete to complete and we eliminate the exists call +I commit this tonight in h if no issues arise +I dont believe there is currently an expectation that the admin interface is supposed to be forwardbackwards compatible +Ill file a new JIRA for it I feel like its out of scope for this one +Added docs for pure Java documentation +This is already committed to trunk. +Extract the system time after extracting the last from the datanode descriptor +Good news thank you for sharing +I recommend mavenIf this problem only happens with please this bug and post the output of the mvn command with the e and X flags +In this preprocessing step the sources are copied to a temp directory rooted at tempsource and COPYRIGHTYEAR literals are replaced with the value ofTouches the following files MMakes svn ignore the new tempsource directoryM Adds a new property initially set to be M srctoolsM srctoolsM srctuningM srctuningM srcadminguideM srcadminguideM srcdevguideM srcdevguideM srcrefM srcrefM srcgetstartM srcgetstartReplaces the copyright with the COPYRIGHTYEAR literalM Adds the new preprocessing step +Yes +It is causing a tremendous amount of distraction when we are right in the middle of getting out the next release +Dims can you apply this to the branch in svn revision +Maybe you can review the tests to see how they might differ from your scenariosIn the mean time I will continue to investigate as well +I see thats certainly more flexible +Please reopen if feel it is en open issue and you have more is so weird since the problem still persists +Throw one more rock into an already muddy water Please see for some more input regarding the schema namespaceApplied the patch +q file will remove in the next patch +Brock what version of are you using? And can you give any more detail on what operation is being performed when this happens +Ive tested with my project and the case sensitivity issues seems to be resolved +Oh noticed it was opened against +Can you check please what happened? looks fine for meso close it +Attached file has an eclipse project with a test that demonstrates the problem +Upgrade and installation instructions seems to be fine for RHEL platformsI found just a small typo for which I filed a new jira seeSo this jira can be closed once the installation instructions for MS Windows and Solaris will be checked +ohh! i forgot to diff the wagon tree +bq +bq +Attached a patch which implements the suggested applied in r +The SVN needs something similar to the CVS +Ted Im trying to setup a test for this can I assign to meMike if you are still around any chance you can put up your input generation code at a minimum +This is a serious bug and should have a higher priority +This is the main point of us using live code examplesIf we really want to save some time we could set it so the sphinx docs only are build when mvn install Pdocs is called + +Should fix the issue +Fixed on branch revision . +This removes the python version check from ant +related to workaround works fine for me +Throttler is still public and we should make it package private +Committed revision for because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Hitting return should create a new paragraph +parameter to choose whether search is via queries or doc id sets +properties files to at build time so we only have to maintain one set of English filesregards did a test and we can also fix this issue by having empty files +And then instead oftrue So that if someone wants to add another encoder it is trivial to do soAt some point we probably should make the more complete and promote it into but clearly that can happen in a different patchIm not that convinced that people will use the typed bytes format outside of python where the library is already present but it does give them an option which is currently not there +If you start it from console you should see in console before installer exit line likecreating Logfile in tmpcould you attach your log to this issue +Attached is a patch wil the latest changes to all Numeric classes Formatting fixes for int for long +Yes the documentation the FAQ seem like a good place +Hi Martin can you please provide us with the steps that youve done to reproduce thisAlso do you mind doing rpm qf and rpm whatprovides on one of the +for the patch +Out of the box Hadoop could bundle or implementations +info tested on Windowsi need some test code to replicate the problemthanksPlease take a lookthanks to rudolf Is the soap envelope canonicalized before signing? or is it implicitI think Axis is doing right thing by removing superfluous namespace declarations which should NOT impact verification of signatures because the canonical form is preserved +Thanks Jon too. +Make that users! We strongly rely on Maven generating our +Do we not have test cases for that? I guess thats something else to add +Move fix version to in preparation for release. +Norwegian mismatches usingThats something like +Patch fixing issue +New version is up mirrors should sync soon +In the meantime Zhong should be able to just use a larger stack as a workaround +listShuttles have items of a bulk update all resolutions changed to done please review history to original resolution type +the behavior re is correct +but sync thread flush interval depends on users setting and also it was a shared disk with read requests +This patch should correct the problems noted by applied thanks Rico +Mark can this be resolved as a duplicate of +Hey Mark I just looked at the patch and it looks like youve removed the updatecreateshard +At verify when the doc is added that indeed only one value was suppliedIn Lucy I expect to fill this role +Please reopen if you have a test case reproducing the problem +To use currently you have to set up a file which excludes from the list of implicit dependencies +Would you like to create a patch to fix the problem? It should be pretty straight forward +Thanks Zheng +Support for discovery agents has been added on trunk and will appear in +I was recently thinking about this and was wondering if tries to parallellize view indexing to different CPUs at all +Triaged for checking repro attached and setting normal urgency +detects Text and writes the bytes directly to save encodingdecoding time +Looks good. +In its part of the core +It also fixes the bug in that fails with WHERE ds +The diff is not between wiki and the version in SVN but between trunk and the branch in SVN +Attaching screen shots of the rather than make comments it was easier to edited the portions that I preferred to change +This patch looks ok to me +Commented on the diff mostly around documentation +I wanted to see what other databases did so I tested it on instances of and Oracle that I had access to +Ok thanks for clarification +I have discussed this also with Pavol it was not occurring on his computer +Want to make sure if it is ok with you +This is an issue with our new SOAP stack that is available in only +You are right olga +Makes sense +send replication requirements automatically without manual intervention +committed +Finished with SBT migration to version +My source tree is dirty now so I am unable to generate a new patch +The error is occurring on the server the rename of the uploaded file failed did the file exist already? You should look in the server logs there is nothing we can do on the client to resolve server can avoid the rename by setting to false but the receiving application has to have the smarts to not process a file thats why we create it with a temporary name first and then rename it by default +I think for the build number it may be good enough to just do manual testing and have the test check the output format and not the number directly +As I understand the code Bouncy Castle is a purely internal dependency no class from Bouncy Class leaks in the public API of +Ive added a unit testcase for this and it now works when the response is compressed and the content size is +Hi Ray Im not sure why this got changed when I fixed +zyro the AMQP issue you raise is similar but needs a different solution because the retry interceptor is low down in the message listener container and outside of SIWe are looking at a number of solutions for that please see work guys + +I think theres still room for improvement here but this resolves the immediate issue +moving to M +The reason being that if you want to avoid large messages it should compress them and try to avoid cant see a situation where you would want to set but if the large message could not be avoided send it uncompressed +Committed the patch to. +There is only to point that is puzzling mefirst I can transfer the to srcmainapplication that is no problem problem side is from maven +Contains the java class files +bq +Shuffle Error Exceeded MAXFAILEDUNIQUEFETCHES bailing about we create a specific feature req create a stable wire format for exceptions to discuss what we and where it woudl be to are you trying to read the results in and create a new exception at the far end? Does it get thrown? Ive started a basic writable exception but not looked at the to an exception +That should eliminate the need for the local variable verified +I plan to commit in a day or two +Thanks +Id say lets not be too hasty about closing this sort of issue just because it hasnt made it in yet +Ive notified that team +It would be better if you could submit it in review boardThanks we create a JIRA for each monitoring plugin and make this one as Task? We can link all the to this JIRA +I dont thing the reader tests do use the master slave directory providers +We could write a custom mini serializer not dependent on Xerces +The zip is a svn export from this SVN URL +We could check different parser implementations defined in to solve that. +Ill see what I can do for CR +I dont find it in apis +This is essentially the issue as. +Jackson is available +The API needs to be opened to developers too +This one will also be removed +The idea is create an interface that allows to saverestore transient properties like to be saved per row +Thanks Jarkko +Updated related identifiers in you. +Setting fix version to next and releases for further study +Actually after peeping at I realize that in a Mojo the should already have access to the plugins classpath but due to the way the classloader is constructed it does not +Can we modify this? Sven +cool a lot for the reviews Todd +Compile and test work without anything in the local repo or with +Strange i pulled the diff from the same export as the previous one +But what other solution do you suggest to make Response object usable outside a stub? Is their a specific reason for making SOAP response message usable only inside the Stub classRegardsTAM Tenfold +ok will take a lookMy question was is the used JSF Runtime the RIMojarra or +Great thank you for your reactivityRegardsLoc FRERINGNew and snapshots are deployed that contain the fixes +Heres project before I added seam dep +will make this easier +added link to the bug to the mevenide page. +batch transition to closed remaining resolved bugs +Looks to be similar is an issue is due to the fact that the copy is not atomic +Im having second thoughts about using since the name returned includes invalid filename characters such as +Every family in the get can be done in parallel Every read from memCache storefiles can be done in parallelSince is addressing paralellism at level is this issue going to tackle both and levels +Patch that adds the terminate just before we close the recovered queue +What I was envisioning was a series of checkboxes above the output that controls what is shown +bq +Ivan I am not so clear on this JIRA could you please give me some more description for better understanding of the scenario +Description is too vague +committed as deaedcfcfddccbed to +Jason please take a look when you get a chance and make sure Im not doing anything crazy here +on Teds patch +this is really annoying +Im not suggesting we add the ability to resume across domains +Otherwise itd be should still keep an eye on this but the impact to Spring Social may be minimal +on the patch but I cannot apply it on latest trunk +and its own jar. +committed to branch Ill check the tomorrow and see if that did the trick +IBM has released a version in which this is now fixed +What the intent was was to keep people from shooting themselves in the foot when editing in the browser by messing up say their file and making the core unable to load +Hehe yes by GA not version +Wiilem are you fixing thisI am bitten by this now working on another ticket +Both stores are provided by the same jar but the store is not the same the directory data store is actually its own code base when it finds a file that can be opened then it istantiates a shapefile data store to handle that file under covers +I fixed the race in replication worker in the new patch +Thanks this is the exception Im looking for +Use both patches in order. +Amusing side node ROME was created by Alejandro Abdelnur the creator of Oozie +Because the file is being renamed or copied just using a file lock on the file doesnt really work +It would help my current project + +it just add +ok Ill do it +This patch includes the previous patch a test in eff +Reopen to revise the impl and port it to so the users can migrate to the new patch is applied and the new implementation is ported to so the users can migrate to the new one before upgrade to +I think I refactored in the right place +Committed w Bryans better wording +Submit new patch addressing your comments +I will commit the patch shortly +final version +It would be great to have a ginormous patch right away that uses an IDE like IDEA to enforce indentation style and copyright notices and such +removing mutation optimization for +just to confirm did you actually test that bit of VTL? I tested it +Works for meThank you Alexey. +Thanks for the PatchGreat idea to integrate Apache Thrift with MVC What do you think about renaming according to the same naming as? libcsharpsrcServerMVC libcsharpsrcServerMvc libcsharpsrc libcsharpsrcCould you merge libcsharpsrc into libcsharpsrc to help reducing maintenance effortAnd please add the mvc option to the THRIFTREGISTERGENERATOR macro within compilercppsrcgeneratethanks! Roger Meier ive changed the code style as you wish please check the attachment patch looks good + +Thanks for the catch RamIm for backporting this to +I vote against guestguest because thats used to mean to access as a guest without privileges +Patch testing v +New patch +Also if you only want parse or only want format why pay the price of initializing the other portion of code +Before the GUI was always showing the oldest available Version as SNAPHSOT +Verified at +to in SVN rev +Please comment reopen if you can reproduce on +Weve decided not to move the code out of as it would not make sense +This is the whole schema +Any comment +Added a unit test for the new metric +For this reason the patch incorporates new methods that are amenable to the introduction of such checkingIm going to apply this patch but I had to Ignore teststesttestas the test XML files are not included in the patch +So I just rewrote the dependencies in and such that the pdf goal uses this batik jar +This XML is backed by schema that is how it enforces it +Hi Simon Ive ran with X e +By default is on +The patch attached passes each member of Notification separatedly to printNotification +When that ratio is below epsilon it makes more sense to declare it I could also see this being a case of caller beware +Just regression tests +Each of the projects posted here are person projects and we wont be able to award a place to students doing the same project so if both Ishan and Dishara really want to do this project we will have to choose the best candidate +Initial version of the patch for mainline tests jar falling test is excluded for are updated to have nicely formatted outputA couple of missed Before and After methods were returned back to the couple of files were missed from the last updated patch gets all the tests running on my machine except for +Looks fine to me +Reviewed by George committed +However I think my suggestion simplifies the code and I do not the extra overhead of calling notify on each send instead of checking a volatile field should be very large +fix pattern matching the SQL you! Pushed upstream. +This is ready to be committed +The problem is that the IBM JDK used has old API +Im afraid the Apr th snapshot did not fix the problem + was based on private functions with small ideal situation if we can use without me the most preferable solution is to ask hibernate team for changescause can provide list of paths to ejb resource configuration files it is seems natural if provide this info this is my is easy to impement feature and I can create patch for do you think about patch to +Committed v thanks! +Should I wait next step for reviewcommit + last patch looks to be of a different jira +yes it now works Try snapshot or build locally from sources +from Pete drls should be placed in the ejb jar root +We need to be able to clearly articulate the goals of the effort and then understand how we can all work together to accomplish them without duplicate conflicting work and destabilizing Hadoop +Merged in dev commit befacbffaafe +bumping to I dont have time to do this right now will do it later +Verified in CR +The problem seems in the way startGrails parses arguments +Should we spin off a separate sub task? Should I take over while you look at the pooling issue +Whos gonna care about the time after Just kidding of course +AmirIf it works there is another experimental option called +Option sounds right to me +I just pushed this +is there something i need to do that has stalled this issue +I have written the patch but it remains untested +Per phone triage with Burr and Max on +I use autowire all the time but I understand that weanies may be a bit old all resolved tickets from or older +Related to this consumer also never receives a message that the connection to the broker is down +thanks you for the pointer anything like this for the custom deserializers +Arjen being a committer of Castor Id like to know what the issue is +I created the issue as Im more used to work with JIRA and did not found any hits in nabble about this problem +It hasnt merged down into version or +Please check the +Agreed +Fixed at revision the mentioned issue with property access identifiers of nodes is still the path instead of the identifier +ShellyThis test sends http requests to the JMX console +The CVS version works +Will be the correct location for? Its used by the language code but I understand a goal of yours is to not have the language code depend on the jdbcapi layer +I am curious on this warning as line is exactly the same case but without any warning +Ive added it in revision +This is basically a problem of recovering partial in files with syntax errors +Committed to. +Gavin thanks for your response and suggestion +For users running the quickstarts some confusion but not a deadly problem as these are just examples and maybe not a serious user problem how may users will want to manually interact with the console for apps that they deploystart via API? Maybe few or +not sure why yet +Have you tried the Streaming Server or even just the regular one? How would this work with someone who already has a separate Solr cluster setupAlso I havent looked closely at the patch but if I understand correctly it is writing out the indexes to the local disks on the Hadoop cluster +because one important comment was missing in the wiki page and needed to be added +setting the fix version to unknown for into this +can no longer reproduce after what I changed to try to fix but not convinced why its fixed +The attached patch solves the static injection problemRegarding issue I guess there is something wrong on your configurationEnsure that struts can see your file +Now T Jakes solution looks nice +After this change the code works for either operating systemAlso changed the the kerberos login module authentication to optional +no new tests added since it only changes a log message +I have added the patch as per the normal process +Heres the reduced YAML that will cause the bug +This also fixes the assert failure in debug builds +Thanks for the help Bobby +Attached patch is based on r before recent changes to scorers +Committed branch and trunk +This patch adds to setup of the system property before starting up a mini dfs instance +Test case that passes on and fails on later a developer please review and confirm if this is an issue +We do not release based on alpha releases either +Brian for sure if this is a bug fix too or just for trunk + +steps to reproduce an often reported issue +Patch uploaded +add dstartupPath to avoid file not found error when launching not in the bin directory + +Please can we make the build archive all of the qatestoutput folder not jtsremote you will need to update the job config to archive this jobs already do archive all of the the test output qatestoutputSince the zip is in the testoutput directory it will automatically get my comment wasnt clear +Not sure how to proceed here +Committed the final patch to the trunk. +So while I am trying to get this available to the community I have to now a few more obstacles to overcome +All it does is configure the to report to Ganglia and then starts up a separate thread which listens for metrics +Ill submit the patch tonight around CET if no one complains by then +Yongqiang was not committed +The bugfix tests include a rule which exercises this +Thanks Sid! +It seems attach the autocomplete element at the wrong place I think it should be appended next to the textfield it replaces instead. +My bad +patch applied rThanks! +Bulk close for +Ive provided the artifact which can be deployed to reproduce the problem +In my testing the only instance of onreadystatechange that I modified was the one within the +It seems I overlooked the Antoni message +I just committed this +But when I look from a user perspective I feel if we allow users to Stop a remote server then we should as well allow them to Start the remote server +This only happens when used with Cordova +That way wed not be filling up the REPLICATEONWRITE stage with a bunch of for the writes +You can add a feature request in Jira hopefully there will be some time to do it or a contributor. +The password attribute in the xml configuration is now deprecated +Jozef is not available +for the latest patchIf there arent any other comments Ill commit this patch +Strange +Deferred to the asyncexclusive attribute in the JMS message service +I am unassigning myself from this bug in case someone else wants to add JDBC update support for other resultset types +The file should be created in the libPath directory +The current algorithm has two parts +Bug has been marked as a duplicate of this bug +Im almost sure youre doing something wrong in your actually has the Exception paged queries not supported hasnve probably mismatch both the debug logs you should be able to see INFO Dialect Using dialect the dialect used by EJB screw up the dialect +if optimization of join orders occurs as normal the test fails +As a first step I replaced and with and respectively in revision +There was a bug in apacheworkerFixed now +Here with no package name we just have to allow the script to be as if declared in the default package +xml file? please read statically properties instead of static fields +Im sending the patch again with the API improvement thanks Juergen plus more test cases exposing this problem fixed in TWMC + +Looks like fix itself that I applied was incomplete as well +and make that as transform too +Gives dedupe functionality for shards based on either shard order in the shard param or by score +In general Im for this +Verified by customer but dont have a unit test case +The puzzle I have is how to retrieve the byte format of the regionName of an existing table? Thanks for your comments I am still fairly new to compared with you guys I really appreciate all you guys insight on this +Im testing the patch now +At some point we should also think about a better name for Source something like or something more self patch adds UOE to codec and makes docValues transactional to prevent wrong flags if exceptions occurI also added some random docValues fields to as well as some basic checks to +Need a demo project +Bulk close of old resolved issues. +Reattach it because the previous one has some mistakes +Thanks! +network load in the node cluster was consistently higher even in idle modeAs Ive mentioned before our tests were focusing on finding the overhead of using counters not on overall performance impact +is very very old this problem might have been corrected in the sure its not +Loading classes from the parent classloader conforms to the standard classloading process and was a deliberate choice when we implemented the component +Can you take a look +ZhengI went through all of your related changes to my previous code and they are looking fine to meso a as far as related stuff is concernedYou just uploaded a new patch for branch +Will commit after regressions +Looks good to me +closing this JIRAIve checked with Daisuke this issue is already fixed inThanks for checking +Konstantin I will take a look at the patch shortlyIn Hadoop summit we agreed that this change will not make create idempotent +Reverted from and are two calls to nextRow in nextInternal +This JIRA is now not part of the release defining criteria for M +bq +Willem please refrain from committing this to so late in the game +As soon as I downloaded FOP complied transition to closed remaining resolved bugs +If no application and test cases are broken by currenty Harmony code I suggest not fix the problem +Before applying this patch must be applied +Committed at subversion revision +Arthur questions +will work +You might want to try upgrading to that. +Attached a test case that show the example of corrupted snapshot passed Adler check +Reference please +Ive tried trashing the local repo too and it still works fine for me +It looks like you didnt do the work to have svn track that these files were moved +And yes the test case may be big because I added the Hibernate so anybody would only have to call ant run +It speaks for itself the are killing us +I dunno but at least were +I want to work on this bug but I couldnt find where to file a ticket +Good luck! Leo +I noticed the host name of the region server was wrong it appears does a reverse lookup and then caches the host name which was causing my issue as I was connected to a different network +Reproducable with +ColmI just update the configurations for both the client and server side to get this test more or less working kind of +is this still an issue? I again vote for making the name s becuse of the persistence model +Can you closemark one of the two issues as duplicate +Thanks for your report and proposed pacthSince this issue is related to an already released bundle I have created to track this +I forgot to say that those tests also fail without applying the patch +Fixed in and merged in master +Manually tested. +Please find the system logs for one cassandra node with debug information +I have changed the test to use the test script submitted by Tars this morning +Thank you Usman. +If they match the OLE header signature it hands it off to POIFS +Committed revision +Still need to work on the documentation +It works on Chrome if I run through development mode +Earlier patch works but since it contains deleted files patch should be callled with E +Seems fine to me and it runs correctly +I will into it in the next one or two days +Copied patch from +Yes please extract the dependencies file +patch can also be applied to branch as modified files are the same + AnatoliyIf it is a bug it should be also fixed in jcr so could you please backport your fix +Oops +Above comment should readCurrent build of USER guide can be vieweddownloaded from +there is also usrconf which should not be the case +Hi BoIn my prospective keeping state from the first version itself will make more senseThere are other questions like how you are going to submit the with the mapreduce jars etcLets discuss those once you have the design doc readyThanks really would encourage you to post a design document as well +bulk move of all unresolved issues from to +Guys please take into account that you probably do not get what you try to reach +Move issues to next release +to DTDWhat warnings from Plugin Manager are you referring to? +Probably a duplicate of this is completely different as RM here is trying to reserve memory on node manager which is more than what it has +We are storing Jackson Smile encoded blobs as bytes in Cassandra but we would like to be able to get a human friendly output from +This should probably also be posted to RB +an ant task +looks good to me +There are more oppertunities to use the string struct in place of char +IMO We should update the version number in the to and respectively as part of this patch +Attached is the debug log file for the makeCall Request +unregister the things specifically registered by that HolgerAh I see +Anything else to be done here +Thanks for the patch +Thanks Andy +Is this something that can be added to branchx? With high fuzz and ignore whitespace the patch applies but then fails to compile +patch tested confirmed and its good +changing the code to return from each thread once it get the response +File and URL have an eachLine method +I do not have an environment available to test any of these but they should apply similarly to windows given the JVM options documentation +Ted That makes sense +To remove the label well need to add a instanceof in Form to not render the label +It seems to be working just fine +Hi Paul I tried to set the content type with but nothing changedCan you give me a simple jsp example with request and response working with euro simbolDo I have to set the property in the VMTnx types of problems are always symptoms of improper encodings and not a defect of Struts. +Otherwise leave it as bulk +My wiki user name is +Can this be fixedo Hang in clustered async store tests +A build snapshot should be available as soon as the build finishes +Proposed dont think the problem only occurs with +Verified fixed in BRMS CR +I tried deploying the attached +Then select JavaBuild PathUser Libraries category and fix and alike references to none existing jar files +Now I tried with a sftp server on Linux and the behaviour is the same +This is not a bug note that readOnly is defined as a hint not as an enforcement +I would keep permissions instead of grants +Please tell me exactly what you filled out in those fields +The execution ordering remains unchanged +UnixLinux has the same max command line argument issue although people hit it less frequently because it is larger +A great thank for your quickness +Rev contains an additional fix for recognizing the error msg on Windows emitted when cannot be found +I will attach a program to reproduce it soon +The only little nit I have is that they seem to mix tabs and spaces +Thanks +Bulk change of shipped and resolved. +Where do you want these kind of issues? JIRA +Sorry I just found a typo when running the tests on another machine +This gets past the job submission step but test still fails later on +Looks like wasnt figured this one out in the end +What do you guys thinkTalking about the next part Im writing about the MINA services +If so you would see entries in the Simple History that would show documents being tried and retried +Just prepare a patch and post it here. +Shouldnt we? What do I missBtw +Fixed by +Thanks for looking into the patchI have added the java doc in public methods and classes +Fixed in trunk merged into +I just committed this +If you move the last in your demo file as thefirst youll see that the first page is handled correctly with no margins +For the future there is a way to running QA tests automatically +Somewhat modified version of the patch applied in Trunk works with and +I think I wasnt clear +Hey when I run the DFSIO benchmark one thing I find is that only one node runs all the mapper and stores one copy in that node and distribute the second and the third copy over the other nodes in the cluster is this how this benchmark behaves usually? or am I missing out something +Thanks in Advance +Good point +The local option is mostly obsolete now that filesystems can be specified in paths +There were two problems with this +We also got to implement server attachement contains wrtier classes for tool to generate C stub and skelfollwoing files should be added to javasrcrootmodulescodegensrcorgapacheaxiswsdlcodegenwriter and was renamed to javasrcrootmodulescodegensrcorgapacheaxiswsdltemplatesvndiff files contain the modificationsregards generation for AxisC currently support only DocumentLiteral style in server side code +Updating release notes fields. +! for the fix as is +Patch applied to trunk and branch. +Should the issue be marked as patch available or is this still in progress? Im OK with this being committed +Thanks Simon +Another reason why I do not use the in Sling +Ill commit shortly +See what I did recently to Aegis +I dont know where the findbugs errors or javadoc issues came from +Springs does not implement Executor it does in +Core failure is caused by but this should not affect unit tests +Posting this patch as a first step to test the schema scripts etc +The reduces would do the shuffle merge reduce call in C and use the existing infrastructure for the C recordwriter to write the outputsA note of caution You will need to worry about +Bringing into +bulk move from M to to it should be possible to edit defined filesets +Thanks Ning +This bug depends on and +and there are +I tested in the current SVN and it work fine. +This sounds indeed like a good proposal which has also another great advantage coming back to a more standard way to execute Sonar +Did you try connecting to mina sshd through putty and test the timeout? Or could you let me know how to programmatically increase the log level? +Also it might be good to create a authenticator and not make it global in the future as said in Tomcats community +See the attached for a first cut at this on trunk +With this change there is no way for implementors of sendMessagerecvMessage to signal to the callers that something unexpected happened or any other error conditions +thanks +ThanksI will check that out for inserts wrong HTML it closes divs which have not been opened +Fixed +The current fix addresses the requirement and usecases we currently have +This will delete all my messages which is not that niceAre there any other solutions to this yet +From pgoracle trigger tables they record it is BEFORE or AFTER as to when they issue the trigger +Hence Trinidad generates the warning message as it couldnt find the association between default skin family and desktop render +Patch for tests update attached +This is a patch Ive had pending from last weekends +Thanks for the patch Ill take a lookColmPatch applied thanksColm. +With you can limit the size of the most recent txnlog to MB +This file is normally under the logs directory of your Tomcat instance +applied patches for trunk and branch +Actually no there is no source code for those +Committed to branch +Sure no problem +Fixed at fad +Im just trying to get the back office apps refactored at this pointThanks for always offering your input Chris +redesigned for extensibility to avoid duplicating the logic in +Hence it is not safe to remove thisOther fixes can be made +Either you get orderable child nodes or efficient handling with many child nodes +We lost solaris native support at some point but have it back now +There can be no default that satisfies the majorityOption +Add the attributes renderHeader and prefer to check the rendered attribute of footerheader component over adding extra attributes to datatable. +Here is a simple test case that demonstrates the problem +Applied patch and added documentation changes +Whats the advantage of this change +How is this issue +In my mind only one variant of name component should present +In case of shell it may be better to run a separated group serverprocess for the clients outside the domainNN is already a bottleneck of the system +This may be an MVEL issue Im debugging this one myself at the momentto try and pin point where its going wrong +You should always use in your documents so that the Xerces implementation does not needto try to guess and figure out what you want it to do +Have added a patch file as the html version doesnt look right +Todd even it seems odd that Seam doesnt support parallel init of multiple war in EAR it is the reality +are you sure this bug is fixedI didnt found any source changes about Parameters +System property set accordingly. + rock! +I am assiging this to which Id like to release +BTW its not only new usersdevelopers who benefit from us using the standard conventions its also like Eclipse or Idea thatll have an easier time figuring out the project layout +Adding to the affects version +Attaching which bumps the last digit of the release identifier after generating yesterdays snapshot +Prashant were you able to test in environments where is and is not available +Jie do you have a sample script with dummy input that illustrates this problem that you can attach +Im not sure if you can detect that situation from C or notMike BAIS is unfortunately not arbitrarily rewindable +If HDFS goes down we either have to pause all Oozie work or we have to temporarily write the logs locally and then upload them once HDFS comes back +Additionally protocol buffers decoding requires you to read byte after byte while both and my can detect the length of the whole encoding after the first byte +OK Ill give it a try +Thanks Amareshwari +defer all issues to +bq +New patch looks great Luca! Ill commit backport shortly +We will rollback this fix so will be fixed +Probably I understood your questionYou have several definition files but only one of these files needs a default preparer +ER based on what I heard today +Updated patchI dont think we need to bother about a global lock for now +The Maven classpath container can contain any libraries as long as they are declared in the +For convenience I moved the patch to the reviewboard +We are waiting to migrate SVN Git repositories to new server until then Fisheye updates wont work +I cant find it parsing these anywhere +All seem to work +You could just apply it by hand to trunk +to daniels idea +But in general when I publish things to the web I want to know what URL will be used to display them +Thank you for you quick answerAnd Im making an application for my workThere is one nasty bug but I have forgotten itOn Monday I will report A bug which is also blocking my developmentIts weekend and I have freePSo dont release it! isnt fully stable +I also have added webui as a component to this issue +Hmm you use tomee? thought it was the snapshot if so can you try the snapshot pleasePS im on freenode openejb if you want +forgot the remove the class +Issue introduced due to refactoring +The same client then performs a read on that write however the operation is routed to JVM +Patch is targeted for trunk +Please disregard earlier patch +This should not happen because FAILEDFAILEDUNCLEANFAILED updates will fail the task twice thus doubling the counters and punishing TT twiceI would say It should not accept any state if the previous state is FAILEDKILLED +Wendy Ill file a CLA tomorrow and send through fax +Often Tapestry can determine the translator from the type of object to which the value parameter is bound but when that is not the case the translator parameter is still required +I tested it and works great as always +Should we continue to keep it open? Should we apply Si recommandation? What do you people think? then Scotts supplied a patch in +It seems to be a bit involved +Thanks Marek +Then we should document them explicitly +Committed to trunk! +So will let him take action + +Can you upload the war file that fails to deploy +I have changed the title of this issue to match the improvement provided +Take a look at least +The requirement on my end is to support queueing a message that will get handled at some fixed point in time in the futureAs it stands now I have to hack around this by having multiple queues each dedicated to delaying a certain amount of time +Making read repair would be a lot smaller of a design wart than what nonlocal indexes require +Closing. +Initial change was made in for making compatible with Spring WS was modified to have the lowest version of to be +Unable to reproduce on with the procedure described aboveIs this a bug or has it been observed on another operating system too? +Qian even if there are CLOSEWAIT sockets the number of file descriptors limit is usually much bigger than this +XML characters are escaped new line characters are removed from error messages extra attribute assert number is no longer present +Yeah it does not make so much sense to use a producer cache in the send processor as its a single producer based +Committed at subversion revision . +FWIW I agree with the proposed I am in favor of this extensionBut I have mixed feelings regarding the default value for the method attribute of the slinginclude tag +for apache plan to address this now +To avoid confusion it should be explicitly noted that this change applies only to the Lucene flexible query parser and NOT to the classic Lucene query parser and hence not to any of the Solr plugin query parsers that are derived from the classic Lucene query parser +Karl could you please open a new jira for a test case for and attach the relevant patch there +Method add desumes the targetclasses and target from maven metadata ignoring eclipse settings completely which is wrong cause does not permit to have a differentiated EclipseMaven target directory or at least lets Eclipse use targetclasses whenever it needs to run somethingME installs its own when a new configuration is created Magma should obviate this ME behaviour extending the Maven classpath provider so that the folder is used for Magma projects. +Defer all unresolved issues from to +In fact Exadels site shows the problem +So all issues not directly assigned to an AS release are being closed +Block size checking could be simplified by making all files have the same preferred block size and each files last block is full except for the last file +from another pair of eyeballsThanks for taking on this important but tedious work +Patch for menu and titles +Its a completely automated workflow with various validations that occur at each step and the ability to reset to a previous step so its a feature we use heavily +r for some dangling missed code on commit. +you say its a little early to worry too much about performance numbers even though I do tend to be an efficiency houndI think I understand more where youre coming from +Assuming this is a topic Im assigning to Bill +Along with files job configuration files are also stored on DFS so we can directly read the from conf files and use them for the sake of authorization +See for further details. +rd the Central rsync recognizes to come from rsync not from rsync +It is not allowed to have more than one child element in at +would it help if I attached a small project showing the problem +In previous time I connect incorrect JIRASo this patch contains replacement for VMRTNULLPTREXCEPTIONand partly forVMRTIDXOUTOFBOUNDS VMRTARRAYSTOREEXCEPTION and VMRTDIVIDEBYZEROEXCEPTIONPatch tested on platforms +Ive added a check in the session fixation filter to make sure the response hasnt already been committed +And hope to backport to +Thanks Ivan. +Minor changes Corrected Naming convention used internationalization for Showing labels Dropdowns etcThanksSumit Screens and Forms in the right file +Hi AaronI understand we can use hostnames in configuration to make this workBut I feel that will be a only +If you have a mainjar and a classifier jar why wouldnt you deploy them at the same time? Its not a good practice to deploy them separately +Updated Smooks Design Page section +Im currently working on fixing that and will posted an updated patch once thats all good +Always use +Test case works with IBM VMESuspected to be issue with DRLVM kernel classes +They are still there for backwards compatibility for now + +This approach is unlikely to work +Thanks again for your participation and interest in Chukwa +Is that what this issue is about +It will need to be removed from the nutch lib directory if applying the patch versus pulling from trunk +Zookeeper has a standard conf file +How is the the value for JAVAHOME? It looks like it could miss some quotesAlso is this something you can reproduce with older versions of installer? I guess its more an bug which has been there forever +fp is the correct priority per the triage on May +May somebody pls try this +They might have a clue but Ill look into it next thing today. +Awesome +Why inside classifier? Are you defining all sequence problems as classification +such an old bug report and no any watchers case closed can not reproduce +fixes both and +attaching a reworked applied thanks! +Can you please tell me any other changes that would be required for this patch so that I can do the same and submit it again +So it doesnt make any sense to use the async store in this case you wont get any of its benefits you may as well just use a sync store +Its necessary to update the links as FAQ is separated document +Also in the vast majority of cases this is not a significant optimization +If you need the fix now install the latest nightly build +Ill leave the unzip and svn add hsqldb and svn ci m expanding example to make getting started easier hsqldb to a committer versus attaching a large patch fileEric +Jan this works with me on a recent SNAPSHOT +IMHO it is always better for committers to have a related issue patch and not a global patch +This should not even mention consumers such as JBDS but be more open to whos consuming it while being more exact on scope +Brad can you produce a test case? What happens if you run with extractOnly? Does it return the content? I tried both that and indexing using trunk and the example per the wiki docs and it all appears to work for me +There should be a way to tag certain groovy files as scripts and so instead of being compiled they will be copied over to the output related problem with scripts that use the Grab annotation in that not all of their type references would be temporary work around is to set the output folder to be the source folder and to remove the groovy scripts from the build path +It would be great to post a message to the Apache Derby dev mailing list asking about this in the past they have been very kind and responsive +Olivier one question code from sandbox depends on version where anything related to unavailable +There are shell scripts in and but I guess they are executed in remote machines +That said no updates since is a big concern. +This style is replaces by newer outline style +We should get that fixed for +was gonna work on it this week but no is a duplicate but has a longer discussion +Committed for nautilus exploring a file should simply open its parent explore command does not work for files actually its disabled +New patch retargeted on latest trunk +Attaching updated patch +change password from user login admin to for +it shouldnt be so hard to configure +fixed +I merged my localminimr junit test into +Lets nail down the design for encoding cyclic references in the serializationdeserialization code and move ahead with an implementation +Ive added escaping of the rendered value +In revision I also fixed some Tomcat warnings about uncleared variables created by the class most likely due to the way it was being used from the classAll changes merged to the branch in revisions and . +You can run it with the following command mvn clean compile exec seems not to be in +There is no archetype in neither of these repositoriesArchetype itself is only in Maven Central and can be updated to use enterprise bits when entreprisetrue any time +I introduced a which is the default now in NamedQueryThis change break backward compatibility at compile time runtime backward compatibility is unaffectedI am not too fussed about the backward compatibility the previous behavior was stupid +Same exception with +added functions substituting POS taggers by Parser this issue fixed +I think the separate contrib directory is a good compromise +New entities created do not get the extra ControllerRoo mix in +So all issues not directly assigned to an AS release are being closed +if no one objects will commit this tomorrow to speed up the build and increase the lifespan of our hard drives +Other places in the code use the default if the property is not set +Its was just that I didnt know exactly what to do so I expanded for the time being +Ill review and commit the patch +That way the users can still get to see the info +Patch looks good to me +The advantage we get is that we merge this reasonably complex patch on trunk thereby avoiding more merge conflicts like I discovered today +Thanks! I was about to make an issue out of that as well but this saves a lot of trouble. +Can you guys create clone for this for the next release cycle and move the unfixed problem to it and then mark this as resolvd please +Id generally recommend to turn the log level to WARN in production +i had only implemented an by memcached but there isnt any for +Thanks for the feedback +I have no problem finding these variables in the variables view +After few minutes the file is broken cannot view +The basic approach Im taking is that delegation tokens will not have a service of ipport +Updated in revision . +Derby continues to meet the needs of some very different applications +This means that JIRA and BZ are no longer on the same machine +Ill see what I can do +Could you please try again? It may have been a temporary problem updating the search indexes or something +See the last line of my revised comment +It will take some time before the update is reflected on the web site though. +A similiar issue happens with the milestone release of activemq +assigning to you to reflect your lsat comment indicating that youre working on it +Patch checked in thanks! +Testing on local now +Though it should be checked +JanThanks a lot for the test case +Reopen if needed and assign to if that is the target release +If youre not happy with how this issue was resolved feel free to reopen the issue or create a new one describing what you would like to change. +Pass in the selector key into +Lets leave it open for reference until somebody figures out what is going on exactly. +Naturally it is easy to create a custom script or shell alias say jirb that invokes jruby S irb. +I just did a clean up work on integration tests now they are use the karaf test helper to set the profile of KarafI will commit the code once the camel release candidate is cut +With defaults no events are seen + for the change +So thank you for your contribution +Im not sure how youve gotten an NPE logged without the rest of the stacktrace +Without this file the dependency report generated fine and the project still got built without errors +Interestingly enough jirbswing works fine for me on Windows and crashes on Linux +Hi +We have a use case for this feature +Sounds good +removed comments +Revisited patch after talking to some of the guys on s client looks good +class type +Do you want to implement it? Get me something that works in its own maven module and Ill give you SVN access +Will open new issue if other ideas about making splits faster still +I dont now why is not marked as resolved again +be verified in gatein master and reopened if still valid +Well the approach might impose new problems +Verified +Many thanks. +bulk close of resolved issues. + +We dont go to the web for validating we bundle our own dtd for speedy setup its possible for users to configure a different dtd in the should ensure that we do have a recent dtd please check this when you try updating hibernate jars for the other issues +I suspect the problem is due to bugzilla status fields not matching the default jira set but there is no way for us to fix that +The patch was committed in revision +I was able to get the PHP client talking to the URL endpoints defined in the serviceI took out the consumer secret generator hack and moved that to +Yeah the plugin is currently accumulating the entire response in an and writing it to disk only after the transfer completes +For what its worth I think I saw a change bundle go through the RI last week to combine the api and impl jar into a single jar going forward. +Testing in progress +The Transactions chapters are different in EAP and EWP +thanksHave a good day +Also present on GAThis really is major pain and prevents our production usage wquerycache enabled +This has been resolved when fixing the deadlocks in MINA and making references to the minaRegistery and minaService static members withing the +Then it was a pigment of my emancipation that you had +This is a startup script for that shows the queue lookup error +A new getRow call in Table using to support this command +The word Pref is ahem preferable since the Preference class includes the User ID Item ID and Preference Value +Patch committed to trunk +patch without for solution that in addition removes the hack onno extra checks yet that validate if calling either method related to children orderingis actually valid +Heres an updated patch +should have this gtg by like I said Im just insuring that I havent dropped the ball +i was wondering what is the procedure for creating a patchIs there any documents that i can follow +It went ok only once without + + for this patch Jakeanother thing I had in mind is a disable all languages within the or just enable production ready languages by default +Its not a blocker +Updated patch for MR +bulk close for +Alternatively the simpler solution is to fix the setConfiguration so that it pulls out the property and then the wrap can use that rather than existing method parameters +This change now logs the errors warnings and info problems +is x faster than the +apache xindice in an component +One problem I see is that only the leader can really maintain its own last zxid up to date +also recipes is a bit more funSimilar comment as doesnt this result in a large number of directories in trunk? Is that a good thing +Verified in ER see in ER see screenshot +So to really test that the timeout works we need to keep the broker alive and somehow delay the response from the server +I have run the TCK on these changes and it appears to be doing better than the previous version so I take that as a good sign +Seems the tests were dependent on the implemention of +A folder named is created under the project folder as well as +Just to satisfy my curiosity are you trying it out with the PE like Stack did or you are uploading something else +I recognized that this only happens when deploying snapshots with a profile defined in +Closing as duplicate of + +I know the heat has kind of shifted away from Nutchgora but it would be great to clarify what this issues actually encapsulates +Some people currently do field collapsing with queries first you get a few hundred documents with just the id and collapsar field +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +everything installs fine it just doesnt run quite right +Each time finlizer threads were active and doing some work +Actually lets discuss this first +Thanks +Nevertheless as i understand the postscriptspecification an eop must exist after a page ends +Thanks Mahadev. +I have included a patch and test in the file +If it is moved to it should probably be renamed to clarify its purpose as there might be a need for an RFC version laterIm now wondering whether the code should throw or log a warningMight be too drastic to throw +This is the initial outline FOAF plugin +Ive created that injects slfj logger into beans +At a minimum youd need to make the NN chroot aware by defining to be another path treat +I have added a corresponding comment to s javadoc as well as to the JPA section in the reference briefly considered logging a warning or even throwing an exception when injecting an extended into a singleton bean + +I just fixed this +You do not delete promotions you expire them using thru date on the stores tabPlease do not create Jira Issues when you want to ask a question use the user mailing list +for now marked as portal but it might be a more generic AS runtime issuesnjezana + +has been released closing issues. +Rob was this fixed or +Thanks for the update Jason +settings as tmp dir is not a good solution using system tmp folder is much better +If user write partition numbers for Hash partition Tajo make by numbers +There is also a in +We can if necessary. +Let us check please +In my humble opinion using lower case for mime types is more +I made a few minor changes to your text +Weve finally decided to provide the ability to go beyond page +Hive is currently packaged with so we should keep that optionI changed the default to false as there are cases where we change the metadata outside of hive metastore +The MR job which I told of in the last post is already written too but I dont want to add a dependecy to Hama +Thanks Pradeep. +Weve experienced Lucene index corruption a few times on a production system +It also seems all docs build fineSince this is not part of code generation the risk is low +From what I have seen from the current code I would expect that the component name is based on the class and the service name is based on the attributeDoes this make sense? If so I have already implemented this +This issue has been fixed and released as part of release +could be a better name this class holds and mutates the numbers that will be exposed as dynamic region metricsIll add the comments and check the line length +See resolution in and . +I hope that could be done by profile but AFAIK maven allow a single profile at time to be active +I took a stab at cleaning this up youre welcome to review and comment +Your suggestion that we farm out the closing of regions to the regionservers themselves will make no difference regards how fast regions close and ditto regards deletes +This is good forThe grid operators can configure an appropriate write timeout value based on their own specific situations +So forget about that ideaLets look at this differently +What component makes use of the slurper? Is it a controller or an event handler +the wrong element is destroyed +Resolving Override annotations pushed upstream as well. +And the disk checker interval is minI think the am is possible to assign to this nodemanager in minAlsowhen the cluster is full of job this bad nodemanger is the only one that have memory to run containerSoit is easy for am reassign to this bad nodemanager. + +Attaching patch with Testcases +Updating fixin information +works for me right nowIll note that lognets download page is not using the proper ASF mirror setup and will rectify this by the time of the release. +After some more thought the only case where this logic is generally applicable is with correlated subqueries against a common table +I have replaced the version of +I think we could do this assumption +JuergenIt seems the snapshot didnt help resolve the ve also tried including resources using both classpathcom +Thank you for email but you sent me the same file +Bulk move issues to and +I agree Id like to see Spring used to manage transactions rather than adding it to Struts +I dont care if maven can cook me dinner or get me a beer out of the fridgeThousands of people can comment on this issue about how great it is no one caresThe bottom line is that people are going to be uncomfortable with it being in our releases and these threads will continue to pop up as long as the maven artifacts are handled differently from the other packaging its just that simpleBy making it consistent with the other packaging people are less likely to complainbecause then its not such a mystery and isnt a separatedifferent product +Even on a loaded cluster the right decision per node may require more global information +Moving out of +Hi JacquesPls allow me complete translation first +Users can now use both Resource driven and descriptor driven injections in handlers associated with EJB Stateless beans +is reverted in rev +Maybe the group query collection code ignores these remote exceptions? +The question is what to do when the sender already has set a correlation id +Could you give use the full list of +Thanks! +This might require releasing a beta version of the EL spec in conjunction with this release +I just linked both pull requests here for completeness is all +I have not worked on open source projects before however +I think this issue should be marked as resolved? +How much raw data is stored that may often be enough they need and no rowcounter run is required +A sent patch is worth at least checking itCan you please check whether this applies also on beta thus whether it is reproducable and the patch fixes it also thereIm not sure whether the same OS conditions apply on the virtual test machine of Win I have available and your one +Patch integrated to trunk. +Looks great MartijnIll commit in a day or two if nobody objects +Just for the record has calls to getItem which all potentially cause the same problem +OK so this was just explained to me in. +I think its safe to apply it before +I could be wrong I have not dug into it very much +Can you take a look at this for +Need more information broker logs broker config client logs connection URI etc +Now the is just a wrapper over and forwarding all calls to itWhile running the test I saw some failure because I hadnt set old jars pathI will run the tests again and will post the result after that +Stack version of GS are you using? I think something similar was fixed recently +Committed at the revision +Runtime detection wasnt changing runtimes when starting a new JBDS configuration in an old workspaceI have fixed this issue +Its only a performance issue on the initial startup rightWe have a precedent set on making proactive on detecting error conditions in the configuration +Maybe this goes without saying but this would be excellent for implementing wikis or other systemsIt would also help in designing distributed systems where there is less mutual trust between instances pulling a new version of a document from a peer is less dangerous if the existing version persists and can be rolled back to +The test failure looks unrelated +has been released closing issues. +HelloWould you please try this patch? Thanks a lotBest regardsRichard +fix pushed to fix confirmed from upgrade scenario +Reopening in order to add tests +Another option is to adjust the tests to not use the example theres not that many of them +were going to need to add a feature to the code in the assembly plugin in order to resolve the source artifacts instead of the dependency binaries. +Detailed responses forthcoming +Let me double check with the assembly +png +There are two common cases where this happens +Yes certainly +Committed this to trunk and +We need to supply the Devoxx organizers with a one color high resolution logo ASAP +rUbuntu testcase smoke kernel +This means that you can just override any observer method with an empty method to disable it +asserts to find where the null Message body is coming will its a race conditionheres a fix +Thanks Ill close it. +Marking this to be fixed in as well keeping it open tillthis has happened. + to go into xfiresrctestorgcodehausxfireaegistype that didnt do the trick +where do you see that rpcTimeout is ti be seconds +. +mr v patch is the for branch only +Block if the number of stores that have requested flushes number of stores unless there is only one store +That was the eclipse formatter +Waiting for your test results before closing the generic and outout are now correct +axisserviceswsdl or by using Admin Web App +I dont think that abstracting out the serverManger and catalogtracker has any real value at the moment +I havent dug to the bottom of it yet thoughI want to make sure we dont blindly plow through an and end up deleting something we need +Branch remotingGA has given birth to the descendants of Remoting including which I believe shipped with EAP which just shipped with AS and which is targeted for AS and +Introduces JMX attribute and a JMX object for querying stream inout status +However this is fixed now and the issue Sahan has found is a new one on hash table +updated issues +thx Mikael Andersson + + +The container implementation must be fixed before +Verified +blockCursor is in the right placeCtrlIns the same as CtrlC copyShiftIns the same as CtrlV paste +Problem is that optimization leaves tablescan in the plan +I guess youre right +Add on commit +With this patch the users do not have to initialize spring by implementing class and running it on +That would fix the retransmissions although first error may still occur in the log +Could the problem be that the verification client which sometimes do not get the connection is started before failover is completed +Shouldnt the system behave the same with debug enabled +Logged In YES useridThis will be fixed when blob handling is rewritten +thanks lasantha +Applied +Also attaching the test output +Most likely we upgraded OL in since last summer and we did not port the upgrade to communityweb copied over the version of OL works +I was indeed referring to the batching that you were talking about +If you want a different behaviour then you need to register a either a different implementation or differently configured implementation +There are very few of us working on this project and as such cant fix everything +yeaprobably you have to ask someone to authorize that account +For previous version. +I figured it out after I looked through the docs again and saw this Note that some modules are not enabled in any plugin and will therefore have to be added to the plugin dependencies in order to include them +What makes you think its not +Ive tried to summarize the issue with the attached release note +Manually tested the patch by throwing errors in and +Thats a step in the wrong direction +I fixed the issue +Im not aware of people implementing their solver classes but they implement the function they want to solve +Original edited patch was messed up +oh +Hi GaryCan you attach a patch here so that I can pick it up and build locally and test it with servicemix test if it fixes the originalUlhas merged to fuse branch +Wow this has been fast! Thanks a lot +Dont know if there is some technical limitation that doesnt allow Start of Remote Server +new SNAPSHOT buildScheduling for next SCR plugin release and closing this issue now. +Agreed on the need for a change names are importantI have a problem with using match on its own because the word is often associated with partial matching +AlexeiIts my fault +Attached is the test case +Committed this to trunk and +Bumping to future for now +This simple patch dumps the leading zeros in a hex number in case when the corresponding byte of and is less then xI ran derbyall and without failuresPlease review the patch and if you find it to be ok please apply it to trunk and branch also +To be reopened if a concrete problem still arises with with freshly compiled +Not really as described in the issue but this at least triggers a failing test on trunkJust does a scanner row count before and after split to verify it didnt get borkedThis test is already one of the less desirable tests that relies on sleeping +is now covered by and context. +Add a unit test to +not for too much rewiring neededBut the concept is good +I was thinking of creating an Annotations Edition with JPA support in addition to the current XML Edition +I still need to update the class as well and then I can submit a patch hopefully in the next day or two +Verified on Beta +Kris should I close this? Im not responsible for the designer itself +bulk defer issues to +I applied it +They are with a varint and then written on the wire adds test for the writable format itself adds a call in that uses a protobuf as a parameter and responseI see one spot for an optimization by removing a buffer copy on the read side but Id like to leave that for later +There was a nice param in that is in but not trunk +Here is a complete cfscript version of cfdump +Any chance of getting a stack trace to see exactly where this occursBTW I think I get it now Camel type converters in each library +You can hit Retry from the Web UI and Nagios should install now +I dont have access to that anymoreas far as I remember I used it without any configuration inside the pom and just called appassemblerassemble from the commandline +Of course this only applies if someone where to finish and it where accepted +I just checked again and I now see itvery strange +See for status of demos on closing all issues that have been in resolved state for months or more without any feedback or update +It seems a common issue +Fixed on revision +Resolved. +Why are we doing this +In this patch removed from default configurations +We tried reducing the number of replicas and the performance actually degraded probably because there were fewer places to read the data from +Ive defined an external Maven Installation and its working fine now +Synchronising access to all data structures is not feasible it would kill performance +well its rather making karaf eclipse friendly +You mean that I should get Spring source and do thisThanks again +I can reproduce with Alphab we probably cannot fix this for Alpha but we should get it reported at eclipse since this just seems very wrong +is uncertain and xml is +here is another one. +Getting feedback and iterating is much easier on trunk then having people apply patches etc +I think this is exactly the same issue as and and has been around in all versions of Guvnor that I have worked with +Problem occurs only when composite primary key have associated entity idWhen entity contain associated entity as property hibernate generates inner join from criteria and all works without problems +Should we close the issue with that +Resolving as dup as per Yangs comment above +Please feel free to go ahead and commit it +Indicated change going into another jiraHence dlosing this issue! +nvm I see it inwhat is the code for in DD? is that a todo that should be cleaned up +The view depends on the privilege that user harry has on t +This doesnt work because there is no message for the java type boolean and textplain +Maybe it could be an attribute in the components attributes collection +I updated to make this an improvement trivial as it does not change the functionality as the FACELETCONTEXTKEY is updated correctly as you said with or without settingit in the constructor per +I added serial anyway but you can exclude and subclasses from your checks +fixed in one go with in +I wanted to name it shorter but have no idea right now +We will fix and update AS this AOP? When +I will commit it later today +I wouldnt have a problem fixing it in a later release especially if removing checkbox is troublesome +Talked to Eugene a bit on this issue +Thanks Isuru for the contribution. +Also you may wish to review update for SQL module +The JSP pages are now to servlets that can access classes. +Please reopen if you have a valid WSDL to reproduce this issue +Closing issue fixed in released version. + +committed +seems inconvenient +Havent check the trunk code but if we went ahead with this fix well need to figure out why there isnt a and ensure there is one +Since the is part of the spec itself implemented by CXF no extra need to be included for +I suspect this is not actually an issue and is actually user error but in case I am wrong Ive moved it into the youre right Ross +Fixing Affects and Fix in Version +We are using of the pluginThe strange thing is that the same version of the assembly plugin works with normal deploys of snapshots but the problem manifests with artifacts built with release pluginI found an older release that uses release plugin and of assembly but still exhibits the corruption of permissions in a tarballNot sure if this helps but you never know +Applied the latest set of patches unchanged +Brian and I are also completely happy to support the installation for anyone who wants to use it and it integrates nicely with LDAP and has pluggable security realms so if LDAP is not ready we can hack up a realm to work with whatever authenticationauthorization info thats available +Closing this issue as requested. +Clint says it works for him + +Reuploading patch with ASF license checked +I forgot that Jira doest link the submit comments with the patches the new patch is +The past aside Spring Security should remain passive between any minor release except for a vulnerability that has no other means of being third option would be to create a new implementation of that is called so you know historically speaking code I can pull into an IDE gets looked at faster than design docs +I havent tested the patch but it looks good to me +part of release. +But we may as well just have one JIRA number +Verified during test cycle we have many examples in distribution right now if there is any objection about current state there should be raised new less general JIRA hence closing. +Brian is used to update ESB category in project examples xml file please review it when you have a chance +Did not use the XML representation +Can you post your config so we can see whats wrong otherwise check again the URL I gave above it must be the sameThanks +yes +Ideally I think there should be a separate setting to toggle for secured pages and components +The the imported script is called which writes data to its own backed by the request writer +From comments seems like this is a and we can breaks apache axis xsdate handling. +Also make implement Closeable +I am this issue since the Release Notes item has now been corrected +the that there is there is a generated file +That is likely the source of the issue +However as a temporary workaround Ive added an entry to that turns down logging for to FATAL +Thanks Remko! This issue was fixed in revision +Same code as in description +Will the patch be submitted soonI have performed the changes you suggested to make Opera render the inputCalendar but I cant figure out the offset problemPlease post the code to fix this as the example from didnt workDoes anyone know if this will be fixed in the release +This is obviously a difference between and +I would like to eventually move to a based solution +lgtm as ve committed this and merged to. +I just committed this. +The save period is seconds but Im actually not sure about the sizeSeems like a good idea to supply as much information as possible without crowding the output too much +I cant seem to reproduce the error even when I try to match your environment +If there is a problem please open a new jira and link this one with that. +submitting the unit test is NOT related to this patch +forgot to delete the file in the old one +This doesnt occur when deployed on EAP with the production profile +Added alias in projection production and improved the formatting of jjt file +Ben ThomasThanks you for the changes +The patch committed fine thanks Dmitriy and Alexei. +FreemanNote that this sample uses the exact same wsdl and xsd files from for which there is a fully functional standalone deployment attached +Ill open another issue for adding caching in the rails version +Need to follow this message on a live cluster I cant make sense of the above logs +I can try to merge it to the branch for inclusion in. +Committed to trunkThanks Rohini for clarifying the difference in Hadoop. +Why are we throwing instead of Exception? any reason +If you do pass that param then we set it for you as a convenience +Closing as this works for me now. +if the FK is between two unique PKs OR Indexes then it must be aAnd yes it shuoldcould be done via revengstrategy but also needs updating +Page breaks inside tables are ignored +this is bc the default is not regionating and that requires primary keys or just a repeatable fid +moved to beta for ease of testing this done in Beta? If not whats the status? +Am I right about thatThe solr user would be because they very well may choose to ignore the warning +This way the widget object is requested one time and is passed through the remainder of the render cycle +You are uncovering why some tests are sometimes flakey +The reason is that there are users of Summary that dont want a single String with HTML formatting recovering from this format is tedious and +It sounds reasonable to have a new Void status which can be after ConfirmedAccepted +This can be fixed by changing the ordering in which the bundles are started +Thoughts? +Fixed +Bumping to +attach correct file +Attachment has been added with description the offending offending tagsattributesvalues are commented +There is no workaround +I committed this to trunk and branch +method is now protected and thus allows customization by subclassing +You basically have to create a mapping for both user and item ids which must also be used by related jobs like the for ALS and the one that evaluates the error of a factorization +Heres the version with the tests +I modified the row bridge to check for the get function on a child handler and use it if available +Im downgrading it to Minor so it shows up lower on my task list +Anybody want to take a stab at this? Keep for or bump to +will handle under the work for using +on patch +Everything seems to be fine +The name max may be confusion since it is not really a max +Fixed tested so closed. +I added the missing usage of the tint transformation in revision +Thanks Jian! +Committed patchs from JIRA from JIRA from JIRA New revision version to branch +How about trying out different message queue implementations based on external dependencies in the contrib area +Anyway this issue is fixed with +I will create a JIRA for that +Thanks Andrew! +One better location to decrypt the password should be in the connector jaxb class so there is no need to create a temp file while starting up the server +The idea is right +Oops saw your commit log still not tested on Windows +Attached is the modified so that you can check how my proposition looks like in real life +its way easier to cover all cases +probably because is currently the most recent attachment +I dont think we will use an Cactify code +This last patch is looking really good +The patch looks good to me +Edward not sure I understand at this point it is an unused variable and it serves no purpose +Adapted title to correct schema version ie +Attaching a version of the which has been modified to use a transient database rather than creating an unnecessary database +BrianI think updating the file should be acceptable for an interim solution +Not reproduced +Michael duh I think smart chinese has damaged my brain for the rest of the dayThanks for fixing this +Testing now +updated patch with the before shuffle +Benson any chance of a test case +It works with +it means my current implementation which opens works only in case if only one sourceresource folderi should consider this more common case so it be workable +All tests are successful on both platforms now +I have just pulled and noticed that the instructions of starting the server have changed so I will mark the issue as resolved +ah we need to upgrade xnio to CR +It is actually a requirement for this to work +The Table component already use For and If components which can be +else if +I realize this is not much information but wanted to share what I have +update a new version to fix some issue +Its pretty trivial but there are way too many booking examples to mess with all this on every app on older branches especially considering that the paging behavior is so poor +Nothing against Jackson its a great lib however the size of the classes were which increased our sdk size about x +If there is no response for a while Ill steal it from him and fix it +The should be intelligent enough to the config file and then start talking to the new JTI guess Amar is trying out Option +Unfortunately it took us one year to get through the patch but fortunately the patch still applies +Im not sure why theyre there though since I dont think they were used in the page +We can continue to use the scheduler but write and ack quorum are separate entities now +This sounds like an IDE issue am I wrong here? If it is Ill close the ticket and this should probably be reported on Adobes JIRA +Moved from to just after merge of old branch into TRUNK + changed them to package private +And for now I dont see any side effectsI hope that this patch is included as soon as possible because our product needs IDENTITY key generationThank you +Yes you are probably right its mainly architectural issues +Thanks for explaining +Various tests showed that attached patch resolves the issue. +Then aims to amend to take the branch name as an argument from the admin job and check out the appropriate branches etcIf that understanding is correct then it would seem to me that this JIRA supersede the work done in though some of the work already done in that JIRA may be useful here +will post in few +Completed At revision +patch looks interesting will try and see if I can get that in for have changed the patch to handle custom added I couldnt find any appropriate API in order to remove the projects can I know if the projects classpath is included +The attached patch fixes the initialization +Thanks for looking at the patch Ted +Fixed in revision . +As a workaround you may have to use the regular since the only thing it stores in the session is the token +There isnt anything that the workflow or scope interceptor can do by themselves because they are doing what are supposed to +I ran into the same problem with Win XPI could run Apache only in debug mode with X +As Gopinathan mention may be that error was due to CONFLUENCE upgradationNow it works. +is still HTTP context whose primary purpose is to keep contextual details about request execution such as those that need to be shared by protocol interceptors +Attaching and a revised version of incorporating Johns suggestion +It is not even commented out in the JBPAPP branch so I suppose this change is done as a part of the patches that are applied to EAP during the build process +Proposed patch +I have incorporated the changes for the manual into the patch for +fixed +The process by which sends notification using email does not work the way the original developers had intended +I think I may have just figured this out +But only exists and has content +However there is no servertools API to get a delta when a server object has changed +Patch has attached +Thanks Cheolsoo for committing the patch. +I havent even started looking into fixing up the last patch Keith put up +it says Note that XSLT requires namespace support +As Fernando noted and are frozen currently so I cannot put the changes into build right now +The approach is similar but based on the new masterrs interfaces ZK classes etc +I kept the names as close as possible for the first attempt +modulessrcmainjavaorgapacheairavataxbayacomponentregistryjackrabbituser +will be fixed for added some basic details on JMX and Spring integration +Ah sorry +Ill try to dig into it this weekend +This is a duplicate of. +Other patches can improve andor fix the content +I will create a MANIFEST file for the compatible to by hand and include it in the build +Just and FYI I am not using a mapping file but I am using an ANT task to generate the source code +Patch implementing all schemes +Tests passed cleanly for me on +Aha +Mamta can this be resolved as fixed? +Closing as per last comment. +I feel like the API should not support this +Test everyone +The tests along with a very basic framework can be found in the directory in the svn repository +Moves the properties and getterssetters down into Field +We can benchmark how long it takes to compute checksums for a block but I doubt it would add significantly to the time for upgrade +Fixed and in r +If an end user really does need a timeout for a particular method they have the option of using async methods anyway +Will try on cluster w square of the number of regions +I dont intend to review andor block this jira +No need for Struts to support this out of the box. +See tag reference for details. +Someone was going to get around to telling me that right? derbyall thanks for your patience +I wanted to test it on an actual cluster first +I haven tested compatibility and Id be happy if someone does it +Also we have to implement oneUDF to retrieve TIMESTAMP in certain formats from stored data +Note that in an email conversation many months ago with somebody I realized that one of your problems with performance was almost certainly that the default threadpool size had been reduced from to threads in jetty +William do you want to work on the +Ill commit tomorrow +If this is only an issue with the AS line I wouldnt worry about it +This means that specifying packages should be sufficient for your use case now +If we consider that as a bug the example should stay as it was thus having a test case for this issue +Andrew whats the status of this +User has steps listed and I followed those steps and did not verify the bug exists +james I have no issue with configuring things to work as they wish +Confirms that these errors are not due to my patch +In Section Configuration Properties in the Discovery Configuration table the description of the excludedContexts property uses the text +I would suggest to upgrading the bundle lists to the released versions and only upgrade them again to the current SNAPSHOT version as soon as the bundlelist itself was releasedWDYT think the laucher should always use the latest trunk version +Attachment has been added with description Patch for has been added with description Patch for cleanupShould be fixed in FOP transition resolvedfixed bugs to closedfixed +Can not reproduce closing. +Indeed as Michal very rightly pointed out this issue is no longer relevant because Infinispan servers cannot be configured with store since they already store data in binary format +Thanks for the patch Evgeny. +The user should reopen this bug when he can provide a stylesheet that shows the problem. +I think I found and roads lead to Rome where these two seem to be Rome +Making it more generic the client should be able to specify a Codec to be used using the Encoder server will write and client will use the corresponding Decoder in the CodecIs that part of? Maybe the only part we need +Verified in SVN branch +One possible use case would have a user running deletetimestamp to clear all data collected before a certain date to make room for the new +I added this to help myself get a handle on quite how this API method works hopefully this might be useful enough to others for inclusion +Builds on +This patch includes a backported +Patch checked in +That service cassandra status problem is resolved by I think so too Closing this issue +Thanks for the info +If you enable the spring plugin then you will get the error +Thanks VarunPatch applied to the LUNI module at repo revision rPlease check this is fixed as you expected. +Possible but it would certainly take longer if were going to break existing clients +Uwe thanks +Lets keep an eye on the test and reopen this or create another issue if fails again +target file +With Each event has a header with control information the event itself and a footer +Hum +Im thinking maybe the cleanest way to do this would be to encapsulate the max size logic in such that it throws an exception if the max size is exceeded +Yes looks good +Ive attached the patch +Jay what do you think about it? Could we just assign this issue to +thanks for the confirmation! +Im not getting any translation of the asirra block at allWhich Browser are you using? Ive been testing it with Firefox and IEWatching the screen it seems as if everything is loaded in the opposite order than you have written +Slightly too agressive on comment deletion +Actions taken up the initializeJob sudo chown usermapred R taskTrackerjobcachejobid sudo chmod R taskTrackerjobcachejobid initializeTask sudo chown usermapred R taskTrackerjobcachejobidattemptid sudo chmod R taskTrackerjobcachejobidattemptid sudo chmod s taskTrackerjobcachejobidattemptid sudo chown usermapred userlogsattemptid sudo chmod userlogsattemptid sudo chmod s userlogsattemptidchild umask + +Do you know when it will be implemented +Duplicates. +Base is more opaque but guarantees data size at times the number of bytesFor default values Im not worried about the size but base is a more standard way of encoding binary values in text than perverting unicode +Most of the failures in the test suite are caused by by spelling typo +CR becomes GA so for now no issues to be put there +Add variable HADOOPCLIENTHEAPSIZE and user can set it either in or +See following comment +If the queues are deployed first then you will not see any error messages reverse the order and they will may be that having a remote DB does alter the timingssequence but it could also be present in other configurations +Petercould u please post a sample soap message from +We dont get any warning message for no package from runtime +Just checked the commit +Zone haltedWill leave data around for a bit before nuking +I understand its important for you +Patch to apply on the plugin projectAdd method information in manipulation dataIt modify classes to add a method allowing to get the component instance from an objectThe patch add the file +It should not block the release +Ben why mark this as WONTFIX when there are several very very simple solutions that could be done +patch testing sees unit test failures that cannot be reproduced on linux machines +Committed revision +Although you are welcome to continue mirror our software we will not list you in our official mirror list + the threading issue was fixed in that the issue is filed for +But the annotation is ignored +Less than week to pushing off more items to +Thanks for noticing + has been rewritten for and does not have this issue +Add trycatch +Rebased the with test patch with cascade db drop fix in metastore client +Yeah if memory serves what counters needed was to replace columns but to fix that we added a replace method directly instead of doing a removeinsert +We can surgically remove it from the JAR +Patch looks good +All deprecated classes and methods have been removed +Knowledge on how to use a C compiler and a dynamic linker is required +This could be done using existing argument parser IIRC I saw one used in V detector patch for example +Changes have been made to make the porthost configurable through the process engine config but Im unsure to what extends this is reflected in the classes mentioned to verify this with the +You need to use CONCAT UDF +Merged to as well. +updated in revision +I have some documentation to clean up and then I will do a release of about the delay but the day job has been consuming most of my enclose a patch that uses attributes rather than elements for versionthreshold ensures that the XML generated is is also bacward compatible and integrates straight out of the box with would be grateful if you could accept thisPatch added +We have a couple of Nexus repositories which are being searched +cant we use description and shortDescription for this +This issue is out of date and will not be fixed +i think jira text means to say work with EPP instead of work with JBDS +I ran twice based on current trunk and they passed +tnx for attached +Path apply thanks! +I havent noticed the original issue therefore we have broken CI tests for the example +Please remove theI just looked at two random packages from PEAR and both had closing tags at the endAlso I dont see any extraneous whitespace +Closing feature implementation should be covered as part of +Can anyone please help me out with this issueUnable to generate client stub for wsdl containing multiple ports +This would be useful +Sounds good +Ben cant find usrlocaldespotsbininitrepository on beaver +sorry I meant make sure the examples gets updated +Moving out of +How do I recover from itAnother similar scenario is when a checkpointer starts to upload the new image to NN but fails in the middle +Christian I looked at your test context and to me it looks like you should be able to reproduce the problem with it +a third simple option is to not save fields of which value are the same as the language default +Which classes should I take a look at +Any way I will try to test them soon +Changes in itestqtest are not required +Thanks Ravi Teja! +Although int assignment itself is atomic if you dont do access inside a synchronization block theres no enforcement of a memory barrier between threads so one thread might not see the other threads assignment +Assuming Clogging +This +You were a little bit faster than me I hope I set Grant for +If this is a Maven problem then it is a very special one +Update patch that addresses Sureshs feedback +Just a detail I agree +Thanks Phil +Im curious how you would use this +I have no good handle how risky this is though +OK I will commit it if the tests pass +The one only complains if the whole request is not a one +you are invoking the service inside a running thread then better to use fire else use the code indicates the send and sendRobust will be handled directly by the while sendxxNoneblocking and fire will be queued in the and scheduled for executionThis fire method is quite confusing for the first time use +The following projects have these processors new project been added for purposes of building the supported client jarfile +Thanks for the patch Laxman! +Note also that Mongrel is behaving correctly now that were not holding references to dead threads in a threadgroup which is why it got stuck trying to shut down +I looked at the stack trace +gitignore +Its not an infinite loop its a deadlock is blocking which is cleaning a socket +Sure doing that now +Closing stale resolved issues +Thanks +patch with typos +The new UI has to be built with use of Airavata API which abstracts all server logic +Sounds good to me +I believe Instead of synchronizing the methods that does the operations like clearaddAll +I verified the fix with my gprasad customer test and conf tests output and output. +It does a sequential scan of all leases so there is some overhead in renaming a file +so people can tell them apart from annotations at a could you check if it is possible to control proposal sorting in Java Editor? +Is there a way to make this project temporarily limited to a small set of people +dims +I could surely use some guidance on how to publish these patches with the least amount of disruption +It is up to whomever is deploying geoserver to ensure these credentials are properly secured +I dont think jarjaring every dependency SolrThe only jar files in question are snapshotunreleased code +Re and this should bypass any unnecessary marshalling if there is no one else in the cluster +Might want to look at compression at some point +Hi Olegpublishing artifacts to apache repo is quite a straightforward operation when using maven +Hi +also the library created should have a name like +Sorry I am still a bit behindAnyway can it handle this caseA custom artifact is loaded before the manifest is loaded +I meant that by default all adaptors should publish results +If is resolved this issue will resolved naturally +Implementation is effectively identical to the same behavior in +I know gabriel is going to making use of that api for this catalog scalability proposal +unit tests and minor fixes +Deferring to not much we can do about this +Thats the point of this bug report +The tests now pass when invoked from the top level dirOnce Anita blesses this I consider Tomcat module migrated +Tattletale says that direct dependants on are have release noted for Hibernate with a link to the unsigned jarWe can recycle the text from that applications fail to deploy because of a +We never were running b +I am also seeing this error on Mac OSX and grails Graeme. +no more +That would seem reasonable. +before May build +One of the things we missed was changing the root logger to ASYNC which is referenced in another issue so we would definitely want things using the ASYNC appender in the long run +There exist some alternate strategies in to circumvent this problem but JIRA is a bug tracking tool not a support forum +Hmm yeah I would agree this is a bug +Mmm I guess +That would allow clients as well as using various standard HTTP infrastructure such as load balancers etcWe already use HTTP and HTTPS as the transport for HFTP and HSFTP +Li Click the Cancel Patch at the head of his JIRA upload your git thingy then submit the patch again +Having an import to an external schema containing the feature type description is a perfectly valid response +I think hes done it +Thanks +hmm +Figure Image is page +If I delete the contents of and then run the as above I get a proper config +The patch is now committed +By any chance is this patched version available for thrift revision +We should create a mutually exclusive nameattribute that serves as an alias to the id attribute +On the other hand the default is relatively big which implies a low impactGenerally if this is about adding a setting which would only be useful for configs which are bad in production then why should it be put in? You have a means to configure it as is +The right to solve this is a of the forward slash so +Move close to final block +By looking at other system procedures and the way JDBC methods takes the table names it looks like following the second approach is more consistent with other system procedures in Derby +Linking to because this problem came up while thinking about that issue +Yeah like Scott says I bet this is the same issue as what that commit fixes +I used a slightly newer build than Max Tools Version You may try that Max or else I dont know +This seems to have already been fixed as a part of some other issue +These test cases now look for the correct value associated with the docID. +Patch committed to issue has been fixed and released as part of release +Yeah the speed depends on many platform dependent details like cache size etc +Hey Ross thanks for investigating this +If the application is still running isnt already deleting the apps current working directory and distcache files out from underneath it? +An Ant plugin handles the artifact deploymentI seriously do not understand why this is such a big deal +Ill do it as soon as I canActually I wasnt using JAVAWSDL but Eclipse plugin +to trunk and +This patch effectively adds a readme file and renames +Looks like the same fix for needs to be applied to mailing listHiI probably found the workaroundat line inchange resources to HTHChanwit it works for me +Removed the backward compatibility methods for Ajax success and failure handlers ajax channel and precondition +Regards Paul Elschot +Confirmed duplicate as reported thanks. +pushing this issue as Im not sure how to fix this for relative windows paths without messing up cases where the value is not meant to be a relative windows path but happens to contain a backslash +So +If you want the latest and greatest will be out shortly +This is partially a Pig issue but it would be terrific if we did not need to recompile a client to run against two dot releases of hadoop +But I think it is not nessecary anymore because I can a attach a workspace to a layer groupTobias +patch was missing a patch due to issues after testing patch a bit hacky since is not a subproject +I think those are the only twoRegarding the exceptions that is a strange thing +A currently does not differentiate a created ledger from an open ledger +e +Did you get a chance to run the test program +We have to fire Metro bug to resolve this issue provide the patch to speed up the need to verify whether this is still true with update Metro artifacts Metro stack related jiras previously marked as for community contribution due to lack of interest no action. +Ive checked it into CVS +Closing as Incomplete +Most have author attributionAlso how does this then relate to the +This seems appropriate to me however I would like verification from the camel riders that it will not cause unwanted +My understanding is that with ant he is running the xmlbeans task twice each time with a configuration of a separate xsd and fileWhen doing the build with ant I had to have to runs of xmlbeans in order to generate the beans +Since there are two solutions available now Cellar and Fabric this shouldnt be an issue anymore +Please also see for pointing that out Enis +always defaults to true in the default caseIncluding support for always breaks the default use case +That is true regardless of shadow mode though since hibernate is a dead state and the node doesnt go live to reset the hint timer until the replace has completedmy understanding is due to the generation change of the replacing node marks the node as dead as hibernate is one of the DEADSTATES +This is needed because it need to be available in runtime. +Thinking this may have compiled but failed during tests I then attempted maven serverwhich gaveBUILD FAILEDGoal server does not exist in this projectTotal time secondsFinished at Tue Mar CLT I dont know if this is a separate problem or related +Logged In YES useridhehe as i already stated in my comment from +implemented +Gotcha thanks! +Here is my http requestresponseRequest URLhttplocalhosttonrsparklrphotosRequest Method Code Not is a screenshot and the error I am receiving in my console +It would be good if the snappy jar is only a compile time dependency but not a runtime dependency +The downside is this one isnt available at Maven Central +Roman could you please grant license to ASF if you want the patch applied? Thanks +This patch splits out axis from and from patches applied by rev +Bulk close for Solr +for the patch +I called the package something like with a files like and Windows which I think is key if I tried import the editor would complain import cannot be resolved +I suppose the patch was meant to be attached to +Rev fixes a problem in the corba +The problem is that this fix requires another Remoting release + attached patch +I have now created a tagsJBOSSWEBGACP tag +location to ROOT? Unless you use the shell to do it yourself I cant see that happening +jeremiah +I will try to sneak in the compiler page into Gora CMS +I also could have sworn that they were there as well +has been implemented for in r +batch transition resolvedfixed bugs to closedfixed +LOL I dont knowI think I meant that for the patch for java mail container and was tired +as per comment it is expected natureso closing +To qualify please run it on a cluster of nodes verify the fix manually and please let me know +Attaching a port for trunk +Can you check that +But alas if there is no plan or never will be a plan to host the Sphinx produced docs on Infra boxes you can close thisThanks +Just so its clear Id like to point out that I made those methods private fora reason +Is this a request by Han language readers +This leads to a +It looks as though English strings are being hardcoded and will end up being inserted in text that is localized to other languagesthe resulting composite text will be an odd pidgin +yes messages are externalizedbut ptBR is not supported yetCan you provide new properties files with escaped unicode content please +According to gavin the only thing lacking is imports +You are deprecating it in the past to remove it in the presentThere is no alternative to as an HA solution without shared storage besides checkpointingIll look at the test cases listed hereI will for and submit a patch soon +This patch requires binary files to be added to svn repo under srcjavatestdataupgrade +I ported the embedded fix to and +Thanks Brock +I reviewed your patch and it looks good to me +Yes register for the dbupdated notification +What are your thoughts? AaronI was under the same opinion as well on PSDP +Could someone please close this as a non issue? Futon already remembers as Simon mentions which I also confirmed on Chromium +attached the correct one +Ive already run into this limitation several times during the development of my project +no problem +on trunk patch +this might have already been done upstream please check +It makes sense to meAnother improvement we could consider is enhancing to do a more general did you mean type search +But work with the SNAPSHOT in the short term +The break will also be very clear compilation error which you must fix to move on so were not making a trap hereUwe are you OK with the rename? I think it actually does make sense that it be in the base class +For this issue I have done some changes +Quick patch to guarantee hash value is always positive +Patch addressing Eniss comment +The test was incorrect +This is fixed in just needs an integration test +I think its needed for +class after checking that there is at least one +here is a patch +Merged this into continuing branch in. +This way the memory required is constant per adaptor +Procedures do not have access to session level temp tables +This table displays instances and requirements +Fixed +attaching DB dumps before and after upgrade management server upload the datadump before and after your to upgrade for us to investigate +If everyone uses the jars from the AS then we dont have to fight with lots of different versions of a jar when integrating into the SOA Platform +BTW the version archive should handle spaces in a better way but not create corrupted har silently +Verified +Double expansion of native path args should not be the default +This is a current version problem +Ill experiment and see how we can make Aries Blueprint and Gemini Blueprint coexist peacefully in Karaf +I take my previous comment back +The POM is generated during the ant build to give Sonar the needed information for its job +Based on committer feedback heres what has changed Binary compatibility has been preserved +I will disable the buttons for CR if the consumer is not +I ran some tests on a production machine running Nutch local having about m records +Actually this is not the redirection but rather how I create the content of the file etcaptforceyes the point of the exersize is to have that file in there so that apt gets to be friendlier and not asking so many questions +responses to move tasks from COMMITPENDING to SUCCESSKILLED +Looks good thanks for the commit +that is useful context. +Updated patch +The version is the ESB does contain the class with the right s issue was related to the scoping and no doubt would have been problem would appear to be the origin of the with serial but this does not appear to be coming from our server +Looks good +No problem +Weinan please look into this +VM is ready who needs access +In fact it was always the intention to make script folder settings a setting but we wanted to wait until a user requested it before spending the time to implement +Committed to trunk. +Patch relative to sure what the patch root should that work is well underway with AS all previous community releases are +I increased the timeout value to +Setup to reproduce the problem + Im able to consistently reproduce the same issue inDo you mean the release? has been committed to branch and above +This is only a minor improvement so no test added +Junping there are lot of changes to digest hereCan you please provide a summarywriteup of your patch? This will help me review it +Any ESB project such as a myESB project created in the JBDS will need to be added last to the server +Updated patch with extra tests it wasnt correctly handling time zones with zero hours and negative minutes. +Bulk close after is released +Please review and commit +There are always several places to improve but at least the public are done AFAIK which means we can now close this issue +So I think this can be resolved. +We use confluence wiki and it has a snippet tag that grabs the source code +Sorry for the noise +As a result when my node crashes or is shut down I expect to lose data that is not flushed +Hi NicolasI tested your patch its working well +Already fixed. +Looks like you have set the apache home incorrecthomewenglechprojectsaxisinstallIs your apache located hereAlso is not outdated and is the new release +For these kind of issues can you first ask a question on the user mailing list and file a Jira only when it is confirmed as a bug +Will attach patch +Will investigate later today +Only two bugs are actually fixed between RC and +For example the following text in YAML filerailsversion would work correctly under MRI but wouldnt parse this +Ah yes you are rightThank you for your work +It will be in the next sync release. +Rather than a major API change the fix has been paired back to just ensuring the current transaction framework is consistent across the or so implementations eg +see last deployment I get an error page when I click on failureIn development I get an HTML alertWhich one do you expect in both cases? I would expect the error page in both cases +is really a plugin +Works well Julien you can forget my last comment +Id really like to see the results of this as well as any other tracingmetrics stored in a but see for why this may not be a good added an to capture the timing at the layer but did not follow requests across I dont forget points out that it would be nice to track time spent scanning past tombstones +The purpose of this tool is to ease the work of the analysts during the process of migrating multiple collaborative long running instances from one version of a process to another +My first guess was the use of the combiner +Essentially we need a new validateForm JS method geared for tabs +Please update your gem to +if there continues to be a problem +Why not allow them to override the default built in simple types +In trunk if I add a buddy replication config element to equivalent to the one in the all the EJB clusteredsession unit tests pass +fixed in r +Mind fixing that so theyre consistentThanks a lot Eli for noticing I will update that change +Roman can you test it please +This diff is taken from within trunkqpidccppsrc and makes make check compile without dieing on not testing return values +Patch is patch is already applied +Hi BryanThanks for committing this! Yeah I think we should keep this issue open for sometime till we get some feedback on how it works +The method you commented should be removed or maybe rewritten and fitted into skipInternalUsing a private internal class also sounds like a good idea +This issue has been resolved for over a year with no further movement +I understand that there are time constraints for the release of JB but please note that this issue prevents the combined use of two of the most interesting features of EE asynchronous servlets and CDI +Indeed most of the pages have been simply migrated from the old doc without xooki involved and thus were using my windows box eol charsIve also set the svn property to LF on those files so that we make sure to keep it like that +Diff for and! Patch committed in revision Note that I removed the log message in case a problem with a header field is encountered +actually I think not +Freddy could you please suggest description for this rule +Hello +No +Yes +Thanks for the feedback +I havent done this yet but yes this should work +Notably Im running Oracle u now +Patch has been submitted +Looks good. +Storing the original compilation schema along with the constraint wont solve it since that would not carry information about which function to resolve in the current schema and which to resolve in SYSFUN +Turned out there was an extra newline at the end of testincludeevent +If you must use an implementation like you also must allow a user to shut it down somehow +Verified on work properly. +We do not use the resolved status. +Ill commit and kick off a build. +This patch includes the option to create a WAR file +grummel +Issue fixed. +May be we should have put this one on +Updated graphics on EDGE. +Version has been releasedLets close this issue. +Better example fileroll ch foobar +Im surprized that we are getting a class that can not be loaded hereRegardsBrian bug has been fixed in Xalan Java +Fixed in +Got it +reopen the issue considering Josefs Josef Could you please attach whatever the necessary class file that I require to the issueThanks is taken and enhanced for performance taking but used today to demonstrate how client fails when called by a client should be fixed now since we have been able to run many concurrency tests with soapsession services and we do not see this error anymore. +sorry +Fixed when tested in build . +This may not be a bug but simply operationsconfig issue +Heres a patch for s +Sounds good btw after this discussion I dont think using JDBM is feasibleAll the directories listed above are hdfs directories +Closing it. +The srctest patch contains +Just tried running the test using Xint and it works fineLikewise the other tests that failed andSo it looks like this is the cause of the test failures. +On trunk it should work to use mvn clean test from the of the project +The fixes contained in attached to this JIRA solve s problem +I will make a separate issue for this +versioning has been simplified in +It is up to these plugins to be compatible with Maven projects not the other way around +Ill do it now and assign to you +I will add a test case for the previous patch +Documentation for custom field types implementation is not very good unfortunately so I need to investigate how JIRA checks if field value is changed or not during edit and how to reflect this mechanism in Multi URL custom field type +Well have to find a more general solution +Well I dont understand what this JIRA user reported a whole lot of when running the client +Added hashCode and equals +Committed Permaines changes to the CR tag +Maybe this task should be renamed to verify serial within EJB +Put another way its time to bury wtf is a supercolumn once and for all +Done at r thanks Sascha + for the patch +You should be able to retrieve dus kind of results with Hadoops API itself +Awaiting resolution +So I suggest you there too +Attached patch changes to get a fresh WGS definition for each feature collection +We should then build a new dedicated site without restriction using the profiling rules proposed above with complete PSML content to match +Brian the rat plugin checks for files that dont have Apache license headers +So I do not see the reason to keep them +Looks good. +Marking as an enhancement as it works as advertised +We are interested in being able to add a labelicon to the users profile so other people can see their greatness +Hi Mithun Any update on this? Thanks +is an interceptor and its points of interception are defined very precisely in the Javadoc +The fix corrects the rollback execution but I havent looked yet into whatever the cause of the rollback was in the first place +The CRC system could +And H is not used anywhere in production so its not a problem to change that caseI am tempted to leave the patch as is with out a parameter since it more closely matches what the web ui does and clients have the option of simply setting the value to an empty value if they need to +Added to release plan +Thanks for your prompt you run eclipse try importing this project into your the classpath right click on and do run as will succeed +Thanks Zheng +That would mean the examples are wrong and all the work I did was to support the wrong examplesSIt still works of course +If I only do what I described abstractly above the error does not occure +Moved to a release too as its a feature request +To get around this issue using the OCM I had to clear the OCM request cache to get a new stream +Im going to rename the constants +isnt true +It would be great to have this fixed in should I commit it to the branch +I did find hits for +etc svn change set +Hibernate JIRA is more responsive and supportive than AS JIRA i appreciate it +Some users want a user or application to be able to read table data yet not be able to access sensitive metadata in the schema an HCD attribute for example +I think you are right +Ill track this down today. +Re binary compability I think its fine to break as people tend to recompile when they dump a dependency version anyhow +the attached patch set the lifecycle of the plugin to initialize cant be done in Maven +With you have to write out each expression for every bucket and do the math on all the boundaries +I have reviewed it +committed +Would it be helpful to also provide argmin and argmax functions? The statistics community would probably appreciate the syntactic sugar +Can we at least put a patch in that throws an error when a unicode string is detected so that this is easier to debug +This patch seems to only account for the most restrictive expressionIt doesnt prune off the rows with columns that dont match the rest of the expressionsIt should do both to be correct otherwise you will get rows that dont actually match your all of you expressions +The procedure works now +Instead of maybe +Actually I dont see a reason to use something as heavyweight as MDThe advantage of using a hash of the query string as ID is that you only ever store one prepared statement for a given query +Robert Thaks for opening the issue +Hi Tim thanks for fixI tested trunk and no more freezes. + +Note that when you use the attached patch your WAR project must have a with a entry +I think the commiter will have to svn delete it +Thanks for the feedback! +But when run along with other queries it seems like the specific queries are not being run in test mode +Dont know what is the status of the issue because Im working on another project +Part of the patch came from Cos patch +Moving this to M as I probably wont have time to do it before I go on PTO +Ill log bugs for those +The change prevents unconditional of tokens but do acquire a token for related operations +May be we limit the merge to unique source type by the translator name or a configuration property +If you could mark this as resolved and well let it go to QE for retest Id appreciate it +Marking as Fix in as a patch has been submitted by Shreyas Kaushik and is in review +Can you please take another look? There are smaller changes in a few other files +I didnt actually think doing it per core would be that easy which is part of why I started out with the approachAny ideas about if the is going to slow things down too much? I didnt do any kind of benchmarking with that +The others should assert valuenull in the constructor +Remember for example that as soon as youre usingan opting servlet the resolution cant be cachedSo for the use case at hand I would rather go with a specialdispatching option servlet than extending our already complicatedresolution mechanismIn addition servlets regardless if they are option servlets or not are mounted in the resource tree +Yes this seems to be a dupe of +Patch was applied long time ago already +in here +However the bug does not appear when is the first documentparsed by the +Changes look fine to issue clashes with the changes in +the class is missing in this have added the class to the previous patch +What is your analysis + will commit if tests pass +Close after release +Suppose we timeout and invoke handle +Attaching the patch +if no exception it will stop mins latter +Committed revision +first on the list is java bI think this is most important ive already seen people on the listmention they are using it +This patch like the first includes some backward compatibility issues for the impl project and like the last time in the same classes changing various iterators to collections instead to leverage the more elegant enhanced for loops +is submitted for Hudson submitted for Hudson tests +Lowered Priority to Minor since this is just a test +implements the fixed slot size disk cache. +My thought was that there is still a single class but underneath it would be multiple most likely +But your comments are worth to be read when considering +It is a problem when server is starting up and http server cant response yet +I had to write my own test app for that saga +If its a thread pool those workers need to be idle somewhere right +Please have a lookAlso please let me know if you need help with testing the patchThanksStepan +looks good +The discussion provides good context and it looks ok to commit those classes in svn then. +bulk move of all unresolved issues from to +Old JIRA for the patch John and sorry that Im so late in reviewing itI think a better way to handle this would be to throw a SEVERE level wrapped exception from the itself instead of handling it in +Side note I dont understand why you map to Jar +Are you relying on this behaviour in some way +this is a very exciting feature +The pending EJB support would be +However its done +Patch for jon gray to testUsually blockunblock runs fine but then it hangs +Dropping and recreating index need to have the DDL of the indexes when truncate is called +I dont reproduce the issue locally +in Hibernate Object should be Integer not int +The Master master stats at Otiss suggestion +Please confirm fix and then close issue +When HC is run under Tomcat it has to rely on the Tomcat classloader to find the properties file +Attached patch and fix the problem +Thanks for the reminder +This only applies to or branches +What if youd also inserted through ? As is your prefix marker is probably hitting the top of the row instead of falling in the middle +Looks like for the moment we have no access to a instance +jdo +I think having conditions as side properties of an expression is awful especially in a read language +Youd have to split the logs and replay them which should be similar performance whether there is or +Neither it is that important +Changed from log to warn again missing tests that are included in previous commit +Hi MarkusThis issue effected Markus +One question any reason you didnt also change the checks in and to use this new facility +adds coverage for good catch +In Firefox is OK +Thanks for making our job easierhad a guy basically refuse to provide a patch vs +Closing issue as version has been released on. +One question is whether the weighting concept is desirable or should be removed +Go back to AppStep +The utility can generated random master keys and encrypt passwords using these master keys for such purposes +are right +I have now replaced all UUID references with or as appropriateI also removed the package as it is no longer used. +Hey Bryan just a heads up that I am planning on moving the jar into libcommon in the distribution +Removed some of the debugging +Here is the and a test for itLet me know if theres a better way for me to submit the code and let me know what I should change or add to the class or the unit test +Is your ban also true for? About TP tell me why other tycho projects uses it over thenI think other Tycho projects are not big enough yet to see drawbacks with target platforms +I committed this +Yes that is what happens in +All done +Added handling of query thank you Font Awesome for making my day complete + isnt any better in that regard +Thats a good idea +For this there is a which define which user attributes are mapped to DBLDAP and to which attributes exactly +Cf When this is fixed enable +Does it make sense to move createUri to and use that to create the uri in as well? I moved it to since it is really security related and it is only used by and +For completeness and not to miss understand it +The nd interrupt test does already exercise recursion into a directoryLet me know if you want me to make further changes +Fix minor issues due to rebase work in patch +If someone needs to reload in the future for some insane reason then they can work through the details of tryload and friends and what happens to processes with open ports and such forth +The patch does not introduce any new Findbugs warnings +See on a dev instance that mmcgrath created for us +Martnez +Weve found that many records that were inserted counld not be fetched in a multigetslice query +I used but I had to both build and install the gem in mode +We now use spring to support JSR dependency injections and so no avalon is needed for all this stuff +Good work Jerome +Else can you provide info on what manual testing you did to verify the fix +you mean just have a property pointing to seam something home? +From what I see your patch makes it fairly straightforward to add a that can be used as the amrmtokens service at the RM +The thing about hacks is you tend to get stuck with them +This issue was just to get it on the map for EAP +I think Im still going to wrap this up to be a bit nicer simply because of the ubiquity of this +make patchesNot all generics are done yetit was just the first I will when I get the time hopefully this weekend maybe even tonight +problem +I developed some osgi modules last year which worked perfectly on felixequinox bluebrint container +Done +Good catch +Updating patch to fix testsValidation on single node secure setup next +messages still happen + Running thanks Namit + +I keep this issue open because of the CSS problems +Add the ASF license header to all files included in the patch +This looks like a reasonable change to me +Yup Ill try and merge up and upload a new patch tonight or tomorrow +is committed +I moved the shutdown hook to the composite cache manager. +I plan to commit this tomorrow +bq +It will be resolved in or right? Any update on the planned release date +Removed obsolete upgrade module and config in and trunk +Rohit you missed my point +on greater than +That should do the trick. +not duplicate just reopened wrongly +Having it all run off the pom have saved my team days and weeks in the past +to fix the javac warning introduced by this patch +Default should still be hadoops default +til we sort out possible issue with Estatio app +Discussed with Yongqiang offline this should be a optimization step instead of the current approachSince at that time partition pruning has also been Namits comments +Thanks Zheng +It during running the jar that they are useful +Thanks for the help with thisAaronI spoke with Suresh about it when I took over release manager for as I was curious about it +I see my mental model of the problem ignored the fact that there were servers with newer and older versions my bad +Again only supports validators so there is no way to generically apply a validator to an arbitrary component +This is a serious problem but our experts have not been able to reproduce it +Fixed in revision +has old boilerplate text +Id be happy with Wont Fix if the error message made any sense +Applied thanks +Attaching a tar file which contains the test case +After all nobody puts a garbage collector monitor and a code debugger in the same program +This has been resolved in the quickstart and will show up in the archetype when the archetype is from the quickstart prior to release +Thanks Alessio +It also hung even on a clean subversion workspace without this patch. +Is this patch going to be backported to also +adjusting subject +Ill attach them as separate patch +I will let you know soon +First off Josh Id like you to try something +Image on left side of installer +The testEnv failure is due to a bug in the +The whole change +sorry +hangs with the patchWould refine the patch before publishing +I do have some nits I dont really like the name since stub usually refers to a non functional implementation +Icons and Drop Down Menus could be easily added to the current Selection Bar +Committed at subversion revision +ai and +do not supports rollbacks +We will add a dummy implementation in the future in the abstract class and only the classesthat need to use the new API will implement the new API +There is a race condition when javas primitives are used to change permissions on a fileThis patch fixes the log messages to be much clearer about what is going wrong and removes the extra chmod of the job directory +Backported at r for +Remote but in the same node +Victor could you check if we cant prevent rd party welcome page contributions from loading in jbds since we control the welcome screen anyway? Max means other options except transformation Snjezana suggested here +The trying to tie the stream behavior as closely as possible to the tx seems to be the most reasonable approach +I did run the test in my local box before submitting patch +Whether you use a doPrivileged on the creation or not the context it has is that of the currentcalling stack only +Fixedthanks fix for the latest from the trunk +The other reason is that its compression is quite good which makes it a nice candidate for in memory filters +Bulk close for release +Thanks a lot Brandon. +Another deadlock related issue +fixes the comments for some ciritcal sections +This is now consistent with the warning message that is presented to the user when a JRE other than is used +I couldnt reproduce this +The script dtomcat is not meant to be used standaloneTomcat should always be started as a Sys V service script which sources the config file etctomcat before calling the dtomcat script +I dont see why this matters +Thanks BrunoYour patch is in trunk at r +pooled property of all derived buffers and its parent buffer must be false and cannot be changed anymoreIn case of broadcasting only we could provide more effective solution as we discussed inWDYT +Can I close the issue +Verified in fix version +That is a little awkward +MartinCan you provide a test case that triggers the stack overflow +that the damage is done +I modified the stuff I did so that it applies on your branchThanksbq +Thomas can you make sure this isnt a dupe of one of your issueswork +Hi Alejandro could you add javadoc for the new public methodfields and Override for the overriding methods +Conglomerate numbers for real tables should always be positive +Im traveling for the next week or so so probably wont have a chance to do so +committed +patch against Will! +Once i login I get only the page with only SearchBrowse issue is reproducable only if I use IE as the browser +Please provide a patch that will satisfy your requirementsWe will evaluate and apply it +I cant reproduce this access against a shared works for me +Please if this problem persists for you with the latest code base. +A user whose job requests for higher resources than usual would decrease the free memory on the tasktracker more than other jobs would +Yep +I do not think that is covered by the SET syntax per se +This issue is NOT about any sort of efficiency gains +vikram granted +compression is also incompatible in replication from a cluster with hLog compression to a peer cluster without + will commit after tests pass +This bug could be fixed by patches to other fixes +Compliance issues are not to be closed until both PHP and Java implementations conform +Patches were too broad and have gone stale +Raising the priority +xml and +LudovicI have built the +I think the right place is in +the last comment can I handle that in a separate JIRAThanks for the quick review +The attached path fixes the bug +does not have a mechanism to pass state across core reloads and thats a change Id rather leave toMight simply be best to track all directories and close them on shutdown +Attaching a patch for the suggested refactoring +it is not related to or specific to target jsr +Made sure that anything in the new model extension framework that modifies a model resource is not wrapped in the transaction logic. +JDT doesnt handle these packages correctly +Basic DDL and general help is in place +Itd be interesting to support the union strategy +Linking Run on Android Device does not work on OS Android SDK not installed automatically on OS am not sure if anything further needs to be done here +btw +which can go up to B should be GB +Should be a critical bug +JacopoCould be difficult to test immediately +I really appreciate the help so far Andrea let me know if I can improve this patch in any way +This is the easy patch +are you REALLY attempting to compile srcorgcodehausjanino as the exception reports? due to a lack of feedback. +I do this all the Ill look at this for im importing existing project from local filesystem +Thank you very much for the patch +Summarizing what components that chukwa includesInput Tools logj exec streamingData Collection Adaptor Agent CollectorData Processors Demux Aggregator Alerts Database LoaderUser Interface hiccBuild +Hotspots are likely better taken care of by increasing replication rather than just moving them +bq +There is a patch so if any fellow committers really want to see this happen feel free to take overThis is part of an effort to close old issues that I have reported +The javascripts should preferably be at the bottom of the page to speed up the rendering +defer all issues to +Else I think we can also commit it and build over it +I like the pattern where we explicitly throw the exception better as then we dont have to trick the compiler by inserting unreachable return statements +Otherwise please get theboard to send us a message saying they want to createsuch a list. +Implemented as proposed +Add a line to explain the name distcpbq +If you have modules that depend on each other they should be a single project in Continuum or you can make a special module in cvs so that cvs will check out another module under your working directory or if you use svn you can use svnexternals to fix the issue +committed +i was able to successfully use h with hibernate using this patch +Ill make this LATER in case anyone wants tofollowup with an implementation +I will look into that but not before next week i am not quite sure whether I have understood the last comment +If were gonna plan a break we should probably plan to split things properly +For example INFO +If time is greater then current time set year to last year +Hi Sergey I do understand your problem but I dont think your proposal is the right solution +We did some rough calculations internally to see what effect an uneven distribution of data into column families was having on our network IO +Testing the patch +Its necessary to force everyone to include icons to the application context or create a server or client side routine to find wheres the icon. +If the library is installed in one of the directories ld looks at it will work without the setting +You must define error handlers before any routes which is what the message in the exception saysWhat you have is error handler route error handler routeIt should be error handler error handler route route +Hi Arun Thanks for your comments +applied to all active branches +Committed to branch and trunk +Setting them causes the to go a bit crazy while the job is running but this is a Hadoop problem not a Giraph one. +Introduced a project for the deployment frontend service the project provides the default checkupdate tasksClose if youre happy with this solution +New patchAccording to custom class should be at but curiously it was +was the build for +Still getting an noticed was available +Sorry for taking so long to get round to this I got hit by a hardware problem at the end of last week and it took a bit of time to rebuild my Harmony world on my temporary workstation +I was getting a orgapachecommonsfileuploadThen I upgrade to File Upload and the problem went away. +Looks like this change was lost during the project split +Im going to try to fix this without having to the thrift server +I cannot reproduce this you can identify anything specific about your project that triggers the problem that would be helpful +I dont think we need to change struct +the guy behind the counter asks what do you want on it and I respond everything then he asks do you want peppers +Patch generated with upPossible fix with and +What if nbRows is M and the rows are MB each? Can this not be done as an option on Scan? Then clientside we look for limit count and as you next through the results well intercede and shut the scanner once we have hit limit +As a result of the work on that issue identity columns are now managed by the same sequence generator machinery as sequences are +Fixed bug and applied patch. +The issue is rescheduled to as it could not make it for the cutoff +Thrift is not bad idea +Closed upon release of. +I like the separate plugin idea but as long as we have a file upload tag in core file upload should work out of the box. +Please apply attached patch for the fix thanks +Over time that has proved inefficient or impractical for large data sets though it remains quite nice for small to data sets +It does not look goodI was trying at other option of not calling send if the server is nullAlso remove the RIT if anything has been added +It is being checked during parsing rather than in the schema validation +Heres the output of diff between the original cpp file and my fixed file +Depending on the length of the Spring configuration we get XML parsing errors when more than one thread tries to create its context at the same time +is not run as part of derbyall +I guess Ill switch to using Nokogiri instead of then. +Already dealt with as report mojos can only be called by a report document renderer +In there were many properties +Good job +PaulThanks for submitting the problem fix and test +This will work with either Hadoop or Hadoop is no longer valid. +And we can find this bug only if we know on what file the exception occurs +Maybe while Steve is dealing with he could make the location of the build files consistentNote that within wsm the location of the build files are internally consistent +I hadnt been adding the hbase group to my reviews which kept it off the jiras +Done +checks to avoid redundant mixinis this mandated by the specification +Nobodys reported seeing this since I made my last changes closing +In this version I cannot possible to append a node from other document that was imported +Could you please post the stack trace +Applied patch from which was an updated version of the! +Also replaced all API use by Java API +Yes it may just be part of JBI as a engine +Too optimistic an assumption? +OK I even removed that partitioner class since its now just a +Thats all I could pinpoint so far +This will also make sure that building the dist tarball generates the developer and user guides too +Please check that it was applied as expected. +One minor comment Should we make configurable +Ill check it out +eclipse toolbar view with vpe icons toolbar +The same problem in Portal +Same here workaround in server view double click jbossin startup poller set webportNot perfect but it work +where DN is the closest to the client here is the plan for handling errors detected by +I determined that the error I encountered was due to starting with the attached file if I start with the supplied and then issue archetype this archetype when installed does work correctly for producing a new project from the additional comment is that the naming pattern is only used when the root projects name is contained in each of the subprojects name +ok needs fixing then +Move fix version to in preparation for release. +Some minor speedmemory improvments have been minor in the cn that need to be to thestable lib +These patches allow the tags such as beanwrite to work identically with their base tags when used outside of a nested context +Please revert this change +Holy log files and configs batman! Please provide me a simple example that reproduces the error you are experiencing +fixed +Closed with the release of +patch looks fine to me +Attaching screenshot of the form +They can even make changes and load new configurations which is very powerful +We would just reuse a standard container +Seems like the was merged with and ended up with some unload calls to the related resulted in source table being proxies and could not be resolved hence they were orphaned by is to remove the unload logic from +escaig Fracci que excedeix una quantitat expressada en un nombre rod +agreed many possible exceptions can be caught from higher catchall point +Closing all resolved tickets from or older +I thought my comment confirmed there were no bounces apologies if it wasnt clear there are no bounces and emails are being sent to you codebuilders addressPlease look at your end once more +In the case where multiple temp dirs are configured hadoop should avoid using a dir when its available space is below certain threshold +only uses LUCENECURRENT when you invoke from command line in all other cases its required arg so we are consistent hereWe should also look at the other in this package +Marking closed. +Related +Its simply useless and its never gonna be completeThere will always be a distribution that will have its own directory layout and trying to catch and guess alljust makes no senseIf there is no JAVAHOME set and java in the PATH doesnt point to the JDK then the user willhave to use and +I cant reproduce this error +I dont see any doc for it +committed patch +Attachment has been added with description prototype implementation for pdfdictionary pdf has been added with description Updated has been added with description updated prototype some more experimenting +Source of the failing portlet +That is fixed in Seam +Testing in progress +Not patch available +bq +thanks Owen +please that the NPE no longer occurs when calling clearChildren on a with no children. +i think it might be a different bug +changes plus caching may put need for this off a while +Must be something +Updated the Maven repository zip to not include this artifactAssigning to Fernando to fix the signature +Moving this feature out of +Attachment has been added with description source code FO showing the has been added with description result of the bug on the has been added with description source code of a bug has been added with description pdf with the bug has been added with description Fix table rendering bug for PDF SVG disregard my attachement it has been attached to this report by error +Hi just in case you didnt get my definitely interested but I think well need you to submit a patch via JIRA and click the grant to ASF checkbox for this to be legally accepted before we can merge it in +I left these out on purpose but on second thought I agree that XFMAT should be added +Verified same behaviour on as a note as per my comment above this being caused by an IBM VM defect subsequent tests with the same VM will fail +Ive moved from to since the latter really is our lowest common package holding JSON support +If you can suggest a better name Im all for it +I think Ill make this a preference that is off by default but you can set it to on by having a startup extension that starts it +Connection loses moveObject deleteObject deleteTree add remove checkOut cancel checkIn get get delete getRelationships applyPolicy removePolicy get gains move delete unfile getParents getRelationships applyPolicy removePolicy getPoliciesFolder gains add remove deleteTree getAncestorsDocument gains checkOut cancel checkIn get get delete +The id fields should be stored as text +Confluence supports CAPTCHA however our last attempt to migrate failed +This updated patch incorporates Stephens suggestion of making the log callback +I may need some educating here +Considering this as resolved as it seems to be working. +Moving to CR since there is less than weeks until Beta +we leave it as is. +See +Quadratic complexity beats any optimizationsTo keep the order of elements isnt it enough to copy everything to a before removing? And produce as the result by default +Can you point me to it +It simply simulates the current behaviour in a different way so that it should be easier to support CR later if we want +And so when the message is consumed and a GET is invoked for this file the lock ois set until the connection will be closed +Marking closed. +Entity A is merged to YNow Entity B would be managed by Y due to the cascading behavior you cited +Actually heres the patch +The test case itself was the problem missing suffix of ULL or LL for unix and UI or I for windowsI am closing the issue. +It should be portable to the trunk and branch +o added support for o added charting samples +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Still we never start or stop any endpointMy colleague discovered that in one instance we hold on to the reference of an endpoint and got the same warning +This was fairly easy to get past as you could just change the entry in the WEBAPPLICATION table to be +I think this patch is wrong +However we have no EAR test of Alternative Specialize so I can easily add some. +I like the Idea +Eyeballed output html locally with Firefox. +I was using this patch last week and didnt encounter problems +The issue isnt the scoping of the injected nor the scoping of the persistence unit producer +Earlier this was committed to trunk +And if an OOME occurs because of a few hundreds compiled querries this means than I have a huge problem elsewhere +Can you please list the steps to reproduce this +Milind is an invalid borat configuration +Since this test already has the capability to test position reads can that also be enabled as a separate test? Especially since that was something that brought out and +The question is whether we should replace any of the old unique indexes with the new type on hard upgrade +For the admin servlet issue it looks like the issue is a web container idiosyncrasy +Thanks Todd. +Maybe we need to run guest on our current dep and see what happensOn the other hand Im very interested in having the Avro discussion so its good we brought this up anyway +Committed to trunk +One with an automated warm standbyBut thats just crazy talk. +FINISH created the desired modelsAlso tested in a JBDS GA Designer and got same expected change the default Teiid DDL dialect selection to Oracle and import works fine +This actually is probably the case we are most likely to encounter +thanks for committing the changes +I did test it and unit test passes +Once I saw why it was happening it was quick to fix it wasnt iterating the parent and child linkedlists in the toArray +I am moving to a later release as this only makes sense if we also have an ESB aware HTTP sure it makes sense to do this for the +I attached a simple fix +what could the problem with the Ross Gardler +Updates for GUT exclude lists +No but its the only kind of path that returns and having geodetic in the class and method name seemed a bit much +Anyone have problems with that +There was another timeout related issue in windows specific code +Any guess when this fix will make it into a stable release +Seems like it doesnt work due to the implementation details the issue isnt fundamental +I like the first patch more as the code is much easier to understand +But not to transfrom is perfectly all right +smile +Ill run some ant tests and check it in the security branch +Whether the methods as annotated private or not users are likely to refer them in their custom routines because they are public +Moving out of Post GA activities +Yes it can be closed. +Some changes such as this will probably have a minimal impact where other proposed changes such as and would have a major impact. +So I prevent all region server to chechin and stop these region server after that point +Obviously this doesnt work for static exports like this +Yes if I modify so that it contains just then the failure reproduces easily for me +Are you sure you did svn update? If youre using git there may be some lag in the replica +Werner I dont know enough about the JDO side of things to comment or vote soI simply defer to you Gregory Martin Patrick and Bruce for your expertise +I am agree the approach suggest by Dennis has several advantages +I will add a unit test for both trunk and patch later +Attached a Teiid Designer log file +describes the exception and workarounds a new pull request which sets the purge attribute to false +Can you add a test case to +Ok great! Then this patch isnt incompatible +include jars in lib directory of release builds for. +I cannot figure out how to make the tests work however +Perhaps Im just over used to things going to the console while startingdeving with Solr but it has become something Ive gotten used toI was thinking we just put the file there and modify any doc to alert that you can also start Solr with a D command to use the example logging config fileI could see going either way thoughThoughts +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian this issue. +Regarding the IE issue Im told by our network engineering team thatthis should be treated as a known error wont fix + +Thanks! I hope this is the beginning of a series of patches +Given how works I dont think you can use +Ive therefore build trunk and checked it inr +Yes a account is a good idea +I think the solution I propose in would be compatible with extensions libraries having other charsets. +Isnt that being done automaticallyOr is this still some left over from the of Continuum +Also includes the fix for +Great I just filed for this +Both XML and serialised formats now defer the deserialisation of attachments properties and body values until in revision as part of the work for. +Fix for this patch has been appliedVladimir please verify +Thank you for your improvement +I see +Could someone possibly review this so we can commit if all is OK? Thanks +Many thanks! +Fixed and pass existing tests and the JPA testsBut Steves eye is a plus +bulk move of all unresolved issues from to +Its too late to do it for the branch +Since we set an expectation that Java is supported by default Im also measuring the time it takes to load libbuildrjavaThe low hanging fruits are libraries that we load but are used less than of the time + file +Added template needed for the Hive driver for the AS server. +The diagnose command fails to analyze some more complex scenarios when package uses conflict problem is detected +If someone else wants to tackle it for M please go ahead +Hi Bruno can you please make a pull request against the current HEAD containing your fix +You mean downloading the plugin source code change its and then build it +Still doing some testing +Should we have some way of setting System properties or additional values in the file +To maintain compatibility with other tools that may be configured to monitor the log file then the value should be set to true this will ensure that the log file written to will be a static valueWhen the file is rolled over it will be moved and dated as per the configuration +This fix causes problems mainly with SAAJ +but the code referred to is the latest visible on this site +The replicantsChanged method of gets called with the values of the serialized objects set in the DRM +Just committed to trunk and. +Hi Daisycould your team review the patch and push it to the repositoryThanks in advance Weihuamany thanks for the update it looks very good +Moved to GA but most likely post first approach is certainly to use the Eclipse extension point mechanism as before +released closing. +Thanks Jeff +Verify all failed tests in are fixed by this patch +Okay can you please take care of this report +The local SCM provider is now a part of the application and it works but its definitely not optimal +So what exactly is the effect that youre seeing with set set to truefalse on Spring? And what would you expect to happen? seems to work fine in my case so this issue can be closed then +Ill try and get an SVN checkout going soon and verify if the new version works for my example if not Ill put together a test case +which would fix any local damage as well so it doesnt really matter if you have a longer local hint window since you have to repair anyway +Can improve on previous code + +Making sure all Ps are at least marked Critical to make them standout compared to P and Ps +Each averages the completion progress of its corresponding phase over several tasks and each is stacked on top of each other +hm +Can we start a dev mailing list thread on it +second try of an implementation with minimal code duplication with with the second patch I have to provide an extension in the URL +will eventually need both of packages are currently not available in a maven repo and the split of content between the two jars may still change before JDK is released +I dont see the easy fix unfortunatelyOkay I think I found it doing this stuff in the bottom of the constructor rather than preRegister seems to work so far +Attachment has been added with description basic! Applied! +This would requiring interning of region objects and we arent going to do that +Similarly for currency and product there should be with available currency and product for reportWhat you guys think +hmm +the only part is that of substitution when trying to read a symlinkif FORRESTHOME is set you can run the script with the solaris sh shell +Anyway the issue is inactive since so closing. +Unless I am missing something simply deleting the build root will not allow this usage to work +Patches were too broad and have gone stale +Alan ping +This is a partial screen shot of the initial state of the application +Golden file was damaged so new was generated +thx +Ill try to create a simple reproducer if possible and Ill let you I suppose that its late to do any changes to codebase for current EAP release +Integrated the crash handler guard against original segfault in release buildClosing. +Making the logs transient would certainly fix out issuesas currently its a show stopperThis occurs with +Then it could override the rowcache capacity value for those sstables +In ongoing discussions with Juergen and Rob about the best way to approach given is in Acegi Security and the simplest solution is to use although this only appears in Spring and is therefore unavailable to the Acegi Security series releases. +committed intrunk rbranchx close resolved issues +Hi Rustyunfortunately you give not enough information for me to reproduce this +in rev +Mark it to be fixed in since this is a regression +Hi ClausSure thing +Please let me know if anything has changed since then +Looks nasty +As to the question you raised Im now not sure if this in fact an egg issue which is what I was originally told +This should never return null probably +Oops I think we sneaked this change in accidentally in another patch +Fixed in the ll backport that to AS as in trunk by you can download the fixed version from that JIRA it works fine with and will be included in +Fix is obvious +Patches committed +Please let me know what you think +Frantiek thanks for the report +So long Jacques positioning works in FF IE IE +I tried update and disable setting with latest and in both cases there was minute gap in server startup +Place the ICUJ jar file and license in the tools directory +This is something I came up with really quick +Is there any chance you could add a unit test that fails like this +These changes look great to me +It should free the String on close +I agree we could add some check point in the file producer to throw the exception when it cant generate the file contentIt could also be the common scenario for the other endpoints +Thanks for the heads up +committed to trunk +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Please see my comments below vmjitrinosrcdynopt contains rn line endings as well as n need to fix to mere n You are right the patch has been prepared on Windows +Integrates with the changes for this issueWill not submit to hudson until we have feedback from others who have tried this patch +Patch with the modifications described +This issue has been solved by setting resolve strategy of closure +Thanks for your workTill now the artefacts are not available in the maven repositories +Although I hate Sybase our company is still using Sybase because there are many power builder applications running +The attached must be applied after a similar set ofsvn move commands +Patches for this issue should be applied after patches for and otherwise it may break other pending JDWP patches +If there are any other changes we would like to make let me know and i can address them as well +We can be a little relaxed after we have out the door and that will give us a chance to reflect and work on the directory structure with maven +not sure why I like to write VERSION in all caps thanks! That will work perfectly for us +Please feed back your opinion +Thanks Bryan I see some of the same thoughts came up Ill look through them in more detail before submitting a patch +Thanks yongqiang +After all the sidebar is the only part with a hardcoded width and the main contents to the right of it grows with the browser window +Committed with revision +This patch is built against rev +bulk defer of open issues to +So I think you are in business +I think this should be resolved can you against upstream +How about this one? This has still everywhere +Note Java Server Faces applications that use ajaxjsf do not work with IBM JDK +Ive added the documentation for this now +New patch has all assimilated test cases moved to a new non conflicting packageAlso contains contribcache that depends on everything else +Keep open because need to add Integration test currently it is not possible to check similarity to the default full text search field of the Entityhub text documentation of the representing similarity is required to allow similarity queries on the full text search with +Thanks Michael at a glance the patch looks good to meI wonder if we can improve the test Im a bit concerned with random merge policies that it might sporatically fail +The system is quite straightforward however some more flexibility in the substitutions performed on the receipt template are required in order to get access to the additional data I need +deprecated support for mylead catalog in anticipation for future integration with catalog is deprecated +We have two competing versions of this patch the one in and a pair of patches here in +The test cases fail with the patch applied and pass when its svn and git ignore lists should be updated as part of this patch too +Reopening to set assigne and fix versions. +subversion repo +language settings NB is for testing localizationen nl fr a lot for clarification Scott! In an effort to look more into why Rave does not do this as desired I had a look upon the class which I think provides the method get to create a widget instance that is used by RaveHowever the class does not provide any method that allows me to also send the locale to Wookie? Or am I just looking at the wrong place +Just tried with eclipse plugin built from sources it I guess its easy now to check why it didnt work in +The CMIS Workbench also supports the Mac OS X key bindings +HiI fixed the failing test and committed the changesCheers Matteo +I switched from to and now it works. +Refactor all of this common code into a helper classIt could be much cleaner if there were functions in Java so I could pass them to it +Also Applies to resolved issues. +changed reported to delete Hals duplicate user. +It will get quicker over time I suspect as we fine tune which jars need to be loaded +This update includes code for using inherited attributes in CTA processingWith this patch update I believe all the requirements of the XML Schema spec with regards to inheritable attributes and CTA integration are complete inI have done required testing at my end and I believe this patch meets all functional requirements of the spec for this featureI am looking forward to feedback to this patchRegardsMukul +Sorry Bryan Im a bit behind on my Jira mail +Thats ok if you decide not to move to EAP GA then for the product we can branch and the make adjustments there +Closing tickets +Attach the patch with the on the trunk as well as branch Thanks amila for the contribution. +The same check is reequired for the Hover in Java +Read the comments at the top of the file for instructions +Can you please use tcpmon to check the soap requestthanksI am currently adding a single space for the soap action if there is noSOAP actionBrett +How about using python style offsets where negative numbers means count from the right +Adds Steves last contrib changes Steve! +Ok for IT was +jruby description from reporter. +Attachment has been added with description zip file which contains the required documents mentioned in the this bug is duplicate +The third option that takes a class should be an that the project supplies not a separate method +I have not tried the code we still run with indexing code but only read what you intend to do which looks fine +The patch looks good to me as is +hmmm +initMetadata split fairly nicely into checkHealth get and increment +But we really need some memory protecting feature +Ill try to come up with a patch with minimal internal API changesSupun +Thats a lot clearer and makes integration easier +Making username pattern configurable via added testcases for itAlso I did some minor corrections in the testcases +Added a tentative patch to the class +as you say we should allow customization of the Type easily done. +According to your previous comments the values of access and modification time are the same when the open file record is written. +Replaced the jar in thirdparty and updated the references to in the project and +This is a forward port patch from webwork +Addressed the comments including the last one +Pushing this JSON example request out to the release in order to have ample time to provide a useful and appropriate example using JSON in with SDG. +I beleive this is fixed by +Merged into the master branch. +i can see it in the Attachments section above. +Thanks Javier I just committed your updated versionCould the live in core? It does not appear to depend on tika +Reason for no test case just naming the thread +Attaching a new patch that fixes the findbugs issue +If the hosts have access to the internet there is no need to set up a local repo +Maybe everyone was just adding that during testing +The patch is generated on top of patch and hence should wait until that goes in +When localisation for an authenticatd user is performed on the server the locale is most naturally associated with the user +When clears we can look into actually hooking it up in +Delete from the lib directory +Thanks in advance. + +You can close this issueRegardsRaphal Ouazana +Otherwise it may wait forever if the interrupt is swallowed somewhere +Keep it on the radar +Thanks for the patch Paul. +Help in resolution is appreciated. +Bean id refactorings also change references to that bean id from within elements. +I called and was asked for all my information +We tested the too with the same result +Resolving issue since no serious problems occured. +Hurray! + problem is visible in demo with accentuated characters in SEAMTEXT fails then and rerenderI have still the problem with +Im sorry I didnt see that one +I care thats why i am to the builder pattern +Taking Im working on a fix for it now + +One more option what I can see is How about not persisting block along with allocate block +What meaning is behind of empty? I have couple options empty at all xml with root node only should have a head and a root node so creation of empty at all file doesnt have any sence here +Is this still the case on trunk + default is closing stale resolved issues +Im pretty sure this is caused by since we compact at least files now +I see another probelm +Fixed in r +Patch to solve and failing tests +Also applied to fix a bug in edit log syncing throughputHow have you set up the? Is it a single directory or multiple? Is it on a disk with an write cache? My test was on an SSD +Submit job now passes in the jobid +Hope it helps +Patch the removes closeCleanup from implementations remove any field storing the reference to a close cleanup method remove any calling of the method remove passing the reference to the method in the constructorWill commit soon tests all passnext step will be to remove the passing of the null reference into the interface +Another quick start that illustrates the issue +Closing all resolved tickets from or older +We can skip the problem just do not adding the property into the component but rendering the attribute otherwise it will work only with jars +I also used to help make it clear that the native profile cant work on Windows and the profile cant work on Linux or Mac +IMO this is not a blocker for it can be completed in +Currently only uses constraints impact candidate choices +Ioannis have you tested with Amazon Linux AMI or custom one based on +Can you add a negative testLook at qlsrctestqueriesclientnegative for an example +means reading foobar httpnamenodeportdatafoobaropGetread +Passed in last runs +bq +If you make a transformation that makes new variables after is done it is your responsibility to do it right and enter the variables in the scope as well as link the to their origin +But heres the problem this patch and my example cover both of these casesI intended to make two cases mutually exclusive +keepalive patch against +I created a simple process but it deploys and runs just fineMany thanks you try a newer version of? I used during my testing +Committed almost as is +Triaged for +That would seem to be outside the scope of a +This will reduce the pain to maintain backward compatibility +then you are going to acquire and release k locks in removeWatches for every write operation +I know of the usage on shell I tried that as I explained in the last comments +Below is a listing of the differences but I do not think it is worth changing the video for this +is there a workaround or a design pattern that will avoid diamond and scenarios without stickiness +Simply call mvn clean +The mailing list is the better place to discuss and reach a wider audience +view EJB beans you have to specify for it to work +This patch updates Nutch to use Hadoop from +The patch doesnt appear to include any new or modified testsAdditional test cases will be in +Bindable is only called on root objects not subobjects +This issue has been last updated more than years ago has no assignee affects an old version of Tapestry that is not actively developed anymore and is therefore prone to be in the near futureIf the issue still persists with the most recent development preview of Tapestry please update it as soon as possible +Is this an issue with the tests environment perhapsIs this test intended to work only with enUS +Daniel please open another issue for that +Dont know what the Hudson failure is from +Since this is really a broker related issue Im removing the NMS fix version +is that normal +Id prefer the whitespace being collapsed in Tika instead but it was for me quicker to fix in in Nutch +This fixes the build which is currently broken +Testprogram that triggers the problem for me +I could not find smaller reproducer +They can be etc. +Thanks Owen I think this is ready +Same patch +This of default namespace akin to the default package in java +Thanks Albert +Rick reviewing old +There is no such class a in guess is that youre trying to use override the jaxp classes without usingthe endorsed mechanism as required by specBut I shouldnt have to guess what a reported bug isso please dont this unless you can explain how to reproduce it. +the correction at rthanks Olivier +Ive solved the problem by pacing the settings outside the method +I cannot recreate the error +Hmmm or did we never port this to Weld? Marko? +NET things needed there as I understand it +Suresh any updates on this +The nominated bug was fixed in r but we still need to get this pushed into the regression testing because it is not systematically testedLeaving issue open +This is a pretty critical breakage +Thanks Chaitanya! +Nice work +I think this might be an issue with FB rather than Java +Updated patch to fix wrong placement of null check inwriteTest still fails +Two comments +Having the hostname anad port is redundant information in the normal use case since this can be found in the browseraddress bar +This still need to be revisited in the future +Plus compactions are still async +Without patch cannot listen for new +That way others will be aware that it is already being worked on +will commit if tests to trunk +Snowdrop User Guide section +Please consider incorporating this patch in your next release +Daryn the patch does not apply anymore +Thanks a lot +The script is invalid and thats why you see the error +I have replaced all tabs and removed some problematic features applie only on I removed this last patch +Reassigning to Gary +Verified in CR. +Ive already done this with a couple others if you lookThanks I added it to the junit task invocation in the definition +Si the intent is still there as it would be a much better UI experience but its not top of my priority list at the moment +I think this is actually better organization +The method is called get and its a load without remove methodThe unit test has been updated to provide a test of the get method +This requires no new tests as it changes only the internal implementation +KiranCan you please attach the entire mgmt log files for furtherI have uploaded the Management server logRegardsKiran +However I just found a bug with the aspectj mojo it cant resolve dependencies +The latest release is still not working with bubbling version since the bubbling plugin references the wrong version of the library +created +Heres a good version +Short explanation followsThe utility class holds the algorithms to alter the generated schema operation requiring that we loop over the XML within a document and perform some change or data harvest is realized with a implementation +Yes this is committed and available as of +Patch fixing this issue in a simple way +and are in that directory too + you reproduce this on Fedora +Moved from to just after merge of old branch into TRUNK +jsvc references seems to be fine to me too for it is an integral part of +Linked issues which will be addressed by rewritten module importer and top level handler dispatcher +It had mostly been resolved before but there was one section on Classes and Interfaces that got structed +Please add your detailed comments if you wish to make the case for reopening it +Created +This issue is a duplicate of which is marked as resolvedSo I presume that this bug is fixed in the latest release +Hmm thats strange I didnt get a notification of your first reply but I did get one for the second +This patch avoids internal server error msgs that comes when obtaining subscriber list and it also modifies the mep of the addsubscriber operation of subsmgr to avoid internal server errors +Awesome work Strong Liu! I compiled it and Im doing the testing now +Yes it is odd but still true +Do you think it be more ideal to create a artifact instead of the tests artifact? That would seem to more closely follow how the other sample junit modules are being deployed +The user can control opening the Central editor with the Show On Startup checkbox +The problem without stripes is that the first mb keeps getting even though it never changesAm i understanding it correctly that stripes in a region would act similarly to regions in a table +Appears to be fixed +Werner you already had this greate idea about month ago +java both in classes and dev no app starting fails today at all +However the WMS client can also send the same filter inside SLD if it is using SLDBODY system but that is not the problem you are dealing with +Ill run some tests with larger sets of files to get a sense of what the exact limitation is when using a larger heap sizeYou can only have so many unique fields when norms are on +I just had a look at the pom and it seems your patch makes sense We use version of the osworkflow but somehow we reference from thexml file +Hi Suresh Thats very nicely suggestions +I like the idea to store uri in the token which is issuer of the token +Umm delete the comment about readToken always returning null +Is there a reason why this is not marked for +Another vote +There is just a high chance to pass such a test with a correct implementation +If we require a discrimator for it will require the update script to be run if you upgrade I dont see anyway around thisE If you read the updated documentation theres a section on indexes there +I think someone was working on adding support for downloading EAP runtimes which would also presumably need the credentials. +You can just call the conversion script by going into the directory you extract this into and issuing the follow ant commandant pathtomyProject +This +Do we know if there are any other places where we consume one stack frame per segment +Issue has been resolved for a long time waiting for potential confirmation from the reporter +to trunk +no need to merge this into the CP +The xml file does not make this as easy to accomplish +remove it and see if it runs everything +Just some trivial cleanups +Gunnar thanks for fixing +No existing code is modified +Full logs attached +Can you please attach a new patch that can work with the current trunk +Appears to be fixed but not closed during the +Thanks! +This needs to be with +has been marked as a duplicate of this bug +The last test failure seems to be caused by the test flakiness or other reasons not related to the patch +Not much that can be done to speed it up but it could certainly be run as a nightly +The data flow is fundamentally different and in addition a non parallel view server might be managed differently from a parallel one +That case will be reproduced if user selects nothing in the step +There is usually a lot of config in there +Source code and of the first submission +I also added the comments sijie suggested +Can you confirm that the unit tests still pass in this scenario? If so my to commit it +done in issue after bundle release. +to the approach it seems reasonable + +Im going to commit this and then shortly if there are no objections +So if for some reason the handshake process itself generates +Its definitely one of those broken windows that when fixed really helps the overall sentiment. +It is challenging to support href with pull model +Thanks! +Pavel sounds great! your comment deserves being in the codewill commit soon +svn diff of is what should happen I believeMade after applying patch +Needed to get the speeds up in indexed hbase +So a test case reproducing the problem would help a lotOleg +Currently I do not have one +FYI a new Sonar Violation Filter plugin is going to be release during next daysweeks +That looks like the same trace +I located and corrected the bug +Adding a test case to would be good +Thanks Dimuthu +What happens if a system administrator uses one VM to deploy two embedded derby applications each of which sets to a different location andor generates custom? Only one of the applications can win +after click any other +Thank you Manuel +Set the file to null and it will be volatile. +Comitters can you please review +Can you please rebase? Thanks +yes we can include this in +The is bit flacky The patch for does contain a fix for that +We cant close connection handles for you due to some slightly silly requirements of the jca indicate if you are already closing all resources in finally blocks if you are this might be a bug if not it is a feature request +Sorry doesnt seem to work always +We could return OPENED for the caseIt can actually happen during transition to failedopen the thread only sees region in rits it doesnt know what open handler is doingWouldnt region be stuck in this case if master assumes server will finish the assignment +ignore that last comment added to wrong issue +bq +Timothy I am not sure if I understand you correctly +reviewed and validated that javadoc warnings are fixed +Ill dig +Thanks for responding! The changes to have been checked in. +You can use the zip from linux +Should be closed when Jira duplicating this one will be closed +Passwords can involve any character +checked in. +This is either a configuration issue or an Oracle specific problem +Maybe Im missing something though +I still sleepIt is clear now +No just hat its harder to fix +Hopefully I have fixed this issue +again for the preternaturally fast response RickCommitted patch to documentation trunk at revision +subject to jenkins +Done in r +Hi EikeDid you specifywicket as a dependency to your project or you usewicket contains all classes from core util and request in one +I will add a standalone ready to run distribution. +Reopening until that issue is resolved +Thank you Ill check them on linux +I also ran javadoc and ones +Closing issue because the plugin only supports was officially +I prefer a single implementation for clarity and code maintenance reason +Thanks Mauran helped a bunch but I noticed sometimes will return a null even though the item in the actual PDF will have a page reference +When the application starts does not throw out a +it does not look like this is a practical problem +It may be desirable at some point to add similar logic for inlined lobs. +DB for instance does not return the generated keys when using executeBatch +Can you do that for branch as well? If not let me know +bq +So maybe we could also look at HADOOPCOMMONHOME which is the lib set that has the +need more input to help it must have something to do with the environment need a testcase to replay issue or at least more info +Ive committed a fix into SVN trunkCould you give it a try and please reopen if the problem hasnt been solved +Probably it is possible to find random workarounts on client JVM but I am afraid we can break more things this way +patch provides a that exposes the bug +I am on attempting to make this class threadsafe +the revision Thanks +I committed the patch +Patch integrated to trunk. +Updated patch which was necessary to make the CTF test suite run successfully. +Use the upstream build once the fix has been applied on EAP ER +Minor refactoring passScore methods now allow access to more data +Dario I have given it some more thoughI think you are right about bullet and in that sense I would like Camel to support other file transports than just +Confirmed +Sonce nothing has been committed so far I will start to work on this issueby reassigning it to myself +So if you enable logging on youll see the SQL before we try to prepare it +Thanks for fixing the test cases +Everything compiles and tests all pass +A couple of questions Which library from where must be downloaded to make this work? I think Adrian has some good points +Committed to trunk with revision +Eli you mention Hibernate Core in this issue and Infinispan on twitter +Therefore should not be relevant anymore +Committed to trunk and +Can we close this? AFAIK we have no remaining legal issues and having this block a release bug for something thats already out seems silly +If we can get this in quickly I can pull it into the that RC +what will be the fate of featureI think we should still keep it as an expert feature as it can be used to replace the implementation of the jobapp classloader itself in rare cases +batch transition to closed remaining resolved bugs +Please confirm once new rbovirt gem is available. +Committed the latest patch on behalf of Myroslav. +Thanks +I have not done it yet but I think it should be done by adding exclusions to war plugin configuration +done + +confirmed that I have seen this with keycachesizeinmb for every node in the just migrated our cluster to here and were having a similar issue when running repairs +Here is the patch to change screen into portal screen +Can you give it a tryThanks GailIve retested in trunk and confirm that the change fixes the behaviour with the Ingres dialectThanks in trunk. +Stefan please comment +Im not using maven +I know Hiram Chirino had some ideas about this when I talked to him about stopping processes from java +Ive configured commit notifications to be sent to though this can be changed if people objectAs the moderator for this list I object +updated the previous patch to the current reality still incomplete +Id etc +Assuming this is fixed in the live environment since the test case passes +I want to do this to learn the process for making contributions +And interestingly suddenly this is moved from MAPREDUCE to OOZIE +one of many related issue for description of adding JPA to New Seam project has been removed from Write Your First Seam Web Application chapter in Getting Started guide screenshot updated +This patch contains a sample SCA service that expose DAS functionality and a client application that consumes it +Fixed in trunk revision +Ill poke at it more tomorrow +I think this should work as is now +It seems like it might be a bug in the server or somethingDoes it happen every time +This patch addresses Dougs concerns moves the common code in abstract base class implements reflect serialization based on ReflectgetSchema +Remove natives from the installer +This is the patch for the header mediator to remove and set the wsa headers to the message in the above configuration +Putting this info in Troubleshooting would help a lot of people +Of course this can also apply to but essentially you are chasing a memory leak right now which make it hard to assist you +bq +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +in fact this is true +Sorry. +Made some code cleanup changes into This will make code easier to understand and maintain +This means that under eclipse current trunk seems to complile correctly +We could request a new repo from apache for the echo plugin and add it to the mobile spec dependency plugin to get installed +The upstream code wasnt port to TC + +Hi Mariusz sorry will need a bit more detail +So I update discard previous and patches +How do you manually recover from this +But I dont think the JVM has a negative class cache does it? That is if you try to load a class when it doesnt exist then move the class into the classpath and try to load again it might pick it up +no. +Hi Ming could you attach your patch to this JIRA if its available? Then Jenkins tests your patch +itexpression is the prefix for the plugin as given in its plugin descriptor. +on both patches +The latest patch looks good to me +committed revision two response writers +Thanks JakeI would like to ask the apache build infrastructure team to install these packagesBut I have still have a question What about a packaging system is there something like rpm or apt +Use dbsetup and +By works just fine over JDBC you mean that through Teiid JDBC correct? That can be because either you are pulling the value as timestamp or string without any conversion along the way +Patch applied at +r on trunk +I think the issue is that varlibhadoopcachehadoop is not group writable and is owned by rootThe directory should be created by the namenode format command +Aaron HADOOPCOMMONHOME and HADOOPHDFSHOME are the parameters that Owen proposed to remove for production deployment to prevent adminuser errors in setting those environment variables +The release note indicates that the new behaviour is correct but doesnt it still allow a pattern to be used which does not conform to the standard +It didnt work I guess is not yet there Ill try again tomorrow +At least by trying to sort the Strings we might end up with a somewhat consistent list then take the largest one +The files located in the archetype jar belongs to one of these categories binary file velocity templateDuring archetypecreate binary files are copied and velocity templates are processedDuring archetype binary files and velocity templates are seeked agains the filtered extensions defined in the registry or overridden with this still an issue to be adressed in the code or is this only a documentation issue that could be linked to +They will not request or use tokens so the secret manager will sit dormant +It might be possible to test it using some or something +Unless some resolution rules automatically map to the table in the from list +you mean enable folding in xml source part +I added a self healing mechanism to the parser to fix corrupt offsets within the xref table in revision The parser seeks bytes backwardforward if the object cant be found at the given offsetNow there is an issue with the length of a stream +I have updated today +Attached patch solves this issueRegardsve applied a modified form of the patch +The current trunk no longer adds the +bulk fixing the version info for and all affected issues have in +This made the most sense to me putting things back in the state you are in when logging into the CLI and only when you have just deleted the keyspace you were in +from me +Attaching new versions of patches merged over fix +jledit has been included in +Anyone who would appreciate further backports however should let me know +Im looking at it to see what I can do about that! Great find! Ill pull your request into jffi and look forward to the sparc jar. +Errors out of order in doesnt appear as though this is fixed in +There will be no release. +But I think this could have a negative impact on performanceGreetingsChristian +Cullprit it expects as value however model is populated with local should prbably be improved somehow to make it more strict it has double values for transaction typesxa xa local local The second value seems to be mapping to JCA +NDo you mind submitting for another QA run since the previous attempt didnt give us results +Method complexity is very important and arguably more important than class complexity +This problem occurs for Solr not for Solr +And I was disappointed there is no easy way to provide custom impl +Can someone please resolveclose it +key comparisons are only useful for OPP which you shouldnt be using +Hi Hans Bakker Im missing some filed for localization. +Committed +Hi lufeng this is great that you for taking this one on +Done +Better schemes to be implemented in future +sorry for prev comment +Closing all resolved tickets from or older +This is the current behavior in geronimo which after making this investigation I think should remain intact despite the possibility of allowing a harmless NPE during the small time widow before the socket closes +Does anybody know what the concrete work would be to solve this issue? We struggle with this in +Committed to branch +A few notes has a pigscript and other files which test registering udfs and parameter substitutions inside macros to demonstrate how this worksI havent been able to upload my diff to because I generated it with git and I think RB wants something + +Fixed in closing stale resolved issues +it works on linux too with your patch +patches added +I integrated the warning. +The syntax in the description shows spaces surrounding the sign but none of the examples in seem to have spaces +Committed to the branch thanks for the review. +Thanks for the patch Marshall +I agree that this behaviour is by design +Why not reuse it? Because most of the time this is not done by the user and resizing the buffer is what takes much time +Should be resolved by r. +I agree this issue can be closed +Will commit this patch shortly +The attached patch contains an updated implementation +Do you think it will address? I havent had a chance to take a close look at that one yet +Initial patch for review +Original reporter unable to produce which data caused this error +Default is false therefore if anybody wants to change classloaderpovider to it must configure it +This patch avoid to create iterator to prevent throw +Nicholas shouldnt be affect by your patch it has to be something unrelated no +first its too late to do this in as it is now in maintenance meaning we only fix bugssecond i dont really like this +I echoed my MAVENOPTS in the attached file but for simplicity it is set to Xmsm Xmxm XXm XXm +When we say when running in embedded mode I always wonderwhen what is running in embedded mode? These clauses do need a subject +Can you comment on this +since we had a seq id index in read ahead cache it would very very useful to remove entries from read ahead after consume +closing in bulk all issues marked as Resolved +hope Ill finish it on Monday . +You need to specify scopeprototype if you want to reinstantiate independent instances there which in this case simply means calling the specified factory method for each access to the alot Juergen setting the scope on the to prototype workes out fineI wasnt aware that scopes could be used for factory instanciated beans as well although it makes perfect sense since is just a bean definition too +Now that work is well underway with AS all previous community releases are +However its hard to figure out based on that whats gone wrong within hornetQ +Any other suggestions +Scary thing is that only a single unit test caught this +I feel like it generates a lot of noise for little concrete benefit +look at time index are two problems in the code +Theres some stuff in trunk now for but would need rejiggering to make reusable +This is a safe assumption for any code in HDFS to make and does not violate any abstraction boundaries within HDFSAn exception is justified when the best handler for it is unknown +nice! I did check java version and its bit hotspot so were good +svn ciSending examplessrcmainjavaorgapacheopenejbexamplescounterSending examplessrcmainjavaorgapacheopenejbexamplescounterSending examplessrcmainjavaorgapacheopenejbexamplescounterSending examplessrctestjavaorgapacheopenejbexamplescounterTransmitting file dataCommitted revision . +After checking the document I found Specifies the maximum number of concurrent unauthenticated con nections to the sshd daemon +Seems to work in incognito mode meaning that SBS was caching something wrong for a long time +fixes all snapshot tests except that some tests are temporarily disabled +Shadow stack area support was added to OPT compiler for internal helpers it calls +and +Ah I do see the pull requests +Question will the current infra Git tooling automatically pick up pull requests on Github for OODT and send those cool email messages to the dev list for them? Or is there something else I need to request for that? Thanks doods +Done in for QA to do here. +At a superficial glance your rewrite accomplishes those two things +Committed to trunk at r +Turns out I dont think this has anything to do at all with usersroles and everything to do with deleting and reimporting JON Agent resources +It would be good to update the site with a better version but the file is still a moving target +Uploaded wrong patch +Thanks mike for looking into this and also for the valuable suggestions +It then continues to call the step handler for time with value which is the value calculated for time from before the previous time the step handler was called +This is broken for me with an English locale +the id of the snapshot +This is the same problem as which does have a reproducible test case. +To be absolutely sure it is a Spring issue I rolled back to spring and its now a side note I also upgraded aspectweaver from to during this same upgrade process +What are use cases for that index if it forces me to ALLOW FILTERING to use that index? +Should we fix this +We could then overload the Java methods and introduce new methods on the C client sideWould be a great project for a new contributor +o files in objLINUX directoryI dont have ICU on my machine IMHO dynamic linking should fail if my app would depend on ICU +Can we include this for? +Unless there are objections I will commit this shortly +package contains the original test case from +This will improve performance and avoid this problem +My two cents +Yes +Logs non in the thread also tries to stop execution early in case of such errors +I did not change any of the code. +He was seeing hfile corruption when region servers were crashing on heavy compactions +Also Enebos comment about IR being beneficial to Ruboto means we will want to keep it +Ok I got it +The changes broke some regression tests uploading new patch that fixes these +Defer all unresolved issues from to +sigh +Ive submitted a patch that does basically the same thing that Chriss does but only affects +I will open a separate ticket to enable the Shindig dist to be svnpubsub +A response was made about marking it experimental +Sure Id be happy to +This change has been tested using the test written by Hiroshi manuallyThank you Hiroshi and Jing +lets see how tests do of these tests pass on local +by the way ive disabled UAC on Windows Vista now im working as a true administrator without any security limitation +Ill take this m testing the patch on the JEXL site +Will resolve commiting in other branches also +Patch for replacing the usage of deprecated classes byNote that i tried to minimize the number of lines changed instead of optimizing and passing Resolver and Factory classes aroundPlease comment +To fix build of VM gc and are required +Committed to and +Then consider a certain limit and increase the index interval as necessary until the estimated index size is below threshold +Im attaching the patch for this issueChangelist updated pom to have the bundle registered as a fragment for moved the from into another package deleted duplicated file from +Permissions adjusted +OK I figured out the other problems there seem to be some classes calling others in deserailizing +That was kind of a typo +Committed with revision +A simple workaround is to use the instead of the regular +please use the latest upload +Interesting issue +Id like to commit this patch on Thursday so if anyone has comments please add them by the end of Wednesday +Per the lock order is JT TS JIP +Its has been too long since Ive worked on this plugin +Closing issue out as part of JIRA cleanup +Commit away +I have attached a patch for the improvement +If there is at least one callback url to a page then it is stateful +The point of Category B is in order to minimize the chance that a user of an Apache product will create a derivative work of a portion of an Apache product without being aware of the applicable requirements +Some discussion regarding permission cache can be found in and +We can decide later how to resolve this +Your provided fix solves the issue as far as Im concerned +Let me know if you can think of anything else I can look at +Thanks Jacky good catch +Hi NandanaI dont think that garbage collection is an issue There seems to be enough memory and there is no noticeable pause for garbage collectionFurthermore I dont find any easy way of deactivating garbage collection +Build problem? It seems that we got revisions number for windows and linux for packages +just committed this thanks Terence. +Still passing +Weldon I was cranking throught these saw this sitting w your name and to me +How about we try to follow the redirect on the clients behalf? If we succeed we can send back the as before if we fail we pass back the and the client itself can redo the request and analyze what went wrong. +on to branch and trunk +The attached patch adds a test case that verifies that the WHEN clause SPS is invalidated and recompiled if one of its dependencies requests a recompilation +Nicks rspec framework is in trunk now searches for srcspecruby +Should be fixedwontfixed by I think +All tests pass nowI added an entry in under Changes in runtime behaviorincluding a NOTE for users +DavidThanks for that! I hadnt looked in depth into the plugin I just wanted to note its existence +The existing handlers arent really set up to do this because the file name is already expanded by the time it is set +No need for IT +Aggregate focusblur event hadnling muct be handled programaticallySimilarly the key events have specific actions for the picklist component moving the selection up and down etc + +Pull request merged +all machines should have firefox installed now ive added all the locations of firefox to the script +Erik do you mind taking a quick look please +Now should be fine +Change GREGOFFSET value to match the description +damn this uncovered a bug in the tests and fixed an unncessary throws declaration +I entirely agree +Done and available in Branch to trunk and branches need a test case that documents and tests the desired resolved issues for already released versions +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Can you please hold the commit +I like the approach that this is taking +Probably it was done by accident +Alright on the patch here committed. +Just a quick note to say that the conjugate gradient attached to uses this framework in case you would like to see it experimentedYou should probably commit this new feature if you think that it is in a good shapeMaybe that people will come up with further comments once they see it at work in the unit tests +Thanks Kim for the revisions +Patch attached +ugg at least its not Windows +Thanks Edison I think we can close this ticket then +But I would like to know where the problem is a revision number would be nice +It holds onto an instance of the builder factory and converter factory and such so things can be completely configured +ping +may be Im wrong +But Eclipse XML validator complains that it is not correct regarding to the corresponding look ok too +I attached it now to this issue +classpath udigtrunkplugins udigtrunkplugins udigtrunkplugins udigtrunkpluginsLogChange over to for uDig release Maybe the bug was in geotools thus the Revision has solved the problemI think we can colse this issue? +Trunk is not a sandbox and getting more developer attention is not a good reason to put something in trunk +problem is in joda got a user report of error trying to enter PM in a and it led me to this issue +There should be a limit in the config file +Yes sorry I ment latests Eclipse +would clean the asset +Im using a non java client and I dont know how I can do debug or the tools that I can use to give some information to help debug this issue +I am currently using an internal version of for testing +The backport improves the content and significantly improves the presentation of commonmapreducehdfs files +The documentation says that you should first just run the testsuite and if everything works ok you do not have to do anything +I dont think this is a bug you essentially moved files out from under a running instance something cassandra cant expect +I also have a depedency on +IIRC it wont wok becuase several other places use wo closing the conneciton with so the count is properly reduced every one of these other instances would need to be updated at the same time +Does anyone disagree +Patch applied to revision +I am using git +We need to make sure our build processes wont be affected by this change +Actually this test can be made into a unit test +But the above is fine too +Added more comments for the test case +moved to does not have to be a dropin replacement forwe have changed quiet a bit of apis already +Committed as is. +No response from submitter I assume this is fixed though. +In particular it seems to work fine using Spring with pretty common setup +Ive implemented chm extractor based on the same extraction algo as chmlib +When the name node receives the error report it lots the error and removes the data node infoA data node detects disk problem at startup time when it receives a rw request after it receives a command from its name node and before it sends out a block report +Now everything is all right +Ive decided that these problems are actually the same +Im working through Camel maven build dependencies +Looks good +There is a possibility that targets are defined inline and then there is also an explicit wire element that is defined +Seeing the sporadic failures again +I just finished rebasing this for trunk but Im going to wait to repost it until I add proper support for importexport +Max I think that only works if you are focused on the PROBLEMS or MARKERS views +example of a jspinclude to a jsp trying to use the actionForm implicit object +Regarding the link the problem is on if the request uri comes as static logs or cluster which all return the pathInfo as +Adding an may be one solution for accessing +Duuuuh +After reading your email a few times I really dont understand what some of the points you made +Yes it got pretty clear now when I reread it +based on Snjezanas comments this was done ages ago. +Oke clear +Thanks for the feedback +you are right that this should be a parameter +Thank you for adding this extensionI still believe that invalid subconfigurations should be treated differently from optional subconfigurations regardless of the invalidation cause and that the silent ignoring should be a switch that needs to be set similarly to the attribute +bug caused Hudson failure +another strange issue in the WSRP interface +Then suggest we look at the two solns and if fb has anything we should forward port bring it forward in this issue or else close as dup +Any improvement in this area with release? +I think for these you can just start committing +The best would be to create another JIRA where you explain that the Delete Tree control should not be used automatically but only if the user either specify it for a delete request or if its set in a preferenceThen you will be able to close this issue and make a link between the new issue and this oneBut you can also prefer not create a new issue as this one is pretty much self explanatory +This issue has been resolved Builds verified working build build +Ive removed the empty file at r. +HelloI want just to better understand this issue is being when using named parameters in my sql I should use? Or will it be deprecated to? when I dont need named parameters yoyu confirm I should just useThanks for the info +A subset of problems has been fixed automatically but it seems to be a class of problems with splitting regions that isnt being handled properly +latest patch looks fine +I tried with mojarra and it is working +updates patches patch patch and to revision +It has something to do with an extension point with an invalid name +A test case is needed to verify this +I think we can close this since nobody showed interest in more years +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Heres a patch from Bill Baggett for this bugIll test and run it and then if it is all ok Ill commit it +Backported to. +metadata +not sure if this is related but I thank it isI open a issues about leases hereAnd I have patch applied to my there is an issue with leases well track it in on Hudson. +for the stuff about clearing the stale connection inside the +Servlet programmatic login will be able to reauthenticate +Marking these as fixed against rather than against. +We wont ever have the AS ligature and just since wed only ever reference the project as Application Server +can we get a few more comments in the code on the overall strategy of the test +Oh and sorry I didnt notice your question about left and right single quotes +Oh well Ill try to make sense of it +Our slapd config contains anon authOn many if not all the ACL entries in our config +The patch failed these unit testsThe patch broke on Jenkins +Only set javahome if it doesnt already have a value +patch some reason I cant apply the patch but again suspect its a local problem +Jurgen as Galder said wed be happy to accept a patch +As suggested by Simone sorry for not understanding the way before +Replaced slfj with jboss logger. +See discussion in +I can provide more but figured Id hold off on that until asked due to the size of each one +Please add details of which processes are using the ports in question +can you please attached the wsdl +updated docs +It seems that its a bug of +The pop up should also show the secondary IP addr +OK thanks +No the results do not occur when the proxy is disabled +If anyone knows of a good reason not to change the default to cachelevel let me know otherwise I will change this soon +Thank you! I used the true solution. +yes that is almost true +I think it would be good to add the assert in the other places too +I have traced the problem further now +This issue has been fixed and released as part of release +Were looking to do many similar things and it would be really nice if we were working together so that were not replicating work +It was only added as a defensive measure +I have a business Object that implements Serializable +If it is not then it would help to actually see the exception +I replaced with which opens standard are not used any more +Sorted out by Jinathan +I got confused by the empty main jar artifact that didnt contain the base classes and decided to solve the problem by wrapping up the test classes into their own artifact +Committed to branch and trunk +attaching sample screenshot of can you please elaborate what are you truing to do with such project organizationIt doesnt seem your example is relevant to this issue because I do see version in child pom +This fails because the CI runs on and that has a different iteration order of maps than and the shutdown hooks are in a map which means they dont have a predictable order +Too bad since I have a project which should go online the next two weeks which seems a bit to close to use a beta release. +I guess the main problem is that CXF tries to access the bus from a thread local variable +Removing JON and CP from fix versions since it affects the embedded console not shipped with +revision Committed latest patches +Actually I guess anything a ith or below causes this +It is probably an oversight +If you downloaded a binary that was built without pthreads where did you find itIf you downloaded a source distribution and built it without threads can you please provide the runConfigure arguments you used? The r option should be none if you want to build without threads +there are times I dont want to use slink but do it just because it supports action would be nice if it has some conditional attribute like false default to have read posts on this issue but not aware of any solution or way around it or perhaps there is +Thank you Daniel. +The compilable examples should ideally go to the examples subtree so that they get distributed with binary distributions as wellPatch committed with some changes +Duplicate of! Thanks! +Well implement this for +Lowered the log level forThe problem occurs due to the fact that try to read only a part of an artifact from the OBR after which they close the stream +Hidid you run a mvn clean install before trying to resolve all artifacts +or always overwrites the bytes of same in case of Clob setString or will replace chars of same length +I deleted it as part of this patch because it was a subset of the functionality in +Logged In YES useridThis problem is resolved with Xerces there are then a lot of deployment failures with getting in the paths instead of spaces +This memory would be less than the data structure needed to be held in memory for only field cache entry for sort +Do you disagree +Starting at issues +In it is possible to properly remove from Introspector only classes that relates to the currently undeploying application +Nice patch + has what youre looking for and yes it is supported by INFRA +TedChecked it on a mac seeing the same failure as you +Ive found that the error even occurs when does not refer the other classes at all but disappears when they are deleted and reappears when they are restored +I prefer to use all of them to check all possible old formats especially those with no version supportAll tests pass +Our developers create eclipse plugins I just mavenize my case they create a eclipse plugin with no java source fordocumentation use to generate all +lgtm +Both should end up with the same +Please submit a patch for two command line options to Sqoop +Passed all unit testsI have committed this. +PHPMD repository has already been cleaned up and improved the big work is now on the PHPCS rule repository +Now there is no segmentation fault however the contains a following errorCannot dynamically link to usrlibusrlib cannot open shared object file No such file or directoryService exit with a return value of Is there some missing dependency +The eliminated canons have not been tracking changes to this test +if tests pass +Depending how fast you are releasing this will shift to next versiondirk +on the latest +The great thing about OOP languages is that you dont need to rewrite entire classes to support new behaviour +Let me know if it needs to go into branch as well +The notification scheme has been setup but apparently not been assigned to the ACE project +This works only if no cluster is bug report is invalid +This is really something that should be happening internally +This sounds to me complicated perhaps my knowledge about http core internals is not so heavy +Thanks! +Moving to Arquillian instead. +To clarify I am referring to the jars under lib that should be deleted along with the +The suggestion you provided for including transceiver count for decommissioned node will not work +Local testing of changes in branch show not only can the feature now be installed from Central but it works as Max suggestedSo Ive applied the. +pushed in Developer Studio +I appreciate the hard work but Im swamped at the moment +bq + +Heres your patch with some modifications +I know Ben at Atlassian has been looking at using it for API diff docs +Robert please help me +Here I send you a tomcat server all you have to do is run it! but previous you have to configure in conf to correct the address of the hsqldb scriptThe context is gvtservicesuseradminpassTry the Configuracin menu then the Organizacion Agregar Option and add an organization! you will see the radio buttons works well accordling to the value you gave them! But after add try the Organizacion listar option and try to update any record and play with the radio buttons value and save then go to list again and try to update again and you will see that they take values you didnt give themThis only happens with subflows I dont undertand why it fails! if a try to map the values directly into action they works well in subflows but I dont understand why it works well when its not a subflowIts clearly a bug! If you need the source code I can provide it to you +Update dependencies table +Any chance of getting this patch incorporated so I dont have to have my own forked build +Thanks +Created +A question though the fluent methods work well now in the base classes so why do we still need them in the individual drivers? If we dont delete them weve got more duplication than when we started! I think these can be safely removed now +The only file required to run this application +Yeah! I like that idea. +Where are you envisioning the weak reference would be used? +We should also fix the edit log output logic so that oversized ops cause an error when they are being written not later down the line when theyre being decoded +There is no maintenance as such no creative input +small patch +changing name to make sure hadoop tester pick it up +Sure rebased for +Our cassandra testsuite can reproduce both +I have checked past test results and this test has not failed since December when we moved the vmware images these tests run on to a faster server +Will not removeNo point putting more effort into this APIWill revisit this entire API after. +will fix it in the next coming patch +this way they will never get applied +Please reopen or send mail to the list if you have more arguments +is used by +The URL for this will be attached when known +definitely it is I guess it may have been while doing FR of some testcases paths +Looks like XHTML or HTML validation +Right now one node had of them and it took minutes to read all the logsThis seems a bit excessive +full patch to filesorghibernatecfgorghibernatedialectorghibernatedialectorghibernatemappingorghibernatemapping +Assigning all documentation related issues to final since they dont represent a change to the fine to me looking at the code of Grails lets close this. +To enable control set Configuration Properties in the Inventory section +I think the need for inheritance would need to change since this utility would be mostly used by operations types +See my comments in. +This sounds good to me would you like to write a patch +So I reverted the DTD change which fixes the invalid comment. +Forgot about these sorry +Im not sure in what kind of situations this kind of clients are used but if a node does not participate in storage and has been dead for an hour there should be no harm in removing it + No I think a better solution would be automatic registration +Possible fix At revision Thanks Darren +Have also a look at +Added a patch with the implementation of the dirty check a verboseIve Quality This is the first time I submit a patch for so please let me know if I should do something differently +I reopen it since automatic upload of assemblies did not workI think that the problem is in the release profile +Virag did you have a solution already for this? Otherwise I can start looking where you mentioned in +I think one way of looking into this is the fact that we already called the doBody method in is an indicator that the file exists +Also modified corresponding Scala components and test cases +I added two methods you suggestedBut please note that MINA buffers are going to be pooled and therefore there can be waste of memory if you wrap your byte array specifying offset and length +Ill continue to look at compressdb performance changes under the other tracker item as I suspect fixing them will require tuning other inlining control parameters not the inline size reduction heurisitcs. +This function gets a to run the command but there is no way to get the server information from it +Giving the job id I think achieves the original issue of relating pig script to the corresponding MR jobs +see subversion tab tested with G +Okay well take a look for +However since weve apparently ditched starting geronimo with a gsh command there isnt complete backwards compatibility anyway +This affects as well will the fix be a part of +Its hard to get it behave +Please add javadoc documenting this behavior to The methods that take an Operation param +Ive just attached it +Instead the natural execution plan is performed +Nothing jumps out +Thanks DominikI dont know whether mono defines this enum but +Should this maybe be set per CFGood question I cant tell really without a use case although +This will be implemented. +So I define a listen address of for one installation and for anotherI can startstop both installations from Eclipse but I can only deploy to the one listening at +Move the majority of and blocker issues to the next point release +But neither has it been reviewed by counsel +Tried this on both Vmware setup and Xen setup both setup list returns url attribute as this format nfsdirdirdirdir +I packported to and Its unclear exactly what would have triggered this bug +Trying again +Im not sure linking to the root is a bad thing we do the same thing for modcluster +We do yearly stress testing with the OSGEO closing all issues that have been in resolved state for months or more without any feedback or update +So we could add a method to to retrieve the and implement this in +does it work if you remove the? that is not a class we ship with wicket +Heres an update +When we try to remove B we should get a me this in the spec could provide a configuration option forthis case +Thanks for the patch Akshat +Is a draft reference guide available? Id like to understand how this works without pushdown and KEY support +Also thinking about it being able to add repos under project properties would be nice too +contains the subfolder going directly into the trunk directory +The patches were applied on the branch revision A new ticket will be created when shared has more Java compliant method signatures. +Added a patch implementing solution plus unit test +Thanks for the patch N +Fix the log information +We should first finish the discussion about that need on the mailing doesnt seem to be a blocker for +Patch for is attached +this is how it works now +So far it looks like thats helping a lot +Thanks Jiri +Did zOlive ever post his code Jakarta Commons? Without him actually doing it I dont know that it is good enough legally to accept itAlso is your last comment such that you think there is a new patchCommons would be all right since this is a contrib and it can have dependencies +Hi Brian could you please review the changes I have done in order to introduce Push demo into ShowcaseThis is quite big commit introducing also profile jbas as separated support for AS +Patch simple enough +is very out of date +After server was started I was asked if I want to change the inactive scanner I changed it to and it changed the value in the file immediately FINE +pending Arunbq +ERROR If the project that using pdfbox try to force the bc version pdfbox will break while parsing encrypted documents trying to access not available objects +Added missing Apache idea whats going on with? Keeps timing out on meLooking at the failed test history it seems it likes timing out from time to time. +Still problem with shutdown for HEAD branch Allow cactus maven to use cargo +The attached file is a patch +This was addressed some time ago +Martin any comment on this? I think we can close the issue as INVALID +I have already sent an email to Transfer API mailing list +you could treat it as a design time exercise like Ramesh was alluding to where there is an additional column added to your source tables and then its up to your transformation logic to decide how that value is affected by joinset operations +Now it is impossible +Interesting +! Patch looks good +We would just need to edit the folder but the groovy script could be the same for all these screens +I see how this helps +Attaching +I applied the and could successfully connect to a secure ldap server that I have access to and it all works ok +This bug occurs with release plugin version as well +Implementation complete and tests successful +but do you see those classes used only to encode the key +The entries variable is now escaped from weblog entries to prevent this problem. +They only accept numbers as metrics +MarcusThe issue still exists even after stopping and starting the System +This makes me believe that it may not have anything to do with caused by absence of order by in this query +Will commit in next few minutes +Attached isA workaround removing from the classpath containerI have created a patch for the first problem +Can you try this +So yes even though a topic is used for communication its responsibility of the message broker to ensure its unique so its responsible for generating and issuing a topic id and not the client +Im not the original reporter but I tried to provide a test project for this issue +tested +i am including a patch for this issue +Added testcases for both kill and patch to hudson trunk test was broken by +Im suspicious of the threading mechanism but iirc there are no yield points between the compare and branch +Image attached is needed to be added as well +Ill try this tomorrow +The summary seems to say youre trying to run the against an DOM node which I think really is a case of if it hurts when you do that dont do that since while workarounds are possible it isnt at all clear that theres a single workaround which will produce acceptable results for all applicationsBut your description talks about running XSL against the styleshet which is a different case there we should be skipping over nodes as if they dont exist and the fact that were processing from a DOM should be irrelevantPlease post a detailed description of exactly what youre doing to what files so we can try to reproduce the problem for analysis and study +I am assuming that commented out parts of the are for tests and that there will be a follow on JIRA to fix the issues there +good feedback I had always assumed apps looking for compression would be stringblob centric +Anyway considering that nobody else had been interested in this issue for three years it does not worth your time +Unscheduling due to inactivity +So I think there is no need to change anything +I think it maybe TomcatBut what for that is not compiling +is this fix still relevant? It looks like some parts of the patch is already in the code +Ill remove the rat plugin as well since theres no special includes to addSent from my iPad +Given that this issue is about code used for tracing only I dont plan to backport thisResolving issue. +In this case needs to inform its parent +This should contain exactly what you see when inspecting the path using the Control Panel +Thanks DiveshYour patch is commited at revision +The content type is application +But youre right the reference documentation needs to state this and currently doesn will edit this issue accordingly +That could be a good idea +So no longer mention of casting a function to a struct or class. +In the current implementation of NN nearly all the refresh work such as refreshNodesrefreshTopologyrefresh donot check the state of NN +I am attaching the hornetq configurations that I am using +Norman would you please take a look thanks +Id like to try the fix but Ive never used a patched plugin dependency +Please check that it was applied as expected +Ok I dont know what exactly changed with juddi that does not make it work like juddi doesCode was added which correctly assumed the publisher was in the db +Attaching the screenshot for percentage the screenshot for MR task graphs +Resolving since its fixed in master and wont make it in the branch. +I just committed this +Attach patch instead of modified source +DoneJun Done +So Is there any possibility of sharing your then I might be able to give you a better answerThank you! repositoryconf contains non ASCII characters but doesnt specify an encoding + +Frank has resovled the issues +Could you provide an updated patch against the existing sourceIf not just let me know +Please ask at users is the user forum +From my understanding the effective change done by this patch is to keep the detached file in the same directory as the original block right +Elliot removed it for some reason +See would be better to have both and both to avoid screwing over projects using the latter and because it is simply better style in many cases than forcing setter use after construction +I will need to take a very close look to the new version when I will have the time but this is an other todo task on the stack after the threaded authority referencing factory work that I would like to review as well +What version of Castor are you using? There was an issue reported with Castor a few months ago that has been fixed +If you defined the prefix abc an element node with a local name of Invoice should have been selected by abcInvoice so long as it was in the same namespace specified by the prefix abc regardless of the prefix actually used on the element nodeDo you have a complete sample test case that demonstrates what you describedThanksHenry +Not needed yet. +I think its ready for checking in and adding to the next beta release where we can see what issues crop up in the hands of users +And it is used to restore applications and their attempts from the data that they had earlier stored +Are you sure its relocate? I cant see cvsclient in orgnetbeanslib allows redistribution +Committed to trunk with revision Attaching another patch which fixes a similar dependency in by introducing an interface and making implement that interface +Sure that is fine +Thanks Jitendra! +OK thanks for the you please also correct my name in the CHANGELOG file its twice spelled Matrin instead of Martin +example of the to produce +Can someone confirm that this is still a problem in Wildfly upstream or on EAP? There have been some fixes made in this area since +When attaching your file be sure to grant the ASF permission to use your code +Thanks TomSo the bundles are not in error but the framework doesnt return directory entriesI will close this issue. +Ted agreed although I might be missing something +This sounds like one of the things that would be interesting to share with the bookmarks feature of Eclipse so there isnt too much duplicated settings to keep track on +but I had a question should we need to put the generated protobuf java files in svn? why not let it be generated during build? in hedwig we put generated java file in svn path but we dont put generated c files in svn path +Thanks Akash for the updated patchHere I am uploading your patch again after small changesAlso there were conflicts in this patch due to recent commits I have resolved them as wellIn future please take an update before creating a patch +but please add a comment above that code block +Attaching a patch with ifelse rewrite +to have the binary written into the response by an action state and then to forward to an empty view state +Ive upgraded cassandra from to then to +Adding a Utils suffix to the name of a class that belongs to a package util seems unnecessary +I may be able to elevate that problem with Apple at some point soon via a customer +Thanks a lotSebastin +So updating the description to reflect that +The link you provided is for the code samples there will not necessarily apply to +I just ran the unit tests and all tests passed except TTMR +Verified with AS updated with current WELD trunk. +attached archive contains c source for my test schema and xml with data +HelloWould you please try the patch? Thanks a lotBest regards Richard Thank you very much for this enhancement which was committed in revision +Thanks +See for an more complete extension to the web flow configuration that requires the change suggested here +Resolving +Also tested on centos and ubuntu +Sorry +Sorry Andy I composed my reply before I saw your note +What I pictured is a service similar to set and set except the users selection is stored as a preference instead of adding another field to the userLogin entity +running patch through hudson +Ill have a look at it for since its pretty clever and probably useful +bulk close of all resolved issues. +The repeated line issue is still present but that is an entirely separate issue +Should be Any row which did not have the given value for the specified column will NOT be emitted +Heres a patch without the counting +SSL cert for the inconvenience +You have patch for trunk so can run it by hadoopqa +Thanks Arpit +Compile and run with JDK +Because they dont support everyattribute does not mean they are generating output +The Summary An example that incorporates multiple technologies metadata tag on this one is pretty generic +For development you can really do anything you want to the eclipse build path or other project settings +I think that maybe the way around this is not to write https in the URL but to write http but specifying explicitly the port used for HTTPS +This issue is closed now +The workaround we have in place that is suboptimal from a security point of view iso to change the tree perms to worldanyonerdigestmesosXXXcrhome worldanyonerdigestmesosXXXcr homemesos worldanyonerdigestmesosXXXcdrwa +Could you retry the experiment with JAVAHOME set and with an empty or absent? Thanks +There is a decent possiblity these problems have the same root cause although the overload broad that was removed in could potentially hide several problems. +I can manually push snapshot tonight there is no automated build system in use +how and why does this fix the hang? The variable is volatile so adding a synchronized block should not improve the characteristics according the the JMM +is the one thats called during column reads +And I think that you can write such plugin by yourself Sonar Plugins forge open for anybody and we would be happy to see your contribution here +Another snapshot uploaded as +the previous patch +Still missing something I guess too late for me tonight +Paths of events that origin from the version storage are wrong and thus are not filtered correctly anymore +Einschulung? +Patch submitted that includes an ant task which is built and used within the same +TedThe rationale is likeI have set of regions S +I dont recommend Eclipse for generating patch file this gave me a lot of problems +Any ideas? our build is complicated and customized enough that I dont think well be able to clear out the phase +This should be a very simple fix +Let me try to reproduce it +There is one more defect created after this bug Management Server stopped responding in a Multinode Management Setup This is resolved as not reproducible +Probably this problem with metadata access in thread while being changed by thread will happen easily with GNU Classpath because its and work concurrentlycorrectly +table is table based on custom storage handler something like +HiSorry I couldnt write before I hate MondaysAs expected RHEL behaves the same way for both unit tests and integration testsIm running again here and the latest I saw was a deadlockOk lets see if I get a backtrace +The reproduce go to +The Invoker class needs to implement instead of +There are probably makes this task redundant +bq +Thinking about that isnt this valid to all browsers? This could be nice default Enhancer aiui +As a follow up Ill ensure pre semantics is kept in distributed mode +The strategy is more or less the same we followed for the +Edited the property to be in line with the existing ones +Please verify +We have pb too +I would need to dig deeper into Axis to resolve this bug +Closed upon release of. +Spoiled by git goodness +If we dont have time for it we move it out later +defer to + for the patch +It also captures all new files automaticallyThanks again for contributing +Id rather deprecate the Blogger API than build anything on top of itPlus AtomRSS works fine for meets the need. +Here is a patch the implements the improvement +Screenshots +So can we then resolve this with just upgrading the exported SLFJ API and without the spi and helpers export +If someone can take minutes to reply to this it would be great +It might have been caused by invalid defined mappings of relationships +Make a script that can be run after installation this script should setup the mappings file and add a standard on cold start take to a page where the password can be added for the admin major configuration to database and have a new controller to manage the configuration +Covered by for and. +is ready to be committed +Hey Lukasz thanks again for this +The fix is still there in the trunk of SVN +Attached patch implements encodingdecoding now possible +Repeating of headers now fixed when resolving +So I wont touch this part currently Ill only add the detection for conflicted usages of Application and. +I see now +To calculate network health we need all interfaces data to average it and display it +Yep Fabioyoure right unfortunately is not a struts related issue rather a commons one +one should of course not put the same classes in the classpath of typedef and in the hibernatetool it will confuse java classloading +last indention change is not needed +Agreed Ill pick up more of this tonight +So we discovered that the interface change is incompatible +New unit tests tests verify that can evict waiting threads from clientpool and reuse them +Nevertheless I believe that the issue is fixed multiple problems in the implementation of the replaceChild method were found and have been corrected. +No the logging error handler does not support all those redelivery settings etc +Going to match version with version of the xml beans library +Simply using volatile ints should be safe here since were just using bit numbers +This version was built from the upstream R branch and includes all previous patches for and the fix Nick produced for this issueA quick test with shows the jacorb subsystem starting on both servers without exceptions now +Implemented the above mentioned safe strategy in MXCL +The website is now live with the changes made uner this JIRA +Carl Sure thing +See attached Interpolator to evaluate it works for me since Seam +The key reason why we cant use the bzip library is that we need to track position inside the class +Hm I dont see a bug hereIts correct that name escaping is a problem but I dont see how Jackrabbit can change that +Nice thanks! Havent tried the patch yet unfortunately it could take a while since Im on vacation until th of April +Attaching list of various settings for +I have already sent the patch which does the parsing and also check the authorization the the starting process +Needs further investigation +Lowered issue type and priority as the creation of new workspaces is not part of the standard JCR API +Please update the patch +The test case should also test for absence of DONE folder for failedkilled jobs +Fixed +Vacuum may because its part of the db vocabulary I dont think compact does at all +Pull it back in if you think different +I think this would be very hard to do in the existing pool implementations in a backwards compatible way but there is the new composite pool implementation that hasnt seen an official release and we could make this work with that in Pool +An empty PROPFIND request body MUST be treated as a request for the names and values of all properties +Many possibilities. +verified it is included in buildfixed groupId for it +on keeping the timeout +Manually done. +Thats true +This one has no code derived from Android +bulk close for +pushed. +It seems to me that after a reset is performed the same value should be returned as what was being returned at initial startup of the EJB +Did you include jar +Initial patch holding a very rudimentary Maven POM +Workaround for now is to just always give a name right +thanks! +If I actually use the constant the code will not compile under hadoop +Im only using it for a ZK Noel Ive been able to reproduce it but seems to fail intermittently +Although I guess it means you dont have to reconfigure immediately to tryYes thats the key point +Let me know if I missed anything +However the kinds of changes that I need to make to are very amenable to weaving +Instead the app will just map objects to DB tables it is the dev mode people use very often +problem is with cfml based for input +Patch that adds Suspend and Resume to chaos monkeyI had to change the way wait worked +We could use a in the same way we do at you already working on this or can I do it myself +another interesting thing to try is run with JMX enabled and you can use jconsole to attach to the VM and get some detail about the VMZK during the election +I was thinking of a goal that can compile using s own classpath +fixed in version . +This is caused by the same datanode registers more than once +fixed in r. +For urgentmatters please contact Amr Awadallah +The list is order preserving as with Mikes patch but the map gives niceaccess to the elements as was the original +Will this patch fix the problem? +has been marked as a duplicate of this bug +is this really fixed +Im ok committing this too +Clearing this comments because theres no problem with JDK at all +Resolving change suggested has been incorporated within other changes. +or sooner If I get a response +If schemas were immutable that error would be impossibleIm not too worried about using zero to indicate invalid hash +Could you perhaps poke around in and figure out where we start to diverge from MRI? Im guessing its in the logic and Im trying to test that without being on Windows but more info would help +The diagnosis above appears to be incorrect +With MRI buildr starts up in +It can affect on DocumentationTutorial +DONE +This must not be the case on windows +Some comments +Ruwan thanks for the hint +Bob of course this is possible from any Sonar Quality Profile to export the relating PMDCheckstyle and Findbugs configuration file go to your java quality profile and click on the Permalinks tab to get what you want +what is your release timeline? Please note that Im up to my eyeballs in a project that also depends on this so the next weeks are looking quite tight already +Werner you are right +applied with and to trunkSending assembliessrcizpackSending assembliessrcizpackSending modulessrcjavaorgapachegeronimoinstallerprocessingSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsSending modulessrcjavacomizforgeizpackpanelsTransmitting file dataCommitted revision +I cannot reproduce this issue it perfectly works for me +commited to and trunk. +I added r that enables this for both windows and unix +Not seen on any platform on last days regression tests +is there a workaround +The is a doc issue that I opened but since it is not sure if this will go into the docs maybe we need to mention this in the release notes +if strings are used it would work right? So a b c could be found but not? A question is whether it makes sense to treat strings different than numbers +If internalBegin is not synchronized the timeout for a transaction would bleed over into other transactions unfortunately this is a major weakness with the JTA API you can only change timeout for the entire transaction manager you cant set it for a single transaction +Not yielding in snapshots might need to be revisited in the future but at least this patch addresses the common case +Closing this issue so that we have an entry in the release notes +Roller included fixes for this in code and added this recommendation to the docs as well +We only require the version to have a numeric major and minor version +Thanks Bobby. +The issue was fixed along with other issues that were fixed before +Seems we werent saving a reference to every attachment in some cases +Resolving as duplicate of +really +Already a plugin and it works fine now. +Yes definately OK +After times the udig is started the log is try to make it clean up every time +Please close if this works for you +a flag to indicate whether TLD should be reinforced for validation type email +My performance testing was somewhat rudimentary but I think it may have as much as doubled performance +Hi ChrisIve deleted the patch on your behalf youll have to it +Im not sure what youre referring to when you say sorting is not working +For I want to test indexs the throughput in solr cloud start threads in the Jmeter solr cloud server Jetty max threads is Usually pressure test throughput achieve the max then keep or down smoothly the average last status is stable +This is going to require me to revert apply and then +If the procedures are created the client will receive locators +Sushanth Left some comments on phabricatorAshutosh I got an error when I tried to cancel this patch +Whats different between a datasource wrapping an and an wrapping an? Is it as simple as one defaulting to autocommit and the other not or is there some interaction between data sources happening that isnt happening for rd party data sources +Paging to disk will only occur if producer flow control is disabled +However I feel for DDL the railroad diagrams are little more clear and are separate from all the DML stuff +They should be close to what is described on the Ace websites on formatting rules +Fixed in r thanks for the sample fileWe dont currently have magic for the subtypes and it would be good to get some samples for those two if possible so we can the readme of that c code The files contain only data +I think this ended up being a duplicate of so closing this. +It also piggyback a change in number of retries in hadoop +Can you shed some light on this for me +People can correct me here but conceptually Mappers and Reducers are single threaded +Movesrenames are really hard when submitting through patches +This patch does not include tests +At long last the most trivial bug in the history of Seam has been fixed +Apparently this only happens if you are using a top level mapping to the files above +Fixed on and closing all issues that have been in resolved state for more than one month without further comments +We can still do smt in UI but it wont be in my high priority queue if we are fine with HBCK checks +The mappings depend on the font +the input array should be of Object type not String which will cause a stack overflow with the proposed because the boolean arg will go into the Object array +The beauty of and the service registry is that the components play well together regardless of which technology is used to defineregister themNow if it turns on that the dependency to SCR is the only problem which prevents you from reusing the module and there is absolutely no chance to include SCR in the stack a patch to the Sling scheduler which changes the code to rely on just the framework API could be an option +on the rest +Thanks Peter! +All we got up to this points are a unless it is configurable and a couple of s +This test has been if no one reports seeing this any more I will close this +Here I am adding the patch with that change also +Thanks Laura +Im running off of trunk checked out from svn this morning PST +can try and get one of our apps to you but it could take a while as my boss would basically force me toget rid of all of our code before doing so +Sorry for the mistake. +patch nuking the bulk api and implementing buffering for standardcodec so we have the same performance +Duplicated with complete log for this issue +The rack lookup in the patch itself shouldnt it be adding a rack only onceYoure right +simply cannot +Thanks Aslak +I probably will not have the time to do this soon +Just through all these quickstarts on as and as after and Kevins changes and they all now pass. +Can we have the latest snapshot uploaded to apache snapshot repository as well +I made the change to but forgot to manually include it in my diff + still uses a customized addressing schema +Particularly Im concerned that the users hang themselves when they get this configuration wrong +Boffo is new to me +Backed the jars out and got back to the class cast exception here cannot be cast to Looks like part of the that first class and thats one of missing jars +I should add you could emulate the behavior you may desire in the implementation with a combination of the and a which always infers I dont see a simple way to make the distributed version emulate the versionWhile I think the version has preferable semantics neither is really wrong +The tree error messages have been updated to treeItem to reflect the change from to used to create the tree +bug +It can be ignored since does not need the test case to be updated +Check the note in steps to reproduce. +closing +Do you think it could be a reason why this IU is on the aggregate site whereas is is not aggregated +Yeah the two approaches I had already got on my list were system property based and basedLets start with that +Clicking the OK button on the error dialog shows that the process proceeds but the view model was not created +Thanks +Its a great way to encourage involvement by people who dont have the special skills needed to contribute to the core engine +I already did all the work last week thereabouts I have been ootosince and Ill commit them soon +Postponed +Committed to trunk. +We could possibly fix this and make it bw compatible by pulling a file name from the factory for instance but even that is kind of flaky +My bad uploaded the wrong patch +David JacquesWhat a pain in the rear! Theres a lesson to be learned here I thinkSo can we establish what resources are disallowed do we know if the google license is compatible? How many files in total did you count JacquesAssuming we cant proceed further with Selenium is there another candidate? Selenium seemed best but Im open to suggestions I really want to shove this along a bit! Andrew +Lets follow up with separate issues as necessary for fine tuning +Addresses the issues mentioned in the classes indicated +Checked in patch. +Id happily accept patches that do this as long as we have a way to fall back to a version if things go wrong +In the case that the snappy code gets compiled into the same +This is true for all of Struts Config but that is XMLI think it be wrong for Struts to fix +That timewarped in there it seems +I think this is up to you Vincent to decide if real Deployable could have get attribute +I start terasort and then fail the +You are right +HiTested with the latest snapshot +Hi EricI cant attach the MB because it is bigger the the allowable limit MB +Reverted my changed +As will be a major release where backwards compatibility is not guaranteed this is ok +The last patch contains the start of reorganizing the and the stuff it uses around the idea of pluggable naming builders +Issue The temporal expression demo doesnt work in either versionLog into Work Efforts application as admin +I had to make a change in since the script was processed after some default java opts were added to the list +Release Notes for will not be at this time +Ill wait to commit this until this patch is committed +Thanks ToddResolving it as invalid. +Wrapping up work on at subversion revision +Running the same file through the version of FOP takes about seconds and produces a perfect PDF fileI tried increasing available memory +Tomcat devs dont like Maven and they do not support the embedded Tomcat very well tooI think we dont need this for now. +Windows Cygwin is equally trouble so before any patch goes in we really need to think about how to test on that platform +For some reason there was a difference in the Manifest between these two for ER +IMO the background colors while subtle are visible in all the mentioned styles. + Add a table property that defaults to the class +Does that sound reasonable +Its not perfect but considering the subject its in a reviewable state imho +Integrated to branch and TRUNKThanks for the patch DougThanks for the review Andrew. +Update on the cluster executing the same DROP on a second node does the magic +Check that there is no other application server running already on your machine dont know how it is configured on AWS +Changes Changes to skip behavior in case of mode in Just to be clear Ill define the meaning of the three constants abort skip and continue abort abort the import operation skip If an exception is raised skip the current document and continue with the next one +Please review +This issue is blocked until is integrated to common +AndreasIf you are comfortable with the suggested fix lets get it in +Added with JBT. +Since Sonar PHP Plugin and file are also supported as bootstrap for test cases +tests showed no regressions +Hi Arun and Alejandro I have uploaded the incremental patch as well as the combo patch +What does NSA believe they are the owners ofA copyright license is already limited to material for which copyright exist nothing else needs a license + +Resolving as fixed as the reported problem no longer occurs +Does it matter? If matters how should i fix it +cglib seems to be brought it for testing with easymock. +Using Java to build again and it is OK +bulk defer of unresolved bugs to +I dont think users are likely to change the value +I tested again with trunk and there were issues NPE on session ending Pagination is wrong and in trunk at revision I will backport after What I really want to do on party side is beginning to work on and remove all this freemarker mess +We are trying to refactor a bit and make available as a feature +Since marks Cluster as Evolving this is not a blocker but it would be nice to fix nevertheless. +This is not resolved we already depend on their jars for startstop so that is not the problem here +Yeesh +I didnt try to investigate the problem +Hi AlesIve extracted the pages into different files and attached them here +wrong fix +Hi Ian thanks for filing the bugs much appreciated! I just wanted to bring to your attention that setting Fix version to is typically discouraged since this mostly reserved for core folks for youre really think that the fix must be in the next release stating so in the description and raising the priority is a better way to bring the attention to the bug +If there are multiple jars all providing that file then the patch will merge them togetherBecause s message parameter is actually a key to an entry in this file it is necessary to allow other file system implementations to add entries to this bundleIn some cases a file system may implement the interfaces directly and not extend +Accurate backtraces probably arent essential and optimising out send sounds fine to me I wouldnt say that matching MRI is that important +list reqsec +The patch for + thanks Stack. +I have thought about it +I checked this issue against the latest version from the trunk but it remains unresolved +srcmainjavaorgapachewicketexamplesvelocity +More improvements could be done +Ill get you an example shortly + breaks fisheye so we have to manually update the fisheye config at the moment +I also checked the MEAD repository. +Gary will you provide the patched Pom and src and I the patched test sources +Apparently jira doesnt make these components link back to this one as being a submodule +Fixed in jBPM +Added an enumThe var added is inside a function so different from the global var declaration at line . +Sergei Well spotted thanks +when you say vc that would be Visual Studio correct? Because vc Visual Studio seems to build fine for me +It sucks having the extra platform dir at the end of libplatformpluginsplatform but it makes sense when it comes to module. +I guess I need to separate patch file +Have changed to and now I can build the projecthave the patch reviewed + globalmosaic +I cant work out exactly what you have changed in order to apply it to the current the trunk version of has changed a lot since I raised this I cant even tell if this is still a problem and Im not sure whether the proposed fix would apply to the current version of this issue should be declined +So I hope that you create a design and implementation that can accommodate the feature module use case for Spring + +Alexey will have a go with it +Could you add a test case for this one +LarsWe will come up with some performance numbers based on our scenario +The +Then you install a plugin that needs to set up another environments block +the callback handler for one or multiple events properties required to initialize the callback handler the handler for events properties required to initialize the event handler +This patch is a valid bug fix +is a stab at the new simple method +It does not look like an issue at allYou can use the help flag to get all the options +Linking this to the original JIRA believed to be the cause +This implies that the actual extraction of the required service information is abstracted in the service and there is no API documentation that the bundle symbolic name is taken for thisI strongly suggest to the method to keep supporting the abstraction +Here is the patch to include snapshot paths also during fsck checkThis is on demand +do you think thats feasible? does it works across browsers +The attached patch works both with MAHOUTLOCALtrue and in cluster mode +Thanks +Yes I would think soCaveat deprecate not recommend for new usage existing column families using the have to stay with it for the reason Frank points out +Let me commit. +and if all seems good then back port +I will cut it out for the release +Also a LTR word for getting this status would be nice +the stup is auto generatet. +They are however still registered with the resolver and available as potential resolution target +I reviewed it and did some basic testing +Should be relatively easy to remedy. +I put this into +While EC security groups help Ill sleep better knowing that Couch is not going to reveal anything to anyone without the required credentials +However it is not safe to make this change on the branch +Please review the patch +Now all tests work fine even when no ant build was done before +Remember to check the regular FTP producer as well it could have similar bug +exist in also occurs +Its the same as the previous patch +Otherwise does this look like something that should be committed to TRUNK +Services within the cluster will use Kerberos without the tokens +Starts a thread that shows a lock table dump every seconds +well the issue is that is and thus we dont generate apparently that default is being ignored by either hibernate or look into it +SanjayPer Todd and ATMs question on can you explain why this approach is better? The only rationale I could see from the thread was copying ZAB +Still same problem with +It would continue to use the old version of the windows until it processes the +Perhaps you could experiment by trying to parse an XML document that contains an external entity thats +Id love to contain it if you have some suggested verbage lets put it in there +org things marked as included in JBDS or not I havent found a good way to welcome for docs to ok closing +Oleg +i files that relate to individual bindings to the include files but the +Patch looks good +Ping It seem to have restored the scan and stuck at r for ages now +Jonathan +In any case I applied the same algorithm for both TCP and UDP +Which env variable are you using to set the max memory? I think its MAVENOPTS you are running it in eclipse as an external tool make sure you set the variable in the environment tab it may help you out +Heres some additional infothe will output a bad library filethe will output a good library file +Updated +I will take a look and see if updatable and insertable are being handled correctly +The same issue also affects EAPEWP on RHEL +I agree with Knuts observation that the THROWASSERT was not adding much value but it also seemed that removing it was not essential to the patch so in the interests of making a small focused patch I simply included the changes to andI believe that the work on this issue is complete +Verified with revision rThank you very much +Fixed with revision +I should have specifically asked for it to be tested +An additional screenshow showing the user menu +This is just an +I was having a shocker with reporting bugs on the weekend it t spend time investigating Ill see what I was doing at the time. +settings folders and files? And how about +Using J version is obviously your call as to how you think we can best get good performance out of the generated Java code but in my opinion and experience youre fighting an uphill battle by generating convoluted code and depending on something downstream to perform a heroic optimization to make it run fast +btw should note that I manually tested this against a running NN before commit as well +Listens on sessionCreating stateEntering and transitionExecuting events +Kathey do we know what release of Derby was getting used +Added the deleteFile optionCommitted revision +Hi MichalIll send you the credentials of a test account +UNOFFICIALWhen running this command I get for +thx +should retry for new patch by sijies change rename should also change its message +I typed the version wrong Im using of the plugin +So if one component has a flex of and the second has a flex of the first one will shrinkgrow twice as much as the first +Reluctantly deferring to due to size of change and lateness in the release process +The change was rolled back so the previous version is still there +I will look into it +Initial support will be provided for the following runtime types Fuse Service Works component provider extension will be provided that scans existing EAP runtimes to see if is installed +So it started based on CDI and Seam modules are extensions to CDI +bulk defer to +Backported to doc branch as svn . +The proposal B makes more sense to me BUT it would mean a change in we would go with solution A for now +Need to tweak file +sorted thanks to gozer and joes +In my example above that assumes that we always convert Ruby Symbols to Java Strings +Ill run the profiler over the test today and see if I can see where the problem is +Eugene can you please retest with Solr or at the very leastThis sounds exactly like which was fixed in Solr duplicate of +well the testcases pass +This is fixed for +apologies +This has been removed +Some more debug infoThe parameter of the method is which is a super class of while the runtime type is +Updated patch +Was able to reproduce some issue in a will take a look +OK thanks! I committed also the build file in revision +Hi FernandoI am reopening this issue and removing fix version EAPEWP +I will leave this open because I think it would be good to backport to although it probably wont make the Release candidate +I havent changed anything else in v so I think could be flaky +should correct this so try it out on your the test after and can not reproduce closing. +I just committed this +Fall through whatbq +I am wondering what is the recommended configuration to use between and +I changed the HTTP error code handling exception message to send something similr to the followingHTTP transport error Server sent HTTP error Not Found +patch fixing documentation merge issues from latest patch with modifications suggested by community would say the updated patch does it for +I feel that Hbase really should request the HDFS about the needs instead silently adding with reflection +fixed in CVS +in addition to the attached version there is one in the controls test dir plus several other ad hoc ones floating aroundClealy a demand for this +for to trunk +so that doesnt really help a reviewer of the code +Slava could you please take a look? +ahh i misread this the first timei can see that it puts an empty string but that is not invalid xmlill check the editor and convince it to remove the attribute if not present +This should be set in the response coming back from the server when you request gzip encoding +Assume closed as resolved and released +Thanks Alan! +ClosingShould QE want to verify theres better performance from Eclipse Juno JBDS theyre welcome to do so +Attached diff file containing changes to and which allow the imagemeta rootaccess value to be controlled via the image properties page +No code change just javadoc. +Hi stack would be very fine if youd like to bring a REST server without agile json to the currently release Hbase +All these tests now give the correct ouput. +was already fixed for the uses already so can we close this issue? +The process is asyn and hence the tool will not be able to do thisEven +Ill post up a list of actions soon right now Im working on a proposal for format changes to support encryption +This issue should be resolved with the completion of. +Preview will be added later +Hi Matt By looking at the code this patch is applicable for as well +I have moved this issue from Spring Integration to Spring Framework +The which allow ping will not be started on another host who dont will be restarted even when they are running +Thanks +Good to hear you are moving forwardBut like you said the pain is all yours +But regardless Velocity is used for general text generation it is not specific to web applications +On Matthieu suggested we dont have a configuration mechanism just yet +Is a duplicate of this or do we need separate fixes for MAPREDUCE and HDFS +This new issue is an improvement over the linked VlastimilDamn thats too bad we need to make a subsequent REST Call +s welcomeCheolsoo are you comfortable with making a blocking documentation patch for this feature or do you want it to be a part of this +Only full publishes will do commits and pushes +I have changed the ESB issue to a blocker for doing the same with this SOA issue +I saw an issue with similar looking backtrace from the bluecloth gem so this might not be mysql gem specific + +Ill start moving the code to use the new ctor class etc +Candidate for SOA +Lin +Committed to trunk. +Therefore should all the dynamic vdbs in the examples have +Thanks again + has an entry for JASPYT library +Our scenario is a bit more strange +Yes too bad that its a solution its not part of your war so you need to customize your installation or figure out how to do the same on weblogic websphere +Ran all tests and verified no failures of tests with in their nameBelieve this patch is ready for review +Now just need to write some tests for it +Use interface instead of interface +Ill script up a permission fixer and tidy up the newproject script +since there is soap action operation has been dispatched when it comes to security phase and every thing works fineLets say some one send message without signing and without soap action +Even if any exception happens in SETSPLITTINGINZK even then the same clean up is getting done +You should now have your mailing lists +Thanks to Alexandres insights I was able to understand this problem +Patch with a test that shows the warnings +From what I understand the restriction is only for lob columns +Sounds a bit like I dont like American lets speak English insteadId say I dont like Korean lets use English instead what we have is Korean and our dev and users are only OK to speak English +Have you been able to reproduce the Autowiring problem? +Im not sure how tightly bound CXF is to the Sun RI +I have seen this happening on my system as well +example to original output isRotatedtextisbrokenintoseveralpeiceswhereas we expect it to beRotatedtextisbrokenintoseveralpeices +added to release notes as a known issue as per the in revision +Hi Liang its really good suggestion + may now be passed in +It needs more prefaceand to go a bit more in depth in the configuration and options of cocoon +bq +I hope these features will be implemented at some later stage +I renamed jar to and now functionality is working fine +assuming that the automated tests pass +Granting permissions to Derby need some write up for runtime permission on get and maybe for on and +Attached a patch that fixes the problem by recreating the compile method marking it as deprecated +Id expect the combiner to ignore the output property and always issue its output in the same format as its input +It can be done but the library used to make the WMS requests does not make it easy +I hope it isnt in my test case +Its on my and already rectified in my local sandbox so expect it to be committed soonThanks for reporting +As Gunnar has already said static method validation is not supported +I believe you are talking about the headless property which tells the AWT system that lightweight components can be made without a heavyweight toolkit +Tom do you mean jrubys irb in Swing mode +metadata +The failed test should be unrelated +Fixed as part of Jeffs work on +Maybe slightly more complicated to avoid scanning all the attachment prior to the range start but thats an optimization +we should add a feature dependency to Karaf http +Shall we do something about it +Thanks Ning +on Scotts comment +Things seem to go well now there are no longer on product pages at leastSo guys its ready to test just checkout the branch and +It can be in a separate class something like +Hi Devaraj if the problem is not in the latest trunk is it already fixed by some other? Do you know which have fixed it +You can only send an error code to let the caller know what was the problemIf you need variable data to produce the error code it might be a good sign that error code is not in a good place +I am getting this issue as soon as I add to a Weld JSF JPA project as well +Hi ColmOkay +Thanks for the patch Lars. +move Environment to implremove from protected static commandServiceso that Environment and things like that can be moved to wont expose the hibernate session in the jbpm api as we might move to jpa later on +After removed the class Hadoop still can be compiled +The BPEL library is cleaned before the examples and the task definition in depends on the library +I tried the modified on a Mac with the flag and it worked fine +Done +Has the diff been applied to svn? If not should this ticket be +Changes ported for Remoting +Remove some unnecessary changes that fix warnings we can address them separated later +There is an open issue for parallel execution +You mean to say that any work done with respect to some should be published in the milestone release notes +Example implementation of how Im currently keeping the in tact with a during serialization +Committed to that last comment +Examples updated. +Bumping to LATER since we havent had any movement on this for this release +The specification update for this feature is brief +Change log level from error to debug for this kind of exceptions +It would be nice to have a tool to do the bulk load automatically +Are there any others who could test this aside from Chris and Myself to possibly provide more insight into the of the tests I can only produce if a I run it in Release mode AND b I run the entire test suite not just the single test +Reopen in case of changes needed +fixed this I ran a build manually and then the project after having renamed it to. +the thank you for reviewHere is an updated In misprint must instead of is must In misprint must instead of is mustOtherwise patch is good +API you could pass any and not only a +What teiid client version are you using? It appears to be a client +There are kinds of node in the image and editsa +Comment from JirkaBut the file in question is better formulation would be You must configure these items manually in process definition XML the release notes page section +Marking my JIRA a dup +The last patch is OK for me +It also refactors the test to make it cleaner +our control site hit the twice but the site patched with the fix worked fine. +Moving to eunit is fun and easy than expected first +I dont want to have to specify an action for each web service call +Wait +He ran it as root and it caused this problem +in +That should work just want to make sure they have similar API so other toolsutilities will automatically work or just needs small changes +Clover is now and disabled bc of the clover license +is out the door adjusting on all remaining unresolved issues to by to +I wonder if we will encounter the situation where a user starts with SNA then ask how to transition the data in it to a cluster. +Before it seems the classpath in contained the clover jar +Apply at federated root +Please confirm that this has been applied as expected +ill try to replicate this tomorrow with a separate +I can however achieve this by using an embedded field which imitates a list which converts its content to a csv String object +The issue you raised actually deals a violation to the spec. +Sorry but I am not sure to understand correctly your usecase could you send us a test caseCreate aggregated javadocs for both source code in the default location and source code in an additional location +Do you plan on releasing such a version or am I required to use Maven to get this fix +use immutablemap reduce synchronization +Currently after the index changes are retrieved the writer is closed and reopened in to be aware of the changes just pulled in from the master +Im still not sure about why +I also dont understand why this should be lacking a test case? +could you summarize what you are up to with this issueTo clarify I mean could you summarize the approach you are taking with the current work you have done on the branch +Test removed +We could at least add a unit test +Its the README that is incorrectThis is critical for testing +When you say NOT write the block locally do you mean not selecting the local node for the write pipeline? That is block placement should use all non local nodesAlso while machine failure is one of the issues there are othe issues to consider as well +Ill send an email to the dev list throwing out options on how to handle it +There is no reproducible test case +For instance today we have Database and +Actually its an output of a multiple choice quiz from my application +This has been released +finally it seems to occur here now as well +Help in resolution is appreciated. +There could be other methods in too that need similar changes. +For the first iteration the data will be dropped +Got around to trying this and I can reproduce the problem +Unless somebody can advise me I seetwo options disable the test for other than Sun or remove the test entirelyWhat do you think? I would have thought he last version of the test would always clear the interrupt flagbut it seems that is not the case +Im having trouble imagining how that would work for a cluster where N RF +Yes quite right sorry about that +The first file contains a lot of changes because of my usage of Visual instead of Eclipse and how tabsspaces are cleaned up. +If no one else has time I will add it to my list for testing in any upcoming release +the proposal to make non ambigous change URL to haveportalgroupsplatformadministratorsNewsportaldashboardNewsportalprivateclassic be verified in gatein master and reopened if necessary +Patches welcome +Committed a fix and added an osgi test for. +reopening to write release note issue now that release note text has been written +I agree the proper way would be providing help scripts to regrant as you suggested +Dont copy over the Smooks Console files +Now it no longer appliesMichael can you please regenerate this against the current sources? Thanks +Add debug comment when includeTests is set but there is no +Believe it or not Im no native English speaker and I dont really dare to write documentation in English +The problem is that youre using AND +Thanks Kanna +Merged the last three fixes to and committed revision Marking as resolved +Set filterTypecustom or set the to false +to removing broken unmaintained code +Thanks Jukka! +That could also give you more control over how the connection is created if pass through is needed etcWe may also still be able to do some form of connection pooling either on a per vdb basis and use identity assertion or globably and add an enhancement to switch vdbs as well +Thats another reason why I do prefer to have the documentation tied to the sources +There are my opinion To keep smaller do not put information that available by other API there Context Request and Response objects etc +This might be the cause of timeout for +Ok it looks like the change to always switch to classpath when jrubys jar is sourced from a jarhttp URL works pretty well +Thanks for the information Mark +glanced over it looks good in the meantime +CharlesI just tested if you add the classpath before the wsdl url CXF will try to use the thread context classloader to load the wsdl fileAnd it will work for you case +Really appreciate it +And without DTD I cant process standard entities such as aacute etc +I fix only those classes which were in Veracode reportI have several other small fixes according Veracode which I will report to you in near feature +When we compile hive this is the right procedure because we want to compile hive against all the different shims and hadoop versions +Examples had to be removed from the assembly +are you guys planning to add this soon +Teiid has committed the changes to the product branch and should be in the CR release +Patch created from git changeset +Closing old issues +Revision +. +This is not tested but it builds and Im running tests to adjust output files +Excellent thanks. +Committed to trunk +Aside from some tree conflicts weirdness with Subversion when updating my patched working copy everything looks goodThis new patch enables support for multiple archive formats adding +WilliamIf you dont specify the WSDL and SEI in the URI that could mean that the works in Dispatch Mode +Upon review many previous lucene releases had no date here and solr stopped putting them in a long time agoIf someone is a historian they can use the apache mail archives or websites or something else to discover all these dates +Corrected summary the bug is not related to rows but to first attribute +never seen it before in any program +Thanks Kiyan +Even when it is implemented JBT is far away from its first stable build +for the just committed this +Patch looks good +Oups sorry I missed the note about populating releasesis +We could add a paramater that specifies a starting lo value that gets added to the current value +It is somewhat rare for people to use it but it does work and it is used +i currently dont have the time sorry +Workaround is to use subqueries +For that set in general properties +Thanks German +batch transition resolvedfixed bugs to closedfixed +works fine in snapshot +Thanks applied the patch +Closing this issue to be included in +Open Groovy fileOn my computer this is when the color scheme gets ignored and clicking apply fixes the issue +Fails if patch is not applied +Regarding rewriting the test I think if the rewrite makes the test easier to understand and still tests the same its fine +As project despot I confirm that Mike is a developer on Mapbuilder +At the moment there is only the regexp based one but new ones can be addedLook for the interface in libraryjdbc module +Thanks Isuru for the correction. +Hiis there any chance this could be fixed soon please? Cheers is up at central +logical expression are a different thing +Patch looks good and is well tested +Updated patch +support a more general retry policy +Finally the journal update is prepared +I like that we not have pluggable distance metric for example +Looks good to me +A dump is also triggered when an uncaught exception is encountered. +Yes this has been committed. +Fixed in the doc source of the Security Guide +Was this after I actually applied the patch +ent file name is incorrect or in the wrong place despite the Forrest documentation +I think it cannot be more explicit +Ive also been told that changes in the for dependencies have not yet been addressed in the M conversion where the scope needs to be set +Thanks goes to Microsoft for this awesome feature! will contact Docs team to put this into documentation +Do report info if anyone else sees this. +New version attached passes the test this time +Hi BrianPatches and suggestions are always welcome please provide one and I would be more than pleased to apply itLooking forward your contribution! in and +bq +See the attached results from +Agreed on the consistency +Thanks Lewis. +Good work Ivan latest patch looks fine +The deprecated repository should no longer be used +Assign to John G +I couldnt have worded it better Jeff +Its follow up on +I think you did something about this earlier but cant remember +Not sure on maintenance status +Whilst I cannot offer an immediate solution if you switched to using node type attribute for a it should work +As a result the ejb cant be loaded at all +Awesome Ill commit it in the morning after usual validation +i think the ledgers should be in a weakhashmap but that is a separate issue +This is lightninginternal jenkins node +This would HUGE +I agree that is just fine at parsing time +I can modify the patch to explicitly decorate classes with this annotation if there is an issue +The tests that are failing are security related tests Dan thinks we may need to add strong cypto to JDK for switched to JDK recentlyEamonn can you investigate +HiI would be highly interested in your solution it would make my life a lot easierWould it be possible for you to post your fixes in the here in this threadThanks a lotMartin is the +I have added support for requestLocation and responseLocation attributes into the configuration of the these are specified then any validationwebservice will use those values for request have also added deployable unit tests into the qajunit area to test validation and location in revision +committed minus the unused method +I have updated the patch to have these constants in the class itself +I have fixed the test suite failure for jdom +When checking to see if the current status is Waiting on our method used the Statuses override of the equals method +Thus marking as blocker for easier tracking +Same for us Hibernate Oracle +Compiles +If it works fine for you now it wont be broken by this patch +Final updates complete +I dont know +nah its not +Looks like the checkpoint rebuilder is done +more or less same issue in that varchar of length rather than clob causes dbms compatibility problems +This version of the patch includes files that I forgot to svn addAlso Ive added Eclipse configurationAll tests pass +IMHO no separate regression test is needed for this issue exisiting class library test reproduces the problem +Sorry misspoke the patch option wasnt given in Jira so theres no file however the patch was mistakenly for Hadoop trunk +This patch will correct the Erwan for the patchI would not be able to get back to this soon so keeping it unassigned once againMay be someone else will pick and commit it soon +I have removed that with these patches +Its tracked here its very low priority for us at the moment +Am unfortunately still getting the same messagegit diff git diff trunk git diff trunk git trunk stdout +Issue fixed at apache and merged to and fuse branches +I have added a chapter and download sample for IPC and added a section on instance preferences +And I do believe the code could use an overhaul as well since its a bit tricky in placesFor instance generating the temporary filename should just use the exchange id as its unique and a much better name than a generated number +Youll be messing w the difference in and append +Passing the port range as the default in the test case should be Ok +Apply from delivered to SVN. +Thanks Zhijie +How big were the messages? Can the shiped example producer and consumer in the examples directory reproduce the issue +an idea could be to have the tool that converts from checkstyle to eclipse outside of this plugin +Please commit it anyway +Path applied thanks! +The patch +My preference is also fixing the tests +Unresolved issues to CR +When does this happen? everytime or +Joe I downloaded the D version and tried it again with the problem document +A database always has a territory when not explictly specified at create time it defaults to the locale of the JVM +You mentioned GrailsAttaching the WSDL could be interesting as well +proposed patch to include as a possible failure +Obviously this is named +However there are two desirable features +Am I correct? Shall I obtain one or is there a key I should reuse +This patch builds a tomcat distribution configured with all necessary dependencies and databases to run DAS samples applications +People expressed concern whether introducing a new exception is the best approach to takeChunhuiPlease extrapolate on other alternativesThanks +I wanted to make sure were doing the right thing +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +If we are worrying about users to be deleting the directory after they create the archives then that is something users should be warned about and not programmed into the codeFor it would need some changes to the Command class that is used by count to use uris for input +Also I doubt well be doing any more new feature development in the branch +Ive moved the issues to Common projectAs for the combined switch of the Servlet API Jetty I am for it but considering the stable state of it might be somewhat a disruptive change but it seems totally fine for line considering the time frames etc +pending jenkins ve committed this and merged to. +Since the description and code is different for can you create a separate bug +Its only related to this issue in that I suspected it as part of an indexsearcher not being closed fail that made false fails annoying to wade through +Will begin to fold this into our staging environment +Ok heres the roo +Would it be preferable to roll this back and wait for to be cut before checking this into trunk +Or are we cnvinced we want to do it +Reconsider why nullary methods are automatically generated from properties +Added Data Role resource under VDB +I was of a split mind on this but if we allowed configuring hash function in these cases then we would have to persist this information somewhere in the data which sounds kind of messy so I decided against it +lib +Can you provide us with a patch that works for you +My patch just removes the deprecation problem +Please populate with your latest release and Ill continue the setupThe area is now populated +Sorry for late replyReto can you please up the site again for me? Now the script returns Not Found The requested URL was not found on this server +In fairness I recognize that JEE also allows these annotations to be omitted in favor of external XML descriptors like iPOJOHave you tried writing classes that work in this manner? Do you establish a convention for commenting the declaration of fields that are managed by the container +I think changes are tested sufficiently by existing unit tests +I have reverted the changes that were associated with the user +As regards a static could be set by a thread while another thread is reading it unfortunately this is true with several other fields in that class however a static field is even worse because the risk of accessing the variable when not initialized is greater +Remko thank you for your quick feedback +No noted progress on this issue +No longer required +Yes +dupe ofThanks +I think one of the advantages of the Parallel Processor is the metadata that is merged from the child workflows +Thanks Greg for reviewing it +Okay is there a forum for this? Mail lists are evel +They couldnt get beyond a CXF client sending a because the Offer was missing a required Endpoint element +Dont forget to click on radio button to grant perms +Im using tomcat +Hi DanielIt seems that we need to file a separate JIRA to complete this fix +There is minor changes in previous uploaded patch now after depress Save button on Add Note screen shown View Order screen instead on Order List screen +Will users get upset though? They seem to work fine +Perhaps there is a way to do this already? If someone could point me in the right direction that would be greatThanks new patch added +I figured out how to edit the wiki page +Updated patch Renamed to Changed URL scheme for to jceks +Our group ids and artifact ids are designed to closely mimic the package names in each java project +Thanks Milind +without any working testcase we cant do much +I Contributions page appears to have been removed in the site documentation cleanup. +Uploading cleaned up patch +Marcel this issue is specific for mongo same test launched against runs I suppose that the fix must be done in the mongo microkernel implementation +Then I realized the methods were not public +Previous fix did not cover the case of a task after exhaustion +For +A varargs method will unintentionally check to see if any array elements are null a bonus. +In general outer join operators do NOT commute and associateunlike the inner join operator +The only changes right now are some doc edits being created JIRA to reflect enhancements requested by Peter Chandler of the FAA project +Otherwise the whole mechanism is pretty much unusable in vocabularis that use unqualified local elementsBoris +I actually find it refreshing that at startup there is very clear path to what to do and not a full screen of mixed my suggestion on trying to make the screen less invasive while still letting users know there are something more to read about +Working in trunk +Agreed +Some comments The path returned by the localfs merge need not be converted to a fully qualified path since these files are guaranteed to be in the localfs The localfs merge is passed an array of paths to merge +Oh I didnt refresh the page before posting comment ignore my comment and thanks Ivan for troubleshooting +But then again I could have just followed the instructions for the plugin +Unless there is a big demand maybe we can bundle something like barring any licensing issue +The connection reset by peer is probably when we dont check if its closed first and Java raises an error because the channel has been closed +Verified by Mark. +now has comments as well +Thats the posters point the parameter is required but a sensible default is provided and Tapestry is complaining that the parameter isnt bound despite a sensible default being provided +for implementing this +The patch failed core unit tests +Although I have not hit concrete problems due to this I always thought it defaulted to falseAnyone else in favor of changing the default to false from? We could check lucene to keep old default for solrconfigs +I have not tested it but I got an email from Embarcadero some time ago that this should have been fixed with CBuilder Update testing shows this is fixed with CBuilder Update +Ive created to cover that part of the issueThe remaining part of this issue that the Admin connections should be abstracted behind a configurable runtime will be retargeted +SeanI created for the issue you raised. +Nice cleanupOne question will we be compiling of the examples every time we build now and running the tests? It seems so going by an invocation w this patch applied we are building an examples artifiact and it is full of code jar tf targetWhen we built before this change did we build example code as part of the general build? Do you know Sergey +fixed in r +We might choose to do this later +This is already available in the latest dev builds +just a note for myself you probably know this alreadyin all three cases I checked the message id causing this was connected to a PUT operation that didnt complete due to +Im not sure why this page is immutable but Ill try to find that out +Thanks for the report RomainI also fixed the other similar situations +Can this command give a warning that container is managed by fabricIm not the only person who has been confused by issues have been reported in forums +Implement Advanced Search for +LukaszI didnt know how to move the comment but because the issue was already reported I added only a comment about the workaround to +Hence we plan to address both these API defects in the time frame +so it would never work anyhow that said I know there was some classloader issues with +Ill test a bit more and then close the +Thanks for catching the changes in +Only NS support possible in +Do you think this should be removed from Java code as well? I will add Btw the wiki entry is excellent +Yes and yes +Tom sorry Ive been traveling from India and hence the late responseI forgot to add I ran locally on since the patch was originally blocked on +They should really be included in the build +jar file for and rAdded a new class and fixed the launcher to use it if not otherwise specified in the file +Can we just move that code into so its fixed for any use and not just? v attached w that approach +Also if were going to ignore the header for we should ignore it in the checksum case as wellAlso did you measure the performance impact +to comment out for now on bits platforms for nowI think we could change this test to use the optimizer directives +Andre? +Ok now but feel free to reopen if you detect any issue about included features +What is the value of wen no is setCheersReto +Solved the problem by using qname equals +Ill test more tomorrow and update the ticket accordingly. +I have attached the code change patch for consideration in the next release +Good catchI have few comments about the patch Can you merge them into one file? Can we use? Its an improvement over without needing to use fastXYZ I think that we can initialize the scores array to maxDoc and drop the resizing in collect collect is a hot method we should do minimal work there +Andrew One for you when you have a moment +James Ive got your point +Documentation updated. +Works as expected now! +Tentative fix +In my case this infinite loop was started when one datanode went down and the namenode started replicating +Which I assume is invoked lazy at runtime on processing a new ExchangeGenerally initialization should be done in doStart because starting it is single threaded and we dont care so much about performance at startup +Finally if this ends up being the last standing blocker we should not block the release because of itThere are a couple of blockers that are PA and that need some review love +Thank you! +since I am currently busy in a lot of other work can you report the results of your manual testing here after applying the appropriate patch? Thanks +Carsten I checked the snapshot version and I confirmed that the changes fixed the problemthanks a lot. +BTW your help on this is mucch appreciated its a really annoying behaviourHopefully FF will helpThanks seems to be fixed with the latest trunk +The replay command belongs to not client so we dont expose it to clients +I agree it is a concern but it will be a little less of concernI think that the fixed number is not very nice either MB or MB +Is this being worked on? No progress seems to have been made +thank you for your early reply and patch +Filed for decorating transitions and events. +That way the class could remain unmodified +with the ONEAR will there be a way to indicate that infinite distance between the terms is OK +Committed to trunk +Ill file a new issue and push this to the repo +Will probably need to change this searcher implementation but this is a priority for later. +Wed have to emulate libcs treatment of opening directories as files or else wrap a native call +Anyway if people would be able to specify the order they might also put their own ones before the +Revision +Dear Kathey Thank you for committing it on revision +OlegThanks for the quick reply +So catch thatOOM then send an exception response insteadDoes that sound good +Please can you update the README and upload an updated patch +Defining them there provides a way for jndi to discover message bus destinations in most portable ways +Since you viewed the logfile in notepad Im assuming you are logging into a file which narrows down the list of possibly malicious appenders to the and theIt seems sometime Im receiving a huge response which cause this issue! strenghtens my hypothesis that your service is logging a message with binary content +Linking to thrift issue +If attachments had their own connection then the primary changes feed could heartbeat continuous replications but I dont like this solution muchIs there any reason a receiver needs a timeout at all for this? If the source process dies or dies entirely on the other end we should get an error reading from the socket +Still need to address core test failures +Please confirm that they have been applied as expected +I still see this error with +Added test in revision this patch runs the test against the improved property index and the test passes +Submitting patch for a QA run +SvetlanaFixed in LUNI module at repo revision Please check this fully resolves your problem +persistence reference turned out to be misconfigured +reading this trread i dont see the actual security issue +This means I think that and are features rather than potential bugs +Fixed r +Resolving after +Yep these features works very well only within Clearspace +A private build using a copy of does not appear to leak files at all during runtimeI will continue to test to see if the solution which was written for just this purpose resolves the issue +However we should look at HDFS too +The DBTOOL should check before doing the upgrade the DB version is lesser than the current one +Right thats true Ive just spoken with Tomas hes gonna supply the way how to test this in his ll try to test dans fix as well +Im testing this patch against a full axis build now and will commit if its good +So how would that work? All requests going to console would be catched by console app? I understand backstage can be somehow obscure but I actualy do have console as a user dashboard in one of my apps +Default behavior updated in +AlexeyIts just an improvement of course it could wait until new build system is committed +But I got rid of that at some point +Heres the places that I found were throwing exceptions +Hiits me who has published the issue and discussions in dev mailing also read the Vincents document of Dashboard plugin and has had discussion with him about that +Thanks for pointing it out Pat + resolved issues for already released versions +In the test the on method is not called +Please reschedule if necessary +Should we put this into +This would certainly be nice to have and can be considered a prerequisite for the fabled but should it be listed as a blocker? Id like to move this down to major and keep it in the hopper for work +Ok see it in tmpjbossws +The code base for and had diverged quite a bit and we opted to not backport many fixes because they were too invasive +Patch file for to enhance JDI functionality +This can be added to the Message Redistribution section +I also updated the READMEThoughts? Does someone have recommendations on a more correctbetter way to do this +This test fails on some other platforms too depending on how it is run +Youre right guys +I will maybe add others later +Slightly cleaned up patch +JonathanI just provided some observation +Any warn message after the reset would trigger log at consoleAlso set the default log level for root logger to patch in revision not seen since patch has been applied in rev +master PR included some stuff that was missing +Found out that the xsd suffix was removed from Axis wsld creation due to an earlier issue +This is because boost library dependencies are a packaging nightmare as boost doesnt maintain their previous releases at all +All tests passed +Also worth noting that this caused log files to pile up in a test +Put a marker in the right bar to locate the code in the file +Would you please try this patch? Thanks very much +Just confirmed that it works fine at using bower do bower first +libUsersjoaDocumentsworkspacescompanyworkspacebin +This is done by having several jsl scripts in trackers +Fixed in r +not to belabor the point but please reference section Session Bean Superclasses of the ejb spec as it provides a very specific example of what is and what is not allowed as far as interface inheritence. +Furthermore we are typically required to smoketest before each commit +Rebased and merged into the branch and into the master branch. +Verified in JBDS contains detailsVersion Build id Build date +may I ask what version of the spec this is +zip names contain the minor revision number the top level directory in the zip only has the major revision in its name +See and ve applied the new lifecycles and artifact handlers here so whatever ejbpar plugin implementations you want to use will have plumbing + +Move the ones back that you actually have plans to do +That did save some time and disk but was causing the jobs to fail +committed +Ill correct it +bq + also change the constructs modifier from private to is now a version to the truck +Patch for myfacestomahawk +This should be tested against an Oracle DB +Not hard to do but requires a bit more than just marking as serializable good +does not have the exponential back off that Eric just added it just sleeps secs between retries +OK Ive applied the patch to both x and trunk and it looks good +So here is a new one +Work is started on this issue +Detailed design document for addition of Data Objects to Modeler and base BPMN model +That just use that in get +Please reopen new JIRA issues for any specific limitations or bug with a precise description of what the issue is or what needs to be implemented +When someone has time they can figure out how this is failing in a few places sporadically breaking builds + for removing +Rev +There is no automated test case written for this change +Good job +I am with creating a branch for this issue and keep working on it until we have all the issues figured out and then merge this back in +Attaching a patch that makes changes in documentation for physical memory monitoring and configuration as well +svn ci srcmainjavaorgapacheopenejbconfig Sending srcmainjavaorgapacheopenejbconfigTransmitting file dataCommitted revision . +You can close the issueMany thanks +released issues. +Mark +Tom is looking at creating actions using annotations which is as far as we intend to go with ESB codebase. +For those people still requiring less than Java for development they should just use the older shell bundles as you have doneHowever I do not believe that OBR should be using Java classesmethods since it is not intended purely for development purposes. +Patch realized I named the patch with the wrong year I am attaching the same patch with the correct year in the name +If you se gmaven and the runtime then it will use the groovy stub generator augmented only to keep the stubs and put them in the right place +done in r and documented in Confluence. +Bulk closing stale resolved issues +Let me start with a fresh server and grab those logsVladimir do you have your logs and conf file to give to Mike +Thanks Madhukara! +Currently letting this bot run against and having local run running against hadoop hadoop all tests eventually pass except for known bad and +I would not have even noticed were it not for the fact that I had some copypaste code taken from about it a bit more adding a static method is probably a bit overkill as well so probably best to forget that + +I was able to replicate this when there were multiple versions of the artifact in different repositories +Thanks a lot + +In this design you almost force the implementers to create a new instance if the user needs to access the solr solr parameters +Patch using the wrapper class added by fix of the issue +And is a static property +Typically printing it as hex bytes feels appropriate to me crashing less soWe do +Attaching a patch which addresses this problem +Attached stack trace +Hopefully this will get picked up soon +The encryption tests passed after I installed the Sun JCE jars +Ive opened an issue about Languages not being encoded but actually most of the other are not class in WPS module actually sets also but the generated document does not contain documents for them +The proper fix for that is it make the have that dependency or remove the +was in reference to your statement that you were just having the acceptable function in the interface always return trueId have to see your logic of how you have implemented the acceptable method in your Action class but it sounds like if a parameter name is not in your Black list then you are having the method return true +do you need a special build now to move forward? what do you need right now from usim not setup with maven to test the fix but you have so i think were good there +I think this is the rd JIRA logged on compilation against The first was and second wasThe trick is that while we make the code compile against we have to make the tests pass for hadoop See some more details in I believe this is also related to is being addressed as part of but I dont think we need to close this as a dup +warnings also removedAs described wikiArchitecture User Interface is defined +Affects Version entry set to empty to indicate that the bug does not affect applications +The HEAD of the frameworks branch for should have this resolvedI will look into updating tests and doing a few more final tweaks tomorrow then it should be good to review merge into master +bq +Update the outbound samples invocation APIUpdate a little bit on the patch was committed to the svn repository. +Would you please help to try it +After looking into this a bit further it looks like there are two places of particular interest where caching of resource and paths will be useful for performance +rvmgemsgemslibactiverecordin sanitizesqlforconditions from orgjrubyin send +Is it by design or adding support for positional reference is on roadmap +there is no reason for a synchronous version +Can you try again? +I later found out what caused this bad error message on our side we did had a cif around one selectItem in a select box and this got via fajaxThis did lead to create multiple selectItems with a duplicated id which then lead to this very error +hasArray back +Thanks Dennis! +Jan could you please try to reproduce in Metamer +Nice +Presumably we could create a maven goal to copy to docs then carry on as usual +Stop server +Ok thank you +That would be the solution all that fixed although you might want to add some validation to the locations page to check to see if its a JDK or a JRE and react +Don you might want to keep this open to remember to commit it again when xwork lands in struts svn as the code was already packaged signed and submitted +position failsIll open a separate bug on that. +If I remember correctly the newly created servers were not having their xpaths added and so none of the xpaths for finding key ports were found +Added a test that uses and found a bug in the process +Trivial patchI looked through the code in this really does not need to be synchronized here +You can change which handlers are loaded for pop commands +I was in deep analyzefixreport bugs session so my mind went toward JIRA +Thanks Stepan +No Release Note Required JIRA pertains to documentation not text in question has been updated to refer to ISO rather than ASCII. +Rob please advice how to verify this +Just noticed theres a CHANGES file too +help lacks the doc info which I found helpful +Resolving with fix version ER to on ER. +Granted we are using Spring Hibernate so Spring handles injecting the into Hibernate +Yes I agree with what you have determined +Marking patch available for Ted so that runs +Thanks for the patch Dario Applied this in r. +Patch includes a basic testcase with Spring setup using embedded Derby database +For people who only want to compile Derby using the first JDKSDK we find would probably suffice +Thanks a lotBest regards in revision Richard please verify it fully resolves the Mikhail The fix looks fine +If you can supply a failing test case and detailed instructions to reproduce this error we can reopen itWhat is this for a nonsens? tony ns provided a test case and instructions how to reproduce is clear +Have the CTAS error out at compile time with a hint for the user to use CAST to specify the type +But I see how this is easy to miss maybe some comments on that constructor would be in order +I am back porting to and! Add which has the updated canon file for +I removed +Dmitry please confirm that this resolves the issue +LGTM +attach my patch +Alternative you say which has parallel behavior what did you mean exactly +Also now do comparison with the expected value as object whose equals method is being invoked +did you try +for it appliedMike +OKIve been mistaken about this +from ircrdaum that patch works +So the issue will not get lost even if the donated code might not solve any issue +can certainly support it +committed thanks +Im willing to guess that if the block report delay is set that high we have other problems besides a slightly granular random number generator +From the JIRA triage on December Kevin Cormier agreed to try to fix this one probably in the January time frame +Thanks Rick +Committed trunk revision I will remove branch now and start to backport +Close this +Even there is such a case my algorithm should not be more CPU heavier than the current algorithm +patch looks good +It would be nice if dependency order was not an issue like when using a typical compiler like javac +Move the fix to +Verified on JBT. +Will be part of the release +Heres a patch that fixes this +Hence the system vm template doesnt have its files accompanied by +Do you when repository will be synched to ibiblio? Cheers +Maxim Ed if you guys are working on this can you please add a testcase and update the patch +In oder to be able to make the comply with the Derby AuthenticationAthorization scheme Implement and which will hold a list of booted databases in the system +This isn t quite right as we do right? and we need to take it from one place not put int o strings in multiple places + please set the fix version +Can you refresh the phabricator entry also +If any new scenarios come up well open separate JIRA issues +Need to perform benchmarking in clustered environment and add documentation paragraphs. +Pull Request applied on the and in. +The sort order of structs containing sets or maps would be somewhat unintuitive which is probably fine if you just want some consistent ordering +From WFS standardThe mandatory service parameter shall be used to indicate which of the available service types at a particular server is being invoked +Translated message properties file to patch did not apply cleanly after the patch for was appliedCould you sync up and regenerate the patchThanks +Yes this is a problem +Great job thank you +The screencasts are now linked from the site awaiting Dave to finish the others +If a profile doesnt meet your needs please comment orNote I also had to set MAVENOPTSXXm to avoid an out of memory exception when compiling with Java . +Can you check if and can be changed to validate this? Look for refs to in those testsShould it be +Darn it +Otherwise these kind of changes could make the development noisy in the branch +Looking at the patch I do not believe this will solve comparison should be based on the classname to avoid any problems with classpath order and different Seam should only have access to one class within one classloader it should be enough to compare the class name +improved in r resolving +The patch for this connection pool change caused the problem looks to be a problem with a perfectly reasonable assumption for oracle does not mean testing w posgis goes well +Attaching +Why? Because it is stupid if user has to testagainst every set of rows in the DB to notice migrationproblems like this while EJB engine has all the informationto prevent bad deployment in the first yeah I understand that its faster to code to just tomake it throw an exception +Hi Marco Can you please attach the testcase files to this issue to reproduce the problem? Thanks +Current state of review is implementation and test missing implementation and test missing implementation finished but test missing implementation and test finished implementation and test finished implementation and test finished implementation and test finishedWill continue at condition tests soon +Just wanted to add that i tried show but the i find emit easier to work with that and i dont know if the query options work with show function resultI need the query options and reduce functionality and as far as i read show did not find anything relatedIf redirect is the answer can some one provide me the way to redirect view url to a specific document +Jconsole screenshots and activemq log files provided for scenario where this works doesnt work +Fixed following regressionsREGRESSIONRESULT SDKAPI FAILEDREGRESSIONRESULT SDKAPI FAILEDREGRESSIONRESULT SDKAPI FAILEDUsing is fine for test domain suffixes as the majority of regressions use synthetic clients and servers for testsFiled separate tickets for the following REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILED REGRESSIONRESULT SDKAPI FAILEDNote that for the following SDK API regression tests to passSDKAPISDKAPISDKAPI variable must be set to CONFIG INT Finally note that for the Cachepart regression test to pass a cache storage has to be allocated MBGeorge +Opinions +Oups not load the good patch +The formatting did not come out right +Defer to +Yet another tweak to make Hudson happy +This is troublesome +Daniel its best if you provided a complete project +We have a few autowire byName beans that rely on implementations and they do not work with Spring +Please change to if you intend to submit a fix for. +Would be great to get this into any update +It is not possible to change EARs module using the JEE Module Preference page because the Maven WTP EAR configurator reconfigures the module based on isnt any UI to change the preference +If an error occurs in the second stage we abort the standby NN since this is indicative of a corrupt FS and should not occur in practiceThoughts +should this enable proper packagin of EAR projectsIve updated to latest svn and I cannot make Run on server nor deploy work for EAR seem to add something to the server but it is not the ear project and sometimes I get two where each have a different icon +A test a fix and a tool for converting output to jasm +bq +point of keeping these issues open forever +Sending srcmainjavaorgapacheservicemixhttpTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixhttpTransmitting file dataCommitted revision +Wiki seems to be in a reasonable state now consisting mainly of design documentation board reports and some work in progress documents +Broke builds as all we did was thread dump +deferring this until we talk about it +What do the other ASF projects do for their wiki additions? Randomly considering it seems like someone somewhere probably considered this +But Apaches track history on making everything a separate project to release faster has failed +This is only supported on iOS on Build and most likely cant be done on all the platforms that we support +Fix has appeared in latest alpha manuals so closing issue. +instanceOf check should be added third party templates should be wrapped into toggling wont be work for templates that are not put into implies that all VPE templates are wrapped by class +fields ctor for convenience which simply passes an array that is used to create a set +swc from our corporate repository +The thread trace I have got doesnt tell clearly that the dead lock happened +Whats the impact of not doing the sync herePatch is fine for +Waiting for verification on Hudson. +The has been deleted until someone wants to write web services tests and a client for it +The use the coding convention when submitting subsequent patches +Please file the JIRAbq +To make this work you need to use class proxies +trusted was the wrong word sorry about thisthats because gateways are persistent this has nothing to do with the location of the gateway +There are issues related to iptables which I suspect is due to the missing files in the routervm which we are already aware of +The URL escaping would not be handled by the parser at all since it doesnt know the attribute is of an anyURI type +Looking at the code it does the right thing +I think we need to have an annotation on the server side which informs Roo that this type should be proxied relying on is a bit too restrictive IMHO +Addressing problems pointed by findbugs +What is the argument against CQLSH? It should be easy enough to augment with introspection etc +After testing fix for Local and Localpart failed on my machine +by +Undone the change in accordance with the discussion about the plugin purpose +Commiotted to branch and trunk +Please check that it was applied as you expected +Verified by Ilya. +As an err on the mysql side strategy I have taken the approach of sticking to the current double based round api +Yes on trunk it applies fine +What the patch corrects The two queries demonstrating JPQL case insensitivity were actually identical +Hi Weihuajust discussed your concern with Frank and we both agree on your suggestionDo you have the fix on your machine so you might simply update the patch with your new test and an additional methodIf it is not possible for you to fix this just return the issue to me and I will put it in line +Yes but you should get the same effect by simply running more Mappers +Hopefully we can work out what the difference is +Mass closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +fixed rev +How can we make sure that this is doing what it is supposed to do? I couldnt find anything in the documentation +I attached two patches that fix the issue for Date and Duration parsing in and any progress on this issue? Please this is very importantIve submitted the patches +Proposed fix +Can you attach a sample project or give more information about when you see this problem +Can one of you summarise for me? The workarounds are excellent but Im not sure I understand the problem +Pushing this back to I dont have the time and the motivation to provide an adapted patch for thisBut maybe someone else is willing to pick this up? +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Hi Arthur I agree that removeChar doe snot check for null have fixed that +i Split supervision in different module +All the code is committed +In general we should maybe open a followup issue to give these docidset classes a base test superclass test +Domaininstances now contain the addTo methods +Ok +Same as +Thank you Maciej. +Closing issues +Turns out this is due to the timestamp it is limited to one second resolution +The changes to the exponentially smoothed runtime estimatortests +Since is resolved that means this one is too. +Its not critical for but we should improve it +The jira number is +Is there any way we could make this an optional part of the build process and expect users to install BND if they want to use it +why are you rejecting this issue without any additional sollutions or answers +Committed we can figure out the karaf stuff later but I really want to get onto the line. +Resubmitting patch tests failed because a port was already in use +Sorry for the delay Andrew +Im just mentioning you can do the move to now and adequate theBut we should deploy a for this +But the sort should not be avoided because the order by is on and not and hence index used for cant be used to avoid the sort on +We could easily give also the raw in addition to the instanceSounds good to me as well +Committed +Hi Howardthis is becoming quite critical for us +done +Andrew do you mind looking at the patch and see if it makes sense for the short term at least +Please to confirm that they are there +It is interesting that you only see this on the and the IBM JRE +elharo since there is a new release and the fix for this issue is in there please verify that this issue is resolved in the new release and close this issue +It is only required to write a cleaner fix it doesnt matter functionally +Well Mazz that is the basis of our configuration verification +I am yet to figure out what exactly is needed for the th th sub tasks +Also needed code in r. +Also the exception we are giving out to users need to be improved +Hi guysSorry for the delay guys PTO and travel resulted in lack of internet +Ok you have tomahawk and tobago +Move along folks theres nothing to see here +Please give it a try and let me know whether it works for you! for the time being assuming that the issue is solved +We could consider backporting to the stable series onceif we test it on trunk enough and fix issues until it can be considered stableproduction quality +I will commit this later +Also violations were coming from different profile then the one used on Sonar server +It calls a CFC directly methodgetFoobarAlso I noticed that whitespace plays a very important part here +There were a couple of other places this needed to be addressed but patch appliedThanks Tim +You are right problem is not in but in myfaces implementation +I havent looked into this any further as I dont really understand how the classloading works +Amongst the providers that are currently in Knox authentication provider is the best place +sorry. +Would anyone please to review it +Sure enough thats a race condition +A new command can be added to that calls get and prints the results +This seems wrong it should probably be a different timeout +Added into the master branch. +Gilles I will hangaround until next are using eclipse noplan to go for +Note you must use Maven or above to builddeploy Roo now +Any ideas where we can place these scripts in gitWhat about build +If someone can post a repro then we can the bug +The modified query searches for empty string or null value to accommodate database platforms with alternative handling of empty strings +Thank you for all your great help in troubleshooting this +Also removes Windows EOL chars +Basically reduce the fluent builders we have overloaded in the and moveadd new fluent builders on the concrete xxxType insteadFor instance the aggregator has methods I think +file to the party still figuring out the build pricess here +Logged In YES useridMoved to feature longer investing in EJB CMP enchancements. +Patch for the example for the example backing beans +would not be involved since a redirect is performed from the authentication filter +Any CDI container handle these passivations +Part of a bulk update all resolutions changed to done please review history to original resolution type +I suggest we reassign to James for placement on +is there any news? +Good job the Hadoop community! +That doesnt work for me either +patch added +Please create another issue for the merge in trunk. +I just start to think if it worth it +issue seems abandoned for quite some time and I am facing problem which is related to this +Committed to trunk. +Juhani It seems like you committed the wrong patch +I agree with your opinion and have attached a new patch please check it thanks +Do you agree +Thanks +Closing per previous screenshot! +Very odd +the resolved issueLahiru +Wow both comments at the same time! Well the shows all components anyway adding another statefulstateless column shouldnt be hard +gschen afaik just click Watching link in the upper right corner of this page +Thanks John for the patch! Your change seems to work greatin my environment so Ive committed it to the trunk as revision . +I think talks about the same issue. +Git links now point to the actual git for +I was trying to import old eclipse workspace into JBDS +didnt get your comment it sorry what do you meanI just would like to reverse engineer my DB schema and have all stuff generated from annotations or +Otherwise I would switch in a snap as it means I could properly restrict which values they can use without manually checking +That is unless we check shas or timestamps +Feel free to assign back to me +With Rob being pulled in many different directions Im going to push this to LATER +I removed the patch file because it is incompatible with the changes just committed by Rafi +Thanks Jason! +For instance KeyValue +version of the patch looks good except a minor comment there are some tabs that cause improper indentation +I wonder if the commit also fixes +ab testing but failed to stand all the time it will failed after the cluster channel conflictsthings like when do cluster reading the channel is not closed +If we had needed to keep the Grab annotations around until runtime we would have needed Grapes as a holder but as of Java we could even hide that away with the RepeatableI still think we should eventually add the Grapes isolation capability so Im not advocating deprecating Grapes just indicating the current state of play. +Thanks a lot Harsh +Adds documentation of the parameter to error logging and test cases for invalid configuration values +Eugeneyes I did a single project beacuse subclipse does not subversionize separate projects +Yeah this issue is gone. +I failed to set up the debug logs correctly in the previous attachment +Not really sure what the point of this issue isThe one problem mentioned in the comments are managed by. +I guess we could make those stats positive only but there is actually some sense to a negative count in this case +Im working on a in +blocksBugs catch exception for searches on deleted feeds fixed too many category elements exceptionYonik please use the last added diff file +Verified on JBT nightly build +I will run a test sometime today with the latest snapshot of camel and get back to you +The LDAP Examples have been added and resolved in. +for i would like a modify property X to value Y method and leave everything else to be done manually a la +Im glad its helpfulIll take care of the work you mentioned todayBest regardssishen +any comments on this? you probably know this best +We can definitely spin this off +Included VCR tests havent been run as of late as tests may be migrated soon anyway +Resubmitting +This was fixed by making some files not look like scripts anymore by removing unneeded s and making the one script that should be executed actually executableAlso was installing a python program but python was not a requirement +JSW is now at with delta pack available +. +Having this allows to build RESTHTTP API that consume files and some form data +Do you see any areas where the DOM code could be improved in generalColm +The drools file that takes awhile to add to the drools package do to the from is now fixed in trunk as part of some general MVEL debugging triageing. +I see that this takes a slightly different angle at the problem such that failure to initialize log aggregation will no longer fail containers but rather allow them to run without having any logs for them +Yes youre right ClausI updated yesterday evening to use Karaf also. +Release will focus more on hflush than append +Committed to trunk and +btw if it wasnt clear Id like the domain +Thank you Brandon. +Applied +Committed to TRUNK +This isnt very nice but it fixed the problem for was against +Elliott filed but I cant seem to reproduce it there +removes dependency to and resolves this issue +Should this be +passed without errorsPatch ready for review +Sergio yes I could test that +I think it will be better to mark as unimplemented likeDIGY +Also I turned off hawtio authentication for now as it requires jaas that we dont use at the momentThe next useful thing would be to enable log viewIn the future we should use jaas for all default configuration and turn back hawtio +read operations number of read operations such as listStatus get open etc +It does not run green but it no longer appears to freeze +Bulk close of old resolved issues. +though that would be very I was able to reproduce the problem you had in the Groovy and Grails joint build but I was also able to verify that adding the cast fixes the problem for I found that it makes a difference if the super class comes from the classpath +Patch is now ready to be committed. +Merged on branch revision . +This issue is resolved with commit bdbbfcbacafadf +Already committed +Check with Len +bq +Fixed in Rev +in +If that means support for skip and top Im in for it as well +we just have to mark the method somehow so the runtime knows that for this method not all classes are resolved +yes +Just the annotations I would like to be kept in a non spi package +I dont think skilled developers will shoot themselves in the foot just the opposite it frees them to choose how to build apps +After further investigation this is an Android bug which means that it is only on the latest which is of all devices +This issue will cause a huge loss if multiple nodes restart at a time and due to the false cluster capacity assumed many jobs will fail in the cluster which is a performance hit. +Is there any evidence that you might be getting an exception on the server side just before the unexpected connection close? Ive seen code paths where a thrown exception causes the connection to be closed and the symptoms are sometimes not so clear pointing more at the closed connection than at the underlying exception which caused the close +Especially on the that are not assigned to you +adds the same manifest to as the manifest presently in and +I also corrected minWidth and minHeight which have as a default value +porcelli do you mind having a look to see if this has a similar fixcause to the issue you looked at +Colin this is required as a part of supporting Rename of Being Written Files +Why would it be harder to distribute an xml file than a set of property filesKathey I wouldnt want to make it harder on the translators or the community to contributemerge messagesNor would I +Moving all issues to the CR +JochenIn your last stament did you mean that groovy support setter overloadsIf so would special annotations be required or doesthe idea youve got in mind allow settter overloads withouthaving to resort to special annotationsIt would be really great if they could just work without any extra fiddling with annotations +Is there anything else that needs to be done before someone can be assigned this to work +Hi I did try to update the validator for USERNAMETOKEN and write my own validator +Tyler +No tests just configuration property changes on the stack +Added patch When compressed pointers are enabled use Virtual Memory poolThis kind of pool implementation reserves G of virtual memory and commits required amount of memory each time new are allocated +Its fixed now Johan you dont need to do anything +Will commit if test passes +Ive seen several times in my development environment Java Index +Could you please try to attach it again? tailing in revision Vladimir please check that it was done as thanks +This patch is quite independent and doesnt seem to have any impact on the yet to be committed build patchesGiri +please and if itworks as expected pls let me know +You cant convey executableness in a patch afaik but when this is to be submitted the submitter shall do it +Thoughts? +Id like to have a best effort PC transaction manager available in Spring +with new the exception is gone +Well just need Paul or another Selenium despot to comment on this issue approving the accountBen +So the idea about calculating and caching the different lists and only recalculate them on is gone +I waffled a bit on whether to go with just two defines but ended up hedging my bets and going with the standard header pattern. +Thanks Vikas JacopoAshish Vijaywargiya +Is there any remaining work for this or can we mark this as complete and delivered in now +Can we look at the user agent string and have omit diskformatversion if the agent reports a couch +I mean we need different versions of Avro in different projects and our supported projects might not want to share the same versionWhat do you guys think +So when you replicate those changes further downstream what would the value of olddocAnd in general anybody can update any document to have any revision history they want using theneweditsfalse option. +picketlink jars updated from to in product +The problem in archetype is related to another bug re right its not the patch but the test +This should in fact help us to preserve API compatibility longer since the mapreduce kernel will be independent of the API +Thanks Simone seen your comments too late after preparing the pull request +due to fix test with JBT. +There is no algorithm changesIn the first issue jthread library was moved to vmcore library +Thanks! +maybe only patched it in our local branchthanks for the directly links to the RPC proxy created stopped this is a good catchTanping could you check again + +Maybe that should only be some hint in the docsCheers of +The target starts up two server instances and runs Cleberts clustering tests against those instances +Good catch +Marking as blocker as multiple tests are failing +I have refactor old wizard and have created swt bot test for my implementation of wizard so I am going to remove this tests +Technically when you add files to this JIRA it already asks that you agree to contribute under the ASL so is that technically enough? people I will upload the files with the ASL header on sunday morning but if you feel like you want to try the translation files you can do it too I dont mind at allSi please tell me if you need a different header on the financials and CRM filesRodrigoPS +Full build including those tests works for me with trunk revision java version macosx +This appears to be a problem with the XSLT file check the transformationseparately with a command line XSLT processor +I have pushed a fix for where I replaced with in +So please take a look +Tested with a an older version of dll file i used instead of +When you open these html docs in chrome you also get the vertical scrollbar as well +Sorry for rev completed in revisions and +Verified by Richard +If this class were to go away there are other possible implementations we could pursue +Committed to trunk +In conclusion there is no good constraint implementation curretly in place +Closing verified issue +This JIRA had a status Closed Done +has nothing to do with ant +Is it ok +Yes I mean the Home Key +If I miss any thing feel free to fix them + has a side effect of reducing the amount of output of ivy since it avoids a bunch of resolve solves clearly this patch doesnt have isnt suppose to have any tests in it +So all issues not directly assigned to an AS release are being closed +The naming strategy should be able to handle that or we should enhance it to be able to +If you are checking the changes into a branch you can make changes to any of the files including the files +Marking Resolved as all necessary changes have been made and merged to issue has been resolved for over a year with no further movement +I think flush is not needed after writing each line in the log +Good candidate for merging to branch +So there was no RS to process the assignI did not think this needed since Id reproduced your scenarioLet me look at your changes +I cant edit comments it should read its only slightly less verbose +Maggagunda since its affecting a different component than the one mentioned in the Summary Id suggest making a new issue linking it as Related to this one +Pluto throws this exception upon the first request that starts a new session +ok the base problem is located in +We do have long running transactions that lock resources + intrunk r r +upload patch for CXF patch had been apply to Revision +DanI havent tested the except for a testcase +Committed revision +Attached initial implementation +Theres no such guaranteeand it causes transient failures +not absolutely esssential but needed for INSPIRE relevance which is the big short term market for Geoserver +edit the properties file you can do this anywhere in your classpath or working directory you dont have to do it in the jar file +Hi guys I assume you are quite busy with a new release but do you see some outline when someone might look into thisThank you very much +Could you reassign this to an appropriate person +I put issue on github +Apply fix to branch +What I want to do is as follows Change to an interface inside index +Sandy author tags are a source of debate at ApacheBy and large I think projects have moved away from using them +Yes I got your point +Fixed in CVS +Which version is this found in? There was an issue in that looks very similar to this that was already discovered +Are you still having this issue? Do you have your Android Manifest +the measure for determining workload should the measure be computed solely based on one region serverOr should this measure be relative to the workload on other region servers +Did you apply the patch? IMHO it is wong anyway to fix the test cases since it means that XML files genereated by might be incompatible when generated by from different vendors. +Please reopen if this happen in current versions. +Fixed in +If you want to use the unofficial unreleased please use the server adapter as its arguments have already been fixed re logmodule +Ill explain why +Thanks SandakithBTW please create the patches from modules folder else we need to search the correct file to patch in all the modules and its +Closing the issue +not sure what validation you are talking about but if it is from Hibernate UI it fails by saying project does not exist as you see I do have project in should be ok here +That is +I tested patch on my winvista machine and ran into no failures except for unitI dont think that failure is related although it appears consistentI have nit it seems the patch has unnecessary change +Justifications tests The existing tests cover the change +Please confirm before CP release goes live if it should be in or out +For instance our interface makes it possible to do a search that will essentially hit almost all of the records +on the client screen Would that solve your second requirement Michel + to trunk. +Dmitry Im a little confused it seems to be a Windows only feature so shouldnt the regression test be added to windows specific test class? Please correct me because if I missed something +Let the cleanup beginSomehow this crucial patch got neglected recently +I would think so +See files attached +bq +hmm +Filed +Ultimately I am confused by this since it seems like JDT is using the wrong value for spaces mode +Shortly Ill commit it to committed to +Feel free to it if you have any new findings about the problemPlease take a look at +Sample configuration file to demonstrate the problem +I can confirm that my code works on the emulator but fails on a Droid Razr M +Thanks RickIf you found the instructions sufficient I think I have what I need from you +Stevo could you open a new issue for this? Itll get lost +Forgot I was just waiting for the fix to show up in the alpha manuals before closing +Ill look into it. +Ok +I dont know where your file comes from +Uploading built pdf documentation for the change made by the patch for review +Queues and topic can have multiple JNDI names now +Thanks +Ill add the comment +This will defeat a lot of the value of this optimization +Fixed in rivision +Quick Path for the Zookeeper Servlet we didnt need the ContentPreview if we are browsing the tree and i removed the escaped Quotes Mark mentioned them on misread +But from the looks of this evaluation so far it seems quite clear that we will be using together with in our current project I dont believe that we get a permission to use anything that doesnt have a company supporting it or that has a milestone label in its distribution package even if its this much superior to other existing frameworks +Please get rid of the inspectorThread and the method getStatus to have something easy to use and understand +That code might be used as well to determine when the new CH has been installed +Default threshold of what? When using the Carmel method the threshold value should be between where means no pruning +Implementing +dynamically fetches it instead +Ah youre right +Updated the patch with Uwes suggestions +Using a simple identity XSL in place of the identity transformer does indeed fix the problem +It is worth opening a bug on this. +Since its a new feature it doesnt really have an affected version +And therefore upon cacheput if a certain key hashes to a Segment that is already full we do eviction on that Segment +TomIs this ticket still relevant today? +This pull request is dependent upon the pull request for +Attached is a patch for this issue +Hope Im not missing anythingWe should look for any other unused jars +was already released that means it cant be fixed in patch that incorporates feedback from Gabrielle Crawford +I am resolving this issue as duplicate of the other one so people can know were to look. +The more versions of the same method that users use the bigger the support matrix getsIm not particularly in favour of having multiple constructors either +I did not get last version from svn +Verified in ER the issue is now present only sometimes so it seems that it is really related to timeout reopening to update release notes. +moving to released +Feel free to try to reproduce with the current trunk and thanks for reporting +The patch resolves the +Was hoping this was something simple but it looks like it is an SQL issue in the authz code +I look forward to it being included in spring core +So it seems not to be possible with asimple prefix but we should have a completely different mapping fileAdrianyour suggestion in is interesting I used an iframe in the ftl to render an external +The patch has been committed to the trunk and branch +These commands only trigger puppet manifests execution so it is not critical but still may be dangerous +I will take a look at possible tests which can be extendedadded for the new peer manager support +The approach looks nice +With maven the is downloaded +Cannot see any problems good to go +It seems trunk compilation is broken after the commit can somebody add the new file +committed this to trunk with revision +Is it too much of a statement to tell people to make sure there is always lots of extra space on the disk that holds their database because bad things can happen if they run out of disk space in the middle of a database operation? It sounds as if that is the problem not file compression +You will notice that the resources get placed into the current working directory instead of the projects base directory +Hi Talat no this patch is for the branchx branch only +This issue is not a duplicate of +Im changing the code +The wrong dependency resolution picks instead of which seems to cause a while loading the wagon which in turn causes fallback to the bundled version. +Ecleemma always shows report for all loaded classes and I couldnt find a way to include in report only is really simple but I have no commiter rights to releng anymore +Let me take a look at it +Thanks Robert +It would be nice to see this to the branch +It takes more effort to not do that and its something thats likely to be done wrong +Thanks Ravi +We might want to revisit it in +Id like to commit this soon +I dont think this should be a blockerlets discuss it on the thread instead +Has the situation improved with +Please correct the status +This does fail with C in all build types and in debug build types +Thanks for this clarrification +Delivered on . +This means introducing an incompatible changeNo I just meant it should not be required to use a JAAS configuration file +This patch adds a new wizard to create a Web Service using +Does this mean a separate file outside the final +Sure +from +release blocker +Bulk update to fix version +Sian I tried the but seems it causes a couple of NPE in seems in which case the connection field in hasnt been inited yetFurther I think the two test classes in the patch can be merged into current tests in moduleslunisrctestjavaorgapacheharmonylunitestsinternalnetwwwprotocol did I missed any reason they are separated to another package +A Brew build will replace that on CR +from me +Where should I link +Attaching as the first one had an extra unused import +Sound good to me +I had never realized that extended the other mocking frameworks to offer the extended functionality such as Mocking Static Final Private Constructor and Partial +Is this reproducing always +As a someone could simply implement an which matches on and then does additional routing based upon the +I have made various fixes over the intervening years to ensure consistency in dispatch to overloaded Java methods. +This patch requires changes in common mapred and hdfs +Can you try to bisect this on the branch and see what commit caused it? Or provide a demo app that exhibits the problem? I still suspect this file is getting loaded twice for whatever reason +Does have to be done in to avoid compatibility problems +Just ask Scott +Any idea what file is openlocked +They run under dash on linux and binsh on BSD and csh is not as I originally thought POSIX compliant itself. +Thanks for fixing this issue! +You mentioned the HA branch has it fixed +not yet reached so no locks taken in the db. +sergey let us know if you think this is fixable before GA +Reason a new blank database was created in was to create a test case and demonstrate performance issues +maybe there is an error and the hostname to sync is mavensynchomeusersoououaibskymavenmreporsyncsshOuaibskyouaibsky +form +The javadoc for getPosition has param but lacks return tag +Id accept a patch correcting the issue +Maxim Red r does not seem to have the issue with the playback of flvs however it contains other issues with the file stored at wrong position when doing aWhich revision of Red is the one that you need to update to in order for your API and RTMPS fixes + +I have spent a while configuring and testing Cache and as cache provider for Hibernate and respectively under +Dear AntoniaIn the process we found several other issues with the SQL transformerencoding did not work on Oraclewe had the need for XML inside variablesThe attachment contains the version we use now with success +The fix looks fine. +is old and unrelated we did MB blocks because we didnt have durabilitybq +it +Wont fix +looks good +Patch for a slight modification to to get the Hibernate unit tests to work. +I was wondering how to fix this till I saw your patchPatch is inline with my understanding of the race conditionIm fine with not having a test for this race +The size of the font isnt a problem +Attached the patch +The out of date information has been replaced with accurate content. +The CDK evaluates the xiinclude statements generating a validIts not worth spending the time to dig into the xml parsers validation to suppress this error given that this file producing the errors is an intermediate file and not present in the delivered artifact +I wont likely get to that until later in the week +I will recheck how to do it right +Ted thanks for taking the time for review +Bulk close for Solr +You must upgrade to Eclipse +The following should work for and the tables that are generatedSET AUTOCOMMIT LOCK TABLE WRITE +The file contains the source code +I have just got maven aspectJ working for hbase +Instead using lazy approach verifying cacheTimeout on accessAlso implemented few comments form Jakob +Attaching the first version of patch +The rd point from my previous comment is invalid +Pushing it to +Tim could you submit this as a diff u? Its really hard to tell what it is youve changed here +Hi NandaniI got your question we need to update the method to deal with all the rules like in your case if its a gift that will be added in cart so we have to remove that product if we unapply the promotion +He started server clicked on then the control tab clicked restart +Are you sure you committed it +Updated the files per comments from MikeI restructured the list in +Ill compare the launchers +A new directory ctestsautobuildwsi has been added which contains an Apache Ant script and config file templates to allow the Axis C tests to be run through the test tools +Ship it! Rishi Ive committed these changes to the trunk in r +However if this is a big problem here which it seems to me that it is Ill take a look at those methods and maybe find another place to put them in order to not have to cast to +Its unfortunate that our build system doesnt let it run as expected +I hope I will find the time to clean up the code and submit a patch in the next few days +Submitting +This is the core issue in this jira and should be a rare eventThe MR app master sleeps for s before unregistering with the RM and reports success meanwhile to the client +I dont understand how else can we do this +A positive test where column aliases are being returned you can add a dummy function in contribtest which returns the same value twice +Test application + are expression operators that are expected to return a +Well probably put it in production as soon as becomes a release +Im setting this as minor as the issue is the followingThe server didnt set the for the Session on the first message because the message was rejected due to security on that reason the following messages didnt get a message set what caused the NPE +wouldnt make sense for this check to be done by the RM maybe via a new method in the scheduler API? By doing this we are ensuring scheduler impls dont miss doing this check +Now builds on windows under both cmd and cygwin thanks. + +quick note that js patch for the png images doesnt work if the png images that are defined in css instead of the html markup +testing zookeeper via mvn Pintegration clean install +Inheritance JOINEDI dont think this model makes much sense in a solution +Includes and and tests +This is where we might analyze the set of andor service properties of registered to find out which events must be sent +So this jira can be closed +Class with utility function and main that calls it on an +When I turn off validation the error goes away +Code and test changes for this issue were committed as part of to trunk and branch +This bug occurs only when the application is started from the Intellij Idea +Tests passed cleanly for me +The data sources that I have mostly seen include document like things that have fields +So this issue also affect the branch and a new merge will be greatly appreciatedIm not able to reproduce this issue on the branch +Test case that shows the issue +Created a cleaner as the default +Are any deletes made in the above benchmarking? Might try to simulate the same change in a small unit test +Otherwise we would only see that the child task exited with non code and we would not have any context on the reason for a failure +Not deployed yet +that is exactly what I had in mind! The duplicate fields check is really cool the return value of map is useful and is a simple check +Added code thats true +like written in the mailinglist the problem is the m in the pattern that can be a placeholder for Month or MinutesThere is no way to detect the meaning of the m +war leave the for the example module for sources details +is solvedProblem was on extracting the Thanx for helping! +Fixed +If it does not exist it is never created +Straightforward patch to add a warning message +patch applied +We are reusing the send cursor result now for regular results which was sending back a portal suspended even if all rows were fetched +Filed ticket for the plugin class duplication +Committed the patch into trunk with revision +it works! thanks for the patch +It maybe though I am not certain that the intent is to provide a way to test your config file by starting up CC and seeing if it validates everything +Usually these applications can tolerate some during the region failover or other time instead of caching them in memory +JasonWe limited Synapse to as the script mediator needs some changes to support but I agree that if you are not using the script mediator it would be good to be able to run Synapse on +Fixed the missing images error Added the Languages Appendix page +This could also be done where the trace needs to dump a large amount of data +JBDS patch applied +This seems to be related to the latest changes +Could you please regenerate the patch? Thanks +Im going to revert this because this seems to have uncovered another bug where map output comes back as OK but has internal error in the body +The issue came up awhile ago Carl Reed said there was an easy process to make this happen as an official OGC thing +Hmmm +However earlier in the log it indicates that the example is being undeployed +Add onmaskclick revision +Im dead itred and I dont remember what was the conclusion bt in any case this has to be fixed + +Heres a patch that implements this +Cant say that Im not disappointed +bq +Had the correct version in my workspace blotched the copy over to the vendor trunk +fixed and should be applied fixed in +Resubmit patch to hudson trunk test was broken by +sigh +Patch submitted in pull request on I get some eyes on this please? Im not able to decide if this is good to go or not +bq +fixed +fixed in proves it. +Now works on Derby DB Oracle and databases +Thatll be fixed as a part of tracking that down! It was weird as hell +Please review it +Thanks Eugene the modified patch was applied to BTI at r +As far as functionality it may make things slightly simpler for the user rather than using the config command +A slight concern I have there is that I am not sure how well a change in that area will be covered by the existing tests since reproducingtesting it in a normal way requires an ant script to be run + +I too am having this issue +Apply +Patch with +I reset perspective and menu is there +Added license to the java event for pt + for the second option +Here is a patch that does what Ryan has pasted into comment +bq +Applied the patch with thanks to Mike +ping +OK +Merged up to trunk revision There was one conflict in some test but they now pass +As long as we have an upgrade path should we add custom CLWe can do the same as with collections +Note that I incentified this issue via Donay on Sept +patch looks good. +The issues are now closed. +What we can do is to make sure that if we receive a cookie attribute enclosed in quotation marks we send it back enclosed in quotation marksI looked into possibility of providing a workaround for this issue in but have to conclude it would require too many ugly hacks or code duplication I am prepared to live with +Changing to CHM helped +Thanks +So all issues not directly assigned to an AS release are being closed +I dont think this is a blocker for +Committed to trunk +max is as it says value in MB +I know is already on a way to be committed +I think its ready +Closing all tickets +is the correct patch +Id like to get some feedback on what other developers and commiters think +I added the files and also changed the name of the +An objection was running flush in a situation would be risky +Im still trying to narrow it down but it would be good if you could confirm that it either works all the time for you or whether subsequent runs will fail +A configurable check could be added and a configurable constraint +I tried to just discard writing log records to virtualFiles but rollback failed because it was flushing the log and reading undo records from the storage +its able to reproduce easily in only HA installation +Should have read the code furtherWill integrate the new patch +Thats fixed in +Running two repairs on the same CF but with different ranges should be allowed +Reopening this issue since the got put into in while is still listing it in issue is now closed +The attached patch adds a couple of overloads of stringinsert for pointer and constpointer that dispatch to an efficient insertion algorithm but fall back to a less efficient algorithm when the range being inserted is a subrange of this +absolutely +It appears this class is not a strict superset the functionality is not there +bulk defer of featurestaskswishes from to +Mitul can you come up with a patch for this +Perhaps we ought to close this as Wont Fix instead of doing anything to encourage running the tests as root +Deactivated and hid the button +While Im happy to have a such a quick response and a response that does take care of it for me is there a plan to fix the problem rather than fix it +Have you resolved this issue? +I misunderstand description of this issue when I reopened it +Please if there really isa bug here +Reopen indexjsp +submit my patch +If you put together a patch with the right licenses and both files itd be my pleasure to push it throughThanks +Im reviewing the current user guide and Id like to know where we can best put the document you providedChapter +receive a request based on v and v back post based on v onlyp change to bind v + Is there another way to execute the post section from the kickstart file in the guestfsYes we could upload for example a bash script and execute it +This was fixed with the inclusion of jBPM +Includes archives but we need to changeem to the archives +ah nice I should investigate what I actually need in ruta +As I am sure you are aware it is about impossible to get a Grails generated war file down below the MB limit to attach +Patch to add Cancel button and fix closing window +HEH perhaps I should have acted on my FIXME last november then you would not have needed to report this. +see subversion tab +Removing the patch from the kit in favour of instructions on the download page to update the AS version of resteasy from to. +Hi AnilThe patch doesnt take into account the effect of using the select all checkbox +If its GFDL then its category X +Yes please so that I can update it today +Note Looks like it might be related to but I dont think so +fixed at r for trunk and r on as the form field gets empty to in some cases like curl d +The backported patch passed all the unit tests in my local machine except which also failed without the patch. +the handler should called after the rcptcmdhandler or the grylisting should be done only based on the ipaddress +Can we change n to something like numPartitions? Other than that it looks good +I wonder why? Let me ask him +Thank you in advance +Please verify +Committed this to trunk. +pull request merged +We still arent building the release notes in an automated fashion +The class invokes the interceptors assoiated with the Action +All services with such a property are registered in the domain and have the type property set toIf there is more than one service with the same property the one with highest service ranking is usedRev Check Tools initial release done in Check Tools initial release done in +the fix version just keeps moving further and further +It has been almost years since a release +Changing to an improvement +Management server start server log from the setup +I simply imported the in my workspace opened it and clicked Set as in target editorHowever with such a big amount of data and such a far away server its very possible that the time to get TP can be elastic depending on your ISP performances and the time of the day +Created issue based on further discussions related to this +That is why there is a flag written into the header +As of right now it seems that the acls are disjoint between mapreduce and yarn +Like maybe just showing the description without expression resolution +I made the following changes +This issue seems to be resolved with the patched jar file +great news its working for your test case +attaching and run fine on my setup +Closing resolved issues for already released versions +Annotations are meant for page specific security +Thats starting to push the scope a bit but Im just interested in exploring the issuesconcerns that people have +In the current state of the project the of Commons tests are running under minutes time +Yes lets do that +The subtask associated with this JIRA represent this information +This is a critical blocker for us +Need to backport it to +Committed at subversion revision Touches the following additional fileM javatestingorgapachederbyTestingfunctionTeststestslang +Applications may be depending on the current behaviour adding a new command to do this functionality seems like the right approach +Ill redo a patch with the required comments and references to this JIRA then try and find an obvious place to insert the information into the WIKI page and make a patch for that also +Patch for branch and trunk +Modifying title to be more general +Had to add a condition at the end for shutdown +Thanks Dib Ill try to get to it soon +You can verify this by adding a project reference in the Build path dialog +Sounds good to me +Closing as per Tamis comment. +Though I can a use case for your patch as well not just what I had in mind for this Jira +Are you fine with tracking the known test failures in their respective JIRA +Hmmm +This is not the case anymore for Spring Integration option is to exclude all of the spring dependencies from spring integration and manually include the artifacts from the EBR repository + +todd argh on the findbugs Ill update them is it ok to amend or you prefer a revert? AFAIK it is wire compat as we are only changing the generated packageclass names not the protobuf stuff or am I missing something here +TimThe only difference between Java and Java is class s package name and so far looks like its difficult to handle this problemThe attached patch use DOM to handle the namespace prefix instead of so it can work fine on both Java and Java +Its set to in case tester arent fast enough to get around to all nodes to start original config has no maxblocktime set which would take the default of s +Please post a link to the BPEL section in the online jbossws user guide +Patch and regression test applied at +YEAH! +Where would this go? ALL the components files? If so this should be assigned to component leads +As an experiment I ran the regression tests with the patch +I just wanted to mention that it would be nice if you could add interceptors either at the end or at the beginning of the interceptor chain +Anyways should the relatively easy to implement this +It can be unzipped under scaitest and run from there with mvn +The server doesnt complain about this at all and works correctly but the logging setting simply cannot be loaded so the server probably uses some defaults +If anyone knows exactly what to do and wants it implemented faster feel free to take overThis is part of an effort to close old issues that I have reported +However porbably for compatibility process this has to be permanently removed from package that is public availableThank you for your helpBest regards to Nicola de how found the solution for and it also worked for youRegards you using Hibernate alsojbossjavassist is Ognl core dependency so it cant be removed from xwork or struts itself then it breaks other users applications +As we all know Hive tables support various other formats such as text file sequence file RC file and Avro file +I never got around to resolving the test failures +New JS File VPE wizard screenshot is attached +what we would like to do is be able to change snitch class at runtime which is not currently possiblenot sure if dynamic is configurable already if not that should be I believe we originally added it to support dynamic configuration updates of a single snitch +I only see a createIndex field but its only used to drive the lazy index creation +Im not sure if we should do the same thing for repair what do you think Jonathan +a is never to use the property attribute but always a nested struts beanwrite +make binary comparator works +also need to add documentation for how to setup arc on your system +Itd be nice if you could even instruct Avro not to put the schema in the metadata for known generated schemas though I dont know if thats possible or not +This issue has been fixed and released as part of release +for all patches +Hi DavidIt should be easy for the to remember it has the lock or not by adding header to tell if it has to lock or notNow my question is how did you hit this kind of issue it could be helpful for me to wrote a test for it +Ive committed it to CVS modified so that it reports the error to a registered as specified for SERIALIZEERR +Somewhere text was put into the tables +Otherwise ill need to try and get access to the hive log for the test runtest is unrelated to this change and is being investigated independently +There are couple of things which are worth considering in the refactor +What Frederic doing is from his client application he retrieves the anonymous svc client and create a new operation for it and call sendrobust with that op name +kevan pointed out that although the fix makes operationMap work it leaves all the other fields in an indeterminate state +Closing resolved issues. +I will try to put something together and attach a simpler sample app by next Monday +We still have to investigate this +Unfortunately well have to clone the protocols in PB but thats okay +Thank you Benoy. +It is not being picked up for some reason +applied suggestion in r thanks +after being able to reproduce it locally +a minimal test application could be useful but I think that in this case we can live even without it +I applied this patch but seems to be more +Its the simplest solution I guess +This is standard Java and should work on any JVMThis does not affect the behaviour +I do hope that the YAML file doesnt have numbers in front of them like you have added +Just to make sure +I cannot remember even what fixture I was stepping through when I saw this nor a do I have a repro +In my case the typename begins with +Yeah from core perspective specific code in should be avoided and should go for wrapper similar to that of inputFile restore +When you load resources from classpath you should use the from +The attached test case contains two projects testplugin creates a simple plugin test references the pluginTo replicate the problem Run mvn install on testplugin Unpack the installed jar remove maven and repack the jar Run mvn compile on is a dupe but I dont have the issue number handy to closelink it +Initial attempt at making the core swapping do the right thing with regard to the JMX tree I am also working on patches for the missing from the patchI really need to get better at this Old JIRA cleanup +I think should be fixed in this jira +amitev as far asI understood from you then you cannot reproduce this issue on later versions thus im closing this bugReopen or create new issue if you find other seam related memory issuesThanks. +This limits the effectiveness of the cache see forthcoming note at +If a region has not got any update for some time its latest wont be updated at all since there will be no flushes +Okay all patches applied to CVS HEAD +Closing Jira. +Duplicate of is going pretty well despite the fact that sucks in +Sure it will be a pleasureWe have a complex network so I could try even some exclusions ldap ldaps etcThanks a lot. +crc files in +Thats the expected metadata then for an attribute that cannot be modified after creation. +With the style sheet these look great +The findbugs warnings seem unrelatedHowever the failure may be related + changes a test case +Removing only xmlnsweb attributes doesnt help to fix test errors anyway +Thanks John +For that I think a vendor specific param makes the most sense and perhaps we could eventually get it rolled into the WFS spec +Merged this to and also committed the patch + +Its a very small change +So we should pick UNICASTFC and enable RSVP only for state transfer commands. +One more one less But its a good idea we can propose that you join the team! +Please open for specific needs and issues +It might not be a blocker if we notify users that they must specify a header width +My plain and simple problem was that a did not have an empty constructor +Fixed into trunk rev +I see +Please check that it looks as you expected +Since it is creating the znode and it has default value ENABLED adding a check looks to be redundant +The was not intialized with a object leading to NPEI am uploading a patch that does the same +HOT FIX srccouchdbprivcouchjs enccharbuf Replace bad surrogates with ? instead of terminating with error +Right youve switched it off for text inputs +I fixed it so that when the user uploads new jars or saves the POJO model it refreshes the package tab if it is opened +big combining the help usage and argument parsing all into one place would make it much easier to keep them all in sync +trivial +This is a bulk cleanup operation to remove Version tags. +Heres a patch that causes the Java reflect implementation to treat Byte like Short with tests +Moved from SWS to SPR +Yes +As I mentioned earlier I am new to webservice thing so please forgive me if I sound stupidI would actually try removing that file and see if that helps at all +Please vote for the issue. +If no data is provided for the new node when using the create command assume an empty byte array +Which may be the case if the close is happening either when the server hasnt yet completed starting up or is in the process of shutting possibility is your server is running out of you can send me a test program that replicates the issue that would be a great help +This came through stackoverflow +patch applied please test +I do think we need to start all bundles managed by fileinstall that are not persistently stoppedThats my pointIn my use case I didnt stop I just temporarily removed theTherefore I expect that when I reintroduce the in the watched folder is restarted +and your comments are very welcome +xml files +Added the default to the can you please address my previous comment either with code change or why that change is not necessarybq +Added license headers to files in commit to master branch bbdffadcdcadcHopefully this is the licensing issues flagged by Andy now dealt with. +It should allow tests that use an embedded Cassandra to use the callback instead of generating a on the fly based on a template +The other idea I mentioned in my previous comment could be called fair enough +Notion of Trash or Waste basket or Recycling will implement a waste basket +But again Im good discussing that but what Im mainly saying is that we cant choose the best solution unless we clearly identify the problems we are trying to solve +rev after release +When the widget is then executed and requests some service must it not connect to the very same server where Wookie resides? So I dont see that this configuration ever needs to be doneAside from that I think the solution you propose would be viable to solve the problem +Ive done numerous comparisons between the web applications to see what might be different but as of yet I have been unable to see any differences that might cause the issue +Due to holidays and a few tweaks it may not be till next week +This caused a deadlock +java tomorrow an give it to you +I just tried and got the expected reply +Ive got versions in our SCM system that I can use to evaluate a fix though going forward to help +Fixed missing take a look +Im not sure if anyone else has been bitten by this but it is the first i recall being reported + +I would write mutations flushing after each mutation with varying numbers of threads +GERONIMO +patch looks good +You have a patch for us fnord +I have sympathy really I do +Spring IDE uses internal Mylyn API which has changed recently +Thanks +isnt going to make it into so we need to get this patch committed without it +If they are only under thirdparty and not in the distribution where is the conflict from signed vs unsigned coming from +There is a separate issue for forcing the srs of coverages +I suspect at Yahoo! youre using some configuration that blacklists a TT after X disk failures +Renaming add to add may illustrate more clearly +Thanks! the source of the duplicates test works well now +This is not tied to a specific release so Im moving it to wish list +asmx files +Mads is correct +If there is a common issue across all tests then we may invest in fixing that +The warning is definitely confusing WARN rint Extender Unable to create directory for Karaf Archive timestamps +If you like it Ill make unit tests +Does this call for release in the very near future? Most Solr users will probably not be able to figure this one out yet it will hurt them +Patch needs a test case +describes some experiments I ran to compare transaction throughput with and without the optimizationHere is a summary of my findingsi The throughput gain is impressive if your transaction simply grabs one number from a single sequence generator +Please review +Or implement for both thread groups in the transport +I downloaded and replaced only that file in my project but did not see any change in behavior +Im not sure if we can do anything herethe NPE is inside WSSJ code Maybe try to report this bug on WSSJ mailing list +I checked M snapshot of revision the bug is reproducible on interpreter +I tried to run latest jboss portal which is with jboss and as expected I see problems witha +I added tests to both of these +sshauthorizedkeys in their homedir on people +Thanks +Ill file a jira to make them all the same to stop confusion in the future + tested but the implementation can be improved to not visit twice the AST +Removed unused imports and variables +please open a new issue for thisThanks +So this option should also be added in the list of useless options in +Clear cache +Postponing this until there are better tools +So nothing this weekend fixed it and something prior to the th broke itThe reproduction is pretty trivial +I guess Phan means the File Explorer of windows +Fixed +I have used an Eclipse feature showing a warning if the serial is missing in a serializableThis catches now all elements I have included in my workspace +This is a classpath issue did the tomcat have an older geoserver installed in it before or has another geoserver installed in parallel +a nice error message like you cant use the select attribute with xslcopy would make the product much more friendly to developers +The all jar doesnt include the which you need +The customer data have been an error on OISG +Please review the suggested patch +Prior to this multiple values werent checked for until the doc made it all the way to the +This may be overkill +It may be final and good to go but personally Id write some more tests since this particular area has been lacking before committing +I plan to commit What means LGTM +I talked to Igor about this bug and although fixing it might be a performance hit we agreed that it is a valid bug and that it should be kept open if only to let users know that its a known problem so they wont open duplicate bugs. +This may not be good because there might be thousands of files needed to be opened by a single client +With everything works fine +Used Curator so code is small +I could be persuaded the other way with a good argument. +Plese test and comment +no response +I removed it and set the termOffsets attribute directly for the field +Yes +But now that I think about it it may crap out when coming back to read even on a recovered file +You are right the function works as intendedI still think its too verbose but agree that its a style issue I guess if the commiters like it its fine +I you could integrate jspx this with or to make it possible to take a mess of JSP and leave it clean and ready to be integrated with FaceletsJSF it would be awesome. +Merged into master. +Because its doing findBundle? Lets assume it is +Now that we have tags and V its not necessary to use out of line storage for cell +This issue is solved for now and further languages will be added on availability of the new languages can be reviewed by other Pootle admins or translation project admins +Where HADOOPHOME is the PREFIX directory of +A patch as in a file made with cvs diff against the project so I can see the actual changes instead of respect to DTD handling why dont we just use the DTDXML handling in WTP +Remove the constants in Table keep the use the Roo backup command and attach the project +I hate subcomponents in jira why doesnt it allow updating fixversion based on parent one +To get things to work best set to something local so that generated results on one VM dont interfere with the others +Would you mind my codeMaybe we should also consider to have a shared implementation for the waitsignalcancel mechanisms for the and the otherwise wed need to implement similar things in two different places +I was talking to Christopher tonight about this +I checked in and the was not provided in the deploy folder by the activemq feature +Any thoughts on this would be greatly appreciated +According to some on our Maven artifacts arent part of the official release so we shouldnt list them on the download page next to it +If I theoretically did have control over it I would just delete the ognl directory and all of our problems would be solved +It also makes the C and Erlang servers recognize the new enum as an alias for TCALLOnce this change is widely deployed we can make clients send a different message type code if they are making a oneway callOnce that change is widely deployed it will be easy to produce a generic proxy that properly distinguishes between oneway and calls +Yes JRECONTAINER should not be placed as the last entry but the bootclasspath entried should be placed in front of the putting the JRECONTAINER entry unconditionally last the class enviroment wont look like the one youre deploying it in +Windows XP +Ok will look up how to do whole project Seam Managed Txns didnt help with this problem + means more than just synchronising changes to variables threads that read the variable need to synch +There will be no equivalent for them in. +The following projects have the wrong artifact id +patch +Thanks whomever did this +Ahh +The problem with adding the skip behaviour like the other goals is that the file to deploy has a default value + +svn +Heh +I cant reproduce the issue on Fedora Gnome +One guess is that this gives many fewer terms +Thanks for cleaning this upMy only comment is that I think we can do away with the interface entirely since it is a private inner interface that just has one implementation +If one of the blocks is invalid the whole page will not compile and you would have to provide an override for every block in the page +Attaching a patch that addresses Owens comments +create a temporary JAR where we synthesize the classpath within the approach pretty much removes any classpath length I would recommend us doing that in the release when I intend to simplify the internals of the Maven Plugin to align it with acceptedMaven best practises +Commited the patch with revision +Alternately could speculation cause this was it enabled for this job +Here is another patch that avoids the duplicate class +We cant deprecate the s accesskey attribute because its defined in +bat file launcher scripts +buildr should work right now. +Pauls revamp of the validation subsystem makes the existing component play well with the rest of validation. +also be backported to This patch also fixesI guess options are the tests in Move the tests into the classI prefer +The versioning might want to support an optional build artifact field so any informalexperimentalinternal builds can have a specification which is readily accessible to the users +It was tested on CygwinXP +Hello TonyThe patch has been applied at revision r +updated the README including links to the incubator website go installation +Given do we really need this? In particular typing in a uuid sounds error prone and tedious vs resetting the bad node +That would be great +To be more precise this was fixed in Sonar. +Thanks Doug +item +This manager should deal only with as in most of the cases it searches for the vms in uservm tableCalls generic to all types of vms user vm system vms should be placed to +see in +I agree with knuts analysis but wouldfeel better if I could cause the IO exception without the code change +From a Jikes RVM perspective its clearly the name space in which the bulk of our code belongs +Do you think its safe to postpone the logic to create a separate thread for server side actions to a later release as currently the upgrade action only does one DB write +In particular youve undone changes that were made by other jiras +Committed with revision +Thanks again +Committed to trunk and branchx +This is great! Thanks for sharing your work +and tests looks great +Removed dependency of to to avoid circular dependencyReplaced with where possible +In fact I guess I can just migrate to the Settings class and get the property back as a it wont be able to find the given path when multiple ones are specified instead of failing with a class cast exception +Socket or EOF exceptions along with NSRE or WRE while talking to a region server can trigger the client to ask the master to engage reassignmentrecovery protocols for the affected regions so I think this issue is part of +error in the NMS client configuration this only happens when you pull the network plug but works fine when you shutdown a broker? It should like a symptom of not having an inactivity monitor yet in +Darren can you please have a look at this at your earliest convenience? the fix is faulty previous version for values need handling separately from all others +resetting P open bugs to P pending further review +As there is a test attached even if in a tgz archiv it should not be the biggest problem to run it and see what happens +Seems like additional synchronized blocks should be added to code +Id be willing to help in any way I can adding tests committing the change etc +Would you have a problem with me taking this on +supports EE only +Committed to trunk +Attaching new patch that resyncs with trunk +Resolve immediate issue by using to generate the concrete class forWe still have a wider issue of documenting how this requirement affects consumers of the API +Other files too find + does not provide context as to which attempt to bind resulted in a CCE a message describing the node that resulted in the failure cannot be constructed +I guess serialization breaks due to being +Removal of what appears to be random characters +srcjavaorgapachecommonscollectionsfunctors +Committed +Thats great! Thanks a lot +Added simplersmaller patch from source +Done +And would be nice if we can add few platform specific QAs as well +Added an entry to the FAQ. +I noticed in a recent patch Maja removed a hardcoded setting of the master task id and set it with get type callsDoes anyone know exactly where the task id dependencies in Giraph are what they are etc? Are there any Giraph tasks that need a certain task id for a job to run? How about Hadoop or MR dependencies in the IO formats needing this? Thanks! +It will have to wait till +fixed a bug in am applying now both patches +The site generation is a more than minor reason that its being investigated as the tool of choice +Or it could be done just as well withlistt SQLLINQ syntax be implemented as a macro like C does it +Corrections needed to doc. +Maybe +could fix that +Please attach a sample project to reproduce the issue +Assume closed as resolved and released +So this is the part I dont get +to using a standard Ive applied this patch in r +Changed the to open with todays date when the input fields are empty +The new problem seems to be a different topic and a new issue. +Can you attach an example project that builds in the wrong order? +Sorry original patch was in error +wrong issue resolved reopen +about the overall intent of example including raw performance and the issues of trying to make it Ive convinced myself to this proposal +This problem is a showstopper for usBTW were using wrappedliteral +Combiners are not compatible with +obsolete +The relational MED extension properties can no longer be removed from a model by the user +a suggested patch makes possible to use WELD wih closed this one as duplicate since is the exact same issue. +Patch for this one +you verify this issue against +I will check for other occurences of the same case in all other +Server is Linux using for the FTP night I did some more work to reproduce +bulk close of all resolved issues as part of closing items. +I missed that part +Released. +LGTM +I support the comments by +just voted for the other jira so hope it will get done soon +view file or the +Great analysis though! Ill try and get this fixed for the next release +This problem has reappeared +that definitely does seem to be a problem +Thanks Emmanuel for the quick fix. +already fixed at some point. +Id like to get confirmation from Martin before applying these +I was actually experimenting with this in the container +Im sure youve all been busy what with being promoted to a top level projectReece + committed to trunk. +Submitting +That all predates my working here so I dont know what support or expectations are in place regarding its continued use or thats a question for Rodney Max Prakash or BenBTW its Ganymede not Ganymed +Pinaki is a better candidate to determine whether the current behavior is correct +See the according issue +Gregory living this as it is +For instant releases Im just running make C compilercpp before make dist and it seems to work +Its just two groovy classes and a test case that extends. + changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Patch committed as r +Thanks Suresh for the review +When logging in there is a checkbox titled remember my login on this computer +I created a small document explaining the rationale for my choice +Preliminary untested patch showing the migration path +Presumably fixed in +This was an easy fix so I fixed it without the patchNext time you submit a patch file could you please click on the Grant license to ASF for inclusion in ASF work button pleaseKeep up the great work thanks +Rossen Im okay with using if I must +Ok done in r +Also please bear in mind that this debate is not about which option you would most like to see its about clarifying what the spec meant originally as it wasnt clear as this confusion arose +The module could even be an aggregate of the actual bootstrap module plus the branding bits with one boot class to set up the branding environment +An exception thrown compatibility bugBest regards Andrew Patch committed in revision +The PM version of contains modifications to clientjavasolrjsrcorgapachesolrclientsolrjutil related to date handling +Nice work David. +Plan to commit this in a day or two if no objections +There as a few places in Oozie that it assumes Unix filesystem changing the only is not enough +Added test case to verify +The pull requests have been executed. +Do you need a specific individual? If so you can put me down +bq +The showcase will work box with no modification required +In that case I will briefly review the code tonight and commit if theres no substantial flaws +Hi ColmThanks for your replyChanceCan this issue be resolvedColm +Ill do it as soon I close the current project stage at work +Revision +It needs to be validated so the user can be told that they typed something in wrong as opposed to a missing value +Manuel Ortiz The strange interaction is as follows As Rule is not fired everything is OK +If it gets pulled by core dependencies then great Ill defer this issue +This patch removes excess replicates from the machine with least absolute space first as suggested by BryanWould be helpful to have this no matter what we decide when it comes to redistributing blocks to nodes with more free space +need review. +Closing. +I think the times add up to about the same p lucasxbinxc d timing c commandlineonly examplesConstructsArray examplesxlibharness p g c Ilucasx Ilucasxpgascommonworkinclude Ilucasxgen Itiming ansi pedantic Wall Wextra g timing o timing p g c Ilucasx Ilucasxpgascommonworkinclude Ilucasxgen Itiming ansi pedantic Wall Wextra g timingharness o timingharness p g timing timingharness Llucasxpgascommonworklib Llucasx lxlib lxrt lupcrtssockets ldl lm lpthread g o asksMaybe someone could see how much time is saved if precompiled headers are used Thats what commandlineonly does right +Having that class in the jar allows users of both and and above to use without having to worry about a dependency that really is not available anywhere for downloadWhat environment are you in that makes this error happen? Is it possible for you to remove this class from the jar +OR Is this failing only on MAC? I checked on linux only. +We break backwards compatibility so this is not an issue at all +Fixes to properly select the to patch up projections for when inserting types +Great feature! Looking forward to see it +updateseq will be incremented for a number of APIs that make changes while not updating documents +forgot to copy the imagesrev +This ones trivial and doesnt require further review imo +It looks like an issue of mismatch jar files + +Yes we already have the display constraint +Thanks for looking into this Alan +Without the fix the tests pass on the RI but fail on HyWith the fix they pass on both class libraries +here i have tried to use various button icons to illustrate theminitially it shows then when clicked on the st time the icon changes to then for the nd time it changes to and i also used mouseOver to illustrate what will happen next +Created to discuss the default behavior +Merged into master. +Ive assigned the linked task to Scott to look at +Sending srcshareorgapachecommonsvalidatorTransmitting file dataCommitted revision +Not bothering with branch +Committed at revision in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Cool +Yeah now you mention it filters are ornery +Heres my first go at this patch +I am desperately trying to reduce the footprint of the public and the set of configuration options supported out the box because it is enormously difficult to maintain such a significant code base with so few people working on the project in their fee time +Incorporates Evans suggestion to mirror naming instead of arbitrary names +Closing all resolved tickets from or older +I had a quick look at the patch and I think it reduces the complexity of the code so it gets a from me +PetrCan you try to reproduce this using beta? Thanks +I recommend convincing the committers to this issue OR you should create a new issue +missed that one +That makes sense +Which I guess it shouldnt be right +Yes the defaults should be the same for all protocols because even if pop is meant to download messages it makes sense to tell the users of camel that does not delete your mail on the server and only reads unread mail by default +IP failover was discussed and we decided that it is an option that will be added +Ill watch for the nightly build to put the new guide on under wwwderbydocsdevptBR then Ill add it to the web site +sorry for that + +The source code is freely available but the ASF license is intentionally not granted since this code is not intended for inclusion in any ASF product. +Deferred to defer of featurestaskswishes from to +If I remember right I discussed this issue with Alex and he noted me that there are some problems with Mac with this my environment it always worked for it works for and for +Id tend to agree that controllers would be more convenient for implementing resource endpoints +Patch applied no comments. +bq +Gray please feel free to contact me if you disagree so we can discuss this in more +Verified that this occurs under Windows XP JSE Ant ivy beta Also debugged to see what was going on +thoughts +this week likely a tagged version in the already open pull request +Just be careful when two consecutive analysis are run with a different remote Sonar server +This patch has been committed to the trunk branchGeorge +It was probably something that we were doing wrong +I should be able to submit a patch for these in a few days +I hope he will soon +Is anybody else seeing this +I dont see any in the test +I believe it is ready to be used and that improvements can be made in separate tickets after this initial issue is committed +in case anyone is interested here is the crude ksh script I used to do builds on a per change basis +I will take a look at it today +to merge +unless servlets can implement more than one interfaceNeat feature though. +Thanks again for all the help +Looks right to me +I remember having to remove all the kerning elements in thegenerated font metrics fileIm closing this bug as Im pretty confident that this will solve your problembut please reopen if this doesnt help +What about this one +oh duh +It seems as though I forgot to request a translation of the description for CXF when Brian and I requested translations a while back +So the user is simply going to have to cast if they want the more specific type +So whats the story for this going forward? Will it be official build for and trunk for now +Because we have no time to work on older branches as well +Needssome more thinking +Looks good to me thanks +I tested it now with RC and it can just load resources from an http server out of the boxTo close this issue we need to update the documentation. +bq +r +enable tests in remove disabled method call in potentially add to the suiteIll keep this issue open until the above items have been taken care of +Checked configuration of some repos improved to render link into fisheye +I have seen this and its precisely why I had to remove all modules except the higest parent in each tree and then clear the flag +No +solved by by +What about just removing the first two test address pairs? It should work in all platforms although it will always cause MINA to bind to eth for a very short moment +Are you sure you want to add a Lucene dependency to Tika? Seems like maybe this would be better served as a Lucene contrib +Wouldnt that be on the AST level? In that case we could maybe use that directly +Updated version of is a simple test class that can be used to exercise the CCC command. +I was unable to reproduce this using the latest EAP build from Hudson running on IE SP on Windows XP SP +Ive added the test to srcit was the output for the plugin which depends on should it be +Just thought it was easier to type was completed by Tom White in. +Committed as svn resolving +for I created to improve mergepolicy configuration +Please reopen when you can show me howto generate the SEI that binds the headers to the resolved issues for already released versions +Please feel free to open another ticket if you end up being able to provide more details. +I went to add some in a JAR to the classpath but the User library wasnt there so I had to first create it by hand +Added back in the two property chunks that were missing and checked it back in + added a log message added a new test added a conf section in the forrest doc +png file I uploaded earlier had a transparent background which JIRA has rendered black obscuring some of the design +I also created to move the functions to contrib +All too glad to help +Hi ShameeraTheres been quite a bit of discussion about the details of the project on this JIRA issue +bq +You can obtain it with maven X or maven error message +This issue shows up in some of the Apache Commons components +Now the user can download andor install a runtime even if the requirements are satisified +I will submit a new patch using feature version range +Its not hard to imagine a FILERETRY happening here and all of them require a session. +If other commands are still paths do not rely upon that behavior +Do you have Windows setup to hide common file extentions? Open up a command line tool and go to USERHOME +And if its false the rest of the result set is the current values for the columns with a condition +Can you run the target and confirm thisThanksBjorn +seems to be to click on another view then click back in servers view +Those would have to be modified a bit to look up the active node instead +A different port may be a good way to implement this +Please commit the patch +Patch has been applied on trunk. +Also as this and a couple other classes are expected to throw various nasty exceptions I had them all ignored in the base class but yonik unignored at some point when he was debugging +Oh Ive mistyped the example here is the correct one OUSampleOUFoo becomes ouSampleouFoo +Yes a duplicate of +For instance if you have properties in a file and used the dialog to edit only one of them other will remain in the source unmodified +I am going to open a new JIRA that requests updating the distcp guide +To support the script needs to be converted from using the old to using the new suite +Hi AhmedThe version I had locally which I patched wasHope that helpsBTW +But shouldnt we have some uniformity +Would be great if we can ascertain how the performance of the new algo compares to the earlier one +This patch added these files + +Oh nevermind +Bug has been marked as a duplicate of this bug +Thanks Shai +Will do this in a bulk change via a tool soon +Not a blocker pushing to anybody know how we could mock the DNS lookup done by java at test time? +Closing. +In SVN rev I changed isPhysical for SERVICE from Y to N +There have been some mods the the JCA connector to defer connection close till a transaction completes allowing a connection to be pooled when a transaction is still in need of completion +Do you want to submit a patch changing the logging level from DEBUG to WARN Al +I havent been able to reproduce this error even on the same hardware. +Instead of expected status codeSCUNAUTHORIZED the user gets a misleading +We should not be locking on the app attempt id +Not sure if the unit test is quite correct but it throws the exception as described in the bug report. +This seemed like a very clear win but Im actually seeing a drop in performance +The implication is that the metaclass of the is not doing the right thing +patch applied in r and r thanks for the fix just the vtest for this at revision +Hi ClausThe original issue I raised was that we couldnt address more than GB of memory when using the Linux wrapper so at the very least it would be worth putting the same note about the memory limit under the Linux wrapper section of the page you updatedHowever I guess Im more interested in whether this will ever be fixed or whether we will have to look at other queues in order to be able to address more than GB memoryThanks need to buy a license for the wrapper if you want to use it with Windows +The testcase should now work fine +At least to play a little more with ivy and then perhaps I could then try to provide something for +I like this +Ok now I see it +bump? +No need for this in Roller +Is there more work planned on this issue or could it be resolved +remove the xfire cxf from the top classpath to reduce dependency bloat +Therefore I personally would very much prefer this code to be entirely decoupled from the standard API even at the expense of some code duplication +adding the example for the extension of custom mapper can already be set by specifying marshaller properties from Spring +But yes you are right using productionapplicationsproductsrcorgofbizproductprice would address all potential issues +Tests pass fine when I apply the patch +trivial patch for trunk +Do you mind submitting this fix in patch format with tests to verify it to ensure that it is not lost in future? Youll find examples of other validations rules in the projects mentioned and in r. +Alljava processes behave this way any jar files used by thejava program stay open during the life of the process + to trunk and patch for committed there as well. +if the datanode dies in the middle of a rename we can have a partial file in the destination as well as the complete file in the source isnt itCurrent code handles this +Trying against hudson +Patch with license Shane +bq +Create Console configuration +Chad also requested some changes to my diff for the JSON protocol +This JIRA is not a problem any more. +Its potentially slow to write all the files on larger clusters and shouldnt be needed in most clouds +This patch make it work for me +I actually tried that and although it helps it didnt seem to remove the issue altogether +Using implementation classes instead of interfaces is allowed by the SCA specThis is actually a very important scenario to support in SCA where you start with existing Java classes which did not implement an interface and want to use them as SCA component implementations without having to change them or create another wrapper class around themThis is now working in Tuscany thanks to the contribution of a new based proxy factory +IMO the last patch is good to go +Change target version to and submit patch since the feature branch has been merged into trunk now. +Force inline of pTLSvmthread and tmnsuspentenabledisable on Windows +bulk defer of issues to +Any chance you could try a snapshot otherwise a better example would be good because the one I just created according to above appears to work in trunk +rid of the invocation of +Uploading a patch for branch +Ill try it +Yeah that works fine now +also if css class order is not important I committed the fix since a stable order could be useful for unit tests +Be careful not to apply it +Heres the patch +The issue is reliably reproduceable but due to the sensitive data involved I cannot share the test case +This is the project directory structure as the example creates it in your workspace I just went to my workspace and zipped up the top level directory called simple +and puts back some of the more convoluted null checksthis one should be more palatable and should be included. +Will I commit it as a start in on this issue as a part ? Im working on the other part over in master rewrite where becomes instead etc +potentially complex and invasive +The use case is smt like this Client picks up events from a reliable event queue processes it and increments the counter +the clipped part where the bitmask is +jacopoas the submitter of the patch im not sure i qualify but we have tested itand are using it in our live systemsthanks luke +But they override your environment iirc meaning you always use boostdatastructure which isnt bad in most cases +I am able to reproduce the issue +readonly If true +Or we should change the test so that it doesnt need in which case it should run on Java ME too and the change that I just committed should be reverted +Thanks Ashish! +Hi AaronI analyzed the root cause for this failureI think as part of some release audit comment fixes Apache header has been added in fileBut this testcase will generate one edits file +Verified +So woo +Please verify in CR test cycle and problem Laura +One liner change to expose via NT Thanks +Committed to trunk +Now that this is in the IDL documentation on the wiki does someone want to sign off on it +I havent been able to reproduce this at all with r +I think that when is committed this patch can also be committedAlsoincluded a small fix to the +Activation has nothing to do with transactions! Transaction code in the is unused and superfluous. +We should reduce the scope of this JIRA to just check the seam dependencies alignment with EAP +You want to make a patch ak +as in the attached happens because title key is used +What you see in your tracing output is that is paging through the i partition fetching rows from the base table seeing stale entries and moving on to the next index cell until it finds enough base rows to satisfy the limit or exhausts all the cells in the i partition +I have to manually move the testprojects directory out of the plugin before packaging it +Does that sound reasonable +The schema was modified to include support for as a value for the value attribute +As discussed in meeting +Jason said If it looks ok I will implement Externalizable in other classes +Thought I should mention the Dfilter option for archetypegenerate in case youre like me and have lots of archetypes to choose from +Thank you for your answer Chris +Pretty sure it wasnt due to confusionI suspect the issue was an earlier on and this was just the symptom of where the bad objects were +We will remove this system property between Beta and GA +Adding AJAX functionality via AJ components The project will be enhanced with AJAX capabilities +This build failure can be fixed by using the attached patch in +SCM dateThu Nov EST Continuum dateThu Nov EST So its a minutes difference between them +Try with the latest release and report again if there is still some problem. +This is the only place that can print the stack trace of the error in namenode +Hello WernerI add the attachment with for this case +I think this will still occur on later versions +Update to apply cleanly with trunk +I think is a good enough workaround +BrianFor now When you modify the source codes in Smooks editor the all graphical model will be removed and created do you mean that you dont lik the editor all models just change the models with the source change +We shouldnt close this till its merged into there a pull request for this on or does it require M +I added to explain the algorithm of H +Per my previous comment this issue was submitted as a question not a defect report +Sure this removed a few import xxx lines from source files to prevent Eclipse warnings and does not need targetted unit tests +I was just pointing out the future pitfalls of assuming too much implementation knowledge +Attaching stack is not a VFS bug +Thats simply not scalable +Fixed in rThanks for the patch +The patch is in too +If Ron can provide a release Ill switch to that +I agree that adding options and features is not the best way to make things simple but this kind of feature is something that can help people to start using Ivy without having to setup their own repository +should I commit + +I have a working Hive Whirr integration +I am adding a patch that upgrades the processing to support the following The engine now properly marshalls and demarshals the detail roleactor subCodes etc +The disk buffer also works this way it may write log to disk even when force has not been calledThere is however a possibility that the following sequence happens during commit processing on the master +The error is Cannot open installer log file at DDocuments and SettingsappedsApplication Datadata I hit OK everything uninstalls and there are no log files to review +Jorn found another memory leak in the UIMA AS service +Good find +Similarly there will be some changes needed for reporting and alerting adminstrators +This is the same stacktrace you posted in +Logged In YES useridI have an RH system with JDK that I have tried thisrecepie on a few times and dont see the problem +I think I addressed this +I am afraid thismight vary from RDBMS to RDBMS. +Patch fixed test failures +What exactly does OWNER INTERNAL TEMPORARY TCPSESSION REMOTEADDR means? ADK OWNER is queue owner INTERNAL is a property referenced when creating exchanges TEMPORARY is a synonym for AUTODELETE +I just committed this +If you pull in the latest sources you would not see the race condition between channel start and create transaction +These entities are actually not entity beans and they should probably be shown in an openjpa specific portlet +The commit will fix your issue and will be in for the prompt reply and for the fix! +Rafi did this change resolve the issue? If it is ready to be reviewed then Im happy to do that if you assign it over to me +Im working on that today. +By default I believe configures the Unix domain socket for HDFS short circuit reads under varrun +Fixing for javac warnings related to using deprecated API from about this more +I am creating some test case to reproduce the issue +I did an ubuntu install and flume downloaded the following unwanted packages +That web page consists almost entirely of links to items that are not currently included with the source distribution viz +Marking as WONTFIX for the reasons above please feel free to reopen if you have further thoughts. +Fix with a test +Pull request sent +Also does add some level of support for inner classes which might be related to problem +Pushed to trunk branches +Assign access to the mapbuildersandbox directory to a greater group of Shorter usernamecamerons +Thanks Rishi fixed in rThe problem was that I hadnt switched back to widgets for the payments menu so it was still using the freemarker template which contained the bug that you reported +documentation in bulk instead of page by page +Thanks Patrick +Preliminary property set for that interacts with JMX and the deployment service has been committed to CVS +And do you have a Sonar Quality Profile explicitly associated to this project or do you use the java default quality profile +The test orgapacheharmonyauthtestsmodule is returned to with relevant comment +Thanks guys for identifying and addressing it +personally Id probably prefer to see all exceptions the current situation was setup out of desperation because GWC listeneres were not fixed after a week of reporting failures to save due to them +The heartbeatInterval is in milliseconds +Aidan can you confirm that this is still the case +resloved in trunk and branch +but it goes outside of the and operates directly on File +Hm +It essentially means that there is no data inside an approximation window at all so why should we make the data appear to be zero? How about a variant express the approximation window in terms of the number of points with weights +new patch attached solved the bug where ssh private keys were not loaded on a fresh workspace +Do you mean the or? or both? is fine to change to provided because it is only necessary to generate the logging implementations when compilingbuilding the projectBut for is not better to leave it with the compile scope? Given that it is a required runtime dependency? In theory users using should be able to use the provided picketlink modules so no problem in this case +If there is a need for automation there would be requirement that the dump should be in standard format +fix think we dont need fix this issue nowPath always set file in front of uri so this issue wont happen +Create new ones when you want features backported and link to is a dead branch nothing to do therec is an EAP only thing +Granted I expect in Java it is not done but I dont think this function to be a performance bottleneck +The statements above are mixing package management with configuration management +The work got done in +the whole reason for that patch is that you dont like the default javah path which most likely uses the one JAVAHOMEbin +Unless I hear objections I will just commit this. +Limitation on A +At some point A receives a message X and during processing of that message responds with message Y +Lets do this one step at a time and open another JIRA for reran benchmarks with K +Patch Christian +severely undermines coprocessors as an extension mechanism and Ill be on changes in that direction +Thanks KristianCommitted revision . +Created for the missing packages have been released +Difficult to have this when building from sources its based on the svn metadata and sources tarball doesnt contains svn a look at the profile in module +I can also enter a JIRA for it appears that using the breaks the jUDDI console +Another questionWhat about? It is the same issue as for +updated patch against for this issue. +On the other hand the symmetry isnt so bad and I dont think that anyone will notice if we do the prepend +Are you able to provide a similar patch for trunkThank you for the contribution and good catch +is this a bug in +this looks good to me. +pushing off remaining issues in order to release version in JIRA and create release notes +Sure sounds good if people dont mind changing to something else I certainly dont mind +Something along the lines of the following would work well as the command line interfacemvn grouppluginidversion there an example of how to hack it using profiles somewhere +I see what the is used for now it sets the step level +See discussion for +I am behind corporate firewalls that cannot access remote repositories such as Maven Central or the EBR +hijun +This taskUrl construction should be based on MR cluster +AllI updated and as appropriate so this should now just work +Thanks Alejandro +It depends on what the ejb profile is supposed to represent +Committed +unscheduled issues to the next release +Thank you both +But I see the bug report states that it fails on all platforms but I know this to not be the case +Closing issues +Im on vacation next week but plan to wrap this up the week after +I would very much appreciate any hint as to why the parentheses I was seeing turned up ever +I am running tests now +version works OK +arun any reason not to close this? I dont think I can repro it anymore. +ready for t we stop putting forth effort on and focus on getting to fly right? It is a standard spec and does everything thatthe can do now that the cocoon protocol is supported +Deneche I think youre welcome to reopen with a patch. +I think we could implement something in itself so that it can eventually be used for making optimization decisions for compactions and caching +Even relative links like +committed in rev thanks! +Do you know what software generated the file? Is it openable by both and Open Office? If you do save as in powerpoint on the file does the new version still exhibit the problem +Thanks Shelly Ive changed the html for the jbossws console removing the reference to EE etc +Daniel Morris opened this issue and supplied the patch which was reviewed by Christine +Moving the build to M wont break Ant projects that use the artifacts +For recurring snapshot we do not need to take snapshots at the same time +will run svn add first next time +Just curious Does inputing chinesejapanese signs work in the source editor +This patch applies to and trunks + +Lets just go with Maven +I committed the patch +Hi just noticed that the ZIP file you attached did not include the XML schema included from your main XML schema +When requesting console however this is the responsibility of the browser implementation and beyond our reach +Since test case passes Im closing this issue +I think the new warning is the test case calling the deprecated createWriter API +Logging needs to be added +Hey Travis I tried to apply the patch to the trunk and it did not apply cleanly +Both running on +It specifies how andwhen whitespace is to be gobbled in a consistent manner +Container will be given up at the lease expiration or when the app releases it on its own +So this is not ready for inclusion. +The issue was on Windows bitAnis what Windows version did you test it on +Thanks again +Erics sample project this time in regular without any Eclipse files +Im going to work on this some more today and report back +code changed checked into HEAD +I cant think of any reason to keep tombstones around on a purely local table +Thanks Doug! And thanks to everyone who helped me out on this oneI think it would be compatible with although I havent tested that or checked if the patch will apply cleanly to the branch +Im bumping priority to major until I have a testsuite which triggers it more regularly +is a test file based on the log files published by Sengor +Just to clarify ourmy customers are not the ones that hold the indexes they are the ones that develop the products and will need to face the real telling him that he needs to upgrade his potentially indexSo what I was thinking is if the abstraction CLI layer could be used to fetch the ordinals from either payload or DV based on the segments version? And also migrate segments gradually +Toshiya yes that was done by design +You can test this by opening it via a file url +I will prepare the backport patch and also one of the suggestions above +If there is no objection I will commit this patch soon +Just provide a link to where the mailing lists areCheers outstanding items mentioned in comments look to have been taken care of over time. +But its not related to this issue at all + I used maven by working with version there is no more exception with eclipse plugin requires at least maven +what for +Anyway you are rightI changed the code to be compatible with using incRef +Hi Grantah ok I see +Patches from Phani and Vamsi that need to be applied toI verified the patches work with and the latest Geronimo branches code form late Feb +I had looked at replicating the fix for merging state but it seems that the semantics for region transitions are different for online merge +This solves the problem of deploying the POM as the POM is the one defining the distribution looks like a good fix to me +Looking good +Sure +To create the Test Works Data to CONTASREMESSA havent had time to check out Bryans suggestions yet but I am planning to dig more into to it so Im assigning the issue to me +a security condition +Attached patch new JAR that has this fix +If it is again the missing CONTAINERKILLEDONREQUEST will be emitted here +I dont want us to try and do any on the native side since that can be prone to bugsThoughtsThis looks fine for WP the only button that can be overridden by an app is the backbuttonAre the android different than the standard js key codes +This patch is a little large but there really is almost no new code +Not quite as easy as the proposed change but the requested behavior is available +Please see the patch attached +I added a bunch of comments referring to in the commit file for so that we can remove the short term codeconfig added in am able to compile locally +Thank you very much for your comments +Never mind appears to be an issue with my zip program on OS X +But i think this concept doesnt even make sense for what we are doing hereI agree +Unfortunately I am no longer sure it actually works +to trunk +I think the next code stage will be to move the localization property data to its correct places +The problem actually appeared for RAR WEB and EJB modules with ve fixed a little differently by introducing a late than never +I think that the default behavior for continuous queries should be to cache everything that is not from a reusable execution +If there are any tweaks needed in future lets address them then +Patch against +The store catches most duplicates but with concurrent dispatch is is possible that the cursors cause a dispatch to be dropped leaving a pending duplicate that the store missed because it had already acked the solution is to have the keep a message audit and only allow one copy of a message to be sent the the +among the Nov +Hi We have fixed the icons lost bugs +You will thus avoid these kind of problems +Thanks Zheng! +Addressed all the the compile patch looks good to me +Applied the patch +Im going to commit Majas change with your suggestion. +Have you tested there? Your patch goes a little too far IMHO adding configuration arguments to the sequential algorithms which do not understand D arguments +So is it confirmed that this is no JVM bug and your patch is really a bugfix and not a workaround +can you try with a axis SNAPSHOT nightly build +Patch that removes some of tasks from drlvm build +Since the JSR isntrcution is not to be used anymore you inline that block of code at multiple places causing the creation of multiple disconnected ares in the source code with the same source line +Hi Claudio Avery and ArmandoSorry for the late reply +I have a working version of and Ive posted a copy of it to the microcontainer user forum for a second opinion +Im not sure that we need to fix this for the release +For if that is not ready we can manually publish it +Thanks Andrew for fixing this up +An admin has to edit the Contributor list but in the meantime commenting here as you did is sufficient +This creates a separate tool out of the logic in the pamphlet generator which looks for spefici information in the release note +I imagine since the files are being picked up by a brand new VFS there is a brand new counter being used +For the case where one was not provided we create a daemon executor so that we dont need to explicitly clear it +The changes only affect the trunk version +Many things have changed for example software requirements were created in the project +In testing +Reopen it if it needs to be fixed. +SteveStepping back a bit what Id like to achieve is to access S from an Amazon Elastic Map Reduce cluster client without having to add S credentials to another configuration file +Update patch accordingly +Im fine with just having the long name +Cool thanks for explaining +is now referenced by apacheds modules +Reviewed the code change +The check should be done before it +We should instead be trying to write exemplary code +Please let me know if something like this is acceptable then Ill clean it up and add a few more details +But I couldntI did yes and gave file to be symlinked as hdfslocalhostuseramarsrifile +There are policy intents and policysets defined for configuring Axis config params in the +This bug only happens when we have environments that share the same S secondary storage +That applies for all of yarnmapreduce daemons +OK let me know if your development needs it to go to a different line and Ill try to merge it in. +The code has been tested and submitted +IMO thats not an option +Can I commit this code? I have no developer access to this work with Mojarra? We need to know first if the reference implementation has this behavior before commit it inside myfaces code +Thanks Shawn +run same with ant run build with to confirm things are working that way still as ivy fixes for slfj dependency for a few contribs +Implemented please +Once I changed it because this stream data source has the problem that the inputstream can only be gathered once and every other try will end up in anI remember the test case was a file endpoint polling for binary files and sending them to the mail sender endpoint +I have improved how this is happening what will avoid issues with threads and concurrency +Patch looks good +Hi Im actually thinking of reintroducing it back in scope of the client api work such that a client code can experiment with various providers at runtime +tests are ok with this new HDFS version HDFS tests are in progress locally but seems to work ok as well +Im fine with closing the ticket now and creating a new one for all the other cases +. +However I dont think its a bad idea to implement so Ill transform this into a feature request + +It is currently broken and undocumented +This was fixed in and if you use the version of the Maven plugin for Grails all should be good +If set to filter remove outlinks of pages which page lang equals entries +can t be since it is not a valid method name in see but do you agree on that expression is badit is just that we have a Pointcut and when we reference it we do it in an expression better to stick to one namemy cents +I think an optional flag on the would be easy to implement +The upgrade is going to be done through and dependency mgmt to override jgroups version from to + +These two test failures look like and I think unrelated +should leave the JVM in the state that it was before bus initialisation pending gc +Some may want xstream jaxb others json google protobuf just serialized objects etc +This is my first patch to Giraph so I will most likely need a little hand holding through the process the first timeThis patch includes the following java files for and +File Attached is the background gif for the legends list +conf baked in as a Roman +Created the Seam Portlet wizard +When two testnodes install a product the first install completes successfully and the second fails unable to delete directories installed in the first install +Vlad mind if I pass this on to you? We shouldnt need flush anymore but we need to test and make sure it doesnt break anything +Ive just committed this +Looks just committed this +Should I update the apache wiki or will you +It creates a server config using the contents of srcresources that dir can also target basically replaces with one that uses for everything and then some EJB tests +Note that the check is performed after autodetectionIm reopening this with a different problem related to have a mgmt interface +Hi Prashanth the patch didnt apply cleanly for me +fixed the issue both in trank and you! +Verified with JBT +more or less updated docs sources will do site upload later this week or start of next week +bq +Seems to work fine as far as I can see at the moment +Steven +Please check the project out and use the included formatting preferences to format the project +How can environment variables be the right way to push installation information around +The jcr installer has been configured incorrectly changed configs are written back to appssystemconfig while the installer only reads from directories named install +This patch move the configuration of the from the to the in shindig javaIt also enable to have different rewriting for each containerCan anyone review it +Alan theres a really easy way to do this since you already have a custom checkbox field for Geronimo +Whats left in the mapreduce API for this +What about that doesnt work +Two screens and info about facets are added in the chapter +I think of using oauth need a separate context path to handle oauth response that would be also a preparation to support other oauth profiles in the future +Verified in JBT Linux . +I do so because Im not entirely sure that it is a valid use case +CQL can model the same data but the structure of the table needs to be different +Howdyheres the JBUGNetherlands logoIll process it out into different file types and sizes +Building VM on nyx +We have two reports currently one that we dont escape low characters and should do and one that we do escape high characters and shouldnt doIm marking this as WONTFIX based on Phils thoughts on the mailing list +This adds the jdbc suite to derbyall +So should not really be exposed +Lets not make a change just for the sake of it but until we have a real use case to be supported. +Oops guess wiki annotations arent enabled +I opened for the further enhancement of configuring this through the Web if the issue key changes for every release you can still configure it in the pom prior to executing the release process +How is this progressing? Done? Shall I review +I doubt that anyone is using them they dont workFor cleanup I think they could be deleted but the original authors should decide +changing the fix version to +patch is no demand for this functionality for almost years from other usersClosing as Wont fix +Finally I am not sure about the static javadoc for the Map and Union types so any review would be much appreciatedBest regards you very much for your feedback LewisAttached is a corrected patch that solves the minor issue identified by Lewis +Attaching instrumented jar +Unresolved as of approach is to force the to check the bounds of the Image before returning it to anyone +robert +Please commit to both and +For example see +Every time it pops back up with a valid header +Tried this again with the updates from the weekend and I am still getting thegcc internal failure +Add armvs in addition to the armv fixed for me on ios beta xcode mountain lion thanks Shazron! +So lets wait a few days and see how this works out +Ive uploaded the updated patch on the current revision + The source code for a simple echoServer to receive subscription notifications +I will take a look at it +Adds two more options to slave configuration in http httpIf both are present then basic auth support is turned on in and their value is setMatthew can you please try out this patch and let me know if this solves your problem? I dont have a setup with basic auth +Please review +. +So Svante pls review my patchyou can modify it if you dont like it I will have vacation tommorrow +This is for manual reindexing of artifacts +In it Ive implemented watch functionality based on a new app setting called as described in the issue itself +Tests passed cleanly for me on +Btw info is the correct log level +Hi AtulNo worries it was an easy fix +And its an internal class so its not risky +The patch looks good to me the only small comment I have is that the severity level should be perhaps warn +Fixed in revision +I think I just used HSQL instead of the criteria API +Of course you may have at the shared lib level and at the level the new jar file split facilitates +Had to revert from trunk since it caused failure +So is there anything we can do about this from Springs side? Wondering about the status of this as a Spring issue +Thanks for the code review Knut +I need to work out how to add a test for this not easily done unfortunately +Change tool tip format to YearMonthDay HourMinuteSecond for Time Series Graph +Patch looks good technically +Thanks for the patch Ming. +Gunther please run unit just committed this +I wonder if it is possible to recover those missing files respectively +forgot to mention that I was usingHTTP protocol +Patch for to convert it to a spring junit integration test +patch ignore my last comment +oops I must have not built after my last change +Fixed in svn revision has been checked in and seems working well +I believe that is because its a function that isnt passing any params +I checked the maven docs you are using it incorrectly is no maven plugin! It is a simple java library to provide search to your application but not to maven itsself +Altered so that the default Dojo blues are now greyed +Is it possible to port this toWe need this feature +Merged to the branch in revision +Sounds like a translation wrapper around Aethers Artifact +Those urls That content +However the current implementation doesnt support that +I get less messages and i took down the ROOT regionserver and we recovered ok +Konstantin you might want to merge this to please take a look +This code applies to Spring as well as the upcoming and releases +Please verify the fix in the latest. +I am trying to test WSDL file from deployed web service so in my case when I am in process of obtaining wsdl file it ends with error I posted earlier +If one or both of the regions being merged is large it could take a meaningful amount of time to complete the merge +Hi Anand Id like to see some of the other commenters weigh in on that one +Attach a patch to implement this main changes +I actually agree with this with that but changing that could cause users to who depend on the checking ordering to have silent test failures +Here are the counters from the two run +We will use the time to fix more bugs and get some review done first +Hi SascchaYes why not +Implemented in SVN trunk +Go for it Jimmy +fixed in +Uploading patch with the above comments implemented +The fix which is required in two places is to ensure that H file are constructed without percent encoding as this is the form expected by H +List of use cases for LocalRemote containers and LocalRuntime interface +Modified the class to allow the choice element inside a sequence that we want to unwrap. +Hi Kris were not going to be able to get this into as we have support customers waiting on that release and this is too big a change at this stage +To prevent requests from being dispatched to web apps that have not been started +Updated patch addresses your comments except the test continues to use assertFalse instead of assert as that is only for Objects +Instead we should attach information to Request to hang triggers to a list of +This is the central class of the publish any object as service machinery +So whats the workaround +Re +Or in my case I wanted to use hsqlb and reuse the results but needed a common location to point them all to. +I would suggest to read the web site check the mail archive ask the mail archive before open JIRA +No objections + important patch for just committed this +The findbugs warnings are not related to the patchAll unit test passed except for the ones mentioned in +After looking into this it appears the problem is the way the equality is defined for +testcase core functionality in need to update functionality will be used in tickets like as well as tickets and via the testcase +That only makes sense if you are working in a direction that the community sees as aligned with the larger goals of the projectGoing full circle Yarns open have as a goal allowing people to try a lot more things much less expensively +thanks again Tim +while it may be true we need more test archives it doesnt seem to help to have this meta bug wittnessed by the fact it has been left untouched for three years now +Lets use real UTF +Instead we could simply document this all clearly so that users are armed with the information necessary to configure their builds to exclude transitive dependencies they dont useThat might be a useful strategy make more dependencies optional and document which features require what dependencies +Updated the hosted version to a member reset for the table and added a note so users know. +from a seem to be workingi will upload patch that adds the new testcases as suggested +Also it would be good to know if the error still appears for you if you use by exactly an hour +running +Im flagging this one as to be fixed for the next alpha + +This is happening in the method I have also tried it in a Test method Im pretty confident it is a test since the rest of my tests are behaving as a workaround I needed to delegate the Answer creation to Java helper class when I obtain the answer that way it behaves as more note this is from the commandline not from an detection logic looks for annotations on the first class in the script +I think it makes more sense to have no template so by default the editor is blank Add three buttons or a drop down or something with items Default point Default line Default polygon which when selected populates the editor with a complete style one that will people think this is a good idea please create a separate issue and feel free to assign to it to me +In order to finalize the improvement for short Git id I send a patched file for the accuracy in the plugin documentation page +Also please give the screen name for reference to see exampleWhat I did isThe rule for select box multiplemultiple is already there in css for Blue Light theme and Flat Grey theme +bq +Min did you really see this bug in? I checked the startup code of is not using this class anymore +I did try other changes while trying to figure out why myproject was segfaulting catastrophically +Apparently there is some new API on type objects in the compiler front end that may be of use in that regard +I agree using the store is the best way +fixed by just need to update to a not too old +This bug is severely limiting especially since the online documentation indicates an AND operation +Could you put the changes to the stage please +I would like accessDone +I am detaching the in a Component methodWhen the method call completes an method higher up in the call stack will call the get as it is returning to process the next event in the broadcast events loopLet me know if you need more detailsRegards + +Uploading the code for resource adapter +Means that the tool is sounds like a description of an existing condition whereas the boolean value is really an instruction isnt it? So Im wondering if it should sayA value of true tells Derby to load the tool +Hey Simothanks for reviewing and for your comments +I need to rework this area of the regression tests are included +Resolved! +Jimmy Any chance of a page of design outlining what your patch does and how you think it all should run? Will help reviewing these big patches +Some small comments now attempts to create a in one of the test cases +Moving Bryan +Awesome thanks guys I didnt think it looked right but wasnt sure +jdkjrebinjava Xemopt Xmxm +Im especially interested in the exact node type information of the troublesome nodes +Is the intention to change the protobuf def to a signed int +I will soon file the bug for that but i need to understand first how exactly this plugin is build +Rewritten the test so that the scenario is tested sequentially no multiple browser windows at a time +So I think we need to catch an exception in the static initializer of both class loaders so we wont fail on older noI should add otherwise this is looking pretty good +Yes this should fix If you can fix the indentation that would be helpful thanks +if the name of the constant is the only one that make you not happyplease give me a hint +This might not be super obvious right now because I believe only the RAW category properly prefixes its log messages however I think this will be useful going forward especially if we add logging codes +The attached archive is a test case to reproduce the configuration of is included Ant script packages a session bean configured with a strict maximum size of first invocation of the ejbCreate method of the session bean throws a first time the client attempts to invoke the method on the remote interface an Exception should be returned to the client +Adapted so that NPE is prevented in case session is not available. +resolved thisis this issue resolved and in the Hibernate picked up by EAP builds? If so please resolve +Found the JIRASince group by is done in the reducer you could just use the trick that is used in distribute by x sort by y when we do MAP and REDUCE operators +It was fixed in a previous bug +yes +However I have opened for future improvement improve to accomodate comment in for Apache license header +Update existing bench example or Create new one is your choice +looks good to me too +HiIve attached a patch to resolve this issue +we cant throw in that case either +Maybe you should wait for a second opinion before closingthe issue +I just committed this +In addition there doesnt seem to be any hidden property to set on the picture in the mediapictures folder so we cant hide pictures that way +The larger issue is that having your security jar in libext will break ANY time there is a reload so if you change other jars it will be affectedThere is a tolerable work around for in just putting them in libext but I think is a bug for and Im thinking that security is something that should likely be locked down +Especially one can think of hosting frequently changed data in and access it in HivePig via +Committed to trunk revision branchx revision close for +Of course switching off the ssh server using the config will make the client not work anymore but I guess that is ok for people +Attaching +Therefore lowering the priority +Revision r the problem back. +Does this happens in too +Added a strcasestr and used that to solve the problem +See the doc resolved issues. +Verified working at rev +corrected component +First attempt at this +Marek please verify this issue I think it is out of date if it is please turned out to a be an issue with how Seam sets up its hot deployment mechanism +I think this works as designedIt may work the way that you intended it but it is really bad from a usability standpoint +It has been a day with no new comments +I checked in a fix this morning +verified on +You need to redo the patch +for the patch +! +This will take some time though +This should be done in both the admin guide and reference manual topics I thinkThe admin guide topic already says Note Imports are transactional +We are not going to add a default to EAP +how do we deal with the test subprojects built by ant scripts? Have these really moved to geronimoanyone know about mitests ? or do they move to geronimo since they require a whole server to run? +Move the patch to close with Incompatible Change flag. +The patch in this issue that was previously reviewed by Morris Kwan and applied by myself has caused regressions and so was pulled out of both the main trunk and the tagThis issue is now open once again. +However what credentials should I be using? I tried my Apache credentials and that didnt work +I just committed hb patch to append branchFor the incompatible problem as I commented on the hbase jira unfortunately changed the syntax and semantics of heartbeat packets +Hi ErwanThanks for the info +Section saysDeprecation refers to a class interface constructor method or field that is no longerrecommended and may cease to exist in a future version +It starts a spring transaction and executes a query on a non existing table +the current system really cranks through a lot of garbage generation +If you dont mind +Modified version of which contains finally blocks in the init and destroy methods +For future reference When changing the sources use mvn clean install Pjavacc in and commit the updated files +I cant reproduce the bug neither in nor in +Another layer of shell expansion doesnt help anyone here +Profile not mentioned in the child at all +Applied via Github fork queue +What is wrong with the patch that Daisy submitted? Could you give a few details +That inside and outside a closure are the same is part of the problem +I cant read the pom without line uploaded a new version with better formatting of the pom. +Im trying to create a plugin which will be the same as the fileTransfer Cordova plugin but changing only the store and writing method but Im a new bee on ios developmentThank you for your! Looking forward to it in +looks to and +BTW if you dont agree please feel free to reopen the bug and continuing discussion. +I did take a quick look at the in implementing this and saw that this handler lacks a lot of the function of the AXIOM one +So I still seem to see this bug in CR from two days ago +Patch to switch TCCL while creating +Otherwise change the Release Notes Docs Status to Not Required +Closing. +I think the short story is that that it should not have ever been used by the engine +Looks very cool +getclass Inline pragma support was refactored in revision +At least its ok with I suggest if someone familiar with JDBC specification can check this problemIf the JDBC specification has clarified that the JDBC driver should do the number round operation in round half up mode then its okay that relies on the JDBC specification to do the round operation +Verified by Han Hong Fang +Running the regression tests now +If you feel you can please feel free to take it backAssigning to myself +WonderfulVerified thanks a lot! +The next community release for AS looks like it will be not so hopefully gets released before Teiid scheduled date +I checked the other build beforeafter this build most with the same test failure +My problem is why in is null when the addon is created and installed correctly? Do you have any documentation for how to add a support for new language? I am trying to add french language is actually correct +Most application servers can be configured to place their classes before or after application classes +Hmmm is not in maven +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Also I have not done any performance testing with this patch +Sadly this seems to be a bug and there is very little I can do about it +It is a strange error +Hello NiklasAnd thank you for the prompt responseThe embedded FTP server I have implemented is only used temporarily during a short session within our application and because of our packaging we cannot create a keystore file that would have to be moved during installation and thus rendered uselessFor our usage I set the implicit Ssl to false so the client needs to explicitly request to use SslWhen creating the Ssl configuration there is no keystore already createdI went around the fact that loadStore is declared private by having to copy the create method in classIs there a particular reason why loadStore cannot be declared protectedThanks again for your awesome work NiklasCath +Its been over a year since the last update +I have made the first patchLike you mentioned the slot may be used in may places and we need to have a careful checkI will read the code more carefully +See for further details. +For the examples I hand the issue back to Tom +Its not so much a bug in Artifactory as the fact that I didnt initially configure it properly +Stack Matt LarsCan I assume that youre Okay with the formation in the latest patch +I just committed this +I just uploaded a new version to address the comments +Thanks Tiago +Passing to Fernando for incorporation of release into EAP CP +Weird I tried applying it to trunk and it worked for me +Attaching the release note +I think it is reasonable to update to the latest one as the SQL server jdbc driver supports both SQL server and +trunk +Any more comments to this? Change stuff? Were using it in production for two months now and are happy with the results. +worked on my computer. +This should be available thru cmd line and web ui +Attaching the patch +This is now driven by the so that the scheduler factory bean does not have to implement anymore +Yeah sorry for my mistakes I corrected themAnd I added the line number to theLet me know if theres something more I can do +Computing a crc per transaction gives more flexibility that allows to read an arbitrary transaction with seeking +This is what is being fixed nowbq +Hasan Please remebmer to set the component field or the JIRA will likely get lost +I dont believe you have voiced a valid technical concern +rewrite so that is shutdown properly between each test +Plushibernate is handling it in xml configuration +If it works as youd like pls close this issue so QE wont need to worry about it +Note that this patch uses checksum to compare if primary secondary have the same image or not +understanding is that the Accept field is used to specify the format acceptable for the response +Workaround provided +Please try this version as weare planning a release soon +No changes have been made but comments indicate there is no issue to speak of +Thats bug brought by +Also committed a fix for +Ideally you want to be able to only build locally the modules that youre making changes to and you want to pull the other modules from a Maven repository using standard Maven artifact references +Fixed under revision +Note that has been committed you know dont need to get the control nodes they show up as action nodes of special typeUploading a newer version of patchNewer patch with testsFinal patch now that the review had been all clear +Yes this is a regression +Updated patch to address some offline comments from Adrian +For me it produced a error with java I have to try +upload a new patch that fix the warnings suppressed the findbug warning that complains inside invoke method which is fine +Somebody also made the point that you cannot selectively override parts of the configuration +Perhaps I should have been more verbose +Thanks Adrianin trunk +The error is obvious and the fix equally soThe error clearly isnt all that obvious given that this bug existed for quite a while without being noticedfixedI think we should have a test mostly to ensure that this issue doesnt regress +batch transition to closed remaining resolved bugs +currently requires defining a rack even if one does not intend to use rack awareness +the patch looks good to me +refactor test code to remove duplicate code +Yes all the async should share the same config including the getcancelrenew token +it would not be reported by examine and therefore it would probably not be processed by Axis +add jaxb impl to can fix it should we do it +Output to console from running mvngenerate task +. +Changes to properly check for bounds + +We need for this patch to work +I think this also means we may need to change the conversion rules when comparing dates and strings I believe it currently converts both to string +This is the I use for AS +I will provide a patch to fix this issue +The connection should not be dropped but the request should fail with error messages +I appreciate the +Can you attach server config tooAlso +So I still think it would be good to weed out this sort of log statements. +Just and without hte OrCant think of an also think about and +Marking as closed. +It was built for +Maybe its worth considering +I have looked through the patch somewhat quickly + +I kept the old test method and added a new one +Would this have issues downstream in stuff like the token codeWhere are you proposing to remove the ports? In general or just viewfs +Should we introduce config param such that can be switched off at runtimeDo you think we might be OK with exposing two nodes in JMX when old metrics are not removed yet +I removed the older file that contained my older patch restarted my instances and the problem disappeared +See linked issue for detailed information +Committed at subversion revision +We looked at metadata in Hive and it is really focused around higher level of abstraction such as tablespartitions etc +There are disadvantages of doing this in the Currently the configuration does not support options for the the configuration simply includes the set of sequencers +Anything else for this patch community +This somewhat cripples use of the new form tags for complex enterprise web applications +So I remapped it to CtrlAltSpace and it works fine +Oozie can also use the same service for data availability check instead polling Name Node +Thank you that will at least work if there is a single processing controlling the directoryI still think the bug is a bug though +Is it the kind of service to be expected? FYI Anyhow the issue has been resolved +What properties are you using +Will still try to resolve it though in the next hours +Sorry Jeff but such discussion should happen on the Sonar User mailing list before creating a JIRA ticket +Im going to link those issues +Checked in fix to Apache +Patch committed to. +may I suggest that the date be in the xsdateTime formatI mean including the time experience shows that sometimes it happens that you need to something the same day +These tests require the stricttck flag +that is the final bit of the puzzle. +Currently with the patch the index can go negativeComments welcome +I like that as I have always done the same in my own code +Verified on Tools +yup will finally start working on it +Its simple to save the timestmap into Lucene then on recovery get the max timestamp and iterate from there over the for the remaining lost rowsdocuments +on version +However the connector framework will be replaced so marking as WONTFIX. +think this is a pretty big bug and poor model for users of libhdfs writing their own code to do writes so should go in and +Thanks Jarcec +Can you add a couple of JMS QA tests just to confirm the deployment outside of ESB? +Will you land this alreadyP +I think it is fine provided that the existing code may also not work for checksums +I think it was +Can you verify that this affects? If its specific youll want to submit it downstream +This will allow ripping out the clunky support we have for strategyoptions et al +It has changed since so check for breakage +Vladimir can you say what deps are causing the problem in those examples + renamed to renamed to Some internal types renamed alsoComments added to clarify the protocol +Please excuse me if my message soundedrude to your much appreciated contributions +Should now be supported thanks for raising the issue +sounds good +you need to put this in the build file because the will be over written everyt time you unzip a new nightly build until this is put in the svn + committed this +So you should be sure your workspace preference are correct +Btw whos maintaining the plugin these days +No response assuming ok. +Note the comment please raise separate issues for each plugin +I didnt know if the framework was already providing additional information +This is a good point about the proxy not implementing the will test it with to see if behaves differently and post a question to WeblogicOracle support about why the proxy is generated in such a fashion +svn commit m Added guideline to add new PPMC member guides guides guides guidesAdding guidesSending guidesAdding guidesSending guidesTransmitting file dataCommitted revision . +It contains a simplified build file reproducing the problem without the eclipse compiler using Suns javac +I added that to the home page +Patch files for trunk and +Here is fixup pointing at new logo on site and in UI +Then the client can dynamically decide how best to make the RPC +The strategy should also be responsible for creating buckets that fit within the minmax thresholds +It does not only apply to uifragment and sould be fixed as soon as possible. +Can we close this JIRA as implemented? All sub tasks are in closed state now +Can you investigate your API implementation to make sure its properly picking up the stuff from +Hmm +Hi Paul thank you for the feedbackWell as shown in the attached screenshot the object is indeed encoded with +Asked goffinet to reviewcommit this last night but it looks like got there first after all. +Please let me know if you have any other comments +Rob im not sure if this is a bug or just a edgecase of how the WTP ui behaves +I wish we could create a unit test to reproduce it but that would probably be difficultI have been meaning to ask you if this appender should move to the async project you created and could it leverage the technologies you are leveraing there? This may be naive since I havent had a chance to review that code yet +It would be great if the jdiff stuff could ignore privatelimited stuff too +The math using the value is the same +This code is committed so Im resolving this but the component upgrade will not happen until +You dont want to make the tail longer +The error you get could be determined by the order in which the finalizers are run +This has been fixed by Glen some time ago. +Logged In YES useridOk this is fixed for TC +Updated patch +Whet is performing checkpoint consumers doesnt receive messages during that checkpoint +Good on you Li Pi + +Aliaksey are you sure that you have a clean installation and this runtime didnt come from old version + +so an older client talking to a newer server with this patch may end up waiting until the timeout +HDFS improvements wont help us here +Can someone assign this to me please? Id like to work on it +Latest Deltacloud depend on features that cant be back ported for this Fedora version +This file contains a patch to replace the use of the regexp module with the re module +Closing the issue then +Attached patch is applied issue resolved. +Committed updated for and in revision +Thanks for finding that +I mean I already use so it is not like they are collected without need of memoryAny suggestions +So Im not really sure to get what you want to display +I may just throw the exception variable away +As far as I know but that is only from what Ive read on the thrift mailing lists the versions should be compatible but I might be wrong +Sorry that there are so many times that sound so similar +The httproxy failures with are due to a in which is using simple string search for colon to split hostname and port addressThis is a test issue only though as is used just as a mean for verifying Apache CXF client http proxy functionalities +Merged to the branch in revision . +Maybe I did a bad deploy or my browser cache interferedAnyhow you can close the issue for nowThanks for the effort. +There are that many people who are unable to use svn? +It doesnt matter what Mail is supposed to do if this is not a bug in it is a bug inAnd a critical one James send a different mail from what its supposed to send +I checked the code for the popup component it indeed does not use the iframe fix which is necessary to bypass thediv over drop down bug of the IE releases I will try to fix it for the trunk first since it is less critical there +I cant seem to be able to reproduce your problem any further clues welcome. +Attached patch My suggestion to fix this +I do not think that is a good idea since it will make it more difficult to tell when the replication was actually stopped +Kihwal Do we think this enough to address this issue? I see loads of it running hbase loadings on +Thanks a lot +AliCould you please help us with a patch? we could really use your help +The constructor also expects whitespace tokenized input text +I can run the test just fine locally using ant testAlso Im confused by the audit warning +AdrianIn the clear class there is a property visibility set to hidden which hide the calendar +It seems supporting filename as a virtual column will also allow capturing file name in query +Louis got this one. +If the server running on port is returning that capabilities document there are some issues in its configuration +Hi JayCan you see this issue in JBMm pretty sure there has been a JIRA on this before and it was supposed to be fixed +So if the counter is deleted multiple times only the latest snapshot is kept and decreased from the merged counter value in read path +Pushed upstream. +Ive added a first implemented for which could also be used to update a bundle by triggering a special servlet from the IDE once the project has changed and the build directory is updated. +So I need to push out and on execution plan +Almost like use strict is allowing access to caller or something now +Fixed +This has been done intentionally +I am marking the issue as resolved for now +They do no good if they pass in dev As dev environment using advanced networking and NFS storage but fail in dev Bs with advanced OVS networking and RBD storage nor do they do any good if devs are required to replicate a very specific setup that requires a lot of effort to create just to contribute tests to their code +Referring to H Cat I would put a cat on it +It is a good incremental checkin for this feature +If the test fails I will fix it right away or reopen the bug +Jeff is this release note correct +As always happy to be in on bug reports +filed for further work. + +I committed this to trunk and opened an issue for the Multi stuff. +Does code completion etc +devaraj arun yes +But in and transactional my case would be valid right? Why not support that +Marking as resolved. +We might as well just use javajboss all the time no need for the fallback as the returned object is the same +But brownie points if you can think of somewhere more visible and relevant +this patch is no longer required as contains the fix for this. +attached patch +btw need a wagon release first +In this patch I have changed all methods of and amd changed the class so that they adhere to the new album structure where we insert albums of local and remote separately +Ill try applying the patch and running the tests +Its obnoxious as is +With the case is different +Do you have browser cookies enabled +NET +tried again with revision still get the argument type mismatchhere is some additional infoguvnor is deployed inside a ear the fact classes are located in the ear is deployed in jboss version I missing something? how could I better troubleshoot the problem +The issue was already integrated before our strict bug handling swiftChangeset of our repository contains the patchChange status to resolved fixed +I will certainly implement it for my project and I do not mind contributing it to Mahout +I will update this patch after that +This is because primitive types are serializes as JSON values not as JSON objects and as a result there is no place to add type property +is using which provides poolingHowever you still have to implement a thread or runnable class to use it in a threaded manor otherwise the connections are serial +Thanks Wolfgang! +Have to wait for the Motorola RAZR to charge but this is looking like a Cannot Reproduce +Whatever +Thanks for working on this issue Laura +I ran and locally based on patch v they passed +closing as duplicate +Thanks to Stefan Betermieux. +The Release Notes text field indicates the workaround but is this issue indeed fixed for JBDS so that the workaround is no longer necessaryThanks the script is also just another workaround and this script should be mentioned in the release notes +Just remove Rohinis fix since this is expected to be no longer an issue in +Here is a new version of the patch +You guys are aware that doing such an upgrade will break gems maven compatibility correct? See am aware of that but is it being used much? I didnt think it worked entirely well yet +reopened just to set the resolution to fixed +will try to find the appropriate change list. +extending only implementing interface +for both the patches +Should I open a new issue +bq +its part of junit and has more methods for arrays if i remember correctly +Do we actually support Janino compilation +Prototype API no longer needed new API is available as of +Non default cant be correctly used +I would need to change my process to set threshold add splits wait a little bit and then set threshold This adds a lot of overhead and a lot more places where something can go wrong +It appears that for some reason is not set in the config +Thanks +Verified in issue was about Servers view closing it. +Ondrej this should be fixed can you retest +how does the plugin finds anonymous classes? do you mean inner classes? Can you post an example +I do not want the simpler approach I want the same approach as will be used in GA +Here is the commit id from which I deployed CS bcdffaeecdddbfeacafba +review board would help reviewers a lotAgreed I tried to create a review request but Im getting an internal server error right now so Ill try later again +Fixed a month all issues created in and before which are marked resolved +After clicking it a popup message appears and describes the error +If we make sure that these jars only show up in hadooplibs dir then the current scripts will add these to the class path only if we cant find any hadoop jars on the system +This issue has been fixed and released as part of release +Seems like we should move this JIRA to Common no? The groups mapping stuff is used by YARN as well and none of these changes are +using the only seems to have been fixed in +Indeed wed need to have a solid documentation for thisIf no one wants it crucially I would like to defer this functionality +Hi SimonI just wanted to confirm that only and are the once which need a fix here +Fixing the validation part listening to project name changes to validate that the project does not contain any remote name with the same name as the one in the field of the Git cloning settings page +There now is an option to turn off that still needs lots more tidying up and refactoring but it mostly supports the access scopes now +I tested the fix but the check for annotationIntrospector null does not seem valid when it is set to a It looks like this pattern is in other places as well to address a similar issue +Sounds good as soon there is a Closeable interface in hadoop I will create a patchThanks +The CVS version now prevents class to be generated for empty groups +Patch fixed to copy all files at the upper level of include directory. +Adding the structure to use Yeoman and adding some documentation to use it +I suggest to make the algorithm simple and remove num +This is fixed in the jBPM SP branch which we will use for SOA is no incompatibility with GA so no release note is needed +Thanks Navis +Fixed with commit with latest build fromWe are able to successfully deploy a VM using a registered ISO in an advanced zone set up with Xenserver hosts. +ability to have logic in pipeline and have simple workflow caps such as have a valve the process as the doesI am not proposing changing the currently funtionallity of the pipelinevalves at all +I also needed this and thought that it would be rather useful function for OBR so I had a go at coding it up +Ive just committed this +Fixed in r. +Where we talk? best is to update the doco on googledocs atm +fixed in CVS please test +Good catch +Well already have two largish logos if we stick with the current template +tested this also +You could disable thrift so that cant happen +Do you know the status of his slab allocator? Is there any chance of it being used for managing the direct byte buffer pool +It would be nice if we had test coverage for these corner cases in the loop but it is a hard one to testAnother approach would be to make the log method clean all the old logs each time instead of just the one passed +If possible +has been marked as a duplicate of this bug +Your outputText tag is also invalid. +Will upload them in the coming days to the website +Heres a patch to address the findbugs warning which was spurious but its easy enough to work around +but we can do it in a followup and focus on the metastore changes on this jira +Tyler is this ticket still validThe changes for probably make this invalid for which Im fine with +That will do! Ill try it right away thanks +The code is actually not platform dependent except for file format in proc +Here are the first few tests plus the test framework +Yes this is a user list issue. +Patch addressing Nicks comments +This patch may not fix the issue +Based on Dmitriys patch the following tests should be run to check there is no regression Also the patch needs to be verified that all of passes +The deprecated plaintext result name is missing breaking old applications +Thanks Yongqiang +In both cases I did get a query plan +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +Now I am up to the eyes with work +It needs to be upmerged though +Thanks GailI am not using Envers but unfortunately trying to get the code from the trunk doesnt workFor some reason Im missing classes and other code stops compilingIt sounds reasonable that could be the same issue as what Im seeingIll wait for to verify the fix works also for meThanksGuy +There is a flag somewhere in s administration I dont remember what its called +The expectation is that will be used instead +Ill take a look at the revised patch today +Thanks for providing the patch +I would take a nd look and ping back am working on patching this for Camel and cleaning up the code in +a config step +Assume closed as resolved and released +Id like to gather some comments and then Ill press on and get this resolved +There are two classes that need to change We need to set memoryAllocator in order to create +But if they can be appended to or overwritten then this information could get stale +Jon It is possible to distinguish whether stdin is attached to a tty but you still cant guarantee its seekable +on v +This should also be considered along with which sizes some of the other Vectors more appropriately +the patch looks good for trunk +The data pulled to retrieve the attribute descriptions is too big when running recursivetrue on the root resource +Patch and wrong application context problem solved +But it is not guaranteed there could be two identical folders with only different names that come into the scenario you described +Also Id prefer to keep a number of tools at minimum if possible +The new patch adds sleep within the tight loop and removes looks good +The patch was applied in the revision +The workaround for this is to provide a to the inbound channel adapter and set the receive timeout on that +Ah I see +I generally dont consider the file extension to be significant once the user explicitly associates the extension with a particular behavior +Will you be able to test with Axis RCThanks should be fixed in the latest nightlyFixed in r and r +This patch should fix the test failure +I just committed this to trunk and +the interest of anal retention I updated and rebuilt my copy of JBPAPPGACP and ran with three different times ms failed ms passed ms PavelI havent forgotten about this issue +Is your test environment is running on? Ours is +Good work on the test cases +Noble and I have discussed his proposal above at revolution and come to some agreement. +what sort of goals are we after just mvn clean install? Any other testsgoals to run +I will be modifying these classes further and will submit a diff tar ball with changes once its complete +The Agent would have all of its fields supplied with at least dummy values +The only issue I have w this is that it is possible for the nightly build to fail which could leave the site stale +This patch fixes the problem + +Proposed changes to the documentation on LRU caches +This performance improvement by the attached patch will eventually resolve too +I dont think this is really adding to a kitchen sink or making the example worseNot trying to complain about the time here but its not like i just quickly slapped a bunch of xml on a JIRA issue to bloat the configTake a look at the patch +From a unit test perspective the patch is ready to go +Ive just checked in some changes that I believe will fix thisCatalogs now have a bootstrapresolver that uses files in the +However that means we are leaving s interaction with this feature untested +IIRC this bug was fixed in Axis and hence should have solved this issue +This is the thread dump there are lots of thread accesing the of attributeCache which making the CPU +Patch applied +id be happy to see your patch committedi also agree with you that the peculiar ordering of the entries in a map appears to mask this problem +will commit if the tests is failing can you take a look +Does silently ignored mean users on systems linux wo the fuse headers get no message indicating whats necessary for the build to workYes +Hi GavinNo problem about the Windows VM +I was not going to allow in the forEach parameter +Ok no worries +See also +After regenerating all the stubs from and recompiling the application they all seem to work as intended with all the expected functionality and no manual correction anywhere +Ill leave the commenting to someone else now but I dont expect it to be an issue with the setup perhaps more with the action being not allowed +Apologies this contains the indent and comment fix as well as better comments on the tests and the fix in more without reformatting all of for dragging this out but it really is important codei think we need a lock around the code where we check for TRUNC and we start forwarding +as per email exchange on th of February reverting back to Tomcat installation tomcat installation the instructions say build but it seems to be no longer available on the we give people more specific instructions on the correct versions location +Component tests hasnt been removed for myfaces jsp +While Logj includes support for this I still want to add an additional parameter to allow the file to roll on an interval +Please try the latest and see if this fixes it as it uses the latest +Reported field names are also invalid +bulk defer of unresolved bugs to +Would you please append a stacktrace +It is cloned from where this issue was defined has a new feature and not has a bug +This is not a Studio bug +ok i couldve explained it betterIt is common to run multiple JBAS instances on a single host when that host has multiple IP addresses +It would be nice to have a version of the CLI parsing library that wasnt based on and OO inheritance since clearly part of what happened here was too much OO magic not making the required usage clear +Verified +Now need to cover it by selenium tests +sub ticket for the cpu peg at startup symptom +Committed to. +Thanks for fixing this so quickly +Patch for which ignores CVS and +New patch does not create a new ERROR status on the Data node protocol +Verified by Paulex. +Thanks Mark Patch for regex was applied at repo revision +Here is a new patch with a test case +is pretty old +Can we have confirmation that fixes are in place for +Once you have them running locally we could convert them to run on EC so anyone would be able to run have tried twice to go up to +Are those resolved and can this be closed now +Then I opened to track changing a columns datatype +Once the formal announcement is made we can readjust to the requirements in the announcement if necessary +That exception usually comes after the actual thing that caused the database to shutdown +Maybe we need to track the forked build along with the build result and detect it has been killed +It has tests showing the issue with correct results after the patch + +Doing it this way anyone that want the change in behaviour can request it and everyone else continues getting what they already did. +Now that work is well underway with AS all previous community releases are +It cannot be a test against instanceof as this will return correct results for and as well +It is likely to be a problem with the test +bulk defer open issues to s the current status of this +Ill assume pretty soon you changed your mind and close this issue as not required +Patch applied in revision +I just committed this +If you dont mind Jonathan do some more runs +Nothing can be actually done in a browser that does not have javascript available +Hi Kimmo RieskaniemiThanks for your supportCan you rsync once or twice a day instead of once a weekThanks regardsHenkrsync has been changed to run once a day +Committed to branch for release +Thanks MarkPatch applied to auth at repo revision Please check that the patch was applied as you expected +Jiri that sounds to me like the correct lookup should be done in the related projects sourceclasspath not in the eclipse workspace +and configuration would most likely result in a new dependencyLets focus on system properties based configuration first and see where it is going to take usOleg +Since this exception wont impact server use so defer to wish list +renamed to +The workaround will bring down all the existing present on the HostThank youChandan +Current design adds extra overhead for every inodeThe design for storing snapshots in the NN in the proposal I posted would add integer startsnap and endsnap fields to every in the FS in order to track which snapshots a given should be included in +ok sorry +closing as issues are fixed in +Xuan thanks for the suggestion +The bundle contains sources required for building the executables setupping the server and running the scenarioThe files in the bundle go to adapters and tests directories proposed in new BT infra +Ill second thatIn my experience on some projects you cant have application configuration files in the classpathSo being able to resolve a location that is outside the classpath is needed +Initial commit in Git ID again if further improvements can be made. +java files contain +Based on that Im dropping from the fix version and leaving it atBig thanks to Arpit for sticking through this tricky issue and incorporating the code review feedback +Next version is either or +bulk defer of issues to +I have to agree with Dan +It looks like our erstwhile squire Hudson has wandered off again +Did not quite get what the issue is? The common container spec token should be immutable after creation right? So why is the simple getter not working? +Thats correct but is what worries me +after all Harmony is an Apache project so we can take and change code from them with no restriction +One option is using the old log files but also the log files that are currently in the region servers +I very much like running unit tests in my IDE as it is much easier to select a subset of tests to execute +Am looking further into it now +applicationjson that is easy to route +Master can create regions and do the moving of data from old regions into newOne optimization were doing +Snjezana Is there a preferred way to implement supplemental projects to a zip that needs importing +m or the root of local id like to be able to optionally configure it to my liking as the patch i sent permits +This error occurred in two situations +Need to check against JPA spec but scheduling for next release to make sure that happens +Hi TonyWould you please try this new one +This quickstart mimics a scenario where we use our webservice components to provide a webservice interface on a backend EJB based service +bq +Duplicates which is newer but has lots more details +Bulk close for +Note the svn push will happen after is resolved. +Checked in to git master +I was going to write in response to your last post that I think it would be a valid approach to set properties that refer to other beans in the application context first and then set other properties +I think nodetool will need an update +is not even a part of the parent AC +Test case may help move this along. +this issue is for the not Myfaces +files rev Added openejb file rev Jason can you recheck its all done +Bulk closing stale resolved issues +Please raise an issue on the Hibernate JIRA this has nothing to do with checked the from Hibernate and this has been fixed in their tree +still doesnt work but will fix that in a separate jira thats likely not setting credentials correctly +Ive implemented this but Im going to leave this issue open to see if we run into any issues +Thank you. +Closed upon release of Hadoop. +It is a really simple fix +Nothing from me +When nothing is provided in these fields is it not intended for trunkIts not clear often having it unset means people dont know what to set it to and trunk issues are set to +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be I totally misread the May solution I would like to this ticket and raise its priority +Resolving this as the issue is not that it doesnt reload it is that it doesnt reload the password fileA new JIRA will be forth coming +Postponed to could you tell us the hibernate dependencies you use on your project? In the meantime I postpon the issue +I tried resetting pwd but did not make any difference +I think there is harm in incorrectly suggesting that those classes are serialization and that people should be serializing them +dev should close this from the last comment this doesnt seem to be a problem anymore +This is not a fix but a workaround +Or our own + +Sorry that I forgot that I meant to introduce this new API to the trunk +Tab is now escaped too +Fine by a good while back +Does that work for youAlso plugin DOES automatically pickup projects from the workspace since +The JON Dashboard Settings Alerts AddRemove Resources page also has this all of the dashboard portlets that have add remove resources functionality piggyback on the same set of pages the only thing that changes is the context passed to the pages +Not reproducable on LINUX and MAC so bumping to issue have not seen updates in weeks and thus it has most likely been fixed somewhere else or notbeing actively monitoredIf your issue is still important for you please try latest released version and if the issue persists or the feature is still missing please open a new issue with a link to this issue. +core schema is well covered now +I cant find a source for that information but I did read it today while researching the requirements +Hi Hiro! Not a problem will do +Thank you very much for spotting and reporting this +The nd one is optional and if provided it will be used to derive the respective target tablesIf the case of only a single table name and no mapping will be allowedOr maybe I should get rid of that option now +Thanks Brett +For status command +Committed +Unresolved issues to CR +Also added a unit test +Probably its enough to be able to specify a global term boost factor per query for all synonyms so Roberts method would work for meAnother usecase is Phonetic variants +Thanks Vinod +The in the diff file has a small typo should beI have been to hasty +Will commit if tests pass +New patch uploaded to address the issues you mentioned +We used to cleanup millions of rows from our that we no longer needed +Im fine fixing it just for CLI then +I looked inside the runtime and clicked the Unistall Facets link and it says that the JRE in question is as opposed to relating to the stuff Snjezana Peco was talking about +If were making frequent changes to the site then this becomes a bottleneckShanti +and branches will be updated in a further svn merge over +Hey Rob just wanted to see if you needed anything from me on this? Ive been really swamped and wanted to make sure you were not waiting on me for anything +Some log fixup +Yeah theres not a whole lot you can do when you OOM other than maybe shut the process down like a C program does when it segfaults. +Could one of the reporters attach a working example POM to reproduce this +bulk close of issues as part of making release notes. +Hi StanI tested again with the new snapshot and I get the same VictorI reproduced your problem +Changed the patch to incorporate the comments +Folks who want to use the tools need an executable jar that includes all of the dependencies +If you finish it before we release beta please correct the versionThanks +Ive also noticed a mysterious file named out in build + +Can we turn off the? Is there any harm if we turn it off? Thanks +I forgot to handle uncache on deletemove and the only cache finalized replicas checks +The second rev gets the tableid from a more general location and will work better when we add a constructor to the vtiTests passed cleanly for me on this rev +If failure is detected mark all pending calls as done +Ok I will but tomorrow +Removed changes to +changed title to more appropriate name and rephrased description +Rather when I try to refreshreload it from anywhere Cloudstack UI is redirected to the Eula license page for which a bug ticket has already been created +Case I used the source of and sended request to start the RETIRED process occured +Ive made fix for this bugIt contains patches +Were depending on infrastructures here linking issues +IMO people dont use the project that way +If you find further issues please create a new jira and link it to this one. +After the entry in is removed but still the master does not know that split has happened +This is a pretty basic fix +It would also be great if you could enable more logging by setting the following propertiestrueand attach the contents of when you see the error +There should be several other jiras concerning this +Nick mind providing me with a patch and Ill make sure it finds it way into the current codebase as soon as possible +A new patch is uploaded +I added some tests +The repo is now up to date revision ACE POM also references the repository revision +Thank you Thomas and welcome to the brotherhood! +beta uses I see the error now +Unfortunately I cant vote since this bug has been closed as wont fix +I cannot comment on the correctness of the notify stuff but on logging the errors +But Im ok with renaming to +Duplicate so closed and superseded by +If you want clean markup then use the XHTML setting to turn off the language attributes +I provided the diff just to share what I did +bpel files +What if its an increment? Or do only puts and deletes go there. +No we dont +Reopening to check EWP +It allows you to override whatever settings you want if you dont like the Canonical defaults +Patch that handles this +maxretriesmultiplier does not make any sense for larger values of maxretriesmultiplier +It contains the code for and supporting code as well as implementations for JIRA tickets andBecause implementations for these are included you can run the performance tests without making any changes to trunk +A note to the API documentation may help +Let me know if this behavior works for you. +thanks for reporting thisthe current trunk has already fixed this +Modifications in Sonar or PDF Report? No problem to use a minor version of Sonar +If that is expected we should update the app template +It would be a worth wile to look at handling code in Wicket and make sure +Any news on this one? did some digging but then forgot to comment here because sourceforge was down and wouldnt let me commit my tests for ve added tigertest which reproduces the situation reported in this issue +Commit ID edffdfcacabde to fix regression in. +Attachment to the +Whatever the format is I think we should serializepersist in a way that the timezone stays with the datetime +Yes that makes sense. +Now putting for example binary attachments to a message header property seems to be a not so clean workaround in my eyes +Lets get the coding out of the way first +If is thrown it should be caught in the original code +I have selfRegistration enabled +Thanks +That sounds like a good plan +when package limiting scope is not supplied using the in option it recurses the entire dependency tree available in the classpath + +Yes this feature is planned for the next version +Well if you create a project with eclipseeclipse the projects id is the artifactId +Attached is the patch running derbyall right now +version of document with CSS for better visual output +Thanks Ted +Thanks Jean. +Yes that will do it +Moving to because the parent task was reassigned to +There seems to be no consensus on this issue besides that inheritence doesn t work here though delegation as proposed by sanjiva might +Replicated this one it only occurs for deleting in a fix for this it wasnt checking whether the property value to compare with was null when looping through preferences +Please check that it works +Infinispan log from first booted torquebox to +But theres one major issue how to deploy the process definition? Can you take it from esb archives and deploy it manually via jbpm console +Using both i feel that they should provide similar this is really minor the main reason for me is that in our issue tracking system when assigning something to someone is always good to say why and say please +Thanks Tim +Regression Test to be addedCan the fix be backported +I have uploaded a diff of the proposed patch +Build structure is all changed now so this is no longer the situation +By looking at the content of the dead code can you make a guess as to what feature it was in support of +Review for the descriptive comments +This is may a duplicate of m not certain this is a duplicate +Hey Kaz +all that it produces is a JAR containing the class files that are in the EJB source + +It is actually an issue +HiThanks for the quick how to +A bug in this feature cause that a session was not ended when an exception occured during the request invocation +Attached WSDL and XSD files to reproduce the have read the slew of problems that is very similar to this issue +Joeadded a sample to the showcase webappPlease try this and tell me if you have issues with this as wellThanksRainer +add and remove projects is not something that automatically deploy or undeploy have to explicitly invoke run asdebug as to see actual deployment on the server similar you have to removeunpublish it explicitly to do an actual undeployment +libgriddevbin +I was able to reproduce the leak on current master with the example application and am digging into the root cause now +I cant reproduce this one +Fixed in . +If it happens again please reopen or file a new bug. +defer to +catch where you process the created exchange +orgapachehadoophbasereplication +srcmainjavaorgapachewicketexampleslibrary +The initial implementation had the notion of requiring ADMIN for any table op that has global cluster implications but then went back and changed that to allow at least users to create tables if they had CREATE permission +The result is that a consumer hangs waiting for a message +I think this is the most satisfactory solution as otherwise we will be introducing some kind of started but not running state that might be hard to understand. +Thanks RobertI tried contacting Owen but havent heard back +Maven +Yes the second request was sent with the same jsessionid but AFAIK it doesnt means that they are using the same hibernate session +I just committed this +Too much is going to make it hard for people to get started +Any idea why the testing is not finding the librariesI submit the patch again with spaces +Can not wait +point we need to reemphasize is why its better to use Widget Form and Minilang rather than GrooyFreemarker couple and Java productivityFor Widget Form vs GrooyFreemarker couple its mostly because it allows more reusability and extensiblity its more compact and clear here experience can tellFor Minilang vs Java its easy to understand that you dont need to compile and especially reloadUsing OSGI for components would help for the reload part and component compilation is not that long +I was hoping to focus on and with this backend category as it applies to ASEAP +I was able to builddeploy the test case and deploy it +Do you need to expose the random feature in the CLI? Can the test not choose this itselfPlease add a unit test for +Rado can you contact me when further runs are clean? I can release a within a few minutes +This problem may be related to issue +Im not sure which is more correctI changed option so it can assert on message tokens and not the message strings from the resource bundle +This is needed forM javatestingorgapachederbyTestingfunctionTeststestsinM javatestingorgapachederbyTestingfunctionTestsmultistressM javatestingorgapachederbyTestingfunctionTestsmultistressChange database property to and add propertiesM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiExecute procedure tests only if class is presentA javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiThese tests are part of jdk suite but cannot be run with JSRM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiM javatestingorgapachederbyTestingfunctionTeststestsstoreRemoved runwithfoundationfalse exclude from these testsM javatestingorgapachederbyTestingfunctionTeststestslangD javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangD javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTestssuitesM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsutilM javatestingorgapachederbyTestingfunctionTestsutilmax and triggerStream tests are converted from sql to java to avoid using procedures with nested connections +Yes same for inner classesThere are a few differences for logger naming and chaining but maybe specific to our usageThis class has been validated in production since a few years now with a LogbackNo hurry for but something interesting to get out of the Avalon Logger ghetto +A sub step is to merge and into a single class +Otherwise this will have to wait until the big refactoring I have planned for next month +Fixed in master +We should stick to POSIX for file manipulation to keep it generic +I would like to contribute on the maven gwt plugin as well +Anybody with the same problem could test it +Project examples are not part of JBDS installation +I see nothingin the documentation that would lead one to expect that result +Based on these two facts My best guess so far is that the and needs JAXB annotations +Oops +Verified on JBDS. +Sure will go with thatThanks for the help +I did run the unites test before and after the proposed fix and only found one existing unit test that failed +Reopening this issue due to ijjj issue with IBM JDK +Read the release notes and documentation you can already switch bytecode providers +For some reason the activation part of the profile is missing in this new pom +I did a major and long refactoring cleanup that allowed me to subclass the appropriate artifacts for as remote behaviourThe initial commmit of remote starting is finally in +Mmm this looks interesting +Right now in the route I have additional step that converts the id to string before setting the header however with this improvement it would do this conversion automatically +Manually tested +Marking unassigned +Logged In YES useridAs the testsuite is currently failing for me I postpone thecommit until this is ok again +Its also fixed in trunk since the above date +Committed. +Continuing. +Convert them to jiras please when you see them +This is now covered in the section of the Roller install guide +this is first step +Thats a classical Either we want to report any possible lines of code but in that case were going to generate some Either we do not want to generate any and in that case we must log a violation if and only if the probability is really important +It is most likely fixed in the trunk as part of in the diff i agree with Noble it looks like his changes in r fixed this as a side effectresolving for now someone reopen if were wrong +But I found interesting usecase with enabled time of setupdate property grows up a lot +The ordering defined in the is still not supported +Introduced openPane attribute into all form tags +I have not reviewed the patch or supplied an implementation to geotools as such this bug is still open +Thanks. +Closing stale resolved issues +That would solve it as well. +Please revert itNamely inside ecommerce there are chained surveys +I I do not see any drawbacks here +We only use quota reservation and weight are present in the fsimage and edit log so we can later implement these without doing a metadata upgrade but are not present in proto format so clients cant actually set them + issue is open quite some time on the Signavio issue tracker but up till now nothing has been fixed thereWhat do you mean with workaround +Otherwise What is the example usage? Theres no alternative class added does it make sense to add it here too +Ill have to improve on that thinking about alternative implementations or make this configuration illegal + +Thanks Vinod. +Im running Cassandra as administratorLast night I found a probable cause of the problem +I think we never can guarantee the order of them +I cant find any reference to them in the apart from them being added as imports +If we had to make timestamp equivalents for each one the API would become cumbersome +Yeah I was originally thinking someone could choose to create the Iterable and delegate the Iterator creation to that method instead now I am thinking of making the method return an Iterable with an option to return Iterables for nested Pairs +But I dont think it is worth spending time on that right now when there are real bugs to fix +The latest release should be fixed so Im marking as fixed +Does this new patch look betterIf so Ill work on some unit tests +Suggested fix should solve this in this JIRA can be used as a quick temporary fix of this problem +This should consist of Creating a deploylabs directory +So I did this as part of a long effort at hardening recovery +Updating diff for engine diff for engine packages under iapi +It is only on the trunk +Added to support specific convert signatures but did not add a general typed based mechanism for other system functions as they generally only have a single form or are highly related such as supporting several numeric types +I want to set the fix on for this since the workaround is good enough + Username dsundstrom +Well yeah I could +C has qnoeh and qnounwind which undefines the EXCEPTIONS macroCompaq C has noexceptions which undefines EXCEPTIONS macroSGI C has LANGexceptionsOFF which undefines the EXCEPTIONS macro +Thanks! +Youre talking about +Patch adds test cases as suggested by Doron +As soon as we switch to JAXP Xerces or later canbe utilized +Setting component to. +This was fixed +I can run ant to comment those features +as a workaround in my case i just disabled the valve since it wasnt crucial to the components involved in this issue are too old to really test now. +The previous experiments are too simple +Jie sorry I missed this ticket before +for the patch +Import will not work without a wizard that reviews the facet settingsI am not sure what facet settings you are referring about +Maybe we could add set as a temporary workaround but by the time you call set isnt it already starting to bootstrap with the value from the config file? I think youd need to make this useful +Raising prio to blocker for +UmaI will change this while commit +You can check it by running mvn dependencytree on the trunk folder +Zhao Is this still an issue? On my box we only bind +Please ignore my previous comment I found the silly error +Not yet +For your first issue xsltproc should be listed as a dependancy of the package +revoverride property no longer throw weird errors in the ui its hooked up to the actual domain model now via EL +Thanks it works fine. +I would like to get this in somehow pretty soon +instead of owner have owner? In the interim I will add a check for owner as field name to move things along in Roo +If we dont have that we wont be able to do anything +Anyone can see firewall rules for anyone else +Steps to reproduceGo to Example main page using Tomahawk theme +IMHO adding crawlparse to Indexer only complicates things since you need to handle additional conditions not to mention the performance hit to process a lot more dataInstead why dont we simply omit storing empty content in Fetcher Fetcher when we get a redirect that we dont immediately follow? This would solve the original issuePlease see the patch to Fetcher that illustrates this issue +Fixed r. +Yeah I think thats a good idea thanks! Ive updated the patch +Thank you! I was able to reproduce current source files from +I am not sure why you see the behavior you see it is not expected at all +Maybe there is one already and i am not aware of itJust my cents +Patch incorporting fb from Sids review +Moving to since well want to consider this in light of. +The complete deployment exception +This seems like a relatively straightforward fix the root project is last on the list not first +bebbaaddededfdcca +The violation errors exist in the controller but are mysteriously lost when rendering the view + +Lets keep it a dumb and simple system put the correct field name and get the correct field name +In the branch using same SUN RI configuration parameters +Thanks +Resolved +I dont follow Ted +Divergence from java. +Do the above add up to a on v of the patchNo more like a +its all yours! +Why do we need a wrapperWe dont have any explicit lock for an application attempt +I think what happens is all bugs are initially internal to Oracle and they need to be opened up to the public +NET Active STS for DOTNETBS and WSASBS while using the Metro Active STS for the METROBS +Copy the license header from another file to put at the top of any new file +I just committed this +Thanks Regis. +Master will assign it once the region name is passed over +Hint contains a workaround for should be fixed now with +Ive removed +Improved the original check in r +Thanks Daniel +We are still evaluating its impact on us and looking if there is any straight fwd fix which exists for us +Complete port of the imageop block to Cocoon +Should be in a release +I have a question that the document doesnt address what if I write in C of another language does the java directory become or cs or what +We should probably add to the book that currently disabling block cache completed is not recommended +Well get this into there any other helps for deploying the Enunciate war into an ear that you can think of +No objection from me on commit +Any reason why you chose to write to a file called instead of just or some such name +I attached a patch to handle link names like entry names +Special characters will not work in plugins right property files of plugins are not loaded in the same way as the message properties of main web debugged loadProperties and you can see that a processed version of the main web app properties will be loaded but in case of plugins the original properties will be used +Committed revision +Tests pass but I need to look it over again +Actually the version number is in the parent pom so I updated it thereThanks JavierCommitted revision +Dimitris can you create a tag to include upgrade in EAP? Is there someone else this should be assigned to +I added them as patches now +Plugin has been created but it still has some issues that need to be ironed out. +If anybody is fine with it should we first create a branch for this upgrade +Backported to branch as svn closing. +My for not doing this more complex than it need be aka +Thanks for the feedback + +I do not have this NPE on the latest trunk +This is already fixed on trunk +Linked to general improvement in error handling for meaningful exception if or attributes are not configured. +I ported about of the tests to C in branch +We should make this do the same thing as Java instead +Im a little confused is this a patch for the Network Client code or for the Network Server code? Or should Iapply it to both code bases +Instead I implemented them so they rather worked on the whole stream at once than in chunks like it would have to being implemented as +Committed. +Hmmm +rootetc usrsbinlogrotate d v etcreading config file etcincluding etcIgnoring apache because of bad file modereading config file aptreading config file aptitudereading config file conntrackdIgnoring dnsmasq because of bad file modereading config file dpkgIgnoring haproxy because of bad file modereading config file monitIgnoring ppp because of bad file modeIgnoring rsyslog because of bad file modeHandling logs +The patches look reasonable to me but I dont understand the relationship between the thrift versus the normal thrift serverIm using CDH and the log messages appear to be coming from the thrift server not the thrift one so selfishly Id like to see the thrift one fixed first +Remaining XML validation errors in +Added which is all patches in one so that Hudson can run it +I think that strings are a really bad plan +It looks like we can check to decide between inetntop and inetntoa and if there are more than bytes just display the escaped bytesIm thinking about just using inetntoa if there are bytes and inetntop if there are more instead of using inetntop for both + +I found the null handling a bit dubious as well +Parser is a big improvement +followed by a comment on this issue +The only way to change it is to patch MylynI can do it if you want +Are there still plans to add this to the the version distro? I dont see it in the distro yet +Yes it needs to invoke parseArguments to take side effect on setting inputPath and outputPath and other options. +Thanks KnutI verified the x b option is there for svn diff and svn annotate as well and filed for the issue +Do you expect another release +Ok so you would like an improved exception message due to conflicting getter definitionsIll try to reproduce as described +Moving to of +Maybe the default value for enums should default to the first declared value rather than ? +This is making good progress +Patch to differentiate between GPLGPL and GPL with testsTested on a local one module mvn project with JDK +Configuration doesnt use to be a performance critical part of an applicationDid you measure the overhead of enabling the interpolation? I would be quiteinterested in seeing the figures I think no one studied this before +The only blip on build instructions was for MSVCI built it as follows configure export XALANCROOTusrlibc makeThank you. +Commit it. +if not we can just purge that page +Would you have a chance to prepare a patch +The username and password can be set in the file +Arnaud defaultfixed values dont need to appear in the XML thats why people assign defaultfixed values +There we only compare the values since this test always writes the same values the index entry is never deemed stale and so we dont ever write a tombstone +Venkatesh may have an additional perspective but the principal use case is for data movement across grids with different versions +Something todo When generating the concrete part of the pair check to see if the file exsists +Bulk move to. +Im not convinced that using an DB is the solution given that the test is designed to scale up to hundreds of thousands of insertions its probably better just to piggy back off a real DB +The idea of the volatile boolean for writes is that it will be set only if while scheduling a flush we detect a over capacity problem +Thanks is a problem in terms of the images +fix along with a test case +I didnt know that it is possible to directly implement Type and plug it into a mapping like a +The second attempt to connect however doesnt fail for the same reason and instead fails because of the +Ill try and look over the old ML archives after I get the base implementation down +completed +I remember this one +I would like to see only the parts that really are these numbers? The factor the new version is faster? Well for the empty string that may make sense but that is not really a relevant case I think +Please test if this works for you Derek +This issue has been fixed and released as part of release +I have experimented a bit more and it seems that the problem is somehow related to the or Oracles HTTP Server implementation +So looking there appears to be a hole in the fix if a recovery is already stacked up and waiting it was not checking the shutdown state before starting a new recovery fingers crossed +Is this math correctDetecting when to unRaid The patch does this using a wrapper filesystem over HDFS +I have addressed all the comments pointed out and have also fixedthe failureI am running tests on this patch and shall post the results of therun by today afternoonI request for this patch to be considered for reviews and comments +this may not be terribly useful +I imagine its either because of wrong naming conventions or because it does not conform totally to some Tycho new stuff + +Thanks. +bulk close of resolved issues. +Fixed with +Services running at this code will reply with delta +Reading the docs on driver reloading it appears to be something that would be quite difficult to get right because of how we have long lived ports open to the driver amongst other thingsSeeing as the driver code is relatively stable we came to the conclusion that the answer to this particular problem may be to just not even attempt the reload for now +modcluster was upgraded by has modcluster which should have these issues resolved +Patch to fix convention plugin for Added new utility method to convert s vfs URL to physical URL +I think the abstraction should provide an api to delete replica files independently on whether it is or so it makes sense to me to keep the implementation details concealed in the deleterUsers might need to know if an operation is sync or async +the omits the namespace from the plugin list does not handle artifacts with a +Did you upgrade directly to or one of the previous versions first? Please check timestamps in your schema and see if they have any future dates +This seems good to me + although I dont have access to resolve this ticket +Thanks for your help and confirmation +I have merged this to. +Changed to use Is now a abstract base class Added missing class Move Bloom filter logic to common utils classes Use Service Providers for pluggable choice of hash algos? Expose settings for memory mark I think you should really not provide any field handling at all + +Leaving that for its not mandatory and TCP is now the default shows good performance using Netty +It is the correct patch though I just forgot that I already submitted a th version of it +Hi MahoutersWe worked on the features requested and are proud to present a first unitWe implemented which can process a variaty of graphs which augments a graph with degrees to increase stability for future processings which enumerates the triangles of a graphThis three jobs and a set of tests to proof functionality and usage a available fromhttpsIn there is a mvn project that builds against +Dprojectcamelsvn commit message Fixed problem when bean processor did not set invocation value at all +harmless patch and looks goodIm surprised this is the only leak there may be others we can find by digging through the code +First review of the patch is very positive +It looks to me like a JIT problem so I changed the summary +I would rather have the ability of using browser history whenever possible +I dont think the program listing for the prefix should be copied into the docs but instead just use a link to the javadocs for the the class and then just include the bit about setting the values on the HTDAlso I not loving the camel casing or the underscore in PREFIXLENGTHKEY but I can live with itotherwise +I attached +if we do it else where we have to adjust all JCRapi method calls that take a path string and make sure identifier paths areproperly handlednow that identifier handling has been refactored to a separate class i thinkit should be feasible to use that in a path mapper to properly convertjcrPaths to oakPaths +I just ran into this with Oracles update on Windows +Fixed in r +New patch with javadoc warnings committed the patch to trunk +That should clearly state that it is not a usage issue +Changed the property name from my initial description. +Are you sure you did specify a password to connect to the SSH server +It would be possible but it means that this command will update the configuration file +Might be good to keep them around though so could get an idea of what was changed via zk +Committed patch at subversion revision . +The test is powered by Arquillian and its Drone extension +sigh +Looking at it wont return false for a standard JRE collectionmap anyway so when it does return false and this exception is thrown the type of the attribute should be useful information +I keep hitting this all the time +which doesnt follow instructions from Max doesnt say that you can Select an existing application +Patch added +It seems to be good enough for our purposes +Miscellaneous fixes done in Running Total functionality now no server error occurs on calling +defer all X issues to +Problem solved +code rvw +Trying hadoopqa for kicks +I would register the extensionsystemstatus as a web console plugin to provide end user status info +Unit tests pass locally +I verified that for has been applied to and the issue in still exits +I changed the Fix Version to +The REST UI does not use HTTP sessions +Honghua My suggestion would make your patch smaller +oki then it seems I at least locally patched it the right way already txsAnd dont worry I think was a big boost again in terms of ideas knowledge and of course motivation which will overcompensate loosing those days easily +Im now only showing items which I believe is correct +If I am getting this right this should occur in any query which accesses timestamp column from table since we need to serialize Timestamp objects in plan +The sequence works in JSON mode too curl u admingeoserver httplocalhostgeoserverrestworkspacestoppdatastoresstatesshapefilefeaturetypes tmp edit the title again curl u admingeoserver XPUT H applicationjson d tmp httplocalhostgeoserverrestworkspacestoppdatastoresstatesshapefilefeaturetypesThe issue does not seem to be format specific +Sorry I made a mistake the plugin works correctly! Thank you so much +Tried the build in nightly for and got the same resultWas there another one I should have tried +Thanks! +Mostly looks good +If we decide to go for JMX we should close one of these as a dup I agree +verified in Developer Studio to set release note note not required for now that release note status has been set +Reassigning to Ted since hes in the midst of updating the documentation and hence the web site right now +You can check it now +I will commit the patch later in the morning then +Attaching the patchContributed by David Lao Bikas Saha Lauren Yang Chuan Liu Thejas M Nair and Ivan Mitic +you guys will have to talk to arjuna is customers use case +but i looked at doing this before and did not find much to work with +Whoever is finalizing the list for webapps can you please update the README to include all the samples that should be in the distro? I would do it but it seems like wed both be working on the same thing +Committed the missing files +Applied in r +Please close the issue after you verified that the actual change solved your problem +Does this method belong some other Class +war plugin build time s war plugin build time s avg +Its bitching because taskType should be amptaskType +If you want I can attach a simple that makes any Map a SetStill I think that using is best here +Waiting to commit on till the freeze is off +yes +I dont get what this is needed forhibernate console configuration uses JPA to autodiscover the entities so why is a file needed +Committed revision +bulk defer of unresolved bugs to +It sounds reasonable for the depend on the libhdfs +If its of any help the repository Im publishing to is a Sonatype Nexus managed repositoryAt least in my enviroment the simply issuing the following is sufficient to reproduce what I said +Could you please update your working copy and relaunch it? If you have further question about the build please open a Jira issue or send a mail to the mailing list +done but is +I ran it times successfully in both debug and release modes. +Could this also be used to implement +Committed to the branch +dont see this anymore with the latest snapshot from Thanks. +Manually done. +This patch covers it +Because of the requirement list I really dont think we can get any of this done without api modifications which would mean +Running them on separate jvms might improve performance isolation +I am proposing that this issue be closed as being a Groovy not a Gant problem +I tried the client with a C service and it worked for records +Suppose the parent tdiv renderer property is tied to a value expression and by some reason a message is output for the inner components +Personally I think this fix should go in and additional featuresenhancements be requested separately +I am currently refactoring the code to incorporate suggestions by Claus and will incorporate all your suggestions into the new versionBTW though the implementation defaults to Blowfish it is definitely possible to use AES as well +The use of looks fine +Hello Would you please try my patch? To apply the patch please +Pavol verified that it works correctly and he is adding metamer page for it. +It seems not work +On editing edit of when we move further to complete the process it display edit with unchecked Billing address is the same as the shipping address option +Are these messages annoying? If there are others whod like them changed to Verbose please say so +This is fixed on the trunk and the branch but needs to be propagated to the branch +It appears that everything is working as it should the difference being how much CPU loadcontext switching is required when the client runs on the same machine as the region server +A call to readResolve as suggested in another issue would solve this problem +I modified the code to throw an exception for the above caseIts checked in r on trunk +to Matts proposal in particular each language shouldnt depend on the build infrastructure of the others +Duplicate of I had originally made this a duplicate of but meant it to be now triple quote strings are highlighted but there are still some problems with slashy am removing the duplicate remark and then closing this issue +I just tried the new connector built with ant +The view should create a linestring with a single point +Unassigning from me so a committer can review the patches. +If no one else claims this for I guess Ill do it +After working a bit with this and registering the listener with the code above Spring still does not invoke the on methodI recon we stick with the workaround by Willem +Thanks Eli for working on part +this is still an issui in V Konraddoes that mean it works for some time and suddenly the attributes become uneditable? Are there any error messages poping up? Could you please check your +Attaching an updated repro with cases for suspendresume of a transaction +Proposed fix checked into will check into head next. +in the future may wrap another list that is not serializable +The release branch is no more actively maintained. +Patch applied on +java classpath +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I should be able to submit some benchmark results soon but first I will put up the pseudo code +The classloader in CP is not a general one +Please keep an eye out for thisRegarding Context constructor have modified it to use as suggested +Im only seeing this when I do SFTP operations is this the same for you +It might be worthwhile to move two existing API roll and roll from the client protocol to the procotol +I am unable to reproduce this +Staszek suggested that perhaps it would be convenient if this patch detected if another Polish stemming library for Polish is present in classpath and if so used itHello perhaps we can provide this as a separate optionThe problem with doing this is what to do if the classpath detection fails? would it fall back on the code hereThis magic could cause a lot of confusion for example indexing on a server with correct classpath searching on another where it is incorrect orIn my opinion for analysis stuff if the environment isnt setup correctly we should just throw an exception +It didnt even pick up the patch the first time around +this is full was not part of the original patch r. +Removed unused templates at some templates in at all template files at r +If test is positive Ill commit to trunkAlbert Lee +I agree this is almost certainly a duplicate or else its fixed by the same solution. +Added an entry to the documentation Start the Chukwa Processes section + set fix dump file contains roughly K revisions allof which are empty +Bug introduced in +Did you guys find and solustions? I am using Axis Thanks years old bug +Fix in rev Fix in rev it +Thanks nkeywal for review + +Yes I think thats a good approachAlthough this issue is not a real bugfix it still has some security related things so this together with the size of the patch I think it would fit in remove add change package name of JSONRPC in small patches to +My preference is to focus on the JettyDerby bundle in the official docs and point to the wiki where the community can maintain information on all the other possible combinations +This is caused by changes where I now actually print out the cause of an exception in JDK whereas before it was just swallowed up into the dark space of the runtime +I didnt change the config name yet Ill do that after an answer to the question above +Once this work is complete I am planning to do the following Port it to branch to make protocol compatibility of branch easier +Adding the memory usage tester the class that is used to find out the sizesof other classes +So when a VDB gets deployed with one of these translators the error for data source is set on the model along with metadata load +DavidI think you believe this is resolved now +Now moving this to the next release. +Uncomment the elements in the contextzip +I changed the Release Notes Docs Status to Needs More Info because its not clear at this point whether or not this is likely to have been encountered by a customer and therefore worth documenting as a known issue +Some are because is not in the path I am trying to fix that +The tasks works fine as long as you set the property in +If Ilya needs to assist in this please ask. +This should be a minor change with no performance impactMy original question was more to the effect since this is already built into the base class why explicitly turn it off? The only substantive reason thats been stated is performance yet no benchmarks are provided indicating how much slower io is when bytes are passed through a native CRC implementation +the above comments resolve the can be added by user if they need itwhen they set a store to No for allow password +Weird +Its because the prefs dialog updates hidden inputs when you checkuncheck a boolean preference +At least you didnt respond to thatThanks for taking your time am not going to get into a argument hereRegarding the bugI did not say that the property you were asking for is out of questionI just didnt think it would work for read only media like which youmentioned in the original bug reportWhat you suggest would probably work for you I agreeIf possible could you please submit a patch with a unit testI am currently very busy with a number of other much bigger things in life andthis would speed things upThanks +return for all of the above sites and would be used to determine what site each page is indexed as +I believe the jndiName property will override them +Committed to trunk. +Out need in particular is for pig hive impala and to be able to query table definitions without each having to be configured to understand the underlying table information + are not always bad they idetify real issues +applied the newest patch in revision Vladimir please confirm that it resolves the check again at revision it is OK thanks +Could do +Thanks Jasonim sure you wont regret it +Final patch including both patches of Wensheng and some improvements of code formating +However since this is an expected event we do not want to discard the log as corrupt +Im stuck at the moment because I dont have time to see if this matters to us or not in the midst of work thats big enough that I dont want to confound myself with changesDid anyone look into the root cause of this +Gytis can you explain what problems this causes when this is absent and why adding the annotation is the right thing to do +They have a defaultsearchbase value set which redirects queries to to the dcstanforddcedu base +It looks like the issue is onlyThe fix has been applied to the updateEntry function but not to the getEntry functionThanks +Thanks +can u explain what the problem is with Option +The intention of filing this Jira is to show illegal casting which is happening in pig +Thats why we should always prefer working with Unicode escapes +What would be the syntax to create userpasswd combos and logging in? Are the permissions stored in metastore are per user or per table or a combo? Do we really need groups? I dont think implements groups +Just something I put together while looking at this +Where to show caption VPEEclipse Status Line +Paco please update we need more information +Move forward to Andrew +Looks really bad revision +Closed by demand +This should catch the most common kinds of problems +How about naming it ? I dont think the length of this name should matter +Indeed wrong backtraces are a problem +It is a windows batch scripts and no unit tests are required for this change +Also have a look at to see how it determines dependency graphs +So choice is between wrapper object or array +just a thought +The stack trace I was getting is identical. + HADOOPCOMMONROOTsbinjsvc What about bit platforms +Ive granted bera permission on the project wikiIs there any problem for you in providing translation for pt instead of ptbr +But the bug is still existingin my grails application +So then what is the correct solution to fix thisIt sounds like this is a bigger issue and this is just a symptom +This was already committed to trunk +rebasing on current patch in sync with current trunk +Do you have an example of this? In the projects we have we do have subprojects and it carries to them so Im not sure what your configuration is +it blows up otherwise +Would you like to take it over and work on? Thanks. +Theres no indication that the version is any different than the new one except by directly comparing the textI think the same is true of the spec web pagesAdding a minor version or perhaps a date that the spec was generate to the generated web pages documents would help distinguish which copies of the spec are newer +To clarify I did find some shortName related code but they are all for backward compatibility +JacopoI tried to find the screen you referred to but couldnt +Running an overnight load test if its still going in the morning Id say were good +Tests ran cleanly for me on +Thanks! +on v +Now that has been completed and has largely replaced the old Infinispan connector these changes can be made + History element cant be parser was not parsing the xmpMMHistory property in the previous version +Joe do you want to take this one as its very closely related to the other issues you took a look at. +Also what about Siteminder +Yeah I was wondering that was going on there +It seems that there exist something between this issue and +Changes look good +you guys find any workarounds +Browser is Firefox on Windows XP or Fedora +I only included h because I had and applied +Sheesh! +proto files refer to itbq +I guess that if you are using IBMs JRE then this is because your target environment is an IBM product +Thats done theAndy +Sorry those issues are related but they are not blocking each other. +I see that we were using ant the last time these were checked and there was a way to run themI dont agree +Michael please review and consider applying to branches +Moved earliest to class +did not update the fixed it in r and a snapshot was deployed. +fixing stale resolved issues +These replication tests fail even without this patch applied so I think this is good to go +Actually I had better revise the patch to reflect the addition of the new file as well as the recent changes to the filesvn stat srcrefA srcrefM srcrefsvn stat srcadminguideM srcadminguideM srcadminguideIt would be helpful to get feedback from John Embretsen andor Andrew +These could be treated in the same way as binary expressions +I started looking at how this could be implemented and I think the problem is going to be +What is the impact of this bug? Am I missing something or is it just that if you add a XML file with a BOM in the project the build will failIf that is the case maybe it is an acceptable workaround to fix such files by hand with the way that presents until we move to Java +Thanks Hrishikesh! +weird +The problem was in the followingCheck for facelets page was removed in but additional check for facelets page hadnt been added +especially since feature types typically get copied around a lot in feature access and rendering we put this into a specific factoryfeature type subclass that is engaged only by the datastore? Not sure if that works for would also be nice to avoid copying the attribute type defs fromANYTYPETYPE ANYSIMPLETYPETYPE STRING just use them directly +If Im wrong +It was released with the fix in +Ive marked as a blocker +I like the suggestion of passing Delete object to the endpoint +I have the same problem +Within JSDL extensions some of the elements are really useful for the HPC production esp +Tested manually +Its possible the lists havent been updated +Theres also a simple integration test in test that verifies that the repository gets correctly started and exposed as a service in those circumstancesSome things still need to be worked on +All tests passed +Will take a look at the wsdl after the release +as the validation takes place an error message is added but the user may not see this local message +Thanks MarkusI did not get a chance to look at it this week but will hopefully do later +yes I can replicate the issue +Garvin +In this case we want q to be by default +Micky +Should be fixed with we should try to remove to see if it works. +It is not necessary to specify the service name as WSDL requires ports to be unique per WSDL definitionMarking this resolved please reopen if you run into any problem +Not for commit here. +A schema is defined when is used but not when using enunciate with no configurationIm suspecting that youve got some configuration to merge your own file in which includes the schema definitionThe problem with adding a schema version to the default file is that it has to predict what version of the servlet spec users want to conform to +Fixed in revision using a file instead of +StefanoWhat is the status of this issue? Can it be closed or does it require more workOleg +If this is incorrect please let me knowThanks issue now release note status has been set +It contains the release builtagainst the codebase +There wont be a release +Adds a failing test for a parameter starting with white space I hope its patch is from the latest +thanks! +Attachment has been added with description sorry wrong bugCan this be closed? As I remember Jay implemented support for PDF destinations a while ago +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I can confirm that it was only section of this document which was edited. +Failed test is actually testing Now invalid transitions +Setting the system property is a global action and using kssl is a deprecated option that was added for compatibility +Patch built on top of branch +Discovered when using an administrator spent some time looking at this today +I have the same problem as previously mentioned how to combine lists of classes in different projects into a single list to be passed to a hibernate configuration +I created now my own test cases which could reproduce the initially reported issues as well as those reported by Karl on Aug +CXF reuses JAXB RI. +New snaps are pushed out. +Maybe we could do something like a register callback with the class +The that is currently being thrown only comes up when we have found their codec and the app has invoked encodeAs or decode but no corresponding method exists in the codec +If there are no objections ill upgrade the ivy deps to Hadoop patch +This is probably related to the fact that your JAVAHOME PATH contains space +Installed Alpha and wiped my +Splitting up gwt client and server side will bury me in merge conflicts +changed name per Trygves comment to +Bulk move for to trunk version +AlanThis seems much cleaner way to set up native Hadoop jobs than the command line interfaces people use today +You going to commit boss +In my tests I was able to successfully set both the prettyxml and json writers as the default +It has been fixed in release +Maybe we cant wait for that +Wolf please choose issue summaries more carefully +Also revised some of the pages like the usage and the examples docs +What do you see when you apply the patch forWhat create column family statement should I give the cli to create this user CF +I will make the changes and test it out a little bit then commit them for +Illustration of table w multiple regions that all have same start key but different end keys +I have attached another patch which addresses the comments +Attached a patch and file for the updates to the system procedure filesThe changes areFilename Titlerrefbackupdbproc rrefbackupdbnowaitproc rrefbackupdbenablelogproc rrefbackupdbenablelognowaitproc +Currently is an abstract class +This patch includes improvement and fixes as follows +Thanks for the patch mighty Alex. +The bug is that the retry handler property should be extracted within setConfiguration and used within wrap +To be clear I know the problem that causes it to fail the complaint is that it claimed you know how to fix itant is a convenience for developers only Im not going to lose much sleep over it not knowing how to handle an out of date thrift compiler +please test closeThanks and closed. +If it is still reproduceable for you then please let us know +Okay with the changes above building with Java works all the way through parent core components quickstarts console but not release +Yes does not physically exist in Hadoop jarPath tmpstagingdaijy +the only thing is you will get an IO error rather than a correct error message saying you dont have permissions +Dont log a huge stacktrace in and in when the cache is shutting down +failed because of the patch again so that it could be picked up by use extension +jar is the default +Since user can only select one profile Messaging and current JMS Server will be mutually there a need to have a standalone installer for Messaging or is it ok to just add a Standalone Messaging profile to the JEMS installerWhich version of Messaging +This is a duplicate of +Issue needs reevaluation +otherwise Ill have to go with one would be really great +This is fixed on the trunk and will be available for the release. +I tried to join two entities using on non promary key column it did not give me any error but when I saw the code generated it used the primary key to join +Do you mean its a deeper wider problem? Then please reopen and explain more thanks +could be faster then this version of for a relatively small number of subscorersI did not include these in the patch because depends onolder versions of with local namesRegardsPaul was indeed not intentional thanks for adding it backTo my surprise the svn update applied cleanly on here and these tests still pass hereant Dtestcase testant Dtestcase test +Here the can be another service within the registry +Hope this one works +Hi Catalindid you fix this in SVN? if yes can you close the bugregards MartinSorry its not yet fixedRegardsCatalin +Any solution +I suggest we dont waste time fixing this Hibernate and cpo are both licence incompatible and thus need to be removed before graduation +Dynamite Damien +Closed +trivial patch to fix issue +Thank you for catching this error Stan +I am quite sure that we could use mostly the same code for servlet jetty and osgiChristian +bulk defer of issues to +Im not seeing a problem +We are nowadays using Jboss and +Looks like any our XML editor isnt activated when opened by on a file. +I dont know if there is a way to avoid installing to the local repo +The usage of the current TM also breaks the other tests +Moved to since it isnt critical for the release + +menu file will also affect the gnome menu when people decide to log into gnome instead of kdeSince this error occurs only in I assume that its fault of not properly adhering to the conventions and this is not a bug that we can fix at all without affecting Gnome users +Fixed committed revision Redudced overcomplicated layering +I hope to apply it this evening +If I had to guess the upgrade maybe makes the local storage pool reregister where it looks like a duplicate +The problem is that the new bloomfilterfpchance in avro interface definition does not have proper defaultI attached the patch to fix it +you confirmed this is fixed? If so mark it as a duplicate please +Im not sure if this is the right approach +I think Mark has finished that already +I tried doubling the delay period and found that tests Test through Test then is a file called qa that contains the following entry timeoutFactor is a multiplier by which delays are multiplied set to a larger value on slow servers COMMANDLINEChanging the multiplier to fixed it for my windows machine +Thank you Amir! +Lets wait to hear from him on the same +bq +added some documentation around terminate and made terminate explicitly +Find out why +I committed the patch to trunk +Verified in theBuild id date +Jos I just added you to the contributor group +Ben assigning this to you +SeeAn option c would check in every so often composing a response and if over a general heap percent threshold just return to the client +updated in prod. +patch for ver this adds a forced option before the list so if you select any in the list it will fire the change +I added code to safe guard against the potential of nulls being used thus avoiding any potentialThanks in trunk +Added by xinclude directive +Thanks David +might be a APIABI replacement for so it is worth a trybtw if you can use the debianubuntu packages since they will be compiled to work on ppc +As discussed in ML Im binding it to javajbossdatasources since its not yet in CR and awaiting new updated patch. +Please investigate the latest failures and close if they are unrelated +There is a patch that is awaiting approval +Committed the patch with revision +We could do both in one go +page is not a special property +duplicate closed +fix committed Ying thanks for your input +However sometimes it doesnt show certain messagesSo everything thats shown is correct +Please keep an eye on the new few builds to ensure this is working properly. +Test close for release +Definitely need more information +java file +Please confirm it has been applied as expected +should go live this ve added Mac specific instructions to the install guide +I factored that code out into over two weeks ago +I can confirm that tests now pass on Ubuntu. +Ive committed this +git you think this issue is valid I am willing to spend time on writing a unit test which mocks the classloader to reproduce the issue standalone +The configuration mechanism needs to be pluggable +If that becomes too burdensome we can maybe utilize Hudson +I think this is ready for commit +fixed some javac warnings +I guess clustered mode is why the code needs to reschedule and resume jobs right? Otherwise my immediate reaction would be to always throw an exception if the trigger already exists +Also I agree with you about the implicit contract +My sincere apologies for being too hasty +Linking in to provide support for partial rendering with Tiles +Committed revision +I got quite a little further with my patch but now Im asking myself if we still need the or if we should add the remaining functions to now with the disadvatage of doubling that code in the deprecated review or comment welcome +. +It does look like the previous patch was applied as it is using +Reopening to deal with the following observations In AS JMS connections are pooled +Ah you are right +changed are merged in now +Last call for objections here I intend to commit this to both trunk and the branch barring comment +Attaching new patch incorporating review commentsThis patch is on top of patch for +Still have tests failing +If you decide to do it in the ipc package then all the retry loops in the might need to go away +by adding an arbitrary sample clause on the dimension table the join result may be empty + +Why is EL expression something that becomes a? I do not understand why that is done +These features have been incorporated into the version. +Something to consider for you guys +Negative boosts to terms are generally undesirable this is really esoteric and trappy +Hi Marcothank you very much for this tool +ahh +Agent changes implemented +Ive updated the build files to include the files in the build jars of the following modulesintegration jaxbrijaxwsjasonmetadatamexmtomssajHowever for strange reasons doesnt include those files but the includes them +Please resolve this when a console release when this fix in it is merged +Thanks Matteo +Hi Ive added a note to the Web page linking to this defect and suggesting that users install Derby which is available from the Derby site +Bilginthank you for reviewing the patchIn order to fix it I found that the should be updated with the attached patch +If you feel the fix wont break anything else Id say go for it +Hi RodolfoHow can I configure Postgre to reproduce the issue? Is there a setting to activate when creating the databaseThank was able to reproduce this error on a recent sonar are the specificsPostgres OS X created the database using the following script Start script drop database sonardrop user sonarcreate user sonar with password sonarcreate database sonar with ownersonar encodingUTFconnect sonarcreate schema sonar authorization sonar End Script I made the appropriate adjustsments to the file in the classes directory and commented out the derby database configuration properties and uncommented the postgres configuration properties +Have you used the stuff in the sandbox area and extendedimproved them? Can you create a patch and upload it? +Committed at revision to branch based on the patch from AshishThanks Ashish! +We need to clean it up for a patchTo make this clear these are not algorithmic changes there are code tuning work performed on the same algorithmThe naming is used to be consistent with the current Lucene class namesFeel free to do more code tuning if you feel it would improve performance further +The quickstart is broken +Make the summary useful by not being a generic statement +Others can be made configurable by the users +Erlend was able to workaround that problem by making sure all the result sets were closed +Ok +It is not obvious to me how to support this for fourth moments at all +In general I like this approach however the patch does not change all the pieces necessary for this to work and i simply do not have the time right now to flush out all the details myself +Also is consistently failing were the raid unit tests passing before this was committed +I added a new benchmark test which sizes the writes randomly on which the hybrid version is awful in certain cases since it spends most of its time in crccombine +This is a duplicate of. +Please test +The one place in cxfbean that is problematic is +is meant to indicate which means the channel has been closed by the peer +not in embedded mode not in domain mode +This should be committed straight away and closed or else left open for +patch applied. +updated patch removes unintentional white space changesThanks for the review Andrzej +Id be inclined to put that in sys +Running it through Hudson +Thanks Zhijie! +final issue resolved in rPlease take a look and see if this is what you were is it also possible to specify the versions in a file or can they only be passed as command line parameters? +And we do check for event sequence in these tests +Closing all tickets +I will commit this shortly +To avoid dependencies I merged the changesA version of this patch without the beta distribution is also attached to this issue +All these services are now configured by specifying the location of their XML property file via + to the patch +The web console shows an increase in the enqueue count +could be that we forget to lock the resources during the add operation +It would also be nice to watch these values as a performance graph +I just checked this out and as Joe said the issue still exists in +It is too much of a burden for them to examine each parameter and decide whether it is applicable for them or not +action and turn off toolbar in editor +Thanks. +Patch with ws changed removed fixed formatting variable applied your patch +Can a committer who uses Avro C on Windows please review this patch? Thanks +After setting up includelibs references I was able to apply the patch compile and run all the perl tests successfully +Freddy This is again on LCOM I guess some decision must be made. +Instead the describe parameter is optional if no alias is passed it describes the last created aliasI didnt realize grunt had autocompletion +But there are alot of hidden costs one need to be aware of unfortunately +So in case you want to use ip as the granularities instead of an app then you can create an authorization system with id ipaddressschemeipschemewhich classifies the ipaddesses into one of the subnet ids + since each commit needs to have been reviewed +it could load config files from outside the war to a specified location +Same for the findAll executeQueryv update on this? +Koen can you look at this +bulk close for +HelloWould you please try my patch? ThanksBest regards Andrew Patch committed in revision +attaching a patch +thanks Vikram +Mark I thought the logj error was happening when executing on Specjs you please specify wherewhen this error is happeningI have seen this once but I dont recall where this is happening +rpm which adds it to the classpath +Ive asked Andre to see if he has any insights into this weird EMF issue +Im close with this working on some final quirks and tests +Sorry wished i could i got the nn log but didnt get the edit log +This could work as a separate custom editor. +Moved to +Adding tests +the following jar command does not create dir entries for a and abjar cvf abI also confirm that Equinox is able to find those missing dir entries +Will remedy that another day. +More work will be done for next release. +Layer name text the colors after general outcry +Leo patch applied at revision r with minor changes thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +The last item from column is then moved to column to honour the widow limit +Patch v adds some comment for the lock +Closing as not a problem until proven otherwise +Does that sound ok +I think it is the processing of this message that causes the SIGSEGV +containing a fix for this has been released and integrated into AS upstream +Im not sure if I set the Fix Versions right since theres no available to choose from +has two main functionality not avialble in +I do not have an account at +Problem does not appear to be specific to Maven +Also if base encoding is requested for the value the key is also base encoded as per the REST specification +All streams are closed +copy over attached configuration file to fuse MB version conf directoryNote I am using and I set maximum journal file length to be MB only so it will create a lot of db log + mistakeDONE in +Here is the latest version that fixes issues introduced with the previous fix. +Fixed added unittests so this hopefully wont break anymore going forward. +would there be jobs per pool which would be ordered according to FIFO or the time theyve waited to get their fairshare +Btw the two main concerns back then were Avoid additional dependency for Validator Avoid potential version problem in case user uses hisher own version of one more reason to get rid of it as part of the prioritization of Validator the shading plugin cannot be used which leads to different versions between community version of Validator and the version included in the end we ended up including a shortened version of and removing the dependency to jtype +Itd avoid we break important functionality in the future again +If the patch needs to be changed which I am going to assume from your trials it does I will provide a new version +This is still a valid issue in the latest the demos and these worked fine +I removed extraneous patches +Passes unit tests here and Im committing. + unscheduled issues to the next release +Specifically the sql statement that gives you a missing expression error would be nice +Take for example the wordcount not waiting for all mappers would cause reducer to emit only partial aggregations +Before when pdfbox was parsing through the file if it encountered duplicate objects the new object would completely replace the old one +These are good changes regardless as they permit arbitrary offline development. +smile +Changes are committed at revision For application which are deployed outside eclipse since we dont have a solution to build the relationship of an application on server and a project in eclipse so far GEP doesnt refresh to add these applications +I have tested this on EC +New Patch +re japanese translation clarify with jvenable should disable other languages +We just upgraded to try again with it? What are the messages to look for +Closing this for now but opening up a followup bug for +Youre right but since this has been closed and has shipped already opened to fix letting row markers pass validation. +Can you submit runnable test case that fails? What version of Hibernate Envers are you using +tests now cover the spirit of the original tests +HiI just want to bump this issue to ask whether its still relevant as were already on Felix now +Can you please describe the scenario and the working environment +i had several problems with accented characters +Also this is generated with git diff p so it should be Hadoop QA friendly +added the new divider code to the encode as well I guess we are done confirm the problem is solved in svn +This is done because is run as a separate process and we can send only text pieces from it +If you are satisfied then please close this bug +Probably made the same with archiva and in Maven Enterprise +good catch on the dropped method thought I had kept everythingalso updated the docs per the comment +Maybe we should simply put the name of the projectjar it comes from into the nameOr in a label next to the name +i am a little bit surprised that you commit a patch before having your commentsquestions sorted outi would opt for reverting that until we reach some sort of consensus about this issue +Accepted for final does advertisement based on the single resource that owns the advertisement capability +The code looks good +tld file that passes the resin validation +I can prepend the FACC code to make the feature ID unique for that library +This looks good to me Ill review it tomorrow unless someone else gets to it first +So this is indeed a JDBC driver or DB bug as you suggested +Includes updates to API classes and impl class +The WSS errata spec removes a xv cert from the list of supported types. +When creating a rep connection edit the values in the connection wizard dialog to correct the text +Nicholas if you look at the trunk patches you will see the sameI see +But we would be still interested to revies your changes Sergey did you get a chance to continue +committed to trunk and +I dont know the source of it it may be wrong optimization for example +Attaching the unit test +Failing test is now unrelated to patch +From what I can see Chrome is still in development and could still have bugs that needs fixing +does not write ids for implicit elements anymore +We may have to revisit this in the future and the attributes should not be considered permament +Not really when I was doing my prototype of moving everything into the class I had set the onload to be true +Thanks for considering it as long as somebodys aware of the problem thatsall I was trying to accomplish anyway +So I will test it again once the commit is made and we get a new build +I will investigate this one + in CVS thanks for the patch + for the patch +Im told that there is no solution for this +This setting of Locale in DVF will be done by after it has finished booting DVF +Need to determine if this is busted in all tables or if this particular table is an oddity +Keep cool +Its even in my project with the aforementionned assembly configuration +Delete the rows in application CF after hinted data is sent +We need to optimize the performance of readingwriting the history data against HDFS later +Does that make it clear to the users? Probably notThe only commonality among those methods is that they access a single row +Ill be working on this shortly starting for installing on my machineRegards MartinThanks a lot for updating the bug and providing the patches +But we cant get info stats by command links dump httplocalhoststat +I can still do both +In our situation the write of the rete tree state enters in a infinite loop whenever An exists node is reached which is blocked and has a childIn this situation an infinite loop occurs in method when persisting the exists nodeThe changes already applied to class in the current development version fixed the problem for our situation. +Looks like I can do it +See the attached tar file +By above in my comment I was referring to the possible expectation from consumers +I think the warning got exposed by the recent changes in +I am able to run jobs now after hrs hrs +ok now you should be able to delete proxied cargo artifacts +It was caused by Rhino JS engine which HTML Unit drivers uses solution is to support whole selectors instead of just Sizzle locators. +I think that it was a mistake to deprecate the old API +I think this issue got fixed by Marcel in revision along with other changes +MAVENHOME is maven related only +Hi Marccould this issue be closed because of a field collapsing which is now in trunk and more matureWhy it cannot be integrated as a plugin? +With applied this patch fixes the problem +native bbox were being updated on the only if the have changed. +I handle that +Closing old issues +event to erase this nothinglabel immediatelyThis makes the nothingabel attribute not useable anymore for our customerThank you for fixing this +Some comments +Hi TomaszI tried to reproduce this problem with CXF +I need to understand this betterThe JAXR API mandates the JAXR client to have complete knowledge of the data +Reading the comment Im not sure I know what the code does +Hyunsik Thank you for your commentI patch based on the lastest tajo sourcePlease check this patch +Right it wont work for that +Proposed patch +Should we add an error code for every initialization of exception? For all derivatives of and? This could be large number + +I have a patch to fix mentioned patch +Can you tell me more about the RMI problem? Ive got some sample code running without Spring that Im trying to use to nail this problem +Anyway we could provide this code for the community +No I mean this issue is caused by the fix of +second patch applied as of rthanks for the patch +Jackson has been released and the is now on a newer Jersey as well +Stream Message is now working in the trunk code and will be part of the releaseCan you please tell me the time line of release? When shall it happen? We can organize ourselves based on that releaseThanksSahay +In way of introduction this sections provides a brief description of Derby +This would result in an unnecessary file in the failed dir +Oh my apologies +Fixed fix version +I think all the compileruntime dependencies have been remove now there are still a few test scope dependencies +bq +As the issue already says because of the get and set methods of it make sense to have for literals +Please ask on the users list if youre still having trouble with this and we can reopen if its actually a bug. +Some bug fixes in coprocessors were very helpful thanksNo need to port this to from our side +I am not really sure if jsdt is must be in dependencies but it at removes so many exceptions out of test execution same about and in ll verify if tests are running fine without adding new deps with just modified files +IT t work for differential measures when period is lastversion because events are not exported in dry by exporting project events in dry run +Hi Shelly this one is ready to be scheduled and will be resolved in the first EAP release that contains both and +The has been assuming a path of to be to not return the root node +Id propose we take this opportunity to further improve the by simplifying its design refactoring into standalone classes using the more extensively rather than queues on every heartbeat etc +trunk +closing invalid then. +all tests passed +Attaching rd rev of the functional spec for this feature +Unnasigned as I am not currently looking at this +Committed to TRUNK +Theres no easy using the same port with the chosen technique +Have your tried it in your deployment +Sometimes it isnt +they are aware of it +this patch updates the javadoc for zk constructiontalks about async naturetalks about thread safety +namely that the getAttribute call is just delegated to to the anywayWith that in place though I think MINA is really going to work for usThanks again dont mind taking this one on if its not being done alreadyLooks fairly simple heres roughly what I think we could do Change the interface such that we bind a Service rather than an address In we add the Service to the When we create a new Session in processSessions we set the Service attachment to the sessionIll need to look at the other implementations as welShould be able to do this today or early tomorrowWDYT +Committed a trivial solution to this issue to lock the parent token parked in a fork node +Patch v adds license header to along with test categoryJavadoc for a few new methods is missing +But the change is very simple it must to be working +Thats a tricky case because since the code bails out at the first offense you dont know if any of the other families would also cause a permission error +Refactoring is necessary though +Thank you very much AndrewYour instructions worked well in that test project and Im using them now for a production regards. +By the way I wonder that why we need to release the new version currently for GeronimoThanks +The multiJoin test passes though +java in your appthe default prefix becomes prop for themIm not sure how works but it may be that it doesnt like not having a getter for the exceptions propertyOn the other hand ognl doesnt complain about this +I would suggest you review your own habits +Heres a patch for this issueIt must be applied from apacheds is for MINA right? Is there a release scheduled for +It solves the tab problem but now the file has no newlines just niosql +This should be okay right +does Hadoop QA not understand files made by GIT? This applied cleanly to trunk as of rev +Did the solution found? Or may be it is possible to raise the priority +The failing tests are not related to this patch +Thanks Patrick! +Aswath When we create sales order Order is created in Approved status +Fixed with a test case +LGTM +Attaching the functional spec as +Lets wait until is fixed +Still need to port it to trunk when it stabilizes +Deprecating existing interfaces while providing the new interface gives them warning that they will need to change their code +Im guessing another hours +This patch is to enhance single puts to use mutate directly instead of multiputFor increment I think it is ok not to use it +So if everybody tells you that Hibernate has an important design flaw in the Criteria API you still keep rejecting patchesI wont keep cloning the issue Im not a spammer +Duplicate of +The advantage of the second approach is that less data will be sent over the network +The +I supposed the default setup should be capable of running atleast the examples and perhaps demonstrate the workflow +It looks like almost none of the work done under this jira has anything to do with the title or original description +This probably will cause correctness of the resultI think your application should be tolerant to this happening and be written assuming that mapsreduces could fail or get killed etc +Fixed with Eclipse is now supported Exception handling is added for the cases when required bundles are not with on Snow Leopard OS. +Reassigning to myself per Gregorys request +bulk defer to defer to +I suggest we dont implement it right now but decide when the feature is actually requested by someone. +HiHere are the fileregards +Going to mark as not you can in the attached screen capture this problem is easy to reproduce and very annoying as we expect to have many security conditions in the final design +Leaving open until it can be into the branch. +Do we need to change the interface or just leave the option to? Delete and the methods it defines +At this point I think this is a real bug in grails +Not a blocker for +The Sonar Server doesnt run if I load the in the extentionrulespmd folder my Sonar Server shows a Http ErrorIs the XML File above wrong +Obviously not normal +Ive expressed my opinion before that you dont need to make any changes in the script +We have some old tests which use Selenium that could be extended to test this but they are currently deactivated +Running with the patch now +So testing has to be done without his help +The runnable task queue on each TT sounds good from the point of view of globally optimal scheduling +I just committed this. +But for deletion the cost of deleting any replica is the sameI meant when replication is set to or higher are we maximizing the number of unique racks for the rest of the blocksTalking with Hairong now I understand that we dont do this in allocationreplication to simplify the codeThanks +This would go a long way toward better usability for ops staffBig +Better have a look at code. +I can confirm I just ran this against a hbase head installation without errors +Henry you want to make this patch available not in progress +Changed from Resolved to Closed per above comment. +Doing it in ruby only helps with shell but doing it in Java means we can also use it in the etc +I see them now +but I think reading the pom if it is there is could add a property to +we had a similar dicussion when the TCK uncovered this +K requests would be redirected back to HUB again which are messages after K NIn order to guarantee order we should ask HUB server to redirect all pending requests when releasing the ownership of a topic +Its also easy for the we would have to maintain a reference from the process definition to its image file +to commit +The reason we are at this state is because the original designers of the catalog used namespace as a grouping mechanism for data which does not really work all that well +This will be an overall cleanup of links before M release +I still get errors when I apply this patch +Ill commit the fix and close this bug +If oldest table is older then user configured number of hours then run compactionItll kind of work but its a hack imho that has a number of downside in practice +for the patch with Jenkins +I did preliminary scan through SPECRI issues and didnt find such issue +you attach stacktrace when continnuum cant start with a bad cron expression +The current API is sufficient +It looks like it would be possible to special case the IG handling in core and since its for backwards compatibility with existing Gadgets I agree that might be the best place to handle it +over an old version of itself it will fail +Once you compile the tool with Axis Java final jars on your class path the problem will be solved +are you sure you have the latest codeI was just able to upload a mb site without fixes done on scp were predominantly in the executeCommand method +in some cases +I do not have permission to update Avro at +Committed revision +Making before I apply your patch does not help and have the same result +Closing please reopen if more work required. +The failed contrib test is in Chukwa and is the same as +No in depth review but just took a few minutes to skim over the patch from a high level looking goodThanks for taking a lookbq +my would convert fine into json or xml but if the client sets the accept header to image what is going to do with this string? Rather than generate some unpredictable exception wouldnt some sort of content type unsupported response be better? In fact that check could come first allowing the method invocation to be skipped entirely +I just rechecked and there are tests that checks the featurein the archetype creation and the project you please attach a test case to reproduce +Configure RM to use +culpa on closing this issue too earlyThe patch and test case provided do indeed demonstrate and fix the problem +Or how to define its own Skin for PDF +thanks a lot the images directly from the we used to have in svn in rev +That step is not always done while Page Context is lazilly as is already released months. +committed to trunk +This patch shows the nd level parsers and their mimetypes instead of listing all the mimetypes directly below the +So we reverted back the factor to and the routers started successfully +File demo patchThis is just a demo of the patch Im working on for the Exchange Listener +Thanks Zhijie +i forgot that contrib and are not run when i do ant test +upstream not a dont believe this is an upstream issue our servers need to expose monitoring ports setup to get it to work. +against +Please take a look at +A tablet server may need to provide the capability to read with key X and write with key Y +Im assuming that the external is in the tree of the main site and points to the location of the sub site so I would assume developers could directly check it out from there without needing to go through the external +This patch solves the issue in running total and ending balance calculationThanks to SumitSurya applied to r +Thanks Todd +Will this parser be able to work for the C version as well? Currently C is using bisonflex but we should leverage the same parser for both if possible +Bean lookup error when executing repo scan fixed in trunk r. +But context root is not +If the creation fails check if the file already exists + for the patch +If you sequence them the behavior is NOT reproduced +Attached a patch that batches fetch of map outputsThe maximum size that needs to be pulled from a tasktracker per connection is configurableThe number of maps requested is adapted +Hi Joseph I really dont know much about the Maven Endorsed Libraries configurator please leave a comment on for Fred to take a look at and Im sure hell get back to youAs for the client generation +If the issue isnt satisfied feel free to reopen this issue +have it mostly done +I would rather clean individual modules from time to time as a task +Hi IrinaThank you for the explanation + +Yikes +No worries at all +We could think of a that opens from time to time telling Upload information about my usage of the software to community +We the original issue +The MAVENUPLOAD request has been completed +Isnt that what you asked me to do? +I reviewed it tested itFrom my point of view all seems to be okSo if no objections I will commit it during the dayRegards the move to a class with public static fields I dont have any problem with thatRegarding the Levenshtein distance I think we can do even better before we resort to such generic methods bring all property names to lowercase remove any Example vs +This depends on which is targeted for +Mark to clarify are you requesting to addremove source models to a live possibly vdb and expect the effects of the addremove to be persistentThis is a thornier issue also if view layers are involved as we have to do more work to ensure view integrity +A test case will be attached shortly +covering thatIn any case Id be ok with either setting the content type manually to something with a charset or putting all parameters in a POST albeit the first one just working because everybody happened to implement it this wayTo be honest Im not too happy about ignoring unknown facet values because this will produce incorrect facet counts when something goes wrong +The problem is solved with snapshotThanks +Or maybe just take the simple route of using synchronized collections for the internals of the tx object? Not sure which one is best but certainly I would not apply a local fix for this +release has not fixed this problem +Attached updated tests from should work now the one test from fails when Castor XML is used I did hovewer work with Castor XML so I will report this as a seems to have done the trick +Either way Thibaut found a way to make it work. +Thanks Stack for fixing this. +Thanks all resolved tickets from or older +version of hadoop are you using? Is it having the patch of already +the controllers command will still return the list of online controllers +According to me the most common usecase would be to create a core add docs to that and then just merge it into the main core which is serving requests +This optimization is required for sequential algorithm like SGD +Were aiming for a the next few days. +We also see many fetch failures but with jetty tasks hang forever +We are running on Fedora Core kernel Hibernate CGLIB +I didnt look into the details but if you dont see the same behviour lets look at whats different between your environment and my this patch again with more appropriate attachment license Grant license to ASF for inclusion in ASF works +Looks good to me +My modifications includes some additional tests. +But hey if allows it +I approvethe fix and applied the patch +Can we swtch to the dev mailing list to carry on this? Would be easier for me +Shortly after Id starting writing adminclient I found out about Fresh which seemed to be a nice alternative +Hehe sometimes something looks complicated its great if there is a simple fix for such a case +Fixed autodetection will no longer fail if an has an invalid. +Try to make and END both be empty byte arrays and compare using reference equality +Created to track this. +the problem with the tranql derby connector is already fixed in cvsthe problem with the jmx service is already logged as the noise from tomcat has not yet been reported AFAIK +Already reviewed the fix +thanks +Can we consider bumping this to the next releaseAnd would it be possible to add a figure to the mapping line between two objects to indicate that theres a decoder associated? That might clutter up the mapping diagram quite a bit but it doesnt have to be only concern there is that for usability it should be EASY to get a hold of such an object and I find it difficult to select an individual line now +If you create a detailed write up on this aspect I am willing to spend some time on that +We also have this requirement delete some specified regions specillaly about the data in HDFS and merge those regions +JIRA cleanup +We now fully support JDK headless mode. +will commit soon. + on latest patch +If you go to the parent JIRA More Actions Create youll get a big form with all the normal JIRA fields +Agreed this is cleaner than what I had +Problem is fixed +Anders thank you for reporting the problem +man readline search for +One thing to add When using readFloat co we should make sure that we set the endianness explicitely in the ctor +Thanks Aaron! +Documents are batched up and routed to the correct shard +the jboss type should be selected as default IMO if not thats a separate bug +Cannot reproduce w newer izpack and Fedora xMay be fixed or may be an older OS problem +However I will correlate the events and the zxids and see if I can get that debug trace you suggested +Based on the description it sounds like this patch might still allow internal Exception types to make it out of the tier unwrapped +The exception should not be changed and just sent to +Thanks +Whats the bar? Releases need to be approved by the Hadoop PMC and more generally the PMC must monitor all activity in the project +Resolved +This issue is not talking about the old but the example config named dismax +HOD can print this info at say debug level or +Another link? It will be probably a change that some people will do +sortFields attribute still present for richdataTable verified at +Updated from svn patch and now guarding the prototype +Thanks for reporting this btw +It should be a simple matter of changing the juddi config to enable this as we do not circumvent the normal juddi is probably better to chase any production issues up through the SOA platform rather than the project as they will be the ones responsible for handling consistency across projects +and then your desired object instance or your desired object instances are created twice +As I say above we really should get this working for +Good job tracking it down +Possible blocks during some tests with my changes for backups are failing due to the keeping the jar file open of the shutdown database + violations the JDK uses conflicting styles +fix is ve pushed a fix for this in aad +This problem was fixed in XSLTCDTM and now propagated to the head branch +coolthanks ruwan im not sure about the windows + +Using the FAQ is a good idea +Hehe +Hello Davidwhats in the case if we use both JAXB and castor and what is with the problemTorsten +Added for the actual move to Spring core as a dependency for the SerializerDeserializer usage +That was the problem earlier Koji +Quick fix to interchange the order of constructor arguments of +Lets resolve this issue as not a bug +Adding may actually hurt the distributedqueue consumer experience you expect to keep working the queue in the face of zk connectiondisconnection without any special handling +Do you think that we can change mxjdisable to mxjenable so that if the user doesnt specify the property the mxj part is not touched? This way users who already use mxj but dont want mxj to be loaded in the embedded Kafka broker are not affected by default +Did I hear someone call that its time to for the release +Patched with modificationsThanks +I checked into trunk +I use +do u need to move +I suggest the main issue of providing a mechanism in to use should be logged as a new JIRA entry +I am notable proceed further +Works great and your change is superior to the one I providedThanks for the quick turnaround +Attaching the Db dumps before and after the upgrade to like all the indexes and Keys that are missing in the fresh install are present in usrsharecloudsetup from a installation +FYI has been committed I am just waiting to restart until no test is running +The samples pom probably does not configure a pluginManagement element for the it should + +Ok It should be in this way so the JSON dependency must not be removedBut I think that we will have to expose a new version of the library package from the Download section for the next release ofThe new version doesnt have to contain the jsr and jsr dependencies +Let me know if its still an issue +Give it a tryThanksMichael +also happens with Grail and Grail well as Java Tomcat m hitting this issue in version as well. +If you run under Tuscany you may see a different set of line numbers +Will commit this if those pass cleanly +ls la which xcodebuild root admin Oct VolumesDeveloperusrbin it was a PATH problem +Will not do this per Bobs comment above. +This really is fixed now +batch transition to closed remaining resolved bugs + +Thanks +the I see more customers finding the route as their preference for a number of reasons +rendering of relative of relative of. +fix version cant be because it shouldnt get into release notes +If not who is the best person to review this? I understand we do not do for WFK so that whole section should be removed +It also avoids problems with frequently called list comands +I tend to lean toward just letting them be for the moment +but yes seperate JIRA +war named directories. +I think I had thatORDER BY behavior in mind when I was working on the GROUP BY code in One difference is that I think ORDER BY can only occur at the top level of aquery not in a nested whereas GROUP BY can occur deeply nestedSo I dont think that ORDER BY result sets will get and in the same way that this GROUP BY query runsI will take a look at and let you know if I have any interesting observations +The struct wasnt quite correct according to some docs I found on the by +If there are other controls at the top that have keyboard shortcuts please add the shortcuts to those tooltips as well. +Yes and I currently using geoserver as map provider but i need to add layer to geoserver map from have no idea you should ask on the mailing list +Creating a copy of an immutable map deleting properties and then passing it into the composite map seems to be much easier for me than having all these edge cases within the composite value map. +Anyway I still see the bug +works for scp scpexe file and sftp +Its standard instance has +Patch provides missed makefiles and +Theres no of connectors +jar filesIf you are unable to deploy the sample apps and run them and believe there is an actual issue with the Struts code reopen this ticket +Have been using this patch for a while in and it should be considered for backport to +I shouldve said that in the first placeIve reverted this commit +Corrected typo in the tlddoc and dtddoc to the binary distribution by updating the assembly module assembly descriptor . +Sample VDB with two models defined +Patch is no longer maintained +used the latest formatter +this was resolved before +Issue raised in title is now resolved +In any case Auditor shouldnt watch for expired +distribution of the values is not uniform +Had attached an incorrect patch earlier +Closing the issue. +Well thats sad because I think would not have been a great deal to add this functionality and it would make much more feature rich so I have to maintain my I own fork but thats not a problem +Changes look reasonable to me I think we should get it committed before code rot sets inRajith have you had a chance to look at this yet +Small patch to fix the patch looks OK and I am OK with having no tests as that part of the proxy has no unit tests + +fix for version is broken too the provided patch does not address this to Denis A +I ran the tests and they worked for me +I reported this originally in the google group here is my infoOS Mac OS X Node Native BB SDK I did setup Cordova with a git clone of the branch and can build just fine +This approach has a few problems the first is that if the index is being updated by the shard servers then the segment files need to be protectedheld so that they are not deleted out from underneath the mapper +The plugin uses compile level resolution so dependencies should afaik be available for the plugin +smile +This sounds like a good idea +looks close for +Passes unit +I found my own furball +This is a good first step +how big are your rows and are you running on a VM or real hardware +What do you guys think about disabling block caching by default on count +That looks like a classpath issue +The first patch doesnt work the attached one does +I mean to make some class called HTML +Setting back to Closed Done after adding release note. +slave nodes logs have been provided with logs for failed tasks +The idea behind the split was to give users the option of either deploying our existing default settings or changing the configuration by adding one class modifying the and warring that with the existing JAR +Libby Im talking to Fil to get this tested on to see if we can use this demo on other platforms +hpp in my case which I just removedCan someone close this bugdaniel +But you should work on a patch for foundation. +Attaching a new version of the release note +No namespace support implemented +Thanks Anders for testing +Slightly smaller with rv thanks folks +Another place where Hadoops secondary sort is useful in Pig is to sort the index entries for Merge Join +Unfortunately Ive got this issue in for Apache Flex when im pasting some text mytextn +Done and upgrade solves regression. +Youre right this is exactly the way to implement it! In my first approach I just made the methods clusterable that were absolutely necessary for some sample application to run +Since NULL means unknown it is not identifiable +Attaching a patch with the following changes +Thank you +If it works I can backport to +Thanks Eric! +Task is in progress the for all components + +Verified this is not a problem. +Now targeting +Should be done shortly +Unless this gets exposed in the configurationapi it will be to cross platform compatibility +Manually tested! +p inf adds mem flags unless something bigger is set and also sets the discovery url +And if someone has a good idea about how to improve the situation with the OOM Im all ears +The actual code still chooses Stephens third answer of as it always has however it then attempts to make that more readable +It seems to be a known problem +Im fine with this + fixed thatAny sample project to show your issue +Thanks! +So maybe files one in seg one in log and one in the parent directoryIt would be good if a backuprestore either copied the file or generated it to get the same message to users looking at the backup and possibly corrupting itHow about for file namesAs to content seems like we should take the chance to be a little more verbose +The instance is used to do the parsing of the newly loaded mapped statements +This issue has been started +Therefore the handler is loaded only if the handler jar is in the root lib folderThanks the patch provided by Kasun with some modifications +Yes it worked in but is broken in. +Jonathan and I are with Max on this +Anyways we can keep the conversation going on this issue maybe its just a container configuration or classloading issue has been resolved for one month or more without complaints +as suggested i moved the to the common labels file +The push notifications came mins laterIm not sure whether there any more issuesFeel free to close this ticket if you dont see anything suspicious +Tomcat is ignoring the duplicates while is not +the patch looks good +attach my patch through hudson +The salts of course are included as part of the password string +we use that heavily for test setup and tear down in Hector +In both cases initializing milliseconds to zero should not be a problem +Unfortunately I can not provide a sample image at the moment as they do not belong to me +If yes should pull down icefaces from maven +Added the same flags as to on and Ive added documentation for this feature in enusersourceserviceswms if you want to review its in svn +false alarm! sorry my jetty was running with and when I start it with I get unsupported versionI need to test again. +Thanks! I tested with the released jar for and it works for both command line and maven compiler plugin. +Thanks again! +This was probably fixed before beta then. +I think this ones should be overwritten by correct libraries because there is not compiled alternative in repo +Reading the explanation of the problem makes my head spinLooking at the code + +It is lacking the validateModel method which causes compile errors in Eclipse and Maven builds +I should mention another advantage of this system over a pure approach having active daemons receiving the logs instead of just multiple NFS mounts or drives means that we can implement storage fencing without any external provisions like etc +Thanks Todd. +You have to set it in as followsEXECJAVA JAVAOPTS GERONIMOOPTS JAVAAGENTOPTS GERONIMOHOMElibendorsedJREHOMElibendorsed GERONIMOHOMElibextJREHOMElibext GERONIMOHOME GERONIMOTMPDIR javaagentGERONIMOHOMElib jar JARFILE LONGOPT CMDLINEARGSIf you are using your own keystore and truststore than you may want to specify +I think the only problem is that doesnt mark its victims as compacting so a minor compaction is free to attempt them as wellThere is a similar problem with +I reopen this issue to rework on it +Im hitting this as well +Hive currently runs slowly on a major drawback +Any help you can provide is appreciated +Removed autoloading from the summary since adding the named jars to the classpath doesnt actually load the driver +Feel free to suggest it would be nice to be consistent in naming and behaviour wherever possible I dont think we should get too hung up on it + now only blocks invoke operations that dont pass authorization check and let go other invocations +The messages are in the Messages View +Which is it you need? Both? Why? If you pass one big file and one little file and ask for six splits should it break each file into three or break the bigger file into four and the smaller in two? How should file size be measured number of records or number of bytes? There are myriad possibilities +It was a simple fix. +Nick do you have any further update for this Improvement? Thanks +Attached patch that changes the order in which the jcrpredecessors property is populated +Btw this is only one part of the needed test suite +Otherwise we need to update the docs to clearly state using the id is not valid +just syncing the patch up to trunkpart of the funkiness i dont like is +I suppose Im not a user with administrative privileges +Okay to merge +Ive fixed committed the immediately apparent null pointer exceptionI was able to reproduce it withhttplocalhostsolrselectindentonsortiddescqsolrwtjsonflscoreshardslocalhostsolrBut this was just a symptom of a larger bug +Empty configuration classes are now supported. +that was what i did first +I will apply his patch and refactor my changes to use his topic and partition classes for the data structure +Closed after release. +Gene can you please download the official Roo release and try with that? If the problem still occurs please attach the script which allows us to replicate the issue +The main point of is to allow the user to create shared modules that are available to all deployments so for example if you are deploying spring apps to the same server you could create a spring module and set it up as a global module and then it would be accessible to all deployed apps +Here have what I made +patch committed to branch and trunk as well. +Have you run any benchmarks +no but you would need to exclude it if it doesnt work +committed v and rmd the TODO line. +Not an attractive outlook Im m resolving this issue as Wont Fix because indeed it is very unlikely that we would introduce a change of this magnitude into the core framework at this point without a very strong are free to reopen this issue and add new comments and continue to add votes if there are arguments that have not yet been heard +Marking as fixed since and have been since and have been closed +I attached a draft patch in the all the dependent relation is kept for those car package and while generating the final import packages one export package for each dependent bundle is added to the meta data of the application with that it should make sure that all the dependent bundles are wired when the application is resolved suppose we should not the cache problems +So I hope that in the long term we can contribute to the project too +I think there may be some more sub issues here but as a start here is my first patch that covers and move issues to and +My preference would probably be just appending to a text file for easy graphing from Hudson +Attached the patch for review +Unfortunately not this comes from EAP and as the name is still the same we cannot block itYou guys will need to filter this out somehow +It is certainly possible to fix this but it would be better if this detail was considered before submiting +Added logic to remove an unnecessary sort from dependent joins that are behavioral the same as +Updated some relevant bits of javadoc with changes related to this JIRA. +soon +This has started to bite us in fairly consistently +Thanks for the patch +Yeah we should DRY ourselves and not duplicate the same data twice +ant test was run manually +Seems as if it would make sense to turn this into a documentation issue and mention in the databaseName attribute topic that the subsubprotocol name needs to be included +If you import a nonexisting method field or property Groovy will not fail at compile time but later when executing the compiled code +Remember the FTP component could have similar if you please try the latest snapshot to test if the fix works for you +Im testing on linux instead assuming that the OOM problem is related to that systems just committed this! +One second Ill put up the new patch +Is this expected +The attached patch is an implementation of the proposed recurring events +Added to trunk pending merge to X onto branch +Also if this is source code updates it should be in RF not RFPL +FolksWe cannot go indefinitely closing and this bug +Thanks! There are going to be several pieces of documentation to add like that with these various +Thanks for reminding +This is all about the cycle such as it isI confess that even with the hints on the user list I still havent managed to get solr running from eclipse with +Thanks Graham +Thanks Brian +Thanks thanks a lot Tim +issue assumed to be fixed by newest build +Best regardsGeorge +avsc +Attaching one line patch to scan only for +Furthuremore I have tested this manually too +Pei tmill wont be an issue just created that account +Moving to like the jgray idea that we use modtime instead of an edit number +from me +Another improvement related to Method and Constructor creation is the Fake methodconstructor +Fixed in svn revision thanksdims +I think this was fixed else where as the test passes +Nicholas Sorry forgot to use +framework re quick! Compliments +Commit the changes to trunk at thanks Hang Hong FangPersonally first step it is better to make myfaces could find those config files +High five +All the requests sent to the broker end up in a common request queue +Is the topology persistence a consequence of us serializing Datanode descriptors to the image? This was implemented a while ago and has probably never made senseWe should file a issue to not persist Datanode information in the image +Bulk close of resolved wont fix cant reproduce etc +Thanks Todd +Hopefully as soon as Tom sees this +Okay here is a maximally paranoid patch +Somewhere down the line IDNs enter the in ASCII so there is no problem there but these tools lack conversion +I added the fusesource repositoy to the +I will commitOK committed with these small changes Replaced conf conf in the comment in that file +Im having the same issue +I am working on it +patch is only definitely available on unix systems +We actually will fix these they have just been absorbed by other issues. +In the patch +Pretty sure I see a repositoryPlease confirm +An example project that reproduces the bug issue +Merged in thanks Lisa +Dart Can you take care of the bits Tom has brought up to simplify this UI a bit +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I dont think more of the same type are a big deal +Tested work of workspaces via +I didnt have a chance to look into this for the past release and I didnt want to rush things either +This fixes the test +Hi Romain! Did you make the update you mention +Im OK with useSchema as well though good documentation will be the key +The error I get in this case isfileDusersGeorgedocdocbookxml The markup declarations contained or pointed to by the document type declaration must be +Heres the proposed patches in r +I guess there isnt anything to discuss as this has already been merged +Resolved after replacing mozilla with nightly build or cr +I can provide an example if necessary +Anil please add to this JIRA all the info needed for signing including the configuration changes needed for customers to be able to use an AS with signed +Feel free to start over if you think autoconf is the wrong direction +Its doable of course but it impact on all subsystem using those threads operations +Marking this one as a dup since has more info. +Ack +Socould you please provide me some other way of specifying thepolicy to encrypt only single elementThanks AbhinavCan you please attach a sample request message? Also how do you specify the client side policyThanksThilina +Id vote it up if it hadnt already been closed. +Is this a regression +It performed pretty badly which is why we committed specialization instead +How complicated would it be to store the checksum file on a separate mount point? In JBOD configurations this will enable both reads to happen simultaneously +Thanks Doug +Verified. +as it turns out the current implementation was completely broken anyway and happily let you deploy ll commit my fix for it but leave this open to make it configurable +and it happens for me only for first search only +Alan can you please review this patch and merge if you think it is safe for +Patch Applied +Perhaps with an ant target to explicitly list out all currently on the classpath for debugging +Hi FilipI am a committer of Apache Synapse and am interested in being able to write to Zip Jar and Tar files through VFS +I have updated the wiki page with your sample +Attached test not crash +I added the patch in revision added the open type font support in revision added the improved handling of callsubr and callgsubr commands in revision based on Jurajs proposalEric and GuillaumeI had to adjust the preflight +verified on ESB Tools +I used your idea but I eliminated the redundant code +Patch applied in revisions and many thanks for your contribution! +There may be others +Line numbers in template classes are adjusted in errors still produce wrong line numbers +Thanks +cannt get on client sidethe error in tomcat log isWARNING Cant find the the request for httplocalhosthelloworldserviceshelloworlds ObserverApr AM invokeWARNING Cant find the the request for httplocalhosthelloworldserviceshelloworlds ObserverApr AM invokeWARNING Cant find the the request for httplocalhosthelloworldserviceshelloworlds ObserverApr AM invokeWARNING Cant find the the request for httplocalhosthelloworldserviceshelloworlds ObserverApr AM invokeWARNING Cant find the the request for httplocalhosthelloworldserviceshelloworlds ObserverApr AM invokeWARNING Cant find the the request for httplocalhosthelloworldserviceshelloworlds Observer +There is a version that I dont think gets run nightly that is even more of a load it runs users for minutesEven then just passing this test once did not really prove anything as the issues often are verysubtle and timing dependent +However it can happen that numbers are stored as strings for example +The classic JPAHibernate stuff +fixed with trunkr r thanks Viola for pointing out the root cause +Thanks Vinod for catching this. +Out of interest what is your use case that requires false +I am guessing that this is one of those magic that was removed from release +Fixed applied in SVN revision +Done +I integrated the fix on the branch as well. +Commit away +Added a simple example that causes the errorThere are routes both with there own input but both use the same output +Works for me +Attaching patch +Is this issue still active +Fixed on and closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Maybe provide some extra documentation for the methods that require this but its really unfortunate we arent allowed to talk directly to the tables in the first place +But Im only able to test little cause I work in a very restrictive environment +Fixed by. +This is the preferred approach going realize that some users may continue to be forced to use parent contexts but thats a reasonable gap given when weighed against making an already complex workaround even more resolution comments on for details on. +the original fix for windows broke the script on mac +The WAR packaging is built in a standard way so it most probably works on other plateforms +I am using the traditional Hibernate mapping xml files +cannot reproduce +Updated patch with changes +I havent looked at the patch but one of the biggest issues with Spans is the duality within the spans themselves +If your usage actually went above GB you would have had a problem +Ill run some experiments and see +It must not have built since since the loadstore func changes went in +Closing issues solved in released versions +Should I wait or complete componentsRahul +ill be happy to do it any tip on what you meant by polishing the code +Either way wouldnt this result in something different +could getChildren carry children? Flavio do you have any number about it? or maybe we need some experiments to ensure it works +Am working on fixing the MDB issues reported against AS will include this in that task. +Of course if we will implement this then would be possible to turn off as any other rule. +I also dont see the database connection leakage described by Keith but theres a chance my use case is different than his +This clarifies that post call is working and has no issues with data as well +Applying these changes in the branch stomped some changes in since they touch some of the same code +bq +Committed +It should then also be possible to override pluginManagement settings in the reporting section +Using the test target on the module fails if the src directory isnt there because it tries a compile sweep on the moudle wheras the build target is overidden and so does not +Yes a suggestion for alternative has been added in the deprecated javadoc +Modified implementation attachedMaven cleantestpackage goals are executed successfully +For our case it is referenced in +Any comment on this one? Otherwise Ill close it as wont fix or not a bug +Thanks Pablo for this new sonar squid lib use should fix the issue you will still need to define the correct project files character set using the CHARSET JVM param when launching the sonar plugin +It has to be able to behave as correctly aspossible even with broken input and in this point I thinkthe best way should be warn the user in the deploy OTOH just ignores the it doesnt changethem but they do not prevent it from working either +Yes Todd I agree with you abosulotelyto slove this problem finally we should sync all transactions on the ative NN to the SBN during shut down HDFS rightso can we open another JIRA for it +Close is fine Lars +Such comments are not supported by the parser which produces an error at startup and prevents from viewing the node dataaclsmetadata +Yes there is a workaround but there are two problems the issue can be a for newcomers to Grails everytime i run into this issue its hard to find the actual cause and I lose a lot of time before realizing Ive been hit by this bug quite bad for a high productivity framework such as its hard to fix I will leave this to then and hope you guys fix it soon enough +Is it ok to close them quietly without logging potentially errors when closing +Integrated to as wellThanks Chunhui +I have attempted a solution to this using custom providers +Doacan that is correct +Hi Konstantin it would be nice if you could review this one +Unassign this JIRA since the project was not accepted +Sample class +take incremental change on top of yours makes it possible to complete the build without Ivy or Internet access if you manually include logj in your classpath +But putting the property in the project itself should work +Its possible it was only a problem because I was running hadoop out of its project directory? If thats actually the case I would be OK with just making a mental note of how to fix the issue in case a user encounters it with some untested version of hadoop +Committed +Description of the solution in comment +Thanks Prasanth! This is looking good +Closing as invalid. +Serializable is missing on class thus wont go over the wire if you use Sort +The docs are wrong +If you want to change to control based on current disk space left only a small change is needed to the patch +for me +I think we covered this already in +Committed to trunk and +Just wanted to clarify +Ive committed this and merged to . +Patch for Amber documentation +Committed revision +Fixed in r +Having some talk with Charles it looks like there are lots CXF endpoint start at the same time which could mess up the CXF we need test case to reproduce this kind of error. +As its unlikely that well include a change like this into the timeline and does not contain any metadata anymore Im closing this as wont fix +This initial version puts the newly flushed memtables into a queue and when there are it size tiers them +Can anyone offer any incites into what may have changed in to cause this andor where I might look for the problem? Also any pointer as to where I could integrate my test case into s test suite? Are there similar rubyjava test cases somewhere +in child the point of the bug was to have the embedder andor meclipse handle error condition better +Forward to Jrgen for review and applying the patch. +You think you could post a sample project where i can reproduce the bug so I can try to fix it? That would help me a lot +Can you come up with a quick test case +See the Admin Guide SSL section for usage +bq +Thanks for Sandy Vinod reviewNew patch add the the comment and expose the whole logDirs to the patch fixed the comments patch exposed the container instead of the log base dirs +JacquesThanks for reverting this patch + comments +Thanks! +Its an relatively easy to implement obfuscation feature but unfortunately is aimed at serialization therefore it is better to leave your example as a nice starting point for those interested in field obfuscation and not to include in the serialization library. +Also added +ForeverLong Term ImpactThe severity of the problem is a NN was consuming more memory than the active NN and trashing in GC +resolved for for. +Fixed AFAIK. +Otherwise the old firstResult and maxResult limits will not be ll revisit s code to allow for of the same object as far as possible in particular with respect to firstResult and maxResult to the design limitations of Hibernates Criteria API I generally do not consider this as a bug of Springs +Admittedly these are not the highest priority but a consistent way of handling this kind of information in a struts project shouldnt necessarily be a roll your own +Ill try and look closer soon just did a quick scan tonight +Applied fix contributed by Brian to trunk and branch in SVN revs and . +Part of a bulk update all resolutions changed to done please review history to original resolution type +Sorry generated the original patch against an older version so it doesnt apply to the current branch +utf files and generate the properties files dynamicallyIn fact I have confirmed thats how it works +In other words the rows are sorted correctly based on what values are in the hinputText even though the values are not being displayed +Thanks for the patch! What about the other components though? Should we implement something like this for them as well +Sorry i just found out your attached test case +The little that it says is a footnote to a discussion of X core datatypes +Applied goodFound out that code style is wrongPlease commit the fix patch applied at +Passing by hadoopqa +Use it as value for XXXserviceclusterXXX caches +added a unit test and fixed a Findbugs warning +Not sure at this point what the issue would be with static vs shared libraries +Committed to trunk. +The attached patch makes the suggested changes +Assing to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +to naming these fields textexampleXXX +We shall have to come up with some alternative mechanism +On one of our demo machines yas the same issue +Remove the catch block to prevent invalid state of the object +Patch +Thanks +Is astyanax the most recent version? Try updating it maybe? +Clients store the last transactions hash check +Committed patch to resolve as closed. +Here is a patch for fixing the problem +Right now were using the tracing flag to determine the context menu so right now I use that to know if were tracing or not +Closed upon release of. +Policy is even worse likely the arbitrary switch to on demand is not very useful on the large queues it appears better to just orddem the whole thing +Would that be relatedI also noticed in the that the writes a stanza before finally closing it +Unassigning since we havent heard from the owner in some time + +Probably its related to network settings +This issue is fixed on the Phoenix trunk +Resolved +The unit test for this issue is also passing +Ive applied it in r +Patch to add documentation supporting integration +Marking as patch available +Unless we want to draw a line +I will provide screencast soon +For this to work you would need to unpack the tools jar before the WAR Plugin is run +Some fields was added to structure so both GDI and GDI versions works normally +It is still open for simple mapping but also not so easily addressable +JIRA is never suitable a forum for asking for assistanceThanks +we would really like to find someone willing to work on this and add some specs to rubyspec for it +Instead of tracking small fixes in the separate Ive prepared this patch which correlates with patch and will be committed to ydist +Seam HEAD code work properly +Changed the patch to that adds test that uses with data in hdfsI believe worked before this JIRA with any filesystem +Yes please commit itIve two independent personal roadmaps Needs a lot of thinking testing and maybe some cleanup to find the concurrency problems +Since the simply invokes the Service via using Exchange mechanism required policies on the bpm service would be provided appropriately +Your fix looks great and is very ingeneous +It will probably take me several days before I can test it again but just looking at the changes it looks like you solved it. +Yeah Thanks! days. +Updated patch +No thanks necessary +Earlier in testing I have noticed failures however I was unable once the above error is fixed. +this was possible because the DOOM is structurally similar to XercesSo + you should be able to control issues now +See my pull request +Its exactly the same problem for logs +It is OK +I found the zookeeper related problem filed and also included a patch +It wastes others +ok a simple workaround dont do gem update system then works for buildrwindows +Instead of relying on the container to figure it out +The problem stated in this issue disappeared after a mvn dependencyresolve +The attached graph shows the performance of a simple SQL query that selects nodes out of a set of k nodes using a simple property constraint +Thanks for the review + +I think the problem is that other products have dependencies on different versions of some of the libraries that XSLTC depends on +View CA Details Shows the details of the CAs certificate etc +The new page opens in the same device as expectedI have open and on a link on a web page and click Open in New Window +Just needs a bit of reformatting to adhere to s style guidelines +Youll be able to put this in a htaccess file +I think you need to update your customer processor to send the exchange to both file endpoint if you got the last exchange from the spliter +Pull request attached should be applied after target platform includes Kepler M. +So this is definitely a small patch fixes a few errors I had in the enhancer section +Please explain what this has to do with Struts +Is there any reason why we shouldnt use a standard container? Low level code when written wrong is much worse than safely using a well tested STL container + tested +years later I still havent seen this error +Justin opinions +do what you do +For Hadoop Common we actually only need to back port the test +This is taking some time as Im fixing a bunch of Canopy unit tests +Alexander Klimetschek that would be my first choice too! +Anyone has any clue on this one I tried putting the wsdl on tomcat to check if this issue is only because iam trying to create java from wsdl which is on my filesystem +Moved Widnows and OSX sub issues to own tasks +Committed revision +Please feel free to download the project file and import it into Flash Builder +Allen you can use mvn to do offline builds by passing o argument Steve Apart from maintaining and by having the pom file it reduces the build scripts maintenance +Committed as r +If the Host Setting isnt set to the VIP for Port and one of your Mgmt Server crashes the UI is still available but the system vms cannot contact Mgmt Server +But I dont find it as the best solution +However we always used a but since Java synchronized types are handled differently +All store files are already open +Uses to get the first day of week according to the timezone +Will elaborate on this theme in the next comments +It is issuing an explicit delete when closing the link regardless of whether the subscription queue is exclusive or not +What a weird thing to miss +I updated the test accordingly and will close this issue as a +Username schaefera +Must install the from not Reloaded +That should be pretty fast +Triggering Jenkins +I feel like Im missing some major fundamental point +Is it possible to deliver the content as? As far as I remember then no escaping of german umlauts takes placeCiao case with Cyrillic letters converted to entities +The problem went away on Windows when I upgraded to use a Java VM with compliance level +Documentation has been updated +Attaching the proposed revision +Could some please review this patch +That type is teiid and dqpThen it needs to display each of the managed component under this type +for the patch and for deploying other pages +bq +The zone has been renewed and should be ok for the next period +I think the default behavior of the shell should be the default behavior of the client which is version unless specified otherwise +fix the error message +Its in the branch so youll have to build it from thereThanksl +The entry is removed from meta but it may be assigned and its hdfs hfile stuff is still present which causes hbck to complain +Attached test cases +Raising priority as this is a BLOCKER issue +The tradeoff is to not allow reserved keywords in the package name since there are convenient workarounds available and error messages now point out any offending keyword misuse +New patch for delete records patch good to me +ok I reopen the issue +Nice fix Jinchao +When I tried referencing jar the build failed as jar aint present in maven repo +Hello allSince its probably night on your site Id like to submit the patch and let it run through Hadoop QA first to see whether any issue to resolve please feel free to let me know if any review points when youre back thanks +In why control the seed? Is this for repeatability of the testYep +UIMAHOMEbin and did not work in the case of +Update for verifying configuration plumbing +It appears that the patch got applied twice in my machine so the config file got doubled +attaching a newer release note that hopefully addresses the concern voiced in on this. +I think we should revert the changes commited temporally while we reach an agreement about how this should looks like +Committed to +for core and contrib tests are due to the current trunk test failures +monitoring seems to be used in in starting a new processor and register +depends heavily on the settings brightnesscontrast will put that screen back on the and try a few things +I wonder if there is still anything left to be done there +I have merged this to. +So the real problem is that the binding layer should be able to marshall arrays without jaxrpc mapping types I would take a look at the since it represents a non generated that we are required to support +And since are part of the markup load process the markup including the langen get cached for faster access to subsequent calls +Hacking OMYeah I agree on this +working on expandcollapse collapseexpand context menu actions +You may have to add the retry logic to your custom code the Tasklet you have implemented or the client that calls the +The idea is to extend pax url aether with a new remote repo config + +It works for me with look at this for cr need to verify if it works on Seam fixed it and cheked for Seam and +any objections +Opened up +Patch that puts back and that sends codes for methodnames across the wire makes PE test faster writing and reading over +Thanks Stepan +Rick I have never used Derby compression I wasnt aware of that GonzaloCan you share with us the DDL for this table? In particular the original CREATE TABLE statement and then the ALTER TABLE statements which added the extra columns? Thanks +Yeah not sure its much savings +Thanks +Before reviewing does this new patch do as previous versions did IIRC removing the memory cap for a region rather moving the memory cap down to the store such that if ten stores we could use ten times the memory we used to use if all stores are undergoing updates +Maybe we could add an in the same way as we provideAs far as we can tell this problem would affect anyone running a versionof antlr other thancheersRobin +understand the problem here +If I featuresinstall the only stuff now arexmlbeans CXF uses as is not in central +Robot seems to want to go on trunk +I will wait until the snapshot build from this is available for download and test againBTW I cant find any info on when will be released as a stable version is there any planned date? +eli thanks for the review +I thought best place for such look up will be in +I reorganized the docs and added a section on the homepage about the linking of source directoires +Id put it in Install Guide +Applied you patch in SVN trunkThanks a lot for the contribution! +Bug the same on it breaks in happens when you expandcollapseexpand see the bug fixed +Undefined can sometimes lead to more efficient implementations in certain circumstances +So you end up with something like thismvn Dtag projectA projectA releaseprepareThis allows multiple versions to be passed for a release +Hi ClausIs this a policy that will be registered against a camel context to provide guidance on howwhich routes may be startedstoppedBTW I recently put together a cool piece of code that does route mutations of producersconsumers on the fly based on JMX event notifications +Thanks for figuring it and making a patch Alexey I forgot to check the shell +Currently the version is hardcoded to be +Im not following the connection of this jira to +Moving to next Fix version to avoid bumping from one version to another on each release. +Attaching +This patch fixes the problem with loading generated code within generated code +bq + +Hi CatalinDo you have register in your script? Why you register it? Is in your current directory +EnvironmentFirefox Hat Enterprise Linux +Late in the afternoon I had frequent errors +I will at least look into it +good pointI was only looking at the impl code at the point I wrote this +Ill try the patch these days thanks +so link is only packaged into the devel RPM rather than the base RPM +The CPU looks fishy Nitin can you take a look why this is allowed +Happened with DB as well +Could you describe the changes? Something along the lines of Extracted the following source code from added license headers +I will open a new bug for this you dont have bug for underlining type references is here +I just commit this +DONE +I will have a closer look and fix the issue when I get the time +I patched the failing specs in bbcfe +which version of tobagothe underlying javascript library has changed in latest snapshot +Reinhard Poetz wrote Could you give me a hint how I can switch off usage by API? IMO this is the only safe way because if others build Rhino the chance is high that the removal will be forgottenWith Rhino from the following patch should disable the invokeroptimization +removed navigablemap! null check and changed some log messages +The problem is arises in the prcoess of importing node SAAJ to DOM +The test passes on trunk as well +Thanks Jerome. +failover implies an active and a standby +in in your error message you wrote documet instead of document +Found the difference +That said the code should easily be backportable in case we also want it in the legacy code base +Yes for some reason the code doing this was removed between and +I face some conflicts with your patch as I dont known danish could your check my manual mergeThanks for the translationsRegards Christhere were still conflicts withsrcmainresourcesorgapachetapestryinternaland srcmainresourcesorgapachetapestrycorelibcomponents when I compare the files they seem equivalent +Resolving as incompleteTry the latest omniauth gem and open a new ticket if the problem persists +Ok +I committed this to +Apologies for jumping the gun with the fix +Sorry meant lets ant clean test pass with Oracle Java on Linux +Since theres not more to do on this issue Im closing it. +Shortly I need a list of icons them sizes colors and types +Is there a unit test you recommend that I can model mine after? Something that uses the +This is going to be the start to so Im going to create a new branch for development +Frankly I dont care if its dont scale and I dont see that problemInstead what matters is that its easier for new users of Camel to get started +sathish As it happens by explicitly specifying the driver the impact is the opposite of your intent +for now if anyone wants to use my changes use the older that only links the dependencies as without the source code attachment. +If it does not what we need to troubleshoot is the debug logs from the data nodes +Fix it +Others use +Thanks Ben. +The reason it does not build has nothing to do with the changes themselves +Not a real documentation +Then startr build likeDISPLAY mvn clean instalThis would be the replication of what is going on on wait should print out all jobs that are running +RESOLVED created extension point created JSF extension for created tests for in +all fixes committed to branch +This is the compiled jar with the patch put in it +Is this deemed acceptable or do you have other suggestions? Would this issue be easier to track if we closed one issue +Currently we promote Writables in +I attached an updated patch now extends and the loadsave methods of nowthrow a +Add a test case for compressed input files to because I forgot to update in the last patch. +Ill commit soon +bq +fixed by in jbossnative +Im updating all download and release notes links +We can add it +Original is the pdf we are attempting to convertThe file is the out from convert version the colors arent inverted anymore see attached image +INFO Recovery time calculation reassigning regions took ms INFO Recovery time calculation reassigning regions took ms With the patch the took time of reassigning regions in the process of SSH is reduced from s to s +I believe addresses this +Created a separate issue which is more think the need it +Bulk close for release +I was wondering we may be able to use somehow in the future if we need to do something to the RPC engineStack if we are sure we are not going to use we can redefine the interface and now is the time +Could you be more specific what do you need? I can think only of a box where you can reproduce +has no change please revert it +Dont have access to this device +I merged this change to. +Please the server and port number +I agree with Burr it is actually an issue IMO +Ill go to legal discuss if I have to but we have dealt with this type of thing before so if the right people respond Id rather just go that routeWhats happening with this one? It would be great to get it tied into the build in trunk +Since there are no takers moving this out to. +The pom is missing from that rev to boot +I saw that fsnamesystem puts blocks in the deleted file into invalidate blocks set in the end of file deletion +If we want to get a samples fixed we should have a different JIRA for that +Cant reproduce this under any combination of types of snapshots suggest just adding it to the wiki if you feel this is useful +There are some restrictions tough Any web protocol resource path that needs to be resolved on the new thread must be absolute or Dependencies on Request or Session scoped objects will not important restriction I forgot It works with NET only +All the partitionwisefileformat + pending patch is only doc change +I agree with all of your points except branch inclusion +The key is we need to flush after the recovery so that we no longer need that WAL we just read to recover in case of subsequent crash +Thinking it through I have a few concerns for consideration +micros makes more senseI think something is wrong with either your chart or your percentiles though +failure is unrelated +Updated per Trustins suggestions +Released few bug fixes found in beta and added more bugs found are and were manifested on certain configuration params when they were set to zero while the default values were is still on possible bug that is reported when using Tomcat but I was not able to reproduce it is realy a bug then Ill release another beta version or +Attached which is causes a uses conflict when being wrapped by +This is part of the upcoming cr. +In other words rather than generating class names such as an incorrect is being generated +This requires the RS profiles POM and to be updated to not include the JSON extension +I havent gotten to it yet +At flush time a is thrown + +That way Avro data can be read whether or not a specific or reflect class is loadedSo thered just be a single +I am working on a patch for and goraassandra which bring this bang up to date +Looks good to me +However I will try to refine the conversion algorithm to keep the original Map as far as possible that is as long as there is no actual element conversion just ran a CVS update compiled and ran my test +Well jar file isnt indend to use with any unpacker programs with UI so this isnt relevant problem in Windows showed those characters as two byte characters +Attached you can find the stacktrace for the same problem when using the web services interface +Users can download the results in CSV file format +Verified on JBT. +This change seems to break the integration tests +David its not there yet can you check on this? Thx +submitting patch +I dont understand why the component for this is Domain Management +I think I can hack a to make things work in the short run +Yes the error is related to the last minute movement of that test into the Fair packageKarthik could you quickly patch this so we can get the build stabilized? If you can fix the findbugs quickly that would be great too as i said earlier these are false positives but we might be able to coerce findbugs into not spouting warnings +The jira shouldnt be marked as a resolved until its patch is committed +Local test passed +please review if this gets no s other than lack of tests +Could you update this patch to the current svn version of? And we need to have some test cases to prove this works as expected +Ive updated description +Moved siteDirectory from to so we can get the true location for site +Cancelling the patch for now +Who owned the source files when you were making the packages +I am not saying there is no way out of this problem but it is like this purposely +Works like a charm +Deserializer implementationWhile it works code is by no means it has to be cleaned up first +Let me know if you need any help doing this +Thus is failing retrieving the jar from the repository. +If we find that file we have to read it line by line and process it just like other hive commands +maven closed so resolving this too +In this case I compare entities that already exist in the database which is why I opened this case +It is not obvious scenario to replicate but sometimes the same is happening for Hibernate Console Configuration File +I did that but I seem to have missed the part +I ran Fengs script except I left out trying threadsThreadsThis was stock hadoop and tip of the hbase trunkThose are pretty big improvements once we pass out ten concurrent threads +Thanks Kristian +The existing binaries are old and I do plan on generating a new release +The change looks simple enough +Is that right? If yes then the situation is quite a mess imhoAnyway +I dont know if a feedback in the session can also have a reporter if so not clearing it might result in +Can you pls comment on any manual testing youve done to verify thisAlso can you pls provide the patch for trunk too? Alternately I can wait for the merge and commit it after +If you are deploying a Web Application you can select the Search Local Classes First option +This patch fixes the issue and also adds time between checks for requests problems +If that fails for any reason throw an including the cause of the failure +I am moving the issue to there +attach an example that reproduces the problem and provide steps to upgrade to and am still observing this behavior +test to demonstrate the problemNote memory usage on interpreter is also very high +And result does not depend on view used to open it +In theory the Configuration supports lazy initialization of any component simply by marking the Plugin with deferChildren true +This is related to classloading +This should have gotten fixed in RC +If you feel like taking this on for sure +tests included +This should apply to as Hudson modified patch for just committed this +Ill post it later today +How much longer do you think the build would take to run with an extension to all tests in that group? that is within have updated the following CI jobs such that timeout periods for the QA Crash recovery tests are doubledand If this resolves the problem I think we should add the change to hudson script failures since doubling the period the client waits for crash recovery to complete +Also there is no annotation present +I dont see a reason to hold avro in lib as core code does not use it +Code changes in patch format +Patch v with formattingWill integrate after test suite completes +This looks like a duplicate of and +fixes this one was fixed by chance when I got resolved +Committed revision +Could you please provide a description of how you have set up on your computer +It is decrypting the assertions properlyAravind +now I started with the simplest thing just expose all the spring beans in all the expressions +Also added missing files +resolving since it is working for us if qe finds issue reopen +In the meanwhile we have already upgraded to + +There was no need for M build to start the server with b option so I expected the same for M tooling +The attached patch is much more elegant than what I initially proposed it simply makes the default implementation of look at the annotation +Are you sure your factory is threadsafe? Does the factory cache or pool sessions itself? I am confused similarly to Mark on the timestamps the objects being evicted are not old enough to be closed by the evictor +Attaching a patch +Alexwhenever is converted to and such its most likely that this is a ln issue where the is interpreted as and not as decimal sign as you intended. +Additionally I have the impression that you search help how to use but thats what the users list is for not a bug reporting system +In current SVN it works supposed to be Stub guys +Fixed the registration of unused namespaces when initial cnd files are configured in the into the master branch. + +Need release note +committed! +I was hit by this feature just recently and was scratching my head for a while until I realized the missing piece +are you using https? Yep +What do you think +which addresses the bug in the earlier commit though this does not resolve this issue since we still have to handle the case where the of the machine architecture doesnt match the of the JVM +Done +Else path is relative +Patch for changes to prefixed Property definitions +its true there is a precedence bug but your example is not correctThere are no implicit +I think this patch has a memory leak Im trying to get a copy of an open source Yourkit license for profiling the heap usage +The decorator appears to offer this functionality. +The key is not created +Kurt is busy at the moment Thomas +Same as before but without the javadoc comment in that stated its an internal committed this +grant license toI linked the wiki to the new attachment +If you do not find a particular class in one of the jars that exist in the client directory then complain +Thanks +Anyone have a follow up on this? Seems like Hadoop based indexing would be a nice feature +Pull it back in if you think different +Ive already forwarded an email to Rob and Guillaume asking for their advise +If is not specified should we include all jmx beans +Proper handling of coordination events in a partial refresh will be addressed in a future revision. +Thanks +DavidThank you for the commit and for catching the error +After the initial connection is made if a send is attempted and is down an exception is thrown +Heres the dialog with a validation error +ElenaFixed in LUNI module at repo revision Please check that this fully resolves your problem +All tests pass with both tomcat and jonas with this patch applied +Are you in a position to try out the latest HEAD code +Submitting patch for branchTesting done ant Ran times Created node cluster +Discovered that was holding a piece of state data methodname +Ive tested this on my Samsung Galaxy S running and I cannot reproduce +We should each spend some time reviewing each others work and then have a discussion to go over it together +Tim could you also look at what is being proposed in? The patch there seems to make an attempt to fix things but Im afraid it probably breaks things in a environmentHow do we break core into multiple bundles without breaking it into multiple Maven projects and multiple jars? Most of us here have very limited experience so any patches you could provide would be heartily welcomed +Giri stack can you help with the publishing this patch to the branch republishing artifacts to validate before that patch could be taken up for consideration with hbase community +and trunk have the fix +There are a couple of changes I am not sure about +In we use spring to wire things together at startup instead of creating these classes on demand so i need to feed them with proxies that delegate to thread local storage +Implemented for async in up to outstanding issues with serialisation of exceptions in managed backend close of resolved issues. +Sorry missed the question +Abdera is an old release that was based on an earlier version of the Atom Publishing Protocol +attached test appreciatedThanks also for the excellent unit test +patch to this jiraNo +Attached a patch that fix the issue +bulk move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +What you get is not one new collection for each operation but instead you collect the operations and if then the user asks for the next element the original Collection is filtered according to the given operations to get in the end no new collection but only the newest element +HiThis should be reopen +One solution is to implement a query execution planner so that it can optimize the execution path to avoid the bad performance or look for best performance +This is planned to get done in +Table file which contains a number that is too large for integer +Moving to the time not have been was implemented no plans on implementing at this time. +okie I just ran the test again passing +If JAAS security is enabled we will inherit the secure connection obtained by the plugin +No reason to use platform suffix in source filename if they reside in specific folders like vmcoresrcutillinux +Please stop this bug +I just tried on IE and I havent managed to reproduce the bug because IE seems to wait for the ajax call to respond before submitting the form +Probably Ales will know if this has already been fixed +I havent been able to reproduce this issue using IBM JDK +Martijn you said its fixed so can you resolve the issue? +User error +I found that upgrading to still works but that I start seeing problems at +Did I miss something? It would be nice not to have to create a pom for a Sonar light project +It is still early days and I think we can ask the users to change their scripts +Other minor work performed on a few files to get them to validatepass ant test again +Is there some other way I should be inspecting the data store in tests to assert messages have the appropriate properties +If every one does the same I want max of n tasks all resources in the cluster can deadlock +Thats great thanks +It was by an bugs in build file +Might have been worthwhile to create a unit test that reproduces the bug in the iterator +Performance are better in is better in +Closing the issue now that it has been resolved. + thanks Shreepadma +Okay I have written a test for this by creating new collections on the fly and I cleaned up a bit and fixed a bug when you have the same shard names in different collections by creating a new shard key that is combined with the collection name +I would like to hear from more people involved even if it is just to say that it looks like we have everything covered here +Hello TonyThe patch is applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expectedBest regards at r Thanks Richard +Closing the bug based on my previous comment +Submitted revision +It is missing in my old workspace +confirmed in trunk +This seems rather arbitrary +I did think about making a test for data first but as the point of this exercise is really to convert test and data is ok as is in java world I decided Id tackle all chained tests together +However I would be a bit careful making assumptions regardingcompatibility with transfer encodingsOleg +Alex could you have a look at integrating into or comment appropriately +I spent some time over this weekend trying to backport and found that I had to hand tweak quite a few merges for just one revision +I think changing the module name on a big fix release would be in bad form. +This patch splits the HDFS delegation token classes into reusable parts and moves them into common +Solution allows to enable this feature and keeps it disabled by default so I see no conflict here +The wouldreturn the child of the first node to the and the would return that to the caller +Mikhail Are you planning to look at this? Ted pointed out that this might be a blocker and I concur +Off the top of my head this could be accomplished at compile time via Ant bydymanically choosing which code to compile based on the JVM version +Hi Tobi thanks for the patchIts applied now can you please cross check and close this bug +The workaround of creating delegate getters and setters for composite key fields is ugly +I deployed the latest showcase which contains examples of all the UI tags +close it +Concrete tasks will be created by our TL +patch switches to new blob youre right +Can I still use and still go around Nexus to get the snapshots +I fixed it in r +is available in ubuntu precise +The URL should have been example +The new one is the correct one +Stepan applied second patch is ok now we can consider that issue is resolved +Well +Ill continue to transition the other files and submit a final patch unless anyone has any objections +Hi Nikita that issue was rejected and also I dont see how it is related to this issue +Complete project src AndreiThanks for your work butSome changes in your patch have already been committed to trunk some others have been solved in a different wayThere are places where the whole method is deleted and replaced with a new one +It does thats why I called the metrics completed not successful +It looks like it can fail in a few different ways +Heres a patch that fixes this with tests +for the patch +Attached is an example of the issue +Issue is resolved +This has been fixed some months ago. +categorize as small test +Please confirm +Add appContext configuration for a flexible web app context configuration +will be known issue for some cases of M on linux +sets a value for the configuration +classpath which has been done at the linked wish javadocpath was supported +Does this need to be documented in Continuum or is there a Redback page we can point to? +Im working on it +exception reproduce the issue +Sorry for the false assumption on what that transaction attribute meant +I got the same issue as informed by Wilson +Resolving as fixed. +If we add a mutex on loading a given file a given thread will safely skip it for and other threads would block until the original require had completed +Previously network server would always call getBlobgetClob for any lob related call +Yes it was certainly a problem with old dependencies to ant or will be fixed as well release all plugins before the beta you can reproduce this bug with the future beta well reopen it. +Rafael could you provide a patch +This does not affect Androids cordovarun command +Migrating bugs marked as resolved before and tagged as version as fixed in RC +Thanks for the patch JrgenI have tested that I no longer get the NPE stack trace at failoverCommitted as revision +all who know the progress about this issue +Comments on RB +closed for +We need stats to make the Source impl efficient so we can +It will close the command window after the build script completes +Checking this in +Waiting on update from WS team +Cant you do the same by implementing a custom? This would be cleaner IMHO +Sorry for the chinese characters + +By default the generated are stored in target folder +A patch to will fix the problem described in this jira +However returning the id directly might not lead to a good hash code +It is not fixed yet +It is more coherent from my point of view to add the buttons on the screenlet that is going to be manipulate +Verified on JBDS L. +Note that implementation is full functional and is part of a real application since May If you have any questions fill free to contact me on fassev gmx de +It looks like youre building binaries on AIX where wchart is a code unit +Eric applied patch and rebuilt sablecc generator a behaviour to verify the use of comments +Server is still launched using Linux distribution native +I have been thinking about this very problem for quite some time +bulk fixing the version info for and all affected issues have in will probably not be a problem with and beyond closing +I dont think it would be a good idea to integrate it as is and later have to change a lotNot sure but would we have to vote on this contribution given its size +Ping +Not introduced by your patch but will be good to fix +Thanks Marco +After the open questions are discussed I will upload a final version of this document formatted as a patch which extends the existing documentation +FYI I have only ever seen this when importing maven switchyard projects +The attached patch should fix this issue +Your right inheritance is what Im asking for +Doc update for new permissions requirement + +Please check if your string length exceeds the maximum length of the column which it is stored to in the database +This is the first time I submit a bug +Fixed attached in applied +Javadocs fixes +I committed this patch with no changes. +Like it is not working as expected +I was fiddling with it and there was no way to fix it +What do you mean by wrong order? When sorting lexicographically I would expect it to return something like etc +It also may be that related to this Safari is not the existing connection which would also contribute to so many challenges in a short space of time +closing until we have a quickstart +it will be their responsibility to keep track of that for restart but I think thats as it should be +Fix version should be I per Jiris comments. +I guess I made a patch for this sometime back +Merged into the branch and into the master branch. +bq +There is no more use of and and thats good +Removes the already registered insert action from the action could you provide your changes as a patch +Looks like since Groovy something has changed in the compiler that causes longer compile times +I added a property called is set to false by default +If users arent using the could have an optional dependency on a module and if users want to use it they can declare it as an explicit dependency so it gets put on the classpath for their execution of MAC +Typos fixed +Any ways thanks a lot +The client does not set appendChunk to be false until the first chunk has been sent +. +Ive reanimated the code that does thread join on shutdown +Fix in rev +The answer cannot be anymore precise than the most precise operand +Attaching and with changes to the following filesM srcdevguideM srcdevguideI removed the reference to databases from since the subject hasnt been mentioned yet at all and then added a section on referring to databases to with a back to for more information on paths +I believe that Matt Diez has automated the Mapbuilder Shorter +This sounds like a good idea indeed looking forward for your progress before committingThanks +Yes it can be resolved +Ran and twice with patchThey passed locallyIntegrated to and trunk +There is one bug relating to not trimming whitespace from text +Also there are currently no tests for the Spring Parser and Im not sure how to go about making one +This issue will be closed unless you can make a small test project that demonstratesreproducedThanks +Attaching a project that illustrates how to achieve this. +Small patch review please +Wouldnt the clientId sufficeAs suggested by Jay and Neha I added a field in to indicate and maintain the correlationId +TT in its offerService loop detected this task as running compared the TIPs last to now and killed it as unresponsive +Patch generally looks good +You can use LOCATION clause in create table statement and create the table in a directory where you have access +Old issue specific to a single applications Java embedding capability +Not sure where though +Also there will always be container for the Application Master which also uses a user configurable amount of RAM +Is there another war file running in tomcat generating those? The stuff is something Ive never seen and suspect its not related to geoserver +We need better error reporting +Looks good to me +looks OK to me +Look at we do need a Row to serialize. +hope this helps +Looks like I only applied to branch +PDF of the of the proposal +Still the proxy to sort but that can be done once bretts set up the application. +Some sort of transport from messages from the application layer to the view layer +Pls find attached the patch containing the Stored procedures related to adding support to retrieving lobs from Network Server by locator rather than by materializing the LOB +Exception refactoring needs to be done with consideration for users of the existing exceptionsSo I would say go ahead with Lucs last suggestion make the enums change and then separately address the exception hierarchy +The fact that the singleton service moves just fine but its just the queue depth that is wrong indicates much more strongly that it is a JMS bug which is why I posted it in the JMS forum +I just committed this thanks Cheng. +I think is used at runtime for generation of dynamic clients not sure if this path to the dependency is correct howeverYes +Solved by. +Setting to future pending a patch +It can be reopened if we get a reproduction. +Thanks! Take it over +The change in is done while I was trying to pinpoint a bug in remove +Move the fix to +Further the fact that we are using an official namespace at all is potentially problematic since we dont have any disclaimers around the API saying it is not official spec +The only use of this call in is in the Log class ie its purely in debuggingdiagnostic code so performance is a me the cleanest approach would be a class and annotation that prevents insertion of barriers. +That works too good job. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Ill start this issue +Verified at +Glad to help +The screenshot illustrates that the width of the popupPanel is to large +This can be used to decide whether certain work like split calculations should be done by the client or within the AM itself +btw for the purpose of this jira we can postpone the batching optimization and just do the structural refactoring and get it done quickly +Thanks Andrzej! +I think this got fixed by Warbler fixes at some point +Note that stores the TCCL as part of creation itself +Yes that would be fine by me +Unfortunately I havent been able to repro manually and ccm has a problem with logging at debug on currently +One more small thing uses double ord and double ord when comparing slots but in fact they should just be ints +Find the attached +Released in today. + +Implementation for using space as a separator we add a space even if a string is empty and so there is a if condition which could be discussed +As we are no longer required to have a banner Ive taken it out +I will take a look and get back to NamitPlease let me know what you think of thisThanks JoydeepThe patch as proposed makes no attempt to cleanup the output directory in the case that jobflow failed +may stick around a good while if any one of a RSs Regions is a laggard flushing +I leave it open for backporting to backporting this to but broke due to a dependency of which has not yet been backported +That should be fast +Alexandre do you also have the SLD you used? It may be SLD used was the default point SLD offered by +Thanks for the quick responses Devaraj and GaryWe have with two messages andI think it would be better if the verb in messages is changed to noun MutationMy two cents +Brd already has this done so youre off the hook +It is not PL issue as the exception is not related to any PL internal cache or even identity component cache but gatein filter performing IDM call outside of valid transaction +Both patches are required to resolve this issue +While I do agree that should not be shipped by default a release note item would have be nice to warn about is not only a build tool I for example use the mail builder to send simple mails so I needed it at runtime +rand with a seed is deterministic +very much RickCommitted patch to documentation trunk at revision +instances are inherently threading unsafe and may only be read from or written to either by the IO dispatch thread or through interface +Did you want me to post changes we made to it? I wont test it again until tomorrow so well have to let you know whether this works +Ill attach a demo which uses a cxf producer endpoint to invoke on a CORBA server +There may be already an issue about that +and the cleanup object would both have a handle to the same event +The SWT error is probably caused by you using Eclipse or with latest hibernate tools which requires Eclipse The out of bounds error looks like the database returning zero length column names reaaly weird. +To be reopened once further details come +BTW am I allowed to use Java generics in benchmark? Or until benchmark should stay on as well? Im asking because I heard a couple of times that contrib is allowed to move to Java its best to stick w +Bulk change of Fix Version for all issues fixed or closed for +Can the orignial reporter please provide more information +Merged to doc branch at revision +The improvements in analysis that I mentioned in the original description and the patch should be applied regardless of. +next rev of the structure completed +I like the idea of checking to see if the file exists before using it +So I removed to class and and changed the templates to use +for setup environment and issue depends on hive but hive is not structured to the new file system layout +Unit tests have been added to mimic the local infinispan server in a remote environment +Knock +one more thing i only realized when applying the patchIMO the compareTo method on the will not work properly +Committed +Hi Robbie can you review this commit +No one has looked at this in a long time. +This wsdl file use so in the wizard page Web service Code Generation Configurationyou should click the check box enable then you can get all the impl for every port based on your comment in the jboss ws forum the easy fixed way is remove the checkbox of enable and add e option for all scenes +methods now catch any not just +Should we really be picking up enough tasks to kill +This is the MarcoAt least for the moment I can say that there is only one tab in line of this patch +Nonetheless this seems like something of a missed consideration so I will resolve this issue and open a new one for Embedded WAR +attaching patch for just committed this +Thanks Mikhail +Committed to SVN +Now that work is underway with AS all previous community releases are +I committed this to the branch as well. +Found another place where the correct state has not been checked In the case of starting SCR after other bundles components of lazily started bundles are not loadedFixed both situations using a new helper method which should these two situations in Rev +just committed this +We decided not to touch at all +Some confusion on my part the title of the issue is misleading it should be something like bind tag requires wrapper command when binding collections +Manually tested +I have updated the patch by fixing the comment +should it be fine to run the whole thing in MAHOUTLOCALtrue mode and bypass any complexity issues from having a separate Hadoop cluster +looks pretty good to me since the usecases where you need debug here anyhow are quite limited +PR merged +I hope its better like this +Removed now that we have Bamboo up and running +Now that Spring has a Lucene API wrapper I can use that to get a usable API +This avoids concurrent startup issues like this +PradeepCan you create a different JIRA report for this failure since it is obviously unrelated to the one seen by Hiranya? +Reopened against EWP +I think I found the problem In the bpd is not set on thegenerated leader area for rule +Fixed on and me I fall prey of the same logic mistake +This would be particularly useful when there is no binding or validation requirement and we just want to get to the view state. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Which is easier for maintenance purposes pdf pdf pdf pdf pdf pdf this also contains some text and directs reads to the official Nutch tutorial pdf pdf same as comments for pdfI would consider the above the core documentation for getting up and running with Nutch however Im not sure if we agree on whether the wiki is adequate enough to store the documentation and whether it is tidy enough as well as being a rich enough resource for people to find and utilise the specific information they require +Im uploading the figure separately and it has to go to the resourcesimages folder to generate the docs +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +In fact no files should be added to the repository without this any more +I understand that we typically dont add new features to bug fix releases but I dont really see a problem with having into it is not really a whole new feature +Please checkThanks ClementI have just checked and it seems to work I can mix my composites and iPojo composites without problem +Results are as followsWith read timeouts simulated for map outputs On Single node cluster with maps reducer and max sec + for the patch +Im not sure theres any good way around it but perhaps Im just noting it here. +Heres an initial patch dropping and converting to a PB serviceIve tested this on a local secure Hadoop setup and had token authentication working for a simple MR job +partial log file with log trace on +Thanks +Try hudson w new version of patch +I think it fixes the glaring issues and has some improved tests +As discussed Ill just drop which relates to accesibility +Isnt this an incompatible change as it is implemented in the patch? Right now getResource would return null if the user has no access to an item whereas with the patch an exception would be thrown +Please verify your code against Lucene Java first +We should in all cases port this automaton from the tests into core +Maybe a unit test would be goodalso does this problem exist in trunk +Bulk closing stale resolved issues +Ive released and updated to include it +Thats kind of an odd error though +Clearly the documents as is are not appropriate for redistribution as they say no redistributionSun will not allow us to take their documents and snip out their to save the typingSo we should just create these fresh from the specgeir +Maybe you could try terminating your script with a new line? Like this x load frogAccording to the Grunt parser a single line of pig latin is not valid +Do you have the source Jar attached to the Jar in Eclipse as outlined in the Spock documentation +Hi ChrisThe snapshot works perfectly +jcrmixinTypesAddmixins which just adds the mixins +I will close the issue +Well tx might be working fine since it uses Spring AOP by default +In fact the surefire plugin is not executed but if surefire reports exist in the target directory Sonar will collect them. +So NN +Awesome! One last thing is to move the class into the existing geotoolsgml package structure +Spanish attached file MUST be downloaded in taglibsrcmainresourcesorgapachestrutstaglibhtml +however Steps to reproduce field mentions only two stable products Eclipse Java EE IDE and Server Adapter Features from stable update site +I have attached the beginnings of a test that I wrote in order to verify the fix myself but I got an early failure +That is really good +Making noise on the Hibernate JIRA would probably be more effective +new patch fixes the issue and does some refactoring on the bookie side to try to make flow clearerThis patch depends on some reason I couldnt update the patch in review board +Search path configuration +did changes for invoke mediator config names iechanged name to +My war is named httplocalhostservices? if were going to ditch the juddi api eventually might as well make it for +Should be in the form of a feature matrix on our public products to compare are IBMs eXtreme Scale and s +Or some such magic +Do we want to push this into the branch as well +Thanks for looking at this KnutYou are right it got a bit convoluted +extshape moduleBEFORE AFTER the extshape module is affected +Opps actually this is fixed returns and in Apache Flex +the testcase assumes that history is always activated when running the testsuite +No need for IT +Im stumped and dont really have much time to commit to this hunting the problem down right now +Could you explain +We have a spec for this that runs in a multicast environment under domain mode and it works without issue +Includes addendum +BM is a new codec for +Is there a clean way to get Maven to support this use caseWe also need to update the documentation to say from or onwards or export the documentation for later use +Implementation of the suggested refactorings. +Tiny patch skipping RB +Closing. +Attaching new patch addressing your comments +attached a patch that fixes the pointers +this was never accepted or rejected and I was wondering why? It would nice ifthis was added to dbutils so that I dont have to maintain this code locally +Ive done some further research +Read the analysis in minor adjustments could be done but in practice the tests show that interop is not broken. +Verified +I suggest opening a bugzilla report over there and providing the Tomcat patch over there +If this is confirmation that the codeisnt behaving correctly and if nobody already familiar with the code has timeor interest I can start looking +sincerly yoursEdi didnt offend me but you can have a conversation on a closedresolved discussion +Not yet +Ive changed this code in the SVN trunk +The new added test passed if we dont mass the Spring with Spring. +Attaching patch that switches to + +I didnt veto anything and this has nothing to do with C +I tried looking at the history of in fisheye but havent found anything obvious +! was removed from the close of resolved issues. +For what its worth I think lazily replicated counters is what I want +bq +Can you provide an example portlet showing how you are transferring a value in the flash during a redirectFrom an example Id put together I dont need the above change to have a value available in the flash after an ajax redirect +Itll be useful to have this already in. +Handler is already in current svn revision +Lets see +merge everything to the only issue was years less than we might want to consider bringing back the original formatting code and fixing itA quick performance test shows its over times as fastUsing the current code in I only got about K formatssec +Yes the print but is fixed in this build +Moving all issues to the CR +Committed revision +This allows us to repeat the run as described by ChengCommitted patch to trunk with revision +if threads but only servers can only see X throughput improvement +This issue has been fixed some time ago and should be available in latest previous comment states this has been fixed some time ago +batch transition to closed remaining resolved bugs +Please definitely take in the hypervisor related changes in the FS for documentation +patched scheduled for next out +Full patch now adding a protoc maven plugin to the hadoop maven moduleThis gets rid of the plugin defs to execute protoc as well as the plugin defs to create and set the dirs in the classpath of the project +Vitally could you attach a page +The second is that in the case of evaluate it seems unreasonable to force someone to catch an Exception that may or may not occur depending on there usage for example passing in a +As usual with date and time issues Id like to suggest we proposing this addition for instead +Resolved per pull request. +We have updated drools so that empty LHS are not allowed. +But either way I expect the last assert to succeed +I have Microsoft Visual Studio +Heres a patch that fixes this +Revision +This is for review not for committing until Im confident that the tests failing locally on the mapreduce project are not related +Maybe use messageId from Message which is what the Camel FileFTP components does if no explicit file name has been given + +I say convenience because all these options can always be specified as Dkeyvalue if you set all then necessary properties +I hope to commit everything today or at least I must solve the problem with missing in +If no one comments in hours or so Ill commit your patch for you +Please revise. +Fixed and checked Seam with AS works ok works with Seam and not with Seam works ok with AS uses + +Done now too fixed. +Thats my fault I thought it said EAP whoops! Thanks for picking it up for me to the set of javadocs will be in CR. +commonsowlsrcmainresourcesmavenrepoowlapiowlapi + +There are arguments that can be made both for and against the featureAgainst Many times requiring a prefix in the colfam implies that there is a table structure change which could replace the necessity for this featureFor Time to the millisecond in the colfam and the user wants all colfams before the year stored in one locality groupId like to restart the discussion to see if this might be something we want to tackle in +Some of the links that were fixed worked when the files were deployed on the Maven site but not when the generated files were browsed +Fixed as part of. +also looks like a duplicate in so far as clients wont be able to make sense of query execution information without having access to the corresponding query plan + for the to trunk +Thanks +Switching all issues that have been in resolved state for more than one month without further comments to closed status +That sounds great Jacek +Patch to widgets are Archived and doesnt have a anymore +Ive fixed this bug which should remove the exception +Due to the GA delay this has been Miroslav +I was suggesting doing it as a client script because then itd be outside of the servers and easier to test +Committed to the trunk as revision Ill investigate merging this change back to +Postponing to the next version +symdatasymtriggerIavayasystemdatapemmoduleavayasystemdatapemmodulemasterreplicaSMavupemmoduleavipemmoduleavdpemmodule master symdataeventconfigreloadreloadreloadWe create triggers on the master node call syncTriggers function and then call reloadNode function to open initial load for the slave node +I just committed this +Assigning all documentation related issues to final since they dont represent a change to the is already fixed in documentation with the texturl The url to submit to either a map contraining keys for the actioncontroller and id or string valueCan be closed +Hadoop was just released. +Since the lock is mostly uncontendedI am not sure what you mean mostly uncontended because I understand the problem description as there are lot of contentions. +Its an improvement we have to get to and its probably going to be a back compat headache perhaps we start by introducing something here and eventually this would become the standard client you want to use +Seems to be missing This patch is for +bulk close of all resolved issues. +HiThanks for that +On a JVM? My Mac has JVM challenges in general but I can also test on Windows and +Please have a look at the version in the repository and report back if it covers your expectation. +Maybe check spaces in name too +I needed to configure a new Context for each thread or reuse an existing one as by default a new context is created for each thread all competing for the same port +I wasnt hopeful but I figured it was worth a shot +Im attaching and the two files andIs this okay or do we need to add some more basic information in this paragraphIf its okay I can put it in some other places though Im not sure if there are any other appropriate places at this point +I would like to look in this issue +All the files in the qlsrcjavaorgapachehadoophiveqlexecvectorexpressionsgen directory are files generated from a templateThe template files are in qlsrcjavaorgapachehadoophiveqlexecvectorexpressionstemplatesThe code to generated the files is in +I already had those changes in my checkout Populating the cache is quite fast +But is running locally and tried doing it +Im using revision +Sorry for the trouble guys +Can you provide test case that shows this happening +I am for adding this to +defer to +What about only logging a warning when parsingI think thats useful +They should not be a part of the framework but should be added when really required through custom extensions +StackThanks for the detailed reviewEranIf you dont have time to respond please let me know +Thanks a lot! +All files found are filtered by anDirectories are not mentionedbq +Dont clear the data container before stopping a cache that cache might be participating in an ongoing the issues in the previous comment +Thanks for taking a look at this Daryn! I didnt think that there were any in those spaces +Therefore unless the problem is widespread +Execution in both cases is a Ctrl during hanging request followed by CtrlC once it finally finishes Execution in both cases is a CtrlC during hanging request Thanks please let me know what I can do to help further if you cannot recreate this +My Gid I meant Git repo not Got repo +Looks good +I dont have a better solution at this point +I found that at least the information about exceptions was outdated +So fix is better but something still going on will continue working +on latest patch +I allways like to provide status +Ask Kevin to explain to Len and Len to write something up +The includes configuration parameter doesnt seems to exist inregards +The patches are generated by simple refactoring in eclipse across projects +my comments in an email to Norman +only fix for if the fix is minor or if the memory leak is a real problem. +Will deal with the parsing issue +Each projects file sets to the +See my comment in +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Stan i guess this is obsolete with the most recent logger refactoring +rightNice catch! Attaching a new patch +I checked in the fix. +I get the same problem +you have an installation job for testing purposes? thats cool +I can confirm this +Assigned to Travis +This is a regression caused by +Robert thanks a lot for detailed explanation +If the changes pass the tests I think we should apply +The problem is that the connection profile wasnt created correctly +looks very useful will read it with interest! +Ive just committed this. +Makes sense to me +Applied patch with a few minor modifications +Committed fix +Im getting the same behavior in one of my processes +My to as well +This is actuall from the class which doesnt have a destinations attributed in the confi +This patch also raises an error if you try to drop a user who doesnt exist or if you try to change the password of a missing user + +In the servlet we look at the content type and select SOAP builder or MTOM builder +Patch provided by Dims applied. +See my previous comments higher above +It would be nice if we can do this unit test along with several in NFS +Very impressive the first time you do it +I think the change may also cause mismatched interrupted resumed calls as well +Moving to fix in the next CP +yes please i believe find bugs and javadoc are likely form else where but Ill do a double check before commit +Your main classes shuldnt depend on test classes from other I realize this is an unusual case + +Your only has a restriction on first soit isnt eligible for anything but the first page +Closing this one. +Attached narrowed down testcase +Is this issue solved with the changes that were included on the test rerunI dont recall when we last saw random failures from continuous integration that were due to CMAES + +Can you attach the logs please +It seems it works fine I just have to make sure some bundle is accessed +Again reverting out that change returns marshalling behavior to normal +as file or directory names at the namenode throwing an exception for attempts to create directories or files with these namesWhile were at it should we also prohibit and in names? I think so +Put in the missing bitsFixed in r +But thats another story. +The same change is needed in +Sorry realised first version of the patch was missing a change in updated +Are you able to provide a patch for the ugly hack or at least a link to the commit? Thanks +this patch closes it before the delete +move the inclideomxmlsec to rampartinclude +anyway Ive uploaded the example code as I shouldve done in the first should either add to or decide that its not suitable for IO and WONTFIX +I will create another JIRA error for it when Ill be ready with example project +OK here are my cents I took a look at sunris performed a couple of tests and verified that is not respected on SEI even if running on the Sun Java System Application Server as expected this is a Sun issue +We use RSA and have a lot of EE projects so we would like to fix that at a highr pom level +v only communicate the rpc address over gossip and fixes a problem in v when the rpc address is left blank +Read first then ask +But I cannot tell for sure as Im not using this one +It notifies about all instances and not about the instance that changed only +Patch test on patch I sent is full tolerant for body all methods can have a body it is to the user to decide what to do with it +Closed again +Another thought maybe File Install has some sort of memory leak +Number of fixes have gone in this jira entry and I think I have tackled all the issues related to this jira entry through them +Set Fix version to +Need to check it in in is done is need now to use it in. +so install a trigger in postgressql to delete the corresponding blob or have the blob mapped as an entity? From what I understand postgress doesnt diretly support jdbcs assumption that a blog is part of the row which means jdbc assumes the blog goes away when the row is deleted but posgressql keeps the blob around since it could be shared by others +Filed +Attaching Hudsons for the test from +to see the difference +Of course we can alwaystell people to use when posting large datacheers bug +short of that I have not idea +That is strange +Sorry for the delay +Hi JeremyCould you create a patch with this fix attach it to this JIRA ticking the box granting ASF the copyright? Ill take it from thereThanksColmCorrection tick the box granting ASF a license to use the fixColm +changed the class to no extend in rev . +Fixed in svn revision thanksdims +are designed to work in advance of the Tokenizer when you need to modify what the Tokenizer sees +As requested from user community this is back into +to commit and will do so within the next few days barring objection +Either provide explicitly a or use a version +Andrew totally will let Pritam comment on that +XuefuI see you want to make changes to have Hive be more in line with semantics +I checked three random samples out of the corrupt rows and all of them had an RVALUE identical to the row places ahead of it +Jody indicated in a private email that for modules without a maintainer any PMC member can review and approve. +Resolve when this goes in +Are you seeing the Starting message? +Committed revision Should the fix also be merged into the branch +Just added them in r. +Logged In YES useridAre you referring to the example config inclient? Just remove the you actually want the srp classes you can get them fromthe jboss source In YES useridThe indicated class is in the and has been addedto the for. +Im downgrading this to minor because we match MRI behavior on line and MRI doesnt appear to have changed this behavior all the way up through HEAD of branch +Furthermore im also trying to figure out all differences between the runtime and the changes i made will affect the following +Hi NickDo you have any tipadvice for helping me on this subjectThank you a NickI have just attached two samples files as requested by youThank you very much +Certainly did not want to crowd up the ticket +No integration tests +Thanks for solving the problem +Cool thx. +By the way is there any possibility to have the active timers list for a process instance +There is a related issue in nutch +Will change to back to bug +Alejandro Yes I ran the steps for reproduce on a live cluster as well and with the fix in place the requirement job runs while without it the slots are soaked up illogically by the other high mem oneUpdated patch with your suggested refactors +it will make it easier for review +Patch applied at . +I made requested changes and commented back +Could no longer reproduce the behavior +I made a mistake +Let me try removing that code right nowLooks like this has allowed us to get away with things we shouldnt +Anyone with EC access who is willing to verify it works? last comment closing. + from me assigning to Pavel to make sure it wont conflict w the new schema. +This is from before all binary KVs +Thanks I must have missed that +Click Add Class Folder +Since the dataset was too small sampler with the RLR change did not detect these small keys causing the unit test to fail +Re replicating the source file I agree I just wasnt sure how to solve thisMy experience with these classes is that they are not so often bug fixing is necessary +We want to be able to do a rolling restart going from to +I believe both the test and the implementation should be fixed to be compatible with RI +Thanks Vikram! +I mean the Group View link that is in the top right corner of each expanded project group +Thanks Sohan! + +Can we add some stuff to for this issue +closing the issue as duplicate of the newly created one in issuezilla +Verified with JBDS +IF you are seeing what is in the attached screencapture then it functions correctly and I will close s in video does not happen here +not critical pushing to are we even putting null into the text? why not just let it be an empty string +I thought that this was already fixed so I didnt even looked at the pom +Is this really needed in core? it is quite simple to roll out yourselfI havent seen anyone asking for this in the mailing lists +Hi BryanThanks for committing this issue and I am closing it. +No the problem hasnt been fixed +All issues other than are fixed +Thanks for the source code patch +Introduce ivy based dependency management first cut +As it stands I really dont understand what the problem you are trying to solve is because all you have said so far is I have come across numerous cases when I desperately needed to turn a simple +I was not using Maven initially and was a little preoccupied +I dont like the idea of building in the reflection stuff +login will be necessary in nowI didnt want to make too many changes within a single patch +Nicholas I have updated balancer patch on trunk +Verify that it is before all getMethod reference +Patch cleaned up for applied on branch in commit +NM me +Specific applications may have reasons to care what the exit code is +Ill attach an updated patch asap +Can you look into it? Id really like to have this logic in +Its ok you didnt know +This would be easier to administer if instead of a master file there is a directory with one file or directory per shard +Is that about accurate +Now that SOAP execution moved into its own module the dependency list on connectors is very minimal +what is the time difference between making the changes in and generating an alertYou have to wait for enough time minutes after making the changes +Version Final is released +Given that the implementation underneath now is moved over to Protobuf no longer works +Thanks Uma! +Thanks Mike +same fails for yum install Loaded plugins security Updating repositoriesUnable to read consumer identitySetting up Install ProcessNo package availableNo package availableError Nothing to do + +I had not realised this beforeIll change this to an ascii string in the meantime and that should fix things +Attached is the junit test modified version that demonstrates the happens in groovy see attached example +Ive committed the patch. +e +checkstylethrough +A better compiler could possibly pull out subsections of code without damaging the flow of the program +Closing! +Fixed a few minor bugs and updated the stuff just probably its own project! +Hi Davidsince there isnt any comments on this issue and i dont have access to the issue tracker I was wondering if there was any further details which you could provide for release notesCheers IsaacThe comments on are relevant to this correct s to the abstract base classes hence no +attach the patch for but I had an issue when compiling +That said I dont think that is currently an option since it would break a lot of scripts +The jBPM JPDL will be able to deploy processes +! +Also looks OK +Can a user get to the logs of a job that was completed earlierAs of now the History server simply provides a web interface for the job history files on the +Committing now +passes +TestedThanks Ronald for reporting this! +patch uses it for seeking to the end of the file in +fixed on rwaiting for Martins feedback before porting to for forgetting about this one +The class constructor has changed in Hadoop and my patch doesnt build +List I saw wrong logo JMX due to inactivity +Swallowing the OOME will make it impossible to detect that the system needs to shut down so I am against any such change +Agreed modmbox should ideally be more tolerant and be able to recover or at least generate a visible warning that the output is truncatedSo are you saying that there is a bug in the SVN mailerIf so does this need to be reported somewhere elseIll report the bug for modmbox +New added for the new test cases +The jar file was uploaded to jcifsjcifs while it should have been uploaded to orgsambajcifsjcifs latter conforms with naming standard in the maven repository upload guide and was also named as such in the bundle Ive uploaded +Sorry it was my mistake I didnt wait for the jvm to warm up +So in the first step all the mapper superstep understands the global key distribution and assigns each peer the responsibility for partition of keys such that there is a minimum of messages exchaged +If someone from Spring Source comes forward and says contribute it and well support it as an or extension or says were adopting Java send it to us then Ill go ahead and do it +Commentsamendments very welcome +you may want to close as well +Thanks Ted +Those could be phantom leaks since if you look at the functions in question they return those allocated objects to the caller +bulk move of all unresolved issues from to +Merged into in revision after releasing +yes +I take it back +This is of course and will need to be addressed at some some more +Verified in ER Plexus Apache jBPM a division of Red Hat +Can be applied on svn working copy by git apply you for the patch +Mass closing all issues in resolved state that have not been reopened nor commented over in the last month +By looking at the error it looks like an encoding error since I did a complete sweep of encoding fixes across the entire rest api I believe this is fixed now +So for now Im OK to live with the extra megabytes +Thanks BrianBrian I had problems in applying your patch +Actually now it does not even need the anymore +hopefully someone else can save time if I document this properly +yes Ive read that claim of passing null somewhere but never really figured what shall be returning null value on our side +I just committed this to too. +DoneMarshall please review as changes affect quite a bit of codeEddie please review as there are some Sofa exceptions with message texts that I suspect originated with you or Bhavani +Reopened still subject of m forced to use Eclipse at work so I can create the Eclipse project files if we should ship that too in blankWhat about the Qucikstart app +A duplicate of for the duplicate +eg +If they are consistent in not adding jira then pull their commit rights or something +Closed with the release of. +Committed revision Thanks Noble +Thanks for idea and patch! +Hell I think thats entirely possible client side since I know the impl has pairings available to it +Can Junit check for indexversioncommits to be same right before and right after coreReload +Looks good +sticking with that +Unless happens to be what you get when you check thecode out of source control +The attached patch fixes a few bugs and also Adds support for so that bundles can be +Can you add a cookiePrefix parameterThat would help me with my issue which has the same rootMy problem isI use the tree to show the content of a folder which can be selected by a comboboxThe user selects a folderThe Tree shows the contentThe user opens some subfoldersThe user selects another folder in the comboboxThe tree is reinitialised and shows the content of the new folder but tries to expand the same subfolders as the folder before because its saved in the cookie. +Thank you Jing +If there is interest in this feature I will write a patch to add could you do that? Maven dependency model doesnt support dependencies like osgi do +Existing implementation actually looks buggy to me +If we can avoid deserializing keys in the listener great +This does not happen in PRODUCTION stage +This fixes a bug wherein counters arent counted across tasks before the job completes +so Im moving it out of the timeframe to reassess +The only way to fix this is to use some kind of sequence maybe not. +the embedded browser is either xulrunner or whatever the user have configuredinstalled and the content rendering is a function of the browsers capabiities and the js +it will automatically pick up any extra features put into the base class I added +provided it doesnt cause performance of the operation to degrade to a delete by query with a The delete actually would not be with a query instead we simple remove all referenced segments from the segmentsN file + will commit if tests pass +to JIRA incorrectly was showing the fixed version as when it should have been changing to be +Changes are in branch +Thanks for patch N. + for the patch +That will ensure correct behavior and no further regressions. +Other supported databases do not have such exceptions +Lorin no problemKristofor Youre right thanks again! Thatll be remedied in my forthcoming commit. +fixed in jenkins is building successfully again +That makes sense +Thanks for the patch. +this task the testing itself is done +by Denis. +userx lougout login as root again +I believe the correct thing to do is to never assume a version when opening an outbound connection rather retrying a few times waiting for the version to be detected and eventually failing fast +This is not acceptable +Ill spend some time looking at it this morning and see if I can make sense of it +This should contain rpc port +Using gives us the performance benefit were seeking in this issue I think the small sacrifice is worth the huge benefit +Am I missing something hereI realize that this is now a different issue than I originally wondered about +Attached is a patch for and which checks if the job is map only and sets the schema accordingly +Konstantin does it satisfy the requirement for block reports I mentioned bq +If you show me any requirement of that kind ill add such functionality. +I committed a partial fix needs further verification before release +We can log the name of the that it is starting +Can you test it out now +Dtest +No regressions found +It would be great if you could make the Name attribute appear on the canvas +The patch for is now available for review and for merging with the trunk +Ive got the same error when deployed on Beta for both struts and as well difference is that errors out at application startup time whereas gives error when the action is accessed which requires validator +We use it in Spanish using and we donthave any problem with themI think your problem is related to encoding using while saving the XML file +Please review please review the doc +It appears to me that the problem is in the class +The +I recommend you try deploying one of the Struts sample applications to your server to see if it deploys and runs correctly +Patch applied for with some tweaks + +If this solves the problem could you please link this issues to the old ones and close them all together +Try Hudson +Opened to track aries issue. +Adds a set method and increases the default split size to +Sending file data +HiSorry for the delay but it would have been done already if you provided a patch instead of the complete class can you please provided it youll make my life easierThanksFred +vmargs usrlibxulrunnerregards GiannisSince this bug is attached to version which has been released it would be better to leave it closed and create a new issue insteadCould you do it please? ThanksIm also wondering what could be altering the property value on your setup? +Yes you can patch the code yourself +Can you provide a patch +Oops I just meant I fixed the spec +is the resulting page Minimal Height is fixed Link icon positioning is is fixed +Hi JinhoIm investigating documents of and source codes of HiveIf you dont start this issue yet Id like to do +We will ship with static libraries for Cygwin +Ok in order to allow individual developers to turn warnings as errors off Ill introduce a new cmake switch for this + is targetedfixed for as well +Verified with my local installation +I think the problem is that if no feature types are requested the targetNamespace is undefined and may be determined by feature type loading order which may be filesystemplatform dependent +Nate confirms even Hector doesnt use it anymore +The List can be eliminated +Updated in ESB codebase will be in next merge +can this be done as part of this patch Kay KaySure Dhruba +push to +Even then you will only be able to know the numberof skew joins in that case why do you need to tag array just increment the counter every time you see a new keyEven if you update it multiple times it is OK +As per IRC discussion with Jason I look at fixing this soon by having the fallback to trying using the defining classloader for resolution which would leverage the servicesimport dependency thats set in the jaxb api module +New patch that waits for startup in +But how stable is trunk? Can we trust it in production +This is already tested elsewhere. +Have run the system test with message size K batch size and compression turned on +thanks chris I have been meaning to fix the SIGPIPE in the tests but always forgot about it +Will commit if tests pass +So we display a message that the user should call the supportSo if is not exposed to the client why does have this hiearchy +This may now be handled better using marker annotations to disambiguate the contribution to Alias +Throwing in my +Closing this as the latest version of Buildr does not do automatic downloading of dependencies +Bulk moved to processing JSTLs escapeXml doesnt make any difference in the resulting HTML code in Visual Editor because JSTL could have only text symbols as escaped sequenceand will display them as HTML codes by default +And yeah I say lets get it in for +Bumping to Major from to be specific +Built without the manual because we still dont have xsite allows us to build things that need objenesis though so we can move the websitemanual bits when xsite is ready +Benn can you take a look as our resident Windows maestro + which version was this fixed +Ill ping themCheersS +Fixed with revision +having the types after of +This shell script demonstrates the problems Im experiencing +while loop +Closing this issue. +Yes it is a passivation capable via Serializable but we do not validate it because is not passivation capable scope Bean is not injected into any other bean in the TCK testTherefore it does not trigger validation at startup +But are they all saved asAnd do we have internally files that really need that encoding when loading +There are around nodes under a parent and it will grow furtherDo you think that we have to increase the columns size furhter? Do you think it will affect the performane in long runRegards very much recommend that make your content hierarchy at least one or two levels deep instead of having k nodes within a single parentIm resolving this issue as a duplicate of that is a generic feature request to support flat content hierarchies +I also assume that is the same sporadic failure we have seen before +Initial diagnosis indicates a timing issue between stopping the network connector and start it again it appears the restart is happening before the connections from the original bridge are +Do you have multiple data directories? If you only have one it may have been blacklisted and marked by a previous issue can you check the logs for anything like that +These scripts use already existing set of templateslicenses +Sorry I only saw this once and I didnt have time to investigate so I dont know what the cause is +If you get any further messages from Solr from posts of fields with semicolons in them can you please attach whatever Solr output you have to ticket? I believe this is a Solr issue but I need to convince Hoss of that and show him how to reproduce the problem +I think its best to start restricted and correct it later if needed +I wish we could remove this but I suspect users would find that painful other databases also retain the column number ordering +skimming the code a bit this doesnt seem easy to be a very a clean way to do this +Thanks. +Looks like workaround for this introduced. +This is not how is supposed to be used +I dislike the use of try finally in Java because it either causes the code to ignore exceptions in the normal case or hide exceptions in the exception caseThis is not true as long as the finally block does not throw an exception of its own +This patch works on the JBDS code stream +The context menu component is attached to the text Context Menu +Also to be clear the cloning seems to give us a little boost too so Id love to integrate that as well +Did not dig further to see where is the problem though +Missing topics securing various bits monitoring messaging persistence +parse is simply a keyword a simple flag to instruct the stop directive to end rendering up to the last parse call then continue rendering +backports distcp part of +trunkCommitted revision +Hm +bump +branchtrunk +This is a candidate but not committed for EAP have just today tested this change in EAP to do my own measurements on the performance impact +Yes it appears that a source attachment is necessary +Hi Andreas just tried the changes the results seems to be OK thanks for checking this. +In FF works +I had one for + +please post another patch with svn diff format +Dima stil it doesnt work as expected and reliable +It does a Saxon DOM String Xerces DOM conversion +A HA NN no longer requires expensive hardware see the issues of is my NN scalable I think this is a novel idea which I would be interested in trying to help +The annotation names Thrown and Invocation are need to choose the names for advice parameters of the types Invocation and Throwable. +Ive fixed this as I suggested now direct dynamic dependencies are always resolved so deliver can do its job properly for any dependency whatever the order of dependencies is. +to Daisy review. +Dont worry breaks are on the todo list +Even if this behavior is unexpected this is probably only useful for testing were the workaround can be was needing it to test the fact index is not destroyed when a new node joins but is not urgent right now +You could also bring this directly to the dev list to get more feedback +It used to work on previous hudson versions for me but I think that hudson has changed the way it handles the build config +SFLJ does print out a URL where you can figure out whats going on which should be enough to figure out how to proceed +Pushing issues to the next CR +that patch is for the looks like an old copy in your ivy cache +Definitely looking forward to this bug being resolved +Moving to as didnt make it to the readme is in HTML convert to a txt fileThere is no diagram for the patch applied. +Can you confirm +the patch is there +Implementation of +Coded added which apears to work. +These are now backed by bounded queues with a default queue size of each default async serialization is disabled as well +I agree that ideally specifying should be enough +seems to be pretty I get the same results as you +Here we go again +However not all artifacts are there +looks like a corrupted repo +Hmm ok +with this patch it only checks when you actually need to use the deployment folder +I am attaching the changedadded html files +Done please test +Id like to highlight how much of a bottleneck this is but I dont want to add a tax on every to do it +I kept the original name as there is something in it that specifically pertains to exceptionsWhen you get a chance please take a gander and tell me if I missed the boat +Thanks Lohit! +Please assign back to me when this is fixed in the EJB code of the xml based metadata currently works when deploying a ws via an ejb +The unit to CVS +For this reason I would choose to use the generated ids if something changes in the class that will make it incompatible with older versions we just the id value +bq +Hi Jarek any more potential issues you may expectSince you gave the workaround can we think this jira is resolvedThanksForrest +Just dont work on this issue until we do have that +Please note that was about a quickstart +here is thine trunk +JFYI +Dependencies have been upgraded to groovy asm junit missing svnkit to same dependencies as Grails +Reported several times already +Having set that our test cases on sorted set do not show any have not looked at the problem yet but I wonder what your preferences are? Change the implementation or the documentation +pending Jenkins +It sounds like a pretty radical refactoringGood on you Good to know we can do it in increments Master RS etc +this is because you use Id +This is a error and does not follow the HTTP specification +Attaching updated patch +has been marked as a duplicate of this bug +We hope the info is complete in Maven final +Looks like was unstable just received a notification it was fixed +If you finish it before we release beta please correct the version +But getting that fixed is not an requirement right +Hopefully some mistakes in the future can be averted due to this patch +it should be left as just timeout +Im not seeing the problem +It addresses feedback from Sanjays review of the common patch on mostly just pulls the bulk of the symlink tests into a common class +Now Ive tried setting it and it makes no difference +The regression is caused by cebbbfcfdfaddbffb +Can you please re open this ticket as it has the required files to the scenario +I only checked the output of a few but all had the same type and order of exceptions +Applied in r +In we can get it refactored pluggable and consistent with the rest of the hadoop environment +I included a patch with that change +bq +Sounds like it is something to look at with Simone has a moasic operator I will mark it down as wont fix +Unassign so a committer can review +you mean as a portal object property that would be inherited in the hierarchy? like root portal page pageI am working on abstracting the error handling in one class the class implementation would look up the SHOWHIDE properties in the portal object and in addition to that would contain a reference to an URI for doing a request dispatch to an error page likejspheader where the resource would be in would be accessed by a request dispatch and we could set request attributes properties with the error nature and such +Could you please attach a patch a simple unit test +bulk close of all resolved issues prior to release. +Thanks DiveshChanges are in trunk except changes done in framework codeWe will wait for someone to commit those changes then I will close that issueIn file is is already exists but those takes date String as the method argumentAnd those methods was not full filling requirementSo new method introduced that takes as the method argument to solve the purposeAshish Ashish! I would like to make small correction so that other developers does not get confusedIn file is and is are the two methods which takes date String as the method argumentDivesh have already mentioned that in my previous comment Divesh +Is more suitable place to unlock conversation +We decided the caching should not be done in the project builder +Please confirm that they have been applied as expected +I built locally and verified the binary tarball contains libhdfs +fixed at with the release of +No duplication. +These tests were committed with SVN +Ive added some extra docs on how the request is matched when using ant patterns +fair enough I will revert the change +ATM we mostly use +This should now be fixed in CVS +no files for x were checked in +The reason not to use the and is the performance +Fix for needs to be fixed in and made consistent with mapreduce +I have assigned it to you +If it works feel free to simplify +Groovy Web Console updatedI had some other changes I had in parallel and had not finished them and didnt want to upgrade till I was fully done but thats now okayAnd memoize does work fine nowWell done and thank you! +Done +strange what version of bash is this? is this on one of the build boxes or your own machine +Assigning to Justin that developed per workspace admin LandryCan you provide more info about how you are editing the second workspace? Through Rest? Through the web interface? An example rest call would be helpful +Unzip the file and use the as specified in the problem description + configuration wich contains configuration +If your response to that committers analysis is to wave them off and send them to another forum then it is unlikely your patch will get committed +This change was included in sqoop versionanother issue with the Avro DATE TIMESTAMP representation as long is when using oraoop the value is represented by a String so the check with the long type failedRegardsAntoine +In this case the bundle remains in the set and is retried later +done +Roman Thanks for catching this +It would be nice to be able to change log levels from a command line utility too +Thanks Serge +I think the CP should know which all Mutations is getting applied each time +Thanks Jesse. +Sending systemsrctestorgapachegeronimosystemconfigurationTransmitting file dataCommitted revision +The issue is rescheduled to as it could not make it for the cutoff +Ales told Emanuel he should have the MC update checked in by end of day Friday June th. +My application doesnt even have any line of code written by me its just generated code +Dan you describe your js as a little hacky but is it robust enough to just drop into a control +We have to release today so we postpone this issue. +They must start with a letter and can only include letters numbers and underscores + changed the annotation to public stable +Fixed with using patch above +I was about to make the same comment about highlighting forbidden terms but then realized that you might search on one field and highlight anotherI think that goes against the point of highlighting which is to make more clear why the engine chose that document but some people have odd requirements +Bouncing forward to the next milestone +Indeed we have a timeout of seconds on our clients +Thanks Rick I agree and will leave that version number be +New problem will be entered in JIRACannot set Object parameter value ERROR The resulting value is outside the range for the data type DOUBLE +Hey Gavin some of our artifacts are about I get errors about the files being too large when i try to add these is there a limit you can increase so that I can upload these +If this function returns an error possibly this command has to return an error to the debugger +version of patch that includes backport of and +Thanks David +two use cases +bulk move of all unresolved issues from to +Yes and I have mentioned that in the steps to reproduce that is not the problem is that when you add the new portlets you need to close and reopen the browser before they are added to the list of portlets to be added to a already appear in the application registry page just not in the Edit Inline Composer until the browser is restarted +Modified his patch to fix those issues +Rather theyre just long functional tests +Apple is a major contributor to so I dont think Java will be fading soon +I spoke with Virag about this before he filed the JIRA +I would be willing to work on a patch for this if no one has enough time to do do +Attaching a patch that applies cleanly on ran forrest on my machine with patch +and Im too lazy too +Thanks again for your patch +As a result I cannot check if my null handling patches are working for good have to solve this one first +Theyre not listed because they are internal commands and the real issue is we have to change them to be prefixed by so that they dont appear in grails dont think its wise to change this now well just document the target ve doced init +I just committed this to branch +I will try to commit the patch todayIf you are using git you can do the commit in some small step and attache them to the sub issue +Thanks +I decided to close this issue it has been hanging around for a long time without being worked on. + +Canceling old patches +In trunk this behaviour is seen on core frameworks when we add extra dynamic imports using aIf something else is happening then Im afraid I dont know what it is and I could do with some extra contextthank you for the fast responses +This has similar overhead to the approach you propose when random subsets of fields are used +equals to +Or maybe I misunderstand +Attaching a temporary patch v for the records which needs testing +As per customer request relayed by Justin +Attachment has been added with description test file to display has been added with description image used in test has been added with description Output of have not tried it with jimi since we are already shipping jai +Hi bibryamShould we close this issue and raise a new one for the specific flex problem aboveMany thanksChris +Thank you for your comments +Checked on EAP ER during some testsManifest in libendorsed says are you seeing a JAR with this manifest versionZIP +Gabriel agree we need to check for these kinds of loops and bail out when theyre detected +I agree that jar file utilities is a better description than database class loading utilities Yes you can run the on both embedded and clientserver databasesThanks! for making the release note so clear Rick it made the documentation work very easyCommitted patch to documentation trunk at revision +rebased set of patches where all the changes from the second Sylvains comment are in patch +Yes this is optional +Thanks for your attention in the details +Please verify itCheers DaveI just ran the test with Camel which is shipped with ESB test passed so Id like to close this this is fixed in +See. +We could provide a better error am inclined to reopen and schedule for +Its not critical and I am generally satisifiedwith the code but putting in my cents in case it helps +Attached patch which removes the test from derbynetclientmats and added it back to jdbcapi +Updated patch Fixed conversion for address location map +Yes reusing existing code base would probably be best +The current resolution is to go with the new Apache CMS as recommended by infra + +patch to fix and provide keys where +The example started up just fine but whenever I couldnt navigate away from the start page +But the earlier patch made that check in whereas this patch makes that check in +I thought about it but it would not make sense for new +Confirmed on master andI think the problem here is that for definemethod were using the block formatting instead of the method formatting +for committed +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +No svn tree still where it was +I could not reproduce this in a live setupFor simulator setup can you check if the Stopped VMs podId in the database is null +bulk defer of open issues to. +I committed this. +Im more of a fan of PHPPythonreal web development languages rather thank hacky XSLT stuff but when in Rome +But the issue here doesnt seem to be that were claiming to throw a ULE when we dont its why are we unconditionally swallowing here wo so much as a warning? That doesnt seem rightEli I think the reason is that this method is unprotected +This class is also present in both upstream jars +This patch against fixes it for me +The converter is for CGLIB enhanced proxies in general not only for ones that are created by Hibernate +Well it depends on the frequency the functions are called and the length of the input +Are there plans to address this ticket in near futureI think we are missing a big step towards adding build reliability across build servers esp +That way instead of having to sift through properties files XML files and the like a web page would display the settings that can be applied to those files +See my comments onHowever I havent tried it on Glassfish and I cant explain why the behavior is different +Hi do you want to update the patch? This appears to have stalled +to next CP quite a number of issues to address not possible for this CP +This call can then be used instead of doing a describe extended query on a table and parsing that blob that gets returned +We are glad for every patch you might provide and help you to get startedRegards issue as every title should be able to be mapped to our commit message and later to our release log. +vs would create confusion btw also using to add an header will create confusion +Trivial fix +It does not have the position for sh commands +verified. +However I assume that the problem exists on the Geronimo adapter as well if so could you provide updated info +Aha so what youre actually asking for is to disable the progress when quiet mode is enabled +A collector forwards an object and pushes its reference fields addresses to task pool the other collector may get the task from the pool traces and updates a reference field +As I have noted elsewhere I do not like the second sentence particularly the is not currently supported part +Fixed +Is this the cause of the cpu usage when resuming from sleep? I think this is more than minor if were getting millions of Ubuntu users soon +This is really a documentation problem will add a KB note once the website has support for such sections. +GA I believe its jboss and ESB +Sounds good +You are right Uri using the requestContext is much more appropriate than using a +You can go higher if you node is able to handle it +I would rather not fix it myself because I am not certain of the service dependencies + +I use it like this nc l p continuous execcat Here was generated by capturing the services traffic +No error but when I update the value it doesnt reflect in the committers view in the source this messageYou currently have valid keys loaded in can update your public keys in your personal details +A lot there! Ill respond to most of it laterAs far as the stale state that is expected +Simple patch that refactors into a class and makes and interfaces +Yes I didnt see +requests be redirected back to HUB +It even seems like the error message is saying this pretty directlyWhat is your thought on what should have happened here +Initial contributionTo Install in use the following command on the consoleosgiinstall s mvnjarbundleThis requires the bundle from be submitted this patch +As its a trivial fix and not a knowngeneral problem I did not backportEDIT wrong rev +Is there something about Nailgun that makes breaking those applications necessary? Or to put it another way if failure is a binary value and all failures are the same why is it important that Nailgun returns exit status ? Why not respect the error code passed into the exit method as does by default? +Thanks Zheng +I wonder if we open a can of worms if we make blob too easy +Lets pull this in to also +do you have a test case we can use? did it work with prior versions of Maven +Merged +Its not too hard to set up and it generates pretty nice output +Thanks Ruel +This patch allows an of which causes all hselect renderers to output disableddisabled when they are readonly +then lets mark it as non reproducible +Issue closed with release. +Please let us know whether it works for you +Does everyone who voted for and are watching this issue using amazon ecsn +No visible problem with in particular the small text isdisplayed for the second SVG +Also regarding arbitrary rangesWhile using allows us to construct arbitrary ranges we must then pick them out of the results separately +batch transition to closed remaining resolved bugs +I think we have an alternative for this +When I went back and tried again everything worked +They seem to mess with debug and warnings levels of javac which may cause confusion in our testpatch program +Please ignore me +What locale do you have on your machine +The address sent to the client to use in connecting to the secondary is taken from the servers +I committed this branch +This probably should go on the wiki somewhere I will put it in the FAQ +Is this what you want? The sole difference is that such a log entry is then done at debug level instead of warn level +This is minor documentation patch for correcting copyright year +To clarify the idea here is to do a read with RR disabled so nothing in the cluster gets mutated by the read in order to instrument the number of damaged replicas +interface provides methods to cast complex types but they are silently ignored by PigSo before this patch casting of bytearray to tuple or bag fails in? That is definitely a fix worth back portingThe addition of deep casting I have a hard time thinking of as a bug since we have never it in the pastSince this is relatively small and did not work in the past I am ok with back porting it + +Maybe we can just compare filenames for now? Its an improvement and would find this bugI think filesizes are unsafe at the moment too +Also Tier is selected before configuring autoscale +To get the namespace of the service you need to look at the name attribute of the wsdlservice elements +I have a little experience with Flash if theres nobody else here Ill review it +UghMe thinks we should either rework Fieldable as weve previously discussed or we mark it as being one of the very few classes in that is subject to change between releases +Thanks Mike! +Ill update the tests to use the same approach as in +Wei Does this still work without a patched QEMU? Can we get the throttling code up for review on reviewboard +I included another patch from +But having a separate env variable for HADOOP is no help there +bulk move to next stream +Applied modified version of the patch in rev +Other than modifying the version of G from what are other changes that are required on this +Please create post about the problem on users forum together with source code you are using + +Without seeing the schema and the failing instance document its impossible to say for certain +This document will at least be the information on how selectors work with regards to destination policies. +Please create a thread on user forums instead thank you! +Draft release notes text statesThe asset position displayed when the NextPrevious buttons were used in defaultPackage was incorrect +Since CFS can decorate other file system besides HDFS the encryptiondecryption is not intended to be a datanode process or another process fronting datanode +No longer supporting jbossesb in Savara. +It appeared to work and seems somewhat less likely to trigger this deadlock but ultimately under load it still encounters this issue +Thanks merged. +Ioannis no need for a full patch +On JDK you should simply not have any such jars around simply rely on the defaults in the only thing I did to my fully working application was to replace and with and +callTimeout is set to a negative value at the first call +So currently we will have to live with these problem Multibyte character inputing is not supportted if OS encoding is native Backspace handling is incorrect if line contains multibyte characters and OS encoding is UTFInterstingly under Cygwin Jline works fine +Go ahead commit Nitay + +The rolling data app is the paradigm that the existing derby algorithm should handle very well with no need to call compressRows in derby in the base table tend to cluster by page in the time they were inserted +Ive noted the difference in the get javadoc. +The ERROR is thrown even if POST is used without any data in the request body +Benoit Thanks but thats the feature that Jan thought I want but dont +In lucene we often focus on reducing the memory footprint but even if we dont safe much here we are still friendly in terms of GC which is my main concern +the theme selector can have another id when selecting theme and other id when i look to make sure that the theme was selected +Ill check it as soon as we require this and close the issue as invalid +Well the patch produces too much conflicts after my two commitsAlexey could you describe what you were doing in this patch +javacc is just released +This was happening because of the kind of BPP we were using I have now fixed this +Patch looks good to me +If I understand you correctly even without disabling is an improvement on performance scalability? If so great and +I believe after discussion Rajith reverted this patch in revision Thus descoping this from the patch will be commited after the M release along with the test cases and patch for MINA threads +The method name change is the only change +Perfect +Ill close this as then unless someone feel otherwise +When I run my test page on OCJ these are the resultsRequest URI avrjsptemplates null +Will verify in the next version of JBTJBDS. +Can you create a test case that loads you actual application context and shows the NPE and post it here +May be the title says different +That way all is recoverablereconstructable from the master table +Thanks Max +So if users of hcatalog do not use we can select all columns for them +I tried this on TRUNK +I dont know the answer so I cant +Just to confirm this is not a bugI was able to add remap directives successfully +the archiving seems broken though its aall appearing under Jul +Verified +Gabriel Volker attached a test case on the gt issue +Added a few config files such plexus integration acegi both the plexus injection is probably better done by using the plexus annotations as the other actions latest testing documentation and missing +Thanks Raghu +This would also be useful when another app needs to leverage a data model that was produced via GORM +Attaching example application +then whether master use hlog compression or not +This suggests it might be useful to provide a switch to disable injection into overriding methods allowing the agent to avoid the superclass check +Could you please check the algorithm there for making sure that no output parameters slip it into the stored procedure call? looks correct from what I can tell +Tons of duplicates reported repeatedly in the PMD output +There was a continued discussion on the mailing list on this topic and people see value in the concept +has been marked as a duplicate of this bug +Hi WeiPlease get the latest Spring Shell build to verify it. +Upgrading the lib is one thing but backporting all the other changes that came around this is another +Consider that now documentedThe journal store will be available in AS +I didnt capture the numbers but it was something like ms without and ms withI ran in a fairly interesting case since I was trying to test the contention of the single socket and it seems like adding more sockets to a saturated network did not improve like I had hoped toCould you paste in your test scenario? I could see that for some scenarios involving small to medium gets that this could provide an improvement to latency +indexer directories and restart the server +I should have mentioned that in our restricted environment we do not have DNS and cannot have DNS +The same problem is observed on +Latest patch submission for Hadoop QA robot as Stack requested +Final fix for attached the final version of the fix for to the JIRA if anyone would want to look over it +verified by Vasiliy +c was unnecessarily merging a constrained type with another constrained type stripping out the constraint +It may be some problem with the order fields are initialized weve seen that in the past but it doesnt happen here on my JVM +like another inadvertent fix perhaps +I think the best is to forget it unless its really annoying and then we could change the xsd pathes in and point to a new location with xsd files +This bug also arises when deploying Blog example into so this example doesnt work on IBM Java +Is there any posibility to removeoverwrite it from pojo file +pitfall I have seen more than once +Wont fix see discussion on +Then the idea is propose a fix to really use that interface +adds new ctors that require version deprecates ones deprecates inputstream ctor that uses default charset reorganizes constants like standardtokenizer and deprecates the old ones +So could you please with +Default values for VPE Preference Page are initialized now +Attached is the code for the binary websocket filter +Resubmitting patch in light of Adams comments where now logging the exception in at a WARNING level +It will help us a lot if you can post a unit test +I wonder if they have a local runtime running thats also eating their memory +Please close if you agree +for refactoring the class and make it immutable with a builder pattern +Indeed I put it in the wrong place +The resolution has been submitted to the board and board has approved the resolution on September th. +It makes sense not to process the same file over and over again +Probably going to be now rather than +Tested on a dataset of K pages and the results were promising +Also simplified the assertions with a loop +This will modify the wicket specific filesJunit test in module pass but for the first time they need to be run with the true argument to regenerate srctestjavaorgapachewicketextensionsyuicalendar file +The property on connector capabilities is use +It conformed the interleaving see above cf +Mmm sorry forget about the of the infinispan module as it would depend on Infinispan and introduce it as dependency for the whole module which is the same we want to avoid from hibernate core +Yep +Again is get related but quite different from the linux s worth a shot to have him try +So working through those first +Once all data has been the migration is complete. +A second diagram this time done on a Ubuntu AMD Phenom II X B Oracle Java update +Seems to work with revision if transaction size is not too big +Also note we are likely to remove reopen if you have a repeatable test case as I am not sure there is an issue here. +Can you elaborate? This is a fairly large change and a lot of it seems unnecessary +Its really strange cause I was assuming all the time that you are workingagainst the CVS repository and my latest updates +Then we checkpoint the recovery point at TIf this occurred then we have messages T T etc which have not been flushed but which are below the recovery point +The patch looks good to me +Verified on. + +Thanks Zhao +i think it may make more sense to handle this outside the hierarchy of something like that would have its own gettersetterotherwise all custom resource stream locators would have to wrap themselves in caching decorators or build caching on their own which seems a waste since caching is very much orthogonal in this case +Thanks +for the integrated to trunk. +Add namespace to yarn protos also cause some compile error in mapreduce by the way this patch add namespace to mapreduce protos +Observed this issue in as well merged the patch to disable the test over to branch with revision +Bulk close for Solr +It looks like this was implemented a long time ago in so closing the issue +It is not documented and not tested as far as I know +Thanks Sergey the patch was applied to BTI at rPlease check that the patch was applied as you expectedBTW is it possible to automate the spec +It is OK to set a Locale in configuration for installations but using the JVM locale is a recipe for disaster +As we talked about needs to run on +Java security policy used by the repro +Thanks but it doesnt appear to have made any difference I get the same error when I click edit in the Operations down the right hand side of the page +Basically having rather than simply baking the constant into a method in class Text makes it clear what is going on to people who do not carefully inspect all the methods of class Text +Have you been able to build on NT with success +I dont think it would make harm if we include it as it was implicitly included before +Identity management is now complete in Seam and next on my list is security +The solution in the SIIDE branch breaks with transitive dependencies +Byteman guards access to the Instrumentation instance very carefully and for good you could explain what functionality your helper wants to access and why +Any problem pls let me know +In other words its only meant to be accessed concurrently after the refresh phaseWe have plans to revisit this for Spring Framework but for the time being Im afraid your scenario is simply not supported +Link to +youre right it should be singular instead of plural to just start it on this node +In case you see it only for kvm +Main menu and project matrix updated so I think you can resolve the ticket. +Patch for to as well +I dropped it in and was pretty thoughtless about it after splitting my two enhancements to thrift +Now uses the date picker. +I attached patch that addresses the performance issues described above +in fsck uses fetch and there is no use option +Ill take a closer look +that the slow writer is simply slow and the test is measuring the systems response to it +create changed to createNamenode in +Committed revision +but Im not sure if its really going to help +Fixed issues with status bar labels and menus +Thanks +Itll default turned to no but then people can set it in their confsI do not like things that automagically do things though in some cases it could be the patch as follows Added a flag to toggle onoff +Backported to +created for dont think so +Im submitting a new patch that takes into account pure POJO accesses instead of reflection +Thats it +we would like to testverify these on our Beta build candidates so ASAP since QE started this yesterdayI can keep QE of the RF archetype until monday if that helps so if you could have it created by friday then we can update jboss central metadata with the new coordinates +for now I did implement a workaround on the jenkins plugin but I still think this would be a good complement +I committed this. +I conclude that there is no bug with Geoserver and this can be closed +this is a critical issue is it not? without the sourceform tabs arent actually +Correction to previous comment +Given that the test implemented here is supposed to be more reliable it should never happen unless we run into a bug +It works find on the command line but on Hudson it fails +A common problem weve seen where the slashes mismatch +Its been a while but I seem to recall trying that +Ah great I should be able to make that work for me +I can confirm that when i set the guest and create a guest user in my OU defined in OUUsersOUVelocityDCvelsysDClocal that I now get You have access to no repositories +Just requires a mongoTemplate autowired +Ive fixed the problem with color +Im not sure special casing to be blacklisted is the best thing to do +Please when you finish work synchronize it with branch +Okay I installed to test this +Ok +Any idea as to when this issue can be fixed? Its pain of neck to add all dependencies of test phase within plugin tag in order to work around this problem +This actually has nothing to do with the project type +However I could see providing a variant which directly accepts a to optimize traversal as much as possible +Cannot reproduce it on RHDS reproduce on beta SP +the defer parameter to ChangesverifyThe patch also adds upgrade tests for the updateXXX case +But its not trival to change to use +Right but already does this conversion is what Im saying +released! +He made my year +Lets merge this all into a single issue Jay will evaluate how much a bottleneck EL is. +Committed switched to rspec and added mongrel pre version to gemspec +the is not removed +Ok I could do it +However the is mentionedin sufficient places +I think this should have a priority higher than minor +Ok the bug has been fixed +Ill file a jira and pick it up since this is something a lot of people have been asking for +I committed the patch to +A few little comments The description for should mention what unit the config is in +Test and proposed fix attached +Nothing more to do for now waiting for team released +so slingFile would be the most logical nameSooo right So we would have slingFile instead of slingDocument + +I ended up to release it internally and get my custom karaf to pick it up at our build upgrade to the new blueprint core should include the fix for this +Hmm I think I got it it seems that the jh option is broken and the launcher passes the value of that param to the java side +It only affects ee tests +Thank you very much! +Let me know if this works and if so should be easier to applyThe diff attached upgraded the voice plugin to dispatcherChecklist Passes Ant Test Passes Forrest Site Passes Forrest Run and works fine in localhost +assigning it to avery +In order for this to work we needed to be able to order interceptors created from the same factory +Means we do ignore ipv +Stan I think we should modify to create a file assigned the name with underscore prefix for a file being created so that can ignore the file rather than modify to handle a +Adjusted the scripts to account for Oracle vs Apple when JAVAHOME is not set +FILEfileAdd the word encrypted +But I am working on this JIRA +Even the estimates are just guesses as potential problems that may arise in the future +Tx for the quick review Zhijie. +Thanks! KarlI changed and class based on your advice +Problem with false is that autoload no longer works Native API access is disabled +The problem was fixed by removing the reference to which is not used at all by other code +There are some non funcitonal changes including reformatting and javadoc corrections but AFAICT they all move the code base closer to the style guide than beforeThere are files outside the scope in gadgets and commons that make very minor formatting changes correcting simple mistakes so although these are not directly related to this patch they are also validSOIMVHO the changes that the patch introduces are good and if I had spent minutes when the commit came through I would have come to the same conclusion and perhaps even commented +A is exactly what wed like +If somebody can give me hints I would really appreciate it +Icons look fine for me in both Chrome and FF +Let me see if I can simplify this +Close issue in preparation for release. +Still reviewing the change +Just be aware that since hadoops users are demanding we behave as if hadoop is think about supporting them while thinking about how to manage this +Why is a blocker +Proposed patch +The creates the service using the hashcode of the URL so you cannot rely on it for dependencies +Should we close this issue +Split into two patches. +I was able to build a release candidate as long as I dont use the directorycwipmcf to build it in +This new bahaviour is in contradition with the documentation that states that an XML element for such a field is simply ignored +Little fix of checkbox SelectUnselect All +stock splitSo for well add the logic you suggested above and for look at splits +but I can do have such a bean with no Remove or Destroy without getting deployment exception nor any errorwarn in the logs +As far as I know you should keep only in etchosts and then you should alias as localhost +This attachment contains a patch which I believe fixes the issue and has a test added to which applies to TRUNKPlease generate patch at the root of hbase workspace +But as it was supported before +This working on this one learning curve is currently vertical but Ill get there just need to persevere +What version of Node are you running? should be defined on all platforms since +The situation presented is not very common but the check for null value before get the size does not harm and do what is spectedThanks for the contributionfixed at revision +this is the same as the previous one except it returns to the client a boolean denoting a successful startstop of the snapshot thread +I thought of preparing before commit +Committed to trunk +When I found the error the people decided maintain CQL specification till it will be fixed +Ok I came up with a cleaner fix where the wraps the with another Runnable that clears out its entry in the map +If I can get this functionality to work it should be relatively simple to do the same thing in the tooling +I dont feel comfortable rewriting these to be instead +after thinking again about this maybe I was wrong and we should change the interfacesimplementations to return int and not long +Actually fixed in but we dont have a version for that yet in JIRA. +Cleaned up +Fix is now in CVS +Further as we plug the NN rpc interface it will help us validate the idempotent attributed added by this jira +Whats the right way to think about this? Can this code be converted superseded? I tried myself but its a lot of change and Im not confidentIdeally these get rewritten in terms of which just takes care of the integration points with Hadoops new + +Based on the work of Weihuas patch I have updated the schema and the XHTML specification to the latest revision released last weekThe update of a half years work seemed to be worth to be added to ODFDOMMichael did not want to update the header for the spec and schema but I did for our work to distinguish the filesFurthermore I had exchanged all leading spaces of the RNG schema with TABS and removed trailing spaces reducing the size from to kIn a second changeset I had added Weihuas four new datatypes Percent fix and changes to the codegen configurationM srccodegenresourcesdomM srccodegenresourcesdomM srcmainjavaorgodftoolkitodfdomtype? srcmainjavaorgodftoolkitodfdomtype? srcmainjavaorgodftoolkitodfdomtype? srcmainjavaorgodftoolkitodfdomtype? srcmainjavaorgodftoolkitodfdomtypeDuring review of the I found old style classdiagrams that I have removed from the repository and from twoM srcmainjavadocorgodftoolkitodfdomdomstyleM srcmainjavadocorgodftoolkitodfdomdomstylepropsM srcmainjavadoc! srcmainjavadoc! srcmainjavadoc! srcmainjavadocThe upload of the latest ODF specification in HTML will follow soon +Attached a document patch as well +If so please close the bug +I see that is is used to indicate the function is a table function +I wish I could use Woodstox however I am running CXF on +If the normalized result isnt larger than the input we can also copy it back to the original token avoiding another reallocation +I was unable to replicate it +Amar I dont find this key or the newly mapped key used anywhere in code in trunk +Id have to get handle on the actual used +Should I remove it or simply set it as deprecated and remove in? In the deprecated case How do I do this in xsd? think this is fixed already +it shows how a persist on a object contained in a duplicates an object in the many side +Ill ask the infra guy on it +included all of the jobhistory related changes needed here so all that code is knocked out to make this a short simple patch +I have not noticed any significant decrease of startup time using HSQL cached tablesI used with for testingThe first launch took ms with memory tables and in with cachedSecond launch took ms ms +In this scheme services exit from the pool early on and leave servers alone to service the remaining requests until they are then also retired from the pool and all weights and pool entries are refreshedIf this is correct then servers should experience alternating periods of and rather than a fixed constant load proportional to their allotted weight +Too early to start thinking of such have a new feature to install Sonar Eclipse for CDT +So please make sure after change the wsdl file you can generate some codes like web service interface classes +This patch catches any exception during fsck logs it and sends an error message to the client side +Do in another issue though +committed rev eecbbbecbeddaffd +unit test and in revision Vladimir please take a +It is ok to dispatch more than is sent if the consumer connects more than once and has a pre fetch +It is a slight modification of the original to include an extra private field a setter method for that field and a conditional statement to decide what tag to use before creating the Span objectJim +This is probably the time to fix this issue +This issue is the same as Links with +Has this feature never been implemented due to lack of demand or are there potential issues that havent been noted in this JIRA? I filed it in response to a specific user case where it would have been helpful +The first cut at this is done. +Another issue nowtrack this +Also ports the fix from to binmapred +Your committing of Mr +I havent looked at the patch but I thought you were moving in the direction of implementing procedures which return values +I think things are working now. +Thanks Stefan I finally took a fast workaround I made finders with half fields each and manually merged into one using a shorter name for the jspx. +Hopefully I didnt test muchYoull have to fix paths in the patch file to apply it locally +Im talking at least ms and possibly seconds +The test case works correctly in so please dont break it before it is fixed for real +That is the lesson out of this issueDiscussing about a separate getset for this posIncr stuff is another discussion for a separate issue +Yes its a bug +Fixed as part of +Is it our responsibility to maintain Mavens POM descriptors in the first place? have had this before +Fix for surefire +Feel free to change it to support whatever you want +Ryan suggested memstore +other image filesThis is due to the fact that if you specify a javascriptLocation it uses that for both the theme specific javascript location and the overall javascript location +I looked at this a bit deeper and found out that there is no other way to avoid the extra roundtrip with jdts as it uses a SET IMPLICITTRANSACTIONS statement to control the autocommit modeIn the long run I think we should move towards using a connection pool which would nicely solve also this issue +messaggi non ven apr CEST Total number of messages not yet translated fot traslation status START sab mag CEST +Reviewed and applied Ryans extras. +The patch is committed to the trunk +im not following here why should runtimes sites always be shownit should only be shown if runtimes are selected correct +can probably canonicalise the ve canonicalised the source file which is about all I can do for the should be noted that the reporter was not using despite the report according to the trace more likely dont have a test case to confirm whether this really works or not. +Feel free to send me feedback before I push on github +I just tried this and didnt see a problem +Can you attach the results of svn diff from the root of the thrift tree +Either way it accounts for a second or two during builds so whatever +Stefano let me know if you have some time for this Ill reassign this to you in that case +Please provide an update on your issue as a comment +Snapshot would map to BoilerplaterTypeObject +I will apply it on instead. +Change package name to wrong value and add some annotations +Moved to future versions as fixing this issue requires component at +Is this similar to what you are proposing? This option is not in though +including or +If you do a full shutdown then the problem could be that X number of serviesroutes did managed to stop +I also have a patch for against the Spring code that enables passing in both a custom and enablingdisabling XML validation +in release + +Patch looks fine +It seems that this issue has been fixed some time ago and cannot be recreated in trunk and in the release branch. +I didnt understand that cygwin dll is needed even for run +synchronized here buys us nothing and should be removed +What is the use case for these exact counts? We do have progress for maps and reduces +The class no longer references the unsupportedundocumented class +We should be able to change it to refer to a bean defined in the attached fragment fragmentGreeter +which feels ugly indeedI would remove the log method if its not used and is too approximate +it seemed we should not worry too much about the frequent small object allocation since Java does a pretty good job on this +Oozie should be a good candidate as well +svn is not handling the directory rename particularly well +this has nothing to do with this issue about hidden dependenciesPlease open a new case with the exact steps to do this +Is this still applicableI need some java code to see what you were stopped developing Cocoon apps for months now +I guess I should check out the Forrest CVS and check that the Forrest version of this stylesheet is still the same as the Cocoon version +Of course is a private classNo its not +If this turns out to be a problem for users when they upgrade to JDK we may consider adding functionality that gives users a more targeted workaround +It depends on the setup performed by tests which run before it in the storerecovery suite so dropcrash is only run as part of that suite +I believe all of the documents have a value for the group field and will confirm that as well +Am I duplicating +Youd need to enlarge internal buffers to be able to cope with that +How does this information match the observed behaviour? I can clearly see two threads downloading the file +Uploading the corrected patch for the trunk +I just used UNION in a HQL query using SQL Server with no problems + is priority not really working? I think there must be another way to do this without as it is not intended for this pupose but the problem is not quite clear to me +I assume this will fix the issue in which case the issue can be closed +You know my story on features in boss +trunk +Karthik high level seems OK to me +The change looks good to me too +In eb. +Is that OK? Thanks for everyones patience +Bad news release msvc build segfaults in without stacktrace VS debugger only points insideDebug build works fine +Agree with Evgeny I also tested and this was working correctly +If Ive understood changing to numeric changes from Double to Long so the error remaines as it now needs to convert from to I meant to say The correct data type for the exceptionCapacity and usedCapacity fields is +Thanks Brandon! +can you please be more specific in regards to partial ordering occurring when prioritizedMessages is set to falseThat is what is and is not occurring with the prioritizedMessages setting is set to false and what is and is not occurring with the prioritizedMessages setting is set to trueThanks +I think we can knock this one out for +Remys change looks good +But theyre useful counters +Otherwise I am on this +Bulk closing stale resolved issues +Cleanup of unclosed issues. +The goal is to have a common response no matter the spell checker so that we can plug and play spell checkers +DavidI feel your pain but it is a bit unclear to me what exactly do you want us to doDo you want to be able to set a as a property of the +Not even in the console output +Would wrapping it in a helpI assume you mean from the performance standpoint +In the method get in the token file is read and tokens are loadedThoughts +will have to fix those up incrementally i guess + +Committed +Ill try to test out the patch this weekend and let you know how it goes +It will automatically forward you to the license administration page once you login and only allow you to use the license second part of this is automatically getting the license update from this will be checked in once its integrated with the network side of this feature +Im looking forward to this feature +htaccess anyway. +For some reason this still fails both on trunk and on +Closed other comments for expl. +I tried that +Can you confirm on your end? +In any case it should be a configurable warningerror +Cant really think of a workaround +I added and back in +perhaps HADOOPIDENTSTRING is not required any more or is indeed harmfulHaving it enables running multiple instances on the same conf dir by running as different usersit may be ok for logout files to separate output from different runs but for pid files its better to use the md without the string +Ok I forgot that inner classes can use the enclosing object but they do not expose itThat means that for the quick fix above the inner classes need to delegate to their enclosing objectsIll try this quick fix next +Please check for technical correctness +FixedDont forget to turn off the haltonfailure property +The purpose is mostly to illustrate the idea before implementing it properly +Our proj use to handle the exception everythings OK +Probably worth checking back there to see what there final thinking was before putting in this fix +to trunk +HelloIm passing through the same problem while running trainnb in a single node +I agree that should be updated now but the properties are still going to remain on the Identity component for a long time to come so there is no urgent need to update examples at this time +Any work for Enterprise needs to be done with consideration for the entire corporate product offering +The are passed from the primary BP thread to the secondary BP thread +Some people might just want the fix for this +Also remember that this is arelease candidate and not the final Lucene releaseI should probably change it to required +This was already being done in +So this is not a bug just a misunderstanding +Ive added support for specifying cache regions through a value for entity collection for example myRegionNote that this will only work on HibernateJuergen +AFAICS theres really nothing related here +Ill commit after testing +Will commit in a moment +My statement about ARM only applied to earlier +Thank you Po. +Also it seems you should get input from experts on the various services before deciding it is okay to whack these scripts that were provided by the developers of those services +This issue was pretty prevalent before we had a fix for it + to trunk +Note that this will NOT capture events that have been filtered by a global Filter Logger Filter or Filter as they will never make it this far +Please Hans could you take a look at it? is how I understand it +I saw some unexpected error since I removed it but I havent had time to look in deeper +Bug has been marked as a duplicate of this bug +Please test if the change fixes your problem and then close this issue to confirm that everything is fine. +bulk defer to +Anyone want to own review this? +I heard talk of the server possibly handling this upstream from us so I didnt want to duplicate the effort plus it would be much cleaner there as it is a bit of a hack to fix this on our side just for EC clouds and to keep track of it across Eclipse sessions +This patch which does create in tmp dir a delete and rename seems to fix the failing in repeated runs +That is not valid HQL +attaching in case max sees fit to apply it to GA +It has been doing for openejb release and release +Still feels like we need to be able to tie the policy provide with the binding specific policy code but you can see the new policies in action at itest and itest answer to this was the binding wire +Admitedly its tiresome to parrot text thats already available in the and javadoc +Patch looks good other than that +Yes I also realized maybe should first focus on java compilation +If the server is fully started it shuts down properly +If you could let me know which ticketstasks I could take a stab at +Ill have a go a some re factoring to remove this wrapping +Hudson will fail since this wont apply to trunk +I have replaced the references to Hypersonic with H +Yahoo! is a Pig shop so thats the current theory on queriesBut certainly if Hive had indexes that would be a very strong argument for using it for data storage +I think this should be broken out into hypervisor specific sections to eliminate any confusion +Ive got a patch on that gets libhdfs running again lets unbreak the test before committing any additional code +backport fix +Should this be resolved +Fix tests attached +Thoughts +Closed. +I committed this thanks for the patch Mladen! +In terms of security I think its correct +Or maybe this is a dupe of +Fixed at revision +Fortunately thats a pretty parameter alreadyYes +Verified on work properly. +You mean removing all units which launch mvnAs the goal of continuum is too launch mvn IMHO we must have units which test fact I was tempted to add units for maven and ant buildsD Toughts +The discussion is at +I would recommend abstracting this code into a shared subroutineThanks +reopening to change CLOSED LATER status +HiI am looking for the same feature plugin site there is a goal site that adds the site descriptor to the list of files to be installeddeployed +While this did fix the problem of missing GPS meta data in the photo now that same data is missing from the optional copy saved to the iPhone photo album +depending on the analyzer you use there is a lowercase filter that convert all the tokens in minuscules +With Farah to see if there is anything left or whether this can be closed +We have ASSUME that works in one direction but not in the other one +the previous patches to include a fix for sync order in recovery andto introduce the ODSYNC flag if available +bulk move of all unresolved issues from to +Yes it is going to be a hard coded compatibility list +I already explained that above +Unit tests the block to populate supportedSchemes set from conf at HAS init time +And do you think this the way to go? By the time the balancer goes to act on the info itll be stale dont you think? It seems like the balancer issue that the boys were complaining off could be addressed by having the balancer not move regions it just moved in the previous balance run? Thanks Ted +I suggest add another way to set up the configuration file where not be needed that the configuration file must be in the classpath and can be location always a secure location for this important file +Thanks +This is working as designed the consumer thats doing rollbacks increments the counter to track the number of time it has redelivered the message locally +in the debug output +In fact the JMS client will always get a regular Session to work with +Yan uses it for component using this package can be simplified by moving all the scoping logic into +I know this patch does not fix the root issue but its not a bad thing to do anyway for performance reasons +Well you certainly arent using the connector in the branches branch +I retested this and it is fixed now. +Similarly deleting and editing comments is discouraged +removed the uwe pointed out this is very dangerous since we cant prevent someone from modifying it +The UDDI are exposed via web services. +We should have a CR soon +That will be unmaintainable +Okey may be this is to much thoughThank you for following up with thisIll attach a screen shot soon of what I thinkBeat +Ignore the last comment +Thoughts +Updated wiki a bit as well +A fix for this issue has already been checked into trunk as part of +Setting to future pending a patch +If we compact several TB how big can the cachedPages list grow? Do we need to change that to an approach to avoid memory pressure from allocating a huge List + please change will invalidate existing files written out in +Yes werent useful we couldnt point to the corresponding NM easilyWe should just rename the header to be or some such +Fixed +Thanks for applying this Vincent! +However at this point I would not recommend removing this code as it would probably reintroduce this bugregards Malcolm did in fact implement this by injecting a custom converter into the OGNL context +I ran the patch through dosunix to remove any in the file +Pavel it will be great if you can review it too +Heres my patch +should be fixed in Rajesh could you help to take a look at this issue as you are working it recently +I pulled the skip code from into so it can be reused in the jetty componentThe option to override the set of ignored headers is provided via the accessors +and pass cleanly now +Fixed in latest trunk source +Or perhaps it was just on trunk gonna check later +Should we deprecate the old ec stuff after this has been committedYes +Once we do we can leverage MCs DI +Patch fixes suspendenabled mode for interpreter +Gavin can you please review? +Ill commit the patch later +Thoughts on that? Would we rather split it up now so theres a distinct command +The difference here is that when Metro is used the service only sends the response to the REPLY URI and just sends an HTTP response to the client without the payload +We are working on the foundation library classes before constructing functions +The code looks reasonable to me though it seems like were already entering the territory of something thats beyond the scope of JackrabbitWhy do such external time series need to be exposed through the mechanism? Instead of registering them with the repository couldnt a deployment simply make such time series directly available to clients through JMX or some other managementmonitoring interface? +The docs have been updated as best I could for M +Move to +Committed for rev +Im even sorrier for the mess +Targeting to as has now become a release with mostly fixes and minor improvements +we switched from business theme to tomahawk which does not has this problem. +Adrea you could close this issue if you have read this +Possibilitiesa somehow the result set gets closed +Makes sense thank you +Yes looks goodSorry I hadnt realised that you were waiting for me +What version of Xalan are you using I cant use because extension elementsare broken but I have not had a problem using a parameter as a Extension objectin any of the provious version +as one more test is necessary +Well does it mean that in year when nobodys gonna remember what Java was Maven will still assume default compile level to be? Thats understand backward compatibility but its crazy anyway +I agree that the recent changes in to apply a pervasive check to all derived classes define an increase in scope and can be isolated from this one +I do not have the ability to create a JMS unit test +Now I changed it to use Create Seam Project contains source src srcaction contains class Open and push Browse +what shell we do with this issue as Werner wants to release? Mark it as resolved +I am not sure if this library is always located under this PATH but I hope that will do it +trivial patch +Remaining two have been moved into subtasks for Rob and Andre to address +I have been trying to crack this nut before but had no good ideas +Will commit now +Next step Install on an Ubuntu instance and see if the test passes there. +Compare to eclipse templates view +What version did you make this patch against? We like to modify the trunk and merge back and Im thinking you did this to since it doesnt apply to the trunk cleanlyCould you possibly pull a tree from svn and patch that? If not Ill try to work out how to apply your change to the trunk +Fixed for both versions thanks! +My expectations are the annotation meta data and deployment descriptor data would have been merged +I am proposing that no code change is necessary +Some more implementation notes The queries for map output locations concurrently with the copies +Barry I have committed a fix to trunk about that genericstrunk +Committed thanks +Pull it back in if you think different +As requested by Eduardo Zanni above can anyone supply a workaround if using +was modified to use simple string keys in +Im hoping to find out in what ways it would be hazardous and what ways I could make it not hazardous +Kicking jenkins +Once you are registered and logged in you will see the link to edit a page at the bottom of the page +patch for version +Adds a write boolean into Put +Can you try to set popserver logging to DEBUG level in +I think that should be flmyvaluebq +It would be much easier for use to work with small test project then trying to extract relevant pieces and resolve missing artifacts in your real projects +Patches contributed to Cocoon must be granted for inclusion in ASF works +StefanThis is great! I just tried it and it works as advertised +I have been travelingThis is a valid concern +My bad I read the patch wrong as removing getJob +If it works maybe I shall append a testcaseThanksLeo +Fixed in AS trunk and working on in both AS trunk and branch +Will try to add a debug message for I am a bit confused +Works ok on the latest snapshot +Will work on it but hold work in my workarea until after is out +Thanks a lot Ben and sorry for bothering about this now it works! +Sorry I think this goes in not Core Components +Maybe this is planned for a future JMX revision already? about using the defaultValue field in the Descriptor of a? +OliverI agree +So execute and check are mutually exclusive? and the dir is used when you check right? Not when you execute? Let me try this last version +This issues both appear when you try to build CRUD forms for recursive try to make your bug reports intelligible by not including speculation rhetorical questions questions about workarounds +Ill push it once Ive aligned with Oleg +But I am slightly apprehensive about the implication of this change this late in the game +Personally I think the correct fix here is to fix +Thats an aries jmx issue an update is for next would like to test the standalone implementation +I dont believe it is possible to create an arbitrary constraint like the value inserted into column A in table B must exist in column X of table Y +Patch looks fine +in the patch form to be reviewed? If this instrumentation is already out of interest then lets fget this JIRA resolved as WONT FIXED + +Fixed in vGACP vGACP and trunk +Even manual requires a suite of operation interfaces which could also be invoked by external mature HA toolsAs a between consistency and performance the block related info could be forward to BN in manners From DN or even client to BN directly or Forward from NN to BN asynchronizedAnd I think the selection between the above two together with forward info in is the discussion in this issue. +However it would be nice to see Hibernate throw an Exception instead maybe something like property X in primary key class is illegal set X propertys to +Well be trying soon however I want to check if we are headed in the right direction +Closed after release. +Patch applied see r thanks a lot Liviu for taking care +Will upload patch when finished + holding a bit in order to have aaime release geoserver beta +Can you please give me an example or any link which will give a clear idea of how to register a custom event listener in Hibernate would be helpful +Link to the issue for transport Support passing username of producer to +This had the of any optimizations made in the application servers configuration +Looks good to me +This still happening +Updated patch which does not depend on to be sorted outIt removes MIN from the Yarn client API and wires the calculation to the MIN from the configurationThe assumption here is that the configuration of the client will be the same as the configuration of the client which typically is the caseAS discussed in we should get rid of calculation after introducing and +Now that work is well underway with AS all previous community releases are +Because G has been fixed at side +Whats the difference between a List or Set of and a? Should there be any methods on appart from the normal ListSet methods? Should there be a reference to outside of to manipulate after it has been given to? Should the elements of be instances of or just pairs of hostname and port? In that case do we need to prepare for anything special in regard to supportThis issue had to purposes for me Get rid of connectionString parsing in +Well likely replace the weather example +patch applied +get should be renamed now that it doesnt get you anything +thx +File extension in Browse dialog for file should be +A complete patch containing the code tests and docs +Id like a principled but that will take time which I am very short on right now. +I should have looked at your jacoco issue a bit closer +I think I need to put here this statement in the interest of full disclosureWhen BOM has been voted for the upgrade to Oozie was proposed +The Metro runtime is initialized from a servlet context listener which of course can only load from the web app classloader which means +Jieshan Does this patch even apply any more +Committed to trunk +Hi AllMy issue resolved +bulk defer to +By default it is as it was originally at Installed. +Maybe Failed to create directory? It shouldnt mention a temp file since its actually creating the subdirectory for a recursive copy +You actually want to set it to to have the smallest possible filter. +Commit +Patched in version of Import for Confluence +The vocabularydublin are used by schemagen to generate the respective sources and as far as I can see they are not necessary at runtime +Increasing priority the same problem was hit in TCK for WS +thanks for your help. +Does this ticket cover user types that map to multiple columns or should I open a second one instead +requires huge memory +I assume that is Ok +Will test and commit +Its a bit kludgy will commit tomorrow if no one objects +Fixed in using has been released closing issues. +I going to give it another shot now see what I can come up with +Maybe we can start a job that distribute the keys to their buckets +Approved for EAP do you have an update on this one +Defer all open issues to +I believe the old comments about previous key locking did not apply to the max scan +Specifically set intersection union and difference and lots and lots of. +There are far too many components involved which are not reliable +The concern that should be addressed though is the timeout that occurs shortly afterwards +Committed to trunk at revision thanks for the test cases! +Is this reproducible withAlso try using Woodstox instead of the Sun internal parser +your right and alex was wrong +It passed successfully +Sorry uploaded a wrong path +Thanks Sijie +Once the particular revision is identified it should be simple to fix +The patch was created against trunk so is for are you able to test out Thomas patch? Im wary to rely on successful builds on my side as they would have passed for when that was released +Canceling to restart the trunk patch one more time +I cant reproduce version of Seam do you use +They changed it recently to add bit support +Your patch fixes the issue. +I committed this assuming lazy consensus +Ill be implementing this for something we want to try so would be happy to take over this issue if youre ok with that +This documentation is needed as a companion to the Argument matching topic added to the Reference Manual for +Since only occur in nested plans I was wondering about the return type for them +Unless someone else on the project has the intent of working on this we will most likely add it to our backlog here at ComcastThoughts +Yes it needs a good spring clean both to address the memory allocation issues that Gustavo has highlighted and the style spacing +Closing stale resolved issues +Will be fixed with the release of M +Fixed in trunk +Thanks Kay KayThis only changes an error message +The addition of version for new instance was mistakenly deleted +the feature is in there and now available in the ui right? There is just the caveat of correctThen is resolved but still is open +Thanks Thejas +No problems were identified on SUSE +In general I think its a best practice for portlet application portability to use the PP user profile attributes defined in Appendix D of the Portlet specification +Committed to and trunk. +Although the patch solves the issue at hand I will like to understand how the testcases that all have unsecure setup are passing without the patch +hidden packs are shown in install panel and the description of a hidden pack is shown in packs panel. +Pushed to provide steps to how to verify this issue with JBDS + Thanks! Is it possible to contribute it under the Apache Software License with the copyright statement pointing to Apache like thisYes my company has given me permission to give this away +Ive added null checking want to tell with Denis about problems with junit tests there +This required a change that effectively excludes reads weve considered for a given trial +In fact Sonar should not allow such parallel runs and should automatically fail if there is another pending analysis +better cleanup solution implementedI leave the issue open until configurable cache value per repo group implemented and unit test added +Need to investigate it before turning off reptaggressive +Hi Jacopothanks for the Input +bulk close of all resolved issues. +IIRC Invalid version means invalid for the standard +Would appreciate some assistance in verifying what Ive done +On a rather slow virtualized environment I can reproduce the issue at will +Yanbo Liang thanks for the detailed explanation yes it makes sense +If it is accepted I will then submit the html taglib patches which I have readyOne thing I forgot was how I used the action prefix to specifically identify in a forward that it is not a straight URL but an aliased URL +Pushed up to master and deployed to npm as + +The bug is still present in +See for further details. +Also the thrift version number needs to be updated +Possible to include it in the future? Without it people will be unable to help diagnose problems and cant build a binary distribution from the source distribution +prefix like x? I seem to be missing how this versions the interface +Collections of collections in general are not supported in Hibernate +Bumping the fix version to when we can add the interface +If the project does not need this library in lib because it is provided by server then server runtime library for the project has to be configured to include it +it makes a lot of sense thanks for the hint! +What is the difference between this patch and the fix done over in? They looks similarThe patch is a little odd to me +I think for that we should open a sep +You can mark it closed +For the repro that I attached earlier today I played a little with the heap size to see if it can be big enough to satisfy the default page cache size of pages but not big enough to be able to stream the LOB into memory +Thanks you for this work Flavio +Heres a deck with the internals that I started +Patch Applied to fix error Released +looks good shame that its not an exception +tested +Otherwise it registers for the notification to do the delayed connector activation as before +updated with the branch + is reliability test suite +You can mark this as fixed +I checked in a fix to adjust the next pointer back by instead of in rev +Result is on the screenshotthumbnail +Split the data into the training and testing set with the splitting logic factorized into a separate class +Yes separately they work perfectly +for the moment I still need to ship in a week or twoif there is sufficient people wanting to help on it we should be able to continue working after the release knowing how you immediately gave feedback when I changed the issue is a good sign of interest then ability to help testingits motivating +AntThank you for applying the patch +This definitely needs to be implemented as a client interceptor +Right that makes sense. +Patch with documentation with minor modifications to earlier just committed this +I build a minimalistic webapp mavenized +Do you see any issue with it +Andreas does it help when I work out a patch for this issueId really love to see a release and will do my best to help you as far as Im eligible to do so +Patch available +Thank you for your +Added inout Test +Ah right +It gets that from which is what gets a list of all servers and then attempts to grab the with the call which gets a connection to a random server but also handles all of the connection caching etc +Do we need to use as scheduler or we can write our own daemon +Btw as for your second point in my method there is NO. +We cant change it for the javadocs parsing but we can define new defaults for the annotations stuff +will update this in next patch along with your feedback on fix proposal +can you please verify +introduced a system property +I did not back port like you I dont think its important enough for thatI have also updated the xsd on the site +So this problem does not happen you run ant reports from command line with with ANTHOME set to portaltools +Ill do it whenever time is available +yes this is unintentional thanks +NDFS the Nutch Distributed Filesystem will be renamed HDFS the Hadoop Distributed Filesystem +Looks good +Can you paste the complete logs possibly enabling the DEBUG mode? I am watching this issue so we can proceed here for now +Many developer desktops are Linux x if all of the native pieces of RHDS are available in versions itd be a shame not to have a build +Filed for a typo here in +Someone on the JBM team needs to reevaluate this code +I took a snapshot of show schema into a text file and removed system KS from it so I will only have my KS schema definition in it +so I guess now we wait for the next Apache Commons release. +This is necessary for the navigation page as that one is included and only changed pages are exported +issue very nicely describes what the desired outcome is whereas the is slightly more diluted +New patch adds in assert to that there are no unrefd files +Thanks + to hover effects for the table rows +first cut at a boring decision table example +Thanks Kathey +Maxim you this New XHTML Wizard but didnt refactor the tests +sorry I missed this one +bug is still present in +I also need to know how to subscribe to the EWS channel so users can install httpd +No point is that the propagated exception doesnt come all the way out of the call stack as remoting and threading get involved by the nature of SI +Unless Im missing something it looks like its doing nothing important inefficiently +New patch added a unit the patch upload one more test to test the web service is accessible +For ZK there is nothing to get the sessionsexpire in the API +Yeah you understood my sentence I dont see the unwrap the proxy before setting the value in the underlying list? The patch just adds a blank line in the set method +It is related to the use of the new soapAction header for CMIS +Sorry I meant to have this patch in sooner but got quite busy I expect to have it soon +Now if you are saying that they have hard dependencies on each other then I understand +This change itself is straightforwardThe test class is also straightforward for as the method to be tested is public +previously fixed. +My understanding is that this would be because we do not do our official release builds with Maven +This rule is needed for step above +Changed requestsCount to s Counter classAlso added the summation of region requestsCounts in This patch is based on latest trunk code +We should agree on that new big features should not be enabled by default +dependency bumped to POI beta in r +There is an that keeps track of the count of processed Smooks Transformations and the time taken to process each +When i debug the code at loadSettings the server started normally without any error +fixed in rev in trunk rev Anita IIUC Jason has explicitly specified all the maven plugin versions in the root pom pluginManagement section +Lets target this forWe have seen to many end users in trouble due type converters couldnt load in their funky runtime container + +Incidentally the withWriter approach now also produces the same result as the long hand approach. +Robert it was your commit that changed the default behavior of Solr and I disagree with that changeTechnically I could VETO but I dont believe I have ever done a veto and I dont want to start nowInstead Ill try and be constructive by going to work on so we can at least configure it +Is it something we can fix and supply a patch back upstream? Looks like a namespace bug +Thanks anyway for your quick assistanceGerhard +Same issue two different ways of demonstrating. +I will close this one when I see the new uses its own POM model in java that can parse common elements such as groupId from both dependency and plugin elements etc +Updated +Thanks Gunnar let me know your views and feedback +This patch adds a new class that implements a trivial stack backed by an array +The parameter is +Shouldnt we be able to play the video tag in +Thats why the callers bundle is cached and used throughout +IMO it depends +This patch introduces a callback interface to remove the dependency +Case became even worse when the same situation happened on the daughters +Patch that adds token persistence in a similar manner to how it is done for the RM +We can but I may have misinterpreted our conversation last week about this +Okay I have attached my changes as well as a text file explaining them +ok youve convinced me +But I wanted to capture the work so far +It shouldnt be in the user doc +With the introduction of the is no longer marshalled into a remote invocation +but I fail to see what your comment was about then +I havent checked other branch +An article on Maven is now avaible at IBMs developerWorks site +hides documents or adds empty segments to the List of merges +Running tests +Will be done in an hour or so +Fixed in Trunk of +can you drop dest at the end of the test otherwise it looks good i will merge it after the test drop to and to and also changed src to src in the reverse test +Please give us update on the issue +My personal opinion is that its not a +Merged to trunk in r. +When a datanode receives a replication request it starts the block in a separate directory +Will work on it +s nightly pmdoc isnt templated and might also be a good place to look for common idiomsAs for java detection its tricky +I thought I could just upload my entire package as a zip file but the file limit is only mb +Tautology +I tried with the latest version from subversion but still had no luck +I already asked for support I am in the same team of a requestor +I will do that thanks +still havent gotten around to syncing to trunk to try this out +I have no idea how to generate the metamodel classes in maven +Are there any news on this? Just wondering if we can still fix this for +Not sure if this poses a real problem though +The intent of fixing listStatus and globStatus to resolve symlinks is about maintaining compatibility by making symlinks completely transparent +Weve had a short discussion in our team about this issue andWe think exactly about something like Tom wrote +Closing but removing silly design from which caused this failure by patching +yes but we would like to finish tests on one of our bigger projects first +Verified +I will look at what other tools I can cobble together and start adding commit revisionslinks as comments in Jira. +Now I know there are at least one more place where a specific xml impl is referenced this time through direct invocation so it becomes a hard dependency over xerces +Done please review +Another way would be changing the strings in +It is by design map is a collection of pair there is no notion of field alias +Let me double check on my side +You only have to look at what is happening in some of the other projects to see what is possible with it. +A minimal web application to demonstrate the issue +So you can see the POM in the directory listing but you are not able to download it +forgot to mention if I run the build so that it doesnt run the unit tests then it succeeds but I get a similar stack trace when I start the server +This issue has been resolved for over a year with no further movement +The second is that devs use some sort of build tool to concatminify their JS into one or a couple of files at deployment +Is this Issue a Blocker or Critical priority for EAP GA? If yes +Ramnivas can you please have a look at this for? I forget to mention yhat im using LTW but spring tests suite precompiled by ajc +Testing should flush these out +Stack I think it fails of the time +cleanup done +This patch sets the correct permissions +Ning I didnt mean to sound negative about this +Unit test +svn ciSending wrapTransmitting file dataCommitted revision +Checked in thanks Joseph! +That solves the problem + does use a shutdown hook +Assigned back to Ted since he is still looking into it +I didnt realise you were talking about binary compatibility here +Handles longfloat values with lLfF at the end in bytes bytesAdded unit tests to +An alternative solution would be to add a flag to the dialect +So following that model I am going to have control the admin rather then just using +The messages mentioned by that issue also have argument names +The first fence is used to get the cluster information +That is an important implication we should think twice about. +I am going to let it sits for a while +Thanks Chathura for adding all the exception handlers I applied the patch to smarruDate Wed Aug New Revision by code walkthrough. +Same patch again though this time with updates to srctestresources and the DDL scripts +ok we should be able to give this a try +Marking closed. +I would prefer to keep this new feature for Karaf and +This is a +Builds currently are effectively broken as the tests are silently failing as the appropriate jar cannot be found +I am going mad here +Hi Delos applied your and it fixes all remaining known issues when errors occur on the GEP side during startup of the server +This should just be a matter of executing with the U option to ensure that the is U clean installon your plugin project should do the trick +Mainly I did this because changing behavior of seemed like it might break anyone who expects null to mean no tables +In order to convert INSTALLPATH to an URL use an own approach or try using path names directly without the URL notationFor one new approach could be a dynamic variable assignment with regular expression filteringreplacements where you can replace the space by an The suggested solution is furthermore if there are really more people which want an convertion of path names to an URL Id rather like an explicit expression to do this instead of using the Microsoft notation. +Hi +The performance improvement was just a few percentThe benefit of the patch is that the code is a bit cleaner but of course that is a matter of taste +We are testing on a Windows server +Thanks Nandana for applying the patch. +Im reluctant to commit this to if it slows increments radically though it is correct +Reopened to add changes toMax please review the patch +Oops I checked svn status for the test code and the two tests that cause these compiler errorsnever made it into the trunk so please ignore this +The work I am doing on the dependencies has received very little attention or feedbackShould I spend any more time on this effort? Am I tackling something that is too bigMany thanks the work that you are doing is well received sorry for not commenting more on it +Closing resolved issues. +Ok so the case wasnt that difficult either see revisionThanks Rob! +So is it enought to check chek project class path and suggest user to remove aj jar to solve this problem +For example if you think that a client application wants to cast the Statement to then the package must be exportedExporting all of the packages gives client applications the option to directly access the classes in those packages +We used to use ftp but people objected as it was harder to proxy or perhaps their environments support proxying ftp +Ahh +Minor problem with the test case fixMaybe you forgot to copy it from dist cp test testThat should be cp test testnote the additional i in the source file was fixed a while back thanks! +Patch removing code interfering with standard Maven dependency resolution +please resubmit when tests pass +This patch demonstrates how to split up the config files using just XML facilities +This issue seems to have no possible resolution and were happy with the current inlining parameters +Closing the issue. +Id like to start the patch on trunk and then immediately merge to after commitThe problem was that the tests attempted to determine a testing mount point for by taking the testing path and trimming the first segment from that +Is it possible to hide the type check somehow? Maybe a static is method onOtherwise looks reasonable +Hi Freddy this would help to detect filters that might be obsolete or need an update +Committed to trunk with revision Closing issue. +Not enough cycles postpone until later. +I started on it but you can believe me when I say that this is far from trivial +There is a DTD in conf conf thats great +Jacques would you like me to take a look at this +OK since changes are all made in trunk I reverted the changes in branch at revision Thanks +Can you add details about test failure +Ive committed to trunk and +Reviewed the patch looks good to me +hi rollback should be normal in transactions +I want to +let me take a closer look if i am not mistaken +Marek I agree that is the real bugbut not allowing outsiders to build meanwhile is a bit harsh +Im against only changing it in the packaging +looks very interesting + looks like a is not a bug because JSP doesnt support facelets libraries +Done and done. + +This has been verified with maven built from svn revision and also the dependency with provided scope makes it work on +Should be fixed in the latest RC. +I dont have JDK installed any more +For the moment MB seems like a nice one +Will upload here when its been vetted more thoroughly +Ill skip these directories for RPM installations to avoid such false positives +Can anyone explain whats up there +The problem is that this sorting does not occur if startindex making the the pages for startindex inconsistent with the page for startindex and corrupting the data setA typical use case is that a client may ask for count and startindex then count and startindex then count and startindex +Need to support foreach for local and other providers the fixed issues. +Nothing else caught my eye as particularly troublesomeKathey +Observer unit tests pass locally +Errr +this file contains the described example. +reg and +Once the patch isapproved by at least two committers itll be checked in +The profile default activation has been restoredPlease refresh the Project Examples section in Central or restart your workspace and check again + +NET Framework are you using +The webdav location needs to be owned by apache because we are doing access control via apache and an if you use a webdav client to access the location and supply your crowd credentials you should have readwrite access at least it works for me and I know it has worked for the Actional client are you using? Finder? Finder sucks for webdav +Ill move the files in question from svn to jira in a few days so that we will be able to concentrate on the first apache release of pdfboxFinally I can see the light at the end of the tunnel +must create window name AND set the content type getting it from somewhere somehow +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +I have committed the AOP stuff for so you will have this feature +Note After fixing but before committing I did run a few quickstarts in both ASJava and ASJava to make sure there was nothing blatantly wrong +Note both product and extras sites builds fine locally w Tycho +Committed to and trunk +bq +No response assuming ok. +The problem was in misconfiguration of test caused by the forgotten management action outcome checkThe evaluation of system properties works nice +Any volunteers to help me get it in? I think its ready for commit +Yes this is by design behaves just like an implicit dependency expressed through a bean reference for which we apply an appropriate startup order as well as shutdown order +I have just committed it to trunk +Maven was built assuming no intelligence on the repository side so it exactly follows the http standards wrt authentication +The linked JIRA is resolved with jUDDI which is incorporated into +The specs actually run with and but everything is stubbedI added a spec to make sure no commit is asked if the version didnt change +Moved out to move back to if there will be work on this in the next months +Package content new images needed by this visual improvement patch applied thanks Adonis. +Fix for as described in previous JIRA entry is checked in. +First all blockers criticals for EJB must be resolved +Verified that this works on Bash +This PR fixes the issue related to the Seam runtime +Thanks Zheng +If user wants have seam model Seam Components view and CA features in projects Seam Runtime isnt necessary +Seems like we will have to implement the original proposalThanks dont see a new feature request like this as a blocker in this project +When I just now made a new suite and created an update site I got for my sample module as expected and the XML was valid acc +Thus the status resolved is misleading and should be rejected or wont fix +revision in build +Its true that the preemption is disabled by default but theres no need to complicate things +sorry I overlooked that oneso there should only be issues with values where is somewhere in the string +Ill have a patch shortly +Incorporates Muirs feedback +A timestamp does provide uniqueness but I have found that when users start intentionally mucking with timestamps problems arise +The new patch looks great to me +website area creaed +bulk close of all resolved issues. +No at all I think its fine as is +Tx +When using a put the shell default to using the pwd +Generally we do releases when someone says they would like one +Issue closed with release. +most of the time i run this test it returns instantlyon the other hand about of the time it takes minutes to complete +I just sent an email to Caucho +Also I tried the patch for +the problem here is you have an empty LHS +Zebras test case for this feature needs to be added to the branch if and when this feature is to be supported therein +The transition to SVN was really +There was an error in the example and Ive updated the caveat with a more specific problem and suggestion +Yes please update affected +accomplishes the end goal only call to the couchfile genserver in order to commit the header. +Thanks SaschaYour patch is in trunk at r +Doesnt change the scope of the patch at all +The AD authority connector also has this restriction +Patch to fix the issueFolks more familiar with the preprocessor code Please have a careful look +I believe this issue has been resolved in the newer versions since I never encounter the same problem again +trycurl XPUT localhostconfigcouchdbosprocesstimeout d and see if that helps +works! thank thanks Mihai for verifying this +from me right now +James created the stuff in to support tickets +The WSDL and complete WAR file have been attached so the error should be easily reproduced. +Alexey please verify +Closing all resolved tickets from or older +passes it DRLVM passes after applying patchI believe this issue can be resolved now +This patch stores an array of not intercepted method hashCodes in our proxies +The changes were applied to files contained within the release +Ive replicated this issue on mac and it looks like it fixable because if JBDS started by shows eclipse icon without any problems +In the meantime the samples serve as a reference point for best practice I guess +I think it will work for me but Ill have to try it out and play with it a bit +fix a error +Its a very good proof of concept though +I am reviewing the patch +In terms of apis as I previously mentioned I stronly recommend you start using the Hadoop Pipes apis and enhance it this will ensure compatibility between Hadoop Pipes and HCE again please consider moving the sortshufflemerge to Hadoop Pipes as I recommended previously +From and optimally users should be able to say which root directory they want projects downloaded by an example goes toSince the standard Eclipse ImportExisting Projects into Workspace action doesnt enable a user to change the destination location I suppose we also dont have to do thatIf you still think differently please reopen this jiraSome of project examples are implemented as zipping up existing projects but that does not mean they should be fully treated the same as import of existing projects doMany prefer to have their eclipse metadata separate from their eclipse projects thus we should look into providing a way for user to say when the project is importedcreated which root directory it is to be done inBTW I have chosen a location to unpack an archive since me doesnt provide importing a project from any archiveA temporary location inside eclipse workspace that gets deleted after succesful import +Hi Uldis could you attach to this ticket your sample Ant project to reproduce this issue locally? had the same issue the next build it was working again though +Hi brett okay now you have me wondering if i should split my qalab plugin into multiple plugins rather than simply one plugin with goals and reports +The change in from self closing tags to closing link tags this has no apparent purpose so I think it shouldnt be part of this commit +Why was this added to requirements? I cannot build astools locally anymore +Attaching patch with additional unit test to catch this condition +Is there a patch +Ive also updated the version of included in the example to your latest version +previous Calendar initializes its fields lazily in the implementation it might be not safe to get the hour without synchronization even if we never update its internal time +I am committing this patch since there is no memory leak for existing RM and this patch sets up the framework to perform failover +If I try your proposal I get an error message from getThe configuration file URL classpath is a a malformed URL +Not sure why the previous patch was failing on your machine it applied cleanly regression passed on my setup +Here we are adding a Java procedure which does a commit and then returns a resultset back to the caller +Well there is only one admin account which we currently define in using roles +Hi When is the Teiid bug fix release is scheduled to release +I can take a look at this +However i have about or so i can see if i chnage through the projects as the same ROOT admin userPlease try it in your lab to confirmThanks I cant reproduce it on the latest +example script that installs EDS +I think we could change the loading of editlogs a little when the NN starts in stage of Standby to resolve it +Is it because these are flaky nodes +The fix has been committed +reopened since the spec now contradicts itself +Do you mean in the upstream Facelets at +Would you be willing to provide a patch that deprecates this set of code +Minor fix to the PaulYour patch is in trunk r +this one works well it started up and I could create someWhat will be the next version that contains the fix? orthanksMatthias +Let me create a different related issue +When I debug into it I realize that rn is the length of the chunked transfer coding +bq +Attachment has been added with description Version with cyrillic encoding in AFM files P open bugs to P pending further priority for bugs with a patch +Igor I disagree +What does your configuration look like +Thanks Michael +Thank you Jing +Added a patch for this changed two tests to add a beginning to the path changed the so that we use the default t fortypes +Corrected licesnse agreement +I didnt notice before but I agree and am surprised its not already in the standard format +Without this patch it is impossible to write a unit test for a class that uses a deterministic random generator created by +The timeout is unrelated passes for me locally +Never heard about it +Committed this + looks good +This is a huge difference +Defer all unresolved issues from to +is now using those new daemons instead of Daily will process a daily compaction only when all hourly would have been done +two cosmetic changes removed redundant import in +Ive changed installer and restarted the build +We could close this issue if the test does not fail again +I think having a JDBC test for this is good as it has already brought to light a serious bug +trunk builds work with the on both JDK and JDK +tested on my Windows XP with SP in virtualbox +So what if shutdown wasnt perfect +good point infact it will break our existing tests Internally facebook uses some tables with column name extendedSol can we change it to show tables extended for now +Balazs thanks for your comments regarding asynchronous HTTP +Im interested in seeing your subclass +I suspect it is related to Nahis changes to support streaming gzip +Verified in JBDS +as reported in environmentGot the full version string +Where is the line between a support question in the forum and a bug report +As I suspected this needs a bit more work to handle the case +If I change both and only changes to Foo are picked up but not Bar +Brings about another bug +This contains the required changes to return the permit if creation of a new bean instance investigating this problem I saw that if a is thrown by the bean the is thrown directly back to the client which results in the client catching a the file contains changes that wrap any checked exceptions with an as well as the runtime exceptions that were already wrapped +Have you forgotten to upload itAlso can you commit the rebuilt jars to mavenjacorb +At the risk of stating the obviousI would suggest going through the chore of converting the page base to +Heres the patch +Can someone please migrate the mailing lists and SVN? Thanks +Duplicate to +So now its possible to have VTSVM test results in commonly used XML format! Its definitely big quality improvement of the test suite +The patch I attached doesnt have the new files I added they are stuck on another computer +The size is calculated on demand for consumerscommitted in more thing just for completeness the properties were not marshaled by default with the vm transport so they did not appear in the size calculation +Ive done some work on this and Ill continue to do so +Please try these fix and unit test +At this point I dont know if this is worth doing +The only simple interface to Oozie is a proprietary GUIWhich Oozie GUI are you talking about? Can you provide a link? Id really like to check this out +Is this still an issue? +Move the abstract entity base class to another package. +Or a whatever +I think Uwe nailed this one + +Closing released issues. +Chatting w Jon Gray he actually needs +JSF ties its views very closely to its validation and action facilities so there doesnt seem to be any point in trying to support this as a pure view technology for Spring Web agree after looking into this in more detail it is quite difficult to have JSF fit in as a Spring MVC +are in the scheduler and should be treated as such +Finally the solution from has been integrated. +Well thats the point projection should fail for all roles basically these are not valid protocolsMaybe youll want to have this as a validation rule and do validation before projection to simplify projection implementation +I believe that the is pretty heavily used these performance increases could give many users huge gains +Manually test added +Could perhaps also happen if you have a verylong table with headers or footers +Committed thanks Jayesh! This documentation is way overdue so huge props for jumping on it. +A lot of cleaning into Pos components and some italian translations + +No I dont see good usecases besides being simpler it should behave the same as the other implementations +Since clients initiating HA state transitions must already handle the case where an RPC to the NN times out it doesnt seem worth it to do soIn the patch attached Ive removed the delayed shutdown code and instead just shutdown the NN immediately upon failure to fully perform an HA state transition +IMHO the Deploy Plugin should not be concerned with SCM stuff these are two different domains +I added the validation if there is no protocol provided for the url +Since the original patch was based on unrefactored code the changes will be easier to apply now than later +Yes you were missing dependencies which are in the +I just committed thisOwen when you converted things to use the Java enum you also threw out most of the javadoc on that class +patched to resolved while testing for bugs. +Passed mvn clean verifyI am using antrun because otherwise I was not able to access the property needed to locate the hadoop jar file +Sorry I have built showcase with older librariesI have also verified that both mentioned issues are OK now. +disable processortest +I ran the failing tests on my machine on a clean version of the vectorization branch without my patch +if not let me know then i can make a known issue for tom +Committed with suggested improvements. +Rescheduled for +So can we have a switch to ignore the mask then? Or do you want me to just closed the linked uDig bug as wont fix. +Committed as r +In JBT CR I had connections to the default server to custom servers +As you mention if folks really have a binary fragment they can use the java or some other API but I dont see a shell tool use case for this +Yep done +Well that should have been obvious +Committed revision +It would be better if the type of update status strategy was configurable + was closed without fixing it but this error is really annoying +Well fix this issue with the proper dojo version and an updated facesIO +In order to ensure the minimal change of the original code and keep the add additional catch statement to catch other except the +Sure but I think this is a source of confusion +Updated summary to reflect that this isnt specific to HAR +introduced a function to do both check and record and used the shared logic in regionregion +EmmanuelThe entire point is that you could have both proxies enabled and casting working perfectly +Forwarded request to George who did the original to Utkarsh since he spoke first +this will go in a future dont know why this never got d I think we got derailed be the conversation near the end +This should in particular make sure that are properly converted into checked exceptions +It wasnt in my test cases so I guess Id just stopped looking by then +back ported fix from in CR. +I wonder whether people have abused this in the past to recover locks they lost +Sort the gbeans in the configuration by references and dependencies and use start not startRecusive methodb +Once the distcp is complete then the snapshot can be removed so the space usage of the tmp files is only temporary +Never mind +I changed the add methods to add and made package private final and removed synchronization since it is guarded by during creation +Please test it +try with version does not seem to be the workaround +This patch moves all memory allocation for initial thread structures from child thread to parent +Youd want your unit test tomix and match Document additions and deletions and make sure it all still worksproperlyThanks +Look how hard it is now +Attached rather big patch cutting over all API usage to flex and removing the back compat layer that emulates API on a flex index +It looks like this was taking so long to complete that the scanner timeout problem was happening +So all issues not directly assigned to an AS release are being closed +It was only considered to be a thread safe class before because it did caching +the test would let us determine if we support it or not so whatever you can contribute that will let me see the issue would be a great help +But the test should still clean up after itself +If you or other users would like to help with that that would certainly be appreciated +Perhaps my other issues were also created but the index was not updated so I didnt see it which would make sense if youre! Ill try again laterBestAlex +Can this be counted as to commit +The first issue has been resolved +also a component may register itself via the Initializer conceptagain eg when is in the classpath the is added automagically so users dont have to bother registering widgets manually only to get compilation errors when the jar is removedit should still be possible to turn certain widgets off +The test is still failing as there is a problem with handling enumeration and extensions +We may need to look at the XML mapping parser as well +Committed to trunk at subversion revision +Si Chen suggested that we should returnFailure instead of returnSuccess in that case +The documentation does not seem to need any modifications +Its a simple test with usage +I have tested this and it works fine +I call out to hadoop binaries to interact with HDFS and I want to capture all the details that occur when I talk to my real cluster + +it writes its own constructor +Made the changeSince this is a patch around main didnt find a straight forward way to do a unit testOne manual test +I can likely erase it by removing the unnecessary connection contextI doubt it the unnecessary connection context is in the noise compared with a networkrpc round trip +This mustve been fixed by the recent set of robustness improvements +I cannot navigate to another page either +Try upgrading to the latest Ruby or. +has been removed to avoid conflicts with the version contained in the release note page section +May want to take out the in my patch +Do you think that is better? Or would you rather have upgraded to only to find out that your code no longer compiles and have to instead research what changed +ok in minutes youll have new version +I assumed that the change to was accidental so I did not include it in the commit +I think this should be fixed in before release +some indications Im getting right now say no +Committed trunk revision Committed revision Committed revision after release. +There Im going to add a default ctor to IW accompanied by an init method the extending class can call if needed +Buti decided to not complicate too much my explanation +A decent workaround is to either not use or to make sure that none of those exceptions are skippable +Resolved in the above pull +If capacitymaximumCapacity for one of the container queues is this capapcity up for grabstealing if other queues need it and this queue is unutilized +Getting new updated QA reports that it somehows magically works in EAP was richfaces library updated to in there by any chance +Committed revision Thanks Spencer and CraigIll keep this issue open for a day or two in case any problems show up with the builds or until we can make the necessary changes on hudson +The latest patch didnt apply cleanly so I generated a new one +Cannot reproduce with cases work fine +couldnt make it for +irrelevant for the aggregation case as it will be last due to the goal being executed in a separate task parent should come first but not incur cyclic deps. +I can try to clean it up but it does require a bit of hackery +Thanks Patrick! +Reporters of these issues please verify the fixes and help close these issues +Or I can omit passing the property values as part of exception message +in hive the fix version is set after a patch is committed +The real problem is that iBATIS ignores the encoding issue altogether in parsing because we only accept a character readerI had a quick look at Xerces source to see how they autodetect the encoding and their method is quite different than what you propose +In general users would setup cloudstack over a particular type of vswitch standard distributed vswitch +This looks good +So we have updated jQuery Mobile to and jQuery to +I have other jiras that have metrics implementations +most recent update is kind of vague and hard to trackmost recent commit is essentially the same thing as the openedAt stat from searcher +a few fellas are seeing this now +Why adding hostport syntax instead of supporting standard notation for URL containg IPV address? Thanks +Ill let others decide this one +Verified status of this issue sure I am not sure about this issue maybe still need to be openedThe patch already here maybe need a verification on the codes. +bcca on master branch thanks! +Surely any code which parsed these names from the strings will break but thats exactly why we hereby introduce xxxID classes +patch applied +Patch that adds dropUlog boolean to merge command +Verified +lets see how hudson likes it +reasonable request but we need to wait until we dont have any open branches that we plan on merging back to the trunk +Hi Michael and devsI attached the latest code for the project and hope your feedback on it for further improvementsThanks. +The dataset or the full log or what the in the confusion matrix is. +Yep working fine for me now thanks Costin! +resubmitting to hudson patch queue +Thanks for the fix! Ive added you as a contributor to Giraph +I like the change in +This means that socket factory code needs to call back into G +an hour before I did the update it was working fineNow I am probably wrong about why it went south on mebut it is a fact it did and I copied the files from the latest SVN for ver I am using the same that was working until I updated to the latest versionI make a copy then replace it after checking against the one that I got from the SVNSo as far as I am concern something is not right +I am seeing this on RHEL and Mac +to +OK I misinterpreted your comment then +Greg dont forget to document how to build the distrosinstaller on the internal Wikis developer tipstricks page +I see a lot of factories and then no changes to actually support +Looks good +jsfJSF Faceset is added to ProjectTested on JBT have we reported this issue? This is something we should push for to get fixed if it even fails on basic jsf wtp projects +on patch v +upgrade for me as for Marco +Post the test results of and against trunk after the patch mvn Dtest test HTTPlocalhost etc etc testTest set Tests run Failures Errors Skipped Time elapsed sec mvn Dtest test HTTPlocalhost etc etc testTest set Tests run Failures Errors Skipped Time elapsed sec +I really think we should spend more time on getting your test case to work in less intrusive fashion +Ive tried to fix some of them but it seems not always possible +Thanks +Have fun +Committing this thanks Maja for the very quick review! +I like your idea Gilles! It would be clean and very flexible +Merged a fix for this changed the grammar file +i am writing test cases for this jaxrpc api classes dont seem to be there for versio example i am not able to find interface must extend +I believe it was that started doing true exec on all platforms. +Using Eclipse Version Kepler id bug in original fix +invitation a lot for your helpI did not see any invitation in my email boxCould you try again +overall Document it now mentions that installation path for hod should be same on all nodes Document Mentions that special character such space comma are not handled by hod as known issue Document Options mentioned in are also present in Default confhodrc Space between comma separated values for has been removed +Based on profiling copyFrom and to are some hotspots +Should we close this one as duplicate as Im not sure we have a good fix for CQL +Ill leave it open but not entirely sure how to really implement this +Ive uploaded as a patch which should fix those bugs. +Attaching latest patchThanks latest patchThanks latest patch +Replaced span with div for emptyIcon place holder +values in a Map + +PS Please run ant precommit Youll want Mark Yonik on your side hereOh I dont mind if he runs precommit or not I dont think contributors need to +Block with one replica is still bad It will be nice to distinguish that from the rest of under replicated blocksComments +The best and recognised practice is to using Strings for all of your form bean properties +I certainly like Nathans suggestion from a simplicity standpoint +And some kernel tests fails on Linux because of +Did that solve the problem? Can we close this bug +Item is harder than I first thoughtOn Linux we can ensure that the initialisation routine is called exactly once by using pthreadonce +I will be out of the office till April th +When invoking the PASS only indicates the events are successfully triggered +Documentation is now up to date +I hope to have results by the end of the day +Closed. +It will help get this issue resolved if you can come up with a test case +And the error is clearly due to you are trying to engage a module which is not available in the repoSo I will downgrade this issues ThanksI dont know what is happening with rampart because I have the mar file in the repositorymodules directory and copied the libraries to lib directory +Thanks MahadevBranch Committed revision Trunk Committed revision . +Just now noticed its an improvement. +The version is very relevant +big thanks to Bernhard Huemer for fixing this! +The patch for itest was applied under r +if we can add a new menu there itll be possible to use a custom propertytester thatll check the project is right mvn clean package is the standard target you run with a warear you dont need to install it into your local repo but onto the app server +did you have something spesific in mind? Id like to investigate further but I still dont have a good enough overview of Pivot internals +fixed i will try to add a testcase for this issue to my testsuite +This also introduces the possibility of errors which seems unacceptable for this APIIll go the other way and just fix handling of the caching because yes its going to be quite wrong if the vector is changed + +That leads to wrong result of edit page permisson check in the +Assigning to Mark for discussion +Proposed patch for this issue that addresses the cleanup code in the Proxy classes +Implements the FTP notifier functionality +Anything thatchanges during initialization should be taken care of by the normal +LOCALPERSISTENT is in both. +I upload the patch to Review board yet +Hi MatthiasIm so sorry but i didnt see the files in CVS system +Actually Im gonna sensor myself for minutes I forgot to look at the issues +with the modest change as described above + patch was intended to help debug the failing CI job however we resolved that with so Im canceling this patch as its no longer needed. +This can be closed now +Original fix had a nasty flaw +Added renameing of files to when relevant +Thanks +What do you think? Do you want to file +Can you repeat the test enabling TRACE for package and attach the log file to the JIRA +I also ran the test case on OSX with Java and cant dont get a deadlock +Hi JonasI dont see the IONA Fuse server types when I installed the latest tooling +The test suite +If we need to create these then create these we mustHere are the files that I know of in our development tree that need to have substitutes created +personally I would use the mvn on top of it to be on the safe side in the long run +were do you get that error? I do not get it on a clean install +Unfortunately for us we are interested in browsing the dashboard using several filters at equal frequency well at least Id have underlined the fact that this feature is not useless for everyone + update of issues with fix instead of the correct +I ran into this issue at work whichis why I suggested itIn your note you also suggest the following features +Attached test file that should work once pessimistic locking is fixed where it acquires lock in another thread blocked after doing so +Ive rebuild my patch so that it will fit in the new package structure +I am a newcomer in the world of databases +You can even add some SNIPPET comments in the source codeAnd then have the wiki slurp the source code from SVN +Please open new jira issue specify version of meclipse attach your Eclipse configuration details from Help About Eclipse SDK Configuration Details and test project that would allow us to reproduce +I attached a patch for the changes described in the issue report +Added test for closing filesystems after localization completes +The attached patch includes a new version of the optimization routine which involved extensive rewrites of the class and major changes to +I am thinking if this JIRA has to be in rather than here +Following the idea of test driven software developmentI have a test document in mind that I can not find and have to ask back tomorrow +The solution with the Filter has the advantage of being centralized +So the intention here is that the redundant constraint be forgotten as far as I understand +They have sponsorships sign ups barcamp attachments etcThe WHOLE IDEA was to get away from that and create a simple name for gatherings of people +Thoughts +Can you give the latest nightly build a try and see if it fixes the issue +Attached the patch for the and supplied patch includes code previously submitted as a patch for which was not yet included in ARQ if this patch is accepted and the issue closed could also be for the slightly funky patch and thanks for fixing it upI had this code originally in a completely custom alternative to which weve used and tested extensively internally particularly the content type selection and the gzip compression featuresI hand ported the changes to the Trunk code and then realized my working copy was out of date I then did a svn update which got me a ton of conflicts due to other patches re parameters for SERVICE calls and in trying to resolve them I lost a few changes and put them back in from memory and obviously my memory was not perfect that day +Looks good to me. +ok I see more documentation there on the possibile customization of the m ok to put the doc for the new settings +with earlier version i am not sure +This can fix the bug that the diff list may be lost when making an snapshottale +Im setting the target milestone to build but well most likely have it done before then +I fixed the file name +It turns out the finalizer has nothing to with the OOM issue +And here is the patchNote that the converter will now throw a on read or write if no atomic type converter can be found for the element class name +Thanks for the patch Alexey Kovyrin. +Ill take a look +Ive added a method to that can be used to override or customize directly the +Sascha BilginI thought a bit more about the default position +Bulk closing of issues that are old and have been released or marked invalid. +Hmmm so what do you mean? Theyre not configurable then correct? Not even the maximumpreferred size +thx paul for pointing this thanks Paul +This is a common JSF problem not a itself +maybe a more general touch command would do it +Hi Markextract the into a single patch is no is not fixed +Finally the patch also catches any problems while parsing the stringWDYT +Have you tested with newer versions is pretty long in the tooth +They are the ones who are masking an exception in annotation processors and it wont even return a failure error code +OK cool +Id actually like to use this myself +This is kind of working now but removing a tag seems to be unresponsive +already fixed +Here is a patch that makes sure to throw a fault whenever success was false even if the operation had no response +I have no idea how to solve m using test but have not found a way to removerelease theitem from locks without the test to fail +Looks to me like its trying to load the off the UCL +Applied patch in r +Still the plugable stripper is on my list and Ill try implement it soon +Touches the following filesA javabuildorgapachederbyBuildTool carved out of the pamphlet generatorM javabuildorgapachederbyBuildPamphlet generator modified to use the new tool +Add and to the list +For all browsers +In my launch configuration I had set it makes the program work +Seems like a great idea +Reopening this wont get anyone to work on the problem unless you are providing a patch +Unfortunately there is other debug code I have in there so it is not so easy +Initial work available with commit eefbfeccbccb +jbpmplugins +My vote would be for not checking in any GPL or LGPL licensed libraries at all +Release audit results are unavailable currently but unlikely to be due to this patch +Changing the cache attribute from true to false should not affect the data being displayed +Added patch to fix the issue +This way the integration with the existing urnjbossdomainjgroups subsystem comes for free including the appropriate logic that would avoid the above issue +Can you add a test case similar to +Added foreach will be a great help for IO performance +Is that correct +Here are warnings about the areas that are going to becleaned nextsrcblockswebdavsamples +Also I had my old benchmarks and I was getting anywhere from a increase to increase on YCSB loads depending on the testcase +I have removed all references to lists of supported databases. +Hi guys I have tested the patch it works wellregards Malcolm all resolved issues for Engine release. +They wont compile now but should run without issue once reference to are changed to +Agreed +I should have marked this as a Blocker since there is no way to prevent Maven from merging these attributes into the final downloaded SVN in an attempt to make a patch for these issues but I had problems with the server and cannot afford the needed time to make a fix +Yes is a subclass of latest patch works great for me +Committed to trunk +If someone really wants to do this at the level then they could use the global config option and register a single instance that has its own methods and lifecycle for providing a +fixed except and am closing the issue +I just committed this. +James has fixed it now +In general I see the solution as an artifact of a poor deployment architecture driven by the limitations of the Jackrabbit architecture +I am also seeing this issue with Groovy Maven and have tested it and it works as expected +Yes I put a blocker to the ADB codegen team +I included the serialization test to verify the behavior of getsetBest regardsRichard +I rollback my commit and reopen the issue as long as the issue is not reproduced +m +no +Many thanks for the information you provided +Now append is also called on the same file +mvn clean +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +In the CustomerSupplier view profile screen it would ADD the Tax Infos Rates Shopping Lists Fin History and Product Store Roles tabs to the menuIn a similar way a human resources application would extend the party profile screen and add the Party Skills Resumes and Employment Applications tabs to itAlthough the party manager is being used as an example here Ive seen the same abusesmisuses done in other applicationsI dont mean to throw a wet blanket on this effort I just think the end result of this issue will encourage even more bad designsI havent tried Scotts patch but a cursory look at it gives me the impression that applications can change the widget models +Moved patch from +Bumped up to Camel this resolved as we are now on the latest version for the release +This patch adds and execution support for the DELETE actions of MERGE statements +Reassigning back to Phil for the completion of the docs +I updated the wiki documentation as well. +Let that be there +Thanks +I will apply the patch and run all the junit and derbyall tests on my machine and commit the patch in next couple days +Ya that is fineAnd the fix resolves my needs +Then use this metadata to determine how to update the bundle from OBR +Ive committed the patch. +verified +committing +The tests fly if no provider is mentioned +Logo now contains a TM trademark and adjusted the trademark statement for the website to comply with the Apache standards +Happens with some patchesReviewed it +Test cases that shows the behavior of +I only created a new fresh project +Theyre almost duplicates in fact. +Just in case you cant tell which file is which the correct one has NOT NULL added to all fields which appear in a primary key +Initial patch for for HUDSON s review comments actioned +Ill investigate the idea mentioned in comment to detach the area and it to the page area corresponding to the last page +And when not defining types or def thats how you set retrieve a variable from the binding +Meaning this just needs some work done and probably isnt some weird bug +Documentation patches for the HEAD +Indeed it does +Heres a patch to for the good to go on this one Sean? Id like to commit today +Committed to trunk and +Attaching a new patch with the new elif block for Hadoop +Closed upon release of +defer all issues to +Save for which was depending on the the code doesnt need this +IS this committed yes or noIf yes can we make it work for +ok to upgrade grails to +The problem is with the dependencies in +Ill provide a patch file to add the new security mechanism +Yes you currently can only choose a WTP project +Thanks Pas! This is fixed +Hi KristianThe suggested bounding of the buffer size sounds fine +it starts in debug modeSame as what run as debug as profile as does in rest of the UIBut I agree it could be clearer so maybe add a Test +Reassigning all open bugs to the development listSorry for the mass mailing. +release bulk close +I will have look and see whether this is broken in the trunk +to set fix version to again with correct fix version +remember to check the report mails or its going to be removed again if in had to find the repository directory in the server +That part of the patch is not necessary for the initial intent of the patch +Sam can you explain what do you mean by gracefully shutdownand clean shutdowns on threadsIf we dont do a clean shut down client willtrigger a pipeline recovery and exclude that data node and isnt this still fineThanks +It shouldnt be present anymore if GPD or higher is included +Closing this JIRA as this appears to be working as intended +BTW removing performance from the component add it if you disagree +committed as rthanks Aniruddha for the patch +I think counting the parents ordinals on the fly is going to be much more costly than aggregating up only in the endThats why we have benchmark noI understand what you say so lets say that we can have which dedups and works properly counting on the fly +and to trunk in Rev +Im not Solaris expert but even if I set the JAVAHOME thething like sudo env wont display itSo this is probably Solaris know how of where to put the env var so it gets picked up by pkgaddI doubt its the user profile +All set +Unlinking from that bug still exists on trunk +Does anybody have any objections if I removed the old FAQ entry about as it would be replaced by the new ones +The listings within the patch look correct to me + on Todds comment +could this be merged to? Does that require a new ticket? +It seems to work pretty well and both automatically connects and disconnects on server startup shutdown +Package has been removed +This is a new version of the Derby release +Noob patch so please indicate if this is the wrong looks good +Oh yeah get is much nicerHere is the updated patch +Im sorry Ben Im out of comission for a couple weeks due to personal obligations +Reviewed +Looked again for the patch +Hi thanks for the report and the very accurate testThis is a tricky one still it surprises me it wasnt reported before +Ive just committed a set of optionsinclude has been out there with the default of checking resources for some time now I did not change the defaults +I extracted all that logic into a new more readable class and cleaned up DW +replied on the user list +Cool stuff +Maybe Teiid does something wrong in the execution of the join +When extracting images from this file I get images +The analyzer is detected as gcc and we proceed to pass invalid compiler options to it +In terms of a argument between Larry and the Foundations other legal advisers the Foundations policy is to avoid copyright notices in the individual source files +thanks Rakeshcommitted as r +The prefix can be whatever wewant it to be and in this case I think we are right in using xsl insteadof out +Committed on branch and trunk. +Includes metric colector metrics for and Includes plugin for Ganglia andI could not create a review request from review board using this patch +I think that would get messy +Thanks much to as there are other small bugs that need to be fixed +You should check to see why you have an image with zero width +Its at least difference +Hi MichaelOfftopic why not ask this question using mailing list before creation of issue +There is a lot of work to implement grouping of maps +I think we need to fix the Jenkins script to run tests at least in the that has the changes and not just the +I would argue it is the html project that is badly setup since it adds multiple instead of having a and +Those wishing to standardize on the SLFJ API could do so as well +keep me posted +Christer Garys attachment wasnt a patch +For example the syntax +Theres no date set for right now +I am able to reproduce the probem as stated using the wagon test will need to further investigation +you need a bundle for each with t the dependency to for the rt version only optional +Looking again it looks like the you had removed the copies in images and they are referenced by +Heres a completely untested patch +the geometry association stuff could be factored out into one such object if we get their api right +leave it open until modello release. +Simple patch please review +What do you mean when you say an unpacked war is delivered? Where is it deliveredThere is a goal that can create a manifest for you called warmanifest is that something that you can use? I fixed yesterday related to manifest entries not working for warmanifest +not onDo you think the attached patch would work? Im not able to replicate the issue but I spent some time working ona patch +In all of our initial server testing for this was not observed so closing for now. +This makes this plugin almost completely unusable can someone adjust the priority of this to BLOCKER +Im assuming the official groupId is now is that not correct +Our handling of a checksum failure makes for pretty abysmal results thousands of sockets used up on host node but fix need consideration +So being experienced in C programming they implemented a smarter data structure to store cached objectsAnd reading their code today I still realize theres a lot to learn from nothing to be suspicious! +New Clone URL git +Logged In YES useridSorry its cvs co r Branch Username cslenkey +for the patch +Ok good to know +Build fixed in path with spaces locally on a buildbot in a and on Hudson +Based on what Sergey told me on conf call todayThis is caused by eclipse trying to start its internal browser to show the content assist will only be an issue on Linux when an internal browser is is not a problem on windows because windowsjava can load multiple browser implementations linuxjava can is to disable internal browser by into what we should set in on linux to disable it +As it stands in all leases take default leasePeriod configured for this region server. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Cybera did submit this patch quite a while ago and we are using it in production right now +Thanksbq +when would you not want different? Its just going to lead to the wrong behavior +Dont have way to generate the obsolete class in the code +Oracle says that including the synonyms will hurt performance since the lookup will have to do an extra join so its best to leave the default to false. +test already was checking for a new attempt being created by checking for an extra third attempt in the NEW state +Also logged in claims this is fixed in the MVEL JIRA. +I will spend little time on this jira +Even if Sonar is now working well could you send me your target directoryThanks to recreate the issue without too much trouble +defer all issues to +We could clone those that are not annotated as a general approach +Thank you for suggesting that +This works because of below +Added a test and committed the changes +Fixed by Git commit fffaafdffddcbedadbd using my approach described above. +When building mahout with testing enabled Im getting lots of out of memory errors that I havent figured out yet +I will open a separate issue for those +SudhaThis is a bug NOT an UI bugPlease assign this bug to a developer +Sergeyyou suggested holding onto the original expception and then rethrowing if nd attempt failedthis seems to create another problem whereby you may mislead the user because you throw the first exception when it turns out that the nd exception is really the one the user needs to know aboutPerhaps simply logging the first exception is the way to go +If I change step to perform an insert instead of an update step times out as expected +Verified on revision +Attachment has been added with description Patch file to remove references +It never completes the caching of revs to use to show in the SVN tab +The small issue found in the code is a sample issue +Whether we actually check for it though Im not so sure about since it could hurt performance for applications that never intend to do XML serialization anyway +Attach. +now passes with asserts enabled +Assign to Javier +up to you guys the changes are borderline many +Good call Lukasz +The username is titmus +On Jenkins those are installed and tests pass +Thanks again for the quick turn around. +is only installed when an Transactional annotation is I guess this issue is really about being able to better detact when is needed +cherry pick to master so it goes in? a patch release is inappropriate for a new feature +I am not sure whats going on +Yes +But I dont like this kind of extra procedureFrom the javadoc of getRowsReturn the number of rows to be displayed or zero for all remaining rows in the table +The code to save performance on has been added +Filed for backporting +The main reason for this is mentioned in +Thanks +By the way Im successfully using M so far so this isnt personally a critical issue providing I dont hit any other blockers in M +so I added a filter for the which allows to use requires +Sincere apologies but I just dont see myself finding the time to improve this unit test very soon +Besides plugins will have to be changed for anyway since the packages change +Partitioners +I wont be able to get more info on this bug as Im working for a different company now +Thanks! +Other server adapter also has this problem such as glassfish +Agree with the above notes +I had the same problem with the consumer endpoint +I am trying to reproduce this issue using srcITunpack project and i cant reproduce it +the patch is quite straightforward and looks good +This of course is long the AS was still in the process of being built? This could happen if this job was triggered manually which i believe it was Started by anonymous user +patch against trunk and the commit for a couple of reasons +If I did not oversee something in the patch then you dont need to do +BTW I also brought the branch with trunk + +We do not put verification files like md and asc files on mirrors instead projects should point folks to the asf source mirrors for those +Committed at subversion revision +Ive removed the bogus entry and a couple other entries that we didnt need r +Patch without the query parent handler check for the same format index style +Therefore anyone using ant dist or even the ant create commands would get permission errors when trying to use +If the jQuery Mobile references are removed the page renders but the styles are gone too +Is this the type of process you were talking about with selecting most frequent wordsYes something like that though i have not thought on this entirely +Sorry but that is still advertising nightly builds to the general public albeit indirectlyIf a developer really wants to find nightly builds they should be able to do so via the developer pages not the pages intended for all users +Can we use reflection to make this change +Sounds like a division by zero somewhere or something like that +quite artificial two make two sets of available +Thanks +defer all issues to +one month without further comments from the reporter +it could have saved me a lot of time creating this replacement And it would have saved me time reverting and making it an abstract class what can you do though? If we could have kept back compat here I think we would have but we didnt find a good way that avoided the issues we facedI think there was a user that picked up a back compat break with anyway a method he could override wasnt being called anymore? So its prob for the best things are always changing and new info tradeoffs always come up +Committed to site src along with Omnigraffle src +Also its easier to say send me your ring output than send me the ring output then the output of info from each node and space used is still pertinent information to have outside of balance issues +Security already fixed via issue is now running with in the libNow we just need to change prefs and capability to RC and run there tests +Howard are there any instructions for Messaging that should be included in this chapter or can I comment out the entire section about until +needs the new class news on this? I just stumbled across this issue myself when I recompiled Trinidad to check out a modification I made +Works fine now. +but the way Im trying to use it seems valid and intuitiveI have an action that sets some beans for my jsp that arnt related to the form +calls for a generalization of this btw +They expire and ultimately are discarded based on the cache settings +Fixed in SVN +hi BrunoFor exactly thats what I meanAnd for user preference yes I agree with youRegards BrunoLooks good at st glance we are embedding more and more Betas +Theyre only working on closed. +Take a look and let me know what you think +Oh +I was even seeing it in other cases but in most of those cases Ive since manually split the load operation so that the objects are no longer instinct says this is actually a current issue on more than just oracle +And there isno way we can get the return type from this +This change is so simple I am not putting it up on review board +need to consider region requirements also +ShalinThis all sounds very good +You can also have a view of the content of the sessionMap at the right of the GA is the version I first comment is the original that was cloned +My concern is more about the possibility of making it optional For in Hadoop at least we should always be clear about who were including +Canceling patch post review +I read up enough on that issue to know I dont want to be screwing around with itSo you have to install ivy in your +I updated source code to work on YARNI think converting Tutorial is almost done +Initial svn import done on the ASF repo +for the patch +Still needs porting to CVS HEAD +fixed +As you see nobody from axis team cared to look at this issue in over a year +Patch for release tested but based on have open issues for +the immediate issue that defrag was unable to read the appropriate header values for storage +Ive attached a patch that solves the problem +Ill try to tackle it this weekend +However I think this issue is a duplicate of +The proof key in the fediz plugin will validate the token against TLS as well. +Solr doesnt have anything to do with it so Solr cant really improve the message produced eitherthere MAY be some tricks Solr can do to recognize when a class load error was caused by another class load error for a class Solr already know about and then solr could generate an additional err message explaining the problem but that would certainly be a new feature not a bug +My fancy new unit test is turning up issues +Thanks Omkar! +Issue created +If youre able to supply one please do so and reopen the bug report. +getThrowable was added in revision +I could not reproduce this +All changes delivered. +In SCA we need to annotate the method with Oneway to mark it ONEWAY +closing issues associated with existing releases +In a seperate jira we should add hdfsclient directory and and split there +This would only need to be done once +I hope we can fix it and p will still update at least that plugin +This means using the crawldb as a temporary storage which probably does not make too much senseWhat we should probably do is to rename into something more meaningful and make it more generic +And thanks to the otherreviewers for the comments and notesI submitted both patches in a single change to the trunk as revision . +After removing the version i didnt require it solved the problem +println to LOGRegarding testing informally Ive been using it internally for a couple of months and Ive worked through the command line options to verify that they behave as expected before creating this patch +We need to document it If we choose not to fix this JIRA +The very simple test case which demonstrates the failure in Flex +So in order to actually retrieve the annotations on the target method we need to resolve that target method all other respects Im not aware of general reliability concerns regarding annotation lookup +I made an interesting discovery! The file in this benchmark is set to sync true which means its flushing after every write +once is committed the latest patch for this will work without any test failures +I can empathize with both sides of the discussion +and one build for updating the siteI would prefer to start with hudson for the continuous build +In FS for this feature proposal is to generate an event on IP transfer from zone to another based on which admin can actually perform required action +First experiment is to replace the double quotes with single quotes so as to stop the string being interpreted in Groovy +Could you give me a link so I can check out +would it be good to have a more stable place to have? Like a build of the for each release linked somewhere +Thanks Ill close the issue then +Fixed in SVN trunk +adds a call to the make in the +bq +I read the log wrong +So it seems we havent reached an aggreement whether to enable this or not by default +Resubmitting + +Confirmed it works as expected in branch +See +workaround in C backend for equals committed in r +A JD in is supposed to have all the entries and any JD from the list can be used to read the journalsbq +The exception in simply does not arise from this logic +The EDS example has been updated to use HSQLDBinstead of derby +Where can we follow developmentsdiscussions on this issue +But what happens then? Would we fall back into the subclassing supportOr could we detect this condition and possibly do the class redefinition to replace the class that was already loadedThanks this initial patch I purposely ignored wrapping blocks of code that require special security permission +It is fixed in the Solaris releases +Added M to the list of affected versions +The cause of this bug is the creation of the facet file for WTP and above +Your change suggestion seems right but would rather itgo in as a separate jira issue and hopefully with tests that exercise it +I didnt realize it hadnt been checked in until I updated to and the fix wasnt thereSo here it is again +It can probably be done a lot prettier I am a complete Hibernate src beginner. +Compared to the term splitting behavior that is evident in the schema for various text fields the creation of additional terms in Trie fields just isnt intuitive +Patch applied please test +Although I have some clue what youre saying with static methods actually doing it properly within a larger work like Solr is something I wont be able to do yet +Thanks for remembering the workspace fork exists +This is interesting Andy as I was facing similar problem yesterday when trying to implement a version of the outbound gateway +Committed revision . +Can you check with? Im quite sure it works so that must be a regression +Thanks Jay useful suggestion Ill update the guide. +Why cant this be in both categories +Please feel free to improve it or send me feedbackI also added a test case in wicket source to test for existance of cookies that were set during a link listener eventGeneral question to wicket devs Is it ok to resolve issue if I am not happy with a solution but the issue is fixed? +Note there are a few other changes that Ive made in that we need to make sure get in trunk +Lets leave that to another jirabq +schedule +please use text t find any log in your attached mgt server log regarding to vm which has a volume is in allocated state while the vm is in expunged stateWe do have the code to handle the situation in case starting vm failed +But its also the box with the most trafficby far so its probably that the frequency of exceptions is afunction of the use of the box rather than the versionIve got a small test app that Im going to try running on the box tosee if I can recreate the bug on linux since I cant duplicate it onos x +And I am now seeing with a runState of RUNNING +It is done already +Let me close this issue now as wont fix +I checked it on IE and firefox and it was working fine +Youre using the it looks like? And can you confirm youre running a single node of Cassandra with all its datacommitlogsavedcaches directories empty except for datasystem +Record or Sample +The mutually agreed upon solution in is more extensive and breaks api compatibility so would it be reasonable to commit this change with an immediate followup jira for the new apis +So for some reason the static fields of the class are set to null somewhere +Please verify +Attached is a patch to provide this capability +new patch with version upgrades +If the timestamp changes the artifact should be installed againThanks! +BTW youll notice I am using Tiles +Looks goodI think we should eventually split this off into another module +POM xml pasted into original description got munged attached is the original example file +btwthe create argument is also very useful for initial creation of the tables when an empty schema exists in the database +I will provide a version with these points fixedOthers any comment or can it be committed with Nicks comments taken into account +Above is so I can subclass and then put the subclass instances into the families collection and have serialization work +However Im pretty confident about that impression +I didnt notice this when refactoringSo should not be set in the first place +this is the with the original problem and it seems to work fine values are preserved +I do this in the persistence annotation project +This is only about successfully binding so there arent host issues to worry about +Since most of the rpc call in are of fairly small payload +I would like to implement in my application by when I add hibernate properties +Thanks drew for applying the v patch +A literal number zero wasalways interpreted as an integer zero and never as a double opr +Sorry yes the Teiid jira can be closed +damn well need to sync the next stale resolved issues +Revised version. +The run did have failures it was with ant +Ill attach a new patch tonight +I think I got too hung up with it being an upgrade issue +So now I use a simply way +patch applied in r thanks Mike! +let me know if I can close the issueThanks! +It seems the problem is in unhandled C exception thrown in thread +Because after text formatting on CtrlShiftF bundle gets unnecessary deleted from bundle map that caused message displaying error +Sounds like a safe and desirable fix forRealized additional issue do we care that this is +Leaving the current patch attached temporarily for viewing only +It seems the sql test hasnt been runned for a long time +Uses the Yoko published to the snapshot repository +This updated patch addresses all of the concerns Todd raised +rev r +otherwise the different return value would make user confused +I had saw that its happening when I put the min max in the +Done +The sproperty escapes everything by default to keep tricksters from posting HTMLJS code as a parameter if they notice that the following page regurgitates it +changed the name to to be consistent with the other system properties +I hypothesize that replication will be on local diskand no real robustness gainNow if there were a way to force off machine writes for a file that wouldbe perfect for walogsSent from my phone please pardon the typos and brevity +Just committed this +Thus this is unsafeint p delete p safedelete p an error because p may not be any longerIf what you are saying is just adjusting memory so that delete will act on the same location the next time then that may hide the problem but it will not get rid of it. +For details see +These have been committed to the repository + for the to trunk and +If you write data to a file its put in cache by default +Attached patch b and committed it to trunk with revision +Robert The forrested one is now gone as we used a chainsaw right? So I think we can close this issue +But I saw they are referenced in s so they seem to be requiredBut I am still wondering if opensaml is declared as optional in wssj some ws security functionality may be provided without opensamlRegards AkiYes all functionality in WSSJ is available without Opensaml +I am not a designer so I have attached a conceptual logo +We should revisit the sites for all plugins that we create usage info and stuff +You can close this ticket +In that case NM will start but will fail after start +Duplicate of. +Please review the included HTML files to approve the commit +resetting P open bugs to P pending further change has already been implemented in trunk +It will only have minor impact on the +Attached is a patch for this issue +LarsSorry for not able to comment on this issue over the weekendThe issue is present in but the fix is not this +In particular it looks like youll need a compaction id to decide what to stop Im currently exposing the CI hashcode which might do the trick +The returned can also be cast to if direct access to the underlying is +Having said that I also would prefer to get such things as a contribution from our users. +for this as incompatible change since it changed the internal url query +They were there until think there was a cut and paste glitch while fixing up another issue and no test coverage for the output at the be easy to fix +Most probably the one providing +on the instance of Bar that was currently being constructed. +sorry postponing a fix since Im still unable to reproduce the problem +I changed Stop Timeout to and then I was able to stop without receiving an observationsI am only able to start if Server Timeout Delay is set to Unlimited +Shaun Can you add the case link test cases passed with EAP CR +If someone wants to help investigate this heres the recipe it worked fine in and started failing in + for the patch +if it has modules and is fully synchronized there is nothing to rename +Do you say that you want to to BOM and samples but keep for testing +Made job retries setting available to message and timer services other than the job executorEdited configuration chapter of the user guide +This causes tomcat and hibernate to be installed as duplicate plugins when it is run. +filtering edits based on the links they contain will catch those fine for us I think +Please let me know if you need any more information +My suggestion if you dont have a very good reason to use geographic go back to geometry +William make sure you add a unit test too + +it happens when you have a ethernet card without any address +FunnyYou say me that its not a Bug +After your suggestion we have gone through the existing entity +Committed to and moved the entry in trunk +Patch modifying to use new api +Specifying a JNDI in standard Hibernate properties should work fine even with reinstantiated at have answered myself the first question because all the methods are static retrieving the datasource becomes tricky +This is a very useful feature +I have concerned about the amount of changes needed to pass the reference id +But Jetspeed has many components and also properties files are managed by each component +Please check that it was applied as expected +HiThis issue is invalid and has been closed +I have to think this over a little bit +Attached patch fixing this +DNS entry created +We are using maven but maven is not complaining just Sonar +SijieIMHO BKJM should use separate root node for his operations +This patch makes it safe to define constant structs that include enums now too +As you can see it can be viewed using the path by r +We have to double check that the env vars are not altered by on addition +In THIS case I think you should be able to gracefully mark the class because the doesnt affect the mapping it simplies relies on itIf there is a larger problem such as errors in the mappings like you mentioned then thats certainly bound to cause more issues in general +Yu Qing and Weijin alread worked out the refined cluster prototype we will document it here and cleanup rebase the codes big change to description of these changes from the mailing listin this big fat patch we have refined the cluster communication in the message level to archive better performancewhat we do make cluster a pure message driven layer no more vc splice on each side cleanup the msg encapsulation and callback implements modified the cache cluster interfacedue to the big change we made there is something we changed in the cluster load monitor hostdb cluster interfaceas our main platform is Linux and due to the network codes we can not reuse we have made some dirty codes into the new cluster that will need more work to get clean to make it aware and cleananyway it is out please join us on the hacking and testingthose codes performs Gbox traffic in our network and there is no limit in cluster performance anymore +So the build is not going to pick that up +Operator overloading is a celebrated feature of Ruby AFAIK +Please review and give feedback +rebased against trunk +WDYM Christian? Bitronix implemented the this same way? BTW you do know it is no longer a requirement to use JNDI for transaction object lookup right +thanks for the steps to reproduce and for the epf file +what you mean by this would also not work with JTA? +I created it as critical because I was under the impression that I was supposed to for the still have to chase it up with promise the next one will have a fix version +Done +This should make it easier to debug issues like this in the future +Its in the JAXB specIs there an bug here? Does Enunciate generate the XML schema incorrectly +I cant see any TRACE log lines in the file +in only takes the path into consideration but ignores the index +Anyway to fix this you need to deploy the channel factory config +What do you mean with including the Infinispan integration? Actually import some code from somewhere? Can you provide some more information +Closing issues up to jBPM BPEL +For Body the attribute name would be called optional and you would again set that to true to change the default behavior +Closed after release. +works for me now +Committed in r on trunk and r on. +It is better to guarantee the caller itself handles the logic properly +Ill make an addition to the documentation to change it accordingly thenSusan +For what it is worth I can actually reproduce this on +Diego Can you attach an example POM and the created Eclipse files? This would help us out +I just committed this. +I reattached. +Thanks for the doc bug report Ryan. +Locally I could reproduce the issue after changing the order of test methods after the fix tests pass successfully in any order +Patch applied at +Bundles with services have to be activated by defaultI therefore removed the bundle activators and unchecked Activate if one of its classes is loaded +Attaching patch b where I have renamed database to is in +If you still think that this issue is not adequately addressed by classpath and compile problems in the problems view then feel free to add a comment. +Thank you. +Is it possible to add methods like add into CriteriaWe use some logic to create required path of subcriterias so we need first to create criteria and then to add restrictions +Like Matview tables +It needs to be movedremoved +But it doesnt work +All passing +Ive not done a build and cannot until I am back home this weekendIf you can confirm a build on a +If you call it from a distribution it will instead just copy the necessary files +Sorry Ive been really busy lately but Ive got a few more goodies coming your way +Can you check if the host descriptor selected is Localhost in the application youve registered? The involvement of myproxy errors suggests that the host descriptor specified is a gram server +for stacking autoencoders and for training autoencodersWhen I have some time I will have a look how to get rbm and mlp autoencoders together +pc temp change your pcs temp folder location in environment variableYour temp location is too long to process. +Add new unit tests +Seems fixed cause cant see these errors in the further builds +Anyway I fixed it and the unit tests and samples all work. +this has nothing to do with Railo this depends on the browser used +issue introduce with the whole method +Thanks for fixing this Kim! +Also when you post please supply the definition of the column from the schema so that we have something to test. +That way you can spend less effort recreating it +If df parsing is broken on some operating systems we could either fix them or declare their df output to be untrusted + +Yes this patch looks better +Pushed to +This does not negatively impact us as we depend on version but it might impact others who depend on older you folks use the maven deploy plugin to deploy these artfiacts +I hope this can make it into the release +Fix to fix +updated title area +We should solve the problem +Added Facelets tag. +Committed to SVN trunk +New patch in which we solved the merge conflicts +Update for trunk +Can you add comment that explains the null check +I tried making the changes but still runs into the object limit when rendering my gadget because of thisI see the value in caching the DOM of the template so we dont have to do that parse all the time especially when it is a shared template libraryI am not sure if my question from above was answered when we ignore an object do we still cache that object but just not include its sizing in the calculation of the parent objects overall size? Is that because that specific object is cached elsewhere in the cache already + I already have Eclipse and Im able install another without problem +Unpack to the modules directory +does not seem easy to share that reference +guys could you attach a JAR file with this change? I dont feel safe building from source code myself +I prefered with nifty corners too +Attaching the patch with r version as no one had problem in migrating to that version +The same can also be done before starting +Committed revision Will keep the fix on as it changes the signature how to configure the quartz component +Thanks for the patch Jon. +Another doc bug with a patch from Chad +Ok in +closing and opening new issue for which is no credentials required to shut down server from will link issues for background info. +Just to clarify the server running scenario should have been covered by clients are required to call close and finished off a couple of gaps where we were handling the close correctly my biggest concern here is the stuck threads when there was no server running as if the client did not connect it is not natural to expect the client to disconnectOne related point the messages exchanged are versioned and the protocol has not been updated since AS so the next release of this library should still be usable on the client side. +LGTM +I think jobs that wrote to the log stream might be tooId be inclined to add a property to a job but keep failed jobs even if the property isnt set +Changes have been implemented with a new task wsdljava and committed in cvs. +Applied patch thanks for the great test case. +May we default it to the selection start index instead of clearing it when the selection is programmatically set to provide a more intuitive behavior on further user selection +checking this in +As you said Alfresco does that with aspects +I applied Kevins idea having a boolean variable locked +Thanks for michael Christian +This class has been moved to package +Looks great thanks! I committed the latest patch in revision and used svn CRLF to enforce CRLF libe breaks in all checkouts +Nightly test results confirm that the issue has been resolved. +Im also fairly against the arbitrary ordering +extends +Added screenshot of custom storage provider showing up in secondary storage provisioning dialog. +Ted is also starting things a little odd +That test is imperfect I dont remember exactly why I put a flush there but I remember I had some issues without it +Please use this second attached patch +batch transition to closed remaining resolved bugs +This patch should be good +It uses Fibers available in Ruby to simulate a synchronous API +bulk close for the release +So the proxy can not rely on the interfaces available on the actual service object +The new classes should have the standard ASL stuff at the very top +This will tempt users to break the MVC t it will be problematic to add beans to all artefacts? How about Services exposed through jmxRMISOAP? Quartz jobs? Sensible default values must be provided for these cases I think just setting null values will be confusing to users +Still lots more work to do but this is a major step +Pages uses the which eventually uses the HTTP state machine which eventually uses the right version to talk to Origin Sever +Closed as. +fixed at Revision Author dantranDate AM Tuesday January Message add to so that all event base classes and interfaces will be included in installer under buildModified trunksrc +For that I think yes we can break backwards compatibility in this toolsWe seem to have fallen into the trap of release early discover when you try to release often that people depend on your alpha crap youre trying to fix +I believe that somehow these errors in the encryption test cleanup is related to the upgrade tests but not sure how +Fewer simpler mechanisms generally include more people +Fixed the error messages. +is the first instance on this server and becomes the coordinator when the cluster splits +For example with Postgres this doesnt work +Committed the patch. +A forum is much more convenient +Is there any expert help me? Thanks a loterror patch failed srcjavaorgapachesqoopmanagererror srcjavaorgapachesqoopmanager patch does not applyerror patch failed srcjavaorgapachesqoopmapreduceerror srcjavaorgapachesqoopmapreduce patch does not applyerror patch failed srcjavaorgapachesqoopmapreducedberror srcjavaorgapachesqoopmapreducedb patch does not applyerror patch failed srctestcomclouderasqoopmanagererror srctestcomclouderasqoopmanager patch does not apply +In that case it tries to permutate the original decision by getting its candidates +The serialization of each of the is done via standard Writable interfaces +Moving to. +A sheet with columns Name Last Name Age +Is my understanding correctThanksJohn +Closed after release. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Remove used to work in JBDS and with EAPEWP +The patch works by setting the deployment classloader as context classloader only when configuring the attribute +Theres no real evidence that it ever worked for example so it also wouldnt seem to be a regression +Proposed SOA TP Release NoteThe run target in bpmorchestration is now disabled +Just tested it +Ill look into both of theseAlso if were using search mode F will decrease so we might want to rely more on stopwords rather than stoptags if it goes down by a whole lot +Was expecting +Closing old resolved issues +A fix is committed to trunk revision As for the startseq parameter it will come laterStill would like to have opinion from others regarding it specially from AdamMarco can you test it? queryparams argument can be used to parameterize the filter functionShouldnt the queryparams string also be included in the function to calculate the replication ID +in +May be it is not needed though + was not fully! Probably I can get my project compiled all of is implemented in categories about category functions +Yup the schema in CR is correct +renamed it to +Made the same change for the element as well. +Thanks to Rohinis help in debugging the AM logs the avro jar issue was isolated +Thanks +bq +Closed upon release of Hadoop. +I was thinking about making it configurable but decided not to introduce yet another parameter +Patch reworked for last revision. +Fixed in +IMHO it enhances information perception and makes pages more comprehensible +could you please open a ticket on to report the problem +Pages dont appear in diagram view +I guess all the classes in them would make it clear as well since there would be something like. +I just created the review board requestsThanksJeffreyThanks Ted for looking the test failures +It probably shouldnt allow the empty string for namesinjections either +added as known issue to release notesWhen using DB as the database for jUDDI DB must be configured with at least an eight kilobyte pagesize for the creation of tables +errors are now propagated. +No we should kill this parameter +I should of mentioned this worked with Rails and TB w +Lets see your config +I will try to find time and work on this +Image of scrollable in Browser +Is a quickly solution i dont know if is a good resolution +Thanks Chris. +AndyWhat I can see from the debug trace is that does not seem to receive any response from the server +The container is missing the dep on module which provides the for the CDI Enricher to lookup +Resolving as fixed. +To this extent the implementation done as of this issue is correct and no further work is required on our part. +It replace the previous broken patchThanks Jacqueswhen you try to update labels files it sorts the labels by the key and inside a label it will be sorted by localesIs still missing to add the apache licence header comment in the xml file generated Im not able to fix this issue at the momentThanksMarco +This works for memvn clean deploy PdeployFor JMX the javadocs and sources are merged into the bundle by the deploy profile +the current fix has resulted in a regression in the release update functionality +is anybody using bundle dependencies that arent marked as provided? It seems like most people use provided scope +Append mode would be fine OEXCL was just a bit more pedantic +It seems to work fine nowThanks! +fixed in I dont see the junit folder under tools directory +Its nice to see the dependency tree complete Patch already committed to trunk r +Second to that what Claus said +As you can the fix version is which means the fix in dedicated branch +Vitali Im no sure what I should look for +So all issues not directly assigned to an AS release are being closed +Add it under the ipc + +Just did a quick look at the patch +are u using a custom +I still however think that dealing with of exception should be upto a and not the callback template +I ran and it passedNot including a test because its a trivial patch so we can address the latest comments +All solved moving to for next check +Checkstyle +That test passes and I have yet to reproduce this locally +A minimal model file still replicating the issue is attached +Attach patch that addresses Gary feedback +As Keith says +bq +I didnt think that was necessary but on second thought it will be good to maintain both methods for one release to help ween users off +add support also for some common standard SQL types like CHAR DATE TIMESTAMP but probably this has no sense on LOB +The proper solution to this issue would probably be to acquire the read lock for the entire checkin operation +Just look at and tell me its doing something thats actually a write operation +Could you please make the following steps clear eclipses log run eclipse with debug options open visual editor to get the error attach eclipses log to this following these steps the log is empty +The related SOA bug has the modified XMLXML QS +When applied to the current trunk it will create the new facets test file with all the old tests plus the new ones +really need this in next weeks +Thanks for the detailed explanation of what happened in your situation +This is not a new term at the foundation +I dont have any issues with the concept but this breaks the build on OS X +I also added some error handling to make problems like this easier to cope withPatch applied +Looks good to me +Thanks Vinod +This should be properly handled by overriding the setAddress method in an overriding +New hbase release doesnt rely on metrics v anymore hbase release doesnt rely on metrics v anymore cancelling +Radio button for single row selection and checkbox for multiple row selection is common expected behaviour of any table +this case linked resources are external class folders +If I remember correctly the jump with this patch was almost . +Completed Documentum stubs working now on +Is it possible to tune castor to allow unmarshalling for sequence of elements in incorrect order +the syslog fileClearly this needs more work +Great do you know when this would make it to the Google gadgets server? We have Wave developers waiting on the fix +Can you add a review board entry +These test failures happen in an unpatched build too and so are unrelated to this patch +and +Couple of thread the attached stack traces root problem is threads are hung trying to take out a write lock on memcache +deferring to as were rewriting alot of the generic type sure when this was fixed but it works in master +As per vkiran this should be resolved. +The idea would be that this would be automatically generated from project or Ivy files +But for now it makes sure that is able to handle multithreaded clientsWill commit this when there are no objections +Em temo que enjoc no existeix en catal +I dont have any workaround yet and I really cannot understand why is released as beta containing this grave bug +Ill look for a flag to make it interactive +So Im changing this to an improvement +In this case i suggest splitting doc into two tasks one for which needs to be marked for and one for Monitoring VPN tunnerl for subtask is target for I would mark the parent task to because every time when people filtering the status for they would mention this as open. +This should not affect running the ant command +tasks log directory to Ive not done this one alone +Updated the fix for alpha by moving the restoration of the project after the child assignment which corrects issues when the unrelated columns as not projected +looks good to me +Andrew will sand down this file and link it to additional documents including our html user guideso Corrects the wording of the webpage as John recommendedNote that the ifjdk conditional mentioned by Andrew is included in the target itself +Looks like the constructors still take aMy vote is for long instead of int to maximize forward compatibility +this is a bug in SMX +I will take further discussion to the forum if you wish +Glad to hear that and thanks for letting us know +There is a chance that the problem will disappear in CR if the demo app is taken from the testsuite where it will be fixed +All of those member variables are being allocated in the getDistance method so no reason not to just make them local variables I think +Displaythe page in a browserIll hopefully get some time to create a quickstart that does the aboveIve been promoting Wickets dynamic approach for page compositioncompared with Tapestrys static approach for page composition posts onThe Server Side and others so hopefully this little hiccup can get fixedsoon I dont want to have egg on my face +I see the Transaction as a atomic operation so you should be able to lock a node after create and save in one transactionalthough the persistence moves to the commit of the transactionI will provide a patch that handles this behaviour with some more tescases so we can see if it works as described in the specification +Therefore descheduling +It also is set as the default if the field is empty +Was for some reason using the wrong Maven version also had to reconfigure the build timeout +I agree with both of your advice +Close Doc issues in release versions. +Ales use or the development forums for ends up in release notes and we dont want to confuse users with irrevelant issues +I tried setting up these properties but Im getting an exception at export time stillcsv excel xml pdf so much for your help so far Jorge! Displaytag is great and these new additions will be great too +Committed to. +Thanks Michael. +Oh well no way to fix it it seems +Thanks +With that change I can modify the existing unit tests to implement a run method where the prickly scenarios can be run against an external cluster as configured by +Hi JochenWere using it to do SOAP with attachmentsMark +Oh fuck Sorry +It is better to remove the logging because neededReplication has already done it +Namit +The issue is actually in and it is fixed in later versions. +Per Elis feedback updated comments to be more accurate +Jan just removed the this still an issue +Florian here it is forked a separate issue for the OSGI discussion +Ill open a separate issue for I think its worth exploring +Project mockup fileUserscheyenneweaverwcnewrepositorydesignjbossorgcodejbossorgprototype mockups are complete +trunk +Good one Ryan +Added code suggested by Mirko. +Works locally +Is this the same issue as reported by +A patch for trunk looks good to me +seemed to have been fixed in beta +as now is default version of openjdk in linux distributions this issue is out of date. +Added patch to webapprailstrunkdbmigrate +This is one way to tell the deployer which server instance to deploy to +I think it should be enough to make both the add and the add methods synchronized +Oh and you spotted bug which we got fixed thats good. +Requests for some tiles just dont return +The url parameter doesnt need to be last it just needs to be url encoded +bq +Fixed for were also initializing the lazily now when actually building the +Im on it +There are now a couple methods that can be used to create a new connection one of which is static and allows for all parameters of the connection to be set there +im fairly tapped out at this pointI like +Just my for this patch +Regrettably derbyall fails to catch this as David has noted earlier since the localized strings are bound at runtimeDid not run derbyall again yet starting it now +Committed in revision +Hi Kai my previous response hadnt really intended to compare the designs as much as correct something that based on your interpretation must have been understated in my client interaction overview on +dodeploy gets added before the progress monitor finishes in See AS server writing a +Thanks Ralf I will take care of all this next time when I will code +Could you please include your +Can we consider C and Mono for this reimplementation +I would like this fixed before the next release +heres a rev with some of the above changes +is an useful API but not used currently +Why do you want a system wide property for something that is only used in solrconfigschema? There is a chance of it conflicting with other system properties tooI do not see the reason against the properties +Closing it. +Do I understand it correct that you say that it is bug when because there is also a space before the xyz? So it should be searched for xyz under field? Am I right? Or how do you expect that to workI totally understand the inconsistency described by Andrew +Pushed transient functionality from to more to do +Also Ive noticed jar in the classpath +note that the above properties are temporary names +Julien You hit errors in the frontend or in the backend +However if record conversion is needed we would need to run with number of tasks equal to the number of splits and the should return the peer index in the function get. +Thanks Prasad +I linked the issue to the portal security interface implementation without JACC. +The rest of the xpath docs look good +images are fixed +Merge was clean derbyall was troublefreeCommitted to subversion as revision +I think a plain text file is good for this type of info +That is a valid optimisation for using in our test suite +Attached v which has the following changes +Ive never seen ext in production +So now the extension framework can use this factory to create the appropriate assistant +Thanks Xuan. +This looks good except that you shouldnt do a shutdown for interruptedexception +Is there any possibility that break the rule each scanner should only see according to its mvcc readpointBecause in current flush logic will be deleted when flushing if there is tag of delete type + +The failed test case is in +But no test cases on the patchIll be taking a slightly different tact for this problem one that should also resolve the issue that the selected is highlighted in other months even if that day isnt selected +Kan thanks for your insight +So under this approach the and would never assume the users identity only task wouldAt a the only user I see area b Task it possible to have a situation that a RPC socket is shared by two identitiesIf yes then we cannot initialize a socket with someones ticket +Thanks Richard. +verified on +Thanks for reviews stack your turn +Committed to trunk +Some times if you change a Hibernate Mapping the Schema Update dont fill this change +I agree +Could not reproduce on my Ubuntu JBDS or local IDE +If you define some new AT then the guide is not accurate anymore + has been fixed +Just remove the defaults +better fix conditional execution of source plugin +I added two basic interceptors which can be directly injected with service objects if a user knows it is a singleton then these basic interceptors can be used for WSRS bean validation +I really appreciate your help +I dont see a requirement for a new bug about this + +There can be more such scenarios +Slav is this WTP causing this or us +why are you obliged to use the row count as the first argumentAnd are you telling me that DB cannot return the value? In oracle you do it returning SQLROWCOUNT or i added those register call to have it return the proper value and if i started to return a different value than rownum then oraclemssql failed +This is why we have a setting so you can disable batching of versioned data +Dobes could you provide some insight on how you wanted this to workBased on what a new shard was supposed to be created +That way people can have applicationjson or or textplain however they want for their application +I actually got rid of uuid completely and just used the same mechanism that it uses to get a random bytes since theres no need to create the UUID object if all were looking for is random bytesI think the code has moved a bit so mine might apply more cleanly +this issue should have been fixed as part of right +Or did I missed somethingThanks +Both Column Label Value columns are populatedIf this isnt the scenario please add more detailAlso what OS are you running JBDS in +This is untenable +by +Resolved Also see er +I committed the above suggested patch in revision after release +so I think its ok to disable the tasks and dedicate them towards trunk? +No problem at all! Thanks for looking into it so farIll hunt it down at my side and see what I can find outThanksAny luck reproducing the problem? for the delay Ive been away a couple of weeks +bq +failure is unrelated +Thanks. +Just in time BillI just started looking at spatial stuff today since Im planning on putting some of it in my Lucene Revolution presentation +HiranyaCan you own the action on this patch since you are involved with the FIX transport? thanks the patch from Asanka A with minor changes and tested the old scenarios and the new scenarios against the trunk +This issue is getting quite old now and is still causing us no end of headaches +Thanks for your efforts! just built a new jar based on the trunk +Manually tested! +Yes were using NFSIf we try to start it using it works fine +If the jta zip contains only API docs then it does not sound like this zip should really be something I handoff to QE as I have no ability to generate API docs +What do you guys think? Should we create a separate JIRA or should I just make the fix together with this code +Changes merged to trunk. +In what case do you want to add rules? Are they any missing +thanks for the review Bobby +I think we should keep the for trunk and if everybody agrees I merge the changes into excluding the additional strategy +So we can add similar fix for this function +Marek is it ok to just add the license file to the trunk of the specs parent? I dont think we want to modify the tag +Get cluster running on loading data +It is tested with chmod +Perhaps useful as additional context for a wider change? +Willem we can set the transacted option to true by default +The final hudson config isjdbcmysqllocalhostsonarautoReconnecttrueampuseUnicodetrueampcharacterEncodingutf +Setting the versions correct sorry about that +In testcase would it be possible to check perl and terminate if there isnt on system running this else it might be difficult to find the test failure +Umask on both local and server is I was playing around with Ivy earlier and managed to publish files on the server so that their permissions were which is what I wantI cannot reproduce that behaviour with ivy anymore though +fwiw the project scheme is configurable by editing it requires rebuilding Continuum at this point though. +This bug report seems almost identical to at the provided with spring it appears that error code is being transalated +I think Ive found the problem and committed a fix into the SVN branchCould you please give it a try and give us your feedbackThanks +Havent see this fail in a good while. +This is by designThe client should not block as it may block forever if there is no consumers on the queueInstead we may consider adding an option you can enable on the seda component to control if its okay for the client to block if queue is full +Tried it on RHEL cannot reproduce it there eitherI can see Loaded Apache Tomcat Native libraryin both andAlso varlibnativelib is listed in JAVAOPTS +Thanks for all the feedback Ive compiled both tests and patches to fix the issue +patch to the subject was supposed to be Polling Methods on RegistryThanks! +Its a regression in that has been is the identified regression the cause of Heikos first exception second exception or both? And do you have a jira for the regression? Im seeing the latter exception only and only when I dont provide the b option to bind to a specific interface +Maybe that file was modified or is even empty +if we keep this we need two new attributes +Do you intend to provide another patch for the service definition? Else I will close except if somebody see a trouble with this change or want to enhance the service definition +If this doesnt go forward we would have to push our version either on the repo or bundle it along with the release file +Of course we have to do some more test on this feature +Removing all the references to XML DOM wc sure makes a lot of red stuff inAll Im looking for here is if Im overlooking the obvious +I can remove You renamed this +stop +Attached patch to merge and commited +Besides this sounds more like a separate project to meMeanwhile just make use of the power of IDEs +Attached updated patch covers one more pointAt Gl Reconciliation screen change the label of the screen from the View screen has the list of fin account trans its label should change should now Gl Reconciliation And Fin Account Trans or similarThanks to Rishi and Sumit +It is for advanced use thoughand is not something that every client would consider doingin which casethe patch has no impact on them anyway +Bulk update to fix version +But I have no strong feeling about it +NOTE this is an example patch that was requested in not a real patch +This is a for but as it can be added later wo breaking backward compatibility it is not an absolute must +With we added an event which should be sent whenever the mapping changes +The plugin is now aware of grammar interdependencies and process the grammars in correct order. +Im on standardizing around a single format and based on everything Ive heard to this point also on that thing being docbookGreat work stack +This seemed to fix the problem for me +is a really unhelpful name +describes sames issue +It should be Trivial +Once I moved the code to a simpler maven project I had no more problems +verified on JBDS +pending Jenkins +But now there is another problem the test sometimes failed but sometimes success +If there are no objections I will commit it in trunk and the branch +Findbugs changed since used in the earlier versions +Im marking this downward +Probably worth a discussion item at our next ff +Hilarity ensues +Attached is the patch for this issue +Checks the namespace struct before deleting that inside guththiauninit method +I created a class that is in both and +Added a patch that fixes this issue by maintaining a hidden property to store the names of the properties that use definitions but that have only one value +Its not only a classloader leak because by leaking the classloader it can become impossible to unload huge parts of the web application because so many Java libraries like to keep static cachesMark Thomas Its not correct to state that having each app use its own copy of will fix the issue +The only problem Im seeing is every time I build a war another entry gets added to the file +OK that helps a lotHeres one question Ive been trying to solve +Duplicate of closing. +If ye dont take it I will +There is a general issue with the way that localhost is being resolved in this release as per the link +Trying to rule that out would require a dynamic assessment +I think its ok to have them duplicated +defer all issues to +avadocs look good trivial nitpick i noticed some relics in such asgood that you reviewed it! I missed those! here is an updated patch +GREAT +Relieving NKB of some assigned issues by setting them to unassigned +By the way I was not able to reproduce my problem in a genuine SMX build so it seems my environment has become somewhat corrupted +I should have created this ticket with a higher priority +Heres what I had in mind at least I think this will do but all Ive done is insured that the code compiles and the current LFU test suite runsLook in the diff for timeDecayThis still needs some proof that the new parameter comes through from a schema file +I had some trouble with my environment not correctlyI have done ant clean before testingAndy take a look at this patch and tell me what you think +I changed the references for and I can run mvn site command now +Mauricio supplied the Portugese translationsJames CobbIf you have a better wording +so in fact duplicates! Close oneD +The provided implementation uses JAXP to realize validationSimple testcase contained in the ve applied the patch +I will write this and post it tommorrow +Further shouldnt that Serializer not extends instead +Closing verified in ER build +We really need to support both cases +This seems wrong +I have applied it in our Java branchCommitted at revision r. +You can manually edit the diagram to associate the call with the +Ill update to and see how we go +Yeah this is expected because new bundles are installed by the feature service +What you think of the hadoopqa failure David? Let me try resubmitting your patch +We need more info to solve this. +The fix does not affect what above move to +We should discuss in the ML for future planning or just shelve it +b is a field in fact brokenBinding is the binding to bIm attaching my open rule +would you please try this patch? thanks a lot +attrib is an optional task so it is not guaranteed that the necessary jar file is in the class we package the jar as well as the ant jar with Groovy then this error is an indication that the user hasnt got the appropriate environment set up +The test class is attached +You should put all the stuff into the same adviceWith and only call it once +How do these tests run? Are these tests run against a particular environment +Maybe this is a critical section to be fixed +We cannot do anything about it but need to be reminded of it +example aar and a client class +Hardy +Yup done +That is what I get when I want to commit it +Committed the patch and merged it up. +I will work first on getting the log outputAttached is the and the database +with startcode and server INFO assigning region contentcedccabfeadfcfc to server INFO Received MSGREPORTPROCESSOPEN contentcedccabfeadfcfc from INFO updating row contentcceaaddcbdfe in region +I manually tested this on a small cluster and verified that a container can be reserved for minutes and the AM can still start the container after finally being allocated it +This is fixed now +With the patch cache and database are in sync +Reopening to write release note issue now that release note text has been written +Please close this JIRA issue +Added path that makes the Hadoop nfs port configurable +Ive created which targets dropping checksum on different blocksize between source and target +Please remove the commented out code lines +Committed this to trunk and +Is this fixed +Sigh +Im going to commit now +Patch can be applied in trunk branch branch +Yes +On the contrary it was great input +a new storage file when an is thrown in method while reading LAYOUTVERSION from the storage file +I fixed the code but it didnt make it into the plugin zip +The max parameter uses shard in the incorrect sense I believeIts kind of a questionable parameter anyway since it doesnt make sense in the context of nodes that may have different capacities +Thank you. +Can you please explain the why these need to be persisted? Delegation tokens are used by tasks to authenticate with namenode +So I gotta try a different approach but I did learn how to authenticate using LDAP now +Created and still on the heap after failover is finished. +In all our tests we test with the filesystem mount options and configuration you have +It just seems that the Identity object is not connecting to the +Thanks for the alert Elliott +Also evenisse pointed out that this should be done so that continuum will not require a plugin release prior to its own release +Mind tackling it +But thanks for the report will raise the priority of the original bug. +If you need it quickly do not hesitate to investigate yourself and send a patch. +This issue could resolve issue applied that patch you could not solve the worst scenario others may commit on trunk so the tag created could not really be the one that you checked out and ran testspackaged do you think about it +Work has been started on that +revision contains the fixPlease verify and close bug if possible + patch for update to iText with patches for update to iText +This would allow all the region servers to start up so that when assignment starts it will be more balanced across all region servers instead of skewed towards those that connected first +Is that a problemThe main problem here I think is that the removes the location of the replica if it was not reported during a block report +This would be one less place to change when standby starts reading from inprogressIf you feel strongly I can change patch to return for nowRegarding testing any HA test will run into it +using next line number for overflow method not last +Do you planing fix this bug in short periodIts very critical IMHO +Two things to note here There will be no public release of AS The only way I am seeing these incorrect launch args is if you are using the server adapter to create your server +But to be honest Id be OK to say this is a known limitation of +Thanks +So we might use for the maps of name countergroup +By the way I committed this to the branch since I believe this dependency fix should make it into. +is also an annotation update +Is the fix now on the branch +The standard will do the right thing +I will update the patch with this change +Hi Israel great workCurrently this only works on integer fields +These defaults have not changed much over the yearsIs it reasonable to have a Lucene index that uses over file handles? Absolutely +released closing. +This can be a bottleneck because Ill have to query db on each request +However restart does not play any role in making the edits log empty +The patch was applied into the svn repository. +Patch to do so +I see that the log also indicates another warningwarning SSL connector not enabled you must set QPIDSSLCERTDB to enable itI have specified the in the config file as well as on the command line to no avail +You will copy hadoop libraries to DISTLIBDIRplatform while the original script copied them to DISTLIBDIR when on Linux +Please categorize this as task as its not a . +Fixed test in trunk +I am going to close this +Fix for this has been committed +Im not sure exactly what Remko had planned but as explained above and on the mailing list there are unit tests for these appenders +Is there any way of passing parameters to the listener +and +This should be resolved in BRMS ER +Tie interface is fixed +Ultimately we just want to make sure that we are covering all relevant goals and usecases in our converged approach +mmmm +Attached change to get around in Event field to bring this patch Patch Version +Will check what causes those failures +Im not sure how common that scenario is but weve had to do it twice once to fix a DN bug once to fix a configuration mistake +Great news! When do you envision a release that contains that fix? +Yes this works +You can group into a map input by host domain or IP and then into each queue based on the same criteriaBTW why not asking on the mailing list before filing a JIRA? Youve opened quite a few which is good but dont reply to comments or questions on them which defeats the object +Jason I suggest we just get rid of the Velocity version and only support FM +I raised to blocker as the solution of it affects the consistency for and future releases +Ive leave to Adam to code and benchmark it to avoid demonstrating my Erlang newbieness +We need to overcome thatThanks I mean was its totaly depend on underline parsers namespace context +Hi MichaelThanks for the patch +The exception was removed +Thanks Zheng +Otherwise the controller wont update the path +misspelling +If you want to change to CFS for you should either subclass and therefore create a new codec or create an entirely new codec +I have verified this behavior on the Apple Java beta for Intel machines +Please add Affects Version so that this issue can be triaged into the correct major version of EAP or EWS +host A sends ping +Its Jons patch only it purges the family comparator from KV and it adds checks to Put Get Scan and Delete that throw exception if family is illegal +Its running a UI for the same purpose already +Fixed with blockers and critical issues are resolved but not verified +Attaching diff for engine packages under impl +Ofbiz web services uses map based results +and which node would that be? I dont feel like looking at logs +Defer all open issues to +cant we use start and end in string format and use +If the tests pass and the module builds from the as well as its own directory I dont think we should worry about it +Thanks for the reminderHey you had better not be working from hospital! +In any case this is fixed for now will be available in tonights +Has the schema been updated on the web? Im not seeing an expression attribute appear in the XSD editor in STS +Fixed in +New patch for discussionImplements a new config controlling the maximum number of apps sent to the ui +very well compliments for the timeliness +Weve already got a section in the config starting daemons and plugins are free to define their own sections in the INIThe second bit is dynamically adding code paths to the Erlang VM +The instances are being created with a flag indicating that the item is URL encoded set true +In this case we must remember we populated when limit reached and restart this population on next method issueI also added a test case for such situation +If every class we generate extends such support classes it makes it harder for the end user to manipulate themA comparative example would be that we should have a domain class that abstracts the Id and Version boilerplate +Sergey MikeCan you please take a look at? The patch for this ticket caused some troubles to applications that use Wicket Tree component without relying on SwingsCan you please test your apps with the new patch inThanks +BTW could you please edit the issue title +Nice the logj dependency on test will be very useful +bq +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I just committed this fix +This was fixed by r. +Thanks for researching this your appraisal was spot on +separate file for properties be generated in case of necessity to mix properties from user file and from eclipse settingsTo avoid rewrite existing file with the same name error message should be in export wizard +Added it to various trunk +Attached the can you describe the test configuration in the referenced failure logas you mention it passes in the Hudson CI job using svn co of JBPAPPandit passes with the test source from svn co against the binary zip combination causes the failiure +dont mess with path either of those alone would solve my issue and not depend on hard coding of if git or similar + +Please verify in vM +is effectively at the end of life +Looks good to me +There wont be a release +The development of the maintenance branch has stopped +Let me know if the overall manner design makes sense to you guysNEXT UP Change in message manager factory to create async managers on demand +Ive updated the issue to set the Affects Release Notes field +also affected to avoid suprises it has been renamed to match its corresonding issue +Fixed in all resolvedfixed issues of already released versions of Roller. +Verified at +in +The feature has been a community effort and discussed for almost months and there is nothing about it its been running a heavy test suite for a long time +will be implemented by item Item will support NOT AND OR operators with precedence +Just to clarify the exploded task uses the target directory +This may be needed in other in the examples folder as well. +on to trunkThanks for the patch RajeshThanks for the review Stack +FYI minimal supported Sonar version is +The bug now is that the second starts successfully right? Why is it hard +In all cases whether it be metadata or the deploy and folders for regular servers ARE on the same filesystem +Documentation update attached +Uploaded a new patch +As well there is INFO logging of the endpoints with the queues and their sizes in use +Well IRI is acting correctly +At this point we have implemented of the changes +Requires updating the version thats being used within the AS +This also makes it easier to keep decomposer decoupled it can depend on and not needBut what about Vector? Vector coming out of Core and moving into another module well everything depends on Vector doesnt it? In what way does this help +Whats the quickest way to build a new jar when Ive changed the file +It looks like the deployables do not conform to the standard and some of the returned pages do not match what is expected in the assertiions +This is about SourceGenerate Hibernate Annotations for java classes action +So anyone who wants to build something that goes into Eclipse has to avoid it +minor fix in the last +Does it make a difference if you run webtest then webtestCan you please post the output so I can see what the variables are resolving you please try the latest version it works for me there must be something about your environment that breaks the classpath +Committed +not to be fixed for +should have been marked +It works on eclipse but not on Springtomcat with the same fileProviding a minimal runnable test case would be dont tell me youre messing up with the dont mess with the entityresolver unless the user specify one +Thanks for the correction on the versionYeah its that empty namingcontexts value which seem to cause issues with StudioYes please open a new JIRA issue for the other errorThanks +I see your point Nicholas +Moved old school HRI to migration package and renamed it to +I worked on this +Oops so it is +Thanks so much for finding this issue and fixing it +What I committed +Pushing issues to the next CR +I have added the whats new chapter to the docs +Correct + +Ill upload a new patch +The have a method set for changing default charset +Closing +Looks OK +Tests are working again. +depending on how many it is initialized with it should use them all for reporting the statistics +If anyone can still reproduce and can provide a fix please reopen. +Can you recreate the patch making as few changes to the test file as possible? This file in particular relies on tabs and spaces being correct +Fixing it right now. +I guess this one is already fixed in and probably also +Closed barring feedback +It does a dump each secs +Ah I did not notice that the fsync call is carrying length now +The call to b instanceof is protected by a boolean variable indicating whether or not the package is supportedI can commit this myself if there are no objections +Go ahead and commit +Hi trouble is that extends +Deepa just curious was there a reason that isJar is a boolean array with a single element instead of just a plain booleanAlso Im currently reviewing Myrnas patch for +I think we can close this issue opacity is now provided by the. +All can use the same static using as in etc +I confirm that the issue is solved on my endThanks for your time Lorenzo +bulk move of all unresolved issues from to +WCF does not need all the data types. +Socket permission continues to be denied to code outside the server such as functions and procedures +Manually tested +Patch made against Hbase trunk with the suggestion made by Luke Lu +I will investigate further +The comments suggest that it may be fixed +Added two new sections to the Reference Guide chapter on repositories a new section on writing custom connectors and a new section on the new graph API +When there is no id deploy failedWhen there is only id set id as process definition nameWhen there is both id and name set id as key set name as name. +then get the full cwd and JRUBYHOME is set to the dirname of that +I was running into this problem before I learned the maven tricks with ivy cache deletion +Actually Im a bit busy recently but as soon I will have some time Ill contribute an other patch based on your suggestionsBestAkos +Merged toIt turns out that cannot be compiled after the merge +Both archives use the same codebase and have a limited set of the River source included +Could you be more specific about this issue +this type of situation makes me want the ability to manually take a node out of the cluster and prevent nodes from gossiping to it +occurrences of using the global variable directly for kinitpathlocal +Do you have a sample POM? This will help in debugging dependencies +Scott is this possible to doI like the idea of integrating the Collector with the in future +I attached a modified simpleinvoke example that illustrates issue +Revisit open +I am having major problems with the samesimilar issue +the java service into Minilang tested it as well +Id bet you could whip up some Javascript that changes the copy buffer to the number without the commas as a +What are correct packages to use +Still need to prove performance benefit. +Where would the new release list beHow you mean? Itd be under the downloads? We dont explicitly call out our releases anywhere but what is listed here under mirrors +Sebastian this is ready to go +Also there are several misspelled words spell check failed on mesages spell check failed on annotatation +Please test against current CVS vbranch +Fixed yesterday +Are we also looking to kill all running containers when either shutting down the NM or when bringing it back up after an unclean shutdown +I added a test to the regression suite butmore tests are always helpful +Im astonished! Jobs such as the injector generator and crawldb updater run at lightspeed with this upgrade! For example the injector takes now s instead of s injecting a single URL in a single record on my fast machine! Ill test it a little more this is suspicious eventhough a test crawl continuous to operate normally +Are you still working on getting this into trunk +I seem to remember something coming out of the discussions about this +This still needs to be fixed but since it is at the bottom of the editor it is no longer noticable by the user +i committed my changes without testing again +Seems one cannot have a dependency on Context in constructor of page? Neither can Click or custom controls +If its possible that a user enters or in one of the fields the wicket developer cannot use the affected URL Encoders +Hello i was wondering if this project still available? I would like to pick this project for my project if its available of course. +Again Im not sure if anything further would need to be done but as for jbt use this is sufficient +Check if the TT has enough mem for the job +The real question is why line is not coming back as +Enabled cache loader config element in and and ran the different test cases without a problem +All these properties have been implemented to be available from the Message class +Waiting for an imminent patch by Patrick Meyer +Applied in r thanks for the patch Dhaval. +derbyall and OKLocaluncommitted replication tests OK wrt +I had to use to avoid the size limits but you can just extract both zip files to the same folder in order to merge their contentsThe shell script will run the necessary svn commands to remove all the old files rename one and add all the new binary files from the above zips +Previously we set the interrupted flag when we first threw the internal +Committed as on using. +when should we know how to refreshbtw +This enhancement will be considered for a release after Struts +in we can transition from NEWSUBMITTED when we receive RECOVER event +Addendum integrated to trunkWill apply similar addendum to and when I get into office +The recursive activity is no longer visible in so closing this bug +Done +I mean use ivy integrate startup script with binhdfs script use as configuration and etc +Willem did you get the latest patch committed? If so can this ticket be resolved +It is my hope to have this done by the end of Oct +Thus it would seem logical that when service A invokes service B with the message would be returned to service A instead of the actual if the behaviour of returning the message from service B to the client would be by design this raises another issue in that the message is not routed identically when retrying delivery +If you stop the inserts and force a STW GC memory usage returns to normal +Next version of patch that contain all changes for that part it is also contains changes mentioned in latest mails on dev list +Not sure when it will be ready for posting. +Heres a patch for trunk which should address the issue +just committed this +Leaving issue open for now until the code has moved to the Java DGM class once that is in place +issued marked as resolved by resolving +The events will occur on merging +If this looks acceptable I will apply it to the other scanner classes as well +reopeneing for and branches +this patch addresses only blob cleanup +Patch addresses the test issue +Will commit this extra logging and upped timeout +Sorry that I forgot to log in +You should be able to use the same datastructure for both potentiallyFaulty and blacklisted trackers +Please dont apply this patch as it affects other server implementations +If were just going to have CF TTL being sugar for clients too lazy to apply what they want then Im not interestedBut if we use CF TTL to provide an upper bound on how long data can live then we open the door for some interesting optimizations +I just thought of a case where this wont work well for utilization +We have unfortunately no real backwards tests for contribs but I see no major problemsI will commit this soon and record merges from trunk so the commits are related by mergeprops +Thanks for the patch Kay Kay +Ill look at it this week +What jdk are you usingI am not able to reproduce this issue +It usually crashes due to a too many files open or Java OOMThen increase the file handle ulimit and heap size. +The fact is we have Maven right now and in my experience it turns out not to be a huge problem in a maven normalized project +bq +I am addressing that issue with +Dan bumped to for now +That div has the value inline under IE and under mozilla +should be trapped +need to test in +Ill try once more +The issue might relate to which also happens only on DIST mode +From Tomout of scope +You need to switch the search flag to true to search those as well through setting the corresponding bean property on your definition +Im using Maven to create the war and such a problem should not append +I HAVE NO IDEA IF THIS IS VALID MATHEMATICALLY but it does show results +htaccess hacking Doug +takes key out of +I instead validated manually using the fake shell +Sorry the component should be Tools not Build ToolsNot able to find out how to edit an issue +Capacity scheduler tests also run fine +Havent had a chance to reproduce this will look at it as soon as the GA has gone out +I created my in code +Bharath it will be apparent that it failed but wont clear why without looking into the tests code +The Google base is a good idea but in order to submit we would need to a google userpass in the installer which is opensourceMay be its possible to go through a another web app or web service hosted at jboss and store the passwords I am thinking that may be we can use the site stats after all simply looks which files are downloaded by which IP and this would give you an approximation the set of files being downloaded by individuals? +Since this is really for walking the Query AST during the rewrite process we should make this very clean in the IF name +Closing per Radus comments above. +applied +Its still broken on AIX +You may want to have a look at as well +Marked as resolved since this was committed into SVN trunk with r and into the branch with r. +I didnt realize that it had been released when I created this issue +Committed revision attached is a patch for contrib and its tests +Is this not included transitively +in other words the name whatever is only for internal use. +Also our archivers like Nabble marc or might want to move their archives so that the history does not get broken or lost +open new specific issues for future work. +I apologize for my lack of memory on this +that works as well +Reverting it on trunk and +Hi Bard you are right I need a different formulation off the static semantics +patch +Specs committed in r +This functionality will be included in but the issue is being punted to when Rebecca Simmonds prototype IDE plugin might include debugger support +I would like to explain what is oprhan log spaceOrphan logs are collation logs that failed to send to network and maxspacembfororphanlogs is the max size of all orphan logs but this option makes no sense when there are other standard logsSo I think Its safe to remove maxspacembfororphanlogs option FYI +Manually tested +This has been resolved by +Integrated to Thanks for the backport JeffThanks for the review Lars. +Knox talks only REST to Hadoop servicesKnox unified and secure REST endpoint to its client for all downstream REST end points of Hadoop services +fixed +For branch +They then probably hope to get a performance gain when they accesses the file directly +Ill be submitting my CLIA + +ok seems as if this is due to a major rewrite as the seems to have been written out of the code completelyThis case is probably better off as a mailing list entry than as a bug report +Patch that excludes white space fixes +here is a patch that merges sorted docvalues without loading into main memory +Now the role attribute for Tiles attribute is digestedAdded test case +The patch looks simple enough and fixes the error +Committed to trunk. +This is gone probably was fixed. +I bet it got fixed on the way +That has not happened in the last years so I think its safe enough to push it upThe code test user guide and doc changes for this have to be done carefully +Targeting to +Doing this will make the start row exclusive or the end row inclusive +. +Len feel free to assign to someone else in your team +There has been no additional feedback so Im resolving this as wont fix. +Ive also updated the and build to check for unsigned +While it does release all of the messages it does not properly close the indices resulting in a handle leak every time a destination is removed. +Will open a new JIRA item for the There is an issue with the JS Chart controls which are currently rendered in the page and does not work as the JS includes are now at the bottom of the page +I think we should have at least a simple test in the Common project otherwise well only rely on test breakage when we make changes to this code +Its just that the wiki doc was really skimpy so new users may not have a clue how to implement it +Henkthere was a problem with my main +I do have the impression that jira will be resolved that way. +Using now +The only time multiple notifications will be sent is if the session has been accessed on more then one node which shouldnt happen if sticky sessions are above was implemented for +I dont think we should create github projects that only a few people can commit to and then link binary jars from them into lucenes source tree +The only problem is that does not really let you edit them. +We should fix this because I also hit this issue a few times +Thanks +Added equalVariances flag to enable homoheteroscedastic tests improvedjavadoc updated user guide to reflect changes. +Wiki is a very nice place for collaboration but it bitrots quickly enough to scare casual users awayAnybodys willing to take a stab at more or less formal docs? +Merged to the branch in revision . +Building on top of Bikas patch +I skimmed it +Reopening for the addition plus fixing OOXML Word docs to also coalesce adjacent that have the same style +Thats true +I used an index of M documents +Doug thanks for your comments +I cant currently reproduce it +Updated with filename change to and some updates to README +Thanks Dmitry for the patch +Thanks JiaFixed in NIOCHAR module at repo revision rPlease check it was fixed as you expected +This is a final version of the patch ready for inclusion +Fixed. +reopen for is? Where is it from? What is the licenseHow is it to be used? +Reading over Phillip says Not to mention that Configuration objects get copied along so its hard to make sure that a configuration change propagates to all possible children +Added external method to set user defined method to call +I didnt know the process +Thanks Owen! +Now the plugin installer gbean tries to start plugins it can find using its after installing a plugin list +From the tests can you come up with what the current overhead is for allocating a chunk of sequence numbers on whatever hardware you are testing on +Attaching patchPatch applied +If it is anything then maybe a design bug or maybe error in reasoning +A version of the quickstart with wicket dependency updated toMay be this was a refresh problem now it worksThank you very very much for the fix and best regrads from Germany +The new names look fine +WillemThe directory that you mention is not located under srcmain but under srctestwrong srcmainjavaorgapachecameldataformatbindymodelfixsimplecorrect srctestjavaorgapachecameldataformatbindymodelfixsimpleHiDo I have to do something now +Released. +Thanks Eli. +Huh? Then why are we considering support for doing that something wrongI guess you have not yet read my comment just it is not a runtime feature only a debugging feature +FWIW the problem mentioend above is indded related to JVM issues +It does not say that only the thread modifying it must be synchronized +A quickstart to demonstrate the applies to + +Fixed with regression tests +All bindings for that service are unregistered but the service itself is not unregistered +Forgot to set the Fix version version +For users these situations will always be clearly noted in the NOTICE file +bq +thoughtsOne more consideration +So double check you are white listing your corporate intranet site in and +With that commented out +Isnt it possible to do this as part of the JOBSETUP task itself +I want to see if I manage to come up with asomewhat cleaner structure for the +Its very helpful for loads +This was written before threads were implemented + +I think this is a valid issue +Final version with some code formatting and agree with Doug +Fully qualifying the export path to hdfs makes the job work in hadoop +Thats something which I had in mind +I made some incremental changes on top of what Ning Liang contributed particularly support for void and oneway method calls as well as a simple test suite +Unit tests and new recipe implementation +Sean The biggest changes are that it uses a long instead of an int for the counter in didnt I use a volatile long alreadyI hope this will become part of the next release +Removed from manual in move of all unresolved issues from to +Everywhere else starting indexs are the norm. +Right click on the file and then Team Create patch +All jars in those dirs are prepended to the boot class path and thus prevail over versions of the same classes +svn ci containerSending containersrcmainjavaorgapacheopenejbassemblerclassicSending containersrcmainjavaorgapacheopenejbconfigAdding containersrcmainjavaorgapacheopenejbconfigrulesSending containersrcmainresourcesorgapacheopenejbconfigrulesAdding containersrctestjavaorgapacheopenejbconfigrulesSending containersrctestjavaorgapacheopenejbconfigrulesAdding containersrctestjavaorgapacheopenejbconfigrulesTransmitting file dataCommitted revision . +It will take a couple of minutsFile parsed in milliseconds +Updates committed in revision +These are are changes I made to get this to work +But and are not +This will return the proper error level when used with the attached +FWIW it looks like a stack overflow since the reported backtrace repeats about lines forever +Fix is the include check in s appendfeatures was not checking the right object +Do we have anything else? Any exception in the log file +Thanks GeirOlav please verify and close +VerifiedLeo. +No test included in patch since it is only about error message and there is no good way to test automatically +Grails now supports custom JSP tag libraries and by doing the above you are overriding Grails built in tags there is nothing we can do about this really +It works without viewfs setting +I have changed the affected locations with and + MarekPlease pick this new build in your seam buildand i have removed all jta dependencies +Patch to remove unused include path options from CFLAGS +Thanks Hazem! +How about the nightly build if we include this on the site? Im not familiar with it enough to comment +More generally Im unclear why we would have null keys in DK shouldnt you use a Token if you dont have key information? using MINIMUMTOKEN for both and values has always been confusing +This patch applies to trunk +Hey Vadimthe problem is the serialization of null which will cause an error +At the time of release or when someone is interested they can generate forrest by passing Dforresttrue +Im not sure why we dont have a response +Im wondering if he used the Sun JVM? I have heard there are some differences between the IBM and the Sun implementations +We put very few jars in our distro +Messages changed to properly information +By analyzing null response I discovered previously mentioned error +For those not present you will get a warningIt required a modification in the data collected during the resolution process +A Writable that I want to check in under test data +bq +Ported from trunk to branch at subversion revision +this is of course not a bug in hibernate +Is this or +Other than that I dont have any comment on test thinking of my little to no knowledge on solrs test frameworkI remember I read something regarding df in distributed mode is a highly anticipated feature I dont expect that can be done easily Im good to have a bug there +All necessary templates were created in order to generate summary table Component CSS Selectors for example with the following columns classes a lot of RF Components havent +Wow +I dont know then +There is however a file as well for end users using Spring withBTW We actually need a better way than having separate files +New patch which fixs s clipping patch has been appliedPlease verify +My intuitions about goals +so fixing a issue is would not address your problem that users and acl exported from cannot be imported with jcrspi. +In this unit test the user returns explicit Response object with the defined media type +I updated and merged +Changed the Graph API to improve the methods of getting blocks of children including adding a new technique for getting the n children that follow a sibling +The user should do this himself +pigcontribzebrasrctesteeorder +Reset target fieldschema in copied LHS fieldschema in so that uid of inner schema is not +See also the duplicate that has an alternative patch and related commentary by Claus Kll +Patch applied in branch +Thanks for the review lads +Duplicate The problem is a duplicate of an existing issue +Committed revision . +Just about to update to another version +To build and run attached sourcejavac cp java cp +Checked into could you please provide a patch for trunk as well +Java doc comments for all four subMatrix methods should read Rows and columnsare indicated counting from to +the patch looks good to meThanks a lot for looking into this tricky issue Todd Colin and Andy +There are no features nor tests we at least need the feature you want in there and with a stable name +Looks good comments on the forum thread +Im not working on this +Brian would you like anything to go to the guides in relation to the examples +Same cause for this and most likely +Its probably running out of memory slowly +Committed to trunk +Hi David +That misses the entire point that user preferences are dynamic and change over time +Noticed when you build Xalan the following is output Project versionD Revision It looks like the revision is equivalent to a daily build number +Duplicate of. +Will get back to you once Ive had some further feedback +Done in vote that the readme should mention the need to run ant jar or ant +Applied +once is fixed the repository will be fixed also and we wont need to change anything in our side +I would be nice to have some test closing all issues that have been in resolved state for months or more without any feedback or update +yeah thats correct +Anyway the next release shouldnt be that far away. +looks liked it downloaded either incorrect file or corrupted file for some reasonAlanDid you face any such problem +For each VMdetach OS iso +Fixed in Rev +Even if they are i think thats a fine thing to do in a release +In fact the above example is hex +Added keyword Thanks Roberto for testing the patch and fixing the DTD +I dont see a doc issue for the property +Knut I will make the suggested change +Patch v makes a minor change to to add a facility for creating a with custom master and regionserver classes by way of the methods +hovers at about files in store +the diff not sure how I would go about adding a test for this as it would require command line decryption and checking of the filename +Hi Artem main intent is to accessed shared queries mapped in the Modeler where you can set column name capitalization using Row Label Case dropdown +Thoughts on this one? Can it be closed +Applied + +fixed in close of resolved issues. +Verified. +I think we should follow HTML specs +No problem thanks for your feedback +We are plugging in our commercial clustering engine into Solr via Carrot controllers so if you change package namespace we will no longer be able to do it +Thanks for the reviews +Ill commit it once I hear back from you +Placing the tree component within a panelGrid and using the paneltreeId comboUsing a similar mechanism as forceId and force like select in a dataTable +Rechecked the test and all seems good from my endCan somebody else confirm that there is a problem please +patch to remove a test case added in as that query will no longer run in accumulative mode +thanks luca I take this +Its also possible that they appear in multiple entitities will which make it trickier to solve +This can happen on any other directory with denied access too +For me none of the outcomes are the desired one +can someone please review? thanks +Not just semantics +A partial resolution Ive added the ability to disable filters as that is the only use case we are currently aware of +Again I can do the accepting testing pretty quickly over here as weve been using Marks patch for several months now +I would like to view the java source code +Did you include all the files in this patch? I dont see anything beyond the included unit test +If not I think it should remain open until one of us finds time to do that +Thanks Stefan your Mirror has now been added and will appear after the next lot of updates. +moving this enhancement to to CR since we have to focus on changes critical bugs in Beta can do bugs in to BRANCH and in Tools +pushed +Ignore the last comment wrong jira +Thank you for checking. +i am not sure he got itsimilarly i was not sure he got routine activity + the current fix and whether we shouldcan use something similar to what was original suggested by this Jira +modified the code and added srcmainresources file +Sorry stupid me +sound good +I am still unable to commit to SVN repo for online docs on codehaus mention that over SSL authenticates using shell ac passwd which is disabled by default +This should do it +If something fails every time then that would be different so let me know if it continues to fail in the same place after +I have used wrong version +I quite agree with some of your concerns and would like to let you know that there are major improvements being done to build system in release +In HQL notps +I cannot resolve this issue seems like a JIRA bug read past this is fixed in deltacloud version fddaeffeaeccfRun command on instance instCommand lsCommand outputThis is where the output would appear if this were not a mock provider +So the test would have to be changed for phoneMe in any case maybe by settingIll just disable for now for JME as you suggest +If you pass the blob directly use the wellformed option on xmlparse or dont use a streaming result then you wont get an exception +There is already an object named edgbindefaultcache in the regardsAnna +Rather this issue is about improving the messages for legitimate warningsPlease and apply the patch +It seems like the regressions werent related unmarking as patch available +By splitting the waiting tasks counts and the utilization of task trackers can be easily computed +With the current design Im not sure theres a good way for the token to get access to the RMs secret manager but yes that would be idealThanks Ill wrap up the patch +At some point we are going to have a custom namespace and that can take care of the magical beans +Hm +When I first worked on the code I didnt realize it was possible to pass the in to the constructor however we can use this to prevent interference between realms +On the FTP site there is a zip that contains the core and the dmalloc log along with a couple other debugjbosssupport got this today +Maven repository managers also allow you to delete artifacts in private repositories while this is not possible for Maven central +I may make some broader changes to some of the other you for testing that change I was afraid that might be the case +please reopen if you dont agree +Added fedora EC compatibilityAdded EC detection option +When doing so the user does not want to unzip the document every time again load the file as stream and parse the XML after every saveOn the other hand I do not think it is a good idea for the user to open the document multiple times with write access due to data inconsistenciesDo you have any suggestion to improve the current situation +Hmmm thinking through it I cant think of a realistic scenario or significant scenario where this issue is relevant +bin srcbugsxmlc Its either way too late or Im missing something because the following tests run just fine +Ill try it with without doing any modification and will let you know +Simplify it to have just ZK path for live broker +So it only issues a message and processing continues +This fixes namespace namespace and attribset +Rebased the patch fixed issues renamed the meta column name +Maybe clean should automatically do this +Right thats probably the easiest solution +But I dont know how onerous it will be to fail on jdoc warning yet +Attachment has been added with description Smal file which triggers this on one machine but likely the problem is has been added with description Screenshot of analysis with our production analysis has been added with description Debug view of when the problem happens with very strange a lot of investigation I found that this is related to the following things JDK patch JDK Flag XX It seems the flag has a bug in patchlevel when using later patchlevels +So now I can see that the are failing +Provided that native speakers of these languages agree I think that this is great incremental improvement +project? system +Can you please provide a minimal test case for this one that passes for you against and fails for you against? Ill make sure to fix this in an upcoming snapshot as quickly as possible +Fixed a minor last minute can use the enforcer plugin now in a profile +These issues have all been corrected for the EAP and EWP tags +Reopen for further changes as part of the the banch and merged to enhancementscon release +bulk close of resolved issues for release. +We needed the headers as specified but I dont recall the detail of what the issues were +This also affectts Sybase +Why is this issue closed? It definitely is still a problem for me +Out of curiosity did logic change in regards to disposeclose and aborting the thread since Feb ? Perhaps related toIt seems as if when you dont set the Dispose on the Session it defaults to s nowJust wondering because what didnt work with my more complex code now is working fine +Thanks Dhruba! +Pls commit and we can start working on an RC for to trunk +It was for verifying testing +for the dynamic partition doc looks good +Yes sorry about that +Thanks for taking the time to explain +Marking for for commit to append for branch +patch committed under so that a testcase can be added +Logged In YES useridAttempting to upload new test case +I noticed this was dropped in the pivot from to +filesdcardmybundles becomes referencefilesdcardmybundlesverify that it is working and that the jars dont get sucked into the internal storage? If so Id say we can close this issue. +to this +We dont do any testing on no versions any longer +Ive attached a patch to remove the recursion +Thank you for the patch brooklin! is this committed to branch as well +Wrapper scripts broken when hadoop in one location andhbase in anotherM srccontribhbasebin Fix line lengths +This WSDL document has an inline schema that imports another inline schema which it refers to by a fragment id +install +Is there still interest in this issue? If so the files attached should be converted to a patch and we should discuss how it works and would integrate with other languages +It doesnt wait even for the serialization CRC and copying the RM into the commit log memory mapped segment +I looked at the replication handlers and they look greatI wonder if we could factor out touchIndex to a static method and share from and +Integration tests already exist +radu thanks a lot for the fixi dont need it backported to v since i was only evaluating xmlbeans for a project that needed to work a lot with the attached xml schemas +Is this behavior acceptable? Could you provide a test case +These imply changesto the docssvn status outputM javaengineorgapachederbylocM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmasterM javatestingorgapachederbyTestingfunctionTestsmaster +I will do that first thing tomorrow! I changed it to loggedIn because success was too generic +I close this issue as it is linked to other issuesand I set Fix version be better to not close them until the dependent or related ones are not it seems like the feature describe by the JIRA is fixed but it is not +Sanne Yes see +That at least will eliminate any confusion for users who attempt to run the example with HADOOPHOME andor HADOOPCONF set +Ive just committed this. +To make matters worse some of the setups are blatantly copied +Yep that seems to be it +Changing between JDK changes orders of filters configuration thats why the problem reproduces on JDK but not on of a bulk update all resolutions changed to done please review history to original resolution type +Sample XML which is causing the error +So the boolean would be what tell if the CAS applied or not +I am on it +Im gonna revert the change as too many cases seems to behave differently +Do you think it will be better that for these events also we retrieve the topic names from metastoreAssigning back to Thomas +We are still having some unit test errors but I guess there are some limitations of the underlying DB that wed have to live with +My cents will be check in the fix +Doing a clean project compile takes around minutes +Right lets go for a warning for nowI like the idea of taking care of the analyizing step +Oleg sounds good +Missed that thread the first time through but happy to help out on this +My test case always passed by itself +I try this in and I get a but this works fine in version +Yes would be very helpful to add a JIRA for documentation +Hey fulin in the future you can set yourself as the assignee to an issue if you are working on it +I would still like to understand what this plugin is supposed to do +Has there been any improvementNote that his ldap creds and group settings look fine +Unfortunate +Fixed in r thanks for the patch +Please keep in mind that you need to develop these stubs in the absence of any JRE class libraries +but Im not sure how useful this feature would be +Bulk close after release of +Id like to completely drop the jca context unless it is really needed +I modified the on Friday so it should be from a distribution too +If we ever want to make this change we should have a design document first explaining the semantic of pluggable placement policy and how all block placement related components should change to support the feature +I tried to apply it to the current trunk and one line of the code that would cause the above error failed to applyIve updated the patch to work with trunk and added a few more safety checks +If you can find a way I will be pleased if you can share it with me +This is copied from the example xconf in the lucene block of a current Cocoon checkout +It is a JDBCJDK feature is that rightI tried to look back through svn history to figure out when these wrapper methods wereintroduced and it looks like it was at least years ago +I restart with fresh install of JBPM Full installer and cannot reproduce for reopening I try to remake my full process and after the error happen again even in default installation I experience the same error. +I didnt know that existed +We recently took care of the split packages so that should no longer be an issue +I committed a simplified version of the conversion code to the maintenancebranch which will hopefully circumvent the JVM bugAnyway you should get the JVM bug fixed +Thanks for the helpful description Knut +push to +This is the fixed patch +I will close this. +Working for me thanks +It is still recommended to provide full type signatures for properties when implementing interfaces. +the patch for fixes all these issues as well for output +Hi Olga just curious what made it fairly disruptive so I can make sure it does not happen againI guess we share the same goal getting pig into a shape it can be used in production +Verrrryy Nice feature You can specify multiple keys for one user just specify multiple lines for that user! THANKS GUYS! +unchecking patch available as fix has now gone in to trunk andmibmsvn commitSending javatestingorgapachederbyTestingfunctionTestsharnessSending javatestingorgapachederbyTestingfunctionTestsharnessjfouTransmitting file dataCommitted revision . +Still OGMs code is not sharing the FGAM across multiple threads explicitly but its possible that multiple Sessions ask to the same for the same FGAMSince the is supposed to be threadsafe Im assuming that this is something that should be fixed in Infinispan rather than forcing OGM to avoid concurrent reads on the same keyTBH Im glad youre on it as it doesnt sound like an easy one +Is anything not working with IE? If so is the current trunk affectedMaybe we can close this one and create a new issue with some details attachedRegards Florian +This and previous curl statements did not work due to reasons I did not fully investigate when the team member who created the the json Stefan Schmidt leftIm resolving this now as complete ready for release by end June Any extra info in actually creating a pizzaorder using curl would be a bonusAlan +Thanks for committingJust curious why were the additional permissions needed? Because tests should not behave different with pitest +Thanks Tobi for the information +Thanks a lot for the contribution Sandy +patch applied + +Tested moving source model into and out of folder previewing that model before and after moving +log in as the new user +to deprecate it in keeping it in removing it in + +I think it could help +Thanks +Any s for this? Its just removing the and protos combining their content up in the +Waiting for feedback before I go onwards +Clone issue created. +Because of code has been changed quite a lot after this issue reported now seems to be solved this problem + adds support for specifying specific rules to be applied to specific locations in the request and response +Is that true? Also I dont understand the reasoning behind moving the pedning syncs up to the servers +What happens if you remove the trailing slash on GROOVYHOME but leave the conf file in its original form +I added the other translations as well +Not all the methods are implemented yet +Committed +Please open a new jira issue and if possible add a mini project to reproduce problem. +An improved needs a bit more of thinking +Ive also tested this patch on Q on a real cluster +There is no surefire output txt as archiving artifacts is failed with +Added to the maven branch below projectsmp. +Finally completed by adding support for uninstalling resources installed for a bundle if a bundle is stopped. +Simply selecting and then clicking away is not enough. +oops i have attached the log file of corz not attacked my bad +Im just not sure about when Errai should do this +Craig my web app is deployed on to shared container with other web apps +thanks you made us about it including the datasource or not will bring frustration so we wont ship the bundle version for HA good thing about the bundle version is configuration this is not possible for the HA version because of the requirement of a unique datasource. +Hey Maxsorry didnt see your comment yes it is a build from source and I just tried with the eap come with the jbds things work wellso this issue can be safely rejected thanks. +I am fine resolving this +The only reason we need to have the activeio in for the tests is for the and classes which arent actually unit tests should we move these someplace else so we can remove the dependency +running tests +unfortunately your proposed fix doesnt solve things +Awesomes If youd need inspirational advice Im on larrykk skypeI do need this issue to be fixed as Im running on my own patch right now + to master. +were always part of the API in +What was your resolution? Have you already submitted your change? Can you please share your fixRegardsRahul +However this plexus code appears similar to code in and and therefore should be a nonissue for Ant +Thanks done! Enjoy. +patch for the is available in the file patch for the issue is patch for issue Hive is of the issue is attached +Its cant be implemented on really so scheduling for +Yes Stefan Studio version will support Apache DS and file +It should simply keep the declaration order +Good point Alan Ive made that change and committed svn revision +Hey Jan I can try to take a crack at it this week +The removal of in should address the invalid Configurationcluster ID used for a client in this case +Hope that HADOOP change to can be picked up this wayIn detailcopy build build tobuildivylibtestbuildivylibtestbuildivylibcommon apply your hdfs patchAfterwards I ran ant Dtestcase multiple times but still getting the test failing +Hi GertI guess you mean we SHOULD add this to the specs project rightAgree with you that we should put this javax +Daryn I think the current version is fine for now and we could create a new jira for the follow onOtherwise this looks good to me +Patch v restores changes to to and trunkThanks for the patch HonghuaThanks for the review Lars +Code has been reviewed by Hairong +Re vs etc +Especially the document parser really loads CPU RAM and DFS while the mappers crunch away +Unintentionally closed +Linking which deprecates the remainder of the spatial contrib that is unused after all the changes +Many thanksOleg +Well have an M next week or the week after with a rather short road to GA following +What features would this component test for that jcrspi dont already cover +Can you lower the priority its no longer critical with this workaround +Gao Im with Ted that you should not duplicate code +Jonathan Would it help if somebody would provide a patch for this or are you against it +CP fixed in revision fixed in revision . +I have a ! in my password and the release always the fact that the release fails when the release is launched through jenkins the password appears in cleartext in the logs even if the Mask Password is activated +To be exact we dont refuse SSH but delay process SSH after master is initialized +Let me know if you have other questions +Aaaand +A lot of people consider wildcard scanning to be a significant antipattern in for instance spring configuration +Merged into in revision after releasing +Should it be updated or lastUpdated so that things match exactly +Bulk Close for all issues before incubation +Added the mailet and the matcher +If we had bindtoallfunctioninginterfaces that it would make sense to do what you ask for in this JIRA +Patch includes this and all previous changes +branch no more under active of a bulk update all resolutions changed to done please review history to original resolution type +m setup properly +changing the xsddtd locally is not a viable solution for this +One oddity is that since and long are both final types it seems odd that any type info would be included with specified settings that is wrong +The file contains the patch in diff format +bq +If possible Id like to test on my large cluster +Although I cannot guarantee this this issue is also the likely source of +There is the same problem when height is bigger than maxHeight +Verified in +I dont have rights to the issue +Also the build only appendsthe console authentication and doesnt change anything else in this file +I wasnt paying attention. +This is something I changed recently to help on in test against SVN I think I fixed it +This means that the caches work fine when first added but as soon as Railo is restarted only one the first one in the start up routine works as the others are ignored due to the duplicate defaultCache +done +So what is bad about adding? any reason why it should not be allowed +I do not feel these stories fit well together +TedThe patch is meant to be applied to branch only +Mass closing all issues that have been in resolved state for months or more without any feedback or update +The new version patch for will be uploaded after get fixed +is the change of the test results +This package is now cleanCommitted revision +Thanks a lotCheers the patch to distinguish between null empty and an actual result value +Well likely take another pass at how were handling the dependencies as we refine how packaging hopefully reducing the number of places were dependencyversion lists are stored +Ok thanks would it be possible to attach two versions of a sample bundle you can also send them directly to me if you preferIn addition you could try latest from trunk as I improved some out that this wrong behaviour can only be reproduced with +The NPE Myrna saw was in store whereas the one you saw was in the language layer +Thanks stack +Yes ant tld that is contained in those versions are supported +Is my understanding correctRon +would be great if you could provide a simple exampleNot so sure simple and are words that can be used in the same sentence An oxymoron perhaps +Filed PMR K with IBM +Theres a quick fix for this available though which involves checking the status of the global component registry to see if its stopping +So we have several options +added checks against the +No need for IT. +From Mike Is it possible that the transaction was recovered prior to the test waiting for recovery +needs a grouplooks good +has been marked as a duplicate of this bug +It just happened +Thanks Navis +I agree that its better to keep this code simple and that a is preferable to a +This may not be a problem in production but it may make several of our replication tests fail so I am changing priority to Major +Case has nothing to do with page caching +Most importantly it adds little functionality to the highlighter component which I believe a lot of people still usingI will add another patch tomorrow which catches the exception + +As a first problem Harmony does not have such file instead they have +I read the code to be a shutdown due to a is a shutdown due to some failure in the serverIn that case requiring a userpassword and permission seems wrong especially since the code thatstarted the server is the code that is calling this shutdown and a userpassword is not required for startup +Done +I think you are complecting what the scanner does with full maintenance +Its basically just to get it on the road map +Not actually a bug. +Sorry no patch just modified sourcecodeAdded message attribute +Patch to use a different JMX type for index for catching thisCommitted to trunk. +Bill were swamped at the moment +There was a bit of funkiness I had to do IIRC it related to spoofing the multiplexer channel factory so it would use the many arbitrary channel configs that exist in the JBC testsuite +Ive also seen this behaviour but only if I have the Configuration Admin service also that everything is OK with the trunk thanks +Heres the patch that adds the missing short and byte sortingWill commit in the next day or two +dependency are found +Thanks Bernd Hi PaulPaul I plan on diving into this for the weekend +Please rebase add a test case and you are in +basically building and copying +Transitioning to closed. +Rebased patch +And you can always unsubscribe once your issues are resolved +Do I need to use my apache email for that purpose +As you said UUID wont solve anything there +Patch contains depends on for by attached apache Earljust a quick note to say that this works like a dream for meTHANKS +This needs further investigation +The bisocket static maps are managed more carefully and a unit test has been added to in the bisocket transport and also in have been turned into static program with a loop in which bisocket connections and callback connections were created and destroyed was run in jprofiler and there didnt seem to be any further memory for cruisecontrol results +Bug is in CXF +Usually refers a beginning university course and is used colloquially to imply that concepts are very basic and should pretty much be known and understood by everyone +Yes Ive closed and made the refactoring subtask to depend on this one +fixed in revision ve verified that the patch has been applied OK +libs +In fact I was reading the Lucene index format which dont specify any limit in the integer lengthEffectively returning a Java int it is useless to parse too many bytes +Ashish thanks for your patch its in rev +This is forever ago + +Im good with removing the ajax stuff since its not improving the user experience the way its implemented +note if someone can explain me how we can do this without breaking every release of seamgen and jbosstools when something gets updated then I am very interested +Fixed some small comment typos +Ported zip sequencer to into the branch. +Now we still need something similar for Windows +Not sure my example is still reproductible with the latest SVN for the Benjamin +The only place it should be touched is in +Using the android command you can edit the allowed memory per application in the AVD definition +AFAIK the paths with forward slashes are valid on Windows +How are you calling the source generator? And are you generating sources for both XML schemas +If the JSF servlet were not used then the would not is therefore an issue in JSF or your servlet container or your configuration +So keep this open +The bad jar should be replaced w the temporary good jar until the Yoko group can fix the specs +Source jar also includes docs +OK merged to the branch in revision . +For the moment lets disable the test +Whoops sorry. +As in tryCache + to committed to branch +Ralph or Gary will you be looking at Nicks patch +Added +If you dont want it to be registered by default just dont add it to a relevant fileIMO if there is no reason for creating a separate jar do not create it +Patches with additional tests for multiple selection lists +Does it stand alone or merge into the as chunk? Also why not put WS into the as chunk too? Im trying to ensure that no git repo contains only a single project because by doing so we ensure fewer checkouts and fewer tagging operations +Pavel mentioned filing some issues with Eclipse +Sorry in either case it doesnt work +Ok but Row was replicated to nodes C and DWe loaded the snapshot on C with so if Im understanding this correctly that sent row to D +We can separate them into +Thank you Raja for the review. +btw this seems related to issue can also unstick it by building it doesnt work because i check project state before add it in queue +This then resolves this issue. +If I get a patched JAR Ill test the patch this week +that looks like memory leak? something is holding reference to some allocated and does not let it godid you run profiler does it show what are new objects allocated inmemory heap and why they do not get GC after each request is finishedthanksalek + +This would shift load imposed by clients from master to zk quorum +I will move the priority to Minor to indicate the same +Thanks for what would be the expected handling? Just catch the Exception here or what? I have no enviroment for testing ldap +in ER there is a single EULA file I thought that the fix would be to just use the same EULA file +I have tested for and it seems fixed. +My assumption is that the conversion of Castors Maven build will be a good preparation step to the conversion of the Geronimo build +Extra time allocated for some learning curve +Thanks Ivan for the patch and Bikas for the review. +How does Karls patch compare to Samis? I havent looked at either yet +Also note if the user doesnt have any security groups an empty list is returned and cached +With Adobe Flex you could take advantage of a Flash Player feature that cached the framework +Seems like it could be as simple as not getting the headers back out completely +add onitemchange event to the panelBar mousemove mouseout mouseover client events to the panelBar onleave onenter client onresize onmove onbeforeshow client events for onresizing event for component client for inputnumberspinner and onslide onsliding client events for onexpand oncollapse events for the events have been added +I have been thinking about a solution that would fix this and +I have to implement this on my serversDThank you for the references + +Please describe the steps to replicate the issue +It allows the unit test play around in the region levelMaybe we shall move these useful code to +YikesIt would appear that my test case was in error and this is not an issue +I had made one optimization in the junit test case to create the indexes after the database was loaded and that is when the problem did not reproduce +Manually I did the necessary changes +If the server is stopped there is no imperative to publish immediately +Currently I am still using karaf. +This bug affects the Release Branch too +never mind the previous comment I was reviewing the wrong part of the log the ran this still an issue using +Feel free to have a look at the current implementation and see if you think there is anything missing for your requirements +I chose the latter because it makes code that deals with this more succinct and it didnt change a public api +However we still should always decapitalize the property name first in this builderRex +Almost forgot should be added to the libphp directoryThe patch to the libphp also no longer installs the +Thanks Vinod +Observe from the above record that networkaclid is NULL +Cool thanks +Please review the attached patch that seems to fix the pb +Done +Am running the unit tests as well right now +It worked fine +Closing old issues that have been left open. +Hi Lukasthx for investigating have no particular requirement for the TOC +Only one path should be taken +Translation +investigating test wasnt able to reproduce the test failure + +The iterative algorithms rename their last directory to to make accessing the final clusters simpler +Hi Ted I just added a profile to the GEP to skip the testsuite as requested +Built my own of vdoclet +Description is incompletenot reproducable directing to mailinglist +This is an accurate statement +Some of the tests fail at certain Similarities so those have to be fixed as well +Great catch since its the collected work shouldnt the copyright be this? Copyright The Apache Software FoundationAlso I agree the license text should match the new style +But coming to harsh comment i thought to raise your point +This patch updates the namespaces to the defaultEnsures that the generated policy elements include a Id with a namespace prefixAdded support for attaching policy references to messages instead of just operationsAdded test cases for all of these +The problem was that if someone didnt add a changenote that information was not saved at serializationShould be fixed in + +Because you made it so? This one is on my list of issues to talk to you about +I will see if i can clear that up in a separate issue +Dealing with missing replicas is a bit tricky with this approach the other downside it has is that it is potentially much slower since validating with MapReduce could cause a lot of data transfer over the network +Processing threads will block on queue insertion but can still accept connections +Works well Dinesh and even better than the PMD rule +This issue is quite old +Fuhwei I cant see any difference between the two patches are you sure youve attached the changed oneReading through the JIRA it sounds like a final patch was intended anyway to remove things like the comments that are there to help review while the patch was developed +Moved back to has been releasedMoving all related issues to the next version +pushed +Let me try and reproduce your environment here to see if I run into the same error +What version of Roller are you using +Verified on +Patch to fix this issue +New checked the problem and it is caused because when ffacet contains more than one child facelets algorithm create a panel that is not backed by any facelet tag but it has to recognize it +Or did you mean the config editor +See the EJB invocations on the same entity requires configurationsas per the FAQ +I commited a patch to trunk in r +Hey Colin +I wasnt able to test it because I was getting errors when applying the runtime patch on the latest Tuscany +If someone would be willing to use it to create the initial wiki page Id sincerely appreciate it +Attachment has been added with description srcorgapachefopimage has been added with description Improved patch for an LRU cache which cleansWe do not apply patches to the branch +Im using +Committed as svn +Fixed encoding +Those two issues are John Wilson fixed I suspect we could also close issue is indeed now fixed so I am resolving and closing there be a unit test somewhere though +now works for by upgrading the console which was temporarily out of sync with the code. +Hi Dan yep just saw the merges thanksNote Nabble was not configured properly so the original messages never made it to any of the CXF mailing lists +We will adopt this as a practice in the. +Fixed in I created a new goal summary which means there is another report page for this as brett report page shows the following general project information like project name homepage +Patch v +Ill fix this +These will corrispond to the various types of testing requirements of the parent pom will generate the actual TCK binary that implementation will use for compliance +I havent given it enough research yet to decide but I read up a bit on it earlier this week and it seems pretty reasonable +This test case application is reduced to what is necessary to showcase the +I merged this to +class missing from above patch +UML expert what next +This patch resolves missing dependencies and upgrades Apache Jakarta SQL Commons to Apache DB + +gliffy for class should find a place on jboss tools wiki or jbosstools to put this info so it is not just hidden inside jira anymore +I experimentally determined that removing the variable was a convenient way to reproduce the problem so thats how I worded the bug reportIll try to find some time to experiment with your patches and see which solves the problem in Nanoc +bq +bq +Thanks Kim! I see the topic in has the title Administrative tools which may explain its previous placementin the tools guide +Thanks Juergen +More specifically the classpath container should be forcing the directory to be created +Probably you refer to versionsI would like to change this to a feature request if you dont mind +pined with label +Thanks for the review Suresh! +The presence of in the descriptor doesnt really mean anything and we should not be assuming security is intended based on its presence +If anyone can take a look and provide a feedback that will be greatly appreciated +Cheers +Verified with JBT. +Yeah thats pretty weird + +Thanks Kiran +We are hitting this again with Sonar +To create the test this issue depends on +The jclouds release is on track for this weekend +this one has been dormant for a while +Test file applied +First page contains page included page in first tab which has the my sittuation header not shown too. +Im not sure but I think this may be contributing to flapping in another case where Ive seen this error +Is this still a problem +Deferring all the from to as it didnt make Release +HiI didnt have much time these weeks to deal with that but since I still dont appear in the contributors list I send a mail to secretary in order to know whats going onBy the way I would like to thank you again Knut for your help during the process of submitting this patch +I have samples for both but I need the ttpsjirabrowse applied to test the Mask case +Please reopen if anyone finds a problem. +No promises on when we will deliver a fix as we first need to hunt down a device which is an expensive venture +I need to make the XML commons issues in Bugzilla first +Committed. +Rob +Above are fixed in the patch +Im closing this issue we need more info to reproduce the problem +Great thanks. +Ron That would be great if you could do it +I fixed the typo and posted the version of patch +Please verify +Bob Brodt has verified the fix on the engineering side +Right +Just looking at the screenshots the improvement is so significant here lets figure out some way to get this in and go iteratively from hereIf there are concerns about any particulars maybe as a workaround in the worst case we could temporarily keep a old admin UI link or something so that you can always do what you did before but with the new UI as a default +This change will show up in the nightly snapshots for +This doesnt mean that I a fan of JNDI only that its easier to write back ends to than a generic JDBC driver and LDAP is designed with better redundancy in from the outset. +Reopening since this wasnt fixed completelycorrectly. +Changes have appeared in Latest Alpha Manuals. +With this issue should only cover the remote case +There are some dependencies between patches and only three patches are independent patches P P and P +Havent yet verified this but it at first blush it looks like a Xerces problem rather than a Xalan problem per se +It would help a lot if we had objects for ids instead of strings +There is no short term plan to put line number there +I know the new UI doesnt deal with cloud expansion yet seeAnd it seems like this handles too +I have added basic support for +I run the same query with limit for a few times and I get the same results +Since this patch has a relatively low surface area and such huge benefits for performance perhaps we could backport it to the branch as well to encourage people not to jump ship for trunk quite yet +However the project Im working on was an upgrade +More fixes for tests +Patch to fix recursion take Still not sure about it +Committed to +As for how to proceed with i really like your idea of breaking out two interfaces one for pre and one for post +I dont think we should do any indenting by anything in the tracker log dump If its parseable and happens in the recent requests console plugin thats also fine I think. +It hasnt changed in years and likely wont +Are you sure these should all be tabs and not possible some of them available in views +Several packages should be filtered by defaultgroovy +I will do it after looking at the test results for the v +Ive made some little changes to the patch and Ive applied it with the revision +I need to find the right code in hive for thisLet me also look into create +Maybe display a reduced console log in the installation dialog window enabled with a show progress buttonlink +Yes we need this reader for example in the bipartite graph matching it would work without thatBut it is enough to add a single reader for each value type so we should be okay with or +Cast the null value +If you finish it before we release beta please correct the version +making summary more succinct and to be done + for +and refactor are one and the same +Commit first step changes to trunk at revision have tried the testing steps described by maojiaAll testing results are good on my local machineGeronimo Application Server starteddeployconnect port Connecting to Geronimo server localhostUsername systemPassword Connection establishedgeronimo deployFound modules car car aspectjcar axiscar car car car car cargeronimo deploydisconnectDisconnecting from Geronimo serverConnection endedgeronimo deployconnect hostname localhost port Connecting to Geronimo server localhostUsername systemPassword Connection establishedgeronimo deployFound modules car car aspectjcar +Werner I was kind of assuming it was changed for a JDO issue +So you can only be testing one patch at a time further impacting productivity +fix this old date issues. +No problem we love users who contributefix bugs +Thanks +We do some tests before integrating it to trunk +Its for Magnolia +I consider it fixed in where you can enable the View Versioning Labels when working with versions of the same project at the same time. +The database created in cloudscape is attachedDetails of attachementThe database attached is created in cloudscape +I had an old svn cli client version installed which seems to produce buggy diffs +Im waiting before opening the ticket against the main maven repo to have this uploaded +I agree +Uploaded this is the patch file over Trinidad trunk for this issue +The issue that precipitated was that if the shared edits dir temporarily disappeared the Standby NN should not enter safemode +plese dont open bug reports with zero information about what you did +The test failures seem unrelated +The new log entry you propose is for the journal? If so im not sure its neededas we already mark the journal after each flush to allow persisted entries to be cleaned up +These strings will come with the translated +But notice this was on a Linux box +Fixed in r +Maybe servlet will make rules clearer +Patch applied +This bug was fixed by +Stack is this code changed removed by the refactoring in progress in the clientI dont think so +This may be the simplest solution + +bq +It was released with the fix in +I have solved this issue by adding Default JRE to the JRE combo +Please check the patch and commit to the branch it is a really nice feature to generate java compliant xmlbeans code +And wasnt that fixed in +thanks Vladimir +Myfaces problem. +But sadly no committer seemed interested in it +Please check the response message and the messages with wsdl +This is a problem +the logic to determine which classes to skip breaks existing apps if there is a mixture between and CDI jars +orgmortbayutil matchesBinary file +Why didnt someone authoritative come in any earlier and tell us this was really a feature? Im not trying to get in your face Steve but with all due respect your reply makes it sound like we were out of line just for having thought this could be a bug +SVD has been once again revamped for and will probably be changed again soonWe need to stabilize the implementation of regular SVD before working on thisPostponing to should be doable without breaking compat so pushing to to low priority +Done +verified in JBT +These fixes will now make the database locker code to iterate on a result set if one is present committed in rThe logic is conditional on get returning null if the lock statement will not result in a some dontCan you verify that this works with your driver +really if you know any program that uses a lock around IO operations Id be interested to know which +So you need to compare with the first tab I sent +There is no unique identifier that is different for each message that is exposed to the user +Should I submit a different patch that calls clear +Mass closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Ive measured performance of various sized attachments with and can find no difference between and +Thanks IlyaPatch applied to MATH module at repo revision rPlease check that the patch was applied as you expected +I think that approach is really hacky and especially as the code is more complex now and also supports sparse vectors +Does my fix look right +I just committed this +I will add thementioned comment when moving it to BRANCHCoult you again crosscheck whether your sample works fine +Any comment or example code +Testcase and patch for binary files in ZK +The attached patch is Sams original patch with following changes Rebased against +I am not totally sure what is happening with the and releases +Those exceptions are reproduced without any patches too +Is this a duplicate +This patch was made against the tag +Otherwise you could have a perpetually bad pom in svn +broker series is EOL. +This is way more extensible and flexible since we dont need to change a method signature just to add a new param +The patch is over my head but providing a super optimized solution to the primary key type lookup problem definitely seems worthwhile it has me wondering if a that works like bug quits collecting as soon as it finds one matching document would be a good idea +Your comment about not accessing the projects admin area was done with a djohnson account you are using and therefore confusingtwo accounts. +In a second step the plugin can be changed to use dynamic references and being usable The procedure is described in the Maven developers notebook and I can do that part +Converted pauls patch to svn based +real life exampleConnection stays open for half an hourAfter I am back and i want to browse the tree it freezes for some time secondsI would like to avoid this freezeI would like to have a setting on connections edit properties could be a field for ldif statement and a field for time in seconds to fire that ldif periodically to keep connection alive +to what Claudio said +This bug depends on and +SorryGail Badner could you please delete the wrongThanks +Done +The reason for this is that to easily and properly support full Calendar compliancy the commons library just for the ISO support class would probably be required +Will we have another type like an which has a life cycle? Im not sure about this so introducing too generic interface might not be a good move +So it sounds like it is fixed +is it too late to move it outAnyway my principal problem is that I can +This will be in the next updateWe would appreciate it if you could increase your update frequency to between and times per day. +An official release is pretty close they are voting on release candidates currently +Well try and work on that for the release tomorrow but thats not a givenworks for properties and bridges +Ok I can reproduce it now and have fixed it as you suggested. +Well I guess I wont be the only german user so the second proposal would be the best way in my opinion +You need to specify the full url for your endpoint and current CXF servlet transport will do some work to cut the full url address into related pathI just did some hacking on it and found it need to cover lots of situations such as address start with hostname or host IP address etcThe parameter could be a good solution for your caseWillemJust realized Dan Kulp also provide a parameter to force the endpoint address start with you specify base addressI just added a test case to show itWillem +Never mind found them myself +If only seed data are loaded these checks make sense +committed as r +Ive increased the timeout value and provided a better cleanup for the MDB tests +This is not an issue with Acegi Security so the issue is being closed. +Sorry for a dumb question but what is the double submit problem +It works now +The default is just defensive programmingYes contains all status information like state progress etc. +has a directory structurebinlibexampleslibwhere as the dev +Thanks VinodI fixed to pass along the event to which is more appropriate +Well even if it is better only in it may be worth doing +Our customer moved to SFTP which seems more reliable so I can no longer provide feedback on this issue +I looked through my email and couldnt find anything but neglected to send an email until today +The change went in as revision Yes it will be great if Kim had any comments on the text or the name of of the readme files +Unsupported compiler version. +verified on + +I thought it would be good to leave the ping timeout in just to be sure that if someone did disable the progress timeout with a we could still detect a task that has something bad with it but I can see your pointI updated the to include information about a value of +Aside from that a stacktrace merged into the exception makes it hard to track down the exception on the +The relative urls used throughout the examples are intentional +Fix to set inputTexts converter to null +I have similar requirement to create a mall kind of site where stores are owned by different vendors +incomplete issueplease reopen if still is an issue and attach a that illustrates the problem. +Therefore it is guarantied that the interceptor is installed +Fairly major surgery on the and classes in order to support dynamic creation of a writer pipeline +The jeeri will also deploy a war that does include the ejb interfaces in the war +Its accessor method is defined as +However we can add a comment where we check for the clientID size to clarify this +resync with trunk +Proposed changes for project list +The bits are still shipped but customers should use the new admin console instead +This still leaves us with the issue of coming up with a versatile interfaces that provides enough information for most providers +luigi can you try this patch instead +Fix keyHash loop +where no problems exist in d really appreciate it if someone could look into this and find a solution +Linked another critical ISPN issue which it seems corrupts cache entries when using partial deltas if eviction is enabled. +Ive committed this +Moving to in preparation for release +The GWT console will focus on jBPM +Waiting for review to Mar +Same patch with better variable naming +Changed parser framework scoring to now include both statements and keywords +Commited the contributed GWT servlet and sample projects in Rev +Triaged for +JIRA isnt allowing me to do it +Ui bad news + has been marked as a duplicate of this bug +The earliest Spring Integration release that should be available after is released will be +Purge this facility from our template and the servlet +Hi Olivier The images were already been attached images added in rev . +Test in description doesnt crash any remaining issues look like a duplicate of +Sorry about last post +and? it failed +Test case added in rev . +I should try harder next timeClosing. +I modified the test case to make sure merge is never called when the concatenation of the two runs to merge is already sorted +The full bundle is Kb with the relevant compendium and Servlet API classesCan anyone suggest good ways of exercising the service? What criteria need to me met for the project to be added to Felix +Committed the patch with minor modifications in revision Lets continue the discussion in +Ill try to build a minimal example probably on Monday +on that last thanks. +ship it +What does it take to get things fixed around here? Thorsten has supplied a patch is it not good enough? Are there issues with it? Is something more needed? I have no idea because theres no feedback or activity on stuff around here +How does Jackrabbit do that +Cleaned up a bit +Symlinks must be resolved relative to the clients filesystem not on the server itself +The attached screen shots show the broken layoutThe CSS class is a decorator it is NOT intended to be used alone +patch to implement requested feature patch seems to handle only +Also modify the maven build for to add the ode directory to the jar file manifest classpath +Also filed +DavidThanks for pointing that out! I didnt notice that file +I had gotten so used to the current names for the groupId node that I missed how new uploads should be the main package for the starter kit is I think that may be the most appropriate +Executed the Roo commandsscript setup provider HIBERNATE database FIREBIRD userName test password test databaseName tmp that Firebird appears to require JDK as I got a javasql with JDK Alan +On Wed at Saliya Ekanayake wroteTheoretically should work no matter how the OM tree is built +This moves the core plugins LICENSE and NOTICE files to the location indicated by Susan +Your fast response is greatly appreciated +Fixing this for xml persistence +Otherwise itll go intoIll let it hang out a little while before committing +Hi LudovicThanks for the quick fix +Thanks Devaraj. +the fox suggested by Pascal above +HI Asela This is an issue +If we got rid of would make this easy to implement +They both use which interprets an Object return as a single model attribute. +Sorry for the mishap Stack +Thanks Alexeypatch applied as expected. +Anybody can point out some patch is available or can you help me in fixing thisThanksNava + +Reopening for discussion Im not happy with the tradeoff that its harder for usto improvemaintain tests at the expense of the special jars +Thanks for the quick turnaround +Can you provide a simple testcase? +Unfortunately I have to revert this change +Should be fixed now please check both the DOM and the clientevent params +Hey guys Id like to contribute to the jruby FFI project Im one of the closet jrubyists +Updated in revision . +So your fix seems to meet my needs +edited show command +Let me get the other two logs +If the managed bean class is a generic type it must have scope Dependent +This patch is not for commit as the upgrade test currently fails because of +Since the issue is fixed for now +I still say that it is better if we just ensure that the two classloaders dont try to use the same location +The Namenode server threads do not catch +Even if the client is very busy I dont think it is expected to loose the lease +The new base test case has now support to only run the tests twice for a Set of test names +X itself can run as a program today but we dont have a good way to start all the X processes on Windows machines +Did you enable icons in? On gnome are disabled by default. +see restorepointintime from conf Jonathan I think the request was to skip a particular statement while restoring from backupExample restore but skip all delete from CF where +Thanks. +Will commit if tests pass +I have a feeling there is a more elegant solution. +This approach works for me and I like how it means that each command now only appears in one list in the script so that when adding a new one only one place needs to be updated +Patch removes connection pooling jars from maven dependencies updates patch checked inThe issue remains open in the hope of finding a way to make it easier for users to obtain the jars that now must be manually copied to libjdori +Only that way can you achieve enlightenment +Still not sure if we really need a equivalent thats in any case much harder because the ingeststatus table has no column at this time containing the repository connection name by itselfr +It seems the changes are more involved than what we expect can you run the whole unit tests before submitting the next fix +The changes to and to check for compression can possibly be reverted +i am able to make it work standalone +Well see about addressing this in +Jukka +Perhaps not in the current release but it should be pushed to the next instead of rejected outright. +I agree we should open a new Jira for the point +In the mean time you can process the filewith a tool like which can read the format +whats your opinion regarding the observation classesIt would be nice to have them in but Im fine with moving them for the time being to get this issue resolved which I think is more pressing +So I am not keen at this point to remove itWe could take a straw poll on the and lists to see whether people are in favor of thisregards Malcolm Edgar +As is will save a bunch of churn +So basically the bean drop down list is never filled with anything +At first glance this looks like a great first step towards ridding LCF of +reopened to fix Fix version because cannot be used for resolving issues +It seems fairly straight forward but the test you added in seems a bit too whitebox for my taste +But we also use Oracle Access Manager and number of its attributes are also considered operational and therefore not returnable when the operational attributes option is turned off +one more novice question Do you want the patch asA a svn diff against trunc +Move to wish list since we wont include this in the upcoming G release +You see what I mean +Something isnt working right for me when I apply the patch seems that installing from the download site may not be a simple test case due to different environmentsWhen running from Eclipse Ganymede I get an extra dialog A feature that you have selected uses install procedures that are not compatible with the current installation support +Metamer was updated so that it is possible to test it. +With httpd turned off the load runs at about K revisions per minute which means it shouldnttake more that hour to bring all the svn revisions in +ResultsTests run Failures Errors Skipped +Can you check if this could be closed +And possibly remove throughout the default indexing chainI like removing this as theres many loops per thread right now its not easy to glance at and know whats going on +Im going to commit that in a few moments +Thanks Thomas! +Sounds good will do +Yes a bit strange I agree I reopen +I did not work on that yetDoes anyone know why this plugin is not included in? Should this plugin need to be rewritten to be included in? I did try few things but the classes are not getting compiledMany Thanks for the help +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I thought I had but apparently not. +Closing this record since is now released. +pending can you check the javac warning reported by the +I think that the skipTests property should be set by the build but I was under the impression that we already did so so perhaps something is wrong there? Our installation already skips the tests AFAIK +Hi JacquesYes I found already existing label as belowPayment Deferred Release Cancel may be there are more label for same string with different keysFor my purpose for Payment string sounds good but others I guess wont be nice to fit inSo what do you suggest should I put all of them in Accounting as so onor in Common and if in common what should be the key pattern hereLet me know your thoughts on this +If there is a problem please open a new jira and link this one with that. +Marc Antoine any progress on this? Do you mind if I apply the changes suggested by the reporter +The installer listeners jars need to be rebuild with +Closing out the issus for the release. +Thanks for reporting +Thanks +Quick update I havent had time to progress these issues because Ive been working at a customer site but I hope to devote some cycles when Im back in Malaysia next week +Ive just committed this +Completed in releaseSee for patch release JIRA +Thanks Bryan! +lets review this in for a possible soon if it turns to be a huge problem +Solution is to stop doing that + +I thought about adding the in the marshal step +Why dont you open up a thread on the dev list and well talk about itIll go ahead and commit the changes +bulk defer of featurestaskswishes from to +Untar and run with mvn tomcatrun create two As and a B and then try out the finder +Actually its on the top page v +And what suprises do you avoid with the fourth issue +supports HAVING without GROUP BY and assumes the database also allows such a construct +Created to move the new method out of Filter interface +Fixed in trunk +Need to apply patch and run unit tests to verify no impact on functionality +build fails now will look into it +It is not the type of issues that should be resolved any later +jrubyc javac appears to actually be broken right now +Thanks you HenriI think that the only active lately committer from the list you gave is Carsten +Maybe we could stick with the styles in my last patchI really like the idea of visual elements floating on sand +Yes this is still an is not because I have switched to Axis that it is corrected inIt is still important for me and I think it is very important for ll try to test with a SNAPSHOT version in the next couple of weeks +The way the http redirect is used is not valid and causes infinite loops that the browser detects and shows tat committed in the in svn trunk +Done a long time ago. +If this is another problem please open another Maurice thanks! +there is something need to knowthe flashing is in another LOGGING thread and that is async from the testingin the testing codes the log file checking is done after the log object is destroyed which follows right after the flushing +Perhaps it could be backed ported for our commons cli version +The fix will be to first generate the byte code for the arguments that could contain backwards branches store their values in temporary variables create the object load the values of the relevant arguments to the constructor from their temporaries and finally invoke the constructor +If you logout after and try again you stay on the main page +Unzip and run the command in the text file to start the app connect over putter and type test enter on the client to get the simple shell impl to send a large payload over the wire to Putty +It is important that we dont do that +Doesnt quite work yet because is not available remotelyWill provide easy and small fix in next attached patch +Modified files svn commit Sending trunksrcadminguideSending trunksrcadminguideSending trunksrcadminguide +I created source jars and publish them too since theyre super helpful for developers +Ill create new issue for that. +There are places where a running average is needed tens of millions of of them without standard deviation + +Theres no way to prevent this though the closest thing we had in the past was a random sleep to stagger the beginning of the replay +File browser dialog was not implemented +I thought I was working on a second level cache +Adding onto does the trick +It is not a big deal but still whyI suppose the unnecessary attributes of SD can just be default values for views +Working test fix is to remove the item during the return event. +Thanks for the quick review +or maybe the proxies are not getting used +We can still add methods to Searchable and define another interface in contribremote + +Cloned for outbound transparency. +And I verified it shows up correctly as in JBDS B +JayThe problem is still here and its not limited to Tidy +It was used to verify for instance +and attach the output +This might appear trivial but it makes it take longer to see which lines are actually affected by the change +Is it intensionalYes it is intentional +The program posted previously works perfectly with this patch so I attach my new version of +So maybe this is an existing bug +Can you please rebase? Thanks +Will fix at Apache and sync to fuse +Just looking for every varvalue attributes of all parent dom we want to be aware of particular tags then we would use our Knowledge Base plugin but it will cost us of KB for every jspxhtml with var could be a very slow process in a big project +Right +Simon has mentioned this already before +The current status is that I have implemented the basic quorum protocol for writes but still in progress implementing recovery after a switch of the active node +Thanks Mahadev +I still see this issue with Maven +Let see if we can include it in I think this feature could be interesting for users as well. +Hope its a little better +However it would be really nice if the view editor was disabledsimplified if the user isnt in the admins listNever tried this but it ought to be possible to suppress the auth prompt by first sending a HEAD request to the views design doc and inspecting the responses Allow header +Last I checked IE and earlier used to display its own error page for xx unless the returned entity was longer than a certain length +Please state the proposed checked in to cvs for betathe main branch should be updated on DEV as of +Give me a couple minutes I need to check in a new comment and help Ill take a look tomorrow +Freddy yes the path was correctly activated +Similar we have adde support so what we call project examples can also just be a pure maven project a little xml to describe its runtime compatibility and any additional cheatsheets you want to present to usersHavent tested all combinations yet so if you see something not getting enabled and you think it should let us know +Nothing else needs to be done exactly the same as EAR projects +I will need a quick start for this +but we may need to redo some xslt or other regex pattern downstream +This new patch is against HEAD as of today +Could you please make a patch from the project root rather than just a diff of theYou will also need to run ant test and targets +Selector is called as part of package build process it can only be applied to package binary +potential dead lock situation did you encounter any when implementing secondary indexing +A patch moving to automata package and hiding it effectively underMinor cleanups +REPRO STEPS Import a EAP sever into inv that has at least one of each type of Datasource and Connection Make sure each of the DSs and CFs have a Pool JNDI Name trait and make sure the XA datasource and the Tx Conn Factory have an Interleaving resource config prop +I just committed this +If a source document is updated while the target is analyzing revisions the source should be free to replace the requested missing revisions with the leaf revisions of the edit branch on which it livesI agree that the only bug is in the HTTP layer and that your patch addresses it +Just browsed through the patch for now and found the following bugIn the position added to indexEntries should be the position in the index file not the position in the data file +bq +It makes sense to not delete that znode at all +A port of the CPP QMF sample +Thanks Stepan +UI makes a list call with zoneId parameter in order to figure out which offerings can be used for deployment +This also works +After playing around with the modules I noticed it found a bea module that I had but not jboss or ejb and after several reinstalls only the bea module showed up +Did the config look correct to youIm using DB but I cant see how that should influence it as the JDBC stuff is fed the wrong this be some bad proxying +into the code some more it would appear that just changing how expiry is used is sufficient +Thanks +And it will not prevent the traversing of the cache but will add calls to a database thereWe should find something better to keep API stable and contracts clearWe just discussed idea of specialized eviction +In Java s the can only made up by wrapped Bytes +Thanks for this ScottThe path is fixed now in the distribution build. +Ive replaced it with apache boilerplate and redone the patch +Moving to for additional study and work on that case +The point of was to steer people away from Derby extensions and encourage people to use standard syntax +OK thanks +Feel free to try to backport to trunk if you wish I dont have much time unfortunately to look at that in the coming days +It works correct on +Attachment has been added with description The proposed has been added with description Test FO has been added with description PDF output before applying the has been added with description PDF output with the transition resolvedfixed bugs to closedfixed +All these are really driven by the business requirementsAnd there is the need for joining one table with another from the values in one table or merging one table with many tables +Daryn Sharp very good comments Quota is the trickest for the hard link +The SEDA branch is more complex since it will allow processing of client requests +bq +Attached corrected patch +Logically youre right and seeing a rash of server implementations here would not be a good idea +As the problem no longer exists I am closing this record that I previously opened. +Its only checking local domains for a valid domain +But we tested it on our system with same java version it worked +In same +See for migrating the functionality to +Pulled out what I now believe to be unrelated failures into +PrakashThanks for weighing inIf possible do you want to show us your proposal through a patch so that our discussion is more concreteJimmyPrakashs analysis resonates with Stacks comments aboutIs it possible for you to collect master and region server logs so that we can verify the analysisThanks +Includes slightly refactored implementation along with improved unit test coverageWill not close this issue until JMS TCK confirmed to pass +Committed patch v to after making the logging fix described for the followup patch +This is really highly dangerous +However the ability to detect the type of a process archive and delegate to a specific parser chain should be pushed up to the jBPM core +I dont know if it worth a wizard +Verified in Developer StudioVersion Build id Build date +Thanks Rob Ill dig into this mystery more tomorrow +Found one potential test issue that could be causing this +I think I discussed this with Mike already +We are using flashbuilder so we are stuck with SDK for the time being. +Isnt it a case of transaction timing out? I guess the large result set is taking minutes or more to load +I checked everything line by lineThe exact same config lines work as opposed to not work before + are mistaken +Hi JD is fine as default but it should also be configurable because the best parameter depends on the size of a clusterAlerting when the log is rolled is good +Running run Failures Errors Skipped Time elapsed secRunning run Failures Errors Skipped Time elapsed resolved issues for already released versions +I suppose I can use an existing library to serialize some data for testing but it may be handy to have another directory under share for that stuff +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Applied +Attaching a patch for which and utilizes the implementation +slip to contains +It adds the concept of adopted test suite vrs underReview test suite for the CTS +Tied priorities sort alphabetically by name. +pMaybe browse page should display all version by repo and offer a delete icon for each +The hang was fairly consistently reproducible on aix so Ill try out the sync on copy change instead a few times and let you know how it seems that its not just the copy operation which needs the synchronization +if it can be improved it should be +Id like to minimize the number of example applications we maintain and it seems like any advanced AJAX shopping cart could just as easily be part of the showcase app +Thank you +Claus what do you mean with should not be buildable with maven I can build with Maven without any warnings or errorsIf its because the old Maven repo style of its dependencies I think its not a problem +Pulled all of the string literals for temporary to references to a static final string into +I will split it into smaller patches for different projects + +We probably need a little extra work here on checking if the URI uses authority +Updated version of Aruns prototype patch compiles cleanly but not tested beyond that +Patch to use the threadlocal approach +Done please back change and made feature is my solution +can we increase the slot count per tracker of the test cluster so as not to hit the clumping? Or maybe just comment out those particular tests +Yes its fixed in +Handler counts helps with thisMoving this issue out of now +Perhaps we could get around this problem by also performing the invalidation in a so that the exclusive lock is released once the plan has been marked as invalid +I asked Werner about this patch and it is ok +Thanks for reviewing it Sid +The opcode conversion code then only remains in xx release +Hi NabeelThe bridgeEndpoint option just affect the http request URI I dont think it can made a huge difference on the performance +Ill commit when finished with this task + Polished in examples to require JDK and Maven +Pull it back in if you think different. +Fix in place +b Schema names can be any URI not any so enclosed is a fix in for that issue +fixed +The message flow priority will be passed through ESB invocations into the invoked services as part of the message context but will not be present within the message during execution of the service +I agree this new behaviour is how it should be! Issue can be again for your fast reply! +patch applied to trunk. +Please note that I also changed the plugin coUpdated the patch based on Toms feedback +Reopen the original bug if the problem is still in Please supply a small self contained document as a testcase +That would be better +Thanks think all these classes need to be as the cache may be accessed from multiple worker threads concurrentlyOleg +The only way to get the quickstarts deployed is to deploy them individually which needs to be done every time the server is started. +Integrated to Thanks for attention Lars +Thanks Trevor! +Matt dont forget to also move the three JPA entities in +Modified the file to add the required artifact. +We will focus on using to calculate rawdatasize faster for +Sure the fields are documented +Thanks Thejas! +For example setting it to be the min of the heartbeat interval +Documentation and unit tests need to be finished up +Well I disagree +I believe the in GORM spec you can implement GORM events in several ways but recommends to implement them as methods so thats what I am handling here +Just to add specifics since a task processor mightve already completed previous tasks when it runs out of space it shouldnt be allocated new jobs but it should stay available for serving map output to reduce jobs that need the completed output +Consider what to do if two routes have defined the same ID currently a WARN is logged +Removal can be handled in a future version +Here is a patch that does a narrowing of numeric types in this case +getting errorTarget site does not exist in the project manifoldcfremoving it from ant build +All I need to do is to bubble it up +The patch is more suitable for and lower versions +David Id like to discuss with you how the subsystem is supposed to behave on these various DM operations +Fall back to first mime type provided if we cant match what is in the Accept header +Please revert this patch +We need for UI to work +by cschneider Adding checks for null and not found cases for the user propertiesM karafbranchesM karafbranchesjaasmodulessrcmainjavaorgapachekarafjaasmodulespropertiesA karafbranchesjaasmodulessrctestjavaorgapachekarafjaasmodulespropertiesA karafbranchesjaasmodulessrctestjavaorgapachekarafjaasmodulesproperties +New suggestionFor JBT list projects for individual installFor JBDS list bundles of features for grouped installNote that we could also have BOTH contributions in the same site so that the JBDS Central site could include BOTH the products and the projects in the same UI +Patch committed with revision number r +Sounds reasonable +Im testing them out now +The most recently attached lgtm +export LCALLbq +IMHO this can be resolved not a very important feature to get +Bulk closing all of these resolved issues. +Not a blocker. +ri works correctly now but pager does not work right because its not able to determine page size. +Patch committed to trunk +Some use GNU extensions and may require special flags andor code changes to work on ANSI compilers +Woohoo thanks Costin +Andy Thank you very much +So in your example if you go to project bar and build from there your patch wont work had a similar problem in our project setup +Completed At revision +No feedback +and would have been helpful here +Tools +Added description of disableDefault attribute for ajaxValidator +Though Im not sure if it should be committed since you get the revision in the json response +I cant imagine how a sysinfo change could cause such problems +the value may be set in java code using set variable and etc +HiWould you please try my patchThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +We still need to confirm this so I will close the issue +No love with the uses tweak +Also to bring in the project as part of the build +Sending sitetrunksrcdocumentationskinslucenexslthtmlSending trunkSending trunksitesrcdocumentationskinslucenexslthtmlTransmitting file dataCommitted revision . +The patch looks good to me Knut Anders +bq +I think the idea is good right? The directory structure doesnt seem to reflect one idea about organizing these files its just where they landed +Verified with JBT. +need to set FUSELIB in the batch file +However I dont appear to be able to change the status of this bug so someone will need to mark it fixed for and +Where do you want the mails to go to? dev +Could be that I just havent had coffee for a few days now though +Committed at subversion revision Touches the following fileM toolsrelease +Heres an update that fixes the reference to the no longer used textSpell field type and uses textgeneral on the text field instead +and work fine and work fine and work fine and work finePlease test againNote that I dont think we have it set up this way on the published website as only changes +Thanks +Thanks and sorry for the confusion. +The only thing missing is that mvn clean ought to clean up any test database files left around +why isnt it responsible for calling rewrite on any of a type it doesnt understanding? bq +There is still more that could be done but all the easy ones are done. +gtr invalid backslash escape at end of stringecho i tr sed s +Hi DmitriyThank you for your clarification +I implemented some modifications to FLE to make sure that it shuts down correctly +Making the changes now +Yep will do +They can change it back by copying toolsresources to serverdefaultdeploy +Just committed it as dfff +Thanks! +Cause there are three class defined along the file +done thanks +Given that the patch doesnt have tests I was planning on running a secure cluster to ascertain the behavior but havent been able to get to itWill you be able to validate the behavior and report the same? Otherwise I ll see if I can do the same in the next couple of days +The fix is trivial +we also observed that some rules fire that didnt fire before and shouldnt fire we are also trying to reproduce it in a single +Block deletion was not observed at all +Removing. +Ive attached this file to help reproduce the problem in a production application that has this error +issue +We should probably wait for the similar discussion in to be resolved +My time was prioritized to work on something else +I introduced a reopen method to since the comparison method closes the reader after use and i needed to use it several times +Fixed with and the massive of issues trying to start continuous ingest +I just ran into this as well +Thanks Dag and Rick for the comments +you are right there we can configure releaseperform to skip the running a long test before tag can be very long i dont know what is impact when other people check into the releasebranchtrunk while the test is going since this change the behavior of plugin it should be added as an option +I have downloaded the sources from CVS and compiled themThe bug seems to be fixed because the example code works properlyIm Sorry it took long time to confirm this +We should probably log them the question is at what level +Generally speaking the arrow on the icon shows that the current editors layout will be flipped +Fixed in and current CVSRegardsBrian this bug down +Will do +While working with the XSD validator on branch I dont get this error +Considering versioning in maven is still being worked out best to wait until Maven to support nomenclature +However allows specification of password expiry with millisecond granularity so one cannot completely rule out the possibility that some applications use password expiry +Polish translation for clicks Tomasz! Ill apply it over the weekendKind regards + the as r +I have a POM to make an RPM that I can added to sandbox +Druba wrote Archives current implementation does not follow symbolic links +yep without specifying PHPINCLUDEDIR CMAKEINSTALLPREFIX and other the extension will build with system php and python +the URL was the issueBTW The WAR generation wizard allows you to specify the role but not define in the teiid properties file +However CXF client and +BilginI have adjusted the patch slightly and am running a unit test to see if there is any issuesIn fact we should add to the krati jclouds documentation that the consumer is configured with as max messages per pollAnd also there is a wiki page which lists the batch consumers I do not think its fully updated as it may lack these new components as wellThat said we should possible ask Ioannis why he configured those consumers to use as max +As you can see from the log the correct objects are created but the checkboxes tag is unable to compare the selected would expect the checkboxes tag to use the converter to generate a string representation of the object and use it for comparison +Now that we are back on the older platform the sites are again stable +Ok I will but tomorrow. +I will leave the other branch updates to the respective owners. +is a more general case of this and implements handling this for ears + improve content visibility +Patch needs signing + +META +bulk close of all resolved issues. +they both seem to assume a clean shutdown +The simple demo coverted to use an XHTML doctype +Cool stuff Yonik! I noticed in the groupBy method that you were instantiating a Lucene filter from the filer docset but you are not using it as argument to the search method so if some specifies a fq it would not work +Can you please confirmYes I did remove from the pom +Repositories need to be all gathered prior to resolution +What really got me thinking that only an HTTPS connector was created is the messageServer running +Ive updated the unit tests that use to skip the test for Sybase to BranchSPCP Branch Branch and trunk +Thanks +Part of a bulk update all resolutions changed to done please review history to original resolution type +thanks Ivan! Committed revision . +Thanks Chris that makes sense +I would not close the issue at this time. +This has been tested and will be documented in the migration document. +Thats a requirement +But we will be deploying this soon and Ill report back if it doesnt solve the problem as anticipatedIll commit this later today +This looks like a reasonable workaround for right now +for this round +I am reworking it a bit so its possible to reuse it for the component asprojectcamelsvn commit message Ant path matcher now also possible with +to force of connections etc +Cool +patch to Ben! +The new changedFiles API can be used instead of computing and storing checksum in this plugin +The latest version I tried was +Released in. +the event admin? If possible I would like to have those fixed to some version and not always just use the newest one +I was a little bit confused about all the names +But I cantreproduce it now +Turns out it is quite a tricky issue there is no standard complex type anywhere in the standard libraries +For anyone whos interested theres two rough areas to thisA my use of csharpwcf +Stepan spelling corrections verified at r +fixed in r + for the change +In we warn to logj that the application should use if the values of the attributes for tmpfiles tmplibjars and tmparchives in the submitted and static are differentThoughts + enables the tests that were commented out +As recommended we will resolve this as Not A Bug for now +I was just thinking do I really need to test anything more than the NN generates the cancel edit for an expire? I would expect theres plenty of tests that verify edit replay for NN and SBN +Thanks Robert! I committed this to trunk and. +commit the fix by Florian at in ran successfully on the patch in addition to pushing the sources after build run through. +Please ignore the second patch for now I will it later +Sure do you want me to try with or the latest? is the version running in where I ran into the problem + +This checkbox is for Inherit parent applications shared library imports +for the patch + +I have fixed this in Rev +Ive attached the patch so that If there are any review comments I could incorporate them +Shall we close this issue or keep it around for a while longer +Were more about hadoop these times +Would you mind reviewing the top patch against trunk +Im also a fan of Andreas idea this is a real benefit if the people are able to run their own distributions with mvnrun so a big +I understand and will await impatiently to come out +I have upgraded my application to and found that it made no difference for me +I had to make another fix for to get working from Velocity template +It seems that patch for committed in r is incorrectIt makes launcher to check for help and X options print short usage message and exit without creating VM +then why not use ab for display and internal names etc in type defs and thrift +This patch does two thingsFirst it adds more correct entry count getters for +Review please +Issues that are really affecting the project include requiring the XERCESCROOT environment simply to detect what the top levelsource directory is requiring the use of the runConfigure If youre willing to look into using automake I am willing to assist in any wayI can especially if we can eliminate the need for the above two steps on Unixsystems +Issue may already be resolved may be invalid or may never be fixed +I dont really agree with making the configurable +I would only include in the latter oneThere is a script in Subversion for what is used on the other servers it might be good to review that +Cheers +With the current connection property setting it is either possible to throw an exception deep in the plugin to show that the path was invalid or to just log an error message in the might be possible to put the event source on disabled on error to give the user at least some indication +Logj does not contain author tags +If you are OK having it in the repository for a day or so then we can do that +for this patch as is +Reopening for PavelPlease take a look at how JBM deals with the test think you can close that now +If with relevant build drl file is not still open please check for exceptions and if possible attach the example project +However the setup was meant to setup automatically with minimal config +There are a number of things the branch doesnt do but would be really useful to haveFor example I usually have a help version option +Should we fix this and open different issue related to running ui tests with tycho? currently I see similar tests problems in base and hibernate as well. +On the server side you can do the same with the TCPIP Monitor in Eclipse or a similar tool in other in the future please use code formatting when pasting code in comments +Who is the responsible to close the? I or some administratorIm sorry by the wasted of time +Im fine with moving COR and COV +Closing all resolved tickets from or older +but hibernate at least doesnt fail on it +It looks weird because it contains only deleted content did you miss some svn add before your svn diff +txt and similar +Patch committed as svn resolving. +So the resolution of is immaterial hereThe change I made which instead of using the last part of the URL path as a file name tries to use the entire URL and encode it so that the whole thing is interpreted as a file name +This fix was necessary to successfully handle the otherwise it failed while parsing the EJB class all issues which got resolved in M +I think expressing these values in KB would keep things consistentbq +Then we can design the type of forkjoin structure Oozie should should keep in mind that the validation is to verify well formed even if they could be done in a differentbettercleaner way +Token operations should ideally not lock the namespace at all but the read lock is at least required to prevent the the NN from transitioning states during an operationChanged writeLock to readLock for token operations +That adds some difficulty +Fixed the fix stability requires getting Java to initialise internal fields. +For safer side Ive shaded guava also +Fix +I guess you use refresh on the browser and that sends the previous command again and gives the feeling that the cli stops workingUse Auto Refresh and INFO to check that really looks OK +Caching the result of as for the bug report! +uploading a new patch for review + +Can you provide an example of this situation? A minimal processdefinition and the accompanying task form that produces the problem +thanks Suresh +I cannot reproduce this is as well which uses an older version of Redback +drb is are no for expect NKF requires a little more care +Made changes to initialize BP +I will create a reproduction of the issue +The jobhistorys setup is a little confusing +Alternative patch that actually changes the default +closed +Uploaded patch for the fixes +leave the issue open until the next release +See also in for a test where is currently ignored because of this issue +That is the intented usecase for customizations +Looks good +Thats not to say you cant use that but we dont support that scenario by testing it ourselves +It seems much better than the current codeI wonder if when there is an exception from which you rethrown in a if we should enrich the exception with a message about the error +gantlib appears not to work +patch that fixes could you open a new issue on tomahawk please +Hello +Expect this build to fail until the dependent jira is done +groovy files being copied into the bin directory +True +Use this one several typos in the last one +Regarding Please justify why no new tests are needed for this patch +In the first example it never reaches get then +Thanks for reading the release note DagI have corrected the spelling error and uploaded a new version +The version is out and fixes this +There is no case I can think of where a +delay during some testing of random reads against hbase we were able to determine that it was being caused by the Nagle ACK delay +Further we needed support from Hadoop to do this +I applied to and trunk +Patch looks good to me +This is not a new feature but is a bug fix +I have fixed the spring context to declare explicit services in such scenario which is ideal +As I already remarked that I would need logs repro steps setup info and more context to get to the bottom of this as I dont see this issueAlso please let me know if you followed the steps I put in for SSVM to be running on the latest code? I dont see this issue on my build and so closing this +Indeed this history page on is currently blank and moreover it must remain impossible to remove the snapshot of a without removing the full snapshot tree +Also need to undo the temp fix to add the Yoko specs that was added to by G + +So heres the current iteration lucene analysis benchmark core demo facet grouping highlighter join queries queryparser memory misc sandbox spatial suggest tools +It would help a lot +is an updated patch whichresolves the merge with the patchI also ran and derbyall and Ive got pass rateIf there are no further comments Ill commit this patch later this week +And in GA there is still this issue for some reasons +That means well have different for Jenkins and local +This of course depends on and Ill put up another one once that is committed if there are issuesIt uses the enhancements to made in to fetch stdout and stderr and tags and sends them out to the s respective streams +And also please add these new confs inYeah please set to a very high number like +This maybe related to +Then in eclipse you can only build the srctestjava folder that way you would be able to run all unit testsYou will be required to have a convention for tests in this location such as ending in IT else maven will execute them in test phase +HTML mail from yahoo triggered SA. +Same patch again this time unified against src +Could you provide the stack trace and the session parameters +Overall pretty easy to do but in some cases I felt I was copying a lot of code with the primary difference being the number of I needed to pass through +Sure I have the heap dump still +Please if this still occurs. +OK given that the expiration interval is minute meaning the st run of the eviction thread wouldnt be until a minute after the cache is started I doubt this is a race condition +How many hadoop jobs are you running concurrently per cassandra nodeBrandon and I never see these failures so I wonder if youre just hitting it a lot harder +use spaces to replace the Tab in the xmlYou could review and try commit Jeff I might miss some background info here since there are commands like osgiinstall and osgiuninstall why do we need another commands with the similar functions +repeat steps +Its too late to add to and this will be going in for a lot JoeReally appreciate the time you tookWould wait on this +runtimes plugin should pick the list of servers configured during installation and create runtimeserver for each server +Vaadin does support an immediate upload feature that allows you to drag and drop a single file onto an upload button +Back to Brian for verification etc +Came to know it just now. +in should be all set now +As far as I understand the HTTP header connection just indicates the server what to do but its not forcing it to do it either side may close the connection and that should not break any of both sides +Looks good +Will this at least preliminary support make it into the spring data jpa M +In addition to the change made in in svn rev the attached patch was also necessary for to successfully build the trunk +These are probably all obsolete +Good stuff +No per resolver configuration and no +I have a patch +We can adjust the number of reps or messages to shorten the running time +There is at least one known issue that looks similar that I will link to the issue +The important things are API and functionality +Attached patch is based on the branch +I think this patch is OK +Please describe what you meanI just checked on the server and the directory distportalspluto is writable by a group that you belong to so no problem regarding file permissions +Thanks Hairong! +Looks like you cannot avoid that since if is not specified then automatically become edits directories as well as name dirs + verify in ER build +for the proposed dont think we currently implement the IMPORTUUIDREPLACEEXISTING behaviour for jcrroot as specified in section +Then I can finally check whether the compatibility problem between emf used by geotools and emf provided by eclipse has been resolved +That works for both firefox and IE +Never +Ok so the scroll bar from the browser is no more present but there is one in the webpage +It includes changes that we made in +passed on a Linux machinetest still failed +is a cumulative patch for the issue +Upstream issue created in JBT +I dont necessarily disagree with that if thats the case +dims + +Thats what I was trying to say +Moved to CR for planning purposes +Next time please provide the files as they would be in the repo with subfolders checksums. +Committed revision +Remy sounds like Rich Sharples is the guy to ask that question +in addition the site plugin really shouldnt execute this it should rely on the wagon to set them correctly +I didnt have any problems running these clients with +Ugly? Dont get me started +Fixed in subversion repository as of r +Removed the if statementSinceString str nnaThe if statement will evaluate to false and we will miss a line +xavier is currently working on these bits in the to add newandnoteworthy to set release note notes are not required for feature requests on new components +fixed in CVS for +But over the long haul it is going to be cheaper in human labor to fix this in Hadoop than to throw more admins at it +ok +This task can be closed +Ill commit soon +but it could not resolve the issue in persistence manager in hedwig +Exploration of Apache Whirr based Hadoop deployment on clouds in the first phase will provide us with the required knowledge about the complexity of the task and problem associated with moving from local clusters to cloud platforms +Pull request submitted +Fixed +Exclude is deprecated. +Initial implementation rev +A simple test case test it with changing +Im moving them to so they can be checked that they no longer cause a problem there and fixed if they do +No its a copypaste problemMy original policy is with only one spIm sorry about this! Mr +Both tests runs fine alone +buildivylibHadoopcommongriddevbin +resolving for real this time hopefully +Im still looking for the root cause +They worked for me on Windows too +This issue is dedicated to the integration of markdown into the review comments. +I would expect the behaviour to return +committed new client build to are succuessfully validated in case of user without a domainVerified in Developer StudioVersion Build id Build date +Note that I creaded at incubatorsince there is no need to remove the dash. +Ill update the patch +One +will be much simpler as it supports reconfiguring the client +Wouldnt it be simpler to always release partition ownership according to topicRegistry +Oh and if you had your heart set on committers rather than committer let me know and Ill fix it +We could fix this by including the connector module as a parent of the shared lib module but I think it would be more appropriate for you to add your jars to the geronimo repository and construct a module that includes your jars and whatever parts of the geronimo server it needs as dependencies and to use this instead of the shared lib +Patch removing classpath from for daemons +Mass closing all issues that have been in resolved state for more than one month without further comments +If we have to keep the hard coding of the bugs and attachment ids Comments in the program said we changed to html because of the changing xml format but I a am surprised that the html is not even more volatile and fragile +I ran the test with the patch anyways and it passed on my local machine +an initial version would be really interesting for d move this issue to for several reasons +Will verify with next release +Can I get some feedback on the patch +comments +Note to include the in release docs +Hm the formatting of the code got lost somewhat but I guess you can figure out how it was meant to be +We still need something for reference +This fix is now in the latest +Just taking Thorstens lazy consensus and closing this issue I have never seen the problem described on any browser and just to be sure tested it on versions of IE IE SP and IE SP without problems. +Ill double check once it makes it to git +cool +I would argue a zipf law distribution probably reflects reality + +This is a small change +Why not useI dont think we should make hard guarantees in a joinouter join about the order + to trunk +Documented in resolved issues. +I reverted this +Hi allWasnt able to respond on Friday but Ive just wrapped up the last of the changes +Fixed in Trunk +Hi BruceAttached is the testcase I was using +Yes that works just fine thanks for the rapid turnaround +Ill test thing and close when I have some free time +Previous patch was in reverse direction +So sorry but Im going to mark this wont fix +slip John let me know if this is not true +from trunk +I fixed it to +I think a networkConnector with a composite url would possibly have the same effect as using the failover transport +This should not block especially given. +Here it is +One of our developers tested it and reported that the same page on which the button is being pressed is displayed even though the action flow should have taken it to a different page +The issue is fixed with +Should be finished +This patch moves the directories into +Thanks David for the hintI finally got a shoot to that and use a file under to specify the codec implementation +Also attached patch for branch +I think the patch is missing the interface so it does not new version of the patch +This was found and the fix was tested elsewhere with HadoopKFS +Could you please hint? Thanks +In removes when Disable Groovy Compiler Gererating Class Files is unchecked +Thanks for the review stack +Michael this is fixed right? Might want to close the issue +Hi HiroI have tested our app with snapshot load and didnt see this errorThanks +can you try to usefalse like described in for? Please let me know if it works for youIve inspected the corresponding sources and unfortunately ODE does not seem to implement this workaroundRegrades +Please somebody look at this issue +Ah thank you for explaining this to me. +Should I keep raising separate for each example or should I group the new together in a single JIRA and patch? Ill attach the patched version fo this one in a minute anyway +to +Patch committed to branch and trunk with little changes +this is somewhat taken care of via the migration. +I liked the what I built as it just has direct access to the Field Cache and there are no other data structures and that if once you load the data in the FC you can do any other search on that field and not have to rebuild anything you can just the dataBut I think all are improvements on whats there but as I am prejudiced and I really like they one I wrote and I think it will stack up faster then the if you do load tests on itJust my Matt +The integration at API side is not a good idea as it introduce a strong dependency between two components that should be independent +bulk defer of unresolved bugs to +bq +No one successfully using LCS has it set to mb +MikhailCan you attach patch generated with +Committed at subversion revision +Closed all tickets +Am I missing something +Id add a countour example as it shows how to handle parametersAnother note the WPS needs to be installed but does not need to be enabled for rendering transformations to idea about adding the Contour example go for it! As for install VS enabled what does the user do differently in these two cases? Is there doc on the difference +Trivial thanks Manish! +The TX implementation has been changed as per Toms comment and the has been hooked wired up for ejb and web invocations. +Whether it lives in the unit test with more or less I dont have a big opinion on +Incorporated from Konstantins list of review comments +Any chance this is browser specific? Im not seeing it in webkit or mozzilla on OSX +We can do that in which is the logical placeSounds like a good idea +Shutting down the NIO provider now shuts down all connections associated with it. +Which trunk did you use? As I tried this patch on ubuntu and ATS and got such error +FYI the corner isnt part of the header it is drawn by the +Any reason why you use a SNAPSHOT instead of a released version +Could you post an archive of the data directory +INFO Failed fetch notification for task taskm INFO Failed fetch notification for task taskm INFO Failed fetch notification for task taskm INFO Too many for output of task taskm +We can just block metadata updates for handle and that should solve the problem +native +Have a look and let me know if you see something missing +Thanks MichaelI committed your patch from Close for all issues before incubation +bzip is splittable but very slow and gzip isnt splittableTo do this we need to look at using for moving data between MR jobs +Im not sure for what to do here because even if the field directions is weel present in entity definition I cant see any use of it elsewhereWhat do you think guys +Thanks +bulk close of resolved issues in. +Patch Added code in to handle Parameterized types +defer all X issues to + patch looks good to me +Thanks Clement works fine. +Yes there is some other problem +I just created +If anyone want to have a look and comment on the direction that would be good +We now have this structure. +Use to track this. +Add scatter chart capability to HICC +So I should have only svn diff srcdocs right +So without further objection Im going to commit this so that I can finish the make spanscorer the default issue +Hi thanks for the report and quickstart +We are still awaiting the outstanding information for the Release Notes on this one +If we add new vector to matrix the metadata of dimensions should be increased +The real issue is tracked at. +If is called in thread it is executed in the that is executed in the UI you applied the patch +Anyway what are you using remote searchable for what is your usecase? Did you think about solving you problem with some other technology +BTW is there a patch for the dashboard? Thanks +Can anybody apply the patch if there is no problem? Thanks +Ill check with the other pmc members and see if anyone can fix this +Fixed on trunk +the ticket if there are any issues with this. +probably need an entry in to set to be the default runtime type chosen and a startup extension for setting up the runtime found in EAP +It should have been committed long agoI am happy to resurrect it +The issue of ERROR XT An internal error was identified by module +A job that asks for more than this number will be failed at submission itself +There has been issues with getting the docs onto that public server in the past this will be fixed for as that is just around the should all be back and working correctly now +Thanks for your patience on this +That said I feel we can take one of the two approaches +Hmmm +Youll need to install mongrel rake and the rspec gem to build the deb +column comparator! key it looks like this patch changes the defaultvalidationclass instead of the keyvalidationclass in the cql case +That would have shown up even if there was just service in the wsdlthanks in current SVN +The mode was mixed and there was users already +Bulk close for release +I guess we need to add order by on partition column for this to work consistently on all OS +Also tables will honor the charset option specified as well. +passedCommitted revision . +I like the idea +Pretty sure I tested with and +Would you please elaborate? thx +Moving to since we have a workaround +This is real patch +You can force Install phase failure by renaming the varlibpuppet directory to something else on one of the agent hosts where you want to simulate failure +I am a bit apprehensive at this point to make this solution public +Committing + +Sebb can you delete the files yourselfHadrian doesnt seem to take action on this issue +Basically it address the case that a region server may not have any online regions +That would be much betterThere should also be a link to mailing lists on the web pageThe tab should also be listed earlier before more stuff such as IssuesWhere Discussion is also present in the page body under Resources it should also be renamed and promoted to the top of the listWhile on the subject the License link should probably also be under ResourcesIts not really appropriate to list it under ASF links because it needs to be clear it applies to each projectProbably best to have it next to Download or System RequirementsBTW some components use Issue Tracker and others use Bug Tracker +you dont say on which page this link appears +Committed at subversion revision +In the above example It fails if you create a mapping named ws in the presentation context and run the from there +For +removes final qualifier from parentName supplies a setter method +I dont think it is the task of the web console to write something back to somewhereRather probably the tool should listen for configuration changes and write them back if appropriateReason The module loads the config file and ingests it into the Configuration Admin service +Please test if it works for you and close this issue +I will revert tomorrow if alternative suggested above continues unaddressed + revision +Per above discussion should be addressed by recent fixes to the builder +Heres the patch that does the job +I tried to apply the patch on Windows +Help in resolution is appreciated +And yes its reproduceable +Attaching a patch which is a combination of and at +Shouldnt junit be included in the tests jar then instead of having it on the cluster itself? Every other program brings the libs it needs with it +tests were added because this was a compile failure fix only +Creating m instances in a loop and profiling shows nothing untoward but something a user might not expect all m instances are locked in memory! +What is the exact contribution URL for the zip file when the problem occurs +I just committed this +Committed revision Thanks Jason +Check the manifest creation in the latest jar plugin +The checksum is verified on the client and failures there are reported back to DFSAre these failures reported to the Namenode or the Datanode? From what I know its the Namenode which I believe doesnt do anything with the reports other than logging the failure +This looks like a parsing error +surl generated urls behave slightly different +After looking through source code I couldnt see any difference in how the infostream file was handled so as an experiment I enabled infostream in my config and restarted one of the cores +The testcase is actuall based on the NIO client this kind of thing is an excellent use case for a connection manager +Closing released issues. +It defaults to +Sorry assigned to myself in error +I still dont understand whats the length mismatch here but one way we can fix this is by not using the when the block is being written to +jar files. +Quicktstart application with custom with for locating resources +its ugly hang when it happens +Cant you use scripting instead +A patch for review +fixed with rev +test and put it the srctestve created a configure the correctly when theres nothing for the topic in the Kaha contains some messages for that topicDestination and Subscription info will be recovered and will be created in add method in that the value of prefetchSize of the is zero +Sorry didnt notice your environment +We should use +id gladly collaborate on this +Change the visibility of some methods of drda protocol implementation to public The instances of som of these classes might be created somewhere else however the instances created in derby network server are still hidden for other packages +Thanks Oliver! The patch is fine +patch that incorporates uwes additions to the testcases and enforces the Version of the source set in the copy if it is an instance ofI will commit this together with +Ah hem +How were you exporting this to a war? Were you using the Project Archives view? Or were you using the export +Its the httpclient upgrade from to that is a hard as the API have changed a lot +On second thoughts we need to be a little careful here +With the change each test will be run for RELATIVE ABSOLUTE and QUALIFIED paths +Please find enclosed a zip file of the work Ive done to get going on Ingres tarball contains mainly ingres components which would need to be added to the server configuration deploy directory on a installation +They are belowFor PDF the standard reads Font metricsFor every font embedded in a conforming file the glyph width information stored in the Widths entry of thefont dictionary and in the embedded font program shall be consistentConsistent isnt the worlds best word for a standard so in PDF we did a bit better Font metrics For every font embedded in a conforming file and used for rendering the glyph width information in the font dictionary and in the embedded font program shall be consistent +It is possible to introduce the dependency of course but it is equally possible to augment the documentation of these classes in to indicate that they are intended primarily for use in conjunction with or preferably by and that more replacement classes are to be found in Commons Lang +I agree about the time but the day is always the current date rather than the day that was clicked on +to show if those dependences actually upversions correctly +the vm will be provisioned as described in +Just getting the ability to do this from JMX would be great +No response assuming ok. +Are there any ideas and solutions for this really big problem +Making an abstract class will make this sharing more clunky +for leaving the original browser menu. +I could not reproduce it and it is not likely that our code may be responsible for it +removes from the exclude list +If you have inheritance involved then creating consistent equals and hashcode methods is not a simple problem hence the reason for making the class can be fixed separately. +I wonder to know that is it technically possible to create customer tag that overcome this issue? Thx a lot +If you finish it before we release beta please correct the version +This issue resolves issue in a more pleasant way +Resolved in +Fixed +Many thanks t Josef. +Here is a simple JSON tokenizer +Resolving +Marked normal urgency as that seemed to be the intent from the triage comments. +Since well be basing Seam on the Web Beans RI this feature request is obsolete. +The changes are actually for another patch I am working on where we can have all plugins in a jar and they can be deployed and called from the jar instead +Here is my current patch +Ive sent a mail to the plexus compiler user list in the hopes that there are in fact additional issues fixed worthy of a new release. +Sorry I have no clue what you are talking about Which mojo how does your pom look like etceditDo you mean that this works well but you would additionally like to force this on the commandline +There are no trademark conflicts for per my earlier analysisThere are other tasks that need to be done such as moving to ASF +a patch for Solr configs which shows the ve added a test demonstrating the bug. +Anyone who is not using the feature would have lost the same amount of dataOr am I the one misinterpreting +do almost follows the same logic should be able to unify that code as well +I also removed as it already existed on +Should we just mark as an incompatible change for referencebq +I will fix this since its just a missing import +New patch looks good +If not please close +The patch looks good please go ahead and commit it +See +Will take a look at this tonight thanks for the patch +I believe that is currently related to this task +I used enums in order to make things efficient so that made to only work in Java or higher +What is going to be selected in visual part if part of next input element is selected in text +Sorry for the late reply! markus +Ok i will contact the autor and see what happen +Also note that the XML Declaration tag is not added in AxisC clients even though the AxisC server accepts messages that have an XML Declaration +committed to and trunk thanks for the reviews! +Sorry about that and thx for the fixes +from the time stamps on the jars theres no way theyd contain the fix + and Xml Template Engines both simply propogate the exception +Sorry about that one +bulk defer to defer to. +The default would be null but and all delegating wrappers should implement that +Can you confirm this behaviour change +Feel free to to by the attached test on EAP. +I have something like this in a plugin that does a few other things as well +applied patch +Looks good to me +looks good +Triaged for +Thanks SaschaYour patch is in trunk at r +Seen back to +This change intends to make that easier +Those benchmarks are completely outdated +can we consider making that compression only featureIm fine with that +Long pending and cannot be reproduced. +The getProperty method now takes an Option which should be a +Sorry about that +Fornow I have concentrated just on implementing the ALTER TABLE syntax as it isthe least amount of workAttached is a patch proposal This patch modifies the parser to support the new syntax ALTER TABLE t RENAME COLUMN c TO newcand ALTER TABLE t RENAME TABLE t TO newtThe patch also contains a handful of tests to demonstrate the most rudimentaryaspects of the new commandI intend to write additional tests later and propose a replacement patch but I wouldlike to hear any feedback on this patch too +Hi FlavioCan you please take a look at the attached patch and let me know if it looks OK +so updation will not happen during deletionThanks for the comments Vinay! I still think our current rename implementation will not lead to this scenario +It never could work with such settings +I believe the preference is to use an existing packaged jar over bundling one for the run +comopensymphonymodulerandom matchesBinary file + +Fixes issues and and replaces the earlier patch that has been withdrawn. +Its definitely better once I get a chance to start replacing the oro regexes some other possible simplifications will probably pop up +Attaching failed test looks like its running two tasks +Can you send me the text that you see there +Advice welcomeback to for success +Ive started using WTP recently and havent had the same generally agree with you and wouldnt mind NOT FIXING this +Later on such system id is used to determine whether its a duplicateBut if your customized entity resolver returns in an input stream or a reader without setting the system id then there is no way for us to tell that b and b includes the same documentThe solution you mentioned works +the base class doesnt run any tests since its just meant to be inherited from +that appears as a separate sentence in the original appears to have been transposed out of the sentence that follows it where it should be the start of + +now with ASF like this has been fixed +Attaching the patch against that I have in mindIf we agree on that part Ill update trunk on the new assumption that the aliases lists cannot have nulls +We dont add those in by default but I can add them in as comment so that itll be easy for the customersHow does that sound + +The library came back though so someone must have linked to it using their +When the file to be open does not exist or is a directory open throws a instead of +Includes changes for also +Committed to TRUNK +Please try now +Are you going to put the PDE plugin at Eclipse? We definitely can get you hooked up at the Mojo project +Fixed in Rev +imo possible to close the issue as not a bug +Release Notes marked Not Required as this affects an internal Engineering release and is resolved in an Engineering release so no customer could have been affected by this issue +Thank you Michael great to see theres some work done on PWR I always loved it due to its simplicity and low overhead. +You may modify your schema without your content and still be able to retrieve fieldsAnd what if I choose to shard across two Solrs with different scemas because I know they share the id title body fields +New version +The patch looks fine to me Im happy to commit to trunk and +Thanks PaulPerhaps introducing some protected variables in the base class likeARRAYSTART ARRAYSEP ARRAYEND etc would allow us to eliminate some code duplication +Comment on what still needs needs to be fixed +KasatkinMe too waiting for this releaseIf you can post us your it would help usThank in advanceRegards Id recommend you before sending your requests to figure out how src really worksIt unpacks original substitutes necessary classes and packs compilled ones intoHave you tried just to run whats the problem +thanks Andy +This is one of those features you didnt know you needed till you see it in action +Heres a patch based onAll CP instances will use a single watcher which keeps the date up to date asynchronouslyIf the watcher get disconnected from ZK it will try reconnect periodically +Missing things client api for secure where should this go? Should we have complete API for the client and have that use secure when security is enabled? doesnt handle failure scenarios +Is it automatic? I also saw a bunch of methods for the primitive array types Im wondering if it would be worthwhile necessary to add first last head and tail for primitive types too +Unfortunately Ive got one of those ones where it isnt betterI guess this is really a feature request for an that puts together all of the files in a directory and splits them evenly +The Rhino and Castle projects and Apache iBatis do include the snk +Also if you change files managed by Eclipse outside of Eclipse workbench you have to refresh workbench to let Eclipse know about those changes +Merged into master branch. +plugins name +This should be documented as a known issue for EAP CP CP +If the array has to be there you need to instantiate it in your constructor +I see has been uploaded already +pushed to cr trunk +No I havent played with the +I am working on this issue nowI have a unit test showing the problem but it will take a couple of hours since my time is limitedThe whole configuration loading needs to be reworked if I got it rightYou do not have a patch arround to solve this issuecheersRainer +In the future the plugin will contain code that is user interface related +bulk close of all resolved issues. +I have moved the to srcmainresources as of revision +If this isnt a blocker then it should be moved to +Good chance this has been fixed with +John why would need to be discovered? All poms in the repository are +is and as such fine for not being proxied +Logged In YES useridI fixed this in someone close this? I dont seem to have close Username lqd +I had several exceptions about illegal format on windowsWill investigate more tomorrow +I see it as a set of projects on disk that are all connected chaining upward to the parent where it has the implicit Super POM parent or is not connected to its parent on disk +here is a clean patch for branch + Luciano + +If its a stat that youre wanting then adding something to the stats module would probably be the best approach +The patch now has only two methods for returning the count of leased compressorsdecompressors +another work around is to get the window handle of a window that your process creates then lookup the PID using using the above getprocessid call returns the correct pid now +Bikas can you pls look at the javac warnings? Thanks +How does the warning help? The above comments say duplication is a possible valid case +to Browse +Sorry didnt realize it was still in progress +To verify UI is started go to ConsoleHost OSGI console and do ss +Its highly unlikely we will be able to query contiguous ranges from the same node +sorry I was wrong filling the issue this is a bug not a feature request +Jian did you get a chance to run all tests with the patch to make sure we arent regressing +PatchI added asserts that trip when illegally has omitTFAP trueand storePayloads also true +you can decide that when you handle this think this is really a cool feature going to boost jBPM +The above code is not going to solve the problem because there is a race +Done. +What I meant is without concrete definition its hard to reproduce this error for each component +Done now +You are of course right on both and This patch should take care of your suggested simplificationThanks for catching this! ran OK with this patch +Log of mvn deploy with essentially the same error +thanks for your fixes +So seems not need to be mentioned explicitly +It indirects theaccess to jpegCount so all the sharea common count +Yup some issues around makeInvisible some around relationships some in relationshipUpdateRemove a few bits of dead code +Yes I know it is not required by spec but do not tell me that it isnt very logical and natural to have +The patch an actual patch this time +Sorry I didnt download the source from trunk so I am just attaching the entire + +Help in resolution is appreciated +changed to minor +Integration tests were updated +Rich checked in this fix in r +this might be a Locale Problem what is the Default Locale you are using +The pom needs to fix tests first +running tests +Even with this error you can start JBDS using studiojbds launcherIve fixed the error in class so it understand both old and new installations +Sure +Anyway thanks for the patch +I will represent AOO and can any day with the above time restrictionrgdsjan I +I had messed up the patch using eclipse + +Please reopen if the patch applied does not work out +Moved to +Did you try to execute the following commandmvninstrument that works but without qualifying the groupid it does not +Wasnt sure which was referred to in the Java sample Ive found a in our directory tree but noCan someone repoduce this or should I close as WORKSFORME +There were few which were unaddressed +Patch update with some minor tweaks of prior changes plus conversion of the and testsTODO +As discussed this approach has been abandoned in favour of JMX +Resolved for a while +Please update to the latest trunk version then overwrite with the file attached rebuild and retest the issue should be fixed nowIve just submitted this attachment to Torsten so hopefully it should be shortly in SVN +Unfortunately I still have a problem with one comma +Im about to merge this into master so it comes available for +Thank you Forrest for the patch and your work on this +Done +Im going to try to write a unit test that exhibits this theoretical bug and then try my fix and see if it fixes it +That will lazily invalidate older cache entries against those tables. +Please verify +Seems like trunk act properly in the shell at least +An own Border implementation +Was not what was agreed on a while back +Please close this one +NPE log on. +See tip +I never liked them and they clutter up the config for no good reason I can see +Bulk close of resolved wont fix cant reproduce etc +Specifically its Sqoop that decides which Avro type to use based on the database column type +Either to remove Internal Error completely or mitigate it to a WARNING status possibly with path to file shown etc. +I cant recreate it neither both on M and M may be caused by different environment +Curious why was this issue marked as Rejected rather than done +Thanks +GreatPlease note that if is applied there would be no need to apply the patch submitted for +How reproducible are these results? One run isnt a significant sample to me +On other suggestions startend key in meta row no objections +Ill look into it next week +Fixed in for +This one would definitely make it into the port. +oops +Early version for jetty +Committed to trunk and. +I like it +OlgaI have committed it on trunk +Heres a question +Isnt GB the upper limit of a signed integer which is what Java uses +thanks! +At least in our website projects here were using heavy concurrent Reader Writer scenarios +Are you running a local instance at the same time? I wonder if I should switch the GC to use ports? Or the random port stuff? Wasnt the random port stuff inAnswering my own question yes +however unless they know exactly what theyre doing they risk corrupting the repository +For my setup the namenode was able to keep up with one file per post and I didnt observe a latency penalty. +Logged In YES useridResubmit with a more recent version if you believethis is still a problem +It doesnt have anything to do with this improvement it would even crash without it because the number of pages is needed for the page flippingPlease do not post to already closed issues! +Fixed in svn revision +You mean this will only be fixed in Web +This will result in less flushing of memCache and make things easier when reading +and if both panels opened +I have no idea what the effort would be to get all these working +not done +It just freezes the for a few seconds thecomes back +trunk +I dont see any reason not to make this change BEFORE the first release comes along +Our experience in testing this with the version of is that if we attempt to stop the with while the storage is out that may not shut them down cleanly due to storage problems leading to further problems down the road +So I believe when the block args passed through all of Rails various splatified proxies they were getting damaged +Do we want to consider supplying open timestamp from the master too +sounds right +It sounds to me like it may be a test error where you a submitting lots of jobs and changing an object in HDFS in between them that is shared by all of the jobs +This might be linked to as this extensively deals with name spaces +The fix for removed support for importing SCHEMA and CATALOG objects and also removed the path value for table entriesThe sample CSV file in the importer now displaysTABLE Table Table DescriptionCOLUMN Column VARCHAR Column Description COLUMN Column VARCHAR Column Description COLUMN Column NUMBER Column Description COLUMN Column DATE Column Description INDEX NORMAL UNIQUE Table Column TABLE Table Table Description COLUMN Column VARCHAR Column Description COLUMN Column VARCHAR Column Description INDEX NORMAL UNIQUE Table ColumnIf you your test using this info if should workMarking this as Out Of with Teiid Designer +bulk defer of issues to +Closed after release. +Very cool contribution +Logged In YES useridSame problem also with HEAD as of +Will run tests and commit +This will not be in but should make it into the next drop. +However this is not a copyright matter but that of intellectual property rights +Updated as followingDouble click on the Monitoring panel entry to open the message in dialog +seems to be the best solution since it would allow us down the line to have some programmatic way of recognizing that there was a problem +Id like to see this task get more traction +Thus I think this issue could be closed as diff from RI +I think we should move forward with this issue and get this feature committed +The default rpc client timeout is MaxValue see +Committed to site to update the Ambari project webpages +conf files to reflect the new location +Please see email for details +Thanks Enis! +Is that ok + +Diggi the patch fixes the issue by wrapping a instead of subclassing it +So localhost means firefox wont store the session cookie and so there is no session id and wave goes around again asking for authentication +Thanks +I aggree lets close this as wont fix. +If this issue is not resolved please let me know otherwise I will close it +We should also include runs with all data items as serialized longs another use case that will be common I would expect +I think its as simple as you are not fetching the System property in a priv block +Additionally would present the blocks instead of the rowsFurthermore in the future if either or has a new format revision it breaks badly +Enis I opened a new Jira to track the failing of this test on Solaris. +trunk patch +Patch looks good to me +Committed v patch to branch with the following minor changes in log all unsent messages maintain outstandingRequests properly on both successful and unsuccessful sendsCould you file jiras one for taking out dead brokers in and another for transient failures due to ZK ephemeral node not deleted in time +Ok on my side I have passed +After a bit of a nightmare this is inThe SVN server timed out on me so I had to break the upgrade commit into pieces +This plugin is not up to date with the patch at and should be updated added test case fixed if above issue exists +Go to the user forum +I thought the real fix should also consider the performance as well since the SGD algorithm is really fast +Should we make the limit configurable? We can but do you really need it to be configurable? It will be yet another knob to turn for the user +Please add your proposal to the wiki +desktop file should have execution permissions to be able show the ve workarounded that by adding hidden panel wich is executes shell script to make +This is happening to us too when we compress files with pbzipAny ETA for a fix? +Seam Config is part of Solder now +If you think that this is ok then Im fine with that. +Attached the batch file that I use to compile and run the test case +Forward and backwards compatibility is needed because during an upgrade via rolling restarts we might have logs of both formats in a single clusterIn we do not have this restrictionMoving out to +Im using your pull request as the basis for cleaning up and syncing the Blob Clob and descriptors +And the Whats New pages are developer created so this should really be a JBIDE issue assigned to me or Tom F not Isaac +Ive managed to fix the offending code +Resteasy will be in the books by then but my tests are passing +If we were to move forward with this as a feature enhancement we would need a specfiic GEP proposal. +Very cool patch thanks +Thanks Chris! +Why do not you just copy the content of temp buffer to the internal buffer? This could potentially eliminate the need to maintain two buffers internally +This may be OK +line still has the incorrect collation determinationI have no examples in mind for backwards compatibilty issues with UCSBASIC just wondering if it had been thought about +Michael Looks good +Ant Editor +It looks like Pinaki has dropped some code via revision r that provides at least some of the infrastructure needed for this JIRA report +The message must be received and processed complete +Committed +However I am working on the issue and can submit another patch soon +This is intentional the is only used for searchRequest that extends beyond the set on the client or if it exceeds the server for a searchNow we may want to throw a dedicated +The point is that we dont want a busy region blocks a whole region serverAs to the old clients right the behavior is a little different +an absolute root such as CTempgrapes that might break with your proposed change but I couldnt get any existing absolute URL of that form to work so I think I concur with your assessment that the change should be +If a service is registered for a static reference the service is ignored +With a Sorceforge project you have to link to them +If you encounter a No exception when using CODI in combination with Trinidad you can use the trinidad support module of CODI +This has already been fixed in CP and will be pulled across into trunk once the CP release has gone can leave this open for now to remind us to check once it has been merged but no work should be necessary +Thanks Ivan +Hi TimIt looks like did not get added can you take care of thatKevin +I believe the is in the correct place +Its available now +Any updates +So it seems like we are splitting hairs at this point +Still working on other bugs so that I keep the debug messages +Randall how does the system manage the exceptions raised by the setters in the source? Perhaps this is an instance where we need to explicitly wrap the instantiation and setting in an exception add a log warn message and leave it null afterwards if failed +Is that something we could do automatically when we create new accounts +PR merged in master and t be verified until is not fixed. +Thanks sami for the suggestion +is an extension to the basic Prevayler that uses JCEs DES encyrption algorithm to encryptdecrypt the snapshot transaction log should probably also be encrypted but I havent done anything with that yet. + +Are there any plans to implement this in the near future +I agree with them all +next because will already use + +Ive checked with and JPDA tests now pass so the problem seems fixed +I have fixed most of the wsdl import issues with wsdljava tool. +Changes in are contained in this patch +Working with Josef to see if the test itself is OK +The interface has been defined and an adapter is available. +Thanks Prashant +Is it possible to work on thisIt would be very useful to have aggregation for PMD reports in projects +Ok maybe add a POST flag so that we can easily flip to proper behaviour after the disaster happens +It was a little annoying when switching regularly between M and HEAD that the files for one were always gone because I did a on the otherStill I can see your point +Putting something like this Client Ack redelivery in something that seems better suited for a higher level +Thanks +They use and run in felixequinox +Of course the use case of one client sends to all is useful I just wish it was not the only one possibleAnyway your context is useful thanks for thatIve tried myself to implement a in your code but for the moment I failed +After patching you can build geoserver with P geoxacml to enable the module +It is better to be assigned asap but it is not mandated +Not going to happen by +this sounds like an excellent plan +DaveYou are right attaching fix to chose right directory to write sstable when compacting +has some typos in comments +Versioning operations like restoremerge will soon be implemented andor adapted to the latest api spec +Bulk close for release +I see you also removed the two driver lines which puzzled me +As a workaround I tried changing my testsuite to avoid dropping the keyspace each time and simply truncate every CF in the keyspace instead but that seems to lead to the same symptoms +You can load the configuration separately by using the Configuration or classes +Marking closed. +Sailaja could you please as per PrachiPlease do not copy the files manually during your testing +is what allows you to then send in values by system prop which is much easier to do for the getting started demoSo I can specify zkHost and all that stuff thats currently on the commandline in? I like that! I havent put numShards on the commandline I did that when I used the collections API to create the collection +Actually does not have this +The SQL state X doesnt necessarily mean that the user issued one of the commands like for instanceij select from t where x sort by zERROR X Syntax error Encountered sort at line column Issue the help command for general information on IJ command syntaxPerhaps we should append or consult the reference manual for the supported SQL syntax or something to cover what the help command is silent about +If you beleive that should be please and contact the program team +Hence the present Spring is Java could now use to improve performance +This issue should be resolved as the code has been merged and is included in release +Please follow for developments. +not sure if this was what was being refered to +Here is a preliminary patch which basically has the content from wiki explaining the multiple categories that affect compatibility my understanding of the current policies for each of them have called out NO POLICY for cases where I couldnt find a policy forWhen you get a chance please verify if my understanding of current policies is accurateAs the next step I can propose policies for items without one +The AS farming service has code to handle this +New version of patch that includes a test +Ok I took a quick look at this and it seems to be caused by bugGlenn would you be alright with me reverting this change? Users can add the true as Chris suggested but this shouldnt be the default as it blocks invocation of the AWT renderer +Can you pull this out and create a separate section on it? Since we talk a lot of and Couriers its necessary to have a separate section on +Sounds like a plan if by remove all content from fs under the tables dir you mean under the tables regions dir +Ill update you in a few hours +That section will still be available for customers in this release +Strange +If we want to change that we should make this an enhancement requestthe fix is change your Im closing it +Does the output actually match Bens description? I think he wants it to automatically reference the appropriate type +A patch is unlikely since users are encouraged to use the public but the rule abovemight be the workaround you it wraps your in a which can be used to initialize the Process Runtime in the working memory itself +The problems with vpe could be related to +We should be able to incorporate the stuff in Lucene now right? Im not spatial expert but this means we could have an update processor that only uses one field right +This patch could be backported if desired but closing now +Those internal nodes should never be printed at all +I looked at the properties for the test and found thischeckpoint influences the plans in some cases becuase rows count getupdated on checkpoint following is necessart to check some plansBased on this I have the following questionDoes really the checkpointinterval influence this test? The indexes are created after the tables are populated so the statistics should be updated orWould it make any difference to the test if we add a sleeping period after the population of the tables? I think this could make a difference if the failures are caused by the optimizer choosing plans based on outdated statistics +Here is the second patch which alignes local memory operands on if they are used in SSE instruction +It works good as well with version. +That makes things alittle tidier and makes accidental use of the value a few orders of magnitudemore unlikely +This session leak should be fixed +file attached to go with java thanks for the info! Ill take a look and see what we can do +Final version of the newmodified plugins heuristic link extractor for files and code snippets +Moving to for produce a better message than an NPE +Also in terms of performance a CLI set statement is smaller then the equivalent insert into +The biggest changes are that it uses a long instead of an int for the counter in and the interval is fixed at ms +who cares if a user wants to override them override at your own risk +Scott Albertine thanks for the test case +Actually I am going to backport this feature to +Attachment fop has been added with description alternative fop not quite +Theres no reason to put it in if is in the testing stage right now +So its que would fill up faster making the master send some open request to different servers while this one catches up or loop until one of the region servers has open slots in it pending open que +deployment is being pulled correct +Nothing more to do. +This was introduced by +Part of a bulk update all resolutions changed to done please review history to original resolution type +The remaining work is to add a test which catches the issue you pointed out with vs +Hi DavidFor such a general question you should post it in mailling listOnly when youre sure theres a bug or a reasonable improvement requirement you can fill jira issueSo please post such question on mailling listThanks is a general question which should be posted through mailing list but not raise a jira ticket +Fix committed to +I spent some time on this patch and have it merged to trunk will run tests and then commit +Theres nothing special that warrants contortions on the client to do simple gssapikerberosThe root cause is the client guesses the principal for a remote NN based on the principal of its default NN +You just have to set min heap high enough for the JVM to force swapping +availableMB is meant to only include memory +This should fix the issue +Except if you look at the test output the failure is an assertion error in the unit test and no NPE appears in the stack trace +No more enhancements to DI model in Seam . +Fair enough fix committed in revision +Thanks for the quick let me know details of the version control system so that i can check out the OliverPlease see if you fix the snapshot so that I can directly get it from the maven trying to build all spring projects when I am running the mvn clean OliverI am able to manage to mvn install the source in my eclipse +FB doesnt officially support Java nor does it officially support Flex Im not able to reproduce in FB or any other +Need fix for +Otherwise how your docs fall into segments can alter your scores which is awful +Im assuming that setting a style is simply not allowed at all now and removed the style attribute but if it is allowed in some limited form then we could put back in something that wiki text will accept. +It does what I suggested +Attaching the final version of the patch +Verified on. +Make sure you are typing your username with the correct case +Maybe this is a Hypersonic problem? Or maybe I dont know what Im doing +Is this still an issue? Can this be closed +Thanks Bobby! +Ive got an implementation of fine grain fsn locking that can be phased in with no compatibility issue +two test failures are unrelated +manish have you had a chance to look at it is another query which army tested before and stopped after +Updated the pull request to include the tooltip for the Release ID group +Fixed and tested on BSC in all resolvedfixed issues of already released versions of Roller. +committed to master +We dont want to split log for all failed servers before we say system table is assigned +It will be taken care of +Farid when you have a moment could you verify that this is still a problem +Mark there is no leak here +If you could test it out and it works ok Ill create a patch for HEAD as PeteChange looks good to me +I can provide more logs later +The attached patch should fix this for reporting this and supplying a patch +Sorry for not looking at it for a long time +And then youre more strongly tying together whether the output and temp both stay or are deletedId rather avoid change unless theres a clear deficiency +Reopen if desired +will try changing some of my dev environment to see if i can force it to show ve verified your fix does not break linux +It would appear as if you have old source code or old compiled classes hanging around cos package no longer contains +Without cache k iterations +The unit test and integration test should be split into separate classes +From commit comment In a has been replaced with a +Shouldnt be prob +ok I will try it tomorow morning and i will give you a am trying to build an application to show you the real problem +Fixed in +I still think that there is merit to having this in +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Otherwise looks like a bug in for me +I am getting this bug relatively frequently with hour time outs +Committed addendum +for the patch +The usehttps property on the addS command was introduced in +Hey Stefaan I am not too sure we are on the same page +On the th there have been quite a few sn commits for so it must be related to this +I assume with all the recent work this has been fixedRam Stack Would you agree with that? If so we can just close this +TODO Update wiki DONE +Does it make sense to add the name of the function as one of the annotations? Then maybe later on we can just register all classes with names UDF UDAF in the classpath instead of explicitly listing out each function in the +Fixed deployment of assemblies. +thread dump from for patch review before the release +Then declare success and be happy that you made the library more useful. +Marek could you look at this and the issue it incorporates and write up a section for the docs? Ping me or Jay if you need more info +As we are all now in consensus I will draft a proposed change to the CDI spec which clarifies the behaviour of in ears +Target release +Your patch is correct +I tried Ant and JDK and JDK GWT and GWT and all permutations of that +Makes me think the project is lively and more willing to contributeSorry for the duplication of issues! +Slight variation of the patch which matches the existing checks forURL and for debugging the problem Dmitri +this to trunk and branches +Simple beanshell test suite for it +I think it its either fixed either never really existed +This is working fine on just tested on my Sonar instance +similiar to the how you would work withHoward Yes would be a good idea +Now that work is well underway with AS all previous community releases are +Since all acknowledge access to the channel is serialized in a then no locking is necessary if this is moved inside the channel +Not sure about the delegate parser though as each file may require a different parser +Closing this as good enough. +The patch solves this issue +When we used the tcp URI the sender was not able to send after we the broker +I did validate that importDirectory will read input files from the given dir and move them appropriately when using the fake shell this only worked after modifying the cached config. +Writables from should be used only when an implementation in hive is not availableAlso the strings should use Text instead of +Quite likely some common bug underlies this problem and +This happens because of the fork mode is not invoked +I assumed the fix for would be enough +nix that first point +forgot to ask +Attaching a draft of what the API might look like +This time with a unit test +The graphs show the average of seven runsthroughput and CPU usage were reported every seconds +So this issue is also not there anymore +It doesnt require a release note so Ive removed the flag +makes sense if it doesnt hurt anything +So pipelineRecovery is not an option +setget of properties used delegation to the internal context so you cannot see that internal variables of the marshaller are not no way to edit a comment +log which would seem more appropriate to find within the NUTCHHOMElogs directory +can you attach a working example +Editorial pass over all release notes prior to publication of. +If we want to be compliant when running with Aegis we cant be preventing duplicate parameter names across methodsSo we need to change something less schema types or in a pinch more unique schema types +In TCP Server impl the value is never set +At that time we had discussions that HDFS side also we need some changes and Stack has already raised the same in HDFS JIRA +The setting is autodiscovered for each file independently +Ill look into them more as well +Please apply this one after applying patch for +If the problem persists on the branch without your local fix can you please file a separate JIRA for it? It may be a bug +The values that are saved to the db when hive loads the data are blank when exported from derby they are blank in the csv and when imported into mysql are defaulted to NULL +Meant not +Doing this would require importing the contents of the repository into the work directory +Thanks for WSDL I wouldve never figured this out otherwise. +If it still hangs try with the above v patch? It adds extra signalling flush is needed just before we go into sync block + +Attached is a patch which adds all xml files in each source module to the list of files checked by and fixes the files that were missing the copyright notice +hm +examplejoinso that maybe spent a lot of time to transfering using hivesql to standardSQLin my opinionin short timeboth maybe +regenerate the patch to include fixes for UDF and Dereference deep copy +Updated infrastructuretrunkprojects as well +If this is still an issue lets open a Jira against a specific version of Tuscany and try to resolve and reduce dependencies and size issues in that specific release +Servers do not know the time a client sends a ping and pings are not sent frequently only at idle intervalsI am finishing refactoring the server side code and I will force the use of a fixed heartbeat strategy for nowComments and suggestions are Im not sure I understand what youre saying +Yes this is not implemented yet +Can you create a new jira for this new patch I mean changes required in current hive code for hive server +string is almost perfect except there is no to the left of the decimal pointAll the other processors do like sowhereas xsltc does like so +In fact IMHO its not important if its a bug or not +I was running the standalone though Ill try deploying the war to tomcat and see if Ill get the for the quick reply Chris +Altered the test cases and supplied a potential patch to make them pass +no objections I will commit the patch soon +the generated script has CLASSPATHPREFIX to accomodate this request +Fixed in +This is sufficiently resolved in removing from the milestone +Know I understand the problemThanks for the helpRegards! this problem still in progressDo you have any idea when it will be releasedThanksVincent Beretti +Hi HadrianI used the as part of an engagement to prove that the solution works +Just to confirm Are these settings liveGBKGBKWe can set up a test case and see what happens +By the look of Daves comment he seemed to have solved how to do it +All the files are issues that are still marked as resolved but were part of the release. +functionTeststestsjdbcapi needs to modify javatestingorgapachederbyTestingfunctionTeststestsjdbcapi +extra lines are there in front of some of the methods remove them to use foreach kind of syntax where ever makes code look simpler +It does however at least at first sight appear to break the Principle of Least guess I am seeing a conflict between lexical scoping and runtime binding +I greped through the report and I didnt really see anything that was related +I dont know if it is related to this exact problem but I have noticed that if I accidentally have subprojects with the same name maven claims there is a cycle +However it doesnt appear that the Factories are looking to see if something is +tested manually +Committed as r as r and as r +NOTE my setup has cluster wide and zone wide primary storages +missed Scott for a long time now +But if we then do a build and target dir appears we are now a dev env +Part of a bulk update all resolutions changed to done please review history to original resolution type +It used to see it on every other run or so so it appears to be fixed +does that mean youre on the patch or the idea +you mean the first release not necessarily +Ive tried this patch and now after the timeout handler the rest of received data is silently discarded by the call to so theres no longer an infinite loopIs this the expected behaviour? I expected the socket be closed when the timeout triggers +Ill open a new ticket for this and get a patch up this afternoon +Could be +just checking the type for it being Serializable makes sense to implement the class loading stuff +Sorry bout that! to the latest a lot for the review Eli +As a side note if the project is a dependency that is also installed Magma should notice that the project is in the workspace and NOT create a magmalnk folder to the unzipped sources. +Close this issueDevin +Mike would you be so kind and let beast break itBeast ran Lucenes corecontrib tests timesThere was one failure but I suspect its unrelated +Removing target release from issues that fit any of the following criteria No known way to fix them Java integration enhancements out of scope for release Other out of scope issues for work done but I want to do more in +This would be a tradeoff between making upgrading easier vs complete self containment +Yes I do not use JNDI I have property and so on +I did not review your patch with that in mind thoughPlease open a new issue for this +The schema is included only once at the start of the file +The same problem occurs also on this is the reason why the server in IDE stops +I dont think we have an owner +OK just tested things out with looks good +Thanks for the good work +Could you post your example for Hornetq? Btw +Im unsure of who causes that should be looked could your team look at who is causing these and report it at GWT and elsewhere +Because this change introduces expressions in and not in we need to add reject expression transformers to the management model to reject expressions in the case that an domain controller is used in combination with a host controller +This should be generated with the tool +tpstats on both nodes showed one active gossip task and a bunch pending +Merged to the branch in revision . +I have applied in SVN rev +Resolved by +shad +Here is a patch which adds gobbing support to +Im not really sure what the issue is +Also for String types we should always propose an empty string +Great thanks guys +LGTM +Fixed in trunk +Namit Sorry I created the patch using a local Mercurial clone of the svn repo the a and b come from Mercurials diff implementation +The issue we had caused by our separate release of unreleased package under the solr was that maven is seeing our repackaged dependency under another artifact id so it cannot prevent that a project adds and because it is two different things +Closing this for now as we believe to have fixed it please reopen this issue if you have a build which fails to Too many open files +Can you please provide a complete code example that reproduces this? Thanks +resolved in revision close of all resolved issues. +I should have the complete code ready to go by against CVS HEAD as of Pacific Standard portlet classes have been moved into their own module which is in the zip file +Ill move the rest of the discussion on the dev list. +This isnt something we get that many requests for in practice so it is unlikely to happen in the near future unless someone wants to contribute it +I think it straightened a few things out +How does this compare to the patch that you submitted to the old Thrift list +Wait I think I was wrong before +Am sure this is due to using a non transacted datastore so closing +The alphas of are close by but currently its focused on compatibility with the new code and a process to handle pom model changes hasnt been built yet +manifest support for Java backend committed in r. +Quickly generated k random inserts into a CF with secondary index but then I experienced that repair got stuck eating on both nodes insteadI did not have time to figure out if it was due to some config issue or related to the same issue though +Eric Riebling replied to the issueHi sorry for reporting something that might not reproduceIt very well could have been a problem running the linux shellversion of the script on a Windows installation where it justhappened to work because there were Cygwin equivalents for everything. +We could change the default algorithm but disabling the tests seemed better +Otherwise I would have never figured it out. +That is youre only allowed to declare one argument to your Principals constructor +In any case he must realize that something is not working as expected +No unfortunately it doesnt help Solr and it also doesnt help lucene +Just make my patch work for a simple union query shown in description +Created a transparent version and set the project icon to it +Youre welcome +trunk patch +At least the compiler should have failed in both cases +The should react on the coming up and then register its servlet +And if I dont agree with the opinion that this is cleaner and thus needs to break the configuration +The fix for that bug will very likely fix this bug +Ive merged the pull request and now the build fails because the test wasnt changed to match the updates in +Were happy to support this if it is known to speaking durable subscriptions are rarely used in combination with JTA transactions which is why this hasnt been noticed before +If the core has the same name as the collection then its not a violation of that principle and from what I understand a install does name the cores the same as the collection +The base attribute and the xslt attribute are using the same code +thanks Ben now I can see bimargulies from Nexus so Benson you should be able to deploy closing ticket +I agree with Brandon but this patch looks fine +Moved to an open version +attached proposed general solution in +EliAhmed It looks like this is causing a lot more issues +above we dont need to store any extra state on disk +You ask for columns and get values in a tuple its sort of what youd expectPerhaps we take your suggestion put that into deprecate the current and move the current code to +Attached new migrated version for peer review +Will upload patch for Trunk +Applied the patch with revision edecdadcfcfThanks! +bq +Please implement this +But I dont like too +Moved to issue fixed in released version. +Thanks. +Ive not confirmed but I think you will need to test this on actually require you send something likemsg nickserv identify the server to acknowledge you own the this probably requires is configuration to send a set of messages on connection I couldnt see this in hte fix +Included is a seemingly working patch +packages from the exported packages apparently +Nice test looks good +We have corrected the issue +trycatch removed as per request +Error reported in +testhealthmonitor failure is due toHowever I am running the entire suite of hdfs tests to make sure that nothing else is broken +on backport +Attached management server log file and cloud db +I dont think so the logs just ended abruptly +It would help if I could login to Upayaviras laptop to test +I have some issues that I want to discuss further more +Thanks for the code cleaning Kousuke! How about merging to this jira +I can take a quick look later today though I cant promise when I can do a full review +I went ahead and did some cleanup on the images +Attached diff removes the extra layer cleans dead and classes reduces some duplicate enums +Can you update to if not +Jason is this a problem with the profile service mapping or the underlying mbean value +This change should be applied to the branch as well. +It would probably be a good idea to do a patch for that version as well +Final patch for review +Resolving this issue as I feel it is done +Im also connected to the internet via a proxy and Im facing the same m using and version of the increase the priority +This was because the script would pass user supplied parameters to Java while making a call to check for the system architecture +backported change to codeline. +I completely agree that is the change I am making +Submitting a patch to address this issue with a testcase +Assuming lazy consensus committed patch in revision +Very basic getting started section should at least help give the user a good impression of how easy it can be to set up a minimal Cassandra installation +Thanks for the patch Guillaume +I looked a little bit more on this and now also thinks it would be the best to make jgoodies forms a required dependency +Im going to close this as Invalid since adding the Apache copyright notices is not appropriate +What is the timeline we are talking about here +deals with the error cases. +When not set xulrunner is available pages are rendered properly but after switching workspaces there is mentioned error +Thanks AlekseyCommitted to trunk +Contributed by Vinay is Ok for me +This is not a symptom ofRegions stay in transition states I think because the master is not allowing opens from HRS it thinks are overloaded yet the HRS are insistent to open +The concepts of that API match with jBPM and the work has already been done +Ill hold on to this so we dont accidentally prematurely convert +But I did some mistake on the commit log +Thanks for the hint version works fine and I finally have no SNAPSHOT dependencies +Im going to keep this bug open but the answer will likely be to add an API to configurator to save state information +Documentation is on the way! +I will also mark the other constructors as deprecated +Fix applied at +file being generated with all needed variables exported and PATH properly set +FYI was committed +Fixed various issues around the integration of the teiid sequencers integration with AS plus integration tests removed slfj dependencies some minor code dont believe the has any tests being ignored via Ignore +Youd likely have to come up with some code suggesting how the proposed feature might work in Shiro otherwise this issue is at risk of being closed +The fact that the jar file has to live on the local filesystem where the job is launched from is very constraining +My opinion is the best way to know if something is broken or not is to run it via binhadoop or binhbase dont let maven help with the classpathbq +Ciemo this issue is closely related to filtering bags in nested issue has the same root cause as hence marking this duplicate the fix for this issue will also be tracked in +Anyone with Java skills who has time to debug the Java server side of this exchange would be very handy right now +Hope that is fine with youThat would be great +Via Andrew OliverThe discussion was that this would not be labs but would be proper but no one ever answered my question +Will fix in my morning +Thanks +Jonathan Tks for your patch I also learned from it! +No response Ill close this as fixedBtw it baffles me when people vote for a ticket but doesnt bother to verify the fix +I will try and upload a test case within a day +I have verified that the session is never closed and the ticket is valid throughout the test +Consequently any portlet that attempts to generate a is failing at the momentThis patch simply removes the import from so that the correct implementation is used +Moved back to has been releasedMoving all related issues to the next version +This patch adds a new ivyconfiguration file called which has the details of the resolverThis new resolver would resolve dependencies from the jboss maven repo when the central maven repo doesnt have the required artifact +I am attaching two patches one for trunk and the other forI have added a new test case for both patches +Committed the SISM changes and the proposed protected method to trunk in revision +Added to title +Please verify in ER on zip distributionRPM distribution contain CRLF line ends as well +this patch adds the workflows to the browser capability of refreshing a node in the tree +In this case I reviewed Todds patch +Its possible that Im simply stubborn but Im not seeing how moving from a documented structured convention to a ad hoc one will better communicate changes and set expectations +I Think returning error to the client is not so good +Patch from rb +Ive had this a few times on my build server too +There is no other way to do so in Windows at all +Setting fix version to look at these in Beta if resource leaks even have an effect +it seems like we do optimize away the entries to the btree in the case where a doc doesnt emit anythingThe other weird bit is that I thought we had fixed up couchbtree to disallow duplicate keys altogether by now +Updated to describe changes in feature generation and the effect on the existing spanish ner model. +The related is +I will fix it and commit the changes +bulk move of all unresolved issues from to +given that we now display all of the token attributes im not convinced staking the tokens will be a good idea visually the biggest downside of stacking is that the row headings get very very complicatedPersonally i would just vote for better visual cues when tokens have the same position + For public installations these files are not allowed +Since natively handles incomingoutgoing attachments it makes sense for the service to be responsible for copying any attachments it might want to include in the responseThanks Ron +Thanks for the suggestion and pull request +Upgrading from XWSS to is not a simple upgrade as I believe the API is not backwards compatible +Adding the files to the VC project and setting vc setting seems to correct this problem in my the is OK +Not sure why git isnt working though +some side changes Rename to Ignore APPNEWSAVED event at FinalSaving state this event is ignorable because app is possible to move from NewSaving to FinalSaving and APPNEWSAVED will come +See comment above and in the zip file for more information +AlexeyFirst alternative isnt OK +on the patch +going to close this as will not fix +Closing issue out as part of JIRA cleanup +Thanks Colton for writing this up its very detailed +But if it doesnt see a header it wont do anything +Thanks for the review and testing Thomas! Looking at the patch for the final reviewcommit +Can you please suggest some fix regarding this issue for nowRegards Ajay Kumar +Bulk defer all frontend issues to +So It will fix some of test failures +Hmm how to workaround inthis case? Any suggestion +Is fixed in recent versions of Maven. +Cool! +Closing all resolved tickets from or older +Looks good thanks +Since im not a commiter uploading patches was all i could doHowever i think base code for hbase shell is quite up to parand it would be better if I could have it uploaded on source treeThat way we can do more collaborative development for command classesPlease let me know your thoughtsPatches I submitted failed to buildIt was clear that jj file wasnt generatedjline library was probably not there eitherDoes anyone have any advice to get it fixed +had defaulted to +First the Active state needs to be cleaned up it should transition to Standby it should connect to Active Namenode then trigger a check pointing +I took a looksee at shell and it seems to basically work +ahh it does work +I need to do a minor change to fix a test failure +Sorry for not looking closer +make sense +This might actually be fixed before +I reckon that you have found the problem + to community level no need to verify +Please fix this issue as well +Apparently having a type main creates a class file with a return type +I misread didnt realize its another sanity check not another way to reference the file +Or other headers +Marc thanks thats a very good point +that weblogic tmp folder is deleted but I can try deployment again check whether that folder exists or not +Fixed with still exists in campo branchVerified on build to put to stable branch the branch QA is currently testing +Screenshot of the change Note the added to about fix version? Why is it still M? Should be Beta shouldnt it +On my setup where it worked and created a vm on cs is having python version of +We then need to make sure the created by the uberjar laucher has the system classloader as its foreign +feel free to reopen or create new one with link to this if I misunderstood it +If you are not using the XML configuration then you would register the singleton instance programmatically via +As it stands Lucene does not support Java Serialization across Lucene versions +Arq hangsThis needs debugging +patch for the +Any more news on whats going on with this patch? Is it deemed worth incorporating into the project properI have a version of Brian Ewins patch that I merged against trunk r +The logger obscures the details of the exception you get the exception class name and its message and nothing else +We can revert those two patches instead +Could you please let us know where to find a snapshot of it +I cant think of any drawbacks off the top of my head other than the fact that the dynamic attributes feature will not prevent users from arbitrarily adding attributes +I will close when I commit this patch +The error text states the problem of this issue +Well write a handler to validate this and inform the web service the following I think this bug can be closed +Any update on this? Spring Integration is in version already. +Actually you are write the java doc look ok +Added him to watch list to comment +Lets not do that +If I cant help I will roll my patch back or merge the both patches +Your changes should remain +No I dont think so unless youre planning on doing what I mentioned in my previous comment +Also there might be code in mapreduce that could be making use of config based mechanism for choosing file systems +Ive added the suite at r +I tried to fix all of them but I dont think that I was completely successful +It is actually easier to just change the names of the variables +Ill update the trunk patch to do the same +I will update here soon +I put LOG statements in each of the catches and am not seeing them in any of my logs +updated version with slightly better documentation for the maven site including examples +settings print find +This issue is fixed by. +Anyway committed at revision with French labels +If a test fails all the time it has coverageThats not true! This feedback and a variety of output from the fails over time is invaluable! Even knowing it can pass sometimes and not others is more valuable information +avpr so that we can automatically use the right compiler +Related to +And my part and part are really only for engine +Updated JIRA title to reflect the underlying the forum thread for the details of the cause +Sorry. +I want you to look it over clean it up a little bit play with it and see if you think this will satisify all requirements +Rob do you have any suggestions on how to implement this? Obviously Denny I havent done anything with it since you encountered your initial frustration a few months ago +I think that looks much better. +This issue should fix itself when moves to the API +I cant think of a better solution but TBH Im having issues understanding the problem at hand +The seam profiles and the jars that it is dependent on have been checked in +Hi Amar do you want to review this one again? Thanks +Thank you Tanping +Make Pig create output files +Now that work is well underway with AS all previous community releases are +Please comment +Committed to trunk. +Ill take care of that as soon as contriblaunchpadtesting builds again +BPEL bbrodt reassigning +Fill maven bug its an Archiva bug +Released. +Because my setup has changed since than I cannot provide any more informationYou can safely leave this issue closed for now and Ill definitely keep posting information if I can +I havent been involved that long with IO but I believe dependency free has been one of IOs aims perhaps others that have been around longer can comment whether this is the case or not + was not not for every new even though the constructor for was being called for each new +No definitive answer has been provided so far so I suggest we consider this issue as solved for the time being. +All other test cases are inside the JAR file +Thanks Ben for your valuable insights +It looks like the applied fix is incomplete +I will submit this as my first patch +Note that my config above is not exactly the same as it will not handle and status codes +It seems a number of TCK tests fail if you follow the spec religiously while others pass +Ah OK Ill go to it and make a patch added +Patch adding the to the http applied in r +discard new andor old transient messages only create persistent state of persistent and transient messages throttle publisher speed up to stopping then when memory size is critical We can obviously combine for example and +I think the main performance issue with clone in the old Token was that it had to do the init before even though clone already knows what size the buffer should be etc +Additional patch applied at . +It would seem something might be wrong with the call to the get method +Is this causing jobs to fail for you or just the occasional task? What OS and JVM version +Debugging SCM related events is not easy +What shall we do here? This would be great to get into release but is it safe to do so +That is a bit odd but the old logic just shutdown the thread pool to avoid accepting new tasks +Im browsing through the applyDeletes call path Im tempted torework how were doing this +I cant reproduce it anymore +However in the old workspace the checkbox seems to be gone forever since I created a Java EE Web project project with the unchecked checkbox to get the demo code +One thing to note though is that with using the Properties object to write the text the current date time stamp gets appended at the end of the warning text +Thank you for looking at this +Fix makes getBT wait for the current Task in that CF if any before returning a new one +This bug is cause by I modified the prototypeso for most of time its I think thats possible some user also might do just suggest to add some defensive code to to avoid +TC on RHEL +Rather than move it I simply added and to the Fix Version +I had a little bit of problems applying your patch but eventually managed to resolve the issues +So i think this is working as intended +Please size accordingly +Specifying an appropriate on the fixes this +we could make this optional +please see +Perhaps I am misunderstanding your comment +Please reopen if Im wrong. +remove codes change in +Udo and I have separated the layoutmanager from select See for the task Please test it now If you want to use a tcselect with a label look at the txselect tag please +faces the same issue +Thanks +QUOTEWe have been using Hibernate for at least two years now in a large ERP application that has been successfully deployed in production +This Fix requires the improved dotnet toolchain making a followup on this applying this we should also port and from that tested the old xsd to check compatibility +For this issue I do not want to change any of the write logic codeI am still debating on the replica class hierarchy +I use a to decide what to with the +Cleaned up and improved version of the initial patch +I was able to reproduce this once but not anymore +Note the comment +Sorry I missed your +I think we can bend the version rules for this as long as we make it very clear in release notes that this is an exception to the rule affecting only users of the new builder code generation in the Java implementation +Im resolving the issue +I think that this is actually a bug in the spec allowing something like this to exist in the current contextOn the other hand OOM is not acceptable either so I will try to patch the code to allow for the behavior described in that thread +Have verified the fix in a production hadoop dont know if the value can be differentBut to be sure I will change the Value also +Startup will be done in but the rest is not going to be done +Thanks Adrien! + +Optimal solution would be building a slab allocated block cache within java +WDYT +Ill upload a third patch that is the sum of both the first two + +It doesnt need two maps +Closing. +how about we name the class? remove unused imports +Calling would be an additional call which is not necessary as far as I understand it +You should use and to search on numericdate rangesClosing this issue. +Id also like to run a benchmark on the new code to ensure it doesnt add any slowdown +Thats the rub +Currently user attributes and metadata are persisting correctly +Option applied on attached patch +Additional svn commands to be executed after applying the patch +committed +Must be a apache config +Startup time is dependent on the database being used and is a non issue anyway in JBM where all new feature development is occurring. +I plan to work on it in the coming weeks +ThanksFixed in rev queues are viewable only for continuum this get added to the documentation? How does it workIm concerned about users being able to see things in the queue for projects they dont have access to +Attaching a patch to set to Affects and Fix in Affects and Fix in Version +Proposed patch +Patch that makes history work WITH written history filesThis can be used in place of the other patch +Linking this issue to +I think we should still throw an exception when the LIMIT clause is used withcomplex SQL statements in +See for further details. + +Please review the patch and give me some suggestionthanks +If you are using sharelibs you can just set the config for and it will work nothing additional to do +I added a color conversion for jpegs using a separation color space in revision so that everything looks fine now. +Close this issueIf you find a bug please file a new issue. +PatrickNow that is out what are we looking at to get your LOCALSOLR to work with itThanks in advance +If we are having multiple clusters then having a proper cluster name defined by admins will be useful +Fixed in rev +Done for seam +I will put this fix on the Roller branch if it applies +is from the SPL +Thanks for updating it with a test case! ve rolled this into trunk at +Could we please apply the same fix +patch doesnt address your comments just the test np I figured the test fixups werent going to be a big deal +These are the new patches corresponding with the new proposal +set AUXCLASSPATH to satisfy and above +Split into and can we please reverse the naming? The JAR containing just classes should be named like a normal library without a classes suffix +bulk close of resolved issues as part of JIRA housekeeping for release. +I incorrectly changed the wrong attach reversed the last change and made the change to the assembly attach so that the javadocs would be attached to the mead repo +Certainly every time Ive seen that error if I manually query the Cassandra database using then I do indeed see that row present on one node and not the otherIn the case of the stack youve posted above it looks like the pycassa client is failing to connect to the cassandra nodes at all I hadnt added any exception handling for that case as I hadnt hit it at all in my testing +Let me be more concrete If the is at testunit it is always ignored by both mvn clean install and mvn grails +Please use libs with this patch +Conclusion after IRC chat with Heiko leave Config Admin section for the moment. +Added file permission setup for key tab files +Applied last patch in r +Found the issue +Ill try something to fix it by this weekend +George which Git client are you using? If it is Msysgit what options did you use during installation? I have tried to reproduce the issue and I see the error that you described but contrarily to your case even the Forge command client doesnt want to execute my Git commands +Thanks for this full and accurate explanation +We could document it as such or get rid of it +Thanks guysDone at rAshish +I tested with on in Chrome Firefox under Ubuntu and cant reproduce bug +On Windows you have to add the exit errorlevel at the end from theend function +Why do we need to be more strict? AFAIK you cannot kill an incorrect job right +Does anyone have an answer so I could start sharing my udfs and hopefully help out +Most probably it could be like schema not set +Selecting or anything contained in that map does not enable the paste action +Closing the issue now that it has been resolved. +By debugging I noticed that has no effect in this OS hence thats the reason why it never disappears +For the upgrade script well need to specify DEFAULT on ADD COLUMN so existing tables remain ONLINE by default +A quick comparison of the proscons of this solution can be find herepros no more parsingnormalizing when searching this simple operation costs around of the global time no need to get the DN from a separate table as we will have it in the entry +fixed and confirmed +I have created a new junit the confirms the problem +I am not sure picking fastest tests is the way to go for derbymats +lhm to lhmlist +Also I could not duplicate this on my local machine +need to change fixed for +no this is not rev I cant find it back +Is this a stock or a patched versionCan you also provide the info level intermediate plan from the server log using set showplan debug +I checked our and verified that in that case the stopServer is only called once +Jasons lazy init fix works nicely for me +yuck +Generated has been overryde manual generator to include conten of a manual cgeated configuration into generated file +Closing old bugs. +Lets not get hung up discussing that here +the current patch will return true for ll look into it +I dont like that fact of creating Threads and then throwing them away +There are a lot of Plugin classes that depend on +All satisfied intents are added to an publication property +When I install Jboss Portlet plugin the issue is goneJboss Maven Seam integration Protlet plugin worksJboss Maven Seam integration alone doesn maven seam integration should not depend on jboss portletWhy it does is beyond me more likely that jboss portlet defines the missing dependencies that jboss seam maven is usingbut good to know +It woks fine for the tests Thanks. +Unfortunately the change for this issue caused a problem in the build +The simple fix is to add a xercesImpl jar into libendorsed I guessI gonna test that and evaluate the impacts with Oracle JDK +Hi Lenny I have checked in a fix for attached file is actually the wrong size the original file was x + +I wasnt able to reproduce it in my RC testing +The transaction was committed and next transaction was not committedrollbacked yet +Change on exception part +Ill let the developers close this as duplicate +Lets create a new branch for this development +I also modified all places in the code that I could find which passed to pass the new constantBasically a trivial change +sorryplease change status to duplicate. +Probably a lot of tests to fix +Just to save anybody else looking it up it looks like the fusedfs binary will moved in Hadoop which is still unreleased as of today +Selecting the feature ids out of the original data source worked fine using the resulting feature ids to update the table works like a charm +It turns out that the first patch was syntactically correctHarsh added something in patch v which wouldnt pass compilationCurrently Hadoop QA wouldnt post back if there is compilation errorHowever Stack wasnt aware of the above and integrated patch vThis is another reason we need versioning in patch filenames so that such mistakes can be more easily avoided + since whatever Ovidiu did probably broke something +Thanks for info Patrick +Could not reproduce + +For now you just have to accept the way FOP implements it if you wantto make sure your file renders youll have to clip the images yourself so thatthey fit into the allocated areasBTW it would have helped if you described you expectations a bit more ahem detailed +I think that there are some collection fields that also contain fields too +Verified in JBT +Maybe just the lines with ignoreCase false +It will just skip calling the method if the condition not satisfied +Looking forward to your patch! Master is definitely the place to patch +There still are lots of false positives +Most wont care aboutIf we keep it as a separate component then it puts an additional burden on people to remember to configure it and figure out where to put it before or after the normal facet component +More is the pity I was hoping to use this screen to learn what makes WCS so highly recommended +Create an abstract initializer class that can except the arguments and instantiate the new analyzer +The snakeyaml was indeed the packaging issue +Closing all tickets +There is already some caching of but it may not being used properly in this context. +Will leave this issue open while the test is still flaky +Ive run into the same issue +Sending srcmainjavaorgapachecamelcomponentfileSending srcmainjavaorgapachecamelcomponentfilestrategySending srcmainjavaorgapachecamelcomponentfilestrategySending srcmainjavaorgapachecamelcomponentfilestrategySending srctestjavaorgapachecamelcomponentfileTransmitting file dataCommitted revision +Attached patch solves the issue +Why are these and a whole lot of docs jiras moved to EWP? Shouldnt it be EAP +The problem can also be caused by using the Sun JVM +Sailaja reopen if you can reproduce the if this is still an issue +Updated the patch with number of reducers for sample job set to +Patch is available for result of derbyall +New patch with fixed test +Basically it is written to test the producer reconnect timeout so it is timing dependent +With those variables I really cant figure out whats been changed +Verified with +The upgrade looks good. +The bug isEclipse Compile Error This project needs to migrate WTP metadata +Without fix this test must fail in +However please submit a failing script +Suspect this is substantially obsolete after the rewrite done in. +Check fix into with revision +Requestor does not resend after a failed handshake +host findbugs make valueOf public +class has been added to the tags with the incorrect following templates are affected ajcommandLink ajhtml ajstatus ajportlet hcommandLink hmessage hmessages houtputText hpanelGroup fverbatim richtree jsp in and revision is reproducible due to fix. +The patchalso includes a simple test to exercise the new constructorThe zip file for patch contains the binary images thatneed to get copied into the test directory netuitestwebappsdrtcoreWebbugsjA svn add for these binaries will be requiredA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjA netuitestwebappsdrtcoreWebbugsjI just grabbed these images from the test for J +That works and I think is the way to handle this by design +Committed +Attached is an updated patch proposal and an updated HTML version of the man pageI think I included all the comments from this issue +In the end thats the ideal isnt itExactly +Thanks ErwanI definitively close this issue since you opened +Committed changes to +This may break partially some applications which download resource with +It will definitely be very useful to have the information for debugging purposes +Attaching full hit the same issue again with ibm and this time much early +But I wonder what I can do to get my launchers working again +Thanks Cyriaque for your awesome work +I was hoping that the final library would be a one stop library for AMs to communicate with all the NMs in the cluster +Hi Konstantin Updated the patch for branchCan you please take a lookThanks Can you please take a look +But first I give it a run on my maven +Thanks +Its great Mark +Daryn has raised several issues of which the duplicates one is a showstopperAnother is where one may not have permission to follow the symlink here listStatus will have to throw when a child via a symlink is not accessible +You are right! I forgot to revert this test with the null and use the extracted mimeTypeChristian could you please apply the change +The forth attachment +Id be happy to review and test some patches +Patch adds to all classes +Fix javac warnings +We do that anyway with the but no need to do anything more unless there are existing handlers with more interfaces. +Test projects showing the classpath issue +Theres no way for us to provide a generic namespace to override system properties since even if we scanned through them all and filtered by prefix we wouldnt know how to cast them to the different values needed not to mention in some cases like datafiledirectories wed have to use CSV or something to represent an array +GaryThanks for reactingI went down another path and used the GA release +Thanx again for taking my issue in considerationI sadly dont know much about CXF internals to really understand why these threadlocal variables arent removed just made a few investigations using heap walker toolIm on holidays at the moment will try to send more information as soon as Im back to work +Being a scanned PDF file I couldnt trim it down +installing with the installer and removing the native launcher what you are saying is that it is better I use hours of my spare time to build a new installer based on an unverified assumption than you using minutes of your time to verify the hypothesis +What I am about to commit +There doesnt seem to be a large of number of files but it just goes really slow +setting the locale to the default locale just to have it changed afterwards if the parameter exists is not very nice I guess so I changed that as wellFurthermore the user should probably be notified if he supplies an invalid locale but Im unsure how to proceed in this case so I leave this for now as it is +Should we prefetch in multiples of the HDFS block sizeThe cap is for the all the prefetchers Im not expecting to see each prefetcher to get more than a few of data +The controllers are already ready and working +Closing at reporters request. +Activiti Designer only works on Eclipse Indigo or a newer version +Yes like I wrote in the first comment its a dirty WIP +Working on corporate approval to contribute this patch +HBCK also does not handle this case +I recommend submitting a trouble ticket +For long time Hama shell may not proceed well +You can see there a workaround +Also feature provide multiple host select that I find usable. +If for instance one row isgb the rest are bytes then compress is likely necessary once that one big row is deletedIf everything is working right no need to ever call compress +Looks like this is already taken care ofCApacheFINALaxismoduleswebapptargetjar tvf grep codeg Wed Apr EDT lib Wed Apr EDT libCApacheFINALaxismoduleswebapptargetjar tvf grep mtomp Wed Apr EDT libCApacheFINALaxismodulesdistributiontargetjar tvf grep codeg Wed Apr EDT lib Wed Apr EDT libCApacheFINALaxismodulesdistributiontargetjar tvf grep mtomp Wed Apr EDT libthanksdims +is this still an issue on trunk +pushed +Does any of the people on the CC list have a different databaseinstalled on their machine and would be willing to run a couple of tests for me +The issue is valid for EWS too +Right now the back up task is generated during the execution time +tests dont give confidence things are going to work in usage. +But of course this code also is from Java and it might be better not to imitate itSo if you strip out the finalize parts of my patch and apply the rest its fine by me +Also it supports Recommender +Patch applied in out issues. +Implemented as an internal patch instead. +This issue may be solved already by someone else closing this issue +Marking closed. +Does this mean I need to setup a connection? Is there any reason not to have a connection in a management service exposing Derby of additional overhead etc +as soon as you give us your feedback that everythings okay now we should be able to make a release forThanks in advance for your feedbackAlso please notice going forward it may be the last version of the branch so also the last version of Groovy being compatible with JDKSo if you can consider moving to a newer JDK that would be a good thing JDK is even already d +Verified patch on Windows and OS X +defeating nonblocking ll tidy up a patch apply it to and post it here for any remaining discussion +Workaround is to edit to point to other mirrors +I have renamed this ticket to Roo Spring MVC scaffolding support for JPA custom primary keys since this is a new feature currently not supported by Roo +confirmed by TomClosing this issue and deploying a new deployed. +reopen it until it get reviewed and AshwinSame issue as I commented for it doesnt work it jdk +NiceSome comments though There is nomain method There is no license header Arguments are not checkedAlso can we link this ticket to the ticket in which this patch was first proposed? I forgot the ticket number +At this point some of these changed Entites have no parents anymore others do +For example layoutFormselectManycenter and layoutForm id for is not that straightforward but not complicated either +Dont know. +About half the source files stillhave the standard apache license header. +This is not a regression but just a particular corner case in this context the value of newcoverage is certainly and this value is considered as a best value by Sonar and so not saved in DB to save some space +Thanks for the patch Ill commit it soon +Erik Going by what you say above you are making an argument that KV and R should just be creating a new class a by passing in a KV and then giving it to cascading sounds kinda painful +Ported from trunk to branch at subversion revision +Thanks. +Support for added in revision . +I sent email asking for a security review +API delete is provided +Basically Ive moved job listener calls after the default exit status value has been set +It contains the scrollbarThe following is a screenshot from Central that doesnt contain the scrollbarI dont understand what you meant by and +Incorporated some feedback with commit ceeabfeeacedbeefbfda +Released in will be part of +Thank you for digging Thomas +The version I have does not reflect this fact +The problem I saw was intermittent but did cause the admin web interface to crash + contains a similar patch to batch the view updaters writes to the btree +An exampleSay we update a bundle that implements some API and the bundle containing it because the API has been changed +There is a marker in HTML editor only +I still see in the call stack I think it is referenced two places in the +Else Ill have a go at it +Henry good suggestions patch updated +Silly diff +We should probably return +This is adressed for application jobs in and for all jobs in + +Committed. +Rick I assume you didnt close this as fixed on purpose +Thanks for verifying Roman +Nice work and nice +Please see patch +Moving to. +Rescheduling to release sooner +Sorry for being so lazy and not applying it sooner +You can request to activate something that doesnt exist but I dont agree that stating a profile has been activated is an improvement to the warning +Was it Cassandra version prior to you used to create a mt keyspace? I cant reproduce that on version +As the error handling and reconnect logic is the same for both a heartbeat message and a quote request message I wont actually save any time by sending a heartbeat message first +Username letiemble +Van can you clone this to Teiid +Comments +very much RickCommitted patch to documentation trunk at revision +and in the end it would be the same +Thanks a lot Eric +On the other hand I guess its still better than having the explicit Xerces dependency +Whoops wrong example in last patch had core for the name +The JCR spec defines the node types and pattern to represent files and folders folder nodes should have a primary type of ntfolder while file nodes should have a primary type of ntfile and should have a jcrcontent child node with a primary type of ntresource and on which the jcrdata property is placedThe current structure is okay for any other storage connector and really is okay for the file system connector as well as long as youre not expecting the outerFolder to be stored as a directory and the to be stored on the file system as a file with the supplied contents +I have mixed feelings about that +I appreciate you looking at this +David has not been backported to eclipse yet +Tagging for could migrate a rootdata volume with snapshots +Oh +I have added a trace +One should be able to easily scan a jobs log for bad thingsThese are all internal right? If so it shouldnt create problems to change them should it +Ill attach a patch in due time +Specifically it looks like it doesnt include all the changes for example doesnt exist +I hope this helps +the trouble is that the dependencies are so out of whack that Im not completely sure which is can try and correct it but how clean are you guys keeping the libraries? Can I assume that if its there it should be included +To avoid the widespread whitespace changes I rebased it by hand and not using git +Verified upgrade in ER to upgrade to to pull in another fix +I meant to actually apply this as a while back and it fell off +There was even a TODO in the source code there already +Hi Rajini sorry for taking so long to respond +And more aggressive when shutting down in doShutdown. +I cant think of any way to write a test case that wouldnt be so fabricated as to be pointless +Good pointI will ensure that this get fixed +Forgot to add that the options do exist under Other +Manik I am changing this to and I also assign it to you since you will be running the final benchmark anyway +Thats whythink better to have a way to specifying seperate files +please post the algorithm and explain why it is wrong +This came back as ofCould you confirm the groovy version you are using when you get this error? Otherwise I am planning to mark it as Cannot reproduce as on both an d what you have reported is problem +Thanks for the help +I have attached the simple project that I used to recreate this issue +Thanks for your time and I do appreciate of it if you can give me a hintMany ThanksOscar +Is this an often occurring problem? Im wary of extending the rules in basic normalizer because they will affect a lot of users and running each new rule adds certain cost in terms of performance and possible of the regex engine +Thanks Rick +If we switch to using javac then we dont require ECJ any more +I dont know how to do think a good example about this is missing in +See attached log +work in progress +properties file realm defined for Geronimo the solution for this JIRA would be really simple We can make plugin as an explicit dependency of inI can provide a patch if the solution is acceptable +However it can be reopened after +Minor correction the sort took instead of +Linking to which documents failure to load in and its child container processes +Marking resolved since the upstream fix was included in the latest rebase +Yes was able to get it to with and without this and the client was the bottleneck +Please see the prior discussion there +New patch builds on Sijies patch but also cleans up how ranges are constructed + OK the above comment is a little misleading +fixed in r +Can you run the HDFS tests locally to be sure before committing +Im an idiot +why its not resolved init is impossible to use the feature now +from me too this seems to have fixed the problem +Current Seam examples do not suffer from this issue anymore +Added plus and minus methods to allow offseting by number of weeks days years months seconds minutes hours milliseconds etc +Heres a patch that implements this +Lets keep the discussion focused on that +Can someone change this +Earl you can go ahead with the MR release whenever you are ready. +Can you create a phabricator entry for this +when the was ther same height as the it would look like this? WDYT +DeepalThe patch I provided earlier had a classloading issue with the way the sample is being deployedI have fixed the problemPlease use the and filesThanks +Two tests always failThey fail even without the patch +Well I think this has been open long enough +Need to remove hudson python code use jenkins plugin the patch for review. +Just moved the class we use on the server side into the client side should fix the issue. +I did not remove the old code in +And then association of Payment to break and it will available for another batch +Verified on CR +With this patch srctest is removed and respective properties are moved to and +Sorry +action File New Other +Please see my previous comment +It has been resolved but resolving it exposed bigger issues with regard to the ASinfinispantorquebox stuff all working nicely together +A value of true remains in the config file +bulk defer of open issues to + +Also ran the standalone hlog tool with a bunch of threads and verify +Work in is an API change in the DSL so lets wait for Camel. +The problem seems to be when I specify defTypeedismax under defTypedismax it is working like a champ +If the problem is fixed please close this bug +No FO testcase to reproduce the problem has been submitted in reasonable time +From my point of view the patch looks good but Im not sure if I overlook all the consequences of your modifications and if they are correct +Nice patch Elliott +Changes delivered +Victor the parts wont work on stable because GML support is only on trunkPPerhaps this is just for trunk +The patch needs reworking and will take some time to do so were putting this off until a subsequent release +I will open another jira for the same. +Your code snipped is the solution for this kind of the patch I thought and will fail according to the other comments on this the patch only should I right with thisFor and also Integer can be used internally +Oops forgot to add one of my source files with svn add +The exception that you are seeing happens because the method no longer exists in Eclipse +This will allow us to step through code +Ill open a new issue if necessary. +That is the reason I just left it to get the review from others +Many thanks +Ill attach my so you can see the order of validation rules +Moving to the dev list as suggested by Kevin. +Maybe its because Im tired but Im unable to grasp the logic behind that +is not specific and still uses Writable because client assumes a writable format +Yes ive understood but adding a method only to pass a reference of impose to rewrite part of the subclasses of Encoder to effectively use the transfer functions +Found a problem with the test case +Fixed +aggregatefeatures +Do you think theres anything we can do about it? If not then we should probably document this +There has been no definitive answer one of the project despots to post a comment approving this account and we can make it happen +In the mean time you can look at the unit tests +Fixed with revision +Attached patch for Jan please let me know if the attached code patch works for you +Committed to trunk after review +is still listed as a dependency in configs which means will still be used a lot +Patch Reviewed patch depends on +Rhino is shipped with official jdks and is mandatory so thats a stupid freebsd issue +Fixing that and the issue Phil noticed allowed the test case to complete +Ok I applied the patch but it seems to break something in test +If the other fix helps you guys than its ok but I still have aNico +Additionally note that when I say Client I mean standalone client +Committed +I take that exact approach with data to ensure all docs for a specific user reside in the same shard and the query is only executed against that shard and returns docs for only that userThe downside with that approach for your use case would be with a potentially low number of unique values that are hashed you could wind up with an uneven distribution of data across the shards +But it would be great if this functionality would be available on the tashi command line + file contains source code class and libraries +When deleting the alias the cpu load goes to seems there is a infinite loop +Because of special interpretation of the +Fixed the bug I also handled the TODO in deleteN +Daniel convinced me I should use the parallelism value from the cross since whats really important about this is how many join groups it creates +Updated patch to address the comments +Pushed staging to official site +For the purposes of whats a Null layer? I can then look through my layer groups and see if any of them qualify +Did you restart the nodes repaired or did anything elseFYI Ive tried removing the row cache but the problem remains +Im pushing for it in logj probably as a standalone companion jar to logj and support in the Configurator classes +defer remaining issues to +We need to remove the test resource coresrctestresources the above changes and rebased +The current code only accesses the Wicket session and Wicket makes sure the response is committed only when the web GUI code is done so there should be no problems anymore. +With respect to refresh and sorting documentation in chapter on webserver has been enhanced to explain the variations +as long as we dont return null i am fine with any solution +Never was +we should address it somehow +The JS include needs to be updated to write to the contents of a HTML div element +but not everything doesnt make any sense to me +We also have a team that is doing PDE development +Oops forgot to set component +Jason +Patch for trunk +also given anyone requesting uploading of bundles will in most cases be providing such details and will also be a logged in user of this jira server there is a clear trail of accountability +The kernel did the correct thing +for this patch +So its ok but still it woul be much easier fo me if I could make a mirror with development components in for example of today and as mention above if I try do do a rsync I get more than GB and most of the space is occupied by +We probably should add a test case for OFFSET in the new test fixture +Thanks Sergey Im looking into it now +Attaching README files giving short descriptions of the replication test framework running replication tests test framework execution properties used by the test framework +moved what looks like changeslargish feature requests to is an ongoing issue moved to bumped to for now +Anyone looking at this issue +Thats fine but in the future its best to include a full description of what is causing the error to prevent wasted time on both our parts. +Yeah sorry wasnt pulled into dist server done now. +AFAIR Achim added a in a contrib folder shipped in the Karaf distributionI dont think its a good idea to embed by default in the Karaf distribution and http feature is an optional oneIts clear that we really need to think about a profile mechanism supporting this kind of situation +I have disabled clover in the nightly build until this can be addressed +This bug has been marked as a duplicate of +Gerhard Was not aware CODI implemented this +Resolved for a while +Not sure if I should create a new jira or just use this one +Changed Description from because uses from system properties +Sid can you please take a look? Simple refactor in Eclipse +Also mentioned in and there +It does also explain why Darran couldnt reproduce the problem so far He had an AS instance setup on a publicly accessible hosts +We can say that in the documentation +Not only does it break this case where the same block body is explicitly passed through multiple levels of calls but it breaks any case where the same blockBody is used twice in the same call stack +I really do not want not to make more complicated than it isIf somebody wants to delete a couple of column families and a couple of columns it can be done with multiple roundtripsNow if the code can be simplified by passing a Delete object then we should do that +This patch applies to the trunk and the branches and +Just add it to the end of the list. +Closed all tickets +The patch has project Dheeraj Sorry I havent had a chance to test this patch +See the related issues for more info +Assume closed as resolved and released +Probably that is not the clearest implementation since the new thing is called a UNICODEMGR not a CCSIDMANAGER but this was the quickest way to implement the prototype +We can update the test case +This missing Affects Version +The full path under testresources is serializationjavautilThe implementation is missing one thing at the moment localized messages for the exceptions +On Win with JDK it takes long only the first loading then selecting other options like hardware and os is without any let it opened while I find whats the problerm with FC and JDK about some caching logic so the information from the cloud is only retrieved once per session. +Done +Im not sure if this is even a problem that can be addressed from inside the framework +If you feel its important to handle please reopen. +or notMiles can you try a recent nightly build? Were not caching the list of type names anymore as far as I are correct and I probably should have tested it with a nightly first before filing the issue +Im not working on this +The attached zip file contains the metadata files for configuring the IDP as the Passive STS in the +I suppose we could generate a dummy requirement for this situation to stuff into the exceptionI will look into it more later +and all ant tests passed +Phillip is going to recreate this and create a JBPAPP issue for it and probably a JBMESSAGING bug +Perhaps the pure q case should be hanlded more generically by this other issue anyone else tested the attached patch? Should we aim to get it in for? Needs some unit tests +for patch +What do you think about this one? Does the performance improvement justifies the increase in code complexity +part of release. +reopening not to Doug +New patch addresses feedback above +Can you attach your full trace +Were reporting an error but one that we can recover from +The stack trace is useful but please supply at least some context around this bug report +I just had to add the edge case of resolving a parent node or path when the target itself is the root nodeRemoved test from the element in since the test now passesFixed in revision . +So the hornetQ work that is already done in is this experimental? Or should the community consider this +They now form the base for our most robust set of datastores and are also the new official starting point for new datastores +Works great for me +Bulk update fixversion to current ui for better ui mock is existing ui is fine if choosing the connection would actually show the values in the text fields and that it would say Get values from Connection instead of Use DTP connection since its not really using the connection once the file have been created +Finally +I ed the fix but where are the unit tests? we had such earlier on for stacks afaik +Id be very happy if youd find the time though +This appears to be fixed in +Everything you read from a database should be considered likely too change anyway +All are down. +But we need to fix the extra job issue before next release +I first got the assigned to a list +JayaThis is linked to the bug that you just fixedthanksdims +Sorry Carl this wont make it in the release which is about to be released very shortly nowMaybe it could get into the release if you are able to send us a patch or a reproducible testcasethanks I can build my own ivy jar with my patch for now that would unblock me so I will see about magicking up the patch +generated from Savara +to Chads suggestion +The test fails can you try with this patch +When would the JSCL be exactly? It wont until the classes that were loaded by it were all +I dont think it was my commit so Im bisecting +Column pagination and row pagination need to be handled separately + on +added a build profile shell which installs the Apache Felix Shell and Shell Remote bundles at startlevel To build with Shell included set add the Pshell argument to the build command +I have applied the patch plus additional correction to pass integration tests see revision +Out of curiosity would you mind posting here the exception +Feel free to use the code +Thanks for the clarification +Exactly this was the reason asyesterday the new Apache license was applied with much less lines +The npe itself isnt dangerous +Andrew you updated the js +svn ci modulesaxisAdding modulesaxisSending Transmitting file dataCommitted revision +Do you have sample data and a command line that you used to run the program +Thank you ant script to start the test suite probably so it bind to the wrong interface +The two properties can only be used for starting the server not for deploy commandsYou can try to set them before server start and see if it worksSo basically the problem mentioned in this jira are all fixed by now please close it after verification +Can you guarantee that it is resolved in that version +The sum of the two retries cased ERROR NSRE and failed MR task +testing failed on my initial fix still trying to determine how to address +Will Seam be able to authenticate with information from LDAP +Since macros can be loaded in a parsed file the specific macro might change each time the page is rendered +Can we break that out into a separate JIRAI went ahead and filed +Smells like language design andor array lib redesign so retargeting to +Fixed in revision with the release of Sling JSP Taglib +Since this is not a Bug theres no Steps To Reproduce section so Ill add that as a commentSteps To Reproduce +This issue will be fixed against the version of the Oracle fixed +rev +Can you supply a curl command or other http request description that allows us to reproduce the problem +Fixed a small but important bug in Scala Option deserializing and did a modest tidying. +I think this needs to be fixed for +For example in a piped command line the input stream is not a tty but it is also not a file +OK I have now also got rid of the dependencies to. +My for your patch regardless +I like your full description but for checkin it is nice to have just a short summaryThanksKathey +Table EnableDisable This has to be better than what we currently have +be on a slow connection and see progress bar show up for central and cancel it on startup +Did you try the include option on the database reverse engineer command? This will cause the columnDefinition attribute to be populated +Im going to work on the patch a bit more today to use the unicode variant instead this allows the full k path length that FAT NTFS and so on supportIll have a look at the function as well while Im at it +Have a nice dayAndrej +I will ask on mojo dev list what to do about it +The issue seems to be a duplicate of so that should be covered in the tests as slipped up twice creating the patch but should be the one to use +The API to is made private but is being used by +Use this unit test the mapping file and the test entity now have the correct Java package declaration +Problem was a missing ifdef for bounds checking +My is taking around minutes to run which seems to be a lot +If there are no objections ill commit the attached patch to trunk and branch soon +putting this into and +Please look in history for what the actual resolution was. +Looks like this should be looked at for +Perhaps you could use some internal caching to prevent the expensive operation from occurring more then once +I think the admin page should be rewritten so that it doesnt use its own algorithm to calculate the EPR but uses the same methods as the WSDL generator +When will be resolved there will be no references to jars from repository but links to other projects in Eclipse workspace so you wont have +I meant it about assigning the issue to me! I havent updated the spec yet + +On restart failed to run system Vms on machine because I ran out of management +No that is not possible +Im happy to step up and moderate +I dont think we need to includethe in the binary packaging since we already provideit unzipped it only needs to be in the maven artifacts +All failures are unrelated and because of port conflicts +On the Oracle Support site there are maintenance packs for versions and below but not for that is the latest version +Delete ActionButton just not working for Archives easy to replicate +The console uninstaller still works as expected +Viacheslav please provide some example to easily verify this or verify it by yourself +Please provide a patch as diff so that itll be easier to apply it to the codebase +bq +The new javac warnings were for deprecated method calls in a test Ive now fixed that +Thanks Alex for looking into this +bq +What jars are you using? We need to be very careful with +Please add it to and post a JIRA +it will be very helpful for testing of partial availability scenarios +Shouldnt that be instead of ? You may have already passed the size the first time you checkYeah thats a leftover from when I was trying to avoid having a spillableRegistered boolean added to bag overhead +Added the comment Ben suggested +etccloudmanagerment or etc this record since is now released. +Please use ajrepeat instead +Thanks +bulk defer issues to +Should continue discussion on css class conventions in you Malcolm. +Changed option to +The thing is Im not against any of the solutions +Thank you. +Ideally we need to be able to act on the event that the leader and isr response is either negative or lost +as to an +This patch fixes the missing help screens for branchI will look into thisRegardsDo you mean this should be closed as duplicate or still the patch reworked +Will commit after jenkins blesses it +I thought we had a JIRA for the index but it must have been closed as will not fix or it is buried under some other heading +The solution is tracking what resources were already rendered to the page +Another level of efficiency can be brought in by preloading the string table with well known strings like responseHeader etc +Linked to general improvement in error handling for +Is the issue fixed inPaul Zeeman I cant reproduce this +That might be because the trunk Solr Connector is the first to use and has not yet been released yet but based on what you report it would seem to be more related to your particular way of developing +Of course I may do some workarounds that for example first serialize the stream using Drools implementation and then pass it to the external component but I prefer to try to do it as it should be +I just ran ant test and it passed +no longer has a method add that takes two arguments +Urls are normalized in Outlinks constructor then validated and filtered in +In any case the entire derby server should maintain some control on open filescurrently the sort module can eat up open files per connections in worst case +Ugh +svn ci serverSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientAdding serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientAdding serversrcmainjavaorgapacheopenejbclientAdding serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbclientSending serversrctestjavaorgapacheopenejbclientSending serversrctestjavaorgapacheopenejbclientSending serversrcmainjavaorgapacheopenejbserverejbdSending serversrcmainresourcesejbdAdding serversrctestjavaorgapacheopenejbserverejbdSending serversrctestjavaorgapacheopenejbserverejbdSending serversrcmainjavaorgapacheopenejbserverTransmitting file dataCommitted revision . +Added logging to onAquire method also standardized logger messages text in +What is happening is that Java is optimizing a method that has synchronizers incorrectly so that certain member variables get updated outside the synchronizer +Please open a separate ticket. +And in fact the fix has already been applied by someone +about naming the hook post? addressing Anoops comments +I seem to recall this reported directly to the geotools list as well some new wms element is not handled by our current parser or something. +What is the current status of these tests? Are they removed? Should I fill a CHALLENGE for that +My only question is whether this pattern exists anywhere else or only in +The end result should be something like for best result on this dataLook at the yellow leaf nodesIf a leaf node can be able to represent more than one vector when the distance between them is below a threshold the yellow would be gathered as a single child to the branch with between siblings +However in the readFrom method the correlation id was not passed to the constructor +The reproducer does not reproduce the issue anymore in WFK on EAP I can still reproduce the problem with WFK on EAP so it seems this might have been fixed by a Mojarra update +Committed r in order to reassign to all issues that have been in resolved state for more than one month without further comments to closed status +patch for patch includes a test case +Once you mentioned everything sorta clicked that probably makes writing the code REALLY easy which is a reasonable goal +I downgraded and it worksThe easyb example itself doesnt work though due googles +update fix version since its not fixed and has been released +For a couple of reasons +And the change is really trivial +As the difference between m and m are the poms perhaps we can set up a systemof symlinks for the artifacts in such as a way as to not duplicate them +It will probably be tomorrow before I getit tracked down and fixed +Waiting on +LGTM +Attached is another rev that does this +Closing bug. +Any update on this +ChrisThanks for reporting this +We now have an abstract delegator and a delegator for PMFCDS PMPIDS +In the where it works you will probably commit for other stuff +We initially did put some new serdes in contrib or rd party libs examples include and Zebra serdeIf you can make it work with existing Avro serde it will also be great +That will also be part of the future workBasically Hive supports all the queries even though rewriting of some queries is required +Need to confused this with another paging issue in GWT proper that was fixed from GWT +How did I register a property path? Is there some otherbetter way I could have implemented the constraint to avoid this issue? Thx! +Looks like is a common denominator of Put and Delete where as this is a container that holds a swathe of Puts and Deletes to apply to a single row they are differentIm up for rename +Sorry I missed that earlier +Ill commit that +If the Entity has Hibernate throws a +New pull request +Last remarks applied +The element is for configuring or properties +Attach patch to include minor ui fix for selectionaction sample pageprefill gadget url field with correct context root +is running ok in our case is not so the error must have been introduced between these to versions +Thoughts +I may have some time to test it tomorrow +a Map solution that allows for unknown json keyvalues to be handled +We should probably support the mechanism hbase provided via QUORUMADDRESS +Yep this should stay open resolved is when the patch is being integrated into the subversionI will take care of this shortly. +Should be ready for +Not sure about the other stores +Forgot to assign back to Nick +Hi Gertthanks for the report +This is correct the there are breaking API changes from WTP to and the plugin will need to be ported over +Attaching log +IS that the semantics of? To return values in a collection property as top level results +Closing this issue +This patch will only compile once is committed and the respective jar files from common is copied ot the lib folder of project +Lets discuss in the deployment forum +OK for your first comment regarding the other comment it has nothing to do with the EAR plugin +Added the documentation page to the page +Thirdly the times dont seem to match what I would expect +Adding pull request for the refactoring work +Using the same classpath container but with a server runtime jars are added to the classpath +libUsersdmitriysrcbin +I cant reproduce too +Unfortunately something changed with recent Java updates and the key bindings stopped working +This is now fixed and available on npm cordova. +On the other hand we do need to test the checksum exception as it is a code path +So it can not cat zipped files correctly +Currently used functionality is bash c command which is supported by standard sh is against trunk +If you as an expert do not agree you can expose your arguments at the EG level Im sure the change proposed is valid and probably it will be included in a further version of JSF but I have to be strict in the spec interpretation more if I was one of the guys involved in the discussion of this feature +ShrutiI think it should work on +I sent a message to the list to inquire about this part of the GRANT implementationIve been studying how DROP COLUMN handles triggers which mention thiscolumn to see if that holds some clues for how to handle whichmention this column +A simple command line Pear installer based on the example code from the UIMA documentation using for argument parsingUses the PEAR name as default installation directory to avoid clashes when different use the same component internally +Much better Stefan +Sounds like a good idea! can do the same +Went through the suggestions on most of them were already addressed +servicejmxrmilocalhostjndirmilocalhostnull Kernel connection failedThe http port never got set either even though the server definition file in the plugin contains defaults for both these values they never get set when defining the server programatically +ESB message service will implement its own behavior cf +Just because has an option doesnt mean other filters should have it its hardly a model citizen +Yep at build time would make sense +Fixed in in the releases +The trunk version of debian doesnt contain the changes from Pauls fix in commit ffcbcefaeacffaaeacdYes it does +Bulk closing for +deployHaving the separate file may be causing problems similar to thosein and +The only way to avoid the issue is to not rely on function replacement of currenttimestamp +trunk +Just not to have passwords in the clear stored in the repo is a big plus +Looks good + +Thanks Mark +which does not sound right to me +This is unexpected +Here comes a little example which will allow to reproduce the the attached example Im pretty sure that both datasources would be which would not work correctly +See Deprecate multiplex transport +My internship at Facebook is overPlease email jsichi for inquiries or ccy for inquiries. +which can give appId dagId and also start monitoring through YARNAM as necessary +The connector module is migrated. +Perhaps you should examine you maven settings +However as you say you have good reasons for doing it this way in +Do a right click on the tomahawk directory and create a patch. +Please test against the most recent CVS snapshotWire log would really help +The signature is just too small to reliably detect the split +im not exactly sure why this is not working in JBT +Agreed +Let the master send the message to the region server or let the region server query the master for safe mode statusIf we do not want to stop compaction while in safe mode then we need to decline to close a region for redeployment while there is a compaction happening +This patch is for HDFS trunk +Patch is very old. +I thought there was a specific reason why i needed to obtain it that way but i dont see it anymore +I guess its underspecified I always assumed sequences to start from zero +Released with GEP so I close this JIRA. +the blocking bug has been fixed submitting patch for reviewcommit +Patch fixing the to hudson +add missing +Sample test case checked in +This is adequate for my purposesbenson +Ill debug and try to find the problem +Related fails because Pig depends on which doesnt play nice withI guess we should bump up hbase version to or +In this case I think that you can check zoneId to determine if it is + +Sure +Quite often patches need to be attached for multiple branchesIn this case should we include trunkTRUNK in the name for the trunk patch +No need for Netty to be on snapshots +All other unit tests complete successfully with this patch applied +Reduced the extensions to submit queuesize and errorlevel +Verified with using CR build +I forgot to mention +Yeah I confirmed with of the other files experiencing this error all corruption issues similar to the first one +I am of course borrowing the idea from the information we have about when the last heartbeat was received from a TT +The back and forth chatter could be an issue +How does your loader know about partition keys? Is it through some conf variable or looking at path stringYes Ive used hadoop path globbing but it has its own problems +r +Patch on top of whats already committed to go the default rackdc route +Sorry for that. +This has happened to me several times and Ive noticed similar reports from other users. +changes look good +which dont have PB RPC but have to be used until +This may cause the VM to stay alive for up to minutes longer than it should have to +Fix javadoc warnings +Hi Philippe Sonar uses the property node internally to manage several occurrences of the same Checkstyle rule +Does it help if we force a commit at the end of each file? We could do that with +Even though this ticket was opened first Im going to mark it as a duplicate of because that is where the active discussion is +Still make public need to invoke create outside Hive +As I mentioned previously this is fixed in my local saandbox +Thanks for the review EnisI just committed this. +Please go ahead and send the patch in +This is implemented on my experimental in the Logj API. +Tried reproducing this over the weekend but was not able you please attach a log file showing the error? log file from ER embedded default server start up +cleanup codes +pending a clean Jenkins report +I also did notfeel like making class coupled with interface +Right +It was pointed out to me that there is typo in the submitted patch file where HTTP is spelled as HPP in file in a comment line +Let me work on other two +Let me know if the format is incorrect +Do you have any idea how its happened +can you try the latest master +Milos thanks again for your patch +This is urgent because the problem has existed since before and the patch has been available for at least two releases without being appliedThe workaround is a select statement with cases if developers notice Cocoon is brokenI successfully manually updated with the changes in this patch +I dont know if the is going to be a nuisance +As for notmodified not being in the index then nothing wrong will happen to it when we ask the indexing backend to delete it right? I assume your point about the fetch time intervals is that it does not differentiate between when a page was last fetched and when it will be fetched +Ive committed a patch +Thanks Tanping +So the approach of the proposed patch is not sufficientI then looked into specifying different delimiters for +Im fine with reverting behavior to the if it gets addressed in as we can workaround it +Filtering on path is disabled by default +Thanks again AntoineSending CHANGELOGSending libbuildrjavaAdding specjavaTransmitting file dataCommitted revision . +Can ignore that +Not sure how this is relevant for Click as the framework only has one category Click and does not print out class names +Similar I imagine to how spring uses its create and destroy interfaces for beans to be hooked into the spring contexts imagine that this issue might be addressestargeted for Maven so on a tangential issue will it be possible to dynamically createaugment instances and add them to the Reactors list before it sorts them? I ask because I just last week started taking another crack at solving the Eclipse PDE build issue with Maven +I guess I wastrying to get it all in before but perhaps that was optimisticI dont remember the instance of a changed method name +Resolution is to deprecate support in favor of ve linked to the issue in RC for removing these classes +I have uploaded as well +The method just returns like it works +Password should not be needed at all +we just never backported to +Rock on then. +What makes you think this should work +Since is fixed I think we can mark this issue as do you agree +Assigning to Joshua to document in the Release Notes +Patch attached +But anyway the numbers are clear +Right +I agree with Mario that this is not a Drools bug as it violates the principle that Drools must be notified of any fact on this revelation its a potential issue in especially when doing VRP time windows which I ve described in. +I got the error because I was following the comment which said that the value had to be one of create update but now it can also contain the value validate and be blankomitted if you dont want database to be updated +This has been thanks for the help! Could you please check SVN to see if everything is OK +Can you explain whats the problem with and our current module layout maybe once I understand this I will also understand your motivations +Also I dont believe we are running with assertions enabled +If the VERSIONS file does not have a codecxxx entry then the NN assumes that the image is not compressed +Pushing issues to the next CR +Hi the site is right you should use cfang as the user name not the +When you upgrade one node and the other nodes dont have describesplitsex yet starting a hadoop job on a newly upgraded node failsAs for fixes I agree +is fine with merging such simple change +Those revisions also normalized the error messages so we can easily change the format to DD you cant call an instance method from a static context if we so choose +the wizards and are not set up at all to have more than segment configuration name +Whew +see the timestamp of the artifacts in the download box from the attached screenshot +Willem is this work done +Yesterday when I was debugging the file authorizor there was some weirdness when the calling isVisible was returning true even though the specification and role was in the blacklist and returning false +its not possible to return error to client instead just of request timeout +Version M has been releasedClosing the issue. +postponed to JBDS we can decide of a better value for the preference +I can see that striping the version would be good but otherwise the name would stay the same +If nobody see a problem with this new functionnality I will commit the MichaelYour patch is in trunk rev +Shall I raise a new JIRA +Dont know why it didnt work on your computer +Say you change the uninstaller am closing since that was merged. +Dutch CONLL files in ner format +Could be good to put a downloads section that links straight in to the download from the wiki page. +Where do you put your hbase configuration +Version +Fine point do you want me to change the patch +That was +If you run through then you take the hit everytime you build the think it would be possible to add an additional SOAP command that would directly stop the embedder +Im using Eclipse at the moment +Thanks Ill make the change +bulk move of all unresolved issues from to +You could at least doc it when committing +I expect a lot of information to be stored in zookeeper more and more as time goes byDo we know which zookeeper release supports storage of large amount of dataI think the solution provided by Ramkrishna should be integratedOnce gets tested and reviewed we can consider porting it to TRUNK +It might be a little complicated but yes. +I am using JSF JDK update +I will check +Thank you +toplinktrue +Your example will work fine for testing +I committed itThanks! +It contains a fix so that email can get sent +Updated checked into trunk and branch +This should probably wait until after the mavinization of HDFS +The safe conversions dont do anything to help the Quantile Function produce valid bins + Logging never fails to select a provider +I think bufsize has way much bigger impact than the implementation +I assume this to work the state has to be kept on HDFS instead of local filesystem +The attachment contains and config file +Use UDP instead and set receivedinterfaces andor sendinterfaces which does the same as UDPNIO +Patch bFixed a broken link +It happens that we chose to put there +sos I have to compile related specifically to compressionI think we have a terminology problem here +I did exactly what you say +Added CHUKWACONFDIR to Removed incorrect test Eric there are some conf files not properly handledcleaned up +Please feel free to pick up this jira +Can I use JDBC Persistence without Journaling as walkaround? Thank you for replay Its very critical for me +Committer! Congrats +Service with action +Ok +Fix available with the attached patch +Fantastic thanks Gavin! +The only effected interfaces should be +Fixed for upcoming release +I couldnt check it in yesterday due to technical difficulties but its in nowRevision Author karthickDate AM Friday April Message Improve Process Cache DesignModified odebranchesRakefileAdded odebranchesagentsAdded odebranchesagentssrcAdded odebranchesagentssrcmainAdded odebranchesagentssrcmainjavaAdded odebranchesagentssrcmainjavaorgAdded odebranchesagentssrcmainjavaorgapacheAdded odebranchesagentssrcmainjavaorgapacheodeAdded odebranchesagentssrcmainjavaorgapacheodeagentsAdded odebranchesagentssrcmainjavaorgapacheodeagentsmemoryAdded odebranchesagentssrcmainjavaorgapacheodeagentsmemoryAdded odebranchesagentssrcmainresourcesAdded odebranchesagentssrcmainresourcesAdded odebranchesagentssrcmainresourcesModified odebranchesaxissrcmainjavaorgapacheodeaxisModified odebranchesaxissrcmainjavaorgapacheodeaxisModified odebranchesaxissrcmainjavaorgapacheodeaxisModified odebranchesaxissrcmainjavaorgapacheodeaxisdeployModified odebranchesaxissrcmainjavaorgapacheodeaxisutilModified odebranchessrctestjavaorgapacheodeaxisAdded odebranchessrctestjavaorgapacheodeaxishydrationAdded odebranchessrctestjavaorgapacheodeaxishydrationAdded odebranchessrctestjavaorgapacheodeaxishydrationAdded odebranchessrctestjavaorgapacheodeaxishydrationModified odebranchessrcmainjavaorgapacheodebpeliapiModified odebranchessrcmainjavaorgapacheodebpeliapiModified odebranchessrcmainjavaorgapacheodebpeliapiModified odebranchessrcmainjavaorgapacheodebpelcompilerModified odebranchessrcmainjavaorgapacheodebpeleprModified odebranchessrcmainjavaorgapacheodeilconfigModified odebranchessrcmainjavaorgapacheodebpeloModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineModified odebranchessrcmainjavaorgapacheodebpelengineAdded odebranchessrcmainjavaorgapacheodebpelinterceptModified odebranchessrcmainjavaorgapacheodebpelinterceptModified odebranchessrcmainjavaorgapacheodebpelinterceptModified odebranchessrcmainjavaorgapacheodebpelinterceptAdded odebranchessrcmainjavaorgapacheodebpelinterceptAdded odebranchessrcmainjavaorgapacheodebpelinterceptDeleted odebranchessrcmainjavaorgapacheodebpelinterceptModified odebranchessrcmainjavaorgapacheodebpelruntimeModified odebranchessrctestjavaorgapacheodebpelruntimeModified odebranchessrcmainjavaorgapacheodestoreModified odebranchessrcmainjavaorgapacheodetestModified odebranchessrcmainjavaorgapacheodetestModified odebranchessrctestjavaorgapacheodetestModified odebranchessrctestjavaorgapacheodetestModified odebranchessrcmainjavaorgapacheodedaojpaModified odebranchessrcmainjavaorgapacheodedaojpaModified odebranchessrcmainjavaorgapacheodedaojpaModified odebranchesjbisrcmainjavaorgapacheodejbiModified odebranchesjbisrcmainjavaorgapacheodejbi +this patch also deletes the pid files once the process is killed +Fixed with bug occured because of nulltype type as a result of type nulltype as in java gen code +the patch looks good to meMarking this as patch available so Hudson runs +This is not really a bug +Closing as this was released in +That is and will remain a problem because there is no CLSSTMT in DRDA +Thanks Ben +Should I report this ChineseJapanese issue as a separate jira or should I keep it here +I think makes sense if the underlying store could be any DB like Oracle DB etc +I removed older versions to avoid confusing +Solid user testing is now required +Is this still an issue? Based on Patricks comment I will mark this not a bug in a few days unless I here anything suggesting that I shouldnt +Im cool with the partial solution for now but please file a jira to fix this later +Committed r +added to the release notes as resolved Different JCA Adaptor can now be specified globally or within each ESB deployment +introduces thrift changes with a stubbed server method +Thanks +Whoa thats great +This is the scenario where we cant locate the service op and then return that HTML page for GET HEAD and DELETE +Thanks +Hi I am in the process of adding iPhone support for my app +Test reproduce this issue +I modified the description to reflect the current state of this bug last week +Deferred to defer of unresolved bugs to +Great work! I will review commit in a few minutes +decompresses the underlying compressed input stream block by block +Patch against trunk removing valueType and adding a test for void return types +Just looked at code and there is no way to clear jvm args so I did a try and figured out that jvmarg works out of box +Felix your patch applied without a problem this time so however you did it differently it made the difference +Thanks for the reviewbq +Hrmmm +Ill ping Andy Clement once its in the bug tracker hell take +Thanks +war from CATALINAHOME +Having negative tokens was made on purpose +The original initializing database +They have logged it with Atlassian and I will post any additional news as it is available +Ive corrected the leakageAlso Ive found that new version works too fast and could exhausts all windows listening sockets +set of revisions for trunk Backported to with revisions Backported to with revisions +Updated patch to fix the findbugs issue and use code in which is now committed +The specifc tests now work. +Please do +It sounds very confusing to me +Cool +Committed revision Now porting to trunk damn +Closed with wrong may be a valid issue for the individual +Patch to introduce an input method to and call it from the menu prevents the form from being validated when it is first displayed +Thanks Stack for the review +Gavin do you know how to work around this? Could it be due to the way hibernates insert the DATE field that could have caused the NUMBER value to be distortedI tested using the same JDBC driver and normal sqls and it was okay +Run another precommit build to check this test again +Im guessing that we will need to update out files and wiki page descriptions for maven integration but want confirmation +Patch with only the pom change +I very much hope that Bigtop will be based on Hadoop and then we can pull this patch in +well it wasnt merged into so he did the right thing +Patch successfully committed. +Removes the global lock from the code path that processes a heartbeat +In all the above tests i had concurrent threads that intern caused application code to get a connection from pool +On a more specific note it may be possible to interrupt arbitrary Java IO operations but probably not message dialogs since theyre not interruptable operations +but is this working with a jasper from of Tomcat? The logger seems gone and the option constants do not exist so I cannot apply this patch +So is there any update on this? Ive been using hibernate for a year now but now we found a bug in one of my queries and to fix it I need a second join to the same association +Attached is the test code I used +Id like to get this one committed unless there are objections +You should see rvztr tkrfrgp +However I added more code in and to check whether the writing API is called when the object transits to its final state +Please make sure that it does not print floatsdoubles in scientific notation but sticks with canonical + was not specifed in either seem to never have been fixed in trunkCR not blocking B +See s comment for the release notes related to the underlying issue at work here +LGTMThere is a minor typo in the comment says The rick is +I have something working locally that appears to be a good fix +verified on EAP +So I would think doing validation as a pre processor would catch a larger universe of errors and give the developer a better chance of providing a meaningful error back to the user +So the state store can maintain the all the tokens batches with sequence number range in each batch +This collation attribute can only be specified at database create time +Work on a database java API with a keyvalue implementation has been going on for some six weeks now and is nearly complete +As you suggested I added this path to forgot ot mention It is for just found a bug in please use the new found another bug in please use this again Sasha +This issue has been fixed and released as part of release +The bug is not fixed for Tomcat and Cocoon downloaded binary distribution today and had the same error +The reason for adding methods for this flag maintenance in both and is that does not have access to which hasthe flag +Same but with typo fixed in the error message must be accessed within a transaction +Whoops file is rollerwebclassesve created a macro in the that will display a categorys image + +Id assumed that it was participants that was null +Hopefully this should fix the offline scrub +Youll just need to change the Eclipse build path to your local location of the SWF jars +In order to not break legacy applications I think we cant change the meaning of currently legal settings for +And added additional fixes on Topic to prevent duplicates being sent if a wild card is usedSubversion Committed revision . +I diffed the script in and and they are indeed identicalok so I did some work on this and came up with a patch but someone with good shell experience should review it to make sure it should work ok on all systemsscenarios +Please in the parser as the other bug I refer to was identified as suchIMPORTANT This email remains the property of the Australian Defence Organisation and is subject to the jurisdiction of section of the CRIMES ACT +However two files with the same logical content can differ in other ways too User provided metadata different block sizes compression codecs etcThat brings up another tool we could provide a tool that does checksums on the binary content of an avro data file +I also tried again to adjust the release notes for +I really do want the draconian I dont see it as particularly draconian we just want all of our version numbers to be in one file developers who put version numbers in child nodes are typically just ignorant of the policy +Patch build has distributed log splitting failing but it passes for me locally +js and +Traced this to returning a null progress +This is essentially what the proxy object would do IF there were a new service model +If you mark the included module in pom as provided +If there are other issues file new bugs. +Hi NemanjaI never used PO and gettext what are the advantages over ou current XML files +batch transition to closed for remaining resolved bugs +Ran through all tests successfully +There is a good chance that this will also work +Not seen since commit so closing +It really is not pretty +Is that what youre looking for? it will be nice solution I think +It might be possible to simulate the issue in a unit test using a modified version of that is designed to hang in certain situations +seems to be already no there is another regression relating to lazy init of dates fixing. +I filed this issue because I got a totally misleading error message when I tried to start my bundle where the root cause was malformed XMLUnderstandable +Since the workspace is set to the default value the IDE doesnt open the Workspace Launcher dialog and uses the default workspaceAttached is a patch +Marko could you take a look at this please +Has that package been deprecatedTwo similar packages look confusing to me +Please bring that to JIRA is not an effective medium for that purpose +Cant repro +This is a for getting this committedWhen I sort or facet the work done up front is reused in some way +And only after the fifth commit there is a change the five rows are added as if processed in a batch +Current version will support both numeric and string values for indextype +Tested this patch by running mvn clean install mvn clean install +Shouldnt it be listed under +I found out that the old query translator didnt handle this in many cases anyway so I think the error message is better than generating bad SQL +I think it would be a lot better to do something like that on trunk where the problem also exists but the node manager uses netty instead of jetty so it is not a full thread that is blocked but just some memory and a file descriptor +The deletion in the bundle module was NOT intended and is not necessaryIll provide a clean patch asap +Ive opened ticket to track improvements to NTLM handling in httpcomponents httpclient +It seems Seam settings page remembers your first seam version and does not allow you to change your first choiceIll fix it. +I am reopening this one because itis something that will come up frequently and should be fixed +APAR link for trackinghttpsupportuidswgIY +it had some stubs that had been already committed +Hopefully covered by +Merged patch with latest trunk +The attached is the updated correct patch has been applied +This would mean that if I submit a with both nodes and the racks that those nodes are on containers could be given on any nodes on those racks +There are no tests included since the only way to test this is to pull a power plugThere is probably some negative performance impact depending on size of files dirty page limits available RAM etc but I think the safety factor is well worth it +Hi Thanks for the will define a jobService and integration bean and will get back to you with the a lot It works nowI am trying to implement Spring Batch Admin +This patch adds a factory command line argIt depends on the fixes to srctestresourcescontained in the patch in to applied with svn revision +Initial patch for review +This issue and has occured because while assigning if we get RIT exception we are not retrying +Jackson seems widely considered to be better behaved +Tried again with alpha and got the same thing +But we are not there yet +Patch for HADOOPHOME that runs a chmod ugox on all contrib bin directories +This depends on +Patch to add test cases to for encryption only scenarios in both symmetric binding and asymmetric binding +Bundles patch to add include missing files in bundle +Attached is a patch proposal to support the statement ALTER TABLE tablename MODIFY COLUMN columnname DEFAULT defaultvalueThe keyword COLUMN is optionalThere arent very many tests in this patch yet I intend to write more tests +Additionally we had to make sure not to call read on the last step of the path because it would be called by the read function +Please report this to the Embedder component in the Maven is the related issue in the embedder. +It looks fineThanks done +have been consumed by this UserAdmin AccountHence Reopening the images and api output with code built of you want to show the resource count on User Details page besides resource limitations right +Id recommend configuring your WAS to use port and also configure your WSDL and COMPOSITE to use port that should get things working for you +As far as know this is no longer relevant and should be resolved already +So in maven it shows +According to a post in the hibernate forum Hibernate is failing to include the yesno annotation so I dont know if all of this belongs in +contains the replacement filesThere are still some debug activities required +As such it does not seems to be a priority for the existing CXF community +I have made some changes to improve performance but you must refresh this page to start with a clean testThe reason it works in M is because the ajattacheQueue demo was not working they are causing conflicts with AJAX processing and both require a refresh to get a clean test. +Ive attached a rough patch from then to show how simple it is though youll have to the work because Im sure the has moved on since thenCheersDavid +new patch renamed setXXX to setDouble setFloat etc +Thanks for reporting Leon. +This benchmark reads and writes files using and HDFS and reports the time +pls use the new annotation instead sad you cant cause its dynamic +Could you please elaborate more? Im not following an exception will be raised in this scenario +Hi DanielThe patch looks goodOne comment I think the Yahoo line that you commented out should be removedOne question the way this is implemented the builtins will take precedence over user defined functions in case of the conflict +Still on the patch + add schema reference correct the referenced pom versions added to the exported updatedI tested the bundle generated and it seems to be workingI have also downgraded the version number to because the project has never been released and it is under developmentPlease test the changes and close theworks as expectedThank YouKind regardsSiamak Haschemi +Now also say it just so happens that the first edit in the batch is from A and the following ones are from BWhen we execute the Put all edits will be tagged with A as he source and thus none of Bs edit will make it to A when it is Cs turn to replicate to A +If you feel the word example implies this feel free to improve on the terminology We should strive to make Solr as as possible +The patch looks good to me +The problem is that we now build the livelink connector even though we dont have a livelink jar +Thanks once again. +Thank you Lorand! +Let me review it +This could also be a good trick for the from Solrbq +Ive updated version of swtbot to new on Hudson Job +So we need another JIRA issue to take this task further should dynamically load and enable support +I didnt know if it would affect anything else similarly +Is it saving time to localize the fileThere are some downsides to this +It inherits all settings from the HTML editor preferences +I was unable to reproduce with +Id rather flush more frequently than OOM more frequently +Verified on Tools +alright i got confused then +I hada thorough look at the sourdes but cannot think of such a solution +After reading and learning more about these issues +Dart your fixes look great +This is a post release. +Im looking at the problem but Im not eager to break the current command line processing otherwise Im unable to get the release out of the door +Committed changes +Personnaly I like it like that but if many others think like you I agree to changeSo other Ivy users what do you think +This issue only affects form that use get as the method and only when mount is used on the page that contains the form +This will require more permissons in the testing policy file for +Thanks Brian! +BTW the patch should use x not e +Thanks +Would it be possible to work around this issue by writing a custom class that contains a reference to an to which it forwards all method calls +The code looks neater this wayOne small comment Should we use latestTIP instead of slowestTIP to make it more clear +you for the contribution Harsh! I have committed this to trunk and will attribute it to you when you are added as a contributorNote I am assigning it to myself in the interim so I dont forget. +I agree we should just get rid of it +out of hunks FAILED saving rejects to file coresrcmainjavaorgapacheshirorealmldap luck would have it Im working on a default LDAP implementation as we speak +Please review +because of the inaccuracies of the above subtracting the actual number of flushed bytes from the memcache size leads to the potential of the memcache size growing over time if fewer bytes are flushed than what thinks is is the memcache +You can probably just close this bug and open another one for more patchesLooking at the host it should probably also be modified to do it in a way now that introduced a launching an embedding API for frameworks +Fixed on the Portal tag could you please tryThanks for finding this + +Verified JBDS Beta +For future patches please dont reorder import statements but thats not problematic here +Replacing the current cache with cache +Attached thanks Sam! +Simpler fix could be to remove Test from the file name +Patch against is attached +Ill create a JIRA in JBIDE for this +To clarify Liyins point above in our setup we have multiple clusters per application and works as the configuration while is the configuration within a particular application +It looks likestruts and jstl should be all it needs +Is this done +Im heading back out on the road tomorrow +The attached patch includes the missing line +I think that is a great ideaJeffs Dirichlet clustering code might be interesting as well +Bump +Add unit test for the new method in +Ok its been about a week and there have been no objections so i propose we close this saying this use by is ok +cleaned we can change the logging level of the warning. +Updated with Security Libraries again Anil. +Oh I didnt notice the extra s +Ill comment here whenif we try to upgrade to the latest snapshot. +Thanks for working on itIm also attaching a stat file showing the changes and a zip file with the generated HTML as is customary +Applied patch as suggested +as part of this or maybe in addition we might want to check when creating a that the prefetch isnt zero and throw an error or at least log a warning since you cant browse a Queue when theres no allowed prefetch +Id like to introduce a dependency on number of user but thats not available until after weve picked a value +Please dont use to ask questions +The work done so far has created a reasonable directory structure with a everywhere it needs to be +Duplicate of +I found the other failure that motivated disabling these tests +Yep NMS STOMP would be preferable for interoperability +No problem thanks Mike +Do we still depend on s? I thought wed forked all of the IPC code by now +LatencyTypeMeanSDththththththCQLRPCPSBBAfter this point I began focusing on PS and BB aloneh +Yeah so then Im going to remove the fix versions and just leave this hanging around because I can see this is going to be more involved than I want to invest time inThanks for taking a look +Check for CWINDOWSsystemdriversetchosts Any network interfaces configured with that address Files in your project that contain this address in particular this issue because it is clearly a configuration issue and there is no further feedback from the user. +Can you try switching the version of POI you use in Tika to be a recent nightly build and see if that fixes it? Sergey has been doing some amazing work very recently on HWPF which may well have solved this beta hasnt been released yet +Im working on a fix +Thanks for doing that. +Added you back in +I am closing this issue as wont fix. +This is not a preference import its a use this directory information to setup the asked for runtimes +Committed to CP branch changed theloadgenerator tcunning svn commitSending loadgenerator file data +found the problem +Please have a look when you get a chance. +Move it back in if you think differently +in +All unit and ee tests passed +Patch committed with many thanks to NikosFYI I made some code changes to simplify the implementation +This did not work well as the JVM only checks memory usage when it garbage collects +Attachment has been added with description Patch for this ve taken a look at your patch and the fix for the font name is clear enoughHowever the buffer parameter for the puzzles me +Resolving this +DhrubaI actually found a bug when I added the new unit test +Perhaps somethings missing in your or +Linking issues and will try to get this committed ASAP +Thanks +Jose are you using the Views or Developer Cockpit plugin? +and thats even extensibleSo maybe we could call the file this as fixed to cut the after release +Im testing the mvn site now before committing +I have the same problem now in another enviroment +It is extended byYes the code could be improved +Applied using as name. +So I guess PBUFTRAILERMINORVERSION can carry value of Ok +Annotations are Bindable and Vetoable +removed from taglib and +As you might have seen we reorganized the Tool Suite and refactored it into smaller projects that can be consumed individually +Reopening to apply fix to the code +Staging support created +Attaching patch to add the maven profile +should be added to wiki +Committed updated nav generator logic which should fix this. +It should be a random sample and should have a different implementation +This is because declaring a queue exclusive limits its use to the session that created it +Why is breaking compatibility acceptableWhy not default it to true and allow those that want the feature to set it to falseI personally dont see the reasoning +I just got really busy suddenly over the last few days +Thanks Vandana +If is true then just do it? Is another set of admin necessary? Shell shouldnt need a new command just do it instant if config var is enabledThe unit test looks ok but it might be useful to include a case where a RS is killed while in the middle of a schema change +assigned this ticket to Blake himself after I added him to the group jira group of myfaces committers +So a different form a serialization did fix itI understand your discomfort in dealing with race conditions since Ive had to deal with solving them myself many many times but I dont know of any practical solution other than reasoning about the temporal logic of the code +Its not just a problem with Spring beans +Wolfgang without an example we cant reproduce it will thus have to postpone it after GA unless you can give us an example +Verified in no longer need upgrade scripts +Patch applied at +First ran randomwriter with block size KB and then ran sort to make large number of maps +bq +For mode this does not make much sense but will come handy for embedded mode +JacquesSorry my explanations werent understandableInto a new patch I go to delete lines about interpolationbecause this class doesnt use any more this code +Antony this is not enough information +However the file is technically undeployed on the server +After my little fix can delete files +and yes i see the typo in the maven config and fixed it locally already +Ah ok i shall go fix that for you +If we did this we should link to a specific revision +thanks shravan! +I would add as a requirement that all this should be completely optional +A workaround for this is to mount the home page at something other than +Hmm no special point +So as you put it the hbase user would own the hfiles +Thanks Ryan +Im open to more suggestions about how to address this. +batch transition to closed remaining resolved bugs +Should we store the statements in the cache using fully qualified names +Hmm somehow I missed that in the book +I made a patch to avoid using the coming with JAI +Merged this to. +I had opened for tracking this in which I created a small +Otherwise the reader will always have a lot to read between two requests +I think this assumption is not true any more fix for ensures conglomerid is unique in +Okay cool please go ahead when you have test cases +Update after +I have added a patch to this improvement Please verify and commit +Please try with the latest build and let us know if it works +to Dougs comment and my ignorance of the existing class Im clasing this one off as invalid +I assume these declared dependencies were on purpose +unclear to me if this change must be made in if it is to be made at all +Merged escape for committed +An added functional test verifies deterministic replay by comparing the job history files +The autoform information is clearly visible in the screenshot. +I removed the flag Resolve Workspace artifcats in Run configurations and it worked fine +Resolve when Hudson confirms EJB CL Leaks to in EJB Proxy. +Simple to do so will be in +In case there is a way to really simplifyunify Query and Filter I want to know tooBtw +If you strongly feel flagging is enough then also I am ok with it +there is no default database the system database should be installed as part of jee features +It should be more intuitive by default +That sounds good to me then the wouldnt change again +this patch fixes the issue and also lets us continue if we find invalid snapshots +Closed after release. +Expanded use of committed in r +Trivial fix forgot to add synchronized to these two methods and missed it in the QA report on +trunk +We plan to use this in NFS work as a project private featurebq +will go through the code again +Given that this seems to be a requirement for R then the change must be done specifically against the TFS and will thus not get persisted in the HQ data model +The Developer Guide now has Embedded Agent section as discussed. +Which isnt the case so far wrt this jira +But I have no idea how big a deal that would be in the java code +Thanks patch is applied +jdcryans Should we pull this into +Similar issue may exist in the portlet as well will confirm and open separate JIRA +seems to work fine +Why is that? This means the Windows code base is diverging from the other platforms preventing this is why changes are supposed to go through trunk first +Dart can you please identify what you did and how to verify it? Plus wed really like for you to attach patches and ask for folks to review them before you deliver code from now on +Ram fixed an issue where empty lines in the test file would be interpreted as Dtest and thus triggering a full mvn test run +Sorry about that guys +This is the strategy used by the Lucene to vector converterb pass around the text +just for the reference adding a project that is a midified Roberts test case that uses embedded Derby DB and sets up the data from inside the unit test + Whats the status of the JCR taglibIts probably out of sync with the latest changesIll check how it works with the latest jackrabbit version and upload some minor changes ASAP +Do you know the origin of the file was it produced by Office or something else? And can you try running the Microsoft Binary File Format Validator tool against it to see if its actually a valid +Actually this can also be used instead of to verify matches JDO come thing of it +Logging was restored for all standalone profiles as there was discussion that standalone profiles even in EAP are targeted toward the developer using the Enterprise bits +I can update the diagram that I attached so that it is suitable for +Make sure to document the fact that the freqs being merged need to contain only mutually comparable values +Thanks Sijie. +Can you check whether there was an exception or an actual response the response content headers +bq +Thats the log starting from the application context bootstrap +Hi MarkWhat version of Linux did you install to? Was it the same version of the installer +Instead of Begins With we should prefer Contains for searching You asked to use Contains as the default option instead of Begins With +not common +John would you like to take a nd look at this patch to see if the changes are still what you want them to be for auto create topics +Yeah the idea here is not that we dont like Struts but that there are ton of frameworks that could potentially be supported under Shiros umbrella which would be difficult to maintain and support given our focus on security +Is this a duplicate of +Does not pass the CP criteria. +I will update both patches today +suggest naming it other than because our convention up to this is that a class with Utils suffix sits over in util package +A patch with test cases in the udiff format as an attachment would be really nice to SVN trunk with some changesOleg +Yeah just using dir copy meant to get the apache commons one Im sure +How does this occur? How does hdfs find out that the inode tmpfoo is dirty and has to be flushed to the transaction log +I feel that your suggestion is more appropriate therefore better to remove other check +Why has this bug feature request been dismissed out of handIf you had a disk logger or any piece of software creating a tonne of garbage on your system you would expect to have a setting or switch which would allow you to control the amount it creates +Tools Eclipse update to Tools Eclipse +If you need to release and you feel this is good enough then feel free to do so and Ill work on a better way when I get back +Replace the casts with calls to the INTEGER function +I shouldnt have relied on maven dependency tracking without doing mvn clean +This issue is not fixed on Apache CXF and will later be available in FUSE SF +On my box I also have a full Acrobat and when I wanted to run Acrobat full after installing the update for the Reader got a warningYou cannot currently view Adobe PDF files from within your browser due to a configuration problem +I can definitely help out with the clustering testing and changes necessary +Closing this issue as duplicate. +I updated the security actions to be consistent with the you check the fix into the EAP branch too? Thats the branch were using for Jopr dev +Latest patch looks OkayCheck out the failed tests +committed to trunk. +Thanks Joe! committed. +Also it looks like your Xml class is not used +we will run some stress test ourself problem is in most cases that mailserver begin to block before they are busy +I have now made a few tests and up to two lines of fixes in I seem to be running fine all with +This patch addresses all the issues raised in Jims last review except for the two questions +Thanks Owen! +If all active branches are using this issue can be closed +Just saw and relevant discussion there about complications +Yeah from my perspective it looks good to go +At first I thought this was a bug but the more I look at it this is like putting on the parking break then putting your car in drive and then flooring it! Its going to smell really bad youll probably break your parking brake and if you smash in a wall it wont be super obvious thats what you didBTW Our tests are included with +thanks for the contrib +ok thanks for reviewing lets discuss about version number +Hi Carlo can you post this to so discussion can continue there +The issue could not be reproduced in a Felix environment so it seems to be exclusive to Equinoxs bundle +It is preferable not to support it by default rather than maintaining a JTS implementation inside the app server code base in addition to JTS as we go forward +I think the problem is that the first space is not added +Nope Im having problems applying the patch +On windows cache fails to pasivate due to +Regardless of how we got here its more correct to confirm the existence of the dc and rack than just the dc so patch to do so +something like that +I think it will require some experimentation to be certain +Looking at the stack trace the referencing module is attempting to etablish a network connection to some database I dont know which one +Ning also take a look at maybe we should move over to using JDBCsqlline +I think we can have a new sub topic in the Samples Setup Guide describing these parameters +Thanks Army for your valuable comments +Committed fix to branch +Disable Java refactoring for Groovy files +I didnt make a note of it on those at the time +Anyway we ARE working on this but its not as simple as it appears at first glance +Committed revision . +Anders are you sure of your comment on adding container extensions? For example the Tomcat container defines the type registered in the class +About patch Is it ok to create diff for binary file such as? If no binary file should the patch include a diff file only or a diff file with a script? ThanksBest regardsIve never tried a diff with a binary file but if it doesnt work then just attaching the binary file and the diff is fine. +Committed revision As I forgot to add the new test in svn initially +However you dont sound confident you have achieved this when you ask for someone to provide some speed optimization +Patch from from RB +bulk defer of featurestaskswishes from to +Attaching a patch with necessary changes +Note that doesnt give a clear explanation of the rationale for the change +the way can you mark Grant permissions to Apache to include the code +Reopened because Fix Version should be set before resolving +Committed to and +Holy shitballs! I just spent the last two hours chasing the same bug OWindows XP bit JDK I can reproduce simply trying to resolve instant boom +In this case the file would not close till the next +Committed revision +Fixed at cBut it would be interesting to see the source code that produces this error +The test doesnt wait for all the nodes to join the cluster so it is possible for two transactions to write to the same key on two different partitions and then one of those values will be lost on merge +committed to trunk and + +In Hadoop mode means that a program loads the Hadoop library and manages its own group of processes +We have a different version deployed on our Hadoop cluster so its almost impossible to perform this testAnyway thanks for your suggestion Ill look into source to check +Reassigning all open bugs to the development listSorry for the mass mailing +Actually this pattern is subject to exception to branch optimization which is not yet implemented. +And yes we continue to lobby for better protection of all of Hadoops ports +It sounds Australian. +bulk move of all unresolved issues from to +If you find further issues please create a new jira and link it to this one. +I am looking into implementing some heuristics that may help us out but worst case will always be bad unless we come up with a completely different algorithm +I have committed this. +We absolutely must be able to add watermarks to the maps as they are subject to strict copyright so if its possible to upgrade the status of this feature request please do +Merge the patch into branch. +Verified during cycleClosing +Attaching my preliminary checkstyle config +Hi Jeff agreed +Thanks Brock! +code develop so far implementing the Mail Transport sender for client + +Thanks Edward but Ive already read Hamas wiki and tutorialsIve already created a raw patch for this issue +changing the component to metastore since the bug has morphed into catching jdbc drivers for the configured in +Nice that now returns a direct reference to payloads instead of copying data +for the patch +This should fix it in. +You should now be able to assign jiras to yourself a problem sir + for patch deletes a function called hdfsService from appcontrollersmainI see that the reference to that function is removed from appmodels as well but I still see it being referenced from isDecommissioning function as a property it depends on +Cheers paul. +doesnt seem to be worth it. +Disconnecting cloud has added value in the fact that you keep username pw url of that cloud +Dave This needs to be in +Its Red Hat Enterprise Linux and the boost from Red Hat for RHEL wouldnt build Qpid +But the gain with bulk assignment is its faster which we wanted inYes +Will mark this issue as resolved +Perhaps we can limit generated documentation to only include things whose annotated scope is public +So this proposal is notI assume you must mean behaviour compatible rather than binary compatibleAs far as I can tell if the code is changed to try double and then float the result will always be a doubleIm not sure that is much of an improvement +after release. +It may indeed be sufficientCan you confirm that it is fixed on master at least? Im going to mark this resolved in for now. +Once the build ends I will publish the results here +And this module is not a standard part of the build in branch anymore +Please your code against the latest snapshot off the branch and let me know if the problem still persistsOleg +Also there are more matrix operations that need different storage structures to be processed efficiently for example editing a single value of the matrix works better if its in Yale formatThanks +Attaching an updated version of that patch that allows you to set JVM arguments for Eclipse launch configurations +Ive just committed this to trunk and +Data transformation We need either figure out how to make and working in sandbox or we can use the Data Mediation service +Planning to commit shortly +Haha +Should be fixed +The report is done +This issue is fixed by but wait with closing it until is committed +Nice Sami! I was going to take the same approachLonger term I think we want to add a write api to solrresource loader +I set to in the new test +if you just run ant test from the its virtually impossible to know which tests failed because theres so much stuff printed on the screen +Incorrect syntax +Good on you Chunhui +But let me double check that +We could go the full way global configuration per request per propertyThis gives the required granularity while allowing to override it in any possible way +This issue duplicates. +Currently the SEI annotation are not SEI annotations can only be used for doclitwrapped services +To used it in webworks quickstart i just rename to to replace the one that comes with ww +I already have more than enough work cleaning up Hadoop +It looks like there might be value in writing out a message for alert +ill have a look +If youre using Cordova CLI then you have an Eclipse project in platformsandroid even if youve never used it +But the issue I think was that when using Hibernate without Hibernate was generating a whole series of error log items because it was looking for and not finding it even though it doesnt actually need to find itIts related to scanning for annotated classes because is one source of where to look for themI hope that makes more sense now +It just does not say what you want it to say +Finally it worked and I havent tried again since +I also tried it on a completely clean machine with an empty local maven repository all without the patch +Well thats one way to close a bug +Moving all unassigned bugs out to to please move bugs back to which you will work on in the next weeks +Improved change detection updated the reference release notes and added +Im not familiar with JIRAs workflow should I mark the issue Closed? +bulk close of resolved issues. +fixing will fix this issue +Not needed +Pleaseinvolve me in this +Thanks guys removing the duplicate page fixed the idea how the duplicate page was created but life goes on. +So do you want me to remove only the host option from the dropdown? yes i filed another bug for it +Ah good +Verified in ER +CorrectThis is supporting my efforts at learning which is entirely in memory +Fixed by simply testing for null + +Any suggestions for a patch +a first patch +I find this very surprising +and are ok to be done so if we have no alternativeBut a +Then again s constructor is being called with a byte positionI think I will commit just the changes related to cleaning up and those fixing the handling of the utf stream in + +atIt would be nice to have the file +Max I dont understand have you removed the EMF from the +A patch to fix this courtesy of Daniel +adds the requested ASSERT +Feel free to add more tests if you are worried about coverage but I think coverage should be taken care of by none deprecated tests what was missing was some check that deprecated version was working or not +Added test project +Sure +In my view this issue should be resolved as Wont Fix +a fix for seek to work reliably when it crosses range boundaries +Yes you are right +snamespaceid was introduced in federation +When I dont use the annotation in the overriding method it works +I do not know if that is the expected behaviour but at least now I can change my code so I remove all the Constructors from my ActionsRegards +The problem is that data is not flushed to disk when get is called +I have no idea why Google couldnt just throw the right exception to begin with but it gives me another shot to contribute to AOSP or Cyanogen. +Updated the patch to use blocking IO for commands so that we dont have any spinning issues with the socket channels +That fixed it Thank you so much +libs +example describe Describes current authenticated keyspacedescribe keyspace Describe this keyspacedescribe columnfamily Describe the columfamily in the current authenticated this ready for review or still in progress +My last post belongs to a different issue. +Thanks for the patch Edward. +Added to ticket that create should be a prefix to the method name not necessarily the entire name +Just doesnt seem think we should investigate the option Ales mentioned on the associated forum thread +Burkay came across this when trying to use the example +You can choose the developerConnection with connectionType parametermvn scmupdate dont see a real explanation of why this automatic fallback shouldnt be implemented +Makes sense +I will address all of them under this JIRA +This problem has been addressed in Spring Spring IDEs Spring Security libraries to include the fix. +bulk close of all resolved issues as part of closing items. +This goes both for files resolved from the reactor as for files resolved from a attachments in the repository are already interpolated +I applied the patch and tested it and it worked flawlessly no Alfredo issues and the NNDN came right up from an dist layoutOne thing I should have noticed earlier I think some changes to the empty file crept in to this patch by accident +I just committed this +Loading the whole list in memory will run out of memory in many cases +Committed in r +Yes I do. +New patch to enable the check only when the feeder has finished and allows for a configurable number of times to exceed the thresholdThere can be a significant number of exceptions due to the return statement used +Longer term we might migrate some interfaces from generic to that package +Moving all unscheduled issues to +All it takes is a cookie policy that overrides parse method and simply returns an empty array of cookies without actually parsing the header +I have committed fix for +As a note to LucianoYou cannot simply remove the module tag +I have appended some codes to judge it already. +Will try to build test and deploy a snapshot later today +Now works for statically typed services +Thanks Robert +So lets talk about it next week +The issue is there because old Feature files could not be read correctly +Sorry I meant the patch I plan on submitting that is currently on github does not have at the and will apply cleanly to trunk +not sure how this would happen but i suspect it might be a configuration error +Thanks for the patch Clint. +Colt looks promising and its under the Cern license which is very permissive +Maybe we should do the same for the WFS one and name it +After all the software does have some input data validation in the code or does everything just blindly fall to an outer exception handlerThe error response delivered is the equivalent to a compiler saying Found error in source fileFactually correct but of limited utilityHow about a message that says something likeInput data error at line xxx of input file yyy +I verified that it properly errors out in both the RM launching AM and AM launching containers cases +Hoss sounds good Ill work this into the next patchStephen sure go ahead and post the changes +Cant you just set the strict flag in the reader to false +Ill need to think about it. +Additionally itd be nice to have some progress report during plugin installation +Say if it has to be set to mins then it is not possible to do so +as far as i saw is rather about supported orderable child nodes in generalwhile this specific one is about the hardcoded string which i didnt want tohard code when fiddling around with the permissions +Feel free to switch this back if you prefer +There was a couple of errors in the message logging part which was detected by the annotation processor +IMHO what needs to be decided upon is what is more important for Tika Server for it supporting all the possible host variations out of the box or expect the users do more work when the server is accessed remotely +Thanks Rafael +changes missed in that last patch new patch to fix +The xml is hidden in the Data Elements section but it is still shown in the Data Types section. +Thank you Arpit +This is a blocker +Can you make sure you have cleaned all of the target directories and retryJim +with Column Families containing INTIGER names where integer named column families cannot be listed due to Syntax error withing Cassandra ver +net client +Thanks for the fix Chuan +So if no other jaxws tests failThanks Jarek for the contribution. +Build log +Nope I havent started working on it so feel free to pick it up +Jira ticket is too wide +we found closing writer is quite slowAny reason why it was slow +As far as I remember someone started translating site info into Chinese and it was decided to put on the site somewhereAs I said I didnt know about Anakias support for +i might need to make a test case that doesnt rely on hdfs testing test to not used tests would hang since threads in executor were not daemon threads +Added performane test for import and clone procedures into performance test suite for the case when repository contains big amount of nodes +It seems in production this works as expected +I updated Robins patch after applying +Please review and if looks good please commit +Hi guys I download the nightly build +Bulk move of issues into after vote that we merge branch into TRUNK up on list +Now it would reload exactly one time per requestPlease let me know if it doesnt work +I forgot to mark patch available flag +Final patch that adds a new test to the CTF suite. +The rationale behind this decision probably was HTTP spec complianceHTTP CONNECT method was initially defined in HTTP specification onlyHowever there are enough HTTP proxies which implement certain HTTPfeatures including SSL tunneling support +But after I talked to Koji it looks that it is hard to an acurate distance between nodes +Backported to together with with revision Resolving issue. +Im proposing the global rrc and a set of datacenters whose cordinators will apply that chance +i still cant reproduce it but i have guarded the code so it should not happen +The server monitoring tasks will now be started from a separate thread instead of the Activator +Added references to the duplicate issue and to where support for rendering mode was added. +I also have the same problem on my project +Thanks Matteo for taking a look +I updated the wiki too +bq +Our JDBC implementation contains some methods which were introduced by JDBCNevertheless those implementations dont claim to fully implement the higher JDBC rev levels from which they borrow methodsJDBC is an interesting special case +Or should we prevent the execution that contains a task to be signalled directly and throw an error +if we cannot Read or write to a KS why allow creating itBecause thats how you add them for new with an rf of zero until you have the nodes in place +Still investigating +Do we have consensus on this +With cobertura report +Generically helping applications find a text file that they need to load in their application is really up to the developer +testAutomaticallynew persistent class jpa name +Created from Preview on Mac OSX everything is fine +Committed revision +No response from raiser +If the locale is null it will use the default locale for the machine +would it be reasonable to assert spacenull in the case where name contains a +I reviewed the latest patch it looks good +Because we check with is it is even more a high probability that we have a programming error if this exceptions happens +Marking this a blocker +In my experience and recent issues +leaving this open in case there are any issues that arise +allows to rewrite norms +I feel that we should use jBPM for this +It doesnt even close the that it allocates +Patches for applying the patchesAre we ready to close this one? from me + +Thank you very much! committed this +Closing due to persistent lack of feedback from the reporter +Done +should be bundled with in the future but it would not happen soon +HelloWas this fixedRegards +Verified +Another solution which keeps the property declarations is to change the synthesize to dynamicIve attached a patch for that here +Its completely ignored +Yes right +Something most likely is wrong with my version on centos +I think adding a note to Known Issues along with this workaround would cover it +HarishThis is really helpful thanks +Sorry about that + affects this too +On Vista it works fine +Fix is now in CVS +Committed to trunk. +out file +Non finalized upgrade is reported by dfsadmin report dfsadmin upgradeProgress and web UI prints the upgrade status +Just add the source to the source lookup which open source code project are you referring too +Shazrongreat thank you! The patch solved the problem +I have worked on this but not yet been able to find the correct dont believe that plugininstall should be changed to do this immediately because it may break behaviour expected by have someone working with geronimo at the moment that may be able to help with this +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content is still an issue in jboss eap thanks +Test case attached the eclipse project of the test case as our tests the application only works with conection pools for each partition +In the method we construct a object for the prediction the goldreference one already existsA simple equals should now be able to figure out if there is a difference between the two if so it must be because of a mistake and then we call the listener interfaces misclassified method +Verified that the now looks up the cause rootCause of the exception and are handled by the formatter when the stack trace is formatted. +Closing this as per Mladens comment +It disappeared like magic +The isnt going to work in the classpath of most projects since it has all of the dependencies inside of itThe in your classpath is not from avro and may not be compatible with Avros slfj +Are you able to work on them insteadcould you please take a look at this issuethanksdims +Looks good +I sent you an example offlist +I tested this on Java with no problems as long as xerces was present +The patch also includes interface and a basic implementation for Orc +Keeping this desire implicit seems to be a recipe for disaster especially since the specification defines different priority rules for the framework resolver +I dont see any particular for this bug the fix went in referring to +Preliminary patch +Fantastic thanks PavolI will look into your changes and accept the pull requests +Patch looks good to me +Actually that doesnt fix it. +derbyall already completed without failures +Ive attached second patch that returns a zero if input datum is null +I committed this. +Depending on your terminal width the lines willstill wrap but those seemed like reasonable places to break the linesPlease have a look at and tell me what you think +Previous attachment in diff has been updated for and also ajdt version +Issue is verified fixed in +No failure +The biggest change is the regrouping of the getQueries to better reflect the way they are going to be implemented +Moving out of +I just jot it down for my better understandingWhen client issues a scan against a table Hbase generates objects one per region to serve the scan request +Sorry I did a lot of file renaming when creating the first patch must have messed things up +Wdyt about naming it Propagate? Because thats what it does basically +is now merged to master linking the Java ans Spring API docs +It already does this no? You can use dynamic methods query the db etc? Maybe all we should add an implicity db connection object for doing direct SQL but thats about maybe only the documentation is looked for respective objects in the binding but couldnt find any +No more work currently planned for the feature +Changing description to focus on the aspect of the thread dump thats significantly different from +The updated DS support is checked into head. +Test cases are working fine for quite some time now +Fixed on and trunk +The problem ishappening because we are directly assigning to the member variable before thearray has been initialised +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +That looks correct +Also would it be possible for you to attach your code changes as a patch against trunk? Which I guess is what you are using +Agree Glen anyway Karaf will be the latest branch we will move to so its not a big deal +please install this in applications +Please close it if you agree. +What I am looking for here is that I strongly believe that API classes should be isolated in their own package hierarchy in order to provide a clear visual and logical separation of the implementation classes +Cancelling the patch +Just for information manually modifying event dates also breaks the new Timeline widget introduced in Sonar +Is it possible to fix it also in +I think we should show all the console plugins that have started +It should be fixed now +Patch applied in revision . +defer all issues to. +Patch generated against svn fro the contribution stefano +I agree with Tim +app that reproduces the you for attaching a sampleNot sure if its because of upgrade or what but I was not able to reproduce this issue +So the total size for the first level CRC is about of the data sizeHow about we compute a second level of checksum over the first level? So for every first level CRC we compute a + push to +Resolved as Wont Fix as explained above. +A similar problem exists for +If you finish it before we release beta please correct the versionThanks +same patch file in hopes that the will run +attaching patch from the top level of the source tree +We could push this into instead +trunk Managed tracer and scheduled polling consumer +I am not sure but it may be that is related to the patch provided above +Meant to attach the pdf instead of the +Ignore my previous comment +lets get the code changes committed and then debate the approach on the deprecated warnings +It looks like project import has failed when you have parent pom specified +Feel free to change itThanks +Adding screenshot of Violation drill down on test system +There is also a resolved bug that minmaxLabels was always overridden with the default value even if given at command line +Todds fix changes significantly but its for the better and it looks like it works +I dont understand what the problem is +Approved for +The new import looks good +diff produced from root of checking out current +Doubt this is AS +fix coverage patch for branch patch patch for trunk and branch +fixed in r in case of missing file attribute a comprehensive error message is now displayed instead of obscure for your report +If I enable rpc metrics it shows for each rpc invocation openscanner get commit etc +Fixed Solr and Orm modules from the patched hibernate search and merged into the master branch. +Can we have this in trunk and tested before that +bq +I just committed this +Testcase is also modifiedAttaching patch after incorporating code changes based on Devarajs Amars comments +we should define a way how we deal with such extensions +The ovjectClass contains the AT in its listversion dn ouobjectClassescnapachemetaouschema metaSyntax meta definition of the Syntax objectobjectclass topobjectclass metaTopobjectclass meta FALSE metaTop STRUCTURAL creatorsname uidadminousystem +Thanks +Ok I committed a patch to fix the active start level initialization bug +For deeper investigation in +Thanks +What I dont understand is it doesnt seem to happen in the usual Seam test suite +This is required to be able to conform to the single namespace for all servlets requirement of the Http Service SpecIn addition just for completeness I reformatted the code according to the Felix coding conventions and cleaned up all import listsDoes anyone mind if I would apply this patch +I would prefer to have a separate set of tests the compatibility tests that are in different packages external only exist not to break the programs functions but test its compatibility directly +There should be no EMF in an EAR +Bob any chance of fixing thisId like to open up the cargo web site to more contributors but Id rather we have all confluence diffs sent to the dev list before I do so +I propose to close this off as wont fix +after a brainstorming IM session with Juergen we concluded that one solution would be add a lifecycle interface which would startstop a component +The test functionalorgapacheharmonytestfuncapijavaxnamingproviderrmi does not require any servers +It is in the initial startup for checking the valid DAG that are on the list of AM +as the following link would need to be changed to reflect thisEither way you need to update to document the new filter of multiple states +Whats relevant is that the user has a token not who vouched for the token +v Advanced java code inside +I removed the DB queries from the sample and I still seethe performance differences using my java code BUT if I run the same testsusing the DIAG option from the command line Process the execution times arevery close +committed as r +The fix will be available starting from mvel +I started working on a patch but found the existing behavior to be more strange than I originally thoughtthere are some bugs in the existing behavior as well completely separate but along the same lines of this issueSo here are some tests let me know what you think +The documentation is thus far except for in the WFS spec +Ill have a look soon +Applied. +Also make shure that plugins ui are installed +I have noticed this failure since we committed +You used New Java Project Wizard which is a bit different than Importing existing project in any case both should havent had problems from eclipse but I have also just used the ant tasks to do most of the things and otherwise just used different launch configs +Whereas in the case of metadata which is persistent across version upgrades backward compatibility is a major issue +Will do a closer soon +fixed completed and passed +Patch moves most of the information in the SCA SDO and SCAsamples INSTALL files to the respective file +Attaching an ij script based on the description of the issue that demonstrates this +But we should test that important errors are handled properly by returning an error code and rolling changes backThis is already the case The aborts on error and rolls back unsaved changes +Changing it to use Felix is fine by me +It would cause the most problems with renaming a package some classes might need to be refactored as well +The build now passes even with is still an issue with validator which affects showcase +for precommit test. +DaveJust curious if you had an early patch for thisAny chance you could attach that DIFF to the JIRA? Technically we cannot copy it from github + +Tomcat works ok +Must be able to open the project and have the table of iterations display the current one +Are people alright with having limits instead of limits? Pools allow you to group multiple jobs under the same limit +unfortunately that is not valid groovy +on the OS X thanks Allen! From everything Ive read Java Prefs and usrlibexecjavahome are still present in Lion +Fixed in Committed revision I will provide a detailed explanation shortly. +pending Jenkins +But if you need to support and your system encoding is not compatible you need to specify both +Committed with revision +That is the way that I am leaning too at the moment +Has this been resolved? The upstream ticket has been closed as WONTFIX +Attached initial version of the which based on the one in Solr +Issue issue +I added some general details about Tokens to the top of the docSorry if Im missing something but wheres this doc? Could you perhaps provide a link +Hmm +Whatever is decided in the end it should involve consistency +Right +Id rather have a small amount of code duplication than have validate do something unexpected like that +Sending Transmitting file dataCommitted revision . +Note that this requires the patch from to work +Found the source of the bug we were discussing the v patch is correct but will submit another patch with the correct way to do things + is still not fixed and its Fix Version is set to TBD EAP so I believe should definitely not be listed as a fixed issue and should either be in Known Issues or perhaps not mentioned in the release notes at all if you think for whatever reason it doesnt belong there +The WSRP content has been integrated into the EPP Refernece this JIRA if issues or oversights are found in the published drafts. +Can you verify the tests work on the test box +Perhaps we can put the warning in for the release. +The encryption happens after the compression in my current patch +I dont see much point in having both stop and join +Applied the rd patch thanks +If represented on the property as it is now +I dont think that we ever covered this case +I wouldnt do that there may well be other unrelated changes in those the jar by running ant build in the source folder of Seam +Just committed the patch fixing this issue as resolved +Applied against trunkMost updates were in the is failing but it is in trunk as well +Pure masterslave removed in upcoming +is owned by the commons group and you can deploy there freely +Other pages could be more dynamic for example hidingshowing different controls depending on certain state +WTF! So you realize you committed some code that just randomly swallows exceptions? Have you gone totally crazy Norm +Only is left pending branch unfreeze after release +Closing as this has been applied correctly in the java branch. +Will update after testing with thatThanks the original problem I am getting still stands with +Instead of initializing VALIDVALUES with the constant values I used the constant names +Removing older field DFSHTTPSPORTKEY +which several eclipse feature definitions are you thinking about here + +for the patch +I cant think of anything that would have changed between the two releases to account for this +Appreciate a quick review there You can develop your patch over +Thanks Ashutosh! +Im sorry I mixed up Range with I just corrected the description and will provide you with source code examples soon +Who knows +fray v +Thanks for the patch +See +Could you provide a simple java project and explain me how to reproduce your issue locally? Thanks +Thank you Bilung for pointing that out +Now both host names work thank which of the two host names is preferable and should be described on the JANINO homepage +I did first changes +tmpcargo is now owned by maestroYou should find out why that happened because itll probably just happen again even if I change the permissions back +Ill report back with my findings and what the configuration needs to be to support developing a Surf project in this fashion +Patch to rotate available credit across all active links fairlyDoes not fix the case were the of receivers available credit still causes a stall +This issue is now resolved +SSH now does bulk assign since a recent change once its figured regions that need assigning +I tested the setProject fix and will submit a patch ASAP +Ill add the error code for device for run anyway +This bug is not fixed yet as the xmlformat rootNode is not consumed and eventually fails in with a bug is fixed. +AndrewIf you also adjust the to set that the node is not expanded then youll be fine +The WSDL generator when its job is done publishes generated WSDL to the docs hash table and the rest of processing is the same as in the usecase below +The reason it works this way is started life as part of the Struts web application framework populating bean properties from request parameters so changing this behaviour would need to be in a major seems to be related to +Verified that the alt image descriptions are now updated along with the images in run trees when nodes are expandedcontracted. +Initially this was a fairly simple immutable object which just held references but it is growing to be more complex now with support and support +Added subsection Write timeouts to Remoting Guide on branches and. +pending successful precommit +Cool lets call it fixed then +Therefore I made some changes to and +yeah I guess +Ive just uploaded same version as is available on review boardPlease accept my apology for the inconvenience and thank you for noticing thisJarcec +Still not a bug +I know that there are many discussions how to persist a message in Sandesha +I am hoping it wont be too much of an effort for you as last time to merge this in +I would improve it first +Please prepare a patch against the current trunk +Maybe this is not a Bean Archive at all or you forgot to add it +Sounds like an option people might want to have and would be nice to include in the std distribution +Ok +See comments in as to why partial state transfer makes no to see this in if you manage to get it ready in time forIf not it needs to be in the latest in order to go into AS +Its technically possible but does not concur with the CMIS spec. +Mock datasource and connection may help I have this issue with +Ive been thinking along the same lines that might help us out here +fixed in trunk +was added when folks wanted to distribute stuff besides the job jar but merging the two might be good. +added check in rit might be interesting to set the length to the number of items in the property but I didnt do that. +This patch adds a parameter to the and allows you to switch back to theThe previous test runs successfully when is enabled +If you can fix this and upload another patch then I can help get this in +Thanks +I have the same problem as Kurt +Thanks applied +Marking as fixed. +figure out how to be smarter about determining whether a metadata file contains relevant info before merging +and have been updated on to point to the reposGithub which mirrors the copies will update shortly. +Ill add a test and commit then backport to and +Will open new issue if other distributed log splitting bug is discovered. +Ive committed this to branch and trunk +Seems to be in combination with import scope +The AMRM protocol is not a delta protocol but an absolute one +If its a feature you really want to see please submit a patch and well consider adding it to the codebase. +This is actually a critical issue +Is there any real demand for this issue? If so please let me know +Closing as Wont Fix. +We would be fine if you wanted to rename jira to issues I dont believe it has been used yet +reopen to upgrade to last version look done +However I agree that it could be confusing +Lance can you update this patch and add a unit test +Then depending of the time of disconnection two things happened +Items are not removed from disk they are merely removed from the key map +There are a bunch of variants +May need to alter mapping xml file path to run +But it seems thatgenerated code is not correct +i think by adding this property into s junit target this would work +why not just leave it there? we will need it for syncing w SpringJ is currently empty so no need to exclude it just to it in the next release +Thats a good suggestion Ive added that test in. +Uploading a new version of this patch which addresses Knuts nitsRegressions were run on the previous version with no errors +done +tests passed successfully with similar testing procedure to +Ive also found under tomcat alpha with and the latest CVS build that you cannot deploy Tomcat without putting in the classes directory +No application context activeI had the same question when I call seam web service in my android projectbut I found the reason is that two different seam version and +Ill keep checking and submit a request if it comes back up +Here is a patch that switches to using gzip f +I added an explicit test to make sure the file is created +looks good to me +I found related issue which seems to be resolved for the next release +bq +I am not sure there is anything we can do about this as it seems to be a Xalan issue +Hi Kris The filter does not support CORS +This was produced on Burbank +So the path in webResources is really related to not current dir +Sounds great +Join execution order is up to the query plan optimizer +Verified thread count +Not that I know of +Filed to track that problem +is there any need for the instance variables statusLine and version to be mutable? Can they be made privateI know fixing this is not strictly necessary but then again is it necessary for the fields to be mutable and not privateIts a lot better to start with all fields private and final if possible and only increase the visibility if strictly necessary +Awesome +Just started using Arquillian a week ago and after getting the framework setup and my first tests written the next thing I did was create a superclass with the Deployment method +Bulk close for Solr +I can do that +Wido Do you mind looking at this quickly? Its probably a trivial thing for you and its possible that Im the one that messed things up in the first place +Thanks again +There is different documentation for which youll see on the Edit Inventory Properties page for a server type which comes from pluginsapacheetc thanks + +Done all accts created log in with key +Use search +It looks to me like couchviewdoresetindexes takes care of +I can see its EAP but please point out which build by setting it +StefanThank you very much for your help! I tried to apply your patch to my local copy but it doesnt seem to be a valid patch file +FYI The CP no longer contains the versions in and as they have been removed as part of only version in the CP is now in the and is in revision of builderseap buildersesb message Dont replace ESB versions of MVEL +When we create the +The since and until values are passed as is to Facebook REST calls which will convert and validate on server side at call time +Merge from Dains work on Geronimo classloaders +This code wont do anything as there is no real default value for the compression +That said there are only a small number of standard mutating methods that return a value and hence even could be used in a filterSo I think we are safe in dropping this test and we can debate the value of a new test later on if someone wants to raise the issue again +At first sight it looks like this problem may be caused by +The only question I would have though is would we publish different jars for every version of hadoopYep +Ill wait until someone else chimes in with any review comments to make a new patch +Committed in rev +settings +Ill test on Java +It should also include some way to include a empty set result if the filter has filtered everythingCan this be rolled into the pb work? +Its been manually tested with large customer files that are currently failing to copy +the part the we somehow need to recycle is the configuration of the jboss server so that the suite package can be resolved in +I assume this is present in trunk as well +This doesnt seem quite right on trunk either so I think its still broken +Instead we can get rid of heavy weight objects like the and the Container itself while copying relevant information over to the +This issue was migrated to +Regards seems like a good choice you made Now things are more clear to meRegardsLars +be referenced somewhere +Herve the plugin xdoc generator is no longer used +Can you review +Everything looks fineThanks Gregory all. +Could you please create a new issue to track this? Thanks alot. +I have the fix in I can add the deprecation +Its possible by preparing an +Well I have exactly the oppositeI am trying to run the of the EAR plugin with and it keeps downloading version of the plugin even though I added the snapshot repo in the settings of the IT +Please reopen the issue if you can find a valid reason why this kind of order is needed in the meanwhile Im closing the issue since the order actually used for the eclipse classpath is the same order used by maven during compilation and IMHO this is definitively the correct behaviour +should we close this handle new improvements in a separate jira +Reproduction comes with its own local web server for debugging web projects +Thanks to AllThe last patch is in trunk I just moved the string into to not show a when there is no information +If the worst comes to the worst and for whatever reason it is necessary to insert a into my code against my wishes it would be less terrible if I could move the cursor after the using the Tab key +We do not want to force splashscreen changes etc +I just committed this +This has been working fine for more than two years +The test I ran was testlibhdfsthreadedI also ran testfusedfs and verified that it passed successfully +The option XX turns off the old verifier +Many thanks icons used for the tree and for the tree updated CristiThanks for the resubmitting the patch its looking good but unfortunately in the meantime some of the classes you modified got changed by someone else so i get conflicts now quite some +Both the initial incubation website and the SVN has been created and code has been submitted +This was fixed in the nightly build +We probably dont need the second statement +If you notice anything that doesnt seem right now please reopen. +We might name the abstract classes andWe might rename the current implementations of and to be andThe and will function like Thrifts Protocol interface +Similarly for ve added a create method and also some other methods which are responsible for executing the individual queries for loading the userinformation and authorities +Fixing the v patch which was corrupt +can also be closed +Why is this issues prio minor +Verified in JBDS H. +That makes sense +CraigYes I agree +Not yet I picked it up from Jesse and working thru it and committing incrementally +Wasnt able to reproduce and since Carlo confirmed that he cant either in his setup Im resolving this. +raise the jira in the SOAP mailing list +Revision +Close it. +Patch submitted by Christiaan ten Klooster +r. +Ive also run a selection of the failed tests on my local box and they all passedIve just kicked Jenkins to +Ill run the rest later but I just wanted to get this up here for discussion +I am looking for comments from the core developer team regarding the possibility of +I wasnt suggesting that we should take the map out of the abstract memory cache +Committed revision +With version Ive added an implementation of the crypt filter dictionary +Here is the patch with a unit test that demonstrates the problem +Hey Andrzej do you know what rev this was fixed in trunk? Id be happy to backport it +So all issues not directly assigned to an AS release are being closed +Proper way is to set boolean flag of in the event of KILL synchronously +I think you lost me where you jumped into implementation strategy +Trunk fix +Sounds good +New Dynamic WEB projectFields Create Test Project Test project name Test package name EJB project name EAR project name are disabled +maybe we could add this to the docs or in a code commentpatch looks good to me +Dear Alex +Working on a fix +fo does not show any difference in output an debugger or print statements is needed +This issue is fixed in Rampart and now Rampart integration test has tests for negative scenarios to make sure this is not broken again +You can follow existing tests in testeepigtests to get an idea how to write an testYes this patch is logical plan only +no default is false this setting was added with ACF a the previous behavior was just ignoring invalid reason to open the ticket could be solved with the right configuration and problem in last comment cannot be reproduced please feel free to change attached testcase in a way that it shows the problem and open a new ticket for it. +I thought it was just the transition text positioning that was not saving correctly +Btw Nicholas it would be nice if you could try out and see if it works as cannot understand what the problem is with these images +Reproduced on JBDS and verified on JBDSBuild +did improve things a lot now the cluster only gets blocked when a node stops +I constantly keep getting this test to fail both from command line as well as from Eclipse so I dont think its an environment setupDoes the test pass in the CI environment +I will try to test on linux with JBDSThanks for your upload +Simple fix + +You can also raise issues like this on the users forumThanks a lot! +Does updating with Admin API provides any kind of indication that this is successful +The new patch uses the checkpointInterval property to schedule the for the patch Hari! Committed to trunk and! +Great +when it was built againstlibstdc +The reason I didnt modify the tests is that we want to catch in all tests therefore the net result would be equivalent +Added a new patch without the new cache just using the existing jar file cacheIt performs as good as the previous patches but is simpler +Its cool but my student fellows and I think it needs a better documentation +You can look at the produced which includes the accurate package info for each jar +There is nothing for us to fix in the code and the docs already covers how touse newer versions of Xalan with the and. +Better documentation about that is too broad i need something more specific +Hi Jingsi Zhu is no longer at Facebook so this email address is no longer being monitored +Thanks Omkar! +Invalid preference page path Speed TracerInvalid preference page path Web Toolkitstays on problem here is that our GWT plugin references preference categories defined in GWT plugins but doesnt declare them as dependencies in because category references are not java dependencies but just a ID used in our +That would help try to resolve this +Addressed Stacks comments +That would be fine +Loan Service Quickstart with additional Bean Service like a bug the BPEL engine is not passing the content through +Our client code shouldnt have to set any arguably internal configuration variables to avoid cryptic failures +Initial contribution of +It is not norms because the test does not test norms +Thanks for fixing both mapred and mapreduce +Sam please answer something +The new and are equal in this regard +bq +I dont believe jar is at fault for this bug +So I bumped up its version number +So far return velocityCount and velocity are gone +But as mentionedearlier only label showsThanks Jag +We missed a part of the patch to fix the jsps +The contrib section of Lucene contains an query parser which aims to provide of Lucene queriesThanks for the reminder +Thanks for the reply Paul! Could you elaborate more on the approach you mentioned? You mean something likeString dataPath +Provided patch prevents goal from puting OSGI jars into extra cluster +Since catalog interfaces are not going to be included in next release uncommit from +I keep not being able of reproducing this error so Im closing the issue +Damn this is a regression +It turns out the the problem is pretty typical for the programs in the +Hopefully I can post comments sometime later today or tomorrow +This means the codec used must be maintained with the message set +Could you see my nd patchThank you +Ive created a xircles account username is hertzsprung +Details in +I thought it was on AM +This is a JDK convention +lgtm +None of the core devs are going to allocate time to fix these issues +However I would have said the change were required for both requests and responses +the svn location +Here is a patch to change back to usec precision +The way to get around it is to configure your HDFS cluster to retry maybe times and with timeout of second +I will gladly let Suresh rebase that patch once this is in +FYI We are also using javassist version +Resolved by. +TPTPno compile errorsBIRT following interfaces have new methods have fixed the issue by adding default methods and committed the changes to the wont break Transformsno compile errorsPortletsno compile errorsProject ExamplesI used the class to fix +Hope to see you out there. +May be I have some wrong conf +Reopening for EWP +PR to JDF merged +bulk defer of unresolved bugs to +on not backporting +Please review +My apology +I noticed that the directory that the IDE is having problems with contains several subdirectories +My bad +added two more tests for testing appending to an file and permission checkingHowever currently is failed even for trunk +There was a bug in Tortoise SVN when creating a patch where duplicate lines of code appears +Actually is provided just as a sanity check for Activity though I dont think it is mentioned there +I can make a separate patch which includes those changes once we decide if mkdirs should pass along the or only return a boolean +for examplewe have nodes cluster +In this way the zk conf is set to previous ledger path and then the start ledger id will continue from last max ledger id +Im just arguing that it doesnt affect typical applications that much +Please review +When it is up the state is RUNNING +Call the option steal or force or something +Cause The value Wed Jul assigned to a date property is not in the specified format of yyyyMMddAfter changing to the proper format code compilesVerified with version M +According to the comment Sunitha is having a look at the codetest +I wont be touching this stuff again no worries +We dont reopen old after a release has been published and the developers may not notice or remember comments on resolved +Fixed on masterbf +Should be passing on Mac OSX +Slava is it our issue or WTPs +I now understand what each is for but I think they could be explained more clearly + issue is closed +Apparently richard has now fixed as of Friday +I guess this is how its supposed to beThanks a lot JoesCheers +Sorry my xwork local copy seems to be corrupted and absolutely out of it should be fixed +very good Charlie thank you very much for all those years of making the best Ruby environment! +commit Checking exitWorker in the loop tests fail err skip time commit Checking exitWorker in the loop tests fail err skip time failed commit Checking exitWorker in the loop tests fail err skip time commit Checking exitWorker in the loop tests fail err skip time failed +trunk Applied patch with thanks to Stan LewisStan do you mind updating the wiki page as well? Then we can close this ticket +I should also note that this infinite recursion exception appears on my installation not the setup I did not go through the effort of creating and wiring a custom for the version +Its a question of using a router that times out idle connections no need to inspect the trafficBut yes we will probably just have to wait +Yes issue was fixed by clearing the cookie +In the case of a split you can fail the job disable splitting or have some way of getting the children laterEven though you would have kindof data throughout a time period you would at least have a timestamp of when that backup was made +It worked for me when I made the exten +I found the solution +I am closing this as not a Camel issues +There are several dependencies which dont exist in the maven repository andor have the wrong groupartifact ids +Sure lets fix the reference count on the next patch +Fixed in revision +ooo really +I remember reading something about filename length restrictions for certain versions of tar +Should it be my project base directory with src at the end +Marc any chance you could contribute a wiki page about running with Jchains +GA now obviously fixed this + +patch provided + +Still getting a as well will keep checking back +Thanks for the new patch +Committed trunk revision Committed revision committed another improvement to use in a finally block to really cleanup the mergePolicy and mergeScheduler no matter what happens trunk r x r +the flushes are still serialized the compactions are done elsewhere nothing changes as far as I can tell +Just to be clear the bug exists in +I am marking this as fixed since the patch has been committed +Committed +looks good +The fix has been applied to branch Its possible that a recent failure of was caused by this race +its possible to do it right now also just subclass and delegate to super only if your auditing conditions are met +If M does it RS could have something opened zk would say OPENED but if the master dies it may or may not have edited META +Interesting I guess further analysis needs to happen for this to be correctI will try a few things to take a look at this +uses for logging +committed the patch to master. +Not sure if its a good idea yet just an idea +Issue may already be resolved may be invalid or may never be fixed +Just changing those to the gradlew equivalents would be big improval +previous title was document how to enable variable history +I just committed this +Suffice to say this could be a bit clearer in the migration doco guys +Weather is terrible and quality of service is irritating at bestSincerelyGeorge Titan +Updated patch with changes based on Todds commentsI left the fsync metric for now but we can consider reporting the time in nanoseconds in the future +All I see is ds If the method is not found or the found method is not declared protected or public SCR must log an error message with the Log Service if present and ignore the methodds If no suitable method is located SCR must log an error message with the Log Service if present and there will be no bind updated or unbind notificationAssuming that the cts is correct I would think we could detect this situation in init and put the component into a state where we wouldnt try to do anything with it +will reopen on April st if needed. +Since this issue pertains to Adobe Flex you need to work with Adobe support to get your issue addressed +here they attached all entities +where it will use the field named id as the id fieldIt seems there are two easish improvements we might do either allow setting the default id field with a setter or look into using the new schema rest api to actually request the id field from SolrIm still not sold on the need to expose a new update methodThats just a quick response though Ill review your patch as soon as I get a chance +Just so there are no misunderstandings about the intent of this changeThis is something I started working on a long time ago +Excellent +Another iteration +has been committed +I ran another installation and the server is startingstoppingrestarting nicely now +Moving them to a separate package heals the tests +This test uses overloaded resolved issues for already released versions +I tested on a table having K partitions +testing the various failure cases like this one is a bigger project than fixing this bug +FTR I attached a copy of the referenced document +Attached patch which contains few test cases to understand the issue +Thanks David! +For the record I am not seeing this error in the regression test run for the branch +Perfect +is for patch specific to pro split code +One is instantiated for each block pool which maintains the state of the scan blockMap and block similar to what had before +The cost of copying a configuration might be too big so the patch takes the first option +There are no settings anymore +patch looks good +part of partcqlshKeyspace select +Bulk Close for all issues before incubation +Sorry for the hassle +The issue is when you stop and start the VM it can move to a different host without any restrictions +It is unlikely that branch has been covered by another test case +Fixed by which involved an upgrade to. +Otherwise I may need a pointer or two as this is all new to me +Look for an alternative to avoid reinjecting a list of merged +fixed in after release +Added as a Known Issue to the SOA Release NotesThe Native package for Enterprise Application Platform is incompatible with the version of Web used in Enterprise SOA Platform +I wasnt aware of this +Finally looks good +Committed fixes in r +Depending if OS was windows setup the initial assembly path as well as relativepath with correct slash format +actually doesnt care at all about the extension of the file +I never tried your soloution + + +Ashutosh Ill take a look later tonight +I just committed this. +Since I have not heard from Matt let me assign this to me +a correction for the test +sent pull think you may have misinterpreted my question +I added a patch similar to the proposal in revision Thanks for the contribution! +Hi AndreaPlease review my changes in +Can this issue be assigned to me +But I feels the logic for setting the endpoint name should be gone within the endpoint factory +I should add that there is currently no separate module +I just committed this +Bulk close for +create an empty in libiclient directory of my JAVAHOME or JAVAHOMEjre directory didnt work +Karthik +The issue has been resolved +Hi WolfWhats the status in this? Will you be able to submit a patch this week? If not please assign this to +Danny could you confirm whether these are all the changes that are necessary +A priority? NoFell free to create a patch +I think that is enough to tell the loaded modules are loaded +What I committed to to branch and trunk +Added thanks. +Which is odd cause the building of the process engine is actually completely synchronous +we would need the server log andor browser javascript log to proceed +This is being handled by merging across from the CP branch apologies to Tom +Fixed the issue component target version fix. +Reworked the provided committed a fix which should solve your issue basically it makes enqueue all events except SESSIONCREATED and prevents the deoredering of writes +though maybe the flush in the test should be explicit since its not immediately obvious it will occur after ops +Ive moved the openjpa dependencies to the plugin and added some of the transitive dependencies as direct dependencies of the openjpa all jar +I created the patch for you plus made some formatting changes to minimize the is a new patch that does the same thing by adding a method in +fixes the markup so that it passes DTD to TRUNKThank you Mathias! +In my experience the in transformer is always called just before a Serializer +But when you set wtjson and have a field that is JSON string it escapes everything +Applied patch to the branch revision thank you. +Is this ok or by default we want to implement security +It still needs unit tests but folks can look at it and see if its the right things to do +get the zipped WC test suite file to downloads +It allows ant to be called with pathtoanotherjre and have the tests run with that jreIt makes no attempt to avoid running tests so it will give errors +I imagine that many of the postgis users probably use shppgsql to set up their dbs at first and Im pretty sure that the latest release of shppgsql declares a serial primary key that automatically generates the fid +Thanks +Not a problem +Try writing a program to do the scan with Java +Hi Patrick the patch is ready for review sorry for response so late +Do you think this needs to be in Nick +The patch reorganizes the code in Redirect by moving some code into some helper methods +Layout looks good will bring up the staging site momentarily +A first start is in rev +attaching the jstack on broker +contains additional tests that are not captured in this diff +Sorry I didnt see the BOM +I believe the problem is caused by your system not +KiranCan you attach the latest mgmt server log files +Cant we just have a namespace called simple system and have tables named and +bq +Renaming JIRA to reflect the actual scope of this issue in the sync issue with lease recovery failing on blocks that are already finalized +Thanks to Jason for the contribution. +Please let me know the path and file name in Jboss +Marcel thanks for the fixes but qomjsr still seem to have dependencies + +Hi Dave! Probably the patch is broken my crashes on that patch +Still dont understand the bulk read API but given that I have indexes with frq files of GB Id like to use it +Hi AntonioYou are right we use an artifact repository as a mirrorYour suggestion fixed the problemMany thanks to youre welcome +In general looks good +I had done the unit tests but failed to include them in the patch attached is the patch that includes them +Just classifying bugs as Keywords bugs that pertain to specific API calls or advanced usage of Xalan +As an example setting variables like LDLIBRARYPATH to include some shared libraries that user has put in distributed cacheI really wonder why we havent done this yet could be useful for profiling as well +Pierre Raoul could you please patch using Seam as baseline +Ive committed this to after running all tests. +Did you try the SNAPSHOT build or +Aaron Thanks for the suggestionsIll have an updated patch coming soon +Ill take a lookHoward +We need to override how the form is serialized to remove keys for submits that are not the active submit and ensure that the value for the active submit is actually submitted +I was curious about why the test only ran in embedded mode since itdidnt seem to use any features that the client driver doesntsupport +If that resolves the issue for you Ill merge the changes up to the branch +actually the LF is in the output file which causes problems processing it +The property dollar is missingI done another test +Ive had only a small amount of time to investigate it and havent found the cause yet +In that case Ill do the same with +we can close it after applying +If corruption toleration is undesirable the feature could be turned offUNCHECKEDREGIONLENGTH is to skip checking the last MB +It would be nice to add the revision numbers you are working with to these comments +The datanode should be listed in decommissioning listbq +Sure I will extend my interfaces from Closeable +Unfortunately it is not intended to work +patch for fixed this issue +Fixed in trunk r with the second option I posted above +Yep I can +It can not be reproduced. +Updated test to check for it +In Hadoop the hadoop scripts change the current working directory to HADOOPHOME to prevent this kind of confusion +Ive applied your patch into SVN trunkThanks a lotMaarten +It would be good to use Hadoop record IO to define such Metrics record which means disallowing such special charracters in metric names. +I will backport later to releases when will be done +New features? No only bug fixes are backported +Yes i can understand that part +Fixed the commit issue. +Thanks Arun I saw it fixed but missed closing this ticket. +set versions of RF integral Parts in guidesdont forget about about Prototype jQuery and is added to the Basic concepts of the Framework section + +Another ping for info about how to sort this one out. +Do you actually believe that you supplied sufficient information for me to understand what your problem is or what you refer to by This issue need to reopen +this patch should fix this issue of downloading someone review thisGiri +So the returned job objects confFile path pointing to intermediate directory where as the jobs actual confFile is present inside done directory and hence the exception +this issue +Let me kow if you want some other information +Just let me know how to best fix it +To subscribe to the list you can send an empty mail to +This screeenshot shows thatI decided to graph the Journal Sync Average Time +and the suite runs in minutesAdded support for timeoutseconds property in +Andrea Ive added three screenshots explaining what I tried to m having the same issue on OSX connecting to Oracle +Myself and Gao had a review on this +Grabbing this as Ive been thinking through some ideas of how CP implementors could expose their own protocols using protobufs +Attached patch file for fixing Tested under Windows XP but not under nix +Fay I downloaded one of the existing test cases and I will prepare a similar one for my problem +Updated license for demux parser +Filling in the reviewer field +Ill come by your desk on Monday to see whats going on +updated trunkdocumentationwhatsnewvpe Seam PDF is now included in the documentation +Need to fix this and I will provide a new patch later +The new mockups look fantastic! Agreed on the desire to launch the new site with a release +The namespace for the WSDL import is incorrect +This looks ok to me although I dont like the amount of reflection that much +Thanks Mark. +With all the options off loose encoding it makes it much simpler to implement a simple openwire client +There should be an override method FS supports umask +the previous path file was bad because it did not have full path names +Please dont remove remove any lucene spatial capabilities until there is a suitable replacementIts absurd to tell users they should have to use Solr to use spatial this makes no sense +Please note someone should look this over +Alexey please verify +I propose to copy the code rather than depend on the class +I dont like having the Manager etc +I guess the problem was that I tried to use a seemingly generic errormessage AUTHNOTDATABASEOWNER but the message text turned out to bespecific to CREATE SCHEMA. +bq +Will do +this is done by Plexus so might need to be considered when converting to Guice +This appears to have been a temporary problem with Apache servers and has resolved itself. +Committed thanks +Sorry +Verified in JBDS H Linux yeah getResources should be fixed no matter what but JPA and other plugins will still be triggered to load when it actually shouldnt +I just realized there was a bug in my previous patch +Interning is likely to take up some cpu time and its likely to be the case that only certain structs would benefit from interning +mmmkay Im close to giving in +It believe it addresses Knuts comments. +Thinking more about it v runs a risk of duplicating an entire data chunk only if a new chunk is fetched from the queue and before setting the iterator to the data in the newly fetched chunk the iterator is set to null + +Tibco processes the request but then sets the correlation id for the reply to the message id of the request +Will commit when Jenkins says okay +its not the case that the standby NN is somehow incorrectly going into safemode after a failover? If so then I agree with Vinay that we should update the summary and description of this JIRA because its currently rather confusing +The new entries are computed as follows The container searches for all files with a +Great work +The patch is for trunk +Back to the topic of this JIRA +Updated in revision . +bumping +Thanks +SanneYah Im considering adding selection for filesystem type +Removing from. +Has this issue been resolved +We can provide that by having an extra field in +Please verify +The questions I have areIs this a good enhancement to have? Does it not enable us to have cleaner implementation of usecases like? objects are very short livedIf we can avoid the creation of Objects that is definitely a better solution +maybe its hard to readIve changed it to Getting worried +Only filters that one can edit or delete are is one side effect As soon as a Sonar administrator update one of those default filters Projects Treemap or My Favourites looks like this user takes the ownership on this default filter and then the other Sonar administrators dont manage to do anything on this filter +Also any changes to should be done in a branch and tested against the TCK +Merged into the master branch. +This patch provides a method to read server response +Vladimir can you look into this +When looking up component mapslists the component registry gets corrupted making it hold component descriptors for components which blows up like this the next time something else looks up a component maplist for the same role +Currently youre cherry picking and remmapping things which is limits pluggability +Usually I prepare that when new runtime version is available and then I use it over and over +The patch is appliedThank you! +Also affects +David W while it may be irrelevant to this topic its very relevant to general business practices +Do you plan to send a patch for this +Seems like restart time gets longer each restart until finally out of space. +Seems to be duplicated fomr could be the same or it might be in a different section not sure off the top of my head +I think this one is done we have a user dev commits and private mailing list although they of course still have incubator in their domain name +add code you try with Groovy? There have been some fixes recently that might address this problem +Promote this to a blocker +userfplugin causes the undefined error as well so I cant have doc ids that concat strings with slashes either + +Default location of the config file is confAn alternative location might be set through system propertyAlso one can use a standard file to specify all needed probability levels in the modifications new package is renamed to class is renamed to system property prefix is made to be fi instead of rather lengthy default config file for FI framework is added missed Apache license boiler plate is attached to one of the agree with the point it is misleading when a level of probability set as an integerI have fixed the config file comment and code to work with the value as a float between and I think adding a special case to handle is excessive +patch looks good +If this is not the case please let me know and I will goahead and send a email to dev list asking to create cwiki accounts now +osgi annotations javadocs with revision diff mechanism the common diff change with rev +Right Oleg +For example the English locale uses the weekdays S M T instead of the shortWeekdays abbreviation +although they also have new functions transform that doesnt exist in the branch +There needs to be a configurable strategy for resolving conflicts +My fix has been available for over months +Thanks Ram for fingering it +According to your description I would say go ahead +Duplicated issue +exception mentioned above +similar issue old issues +If you that describing your change Ill try your patch against your release note +Maybe a row with a blob that has actual data in it bigger than this number +The patch introduces create mkdir in classTest case also includedAll unit tests look good +Any idea when will be this fixed +Commit at subversion revision +propose wont fix not part of core. +This issue is not as bad as I thought +Sorry for the stupid question +However ultimately we the logic in to avoid the lock strategy of +Per Stack script to fix available in but missing a few lines required to do on DFS fixing +Jay can you please plan what other needs to be done for testing archetypes? +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +The two patches here would seem to have resovled a few of the failures in particular +A fix was committed in the trunk but we need more tests +I cant reproduce this bug +I agree that should be used for vertices without a value +Have you read my entire comment in the bugI have stated clearly thatThis may be because in case of edit ISO we show extractable and featured field as editable to normal user which normal user is not allowed to do and api passes these as parametersIn case of template these fields are shown as non editable hence API passed does not contain isfeatured and isextractable fields I think normal user should be at least allowed to change the name of his own ISO the way we allowed in case of templates +added a quickstart you tried with snapshot +Ive committed this! +SVN revision contains a minor fix for Windows boxes +Then I can add a shim class in to wrap the old version with deprecation warnings supressed +This is a tricky one to deal with +This is the same issue as though I like the idea of a interface +It should work but I cant test it against your database +Attaching the patch discussed in IRCIn summary it is necessary to add a default with every new field independent of whether you are using a union to make the field optional +I dont think that putting on interface is an right way +Patch applied thanks! +It works fine for me +changes the identifier to and removes references to in +Patch for for sure that this patch does not generate additional javadocfindbugs warningsAlso from what I can see the trunk build does not automatically build. + +shouldnt we be using the default umask for the part files? and the owner would be the one who runs the archive command +I guess Ill just add a supplements file for the UnnamedUnknown projects +It would be nice to add the same feature to the concurrent caches although I expect it to be a little more tricky +If there are no other comments orobjections I would like to commit both patches tonight around GMT andcontinue with authentication cleanup on CVS HEAD +Its a limitation in JSR +Updated with a fix for the class diagram image +Manually tested +If form data includes doc interpret it as JSON and treat it the same as a POST update to the document +Attaching testcase so I dont loose it +Will fix it shortly +is this with Maven? I was trying with the +The Sun ORB class references cause problems with both IBMs SDK and Suns SDK so this is not an easy problem to attack +Modified according Zhengs suggestions +I will attach these results soon +FolksI managed to track down the cause of the problem +Thanks +eliminated paragraph on narrowing conversions of value classes as part of getting rid of lingering references to value casts and xcdto +You find it attached +Also post the javascript output that is causing problems with a description of how the browser reacts to the script +i give up on this one +I also feel the need to point out that pulling jsvc out of the distribution should also require pulling other binary components out of the distribution for the exact same reasons +IMO checking whether other is an instance of Foo and acting on that will cause very confusing behavior so that idea can be IMO it would be much more consistent to go through the setter all the time and use property when one really wants to access the instance variable +Let me know if theres any useful information I can add +Linking to because this issue was disclosed by the localization audit +What kind of JVM are you using? Have you installed some custom cryptography provider +Patch available in branch commit fcefabefadffaaebbdb added apache rat as a dependency in lib added a ratrunnable gradle task configured exclusions in +Some comments Did you try to implement the Samplers classs getSample using the super class implementation? So all the added back code is just from? It did looks like so I except the class definitions which implement from mapreduce +Thank you for the feedback +As mentioned above the TCK for CDI depends on thisEven if we close this issue now we will need to deal with this problem sooner or later in order to pass the TCK +In thedebugger I can see that the compiler erroneously inserts thetransitive join closure predicate but I havent been able to forcethe optimizer to end up with a query plan where the bad predicate mattersI was thinking I might try to use the optimizer overrides feature toforce the join plan that I want but I need to go study that feature firstIn the meantime I attached the first version of the repro programin case somebody else can see how to trick it into reproducing the bug +See for one of the affected cases +Is there anything to do here for? Jason are you still around? +If the Add child action of Add menu is disable they wont be added in the Add menu when the popup menu is the selection are changed the Add child action will update so the Java mapping cavas is first painted if uesr in the panel the selection wasnt change Add child action all are disable becaues the selection if null until user on the panel to fire the selection change is a funny thing If we on the Java Bean figure it will be selected on and the selection will be trigger but if we on the panel the selection can be trigger +I have that patch which fixes NPE issue but I still see two testcases failing on the assertion in the same line where NPE used to happen +Patch file with the +The issue is incorporates with so they should fiexed or not fixed in both +from me weeks have passedPlease in advance. +Eriks I guess you will find all what is needed in the last Christians commentThanks to both of you +Uploading cassandra init there be a spec file to go with this? How does this work relate to +What ever we do it should be something that is also pluggable +Attachment has been added with description that demonstrates the issueAt first glance this seems to be due to a rounding issue +I have been running for min and there seems to be a little memory increase on the heap +excluding only frameworkcommonconfigThis would need framework rights so keeping this issue opened +the attributes from tables must be deleted only by dev team is added to on +Sorry i didnt have any comments there +During normal execution will be invoked only when a tracker is lost +Im just worried that the timeout could happen on a slow datanode andor client and then well have more mysterious errors as one thread deletes block files out from under another thread. +With I might end up debugging both ZK and thats why my hesitance to upgrade +It should be OK now +EL expressionevaluation is in mail templates just not in the attributes of the variable value element +What you need by way of review now +Fixed in trunk closing stale resolved issues +As a beginner i think reading them is sufferingCan you add more details on why the documentation seemed confusing not clear or hard to read instead of just saying it needs improvement? Better still you can also post improvements to the documentation +Yes +Some recent work on the eclipse side may allows us to do that +Some other tickets talk about this but I do not have the number off hand. +I agree with you +Fixed in + Username nobody +Assume closed as resolved and released +The latest patch looks fine Kim +This could well be fixed in which is almost done now weve tidied up the deletion stuff a bit fingers crossed +Part of a bulk update all resolutions changed to done please review history to original resolution type +Thanks Dave! +It would let to have platform independent build that could be executed on both platforms windows and linux +Jozef please verify it in CR to into this +I love the poem and great initiative and imagination Plus beer is always is looking at this closer and will comment with more details +Meaner testing will now commence +An option where the analysis would ignore methods fields with an Deprecated annotation wold make sense to me +What date is your snapshot from? or The edit roles feature is fixed in the example app as of might want to open a separate issue for remember me or change the subject of this one +Fix the test I broke +Hi Daniel did you run a source format? A lot of formatting was changed in the commit so its kind of hard to find the actual changed part +This patch effects only fault injection tests +I will commit soonAnd how about +modify to extend only this abstract class and copy over the hashCode equals toString methods from +tests included +this patch adds the check at the server only on create +There is a JIRA about this already I thinkFor these reasons I dont think there is much point trying to patch the existing version +META +A note on keeping a cap on growing logs +Closing this as a dup of +I notice that we are attempting to use the provider all the time which appears to be where the locking happens +If you can provide more information just reopen it. +pushed to with with JBT +Individual measures can be added as new issues +Okay that seems to make sense +So are the patches attatched to this issue still relevant +In this specific case the machine was performing a repair and then restarted +Do we need to them to git manually? Or are you planning to scratch the new git repos and redo them from start this eveningIt will be highly appreciated if you tell us the plan and estimated amount of time for this whole process +It puts the same class in targetclasses as well as in targetinstrumented +If its not can open a bug against it. +So maybe my suggestion to rotate it isnt a good one +youre right + +HiThe above user email id is not valid +Heres a rebased patch for master +Regressions ran ok +the issue with the dialect field was that on mac you actually have to double click to get the combobox to go away again +With extra condition for providing the query plansSince the problem does not appear for you in its probably not due tobecause I believe that was merged back to +Thanks for committing +This other patch seems to work too and avoids the Java call +Part of the point of supplying a WSDL is that you can be more explicit about which address people should use +If you finish it before we release beta please correct the version +All uncompleted issues are now moved to the new To be determined for future releases category +Raised the for workaround press to unblock the terminal +Detect architecture by uname you do the patch via simply svn diff on command line and not via IDE specific output +It happens to me onOS Win JDK java version Eclipse Helios Service Release Build id FeatureRegards +But it turned out it did not work +resubmit the patch to kick on latest this to trunk and +I just committed this to trunk and +Updated patch +Thanks Mike thats a good point +About the files i am not really sure +The problem is that we create a proxy objects for EL elements that dont support and interfaces and we make cast to in of proxy attr +patch for trunkincludes will test getting delegation token over HTTP +Remove interrupt since the thread now wakes up periodically +Works as youre using CXF the CXF runtime doesnt honor the parameter names renaming them to arg arg etc +its in branches in SVN +if a user provides an BooleanasBoolean method or wants to hook in the invocation of that method is it legal for us to ignore that will and bypass it completely +This bug has been fixed in Xalan Java +Moved to to think its a very complex thing to implement +I came across similar problem recently while developing Arquillian extension and testing against managed AS +I simply limited the number of fractional digits to as mentioned by Martin SchrderThanks to all for the input +The testsuite completed without any regressions from this commit +And the version references you have provided seem to indicate that you are running with this code baseFrom what I can tell these tests are using the same mechanisms that your tests are expecting +Scheduled for +could you explain a little what you suggest +Ditto for table +Ive attached a test project which helps to illustrate the issue +Im not sure this makes sense as the is for JSF requests and not relevant for other clients like remoting +with the details of the Tika exception callstack which would help us narrow down why the NPE was triggered and whereAs per your comment Ill close this bug +Neither way can I replicate this +Same introduced in +This way the IP address is not shown +Thanks +So this rule should only log an issue when a local variable hide a class field +Thanks Miyuru for the contribution +bad idea after all +This is then testing the intended functionality ratherthan a side effect +On my mac they hangfail +This means that the normal way of dealing with inner classes for content assist and organize imports will not work +shipped in +Applied today at revision . +Workaround for current Grails versions create war with grails war command +Verified on +Attaching to fix the problem that breaks the docs build +It is a significant issue that application designers should be aware would like to see the issue around columns that can contain arbitrary Unicode values called out in some way +bulk defer issues to +What do you think a good parameter name would be for this +I looked at it briefly and it seems OK +I think this was introduced by get assumes that the sstables we check for overlaps are part of the live set but now we can validate over a snapshot instead +The way I start the way my install goal is I believe maven launch config been moved to Run menu in +OK I is better than old approachSubmit patch +I dont know which jar was causingthe problem +So now that this is in Lucene we likely just need to update the Factory in Solr right? As well as the Lucene libs if they havent been already +you should set useDLQfalse for the +There is no good way to solve this problem +OK only outstanding items for clearance are +files should be created in a place that SE linux allows +But go ahead +Please. +API has not been modified as it is explicitly doing a conversion to String +I just committed this +Ive attached a patch which creates a based on a Node and then uses the to create aThanks to Dan for showing me the wayAfter applying this patch Aegis bindings should be able to interpret annotations and serialize them into SOAP faults properlyThanks againZarar Ill give it a shot +In many conditions they are used for access tothe repository but there is not way to enforce it +These failures are very environment and timing dependent +This issue is fixed and released as part of release +Improvements will commit soon separate corecontrib and tag tests completely +wasnt it already shown currently when login fails +Patch works fine +Final no? Aligning to the current ARQ version would not make much sense as these have independent release cycles +DhrubaAkash is working on at the momentFeel free to give this nice feature a try. +but the site is still incomplete +But minOccurs means that the tag can be omitted in the envelopeIn my envelope as you can see i pass explicitly zone and parole chiave but i leave them empty +I just committed this +Ive verify mvn clean install and Ive tested this patch on a real cluster with queries +Thats sound good +Otherwise a problem that changed all of the data to s would be counted as fine +Currently i have updated to point to the Wiki link +This patch fixes and test cases from last time didnt make it in. +Removing myself as assignee as I wont be working on it in the forseeable future +Nice! Hopefully Ill have time to try it out soon +Assigning to myself so I look at this tomorrow +someone explain it more clearly. +name jars ls r root root + +Committed to trunk +for patch +Please verify. +Thanks for the work Haohui! Ive committed it to. +The first part I think we all agree on and we want to get it soon +And this isnt related to what I was seeing because I was watching the pages of that were all cold slowly become hot during compaction +I verified that it worked for me but would be good for someone else to check. +Are you running this on trunk? is in trunk and is not yet out I feel we should not worry about it until the next Hadoop release is out +Thanks +The reason and is required because configuration validator requires to construct object to find out number of queues present in the system +The Java build issue has been resolvedReplacing reference to jasypt with a release version +Marking resolved. +Before starting to use statements in place of file I would like to fix +Looks fine to me +Here is a patch for the another for branch is worth mentioning that I the whole trunk for this property and got no result +Can you take screen shots of the maven configuration and attach it this issue +Since we are removing the tabs it is necessary that is ticket is no more relevant as the Clouds service is always displaying information at file level +No other software on the exact same OS was affected +bq +did not fail and failed with or without the patch +I will apply it at some point tomorrow +Is there some reason that this information is not updated for all destinations +Marking as a regression +A new one will come later +Id also like to know more about the apps deployed +Increasing maxmemory to MB +Closing assuming it was fixed with a lack of response +Running through Hudson +Since the last log entry INFO there has been no change in the vdb status +I add assertion in GC to find where object with null vt added to the pools +Make sense +Somehow though i got the feeling that something is still missing +Jocheni can get you an id for there is a instance there +Pull request was merged by alesj +tail is the only exception that comes to mind +parameter can be left as they are +Sergey you can make the patch yourself from the classlib directory by typing svn diff c where is the revision number from Tims comment above +Verified in JBT nightly +Weird I swore I committed this class which fixes it +Would all committers please review +We are already no more compat since we use new logger feature +Assigning parent task to you Lorin +Note there was a minor conflict in the math I can repost the patch with the conflict resolved +I will have a look +I think you should reopen this also agree that this method should not do any kind of conversion or parsing because here I expect just a copy nothing else +I can confirm this issue also +Updating target version to the Aggregator strategy and the + +The patch looks so much better now +Im making a guess that this is no longer an issue +Otherwise running the query should result in a property not found exception if there is a matching row +Ansgar I just love your idea! Really really cool and usefulIm very curious to see how you generate tests dynamically +That makes a lot of sense +If they are implementing paging they are already keeping state about what page they are on +Also setting an explicit and short does indeed prevent the queue growing +The underlying class does support different types. +Please commit if tests pass +Im going to close these +The promised patch never materializedOne point though is that is not in any case so youd be unlikely to gain much in performance by going multithread on an already connector implementation +Issued resolved in commit +We added a test to verify it. +We are currently voting on which includes this fix. +Request always executes properly but Request abends with a close connection errorIf I take out the close statement both requests execute properly +Jimmy want to attach patch with so hadoopqa runs +I like the scheme that Richard proposes since it is a light weight way +to +Ramesh I will test Steven No there is no alter statement +Resolved +Yes is clearer than HC. +Fixed in SVN +Ya should be easy to doIs this patch ok? If so i can go ahead and prepare for also +Dont know if in is a consideration at all +Touches the following filesM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrda who is actually producing and contributing these translations? I kind of doubt that youre bilingual in all these languages but if you are my hats off to youmostly Id like to avoid any perception of commit by proxy which Apache doesnt do but there are ways to contribute work produced by others +I think its wise to remove the mention of how mechanism names are encoded since theres no need to reproduce the SASL specification in this documentbq +Thanks Aaron! +and regarding class not found error why dont we use some test BSP job instead of of examples package +no longer carries an instance in TRUNKMaybe we should do this in too +Reopened to be sure we have investigated all the consequences of the fix +Verified in CR +Running patched version of Hive with Postgres as metastore shows no problem +Also changes a couple test cases because they violated this rule +Reverted the fix in SVN . +Thanks +Closing Jira. +As far as concerned we dont need these occasionally created beans to be alive in memory for everHow would dependency management for beans workI have no idea about the management of it wrt spring framework but it can be similar to a cache eviction or the container which holds bean can use javaHow do you see properties working with setter based injectionThis as well i have no idea wrt Spring Framework but it can achieved using proxy or like hibernate ORM kind of property lazy initeven is a good choice for but it wont address all my concerns too much of change is required in the existing codeAnd more than my requirement at hand i have raised this improvement to see spring more powerful than any other IOC framework +Thanks for the patch Elliott +Regarding qtupdate what is your opinion on a compromise way forward? I guess you mean on +Applied patch into trunk Camel Camel branch with thanks to Robert. +Thanks +I dont think this will get in because I underestimated the scope +Feel free to mark this resolved once the patch for is applied + +please try this patch thanks +Ill get to this at the end of this week is that ok? Ill include it in a bunch of other updates +nevertheless this isnt a valid response because directly editing the xml is a valid use case +Ill look into this but probably not much before the end of next week +You should have gotten something much sooner +I have been thinking about this one for some time +I loaded the stock branch and this problem does NOT occur +Moving this out to please move back to if this will be worked on soon +Fixed in +Michi +Well take a look +Reassigning to Martha to proceed with the build +Follower need to log uncommitted txn sent by the leader to its txnlog during synchronization +Changing the title to reflect the existing bug +Committed to +Fixed in +So if we kept it as they would be largely unaffected? I moved back to that +Trying against hudson +Changes in ant with no arguments prints out available targets +If the app starts completely and can be used the problem is fixed +So if there is any difference it is within the measurement precision range +are missing the signing of the bundles is done mannually right? it is right now also missing for the child repository +is in ZK +Committed. +I am not sure if the WAR generation gives you access to define the role name or not +Im using Tomcat +I wonder why the HDFS namenode was not using the group information in the first place it seemed like a pretty reasonable thing to do +I get status on this? this is holding up a release for us as the repository isnt reliable nor is it syncd +I assume you muct have the piece of code that caused this problem +This patch defers dispatch on a slave til the dispatch notification so that the slave can honor the dispatch decision made on the master +Under which circumstances exactly do you want the redirectAt the moment we redirect for welcome pages as we know they directly map to a directory +why is it a problem? You just put the +The check that resets the index to can happen all the time +Mass closing all issues that have been in resolved state for months or more without any feedback or update +If I may ask a stupid question why does logj need to explicitly stop the root filter +You are right this is not a typical ajax request but it is a script triggered submit +Ill fix that before I commit +There are two kinds of crc type the enum and the int values defined in +Disconnect is totally different than session expiration +This is in line with keeping future maintenance in mind as Owen had pointed outFor now what I will do is that I will commit a portion of the patch that handles the negative counts of mapsreduces and commit that part as part of +This has problems which still need addressing on trunk much less moving to branch +the leftonly class might should be floatright and the rightonly class might should be floatleftDoes that work +If your usage actually went above GB you would have had a problem +Now that looks like a bug +Ive tried to the element from the WTPs Palette to its the same problem in JSP file +Id make a new srctestjava class called something like in the package +It may be good to provide some background on what these annotations provide +Does not hapen in EAP. +testcase attached +added parallelization of dependent join queries based upon the user request source concurrency but with a hard limit of as a preventative messure +The patch with regression test +I also changed the INFO log in checkDirs to WARN + +I agree with Alex about having those new in a separate bundle or maybe +Closing resolved issues. +Revisit open verified does not fail even with a logging service in place +Is this another file you are looking for +I can see that the latest SVN has them both +This patch doesnt change anything in +Please use the free forums or red hat support for support issues +Good job! Thanks +This patch was created after the following patches were applied so to be safe it should wait to go in until theyve been applied first +What are the command line arguments you used? and can you check with a nightly build or with a RCI generated the code with the current code base in the branch and it works fine +Patch for +fine for me then +Perhaps we should just copy the framework version into OBR +I applied your patch and now I can run a cobertura report after integration test +Any feedback on this yet? It seems a significant regression bug +What about adding an verify attribute to the database tag that defaults to trueif absend + +For better performance a LOGINUGI will be stored in +I have merged richfaces branch to master +I simplified the language in DROP VIEW from is disallowed to fails perhaps that should be changed to generates an error +Attached a full dump +Updated patch +r integrates of switchboard into branch +Ying the change looks good +The new one does +And instructions are provided on how to remove let me know if this works for you and is of the text probably needs to change but what is there now should give you an idea +There is a reason we ask for those +Please confirm the current status of this issue and if its incorrect change it as required +It happens if there is any file anywhere in the archive +Martin any chance you can add a quick test case to the patch? For a good example of how to set up an h datastore in a test case see in the wfs module +Postponing to since works fine on external browser +committed to SVN SVN I have verified the fix by running a test +Committed to trunk and. +I have tested with patch from +so the relevant patches are now and it! on the old version +Overall it looks pretty good +The release notes on this ticket should be updated to indicate that in addition to the improvement it also fixes the issue +Yes Im trying to build a reproducible test +Bulk close of old resolved issues. +and making the user mkdir all of the storage dirs before running the NN format is pretty unfriendly +two months later is there now a fix to this problem available? Or another plugin with which you can copy around in your project at wish +Im moving this to as theres no urgent need for this in +Cant reproduce please provide a sample the end of the day I will post here the sample project + patch added a protected resetBuffer api +The web module builder tool is complete and tested +It works fine for me +Seems like there still are issues with the patch as attached +I dont see any point in repeatingthis block more than once and have removed the repeats in this patchOut of curiosity does anyone know if there is a configuration setting that will make this testpass without modification? I tried setting true but that doesntsuppress the tablealreadyexists warning +Have you already looked after this? I am waiting for this +committed to issue as it has been released as part of release +diff for for javatest the patches applied +Again this is an old issue that has been resolveddiscussed many times before so please discuss on the forums before creating JIRA can it affect versions and when those are not released yet +Yeah that sounds about right +Done +This fixes my problem thanx +Not only to get the correct subclass by the way but also to get the correct message text +has a wrapper class but I dont see how it would be significantly better if Context were an interfacebq +Cant reproduce this bug after is committed +agent crashed with the sigar patch +in the PageControl constructor overriding get using lazy loading in onInit and onProcess +The patch solves the problem +RameshI moved all the resource adaptersconnections over to CR with JBOSS EAP +I dont see how the scope or the origin of the code affects thisIf the class will never be extended then make the class finalIf the method will never be overridden then make the method finalOtherwise extract the code into a private method as originally suggestedSince the fix to make the code safe is easy to do it seems to me that it is worth doing to avoid potential problems later +Patch looks good +I guess we need to apply the same logic also to other Tika parsers that may deal with relativeSince we in any case need this functionality in Tika would it be useful for Nutch if it was made available as a public utility class or method in? It would be great if we could avoid duplicating the code in different projects +I have a question regarding this fixI thought an alternative to having in javatools was if you have a property for junit in a properties fileSo I have a line like this in my junitlocaljunitThis way I keep out one more in file and I need only one for a number of branchesSince the fix for this bug went in the build downloads which I think is pointless in myDid I do something wrong in making use of this property? Or is it ok to change this so it will only download if the junit property is not setthat seems reasonable +This doesnt appear to happen locally +I could live with the synchronized issueIll provide a patch if you like +Other approach is to to check if the has delegation token using for that particular hbase cluster and if not fetch the token +Closing the issue as the first version of the test has been completedMore work can be done for instance investigate the performance with the client driver and add newmore tests. +OK after discussion it seems this issue still needs to be fixed +I was talking about the Work Log tab at the top of this pageDoes placing the Railo logo on each of my websites count as sponsorship +But there is still a issue there and I am still investigating +I simply do not understand how this ivymaven repos stuff works +Marking Patch Available so that Hive QA can pick it up +I added the test cases. +I should also note that my script is a batch file DOS script +This is not for commit here but it is illustrative of the approach being planned for trunk as well +I followed exact steps as mentioned by Hans +unticking the option where? what option? Are you trying to disable the WCS service +Hi TerrenceI would just like to clarify how do you Restart in AS +Done. +Im sorry that youll be out of town Ive setup a dial in and would love it if you and others that are out could attend +Since we have a different case here JMS API used against a JMS compliant provider I would like to solve it through automatic adaptation in +MingjieAndrewThanks for your reply +I didnt think to check for a parent pom. +I think we can just add back the row +That is strange too because I dont see a either +close off release +Dear DanielSorry I do not have the time to carry out any more testsMy intention when sending in my findings was solely to make them available for the Apache CXF development teamMed vnlig hlsning Best regardsRuno Barrdahl civ +doesnt fail for me with or without a patch +I noticed that the rackdc properties file is in the source but it doesnt get built into the debian package +Ill fix itPlease dont resolve issue until you test it +Thanks +Prep for real patch with unit patches are against +Ive already had to put some flaky hard drives off to the sideFor now Im really in wait and see mode +I arrived at the exact same problem inside my build but dumped out my environment properties +How about attaching your utilitity for consideration insteadBrandon? A lot of people feel the same way about disjointed validations +sbt update look right to me +The change over to the new CRS implementation provides an bug needs to the hsql epsg database to provide valid areas before we can begin work +stack do you mean I should submit the patch for as well +ll run ant and fix this +We would really like to have a solutionAny comments on array setters and making these efficient? +Do I have to check out the revision that dates to to get this to workedit nevermind the answer is yes +No worries thanks for your help BJ! +the four files log and snapshot should go into srcjavatestdataupgrade +Running all functional tests +Maybe to clarify the point about using the logj api that only this got logged only happened when we configured it using the at least the container slfj backend and frontend should be taking precedence even if slfj backend are bundled with the deployment +Here is the patch +It is merged +measure the time to display a dialog box with an without the optimizationThat way we have something factual to detemrine if the optimization is worth it +Hi Andrew thanks for the patch. +Michal is gonna build and test with my topic branch and see if the issues is fixed. +created javadocWill update the patch shortly +Ill see if a more recent derby version fixes the problem +from all expressions in +Only changes made were minor test changes that affected test timing. +It turns out that the thread context classloader is null when using JNI on the Sun on certain UNIX platforms and that causes the EMF failure +mingjian Can you help us? Li cant reproduce +This is the same issue as +Is this still an issue? Do you expect any help from my side +Thanks +Thanks for your quick response +Attached the that is most probably a dupe for applied. +verified +The jBPM stream is not going to be supported and should no longer be under further development is apparently happening on the stream with the latest version currently being jBPM and releases should be pushed up to their download site today +Patch adding a corresponding integration test. +regarding +Shouldnt it be used to figure out if the send operation needs to be retried? +Ted Thanks for your help with the testing +That said it can still be done +The default setting is jakarta BSF +I assume you changed and that I need to check out the CVS +Confirmed that this only happens with applied +This is per HTTP spec requirement Found +Oleg +Im also with Jan on this +Its the project responsibility to add it to its classpath. +You can change this is you want to host the in your own companys domain but by default the addon creator feature is intended for Google Code hosting +Using Struts for portlet development does require a different setupstructure of an application because of the restrictions of the portlet spec +The table type will be whatever type you specified as the default for that database +Can we get this addressed in +Is this still a problem in +Looks like these are deprecated nowShould I look for a workaround or would Tapestry fix this at some point +Has the patch for the patch been applied to trunk? Im running trunk and the cache does not seem to work at all and I cannot find it in the svn history either +for the simpler approach +It can be initialized using web config parameters or as a +Just setting open to the community for the moment although in the next few weeks I am planning to run through the pending issues relating to domain management security again and schedule against releases +The tip may confuse the usersThe other combinations a valuable for me +Done. +Actually I am using feature described here I am binding many columns to the same variable and it works fine +This patch contains below changes +Gnodet we can only backport it to the branch as this is where all the new stuff was Fuse MR has no branches and has not been included in any Fuse ESB releases +samplesjunittest +I just wonder if it would make sense to disallow using this construct inside nested the outermost bean might not be intuitive especially since it is not forbidden to specify an id on a nested bean +Thanks for this explanation +It looks like a diff on a previous diff +Secure connection to a proxy can only be needed to secure proxy credentials in case of BASIC authentication +Also we should add yarn +Attachment has been added with description implementation P open bugs to P pending further priority for bugs with a for the patch Peter +Committed branch and trunk +How do you envisage this being configured in a way that makes it clear they are test ng properties without making th eplugin too verbose +is the functionallity there +Committed +I have created another patch for the Developing Apache Harmony Code with Eclipse page that includes Svetlanas changes and also makes the page relevant to DRLVM developers +Patch that adds a timeout to the bootstrap script +Would you be able to add a unit test for this new functionality +This is happening even after adding deply we many need to disable this test case +I just tried running the script you provided with default and iso and in both cases the test script executes so I am marking this as resolved +Your JAR seems to be corrupted +Verified in +Works for meThank you Oliver. +Which object was null? Foo or bar? The Assert class gives the developer more informationIm not trying to push this btw +This is an issue that arises with the frequency of the clock but it is in fact a runtime issue +Thanks +Hi Thomasyou have reason probably its better to leave this for the release more testing is requiredThank you again in the meantimeBye +I bring this patch up to start the discussion how we should handle this particular caseUsing version would not be that much of an issue as all Analyzers using a do have the Version class already +Patch for +Triage for m wondering if it makes sense to simply make the synchronizedThis would make the servlet serial +I think theres just the CRLF issue which needs a resolution +Says improved things +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +will test and work out any kinks during the release process for plugins and some test artifacts next week +Fixed in good enough for me! Thanks +Right thats why I said Easy expose gBT +Closed upon release of. +But might it be better to simply modify Task to use directly? Theres not much left in Dispatcher at this point +I really think that this is better done by using a separate and +Ill be sure to use the forum first next time around +time +Thanks Cristoph +All is fine +It would be nice to see one or both of the workarounds as well +Hairong it seems like the test for this fix has been posted separately in +Ok ill make the final vote for it and it will be included in +patch Tony the patch was applied to LOGGING module at r +Closed Release note not required per comment by Mirek. +Patch +In particular this causes an exception in the ASM local var sorter code when a frame entry refers to local var slot which has not yet been written +Seems this can be solved entirely on the Infinispan think this issue has been resolved at the level please reopen if this isnt the case +Ive changed the default htmlEscape method to use named entity references as this does incur a performance penalty over numeric references +The TCK tests will be run as part of the CP testing which Rajesh will oversee +can you check that also please +In embeeded Jopr was link to old was an issue and so does not require QA from the Enterprise side +A mass update like this just adds regression risk for little benefit in my opinion +Revision On Debian Squeeze bits In a room add file In the popup window select file to upload Click Start to import The file upload and the conversion start The popup window stay open forever with the line upload OK converting +Robert +once we get another test case we can hopefully put this to rest for good +A patch for removing the dependency and making corresponding changes. +It has got a new timestamp +The patch looks good +Thanks Pauldo you have any idea when this patch will be applied to the branch x +patch applied in r with thanks +Keeping the issue open to track the commit on trunk. +Here is a patch which only moves to the INCOMPATIBLE CHANGES section for +This is a deliberate decision +The patch failed to apply to trunk +So I am not able to get the output file name from the taskid +verified in Developer Studio to set release note note not required for issue now release note status has been set +verify in after hand off to against. +Magic number is always not good but anyway its not a problem +The failing tests purpose is to check the max read OOM prevention feature removed by +I already respect you a lot for the time you do invest +Because of that we havent been able to load the permission +JucaI am still seeing the two unexpected failures using the junitinstrument ant target +Sorry Mark the patches has been for a long time and no one response the issue on so I think I should push it a littleAnd I just committed and other patches are still in discussion I didnt commit them +Since we use already why not put all the test configurations into a jndi structure +Attaching scripts to create the table and to reproduce along with sample input data +This looks good to me +Is all this worth it just to avoid stack dumps in aborted speculative task attemptsProbably so although Ive not looked at the details +Please create a new skin family for desktop renderkit +That will be helpful +Hi OlegI would mention the optional requirements after the hard requirementsShould we split the description into build time and runtime requirementscheers Roland +Patch resolving the for the contribution +closing to clean up the outstanding liststill would love an update on it +It doesnt work it results in a test failure as well as similar exceptions when I have sharedLiblib in my +Ouch I accidentally checked in some test code. +get unwrap etc +We can fix branch release +huh? Only if you select the library +My bad +occurs when an import statement has a start position greater than and a length of less than +Here is a version of this consumer endpoint that passes all of the unit tests and also appears to work flawlessy for us +Spoke too soon apparently this may not work for so the shims approach should be fine +There are good use cases that bypass HTTP altogether and we should strive to make the replicator etc work with those instead of adding arbitrary limitations based on HTTP implementation details +Thank you Plamen. +These work so far for me +Hi Nicolas I will work on the branch patch +Testing also established that even in large clusters it is still best to start all the Datanodes together and not try to stagger them manually the Namenode does not have difficulty enqueuing the flow +Actually turns out that was working for attributes but not for elements both for and for + code to handle defaultfixed values needs to change and should be handle in is being dealt with as part of work on other issues. +However the memory capacity per slot on one type of machine would be larger than the memory capacity per slot of the other type of machine +I figured it was cosmic rays but now I wonder if theres a maven subtlety Im missing +Im going to revive this and if it works fine for trunk then we can use the basic system for RT eg + +Thanks! +it seems no one is against this I will clean this up add friendly deprecation warnings to the patch +Here is patch for this issue +CR build is already submitted +It would be beneficial to add a section describing the registry interceptor mechanism along with descriptions of the current interceptors +Fixed in r +that is why i referred to it as a partial syncim really having problems trying to reproduce this error +I look into the code of our which works actually like a hash table +FrankI have attached the updated Reference test case and schema file +FutureCurrently there is a fix in CXF to get around the jettison issue of having to refresh the CXF bundle afterwards +Then I had a look around and noticed that a there already exist even simpler s in core and b they dont even have any dedicated test coverage it seemsHonestly I was quite surprised there was no build in functionality that did this already +Others depend on anything from Java to so it needs to be done per component +The problem was that I had to modify James to use the same authentication store as several other applications and because the alg was so strange it caused a real problemI dont think that the one it was using was all that bad but it was very tough to changeId request at the very least you allow the administrator to specify the method to use +We could add a more informative message to the logs to help debugging +Please review please review the doc +It was like that before my patch +FrankI think Kelvin meant detach the target dataobject from the tree serialize it then put it back to the treeThe problem occurs when a dataobject in the middle of the tree was serialized to a Java object the Java object will contain info that was not serialized which means the parents of the target dataobject wont be in the Java object +That fits into ctxs picture + for the patch +v attached +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Hi Nick can you please provide issue tracking for JSF RI issue +OK +Thanks +were backported to as well adding the latter fix version. +In my own case an isnt sufficient as i need to have positioning capability without the insertion capability +I think its important to work on this details to make the Stanbol Java API more usable +Attaching a new which adds caching functionality to the and the there still the Advanced subclasses which might need similar treatment +in my case I have more weak computer and more pure internet connection this could be a problemAlexey Ive provide steps to reproduce +Yet this could change in the future who knows that way we could get rid of the deprecated methodes little quickerFrom a backwards policy perspective we should do it that way +Its not a licensing issue its a if you change this for everybody else most code in the world using Xerces on Windows will crash issueIf they want to add optional build targets thats fine but as a default its entirely the wrong choiceNote that MS itself now defaults builds to the DLL version I believe and has generally steered everybody that way so that they can properly address security fixes to the runtime +Very often a Configuration class will contain both methods as well as other methods +ok taking a look +Read M rows columnsh +I agree with Eli we dont currently use the abstraction in so it can only use a single shared dir +Resolved as part of. +On my linux laptop it fails after connections are queued when it runs out of threads +Here is a patch for this issue implementing the solution described and applied the patch +Attached patch which makes to use instead of +If you do want to work on it I would be happy to review it and commit it after your upmerge +It seems to me the patch targets for a different problem +this is the simple A B class test case the works +Committed. +We dont do any merging hereDo you have any sample project to demonstrate your trouble +Thanks +oopsNote that since is always escaped this can be done using sscanf to read the two digits and convert to a byte +The zip release file is put in a different directory from the tar release file +But DBRE assigns uniquetrue on each find attached my roo project +Its also bad that we have abstract base classes and services in the same bundle +Either this needs to be resolved or the changes should be undone and we accept this as not fixed in problem is that the and are attempting to use jndi from within their createService methods +Well I see no any difficulties with this +Hiyou have to additionally install the core pagesRegards Florian + is for trunk other patches are for the branches respectively +Snapshot has been deployed. +Most likely wont do a as there are some other issues that require ABI breakage to fix so next release will be and ETA is when its ready +Moved to post DamithaThis sounds a good improvement +OK I will +WSDL document used for the faulty WSDL wed like to close this JIRA if its done +But for parameter based configurations it is required to include the corresponding handlers in the and which should be included by default. +The old version is wrapped internally so that it works as it used to +The onloadend event is not returning the result +Is it ok if we have those for or do you think they are blockers for? In the case that we move this new functionality to it would be great to create new jiras or update existing jira to reflect it +Im no expert but AFAIK the sanitized poms were the ones with the substituted versions? Why not keep the parameterized in the normal +ToddCan you upload your work to JIRA now so that we can review and merge it for +We may want to post patches to this JIRA instead of open a new JIRA. +I liked the test case +I dont see this as a problem at all +xcss format will be not used in build +I dont have an arquillian test but the attached maven project is pretty easy to run +by default it would work identical to jeffs patch but with the possibility to override it +This patch changes only tests +are not found anymore +Thats why youre so pertinent +Please feel free to sever this link if these are not the same bug. +It is NEVER safe to CFLAG and LDFLAG values +Did the tests pass? Anything else I can do to help +Same here +Do you think that the IO might be faster if the Vector was replaced with BLOB IO via directly? That is what I am working with right now and I did observe my indexing time for a sample data set go from seconds to seconds +I think number of splits is a useful metric to show but doesnt deserve an API +Ive just committed this. +Removed in revision +batch transition to closed remaining resolved bugs +Oh thank you for correcting the purpose of this issue is to rise and expose the idea so yes write the by myself is an option being evaluated +Attaching patches to return the JDBC Driver version number from the manifest file for both the Driver class and the class +patch committed to trunk +I have triedfoundJvmecho Using JVM PRJVM C EXECUTABLE ISSERVICENAME start stop +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Now that work is underway with AS all previous community releases are +So any exception in job and step persistence or exception in commitrollback will hide original exception and sometimes cause diagnostic a bit difficult +Thanks for the schema file Len Ill see how we can get this fixed up +r +Hadoop command with libjars two files seperate by comma without a double quote on the libjars parameters and it passes with this patch +That was not clear for me from the docs +Patch integrated to trunkThanks for the patch ElliotThanks for the review Stack. +Sent email to dev list since this would be a JS change +already committed thanks namit +It is more important to have a good design +I see no point in breaking existing code +Updated patch that drops unnecessary interceptor and with more complete event handling +Cool I can reproduce on jruby +Now looks for the root folder containing the file and uses that as the focal point for the security artifacts +Before I would return only the configuration the resolver was asking for +I am very concerned that the patch for this issue is for the test and not the product +I appreciate for your kind comment +Apache Maven project for historical reasons has most of its JIRA projects at codehaus +It would solve the problem of in file names but just create a problem with name transformation and use confusion +Also note that the fix is not in +Running tests now +There appears to be a problem with the deployment maybe aduplicate version of FOP or a problem with the servlet containerThe information given is much too sparse to do anything +I dont think were seeing any emails on the dev list from JIRA even though issues are being added +Id like to take a look at the new functionality +The regenerated patch file is and the corresponding svn status output is +Carl thanks a lot +Note however that s javadoc must be improved before it is made public +has been marked as a duplicate of this bug +AntonioCan you please resend the patch? It is not compatible with the latest trunk + on +to JDK support actually it is fully supported +Fixed with the committed change +Instead the following will work essentially you should expand HADOOPCOMMONHOME yourselves for now HADOOPCOMMONHOMEbinhadoop config HADOOPCONFDIR jar HADOOPSTREAMINGJAR input mapper pathtocommonhomebinhadoop dfs ls reducer NONE output via + +addresses issue including test +We did agree to produce a hive release in less then months that did not use the trademark +The sorting issue doesnt account for the difference in reduce input records rightNevertheless ill see if fetchTime can produce a suitable hint for multiple segment inputsThanks +I think throwing exception is better in this case +While this is not damaging it is still redundant +At the very least the cookie management code needs fixing +bq +TalluriSailaja was it a multi node management server setup where you faced the problem? Sailajas log shows that it was multi node MS but Talluris doesnt +How people are supposed to use it on Windows? Is there any workaround +I then started another client to create the same file +Patch added prefer to surefire configuration +If does not set it then some internal use of may not work +Here is the patch with the following changes +is not a bug iterator order is not based on the order you added features instead it is based on the spatial index used for the shapefile +Duplicate of +Fixed as part of this record since is now released. +Can we add a test for this case +Fix is in the RI and the TCK test is working +Alby and I had a think and I am implementing a fix where the DT is just stored normally in the memory manager rather than being in a global document +This is a duplicate of an item I already have on my task list. +Lets wait for Hudson to run its tests and Ill commit it to trunk and unless anyone brings up concerns +I guess you dont use really the last trunk version because I just retried with a totally clean instanec and I cant get this URL working nor the help disabled the url which I reversed on my demoD +Remove the synchronized keyword +Marc could you give me an example of someting that works after this change but not before +It contains wrapper for STDMALLOC and STDFREE which is switched on by MEMMGR define +Removing has side effects and afaik we havent tested the system without it +Ignore the last comments and see the latest attached a slightly different form was needed to specify theThis client demonstrates that it is possible to fully define a security domain using json over the HTTP management interface +Thank you for your help +Uploading a patch which applies to app in branchBefore committing this will reset mrprotos on the branch so that the merge from trunk is clean +Jeff do we want to set a fix version? I find it really useful to know with which version contains which changes +Yeah our belief is that the shuffle process ends up making a lot of the getpw calls and we have already seen a couple of ldap servers outages +We only have a snapshot until the Alliance releases the final companion code for but this snapshot is more recent +The patch makes the method not synchronized and then first synchronizes on the and then the +Applied your patchThanks! +Attached patch to remove the extra loggingIll commit shortly +Cheers in CR +Its been discussed before +Thanks for reporting and the patch +Looking at the code I think I see how is not quite fixed +I fixed this recently because I was sick of the extra output during spec runs +I think this should be fixed for the next GEP release +BTW underlying issue may be a JRE bug +The following explanation explains why the folder is still there at step The reason the folder is not immediately removed when you remove it from the project wizard is because WTP will not remove it until a publish event is you remove it from projects it is marked as removed from the server but not physically you click the publish icon it will the server with what modules still belong there +When I click on the Resources tab and expand Blogs and Articles About Derby in the left pane I notice that clicking on the links takes me to the Blogs and articles about Derby page but I am not positioned inside the page on the correct headings +For posterityThis bug was really hard to trigger as it could only happen if the pools first initialization fails bc of incremental recovery failre when a transaction is active and the min pool size is +supporting groovy made enums on is possible we would just need to replace the base class +This caused the subtool to fail when being loaded +Can I resolve this as wont fix? I dont think we want to permit multiple records with the same name in a schema +Seems to have been fixed +Upload a patch against the last patch is only for since there is some difference in between and the last patch cannot be applied to directly +You should be able to use concurrently and not have them interfere with each other +Forgot to close. +This would then be database agnosticthoughts +maybe every th timeBut I dont understand why +The patch looks fine thank you paulex +am closing this we are not backporting this to and its committed stable on trunk +there would be lots of files opened for writingreading +On the other hand Im unable to import the old ova to so there is something wrong with ova format or specI think in the vmware logic where were importing the template to spin off a systemvm we should ignore any such error +Nope I left it open because of the version numbers patch +Thanks for the patch Kihwal! The patch looks good to me +The exception could be sent if we can find a way to detect if both components are sharing the same thread. +And then tasksjobs start failing +We will! please contribute +Oops felix I missed your responseYes I do agree that may be there is a problem with the testcomponentfactoryrequireconfigurationI actually started to modify it in order to reproduce the problem but I have not yet finished +I dont think there is anything more to do on this issue. +to be clear that mail +I think all tags are gone nowI would appreciate if you verified this and close the issueVersion of the patch commited to trunk with revision +The Follow TCP Stream file looks perfect +We should have this completely defined by M so that all there left to do is add components that get developed +The fix for this must be preserve backward compatibility +reviewed +Keys must not be a +Well regardless of version you need backwards compatibility +I am resolving this Cannot Reproduce since you could not get it to happen again. +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsermodel +Updated to apply with trunkMoved the valid values for to a common package as a side note when you make a patch its easiest to deal with if the path is relative to the solr root directorysrcjavaorgapachesolrsearch is better thenUsersekellerDocumentsworkspacesolrsrcjavaorgapachesolrsearchI am new to the list and to Solr so I appologize in advance if I say something sillyI have been playing with the field collapse patch and I have a couple of questions and I have noticed a couple of issues +Uploading the output of the test from the Hudson run of +The current behavior is too surprising +Ive fixed this in the trunk +I agree for very large fields this is expensiveIn fact your patch does not handle this case and I agree with the others as its to heavy to implement and adds back the crazy complexity we had with lazy fields co +but it seems to be using a clone of whats in the postgis datastore? Cant we have just one and maybe create a template method that subclasses can override for building this +Oh this is the GUI console? Not the cli? Ive no clue about that +Shinichiro you should go ahead and commit this +one last comment protobuf has an RPC framework it just requires one to provide a transport + was solved in at rev to +Any request for a high priority compaction is assumed to be under the premise that it is the most important thing to happen +A patch for review +Probably under the jbossas projects directory +Those potential improvements should be evaluated separately and individually afterwards +The attached patch factors out a few methodsfields that can be used by MR and backends into and +Thanks for the review Stack +It seems to me this will lead to confusion about updating those values or adding new ones +In JSF applications you could do the same using the +Cant remember how to put the state of this issue to In progress not a big deal anyway +try the latest patches for this issue +Now it works as the update frequency says in the descriptionI set the alias to fit your requestThank you tooRegards looks good now close issue m gonna change my address because a lot of spam coming +Thanks Knut +What about displaying loggers only with the verbose mode +As there has been no feedback in the last days I will assume the patchworked and am closing the issue. +Still a bug in r +Performance has very low overhead at the moment and I dont want to increase that +This is supported by reasonable RPC implementations +This should not show up in CP tag +Im a little conflicted maybe this is something that should be configured the IDE level but it shouldnt hurt anything +Hey GuysAwesome that was the intention for Tika from the beginning Jerome and I originally proposed this as a downstream feature and I think that the time has comeThanksChris +The field should be marked final +You cannot check that and take the decision because if there is an error handler defined in the sequence this logic is not going to work +I was assuming you but it doesnt really matter +related work around that I used when I was getting null pointer on in integration test. +This hasnt happened in a while +Do we need aIf we want to distinguish the different cases between retriable not retriable dont know then yes +out there who will be surprised when they upgrade to if we leave it be? Or users of +I implemented better handling of All issues types select item +I especially dislike the part of the error message that says parachute is non existent +Disclaimers might actually confuse the users +Norman can we do this? Its just a maintainence we can +Ill try to replicate the issue on my linux box and Ill get back to you +This test patch should be applied on top of the previous test patch +Im using zips exclusively in GVM now. +I opened thank youAnd what about the missing argument names in the code assist window even in the case of this issue? Should I also open another ticket for that +Id like to commit this to as well +Ok I see +task is done. +Gotcha on no I will keep that in mindPardon my initial mess of a comment edited for clarityBy ignoring in your try catch blocks for takeputcommitrollback there are going to be responsiveness problems since most transaction implementations are likely going to be using some kind of blocking patch awaiting an updated patch soon +This bug is realy ugly cause this way its very difficult to have a sourceforge project site build by maven +But will try also changed the setting to so this one can be closed out just have to turn my assumptions on security on their head then it all makes some this is correctEverything is closed then you open doors from top of the tree to the bottom +This patch has been updated to provide functions of extracting RDF metadata from the embeded documents +thats why the default isnt arbitrary its a function of the weighti never said it was a reason why queryNorm was there +This is most likely a class loader issue The Spring bean factory doesnt seem to know the interface that the implements possibly because the bean factory is loaded by a different class loader than the t we had a similar issue reported recently? Rob could you please check this out if you have the chance to? do +Attaching +Julianseems like could also provide a + +VerifiedLeo. +Kevin wants to look at this himself +Is there a full stack trace for thisThis is resolved +At least for users it does not make sense to depend on karaf main or the felix impl +Please find the updated patch with tabs removed +big for +batch transition to closed remaining resolved bugs +I guess youre putting the onus on me to show you that not working concretely +Committed. +Postponing to the nextI am runnning into same issue +verified +Since is currently the only value supported for the language parameter that fact should at least be stated in the javadocs +Having implemented a fix for this I discovered that the fact that the message is incomplete gets lost in the upper layers so the client ends up seeing a partial message as if it is the entire message +thats actually the opposite purpose +I just committed this +I would call tell the developer what happened when there is a fatal error a significant improvement in functionalityThen lets add JUL configuration to the example server setup in the documentation +to Thanks for the patch clockflyThanks for the reviews Sergey Lars and Anoop. +g files any use at all so just edit it directly in eclipse. +The only thing the force split request does differently is ask that the max file size be effectively considered to be +Test cases can be added to functionTeststestscompatibility and the test itself should be run by invoking the Suite class in that directory via +Thanks Kim +The apache parent POM were using already enables the GPG plugin so theres no need to define it again in the project POM +for patch +I just encountered this issue on Seam using example +Apache project services generally must be hosted on an +Rob is that a bug or simply how the model beneath looks like +The masthead and areas grow with font size changes try it out +Mark Namit requested that I add negative test coverage for each of the exceptions that are thrown by the modified version of executeUpdate +This is a funny one +In my opinion it is different situation to let Openshift create your project with default settings but when I have my own project I do not want external tools to change my files and settingsI do agree this issue can be closed and Ill prepare a new one +release bulk close +I improved the area with rPlease check the change in the Subversion Commits tab and give feedback whether there is something more to improve +That will not work to edit Calendar Events with a user that has no timeZone set +Thanks a lot patch applied +It had two logs to split +twas for the greater goodI will look into reintegrating ripple and see if it works now +rpmbuild has bug to use long file names with dash character +Come to think of it we can still increase the configuration entry from to say and see if that helps +Glad you already noticed itIve resolved as a dup +Since this is a trivial change Ill just do the fix commit +In manual has been displaying and in alternance +Patch with tests +Already fixed in last version +I modified the example application so a I could run it in glassfish v +The annotation processor API is only official since Java +Thanks for reporting +Hi SergeyAny change you append your whole SA projectAnd did you already specify your version as to pick up the latest fix? Freeman thanks for this information it all sounds perfect Ill reopen later on in case Ill see the problem againthanks Sergey +As to the record issue Rajesh mentioned I think it is ok to error out +You may want to open a separate bugreport for this because this one is already resolvedOrtwin +I cant reproduce this issue +Brian would you mind coming up with the patch for the Ganglia context +Proposed fix is attached +Please close it as invalid +Id prefer the first option and factor the other conversions into a separate class + +These benchmarks were run on my laptop and averaged over ten trials + +to trunk +Applied patchInstaller in EULA are working but they do not open separate browser +Perhaps someone has done that or would like to do itReopening since it appears there may be a way to fix it +Anyone have an easier way +Closing released issues. +Dozer test wants to load Xml resource from classpath +That seems to work and will solve our immediate problem +but this is not only issue through the generated web layer also not able to listaddedit the rows for the entity which has composite primary key and so this seems to be bug rather believe I missed this update in initial bug logging + did you assigne this JIRA to me? Im not involved in AS development nor author of these testcases +Reference Data is typically stored in nd level cache but I wonder if it would get there without the connection through st level cache +Didnt get much of feedback I was looking for dev +Patch v is same as v +I made the patch added a few comments did some casual testing and committed it +Uploading latest correct patch +My apologies +Is it worth making this change and causing disruption to users? +Role Permissions instead of User Phone Numbers +Yikes +if we had +Using service names sounds good to me +Are you raising thisas a neatness issue +If debugging is enabled the Call object will serialize the request into a string and log the message +To simplify the patch can we agree to add a method that returns a? This wayapplications may have access to deleteDocs and not be encumbered bythe method +I will move this to the EG +Then we should probably add a version of getChildren that takes and returns an offset so that getChildren could be called multiple times to get all of the children +Both are for clean up after some issues which sounds to me better to be in the final block so that its done all the time if not open successfullyWe can address this in a separate jira if you prefer +just linking to make it easier to see. +Hithe only workaround I found was to remove debugging args +Updating patch on review board would make review easierVery well + +And as much as possible we want to permit folks to build new data representations that might map Java types to Avro types differently +Ill remove them from the filtering patch +Justin it seems so +I made the change an committed it +implementing is solved +Remy can you give an example for when a user might need to change this to true +The +JornWill close this one +Sure we need to fix it and give a warning to the user that the plugin will use the default mapping instead of the blank one +This could bring the versioningrevisioningbuildnumbering into trouble +The coming release already contains the fixThanks confirmed its now fixed by r +Ive committed this to trunk and. +From tests it looks like we should not overwrite the project file and then the user feedback on authentication error will work +pushed the revision closing the one of you send me a private JAR for this one +Thanks for the format suggestion +The new code directly sends the event to the dag. +Went over the patch and it looks correct +ll try it again later on +We may want to postpone this issues since this kind of testing is done by automated tests in QE process +Mine is also a single sharded setup while the issue markus mentions is for multiple shards so Im not sure if this is really a duplicate +I also would like to add that after staging the directory specifiedby is empty all the stuff created can be foundin the directory of the parent module under targetstaging +Yes thats the technique I remembered +fixed in SVN after release +If I were doing this project I would avoid the upgrade issues by supporting both formats in some form +I attached the plugin +Sure +SVN havent done performance testing still algorithmically get is improved +its taking me a bit longer than I thought +This chapter has been removed from the EAP Admin and Config Guide +Is there any more information you can give on this? The input being used or any more of the traceback on why the input is not being read? Theres an underlying error somewhere thats causing the failure but its not apparent from the gist what that underlying error is +That would be I think +not a can you take a look at this for me +Luc can you testconfirm that it only tested this in VFS not actually using any JDBC features. +Heres a patch which addresses the issue +Thanks Oliver can confirm that this is fixed with the latest SVN since the reporter confirmed the fix. +Ok +What is the timeline for making this refactoring happen +I assume the you added here will not interfere with any existing thats going on within a +I had thought we were going to need to back port some of the enhancements to to resolve but this did not end up being the case +It appears still to be an issue with I tried the test case in the description and it evaluated to for checking and for the patch +I tried my best to be evil to it but failed The nightly tests also seem to run through the tests with no issuesIt looks like a winner +something is amiss with this patch the last part applies changes to a file which isnt there +now rolls back the transaction so the db returns to its previous state regardless of whether the trigger create succeeds or failsProblems with lock timeouts in the rollback tests continue added a testcase from +Ill take a look +Ive committed it. +Ah yes this is from Hibernate one of the reasons why we originally changed to debug logging for that case +Fixed. +I just ran into this problem +Care to contribute such a feature +For decimal values do we get enough precision by converting to double and return the result? Or do we need to create custom decimal add subtract multiply and divide functionsBesides resolving the decimal question I still need to add date time and datetime to add and subtract +Tohu itself doesnt require this class but the used in some of the examples doesSo youll need to add this jar to your war too +Patch attached +I just hope that this is not a deeper issue with the way inheritance and aggregation are implemented so that this issue doesnt pop up at other places as well. +So I think it is time to move on to Jetty or even as has beenproposed elsewhere along with upgrade to servlet API. +Updated the patch to apply cleanly to current SVN Committed revision +SVN diff with the changes. +The key should be but the exception says that the key is PRIMARYRadim can you replicate this in a smaller scale and generate some logs with TRACE on for packageTristanMircea can either of you have a look to this +ThanksAndriy +Rather than a configuration option we should just change the behavior +Then from there introduce some retry logic +Uwes jenkins runs on windowsthan the freebsd one at apache and what not +Can you try building that against your copy of Xalan and see what happens +But as current my job is not on hdfs I cannot keep track the updates in time +This will make it much more user friendly and useful than the existing issue tracking systems +code review +Ported from trunk to the following branches at subversion revision at subversion revision +I dont think either of these could ever happen given the realities of how SSL connections are used +after remove findbugs warnings I just committed this. +Perhaps these classes should have an option in the constructor to select if you need strict accuracy or not you needI think ULP is a little misleading since it seems like the overall absolute accuracy is within eps in my tests at least for sine +Ive made an initial drop of the jaxb collection support +Fragment host symbolic name updated. +Committed revision in nutchgora revision in trunk as well +I agree that sources jars would be helpful +Proposed patch file of patch of why would you take the principalDN and replace it with a constant normalized DN +Thanks Rick +We would either need to rename trunk or use something like sitecmstrunk +Since theres a separate link on the top I added an extra page like I did for the rest of the missing pagesAdding ApacheKalumetsitesrcsitexdocdocumentation Adding ApacheKalumetsitesrcsitexdocdocumentation Sending content ApacheKalumetsitesrcsitexdocdocumentation Completed At revision +We need to depend on this package for Commons VFS on which Apache Axis Synapse and many other projects can then depend upon on in the near future +The class together with a test case and an example is added to contrib directory +Assigning to Len for now because QE are going to try this on a Windows box +Sorry for the inconvenience +Iassume this os +I added the license header on github +Yes I just did that and it worked +If you have further problems drop us a note at either the Archiva or Maven lists Ill close this for the time being +Ive committed this to trunk and. +Jason and I saw this happening with Collections but it does actually happen with normal attributes as have just run the renovated tutorial and it fails with normal should be committing the tutorial today +After running again it turned out not to pass + +The wiki is not working for me so Ill upload the wiki comments here for now and will add them once its back up again +And visual and source should work on first time you open a new document it should show either full editor with vsiaul or full editor with source +alejandro making it work is not difficult i know +Bumping its status to a blockerRegarding possible solutions we can have a at RM like Todd mentioned or we could have a simple JS in the AM which periodically checks the liveness of AM UI and redirects to when AM exits +Thanks for the updates +Actually what would be really great in terms of reducing calls to the redis server and ensuring the correct database is selected woul dbe to somehow get the connection pool to test connections on borrow using the adequate select command instead of ping +Attached JMX property provider should now know about the complete set of JMX metrics available to us +Fixed the system property issue replaced with +there might be lot of files hanging around but they can be created in tmp which will get cleanedanother option is ashish is implementing session level temporary tables in metastore +Thanks! +Inclusion in JBDS will be handled via +Still present in Cordova this happen on a real device? Emulators dont matter! +Ive noticed that is now flaky under windows + +Thanks Evgeny for your help the problem is fixed +Anything left +Thanks Tim +reply again to one of those messages this time though send it to me instead but change nothing else about how you reply +Thanks a lot Ashish for your contributionThanks a lot Nicholas and Brandon for reviews +Full debug log +I hope IE is fixed too but I didnt test. +however i did manage to reproduce this just once on my Fedora box following the given is more a windows specific problem as its not always reproducible on Linux +This patch includes the previous patch and adds some testing of the first composite key class Bill +I think a good way to implement this would be to allow a client interceptor to initiate a request and await a reply it can use either the requestreply type or some hidden identifier to establish that the reply should be delivered to the interceptor and not a waiting client +I want one +They were here just fine in my local but somehow there was no log of them anywhere even when I started digging through older revisions on +For can we get it done in since it is singular at lease the path for tables +Just skip execution for such module +Just committed a test case which seems to reproduce the issue in rv testsrctestjavaorgapachefelixdmtestintegrationapi new proposed patch which allows to propagate adapter as well as aspect service propertiesAll tests seem to pass OK including the new +Thanks +packages Id rather have everything private by default and only stuff in +When I run it it deploys to the tomcat server and shows me a listing of the root directory where I get the css img and jp folders +If thats of interest I can put that in the mix too +However langc dist does not copy the final artifact to the distcpp directory where the release artifacts are meant to be collected nor does it build the C API documentation and copy it to the buildVERSIONapicpp as was done before +And users can specify share in terms of a physical core +There is meaning to let it authfalseAgreed +Shelly I was able to reproduce with the EAP ER build +ll take a look +HiSorry for the delay in replying I did not see this comment for a few am not sure why you want to have this code or some variation on it included in byteman +Thanks AdamIve removed the entity and two fields from the reports that couldnt stay there +I do now config my deployables in the deployer section and no more in the container config section so deployables will now no more be deployed on startup of conatiner see appended patch +Committed revision . +To find the next block pool to scan the datablock scanner iterates over the set of blockPools +Any word here? Has bumping up the memory worked for you +Thanks for reviewing Daryn +If adding the list included making a hit all the way down to the update sequence btree Id force the overhead to the client instead of including it in every update notification +In the future if we wish to switch to a different repository implementation other than we can simply provide a different wrapper class called +I tested this and it works great +Moving feature out of +No failures +updated the review board +See for forward secrecy for other key types. +Well Haruhiko If your assumption is right you should be able to verify it by creating a webapp and a small script that sends a lot of requests to that webapp +This certainly uses more memory but in the memory usage in the JT isnt as big a deal right +Updated the patch for SP and combined the and patches into a single patch +I am concerned about the DB version of the since it appears you moved functionality into the constructor of this persistent class +I guess is configured to listen on and thats why it logs this messageProbably just the logging level should be changed to DEBUG +I see RWCCs beginends that are in trunk but not in +They are now starting with +Fixed in revision Jukka can you please merge this change into the branch? Thanks +just close it cause redundant with. +If you between step and just reboot the database using or higher step wont fail +You reopened a jira and made no comment! Please provide more information on why and new steps to reproduce +Hi BerntI have looked at the patch and I think it looks solid and the test seems to cover the relevant casesCould there be existing applications that somehow rely on the incorrect behavior? If so should there perhaps be a release note? to revision +These blockers and critical issues are resolved but not verified + suggestion hereCurrently we have a crazy reuseable reader in +Are there any plans to release a fix? +false alarm What was cant really be used for this +It seems to be intermittent +just committed this +The patch is fine +And the serialization logic was a copy of that class mentioned above +FellasDo you want me to clone this issue off as a Docs issue so you can Resolve the issueI can do the docs elements on the cloneCheersThere was issue with JNDI and dedicated backup but it was solved wit commits and into jbossas repository +as per upstream code released to +ThanksCould you run the unit tests? Hudson did not run any test in the previous build +This version compares classes rather than class names adds some javadoc and improves logging in the test case +Basically I am just looking at showJobs as being equivalent to showPools and hence should have the same synchronization constructsIf you are modifying the patch to call get and introducing the synchronization around the scheduler please make sure that theres no inversion in the order of locking between the scheduler and job tracker instances +release bulk close +After talking to Konstantin we decided to create the throttler on the fly for each file transfer +let me fix them as part of nd iteration +Cool to see the positive process on this +Ill wait until Sunday to commit +I will make a test +Ill likely it when I implement +Solution with custom properties component overriding +However please keep the source for the tests as close to the originals as possible to ensure the tests remain accurateIn terms of the requirements around code code coverage the jts exclude of remote is the thing that you would remove from the codeCoverage profile so that these new junit tests would then be run under emma +please fix it in RN tooThanks +These two constructors are for testingThe other constructors accept Store argument which has reference to +Part of the fix is to provide helper to check for the form contentType to get the desired FORM might be better +I think the problem is coming from the fact that the attribute declaration is missing the type attribute +Thanks +Could you please add more description to given that this is will be used for annotating the entire code base +Browser traceback after running a flowscript sample with the debugger enabled +Getting a test please change the javadoc for cacheKey parameter +The JSF version number showed incorrectly for one of the because it looked at the resource bundle from the wrong second solution doesnt sound good either since you have to configure a separate other ideas +This is going to be a pain to testreproduce +Its really annoying for our users who are trying by default to do it from where the jar is +everything looks good except your last bullet Context must be marked as for and as Unstable +All elements in the JDT Compiler AST that can have doc comments attached to them have a javadoc field +If you disagree revert the change and reopen this issue +Please reopen if you can reproduce the problem with Xalan +Net FrameworkI use ant to create them after adding the lines to the file to tell it where to find the key file. +actually the patch before did not work for tarballs +Hi MandanaThats what Im doing right now but as someone who has a workspace with projects in multiple environments its nice to have the IDE switched to the correct perspective automatically which is currently not the case with X +no the white one is wrong +Moves the default setting for NMLOGDIR to the file as per Vinods comment +Well if it breaks those unit tests it is not good +I committed the patch to trunk and and +is used for different purpose +Ignore the closing stale resolved issues +I not sure if our put and get style capabilities are actually spec compliant +buildivylibcommoncomm +I added a EAP server and various other IDE setup activities and now when I go back in to the add server screen I see the WTP servers available + tested! +From a users perspective and for a coherent platform standpoint i dont care if this thing is a EJB a MDB service or a CDI Bean +Logged a few Jiras to be done for another release in the next weeks +Is this issue really specific to SOAP? So I could use the soapAction attribute alone in a WSDL for a SOAP binding and it would work +bulk close of resolved issues. +Im an idiotI didnt see the setting later in the file +I am reopening this JIRA +But it seems it works fine through SOAP even though the WSDL is not working +I test with trunk and it worked fine will test that scanrio as wellThanksDeepal +It is attached as metadata to the artifact not as an attached artifact +Clarified docco the reference to was wrong. +This problem does not occur in version of +Reporters of these issues please verify the fixes and help close these issues +Theres no new test for this otherwise looks useful feature improvement for snapshot +Had a quick look at the patch overall it looks good +never mind I get it +Submitting it again to see if it is flagged again +pull request has been created on github +This is a libevent topic and not Windows Port relevantI use libevent on Debian Squeeze and this version does not provide evutilsockettfixed that with the following define within libcppsrcserverdefine evutilsockett int one of the following defines should control the evutilsockett support which one which versiondefine LIBEVENTVERSIONdefine LIBEVENTVERSIONNUMBERCould you provide a updated patch without windows headers and with an updated evutilsockett check rule +No code was removed so that if the underlying issue is fixed in a future version of POI we should be able to easily bring it back +Committed to with revision . +Im looking at the perf setup currently with regards to a couple of other around this and that might help understand what the problem is +Patch for xWork to report OGNL errors +for merging that fork +I took a look at the +Are you saying that it is not possible in mighty SWT to set a minimum window size + +It happens because the value in the source tree is no longer used +Attaching patch a with a suggested update of the name and description of the Maven artifactReady for review +Publishing never got a error message and was alwyas silently logged +Andrew does unit test pass for you? Fails for me when I apply this patch +Can you expand on the difference between a Prepared Statement anda Logical Prepared Statement? And also on the difference between aConnection and a Logical Connection? Are these concepts in the client? Or did you introduce these concepts? Thanks +Should be not edu +Fixed with svn revision +the extra config field do you mean the ESB runtime selection field on the last ESB project wizard page? the field is used for the user to select an standalon ESB runtime if we dont need to support the standalone ESB runtime we can remove the ESB runtime definition preference page and this ESB runtime selection field as well +Trying out the examples seems to think well need to keep a close watch on the use of the Ordered interface +Applied svn revision thanks to add the right Fix Version +Punting to +Sounds like the right thing to do +I dont think I had a problem accessing the database because if I changed my schema file I could see the results reflected in the database +Would it be too much trouble toask you to upload the castor jar with the patch appliedI am happy to test it against one of my applications +Here is patch to fix this +I also tested against and get the same wiki markup to preserve code formating +Have you tested on any other devices or simulators? There are similar reports of this issue on this simulator being solved by running the code later in your initialization +Incorporated Toms comments +Can you we use that? Make it more prominent thereOr add a simple Servlet that from any master will return the URL of the active one? Would that be good enough? Would be an extremely simple change. +Surprisingly this test seems to pass when running with Maven and fail with Maven +Thank you Brandon +The true solution is to pickup only jar file not other at revision and +Ah okIm getting old +Why dont you just use the CMS and be happy? It supports exactlythis kind of layout with +Feel free to submit a patch with unit tests attachedCheers +Do I need something else to add this attributes to tlds +Closing issue as it has been released as part of release +I workarounded some issues coming from calculate I will do to add them back +Patch attachedThanks amila for the contributionRegards the resolved issue. +Do you think you could make the pig contrib project get the dependency w ivy +All squared away now thanks very much and sorry that I didnt update this after getting help on IRC. +Im not sure about but for the way we enable this is to allow the user to configure in their own Invoker +So where is the list of source directories hiding +The correct way is indeed without the trailing spaceThe problem with Spymemcached was that once I removed the trailing space I also removed the separation with casunique and so some tests failedIts all good now +YesKeeping multiple versions in the computation is exactly what I had in mind +Done making the changes in trunk +ve turned this issue into a request for compatibility with the WAS JPA support rather than actual WAS +it still merges into trunk cleanly + incrementals are now on slim builds but we still need to generate the EAP overlay +Im not sure what the correct behavior should be for here though +Thanks guys +This should be fixed in incremental build or would you be willing to try the latest incremental and confirm it works for you? You can install it with lein immutant install just tested it and it works great +Will attach a patch soon +If the problem persists you can reopen this issue. +I eventually found comments his change in whats new but I had to do a lot of searching before I found itOur current workaround is to copypaste existing definitions in the file manager and then editdeleteinsert in Eclipse +Did you ever find a workaround? +Sources of Mina HTTP forreviewed the patch the code is ok but miss the Apache licence headers some javadoc comment and unit testIll do the integration after the release. +We may improve this laterAt last I remove the UDF because it is the same as +Each lookup creates a new initial context and does a lookup +Have tested with jbossas branch and is working for unified invoker ejb calls. +Logged In YES useridUnfortunately the web console is not really usable becauseevery time it requests something from the missing the tree view is reset all the leafs are closed +Added in +Thanks for the pointer on this Erik +bq +No I dont think its struts I was hoping it could be something like that too +Thanks +C support has been added back in for +That it works in and is a good thingWe havent changed anything in this area in a while so its a good thingWe might need to double check on the line thoughBut if you havent migrated to I encourage you to do so anyway +s sentiment path exclusion should be a standard feature of mapping +Just wanted to add that Id be interested to checktest the hibernate port +If it didnt want to do any searches and simply return Hello World or add a couple parameters together it could +Borris sorry that more comments are here +Had to revert to plain Java instead to make it work +Ok its not a bug +plugin jars no longer on classpath for +nested loop is unnecessaryRemoved that +These concurrency issues are addressed in a follow on JIRAThe data from the history files could be stored some place else in the future but it will always start out being written to HDFS +Maybe a test which continually openscloses a region while making edits and after every close verify its not in the map +We are much better in terms of perf now than at the begining of this and shoudl get the same perf level now with an instanceof removed +whatever the framework can do using the recordreaders new interfaces the recordreader can do itself +Committed at subversion revision and ported to the branch at subversion revision Touches the following fileM javabuildorgapachederbyBuild +Nigel Daley notes on the listNo this seems to be a bug in the patch process scripts +Ill go change it nowPeaceScott +With committing them to the main line directly there is a risk of main line getting destabilized +This has been fixed for some time and especially it works in the +I just need to add some unit tests +Now it will work even if you dont have path to but you have JAVAHOME system variable +points to the groovy executable this issue is about getting the path of the groovy script like perlrubys FILE variable +Jody see how the port has been in postgis for an example of how to give it a default value +Are we doing anything about this? Or are we done +TedI just put the else for symmetry +Oh really? must have been using the search tool wrong because I couldnt find themYes sure Ill get onto that now +Then the script will be simply evaluated +Updated README +I believe it will stay at until it has shutdown all active bundles with start level +Also why the change to the classtype check? Its not wrong just not as flexible as before +which did actually cause a unit test failure but went unobserved +And just to be clear embedded Derby and server started via API are not affected by default +Please wait for my refactoring to be committed in +Just to clarifyBy meaning new format and old format we mean that if znode is not available after ENABLED then it is old and if the znode is available after ENABLED then it is new format rightI can understand how the master will make use of the configuration +Just noticed that youve attached the demo without granting an ASF license +The result is that if the regionserver crashes before flushing memstore to a store file the delete is lost on recovery +The way to do it is to use black pixels outside of the logo area so that the outer area will be stretched +Apparently Im a victim of an overzealous spam filter +Mriduls suggestion looks good to me +r. +Hithe problem has been solvedWe are intrested in bugzilla as it would reduce the time taken to transform the xslBut i dont see much better performance if i compare the result with XT transformerfor example if XSLTC takes ms to transform a xsl and xml XT withoput caching takes aroun ms and XT with caching takes mswhat i feel is we can improve the XSLTC performance if we could cache the object and use it intead of everytime creating itBut i m not able to do it as i get the result only if obj is created after u set the propeties for like writerencodign there any workaround for reporter of this issue has commented that it is fixed in an earlier release of XalanIf performance is still an issue for you please open a new issue with at testcase that we can profile +this isnt all that important at the moment +While my job is running I notice a bug remove the old cache file push a new one to hdfs and then launch a new invocation of my job +With Hessian it was failing but with Hessian and Spring hessian is working fine +Mark this seems to be working fine for me? It sounds like something isnt getting cleaned up from one of the previous runs +Still need to test with latest trunk from hadoop +I have just committed it +Marking as resolved since this issue is fixed +Patch looks good +Closed after release. +I also get the version numbers from the script just like the autotools build scripts do. +We have too many of those +fixed in revision Anton please check that it was applied aspatch is applied as expectedThank you +I think that may be the right solution +Found that problem existed in original implementation +Fairly straightforward patch +Fixes bugs with replication +Thanks. +A bit of experimenting showed me that currently we support the change of an existing property for a static import +Sorry the class is actually but not +Jun Aoki apologies if my request was seeing as asking for additional tests I was asking to do a test running the examples to verify things work +It is probably a good idea to change the default sample to show the support for this feature +Fixed the output handler to break up with special characters andto escape the special character +Pull request submitted to master only +Incorporated Dougs latest review commentscan somebody from dfsland please review this too? Thanks +It would be nice to have an option to fsck to delete the last block of a file if it is of size zero and does not have any valid replicas +The issue has been resolved and will be applied to the next release + +Brilliant +The code is still also statically inside the tag files but the first registration winsRich can you review this change and then close this bug +So far I havent found a good way of leveraging them in the implementation +OK I will add a new option UNICODEAWARE which will switch conversion ON +Ive added corresponding clarification to the reference documentation +It can be a regression or jenkins being under load +I guess using styles forces you to disable the global services effectively +That means this feature is pretty isolated and implementation must leave othre the +I suppose that if everyone else who runs into incompatabilities like this is a Java developer and wants to learn how to write Mailets then that would be a second optionIn any case we can agree to disagree on what solution is best +Why was this bumped? I am under the impression that using it is just adocumentation changeSent from my phone please pardon the typos and brevity +Completed the code removal documentation changes are pending +modify +samza extension +Complete in master rewrite branch keeping as patch available until merge with trunk +ChatreeThis sounds already quite easier the component part should be abstracted +Also can you start working with CXF please? Quite a few fixes went into this specific area though I can see this is still for the info +Have you come to any decision regarding the fix version? Will this issue also be fixed in the branch or event the branch? At least applying it to the branch seems quite important to me because will probably not be releases that soon right +dont think I can assign to myself but patch is +How would I have known? Where is this documented +I had submitted the patch to Hudson to ensure indeed we are getting IPV address +Whoups the title should read Resources versioned in custom environments even though it should be disabled per default +If we want we can wrap some of the common access patterns to a utility class but I would avoid introducing any more indirection than that +Its the same patch that I sent before but it also removes the specs tagged as my latest patch broke some external tests +changings added bundle version as added exported package version as added generate folders to svn Bram I dont think it is good to specify the version in packageinfo because mostly its just pick up the version you specified for the packages in the appropriate Bndtools view +Yes Id like to have confluence +Thanks to Myrna for helping me with this +The issue is dependent on +Moved back to has been releasedMoving all related issues to the next version. +You may also checkout the code and see for yourself. +The patch has removed those files from the build and replaced the calls with direct methodsI will make change the versions to Apache hadoop +Only the hinputText Fields convert the umlauts with every Submit +Closing all resolvedfixed issues of already released versions of Roller. +Verified by Ilya. +custom action would be good IMO +If Mahout integrates against a different version of Hadoop then the wire protocol may be incompatible +And yes well continue to look for a better solution +has an additional test and is updated to fit the current code base +So setting CACHENONE makes things somewhat worse +Let me understand Jeffs patch also and the idea behind it +Raising the priority since wont come out until this issue is fixed +BJDo you know if a person is offcially doing something around the email stuff in +In the test command you pasted I believe needs to be replaced with to actually get it to use theIAC does it make sense for to make use of config objects returned by the getConfig objects so that similar changes in the future dont break unit tests +Unfortunately we dont have direct control over page layout wed either have to insert new content to push it down onto the next page or remove content so it came back up onto the next one +Upgraded Teiid jars to latest ER was released. +Changed implementation such that a session is always created using an admin session in Rev +Just have to get through an interminable conference call first +Not sure what you mean by explicit routingI mean where the user has placed a prefix and ! in front of a key value +Good candidate for merging to branch +It seems the exception occurs when the JMX elements are combined with a chain +bq +until you manually activate that refresh the project examples section in Central or restart your with +I thought we arent linking features to releases now until they go in since releases are now not +Id like to commit this unless there are any objections +I directed the request to the tcp mon and I could see the wsaAction with the correct namespace stuff without a problem +Thanks Eric +This seems to fix the bug +The usage looks really flat. +I can see that it would add more methods and make the interface look more complicated but log are nice ways to generically implement some features for some clients +Agree on +There is no file any morer pchakka users Sep pchakka users Sep should have mentioned this in the patch comments +Also can you give me a use case for why a user may not want to see error markers on files? It seems to me that they should always see them if the groovy files are part of a build in a groovy project +Closing all resolved tickets from or older +See resolution in comments +Without fundamental changes to the codebase is there anything more we can do? The answer may be no +We could look up the action config in the mapper but then it would go through that code twice +Youre right my bad +patch committed in rev I will close the issue when I will have add documentation +Factor out a simple in API similar to that proposed in the future support for RFCBetter to use an object rather than a Locale so that multiple user and client preferences can be easily added +Ive updated the patch to work with the changes in +Thanks Dushan and others for all the help +Matt the failures is not related to this patch in HDFS trunk failure case +Bug has been marked as a duplicate of this bug +this patch should do the MatthieuYour patch is in trunk at r +Previously the generated PDB files for each project were both named causing one to stomp on the other and making it difficult to tell which PDB belonged to which project build +Collation only applies to character string types +Some of the specs are segfaulting on Ruby on OS X Lion via RVM using bundle exec rake gem +supports itthe changelog for does not mention adding any new feature like thisAnd versions before are not supported by anywayI tried now v and it also supports the call next value for seqname syntax +Im not confident of the Quartz clustering since the first time I enabled it I ran into a bug +This code is one hairy bastardBtw nothing personal Koji +Hi Benjamin Thanks for this report +Thanks Lily and Kristian for this collaborative effort +IMO we can create a new project with a basic implementations +any chance you can upload your model or repo etc? +and I rolled back to and now it works +I agree with grant that this syntax is more clear then using for each bucketJust throwing it out there +Actually this problem only occurs for the JAXB RI binding so it should probably be in the databindings subproject +Looks like a view +Helloystein youre right the patch was relative to a trunkjava folder +So lets try to keep the chatter to a minimum and put class names and method names etc in their own issues +I have the same problem and upgrading to version made no difference to me either +NB the custom JAAS module should also import resolutionoptionalto be visible in Karaf. +Appreciate your email contacts re +Thanks again to Rhett for the patch and sorry it took so long to apply! +Tested with hadoop and hadoop and noted that the new parameter gets written for hadoop and above +Commit changes to trunk at +Highly appreciated if you could have a look at related issue +The previous was for the AS classpath +This is related to adding win bit to list of supported platforms in as is for +I will check some points before committing +Yes this is an oversight of the initial fix. +Multiple corresponding to multiple columns may be in a log entry now and not all columns may have the same scope +If you find an issue then it can be revisited. +This is a documentation only problem the actual code in the +thanks for your effort. +The OO example is only one level of extension +Ill test this an commit today if all is good +Would you confirm and report the issue has been solved +Hi John I couldnt see the deprecated tags so I just added them +Looks goodId suggest adding unit tests for +Merci +Since this is JVMTI problem in DRLVM Im changing subject and link this issue to DRLVM component +Thank you very very much +Fix for the branch +Thanks Raghu +Made the modification and added some javadoc to the delay timer classes +Hadnt looked much at scepexe because its not scp Im seems this could be tweaked to work for me +I ran your extensive tests with my changes and the tests worked fine +or Maven will play havoc with your version ordering! +The last few forgotten copies should have been fixed now I hopeI have also added a few more constructors using and improved test coverage which lies now somewhere between and for this class. +It seems like what youre asking for is something more general than this rather than allow just one ordering of values within a key they can order values in many other ways +I think you could probably make your own version of it and jam it into the framework through the xml configuration + added +I dont know how other database handles this issue +Looking brieflydesruisseaux in Confluence is not a member of geotools groupsThis is because they have not set their Confluence username in Xircles hence no mapping between the two that is done +BTW patch attached does also contain test code +Can you verify it again? +looks like problems exist in org plugin there exist folder resources and in this folder exist file +Also add a component for work has been done in the servicemix web consoleThough features are currently very limited +Is it repetitive? If you open the server editor for that server does it have a runtime? If you edit the runtime does it list a home directory +Attached patch that makes public +This needs to be committed will be momentarily +This is an problem not ours +I put a redirect in place +We cannot trim that +Screenshot showing two different problems in here I believe Different behavior in from Ganymede that im not sure why that would not include starting jboss since that is what done for an app +Fantastic! Thank you Juergen cant wait to start using our new projects with Spring and Java david +but if we wait for it then it may never come right +No tests added as this is a minor change to a log statement +Barring other commentary Ill close this as +do +Either way using handleFaulttrue doesnt change the outcome of the test cases +How does that differ from the Down description? Thanks again for this +Unless we implement joins in query there seem unlikely any efficient way of counting the number of features +needs to be reviewed for as a regression or closed if this is by design +Thats where we step in and make it work +WSDL used for the test +committed to trunk +By default assumes that a file is from HDFS and that was causing the exceptionThe solution is to the entire path to the which allowed to find it in the file system +I defer to your judgement on thisSo I assumed that you are still not convinced +Why do we need more than one datanode and nodeManager in a single node setup +Without the patch I thought and will fail according to the other comments on this bug +New build is available for and +Either will die or we will modernize it and release a non binary compatible version +Marking closed. +Ive downgraded this to minor and allocated against really does have issues though its only really evident when demoing as thats the only time youll instantiate simultaneously twice for the first run +Looking forward to +Closing. +assume fixed since no further reporting +and were all related to the same issue which has been fixed thanks to Simones patch. +Our vision is to build a full data life cycle management stack that encompasses data creation notification consumption retention and security management etc +yep same as other watchers seems reasonable +Fixed in trunk and +I just committed this to trunk and branch +While not strictly necessary we could also add timestamps to entries in the image and edits for validation +We usually review patches within days +committed the patch to fix the potential bug source +But when I use it with Xalan it fails +It seems betterPlease check the +Adding the new snapshot as a compatible version and will rerun the EAP CP testsuites for both and +Does it make sense to make the regex configurable that defaults to what Swapnil suggests +Fixed with new gem in acf. +Early version of the minor note currently Create and Delete snapshot operations look almost identical +Should I be looking at another bu that has the info you mention? Or is there a log of the IRC chat +This has been comitted to the trunk and the branch +Thx +Should it have been backported? If so please create a new issue to fix it for that branchThanksGail +statements +Ill get this in once bug has been committed +Fixed at revision + AndreiI would be ok with this patch if you moved the functionality into the protected computeAction method instead +Apache headers testcases some warning fixes +Thanks you Hari. +Thanks Todd! +Fixed in trunk and Branch. +Can you please submit a trunk patch? patch looks good to +batch transition to closed remaining resolved bugs +Changes introduced to by add in to your diagram so can contrast the two pictures? Before and after +Attached are html versions of the updated files associated with this issueI would like to create a general topic that explains how Derby handles collationWhat I need is advice on where to put that topic which is the best book to put that in +Check r and give feedback whether it serves the new need +This is no longer an issue due to the way that apacheds dependency is specified. +It would in effect intercept all the calls to the filter +If its OK then Ill begin work on moving all the accounting documentation into the Docbook formatThanks i did not know about the first patch however my comments are also true for he second patcheven is pretty old because i worked heavily in that area the last few daysthe format of the xml seems to look ok thoughRegards Sharanyour patch is Committed revision one comment though rather than empty a file and create a new one rename the files if you are using eclipse +looks like it is compatible +I have the configuration and parts done +Thanks Ben +isnt that what the autoSubmit parameter was for? thats how you described it in the mailing lists? +It looks like the lang param works if a request is made to a controller but not otherwise +Any chance Mohan can provide a fix for this +OK I think its the sequence in which the test cases are executedBetter to fix it as its the difference between Java and +Added ftest infrastructure and web services ftest +Updated patch +dims +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +bulk close of all resolved issues prior to release. +gif file was passing through a text filter in the build +Assigning back to Mike +mapping of usernames to apache import into main repository mentors have commit now hopefully they should be able to give access to the appropriate people. +I think you can remove the check for taskInited +It is now called RTNREPLACEIMMEDIATEVikas +and it wouldnt affect the model as far as I can ll attach the code as soon as Ive tested it +The method get was implemented in the classes and the patch to the branch revision thank you. +Waiting on +Yes exactly +Just used a default path instead of +Have you tried ulimit s +Defer beyond beyond release +Thanks +Problem is that we dont want a hacked solution +We cannot predict the future but we do need to put an upper bound on how long we try to do something otherwise there will always be corner cases where we can get starvationIt may also make since to use some statistical heuristics in to try and give up sooner rather then later if someone is asking for something that is really outside of the norm +Verified by Alex Smirnov. +The same rules apply also for releases that include such externally developed components +Which sucks if you are building every hour +Thanks for the help! Guava docs indicate it is on bit machines that this code really shines +Bumping up priority +This one I currently know about are +If you want to call the response on your own with your own infrastructure +I made a few code changes that need to get merged into AS master git repo +would automatically downgrade communcation to marshall call between incompatible class loading contexts +Implemented in revision . +I ran all unit tests +Yes is obviously needed though we ran heavy loads for a week or two before it first failed +If the traffic is high so that the time it takes to fill the log buffer is normally less than the timeout this means that log will only be sent when the log buffer is full +Its been determined that the efforts involved to solve this issue are not worth the time without a profile notion this would be rather complicated to set up as there are some dependences on thirdparty as well. +Changed Target Version to upon release of +Note the user will have to set the location of the rat +Would the same apply for Command line as well? mapred queue info +vmnum bytesCRC MBsecCrc MBsecCrc MBsecCrc MBsec MBsec vmnum bytesCRC MBsecCrc MBsecCrc MBsecCrc MBsec MBsec +Patch patch was applied +Fixed that +Note that Xemopt should be specified as the commind line option +bq +I forgot that we create different dir per test JVM +This test fails for me sometimes too somehow we should get to the bottom of it +Keep in mind that this code is only executed in mobile environments or using some renderkit extensions +Suggestions are very welcome +So you still need to use JDK or my patch for JDK +Ashutosh Chauhanthe problem is reproducible on trunk branch branch that set truethe patch is for the version that uses and so branch branch and trunk they all can use the patch +This might cause some unwanted results and is very misleading +Please close this issue if you think that is the right thing +I fixed the build failures on this morning +As for the deprecation approach I think this might be overkill and a pain to addtest a ton of in the code +This makes DV unsuccessful +But besides that nothing really speaks against that idea I you dont serialize the Closures then what is it you want to serialize? You would be fine with an empty Expando after deserialization? +Ive applyied the patch and its working fine +Thanks Omkar +posting relevant ant build scriupt snippets also could help +Fixed in r r +Marking as fixed. +Aaagh +This should fix itTrunk has the surefire set to always for all tests +Nope reopening the issue +Note that very same container also used for runtime classpath when running application or tests +The ant build can work with that structure just fine +Uploading a patch that passes the attached test by Alejandro +Bulk closing stale resolved issues +The WSDL is You are writing an STS +Theres no to the broker unless you have prefetch set to zero +All dont think this is fixed see attached screenshot +nice +Patch is fairly involved not sure if more reviews are necessaryMoving to common makes senseDoes it make sense to compute the epsilon +any updates +It is used at startup when the datanodes sending the blocks to the namenode +Changes went in +Im using mvn help on the child POM and seeing the behavior I expect where the junit dependency in the child overrides the scope of what is stated in the parents depMan element +You executed in mode +Will close for now since there is no clear action and because we are moving issues gradually over to Github +Ive tried with the default java project and one cannot run the Open test action either +Extremely impressive +Given Remys comments and Charles of JON had similar ones I think Im okay with this +I tested it and went through the Chromium issue and Im going to have to resolve this as Wont Fix since this a Chromium issue and not directly our own +It should show your custom error message page +Im afraid youve missed my point +I made a mistake to fill the fix version invalid to This issue is active for made a mistake filling the fix version invalidThis is a active issue forWhat I meant was that in this issue worked correctly and from onward this issue exists +Should documentation be updated for this enhancement +After Matts comment I think there are no more objections +Final patch attached +the message isERROR Error processing element CLASSNOTFOUNDim just using a simple eclipse java project with a testing class with main and a custom appenderincluding the in classpath. +Oh if you insist +It needs to be talked over by the we decide to realize the new wizard we can reuse the new wizard codes for this for this issue the easiest way is to fix the log error but when no dynamic web project in user worksapce we should disablethe Next or Finish button in the wizard and give a warning message we need a dynamic web lets wait for the inputs from the team +I removed the slash in the packaging configuration and ran the packaging task again +I would prefer a separate jira to do that +Take a look at the OFBIZ jira workflow I added a Submit PatchPatch Available workflow to theirs +If we want to have possibility to seamlesly port applications to Grails this is musthave above trick to attain bidirectional association works great +In it is part of the calling bytecode +The problem with using a concrete filter kvp parser for the given wfs version is that since you have to feed the spring beans with the service name and service version they apply for when you submit a request if the request shall contain both the SERVICE and VERSION parameters in order for the proper kvp parser to be catched +We wont need this any more given. +Going to close this one since fixes it if you use the browser history. +The problem can be overcome by using Suns implementation along with. +I am interest on remove labels and compare version +Marking it closed. +My comments for are applicable here as well if such a change is introduced then an appropriate environment has to be set by the tests beforehand +Jobs arm twisted Oracle into releasing and supporting future Java releases +I believe this should work now I will test it out +Tests passed cleanly for me on +visual element can be retrieved through getters closing as wont fix +This looks good to me +Unittests +Reported by another customer +i got it thanksIs there any way to achieve the behavior i need? would it be possible converting the to an to get this or is also not a buffer? Or do i have to create an byte array which reads the stream until EOF and handle with the byte array? +And this note from confirms it! tatu There are some potential problems with this Escaped chars are not handled +Fixed in apache trunk rev This will eventually make it into the next major FUSE release +by the waywhen I say I run the application in local machine I mean run it on the local file system instead of from a web server +I think PKCS has no problems in both Linux and Windows. +The running distcp if it couldnt find the original file would look for the same file with the +Also following on from the Pranav suggestion for now can we not just let the scanner exhaust the current row only against whatever was in memstore rather than let the scanner run to the end of the region and then when it reaches the end of the row then let go of memstore reference? In other words only at row junctures update its memstorehfile scanner set +failing with +Thanks Chris! +The other stuff should be a new issue +That explains why I was not able to duplicate the issue in my environment which uses only a single is most likely a problem in and not something directly related to Railo or the Railo Installer itself +Please advise +New patch to follow +i took by random choice could easily be checked against other dependencies +Please review take checked in +qt +Adding parameter depth works +Not capturing the proper exception at the client side is a problem with and has been fixed with the current axis trunkAs I have explained stub code is generated using the wsdl and if there are element name as with the fault element names we need to change the class names to avoid compilation issues. +We can update the docs to reflect the changes +Good idea +Im taking a look at the changes now +Ooops I just remembered that the list +So Ive tried to change it into one like with no luck +I had sent several messages to derbyDev about two remaining tasks and this is one of them +I tried this on Seam DVD Store demo with EAP and what happens is that the browser window is only opened once you remove the deployed earSteps to reproduce +Soo keep it up. +I just noted the checksum failures in this last test +A that shows the problem +Hi Reid sorry for the delay but thanks for your contribution +I will post something to the mailing list to get this issue +I will commit this tomorrow if nobody disagreesI will also have a look at +As soon as the guide will be available on apache server I will make clear on that the project is moved away +New patch reverts public method to private and adds comment explaining the simplistic removal of the semicolon. +I see it now +I envision this as a hyperlink type control in the on the actual state name next to the service name +fix verified by submitter +I didnt see this since it was not set to be fix in cr +The idea is that the embedded runtime can be updated without touching the other code +Fixed both the issues +It should be patched with binary option because the file also reflects the changes for binary files +I would prefer to delay stuff like that until we have more test cases and we see patterns emerging +We can outline those and get to agreement on the basicsJims request that a file exist if its producer dies before a close sounds like it would be easy but then we get to flush semantics +I decided against that since the URL can grow very large if you ask for many nodesThe other parts I can look into but Im not the java expert so if anyone else feels inclined to do that feel free +Ill have a look as soon as I get around to it +Links in a filesystem are normal +Closing issues which have been released. +There currently is no token just userpassword +Checked in a fix into develop for the font size issue was just an issue with the mobile theme CSS +last release seemed to work smoothly +Its kind of like a password to get more data +Resolving issue. +Sorry for the confusionCan we issue the problem as an error instead of a warning for future releases +Please try +Running tests +it looks like the log dirs may need to be created with the correct permissions as well +Thanks Raymie! +We should probably add a test case if possible so this onedoesnt pop up again +Using task counts is not sufficient +Is this ready for review? If so please click submit patch button +Seems that the file is cached and so even when its refreshed from SVN it doesnt get into the build +Thanks +Ive moved all the metric examples into the Analytics docs +We could do that as a step toward the surgery Gary suggests +When I mentioned others getting involved with it I wasnt worried about Hans not having enough time more about Hans not being able to read peoples minds or have access to the business experiences of other peopleAbout the thing I didnt mean to imply that such a thing really existed +No problem +This issue can be hit when we update to the offering which does not have the services of the initial offering +Woops +I dont think parent versions should act like dependency versions +to postpone that to +It was suggested to me to open a JIRA ticket as no one could help +I was able to reproduce this bug in JBT but it works in trunk. +Committed proposed proposed solution works so we can close this issue +Well yes the correct fix its +Committed to trunk +I think if we dont like booleans for this silly reason then we should just use an enum like the old APIOK Im attaching a new patch using a NO enum passed to and +The checkin to again adds a second parameter to the constructor which breaks compatibility with JEE +Is there an easy way to statically generate all the code from lombok and then remove lombok from that project +Updated patch Added info fro non inlined methods +The JPA one I suppose. +will resubmit to +Last time I did that Jenkins didnt pick it up +Sebastian I might have gone overboard when describing your process as esoteric +It would be nice to hear other opinions on the matter +docs are of a bulk update all resolutions changed to done please review history to original resolution type +Ill open a JIRA for it tomorrowRegarding the memory limit we would hit it for more than a couple of hundred million ledgers +Build now knows new property +On the first stage only plugins are copied to the plugins folder and file is updated +Like I said in the comment above I wont be able to test it until next week +Lets first get an agreement on the exchange +YesI agree that in response was not appropriate +This fixes the issue although it does change the names defined so it may break existing code but I cant see any other way around this +Committed thanks! +This looks unnecessary +This was addressed as part of the commit. +I am updating the patch for the record +Marking as resolved since we cant reproduce I really meant I cant reproduce on rc because a similar issue has been fixed in rc +I will give xstream a try tonight +Trivial no patch needed +May I ask whats the most worker tasks youve run on a single ZK this way +Marking closed. +applied your patch txs Gerhard +Added in r. +I can help working some on this +Because there is now way to asses if someone else find a way to create a real splittable gzip codec Ive changed the name of my class to The idea is that its clear that the intent is to make gzip splittable yet a workaround is in place that skips data and seeks out the start of the splitNew patch for follows shortly +i do have a latest patch for this jira +WillemThere are two files in the resources directory +I also think it is a very good idea to be able to configure it off at the namenode +This issue is very old and we miss a reproducable test case so lets close it +Did I say a Float is an Int? Thats obviously wrong +be warned that this is a draft only and was never tested +Its look like you are using form auth below? Can you debug and see if the is properly invoked for this request +The patch is in the fuse ESB trunk +My blog username committers who may be interested in bloggingjmhsieh esammer nerdynick mlaiNote These are ASF user ids and not blog account handles. +Or are you suggesting I create two silly wrappers instead of one? No need if there is no easy way then forget about it. +Marking closed. +categories added be committed +We have a large set of simulations and tests where zookeeper is a small but important part +Here is the missing stuff +Would be great if folks shed light on this +Thanks a lot Harsh for noticing itI will resolve it as duplicate. +Please that this issue exists I think some other fixes also invalidated this issue +I think its general drift is fine +What about JCR and +William Dekker Upgrading to Spring fixed the beans problem for me +Thats cool thanks +Code looks good +It will be fixed in the release but in the interim well put out a sepatate download of anything necessary to run the trailmap. +Im certain you are running into a problem with gcc and a new compiler optimization +Thanks Todd +I think its ready to backport to x +As I mentioned before if theres an error opening a resource in a restart scenario it will be thrown from open and therefore isnt skippable +Checked in to trunk +though it is fair to say its not a pretty patch + thanks +Someone was not agreeing to me because it was not the letter of the spec my main argument was ease of use and the fact that it is trivial to do +It looks like the code is trying to handle the problem of how to make a sort avoidance check at this intermediate point in optimization +Until then postponed to +probably can throw in logical assertions at the same time. + NoteCause The log message for an invalid state SOAP fault says that it is cancelling the participant when it really compensates the participantConsequence The log message does not agree with the actual actionFix The log message now reports that it is compensating the participantResult The log file shows the correct information. +One alternative is to add a code point so that we have a way to indicate to the server that an error has happened +We need a slick diagramming tool for this and perhaps it would be nice to agree on one and use the same tool for consistency across the site +Does anyone have any ideas on the way forward here can I assist in any way even if only testing on my Windows system? +Priority set to minor +Added capability for creating scripts +Thanks +Another day perhapsSo much to do so little time +committed to trunk +Im curious about why this is caused by a table only having delete operations applied to it +Do you see any issue with itYes +Yes Len +I have just committed the patch you proposed +However the missing ones were in Ecommerce +FixedThe space navigator is designed to include once not allowed for end users to add to a page the toolbox will not display the portlet +workaround manually remove the hibernate plugin +Attaching the straightforward patch to fix this issue +Integrated to and trunkThanks for the patch RajeshThanks for the review Stack Lars and Ram +The server cant possibly know that it only knows what the browser tells it +Use the correct JIRA type +I actually did open a new patch for this issue on trunk which is linked here +can use children instead of similary for +If further testing around shows that this is not the case this issue can be reopened. +bq +It looks good +The exception handling in the network server should be taken into consideration in +changed from resolved to closed +Hi Antonio I ran the test case outside Eclipse using your and +Maybe you want to looksee Jon +I think things would help in understanding this Have a trace flag that tells us if the prepare ended because of timeout or because of plan exhaustion +Cleaned up the github stuff +Have a look for a better understanding +Bulk move issues to and +An interesting issue is that when starting the program with r profile performance is greatly IMPROVED +r pctony pctony May r pctony pctony May r pctony pctony M May Essentially +I guess one use case would be to put in a map and not specify a comparator +Chang I am quite confused +Also move the wizard to be a standard Import wizard showing up when you select Import +Often times these files are quite large and the cost of sending them to the system is significantIn my own use case of the handler I imagine the policy would be the more helpful policy +pull request is applied but it is actually not fixing issue +I agree that it would be very helpful to add a sentence making that classpath information more explicit +this is a small it +Dave the CI build is currently running under so should apply to any changes associated with as well +reformated tabs to spaces rethrow an exception during instance need this to be merged into as change looks reasonable +Thanks Hari! +great stuff but I think it should be provided to the user automatically + +I know I have a couple +Thanks +However the warning message is easily get ignored and skewed join will continue with something very wrong +comments on good to me +Removed the old versions and reinstalled with the trunk solved my problem thank you very much. +We are looking into use instead of JPA to provide persistence out of the box with CamelA JPA is really ill fitted for a map based kind of repository and is used by and thus well tested and really fastHiram has helped getting the bits togetherEnd users can still create their own JPA version if they fancy using a RDBMS +A volunteer would just have to add a directory and copy few lines in the. +Updated draft after fixing typos and partial rewriting +This ensures that Camel can handle its lifecycle and rd party can plugin a custom provider for thread pools such as from a JEE server +Do you maybe have overlapping bean definitions using the same bean name spread out across your three context files? In that case later loaded bean definitions would replace earlier loaded ones of the same name +for the patch +Closed upon release of. +The test is test under +Looks good to go in +This makes the logic a bit complicated +If support is to be created it would be in a separate project either or Im closing this issue and hope to see it reappear under a different heading at some point +Heres an example of one that worksArchive Length Date Time Name filesWell adjust the system to be able to handle the bundle with a +Ive started looking at a port to Avro Ive done some initial hacking and it will definitely be a patch that touches a lot of thingsTheres a bunch of cool abstractions in that will let us do this a little more easily like the pluggable velocity templates + +Hi TihomirI agree partially with you +srcjavaorgapachecommonslang +According to you it seems I was on the wrong track anyway so unless I get put back on the right track Ive gotta stop working on it and be happy with the workaround +I think a separate issue would make sense +Its not a matter of finding the previous version +At least theyre now in the capabilities before they were there but not advertised so you had to know there was a rss shortcut +So I suppose itll rather require some special setup in Hibernate itself? for reproduction project as requested +Fix in fixed in BRMS ER. +Workaround write the password with a backslash before the quote caracter to escape it. +I dont see this testing against trunk perhaps you want to try with a later release +fixed comment and committed +Leaving as critical since this is a quality issue +amr +It looks like the fix for can also be backed out now because a placeholder for an autoincrement column or a generated column cannot appear in a result column list that far down in the node tree anymore +So far these are the modifications Ive made Removed the Ant Script to reduce maintence and duplication of dependencies +Submit the patch for another run +The default should be false since thats the default in Jaspers +In manual testing it was found to be a permission with the folder permission due to known issue with QNAP server here +Please review this patcha few issues arose Validators usually dont implement equals and hashCode either so does not work correctly if only the validator is different does not have any attributes that are not used in so all are compared but I think it would fail if two different subclasses of were compared +The impact seems to be that the capacity scheduler is actually initializing more jobs than it is supposed toThe fix at first sight appears to be straightforward too in the capacity scheduler +I think we can change the fetchSize to per default +NET web servicethanksdims +Fixed to fix the deprecation changes look OK to me too for the most part but the new files you have added are using tabs and the spacing appears to be off a bit +reopening test cases fail real issue is that it is possible to save a renamed node without saving its parenta should be thrown instead indicating that the parentof a renamed node needs to be saved as well +Im working on making changes to the load function interface as part of the types +will commit if tests pass +Uli Gav Still an issue? Or perhaps give me more info as when I created yesterday I didnt get a an issue +So I think it belongs here and will be used byI will post the corrected patch +Ok Ill download tonight and tomorrow +such a string field could be used for sorting as well as facetingAny other ideas for a good nameThe conventions we already have involve appending an s for multiValued dynamic fields and Im just worried that trying to do something like append a U for unstored or an I for starts to make names too uglyMaybe something slightly more radical is called for perhaps we should never store copyField targets that would allow us to use the same dynamic field multiple ways +Id fix it myself but the socket code is quite the maze +It was a deployment issue and I had notdeployed it right +This is fixed everywhere now +It must not be that important Reopen if you want it back. +bTable t is similar to t but without a BLOB column cTable t has an update trigger defined on it which updates table t in the trigger action +The changes made to the pending cmds are to facilitate a redesign of multiple commands wo constantly breaking hdfs tests each time a command is updatedTook out some of the TODO items related to standardizing +revision . +So if you dont have the time to work on this + +Any ideas on what could be happening? It sounds like is the only one that is modifying these files zoie only seems to be reading from them to make the backup +It is consistent with current behavior +No need for a separately named patch +Im putting it into STATUSBacktracking This isnt even in trunk yet! moving to +committed in the closing all issues that have been in resolved state for months or more without any feedback or update +This is deliberate as the version element is parsed along with the groupId and artifactId to determine if the dependencies are present so that they can be removed if the user decides to change the orm provider or database +Do you need feedback before May Im going to mark this as resolved because this problem has been reported several times with Axis C but we have not had any reports since the various fixes went into the buildsWhen you get an opportunity to test this can you then move this issue to closed or reopen as appropriateMany ThanksAdrian +Im about to commit a test using a jar file and a simple script that adds it to the load path and tries to require a file out of it +Is there any progress on this? Do you need any help to reproduce this bug? This bug is a show stopper for us because we cant upgrade to Ivy +Hibernate passes the Sun TCK for JPA +It looks that the problem has not been a hack in to get around the problem +Update the wording to add Jira number to the review board entry +With commit bcbaeecdaaddacaf has been removed and all MD lookups have been moved from the type to the. +Can you dig through JT logs and see if any such issue happened +The LF is used only in source code +Not sure if it is still a you provide a complete stack trace and some description of what your app is doing when this happens or even better a sample that shows the problem? Ive spent a lot of time looking at this and still have no idea how it can occur +The code looks good +to commitI dont have a working doc working set for the moment so if some other committer could commit that would be nice +If it is null repopulates the cached method from the propertySetb Changed to for the manifest generation +Work complete awaiting PR for request merged +Change message to indicate timeout +The DragnDrop should only be activated when the is loaded +HQ is another name to be need to test before closing this issue +This is important for QE verification and use of stage server in general +So it is also when the top project pom defines a parent pom such as an organizationalcorporate pom +Alexey the best way to start with this and the way that will help get this fixed in Lucene core is to write a unit test class that does what your code does with and and shows that the test fails when a custom Similarity class is used +Patch for buffer reuse +all issues to the CR +People using this can just send explicit queries designed for this Component +Updated testcase to repeat the filter queries in with configuration turned off +haha of course its the one I assumed it wasnt +WHY? to change a few files from one package to another? The user may not even care if the entire all classes are put into one package +A new version is attached +Possible patch for trunk +And bitmap will have bitmaps for each countryAnd if hive can support skip rows the bitmap index will be very useful +It would be a good idea to use something bigger such as K or even K for both sizes +Some comments on the patch Lets not break API +I just committed this. +This seems the result of a failure to recompile a changed source said this isnt quite the same issue as and which both deal with recompiling dependent files +Attaching patch a which produces the errors mentioned in the above commentNot for commit +But in any case you can see a approach to handle normal and empty xml elementsI would appreciate if you can review itthanksregards aki +FYI upgraded to work with both the and Avro + screenshot of Avro committed +Any suggestions or concerns +I misunderstood you +I believe syncronizing a clone call will be faster than the syncronizing method used in this patch + Helps better understand the problem I am facing +Will follow up with patch for Xalan branch +Only patch attached here +The problem happens when the constructor sets a field and AOP intercepts that field write this happens the field write wrapper is called +r +these tests are now working in my IDE +It is the basic property of NIO which is not working properly for my configuration +Have you got the same error too +Do you agree +I modified sortby to do the aggregation and yielding in a single pass and now all is in ccc +Bulk changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +no prefix is different here +runnning test case jdbcapi exposes this didnt notice that doesnt extends fromHow about creating a in Sqlca itself with out going to with information available in the exception? It has and error code +After reading blackdrags blackdrags last comment Im closing this. +Todd thanks for taking a look +Yup thats the feature +Updated and committed to trunk with revision +forgot to close +Doesnt this approach mean that as soon as a compaction starts your severity will get set to whatever the compaction calculated? I think you need some kind of flag that says this is manually set dont mess with it +However most of the JDO spec in relation to querying is not supported either +Thanks Simon! +Created a separate issue for the references issue because it seems unrelated to the thing +I did as you suggested itAlso as you suggested I added buildpig but also left the pig since within the release the jar will be in the project home folderI did not change to use the fat pig jar since as far I understand we all agree we want to move away from that +Thats probably good enough +Attaching a patch that moves the definition of primary key constraints from the table definitions to separate ALTER TABLE statements after the tables have been populated +set fixVersion to fixVersion so LATER since very unlikely that there will be time for this +Cannot reproduce for long time. +Spark patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Thus its more the object that is being chosen here not the name string +This is the patch I am testing now +I have attached it +That line has been removed +Done in trunk +These are some preliminary unit tests for modules and modules +Thanks. +It seems it is Dali bug +I do not think new directory and +As outlined above +John S did this +hope itll make its way to the website soon +This subsystem is still under discussion so moving related to CR +Thanks Anuroopa +I think that this may be an Android SSL performance issue which may mean you have to find some way to hide the iFrame until it loads or use a splashscreen or something +We still wait for a fix so doing it from a parent would make it work too +This gets the clusterhosts api working as + +Id suggest to name this or to make its nature clear right from the artifact name +The issue wont be present as the new style SOA UI has in ER +Line in does contain quotes around teh PATH and yes my PATH does contain I removed those from the PATH and continuum started up fine +Resolving this JIRARemoved label no need for that IMO +I have taken ownership of this issue and will choose a suitable point to schedule it +You dont mention it but it should be be an integration test no +There are some +None of the seen items are recommended for a user +But this can cause funky scoring for homographsEven if the analysis is legitimately different tf is wrongMaybe solrs removeduplicatesfilter should be recommendeddown the chain +Im putting them to good use to create a test case +Looking at the index using luke I can see over lucene documents with only deletions +Committed to trunk +Nice catch +to correct the fix version. +Committed +I think its better +I suppose we can close this nowHow come those classes end up inIve searched through the build files and only jboss being created in server. +Looking at this further i think if we return a status then the user will loose the clue of setting the use to ture from the previous error messageSo the best option is to leave the current error message as it is or we can return a error code and log the current error message in the console +Same thing +Appears to be related to +Since is not going to be included in next release uncommit from +was a different use case +The file was erroneously modified to include the context listener +Thanks for the detailed test case +Please find it attached +Please change to if you intend to submit a fix for. +Please have a look and comment. +Some basic checks revision Thanks Noble +Ive made the changes to the and included unit tests for each of the new properties +Constant name denotes the code version documentation denotes the actual changesbq +Verified in jbt build. +Scott added comments to the forum post listed above for more information +Thanks Yun! Thats a good catch +removed trailing spacechecked javadoc one remaining not from this patchnot sure why compilation failed when I tried locally it worked +Yeah a couple of test runs today and it passed. +However you can cause Derby to dump the class files to disk by setting the following flag when you start the VM I dont know any other way to force Derby to dump those class files to disk +I also recompiled the Hibernate with the suggested fix as this bug also breaks our production system +Setting to resolved as unable to reproduce +Im going to put this in a new category called examples +Talked to Lorin apparently youre working with him on this? Ill take a look at it but this could use more eyeballs +The problem was caused by upgrade to the latest rubygeems but not applying some additional fixes required by you try to verify that the bug is fixed in the latest code from subversions repositoryThere is a number of workarounds for the problem +Please provide a simple test case +I havent had a chance to run any benchmarks for the trunk version but should be similar speedup +I buy that +Jervis +Fixing RN fields +patch attached feedback welcome +Thanks for your persistence and attention to detail +Thanks MarkYour patch is in trunk at revision Trunk because only bugfixes get to the release branches +You wont be able to run both jetty and tomcat on the same server if you have ejb web services but I dont think that is a major limitation. +The job dont find the file that were created on prior execution and cant continue the is the right scenario sorry again about the confusion +It looks like Ive fixed the Seam tests +Ill provide a fix for that. +Jochen If instead of we have wrapping a then initially both have the same source location then when token sets its source information on the the wrapped remains isolated and still continues to have the correct source information +actually should be moved to before we do that fixing this would be good or at least I would like to understand why it fails since the plugin depends on does it not? why is its menu locations not available when seamgen loads +And will create per application a sub task +Thats a good point it should not I will modify the current unit test to check that and correct it after +perfect thanks for clarifying the issue! +Ive also experienced this problemCan we download +Please reopen it if you face the same exception again in M +but there are a lot of updates in version anyway +Not an idealsolution but it works +i dont think we should introduce a file dependency by default +Oh now I understand what you mean +Uploaded patch +Never mind Jay I see whats happening! Thanks +for the changes to improve stats collection +Why is this an issue? Shouldnt the pom included in the jars being interpolated? I am missing something here +This will let you rapidly display all sorts of calls +done. +Suggested good except for the typo for reviewing +Targeting this at should be fixed in Axis +Here is which contains a fix and a new unit test caseThe fix is in where I replaced xsdID to xsdstring on the queue name type has been modified to insert a new unit test method for this bug validation +Yea welcome! Ive also assign you Example Using JPAIf you dont think youll get around to it just let us know and well put it back into the poolDavid +I think we need a bit of upfront design and then well be ready to work on a solution +Its working thanks +The set method now properly registers the error channel including the creation of a dispatcher +Removed unnecessary code from metamodel plugins +Tested both trunk and branch version. +Any objections +Attaching a patch for the simple naive approach to the problem +Heres a shorter example that reproduces it +Whats the argument against doing this automatically? Im not a ZK expert so I may be missing something but I do think it would make it easier while developing and testing as you dont need to start zkClient or anything like that to create these directories and you dont need delete zk data directory restart zk nodes etc +We need to add some test cases +Cool! Thanks for your time on tests +It would be interesting to see if it hangs +But Jetspeed has many components and also properties files are managed by each component +The bug was fixed +correct version staged now at the same URL + +Fixed by Daniel in r +When finished reresolve with the new fix versions or label backportrejectx as appropriate +Sorry Im wrong +I probably have somehow messed up runtime and it seems thats the reason of not generating version of WS +Im attaching an example email so people can comment on parts they like or dislike +karaf templates +Yes +Thanks +If the task type is reduce task basePort will be the debug port for reduce taskThere will be no port conflict issues even a reduce task is started while a map task is still in progressBut this is possible only when a particular range of ports are free +Screen shots for the something is wonky with this patch +This did work in beta +Verifying expand operation in the installer +I just committed it +Sounds good! FWIW I figured wed let the patch sit for a while +Looks good to me +No it is working because I disabled querylayers plugin +There are plenty of forum references on point +Scott? +RAM For this case +xsl need to be included? They seem to be a different purpose to this pluginWhat is libI gather that everything under srcdocumentationcontentrep is your example data which would not be added to Forrest +Please add your detailed comments if you wish to make the case for reopening it. +modulescommonssrcmainjavaorgapacheairavataregistryapiutil +Thanks Nigel +hmmmm something looks wrong in the generation script +Once we have something like that I think we can revert this change +note in some cases we will still have to use the throwaway treemap or similar like the patch i uploaded doesbut we could then know these two cases up front someone enables ignoreCasetrue when binary sort order of the charset! utf binary order +Not GORM which is in Sorry I think its not you need +Support code for + +Ive fixed the problem with messages with no keys not being added in trunk +Thanks Isaac +This only affects developers with sphinx +The proposed change would not add any new functionality and it would add a temporal coupling between the call to createXXX and the call to you can better describe your specific problem then maybe a solution can be developed +Also I have implemented a simple failure detector based on the class and adapted this class to use this FD implementation +Why was this marked invalid? I posted to the user forum first and Matthias asked me to open a JIRA issue +Thanks! +I just figured exclusion support might come useful someday. +Good call it does look to be fixed +Has already commited to the trunk +So all issues not directly assigned to an AS release are being closed +See my comments on for fairly detailed instructions on how to upgrade a plugin to suitable versions of Velocity Doxia and +Hey there was no last name mentioned so tons of possibilities here actually +by commenting out seemingly unnecessary calls to Closing per dev code change. +FFI should make it simpler to keep up with any ZK API changes and make the code more conveniently portable to and other platforms but that might not be a big deal ZK is at and the C API has been pretty stableRe ASF vs github I think itd be great to have Ruby bindings in the official distribution as soon as the bindings are worthy +This is too late for. +The editor may not support moving between collapsed and expanded state but I believe it should still provide the correct visual representation if it loads a model that does contain a collapsed process + +Arun I still disagree with the cores being an int +I just tested and trunk and they are fine +Resolving as fixed. +they are about to screw up the last good IDE on the market +The class could be renamed as +Everything works fine now +EJB portlet initial versionThe file should be extracted into the pluginsopenejb directory +here is the sequence diagram +bq +currently murmur is slower then murmur and has design flaw +The last two testcases seems to be an unwanted white space change +coprocessors dont really look much like Bigtable coprocessors which are not really well defined in the literature anyway +unit tests VladimirPatch applied to TEXT module at repo revision r with modification to the tests to fail if the expected exception is not thrownPlease check that this fully resolves your issue +I will leave it to Jean to figure out why you couldnt since Im relatively new to this myself. +attaching a patch to include a local reference to the file in the you elaborate on your externalization proposal? You mean set the pageflow schema resource in or something similar +ThanksGeotools Committed rGeoserver Committed all issues that have been in resolved state for more than one month without further comments to closed status +Wondering are there capable client installations that do not support SNI? Can you run on and IE? My gut instinct would be that if they have gotten to why the fck are they still running XP +I know of a person working on this Ill ping him and see what his status is +This feature already exists +Igor could you upload your project to this issue +I committed this to trunk and +Your comments seem to be a good solution as far as I can tell. +we have to optimize this too. +so perhaps is coming in as encoding +The cd command can select a java resource in the project explorer but it doesnt open the resource in an appropriate editor +Are you using the jakarta parserRainer +Ran the hcat tests and they were successful +The proper way to fix this is to have the service proxies implementing equalshashCode accordingly considering proxies for the same target service as equal +We dont support module reload use cases +it would be also interested if as an additional step from top down if JBIDE can create the hornetq xml definition of MDB JMS cents +But if you could check it out too that would be great! Its super obvious that somethings gone wrong when you see it and I assume its on all Community pages as it effects the main! Cheyenneis it still occur after latest releaseThanksLibor +Integration test added +Please refer to this issue which got fixed to BJ Pranay and Ashish +Hmm the patch doesnt seem to have taken effect +Dont use this yet +hmmm I do think this patch can prevent this crash +Attaching DB dump +We need to redefine replicationFactorreplicationFactor has always been the target +Will try to include the patch for +At least it used to do this +Duplicate of +Give us some time to get back with a fix for that +Please resolve as such unless you have something else in mind. +From playing around with the suggester i figured out that looking up multiple terms is possible resulting in a ORlike list for each term +This might require few API changes with possibly an additional flag +is churning +I created duplicate targets and that seems to cause problems with ant but didnt with ant +Thanks Viknes. +Reviewed by Barry L. +forward all email to jacmorel will be forwarded on to that addressIll take a more considered look on the weekend +Per Daves comments and lack of additional information it looks like this is no longer an issue. +Fixed ordering for the tests +is required +In the meantime my machine is up on the blocks and wont be repaired for a couple more days +The info is added to the guide +While Roo ships with Spring JS Dojo as default we hope to see alternative tag libraries offered through community +This is a potential regression which we should examine +Attached patch addresses Konsts point above +One more thing +I made some minor changes in that I gave priority to the in the template over the set. +I just republished the assembly snapshot +Also marked as Optional +I dont understand the implications of changing the isolation level in the middle of a transaction +The configuration you are using is known to work so I suspect something else is going on here. +proto to the fileset in my next version of the patch +Attached fix +Eric I think this should fix the problem youve found +I suggest we change the embedded driver to do the same and not throw an error if the user specifies columns that are not identity columns +Then I went back to my application wrote a test class and ran one method that I know doesnt work without transactions and it worked as expected using Spring nightly +moved fix version from final to P since it is cluttering up the release notes for +We still get the out. +Although this has not been confirmed by the customer I am increasingly convinced that it is fixed by +ThanksAlso it might be good to run a version of the sort benchmark validation during unit testing so that we exercise those codepaths and find things like this sooner +If the settings are within the st page it works fine but it extends to page the error surfaces +Help in resolution is appreciated. +this patch can also be applied to new patch merged with +is fixed on the branch which may explain why we only see this issue on and +Seems its env issue and resubmit the patch +Patch for the improvement of RobertSorry +bulk move of all unresolved issues from to +Ill commit this when Ive had a chance to do a little more testing +I will change both of them back to the way they were before +If each of the factories would access a class that is only available in that specific version it might internally fail with a +Not sure thats a good bisection +I havent been able to get my WCF tests working Im not really sure what the goal of the WCF support was so it makes it more difficult to add that test nowDid we come to a conclusion on how to fix this issue +Thanks I verified it +moving to +What do you mean by was not in the runtime capability? Do you mean when it is not exported on? I dont think you could mean this sinceif it is not exported then no bundles would resolveI guess just give me the specific steps to reproduce the issue with the above file +I also dont think each attempt is tracking its startfinish time so those would be nice enhancements +Thanks +The root cause is explained in which indicates what needs improving +now checks for the existence of system property +No this is not the right fix +Moved into we need to have a conciseconsistent story for this class +It will be great if you can review the patch before committing +Anyway a retest of this bug shows the expected output now +But not found +Hence marking this one as wont fix +closing since fixes are in released code +Thanks! +Thanks for the help +Will send out an email to to see if it sounds good everyone +Committed to sqoop +Correct which is why I have already created a JBAS issue which this issue is linked to +It looks like Im not going to have an oppurtunity to resolve this any time soon +I agree +Do you happen to know if there is anything in your application code that can close the underlying io session bypassing the SSL layer? The only reason for an IO session to get shut down by the IO reactor is due to an unexpected runtime exceptionOleg +This should not be the case in the code as it is now but we could have webpages created with the default status somewhere else in the code and we would not be able to differentiate it from status used by the redirections +Shall we Resolve this? or am I missing somethingRegards issue has been Fixed and therefore Im resolving this issueRegardsSenaka +Looks straightforward. +modulesimageiosrcmainnativejpegencoderunixIll check this again as soon as I have time frame for it +I will give it another look now that I have a better knowledge of myfaces +I seeLooking at the usages of it looks like we can drop that constructor initialization entirely +Ive committed this. +put back the passing in base +No time to checkin though +This is my understanding +I still have this issue with + +if this version is ok for you ill look good I will try to commit evening +I manually tested this +Confirmed through junits that the trunk code will give a random secret when ever a is usedI will include that fix into my this is an interesting JIRA +the other two points mentioned by alan are still im wondering if we should try to use source code control so one gets the changes of the otheremmanuel? are you reading this +Migrating bugs marked as resolved before and tagged as version as fixed in RC +I dont think that these protocols provide any guarantee on message order +Do you want to open a new issue? If you want I can take a crack at it but probably sometime next week +I will add new itests using Pax Exam in order to bootstrap a Karaf distribution deploy features on it and check the behaviour. +I agree with the point about +Added it to the patch and committed it to master +Its back now +In theory that should be the same as running out of memory and having it killed off. +does this still need to be fixed? i understand longer timeout is a can be closed as it not reproducible in QA and since also the increased timeout works. +But since Configuration is already permits overwrites this is perhaps aExactly its already a class +Same for I have been unable to reproduce this +Ie creating namespaces for tablenames with +yeah this is needed for m going to work on this +This default configuration may or may not be the same as the one used when the cluster was started + +Can you explain when you expect an unset or incorrect state +from ACF docs Boolean value +EDIT TYPO IN NAME Thanks MartinYour patch is in trunk r rI wonder why we committers are so long to commit such patches +and what do you suggest +Shall we change that also +Thanks TianIt seems like this one is ready for commit based on from both me and Andrew +Well consider this as an enhancement for Spring JuergenThe spring form tag is agnostic of where it is being used it doesnt know whether the jsp in which it is used is being used by AWFC or SFC or ACThat is the reason why I thought that the defaultness of handling the page attribute must be in the AWFC +Pushing to since this just hasnt been our priority for +Trivial fix +Test case to follow +Or is there a way +This caused somesupport nightmares especially when a single database problem could prevent access to all databases +I need to read through and digest itIn any event moving to +Patch looks good Lars +Oliver our original implementation registered our custom converter in a static block before any application code was executed +We fixed that and now we have one Name Tools for such menus in JBDS and JBT +on both Idea and then tried running ideaclean and ideaidea on the same project +Im trying to narrow this one down but its taking more time than I expected +Is it possible that this failure is the same as? Do the nightly tests run with just in the classpath and not +Attaching patch for +FWIW beginners can supply patches anywhere! The INTRO label is just meant to help people get started +It is ready I will commit this later +This has been added to upstream +Still a tiny bug to solve on the sonar plugin +Its ignorable for think this is now solvedPlease reopen it + +Thanks David +ah yeah agree otherwise it would have to support both naming schemes again +I didnt see any related commit was it done +cnds its adding +Yeah that sounds about right +I would think just disabling would give more benefit than without disablingIn any case already existing benchmarks might be able to show any benefit this gives +max is cumulated in the for loop so the opertaor should be +I committed this +Folks who wish to be able to do things like this in user data should use +Created patched based on Eddie Bushes patchbut he still gets credit + was copied from mimej and its javadocs are wrong with regards to nullability of mimeType and transferEncoding +Where seconds timeout was sufficient needs now up to minutes and more to get executed +Patch looks good to me +Is it kosher to move patches out of zk contrib and into independent repos? And what would make it easer to collaborate on in github vs +Implementation does not resolve inheritance hierarchies in implementation +Resolving with fix version as the release notes should be fixed there +Which should be a rare case +Ill attach it +At revision o Throw deployment exception for errors in parsing validation and constraint mapping revision o When the modules are packaged in an EAR additional processing is required to obtain the validation descriptors o Each module has its own + +Well be upgrading to in the next few days +This patch also contains a fix for a boost problem +I will look into it some more +Ill keep this issue in mind while working on these. +I can also verify that instead a warning to stderr is printed which seems to be the intent of the patch +Suraj Do you have a patch for us +Good for me +Which is something I would not want to like +all passed +Thanks for the patch Gentaro. +At the end of adocument add wed flush the like class atwhich point the underlying data would become available forreading +Committed. +Let me know if you want me to remove it because you fellas want to grab it +Works with but fails for and all later versions + +Any chance of fixing the char lines and the javadoc issues? Let me review it again +Was getting overly exited +like stop force if the bookie doesnt shutdown in a timeout period it call kill +Done. +The docs do mention that a restart MAY be required +I have release on my PC +The problem is related to the which stores the constraints internally in a +if we can have samples with a single feature and samples with multiple features thats perfect +Update of original patch with doc fixes and an alias of mozilla to gecko for agent of patch for NOTEIf you backport this fix to another branch then be sure to backport as well +It was released with the fix in +Refreshing requires you to hold the global lock +Committed lets see whether this fix solved this issue or not +IMO we should solve instead +Can I assume the param to constructor is always? could cause extra copying that we may not had before +Code cleanup to confirm to the coding guidelines +Can you attach a patch that adds layout unit testsThanks has raised to add a unit test for the changes in FOP so I will go ahead and commit this part of the bug fix too +Thanks kw +I have always thought that we should get rid of it as it is not really needed but if we want to keep it so that we do not have to move the entire synchronization logic into the server then lets at least make a step in the right direction +Committed as r. +My Firefox now recognizes it asThanks a lot +Thanks! Ive applied the patch all seems to work fine for implemented +Thanks! +please move method back to it original location in +Ill try to do that and also see if I can squeeze in an extra test in the nightly runs +BTW Id recommend switching to Maven or even + +Fixed in + runtime error you want to start the debug modeLine Error The object doesnt support this property or following microsoft script editor cannot show anything +So to me it makes more sense to introduce a new create method with of the existing file as a parameter or a copy method with an option to replicate fields +The edits file is not corrupted +The filesystem would be intact only minus the changes from the outstanding edit events does that sound correct +Where is the NPE coming from? Can you explain +Just to be sure have you tried removing the +nice patch +exclude more intermittently failed and unit test patches are appliedPlease verify. +closed this as the description was incomplete +As mentioned in the javadoc and Cron only work with Quartz +Ive moved the repo from over to and made what I think are the required changes to the module names and files +Thanks it works nowHad to remove from my local dirNow it works properly +The lists seem to be totally broken now +This was tripping me up too +Reverted from +I will try to isolate the problem further and keep you guys updated the following days +Not. +will run into the same duplicate class problem you did run into when using the new with out ant clean with the last one +Thanks Hairong! +I had an idea to resolve that issue but unfortunately it didnt work out too well in practiceAs a fallback Im now solving just the XA case and leaving the multiple store operations in place. +Fixed by +Which do you use? either +I tried to do a search before creating a ticket but for some reason didnt find the issue +Anton if it works in some files but dosnt for others it means Model does load aj beans +Hi SamIts mid next week already Can you approve this requestAlso please let us know whether this approval is specifically for or whether we have permission to treat other Category X modules in this way tooMany thanks +Looks like well need to use the Fabric to find the private key to use for a node though I wonder if you can figure out if we had a private key how wed specify it in the terminal +The nested scheme is pasted onto the front of the authority with a dash +It may be +Congratulations on being the first person to use a generator flag +Closing the JIRA. +Considering this as resolved for tracking purposes +Thanks Dennis +So we are going to move everything currently under contribanalyzers to contribanalyzerscommon +committed as r +and the can give that backThat way the code would know of the that the new one is dependent onOnce that is available the can reorder the insert actions first order by go through the list and move Entities with Parents to the end of the list now move all entites with no parents are at the beginning remove those toplevel entities from the dependent list of the Entities that contain them as parents +cdh codetest snippet would be great I looked briefly through the code but I cant find anything can you comment more on user impersonation how would you do that through the Hadoop command line itself +Fixed in rNow managed x uses commons math close of issues as part of making release notes. +thinking about this more if an aggregator mojo is bound simultaneously to two module poms in the same reactor where should it execute? If its bound to a pom at all then merely the root location of the reactor may change whether the aggregator runs or not using the logic from revId m commenting that code out until I can think about this more and come up with something +The fix has been verified with previous builds Hence closing the issue +You may want further logic to force a particular CRS to be returned but that seems to be a seperate +Fix for Revision Fix for Revision +Some views only are given the document +Thanks for your contributionIve just committed it +The comment there says This interface contains the original static final int constants which have now been replaced with an enumeration in Code +proper version +Applied patches in r +Any news on this issue +That would be very convenient +Ill make sure he fixes it +Is it same as? adds a configuration to know whether job needs a and +The updated working patch for the v branch +Ive observed the same problem with Giraph with Hadoop and +I am wondering why my Eclipse was not complaining +Tagging for t this be an enhancement to +Lets dont change the messages here +But right now maybe the Boyeux should be shared in a different way? By exposing it as a service maybe +Any chance you could fix it in Spring Framework final +It is the automated build process +Marking issue as fixed for +Interesting suggestion +Spring upgraded to to Spring +This could not be done until had removed the documentation +Thanks again for reporting. +I held off while I was looking into +I have committed this +I have no negative ratings +Files have been updated +Still a problem in. +Its OK +Refresh or rerender the table several times memory being used +The idea is to have a cache value to not regenerate the merge index for all requestsAnd to cleanup temporary directories having a session listener which will cleanup existing tmp index when user session end +Just need to fix a few compile issues caused by recent chanages on trunkcommited to trunk in revision and that I think about it +No +A complete thread dump of the application while running +Apologies for the inconvenience and thx for the extra info. +Oops attached the file wrong files +Thats basically my viewI did look at the hudson plugin and it has similar problems with supporting multiple sources and multiple build jars +I omitted then passing the password as argument and had to enter the password after SVN asked me for it +Final is released +I tried with the attached and several other xmls and all of them were correctly imported. +The view ID will not get out of sync with the session because I am only hitting it from one browser tabThe way we ran into trouble was that we hit two different ports with two different browser tabs +dataList removed footerClass headerClass at +The fix above has a slight issue but when the correct fix is applied the resource does become available +No problem about the delay we appreciate all the effort you have put into the Tomcat migrationThe results test migration looks fineThe jasper error can be ignored +from me + +Simple patch that adds a configurable time to sleep before trying to lock a region server +Another option is to use reflection to find all public methods and dump a description into gold file that we check in together with the testRunning the test would involve regenerating that file and comparing it against the checked in version +Updated main trunk Committed with revision Updated branch Committed revision . + pooling committed +It works as expected +Yesterday I updated to current cvs +Make public +subversion patch for +Does the ignored blocks affect to the metadata of image + +So anything I see something with security and key length issues then I think about that +This code is not actually part of SWF but currently resides in the Spring sandbox. +just cd to a specific module and run maven Dtestcase testsinglethanks dont want to run any tests +thats the big thing people miss out on in terms of steps to get stuff acceptedCould you please attach a diff instead of the full java file? That allows us to see what actually changed +Thanks +If you disagree please reopen? +Let us know when you can make one and we can reopen +However I encountered a problem with serialization Im serializing twoclasses one of which is a subclass of the other +Linking to the HDFS issue +Users have to sign up to oracle site +I dont think it will take much to port that to Steve +the root of your project is not part of the classpath so we wont see it +Could you then test it using on your system to see if it works +Thanks Andreas seems like i missed the other reports while I was searching for a solution +Sound ok +Gregory could you specify revision on which it starts to fail in such way +This worked in M +In the current implementation the caches are flushed when the total for all columns reaches M +Please close this issue +is more safe as if a child reader is closed also parent readers have mostly no chance to do anything +If the user changes the location of the server or runtime the filesets are outdated and not updated +changes devoted to the dev tasks were made in the scope of additional changes are needed +Thank you KojiNow I am running all the tests with your fix and going to close jiras as soon as I verify that they are fixed +Hi Carstenif you did not provide any patch we could close this incomplete issue +Can you comment +Needs evidence works on real cluster and that edits are properly recovered before commit +Great thanks! +MaxCould you please provide me with more details +done in m. +This button is displayed in the dialog box that appears when you select a global in the guided editor +The issue has never been in +Also I put the images in resourcesadfimagescasablanca +Still wondering about the reasons for the other problems regards Sh Mais the webapp that is currently posted now your updated oneJust checking to make sureregards MartinSorry for the late reply +Does such an option make sense? I assume currently modifying any file triggers a reload in all browser tabs connected to the serverTheres also the AddRemove option which just says There are no resources that can be added or removed from the server +HiDid you executed analysis after change of severity +below a commont intermediate folderI committed the dojo source in the extensionsdojo module in Rev +the hello world plugin why do you say its an improvement and not a blocker bugit seems i cannot use some default plugins in sonar with these releases and cannot add more plugins +See hang screenshot where JMS tree just say loading +Completed in the pull requests fork +I sure hope not! This is definitely not a security feature in any way. +Alternative patch that uses a new native to distinguish the server socket case so the soreuseaddr can be set as default for Linux +Looks good +Therefore no need to pass it as rd argument to the filter function +Per Teds suggestion Im uploading a second version of the patch which breaks up the method into two lines due to the length +fixed in rev +Perhaps if the annotation can only be placed on a field or local variable then the compiler should treat it as an implicit def as it does now +I bumped the RPC version by one since a method was changed in the master interface +Using includedocstrue saves on disk space as you can avoid duplicating the data but does mean that the document is fetched after the view has run potentially after an update or deleteI cant immediately see why the document would be null in your case since you say you never delete the document only update it and you arent emitting objects with id or id and rev values which would alter the document retrieved by includedocstrue +for the review +Have just committed the compatibility matrix section to the release notes last section I think is needed +When namenode checks datanode status by calling get The namenode goes over its include hosts list and tries to determine if all the hosts in the include list are all registrant +Thanks! +I think I dont like it +Have not yet looked at a cmdhelp packaging. +To fix this well need to associate a weak affinity for the stateless bean invocations +I dont see why this is a problem +bq +This is functionally wrong + Thanks Rakesh +verified fixed. +If not pleaseThanks Eli and Nadav for catching this +It looks like this patch will fix the daytrader sample app but I have no idea what to do with this file +Cheersve added group write permissions to the appropriate +Great committed. +When you must run some task to make it happen the question is then which task +Defer to +Updated release notes fields +zip has been provided since EAP +added licences phunt will dosunix all the files +bq +It is a big broken though because the may not be exact without having a specific entry for save with a propagation of REQUIRED +I assume we dont want to do this in +Ill see what I can do in the next few days +Can you confirm that you have a work around for the moment you can simply copy paste the surefire code. +Please excuse my ignorance but which branch did you fix it onI checked out the trunk and it does not seem to contain your fix +If we are committing a batch of rows we should do the retries with only row so that we dont OOME region serversThis way a single row commit would at least know what failed +from experience without a getter +Development is continuing by bringing up the manually +Big thanks +Im running JSDK +Done now it works and can be deployed without any modifications to Tomcat and Jetty ll test with jboss AS and see the trick is like in the war assemblies +When I made my recommendation to close the bug I tested against SVN +Pretty annoying +Issue may already be resolved may be invalid or may never be fixed +Ive seen several cases of nodes failing to keep up with write requests where the commitlog was the bottleneck +Done with this cleanup effort +I have finally been able to have a working implementationAs a first working version and the complexity of the feature it needs to be tested in a daily use +No bulk moves from trunk for this release +It standardizes output as UTF whichshould be able to represent any character data +Ill investigate more tomorrow but as of now it doesnt look like its yet another RCL assumption gone bad +is currently running Is this issue is reproducible on +exemple on Query exception etc is now done for all the explicit +The test failure is caused by the same typo is expected in the unit test +Heres a version of this patch that adds a test +so be sure to update merge before taking the patch further +Looks great for me +Same issue as for ghost project deletion or project deletion page once clicked the user is not redirected to the Migration in process page +I need this facility +Changing this bug to a blocker +Can you describe this feature request +Tried a couple of times this week to get the Macbook without success ill get it eventually thoughGlad to hear the buttons are there +ok Its now also possible to say allowEmptytrue on the spec element to say that an empty field should be allowed +And it only took four years +I made a temporal fix for coming release +Trying once fail erratically canceling s try it again +Attaching the patch of a basic first attempt +This means that you currently cant define multiple or without idname +The NPE should disappear has already been reported its fixed in the nightly snapshots for several weeks +Fixed +I am also having the same experience as Roger +Now that you explain what you want to do I understand the requestYoure right theres no way to do a custom +I cant reproduce this +Closing issue. +In some cases I even saw one minor commentInitialization for stats can be done inline +Number at the beginning of the line is the return value of the previous linepheme trafficserver start pheme sudo trafficserver startpheme trafficserver stop warning failed to kill Operation not permitted pheme sudo trafficserver stoppheme +No disputing that it would have been better to use an abstract class than a concrete class in the design I guess this somehow just slipped by all the people who had a look at the original draft +Fixed +Added more test coverageThe new test adds documents with ids and validates that a single maps to a single shard only +We should have a better way to avoid this +Okay I used ant without parameters and for database settings +Im going to prepare similar patches for all modules +In BRMS the RHS works like a BRMS the RHS doesnt also applied to and branches. +Sorry I dont know how tochange this in IDE after project creatingCorrect +This document only change +pulling standard aggregate functions into the grammar is a slippery slope +If the RS does it that means that in case of RS failure the masterzk view could have a region as OPENING but the RS may or may not have edited META +resolved conflicts +Notes and explanations are in the WINDOWSNOTES file separately attached +Elevating to blocker +marking as Patch Submitted so it doesnt fall off review queue again +Done at r +Also do we have additional jenkins resources I can test on +default max time between clicks for a double click is ms +The hidden request parameter is used in JSF to detect a postback even in a environment +wont fix. +Who uses anyways? It is not my code i do not have the guts to just kick it out or move it away +This looks like to me can you please try latest from scr plugin from svn +binary mode do you have an example +Hi Stack +Unicode escapesUgh + +Maybe something in netbeans itself changed +Fixed as suggested in nightly build +The type handlers will now be registered explicitly for both primitives and their respective wrapper class. +Patch can you please approve applying this patch to the geotools repository because it concerns a fix of a bug that could affect others +This should take care of the failing tests +Ok it is good to know that this is expected beahavior +The logic looks correct as long as it can also handle +Marking this as resolved +A comment perhaps best addressed on another jira is the questionable utility of the overall permission checking for this method +I am far from convinced that this justifies the added computational complexity of walking up the tree and checking all for a that might retrieve a particular markerI am of the same frame of mind I think thats neither feasible nor really necessary +svn ci Sending srcmainwebappindexcommunitydownloadSending srcmainwebappindexcommunityAdding srcmainwebappindexcommunitynewsSending srcmainwebappindexcommunitySending srcmainwebappTransmitting file dataCommitted revision . +docx etc +There will be things which are unknown to me and have to be address when it comes up +Hes asked that I go ahead and submit a fix for the iphone ipad issue +Profiling is deactivated by default thats why the logger rails must be set to WARN by default +I fixed the code some time back so that server do no longer require SOAP Action to figure out the method to invoke +will run tests +Please reopen +Ive committed it +I agree with you +Thanks + +This is the best I could do after years of looking at this methodThe duplication is unfortunate but hard to refactor because theyre doing slightly different things actually +Will close this ticket and open new tickets for any specific scenario we want to handle later. +Please open a separate ticket for this issue. +See +This issue is fixed and released as part of release +I have committed this +Did you mean to revise this one as well +This is very important +Privileges issue +In my tests this commit fixed the problem of Thread for releasing the fix +Rebased against +This is now fixed by Mikes cleanup to etc which fixes this combinerewrite close for +you could use the existing test repository although there are no guarantees that data in it will live for long +i think durability guarantees of local file systems are a bit different from lose of local file systems because of how failures happen if a machine fails all processes on that machine also fail +Ive seen the failure today as it is described in the first way on Windows Server xPentium x IBM HT +AdrianThanks for putting up Image +Thanks for helping to test this Amir marking this PA +verified fix by testing with groovy beta code and postgresql! Thanks for testing that +It would be great we can use it in containerwe need add and bundles seperately and all dependencies for swagger jars which are not bundle yetI will add for any bundles if I will add a featureexample for Servciemix track Freeman any chance to add this quickly for release? Or do we postpone to release +Attached patch adds a few more error codes to exceptions inside +This patch fixes the additional problems with windowsAlso this patch is includes all the code from the other larger patches +Rajith I believe this is now fixed +Added boolean check as Ramesh Final is released +Stephen lets continue this discussion over at and Ill be committing the patch attached until we find a better solution. +patch from davisp is fixing the issue however now the attachmentnames patch is failing +I think should stay inside +More of these will see the light of day perhaps +So just have patience +start cont other start other and the sequence patterns which are not allowed +Then i ran the full build of axis with tests it did build successfullyI have attached patch file with this +This is solution BGood pointPPlease interpret my comment then as pointing out that making this change optionally backward compatible is pretty trivial so why not? Some admins will be unhappy about this incompatible change while few if any will be unhappy about an extra configuration variable they need not concern themselves withAlso to Todds suggestion of a format string versus straight prefix +Readme and FAQ are updated +Im writing a patch to fix it now +I also converted your test code into a unit test +I presume this class will be removed from Axis binding as we do not want bindings depending on specific databindings +Nice Daniel +have decided to reparent to and leave open +The test then sends some queue or topic messages in a round robin fashion checking that each message is delivered to each consumerThe source code does not conform to the coding standard in +Need to check whether it worksAny other updates on the patchrequirements if it worksThanks +JNDI should not be creating a new channel per context +The code would be cleaner +Thanks for review t the RS not check in to the master with an RPC until it is available +Whats the status on this? Has in been fixed for Hibernate ? This is yet another +It may be not enough to store a reference to the metrics as the HTTP connection can be immediately leased by another thread of execution which may alter the metrics +Document the secure processing feature in the features page +I could have sworn I had resolved this on Saturday +patch is attached to the jira also +Ari S +Please do not use the embedded wsdl and xsds except for use with this sample project +Id rather do some heavier testing and also make sure SLES work properly +Added a test and fix for this issuePlease test and close if ok +I think I misread the will be consistently not called in both cases +As discussed I dont think this is a valid use case + +I have tried turning producer off but this made no difference +The upgrade test errors are logged as +Nice +It was not an ADS problem +However why dont we just store the reporter reference in the base Operator class rather than the specifically? If we run into other cases where we have to add progress indicators this will make it easier +Steve AFAIK Pig uses the mapreduce API the work done for binary compatibility only covers the mapred API +Committed on trunk as svn including Knuts proposed simplification thanks for looking at this +UnixAFS permissions are ok +ported Metadata framework over from Nutch including metadata container class spell checked metadata class unit tests +Merged onto the branch and into the master branch. +Please check the current you areThe problem in this bug doesnt exist in trunk after as it has been completly rewritten and uses an arraylist to add extra items +Attached are a minimal xsl and xml files to convert to PDF +Thank you +However all pages are updated on save no matter if I browse something at localhost or at +I think these characters should be replaced with spaces as a matter of question Does the take responsibility for this or should plexus change their class to replace them? I think plexus should be changed anyway since it currently allows invalid jar manifests to be created +If syslog address is provided to ringmaster and hodring section then their log is going to syslog. +A tip from Denis Golovin we may use plugin for this dialog. +Im fine w all being under rather than under subpackages under especially if its loads of work +This patch applies to the trunk and +Anyone using the patch should understand that vanilla can not start on a that used this patch before the start. +No complaints from either +skip the two levels just allow setting one type and if it fails then print an error its the users job to change it to one that works at that point +The writer concept seems for me the best solution for thatAnyway great work! When it comes the time that are users complaining because of additional features u did a really good job +I am not clear why application writers would need to do this +Oh Im asking because Im trying to help out +No changes were made to AS +On my opinion is qt is for the birds get rid of it +If we find problems with a given migration Ive had it take substantially longer or be a quick fix it depends +We decided its ok to keep as it is now +I cant tell since Im not an ivy expert but it looks like the groupId is +Oops I attached the wrong patch +I am going to commit this to and trunk very shortly +It is not a MEAD build so we need to recompile until we deploy the repolib wrappers in the next EAP cycle +Fixed in revision . +You could match the font on the creation but the trick would be matching the styles when the font size is increased ok forgot about resize events +it can chew up resource pretty badly when u have it inside uirepeat etc +Could we perhaps also merge with by giving two lists? There really isnt any such thing as a +I am not sure if you intended that to indicate that you need this change on those branches +Thanks for the code this should be fixed +The two java files are the main classes the zip file has these and the various support classes needed to get it running +the proposed name change +Changing the Phase of the to PREPARESEND makes sure that it will be executed before anything is sent to the server. +It fails +See for details of theSee for details +Resolving as duplicate +New tests not needed javadoc improvementsCommitted. +bq +I tried the snapshot but the runtime libraries are still snapshot works fine +I should also note that the maximize window state works correctly in IE +See +Also fix a javadoc typo +Hence closing the bug +Nevermind Ill stay out of Nicks hair from now on. +If you dont have a server where does the Directory take its files from? If its from the local you can use to load the files from a. +To really fix this bug should the eclipse plugin not read the workspace file +I fixed the hard code for the host attribute in +Given that the classes in question arent serializable it would seem to me like the correct change +Attaching a patch with the solution originally suggested +I am working to chase it down but I dont think this patch will make it in for +I will try with the newest Cordova branch and JQM and let you know the result as well +is there any way to break this K patch up a bit for review? Maybe one for renames and one for manual edits +Patch adding and fixing up usages in Trinidad to use so that the are cleaned up at the end of the request thread. +could anybody tell me how to solve the problemsend to ! +Integrated to trunk +Nice example of environment differences +in +has a simpler implementation as well +Its currently the remote file URL so you need the repository url in order to be able to create the task in the first place +Ive committed what we have so far to help move things forward in +Abstract base implementation stub I used to implement some of my user types +It as a syntax error +Was there any reference to the SQL spec ever found to ensure we are handling the issue correctlyThe count of number of rows in SYSDEPENDS seems like a kludge to me checking indirectly for some state rather than checking directlyWhat if the view is dependent on some function that is not owned by the view owner how that does affect the solution and the desired behaviour +cases access to resource which have same content same every configuration but response is different between webdav and in web application access +If I have to mention I have failed in finding a simple explanation +Jeff GordonURI class in is a complete mess none of the existing committers would touch even with a barge pole +Chris many thanks for your patch +Thomas please address this issue as you see fit including resolving as wont fix +The fixes in jclouds snapshot could be patched inside whirr once all is verified +Incorporated a proposed workaround change code logic to test for the presence of issues for beta . +Assigned it to you +please mark it for later if others disagree +Solution using from axiom snapshot instead of +I dont know IMAP well but you may need to do something link ping for keeping connection I guessAnyway I agree that its not a issue +I built with Java and and both Maven and +Patch +Ill go ahead and close this issue given the explanations above +All this issue and did was codify a way of doing that into the framework which is IMHO a good thing rather than using an solution which as demonstrated didnt really work for all users +Looks great for me +However I am not sure they will be used that much +Id love to see as many possible valid and invalid XML document instances attached against this issue +Of course it did not +I think that for this well need to clarify which belong to the AM +I did a LIMIT on one of the data sets to be joined too +Maybe we should name the accessor methods instead ofThats a good idea Ill do that. +Committed patch version b to trunk as svn +Ill go close that +Modules has no idea about AS configuration files +I have attached patch to disable this +Patch is looking good but Im confused by a few thingsAre you starting all the servers at the beginning? Or do the ZK servers only actually startrun once you kill another oneThe idea for this is to create a ZK quorum of servers and then be able to kill individual ones +Are there some missing test code +the same project works in BRMS changes into the BRMS branch +So everybody could be rest assured +Configuration javac parsing of result of compilation preservation of results of compilation fix patch by Peter +Are you asking if hive can export a partitioned table to a flat directory? please include more details in your report +I hope it will never get fixed +It was more of a do we really wantneed to expose all the properties relating to via pattern convertors? +Patch for +Thank you for the comments Rick and Oystein +Perhaps the should redirect the user BACK to facebook and ask for authentication again +There is the same error in other panel components accordion and toggle panel +Hi Andrew the issue was spotted on Linux machines under two circumstances When debugging since debugging port is openedclosed soon after startclose to the end of JVM process +patch for urlCan you give us some examples of how were tokenized before and how they are tokenized with your patchFor examplehttpbazbarcardarmarHow is this tokenized with your patch and how was it done beforeThanks +So I have to parse the XML of the WSDL to figure out how to request all of the code be generatedIs their at least an all switch? +However we cant do that since youre not registered as a project memberFor now the one who gets the patch accepted will get the credit for the fix +Fix merged into trunk +Nope +I dont think its a bug at all +If you got that exception it means that no other servers are left on the client side to try +I think we should encourage people to expect it to behave as it does in karaf so that you get back to original server state before you installed any applications or bundles +Now I can go off and try to investigate a little bit moreAs I mentioned theres a very good chance that I screwed something up in my unit test because when the test is complete the temporary folder with all the indexes is removed +testing against dev build from I still needed to use the workaround you provided to get TB to recognize the hostMy setup is this Im using a +Yes because on start up logj loads by default +Going to open a new JIRA to handle merging the concepts and cleaning it up for the previous comment. +Revised patch with both nits covered +so afaik I can see this issue is resolved not by removing all the tabs but making the source page default in case VPE is not working and showing an error if users insist on seeing the visual tabThat is a good compromise IMO +What version of xlC are you using? We compile Xerces against xlC on AIX every night and dont have any problems +This has been backported only to the Fuse Camel branch +First it is not obvious to me that the right place to implement this is in DBCP itself +The URL is not valid and cannot be loaded +Can we make this for? thisnext week +Is that ok +Patch to lock the socketWriter to protect against concurrent access +The error messages would be easier to interpret I think +Ill try to get it done todayI suggest that this Jira is assigned to Dona before committing this +Actually Im ready to contribute this enhancement to Jackrabbitexcellentbq +If you are using I dont see how a property annotated with Lob and declared as a Java Clob can work with trueIm using with Oracle JDBC which does not support the contextual LOB creationWhat version are you using? Could you be using a version that does support contextual LOB creation with falsePlease check your log for the setting Hibernate detects forThanksGail +Here is the file that shows the screen error. +The build lets me build under and if I apply that patch it compiles and tests fine +A diff to use the for the +We have to review all JIRA issues again before doing a release. +Thanks for the patch +The fix sets the time to UTC String format in the GMT timezone +When this happen the table datas inserting is in wrong order +I still need to forward port it to trunk +This is just a bug in compare the query debug output dontstems bees to the output of dontstems bees and you will see what I meanThere is nothing wrong with the lucene filter here +I cant reproduce this problem there +The service provider functionality and Classloader visibility might be better controlled with an service instance +I am certainly interested +Integration tests were updated. +Committed to branch +This is definitely the same as which from my testing was fixed in +This is also a Specification Compliance issue +Hi Luke +We cant change the API behavior at this point as it will break third partys UI and Apps built on top of CS. +Could you maybe provide or attach a test case for this? This would be of great help of isolating the issue +The dumps should probably be on a session basis +I recommend leaving the timers on +As for encoding aggregate functions i suggest for now we follow the pattern of the rest of the class and just deal with the duplication and when we come up with a more complete SQL class consider applying it globally +Reopening to write release note issue now release note text has been written +Uploading again to rerun hadoopqa +Trunk committed revision +No real need to build the package for that removed the build phase and it now gets the same data from another source and doesnt give this error when some other asset is broken +Just execute the +So I created this new patch attached which compared to the original one takes care of wfs as well including code modifications and adding tests moves the consistency checks between the number of view params maps and the layers count in the kvp parsers instead of spreading it out in the code added a very small change to the docs so that people can learn what is possible to doThe patch passes a full build and I also tried it out interactively finding out a useful feature of it past allowing the same param with different meaning in different layers allowing the same layer to be requested multiple times with different values of the parameter this allows to make maps like the one attached with this requesthttplocalhostgeoservertoppwmsVIEWPARAMSpeopleACpeopleALAYERStoppAvpgstatesCtoppAvpgstatesSTYLESlineCpolygonFORMATimageFpngSERVICEWMSVERSIONREQUESTEXCEPTIONSapplicationSRSEPSGABBOXWIDTHHEIGHTJustin Ive cced you in this jira because despite the testing the patch is large and now involves WFS as well +i reverted the export and moved the utility one level up. +Attached is the final version of my work to add SSL support to the Java client and broker +Rather the different trees and jars can overlap in the java packages they contain +Its the thread causing the change that pushes out change notifications which in turn drive the send back to clients in a PS callIf we have a problem and are tying up a thread they youre fear is very valid and need to correct it asap +Can someone review this +Anyway it seems the issue was related to the old NFS server +Is this issue still in progress or was it completed? Tx is still working on some material for this +IT added. +This is still brokenValidation failed validity error Syntax of value for attribute linkend of xref is not valid validity error IDREF attribute linkend references an unknown ID gt +yeahbq +Assume closed as resolved and released +Sorry I finally find some time working on this +will commit shortly +management session of the parent build works that implement still fail something with API to be excluded from the build working on that +examplestarget in ceeeccfcbfcalso created a jboss profile with the clean package Pjboss +And the converted page +Patch does not apply +doesnt apply +Ted what is that you are reffering to? +The text in the static regions does seem to be layed out properly dragging over the text in my PDF reader highlights the invisible text +pom files and missing features are addedready to include in update in last continuous build and Will be available in next nightly and aggregation on JBT +Now that I know it solved your problem I can commit the junit test +groovy will display ivy logging messages to +Im writting the patch for this nowIf you could Does someone assign this case to me +Thomas Id prefer removing the code then commenting it out +If you want I can take a look at it tomorrow at work +bq +bulk move of all unresolved issues from to +Attaching unit tests that show the issues +Wed still have access to through but we want to discourage direct use of it from compute +The run goals are typically using during development so it becomes cumbersome to type httplocalhost for example +That is really interesting because in my tests using a byte oriented state machine for parsing csv data can be nearly an order of magnitude faster than using strings +Note Ive tested the final patch on Win and Linux IA and didnt see any failures +Committed those now +Again It is a great Work +MichaelThanks for the patchOne minor nit Can mrinput and mroutput strings be and reused across both setting statementsFor the test question I think events outside code approach is better +Hopefully my english isnt that bad +If It is not solved please. +has been marked as a duplicate of this bug +The xml format of your example looks totally free +Patch should be applied in the directory off of trunk +implemented the server side change +I fixed the s and will commit the patch to +tests are in progress +That said its tests so it doesnt really need to freeze when the rest of us do +I can confirm were seeing it on still too even on sstables that havent been upgraded from older versions +this is fixed right? +Lets go with this for now Mingjie +Yes that syntax worked +can this be closed? +I simply used to create several documents and then search for them +Fix committed as r test case derived from Nick Wellnhofers submission committed as r +packages content that throws CCE +Scheduling to a proper release as to not bump from Beta to Beta +Converting to issue with commit of. +Bulk close resolved issues +Closed. +you will need to do ant realclean because and files were modified and please dont forget to commit interfacethrift modifications +The Struts Example can be used to demonstrate this problemClick on Register with the Demonstration Application from the initial screenThen click cancel without entering anything in the Registration panelI get an alert that displays that the From Address Full Name and Username are required +Propsal to make mail processing more robustException handling moved into loop +screenshot for irb +This the component I created to support rich text in our application +You need to your font files using the enc ansi option for +Windows Stefan I think the solution to this would be to look into the PK object to see if contained any fields and add them to the eligible fields list +I ran the benchmarks no measurable difference makes it easy +It actually only provided a default value to be passed to the getTemplate method of the Configuration object in the endpoint classAnyway I took a closer look at the source code and realized that there is an even better solution which is going with itself which defaults to the platforms encoding when none is provided +Jing Haohui thanks for your comments thus far +But I guess it might sound a little misleading if the no longer shipped items are still going to be supported or certified +The unit test was not successful in replicating the problemI am requesting that someone execute the short simple steps that I have given for replicating the problem +Thanks JasonI put this into trunk and +Fixed by the changes doen for. +Others please chime in +Marking it a blocker for just tested this patch against trunk trying on mentioned problem +Dima didnt you have a similar error fixed +This should not be hard to fix would you mind trying contributing a patch for this +added support for javabin +My WSDL that wsdljava of Axis generates additional Java classes from which wsdljava of Axis and above doesnt anymore +I have not yet tested it but according to the JMS documentation it should work like in SQL +No failures +This creates a unique way to refer to a document for editing and updating individual documents +I guess I should have looked at when I did it on trunk +msec SLOWER Djava Test create MB file +Fixed my issue so that simple ajax call from ajjsFunction would work in IE +Is there some deadline or timeframe within it should be submitted with the next release in mind +It was released with the fix in +add the following to the release notesClustered home proxies for EJB beans are now supported +Thanks Yorik this could work for my purposes itd be nice to have a choice on a basis dont you think +we need to get a test in +Attached is ptPT copy to frameworkslocaleptPT in your SDK to fix issue. +Patch coming for review +Also if similar thing applies to other tests implemented recently or in past then we should find and fix themRegardsAnil +Regarding different coloring for out and err I dont think that implementation should get hindered because we have functionality to capture both out and err in one class +Fix in trunk r and branch r. +The problem has fixed together with +Search tag elyogragspringclean +Env vars are brittleI suppose this is on Windows +thats more like a borderless modal dialog but these arent really used for editors that require extra clickingThe filter buttons could be used for it though but might be better to use dialog if buttons supposed to be used instantly +Attached is with cacheinspector debug enabled +At the moment only the NOSQL one is working. +A better solution would probably be to compare schema versionsAgreed +Assigning this to Kevin for now as I havent had a chance to finish it + Britt +Sorry for the noise +Currently we are throttling based off of a given number of bytes per second +Attaching used service configuration +Jenkins should not run with extra code +Thanks Brahma this is building the tarballs using user root or using the packagesThere are two copies of lets update both to be consistent +Its a host name problem +This is the reason why the strategy returns an int which can be used to differentiate one char and two char cases + is done +Marking this bug for this for the changes may be too invasive +It doesnt appear that this is automatically updated since the last update is May th +Patch looks great +Have some text ready to add in +Thanks +This ticket is a blocker for alternate file formats +Ive added some unit tests +I was under the impression that you didnt want to remove Forge from JBDS because we wouldnt be delivering exactly the same solutioncommandsetc +fix version needs to be changed +hope this workedshould be a diff between my working copy and the repo version of inI added an AT integerAttribute with integer added the OC extensibleObject to the test oranizationalUnits and gave them new integerAttributes matching their number +Since I dont have perms to close this ticket if someone else could please do so then that would be great +already has provisions for this by allowing you to define a cookie domain +Im going to be out of town for a couple weeks but if nobodys figured this out by then Ill take a look at it when I return +So far the slab unit tests passed on and LinuxIntegrated to TRUNKThanks for the patch Li and Todd. +Note itis not as simple as shrinking the array in dropSort as the array index is currently the sort id +What comes in two years when we have even better ideas? NNG? As I see it a new API for should either be extending the old or providing a full new implementation if you flip the switch +Ubuntu bit cores GB memory no java options thus it configures itself to run in server mode with parallel collector and access to +If it does not work please comment on the error +Of course we dont have stats right now to do so accurately but I think this is a step in the right I generated the patch for the latest trunk +I can also confirm that we were seeing the same issue with localStorage not being saved on first run +Thanks Ning + bundles have separate classloaders so the problem is how to call with a classpath of a resource defined in another bundle +This bug has been marked as a duplicate of +SamiAbout your questions thank you for looking at this plugin +The stacktrace and the proposed in revision +Verified in in +The patch was approved since it got one from Edward Yoon so the total was two binding . +I made a new issue for backport to +Maven Managed Dependencies library container does not work this way and I am not sure if we can disable choosing out container from there +The variant in DGM is deprecated so I am not too worried about it but it does mean that the proposal will appear twice in content assist +Code is fineThis portion of the javadocbq +bq +Just pretend the token in the ring is N +covers the test failure +Yes that came to my mind as well and verified that it was not a problem at least with bea consumer +correct a testcase on Antarctica whose contry code should be AQ rather than AE +Thanks Nicholas. +Try it out and let me know how it works +We could add in something saying how much RAM each container has requested to indicate its usage +Strange because the task isnt in Started mode +Needs have executed the with HTTP and also executed a few tests on JBAS and everything worked fine +Ok what should T be and why? +The patch was committed at the revision +Hi SylvainI think that you left out of the patchRegards about that RickHere is a th version of the patch +Suggested please wait before doing anything on this +Working with Cedric on this +Are you planning on using an existing JS library +This patch has some changes from included also so I will have to regenerate this patch once makes the trunk +Mayank You should Grant license to Apache when uploading patches +Looking much better +Given this we should definitely figure out the regression and get it fixed +We probably need to rethink the header if its on a mobile device +I just committed this. +Fixed findbugs warning found in earlier patch +they will need to look at the description +With the build broken I cannot test this for tomorrow +attached patch +John I think this bug has not been fixed yet +I have a simple patch waiting for this one +Client should not allow to open any new channels after closed it right? Not sure I missed something hereCurrent patch also looks good and ensuring all executors stopped before waiting to release +Ive committed my changes to make this work as a New Wizard +Do you want to accept this patch or do you want a different one that updates the README +Updated the summary to reflect the real issue +I wasnt able to reproduce it if it comes up again we can this issue. +Please add xiaol to Hive as a contributor +foo bar version same config application debug log on JDK +In SVN +Because of find bugs warnings +Need implemented as either additional or default to complete this +Nick can I ask you a question before assessing the patch +OMGYou were rightI have spent hours and havent seen that voidSorry for creating this bug and huge thanks for very fast answerBut at least there is one advantage of this report others could find you answer and my stupid problemHave a nice day! +pushed +Youre talking about a QD patch +No reason that jar shouldnt be part of Solr but it needs a version after Lucene you imagining that it just be included and that the users must use include a lucene analyzers jar themselves +defer all X issues to +Closing +This one requires changes in the archiver so fixing it would require releasing other modules +Seems to be slave related anothe build on bits platform is build on different bits slave is fixed on both x and on Hudson job +so we can get it cleaned up and off booksI couldnt find it either +Added the example code and unit are currently no plans to extend the Velocity support in Spring +Files with underscores in path and not in names are excludedThis is wrong when the container is put inside a directory with underscoresI reopen this issue +CSS comments were removed +Added a special message for datanode cluster +The patch looks ok +Committed to branch +And do you use felix or equionox +If you like I could come up with a patch to address this issueIn project we confront similar problems quite often and have a lot ofcode I can draw upon +A change in walk order shouldnt affect anyone because the order was never guaranteed anyway but at the least we should have the same sort order for both directories and files otherwise the scenario youve described in the issue description can still happen +ve just committed this. +Personally I prefer to fail user with NPEThe reason user that receives null thinks that there is no available provider for the submitted inputBut actually this is not the case since user has passed null for a required parameter and this means there is a bug in his code +Kent shall we put this to our wiki? Might be easier to edit and track the progress +Closed on the request of the issue reporter. +Remove redundant code in and +May be warning or additional comments above mutated members +Adding this to wishlist +Removed the dependencies to +I didnt have time to workout the locking issues +Fix for issue including integration tests +However we also reviewed the current solution and where we want to go and filed to help us get there easily in future +bq +Because it sits outside lib folder in Hadoop home directory +I have created for the test +Patch looks fine for now need to have a closer look +I wouldnt want the log to be async +The problem is that this check is implemented in a manner such that it prevents blobs larger than k to be written to the DB when SOLR is enabled +Moved to next milestone release +Fixed with commit with latest build +implemented in r along with a general integration test for slingvanityPathproperty name is slingredirectStatus +Let us know when a snapshots available and well be happy to try it +In your last comment there is an exception thrown MalformedThe reason is that the charachter can not be used in an but the request isTypeStatisticsEndpointNameinternalThis one comes directly from the plugin +r makes the above changes +Is Transactional imported from? That might be the general Persistence Extension is not yet ported to use new transaction extension and has its own JTA implementation +There is a parameter to the that tells it to scan for a port so we just had to change the value to true. +Ay ay Code Dup PolicemanFrom perf standpoint for real in that would be but here its fine as Tokenizers always buffer +I decrease this number from to to accelerate the unit tests +Youll need to talk to Adobe about that version of Flex +Hmm that means users will always need to specify when building a project with Hadoop right +But do file a jira describing your findings so in case anyone is interested to further pursue this will have some context to begin with +bq + +it was done and succeded + and have been fixed elsewhere +I didnt even know there was a verson +I have clicked and clicked closed and the file stopped and Eclipse +Fixed. +I guess there isnt anyone looking at this bug aside us +Ive tracked the problem down to revision +Aleksey also did some testing on his old laptop in disk bound scenario +b is working properly with +Assigning Sarah as the reviewer per the JIRA meeting July +There are reasons to support both configurationsThe design doc path allows rewriting proxies to treat applications as resources but the shortcut version is easier on users eyesWe could have a big argument about which is better but frankly they both have their place +Adding the documentation and release notes flags +May be broke +It was a hard one to crack as it took some debugging efforts to find the culpritGlad it works now +About the collapsing issue then it just happens after a few resizesrestartsreloads havent found a patternCould you attach a screenshot +what does JAXRPC say? does not implement JAXRPC +If you believe a different fix version needs to be set just do that. +This close to I am tempted to wait +This patch checks to make sure the collections object isnt null before it starts processing it +In that case the pool shuts itself down. +That doesnt work with the ESB plugin +Yes I agree that choosing the default realm as the one where image belong is good solution +Updated patch fixes the spelling +Thanks Mamta +Is there a compelling reason to keep the as well as? If not I suggest be deprecated +If were not going to wait for everyone we sent a message to should we allow proposecommit to accept ballots newer than the one it has promised on the theory that our proposer wont do anything nefarious so a newer ballot must mean that we raced with a pending prepare +Is this still failing +Patch appliedThanks! +Well take a cant reproduce this with attached test project using recent build +Ok if makes sure that blocks under construction do not add to blocksMap I will treat blocks whose length is inconsistent with NN recorded length as corrupt and datanodes stop replicating them +Can you improve logging for server instances and DC so when a client remoting message is sent and received on DC it is logged in trace mode +Attaching patch for trunkPlease review and provide your comments +In the end I decided to go a different route I am providing a way to configure your benchmark from a Freemarker template +patch applied +The build error occurred file and I fixed it atAfter I added +Hi as far as I understood that requires writing a implementation of +This means that each to line should only have one email address +Im not a Java developer but it seems like visualvm works some magic to profile things locally that it cant do remotely? I dont know what other information I can provide that would be useful +Max please review my patch +This was a mistake +Ideally we would get out of the object serialization business and delegate to thriftaveromsgpackxstreamwhatever but that is a long way off +This new patch should fix the problem +Whats your wiki name so I can give you access? Thanks +I committed this because it doesnt cause any of the existing tests to fail but it would be good if you could provide another test case that tests the new methods +This is similar to how or has used capacity defined +Noticed a difference in the way was defined in our environment versus the build environment +Im moving this to fix in R today our gives all users webdav role which allows access to the repo +Thanks Evanthika +New class source t it be possible to configure the classloader to use +You have to follow all steps from point +But when we do reach this limit we should load shed from the purgatory like you suggested +Both look fine to me +Patch v is checked in thanks Swapnil +It appears to be a git patch +Overall the patch looks good to me +Does this build include the fragments fromThe plugins directory should contain or +Thanks +In the meantime please dont install the extension. +Overall this is going to require some refactoring on how the EJB annotation processing are implemented +I have committed this +So I went ahead and rolled back the removal of the Ordered interface so the filters now implement ordered and I put the configuration code back to where it was +Thanks +A KFS archive would beharportdirabwhich would name the file ab within the archive kfshostportdir +Developers add a marker to a logging event by including it in the logging method call +The current trunk has deprecated warnings from the mapred package +Its probably +Alexis does your patch support this? This will be a nice featureEnis can you describe what the problem is with the current implementation of the Cassandra backend? I remember it was related to but it would great if you could give more details so that we can be sure that the new implementation does not suffer from the same issues +Test is unexcluded at +Fabrizio Giustina wrote the maven eclipse plugin DOES add addiction paths for generated files but only if you are using maven in order to generate java source filesOh nice! Thanks for the informationWhen I created the attached patch I first tried using the Maven plugin +Therefore must allow the principal of the primary as well +The class is not included in the distribution +has a release that may solve your problem. +Fixed in r and r. +I filed for the test util refactor +Im looking at the issue now +It works like a charm +This issue has been fixed and released as part of release +looks really good +I was looking in the wrong place +Checking +Sorry not to get back to you before I was out of town for a couple of days +Updated the installation instructions in the trunk and on the branch +I did not have access to a instance to test the migration tool properly +Running the example is something I frequently do so having it in the ant script is something Id find useful +Thats not incredibly lean now but would be even less clean whenif the build is actually properly split up with different projects and source worked around that issues excluding them in the module they are and including them in a higher level +it is in releng module of jbossideI recommend try and talk about this on the jboss ide maling list or dev forum since it is actuallythere the packaging occurs +This will now hang until it times outSometimes the problem goes away after restart some other times it persist across restart +Well in that case . +So we do have the Stat object right? The data is optional but the Stat is always fetched +Hi Milosz Ill apply your volatile patch later today and let you know what happens + test close if verified fixed. +Example for reproduce the to reproduce the having looked at the proposed patch I think we can fix this for +HiCould you please post the code and any jndi properties you are using in your test so we can reproduce what you are trying preciselyThanks the error it appears you are using the C broker and not the java broker +For example in and in a few others +The java column is the pure jvm scala implementation in the patch and the native version is which uses JNI to call a C CRC library +Attached is a patch which shows the scenario I believe is occurring +so in the scenario above clients affinity to jbC would have been replicated to httpB +The gettersetters seem to be just noise +Has this ever worked or can you make it work by coding everything by hand +But thats really more of an internal code organization thing +Committed to trunk +Let me know if anyone has a better idea of what these things are or where theyre supposed to go +Hi AllI agree with the interpretation of when an attribute should be prefixed with the namespace +If Central or Extras provide a and the user wants to roll their own with the nothing stops them from uninstalling and going off the menu for some a la carte wackiness +The fix for this is included in the patch for superfluous lifecycle entry causing tests to be run twice +Updated release notes in svn. +It works then well for any property that exists as meta property in the meta class but if there is for example only a getProperty method that does define such things then there is no meta property object that can be searched for +Dup of. +It crashes for me as well exactly the way Yunkai describes it +Attaching a new patch that just formats the new method as per convention +We can pull it back if done before +Now does a refresh for the entire asset view +The more generic issue may be that native implementation of write performs output not +I made a patch for and to let producer close the connection when the connection header is not keep alive so we can let the real clients decide if they want to close the connection or not +The only disadvantage is that we cant use different implementations at different field mappings +Attached is a draft of an implementation that uses a to bind the cache to the instance and a to provide LRU functionalityDisclaimer Im not fluent in Java or OSS contribution so there may be holes or bad style in this implementation +Karel is there any other trace of the failureA record in browser console selenium driver log +Ah I see it occurs when the module itself runs the main method. +nice minimal touch point wmemstore and I am also happy to note the chunks are being dereferenced during rollover +The Task issue type is rarely used +the project must not have any notion of JCR asit is designed to work with any kind of webdav implementation +Its likely that you will find your patch committed tomorrow morning +If this is the best we can do then it is really no security at allWe have to account for the fact that a user could have bookmarked a page when he had different security roles or security polices changedThe fact that a user could navigate to a bookmarked page and bypass all restrictions is unacceptable +to come in late the patch has gone stale +I believe both pmd and findbugs are on maven repos so one could use ivy to fetch them automatically +Thanks for all your efforts! +will do good point +it seems to work fine these days +Its probably a bad idea to be doing recovery in the ctor or anything else that leaks this to other code that assumes that the tablet is fully constructed like the MMAgreed +We could make a patch to add buildcontribclassesI agree with you Micheal Stackbecause i dont think i should make a new stater all over againThen can this stack handle this issue +Sorry about that +I think now that the test class contains just this method we can consolidate all the helper ones into ? It was required before cause there were two test cases +Not been able to repeat these can use save as to copy models to other projects +Close for all issues before incubation +keeping it open for more enhancements +added to CVS HEAD please verify. +Brett the idea was to register these with the +Thanks a lot for this works and works great. +Ive added some more inventory related descriptions to the entity +The attached patch spits out the command before executing it +Ive committed +Andrew Grieve will be looking at it this week and I believe this NPE will be dealt with as part of that +The reason it isnt wired into the icon checks is that we pass all the conformance test cases anyway without bothering to extract and test files +That looks great thanks! I replicated the database but cant figure out how to start the can you help? Cheers Chris +the earlier patch doesnt compile on hudson machine and needs specific include for netdbh header file +StackWe have not fixed +As we are tentatively planning inclusion of most or all of the TR feature set in the release I am setting that release as a fix target +I reviewed and approve the above patched for inclusion in branch +Fixed on revision on trunk revision +Heres the simplest patch that fixes it removing any executor thread limit when in ZK mode +for the new patch +Before I left on holiday I had an integrated into and mostly ported over as a FATE repo +The man has spoken I now see your point +I am going to take another look see if I can work out what is going on again +Care to submit a patch +No worries were pretty busy too +Im working on it +Rename Copy to ditto for Move +to placing this information in a file of some sort +Thanks +Thanks Lars +I retested this in alpha and it is working as expected so assuming this was resolved as indicated previously +Go ahead and file a new JIRA while we decide whether to close this one +Reopening +for the patch +Nor would T risk it if it were inside a JDBC call since this thread is not being interrupted +I think this ticket is stale the functionality described is already provided by in a modern release. +If anyone has alternative ideas please say so +Having authorization may save some resources by avoiding elaborate downstream checkingJust to clarify I meant resources may be saved in the cases of dealing with +Good points it sounds like the behavior should have a toggle ivarThe attached patch brings back the original behaviour as requested by Chris PimlottNOTE This is another way to unearth the bug in so well need the fix if we accept this patch +Nice idea +I am adding unit test that demonstrates the problem +If the devs agree then it should be patched for as well add a configuration option to enable your workaround so to prevent breaking other users HTTP authentication workaround was necessary for my work and introduced an expected behavior +If no is present beans are have not tested this extensively +I would think the same would apply to contrib +just committed this +It wasnt rocket science +If the database value is empty Id suggest using the configured default instead of failing the check +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Logged In YES useridUCL has broken our web appears to be hanging in the same loop shown for the main thread in Michaels thread +I had this patch for a few months now but no one told me how to test it in junit so unfortunately I havent provided a junit test case for itSice this patch is old I dont know if it still applies to recent versions +The ideal case would of course be for a JMS Connection to keep its physical connection alive for as long as it is in active use +Ive kept the copyright statement intactAll other files were either already part of Thrift or created by me +As long as the was previously initialized the workaround was fine +seems like a traumatic change wondering if this is really the only avenue +I had been struggling with a name for it +The mode were using is FULLXA +Use Names constants through out +if you filter binary files they will be corrupted +New patch is uploaded +Site has been revamped completetly +Done +but i could not as far as i can tell thats actually a defect of the DOM parser the spec says it should be thereEither way it should be safe and easy to check the Node type and only include nodes that arent comments or processing instructions ill try to do that tomorow. +Actually I should say this change doesnt do what the JIRA title says does it? This scripting will run a report but wheres the bit that enforces policy checking +When deployed this bringsup the regular webservices but also an RMI service +Reopening targeting for and assigning to myself per direction from Barry +Thanks Lars and Stack +I think it is safe to assume it is a regression from since it involves +Patch is applied with thank to all tickets +I am out of my league here but can be make the a variable that is passed down from theIf it is null then Defualt toPlease feel free to submit an improved patch +Support to deploy Singleton beans +is launched in a seperate vm so its actually not failing it is the gcjvm that is running RHDS that is failing +I heartily agree with Antony here please reopen this issue +If you set the treshold to then this means that the id generator is able to serve ids +Gone. +No reason why that couldnt be done though youd have to decide whether that is something youd republish every heartbeat or just when the TT first registers +Patch adds case to handle relative path consisting of exactly segments +Stefan you are right I somehow thought that both the new variant was also supposed to allow +Thanks. +Implemented in r +I have also attached the LICENSE file by renaming it according to license naming convention +Attach patch +Question Should we then improve the metadataKey handlingCurrently the metadataKey is build using componentType componentName profile authentication though the metadataKey will be too ambiguous +I am investigating +Is this related to +I will be a moderator me as a committer too ehatcher moderators is quite enough otherwise people will just be duplicating effort with little benefit +same problem here +I run some tests and it looks like it works fine. +wdyt? +We need to get the platform guys to make those to the jBPM configuration and the QS H database changes for now +Closing. +fixed version now looks like of applied the patch in revision as proposedThanks for the contribution +Even if the calls didnt create a field error there might already be a field error around from an earlier note that you can choose to only show the first error message when rendering your view +Please feel free to reopen if you think it is not +Do we know what the performance cost is of making TokenThat is the key question +asserts are to validate preconditions and other assumptions made by the developer which includes the check mentioned by Dave here +I dont mean creating I mean integrating it with JSF +and are flaky +As discussed on IM theres a cost to scanning all results but it doesnt mean extra time waiting for RPC responses since it already waits for all +For Iains issue this is an issue of your application not taking into account the good as both ISBN and EAN are unique identifiers as well +By the way I did use the deploy mojo to create the that are currently deployed and unusable +see also for more believe this issue was fixed as was closed and weve since done a lot of testing on the samples +The patch looks good to me too +I will make this change and resubmit the patch +I just committed this. +FILE Change this Specifies a new boot password for a database + thx +I will submit another patch after testing against the clusters +ah yes +Does anyone have any thoughts on this being added +Patch from Andy +Is a branch? If so you will need to push itThat sounds like a timing issue +I do not expect anysignificant changes before the final releaseAbout the quality of documentation Do you have any concrete complaints? You arealways invited to provide patches to improve the docs modified the so that we have now two links to Javadocs One tothe latest version in SVN and one to the releaseI manually copied the apidoc on the server +The tooltip location for pies was not always correct +Just remove the assertion +I just tested my example project on and Im still having the same issue. +Then the RS is let inWhere are you when this has happened? Just started? What session has expired? The RS in ZK +I notice that it is not deprecated since there is no replacement in the new API +However +Thanks appears fixed in most Xalan API tests so far +It is not your fault but that is sure an ugly name on a method get +I will take care of the property while attaching the patch for +We do need to make sure that we run all the tests especially on branches before committing because branches need to stay stable +Closing. +Seems like I missed something +As far as Im concerned the issue is resolved +Patch applyed +for the on the directory for nowIn the medium term we need to remove the job client doing anything to the system dir +From Jonathans comment we also want stats like of rows total of columns avg columns per row total of versions avg versions per column avg versions per row etcPlease let me know +So Ill fix these tests and commit the patch in this morning unless anyone has objections +Not quite +The testsuite test for security annotation on annotation includes The testsuite defines two rolesuser and three user with role annonymoususer and adminaccess the EJB methods +Thanks. +I agree with you about defaulting to the old behavior for now at least +for server logs see report this test there were driver threads distributed over driver nodes RCM per Hot Rod asymmetric cluster tests Im adding are including a fix for +Effectively checks the vertexId while trying to assign a container +Then I sync with BRANCHOne question Is it possible that after a player has been chosen he doesntappear in the selection list any moreKeep on your excellent workBestm not sure what you mean +So outside of this it might be another type of error outside of the Cordova code. +My search result when I allow potential t reproduce what you see +If there are files it doesnt know about it leaves them on disk +David is this still applicable +svn statM javadrdaorgapachederbyimpldrdaCan someone please commit this patch +Argh +Forum reference was to the second pulling HTTP during config +Are you running Ganymede? Im seeing this problem in Ganymede with the latest version +Does shown in the JNDI tree a display issue or is it missing from the JNDI treePlease explain more +Make sense sorry +I attached my version same idea adds INCREMENT DECREMENT and COUNT +This would be very nice from a dependency management perspective +There are blockers that are issues with weblogic that will keep seam EJBs from working +I think we should never go and just return true here we should rather fire up a true error here like a corrupt index exception or something like this +From your description it seems like you are pushing the edge where derby decides to use optimizationor not +For Panel Stack there is some issue with the code that the builder plugin generates and Ill look into that some other time +Attaching the patch refreshed for that branch +Please provide a working code example that isolates and reproduces the problemKeep in mind tht has been in use on thousands of production systemssince and would be useless if it corrupted file transfers +The patch failed these core unit tests +Any chance of your taking a look +Works great here now maybe stronger caching than just a weak hashmap +Lucene doesnt have an place to store arbitrary information +Solved thanks for reporting +Got those but do we not include known issues and workarounds if there are any +The changes to can screw up when you have explicitall +Thank you for working on this +The attached patch fixes the issue for me + to getting rid of the +There were a lot of changes in the part since so I guess we have to look there for a possible performance loss +Finish reading from Reader only when return value of read is less than Test Execute derby all and found no error except for andSending javadrdaorgapachederbyimpldrdaTransmitting file dataCommitted revision +If it was it wouldnt work for anything and would bleed everywhere +Worth doing +ASL release process is already an enormous pain the in rectum and I am reluctant to make it even more complex because of an optional component +I assume I dont have to patch any code that says it is +But there exists a library Jansi which provides a simple way to clear the promptterminal +there are known limitations when combining identity pk gen with sequence pk stale resolved issues +Yeah I wonder if we can introduce an attribute in the xml namespace so you can configure it on the kahaDB would be great thank you +Yep Ill update with . +If you go back and look at last years discussions regarding pluggable handlers youll find the following in the archivesThe proposal I put forth wasnt just but pluggable handler packages +So I think that no need to update the current exception implementationThanks +Its not clear what is the correct course of action is for this test +Please change to if you intend to submit a fix for. +attaching AM logs +closed. +Thanks Konstantin changes are good +Issue which added aggregator now in r we have a properly failing test using the invoker plugin +How much RAM is on the machine? Have they tried adding more +I have resolved the requirement for reading WSDL from a DOM Element but in a different way to the suggested new readWSDL method +I applied Izaaks fixup of LNs original patch though little discernible improvementRunning the with the patch we spent about less time compacting in total but on test completion there were data files in the filesystem as opposed to when I ran without the patch +This looks like a better solution +The same works fine using curl without knox in path +The issue is that JSR annotations dont carry the necessary metadata for this wsdl +Submitted Sebastians patch as it seemed reasonable and hes reporting success. +And its marked as known issue as discussed here and on mailing list + is indeed in the wrong place +The issue wont dissapear from the database only its status will changeThanks +WRT yes I see that it supports +Not sure its worth it +If you find use for it please do include it somewhere. +On another note it still seems like a bug to me that UIMA does not behave this way +The selection of the differential period should be available only when defining the filter +work going on here? Otherwise I take it +Patch reattached from bugzilla +Baby steps towards real solution +Hopefully I havent made things too confusing with + + to trunk +I never tried that scenario. +And the bigger the cluster is the faster it will beThis version is for review only +I understand will only do one direction +I tried Saxon and its doing the same thing +Assigning this to Rich to do the actual Patch work +Now it compiles and runs +Weve seen similar exceptions when loading data that contains text with the column delimiter in it which produces shorter than expected tuples +Tokens are not cached +You could even make an alwaysraw binding prefix to do what you wantBut its not something that should be changed in Tapestry thats my anyway +though +Verified in +new patch updated to latest trunk with cleanups +This seems pretty useful +These tests are going to have Ignore added to them +No further advance can be done here in branch +Dont conflate their versioning w the suggested WAL log versioningRegards PE data its data is not amenable to compression +it shouldnt +I merged from trunk into ran the libhdfs test and committed. +Vijay you need to run with two places to see the bad place exceptionI think the exception might be correct but I want to look through the code more carefully before deciding that +If you just enter at the start of the path name how is the parser supposed to know which drive letter to look at? I imagine that the behaviour you want is for the parser to determine which drive the command is invoked from and use that drive I would suggest thats a bit much to expect however +However I dont see any compelling reason to have it there especially if there is no other need for a custom for taking my input on the view code Ive enjoyed collaborating on this +They were thought very early on when the Hadoop DT token story was still not fully backed and we had idea that the client would have to give the DT token to oozie on submissionWe have to see if supports making the type change else the ooziedb tool will have to handle the change for each DB type using the corresponding sql + for the change +EWSRC both TC for RHEL +This bug has been marked as a duplicate of +Configuration of bean connectoris done via and +To cut this down change your server timeout to something like and then your server will be marked started in seconds +I agree on the consistency front Robert Ill go with yours it looks is it ready or need to make any last changes +reopened just to set the resolution to fixed +embeddedejb +Working in I found that svn was not native in next files which was added in this taskadded filesjavadrdaorgapachederbyimpldrdajavatestingorgapachederbyTestingfunctionTestsmasterjavatestingorgapachederbyTestingfunctionTestsmasterjavatestingorgapachederbyTestingfunctionTestsmasterjavatestingorgapachederbyTestingfunctionTeststestsderbynetjavatestingorgapachederbyTestingfunctionTeststestsderbynetjavatestingorgapachederbyTestingfunctionTeststestsderbynetjavatestingorgapachederbyTestingfunctionTeststestsderbynetjavatestingorgapachederbyTestingfunctionTeststestsderbynet +Committed to master. +i felt a bit strange to me that the indexhookprovider used to be merged both before and after the userinitializerIm having trouble following what you are saying +corelibmanagedcompile +If you want to have data in your DB then put the creation in Bootstrap +All the regression tests passedCommitted revision . +ssl certifacte needs updating open a ticket +So adding log output in dev mode would actually be the quickest and most efficient in terms of getting something into the codebaseIn that case Id say go for it +to patch committed thanks Richard +Can you try with a more recent version of CXF +The test also depends on the ability to set on +The next FAQ will be Where do I get a copy of + added +Its a style thing so you and I can debate on it forever but I think that we both agree that its important + for this to trunk. +I think this issue can be resolvedclosed the new layout of the site has changed and this is no longer an issue. +Duplicate to +Patch committed to trunk and branchThanks Daniel for the patch and Eugene for the review +My patch replaces the by a declaration order thing +Im going to try showcase with latest AS snapshot +I tried it wo a profile and forTodd I updated so that our build now uses this new flag insteadThanks for the patch Joep +Tested locally seems ok +Other tests also run fine +Thanks +Any chance for this to be merged into trunk +Calling it fixed since the original code is now ported to Java. +Well its really unprofessional not to release stable releases once in a while for a product you ownmaintainAnd if they cant see the value in that then they will loose out as people will go find something else +Just need to put the xml path to the main method of this class +Thread dump where you can see threads of Update getting created possibly for each bundle which asks for ve applied a potential fix in rev which splits the resource resolver factory implementation into two classes a common singleton instance shard by all instances of the factory +I am asked to undo this patch +If that is the case threads will prevent garbage collection of the scheduler tasks and eventually block shutdown of the jvm which seems unnecessary +guess annotations guys dont like codegeneration +It is fixed with a newly converted test in +This parameter is not supported by JRE +Bulk closing stale resolved issues +I have an interesting remark it works with just tested a hot deployment of the example EJB with the following configuration GA GA JDK it did not exhibit the error +I guess we target WFK with this +We changed behavior in startup earlier. +No hard objections but one worry that it might become confusing to use +I cant believe we missed this! I think we should do a release to get this out soon +Jacopoi have no idea what you want to achieve with your questions +The patch involves the following +Uwe let me know if youd like me to work on a patch +Hi a sample is welcome +XufengCan you attach test suite result for? patch for trunk +Matt please test this version either out of subversion or wait for the beta +fix some unit tests +Archive of on machine A containing config you verify this occurs with TRUNK +I agree we should set Java explicitly +The patch looks good to me +Well +Closing this since has been released +Awesome Ryan good to hear +This configuration is added to the SOA platform by the platform build process and has never been in the project +I believe it has to do with some stuff that was added for versioning purposes +r jwaldman linesChanged paths M myfacestrinidadtrunksrcsitexdocdevguide review and fix values for platform in skinning documentationthanks to Anand V Nath for the patch +I have since moved on from the project and am no longer in a position to test this personally +Sorry for the late response I did not receive an about a comment being added to this ticket +Also it does the same for names of components in the extends meta hierarchy +patch added +looks good +Right they are not meant to be +RonI had a look at this and I cant find the cause for theAlso I tried to deploy your example to my SNAPSHOT but you are using Fuse ESB andthis doesnt seem to be compatible with the standard apache servicemixSorry seems I cant help here +This is going to be a fun one +Thats not the original bug report but it is where this is nowThe test program now runs fineUbuntu +Eddie you can pass this to Rich if youd prefer but thought you were most familiar with the use of the +Code was being reviewed in +I do agree with can keep it in. +BULK EDIT These issues are open to be picked up +This issue is as a result of so closing it +If assign element it is ok +I think this problem must have solved in some other solution +So for now I will just update the documentation +Bumping to because I dont have to research before investigated the code and the issue can be closed +Thanks Raghu +The error logs are quite similiar I think its the same the hardware is alright cluster is small but you only have client inserting rows so it shouldnt be that badRegards its weird to have a conf missing from care to open a JIRA for this? Thx +Slip due to time no its done +my id is are in the incubator unix group. +i just checked +Jens would you mind joining the discussion on our developer mailing about design issues like this one +I added the JAR to my project but I suppose it is not that easyCould you help meThank you very much can you give this a try with the latest snapshot +Doesnt sound like it an issue anyways +Hi Stepan Maybe we shall ask bouncy castle guys how to deal with it +Even before it was set to +Moving out of as it would break the API. +Committed the patch to both trunk and +So parking them into lib or corescollectionlib does not work? The scripting engine will not be dynamically loaded +Thank you Jakob +Thanks for the great work on this Mike committed the r patch +I experimented with enabling widths but this will require substantial changes to the calculations that calculate the column widths across the multiple html tables used to implement the m going to bump this to RF where we will look at a of the EDIT code to work with the new standalone javascript widget architecture. +AliCurrent SVN has support to launch activemq and run the old testcases +This is coded but Im unable to attach a patch +Hi Eranda thanks for the updated patch +I moved two static member variables into a method body but dont think that was the issueThe issue I think was that for some reason sometimes an Ngram wasnt contained in the table of transition probabilities +fixed in build +The only cost is the task serialization time +This patch made it much easier to output JDOM to a stream using s patch looks good to me +Itll probably be announced tomorrowAre the Lucene test results with Harmony accessible on the web +Ik just reviewed the patch +Patch submitted +Both of those problems relate to sharing code to process the actual bytes from a Chunk and that can be built on top of this change +Is all that is needed for our usage +A related issue is the sourceforge bug of +A DSA key size may be multiple of but it possible that it is not supported by an algorithms provider +It might be this one you have in mind this issue is same as +This can easily be extended to a service level or service group level parameter +But lets see if the problem disappears I will know in around an hour I hope +New patch on top of Jeremys svn property fix +This patch moves the erlang thriftjsonprotocol to the contrib folder since the external github dependency on jsx for this routinely breaks the build +Keep em coming. +After this bug is fixed I should update the functional spec and highlight this behavior for mentioning in the user docsa If you try to retrieve a UDT whose class isnt on your classpath you will see a class not found exceptionb If you try to set a UDT parameter with an object which isnt an instance of the UDT class you will get a coercion errorThanks +Thanks for your helpful bug reports I think the server functionality in the plugin is much nicer as a result +The way I see it there should be a bundle exporting the interface implemented by +Ive updated the issue title to better reflect the issue the forum comment describes +I think it does not because for MUSTPASSONE it keeps the min rowkey as the hint while for MUSTPASSALL it keeps the max +Applied the patch with thanks +Compiler warning suppressed at subversion revision +If this effected any improvement in running time it would probably be noise at best and likely inferior to a better configurationI dont usually like tagging types but the interface will not only resolve any case this would but could also help with impl table stores etc +If I cant by tomorrow another committer will need to as Ill be unavailable for a couple weeks +Give the admiin full power I say and provide the optional switches to filter for convenience. +Arvind this is a good one to work on next +Any inputs +Thanks for the quick update Amar +I believe SCP server should be removed honestly +The target date is fine however the snapshot is not ok because we dont know what we are testing against +xceiverCount should be set too or there will be a stack memory overflow or reaching other resource limit +Can you try writing the test so that the types match accordingly? What might be happening is that iBATIS might be attempting to use an Integer to fetch a number from the that is too large to store in it +I take it back +Do we need to upgrade Tomahawk in order to use JSF? Thanks +It is only useful in boundary cases where the default bidi algorithm behavior doesnt process boundary weak and neutral directional characters in the expected manner +Is that enough +The hack is a possible solution but there is a easier one +So this will require some container hack +Thanks Chris! +looks ve committed this and merged to +Thank You +Could someone remove the RESOLVED status from this bug +Adjusted tests to accomodate this weird issue in Android +for and trunk +Fixed +Werner unfortunately I was awaiting feedback to the commentquestion above +Thanks for your support Guillaume +see for that. +Hi MistyIve got this issue during testing of EAP ER on Sybase db again +Hi Olgai attached a patch which is against revision +Thanks. +I am glad you are paying attention to details +In case of real eigenvalues all entries are +I dont see the By keeping it around all the time on the component we can always lookup the active consumers part in the fix +That same fix for the percent is supposed to also be applied for the XN transactions completed part so that it shows up as +So if your DNS server responded with then your host principal would be host +rename to +The updates have been verified in the ER build +Can you please confirm what version is appropriate for EAPCheers Core Ref Guide up on stage now +let me try to rebase this +Bulk move of M to is moved to next move from to to to +Thanks alot kons +Think I just found another keyboard shortcut or something +the additional ESB definition preference page is for those who only want to develop ESB with a standalone ESB runtime so we should provide a way to configure ESB project classpath with it +There is no coordinated effort between the and me things always change and we dont guarantee any compatibility between the two +Ill commit it soon if there are no objections +So I thought it would be a good idea to identify a problematic bundle and not try to install it again and again +Attaching an updated patch +looks good +Thanks David! +jason Is this still relevant? Or is this all fixed and done on trunk +cannot replicate it on recent Kepler rc build +Attached is the bundle containing the fix +Max I think indeed that running in the same VM will be tricky though maybe not impossible if it happens in a thread of its own with an appropriate classloader +for this I am going to a new jira to increase timeout +Further edits are likely +gant files are not compiled automatically +checking the number of set bits etc +should this go into +Closing for for just committed this to branch +A totally missing jar will however cause a while Jasper is trying to compile the page +I need to send a proposal for this event +Not all Hive artifacts are pushed only those needed by Oozie +reopen for fixing relevant so rejected. +I still dont see the questions about what do you guys see as a legit components being answeredbq +what is meant with this +The also uses a Callable for starting instances and also for cleaning up the failed nodesIn order to be able to tests using mockito I had to extract some static constructs for example the creation of map is factored outside of +Looks like the issue is that it cant determine data size +Manually testedIve just an issue with the Project open reviews widget which doesnt display anything with Derby but I guess it would be better to write in order to detect all possible issues due to DB +Chris thanks for reporting this +replaced the original patch with the the latest +As such were going to hold off on solving this issue until after +Pending application to trunk +Committed to with revision +Thanks Aaron for your response +any news on this? Id like to start making modifications to some of the samples code for support +making them public would potentially encourage developing an inefficient query processing engine in which generated the AST and then the walked that AST +Not working in trunk for me +Though it can be done when using the binshell script or binclient script +I think the ideal scnerio would be A way to add behaviour to the render cycle which allows a contributed service to determine which any given component in the system is passed + +It seems necessary only that be processed before the bug exposes itselfIts not necessary to process with the same parser but it is necessary toprocess +Reporters of these issues please verify the fixes and help close these issues +Did some additional tests +The bug was originally reported against so it must have been fixed in the meantime. +This is pretty easyI agree on synchronously waiting on the CRM in the listed scenarios and a CV would be a good way of doing this +group partitions according to Table +Looks good +Yes the default is hours +Please ensure your server is up and exposes its management ports via the system propertyThis is not a real critical error indicating the servers general health but rather it means the mbean explorer cannot connect to the remote server via jmx since its jmx is not exposed +You can patch it locally but were not interested in maintaining that code. +The discussion on this ticket is more for the overall proposal for two separate releases and roughly what they are +Would you be able to get it working there? If not Ill take a stab if necessaryThanks +I think I found it +Have you tried setting the poll within the configuration parameters for the +I have run Freemind of version +Good jobI would suggest the inclusion of at least the height and the reference system usedGreetings and thanks for your work +Maybe its not related to thisJon Thanks for running our proofing +Please verify +Obtained usingsvn diff +it is very inefficient to do a class loader lookup by reflection instantiate an instance and call a method on the samecan we quantify that +fixed patch for the patch to have two separate +Hf however the question is if the input is broken or the reader +JeffI am afraid we can not do that because has a dependency of we can not make them +Hi Chrisjust curious +Somehow missed the earlier questions +Thanks BJ Reed for the patch! +Hi MartinI was just trying out your patch and was tripped up by the references to an constructor +I have already finished with solving the issue +If you filter out with a filter before that checkVersion is never called if the filter returns INCLUDE itll call checkVersion and incur a seek +what would the other dependency beim sorry to hear that this is too much especially since i put quite some effort and time into it +Since there was no further input and I cannot reproduce this problem Im closing it for now. +latest patch after code review +This bugs priority is marked as Minor but its priority may easily increase due to the of creating new XML schema global element declarations as described in screenshot of Tests results before the patch +I just set a property AXISHANDLERALREADYVISITED after the incoming phases are passed from sandeshac +This is not an issue in as the replication logic has changed +OK my version is now working +Java code looks great +That looks like a better approach because Hibernate Search might get validation groups as well so they can be linked +Thanks Kristian +Thank you Arpit for the review +If there is then we assume that the writer is being scanned and only process the first item. +Thanks for the patch Ill look at it later today or tomorrow +But we would like to be able to recreate the original issue before closing the issue +For clients that dont like that they have to write wrappers around the thrift API +If you make it public it will lead to more help because more people will use it +The struts and dependency plugins are examples of this and the reactor takes this into consideration when building the plugins +The second patch has been committed in subversion repository as of r +or calculate approximate size using avg size of few messages +I guess the work on this one stopped? You attached the patch but also said Let me get it working there first before adopting it for +So easy fix? Weird it has taken this long to show +Im enclosing a copy of the larger pdf file +Committed patch in revision +I used any anyType as an input parm w the style and it the WJ gend a input parm +Geotools provides the version and geotools works with the version +Committed +Reopening to add a system switch for padded comparison +Ok heres the proper functional document +I need to look into the issue before we go further though +Thus it is wrong to interpret r as a line breaker +Can we help get a latest build kicked off to see if that succeeds +However instead of doing this it pulls the files that were not whitelisted +Sorry for the checkout done +I updated the patch +This column only makes sense in the queryAll context query will never object where isDeleted is will override the isDeleted column using the default value property +If you have any questions please let me know +Awesome AndreasIf you dont object Ill add Mr Tiger as a Rendering test case. +For solution see as in cvs for +Patch applied! Thanks! +So do I wait for another iso image or can I go ahead and build from source code and verify that the fix did not break anything else +I may apply this tonight if I get time +Do you plan to run more tests or is this ready to get in +New patch that also changes and my bad I was applying your patch to the X branch instead of trunk there it works like a charm +Best to get this change in now rather than laterThis version does not include the kill method added to the regionserver by previous version +Yes we have assertions disabled +The schema name derived from +Will double check but I expect double row in message with content in case of copy +Changing it per document is pretty much out of the question particularly since its too late by then and Ive lost the initial memory allocationThe report Im referring to is a different bug that I filed years ago and we closed after changing the allocation sizes +Status seems like a reasonable level to me +The init method then calls loadServices which will create the timer if hot deployment is enabled +Can you maybe give it another whirlAlso have you seen the class? It allows you turn a Document into an for xfire to parser +Better post links to jobs builds which are marked keep forever so the metadata doesnt vanish +Thanks alot. +im right +I add a test case to cover this but exclude it +As discussed link to added +I dont have access to this in github but if someone else does and can make the change to trunk and save then awesome! +Somethings bugging me about this once because looking at the code Im not sure why its failing +Support for handler methods is added +The patch has been appliedPlease verifySian could you please create patches against Harmony class library or module root directory as it required by Good issue resolution guidelineThanks in advcance +Only the latest version supports the FETCH clause +I have added check at both places +It will identify itself as follows java jar Version passes in hudson. +deferred until CP +Sure thank you +The knowledge of Bugzilla among the Maven devs is probably lower than it is for JIRA so we would probably need help from the some Bugilla in the community to pull this comparison to Maven SCM is both a good and a bad one +Im wondering what kinds of issues those were and am thinking Howard must have thought they were worse than this +Thank you! + +Move hadoop archives into and harfilesystem into common +there is no point to spend thime for this now +Created +What Nick said also metrics like this dont take into account s elastic nature +I agree with the fact that there is no way to configure infinite fetch failure backoff anymore as part of this JIRA +It may be an bug it appears that the final buffer is exactly byte too longand the splitting of the buffer splits it into a byte portion and a byte portion +In your step you have added a zone with a Xen host in adding S image store our code will download specific system vm template based on current available hypervisor and thus download Xen systemvm template to your s +from what I can tell this was a commons issue not a Maven issue +We need to revisit style based label just this bug simply says we need an event model for style changes +Im all for making the compactor smarterGreat work FilipeI wish we could accurately make this equation work filesize datasize postcompactionfilesize but it seems overly complicated to try it would just be a nice API behaviour that isnt required for any of this +Could not reproduce in either case +incorporated comments +Thats right the duplicate is a shallow copy of the buffer +Its up to the user to add one if they so wish +Thanks +Jacopo thanks for fast commit +Initially only one node of a two node cluster was impacted +This shouldnt create any backwards compatibility issues +looks like Ill have to change the way we test it +I will try and do a patch tomorrow +Should be even easier than that +I also tried on but failed with same errorLooks like a problem with the STR transform Ill take a look at it when I find the timeColmHi NitinCan you run this again with logj logging level set to DEBUG and post the debug logsColm +This feature is really needed especially when appending items to an existing order +Can you please set the Affects Versions field to the release where you observed this issue? Or did you observe this issue on both and +bq +to running on Android too +Please let me know if that is not the case +Heres the patch I just committed. +The flag by default is false and it wont try to autoscan your project for hibernate files +oops +Ignore my earlier comment +Why does adding fixEOL after replace not work? If the EOLS are incorrect shouldnt the task work afterwardsThe reason the patch was not applied verbatim was because in the posted state it attempted to fix EOL on specific files on Windows regardless of whether or not replace tasks were run +courtesy findbugs report +What would an appropriate test look like? I dont recall seeing a test for s borrow method but I could have missed it +how to do thisSo instead Im going to now download the latest and ensure that we support the latest micro format +For those that dont know the approachGiven that exec is supposed to replace the current process exec of a ruby program wed like to stay in the same JVM exec of any program wed like to avoid pumping streamsI propose that exec instead of immediately deepening the stack to call the program in question and having to pump streams and other nonsense should immediately take the command information and throw it in a Error +Patch is committed. +At first I think using sed would solve the problem however I found sed is not good way when trying to make the pattern for sed +I though that I fixed this trafficlogstats crash in ddfebcafdbefeedd +Rather this patch is intended to enable developers to work towards that goal without maintaining a separate branchChanges in this patch include permit Ivy to find unreleased artifacts in the local Maven repository for projects +I am wondering why you are not using the RPM plugin that is already built and working for a number of people +Once the software is stable the wire format should rarely change +OK my bad then +Ill put the optimization in another patch with a separate issueI suggest the title of the issue be changed Id do it myself but not sure what the reaction would be so erring on the side of caution +But left as +Since it is possible to use the existing configuration I dont want to change the namespace to support yet another syntax all you need to do is set the maximum sessions to to allow unlimited logins +But Id still like us to squeak it in if we can +Please add affects version +I am running BTW +deleted that method and did import static Its my mistake when by IDE +Let me know if you have already made progress + support the out of box if we dont miss up the spring and spring jars togetherI just added an unit test to verify it +Patch applied +Uploading the right patch +Even if it would be interesting to know execution times under different conditions I think we can life without them for the time being +Hi JamesIn this situation I would go for option of Roshans suggestions +thanks KT moved it to jboss tools and Sergey will look into it +NODECONNECT +This issue is also a duplicate of +Ill definitely try to have a look at it next week +Could you please check this +These blockers and critical issues are resolved but not verified +New tests look good +fixed test +I was using RC +Thanks for verifying this issue Dominik +the result can be used directly to query the sessionsevents table +Thanks SujaySome commentsquestions Where will this code live? In? I think itd be useful to attach a patch with this code integrated into Bigtop source +Would a quick meetup with the people involved help iron things out +If specified as it indicates a request for the IPSTS to the user for authentication before issuing the tokenSo if the RP passes wfresh what should the subsequent Expiry date of the STS issued token be? According to the above it should not be longer than the given value of wfreshColm +Since the datasources have their own implementation specific names perhaps one solution for someone wanting to tackle this big project would be to create a new set of Datasources with new names that have the consistent behavior. +Better patch +Hitesh you have a point +Basically if the returns null we treat the result as empty string +Brian if you use the same cache container does it work? If so can you not just do that +Thanks Jimmy! +I mentioned it in +Integrated to trunkFrancisCan you attach patch for? patch +bq +Not sure what is the reason for that but Ill try to find out +This is a nice idea +I agree that a better test would be useful but I think your checking the manifest is a good and adequate check for the successful backport of this change. +The Database one might have the same defect though I didnt test +for following up on this Jim. +rev in the documentation developer guide +Thanks for review N +makes sense also needs to support as well +I suppose +Should we remove the header System Tables +Question is it only removes server if HADOOPOPTS starts with that +Just public methods in the list +I think you are probably right and that theres not enough justification for a totally new method for Lists so I am closing this issue. +Can re refactor out the duplicated code into a shared area +Use it rvm use +The code change will be limited +As per the specification section it is a producer error to sent both a markup string and markup binary as markupString and markupBinary are mutually exclusive +The test result say core test failed but does not say which test +Thanks ScottIndeed Ive been doing too much PHP and javascript lately +It will require adding more new states that will make the whole thing more confusing and +ah! The quotes +Posted a patch to abort the task status if log splitting failedLet to handle the duplicated status entry +OK I just committed a fix for the null pointer exceptions changed to use internd comparisons and a little restructuring +This could cause setwatches never to be set if the count is initialized to for all the three watches +The next version of XNIO is going to be using a Classloader based mechanism this will mean that the SASL providers will take priority just by being on the classpath manual registration when using Remoting will no longer be required. +fixed at revision +Correct migration +will commit after tests pass +following example is commonPrebuilt ICU binary are only available compiled against the release to manufacturing version of the Visual C runtimesRecent MS patches for both Visual Studio and security fixes for the runtime now require compiling ICU against more recent runtimesICU does not compile against newer runtimes but ICU does +Hey VishalAny chance you could get some sort of test for this soon? It looks like it will be a merge conflict with so it would be good to have this in first since it is a rather nasty bug and Id like to make sure we dont make the same mistake in the spaceThanks +I think Amber should get on with their lives +that should hopefully fix the issue you have reported please test it and let me know so that we can close this issue +last version of the patch +I forgot +The new flag images might be a different size and require updating of baselines and Asserts +see it +Addendum for trunk demonstrating how master can respond to new exception which cannot be ignoredOpen to better naming +Im all about classic +Looks good +George your patch and test case look goodFixed in luni module javautil at repo revision Please verify that it fully resolves your problem +I think the warning should be there +Thats sort of unclear actually +The only one issue I want to address is that is in support package in which users shouldnt need to access the classes +Thanks for reporting Gardella! I have updated the WIKI page +Hi Tian HongThis is a partial backport of which is already in trunkbranch but missing in branch +Additionally if we in advance we simply end up with inefficient compactions for example in this case we could have compacted small files in one go but would have ended up compacting and if it spent less time in the queue +Its recommended to use C switch +Is this is a known issueI have attached my client code for your referenceOnce again thanks for helping me solve the problem +Xmx OOM? no no no no no no no no Xmx OOM? yes yes no no no no no no Xmx OOM? yes no yes no no no yes no +cant reproduce the for input +But all other setters in that mbean start with set +Ive also added blacklisting mechanism +Reviews are welcome +Test that the tells the system adminstrator that one volume of a datanode is unavailableAFIK currently does not support functionality of explicitly indicating volume failure in data node so we can not include this test caseThis is good to have functionality opening new Jira for it +Ok we found the problem +This issue was not scheduled yet +Thats fairly completeYou could write a new job to finish the recommender computation too +In that case the will log a warning on startup +Targeting for now +pushed + has been addressed Im okay closing this ticket if we can get through a day of randomwalk without hitting that issue with +Ill let you know as soon as I have something working +For your reviewing pleasure +if it still happens please reopen the ticket +So things like this really need to be reviewed before theyre made +Hi EliThanks a lot for taking a lookInitially why i did not provided separate tests here is will completely So actual testing will happen through thatI agree with you +How does this compare to the version +Just have to do backup and these other clean up issues + +as r to trunk. +And before changing that I need fixed in the first place. +What do you think of this version? I broke it into two patches eliminated some copypaste code and moved the function docs from the client to the interfaceOh by the way awesome patch +And the try catch block added may not be needed +In future if we do come across users who run into this problem it will be easy to track this jira down since the NPE stack trace is clear about where things are going wrong +SPRINGCLEANING + +Bulk closing stale resolved issues +Following the conversation on the issue I attach a patch with a new and a sample application +Havent had a deeper look yet and Im not an official developer in any case only collecting patches for my own uses but it sounds too good to be true +Currently its both +That logic is no longer needed since the now works +up to your sensitivity +still has a bunch of Root related stuff +Config management for envs is already burdensome +There is already a but I dont think its working +I guess the new main package is to be added as well +Reduced priority of non critical issues which have current m not sure about the name maybe instead +Should also replace uses of Guavas patch that switches from Google Guavas Charsets to JDKs +Attached patch to fix docs +For details on whats going here you can take a look at the cms security code in the package and the class where the security data is queried is the did notice that there is a small bug in the logic which uses optimized Hibernate Query when the being used is not the this logic should be Use optimized Hibernate Query only when the the the core DB will fix that as far your requirement is concerned you just need to provide your own implementation of the Identity Module which takes the approach of splitting the data between multiple databases +I dont have the permission to set it now. +Need documentation specifically warning users when attempting to load test their application with seam debug on +It would be good to have someone independently verify this +for the patch +new version of the patch fixing +As this worked on so i am sure it works with Shindig +But first I want to fullyunderstand the valid uses and dangers of this patch +This was fixed +Works for me now +But I suppose the order in which it load stuffs is very platform dependentbq +Thank you John +bq +But you can improve its chances by casting a vote for itve come up with a fix for the first part of this bug +patch file on that fixes the specific issue you faced +Closing this issue now +Ok +Committed to sqoop +On a personal note I think excluding length from packets is risky +ACS feature freeze date is November th +patch with at rThanks at please see commit log +Dont ask me how I did that +The problem being that theres already a bit of native code that gets compiled via Maven plugin and Im no sure that plugin allows for multiple native subprojects +Theres no history on this file in SVN right now but the and current trunk versions do not include the tags reference +One of problems is that this test needs database in previous versionEngine implementation of the versionwhich are testedcan not create database of previous vesionThen how to serve database in prevous version which are upgraded in the test must be considered +as mylocal repository and the error has not occurred again +Could someone please try this patch +User can specify gdb parameters in +Testing now +Removed unwanted dependencies +I checked in a fix for this issue +Fixed a couple issues with maxApplications per queue settings added documentation add more tests +Ok good point. +Could you change the message in the Note on snapshot deletion section as followsAfter having deleted a snapshot its still possible to recover it as long as a new project analysis is not run or as long as the deleted snapshot has not been created more than hours you also use a bold type to display years and months? and pushed +Maybe someone who uses Solaris can verify Im fine punting Solaris to a future change as wellbq +In the admin can currently delete them that is fine but renaming would be nice +I have the same setup and increasing the stack solved the crash issue with +No one has complained about it up to this so well wait till someone doesId just removed it because didnt think anyone was using it and I thought it was broken. +An application can be using the jar that I uploaded with this issue +Not with jing I hope the error reporting is atrocious usually amounting to something bad happened around line X +war and configured in the example solrconfig +fixed in head use tls wrapping function +Forking is not working on Windows +it works on the IDE and dont work when java jar invoking +Click Delete on the newly created blipIm pretty sure this bug occurs when the youre editing a blip in a reply thread that has no other siblings and that reply thread gets deleted +Commited +HiIm new to Solr and OpennlpI have followed the tutorial to install this patch +Very helpful +While I was in there I also made it illegal to drop a schema that contains sequences +FF works fine +The parent property of a subworkflow doesnt point to the ID of the parent workflow well need to fix this in order to handle subworkflows properly +bq +Fine with of a bulk update all resolutions changed to done please review history to original resolution type +But if your git repository is current I can pull the stuff from there +Updated patch with some bug to add joda support now prepped for inclusion please review +Pro RHX Edition +Thanks Jan! +to clarify futher selinux uses the webpages for its teststhis mean setting up the nightly builds so selenix can be runthough you say there are many ways to accomplish this I have not seen anyI would be glad if you point out such methods that are available +If the attribute is not present the default should be parsexml which includes the docbook xml which is what I believe is currently happening +Please confirm the current status of this issue and if its incorrect change it as required +The code in question is currently not used by exec logic so while it is good to correct it +Can be implemented for Facelets by using +Anyone knows how to fix it +Are you also saying that the is redundant or poorly designed or would it suffice just to move it +Reassigned issues to Nick and had to set resolution of all of them to done +looks good to me +Needs to be updated +is another possible +Thanks for trying enis sergey +Thanks Namit! +And wizard is working +No changes at the moment to the BSP model but look at not saving all the messages to be sent until the end of the superstep +Works for me on centos +that it has special treatment for Errors +DOM Level Events is being worked on again now and DOM mutation events are being reworked AIUI +stat file for diff +Committed to trunk with revision +I meant that since these parameters are specified in the constructor theres no place where these parameters are reset when a compressor is pulled from the pool +I just added another fix to that should allow your second syntactical construction above to work properly +Im seeing a bug related to this patch going in +The proposed syntax is described in +Charles is correct the configuration will force a can also restart the agent and that will force a scan as well +Isnt it better to make OBR SCR a separate pluginsOBR is not into core spec and SCR is IMHO rarely used +Aaron I have patches that apply on top that fix HSFTP to work correctlyGreat so it sounds like the testing I did should cover all the pertinent casesI also ran all of the tests with this patch applied and the only one that failed was which Im confident is unrelated to this patch +This is first problem I was blamedI recommend you can use report tag in screen because image will placed inside s runtime directory +Manual testing code added to UGIs main method +I was wrong +I really agree with Yonik here +Will attach new patch after that +Thanks Kim +Ive just committed this. +If someone identifies a use case where its absolutely required it can always be resurrected then. +Sending Transmitting file dataCommitted revision Luckily velocity specified logj as provided so we wont have any version conflicts +BenIt was that was causing this problem not the container you use Rational Web Development Platform and its JSF implementation you will run into this it will be fixed in +Max did you forget the screenshot +What about the roles that dont fit this category though? Like the usersys admin +EL functions are a special syntax +For abstraction purposes wouldnt it be better to have a thread pool executor to decouple the zookeeper event notification from the execution of elections +Really sorry for the delay I was focused on GSOC implementationWhile looking for an option on Windows I found that Java doesnt provide a way to control the prompt +But doing so means customizing at the level where the API template is instantiated which means creating a custom service provider and connection this issue to an improvement story where the desired improvement is to make this customization of the request factory simpler without needing to customize the service provider and connection factory +So better documentation warnings and error messages might suffice +Not sure really what happensOK maybe something else I can try try using different test frameworks in subprojects +Tested at the WCF plugfest March +Yeah I didnt see before that already fixes that there +Only problem When not logged in to hudson one is not allowed to access the workspace directory +I guess we can check the audit log for the trash expunge time as an alternative solution. +Yes I wait until nexus indexing is finished before asking for local encounter the indexing problem both from Windows Preferences Maven Installation Reindex local repository Maven indexes view right click on local and update indexFrom time to time both ways work but its quite rare +ContextI tried to reproduce the issue which you described in this ticket +I assume when you say row level statistics you are referring to table statistics +Why is the fixed in quotes btw is there something specialfishy about the fix that went into +There is an issue with the packaging page +html files to +bulk defer open issues to defer to not essential to address for +resolving and reassigning +Here is from . +Len Id reassign it back to you but your name doesnt appear in the JIRA assignment list +Have you done any benchmarking to see ifhow much it improves performance +to commit if the regression tests pass +I also couldnt see how to tell svn that a particular encoding goes with a file +I also set to default in exampleMany thanks Sebastian! +has been resolved +Closing resolved issues. +Merged to. +Attaching final patch committing under Juliens Three changes added Missed a test class earlier which needed the static string moved to +Fixed with help from Hamlet and Jochen +Im not sure its within scope of Accumulo to do key management at all +nice +When you have more than one version of Mojarra you will see in the JSF startup message +I added and to cover the resolution of bridge methods +There is logical error in RCE +RE small details like that I work out at commit time +Not reproduceable anymore with iOS fix click on your projects root icon in your Project Navigator then the projects target +Sounds like a good starter bug for someone. +Lachlan can you reproduce it on a small project? It would help if you could attach it here +I looked at this in depth and think it is going to happen to work in many cases but not all +HEAD on branch resolves this error +Only thing to checkchange is noting official Flash player support and confirming what versions it has been tested on +This should be a unit test of the specific map usage pattern +Thanks guys for your commentsPaolo I will take a look at Jena RIOT for inferencing +Discussion has continued in resolving this as a dupe. +This is not the cause of the issue here but I wanted you to be aware of ll look at your case now +On top of that please provide reasonable operation descriptions. +The patch should not include binaries these should be built somehow by +Thanks for the contribution but this is already in for +many dbs dont even guarantee that subsequently generated ids will be successive numbers. +The class renames and class editing within a weaver can also be customized through the weavers methods. +Tested +These are thread dumps not stack tracesStack traces would in general be dumped in logs +has been marked as a duplicate of this bug +I tested using the version as defined by the tag +I added the following to usrlibrubygemsgemslibhelpers output lsof root output mount root output umount f sysrootdev root The lsoft output shows no mention of sysrootdev neither does the mount and the umount doesnt alleviate the problem either +Patch applied under revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Peter you already seem to have a JIRA for updating installation guide +I think to represent this returned result we need to have extend That makes sense +Here is an example of how I am testing url mappings +push out to +Are we expecting AS to be in TB +Closing old resolved issues +revision all old Resolved issues +Tomcat version included in is known +Ive created a new JIRA which describes the more generic solution +also required +I cannot reproduce this at least on +One solution for allowing the connectors to be loaded from the GAC +Is this still relevant +This requires more tests so move to +Passes all the tests +Initial version of forward port patch +I dont think changes to DNNN RPC are considered incompatible changes going from eg to since they dont affect users +I currently prefer the way +Implemented for file upload and download using NOOPMost ftp servers dont support concurrent control and data channel access in which case the replies need to be retrieved at the end of the transfer +So whatever script is launching EAP can set those environment variables before launching EAP +Its ok though +I think I will use execution option in most cases +But it does not include file system status like file size etc +Until the resourcepublishing split is put in place I would just consider the layer and its resource a single entity +Maybe someone can have a look at am NOT too sure wheter this is a problem of the site plugin or doxia +Rebased the old patch with the new changed in trunk +in which context? +Of course POI block IMO there is no need for adding further cross dependencies +to the illustrated package structure +Committed at subversion revision and ported to the branch at revision I have not been able to figure out how to trigger the problem case raised by Knut +Heres a pretty simple test that uses to try and modify column families +Id be happy if you can your tests so that we can resolve this issueThanks all this works now +This is in the JOPR code +shame on me! Thanks + +What needs to be done is just svn rm commonsrcgenorgapachehivecommon +committed patch with minor extension to include in r +Whats the effect of marking all sstables as compacted? Add truncate to the JMX interfaceAdd truncate to nodetoolAdd system tests to test nodetool JMX +The fact that the CF for schemakeyspaces is marked for delete worries me but perhaps that is an artifacts of the schema merge +The change you made for fixed the issue for oozie + function for converting a file path to UNC +Moving out of +What part of the java serialization spec does the code introduced in violate +Happy to figure out how to do this in SVN if thats what you need +There are some outstanding issues still linked to this task +Thank you for reporting +The other items under buildrequires alsoneed a compiler +Sorry I confused myself Looking more closely was about column names not column types so that got fixed in so one may never need to define a importer property +Happens about an hour after enabling it +Maybe you could look into submitting a patchThanks Lin Verified in build the problems in Chinese admin console described above are already fixed by Jeff +The fix it to restructure the tomcat testing moduls but a work around for the release is to just delete the modules in distributiontomcattesting from the staging Ill remove them for the beta to Beta +If there would be optional artifacts izpackexe and izpackapp with classifier ZIP it would be greatFrom there it is possible to extract the required artifacts to the working dir and do the script python is required for systems but it is written down in the docs nobody will complain about it +It seems that problem is in the locking engine +I just encountered this bugIt seems that this works just the way it should but you have to use a comma as the separator +This allows to replaceor extends the default implementation with custom actions like postprocessing of some of the objectsThe advantage of this patch is that the user of iBatis have much morecontrol of the creation of objects opposed to the post processor interfaceFor example he could replace the whole object through a proxy etcIn the patch the factory for is configured in the config through a static class +i will take a look and update the jira +We were used to the fact that the plugin name corresponds to a certain build so same string means same build +Patch which can be applied to examplesjms to reproduce the problem +looks like martijn committed to I committed the fix but couldnt close the issue due to unavailability of the JIRA server apache +fine by me +Can you open another JIRA if you think thats really necessary +Help in resolution is appreciated. +Lucian I suggest you can try to use a Server with higher version it looks like a problem of your can you try to use or +No need for IT +Based on the above understanding of the current work going on for is not going to fix +Assigning fix version +This is working well for me locally +Committed trunk revision Committed revision Lucene is not broken it only has the private Exception. +Spoke to Kishan Sailaja Customers cannot use their own system vm templates +Please check that it was applied as expected. +You fix the product instead of documenting how to understand itRequiring users to edit an XML file is a separate issue but I think it is a serious problem especially because any error messages show up in the server logs +Please +This patch fixes some specs more +Patch applied at +statically initializes the +This issue is fixed and released as part of release +Split hlogAdd suffix +This is cause of the implementation of +I will check if it is possible to use spring services and cxf services with the same transfer objects +This should allow us to figure out where and when that exception is see +Added support for default delimiter in UDF and more tests using embedded +Changing title for clarity +The patch I just attached makes a test in fail because it tries to remove a field through the iterator +Thanks Patrick +Also just tease out generated code and just post your changes on RB that will make it easier to review +I adjusted the Encoding class to use in revision as thats the logging framework were using sinceAs for switching to slfj or jul I agree with Mel that the benefits of doing so are not big enough +Ill commit the changes then +The work was already done it just needs to be exposed for tooling +Fixes the condition to be stopped OR interruptedUnsets the interrupted status if it is set before calling some comments about the interrupt status reset +Assigning to Tomohito since he was the one who wrote the fix and then resolving and closing this issue. +I know this isnt the case today but I just think its dumbNot weighting prohibited clauses would needlessly break certain types of queries +tested and works fine +A struts application supports any encoding thanks to but a json string submitted and received by has to be encoded +Looks like this work might overlap with some of the work thats going on in though that work is so far. +Patch looks good but it would be good to have a regression test if possible +This should be as easy as adding the following on the command lineCmydirThis will not resolve the chmod x issue as Im not sure File object in Java had a setExecutable untilThat said if you wish to build and skip unit tests you can pass true and that will stop maven from running the unit tests +with the latest revision this issue is now fixed +Hm Im not sure Im getting what you meanIve added some more tests to to document the current behavior +Patches were applied at the revision +Attempting to start two instances of the AS server using Service Binding Manager will result in an error +I could not reproduce any of above mentioned errors after building components and showcase snapshotsSo as you guess the upgrade of Atmosphere fixed it +without any changes on their end +The spurious warning issue It is indeed fixed + +The code looks great +Would it be possible to have a unit test +If you fix experts could give it a whirl now and see if theres anything to be added before the release please shout +Its working +Main Points +Will definitely look into this +Let me know if anything is not clear Mike +Done +The machine has core cpuI then tried an sane trunk build with the same jvm same machine unmodified and it failed out of timesThen I modified line as you said with a timeout of seconds and there were failures out of runs +renaming the issue to reflect the actual problem +I just deleted the patch to and am nowIt is the same file it is just that I forgot to gran Apache the license for inclusion +Just checked we need the handler responsible for the HTTP to HTTPS redirect to be moved before the handler responsible for HTTP upgrade that way the redirect will always a later point if we support STARTTLS on this connection then we would still want the redirect early as it would check if the connection had been switched +Removing duplicate documentation on sortLast and adding pintpfloatplongpdouble to the list of types which support this feature +Patch for Integer Decimal Long Int Short Byte Float is normally in Justins area though this is not the module Justin maintains actively +TedBarryAt least what I inferred initially was that this is an XML document model scenario see the attached vdb +We havent tried doing a new install and havent looked into what happens to during a new install +When looking at beta I can see the same problem as before When I invoke Red Hat Developer Studio Add JSF Capabilities +The deploymentmachinery can process it write the actual topology file and delete thetemp oneI do believe that there will be a need for additional tooling at some pointI could see this tooling manifest as a few different options REST servicecmdline server startup mode etcI guess the question becomes whether we want to create a new Jira forrewriting clear text passwords in config files and defer this generaltooling jira until later when we know more about the HA requirements andkeystore locations +Lets get a first cut into Camel and then we can expand on this EIP later +Lets go w that +But yes if we go for KISS clicking on it seems wiserWaiting for opinions +committed thanks +This is still valid we still have some unit testing dependent on canned derby databases +This specific scenario is hard to verify in a test I think +If it doesnt Ill revert the change but if it does then the fuse team can decide if they want to commit that minor change to apache +Turns out you need this set anyway for mavenIf not availabe the batch file will set to use the windows search path to find +The metadata is versioned it just doesnt have a global version +I cant reproduce this +The attached test case simulates its logic in the simplest manner +our patch for wtp wont solve this +Changes javasrcmainjavaorgapacheshindigsocialopensocialmodel and javascriptsampledata to make consistent all uses of displayValue +It passed when I set JAVAHOME +Moving this JIRA to TBD issue has been triaged for EAP by the EAP PM team and it is downgraded to a Major. +Add some Code annotation and test patch the failed test is not cause by my patch +rename patch name to +Fixed in! Cant wait for +See comments above. +moving fix version to out of +Plus invoking on the client side ssh user pommes works as expected +Please disregard my previous comment it seems a later release has been performed. +Please do not submit this until is submitted +Patches for ys branch +Unquoting both seems to fix the problem +I cant think of anything else we need to do +I would go further than this and suggest a facet or attribute that you can render the doctype yourself for moreI think thats a great idea! I really like the clean approachThe question is how to go about implementing this + +Please review it +Thank you Steve +Quickstart reproducing the fix +Targeting revision +Removed the word full from the should remove all references to the BEA JDK should replace all references to Java with references to Java as Sun will EOL in October +I applied the last patch pasted by Ben up in the HADOOP issue +done +In addition foreach does achieve the intentionA foreach A generate a as achararray achararray does not actually do the type conversion +Angel not taking anything from the issue youve raised Im wondering if context needs to be a session attribute if it is present in the URL +Of course tests run slowly with +The Region Digraph bundle the needs to be added to the local repository +As per Martins suggestion reformatting of the code was not needed +Committed Susan Cliness patch revision that modifies integrate to clarify which Eclipse releases can be used with the UI +I think another executor pool is helpful especially it can be used to assign those failed regions in parallel tooHowever for now probably I will just create another zookeeper watcher instead +Updating with the corrected version which fails without the fix and passes with it just for the record +Okay unless theres a third version of that file its fixed now +The next module will be security +The python code which is also using the HTTP protocol works just fineAny help or ideas would be appreciated +this issue related to Tomcat installation and not to plugin +Im checking into whether or not this is intentional but in the mean time I think the patches can stand as they areIll leave this for review for a day and plan to commit on Thursday +That is why I was not sure which one to keep +This is fixed already in the latest code base. +Is it in this case a good idea to close this as WONTFIX and open a separate one regarding this findbugs warnings +An earlier update process had apparently died and left a stale lock behindSo the update process is jut bin or it some other scriptprogram that eventually calls binIf its the first we can write the pid of the process running bin into the lock file and let everyone else check for both the existence of the lock file as well as the status of the process doing the updateIf its the later then the parent program can take care of cleaning up of the lock file in case of failuresThoughts? +It would be nice if someone could review the refactorings +Ive just used this for the first time in ages and have immediately hit the same problem +Not for commit here. +Which isnt a big deal +I would rather push this sort of thing out into the DIH Solr Cell for example +Took all of Konstantins advices +Is there a patch here for your latest? I see the diff but not the patch +This requires SSL tunneling by layering sockets and Im not sure how accommodating Mina will be in this regard +I dont understand how a readonly list equates to software licensing considering that many apache mailing lists are moderated and there are private lists for each community? +well since the uuid is created immediately and the node is referenceable as welli would expect that i could also get it by uuidsecond we have an issue with a backwards compatible user mgt implementationa NEW user used to be accessible by for the editing session even before commitingit +Why does this affect the other quickstarts +Just retested and this issue is resolved. +Generated new patch for trunk +If it starts to catch Ill be happy to offer help with use of the automaton API etc +We could initialize the list from a config file in the case of persistently bad hosts +My fault +Yes that worksCan you check the value of this preferences setting? If so please display a dialog with instructions how to fix it when the setting is wrong +If you need to do this make sure you turn shadow proxies off for +MahadevThe four test classes excluded from the new patch because of their dependency on classes in the newly created examples mvn module are +Please close or retarget to a different fix target as appropriate +As only tests were added Id say that camel works as designed in the already released and right +True +httplocalhost attached is not a depends on test case using workaroundRunning Tests run Failures Errors Skipped Time elapsed sec +create keyspaceText is used when creating the wiki help as the sort order and heading +src implementing dont understand what problem you are trying to solve +The message has been updated and now contains the date of the snapshot +done +I didnt see anything else useful but heres the whole thing +Whats an exception look like if you pass an unknown command +So the issue was in my configuration not in jBPM +need to add a section about adding DERBYHOMEbin to the path so that the scripts can be executed from any directory + +Also for togglePanel accordian inplaceInput componentControl +However Bill Burke has indicated he regards it as less of a priority than stability and manageability issues +The patch drops the association between +change fix version +I am using the solutions above for now +This isnt a real issue its more of a usablility issue that can wait untill is out +And committed to and +correct usagegroup so on +At this point given there is no requirement we can keep the interface and not to honor bypass for createTable etc +Grr +We shall find a way to address them + +The patch will be slightly different due to the differences in the use of and encryption support +When I submitted the bug I was using Eclipse +Now changed the return type to boolean as per you satisfactionActually I thought that whenever we execute some commands it gives some error code like for not succeed or for succeed way I have implemented the method but I have provided for not succeed instead of +Attaching project that demonstrates described behavior +Jsp file referenced in initial defect submission +We should be able to choose which configuration and which type +Yes and no +The other dependencies can be added back as theyre needed +Also if open already in master could it be open on a regionserver? So this assign will mean double assignment +The test case checks it against the expected number +I believe that the changes Brian and I made to the Parser fixed this issue +Hi DougI compared your patch to the existing Java code and I see divergence from the Java code with the extra code you added +I can verify setting false in studioconfiguration makes the errors go is what has changed that require us to add that +Btw this will also fix +At least that sounded like the only reasonable explanation based on your description +It also adds two new list transformations and +Will create a separate issue for them +Yes the Tomcats web console which is available at httplocalhostmanager Sergey the latest fixes are in the code base +This issue has been fixed in the PHP implementation feel free to close the issue ifonce its done in the Java in both php and java +Alexey please verify +This should not block release. +The proposed fix doesnt change that existed behavior. +Everything else is suppose to use it if neededSonar handling is done at so we dont need to do anything in the Hudson but rather hdfs would have to be added to that server much like Avro etc +if you put a getcount in between the the remove and subsequent slice it passes +may not be a blocker for +We have to ask them to release a new version ASAP. +Yes +The old edit of has been removed and replaced w Todds new way of doing it +This patch is in Work In progress status +I volunteer testing and review +duplicate of +looks good just added an extra check for index out of version +Not sure why but this seems to be working now +Harish Also can you get rid of config variables in which were about size of persistence byte list those will become relevant after this patch +here is the new version of that patchthe old test doesnt catch it because it doesnt add enough data to cause a flush which is where the commitlogheader corruption happened +typos i mean closure cant run properlyIf there are s jar file in tomcats lib folder and its loaded before Guava is loaded Closure cant not run Collections project has become Guava +Thats making finding the real changes hard +Thanks! +Adds fsm +Lets just get this one in +I dont know enough about the aether library to know what its trying to do or why it was integrated in +See as an example +This version defines a Listener class and a class that handles listening across multiple sessions +Let me rework the patch tonight and see what I can come up with +It was released with the fix in +so added to svnignore for moduleslunisrcmainnativeluni and modulesarchivesrcmainnative OliverPatch applied at repo revision rPlease check that the patch was applied as you expected +OK I see +In short one might have the right to grant privilege to others but not the right to give other users the right to make further grants if WITH GRANT OPTIONS is not specified +Works for me + +All tests pass now +Thanks Dag! to commit +According to Rob the basic implementation is done +flavio Currently if a server is on leader election and doesnt receive a notification for some amount of time t then it sends a new set of notifications if at least one server has delivered a message from the previous set + classes query implementation unit to knowThanks I miss something but what is the usecase of this query? I guess you have a good reason for adding this patch could you please add some documentation to the class and or to this issue which explains what you are trying to do? JIRAS lucidimagination link broken and it seems like the issue can be worked around. +Anyone working on this +I think youre seeing it wrong uses the entity id +sounds like this is fixed now +A bunch of work has been done in AM since +Unfortunately its difficult for some of us to test these container specific issues. +So stop talking and implement it +Current cordova info output appears pretty disjointed with a lot of extra vertical whitespace +Owen the sizes of the spills may not necessarily the same for all spills +Drat +Moving to M +My local works fine as does but our stdlib and my copy of Ruby both ll see if theres a good way to add a testspec for this too +Another thing is that in EAP you can set the password in the two places +These same issues existed in the error messages document I reviewed +Im guessing that this will be a problem for anyone using a javascript client +updated to include release note. +Trinidad Version branches are no longer maintained and this patch reverts a previous fix which has already been released +Why schedule a build if the component is inactive? And why days if its that slow? Why not weeklyAs to API changes that break downstream it happens +Marking as blocker as we need to freezeAt this point i dont think anyone is going to change their opinions already stated out on the mailing lists +It does not contain any class files but just the merged packages without a header +You could run all tests with verboseclass and then grep for locations which were actually used +Merci. +Removing from the critical path and leaving this as available per Nicks Oct comment. +After deploying a BPEL project these tests check the status and fail +This can be overridden if desired by setting the publish property. +I did go back and look at the original case that made me worried in that case the bad document is K and the matched term occurs times in it +Jira is not the place to ask questions +Ill remove it from the next patch +Eventually we should figure out how to configure Forrest to build things in the build directory but until then this is a good patch +how about suggestions for the tag though +is only with right? Because I tried in trunk and got no errors + +fix test issues remove stray think the batch API is essential to get good performance when adding or removing a lot of entries at onceRe error returns I will create a Java object protobuf for each request so that we can add more fields later +added default implementation. + for the to trunk. +Uploading correct patch +couchhttpdauthcreateuserreq and couchserverhashpasswordadmin use a different algo to create the hashed password in current +I just checked its in branches as well as trunk +A quick workaround to let it work by default is attached + +I fixed the bug which covers problem with the single line comment and empty parameter in that was a fast attached code runs but formatting fails also +patch v +This patch was incompletely applied +Almost done Ill commit them in subversion repository as of r for branch and as of r for issue as it was included in version which has been released +Incidentally if this still is an issye is your Spring application context being closed? Do you see any logging when you undeploy the WAR? Which mechanism are you using in to load spring +I have increased this to k +Committed revision +I committed the workaround to in rev After discussing with gnodet on IRC his suggestion is to disable the weaving by default but allow the users to opt in for classes where they need itPatch supplied. +Put in the restriction for trg and src to be in the same directory +stat file was ll do a lang formatting review for the +When using a theproblem doesnt seem to occur +DONEOK +Ditto for do you have a patch file of the double quote changes? If so we can put it in easily for +Committed for support for attribute types declared in Simple and complex types +No +Now I understand the issue +A few of the tests in this test class get incorrect number of messages +Submitted patch to add MIME types to with default MIME type being the existing application +IMHO we should close the issue now +See discussion in the pull request +are still very useful as internal mechanism +This are the charts from a very introductory presentation I have held at the IBM Innovation Centre to give an update about progress of my work +we need to add some more details to thisThere has been a change in the add cluster API we cannot specify the overcommit values at the time of adding the cluster +any news +We can probably clean this up but I think this can probably wait until after + +First versionThe static maps per region in compaction request might as well be stores in region since the code to populate them is called from storeThen move all the compaction stuff into separate class from everywhere +We do not need to worry about atomicity when rolling images +To avoid this case one solution could be to check if the job being queried is inited +The merge looks simple but Id prefer if you did it as you know these changes better +Was this really against or if not could you try it against and report back +Note Theres a whole bunch of utility methods below you need not use them to use the general idea ignoredelete first version of you could solve for record types by creating a reserved name such as hint or some such thing with the only caveat being you would no longer be able to create user fields with that name in python +Upgraded to parent pom which allowed removing most temporary additions. +Verified by Spark. +Looks good to me. +this patch fixes the in +Now that work is well underway with AS all previous community releases are +updated description to reflect decisions match was committed along with +Our closures cannot yet handle null as sure about your solution suggestions yet but before going into that doesnt that make it a JIRA item for grails instead? I cant see anything wrong in this case that is happening from Groovy per se +Marking fixedclosed per Dhrubas suggestion +Patch is in CVS +I have added some error checking to ensure that the principal components are installed and have uploaded a new snapshot +I am also creating the with xdoclet and running them on weblogic but no problem occurs there though the same jdbc version is being In YES useridHiIt worked for me with latest cvsYou had six I only configured thefirst so it failed further down the jspRegardsAdrian +Committed toThanks Jeffrey. +Thank you +Thanks Suraj! +Script to fix various missing SVN propertiesCommitters should check that they have set up the correct SVN defaults +Any chance you try +Fixed in Rev +Here is the change I made in +I know about it but theres not much I can do at this point +The implementation has been consolidated to the and should be a more generic solution +From the original description I understand that a read request hits a bookie that has never seen the ledger of the request before consequently the result of the read is null +my patch for this am not sure if is really supported any more +Closing. +Hi Joel the staging upload went through at first attempt so thats encouraging +So even if you had a file extension it wouldnt tell the whole story +Thank you Rupert this issue seem to be resolvedIm building right now fingers crossedCould someone mark this issue as resolved? Im not too familiar with Jira dont know ill do it right +committed to both trunk and branch +most likely Dmitri never added these methods onto because its an interface +The doDrop and doTruncate properties were set up so that the application would perform as before if the values where not changed +Found this post googling trying to understand why Im not getting authenticated +Added support for other data please note string value returned by custom converter should exactly match the pattern set on calendar component. +Ready for commit +Sound it interesting to you +The code was refactored to combine two visitors into one +After the validation above I got the attached image +Added writeField method also includes the ending CRLF while the previous didnt include it +I dont think has ever been tracking sub deployments descriptors +I believe it will be fixed as part of +Tools are meant to be reusable while each webapp coder writes its toolbox file +Sorry for the mistake +My suggestion would be to collocate a file name alongside the jar that explains it +applied +somewhere +Tested and works well +I think andor is the culprit +Its going to have to be a JVM option for now until we can figure out a more elegant way to handle it +There is still one call left to a logj log method in +I was just making a suggestion to change the pom for the assembly plugin so that I dont get all the debug output when building +Is this easily reproducible +Renamed class to. +Internal pool doesnt exist in fix is only applicable to +Ill add some in the next patch. +It seems like Im not the only one facing the problemJesse could you please reorder the repositories? If the opencomponentry repository is placed before Howards repository the problem doesnt exists anymore +I cannot reproduce the seems that we have fixed the issue amoung the other fixes +it is expected discussed with Trong +With these I have the working +Hmm actually this may result in stale nodes when one of the creates fails +sources jars are only available for release not for is your case about having sources for SNAPSHOTS? +I will follow the these guidelines for future work. +order not deterministic +as what type is definedShould be one of these to be quotedcase TEXTcase CHARcase CLOB +Hi Sanjay +on introducing in +Its working well with that fix +What would change is the way the resolve of this container is specified +umask on cmd prompt yields umask in code above is newPermission in +When do you do that +Verified in ER +Indeed measures at fileclass level are purged each time time a new quality snapshot is created so we dont have any way to compute variation + think the stack trace is useful for debug +Found the issue with Durable seems to work now with my Broker doesnt seem to work against trunk broker though +Simple patch +I posted a review on reviewboard +very niceAre Accumulator and Algebraic functions supported +Jesper +Hi BJ Ive fixed this problem +So there is a compatibility issueIs ACF behaviour wrong or should the settings be writeable with gettersetter in railo +would be nice to have it on review board +Please check it! Thanks! release note looks good to meIll start the work on backporting the patch to older branches +is partly used to test the embed document some of them have been adapted according to this bug fixAnother issue that the commented lines can be removed because it is duplicate with the loop +OK what do you mean use? In qa? On our developer boxes? for CI +Can someone this to use the cli instead of python +I have no idea either especially after I found out that the test works on my home pc which has exactly the same environment with regard to mvn version as the one in office +roger thanks for finding out about the missing patch please find the missing patch for attached +Jozef would you be able to write a quick test for this on Monday? If you cant find an assertion mark it implied +Didnt make it into mentioned that Anil contributed a fix for this +Subin Looks like the directory useroozie does not exist +Alternatively you could set it using a a namespace change is required the change would most likely go into +Thanks +Sylvain this looks good for now +i still have to port this to but it is fixed for there a workaround to get this fixed for? +Just wanted to make sure you were aware of the issue +If you have any insights please share +Ok I obviously this change in +This should also make the recipes more portable +If I change the clauses to just clauses the problem goes away +Just needed to update the JSF attribute to the new value which RHQ is using. +everything for which there is a support moduleIf we go w Servlet we wont be able to eliminate our Cookie implementation IIUC since still doesnt support the httpOnly attribute +It is really a trivial fix but it needs to be done by SUNAdded javadoc and removed unnecessary class +Dear PalanivelrajanThanks for your replyThe following files should be in the folder coresrcmainjavaorgspringframeworksecurityauthentication following files should be in the folder coresrctestjavaorgspringframeworksecurityauthentication following files should be in the coresrctestresourcesorgspringframeworksecurityauthentication Waiting for your commentsBest regards PalanivelrajanDo you have any newsWhen do you think you will be able to look in itI just know that generally after comments I need to do additional monthDuring monthsWaiting for your commentsBest regards Michael +I would suggest that this is a little more important than minor +This would be on par with other implementations of JPA that are coming out such as +Hello Pavol Oracle JDK There are about concurrent usersThanksNeither of those parameters are set in my do see that I get spikes in the threadcount but that does not seem to block the system +Fixed in revision +This was a side effect on my work on SPNEGO OIDsThe problem has been fixed changing the internal type of form OID toThe test has been addedAnother test has been fixed accordinglyThanks van +There is a similar issue where there is a cluster with only a datasource and a queue +unless someone objects +Code has been added on committed revision Sun Oct EDT +Verified in CP CR +The main problem with trying to enforce prefix consistency is that does not add prefixes to executable scripts it installs +for that +Michael Thanks for applying the patch Best regards Wolfgang +Patch file that fixes this issue by using selfadjustFields for filtering fields +applied +I tested timestamp and date it look good to me +the dependency on Progress it is only available to their customers which will need instructions for wrapping and deploying it in SMXI will work on the unit tests and Apache License issues right away. +That seems like a good solution +Otis this really isnt a bug +Attaching a patch for exclusion of from group Colin and Andrew +There is no need to assign issues to somebody in general but it would be great if you are interesting in working on it and provide a patchThomas +The issue was with iN message and its use not fully implemented for the local names +It just indroduces new instruction which can be used in other optimizations to allocate space on the stack +Its basically an internal refactoring except of the additions to in rev . +Can you provide a bit more detail here? A screen shot maybe? Im not sure what youre looking Smooks Run Configurations are added to the Run toolbar button list if successfully created +I will include this patch withthanks flavio +Looks good +Evgeny I did not find any simple solution to keep or properties and use a skip to control them alsoSo I came back to my first solution thats simple and worksMaybe profiles may be a solution but I make the quite more complex for meSo I let you either look at it or resolved it +svn commit m Updated Sending Transmitting file dataCommitted revision . +Close all resolved issues for older releases. +Thank you Mikhail +FWIW the aggregation framework was move to due date is now in Oct +Hi GlennSorry! I have asked the question on the list nowThanks taken to list. +Patch to fix the issueThanks you please give the steps to reproduce the issue +I check +Sigh dont think thats how it works +Had a look at a too good +If were going to implement something for mapr we should just call it Mapr +preInsert is executed and it works when the property is defined in issue was introduced with version it was working until version we will now check all the changes in this version +If we replace Deprecated with some other annotation then our users will lose the automatic warnings +the update appeared on the site a while ago assume weve covered the issue. +for patch +I am closing for now +Tested +The merge functionality is much improved +there is no clear path forwards despite much discussion Im closing this until somebody comes up with a Fix For All Browsers And Services +Could you elaborate +d by Stu in IRC +implemented as mentioned in the comment +this is definitely WTP and actually I believe its by design +Ill resolve this issue nowThanks Isaac +Whilst some of the instnaces may be necessary there seems to be far too much additional being added within the general implementation methods in etc which are used to support the JMS methods +committed. +It requires no changes to other commons net code and no one is forcing anyone else to use it +I guess we can add a special case for any path that ends in crawldb but this is not a complete fix either as someone else may rename his crawl database something else +I think I just fixed this see commits from a short while ago and my reply to the fail of replication handler test on the dev listMore details to follow +Currently not we need to update Clerezza +Yun we are not actively patching right nowI will review the changes for trunk and in and trunk +Ping +last patch was stale and didnt have the randomized string fields for testing +Dont care if you want to change it. +Can the operator not do something like add a check hasProperty +Guvnor module does not depend on directly anymore +the plugin could be clever and cache it for offline exactly is the difference between this and +Im taking this Ill add this capability to the ant tasks +Not critical so for +Patch created at top are a lot of merge conflicts here +So if I am working on feature module I can use the Spring MVC Test to just test my controller services repositories all the way down to the database simply from the the IDE +Bulk close after release of +will be probably moved out of is not a funny thin it is quite a funny world Well need to shift post +It is onlyattached for reference not for inclusion +Could you also include the sample gadget that try to use this template to help debug the issue +Suresh thanks for reviewing the patchI have committed this +WDYT +If the databaseis booted by two class loaders database corruption can occur + +In this case tombstones are compacted before the data is ever compactedCase data is in sstable tombstone is in sstable +The first for the outer read of the large data set and the second for the writes +runs as a separate service started from the operating system +That one in is not hit very often but fine lets change it too +Closing old resolved issues +Id ignore stomp issues for the moment unless you are planning on using the stomp protocol which I dont recommend as the openwire protocol performs better and offers a more complete set of features +Closing per Garys comments. + +And we are using in our clusters by configuring higher value than default of m +heres a test case also +The first three fixes should be given a higher priority +Amarsri and I discussed the test case in the original patch and decided to retain it as a regression test for the bug that triggered this JIRA in addition to more targeted unit tests that I wroteThe new patch has the following changes in line with the above decision Makes editorial changes to the regression test adding a few more comments refactoring for more explanatory names etc +This patch stops the BIRT component from blocking context sensitive help for the Accounting Facility and Order components +I can do this but according to Mark only with a new issue and patch +Thanks Micah +We are in the process of removing the unique constraints from the server and service tables +The array wrapper beans must be included in the deployment +This is the final list of the current fixmes related to release +Repo is writable now enjoy +done +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Attached the third +Assigning to Fay her patch is a better solution than mine were +Now that we have tail append headers we could do this pretty easily at the batch level IIUC +the problem i have with special handling for rollover to make the epoch change faster is that the code path would almost never get hit and the path is so it would never be hardened +patch fixing download url for I missed that fixed at rPlease check that everything works as you expected +shouldnt the fix version be the next major release for a change this largeAgreedIve been trying to figure out why this ticket makes me uncomfortable and I think Ive figured it out assuming each client has identical request distribution which is the case for almost all applications pulling the cache check out to a separate stage wont actually help clients progress any more when you are io bound since once they issue a they have to wait for the physical read stage anywaySo while theres a certain ideological purity of having cache check on a separate stage in reality it doesnt actually help you while still adding a fair amount of complexity to the read path +Can this be closed or better yet updated so we know the current status? CheersTony +Just committed to the +Logging happens in +The machine that it fails on is a Microsoft Vista Business with maven +Bug in my http client. +It seems that every call to the reallyExec method have a as the value passed for maxRows +Test and pass locally +Fixed +I think rest server has to get the data at first then to find its sizeYes +Manually tested +Another Emacs related issue that you may want to consider +A straightforward change +I guess we could to avoid bumps like this one in the future +I just committed to trunk +It seems you have changed tabs to spaces which while I understand it correct based on geotools formatting conventions makes it nearly impossible for me to tell what has changed according to the diff +In effect I would like it to be the case that if the dependent resource is available it will be started first but if it is not available no error will resultIm afraid that if we generate no dependency theres no way to achieve the start first if present behavior and the optional resource may start after the application that uses it even when both are on the list to be started +This problem appear when i install and run the exampleI found the which contain orgapachehadoopyarnipc is under HADOOPHDFSHOMEsharehadoopmapreduce diror shoud add HADOOPHDFSHOMEsharehadoopmapreduce in? +Hi DanielThis issue is missing the fix versionThanks +He cant use a snapshot as the complains about thatBrian a workaround is to do an internal release if you cant wait +Perhaps one idea is to use a visitor pattern to decouple tree traversal with the operations being performed +Im zipmerging now and Ill go through whatever missing dependencies come up +It is however up to date now and will stay that wayThank you +I think we should be careful to document that moving forward instead of changing existing commands new APIs should create a new COMMANDXXX command instead of bumping the protocol version +Its actually worse than that everytime you refresh you occupy one more handler and eventually they all get full and log splitting stops +Latest patch requires +Brian we can manifest generation to the another phase +The problem is that we will have breakage that wont be detected until HDFS is updated +so there is really no harm +The invocation occurred using a ruby object implementing a java interface that was passed into java +This is related to new that is copied to libendorsed jaxws jaxrpc and saaj proprietary maven and from maven repo. +Perhaps just add this test case to? Since we currently only support a single shared edits dir the test could be made more realistic by not configuring several +just saw that this is in and not jackrabbit +Happy for someone else to take it on or if its not relevant anymore to close it out + +I will upload final patch with test cases in some timeMost of the code is borrowed from will be covered in. +The fix for should solve this problem +My warble reverted back to an old goldspike. +Change looks good +A couple of things +to the Sun proprietary license not being useable +Its better than before btw but still not perfect +was irrelevant when were just packing unlimited in memory +these new edits are not going into a WAL at all? Yeah thats right we talk to the memstore directly rather than go through which does the +submitting +Done as part of commit +Checked in queryAll for the connector portion +I guess the optmizer model might not be wrong if it is actually the case that if there were rows in the outer table thatit would actually not do the full scan +is needed? dependency of +Which browser currently uses such whitespace compressor? Sound like more acompressor problem which should then add in the necessary +So we need a patch that covers all those +I think it is better to switch to extjsWe can keep extjs for as it is closer to release +This has been committed +I havent dug into why that is yet though +I dont think is the place to add this from a designpoint of view +as well +periodic bulk close of resolved issues. +Patch to fix bug +It would be great truly if we could have this sort of magic just work for all cases +This was actually a bug fixed in +the dependencies and info have been in the release notes. +Youd also might want a class to hold name data type etc +thks +So all issues not directly assigned to an AS release are being closed +patch looks good + +errors are logged at info and then on the second attempt at ERROR as it means your network is really hosed +Fixed in r +Just to let the test suite run without problems on any of the by me +i share jukkas concerns to automatically remapping the new namespace for sessions that have already mapped some other namespace to the same prefix +New patch with comments addressed +Thanks Emmanuel +This patch limits number of concurrently running applications +I have no problem with them being public +Many thanks Joe for contributing itTo my best knowledge the only standards that ASF enforces is licensing +Working on to test it out in cluster to see the metric in Ganglia +The latter considers these errors critical and does not let the application run +Closing old resolved issues +Indeed thats what Im planning to doJust made a comment here to keep track of my investigation and to keep people talked to David online and he explained that its not version thats in been added to each individual object but that each object is written separated in blocks which is what is adding extra information +Merged +See commits cafcfaddafbcbde afcdefcdbdaefaf +Now I cant duplicate it +Yes the plan is definitly to make these tests like the other new tests and to follow the pattern from the upgrade tests with respect to how old release jars are foundIt appears I will have time to work on this nowFor the first version of +Craig Any update on testing with linux? Based on that I can make the patch availablethanks reason for the failure is the missing on the build serverIm going to resubmitting the doesnt do anything to the libhdfs targets unless the two flags are set +We could have it to move the label closing all issues that have been in resolved state for months or more without any feedback or update +This bug should be closed. +Ill look into it +Dont know if this is a problem in +The Java file remains unchanged +can you turn the printlntoStrings stuff into tests using assertColumns +Heres a patch that fixes the tests that were moved over +DB was used as database +Oh its a real line break causing confusionI agree with you that it isnt terrific it makes it look like many more changes occurred than actually did making it less easy to verify changesWe have a somewhat similar issue with the derby web site +Previous comment answered the question and the new API was added by commit so resolving FIXED. +Patch applied at . +This one is vital to since the key and length are no longer located where unversioned readers expect them to be in particular they will be compressed with nearby buddies +The database file is closed when all connections are closed +has been marked as a duplicate of this bug +CLHM is implemented as a decorator to CHM to provide the bounded capacity semantics +new patch for removed previous addBlock methodchanged versionchanged log level in to debug for the node exclusion operationrefactored to junit +In there is no way to disable new logical plan with or without this ticket +See +Some specific access to the session must not go through the close flag Emmanuel! Im checking out now but probably wont to do any serious testing until Tuesday +Foreign key addition statements generated by this dialect do not add an index at the same time thus preventing duplicate index creation. +Done +This particular issue no longer seems to be a problem +You definitely have to throw an exception if someone tries to make a type its own parentI dont see how a method called is will help +Using would allows the user to ask for more eventually +Why is the rationale for a reference to a file? + +Thus README INSTALL and sample files +I search for all instances of in the trunk and all these appear to be correcty unwrapping the exception +But you are right this should be taken care of while processing the report on the +closing as fixed. +OK I think I understand why it does this but it seems a little odd to me +Fixed in +If one of your elements in the map has the key length and you try named access you are out of luck but this is an acceptable limitation from my point of view +Attachment has been added with description FOP generates maps for CID embedded TTF is a patch that makes FOP generate maps for CID embeddedTTF fonts +There is talk afoot of radically input validation +Thanks JacobI think the core links on replication page were meant to go to the admin page of the core +Can you provide the exact steps +Can you provide a test case for this +I will take a close look this evening but it looks like the algorithm you are using should be exact +Here is how the does it +bq +but also couldnt find the actual test case in the header +This is already an issue if you have one pool with the executable process a scenario we face in camunda fox for sureAssigned to us. +SvanteThanks for taking the time for reviewing and feedback +Doctored copy of input is not +Thanks in rev Malith for your contribution +NET naming conventions and changing this may produce confusion +Not sure about the others +That solution was practically a complete of the C library however +Koen weve taken care of the ESB side of this +Therefore whatever you put in as its value should evaluate to an outcome +Does this include separate local repositories for each project group +I addressed issues +The title is a cryptic means that I want strftime to support T +Do you have a link for the spec? +OK +Looks like a bug in the trunk too +I could update the patch to add to the top of the always list rather than the bottom +Radio now extends from +I am wondering whether the partial fix is sufficient to close this issue or you have a use case not covered by the partial fix +Thanks Priyo +If we want group by semantics this will need to be done first +moved from drools to jdocbook I remember correctly Sean Rogers also had this problem with docs +I will reassign this issue for verification when the docs are available for review on the +Start EAP server +I was going to handle it as a part of this one +This patch incorporates Dougs comments and leaves and alone is restricted to instrumentation within returns path state back to the Caller by returning a subclass of with associated path state includes more comprehensive testing code +Ive committed this +The resultset should not get closed as part of the commit +Yeah I saw application error traces coming in as you tried to edit that page + +Sorry it was that I was thinking of for the Hudson build +David Im going to tackle this one by simply moving back to a contrib module +Dan ulimit u is the same as the soft nproc value Ive suggested above Scott tried as well with M of as thats where the stack is allocatedIt still doesnt build for him +I have updated the issue with my analysis +Fundamentally we need to be able to answer the question of whether the lease recovery request is handled independently from the other +Implementation of +This is great to have! Can you please convert it to Forrests XML markup and construct a patch that places it under srcdoc? Thanks +Moved the two subtopics of Changing the properties programmatically which were short and deeply nested into their parent topicThe following files changeM srcadminguideM srcgetstartM srcrefM srcrefM srcrefM srcrefM srcrefM srcrefM srcdevguideA srcdevguideA srcdevguideM srcdevguideA srcdevguideM srcdevguideA srcdevguideA srcdevguideM srcdevguideM srcdevguideM srcdevguideM srcdevguideA srcdevguideA srcdevguideA srcdevguideM srcdevguideA srcdevguideA srcdevguideA srcdevguideA srcdevguideA srcdevguideA srcdevguideA srcdevguideA srcdevguideD srctuningD srctuningD srctuningM srctuningD srctuningD srctuningD srctuningD srctuningD srctuningD srctuningM srctuningD srctuningD srctuningD srctuningD srctuningM srctuningD srctuningD srctuningD srctuningD srctuningD srctuning +Problem fixed in revision +Youre right Joris we could fix this issue on Sonar side +Can we please take this general discussion to the mailing listIf is easier to respond to the various points made there and the scope of this important discussion is well beyond this issue +So the main workflow is completed Selection Bar was moved out from the VPE plugin +extends +Now the is catched and a is pushed into theBR great to see this being fixed so quickly! +Hi paulex the patch works fine + +Fady this issue came to light in a quickstart which Burr Sutter cause of the error is an attempt by the ESB integration code to deserialise and access jBPM variables without creating a jBPM are currently looking at the ESBjPBM integration and this issue is one to be addressed +true true + Roman +Hi Aidan does this change work ok for you too +Assertion failed error will surface and the vdb is unusable and the only wait to continue is to restart which is not acceptable +This could actually be anenhancement to branch which you are reporting the bug againstWould this approach help you +I apologize that this isnt the most direct example but the stack trace looks the same +Verified in EAP ER. +Thanks +Seems to be there are some bugs regarding the service operation implementationstested under branch +I ran jcarder with the patch applied and still detected some deadlocks attached will detect if a lock order is inverted regardless of which threads are running ie it cant tell if lock order AB is used before a threads started and BA is used only after its started +Checked found that when ringmaster section syslog address then only ringmaster log is going to that log +This bumps the last digit after generating the second release candidate. +If you have the same results this means the console has a bug +Can you try the patch attached to this JIRA and see if it applies cleanly for you +Heres a possible solution which uses factory methods instead of public ctors +Ive applied this the fix looks had to modify the Selenium test a bit to work however. +Guillaume I applied your patch +Also built the document using forrest and checked the documentation changes covers newly added image viewer functionality +Please verify +JIRA Cleanup Closing Old Bugs +I found that I had the in c + +field reference jpa class +The site sources look a bit funny as a legacy from Confluence +CoolThe dependency on can it not be scoped for test as it appears as only used for testing +This revision of the patch cleans dead wood comments from the patch as Owen suggested and adjusted error messages to indicate a property can be set at Ant command line or in build properties file +I just did a rerun but this time have DEBUG on in DN logs so if you need anything extra just say +No they dont take this and for damn good reason SRSSWS contain fieldinfos +Like I said this really shouldnt be tough compared to what Enunciate does to generate the whole webapp and servlets +Just drop me the patches and I am merge them in and I am willing to assist with the code as wellWerner +The proposal sounds like a feasible ideaA related problem is that a Namenode server handler thread will remain busy during all this time +The issue is we want to break interfaces once so we dont want to introduce any of the interfaces now +Should be fixed in Im fine moving this to Beta since our major concern was if AS works with JBDS on Windows +Removed link to as that JIRA is closed +Is this supposed to work with Pipes and Streaming? Do we know if it does +Folks any objections to committing this patch? committedOleg +Remove the which is added to mark the end of stage +so much +wanted to get the content written and up somewhere we can debate where this belongs on mailing list +It took only two weeks +Whats the issue youre experiencing with prefixed patch Cheolsoo? now I see your issue +Ive attached the files to change the site +I have found another bug +This issue can be solved by the following steps on the project +Thanks Ahmed! +See log that despite server throw exception the test runs complete! I havent figured out a way to trap the server exception and fail the test yet +Overall looks pretty good +So the URL is a reference to something on the CLASSPATH? Couldnt you simply add the String since that would do the same thing +Some problems +Just to be clear on one of my questions above I meant where do my build artifacts go when I run mvn install if I am not using a repository +I haven confirmed that the error is in the validation configuration because once you change it to required everything works as expected. +If the DEFAULT format were ever later updated that could invalidate the users formatI meant that the ctor should be equivalent to PRISTINE or perhaps PRISTINE CRLFTo fit in with the fluent API there needs to be a static method +This patch seems to fix the issue would love to have it looked at and get some feedback +I just committed this thanks Ari. +Patch looks great +As a workaround just use Hudson instead +Now we are working under separating xmodel from generation dialogs in vpe because its very hard to support +This bug needs to be addressed somehow +Thanks +if you can reproduce with rc please include what you did between starting it and stopping it +please attach the changes done to compile using vc as a patch to this jira issueThanks patch is now applied +New chapters always start on the front of the page +The browser was to locate class file location and not a class file +added so you dont need to add +The patch that changes the sleep cycle of to second intervals +As has been resolved upstream the workaround for this issue in the historyserver needs to be removed in. +Would you be willing to continue on this Stephen +Older hadoops usedNow they useTo be safe I just left the old one and added the new one so we have both +tests are done during development but compleatnes goes beyond beta as I will add some more now its enough +Fixed in version. +I must be missing something +But Im not sure that it should use a separate cleaner thread +Combined patches and used to strip the leading directory namesThis patch includes both the test and the fix in a single file + schedule for +I am going to apply and checkin the backported changes to these issues in order +And SVN diff should be able to handle whitespace issues just fine +Yeah that sounds greatI will need to learn how to write +Maybe it would be benefitial to open some followup issues for the agent integration issues? Issues with a smaller set of features may help to tackle the problem and encourage other developers to send in smaller patches like Ive done +I disabled the integrated agent controller and now I can select some options on the monitor tab as expected +I pulled as of yesterday morningMay be this is seen only in secure clusterIn the middle of more testing nowWould pull the workspace in a few hours again and repeat the testHas anyone else tested with secure cluster +Why is that an issue exactly +PMC Chair Karma is already steps done the should be live by UTC +updated patch includes tests showing the parser working even with external field filed when the field name has whitespace and quote characters in revision . +Im worried that well recommend a version that will later turn out to be broken and wont remove the recommendation text +Any update on this one Gao? Im moving it out of till hear otherwise +I think this issue is solved so closing it +Thanks. +Looking into it +I believe that breadcrumb is already the jhist file +Move to +As Don mentions the vast majority of users wont need to know or care about this +Verified that passes with MB heap +There are a couple of things going on hereThe error Could not query url httplocalhostmanagement error HTTP Error Internal Server Error is generated because gathering results in a unless it is configured and running +I think a time threshold setting would be a nice and its fairly easy to implement +You can try it out Lance to confirm +by analogy with I reopened this issue +here is a patch that adds the license file to the hashtable implementation +committed to trunk and closed the wrong issue nothing committed for this we can remove the jsvc dep too +Without a patch these issues wont see any traction. +It works and resolves my earlier classpath errors +We could place some tiny small bars to separate the different sections and round a little more the boxes on the right +it breaks compatibility with existing code specifying files but with some small amount of additional work could likely be maintained by changing File specifications to local filesystems URLS on the fly in the internals of the session builder +used your changes reverted my change in +Internally it just converts them back to the correct named value +The workaround for this should be pretty effective even if it isnt a perfect solution therefore downgrading to minor and moving to since the fix breaks the clustering tests +is a Maven plugin so you should check that this plugin exists in local maven repository and used by your Hudson +So anyway thanks for merging this in +Example how HTML Forms could be handled by an optional behavior +Forgive me for asking a pretty basic question what is the purpose of this JIRA and what problem is it trying to solve +No IT on that feature as it relies on external authentification providers. +It may give us a burden of the distribution of GiraphWhat does anyone else think +Thanks for this new patch Mamta +This patch fixes the bugIt also refactors so that it overrides a new method that has the essence of the delta needed over its superclass +By the way forget to mention the error I got which is not a lot and does not show the stacktraceThe Reference for URI cid has no +Alexander nope +The fix generalized the matching. +Closed since it is not considered a bug. +sln files they arent required are they? can I ask you to submit an individual CLA given the size of this contribution +Especially when you developed an application server framework that is used by other developersYou usually told them send an request and add handler to deal with the response +This solves the issue and the javadoc task completes quickly on my machine +Why abstract it on Android but not iOS? Sounds like exactly the feature I need +Can you tell us how to reproduce? What kind of degradation are you seeing +this is the issue with the param tag but not the a tag. +Is there a fix for this version? There are currently no plans to upgrade to version +Now that the Jenkins job is failing with this error without an obvious report saying which test has failed +I am currently disabling all unneeded test plans to get a feeling whats really required and to take it from there +Suggestion to fix this sounds good for merge +in I actually have added a check to the NN startup that makes it fail to start if HA is enabled +Closing Jira. +Everything else is junk +Checked in the files for the new page +Suppresses a diagnostic message in the generatorTouches the following filesM javabuildorgapachederbyBuildM javabuildorgapachederbyBuildM M Committed at subversion revision +I just saw the it tests in now Ill see if I can write one for a flat +Reproduction test have managed to work around this issue by not using Jencks and instead using Spring and fact that consumer stopped receiving messages was caused by the transaction has not been started error this was caused by the default JMS template in Spring would open and close the connection for each messageSwitching to the single connection factory where a single connection is reused works around thisit is possible that using will help out with this as well when using Spring managed transactions Filips comment is this still an open issue +The attached patch is a partial fix because it only takes into account that the transient entity is already in the process of being saved +May require data inside +removed the incubator section since it was merely a local cached copy in my browser +Can you please verify this is resolved for you +One could make the case that none should be removed altogether but I didnt go quite that far +Updated the attached test to spawn a fresh jruby process to make it fail independently of other tests +The rough code structure for jdb + +I raise to JIRA to Task +Thanks for testing +I have to merge the two patches since Bryans patch does not include all the parts from my patch +patched a secure hadoop deploy with the patch and now the secondary namenode is able to log inQuestion if the HTTP principal fails to login should we not stop the secondary namenode server? I think we should do that as the image calls would fail without the if the HTTP principal was not available +I lowered the priority though since the reasons for opening this bug are mitigated in. +I would imagine as CDI user you would like to limit the search +This change has been applied to branches +This code starts with the comment If the row is there already simply undelete it +When region is not handled by any RS but it is there against an RS in HMs in memory state? Any other +Could you see if this problem still manifests under Tomcat +How can a rule add and remove rules? currently there is no key word or built in function that a meta rule could use to create rule +Otherwise all future test runs will show the findbugs errors +We need to make sure it gets updated on a platform upgrade +Any update on this? +AMPM one off ci m Applying Robert Scholtes excellent unit tests from and his second patch from fixing a discovered boundary case for AMPM in roundSending Sending srcjavaorgapachecommonslangtimeAdding srctestorgapachecommonslangtimeSending srctestorgapachecommonslangtimeTransmitting file dataCommitted revision . +Probably for legacy reasons +Ive just committed this +Ive just committed with your because such a patch is very likely to be conflict with other worksThank you for your contribution +I think I can resolve that as a parms parsing issue +Closing. +Adds an additional called to the roleSets returned by get +Uploading version +Imagine the code that invokes virtual method inside the loop +fails doesnt seem to copy jars into mr distributed cache properly +AFAIK the xbean problem will be resolved in the next release of xbean also. +Its a pretty complex class so it would be good to have some good docs on it +The UI no longer requires this functionality. +A +Antonio Fiol Bonnn OK to provide element containing the DN for each entry returned in Please create a new patch with the relevant parts of the original patch is for a too old version of +About the automatic update of the Problems view I get the following error when changing the severity level nullErrorTue Dec CET Unable to update marker severity +Why do you need to demand that certain things happen as you prescribe? Why do you need to make threats about revoking licensesThis issue should be about consensus not bullyingAre you kidding me dudeI hope you start working with the community and stop trying to step on it +It is a simple change to add a missing else keyword to to keep the BASIC authentication message from being sent twice +A new version of instantiator that fixes a concurrency problem +In the new I can see following imports and giving compilation error because new imported classes are not found +VERIFIED +patch with some clean up avoids extra tableexists check multiple times in disable and drop ruby methods check table enabled before truncate +Case denyThe global deny rules is not added because for VR and SRX the default egress is DENY +In this case I would suggest testing again against a more recent ve made a change to to lessen the impact of this issue so it is no longer blocking us +Fixed in rev +I think we have concurrency issues in the table? Maybe itd need to be rewritten to use +Anyway thanks Ruchith for finding this bug! +For an app attempt that has not completed then those fields will be invalid when we read the saved object +Implementation done +reopened to remove release after removal of release notes +There is one more issue with the action bars +I dont think the xsd precludes you from specifying an id so does that mean it will be ignored in the context of the chain configuration but used for message +I think you can deal with this +After a reboot the test works either with or without the patch +For the record I am still confused about these analyzers that implement snowball algorithms why do they exist when the same functionality is in contribsnowball? +Now attached test passes +Alright then i think this must be related to +As explained above the issue has already been fixed. +Just because the current implementation ignores the return values doesnt mean theyre not usefulIs the recommendation now to throw a instead +according to bertrand the issue is easily reproducible +Im sorry but I dont see this as a struts bug +I posted this to spring +In reality the only special sauce for deleted docs is that theyre not considered conflicts +committed revision committed revision +Can you confirm that this is still an issue with Struts beta? If so please provide another patch and I will review it and commit it +See for a related issue where something similar was discussed earlier +Leaning slightly towards less code forBut if you think readability at the server is improved even if that exception is ultimately not passed to the client +slip to didnt you fix a bug similar to this one recently? Is it a different one or +bq +You can try using wildcards in the resource path +with the following caveatI think not implementing readField and write is a problem as MR users may read a record in a map filter out those they are not interested in and then immediately write it back out +Why do you think this code is wrong? is always the last interceptor and returnConnection on it will only be called if any pooling interceptors have decided that they want to discard the connection +Added tests cases illustrating patch includes my previous changes plus changes for the documentationI really would like this to be considered as I think it is a change and makes the configuration process more userfriendly. +If not a new portable mechansim can probably be built based on +In our tests we are killing the active namenode every mins and some times we see that after the transition from standby to active the namenode goes into safemode +I just committed this +We need to resolve all remaining jline issues on all platforms for the next releases or have it gracefully fall back to no readline support +if the path is of this form then the upgrade code will automatically create a user with name yyy and make yyy the owner of this filedirectory +Ilya could you describe how I can reproduce the failure? How I should change my rigths to see the failureThanksStepan +Feng can you please do the honors? If not then I will within hrs + +For Im still working on the unit tests +Forgot to mention the module should be able to participate in an XA transaction if appropriate. +At least the history search +That is confusing Sorry about that +Will tab Tree as in our other XML editors be of help +I have the same in fresh installed JBDS by +I spoke too soon about the special string being unnecessary +Can you provide me with reproduction steps to get up and running with the simplest possible app that demonstrates this issue? While youre at it please double check against +If you are able to reproduce the problem please it +just reopened so the utils part can be in the http and web commands +Now that work is well underway with AS all previous community releases are +Bumping the severity because Ive been getting this quite often and a deadlockreally prevents this from being used in a production environment +This way it matches with prior behaviour +Attaching the working view of consoles of SSVM and after rebootingdestroying the having the master branch setup I had observed console is not coming up in my setup similar to the view posted in above screenshots +onitemchange mouseouverout mousemove +Thanks Mike +Thanks MostafaI actually got a bit confused earlier +So thats a no on tests passing +Thanks +cvs dpserveranonymousvarcvsshards logincvs dpserveranonymousvarcvsshards checkout bugs with invalid fixed for release as fixed in. +These are the values used String SIMPLESECURITYAUTHENTICATIONVALUE simple String DEFAULTSECURITYPRINCIPALVALUE uidadminousystem String DEFAULTSECURITYCREDENTIALSVALUE secret +Added a very simple patch to workaround this +Will merge it into next +Could anyone check it under windowsAnton I hit upload project so youll have this version in several hours +Review as part of EAP QE +still an issue maybe just build from the wrong tag +The common way to test it would be to check out the sources from the sandbox build the projects and either test it in a code snippet or in the IDE. +There is no mess of symlinks +Proposed patch from Gary Tully can be reproduced easily using a transacted consumer that rolls back the transaction +I ran ant javadoc with and without the patch +Fix in place +This is moving away from our notion of embedding the samples and fits more with the requirements for the jbossesb ll create a new ESB issue to allow the SOA deployments to configure the target in the same manner +Verified that the problem is that the working dir is made into all lower case which is a of the fix forWould like input on if changing to use canonical path for comparison would cause any problems +bulk close of resolved issues. +not sure it is what you want but imageVersion will lock this in without having to code image per region +With the release being delayed whats the current plan for this fix BOOLEAN to match the spec or continue with removing it from the branch? Seems like with the more time available Rick you could continue your original itch to implement BOOLEAN +Thus it is moved to TBD EAP to be triaged for the next EAP release +Or add one +Thanks +updated +But perhaps runs worth gives enough info +Ill give this a shot within this is retired any way it is best for team to provide a service retrieve a list of available version key a pair of groupIdartifactId for to be ready for karaf thanks +Attached is proposed patch +So this general issue is going to be closed +ill try to get it in for but it might move to +If you dont specify the trailing war which represents the artifact type then the deployment agent will look for a jar and not a I think that the behavior is pretty much expected +is the patch for Commission Run service with test caseThanks to Rishi Solanki Sumit Pandit and Awdesh Singh Parihar +Then why dont we run the command twice catching the error the first time? +where does come from +No no it is only the change notes you add when you save the page and the that did the change not the actual contentThe plugin does not show diffs +Attaching WAR for reproduction +Moreover this issue was closed dup of which was closed as a dup of this one +But how would open succeed? It turned out to be a synchronization problem +From the image edits that Ryan posted to it seems to me that HDFS may have missed some edit logs +Not sure if it would be best to create a new mojo for this default behavior or work it into the existing mojo with some kind of flag +Oops I meant to resolve not close. +So how can weconfigure besides system property? System property works at least Itried it by adding it to +Bruce a patch would be best even if its trivial +bulk close of issued resolved in after X release notes were generated but before XDT finalized their release. +pptx file using MS Office Save As from the +To fix this I think wed add methods that create properly encoded data and report its performance +Immediatetrue doesnt skip validation +Pending resolution. +This may require the tests to be converted to tests rather than the currently annotated tests +Works greatThanks for your helpRomain +Can one of the Java committer take a lookAside from that patch looks good so we need to remove rename the dir rhmCarl +Sergey Spec you dont allow me to help the project which I so obliged to just because my server is located in USA isnt it +we are using the apache wiki for install and use will be out of the office Aug th. +Definitely in but maybe earlier. +Attached is a ij script that automaticallyexports and imports data from tables it creates byitselfIt tests the procedures andTo use it to test imports in replication just do theinitial replication setup in ij and then dorun I thought the script might be helpful for peoplewho wants to test imports with this patch withoutgoing through the creation of the tables and filesnecessary +Please package your code in a proper quickstart app so we can debug it and fix the problemThank you +Slight changes moving bucket initialization out to thread creation +Marking this as a Documentation issue also in case the resolution for this issue is just proper documentation of the problem wrt databases in a jar +Its disappointing +See this JIRA for more info Aaron for letting me know +The colon isnt the only issue +michi and flavio are you guys good with this? please express concern or if i dont hear anything by tomorrow ill assume the latter +This just isnt how is designed +Thus it could be bundled into JBDS +bq +Here is a new version of because the old one could not be applied correctly. +FlorianThe behaviour of of course is incorrect +Didnt work under SEAM Icefaces this related to the conversation never dies +Its almost as though we should chunk from hfile all the ways on out to the clientTough one is on the way in +Include with it a brief explanation of what it does +solved in +Isnt it absolutely pointless to repeat the same information again? I dont think that twitter is a good example here +Committed revision +More meaningful changes have been merged from the branch for some time now so resolving this issue +Now I have BZ karma I have ported this to +patch for trunk +For the pull requests in they have been integrated but the requestor has not closed the requests +So what needs to be done here to get this taken care of +If that doesnt can you try maven or maven +Corresponding test environment for NetbeansSee in testQUBsrc investigationsThe first problem actually happens if the client stub is used inside a simple web client that is not using the skeletonSo it seems that the For input string issue is actually happening when the stub is used from a servlet context and not happening from standard java applicationWhat is wrong with this web service? Why the client stub doesnt work in a servlet context +Just for clarity +Integrated first suggestions from the review +Alex can we close this + +Will do +looks good to me +I would expect this to be a separate module +ant build task updated in svn to fix these in RC the directories arent there now. +Because we want generics to behave in the same way for classes and structs we have to avoid making unnecessary distinctions between structs and classes in the type system +I will apply this patch and try it out +Any update on this one Paul +Please close this bug if you are satisfied thanks +Component metadata values will not be ignored for bound components and will take precedence over values initialized programmatically. +See +There is a int overflow for setting multiplier where the type of multiplier is long +Patch removing apache from all Solr artifact namesI havent tested it yet this is just the result of findreplaceTesting now +Well be testing this internally over the next few days +Is there any update on when this will be completed? We still do not know how to test +sorry for the delay pipes is now mavenized +the change should be better done to attached patch what you are suggesting Jochen +Brent already tested this thanks man +However it all revolves around XCAS files and does not work with a directly +Jan Can you please work on this? This should be a good task for you to verify manually first then then by addingupdating Metamer +I just stumbled on this too the documentation is a clear contradiction in this case and makes for a bit of confusion in constraint definitions on CommandDomain objects. +I experience the same problems on Linux +Jay Is there an example of this? We need to make this more obvious and perhaps provide some examples of how to use it +Follow on issue dealing with two remaining failures is filed as +Committed to branch at revision +I have confirmed that the problem is fixed in. + +I am curious why this takes so long to be fixed +Let me know +Suggested Andre I was wondering if you could have a look at this if you look at the screencast the combo behaves very very strangely +You get a runtime error is that not OK for youThe problem is that we could provide a compile time error here but only because a static field is used +bq +Looks good +Patch committed rev dbfddaeabfcccf +Committed to trunk. +The new patch is attached +This is the only case where we see concurrency problems when prepared statement is a class variable +Assigning to you Snjeza as I think you can fix it way quicker than I can +Colm could you please explain this issue a little bit more precise? Sample +Hi NagarajThe xml size is quite small no more than KBThe memory leak happens all the time whether during load or not +Also known as Dynamic Attributes +Once its worked fine the changes will be applied to shot at fully utilize the plugin in openejb +What about Servlet and? If the specifies one of those versions how should the container behave? If the container supports does that mean it does not support or in regard to compatibility? If I deploy this on a servlet container I wouldnt see the same error +The last failing run was last night so I think we should maybe run the test suite locally with and first before reverting parts of my commit +So the reason of the test failure is explained by the K restriction on an environment block for the new process creation facility provided by MS Platform SDKIt leads to More data is available error rising during a process creation leading further to The creation of the Process has just failed +zip files but we should probably provide a description for the in both versions +So all issues not directly assigned to an AS release are being closed +This sounds like a quite critical problem to me +In the log above there is race with threads +Have you tried using the library +I dont know if Roger integrated it or not but the Weld nightly works fine with my patched integration module +most of this was actually done in bb by having hiearchy mode and content filters. +Bug fix +Applies cleanly I suspect there is a charset problem +So I am for the suggestion +I have changed only column for am not sure is it needed to apply component prefix for and other db objects that might be component specific +Revision Aaron +is this fix going to make it +Fixing this will also fix since I have no ability to complete that +Ive retested with the original build and its all working fine with +Ill get to for of which you also opened +import is good practice to limit the importsGot it! Thanks for the clarification +Ive seen similar performance issues when there are a high number of reducersChristian which linux IO scheduler are you using? Can you try switching to anticipatory and seeing if the problem improves? +Stevo First thanks for your help +Looks like this is intended behavior +Thereare still issues to be worked out before we can remove those +I considered adding the Override to most things but I dont want to make applying more difficult so I think we should wait. +Not seeing this in the code a patch would help +I hope its ok +Remove static field should be accessed directly Eclipse warning +So it would be cool if there was a means of having code executed just after a user has been authenticated in order to load roles etc +I verified with CVS and it looks good +The class will now produce warnings as it parses misplaced ignored elementsClosing as fixed to be part of. +Please review +The build failure seems to be some problem on Jenkins side the issue is being tracked by. +Resubmit to jenkins +Sure thing Claus +Greg Noelattached shows an idea of how to make edt police more people friendly what do you think +so whats the issue here? do you guys want to keep itjim please elaborate +The project doesnt publish nightly builds +Ok thanks will check unit test results to see if the stack trackes are gone away in current unit test runs and will close it +Minor update to remove unused imports +This is a very uncommon similiar events affect other I designed in boundary testing for HDFS. +I am reluctant to change the stored format of Derby +Can you please provide more details? +This patch exposes to +Thanks Hairong! +What do you think? +IMO we need a separate column for +At the very least the mapping code +In any case I think this should be treated as an internal performance optimization and handled entirely within so I dropped from the list of affected components +Now that work is well underway with AS all previous community releases are +No change for LinuxSolaris etc and these would use the default implementation of the JDK through the native callsWill test this through later once I get time +I will also download the project and see if I cant patch it myself and make a contribution to an already great project +Eugene please let us know why youre removing all new comments in this issue +Integrated to and TRUNKThanks for the patch JonathanThanks for the review Todd +It isnt correct yet for running the test as it gives terrible results but I am hoping a second set of eyes will help +branchx rev +Will update the launchpad to once the framework has been deployed to the maven repositories +But in QE environment we cant change Ant installation or settings. +Hi the PR upstream does have mismatched packages +come to think about it I researched this a while ago this will create up to at most directories then it will recycles those directories so it may not such a big issue +We could also consider a defaults element which could wrap a block of configXML +Does that mean that the original error is not occurring anymore? Or just the +I have everywhere except this is the only dependency I have left pulling in the old +I can see using etc to clean the code up some +Thanks ChristianI have applied the nd patch +Committed! Thanks Liyin! +Yeah that builds zip look fine to me as far as I am concerned this can be marked as resolved zip pattern is testoutputTEST +If you go to a jobs configuration page and search for submithost then youll find both the hostname and IP address of the submitting job client under and respectivelyMoving this to the MAPREDUCE project since Im assuming this is relating to the server and not the YARN application history server which isnt part of +Sample data that cause the error +So we cant remove the xalan dependency until gets fixed accordingly. +BTW this worked fine in version and +Its correct that this version is based on Oryx actually Signavio Core Components +thoughts +I am no ivy expert but it looks good to me +For example java package name starting with upper letter +Thanks! +Thanks amila for the contributionRegards the resolved issue. +I was able to start one cluster up but one one other cluster I had this issue +See +However it will be looked into to have delete the header blocks + and to support database as an entity locks for database in inputs and outputs database in outputs user must have All permission to execute the query +A good example to think of here is the field +I suppose this is still unresolved as is +Thanks TonyPatch applied to LUNI module at repo revision rPlease check that it was applied as you expected +trunk should correctly report darwin for both values as of revision . + +however i will be getting samples the day after +Hi EdgarMaybe it would be a good idea to load the SQL requests from a property file so that tables can be renamed without having to recompile the codeThis probably requires to only use but then again this might be a good idea for performance reasons too no +Should be good now +Thanks for the review Jody +is too small may be bump it up to existing number of RPC threads or something in that range +Image example of my to fix readds ulthe issue you report is due to a recent rework on theThe html of the menu has changed in structure in order to be compatible with a menu theme +added +I was just pointing out for the two issues Kathey raised it dodnt seem to me that there was a problem +And the serializer in most cases only outputs the elements it get reported some of those serializers will go crazyThe reason for this is because SAX is in general seldom used to generate xml documents its more XML parsers that report elements they found in an XML document +committed this patch to the master branch +Demonstration app to show the problem +Verified again that doesnt affect the success of the tests +Though this information is in the POM there is a benefit to having it included in the findbugs plugin report +I tend to include this after review at the final patch +Never mind my last remark I was tired and confused +right? But HDFS is not the only FS we wish to support and not all of these will have a sufficient CRC system +This is sad but I will open some issues on the plugin to change thisI developp a bit more my sadness while reusing the mojo inside anoter mojo we can not use the parameter injection via javadoc annotation so the code is not reusable at allCorrect me if I am wrong +This is basically a dup of or +This delay is expected +You could add the functionality to your scanner plugin but Im not sure it is the most convenient thing to do +Setting version to AND to remind myself to test this in and if necessary apply patch there too +All we need to do now is stop the CLI needing and well all be singingUpayavira +But the patch will need to be modified because it affect the code which does the injection on every request for both singletons and resources but we only need to do it once for singletonsI have to think how to do it better especially given that also requires supporting the injection into Application instances so we need to deal with injecting Applications and also with injecting contexts such as etc into the injected Applications ouchthanks Sergey +Note that I have tried the same test with AS and it works there +Setting the Fix version accordingly +Committed as svn resolving. +Could you this on your end? Thanks +and I are exploring to setup a Nexus repository manager +Just running scrub leaves Cassandra with that behaviour +Yep I definitely agree we should commit patches that the ceph folks need for integration but the actual ceph FS implementation seems like it could be implemented externally +I discovered one other step after applying the patch youll need to chmod x scripts + +Geoffrey can you provide an ARQ test for this +I do not agree with you +So Im actually using Forge and the scaffold on an existing Java EE project to add CRUD +Either the scheduler must be distributed as well or in the simple case the batch of commits queued by the scheduler must be retried in corpore +Empty string null +It happened before too +Easy enough to change this however. +The patch also adds the new datastore timeout setter and getter methods to the signature test input file +Thanks Enis for reviewing this one +Thanks Leon for the patchits committed to backport in for in at rArg +Resolved after struggling yet again with some strange mavenplugin behavior +So delete would win automaticallyCreated a jiraupdated a patch so we can ensure that the winning update is determined based on the order in which the client adds operations to +Attached sample project that appears to show this adapted your example to look more like mine removing the last fajax makes the popup alerts appear +This was not a bug but wrong usage of the dynamic +The single quotes are never stripped its only the double quotesI tried to remove the stripping it broke two tests test in thats the test of it checks if the quotes of an argument are removed after the parsing +Hi ElenaFixed at r please check if fixed as you expected +LDAP is covered in this guide +If theres bandwidth in the period I think we should either give the Thrift gateway attention or deprecate it and remove it in favor of the CC client side API +It uses a different platform name than the rest of the build +Have to revert this change because now the BPEL resource cant be saved changes in the Design view are not reflected in the Source view and the editor is never marked as dirty so it cant be saved +Good catch +Absolutely! It must be that I did not apply the fix on the relevant branch then +Patch had rotted +seems the right macro is MACH its already been used in other parts of the launcher +This patch makes it generate the complete list +I will update the documentation to reflect the change. +Thanks for your review Pat and Camille +This problem also exists in version for the report it helped us to add synchronization to xsltc +I have put the negative value for fontsize i am getting exactly the same error at the same locationYou can use the attached sample application to reproduce the issueThanksRaghu +Theres also a patched you can use to test whether your issues are fixed +It still didnt work +Thanks Maarten +Hibernate will then try to fetch a Connection from it on initialization to determine database metadata which is the failure that you see +comparison should be changed to compare TINYBYTEARRAYSMALLBYTEARRAYBYTEARRAY +Closing jBPM issues resolved in a released version and reported by me. +Its not at all obvious to me that this is a clustering bug +How can I run locally our maven tests against +I just explained the problem in my comment +will test and commit if tests s the reason to change +Is this commonly modified by users? The general practice is that youre allowed to have configurations that arent in the default if theyre rarely changed and arent going to be guaranteed to not break between versions +I see the region deleted at +the latest patch looks good to meThanks a lot Colin +Hi ThomasI am using community version doesnt Enterprise problem this jira is the same of my problem? resolved issues for already released versions +the sample signenc is correct whit cxfI dont understand what the problem is could you please clarifyColmI still dont understand the problem +And now I want to close one of them +Patch that only caches the parameters object +Thats why we ceased to create this file +Continue investigation +Ran patch x locally without seeing any weird output or failures +Looks like this change will not apply directly to as fix for is not in +One do you have another driver you wish to use? It would be good to test this with another driver +Regarding sleep calls I meant we should not be adding any new onesThanks +Excellent! Ill remember next timeThanks +its easy to clone an example file and appendmodify the default description to add site specific notes +diff file which should work for Shuller thanks for the contribution +the patch on trunk +A cleaned up version of the patch that we were using for running scan benchmark on a clusterAlso added a simple test +Thats right +When the first assignment reports back to the master that region is open because of the state in the master it thinks it a rogue regionserver coming up deploying regions it shouldnt have +It would be quite difficult to track JS issues with the minimized version +Not really sure why not yet +Bug in property view when the server is not connected. +Hmm my bad I accidentaly removed some part of the srs lookup be fixed now +JIRA takes care of greying out the older versions +even without your fixTesting with your fixStill no exception +The scripts with BIGINT works fine on a current ASE installation +Under the Project Documentation and before the built by logo. +Its not clear to me why a new transaction is being created when there is already one active here +is weblogic satisfied with a package private constructor +Looking at the threads it looks like the clients are stuck in wait +We can partition the work among volunteers +Right wed make it clear the collection is unorderedIt just seems like since we are building up this collection anyway we may as well give access to the consumer +Assigning to Laura as she is the new lead for EAP docs +I am tracking it down now +Consistency between client and server configuration +Please review another patch since patch b cannot be is causing invalid thread access when started from server +to include the class but not export it +Requested Derby RickYes this indicates database corruption but can you tell me under what circumstances this might have occurred +Please make patch starting from the source root directory +Could someone test this on a unix machineIve just realized that simply renaming a page isnt a problem in my installation so this bug may be a windows file system acl issue only +Although it sounds simple there are many ways to approach it and finding the most appropriate is what makes it tricky so well revisit this in +Shane Ill pass this off to you for comment +Given how many nasties are floating around the network these days I havent investigated too closely +Some interesting stats +We will make the change first thing in and then backport ifwhen it proves stable +to trunk +Heres the pull request for +Greetings +I did have this in the back of my mind when writing this but it seems an extra feature to me and would bloat the scope a bit much +RF was just released +can we close this now since jbossws has been updated to +Thanks Philip! +We now ignore non existing jvmenv variables if you have ignore missing location enabled. +I apologize for accidentially creating this issue twice +Additional seam runtime War projects for action fails +Loose scripts as you call them are usually standalone so are never referenced anywhere so that can be fine in those case to use whatever name a user see I believe that nobody ever uses such names to avoid the problems where you cant reference that script or the console is not just for loose scripts but also for normal classes which will be reused referenced and which will need to have a valid if we follow this issue were restricting users to use valid names but I dont think it should really have an impact as weve always tried to tell our users to use valid names in the first I expect the impact to be very m happy to have the feedback of others on this if they think Im wrong and that its common to use a name. +Verified on JBT build. +Will commit if all passed and if passes again +Essentially name takes a deployment name that requires an extension if you dont provide one we assume its a stanard jar so ears will not deploy correctly +Found while testing rolling restart on +trunk all resolved tickets from or older +Re Apache Jackrabbit +I just noticed upthread your I am not a javaprogrammer which explains why youre not using the regex thing and a few otherthings Id wondered about that a java programmer might know about but you maynot +By using the new config parameter this issue is not a problem anymore. +Resolving for now +Im going to try to see how far I can get without changing the existing code but good to know I can change access levels if I get stuck Thanks +igor removed thanks for the hint! +I really dont think anyone will miss it +There is now a way to be ok with the suggestions and if not ignore them in case openshift changes the dependencies +Adding more than role object to the core object results in a +So different attempts to run map would result in different eventsThat said however we probably should make another event lost or something for tasks that are lost because their output had problems or the task tracker was lostWe may also want to flag the complete events of lost tasks as obsolete so that reduces dont see them and try and fetch their outputs +batch transition resolvedfixed bugs to closedfixed +Theres a wider audience for discussing this on the mailing list +How about add metrics stuff in addition? I can see a Chore in being a useful place to drop in other stuff over time +Perhaps for avro we might consider more direct support for optional elements at the schema level language mapping level and also perhaps in the encoding +msi installer artifact should put the LN files in the top directory where it installs to +Perhaps we should allow this and both localize the file and unarchive it? What do you think +Done! Thanks Jake +close this as it had already been fixed. +Thanks Anupam +That discussion could be on orThanksdimsThis is fixed in the release +Initial patch +When I said The returned data structure doesnt match the definition in the WSDL and WSDD files I was wrong the struture is correct but the data is missing +Resolving +All tests are currently passing Ill post my first patch tomorrow +can you try do this with plain and see if you see differneces +what are you jira +patch for issue for issue good commitedPs i dont mind if you batch up a couple of naming constancy fixes in a single patch some times too saves work +start nodes +yessorry for the confusionPawel pj +So there is no possibility to access T at AE level right +They may have other dependencies in myfacessrc that are required for them to work at runtime +bq +Marshall with AOP exluded from RHDS and future dev for now is this still a priority +The propagation of the logj properties file is via HDFS and it doesnt look like it is handled correctly +I think its ready to commit +If you think it is ok I can commit the patches +Will try to finish the over the Weekend +patch to add check for invalid socket on readwrite StanleyPatch applied to LUNI module at repo revision rPlease check it was applied as you expected. +Bulk closing all of these resolved issues. +Is this problem reproducible? If we get a test case that shows the issue then we may be able to identify the root cause and fix it +Thanks Cameron! +Closing the supertask. +youre right that the menu option is not there +Is this the latest version we can use +Applied branch and trunk. +Couple of notes just to follow up herePostgres is BSD so it is fine to includeLGPL dependencies are optional and may not be included in the distribution +Can we close this? Does anyone have any interest in fixing this? +We may need to have a distinction between operators that have plans and operators that dont +since this is a relative major change I think either Avinash or Prashant should take another look at it +OK I opened for ngram love. +Initial comments over there are addressed +See line comment Invoke the to add entries for block components +Thanks to the reviewers. +Daryn different unix have different rules for the valid usernames still they have rules +Duplicate of ari +I share your feeling that where ANSI does propose solutions we should consider them carefully +prop where the last one currently fails +Test failure in patch for review +What do you mean by it stays here till it times out? What is the message +we should not change the behavior of a performance test in a point release this would break the comparisonsAdding an option would be fine +This appears to be working and has not been commented + +I notice one more thing here +can you try to upgrade to seam that allready includes this patch +As Register s visibility is SYSTEM this node and its children are not showed in navigation portlet. + +Having said that lets hear what Bruce et alias have got to say +yes but I still need a way to enable the lucene stuff +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Ill touch and note thatIm going to leave this open as I want to get the tests into our framwork somehow +The fix will be a part of release +Thanks Erik ill commit this to trunk soon and we can patch against that for distributed can be back ported easily I dont quite know the ropes on when and how stuff is backported +Created a patch according to thisFixing would be the best solution ultimately. +Slightly different after all +should be fixed on commit i run my ide from windows so the diffs will have rn in them if i understand how svn line munging works +Unit test passes and our production system no longer exhibits this bad behavior + wasam hesitant to move forward with this given that many of our previous releases are missing from the archive site +Jasons rationale for this change and subsequent explanation sounds reasonable to me +Committed to +youre right says a bit am afraid there is currently no workaround for this +Attaching +The issue is happening only in Ganymede +Resolving for Daryoush because he doesnt have full Jira rights yet +I dont see an id on your publishTo line +I was erroneously using a release candidate version of Hibernate Core instead of the final release +only warning is solved + +Attached the test case +OSOA closed this issue with no action and will not update the spec here +Resubmitting with the proposed fixThanks for testing the patch Travis + has been incorporated in the C compiler from its first commit the current r +fixed issue with undefined javascript object in MS IE. +Would be interesting if you could create a junit test case that can reproduce your issue +Can you please either a create a separate patch for the formatting changes or b create a patch with just the new tests +theres no intentions to create a branch for each bug +The RESTAT spec should follow the JTA spec so the former needs to the spec is correct since the status is so this JIRA is invalid +Thanks Farah for the reviewTake a look at the screenshots and let me know if it implements what you recommend for the installer +The javadocs were also incorrect so I updated those as well. +Lets make sure this is completely fleshed out for Beta +Then generating a patch file that I can use is a matter of doing svn diff +This feature has been implemented inPete if its still doesnt work then please reopen it as a bug with step to reproduce the issue +Some observations A significant advantage of FTP over HTTP for access to HDFS is that it supports directory enumeration and other operations +Verified by Mark +Creating an issue with priority Blocker means that every Sonar user is impacted by your issue which is not the case +possibly uneeded jarsICUJ is only needed if you want to support like arabic +I would also find having the configuration options exposed very useful +The only thing missing is the additional work to expose it in Annotations +This is just a marker to clean up +Your second patch has now been applied and a new SNAPSHOT has been deployed +Authentication is done at NN +Aaron thanks for the note here so that I would notice the new jira +This is checked in as revision . +This operation will be based on BSP instead of MR. +Ah it is only a WP issue as WP changed how it handles the basic getPicture function +should be fixed in and webtest plugin release verify +Thanks for the patch! Ive simplified it a bit and applied on and trunk let me know if theres a problem with the modifications +If someone has a patch set that works please find a way to share that with Dave and the team so Oracle support becomes a reality +Are you talking about WTP Servers View or about Server View +committed thanks pranav +I havent retried it in a while but if you tell me its fixed Ill believe you +I think we canupdate the CVS since the change implies only library updates +trunk its only a few options you can configure on the Timer itself +If you like the changes I would be happy to supply an equivalent patch for Hibernate +Im building against Java locally and not seeing any of this +This way if those functions eventually get magical ability youll benefit for free +has been marked as a duplicate of this bug +If you care to try it on a Wiki where you have protected access Id be curious to know if I broke anything +I will keep my eyes open for your commit and let you know how things go. +Extend the existing resources for that +how to write a proper Gzip interceptor +Any update on this? +Attached a final patch that includes your suggestion about renaming the variable in the styled shape painter and adds a missing header +Is the only place you see this in? If so please close as Out of Date and put a Fix Version of +Since it the approach sounds reasonable to you here is the complete patch with unit tests +That part of the API has now moved to reasonersbase +Is this just a matter of taste or is the current text erroneousNote that other languages dont have exact translations of the English original texts eitherWhat about the other resources in? Could you check them too please +For a single build these would be snapshot references until release time when the release plugin would update them all +The fix doesnt show up on subversion commits but was checked in with revision +I would like to suggest that should have scrolling capability because it is very painful to add it later +I think the behavior of harmony is reasonable here +Committed at subversion revision Touches the following filesM javademoD javademosimplemobileD javademosimplemobileRemove the demo appM javademoD javademosimplemobileM javademoRemove documentation about the app +Corrected to in calcSimilarity call +We could probably do the followingModify the scope of to do exactly that and then reopen this jira and revise the fix to just accept a boolean query parameter to determine whether to include status or not in the listing +Eclipse lets you remove all unused imports in one shot +bugfixThanks approach to the problem might be to eliminate all mention of soft upgrade +Firstly why do you have to update a bundle thats managed by fileinstallSecondly assumption that the location is a URL is a problem in itself +Merged into master +However I would like to provide full support of with XML nesting elements +AnneI am not certain but I believe permission checking has been left out of order screens intentionally so that anonymous users can create ordersI dont know either way +The successful jobs should display success +On page before you start covering each of all the scripting and expression languages it would be nice with a segue about likeWe will now cover all the languages from table in the following with a brief description and example +Same as my wiki page on +I just noticed the difference between what you did and what I did I removed the default namespace from the instance document not the schema +possible fix is to scan for the ID if not found at the head +Thanks Sreekanth! +Give us a few days! If no progress then well move it to +blocker forClint you down with that + +To be clear above and comments were forCould have the issue fixed +For reducing the connect timeout are you speaking about? Is there a way to change it for connect timeout onlyYes I was referring to in trunk +I used the current trunk version patched with and and varied the buffer size of the cfs reader +Closing. +Preliminary property set for topics +I would very appreciate this feature or better said standard behavior +I have a patch almost ready that cleans up unchecked conversion warnings in package as well as the deprecation warnings in the tests +Thats the one that actually effects me +Same here +To enable the new class loader simply use ve inverted the default +Youre must welcome! Happy to help this great project +You should change the packaging to jar to solve your problem +Regarding referencing a temporary function from a view I guess we could just allow it and leave it up to the user to reregister the function in a new Hive instance before referencing the view +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +As Ive recently had this pop up in one of my apps +This issue should not be closed without being resolved +can not reproduce with keycache turned off on think i just reproduced this with key cache turned off +show as being fixed in +lets do that in back to was done a while back. +Patch applied at r thanks for your contribution please verify +Can anyone confirm that with trunk they see different values for and docFreqI can confirm that the variations disappear after optimizing one of our clusters +There are plenty of places we could add more null checks and clean up the framework code but doing so makes the framework bigger and slower so we want to make sure we spend our time and your time wisely that it actually makes a critical differenceOtherwise we should downgrade it to minor +Id say thats a resolution to this issue. +Seems to be working ok +I have attached a proposed patch to which also addresses this problem I believeThe proposed patch is slightly different from that suggested by Saurabh I modified thebind method in to populate the defaultInfo as needed +So either theres some bug in this cycle or your retryMax is greater than +Fantastic! Builds successfully for me as well +this works but might be confusinginconsistent though +And the migration from iOS to doesnt work either +I would vote for the JUL bridge if we suppress the messages might we miss something real +This issue has been fixed and released as part of release +I meant the registers we use for hardware exception delivery +It uses pure Expert API no planner needed +I will see what can we do in the code to have the stack trace printed no matter what kind of build we are working with +Thanks for the nice test! fixed in latest SVNthanksdims +For this run of about hours billion nodes were ingested and nothing was lost +Then you have a th TAR the TAR +Updated patch to trunk and closed filesystems in which fixes the other filesystem leak +Additional info to Mirkos comment +the patch works with no svn revision available +Wont be moving the issue to the makes sense +I can unfortunately replicate this +Committed to trunk +It is a possible added utility +This has a dependency on authenticating these users +It looks very much like a Spring problem +I should also add that as far as I know doesnt extend its support to and hence does not provide an equivalent ofIts clear for me but confuses at first useMy apologies for disturb +As far as I know that would meet all our needs +Thank you so much +Added a test to demonstrate the Mike +is taking care withIt seems that there is no JIRA issue for yet +Thanks for the patch! +Now that there are releases around Ill try it with and see what I get as that is where this code will end ups no need to test it with old releases +I think defaulting ajbuttonlink to form is a good idea +Wing Yew is right this is a JVM limitation that is difficult to work around because one doesnt know how the code is going to be compiled +so we have to merge settings file +Will commit when tests pass +Any idea how we can proceed? +verified CR +It does verify the indeterminism nature of antlr +deploy has been done for correct? The creation of a datasource is there anything that we want to do there since an embedded datasource connection could be created just like any other datasource +Thanks +The root context is supposed to be managed by how do the action support classes get access to web application context +There was a rumor before that a WAR file within an exploded EAR worked but an exploded WAR within an exploded EAR did not +class annotations should not be propagated by the Java runtime unless theyre Inherited but the same does not apply for overridden methods +The second error you are seeing is because you are missing inyour CLASSPATH +I was talking about Christian Kberl not Roberts one +Just to add the process works in in error +The attached unit test can reproduce this issue +Issue could be resolved but may not be see for further details. +I see your point +Patch for yahoo distribution security branch +was also applicable to +Fixed +and another that is then used by the WAL file streamingIs there some reason the WAL encryption could not use a compression codec +You always need to make sure that if the return value of the rpc call were to disappear into thin air that the problem would be detected eventually +It now works as it shouldYour issue seems to be a different one +We can work on getting this in for +I am just wondering why these options are not set to true by default if it is so critical. +org build script good to me closing! Thanks +It is to fix the issue raised by some beeline users +I would advise against removing tests from the default build sounds like a recipe for people forgetting to run tests! Why not just add a few profiles to make it easier to select what you dodont wantmvn install builds everythingmvn install Pcontainers builds everything except containersmvn install Pcontainers builds everything except of making profiles that are special to our build that needs documentation to be used anyway why not just teach the userdevs about the features that are in the build system we use +Tiago I looked at your patch and noticed that you are using from Apache Commons lang library +Its odd though that this is not working +Some of the stuff is not implemented +Please create his account thanks +There is no specific unit test +Is this a change that is specific to the branch or is this something that we should commit to trunk +Mass closing all issues that have been in resolved state for months or more without any feedback or update +for the patch +Committed to branch +can handle the large query but the delete is so large and slow that it takes forever to execute by specifying all of the in the relation table explicitly +If you are able it would be excellent to get your contribution reviewed tested and hopefully committed in the short term +Though I think this has always been true as the last batch of content server fixes only targeted dealing with cache controls for browsers correctly and did not touch the resource resolution code +The issue is that in the current patch we first calculate how much needs to be removed then we update the size of the archives then we delete them +So a upgrade to is also needed +Moved to. +Something you might want to take a look at Jervis +I added a fix to trunk can you check it out and try to build now? I have no Sun machines to test on +The only difference is that I think it is better to not expose and to users beacause certain fields like actions etc are framework level +I am guessing that since this conversion happened many months ago and Xalan has been using Jira for a long time now that it will be difficult if not impossible now to find the garbled data without hopelessly screwing up the database +If you do not believe just look through the changes list for any version +I am going to leave it open until the user who reported it will be able to check the patch +Applied +Thomas is porting this module to is the list of issues that we want resolved for final +JD TedI have included the suggestions except the The new patch still have a few offending lines +HiI looked at the patch and now I realized LDAP encoder generates arrays of +Oleg please try this patch +Could you please take a look at this again as it has a different semantic from the other ? feature requestIm concerned about consistency with the ? operator +Uploading a new version of the draft functional specification version +I attach here a patch with a POC test case for your consideration +rev fields that read regular expressions like pattern should be literalOnly in other words does not read because some regular expressions are not readed correctly if is activated +Taking a look right now +Never mind build is working again for me in trunk +Ok just wanted to be sure i understood its current would be cool would be being able to add styles to external files AND be able to editupdate them too +Your comment is not readable as is +Patches checked in with some minor tweaks +It seems that this enhancement is atm out of scope for Sling therefore closing this bug. +Thanks Mikhail +Does this feature have to go into? Can we do bug fixes in only +All other resources including and the custom dojo build are now in a separate module called +Fix verified. +So the distance calculation for null is wrong +patch to reduce size before releasing lgtm +Tests are not required hence and the single test failure seems to be due to something else as my patch touches zero code +This year I have been programming with Jetspeed Websphere Portal Express and JSF to check the old implementation may be useful to build projectsWhat about to create the typical greetings application + +Financiers and Marketing are driving it and sometimes it really scares me +Cancelled and retried and it worked +Thanks +The failures are when getting the object by id where the oid is Person whether validate is true or false it returns a Person instead of a +We have seen one other deadlock in the past when trying to lock the Configuration object +hcatalog pulls them because it externs hives svn. +Sample +Adding it here as a place holder in case we still need to refactor in this manner +Attached patch to update the Patch +I receive no email responses from the server +verified +bq +What Pi suggested is a dedicated SAMPLE operator +Would it be possible to enable JDT weaving and see if that fixes the issueLet me know if that Ive added change to Spring IDEs integration with AJDT which should speed things up +This wont preserve history in SVN so the rename and other changes will have to be split into different patches + far as I can tell there is no reliable way of knowing when user stopped dragging the window then cursor is outside browser window +On my system it looks like upgrade tests run first and thenthe encryption tests later +ve test move query to as a ve test to trunk +Ugh and certainly Spring should be whittled out +All the quickstarts appear to be working now in both remote and managed mode +May be not all properties +The patch doesnt apply cleanly on +This would include things like updating adding to test plan for release etcThis is not for the actual testing of the component for releaseThis subtask should be broken out as its own task and not be directly in the parent +If it cant find the dojo compressor it carries on which is what Im assuming happened back then. +Thanks for pointing that out! Im contemplating chocolate +When you say unistall do you mean delete the folder and reinstall the other version? I was not able to find an unistall option in the start menu or in the Windows AddRemove Programs +Committed to with revision Committed to with revision Committed to with revision . +Im probably going to make a start on doing this Sunday th any problems with that shout up now +OK +Also we areinterested in this feature for branch and trunk +line of the diff +Could you be more specific +I believe the motivation for this escaping was the use case of filtering properties files where backslash and colon have special meanings +Im sure a committer will have a look at it soon +Users I know have lost hope trying to tune these sorts of params by hand. +Daniel do I just post message to RegardsFred +What do you think +Reopening then closing to deal with migration bug +So here the patch for replacing directly with sources +This was useful in testing and would also be useful in Example Generator later on +for the change +bq +Thanks a lot +Bulk move issues to and +It least I can search and then my words highlighted in document +Will port to HEAD later +No just field collapsing +cvs is up now +It went in with the patches for. +is different enough from that we just didnt have the time to get it do provide suggestions on how to upgrade via warning messages in the editor so thats there +On windows client gets connection refused after happens immediately +I dont like having an arbitrary character like signifying a sort type because it does not explain itself to a user +Change the assignee to myself and reassign issue to version +I unchecked these two checkboxes and tried again Im attaching the new patch not sure if this will solve the problem +It seems that the preflight sutff still works +The IOE says what the error is +better summary for this issue would be Bug in static method in DIGY +It seems like some sort of warning should be omitted when the user picks an alias that clashes with a native name +the difference is that it would rarely happen since the failure state of edits logs and image dirs was coupled +Attached the file from sf so that it cant get is a version of the file from the fall I didnt update my code to the code from the trunk versionBut it seems to fix the problem I described abovePeter +Paul also pointed out that in this edge case couchkeytree will need to deviate from its usual logic of always preferring the existing value over the one in the inserted path +Please reopen if it happens again. +btw varchar is not supported by the jdbc driver because of that railo pass the value to the driver via setObject +New patch addressing additional comments in +exporthomerayeesSCQAAUTOprimary exporthomerayeesSCQAAUTOprimary +The cast to is unnecessary and a cast to should suffice +We can reopen if theres interest +I wonder if now would be a good time to do a little more sstable cleanup? Here are some things I came acrossFor what its worth the nd and th issues are handled in +Post a listing of the index files with their lengths ie ls la +Today Ive managed to reconfigure sybase server and it seems to have case insensitive identifiers now +So instead we should restore a storage without storing creating new fsimage +Anything new with this feature? +btw works fine on fine on Windows XP +Added attachment for test caseAnd one for the case of adding a paragraph to a section and then changing it to a Heading +After stepping through the current code carefully in the debugger and trying several failed attempts at changes I learned several more thingsFirst as Kathey observed on the mailing list qryrowset is commonly when clients perform a fetch of result set data +It uses a private static createUri to create the uri +Verified with. +I think this is much better if you want to see what the test strings look like now have a look at ant test Dtestcase Dtestmethodtest true +I think Ive now spent more time on this patch than I did writing the code +Could you please link the task logs? Our unit tests seems not to be really accurate +It just knows that thefield cannot be added to the given I might try to add a way to get the reason from the +Thanks +The jQuery is not developed or maintained by the Roo project team +I also create a var and scope also join the tld of the new patch +but the relation is still not saved +We can close this issue +Nice work Ken +Spring does not influence the SQL generated by Hibernate in any way +yeah the optimize works for now but i hate it because really we should try to avoid this in our tests now that IW settings are randomized +hit them hard again +There is already a digest attribute being used to address +Thank you Sean +This fix addresses a significant functional bug in SASL authentication +Thanks James + +all +One update to my last thought +The fix of the bug was commited to by mistake +I have an explanation +question leads to new questions +Cancelling stale patch +ApologizesI it to look much closer to the route +This would also break backward compatibility at the package private level but would revert to a solution with essentially the same scaling perfornance challenges as the per instance Timer in the patch +This is the same as Uwes patch but with three additional teststesttesttestNot sure if its worth adding them but they do pass +Actually there is a dependency for WSSJ other than the SAML token validation +File used +I will proceed after releasing MINA +But I think there still is a race condition with markedXid +Recovery worker takes an underreplicated ledger fragment and rereplicates itEach bookie is running a recovery worker so the work of detection and rereplication will be distributed while the auditor for checking the bookies will be centralized +Maybe its not bad + +OK looks like my commits fixing the two tests were reverted so no wonder at and au fail now +class will not see anything which you defined in your ivysettings file +It might be necessary though to rethrow at least and in of the +Thanks. +I suggest the inclusion of this in the javadoc and the change of the default behavior inThanks +When you click on the it changes to none then the port is no longer in the combo box and cannot be entered +bulk close of all resolved issues as part of closing items. +Regressions passed +All youd have to do is check if stdin has anything on it read it if it does execute it as commands and then exit +Attached patch file that contains the changes in the projectSummary page +So if tracking per client is high overhead wouldnt that apply to the trees? What is the normal arit between clientstrees? Again we are talking about clients that store something not the ones that refer to the nodesPatrickApplicationService makes sense but how would we reliably know the app name and also deal efficiently with duplicates +Seriously a second question What about alternative languages? I mean maybe Java does have excess baggage related to its quirky semantics but could the raw JVM support a implementation of BQ without leaving the JVM +An alternative patch to fail the job if non +So long as each batch is flushed before commit there arent going to be any silver bullets that will result in huge performance boosts +Thanks JrgVery clever and clean +Anyone have an update on this? This appears to be one of the longest jira threads Ive seen +comma added +what part of I might be persuaded into having the different strategies as a configuration option but until then you can just write your own reveng strategy and get around it that way did you not understand +Working version of vv and folding in the compose method that was added to as part of which makes done as part of thisStill have to test writing which will be tonight probablyAlso Brandon had some ideas about optimization +Windows bit installer +It has the same patch +Dont know what is causing it +Thanks for the update on this Stack +We can add thse as they come +Can we close this issue to be included in +The problem was the correct way to orchestrate a region and compactions wasnt obvious just took a bit more poking around beyond the obvious methods +Thanks +Sorry if I am dense bit that means that you can run with normal JRE and no more scala stuff is needed on the system +The patch only modifies the sh files that havebinsh and makes itbinbashIve noticed that some of the +I like it because its quite visually distinct from the project logo but obviously sticks with the yellow elephant and cartoony theme +Should we call validate even if we use an anonymous user +Thanks +This was because Seam targeted which included Tomcat which is a servlet container +You need to grab the Capability from the wire to get the package name +Please report any issues you encounter +usrlocaldespotssbin only root can run it +Noting additional found by manual inspection could use a header comment as usual +Committed to both trunk and branch with some simple test fixes +Patch from from committed +They can be annotated with at the type level mixed and matched with pure annotated controllers and even get autodetected using the Controller stereotype so this should be pretty convenient as well +That would be nice +I know this is low priority but has there been any more thinking about how to address thisThanksGeorge +is fine. +Can you attach patch for trunk? TedI have already attached the patch +Thanks. +Fixed withttpviewvcrevviewrev +Marking as resolved since this appears to have fixed the issue and since we now have a test case for this condition +is this the right place for this files +Committed to the trunk At revision +Arun Pbintar is now Pdist Dtar +I just checked our internal branch and I did correctly there +I have updated the document that you are referring Did you able to get hold of the latest document Its uploaded to the server before the I resolved this issue +This patch enables the use of the property for dependencies +It appears that does not look there when you run tests from +Dprojectssrcjavaorgapacheddlutilsplatformpostgresql Sending Dprojectssrcjavaorgapacheddlutilsplatformpostgresql Transmitting file data +Move it back in if you think differently +Are these new knobs at the job level? I think this the wrong direction +Gregory spelling corrections verified at r +actually already working +I would like to get the failed tests down to zero could probably make Context immutable as part of a larger revision +see class unloading design discussions on. +the failover transport will recreate temp destinations but an activemq connection tracks a temp destination deleted advisory +While is synchronized unlike the JVM is smart enough to realize when a variable is only accessible from a single thread and optimize most of the locking penalty away for anything thats heavily accessed +actually I could but it would be a single threaded test +This is a clean and lightweight solution +so we have real package names? Then it would be real sweet! +workaround use ant changeset support to detect if you really need to run the code generator +I will run tests again + +The kb patch is the original which doesnt quite work +However the resources are installed in the repo along with the version resourcesThis worked differently yesterday +whoops v had an error see v instead +I like how the proxy server main method takes one option the name of a config fileBecause were relying on Hadoop Zooekeepers classpaths to provide many of the dependencies that we need to run the tserver master and init processes using accumulo mini will only work if the two are installed +Its strange because the patch didnt get applied in trunkBTW +Moving this back to Are you close with your patch? As discussed in this thread we should just replace Fetcher With Fetcher change Crawl class and check that the tests pass +Regenerated patch with the current trunk +Please verify and close this issueThanksColm. +and published a message to with Persistent Delivery +I didnt implement any support for write operations yetIf someone is still working on this Id be happy to join +rc release will get put out once we get updated +This is attempting to figure out when we do our scans if we are missing somethingJD and I originally thought that the row for the first region was still in META but for some reason scanner was missing it +Thanks for me it seems natural to compile I should suppose that the classes that arent compiling will never be used in runtime +Each chunk within the buffer would be checksummed immediately prior to returning any data from it +I have written quite a bit of code in Upgrade to handle this case wellLooks to me we might be subconsciously trying to fit to match what already does in +use a subset of the documents for example as a training class more normalization of the term vectors by max frequency in a document or by total frequency in a document +bq +For smaller keys it seems likesaving the position is slightly cheaper whereas for larger keys itappears to be slightly more expensive +Good point the new dispatcher originally did the same as the old streamed everything out to a file and then i switched to a pushback stream +I dont want to be responsible for it for all eternity though +I believe the notifications are now scheduled properly. +I would not deprecate for now +Thanks Prasad! Anything left to migrate wrt the module +Thats pluto specific deployment is app server specific and has been implemented by other plugins +any ideas on your side +Removed bookkeper doc files and bookkeeper references in and +Looks good thanks Tim +Had made error while printing out message which was caught by my testcase itself +Currently any source Procedures marked with FUNCTION TRUE are treated as sourcepushdown functions by Teiid correctAt one point we discussed the idea of any UDFs being defined as virtual Procedures marked with FUNCTION TRUE +It didnt seem necessary to me that I needed to calculate how far to jump call jump then be checked again +revisions to closing all issues that have been in resolved state for months or more without any feedback or update +what it doesnt say is if nothing fails but the column isnt found skip or pass through the entire row +Sane should give a line numberfor the problem in openmay help to provide a ls l of the seg and log directories + +One of the are +Changed notification address to dev build to port as the and builds do run at the same time sometimes +I meant to add the new functionality in the existing +Thanks Amar! +I wouldnt think so +Thanks to Jakob Korherr for this patch +Validation can prevent a transition causing the flow to the same view +Ill add them now +I havent checked what they are but I recommend we follow POSIX conventions +not so trivial for me +Please cross check and close the bug. +That it why it load so much +This is hibernate tools issue and I sent pull request yesterday to apply the attached patch +I believe this means that Eric intends to not release without this +Yep thatd be nice +Ill check this in +Currently its assumed that underlying loader is providing data in a way its expected +It could be beneficial to research other databases in that matter +I will determine when well get a fix for this +Fixed patch was committed +This looks great +Did you launch a new project analysis and clear your browser cache + is a similar problem caused with the services and concurrent sessions +Here is the use case Assume that in a table there are two column families as CFA and CFB +It also fails with expected value +Shawn perhaps you can provide an patch for your favorite OSIf I can figure out how sure +Thanks for the bug report and especially for the patch with test caseyour patch looks correct to me at first glance but ill review more closely and try to get in for +on Brunos patch +It makes sense to remove app from request +Denis I found a problem while using DTRACETRUE youre shown variables in part of the installation window and so you cant add any runtimes +Dag could you take a gander at this? I will run tests later tonight +Guys I have created a very basic patch for this issue +This could be a first step to get rid of the at all by first shrinking it to the absolute minimum +Applied patch at r +bq +Username kenyee +Issue possibly exposed by addition of +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The will automatically do that once the request has been sent so there is no leak here +So I simply reverted the change from the branch in revision +Attaching new versions of the two patches that I think addresses Dans first comment on the addition of permissions to +Workaround for secondary sorting with display tagIf the list of objects is sorted from the backend that sorting is retained while sorting on other of this workaround is that the secondary sort column is fixed. +I wonder if it would be possiblemake sense to make this a tokenizer instead of a charfilter one token per sentence + unscheduled issues to the next release +buildivylibcommongriddevbin +there is no in the hotdeploy folder +Has anyone studied the impact on the NN yetWo having reviewed the patch I initially dont think this could be any worse than ls R +It would make building a normalizer a bit easier +From your branch git pull rebase upstream should help do that +It seems the failure has to do with the binary property being multivalued +Please update the patch with better comment and I will commit it once tests added extra comments in the to explain the modifications +The design is treating jmx and ganglia as first class citizen +Thanks +The solution is that all of Gebs exception classes need to be rewritten in Java. +Closed all tickets +Importer screenshot attached. +Verified. +Andre I was able to manipulate with this account normally today no timeouts Yesterday I was WFH and connected via VPN to RH network could that cause the timeout problems +I committed thisThanks for the review Arun +I have no issues when running the app on Tomcat from within Eclipse +When a session is expired it will be evicted from the cache +I would encourage you to run a test case with similar data to what I have and try to determine whats happening as Im pretty sure youll find the same problem with a similar environment +The cost of the variable width types is really small with the vint encoding + +Bounced Sonar upgraded to including pluginsRebuilt problem build now works. +to trunk +Please try the latest version from trunk and close this issue +Fixed +Yeah +Eric says the github branch above was kind of hacky and it may or may not be useful +Google Analytics can now be entered through the Xircles on the project settings screen +Thanks BrianFilip can you confirm that the testing mentioned above is successful +Not sure how to get that to you though since youre using Maven +Not sure if these are the same issues or not +The fix is in for release. +Launcher cannot load VM DLL since it fails to find some symbols in it and therefore cannot link it properlyThis problem was discussed on the mail list and it was decided that too much has to be changed in VM to switch from modern XP vectored exceptions to older structured exceptions so most likely this wont be fixed especially since windows is quite old and isnt supported any more +Right now there are a number of them +Attached examples with modified Hello World example +OK low was removed from now we have only highestdelivered and highestreceived left and can therefore apply the optimizations described optimizations are going to be done in +Id like to work on this new feature +After lot of inspection over the spec and Hive grammar I decided to implement TIMESTAMP asa primitive type which will be stored and retrived as a string of characters +The attached patch fixes this issue and adds a bunch of specs for other unexpected r +Good catch Virag my bad here +Is there a specific reason why that was never implemented? The default regexp are good enough for me I would prefer not to use a rd party library for regular expressions +I created a image to show what happened +in +I then compiled with that component +Antonio Suh was joined to this project +If theres more to be done well open another issue +I did mvn test and passed on same platformAlso i noticed that for rest of column families which are not using supercolumns with TTL expiration live ratio reported is about nothing gets close to committed then thanks. +bq +shows the test case try to open mapping for Customer orgName as result is opened but orgName is not selected it should be selected +The only catch is deciding what to do with the connection if a remote checksum failure occurs after already starting to stream the content. +Attached you find a patch for review including the above change +This patch adds interrupter thead that waits until thread locks on condition waiting and then interrupts it +Hi ChristianI was planing to send a mail for the second unit test but it takes me too much time to merge the patch due to the network issueHere is my explanation for it because Camel is using the to delegate the life cycle invocation it is also in charge of do some clean up work with the +The element now provides a attribute which accepts a reference to a instance. +Fixed by r +Do you have a patch to fix this behavior +committed to trunk. +So I confirm that this is not fixed +However there are issues on the executors as Im not maxing up the IO caused by Paging +We can delete the manual steps handles this +skipSample and tradeJob were using the same job name so the second executed was closed as resolved and released +This file is the updated version of Cjdevrttrinidadsrcmainjavaorgapachemyfacestrinidadinternalrenderkit +I get asked to install the mortgages example repository +Hi Michal Sorry I cannot find slfj in only two jar I could findeartargetlibeartargetlibCould you please tell me the detail places where you declare the slfj as test scope + +it looks good +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +so there is a mismatch between our encoding of the class path with +Using templatepath solves the problem +The tasks should throw an exception when they try to create new files since their directory will have been removedOf course there is always windows +How are you using this cloned as another maven plugin? If you attach a simplified project that reproduces your problem then I can take a look +This is way better +Although I perfectly understand what you are asking for is reasonable I am not sure if this is what we want! Goal of this operation is to bind to all +And that is why I disagree in modifying the get code in which really does not belong there +Patch appliedboisvertsixtinesvn svn commit m Regression baseDir system property is not set when executing testsSending CHANGELOGSending libbuildrcoreSending specjavaTransmitting file dataCommitted revision The reason why the original patch didnt work was due to using instead of +Claudio Thanks for the suggestion I will further investigate the issue and provide an update when I know whats going on +Deleted the old patch attachment and added a new one +within a web framework like wicket or to pass it over to javascript or JSON you have to manually the value returned by the map method otherwise it may get doubleif the implementation of the API method is not changed at least the javadocs of the map method in the interface have to be updated to document that the paths returned are and they cannot be used as reverse operation for the resolve method +The Annotation should indded only used by Experts I will add such warnings +Would like a better fix +This test input shows the bug when run under JDK producing PDF output +Otherwise get will substitute the system principal +The problem was that we were looking in +LiangLatest patch didnt get through hadoop QAMind attaching patch again +But the main task is accomplished. +Committed to and branches. +Ideally the repository itself would just collect all the interesting information and a deployment could then decide whether to make that information available through JMXThe ways of making such information available should be flexible enough to allow for new bits to be exposed even without changing the +Temporarily reopening to update release note information. +Oh we should also include a unit test that verifies this behaviour +slip John let me know if this is not true +Sandy what do you mean by? Which clientMy bad +I think this one is better left for the Piccolo guys to deal with +Qualifiers arent compared there either +I had the same problem but it turned out that it was operator errorI was trying to run releaseperform before running releaseprepareHope this helps. +Fixed long ago +Attaching the database after a failing run with true the from a failing run +Patch compiles +What caused the issue +I will leave this issue as open as I need to revisit the tests which are dependant on a running server these are functional tests and will come into play once refactoring is nearing to a completion for the first phase and we have set up scripts for the server which also deploy the esb and run the functional this issue be downgraded +Why not create your own annotation and validator class for your projectOtherwise the validator project will get packed with custom validators +I think Len is pointing out that the examples use real names and therefore might generate some comments from some supporter of the above +I also like the idea of moving the grouping code into a configurable ranking module possibly as a third ticketThere are a couple of other tickets that Im close to finishing up and then I can revisit this one +The group id is as set in the sync script as of rev closing the issue now +Ive requested translations from David and Luca via email since it appears they dont follow this issue tracker +But maybe this can be done in a fast manner +Its not listed in the jobs though +No there is no chance of manual intervention +Thanks Chris +My certificate certificate setup is working correctly +Lars Is this what Dave Revell was going to add at one point till we discouraged him by telling him use a filter +Ive committed this to trunk and +Closing with the release. +which is no different to an array with a sigle null value +Closed after release. +Justin you expressed some interest in this so Im passing it on to you just trying to clean up the unscheduled stuff feel free to move it back as you want +I implemented cmdarray logic for popen and friends some time before RC so I believe this is fixed as Roger Pack points out +Again I am not able to get a combination which works in all build cases +Fixed +Bumping up the severity +I have the same problem in FF when scrolling horizon offset the tooltip far away from the original cell +I have upgraded my It is great for dont think you have to refactor the advisor + +there may not be time to do this in the +Its now these threeReopening please check +Language based ordering is being added by of link to issue for enabling national characters as collation is independent of national character types. +bq +I think that we should life the character restriction in +there you got a point +might be better to get it review so that we can capture issues earlier +Work to date pushed in aea +Help in resolution is appreciated. +done for subTable +After looking into this for a bit then the JDK does not offer good for being able to do custom logic when rejected by which you get access to the inner details of the runnable task you submitted to the thread pool +Im closing this as all the connection factory properties are available and transaction batching is something that needs implemented in the ejb dont close tasks as done if theres nothing being done. +Sure Mike definitely interested +Looks good +Hadoop was just released. +Thanks BryanCommitted revision . +It is not easy to see the other hosts that are available +Reviewed the content of Smooks doc sourcesdeleted all the old date screens and +the same patch +I will go ahead and commit these changes shortly +Appears to have been fixed at some point before r +Last patch containing set method forI have intentionally avoided the full set name +Reopening to change the Fix Version to future since tooling patches will not go into the BRMS release +on your +Unit test code and output from a run with messages +If you beleive that should be please and contact the program team +I tested this and it doesnt downgrade the bundleI am testing with JCR Installer version the can you please test the latest trunk? Im pretty sure I fixed this problem some time ll have a that this is bug +committed +Thanks +Done +I still dont see what harm that dependency causes aesthetically or otherwise +see +Deferring yet again +If SCA CPP is currently then the samples instance docs should be modified to be valid with the schemas in Tuscany source control +The patch looks good to me +On recovery does it give correct length of the file +Committed to trunk. + +Portlet test forward +The tests trigger the in and return the correct values with the patch applied +This seems to be an issue with the SOA platform there is a jUDDI configuration in +Hello AntonioThat sounds good +Patch will not clear the blocks in this caseSo I checked the rename case +In revision I disabled the check for protected items +Running the tests shows the same no of failures as before applying the patch so assuming it doesnt break anything +Remember to set Components in future +Committed the patch to trunk +Committed to trunk +Set usecapturelobs on symtrigger +fixed CLI overrides of POM properties and added another integration test for this case. +My bad +we would just be guessing and would probably end up with a skewed list we please just get the artifactspoms published to central as they are attached in the bundles +making it even tougher for most people to change policyGiven the above should the system record the policy in the fsImage to prevent it from being changed? Similarly should the balancer check to see if it has the same policy as the NNThis can be done +Or we add some upgrade code which checks the plugin property for the old value and replaces it with the new ve applied the after release +So does that mean that it would still affect regular +However in MPING discovery messages bypass transport protocol and are sent directly to network interface with appropriate mcast address and port thus causing interference +You are right +Ok please raise another jira for this improvementIll give a look for the next version. +I just checked out a fresh federated build and was able to sucessfully run so as far as I am concerned this bug is fixed +I suspect the new updatedelete functionality would make it much easier to add support for synching of Lucene indexes using database triggers +This would happen whether the user did an undeploy or a remove +I would be equally happy if cqlsh can provide me the information on how to format hector and thrift queries to throw data into this column family +It might be possible to have a base class do most of the work and the two subclasses simply define the properties and the mechanism for creating the +This issue is fixed and released as part of release +Next to Availabilty +and many thanks to Mark for working through so many iterationsas we explored the different approaches here +Looks good +Release is completed. +made some modifications to notification handling for annotations +Back to you Max I outlined the issues remaining I had to you in an email just a couple of classpath problems with drools. +What is the problem here +Craig this is very arguable +out files could not find any +Unscheduling due to inactivity +Fred I think the URL you have is a timebombed URL available only to and Mark N are working on a direct download URL +The library will now be deployed in the lib folder automatically at deploy time +I agree that its just those counters that are getting set by this code path +Specify to identify the object using only the objectID so it looks like it would be possible to verify whether the owner is recycle bin +See for further details. +What about the validation of the keys that user has provided in the xml for the keyRight now as per the code it will add an entry with that key and value as null in the parameters Map +Ok to WFK +Fixed in revision of buildersjbpm resourceschematoolesb message Uncomment the identity mappings at build time and do not change them in the schema tool +Minor editorial comments on the doc changes I think it is better to say Users can instead of User can In the word different is misspelt +I realize that this is a HADOOP jira whereas it probably should be HDFS +Yes it is still problem in trunk +Do you want to tackle the issue? Or maybe provide a test case +was just create to get the discussion goingThe design we are proposing is to let send the length +thanks for the patch Viet +Thanks! +The logic for filling the combo is put to method initialize and probably need to be improved +Can you create a request +I think these additional version management functionalities are very crucial especially the ability to restore a particular version +I dont know if this can be avoided but just removing the repeat and expecting the subclass relationship to work out the details did not seem to work reliably +Let me do that +I therefore kept us requiringFurthermore I removed the upper bounds that were excluding since we have Juno ahead +Rev Fix for AS Thread local pool is now backed by infinite pool to handle the life cycle of beans +Fixed the nit and committed +No reasons to keep this open anymore +Any information on how to enable this special Maven plugin in Maven? + needs updating +I figure I will make the reference count incdec in the methods and then if the reference count ever goes to we update a timestamp to indicate when it was last used so we can purge oldest objects first +Make sure to put an author tag into the files youve changed or added. +Thank YouIm new using this technology and its nice to see the problem Im facing is solvedCould anyone tell me where can I get the jars files with the solution implemented? one note myfaces jira should not be used to discuss this problems +Closing as cannot reproduce +We should fix it in too +Current version we use is which in general comes out of the box +Please find attached the wrapper log +Is there any way you could post a small sample and bean class with annotations +Make sure that if theres any error the directory is moved to an +attaching clean patchThe is failing maybe somebody can take a look +I would prefere this tests to go into CTF Testsuite rather than examplesAgreed +Changing mixin types must perform ofall properties even though when they are not modified etc +Dont you think it would be better to use one of the classes like? This way you can have listeners attach and detach without explicit synchronization +That would be fine with me +example use of Sandro this issue has been already fixed +This is however a message that james did not like for the version I reported +You will find that line just after the comment Regression test for +Yeah but see my comments on the issue +HeyIm starting to get concerned by the hacks upon hacks upon hacks +Bulk move issues to and m also interested in this +Moving to +Should be permanently fixed with commit adebbaaacbfebdae +Do the java classes need to be compiled as Java source and target? It seems like we could just compile them as compliant as wellIf they do need to be Java compliant or thats the desired outcome should we just move these few classes back to the archive module? They seem to be fairly minimal adapters for the Packer and Unpacker interfaces +It may be less scalable to keep all the logic in a single service but I have yet to be convinced that the jobtracker is a scalability bottleneck +Heres a regenerated version. +Do you have an implementation in mindalready started +The bug can be fixed even if the stripping behavior is retained as it is + +Is it fixed in but not in +Looks like indentation issue +I just realized Im assigned to this dont know what else i can do here +patch to current LeonYour slightly modified patch is in trunk at revision +Hey JasonMy tests are inconclusive on the patch posted above +I had a similar issue and quickly solved this by giving using the right parameter for the avg function +To take Patrices example right now we emulate this behaviour by specifying the listener B as a parameter of the listener A constructor +Applied the minimalsimplest workaround in c on master branch +Patch checked in +This is bizarre +Implemented via SPI. +Rob C +Frankly hbase already has this covered +Needless to say applying this style to a layer doesnt work +On my machine though we expect the container to be long running it crashes +Depending on their release schedules they may not be in a position to release an updated version quickly +Lots of frustration so far but I like the result +The fix reintroduced the issue where the log messages are displayed to even after on logtail +from me +With this patch parsing occurs successfully in a latest build +committed on branch on branch to trunk for doing this Elliot +Please update the Affects field and leave the field as Unknown +Can you please check into this? Thanks +I dont see an way to support this feature without rewriting the entire supports mechanism and breaking backwards easiest way for you to solve this would be to subclass and override the supports method to do what you want to +Another possibly related hang +This appears to have a fix and could be resolved in timeframe +Tested +Wouldnt we want both to work the same way +I missed that detail +Committed in r to the JBOSS branch +I already d thisPuts and Deletes are timestamped +For such a simple change I doubt tests are required +Yeah thats what I have going in the changeset Im working on now but it would also be nice to see that the monitor is making the right thrift client +we need to validate index targets one way or the othershouldnt this be done by calling +Will commit to trunk and. +Youre in the correct is the problem you are seeing +The exception suggests that the header returned by the server is incorrect +collapsing is now disabled +AFAICS the usage of the maps in the slots array of the are properly protected by the writerI will additionally protect the maps with a synchronized block while looping over the key set +Claudio youre issue has to do with Java Sets and hash codes +We might have to do a mvn install before running find bugs +It seems that it should have been this way all along +Therefore will be closing the issue +This patch moves and related classes from hdfs util to hadoop util +From the above discussions we seem to agree on limiting this to the ntfolder slingFolder and sling node types +SVN Revision Added support to wait for the to startstop +Date Sender malkimiLogged In YES useridI have the same environment plus and found two issues that cause these problems +My +Same crash with +Let me take a look and see if there is a fundamental reason for this or it is just a bad test +We want to make a friendly home for cool new work but can also accommodate downstream packagers who dont want any kind of support implied +Note that in any sane DBMS a unique key is implemented with an index in any case + correct Ill take a look at these this weekend +now includes findbugs exclusion +Resolving this issue as I cannot reproduce the failure +Thanks for the finding this Suresh +srcmainresourceswebapps static jquery +To avoid possible duplication of code perhaps add a boolean to the method signature for whether to verify the token? Or maybe pass in the NN object or null to indicate whether to verify the token +it syncs with the codehaus repo every hours or so. +All right +We just dont have the capacity to refresh everything right now +But so far I cant access the value stack from the action so for now the solution would be finethe patch you submitted +Because they are systematically managed by IDE +I think we need to address to fix this in my opinion flaw in the design from a Maven perspectiveThe good thing working with these archetypes is that we see what doesnt work well +Duplicate of duplicate. +As I read the enumeration doesnt have to reflect the contents of a queue because of message arrival expiration or consumption +I am in constant touch with carlo and have finished all the major portions of this +Closed +Yes good point +Thanks Mike for catching and reporting this issue +Moving to has been deprecated in favor of Avro in Hadoop. +Fixed and unit tests added to test it +Is it failing consistently or only some times? In any case error looks bad could u please zip the this test directory and post it into the Jira if possible it might give some clues to find the root cause of the error +Hopefully the meta information of an application will not be big and will not terribly affect the IO performanceIn addition we can do application level cache to avoid accessing the secondary storage system all the time +I feel we should consider removing support for paths using as separatorI think this is a good approach but it will break things on Windows and we should file blockers for these regressions or fix them here +However we can use the existing approach by which JT persists the competed and dead jobs to persist the jobs in JT +Will submit a patch with tests a little later +I can reproduce this on the latest +No luck +no problem remove away +just committed this to trunk +If that happens the cluster is without loss of data +Im going to comment them so future developers wont think the typesize mismatch is a mistake then commit to and trunk +Was there ever any documentation written on how to access the EXIF data from the JS side +Havent heard back from the original reporter and not able to reproduce here so closing +This patch should be against and trunk since is going to be against +Id love to see this one resolved +fixes minor bug in my earlier patch with respect to configuration settings and calling getjobcounters without job being initialized +Doesnt the retry policy mechanism let you do this with things like timed intervals exponential backoff etc +cant. +A diff showing changes in with the version number and test changes on EAP ER +Could you please review and applyThanks comments from me patch applied. +Applied below patch adding two project files insrctoolsideeclipse and adding a new Ant target to copythese files to the root directory of the will have to restart Eclipse to benefit from these project configurationfiles +Sorry Jeff the procedures dictate that we must have a product branch we arent allowed to do productisation from trunk +aeac is displayed in log now +hmm prefetched messages are eagerly dispatched messages +I think I had left the setter in from when I was doing some testing since it wasnt actually being called while I was testing it this morning. +Is there any other option here currently? Looking to the future may have more capabilities at optimising if it understands the distinction between mutating wrappers and ones and how they are composed +Thanks for the test case +This patch substantially changes the API of the gamma function to make it compatible with the wikipedia definitions for the gamma function +Whatever +Thanks Bikas +As progress reported for all the pipes task would be +Thanks to your comment +That will greatly simplify the user code +If the user says apply then and then should we not preserve that orderingI also take my words back on the issue existing in the current code +And it helps a lot to reduce the cpu used for mem sort +Used as an additional check beyond the new tests to verify the issue and effectiveness of the changes being made +JasonURLs are nice and all but theyre fairly unstable +The element name output should be a hint at its realm of influence single files +Let me do some profiling here +Will follow up there on. +I fixed that the path will be matched but this is still not connected to the properties system +bulk defer minor issues to +This doc is finalized we are going to be releasing tomorrow +only init flow scome cases we cant distinguish meta region server +It may also have the changes for passing the handler nameThe should take the Context as a param and the does not need to make a copy of the attributes +With GS logging in geotools developer I found the solution +was created on +The seems to be more stable it is now OK or we need to be better? Isstill there a leak +Please close this issue if you are satisfied +Philip Your comment is not very clear +Moreover according to change in r in case if property not specified we will print null which is IMO not +This bug was found at binding at +I dont think the is the correct place to implement positioning +I took your starting point and filled it out a bit +It will help to batch more puts for each region server before sending out the rpc request +No there is nothing to change for other part of the command +probably wont do this soon as it makes stuff unnecessary complex +Heres a patch that allows commitet and deployed +Can you please switch develop to be the default thereThanks everything has been reset back to master can and have develop as their default branch +Ive committed this to. +Please update the patch and make sure Object works too. +Applied remaining changes from patch in rDid not touch the simplifications for equals methods as this may collide with findbugs checkstyle as well as some string concatenations. +for urgent issuerelated to my jobemail eugeneyumobile no RegardsHao +I didnt do that in my patch person or edit and so if the prefix notation helps please raise that issue. +Now the version is incremented +Still havent found a solution for creating a valid patch +Just in case anyone else needs a fix quick. +We will try and resolve it by tomorrow + server is not actively maintained. +As a note there is no way for me to recognize when a runtime selection is changed which means my Error message stays up and present until someone the proper deploy and folders +The branch should be used as the trunk and the tag should be treated as immutable +Copyright? Red Hat in header in IE strange symbol presents +as Werner said it must be using XKEYIDENTIFIER instead of SKIKEYIDENTIFIER +I kept the code for it just never calls it but want to have that as a seperate issue and possibly only do it in very specific cases +which is not a good solution +What are we going to do with this? Mark as wont fix? I also prefer regex as the solution +Im afraid that do something similar behind his back might create confusion to some users +Resolving since modern versions do check per Marcels investigation. +no its fine Ive added the missing revisions from the old SVN Thanks +trunk The scheduler will now keep the jobs running if in clustered mode so they can keep running on the other nodes +Thanks +These findbug issues are not related to this patchAfter applying the patch nodes page displays as shown in the attached screenshot +Thanks +Committed the original by Hoss +suggested change in +I updated the description because its not only with sun jdkI committed the patch with revision +I tried a minimal impact code change all the referencing module unit tests passed so I hope youll find it useful +committed +The surefire report attached by Udai shows exactly what I have seen here at home at the weekend +I will backport later after +Ive redeployed the site using the latestWait for the sync until the footer is correct. +Verified that it is OK now. +Clicking does not do anything until you configure at least one clerk +I figured out what is going onThe problem is that you use but it doesnt contain facelets tag lib for f and h +Since the patch is getting big can you put it on RB +I seem to be making a general mess of this JIRA entry +And I was able to successfully run the ll consider committing this +Implementation of service quick as per details mentioned Moved ecommerce map processors in party component and also changed references at other place +Lets start by implementing +Hi Cheolsooplease accept my deep apology for keeping this issue withnout notice for such long time +The task is too old and now the overview tab is no longer included in the Smooks no changes are necessary in the Smooks guideInvesigated Split the Smooks configuration editor into more tabs +All subtasks have been done but one item which has been judged a issue +testcase for isolating the problem +Simple enough pb change +Committed to and trunk +Maybe + is based on which is not in yet +Then maybe NO WARRANTIES ARE PROVIDED +changes applied +Nice digging N +Want to attempt a fix +Thanks Nick +Jacques could you let us know if it works? Thanks +Sounds great Ill assign this back to you then +As no memory leak occurred I assume that this leak is reallyrelated to the existing and not to the DOM implementation +Patch committed +I have a few small changes Id like to see but Ill open subtasks for them +Fix has been applied to branch remotingGA +My previous comment was mangled so I rather try to attach a patch +Removed from thetcunninglocalhostproduct svn commitSending productrosettawarresourcesTransmitting file dataCommitted revision +Ive been experiencing this bug with Portlet Bridge beta RF but it seems gone with beta RF tried the solution given by Dirk Pitt before I worte my previsous comment +Hi VincentCould you submit a simple nunit test case showing this problem +Thanks for fixing the previous missed clean up as well +I think we can move the snapshot information to the developer sectionBut now I just cannot update the wiki page as I keep hitting Non Clustered Confluence error +Is it better now +It has never really changed from release to release so version information is not really relevant +optimized build could be optimiation was merged to version +to trunk +Moving to +for the patch +Opened to follow up. +Regenerated patch based on r +Committed revision Thank you Jan van der Lugt for verifying the fix. +There must be a bug in the layoutmanager that belongs to select +It comes with and active hit collection cache et c +You may have some idea of how hard that isI dont! I dont know what the licensing issues are for such a plugin so I cant say whether this work is even possible for our Apache community +In java the connection interface changed the patch attached is for but I also have one for trunk that implements the methods required for java +I had to apply several other patches in order to get it in sync with current version of DFSIO +Do you have access to such an installation and if so could try to email me the dll? It will have to go to +Also this is unlikely to be done in although this probably needs more tasks added to it +Ok since three subtasks are done this issue can be considered as FIXED. +Its essentially sg in the file +The fix for the other broken bits are in with new code is also significantly faster +Patch looks good to me +If the client does not have a writer or if the client has a write but datanodes in the pipelien hang this patch does not solve the problem +For example the value would be changed from Mevenide to for the feature +It may be a good idea to move the whole logic to a separate class so we can make separate out the logic and make it cleaner to read +Its invalid +Help in resolution is appreciated. +To fix this perhaps the namenode should refuse to represent more than one copy of a block from a given IP? If a second is reported the first should be forgotten +I verified with +Converted demux to use flag to enable libjarsThis change enable the ability to pass in libjars for standalone demux and demux running through demux managerPlace the parser jar file in CHUKWAHOMElibdemuxYahoo is shutdown over this weekend hence I cant test this code on my cluster +Agree command would be great +My apologies for the confusionReopening this +Please have a look into this +Patch committed to trunk +So here is another fix to The seetting of vistied dir needs to be added a to make the problem disapperar. +Did you try the latest available version? check the latest version if the problem remains please attache some test case with more details +Could you add the data needed to reproduce this problem +Ive tested with and the samples block can display all the image samples with the new implementation. +PATCH NAME NAME Application DESCRIPTION conflicts with the use of commons logging at the web container and web app level need to be DESCRIPTION conflicts with the use of commons logging at the web container and web app level require the web container to preload the tomcat classes using commons logging and an update of these classes to use static references to the commons logging Log so that logging can be initialized during the tomcat deployer start +a panel with is used twice in a page +The corresponding ESB issue here is is resolved in the CR build but that has yet to correct SOA issue is already linked to this one. +Im testing my build now +I dont have TRACE enabled logs for +Improved version now uses stuff rather than having tests mounted in missing license headers thanks Jacques +if we end up committing then this is no longer needed we can resolve this as wont fix +Thanks very much JarekI will try to find whether there is another way to avoid the infinite loop. +not an infra issue +It was an oversight that mkdir did since the p flag was there +Patch was same as original but rebased with trunk +Heres an updated version with a few changes +is not integrated yet either +I just committed this +Closing released issues. +BOOLEAN is still not allowed as an argument or column type or as the target type in an explicit CASTM javaengineorgapachederbyimplsqlcompileBOOLEAN is added to the list of JDBC types which have both primitive and wrapper Java correspondencesM javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangSome unit tests verifying the correct resolution of BOOLEAN valued functions +However how do we know the config initialization has finished so that we can report unused properties +If the compiler hypothesis is correct the speedup could be for map task and the speedup for whole job should be larger correspondingly. +Im considering Map growing as an incomplete feature in recommending an upgrade to Spring for proper consumption of that feature +I replied on IRC but guess you missed it +I changed my mind +Lets make a new issue to make these public in +Just needs to be finishedI can try to work on this over the next couple days +Merged into master. +Hi HarshThere are actually a bunch of changes in srcmainpackagestemplatesconf too +However if the first and second invocations are for the same id then only the first message would create the process instance and the second would be correlated to the same process instance but that process instance would be expecting a goodbye message so would not process the second message. +makes no sense to me but the damage is done shipped with thatJustin opinions +They will be different re right +This test does not run with cargo plugin +Did you write these descriptions yourself? If not and they are copied from microsoft documentation can you point to some licensing information that says we can copy the text without attiribution? Generally text such as suns javadoc for javaee spec interfaces is very much under sun copyright and we cant copy it into other spec implementations without some additional license I dont hear back soon on this Ill probably remove or reword the descriptions +for normalisation etc +You would have to change your config to accommodate only MB of heap + for the to trunk +Thanks Doug for looking at this +I seeThanks Sandy +Thanks for the contributionGood to hear thanksCheersBobby +Iwill see if this applies to as well + depends on this +Hey Senthil I have a patch for this mind if I take this over +Typically for a backed byt such backing youd have to be extra careful when using it to never alias a Column returned during iteration because that alias would be invalidated by the next step of the iteration itself +Patch that fixes the problem +verified by Svetlana +A prototype of the JAXB in GEP the sections of web deployment plan editor can work fine +Added unit tests at Apache and syncing to fuse and branches +I have removed the use of tag from all the components so I close this issueMarco +Yes but how do you fix it? I can certainly add a more descriptive error message but if getDepends returns null something was seriously broken in the parsing of the xml files +Damn +patch applied see revision thanks for contributing! +Next stop Jason +Ill mark this as incompatible +Both issues were fixed in +That private variable is used to copy the file to the local repository during install but again it is a file we cant sign it AFAICT +if that works for you please resolve the Heiko this looks better +Proposed for the patchWould it be possible with an unit test as well? Or can you point to an existing in that would test your patch +Attaching tests for and a first test revision Will be extended +Thanks that looks like a viable fix Ill double check and check it in tomorrow for you +Now all messages that come out pool edits are shown in typical ESME fashion. +on latest patch +I would also like to see a generated serial but L or any constant is wrong it needs to depend on the class signature +Good question +By the way has library in lib as JCR depends on it +The above results are not very consistent either sort on nodes was faster with the patch while randomwriter was slower with the patch both of which show little correlation with the and node benchmarks + should report the expected node name rather than its storage id +Hmmm it is complicating this issue by considering the stuff +good job +Is it just local file system? Or is it something more specific +How about we get rid of the old API of using the regionName and fixed on using only encoded name +Potential patch for ve solved the problem by adding the relevant classes to the source tree and still have framework as the dependencyAdded first implementation. +Im moderator of dev list and I have double checked if there was Continuum report on build +Attached a new patch that takes another approach +Activated again umlauts wrong again +Test demonstrating error +SAML Token processing is done by Username Token processing is done by Fix version to +Nice! was just planning to look at it today and its already done! Thanks for fixing this Teddy +This is confusing me +Try to vary a number of rows in the model rerender the table and try to scroll it +It makes it a bit easier to track down issues if preconditions are verified at the start of a test +Is that advice from a lawyer Max or other expert source +Lets postpone it again +Claus is this still a problem using a shipping ESB build +It may not include compression so may be better +Resolved with SVN . +But please ask the guys their version and Ill update the demo with any additional usefull parts of their example +If the point in time metrics are available from JMX we should use it insteadLook at all the host component metrics and make sure that they are obtained from the right source +Fixed in works for me +Let me know if you have any problem running these + +Sending srcmainjavaorgapacheservicemixbeanSending srctestjavaorgapacheservicemixbeanAdding srctestjavaorgapacheservicemixbeanAdding srctestjavaorgapacheservicemixbeanbeansAdding srctestresourcesTransmitting file dataCommitted revision +Before that I dont think we relied on the remote address of a socket being the address specified in the config +Tooltips seem to be pretty self explanatory +Now that work is well underway with AS all previous community releases are +and the private Coordinator class is unnecessary in that case +Please attach a quickstart that reproduces it and we will reopen the issue. +So we cannot package the war +to commit + uses so this will be necessary to support all of the file formats that uses +Many of the issues you are creating are really questions and are not yet completely formed enough to justify a JIRA issue +When using this with +Sure but users should get the maximum security out of the box. +when the above requirements are met +strange +I think itd be nice to have a patch that just made sure that anything that started with +I saw it saying tuesday result and assumed it was yesterdaySo at this point just have tinderbox nightly result which passed and no other nightly resultslooks like i missed the window for the sun runs +Thanks Harsh and Todd +I dont know much about JSON but it is my understanding that it is all generated from the so Im not sure what the solution would be +Bulk of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +I have applied this patch and cleanly compiled against Mustang build +Please let me know if something wrong this is my first commits inThanks MarcoHave you looked at the ercetly updated ebay store do you think we still to have them bothThanks +Ill reopen it if the bug appears again. +If you give me specific details on the tomcat version I can try that one out too +What are the observations? These experiments may lead to expected or unexpected results and would be a great addition to the design pros and cons +Committed at subversion revision +Sorry +It actually affects more than just the dfshealth page so this should provide much improved performance +Ah yes I thought that the comment is going to be linked to the patch which is attached the patch did not apply clealy so its hard to say if that just messed up something in the lione numbers but I am still not able to match them up with what you said +bulk defer of issues to +Therefore I have to reschedule it for +Hi ToddI have just tried to reproduce this today using your test program +I agree its a slight wrinkle but I insist that when searching for objects you should use queries. +Hopefully we are done +Yes +So I prefer to not add this feature until IDEA has been the jdkLevel for the plugin is a good idea +I did also remove the security test because it didnt do anything the other test did not you very much for your help! about output if you scan the code you will see that getOutput is not used anywhere +Supports two modes Start daemon with an existing rpc server for daemons embedded into Namenode process +Thanks again! +Maybe could be more explicit about setting values to but then if I dont bzero it we could miss initializing something else later +And they go always to the class directory of the jar fileI dont know if this is eclipse exclusive but it should work well even under this IDE +Closing the JIRA then +There is a patch for which hasnt had any TLC in a whileIts been associated with which looks like a major change that could take a while to get inId like to see if we can get movement on as it would help with some of the scope madness Ive had to dealwith and resolve the issue once and for all +Even at jvm process termination the locks on the files are not released and so the files are left behind +This patch appears to contain code from the C Boost library +check for has nothing to do with pstree +Thank you Erik +Thats a good idea and I think well proceed with this Ive already moved the timeanddate example from the international module into the general examples because of the same issue +It sounds to me this issue is more general in nature +Works for meThank you Alexey. +Implement applied +If it helps on runs the version of the exampleThe daemon does nothing but answering with the example response I posted in the issue +If you have time Martin would you be willing to submit a patch to for the site with the appropriate documentation? Otherwise someone will need to pick it up when they have free time +bq +However Im personally in favor of keeping the internal commands simple and making more advanced commands as closures or as separate bundlesIf Gogo closures arent powerful enough to make such commands as a closure then perhaps the real issue here is if we should make it so +All worked as expected with no template changes at all +Its better to fix this in next release because to fix it we should change class and its class used in many other templates of vpe not only for html tag img +yes Gemini Blueprint is incompatible with Spring +Marked it as resolved +defer all X issues to +I assume there are temp files that need to be cleaned up when testing withIt might be worth opening a separate Jira for this. +There is nothing blocker here +That is what it should look like after getting changes on the we change this to use an HTTP POST body instead of request param +CarloBill know about it +The add is synchronized so if we switch to it directlythe cost will be huge +Toggle menu not available for VPE in +The changes look good will merge if the tests +xml files and now it works. +Fix for typos attached +For now I think their best workaround is to simply select the error view for rendering executing whatever model preparation is needed before returning the +menu action group into the existing Run menu Restricts the Hibernate Code Generation +First cut +Memory reference think that the bug is in +Ill check out the logj docs next week to see how to do this +Maybe he could contact Nathan or Will to fix missing artifacts? +But more importantlyb revision stemming should do the second merge for us +Yeah this issue is now resolved through and has only been backported to the branch where the problem started after the fix of. +Thank you! +Thanks Mark looks like its the same issue +Closing issue. +Resolved. +Need to check into that +This patch changes to be false when a disk error is detected while receiving a block +It contains WSDL and WSDD for this problem. +So we may be a bit to conservative from that point of viewThat said wed identified other dependencies were we can relax this constraint a bit and well discuss releasing an early supporting slfj should upgrade Camel to use the slfj API as the other Apache products will migrate to that version SMX AMQ Karaf etc +If user wants to restrict they will need to modify the configuration which will then enable the feature +If you disable the Spring plugin then itll work fine +Are planning to take my job? Release train has started with yet again in efbefdbbcecfbeffa thanks for the patch +I know EM asI like the idea of general EM framework +It looks as though the thread is blocked waiting for a response from the broker which can happen when the failure occurs +Everton thanks for the test case +extthriftprotocol +good point ill take a look and the places where instanceof checks might detect null values and replace with null checking +I suggest you close this bug strip down your code toa small example and post it onto the list +The project does provide a wicket component so it must be a legitimate part of the release +does the server actually detect that we touched the descriptor? It should restart the application +Applied to Branchx and trunk thanks. + +Thanks +makes the sstable path absolute which should prevent this situation entirely +noi do not compare this with asynchbasebut this feature is simple compare with asynchbase +Error occurs when reloading all tabs. +Added support to get extract the image information for each compute node +Itd be better to have mechanism instead of sleep +Checked in to trunk with thanks to Ton Swieb for the patch +manual task are not acceptable this needs to be to be please provide a GPD maven artifact that can be included in the distribution this is available please assign back to me +Actually the problem seems to be specific to injecting into views +Test is passing now. +Thanks for the patch Todd +The fix would be for Hibernate to generate ANSIISO Standard syntax when the datastore is DerbyI have lowered the urgency of this issue +It sounds about right +Might add thatIf you store the head of the log you dont store the failure that doesnt sound right +I want to be throwing SQL at my Pig relations to verify them like crazyThis sounds like a new use case +Very good analysis thanks +This issue has been corrected and the fix will appear in the next publication of the Admin Console User Guide +I am committing a fix that will go into the next snapshot build that adds this new Long list length setting to the Preferences Groovy Formatter page +Changed to use a method +This patch contains the fix for the Fault serialization and also the unit test for the fault messages +I had it fixed before I noticed you had provided a patchAaron +I have addressed your comments in which Im attaching and which I committed at subversion revision +Yeah but I cant edit it +Attach with doc improved +So if the path exceeds a limit the first limit characters are replacedBut then a path of lenght limit would still exceed the limit + +When raising an issue for a specific user it helps considerably if you include the name of the user +It have done some minimal testing passing no parameters passing r and passing other parameters to ensure they get passed on toIf we want to go with an option like r these my two patches should cover and to The fix is present in +Attaching the patch with the Map solution +m asking for comments on this patch + +Also this doesnt warrant a new constructor in for seeking +team havent reacted against it as far as I know but I also is not sure they realize what it means Thus its important we get a build out that gives the various parties a chance to try out the experience as users will see it when installingstarting JBDSNick will the JBDS integration builds point to the staging sites +Closed with released. +I was able to reproduce it. +Simple junit test +If you existing endpoints and provide file name as a header then that works fine as it can existing producer from the producer pool +Be sure to consider +Thanks again Florian +Doesnt yet hide typing but you could pipe your username andor PW to it as looks like and may be talking about the same problem +maybe fmt is not used so dont force it +Patch that fixes the issue for me on single node secure setup which is now same as setup from the POV of security of the communication channels +Can you please look into this as well +patch to trunk and branches. +I need to install FC on one of my servers so this will take a day or two +I will try to find out the server vendor so we can file a bug there to get this resolved +Thanks for the patch JesseJesse any chance of your adding or porting a test back to with a prescription in reference guide or as a release note on how to run unit tests on only for the likes of the dummies like myself and MattThis is great +Thanks Kim. +There must be something wrong as it seems to work +bulk close of all resolved issues as part of closing items. +for the latest patch +The only remaining warning is for aopalliance which doesnt have a license because it is public domain +Replaced files after making minor interface changes. +The server log +bq +The buttons shouldnt be enabled since it causes confusion and that can be fixed +I like the idea of doing nativeascii in you know if there will be any negative side effects of the change to? What was that attempting to addressIs there anything else necessary to get this issue resolved other than the above patches and the nativeasciiing +It seems to me that the real fix we need is to turn the assert in into an if statement so it cant be turned off +Attachment fixes two issues +This should be a blocker +patched applied and tested with and +My general thought on this is we commit +Presumably this complaint could also be an interpretation of the fact that the regionating index is cached between requests even if the feature limit changes +Thanks Alan for catching this +I report the same behaviorIf there are on the JSP page for example two hinputText components from which the first is requiredtrue and one hcommendButton with action leading to the same page then +In flushcache comes before compactionRequested +Epoch timestamps have no concept of timezone they are all according to GMT +Looks good +My assumption was that this was a duplicate of you provide a repro case for this? Ill look into it. +I noticed that some runs wouldnt take long depending on the order of the bundles when calculating uses constraints which changed since they were pulled out of a map +Should the new methods be added toIn the case you described above should we allow the path for ver of the jar to be purged from the new classloader map +I made this snippet in one of my projects because I needed it and since it seemed basic enough that it might belong in the core I just wanted to share it for others who might need it also. +If they need the complete then can do the next query of the specific job +I do not believe it is possible to completely generate conversion code but appear to assume it is. +passes +Look for group by orgid +Im just not on the ball today +Florain sorry Im probably being stupid but I dont understand the point youre trying to makePaul what happens to someone who understands the danger of not filtering raw user input and has taken care to do so +Removed in order not to confuse anybody +After upgrading to CR it works +r +Please check that it was applied as you expected +Bulk close after is released +Related to closing the gzip input and output streams on the requests and responses if compression is being used. +Thanks for the review Suresh! Update the patch to address your comments +Since the binaries arent published to Jenkins CI plugin repository I would consider the project invisible to the majority of the public. +You could do that inside Ant but not inside GumpBut thats not what I wanted to say +Fix checked into both and trunk +but they are not used +Supun +Is this my assessment correct +This is fully supported from a Spring perspective and equivalent to the solution in Spring said you could of course also upgrade to Spring right away +Thanks +Because Ocean makes use field caching per segment and new segments are being created all the time something like would need to be created to fit Ocean into the existing Lucene field cache framework +I will apply the fix there also +One more issue remains updating existing data sources +Yes sorry I forgot to mention thisBut a header where this is stored would work great and probably there should be a config item if you want it or not +Thanks issue is fixed and released as part of release +Feel free to switch reschedule if you feel appropriate +verified in JBT +Can you provide a test case for this or something that shows what you need it to do? Removing the relative path seems like an odd thing to do if not a dangerous one since it would then start trying to load things from elsewhere in the load path +v +Check into the site images along with the other commit im missing something but on my machine that eps looks just like the normal solr logo no powered by textindication anywhereIs this the attachment you ment to post +Sorted +The test failure is please +Ok found a way to get those in the repository +Attached a patch that solves the Problem for me +Thanks for reporting however could you provide an example or testcase to understand better what you want to achieve +Created a new trait Config in common package it has a method that can validate a clientId or a groupId +Commit the fix in revision +Well see how that goes in the mean time I dont know we need equalshashcode as we do not store these in a listsetetc anywhere +Moving this onto the wiki +Updated patch with my commentsCan you please take a look add a test for REBOOT too? Thanks +Please let me know if you need more details +Another thing I would like to add is that code completion on the context files is also throwing the same exception described in my opening comment + +Closing this bug down. +Ill schedule this for Spring although it will probably take till Spring if the necessary changes and their potential side effects are too need to use remote callbacks and support for this by Spring would be very useful +Hudsons complaints about findbugs and release audit warnings are taken care of in this new patch +In what release is it likely to appear +But if so how can those tests attached work +Thanks Harsh. +Namit I left comments on +cant seem to go directly thru LDAP instance without requiring user principal to be in Kerberos rightCorrect its fair to say direct use of kerberos is the only option today as with the case of Hadoop in generalI still think this jira could be resolved +Thanks for marking this Existing Application Impact Release Note Needed and thinking about existing usersI agree it would be good to document +Thanks Jacques +This is all clear. +I dont know how to set? is it about +Pull it back in if you think different. +agree! working on a patch for this please put this issue into Patch Available state once you have resolved all the issues and the new patch is a better really do apologize for not catching this earlier but when I was about to apply the patch Ive noticed that you comment out blocks of code +Fixed Chriss nits thanks +Fred you set fixversion to here and we past the feature and codefreeze for anything but absolute critical changes do you expect and when for this issue? OR is this more a issue +I was just worried about a mostly unrelated improvement getting reverted due to misunderstanding +Minor issue for now +Were working on fixing the other bugs and will update the Test file when we fix them +bq +Patch file +This is a script used to load the test data which gets the results above +Lisa did you just complete the analysis or did we commit these changes at some point +I dont think its specific to a project its specific to a launch config +The ultimate guide is in docbook format under private CVS +The current approach is to have development edit manifest information in only one place and have that work for both maven and Eclipse building +Any thoughts on thatIn the adaptation Im working on I have the entity engine creating the Funambol tables and loading the seed data instead of using the SQL scripts +I have added unit tests based on commits of and this works fine on trunk and branch +Fixed in CVS +was first documented as comment of +Thanks for review. +In addition I think you should create a new issue with regards to the discussion about an error handling strategy as well so that this issue can remain closed +Each Uima As service uses a new component to send outbound messages and a component for receiving incoming messages +Ive just committed this +actually set fix version +wops correct DEBUG not WARN +Please fix +See for examples +can you provide addition information on how you got therenull privilege name doesnt makes too much sense and i dont see any reason why you should getthere +One thought if you are getting a race between messages with replyTo and temp destination creationpropagation via advisories ensure that the network bridge that propagates the target destination of the request message is also the bridge that propagates temp destinations +David is group assignment covered in the new console? If yes you can close this case. +No need for IT. +Try it out with the release candidate +Some basics Ability to parse +Well agree +Im happy to have this as wont fix. +Ive attached a patch to change this rule to also check for the parent +Im not sure what code path leads to this particular notice in the implementation of you do not directly implement the interface +Tested +Id have to dig around my network here at work I know I have a ACF server still active +I meant lets not move any of these configs to zk +Thanks for the feedback +Hmm its interesting to see that all instances come from the same thread so maybe something went wrong in one the requests and from that moment onward all the requests fail +I have included a patch which incorporates this logic +I have committed it to the Append Branch +Thanks issue is closed now +anyway I am currently traveling around and have very limited access to my code and testing environment +Good luck with the quick part there seem to be a lot of changes since you made it and lots of tests fail even after it gets compiling again +I bet this issue is playing a lot against us on small clusters under high load at least can temporarily help fixing that +Neither in the query string nor in JSTL expressions +Thanks! +Not meNice finding +Thanks Jing! +Thoughts +Svetathanks for doing the archiving SUMMARY applying the patch +Id like to propose this gets committed to its only a change to test code and the tests sporadically fail due to it +Fixed in rThanks for the report +this patch looks good to me +Verified upgrade +I didnt realize the message loader and the transcoder were separately configurable +This patch removes the unnecessary th fetch failure eliminating the race condition +If this is your only problemhomehossmantmpexamplesolrconf l command not foundudo no passwd entry forthat is easily fixed since its caused by in the distributed file +Verified in +fixed a chekcstyle +ThanksWe dont assign things as they come in they get reviewed as people with time look over them and our priorities tend to be on regressions and bugs with large numbers of votes +Regular JAR stuff just works +well it seems it might work for an NFA tooSorry yes the algo doesnt care if its N or D +enables assertions in during the junit tests +The comma has been changed to being a reserved character in the query component therefore requiring commas to be allowed in +Another option we have is just try it with las available installer instead of current one +This patch fixes the problem and adds a test case to ensure that the empty string is correctly handled +Changed configuration of to be a simple list. +Thanks Tom +Next time I see the problem Ill switch to TRACE and post back anything interesting I find +I also changed it to make use of the object and store the reference to the window object with the window div not the content div +Are you testing this patch in the context of other local changes? if you maintain the same name for subsequent versions of the patch JIRA automatically keeps track of the most recent for you personally I find it helpful to check out a fresh copy of trunk and apply my patch and run the tests there +Yup Ill do that right away +commenting +What Ive actually found is that Spring does not allow a value for concurrentConsumers +Im sorry I missed this in reviewThough the current patch works for this case with only datanode pulling it from the is closer to the intent of the test and doesnt modify +Deprecated that exposed functionalities outside ofThe Definitions interface and its implementation are still ok since they possess a piece of implementation that can be overridden +Without a normalization step we often end up with multiple keys that should have been the same such as CD Cd and cdThe patch is a first shot of originating term support and does not attempt to solve the missing for collation fields which I find is a wholly separate issue +For the while loop I moved things around a little in a way that seems more clear to me +In general mixing issue can cause this kind of situations where one part is easier to fix than the other and so the issue could stay open for longer then needed its better to track them separately as they are anyway separate issuesThanks for fixing it +Closed after receiving no further comment. +do you have an example where we do thisBy default wicket doesnt generate id i assume we are doing this for ajax to be able to tell whether this contribution has already been made? if that is the case could you please change it to something like wicketuuid or wickettag or some such +Submitted new bug +Assume supported now +if you are going to go that far then you may as well use the containers infrastructure and create the entire wicket filter inside the containerigor +I am cancelling this patch because this patch is only for patch has been deploying on our cluster for a long whileI think it is good to be committed +Thanks Jakob! +I can only reproduce while running in the container +Instead we could defer the binding on the backup until that backup becomes a primary mark them primary only bindings. +Thanks for the quick turnaround I look forward to updating when you get the fix deployed +Its OK +I have pushed the change along with a substantial expansion of the specs that accompanied it in r +Sounds like this is a race condition so this patch wouldnt be sufficient +if a job tracker fails while running a job youd want the failed jobs information shown in the history so you probably want to log a job in the file as soon as its launched rather than when its completedThere are several links that can be drilled down into from the job information tasks failed tasks task trackers etc +Hi Jitendra +this issue it was resolved long ago. +Im not sure in the end if it really makes sense to support this feature one advantage of the is that one gets the put and replace method available on the backport utility already contain support for so one can use the code transparently wo depending on the underlying a as a normal Map literally kills the implementation the issue should be closed as wont fix +Attached is an updated patch Synced to trunk as these PQ rewrite methods allow setting of the size Renamed to Please review I think this rewrite method would also be helpful forimproving Fuzzys junit tests Testing that scores are correct etc +Anyone wanting to build a bit extension would have to manually specify sparcv +It seems to be stuck in IO +In HHOM we first check if the schema matches in gossip so we can avoid the rpc in the common case of no schema change then fall back to waiting up to RINGDELAY for agreement via rpc before delivering hints +It may be possible to get it down to two commits delete extpaths introduce javadocs publish add extpaths delete javadocs publish +I guess the doesnt really need it +This is definitely something we should include +Im not familiar with your release process +Moving fix version back to since we havent decided to release without yet +If you cant move up look at for one route as to how to go about identifying and working round the XML parser versioning issue on your own machine +This could happen in a case where all maps complete more than MAXEVENTS reducers complete but some straggling reducers get fetch failures and cause a map to be restarted +Holler if you notice anything public and undeprecated thats still undocumented. +Patch that I am going to apply +contained sanefalse and debugfalseApache Ant version Sun an urgent issue moving this out +I used the external test case to test it not ideal but this needs to be fixed for the pending release +I think we should start with separating where the zip files is stored and where the container installation is unpacked +Sorry for the noiseI will try to do more research on itThanks +It was just a hint to the user that wicket doesnt use themAnd that was what threw me off as I couldnt understand why the image was only displaying on a different nd level mounted URL +From an API standpoint Id like to see a static method to toggle the safety check and static methods to get all the keys currently in the cache and to evict specified keys or all keys for a specified +I just figured to log the issue for others to vote on if they wanted it +Patches for both trunk and yahoo dist security branch +The using the embedder helped us find most of the problems +The zip is not currupt zip allows multiple entries +Patch available wasnt needed +Assume closed as resolved and released +Attaching converted test for this issue note that one testcase for group by that somehow ended up in was moved over to a new + +For the record i have no issues putting plain blackslashes anywhere in the template except in interpolated strings +speaks only of a solution +Any better suggestion for name? +Works for me +I will be making some updates for the current JMS adapters so that they can accommodate both transport and adapter mode as described in this thread +Fact is the server has no memory arguments passed in +Forgot to remove a tab +Added test case per code pasted by Ron to forum thread. +Committed to trunk with revision +The new version is now used in the Crawl class +Someone recently reported a checksum error from the didnt they +This makes it a lot easier to spot the stateful components in pages with lots of stateless components +Please disregard my last comment which was intended for a different Jira +Should be fixed at revision please verify +Patch to revert +Well attach jar files and update build scripts +It seems to me that perhaps this issue should be Blocker and not Minor as markedIf the upgrade tests are broken how can we release? Rick do you have an opinion as release manager? Deepa could you summarize any analysis you have done on the diffs +heiko for it is sufficient that the reporting failure does not block proper working of the rest of the console +MaybeAnyway its not a big deal +The exception is still available in the in header map with applied with huge thanks! +should we change the default value to be min +refactored the logic to minimize the direct dependencies of. +hi Sam Ive copied optjavajdkSunjrelibext + fix another case of the sameThis patch is covered by anything left to do here +Most hashing is incremental so if feeds the last state of hash into the next datanode and let it continue updating it the result will be independent of block size +Resolving this as I just merged this to trunk +Build and test using the shell on Firefox and IE +Not required as these tests are RH internal. +Its hidden in the FAQ +Ill commit shortly +Could you please check it again +not sure however +Here is the result of svn merge reintegrate i dont think its necessarily that easier to work with unfortunately +Logged In YES useridSimply to reduce the minimal footprint for +This feature is a common goal for a number of projects +These are the things that uses to orient itself with respect to addorganize imports +Why we need to mark tmpLoader? Is it to support xml mapreduce plan? Or it is a separate thing +Please leave the case open until I figured out the problem with +Console logs Thread dump and thrown +Yes run log splitting on an offline clusterIf it was a clean shutdown there would be no logs to splitIf it was not clean shutdown the cluster cannot come up till after logs have split anyways and do it as part of migration thered be no messing w figuring has meta been deployed or not +leaving also as because many sandbox projects were updated +I am not commenting on how it should affect the benchmark proposal at allWhether it should be in bench or contribmisc this would be a call for the committersThanks is fine I think this would be a neat task to add to benchmark w specific documentation on how to use it and for what purposes +We need to at least allow one not to have to specify a mainClass for it provide some doc on how to achieve s write a it test for it +Thanks for the reviews Stack Jesse Matteo Ted +This patch does some extra checking around the code that reads in filesIm not sure if there are other failing cases for other config files +Please validate the patch +We wont have time to get this feature in +Id like to test fixes as soon as you think theyre ready +You can see that downloading Hibernate source and looking at for +I suppose the minimum we could do would be to check the version of pyro used and provide an adequate warning that they need to either upgrade Spring Python to switch to or install the older version of Pyro +Ive been running a similar variant for some time although without the synchronized additions +We can wait for fix for to be committed verify it and then close this issueHowever several other tests of jdtdebug suite still fail due to other reasons +ie crash the updater then the manager restart one with new ddoc +lib +At least the changes log section of spec is exactly the same with that of specDoes anyone know how to report these problems to JCP +LarsThis issue cased by think that it is not similar to in at least +Johan was there a reason that this config options wasnt added to? +It sounds like such a component would require more Hadoop than Tika expertise and would therefore be better maintained within Hadoop or as a separate external project +Closing as duplicate of. +I fear that I didnt create oneYou can look at Dains code or you can wait for me to cook up a test which I will endeavour to do this eveningMy recollection which is sketchy is that Aegis tends to define types for params whether it should need to or not +I have also created a patch for the karaf features file under camelpackaging and am attaching it asI have verified that the feature installs without issues in Servicemix CheersAshwin +Not sure it is on all the mean time can you put Java in another place +Zhijie thanks for the patch! +I just committed this +We can change Scheduling Info title to be URI or some such +bq +I dont think anyone has come forward to work on this and have no evidence that will happen +I agree with Gunthers conclusion moving this out would be the immediate fix for performance reasonsBut a profiler is an advantage in many ways considering how something like yourkit takes mins to run through a min query requires a license key to inspect the outputThere is definitely middle ground to be found hereThe operator hooks need to be removed for sure since it pollutes the common and allocates objects in the inner loop +Fixed for upcoming release. +This is fixed in Hpricot but Im not sure who is responsible for releasing it +If profiles arent designed to be inherited at all what is the id field for? I thought it was done for profile merging during inheritence +I havent started on it +Well according to the commits Rohit took the time and fixed that already +I assign it to proposed solution didnt handle various generics cases properly +Its resolved because Peter has filed numerous duplicates of this issue which are more concise +Domain whitelist enabled back. +Submitting to test files +sorry I forgot to mention +Fix was checked in when the previous comment was logged. +I will be happy when removes all this deprecated stuff +What do you think +Good statistical methods +Hopefully the AMarkovLevenstein implementation I talked about earlier will be around sometime this summer +trunk Committed revision branchx Committed revision +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Either way I thought the issue should be filed +failure is likely unrelated see +Thanks for the quick response sounds like a sensible solution +WIP in second example is a frontend problem extracted into +Fixed +Submitting new patch +It is caused by Spring AOP CGLIB proxy mode the generated CBLIB proxy classes have gone out of control Im testing a workaround now +More effort is need to reproduce the problem although getting off topic for this issue the problems I am facing with Xerces are related and typically caused by clashes with other libraries +This has blocked me quite a few times +This will primarily be done by Ilya and myself +Przemyslaw Ill take a look at the patch later today but you really need to grant license to apache for the patch to be included +Thank you Ashish +It should compile without problems and it follows closely the best principles of Maven and the documentations of both and Clover plugins +The patch as submitted is not complete we still need a way of shutting down the context when the shuts down +Please reopen the JIRA if you find it is still for you +almost looks like i thought this mostly because of a ui bug i think perhaps its showing green for a moment when it should not +Is there a reason for this +Thats the intended behavior + which implements this be so kind to provide a patch +arent a region level configuration they are a cache level configuration +Does that agree with your assessment of the problem Paul +Id rather specify the Mapper class instead just because maybe someday someone subclasses puts the implementation in a different jar and then this line wont workI can set it to for now +Thanks for pointing that out +Updated patch to fix a test failure + +Fixed avro refs +The height and width are fixed and there is no scroll bar +Excellent! Thanks for the quick feedback. +did not check this properly +I removed all the java includes and javaimports and I get the same error any suggestions how to investigate this more + +Working on this currently +I checked and there are no explicit mentions of the datatypes +Patches were too broad and have gone stale +Probably the retrieve job and the classpath container should be more separated in term of configuration having just in common the and the ivysettings +Should be verified in gatein master and reopened if still valid +I would have preferred a single procedure with two parameters schemaname and tableorindexname +I think for were at least doing some simple mapping but the main issue isThis will require some updates to and POSSIBLY some updates to the JAXB databinding +It would be fine the way it wasThanks for the reviewAre the changes in specialpurposeexamplewidgetexample really needed related effects +Do I have to tell everybody not to use default namespaceswhen they send their documents to my site? On what grounds? That their documentsare invalidBy the way I xalan from a different mirror checked thatthere is no some older +Solving this issue also need to take into account the current duplication between groupId springframework and +this was fixed in Classpath by you Ian +that line was removed by accident +perfect. +Moving it to RC to ensure that is addressed +reset +Gavinafter two monthsproblem still is problem +I have committed the patch to and trunkThank you Arpit +srcmainjavaorgapachehadoophbaseutil +It seems less silly now +Removing from remoting project and CVS all together +Based on which criteria does your throw the? It seems that things get invoked properly but that your somehow encounters an optimistic locking failure and decides to throw an exception +I tried this patch and this exception is no longer being displayed + +Patch version a million for the patch! Committed! +What Im committing +error log with JDK this helps attached my full Eclipse IDE config info running on JDK Note that crashes only happen for AS +The fix will be applied to the current SNAPSHOT of +The previous patch failed when the had the id fields set to +And code that is using a usually has a way to set a custom configuration +I think newlines would mess up the CSV output of the reader tool +Random makes it very hard to reproduce an error I have had too many of them fail sporadically and have to dig through the logs to find the seed then hard code the seed +less than one hour from report to fixThanks Gabriel +The runtime does not provide a way to clip so there is no way to scroll within a container that is not full screen or is occluded by other display objects +Part of the tutorial has been review to run the example top of Apache Felix Karaf and Apache Camel of the tutorial has been updated to run on Apache Felix Karaf +Thanks Jeremy I think that using the logger in this class should be OK since this class initializes it in a static block +done for nbmnbm from build lifecycle and added a nbmdirectory goal that creates the netbeans directory structure required for the nbmcluster to work +Hadrian are you sure is an bundle? And where is that bundle jarThe features file should also be updated +but I did not like dialogfor Hibernate Mapping Diagram is more suitable +Making critical and bringing into + This bug has been marked as a duplicate of +patch that fixes the +So please report them back that this is not a Spring issue but a rather a Commons Logging one. +supports Digestand NTLM as wellOrtwin for the hints I will ask the userBen +Tim I know you are on your holidays but should we apply this to? It changes behavior so its like but changes it +Would HBASEREGIONSERVERMLOCKUSER be a better name for this new variable +yes I fixed this +Why did you change checkPermissions from a public static function to a package visible method +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Hope that the discussion on other approaches to secure the DN process is fruitful in the other jira +I am guessing that extracting everything into shims would be rather difficult +I met with same problem tooI think that following is cause of itFOP puts configuration on static field of the always applies configuration at every time when object is composedI altered code of Cocoon to configure FOP on startup timeAnd it worksIMHO It should correctThank youExcuse my poor EnglishTatsuhiko +most pages are larger than a standard screen so usually you hit an ajax link while scrolled down on the page +You were right +I split it into tests now so theres only tests per run per instance +Committed to all three branchesThanks to Anoop Ted and Stack for the review + +The other is the plugin with their download options +patch with teiid JBDS Linux +This is not specifically a Ruby problem +Should be verified in gatein master and reopened if necessary +Thanks Vasily and Dmitriy +That looks like what we saw Enis +The necessary coded has been added to and the wiki page has been updated to show how to make use of the SSO token generation capabilities. +The dynamic reload functionality itself is the reason to integrate a scripting language if you ask +Closing as this change would need to be made in the specification first +I am an observer on the WFSFES SWG +but is not getting reopened properly +The patch file for the module +Fixed the BB platformIm already setup to generate the docs +That would be nice +obfuscaten +You need the tag so that the customer gets the information it has been fixed. +Ill try nstest with the branch up and will report back +The difference might only be slight but if youre reading a million lines from a file the extra time to copy the array could make a difference +All done we might want to revisit this before making the as I reverted the solution for the logj problem +will commit when tests pass +Okay that makes sense to me +Other messages do not work to but if select Default Configuration for Runtime the message appeares and then descriptions for other configurations work too +I cant reproduce this +could have a bandaid just for the specific use case you mention but there is an easy workaround using a delegating reader that is also a +Will set back to Closed Done shortly +Attached the server is the problem starting the all config seen only when started from within the test suite? or is it seen also when starting the from JBOSSHOMEbin and invoking +This issue may have to be revisited as using X option has same effect as running DONEPROCESS DNODETACH and would also allow debugging but no ONEPROCESS option is set that can be checked for +please take a new build +It was relatively easy we used the bouncycastle cryptographic library +Supporting a proper HTTP error code requires being able to catch exceptions that are in +So far I do not fully agree with that Groovy not generate implicit getter for fields which start with one capital letter +Im preparing revision of this zip right now +r +Response Handlers Handling RSTR responses and applies the reules based on the RST and RSTR parameter instances specific rules Entropy andResponse Handlers for ValidationRenewal Cancel Bindings +Blocker for the next release +Ted do you have something youre interested in committing for? or is this done already? or to be put off to a future version +Would be great to have more information in to the client +SureshIf I understand your comment correctly I should rename to and omit decommision status comparison +Beehive now includes only both in SVN and in the distribution +No this is not done yet +Both things can be combined for sure +attached imagesimage is the gfsstoragehbasewebdatasizemapfilesshowing the reference for the splitimage shoes the reference to unknown until it can be looked at again I thank it might have to do with the latest patch from so its not related to trunk yet +Verified by Nathan +Added a workaround +Do you need the calendar variable in +lets see if we can resolve this for in svn rev +yes it will be shipped as part of. +I chatted with Tom about this and we do depend on a feature thats in the SNAPSHOT version and not available in a released version +The fix looks goodBest regards by Richard. +seems like a strange decision in retrospect +Thanks Marat +The docs I was producing is actually for multiple books so not sure if the multiple books mode has the same problem +everything is broken no motivation currently to make it running +Fixed in acf. +Thanks for the patch Kevin! +Danny thanks for looking into this so carefully I think you definitely have described whats going on in this now +I believe this fix should resolve the issues you encounteredCheersAshwin +in directory where drlvm sources reside before applying the patch and building againIf this failsould you please attach build log to this JIRA +Why was this marked Implemented? If its committed to a release branchtrunk lets reopen and as FixedI marked this as implemented rather than fixed because its a new feature not a bug that had to be fixedBut Id be happy to change the resolution if you think it makes more sense to say fixed +Reviewed itThe propertysconstants are equal and theres a hostport pair in the configuration +I was unable to reproduce the problem after a number of trials with varying numbers of threads and transformations +Great Egor! The patched fixed this issueHowever this does not fix an issue reported in +I apologize if Solr JIRA is not for this kind of questions but I really need help with the Analytics Component +I correct this +bulk close for the release +No chance to start with bypassing this bug +Screen shot of JBDS starting up so you can see what version I am using +can be used to accomplish this to some extent +jar into the release +Ill backport to soon and we can resolve this +Committed revision . +Some day wed like people to manage their cluster in a more scalable fashion than bash scripts so keeping those scripts stupid seems like a good idea to me. +A few questions rather than excluding specific files why not include just +Its only silent on a which is fine because thats the state you want anyway +I like the simple approach of just specifying a single integer value for allowable skipped before stopping the batch run due to large amount of skipped records +Dup of. +is these jiras are related +Ive created a ticket uses the maven assembly plugin to create the src and bin tarballs in disttarget +Patch for the issue. +Can we get a component update for this +Add ALLOW JOIN FILTERING to CQL clause +same thing with tomcat on windows t you think that it can be due to the contextPath that we have in tomcat and not in plexus? Otherwise I dont see +I think if its a loose coalition maybe we should hold off on adding the icons to the banner +This would just mean expanding the detection and treatment of unset unions in a bunch of other places +Additional fix r +modified files in of files deleted from for fusion +For instance the above dependency with provided scope could be easily replaced by Eclipse variable PropertiesJava build pathLibrariesAdd LibraryServer Runtime +looks good +I have made the update to the code I will submit a new patch as soon as I find a USB import functionality is tested as a part of the JCR test suite that is automatically invoked when you run maven jar to build Jackrabbit +Committed patch to trunk with revision +This is expected behavior the rail reference accessed in the async is remote and thus throws a which the finish then aggregates to a +If I see more occurrence of Lookup or search for find forms then shortly upload the patch for this fixesThanksChandan Khandelwal +If resources are released then the NM would tell the RM about the newly free space on its next heartbeat after the resize has completed +png is even better than gif +Looks like this never got submitted +Thanks! Eric +If your admin allows it the recycle bin can be reached by users but they have to go to the menu bar PersonalRecycle BinAnd yes the Recycle Bin does NOT appear in the UI only the enterprise WS does +I really wonder what can happens +The doxygen configuration files to create C API Reference Manual web pages when applied to the snapshot or Version source distribution +Marked as solved in +Thanks Ivan. +Re feedback made a static class within which is now used by both and +pushed +Since I wrote this cool technologies like jstl have come out +Committed revision with silence lazy consensus +Im then +Bumping to since its still ongoing and in Nahis court +For we can pull this particular part from +attached is the filter turned into a patchhowever I added an additional random test and it currently fails +What do you thinkFor this immediate refactoring perhaps I should revert the current exception handling and create a new Jira issue to track a modification of the exception handling +This patch backports the utility and framework changes required for supporting the same +I was wondering if there is possibly a new configuration I will needRegards Username airsquig +Yes standard properties souds better +Ive updated my patch +classpath Eclipse to JBESBCP with revision . +Thanks. +Micah first of all please note that all meclipse poms is still work in progress and likely going to change +Thanks Steve +to trunk. +WHERE ANYWHERE ANY +What kind of policies are you trying to add +Update metric group to match style of from +I agree with this requirement +Can we do something about the +Just so we know we are actually huntingfinding the same issueSnejzana are you able to reproduce this one +Otherwise we open up that entry and grab the mime type from that and return +When youre talking many hundreds or thousands of nodes there are always exceptions +and please go to Integate with weblogic is very important to me +Because its an artificial limitation from lucene truncation queries expand to one clause per possible term in the index generate enough of these clauses and you have a really slow search +However this condition only performs a candidate classification and not a match on the document +This patch should fix the compile error with the wsdl jar +but its definitly not blocker as i first thought +What it actually does is remove the link from that row into the JBMTOKENVARIABLEMAP disassociating it from the process instance and therefore hiding it from the cascaded issue with the changing type is similar in that replacing the value in the also disassociates the current value from the process instance +Im closing this because it is so old +Resolved +When is added as an order item its component products are listed as associated products with the order item +its not +Imagine a cluster and and we have created a split composed of small files that all have a replica on a single node +Seems to be a duplicate of +Requested logs and configuration not provided the provided trace makes it look as though the logger is stuck trying to access a network resource. +grails has been causing a significant amount of problems with fisheye over the last ve just had a special fisheye loaded to fix this but Im not sure if it will believe the index of grails only completed in the last hours +similar patch closing stale resolved issues +Two look different scenarios to me +Depends on a new folder for or something like that for this new version so we can keep the others the hello world and the cordova quickstarts do we want the same split? the cordova one would just be one folder since there is a shared folder between the ios and android one + +Thanks +Iow as I cannot reproduce your problems anymore this bug shall remain in itscurrent state +then press the back button to go back to page A and click any link this will force page A to be deserialized from its pagestore causing the same serialization problem +Works now +I rework it based on original mdb sample from as its not an issue +Please correct me if I am wrong +If it resurfaces Ill comment on +It is a bug in Netscape and that the user does have the ability to work around by modifying the fileObjections +where does the userdata go +Fixed a bad match error in a case statementcheers +HiWould you please try my patch? ThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +if you view the source code of the page that is generated you will see that the dropdown does indeed use the product for the value of each option the product store name is only used for display +The old package is deprecatedChanges +I had reservations about changing cli parsing characteristics for and this looks like a much better solution +Sounds good will you make a patch +grants patch but with the logic to handle revision +Burn check if this is working fine now +patch attached +There was no message in the log file regarding deployment or no deployment of the ejb +This has probably been fixed by version see related ticket references. +Here is another patch +I didnt see it very useful to include information in the scheduler events especially as that is captured by but we can add that in later if we feel its important for our analysisI think some of the information is not captured by the jobtracker instrumentation at a job level memory based blocking for instance also our initialization logic is different +I refactored the tests a bit to test all combinations of cache onoff and localglobal macro namespace +A very basic implementation of transaction timeouts is done + is to be fixed under this issue +Problem solved! An argument to lp was necessary as you suspected +It was originally marked NEEDINFO because I wanted a sample program +Bug was introduced in +It rebuilds the URL with querystring parameters properly sorted +Style +getValues renamed to get to reflect the change +I plan to leave out E and Pi +It has something to do with the new +Please check that the patch is applied as expected +We need this to work on and on Y! secure Hadoop +Thanks to Samida for removing this +I agree with Stan +JacopoI clicked on the link you provided and the page comes up immediately +I created a new constructor which has parameters the fourth parameter is for explicitly specify portlet application contextName since in this case the war file name is not equals to contextName +Mind doing that +Manually tested +I think the first comment was useful +We should instead consider making a new singleton class that provides the static mapping and is used by the test infrastructure +Its no longer your commit to do anything you want with over the objections of othersUndoubtedly you would now revert any commit I would make to rectify the situation and fix this bug +All right then +the patch has be all tickets +So far nobody ever mentioned this again +Bostjan Lah thank you your testcase is very nice and helpfull helped me a lot +Moving to +With the fix authentication will be prevented when password is blank AND using simple authentication +Hey Jukka Interesting makes sense but oddly why does Maven allow external repos to referenced at all then if they want you to just put all your artifacts into central anyways? Ill send an email to the ers asking if they would be willing to upload their jars to central and copy so stay tuned +Ive also pushed this to +I cant see any effect +committing +Fixed in r. +A project cares about backward compatibility to the same degree it cares about its users +Patch for Pauls better DLM + character encoding causes an exception +I do not like the idea of overloading the set mojoI would prefer to see a separate increment or inc mojoIt is easier for users to understand why things work the way they work if we keep each mojo doing a single functionConsider the case where you move to a new project now you dont know whether that project has the magic property turned on or not and the mojo starts behaving differentlyA separate mojo goal makes far more sense +Must have been a bad system configuration +Ill gather further opinions on this by pointing the Spring developer list at this JIRA issue +Thanks Kevin. +Removed all optional changes for base functionalityThe patch now only contains changes for retrieving of a published message and nothing else +One commentcdevcsecureidentityo Last paragraph connection organization connection authorizationThanksRick +You can also try Commons Lang this version is not affected by the bug +It means they need to do a clean build had built a overlaying model before war support that worked opposite +So I needed to add a dependency for the new jar +I used and changedSIB code +Any thoughts I would like to have that +I think what Hiro meant was do you have a reason to require specifically rather thanWe encountered this bug too but worked around it by doing bundle update which resulted in requiring instead +on patch +Ive committed this fix into and trunk +Now Im trying to set up a Windows server machine +is this ok with you +ll see on it +You should also have control over session serialization in your servlet container +But as those actions are self explanatory tooltips for them can be omitted +Jonathan do you have a screenshot of the layer group configuration +It was fixed in M so you can try this release +Its excellent +The AFE key is now in the KEYS file however it does not appear to be available from the pgp server +No response assuming ok. +The patch also adds a best match unit test similar to the example posted in this issue. +I think diskPersistenttrue by default right +I havent looked at the source but from the sound of it it should be pretty straightforward to fix +Thanks +It seems to run fine on Redhat too +We can clean them up when we merge all the parameters to the coprocessor api +This looks like a bug in jvyamlb +Tried the given suggestion +I also noticed that did not compile without having and on buildpath +Probably that will need to be deleted manually +The only way to install JBDS is using installer or installing clean Eclipse SDKEclipse Java EE IDE for Web Developers and use JBDS Update site +Whereas when you cast to uintt the expectation is that the data can be interpreted as a contiguous array of bytes +I know of one mac box that builds fine which does not have have java installed at all +I was referring to hadoop appends because I think the design of append masks the OPs issue +One more point is that some of the subprojects in mapred seem to be using as the dependency and som as the dependency +I have attached updated patch +It seems that you have an old revision just call svn update and rebuild is already applied on trunkbranch +should it +Second try +Dulpicate of +Heh do I know +Great sounds like a further optimization for both sparse and files +It will also eliminate the allocation of a temporary Long object in both get remove methods +Hey guysThis is an old ticket we have discussed this on the +XML jars are notoriousDoes this work with other versions of Xerces? Does it work with Struts deployed outside of an EARIt would be hard for other developers to put together an environment enough like yours to test accurately so if you could help better document the nature and scope of this problem it would be greatly appreciated +I will post preliminary design docs soon +Also we should wrap the code in spinnerStart so that it is in a Runable and then execute it on the UI thread +The copy rights still need to be updated or added for the following files +Please provide an example of the typo and the error message from the command line +I just committed this +also with progressing with more collaboration between ASF projectsWell this discussion should probably be on some dev mailing list +to this patch for and trunk though +For Arquillian and managed container in see. +Struts logging will not be changed +Please reassign accordingly if no action is needed from your side +generated and attached the patch +I think we should do both + is part of the distributions +yup +patch looks great and the testing you did sounds good +That would make sense but with the current shutdown logic of the server zkclient is closed before controller is shutdown +I do not want to block the release on these changes +Jason reported something similar in +Since GET requests are stateless the client can be redirected every time he needs a new dfs blockI suggest pushing this all out for a later version for simplicity +this is the same issue +Seems part of this patch was committed under in r nut not all +I am not sure whether it is legal or not to reference a type from a namespace without an explicit import directive for that namespace +Btw is a perfect sample why to NOT allow all things to get changed dynamically +Verified that this scenario no longer leaks handles. + +translates into no timeslicing and any positive value will give you timeslicingTimeslice is a crude mechanism that if maxThreads i set allows more than maxThreads connections to share the avaliable number threads +after further research this may not be possible +Fixed with a test in dd. +I believe Directory is overused in the current code +Fixed in trunk +is a rebased version against the trunk fa +xcss at +Incurring at least m of cluster downtime is not an optionWe need another solution +If added your patch with a slight change +Patch applied on looks safe + +docs contain instructions on how the jpdl xsd can be added to the eclipse catalog release instruction contain an item to upload the xsd to the docs +Committed this and pushed to branch. +If the fact is yes and which have better performance will be a better chose than Vector and +Thanks for the patch! +However I did some manual testing and things seem to be working as expected +This will also address some of the other tickets created for +Heres the back port +Moving back to for now +class suffix why not treat them the same? That is just return false when we detect a malformed +Not sure how we should solve that +Patch applies against trunk +One of the dead ends I experienced with Tapestry and Tapestry was trying to implement every suggestion +As far as I know no its not going anywhere +off theres definitely no such thing as too much information in bug reports all this is definitely helpful + +Do you have more information on the problem you hadcan it happen wo time differenceThanks +We need to discuss whether well release this publically at all given that will be merged into issue is no longer valid +There is workaround for that issue as explained with the s the reasoning behind resolving this issue as Wont Fix? We are experiencing the same warning on RHEL when using jsvc user parameter +Thanks Amal +If nobody objects I will do that in couple of days. +Thanks for finding those other issuesThere was nothing saved in the fail directory for the multi test +and the policy clearly states no ASF copyright notices in source files +maybe it only happens on a clean workspace +Committed to trunk and. +Feel free to close this one at your as the patch has been committed. +Hernn. +As a work around you can explicitly set the report date to start after the time the folder was moved +And I did insert the spaces before commit +Should not be addressed via Hibernate JIRA +Thank you for your improvement +this patch will cause drlvm build system to be dependent on successful download of from Robin Garners site +Maybe jgray can chime in +Suggest we reopen or open a new issue to track further work on this +v. +regardless of how much of what is currently in that we can push into zk the basic information of what zookeeper machines should i talk to still needs to be something we tell the machine so it can go talk to zk and get all its detailsThe fastest and most consistent route to getting in zookeeper is to put it there just like +Jason is it sufficiently covered in your docs + for patch +Try to bump to +Committed thank you Shingo +Ok thats fine +fixed! +No were removing DEPS but not DSS +Several more commits and porting to txmanager trunk +Chris is working on this one +Attached proposed patchIf no objections I will commit this patch soon +Looks promising Ill have a deeper look in the next few can confirm that your workaround eliminates the described issue but there are a couple of other cases which dont work +Happy to remove from upon commit of the patch +bulk close of resolved issues. +So the fix here is just a matter of incrementing the version used +But then again it would be a lot cleaner than what it currently is +Please list all the proposed new methods in the API or functionality being requested +Committed revision . +if the sync fails and this is the only location specified in then the actually shuts down! so technically you are right that a client can see a newly file in the namespace even before it is written to the transaction log +unit tests passed +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Committed revision . +Verified on Tools +I will take care it from next time onwards while raising the is my suggestion for does not support to check the revocation status of the client certificate +Matt could you please confirm that it indeed increases the performance +Adds recommended debug output suggested by Mike +In the mentioned case we should be able to tie the instance to a dispatcher so that its destory method would in fact be called only once +This is default ADB +Well as I said I dont see the original implementation as a bug so if all others accept the new way maybe as an improvement we will check to make sure the dont affect our current code +It is also a migration headache like if you are depending on to warn of Cartesian products and we introduce a new variable the user is no longer protected +The problem should be solved now +remove extra eclipse in path to notice and license in use proper line ends for committed in rev . +The correct file name is fixed these +Emond my big thanks for that! +It is not yet working for type and constructor completionsThis problem has uncovered a few other inconsistencies with calculating locations for completion +Thanks Zhijie! +Whats your input URL? It looks to me like you are not sending the request to the correct Request Handler but am not certain +So to which layer should this issue to be addressed? In limit operator pig layer or layer? +Removed the deployer xmbean descriptors from Branchx and is used by ve left Branch as is we are testing the VFS changes. +A clarification was requested here by QE as these two statements seem to be at odds with each other +has this one been fixed +but for a bigger change it would be better to put it on review board for comments and reviews +Make sure youre using the latest snapshot version of Spring Data Commons as well as the API has changed accordingly there as as well +Thanks very much +I will take a careful look in the next couple of days +i have a couple of questions +Reopening till it will be determined if it should remain as it is or changed to lower differentiation between httpd in EWS and httpd shipped with EAP fix solves the problem plus see comments see comments above +Dave which filters are you using? Is it possible that filters are working correctly in +To report problems with a EAP release open a support case and support will work with you to get your issue AS is an unsupported version of the application server so there will likely not be any fix for this issue in that EAP CP contained the fix for this issue in so Im closing this JIRA. + Make browsersim button aware of when current selection is a module under a into with JBDS +Infra has deleted the files from in wwwThey are still present in wwwapacheds +Easy tracking of lead for a product Better visibility of opportunity for a product Easy conversion of lead to opportunity as we have product specific data of the leadWe didnt find our suggested example in any SFA module +is not supported +Mirror possible? +It is possible that trying to close an that has not been properly initialized may cause a hangI was not able to use the + for the patch pending Jenkinsbq +Fixes all trunks present javadocs todayAvery Perhaps once you have a QA Buildbot like other projects have you can use the javadoc hooks in it to a patch? +I just ran my more complex test successfully x in a row +fixed +Vinod Kumar Vavilapalli do you have any opinion about service or processI think we are having same discussion in lets continue there +Moving to +Please select the correct project when creating JIRAsPlease close this JIRA and open a new one in project AXIS +See testsjdbcapi for an example + is for branch +So I exported the project and will attach the zip file +Ill add make files soon +NOTE this is not actually patch available +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Look forward to trying it. +An easier solution could be to just keep track on the in and wait until it reaches then you know all the flushed exchanges is done +ThanksSveta +Oh I forgot to mention that one interesting thing I did was to get rid of the disparity in handling between documents retrieved from the parser and those created directly +Do you have logs from the times when this happening is there anything in them to give more insight? Have you tried with a newer broker release latest is test case and no logs +patch looks good +it appears this was already fixed in but we should incorporate the tests in this working on this issue please also check the links to the of fixed and tested. +Looks made those two changes and committed the code is going to look at this as part of would help if the ajpoll tag wasnt commented out +Please link in relevant productproject bugs to this when there are news +It was testing that the node is null or the class it was just explicitly casted to +This caused the JVM to throw something I cant really figure out what because neither Eclipse nor is very good at explaining it +Kicked back in community land in both and transition all resolved issue that did not see any further comment in the last month to closed status +Second attempt in attached The interceptor now uses Camels type converter with a new interface instead of a conversion map +bulk close for +Validator checks and check if file andor directory will only be done if a value was entered or no value was entered and allowEmpty is false +srcmainjavaorgapachehadoophbasemasterhandlerr tyu staff Jan +And unfortunately Springs Jira setup doesnt let me edit my own comments and bug descriptions +Stack agreed with most of the commentsIn a java interface all the methods are public by default that is why I havent mentioned the word public explicitly +however the other worries me eclipse requires junit to be declared +Sorry Rajesh +Probably related +javadoc warnings are unrelated to this patch +smile +This would enable processing of input CAS and its children at the same time +Link to previous bug +Once I sum them up on my end Ill reply backPlease also look at the issues raised in and the reasons why we did not support windows paths on HDFS +That is right +Fix by removing occurrences of trim in coercionssrcmainjavaorgapachecommonsjexlsrctestjavaorgapachecommonsjexlCommitted revision +I like too +To have an system scheme in the properties component +Im intending to work on this but wont get to it until at least this weekend +The session registry implementation is not intended to survive a server restart and adding in these kind of patches is not a proper solution because it allows the user to exceed the number of allocated sessions +released closing. +Since Maven is designed to give that report we have a just want to clarify that this issue is important for various related initiatives +I would also extend from using the same pattern +A new patch added explain extended break the test to tests +I have done testing with respin CR and verified that the issue is fixed also that there are no other issues. +Thanks a lot for the reviews Todd +But I imagine we can store them in serialized form in some column +Attachment has been added with description Excuse me but I still dont get whats wrong see pdf generated by has been added with description Rendered with Fixed withBut see my attachment which indents incorrectly with +Leaving this issue open so that this note can be removed when other platforms implement the field. +As per discussion with Zheng fixed the issue from the Hive Server side rather than in the driverAll tests pass +So in your case you could try to put your applicationContext +Actually for this stack trace it does a pretty good job of telling you what happenedgot exception trying to get groups for user id No such userIn other words the username that hadoop is trying to use doesnt exist on the machine and therefore cant get any group information +I guess with a good explanation in the migration guide it should be OK +In r +there is no dispute hereIf there is no dispute what exactly is Mike talking about +Attached patch simple think its more complicated than that + +I see that makes sense +How did you avoid it +Sample project with this issue +Attached above is a different take on this problem +Can you reattach the file with that flag set +Committed file revision +This is looking like it is going to take a bunch of work for us to move on to hadoop given all the test fails +Even I saw the similar problem to adding and twice +batch transition resolvedfixed bugs to closedfixed +will make it possible to load other Journal types +Attached file is having complete information about the issue +PS I realize that hibernate itself doesnt allow the typedefelement in the but I view that as an oversight and I see no reason why Spring cant provide something that wouldindeed be very useful to the user +This Patch here requires to be able to create the specified path if mustExistfalse which is very bad why do we force this validation and only if that succeeds run the custom validator? Why dont we put this validation to the custom field validator and let the packager decide whether he needs it +The patch itself looks good to me but I would like to see some tests added or a justification why no tests are needed +server with client and server with client +Im going to assign this one to myself and set the milestone to depends on isnt actually an error this can be worked around by changing the timeout variable but this needs to be documented better somehow +Just committed +Again thanks for doing this AndyLooks like the latency is indeed reduced greatly by around judged by the meanThe slower instances now hover around ms as opposed to ms before and the other instances around ms as opposed to msIm still trying to interpret the th percentile numbers +Thanks Eric +Golden file updated +Resolved and included in BETA +Canceling patch because it is a duplicate of +Knut I plan to backport this change to this upcoming week +Could you then try out with the latest +Small updatesJust one question The only Analyzer left is +Closing stale resolved issues +Tests passed so I wont revert this +This method was being called reflectively looking up the method was failing silently at runtime +user reported that this works now +But I have a bit of a problem +Usually jar files can be found in lib +Thanks for the pointer Nicholas +I am currently facing the same issue and not able to migrate to GA yet + Sounds great +This would make start straightforwardI also like the approach of namespace objects +Please create a JBSEAM and assign to Dan +Can someone look into this issue please +Of course anything that removes the need for to have to worry about which parser to use is goodness +My cents to this bug with meAdded support to copy the expanded directory to expanded folders. +So check your and make sure all dojo js file requests are going through struts filter +Oops turns out I got this wrong +e versions worked fine +if one set a value of the new config parameter then the NN will not exit safemode unless that many have checked in +OlaThe policy right now for is that is likely to be the first really functional release and so currently tracks trunk for for the fix A bit of a softball for you though no? +Right architecure for amd is x +Thank you! +patch looks good +I will review userroles test for detect problem +This issue has been fixed and released as part of release +Done +reopen for proper looks innocent enough my questions are only if TEST scope is the right scope and if this will have a significant performance impactBut I guess all dependencies are already loaded at this point since eclipse JDT needs it before we do +fix for the comments Andy! I will address your comments and continue the cleanup +We are putting our main efford into JAXWSJAXRPC tools issues may handled by community contribution +Enriched exception messages in item and validation errors are included in the string +How this is described to work looks great +I checked a fix and test case into the branch under revision r +Could be integrated w whatever happens in here +do you have any idea where this is coming from? Id like to have it in French +That way we can still see the messages and we still have an option to see the parameters if we want to debug that +Thanks that does fix the bug +Patch to this issue +Also it seems that the patch actually adds the dependency to +I have made a few minor changes and added some comments +And that is being the stable release of we do not plan to look into this in +Err sorry looking at the wrong patch +use configured preferences problem is that the more we add the more complex it becomesCurrently it behaves like this If there is a request parameter use that as the list of media types acceptedrequested Otherwise consider the Accept request headerThen build an internal sorted set from this value +Yeah this seems to have been a version mismatch with the Hibernate library +Ill provide a patch later today than +The WSDL that is imported from the main WSDL +Of course assuming that each step takes the same time but thats as good as having a rough ETAIve looked at other progress tracker apis and they usually provide both +In any case a request timeout is hardly at the same critical level as OOM +Moved this to JBT since it is the drools plugin that is failing to handle the case of missing I dont get is why it complains if it is running on the same workspace instance krisv is the runtime info stored outside the workspace +Documented as a known issue in the EAP release notes is responsible for cleaning up the after the end of the transaction +I tried to merge the patch in but I got a conflict in and after some struggle I surrendered +I think you accidentally omitted the package since I see references to it but no code +apologise something must have gone wrong during issue creation +Adding to Brians comment +An explicit goal of Mumak is to work simulate using existing Hadoop simulators +a build lifecycle please give it a it works! Thank you +I am sure you wont have any problem in building a working jar by your own +I have now marked neoj as excluded dependency in the REST POM +Ive managed to reproduce it +These are the missing filesframeworksprojectsautomationagentbundlesptPTframeworksprojectstoolairbundlesptPTI saw some other files somewhere that were part of a ptPT bundle but were not translated +Will check it out +But the async delete just seems broken right now +commit edbbcedefedcecThat next commit fixes so file urls are not scanned through whitelist unless explicitly referenced in the whitelist access origins in +My project is included as an app that exhibits the error +comments on phabricator +Thanks AnneYour modified patch is in trunk at r at r Instead of copying labels in I moved them to less and more consistent changes in the future. +Im in the trenches right now with my project or Id take a look at this +Backported to. +tested on osx and also pass on windows xp +The fix is a bit too consequential for the current state of for the attached unit test +Great job! Changes are present on closing. +Lets be careful we dont need to introduce explicit deps on configs from HDFSMR +We have to compile M ourselves for our cluster on +It just seems like a activity to me +Ales correct +info on how to apply patch +Can you please investigate +This happens because thread votes for rather than itself as it has already received a full complement of votesThe race is therefore that thread collects all its votes and updates its own vote before thread is responded to +It wouldnt hurt to add an example command there though +I have a fix for this one will push it as soon as I have net accessbtw I found the implementation of params andor return values was broken in as well it supported them but they didnt actually work +Updated patch maintains API compatibility +Hi KristianThanks for reviewing this patch +Did you see the zip file that comes with the distribution? read the docs too tells you how to patch AS with the latest resteasy distro. +They download and execute a number of reporting tools that arent included in the final library +When I deployed a VM with only GB disk I ran into the same error +We cannot reproduce this and the report is very old. +The static constant CONTENTID is introduced with r in trunk and r in +Thinking more about this issue since we already released with this issue marked as fixed it might be better to open a new issue and close this one +Removing the use of the and using the driver class directly to avoid any issues with classloading and driver registration. +EAP both withwithout auth +At first glance it seems youll be able to use a single codec how are you planning to support different compressionencryption codec combinations? Or it is already there? +classpath and +Slating for the General Backlog then. +My problem like the original report involves running ModulesForge in Dev Studio +Committed thanks for the patch Gustavo improving the tests helps. +This seems like a good idea as we assume VMProcessors are on independent processors +I attached file +Are you suggesting to limit the number of NEXT VALUE FOR clauses per row or per statement +Committed. +Will commit if qa pass +I think this one has been solved lately but I dont have a way to confirm +Ill create slfjs prototype this weekend +Bulk move issues to and +Its broken +To use it a user will need to manually add the directive +Without finalized if we commit the doc it might be confusing +To fix the regressions well switch reptaggressive JIT option to true by default this will disable changes made in commit and the regressions will go away +This turns into a change +RecontextualizeReconfigure Reparameterize RecomposeThat is where you would handle things like that +I also dont have itch to look into this issue for though I think this is very useful functionality +The specific kind of validation error may make a difference so youll probably need to attach the xml file that youre using as wellAlso please be aware that we have limited resources that are focused on the current releases so its unlikely well make fixes to the codebase unless we have volunteers or patches for that +I think that we will probably need to build something to keep up with the stats that will be a little easier to use that also contains some amount of history +Mridul could you provide a new patch according to s code styleSijie Guo has hopefully answered about the JIRAsbq +Pushed to master +Done. +I agree saveNamespace should throw an exception if it fails all to save to any image directory +Would be great to see this fixed in thanks. +Thanks Stack +I suspect there is a jar prefix present +at least define the limitations here for M +Thanks Julien for commenting. +Verified in reopening to update release note info. +May be we should add some test cases for +Since is not available in you will still see this error when running as a Groovy script inside the IDE +Added dependency to POM +I dont believe it will cause issues elsewhere within the java client +I think its a hold over from the Mac files +It seems as though is nearing its EOL and we should be upgrading to ve been on vacation so havent been monitoring this +Im wondering if theres some environment variable +We were not yet able to reproduce this issue but if you were able to work out an example application or maybe even a test case showing the issue would be awesome +Most of the time this will cover all corrupted files in the system +Magne Can you provide some test project with steps to reproduce +If the clients would wait until they got their turn then the Java server would throttle the whole process +TihomirSure I confirm the Name attribute can be set +See and +Brian we can show New Action in ESB editors menu only if class is in the project classpath +We havent really upped requirements. +Once the locking code correctly uses locks as described in then these tests are going to require DB specific setup and should be located under +Anyway that will have to wait until at best sometime next week since we will have to rollout updated clients that can speak and first. +The new trunk patch corrects the typo in the comment +bulk close of all resolved issues. +In this patch Ive added an integration test and checked that it works on cloudservers +All modules except acceptance tests have been implemented and run to completionPatches for the acceptance test build will be delivered subsequently +Done +Uploading a patch that with all your current comments addressed +Im just thinking about what can be done to minimize diversion from what is expected to happen with normal classloaders +Thank you for the code formatting dont know if its relevant but Im using the EL Implementation just to be able to invoke will download the issue project and I will try to reproduce the issue +Patch modifiyng the Maven iPOJO plugin to be consistent with the maven bundle plugin +Fixed +For CSS we could prefix classes with clkselected or cselectedIn the Javadoc we could mention the CSS classes used by the control similar to the way we mention and CSS imports +Sorry that was my bad + is a nice class that seems for sharing between unit tests of multiple both Local and HDFS +However this seems to be fixed if we update to of the bundle plugin. +I also came to the conclusion that the is fundamentally flawed in the way it matches schema systemIds +Attempt for a fix attached +I tend to agree with Bill +hopefully will write one covering the scenarios +Latest copy of with the proposed fixed made +Francis and I paired on this today and are fixing the poms in. +I did some refactoring to simplify the debug +Now with the new Mercury jars we can fix this issue for the for the release in the branch +I found the source of the problem the generated +Hopefully you can just copy paste the beauty you made above +This bug is directly dependent on what fonts you have installed on the system compared to what language you are trying to install via the installerIf the language you select is not supported by your system the generated within the will have the incorrect paths and hence be unable to remove the shortcuts upon uninstall +Fixed on and trunk for Oracle still something odd happening on H +Thanks Ant Im getting it now and trying it out +Can someone update and resubmit +pushed +Why not just wrap them in and add a note to the javadoc? Adding an exception class plus a package seems like overkill +Complete. +Can you please look into it before you close this issue +It would extremely nice to show that we can take existing junit tests extend the test base class and have them just work +This appears to have solved the problem +will test and commit +to proceed with the issue is necessary to get answers on questions above. +And actually after we do that we should perf tests of the caused serious GC problems when I was testing automaton query +Removed unnecessary move of the unnecessary space unnecessary format changes +My problem was with. +Patch attached +pls check +I have used it and would like to use it more +Not important for though +Hi OlegThanks for the info it is much appreciatedWould you like to enlighten me if Google have a reason to not upgrading orfixing this issue? It seems that this has been a long standing bug +Applied patch to trunk and branch to resolve defect. +Its working now. +Allen Good to know and thanks for the input +End users can register new indexes using for updating indexes trough proxy the index updater is using Mavens wagon infrastructure and is running from maven embedder in other words it is using exactly the same settings as Maven so you should be able to put your proxy and proxy auth config into i understand to be able to scan a remote repository this repository must be fist scan by Nexus Indexer CLI or be served by Sonatype Nexus repository managerLike my repository is served by apache i must uses Nexus Indexer CLI +Sockets in TIMEWAIT have been closed +Heres a stab at using a LRU cache +JcodmitriyWill factor out copied code +cool +Well deferred can actually solve this issue and It can be also applied to branch +Verified fixed in ER build. +this will be fixed as part of the refactoring being done in code has been checked in +wouldnt everyone think that you could load properties from one of the provided scoped dependencies within a maven plugin or am I smoking crack + This bug has been marked as a duplicate of +Analysis LGTM v attached with a bit more cleanup +Works now as expected +Ive attached a patch against the new trunk on if it makes it any easierEdit No worries of course +Matthewyou are rigth +What exactly is the error that youre getting? Spring should actually allow for newlines there +Unable to simulate seems fixed fixed in JBDS h Fedora +A node wouldnt have missed A because of a compaction but because A was not inserted there since S is not built from the existing data on disk but incrementally with each change S really represents the combined history of all the changes performed on the token range since the last repair sessionSo nodes dont have to scan their data to build S they simply start with S when incremental repair is first enabled regardless of their initial differences and start again with S after each incremental repair sessionBut it is indeed possible for two replicas to have the same data but differing values for S for instance if a replica gets A and A whereas another misses A but gets A this would lead to some unnecessary streaming even though they both have the latest value A +Maven project used to reproduce the Ill made that behavior off by default +Sorry for the false alarm +It works but it would be nice to have wrapper methods because the api is crude for simple tasks especially when it comes to scanners the primitives are java objects and even though jruby makes a pretty good abstraction of them its still much better to manipulate true ruby objectsActually I have some code I could share that does that and it would be nice to integrate it directly in the shell +I was able to deploy a M artifact to Archiva using deploy Maven OS X Java you need to raise this as a bug against Nexus since it doesnt seem to be a Maven regression +Let me get the component chief to take a look before I commit +However I am studying how it hangs together because I would like to use it on other projects +Having that you have the jar information at compile time and then you should be able to do the compilation +srcmainjavaorgjbossforgespecjavaeejpaprovider +Will this be included in the release +it is fine thank you by Paulex +You had the CDH problem too? I thought it was just something weird with my system + +Rather than a patch to make life simple heres a ZIP of the changed files for this patch +Please verify they were applied correctly +I wonder if its possible to move the finalizer logic to +Awesomesauce +The performance improvement is pretty impressive from ms to ms per queries +If youre still getting this issue with that version of Cordova please +Can the following scenario happen with commits? A reader reads which says the index contains seg +Its really helpful as it allow us to examine occasional intermittent test failures that does not occur on our local boxes +dont you siblings talk? +I will revert the patch since there is an incompatible problem +Moved the changes here +I agree that the correct fix would be for Commons EL to change the attribute name +Patch attached +A fix for this has been committed into Qpid trunk at rev A test case for this will be commited shortly and will be back ported to the M branch subjected to the release managers approval +Do we want this in? Seems useful for presplit tables +For example there is a inkmemzero but gcc inlinesmemset for constant sizes which means it is definitely slower than the simpler memset codeI would say that either we benchmark the difference or just move to use the standard callsfor safety and simplicity ifwhen these special bits of code cause trouble +seam supports this cyclic dependency at runtime but the proper way would be to seperate the to be in the jar instead of +Its quite a long list of changes as Randy mentioned on the same ticketI think it would be a valid patch if you want to go through it and make both work with the same code base +Its not ideal as the are exported jpegs +Capitalized LWG in Summary according to convention. +The other issue is security it now becomes possible to xinclude from somewhere else if you arent careful +Please review. +Defer to defer open issues to defer to not essential to address for +The patch should take care of the problem good to me +Hi thanks for all the updates Ill investigate and get it fixed +Done for M +Fixed in head revision +However no messages will be delivered to the +Chad Powerset is trying to get code to dig in +Looks good + +fixed now +Also in each of your groups in the screen shots above you are using the group title as the instruction +A checkbox Store PSML in the database? was added to the Components selection screen with a default value false +Jo! done revision + +So the problem seems in my case to be related to having or more open through a given connection +Once this patch is in it can be done bit by bit and eventually remove the old methods +If were splitting up the backport and then improvement make a new ticket for the improvement portion +Patches for all other branches +Based on Burrs comment I am marking this resolvedShaun let me know if you have any questions +It works fine with CP +I couldnt even reproduce onCould you please specify the batch used to solve this issueThanks +A measure is a quantity that you are interested in measuring +Sean and I both half heartedly tried and both failed +For the Getting Started page ONLY please use instead of +Please fix grammar mistake in bug title +BTW it is another schema but the attached one +your diff is somehow corrupted +Applied the patch onto the trunk thanks charith +This is same to yhe issue I hope that this issue will be removed. +May be it can close this issue since I cant reproduce on trunk +I have also replied on reviewboard +so files I expect +The document briefly explains the design and test +Heres the patch +Im only attaching it in case someone wants to experiment with it to find ways to expose the bug +JT will go about the requestsone after the other +You cant simply continue and read the next row of the query when this happens +Changes look good +Should I create an enhancement issue for this +productservicessoapsrctestjavaorgjbosssoaesbactionssoap do you want to do with them +I just thought of something as well we can try to use a Zip stream to compress the output even more +you agree that this can be resolved by a documentation fix to ensure that Oracle g and above datasources should not be configured with set to true +Xaviergit is perfectly fine butgit is not +Can you check the server logs if there are any errors +The relevant jar is the wiki page has been updated to reflect this. +Wed actually need a submission for the common data and one for the artifact +Devaraj and I have run test suite multiple times +Changed Target Version to upon release of + move Diff and the related classes to a new package +Please review and apply this patch instead +Marking as resolved since there have been no comments on my previous update +Oops small bug in the last patch +can you give a bit more details on the machines you are using? what are the cpu and memory size? also what is the throughput of dd ifdevzero ofdevmapper? is there just one disk where is a partition on that disk and you have another partition for the snapshots and logseven if you dont have swap space code pages can be discarded and loaded on demand so that could be a potential problem +so perhaps a more general fix is soon coming rather than making local changes in the console +A plugin would do the job but I think there is more to this than simply scrubbing the repository periodically +Sorry Im new here +needs a little more work to be useful with the list support for rules like +Minor comment javadoc for says Test balander +It verified my guess +This process can now be automated anyway so at least the tedium is removed +I dont know where is the package +I will in fact fix this but I am going to need stacks help +New patch fixed a typo +I run core tests manually and they all passed +Applied patch in +I added stack proper stack updates when temporary id is replaced with natural id +IMO it would be great if we have the comment follow the codeRest are good +Works for me +It looks like the culprit is with consumers that may become inactive for periods of time +Yes this gives me access to the information that I was after thank you! +I just committed this +Its how IDE integration loadsimports projects and updates the POM without executing a goalphase +We experimented with some code in the that adds a shutdown hook and deletes the directory at cleanup +Is there some reason that you cant use the friendly constant +Reopened and changing Fix for version to not forget uncomment test once issue is fixed +Hive does support commentsTake a look at +So whatever the reason for the original FIXME seems to be working in mode +Leaving the issue open until this rough first version has been further refined +Tom that is correct +Thanks Lefty the document for this Jira looks good for me +I just committed this +Changed it to Fixed other comments also +Logged In YES userid Thu Aug PDT Thu Aug PDT classes Thu Aug PDT classescom Thu Aug PDT classescommsei Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseiquery Thu Aug PDT classescommseisql Thu Aug PDT classescommseisqlstoredproc Thu Aug PDT classescommseisqlstoredproc Thu Aug PDT classescommseisqlstoredproc Thu Aug PDT classescommseisqlstoredproc Thu Aug PDT classescommseisqlstoredproc Thu Aug PDT classescommseisql Thu Aug PDT classescommseisql Thu Aug PDT classescommseisql Thu Aug PDT classescommseisql Thu Aug PDT classescommseimsSQL Thu Aug PDT classescommseimsSQL Thu Aug PDT classescommseimsSQL Thu Aug PDT classes Thu Aug PDT classes Thu Aug PDT classes Thu Aug PDT classes Thu Aug PDT Thu Aug PDT +Thanks. +Attaching a patch that adds codeblock to the list of tags for which whitespace should be preserved in +I went and grabbed the release and see that youve got searchPaths AND searchPaths +were backported to as well adding the latter fix version. +I just updated the subjectFor other reading this +I also moved the clob version of test test case to to keep the pattern +For your second comment we do not put users jar before +Committed to. +Tom and Owen I took the liberty of adding you guys to the watch list so that this issue can move forward +Because once the code is in it is very hard to get it out especially when you work in an open community +Thanks for the reminder +afaik +After discussing this with Bruno we are setting this to wont fix. +I modified our construction of to include the ruby trace as the Java trace and the ruby message as the Java message so this should make exceptions coming out of BSF and a lot easier to understand +Heres the equivalent patch for trunk +Patch with review comments incorporatedbq +Note the patch would also solve +any details available on this one +Can you please rebase this on trunk? Thanks +I dont think we have to fix that for GA you have set spring nature on the project so its not blocking issue IMO +Any CI system can set the CLI parameter and embedded integration can set this option in the execution request +Let me fix. +I could pass in the prefix to the handler so the solo reason of constructing the exception message +Hi Thilina do we need to keep this as a blockerThanks think its too late to apply this patch so lets do that for next see that the serialize method now calls its which does use +Skeletal program that performs a simple test +Thank you Dhruba. +Importantly though the order is never left to Map implementation to dictateRob +Ive committed a patch and the supplied scenario now works for me +Nice does this mean the results of branchcommitmerge operations all go into the same segmentBTW I find it very confusing that we call it when in fact it is a implementation +are still some issues with using the Gwt in a project but I have found no new problems introduced by this patch. +Id really appreciate it if someone could chip in and test if the problem remains and possibly even update the patch +Ive committed this to trunk and. +Attaching patch for replacement of appid in +Do you have any logs of the servers so that we can try to narrow down the issue? In particular Id like to have a look at one log file from either +Updated patch with formatting issues resolved +ExcellentThis did the trick +Id like to add check of Null Point in to walk around this issue +Hi PradeepI applied the patch with revision efddaeaeeebaThanks for your contribution +Linking with +The fix for is the source of the problem +Patch committed +patch looks good +a token was not set in the initially when nodes were added the collision errors appearclustername FOOClusterinitialtokenautobootstrap truehintedhandoffenabled trueauthenticator authority partitioner datafiledirectories mntcassandradatacommitlogdirectory mntcassandracommitlogsavedcachesdirectory mntcassandrasavedcachescommitlogrotationthresholdinmb commitlogsync periodiccommitlogsyncperiodinms seeds diskaccessmode autoconcurrentreads concurrentwrites slicedbuffersizeinkb storageport rpcport rpckeepalive truethriftframedtransportsizeinmb thriftmaxmessagelengthinmb snapshotbeforecompaction falsebinarymemtablethroughputinmb columnindexsizeinkb inmemorycompactionlimitinmb rpctimeoutinms endpointsnitch dynamicsnitch truedynamicsnitchupdateintervalinms dynamicsnitchresetintervalinms dynamicsnitchbadnessthreshold requestscheduler +MarekYou did some work against this right? Could you update here and let me know what you need from me? +nitpick I would prefer the options be labeled in the help as something like f FILE rather than fFILE it feels more clearIf you could put together an actual patch and update the docs that we can actually push into the code base that would be great +trunk Committed revision x Committed revision Thank you Massimo Schiavon for opening the issue and review the patch! +Currently when we recover the entries from a bookie we basically send all the read requests to the other bookies at the same time +Grails will check the apps in bundles first and fall back to plugin provided bundles nothing is specified by the app. +Follow issue + seemed closer to what I want rather than +please have say something about this special version in the pom +Casting to Number would be safer +But for it we need to extend the odataj framework +I think I saw in another JIRA that the maxThreads was set to or some other low numberI regular run with maxThreads of on my little test boxYou must be confusing something else here +Patch with for the patchI used the u for the uni code chars as people in eg China etc +Removed all LGPL dependent libraries and include readme and license files for all dependent libraries +I realized the messageTemplate is final so a constructor for setting it might be a good option too +If it looks ok it would be great if you could commit this change +Although I have not dug into it I assume this is due to having polymorphic paths for looking up variables +You can use edismax for thisSPRINGCLEANING +Ive got approximately the same case under the AS +I dont think its a good choice to use such a remedy as its simply a very strange behavior I wanted to point out that the actually working example of the pop up in the showcase just stops responding to the tab key after removing the ajstatus and calls to the pop course I prefer the bug to be fixed but meanwhile I need every workaround I can get +Would this be possible +Thank you Kristian +So no Exception will be thrown +And What about? Id suggest the following can receive ODEIntalio headers but sends only ODE while sends and receives bothHow does that sound +this patch adds ivy file from the base dir to the tar ball +So that is necessary +Thenwe wont have a special case of checking for a specificattribute name of labelStyle and label and dontintroduce something inconsistent in the current implementation +I encountered the same problem +Many things have changed since including the API +Will commit when tests pass +I assume if there is no actions block it would be same as implicit +I tried to use custom Directory implementation which doesnt use file system and found that there are some places in the Solr code which dont let to do this cause they instantiate +Thanks for the quick review Knut AndersPatch b is identical to a except that I ditched the changes in +Might make more sense if they were called clientParams though +handling was introduced to in +Dmitri can you describe what is supported here and some simple test for verification +Yes so Ive noticed it while testing +on a layer and choose Properties +Committed to and branches also +All looks good +CDATA section bad doc TAE related code removed completely +I have completely rewritten fragment handling and will commit it shortly so well have to see what impact that has before trying to debug this further +I like that idea +So to summarize this issue the fix for introduced a regression that bypassed context filtering for the ROOT context causing the ROOT context to be enabled inadvertently +I tried this and all is ok +If this continues long enough the session may time out on node B even though it was accessed on node A just a moment Even if the session accessed times are kept in sync between nodes each node independently expires sessions +FYI couldnt access updated repo at git and content on github mirror was out of Mike! Ive merged that in. + for the patch +I ran and worked fine as expectedIll close this jira once we commit similar fix to trunk +Thanks Rahul! Sha beddecddbbafeca +Also modified the unit test to check serialization of control characters +Resolved in revision . +Marking critical since either there is a bug or change in behavior on Eclipse JBT or in Eclipse JBT +I think I can fix this my end by updating the MVEL Factory think I have a clean way of doing this just debugging it now +Ok running the nightly build seems to have fixed the problem weve had no repeats after a couple of days of heavy use +I think both Ram Teds suggestion should work assuming the duplicated notifications come in together +Ick +So anything needing the public API infers that youll also have EE stuff available. +or we could do a toUnicode for outlinks or directly in the fetcher +close issues for release. +Remember the execution point is in Javascript not java +Submit for review +Better to have consistent behaviour than to end up with this conversation in the users headok Ive made states now I need to connect a line from a to b draw a line from a to c +reopening for possible backport +Yep builtin support for this sounds cool +Since there a number of ways to register timeout behavior for both Callable and if a timeout occurs and remains not handled I suppose we can agree that the completion was not normal and return as a last resort without a response header thus advising the client not to retry +Dart pointed this out to me since I posted that comment +This implies that they will load the users and should probably be done in a separate process +Also I tried different mechanisms to reproduce this problem and I was not able to +ClausThanks for polishing the codesI think we can close it now +I have started writing a little code for this which I can post here in JIRA once I get it just a little further along +moving to on JBT +Looks like this is fixed along with other private gateway issues +Youre checking to see if its transient +I also see the NPE when I create the database with trunk +Reviewed with minor tweaks checked in. +on +Thanks Bill! +Step should not be necessary +Ill post it as soon as I have it completed +Marc wants a simple task +The test to reproduce the issue +Patch for Hadoop added. +Seems great thus far though +Yes possibly a flag in which we default to Wont Allow and let Jim overrideEnforcing helps keep pages with standard names which looks good for the project so newbies should be with that by default +It is always good to have tests such as theseI had no trouble applying your patch and running the two modified tests +Lets not do it before it makes sense +I appreciate your contribution! Ive committed it after some trivial modifications such as coding convention +I hereby license the attachment to ASF +PublishedFor a couple minutes after I published the book cover image wasnt downloadable from production though the rest of the site had changed but now it is +Hi YuriIm getting this same error these days +Sangeeetha Can you check what is the default offering that show up in UI for this setupthanks we try to add a Basic Zone by default UI shows for Network Offering +I closed with a fix for the + +Im getting pretty good performance with it right now but its not setup to be distributed yet +looks great! +We need it because we always update them from the product +With this parameter tenant will be created at minutes without it sec +Looks like this might have been fixed today with +It seems like you should configure Pax Runner to not delete the cache for each run otherwise youll run into this issue with every bundle that correctly saves data to is private data area +Manually changed the patch +Patch appended please review and apply it if its ok +The two gzipped files contain a reproduction sandbox for this defect +Now that work is well underway with AS all previous community releases are +This jira has been closed as part of a review and archiving of the aj project +Not sure to understand your point Bob +This version should be compatible with jsdk +thanks for testing this +Len This information from upstream has now been added to the Programmers Guide +They would obviously only need it if recompiling things though +We just point ivy at the tarball location +Bulk closing stale resolved issues +try to merge small files and avoid merge large files as much as possible +Any new on this issue +will wait before closing until a test case is written +This last patch is looking good +But I have also tried to run goals under and still cannot reproduce it +I will go over the patch tomorrow again and straight out javadocs etc +The default value for the codebase +hmm the two most recent patches are identical best I can see +This patch excludes uniquely versioned snapshots +Based on the above comment resolving this JIRA and scheduling for CR build which will have EMBJOPR release with IE works fine +Can you make it clearer that that is how it works in the shell help for get and scan? I notice you are following the convention of the other examples and they do not give you much leeway explaining your format which is a little more involved +Thanks for the intelligent review comments +Could you please include a benchmark for this in the patch +Closing issues before next release. +need to use Eclipse release to get that GA issues on to CR since GA shouldnt have issues +JeffDo we need to fix this bythanksdims +Attaching updated patch that adds support for relationships +Attaching a version of rebased for the branch +Hi DaveCould you attach a patch as a file so that I can give you credit for itThank you +Thats a good point and I can certainly see the value in that but that doesnt really have anything to do with how Im writing the data +I dont see any reason we couldnt try to support a newer GCJ that works with Java stuff but with now getting into all Linux distros theres a lot less motivation to do so +I can reproduce it +This is mostly a dirty way to achieve what should be done preperly with the real life the should be enabled immediatly in the first line and should not be changed during the lifetime of the execution +I think your work is overlapping a lot and you two need to sync either working on a single patch or on multiple smaller patches with serial dependency +Then again wrt dependent scope you are right the semantics of unscoped dont Allen said I dont care if its consider hackish because the idea works for s a bit unnerving +Now Guththila is similar to libxml in behaviour +This issue should be superceded by +Fixed in +See the linked +Sorry to bump in on this issue but the in the repo does not match the file contained in the file downloadable from sourceforge +Done all except for Eyebrowse which is not operational +I agree that having this as a config option is useful +There seems to be a few database setups which seem to fail due to wrong user credentials or drivers +Another change is to update datanode info in when it receives a registrationThis patch also includes various changes to name node front page +I fear that the tomcat fix will not help us +This sounded like a real cool solution when I read it but mixing only applies to functions within the same class +Will investigate the compiler warnings I get the same results on a tree w and wo this patch +Dave and I discussed this and decided in general that we would remove classes in rather than deprecate +In branching shortly after and on a separate jenkins build I find that I cannot get a stable build there as well so the recent issue are caused by +Code is in my fork of Weld needs merging to to master +Hi Patrickcould you please upload a patch for thisregards Malcolm +Normally that is not a problem since I share the zookeeper client across many threadsservices in the application +h adds a test failure and javadoc warnings are not related to the patch +Seeminly not of interest to anyone. +Patch committed +But this one is quite definitely reproduceable +As far as I have seen no Derby class implements Runnable +This is not a project. +closing for +If template content will be showed the issue is resolved +However it would be nice if as long as the status is In progress only one of us would touch the files +Adrian yes please use a configurable property as this will be more robust +Its also in the spirit of core supports only raw bytes for data +If you want I can create a patch for the AS build and attach it to +They look good +So is there an explicit change in URI handling so that relative paths in the are expanded to the full URLfile name in the cache +I dont see the review board item for this +See for the original that helpsJuergen +Make whatever modifications you need to I can always rebase off of whatever changes +Thanks! Ill comment on this bug if the problem reappears +Ok Stephane we have just to find the time to have added an it for this issue +And you were directed here and now know exactly what is going on +That is correct for the resource bundles +It was a coincidence that I tried it just after midnight UTC such that it passed before +Havent tried other renderers +s flasher behavior +If there are no objections I will merge this patch to and trunkColm +that mojo gets the dependency list from the maven interpolated project +Nice patch committed to trunk +Jan I see this reopened with no comment can you explain +ok megre successful this dont agree that this is even close to closed +commands +Its a right pain to see item after item +Attach a patch to improve shell +Quantify the summary and edit description for readability +I started looking into this today and realized there are additional problems with the stats faceting code changed in as it relates to tried fields and the original problem report +The patch has applied at r please verify +So somehow this needs to be fixed +I use FileImport +This is reproducible in the latest trunk +Quibble do we need to ship mockitohamcrest in the libs? I thought we only needed those for crunch unit tests I wasnt aware that any of the bits depend on them +So i already thought of generic visibility in reflection code and wrote a simulating tests +the commit has been rolled back +If we afraid of possible incompatibilities and not full support of JKS or PKCS why not to let user chooseWe can specify keystore in configs or choose type from available on current VM +Thanks for the quick SR releasewe only regret that this issue has not been integratedsolved in we will have the fix as Quick as for you please inform me about time scale for this correctionRegards +now runs cleanly +Its not reading a message when it goes boom its trying to render the schema into WSDL +Great Ill add that fix and the other one and check it in +Reopening issues to set version. +I think it was either user error or got fixed due to other changes in. +What is the build output reported in the build result +Also even with the upgrade we still need a way to filter which attributes we want to plot per mbean for system tests and organize the graphs in dashboards per role +The should apply to t show any benefit after fix closing. +Thanks Finn your solution of invoking get from get looks good +If the CTF just skpped those tests it wouldnt even try to compile the code for those tests +The ivy and pom configuration of contains a dependency on for backwards compatibility +Both test cases compile as of r +Please look into this +Verified on +Currently when defining a resource in using this adapter I need to define the as +With this version the patch in can be used to enforce CPU weights +The fix was to delete all timers for a given process instance +Ive also just uploaded +IMHO its simpler to use Servlet API for loggingBTW +Release seems to have this file and I dont think is going to be so I guess this issue can be closedIt would be interesting to know though why doesnt have that file +Is this any different from other bugs reporting that the pids are wrong on Windows? +seems like a different to x and after release. +Isnt this a bug in RI? Shouldnt we just report the bug to them rather than emulating their bogus behaviour +war file +Updated patch with one more fix this fixes all our current complex schema tests +Patch with a fix is Robbie Could you please review and commit the patch with a fix +Ive created as a separate issue +If you want to properly normalize the strings just use an Analyzer for that language +Dhruba in addition to the block replication complexity associated with snapshots as we discussed yesterday rebalancing a dfs cluster introduces another complexity +ok I swapped the arguments so that is always the first and added the since tagsI didnt add the is method +Added unit test for Exec deserialization +if you want to get the documentation part of the wsdl you can create an object using wsdl builder. +They just the unauthenticatedIdentity tweak on the serverStill no reported problems +For instance what about? the velocity and xslt subdirs? dictionary files? Simply uploading the and files will fail if any filter references say +Fixed at some point along the way +landing now works fine with BSF because it was. +We can alternatively make it configurable by asking from User +Applied to trunk and +Ok I will try to contact with someone to do this +Thanks Mikhail +This isincorrect and needs to be noted as such +Btw +Manually tested with location of conf directory + +Yes we should we definitely have to do it before putting it into productionNeed a new JIRA for that +Please can someone apply this patch or at least say what more is needed for itto be appliedIts a very simple class and it doesnt affect anything elseCheers +the updated patch looks good to me +Ill have to think about this +Ive updated the pull range for my request after cleaning up the unnecessary differences +This needs to be fixed in Glassfish as well +The callback function is called by DWR once the requested tree data is transferred from the server +Would it not be possible to register the Hibernate domain classes in do? Then a plugin could depend on the Hibernate plugin and still do their own bean registration in do +Issues resolved in were backported to adding the latter fix version. +Could you please generate this file for testing with the hope it will be the only file to be replaced +This cr can be close +Left some comments on RB +Solved this +Attached patch +Yeah was just add as since it was during bug bonanza +Ant Im moving this to Mx since Im not sure youll have time to clean up this sample so that it can make our release +Will work on this +Whether it is CRMLRM from Linux HA or both need this +Disable filtering +Can you provide some sample code to help us reproduce this behaviour +We may revisit this for a mutable token interface in the future that would include the setters but for now this works for us +Thanks for contributing! +Updated patch to include all suggestions +Please try using the CLIs batch API and submit all the operations as a batch +Opinions wanted +As Devaraj pointed out the actual fix could be in HDFS andor COMMON +Then we can try to find a design that covers all the requirements +The test failures appear to be caused by either existing issues or Bind errors just like before +you mean with GE betaIf so try saving the +Reopen the issue if it is still a problem. +Phabricator doesnt seem to have any branches so gets difficult trying to get it from one of their trees before the version bumpWill try posting couple of days later +Will this be getting resolved in the next few hours? Today is release release +Is this related to or a duplicate of +Also we need to remove +Please svn up and try to generate new patch +Second local execution ok as well +The runtime exception is the one that should be improved + +These things become hard to debugleave aside DIH because it almost entirely needs its threads and it uses it +so for doing this +Attaching patch for earlier version of hadoop +looks revision +The earlier patch had incomplete file paths +My initial suggestion with waiting for all responses before sending the rollback is already in place as the is set to GETALL for prepare only situation I can think of in which the problem can appear is when we get a replication timeout before the prepare message reaches all the participants +Please kindly help review +Nice let me know about your findings and if that current remote container can help us automate with in CI environment +This task is related to +I assume this is about script resolution renderingIn general I think with a modular setup and you already have a rich framework for feature flags bundles and services can be seen as featured and deployed started stopped at will +That should fix it +No I do not have a real fix +I looked at the rat warnings a while back iirc they didnt seem problematic but we should get these to zero in case real issues are hiding. +Those are actually the sameWicket +Of course Im curious about stop parsing what is it good for? I think thatstop renderning would be much more useful not only for me but for everyoneYes I need it! o I can offer you some help althrough Im only user ofvelocity +I found that I can switch between versions without problems +Hi Yun Feng thanks much for working on and fixing this problem and providing instructions so I could reproduce the failure +This ticket now has votes and watchers +Thats really quick! Ive tested this successfully on Windows both with and without the ANSI addon installed +Maybe in a wish task or an the time being lets apply this patch on the trunk and get rid with it +Hi Atanu according to the log file the unit tests doesnt seem to be executed by Maven + Hi Justin I have already attached a new patch for the new formattervalidator filename +I would have some use for it too +Would be good to check again with ve noticed that Remember me only works for the duration of the session +So has this issue been fixed and I just didnt notice? Or did I say something stupid +Check the logging output after running the test +Note that the change in actually is a behavior change and perhaps an incompatible one +Thanks for the comment from experience Koji! I think we can use the most latest patch then if it looks goodRe Docs I didnt find PIGCLASSPATH documented anywhere and wasnt sure which doc page to add it to either but if someone has a clue Ill doc both PIGCLASSPATH and this on the pointed page and update the patch +Please consult the Teiid team on the discussion forums before assuming something as defect +paranoia aside I have a much harder time time imagining use cases for this than I do for a data file tool +But looking back with what I know now I wish we had gone for it +Hi DhirajThe typo is corrected now Thanks! +Committed the modified patch with minor formating changesThanks for the contributionId like to add you to the list of contributors +The plugin is currently following a different release cycle even though it is part of the ESB source base but we will get around to it as soon as we will see if I can get some time to look at it this weekend apologies for the problems this is causing you. +After clicking it no error occurs however the newly created server has assigned no application and no deploy project +You will run into other problems when the scm url is inherited and the location in scm! artifactId so better define the correct scm url in the child pom +script updated +This fix should be available in the latest nightly snapshots Unfortunately we do not have the required rights to reopen closed tickets so you may want to create a new ticket pointing to this one to make sure this does not go by unnoticed +We are specifying wsdl locationOk +Ok when I did the jira the ear deployed fine to AS and EAPIll check again +I tried something similar for jetspeed portal permissions and it worked fine +Need to check the types listed in can +AnthonyYou and me know that our customer have special requirements +for global onCompletion and the routes onComletion can override the global one +Attached is a patch with suggested changes in some of the functions and some new tests for removeFilter +Yeah but Im using already +The demo also ought to be modified so that it implements both a BA Participant Completion participant and a BA Coordinator Completion participant providing a simple example of why one protocol would be chosen over the other +Will you provide another patch for these +The title in are replace was already fixed +right revise it accordingly. +forgot to change the log again +Thanks +With your property can these checks be moved to the modules which need itI would keep it like it is for now +Patch to allow wars defined via webapps config to be deployed +Ive got a copy of the proposed Tiles so will try it with that as well as the one in our EBR +redhatewsshareconf back to Permaine to decide where it goes from here +Marking closed. +they seemed to be unnecessary and might have been left over from when i started hte project and needed them to debug +is this a duplicate of +I added a color conversion for such images in revision There is one issue left on the second page +Ive confronted this too +could you please provide a quickstart that demonstrates the problem + +Thanks for the report Horst Ill work around it +They look sensible to me +SeanI see the issue and am working on it +defer all issues to +Please reopen if it is still broken. +implemented in svn test coverageTCK please create separate +Verified on JBDS +Another idea which you may want to experiment with at some point is to write a equivalent that does readahead or buffer filling in a second thread +Bug has been marked as a duplicate of this bug +We can address that independently +This fixes the problem +OK +From the ejb specThe timer service is intended for the modelling of business processes +I dont know what other fields your has but you should make all of the fields either required or optional +it is very easy are executing the script from inside the bin dir and telling it to look in the conf dir but you didnt direct it up one level +The following files are missing headers package package package package package package package package package package package package +. +proposed patch +We can add another property to to accept a code +Please do let me know if you see any problems with itI agree with AllenTodd that since Todds patch is an optimization we can get it committed even if this optimization does not work on platforms +Yeah we should backport it +We currently have the an executable JAR containing Felix and an embedded management agent that takes command line options to define identification and discovery properties +I just after the commit of which fixes the failure +Shortly Ill commit it to this belong to Common but not HDFS +There was an additional head element around the publishDate element. +Fixed the unintentional commit of cruft in to +Yes I deployed my EAR successfully to apps +Mind taking care of this one boss? You have clue on what metrics we show and when they were added +Plz review thanks. +Duplicate of fixed on of fixed on wasnt looking for a Tomahawk bug report in Core +This test can run +Ive updated the patch to point at the new some Jamie +any followed by an invalid escape sequence or just a solitary will have the same effect +It should get pulled into HA as they pull from trunk regularly. +Should be fixed soon time permitting + +Thanks Alejandro +Resolved from CSS are now +I believe drivers might have to handle peculiarities amongst their sizes until we derive a definition for CPU and VCPUSuch as I dont believe we can specify a CPU frequency as a property of a node typesize since it is not concrete +It is not executing all cases not having a break +The file shows the changes between the old source files and the fixed ones +I will follow up with him +trunk Marco if you build from source you can try with that commit which should help fix you issueI need to add some more unit testing but wanted to give you a chance to test it on your end as well ASAP +That seems to fix the problem +der and sorry this was done weeks ago just forgot to close the issue. +we dont have anyone in the team that can have a look at this right now + +This is not required for Final moving to that work is underway with AS all previous community releases are +And sorry I actually missed point about avoiding this in the Boilerplate section +Thank you stack for your help +a few lines of code specific to where the is +Paul as you create the zips for EAP can you please handle this? Thanks +Something like this boss? I tried to come up with reasons not to tag each Put and Delete and make it configurable but couldnt +Fixed in revision +Let me investigate a bit and Ill get back to you about thisAlthough its still strange that behaviour changed from +I guess we could catch stream exception and return false for both at least that would work for WAL +There is nothing we can do for the time beinguntil we move to another library or the newer version of gwtext fixes this issue +Thanks Amar! +Do you know if the name attribute on that actually works +Moving back to resolved +WARN Used bytes of buffered edits waiting for IO threads +I use the default settings. +Thanks a lot for the review Todd +Issue marked Release Notes NA because this is a component upgrade and they are not mentioned in Release Notes +Even though these failures are not correct as you pointed out +Sorry Ive reread the IOC code and youre right +One thing I found about using inlined managers is that it will have to introduce into the blueprint application code +We dont need to do +Tests passed cleanly for me on +Also did some refactoring of code +Proposed patch for this issue +If only change the name from to it is easy +This is not commercial software and you need to adjust your accordingly. +DocumentLiteral Wrapped DocumentLiteral Bare RPCLiteral the dist folder there is a complete war file which can be deployed at have testet all webservices in a +I did see the lowest levels use nolock though +use perhaps? A bit wordy but descriptive +Thanks Ram! +So it shoudnt remove any bits just add them +Patch for this issue +I switched to the standard analyzer and the problem went away and indexing also seems to be running faster than before +Reopening this because this optimization is not safe for all comparators + +However I dont think it should be a requirement to add application level classes to the brokers classpath in order to drill into individual messages +Does other databases then support SELECT identity +Committed patch at svn +Since this is a trivial logging change that fixes an annoying issue Ill commit to both trunk and pig branch unless anyone objects +So now instead of throwing exception during putObject overwrite I do a relocate first +get +Thanks Vanessa This should be documented in somewhere +Im pretty sure it will be practically impossible to find and fix all security holes +See if you want to move the code here +Bulk defer all frontend issues to +Quite possible the failure is unrelated to this particular patch but worth looking in to +closed +The bug was actually due to a test model XMI file containing a malformed annotation element +So I coded some classes to replace and for the good use of SSL scheme inIll give the sources and the compile classes in attached files +Done +See Erics comment above +Ill get it into my client and update it accordinglyI didnt see a failure in that test but do see the problem it uses a random number generator without a seed +They have the same file name on the host File System +WHERE EXISTS +It would also likely break quite a few is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Closing resolved issues for already released versions +I could not subclass because f would not be visible +That said one of the goal for is drive towards a state where incompatible changes are not allowed after itbq +Right thanks Ill take it to the nano list +Will commit shortlyThanks again Mark! +Fixed in German and Russian localization is missing so Im not closing the issue +updating ticket for current release as of in trunk and will cover documentation for allowing configuration in +Will we be able to do almost the same thing with blocks? I mean can we list blocks we are all related but not referenced in the META +No misstack. +With updated build test passed on platforms +the image i sent you in June was generated with an early so I guess something broke again in the meantime +Patch has been committed +Added selection to second page of the and resolved the widget issues. +Thanks for catching this! Please let us know if you find any other issues +The latest patch lgtm as well +Hi Eugenefrom an external point of view its difficult to say whos the culprit for this issue but what i can say is that with meclipse plugin installed the problem is always reproducible following the above steps while i couldnt reproduce it after uninstalling meclipse +Reporters of these issues please verify the fixes and help close these issues +Im not suggesting using scala in C nor anywhere +srcweb used by the containerIt uses by default the Console Appender +So we either contribute our existing line parsing to or we transform to GML and use an OL GML layer +Gets set when trying to close a region +This was a problem with unquoted paths in exec commands in the Android build scripts they have been fixed. +We are unable to use it +but probably hibernate has to access the proxy using the interface and therefore it has to reside in the same namespace as hibernate itself. +thanks for the test caseissue resolved in r a timedout or failed connect attempt needed to close off the transport to kill the reader thread +sorry +Updated specification to lift maximum role name identifier length to and to reserve SYS prefix for role names +Kathey wroteSo then upon reverting to will the Client send too many bytes for the stored procedure and thus the JDBC call fail? If thats the case I think option A is not an optionIt is the other way around the client asks for a number of bytes the server has a limit for how many it canwill send in one go +Also a simple fix. +PeteDan Done see Could you add a hack that ensures that when we copy the file for Seam versions we add the following to in the test project +creating the healthcheck policy + +Removed businessservice quickstart in revision buildersesb message Removed the businessservice quickstart as it uses EJB +How or why the above problem can occur with a standard Converter like the escapes me +Sorry for my blindness! This bug can be closed +I dont know +Thanks Nico for reporting this and for the initial implementation +I have also let it run for about hours continuously and I have manually inspected the logs +both tests are the same +Can we wrap it into a pom file and deploy to perhaps apache maven +In this case it will be damn hard for me to get an overview over the actual changes +Im going to provide updated patch for Windows in next few hours Ill take your note into I see then its fineThe only thing after the patch is complete we should not forget to file any differences our implementation would have with RI +Then please submit a proper runnable test case +I wonder if there is a way to test this maybe adding the codehaus repository using https as a protocol +Ive never known about Pichais work before I started this project +Disabling for search is not really possible +Thanks for working on this Gopal +Issue confirmed with a unit test +should be fixed now +Yeah sure +Karl sounds good +Ah then if this is fixed this JIRA is no longer relevant +Should be easy if we model something along the lines of +binanyserverPlease ack many thanks in advance +I have attached the patch for the described issue +I am not sure that the failures in the unit test is connected with this patch +I checked out HEAD of added into the project und said run as +I think the option is the best of the suggestions +It seems like most of the Store not just compaction should become pluggable because of the scanner changes as Sergey mentions in the notes +Sounds great Charlie! Im going to try to give this plugin a whirl this weekend and Ill let you know what I find +Attachment has been added with description the has been added with description sample program to reproduce the transition resolvedfixed bugs to closedfixed +The two Javadoc warnings are known issues and I believe the findbugs warning is spurious you fixing this in the same JIRA Aaron +published +I have set it right +That is great that you have this talent and wanted to contribute it back +Attaching updated patch addressing Deepas and Dans comments +Run multiprojectsite form the mavenbuild directory to reproduce the bug +So no real problem this issue can finally be stopped Thanks for all the help +This patch has been applied in the revision +flexuble enough will provide alternative patch today or can you instead close this and open another one +Now it shows the correct patch +Is a fix likely in the release +Will try adding a sleep after deploying to the other nodeundeploying active node +Realized last night that I am actually using the helper class for Derby test harness to reproduce the problem +As you can see in the source code I just added an attribute some access methods and I use it in the execute method to add params in the same way the update were addedIn order to add headers param I guess that the modification will be heavier and we probably have to transmit pairs to the +This issue may be +Attachment has been added with description Sample FO file showing the P open bugs to P pending further review +This is actually what I wanted to avoid discussing along with lease recoveryLets commit this and come back to the test directories issue later +Will commit tomorrow afternoon I meant to say +I agree with the comment Eric made +Please let me know if its a configuration issue so that we can close this issue +In this particular case I would expect the completion specific to the type to show up first +The reason that we break all the way out of the scanner iteration loop is because it is impossible for this scanner to ever yield another that might match the current table +I will fix them sometime soon +Currently SQL authorization mechanism is looking for routine descriptors to create and hence fails as they dont really have alias +Marking patch ready for just committed this +the TX will be rejected +Looks good to me +But I didnt find any +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +We actually do have features that are written per the specification that are failing several TCK tests because of bugs in the TCK tests themselvesDo you have fixes for those test cases? It would be great to have them in so that they can be easily incorporated to JSR +in that case please rephrase +committed v to online snapshot branch +Vinod comments +Bug verified +Agree about the naming convention being weird will discuss on list if we can find our own convention that makes sense +bulk close of all resolved issues. +Transitioning to waiting for is now installed on to Jake for fixing Minerva all done. +Thanks MohanThe patch was applied to java branch at r please verify +Committed this to trunk and +In old days there used to be an optimal balance to be struck between flush size compaction size and region split size +I have to take that back is in package and should therefore not be used to configure the behavior of a specific implementation +the mdsum of that zip doesnt match with what is in MDSUM file Im checking with +Thanks for the patch James +Do you know which version of commons pool you have? is javadoc in If you sure about they are the same you can try Phils solution +Commit changes to branch at Allow the dynamic import is added in the shared lib bundle +Should we try to fix that w this issue? If so can you take a crack at it? Thanks +it would be great if you can really look at the patch and express more concrete technical concerns related to stabilityI have reverted +I will look into this more +Our Business events are partly knocked out +Could you please reformat the change +Thanks Jiaqi! +Simply add a global parameter filter interceptor with session as one of the blocked parameters +Chat with a jboss web service guy he said we cant custom the data folder to put wsdl files +For example a thread pool have max threads if you firstly submit runnable which are all long blocking like the way in the addendum and the following submitted work item wont get run until the first completes even though they could complete very quickly +Sorry I didnt know there was already such an did a search on all issues but found am using looking through the issues I presume its a duplicate of of +The fix +Jenkins seems to have been down and this patch isnt on the build queueCan you try again +Thanks for the patch PaulexApplied to NIO and modules at repo revision rPlease check that the patch was applied as you expected +This great new option async would mean our application can continue to serve new requests even when others are delivering and so having it available on JMS is invaluable to usThanks is now a new option asyncConsumer you can enable to let the consumer process the exchange async +Please try nightly build of Axis branch +Agreed about the naming good point +I checked and I still had the results files cluttering up my disk +Exactly the same applies to javadocs +Not sure with other locations +raise the severity since the webconsole doesnt load at all +Sounds reasonable +I am now confused as the Java Source code supports the encoding and ISO which are the same +thanks for fixing it Virag +Actually made a small change to the patch +Should address javadoc and line length warning +Ill do some more digging +I think the if the two patches Ive attached to this issue will be committed we could close this issue in fact I think that the main reason for implementing the ability of selecting the options to display was to make the forms less confusingWith the select box approach this is addressed +public cache files should probably use the username of the TT itself +Im upping the priority on this on the basis that the XML output causes IE etc to barf. +My solution was to guard both files with the corresponding ifdefFinally to build for MSVC you can just pass the pv in your invocation +Did ant test pass +If can happen somewhere this means our unit tests are insufficient and we should add coverageI think it would be appropriate to add that functionality ifwhen we need it +move reporter from Steve to defer of open issues to +It may be safe in special case +On the other hands datanode cleverly stops itself leaving a log on the namenode +I will work on this +Yet I tend to leave out the variant and open a follow up issue that converts the case to use a skip list +But it is a start +Fixed by Henry Story +If you give me a brief direction on what you think the issue could be then I can continue by implementing a work around + +The failure in main is fixed by adding a xalan test dependency +bulk move we can reopen if we ever want to backport to +revision next we need to look at and doing some ant refactoring to enforce the presence of these files and i think changing the output generation directory from dist would be good too so its not misleading. +BTW regarding defaults Im OK with deferred flush on or off so long as its very clearly documented in the getting started guide and we provide an example config for absolute durability as well +Please check this +Bulk Close for all issues before incubation +This is v only the pb stuff was regend +Thanks Thomas +Fixed with +The file is picked up again if its changed such as length etc +Looks good +Resolved for a while +Same fix after testing +Sample java class and its test Atanu in fact you should not have to configure anything in order to get your unit tests executed +Yes this change forces every constant redefines silence +Thanks And RegardsSumit Harsha and SumitHarsha your patch is intrunk r r r r simle vs Delegator conflicts I gave upGreat collaboration + +Meng how about your work on shuffle optimization now +thanks flavio. +bulk close of all resolved issues prior to release. +the patch implement mechanism of loading service provider of prefs module which described in specadd following files for testingmodulesprefssrctestjavaorgapacheharmonyprefstestsjavautilprefsmodulesprefssrctestresourcesconfigservicesalso modify to run the test in a separated vm with different classpathWould anyone want to try this +If you believe comparable functionality already exists please state the mapping +It takes focus after interface is used +Bigger problem missing from the results for the first problem +Patch that fixes the Philippe +Apply ports the file missing from +Doron do you plan to commit this soon +Yes the issue is that the tomcat class loader does not honor manifest classpath references +The prefetched messages are now rejected on closure so that they can be close of that have been completed or are now invalid. +trunk +Addition of to list did fix the problem +Looks good +Would appreciate advice from someone who is better at reading these tracesIt looks to me as though many transactions have shared locks on the SYSCOLUMNS row for the identity column +If HA is enabled when any of the ZK operations result in the RM is automatically transitioned to Standby state +Your help would most definitely be appreciated +attached available +Attached file with the layers in question +This would explain the huge increase in memory +A backwards compatible patch is welcome +No havia ents mai qu significava +The some ordering Im continuing to work on this offline so please dont spend too much time reviewing the code +Please fix this and get the fix to Sun for inclusion with a Java release! I havea customer who uses frequently in text data +The plugin is using whereas the server is using +ShardB has Num Docs Max Doc on one node and Num Docs Max Doc on the other +increase ones salary +Setting affects version appropriately +But the site is not distributed in that form +It looks like Donald Woods also noticed this problem and provided a fix in +The idea is to make the listeners work in similar fashion to Servlet Filters +support is in latest build from the update issue is still a problem even on groovy so suppression is the only option for warnings related to that right now +Thanks Nicholas and Anupam! +the server itself at least AS provides the environment +If there are no objections Ill commit in hours +Here is what I got when I updatedA srcsrcA srcsrcmainA srcsrcmainsrcA srcsrcmainsrcmainA srcsrcmainsrcmainjavaA srcsrcmainsrcmainjavaorgA srcsrcmainsrcmainjavaorggeoserverA srcsrcmainsrcmainjavaorggeoserversecurityA srcsrcmainsrcmainjavaorggeoserversecuritydecoratorsA srcsrcmainsrcmainjavaorggeoserversecuritydecorators +No apparent loopsWe would need a repeatable to track down the cause +Tested under load and splits happen again +bq +Nothing is displayed on the console +Ah sorry JBI fixed and found they two are duplicated +Use this patchI can run correctly +Theres an option to do this so that its compatible with +Closing issue as it has been released as part of release +Its simply an API called listAPI which returns a list of their names descriptions parameter info and response info +BummerEven if I were to accidentally the resource portion of the start line I assume I would just get a not found error or something else that will hopefully be harmless +have some improvements already preparing a test patch enables counting the number of queries in the JDO tests and does a minor improvement in the getRoles have changed pom and logj config to print the SQL statementsworking on bigger ehcache cache that allows customizing cache sizes for each persisted class with an example config file for redbackTest enables the cache and checks that the number of queries is lower as expected. +Basically it now handles null aliases which I had never expected before +Thanks Tilman +It was released with the fix in +It does not make sense to turn it on in productionIf you want to do your own way of managing definition refresh do it +But Im open to suggestions +If is caught by a method it is important that it be so that the thread actually dies +Is my understanding of this issue correct that the only thing were trying to fix here is the fact the clients are not retrying attempting to talk to the active NN when it receives a safemode exception +This patch includes the Blueprint Quiesce Participant implementation from Simons patch above as minor changes were needed to compile +Closing. +Fixed with approach should be resolved without creating a session. +If eclipse uses the same configuration option +Committed revision +IE seems to work finewith just inline moz doesnt like that but seems to work well on a div with some stuff inside of itsetting on an unwrapped table is trouble thoughGiving the mozilla version the unsupported vale of results in a style with no display value that or simply not giving the wrapper a seems to work but Ill hold off on checking that in until its tested more +Hi Benjamin +I think this has been fixed a while backTo be reopened if not +Which isnt what I was after at all! Explicitly labelling the braces as a scope produces the NPE the only alternative I know about is to change loop variable names whenever the NPE +All of them will have same patch which will include all the fixesRegardsKiran ship bill ship and JacquesThe cart was replaced with screen +The guides were being sent to BREW but not staged properly +Not blocker nor critical +Never mind that previous comment +The addition of the session attribute feature doesnt change the value of this request +My mistake +Uploading again with Grant license to ASF checked +Ran consecutively x on my local machine without seeing a failure +lib +Also attached is a design document which explains the changes incorporated and the classes and the build files affected due to thisthanxThere was a small problem in the previous patch while doing a svn diff +making sure is used irrespective of security +initially on the issue persisted on I downloaded the test ear locally and will try deploying it +If false was set on both an action and a script Setting on both an action and a script is not a indispensable condition +We were thinking about ways of enhancing the type completions to show CDI specific completions +Lets wait until after commiting of bug and then separate the two patches +Patch applied at +I can commit if there are no further comments +Attaching that patch +Woops I was wrong here Roberts test is random on each iteration it replaces any Ns in the pattern w a random numberYeah the terms are equally distributed though just a fillThe wildcard patterns themselves are filled with random numbersThis is my basis for the new wildcard test btw except maybe definitely want over unless you have better ideas +Assign to and bump urgency +bzdownloads +Done +Hum no this already has been fixed by jdeolive a few days ago +Verified in +RolandI think there is still time to fix this one so it could go into alpha OlegI dont have cycles available to take care of it in the short term +This was created against branchx but trunk probably wont be much differentIMHO ready for review and possible inclusion +Includes unit test +Nope its not because we dont know if its installed its because is known not to work on all data +Agreed re specialized exceptions +html +Or adding additional admin facilities for manipulationrestart which is covered by +Why do you expect it to be? Its the next so it must be a SNAPSHOT +Now it works correctly and the Findbugs plugin does not get corrupted any more +FWIW I agree with Doug on this I dont see that subclasses would buy us much in terms of functionality except for the sake of purity of OO approach +Marking this issue as resolved. +Patch which renames the endpoint into commit this tomorrow if no one has any documentation are you thinking about specifically? Well need to document the whole of anyway and that would be part of it +applied in revision Denis please check if it was applied as thanks! +Otherwise I will likely get to this some time in May +Alan many thanks for the test case proposed ve checked in a fix to the CVS +The above bug was opened to track that issue +Hi Saphen since you are investigating that could you please help to give some comments? Thanks +In that case I will have to oppose this change since it is attempting to introduce a new semantic specific container type +Add a qualifier for intt +How do you detect that someone did this? It would seem like data loss or a denial of service +This is an issue for the referencing module in +I mean even if DSMP returns an error or something the plugin shouldnt try to download the same artifact more then once and as you can see at the timestamps its really pumps! I mean were talking download attempts per second +bq +If we are going to use enums I would do it in a release and ditch the command ordinal compatability +This might be caused by the JMS Connection getting an internal timeout and not recovering probably without the getting notified in any way +Ill take advantage of this JIRA to do some refactoring of the code that finds parity files +I have attached a test case which basically creates the dynamically and then serializes it to a temp file +According to the code it does not do that thoughSomehow is subject to racingIn the case of that happened when expireSession is called before the connection was actually establishedIs there a way to check whether the connection was established first and wait if it wasntOtherwise Id say we disable this test for now +The only thing I created specially for this issue was the java fileI have just tried again and it generates the same thing +Ive modified it so that it checks that the message context is also not null. +Thanks for having text there to begin with Andre its a great help the issue with updated release note text +With this patch is now compatible +This is a duplicate of. +need assessment for is an Eclipse plugin and the behavior seen is expected +If we keep a project structure then I think that it should remain under archetypesRegarding the integration test +It seems that now all memory leaks are fixed the memory consumption did not increase after processing K in our test system with the latest revision +Need the refactoring to support configurable parser pool implementations then can use the impl +This also should add the current port as well +Thanks Owen! +will commit if tests pass +Fixed thanks. +I made a mistake with the domain config there since I forgot the default was not appropriate with the extended access log. +using jvanzyl as thats the ssh key we have setup to connect to apache +Further editing will be necessary to make the documentation andcode coherent but once it is applied all tests pass +all resolved tickets from or older +I was looking at the CVS version +the first using whitespace and the second using comma as separator let me know which one +Shouldnt this be looking for the default realmThe default realm may not be the home realm of the current user you want to get a service ticket for +Is idempotent? After a cursory looks it seems like it could beIf so we can just call it for all children seen in node +Thanks + +There will be alot of different packages I assumed that cycling through them all would be undesirable +appears fixed between time reported and doesnt work +We need to review to make sure it terminates properlyIf the gzipped file is not setup for splits we fall back to not splitting the fileAn unsplittable gzipped dataset can be converted to a splittable one with the following Pig Latina load store a into +especially when you dont have control over the types of queries being executed against the engine +Im guessing that youre asking for review because there is something about it youre not comfortable withThere are a few things I can see that are small issues with it as I read through the patch but the one big issue I see is that its a bit of a pain to change and perhaps a bit of overkill as a solution to this problem +closed shipped in +Applied to branch in SVN rev +Im reopening this I dont believe the breaking change is justified any more +Committed revision +java file for the class than have it as a static inner class inside +Heres the patch +replace Final source file with this version and rebuild +Resolved for a while +will apply when svn available +Hi Sergio Ive tried that out +Things have changed a fair amount since this but my first comment still is true I think +Yes its less visible in the footer but I think its something that user who are interested in will search for and so find it in the footer as you did for Ant +Ill add a new entry to the unstable build site very shortly if anyone would like to verify the fix +Struts may be able to do this and you should check if it can +Usage of and related Alexandrejust committed. +This rev of the patch adds tests for correctness features and garbage collection +Actually noticed that we fail the unit tests in the exact same way as MRI +I just committed this +bq +I installed the patch and ran the TCK +The outer read can continue for a much longer period of timeHaving looked at this I think the best solution for this type or requirement is to use a second instance of the +The patch contains the changes for Kans review comments +This does not affect the outcome of the event unit test +Patch applied with slight modifications +I dont understand either and without reply we cant do with Lukas Velocity cannot automagically know youre trying to represent a directory structure then do the replacement if you dont tell it to do +As a result the handler for multipart types cannot be located at run time +Update this link and make the note an Important to highlight it +Please code review +If you really want to do this then we will need better reporting to the adminstrator so that heshe can come to know that this datanode needs tending toIn the existing scheme of things if a partition becomre the datanode shuts down and the adminstrator can see it being listed as dead in the HDFS UI +Closing this issue was rejected. +I think Im a bit confused as to how things are obviously working correctly for you and not for me + +This version of the patch increases size of a block up to Mb and file size up to MbAlso it now tries to locate last block of the file rather than the first one +Thank you very much! +There is also the file used to generate class with the castor maven plugin +Verified with latest automation runs +Awesome +Is this something for or is it not relevant any longer and should be closed +By looking at the diff of wouldnt we be able to go around that issue by tweaking CLASSPATH and HADOOPCLASSPATH in mahout wrapper +Thanks this is helpful +I want to use svn move command to make this change to preserve the history of the file +Ok I left svn rw so you could do that but Ill fix that up then to get this issue closed +Will commit if tests pass +Sample is also available +HelloAre there any updates regarding issue? There seems to be almost no development activity in Velocity community +Just for completeness a dump of the table Pieter how many Sonar analysis can run simultaneously? Can two analysis on the same project can run simultaneously +I couldnt find the cause when I took threadump it shows the bkshutdown is indefinitely waitingAnyway Ill try reading NIO documentation about the possible causes and will see our logic once +No response +Is this bug what is causing problems with when there are labels on the map? if the user pans or zooms the map then new labels are drawn at the correct place but all the old labels are also still drawn leading to cartographically poor results. +Attaching updated master file for langsubqueryFlattening +to verify +The thread session buckets mutex are converted back to thread mutex because the netvcdoioread require the conts mutex should be hold by current thread +Patch file replaces all earlier patch files +Ill try to fix this by further restructuring single step in JVMTI both for x and x +The application deploy successfully if the following classpath entry is removed from the file located in the ejb project However removing the line breaks imports in the IDE +Engine now parses as Long release +Touches the following filesM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javatoolsorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javaengineorgapachederbylocM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrdaM javadrdaorgapachederbylocdrda this to Blocker for the release +I just committed this +fot the patch +Thanks for the fast review +Defer all unresolved issues from to +The last time that this code compiled was well over a year ago +I have modified the title to cover all changes to skip +seems to be fixed. +This patch just checks that information too + was dropped and replaced it +But we will keep the JIRA here now +I had to change the instanceof evaluation to but once Id done that it stopped the errorsThanks for your help +Bugzilla cleanupFixed in FOP transition resolvedfixed bugs to closedfixed +error is still present in Wicket you very much! Ever since Ajax file uploads were introduced has been a constant source of problems for me so Im really glad I dont need to set it explicitly any more. +Should have known it was too good to be true So clinic is the reference layout if you will? I might have a play with it this weekend. +So basically I have a normal path that does all the depends checking as before and a few fast path targets that avoid it +Today I tried this with the latest JB Tools Nightly Build and it seems much easier to get the Associated Port list to open +Ill let hudson run through it and will go ahead and commit once it s +is out you know? +Thanks +The one that fails has an at symbol without curly braces +now looks for the string not supported for command allowing for the instance to be created +Need some Forge tooling for this and also need to discussbrainstorm on how the component is configured and used +please review the patch i like to commit this soon +This patch deprecates and in updates all places in Solrs core that reference these classes to use the ones in LuceneAll tests pass and this patch is +here are the full wsdlI would need the bindings files as well +Hows this +Especially since this goal respects includesexcludes to define the test code that should be part of the artifact +They dont assert anything about the owner of the conversion fromjcr names to xml namesSuggestionIm using to develope complex taxonomies and resource hierarchieswith strong and complex query need +comment updated by since then +Jesper the reason I dont want to write the parser by hand is that the Configuration bean changes from release to release +I was on vacation at the time of your last comments and than forgot about testing this +Are you fixing this by tomorrow Jai +wooksuh Thanks for the report +With the being hammered in a real world scenario your findings indicate that the slow to main memory dwarfs the extra logic for extracting the values from the packed structure +test Failed because of a timeout +There is one simple test and currently the only way to choose which nodes the collection is put on is to give an existing template collectionThe test asserts nothing at the moment all very early work +see the above comment for a solution and anyway Grails is just using Groovys so the issue would need to be raised in Groovy Jon for the sorry yeah I should have relised that it should be in the Groovy +Fixed via issue has been fixed and released as part of release +Earl please provide a testcase and more information so that a Xalan developer can reproduce this problem +Even after pulling latest changes nothing breaksAny thoughts on what this may be +I have introduced a util class which will supply the applicationContext on when the service in invoked +I added an addition to test in to show that when the unsubscribe is sent with the correct headers it works as expected +Thanks a lot Jason +Additionally I agree that a solution would be idealIm tied up on projects but please update this ticket when youd like more testing to occur +The test has been disabled in revision +Just add more info to then if its really the same issue +Hi JulienCan you provide the sample code which will lead us to regenerate the bug you have mentioned +I just committed a fix for the java problem +But looking at it again I was wrong +Should this issue still get address beforein the fix in can be set when the issue gets resolved +Moreover it should be possible to select the Rating metric as the Treemaps Color axe +Heres a for this behavior +for the record i did commit this to maintenance fixes to on on +Committed fix to trunk to remove excessive error logging +Once the delta reaches this point the script exits +Ive never seen a blocker improvement so far +Please ignore. +I mean if theres a translation there thats not working super well then someone else is more likely to fix it up and step up as maintainer. +I understand what youre suggesting and wasnt concerned about a new FS instance for each test +patch looks good +We sync from the codehaus repo Castor devs need to fix it on their. +The solution to this minor problem is for Roo to ship its own implementation of the interface +the from is may be not totally accurate either +Here is an updated patch that maintains the sockKey even for immediate loads +I am getting the same issue with newly analysed projects +Any news on this +This patch is only for fixing current test failure so no new test is attached +This test file can be added to the list of tests in +Alright good to hear that it works for you nowJuergen +to commit this please remember to update to switch to the tika plugin perhaps add a comment about the deprecated parse plugins most people look here and not in the where this change is documented +Patch applied +Hello Harrythanks for your replySorry for my bad policy +to reflect final as fixed because classes from generated sources are now included but we cant automatically capture random binary files +hasFeature is supposed to be the union of features that the container rendering server supports and what the gadget requires +There is a typo in the documentation +No analyzers are in Solr and the caching and other parts will be moved out + +SelectCreate any Sales order the status of this order should be Approved +Both should be weighed against their added complexity but the improvements offered by optimizations can and should be quantitatively measured before theyre committed +Issue reopened to correct Release Notes text as per JIRA +Axis Code generator still throws an Exception when trying to codegen for the Trasaction Specification +All replication based tests are passing +They will grow I can assure you +Attached the patch with test assertion to reproduce this issue +apiSending apisrcjavajavaxjdoSending apisrcschemajavaxjdoSending tcksrcconfAdding tcksrcjavaorgapachejdotckapipersistencemanagerfactoryAdding tcksrcjavaorgapachejdotckapipersistencemanagerfactoryTransmitting file dataCommitted revision svn commit m Add property to PMF +Only if a is available the main is using this one +As a solution therefore we force handling deprecation on all keys before the gets written out +First we had to deprecate its constructors. +Thanks Nicholas and Bobby! Bobby pointed out to me that the patch must apply to +The required destinations are described in +I am bit confused about two patches +Corrected couple comment and log messages minor code cleanup +The properties that can be configured are collected together in coresrcmainjavaorgapacheaccumulocoreconf where each is given a name default value type and description +Changes look good +However please comment whenever you see fit Ill be more than happy to incorporate comments from all reviews + is responsible in to fix this problemI dont understand what you mean by validation fails in +What I mean is a combination of lists common individual this is equally flexible but much less duplication +Alright then on v +Fixed. +See comments. +Ill integrate it +Thecommunication between the back button widget and the form handling flowscript israther hacky and need improvementReworked the calculator sample so that it use flowscript based navigation andadded a multi page form example. +Im trying to do a similar thing +Also its a new feature and not a bug +I will also attach the deployable archives +Attaching a simple sample program demonstrating how to access Derbys JMX features programmatically from a separate JVM on a local or remote machine +If this is not the case then it wont do +Just because you know about an issue doesnt mean it has been fixed +Thanks for reviewing +If we feel this wrapper class is not generic enough we could go with the new patch +Jacopo understood +might be some copypaste issues +is a duplicate of +? Agreed originally it wasnt there but its good time we add one +You may want to add an option to the appl file to enable or disable exclude docs just a thought +Add some intelligence or some mathematics to figure out which path to take under what condition +Well I dont see that on the link you provided +See for details. +Some of the in this project have a dependency on arquillian so they cant be built as part of EAP +Btw the parameter has no effect on the decision to merge with the latter does not support sorting anyways +Also tried w blank value for TARGETFOLDER and job failed as to Fred for next step +Otherwise it wont be the last time we discuss this problem +Looks like this support has already been coded for Twitter Status is a duplicate +Any chance this could be scheduled for the upcoming release as it contains tests as well +Additionally CAHR FOR BIT DATA also does not match for the modes +btw it helps when you run in one of those big containers where you hot deploy your Camel apps +Attachment has been added with description test example uses a larger image to better show the differences in file size of the resulting PDF +An alias should not be allowed if the name already exists +Yes I agree with this format +Suddenly decorating stati content works with the default and serving static content from jars does not work anymore +Currently we use the part before the colon as a kind of domain qualifier likeosgi or admin +The documentCache filterCache and query were not present +Closed as per state in Bugzilla. +OK I figured out a fix for this +This patch file just contains some tests of the version of the methodsAt a later time I intend to combine and to produce a new proposed patch but thiswill probably not happen until has been fixed since that patchwill supercede the embedded patch proposed in +Hi TuomasThank you for the patch +Thanks for the is a blocking issue for me as it seems to prevent creating new users either in Bootstrap or in normal operation +Moving to for now as not sure how much additional work we want to do around vs focusing on BPMN might depend how quickly we can get the BPMN tooling up and running. +to changing the language on smallest I used the same in the code +We are still seeing this intermittent failure although not as frequent +I use Spring so that I can run in a container but have to use a name so that EJB aware containers dont find them +Yes I think it should be and Ive added the effects docs checkboxThanks a lot for the patch +Peter would you mind if I close this issue and resolve +It then occurred to me that wouldnt this be a faster way then doing the range queriesSometimes yes sometimes no +The workaround wont work in our case since resetting will make the server unusable +I committed the change to trunk and +what are the supporting reasonsIt gives transparency +assigning fixfor so I dont forget to evaluate +Rails guys are looking into this. +yeah sorry +HiramThanks +Max please take a in order not to definitely looks much better +i have run your testcase problem is not that continue behaves like a break problem is that continue interfacts with switch +Perhaps this is happening in distributed setup? in machine setup output is done to log file see this happens only in a distributed environment +Ive absolutely run into this limit before when running web apps on the same host +Patch b addresses Knut Anders commentsKnut Anders wroteI checked the code in and it seems to correctly request more data if the string is truncated so the fix should be fineYes truncation is what happens currently although at a different place in the codeActions taken for the comments Introduced and a comment +Committed revision . +Ill have to test to yes this is Eclipse WTP behavior +I completely the inheritance hierarchy +I commited the change he proposed +Hopefully they should be in soon +Diff file to enable uncovered lines to be recorded for PHP projects +Gautam could you please check for every successful run of was there a failed map task? I just want to check if the described problem occurs whenever there is a failed map task +Minor Use Records instead of Also can you please mention how youve tested the patch +What kind of disks and was write sync enabled or not on those disks +Committed to branch. +when I am back home +does this affect the HTTP API +is there anything I can do +I intentionally didnt do that for PWD and CD +putting the patch up for safekeeping until I get things working patch looks good to +i dont think we need any more timestamps +Updated readme in test case resolution to this issue was to CACHECONSUMER as transactions were this issue. +patch to make sure that whatever we commit here we can at least test that the ngram filter doesnt throw an IOOB anymore +Hard to mix versions and be flexible +If different class loaders are used to introspect the model compared to the ones used at runtime to build messages that are intended to pass across the Tuscany wires then there will be problems +This was resolved in +got a patch to make something that works on java to +Cleaning up old issues +Setting to future pending a patch +Ill backport this if there are no objections +Lets see what comes out of +text looks OK needs to be in javadoc format if we want to get it into the to javadoc added as per java standards +Thanks Ning +Hi IvoCan you provide me your configuration that you use to enable the hornetq object store pleaseThanks need to say that I verified it by the presence of hornetq object store subdirectory in qa directory +Attached are test results from run with +HiWhere is the Server Runtime linkIf what you said is Server link on the opened Service Deployment Configuration wizardyou can get the Servers list It has two trees for servers if you create a server at first it will be displayed in the tree +Regardless of the problem caused by thread stack size those approaches look promising +Just wanted to say that the test needs to be updatedI put that comment bc in the following three lines and features are tested +Dont you want to know when something thats suppose to be there is missing? I suppose there are times when one doesnt +Hi Dustin I fear I dont have access to that file any more as I switched jobs since thenSince nobody else had this issue and I am the only one watching it I suggest you just close it as not reproducible +It seems appropriate to suport list partitioning now if we can sort the partition column and distribute the rows to the reducer to write +Files have been dropped as a result of the SVN reorg. +Removed unused issue is outdated now +I think this is fixed now. +class is available to the test case +Raghu I agree with what you said +Please add him to subversion +added to the release notes as resolved compilation error would occur when attempting to access context or process variables in code constraints +But it is fine with me +Training data is available bit by bit not all at once in these applicationsIt would be reasonable to have a convenience method to train from the rows of a matrix chosen randomly but I doubt that would be the wide usage +If we did that we would need ordering +K I will pull this in sometime today run through some local tests and likely merge it inThanks Lorin +I have discovered that this is not a CLI but rather a Core problem I will change a title to reflect that +Having it churn inout all of the ones introduces a lot of extra addremove operations for the common cases in return for saving a bit of memory in the more rare cases +fix added to release. +Should I open a new bug for the filter issue +I rarely find that people take the time to work things out so wellAttached you find a patch that fixes the problem +Thanks will wait for the testing of it by Community thing to note down is the current development version is working with Ganymede +I was wondering if perhaps instead of allowing to pass a createtruefalse we should use an enum with values CREATE APPEND CREATEORAPPEND +we could determine the walltime of the process by making a system call +Added code to to separate data source based errors as runtime errors and kept the previous validations as such +We noticed when testing another part of the system that we had SYN flood warnings in the system logs +Rebuilt with the Release tag fixed. +Shall I go ahead and check this inYes please +Will keep you updated +Hmm perhaps a tad unintuitiveI guess it would be better to use special but slightly uncommon text names +Marking it as cannot reproduce +Reopened to fix uiEnd offset issue reported by Nico Blodow +Coming back to the constants should we have the full name of the property available as a constant +Let me do that +Additional input from Hadoop PMC can help to decide the right path for HMS +Otherwise the code cannot be compiled +This bug doesnt reproduce on rev +This was also indicated in Owens comments above +I dont think its the same issue +Is that something we want to do +Yep +The web page still refers to bare and single build +Is there a separate source tree for Beta +In my scenario it is always happening when fetching multiple column names from one row of SCF +Fixed by and +Vinod you mean +a typo in the nightly patch +Please verify +Downgrading to may allow you to workaround it in the meantime +I removed the Fix Version as that decision is neither of ours its Manik Surtanis +batch transition resolvedinvalid to closedinvalid +However there are other type of problems like precision lost in float in ganglia metrics library +has been marked as a duplicate of this bug + +It may be an sbt issue +As we just backed out this should run successfully again +As comments above and in show the current behavior is an intent. +Its just a number +The choice will depend on the ordering of matchers as in +This patch contains a good code which can be extended to include more rulesIt passes mvn clean install +This patch resolve Exception missing in +That seems quite sensible +page with Continuum in Firefox +bq +Logically the transition to REGISTERED is not complete until the service has completed registration +I agree that this issue appears fixed +This worked +Would be nice to have just a flat unit test against the incrementCversion +Thanks in rev Malith for your contribution. +Would you be opposed to that +However Im hesitant about modelling the result as a vector +one selection we suppose receiving all reponse data the other we suppose not receiving all with some tcp errors +More info +this is already done +As long as the patch is not fully applied we should consider this a serious leak and raise priority to Roadblock +Please +Patch attached +we need to incorporate the factored out connection factory into Spring and document it properly +An installer is typically made to install something into a chosen destination which in this case is defined in theIf you remove this anchor from the installer its expected that it cannot work in its current specification therefore its not a bugCan you provide an example for what purpose youd want to create an installer without a +Committed to +Thanks for your efforts +All profiles that are active by default are automatically deactivated when a profile in the POM is activated on the command line or through its activation config +Thanks for the detailed powerpoint explanation +Ive documented the Preferences Page +Id like to include the fix in as we will likely use the plain skin for RF components while we sort out the Bootstrap theming +You can close this bug +Thanks VukIll do it +would be better name for the configFor trunk patch the new classes should be annotated for audience and stabilityWill take a closer look at the patch +Scaling down to a Wish +Moving to + +works for me +Youre welcome SimonPlease reopen the ticket if you find other SSL related issues when pull replicating. +scratches head and thinks +I left it alone because I wasnt sure where it would ultimately end up +Resolved for. +Im thinking part is ok and the binarysearch for meta index might not work because im not sure if the metablock names are actually sorted +diff now applied correctly +Attaching new patch for trunk with review comment of Vinod incorporated +I removed the commit and the test finished in about seconds same as if it ran unchanged +Quick note the update to CHANGES listed and instead of and . +Actually it is very easy to reproduce which you should have done +in the next patch you post +Linked in distance filtering and data format issue +How to handle different analyzers for query? Guess not possible in current design +Trivial patch applies to both trunk and +Looks like a great start MarcoOne small note service names should start with a letter just like Java method names do +Could this be an issue +Yes youre right +Deferred to defer of featurestaskswishes from to +I imagine Torsten was thinking broadly when he created this issue with such a vague title +Can someone check? In the meantime Ill ask somebody to take a look at the patch +This means creating a completely new breaks the initial premise of reusing RHEL artifacts for producing platform wont do it +I will fix the unbind issue +Attaching the patch against trunk that removes everything related to the replication state +See the listings below you will notice the difference is that is missing ls serverdefaultdeploy ls serverdefaultdeployAside I thought that was going to use Drools not? When this changes please note that should be used instead of +This file contains an operation that has a module tag +Ive attached a patch which reaches explicitly into the bookiewatcher now +Closing +This issue is a duplicate of +We also do not test IBM JDK this should be removed as well +If there are specific issues new tasks can be created. +So I think we could mark this issue as resovled or cannot reproduced +If this satifies your query can you please close this issue +The patch is ready for review and commit +Is the DBCP project dead +Anyone interested in implementing this? Should we move this to the wishlist +Jonathan the issue which I described was when the nodemanagers are lost and not killed +Funny I used various mapserver generated indexes lately and they all worked fine +Tomcat +I totally disagree that this issue is resolvedIf I have use attribute maxHeight +Bumping off since no and no more details +Im pretty confident that I checked in working code so not really sure whats up +cause unit test errors +That means that your product Id has absolutely nothing to do with your suppliers product Id +if you do select new we cant know how to get p p and p from that select new object +To me they are different things +this approach seems error prone to me because multiple launches can happen for a task even though there is completion +I didnt check up to this point +This things it works after restarting the server and after closing and opening new browser I get the same exception I did like you suggested I removed from classpath and have one under libRegards resolved. +In f. +Any chance you can check Problem on the command line with Groovy? There was a bug related to this fixed in that version but not sure it was exactly the same issue +Do we just need to commit the patch and jar? What about the comments about maven and JNI issues +strutstests +Moved to idea whether this is still a problem but we better check +Im not convinced that this problem is fixed for GIT eitherIts quite possible that Surefire always invokes releaseprepare before releaseperform +Looks like the fix was somehow undone. +x +Whoops sorry I did not read properly your code sorry +The module impl can go in the package +Im not that familiar with Netty but if benchmarks show this is a winner Im all for itPredicting the buffer size sounds good +I downloaded version and spring framework +Handle cases when the table is not in enabled state +Rick can you try replacing the in the script with and let me know if that fixes your problem +It Is NOT a BUGD I thought when I create a table in hbase it Znode should be construct in zookeeper under hbasetable sub folder But Our team members understand when we disable a table zookeeper show it under hbasetable sub folder +of them keep failing for some of the functionality isnt yet implemented +Do you propose that we move core code to a different? Apaches might be a better choice for a standard logging API +Creating webappproxy directory resolves this problemAttatched fix and test for +ChandanPlease provide database dump +Patch looks good +Im on the jetty ML and the beast is not yet fully ironed out +release bulk close +Why the Wont Fix status? Has the fix been rolled back +Any other reflect package refactoring will be handled via a separate issue +specifying the set of supported languages assigning a priori language probabilities specifying max text length etcIf neither is an issue then I could roll this in pretty quickly. + unscheduled issues to the next release +This is resolved for EAP and EWP +Hi Jay as far as I know AGSEC has dependecies on AGRAD or SIMPLEPUSH link them doesnt make sense to way this issue was already solved for this reason Im closing it +And Ill be happy to help. +I was hoping to go that way with pretty simple helper classes that expose update methods +Attached patch implements this +The attached patch fixes this issue +oops +Reopening because the problem wasnt fixed in the root cause +This should be resolved with the upgrade to +reopen after am not sure why in both cases a null array would be expected +Normally it will be fine in continuum if you have nothing to do after mvn can also call mvn with call like thiscall mvn clean installbut in this case im not sure call return the correct exit other possibility would be to copy content in your bat tried using call as you suggested but it still cuts off the script +Thanks Russell +Hi Can you please post more informations? a stack trace and if possible the XML report that makes sonar is the result screen in hudson the xml file with the Problem and the some more informationThanks for Cdricthanks for the infromation I am right now testing the new +Attach a patch +I am going to assume that UDP is really being used as I look into this +I against throwing the exception inside a finally clause while I agree it would be better to at least log it rathen than swallowing it +Looks like the patch was applied back in December +Let me think about the priorities and get back to you +clearing patch available flag until new patch is available that does what Pi suggested +Dan was this ever done +Moving to next release +Thanks just wanted to do it myself now +Ive never submitted a patch to an open source project before so please let me know if I did not do it correctly +Looking at the data model the relationship is not unidirectional it is bidirectional as definesProperty objectRegardless removing this doesnt resolve the problem time to dig a little deeper +Verified +What is your problem in concreteCheersChristian +KeithI uploaded a new patch with simplified version of no further comments. +Error log from have a nm +Im sure that nothing was changed in because we are using released version But Im happy to know that plugin now works for you. +eLooks like I had e on my update sites list +I dont think the cost of an object and next pointer should be a big deal for a short lived object +note there were of these for just one project +Make sure you run maven site with enough memory Ive used export MAVENOPTS build consumes far less memory after updateing checkstyle config +I agree that is best +Ok Thanks LarryYes I have pushed the changes. +Solves startup problem in hotdeployerAttached patch please review and commit +Again looking really good Brion! Sorry it took me a while to look again but I should be able to give you quick times going forwardSome more commentsUse a for abortTriggered instead of a mapMake the class static and put it at the class level instead of nested in a functionSet the total and lengthComputable fields for download by looking for the response headerTabs vs spaces has been an ongoing issue for our codebase because we have no checks in place +Adoption from the original patch +Again though kill should be a very rare case +This would not be encoded when servletsuperfoo would be replaced by servletfoo since the servlet connection then thinks of the connection as a normal one +javadocs for gettersetters are kinda redundant especially for codegenerated methods +Committed to trunk +I am not sure how that should be handled when using solr in cloud modePerhaps we should just remove it? I mean its a core reload right? Why do we need to support that through the create cmd as well +I have editted the svn comment to fix it +Works great on linux with the MSR +at r at r +trunk Fixed mina and sending files +This is the updated patch that fixes issue as well as this issue +I would like to submit this patch to svn trunk +dave +This is a very expensiv thing to do +I think that the actual are done in bulk so that combining can take effect on the client side +Thanks Eric! +Could you remove unused imports in +Large numbers of inserts per transaction is the usual case we recommend for users that want to insert as fast as possbile +examples is gonedocsexamplesnetboot obsoleteserverdefaultconfbootstrap no defaultconf no defaultxmbean some descriptors are gone +Making subtasks for each of the above mentioned jira issue and all the sub tasks under this are related to the system test subtask of to full issue so we can close this one. +The is no impact up to now +Since this is a bug fix I think we should commit to +As part of these changes I have avoided overwriting any existing header values +I believe this is fixed now +Same thing +I attached thechanges I made +thats s a simple fix so Im surprised the maintainers wont take the seconds necessary to make it +Bulk closing for am facing same problem in solr +I checked this into too. +Can you please describe exactly what situation is not working? Thanks +Dawid Please see my comments on phabricator +Fix version is set by a committer after the commit +Documentation has been fixed as well +Im waiting to hear back from GF folks but after spending a couple hours walking the heap analysis I see basically no references staying alive that we can correct +Ok the problem that you are experiencing is that the web browser has not been shut down so remembers the username and password previously entered on starting the new installation of to the web browser this looks to be the same server as you are connecting to it on the same address for this reason the web browser is able to respond to the challenge from the server and the user is successfully the actual HTTP requests there is actually no session for each request authentication tokens are exchanged and validated for that shortly I will be switching to some tasks to move the authentication from the web browser into the actual GWT console this will eliminate the web browser caching of this stage the scenario described is being caused by the web browser so I am going to close this issue however this will be changed within tasks we already have scheduled. +Im hitting a few regressions with this patch but when ran individually the tests pass +Sorry Timothy that did not work out +This was a misfiling of when I was still becoming acquainted with the JIRA system +Should I clone the bug +They look fineThank you +I applied this to and to fix and to fix version. +I believe that is the contents of Thirus two patches +Thanks +Thx ryan rawson for great helping +I just committed this +if you havent heardof JDBC nows the time to look it up +Lets try this +Wed just quite like a resolution to this issue +Included in release. +The test fails on checking accessibility of the public method of a local class +also not sure you did it quite against CVS HEAD +Windows Ultimate N bits +Yes it is about steps optimization. +This patch allows to build xercesc libraries with native xlC compiler on AIX Vitaly +I will update the status later this afternoon +I will look into it +to closing the issue with no action +is great! How about platform supportAlso are the relevant UI pieces done also needs to be worked onThanks again +Track component upgrades EAP diff result is attached as part of file +Currently installation of GWT from the JBDS Extras site does not work because the Extras site has M bits which did not include GWT +Added file Test for the issuePasses with interpreter fails on jitIt seems that jit makes wrong single step when stepping through the native stack frame +New issue all the admin console are of Jopr consoles not the branded console + and all examples now switched over to the new map examples might be changed but the tutorial still mentions to use which leads to unresolvable errors in Quickstart due to the changeshould be changed to M or SNAPSHOT +The instance is obtained from a which can be specified by the in the Before execution phase depending on the testcase being for your contribution! Im not a fan of thread locals at the framework level if we can avoid them so I implemented a different mechanism in revision using optional interfacesThe class is a which does nothing by default but if you want to use thread locals in your project inheriting your own rule class from that one should workPlease let us know if this works for you and Ill close this issue if it does +opensaml patch with basic saml of previous patch includes new saml source nice to see another effort at thisThanks for notifying I missed some files in the patch +Updated patch with utilities support for hup once is changed by the admin +Attached the script used to do the census +Should i revise this patch and move the logging into Seq source instead +I dont think that were going to be able to republish the poms +Proposed updated +Ive uploaded a slightly modified version of the patch +Working on unit tests +Everything is ok closing. +Added fast version of some functions to inline +change between Fixed one bug related to deleted keys exposed by the above test case +I committed the patch to +Alex thanks for fixing this +I forgot to addif the new btree approach is checked in should add tests to look out for bugs like which came about the last time a new but similar btree was checked inThe new backing indexes for deferrable constraints are not sharable for now so we should avoid this class of errors +If I change this test to be a Solr test and use the classloader from the core it seems to work +With a little more documentation +Yes there is an issue with the patch +Thanks Tony! And nice meeting you yesterday! +Tested manually that the patch fixes it +That method creates a new so the test actually has to write to tmphbaseusername +Im also worried that really the only testing is using which contains only queries +I am now done saying it +There are some limitations still that would require some testing to be sure everything still works +In fact parent is never written to so definitely unhappyI suspect all uses are wrapped in so its never blown up +The work on public render parameter for branch is finished. +Thanks DeepakThis makes sense indeed your patch is intrunk r r +See next attachment for focus on a real problem lets just redefine this taskI would be extremely happy if there was a way to make this possiblea +we should change our example config file format at well + +Impacting the entries when modifying the schema is way too complex and costly. +Verified by Vladimir. +It does not log a stack trace anymore though but just the exception +Patch is incomplete? Maybe you forgot to svn add +We now use the released version of jmDNS from maven central. +Close for all issues before incubation +A screenshot that shows the blank and outline view +That means to write a new APT tool implement all classes on top of the Groovy AST +Here is my suggestionIn case of cache key collision we can verify that the contents of the Cacheables are indeed the sameIf yes return the existingIf not both Cacheables are evicted or throw as we did previously +Bulk close for release +Transitive is for people who work with and eager to use something until we work out transitive dependency support in the framework. +rather use derby +Fixed in rev and for in rev for +the workaround described above works for me! great thanks +QE for this book currently scheduled with Andrew Ross for Friday th allow for final edits all feedback is required BEFORE Thursday th + and implementation were done for the Preview release +I think you mean the former +Heres the final diff +Ill be more careful next time. +Is there any documentation for this feature? Is it supported in the jPDL designer for EclipseThanks +Is this issue resolved? I believe the code is already delivered to trunk and build system is now making bundle from maven bundle plugin +Bulk close of resolved issues of +you can use them as locks without worrying them existing after the session +Assigning to you instead for resolved +Where exactly would you expect it to be marked as optional? far I see the change to test scope was made in time for RC +I am not sure about but why not custom producer might return it as the converted value for some property nameIn our tool we have never validated type of config property injection. +But why not relying on the jre as it is the base API and what is provided by the jre API is expected to be usableThis would imply having Karaf with all jre packages set +Wraps the timestamp in a Clock object +I see now that I can attach files +this issue applies to the product itself hence it can be resolvedsolved Ive just verified the fix in b and filed for JBT to cover the cases in features coming from as requested by +properties file +Thanks for the pull request +Lets discuss about that later +Please find attached a patch that should solve the issue +Whenever I need to export something I use maven +Is there a way to change it to publish to local maven repositoryRaja +has been marked as a duplicate of this bug + +reopening so I can modify fix target to +That certainly hate to keep pushing back on this issue but Ive spent some time with the war youve attached and unfortunately cannot dedicate any more time to trying to reproduce s needed is a simple test case that fails out of the box demonstrating your problem +One possibility is to pull the redirect handling out into its own decorator as well +Hey Gabriel Im for this but Id like to have the same semantics for the pipeline so that my unit tests would fail quickly if I was doing multiple passes over the data using the same Iterable instance +Consider this scenarioexternal SOAP client SMX consumer SMX provider external SOAP serverUsing the new SMX http component +This changed in trunk +It has been committed in r. +For instance if one of supported version is version url for sending requests should bi httpaaaapiv +Broker Bulk Message setting the producer window size to Maxit seems running yesterday night +Hi SadiDid you mean to create this bug against the Hibernate Shards project? From your description and the stack trace I dont see how Shards is involved +Please see the attached file for more info +Second draft of the patch +Fixed at some point +It wasnt even created +Remains to be fixed in tck +Thought I was working in a branch but I guess doesnt handle topic branches the same way +reopened these issues so that I can move them to Future and remove wishes +either use proper XML names or use the expanded formati will provide a patch +if sort type of closure is already operational then obviously theres no need for this +I tend to dislike AOP but without interceptors I guess Im forced to use it +Merged previous patch from trunk to branch at subversion revision +Move issues to next release +steps to reproduce svn update ant mvn N Pbootstrap install mvn install Its been a while since Ive checked Ill give it a try again thanks. +Please review this +But since this was marked as resolved after was released based on one of the comments above Ill just reopen this +As such I think you should go ahead and revise the patch to have a more conventional API +We discussed this during the JDO TCK conference call and came to the conclusion that JDO should support multiple methods being annotated with the same instance callback annotation +If there is some guidelines on how to make the component more I would be glad to help in this process +Hoop fully implements filesystem over HTTP transport +will completely chage SP stuff soon making this unnecessary +A simple looks good Earwin simple addition +Per Burr Sutter on this is a blocker for to Beta since that is where these should be verifiedresolved +My issue was created based on this practical observation and by which means I have provided a solution within other frameworks than Spring +Unless and until this feature is implemented I suggest making the error message for this particular case more specific +Now that work is underway with AS all previous community releases are +Maybe the assembly plugin could invoke the with configurations given by his assembly descriptor +Hopefully thatll give you what you need +this patch uses and sets the mode per this still exist on HEAD? Right now the only code that has a mode symbol is in packagingWhere is the API that should allow setting the mode +Opera is not sending this header so the field is indistinguishable from a regular text field +Jan please verify and close this issue +I really dont see the point in trying to support running other webapps with Solr it should be the only thing running in its process +However that seems more confusing to me than just forbidding the use of weak hashing altogetherThe functional spec should be updated to reflect this changeThis version of the patch touches the same files as the previous version plus the followingM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceNew message raised if the user attempts to manipulate a NATIVE password when is not setM javatestingorgapachederbyTestingfunctionTeststestsupgradeTestsThe upgrade tests now set if it is not set already +Today is a numeric value +Where we can talk about it and improve it? It will be here on this comments? Ok +Now each and every update for the main collection is sent to the server +this work for you +Also this will add some positive points to you during the selection process +Changed confidence level to significance level and added explanation of computation in r. +Not exactly the suggested patch but the same in spirit +Thank you! +Marking this specific issue resolvedComment regarding the test harness use of properties still applies +I tested using Tomcat +After applying lastest patch Ran Sort twice and did not observe this issue anymore +Thank you very much! +Nick why did you mark this as a duplicate +Promote this to a blockerThese messages can be ignored +It would seem wise to get it off the queue and into so that it wont get stale +This applies to builds on two different machines with OS X and Windows respectively +I suppose you have it here on the doc for get +We use spaces instead of a tab for indentation +However I did make some additional minor changes due to other commits in the trunk since then +If I missed something let me m moving the timer change for the next CR release because well have more time to do proper testing then +Nghi maybe you want to check the bookmark dictionary and compare with the PDF spec +This will correspond to the skeleton of the final patch +closing it +Sample test output showing the presenceabsence of the I noted +Thanks Dan for the contribution +Attaching a patch for the trunk that adds all the offending lines toI think this approach is better than adding more canons for wed start to quickly proliferate them wed already need one for ibm and one for ibmRan this test with ibm jdk ibm with embedded +Also there were extra lines in the master file in the previous checkin +This patch implements callQueue as a blocking queue +Hey Max in r I committed the latest patch with those tests disabled in for now +So though the command line produces some output bits what gets generated doesnt work with the bits used by the toolingThus the need for the patch to strip the from the port type name +this needs to be done but not really for the actually want to talk this over with some of the folks who do this for a living +The nd one is a small read optimization but its a bit involved so its not worth bothering with in +we should either include it or change the demo instructions to tell people to get the demo sources from the source added to the instructions that you should get a source checkout +Committed to trunk. +Paul thanks for the contribution +This is a WTP bug which have been fixed +The getting started leaves a bad taste in your mouth when you hit this +Unfortunately not all the experimental plugins build on all our supported platforms +Right after posting the patch I noticed some goofy indentation and stale javadoc comments from an earlier version of the patch which declared more exception types +Bug still there with and you make the change above then how do you add a length to a Lob property +Please review your items if not ready for please move them out to future release +Add a new jbossws facet version this version is used for if the target runtime has the extension option because doesnt support the +Please check that the patch was applied as you expected +One minor nit is that a new configuration object is being created here which i think is unnecessary +Is this not strictly an improvementfeature? It seems like it doesnt belong in stable branches +Attaching the latest patch +I may remove that naming code in the future +Linking in the refactor JIRA +looks good to me +Last failure was due toI removed trailing spaces just as an excuse to run the tests again +When the branch is setup I will fix it there +Hopefully the same is true of this fix +Also we need to make sure the new zkclient jar is backward compatible +Cheers and Thanks +This will need to be dealt with +Vladimir you are right +The above approach is more complicated but does enforce true separationIn the long run tackling the disentangling of the lib from the system is a requirement but the extra complexity requires more careful review +done. +This issue looks like its still not resolved +Thanks for the great spec Francois +bq +looks like the error is coming from inside the yui library so we cant exactly fix it +If it ends in ZZ then its not part of a literal so theres no concern for clash +if data did not get routed it should be in symoutgoingbatch with a nodeid of +A subsequent nighly snapshot should include the fix +Stay tuned +I dont think depending on a third party Erlang outside of the Ubuntu ecosystem will be kosherThe ESL package would be perfectly fine for a user who wanted to build from our release artifact on an Ubuntu system. +Attach a project that reproduces the problem with steps to should say in my case it seems to happen under heavy load and only occurs once in a while +Then make sure that the resource name is no longer selected when opening the issue details +Thanks +Patch to resolve subscription load after to head +If the expected parameters then we will use the body as is +to what Pavel states +This appears to be fixed as of at least I cannot reproduce it with either of the attached test cases +I have a fair amount of experience with maven +He fixed it. +Should we apply the patch from into the SOA platform CP build +Uploading patch b which is the same as a plus an extra commentCommitted revision +I will plan on committing in the next day or two +One thing to notice +on the patch +Resolving this for now +I also notice that youre not resetting the transform after drawing +The prefix on the old Fuzzy is just to get around scalability we could just deprecate and create a new Fuzzy without it +romain +Might end up just closing all of these as Out of date +I looked at writing tests but its awkward killing a server just after table creation and during assign +Will resubmit the patch +move underway should be done in an hour or two Ill check again later +This is fixed way back closing. +Thanks! +It may happen though that a query selects a node that still exists in the persistence manager but is deleted in your session or transaction. +We dont use Mitosis anymore. +Thanks Dag! Committed revision . +we have also been improving logs +Looks gooda suggestion for future patches is to put the Jira number on the patch file namebut no need to do that for this onethanks for all your effort +That should provide a better diagnosis than a NPE. +This would help me reduce regressions in while I am hacking on itCheersOleg +Ill figure it out +is eclipse core java compiler therefore it is signed by Eclipse +Attaching instructions in file with description how to enable and configure both features +The real question is does it make more sense to invest time in rather than further complicating FVH? FVH works great for simple phrase and single term queries but it has so many corner casesThis is my feeling too +Fixed in both trunk andThanks for reporting this issue! +Is the checkstyle old there +The javadocs for says if instance is not an entity or is a removed entityThis entity is clearly not an entity or a removed entity I agree +Again for next time +this isnt going to make it into I dont think this will make it into +Perhaps this is a bug that does not occur if there are certain dependencies already in the Maven repository +visual web tools is checked +You are right bug is fixed +dodeploy after all submodules have been copied overIf we are adding after then what we are seeing is a weird bug with the progress dialog on OSXIf we are adding it before then this buig is clearly still valid +I fixed the reactor order in the patch. +So +True +If your databases are already in sync then it will stay there until the first new write +it will be a great one to get fixed and help make execution more reliable +removed ZK path regenerated the patch after rebasing from trunk +Initial tests were successful but need to test furtherThnaks to peek mechanism implemented to parser early this year this taks was relatively easy +If thats the case let me reassure you that it will go in tonight or tomorrow late afternoon +Work to increase performance has been done already in the code stream +Need to add integration tests that are performed on the adminapi +Changes to are already in TRUNK +Rob says he fixed this yesterday +Are you planning to fix this today +Andrew I left some comments on phabricator +The repositories are only in parent pom. +This was corrected quite a while back +Its a pure luck it even builds I wonder if the your build script can have e added as bash option so that build fails on commands returning errors +If I understand you well you mean that this is NOT a bug rightMaybe its a silly question but why the code generators behavior differs between the skeleton and the stub? Could it be possible to have an additional option for the tool to handle the header in the skeleton like in the stub +JordanI think your ticket was but the issue seems identical to +in rThanks Filipe! +has not been updated for yearsIll check in your patch in a moment +The implementation stores downloaded resources from external sources into a downloads directory of a projectThe downloads directory is not deleted on a mvn clean by default +SOAP or +I have just committed an unit test which demonstrates the issue Its set to be Ignored +Just in case attached updated patch with changes described above +No it has not been committed +load +resolved in trunk and branch +Hmmm why did you disable the test? Everything works fine for me. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I dont know if weld can do that but in theory it should beWhats important here is there is no need to do anything from JSF side if and only if the bean container deal with the scopes +We dont want pending transactions to hold up Region splitting so if a region splits the transaction will have to be retriedIf you are simply doing a bulk upload of data without any other readswrites happening then I wonder if you really need to use transactionsIn general your code should handle the when trying to commit +I tried running this WSDL file through wsdljava from CXF +probably files versus console +Does this need reworking? is theres some concurrent container these days +So dont really see a reason for this +Do this when we upgrade to Icefaces was already done some time ago. +I committed this to branch +works for agree that an exec as a simple to config option should be required +Looking forward to this +Uwe thanks for taking a look! Ill fix this +I think using a default context is probably a good idea +I can see that otherwise theres the possibility of infinite recursion +Fix in progress +In this case speculatively fetching the same block from multiple datanodes or speculative will probably reduce overall throughputIf however the number of readers is substantially smaller than the number of datanodes and drives and access is known to be sequential then speculatively blocks could speed things some +You are describing a very different issue from the installation problems aboveAre you seeing this behavior in Java code or in Groovy code only? It is possible that this is coming from the Groovy Information Hover +Saksham when you say new template do mean the new systemvm template for vmware for you on +This is not a bug see for enhancement request to allow sending of IAC commands +Can you open a new issue against seam tools for that and link to your PRThen please close this issue again +That shouldnt be a mandatory s a it right as +write it down in small are you considering being more EJB user only friendly regarding generators +This time they told me something different +Otherwise when I get some time Ill see what I can do +Oops +Putting aside this specific test case for a minute if you apply your patch to do all the rest of the regression tests pass successfully +Is is so? Wall clock time would depend on what the bottleneck is +I had some rough idea for this when I opened +Any suggestions would be welcome +Fix applied at +Is it acceptable that we leave our version as is? Or perhaps just change it in the PR to jbossasquickstarts +Please could you verify that everything has been applied as expected +if you get it working well with we may as well submit it back to since I know they at least would like to move toward parallel threads in the future +Views are supposed to capture this idea of changes to the group of processes executing a distributed computation +So instead of the steps in the description you can do this Create JBOSSHOMEserverSERVERCONFdeployorgapachecatalinastartup Extract the in and copy into this directory +Thanks +This issue is not a blocker nor a critical for EAP +Hi Jan can you attach the schema which caused the issue please +how will you go and change users its a Beta release +Merged to doc branch at revision +When searching for a lanquage it would be of a great help +This is clearly a bug +Features look fine wo any changes neededVerifying w a local build now +Just a direct get set into the Proto object +Someone else may want to work on this +At least I think we need something that couldreturn all the data for a feature in one JNI call +done +Hey Campbell What would the maxRetry be used for? If wed be counting the number of failed connection attempts it would need a cleaver mechanism of parsing failed connection responses otherwise we could be retrying on impossible to recover scenarios such as wrong password invalid username +I agree Im working on a similar optimization +right even for something simpler than phi wed want to sample say x per second at most +A set has no ordering +Can you confirm that the problem lies with transitive dependencies +Have spun off the wizard changes into and will address them myself +Maybe we could call it or something for specificity but Ill leave that part up to you guys +unzip the attachment and cd to sample +The real fix maybe too much hard work as this time. +Still happens with Castor with exactly the same messages and amount of for your think the issue is that some anonymous types now get generated not only using their their parent element or attribute names but also their grandparent element names etc +Can you elaborate? Is there a related ticketIs there a way for the user to change their mind about their entity groups? Say a group grows too large can I manually split itbq +Sorry I meant the patch attached to fixes this +Much easier to debugSELECT FROM joinOrderFIXED FILES F constraintFILESREPOSITORYPATH FILECHANGES FC constraintFILECHANGESFILECHANGESET CHANGESETS CS constraintCHANGESETSPRIMARYIDWHERE AND ORDER BY DESC +In the integration tests I currently install the which failes here and there when jboss fails to start up +Can the link be added to as well +Ill hold off until after some testing with build to officially close this +committed +One more question That would be very nice if you could attach the patched version of? Thanks a patched version of using patch provided earlier +Initial commit Features still missing Automatic archival of journal log entries to save space Using a database as backend for the for geting started on this issue +Any news on that? It would be nice to include your changes in incoming release! +I see that now bullets are rendered in Confuence greatCould it be possible to have different types of bullets in function of indentation +Yep I meant to click the donate button but forgot + +Seems good +Ive created a couple versions of the logo pulling colors from the header +and btw +Fixed in commit eefb +Committed +Wonderful Brian this fixed the issue +There are pros and cons to this +Indeed I wondered there was no CVS commit message when Sylvain closed the bugI thought it will come soon and forgot about it. +Saves me time since you have already done this analysis +Agreed +Consider that in a multimodule configuration I am sharing surefire config and define a set of groups to run +However Ive added a check in the as plugin so it doesnt blow up when it does occur r +patch that moves to m not sure why the itself is buried in the Parser +Xuting this error was seemingly caused by one of the moderators for the mailing list having a mailbox that was full +I didnt ask for a pom you did +on talking up the peer aspect of partipants +By debugging and looking at their ant task Ive found the current thread context classloader has to be set as well with classpath used to compile the report they use this classloader for report design checks before actual compilation +It could just continue to use the existing Tapestry Component +Names of exclude lists were updated +If the test is full featured enough that it could be called as a tool then there is no way that it is a true unit test +I think we should spell out in more detail whats allowed +Might close as wont fix altogether +Shouldnt take very long +FYI Cloudscape supported a CALL statement but did not support procedures + Username duchin +Busy times here right now so gimme some timethanks +Thanks to Tyler Hobbs for tracking this down +that also looks like a separate issue that can be dealt with laterSo it looks to me like this is ready to be committed +The failure of and are unrelated see +Is this ready for commit? Lets do it then its breaking builds +Eliminating unnecessary stack traces has just increased the number of cases where this can occur +It is some packaging issue +Ive contacted Apache to look into whats wrong on the hudson machines but so far weve not heard back +In federation we wanted to use for sharing a single checkpointer for all the namenodes in a cluster triggered based on crontab +It seems like isnt being set everywhere +Do you have a file that you need it to read? Make it small and we can add it to the test not expect it to do user defined at this also doesnt use hints because Im not certain how the user passes hints in right at it and comment on the results +This patch should allow the raw stack trace and error to be returned if the response is not XML +So the data is sorted by string first +The attached patch fixes this problem +Uploaded the wrong patch +I dont see any references to in any of the capabilities documents +SteffenI suspect it may not be possible to introduce this functionality without losing ability to pipeline HTTP messages +the problem still in problem? You mean the xml is still commented out +It could have been created as a patch and that is far as its gotten +There is also a problem with on windows it rotates the log files incorrectly since it operates with wrong info about the log file sizes +Reapply r in the upgrade see and this was done though the of r isnt linked here +Javadoc crashes with out of memory on my machine I dont believe this patch is to blame however +Throwing the is probably the least controversial solution at this point It does not interfere with the delicate workings of the runtime and is more useful than a deadlock anyway +Will there be any issues if I try to port this fix to Hadoop branch? I want to do this since it would make to easier +although no full consensus has been reached featured implemented functional and released! +I have created a new issue +I think we could just comment them out instead +Yes I added a table with the transferExchange option to hazelcastseda +Sending to you max for approval +This also means that the WSDL can remain in the AAR file as they no longer need to be on the CPThoughts much appreciated +pending Jenkins +VenkatFYI we dont use the fancy stuff from JDK as the code should compile and run on at a minimum +Also at the time of reset password the current password required and which is passed via email in encrypted way +Posted here so Chris Howe can proceed with his work on March +Slava can you look into this +Thanks +In the process we make others sad because we trade exact compatibility for better perfDo we sacrifice the good of the majority for the good of the minorityAll that said I will play with an additional check to see if its +The HTTP element charset defaults to whereas the default URI charset should beI see no way around adding a new HTTP parameter specifically for the URI charsetNew patch attached +So the Commons implementation doesnt sound as an happy choiceWhat about then to keep the and disable cache for the request parameters? Such that the cache cannot be overflowed by a malicious user +This data subset would be a part of the entire user experience not just reporting +O colaborador em questo j no se encontra associado a empresa +combine with dbunit to prevent having to redo security tasks for all suite started will grow as things are fixed +It looks like you are indeed reproducing the deadlock +testsuitedbdesign +do this I will have to revert the change made in Alan +For the sake of others watching this issue Upayavira and I have agreed to kick off this process early next week +batch transition to closed remaining resolved bugs +Thanks! Can you please attach the zip file to the bugzilla issue so its trackedIts be helpful if you could do a diff +Closing as a result of the Selenium incorporation via +I also added a unit test +Youve attached but youve cutpaste the two files that it uses and and Im a little worried about a further cutpaste from this issue into my own +vote for inclusion Rob +How do you know this is the expected behaviorFor specs you basically need to use your judgement +From looking at the code in WSSJ I dont believe it can although I could be mistaken +Patch attached +Patch to use Resources Servlet in both cases +We have our own validator for EL in seam projects +Added some comment to the RB +There have been no comments so going with my view of WONTFIX on this +I wont commit anything until I hear from you +I think having more information available to the user would be a good thing +Ari I had looked at the class that you defined in +All he did was patch his version of source code and rebuilt it +I also backported one of more extensive tests for to +I will add the new examples to directory bpelsamples open issues to CR +Oh okay +pushed +It seems to me that porting from pystress wasnt such a good idea needs +In fact I found that the behavior is a race condition when the strategy is used on both the subresolvers in the chain +Note that this patch doesnt attempt to remove the hardcoding of the protocol version to +patch with all files passing fine now. +Proposed patch applied +Looks like this issue is a duplication ofAs for solution IMHO we can solve the original issue by creating appropriate files without touching the code at all +The warning is you attach more log +The inherits from and should not return a +There is still more to do but I want to get this big patch in sooner rather than later so as to continue with more changes +This looks like a dup of +Otherwise dont whine about what time other VOLUNTEERS do or do not have +Im pretty new to tags so Im obviously missing something +We have already fixed similar bug +And if the attempts of a tip failed four times due to such factors that will kill the tip thereby kill the job +ChinthakaHow did you start the serverI mean did you start the server running the shell script in the distribution or did you start it using an IDEDoes anyone else having this problem because it works fine when I checked the distribution while agoI just checked POJO client and it worked fine for me +Committed thanks for the help guys +jmacgill my first shot was to go new layerjmacgillthen I found I needed a map so I went new mapjmacgillthen I found I needed a project so I went nwe projectjmacgillthen did the mapjmacgillthen went for a layer and was stuck wehn the new layer dialog asked me to pick a layerjgarnett Sounds like quite the naming problem there +Good catch +This was fixed before +jj file +As serious as this problem appears to be it is too risky at this point to land a fix for it in + +Can I ask you which part is unsuitable for? I could not get it. +Please close this issue if this fixes your requirement +Which version of design is currently full approved? to have consistent look but I agree with other people that looks better +This could be a new parameter of the Sonar core plugin +Go ahead +numbers file with Charts embedded +Assigning to Jonathan so he can apply to CP and then close on both branches +I committed this +If someone feels it is really necessary I can change parts of this to use dynamic buffers +Based on JZs comments Im closing this as Wont Fix +I had to revert the new test for now. +Since it is handled differently by the new framework +Attached final in revision +The problem of this approach is that when I insert a single string wrapped in a DCT as column name after it is no possible to retrieve it +Can you generate a patch for the line +Tested +Fixed in r +If we go down this path it makes sense to our constraint +ps the current patch works with the iphone os client library although it always uses the canonical user as the viewer because the sample shindig doesnt have real auth +its not working anyway +Roo with commit cadceffeffcfbd. +IMHO Im not sure how this can work with all providers +Pploading a diff to that includes a fixture with the problematic query in this issue +These patch files are for Codegen plugin and Service archive plugin +This patch converts source and target levels and to + was a similar issue +Cant think of anything else to test currently +Will respond soon with some comments +Nether option is a great way to handle missing data but do we want to force one approach on the user? Is there way to increment the full vector of values and account for missing data on one or more variablesThanks could allow in the input vectors and skip updating the bivariate covariances for pairs including a +Confirmed this works OK now. +What we dont handle are things that dont extend Exception such as Error +Thanks for help Suraj and Apurv +I am still showing it as failingI dont think the whitelist changes are in yet +Setting the fix to as this related the modularity of Groovy +Thanks +But the code is case sensitive so APPLICATIONXML would be an additional entry to applicationxml +The patch removes the only currently known problem related to replication of encrypted databasesA separate jira has been created for testing replication of encrypted databases so this patch does not include any test cases +The trick is keeping the markup and code separate as far as is possible +Here is the diff for the patch version +Without that change tests were failing when run against jar files +Fixed in revision +Library names changed etc +java file rather than the eclipse project view so I have no way of knowing whether it was the checkout or eclipse that was screwed up +Damn it +Thanks Zheng! +Thanks Nigel! + applies only to the branch applies only to the trunk +Im attaching a fixed patch +I need to know the Unix group to request Mark Andrew Davidson new account +for turning it off in the release tarball good idea +So Im ignoring the formatting issue since that has been fixedHowe the returned null is a legit bug +Andrew patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I believe the gold file was never updated for this +This is worth investigating before we release +as the problem of the miaximized dialog is still not recevied any updates from Koen assume it will be fixed do you have more precise info +Well As Olivier said I think that its better to keep this dashboard devlopment oriented and think about another plugin for more quality ISO and report we want to offer this kind of functionality we dont just need to store full data in a database but also manage refactoring what to do if packageclassmethod moved or deleted have a way to design a quality model thats to say a mapping between ISO and metrics from PMD and co provide a way to aggregate metrics data for a packageclassmethod or ISO category +Assume closed as resolved and released +Thanks to Brock for the review. +Hi Jodythis patch ports the trick over to the it does not make the as fast as the shapefile one it improves things considerably for the case in which a lot of features end up sitting in the same pixel and does so regardless of the datastore that backs the you try it out with uDig and provide some feedback +Do we have a Jira to track SQL injection issue +with minor typo fixes sthenthan and sgreathergreater +BTW all I filed to fix this inefficiency. +Im not getting it in so possibly it has been fixed already +others have cleaned it up alreadythanksdims +The component attribute map would evaluate before the CC attribute EL resolver could +That works too +Let me know if you need me to do any more changes +We should be the jar and complete jar files before we release +Why do we need to add the queryConverter definition outside of the speallcheck search component? Is it going to be used by any other component other than this +I think these might be destabilizing changes +Youre right! There was an escaped slash +Hallo Tonythanks for looking into this again Tony +or am i not seeing what you actually want? +I have only added minor edits +get has a new signature a +I also wish Counters could be a part of normal only so that we didnt need to do another separate query to fetch the counters data +In there was only which was actually in Core +This jira removes unnecessary listing at the client side +Reverted changes until further consultation with Mircea. +I dont think so since this was encountered on problem here is that node A removes the fat client but node Bs timeout hasnt been exceeded yet so it still has it +Offending machine passes time test now +would require the following changes Make the httpClient protected instead of private or provide a getter method for the member variable Create an interface for the that usesI like having use a interface i think thats a cleaner implementation then whats there today +This is caused by the schema definition +Fixed +Possibly but it is necesseary to check when spaces are in the directory name. +mistakenly Im going to get started on the content spec for this guide +Do you still need a patch +ConfusingIt also looks like the method wants a more plain version format like Any help is appreciated here +Ill try to squeeze in some time this weekend as Im starting another project soon that will use. +This is a bug in and it will be solved if you decide to upgrade the minimal JDK supported by Ivy +Committed to trunk with revision The test is still disabledLeaving patch available flag set Ill look into getting the other test committed as well +ser mvn coberturacoberturaIf yes I would want to ask for input from others on the mojo dev list +scratch that +temporary fix +And HDFS as well +I request a programmatic way to do it via the connection parametersAnthonyConsider implementing a custom and initialize the socket as you see fitOleg +contains class developed byMichael +If anyone has a pointer Id apreciate it. +Rebase for trunk +Looking forward for something to commit Re If we set an expiration time on the cookie it is not a session cookie any more +Would it be possible to pass Hints in some subclass of? If the actual is not an instance of this subclass null hints would be assumed +Resolving this +We should allow modifying system table contents but not changing the schema +I just checked in a simple worker monitor API and a wicket project that lets you see what is happeningIm not totally convinced with the API but it is a start. +I believe we still need a note to the effect that it is public domain +The patch doesnt change the default it only allows it to be changed +FYI the release is scheduled for late AlexSorry for the late reply Im attending Javapolis this week but once I return hope I plan to address the issue is on top of my guys! Ill continue my research in the meantime then +However there is definitely still something wrong +Thanks HenryCamilleRoss! +It works fineThanks +I guess I have to test different drivers once again as I did some years ago +provides the total load of the cluster +Brian thanks for your offer +The sanity checks are pretty important +Reduced priority of non critical issues which have current is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +I think I need to add an alterindex function to the which I think requires editing the thrift filesIm not sure if that is the correct way to go about that +to and trunkThanks to Lars and Anoop for the review. +This is a good idea but you could accomplish this yourself +Iwant to do a Drools beta in the next day or two which means Im going toneed JCI snapshot on ibiblio and for that I needed a committed Factory of somesort +sounds complicated now +If vertices and its objects overhead are calculable thats acceptable to me +I was thinking along the lines that the same script multiple could only be called multiple times if the corresponding variable is not already used in the same Screen +The tool I am using is developed by another team and I dont have the source code +thanks Uma for the patch. +Committed revision . +Adding attachment +Just a quick question before I take a more detailed look +The documentation can be fixed anytime but this issue of putting into the distribution cant be resolved until is resolved to make build and test able +The issue is that if your function takes sometimes two and sometimes three parameters and has get it is likely to result in an error +Ok so i wrote both as optional features +Thanks Daryn and Todd! +Yep thats better + +Added pull request for the comment pull request with the docbook part explaining the user roles format +And that guys got SQS too +I just committed this +As always I admit that my writing may leave much more to be desired +Attachment has been added with description potential has been added with description revised bug has no FO test input file on which to perform a test on current trunk the obsoleted Simplified failing fo file is not usable due to failing for other reasons please submit an input test file to obtain any further action on this bug otherwise it will be moved to resolvedwontfix alternatively and better close this bug and file a new bug based on the current state of trunk providing new patch and input test P open bugs to P pending further priority for bugs with a patch +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Now that work is underway with AS all previous community releases are +I assume the configuration boolean would be something like +It looks like you must have a +Scanners already take filters +Attachment has been added with description The orignal fo problem was created in revision +Jason would you be interested in coordinating this? My thought was to ask some of the seam module leads to write one or two to show off what they thought was an interesting problem +Can you give me some suggestions? Setting to zero does increase the latency +I can understand +Sample data to correctly display Accounting organization and parties in HR plus cant developp tree whatever relation is given to party for departement +Here is a patch with the dependency added +This method has been removed and this error should not occur in ER +Please check that the bug is fixed for you +Heres the patch that I described when I opened the Jira +Changes made to are also ignored for the same reason +It seems folks would like to make this change so that it can be included in +Why not this purpose +Hi Martinyes sorry +Cant we just spawn a thread to flush? Thats how i was doing it at the beginning in my prototype +It was a problem with running the web server from Eclipse +change status from In Progress into Submit Patch +Thanks for adding the comment to the code as this is a tricky subject and the comment helps considerably here +In the capacity scheduler case there are some contents in the default file as well which changed in +They would then presumably be able to run the mvn execexec with a parameter for their custom class +Thanks Thomas for having a look into this issue +Retargeting to +Im not sure what you exactly mean I think it would be fine if the installer would still work with JCR I see atm no need to tie it to JCR clean up the Ive created an interface and two implementations in revision In revision I changed this back again to a single implementation Ive added an enum type for the various resource types to Next will be a detection mechanism for the resource type this is currently a little bit spread around +Maybe providing a character encoding to the filter could enable it to compute the needed byte offsets +At the moment we use our own plugin for this +Forgot to add can we get this patched on trunk as well +fixing test case failure and increasing test case timeout. +i made the following changes to the original code add generics and other java things add junit tests add polish analyzer with stopwords list from carrot repackagingyou will need to extract the stemming table into resources folder right where the Polish stopwords are +Thanks for the +Committed w light modifications +Thats strange but maybe we fixed it in master and missedCan you submit a patch for it so we check what happened +for the trunk and patches +Fix in trunk r handle virtual repos in webdav browse add new for virtual reposLeaving this issue open until other people aside from me are able to test this further +Rayees How to rerun this to get a db dump +Note that is not currently configured to use HTTPS so browser like HTTPS Everywhere may cause the download links to stop working +Hello is this project still available to take up at Google summer of code +That was more one of those things like if its really easy otherwise well put it in a future JIRA IMOBut yeah I was just thinking of doing it in the browser before uploading +Fixed on master +Damn Java +otherwise good set of null column values is a bug not something to be true +Clean out of +Text extracted with the current version to resolved +I just committed this to trunk and +Note to Self +Patch committed to trunk. +Patch committed +For the first time I can get the mdb program to be deployed onto either or server from within eclipse. +Much of these are supposed to be fixed in but a quick test did not show a significant improvement on my machineThere may be some other bottleneck though that Im not seeing +hadoop wiki is also unbearably slow pretty much always +Corrected some bugs in the new code when testing in a cluster +Tests includes assignment to raw types lower and upper bounded types. +But it might sometimes have an unfortunate side effect to do with logs not fully flushed yet +I have en and US so htat shouldnt make any difference +Not sure it can work as hibernate user types match one row and localized string matches several rows +Right now we cant show Code Assistant for components from taglib but we can show one for attributes of that component +I only briefly scanned the patch but Ive seen Ted mention LLR several times now in relation to several different things +looks good +After that current code of jdbc driver should just workIf we have a synchronous api the clients such as jdbc can fetch results after the execute immediately without bombarding the server with so many calls +FYI when I remove the explicit dependency on SD Commons Core and run mvn dependencylist on my project I see jarcompile in the output. +There is no bit windows for Eclipse that includes xulrunner thus during installation where we need to call out to eclipse to prepare p metadata it is using the bit eclipse binaries so yeah use bit jvm to do this +I certainly dont remember seeing this issue with Oracle deploys but Ive also stopped installing over a year ago +I also prefer method attributes but I first have to look at the impacts on design where I have not taken concurrency into account also think shipping a release soon would be a good idea +where is the stored? is it a flat file? where is the file locatedis it versioned? +Patch looks good to me +The source now supports tapestry versions from to and +Patch was commited to trunkwhere developping +Ill fix it in the next patch +Do you mean that some of the last bits may be set in the newly allocated object and VM clears them? What is the value of objinfo after allocation is done? Are the higher bits clear? If yes then clearing them is not required +Seems to be fixed +pushed to and I just realized that there is still time to get this in so I added it to fix versions +Lars the fellas up on IRC said they needed your v applied so I went ahead and did it +Im not sure if you saw the inUse in the file data store what do you think about itAlso I like to make sure that the database data store can to the database automatically if possible the feature of the bundle db persistence manager should be usedAnyway Im working on it but it will take some more timeThomas +Keith can you review pleaseThanksRobbie +Wow that was quick Erik +updated patch for timeouts to trunk the patches since now needs a separate version +Another example is AJ +Its even more related to the core functionality than the well known DIH issues like +bq +How so? Maybe you mean security requires the native libs? Yes you are right +However Here is the spec which states Scope can be specified on either interface or impl +Am I missing something +You may cover the whole screen with anything you need in this case. +Created patch that exposes the hidden variables +that simply address this issueLahiru +reopened for fixVersion +Yup that all seems to be working thanks! +Thank you for this bug report +This is a duplicate of of. +Moving this out to please move back if you will work on this any time soon +The component in Camel onwards support async routing engine +lets see if it passes +Integration classes are moved to test folder and dependency set as test scope it works fine using the jar from repoIf trying to use in compile scope it references +Attaching +Is it planed to update the site plugin with the new version that supports the poweredby tag? +This only happens if the FIRST PDF is version! When the first PDF has version and the second has this problem does not occur +Will get back to you. +Here are the and the localshared settings file for my based repository +Patch Committed revision Thanks Anurag +Thanks for getting to this so quickly! +Also note that PARTIALSTATESAVING setting changes the behaviour which is always suspicious +for the patch +Would be nice if didnt have public API but probably hard to do +You should reopen the old issue instead of creating +This is because there is no really sensible natural way to compare two maps or sets +Which command may be absent it is better to use type instead as it is build into bash +Rams Comments +This brings up the issue that has been mystifying me about the recent move of dependency info into the dependencyManagment area of the root +Cool +same fix test case that fails without the fix +Thanks Alexei +Tests failed because of wrong Dds +STEP srcqpidmanagement Apply the patch inside QPIDCPPROOTsrcqpidmanagement +Well track the current status in +After doing this the previous two scenarios that failed to recover now work +The fix doesnt seem to solve the problem +Part of that processing involves producing hidden markers embedded in the html for PMC Chairs and PMC members +It tests both cases only the lastmodified updated and content overwritten with exactly the same amount of bytes +have a directory containing a file + +Regression tests pass +Ive downloaded them and both jars are the same +Thanks +Can someone please let me know on thisRegards James +Hi Rao I have not been able to do much further work on this +Thank Tomasz for the patch and Nicholas for the review +Moved to HADOOP as it targets all daemons +also refactored to combine the with the authorization visitor and unconditionally run the visitor +Although I do know of one user who only utilized Local Mode +Agree with Stack still flaky will look into it more later. +bulk close of all resolved issues prior to release. +closed +First version of code +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Patch to fix output escapingThis replaces with a method that escapes not only double quotes but also backslashes and newlines +have some templates that have arbitrary recursion depththat may break if a fixed recursion limit is enforced +Could you also change unit tests to use it +added the necessary files so that the Eclipse plugin can now be built with Maven . +Can you regenerate the patch? I am getting some conflicts to previous according to Namits comments +on this being a blocker for We have seen this fairly often but it is not reproducable +Moved to can you take care of this since youre in the deployment code anywayThanks +Working directory can now be specified through the default value obviously remains +Genericizing your example I think Hennings solution if we hit a security problem try the interface is the way to go +testcase for input +Will add the error message and a negative test case +is the default tool setting with keytext and could not work correctly with Cant find resource for bundle resources key zhCN. +Commited to in this was committed as not +Only if adding the character x is done before then it shouldnt be a character that is encoded +Updating the patch with vinods commentsThanks +What do you thinkIm importing serdeConstants for the get method which converts Thrift types to Hive types +Reproduced in sometimes fails to update class path in ejb project +I will do it in a few days Im overworked +This is a known limitation of dotNet plugins so Ill let Alexandre to continue discussion in this thread +Not getting done for +Fixed in on +Any idea how http compares to binary performance wise +Refactoring usually implies a fundamental changes whereas this patch primarily creates an interface for methods in objects +Moving out some bugs to +Closing. +I think this issue is too generic to be doable most external widgets already provide their own ability to be incorporated in websites including blogs. +See test screenshoots on how to reproduce this problem +Note I dont seem to have permissions to close someone on Incubator or Etch karma should do that as OK +This happened with a data dir I was given I have to try and reproduce with a standard one +you cannot and update w a given uuid on multiple replicas +Is Slava right or +No i have started using the pattern attribute in my retrieve target +Linking to the related issue +Please make much smaller patches. +However if the UPDATE touches columns which arent mentioned by the trigger those columns turn up in the row passed to the action +Marking as closed. +Is this still the case +If the time is not tight I wish that we could include this improvment inAbout the menu location I think we could create a new menu itme in the navigation tree lets call it alias configuration and place it in the Server categoryAny comment? Thanks +On reflection the change to in the patch may be a bad one since the fix to means you can OK the page with a file in the user settings widget it may be intended behaviour to default that value to the standard location for the user even if it does not yet exist so that it will be picked up if the user ever does create the file without them needing to change anything in Eclipse +Server does not go on to OOME as weve seen in the past +Then we could progressively use it in the FTL files but for this I can wait for your work doneAre you OK with this +Please continue this thread there not here +fixed in beta and dont know why this fix is not included in betaa according to the dates plugin was built on and the PR was closed on +The original problem reported in the original test case is fixed in +Ill make an update here +although bit flag is instead of +Yes I was about to write a comment about it +modulescommonssrcmainjavaorgapacheairavataregistryapiuser +patch looks good thanks Uwe +Was this ever resolved? +r may fix this according to Fred Toussi +Moving to CMS would be a very good step +This bug is closely related +The point of the patchthis issue was just missingSince I can edit the description of this issue I have moved the quote of your mail into itDid you have a look at the XA stuff in ctx thats already there? I know its not much just the mapsI also worked on but didnt find the time to complete it recently +I therefore checked in a different fix +It could be a duplicate of which been fixed after release +Roo currently treats all DB configurations the exact same consistent way I am just not sure if we should change that if there is no real this ticket referring to the argument above +Why not just have the shell do a list and then send relevant removes list takes an optional path argument so that the client doesnt have to be bothered with seeing results that dont pertain to the given pathWe got rid of the batch operations because they were a big headache in general +This issue was triaged on July patch looks good +As from Roo future versions of STS will need to use the new Parsercomplete API +Fixed the trailing meant the line length +s is will then use this method to return its value and will not use the saved sessionMode string +Thanks Junping and sorry for the noise! I reverted so this is no longer a problem well fix up this test as part of the next rev of that patch +Can we close this jira now? Since has provided a way to exclude jars from scan +I would say we should open a separate issue for the existing binary protocol accelerated and then just not do it when we decide to commit +Related to incorrect merge resolution done by commit baabdafabc for network ACL feature +Hi Werner! Did this already +Please close if it works for you +Sent a note to to solicit more opinions on this issue +The bug exists for foreach also not just ateach +The issue manifests pretty quickly if you use RAM directory and exclusively pull NRT readers +Is there any work around for this? This is blocking me from using the release plugin + +No progress on this yet +appled the patch to the trunk +Status is fixedclosed +Basically it is providing metadata to the clientIll look through the code in the next week +Heres a patch that removes tabs for moves former to the nav menu adds a related project listing to content of the front page updates the list of PMC members adds redirects for former subprojects to their new looks good except that we need to wait for their new web site to be up before we redirect their current url to it +Apply the patch I verified the problem has been resolved +You can see that for directly in the fix version while is a subtask of an issue with a fix version for Groovy +Patch implements the interactive shell +This gives immediate visual feedback showing if the right settings have been chosenThis means that when the user selects Audio only or No audiovideo the video feed should not be shown +Its a severe data integrity issue +The patch moves the action of writing out into a new write method +So unless you find a reproducable example in pure Groovy I fear it may be the interaction of Grails which may be leading to such an infinite loop +If no change have been made no date will be displayed +closing resolved issue for release +Ive filed with the details. +But code wise the implementation is work Robin +The type of the component should not be determined by the type of its owner +Forgot about that +The JVM locks the jar files when the installation verification is executed +to everything but the changes +Thanks Knut +Sorry the correct project is this one +Indeed it would make sense to put these checks in place on a higher level such as the new related infrastructure. +I use Studio +Also Jacobi rotation and are probably interesting on their own +I am indeed curious about performance! The current policy provider does not process OS policies but only DFS policies thoughIn case the performance impact of these mechanisms is unacceptable we can easily bypass them +Then yeah I think you implemented this correctly +Maybe we should handle this case special go find the latest and add the delete with that timestampOther thoughts we cant have that takes a Get object because there already is a get in +It has been awhile any luck reproducing this after taking the fixes for and +The limitation is very understandable +Nigel thanks for pointing out +andInterceptor bindings may be used to associate interceptors with any managed bean that is not itself an interceptor or decorator or with any EJB session or bean +Thank you for the review +Yes you are rightwe cant increase the limit to close to regions +perhaps its a good one for us to incorporate into CI since it seems to hit all the fun eval edge casesIll look into and try to get testsspecs in place to keep both of these working. +Then Ill check it in and do extensive testing with + are entries that really make me wonder about the home life of James Linda. +You are right +Did you modify the file before or after starting the server +i definately want to update this in the filesand will post very shortly +I will commit a bug fix for this just committed a fix. +Attachment has been added with description Example fo file a table with has been added with description Patch that fixes the problem for me +My exact configuration is in the initial bug report if you need more informationThanks +Thats the same thing we do for table functions and it would prevent conflicts with future standardization in this areaAs to item above Id suggest that we go for option a in the first iteration and revisit it when we have more experience with how vararg functions are used +Hi ColinThanks for taking this forward +Ok thats not a hard limit because the user might use a more efficient encoding in terms of space but that should be fine IMO. +has been all day +Patch looks goodDo all tests pass +Karel any update here +part of version +Recreate the patch based on the latest patch looks good to me but I have two minor comments +Ok its more clear now +Including a long debate in the costs obviously reduces the number of features we would choose to backport so I think its important that we trust our committers to make such a decision and delegate that responsibility. +This doesnt happen with SP all senders complete with ca K messages +Andrew I wasnt sure if you could reproduce this +is there so that on get call we return the +Scheduling of a timer is delayed until commit time when a transaction is solves the problem of receiving a timeout while still in the create this change a secondary issue is what happens when scheduling a timeout +The settings will be savedloaded to the project metadata +This way a choice can be made between consistency and performance +Hi SvenI need much more time to fix the regression +thinking i may miss something +Id love to use this functionality I sadly assumed it was already present in Spring Social! Is there any indication of when this may be implemented +Karaf makes my brain hurt +It throws an exception with a meningfull message now +Taking this from Daniel as discussed with him +Could you state that constant language is in Camel +I just committed this +Committed at subversion revision When the Reference Guide is built it is assumed that the source for the Developers Guide sits in a directory right next to the Reference Guide source +Manually tested! +Also note that with JDK update the whole test wont work due to changes in the internal implementationI agree that the two test classes are just a hack +So its not super easy to implementHowever its most likely not on the timeframe as we got major goals with and Async stuff to work with. +Confirmed in and using mxApplication error occurs top is null +Hibernate version number corrected toCommitted in SVN revision Next step complete one more JIRA relating to this guide then Brew it and get QE to confirm that this issue and the others have been resolved +I dont think the potential danger of this optimization is worth the chance of regression and I dont think RINGDELAY is something people consider a pain point. +I think we are pretty much done with this issue alreadyBefore closing this Id like to move the new classes from within to appropriate subpackages in + accepted property name +Is there any movement on this? Was this issue folded in to a later release? +Yes please I think it should go in both +I added a after each row and found it to be almost three times as slow as the run without the free ms +Please set the requested fields so that the issue can be properly triaged or it will be rejected due to lack of information +bulk move of all unresolved issues from to +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +I will make new generic iterator with abstract method +Great point Alejandro +I am looking +Its safe +Unfortunately because of technical reason it is not easy to do without big changes and this is why this solution was when the current error message is logged we know that the language is not supported but we cant get the list of all supported languages +Trying to get it running inside Maven will result in putting some Ant or Groovy scripts in the dont see any benefit in integrating it in Tycho or a +Patch attached +This patch adds a check that resets connectindex at the only point where connectindex gets values +The counter update path is almost as fast as normal column update pathI have no idea about the detail of the removal before incrdecr problem +For any existing file in the subtree the NN would need to migrate blocks from one storage tier to another +The results look good so far nice catch +because policy is not yet resolved +Has there been any movement? Perhaps I need to take a better look at +I am closing this because I do not think we have anything else to do for the original report +What do you think +Which JDK version +Committed revision +It was nesessary not only consider the methods +Patch committed made two minor changes Moved the class comment to before the class definition so it is a javadoc comment for the class renamed the suite to match the test nameThanks Manjula +Ah maybe this is unnecessary after moving to UNICAST +Perhaps the best thing to do would be to add a getter for the on so we can easily reuse it if all we have is the +instanceof is one of the operators directly implemented in the Java Bytecode as a separate instruction +I just committed this +However the important thing is that the debian directory should not be shipped with the release tarball but in a separate diff +But I see the same list of input methods as its described in the current version of the guide +From reading the comments in it looks like clean will be removed in CR am I right +Grid can you look into this and see if you cana Duplicate it using JBDS and document your resultsb Duplicate it using the latest JBDS build and document your resultsc See if you can figure out why its throwing the errors in the log and provide a patch for allI reproduced this issue +meantime keep original jsp and servlet files for performance test script work out. +Updated dependenciesR handles arff as wellYes +Sure heres a patch with the name as you suggested +I agree that that synchronization is unnecessary +So under Jruby the is basically in a race to finish before the main thread gets scheduled again and exits +I am reopening this since we are still getting failings in the interchange test +Have installed version to replace RC +Anita here is my patch for +Thanks Devaraj! +Applied the lasted patch. +I think r is the prime culprit +rejecting for to be proposed for will remove the absolute path then +We can use We already use it in some of code +What is the motivating force here? Are there specific updates we need to pull in? Keep in mind that the implementation we have is working and tested under the hood and upgrades for upgrades sake can lead to incompatible runtime changes across point releases +Testing comment +Thx +Please test this since I dont have a SAPDB install +That is the executes these methods only once using retry cache functionality +as afaffbdcededdbff to +I guess I feel like users shouldbe able to do this if they so desire since this limitation is not necessary +I found it useful while troubleshooting since that marks the time we received the request from the socket +Some test changes some setting without reverting it back afterwards +Logged In YES useridIve just downloaded the Release but the Bugstill remains +Better sort stuff adds a new doc maker that adds a random sort field over a specified range of values allows choosing of sort type includes sample sort benchmark algorithm +So what can be done +Added formatter ruby module under bin +Alternatively we could have apackage javadoc in mimej referring to these classes ideally withsome example codeThis is a TODO contributions accepted +Patch applied at +Reassigned to Sean Rogers as I had left the project +I dont understandIs the issue already in the coreRegards Vaughn +Or do we want to have all for also available for and additionally another CP for subcomponents? I dont think that would make another approach Create only for which would be also applicable for +Release notes were updated during CP. +So closing this as duplicate. +It has nothing to do with +But that can come as cleanupit doesnt need to block commit of the current patchesNo other comments to make thanks again Thomas +and made classes in there use DOMAIN constant in nethyperichqproductservletfilter to refer to +A comment on the general approach +still needs javadoc +to trunk +I refactored the tests so less state kept between testsI also added Map of failover regions w HRI as value so if present we dont have to go to +The can be simplified heavily when there is no coordination needed +I was not aware I could specify the version to be used by RampartCould you instruct me as to where I can specify this +a few are now available and run on hudson +Some features can be added to easily +I regenerated the patch and submitted please check +Were currently at version. + file to change the from the instances directory. +Reproduced when HTML page without in the of a bulk update all resolutions changed to done please review history to original resolution type +Committed revision . +Here is my technical reasonJust pointing to two non concrete comments about whether this should be a Lucene feature way up in the issue is not sufficient +This test fails again +Thanks Jun +Attaching patch that implements this with unit test +Server started with fle not set or set to false the server listens only on one IP +Patch that implements the helper and trivial tests +Max if should change something there please reopen it +Source for the Request Processors +Committed by Mike on in r +moving to becomes GA so for now no issues to be put there +Thanks again +Will be included in CR and Final +resourceswlm resources is a the special path for Wicket resources in +As I understand the bug Evicting the proxy for the superclass entity prior to executing polymorphic query will NOT solve the problem for sure +I close this bug as the fix is now in created as a clone for of this bug +Changing the summary to fit with the requirementThe requirement is fine we might be able to find a good way between and Java +but not documented +Thanks thats pretty clearIll submit it to the team as well as the bug is in their source +That reveals to me it is more than just idea what change on luna that causes this to suddenly to fail +The way such conditions are handled right now is correct and conforms to the FIX specifications +sorry meant committed + for the patch +The logic is a better fit in the engine in any case since it already has facilities for this logic based upon dependent joins however it will not be directly related to the exact length of the salesforce query +I can do it in batch if you supply the appropriate sizes. +Yes they can add that +Thank you ChengPatch a makes the test report the seed when it fails in an assert +Once they tab off the message appears +If First is intended to be a wild card it should be +Would you please check the mailing list messages and let us know what you think +This was exactly what this fix attempted to correct and it even added test cases for doing exactly this +The trunk which will be the source for does contain projects files +yes something like that +generic generators test case previously attached to got it to work but the code is pretty ugly +I then left all consumers over night for over hours +It could look a bit complicated coz we have also abstraction model in levels to enable generalization of cppProject for linuxwindows and staticdynamic librariesexecutablesWould you like to take a look at the project? +I will reassign this issue for verification when the docs are available for review on the +Thanks a lot for your reviewsDaryn will redo the patch according to your suggestions +See not apply to onwards due to a new +You dont want to dump a ton of finely polished work that has been fully documented and then find the community has other intentionsEven a patch that does not compile has a lot of value! Just specify its early state and limitations but it lets us know the general direction you are taking allows early feedback and helps any committers digest the work +I did some cleanup +Found another case where this is needed when building the that often includes a which packages things which are not eclipse plugins into one so any of these need this removal as well +Moving to I am closing this oneFeel free to create new tickets in the future for any work on sorry I though this was the initial ticket for Obvious this is for using there any demand for this? People can use for close old ticket that are not in demand. +So for instance on an index on columns ABC there are actually cardinalities calculated AABABCI agree that the calculation of cardinality of ABC couldshould be short circuited for a unique index +Attached classes to demonstrate the issueSee comments in for details of different behaviour between for added these checks along with some additional similar ones. +Its too specific for Langsdomain of improvements +Aggregating disparate counters in a shared variable only to push its deltas to another aggregator is not an expected approach +Considering this issue as closed. +However we still get failures related to the correctlySpelled flag and with the use of alternative +Mike you had it right +Pushed to master +If you are inside a MTQ and you want to do the rewrite caching or other things you can do this based on the IS +Given that I think that a functional spec would consist of describing the delta from those rules needed to support varargs +can you git pull? I see this as already fixed + +an image next to it +We are using the releaseHelp +Not sure if these work with the mapping code +thanks Ivan +Ive just reproduce on Cordova from +BrandonCan we make the new commands case insensitive? We can log a different jira to make existing commands also case insensitive +Need to resolve that issue too +Perfect closing this one lets handle it via github issue. +I agree the log message should be more clear when an artifat is not retrieved +Thanks AdamI will then certainly investigate on the APAR issue +I just committed this +AS boot time is again acceptable with this fixother boot time improvements will be opened in separate JIRA! Whats the Fix Version? +Closing the ticket out +Patch that fixes the twiddle andre +Are you planning on having the AM registerunregister with the RM for diagnostics and exit with proper job history as part of the fix +There is some chances that this issue is fixed in geotidy +Documented in the EAP release notes. +preference +just trigger the Ill take care of it +I can create one if necessary +Ive attached all logsThis a single node cluster +Ill prepare a new patch +Just assigning the JIRA according to the component +The patch went stale +So has a pom has a pom etc +Dennis the issue looks more or less correct aside from the anomalous behaviour with the UDP JMX bean described should we close and open up a new one for the anomalous behaviour +surefire seems to work in last build from was unable to reproduce this failure in the past withI couldnt start Tycho build locally to check itCurrently threre is no such a failure on Jenkins neither in trunk tests nor in M branch +Closed because it is irrelevant for Spring core + unscheduled issues to the next release + is an oldie not not a goodie +Full rebuilds are done with DIH because my application cant touch it for indexing speed +And removed the line skinpart comment in the asdoc +most definitely not a waiting for a test case +Would anyone else like to weigh in so a decision can be made +Going to need a test case reproducing this +Seems that this is not addressed as a bug +Need to test this in like a test +Include patch for and upload as patch +I just need to get clarification on how to setting up the correct public repos would be the way to go +Looks good +Thanks! +Did grep on complete source code +Added remote +Please attach and Im sure we can find a way to commit to trunk +The state of this bug report is a bit confusing since it contains multiple reports of issues with similar symptoms +Mat is right xml that you depend on does not exist in the strange because Ive depending on that file for years and nobodys reported a problem +Please close again once you have assigned a Fix field +Why use an empty exception to record the time the renewal thread was started instead of a timestamp? Otherwise the patch looks great +Punting to post RC +Kindly share the output of and system tests +See the installation instructions +This was actually due to a bug in not SheetPrompt +Might as well as I had to touch the proto files +There is a know problem in this patch File upload gives write access error when Ofbiz is run using Derby database +We could allow the user to define a macro to enable this thread checker code also +You need to have an existing maven projects +Hi folks any interest on commiting this patch +Heiko should this be resolved against +ok confirmed by marshall and the process is like thisCreate is actually deployrun as +The table where we were seeing these slow scans was definitely a tall narrow table +I dont see any need for iptables and httpd redirects with ssh gD +Added documentation in the Reference Guide for the Teiid VDB sequencer +Will attack Oracle myself if else grabs this +This required some changes to make it work such as shutting down the Jetty server to make sure it released it port between tests and changing where the admin system properties are read so they are refreshed when a new server is createdRebased on trunk and addressed review comments +r sobryan linesChanged paths M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadinternalagent Add IE trinidad agent support Thanks for the patch Gary +Is there any way for you to reproduce the scenario that showed the issue in order to verify that it is fixed now? If not thats fine too this change is definitely an improvement +Changes looks good for meHarikrishna +I had forgotten that attributes are optional by default +We can close it. +Thanks Sijie +I am running regression tests nowTouches the following filesM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceRemoved messagesM javaengineorgapachederbyimplsqlexecutertsM javaengineorgapachederbyimplsqlexecutertsM javaengineorgapachederbyimplsqlexecutertsM javaengineorgapachederbyimplsqlexecutertsM javaengineorgapachederbyimplsqlexecutertsReplaced fragments with hardcoded SQL object names +Patch do prevent invalid repository object to remain in the s for the patch! I committed a slightly modified version to Jackrabbit trunk in revision and merged the change to the branch in revision . +Yes I think that my perspective clearly +Patch committed to trunk +As most agile places will have week iterations days will help us understand what state the code base is in after an iterationthanks +If you still see any problem please. +I am running it on Windows +Put another way the value is pulled from the body content only if there is no value attribute +Any following discussion should better be done in the user forum +Please put a comment explaining why this is disabled and whats the issue you saw in case we explore this route in the future +Well likely need a testcase for this +Ivan Szanto at find the attached patch that makes it work with and still works with will you put this patch in the final version +Aaron Pestel has experienced this as well +Which use case does this target +I agree with Mark. +Apache Directory Studio has now been releasedLets clean Jira and close this issue. +This patch is identical to the test patch submitted in +Example of what we can do with previously posted patch +This is part of Bens patch submitted to +David asked me to assign it to him +Test Executed on jdk rc environment and found no error +Patch for RC release is attached here +bulk assign all unscheduled issues to next release +That was imho not really needed because the name of the data handler may change but the file name of the attachment must stay the sameLars +Thanks Stu +Im nearly finished here +Yes I meant Beams confirmed the problem when using the above +Let me know if youd like it and Ill email it to you +My personal hope is that someone will recognise that this interface has caused and is continuing to cause problems with every release of Axis and that it will become a test interface thereby improving the quality of Axis for all users +Fixing index limitation problem +A relatively untested patch +Im doing some work in this area +We probably need to add a unit test that writes and reads a file with larger blocks currently the tests all use small block sizes +Usingrefactoring over time I think we need to remove as much duplication aspossible +The patch merges changes in vmvmcoreinclude +This is related to which Ive punted to +I would request you guys to provide an initial feedback while I proceed with the code cleanup +Thanks Sumit Pandit +It seems what you asking is an IDE feature right +I have done all where I was able to do something +since tPH is a plain there is no relationship between the updates and the reads +This has been superseded by support in Spring as well as Hibernates event listener mechanism +Add a unit test to verify that we catch this case +Maybe the PDF file uses a specific font that is embedded in the file an maps to images of karakters instead of ASCI or Unicode characters +Verified in ER +It seems that it is an old issue that was not discovered yet neither by a tests nor by manual testings +Same issue for me dfeist +Can you put together a little test project that can be used to highlight the problems you are having +There is a method to slove this problem +It will make our life easier we track site instead of +The concurrency bug will be handled in separate and +Proposed addition to Category X sectionThere may be situations in which works under these licenses may be distributed in an Apache project +Fixed tests again you need to start the filePattern with a if it is absolute +Thanks +Thanks Regis +Ive modified the patch and have resubmitted it +Ill add a WSDL as an attachment in a second but if you do the following with a instance you should be able to repro it locally +The problem is that to the casual observer it would look like these tests passed in CI when actually they where skipped +I took a glance looks good to me +Thats why I commented out Ok now I see the problem +Another step closer to bring to java clustered worldBTW +Per discussion on this is invalid anymore +Testing it a bit more +Thanks Joerg +Wonderful +Its that all direct dependencies are in Maven Central but other dependencies may be Im not really sure we need to change much here +Were using the for an inHouse iPad App to display Videos PDF etcWith CB the video fullscreen works as expected but were want to switch to the inso Ive tested this +Aslak where is this on the roadmap? How hard would it be for a contributor to fix this? +push to +It is because is raised by IO error and will lead to connection close +Thanks Alexey! +Robert Ill leave it up to Keith to make his preferences known to us +Hi GustavoI think the new doc is ok +Change to and associated my opinion the positives of this change outweigh the negatives but is it worth consulting the mailing list and canvassing opinion before committing +I think I have them all this time +It includes what Adam suggested in the last comment +Duplicate of JamieDo you know which version of the ESB this is going to make it in toThanksSen +committed the proposed fix thanks! +Fixed in rev all issues +Removing fix versions as from the comments it doesnt seem to be fixed +Yes for the above data everything is done on a +Ran the patch against hadoop which passed as wellIntegrated to trunkThanks for the review Matteo. +I committed this +The patch which allows overriding of header +Im using Ivy +Closed after release. +I have filed jiras for addressing all feedback plus and additional one for reducing buffer copiesThe one left is varint and int inconsistency this isnt a protobuf vs writable issue +Version Final is released +These blockers and critical issues are resolved but not verified +may not be accuratesetting useDLQfalse disables DLQ handling on MDB side +Still not right +Initial Donation Flex Framework +Groovy is out +I didnt find when I looked for it +attaching I have it for inclusion in +The patch has been applied successfully except for the file implsrctestjavaorgapachemyfacesel which I dont find +Thanks for the patch Colin +RTC task all remaining issues so I can release in JIRA +Attaching a new patch as my previous one has an issue with the + + +This is the patch I would like to submit for the fix +In any case it is a real problem since we see a large percentage of our potential customers sign up for new accounts and then never order anything +Reassigning to Martin who has better knowledge on inner workings +Patch proposal based on version +If you want me to look at something please attach a patch to an issue in Jira +Patch ve deleted the patch because Im going to put a new one in shortly that has some more changes +Try to upgrade toThis works for meString smtpUri smtpusernamefreeforallpasswordfreeforall FromfreeforallTootheruser +The issue at hand is not whether Oraclemonitoring tools should be able to handle SQL that differs only by alias we know they cant and thats the world we live it today +Verified +I was able to exercise the move case but I cant figure out how to exercise the online case +I think that +Also tune down your to seconds +Thanks for all the work you have made on this feature Simone +Ill just edit +Solr tests do pass with this patch on branchxThe little bit of poking around that I did in the UI on did not turn up any noticeable problems +Thta fix seems to be invalid and not required +Any objections +As a workaround you can simply manually reload the page and get the updated bundle details +Rejecting as this behavior is expected +Thanks guys +Moving out of +AS M changed their jmx signature for shutdown +Test program to reproduce port only to +Public now since fixed. +Sorry this one fell off my radar +SECURITYGROUPNAMES secgroup secgroup is this line correct? It looks weird I dont see this kwarg being split anywhere +Wanted to revive the discussion +I think it could be better to call in the constructor of instead of the constructor of this way it wont depend on the commited to and +After adding the sychronization fix I the test about times without failure. +I will run with just the patch but am woefully out of time today +Update draft can sort and generate hfiles for bulk load but havent test it for multiple regions +makes more sense for users because user doesnt expect creates a new file. +Im working with the Mesos team here at Twitter on thisfeature already so that works for meThanks Ben +What error you are experiencingIn addition I have attached the README file so you can follow the steps to run the Management ClientMany ThanksI am looking at and I am seeing the following errors cannot extend as it is not within the same package same for The above issues being solved has a type mismatch issue at line I may not be looking at the right code though +I prepared a patch +I restructured the description and more importantly added the Analysis if youre watching this issue you may also be interested in watching this for M pending the outcome of +Are you all using a unix fileysystemIf I add Csomething as a deploy directory either using or using the jboss tools directory picker no deployment is done from this directory +So the bitmap index will not be very useful +If this is not true please point me at the code so I can study it +for the patch +Pls feel free to ignore them if you want toResolved conflicts of no real change from +Shay I havent looked at EAP in a while now +Based on Kohsukes comments this is an issue in javacapt + +Works great i a regular table +If a example project isnt feasible at least the POM snippet showing the bits for the Shade Plugin and the corresponding debug output from Maven would be appreciated. +Sure +Im using checking the plugins +Fixed typo +The crypto components for support for with HMACmd are now supported in JDK +dev as the mailing listUsernames from Cordova that I know have created accounts agrieve shepheb +This does not seem to be a problem in reading the instantiation code. +I guess I shouldnt test stuff that I never use in real life +Working on a test case for fencing +Parameters are not supported inside XMLSERIALIZE and thats what the new test cases are checking +Ive attached a test case +Id prefer independant xdocs for each bridge component in generalIn this case though maybe its more useful when accessible directly from the main site. +The interface of component The implementation of component Composite define file Define the interface whith WSDL The client to test the under +The test failures was due to a logic error in my code + +Ive tried using the maven changes plugin but getting nowhere +I updated this morning and in one case an error like this went away +Added this validation in Rampart +We will also need to get writers to document it. +can save an appreciable amount of time by checking for schema during the delay and then short circuiting to the is check. +patch is for not trunk +I think this is not relevant any longer uSling and Ujax aint no more +this patch uses jquery to generate the proper requests to the field analysis request handler then applies xsl transformation on the response to render it appropriately +The partial fix is now merged onto the branch too +Thanks Eric! +cheers. +This solution assumes that all client machines have synchronized clocks +relevant conversation piece under JDK screen in WSRP screen note the WSDL box is checked but the fields dont match Mozilla FF +HelloI am experiencing this same problem with which I believe comes with version ofIs there a known work around for this problemI cant upgrade to at this timeI tried copying file to the lib directory of the server but get other types of errorsIf this problem was fixed how do I upgrade my to a newer version of the EJB jarThanks +done Only difference between patch and what went is is manipulation of boot CP since it was half way it was depending on in deployjrebinlibboot So removed the line and will take up on dev list. +We have test the patch in tooit can fix the problembut I think its better to have a reference in catalogtracker then there is no need to use to invoke a rpc +Unfortunately we are no longer using AntIvy as the work done on the Maven build structure is proving to be working wonders the Ivy stuff would therefore need to be dropped and Maven integrated +I have committed this +No +Added a mean shift clustering job and now it works for too +Hi Jervis thanks for the information +Thinking on this a little further I do think this is a bug and I do think it warrants going into The original range and prefix queries were broken and I fixed them via + +hdfs or spilling messages to the target server? Or storing in +clone method thinking there might be something fishy with that but nothing came out of that +Workaround existsDont use web console move option to replay a message from +Thanks for clarifying that for me Taro +Adds docbook render new Docbook thanks. +Further use of methods in Git ID overloaded method taking in a description string which will displayed as additional information +Thanks for the hint Felix with this I could go back to the latest jruby in revision +Thank you. +Those can be viewed individually too on specific managed object page. +Simone already had a data structure of supported file extentions I used this to check files prior to them being handed over to the annoy gdal code producing Error +It seems we would either have to create some management representation of or expose the enabledisable counter as operations on the queuestopics +The XA datasources default config in Admin Console was to use a domain but the domain name is not specified +Hi JacquesThanks +However doesnt Eclipse have the ability to attach actions to individual words inside the editor +it seems to succeed on linux +The Activiti Modeler is not shipped with Activiti anymore +Not yet +bulk defer of featurestaskswishes from to +Maybe the best solution is applying the patch in and then make the error message more useful with other debugging informations +Thank you very much Saravanan for the useful information +Committed in trunk and branch +I also added svn to all files in trie andBecause this is not yet committed the patch may still fail to apply but I will commit in the next few hours +Cool Can do that If the name is really long it gets cut off +Althought I would try to make dependencies on jars onlyThe question is who would want to depend on any individual ACF war files? If theres a need then fine but I dont see one here +I think SWT GTK is stable +Not sure how difficult that would beWhy is this an option? Surely a is a serious bug and we should abort if that happensIs this intended for production use by users? If its primarily for debugging then maybe integrating valgrind or address sanitizer memory poisoning is an alternative +I will submit the patch todayThanks you please review the patchThanks issues Kristian I will update the patch within a day or two +I slightly prefer to leave the mechansim length at since it simplifies implementationTotally agree +Slow clients with outstanding requests are handled by +I currently try to integrate his xsd generator with the +Straight forward just committed this +With the new publisher the hierarchy traversal is taken care of at the publisher level and applyfsdata dont have to worry about it +I will commit the fixes in a few days along with a few more tests +Oops sorry Susan I changed one too many +Straight forward patch +I have similar problem +Andre is on PTO issues moved to in Developer StudioVersion Build id Build date +Dupe of as cleanup +Hence well have to build the delegation token system +Please review status of this issue and close if finished +Looking forward to play with these new search features +Updated the dependencies in vdoclet to and +on dougs suggestion +And covers the entire possible ledger id range when doing gc +Linking upstream Cordova issue +I have removed the LIBNAME assignment done in the srcxercesc makefile +obsolete +Bruno thanks for the feedback +Hence I mistakenly updated the description that Mojarra and behaves similarly in BB +Same error didnt occur with LibxmlNormally +I suppose its stable for client as it has been announced the but for core? JacquesThe http +Closing the bug. +Otherwise wait for the release which has a version that is catching all possible throwables to log why this can occur +This allows added and removing search servers on the Rohan. +They have disjoint data and should not be affecting each other +This issue has now been fixed under so I guess this one can be marked duplicate and closed +Updated the alias generator to account for references when supports is turned off +adding someone will need to apply to branch +Thanks Aditya +I am fixing all the issues related to TAKILL at +Things seem to be just fine +Thanks +Update +Reopening pending resolution of the design issue +Quick question do you always increment by the same value by any chanceNo sorry just happened to pick that example +I wonder why the usage of col is considered normal and not a hack I feel that providing such solutions is not correct to any client +Well have to rely on JEE test environments which is already part of the itests anyway +Yes +We do not have a patch and this is an existing issue +With this patch the aforementioned spec passes and it doesnt break any of the other specs included in the ant let me know if Im missing ll wait for the stdlib fixes from upstream +Backported. +This would solve a major problem of ours +Attaching the with the feature implemented I hope In the end this was quite simple and no need to update Clirr was necessary Vincent was right The new feature Ive added is illustrated by the integration test srcit look into the of v underneath there +Is this fixed as well in the recent snapshot? Thanks in advanceUsing Studio C compiler and using gmakeThanks in advanceMario +I believe this change went into revision +I just wnat to be clear that I spent a lot of time thinking about this +done for disabled attribute +It said connection refused +CSV was base functionality mostly complete but not quite ready to be fully promoted and added to showcase +contains a workaround description thats SPI confirm updates when creating the release +Thanks for review E. +I need to implement a dialect for Tera Data but I will be using Hibernate +There is quite a lot of overlap in the code for the GZIP and Snappy codec in +fixed by using and owb shipped in +Thanks Robert I was looking for testing sharelib update command with HAtest helps. +srcjavaorgapachecommonslang srctestorgapachecommonslangSending srcjavaorgapachecommonslangSending srctestorgapachecommonslangTransmitting file dataCommitted revision +They are of no use except when we are explicitly trying to debug something in which case the user would be aware about this configuration file. +Hopefully someone else can retrigger I added you as contributors +In fact the part you cited just confirmed that itspossible +OK and Sorry by the DavidCan you try again with the X maven option +Menu Items were incorrect in previous versions of JBDS +In other words when I will have completed the Junit testing +Text corrected as suggested. +Ralph is there some example projects in the branch somewhere? I just want to see what you see as the layout +Hi Ronald I cant manage to reproduce this issue on my side with Sonar +Please add your wsdlxsd code samples and then we can reopen this bugthanks have not been able to recreate this issue with the latest nightly +If you could get permission to attach these to this bug entry itwould be appreciated +If there are no further reports in the next hours Ill assume the issue is resolved +I am not sure what your cause is exactly but I give you benefit of the doubt +Thanks E. +TAR like all othe archiving formats doesnt need directory entries at allIt is very common to create directories as needed when extracting archives. +Its not used to replace +If we choose the example documents to index wisely to be +Good idea +We should open a separate ticket for making sure that optimization does not break on nondeterministic operations +This issue hasnt been completed fixed yetNeed more investigation +I had a look at the code and couldnt find anything that would cause this +My bad +I just committed this. +Wait we lose entirely? Or just for get? or rather its impls +Hi MithunCan you give some more information about the problem? For example could you attach an example that triggers the bugAlso does run correctly for you +Added it to check this +wont fixthis is the AJ way +Please review +patch for SWF as a duplicate of the earlier issue +Once we are happy with the impl and the xml +Attached is a patch for this ve changed the to pass the natural and the short class name to the scaffolding short name is still used for the g have more advanced in support for names now +Just came to my mind the patch includes the fix for a bug that i encountered while using this plugin in a project +I was hoping we could instead reuse the code from the logical optimizer rewrite and thus make it much easier to make this change +false skips only unit tests. +x +The issue is caused by current take the exchanges exception into considerationI just did some change in the for to hold the latest exception in the exchange and added a unit test in to test for the propagating exceptions testing +trunk The major work is now done only a few bits left +At this point Id just say lets ditch SVN references from our pages and go with Git +At least the current interface is fairly simple and there is always JavadocsThat said I look forward to seeing any patches you may having demonstrating what you have in mind +See +So youre saying this is not a bug +An equivalent patch for is attached +stack Okay all the points you make are good and easily implementable +Released with configadmin +Additional cleanup of the xaState is required and has been filed as +It seems out of scope to me because the summary talks about reading two specific properties listed in the description and the description talks about using existing code with existing privilege blocksYour code addresses reading all system properties and adds a new privilege block doesnt seem to overlap in my mindI think the patch is fine and can go into the trunk but say I wanted to merge this up to branch I would be more comfortable merging just the specific fix rather than this additional change that can change the behaviour +Thank you for the excellent response on this bug +Patch ran fine for me w feel free to commit wthe SBT file removed +Good catch Ive updated to remove the additional call to Lets see if this fixes the issue +Then this JIRA can be closed. + +Example the revision . +Duhhh lazy me +In order to do this I think the BPEL integration needs to be in a separate plugin +Id have thought that all objections having been answered it would have been a done deal then +TODO unit tests as necessary +The attached file implements the specified test +Fixed in r. +Dont know if theres a clever way to do so +Still reproducible +rename has to has in case the logical use is the same +Thanks Thomas since this is a BatikSVG issue then it is not a bug with FOP +Agree with Julien. +Nevertheless as you probably assumed any single quotes in the password will still need to be guess that this problem also applies to the user name +Abstract object methods +Its possible there is still some bug in the stuff +This is so crazy! Why did we never hit this combination beforeThanks for fixing although I see the not really as a bug fix it is more a new feature! Maybe explicitely add this as new feature to +We will remove the create code from so all exchange id is generated the same and thus it supports the pluggable generator +fixed with revision +Marking it for +After trying some combinations it appears that it has to be a combination of ivy and artifact pattern +It looks like the stylebook might have failed when trying to generate the graphics for the docs +Perhaps the dev convenience version could skip running HTTP tests with and total replication +I am on this +Why is it okay to call getObject after getBytes but not the other way around? a OK getBytes getBytes b FAILS getObject getBytes c OK getBytes getObject d FAILS getBytes getObject getBytes e FAILS getBlob getThanks for any advice +Hi Phil Thanks for your investigation on this in advance +No actually it seems to performs worse but the same is true with a Java client +It included a simple search method for demonstration purposes +Once I pushed onto the classpath things went awayI think the hadoop core tests arent noticing this because Ant ships with xerces gets tht when you tell junit to inherit Ants classpath +A test with FF revealed that is does not occur thereSo we will have to wait until our IT infrastructure guys decide to upgrade IE +openshift +Its the encoded region name +I believe this were already fixed but Rob please verify +need to have a look at using prepared statement params instead of a string reprentation of the current value +add to the top level which should include repoclean and later more +Please update the patch to trunk +you configured the cluster connection to only retry for timesI will see if theres anything we could do +This issue is fixed by instead of updating dashboard portlet we only update fragment of gadget that has updated preference by using ajax requestCurrently dont use response No Content for any portlet requests so gadget is the same behaviors +Fixed through the use of a public +If it works for your setting then you should be good to go on this is expected +This cannot be fixed until we implement duplicate message detection which is scheduled for therefore I am moving it to have commented an assertion onmulti and test +and now with explicit file mode +Hi Guido thanks for applying +When you deploy the SA youll see a number of exceptions in the servicemix log +Seems like those two things should be independent +I think you should investigate this in the community forum +imo gang scheduling is a bit different beast +Maybe something for Vladimir to investigate further +Just use TRACE for debuggingMore to come +You dont control the context and cant bind exactly where you want +Im closing the issue but please give RC a try and let me know how it goes. +Tested +Thanks Bruno for taking the time to fix the remaining issues +The parser productions look good to me +How do you imagine flow control if the network gets slow? Would youblock a transaction whose record would overflow the buffer? There are at least two simple alternatives for how to handle a full replication buffer Stop replication Block transactionsI think the first alternative would be the better one since blocking transactions would mean no availability +attach a patch to fix this issue including a test case +Make an abstract classThe bigger questions Do we plan to cache the column values themselves at CFS level? This seems to be the more effective caching mechanism +Ill complete the rest of and commit it. +I think a Decorator class would be a good place for the method you added + +Marking fixed. +None of those versions have been released yet as far as I know so I think they should still be marked as unreleased +I make a anttask already to move jakarta +handles file exceptions and Form handles sizeMax exceptions +Not clear if we still need to +Then would you suggest a new enhancement request to change the errors tag toprovide the same iteratorlike functionality as the messages tag +parentchildgrandchildbq +bq +Going to resolve it now. +Thanks CarstenYour patch is in trunk at r +Oh +Closing resolved issues for already released versions +d is needed and will be done for but c would still be relevant in case the user is not using a connection profile +It is problematic becasue this user does not have a reaonable default page in the root folder +The newly added files were missed +Resolving. +has a timeout value that is too low for a slower dev environment like a VM +Its as well better for people new toJust my cents +It seems like a bug to log audit events for a single logical operation +Using Configurationget instead of +I committed this +I havent run unit tests and with my changes to the patch +Further investigation is needed and the IT for enhanced +is not supported for EAP +Since we less or more cancelled we will not fix this in +Thanks for all the good info +As your comments indicate this patch introduces a new race condition We will kill another users connection if it sneaks in between the authenticating boot and the encrypting boot +I confirm the issueThe null points probably come from the algorithm thinking some boundaries are not closed +On a relatively quick inspection it looks like it is not +Commit r resolves this issue +This only affects slow platforms like the Android default ARM emulator +Hi KristianHere is the output with the patch +Thanks Colin! +I have opened a mail conversation with bwelling +Running the whole test suite with new +For the next version I can chip in with this extra work +Is that error still there perhaps in a different environment +This change has been reverted until implementsthe replacement and it is in the appserver. +bq +thanks for your help +We will try out the patchFor the record after some more discussion with Ben I realized that none of the workarounds that I was trying would work +So some errors are deferred until the class is linked for imports +QA testing is in progress +I should have said in production we were using a second delayAfter a period of close to hours we got the stackoverflowI had made it one second to try and replicate it quickerI will see if I can replicate it with a longer sleep period +Here With services component in the URL is required however it was creating WSDL correctlythanks again +Just as an FYI we had another user in Abdera impacted by this so we had to do a work around in would be great if a new codec release could be done to pick up the fix in this jira. +test could save a few lines if it uses Assertfail after test instead of using gotException +For me its a real bug +applied cleanly to and test passed +Marking closed. +This bug has been marked as a duplicate of +The fix discussed on the forum thread has been applied and the unit test now passes. +Hmm its sunday and the sun is shining and im sitting in the shade with a beer and a laptop and cranking on the sample to use AOP stuff to do audit backups with camel endpoints all configued using the xml and aspects +uses very large buffers when sorting and merging in order to minimize disk seeks and we dont want everything to start using such large buffersSo why not just pass the missing parameter down? Ive attached a patch that does this +This is now being developed in +StackI prefer passing as constructor but that needs changes in many placesActually i had ways of doing this +in JBT M +What was the error in the server log +I wonder how many customers are struggling with SSL issues when they shouldnt have to but dont realize that there could be a simpler solution +As storing location in metadata and the object in one of journals now is not enough info for recovering +It should be only in a few hours +In my eyes this issue is gaining far too less attention It causes data corruption which in my eyes is a blocker for using the software in production systems! We only can use the library because with every new release we patch the jar files with the fix attached to this issue +Yes I do on my linux machine +We have said in our NOTICE that we use it under the terms of BSD license which is compatible. +Documented as a known issue in the EWS release notes +JPOX doesnt support embedding of this type +pdf files +In change to fileCASSANDRAHOMEconfIm hopeful theres a less ugly way but this seems to work +Does this include the bundled slj in? +Updated the patch based on the changes on the trunk +I think that the tidy process should just be opinionated like Apache Maven is and remove all spacing and reformat it with spaces +thats a clever way of doing it +This is almost never what one wants to have happen and it will almost always require a configurePager tag just to get paging runningThe alternative is to have the default behavior be to show rows per page with a pager and be able to show all rows by using the configurePager tagIs your goal to be able to just display all? If so would the latter solution be sufficient +why would it do that +System tests failed because of the timeout in compilation +Closed due to the release of Tiles +Like Maresh indicated that we could use something likeALTER TABLEVIEW +removed now only would be in EAP with CR. +If the security identity of the caller has notbeen established the container returns the containers representation of the unauthenticated identity +In fact a status would be great +Ie +We have checked heartbeats internally and it looks like good solutionWhy It is impossible to setup c client to create new session while reconnect failure +Received confirmation from Tyronne that the change applied is issue +Wow +Please apply the attached patches inside the directories given below +In such cases the jars should be under HADOOPHOME and HADOOPHOMEcontribgridmix +We do not regularly change the format so that is reasonable +the file that I downloaded onto my box then consulted in order to get the build goingSo +fix javadoc warning rebase and some small changes +Please add javadoc to prepareModifications method expecially the parameter number Instead of using names file file and file names such as file file file better? Make the following methods static and move it to a helper class say createSnapshot check Move some of the common utility methods in this class to a new class +The class imports the though this is probably just minor because it is not really required to compile the class but only to resolve the javadoc referenceTrue +I have not seen this test fail since I increased the timeout as part of +Pushed into as this is ongoing into as ergonomics is an ongoing task +Just upload a patch for ones might still use the old versionRex +The problem no longer occurs +If there is a problem please open a new jira and link this one with that. +Oops sorry I guess Im talking about two scripts +It effects the number of artefacts that need to be managed may impact class space resolution due to the availability of new candidates requires configuration changes down the line etcThe preferred solution is not to change the modularity semantics of the artefact +I will try to think a way out +Closing all resolved tickets from or older +Hi Gavinthanks for your feebdackIf ever you find it simpler to mirror the whole repository instead of not only trunk feel free to do it +Resolve outdated issues +Could you and let me know if the new SNAPSHOT resolves the Message probably expired null problemThanksHiram +Making patch available +For example in showcase httplocalhostvalidationEntering for the int field will show this error must be integer min max if suppliedEntering for the int field will show this error Invalid field value for field integer +Yura is this expectedif not is it something we should fix or minor to push of to +Id like to suggest to postpone this until or later if this is not really urgent +If not let me know +Setting Fix Version to future for issues without a set fix version +Retry hudson +Version has been releasedClosing this issue. +There is no snapshot repo for only for M +reopen to set fix version to again with correct fix version +Replacement patch that changes the xml rather than the generated was a typo in the previous version of the patchSorry for the noise +The actual design and implementation is event based with fixed rules for how event notifications are propogated to the component and its mixinsI think neither solution can solve all problems +The build seems to be working fine pdb files are copied in place +Turned out that windows cannot save signature files quick enough +You can close the issue thanks +The committed patch makes this test pass again by avoiding the issue +Double check wMark on this +It turns out that I was using the version for my tests +Patch applied +I fixed that in the build which only takes care of the RHEL builds these should really be removed in the tomcat zips which the Hudson builds can you please fix that in the tomcat builds? Thanks +committed to trunk for I added a small fix to not add an extra if context root starts with already +You may want to ask this on a forum +Instructions below say that choosing Java rename will also rename the file and Groovy rename will is not ideal +No it didnt work. +Features maven plugin can not create a kar with Maven due to version dependency +If Aperture generates that actually work Id like to know howIf it turns out that Aperture arent worth copying then I think what we want is to use the symbol to separate the archive part of the URL from the path within the archive +Changed to bug since Ive not found in spec that it forbids use of valuebindings for validator and converter properties +Thanks for the quick reply +BenjaminThanks for taking the time to look over my latest changes +ie client could continue just with remaining good nodes +This does not look like a complete fix since this can happen at any store operation +This bug should be fixed there too +should now be entirely free of anything referencing any XML +Unfortunately this patch conflicts horribly with recent changes to trunk too +Theres also a comment at the beginning of the class Is not +Fixed +I was planning on doing this in our library somehow but being able to delegate rendering to an implementation of a component would be great +Small patch +Thanks to Jakob Korherr for this patch +After some discussion with Jonathan it seems that Erik is no longer actively working on this patch +So why it closed as wont fix +Old JIRA cleanup +Success depends on interaction of jobs that run in background while wizard performs its work + +Hopefully I can improve this case +Can we spin off other specifications for JMS CDI Extension? Should a CDI JMS extention be part of the JMS spec +Closing all resolved tickets from or older +Thx for your help +The subversion version doesnt appear to contain this attribute can you check for local modifications please +The sentence is removed as you can now bind lifecycle interceptors to both constructors and lifecycle callback methods. +Ok Heath go ahead +Or it will break hdfs and mr. +This save us a byte in each crawldatum for nowI tested the resolver tool with a mio crawldb and in average a performance of IP lookups sec per box is possible by using threadsI really would love to get this into the sources as the basic version of having the IP address in the crawlDatum since Im working on a tool set of spam detectors that all need ip adresses somehowMay be lets exclude the tool but start with the crawlDatumAny improvement suggestionsThanks +Hey TaviTry passing null as the field +Code added in revisions for and for trunk +If anyone else has comments let me know. +the patch is same jira patch is only for branch on ibm jdk +Thanks for the review UweI will commit the bugfix shortly unless there are objections +Attaching one liners for trunk and +Thanks Sadeep for the contribution. +Meaning that even simple identifiers should be turned into and each field should be used as a criteria in the searchIts late so I havent tried a couple of my thoughts but Im attaching the test case in case anyone has some other ideas +Verified in JBDS GA Linux JDK +We only need them in Searcher +build the latest Harmony jre +Typically before the start of a scheduled HDFS maintainence window +No not yet I wont be able to do an update on the staging systems for testing until next week at that time Ill build from CVS and push it through its tests +I observe the expected behavior mentioned in the bug report +Applied in r. +In any case where an old server would return null it would have thrown an error before responding +Awesome! +It could be a nice feature indeed +Ive just committed this +So then we remove redirect +Attaching a patch to look at later when I have more time +Now that the Tuple protocol is getting the same test treatment as the other protocols I removed redundant test cases from +In any case if the task form is not available the link is now disabled +Exactly what I am talking about please get vmware or someother way to reproduce the way in which at least of our customers develop +If you need the Portlet Bridge fix pick up the code from the trunk +Will post a new patch after rebasing +solution exists to get the same results as this proposal +So I downloaded the release and verified that Crunch patches tests with it IF you download the source files and compile it yourself usingmvn clean install and then run crunch asmvn clean verify The default maven jars go against the version of Hadoop and will fail if you run them with Hadoop +Rob I guess you did not move the core codes declaration of wizard groups to +So Johannes patch is a solution for current codeThanks +I was able to upgrade correctly on trunk. +Servicemix has switched to maven the recommanded way is still to build with maven pom files for maven have been checked in but the build is stil not very stable +Thanks +Please submit subsequent diffs against this branch +Seems that there is same problem in Hudson see but should fix it. +Its ok with me to make the change and ignore deletes +How in any way is this screwing with the community? First of all having a separate version of the scm provider prevents you from having to modify the release plugin in order to pull in the scm plugin and second the code works so why not have a tagged release today for those of us that dont want to wait for Maven? Open Source is about the freedom to be able to scratch your itch and share the solution with others not about some ivory tower club dictating what the masses use +Still cranking along as I can sorry about the wait +Cocoon when you have directories with spaces +Made it Unassigned because there has been no activity +Ive copied it from +All SPARQL Update operations and new functions in SPARQL Query are now supported by the preparser +bq +My intent here is to let other people know that this can easily stop and there is an easy way to do it +The method youre modifying in your patch is to really fetch the last add confirmed so your patch actually breaks its semantics +The portlet archetype was updated in r and the dbportlet archetype in rBoth of them build and start correctly with mvn jettyrun but the apps dont seem to do anything when you click the links to go into edit mode +All that was needed is in the +can ignore but apply and once is applied then these will work properly +Ill add more comments to the JIRA when Ive tested it more thoroughly +To avoid the messiness involved with in a future webhdfs I wonder if the presence of the doAs param is sufficient to alter the semantics of the user checks +this should not be used without models +If its a matter of big importance make a GSIP to state this in the coding on and trunkNo contention with aforementioned coding conventions. +I would also like the administrator to have the ability to set n periods with a warning on the config page that each period has X additional impact on analysis +Sounds interesting +So does that mean you think and should be closed as Could not duplicate +DONE +I have exactly the same fault on both and +Committed to and trunk +Keys are then stored as data elements in maps associated with the nodes +First system selects the product +Also I think comma separated testcase is nicer than adding classes with isnt itThe missing piece is to either ignore qfiles that dont exist or only ignore them when multiple tests are specified +JBOSS AS of +The HTML looks pretty +Using we cant implement the full complement of IO operations including rewind +The problem occurs in the test of +This bug is very annoying +bq +Here finally all whitespace changes in one patch +Ive just committed this +This issue was fixed and should be in resolved state. +Ill link a JBPAPP to that as well for EAP +I fixed the patch for distributed search +I recommend adding a static constant DATANODEXCIEVERSTOPTIMEOUT +Removed the NTLM codeRemoved the configuration record member transparencyenabled and modified the use sites to use either clienttransparencyenabled or servertransparencyenabled as appropriate. +Thanks AndrewPatch applied to NIO module at repo revision rPlease check that the patch was applied as you expected +Tim thank you very much the patch is just right +Just add op constants and backing implementation in for MODIFYTABLECLOSE MODIFYTABLEOPEN etc +We will need a new ticket that provides an interface +Output of diff attached as requested +thanks pointed out by Mary Thompson in a comment above there is still one case that has not been addressed In the case where the service throws a fault there is no axisMessage in the message context +It failed later about not being able to find a fileSince Camel there has been changes in maven files to let it build with Maven and hence why it may break +But almost every VPE test use this approach and changing this would lead to a big effort +I dont think this is a confirmed regression so should not have the Existing Application Impact box checked +Do folks see real compatibility problem hereNot sure what a real compatibility means +verified on +someService +I thought this was done +Sorry The status should be. +Thanks for the really quick fix Perhaps this requirement should also be documented for example in section mvc in the reference manual +Release noted this problem for upcoming release +Server log attached +bug fix test cases +Dart I just took care of the +Not all are finished but most are thereYou need to set a property in to use them +Yes I knowAnd Im using and still get the same exception +This also affects version +There are a number of instances of this in the package +that this issue is one with will close once committed +I guess that this should be part of by +The result is that some facet values close to your limit boundary will not get reported for aggregation which will make your ultimate results somewhat inaccurateIt was my impression that other features of solr when there is a limit to combat this situation? For example if you specify limit the distributed query might have limit or and then during aggregation it would be limited to the top I am working on similar functionality for this patch +Steve could you share an example how to use as an extension point so that it works universally well in Oracle and Sybase with? + +Release Branch is also an affected version +HiIn the IRC the day before I had mentioned that I will attach the diagram that I am proposing for SCA Java for community review +The workaround of using an explicit import package is simple and straightforward +Committed to trunk. +There were significant issues found by QE in bpmn and Fuse and teiid has a bunch of fixes pending as well +RSA goes down and +Currently only broadcasts changes to Nodes and the implementation maps Nodes to categories +ThanksApart from that Im all for improving the jcr resource listener for Oak and leveraging whatever is there however we need to make sure that the jcr resource bundle works in both situations any jcr repository and Oak implementation so its really isnt more to my ff conversation with Michael than what I mentioned above he just showed me how the new Oak observation stuff works +If you need to test multiple permutations you can use multiple executions of the surefire plugin with different settingsAs for other common properties common practice is to use a common base class for your testsIm marking this issue as wont fix but feel free to reopen with additional informationargumentation. +Attaching I see +Trying to fix some tests that are failing +Revision some small changes. +Wed really like to fix it +Is the the only issueI can prepare a new patch that works around that easy enough +Yes Roshan thats a good idea +The config scripts in the bin directory assume the bin directory is named bin and that it lives in the hadoop home directory so thats where the test scripts extracts it +That might work but the fact that its not really an dependency bothers me +But colors a bit dull +I checked the JCC which doesnt support columnIndexes but does support columnNames and it ignores columns that are not identity columns for columnNames +Postpone to +I suppose factories in Droids are meant to be used in a write once read many mode +Linking to issues that are for this currently integrated in the master branch? If not then either we finish and close it or punt to as we have enough patches in flight +I assume the CLI appears to be hanging? In case you can reproduce this can you grab a? It might be related to some remote connection close issue we had +I did some tweaking to the both perf +I checked it and it was a typo +Thanks everyone for keeping the patch updated with newer Hibernate versions +a portType names Service as well +is there actually a problem? I think the stack trace from axis doesnt actually mean anything is wrong +What was the consequence of this issue +The ZKFC configs can all be suffixed with nameservice or looks pretty good to me Todd +bq +I think a better solution would be to just remove the offending header from +Problem seems to have been fixed in reopen if you still have issues. +If a job is still running and cleanup task can be launched then scheduler returns a cleanup task from assignTasksAnd Cleanup task is essentially a mapTask +The one that is added by will have its qualifiers replaced with synthetic ones by the default bean extension so it will not conflictWhen debugging I observed two s for one with discoveredtrue and one with discoveredfalseSending deployersrcmainjavaorgjbossweldintegrationdeployerenvbdaSending deployersrctestjavaorgjbosstestdeployerstestTransmitting file dataCommitted revision trunkSending deployersrcmainjavaorgjbossweldintegrationdeployerenvbdaSending deployersrctestjavaorgjbosstestdeployerstestTransmitting file dataCommitted revision . +I think its just a matter of adding a new dateformat THHmmssDoes that sound right +The method has been added time ago its available for sure in the closing all issues that have been in resolved state for months or more without any feedback or update +Updated the patch to trunk +LiyinCan you try adding JIRA to theI have to deal with copies of each review email in my +Both of the proposals change the attemptid format although the first one is implicit +resubmitting the issue is still blocking the use of the STS service in Rampart +Still we should probably work with what is correct. +Thanks for the detail +Sorry +The current code does not support a fallback configuration +has been marked as a duplicate of this bug +The problem is not with but using the maven taglib plugin +I think that there is value in having validation without requiring a connection to the Sun server where the xsd is defined +Is it not better to apply all patches at the same point in time +If we document from the beginning to use digits or less in your schema design unless you need more thatll help +small comment is not enough as a comment I added the description of the issue to piper branch as cfccdcbadddb +Nothing would be worse than the total runtime of junit v s testng tests to suffer throughn again +Obtain guide from comments have been incorporated but this ticket should have been closed years ago +So the point is the handling of blank needs to be treated differently to the handling of actual user which is the issue that probably caused the regression this issue caused the regression which describes how an empty date field cant be just noticed this issue also I was going to post it as an issue when I saw this. +Is it expected? I have removed authentication as a workaround even though I would prefer having it +I am heading out for tomorrow morning but I will try to look at while I am there +Hopefully well have more info by back to +Actually if using jvm mode and no path is specified prunsrv will read from registry HKEYLOCALMACHINESOFTWAREJava Runtime Environment or Java Development KitOtherwise if using java mode and no JAVAHOME environment variable is specified prunsrv will also read from the same registry as above +I will commit this over the weekend +shouldnt be in server package +If my understanding is correct hive does not use a combiner anywhere +Application of either eliminates the failures in all the builds I have performed +Thanks Anil +Done +I see JRUBY was closed +I cant verify this anymore because the issue is rather old and I neither have nor the app it occurred with running anymore +any additional info here? Otherwise we will close this ticket within the next days. +addition to the issue reported here will return duplicate in a heavily single instance +If its POC then you should have stated that in the subject and comments initiallyIn any case whats the concept trying to be proven +When Pax Web gets released we can upgrade to that to give better logging in this scenario cfr +Committed +This Fixes Fixes +I think I know what went wrong +Marking PA to get HIVE QA to kick in +Finally the full build requires around +Agreement to switch from Strings to in +Scary output as usual but i dont think it explicitly failed any tests + +committed the latest patch to master +Patch course numerical instability still occurs for large powers +It works only for terminating jirb when there are no any characters on the input +the upgrade instructions are the same among former releases +This is corrected in AOP +While debugging I found proves of this runing with the server jbossws logs at INFO level reproduces this issue +We can just do an SVN merge +looks like fixed due to +thats where xml Configurations are being hit concurrently for different purposes +Manually test added +Is this patch still working or do I need to rebase on to some branches +Added support for Aggregate CAS Multipliers +lgtmId say commit if passes hadoopqa +David Ill come back with feedback somewhen tomorrow +stopped setting the MIN in the local clusterinfo +file using ant test Dtestcase Dqfile Doverwritetrue +Hopefully the last version +Would be best if we could simply efficiently use +If you set the system property named to true line breaks will not beinserted in the Signature +Sounds reasonable the UI shouldnt show the option if Shindig isnt installed or show an appropriate do have the true set in +When you say reports the number of collisions thats always zero right? If not then Id be a bit worried +Please refactor the classes into their own files +Turning this from subtask into independent issue I think thatll be more usefulThe ultimate goal will be for the tests to get converted to junit and run with network serverclient and that any bugs should get loggedAlso note that differences between embedded and network serverclient can get linked to +JarekI has this issue using the snapshot with your fixIn destroyed callback case only JNDI lookup failsFor more detail see my atteched tested results +Uploaded File Description Is the mapping file that was used to create the XML Is the results from using mapping with cayenne + +See also the FAQ entry about memory problems +These patches provide some new classes that allow different configurations to be used in an application depending on values that may change from request to request +Now it is clear this is possibility to add files into cfg which is out of it if there are such files +Okay my original complicated example also fails when I eliminated the particular named template call actually the flake out was odder but it is still confined to the callsIll see if I can narrow this second instance down to a small example +Making critical as this is something we really need +bq +New Jira for wrong report data +Ill open another issue for this + patch changes to tmp in the file +Oops fat fingered from mobile +With fix for this issue has been resolved alsoI am adding test case with four connections on to Test test +starting Chris you can also remove akpetteroe and dougarthur and then close this per Bertrands request +Reporters of these issues please verify the fixes and help close these issues +My full name on the CLA is Richard D Jackson +Should we blanket disable them now and then reevaluate once the framework code has been updated to use only New API objects +We cant do this yet as we use the same POM for both building Groovy and in the created artifacts +Kathey wrote We now get a timestamp before each line which doesnt look quite right +Aww I was about to work on it PThanks for the fix Enis +Runtime may result +I thought patch is for branchIs it going to trunk +Fail one of the job while setup task of the job is running +Else it prints out the warnings and errors +Still problematic at RC +I want to be able to switch between and my new implementation by configuration for use in the Slide project +As I said we should provide a parameter wrapper to handle this issue +thanks Willem +note this change in behavior causes it to break when using something like testng from eclipse where there unfortunately can be duplicates on the classpath +Between and a lot of changes and new features were implemented +Updated patch has implement pushProjection +will be replaced in by a jspwiki native implementationIt will mainly aim at a simplified snippet definition and IE compatibility +branch version . +Is there any way to get this into? I know your about to release it but this seams like something that should be fixed as soon as possible +Patch to correct test logic as per you please review and commit if happy +I get this exception if I execute Project Clean how I can reproduce the issue? Where I can get this project +It is already done by assembly now lets go for a simple setup +This way is simply not usable for securely hiding anything its more for lets say visual customizingThe very next issue in this pipeline would be NOT user Simon is editing the page then he will see the sensitive information too +Fixed in SVN. +Thanks Mike +It is difficult to get our review board to approve libraries +the initial version of the sample application +In that way the previous behavior will be preserved +Should have seen this much soonerIm running on trunk with the patch applied in a loop of times to confirm +The only thing Im not too happy with is the static header parsing youd like I can utilize to do the job +MBIGNORECSCSERRORS patch applied in above patch is a workaround not a fix +That way previously written applications can override the default library contributions and contribute prototype instead +Will revisit when DN fixes are available +Seam has got its controlled life cycle for components +Attaching a patch with an additional test case just in case +Sometimes too many indirect approaches can actually make things more complex without adding any benefit thereby making things more error pronePlease dont get me wrong its wonderful to see effort going into this code +system prop +for you test and appreciable comment Charles +We get requests from time to time to support less mainstream algorithms but I dont want to add additional external core dependencies for what is really a niche requirement +Ill commit it soon if there are no objections +fieldName country type +deployed jms destinations are manageable under the deployed archive resource +The logic is that if we have an fsimage then that file includes all data up to and including txid +Done +Not having a Sybase implementation I cant check this +tests and ran successfully +There is stacks of info that we can add which will enable to assist themselves and others to assist at infraThis dev section addresses various audiences +So far my tests seem fine so Im not tickling whatever you are seeing +Test set Tests run Failures Errors Skipped Time elapsed. +I can confirm that this indeed is fixed if I unzip this until a folder wo a space +I confirmed this on my machine so its definitely still an issue +What would be the result you are expecting +Reviewed and pushedClose it. +Now I plan to include a get method to the pos tagger factory deprecate from the train methods that requires things that should be accessed using the factory check if old model still works change the CLI to allow passing a factory class I am thinking of letting the handle the ngram and postag dictionaries artifacts and serializers +This issue opens the question if we want to add any JDBC drivers that are compatible with LGPL + +Better pipesstreaming workflows are explicitly considered in one can imagine an implementation of the or loading its user code in an implementation written in the native language +I realize this would mean bringing jakarta into the set of jars required to run tests so I think not a good idea +If we could remove Ruby from our automakeautoconf files that would definitely help. +Committed revision with the false check changed to Thanks Shai +It sounds like I have a task +It considers both reads and writes as the same cost to datanodes +Now using the it only checks on container completion which never occurs because no containers are runShould this be fixed in a separate JIRA or here +Looks good to me +New version of this patch +Commands and Webconsole worksI encountered some issues that I think that are not directly related with this patch +Greetings there any news hereI made recently some changes to the eliminating the duplicate static String definitions and importing them from insteadFrom my point of view there are two different problems here missing support for some field types Is this still present from your point of view can you review this with the latest snapshot? This should be done in implementation cleanup This is still present as far as I can see but can be moved to orLets move a bit with this issue if you can +Thanks for all your help guys + +Patch for uses displaynone instead of visibilityhidden to hide the calendar uses displayblock to display the calendar +Checked at AM November th nothing wrong and e is running well with local and remote at PM November th with revision error has occurred +log file using JDBC drivers +Will check it in soon but there is one remarkable thing duedate on tasks is not parsed +jars copied out to xwwwrepojars +It corrupts two of the blocks and removes one of the replicas +I think we should fix this for since it fails to build + + branch no more under active development +My statement was related to the default order in since that was one of the problems I fixed this morning +Removing the patch available flag as it is very likely that a new patch will be attached soonI need to consider the options and the throws clause needs to be fixed anyway +Thanks for filing this JIRA Jun +increase is fairly substantial for large namespacesPlease look at the overall increase in memory usage instead of increase over used memory +Sorry if it caused you some problems during your testI could try what you suggest but I think it could not be considered like a final solution +Im not sure of the rationale but SLFJ does not provide a way to change the logging levels so the only way to do it is with the underlying logging system +We opened and fix Eclipse bug specifically in order to make it easier to create an +Patch looks comments +The failure scenario Im worried about is that writes the data and then the collector dies in a way +Environment is Windows JDK and Activiti +Do you have any more details on what this would look like +I was sure I had committed this +Should be fixed in further feedback assuming fixed. +check select also ve cheecked on +Looks good to me could you add a contribution statement +on patch +Its an intuitive thing to do even if its not documented I would argueIn any case it would be good to mention this in the reference documentation +Verified on Tools +Do not forget to set resolution +Ill keep you posted +The corresponding test is added as well +It makes no sense to replace a serialization protocol because of serious short comings in it so that we can have wire compatibility but then insert parts of that old protocol hidden in the new one +Shots before and after click +Obviously this has relevance to so theres cross over there +This is a good idea +Especially important because is not provided by Hadoop and some others +Patch to fix Ryans observation +Why not add a test case for skip +DNS just started working again +Once youve done that try the same thing with +Thats not what I had in mind +If there is further discussion on this topic desired please use the Roo forum +attached the deployable think the check which verified error pages called either complete or dispatch was not in the right location. +I really need it to happend for +Applied +Thanks SANNI Lookman for the patch +Document the use of WEBHCATCONFDIR in the docs +I refreshed the project in both and Im getting still the same error +it is not related to this patchI agree that it is not related +This change looks good +Search now weights artifactId higher than any other field +Not sure where the problem is but the quickstart works in +to Owens suggestion +Rolling back patch from lets testcase run without errors +Patch checked in +BOSH has been heavily reworked in TRUNK +Built using cygwin +Documentation was improved +did you eventually find a workaround or fix? Im having the same problems and the only workaround so far is restarting the application +Attachment has been added with description The patch as has been added with description another logging applied +How about patch v +Describing the debugger issue seemed difficult since there was so much to setup to get to the place I was Ive started using H which doesnt have this t say right now when Ill get back to it +Committed to master. +I havent had a chance to work on it yet i was hoping to get started in a week or so if work commitments allow +in c lib there is a that does what you describe +Closing as nothing to do here apparently +The upstream hudson job is no longer failing with these errors +The new code was similar to the existing code so these issues here would also apply to the other goals such as run embed etcI assume its from the very early days of Camel years ago when James created the plugin that the coding style back then wasnt as used to varargs etc +Reassign to a release that you feel comfortable resolving this issue in or leave as is +Please refer comments in +Code looks good +It looks good in svn +I mean the config dir not the config file +Can it be fixed without upgrade mentioned on forum? What side effects then +This is working solution of what I was talking about in mailing lists long time agoJust got some time to do it +Yep lets give something nice +connection +I have not yet looked at it enough +In the past years Ive applied some patches to increase the probability to be able to stream only the message without loading but this still does not happen on every case +Please check for the and were properly generated when a is used inline in a Proxy Service +This issue should be fixed in and branchThanks for your patch on master +This issue has updated all the jars except for which was changed by +Thanks Dyre +All actions should log their action should only be enabled when item is determined to be to fix in since has been shortened and made more focused +Thank Jonathan Hsieh for your commentI do not know what the word Could you the same v patches but give ASF permission to include meanyou can modify the patch for branch and other branch +Separate issue would be good this one is long enough +Obviously only space and carriage return are allowd as whitespace character when parsing those inline images +Savas I will attempt to create a test case and patch but Cargo is entirely new to me and this will likely take some time +Please if you want more classes made package privateOleg +Servlet type has been removed completely as of HQ build +See +Another one to add to the list one of the committers have a look at this one? Should be a simple and none intrusive change to makeThanks +This fix is supposed to be working on all the browser with CSS support so it should be working on IE or but I dont have the environment on hand to test itPersonally I dont think its necessary to spend more effort on fixing this on IE since IE is an out of date browser +Fixing javadoc +That would need to be handled +OK I have gotten them all to pass again so I will file for the failures for the patch. +Had to still suppress OGNL error when source is null +jpeg extension now +The error means the replacement context is badly done and should also have the same JMX compatibility for +Unless there are objections Ill commit tomorrow +I think its related to the key of the would have expected something like and not just unique +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Since Jason is possibly still in transit I updated the pom +Maybe an annotation value or something in the test can specify the port if you want to override the default +Note its possible to publish snapshot jars to the apache snapshots repo from CI we do this in +This is the real bug that we must fixMy proposal is to move typeId generator to compilation local typemanager +m sorry if I ask dumb questions but Im a bit foreign to Springs solution with transaction managementDoes this mean I cant use the JPA version then +tcomponents has +I think we can just remove that exception and move on with drop or do you want it to until all compactions finish? +Applying this patch the report set with id aggregate no longer is inherited +Proposed changes doesnt match symbology specification +Thanks Hamed! +Thomas can you please have a look at this at your earliest convenience? Would be good to address it in time for Mondays release +A benefit from the new types you can realize today is on the iPhone +Thanks! +How about moving them to a separate private method instead? Then itd clear how they are suppose to be called +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Hi Ging yanYes we do not support right now in the query layer even though the support to create those tables are there +Sounds like a good idea to have lock reads +Checked in fix on trunk +Over the past few weeks I have spent a lot of time managing my clusters I try to do these type of operations in the AM so they are less performance impacting but I have a very low sucess rate with any movejoinrepair +I agree with Ari this should have a test case +http dump +The test for this really needs to cover the entire testsuite and then some +We might want to rebuild so it doesnt have the dependency on the glassfish pom +set to MB +Attached the patch for these updates +Nice Ryan! Can you open a separate ticket for fixing the deploy task +Trustin David and I modestly announce the availability of Remoting is an incremental release with dozens of bug fixes and several new features servers can be bound to multiple IP addresses can run in the presence of a security manager greater configurability supports addresses improved connection monitoring server gets client address in invocations Ron +I think the problem is hdfs and hdfs are actually mapped to the same machine +But hey full speed ahead +Ive coded this and Ill go ahead and commit tomorrow unless I hear otherwise +Changes committed for hadoop +Patch for trunk +Same with containsKey x +Added test case with the described scenario Fixed databinding to properly consider arrays when creating JAXB Context Fixed atom binding to avoid data copying +Closing for now +Added the provided +i think such change is better suited for wicket instead of a minor version release +Committed revision Thank you for taking the time to get this working +Doh the add method CAN get called in that block via the select +Attaching a patchPlease the correct release audit warning and the eclipse warning should be unrelated +In theory you are right +Jan can you verify that resource is the correct usage for loading styles located in the web apps root? Is this still an error fro or is it related to? Please see +If this change causes the HDFS tests to no longer log anything does it also stop HDFS from logging anything in production +Ive used ngrams from cp and wrote custom byteMap +I still believe that my patch makes sense +svn ci m Applying Brian Egges fix and test from Sending srcjavaorgapachecommonsclioptionSending srcjavaorgapachecommonsclioptionAdding srctestorgapachecommonsclibugTransmitting file dataCommitted revision . +I have tried to optimize synchronization blocks in and now am happy to find this patch may i knowwhen this patch will be officially included in OGNL +I noticed some of the same issues during my refactoring for Milestone which you can see as +Push out to. +Im also not even sure how we can get into that state Ill try to write a unit test +Great thanks for the feedbackPassing the different event type back directly and firing to the success or onprogress based on its content should be fine Ill try that +can you provide a patch? would be great! thanks Im newbie in Java and JSF +Docs need to be updated. +Patch for and +Attached a SNAPSHOT working in by +I think we already discussed it some for +Cant wait for more contributions! +Is there any update on this +Marking as resolved. +Please give us more informations about the context in which you found this UMR either the XML file you were parsing or the line number of the source file where Purify found it +This also makes sense because as ASCII these URLs are longer sometimes much longer +Changed priority to am sorry its Task controllers +FYI the issue was an incompatibility in the dependencies +commit to trunk at rev thanks +Hadoop was just released +Applied patch in source repository +it is a bug but no critical bug +Committed in Rev in trunk +KarlI made TVOI serializable Committed revision +i installed the last version of get the latest Ruby version etc +That doesnt remove dependencies unless the user explicitly uses the purge option or runs something like +I fixed the problem with the implementation based on recursion but I will keep the issue open for now since I suspect that there are more problems +It should be disabled by default to preserve the current behavior +As of now Attachment is now GPL Licenced and as such cannot be included in any Apache project +bq +Ack nm my local copy wasnt syncd +I believe regions on filesystem is a simple count of the number of directories under table for each table +Im working on it at the moment Ill post documentation as soon as I have it +Updated with bug fix for scanner instance cache control +. +Thank you Roshan! +The patch is quite old and doesnt apply cleanly this days +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +I think we probably want to make sure that has the correct data when a peer transition between leader and follower +Removed unused field +the above is not too readable +General recommendation to all users is to not use is no acceptable! We work in companies where they impose proxies that use such authentication scheme +Hrm +That would be so much better +Maybe the shouldbe changed to determine the encoding in the same way as the does +I believe this issue is resolved by the patch in +We have many other log entries where both values are higher and dont differ by +Reopen as there is a regression JAR dependencies are no more appended to! +I am all for adding this goal but it must be compatible with maven +Any option can use that even types that are String etcThe dont need to be documented in all options as its a general idiom in Camel +IMO this should be reopened and fix version set to +It just has some limitations +In current patch Passing Operation enum to post +Fixed in rNew version deployed. +Thanks Rajesh +Reviewed all these issues and marking them as closed. +Things should only be registered in one type mapping +It works +A non defined Deployment has another meaning in Arq Core +Its really not the fs that knows how high to set the replication count since the will know the number of mapper tasks +I tried your example and is working fine +I attached the info stream output from theBasically I copied the output from the log file +bq +fixed in svn +committed the patch from Amila +Thanks Ivan +Should be finished within the week depending on how stupid s integration features really are. +This needs to be in due already merged in EAP Seam the is now out so closing all resolved issues +Looking more closely at the javadoc saysAn application should catch instances of this class only if it must clean up after being terminated asynchronously +After several runs in a row the Generic performance decline looks consistent but smaller than my prior report about to on most tests +Based on that information apropriate API migrateVM or migrate could be decided +After regenerating we might want to check whether these attributes now have default values for all elementsSub chapter Attribute Name drawtype styleposition svgcx svgcy svgr svgx svgx svgy svgy table table tableorientation +Hmmm still seems to be a problem with entities +Hi Sanneunfortunately the project in which I used infinispan and mongodb has been suspended so I didnt made any progress improving the code sorry +For me with this patch I can build the modules and run the test suite with both java and java +this is the standard old vs new api mess +I have attached a sample script Another question I have is that why this behavior is different from +Hi PhilPatch applied +It looks good Sijie +copy is anyway done from lib to distlib for these jars +remove unused If youd like to get these in you might want to do it before +Anyway thank you for the explaination +Attaching fix for testcase on top of patch of earlier version of hadoop +Everyone gets confused about this stuff and it leaves a frustrated impression before folks have the chance to really see what Giraph can do for themThese details matter! Thanks for getting to this! +Thanks Yongqiang +See and patch I just checked the class and saw the javadoc hasnt changed in regards to required +marking as upstream. +In our approach they need to be added +Good idea Raphael Ive created ticket. +Bulk move fixing the version info for and all affected issues have in all issues not touched in a month to after release. + +This patch adds a testcase as suggested by Nicholas +In fact we already have googlebot and we know nothing about its parser or when it changes +Then we can do the rest as time permits and also encourage new connectors to use the new approach +also hsa the wrong else was deployed +None of them are related to this patch +Term Vectors still failing +Im assigning this to Jim has access to a Windows host to figure out a solution that works for both and a workaround removing the space from the EAR file name might solve the problem +This change continues to use the Iterator approach and preserves support for multiple A records for a single host +This code was deleted during the merge +Reviewing the Russian translation will most probably also require improving the English original +If not maybe ill try passing it through portletsessionsJoseph it seems even if we just shorten the var argument length if the user increases their input enough we still will run into this issue? It seems also with the way pluto works and the response it send back might be so big that reducing my var arguments by a few char may still greatly fall short of resolving the limit +To support variable substitution for those the work is anyway needed +I think Jitendras comment makes sense +Bulk Close for all issues before incubation +This situation means that for some reason the XML parser installed as the JAXP parser inside the environment refuses to set up schema validation +Remove it from your builds and it wont be an issue +Tested on S cloudfiles +Attaching the patch testcases and the two changed files +Patch to add proxy feature +I deleted all of the other tables and then it spread the tablets out evenly +Fixed typo in my example from cas to xmlns +I was just bit by this after spending several hours creating reports in the latest iReport +If you want to give IBM my email address it is wesw +I think Id prefer suggestion to be consistent with the implementation pattern +It may be relevant that I am using Windows XP +Fair enough +Must have removed it by mistake while creating the patch +Is this a issue in +I made a try removing it from the lib directory but then got Caused by +there is no community release +Is there a way to set per application instead of a global config +This patch has been submitted with the permission of my employerThanksMohammed Azeem +Tried using one of the embedded fonts to write the text but got the same errorThis error is the same everytime For input string +added modified will be fixed in the first jBPM release +Lets hold off on this for now. +patch updates test for +Unused imports Remove all +Please create a quickstart application that reproduces the problem +Bug fixes code cleanup since last revision + Changes name to displayName in Use this filePlease ignore since it does not contain path information for the file modified +Im reopening because one of the sentences is incomplete +No longer applies to trunk can you update the patch +So we have to introduce a new indexing version and coordinate with the designer to determine old from new values +Added link to duplicate issue +Headers are for fixVersion +DavidDo you know if thats possibleThanks understanding was also that these messages are not an indication of real problems +As far as I can tell the changes that were made in dont make the build any faster +No problem +Thanks for your review SiddharthI fixed the removal to use attempt and to launch only if its target is map task +All that huffing and puffing for nothing +patch for applied +Some convention is needed like how to specify the cluster that the file is stored on since we want one hdfsproxy to proxy for multiple HDFS clusters +enabled and accessed by some link in administration console? +This latest version of the patch adjusts the unit tests to cover both the new old version of the readers and writers +How do I achieve this + +Tested +patch applied +Test attached +Correctly is an interesting word +I have committed this +Sorry this was fixed a while ago with +Thanks a lot +completely forgotten to check this one +I can open a separate JIRA for that and commit a fix after we get past the issue this one +Verified + push out to +I think it needs to be acknowledged explicitly though +Patch applied at +Patch to resolve this it buffer might imply that its direct access to the underlying buffer +I wonder if this should be fixed also in branchProbably so only if we make a but not needed if its gonna be a +The original patch wasnt general enough to work on all platforms I think +Cheers updated launchers so that they extend a base bundle list +ANTHOME is unsetJAVAHOMEetcalternativesjavasdkCLASSPATH is unsetPATHusrkerberosbinetcalternativesjavasdkusrlibccacheusrlocalbinbinusrbinnfsjadevintHADOOPuserbinI do see BUILD SUCCESSFUL at the end +Scott if avro classes are used directly in then an explicit dependency is the correct way of doing it +Agreed that this sounds fine +Furthermore the user can now install or receive detailed information about addons by providing either the BSN or the bundle ID. +This is not expected to be part of the JBAS this will involve adding support for dynamic resource types to the Jopr platform which is not a trivial feature to add +That would make it a lot nicerRecommend we delay this until +Today the unique setting in the determines whether to expect a unique query result or not +In reply to comment I have reproduced the issue by editing jars +Remember that this is tied back to the hudson builds which are most useful when tracking the branch not the individual tags +Try setting this up to MB or MB +Nice works fine so far +Tests passed cleanly for me on +I had to install geoserver in order to use the preview option in geonetworkHow can i solve this problem +I committed this. +add a test to produce this issue at Revision Author dantranDate PM Tuesday December MessageAdd test for Modified mavensharedtrunksrctestjavaorgapachemavenshareddependencyanalyze think it is worth to rethrow another for the shake of better error visibility +The applied patch does not increase the total number of javac compiler warnings +This problem seems to be cropping up a lot This bug has been marked as a duplicate of +Im investigating your latest trace now + +Can you ask someone to commit this +Reassign John Peterson to this issue +Messages in brazilian spanish of messagesptBR for Spring RCP V I updated the file it has some patches and some additions +If you disagree simply close it back +bulk defer of unresolved bugs to +This one focuses on summarizing the document relevant to the individual query terms and the API reflects thatWe dont need to have a solution we can have choices. +Thats the script error report from Firebug plugin on the lower section of the screen +Fixed by r +Any success here +All I did was modify the of maven to get it working. +I attached a patch for adding a persistence option toYou cant seem to acutally run commands in the file so I just save the aliased part in thereFor now aliases only get appended to the end of the file its not possible to remove them +Hi Jeroenthanks for the note closing issue regardsHenk Penning +Having the eclipse artifcats deployed as maven dependencies worked out to be a huge improvement +Typically a column will only have ttls and theyll be updated in a fixed period win the ttl which is very useful for our case +Thanks committing this means I can remove a custom branch with only this patch +the change above would require a modification on the public API +Cairo is available for the platforms its required on we should use the system one +This patch will solve the last minute +for the patchI also agree with James but for now we dont have yet neither systemd nor upstart scripts +On x it doesnt work of course +not related to issue +Why is the custom message that comes from Checkstyle ignored by Sonar? Is it only for now? Will support of Checkstyle custom messages be added later +WDYT +I put them in same piture +I assume something tries to be deleted that no longer existsThe indicates the undeploy still succeedsSubsequent redeploys still appear to succeedThe leak reported here still occurs regardless of this WARNSo I think for the purposes of this JIRA and this leak reproducer this WARN can be ignored +bq +That failing test which produces the above stacktarce is in the Hibernate Search project and fails regularly though not consistently and on windows only +I completely agree with you Marcus Instances should get asapIt will be however also very easy to have the Management server support for the API since that is also neededIn public clouds you also want you API to be available over without a proxyStill the need is there for having in the Instances. +It just creates a need for additional documentation tests and requires new lines of code in the parser +I have created a custom script that adapts the distribution with jars containing suffix so it can be compared with distribution with community artifacts +Please move tobago from reposasfincubator to reposasfmyfaces +Unfotunately it is necessary to look for rewritten +However feel free to close it if there is no community interestSkip +Hadoop is about sharing a cluster with other usersI beg to differ +The ident will likely cause those to wrap to the next lineOtherwise looks great +Lets split into smaller tasks. +I also ran all the unit tests and ensured that they pass +Assumed that it was valid to pass in a String to a Date field which it is not +This blocks testsuite +sorry wrong removing version again was a doc update for only cannot be backported to +Committed trunk revision +This is the installer generated script showing the options selected for building this server +Andy you have to use Ganymede with this build +To do this Steven it would be very helpful if you could attach a test showing the problem +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Thanks binlijin +doesnt need to as it is tied to Solaris and is the only job that runs there +Diff of the modifications +Lucene writes thesefiles roughly in sequence so if something bad happened in writing the postings then the +not compacting +In the alpha we want to just refactor the original portlet without modifyingaddingany new is possible to have topic watch in Beta release and also new design needs it. +I then applied the fix and ran the complete set of integration tests on the whole apacheds project success +Not making +This or something else in the setup Roo does should prevent this problem +Syncing patch with trunk +I see that Jaxen actually builds with Ant and generates the POM file +batch transition to closed remaining resolved bugs +Before the creates a new proxy it calls this method to clone the DT to the DT its about to connect to +See unit tests for examples +Tony can you update the status of this JIRA andor close it if it has been dealt with? +I could disable this explicitly in the file but im more curious why are we enabling query caching by default? That feels wrong +Now the result makes more sense for me +preserveMulti implies getting values back for every field value hence we always should regardless of what snippets is set toIf we arent changing the request format it still seems like the most sane thing is to show all field values and have exactly one snippet per field value +Committed revision +Patch checked in +Jeremy can you take a look at committing this as Ill need these changes in trunk before I can post a patch for +The relevant classes in the release branch have been patched address this issue +This work is now on trunk +The patch has the fix for the bug +Is there broader interest in this +As I said I have not been able to figure out how to create Knuts scenario but it may arise when we boolean literals +The correct link type is is related to +Im open to suggestions in that department +I also have to add to the regex + +I can remove this validation if you want +has been resolved again +yonik I see a bunch of commits on this issue can we resolve this +Ive applied this patch to workaround the problem +This is primarily a convenience method for a lot of the testing frameworks +Attaching an example patch not to be committed to branch +Scheduling bug to be fixed in next am not that into developing test cases in Java but the test should do the following create instance and get flatpack of Widget A create instance and get flatpack of Widget B get instance of Widget A and get flatpack for it +The same is logged in Startup msg +JBDS notifies the user during installation that it requires Java +I want to know whether there is any other along the way +The To element represents the endpoint and not the action what is the complete envelopeThe purpose of specifying the header is to provide the endpoint with a shortcut to identify the task without having to process the SOAP request +Sample project to demonstrate that with Maven the transformed POM is installed. +This is resolved correct +pushed upstream +Resolving +As I have stated in my st message the back slashes are doubled which is correct +I simplified the benchmark somewhat and ran it on my bit home machineThe relative results for are very similar to the ones above so this benchmark is a good measuring tool for me +Joe is this related to the bad history plugin on Android +Please find the attached +The changes I made this far +source for the new accessmanager to provide integration unit tests for the accessmanager patch to fix some minor issues add some simple UI pages to the of the patch includes a couple of sample UI scripts for the bundle to provide a starting point for some html UI to addmodifydelete access control entriesIt should work for any JCR node by using the +doc is going to ship post doc will ship for have now created the JUDDI Registry Guide based on the upstream branch + +Disk usage info du hs home M homebertrand homecdemos homedaisy +This seems reasonable for most use cases +Is there any ipv going on in your cluster? Thats definitely been responsible for weirdassery along these lines before +Ill double check that the preference ordering makes sense +If you find particular exception messages confusing then we should fix things so that a more informative exception is thrown rather than try to translate the messages to something more informative +Attached is an updated patch that works for both Solaris and OS X +This time it is truly writing exactly what was in the recommended item object +It works! Thanks Martin and Sven +So that module is still running but new data has been written to the config store and the config index updated accordingly +This looks like being a duplicate of +defer to +Thanks Lohit. +I cant repro the exception outside of Maven! If I run from the test works fine +I think its mitigated in with the diratime addition though might make sense there +This will make our dependencies shift to the correct versions as per our parent +Im still trying to narrow down an easy reproduction but for now if I remove from my Gemfile my Rails app no longer retains two handles +Thank you for the link +ok I figured it out I had to install package +I Think this example will have to be completely in order to comply with the MINA API +Patch looks good but you have lots of white spaces in there +sounds like it should get fixed in geronimo +Attach it to this Jira issue add checkbox contribute to Apache Foundation in the upload boxThat is the standard way of contributing +oops forgot to setcomponent Persistanceeffected versions if you take Party out so Organization doesnt extend anything you will still need the belongsTo but s children will be there +There has to be a way for a user to specify implementation specific parameters per job +next iteration this time fixing most of the Byte variants to only write open one file at a time +The patch looks ok lets commit it +Hi SibelDid you try with CARGOThank SibelNo CARGO only has in which you describe the users passwords and rolesWhat is expected is that CARGO would automatically create the rolesDo you get any error messagesThank SavasCARGO does not create the roleI get the same results as described in my first commentAnd I did not get any error messageThanks Savasthank you for your supporthelpIt is working fine nowTesekkrler Ill then close the issue +The attached file contains both jsp file and its corresponding bean +To make the initial implementation of the above steps easier Ill begin with dropping the cache handling completelyOnce all other features are completed and properly functionally tested the cache will be put in place +Jaroslav what server environment are you testing on? AS ? EAP +Probably something changed in the implementation of dialogs +Would be good to validate under your own configuration using the next nightly SNAPSHOT to ensure I got it all sorted +pending a hudson run +Please see the in the enclosed zip +We can replicate the problem you service provider has been notified and is looking into to Jira solved for now. +I think having the ability to change every lock application to have a timeout is a good a define is like using a global variable theres no way to control scope. +Otherwise it delegates to I dont like this at all and I would love to find another way to do things + +This version implements a Jim Kellerman suggestion that does away with synchronize on createTable method and instead keeps a lock per table currently being created +Closing. +can you provide a patch for TAIWAN and TRADITIONALCHINESEthat would be great. +Javadoc of stubs appears to work fine for me wo any additional configuration of the. +Updated patch contains missing file note that the following files must be deleted when applying the patch +applied in both head and +I have tried Tika The spacing issue is solved for the two test files Nothing has changed for the Windings characters +Additional information it happens after flushing they appear in pairs in different threads +There is still a related issue default entities like ampamp get inside head leading to invalid html output +Commit to branch +do you have a file that reproduces the issue you could attach +the auth stuff could be nested here too maybe as properties rather then needing a +Just for your information Oracle implements this feature as followsRows in which all indexed columns have NULL values are excluded from index rows that has value in at least one indexed column shall be unique +Applied patches Tong please test +This change looks to me like it will prevent me from doing a rolling upgrade to of region servers while my client is using this filter on scans because the filter will fail to implement the changed interface +Applied patch r +Deployed to production. +The solution to this issue has been attached to CVS now +Also see . +Closed upon release of for the correct file +Thoughts +Thanks for raising this! There is indeed a potential race condition there +I have difficulties running osgi integration tests +We might need to disable one or two unit tests +No immediate comments so marking patch available +Youre saying that you have Ruby server running and that when you use Ruby clients one always waits behind the other but with Java clients the two clients interleave their calls? If thats what youre saying then thats pretty interesting +Ill get to work on that and report back if there are any problems +Committed rev +Sure Im on it +You would also see that this thread is marked as RUNNABLE probably because it is writing out its bufefr slowly +Does that count as a little bit +Images showing gadget where only the context information is displayed for the situation and the REST client showing the values from the server which includes the properties +so I still need to implement the proposed change in the handling of small reads writes when cache policy is explicitly specified + of a bulk update all resolutions changed to done please review history to original resolution type +Runtime artifacts will be removed from switchyard tools plugin but might be possible to leverage switchyard bundles when available. +Its actually better to do that at the and level +Why do you want to override these +Here is a first stab at a refactoring it is without any additional test coverage at present and may well be a bit too much to swallow +Changes the threshold for warnings to be the length of the maximum password lifetime +Checked into types branch thanks Daniel for all your work on this +Still run out of memory +Fix will be commited to will be in CR +Ill investigate ifhow we can fix it +Java defines equality through equals +I have extended the example with some Locales and Messages I am not sure if it makes sense to add these to a JSF app though as JSF has its own mechanism for specifying available locales and we handle those in Seam Faces +Are you interested in implementing that +WDYT +Please address this commentbq +Is there any workaround? +Updated patch to do this in a more consistent way with how the scheduler works already setting weights to when the job is not using a particular task type +Any ideas on how we can fix this +Oops +Or would we still buffer the deletes then onlyapply them only to the new SR created from the RAM buffer whenflush or getReader is called +Maybe we can provide a way to customize the data after the default data is loaded +Add into fix versions +Whoops +TomYou are correct! I created an off by one error with my prior fixPlease take a look at the newly attached patch +now includes ddl for derby test too +It causes no harm onWe may also want to switch the setting in back to false +First we should not assume we are running a good convex landscape +This patch solves the problem and includes specs and fixtures that show it without import any external dependency +See on. +Simulator source code +Looks like the change in Common cant be applied in HDFS JIRA +This is the corrected one +Thanks Sid +I havent had any time to really work on this in a while but I did bite the bullet and join the pivot mailing list and figured out my issues with making a fluid resizing layout which is sweet and will hopefully motivate me to make some progress here soon +Please resolve with comments on how you verified +We should be consistent with Java but more importantly consistent between the Java and C backends +Committed config change to trunk and branch and rebuilt public +Issues of this type should be filed seperately and brought to the attention of the compiler vendor. +The should know they are required to upgrade to the CP once it is released +Yes I can empathize making everything work properly is a challenge for each database even though its supposed to be transparent That specific thing shouldnt be a problem though but if you have specific difficulties feel free to ask for help on userAnd btw once you get it to work on Oracle a little How To would be much welcome! + +Yes it would be awesome if you can try it outIf there are any issues please reopen this Jira. +this just stung me took a little while to figure out that adding the setter to the interface would resolve the issueplease fix +restart the producer to send more messages +patch looks good +Attachment has been added with description Patch to fix applied +Can you have a mtg and resolve it tonight +ah Just seen this issue here +Attached a patch with fix and test +Doing additional tests Ive noted that if the rect has an y the gradient will be painted from while if the rect has an y the gradient will be painted from doubling it +on the patch +Also why is the Guava lib in solrlib as opposed to contribclusteringlib +It could be the same issue as I had before with Equinox frameworkIm going to fix that +OK no problem +Sorry bad comment above +I have also run into this issue in backing up a DB +Please let me know if you or anyone else have any further questions +All will then use the same cursor name and we will hit this bug if only the testing changes are applied +This is required in CPThe fix will not be in EAP CP so it needs to be patched in by the SOA build +war contain duplicates of all +Really appreciate what youve done +is awkward name yes +I need more detail to sort it transaction recovery manager in the app server should run every seconds +Thanks +Im cool with one question why remove the serial version UID? My IDE flags the class with a warning because it doesnt have it +I want to get larry and clinton to chime in on this +Submit to hudson +I cant find the host parameter in the history anywhere? Am I missing something here +In trunk we can find better ways to do this +Thanks Ramya! +Patch updated with whitespace fixes +Spelling fixedRegarding having a troubleshooting section I think it would be better to wait until we have seen issues in the wild and add them to the troubleshooting section as needed +hereThx and cheers Katrin +WhoopdeeeLet me post latest up on rb in case anyone wants take a look +Changing the marshallers to put a new line after the major elementsBasically trying to get our marshalled output closer to our original files and vice versaTheres more that can be done on our standard configs +But would like to understand what problem it solves +Fixes truncation by opening file in append mode +Very good policy Reto if you want to completely stall the project +Thanks we finally managed to get it work! +The startup would be really fast on that case +Thanks for the comments I will update the patch accordingly +Patch reviewed thoroughly on RB +I updated the building instructions in r +More importantly itll continue working even if one by purpose or accident removes the final from getStyle sometime in the future +The question then is if an agent is restarted and a request is issued during the restart we might see the request skip the host +The main advantage of changing the index formats nowadays comes from better compression and this patch gives a higher decoding speed at the cost of less compression so it is probably not worthwhile to pursue this any further. +The rationale for the modified patch is that in the first level under target is the mounting point and from there one the directories are considered Java packages +Or maybe theres an option in commons cli to take everything after an argument +Indeed it would better be a regular attribute +The attachment uploaded at supercedes the one uploaded at version now checked in +OK Ive attached with that approachI plan on committing this in the next day or two if there are no more questionsfeedbackThank you Ning for this great addition and for persisting through this long process +Namespaces needed to be different. +patch to Configuration so it will return the local host name +It happened when I do a deploy and call an action +Could you add the following info message when a leaf is reached Zooming in is not possible as you have reached the bottom level +Code committed as is +what is in the +So to help those users we could create a basic or add a big warning in the documentation or log a warning message when a large number of readers is opened? Just to make sure they dont run into the unclear OOM situation and conclude Cassandra is unable to handle a few hundred GB of data while its actually just a poorly chosen default config that has already been fixed +Thanks MarkFixed in at repo revision Please check that it was applied as you expected +Acceptance tests with all implicit types at once and in a loop +Would anyone be willing to take it on +Is that unworkable +Indeed for maven the pdf generation fails with a duplicate id +ivycachexmlencxmlenc +Any update on this +Im experiencing this issue without using Camel +Hoopla +What is the impact of this issue? Is there a specific commit that resolved the bug? I havent noticed any issues with a node after it encounters this bug but Im gun shy about ERROR messages +The test was supposed to be ignored when Xalan was not on the CP +The test looks good but lets not have waits like the following in the testYes changed it +after applying the patch you have to run ant jflex from modulesanalysiscommon and ant genrbbi from modulesanalysisicu to regenerate +Committed +Also It may be helpful to note that I cannot reproduce this issue with a +there hasnt been much interest for this and since the module hierarchy will be flattened for wicket post anyway we dont need this anymore. +Actually I need more time in order to implement the same feature because I have to survey Antlrs error handling +Thanks I will try locally as ve got a new patch +Wow that was quick! Thank you +I will set that up and see if the tests pass +patch doesnt apply anymore +Im all in favor of removing unused code +In my environment just got for errors without the patch and for sucesses with the patch +Reopening for more investigation because the amazon provisionr live tests are failing on master now +Thanks for taking it on +Advance zone with SG is not inAnthony +I worry about the performance for thousands of cores functionality. +Id focus on introducing this functionality in jBPM +For It would be nice if it asked Do you want to change the name of the variable? or Create a new variable In the variable combo box if you input a variable name then push enter the system this is a operetion of creating a new variable as think we shouldnt encourage user to change the variable name in this place because the variable has global scope +verified by Vladimir. +has already been released so theres no chance it will be fixed in that release +b +FrankI believe your patch is solving a different problem so probably it makes sense to file a separate Jira for that. +Patch with the changes needed +Patched build was already fixed. +Im not familiar enough with ipc to know which way is best or if we are stuck +Looks good +Applied the patch with thanks to Alan I also fix some CS errors of the patch and updated the File wiki page +Thanks Nate +And it starts using the high scalability counter class +This might be the same as +aar file +Yes thats what I was proposing +Please take a look +Updated to match review +I tracked the problem down to +Pushing Fix Version to v as this isnt necessary for beta and Apache still hasnt given me access to the svn tree +also seems jmxpoller was never meant to be a shutdown poller +These issues have been resolved but never closed +I would like to take a look at itAdrian +No new IT as its covered by existing +The duplicate ids created is a problem IMO +Committed to branch and trunk thanks for the review Stack. +If not feel free to submit a patch that changes charset representation from String to Charset +Now that work is well underway with AS all previous community releases are +new parameter has been introduced escape +If new specs fail for env please open a new issue for this. +Patch was applied OK. +I will address them in +This program reproduces the problem for Clob +But should not conatin new features so move to +A daemon on the build machine is subscribed to the mailing list and processes the mail using this script +Thanks Owen +This can cause your books to go out of whack and should not be allowed +Thanks for the patch +Obviously the unit test failures are not related +Your comment included It wouldnt catch any refactor errors such as renaming the controller method without changing the gsp +looks good to me +But I want to knowwhat version jboss tools or jbds will use eclipse bpel instead of jboss bpel? If it is very soonwe dont need change our jboss bpel codes just use eclipse bpel directly but if not soonwe should fix this issue in jboss bpel codes +While AP or GB means something to a VPF user it is just a random five character sequence for someone who is not familiar with VPF +I just need to add ability to queue requests responses to the default protocol handlers +Downloading from the Drools website at the momentI just tried with with Ubuntu FF and Tomcat without problem +Thanks Ivan and Dennis for noticing the compilation problem +Or else if you can send the source code we can have a lookThanksDiluka +you are right that was my fault +If youd like to propose a fix Edgar Id apply it +Getting this in so that we can finalize the soon +Ah the joy of lucene vs solr politicsHeh +I meant to change M. +I confirmed that refreshing the bundle will cause the error +Posslbly +bq +Contrary to this Karam never saw this on a cluster +Attached management server log file +DONE +Renamed to better reflect the actual need and approach +Looks good +However based on my experience with spock and AST transforms I am thinking that many of the problems could be related to that +Thanks Pradeep for all your work on this +Prolly fixed in Nates Grand Refactoring that Nate will commit after +If you run it with the old version you will get collisions +Committed to trunk +It is actually a snapshot from somewhere between Beta and Beta +Thanks for the report +With patch in the patch above becomes obsolete +OK works for me +Release Note for fails to drop if constraint names change +Patch integrated to trunk. +BJI couldnt find that link in the catalog application +Committed at subversion revision Touches the following fileM javatestingorgapachederbyTestingfunctionTeststestsupgradeTests +has the same comments introducedchanged in this patch +I hope that assert is from the SSL bug James is working onAlan You want me to look at this +If you want an example compatible with EAP included in JBDS you should choose the Seam booking other booking examples require EAP AS and Seam to errors they are created by the XML validator and arent fatal +Is this necessary for any of your wfsv demo work +If we dont use a Pool of how do we solve the problem of reusing? Isnt that the goalFrom what I understand we cant reuse instances that were mapped to annonymous inner Wathcer if we create a new instance of the recipe right +stocktraderdotnetorderprocessorwhile the Buisiness Service uses +you rebase and when you have a few minutes +I saw the same problem +I think rotate only if needed is too aggressive +Committed revision +On commit I made a very small change so the tars extract in directories +Fix for fixes this one as nowxdhadoop config fs namenode webhdfslocalhostxdhadoop fs ls Hadoop configuration changed shell +If there are n and s flag without k the generated artifacts are placed in output directory. +I guess you then still have the factory problem to deal with then so could be you end up having to wrap in any case +Attaching my +Cleaned all length files removed +Closing as cannot reproduce +with the Iterator back to using booleans +My machine at work so far does not seem to have the problem even using the same test +Two threads concurrently writing with the same name may raise issue +Thanks KonstantinJust fyi for other people trying this test I had to set to in my because BLOCKSIZE in the code otherwise the test would hang indefinitely +It would be nicer if this was just part of the boot message rather than an additional message but not sure that fits in the code flowI would be concerned about the last part suggested where we just ignore the flush errorIt would be good to know if the problems with this wierd installation are also problems with the expected read only medium +Thanks for the review Nitay +Hello KevinHas this been added to the community docs yet? Please let me know and I will then add it to the product docsThanks platform documents differ from the project documentation in this area +I understand +Fixed and remove pom packaging +The target ant prints a description of how to use the quickstart like things to do before running +Being non optional and having the groupId it caused some classpath havoc in taseree + +Do you have a bug to report? I believe the only mistake here is that the code refers to the prefix as the uri +Sorry I failed to reproduce it with a unit test +simagingimagine +Patch applied in r with following changes improved javadoc added a test caseThanks! +I think supporting package as a resource makes a lot of sense because it is widely used practice to lookup additional resources like images and css files based on root location +Some one working on this? +I wonder if its time to ditch the separate diskmemory encoding feature as I have a feeling people dont use itNot well enough versed to say whether or which +most likely doesnt compile but to show my thoughts. +for the patch +bulk defer of unresolved bugs to +Change done with the exception of Unmodifiable decorators. +The issue is if we do not update we need to add more checks to ensure it is already done or has failed +This looks like the mapred vs mapreduce set of but the difference is that mapreduce mostly replaces the whole mapred namespace whereas Cell is not +can you add a testcase +I think it would make sense for it to work without the semicolon since embedded doesnt require it +it works! I guess its partly my fault for not using proper element format in my wsdl and I didnt know because nothing was giving me an error +When these steps are done I deployed ear in geronimo server and navigated to DB Manager and it showed that initialization worked +I think that is a more general solution +derbynetclientmats had failures also seen in nightly testing +I didnt say its a good argumentBut I dont want even bad arguments to hurt adoption Extending cqlsh to handle the metadata introspection that the cli does would be about as much work as creating a new cql shell in Java with the existing introspection code from the cli so if we do this now it doesnt cost us much if anything s actually a more meaningful reason why avoiding extra dependencies is a good thing the most popular dev platform remains Windows where dependency management remains firmly in the dark agesIts kind of a black eye for the experience if Joe Windows tries to follow the README and try some sample CQL but it doesnt work until he goes and installs Python first +I which was clean +looks cool how is it communicated back to eclipse +I think this issue is probably unnecessary the work of moving cores out of has happened elsewhere +Encoder rather than etc +If youd like to see your problem addressed before Antons problem then Id suggest submitting a separate bug report since this bug report will not be addressed prior to Struts +Recreated patch this one should the patch to the branch revision thank for your contribution. +Question if the docs get updated the built versions do too obviously +Different conversion methods for list and show views +The issue is why not and how to fix it. +This patch involved massive rafactoring +I am for commit Seb +You mean how it works in? It is simple +Alex If you agree please resolve it +Ill open a separate bug report for that +Thanks +I am not able to reproduce the issue +out that those classes are obsollete +As we go forward there are going to be sets of standard annotations which a custom control author may wish to use but which are not defined with on themIn this particular case the is something that is intrinsic to the development of the Service Control if you want to change it you will probably at the same time need to change the source code of the JCXI dont want anyone who is not the JCX developer to be capable of changing this +There three different this issue will be fixing Convert the Embeddeds runtime engine to use delegating class loader instead of the currently in The directory where the runtime libraries are dynamically loaded from should be The extensions mechanism needs allow configuration to specify multiple directories +Code maybe experimental still it needs to have test coverage that documents what it is supposed to do +Where is this mentioned Schema and the attached file +Duplicate of +The related tickets fixed support for not Embedded +AdditionMade my own Renderer and changed setVariables to fix the wrong values works fineThe only thing i cannot get to managed how could i render the scroller with the new values correct values it still shows the old wrong viewAny ideas? kindThe scroller gets rendered correct but the dataTable not +Committed to JBESBCP with revision +merge with skipping conflict artifacts is added +It was working withIn anycase this is a bug in KarafSMX and needs to be fixed there +Excellent idea +Committed patch with revision +The second pom attached was incorrect +Integration test added +Confluence done +For trunk the path is srcmainresources about that +Makes it easier to navigate browser history +With the fix I think this is much more likely to occur at a user site +I think we should integrate checkstyle sounds like a good idea to use checkstyle +Darn +Fixed +To fix this properly the quickstart may need to be +is a mandatory element also you need to use the attribute to tell if the is complete or is really a forum discussion but Ill let the JIRA slide since it shouldnt throw a NPE +It is required to add extra check to identify documents separately +However Yoniks concerns on the current patch are valid and we cannot go ahead with the current one +There was a javadoc warning +Thanks Curtis! +The releases are built on linux and as such use the standard linuxEOL when they are checked out of CVS +Move toIm afraid the solution will be something like the one implemented in Tomcat and Grizzly kill the selector and register all the keys again on a new selector +Im going to check and remove my change +Fixed the location of build class files +Sander Temme installed and Ive switched the configuration for JIRA to use modproxyajp +MichaelHaha yeah thats exactly what I did +Log from server and from test +Could you please try again with this latest patchThanks based on Juns suggestion to reduce the sleep timeThe change in this patch is in the validation stage the script will compare the replicas data sizes after a sleep of sec repeatedly up to sec max +I think JIRA messed up the formatting +Will see if I see it pop up again. +side effects none found yet possible improvments remove the limitation of publishivytrue we might publish a limited version of the file containing only the listed confs +Note These are all basically the same issue and all fail at either line of or line of +Since Hive is badly broken on I think the prudent course of action is to revert from for now +The same with a POST request +Try it on patch build server to see if ms wait on regionservers will work up on jenkins +and I cant seem to edit it +Fixed in backport in fixing the issueAlso updated the in the Manifest as some code is using the JDK API +It would also be great if Hudson tested with jdk as well as jdk +I took your changes and modified them so the commit is mine but your name is in the author tag for the test +But its not clear to me whether we need to track the community and the project feature or just one of them +Looks like trunk has the same problem +And for testing it seems original bookkeepers code practice is overwriting getXXX +Inserting that call fixes the +it worked for me +AFAIKT this is not specific to the enforcer plugin +tests pass on branch +Accepting and starting work on this issue +My previous comments remains applied your fix and everything works fine now +tnx for Barneywe are planning to make a Magnolia Extension for Railo all your experiences on that are more than welcome. +was maxEntries profiling is done the results are provided +Otherwise nodes will be timing out all the time and handling the reconnection seems like a pain if Solr needs something from after a GC ends its going to have to pause and wait for the reconnect +This patch addresses Gabriels comments here and Matthias comments on RB +My bad sorry +Attaching +I added the following comment The corresponding XA connection association state is changed by set inside the call to throw according the error code of the to be thrownHopefully it may help someone in a future not to miss it as I didFeel free to ignore the patch if you find it unnecessary +I would also like to move the CLI scripts to a scripts or general clean it all up +And it should never ever catch Throwable please specify the exact Exceptions and handle accordingly +Erik Id be happy to write a test case +Manjula please close this if you are satisfied with the resolution +The i option must come after the generic options +Fix two more places +Please feel free to comment +Igor please verify that its working for you in the next snapshot + Username pilhuhn +You will see a list of settings for the composite service including a tab for bindings +I am puzzled about this one +These have all been added in. +I am planning to reviewcommit patch if any one else is reviewing this patch please let me know +dont want to take up bandwidth with stuff thats not really even that critical. +bq +Well +The serializers for XML now use the serialization and have been fixed to produce the correct output for atom +Allowde Assets get added from andIf people choose to override the default they have to live with being responsible for taking care of security +this issue depends on issue since they share common part of and checked to allow a variable number of operands + +Order now go ahead and just orderHaving this sort of automation could lead to automation of supplier orders manually sending hundreds of orders is simply not feasible for anyone doing a decent volume of businessAutomation might also include an upper limit block the order if its above a certain limit +After hes checked in his patch to change file organization we can this desire. +Committed revision +No this seems to be resolved +FYI we are using spring in a Swing application so this overhead is greatly impacting the launch time for our application and causing frustration with the user community +Closing all issues fixed previously after a brief review of each +we are currently running successfully and didnt dare to upgrade our production cluster to version +Fixed at dafcedfde +With Metamodel syntax it is possible to Join an embedded object to its parent and to use the resulting Join object to further join collections and reference contained in the embedded object +Should be verified in gatein master and reopened if still valid +Youre right I updated the title of the in an IT in r +Resolved in the CR build. +I just committed this +Ive proposed text to +Hmm I thought I sawchanged this somewhere +The patch appears to make lots of changes to but it doesnt +Andrea the components are now committed and +Is this an oversight that needs addressing? I think is private +Committed revision +Attaching the patch for this +It does still need more extensive unit testing but it does seem at first glance to work with concurrent evaluationYet again if you or other have more advanced xml documentations and xpath expressions to use for testing that would be appreciated +Please feel fre to reopen if you can provide a test project that would allow to reproduce this issue +I would like to see how the spec is developing and how many people actually ask for this feature +Since the UIMAAS need for this properties file is minimal +There wont be a release + +IgorHere is some more information that is hopefully helpful +thanks Ivan for this work +Whats the scenario +Im aggre with youI will look to integrate it in JSTL +youre right there is no check in for illegal xml chars in re of the opinion that this would be a responsibility of the xml layer to do far as I see woodstox doesnt check that for performance reason I dont know if other stax layers the conclusion is wont fix +Help in resolution is appreciated +The test would fail + +Patch added +Attaching the log of failed test +Thanks for taking this on Bryan +Ill keep the test running for another minutes to see whether memory increases +Sorry Marek +This is of course working perfectly good in camelWhen I put this converter along with POJO declared as a bean to do some conversion camel is trying to convert this POJO intoAttaching the stack trace and +Attachment has been added with description modifsgetconfigure has been added with description Moved creation outside of the iteration +Also we are reporting the information back to a master called Collector and resolve the memory issue in the jobtracker level where resolves this issue in the tasktracker level +Since a timeout warning occurs when a user attempting to the system I dont think creating a new session at this point is such a big deal +If we get that base in here we make the currently committed stuff sensible and the current shard splitting code will start working against trunk without anyone needing to do the digging I did again +I am almost certain that I used the current cvs versionI did not supply any version tag simple cvs coHow can I verify the versions of the libraries in question +Leave till +Sorry about not providing more information Gavin I should have said that the problem is hoping that posting the stack trace was of help and that it gave enough information +Is it correct behaviour +Thanks +Ah I got it nowIts definitely a bug +there was a logic bug in these are both also fixed in the patch +Ill comment that issue and close this one +Given that its giving wrong answers today anyway Im fine with throwing an error. +Please let me know what you think +srcmainjavaorgapachewicketexampleslibrary + +The two issues should be coordinated +artifact that is added twicec +For things like this please ask on the user mailing list this is a system not a help forum +If we encounter connectionreadout timeouts or other exceptions while determining whether are available or not we still display the links assuming they might work later + +Makes sense to me +Not recommended for other purposes srcjavaorgapachelucenesearch Added some helping code for tests by srctestorgapachelucenesearch explicit imports extend srcjavaorgapachelucenesearch Regards Paul Elschot +If given a sufficiently complex workflow my simplistic node placement algorithm cant guarantee that overlaps wont happen +So you can leave it as it is if you feel that there is no added value in the omission of getter methods for protected fields falls in line with Springs testing support back to the legacy code base I am closing this issue and leaving the code as is. +strange but thanks to your hints I found it and now it is working correctly. +The models are still compatible the check should be adjusted to allow the loading of models +Ive deployed a modified version of ngircd to the forge +The machine I got this on is fairly old and so fairly slow perhaps that matters +When we are defining behaviour for operations now is it possible to implement it in compliant way? requirement for operations statesFor operations an INSTANCE MUST NOT return a HTTP response that contains an envelope +Token now implemnts but violates its contractI dont think this is correct +Your changes are fine IMOAshish +Each instance is lazily instantiated on the first CAS deserialization from XMI +padding corrected +Sounds reasonable +I cant reproduce this even with the seed revision etc + +Ok +If everything goes okay with some new to cover precedence orders then this can be +Hi DeepikaGentle nudge Can you give us any more feedback on the questions which Kathey and I posed in March and April? Thanks +I understand job execution creation needs to be atomic but I dont think it implies the whole logic around creating job execution must be handled by repository +I could not find a way to reproduce the bug with only one of the two ingredients +Please submit a simple runnable test case +Given the existing support seems like best solution would be to return same error number but different message +Thank you very much +The latest patch kills only the last task that started if the sum total of all tasks memory usage goes beyond the configured limit +If you stop remove route then it will be disposed from all resolved tickets from or older +In this case also the stored procedures will not bepresent and will result in the same bug +Thanks Carl +The implementation for IMPORTUUIDCREATENEW is in my view not quite correct in Jackrabbit and Oak +Going to leave this open as I think we can really take this to an interesting level of capabilities. +Any help you can offer would be yes +Could we not declare both methods in and deprecate the old one? In trunk we can remove it and only provide List +I added a testcase to avoid future problems with that +I have moved to +Well +not a blocker for moving to the fix release spec for the time being we probably dont want to do this till v timeframe anyway as its w compat. +good catches +With this patch an observer logs committed proposals coming in the form of inform messages +Relating this issue to the parent pom which must be adapted for the deploymentActually the recently released Apache parent pom contains much of the release definitions we currently have in the parent pom so upgrading to that parent pom might also be a good idea +Ive got it zipped up and can attach it on requestBy the way misspelling in the bug title +All of these should be deprecated and marked as being removed inI will apply patch tomorrow or Friday unless I hear also incorrectly marks numbers as HOSTFor example on line of the type of is HOST when I think it should be NUM +Igor can this be closed or we do need something else +This patch no longer applies to trunk +Here is a patch for was actually committed before beta was released just forgot to close the issue +reloaded so that the builder sees the change and gives its was no longer applying I updated it and made sure it applies and compiles test time is a to trunk +Sorry for not being able to find anything out previously +Merged with latest trunkVinod can you take a look +Fixed and added a test to avoid this in the future +Please could you verify that it has been applied as expected +added test query result +thanks for your quick work. +It looks like the second problem is specific to your code +The code has been unified and all UI functionality has been merged +pushed +My apologies for any inconvenience this has caused you +svn ci m Applying a unit test for +is the the proxy serve up to the a result of a issue +If somebody wants to work on single projects only he checks out only this project +BTW who has reviewed the patch? It seems there is no on the patch +in the night this seemed to start working again it may not be a problem +Correcting the branch patch +Basically for this to happen the one need to import from the other +I was seeing the running it up in eclipse +Thanks Stack +Maybe you can consider before moving the console over to Apache Felix will be moved to the FELIX projectAnd the part of this issue is of course handled by from the release because the web console has been moved in the meantime and the OBR support will also be moved but will certainly not be part of an initial Sling release +The bug is in the Oracle compiler +Changes look good I am wondering though if we can change the copy constructor in to make copies of canonicalMap and and then the code to setParent can be removed from the patch +Testing with the artificial the server eventually goes down on OOME +Fixed for +So we use a special ldap user who is binding with the password given in the configuration who then follows the ldap link to retrieve the user password and checks it +Also added a section on installing bits from the vendor tree +can someone verify that it works? I will merge it to the branch but I dont have windows. +Dont commit yet then +Pat did you forget to resolve this +We are currently experiencing them on all automated patch tested +I will provide a test ASAPCheers TatuI have attached a unit test +Will the be updated in RHEL? +I successfully retested the case in the current version hence I close the issue. +AXISLOGUSERMSG and AXISLOGERRORMSG without AXISLOGSI being provided and without compiler errorI fix my code and everything okPlease drop the patchThank AlexRegardsJM +It fixes the cases outlined in this issue but according to Mamta there may be more places that need to be changed +OopsUse this as the patch for the. +Unfortunately uses the other +sijiefor consume request the changes break backward compatibility +Although its not explicitly clear it seems to me the intent is that it should return no results with an update count of zero +Checked the project buildpath and did not find any evidence of the Groovy it to be disabled in the Groovy Compiler Preferences? +Or dont use DOCTYPE refs. +Mixed up the two because building Lucene with IDE felt wrong to point the IDE to the ant directories for so I placed it under Lucenes libAnyhow if lib is to be kept but not wishing to pack an empty dir could create it if it does not already exist +Added Andy Clement as watcherAndy could you take a look at this? Any comments would be appreciated please assign to yourself if you can spare the time to fix it +The problem is with deferred node expansion when this is on the Node appears as null instead of a Text node with an empty String +Just confirmed that this NPE bug still exists in of have found a similar situation but managed to narrow the problem down to a mapping being changed from to a component +test a new TCK test from my reading of the spec it really is a conformance issue +Im sorry but Im not working with portlets any more and setting this up to test it is not high priority +Hoss Man Oct PM the equivalence described in the javadocs is one of visual character equivalence not of semantic word equivalence that would be a lot more complicated +has new translation stringsThe other classes are comment additionssvn Completed At revision srcmainjavaorgapachemyfacestrinidadresource srcmainjavaorgapachemyfacestrinidadresource srcmainxrtsorgapachemyfacestrinidadresource srcmainjavaorgapachemyfacestrinidadinternalresource srcmainjavaorgapachemyfacestrinidadinternalresource srcmainjavaorgapachemyfacestrinidadresource srcmainjavaorgapachemyfacestrinidadinternalresource srcmainxrtsorgapachemyfacestrinidadresource srcmainjavaorgapachemyfacestrinidadresource srcmainjavaorgapachemyfacestrinidadinternalresource + +Thanks for applying the patch Antonio +Bulk move definitely Avro instead of serializationbinary format +I moved the processes to each connectors and I inclueded JDBCBut Oracle jdbc driver couldnt be downloaded automatically +Note however that I would prefer to now change the jars in which the distribution is organized in +Ill also take a crack at this issue +Many thanks for the patch Patrick applied now! +Applyable patch +Fixing that might also correct the segfault as there could be other build issue going on but I cant really say + uses the passed array instead of copying it +Rescheduled for +Put in a conditional fix +Michael can you provide any details on how to reproduce this? Thanks +Isnt the effect the same as setting mergeFactor to +I will send code review for partuzasoon +Do you expect that the complete group map is displayed after or only the node that you changed +I am not sure the right number thinking about proposing meg vs +But if we want to disable the client generation for now we cant yet +Updateable s are a JDBC feature not SQL +I couldnt think of a better name. +Simon I let you fix the design issue reported by discussed were not going to fix this last limitation in this version +actually I spoke to fast +Other than that Attaching an to update +Complex example is no longer inlined in manual +Vinod the +We already throwlog an exception but it is up to the Object factory to complain about a class being invalid +Nice workPlease add a fat release note on what this facility adds and do some cleanup of the old issues outstanding +Eks can you see if the changes look OK? ThanksIt looks perfect you have brought it to the commit ready status alreadyI will it try it on our production mirror a bit later today and report back if something goes wrongI guess I dont really understand the need for Fieldable +Pity couldnt be fixed better but yeah would need migration script +Thanks +Ive created tests and a separate ticket for API documentation and for adding report functions to the API +Ive made the additional changes to remove traces ofI tested this patch Can you confirm that this patch worked for you? +Does this work on OSX BSD and Solaris? I seem to remember this being proposed before but killed for compatibility reasonsI checked that m and n were supported on Solaris +Here are some unit tests that show the problem exactly + maven project showing the bug in wasnt aware of the fieldname qualifier fallback so I assumed that wiring was purely performed by type instead +I committed a fixed version of my patch that adds a for toa for ranges that are from Fixnum to Fixnum +batch transition resolvedfixed bugs to closedfixed +Im going to try to tune it a bit + +Should be fixed now. +No need for IT. +When is the release date for that fix? Currently our fix is to overwrite the Component class and remove the hexing +If you specify the namenode address correctly this error wont occur +That is just replacing another hard coded setting in the current code + +Verified in +I remove this issue from release notes because it was a regression during the dev sprint. +Sub tasks complete. +Oh I forgot +Thank you for showing how this can be implemented Sebb +My apologies +Otherwise you write TCK for Weld not for other implementations +src added to libcpp is missing from catch merge error +I thought you wanted to forget this oneI will test tomorrow it is late and I already have about hours of work in todayI will comment again later tomorrow +Maybe we should update the wiki page too +for the patch +Right now the has the following behavior if you define output fields using flfieldfield as part of your SOLR url request then you can control the way columns are displayed and ensure that there is a uniform set of columns output in the CSV based on your indexed documents if you dont define output fields the response writer simple assumes that your documents indexed have uniform fields and will try and simply write the output fields based on those fields indexed per documentThe assumption might be limiting but I it was the best I could think of in the case that the requested output cols arent uniformly specifiedHTH clarify what youre seeing! Of course ideas for improvement are always welcomeCheersChris +If the wicket tests are ready we do the svn merge +I dont have a TIBCO license so I dont have access to the at least being able to compile the project will help a bit +This would be great to get into +I agree but I guess that leaves us with the pending compactions returned by the mbean? +There is no circular dependency this bug depends on + change to pick a different directory for use by zookeeper mini cluster +Merged in dbd +Thanks ClebertAs much as I dont like doing assignee tag please can you confirm that the users config is set up in a manner that would allow the RA on server to return the list of indoubt branches on server also +Of the remaining methods are based +I suggest you move the rest to the next section +Ive just tried to integrate solr and Nutch following a fairly clear explanation +Confirmed as still an issue with Apache Flex values go from to not to +I am working with the code base to work in some option or setting to allow to turn off these specific stty options +Attached is the pom used in the Documentation Guide +Burdening the hadoop YARN or communities with the maintenance of all those projects is clearly a road that we should all look to avoid given our experience with the contrib projects in the pastWe have an opportunity to set a clear precedence here +Its fixed in some of the new releases so Im closing it. +Mike I just looked over the changes files in the pull request and have a couple of comments +This patch is an improved patch libhdfs now builds OK but doesnt build as a +Both the send andreceive methods can be outside the HRXAHP if need beIll probably have another look at the code thischeers RolandPS Its in the mail +However its using the mojos archetypeng rather than the default archetype plugin +Adding children to soap header elements already works + +Adding here other urls describing the resources would be also interesting so a client could discover themsmall bikeshedding I would replace admin by ui there +At this point wilma is not a valid cache name +Will attempt cmd line run withbingiraph if of ip testGraph op testOutput ca next +Patch fixes and to pass in credentialsAlso changed to give back Credentials which may have been acquired during split creation +fix on +I dont think adding an executor is a good idea +I guess I just blew the check there +The installer writes one configuration or the other depending on the selected AS issue resolved. +Attaching latest patch again +Thanks for the contribution Rhett +Many thanks Scott! +Updated the patch as per Sis comment +Its true that open will hold onto thoseHowever those segments should be the smallish ones and they should be merged awayIts odd that in your case you see them lasting so longCan you give more details about how youre using it? Have you changed the default settings? Are you closing old readers after reopening new ones +You should ask at legals great newsIf the authentication architecture is at all similar it should not take me too long to do this +In contrast creating a dummy method or failing that removing Roo nature from the project seems less straight forward +eg +I dont think it will be interestinginstead just make seeds less reproducible across java and or other jre implswith different of +Ill commit tomorrow +Here are the updated patches for the bulk their meaning is the same as for previous patchesBesides rebase to latest Harmony state these patches also contain last improvements in Threads module and fixes in suspendresume tests +What is the problem with the checks inI just checked the code the problem is that it always seeks from a freshly instantiated terms enum +I commented it out to make meclipse happy as it kept on insisting I was not using eventhough I am +As in the first case you just need to plug the jacoco agent to the JVM. +You can run it with Junit. +Thoughts +Anyway its fixed by openejb rev . +push to get ready for RC . +Logged In YES useridAre you sure only one client can consume a given multicast packets +Part of seems ok +It is important to resolve this I have been contacted by others about this +Instead of using updateseq from GET dbname I instead grab lastseq from GET dbnamechangeslimitdescendingtrue. +Logged In YES useridhow? Im working under Windows with this version +I dont agreeThe and icons should be in the same style as they are opposite actionsFollowing tonis logic that red is reserved for bad things we might want to reserve green for good thingsIn that case we can use same colored orange icons like +The problem was that the test data was not seeing that this should be using the test seed and the test is notdeterministic +Forgot to say that I also believe a subset of the issues mentioned above is also valid for the otherIt should also be noted that there is a stack of streams and I do now know at which level the real bug is present +Its uncommon to use triple single quote but if anyone use it he gets an error +There is an implementation can compile but cannot run +now uses apache pom version remaining is to release to version and then update the other tiles pom files +I will wait +stderr running ant test after the change +I personally dont think we should rely on NFS but rather we should develop our own network protocol and service by which edits are syncd but I might be convinced otherwise +Please reopen if Im wrong +We would need a jbossws uploaded to the so we could switch EAP AS to use when could you get a build uploaded? The EAP AS CP freeze date is Aug +Latest patch to incorporate the fix for Jira issue created viasvn diff new attributes added please checkThank you much for participation! Ive committed your changes +Why the th one? Anyway keep me posted if you find a way to reproduce the problem +Hello Costin and gang Thanks for the great work thus far! I too tend to agree with Ned that the behavior is slightly +What you described if the parent has a version dont allow a child to override it is +Havent tried this out on trunk yet +for the based rewriting approach with Pluto uPortal had implemented a similar approach to deal with multiple versions of files +Thanks +What was the idea when we committed alfredo? I want to ensure Im not missing something in my understanding +Attachment has been added with description CAPTURE OF MY me the pdf is lighter and svg more to the has been added with description circle not overlapping the has been added with description bug has been added with description my screen again for the last example sent by to example renders fine for me with the currentbranchessvg batik code +I opened a jira just in case we forgot it +r +I think this is just a issue +We should systematically remove all of the autoconfautomake files and regenerate them in the build directory but this is a step in the right direction +Not to be overly cavalier about this but +Insert scrollable in form +There was a lot of logic in my first patch around checking the permissions across the union of all pools +someone with hive committer access should try this to see if we are able to publish to the nexus staging repoI tried that but Im getting error even with proper settings in +Fixes default spellcheck configuration AND adds spellcheck to default select so that the Spellcheck checkbox worksThis is a superset of +Closed after release. +Attached a candiate patch for this bug + +This version includes support for soap and rest endpoints and includes unit tests and samples +The callback will be executed by a bookkeeper thread so you really want to give control back to the application shortlyMy preference is to leave the bookkeeper client as is +Currently there is a dialog wich tells the user to wait until the AE has finishedBut now the log output of the running AE should be displayed to the user in this dialog +Manik if you seriously havent been able to use the dropdown then I think we have bigger problems I know little issues like this can be frustrating but please try to report them accurately as itll help us to prioritise things accordingly +Though Im personally not quite happy if I cannot make it cleanerFor dataSrc I might haved added that and I forgot to remove +Take two based on RB feedback from Matthias +JS Error formscalendar is not defined URL httplocalhostsamplesblocksforms Cannot set value of field with an object of type URL httplocalhostsamplesblocksformscarselector after selecting the model make and type are resetActually in my version I get an error when I change the model +It lets you implement custom protocol handlers with ease and in a way thats easy to read and understand +This should also make the design clearer that this are and not yet what these wrappers can produce. +This issue has been fixed within The existing isnt valid +Add to support method lookup termination if a class provides suitable methods which are not accessible +Ok i tested in same behavior when you select a purpose type the upload works fine +Thanks for the review stack +okay will test and commit +Code looks good +The really odd thing is the metadata indicates the procedure is but only getUpdated actually worksCase sensitivity sucks +HI I had the same issue and I upgraded to +Feel free to reuse the code if you want just beware that I didnt test it much. +However the locale of the error messages and tools is not determined by the locale of the database +I misunderstood that as being able to cache across requests +For some reason one of our applications with Spring Webflow Spring is creating this stackoverflow +I am certain this cant be reproduced because it involes JSON parsing and this has already been removed from HEAD +Still all things are looking good +I hope this will solve the problem +Is his attached not sufficient? +Trying to send one of the attached to the printer results in an application crashI am unfortunately no expert on images in PDF so cannot say if we can do something about it from the FOP endAnyone +jar deployment. +Will you backport this added the new velocity version to the Maven poms +Any grammar changes will not happen before updating fix version +User Task tab has Actor property +I checked and the JVM is being set correctly by the Maven test runner +We cannot say that its a concurrency issue because the test runs in a single thread +Thanks Konstantin for the review +Or should we have an expireBefore parameter like gcBefore so it can stay constant during the compaction +With RC there is some other problem with the restart which causes the test to hang instead of failing +Ive just verified it on mac. +I can help you to reproduce the issue if its necessary +I ran derbyall and one test failed datanet +The dfsck is done at the client side and hence does not lock the file system +If you are sensitive for memory considerations this is the way to go anyways +bq +That seems reasonable +Instead it is either the responsibility of the skin or the caller to apply an appropriate focus traversal policy +The user is responsible for using proper data structures as argument to this method +I will create another patch to make the changes for item and in the wiki as and when my hands are free of other blockers +The alternative is to iterate over the directories looking for the highest n +png showing missing method +was able to launch an instance in vsphere through the deltacloud GUI +Are the configuration parameters different from what it used to be before for configuring short circuit reads +With only those developers that learn API will be able to read the source code generated by Roo +The patch looks good + +In that case a subsequent handler could still set result to OK or generate some other result +Now that work is well underway with AS all previous community releases are +Maybe I need to do something to creatre a better formatted patch +Closing +I should add that Ive been testing on a single node RP cluster so far +changed code from logging warning to throw exception +patch looks good to to trunkThanks Josh! +The usecase is that we use the apollo broker to realize the interaction in additionis the total number of the apollo broker no limit +Beta doesnt handle zero length in the iterator tag +wow this is getting pretty close to collection clustering and other candies somehow to mahout and its thereGreat job and great direction for solr +Marking this as done as we should do a feature request per quickstart +It is not a browser issue but the that is to receive this requestWhen I get the parameters from the request +Right now Im getting a ton of these errors +Fixed by Mikio in some previous commit +Are we ok with that? I think we are because otherwise we would make users lives a misery +But this is just meant to get existing indexes workable with the new code +I have been unable to reproduce this on an iPhone running iOS but can consistently reproduce it on an iPad with iOS +In addition I will commit before making a new snapshot release available +Use this jarfile as a replacement forlib this solve the problem +ivycache so that everything has to be compiled resolved from scratch run grails run grails functionaland it works there any way you can attach an example project that reproduces the problem +I didnt even know that was a valid expression +I am inclined to agree +why not just stick with extends +Just add the Junit jar to runtime scope to add it to the libs folder +This sort of change should really be discussed on the list guys its a pretty big change to the xstream stuff +Adding new patch to also incorporate fix for which depends on this one +but if it isnt the fault it is even more misleadingWhat we can do is what you have done place the stack trace on a bugrep where it is searchable +The only viable solution I have seen so far is the given patch to UIMA to enable allowedValue merging +Hi PetrIve just committed an improvement for this regressionCan you test with latest? Or attach a quickstart so we can test itThanks +Works like a charm for me +Committed this patch to trunk +But it wont be easy +Committed revision . +The sorting functionality in richdataTable is very restrictive +fixed +This seems a lot more sensible to me +mvn site install file mvn site are right of course is only a problem when the artifacts are not in any available repo local or central +You want to add to supporting projects kay kay? +So more or less the feature that we are looking for is to be able to do both failover and round robin +This adds support to the port offset feature for all the containers using the containers using their native can I get an access to SVN so I can commit the code and to the Wiki for the documentation? is the latest patchFor the issues you raised it was a copypaste issue +I am sorry I cant not get your instruction +According to I have added fixes for all the project examples except fixes for the bpel and esb examples +I committed it to the branch. +Leonhard your use case seems rather different from that of this JiraI presume that you are referring to the generated phrase query boost being a little odd or maybe that the phrase boost should not occur when the terms are queried against fields not listed in the pf parameter +Thanks CarlPlease do the including wiki page for code guidelines eclipse setup as soon as possible so that our developers can follow the new standard that you helped set up. +Moved to in +This lookup is now only done when the connection factory is first accessed or when the connection returns an error +Lars Thats another issue +looks good to to and trunk thanks Julien +on IRC +The point was meant only to bring your attention to the fact that there are no guarantees that someone wont change the configuration from multiple threads +No worries Avery +Well likely need Max or Burr to chime in on why some of those things were split up +The inline feature impacts some shindig API +is only visible when Seam perspective is there it is that the only place we should enable it by default is in Seam Perspective only if users ask for it should it be shownSimilar to what Open Type +As this change aint going to the repo should we close this one +Thanks. +TODO mapI think the TODO here is about handling other kind of date formatting +Looks good to me +Any update on this +If the is used then binary properties in in the bundle cache are stored in the temp file system. +Verified. +Cant test the code right now +This work will be done as part of transitioning Grunt to antlr +This got implemented some time ago. +This is converting a static byte array into a String I dont think testing will add anything +made to work with +It might be easier for you to do it that way if possible then we could add your new version to the framework and make it a configuration option + +and Test which generates JDO descriptor included +bulk close of issues as part of making release notes +I will then try to increase the speed of log shipment +Hi Keith could you review these changes pleaseI have removed the display of the unsupported operations from the console and moved the old methods out to a legacy interface used only by the console when connected to older brokers +FWIW the Hits API was originally designed to support desktop applications with a scrollable pane of hits +I still dont understand what you are trying to achieveIf you cancel your data and go back to your list how come that you have old values in yourWhat you are describing to me is exactly what is about it does for you and your beans are kept filled with your data as long as you need them +stricly speaking does not require hadoop or am i missing theres a transitive dependency depends on and that depends on hadoop +For example lets say we have a key of mybean configured in the to be excluded and that value is set to null on the map +File that causes Imaging to throw the exception +Will be dealing with in a good time to fix this or understand it +Simply renames locals to match established convention so that intro will compile again +You can simply build the Axis from the root as usual +What does your pom look like +its all we need It depends on slfj which we already haveThere are other features like unlocked reads etc that can be very useful but as of now we are good +good to go thanks wiki updated +test test issue. +The sitemap and and Index are pretty much the same although I did not look at Ross work not sure about citations but might be related as well +Stan and LukI use for beans that belongs to the UI layer and its great where JSF plays the main role +thanks for fixing it +Patch committed with addition of an updated +That happened after upgrade to JSF +Committed missing files +Also what makes you think that if it is a class then it has t have a main method? I think instead the ant task should use the same logic that we use when we execute a script from the command line +Thanks. +i didnt found the solution any of here can you send me the solution for this issue? it will become a great help to Sripathi please post your query to the geotools user list describing the problem and your system +The other changes I think are too specific as I see their content evolving over time +Please with a list of any events found not to be working +I attached an altered version for testing +How about putting security related classes in a set and exclude them from unloadingWhat about any other type of coprocessor that might be registered on multiple tables but not on all tables? Security was just a good example +Even though fields are renamed with alias name they must be available in WHERE ORDER BY GROUP BY clauses +Are you willing to contribute one such toolThis is also a dupe of where a couple of external solutions are already posted and those authors may be persuaded to contribute back I guess +Could you please give a test case that fails at least under and master +So all issues not directly assigned to an AS release are being closed +Should we raise exception content not found or would be better to hold some history and throw exception content has been deleted +attaching screenshot +Set to x to get the old behavior. +view files. +Will respond by EOD today +I am going to ask something really dumb that I will probably regret couldnt we just lock the python GIL inside the initializeClass method +Event forwarding can be further configured to use batch +Second commit fixes this issue +Its a little bit redundant but better than maintaining a lot of cut and paste code +it was never in the main repo +Ive tested with two configurations Trinidad Tomcat and with Trinidad Sun RIIn both dialogs launch without issue +Closing old issues +I dont know how you feel about this +The patch also adds tests for function comparisons and order by expression toMamta could you review this patch? Also if you know of other cases that might fail with functions please let me know +sha files correctly and deleting the +We can track the issues that failed the import from the logs so maybe we can just clean it up in the CSV file or we can just file them manually +There is one WS handlers setting that you do at the EAP level +is needed to to return in the get call +We are generally deferring significant LF changes to or later when we plan to invest time in doing a full overhaul of the UIIf the current version works for you my vote would be to leave it for nowNice work thanks for the submission +Rat complains that and are missing the Apache License header +Thanks Dhruba! +Ive applied your patch into trunk +The request processors would be the stages +Any other ideas +We have already moved away from RPC protocol version right +OK heres another try at the patch and the corresponding pageI feel that the docs are significantly better than they were before and they nowaddress the issues raised in the description of this problem as well as severalrelated issues raised during the discussion +Fixes the size in the fetcher thread logs and does not update the offset and retries again in the next loop for that topic partition +Hmmm I had missed that required a recompile +We should keep our eyes open for problems this might cause +bulk defer of open issues to +Keep arrow right image as green arrow and rename minimize image to collapse +Problem reproduced on Mac OS X +Knut Anders Hatlen reported that this failed for him on a couple of platforms for +I have a few comments and questions based on the interface proposalwip implementation youve put up At a high level I see a lot of similarities between what you are proposing and hedwig +Verified config in IR and docs in project +Feel free to provide a patch for others to review +So to me it sounds like you are reporting a bug? If so please create a Jira and attach a test case +Thanks for adding more tests Lily +bq +Looks like we need a first pass over all classes before trying to pair up a relationship of this type +to say Im working on fixing the resources approach using resources rather than plugin +Ive added the fontbox part with version Ive replaced the copyright notice with the ASF license as discussed +I did some research and I couldnt find a way to get a valid base price that made intuitive sense +Witness your Ant patch is not very long +Thanks Matthijs for this notification Ill give a try during upcoming development sprint +Client certificate is available in the keystore +Hi AlessioSee attached for a patch for this issue against trunk +Sorry for the broken test case but hey at least I learned something +patch to update dependency to +Im on the patch but a review by Andrew or Gary would be better this needs a release notes and a fat warning in the documentation as someone upgrading will have to change its settings to keep hbase running in secure mode +I verified mvn clean install +Closing old issues +Currently we already use some Java API so we just accept that fact and should check against Java Java API use in Rev +After explored the issue is designed to be invoked by application scoped component +This has now been commited not to core but to contribmiscellaneous +Adding a fix for those mentioned issues +Oki will move this question over to the EJB EG. +If you only consume the changes feed up to a sequence number before the stable replication checkpoints you wont be destroying revisions that havent replicated yet and replication should continue to function +I was concerned that you didnt feel like everything had been addressed but it appears that is not the case +I definitely agree that turning off the healthcheck would not be desirable just excluding them from the stats as you mentioned would do the trick +Just because it might help to list out whats noticeable +I already reproduced the same Exception here and the above patch at least fixes it in the sense that my code for reproducing it would stop to have ExceptionI committed a patch on my master branchThanks to everybody working on it +Thanks +However I believe the approach is correct there were no negative comments and it has not caused any test failures +Release Note now attempts to obtain a new sessionproducer for a request if connection failure occurs instead of immediately closing it +A fix for this has been committed to Qpid trunk at rev with the help of Rafi +Most of the code is fine but there are a few small things that need to be changed +er that should be overrides since there can be +I did some searching and it doesnt look like this was supposed to be overridable +Yeah please submit a patch here and Ill commit it +add override method find to +i wait for your patch +next step is to handle findings according their Annotations but this is quite easyTo handle the AST the is pretty good in the checkstyle package somewhere +One of my users is also facing the same issue +bulk close of all resolved issues. +As I see it right now I can only add conversion to explicitly named channels unless I am able to add a or a customized is there some way I dont know yet? Would be glad if you could provide me with a workaround +Move to +Thanks for reviews Jesse Ram and Ted. +SeeSo Im tempted to say REST connector is ready +Some servers sends spaces in +Thanks Yuki. +Full source for the Acid Test +Patch available +think that is all the more reason why it needs to be done right and not just be a good startThe fact that it has been around for so long means that the good start is gonna take longer to happen +Works with tomcat +Additional patch was applied at r. + its not compatible with Hibernate Annotations in AS OOTB +I committed this. +Will also use this issue to work on jar manifest files +Good + +Any progress +I change it +Jan IIUC none of the problems you have mentioned in attachment is a regression right +Aha! I noticed that if lazyInit is set to false the constructor for will fix this problemInteresting +I do not think that making the user aware of whether the replicas are created or not is a bad idea +Hiroshi youre right! The synchronized is not necessary and since in forceInit something has been already added into the comparators +I think this is safe as this is done during init time when the NN is inactive and it was causing all the and startup failures +Grid please look into this when you get a chance +why are we dependent on remote dtdsxml? All these should be local so you can build without network connection +in release +Additional sample will be added soon. +I also tried with the version and the issue is still there +Hey all can a committer take a look at this patch and let me know if theres anything else it needs +Committed revision +So Im wondering whether it would make sense to add to fixVersion as well? +Trunk version works fine now. +This patch restores the confirm that test compilation passes with the attached patchThe patch looks good too +We will work on this BNF for parser implementation for Query Objects +Fixed +bulk close of resolved issues from releases prior to +Two weeks ago we ran out of disk space on an NFS system +I think this cant make much sense to can reference other cep technology to solve the problem such as buffer all received events at certain and them. +Thanks +The implementor creates whatever implementation he wants if someone else wants the other hell contribute it +Ok well please attach an example project with steps to reproduce +Canceling the patch +Fixed in trunk r. +If we did the repository would be or times as large as it is now +New patch +Gabriel I had the exact thought that you did and updated the patch wjust such an initializer +The patch failed core unit testsWhats wrong with the core tests result? It seems not caused by this patch +So you can close that ticketThanks for your quick answer +Close this one since its verified +This is really odd +Thanks Tibor! Thanks Tom for reviewing +Dirkjan if you dont use the replicator database does the same thing happensThe only case I can see that happening is when the target already has all changes from the source in that case I think the activetasks status is left as Starting +committed! +So this works with any number of standby nodesI see that makes sense +But is this not also a problem for the general Maven release? Is it necessary to apply this patch for Maven? Or are there no plans to upgrade to +Either way this was just a suggestion which came to my mind +uploaded and put relocation poms in old aspectj! +Any neat solutions yet +Without this would render it useless afterserialization or as you point out would require the to loadall of its related objects before really think in the long term the whole approach should bereviewed and maybe bought in line with the approach being adopted for lazyloading +The problem was antlr not knowing about and other tokens. +please dont reorganize these into a wildcardHmm I thought I disabled that in +Looks good Roshan +I will inspect and apply them shortly +It may be verbose to have similar description in each pages but I think that verboseness is prefered in the reference manual +It would be great if we could eventually do it via POS but thats not important for us right now +This has by no means even been tested I do not have access to resources to test on a development Cassandra nodecluster atm +It appears we try to select +All ant tests except the following passed All of the above failures are known issues +Just thought that we might be able to spare some users a data corruption by recommending that they experiment with Java +Removed at request of owner so this isnt applicable anymore. +Im not sure my patch fixes all the jQuery related issues with the CMS +Just released if you want to use it +Thanks for the review StackCommitted to. +patch submitted +I moved this to the next release as it doesnt appear anything happened with this JIRA +WDYT? What is the status of new workflow interpreter code +I totally miss what the issue is +Looking closer at it looks like it is a duplicate of? Is it +would like to see this feature implemented +Is it possible to ignore certain types of failures in the reflection analysis that groovyc is doing? +All audit events are at INFO level +It allows for new functionality without adding logic the the base class and therefore minimizes the chance of breaking the existing implementationSo yes other than the fact that I have created a very small class and added to class proliferation I am happy with the +Doesnt actually build since were not actually using CLRU anymore but takes care of some boilerplate for whatever its replacement is +Fixed in revision after release. +As you observed and I confirmed if this assumption doesnt hold then the scheme doesnt work +Added the images to the feature shared folder +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +This issue has been fixed and released as part of release +cnexampe and cnEXAMPLE are equalsThere is no difference in the two so its plain normal that you get only one DN as a result + +then set CONFIG INT may help you please try it +Ill be uploading a new patch and the unit test in a day or twoFeel free to work on something else in the meantimeCheersPaul Foxworthy +Dhruba as you can see contrib got a +Simple quotes inside a description must be escaped using a as specified in RFC You can report a bug on project +forms auth bundle +Thanks for the reviews +Took care of all the points +Thanks Thejas. +Thinking would be better +I forgot to attach the said patch earlier but thats fine because Ive rebased the thing +Can you try to reproduce this issue using or if you can provide a program that demonstrates this problem that would be great as well +Attaching patch for review +Type quit or exit to quitcassandraI would like to be able to specify the history file location will that be added in future releases? +Update the patch to address Sureshs comments +I was involved in reading the codes and the source code at one stage and wanted to update it but got drawn onto other thingds +This is a temporary solution until we plan a different approach to do this +God this is annoying +Im looking forward to your patchThanks to internal you check out that your patch passes ant clean +This patch is to fix the remaining regressions on embed ejb +We can upgrade to a new version once it is fixed there +all hover classes should be removed and should be added instead +This involves creating a custom implementation of +committed fix +Very generic functionalities used in both have been put inSo I would suggest we keep them as seperate +For anyone to try and help wed need a way to reproduce the problem +the alternative would be to log the condition and fail the request with an or +There is no activity on this since quite long +Patch applied. +For one to to free locks that have never been freed because some thread forgot to free them + to the idea +Thanks guys. +Ill add some more details to the docs +If not then it goes local +You get a Job object that you further configure and run +Let me know if you are strong about this +Right on +Why is the resolution of this issue rejected? Is their any workaround known? Is anyone working on a patch +I had a suspicion it was related to but I backed it out and the problem persists +The origin of the current client exception is that it does not yet know about the temp destination but the broker should at this point. +accidentally assigned to be fixed by upgrading SSO component to +Committed vastly improved scripts with revision +It affects also and the Datagram support +So if you have a patch for trunk with test cases Ill be happy to review and patch was intended to resolve two issues on +I tried dumping the database yesterday using and it workedHowever all table names were converted to lower case +updated +Ive committed this to trunk and +This may be interesting but perhaps not a critical thing at all. +Painful Id imagine +Brian you should really be logging the work against the AOP project not the MC +As to the method is the embedded field will not be set again in this method if it is already set +I am asking why we should not be doing conflict resolution at all +The patch looks ok except for the uncertain number of bytes in long type on different platforms +Did you set the parameter ? makes you think the server expects? like the right fix to me +My goal in is to make current camel examples work better in Servicemix +Per discussion on the user mailing list the proposeed solution will not work and is withdrawn. +in +I think is a reasonable number. +cool working fixed it +Will commit later tonight +I was planning to do that after querying +v patch for TRUNK with the correction. +Namit In the first version of the project which this patch implements the only way to trigger stats gathering is through an explicit ANALYZE command +Theyre not enabled by default as far as I know +That is it should print as that not +Again Ive documented this as resolved but will leave it up to others to resolve the JIRA. +Minor functional incompatibility yes because existing catch blocks may not catch it +err regarding setting it differently on individual nodes creates heterogenous clusters already whoops +and this can only be checked if the or variable is set +Selection bar disabled view +Actually Im trying to get every gsp within the app and its plugins to use the app layout but it is picking up the plugin one when accessing a resource from the plugin maybe it should apply the app layout first and then the plugins as a inset layout +Ive recorded this idea in a comment to which tracks the request for a public interface that exposes the data dictionary version +fixed r. +Please let me know if you need assistance reproducing this issue. +A mere potential threat of a lawsuit is enough for me +Meantime Ill make the change here in to add write time +I wonder if its worth it +Resolved thanks for pointing this out +Did a complete rewrite of the README and added the links to the download pages +Just to clarify did you hide the attributes in the branchNevermind I see you createdresolved in have linked to from this issue. +sync should be default true as its confusing why camel doesnt wait for a response when calling a remote server +The patch makes the protocol be set and duration and retention be displayed +id appreciate any feedback with this +This issue is not approved for SOA CP +Bulk close resolved issues +committed into rev +If I hear no objections Ill start cutting release later today or tomorrowOleg +What we already have is the server scope that holds a lot of this m not in favor of this for this we have the action info with cfadmin and the server scope. +Here is a patch to check for a null return value +First draft of the documentation has been created +Unfortunately the change was large enough that it could not go into. +which makes the two concepts much less useful +Patch Renamed to Made consisting only of Fields Made Field implement both Storable and Made iterators to extract indexed and stored fields from document like the solution but we will wait for Mr +I was focused on splitting sources from libs and forgot to commit that oneCommitted in revision to also add its license file to the legal directory as is done for other dependencies +Third the code in general did not properly account for the case where there is no result set defined in the metadata which is somewhat the crux of this issue +This is a substantial amount of state it can encompass any number of clients and streams and even requests and replies which may have been partially this ties into graceful shutdown if a connection is gracefully shutdown then all clients and connections are closed otherwise the information is all kept +Use this one please +These two files in trunk still have dependencies on +Committed +Here is error log +Thanks for spotting those too! Indeed those can be replaced by array access +patching is almost a on and trunk +The code can open the dialog filtered to only runtimes rather than show all available +Merged. +could be implemented by directly calling a native method which fetches groups via posix apis +Attach an example that illustrates your particular xml override scenario as I have tested being able to override the xml parser implementation in and it works as expected +well look into it asap +Here is a rewritten patch based on the original written by Alexandre is based off svn tag only modifies one method in one file and as far as I can tell uses current code someone please take a look at it +In other words this change created quirky screen behavior and it is difficult to troubleshootFrom my perspective this code should not have been committed +Wasnt the menuinheritedRef stuff supposed to handle all this? +Not necessary after the improvements done in. +If this is a temporary issue please reopen this report when it is fixed +So fixed for GA +Did you verify whether it stops on compile error but continues on test failure? If thats the case Im on this one +Ah I see this was actually user error +Vikash first of all thanks a lot for taking the time to help me +Defer to least for the C backendruntime I think this is now fixable since we compile all of the handwritten and generated code together +all this needs is stablekeyspace in interface +I will update this soon +It is possible though rather unlikely at this late date that everyone has misread the spec +patch to stop transactions from being generated when in mode in module applied on uDig trunk r awaiting change on +Please verify +Im now start to rework my JoachimCould you please give me a simple example using new annotation processing APIRegards Joachim is on holiday this week + +FYI this issue was discovered debugging differences in SOAPR between MRI and +But its probably not worth rewriting all this for now unless someone gets really bored +Dejan is this not already fixed +Hi KihwalIf the NN already has enough to write to a new JD will be synced asynchronously it will start syncing and accepting journal at the same time +Rebased and updated patch attached +eccbd +Because of Delegator this component was removed +But that is no declaration or definition +There is already a focus attribute on txin set focustrue at the element which should get the focusIf this did not meets your requirements maybe we should interpret focusfalse like you describe the nofocustrue +Thanks for the patch. +Can now handle a wider range of Composite Uri formats +The solution is to add the q parameter to ncThe reason is that by default nc closes the connection as soon as the stdin closes +I agree +Hi folkscan we put a high level design doc or draft here then we can push this important feature ahead +It should be fixed nowD +SVN +to girish sathe hi i also try to implement to sub tags of join +Regression tests added +Ok Ravi but if you really plan to analyse Mb of java bytecode is really not a big value +Thanks. +changes to srctestcoreorg +It looks good to me and Ive applied it in revision . +The same result +Patch with more test cases ANDOR logic for MPQ is combined and test code made simpler +Reassign to Devin +I know in this case the left operand was the upper call I dont know what the right operand was +Sounds good to me PaulCould you open a separate issue and attach a patch +The ejb client is now able to deploy stuff using the profile service client. +Commit as you please +Wont fix issues should not be assigned a release as it makes for a confusing release notes. +This is why a added this issueMaybe you should add your comments on the on thread to get the opinion of the others +This was causing issues with the build +By analogy with set there should be a have changed the way in which default and set work this should not affect any existing build but significantly simplifies the way things work internally +Since the goal is have this library encapsulated under the parent package springpython the first subdirectory underneath trunkspringpythonsrc must be springpython +We can unchecked warnings into several categories +Ill hold for this +Abbreviated long subject line and assigned to the right component +In the context of criteria queries there is indeed such a thing as an eager second select +generally speaking parsers become unmaintable in projects are peoplewith the required skillset move on etc +The in the file explains the basic approachThe following is a summary of the class hierarchy A simple interface for factories that retrieve passwords from data sources other than the default configuration file +This patch allows to be used with binary data +After s recoverLease is retried and succeeds +This is being handled in other bugs such as is being handled by other specific to individual issues or features for BRMSGuvnor. +Committed as svn resolving +so implement a workaround for windows with creating git add +Hi ReneI just have quick run with the test and patchI found the patch doesnt fix the old version of and new version of is passed if I dont apply the patchCan you double check itRegards patch into master and branches with thanks to Rene. +I guess you could compare it to data in an repository +Reopening to figure what is going on +Ive been doing some major refactoring of adding some significant enhancements to the Selenium support +Ted comment is fine +Hi AndreasIm trying to do this without maven is that possibleThanks P +Any client or region server lieutenant wants to take a look +But with an AJAX call I get another NPEYes I know about that Im working on it +rpm add chukwa at the end of CHUKWAHOME so the path in script is up startup script and RPM build typo +How to not binding with juel? Because is a part of maybe sometimes later every servlet container will provider their own implementation +However since it has no need to reset the device we can make our workflow more efficient by removing the need of testing dhv reset +The references are different for the book they should befigure and issue but very low priority +Problem is in class +Well Anton still sees timestamp in the future and is still unable to do schema updates it does sound a lot like there is a problem and its related to this ticket +Perhaps this should get a whole new JIRA +They have one big job that has all the reduce slots gone sometimes for hours at a time when a smaller job has all of its maps finished and just needs a handful of reduces to go +If so perhaps we should address this bug in a separate ticket for trunk +xml files its a bug + +Marking closed. +Lets add unit tests after is committed +Also if you have a patch for MFQP unit test class please attach those too +Sorry +Closing the issue now that it has been resolved. +Looks resolved in current release +Comments on RB +Please could you verify that it has been applied as expected +IE Jacques hi Brunoi hope u had a few nice days +If I create a new Smooks config add a Java Mapping task select it and just start on the canvas it still takes me two or three tries of to get the Add menu to appear with nothing selected +I have two authentication provider the one provided by j and the one implemented by my own +The history here shows the discussion that seemed to indicate it was closed +We may need to discuss more about this did we open another issue for the behavior change in? I dont see the need for so many changes if we are dealing with things separately +Ive run it times in a row now wo any case I dont think this intermittent failure should be a blocker or even really prevent closing this issue +Next thing is to create a that will provide additional javascript resources images css m thinking that this class should be part of the UI package what Im looking for is suggestions as to where we place the physical resource files and how they are exposed via the resource servlet +I know it was a mistake but this will screw up people that depend and use that artifact id +I would tend to think that the issue is caused byGert indicated that when tests are hanging some messages are left on the queue and never dispatched to consumers even if those are active and waiting for messagesI will disable the tests for now but we need to keep this JIRA opened and fix the problems +for closing it +ill add inputa validation test methods to my existing tests +Hi Len Rajesh told me the current settings for Windows have been able to run the tests without any problems and suspects this has something to do with +closed with release of +I take it this is a low priority since northings happened for a whileIm thinking about doing some changes in my programs and unfortunately a few of them would require this to be implementedI dont know all that much about how this part of hadoop works are there any outstanding tickets that need to be fixed before this can go ahead? is important and will get done +Thats why i didnt find my +Didnt grok your earlier comments on recursive schema references +I think that the basic stuff are there +Thanks for the patch Knut +Thanks +Im not sure we can do anything from Roo to work around this but I am certainly open to suggestions +Ran into this againEditing the title to correctly reflect what we want +Patch for you +Upgrading this to is to get url before GA and have updatesite content for our plugins initially +The MED project builder might actually issue a warning that the Model MED is not the current version in the registry +However it feels like a bit of a bodge +And each of these seeks would also involve seeking the associated CRC files as well +Attaching a slightly simpler +Yes using reflection might be an option here but that would be rendered useless if the features of JDBC could not be accessed remainder does not need discussion does it +Can you please verify? Thanks! +is this really still not fixed or was it abandoned entirely because of new versions of Hibernate? +I have no problem setting up an ACL limiting access to selected PPMC membersand the Apache Security Team however the IPMC must be able to review and participate in the activity in those areas. +Anyone want to have a swipe at fixing it for him +There is also a corollary to this issueif the pluginManagement specifies the same version as the superPom then the code to unspecified versions aggressively an integration test in m unsure why I thought this was a enforcer plugin is what you use to trap a plugin without a version we have a pom with a plugin referenced and the version is not specified in the pom we are running on then the pom we are running on does not specify an out of date version +Its easy to reproduce it with a recent Tomcat and +Bulk close for release +what is searched by default and jbds would be the only one having the discoverysoftwareupdate stuff configured or possibly different discovery location +Many thanks Alf for contributing itOleg +This is my number one feature request and the only thing Im lacking in Forge right now +Thats something that has to be considered and fixed with whatever is appropriate for the application +Sadly the functionality isnt on the roadmap just yet but wed certainly love to hear your ideas on how we might deliver this data down the road + +Ill point you. +Hi you all Im student from Brazil if I want to participate in with this project how do I find a mentor and where do I find any examples of the being usedThanks +Fixed these anchor issues +The errors and whatnot have been fixed on should create a new jira about the add repository process in light of the configuration upgrade phase +I must have forgotten to reopen the jira + patch it adds a guard while creating the znode +Seeing this a lot lately think I may have tracked it down +That aspect can be handled outside of the engine by an applicationThere are just too many different ways to do this +The problem with the integration tests is fixed in the class in Rev +somethings wrong if introducing more reduces performance +Please looks good +But what I do not understand is when you change the bit of code at least change it so it is going to be useful +Committed to trunk a while back. +You can have two references from the same git repo but youd have to specify it twice once as argument and once as argument Yeah thats what I figured +This has been open for years and the general agreement is that theres no straightforward solution until browsers are fixed +The software discovers that the input data it is receiving is not what it expects +New patch uses where indicated +Use another parser +this specific issue is solved opening new one with the new found issues +Patch to refactor common code from test +bq +I meant to review this earlier but got busy fixing and +Can you open a new issue +Since then I have not tried using the toolkit againExactly what did you do +I agree with Ken Roo give us an uniform development tool our projects are DDD and independent from UI layer currently we can choose Spring MVC GWT +As I understand it you are resolving composed process definitions on the database +Debugging becomes quite cumbersome without this in place esp when working with external source references +done to be release in problem was that your projects somehow form a loop +Patch and tests for again w an actual committed this on revision +Also I already have this implemented I just need to break out the patch +Alan is the release manager +Good point +Should this be a post item +Juerge What will JBAS do if it comes across a dependency entry it cant fulfil? IOW will the entries be backward compatible with older JBAS versions +Checked in patchA couple of notes for the future All patches are going to be checked into the trunk first even if youre using its much easier if you give us a patch relative to the trunk Your patches should ideally be relative to the current state of the trunk +That was the first thing I was thinking about when I heard about Flex +In order to fix I made it so that Axis projects are laid out more like the mini projects +Users with very large files can still pass a temporary directory +It would be awesome if you could attach a small app that demonstrates the problem +How does this effect +Hi JohnDo you have any news on that? is suffering with this we extend the POM to allow to distinguish between different types of extensions in order to tell which of them can loaded sometime later during the build there is no path forward +Another thing I noticed and feel free to punt on it but may be useful if get has an info log if it fails to use the unsafe comparitor + includes some extra python tests to exercise the bug +As each key has to be unique and a hash is a unique value representing that key by storing the objects hash is an instant win on size automatically + +Thanks! +Thanks +bq +It should replace +Regression test patch fix since from my point of view this issue is not very important and this specific case isnt directly mentioned in the spec lets just be compatible and comment out the tests +Thats the initial code of which requires the properties type to be convertable to StringThe code is years old +Can someone else test this patch? I will check the logs tomorrow +Got the same problemHey guys this issue is extremely important for the Asian and the European markets! We here in Germany have umlaut characters like they are part of our language! We need to compile them correctly in Groovy! The Asian people have many many more special characters so that they also need +Thanks Uwe these changes sound great +this looks like it has been committed +There were no checks that involved mismatches +on the if you can get it to run +svnpubsub is now active on the Tika web site. +Why arent we using the same shutdown script as EAP +OK removing the hibernate classes from my war fixes this +Thanks Kim +OK I cleaned up the formatting +Sorry about that this will be difficult to resolve in the meanwhile mount the stateless page using regular is a stateless checking code now can you please test if it helps your problem +This is now complete +the invoker component seems to have special setters for profiles and properties +Found some badly named test data directories in the process +HTTP proxy should be quite straightforward in contrast +is constructed using schema data queried from LDAP server thenit cached all scehma data in schemaTable its fast for searching +Check the CVS head its working for me +Ive added a junit test and fixed the problem with a solution very similar to what you suggested +Yes as I mentioned in the comment Ive tested this with EAP which includes Messaging latest CP release +I would rate this patch as critical +We need not have as separate making TASKTYPE as CLEANUP will solve the think the refactoring of apis should be done as a part of +In all cases I believe we should only store a single UTC timestamp per value +Hi MounirCould you please have a look atThanks a Julieni will have a Mounir! +Thanks Ashish your patch is in rev . +The behavior is undefined for this case +Having said that theres no harm in separating out the class into separate classesBTW Would have to separate out and too so youll end up with classes +Im having a similar problem loading a pyramid in EPSG in Is there a fix or workaround for this issue +mockup and hat shirt looks amazing +That probably explains seeing server adapter at the level +Not yet tested but seems to be an obvious fix +I guess I need to try the opposite solution +Here should be Thanks Tim for assigning the issue +Hi Pavel Can I mark this issue as resolved now +Here is the new one +As I said the version from the zip file works fine +It now compiles +Considering this as resolved for tracking purposes +Get rid of the WARN w thread dump Id say especially as is where it comes after reporting were not going to use +na +I dont think a global transaction can have holdable transaction under any circumstances +but as DOM parser we can add getting SAX parser instance like methods to for generalizationbq +It seems the sources there are not consistent with the binaries +There is nor order + Removed reference to nonexistent usrlocalbin +Where in the source tree should stuff like this go? +Typo I meant you said I had missed adding it to tmpjars +But maybe theres no way around it since it sounds like the whole point is you cant do OR mapping on a column since it never becomes part of the Java object +I am working on a patch and will provide it comments Could you redo the patch by just running your diff command from the checkout root? This seems to be some syntax +added a demo is a screenshot of the demo gsp page +The Edit dialog contains the Browse button that calls a native dialogYou can programmatically add some runtime paths and test them using the Edit dialog +if the JCR API takes a then it should work with any implementation +IMO this is not suitable for a CP release as the changes are too extensive so we need to keep this customer in an branch until someone convices them to upgrade to or +Does the feature we need exist in that version? If so would it not be better to use that or alternatively wait for the vote to finish and use final so users and devs can more easily get a hold of the same version we use. +It still might make sense actually to do that buffer the items skipped and call the listener only once at the end when we know that we are going to commit +Fixed by +its failing on the processor from +Sorry was missing from my patch + is a for branch where its fails +Please let me know if it looks ok now +Verified in +the problem is that we are using DK both for routing and for local key we can skip the hashcode change if youre worried about boxingYes thats what Im referring to for kind +Closing this as not a issue. +My intentions were to provide complete fixBut Thanks for Providing the implementation for this so quicklyWith this I will try to contribute moreRegardsSantosh +There hasnt been any activity on this issue for over years now +Do you call disconnect andor close or do you simply kill the JVM and restart the whole node +Instead it traverse only the nodes that the block is located +The most common use of the is to iterate over the values +I had a brief discussion with Simon and Raymond at a Tuscany call before +Please also add more description on what are actually done in individual instead of Clean up Xxx +ThanksSannna +is isnt documentedWith that fixed and some cluster load testing Im for commit +Igor please reopen it if you are still able to reproduce the problem +fixed alreadythanks Kasun for verifying this +Improving title for has been fixed thanks to Trustin. +triaged July vmware I have not run into OOM if there is sleep and gc after every iterations +Seems OK +updated yarn and mapred only the patch with common changes +Sad to see this not having been Shell seems to have got off the ground and then lacked a bit of simple polish such as the that got put in Maven Central being a release that this issue likely to be resolved any time soon? I see no good reason to treat the removal of the custom as something that must be postponed until when if anything it will just continue to cause problems for people wanting to use the public release. +I just committed this +Clebert I dont see the actual error listed in this jira anywhere +The suggested fix +This is by design +Often a framework of the size of HTTP could have been large and thus much work to do +Leveraged the Teiid class which was overriden with the Teiid version +We just encountered the same problem which is gone after switching to Hibernate +Im reapplying this change to my local machine will do some testing + +Please refer the attached file +Bryan what do you think +yes not even sure axis support in WTP supports anything but tomcat +Ill commit this +If a version we maybe should do that in another issue +found that a job would fail if the process that started the job was killed +Merged to branch at revision +Just attached a slightly updated fixed in both and trunk. +Nigel I am not saying that throwing at the command line is not a bug +Hey DarynIve not verified this by self but does trunk remain unaffected by this issue +Will start working on multiple outputs for newer hadoop for written by extending the class and provides apis similar to which allows the features of for Avro +Nik +I say should as I cant seem to install from the Update Manager anymore +Problems will occur when some threads try to use a instance of +It seems that there are no issues using bootdelegation at the moment so Im closing the issue. + +That makes sense +However Ive added the missing documentation piece to the trunk +This isnt technically true since the files are locally stored +Just to add +Patch is updated +Do you have any of the instructions for how you deployed VCL on Grizzly? I am working on the Massachusetts Open Cloud and would like to do some testing with VCL on our setup +Thats exactly my impression on all this too +This issue doesnt apply anymore to Lucene Directory users since is effectively a workaround for this but I think it could still apply to other use casesStill now that I know a bit more about Infinispan internals Im wondering if this should not be documented as a limitation of SKIPLOCK instead of being fixed as a bug in absence of a lock Im not sure how we can mandate proper orderingMain problem of having now this workaround in Lucenes Directory is that we have no more a test case +Unfortunately logs are small due to a bug in the way theyre grepped +Looked through the dependencies +Corrected broken link to pax patch of Christiantrunk +Early version of the patch introduces state machines for activestandbyI will update the design document in with details and to this jira once the implementation is done +Partially need to look at the creation through a JBPM process if not using jBPM +The encoding actually goes through the properties from the schema +Yes sounds all good +That patch does not seem to fix the issue for me +One minor correction to configure without sqlite +Verified a scenario where i had multiple level indexes and was able to get the correct midkeyPlease review +Please reopen if you still see this issue +I removed the earlier attachment +Aah got it licked now +Setting to future pending a patch +There is now because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Thanks Giovanni for the report! will be solved in next release or you can build a snapshot of latest master +Apologies +I dont think that should be blocker +Looks interesting +Commons logging as an implementation has its problems but the API is good enough for our purposes so it doesnt make sense to break anything +Hi BensonI thought about itThe reason I didnt was I started to wonder whether in fact all super types that are in either package or should be excluded? I thought Id better leave that sort of decision to someone that knows a bit more +Ill deliver this to the and streams first thing in the morning tomorrow +Simon might it be related to the usage of a new settings API +Please check CVS to confirmI would appreciate further patches for Equals and builders if you have a chance +AJAX requests cannot be on form +spec +Certainly! Ive added your change along with a similar check in the toolchain to require a minimum of of Swig if language bindings are built +I added a simple test in that tries to fire up a spring configured server using the xbean stuff +I can try to make update for +Fixed in r +Alls well. +Attaching report from derbyall with patch +Does this help? +Looking good +You are right +It is also included in M +Im not sure if the failures are to do with test set upconfiguration and not the actual test? Logs are all showing the similar output to belowPass Fail Error Apr PM Finished javaschemapackagestestCase Failed +One minor comment should also be final +ant results for both mapred and common return all s +Im not sure if there is a way to check it using JAXBAnyway I still think that should ignore the generic type since I dont think that it somehow changes if the object is readablewriteablebq +I had to add a wildcard property to the and have parsed as an unresolved element of the wildcard +and I think that WW needs in mechanizm for configuring builtin themes +Ok that sounds reasonable +Theres the chance that people might accuse of being sneaky to improve benchmarks but really thats kinda what delayedcommits is doing +For Arquillian we would want access to Servlets Name Class Mapping Name Remote and Local Interfaces Endpoints Rest Endpoints? Security Domains? etc +After looking at it for a while I have a general idea of the problem but the fix has a high probability of breaking things further +I find the interceptors that are on and other framework classes to be invaluable and was disappointed to see that the does not have one +Only working with one may need this +Just think about users who wantto replace one particular rd party library because of a bug fix aperformance improvement etc +We have nothing to distribute +Thanks Natty! +classpath changed +and ROOT on separate servers. +Ive addressed this issue in the branch +I wouldnt say so nick +The relevant difference I can think of is that we can cache only the identity of the item rather than the item itself but Im not sure how often it is useful +Duplicates which has been fixed close all issues in resolved state that have not received further comments in the last month +Thank you for your improvement +r also +Giving the snapshot a go now +So if there is data it chooses to disregard it should be be done at the recommender level not the evaluator level +Committed for nutchgora in rev +Patch above is good for for +Efficient of snapshot and materialization of snapshotsMany of the things you are repeating again about space and creation efficiency etc +That would be nice +I will need to supply the necessary hierarchy reorganizations and the changes before you will be able to post anything like a complete patch +pass and derbyall tests +Committed copy of current trunk book and site to branch +Fix javadoc warnings +sorry yeah its been fixed we still needa tutorial though +Without the library many tests in node manager and job client failed +Ive attached a zipped keynote file +I got this one implemented now +Reference detailing motivation on this issue Bloch Effective Java Edition Item changes in comment +commited patch with revision +I intend to run the parallel testing patches on Windows tonight and report back tomorrow +The blocks that have only one replica have the highest priority followed by blocks having less than replicas and then followed by the rest of the blocksThe data structures that suggested by Konstantin has the performance advantage +Thanks Jarcec for the contribution +batch transition resolvedfixed bugs to closedfixed +This is doable but involves an RPCWe anyway do an RPC to get the right? Isnt the length always present +Ill leave the subscribersas is but have moved archives back over to where they belongThe other requests will be modified accordingly +We dont maintain serializer code in more +This is a patch that can fix the atomicity violation bugs +The patch attached is much like Johns suggestion except that it doesntclone any properties +Added a test using that to expose the issue +isnt using a so it doesnt appear to be thread safe +added to release notes as resolved The jboss field of the class is no longer declared as static +closing even though I didnt get any answer on this its not an issue. +This is fixed as of r +we will need to check this in for people who dont want to install the C compilation environment as we had discussed previously +jruby S jirb would be correct but were still missing ENCODING support in the parser +Add unit tests plus fixes a small set of issues introduced in another checkin +hamcrest is a transitive dependency of junit well need to exclude that specifically in our poms +could you give it a whirl and let me know? Thanks +New patch dequeues p after the write completes +NEKO version downgraded to because the latest one was causing issues with RF unit tests and in runtime. +If youve gotten this far then is no longer an issue +based on current CHUKWA just committed thisThanks Jerome! +Stuart please make a new issue if there is still stuff to be done specifically for those changes +Thanks a lot for your valuable help Adrian! +The patch makes sure users cant set slow start to correct so it gets rid of the exception for jobs? I dont think is addressing the issue of starting multiple stages at the same time +This could be an ongoing tasks for CR +In fact we should make the public and not package private +Using this jira for very basic reuse within the same vertex +Or may be enough disable gwt compilation for this plugin +Patch for +Will commit now +Seeing it completed before GA is a good goal +Thank youGood luckLeo. +Verified by Richard. + +Added to release notes as resolved issue A backwards compatibility issue with serialized messages from the ESB being received and read by later versions has been resolved +Now I have another kinds of errors but I will study to it tomorrow +If you find further issues please create a new jira and link it to this one. +Oops that should read This task is now complete +See for more discussion +sample file to run +This was logged running against Final which is what in Teiid Designer trunk +You mean rollWriter? I meant the fat log message we make under the log +Im unable to reproduce this +Nitin can you please fix +Sounds reasonable to me +What exactly has to be done +Fixed in close of all resolved issues. +Sure thing i think it would work now +I do think the fact that splitmb sizes more in keeping with HDFS block sizes should be possible but are not +Cf were created in and cql describes them just fine in cqlsh in you creating in though upgrading and then describingedit i should clarify these column families were created prior to on a variety of versions +I missed this one +Closing at this point with no further testing or investigation so we can include in the release. +We will also clarify this in +you can estimate how muchyou will need by looking at the log directory of the db that had the problem and see the number and size of the last log file there and multiply to give approximately what all the log files will take +if they are valid the job is relaunched b +I dont believe were planning on rebuilding the installer for JBDS anyway based on earlier conversations +We must use the same mechanism like ejb and war plugins +Patch to project to fix the have attached a file with a workaround and granted Apache the right to use this patch for a future correction to the issue +Tests how many files the system can the changes Michele please test it again and confirmThanks have same problemhow to solve it. +Same problem with doesnt seem safe I will wait for some input from the Swarmcache team before I change the provider. +Lars and TedThanks for your reviewsI was also not convinced in adding the instanceof check for NPE +java and +joint TEST confirm I have the same error. +The problem was that we didnt mark all of the catalog entities as +any news +reran the tests in Eclipse with that jdk as well as a sun jdk and openjdk test worked fine +This would be a major pain and clearly its not going to happen +Mostly simultaneously +Changed issue titleSo digging the serverName we find when we scan is that of a dead server since processed but the region has just been opened on its new server +Not so fast still needs tests +bloom filters to implement distinct +Added custom collectors to +I did have the +the file handles should not have been open in the first place fixes that +resources +So what are you suggesting here Feng? That Talats patch is OK or That there is more required here +lets see what JIRA now provides special handling for plus signs in released in. +Checked in parserchangev patch +Patch has been applied and committed +tests are a bit broken just now +Rebased to trunk +JianBikas shoo get off my ticket and move onto +This is the situation that the test case is causingIn this case the global temporary table has been declared to not be logged and on commit to delete all the rows in the table +I also confirmed that turning off invokedynamic fixes it +clean AmosPlease can you backport this to the branchThanks! need to change dependency version to and XTSlocaljunit dependencyI have backport to branch and it needs ant in PATH and export JAVAHOMEIm not sure it needs build as I can not find the test commands for xts unit and interop from httpalbanyviewNarayanajobI have updated the config on the jobs to use this new script thanks AmosPaul please can you give Amos a pointer on the xts commandsm in the process of getting the commands from Ivo to automate the XTS unit and interop tests +Marking bugs with invalid fixed for release as fixed in. +Thanks for the explanation this is terrific design Ross +Understood will update the patch shortly +Please let me know if you got it and was able to reproduce the problem +This appears to actually be an slfj issue +Have you tried the export JAVAOPTStrue while this option is on Geronimo will not scan all the classes in the classpath +Adding JVM heap should not be a problem +But I think we need a new test for circular referencing regression +Please verfy +Patch for the GUI stats handler to follow +yes +Sometimes refactoring or redesign results +To be completely safe with them wed need to copy only the part that is suitably licensed +Ill attach a new patch that follows the suggestions from and shortly +If this feature is done can you change status +Dan I suggest you take this to the forum if you are still having problems. +patched jar ready for testing +Juan do you want to take a look at that +You dont need to extend it right? The patch gives you a version of what your patch did only wo the special data member and the extra setter? Thanks Micah. +I cannot reproduce the problem but am changing the status of the issueto ASSIGNED +Kill the attributeIll take care of for the process train since I know its code +Thanks BrunoYour patch is commited in revision +The code you gave explains it perfectly back to the original point of this bug report On the close of a resource why is it necessary to delist with TMSUSPEND instead of TMSUCCESS? The code Andreas just supplied indicates that would have no problems with TMSUCCESS but the comments in cvs indicate that it was that instigated the thoughts +So its correctly fetching data for that module but same result as I was getting on parent project +I have attached the patch for Message Queue Component for Nuvem which includes the and the relevant interface implementation of the interface to and the test cases +Im realy quiet for the moment but i am still alive So i wish a happy jump into +Once the metadata was fixed the query ran fine +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +Integration the beans discovery with Geronimo artifact deployment +Thanks David if you could pull from github one more time before turning it over that would great +The should not be tied to neither core nor SOAThe problem here how to handle different setups for at least full releases and milestonesnightly builds we could handle with a system property +I cant reproduce this right now and the implementation passes the tck which tests this as well +Someone suggested using EMF as the modelling framework +batch transition resolvedfixed bugs to closedfixed +Close enough +In any case I will file a new bug with the newer patches when Ive got further down the road +This is a pretty simple fix though it was a tricky bug +It occurs from not enough memory left when you take all of the servers RAM and subtract the heap +Thanks Chaitanya and Rahul. +Thanks +Fix attached +I think I have fixed the issues you described in our chatI have created a new update site and attached a new sourceI would have one suggestion adding the Select allDeselect All button to the Edit and Search Runtime Path dialog that would enabledisable all runtimes +No tests since existing tests would break if the simple interface change is incompatible +Moving to since we dont have the necessary infos to reproduce and fix the +the patch on the trunkThis patch applies fine on my machine using patch p i +Nice idea we just dont have cycle for it +If you could easily replace Kerberos with a mock all of the unit tests can be run with security enabled +Oh and Im using a Grails from build server as of today +I have applied your patch but it seems the stuff is missing as I couldnt find for having one configuration type per container its a limitation we can lift +I removed it and error is could you take a look at this +For samsa module fix is more complex because of using lot of copy instructions in ant build script which do not support if clause +No response assuming ok. + Bug has been marked as a duplicate of this bug +should be fixed in the JPOX SNAPSHOT as of AM PST this test still fails on application identity and succeeds on datastore identity +Fixed guess Im simply not entirely convinced by the description of this jira although I have a guess +Im not exactly sure of the best way to fix this +TTMR fails for me on branch without my reverted hadoop libs +Likewise! The problem I was seeing is gone in the current snapshot. +already has the info duplicated can you please remove it and retain it only in Environment +Once web services have been changed to use the existing AS injection framework CDI injection will start working automatically +Yeah feel free to reopen and edit the title to be more specific to that and someone will get on it +note that not all blobstores return an md on GET +Basic testing of the Viadeo module confirms that it will work with this change. +Will check this in once Jenkins gives its blessings +for option +the version field of module type doesnt accept a version will add unittests for it +DougThe rethrows the error +Thanks Marco for your comment +Ok committed to trunk and! +Pull it back in if you think different +Would you like to me to provide a test case to help facilitate a fix or is the description sufficient +For and we will get rid of them +Why is the verify method getting called for every app? Shouldnt this be only once? This could explain one part of the gazillions connections we are creating to during +If not Ill do it for +The result of the function is a large integer +Can you move these tickets out to Future since these do not meet the criterion for inclusion in +Prepared fix for displaying images as documentation notice +All work fine +Patch for applied +For example line is in the middle of a comment in current codebasePlease redo and post a new patch this is important +I did but replication to m repo seems to fail +I still havent ported all the ones from +There are times when I would also like to build my deps if I am working on asjmxopenshiftblahblah but this is not the usual situation +I faced the same issue while creating instance using API where component has its own instance declaration too +Revised the patch its OK +BTW There is too many readwrite operation +This is an IE on codehaus sits behind an Apache proxy +Can anyone from Spring address this issueThanks were using Spring Using Spring has finally been addressed now for Spring Framework RC along with the changes in and will also get backported to +Harness n bloat Dacapo version reproduced the issue on the build r in Xemserver modeThe failure is not reproducible in Xemopt mode +If the problem persists please reopen +I am supposed to get this fixed this week +Id go ahead and pull the fix and resolve this to master +Jeff Turner Feb PM Im looking for a way to send anonymous email from a website page to our mailing list Why would this be a good thing to haveIve had feedback that people are not posting to our mailing lists for various reasons shyness not wanting to be seen to be asking dum questions not being seen to not know things they feel they should know not wanting their or their companies name email address to being seen in the archives etc +Not yet thanks for suggestion. + +I just dont think you understand how open source works Jan +still lots of work but its a start +Hibernate is upgraded to in Sonar +This is fixed with the latest JPOX SNAPSHOT. +open the domain for transfer and send the auth code to root at you already get the auth code fromIf not I could ping Janne Jalkanen to explicitly ask him to do sokind regardsHarry +The config option is working for me now so that has been fixed +Bulk close resolved issues +Ill keep looking into it but at this point I want to hold off committing this patch +trying to run hudson just committed this +Committed to trunk +The test case should do the following Submit jobs as a single user + +The described with some functions removed for simplicity and my test file +Thanks. +I am closing this but would like to see the test case converted to use the from the test client to locate the services under test. +Consider the case where only one thread modifies the map +Patch committed to trunk. +Im not sure how to edit it +The core sort class is adapted from one found in the Project Voldemort +Yes I am working on a patch which when it sees old +Argh I missed one place +Thanks! I patched it a little bit differently but used the same concept +NO SQL is hidden now +Add some javadoc around the API to say that a duration of means that the servers configuration of stale interval will be used +Im having problems testing this +Again this is awesome! Thanks + is the original patch from Coreys pull request +Fixed +Thanks Laura +The appl flag that Steven recommends makes sense +But now that I look at them they indeed look like a working module of rolling followed by a working model of recovery +What do others thinkIn a generic topology theyre essentially different levels of +I agree with Robert and Shai on this issues and that other folks should just pull out the class they need +What do you think? That way more people can add patches as necessary to style it +The websphere entries are added to the generic factory +Finding it based on the error message thrown is some trouble especially given it works when the injection is in an injected child +Closing as wont fix for now can be reopened if needed +Here is the required patch for this issue +Updated pull request. +Im not sure what to make of thisI would have to agree with you thereI think I did a test recently just to generate B entries in a table M rows with and in each row +But I can be patient on it +we should explicitly support them in the protocol not as string hints +In my case this prints false when called from the ll try to look for it a little more comparing with sample project +You can look at for some of the implementations of it. +I added a comment in the parent about this +Its basically only extending what you get but its not hiding anything. +This problem gets aggravated when is in use +Apparently Simon fixed this issue with another Bitmap recycle fix in the CameraI run Cordova but with Transformer Prime I get exactly the same Crash log mentioned in the description of kzusyWhat can I do +That said it would be a good idea to recheck for process beforeafter attempting to execute +I dont remember how the rules for ASF committer access work on a basis but if it helps Im already part of the ASF +Need to ensure that this does not cause any testcase problem +created to track refinements +Ted committed this a few days ago +How would getpagedslice need to be to make that sane and is there any short term solution to get this fixed for +The counters need not be maintained by the at all +JPOX CVS also now creates wrappers for serialised fields based on the declared typeThis results in the test passing +if the first character of the termBuffer does not match the automaton it gets rejected +So with this in mind there is no need to support +tgz format appears to have the proper permissions +Then when you hit the second webapp it is seeing the first webapps cached resource guess the only way this could happen is if returned the same classloader for both webapps +So CS is using the same path as the host for the datastoreLet me try to get it to work +Nice works for me tooCould we make the URL a part of the input metadata using the RESOURCENAMEKEY? This way Tika would have a better chance to detect the type of the document +Hi PhilThanks for your response +When that occurs the blank namespace is all the gets to work with +Some information about the usage of panel tabs is described in see no changes in +OK +I ran the offending pig and it seems to work fine +Joe is there anything else missing from our side? Thanks in advance. +changes look fine to Did you test the patch? Can you post the test results +The indentes did not get saved and I do not appear to have edit privilges for my bugs +Would srctestjavaorgapachehadoophbase be a better place to hostThe reason being that it tests both mapred and mapreduce input formats +Of course the interception mechanism is left to the interception technology. +Forgot to generate the patch under the trunk root +Sorry I am a little bit late +thank you for providing comments +Jeff when Nick fixed the build issue you can try to use the trunk build to test it +Cant check it right now but I remember that trying to scan for ranges that I knew would be on separate tablets vice the full table resulted in the same issue +Initially these annotations were defined not specific to JPA instead can be used outside the scope of JPA +Sadly I was wrong about the get +Updated and corrected patch +Both actuallyIt called Cant disable attribute splitting so unable to disable splitting was main this applied to the experimental branch +This was causing problems during hot deploys where the service name or service category changed and duplicate were registered +Sheng the table has to be as a part of upgrade as the feature was missing in +Patch applied thanks +committed +The test case was already disabled on Java and higher but the patch also removed the codeAll the regression tests ran cleanly with the patch +Checked that doing a CtrlShiftO in Eclipse puts immediately after Map +or the samples copy their war files to Tomcats webapps directory +is OK msgs is slow msgssec! So the regression must have happend between and comparing the snapshots now +See also +The current solution is a massive improvement and is worthy of a minor release +Adding style planner conversations will be investigated later. +Change deleteLedger to versioned delete +Was a bug in eclipse and not in meclipse. +Is there any reason why the H support cannot be fixed +That is the solution I have adopted by using subject in that particular view +Thanks Namit! +Mark thanks for your contributionSeems like a to me and all tests pass with the patchId like to commit this unless anyone has objections +Please use the list to ask questionsFOP currently cant break blocks larger than a page which should be kepttogehter +Hi JrnWe have the and in some tools like +I had started this a while ago but hadnt finished it +Vote thread has been started for a month but branch still not merged with trunk does that has any blockers here +Added new tpstats command to after reformatting braces +This looks greatPer our conversation Add in upon application of Spring Nature for an existing project Above should pop up a dialog allowing the user to customize which files are actually included only if there are any candidate filesThanks. +I uploaded a new patch that addresses Amars commentsbq +bulk close for +I had a quick look and dont see any issue in abstracting interfaces from and so that implementations can be customizedIll provide a patch for that +good job mahadev +This enhancement will be considered for a release after Struts +I will commit the document patch after some review +Could you rebase the patchThank you +Checked that patch in for Daniel since Id already applied it and tested it +Could you show me more of the log? I would like to know where it get confused as the for the should not handled as tag +Looks good +Ive committed a first version together with a test case in revision Ill leave this bug open for further this to fixed now after some further testing +All handling happens in +Committed in revision . +generate is a bit unreadable +contains fix and the fix for the second problemThe second problem was I didnt mark ebx and ecx as clobbered on testes the fix with before submitting a patch +reconfigure to be a regular component taking factory configurations and requiring configuration for SCR to create after release +Why move it to there when the proper location for the global setting is through the menu be annoying to see that icon show updisappear everytime I switch editor +We are doing some timings and we may reduce how often this happens if it shows up in profiling +Closing for now. +Hi David does this feature work for you +If you think they are critical for CR please them +This should come before an iOS implementationWhen youre ready for this to be merged please file pull requests for them through github +There is more than one way to solve this issue +Should we go insteadWorking on a fix for +As far as I can tell the behavior in the AST view is no different with or without that extra line number in the class file +Ill let you know how it works out. +I am not sure if this is necessary +So stay by +Feel free to reopen with more details if you encounter the same problem. +maven is now the official build tool +WSDL file hope this was the one you meantI will take a look at your bug solution proposal for my second error tomorrow +I tried times +Just that he has to do thatOkie if there could be some use case with post hook also then on patch +That would be a nice solution +Need new templates for them that provide x file line for Main template all XJava debugger issues +This is NOT a blocker because the correct solution is to use isolation which will be on by default in +Lets have this in patch for +Hey Kristian Sry I think I misunderstood you earlierI added the to surefire not and got this error again todayDid you mean that this also needed to work correctlyUnfortunately I cant switch to just yet as I am locked to using release only versions ATM +There may be kinks to work out when actually performing the release to the Apache Staging Repository but someone with LDAP access and the ability to GPG sign artifacts will have to drive it +Ive attached a patch that fixes that to make the be +Thanks Billie. +Linking to the Akamai issue as usage of akamai will add constraints to how we deploy In order to leverage Akamai benefits files has to be in the static folder +Added test for new serialization +Im just testing prior to applying it as a patch +Thanks +Moving to GA for now but unless we get something back from the JON team we will have to push this further out +Yes this is the reason why it was never implemented I am currently busy with other things but this issue is on my radar +The page validation was taking me back to page +in this worked didnt see this happening with the provider and Groovy +I had to resore my system and tought that my local file version was the most I noticed that in the online documentation the imagemosaicplugin tutorial is missing in the Tutorials module +However it is a change in wire protocol so I suggest we file that JIRA and handle that part on trunk +Adam and I are dreaming up a task manager sorta like cron that lets you define arbitrary actions to be run on events they could be time related based on stats counters or just continuous +Let me work this out and upload the patch +I had to use the workaround. +v should apply trunk +Needs a unit test +Remove them from the plans for those test cases and replace them with the correct operatorI think is better though its a bit more see failures in and when I apply the translator patch +This is a duplicate of +This zip file contains the first version of the definition of the services and their is long gone so guess this can be closed +Committed changes in Rev +Manik if Rado adds TRACE on that should include too +In the first case the java class would be used and in the second case the compilation would fail due to a missing class +I dont use ADF either +Deploying partially copied files is poor behavior it disrupts the server +Now I added more unit tests which catch this bug and exercises larger portion of blocking encoderSo the new patch has the original change Ravi made plus the new unit tests +Sorry just got back from vacation Im currently buildingrunning tests on RC in Bigtop will update you on Wed +has been released just three days ago so will take a while +But that is a good point +Hello DeepeshHere is the patch for automatic cancellation of replacement order if return is not received within daysFollowing are the steps to test this patch +Is this correct +I also think it is unlikely that a given application of clustering would mix sparse and dense vectors though it would allow us to make the particular encoding be automatic on a basis +You can make a formula for style width to calculate the width based on the current scale If you really want what you describe you would make a double line river and then switch data sets as you zoom out to replace the double line river with linework when zoomed out far enoughCheers and thanks for the suggestion. +Set maximum heap size to make sure the hash table will have to spill to diskjava jar ij far as I remember scrollable updatable result sets may also spill to disk +Thanks for the patch Michael +Reasonable improvements have been made to pick up binaries from known locations within a linux environment or when building and running from an svn checkout. +should work for usLatest patch looks good +Why dont we use the trycatch to just whatever exception might happen so we can handle it gracefully upstream +Ales is there a plan that backport this fix into EAP +Use local to refer to the +Ill be careful to check these when we get the next code everything as is Ill take the hit this time around when I merge the code drop in +Found in moduleA and moduleB +but for testing I am thinking of using reflection to add the from control channel socket to JSSE providers cache based on host and port +Could we avoid creating duplicate jiras +Done +The patch looks good +The source of the problem was that a Java version I had installed on my laptop in opt was not installed on the QA labs +Though the symptoms of the two bug reports are slightly different the root cause of this is the same as that of +Then anyone could build the tests from with required changes and run the junit tests againstSince branches are cheap in subversion this should be easy +Javadoc and source jars are excluded when we create the aggregate +Its our IDEs not being able to resolve the sca schema +This is the log file of the build outside eclipse from Shell with maven +However a user can overwrite this values by using command line argumentsIm happy to provide a patch based on the result of our discuss here +And here things go right +It is illegal to set prefetch to and set message listener +It is happening because is declared in site line contains +reopening to change version. +For example uses stat cookies to store access rights cookies for and you cant login with +I am reopening this issue to investigate why the xml beans code generation is producing over files almost all of which are the xsb binary schema files +The given bugfix works perfect for include in the next official of fixed +Reaching out to some Sybase experts now to see if they have some alternatives +More generally I dont think Im really a fan of breaking out linear logic into a separate function anyway +That also explains why there are no test failures it only occurs when deploying the service with +A value of for this parameter means compute the delay based on heartbeat interval +I added a new test case for this behavior in +This issue has been fixed and released as part of release +simple patch to fix a basic problem the fact that a thread ref may be null during shutdown +You know the project much better than do you have any further comments on this? What exactly is used from adaptx for the test framework + I have ran for times without seeing a failureFor the first patch I applied through Eclipse +bulk defer to defer to +It would give the impression of having closed the other result sets opened in the procedure because gc would in most cases close them implicitly before the table was attempted dropped but not alwaysThe attached patch makes the procedure close the result set explicitly to prevent the intermittent failure +Lets add a wrapper for get something like get which is akin to get so all the callers dont have to pass the two keys +Everything seemed to be working correctly tome with the XML Junit tests so I committed to subversionas revision +heshamWhat exactly did you do to create such a pdf? Can you attach your sample +This is implemented in commit on trunk +No other changes to the code baseHoss Do we need anything more in this patch +This patch causes the kssl static block to only change the ssl ciphers if kssl is being used and enables a much wider range of ciphers +If it passes I am planning to commit this patch + +Verified on Windows and OS X +Ive added a new event handler to address this +Thanks Edward +Hi AaronNoted +M released +Moved the exit condition out of the OliverCould you supply unit test case to reproduce this issueThanks in advance +See +As of jline is now +Patch generated against the branch +The pointseems to be the in a +Not really they are only viewable from within the admin interface +If something was added maybe there ought to beAlso I just recently added creation of which is subset of core that excludes annotations but is otherwise compatible with standard core jar. +of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Here is what I did to run into this errorBuild syncope from trunk +I think things are more in sync then there were though +There already exists dependency from in vpe +That would to come late with this but I just realized that it would be nice if the Javadoc stated explicitly that the declaringBean parameter may be null if the annotated member is that to the javadoc +So I handle this scenario by removing all roles from the user +Steve Give these css files and images a try +Merged +Im probably be forgetful but +For the release we could just update the parsing of the metadata used by the to pick these up from the if its seen +Hi DeepakYouve found the way to go ahead thats perfect +Tests are broken due to +As is out I guess development for is this patch be applied now to the new branchll take care of Java enums with than creating a new issue I made up my mind to reopen this one. +First attempt but things are still not working as mind we just dropped the usage of from normal builds only the release needs it now. +I just upgraded to Fedora will retry and see if its still valid +Let me know if I should hold off on the commit until the python issues are resolved +Resolving +I tried to delete the stored annotation viewer preferences but it seems that there is another place where the preferences are stored since the result is the same and the entries in the style editor are still availableWhere do I delete all the stored preferences for the annotator viewer? +Final patch +Thanks so much for the unit tests +Dont we have already enough problems with Jisp +Thanks! +Please try with a Roller version +Fixed and IT added +I think we know the root cause will update with details soon +I did polished the code a bit to ensure it comply with CS and cleaned up the to set scope to test for deps which are only used for testingtrunk I thought I had to wait to sign up until the ICLA was acceptedIve signed up now using stevebate as the user ID +Thanks +Been running in our prod cluster for a couple days without issue +Closing tickets +Patch applied at +I still cant reproduce this failure +Also Im not sure if it is too late to change the resource bundle keyString +sure +Bumping issue any news on this side +Please use the pom file from issue log file shows two bundle uses class from meclipse and this class has been renamed to some time ago +if ZK properties change it should not restart other patch + +By default a programmatically defined SDO property is an XSD attribute +iow allow specifying start and end keys +Well it shouldnt be build another mgmt interface but rather converge on a common one. +extracting from a directory much closer to the root of your filesystem should solve your problem +it wasnt clear to me from going through the patch how the is being parallelized +closed +Hadrian as Christian intends to cut the last branch release by this freiday thats not much time leftI believe the provided patch does the trick but the last proof would be using it while doing a dry run of the releaseI think we all would love to see a proper HTML PDF by the next release tar ball +We could maybe start to add JDK only components to Camel and have profiles in maven to only compilebuildtest if running under JDKWe do have a CI setup which runs all platforms on JDK and +patch to fix issues hopefully in the correct format +Slightly updated patch Flavio I believe you do need to instantiate as is not static and therefore cant be instantiated without an enclosing instance +The jbosgi bundle version is actually the servlet api +Attached test patch +Many thanks Markus for contributing it +Applied in rThanks for the report and patch! +The two should be the same +For beans of type we should handle target beans defined via the property target too +The patch failed these unit tests in The patch seems regressing on Linux +is not a config property that is why we removed it from the +You have to handle the case when timeout value becomes negativeactually I think the get method has logic errorYou should update timeout when catching exception not the other way aroundThe easist way to implement the logic is to measure the elapse time difference when you catch the exceptionIf the elapse time is bigger than the given timeout then throw the exception +One more point which I discussed with Vinod +Do kindly reviewThanks for reporting it Trent +From my experience working with Bean Validation are quire well designed in order to provide interfaces and minimize the actual wiring of implementation to couple of placesI know the Hibernate Validator specifically says its Apache licensed +How do you define progress on tracingId define it as us making use of the lib +Moreover as a result of these bugs I changed Roo to specifically use the dialect in the one thing that might be nice if possible is to warn about this when performing the DBRE task +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Moved to after Beta is tagged reflect changes in EJB Core as Module in AS will not be ported now using the Plugin +Tested and its working +I have attached a test case for the problem +Make configurable through Configuration +A +I believe that fails because it tries to bind to the same ports as the used but did not releaseYes I believe that is what was happening. +All new related feature requests and bugs need to be logged against new Jira tasks. +GO GO! +Committed revision in trunkCommitted revision in nutchgora branch +Will upload text versions next time +albeit this is more work but I would think it would be well worth it so that other developers dont need a seam runtime on their box and can just check out from source control and start programming +Block report processing should ignore blocks that are being written to via. +Can someone with privileges please add Profiles to the affected components to make it easier for others to find this issue +In trunk we have +They had to be resolved for. +You are absolutely right that the changes should have no effect on a correctly written XML Schema ParserThanksdims +Please reopen if needed +Can you generate a patch that is less git specialized +The patch to exclude the test on other platforms is attached +Sounds good +The new ones will not have the do not cache flag set +Ive committed some cleanup changes that should make it easier for a consumer to add default xpaths for their custom server type whether or not +this is wrong waya right way here is use reference to with absolute path +HeyCan someone take a look at getting this patch into trunkThanksJeff +bulk close of all resolved issues. +has requested you rebase +That would make it easier to switch between different versions of thrift. +Its behavior is back compatible with but we no longer return it +It is old and obviously has not been with the later from Sun +Argh +This should explain restrictions in general and then per component if applicable +MarkusThats great news +So for him and me thank you for your work +DaveDo you have more patches coming for this bug or should we close itThanksDavid +From that perspective Id prefer the behavior +Can not reproduce +I will check that but may not be today +Yuck +Ive opened to remove the limitation for IN queries but that part has never worked anyway so its more of an improvement +Currently using the snapshot and many of the problems are gone +done. +Updated the diff on RB to the latest trunk +open a thead on the dev list +coretargettestlocalruntest +Assuming this is a issue with connection pool I reviewed all the settings that we are doing and these look fine +can not start with a letter so +Well Samuel youve made your case +so this issue is still fixed in Beta but just not verifiable right +This is surrounded by conditional compilation so that it is only turned on in the specific compiler and versionSo the warning is still active and reported but wont cause a build failure +problem reported to be fixed in +Fixed the +So you could this extra codeIs this intended for a commit? or just food for thought +I didnt see any reference to lostremoved nodes in the RM log +The expectedactual values are backwards in the asserts and the test belongs in but other than that this looks good +This also copies the DB driver to the default server and not the production server +It sounds too restrictive to me if all of s cells would be the same size +But beware that this is not supported and no guarentees that itll work in any future releases +Ill take a look +Look at +Thanks! I committed your changes in revision and followed up with some further improvements in revision +OK it works for Jackrabbit but Is it JCR spec compliant? Would it be possible to change the application? What is the impact if we dont provide compatibility with Jackrabbit hereProviding Jackrabbit compatibility should be possible but I would rather not unless it is a really important use case +The question here is why there are bookie failures insideAny logs regarding to that failuresBTW if you are running it throw eclipse you just add the below property in eclipse srctestresource file +Committed to trunk and +Costin your thoughts +Jason can you take care of this +Thanks Chelsoo +fixed in to the branch in revision . +Cheers What url do you want in the REJECT message +I uploaded a patch created from trunk which I believe resolves the issues +Marking it as a blocker pending further discussion +This is a one of the core jars from the AS +If this can be done quickly lets do it +You should never actually see these in an +we can leave it open but marked resolved until the next release +Added patch to fix issue in tomcat +One small change we only check the first map value for map and the first record for collection +Please close Dave if you are OK with the fix +Is this issue reproducable with one or more of the following formatters too? bug is in so it affects its inheritors and as well +The enclosed patch fixes no less than three different concurrency issues in this class +API coverage as of Aug +Odd enouth I can not reproduce this issue on my machine either +I would like to stress that the test case does work against +Diff file after compatibility is not currently a priority for HadoopIs there a platform you require where Java is not available +Committed to trunk +No need to add new test committed this to trunk and +Approved +I just checked seamspace anyway +If it is then it needs to be converted +I think this has been reported before but I havent searched bugs +There is no response here for a while +This finally means that an aggregator Mojo would NOT work as expected when using this do NOT want to start the discussion whether this was a misleading design decision or this initial issue does NOT make sense and the intention might be addressed by + to trunk +Hi I am unable to reproduce this on AG trunk +The patch will remove document generation logic removed xdoc directory at the root level removed and edited the to copy plugin jars directly from the modulePlease review and apply +Tests are also is no need of applying any other patches +Security what about having whoever has write access to the file can do it? This would let the owner of a file check itI dont think so +Here is another the hash code in the constructor is a good idea for this class! The code is thread save automaticallyBut I would keep the hash code computation in a separate method to keep the code clean and comprehensible +I vaguely understand the NOOUTPUT optimization +The patch includes fixes from the previose patch +ramCould you give some comments correct me if anything I didnt consider +Thanks for spotting this Alef fixed! Unfortunately this meant adding additional state to namely the root object at the top of the path which is only needed for proper error messages +So the exception is based on configuration not data existence +My gut feeling is that the instability is caused by slowness on the Apache Jenkins servers +My apologiesThe svn stat command produces this outputM srcrefAlso attaching the HTML output file +for reporting it creating a temp file every time is not very optimal but if no other choice +I ran similar tests as the original poster and in development mode its very easy to go OOM in a matter of minutes at a pretty low messaging rate similar tests in production resulted in heap usage remaining at a plateau over a matter of hours +to trunk +I will that +Technically I dont think you should be passing in null there but I suppose we can check for it +If you believe the issue is not resolved please reopen or open a new issue +I believe it is still worth looking at +Escalating a reinforcement and Infra level is overkill first cut at a reference checking script +In CXF we have recently added a profile for Java which adds the dependencies which are only needed for Java +to making the annotation per class its too hairy to do it per methodWe just have to remember to move hog tests to their own classes so we dont lose coverage of any test methods sharing the same class +I get the same thing +Even if its unlikely to change if someone does want to change it theyll thank us that they dont have to change the coderecompile to do so +Old JIRA is actually a rather nice stemmer if you ask me +Double check about character escaping +John or Robert if you could please test the fix Id appreciate it +so i think the Logger and some other related classes should be serializablewhen i use no exception happensif some people migrate the logj from to they need to change some codes +Now Struts is logging this at ERROR levelI see this issue closed but I think it would be better to come back to log this at WARN level wouldnt itTemporarily I have avoided this annoying message adding this line to logj file because the message was appearing with devModefalse too FATALThanks +That would be separate jira thoughDaryn My take would be to hold off with a new Jira until we have a user scenario that justifies the need +Sorry disregard previous comment regarding Cannot join tree represented byThe previousParent property in my object was not being set there any progress on this issue? +Is that correct or should the component lead be someone else? Or better +Closing this one since it doesnt fit in with the current caja integration anymore +Moving out of +as long as it still allows me to set my own nameAlso what will happen in the case +thanks Jeff for the patch! +Good point +If you dont want people to make complaints when something that seems like it should work doesnt work then perhaps you should rethink even having a JIRA system in the first place +It makes sense in this case to use the annotationBut there was one more thing that I observed in most of the test classes that extend +The inconsistency is confusing +We shouldnt duplicate information in the portal and after all the portlet is what knows what modes it should support +Please explicitly state the versions in the dependency elements for the sling artifacts +If you agree then Ill attempt to create a patch +RE hardhat vs I could see an argument for using the mainline logo instead of the hardhat powered by logos tend to be the main touchpoint for the brand with things like httpd and probably hadoop so less confusion about which is the main logo might be better +Incorporated feedback from Stack +Looks good to me +Checked with Daniel and hes happy with the doc now +Good idea Randall +But not so sure now +So how do you handle timeouts in the case of stacked sync events then? Are you still preserving the order in that case? I guess i will have to find my old tests and look at your patch +Improved patch with comments and code formattingwhite space fixes +As far as I can tell there is no problem. +Works correctly in. +Any status on this? We would like to migrate from clover to jacoco but this issue is keeping us from doing that +It could also be used in the shell to do stuff like debugget debugscan and then it would be easy to debug performance of requests +Francis would you be able to submit the patch for this +Thanks assigning back to you for the next round +If we arent supporting secondary indexes for super column families in the near future then explaining how things work with is as simple as saying you cant use a filter on a super column family and throwing an IREI do think the clients should do a better job of supporting paging given the number of questions on it though +If it is on Eclipse first asks the previously opened editor if it can show new file in it +I understand but we dont know if that is caused by the fix not working or by something else as exception indicates that something was send to the node this could be a separate issue +no exception was thrown if we are here so it was a successI think the last thing should remain the request time for back compatibility and easy scanningThe result parameters should go after the input parameters separated by some sort of delimiterto be nice to machine parsers +This avoids slowing down the speed of the test response +In that environment a service like a database may be swapped out frequently in order to make the device responsive +Stan can you comment on this? I assume its lack of error reporting of the management layer +I just now tried releasing version of MSQLJ but I get a similar behavior as explained in description +The client would use the pattern I outlined above the only difference being that repositoryA would be a remote RMI or Davex connection instead of a local cluster node +I would rather not keep such general targets in jira +Rolled into a small point release with other fixes. +AlanI have attached a document describing the PTF Windowing Specification classes +The server then think that your are in the same session +Verified at r. +No one on the Cordova PMC can edit any pages +studioeclipseplugins +added Dmodule option halving test prepare time for single test +No sorry +Which means there can only be one and file +Changed run script +Unfortunately I put thewrong version on the bug report +EdellI made the updates you suggested +The title is not etirely correct Eclipse seems to work +I applied it to the LUNI module at repo revision rPlease check it was applied as you expected +Thanks Lars +Now that contains its ip address not its host name +At least the default GET handler for NHTTP transport does that +Ulrich can you check what Binary implementation class youre getting +Good catch +I would love to get rid of the monkeypatching but as far as I can tell if we start listening for the various device events wed be using up more resources +Added mysql support for bit operations and created a mysql specific translatorcapabilities +Thanks for the patch IvanJody do you want me to apply it or want to take it + to trunk +Fixed with latest commit for + of +I take it this issue can now be closed as the tunnel code has been pulled into the github repo? +If I understand it its still busted for meFORRESTHOMEmain cleanFORRESTHOMEmainFORRESTHOMEpluginsant cleanFORRESTHOMEpluginsant testThe todo and changes resources of the dependent projectInfo plugin are broken +Just to give some update on the issue +The stmt below that would be enoughThanks for the reviews +Make sure to test both in memory sorts and sorts that go to diskAnother issue that will happen if users use varchars instead of clobs is that varchars are going to read everything into memory while clobs are optimized to stream rather than read into memory +This seems to be ok now with jboss +Thanks Amareshwari! +Ill let a committer close it +JAASaction +I will keep looking for other Issues to see if I can cover any moreIf you can indicate some other issues more appropriate or priority or if these two do not seem suitable for this gsoc please tell meI have already mounted the environment Derby and I made some tests on it after building it I have the intention to over this week after some more testing begin make some changes on the source related to theA greeting and thanks +Updated Remoting Guide. +Perhaps they could be copied +I get a number of errors when trying to run the tests with this patch +I hope the current solution is enoughWe now use a local stage repository where we place compiletestruntime dependencies so that we dont bundle resorces retrieved from remote repositories in our releasesMaven Plugins are still downloaded from the main plugin repository and are not verified I hope maven will give us an answer to this topic but in the mean time this is the best we can do unless we dont want to copy MB of maven plugins in the stage folder +As yet untested except to verify that it appears to stop the connection leak when opening and closing a couple thousand connections +I wouldnt mind but it seems like Varun even volunteered to try out the last version of patch but the patch was committed wo waiting on his experience +Today we achieve it via stateChanged event listens on state change events and shuts itself down +That should work to avoid this noisy exception. +I always end up creating setters on domain objects only because of Springs data binding + Reverted the contents of to its state +Closing as wontfix no need to do this for imo. +We must supply every schema that we compileIn the m plugin I had to supply an entity resolver and a catalog manager in order to help xmlbeans find the correct local schemas +Previous version did not know anything about TLS so we cannot do anything else than discarding itThere must warning about this but that is it +Thanks. +I cannot login to Bamboo either with my username alitokmenIt always says that my password is incorrectMoreover if I click forgot my password on Bamboo and enter my username is says that alitokmen does not existCan someone please check +If you find issue with please report them +Deferred to +and ROOT regionsFor now we can keep in ipc module +I think the most intuitive method would be to remove globally from the list +Revision with fix is +Changes look good to me. +Removed affects version incubation as this is a redundant version +Opening a new defect with the proper version etc. +We should not be inferring types for constructs with an Native annotation and should instead be forcing the user to provide an explicit type declaration +has done +Can anyone verify this +This is a forward port which is very straight forwardIt was checked into on agree this needs some docs +And its IMHO a good idea to add a text saying that This exception is ignored +But Millind helped me test it and it works +This is problem of Lucence search engine it cant distinguish between underscore character and space +Its probably not a good solution though as only properties coming from ant are available as system propertiesSomething better would be at least to ask in order to retrieve the needed property its probably possible to query via cocoon in some way but I have no idea how to do it +This implementation also doesnt not support rehashing so I wonder why dont we use one of the templates +First integrationSending assemblySending assemblysrcmaindescriptorsSending assemblysrcmaindescriptorsSending assemblysrcmaindistributiontextetcSending assemblysrcmainetcSending clientSending clientsrcmainjavaorgapacheservicemixkernelclientSending gshellsrcmainorgapacheservicemixkernelgshelladminetcSending gshellsrcmainjavaorgapacheservicemixkernelgshelladmininternalSending gshellsrcmainjavaorgapacheservicemixkernelgshelladmininternalSending gshellsrcmainjavaorgapacheservicemixkernelgshelladmininternalcommandsSending gshellsrcmainresourcesorgapacheservicemixkernelgshelladminbinSending gshellsrcmainresourcesorgapacheservicemixkernelgshelladminetcSending gshellDeleting gshellsrcmainjavaorgapacheservicemixkernelgshellcoreremoteAdding gshellsrcmainjavaorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainjavaorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainjavaorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainjavaorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainjavaorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainjavaorgapacheservicemixkernelgshellcoresshdSending gshellsrcmainresourcesspringSending gshellsrcmainresourcesspringSending gshellsrcmainresourcesspringSending gshellsrcmainresourcesspringAdding gshellsrcmainresourcesorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainresourcesorgapacheservicemixkernelgshellcoresshdAdding gshellsrcmainresourcesorgapacheservicemixkernelgshellcoresshdSending gshellitestssrctestjavaorgapacheservicemixkernelgshellitestsSending Transmitting file dataCommitted revision +Presuming this relates to problem with static analysis tools is that they dont have appreciation of a dynamically built interceptor stack +It should be pretty easy to add tests just need to do a full load of a cachable object and then a range load or some such +Im really at a loss to determine what is the root cause of the issue +The helpdescribe mojo is a chunk of core logic in a plugin +Too much dead space on the margins of the screen +A test case always helps +This one is fully satisfied now that we have per workspace styles layer groups service config and admin all issues that have been in resolved state for more than one month without further comments to closed status +Yes that method should only take replicaId +Why dont you use slfjlogjMDC? I think its better than to add some code to print the user in logs +Many thanks Suresh for the reviewI have committed this. +I was able to Transform the XML by changingI used Instead of +Maybe we should deprecate this class +SebastianI changed class to encode blanks in the query part of request as +Last file for the patch +Pavol this should be fixed by which has not been released yet +to which is resolvedClosing issue. +Completed as revision . +bulk close of resolved issues. +can you verify that this works w jdk as well as +Ive done some more work on POI and Im happy with the output for all the test documents I had to handOnce weve had another POI release upgraded the Tika dependency Ill enable support for the older word format extractor as see POI with this change is released and the tika dependency updated +Reason for reopen +Could you please list the tests that fail on Windows after your latest patch is appliedCould you please attach log files for the failed tests +The tag name that the plugin suggest now contains SNAPSHOT +The cleanup thread will clean only the base directory which exceeds the limit +The test works once contention issues with plist creation that underpin the file pending message cursor were sorted in +It may also make sense to mention the SYSCSUPDATESTATISTICS procedure in case someone needs both compression and update of statistics. +I think all of the stuff I mentioned in my comment above would be set for a later release +this test is all resolved issues due to a successful release +The most likely cause is that the file does not exist for streaming +The only purpose of is to qualify EXTERNAL when declaring procedures and functions +We have seen a very similar issue in our environment +Please add this skeleton to package and try thank you +Previous patch was generated from a revision that did not include the most recent set of changes in +Strong yes this does affect EAP +Currently the merge policy is implemented in ways that assume certain things about the existing list of segments +Since the group failed to provide a list of committers I cloned the list from Jakarta Commons +Thanks for reporting +has issues with lookup of custom properties which causing Cannot find a definition for the property named This seems to be an escaping issue +the reason why fails is because you moved the num out of intoWhen we buffer del terms in TermsQueryQueries we grab the current num as the docID upto to record when it comes time to apply the delete which docID we must stop atBut with your change this value is now an undercount since num is now acting like numOne way to fix this would be to change the delete methods to use next instead of numBut I think Id prefer to put back num in get +please to Rich +Contains to show bug in unmodified codebaseContains patched classContains adjusted class +Filter only those properties matching the criteria +What do you expect to happen if filter will pass it through? I think it will fail because no proper CORS response will be returned to the browser or does it work if you modify the filter as you suggested +This patch should fix the errors in checkstyle for +I have reviewed the patch provided by Brian The changes made look good to me +My preference is to keep a bit of separation however +i have missed that +It works with list arguments +I guess it is strictly related to Oracle link and see you are the one who fixes bugs please look at this one I tried to fix it but no one responds +One little issue I worry about the providers documentation +If this patch fixes the problem for you please just add a comment saying that +Second patch that fix is not existing but instead it isCan be committed +The patched file takes care of this +The schemas have been moved from metadata to jbosssx +Bug the mail list too next time +Patch that handles this exception +mrepositoryhomesonar +I am attaching one of two patches that will make the role configurable via the file +From now on an info message will be logged and the problematic character will be ignoredThe problem with the thrown exception is that the application logs can be easily flooded by an attacker +to represent the current directory +In my opinion the node health should be reported as unhealthy which will attract the attention of the Script DeveloperUserThis isnt how the feature was originally implemented problems in the script dont imply a problem in the node +Also there is always the option of having multiple methods one delegating to another. +but this patch did not contain class +We would very much prefer to not have to specify that flag to our own code so we can run on stock JDK +Noted that injecting other beans can lead to behavior this effectively just clarifies the spec and highlights that an impl may choose to error at this point +If it produces a dita file that will actually build Ill attach both it and the resulting html to this issue +I will get to this tomorrow +Pull request merged +If that enhancement is taken onthis should be fixed thereThere is also a push to have people move onto the JSTL type tagsIll leave this sit for a small time to gather opinion before closing +If we can add anything to this great +Attach a patch to document message filter features +Yes for Archive files there is a JiraI am working on that +Is the right patch? It looks like eg containsPath is missing a curly brace +sql script as there can be no connection but if we catch the output in that mechanism we can check for the string and Driver not found in the output +Perhaps some of those treat the encoding differently +there are no leaked file descriptors however the XGC image session context does not provide mechanisms for programmatic control over retained image source memoryso this is effectively an enhancement request to XGC as such it is not an FOP bug per se but an XGC issue +address both +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The reason for this is because the number of error messages to display would be very largeregards Malcolm Edgar +went for the solution here for now +I think it is reasonable that is granted permission to create +As we were not able to reproduce the problem and we didnt heard from Thomas anymore Im closing the ticket. +It seems like the attribute is getting lost inside Dojos code +Improved the warnings logged when closing context in unstructured manner or from another that unstructured context closure succeeds despite the warning +This operation changes working directory so it is required to destroy ftp client after this operationIt is possible not to destroy ftpclient if one can find the way to restore working directory +Note that we are considering a new type binding library in a future version of Roo such as Eclipses JDT. +The Basic models referential integrity feature only works with content stored in the database federation of multiple sources requires performing referential integrity at a higher level than the connector +looks good All in are beign ignored but that is the same as current a new patch as there was a non determinism in an existing test +I wanted to double check it was ll check the other and reinstate where missing +The I suggested as a work around will only resolve the issue if there are no existing log files in the directory with the +Please review +Solution is to add in to dependency testing now on IBM then on other for compatibility +I just committed this +Looks good +Sorry again for coming in so late on +Integrated to as well. +The query above never puts the value in itself +I would imagine it would be better to have the new metrics take full advantage of all of the stuff Metrics provides right off the bat +I just committed this +Errors must be +Im wondering did you find this sample somewhere +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Fixed in nightly build +This is my sample demo which is not working +Thanks Suresh +Or wait for delay scheduling to kick in and get scheduled on node +To fix this issue you should be able to engage addressing for proxy services which is an improvement stated in +Net version +make +Ill start investigation on this and update the JIRA and mailing thread as I make progress +The default merge function would prefer summary tuples document bodies REVMISSING markers +Hey Max Im a little confused over what you what exactly specified in the docs +I cant reproduce either +Done in revision +But the type of destination will be checked afterward +Mikhail L most such reports are nearly useless without backtrace uninitialized values come as arguments and it is hard to find real causeIn particular those warnings in JET are caused by managed code and Im not sure how we could trace them without manual debugging. +defer all issues to +Version id date her eis the datasource for it +It is currently something that is hard to fix and will need some careful thought on how we manage the session in the sitemesh layer +running test +If everyone else thinks this merits a package I guess I am OK but I wonder if the package is not actually datamining or somesuch +I have autoCommit set to with openSearcherfalse +In slide titled Test Results of First Usage TPS is write request per second and QPS is read request per second +Patch applies cleanly but jars are getting published in hcatalog namespace instead of hive namespace in local maven repo +So if there is a list of hooks they will analyze the parse tree one after the another in the order they are defined +The source for that got added in +Perhaps he can shed some light as to the version. +Ill remove it +I have fixed the hornetq dependency you now have to either make it implicit or fix the other modulesI would suggest making it implicit as users may eventually add their JNI and bump into this issue and thats a hard issue to find +thats because has nothing to do with the issue at hand +The Ant script exists to do things that the Maven build cant by itself it is not duplicative +Its simple patch so Im giving my here +Patch a enables the client testThe poor performance has been resolved by other patches most probably andFor me takes a little less than minute with embedded only and around minutes with both embedded and client testsI plan to commit the patch tomorrow +Does it make sense to print some log every so often to let users now if the application is stuck in newsaving state for a long time? Say every times of theCLI is the most common user of +Committed to trunk and. +Thanks Jirka +Preliminary patch to bring in via hacks +but make sure to nuke your +let me read over the code +Im sorry but I did not see this when it was filed in June +classpath and +Heres a quick fix +New patch No additional changes in this patch it just freshens the previous patch to take account of the recent work on label obstaclesAndrea as per the previous comments this patch just differs from the original one in the get method which now queries a to get the number of feature types being used +The behavior is the same on Linux and Windows +Does not group provide a better connection and or CVS server system? I use and it seems to be a good app server so farAgain if you have what you are sure of as a working jar please let me knowRegardsTony +Ill stick the value in the endpoint for now +Looks great for me +This requires seperate ejb and profiles then +was about to comment this feature have existed for ages good thing you found it +Could you please attach the process once again? Does it work ok when using constants instead of variables +But if user try to run a project with errors then it seems to be resonable to warn him +Committed revision . +Committed to +youll have to do it as part of the platform integration code if you want it +nice to get this cleaned up +Also unassigned +Implementing this would be a good learning experience for me but I wont object if someone else wants to take overThis comment is part of an effort to close old issues that I have reported + +When I was fixing the I did notice that a surprising number of node types were stubbed out +Done. +Thats true and I agree but we should still try our hardest and give listenaddress a try +helei please enter a new JIRA with this issue reference this jira in your description +Last response from me expecting close to this issue soon +Linked which seems to be the same issue in iPOJO +This pull request appears to fix this one but not. +Can you help commit this patch to trunk please +Works fine in HelloHelloI dont know it is make any matter but In both Im using indy +Then the parameterized constructor can go away completely +Improved commonswebbase initialization in revision enhancer integration tests pass on my box now +Can you check the GC for this process + +It can also be overriden to cover the empty string of the not set of and the hypothetical scenario where it has to be some other valueThe use of special magic values is necessary because Mojo parameters unfortunately combine the empty null and default scenarios for strings +Aaron I just have created for compareTo +Ops it took too much to write the explanation you already got it +Committed revision on trunk +Ramya if you are still having this problem please feel free to reopen the issue and PLEASE give us more information so we can reproduce it if so +are you gonna have timeitch to finish this soonish +As part of this need to separate out the public API and the impl in the code base as well +Do you agree +Jon had to append Test to if we were to run it in the builds +To be clear the other node that wasnt rebooted was not active at that time +Feel free to provide feedback in case you have things that youd still like to see or have changed. +Which branches should I generate patches for +This is fixed by the patch in +Closing issue fixed in released version. +Added m not sure if will even continue to exists as the VDL deals with more variations between JSP and Facelets theres no harm adding it for now though +used to drop and the endpoint +Any idea about howto reproduce it easily +My patch tries to make the check as low impact as possible +Your patch is in CVS +I think the feature of running on an existing eventbase is an important one if there are any users that run this in a scenario that is not then this would be a for makes sense Thanks Pavlin! committed +jar no +Removed from fix versions as does not have the jmx option +I suspect this is due to a race in +Users will have to install JBT with indigo me +I just committed this +Ive found that when you enter a broad type of issue like this it tends to be pushed off version after version +ManoCan you please post details of the problem at users forum +for branch and trunk. +Thank you +Brian when you said bump +Done +From my own tests it seems like the indexer was leaking connections because no matter how large a connection pool I started with eventually it would hit the limit +sorry for the noise +Just for the halibut? +Thats fineIn first version we can disable region splitting +This issue is fixed by +Thanks for patches +I havent been able to reproduce the exception since I rebuild created a couple of Wars and Ears and all seemed to be created as fixed +Erik is there a review for this? If not can you create one? +Thanks Nathan. +There is an issue w this patch related to unordered overlapping spans that still needs to be fixed +Thanks Alex! +Thanks Devaraj for the patch +In order to be applied the failures you mention need to be fixed +Thanks Namit +Yup should get all jira traffic and should just get createresolve +Posted from Bugbox for wasnt taking a picture rather I was starting an app and I got the exact same stack trace +This is another argument to keep velocity for jnlp generation in favor of using something like DOM +Infostream logging should not be enabled by default in example thoughtsGreat idea lets put this in for sureI think the config actually does belong in though +Also does not support multiple resource bundles +I will have to the changes and will update the ticket when I have done so +This way it will not go through the process of trying to resolve the url within the current portlet which will not allow redirection to a different portlet +Attaching API logs and management server logs and DB the issue as it is not seen anymore +New classes as well as changed classes will be formatted using Alex this issue is about source code of the plugin itself +nevermind aux is a reserved word in windows and not in git +We shouldnt create a blog for each project space until we have agreement from the respective project leads that they will use it +The semantics of implicit join changed slightly in HB +zip file +We want to come up with a fix for this problem +They have a nicely isolated dependency grab but the compile depends on it +Wondering if this goes far enoughLooking at the +At this point it feels like were better off unifying our scripts instead of introducing yet another copy of one +This is a feature that is currently working with my OCJ application server +Consolidation of the plugins guestfs sh calls means this only needs to be done in and only affects +Also were not fixing Android issues in this release were building against Android and turning on Android features +for the patch +To what extent should the configuration file alter the behaviour of the query server? Shouldnt this state be contained within the design document as much as is possibleIf significant configuration is required for a specific application to what end is this going to be beneficialproblematic? Ideally youd be able to deploy a design document to any server and have it work correctly without needing to restart or otherwise tweak the configuration of the server right +tried with with windows chrome and windows safari works fine. +sorry +The following operations will close the file +Eek +Suresh Srinivas Sanjay Radia and Arpit Agarwal +Patch created against devtoolstrunk +You are welcome to vote +If we wish to add such checking we should address it in a separate issue where we can benchmark it etc +changing that behaviour will break existing change how you build stuff for example a client im working atm has to change the build process otherwise the complete application will break they do not set this flag explictily but suspect that this setting is set to can to this for next release but not for current release rc release cannot change behaviour +We dont have to get it perfect right now but some separation would be good +Thinking about this more this will of course only happen if the value is being pulled from the body content multiple times in a row +Itll be included as part of CR so marking as Resolved +For the patch so far per my comments on dev I think this could be pushed into the interface instead of here +OK +I thought that the output from the Octth triage meeting was that we should enable the compatability testsuite for CP branches so we could reliably test whether a CP fix broke backward compatibility +The UGI must simultaneously support bothMy general thinking from before the summit has been a client UGI should do JAAS login for a given +set has to throw ISE because we actually perform the connection during the constructor +Spoke with Ray who said the current patch is not threadsafe +is for illustration only will not be committed +For a first sight looks to be working nowThanks a lot +that one is correct isnt it? +Im tempted to close this issue as Wont Fix because I have no way to test and verify this nobody seems to be using this this issue has only votes and only watchers the original reporter mentioned he noticed only marginal speedups +changes you need me to put it back in +Ah +missing assignee +JarcecVenkat +This is the stack trace of the error it throws while packaging the project +This email fails to open +Committed to trunk +Can you check with the latest trunk version if the problem is solved for youI will then start a release tomorrowThanksdue to my vacational out of office i unfortunately can not test your commit +What do you mean with breakdown +Ok I think we are talking about different things +arun I assume youve meant Youd have to contribute to Hadoop whatever you want to use from Oozie right? If so Im good with that I dont care where the code lives if I can use it +For the life of me I do not remember if theres something that makes sense by all means change it up +java version whose issue is it? Is it the apt issue? There are two ways of installing maven in Ubuntu one is downloading the maven archive and point to it and the other is the standard of doing it via one of the package managers +Implemented for an internal specification but submitted for consideration as a starting point for something for +Trivial patch attached +Should we perform rebalance if the cluster is not finalizeUpgrade dIf yes maybe one test case for this +See also Head console thingy or Mesos UI +Hopefully that will be seen in the results of your serialization benchmark +Andrzej Are you going to add a way to automatically add an index to a Solr core? Are you planning on adding test cases for this patch? How does one set the maximum size of a generated shard +htaccess file +A lot depends on the actual structure of the subtree +Thanks for the quick review Dag +Correct it is a synchronous construct so could should be allowed to breakcontinuereturn from the at +Otherwise buddy node may start accepting replications from data owner before state is received and the state may overwrite replications +Committed revision +The problem does not seem to be related to the use of the rtrim predicate because I can remove that such that the ON criteria is and and the and portion still gets dropped from the query plan +Patch committed revision +Can you please provide the log file of both server +I verified it with a fresh install of GanymedeTed might it be that the update site was remembered somewhere else? I had to enter it and I would doubt that a fresh Eclipse installation would know about the Geronimo site where should it know that from? +Can you show us the code? How are you specifying the character? What font in specified? Are you sure that that font provides an image for that character +Patch applied in trunk r +So it seems logical to me to put them in a separate module +Thanks for patch mighty Andrew. +shouldnt the stub generator use fully qualified names there too then +I will commit the patch without going through Hudson +Update the title and some one check this mapreduce part? Am not very sure as what is happening here +You mean you sent a request to but didnt get a response? Make sure youre subscribed to that email list first otherwise your email gets silently dropped +we might have found a resource that is going to work on the webapp +Solr does have timezone support it supports UTC +Now to use session scoped temp table we are saying we need the same connection back +Will commit the change this afternoon once full testing is completed +I think there is some call back that is used to check duplicates +Claus if you agree I would close this issue +Which version of eXo JCR do you use +I checked this and now it works fineRecently we did a change to to support character encodings correctly +Added +Well spotted Robert +Fixed in branch add TestAlso I add this test to branch to prevent regression +Moving to M while the debate continues +Thanks Tom for filing the Jira on this issueDoes anyone know a possible alternate gzip codec that could just be plugged into Hadoop? I suddenly have terabytes of gzipped data that I dont know if I can trust hadoop with and it would be a big pain to everything +Thus a new fix was required changing the fix version for this issue +Patch committed thanks Gunther +Several patches are up for review that make and as well as variable YARN versions YARN Giraph profile work +Added class that implements for the same fixes the does for pessimistic is now agnostic as to whether optimistic or pessimistic locking is used +processDay function should delete the previous day directory if the previous day directory is empty +Otherwise well follow the normal parent order +I also see the POI jars mentioned in targetclassesDEPENDENCIES but those are also just version numbersI see the and ooxml packages in orgapachetikaparsermicrosoft but those are just the testsWhere does the POI jar go and what needs to be configured before running mvn to build TikaMy guess is that my ignorance on how maven works is why Im not sure what to do but based on using ant Im used to putting something in the right spot and possibly changing a property to make sure it is looking at the correct spot +I switched up the wording because it created a long noun phrase so Derby was moved into its own clause at the end of the sentence +Removed usage from and related classes in package util in r +Drag a simple Dynamic Web Project on the Deploy only server in the Server View +Remove author tag from +It seemed safer to detect this condition and fail fastIt is unlikely that this change will make it into as that is a managed branch +Andrey thank you for the contribution. +If you are changing things that dont overlap than no problem +Implemented a pure test for the and fix the bug that was causing it to fail +wont have time for this until the next release +Will remove from hdfs after this change is committed +The primary hooks we have today come in the form of the Source Channel and Sink interfaces +Ive a patch for HEAD +Im running on DebienLenny with a custom kernelThe problem appears to be reproducible on my specific environment +To help clear up the design further the answer to another of the questions I asked would help how will the data be used? In other words what will be evaluated or when doing evaluations what data and data relationships need to be considered +If it doesnt work then please reopen this issue and if possible provide a test case +By ownership here Olga didnt mean taking the code out of Pig + were right Dinesh this rule should log an issue only when a method explicitly throw a checked exception but not when this method only propagate a checked exception +Do we want to have those changes too? There is no test but it sounds like a good idea to have one no +Patch that converts hfilefileinfo and removes +Was looking at the one on that hadnt updated with your commit yet +The easiest way here would be to switch the execution order to first vertices then master this would make all of it consistent and it makes just as much sense as the other order does to me +Thanks a lot! +using and everything works as expected. +FYI Ive attached a patch to that allows updating to which does not generate code with variables named enum +The a patch looks like a good precaution to prevent hangs in case there are other similar problems + +Updated patch that uses now that cookie auth has landed +Time to close this one + +please set this to resolved if this is in the last build of Ref Guide is in zip and GUI installer dists. +How about dealing with the Caveat Emptor model? +Pnative vs +Fixed all file based configurations can now be loaded from the classpath tellus if it works for you Joe. +If that is the case we should revertHow come this was not detected in Jenkins build +Thanks Joerg. +Applied. +Of course optimize would get a different message thats this is deprecated and the user is not want to call this +You can use multiple on the same orderinvoice +The reason is that the client driver has a finalizer for the Class which when run also closes the Reader passed in +Reopening for setting proper resolution status. +Only files in accounting application are changed +Oliver I think Ive come across a problem with the Querydsl support +Ok now I understand +Compare betwenen and if you wish to confirm yourself +I dont think this should be a showstopper for the release since its been in there for a while +Relevant stuff ported to in rev I cant find out what if any openejb version is supposed to go with g so if the correct version is not openejb may not build. +Please review the rest +Thanks Dmytro. +What OS are you running this on? This test is a bit finicky it actually is suppose to test that the back pressure feature is working correctly by nearly running the JVM out of memory and letting Blur react by killing currently running queries +For the life of me I cant get review board to accept my diff +See for details about the release available on the public sonatype repository and hopefully on central soon +Ok the failing test is because of +This is not Junit Test Yasin +Handling of JAXB conversions is tricky business +Oh sorry +If you move the handle it will jump back as soon as you release the mouse button +A working test case for watch transfer with multiple hosts +My class hierarchy has an unindexed subclass since this uses Hibernate for the query naturally it gets all subclasses and attempts to index them +Hi BorisA few little niggles +forget +I am able to reproduce on my Mac but not on so it looks like still have some +Moreover these times are also recorded in the job history +Seen this happen occasionally in Jira though only previously when the workflow was being mucked around with +The more complete option would be just to make all of that more pluggable so things like the classnames on the Interceptors and Features annotations can be used +And a mvn U does not help ether +The tests were marked broken for a moment while some further work is done to correctly enforce the assertions being tested +For the AS yes to avoid the WARN in the doesnt exist in so this jira issue is stll relevant albeit less important +There should be a top level HDFS with the servers in it +In the meanwhile can you not set the TCCL to null in your Activator +If the patch follows the same coding conventions as Geotools does not unecessarily contains reformats and has a test yes its likely to be accepted +A simple test case on the SPI module should be able to assert if this case is still present + again +Fixed thatBy the way none of the minicluster configs are in didnt include the new one created here either. +It does minimize synchronization over MR +Fair enough +bq +Making it critical this is about making it so can we can do schema edits online +Could somebody explain to me why we create the node first and then take it from SPLITTING to SPLITTING? Just to trigger the a node changed event in the master +henry are you going to add tests to this patch +My issue is that if the tests have nothing to do with jBPM why are they called? Do they test something related to jBPM and thus require a jBPM runtime +I tested the patch by running distcp to write to swebhdfs +However this bug doesnt really cause a problem since it only occurs when doing unaligned reads +While I dont see a very common usecase the entire codec stuff is more or less for uncommon usecases +So Im voting that location +Attached patch is ready for review +A lack of response by February will be taken as consent that we can close this resolved issueRegardsBrian certain that I verified this a year ago before the from Xalan went out in Xerces +r on trunk +This should be fixed now. +Thanks I can confirm that now builds cleanly on our buildbot as well +The is recognized correctly but the submitted value is lost so the user doesnt see his mistake +I dont really care where it lives and it sounds like tools makes sense +Marking this as submitpatch so we can review it +Done in trunk at rThanks Stephane +BlobClob will not work correctly in soft upgrade mode if this is not corrected +Patch applied +Thanks for the reviews +Hopefully we have that worked out with the fix for I believe the upgrade tests test test metadata make sure that everything behaves properly +Thanks it now works for both Vivek and myself +Can you give an example for illustration +It also seems to me to be generally good programming practice for a library or framework irrespective of my particular there any other areas you can think of where closing the application context will still leave things around +Hi HowardIm not familiar with implementation so well but I think this approach could work +Too many tabs open in firefox +Hence it wont be using the local filesystem +Fixed took it for a spin with wfs datastores which are slow to load and yes it helps +I am strongly against breaking tons of languages for some sketchy optimization of englishIf people want to turn on this optimization they can turn it on for their field just fine +Verified on the docs stage. +Looks like it got even worse recently we got a situation where the was treated like if it was the RSs own and it FATALed +Hi Andrew this is kind of urgent now when it can be fixed +Clicking the grid always returns the incorrect selected items +Dhruba Thats an excellent question +The fact that one attempt to open the ledger and recover fails indicates that there has a been a concurrent attempt and in this case I would say that the client that gets a failed request should actually back off +Ack are ready to take the staging site live +Discussed with Owen on the issues +Thats what Im about schema and state changes that are +Refactored the API to be independent from JSP and from other technologies by using varargsThere is still a JSP dependency in but this will be addressed by. +Thanks for the reminder +So for these platforms this exception could be still relevant +In I believe that we should omit the restriction that there can be onlyone Struts front controller +Im going to close this as Wont Fix on the assumption that well make use of the script when we finally get around to implementing. +Please attempt an import with this new file +Committed a minute ago. +Is it possible to change linebreaking behavior from Knuth model to simple in my +Gary yes I know its new my concern is about the next release after adding bmpmI am concerned that the bmpm API is not stable and that an early release might entail an incompatible change laterHence the suggestion to discuss if that would require a packagemaven name change given that few external classes would be using it +AdrianYou obviously havent looked at the patch so I guess were talking about different things +All the tests ran cleanly with so I committed revision . +Leo Ive debugged through and it seems that apply is not even called for my fragmentPlease note that my xhtml consists of uicomposition and the action is in a uifragment +Test results T T root S S root P P root not conclusive at all +Any properties that are defined in your bean definition are ignored +Im not bothered about needing to have the logging subsystem installed in order to get this functionality +commit the fix At revision on buildThe problem still exists on the two conditions belowInstall web application via cmd and uninstall it still via cmd then hot deploy itInstall web application via admin console and uninstall it still via admin consolethen hot deploy it +Joe installed and that resolved a missing dependency +Alexadnreany progress on it +testngspringide possibly even our eclipse dependenciesHow do we handle Ps total lack of optionallityOnly have required stuff in the product and all optional plugins on the updatesite +Also decided to modify the standard project template the same way +Ill run the relevant tests and check in if all is good +The code to registers the providers was missing from the bundle +When the is updated to send the message header back in a response header namevalue pair I assume the name will be +Are you also using? In case its a more recent one could you attach the stacktrace +Thanks Marietta +Maybe moving stuff like version information to the parent pom? I feel like using a dependency management section in the parent pom will allow you to simplify the lower level dependency listings +This time it is REALLY rebased for trunk +It is a bug report because an advertised feature is incomplete as currently we require the use of a special message format for tracing and do not provide a way to create a message in that format not providing a way to create the special message format is a bug +Will be release in in released. +Can you configure please +OTOH Id want the namenode to be able to move out of safe mode with minimal restrictions to avoid unnecessary manual interventionOTOH I want to avoid unnecessary thrash on startup while data nodes are connectingSo any solution would require both a threshold and some timer once that threshold is reached to avoid unnecessary replications the second the threshold is reached before the last datanodes connectId go with a configuration requiring blocks are available at least one replica allow one more minute for stragglers to connect and report their blockstypically this would allow up to two nodes to be missing but if by some magic more nodes are missing and I still have all my data intact such as when we implement rack aware data placement and a rack goes down hallelujah I wait one minute and start +I dont have data in your projection but only some sample data in NAD which consistently fails to generate a +bq +All committed now from r to r. +The only problem with this was that it requires us to modify all our deployments which would also be required if we went with the solution +So I think the issue has been resolved about one month agoDo everyone agree +Doing this would be trivial using or simply by reading the property files into a java program and writing it to single file +Still the artifact is missing +they could also switch to Core should we generate files tooIm on adding files that are not needed in default behavior +moved the appserver plugin to the phase to guarantee ordering +So personally I would not consider addressing the copyright notices in existing source code files a blocker +that would be fine with the default of a think that backward compatibility is important some sacrifices need to be done too +To me all the other capabilities other than filtering by a Spatial Tile Field are already covered through normal Solr capabilities so I dont see much benefit to inventing new syntax and more code to maintain even if it is slightly easier for the app designer to consume +Definitely Equinox does this differently than Felix +Ill commit this to trunk and branchThanksMichi +The first version is correct and works for me with FOP and even FOP +Its a pity this page wasnt added under but maybe we can get a wiki administrator to move itIt should be okay to have it moved under by then +Finished! Code coverage is almost at its maximum +I run and derbyall is running now +Thanks for reporting this issue +bq +It looks like it resolves many of the issues that were happening with though +Here is a patch to fix this problem Add a JAR containing all the classes to perform options and arguments parsing +I removed a println statement fromThanks for the contribution Patrick! +For summation of float double should use to do +Fixed in remote resources plugin and in the bundle +I guess this is trivial enough to be added to a testcase +We can mock the datagramSocket to capture the packet with mockito +updatting abstract +Or perhaps is should be a blacklist that could do the match +is really well thought out and put together a pleasure to use +patch looks committed this to the trunk but on the branch I was not sure whether this patch was really required +You can qa and ask them to kick off run for the CP branch +Hi Joe Thanks for following up on this +Applied at r thanks! +Committed the change to the branch as revision Need to update the fix versions as JIRA isnt showing me an appropriate fix version right now +The generated +Thanks +The classes in the stack trace are part of Xalan not +The definition I normally work against is how much work a team can get done in an iteration your definition appears to be closer to how much work a team can get done in a real day +With regard to the default encoding I must tell you that on my system is Cp which you can see on the print of the sysprop attachedI still want to point out that the extraction problem does not occur on all PDF documents but only with some +These classes carry all the information to map between XML artifacts and Java classes incl +Please use both these patches +Is the user that you are using to run the script in etcpasswd? I assume that it is but if so Im not sure why the is thrown +Hopefully fixed on both trunk and issue has been resolved for one month or more without complaints +Note that just as happened for Jason in the clientservercase I get the confusing unarchitected and conditionexception whereas in the embedded case the XSDG exception is deliveredstraight to the program so just as Knut suspected this appears to be a weaknessin the client exception handling +looking at the code it seems like it should happen everytime the sync request is sent to a follower but that doesnt seem to be the case +Reviewed by Alan and Gordon +This bug appears in as well +had failure which was not reproducible when it was run on a standalone basis +I am resolving this as Wont Fix +See the for usage +I think I do like the idea of limiting by bytes instead of count as does +No intermediary to be a potential bottleneckThe notion has a reasonable argument but its a lot of work +Its improved but I agree that it will need some additional documentation later +Fixed on did not need any fix on +Thankskarl +Similarly use instead of + +Fixed once these were changed to +These schema primitives are all things the schema files do not contain unfortunately +Ive just committed this to trunk andThanks again for the review Eli. +The patch make use of the Hth that would make the test pass however the test is about ensuring that the binding occurs on private variables +The reason for the MB patch is that it includes a sample text file for the tests to useI have done applied the patch myself and it appears to be working correctlyI dont know why the core tests are failing or the contrib tests but after looking them over twice now I am pretty sure I can conclude that they were present prior to my patchIn any case here is the latest patch +Reassign to a release that you feel comfortable resolving this issue in or leave as is +There are occurrences of use of raw types in Java files +The determinism does not imply a vs approach right? Only that already validated objects dont get validated again +But then it will require maven or greater to build +As the CI servers cannot download the SNAPSHOT bundles +I am seeing the curly quotes. +Upgrading the ehcache version to get the benefits of the reworked shutdown hook has has been commitd to revision in archivatrunk +Merged to. +With the above scenario where wehave extending is the following statement correctIf we delete an instanec of we have to delete all instances of share the same identity +If thats really what you want you can still do it by copying another nodes Schema cf files but its basically never the right thing to dobq +Please find the mail with attached details +I will take a look +There is no problem I understand what you say Roy +i think we can +It has maven deps on jug and xstream +Fixes +So more than likely the underlying problem is a userpassword issue +I must be missing something here +sample file see the method create or splitContainer +bq is parsed with the defType specified parser +Our one is based where as the JDK one is based +Thanks +Patch for the would be handy if it output a list of the segments it generated either one at a time or a list at the end of generating all of them +Its true for other plugins as well +Thanks Paul! +Sampling rate is currently left explicitly to the user via correct +or something like thatDunno whats the purpose of the +So I skipped +I dont think this has been applied yet correct? Im reopening +A quick glance at the code indicated that the JDBM commit was only called during clean shutdown so this issue was already on my list +should fix this +Unless anyone objects Ill close this as WONT FIX +And hide all constructors by marking them to private +Does anyone have a workaround +Changed the names in this task title to match reality. +These are really the same +I was able to retrieve the index and browse its contents from meclipse +Per Bryans request I will be reviewing this patch also +Fair enough +Let me do this +Therefor it might be useful for people who want to fix this issue cleanly +Use backticks They will be replaced by the appropriate quote character depending on your DB +Right +assign to Daisy review +Extend example code to include generic class implementing generic branch merged into trunk in r. +Sometimes mail client are smart enough to send my messages to trash so Ill attach it here too if anybody wants to improve on it +part of release. +defer all X issues to +You are changing it previously the message was easy to grep v now its very short and maybe less soAs for k +This patch addresses the issue with test duplication entries +However we do want to upgrade occasionally + +Verified +The for this is when an MDB is connected to a remote cluster +Ill try to update the the Forum Tim Fox wroteIf clustered is set to false then none of the cluster related attributes should be parsed that is the way it used to workI believe Clebert added at a later date +It looks like and are the jaxb jars on the classpath +Gave separate error messages to each of the possible failures for UDF method loads. +There are strange errors when downloading from to delete it from local repository +This issue is also occurring on containers but only when running automated tests +Dont use this one +Please if this is still an issue +I created a clone JIRA so that we can fix and verify separately +This looks to me like a GF issue as it should not serialize transient fields which appears to be doing +I think there is some other problem so this simple filtering doesnt help +I believe the current mechanism is already subject to that problem theres already a genserver there with a handlecall implementation that could become overloaded in the way you describe +Hum first time I see this issue +I have Category no matter what perspective is open +Applied patch into trunk +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Let me know if this works for you. +This issue is related to can test it by removing the clean option from +sure! patches are welcome if you could provide a patch and attach to that issue it would be much easier to integrate that functionalityTIA +Can you clarify what direction Roo is going to take +was for the AS plugin and this issue is for the AS plugin +Turns out that a simple binary compatible fix was suggested by Dennis Handly in +happens with +This feature was implemented by VPE team recently +Browse to any page node in portal +An updated patch +Both the Job Tracker Instrumentation API didnt have a method for finalizeJob +Good +The blue graph displays the old content model where group membership is recorded in a multi valued property of weak references +Consider an additional patch that renames dthreads to alldthreads so the naming is consistent with allethreads +Yeah this is definitely a real problem that is leading to not all messages being replicated +they dont get deployed +Hi Aslak Im assigning this to you or am I commiterof xdoclet ? Did I missed something +That will not work +Sorry attached a wrong version the first time +Assuming the new datetime library comes out with Java and that Lang is dependent well revisit all JIRA issues then. +That would make development much easier for me and there will be less patch mistakesBtw +Please feel free to close the bug if the problem is already fixed Daisy! meant fixed on my way home in the train where it was quite annoying to stop repeated testing through this issueI will break the rule to create a patch for commenting the two lines as the patch is easy and more complex task are waiting +Using reflection Configuration validator finds out methods it requires to validate +I have checked some JBT Hudson jobs and they are blue and using Tycho so Im closing this Jira +You get back if you dont set any incoming charset and you get if you specify in the incoming metadata +I also traced down from where the synchronized call is happening looking for a potential race condition between reads and writes +No further feedback +I dropped the diff since it looked leftover from the previous patch and no longer needed +Verified by Mikhail. +Wish me luck +Its not a bug in +solution remove clib native libs from the PLDLIBRARYPATH to force usage of JDK encoder +Vlado Facelet HTML Validator is a part of Eclipse WTP +This is actually quite an improvement + includes back all tests for all platforms for DRLVM only +BTW do you mean that it is a bug in my code? If so how do you propose I fix it? Also could you reproduce it? +Any suggestion on what the separator should be if not at this point is just matter of picking one symbol that works and replace it in the regex +So an assembled product might have component products but each of those components will also have a single component that represents the raw material that the component derived from after manufacturing +This would also make merging the patch set with append substantially easier +Doesnt work +The issue is not reproducable on the sandbox jspwiki +can override conf directorieswe are talking about very special kind of file actually +Hi Jimmy +It does not require changes to the code and very few changes to the +Dropped an email to we cant do this for unless someone steps up +derbyall and junit runs on the branch with the patch were clean +This can be serious and hence we need to take care of this +How about as the package for +Mass transition all resolved issue that did not see any further comment in the last month to closed status + +Btw Im runningAmazingWere you able to measure garbage production +Transitioning to closed. +Maybe retry in reset is better +verified in JBT +When user types value into input field it is immediately committed to model and validated +Committed to trunk. +I looked closer at the test problem and it turned out to be the problem of shutdown patch not TM problem +Aslak Exactly. +As I think it is good to write a sample for demonstrate this XML Stream API base approach +Added doc issue this fix as I realized Android does not do this +Done +The sitch is and is by default true +On reflection it reduces the usability so Ive put that back in +Thanks Tim +Thanks for the quick fix Andrew! +attach a patch having request timeout rather than channel timeout +thx a lot +If the security is not an issue for the Server then it does not make sense to keep the local host by default a lot but if it is then opening it up completely by default does not seem right it would seem reasonable to me for users actually having to do more work in such cases with the host calculation requiring the least of effort and with setting up the server certificates taking the most of are more or less my thoughts +I am really sorry for bothering you. +Bulk close after release of +Committed revision . +Committed as r +patch looks unit tests locally +create a new patch with subversion +Thanks +All Done +Also my earlier objection to Curator is now removed as noted on the other JIRA because Curator has since entered and graduated ASF incubation to become Apache Curator. + +to an else +available on as r . +Attached patch for the issue +My bad +Let me know if you need anything else. +Notice the invocations from handle +With the the authentication is not required locally and I dont know how to force AS to request it. +For this particular test case what is getting us into trouble is the Unroll annotation which is really only useful when we are compiling to CUDA +Ive spent a couple of hours reviewing the patch and am not going to be able to commit it as is for the following reasons +This issue has been identified and resolved in the latest snapshot +Fixed in trunk at two files attached to this JIRA are the diff to add JMS callback support to the release and the prebuilt Tuscany jar for with the patch or not a problem or record this work item somewhere else +Plus added missing dependencies reported by dependencyanalyzeModified mavenpluginstrunk +Patch looks good to me Ill try to get it into Seam branch this week +We store in the details part of the authentication object data which can be changed. +This is trivial and should be committing +Attaching a new patch that includes the suggested ctor changes +will commit it later +The problem with the installs on the dev machines are likely a combination of the filesystem being NFS and the way that the fixcrlf ant action works +Toby can you look at the attached patch re testing this feature? Im getting a test failure on line and Im not expecting to +New patch somewhat works around the issue by reopening a new searcher with the new writer right away + +Thanks everyone for your feedback committed in rev +Maybe Luke will see this and know why. +Owen on introducing in +Issue error +See the CREATE statement for the database object that you want to grant privileges on for more informationFor CREATE SCHEMA I changed the syntax per your recommendation and reordered the examples so that they would match the order in the syntaxI need to look up the info for SYS and will get back to you +FWIW Ive found it most reliable to convert a date with no time component to noon UTC of that day when a Java date is needed +I think being a bit lossy here is fine and not worth the overhead of synchronization +I was going to go through a conversion of Castors Maven build from Maven to Maven before pursuing this effort +SU has a solution already +verified on +Not a bad solution and will likely work for usI was hoping there could be a more dynamic solution since it seems like temporary tables have some dynamic metadata +Fixed in will check the arguments shouldnt be quoted +I am thinking for some time now for the best strategy to cache lookups for Constructor and Method instances that do not create a memory leak when their original classloader is dropped +Oh forgot to say that I stuck with because the master is still called +Patch for +The last stack trace I got was based off the HEAD from CVS not final +Adds missing +Removing the sys out which was accidentally added +I seem to recall some discussion on the list about this issue +Even though you will not get a compilation error it breaks methods that worked fine before now throws an for our custom where it did not do so before this change +I have attached patch for this issue please collect this if there is some improvements then let me knowBrajesh BrajeshYours patch is in trunk rev . +Well peer review +Is it required to use rollinterval when using data bucketing? In real world you dont want to use rollinterval to generate many small files +if anyone is working on this speak up now or I will +I also inserted a missing article I changed with request to with a requestIm attaching and and will go ahead and commit this final patch +is not finding the generated bean class the first time the quickstart is run but any subsequent time this works fine +Considering that thread safety is not a goal I think I still prefer the current implementation +Heres a new patch with more consistent whitespace as well +Changing the Inject to a Parameter with a default will solve this problemIm nervous about making the framework any more baroque than it has to be +The project has moved to using subversion +Attaching release note +Since permission related Jiras depend on this can we keep the temporary hack until is committed? I have an explicit comment about itupdated patch attached +Daryn would be possible to prepare a patch for trunk? Thanks +If thats not the case please +I am implementing Sameers solution +Ok I have locally tested the v patch now with the existing cores tests and g codebase so I now intend to the fix into branch after doing last bits add a test case for this new multiple extension thing set the multiple extensions on the from groovyc just for information sake for nowThis should give others some time to play with the fix with other extensions +I think we can aim for. +How about consistency? This would allow maximum availability +Against svn +Thanks +This is quite often happening because its advised on forums to include HIVEHOME and HADOOPHOME in the PIGCLASSPATH which is then passed to +Specific problems will be fixed as they arise +Heres a patch with a working test +But for now it helps me make my error messages more consistent +Verified in Examples work fine. + quickstart demoing the problem +Ok Ill make the necessary changes +This way when the cglib enhancer selectes the classloader it chooses the POJOs classloader instead of s classloaderThis fixed my error where the hibenate classes were located at web server common libraries and in the web application itselfThis problem seems to be present in hibernate as well I guess that this fix would also apply there +Uploaded and for this +Alright so its not since this goes through the +Patches supplied to to store previous requirementcapabilities and resolveupdate classpath only when all issues raised by me +This suggests that the problem arose in one of and +Reassigning as I dont have access to perform this +Being able to pass through a schema from another system is a powerful use caseLater we can add to the specification how to encode +Seriously we just need to pick some terms define them and use them consistently +Need to rebase it post +Attached is the patch containing the initial crack at handling functionality on bean created using the and implementation attaches a annotation to the and relies the to add the to the +Verification fails because theres a newline on decryption +That seems like a reasonable approach +I guess it is a duplicate but even when I used the master version of the plugin I still had the same results +trunk updated in revision +Did you check with Opson that +Added NPE check +Its also done now. +Patch committed thanks Erik! +NET SPNonetheless +It is ok for VM if further inlined methods code is also inside this regionIs it possible for JIT to report codeSize in this sense +I looked at your activemq config in your test case +change in enhancements release +So I guess this does actually work +Im afraid this one was closed too early +Which tests are failing for you +I suggest you go with the option in the options you have listed +I dont know what other settings fail to be read correctly and adding a newline at the top of the file definitely resolves the issueDefinitely could be a windows line endings issueThanksWill +At Kevins suggestion Im going to change the name to Cached +Looks good +Added new patch to address the missing scripts issueThanks Ted for pointing it out +The Editor Id issue is fixed in master +I did not know one already was thereJonI am working on writing test cases Will submit it once doneThanks guys for your reviewsTedwill address your comment in the updated patch +While I still need to review it and push it upstream I did remove the console redirect on a PR I have open +Works as designed +Fixed by updating a single boolean flag +This is on the current stable release per the URL in image provided +for the patch +Thanks for suggested fix +Fibers are fixed in +Ok I can verify this +True the blockFor semantics so far has simply been a guaranteeIf we want to guarantee W R N consistent HHO nodes cant count towards W +The SFSB service name is serialised which is used to get the which has enough info to the interceptors +I committed the patch to trunk and +Agreed remove unused but why is the classes fully qualified +Attached a patch that makes the namespace registration in as as possible +Even with the favoured behavior wont workFirst request to Home will be OK but then will create encrypted urls for the in the page and from there on the url in the address bar wont be bookmarkable anymore +is the only one I know about +Thank you for reviewing the patch Kristian +It does not make sense to have the number of frozen columns the total number of could resolve this issue by either issuing a warning when frozen columns total columns ignoring the frozenColumns when the frozen columns total you investigate the behaviour when a similar configuration is used with a single EDT +I have no clue why this only happens on but something is definitely wrong in your applications request handling +Want me to open a new issue or can this one just be modified? +Fixed with the issue +Attached test to demonstrate the bug +Fixed wrong resolved issues for already released versions +Hi MamtaYour plan sounds good to me +Hi Kevin KevanThe current implementation in handles foreign key constraints correctly and resolves circular dependencies +Once we get to the stage of deploying the app however we need to know which component is gonna step up and provide the serviceYou mentioned that your deployment had an implementation element under service but not in component +Im planning to merge the security related changes to the branch so that is not released with security issues related to JMX and to avoids future backwards compatibility issues +It appears to be partially related to this fix as Nahi found s imap is not synccloseing the socket so they are left for GC to take care ofIn s case leaving it to GC seems to leave dangling sockets around anyway so where MRI eventually does GC and clean up the sockets leaks foreverThe imap patch fixes it but Ill file a separate bug for the dangling references +Ive toyed with doing this for a while but it almost seems more confusing and error prone. +So which information do you require exactly +Within and keys are created on demand +I didnt use a but just a to avoid a dependency on JDK and the error message is ok without the complete list of visited properties +As far as is concerned both answers to the question are right +Can you guys point me to a link with documentation related to hive indexing? I want to understand on how the index works as now the query without index takes around seconds and with index takes seconds +Lets get this in and close the ticket +We do write a log file for this if we want to scrape +Jobs itself are created and started via a webgui +You going to commit then Jesse? Mr Committer +We could certainly wait till +Eugene I closed this issue because the issue seems to be fixed in the latest +This issue is similar in visible effect to +The response body contains only one character +I have based my reasoning on the bug description and comments +I think I will commit the patch soon except if someone see a problem with that of course +Fixed in trunk +Closing this one as a duplicatebq +Is it truthThank you this should be done +Doh this is an interesting one + +Can you also add a testcase which I have above +No thats not normalI just took a look we had builders called maven one was maven running maven and one was maven running maven Confused yetAnyway I renamed the maven running maven builder and reconfigured one of your builds the build was successful. +This will read and write frequency list to Distributed Cache +Im pretty sure if false and onloadtrue will repro the crash +Should we call these custom env +Measure retrieval will be removed in Sonar Eclipse so this issue can be closed +update with suggestions from stack and to address nits from review +Closing as committer. +Thanks for the links +htaccessto set up moredifferent redirects. +This is no longer an issue +Ryadh +Llewellyn said on the dev listvery true the initalization line should read +Committed to trunk. +The fix would be trivial of course but I dont know if you had a reason behind the change so I will leave it with you +The problem was that the file was but it was not valid +Comments LouisThanks for your comments +Logged In YES useridYes Im using pessimistic locking never got any performance out of the optimistic +I cannot reproduce the error with the tests +sure ill reduce the priority for now +Both of these types are XML mixed types and the preferred JAXB binding for mixed types is a wc DOM +Shrinkwrap updated to using the Shrinwkwrap BOM +Is that correct? or Did I commit the wrong patch? From a cursory glance the former seems to be the case +Hadoop is designed to work with code thus is not able to defend malicious codeThere are just too many ways to damage the clusters by malicious code through Hadoop +I would love to see tapestry described as a superb fast stable tool but lets be honest who will use this framework for serious job if it changes constantly each month and number of jira tickets rises exponentially +just as we expect sample projects or at least enough data to reproduce internally for our bugs we should do that for external bugs even suggest you add a least a sample project files to those issues +This patch is just code refactor so I dont think it need adding new testcaseAgreeI will commit the patch shortly +mrepositorylogjlogjhomesonar +Ill try to see if I can test this tomorrow +for Karthiks approach +Wed have the same issue if a user were to set one of the parameter values as null +Verified during test cycle fixed. +This is the +Only the text wasnt setStill lots of flying around will comment more as I discover +to commit +Fixed on trunk +Still looking for a way that this can cycle without freeing all locks +See The Java Language Specification sections and for detailsParameterscls the Class object to be checkedReturnsthe boolean value indicating whether objects of the type cls can be assigned to objects of this classSo basically if I would like to create a new request for instance which extends somehow the original provided as type for the cycle above Im not able to find my new Class anywhere due to the wrong direction of check. +Then the current test case of can trace the problem of saving the instance twice +VinodI just glanced at the latest patch I did not read it in detail so if you say it covers that case I trust you +UmaIt looks like it fixed at +This is no doubt related to the same singletonprototype issue I raised in comment to bug +What could happen though is that a test that formerly used that was run in shared VM with a test that messed with the environmentedge would not potentially have problems if we switched it to +Actually the two translations are semantically equivalent one is simply more optimized than the other +I just committed this to trunk and the Hadoop branch +Reopening provide QA steps to verify this issue +Theres easy workarounds for this and it seems like an edge case to me +Bulk closing stale resolved issues +Yeah I see this as a usabilitysupportability issue. +Are planning to make any changes to the patch? Once you think it is ready I can commit it after sending it through hudson etc +There were other similar issues with bad values for ints doubles floats class values etc +I found no way of setting a virtual host name on in Java +I cant let the subjective sit there suggesting some weird corner case offormattingTake the case of an app thats localized into three languages and you as ananonymous user visit the site +That seems like a sensible exception if reading data from a different versionIts not what we are seeing on jenkins though +Hudson shows this either not fixed or causing a regression +However the explicit table lock doesnt help in your case +checked in +This is out of date I suppose. +Im not surprised I needed only confirmation from you and Ill reassign the issue on didnt think that it does anything with Seam but I have explored the problem +Moving to + +If for some reason that is not possible then I think that the javadoc should explain that this preliminary step is needed +For other types that are not defined the outcome is similar +Basically It will directly read and write data sets on local file systems and HDFS via libhdfs +Its functionally identical AFAICT +More work to come +The vm should be pingable +Okay we need to extend our unitary test to find others cases and I know its tough +As you say the problem lies in the fact that variables are directly accessible in the tag file with a tag gettings its own +author tags are discouraged by the ASF board of directors and Logj doesnt use them +Fixed in trunk under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +I wonder if this is the cause of this bug. +Blocks +r +Good questions Mike! They say a test is worth more than a thousand words so I added tests and they pass +Attached is a patch including the simple test +Can this be resolved? Looks like check in is made by Likhita +This also contains the required copyright statements for and which I also found quoted in Harmony +I actually get the same error with a single band image created through the same process +New patch even simpler. +Just to make sure this is clear the package in question is trunkr x +Forgot to add the ASF headers +Correction I just discovered wfs tests were working becuase they did not set the forcexy property in time and the referencing subsystem was already initialized with the latlon ordering +Looks like this was fixed. +Committed in +Is there something wrong or a negative impact in that approach +No tests since this is a change to the pom +Committed to and thanks for the reviews guys +Jayapal I see a number of discussions around egress rules please add a link to the discussion around this particular feature specifically where consensus has been achieved +defer all open issues to with made a mistake +I just committed this +Its fixedBR into in revision after releasing +Patch accepts the state as SUCCEEDED and ignores user request to kill when TT reports task as SUCCEEDED and there was user request to kill the taskPatch has been tested for the same +Pong +All operations of changing schemas were not affected +mike this looks good while I am not sure I understand all the comparators +How about you put this into standardtools and then we close this JIRA +We could consider this sometime in the series when the existing Faces scaffold would be written to use the templated form +btw thanks for looking into these details +can you pl post results for this one +Unfortunately my thrift build setup is so botched at the moment that I couldnt test it properly +As indicated above. +Reopening and deferring to next release due to some issues with clustering +Workflow AM can reuse containers across jobs which is a more general case +An explicit flag is the way to go then if it is really needed for performance reasons imho +Ahh meant to enter this as an enhancement not a bug +However I have now fixed it so that it should recognize a switch to any folder already mentioned in the editor before startup +bq +Try to use to use Red Deer configuration settings instead of +The Services WSDL +Its a reasonable idea but seems like a for +Thanks Juhani for sharing the work you did on this +m and make active by A mvn deployrm r +Change because WSAS has not patch creates the config settings need to connect the Trade Client to Http PHP Http and WSAS HttpNext step is to do the same for BSOPSYou will need to in the following orderorderprocesssorbusinessservices for applying this patch Drew. +The windows build now works fine +With the empty closure we are erroneously using the outer scope for content assist +Cancel patch until we sort out Ivans issue +Thanks Doug +Git will then create a new one +Pull request has been merged into branch +Thank you very much for the info +Thanks Brockrev afbbfbadfbfcc +wicketsrctestjavaorgapachewicketmodel +where button invokes a dialog for editing not just content of the text field but set of properties of object of which text field contained only name +Yes look +Anyway we can change this name to +Starting final review pass +Could this or a similar technique improve searches for things that use both leading and trailing wildcards like a search for ildcar? +patch includes the +Freddy please review +Ive got quite similar requirements and for me it was sufficient to replace with one that uses stubskeleton pattern +to trunk thanks Dmitriy! +NPE is fixed +No point doing that until developers indicate the approach is ok +Verified by Alexey. +Did you use the sample that comes with AxisC or wrote your own +In that case I think we should suck it up and take the backwards incompatibility to fix thisIf we violate in general that is a bitter pill to swallow but its still a bug that should be fixed +jd Yep youre right. +No response so closing +committed also to +Reading directly from ZK is bad because its putting duplicate code in a number of places which need to be updated if theres a change to the ledgersavailable format +I dont think this will fix your issue but is really more stable +You might want to open a new issue for that +Added classPlease post new issues for the other things +i will test this week and have some results out to by I need some instructionsI guess i need maven never used it but will get itafter that what do I do? need to build top level pom or call all sub projectsDprogjlogjrgoersDprogjlogjrgoersDprogjlogjrgoersDprogjlogjrgoersDprogjlogjrgoersDprogjlogjrgoersDprogjlogjrgoerssrcyes im on windows have used ant but not build you run mvn clean install from the root directory of the project +The order should be SC I think this is done. +Committed to as well +Also this is an improvement only specific to HTTP transport and no means for other transports hence I dont thin introduce a new API is not wise thing to do here +Committed trunk revision Committed Lucene revision Committed Lucene revision +I would only suffer from the extension processing time which is not that a pointIf I really want to deactivate it I can just override the part where Weld is finding all the extension and manually remove the one I doesnt wantneed +The tool tip with meta information should contain Captured at with the value from Captured at input field instead Created Date +Patch in cvs +Overlooked that there have a readFully function. +It was there before I understood the code and I was just marking where similar data was being used +Thanks to Kan +bq +It makes the second try to get a block from the block cache not count as a miss +Patch applied to r +QA teamAffected +Unit test for this scenario +Can I get a to revert from someone who d it? I should probably do it in a different jira too. +Ive seen this in FF as well for typefile we need to disable onblur validations just do it at form submit time +Merged to doc branch at revision Merged to doc branch at revision +I tried from review board but still got conflicts +Try this patch now +Fixed in revision . +Attached a patch that implements this improvement +The best option I see is to utilize the pre attribute to do validation before the XHR +I had forgotten this +Also need to handle List types + lgtmOn my system the ordering happens to line up but youre right theres no reason it should and it looks like it doesnt on your systemThe Hudson error is +Just today Im trying to isolate an issue with Groovy beta and Spring in my application +I ran into this exact problem today in ivy using the ivyinstall task in ant +Actually the patch only solved the first failing case in the repro so the patch is insufficient +All tests pass even +ignoring it could be a good first step +Fixed +Maybe even using something else than the if its the culprintI added this to the known issues to Camel release notes +Scott good stuffnitpick if Im doing a clean it should run mvn clean test rather than running mvn clean and then mvn test +I use EC +no you dont need an activator +OK I created and set the fix version to and +I am curious about how you think this patch makes things worse than before +If it only occurs with jaxrpc and is not a critcal issueit can be resolved as wont fix +I tried that yesterday too +This should be good to go +configure builds with the XML Exchange if and Xerces are present +Use Ping instead if you really think its a good idea. +Thanks Bobby! I modofied my own code to include a statement in the catch clause +New tickets should be created if there is new requests for improvements new features etc +If the bundle is built and installed to the local repository the patch should fix the issue +issue states the same +This is pretty much what you would expect since Ivy should be using its local cache to resolve most of these dependencies. +There are other sections editors and wizards that have to correctly handle the knowledge about the sharedlib being used +Still working on the part of things though! +to update to. +closed with the release of +A high level design doc +Given the of this ticket Im going to triage it into . +Good news A test on some of our internal data shows that the storage is saving space for us +Shouldnt be encouraged +use this one instead it fixes a bug in IE +Things have changed now with which needs to ADD docs to slaves +Fix confirmed. +for the setter +A modified version of the patches were checked in r on Qpid trunk after reviewCurrently we only check idle state on the incomming sideIn the future we plan to add code to send a heartbeat when we reach the idle state on the outgoing sideI will keep the ticket open until this is completed +Fix is in James by issue m sorry but I was using version RC and still saw this issueThe problem isnt looking up an address but what it does when it cant look up an address. +. +Doesnt seem to happen when I set compression to Gzipset set +That could be done but that will only get you so far +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Committed revision an xsdJars section to the configuration to point at artifacts whose xsds should be compiled + has never been in a SOA other jars that are in esb are also not to check if they are needed +the Changes required in Client is linked to a wrong page and the original one seems has been removedI suggest create the complete steps for creating the client +The has been completely for and no longer uses the at all +No problem thank you and good week end +Yes I would still like to do this +prototype echo clientserver using eventmachine and looks like there was never too much interest in this issue and Ben hasnt done any further work on it to my knowledge +As for I am waiting for a patch when you think that you have a good fix +Steve Thanks for the insight +to the idea Sangjin +The will then happen a new epoch chosen and the lower bit thereby reset +The has been written strictly with and in mind +Wait I was wrong you simply cant add a new method since it will change the codesNot so sure about backing out +These appear to be the same issue and there has been a fix recently on trunk +There is currently no way to properly reflect this into other sessions +Any work already done on this issue +Adding a link to since this may be the cause of those issues as well +The tests really need a fresh metastore schema for each run +However as Sonars profile manager only allows to have one extension per class name we can only use each extension we also need this issue to be fixed +Could you also try to use the? I would like to know if this is a Commons Http Client specific bug or not +verified thanks +Sorry for coming to this a little late +OK thanks +Raised to critical because this prevents us from running the examples against the MC could you perhaps please have a look and advice +Works fine when file is used for update on MS Win any idea whats going on here +Postponed +pom and log file for run with no executorId and log file for run with invoker executorId and log file for run with executorId ve gotten around the NPE using the invoker mavenExecutor by modifying the code to handle the null ve attached a patch. +there is nothing compressed +Use queue to avoid parallel AJAX should update suggestionbox together with input +this is done since a while ago closing +updated to test failure is unrelated to this patchIt is difficult to write a testcase for this patch +It probably only Major or Minor from a users perspective. +Its currently too much blue in that screen I think the kind of metalic look for the window topbars and headers is also outdatedI might experiment a bit with it the coming days +The protocol allows null values +No version logic for already deprecated classes they are deprecated so no support any more +JDK does by default +fixed in revision +There had been a check for but I believe this fix will include that +The controller on the other hand should ensure that it deletes that path when it has completed processing it successfully or unsuccessfully +I presume that these control chars are some chars that are coded in your classes and PMD simple recopy them and creates a bogus XML report +Thanks Charlie! +Also I forgot to throw this helpful bits in the bug report +any news regarding this issue? it is possible to implement this like the rubygem does +Sian your patch is even faster! I think its all the finals that you inserted Lets go with that one +class and +Hooray new test! However in v it still wraps with the and then instead of just wrapping with +This patch extends all maximum identifier names to It also renames the into the constants it contains from DB to! When will the next official jar drop be of Derby which includes thisAny document how to incorporate this patch into the sourceThanx againB +I wasnt able to reproduce the issue in the webservices module it was probably fixed by some other changes +Oliver would you please provide a status update? +This is most certainly not a good idea +This seems to have gone away uncertain why. +Fix applied in trunk along with a test case. +It brings and with it and it moves a couple constants aroundAfter this jira Ill move the base classes and then add the module +Wonderful work! The new revision works well in my Chinese environment +Brian here is that jira +Thanks for providing me useful inputs regarding the problem + providedAdjust database settings in srctestjavatesthibernateinnerBuild it with Maven with in your repositoryTest works with but not with +Reopen with further input if it still doesnt work +Looks good to me +BTW themes should work even better with web app proxy as your preferred theme works for any by default +Its just something I tripped in +will be setup on apaches hudson +There are exactly mandatory arguments and that is why I believe the test should be! +Yea assumed it was for legal reasons +Is this the issue where the user receives numerous prompts of File has been changes externally +Please note there is one patch waiting for reviewcommit +Take +Patch submitted by dafu. +I tried to apply the patch but it didnt fix the issue +Fixed under the fix +Ill close explained how this is expected Hadrian Tried to demonstrate HQ showing an autodiscovered service as unavailableby doing the following +thanks in advance again +chapter page g joined words onto the serverfrom your +Is that correctYou indicate that the checks might move once the code has settled i think the comments in could be enhanced with your knowledge +yeah theoretically stale values are expected any way +Heres a patch that removes the use of temp files +Hi Mamta I noticed the grant to ASF checkbox was not marked with the attachment +From my long experience this is never a problem +Now seems to me it is working fine +Probably best to backport that patch wholesale +Manually tested +Thanks for the patch! +If its not in the syntax specification it would seem to be a candidate for a custom extension function +This was listed as fixed but the code that supports this only exists in +Changes from Teds github repo committed to Mahout trunk +When a replica is detected to be corrupt it is deleted +and among many others +Assigning the issue to Hvard since he was the one fixing it. +Is this still an issue with latest master +I will take a look on this issue +This patch contains the addition of module for photark trunk +From what I understand about VS an +If by some chance this is still relevant please feel free to write to the listThanks all resolved issues for older releases. +is obviously going to be fastest especially if you require rows to be in sorted order so you can basically stream directly to an sstable data fileNeither +Thanks for verifyingconfirming tests pass +Yeah Im sure we could use more and better tests +test? Whats that beast? patch provided by Nicolas. +bq +Thanks for reporting Stewart. +Patch added for branch patch for patch cant go into because it is new functionalityWhy is it in chukwa rather than some where more global? Does it depend on chukwa +Correct the screen as explained in the issue description +See +As far as I know this also means that the connection can not be used to stop replication or perform failover either +Ill not volunteer to write it though +Include the running version of Hadoop Client version and revision added to outputShould we print the server version instead of client version? The client version can be obtained by +There are a lot more String values for encoding that will throw an +Our current problem is that if we generate the documentation for each application separately only elements referenced from outside the current services schema will be included in their respective XSD files + spot +cant reproduce +Done +Please reopen when you know how to reproduce this. + +Canceling patch as package cleanup will invalidate all current patches and will require patch regeneration +New protected methods are documented +Unfortunately too many mustella tests were failing and I dint think that it made sense to progress without fixing those +related it possible to have a snapshot data with all sstables mergedcompacted into one sstable +did you made some more progress already +no ve done all this +Add an option to the existing command rather than create an entirely new command +defer issues to +Thanks! +That is indeed the correct fix you can even commit it to the branch +Im looking into those now +Interesting yes good idea +there is no hdfsxxx +Sorry bout that +ok thats great +This would make the plugin ignore Excluding startup is an advantage because a plugin wont be started if it isnt usedNot following how that relates to this? These plugins are installed and supposed to be used +how important is it for your or your org that we get fixes for this in? We are trying to close down and seems like this might be a candidate for a release +DIEC glosses recar as saber greu +Fixed in r +Roshan how about applying the fixI would fix he cases with Short and Float too or not +My debugger points to a similar socket reading hangup and I did not suspect it could be the terminal causeI wont be able to continue investigating as the feature I was working on that led me to this was canceled but I could supply the code to reproduce it +it now appears that this happened between and +Id like to present my apologies for waste your time +Thank you to Chuan for contributing this fix. +Thanks Mike +Hitting refresh loads all the new processes +Thanks very much for doing this! I will try to review it in the coming weeks +wondering if this would be better served as a github project so it can be collaborated on much easier +Stephen thanks for cool patchI am getting this into the kit nowDo you mind helping with wiki documentationIt should be a child page at the Data Format section +Added further tests to and applied to JTS +Javadoc still doesnt work for me +Steve Im saying its fundamentally broken because if you hit the sequence twice and make any sort of assumption that there is any sort of cohesion between those selects is fundamentally broken +Furthermore the associated generated encryption key is not stored in the properties conglomerate +The implementation should not affect the sessionfactory normal runtime behavior and im not sure apporach is ok for this +Im gonna retry Joshuas quickstarts now +Thanks a bunch Boris +Nothing new since M +Took minutes just processing log files +I didnt read whole text since you posted whole logThe only thing which could cause this is Arquillian wrongly interpretingprocessing the maskAslak could you please look into this +Fixing Affects and Fix in Version + revision to trunkThanks Guy! +Resume file as pdf which cannot be extracted +Bulk close for +I think that may be improved with a suggestion that it might be an SSL server on the other end sinec it also may be some completely different application talking an arbitrary protocolMaking servers listening both on an and an plaintext port should be a pretty straightforward to implement +My previous patch attempts all had problems and then I found a way to change my code so it works +We should also feature more prominently the fact that constraints can be and as well as that the corresponding validator will be injected with dependencies +on the patch nowI suspect its best that we follow the correct Unix way here things were put there for a reason +I just committed this to trunk and branch +You could have the application server run a single embedded agent application that all the other applications could log toNow the opposing viewpoint is just that each application can run an embedded agent or IMHO this is really a use case for a traditional local agentThoughts +Sounds good +diff u version of the am attaching an updated version of the patch for both this bug and +pmd +Removed Zookeeper integration as a blocker +I agree with Dave +Hopefully this will go into the codebase. +Extract sll the description texts from property file to localization property file and provide translation +bq +The PB stuff would be better left to you I imagine as are any additional tests you feel are required +I was interested in the part where it says if Secured is applied on the class then usage of does not have gone in the loop of downloading grails its to get the jar that contains Secured class +With the release of EAP pending all relevant JIRA issues which are open are being reviewed +Running the modified the problem can appear whether using or +Jiggering test so can have two filesystems in place so another can grab lease of former to recover files of killed regionserver +All done the fix was just to make txnUtils more robust to getting null values in the response JSON +over all +If the modularization shows that some classes have to go to different packages then thats probably a good thingOn the other hand modularization is certainly not a must have for me +If you want to output a property value then use the property tag +Committed to branch +Reopening to set release note is already documented in the release notes as a new feature of the VPE +Is it possible to write a new regression test with some Long values which require andor bytes to encode which demonstrates the bug and the fix +Lets close this for now if you have comments about reconfig documentation please comment in or the shared doc +Implemented earlier added to config attributes are still unimplemented +All the regression tests ran cleanly with the patch +attached a non git style version of v of the patch +Attaching another patch with the correct comments +Reconsider for generics version +LGTM +As said you can already do this using a custom decorator. +When this message is displayed would be better to not display the title of the chart if one is defined Could you integrate the last version of the widget description Display a components source files in a Bubble chart +Glad it worked out for you +I tried to find a way and explain it how Castor could avoid closing and database sessions +page parameters were done in +HiWe experience the same problem in final in the latest snapshot +Then you can implement the BSP message passing latency and bandwidth measurement tool +What is hidelink in your code? There is no such component +Eric I think these two patches are equivalent except perhaps for performance +Looking into how best to fix this edge case +Please reopen the issue if its still reproducible +Unable to reproduce with latest problem is setting of the IE browser on the Hudson machine dev +orgapachehadoophbaseipc + +Changed to JSF RI because this version is contained in all our demos and also in M. +Another change Id like to do is to share the same bulk assign code +If you finish it before we release beta please correct the versiond like to see this change in beta +I merged Arabic analyzer and existing Brazilian analyzer in contribSince changes are much I post the result as a zip file not as a patch +The is either to always use a different groupKey or to use separate tables for each Message Store +thanks +The patch has been appliedPlease verify +Thanks! + +rule files +Adam thanks +This is +Thanks do you know if we can replace the jaxbintros in the WS with the jaxbintrosCheers tried updating the library and the testsuite passed +Give me a week and Ill try to get something in the commons package that you can begin prototyping withOr you can wait till Portlet which does add fiter support +just checked in a fix that seems to work for FF and IEMatej is checking some more +NOTE The default messaging mechanism is still JMS using internal Broker +This approach works on Mojarra but still encounters the same NPE as before +The balancer was made pluggable and that feature shouldnt go away +I havent been able to reproduce locally yet but the patch should fix the problem since it is essentially identical to s root cause +Benjamin what you have said is true of Maven but not of Maven +We should resolve this for +To clarify the aim of this JIRA is to solve the use case where lazy deserialization can be done using the cache loader provided by the user +Merged the two test cases and ran them successfully. +bq +Attached is an implementation of a custom store function that splits the data dynamically based on the values of user specified key field in the output seems to be failing during compilation as my test case defined in package is reusing certain classes from namely Util and +Therefore if you have more than concurrent requests from the same client they can displace some states from cache that causes +salaboy the error reported happens at in the server ve never had time to fix this issue and test it against postgreSQL +Thats right Si +Thank youLeo +ejbhibernate + +Sounds reasonable to meWaiting for Pauls review of the rest +Its a subtask of or did you mean something else +Can you test setting up and starting an EAP server in JBDS without TeiidDesigner in itWant to make sure its not a AS Tools issue +Comitted with tests +we need to have stepped on children invoked before parents so formcomponentpanels work +William Did you really mean to attach the same diff as for here +See for codecache issue. +Could you regenerate it +Is this still valid +Resolved thanks for! +We need to move from verify normal operations work in the face of failures to normal is overrated +We will have to build it in a Hudson job as part of the job I guess +I fixed this issue by adding the before the but this is only a dirty work aroundIn my project I have to use the to startstop the Seam lifecycle that is also needed by thePer spec we have to fill in the operation name in the context when passing into the handlers +When running on users see failures for the projects that were working on I do agree about injecting dependencies and that issues should be separately logged on the xmlbeans plugin. +bq +CXF SOAP faults on invoke now handled +Pushed to master +Permaine my understanding from your last comment was that certificates will be generated inside EWS installation directory +Issues resolved in +This version of the patch rebases onto the latest patch from and exposes additional read write methods on +However I donot see that it should be very difficult to write a small test programthat opens a connection and sends a message using you new interfacesI will factor in tests for this code +Damn! Too many at once +Thanks Kim +I am however unsure about the difference between no namespace and a namespace being the empty string +to trunk +Can you run maven with the X option and attach the stack trace to this issue? Thanks +I think the current box serves the purpose +Actually I was the one who requested this normalization in TomcatLets see whether we can convince the Tomcat devs that the urls above are OK +I will attach a new patch that implements thisI first thought we could also move the cxfbean package to the new component project +If we dont who are we to assume that the source isnt just slow? Doesnt seem like there should be a timeout at all to me +Can you check to see if the is also empty in the server repository before you try to convert it? If it isnt empty can you attach the application to this JIRA? Thanks +mail sent to and soa production team to clarify this issue +Final version which grants rights to the ASF +I will do a full build to ensure no side effects is introduced +RPMDEB works as described in Owens comment no need to specify environment variables +Please check your settings + +Previously vhosts mentioned in would be automatically created +Merged the into the master branch after rebasing +Sorry +Attaching a new is the status of this stuff Christopher +If not we should add those to the METHOD call +Please reopen if its not solved. +I have opened to track the issue of default constructors in particular +Only one of them can win +This looks good Ning thanks! I will commit +patch looks goodIt seems that hudson is not working +Roo only checks for annotations on types not on annotations themselves like your verified that annotating a test type with does cause the annotation to be removed from the ITD so this works as designed and has done for a few years! +but actually done. +I changed the peek to getand added proper invalidation you try the attached jar and let me know if this works for youIf yes Ill do a proper VFS no Ill keep investigating +zip attachement in an Eclipse IDE andor environment? I did click Build All on but magically nothing happened I could compile by looking for eclipse plugin jars it depends on +simplified forbid to only handle the update case removed the test cases for from seen in earlier patch versions +Execute onclick of facet command if set instead of submit the actionId +resolve the issue since it is fixed in current trunk. +It didnt seem to work +It becomes more clear to me now +Add Waiting for CR artifacts are synchronized with Maven Central +group by abwhere data is bucketed by abcA mapper might geta b ca b ca b cin which case the current algorithm might not workAlso change the name of the variable in is to bucketedCols instead of +Yes this is because we open file from Web Project have preference Tools Web Editors Always use Tools editors with Open option which is used by Web Projects View +I had a look into Ambrose and it could be an initial step to consider for sketches for the visualizer also we have to consider some other visualizations options +This generates correct WSDL with the current trunk +You will need to edit the files and remove the references to the JMS queue Timer bean should also check our file and compare it with the current default version in the jar file +So they are just ignored +Hey guysIs there anything wrongDo you have an idea about when the project will be updated +As I write in the issue I have tried to seperate the two by space comma colon and semi all cases it seems as the compiler tries to open the as a single file what naturally fails +I committed the patch to trunk +declare statements are not covered by the proposal if also a rule element is possible +Thank you +Committed +Hence marking the defect as Blocker +Hi Krzysztof I am not sure I follow you but I did clean up the fault checking a bit +I was going to do this by parsing the XML file to figure out the names +Im running into this same issue on OSX using RPM +Fixed now +Defer beyond Release +It would be good to leave at least deprecated dummy subclasses in to avoid breaking things +big patch combined all the individual patches +Hmm Ive never tried ediing so Im not sure what to expect thereThis is a bit more work but you could try adding to the lib directory of the plugin and then edit the plugins manifest file to include the jar + +please could you update my public key aozsihyjvOlyDg +Attached to Paul Rivera for provide us this patch +Fixed within +For the other metrics you may want to wait for to complete and do it in another jira +Assigning to the reject button should take this issue back to Open the accept button issue should go to resolved +Committed patch +Can you do that? If not let me know and Ill get around do it this week +Not sure if it is really related which I need to find outThe issue is that I found that tests in a same test class are executedvery randomly +Hi is there some advance in fixing this bug? Thx +Bulk closing issues resolved in JEXL. +Fixed by SVN revision fix version as this hasnt been merged to branch +will be part of next is empty or undefined in no webdav will be enabled. +Changed the text to show what is actually shown +Of course add line numbers is set +Your patch doesnt match with the trunk could you provide a new one? Tx +Amardeeps patch was completed intrunk r r r +Hm seems the switch doesnt have any effect anymore +Mark Ive noticed that the JIRA is assigned to you +Added a better error message for now to indicate that class reloading is not supported so far for such as vfsTo add proper support needs to be extended to support not only vfszipToo late today will check this m bumping this to because I just checked the xwork source and it doesnt appear to have been updated and we just got a release yesterday so I dont want to hold up Struts waiting for another +test case will be broken if entity will use Transactional cache strategy +Committed to trunk +Thanks +All scenarios are accommodated if it just werent thereConsider this too +Is this basically about changing the slowstart to be nonlinear? ie instead of just start reducers when x of maps are complete factor in the total number of maps in the job as well +Ive attached a patch to the myfaces issue +The bundle refresh problem has been taken care of in +So all issues not directly assigned to an AS release are being closed +FulinYes pls follow as Ted suggested +This way we do not have maintain different I can see the impact on existing application which heavily depend on current behavior +Its unrelated to this JIRA +push out to +I have added the content as a entry +Applied patch and fixed some mixing french translations. +For this blocker bug we are only disabling those options visible to users +Updated README instruction for generating site content +It should implicitly set committrue +Justin should review this one +Great bug find +For example with the ROOT region we used to get it from the master +prune columns the logical operator belowI have to leave the logical plan in an inconsistent state in the middle +I guess ant task validating all in Seam srource snapshot would be very useful since it will do the check repeatedly +for each authors company? Make a choice and stick with it +I have also spent some time looking at the jdt core plugin whee the extensions to the classpath container initializers are being used +Chris feel free to file a JIRA for rolling of stdout and stderr and we can look into what it will take to support that properly +If you do so please do not forget to set this scope default access to private as well. +What Stack said +With is fine for me +Committed to trunk. +Attached patch by Xiangtan and me adds unit test coverage for the new behaviour in +it works ok +Im asking because that also means that user will be able to specify as many timestamps and consistency levels as he wantsThats what I said +Just today we were talking about having our own ids for workers because of some other optimization maybe that can also be used to fix this problem +Anil would the new attribute need to be documented for all customers in the Sec Guide +But now I couldnt reproduce the issue +At present it is very easy to find out the last version of the file in TrashAnother option would be to make the Trash client retrieve the contents of the Trash directory and then scan what files in the list +If true we can use them directly +Here are the provided by peter royal upgraded to the latest cornerstoneexcalibur and a patched to make it runningWithout the patch it does not start due to validation errors in the of the Peter Royal will tag the sources in SVN Id be comfortable to put these into the test environment but I dont want untagged code in our builds +Fixed with revision +Can we further break them down to the stats which will be collected automatically at insertload time vs +The results show a increase in performance for this patchsdeleted docs skipto code + Also the conf is in a strange place and has been refactored + push to +It will be easy to compare the clusters produced by spectral vs rest of the clustering algorithms +It also contains the serializers messages an message services +The exception was indeed misleading and is more correct +You cant rename keys in JIRA rather you have to make a new project with the new configuration and move all the issues over into this new project +Committed to trunk +Is it a fundamental requirement that any of Derby object be case insensitive unless specified by enclosed quotes? Yes when in the SQL language +Thanks +Worst case we could use it as as stop gap solution till a high throughput file channel is availablePlease share your thoughts and if you agree i believe the concerns you highlighted could be worked on in an incremental way with your inputs otherwise please advise on the correct route to be taken here +Speed wise the problem is that the WKB format is encoded over the wire as a set of two hexadecimal chars for each byte and on the client side you have to parse them to get back the binary version that can be handled by wkbj that exaplains why is as slow as wkt +working on removing the usage from s an updated version of the patch which removes dependency of from +Note that it is in the JDK orb IDL processing rather than default build does not build the JDK orb but CI will which is where this issue was first discovered +I see the source of the problem +However we should either remove that logic entirely or find a way to make it work properly for +Any update on this +Id say that the HDFS is such a system that you do not really require backups as long as you have active node monitoring and for a few files its alright if you do itThere are reasons why people still make backups of files that are on mirrors +if you turn on FINE logging you might also get some more information about it +We dont need to suppress tombstones we just need to make sure that any data that the tombstones were compacting is new enough that we dont need the tombstones to suppress them +Patch applied thanks +Thanks for the attention to detail. +Added test cases from. +Any new updates? Very interesting ll need to move it for we did not have time to polish this idea +Committed at revision +Attachment description Exception message shown in the stack trace not even visible in the eclipse error log? If not can you try running with the eclipse console view and see if the full stack appears there? I think it will appear if you launch with eclipse consolelog +to FIXED +It seems that has already dead for some timeI can not find any code that deals with intrinsic at +I agree that we should search the whole hierarchy till we find one that implements the +OK got it. +But yes it should allow for this + has been marked as a duplicate of this bug +This is the version of that should be included in EAP +And then i tried experimenting the useNaming setting both of the config to true and false but still the same FNFEJust a wild guess from googling around if this has something to do with the classloader classpath where it couldnt see the content of the war that has the +I am definitely open to other naming +I have modified tests accordingly to handle the same +Dmitriy Moving the get out of the compare method should give a significant gain for the case that you were seeing +Running into a bit of fun with not picking up the CRS by default with the first include +failing because its not currently upFix as suggested above is to retrofit Editor and Root to retry on failed connect to server +I dont have this problem you use the windows installer to install GroovyDo you have GROOVYHOME set +Will conform to the cluster its being run on +It works for me +And of course daniels docs are still rendering fineThe were several problems to solve The clippingpath is part of the graphicstate +After you execute mvn jettyrun try the commands I listed before +It would open up possibilities that would require scrolling through the complete result set several times +We can open a new issue for changes after review. +Patch for trunk +looks like it just needs docs +Committed revision . +Owen the critical point is that in order to use Maven we have tocompletely replace all of the build scripts and probably write abunch of Maven plugins too +Merged to the branch in revision . +now the connection url without localhost value +need to be rebased on trunk +these were committed to +Thanks Doug. +The first case works because some special case logic has already been added to check the immediate parent typeAs for a solution we can either Look recursively at the parent types until we can determine correclty +Dan askedWhen Derby is shutdown the autoloaded driver does not remove itself from the driver manager even though it seems it will not honour any connection request until the embedded driver is explicitly loaded +test use the number of files in the logs dir to determine the map count +Hi Brian Ive opened thanks. +For the backdoor you are describing John suggested there could be an admin mode which will skip these checks in metastore +Setting to future pending a patch +A patch for which enables these extra paramI think you have already committed this patch so I mark it as resolved. +I agree +Closing. +Thanks I was just going to ask this +Which are the recommend to set throught + caveat RSA usually breaks for some reason +Documented as a known issue in EAP release notes +Thanks Elliott for catching this +Will this get fixed please +If the above points are implemented make this website the root application as a one page introduction to the ofbiz systemThe advantage of the above is that the site can be maintained from svn as the rest of the system +Hi YashAny updates on this issueThanks +I put in a fix in r but I am leaving open until someone more familiar with this code has a chance to check +This will be using a new compiler that is based on JDT +I suspect that due to the schema visibility requirements of random access there will be significant differences not just in implementation but APIAdditionally it may be possible for an index file to encapsulate all the random access concerns and use the above format for its data storage +Now we need to insert it into the namespace and we need to ensure that new calls to create the same file should fail +Talked to Todd on IRC this maybe similar to same as ll take a chance and resolve as a duplicate of +Typically wed been waiting until API incompatibilities forced a new groupartifact combination to permit otherwise incompatible versions to live in the same classloader and we try to preserve backward compatibility between versions whenever possible. +Nick sorry for taking this on so late but theres too mayn things to do withso little time +I did not get to review this today +Niek do you want to upload a patch for this +Removed Managed X since we can compile this specific test case now +Bhushan can you refresh your patchI am getting some merge conflicts +Please let me know if this works in your environment +The is included in most modules +Ah sorry I forget there are two motivations really dont like s kloc either +Throwing the exception seemed like a better choice +Can someone review the port please +unzip the zip folder in srcplugins in Nutch src and do the patch +Fix for jdbc driver +Committed as svn closing +bq +It does but you have to specify the target charset while constructing it +this issue just got worse for me +I would probably just make it +seems tho +Updated webserviceconsumerwise to reflect new smooks behaviour + +executes commands until COMPLETE OR ABORTED is returned by a command +I just improved that traversal algorithm. +However it would make the build system more complicated and probably cause confusion for users +We have other priorities right now +Attached a patch file that removes the behavior for registering namespaces as it was no longer configured to use that behavior in our codebase since the JCR API was supported +I like this +The policy code has also been refactored into its own package +Corrected the title +the patch looks good to me +Thanks +I think returning a split for an empty input is not a good idea +Patches applied thanks guys! +Source code can be found at the repoThanksIll try to check your patch adding the change I want to confirm the places where I should add Wysiwyg Editor +Escalating to Critical as this issue effectively blocks from running on Geronimo on Harmony + as incomplete since theres no reproducible test case and no followup +Which would be some perfect new shell that is not so married to bash and hadoop I imagine this new job shell would do explicit class loading and would reference nothing except HIVEHOME from the environmentOption +But not all users need four server adapters is it necessary to bundle four server adapters into one distributionAny comments +Raghu how is this affecting? +The last I remember while implementing security for YARN components I added IP addresses instead of hostnames because I ran into issues with using hostnames at the RPC security layer probably related toIt will be ideal to replace IP addresses with hostnames everywhere +Main page has a listShuttle which wont work +commited patch along with some documentation additions and test cases. +Currently looking at this +I think that makes sense +closing it +If nobody objects Im adding a an explicit subtask here for that +depiste the failure mentionned by jenkins the patch is visible in the source code +Issues re +Please do so +to answer Stig Lau As you can see the issue is targeted for so is the version we are going to fix this by introducing several breaking changes +OopsBTW your test case was completely broken it assumed a database with data in it. +Thanks JuhaniRev dbceaeafebbfdddceff +If theres a leading slash and contextRelative is false we should prepend the module and context prefixes +The escape character in is +I have a few ideas but it would probably go a lot faster if I had logs to examine since I cant seem to repro it myself +confSystemLibraryFrameworksVersionsHomelibUsersrjurneypigbin +Cannot test that at the moment will check it when I find some time +if you clone the plugins please use the next branch which uses the API and should gemify the versions as well +done +I cant see any issue at this end nor any connection attempt by youCan you please use ssh vvv +Downgrading priority as Jozef thinks this is not a blocker +It could be interesting to add negociation version and let user select what version he wants to negotiation should be supported +I really wanted someone from team to take a look on as anyone got the task I will assign to me +attched patch in sync with latest code in patch and it passes all our internal tests +merged into producttrunk in r +Write the test case +I just checked but Im not seeing it in +There are better ways to optimize for that +Sorry the bug is actually related to Ubuntu repositoryI had a similar problem when compiling project +original filaname for the note +I guess the fundamental question here is should this one be extending at all since its ctor takes int valueCount or should it have a different base class +Can you please give us more details about how to reproduce? because up to now we cant What did you modify on your vanilla EPP and how? a local or remote DB? if remote a wired or wireless connection +If we use external classes we cannot dictate that class to implement a +As mentioned before sometimes it is not easy to patch Hibernate or other third party libs +Yep Ill see how we can move the information to +This job uses the map phase to evaluate and copy changed files +That should do itWhat have you been running to get to this problem? This is a side effect of and is extremely hard to reproduce +It doesnt sound like anyone is stepping up on at the moment so they can refactor as necessaryWill also add in the necessary testing +Updated URI in settings cleared cached pages +with a subsitemap generating RSS for certain pages +Ian didnt you put in place a generic error handling page which would prevent this sort of error from getting displayed +So we should get a common format on all platforms +Thank you for point it out and submit a patch +Oh I have had many issues with codegen so I dont really even know if the patch solves a problem in the code or if a misconfiguration on my part resulted in me trying to hunt down a bug that wasnt there +The version is Hadoop appears to be a dupe of +It should not cause any errors +Jan what else would you like to see in order to help beginners along? I completely agree that having an effective channel for beginners to get their feet wet is something that will help everyone +Loading websites shows the same problemIt only appears to be a problem in the situation where you change HTML pages and only on iOS +Ive rewritten the servlets that rely on the url parameter to use the additional path part is a little bit tricky If you configure a project it might be possible to change the projects name +Updated patch to address Todds comments +Also this is more involved than I expected at first +Thank you for the pointers +The current api command fails in my view on this one as it only offers an all or nothing aproach to getting processlog instances on a process instance +For people who are paying RHT for support we want to be able to provide additional value were developing the problem reporting API to enable this and are going to do a lot of work on problem matching and associating solutions using various data sources some of which only subscribers can access +Could you build hibernate plugin and test changesAnyway I am bit misleaded by your way of definig subclasses +anyway take it as as worked for me feedback the way I wonder whether requires is really required for the +because of the backward compatibility issue that has been identified this patch uses different escape chars only if the OS in windows +Thanks for reporting this mismatch in behavior + +Modified both marshalling and simple Web Service inbound gateways to unwrap the cause Exceptions from any Spring Integration test the trunk build with your code if at all possible to verify that your Soap Fault is being returned as expected I will try it on Monday and revert back +So a mechanism to bring the out of sync audit logger in sync will be needed +Move to which will focus more on supporting the switchyardjBPM execution environments +i fixed the errors you noted i just uploaded the wrong file +and +Clarified the instructions +thanks rodrigo +abortAfter function works as advertised other onError behaviors still function as they did before +It may require quite a few changes to make it work on WinWorking on it +I tried that and it works but the problem is that a small width column is created in title +I dont see anyway around this. +Well guys for logging libs my philosophy has become hold it until it becomes too tightBut if you want to change it to slfj I dont have any objection. +convert this image withjava jar m imagejpeg d q dpi and withjava jar m imagejpeg d q dpi to see the effect +Thank you for sharing your thoughts +Hopefully there will be compatible annotations release soon too +Verified by Leo. +That is by the time we check the validity of the input string its already normalized as per the whiteSpace facet and all those characters have been replaced by spaces and then collapsed +Should fsck perform some sort of HTTP client failover? Should an admin be able to run fsck against the standby NN to make sure its in +If HADOOPHOME is specified then it overrides hadoop in the path +Fixed in revision +In the previous versions this was being ignored and the image had no transparency at all. +Two approaches here Mark the fields as updatable and insertable false as initially requested +It was suggested that it may be an SCA proxy problem +A heartbeat is a special packet with seq of +Snjeza could you detail here how a random wizard can work with a project example descriptor? cause you got me puzzled here +Closing this jira. +Is there a reason we have avoided pulling the logs directly in YARN as well? If not should we do this for both the AM and task containersI see your activity on and +No tests added because the plan generation is correct and the final output is also correct nowThe main difference is the execution path at runtime +index blocks could be treated with MEMORY priority in the LRU cache +Or should there be a more common issue for rule problems with inner beans +The smoke tester with the latest patch succeeds +Fix verified +ASM itself doesnt guarantee backward compatibility +Can you attach the requestresponse +Hmmh +With ian to are now shown see screenshot. +The legal concern should be treated as critical +I would prefer to have less blocks for handling exceptions +Added a authenticationauthorization provider to be enabled if and only if the class is on the classpath +In GANTHOMEbinstartGroovy a stream edit that should have happened hasnt +I mean one cannot use an updated commons while the others are left behind +Run this and Ive created some data a comment and click Click to Add +The whole footnote is dropped +I was thinking we would do it after the first that resulted in a tombstone +I am resolving this for now as it appears to be an issue in +We now prohibit entering text into the file input via for IE +push out to +just committed this to trunk +Exception in thread main Bad version number in +will be fixed in in revision is a bug that affects the internal CR release Ive already changed the Affects version +We should just drop the isUber stuff +Im guessing this is a typechecker error +Is that a mistake? Are there any ramifications to changing it? I agree with Konstantin that this stuff is very dangerous but also agree with you that its important for HA to fix it up +This issue may have been fixed in the upstream kitchensink quickstart we should start by updating the codebase with the upstream code +Ive submitted a patch for this issue to the trunk in r combined with +resetting P open bugs to P pending further review +Yup +Perhaps we might support a binarydatafragmenttojson but Id much rather see the data file utility first +ok well i was stirring up trouble recommending this filter from Solr to users so they can index hindi etc in the short termits a pretty good workaround for the missing unicode support in lucene but hopefully this wont take much longer to fix +Oops I just realize that some hdfs changes are already accidentally included in and committed +Talked offline with Ning +But we are looking for list of offeringsFixAdd condition to look for empty response +Similary was added to wrong section. +The other use case is the that insists on using a round robin load balancer +This issue was migrated to +in that case closing as there any plan to fix? +Committed revision The framework to use is specified by the properties in etcIt can be switched to equinox or felixThe main jar has been modified to dynamically hack the classloader and add the required bundle into the classpath at runtimeIt seems there are a few problems with some of the gshell bundles that do not start correctly thus a few commands are missingNot sure what happen yet +I just want that my group is used for counting the numFound resultThis is really weird +Committed to trunk +I forgot to made changes in here is the patch for. +Should we close this for now or you have more goodies coming this way +This is not possible to do for kind of ejbs +a new improvement support multiple configuration back ends added +bq +Grid looks like you have a hardcoded src destination instead of the actual folder location +Confirmed and exclude lists updated +As Patick pointed out that transient variable should not be used in because dict is used by cuncurrent threads +Yes it seems the group has been deleted worth trying to contact the developers directlyvia sourceforge +It seems like it would help to have the datanode generate a unique identifier the first time it is run and save that in the data directory +Trace B happens because needs permission to set the context classloader +Atmosphere should differentiate between failed connections and normal disconnections +Please pull the master branch and test it again +pushed +Good idea +Great test case really helpful in narrowing down the various issues +Why two steps to edit user profileWhy not directly link username label to editHere is a SNAPSHOT of the information panel in Firefox +built via are probably common but not necessarily those that parse the local date format to a and only these are affected +Yes I agree thats much better +Rebased and merged the second into the master branch. +nice that it works on linux + +Im curious why its a bad idea to create overridable methods that are called from constructors? Is it because a subclass method could create objects with references to the created object? +A fix is in SVN +You have to point it an eclipse install using the property or use the default specified in the file +fixes findbugs warnings +Well make sure it comes into though +Ive verified that it causes no unit test regressions and fixes the problem for us on Linux +This feature is also what we are thinking about We will probably integrate a annotation editor with jboss Web Service tools and provide fast track wizards for adding annotations +The problem isnt itself +but I wonder if that shouldnt be an additional process to what youve already madeTotally agreebq +How this is configured is very implementation specific +With the development of this one plugin all major and data stores could be supported +ve just committed this +I have the same rehashing problem commented by Erik Salter +srcSending srcmainjavaorgapachecommonslangreflectTransmitting file dataCommitted revision +i just committed this +patch available making it a candidate for +Thanks for catching it +but in the meantime Untappd whats your main componentslib setupFor Twist its Zepto UnderscoreBackbone Hogan templates GPS watchPosition stop action Bing I dont see this in my most recent crash log that has the ANR +On my Pentium M laptop under Windows XP for I have following results gc v gc v Fix proposed in +Hmmm +this should be removed in this patch +could be removed from the JTAX which would instead be able to reuse the JTA version directly +Thank you for your reply +I guess one way is to generate synthetic parameters and collect values in a map passed as an argument +If still relevant reopen or create new issues. +May or may not have a good resolution yet but should be filed against Warbler instead of. +On the other side information about rd party plugins is now little bit confusing because installation is working without manual installation of these plugins so if somebody decide to install this plugins manually because its written on mentioned web page its additional work for himMaybe we can leave links to rd party plugins there but change content of web page to have these links there as information not as installation instructions which are necessary to be done +Committed in r +feedback incorporated PDF checked doc has shipped for GA so Ill set this JIRA back to Fixed in GA create a new JIRA for CP with the last set of comments +In retrospect I agree about moving that to the error handler Ill do that next week unless someone beats me to think this can be closed. +was a Struts class back in the Struts days it was then factored out into Commons which Struts then usedYou can find in the package in the version you downloaded +The tags should be serialized with the KV inline so can be read with the KV data in the same read opWhat Im doing now for backwards compatibility is write the value length as negative integer to flag the presence of tags and store the tags pretended to user data as part of the value section of the KV +Either the builder is missing from the install or it belongs to a project nature that is missing or disabled +The added tests pass with this patch also +initialization cannot be moved out of the constructor without moving ledgerManager construction out +You need to know the integer identifiers of the operators and the layout of the token queue to craft expressions that cause the infinite recursion that leads to the. +UsageRun this in Agentadd This will generate system metrics stats every seconds and serialized in JSON format +Its not a bug so it does not belong to the issue tracker +I have committed this +Which release do you guys think will have a fix for this issue +Thanks SteveHeres a new patch w that TODO done +It is due to the original beans module does not export resource directory as the source directory in eclipse as I reported in +No maps jobs are valid +Thanks +Closing tickets +Amen to all the above comments +In this case you cant specify a home so multiple containers cant be distinguished +Problem is that palette contains components from both RF and +Try this marshalling fix basically prevents the NPE but we can end up with a view thats greater than the digest +with ANT +The and trunk updated for the latest trunk +Moving unresolved issues from to +Updated S +No extra tests +seems taht my other apps on the classpath are not having problems with it so I will remove them does continuum need from the system path +Ok here is a patch that throws an exception when there is a problem in the start of jetty server +CR becomes GA so for now no issues to be put there +Patch file +better exception message would be great +IMHO or may still be confusing +Yep I can confirm it works +Dont worry he already had an account +I guess it will resultin a rework of the parser rules +Includes updates for these files ve attached the right patch file for this JIRA +The XID was preparedcommitted XID was preparedRetryAt this point no matter what happens XID has to be committed on recoveryRecovery on my case was not working +The application server doesnt resynch withthe persistent data store if you are using the even with option D in effect +Sitemesh filter before Struts filter after struts dont see the problem here looks like you had a JSP error that either was or wasnt decorated by sitemesh depending on the filter positions +Lets do it +This implies that the jcretag property is required +I have opened to address this +On Ubuntu x Im not seeing the incorrect output +This patch eliminates the use of +for advice about this +Need that one confirmed +done +Yes thats exactly what David is asking for +Thanks Chrisrev +For example to run on an server set to truerepopen de bug if this solution does not work for you +Matt Im scheduling this for as I think it needs a little discussion as it does alter the behaviour of beans + for the patch +There are instead separate and targets in ant +This gets rid of extra bytes at the beginning of the readme file with the earlier patch +Passed on passed on and Oracle as issues being marked closed in conjunction with the release of. +Every once a while mvn fails to run any tests after looping but Ive determined thats mvn being mvn +Now that I think about it multiple languages would be handled at the Ant level not in so theres unlikely to be multiple Properties objects in a single +closed at teds suggestion. +Thats because there is no synchronization between the task that updates the server view and when the given module is being deployed +Applied for and. +not blocking enough to get fixed in removing setting to to +The entity table that is to be updated does get referenced by alias in the but the alias is never specified on the table itself +Nice tests +Especially the USERHOME file could cause troubles for doing aggregates cvs operations on the Project files node +Fixed in the trunk for +bq +revision . +But is not handled correctly when processing the expressions +The attached file is a patch fixing the problem +danielsh Regarding the content No everything that ends up in a release is held in the subversion repository +While it is normal that the client is actually not dependent on all theres no automated way for Maven to determine which deps are not necessary for the client nor is there a way to specify this in a separate way +Updated patch with and +I believe r on Branchx from last Thur is my last commit that includes anything that would be relevant to these tests +It is not the security layer +Does the assertTrue message really accurately describe the exception? Thats actually not the assertTrue message thats asserting that the exception message contains that stringbq +Thatd be perfect Ben no patches for the issue +I just verified my fix on an independent machine with a clean checkout so you should be good to go +The problems I identified in my review of have been fixed +Maybe we can call beOnce this is done we can look at creating an library that will provide a single object that an AM can use to manage its containers + released issues. +reverting that change to fixes it now the question is do I just have something setup wrong since obviously returns some really low unfortunately when using it with Cascading I still get eternal iterations so itd still be good if someone could take a look at the patch +proxy error double all resolved tickets from or older +Duplicate of +Update to remove temporary file creation and import problem is when using the zis directly if during a parse the stream is closed you cannot access the next file in the ZIP +However if the first chunk processed correctly it seems that it was successful presumably because it was part of the first business transaction +Attaching a patch for the same +It first process added servers and then bundled server +Marking as closed. +jee subpackage +Damien Lecan Sep AM When this plugin will hosted by codehaus? Exactly it would be helpful +I agree Woonsan but once we have support for dynamic reloading on resource bundles the problem in might be dealt with tooIf I understand the report from Frank Otto correctly he only modified the resource bundle and on redeployment those changes were not picked upWhich would be correct as currently we run a checksum on only and so with those still matching up we simply assume no changesIf the only thing Frank wants is merging resource bundle changes dynamic reloading from an admin UI would do the trick for him just as well. +Forgot a few files +The addition above is when one wants to use it in the default webapptask page not the customizable one + has been marked as a duplicate of this bug +if the incoming xid is lower than expected you lose the outstanding change which may be the next one in the pipe which will result in another warning and skip and so on +But I get the same with yesterday mornings build of X too +We still have to create the in svn +I dont want to materialize a collection +The method doesnt make a lot of sense +Committed +A unit test would be good codifying this perversion since it will be brittle being not whats expectedIm against changing the behavior of the default case in +is another related issue +pushed upstream +Attached patch including IT +Ive put my hands back on it +I had a brief look at the current state and I think there is something wrong wrt +Adrian can you throw up a comment on here when the changes from are set since you mentioned it could change +Incorporated all Chriss comments +eg try using jaxb unmarshalling which would require on classpathThat should cause this exception to be thrown +David thanks for uploading your classloader work +Last try +to Todds idea of similarity with for the fix +Pl +try this brokerURL tcplocalhost brokerURL tcplocalhost +This is still happening to me with the Galaxy S Android Cordova +For me both are OK +Maybe some packet gets corrupted in the network under high load and the underlying implementation doesnt recover from thatI will however turn off compression and try to generate a corrupted file so at least we know something else +The Vm deployment fails if there is no local primary store +We have to generate based on the class only there which is where the specific uniqueness problem comes in +This patch upgrades the jar and native libraries also fixes a stragler unit test in plugins that still used add +Since this is really related to error handling should belong inI think the cleanup should be a separate issue +Many thanks for contributing the patchOleg +I have not tested explicitly but code seems OK +Will add the new page later to wiki and also the release note should have a link to this one +Its freaking annoying that the deprecation text cant go in the Deprecated annotation +Unzip and run test scripts three of five runtime tests fails +Can you post the reviewID the status shows Ready to Review +All the links link to the GEOS confluence space. +Abandoned no longer necessary. +I am OK with not having any new unit tests on this one +Even if it can not find any applicable hyphenation rules FOP needs to enforce some kind of hypenation to avoiding overflow into adjacent fields for long words in only UPPERCASEANDUNDERLINE +By definition if the field should not have a length of it is required +I tried running the test again but this time I did not end the conversation before redirecting to the search form +Ill continue on the other ticket anyway so that we can close this oneBest regards Serge Huber +The second failure is because the usage is now appearing after the expected error. +git config true FTW +This is using the latest WF +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Files relative to my previous comment + +Everyone is still invited to review this please reopen this issue if you find any problems +This would be particularly useful on Google app engine +What is referenced n the +I needed this too +The only question that remains is should the list return the parent or not? In my opinion no since we want the children and we know the parent. +Im not setting it to NULL +As for the timer service javadocs I need a fresh pair of eyes tomorrow to see what new javadocs have been added +Bug has been marked as a duplicate of this bug +I have not played with it yet +I was able to increase the performance of the stock quotes refresh and now I see very frequent updates +Rosta Sounds like a regression +tested! +rUpdate the NOTICE file +have you considered adding a test case that ensures that a JN which experiences this scenario will return to participating in the quorum after the next finalizenew segmentI plan to add this test as part of +I will downgrade back to OGNL and hopefully get my things back in marching orders +Fixed the two locations I found with inconsistent locking +Hi ClausI have been given a rather urgent assignment by a customer that will have me completely bottled up for this week +Same to you if any one can take this up will be good to see this fix in next release +Thanks Nick +The patch is available for and am porting that to Hive and trunk +This issue will be fixed with the next update of +Done +is on. +Thank you Daniel for the review! +Im not seeing a crash +I would prefer simply trying to solve reporting for CMCE we only have information on those tasks because we have to tell us about it +bq +Creates the test image in pretty much the same technical context as the real ones +could be a new Feature Requestand will be implement in futureThanks +This patch fails in releaseAudit for two new html files +The tests i made with different sizes didnt show any difference on performace or disk usage +Couple of inbuilt method exists to parse the XML and fetch the values from the child tagThis will provide easy way to do the job instead of writing whole code from scratchPlease let me know if you are having any more questionBTW Thanks for your concern +Dup of. +Removed and from Roo +Jeffrey if you spot any problems please reopen or create a new issue. +Where should this be documented +Because doesnt have too many hooks for extension creating a custom to handle Twitters way would involve duplicating much of the existing class just to use a custom implementation of +to An important feature provided by a managed environment like EJB is transaction timeout which is never provided in code which should read better +They worked for me in but not in tc Server type was already specified for element forminput +Date formatting and parsing also tend to be surprisingly expensiveSo if we support passing NOW as a date string it would be nice to also support standard milliseconds +Thanks for testing that Todd! Assumption is the mother of allThe code in question apparently never worked for years! I dont have a ACF test environment so couldnt test it +request to fix in the hudson sonar plugin uses maven which means sonarsonar fails in my CI environment +This can be resolved for now +Moved from to just after merge of old branch into TRUNK +Will commit if tests pass +I took a look and didnt have the same level as Components in Camel +That way checksums will not be used just for HDFSHDFS but also for SS etc +patch appliedthank you +Lower priority to keep list of date +Quick question +Thank youI think we should also backport to +sharetestdata was just where I placed the file already attached to this issue +bulk defer of items from to. +attached a patch to support parsing of rotated patch +You can actually just deal with the situation you described by making the thread pool larger which is all that the QE team should have done +A workaround is to configure the with a single default queue +When I get a chance ill do more testing with GC logging and MSLAB enabled +Im not sure if that was Pauls original issue but the groovy script has been rerun on both and bit without issue. +I want to see the NPE and its stack trace +In trunk the feature is enabled by default but in is enabled by a context param +the task is done only partially for needs to be moved to the next release +Great progressOnce the patch is in place and the UI test for the authority runs successfully this can be merged into trunk +Do you have already some idea about how to implement for example the ipod example from the javaranch threadTo boost the object at indexing time you could want to give all accessories a lower score than the core products you would need to store some flagBut this is indexing time boost not sure this is what you meant +Now that we can reproduceit I think we should fix thisRahul +Since I made the mapperClass attribute optional the adapter registry needs to not log an error when the class is not found +There are different control rows depending on whether it is a leaf or branch +Hi PaulI just reviewed your document its good to see your document explains the wholeSo I copy your whole document intoThanks for you contribution +Got applied at some point as Ian pointed out so marking resolved. +Moving to since this requires a cfgdocumentation new eviction configuration element in. +should verify the issues case +It helps me to understand yourthinking and describing it as you did will also help future students of the code whoare trying to understand its behavior +Ive updated the wiki page to take into account Dmitrys suggestion on using Typed Bytes +is the patch with OFFSET removed as suggested +Strong should we add oracle to the release notes next to H +The pull request does not address all of the problems +Would it be worth adding that Option should be prefered in an effort to push the code to more idiomatic Scala +I wont have time to spend testing this patch more before tomorrow evening the point +Net plugin Or launching Sonar analysis with java +Verified that can be applied to the latest branchVerified that the DB Pool Wizard now allows you to select both DB jarfiles +One other problem with the direct link is that it is not easy to build the complete docs if you want to change both sections but I doubt thats a biggie +has its user +Sounds like content as well +Should contain everything from xhtml and set templateDir to themes +okay I think Ive addressed it all will leave this open until im I think its good that misc issue was actually in the last release so I think I made the right move with the rename +Maybe debug or trace logging could help? I can enable it and restart cluster to reproduce the problem +Martin please fix the test case +Surely I will do that + +bq +In that case we can consider this module migrated tooCheers +HDFS return the hosts for the file blocks whereas a file status type from a or local file system would return a single host all by the same interface. +I didnt realize that the calls to this method are generated +To get going just add all the libraries from the s lib to this projects libNext step plugging in Hadoop +Thnx Rod +Oops +You might just see what they need and make sure they can still do what they need with any new version +Thanks Adrian +Yep thats how I found it in the first place +It does not mean that the changes are unacceptable +Assuming the actual images are the same then theyre to commit from me as well +Thanks to Zdenek and Ernie for preparing the fix +Whats your usecase? It might be better to detect the encoding automatically rather than having to specifya complex eclipse plugin configuration section in each pom +Was going to suggest volatile instead of but looks like the get is fundamental so withdraw this nit +Do you envision using mixed configuration where some depends are in the list and others are specified in the plugin conf? In general I agree that it would be more usefull both ways +This is the point where the stream is closed and the data is lost +We went from Our workaround is to disable the to +There are no functional code changes. +can you tell me the server ids of each of the different serversalso in the data directories the accepted and current epochs are all b but your zxids all have epochs way less than that +Since just assigned to will not make change to +The issue is now fixed on RHEL thanks +This work is complete +Certainly you cut on a one line method +Well I tried to run from and the result is the same +However I also tried with the age field which is an integer type with no luck +IMHO I see great benefit for adding this please consider +Yep just ran your tests and it does look like some of the socket changes made to MINA has resolved this issue +Thanks Enis +Remove old related +Thats the point Vasilios! The application that the team I work had developed didnt pass on the penetration test just because of it +Thanks Juergen +Already done +Take a look at +Please give it a try with a deployed great thanks! +Ben Im not advocating modifying the listeners for either retry template or repeat template +Perhaps this could be considered the object store supporting the JPA JDO and any kind of data source +Maybe these alternate views can be part of the plugin architecture +I changed it from Improvement to Bug this morning +Opening a new reader should actually be slightly faster due to fewer files to open +Note that this does not affect the wcf component its solutionproject files must remain since currently doesnt have the support required to generate the C +One could do this by adding the filter as a required clause to a in and see whether filtered queries are faster with that implementation +Workaround provided sees comments +And in and also in the docs +Null row being passed to Get +So if no objections with the patch itself I will apply and close issue +Support intercommunication bewteen widget using publishsubscribe pattern +Im doing the same thing with and Ive used struts and now from and Im failing to find the exception as well. +Cheers all issues created in and before which are marked resolved +can you outline precisely what youd like to do and why? Then we can try to map it into usable Axis functionality from thereThanksGlen +Marking closed. +We could think about a proper approach for that in enterprise quickstarts have been rolled back to their working versionThey canwill be updated separately from JBDS lifecycle +Created utility method and used to process excludes for both SSL and non SSL ports +Verified manual test case +Patch for adding documentation about server side pseudo classes and fixing the bug +maven example which can handle along with jruby in the build classpath +use ant in the installer and build deploy from the sources of these demos in the seam directory in EAP over to example is still failing to deploy in +Hello Morettothere are some things to ask and to tell about this problem +This leaves the project naked and no longer producing a jar and so shade is flummoxed +as wellThere is some logic in that prepares a Map with the data those template engines can access +Please confirm the current status of this issue and if its incorrect change it as required +I think its not +Cant we simply remove the from the library moduleBTW I still find the current structure better than what you propose but this is just my opinion +Do you have some concrete examples for what youd like to achieve here? What kind of dependencies such a transformer object typically has what state it might operate on etcm working on that +Prasanna can you confirm if this is not seen for was only on master +In this mode one would expect that inserts with autocommit off and on would be in the same range as syncs are not happening but the difference here seems to be more than the approximate estimated range +bq +Thanks John! +on Linux bit and Linux attached file with error screenshot +If no algorithm name is specified we could fall back to the algorithm to stay compatible with old databases +The last statement in the script failsit is the original statement which disclosed the original problem +Previously we used to return the firstlast value in the window if the evaluation of laglead amounted to an index that crossed the boundary +Like I said on the ML in that case you would have to do a read that waits for all replicas +Unfortunately things came up +Corrected script touch before adding it +Im not sure I really like +Triaged for +Canceling the patch to address Erics request +This feature is done. +I wonder if the problem shouldnt be caught by in +Thanks for the reviewbq +I just committed this +I only want you to use the default hashcode and not identity Hash Code because it is the same of hashcode stays not implemented +Also both of the truetype calls occur in separate methods +scheduling for the commons issue has been released in +I cannot reproduce the problem +Ugly but works +No problemThats a good idea + +Weinre benefit to you it is not closing it reloads after refresh or url changing +A has its own trunk branches and tags in subversion and releases separately +The other is to explore JMX +style was fixed in revision may I close this bug now +See attached +Any other opinions +The messages currently stores all the messages across all locales +sar files archive going? For example if you have a plain old java project youre filling with +Updated Framework Container and Native to use this new version. +Should not be difference as we use Dalis validators almost everywhere. +Closing issue. +My new plan is to keep the copy but do it in instead of in Store +It should only contain jar heres what I suggesttestdata where generates the JAR for the adapter classes and generates the RAR and depends on you think you could modify your patch to reflect thisLet me know if you need any help +Think about this examplepublic class extends extends Is that valid? of course not because the navigation handler needs to be registered in to make it available +now fails +Yes we will include this fix and for ews +Cancelling patch to provide same with +thanks +In the README update REQUIREMENTS section for Fedora to include libtool and pcre +The model project this with JBDS CR kit and it is not reproducible +So the issue appears to be during dynamic compilation +If Teiid were to reject it with a message indicating it has no visible models that would be great +Yup + +But then I didnt write the code originally so perhaps I misunderstand itWe added a test case that causes infinite recursion with the old code but appears to work with the patch +The caller may forget to set it and then lead to a bugHow about the other parameters? How to determine which parameters to set before passing? I do feel that the new class makes it unnecessary harder to use the methods since the compiler wont be able to help checking the parameter lists +Thanks +thanks for the test case but we cant use it and probably shouldnt look at it? Can I have you your Entities the test case and check the Grant license to ASF for inclusion in ASF works box when you upload +with day TTL and in table +I added fix in rev in Qpid trunkModified the to maintain the timeout value and set it to any XID in the start method if the time value Also the XID nulled after commit rollback and forget to prevent a timeout being set on an invalid xidThe setTimeout method will only set the timeout if xid is not null and if the timeout value is different from the previous valueModified the test cases in to adhere to the correct behaviour and also added a new test case to cover the issue mentioned in the JIRA +Checking this in +The way the works is frankly stupid +Assume closed as resolved and released +The Cloud Foundry has been excluded from the distribution in commit James it is a good workaround but I think it doesnt solve the issue is an classloading issue imho there are at least ways to solve it redessign to inform user the API should be installed before be able to use the commands Create an bundle activator for this to check the availability of the cloud APII think feature is really good so please the issue to solve it +Thanks in advance you for the review DevinI have pushed the revision to our master and I am closing this issue. +Also please run your altered tests against the profile mvn +validation file that has rules in preferred order of error aresrctestcomopensymphonyxwork srctestcomopensymphonyxwork srcjavacomopensymphonyxwork srctestcomopensymphonyxworkvalidator +Should figure this out before npm update +That patch also requires you first apply +RC does generate commands in blue correctlyInitial spaces in the code lines however are not displayed because the browser collapses all spaces to a single space +Having the same tried everything I can think of no JBOSS IDE for eclipse is not loadingAn error has occurred when creating this preference page +Hi allThis is a first version of this issue +Updated +The problem introduced by is that the lob mapping keeps BlobClob objects alive that earlier would be gced when it was no longer referred by the user +This Zip file contains a set of the +Im not sure if it can be done from ant so either you use maven or you do it by hand but in any case the files have to be in the right place not like last time +Attached a patch for trunk +Hello Cezar Thanks for your reply we found out the issue it was not an issue with XML beans issue was with our ehcache implementation our flushing mechanism was not working and this caused those objects to be in memory for ever +I just committed this to both trunk and. +The newly added test case fails when running as part of unit test set +Resolve the parent issue to remove clutter. +JoydeepYoure definitely correctBut in general there are several problems with the system from a software engineering point of view Naming conventions dont exist +Sorry Igor my badIm developing a way to provide somehow components + +BTW this fix will not be to the stream of +Therefore I introduce an period +Not sure why Solr does it on +Updated reviewboard patch has most of the review comments taken care of i think +All DTC fixes in one is neither a new feature nor minor +Its not a bug in PHP +does this mean we can specify uuids in form or we have to use u notation +Commited to history r +What was the problem and how was it fixed? When will be a fix available? +My last issue now is that when a good connection was opened but not necessarily used right away and the connection was forced to be abandoned in between I get another new error +See sample code +Pushed to with JBT. +At the moment I have no more ideas how to improve the performance with the current API +To compile the program I had to remove the libraries Bernard included and add the followingFor just the Glassfish library that is added when you associate the app with your application server in and JDK +But at the moment Ive checked in good first stab at getting this issue resolved +close issues for. +The main difference is basically who is responsible for setting up the CAS +Thanks for fixing issue +Improvement of previous patch +I might just write a simple extended operation that returns immediately but sends a notice of disconnect to the client before shuting down the cients session in some number of milliseconds +Thanks jing for the rebase and addressing umas comment +Look into the and directories and you will see a directory named home +Found and fixed the issue +Duplicate of +It would help to narrow down the issue to one of these two sides. +Updated the base zip to include the endorsed versions added the endorsed config to and enabled the EBWS SOAP tests +Affects both and modules +It needs changes to the execution codepaths in +I am on you move up to r all these including case should have testcase that do similar things which are working ok so i guess there must be something special about these contributions + +but until cassandra ups their versionI dont think this is actually true +Perhaps we could get the favor of a releaseIts classes it looks like and its Apache licensed and there is no release cant we simply suck this into our code base with a readme and JIRA about removing it and switching to a release when one occurs? Get the whole dependency thing right out of Mavens claws +This does not include significant changes from the last patch except that it has been integrated into directly instead of placed in a contrib directoryI have included the ee tests in this patch but they will not yet work as they need updated to deal with changed paths and the fact that I removed some sample jars they depended on +Should I open another jira issue as David might be working on and submit a patch to this jira issue? Kindly suggest +Verified locally that the new test passedIntegrated to againLets see if the builds pass. +For API usage this will be trivial because the client which calls the API can read such kind of a file and pass the comma seeprated values +There is a documented test in +If we do something like this and Im not sure its something we want to encourage +xsd file served up successfully to the browserhellohellowsdlresource +See attached +Reopened to review if the fix should be committed in Beta +As to whether folks are using filters the answer must be no +What would the advantage be for having tooling depend on all of Forge for implementing what for now looks like a simple project template +However above the px mark the image width to height ratio must be greater than +Issue updated to reflect the actual problem +This is as you would expectWhen you click release then click either change button events are still received +just committed this +It seems from the discussion we concluded that if we are going to make use of multi we can enforce as minimum runtime dependency but as it is is not needed +One minor nit Can we avoid involving in get? It probably suffice to create a and create a on top of it +Enough testing has been done for now +Got you +best news is that sigma is now responding and if I run I can see that the guard lets the remote style test body run and the test succeeds +I know +Im going to look into this +Looks like it is wtp bug +The rd example is incorrect behavior +to generate the patch Im on linux but I never used windows svn client tools to do itRegardsFreeman +Resolving as incomplete for now. +This was disabled from the POM +Patch committed +fix is confirmed +the original services listing works as expected and can be used with the example or the example shell script in the comments +What incompatibilities did the redesign introduce? At the least we should document them +If there are any remaining issues or new issues please open a jira against. +personally think that we should get rid of the hbase +While the API of suggests that there is more than a single provider the implementation uses exactly one which is also the manager implementation +The network sever part of this bug is fixedThe issue of the having null arguments has been moved to +Dont the Apache Zookeeper project want to be compliant? Seems like a fairly easy fix in your ANT build script +there are projects for em +It amends the strip test to check that length is positive +If you have an immediate urgent requirement for this I can adjust my priority list to do this sooner +The patch has been checked in +Closed as per state in Bugzilla. +Something must have changed after that +Attaching patch of corrections to srcgen tests so that they pass XML comparisons that were not done this is adding correct xmlns declarations to the root elements on the input files +This directory could then be added to the project properties classpath and these classes would be available for use by the Java classes in the same would at least allow one separate output directory for the Groovy +map the virtual server in to the alias found in the +We dont advertise support for security to work on anything but Linux +The patch corrects the problem +you can reopen and attach the patch to this one +Deployed and live +Then after the build process completed I checked the target folder and it contained non only the test classes but also the classes from srcmain +The size configuration is still Do you want to change it now since not many people are using it yet? +batch transition resolvedfixed bugs to closedfixed +Therefore this could well be a Solrj bug +I know this patch works because of the yarn test I ran +See also fix the two at the same time +Nobody else interested in this feature? No update since +That could indeed be the case +That is wrong it has to be marked for Seam RajeshAll databases and JDBC connectors are listed in the New Features General section of the Release NotesCheersJared +I have one more question If I create Seam web project wo setted Axis runtime I got an exception and project isnt created successfully +We cant break the standard usage for this use case +For adaptors mapped to just Iterable having the simple Integer s default converter provided and not needed to be added explicitly +DevenPatch applied at r please verify it was applied as you expected +I only know the corresponding internal address +Added Christian to the watch list +Will someone please review and commit this patch? Its a super minor issue but super annoying +Right now its equivalent but later +Applied patch in version +The attached contains a with the code changes and a with test changes +Yeah it probably would be a good change except it will be irrelevant with the new stuff +No problems reported regarding the fix. +Committed a fix for this issue +Attaching patch to to address concurrency problem in constructing date headers +Good catch John +In some cases the crash happens at second try +What would you proposeOne possibility would be treating this somewhat like Category B? +Sure lets do that +Such a cleanup approach will hide those bugsI personally would not optimize tests for the time that they failI also think that youve found an issue in the clean database decorator and as such it would be good to fix it centrally there rather than in a single test +If you experience a problem then you mighteither rebuild struts yourself with the jars included from the release or simply replace the nightly versions of thecommons with the version of the jars + +The only reason I can think of for doing it in is to actually have the generation be complete but just generate only the files that pass the you considered how the ant taskseclipse ui will be able to use this +Committed +If this is by design then Boo needs to emit the proper error message instead of trying to automatically specify the default constructor +This is done for all setups since Ive run into this sort of template thing elsewhere as well. +Is that OK by you Todd +Then this was fixed in according to the component matrix +Can we even do this in without an change +Another use case is creating annotation proxies for XML and programmatically defined constraints +I like the additional attribute lets see if anyone else chimes over the next hours or so +Ill take a quick look and will either commit or provide comments +Another complication is other Apache Java projects which have patched jars in their dependencies +Im having the same problem in +Without it it locks in the tearDown step +Hi thereIve been using some simple scripts to manage dependencies in my grails project so when I saw this ticket I thought Id share! is a very simple Gant script that just checks for the presence of and in the project root folder +What more needs to be done before the changes are moved into the branch +Ill commit it then +Ill silence these warnings in my next patch but Im not sure if Hadoop QA wont complain about that +I would not remove seamfp patch seamfp tests from tests if we would like to test backward compatibility and test our new code with old versions of seam an EAP +Try the version. +Yep +Moved to this version of FDSOCK to see if it can deadlock +I am attaching a patch that includes the required checkstyle changes +Im planning on basically taking sauls ideas and generalizing them a wee Ill be contacting you when i get to it +Hi KojiIn the filelucenehighlightersrctestorgapachelucenesearchvectorhighlightI have added a TODO regarding empty spaces being addedIs this something we should deal with? Thanks +The raw iterator shouldnt have to allocate a and a per entry should it +Added beforewrap again +Patch integrated to trunk. +Patched the LICENSE and the NOTICE files +Notice how the contents of the auto completion list change depending on whether it is the initial pop up or a later auto completion popup +Attached patch didnt apply to trunk +The updated files areFilename Titlerreffunctan TAN functionrreffuncceil CEIL nad CEILING functionsThe patch is called The file is called +Committed rev I also added randomly calling it fromThanks Mike for the review! +If someone wants to do something for lucene users fine +We could change our to declare the newer which we are allowed to with our entity resolverI have tested that locally and it works with the embedded Jetty and with an oldTomcatAlternatively though not as good we could do a trick with the entity resolverso that it delivers the when it is asked for the I am ready to commit the former solution +Given it is unstable I say just go ahead with the plan to make it private and have the deprecated getter there is a common theme of returning the object return this on all setters +This is a server issue +I checked deeply the issue well it turned out that here is no issueYou must have created the properties file by hand hence some parameters that control the reader were missing +All effort goes into jBPM +Can anyone confirm this issueHiin order to confirm please attach a short test case +no commit for branch +Reopening so that missing line can be added to the messaging bisocket connector config +No test case provided and I cant reproduce any leaks using the information given. +There is something wrong in how things works here the is useless as its obtained from the build environment where Hadoop was built see +this is a Shark libraries of version +Also dropped from dependency +To require the manual creation for usages sake would be acceptable but such a case should be detected by the and reported via Exception to make the behavior more predictable +To enable that select Quote SQL Identifiers for a given in the Modeler +Thanks Regis +The comment in the file is correct but the displayed text is incorrect +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues updated This updated patch also outputs the pubDate for the if the field pubDate or timestamp is indexed in solr. +Its OK with me if you check in this change with the current algorithm +Now that this has landed in tree I think it would be good to start splitting further work into different Jiras like we do for the rest of the code base +Therefore Ive created and to document these concerns for the Facebook Twitter and modules +Here are the steps to see the problem + +Could someone add me there +Fix and where this broke the code that cleared out the old cached +That should not be the case right if the is inside quotes it should not be taken as a named parameter stale resolved issues +I think to properly address this issue the current implementation for CREATE TRIGGER needs to handle null compilation schema +I would still vote to leave the flag in place for Object types in case a consumer of the response finds it helpful to look at the property +Global ecas are a bit special +The commented out lines should be enabled after auto casting is implemented +Lets do the matching fix. +Verified with latest build +sense patch against limit to years exception contains requested vs max limit added to I would have put MAXTTL in rather thanBut in any case +I can try but I am not sure I have the time at the moment +Therefore it is sufficient to add a simple if statement to fix this issue +See new comment in +Lets wait for a bug reportIll try to fill in some more explicit dependencies +Remaining features are not a priority +We stil have this problem with upgraded Axis +thanks nicholas. +Have you tried this solution yourself? The parseHtml flag does not affect escaping it is only used for validating that the HTML is +Not a bug end user issue +Could you please remove each of the bundles under the groupId and then my newly created bundles back to the repo +See for a discussion about the versions of domj and their issues. +Thanks. +Attached it the updated version taking into consideration all your comments +It is redundant with the written in the jar +Done +This was included in but not shipped as part of the release. +Snjezana you had thoughts about this one +checkcheck is one of the key features that differentiates from other key values stores its a shame that these are only available to Java clients since they are missing from both the Thrift and REST +Allows for proper mapping of a to an document type schema typeAlso added new for document schema types +At some point I may get around to it as part of the effort to have fully integrated with AS +Hope thats OK + +Upgraded to version of Jackrabbit +Removing so the jdbc works correctly on windows +It can even lead to an unstoppable job in the case that the job flow contains a cycle and the cycle is already complete when the execution starts +Fixed to log the blacklist additionsremovals in FifoFairCapacity schedulers +Graham can you please attach all relevant files +on current patches +FYI Were planning on implementing this feature via a delegating translator and +Im now thinking a properties based mechanism is most appropriate here +The quinit screenshot is there right above the application in updated the clean comment and other other I think we are ok with + +Great performance results show bug jumps in jacks performance and modest improvements for lusearch. +Fernando email sent to you and PM list +Thank you Mark! +I was not able to create patch against head revision so I created the patch against source codes downloaded as zip archive +test seem to be ok so far closing the ticket +We might just end up with a different pdf generator dependency though +Sujested patch no think this must be bug not wish because now with Jira Changes plugin is not working any more. +I saw your post and actually bugzilla told me collision detectedSomeone else has made changes to this bug at the same time you were trying toThe changes made werepretty cool huhanyways it seems like that for some cases your are using probabilitieswhile in the example I mentioned its not calculated this way why there seemsto be two ways to calculate this? is it for performance concern? personally Idrather get the accurate answer than saving time and Im currently doing theslow looping thing to get the right answerthanks +Thank you for the patch Raja. +Please include it into tomahawk +My thinking is get it working first then extend the exposure of symlinks in fsck results in a later JIRA if desired +Trying to do a file scan failed with the following errorLast Error matcher ignoring dir QSo maybe using subst isnt a good enough test? Will try to get someone at with multiple real partitions to test this +When half the session timeout goes by without getting a response from the server its connected to it prematurely disconnects from that server and tries to connect to a new one +So merely requires application developers to implement another interface and then invoke a setter on Application without freeing them from the necessity of subclassing Application +Problem was caused by resolving type systems imported by name + +Appears to be a bug with Maven +Thanks for report MarkMaybe a reason to begin to turn to Java +I forgot to give the following additional info I was initially assuming that this is indeed caused by an issue in +It would be great to handle these situations but Im scared to even think of how +The Java Principal does not have an accessor to get the short principal +bq +Assigning to Steve for review +can we include the same plugin more than once in a feature but with different platform filtersGood idea if it will work +Indeed I am down with this syntax +The description of the JIRA and the patch should be updated thenbq +We promised it in the release time frame hoping that would be enough to salvage the deal Rich you need to schedule this so we can get it into the EA +Marking as Intro +Thanks works fine with this acocrding to the reporter +added folder into bundle jar tested on local on JBT. +Patch to and for reporting +The code that was failing before now passes its unit tests with this patch +Markup attributes are now unescaped while parsing the markup +I think this misses the primary meaning of cabal in English which is a group of people rather than a thing +Problem is not in request from ATS to origin server but in response from ATS to client +So the problem is that the extra search match is being renamed in the first stage and then a second time in the second stage +To avoid the confusion two alternatives way I could image are +Nice work Sid! +I tried to find another solution to avoid the upgrade of the ant dependencyNow VFS will shutdown if a target finished and a refcounter reached This refcount is incrementeddecremented during It works with your test cases though could you please check it out tooThanks +John does this include emails we are sending out forNormal case case emailsnotification software emailsLee brought this up so adding to Jira for clarification + +Thanks for checking +Refactoring the proxy fixed the do understand that there will be one polling loop per listener but it is great to know about the JCA option +Just merged thanks a lot! +Im not actually sure if this is a proxy standard +Hmm +Further work but no patch yet because it requires a bunch of test tweaks I have split timeDecay into evictDecay and warmDecay +There is probably days of work left +The patch looks good but needs a rebase against master +Patch provided by Patrick checked in +Because of this easy workaround I also reduced the priority +The real name of setValue is saveValue +Actually the plan has been to maintain two branches for the Remoting releases +function well with sequential writes sequential reads and random reads so if this issues is fixed they should work well with Lucene +The world is more complicated than that so if you need something else theres no option but to use implicit grants +provides important bugfixes over the patch for the branch +IMO the problem is that the itests try to use which is definitely not available yetProposed fix committed to branches in rev +My goal for the project is to be useful to developers and I suspect that many developers are going to become TezSpark developers in the coming years +This is already a big No No for me +can be closed without consider it +modified file listSearch SearchSearchSearch +Whether you think that its not real problem please close the jira +This is no longer a valid thang +I dont think it will be ever possible in a web environment to guess the absolute URL +Thanks again Pradeep +Or should I be looking to a different place? +i was going to look at it dont you have enough others to look at Given the limited testing the generics area has had I would not be surprised if it is easy to recreate without source +The attached patch adds support for nested structsThis patch also fixes a bug affecting the schemaType of the schema element corresponding to an IDL operation parameterIf a struct definition followed an operation that had a struct as a parameter the schema element corresponding to the struct parameter was erroneously being assigned the wrong schema type + The following authentication are configurable with the JAAS modules +I actually do think these bugs are related because it is related to the forking and my case is causing a different Exception +Do you plan a release or sothanks your fix is just fine + +The skipped range is divided into two halves and only one half get executed +However its the correct behaviour +Hello Dibyendu +With ERcxf if uddi lookupregistration is set to false in the modified dynpartner example worksSo will create a separate issue for the ERnative port not found problem +Is any one working with this ticket +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +NB JIRA interpreted XML comments somehow +Patch looks good want to have a bit more clear ideas on dep stuff +Looking good +Hodring does not create the dfs directory though it supplies the name to the jobtracker +I was able to pass the testcases related to and +This is fixed although the solution is not very elegant +JacopoThis souds like a rather complicate solution to a so tiny problem to me +Removed the reference to the metadata table functions since this problem affects table functions too +Unfortunately Im uninvolved with Grails at this time and cant help you further. + +Perhaps we can add links to the scripts sources in subversion as detailed documentation +That might help clarify to users how we name our +Thanks Enis for taking care of this +Do you want to take another look at the patch +The draft of the is attached +patch looks good +No Tests included +Thanks again JimWe need to remove the offset code the right way to do offset in Cassandra is with the start key +A would fit better +what is this patch for in this JIRA? Its been closed months ago +Thanks for the report! This was an unfortunate side effect of the new ability to let a custom editor convert a null input into some other +Rather it should like fqs default to the lucene query parserSounds reasonable for fq and bq to be parsed the same way but it seems unrelated to this specific JiraAnd this also begs the question of why bq shouldnt simply be a global query parameter like fq +The parsers are still public for total control on the conversion +If Im wrong please. +This will help the common case of a new user whose datanodes failed to start and is confused why he cant write blocks +Added attachment fix +Extract this maven project and execute mvn clean verify to show how works okay +New issue created to track adding autocomplete functionality to the richselect component +Is there anyway I can fine tune thisThanksVu +On Windows default STL allocator just delegates to system mallocnew +Other than that straightforward test +Would it be possible for you to see if the same problem occurs when you are using version Hadoop +Resolved on and moved in another ticket the part related to other components. +Committed to trunk and +I believe the javadoc warnings are from which added goal and phase tags that are generating warnings during javadoc generation +Ill need to go get my sackcloth outfit first so give me a bit of a headsup. +Added patch for of vof problem and description tagBy the way we still have outputPath left as is in +Fixed in r added method for deleting the artifacts in the db when a repo is deleted also deleted repo scan statistics when the repo is deleted created DAO that does retrieving deleting and adding of to the db added and updated tests +Fair enough +I have spent a lot of time with this patch cleaning up complete changes formats missing dependencies and all kinds of problemsIt is a really bad idea to put together such massive amount of up patch and updated to current svn some reason the xml files show up as completely changed in the patch although only the first line looks like the incorrect URL resolver is used +test only issue will be included in EAP. +Revised patch in sync with trunk that addresses more files in srctest +Do not commit until smoke tests JIT team may have already fully implemented vmmagic support in the JIT +Patch is in CVS +Patlateg I already committed tis issue to both trunk and branch +we can try writing a test case +maybe we should use description instead of comment in a tableGood work +I had to invoke on the value for the substitution to take place +moving from svn to git +Thanks Jing +I would actually prefer to see that moved to an external or something like that +It would tie current trunk code to release functionality +Is this a duplicate of +Requesting thoughts from other users who having similar usecases +The goals are fixed by removing the aggregator annotation from the class documentation +Adjust the logic in your app so that authenticated users end up at a standard welcome page in which you can standardize the loading behaviour so that a single session and security context are always established before proceeding to parts ofthe app which may submit multiple concurrent requests +JBDS product poms already updated along with JBTs build poms +I would rather vote for or +Bengt what FTP server do you using in production? And what OS are you using +Im happy to expand the check to all versions of gcc +Ive committed this to trunk and . +Slightly updated patch as I missed updating a javadoc that was throwing and now throws +Maven is running on a VM which is a Windows application that does not know anything about Cygwin paths +you woul dhave to build the jar plugin from source +This will prevent this problem +If your serialization format produces a dump for a particular object that happens to include a character you cant use that object as a key or column name in any of the methods that expect a concatenated string where has special meaningAll the values need to support the full range of possible just byte arraysMaybe I am missing something +I am not familiar with this one +Patch committed. +We do not want to resend an indexing request especially one that failedThis seems to come from some kind of protocol error +Im finally back form my holiday +Secondly it needs more tests +Here is a patch which simplifies the probe for integer literal syntaxI dont know that it will help +Ive looked in the release notes for recent patches but theres nothing obviously in that direction +Builds local and passes all tests +This should be fixed now +try to solve the problem +A came about this because I updated an attribute and got an exception +Best regards by Andrew. +I would like to revive this patch? Thougths +Still waiting for a sample to work on +Should be good to go +Fixed in r r r +I am using the same eclipse with same set of projects for for a long time OOM never memory settings are the Eclipse defaults for Mac i did not modify themmI believe that OOM is somehow related to the bug that i saw +Alex if it solves your problem could you please close this issueThanks a lot Jason for the troubleshooting information +in M +Im working on new patch +Justin Do you have any tests within dataGrid or any links to code where it is in use? Im assuming it is being called via hasMetadata get get? and AMF encoding uses this with the include option may also fix some subtle bugs in those +Updated patch incorporating Olgas comments +We really like this version +The regression tests are in lang and in derbynet and there is also a change to util to support the new lang testsI think these changes are now fully ready for review +In current implementation it would throw if the missing but it couldnt detect the configuration file has corrupted or has being truncated +Each UIMA AS component that needs serializationdeserialization creates its owns instance of +Moving all issues to the CR +thanks peter +Yes thats great +So its still nice to do but doesnt solve the implicit iteration problem +It will be upgraded once the distribution maven plugin will be available +What problem did you see with this? +To test it change any of the test case to throw +Files in this tar arenamenoder mmao visible rw mmao visible r mmao visible r mmao visible r mmao visible r mmao visible r mmao visible r mmao visible r mmao visible r mmao visible r mmao visible every log is edited to be centered right around the start of the phenomenon +Last time we discussed this I was told the default is automatically calculated by WTP as latests versions of all facets that it remember as if there were a reason to us having them by default but as long as we have a Seam and Seam named config it should be fine to keep the default to a minimum +From the description when is resolved it looks like this issue becomes a +Done. +Then this JIRA should be a bug +I have updated to have a new way of getting round the problem which is still there +From this point one can assume that its available +Im happy to have merged back into the core now that theres a buildr committer that uses IDEA +How did you get like thatMartin Thats true +Heres the resulting file which contains packages like and from a provided dependency. +Ive deleted previous patches +just need to add k to the hive grammar optionsI tried removing backtracking altogether but that is difficult due to some left recursion in the grammar and it isnt clear how much more benefit it would give +Otherwise the values arent going to be substituted +pending jenkins kicking off jenkins manually +Eric can you comment on why the kill isnt killing the process? Didnt think we were masking that signal +Yes it is definitely experimental +This could be a joining bug +we had so much problems after upgrading to we thought our addded features was the main culpit +From the exception pasted there is a hint that we cannot deploy a vm with ghz cpu Failed to power on VM +I tried a couple of queries in ij just to see how it worked +Thx. +I agree w your analysis that this is necessaryNit The st writeFully method javadoc says and uses fc but the arg is a +Perhaps fsck too will fail if used from a machine other than the s one +Can someone please review my last commit there to make sure what I did was correct? This is to get the builds going again +Hamlet is it correct that you are really wanting this only on properties +Youre probably hitting this yes. +Latest snapshot looks good with my tests +However same results +But if it is set by the user and it contains spaces +Could you give me an answer to my issue? only highlights the fields with the errors +DenisTruncated chunk issue can be reliably detected and recovered from only if the end of message is delimited by the end of stream condition +Opened covering this. +confirming default parameters are well presented on the UI for +The same happens on Linux it seems that java used to run Eclipse is selected in servers launcher not the workspace default java +Standard Roo code doesnt generate a aspect +I did not take into account that the for the package is already created in Ctor +Affects +Anyone else still having a problem +As long as the is created using the right configuration object the job should be picking them up +Login is basic for a professional web page +Thanks for your help guys. +This will launch tests with no launch all tests in project just set your configuration to run all tests like it is shown on the third in that case will run all tests he will find in project even those which were excluded in maven. +Ooops yeah crud +The can than just collect global ordinals in a impl and the can just iterate from this. +I guess all I meant is that since people tends to get touchy when you limit the thrift interface doing the minimum validation so as not to crash CQL could be an option +Attaching the mv script and corresponding patch for jenkins +Commited patch to resolve defect +I apologize for the oversight on my part +Dan is it fixed by one of your issue +Bulk move issues to and +are you sure it was the latter +Forcefully killing the network server is the only way to recoverNote that there is a background job in the system that is running an UPDATESTATISTICS every minute as in Derby the automatic statistics update daemon has problems and cannot be used +Test cases included +Any of those approaches you mentioned will work will be released in need to test customized filters document but thats not strictly part of this task +NadavI suggest to keep Lucenes because it is a very central and highly optimized part of Liucene +Good job +The correct information for installing Java via RHN has now been included and the difference in steps of installing on a Red Hat Enterprise Linux machine using RHN over a generic distribution of Linux has been made to avoid confusion +The Ruby clients wait but the Java clients interleave their calls +the same screeshot as in the same screeshot as in the to reflect issue +Patch and unit test +Making sure that the tests cover all the cases is important +The source jar validation has been turned off indefinitely because there are too many exceptions to this rule +Thanks for reviewing Sijie. +I just noticed that in the XML Data Import Readers screen entries fields are a bit too much at left but this does not come from your changes but from the inner ftl file +If nobody objects Ill close this issue in a few days +great Charles! I can confirm that autotest is working. +I will experiment with some typedefs to make the code build with to see if it can be done +The test cases however run against the yahoo endpoint so Ive excluded these test cases from the build for the time +That isvalidutf function is definitely not at all near correct +This approach doesnt rely on stdin being read by the child procI could see a sudden Erlang VM death leading to an OS process being orphaned +If you use do you see any numbers or they zero there too +Would it be possible for you to submit your ideas as a patch which we can test within the community +Changes made plus some other polish am moving all the xpath reading to solrconfig +Thanks for the work on this Babak +Yes I know but thats how it was designed +VM is ready to get a project admin +As Cho indicated seeing a bunch of skips indicates Ive probably got a problem with the configuration of my test system versus a bunch of failures indicating theres a problem with my product or testsAnd as long as Im making feature requests in the skip department Id like to also add a request for a Skipped option in the Filter Results dropdown on the analysis page or some way to again separate the skips from the failures +I think its important to add comment on wiki page to say this page have been migrateTo be sure to not forget Just a sentence on top of page + Username carlschmidt +Please grab the latest SNAPSHOT and confirm on your side just wait till Jenkins will comment on this issue then the build will be available +As for mysql we can make this happen +If I understand correctly we need to add one more command by calling which will be opened one more VPE instance and placed horizontally with existing one +Max I can not reproduce it can you attach a screenshot +Doing it this way for a bit will inform what kind of macros or other context helpers we need to make the view as clean as possible +The patch also include the patch for converter message and EL of issue + In which case Id prefer the latter in order to avoid confusion with the other create methodsok +should be +I will come back with more information +That may be a case where you could dynamically change the memory channel setting and get away with it but more often you cant +Also changes to use the created object +Pull request merged. +the location of the +I will wait for a day or two and see if whatever mirror script they are running pulls over new changes +votes no news +the issue still exists reproduce +Looks good as such one question left isbq +Thanks very muchc +Hudson ed it as well resolving. +I tested this by building the project with it and it appears to work fine +Committed revision +KyrillIts definitely on the radar +Thanks Sriram! +Ill do the writeup and pretty benchmarked graphs and stuff tomorrow +We need to fix the documentation for this not the updated theThis fix was made in the trunk +Folks who are serious will use something like Kerberos which is well suited +private list created with olga and gates as mods +Looks good to me though we need to also find a long term solution of running junit in parallel in order to get out of the habit of bumping the timeout value +I just committed this to trunk and +Committed to and trunk +The Major item needing appends in is so we can handle crashed regional servers with out data loss +I have opened to address this issue +Please dont mark it fixed until its been committed into the SVN +Thanks reveiw I will fix all the comments +Actually tag has it in name +Store the number of files under construction rather than the number of leases +bulk close of resolved issues. +Yes i belive that its related on some parts +Cleaning up old issues +need to test as we did for trunk +Our style class output code could be cleaned up a lot +I forget how it is indexed off hand but you could work w that maybe +Yes thats what I guessed looking at the code +This is the test case for the +testsuite +So if a commit or optimize is running it will not evaluate for auto commit until they complete +Would you like me to check it in now +bulk move from M to to in +The container is broken if it requires application scope objects to implement Serializable +It seems there is no other place for doing this as well +Hey Paritosh why dont you take over this issue since you now have committer karma +Patch to attach javacompile as a testtest prereqsthis patch also solves issue was intoduced from the fix of +Marking this an incompatible change so that it gets attention +We now have a test which demonstrates the issue +Just tried this out and it works just fine +Specs have been enabled in specjavaintegrationcoercionspec for this logic +All tests pass even with +Updated patch that contains the method rename for +Incorporated into the release notes as per Rebeccas comment in the General Known Issues section of the Release Notes on stage docs +Cant reproduce on my other cant find it at the moment but I did see mention that HTTP BASIC auth was broken in one of these developer releases of chrome not sure if this may be related +Hythis was a bug in the JSF RI and worked in myfaces andIn the the expression is checked if it is literal and if it is the literal navigation outcome will be addedSo i think this issue was never a new feature in myfaces or i have missunderstood samethingRegardsMartin wrong inside of the impl details of the JSF RI not really a bug +I have tried to follow the work but there has not been much discussion on the list about the issues +The options used were the followingisolationcgroups cgroupssubsystemsfreezer cgroupshierarchyrootcgroup I just deactivated it since that host seems to be the only one affected +I then left this window openI then used the New Window function in the file menu of SQL WorkbenchJ to open a new window and connect to the same database separately +Hi MattWhat affects version does this relate toCheersJared +I found the same behavior but its the in Tomcat which tries to workaround a security hole that way +v removed the unnecessary block setting launcherconf to group name + is have checked the and it is not empty +Marc Right you are! In this case I would imagine its the giant unicode hash causing the failure +Please feel free to reopen the issue if needed. +Ive committed it to CVS +Dont forget to utilize new daemon management script that now provides + +Thanks for working on this but before you do start to make modifications please assign the issue to yourselfFor the changes themselves I dont agree with the creation of those many localized messages We have been trying to rationalize and reduce the number of those by removing duplicates and combining several ones to convey the full explanation of the problem +additionally you will need some logj configuration in your with the grant license +Andrew this is quite your turf +Patch committed to trunk +David could you confirm that with this ticket is useless +fix that one investigating on different order of graph creation scenarios now +I will have to take another look +Its such a small change that people can recompile Flume if they want something running against the old version +Lets just add them as javadocs to the interfaces +There is already a tool for that +Probably need to fix this before the next release +Im trying to compile but i really dont know about Eclipse is there an easy way to know the dependencies with other jboss tools modules +The following patch to allows garbage collection of event listenersHowever updating listener counting in is far less trivial than mere replacing the eventListener by aFirstWe need a that queues all we need to keep trace of +Some bugs will report in separate Jira tickets +Updated the trunk as well Updated version to to match the highest versioned +Alessio we had a similar issue with the interop tests caused by the JDK +there aint nothing to do if you are addressing you telling me you are actively doing nothing Ben +If you still need a way to do this with the new design please reopen with more details about what youre trying to achieve +You solution makes way more sense +now contains a mechanism allowing any plugin to have a history. +What sort of recommendations do you have as far as config changes when upgradingI am using the JDK so I would I be OK with the presence of JSP in the version? Solr is not directly reachable from the outside world it is used on the internal network by our web application +It will give you greater flexibility +Andrew you are totally right I misread the issue +that should be completely documented +Wed love to hear more feedback +There doesnt seem to be a good way to deal with the python lib path dynamically with the rpm plugin +The Reviews global service doesnt seem to support differential mode +Committed to trunk +Draining events immediately dont quite remember why I added the option for a delayed drainThe delayed drain was already there on trunk just not usedbq +Browsing through the file works for me. +Ill take a crack at updating the install script once the Solr build process produces sources and javadoc jars in addition to the current binary jarsI think it would be better to integrate maven deploy with the ant builds similar to how Lucene does itIm going to start making the changes +Assigned to the roadmap of Sonar Eclipse for evaluation +If I may comment it would be great to generate the latest version of the docs instead of every version and then give the option to generate a specific version of the docs if thats one of the requirements +The logging stuff was fixed a long long time ago +No problem the manual is kind of large +Resolved at r +fixed in +Before retrying could check the table state in zookeeper and abort if it is deleting +after running the tests +Fixed erroneous character encoding setting on the response in the class in Rev +In fact the datasource should be different in that are different +added some fixes some integration test cases and an example +Lets create separate to address thosemostly minor commentsInline Commentsqlsrcjavaorgapachehadoophiveqlexecshould add Override hereqlsrcjavaorgapachehadoophiveqlexecnitpick start with a capital letterqlsrcjavaorgapachehadoophiveqlexecnitpick please fix the comment hereqlsrcjavaorgapachehadoophiveqlparseimported twiceqlsrcjavaorgapachehadoophiveqlparseshould stack be parametrized hereqlsrcjavaorgapachehadoophiveqlparsejavadocs for all of thesehadoophiveqlparseThese can be packaged intohadoophiveqlparseSince all the process calls are being made using reflection we can get rid of all the process functionshadoophiveqlparseThis can be made an interface instead of a classhadoophiveqlparseIdeally this could also be done through a second pass walker or it could be done in the same walker as abovehadoophiveqlparseSome description would helphadoophiveqlparsedouble inclusionhadoophiveqlparseprobably this can just be instead of taking the class name and then getting the class out of ithadoophiveqlparsejavadocs for each of these variableshadoophiveqloptimizerjavadocshadoophiveqloptimizerWhy is there a mapping from null to currTaskhadoophiveqloptimizerCan you explain what is being done here +SergeyYes CXFRS client is default to use we can remove the set header code out +Error occuring still in branchpls refer screen shot attached +Max the first part of the problem I already solved +Patch to remove argument case for fixes to I have identified the problem too and have the same fix in my local copy +The patch is now applied to trunk +Fixed as part of. +This does not seem relevant anymore +We have moved the Perl bindings to the qpidmessaging namespace and updated them to more closely match the C to keep them consistent +Mr +As do the checkstyle changes +If the regressor has a changed value then we would say it is not a constant +Guys lets start from the test projects that would allow to reproduce this issue +added a missing sure security logger is enabled for jobtracker +Also added tracking of batch overhead and refined the crf function. +If we arent going to remove them then we should at least deprecate brought the patch up to trunk it was very out of dateWill try to make it committable soon +Ideally this process would preserve defaults as well as doc and arbitrary user properties +Logged In YES useridThe transaction is no longer stored in the Invocation mapsthat are marshalled. +Changed with r. +Assigning back to Jervis +And maybe hashcode should be replaced by identityhashcode +Eclipse plugin works for and also +will this bugfix be ported to the future version +The client +Following the thread it looks like quite a few people have favorably reviewed the patch +The if not JAVAHOME goto got is the check for the env var +If not specified this information is printed at a debug levelCurrently the additional bits printed are the Ringmaster Mapred and HDFS RPC ports +Several would not pass as written so Im not sure if they were just overlooked or if they tested behavior youve fixed elsewhere +Now works without ajform as well +Unfortunately I didnt pay attention and lost that code +But there is nothing yet to share +It was my understanding that given a commit interval of the flow would look likeread processread processread processwriteHowever it looks like its nowreadreadreadprocessprcoessprocesswriteIm just wanting to make sure I understand it correctly +With shutdown S temp is cleaned out with M is left behind +More to explore this week +I found some use cases which might make the patch obsolete +Also it looks like the was introduced on in revision +Committed revision Thank you Daniel. +Has this been looked at? I added all the information you asked for to the Environment field +Status update Sept Progress On Due Date Fuse previewThe below test cases have been implemented and committed to nightly CPI testingBasic following test cases require based clients for automated testingll add Jira entires for these note has not been resolved +This is a feature request +I should set this value otherwise sessions of will not be closed +Please keep in mind that keeping all cores in the directory is not feasible if you have s of K cores +Thanks Ferdinand that is added now +Jostling around the release goals and numbers +create an EDI Parsing resource for parsing an EDI message +A patch for review +CtrlshftT should work on any perspective that have the Navigate java resources actioncommandset is independent of what editor that is active +bulk defer of featurestaskswishes from to +Committed a fix for the must have a root thats a valid readable directory problem to the revision +What do you think Mark +The default should always be none +Changes to enable working with filters for subqueries +Thanks a million for the patch! + +Can sombody tell what kind of patch is supposedly available? In the error still remains +Looks like we agreed with Dhruba that whether is separated from or not it should be reorganized in order to support locking he is working on nowI am still not happy about how it is separated in this patch because it reduces functionality and introduces code replicationI propose to leave it as it is in the and open a new Jira issue for separating theI volunteer to do the separationIt is important to proceed with all other inner classes +Its no bug the regexp syntax is optionalFieldstuff +FYI The expanse attribute has been depricated +With that said Im not against the patch +does indeed cache already loaded +I hope some JBOSS users can confirm this fix +Patch commited +Can we put a customer case to this +Im not exactly sure what happened in this case as Im just documenting the poor error handling by the AM on a job I was asked to analyze +Can you try this patch Ryanbq +I think it can be safely backported after the branch +LatencyTypeMeanSDththththththPSnannannanBBnannannannanh +If versions do not match both status and version are synchronized +Since Derby used to handle interrupts in a better way before we switched to NIO would it make sense to change the issue type to bug and flag it as a regression +I think it could and possibly should be in the exception handling at the higher layer +I am putting this JIRA as blocker for EAP +This would be a method of some sort that we pass in a Confiugration into that either mutates it? I dont like the idea of the mr client layer needed to know which settings to extract from the minimrclusters config from an encapsulation point of view +Uploading Stacks patch againBTW Stack do you want to open a RB on this +Thats I am talking about dont clutter API for such useless and inconsistent out +addendup to patch instead of adding hashcode to classes that override equals unnecessarily rm the equals methodinstead of adding null checks to variables that Should Never Be Null add assert! nulllook ok +Thanks for the review +I believe this is a simpler better solution +I havent thought too much about how to do this +Basically it supports the case that WSDL is used with a java interace annotated to be BARE +tested against latest and all seems well +Consensus has been reached that Creadur would be a suitable name for an Apache TLP. +many things have the CR feedback essentially reverting some my from RB +Links updated and pushed to website. +Looks straightforward +But I do think that this issue is just about adding hyperlinks to the website +The accepting thread notifies the reading thread of the newly accepted connection through a pipe +Oh didnt notice Ed is already looking into this +unfort +This isnt a spring issue NH cant handle that type automatically use the overloaded method that specfieds the types of the parameters in addition. +This issue is the same as +The content and api has changed completely so its worth taking another look even if youve read the previous version +Patch applied thanks Lukasz! +That attachment belongs in which regards module deployment etc +Anyone should be able to just switch to an https version of the URL +Thisd be complicated with JVM reuse +Would love it if any Excalibur refugees could explain things better there +Setting component. +Same errors on OS X! that bad but at least I can go back to my Mac +You can see the code style guidelines for Felix under the developer documentation +It would surprise me if Rubyists everywhere would not react to hearing of its summary amputation in in the interest of shaving a few milliseconds off some with and scoffingElimination of Fixnum operator overloading is a very reduction of the Ruby language feature set and I see no justification for it other than a need for speedOn the other hand if of developers are really doing serious and truly need this functionality then so be itBut my suspicion is that this number is rhetorical fiction MichaelMy perspective is that users who require that extra few milliseconds are the +The methods with deprecated tag in the method comment should not be usedI propose that these deprecated methods be removed after tests are done. +I dont know if i will really get to it amidst the busyness of summer but i have been hoping to start pushing a release process by summers end +We shouldnt cut this off at some arbitrary number! We should just as soon throw expcetions in Solr when too many QPS are reached +The exception is thrown probably on due to a to low transaction timeout value +Several changes have occurred on the trunk related to recovery and initialization since this issue was raised +It will need to happen some day but we can wait to fix it. +Generates schema defns for dynamic sdosI have included some testcases for theseI have used and Axiom libraries to build the schema definitions and then to add the sdo annotations to the generated schema +depends on it so the reflective emf parser adapts properties parsed as integer when the EMF generated one is dependency over so the model for is type safe +Had to make a small fix to the test for postgres and include a default value for the column since pgsql wont let you do that all in one be fixed in +also it would be good to attach a processdefinition producing the problem in that tried reopening the issue myself but I dont think I have permissions for it +Then wed have to save more information on the exact directory structure for each partition and it seems to get more complex then it has to at this stage +So you have to add the setting +If you throw an exception +Logged more info for put +Could it be optional +Sorry for the inconvenience +A test would be to have more than one paragraph and see if all paragraphs get selectedBTW I think a workaround may be to call selectRange before calling applyFormat but that will move the selection. +Please add affects version +r patch applied with thanks. +Any update on this one? Astrid did you solve the problemJustin what about the online resource? I see its still as it was when Astrid reported it +Significantly rewrote the bug description once we determined the origin of the problem and created a reproducible test case +This patch cannot be considered because it was not granted license to the ASF +What if the next product is released next day +The root cause is that the ear is packaged as a osgi bundle and the resource put in the root folder cannot be accessed by the bundle classloader +OK I figured out the issue +bq +Thanks +the listFiles is probably related to this? likely a timing issue where the filedirectory is deleted out from under you +maybe we leave them alone but for now i just want to make progress in contrib on a unicode support +I fail to see where this is a Struts problem looks more like an encoding problem to me +also duplicates +ISTM the easiest fix is to always stream from the node that will be removed from the replicas for each range unless given permission from the operator to choose a replica that is closer less dead. +I put out a new snapshot of the JSF Deployer that should get picked up by Hudson soonIt looks like fixing the Weld failures is going to require some coordination with the Weld deployer +Please read the full issue descriptionThank you +I ran our test plan again and all was good at the usual threads level +I updated JIRA description with some that I know are part of the backport +Thanks. +So lets clarify itOn I keep it like that +Okay then please close this issueIf anyone else encounters this just add an exclusion for to the dependency so it wont propagate to your project and hence not be rejected by maven enforcer. +How does this issue relate to which proposes elimination of client side cachesIn view of is this change still necessary +Has there been any progress with this issueI am using Hibernate Envers in a large reservation system that is being based on a legacy systemThe old versioning mechanism effectively had revision groups +done +Thanks! +we can take a look for but a patch that is classified as a dirty hack with no tests in critical functionality is not something we can just apply +I need ledger ids for a different reason simplyI have multiple writers writing multiple ledgersI have separate readers reading those ledgersI dont want to maintain state between them about what the list of all ledgers are since BK already has this stateI am currently using some gross code that reads the ledgers directly from ZK +Its up to you if you want to add it or not of course but either way would be fine by me +Here is fix +Maybe it is just a missing slash in the patch file +Gabriel can you have a look at this one +Change to take a new rd argument autoincrementValue +If you finish it before we release beta please correct the version +It wouldnt be easy to change that my opinion the fix is simple is renamed to we would want to use the same code for Kepler and Luna we would have to remove internal API which would be difficult to do because other parts of JBT also use it +Isnt the file bytespercent transferred information already thereAlso we still dont have a way to associate a call to sstableloaderbulkload with a specific streaming session right +I made a few changes to the overall patch +JPA is currently unsupported +Will somebody try the patch +Copy files from old to new package +pete +If we consider the gcgraceseconds time to effectively be a max time out of sync then it would make sense to apply it to hints as well if a node is down for longer than gcgraceseconds then there wouldnt be a good reason to deliver hints to it since it should be considered dead +Anyone know how this implementation is coming for Android and iOS? Any idea if its going to make it into the next release +Updated in SVN revision . +It would only break anyone who put a folder in their lib directory manually +That probably what has happened +BULK EDIT These issues are open to be picked up +Just updated the javadoc for +Perhaps we could change the deserialize API to be optionally passed a length or somesuch +If opening an SSH socket on the server is the only option we need to clear that one up first +It also requires that an be present +Todd our internal branches do not have Avro support so thats a little overkill for us but might be good for a subsequent task +gcc and gcc arent binary compatible and we need to build our code with both compilers +I think were up to date in trunk and now +Maybe we should put some kind of block in there until template initialization is complete +The memory leaks and other issues originally reported were in the test case not in the +OK this works for me but +Well it could be that the was missing from the patch +Not if you also include the sequence flows or am I missing something +So Ive taken the code in the original patch and updated it for Solr removing the exception and adding a couple of unit tests +AS in AS +Committed Scotts patch to trunk and +We dont see it even trying to connect until s after the session was closed +I will reopen this issue as well +It looks like the above issue to create incremeters has been fixed so we just need to add the additional schemas and update the with code changes and SQL scripts for Sybase and SQL Server +I got around the issue by just passing a reference to the scheduler into the beans property on the exporter. +Attached a copy of the PDF in question in case the site is changed +It would be better to add a test case for cogroup as well +Ill check this in +too +Please contribute a patch +Is this optional? I mean can we leave the default to seconds +Forgot to include fixed that as well +So one could do a dump of the block map and at the start of the dump record the transaction id +A little more work but the right thing to do. +Can not reproduce on WindowsIA +See But you can use them in your IDE without problem just dont click Update in the Target editor its not clear what it does but thats probably not what you expect +The same should be done for the iv method in jruby +Ok fair enough ill tweak those rules when i upgrade Nexus to to minimize any downtime +Should be fixed now +Steve Firstly i am sorry if my questions have troubled you +changed fix target to unknown because it looks to me like this was not a bug and therefore shouldnt show up in release notes. +This also adds an to complete the set of adaptors for all of the currently supported protocols +client package and mock out an which wasnt very nice +why do you do that +Does this also need to be delivered to the stream +I just realized that my changes werent showing up because they werent beneath srcmainjava which seems to be the only place the changelog report is looking +Yes it looks really good +I dont really understand very much but i thought I understaood it enough to take a stab at the changes paul suggested to incorperate this type offunctionality into usingNot suprisingly some of the new tests I wrote for this functionality dont pass +The old one between and that is a or speedup +Cannot reproduce the issue under reproduce on livedemo with BETA under the issue is gone when using at +Removing automatic repairs is inline with some other improvements I have in mind so Im fine with that +I am getting the same type of error +Ive just committed this +dll from nantbin +As the form says Only to be filled in by the assignee before starting work. +Please mark it patch available once we branch +The alternative is we dont use fuses option handling capabilities and use our own but i think thats even worse +In this case I dont see anything wrong with that +The patch looks good to me +wow thanks for the detailed specs +as discussed on the mailing lists the official Tika project direction is that we will support Java. +thanks since is now released +Let me clarify The final state of the application is fundamentally different from the of the application +Attached +Im going to commit this momentarily +if subunsub isnt frequent it might be OK +It looks like the escape character works fine and the key is to use an Analyzerthat will not throw it out +Where can I see the changed documentation please +huh? queryNorm is passed down by to all of the sub clauses +OKForget it anyway its commited +Claudio for the review! +It should be possible to scrape and use some heuristics to generate a maunal verses automatic reporting schedules +and I think I have seen that myself +New patch including all suggested changes +The dependencyManagement element is not for managing plugins or extension +patch ive attached is just a simple patch for cr +Okay after playing with this a bit more it seems like I get either the prepare notification or the normal one but not both +Meanwhile I have looked into the source of and it seems that maybe dialog still should be improved +as far as I could see at the moment all types are allready implemented +A lot of issues regarding the network communication were discussed in make sure you are up to date on these +Hey! The after the was not there of course +It makes sense Dan I will update in that way. +here it patch to trigger Jenkins +Can I access Jenkins to download an compiled version of the SDK? I have no time to set I an environment to compile the SDK myself +The attached updated version fixes these two issues +Sorry remembered too late that svn doesnt have an equivalent to git stashPlease review +UpdateAdded an indicator message on the progress page to list an exception if one is caught and a back button to bring the user back to the plugin install home page +and committed! +Closing issue as it was included in version which has been released +I should useAfter I changed to everything works fineSorry about this +Not sure how to tell the exact Console version but it displays in the bottom left corner +Here is the patch to fix to recent code refactoring patch need to be updated +Im trying to think of scenarios where the connections options would still be required +Would you prefer that something be implemented by a best shot in the dark style of guess +First version of the patch +Thanks Luca +A converter provides to annotations one to declare it as a converter plugin and one to identify the keys that it matches on +Ill commit my local changes in the morning +agree with +Weird but have noticed that not too often but twice +So a usecase for why this is a blocker is definitely required this close to GA +Changing it to httpsFlag would constitute a compatible change and require to become +Revision improved navigation by displaying requests methods and partial path. +This could be handled by attaching active connection to current thread so that the nested calls would always obtain the active connection +We tracked this down to the blackberry NDK tools not being available on the system PATHThis section in the docs got lost before publishingAfter installing the Native SDK its tools must be added to your system path +Not sure why I cant vote but one vote from me. +Does anyone know why its there +I will update my svn copy and commit second final patch for this threadcheersandy +Unfortunately the code hosted on the link you provided is released under GPL which is not compatible with Apache licenseTo include it into it would be necessary that you attach the code in question to this issue and all source files with an Apache license header or you update the license on the project hosted at googlecodeLooking at the source itself it looks like you did start your work from the original SunOracle code which will make this even more complicated +The attached patch fixes this problem +Looks great +Moving the issue post EAP GA +The order they come in is basically random so I am not surprised to see that it is different between the two logsI have no idea what is happening with the standard deviation so I will look into it later +SVN Revision +Should we catch the exception and provide better message +I finally had some time to fix the issues in the previous patchPosting new patch that shouldnt throw any nasty exceptions if the user doesnt provide all the needed version with error handling etc +Please try again with fresh checkout and reopen if its not that was fastUnfortunately neither of the approaches worked for meNote that I applied them in my targetconfWas this as intended +You sound like you have other concerns +All Owner finders get the Find by Pet header due to it being in the Pet section of the menu +Ive tested the above changes pretty much exactly as above in and and everything so far seems to work fine in both test and live environments +Your patch makes it uniform +Will commit unless objection +It is not fixed there use or build jbossweb and change your jars files. +Thanks! Looks good to me now +Is this not a feature of Hibernate? So I can open a bug there right +This is moot as we can tak a message as DONTBUNDLE to bypass bundling +um this shouldnt have been committed to trunk I still havent looked at failure I will probably do todaySorry for confusion +see how that works out. +Re this why twice? comment in test case I forgot that there are sorry +should note either here or on the page that performance optimizations should be validated by benchmarks. +Can we rename to run JBM test suite and set version CONTINUING +I have fixed it one moth ago AFAIRPlease when creating an issue dont think that all the fields are optional +This patch is based on Olegs latest one with a few changes Message Multipart and are disposable but not all Body implementations have to be +I was going to respond with that but then noticed the type was wish +This attachment contains the dependencies packaged as uploadable maven central artifacts +All of patch looked at except the classes in the builder package +this patch will fix the follow bugs The happens in the constructor of because the System property is undefined +Its a waste having it as a comment in JIRA. +Moving all issues to the CR +Im really looking forward to this enhancement! Thx +With it all of my unit tests are green in the four combinations of root servlet pathspecific servlet path and root application contextspecific application contextI think this needs to be included in the branch if it is to become an RCThanks again David. +Just to be clear this is bypassing the agent process +Well reopen if theres a request for this target. +There appears to be a bug on delta CAS serialization for content added to a named view +Users can take care of it if they knowOn the other hand I think this is more a feature request than a bug +Fixed in root cause of this trouble is the removal of the empty catch block in which turned out to be painful but important. +Thats great! Thanks +Ive commited a so everything that is in Abridged is actually available from the subparts +for the start I managed to add code that allows to run the project or single class +Hmm +Can you break out the logic somehow so its not duplicated in places +patch for trunk +the issue because the floating White Stripe is still persist in page of Welcome with. + as well as are definitely not thread safe various methods of those classes incrementdecrement things like size only thing that bothers me here is that throws not very nice exception in such maybe it is good makes such errors clearly visible +Is it sufficient to only check for scheduler shutdown before scheduling? Im thinking of the multithreaded case where we check if the scheduler is shutdown and it isnt but then another thread shuts it down just before we schedule +Screen shot of Friday the th being displayed in Calendar as a Saturday for the Australian time zone of GMT just committed a fix please with that one +Providing additional constant isnt an option basically its a bug as other options are overwritten +This one doesnt +xml whereas it should be declared on. +The topology change is started on A only after the invalidation has been sent to the time the entry was requested from C on A no value was yet present on A but C was registered as requestor +Ported a class to this style of logging to provide an example +I just committed this +I know you are still waiting for a test case +Client libraries such as this Go library are no longer being maintained in the main projectPlease close +the Topic is VERY stupid +These have been committed and Avalon deps have been removed +I do recall a with regards to the class that uses that has been resolvedScott if youre not using the latest code from CVS try synching up and run your tests againFor completeness I will add your unit test to our test suite +trunk . +Im using Windows ll look into the path separator issue tomorrow and see if I cant produce a patch for it +No was intentionally used +Evgueni should be compiled from source +yea fabric can be a moving target a lot of the time +guess we need to supply a tool to rename sstables files if anyone is on longer namesWe probably dont need to do anything +Sorry was the change for Blocked Tests just house keeping or is this still an issue +My initial cents Im not convinced the effort is worth it at this time +However Ive done quite a bit of work on related issues recently +Just getting it +right theres a chance someone is using it for both and has both metadata files not recommended but dont want to break it +A quick test project is attached +Plan to close this JIRA if there is no objection. +Hiyou did it the wrong way around at least as i do it +jar +It has been too problematic and can only be solved with a better type binding library +For some reason the deployer is trying to undeploy which does not contain anyIssue this is a violation of the xsdIssue the attachments should not contain the metadataIssue the deployer should only use processed Happened after EJB deployment failed because of duplicate component in a copy of the same EJB being deployed before. +For starters could we add the filenames to the Exception? That way users could submit broken files to this ticket in the future. +But then the application could have actually used the again considering the fact that it still remains valid +As before extends Scanner entity to support filters +Pls remove +Resolving would require a patch to the Mojarra impl bundled with EAP. +Eric this doesnt apply to trunk +Otimizing should fix it for now +Smooks editor is no longer being developed +database x has a bigger limit so derby is not compatible +Logged In YES useridIs it clear enough now? Please let me +sure lets get it fixed today +I will check the TRUNK and versionDid u check for the trunk and +Should be good now ticky one +bumped to +See But Runtime Detection doesnt recognize SeamAssigning to Snjezana to fix it +This is something we would like to contribute +Good catchI think Ill leave them in images and reference them in the css as images instead of images +bulk defer to +Original commit is fine +Damn I thought we got rid of these all already +Committed to trunk +Clarification test fails when log replay is turned onWait +fixed by revision removed reference from to changed histories map into a resolved issue +Do you need it applied to the branch as wellcheers the way atom push tests are passing for me if you could check whenever you get a chance what might be going wrong then let me know please +Ill give that a d still like this to put the precise errors on stdout though +I forgot to move a couple of formatsEverything thats an IO format at all is now is consolidated in the lib directory all package dependencies etc +Can you break it into separate jiras reduce inputs for limits should be an optimization stepandSupport fetching for simple queries with selectlimitfilter operations onlyI think moving the limit stuff to an optimizer step is an excellent idea but it might be easier to reviewcheckin was thinking about it more what I said above cannot be performed in a custom hookIn a hook you dont have enough information the type of the query stats etcIn a hook it is too late +I believe you are right +Added a comment and a regression test case and committed the changes to trunk with revision +Thanks Kanna +Those test cases seem quite useful and relevant to me thanks for sending them alongIve updated the patch and verified that the regression tests passIf there are no additional comments Ill submit to thetrunk over the next few days +Unassigning and marking Patch availablePlease review and commitPatch provided for sorted listUnassigning please review and commitThanks for the patch Rakesh. +I dont see a thread trace from the replication handler +Thanks we want to make these String more efficiently by directly using Text instead of StringI saw the comment in the code but shall we do it now or later +exerpt in VERBOSELOGGING mode +Just to get an impression on this issue In which cases do you not want such a behavior? Where could it be disabled during options building or during parsing +It displays an error dialog that tells the user that is busy and that he should try again in a minute +So it cannot be included in Nutch distribution +The commit message referenced the wrong jira id +Reg test failures already registered with name HadoopservicenamesubControl seems to be one of the reasons for failing indicating an env setup test fixture issue instead of being related to this patch +When Ive got this functional Ill and you can have a look at whats there +Im confused +This seems to be an intermittent error +But the code seems to try to shut down the dbSo I think docs are correct the spec is out of date on this pointI have investigated a bit and it seems what happens when the connection is denied on the slave is this The slave receives the stop message It stops reading log and breaks of recover The boot halts but the db doesnt really shut down completelyI havent been able to determine why it doesnt shut down fully yet +Reopening to follow the PR workflow +No test needed for this simple change +Other optimizers updated in revision +Attaching client patch for tabs that have crept into files +Better to be safe than sorry +Teiid uses a Expiration based Cache in clustered mode +Just complicated and confusing +This will do for the final release. +Enjoy +Thanks! +Small fix +Waiting for the is not a good idea +We wont create this temp tempDir rightwe do a mkdir just in ant for this parent dir +I think this one is done now. +Its not clear which is better but my preference would be JSON because of its relative simplicity and because otherwise people will be tempted to want us to use full SOAP with etc +For licenses ASL MIT BSD would be niceAnd yes the link works now +continuous ingest test runs on a test cluster with the of entries plotted over time +Ill leave this code in place in case since it makes the test more robust +Build tests pass +Attaching s cleanup issue +Ill upload a patch for that +They only differ in device and inode numbers and timestampsSince the archive doesnt work for your Suse system could you please create a cpio archive of your our files and upload it here as well +I created a new patch for thatRegardsChatree Srichart +It was released with the fix in +IP as fallback +If I hear no objections Ill make add an header per default +Heres a new version that updates the spec to describe messages and to distinguish between stateless and stateful transports +We decided to do it the java way in classes +Such processor could save on initialization which for some cases is quite important +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Revert patch produces code that looks identical to pre and. +Typo in summary line +looks good to me +closing again tycho already reports missing features +Committed to trunk +I have also come to pretty much the same conclusion as Haohui that adding a REST API to serve the data and and using templates and AJAX is the way to go +Let the implementation take care of which options it wants to read from fdOptions +Thats very nice thanksIll consider adding some config options to enable those settings for the next release +Actually I checked the source and theres no prev pointer each Node has just a pointer to the payload and to the next node +Theres no such thing as attributes +The issue with the latter is that if the broker crashes btw the two operations the broker is left with no checkpoint file. +Why they do such Error messages when it should not be required? +Marking as resolved +Maybe this could be done as an enhancement to the existing goal. +Mr +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Look at this for +If i zoom out farther than the CRS envelope i get and also +The only thing I did different was select yes to installing thats bundled w RHDS +The method will be removed in the next major version +I would much rather externalize some methods into public static methods if needs be but then again one can just provide its own implementation especially for an optimized case like that Spring DM uses the Spring conventions of creating the inside the thats the main entry point for configuring a different m opened for suggestions but Id like to see some benefits before opening the API since were really talking about searching which is neither pretty nor generic +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +Also per default the laoder should be the same as what get produces +The proposed fix doesnt do the job +Thanks for the review SureshI plan to commit this early next morning if there is no more comment +Ram can you pl post results here +Looks goodNo test needed for code removal +Difference between Java and Java and original SVG file +So when a delete is applied while compacting you want to store somewhere else what has been deleted just in case +The presence of the +Hi FrederikI call the in the which is itself called by user root +lets change the close for +In this case all the outputs will be merged to disk +Just wanted to verify I checked out the latest trunk and looks like you got them all +patch looks good +New JIRA is so future +Can you confirm that these are the only issues that youre still seeing in the view cart screenFirefoxQuick Add right float dropped half a lineExplorer Quick Add heigth of a bit widebuttontext class not consistanly rendering top and bottom fixes the height of Quick Add of the in IEfixes the dropped right float in only issue I have with the logo solution is that most people arent going to be displaying ofbizs logo as their header on their ecommerce site +A patched is attached +Thanks for the patch +Im not a big fan of including arbitrary sleeps IMO we should try to write tests in a way that is reliable +I do not see the issue with version +Should the members fs trash current deletionInterval etc be moved to? quick comment Should be? It should still retain though +Same issue server restart the +Mark fixed this. +please provide the sample project you refer to so we can be sure we are talking about the same thing +This was already addressed as part of and fixed in was not resolved in jsp hot swapping was fixed by jsp debugging is still broken +What are some other peoples opinions about the default settings +Anything I can help? I can use my app to test +Verified +Is this open to community? Can I take this one +discovered a slightly larger problem dealing with dynamic copyFiels that i spun off into fine Hoss go ahead +Bulk close after release of +Straight forward port of work +cleaned up code a bit +Quick question did you also port? +Should have something within a week +This means it will not scale well across many collectors +Wasnt seeing the full pictureThe reason we dont see the error is because the core doesnt load and abort never gets set to true +What does inform do? If that is set will you see the warnings in stdout along with the messages about Creating classes for +Ping me with your email address +regarding your patch do you have any figures on how that behaves in case of millions of users and whatwould be the possible impact of the synchronized map +Thanks for reminding about this JanRegarding your question no its not needed +committed this to trunk and +a close is deferred if in any transaction +It looks like the right commit +Id like to do the same to x but Im worried that it isnt OK to break code where is explicitly used +Same diff with trailing whitespace left in +Thanks +This is not Major obviously +Thanks. +Are you sure? The version Im looking at has the cache a static variable for the result +Thanks for reporting the issue can you please elaborate a bitHow is it broken? Is there an error? How can it be reproduced +What you mean under Dunamic columns? hcolumns or richcolumns component? I tried to use both and give same result all facets rendered accurate under columns in case when column rendered and facets not rendered in case when columns not rendered +Marked as out of date since latest version of WTP has the fix +Im attaching and a new version of the HTML fileI wasnt sure exactly which example output you wanted replaced so I just added the new one +Or it was at one point once we have some more thorough testing and some more fixes we will have to see how that turns out +please fixed with rev +only one of key and value will be useful and the other will be nullThanks Doug that makes senseWhat about the? Does the same condition apply? I can see ignoring keys for the output of the reducer but what about the output of a map +I noticed the serializer was outputting the xml declaration which is not what I intended +The nested class is not necessary +Were not going to create extra deliverables forthanks Roland +In order change the default simply set falsetrueThis a global setting a per page setting has not yet been implementedPlease this issue if you think the solution is not sufficient. +Unit tests pass +Should be just a change to the error message +The only hurdle is specifying the template location in the absence of a solrhomeconfvelocity directory +Since is committed I think these should be provided only in new job client apis +Actually they should be instantiated at parse time +Running without the write benchmarks improves the first read benchmark significantly in that case the lzo block compressed +committed toCreated to follow up on bulk multithreading +I think this might already be fixed in the SVN version of this plugin +One is picking up on the current toolbox icon whch is part of the old logo but transforms that into a more modern look +The old patch was in the exact same way +The method is used in the setup of ezc suite to create a fresh configuration object for every test case +Remove useless initialization of rdbi +At what levels of bitset sparseness does it make sense to use random access? I ask because sometimes Solr actually knows the sparseness of its sets +Rob could you please take a look at Snjezanas PR +Thats what do for thrift now not avro +We might as well wait till the morning at this point no matter what really and that will give anyone else an opportunity to chime in +I opened for the search mode discussion +Please verify +I will try it out on since that is the version we are currently usingVishal +My understanding No changes in files should be required +Instead your application needs to make sure that all relevant parts of the object graph are +Ive committed this to trunk and +Thanks! +performing based on its recorded schemas c +Yes the issue is fixed +contains conversion from doing string concatenation in loops to using stringbuffers +Can be done as a post commit think the final decision is to put under derbyTestingsystemoptimizer +Like the functionality of the region please add workaround version changed to in order to provide more convinient workaround +Yes +Added a patch that picks up jars in ivys build folder buildivylibPig and the main lib folder lib +Can we fix before we commit this +Nahi Did you have a chance to look into the Ruby Archive projects mechanism for handling this? If they have chosen a standard way to load libraries from within an archive it might make sense for us too. +Probably in new version of XR will be working on Mac OS x +Thanks KnutCommitted patch to documentation trunk at revision +computing ranges +Pull request sent +Attempted fix preceded by cheap hack of URL protocol when dealing with a jarCommitted to trunk fixed on our buildbot which builds in a path with spaces +Oh one more thought When we get the PDF generation thingy working again we should include the PDF and HTML versions in the distro +Other than that looks good +I can then implement the changes we would agree on. +Assigning to you back so you can check and resolve this issue +Folks any objections to committing the patch? good to me +Targeted for RC pending more input are right +Changed property file location for inclusion in bundle jar by M on revision +Hpricot trunk now works with again so the next release of Hpricot will be compatible with +I changed it to Could you please provide a sample test case so it the problem could be reproduced +Just needs to be verified. +In the User field enter the github account you used to authenticate yourself on + +Coordinated with Sam + + pending tests are included but I can confirm that the presence of this patch fixes the test which had been broken on since the commit ofIm pretty sure the failed tests are unrelated + backwards compatibility is a requirement here since the is a common point of extension +Thanks to Mathias Werlitz sorry for the delay +When the directory containing the open files is deleted +Therefore it is not a test that has been designed for browser compatibilityI currently dont see how I could propose a patch that would have a chance to get accepted as you seem to defend the full correctness of +I have tested the wicket filter and there are no file handles any more! Thank you very much this will help me a lot with my current production issue on WeblogicBut I think I need the Servlet variant for a short time because IBM doesnt support filter in my environment +Tried with Java also installed well and loaded the for clarificationI dont see any mention in here verifying that the GWT facet feature we got has been verified to work +Update wiki with the mail overrule strategy TODO +this may catch an issue sooner +in +Now getting only deprecation warnings and jetty appears to run the app correctlyHave you tried test? It looks like we are missing a few dependencies there +Thanks. +Ill work on getting it in today +Juergen this has been fixed in of both files +Er or you could get rid of MDP +If we are going to do this rename is the time to do it +Is this a regresssion from +There are configs required in NNs There are configs required in NNs patch that wraps the last patches plus some additional edits to the source install doc +Im doing so in this patch +It can be the last optimization step and we can assume that map join conversions etc +I made to add it to or if this RC passes. +That appears to be a valid version so that would be good workaround +Confirmed downgrading to fixes the problemThis presumably means that the problem should be resolved with the latest release of annotationsIn the meantime the annotations page should state clearly that is incompatible with and will work only with +Please create a separate jira for that and well track it there +Youll have to try it out to know if it is faster for you but if you do try it out wed be interested to know any performance changes that you seeRegardsBrian thank you so much for your timely response +This is more a duplicate of please can you +I had considered added the term localized to the method names but Im not sure that would be a clear enough distinction +specific comments will follow +Did you see it going backward or out of order on SMP EnisThat was on windows but I believe it should be the same in linux as well +HelloI can confirm the issue on jruby on Mac OS X +attached fixes this issue +But we need to make damn sure that path is not used for execution +Hey Arup what is the schema of the table youre writing into? Is it partitioned by action +This patch seems to have been committed to trunk +Otherwise if only the exclude is defined make the include +Hence the need for a separate patchI have copied the changes from here verbatim to verified the test passes +Ill make a note of this on the wiki page +Greg what is different in your version of Hector respect the main branch +Mike I have have not started on the subclassing from yet +for the patch +Note this is for the global CF validatorcomparator not individual columnsAlso Adding ASSUME KEYS of +Arun is still investigating alternatives for map input bytes +committed in revision for branches +Arent those just the statuses for a party assigned to a and not the itself? There are quite a few more even specifically for a task +The Delete and Put would be grouped and then written to +In addition this patch passes the all unit tests +the importexport handlers should also have access to the resource that is being imported or exported itself not just the backing node +I click on Add to Group but the jar is not added +Ill apply the change to as well +should read Note the insertable false updatable false on redundant Column fields +Im wondering what the use case is for thatMy suspicion is that well end up with a Csv parserwriter in Lang at this rate +I suggest sorting out the migration path in another JIRA since this one is pretty unwieldly already +I will do it right now and give the result +As a side note I think we need to update the to the correct build number and date +Then a splitter splits the payload into seperate payloads +Actually Yasuharus idea of quoting strings always is not a bad idea at all +Closing all old Resolved issues +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +This is probably a bug in modjk +Fix in CVS +Also with security enabled jvm reuse tends to get used even more rarely +What is downside to this +Is your router installed with dhcprelease binary + +I just committed this +Without looking into the code Id call this a bug +SGTM +I dont really want to split them up right now but I am happy to file a separate JIRA for tracking purposes if the community decides this is a direction we want to go in +This relates to the overall package renaming task +Please try latest CVS +If memory serves mochijson did change its behavior when we switched from r to r to only support integers up to +Converted the attachment to a more common zip format +patch and test +I think a new JSR should be filled to accomplish this +Here is a screenshot fromI have created a cleanroom version of the test application that I am testing with +A reasonable assumption to make while computing used capacity is to assume that for all in a cluster the amount of memory per slot is configured to be the same valueI am a little confused about the above statement +Thank you Ram! +Such a class already exists in committed earlier + +Some comments patches have to be against trunk public api cant be modified we need to keep backwards compatibility need documentation for instance that setting http proxy will affect https requests unit tests would be for maven components for maven site is too complicated for Maven +Committed to trunk. +develop addin is right next to Compiler in the list +Assume that all support a GET parameter committrueI think we should assume this and fix anything where it doesnt work +Indexes one folder from IMAP account +So Ill give this solution a try and see if it worksThanks RonI think that with Eds suggestion we dont want to use the extended +Testing now +So its not the end of the world to add a dependency there as well +Fixed in trunk with tests +Thanks +to will fail because now when installed will up the version to and the branch wont have the upgrade path from to defined until the fix for goes in +Thanks +Make it +The command line that I used wasscomp d out srconly and I obtained the +can i bug you to upload the wsdlplease send code snippets that you want me to use both on server and clientthanks for being persistent and helping with fixing this bug i checked in a fix into CVS +I will do with. +So this is no longer a crasher +Kissing patches +These consumer have nothing to do and therefore would always trip the slow condition and eventually be closed +As long as you have a central repository either one will work with similar performance characteristics but having a single integer is just simpler +For some reason this has patch available but that fact is not reflected in the status +Those are all covered in the reference guide +Thanks Shai +Ive been overloaded with other things and am only now getting back to this +please use this file instead of old +All ESB plugins have Java compiler compliance level +I agree with Joels suggestion +this is because on the Mac is not serializableexternalizable +So I am tempted to push back here and close the issue for now until we have a real need for this +If it was not it will use contextName which if you look at the original start method does the exact same thing but within the method itself +yes +Yes that works +on on wontfix as well +This is in the next code drop +The entity class may have other constructors as well +I was told Fernando would be handling this +Initial fix for this is in revId Im coding up an integration test now that will guard against this happening again in the future +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Attachment has been added with description The patch as had erroneously used instead of an error pointed out tome by Arjen Duursma and now corrected in +Need to work with additional prototypes to flesh this out further +Many questions connected to visual design are solved and will be solved in the network of +Your solution works at the moment but it will be better when android is fixing the problem +At some point AS might switch to hence this jira +Its a huge performance boost and people use a lotAgreed that it doesnt really need to go into if we are hoping to get completed for that release +Some other perf tests around querying took a nose dive on linux as well though +Fixed in trunk and branch +In that case the javadoc of needs an update +Could someone with karma please close out this bug report? +I dont want to close this prematurely in case theres actually some work to do here. +That might have an impact +Load divided by divided by iteration divided by work time +Good for now +We can resolve this issue by reading through the to figure out the value later increase the Web Connector and port with the same value +fixed test +If the index does not change too often the overhead by a delay in unmapping the files at a lter stage does not matter +t hesitate if this occurs with a more recent version +And exactly one tailer can get started for each file +Attaching the patch with suggested improvements +why does getLogs not support XMLOops leaving in was a mistake +This is fixed now in the redesign +Do you have an example projectfile I can try this onYou are still saying it will create either a wire binding or value binding but the UI is stating the opposite +Taking off fix list until we have some steps to reproduce +ThanksFixed as per suggested patch +Thanks +the same problems with three nodes. +Building up an entire array ofstring entry names is wasteful if youre just going to iterate over thearray later +We have not required a branch in the past +As per your fix The actual instance of the interface type is injected if the resource is not initialized and a proxy instance is injected if the resource is already initialized +added section Iteration components Ajax attributes with description for attribute. +ClausI added a lot of detailed logging at the debug level +Done +I think that Izpack already has a ports validator but well have to write a custom one modify the existing one to take into account the offset +Thanks. +You mean rightI remember there was a specmojarra issue regarding it do you have the link to that ticket handy? +Id rather not reference eclipse codebase from config or goalsgrabber +Is it possible to move the mail store out the of the james path +This is ok because those edits were coming from a WAL anyway + contains code changes for java +Documentation now done. +In Firefox I noticed that the http is not selected by default +Default JMX will use one more extra random port along with the port configured via +Im marking this as WONT FIX +Were just checking that it reads in the string +have committed now for both trunk and. +I dont think that css layouts are the way to go yet as no single browser besides opera pass the acid test +If we tackle this at least in the way you mention through the profile service then we would be going against what our customeruser feedback is telling us +Thanks Simon Ill +Attachment has been added with description testcase demonstrating the P open bugs to P pending further think that if the inline is replaced by a block we should still get pages but we get +Accepting for review +Good question I dont think looking at requests is good enough +This patch adds optimization for diamond queries as part of optimization +See comment +Just feels rightEither way this fine +Other than the hiccups I ran into your code works great for what I guysJimmy +This patchAdds the nv option toEnhances to assert that the proper vector types are generatedAdds to validate the proper option behavior and vector typesExtracts random document generation code to utility class +Disregard this I guess +Attached test case +Patch v is rebased on trunk +ok that sounds reasonable +There is some confusion that we have more time than we do +This issue is addressing the possibility of multiple concurrent access aggregator in a distributed environment +Browsing through the code i believe we can conditionally append whitespace only if the element is in +because it is seen consistently and after i upgraded the host to the snapshots were getting created successfully +added that also ends up testingChanged in to useChanged memory setting of AM in because there were occasional failures with the old setting +But the correct solution in my opinion is pretty serious changes to the concepts of Loader and and +Anyway the config for it is in the apache pom with unique snapshots turned on so we are now building with timestamped snapshots +Committed revision +Too bad we seemed to have missed it for +but are you sure the entire line was selected? I tried again and the issue exactly m using jdku on the run and run selected commands into two commands with the shift key modifying the run into run selected in its various key combinations. +Something similar to the message that says Html Syntax Allowed but instead would say Comment Moderation Enabled or possibly something a little bit nicer +since with the SAX parser Ive only coded tag comparisons by extending the default handler there is nothing to be added to the from parser +Thanks for committing this patch +A screencast would be very helpful +This is the query plan produced by the optimizer when matchingtwo terms in a table that contains only the revised unique index +Done in Rev +Note JAX RPC does not require data types to implement thus this should not be an error +Uhh please calm down +JiriYou need to to select the background then the menu context will have done some modifications to make the action can active the background and it can work well in Windows I will find a way to fix it in Linux +Freddy all done +Aligned on in r on trunk +Fix for the problem +Outgoing responses are not validated +But if we concern about the backward compatibility and future support I will make both quota and count to be longComments I prefer to make the checking and updating count to be an atomic operation in stead of breaking it into two operations +The issue of reported has been fixed in other issues + +Closing resolved issues. +As we have the to resolve it we dont need add addition option for it +Please dont hesitate to do this on your own next time +Thanks Bikas! +Which seems a bit silly as its JMS +but it is the place defaults to when creating new repositories cloning repositories to +for committing to just committed this +Except that the class is dependent on which is in the +I wouldnt think that it would be too expensive though Ive never measured the cost of the deadlock detection and if you have to wait for a lock youve lost in any case so one might just as well do the extra check +I cannot reproduce this from my local vmware setup +As soon as Spring Framework is released Ill be able to spend more time here. +The thing I wanted to mention is that it has nothing to do with the project settings or so on its all about whats happening in the www folder within the webview +The change is to allow jsvc to be picked up from HADOOPHDFSHOME instead of HADOOPHOME +Hi MarcGreat work thanks! So as a summary for the streaming code memory consumption appears low for encryption decryption signature verification cases but high for signature creation +Yesterday Fernando asked me to retest the EAP CP using and instead of the jbossas and that were in the test compose +Hence the priority is marked trivial +Fortunately thats not a scenario normally happens to a real clusterLet me check if we can prevent from assigning a dead region out of SSH from the master +The file can be parsed by Tika without error now were on POI beta +literally replaced the mentioned the and javadoc mentions with their fully qualified names and the to represent the keys mentioned in the bug report +Do I need to switch to svn +JSON Support for Axis I didnt feel like waiting around for someone to answer my question so I went and got the original patches to build against axiscThis is unfinished and has a bunch of things that need to be cleaned up but it at least compiles +Can you tell me what steps did you successfully reproduced and in which step you cannot continue +At present I cant say that we are ready perhaps in another week there are still some issues on the CTXS side of ensuring no customer data gets transmitted with the transfer of to waiting for folks quick update for infraAs a project weve decided that we are going to defer migration for the immediate time frame and would like to request a project in ASFs jira instance be created in the near term for +Can we confirm whether this is still an issue and either close it or move it to +They both eliminate the facet data caching +Im going to need help understanding why some write paths isnt a sign were Doing It Wrong +Thanks thats great news +Didnt someone have a patch for this +And what versions of the various software are you using? eg AMQ JDK etc +In order to support SQL collations in the future we will need to be careful to design this so that it will be clear that the string columns in the system catalogs have a different collation than string columns +Arun this patch no longer applies to trunk +newest patch + +The patch for the Admin guide Network Server properties topics looks goodPlease let me know which diff files are ready to be commited and Ill commit themYou have made a lot of great improvements so far and I want to be sure we implement themas soon as the diff files have been looked over and approved by someone +Direct use of would technically be possible but discouragedAs described above there is a way to converge on the SLFJ API without breaking compatibility +Ah yes didnt notice that one +You need to host it in your own repo the policy is no snapshots in ibiblio +QE is busy with testing right now +Patch with the same information +Committed revision +the fix contains both JET and OPT fixes and regression test +Then the problem continues to occur +Done +Post another INFRA bug when we get the new ml set up +Lists createdmoved +what do you think +not a blocker +Bulk close after release of +Thats exception from the common behavior of passthrow attributes +Ah thats where the problem comes from looks like its a problem in Facelets implementation which saves out the state at this point +in the case where the +Patch for trunk +I included one more benchmark for file opens +Thanks RichardAnd thanks Nathan for adding testcase for that +Dan can you look into this one +If for example mongodb has some natural support for compensation or retry etcNormally for the producer you would just throw the exception and then the routing engine will take care of itAnd for the consumer you would normally have to do a try +Ted updated fix version for thanks +So for Karams request +Thanks for your reupload the patch +Slip out of I think this is largely done with our Hudson runs +Hi bill I applied the patch again +Marcos if you can run the tests again using the latest sources this should hopefully turn some red squares green +We should also solve the problem reported asI would also consider running both versions of the stress test for a while as part of and derbyall +This is not possible since there is no way for the tool to know whats in the controllers log +Fixes NPE in desc database +Thanks for weighing in Gabriel +should have been named and Well +Galder is the upstream AS jira. +But yes we should make this painfully clear in the documentation +In this patch Ive captured the max timestamp and stored it as part of the stats file +TedYes I was +We have raised this discussion as part of another thread before without deciding whether or not we want to add file locking to file appenderslogj does not have this and I have never seen corrupt log messages because of such race condition however it sounds reasonable to add file locking for file appendersThe main issue is what should an appender do if the file is locked by another process? Should it skip logging or should it wait for unlockI guess file locking also have a very tiny performance decrease because it need to do another system call +This is an MVEL regression have put in a pull request to MVEL to upload a test +Closing since change has appeared in Latest Alpha Manuals. +Dear Juergen Sir I got over this problem with help of your comments Thanks a lot +out files with missing built on Jitendras system +It would help a lot in order to test WMS general behavior to extend the OL preview with yet another config in the drop down usage of WMS vs WMS protocol +Closing wont fix +committed to branch +Anyways thanks for news Regarding delegating the functionality I had already done that change for both and last month +on patch +It will be the best solution that database manage the synchronized has sequence maybe +Please apply the second patch after applying the first patchThanks the patch +Ill add this to the next point release +Closing the issue now that it has been resolved. +Do you just the one instance of Roo shell running +Strong yes I think this can be resolved as Wont Fix +Still not tested but I think it is a good idea are there any objections around this? Jeremythis sounds like a great improvement +merged into openshift removed bootstrap profiles from pom are theyre not used. +A few immediate ideas that come to mind areCreate a new target for the compatibility tests in the main build file and call that rather than the target currently usedCreate external files as you have now but include and call them from the main build out the and versions into different files and all tests will be executed when the target tests in Note that the tests target will execute a lot more tests now. +Also I apologize in advance for this request but would you mind posting a review for this on phabricator? Due it its reliance on this site still doesnt display correctly for me but Ill use a different browser just for this request +If you have a service throwing an exception you get the exception wrapped from the service and then another wrapped exception from the Controller so to get to the actual exception that caused the problem you have to unwrap it twice +Thanks for the workaround +May the problem be in the endorsed libs stuff for xalan? At least this seems to be the problem for +Can you please rebase? ThanksGiven this is not an issue with should we still commit this? Im happy to but not sure its useful +moving to M +Committed revision Thanks Scott for reporting the issue and reviewing the fix. +We have some support now +Committing this +This is a known and expected dependency of Serialization +Im ready to submit a patch if it will help +In any case correcting the copyright is the most important thing +Now again the is usedIt seems that the problem lies inside the component +Ported from trunk to branch at subversion revision +The next build that should have this patch ie +As I know there is no default SCM bundled into eclipse +It might sound like job runtime increased by but it was less than th the original job runtime of the pure Java codec +I suggest to exclude this test up to end of investigation +If it is because the files have the wrong permissions then we may just need to ignore the files +Tried a unsafe operation +Applied patch in revision +Thanks for the quick response and sure I will attach the patch after making this change +Looks ok +maybe? m rocks I thik that anything that will push it a little more is muy common classes between the plugin and repoclean should be factorized out to a library see related issue +bq +Already clarified +Then they should be a seperate moduleIn there is no support for modulesIt would be nice to keep implementation roughly the same for and +Patch applied +Thanks John! +Whats the exact issue +Sticking non field controls inside a form is generally a bad idea as it break the basic HTML GETPOST model +I just spent some time trying to repro this on rev +This is fixed right +Technically its even wrong to use that char specifically if at all we should use the constant from +Jamie is cutting Karaf currently so there is a good chance it may be released in time before we start cutting Camel +Code change looks fine to me +Yet. +Thanks for the temporary fix! I am able to the issue now and continue progress +The documentation is already included so I think I can close the bug. +I am pretty sure is ok actually nowGood news! Thanks for performing that analysis +Fixed and a new has been deployed. +The issue is solved +The parameter has been renamed from language to languages +Thanks issue has been fixed and released as part of release +Modifications to be attempted on the next CP case patch +Can connection profile give you the name of the jdbc driver class name +Edited previous comment +And all tests passed on branch also + And wow you replied within hours +Actually the problem might not be in the rename +here and in bug +A modified patch applied +In my setup I got lucky to have such a partition combination +I have reviewed your patch and it looks great! I think I can check it in soon +I will upload a test soon +added to release notes as resolved The class now uses logj for its messages instead of sending them to. +committed w updates to comments and some inlining ofcan you rebase r to trunk +This is also transformed to colspans in HTML +Thanks +Verified on latest commit cccaffdbcdecTest Passed +And from John mail and the meeting with Max for the beta release we dont include the plugin into build +Btw Ive recently modified the configuration of the jetty client to use non blocking nio selectors instead of standard sockets and this also seems to work much better and does not lead to a freeze afaikThis does not prevent us from applying this patch of course but i thought it would be worth mentioning +Probably the test needs to be extended to at least one batch opLet me also run some larger scale tests on it +I tried that +i was so eager to see the home page and any one like me want to see here it a bunch +Can you confirm it is fixed in the CP branch and that it will be in the +just tried latest CVS +Steve how about adding a note in the file with the workaround until it is fixed upstream +Indeed Freddy this is already the case +Ok +This seems counterintuitive but after searching for a more elegant solution this is the best I came up with + is a specific instance of the general issue reported on as Done but the right resolution is Wont Fix +I have opened issue for that +It is a IBM VM bug and there exists a workaround for using a policy file +Did you see it only in Java Editor? If not only Please provide steps to reproduce +Another effect of added is that in Internet Explorer the Wicket output of exceptions doesnt work anymore +Nick I have reviewed it in JBDS Alpha +Thank you for attaching a program +note hbase config doesnt seem setup for cloudservers conceding this hasnt anything to do with this issueWhat do you mean +So we discovered that the interface change is incompatible +I dont know how people think about being able to use the stuff without spring +It made the order of the layers consistent but the ordering of the wms layers is reversed +There are things I think we can improve furter also use reflexion inside of force touse also when property is absentIn addition you may have a look at typos in javadoc +Documented in the API documentation for the property. + +Unit test fail in newest hadoop +Could you please replace the strings by org in your code and test itIm embarrassed by so many mistakes you have fixed for me +Moving tests that will require a full test OUT of for time purposes +svn ciSending modulesinterceptorSending modulesnamingSending moduleswebservicesSending Transmitting file dataCommitted revision Thanks Henri! +Reopening to correct the fix issue info. +The only NSSSSL option that is is the certificate name which is now settable as a connection option +Canceling. +This never seemed to generate any demand and the discussion has also grown stale +Can someone at least respond to thisThis seems like a very key feature of that is missing +If you are on for it be my guest +yes move it in soa +Will see if he gets either +Initial tests show speed ups in the range +I just committed this +on v +One simplistic option would be to just include the console in default and if people want it in other profiles they can copy it do I get this war from? Is it in the maven repository somewhere +Tested on EAP tomcat unexpected file deletion andor alterationNOT FIXEDDeploy from and than check content of +I will explain why +This has been fixed in archivatrunk +Ive looked at the sandbox code but I havent a clue how to use it +Closed issues related to Roller release. +hiSeanmaybe we need order it in the reduce job and need a reduce implementationmaybe mapreduce ruffle the data in the phase +Committed revision Resolving the issue. +Ive been dealing with Web apps too long +Many thanksBest regards by Andrew. +Once I do the next mdep release Im already planning to move most of the code to a common module +Java +fixed +This is also causing serious problems for my organisation +httpshudsonviewStableBranchjob +I have cleaned up my is probably a command line parameter to tar that sets the file permissions +In we had exceptions initialized with another exception as an argument which internally is similar to calling initClause +bq +This will work but you need to duplicate the and therefore maintain it on your own even if there is a small chance it will change over time because it will imply a migration processOTOH IIUC with the proposed patch its not possible anymore to implicitly use the default engine +Lucene codebase is impacted onlyMichaelPlease consider it for commit +Thanks +If this sounds OK Ill create a patch later +Thus I believe this issue is fixed but would be great if QE could check the full rSE flow and confirm +Was thinking this might account for both cases all through the mechanism +I am not sure right now but I think they are before numbers +You could do or +Hi Aaron +Either way needs to be fixedId be happy to shepherd you with the patch process +A few changes in this patch Made the interface static Removed logRow method +I think the code in the is wrong and I can write a lot of examples +Is this a case of? needs updation as Hemanth pointed out +Thanks Alexey +Issues are already fixed in the branch and trunk +The chapter you reference is now its own guide for EAP and will be released as part of the EAP of ticket raised in the JBPAPP project +went almost a day could it be something to do with that runs once a day +The modification logs is a rolling log file +In the Seam Dev Tools Reference Guide there is a sentence Lets create a WAR project using the New Seam Project wizard +I think this should be a Hadoop problem or communication problem between Hadoop and Pig +The other option to fix the build would be to just revert the changes in the jiras mentioned in the previous comment +Awsome! ill fix it right tom +That logic does not do a proper dynamic dispatch to in all cases. +Only two bugs are actually fixed between RC and +I would be glad to make the fix will wait to hear on the outcome from ML +Robert and I committed more fixes to the remaining tests to the branchI also added the test of to this branch and fixed +then this can be a candidate for BUT this may need to wait until we fully support Java based on the following information I found The IBM seem to only support PKCS httpdeveloperworksjavajdksecuritysecguidespkcsimplDocswhile the IBM include support for PKCS httpdeveloperworksjavajdksecuritysecguidesjsseDocs +The patch doesnt apply to +The problem encountered by is likely related to these tests +Ways to fix do not cache add some notification mechanism to which should track if particular property value is changed and let instance knowMy preference is for until actual measurements can prove caching is useful here +But this has been committed for so long its ancient history +Regression in +Here is a version of the patch where the stderr println has been replaced by agent tracing +Ted before you add this please cleanup the braces formatting to match the rest of Struts code +This patch allows the user to ignore the corrupt object. +I relied on that FK constraints would warn me but the particular one was missing +Does not fail with the patch +Some offerspecially designed condensed or expanded font variants +Since this seems to be OS X only well need to figure out whats going on there +Looking at that none of those settings should be influencing the management HTTP interface as that has its own definition +Looks very similar to +Resolved some time ago. +To use this feature you must xsd schema for web subsystem +Oops the diff is for +Backported to branch as svn +And Ill try to provide a sample project with the issue after +Please check that the bug is fixed for you. +I need to be able to duplicate this if I am to fix itThank you +Since the EL name still have to be looked up dynamically why would this be faster than +a few reason were discuss not to integrate this out of theo box +Assign to during tidy up prior to the issue now that it has been resolved. +As it stands without this feature not sure if they want it now +Pavol I checked in the client jar from SOA ER which adds patches to the +Agree if it is too expensive to implement we should forget about it +No but I didnt notice it when I released maven and its ll need more information to track this down but one possible reason is that the location of is as xdocs in the jelly script so this wont work if you set to something else +This presents a problem and explains the necessity for this feature +If there are still issues with it well open separate bug reports. +Mikalai Your name appears on my dashboard now +However this issue appears to be resolved already in. +Ive tested this on revision thats too old for the fix you need to come up to at leastIs this still a valid bugCan it be closed? +A message id that is not in the form of a UUID can be unique +Now we catch the and redirect to home page and add a Session Expired message +Closing resolved issues. +There is no reason we should have this arbitrary number at the end +Delta patch incorporating changes reflecting Chriss comments. +Starting conditions for the failing test depend on another test that modifies source +You can override the location of using the location of the properties using the location of just the templates using +Try this one should have the whitespace stripped +Testing from Alex on +not workign correctly on Trinidad Please reopen +Took a look at this +I am still reviewing this patch but two preliminary comments +Test case introduces a and jmock dependency but this is just for testing so I think were fine with this result. +Took some digging to see where this disappeared since it happened before the modules were created +setting to since easy upstream and in JBDS GABuild id +I can help set up apache build of the branch or do it over on andrews ec +Added a test +Im also not sure why you would load this many cores at once I thought the whole point was that a subset would be loaded and the rest on demandlazily so mostly you will just be enumerating very quicklySome OSs do slow down with many thousands of directories in one dir but you will have the same issue with s of core regardless they all have a solrhome and index dir and at that scale you probably want to create some form of hierarchy strategy for foldersIm not dead set against some sort of plugin here but you have to weigh that vs the complexity and cost when users start using it and it makes future refactoring more difficult +The and now only deal with bootstrapping Grails for CI systems external tools etc. +The client will need resolve the issue with the duplicate or invalid client Id and then create a new connection you cannot reuse that connection +I used a simple waitsleep loop here since this method is called from several places and it looks like a real pain to try to do a Future or whatever +I was using a multipart schema as well which consisted of main XSD with includes +Jason can you please followup on that +it works as u suggest. +Theseeach project have daily updates defects redundancies line of codesper class and etc +However i cannot see what i am supposed to be seeing +Even if the second bean defines a producer or observer method the method will never be called +I dont know if we want to back it out entirely as it seems to solve a race condition that would be good to not have +The current writeup doesnt propose how to handle compression elegantly +The issue isnt related to Runtime Detection but to ASIt happens because the plugin isnt started when calling the Runtime Detection preference pageIt is started when calling the New Server wizardThe plugin adds that creates the default fileset +Please let me know if this needs to be in too. + +I believe Ive fixed the javadoc warnings with revision Ahsan when you get a chance can you update your svn sandbox withthe latest revision and see if you can build and run some tests toconfirm that things look correct to youthanksbryan +I will be out of town all weekend but will have some timenext week to help out +Yes it is but admittedly its a little obscure +We dont have Cloudstack Platform anything installed if thats what you mean +Same hereError on removing indexing context local Cannot delete Cproject +Working as guess this is what just happened to me though I inially associate the problem with +Contains and +If it cant happen automatically let me know and Ill write a howto in the wiki for this +Having it be a just seems confusing to me +The two remained are also failed in other OS and are not related with slashes +The alias was also removed for the destDir variable in +If we add one it means we have to keep it up to date to match whatever one is pulled in by something else we do depend on whenever we change the version of that dependency that depends on Jetty there is a risk that all the orbitjar fixing dependency entries are now pointing at the wrong versionsWith that in mind I think a feature to specify custom mappings would be incredibly useful so users of Ivy would be able to quickly work around these simple cases with a mapping to tell Ivy that orbitjar +Patch checked inOleg +this is where i think the cluster connection gets terminated suggesting some sort of a problem the ats server is having with a peer +Affects doesnt work with Long converters +This is mostly done but moving it to to track the hibernate issue +Now with regard to the exception are you certain that the maven folders are causing this? Did you try unjaring removing them jaring and redeploygin +This should be fixed as this was the intention of uncommenting the ASYNC appender +Ive ran the test more than times +Extra space is needed to allocate +This is an improvement and therefore would not normally be considered for a maintenance release +Removed karma client dependencies and load history option from xbaya +FIXED +Seems to be working fine enough but its still quite a bit of new code so committed only on trunk +I see no reason to remove the String only constructors as long as they are well documented +It is only error in the tag file wich allow access to this attribute from page. +trunk Added Spring DSL support and thread pool configuration in the Java DSL as Multiple toAsync in same old resolved issues +Migration to done +This patch must be applied first to allow the real patch to rename to +first copies src to tmp and then rename tmp to dst +its an improvement to be sure +implement a cache that sometimes threw locally it would fail the CLI tests since the locally thrown version would not start with while the remotely thrown one would +FIX Added the missing codeblock in the generator +This was much easier to grok once the tags were refactored in to components +Need to fill out a few more tests +Will go ahead and commit the changes into into with revision +Sorry I cant remember which pullAnyway Im marking it as resolved +I did some testing with RC and it was consistent with GA. +It might be possible to do something like You havent closed the slash which started with Foo +Originally I hasd distributable in but because there is no clustering in AS yet I removed that from so its not distributable +There should just be one for param value the current value in a list iterationor somesuch +It looks like if a map output sorter has to do a merge it crashes with a null pointer trying to create the temporary file name +Hi ArnaudI have fixed it this morning +Regarding metadata Every partition has its own Storage Descriptor object which contains information about file descriptors serdes and other storage related information that a table can have +If you find further issues please create a new jira and link it to this one. +Thats not what wed need +We can definitely do that for the release following +These are optional and provide some debugging information +Although it seems to me from the annotation point of view a given class can only have one Entity annotation which has only one name +whichever ends up being cleaner to implement and useevery container using shindig will most likely have a different set of fields that they support +or at least for the merge stuff +At least Ive never heard of such a request +The stack trace shows its a map with an in itAnyway I hope this issue report will be useful for anyone +Then is this the same as +The website just hasnt been updated. +And just map the properties using old fashioned java code +im interested in this issue +We have to consider both acceptParams andCan you be more specific what suppose to be done +Here is a thing if you close the input stream of the you will not get a readable from the message body of the exchange which will be processed after thePlease remember to call the close of the If you want to make sure the temp file to be deleted + svn ci adventurebuilderAdding adventurebuilderAdding adventurebuilderAdding adventurebuilderNOTICEAdding adventurebuilderAdding adventurebuilderAdding adventurebuilderAdding adventurebuilderAdding adventurebuildersrcAdding adventurebuildersrcetcAdding adventurebuildersrcplanAdding adventurebuildersrcplanAdding adventurebuildersrcplanAdding adventurebuildersrcplanAdding adventurebuildersrcsqlAdding adventurebuildersrcvarAdding adventurebuildersrcvarsecurityAdding adventurebuildersrcvarsecurityAdding adventurebuildersrcvarsecurityTransmitting file dataCommitted revision +I havent tested this with earlier versions yet +Alright v looks good to me +YesStill for the long run it would be very cool if we can get some kind of counters that are sychronized across the cluster after a barrier syncThis will be a lot more convenient than the tag checking of the messages because you then just increment an update counter in your code and after the sync it is guranteed that the counter value is consistent across the clusterOr we can add some kind of a method of syncCounters which has less overhead than a normal sync +This would only be supported via weblogic +requires also additional files likein etc folderDont forget to copy them too if they exist + for the patch +The code change provided works fine +Old JIRA cleanup +The reason to provide no annotation for this functionality is that you may often not be able to declare it at the super type since it is provided by rd party or the JDK directly +Rejected + +it looks like we also have to output characters we find inside the sl tag +Ive made sure to update the version number there for the final probably wont bother with fixing the RC since final will be released in days on Nov +Perhaps can be addressed when the proxy API is revisited otherwise we might bump it out. +From the log it looks like C is started twice in a row +I think at the very least that should match the behavior of the gem +Once the file system has been detected as being unavailable dont recheck it +Ok thanks jacques +was resolved already in text currently in the is now optional +Alexey I somehow missed that thanks for notification +Using asynchronous JMS Continuations supports this use case and seems like a useful feature especially in a true ESB environment +It was actually started there as the first Main class we offered in Camel +Its working +Wrt requirement +Uwe is it a problem? +Thanks Dennis looks like this can be closed out +We will release soon +Rebase again +What about +Quickstart +Now generates the asynchronous client in addition to the standard one +nnHOST and HOST must be taken care of by the dfs clientThoughts +attaching fix we now check for a null value of namespace declarations and turn them into empty strings before adding it to the attributes list +Aside from my other comment above +For what thats worth the patch at fixes that too +Fixed in +Do you mean that the Tools Mobile Browser Simulator featuresplugins incorrectly fail to declare ALL their dependencies to other and featuresplugins? Or do you mean that the plugin isnt correctly adding itself to the perspectivecommand groups +The constants are now protected. +these two naturally throttle the rate to around deletes a second +However it is not clear to me how to fix this +I am reviewing the patch now and really like what ive seen of it so far +before my SDD crash I had a basic prototype implementation of adding accessibility patterns to the jars was rather trivial to do +Thanks Ted +Max allowed age I Current disk usage +Specifically where are your classes that are annotated with annotations what exlusionsdeps you have added to any subdeployments or you using any other library that would pull in resteasy via some other path +Server side work for the given WSDLPlease use for deploying this service +If someone finds an example where the current two mechanisms wont suffice then we can always create another issue with the specific changes needed and why. +No errors appear in the server logs or agent logs +I wasnt sure what the protocol on changing the expected behavior for was for +This sounds really really fragile +Tristan assigning this to you since youre already looking into Ok? Cheers +I hereby confess that Ive been wrong about Namchinovs patchIt produces horrible results see sample bellowThis patch shouldnt be applied by anyone +The problem is actually not with the space but with the In URI is treated specially and we wont be able to support without changing hadoop code +is flaky +Would prefer not adding additional configs though +Why is the instance metaClass property sometime of type and sometimes of type? Thats the problem here not the existence of metaclasses +recut version of old was this committed? I dont see it in the log. +If youd like to help provide a patch for it would certainly help +Ill try doing it with extra trailing directories and see if it fails for me in the same just to confirm this is on right +The reproducer also includes the hang when you supply wsdlLocation in the annotation +Closing +It would still be good to understand where the problem lies however +Can you post an URL you are using and turn on debugging on to catch the exception thrown by transport +This might happen because of VMs behavior or JPDA agent which is loaded into VM address space +We need to change it as soon as possible because the affected test is part of the TCK +Please correct me if I am understanding this wrong the logical name id has two uses HA case to differentiate the we need one ID per RM for this +Also FWIW a connector like this does not invalidate at all earlier discussions weve had about including a pipeline in itself +Diff against Grails strips out comments empty linesUsagegrails war issues that arent going to make it into to +fixed whenever HAN is close when HEM is embedded there is no HEM +If you think its ok I could set constant score rewritetrue in this caseI agree that it should just be left up to the user +or provide me with such a patch +Hi I tried with forge today and I can confirm that the issue is gone +bulk close of all resolved issues as part of closing items. +updated diff to reflect comments since my first diff +Very unlikely +Willem do you mind fixing this +I guess I had the wrong beta versionThanks +Marking as could not reproduce +For example shard gets split into shard and shard etcThe node names are created as collectionshardreplica collectionshardreplica etc +Fixed in revision +Unfortunately Hudson again didnt generate a proper but different tests failed now so it is quite clear that the failures are not related to the patch +Moving out of +hmm ok +seems trivial to fix +We have tables and partitions servers etc +Thanks Stefan for explanationIts indeed not that intuitive at least it wasnt for me Therefore changing the issue to type Improvement +This is because MRI looks up the executable on the PATH before attempting to run it whereas jruby passes the argument array directly to the shell for execution without first verifying whether a matching executable can be found on the can be fixed at the same time as by making it so that jruby also searches the PATH before executing raising ErrorENOENT if no matching executable is found +Stop reading can be done by setting toleration length and quit without saving simply means that NN cannot start upFor edit log toleration admin could first set toleration length to zero +Bulk close for release +so looks like all the +maybe it is not complicated but i didnt get it right away as i hoped. +Actually the previous way did happen to work since wed take the short exception name and search the exception class names for it but I think comparing the full exception name makes more sense +Theres always room for more +This is broken now in stable branch +The description says that All files below this size are always included into a compactionI had assumed this setting is to quickly get rid of the smaller store files +Recommend you separately compute the messageId and messageSize and pass it into generatechange since v +this makes sense to me though any objections +It should be noted that the caller should never alter instances once they may have been used in a request +Added a warning message to not use and switch to the new filters +Marek says he has applied the workaround to the examples in that exception I said was happening I may have been on an altered config when adding the fview it all works perfect now +Could you document these new parameters and put in reasonable defaults and update the patchOne thing Im not certain of though is why we need to pass in the number of map tasks isnt that a config thing already when you setup Hadoop +But this is not what I really want +Well that was completely useless +Manually tested! +To continue it is incompatible given that has been declared Public +Added test case and committed to and trunk + resolves the problem +Closed upon release of Hadoop. +There wont be much delta and its probably easiest to merge with its important to lockdelete HEAD +Committed as r +Please let me know if this is not the case. +The coordinator itself is thus effectively a replica for the useful part of the operation so a single other machine should be adequate protectionUnfortunately we dont have a way to say store this replica anywhere but locally +Thanks Owen! +I have circled in red the links that are the sysadmins fixed it and there was much rejoicing +management server log and database dump +Robert so did that patch fix the issue? Did you have an occasion to test +Pushed and deployed backport. +Thanks Vinod. +Would it be possible to at least capture this NPE and return an empty result or call the error callback +The previous patch didnt mv the now variable into the lock fix it +example workflow succeeds with the patch +Indeed +If you are happy with the changes we are good to go. + +Attaching and with the following changesM srcadminguideM srcadminguideA srcadminguideI decided to give the new topic a more positive title Maintaining database integrity +However its way better to write doco while writing ocde for two reasons its fresh in your mind doco is a part of the codeTo be very clear its all about progress +Not sure whyWill be working on a quick pull request with a some documentary comments though +will hopefully not be too long since were likely to focus on upgrading Infinispan Hibernate Search and other libraries we wont want too many other featureschanges in there but this is one we can include then +Thank you Junping Du +I tested mvn gwtrun with Firefox on Linux and Mac as well as Firefox IE and IE on Windows with no issues +Sun has posted their evaluation on the bug above and accepted it as High priority +For I believe it is just trying to compute a value that is randomly within an interval of size randomNoise about defaultValue +Thanks for the confirmation +Cool Ill keep watching this JIRA +Could you perhaps attach a test project that works for you +Use that one +You can demonstrate this yourself by running our sample and add a like finder for Pet +Heres a patch which fixes the issue +Adrian what was the svn revision that the fix was comitted on and on what branch +I opened since I want to address changes in general. +In that case the application requires a or properties in order to create the table +Caching would help so maybe its not a big problem but if schemas are small and fast to read then it shouldnt be bad to put one at the start of each block +Without the HEADERREADME the site is more difficult to navigate for some users +bq +changed Container constructor to take individual parameters +For the bundles we already have an existing clean provided by the frameworkHowever for the the current is not very easy to use and doesnt reflect the config commands +It seems to appear both when you shut the db down and when you drop it +NET on this issueI think we need more interop test with +If it later makes sense to increase the default byteschecksum for DFS it may also make sense to increase the default for all other filesystems too continuing to use the same parameter or it may make sense to split this to a new parameter but for now I see no reason to add a new parameter +bq +In that issue he proposes a kill operation that will end the conversation at the beginning of the request rather than at the end. +Patch fixing this issue +This sounds great StefanAlthough I do not know whether the effort is worth it +It is the that we should investigateI just committed this to trunk and +applied the from resolved to closed +Id be happy to introduce in the future a log or externals role so external scripts dont need to have full admin credentials +Would you mind posting the relevant parts of the domain class as well +Forced encoding in writers validator and catalog resource loaderPlease review attached patch +This requires pluggable auth methods in my case adding support for SASL PLAIN as a substitute for SIMPLE +Thanks for making me a contributorNo +Rolls up both the other patches and removes the rest of the tabs from the java my test failures apparently related to this patch Im getting the same thing after rolling it all back +Ill fix +Whatever format people commit it will be fine. +If no master RS should have its watcher called +Looks like we patched this in parallel! You added some tests which is greatMy patch is a little different +To get the encrypted password please use class +closed +This patch includes the Axis environment import and the tls classloader fix as promised +min you may minimize the impact of shuffle seek overhead by limiting the amount of input data and choosing a small reducer number +I will asked id +deployed the rpm after this fix and did not see the above issue +Thanks to Michael Kurz for this patch +thanks for guidance on this +This will require that the user grant the createevent extended permission to Greenhouse. +Dave I assume you have seen my comment above yours right? How about looking at the sources yourself and provide us with a patch rather than moaning at us? This is open source right? Then just act accordingly +I have created tests for this and +Reopening to fix resolved field in JIRA +Are you going to submit it for review +Could you provide us a complete example? Have you got a chance to try the same example onthanks in the properties disappear then of course after you restart you dont have them +I think this is sufficiently fixed for now. +This needs more testing for. +Fixed as described evicts its cache entry if it cannot detect the reordering of the child node entries that took place +When I checked I observed was that header is altogether missing in the request +This is an interesting observationIt would help if you could provide some further details why did you mark this as clientembedded difference? do you have a repro that you can share? when you indicate you start a second instance of the program how do you do this from the same program or do you start a second jvm from a command line? how did you track the creation of the daemons +It completes in only minutes and has no permgen problems +This patch corrects s not complaining for me +For great justice make a superhero based on +I was pretty sure that the fix would resolve your issue but Im also glad to hear that it did. +These changes caused map reduce jobs to fail if the process that started the job exited +Who is the release engineer to talk to about CDN +Patch also updates tests with some style changes mostly relaxing private scope to default +Fixed in and released in +Fixed in git master +This really should be fixed +This for sure requires a new releaseversion bump of the parent pom we cant stay on forever which is used by branch too +push for maintenance sorry +Let me know if you find anything else before then +Uploading a patch that fixes this test break by fixing the test class path in jdbc +Very likelySo your Win and linux are on the same machineWhat linux distribution did you use? I got the exception on linux too +As that time to recover increases exposure to multiple HDDnode failures increases too +Assigning to Misty to get this identified as an update for EAP +in my apps there are many fields that are stored but not loaded for results displayHowever I do see the convenience of specifying just fields that you dont want to load rather than that you do +That allows users to override it specifically if they want to +Conditionally exclude this test on so we get a clean testsuite run in all cases. +First of all we need to think through what message commits mean when there is only replica in ISR +I was able to reproduce the little test you did with the groovy console +The core test failure is due to C compilation errors +Other developers may need this for local job submission +OK this error isnt simply schema validation +WDYT +Please file a separate jira on this if not filed already +After reverting the trunk patch passesIn the future please either run the whole test suite or let Hadoop QA finish the job +Done with committed revision Len for this updating +Using the above steps I see the same behaviour with datasource and connection factory without deploying anything in farm +Even more perplexing the issue does no appear if I invoke directly on my code and test data but does occur when using as viewlistshow function +has been marked as a duplicate of this bug +See +Merged to doc branch at revision Merged to doc branch at revision +The attached patch is a proposal for one way to modify our launcher to make it easier for people to create custom launchers +I have exactly the same problem which prevents to use Maven in our BrettThe ftp deployment works now but it is really slow +verified in + patch looks goodDoes trunk need this +It appears to not be related to this fix at all but I could not find another JIRA for it even though I found others where it also failed for them +I guess the vm must be silently swallowing +Screenshot attached +The case of a jsf bean using an explicit should be handled now but there is a more general resolution problem with web components not specified in the that needs to be addressed +on it + at +Im looking for a tool that can be used from the command line or web UI to generate the composite diagrams from Tuscanys representation of the composite model +Not sure why that should be but there it is +For the user this means for this feature Project Archives is now a dependency +ManikI worked up a new round of logos for Infinispan based on the our brief chat about going in the direction of an optical illusion or symbolism around me know if any of these are going in the right for V is still winning for me although Vb is pretty nice as well +Need to test that it doesnt break rolling restarts +How about start deprecating it in? If any new development based on results in unique features before next major release we will itThat sounds like a good plan by me + for the patch +Below is the information provided by AngelaPreferred userid or acymbalakFull name Angela CymbalakForwarding email address CLA is on request sent to root and created in +Are you sure you are looking at the right instance? Maybe you need to clear your browser cache +Im looking into the correct fixes for them +Yes I understand whats going on +could you check and maybe reinstall lzo? Thanks +Release has been cut off and the VOTE has been sent +Seems to work fine for me +Unfortunately it is not possible to handle it on servletDestroyed first and then on attributeRemoved because there is no way to know when the session object is being destroyed from the listener so the best bet is let the code in just one place. +recursionOrder attribute addedCorrection id of the adaptor is used not the last is the default +Please review and if looks good please commit +I would prefer to see someone fix the wiki page with the misleadinginformation and have it corrected to explain how to obtain and usethe python markdown package. +Will look around tooooooThanks for feedback +definitely +Agreed on all points. +My broken project +Synchronized blocks on the static df variable +Hi OlegHave you had a chance to look at this issue +Could you say exactly why the current way is bad and how it should be resolved +i bet this could be fixed by adding an extractTerms method to +From here on I can pick it up myself +Steven I think the IDEA configuration should use a project +It works for me + +Ultimately only final people will be using this soon +The fix looks good +Note that you will want to backport the corresponding docs change too +Isnt this one now a dup of? +Version has been releasedMoving all related issues to the next version. +Patch comitted in rThanks Robin for providing! The fact that your improvement is covered by tests moved your patch to the top of my personal queue. +Great +Your patch is now in CVS +I just committed this +Still need to fix up the build to pick up the new files +So have patience and cut the drama +batch transition resolvedfixed bugs to closedfixed +Added documentation to the clustersetup page +Closed all tickets +patch applied to all issues raised by me +adding marshalltrue to the vm url ensured they were part of the calculation. +Run for a couple weeks with no complaints. +I was thinking about it but the parsing code is so ugly thatI dont dare fiddling with it +Fixed by after release. +No integration tests +Thanks Gavin +Retry +Well youre telling me that you want to tell a user that there is no matching in this case that its a more intuitive solution +With permission of my employer I provide this patch with license to the ASF +bq +But this does seem like a lot of work +Have tried with Pluging Plugin still no luck +The value of name passed back to is the +Also marking this as against +Ive addressed in the attached patchfor we are not eating the +Yes this is clearly a typo +The console is an addition not THE central entry point +I was able to locate the problem +is NOT affected +This patch does not allow the deletion of the default project goups build definitions definitions in project project info will create an improvement to allow edit the default build definitions in project group +Verfied in release closing issue. +Also the properties file will change to an XML format as the +Please check that it was applied as expected +Proposed patch to set the thread context classloader on thread creation +Closing out +I dont know if it is possible to migrate issues across jira installs but should probably be STS +Remark the selected rows are only restored when theres no setWith a you have to this on your own +That is as opposed to the current approach of immediately instantiating the proper type at parse time and cloning those instances later when the area tree is builtI think implementing this wouldcould get fairly involved +Three events are dispatched to make this happen a COMPLETEDCONTAINERS event is handled by the which dispatches a KILLCONTAINER event for each container to be killed which the handle by dispatching CLEANUPCONTAINER events which are finally handled by the which tries to kills the containersDoes it make more sense to use this chain events or to try to call the kill code directly? For the former the issue would be how do we know when the cleanup has been completed? It looks like have their state changed when their containers are killed so the shutdown code could monitor them until they all reach the correct state but a fair bit of plumbing would be required for the shutdown code to be able to get to them +I referenced the wrong issue its not +This patch also fixes that jira +Hi Stepan Verified +One more iteration to deal with the fact that Windows does not contain the killall command +for the new path +See revision +Later when the files lease expires NN will close the file and those blocks will be finalizedThe idea is to start from the same state when it was down +Also need some coordination around splitting and compaction +New patch reverted all files with changes +submitted patch by Thorsten Berger. +Do you have some link or reference to learn about this theme +Though its the st endpoint that creates the queue that determines the queue size +I agree the two compilers should be doing the same tihng +Example project that uses the sfileupload tag inside an Tab +It would be nice to help with this but I dont think Ill be able to +This issue has been fixed in the PHP implementation feel free to close the issue ifonce its done in the Java in Java by Kevin Brown. +Thanks Andy +So it should just work to fetch N rows and have the loop check for N and just return from the middle of the loop and leave the scan to get the rest of the rows on the next scan +HiCan this issue be added into EJB RC please? I did not see this issue listed on RC RC now supports use of the Properties Service for deployment descriptors this provides this requested functionality +committed this +Will commit if tests pass +Looks like a trivial safe change +Changes committed with commit revision to update and system properties +We need to change the serialization API a little to support to read and store the case flag +Then Ill add the hack to the app +moving to installer project +There is not any record in the log back to version makes everything work again +Sanjay Test cases still fails can you please have a look +ill verify this once is out +Feel free to reopen if you dont think they are duplicates +In particular the test case youve provided is already covered by the test Ive mentioned already as well as the unit test test +Peripheral code could convert a Map to a Mapstack and do whatever magic it wants to do with it +The question is should this be a default behavior? When I accidently shaded bouncy castle boy did I get an unpleasant surprise +The unit tests pass +If its OK with you Id like to tackle that review comment in a separate JIRA +can you check if existing VM can access outside +now only displays published widgetsShould the users page also check on status after they have been added +If this is a new feature it should really go in the release +A typical facelet page in my typical facelet taglib file that lists my facelet which includes a reference to the better to move for move from M to visual part better visual rendering was added but ca not finished yet +Sounds good +A preliminary implementation of the algorithm to provide version consistent IP addresses. +Verified on the metamer application that it is ok now. +Heres how I see the line aggregators working +Vitali please fix this issue in And why is this problem suddenly appearing in? what changed +Ill default it to UTC though +He acknoledged the bug and linked to a sitethat claimed there was no fix prior to me sending him the fix +Moved the working directory for the tests to target +I have added a patch to set an upper limit +I ran system tests but they failed and all the kafka server logs were empty +Patch is attached +I will be reviewing the patch for but also wanted to know if a different patch was coming for this issue or is it just waiting on +This is an old version +It does not invalidate the case for an atomic renameAt some point we may fix the problem of reader leases +Thanks at Karl Baum your monkey patch seems to get me running +Many thanks for this enhancement +Hack to pump EOF from shellIn to process Tim and Pierre +You can keep on submitting patches on the plugin and if the code is the way we like it and youre showing comittment to maintaining the plugin we might give you commit +Trunk patch at revision Patch not good need to rework begins to be old any update? Shoud we not close +When Ive ran a new ATS Ive found that all previously cached objects are +Please provide steps how to verify this Jira or close it. +The class has not been added to SVN along with other changes +we have fixed some similar in meantime pherhaps this had also affected this the faulure still occur in railo please attach a small example info here +I havent seen this problem recently +has been marked as a duplicate of this bug +DRT passes +I also added some checking in the code to issue a warning if the name specified in the annotation doesnt match the property name and sparse is not set to true +bind methods but some might be inaccessible due to missing dynamic imports but the next time they are looked up the dynamic import was satisfied so the method became available +So all issues not directly assigned to an AS release are being closed +As is I think its better than it was before +So maybe mark the public createMarker as deprecated and move it to in the next major release +I created the which encapsulates almost common task for writer +Committed the patch to trunk andThank you Stephen! +are the patches reversed or did I wake up too fuzzy today +Patch applied at and . +Thanks for the original patch +The patchPlease apply from Windows platform +Im personally worried about anything that starts to pollute stream definitions with deployment metadata +The same question applies is it ok for a network configuration call to span various seconds or do we want to add some parameters to it to force extensive SRS lookups +Approved for. +Tried with JNA and without +Just let me knowRegarding the importance of the different sizes I think the next step is to instrument the CRC to log every invocation to a file and run some wordcounts or something +There has been no activity on this issue for years +Added a patch for the situation including IT and documentation changes +However +I committed the patch proposed with the test done +bq +Hi Dennis whats the next steps here? Is there anything you need help withA user was asking where to file an issue and I wasnt sure whether to point to the old tracker or wait +patch to use a instead of adding directly to the writer +Mark they would get large fast but i think we only need say Ah right you mention that above in the summarybq +Matteo Want to update the patch as Andrew suggests? Thanks +The reason I did this was that I could not mix xsall with xssequence from which is a superclass of +Posting such a huge patch makes review code reuse etc +The actual issue is caused by the fact that the proxy passes as is to the provider which confuses it another JIRA will be created please link this new JIRA issue when createdThanks +SOLR suggested improvements have been moved to +CP fixed in release fixed in release +Moving all issues to the CR +Keeping the existing approach for the efficiency reasons described above. +Hence closing the issue +should cover this for the most part +The level of reuse tends to be when a whole page of data becomes deletedThis is not true +Beware all ye who venture here the sea of XSLT is deep and wide and unforgivingThe story goes something like this the generated ids for topic references were not being retained and new unrelated generated ids were being created for the same topics as they were inserted into the final document at a later stage of processing +I agree that a behavioral change like this seems on balance to be a good choice but I agree that it can wait for +What is the timescale for? Around the same time as the Spring final release +If Axiom worked correctly this would no longer be required +We definitely need to sort that out before considering putting this inCan you try running a test with concurrent clients each reading K chunks with the pread API? this would simulate an workload +This is similar to character columns coming from system schema vs user schema +The temp file for the still exists +it is actually is only for apache infrastructureFixed in svn +Needed time to integrate fix into Apache CS +For consistency modified the newcustomer form buttons though it is outside of the Profile feature since they serve the same functionI did test viewprofile without authview and I did not see an impactIll submit a new issue as an improvementTom +This patch resolves the described problem the unmarshalling now respects the whitespace preserve setting either set via the Unmarshaller a DTD xmlspace test cases will all run sucessfully is either one patch file for both files together or one patch file each in the ZIP +The attached patch implements this feature +Thanks for investigating these issues in such detail +probably only a unit test configuration problem +Thank you. +After applying this PR parent issue will be to master +Would you please try the patch? Thanks very much +Assigning to Yongqin to verify +Descoping from +Im more used to using factory methods or builders for such havent seen this chained construction until now +not all the combinations are supported +You are right getterssetters should be prevented but what about BindableVetoables? We should not allow the combination + +Looks good to me. +It is not applicable but I merged the changes by hand. +Hopefully this isnt seen anywhere else +The test be fixed now in SVN trunkCould you give it a try +This must have been fixed between the time it was raised and the current revision + patch to fix headers +Ive fixed error in assemble target for tests and restarted the build. +Btw esb is deployed to the all transport has been rewritten +If its not null thats probably a bug +Fixed +Done +I committed this to trunk andThe same problem exists in but the patch is not applying directly neither twoKihwal would like to make a patch for? +about that git hadnt pulled all of the latest changes for some reason +Look at this and suggest me any required improvementsThanks do have an IBM jvm but Ive not been following your code coverage work closely and Im not sure what I can do to help +Also theres no on for any changes to the menu +JSF takes in consideration the label atribute of the validated component +Also we can warn about this while falling back to old reader +Hi ArjenIs there any way you can give any guidancehelp around how to get around this problem or ifwhen this will be fixedWe really appreciate all the hard work that goes into this well take a look at this for +Closing all resolved tickets from or older +The new Patch is ok now +Good news Lukasz if needing me help deeply pl +If deploy only works with sshkey can we set features to an empty dictIndeed we also have eximportkeypair calls in the driver I will add the importbykeymaterial but I dont think its very intuitive +You should consider to give us a clue about which plugin you are talking about +to make sure that we dont forget about this one +jlindqui Domai Feb jlindqui Domain Users Apr +There are failures using Oracle JDBC because of is set to true in etc +this should be easily fixed setting additivity to false in configuration of the security classifier in the logj conf +More to come very soon +Fixer is unhappy with level of stress testing but increasing those levels should be in another task. +Moving out of +My comment Both appear to be flaky and unrelated to the change +Here is the +bq +Since value y is not allowed the commit is rejected and dialog appears prompting user to input a correct value +pending JenkinsChoosing a k ms timeout is a little odd in that its not a round number of minutes but it seems fine to me +Unfortunately Continuum seems to be broken at present no builds have run since July meant Continuum itself was brokenHowever it now seems to be OK again. +I will check this later thanks for the reminder +It seems to work but maybe I overlooked some things I dont know ofI included a basic test for the post method +That would keep things +Did a basic test and it works great +Can you check if it still occurs with version? +All requests succeded and data was readable after that +After investigation there was too much risk associated with the change including the possibility of effecting current behaviors to be addressed in a point release +It looks like we do not record it in journal as we run execution +we should address the real issue insteadDo we have a plan or even a vague idea of how and when we are going to solve that? In fact I do have some ideas in NGP for solving that but they are way off in the futureIm all for going for the root cause but who will do it +No big deal though +and have patch to address this problem for trunk +bq +Adjusting urgency and priority to the standard levels +This behaves thesame as before +I committed the patch. +I modified the files srcadminguide and srcadminguide to document statement caching I hope correctly +We really have no idea here +Hopefully that error message is informative as well +Can you also add one or few tests for sort merge join +Needs Forrest doc Javadoc and a sample but the code is in place. +Fixed +lets move on with this issue then Ill create a separate JIRA for the tests and well copy them from svn history +I just integrated the v patch into and trunk branch +Does no one actually use this feature of cargo that it has gone this long without so much as an additional comment +This is really an application server level issue not specific to the ESB and material that briefly covers this has been introduced into the GSG in Suggest removing that section from the product ESB Admin Guide +Maybe a misunderstanding I dont create a new process instance I really change the existing one to point to the new process definition +can also be added for the affects version attribute +aMaybe it makes sense to save when a Throwable is encountered +Hi DanI believe the way how dependencytrue works is that only consider the bundles in same feature not include the feature dependency so if no other bundles in feature itself explicitly import package than its not installedFreemanInteresting +Uploaded failed tests after applying +I agree with Bela here +Those two are supposed to be changed at the same time. +But the data pointer of this breakpoint reference is kept the same +It doesnt fix the problems Dougs having +Thanks Hitesh! +Hi is the tag for Common available to be included in the EAP +Cannot reproduce DRLVM crash on rev +Thanks Arpit for the patch here! Interesting approach overall +The second attachment contains the first page of the converted tif +Would you mind submitting a? Having problems reproducing this on the latest from I have tried this on and +pending JenkinsColin could you please set the affectstargets versions appropriately? Thanks +Thanks MaximTried and installed ffmpeg etc again but still the same issue +We can definitely print on the CLI that apps may get stuck after this so that we suggest users to use yarn application kill in those corner cases +Not sure why I bump into it only in the testsI tried to copy into collection in but that did not help either +There are still a few things left to do update all poms to use the tfs provider create a site documentation and update site update the scm provider matrix in our is expected to be released? Our provider contribution will be included in that release and we would like to tell our customers to watch our for the release +Im not in favour of adding another component when we can easily make this a simple style addition to an existing componentMostly because we end up duplicating a lot of code unnecessarily +Id vote for the synchronized map +Fixed in the issue now that it has been resolved. +It sounds like a fine idea +Please use a seperate src dir for your autogenerated stuff also alot more cleaner if you have a feature that only one person has requested and you have doubts about it and you wont do it soon it is better to just reject it +Move minor issue to a future release. +Sorry I forgot the tests +Rosta please verify and close this issue +Will do so log message to identify controller with more than one path and also allowed process to continue in Git ID efeeddbedcce. +I will comment with another update later today or tomorrow morning and might have an ETA at that time +Suggested drts and bvts passed locally with this fix +I took a quick look at the xml grammars which Andrew mentioned +We are going to use for that purpose. +Hi can you please attach all relevant files rather than inlining them and also attach your domain classes as otherwise I wont be able to run your test +to execute it Run as Test +Cool +Thanks JonYes +Consider a case where after running the cluster for long we realize that we need to change the slot ratio cluster wide +and thank you for doing so +Half java half script +Hi thanks for noticingThe problem was not url the problem was the null +Not sure if that is done by an external system querying jira or some type of jira hook +This is a trivial zebra jar version number change intended for APACHE trunk +Failure analysis is progress +There are still mismatches currently jbossweb has version in src and shows in doc +Henri I think we can close this one I have quite recently added a few servers +Why remove the installation of the headers? Otherwise looks good +bq +Hi GrzegorzIve tried the classlib one it should work now +. +Having consistency is more important and either one of these patches will improve that +Note that AFAIK using this feature requires or later at runtime. +The theory here being that the multithreadedness is causing seek contention when loading the sstables +Implementing the interface is not for the pure mathematical elegance we do it only if it serves some purpose. +should be private since it concerns a private repoHowever skimming messages sent to committers in my archives I didnt see any that could not be public +Thanks John for testing it out +the patch looks good to me +If we want to do our own thing then we should just come up with our own API and not use the package namespace +Im now getting a on target rate or better +The worry is if any of the qpid or boost libraries themselves link against the static versions of the runtime libraries so that two runtimes run simultaneously. +Please modify the attached quickstart that it shows the problemThe related code is in +I have added this in my workspace and it appears to work but there are still some issues I would like to iron out +change a better it to blocker since lots of package service tests will unless it is suggestion Johnny! Care to provide a patch +The only thing it has to do is remember this autonomous decision so it can report it to the coordinator is all documented at length in various standards and books +This looks really cool but seems a bit specialized for inclusion in the platform itself +The guide was checked +Thanks for suggestions +I could not reproduce +Actually we were only talking about exporting here for its the job of the JCR repository to take care of consistency during import so we dont have to care about itHowever your scenario would not work because would prevent the user from saving modified and call an error +get the Connection off the statement +Havent done the precommit thing yet since Im not ready quite yet so Im sure there are a bunch of things like javadocsFunny you should talk about stressing it I spent most of the day working on exactly that see below +Uses the resolved hostname of the ipc connection and rejects unresolved hosts + reviewed file +Thanks Aleksandr +Richard could you please try this again from trunkLet me know if it works with the fix +depends on so it cant be moved in its entirety +The footer has odd size of height because the font on the demo page is Verdana not Arial +Should we create another ticket for isolation with timestamps? From a user point of view the fact that timestamps are even involved in establishing isolation seems like an implementation detail at best +The component tests double the build time for that repository so I made the tests conditional on the run profilemvn clean installThanks very much for the contribution Mario! +Another little glitch i just noticed is that aparently with the new jetty configs JSP support isnt enabled? loading httplocalhostsolr works fine but httplocalhostsolradmin gives a error JSP support not configuredYepp just created for that +Diffs against Subversion HEAD are preferred. +Committed revision Committed revision +I mean next brew build if you decide that this needs to be fixed then CR otherwise next what is the status on this? Is this change in CR? I didnt do Seam build and it is minor issue +fixed by using ntaddress as ntbase replacement since it does not allow childnodes and has no mandatory or autocreated propertiesCommitted revision . +Even if one would be useful the lack of it isnt a bug +until we decide on desired behavior +Documented as a fixed issue in the EAP and release notesApache HTTP Client has been upgraded to version to enable corrent handling for let me know if this is not the official version number +HDFS no longer seems to need slfj and we can explicitly filter the slfj jars from the mapreduce project +The src folders will be enabled on the profile that gets built +Should the Java facet version be set to the same as the runtime or the workspace Java version? And what happens if the two arent in sync +Updates to the manual has been provided by Milosz Tylenda and committed under following revisionstrunk r r +Surprises are not ideal +test in revision Vladimir please check that it fully resolves the thanks +but checked the output and the code to confirm +Scheduling for so this gets looked at this work right for the WFS operation? If you havent tested could you test it there too? So we can be sure its just the +I suggested to get rid of the refresh link and to have the table listen for new projects noted and commented in other jira +RahulGuillaume has changed the karaf build so it no longer requires the sources +Ive hit some challenges getting the full test suite working going to post about it now on the mail list +See for more contents area for the section headings in the no longer match the sectionssince jdk was added as in the middleIts also a little strange that the optional jdk was added as section in the middle before the required sections of jdk extensions and junit +Actually just tested and its a problem when you take down the lead VM as well +You would simply point the script to the right dojo root +Performance issues are very hard to test in CI +If it works but having all of npm proxied is inconvenient we can looking at some way to override that with a proxy +Attaching files for reproduction +Sounds fine to me +This fix has been committed to trunk and x +Ive sent a test jar to Lin to see if this works for Oracle +I get this with Java but not in Java +are you using eclipse +Looks like the row cache is returning a column family without any columns during the second pass +Thanks Mirdul committed in trunk r and r +Maybe the expiration time in the hash is enoughRe would it make sense to store the tokens in the repository? This would fix the cluster issue would it create othersRe would that be a single custom attribute? with a customizable nameRe perfectAgain the proposed bundle to integrate the suggestions +Do the two projects point to different Nexuses? It is possible that one of your Nexus has a pointer to a repo that still has the wrong version +The last thing left to do is documentation which Ill get done as soon as I understand how this works +Two extra logs at debug level no +Committed to and to trunk. +Or we can just attach the exitStatus to the end of app file name so that we can avoid more IO operations +In the generated project the datasource Ant target was copying the file with overwritetrue +no need to add to folder. +This is not ready for prime time yet but Id like to solicit comments for the general idea +This issue was resolved in revision +Ted would be nice if you committed class with its dependencies form I cant seem to find it in the trunk +If resolved before is released please assign to +oops! Managed Reservations improperly killed by health monitor +Id have to check if it is passed to in some special way to also affect the method level there but I doubt it +All other are okay except for and +Now I see that bug is not a duplicate of this bug +This should be closed as Duplicate instead. +If not we put a note in the code so that people wont get confused when adding new types of requests in the future +The bug probably only surfaces with inputs +OopsRemoved comment about running of course I installed the branch and that is now running on fine +Very slow +Is there a way we can keep a sanity check here? was not so long ago +Ok Ill take a look at the code +With all of these cases you also need to ask if there really are any valid use cases for the collection in this case +The fix in will fix this issue +It works for version of the plugin do you use +Please review and provide your suggestioncomments +we now display a friendly error message when this happens and proceed with the invitation. +Im not sure if you are voting for this to be committed or for it to be solved following Mahadevs suggestion +The default buffer size is bytes so sending a stream greater than bytes results in or more additional packets compared to when tight encoding is disabledIf extra packets are generated the total bytes sent when tight encoding is enabled becomes larger than not using tight encoding +This is documented functionality which is just not there +Hi ErandaIf you run the reproduction script with does the problem still occur +See CE thread on Generic +Theres a plugin for this in the registry. +Hi PatrickI just checked the online AJAX Dynamic Select example using IE and the AJAX part does work +you confirm +I will update the Envelope diagram in the docs to happily show two arrows. +Oops my bad + +Can you test this with upstream? I think there have been some entity bean concurrency fixes since +Still Axis is based on jdk +is already generic lets put it in sayAlso we need to point to this for completed containers +You must be looking at something different than what I have +Can you attach your eclipseRAD configurationGo to the same place where you grabbed your feature info and click on Configuration +This list sounds about right to my mind also +Read documentation +I will post an RC build shortly and you can try it +Looks good more clear than mineThere is one error however +Minor changes were needed to the optional join logic. +Diff for changes outlined +Im going to do a full review soon and Ill report back +failEmbeddedLocalptestReplicationLocalpbigInsertimmediateI have zipped up the entire fail directory available upon request +committed +so when i do unapply both rules sould get unapply soremoving all adjusment created becauseof above promotionMy question on unapplying promo code both rules dnt get unapply +This should be a custom whatever is the right thingThis would have to be above the distributed broker +Or let us know if you only hit this with the Fedora RPM AS installation +You dont need to keep Gb of logs you can perfectly decide to keep the log maximum size to say MbAll in all the exception that bubbles up is a message which is logged +Did you configure that yourself? How did it get through the validationIll add a guard to the code so it just doesnt schedule +Latest set of test set of test reports with the SYNONYM issue fixed in the testing jar fromShows errors and failures +Thanks Chris! +What about the fix for the current versionThe patch provided is very simple and causes no problems if a project description parameter is not it would help get this fix committed I might also be able to contribute a simple unit test patches to documentation etc +tablecould you please consider parse this syntax +I also added a small test case to check for this in the test +Thanks for the patch +This has been done please evaluate +I expect Ill be ready to post it here within the next week or two +Here we use an s metadata to avoid the MVCC altogether. +So should be goodMaybe than Append can be simplified to only implement Row and not also Writable +The current code only assumes one popover is displayed at a time and we access it and the callbackId via +Thanks +we need a test case to see if this occurs in or current trunk +Ive looked at it some more and bundles are rewrapped but thats not going to help because the wrap method is invoked on the original item not on the bundle +Please apply patch +If you have security concerns you should instead write your own entity resolver so that you can approvedenyredirect loading of external documents according to your own policy +Ok fixed now +Patrik is this still a problem? Have you tried it on the more recent greclipse for eclipse +Thanks Nathan +I dont think is resolved what has changed +Furthermore since it can have a tiny heap and few threads the fork will be much cheaper +Duplicate of +This should be fixed in meant replacing by above of course. +Thanks Prajakta! +patch applied. +Its really nice to see such quick response +Feel free to reassign this to a specific release but only if you have the time to do that work is underway with AS all previous community releases are +I looked at the patch too fast and saw a in it so I figured you were controlling those but I see upon looking with my brain cables properly connected that it wasnt that +A little bit of manual but Aslak is writing a ruby script to smooth a little of it. +Yes I think adding the config options to the Main Servlet makes sense or creating a new file upload handler component which is used by the main servlet +Creating a new module is another way to approach the problem but it seems like overkill to me +I think the best approach is to suppress the isPresent bit stream in the case that the entire column is present for the stripe +Its currently where we hurt badly and one of the things hopefully this patch can address +A simple message should be enoughI will review and comment the patch and ASAP +Generated and checked in Sybase script +Progress reportThings already done INSERT DELETE except DELETE FROM DROP ENABLE DISABLE SELECTThings to to CREATE ALTER Various formatting methods in SELECT +Will commit today or tomorrow +Attached two logs of a good run and a bad run of the gist containing the steps to reproduce with cranked up to TRACE +This patch fixed previously failed test case +We may want to add a signature table as well and way to verifyintegrity and whether any rows from the actual audit were deleted +I dont have a very strong opinion on this so whatever you guys decide is fine with meI think if it does evolve as part of Tez at least some changes are likely to sneak into Tez branch without going to trunk so they might diverge for a while but if we are ok to take that chance thats off full unit tests now +New patch where the ppr is controlled by a boolean parameter +Patch applied to r +The shared method spends half its time checking if message is null so youre not really extracting out common code at allAlso when you upload these changes could you put them in reviewboard +I will change the name of this issue to something else without the word in it so you dont think this has to be some form of gzip but instead is about alternate storage options that will benefit particular use cases +We actually dont use the tests +the issue is still open so what made you think it is fixed inI got no answer nor comment from people in charge and the issue has not been assignedSo I got no way to know if the issue has been reviewed and considered to be a real one by the teamAnyway im happy to finally get an answer from you. +The official jboss shutdown script both use JMX to terminate the server +The actual performance diff probably does not matter but for locking would benice if we could get the optimizer to pick the unique key index in this case +Now the distribution references one or more features and defines the assembly part +I went through the linked issue and the revisions mentioned there +Patch posted july is a partial implementation there is a bug with boolean fields and variable length fields but they do have partial implementations +Committed +Heh Stack and I were saying its fine for +another option would be vv with v for being verbose and vv for being more verbose. +The patch looks fine to me +Is there another way? this is really not would be nice woudl be if the parser could punt to the semantic compiler in place +Same problem on OSX using for all eclipse groovy componentsDebugging works in that it still runs the script but it behaves the same as a run rather than a debug +Ill revert that change +Verified withI have refactored the date control forThis bug doesnt appear on my local snapshot I have to resolve one issue on the facelets support code until I can commit +Hello GeorgeThe fix looks good thanksBest regards by Andrew. +Im marking this as resolved as Rajini applied the patch +DNS setup project created cvs hey +Done. +Any movement on this issue +Without that evidence though who knows anything +Committed to trunk. + +Also polished the files of the examples. +I checked the patch and committed some changes +The exit condition should be i +I am investigating +This could not be reproduced with more recent hive +I know the exclusions thing can be a you aware that much of javaconfig has been moved into Spring core for? You may not be in a situation where you can use the Spring milestones but I thought Id ask to make sure you know. +Whats the problem with not guaranteeing that wBuf data is sent on to the underlying transport without an explicit flush +Alan I cant seem to find the test which was failing for the old JDO upgradeThere was some problem when we upgraded last time + +Hi Claus Jon I have added the Camel Printer page to the documentation under Components in the Documentation WikiI do not however know how to add the Component to the list of components on the Components wiki page +Create a new Seam action via New Seam Action wizard +See these steps +Whole index replication requires a hard commit first +Making the attributes argument would have been better +That way people will feel free to extend it to hold all sorts of data in time +Yes I agree +Canceling patch I didnt look at the messages printed from the Strings in the class +Sorry edited the wrong bug closing it again +Im churning the change to branch as well +I have large number of snapshots and I would like to really delete the snapshots from the database to use less space on disk +For functionalisms Id prefer things to be simple and general rather than typedspecific and fast to enable any kind of object to be treated in a similary understand the rationale behind this and agree when youve got a function that can take multiple input that does not share an interface +Hi DagHow about this text for Invalid setting of the property +Please try this patch +The recommendation that Chris mentioned works for me to update BLOB dataThank you all +I think we should allow the sorting to be pluggable if possible +This will be fixed in the final release. +LGTM to merge. +of course the just have to change the build script but such a chnage should not be needed between patch updates. +Thanks +Murmur Murmur backwards this fix been finalized +The hbase jar changes are uncommitted leftovers +does here +I think option is more favorable +If the shuffle is based on networkshould the reduce start before mapDid affect performance +Thanks Elliot! +Micha is right +and trunk didnt need changes. +My guess is that for some reason truezip wishes to create a but that fails because its constructor has some assumption of swingawt being present question thus is why it even tries to get thatAnd what is so special about creating a normal jar that needs key management +Bear in mind that youve got leniency turned on +Because they assume theres a built AS in buildtarget which is not true in case of running against supplied ASEAP using +seems tabPanel issue +Also defining an absolute expression should also work +Jinghaizh labels for party pos and product components are in trunk at revision +sha on and in the latest code it no longer occurs +I will check the patch more carefully and apply it if it is OK + in TRACE modeMinor difference with described scenarioThere is a java consumer on DLQ that redispatches the messages to a queue +Thanks Marc +Thanks for the review Josh. +SaschaCould you confirm Erwan pleasebased on the screenshot from Sascha Im closing this issue +done in staging site +Cyriaque have you tried declaring the serializer in your projects file? I am not sure which one takes precedence +trunk +Covering CLI first would be better +No progress has been made +Im not sure how reliable the test is as it probably relies on the browser reusing the connection +Mark do you want to create a patch to the spec for this +If you find a bad repo please open a jira with a specific problem. +Added a test case which makes the error appear in current implementation and fixed the bug +Can this issue be closed as fixed +I am unable to reproduce this failure +Thanks Uma +use old cglib withLeonidas +Configuration override is still absolutely the way to go +Also I am very nervous about the possibilities of deadlocks if the can block +I think that OWB behaves correctly +Actually they are not specific to Spring MVC I am looking for a place to put some code that is executed when the application starts +If there are lots of apps in the apps table it will try to fetch some of them using AJAX +However this wont work in the future where we might not be able to enumerate all possible from the client when we first get the DT eg in a zookeeper based setup the client may only know the address of the current activeA question I need to look into when the JT renews delegation tokens on behalf of the job does it pass the jobs jobconf or the jobtrackers local conf? If the latter it may not have a local configuration that has the same clusters configured which would be problematicI also need to test this on an actual secure cluster with MR to see if it really works in practice +Fixed by upgrading from ConnectorJ to +SamTodd can you pl comment on whether this bug exists in Hadoop trunk +We would really like it to be fixed as it is the only reason why we probably wont use Sonar despite its useful version is the fix scheduled for +I am still working on a systematic regression test for this behavior but Im because of Java One this week +Will reopen if seen again +You should enable that. +resync with current trunk. +I tried the original code again with spring security and it works now +Checking for missing output files would require keeping a list from the previous run +Also the issues with combiner runs zero or more than once has been taken care +I am working on a patch to solve this bug however i am unsure if we need to change code in or classes? Does anyone have any guidance about what should be done to implement changes suggested in previous comments +It really depends on how you define real world size +If we think we might want to support UDP in the future but not as an immediate priority we should probably get the implementation in now and add the configurability later? Just to make sure the work to portrefactor doesnt go to a waste only to be done again laterEither way the committers should probably make a decision here. +Account created +Let me reopen it to address Niklass idea +Its limited to and +Sounds good to propose a defined resource lookup of no other resolved issues for already released versions +Thanks Steve +Duplicate target buildmagicclean in toolsetcbuildmagic tasks since we are not going to produce another release from that branch. +It also did not clean up the targetclasses and target directories which are just required temporarily by the assembly to do filtering and copy filtered resources +Attachment has been added with description initial has been added with description an example has been added with description alternative corrected and updated version of the proposed patch +sorry i got confused amongst the different issues +Note to test this patch the example doc Im attaching to your solr instance using the mods to the example schema Ive attached +Thanks Rick and LilyCommitted patch to documentation trunk at revision Will do another patch for branch just in case updating the copyright year only of course not the release id +Please read the scenario asL contains only BK +In revision I simplified the EBA code that deals with deciding which modulebundle should be refreshed +But the eval echo +To give you an idea of my app size I have Seam components of which are actual +Looks good now and will be in the next update +Thanks +This diff should be used to correct the file that got into Beehive +Please verify whether it is what you expected and close the issue +Each of the old project is contained within a folder in this new examples project + the change looks good +Sorry for delay this code was refactored so the pb is not exist anymore +Wikia folks are seeing this pheomeon in their cluster +The problem with this patch is that center tag should center the content with or without anderlying css +VM created can you upload your pub key for ssh to reposinfrainfrastructuretrunksshkeyspeople then I can give you access +wrong t this fixed in + +Thanks for the observation +The part of the failure looks as follows java langtrue Start predicates a ERROR The exception Column not found + Username prudrakshala +Arent we only interested in the regionserver carrying +TC is not working as expected need to be improved patch havent found anything apart of test failing on derby +Thanks Gary. +Ill comment back with what I find +Hm then this still seems a little weird + +properties files to +Im retracting this againP hehIm hoping to take this in a different direction that wont involve WS! +Thanks Liang and Nathan! +No Im just trying to understand why the HTTP transport should have a default behavior that violates the HTTP specs +Test code added as part of code added as part of +I was going to look into this for jirb +Glad to help +dims + +I had merged the patches and added some extra testsOne observation When marshalling into DOM node one test fails preciesely test +Attachment has been added with description Patch for majority of transform issues with patterns has been added with description Update that support patterns with overflowvisible +A patch was submitted inTherefore I consider this as rslvd! +The fix made by has now been applied so is not necessary +Fixed git prefix +this should be called transformersOnce completed please assign this to Nick Boldt to get the proper requirements into the build +I can check this is the right solution +Verified in JBT SOA Tooling site from build H and JBDS SOA Tooling site from build H. +Ready for review +Its much more likely to actually make it into that release if you try your handat a patch +See for listeners signaturesBoth attributes are of type +So good just as you saidLooking forward to the next version for this +FWIW it wasnt the second time +has always been broken this way it is broken by designcould somebody have used the broken behavior and relies on it? Just making sure its not a bw break somehow which we should document +Im investigating more deeply now +Could be done via a common utility function instead of sharing the job conf. +Closing since issue was resolved over a year ago. +changed to git to this ready for review? If so please change the status to Patch Available +Added the amq +Drop use of +We balked at having to provide an array of Column when only one element is expected +I already have some documentation but need to update it before I share it +Is a warning for every request a requirement +Could you create another jira to track this +Currently it needs to added by hand +Good thinking on not setting it to save an extra field +weeks +r with additional adjustments by me in rev +I merged the patch to. + +Attribute true +rPostgres adapter overrides the alter statement to resolve +Currently we have a type byte in each KV +Lets just be careful going forward to ensure that the tracing library can easily be broken out when necessary +No looks like this has been fixed in another JIRA +Thanks for the effort +For some reason I wasnt previously +I tried reproducing it again by applying the FXG component via CSS and it worked this time +diff u srcdocumentationcontentxdocs srcdocumentationcontentxdocs is the machine name we should use since it is the friendly that is now fixed in the source. +we need another GS release +The parser looks for functions registered using registerFunction to retrieve the funcSpec +Ok? Thanks +it works as expected thanks! +Hi MatthiasI see youre using +weird +Sorry accidentally hit return and submitted this at the code it would seem that text has never workedits not listed among the attributes for this theres no setter method set +We are rolling the WAL periodically but if we do not flush we may end up with a lof of hlogs to recover from in case of failover +Hopefully Ill be able to attach the MB +Any working solution would be appreciated +they cannot accidentally be altered by concurrent executions +Yeah I have found this really annoying in the past too +Hmm ok +Yes it should +Im a little confused about how testing works now + +When youre using a continuations based flow execution storage you actually have multiple physical copies of the same logical flow execution +on commitI see now that the effect is the sameIn ORH wed run the disabling code regardless whether we deleted znode or not and whether region in RIT or not +There are a few things that could be done much better +Vikas use the mailing for help with Camel +Was fixed at the beginning of November +Attach patch merged with latest trunkThis jira is marked to be in +I would like to only allow it to return so that we can have better assertions +Images which illustrates how tags could be drawn below token annotations +For the best experience with all of this I would recommend picking a direction and making sure its the right one before charting a course to it +though is expected to work almost completelyLooking at the patch file i noticed you also proposed some transitive changes like converting pure library projects to osgi bundles +the script of is not persisted and we will decide in what we are going to do with this in the context of a reworkexpanding of the node types +I know this is probably the last thing you want to look at after you get back from your vacation +Please reopen if the issue still exists +I ran the test suite against a local server and could not reproduce the problem +Updated patch based on user comments discussion + contains an array of my business managedBean contains an array of my business of a bulk update all resolutions changed to done please review history to original resolution type +The problem is that the latest version of spring in their repository is whereas the latest version in the maven repository is +On Lars suggests using two filtersFirst changes with a certain naming pattern into request headers +Thanks for opening this Jira MahadevI havent looked at the code does this happen for both C and Java clients? +Ant task would setup the environment for them so I wouldnt imagine people having a dependency on the actual script +Adding a conformance check is nice +Sorry for the silence from my side I was on a business trip in US for weeks +Should be moved to the MNG project as JEE stuff needs to be put in the ASF sandbox +Patch from Jaya is applied. +Committed revision +Look current maven trunk its possible to build it with mvn andor if I had this project as a m project I cant add a new build definition with a ant think some projects have different build mode ant maven maven and want to maintain this +Either way it is my faultThe test that caught this too is +crash server + for the latest patch +The first cell you sent to got the current time as the timestampCan you please confirm what was your expectation +Thejas will be helping you with this one reach out to him if need have modified the code according to Thejass suggestion and added a unit test +Gary yes tried that too +Is this is still a problem? If so it has to be like there are no matchingRules specified for the attribute and the server is freaking out +Thanks +What is the reason that you set the initial size to +Closing resolved issues. +patch for generated patch +Postponing to the future +What has changed and where +here are some initial thoughts on requirements for applications like hbase +The rule of thumb that we worked from was the Source was and interacted with external resources whereas Item concentrated on the objects +Im looking into adding tests to the test suite to confirm this behaviour +I also thought of another option for this +Ok this is fixed on gt head +In fajax case it is clear the client behavior applies to the composite component not to the contentFor other specific validators that does not has a similar intention like fvalidateBean should not use wrap mode +Eclipse tries to load them withrun mode +this issue as a new feature rather than a bug fix. +What do we have to do to include tha patch in the releaseIMO the minimal amount that needs to be done is map to map to +Btw I see your point regarding the implementation of +Fixed in r +Ok great thanks John +probably you need to do it by hand +Thanks as requested Ive updated to Railo but the problem other information can I give you +Can you provide a link to the definition of what it should do +Thanks for review stack. +Yeah you are correct I have tested in AS Final release with the hack you have done and its working +Ok I can confirm that after upgrading to where was applied and after performing a scrub we were able to bootstrap in new nodes again +Extended IT and fixed piggybacked bug report in r updated snapshot published +Please reopen if you find any further issues with the or newner releases. +Thats because CMP engine is started after all thes are created +Youre absolutely right +See the package in the sandbox +And kudos for the fix +Hi JoeIm unable to reproduce on Postgres with trunk and release + +Proposed fix dont use an object for passing bindings +verified by Paulex +rename from afile to bfile This should read rename from afile to bfile bfile on time one is moving from current state to new state one uses an atomic rename +Joydeep is right + +We have to downgrade to Jetty +Thanks Ravi and Vinod +ThanksShivendra +But we can cover that with another ticket +Defaulted to the first of the month when the date string has no day that rcalendar behaves as Brian described +committed. +It must be something else Are you sure that there is no other error within your codeRainer +Oh silly me I see that you did attach your test results +Hey Juraj this will be awesome +In particular all accessors in appear to be boxed whether the value is nullable or not +This patch contains changes to the core runtime to enable callback support +However I do think we should be able to integrate this with something like +Memcache use is higher now because of the use of percentages instead of hard limits of course +Like I said above Im fine bumping the version +The from tests is expected because this is only a change to log lines and hence no tests were included +and i think MVEL is best candied +The patch has been appliedPlease verify +Need Pyro +applied in revision Alexei please take a fine +Im on the latest patch and will commit the previous patch unless there are objections +So are you using a Java term but not talking about Java +New tomcat support is the way to address this. +See for further details +It calculates a precision values and rounds the result accordingly +It moves all of the based stuff out of so we can separate the more stateful writing code from the less stateful reading codeRunning through an internal test queue now +See also recent discussion on dev list. +Added patch +So we still need converter to convert our Jetspeed per entity db preferences to defaults psml preferencesAnd I think all Jetspeed users that use this use case has such problem tooSo can you create such converter? I will reopen this issue as its creator +This patch fixed the problem +Definitely an important fix +Adds a few lines in that reproduces bugPatch fixes bug and test now passes +Skipping missing properties looks like the inverse of the problem here which is dealing with properties that dont exist in the Config class We dont want to skip over errors just turn the stacktrace into a friendlier new patchDont skip missing properties print them all out and terminate +Thanks its nice to have methods with java doc and comments +Yet the other should the be seekFloor noAhhh right we had discussed on the dev list +I dont know whatshould be done for them really but looking at the Java skip code theyarent handled there either +It test a usecase whth dependent objects and keygen +So the script could be in a separate file I see you deactivate yarnmapreduce on boot +This dlldesc is kept in axissvcSo next time same service is called I get that dlldesc and use its dllhanlder instead of reloading the library again +I would suggest using a symbol to store the context path and just keeping that in sync with your actual deployment. +sounds good to me +Fixes the issueStill need to Add test cases for kill and various states of Fix the to send a kill signal to +Alan do you have a for this need? And frankly you can do this in Java code and use a method call expression to setup in the XML DSL. +Thanks Ahmed! +Again this is fixed in SVN already +Lets get it done for +NPE was fixed +Thanks for the investigation. +Meaning how does the directory match for that attribute if there are no matchingRules associated with it? I think the attribute MUST have an attributeType definition supplied in the schemaextensibleObject just allows us to add any attribute to the entry +thanks once again for ur efforts +I think it is not reasonable to assume that a file name has an extension in order to be used with a relative path +It would be great if you can provide it +Yeah I was afraid of that +With leave everything as is I only meant the second part +And Ive got the exceprtion +My second vote is for Maybe we should try upgrading Selenium or searching to see if this is a known issue +The patch includes the first two examples and is completely independent from the rest of the projectI included the repositories so that a developer does not have install the snapshots into their local repositories +If you think its worth it then yes I dont see why not +It would probably be possible to do better than that if we happen to have more time to spend on this is some remaining work to do in before we can use the new API in +Max suggestionnothing but presenting the primary functionalities up front server configurationstartstopdebug deployment and archiving and then walk through an example of each and then have a reference over the views after +it takes some time +More info in a few minutes +Yep looks good on my Mac as well as Windows on FF IE ChromeAlso Fedora on FF and Chrome +also include Project Undertow +The attached patch seems to be the minimal requirement to resolve dependencies and their repositories +fixed +Can you provide th CQL statement you used to cause this error? Just any wrong attribute name will cause this + fails because of the bug +Revisit open of Date see above +Has been fixed now in trunk +I keep following the project I will syncrhonize the sources and certainly I will try to contribute the part with JMX support compatible with +moving to M and still hoping for some feedback +that removes thriftfs and its build references +I wasnt aware we had any linuxisms in there +Presuming found in versionWe need to investigate this further as the current is largely acutnpaste job that hasnt seen much testing IIRCTargetting the beta meant FIXED +Failures also seen in DB and Sybase databases +Please could you verify that everything has been applied as expected +Sure we want to expose this? Its only useful in limited contexts +Bulk close for release +Are you referring to replacing the boolean force with an enumerationThe enum called flags +Looks good to me +Which could be relying on a server id or not +Ok I missed the fact that theres already a parent ticket for that +What if there is no longer an issueHaving the operator clear out the blacklist files on restart isnt unreasonable if thats what he wants +Cant you use a setRequests function? If you didnt take requests in the constructor it looks like your patch would be half the sizeOtherwise patch looks good Ted +Sorry about that +has container listener configuration using the listener element and sub element with value LIFECYCLE +HansenThanks for fixing this one Can you please provide us with a fix version and a tentative date when a fix will appear on the download page? TIA Wolfgnag +Here are the results Number of producer threads Number of partitions Batch size Message size Kv patch draft patch Still a performance makes sense nice catchI think to make an informed decision we really need to also test the acks case with say threads +The two are conceptually exclusive but its impossible to express this in maven. +an additional comment The fact that the pw hint refers to an auto generated pw makes it easier for people which knows ofbiz to get unauthorized access to the system +Manoj Could you please attach the xml payload that you trying to parse? What is your underlying parser +Bummer +If were going to take thetime to split up the jars appropriately lets document the jars for thecommunity +Another way to resolve this is to use in init target of instead of +Patch attached +The retrieval of value for staleInterval is in ctorI am not familiar with trunk code base +Hi WernerI havent tried with the latest release of castor +how to handle all of the previous change entries this affectsI think they should all be fixedIe the is something users can read to see whats changed since the last release not the account of the series of iterations we went through before releasing +one option is to only restore the content of the item not the whole node +If youd like to discuss this further please contact me at chris DOT beams AT springsource DOT com +I tried to take a shot at this issue +Thanks for the patch Knut +I would like to request that this bug be reopened +We are doing research in IR and NLP IMOId like to use Reuters for some classification work too so on it +kinda ugly +v of patch rebased against current trunk +I attach a patch done while configuring eclipse +can repeatedly get this error +Added support for catching exceptions that occur during the marshalling of exceptions +Patch doesnt include +One thing I dont know is if it should be the job of server side clientserver code that you are looking at to do this of if it should be the job of sql execution engineto recognize an input clobblobstream cant be read twice and it should do this work in the case of an update that might require updating more than one rowI think we use to have generic embedded issues with blobclob where it would instantiate them into memory in cases like this +It is parent builds you need to how the hell can I know whos using +What is the status on this? Can we proceed with the release +needs verification however +Updated rebased +screen should be remade as the current ones dont meet the usual format used in other pay attention to the changes described in the linked are adjusted according to the format used +However was already committed so committing another patch under it would confuse the changelog +Brainstorming How about set the min to +Hot standby functionality acheived by shared storage approach in now +What happens is that various threads are started while the server has the root user id + +Existing implementations should ignore the name but they should also generate an error if a union has two bytes branchesA dynamic language needs a way at runtime to distinguish whether a or b is used +Please give some feedback +I have linked the associated JIRA that triggered this request +got it +Closing this record since is now released. +we could add separate configurations for fallible audit loggers and critical audit loggersI was thinking along the same lines +It looks like that was an unintentional leftover from one of my checkouts +But separately each run picks a different subset of the tasks from each category to run so results from one run to another in general arent comparable unless we fix the random seed it uses +Bartosz This issue can be resolved +Patch attached +setup tasks for docs js tests and then implementations. +bettydramit Is this still an issue on +nick in doc with red color highlighted attributes to be hidden from complete return to gleb to describe the other attributes +Produced a patch from the temporary branch +Thanks DougI will have that and junits done by monday +Ive just committed this. +for the current patch +I just committed this +Committed revision . +My expectation would have been that running map tasks would do almost twice as well as running because we have machines each of which is dual core +Test case illustrating the problem +with that change +I applied the first patch but now we fail later the next time we try to load the filter w same +Added some dispose calls to the branch based on these changes in revision after releasing + +Hi ManoharIm testing against C service and this is working for me +My Eclipse seemed to go crazy +The version info goes into the footer and is generated from the version property +I think add a button to let people can navigate remain part or the url is ok to solve this problem +jchris oh I ignored this was in place +There has been some changes to the decoration layout code in the meantime could you try +Closing as no specific major rework is planned for +bulk defer to to +This was fixed in r. +It looks like there are some minor issues such unnecessary imports and some formatting issues +Reopening since the comments mention plenty of repros +There is an automatic redirection to the image drectory in the +ZuhayrIf I use this patch should I be able to package an artifact with symlinks? Could we use JDK features to figure out symlinks? +Merged to the branch in revision . +Otherwise it prints out error message and aborts the test +have committed this to and +Benoit any ETA on a proof of concept there +your solution is more involved there will be a lot of changesyoure right +dohThe other issue I have is about duplicate MaxI looked at but Im not sure what the problem is on that ticket +A new patch added inputsoutputs for rebuild index +Since this issue as about the path to an overview file it may be related to +This works for me think it is wrong to disable rendering of since the JSON renderer is intended to render resources and its properties without respect to their actual kind +register affects contains calls but not get! null calls +Please see attached patch for a possible solution +I just read back a bit and see the issue about SERIALIZATION across clientserver +see my patch +Thanks for testing +committed +fix the in the xml in and change BUFFERSIZE to long in as well to force the promotion of the position parameter +Seems like in this case we need to give the a little bit more information about which constructor to use +Committed to trunk and branch +This is not really a viable option +The Default has now been switched to AES bit encryption +Ive committed r to the community branch +Note that some commits for this issue were inadvertently committed under +You could but benefit is little probably not worth the risk +If someone wants to add a flag and allow this to happen then thats fine +bq +Patch checked in to trunk on May by Paul Lindner. +Execution of the programm does NOT produce any exception error or relevant output +Ive got concerns with this patch +Solved as suggested +Ok so we would need an EWS project in JIRA because EWS has nothing to do with EAP +Sure Ill prepare an IT for you. +One of the major reasons people use libhdfs is that theyre writing a app and I would be really surprised if this client can match the performance of the one especially when read is enabledAn additional concern is that webhdfs doesnt seem to return any checksum information with its transfers +Updated patch with few log statements for easy debugging as per Davids suggestion +TimThats unbelievable +Agreed +It works fineThe patch has been applied +Hows the sampling the needs going to work going against hbase? just needs split points so Im manually writing out the necessary instead of using the Sampler interfacebq +Id integrate your patches but I may not have an opportunity before I leave +I suggest not worrying about time based reconnect but instead fail the operation if all the configured nodes have been tried and all have failed to accept the message +Closed + +Passes tests locally please review +Without doing a lot of narly work Maven wont work with current versions of Sonar +JDK is not supported with Karaf +Achim you are correct are what happens in the shell +reopen to set fix version to again with correct fix version +No response assuming ok. +PR applied in master and updated with the new descriptorThanks +Its also not clear wed want to use this approach since the resulting term index density can then vary substantially +This patch changes the version of webdavlib toBut I cant find them on ibiblio am I blindCould you please tell me why it was needet to upgrade +Verified in JBDS +Can you attach a sample +Could you give more details +Well over to Glen thenSimon +Moving out of +obviously the shouldnt mention Base +rev also includes minor cleanup +This solution is kind of a middle ground +Thanks Stepanthe patch applied as expected +This patch fails a job if the jobtracker detects that the number of splits for the job exceeds a configures upper limit +Try rewriting the query to remove complexity +Thank you so much for that tipIt would still very much like to define the sets of dependencies needed for different modules and apps in our system +I reworded this a bit +For each attachment we write we instantiate a couchstream genserver which is then closed after writing the attachment +Should we disable this flapping test till its figured +Make to NOT blindly heartbeat irrespective of whether previous heartbeat is processed or notFiled we also add some additional flow control within the RM to prevent this work from getting into the event queues in the first place? Having the clients throttle on their end is important in the short term but in the long run we need a flow control strategy that can exert back pressure at all stages of the pipeline +I just dont want to discount a potential performance optimization without actually running any benchmarks +Are you planning to integrate them +this is a bug in CVS when you checkout from a tag that was just created +Thought this jira has been addressed somehow so close it +Doesnt look like a bug to me +So I think this should patch in commit for +But in any case you reverted it almost right away so if that wasnt accidental the regression was likely easy to spot so well see soon enough +Patch committed +I dont seem to be able to reproduce this problem using the debug jars and the instructions you gave +thanks for the logs above +Ill review this shortly. +HTTP sessions are cleaned up but the underlying sesssions are not +Attaching a repro which Knut created +Without the schema a client would specify sortfoo desc and Solr would have no idea how to do that +I believe that this is triggered by much faster block replication scheduling introduced by +Double commit my fault +so not the right approach for using as just a transport +libhomeedwardworkspacebin +Hi Eli I have some ideas +For more fixes youd need to create a new issues plz +that would have been difficult to track down from here! +Marking fix version +Moving to next release +I dont want to remove this option at the moment as if we need to change the way identity model is mapped to picketlink or provide some migration tools this info may be good to have in database or enable to change implementation internals easier +could take a few hour before it appears on +This is still seen in suggested based on the reference documentation +Its easier to leave the filename the same but its definitely not just statistics anymore +New patch w changes described above +I intend to go ahead and commit this since its a trivial change +If you want a command to be executed in the WTP editors you need to use the context +I let you decide howwhen the VFS integration is done +Its resulting in errors in one of my XML parsers +This also occurs when reading from a file not from +The vencat operation is very limited and meant to be used in very specific contexts +please shutdown tomcat gracefully +replaceNode functionality added +My gut feeling says yes but IANAL +Use type strings internal type string and display type string +This is because they are not to be used as java classes only as within the context of a SQL statementI hoping to do some work in the future that allows a table alias to map to a vti so that the diagnostic tables would be something likeselect from and thus the vti implementation details are hidden +Can we please have a resolution to this issue for +It looks like our current scripts are working kind of by accident +I suggest closing this issue and in the future open specific issues per provider +The more I think about it the more I agree that reflect should be independent of Avros runtime classes instead using Java classes +And yes there seems to be a problem with +Removing this line fixes thingsI forget where I read this recently but closing the stream explictly can causeproblems +Resources are not an issue my machine as G of memory runs on and theres no virtualization in the picture +bqAny reason why NMPORT is set by NM but not NMHTTPPORT? Looks like we can remove NMPORT and use NMHTTPPORT instead +I merged into +We are just iterating in trunk +svnignore +I agree not very nicePerhaps opening another ticket and targeting could be an idea +Patch integrated to trunk. +I committed this. +Hithere is a mistake in the patch you prefetch address as an integer number but not as a memory area +Maybe it was different on previous by me Sergey +This bug has been marked as a duplicate of +Unblocking this +If you like the idea of it could be used in many other places in Giraph +Here is the proper patch +Did you reach any consensus yet +I dont have a problem with the changes from this being reverted and rescheduling this for a future release since IMO its a relatively low priority issue +The POM includes all sorts of noise that just shouldnt be required per Chriss point +I think the more important thing is the performance increase as the key comparisons are very expensive +the patch into master and branches. +All these warnings look like theyre caused by the cadvise bug documented in + +Made some additional changes based on offline comments from Devaraj and Owen + +Apache requires that we vote on the sources of a release +Do you plan to submit toYep thanks for volunteering +See for further details. +I will update some documentation in this senseI have assigned this to Misty but I am not sure how the docs process worksIm Writing a mail to her with a link to here +If there has been a plugin written for this would it be possible to get the code added to the wiki? As we have both and for text and outlink extraction for html format I dont think that this plugin serves much purpose for the average user of Nutch +I just committed this to and trunkthanks! +fixed faulty column size detection for decimal of unclosed issues. +Sure? Attached list stops suspiciously after fixed everything after M already +to CVS HEAD and Oleg +uses a gate to wait until is re right +Should build clean and pass sourcecheck +To me root could mean either the server root or the webapp root butprobably doesnt mean the current URL minus the file part which is what I thinkyou meanperhaps dir +Cleaned the code of the patch +Added change mentioned above to branches and tests +Thanks this case we had tests with wrong results checked in +vpeplugins +This file replaces the default etc file +The synopsis is the same with +If the versioning support is required in Windows please file a jira for that +updated to work with trunk +Try Tomahawk and Trinidad and see if that setup a new server and was not able to reproduce it with Trinidad I will close this issue +Fixed testcases and reaction nullempty path for constructor in Rev +Can someone please assign this ticket to me +on plane right now and missing dependencies so cant test this but in tychosurefire code is looks like useFile is hardcoded to true but there is a redirect that might be worth should try get this fixed in tycho if still the case +Retained object sizes as captured using Java profiler demonstrating how significant this leak is +Thanks for the scenario file! +Regenerated the patch as per Ashutoshs suggestion +Thanks +That leaves the addition of the regionhistorian column family and the bloom filter cleanupPlan is to make the migration doing bloomfilter cleanup and migration of HTD etc +Is it possible to do some xml include and put them in a common file to avoid missing it in the future +It would be better to match the prefix in a fuzzy way but the current approach does not do this and I see no solution how to do thisFrom my code analysis the remaining code to do prefix matching is identical to so I wonder why we have both at all? Lets remove the Jaspell one completely its just the same concept with more ugly code than theI will close this issue as wont fix. +That being said the Callback Server is one of the most inefficient ways of implementing a Native to JS bridge and well probably get rid of it in a future release. +The comment is relevant because implementation of this behavior is done in exactly the same way and has already been implemented for findbusiness and findserviceThe issue includes an attachement that can be altered and used for testing this functionality once its implemented +Is it possible to add another constructor to which is including all necessary parameters? It would be really helpful because currently I have to create a request by parameters and passing inside +For the C backend it would require a significant change to the way the code is structured and the breakdown between and generated code +I have just updated for all +I have a use case where users can create tables CRUD screens validation for the fields etc +what Randall saidGreat job PaulOne caveat thinking about how the comments are read over time recent will not convey they same meaning as it does today maybe we can make the messages more explicit and have them include version numbers of what works and what doesnt +Ill resubmit without that +Attached testing app WAR +In the the design of I made to be compatible with the original of where the pathname is a +Please note that it would be helpful to provide a real patch instead of the modified source code you have add unit tests. +Attachment has been added with description patch again +Also please note the timestamps between the messages +This version includes a cleansed of Tuscany references +The patch should be applied from inside the folder +Maybe if you get these complex expressions it would be better look at Rules dont think that the expression field should become a clob or other longer text field by variables should be known in the expression language +marking as fixed. +Logged In YES useridcan you tell me how to turn on full debug in so I can see what is going on and report +Running full tests +moving both asm and metaclass jars to containerlib solves the conflict +Thanks +This fix version means that in release notes for has to be written that Guided Drools Rule Editor was removed from JBT +it seems to work +There may have been some other small bits in there but for the most part thats it +closing because obsolete +Will the patch +Closing old issues +HiCan you try patching your source code again +I have the opposite situation +In we only get a set called from exception +Have tests been run on both Windows and Linux to verifyYes Ive the tests on both Windows and Linux +If you have a look at the plugins which ship with Nutch these are the only ones required +See also and +It has never been applied +Does somebody know any workaround to get it working in Glassfish container? +Any ideas for an external APII like the latter option more +Go for it +Committed patch to resolve change +I have optimistically gone ahead with a fix for this that passes all tests but removes the old loop break logic which I think was probably patching around bad loop behavior that no longer exists +I have created a new JIRA . +Please apply first +Since it is likely the application Flume is embedded in will be managing the configuration delegate the reconfiguration handling to it +Committed as r. +Pattern objects are notoriously heavy to create +This is a minor issue as I can resort to the same as I did for biggest dependency is ASM is not in the see Jeff Jensen has put in a request to the ASM team to publish the will put in a similar +ok clarified so I will change the validation on the Console side instead to ensure we do pass to server values +Gotcha +I just looked at Dans patch and it is certainly much better than the one from this gets applied soon +Closing resolved issues for already released versions +Fixed and merged into branch +Hello Andreafollowing your explanations the pixel gutter makes sense to me now +How consistent is the list of attributes from Livelink instance to Livelink instance? In other words there is undoubtably a set of ones that cannot be changed and there may be a set that can be added as a customization +Thanks for clarification +patch hbmddl ant task update all hbmddl properties setup ant execute test all is ok here +This is implemented in a branch of mine and an initial review has been done by Brian +We could add some credits in the controls Javadocregards Malcolm BobI have had a quick look over the tree control work and it looks really really good +Attached is a patch to fix the debian build and include the gem in the debian folder as it has handled previously +Oh I see +Assigning to Denis not sure who should own this +Note that this may even need to be done even with the updates in the WSDLJ +Add Trinidad to the list +Based on the parts changed it does not seem that you put the support in for the handling of the requires attribute in the XMLSCDL +Do you see any downside to always generating the setters and getters for fields? It certainly doesnt hurt to have the extra methods there if you dont want to use them +It uses the passed in version for consistence checkFor other error situations in it seems either it should be taken care of by the current system or it isnt unclear how the sytem could run into that situationThe patch ia attached +Oracle database lock query result +Additionally upon successful completion of a deployment to a set of nodes then a customizable deletion policy like thing will be default cleanup the old cores on the system +! Thanks Sean! +Fixed in r of spec +Is anyone opposed to using a custom published artifact on +Grepped through to make sure there are no other versions +Mass closing all resolved issues that have been closed for more than one month without further comments +And then I asked what the installation looked like and it was a chopped up Maven install +Linked to patch that ramps up compactionssplits more slowly +Committed as part of +That collection is held in the individual overviewDocsbindingTemplates contain a collection of instanceInfos +The generated invokers will need to call the method to obtain the bytecode from the generated method when the equals is +Thanks for the patch Ashish +Also all UI Repeat samples in Metamer doesnt work in another template +If you have other suggestions Im open to hearing them +There should be only one way to create a server dialog is merely a wrapper around the Server creation dialog box and simply allows the user to modify the JDBC userpassword and Teiid display name +Is it true that the endpoint must be able to deal with that ambiguity? Yes when an array is reprented as a single element definition in other words no wrapping complex type then a null value is semantically equivalent to inserting a single null value +This option can be configured using the elements in the +Depending on the Maven version or JDK vesions it could just be getting a different mix of and transitive deps or similar +In our next meeting can we review my code and the xquery specification to make sure I am getting the details correct +Attachment has been added with description Revised patchPatch applied with rev +Additionally I will not be able to commit this or review or commit any other tickets in the next days due to very limited Internet accessI am on this one as is +Yeah I was a little curious about that too +projectbootlib +Documentations and the relate to Resequencing Message Processor added +Ill close this issue as fixed as the database engines I liked to adress now dont show failing tests any more +Since gc code was slightly updated I removed from the patch unnecessary lines +Backported for function in hibernate dialect use the following four will create a field of type bit in the database +The attached archive contains my testapp and the details of the issues I have found +This will be handyWe have a coldstart option in our App but the process is a manual delete of the Felix cache dirs prior to launching Felix +we could instead put it in srctestlib or somesuch +The first is the missing assignment which is fixed by the patch +Reviewed by Alan +Thanks Vinod! +We are guaranteed that both states will be received together however these states are held in a map then iterated and applied as well as notifying in the same loop +Can you share the your projects +It does require that you explicitly list all the deployment plugins you want to use including +I will throw it on my stack of bugs I want to fix for +The would have to be changed in it is independent of webhdfshoopLet me recap all outstanding issues in a follow up comment +add description Creating the Component with a Page Tag Creating the Component Dynamically Using Java Details of Details of Usage Customization Skin Parameters Redefinition Definition of Custom Style Classes Relevant Resources Links +I can imagine the characters being in an install path and as part of the install the application would need to generate a file with the aliases which might be a pain but I am just speculating on usage +But not in Pig code but in loader +It just so happens that this issue brought that one to light +A quickstart that demonstrates the problemEnter a number with at least digits and submit +bq +the cleanup is completely broken +Or under like the S codeOops +Resolving. +committed +This seems to work +Shalin I have updated your patch +Any feedback on the patch? If not Ill go ahead and commit it to the contrib section +Please review and give your comments +Thanks +We should fix that to not create dangling nodes instead of having another step to remove it later +Ill put up a proposal first over in +This causes active NN to exit when backupnode has an uncontrolled shutdown +Without it I will need to track down every place we change a persistent collection and manually sync that with Compass which is less than desirable when all other changes are synced automatically +It would be nice if both approaches would be equally powerful +we are exploring options today so hold tight +The new models are committed to the trunk in rev Since the models validate the configuration client code can be simplified +What happened to this issue? The pull request is not available any more and the wording in the spec still seems to be wrong +and closed Version attribute. +Well need the version of oracle server and driver plus ideally the content of the objectstore and the data from DBAPCPENDING +Are you using a custom file by any chance +What are components Exchange objects also provide a means of correlating messages because related messages are stored in the same I dont see it like that +So you might see a more informative stack trace now then the one you posted initially +It will compile and install the plugin in your Maven local repository +ear in r +This has been done for the unit tests but we still need to run the broker itself and the test clients undervalgrind during make check +Passing to Fernando for incorporation of release into EAP CP +this error does not occur with current axis am pretty much to axis and for that matter web services +Eclipse has Checkstyle PMD available as +Did you try the original scenario with GEP and it failed? If that is the case we should either this JIRA or open an new one +Its a transitive dependency from to which relies on aopalliance +The block id can be logged with the blockname tableHeres one way to migrate initially the id is zero +I am wondering though whether it makes sense to use Immutable with +This issue is related to which also seems to deal withfailures of failover attempted before replication has completelystarted +I guess its up to Eric or other people who need Framed +applicationssrcwebappA +The and spacing problems should besolved in a separate effortSimon +marking this for M for at least seem to me this functionallity is rather critical to allow VPE to prevent us from having VPE choose what templates it uses based on namespaces first and then defaults next +It is necessary to implement mechanism by css layout styles and selectors +IMHO they shouldnt be present in richtab component wdyt? They are not present in toggle and accordionItem +no reply presumed fixed +But a unit test is not trival so it is not requiredI just committed this. +But if a managed object implements it it must provides the corresponding Stats The Performance Data Framework consists of the model whichany managed object may implement the Stats interfaces which specify standardperformance attribute semantics for each managed object type and the Statisticinterfaces which provide specific interfaces for representing the commonperformance data types +SvenThank you for your fast reply and explanation in detailNow I have confirmed your fixed the problem +This way we can avoid calls in each method of ProxyThis approach implementation gives up to x boost on attached micro benchon patched versionxxtrueresult on clean versionxxtrueresult Ill look at t see this before +How can I help resolve this issue? This is a real showstopper for me because the remote repository is another Archiva instance within our company that is used frequently and this is affecting us pretty severely +what is the license +Pushing to + +The JPA spec has an example in section that illuminates further +Thanks for the patch +The most interesting is webhelp which was developed as a Google SOC project +It seemed to break only when there were enough factsrules fired to cause the ASM optimizer to kick in and the class I was having a problem with referred to a class that had an object tree several levels deep +The bottom line is that the main impact is given by different default heap size +But the defect is Dependencies report throws when a transitive dependency has system scope which is still whats happening +Luis are you there? Can you give a status on this +We have less than weeks before we cut the branch so lets get it in now so there is time to test against all and the +The changes on this JIRA were the cause for the failure of this +Lets try again for patch includes all the features of the last patch posted to as well as patch is for trunk +oh also I added the default parameters as tucu suggested +Fix silly mistake in memcache where tailmap would be empty because of test comparing rows was using the instead of passed in row +Manasa Can you test if this issue still exists. +Thanks for the quick fix +It uses less virtual memorybit systems have an abundance of said valuable resource +Thanks for the reviewing Jon. +Preserves existing public methods adding overloaded versions of getHighlighter get and doHighlighting +This new behaviour breaks our client +Will open a new bug that correctly describes the problem +I do not believe that this works in either but i will test it on to be sure +for second patch committed to trunk. +Is it really ok if you try to read data halfway catch and keep the instance to continue reading next data from the stream starting at the halfway point? +If you are building your own libs you need to build via which will generate the info files for you specific to your build +aboveorgapachetikaparserhtml point thanks! Fixed in revision . +Can you please create a new JIRA for this issue and comment the issue number here so that I can follow up on it? +Merged in revision . +This will hopefully make the attached release note appear in the release notes. +reattached patch with test and example pictures patch is still based on applied on and trunk thank you for the contribution +fixed in trunk +It is recommended to just use a new empty folder for this but you could also use an existing one in which case it will override any jars that might already be there +Hence I suggest that for the moment we continue to keep development separated out as it currently is with trunk maintaining the version and the maintaining the version +I have attached a proposal for how the build scripts could automatically set and based on the operating system and JDK vendor +We can document it clearly and people will find it and use it +Committed to trunk +We should track that in a separate issue though +Sending Sending srcmainjavaorgapachexbeanblueprintcmSending srctestjavaorgapachexbeanblueprintcontextTransmitting file dataCommitted revision +Currently slinksbutton have no support for actionlisteners +Fix is committed with regression tests +I tried to convert the name of the method but found issuesPath cannot reach a particular object for Set and Collection already +Also tweaked Sybases Relational Model Processor to turn text type to CLOB and image type to BLOB and back again during export +Assign to during tidy up prior to +I think this is consistent with nearly all other Web frameworks in all other languages +Since the user reported that this issue also related to not having the file generated I am attaching a sample one +It is noticed that uncompressed references require uncompressed objinfos which need to modify VM also +Yeah I know it can be coded by adding new conditions in the implementation +Committed at subversion revision . +test patch to address to the problem +I think this is a blocker for GA but not for Beta + +fatalbean exceptions are now trapped and logged before being rethrown +This is an unrelated patch +Thanks for looking into it +Each tenant has a relatively small set of records to search and searching is never done across tenants so this reduces our memory footprint on a search considerably +Patch applied in Committed revision Thanks Shivantha +I understand we need to generate the javadocs and package them with the packaging but I am wondering what is the requirement on source jars? Where should they be bundled if at all? Thanks +The errors tag is a specific type of message and really has no use now that there is a more general message tag +to a deployable descriptor format as a feature though that doesnt alleviate the issue with allowing configurations +Glad you got it working +Here is the updated version of patch for the tests +I updated the specification to incorporate these changes +Is kryptonite a cluster patronized by yahoo employee? +That should avoid the entire problem because nothing will be serialized that can be changedIm still curious what could have been the cause in this particular case +enhancement You can simply register your own version of the it will take precedence over s internal one +Id actually prefer to close this issue for the time being particularly if its might take a few months to follow up on it +Samsung is one of the goofier ones +I committed this +if still broken please reopen with a quickstart. +I have worked on the WMS wizzard to provide more helpful error messages +However a configuration variable is added to and which enables the to handle custom password types +Someone disagrees +Tim spelling corrections verified at r +Passes check style checks +Now users have to copy the plugin and make an internal version to release their projects +What I see from the screenshot is that a different area appears to have been shownAs usual wed need data and requests to reproduce +Complete with commit ffaeedcbccfdfcaeabe. +Any ideas if the issue will be accepted? I am also analyzing using sonar lite mode +Is it? I dont have any specific counterexamplesIf you cannot easily find an counterexample does it qualify for quite specific +Shall I mark this as a wont fix? Even the idea of pointing at codehaus +Bulk closing stale resolved issues +I changed to use it as an example +Maybe I hadnt updated my Git repo indeedManually tested +Nudging this issue to ask for review from a Pythonista andor a committer +Then appends the separator instead of + +I copied from Branch the tests are passing now +thanks I will commit this patch +We still have the issue of deploying to AS because of it already bundling resteasy as mentioned above + +It would be great if you could try this +First draftRefactor code to prepare for work mentioned above +The fix lies in the configuration of one of those containers +Ive downloaded the ER today and tried it again works fineThanks Jeff Robert tested it on soa and soa windows machines which have wink +Karma granted on +withThis project has retired from the Apache Incubator and is not developed at the Apache Software Foundation anymore +Its a bug + +Its trying to look into local repository +Lets postpone this issue to next way we can get the server selection into the main start page +Closing + +Change looks good assuming Jenkins comes back OK +To clarify the fix breaks the design because a bootstrapping node that goes down temporarily but completes bootstrap will not actually have all the writes that happened during bootstrap on it +thanks +In reply to comment Now I think we could simply create a separate descriptor xml for each component and then use the to just process changed files and regenerate the xml for thoseSounds good to me +Im on vacation and will be back on Monday October th but Im watching this from hereVictor Thanks anyway because I dont have the patch here on my laptop +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +looks very useful to meIt would be great if we also had an iterator that runs over all the parts files in a directory that would allow us to simplify a lot of code +Apologies missed the fact that maxThreads is to be configured on the restlet component and not the endpoint +Unit test reproducing this I forgot to mention that both and seem to be affected +Updated the wiki I can edit the text and HTML and send to Bruce to update his site +What else has them? Postgres +Ive replaced the original with which contains nothing but the serialization methods +I am rephrasing this comment +The more validation you can get off before you go the better +Commit away +This will be addressed by the patch for is now taken care of by frameworkAlthough at the time this was opened around it was addressed by refactors at +I included a comparison was actually done on purpose this way since the map is under our control and populated only with Ill fix this to avoid any confusion +No futher info is needed +May be if he is fine we can merge both or we can seperate this seperately +testcase is suppose to be covered by however a multiple packet writes creates some strange problems in the test which might be taken care off separatelyCould you please open a jira on this? Is it a problem with the test or the pipeline implementation +if we cant connect to we should ignore the test hereits not solrs faultIll commit shortly if there are no objections +Is there a way we can remove the instanceof checks in completely? I think seeing these is a sign there is a serious problem in the spellchecking +Id like to see further refactorings I see that the code is using indentation levels whereas Lucenes standard is + +Another example I have seen a couple of times is projections attempting join fetchingselect from Animal a inner join fetch for throwing exception this query is semantically incorrect +yes the request result is null +ThanksFuhwei patch in revision +As an unskilled Java developer I dont have a proposed implementation +So youd potentially have less in total +Ive started this but needs to be resolved before this can make progress +Whereas the Ant tasks for Maven only download the that are available in standard public Maven repos +We need to get the release out now +YARN is the default framework now as is documented in +Given the issue of SeamVFS performing an excessive number of scans its important for the release that we delay Seam initialization until first request on the patch lazily loads seam on the first web you try with +this patch fixes a couple of places in the configurations +One issue with doing this for the branch is that it contains dependencies that are checked into the kafka source itself +Especially considering it negates expected benefits of trying to cache this information. +Stepping through the processPartial I can see that the lifecycle never proceeds further than Update Model ValuesA simple app isolating this behaviour will have to be created and the lifecycle stepped through in greater detail to determine whenwhere the error is occurringGiven that this only occurs with PARTIALSTATESAVING FALSE Im bumping this issue due to current resource constraintsCody thanks for filing this issue well get back to it when time permits. +initial investigation shows the issue is actually with BYTE type values rather than with per se +We are kind of trying to address the streaming issues with MTOM caching model +this happens on every default instance also in the seeing this issue in CP CR +True that it is not found on mediawiki and it is a matter of personal taste of how one structures your if you compare the Controllers Interceptors on to one cannot find the hierachical structure used in how do you navigate from Controllers Interceptors back to its parent Controllers? I regard this as vital information as the reader may have entered the site through a web search rather than navigated the structure from a starting point + +I understand the issue and you are correct theres no way to do this right now in the code generated from AbatorI made everything private because it is somewhat fragile code and I didnt want people getting into trouble by calling methods when they dont fully understand the consequencesId be open to adding a configuration property that specifies that these methods should be made public something like allow which would default to false +Its not at all possible for me to show any warning when the user clicks on the open launch configuration that I can see +Excellent Ferdy +Hi DavidI think that this would be a really useful addition to Hive +They will be automatically downloaded from if needed during the testcases +Thanks. +No luck everything works fine here as far as I can testI do not get a blank screen +done +Many thanks for contributing it James +And it might turn out that theres even no possibility of a performance increase from skipping the rest of the code in the method +this is a patch thats needed for the in need to move these contributions and patches into trunk so they are part of the build +provided test doesnt work against the version specified +Closing old issues +The expanded TCK test data exposed this +Please let me know if there are cases where this assumption is incorrect +Coding Rules Compliance +Will commit this based on prior +The dependency resolution had been significantly revorked in that version +The one objection I was going to raise was the delete markers story but you got that in your footnote +settings directory of the projectIn contrast I have got a Seam Web project created by JBDS in workspace and it has got in Project properties Builders also Seam Builder and Hibernate Builder +The original resolution may not have been done but as the bulk change required a single value this is what I chose +Looks like recent changes to openejb obviate the need to patch +The sooner it gets in the more quality testing time people get to have with it +Its ok +I think there should be an option so people can turn onoff if the auto table creation should be enabled or notThere may be privilege reasons to let some DB admin user create tables and not let applications do it automatic etcAlso the with table created is a bit unclear what it did +To have the alphabetical order the user will have to confiure it in its project preference page +But thats a matter to evaluate after the work is done if someone is willing to put forth the work +xml file you need to modify when adding new data or schema installerdata is a script that is used to upgrade the schema from an old version to the current version. +Please send me any specificsThanks released closing. +I dont mind it either +Yes you are right +Thanks for the patches Heath +JIT and EM parts are OK +What is the exception you are getting +Last chance +on not releasing without this +Gabriel next time please avoid reformats its hard to see what you changed in the test file +Fixed release audit warnings and a few formatting warnings from checkstyle + +I still havent managed to resolve this issue +Hi Justinasyou should have each needed XML under solrcontribuimasrcmainresourcesorgapacheuimadescMaybe I need to fix the antIll inspect it thanks for your feedback +Nice work +Lot of new things were added in they have not yet been incorporated +If there is an array needs to be skipped and that happens to be the last field of a record and the record is contained in an outer array it does not get skipped properlyThe test patch has the test that catches the bug and the main patch has the solution +ok lets push this to then +We could fairly easily add a includedeleted option to the view options parameter the same we do for includedesign +If there is a problem please open a new jira and link this one with that. +Thanks Siying! +Attaching patch to do the change described above that is use a simple string map for the STARTUP options +If there are no objections Ill commit in a day or two +Sure! I installed via Eclipse Marketplace Spring IDE +Can this be rebased +AlexIt might expedite things if you simply attach screen shots to this JIRA task of the screenshots of the installer workflow attached +This has been broken again by the removal of this problem is more severe than the double menubutton event +I cant reproduce this anymore in Eclipse and UI tests we have for this use cases do pass on both and +So we keep a standard entry point for the performances tests +adjusted we need the same kind of upgrade to be able to use xmlbeans with the latest plugin version or only changing the plugin dependencies would do the trick? + Username djencks +Not to mention that issues marked as RESOLVED could still be reopen and commented on +Firefox seems okay +This patch corrects the regular expression errorIt comes from the +yep that was what Ive mend by providing a patch to felix adding an option to not including the into the print +Integrated to TRUNKThanks for the patch Nicolas. +tomcat module needs security module +This second patch will resolve the inventory primary key violation that results after the first patch is applied and allow for picking and packing of shipments containing order items from more than one order +Attached branch and trunk patches +Logging was rewritten to use Commons Logging in +I cannot reproduce this problemCould you check if there is no in the parent directory that introduces such a loop +ok this includes fully working conditions loaded up +Also added documentation on the different site building options +so seems it doesnt need to bump a minor version +Congratulations +Hi SebastianCan the property be used for thisElse if any code modification is required all patch propositions are most welcomeThank +Odd +So now it would be easier to clean up logging Yes +Closing as this was released in +I dont know how to setup and run +this solution make use of the Axiom to represent WSDL and therefor much more efficient than previous solution +This needs some tests as existing tests pass without the patch +Not able to reproduce on latest master code with xenserver host setup and version is +I second Nevos comment +Here is the updated patchDont forget to add vmtestssmokegc to SVN +This updates the release summary to describe the release and regenerates the release notes accordingly +since the suggested change is i suggest to apply the changes to instead +Need confirmation whether its in branch +pushing to to affect the release +patch looks good +A Snapshot is deployed +This problem whitespace around the ampersand being lost while parsing xml seems to have been outstanding for a good period of time +The directory where wagon places temp files and the directory where the CI is run from are in different file systems and doesnt function very well across file systems. +patch coming soon +docsupdated +Marking closed. +The patch for allowing password changes +anything we could report as an issue to eclipse +So Ill revert to its former self and just deprecate +Are all these statements correct? Wants to make sure I understood all the nuances here +java regression tests go in +TrivialI was starting to read the latest functional spec and noticed the date and version of the current one was not updated to reflect latest version and date +Ive seen the same problem +Ive built it and run the SDO tests successfully on both Windows XP and RHEL +The current regular expressions from the project used to parse the svn log do not match when the author contains a space +would be great if you can have a look +Thanks Zheng! +Note that JAVAHOME could also be overridden through etcenvironment or bash profiles. + +a +With revision the redundant privilege problem is resolved +Its linked to a similar issue for Lucene +IMHO the contribution is small enough not to warrant this +Please change solr to match this +Nothing for to finally get off the ground with the new versionWith respect of the small amount of time were having there would be appreciated a tested patch which works for you even if there are small changes +Sadly though the only time we want to put something into just without going into trunk first is if there is a bug that only exists in +We should probably add tck tests to verify that users can call the new constructors +I think were voting for the concept and then we can adaptrework that as necessary +However using the jdbc driver out side the VM to execute queries is not completed +to Ill commit in hours +I prefer a link list perhaps even across the bottom of the page as that seems guess a choice of those two and perhaps the centered along the bottom option with the links in menu option is best +we should perhaps special case it to stop at? Alternately wouldnt the Stringable annotation go on the UTF classIt does seem desirable that generated code is able to work with the reflection API +for extracting annotations +Uwe Can you describe the tweaks that got you past this? I suspect theres little we can do other than specify a larger stack size since interpreted consumes many Java stack frames +Which mailing list and how many is to lowCurrently Open Files is set at +When a soldier deserts he simply leaves his post +Thanks Ahmed! +Alan are you the guy to look at thisDave are you sure the STS step is required +So its more about breaking it thanks +Besides that would break BCMarking this as invalid +FlavioYour recent rebase didnt work because was backed outI think you dont need to frequently rebase +Approved for +This will make collision even less probable +Verified in reopening to update release note info. +If there is an error in getting metadata the client will get an error code +I posted more information on the Google groups thread also +Thanks for the suggestion +It could be accomplished with FEST as well it too is capable of running tests on xvfb +enolan Applying Alexey patches all tests pass +Looks like a good option + +an aliasing feature of that particular reader implementation +Thanks NevoDid I understand wrong or is this only possible with one way messagesIe its not possible to do regular RPC calls between the two servers +Assigning back to Dag. +fix failure do FC symlink resolution in FC not in AFS +Hope this time UI people will address the issue. +I do not think this needs to be first class data typeThere is validating schema for JSON right? like XML Schema at least I did not hear about it +This file only shows hadoop openfrom you probably mean PID +In EAP JMS destinations do not get discovered because of the exception above +Dave +Please confirm +Kristian please roll out soon as we have to use a now and the forbids using SNAPSHOTS for official releases +does have a method for shrinking as well as growing +Robbie could you review this commit please +I manually tested by running the UGI class from the command line with and without the env var set to true +Thanks for the fixed patch +for this patch for now +Can we drag the sentence The property in the above two files should point to two directories in your local file system +Attached patch stops this heuristic from being applied to file and fixes the build failureYou might prefer to widen this exclusion I bet it would still fail if I built in a file path containing services but you cant save everyone +All I can say is that I heavily tested the logging output for ant test so I have no clue what has changed at the moment +After some database tweaking now it works even faster Just seconds on a large dataset in a production database +superceded by the templated sdecorate stuff +Please let me know of any concerns +Assign this to Rajesh to suggest how we can proceed incorporating the WS suite in the regular QA environment and make it part of the AS release the release I think we dont have to block on this task as long as Thomas is satisfied with the results of his local hudson runs +Succesfull and releases were generated from this +Although the nd replica is placed the rd replica is placed on the same rack as the first replica +Ive got a patch almost ready that brings in the stuff +One option would be to script generate some obfuscationUghShould we just spin cqlsh out to a separate project? I dont think it belongs in the dbapi tree either +Looks fine for me +Checked with Safari at +Rather we could have it throw an exception if HFOF is passed a KV with LATESTTIMESTAMP? Ill open a new issue to do that. +Add create method in Repository Class +Your patch looks fine +just discovered that on bsd is a symlink usrlib lib has always dev packages installed because bsd do not splits include and other stuff out of base packagesdetermining correct version number at runtime should be fine +No specific QA required +Paul in decompress I added inputSize at the top so that the header is +Attaching an updated patch which is still not ready for commit but it fixes two queries listed earlier which were failing with st patch With the query using column a in the select query I was looking for field name in which would be null for an expression like a +Itd be nice to hear your plansthoughts on this as early as possible +Marking as resolved in +Hmm is there anything special with your action implementations? Can you share your bean definitions and corresponding implementationsDo you get code completion or hyperlink navigation on bean and method attributes under question +Thanks! It did work that time +any good + +that makes sense + +All relevant calls have been removed +Updated the logic to unwrap that are from Teiid. +done +This passes mvn clean install right? Ill commit it +Nice job Aaron De Morgan MyersConjunctions and disjunctions are my jambq +You have to use versions of and jar files +Reason this happened +Clint please take a look at this +See also +Heres a patch against svn revision +Thanks Yongqiang + of Hudson run on the branch overall +maybe it might be possible to change something so we dont have to run up against this issue +I dont think its critical that this issue is fixed in but it should be fixed before the release +Arquillian based test case that shows the problem +Committed thanks +Any decision we make here should also keep that issue in mind +AndreasCould you show me the URL that I can download the latest jar file? Version or at least +shimssrcjavaorgapachehadoophiveshims shimssrccommonjavaorgapachehadoophiveshims +And probably yes it will likely need to be fixed in as well not sure if people want it to be +This is Jasons patch with some additional fixes. +with tided code and some unit tests regressions will be easier to as dependent on as you need reliable determination of for this to work especially during testing +Here are the traces +Fixed applied in SVN revision idea if this will be fixed for the release? If so what does the timeline look like? +How do I query for the progress of a jobcontrol object that has been submitted +jsffeatures +So I use this issue to hang all issues related to balance +was a decision made on this +Here is the case that works fine +Patch attached with fix and unit test demonstrating the problem +NET platform equivalent of the Java is +Just merged thanks for the +Im gettning the same exception using java but when I change to java I get no exceptionSo it seams to be related to java +Thanks +Could you attach a simple project showing your project setupWhy cant you just use the packaging in the child +It means something to control walkers +Please check it looks like you expected +Open web services + +Perhaps some data does not benefit from this sort of optimization +optimization! With this shall we bump up rpc version number +Updating resizer on window resize event was added to s event to test it on Mac OS because of selection rectangle repaint problem during resize event +Variant of patch committed in r. +Subversion should be stable +Patches committed Sending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetSending srcmainjavajavaxmailinternetAdding srcmainjavaorgapachegeronimomailutilSending srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilSending srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilAdding srcmainjavaorgapachegeronimomailutilSending srcmainjavaorgapachegeronimomailutilAdding srcmainresourcesSending srctestjavajavaxmailinternetSending srctestjavajavaxmailinternetSending srctestjavajavaxmailinternetSending srctestjavajavaxmailinternetSending srctestjavajavaxmailinternetSending srctestjavajavaxmailinternetTransmitting file dataCommitted revision . +Directory listing using DIR N S is can you look I cant reproduce and cant see why this would go wrong +re docs the main issue is the usage information for the scripts themselves not sure about the forrest docs but yes it would be good to update those as well if possiblenecessary +The script can only work reliably with an unmodified installation +patch stopwords file moved to resources folder and test to ensure it loads +I dont think this is failing anymore +I think that at the moment this would make more sense in the Terracotta Wicket module which would then have a Wicket version dependencyIf there was a module in Wicket already this implementation could go there but it doesnt make sense to create a module for a single class Also maybe some of the developers of this have commit access to on work directly on the code +Its still a pretty bug +Fixed conflicts with trunk +Yes my fault its not system mode but system property +We cant just assume theyll work +Okay well thats better than nothingCan we get a special notification for unapproved changes? Or do I just have to watch the space manually +That was very quick turnaround +This took a while to reproduce reliably but its definitely a big problem +Hi FreemanThanks for the fix! Can you let me know the Apache issue that resolved this oneRegards Freeman. +perhaps something like this will do the trick +StatusPart A is still to be donePart B is done once need to do again just prior to releaseBecause we let the release slip past the date of November we now need to fully meet the new policy +This is however just a different way to view the same information +Mind merging to ? Would be good to have the annotations there as well +Like thismvn installThe same syntax could be applied to skipping individual executions +Luca I was wondering if you could clarify a few points of your last post Im not sure I follow the argument about append +Changes have been checked into the product branch for the ER release +Sorry about my mistake I meant MarceloYour patch is in trunk r r r +Rather the scout codebase ends up generating a am working through this now. +Thanks Sharad! +Right now Synapse can work as a full REST proxy as expected but of Axis code could affect this very muchthanks a patch that fixes the issue +No worries please start one and lets chat there. +Ok no problem + +There is not a technical problem or so I put forth here +Fixed as part of. +Once we do so then the fix will be to ensure that the algebra operations establish these properties appropriately via simplification from where possible +and it would be really great if you could test itIm afraid there is no other way +Thanks Jason patch updated +Just removed the Web reference as it is misleading Web is completely independent of the security realms used for domain interfaces the domain interfaces also share the same definition between http and native invocations so referencing a connector within a subsystem that may not even be in the same process as the management interfaces are exposed from would not work +in theory it might be needed only in case that you have JDK installed but you are building with JRE +I am still confused about the exact conditions of this bug +I will wait for Jaroslavs comments before closing this JIRA +fixed in rev +Are you planning to create another JIRA issue for the same problemDont worry we should handle it here in this issue onlyThanksAshish + resolved issues for already released versions +If any of these updates are done in EAP and merged into please update the fix due to inactivity +lets not put anything in to b that potential breaks the build +Ive made some changes to continuum release and webapp components and it seems to be working fine now +It would solve all problems but it seems like a trivial change to me +Fixed documentationSending docTransmitting file dataCommitted revision . +A number of people worked hard to implement something equivalent to the symlink feature in +Thanks for confirming +if they fail on server restart the server believes something is wrongWhat is in the logWhat is in the +This is an issue of WTP +removed src directory from +If the return code asks not to include the keyvalue the return code is just returned +Hi Supun Can you provide me more details about this improvement +This remains as a doc until then +Anyboday has seen this bug in and wants to commit the fix toA consequence of this bug is that it introduces false alarms +is unusable as is missing +Thanks for the patch Supun! Ive added you to our contributors in JIRA and will add you to our people page as well +timestamp can move out so that the same ts is used across all local dirs +Passes the CTS tests +So hopefully this will work in I will mark this as Resolved in and will check it there +As this is technically a little bit difficult and as he have other incompatibilies I set this to wontfix +All that is need is a try catch block around the get method call +If you set the version to and generate to false it wont complain anymore +was probably the culprit +committed! +I asked to the developers of the project if they cant switch to but they dont want because they were not working on the project who use castor and it seems that its difficult to know the impact of the castor far away to castorCant we be inspired by the dependencies from? I know that I do not require the source code for maven and I can exclude it to change the just to say that its more difficult to know if we can change the version of castor or thanks for your offer to help us but it wont be possible also because the project will finish soon and they dont want to pass to much time on create this issue to know if it possible or not to add castor in a public maven repository +Consequently I would like to defer this change until is out +I just committed this. +Your steps and sound like a good approach to me +Commited and tested these files are now in the javatest directories under +Why cant we leave it in the CDK? The framework already relies on the CDK as a resource +I believe this is not needed anymore since r and sequel +yeah I will look into it and update the is failing because syncEnabled is not enable by default when object is created directly +cr i assume +More name discussion +Thanks David Reepmeyer +Im seeing some openejb failures + +Can you attach a file which not work +classpath files to help ease people into using Ivy and when theyre not used to it +sounds like to me +Need to get some thread dumps to confirm +I dont follow this patchIf the installer adds EAP specifically then that should not be added to the runtime path scanningIt should just be installed and nothing added to the runtime scanningWhatever the installer adds to runtime paths should always be treated as any other runtime path search but the patch seem to say if the path added to runtime path EAP install path then dont ask +any changes which could be needed in CDK should be is proceeding in the branch verified issue can be closed +This didnt happen until I did my svn update +As you see creating a spring job executor is a little more involved form the source code you can see that is some duplicated code between and +SSO plugins do not work anymore because they dont provide passwords +If we want to change that behavior we can handle it in another JIRA +OK thinking about it it is an OWB bugcreate should not deploy a managed bean +Please find attached a patch to fix thisMichaelthank you for reporting +The page has been updated to reflect and the test page revised to as of the text had previously mentioned as a target for supported modules +it should work now +Alejandro Good idea on checking with us first +Adds the accuracy functionality to the Lucene spell checker and also adds support to Solr +Should be fixed once Deng applies and we release a new redback for. +The majority of the failures are related to clustering +If something goeswrong the last started EJB container which provoked thestart of CMP engine will be reported as guilty +You can easily change them in the configuration when you actually deploy a Geronimo serverThanks again for your help +Most of those open files are disk and in your link Raghu youre talking about network IO unless I misunderstood +Thanks guys! I just now reproduced it with the gzipped log suggestion from Patrick +Attached a patch which fixes the bug Ole found +Itll produce results but should not fail +Without a public maven repository of our own we would end up maintaining packages for many projects that are not our you and common mistakes I have other repositories or pluginRepositories listed in my POM is that a problem? Yes the central repository must be self contained which means that all your dependencies must already be in the central repository +This makes it much easier to override things in that used to be in +Fix and tests attached +sigh +Prepping +In view of the fact that the code has changed substantially can you review this and decide if its still required and submit a patch +It would be good to add a test there too +jdk target? really in +A common solution for this is to have a trash directory per user +For projects you can also use the artifactId format which will take priority for the specific module that it applies to +I agree +But even without it once I install the second port becomes empty ie +I need to release Maven SCM first +Attaching a new release note +Plain text is better since Velocity is not a graphical library +I have reverted back to ehcache and it works now +Which hibernate version is it fixed in and does this address +Great patch Chao thanks! Just committed this to master. +It is not safe so they should not be allowed in the batch +are two new projects for the ORB adapters built seperately from the openejb patch file is applied to the openejb tree +We havent given up though were still investigating possible solutions +start working on it +With the information provided I have no idea where to even look as to what the problem is +No only the I listed +bulk defer of features from to +We are using postgreSQL database +Another way is to set the torque server configuration keepcompleted to a very very small value or +Im resolving this issue since all the patches have been committed and there has been no activity for more than three years +Modcluster does failover with Tomcat backend on Solaris SPARC correctlyand it doesnt work on +Ive just committed it. +Right now we have various combinations of file and jar sprinkled all over the codebase and adding classpath now fails to trigger the filejar URL logicIm curious why none of our jar globbing tests are failing as a result of this +Octave also noticed that if epsilon is used the exception disappear +issue is more like a bnd tool question not a Camel issue +Sigh +This output writer makes this problems go away for perl since it uses the right escapes and unicode strings end with the utf bit setThe patch is small and has virtually no impact on current codeIf you need something else dont hesitate to let me knowThanks for the comments +Range and interval semantics has been a bit updated in Groovy so if its still an issue please reopen. +just sLucenex for the addIndexes testswapping in to found another bugant test Dtestcase Dtestmethodtest effeaabffeabffdac DargsDont know whats going on here +Fix for allowing same username for all services +Fix version is now correct. +For example I am concerned that we havent tested database access functionality with the new transaction transport +comments onThanks for your reply +So insead of outputting anillegal namespace declaration xmlnsns and nsAttrWhatsup we simplyoutput AttrWhatsup +Pavol it looks like you are using old Teiid components +We may also need to split each collection into its own node and the just holding the names of the collectionsWe actually used to have a very fine grained layout similar to this it had some advantages but it was very expensive it turned out because of having to do so many calls to load the state +Attaching an updated jpg file +No problem I will wait +patch applied +is an existing class +Next time we need customization of converter behavior we would have to introduce the fifth parameter on the register method +This looks fine +MarioCould you commit this fix? This is required for Apache SynapsethanksCan you apply current fix and redo it in generic way later or we need to wait for generic patch +the general reason for a is that a dependent class is not found +Except the HOD bit that I dont know much about message format I didnt spot anything wrong +It terminates and thus doesnt end junit test properly +Try to run just plan maven from the directory and it should download that particular kernel version +jenkins job passes +You havent given any explanation +Linking which is the issue that encompasses enhancements to the Fuseki I suppose we could make the change so that the query string is preferred as the source of graph name and we try the transactional update when it is received that way and the dataset is transactional +So I guess the only way to do it would it in is there any way you can update glibc to the version snappy depends on? You can check it by ldd v +If the servicestate uninstalled and wipeoffdata false only the rpms are removed +NAKACK will also need changing to NAKACK for jgroups +hibernate database node is NOT for general database browsing +Thanks +Doesnt seem to be an issue +Changed title to be less prescriptive +Sorry it didnt apply because you had to first apply +I have seen this as well +What course of action do you suggest we take on this +I had resurrected an old computer on which i must have not removed those old from the local forrestbot deploy directory +The changes to look good +Nice graph! +postponing +Additionally if you have various interfaces you could use them +If Im right and if theres no way to fix this at least you should update the plugins home page to warn users of that they should not use the plugin +To be clear Im too much of a maven incompetent to volunteer +Gleb what do you think +Thanks for looking into it guys sorry its taken so much of your time +Updating the patch with that please reattach the patch with explicit license grant for ASF +Having poked around a little I saw that the pooled service model is using the same pattern as the threaded model should this also be have the same changes applied to it +We ended up with event support utilities +You need to go back to multiple lines when building the macro call string concatenation is avoided for a reasonWhat would a user do if they wanted to remove or already this help field? Not criticizing just putting it out there as a question +It looks like Hardy pushed the changes from the pull requestCan this be resolved now? +The settings were ignored because the fix for was causing an override to the surefire system propertiesThe solution was to use the append element +Closing this. +Works well but the fieldcolumn userLogin should be renamed reporter +Or should I revert back to an old production version? I need to work around this bug somehow and I cant use development code. +I cant use it otherwise +I will try to test that later on today +Sorry I was somehow like completely drunk +WFS is not and Im not sure we want that to be either +Really appreciate your help +we can now handle quotes around charset in jetty http the patch into branch. +Looks like could use the change would like a test case that gets into that code first +The results are as follows works identically on RI HarmonyIBMVM and HarmonyDRLVM +they dig into the object model to perform query parsing +this was done over a series of other commits +Will this allow access to the at a time when it is possible to add beans and annotated types CDI +Thanks. +Tim did the icuj upgrade independentlyThe unnecessary classpath references were removed with +FYI Ant does the things you described automatically from WHATSNEW The header in the default manifest now contains the JVM vendor and version according to the jar specification +DavidYou seem to be focused on the maven release so to speak I am focused on the normal release that people download +As the issue has no impact on the correctness +It works for both enhanced and unenhanced persistent typesPrimary limitations the entity classes in the entity manager must be Serializable cannot serialize a pessimistic transaction cannot serialize a broker that has been flushed cannot serialize a broker that has affinity to a particular connectionI have not done any testing in a managed environment whatsoeverI plan to commit these changes after making a jar available to some people for initial acceptance testing in the wild +Daniel is this a dupe of? Sounds like it +One thing is import list is ordered +Patch reversedLarsWhat do you think of removing the limit checking fromEffectively maxSize would be agrees to obtain consensus from related parties and finish code change +attached is a revised version of the patch that adds path support to Ivy standalone with consistent behavior with or without Ant +Do I have permissions to do so? I will try to come up with unit test although I have tested it in my use case +Is there anything else that needs to be done to unstick that? +And now I hear about switching from prototype to jqueryplease decide here what the tapestry is stable framework for wide community or a for few developers +Version of Javassist bundled as of Hibernate +Is there a reason to keep it open? +My comments are the same as in mark the spring dependencies as optional and please upload those +Note with Pool and DBCP this is all done in DBCP +Wonder how the security context can ever be null? +Were on these days. +The master file is updated to reflect changes in the metadata test +I think we should just close this issue out +now that I study the client version output clearly is there a logic issue? It ends with the followingShutting down engine via this URL jdbcderbylocalhostusersysadmpasswordshhihtybbmshutdowntrueBut I thought if you were running instead of embedded you had to shut down the database but not the engine? Im probably confused again +Packages are already split so this is not really new +Merged to I believe the remaining problems are with the JCC driver and this issue can be closed +Subclassing is not the right approach +Logged In YES useridHow do you disable automatic discovery +This is the the payload in the new method id and jboss serialization scheme in this is not relevant anymore. +Still the deserialization logic has to be optimized +Christian seems to have objections related mostly to the solution and to me it only means that we need to find a better one +If you use a durable sub messages will be discarded once you unsubscribe it +Terms from this field will be highlighted regardless of the specific field you are highlightingOnly file to worry about in that huge mess of files listed above is +thanks a lot Gavin very appreciated! +Why not help Apache SIS grow and contribute spatial base as a module there +Note that without it we get some of these zkid messages +Initial versionjitrinoemencoder project filesSee README file for details +Now that I implemented and export the Servlet API from the bundle can we also close this issue +See +I think this breaks set semantics for integers and strings +How does data get loaded into standard +Report facelets issues in the facelets issue tracker +Looking at this further there are a number of headers that would be bad to include in metadata +Before it does not do the check before this counter is decrementedWhen a dfs cluster is started if an early started data node has stale blocks the name node will ask the data node to delete the stale blocks as the reply to its first block report +An example of a handler that might have complex information is a handlerNote that in the proposed solution the savingrestoring of a requires an explicit action it is not automatically performed +BenThanks for stepping in Https was the default assumption right +I think that the module needs to be fairly robust or else Im going to have to do something like duplicate it into my project to make the necessary features available which I dont want to do +I would like to get the simple implementation done ASAP with all the tests developed in parallel +If theres anything buggy let me know +This way its not easy to check if this is really a zip ll set it to fixed still there is no real solution for it beside ignoring the exception which I consider bad practice +Lukas please assess and scedule was able to reproduce the issue on Android need to dig into that more +Moved to next milestone release + +Yes i plan to fix as part of the update +Modified the javatesting to reflect use of instead of with revision . +Sorry I have to close it +Hmm I havent changed the behaviour there so I checked and yes passing columns will return all column families +Verified as correctly completed in EAP CR. +tools guys will know better +But that is not really a bug +Actually I was hoping for this +Continued in r +The workaround is to use Jackson. +We also need to mark the variable as initialized in async to allow the Java backend to implement such initializations +is the format proprietary? Or is only the compression algorithm covered but not the inflating of the files +Committed thanks! +pluginconfigurationtagletArtifacts for the are not checked tooWould be good if the versions plugin could somehow determine that plugin configuration contains a list other maven artifacts +We should try with Google closure See included from branch +Javadocs are for developers not users +This issue report is pretty sparse +in SVN trunk and branch +Fixed in trunk +I am using a dummy in the testutils +is the result of the second startup +Removed Spring dependency unsing inDevelopers using iBatis with Spring are recommended to use +Cannot reproduce +Thank you Mikhail +may be related wrong location of the my comment inThanks for your suggestion about EAP ER +For objects use the corresponding k attribute +Rayees there was another reboot issue on which you had filed earlier +Jimmy this looks like a duplicate of +Yes I got that you said it expected it in the same read +So resolved +Hi Im reopening the issueJust a few observations Apache Directory Studio does actually recursively remove all children +In many of our projects we are using such kind of before package phase +Does the logic look correct? Good feedback thanks! What was the noise? The newlines? Personally Im a fan of whitespace as it makes if blocks easier to read but I can remove that if its what youre talking aboutThanks +It is to allow the use of the same outputs from different components running in a Map or Reduce without having to pass the instance as parameter +If you just dont inject any Faceboook service in your application no instance of it will be created rightAnd even if that is a problem for whatever reason why do it at runtime? Isnt deployment time enough? In which case you can just split the module into submodules and just let developers package the application with the services they need +Well given that this appears to be the first case where we are putting a value in there that doesnt correspond to a wire value it might be the time +The wampler data +To do that you need to up the rpc version +To Chriss comment even if you do everything in cygwin youll get mixed output files because cygwin doesnt have any impact on java +Victor heres another welcome page issue to look into for a later release +For documentation what you would need to do is add the dropbox connector to sitesrcdocumentationcontentxdocsenUS +The fix works thanks +Okiedoke Im convinced +Paul can you produce this release for us +Version Released +Please review attached license file then we could commit it to svn trunk +Committed to and +Is there any way to make rolling upgrades work? Even making the snapshot dir configurable would not allow for thatI was thinking we need a version lock or upgrade lock to block certain operations while rolling upgrades or rolling restarts for config changes are happening +Verified +Logged In YES useridYes I see it +Correction +The failure reported earlier was not related the changes in this patch as well +Actually the bytes AND propertyDescriptor parameters both are nullmodel parameter does workThe propertyDescriptor is and populated to the contribtable +The current approach is about as and straight forward way to deal with the immediate problem. +and it is working without having to change any setting in the DB +More or less followed the code changes proposed +In this case there will likely be a closed set of to versions +This is how is meant to be. +We still need to avoid the class loading imho +There are some unresolved variables. +Putting my vote on this oneAbout of our filters were broken by the change to making filters work only over a single segment and our hack to make it work in the meantime is not particularly great as it makes our filters only work with a single reader +bq +BTW the documentation needs to be updated to reflect that Document is also a supported primitiveIve searched the mailing list for +We are doing an upgrade of Thrift for the same reasonWe will also be moving to Avro once that is out but a move to with sASL support in the meantime sounds good +fixing the summary to reflect the problemadd description from mailing +If the proposed change in gets checked in before this branch gets merged youll need to make a change here +Then the method declarations you are using are not a valid syntax +Fix that bug during jive upgrade plugins is no reason to have pnbio column as s change that to text and remove shrinking updated value in code before saving +Patch attached +Thank you! +Then on the user name text box you just type the user name and if it is available then it will show the error +for proton +It seems that uses which is a wrapper of. +Comment on tomcat builder was supposed to be on not this issue. +Ive checked and this does workvar barString foo +Its now corrected and Ive fixed the problem. +I have tried to fix what I could. +middle is working instead. +New patch Added contrib folders to eclipse Added javadoc entries to Fixed Javadoc errors Upgraded test case to use schema final patches which will be committed now +Its simple enough to clarify that this must happen in fact this is only an issue at all because there is a bug in Weld +Resolved for you Esteban I can see its done +Moving to Beta ll work more on guides just after beta release as there will be no more coding tasks only eventual bug fixes if users or I will find something wrong in Forums. +It will be difficult for me to keep this in step with the changes we are makingto the Actions +Is the fix okay +Sure I will +ThanksI realized that I could create my own patch and diff that to your patch +Verified on ER +i changed the check to log warning message. +FWIW The version of on my machine is Could this be a sync issue between the EU and US version of the repoGary +The Patch is attached +At least I did not see anything in the documentation +im no expert here but this approach appears to be sacrificing the benefit of ever using jute for rpc in favor of the benefit of not breaking existing ddlsNot really +I made the change in the build script so now contains a manifest classpath instead of containing all the client jars +New Revision because I just noticed that in instead of Im rendering +Applied patch with thanks to William +We used to allow commenting on closed issues +Patches welcome +Defer all unresolved issues from to +New patch that fixes the failed unit test +I just committed this +I reported the wrong functions +Through the shell that might be awkward an unexpectedCould store that information in ZK but thatd be strange as well +This is slightly entangled with the nd later patch to that will be needed to apply this one cleanly +Added docs under the Developers Guide and referenced from the Caching Guide +I think we should release now +Rebased to trunk +The problem is with the plexus +Couldnt a query server developer just take command line arguments to their query server program +Found another memory leak caused by +the patch looks good to me +The cost of emitting the point is trivial +Verified working in WAS +Patched in Scroll Exporter Extensions plugin version Now installed on STAGING instanceWill be rolled out to the PROD in next release +Thanks and sorry for the unintentional noise +Removed reference to now +I have implemented a version without temporary file dependecy and committed to +Can you delete comspringsource from your local Maven repository and try it againAlso why are your log messages showing you building? The trunk is +Hi +I hope your ssh issue will be fixed soon +Thanks Hal +Thanks for the patch Juhani + will commit after tests pass +Ill see what I can do about that. +Thanks Saurabh! +Mamta I updated the attached files with the information about how Derby processes LIKE when using territory based collation +Does that mean we can close this ticket +My question is that is this a real requirement and if it is not isnt it enough to declare it as an implementation limitation of Managed X in release note + +I merged the patch to. +If you believe any of the above is incorrect please let us know +The message handler received the message and sent a growl message to the user note I have other user case that follow better the scope of general Messaging Architecture +Maybe this will address some deadlocks that weve had when groovy code gets invoked too early +close off release +The release note text is incomplete it only says groovygateway quickstart fails on I think you must have been looking at an old version of the Release Notes +Thanks a lot for your work anyway! Im sure that will help in the future +Thanks for the patch Omkar and thanks for the review comments Sangjin! I committed this to trunk and +Couldnt you achieve the same through different field naming in your schemaFields firstname lastname fullnameAlias firstnamelastname to make sure I understand youre saying create a pseudo field which we use for querying the actual fields? so basicallypseudofieldrealfieldrealfield Known workaround think it would cause confusion if who is both a physical schema field and an alias for itself plus something else +For simplicitys sake I put the plugin execution in the POM so it will execute for all +This should be resolved in that is on npm. +One of the reason the failures with RI is that RI will not add start parameter in the header for it is an optional one +I will do it +And you will have to regenerate this script +The EC thing is just to avoid having to have a properly configured locally with correct test data at all times +The CXF should pick up the Address information from the WSDL I will dig the Exception to see how to fix it +Will commit shortly +Please see if it makes a difference for you +svn ci m Removing old Range classes per +Thanks Sanka that was a great tip +In any way this is not the same issue as the one originally reported +The interesting thing is doesnt even use the new style sheet so I cant understand why that would have an effect on itMy recommendation would be to move the popup calendars styles to both of the main style sheets then style it to be the correct size +Thanks +We definitely should create tests to cover this functionality + +it into your own svn at a suitable location such as +The rest of this patch is Davids and nothing of it is changed +Ignore last comment +Patch for to fix the replace the existing in serveralldeploy and serverdefaultdeploy for will not be officially integrated in an AS binary I think. +comments on Phabricator +pushed +There wont be a release +RC provides no more functionnality to set up a HTTP proxy from the preferences +Is this just a test bugI dont know +very bad Struts was designed to instantiate action per request strange think can happen in other way +Logged In YES useridI can rule out the simple matter of exceeding max sessions and I can also rule out leaking sessions +SbastienCan you please upload YOUR wsdlthanks WSDL file containing some attachments using MIME type application +Also note it really has nothing to do with Nagios +However there are release audit warnings +Cedric do you want to commit it or should I +Closing the issue. +This patch will disable Tomcats JNDI creation +please correct me if this is wrong. +For example if we consider every code change a version we would have the following If I make a bugfix do I modify the copyright? Yes +The first one wins all of the others get the same attributes as the first one and going to add a test and a lot Andrea +So it is not possible to have a filter on the exact correlation idCould we perhaps create the correlation id in a way that allows us to set up a filter that matches all messages for a certain conduitThe current correlation id works like correlationId host userName threadId timeTheoretically it should work if we use a correlationId like host userName threadId We use a spring for the listener +Can you please supply the assembly descriptor you are using and the plugin configuration +Your logs suggested that you are using older version of HDFS +This module can now be built correctly using either or version jdk. +We have another issue wrt links in which could also end up affecting backwards compatibility +It did not run all unit tests and then declared a test failure +slip for later but keep. +Do you confirm you have others extensions definedthanks important thing this error stills exist on also have this problem on some projects with Sonar and maven or +There are two versions of and currently being usedjar vsjarjar vsjar these duplicates or are they both required +Will help debugging a lot if someone reports an oddity +I think the most appropriate way is the second way since we want queries like select distinct c c from t order by c to be validThe patch contains the new modifications +Merged in in r. +Closing this record since is now released. +Eventually this will yield the sameAlbert might be the case +Thanks Claus will do +the bootstrap works since applying the fix to make the modello version inherit in +Great! Tested and it works now thanks +We will work on fixing this in subsequent releases +I wasnt expecting it to be accepted yet I thought it was still a little roughThe work is far from being done +Is there a reason you chose not to use the get facility that is commonly used to locate the appropriate meta regions? Im not sure that this is a problem yet but it is my suspicion this could be related +On other DBs the fetchSize will be + was the feature that added this you create a unit test and even better a patch for the issue +Do you happen to use bean definitions with ? had some side effects you have a chance to verify this against a recent nighly snapshot? It would be great to know whether some of the other fixes in have an influence on this I am not using on any of my bean definitions +Vineet have you found any problems with null approach +I can post the code but it quite trivialIt would be nice if this functionality could be integrated into the stock BLOB mapper +todd dhruba as our practice the duration of global lock does impact the performation while it is insignificant unless there are many concurrent pure namespace op such as touchz or mkdirand I think we assign different sync priority for consistency requirement +I have not tried the patch of wenlong yet +SorryFind attached a patch that at least pass my manually testsNow the port for jetty is configurable as well the code already had trying to load a value from the configuration file but the node in was missingBeside that I organize some how the ports that are default used a bitI suggested following portsndfs namenode ndfs data node mr job tracker mr job tracker info port mr task tracker output mr task tracker report Well now all of this ports should be configurable individually +The patch looks good to me and thanks for writing such a clear and detailed comment +The only way is to execute a single test or add the tests to a testsuite + to Brian +I think the best choice here is +closing since fixes are in released code +Peter can I copy your Radiance HDR code into Sanselan? Is it under a suitable license +Instructions are in file named HOWTORUNREPRODUCER +thanks +Im not really sure we still need this. +I will work on it soon. +for branch please +Nothing like a minor emergency to get oneself familiarized with an projects codebasePriceless moments Watch for over the next few days +Looks much better +see hope theyll find a way to enable required facets for selected one instead of just pointing that some required one should be problem was that the gwt facet were only saying It needed java OR web facet +It does need to be documented in my opinion +I didnt delete the from derbyall +I committed itThank you Thejas for the patch and Ivan for the review. +Committed to trunk thx for review sorry for missing this +Ruby folks never implemented RPC interop tests from what I could tell +I think you are rught that we should probably have call back methods with lockacruired and lockreleeased methods +Benot what other tests do that? The only ones that do a wait loop I can remember like do it to wait for something to happen and not to fix an issue like this oneAnd as for the nd file descriptor its ref counted like the updater file descriptor +Regarding your patch so I think there are some issues with it can we discuss it via skype +Sample Turtle file with a BOM which fails to so it can be marked resolved +Removing backport target until we have a patch for this +But to simplify future merges of updated standard lib from MRI it seems that the better option is just to leave there in exactly original form it wont be used in but it will be easier to upgrade +Isnt existing supposed to take care of this case +Thanks +One possible solution is to ask the JSR EG to declare those abstract flags as implementation variants +Closing all issues that were resolved for. +Is this JIRA unresolved pending commit? Just checking as it shows up in my filter +Verified on QA it is done closing. +Closing issue. +If you have a specific problem with any of them please raise a specific JIRACheersDavid +rpcencoded is also prohibted by the basic profile due to its inherent interoperability issues that cannot be of that jbossws does not support multiple soap body elements. +I know Im going to fix progress and build results in trunkLeftover issues are build can see another issue not listed here yetSometimes there is more than platform suitable for installation +Patch removing the chained Id like to commit this soon if there are no objections +Not sure why Bill says its low priority +bq +wrt cloud mode theres probably a lot of edge cases we havent considered when rollbacks come into play +Is this working for you and others +David when you use buckets did you see bad performance for lookupsI didnt try in any serious way +hum +Can be closed +Properties support fully internationalized names and description +Hey PatrickAre you still planning to work on this patchThanksm happy to take another stab at this however Im not very familiar with pypidisttools pkging +Just a heads up +I tried to add the host in uDig with same result as above cannot be cast to m not even sure the module ever supported WMS but anyways assigning to the current connection now works fine and Ive tested a few servers last week with success +Its lame iirc with the right software you can crack it in a second or so but it may be good enough for some uses and it would allow access to files produced by others +Adding the file here +Starting point for fixing this has been committed in revision +No I dont think this is ready to be resolved +Are you working on this at the moment? If not I was going to take a look if thats ok with you +patch applied txs! +Applied at r please verify +bulk close of resolved issues. +A fix has been committed to archivatrunk revision review of the message text would be useful now +Agree this is not ideal but it is not just flow control it is any memory limit because the default behavior is to block a send if the memory limit is reached +The patch should address the problemOleg +Patch for yahoo security branch not for commit +Commit changes to trunk at r +Unlinking from patch got stale as well +I need timesQuite staffed with companys work at this Adam yes that would be awesome +See + test +Is this a recent fix or something +Afterwards I studied and discovered the synchronized I am a fan of reentrant code so the best thing will be I make the reader you should add some info into the javadoc to avoid this pitfall +Thanks for fixing EmmanuelFelix +This feature is mainly for performance and convenience yes +Apologies +Forgot theres one in srctest too +Fixed applying Norval patch and fixing waring on trunks. +I love the war story +However this does not seem to be a problem as far as the libraries in the packages are concernedI agree this does not seem to be causing a problem here but should be fixed in another JIRA +Ciro try with the latest source? Where +Using Maven here is just one scenario that could cause this + for the patch +bq +This is an incompatible API change +bulk move of all unresolved issues from to +Luckily i found a better jdbc driver for sqlite that actually does support fetch size so closing this as wont fix +patch looks good +It is done on merge outputs via null to keep the writer from putting in sync blocksIts actually more than a measure the syncs make the computation of length harder and temporary outputs are +Prepending file in the baseDir also did not have any effectIn the end I prepended the full path in the file to load the images correctly +its just an example +Reopening the bug based on the recommendation made by Fang in her comment +taking logj appender route or other +kill +I will apply it and test locally todayShould we have not reverted the previous one and applied this one to the previous version +Interpreter service is integrated into airavata. +Sounds good +This issue is relative to. +scheduled force dont care which way I have to do it but some way would be helpful +Is anyone working on this? If not maybe I will pop the code open and have a look. +Unfortunately the logs without TRACE do not provide enough info to investigate this +The patch for branch is ready please help to review and follow the steps below to commit it +the moment after the first checkout I modify the default build definition of the project if the project have other build definitions they arent affected but I dont think a user already set some other build definitions before the first run +The changes related to cleanup as a separate task looks good to me +HiI tried to simulate this issue with and I miss messages messages are in delivering state are totally lost +A release note would be good +I dont have time to work further on this issue +r +I suppose next step is to pass on the data to some algorithm +The +In that case some folks dont like subclassing and prefer an interface and static methods +From the user prespective this will not make any difference but interally the whole size is reduceEarlier the size of for dojo tree children was chars and not it is reduced to chars for my Geronimo server +is still a Writable? Is that because Exec needs to go over still GaryOtherwise Im not whats going on but patch looks goodThanks Gary +on the patch and also verified this fixes the bug +Ok i will try it out and let you knowthanks +I have been planning to work on this in a while +groovy filesCommitted at revision . +Thanks a lot! +This patch incorporates all Konstantins review comments +The thought being an equality check between objects of different types should return false and not raise an exceptionThe unit test assures a is not thrown when comparing objects of different typesFYI by different types I mean objects that fail an instanceof check +This works fine after resolvinf even if the suppress duplicates algorithm is enabled +Moreover it should allow far more efficient seeking which should improve performance a bitCan performance improvement be quantified +If this doesnt make any sense on the other platforms the platform should be removed +Applied branch and trunk. +One thing I want to be certain of though is what jar dependencies would be necessary for your implementation and whether the connector you have built is indeed as as the one it would be replacing +I understand if youre puzzledI wish you happy hunting +Does it have to be public? You should usually be operating those objects as instances of the DOM interface rather than referencing the implementation class + Bug has been marked as a duplicate of this bug +This is not small because was used a bit everywhere by the CQL code and since thats a refactoring ticket I used the occasion to do a few renamings here and there but overall this is pretty simple stuffs +I committed this +The patch looks good to me +failed tests exist in the trunk +This passes the current tests +I think this is another good reason to bump the version of debhelper used to build the bigtop packagesJames if you think you can help us get this effort underway thatll be very much appreciated +Since its Brians deployer that spits this outIll assign it to him to have a closer look +Attached a first draft +Do this first for! +just ran cassandra tests from trunk whose default test uses and also the same ami without auth problems during bootstrap +This also happens on Weblogic SP running on Redhat AS and NOT resin guess I expected thatp +I dont think this applies to the new client. +Hi JeffThe JSP stuff is still about a week out +Just putting it on my back burner for now +In addition for people it would also be good if a project can just reference a few on the file system or in the workspace which contain annotators or resources +If they want reliableconsumer failover they need to enable the inactivity monitor +Not affecting later branches +Hello PaulexThe fix looks good +bulk defer minortrivial issues to +This changes to want update and depricates the other flags public boolean overwrite true Deprecated public Boolean allowDups Deprecated public Boolean overwritePending Deprecated public Boolean overwriteCommittedIf any of allowDups overwritePending overwriteCommitted are set it will use old style calculations +I guess you can close this issue as Not Our Fault now +this issue is annoying +Thank you. +Thisd be a funcrazy noob project +If would be necessary to rerun the profiling Ill reopen the task. +Please verify +That the installer switches to command line mode if the X display cannot be anyway as Julien says if there is a more suitable Groovy package for servers then there is perhaps no need to do Robin +And additionally committed to updated. +Deferred to CR +So I suspect even though one level of currying was working if the of params was wrong then possibly further currying even prior to the recent changes may have broken things +Metadata is now only supported by the and youll need to cast explicitly in order to use it +Peng Can you please reupload the patch and grant license to apache for inclusion? Thanks +One thing that is missing is as per SQL spec in a comparison operator as far as there is one operand with collation derivation the comparison should work +Turns out it is slightly more complicated given that we conflate principals and keytabs +Running jdiff should give us enough warning on this. +Committed to trunk and branch +was incorporated by this does the profile service support the previous netboot functionality? If the profile service variant of netboot is already supported in AS where can we find documentation about the configurationThanks in advanceDirk Daems +The dependency to was changed to the new version +Reporters of these issues please verify the fixes and help close these on BuildAdded an instead of ANOTHERVMDCclus and it worked. +Unfortunately when I make this change I can no longer run groovy scripts in the Java launcher +Yep +I could not reproduce Closing for now +For clarification Accumulo does not support negation in the visibility expressions +The user has can change these +Patch containing the image in SVG format +And thats on just to make sure. +How about we mark it down a bit as critical? Lets revisit this after we cut the RCthanks it quits! for this bug +Changes svn statusM srcref +this one will be too hard to fix in i think +Other web frameworks with binding capabilities will instantiate the nested object graph as needed while binding +NiceFrom now on I will add unit test module whenever make codesBTW Asymmetric matrix could be appeared +Quick +bq +Any ideas on this one? Should we start looking into Eds Avro patch in order to start upgrading libraries +because it looks to be the same algorithm to me +Please reopen if you still experiencing the same issue and can provide more information that would help us to reproduce +John Which of my messages are you talking about? Also whats critical mass on this jira +Now that we are considering monthly releases this wouldnt be an issue at all +For parsing variable index number doesnt need to read POINTERSIZEINT value +Freddy It is not clear to me if we only need to add the depreciation link or if the behavior of the rule must be seems in this case that behavior has been updated +The add script basically contains a stack upon which new events are pushed +The first commit corrects the spelling error while the second commit corrects a compiler warnings from recent commits +Just need to remove some proprietary pieces +I dont have a test case +This is so that Wal and arent brought in as dependencies +Youre going a long way to keep also already did this on local working copy and got stuck with JAXB API I find the time maybe I can be of any again +I thought I would try this with DB and found that after the global transaction commits the GLOBAL TEMPORARY TABLE was no longer accessible +final will resolve dependencies one time and much more like other plugins +Thanks Harsh +I propose changing the name to something more generic solrclean or solrpurge etc +Stack Thanks for reviewing Ill post a new patch shortly incorporating the with missing files +You probably ought to be generating a paragraph though but to be sure best ask on the POI user list for requires HWPF document with you Nick I didnt know that the Closing Issue comment goes here +I think should not be starting a new segment when ending the last one failed +Actually I did an automated process for the help And then about Ivy doc I opened. +But yeah for completeness sake we should probably add for the review. +Even a MB seek will read MB data that should be skipped +Use of temporary files plus data seems to be what makes the RSS connector significantly slower than a file crawl +Resolved with recent change +they are all CDH based +I like it the alignment looks good +The final fix on this is dependent on +Ive exceeded my and have to now get back to working my app +bq + on the separate build targets +upgrading priority for release +In the case of the MQTT code as it currently stands when the subscribe occurs the code uses the client Id as the subscription name as well +I see the idea of splitting up between cleanups and essential bits was a good one but it looks like the patches really both contain essential bits and cleanups +Please open new JIRA where we may come up with better idea +Closing. +I will do itSo all we need now is a GC before after screen shots +I think it makes sense to move the extended elsewhere and remove this test to avoid confusion +I see that youe updated the names of the jar files in the manifest +CSS dialog and CSS Editor used the same model so it changed when user changed something in dialog its changed in editor too +Patch from committed +Fixed patch formatting reverted and committed again at revision +Tests pass cleanly for me on +I started reviewing it but sidetracked +Thanks Gary will first test with the +Currently you dont even get anything in the response thats useful you just will find something logged on the Overseer +Deleted the test it has been excluded for over years and the functionality is covered by other tests. +I think the new patch is the proper behavior + +Merged in upstream by Jason +Fixed +patch committed in + is the current names as they appear in WF using +I dont think it is necessary to use for +This way the key buffer would be cleaned at a proper moment in time +Looks like is currently being pulled in transitively and is not dependency managed +Thanks for the help Darran +The pb is a bit different that what I thought in fact when the config does declare the interceptor its not loaded +The nfs related problem also seem to be related to what we do in in the local mode +Discussed in EG meeting there is not really consensus in the community about the right way to resolve this and were running out of time to add new large features to CDI +Updated patch +I built the jar based on the zip directory so well have to clean that up a bit +No thanks Ill send in a new patch shortly to hopefully close the Issue +Triaged Sept +And it is not planned to port allthe features to the jdbc pm +The patch applies to trunk and and +Milos we spawn JVM for Maven launcher and set current dir set to pom location so it would be kinda surprising +I think it is not good stream +Another problem in hyphenation +The only issue I ran into was a significant number of maps and reduces failed because they timed out trying to establish a connection to the AMThis is new +the build number isnt a try it only happens on d like to consider the revision number as well though in Continuum we might have multiple profiles and only one of them should be able to use that for counting +This is what you get for mixed content +I think calls this a Checkpoint +Attachment has been added with description I have attached the pcl with underline issue from linux OS +We are not ever going to use EMF +Im stil getting the same error with +Problem with the Gant launcher on Windows +So could we have fixed this ourselves by using the Action Delete Cache +This patch has a protocol change +Please attach individual files instead of a big zipI dont really understand this so Id like feedback from othersHowever I dont think we can put a k JAR file in the Thrift distribution nor make it a dependency +Attached patch +Reviewed by Ramesh. +Ill look forward to the release of MRRP! Thanks again Elliot +This test case fails when files have been removed by a compactionNot to forward port Todds fancy WAL writing +Maven project to duplicate the a bug iteration component specific you should explicitly wire component state to the model beans +Thanks Tim. +thanks +Im not happy with having a implementation in Jackrabbit +shipped in +Hey EugenePlan to address Todds feedback and update the patch for trunk +Fixed +Is there a way to fix this +Committed thanks +Here is the workench Ready to use and to reproduce the error +Hi RaviThe Geb plugin doesnt depend on Spock at all so I am not sure what you are referring to here +Looks much better +I applied and committed the test patch in with revision I dont personally feel the need tointend to backport this to although it probably isnt a big deal if someone wants to do itThx for the patch Mayuresh. +If it is actually implemented in Derby then yes I think the code should be removed too +As Justin already stated this is nothing we have to have in jira. +Do you have an example of the config that causes this bug +dump is the output from an attempted runThanks like this is because hasnt been implemented yet +fieldName postalCode notNull sizeMin sizeMax add field string class +I need this package to easy configure SSL and i dont want component and this dependencies +Is there a reason why the default mount for is +And also with a postgis based layer there is a bug in how the position is computed but a different one +Larrys First publication of this version is deeper in meaningWed need further questions on what a version is +However it doesnt work on Mac OS w GeronimoMac OS G works fine but G does not +migrated to JSF facelets am closing the issue +Increased memory +But ANTOPTS +Then when the redesign comes out your code can switch +Thomas reported a flaw in the use ofReplaced usage of with plain old synchronized methodssvn revision +As Kevin commented this is a jUDDI issue therefore adding also jUDDI withing SOA component +Yes +leaving it open in case it still makes sense then +multiple are created instance on the index side and on the search side so its better to pull all the setup you can out of the Filter for performance reasons +WDYT? Will you locate and fix all the plans that are affected by this change +and versions of Nings patch are attached +fixed in current head +This is still valid when trying to install from apache machine there seems to be a timeout and when downloading the big jar the connection kind get timed out and the jar always get corrupted +Fixed thank you! This will be in the next SNAPSHOT and Release +This issue is closed now +So the scenario is correct. +for tests I use with Windows bitsIt look like there is deadlock somewhere +Can try checking the repository if its up and running the project creation process again +So all issues not directly assigned to an AS release are being closed +Reversing order of commitJob and storeCleanup calls in to maintain behavior under +If you finish it before we release beta please correct the version +Ah there was still a bug in which I have fixed +Nightly builds are working again the doc should be up there wait for GA to include it in the build +This seems contrary to what the schema permits +and that theyd have their own means for exploiting the newly added APIGo ahead then if you think it important or useful for +Plan is to Break process into unit action processes +I was aware of the persistence problems in former releases but the current version works fine for me +I totally agree with you that building a child project will not be possible if only the toplevel pom reads the properties still think that the patch is valid the promise of the properties plugin is to behave the same way as if properties were specified in a POM file +bulk defer of issues to +When crawling a terrabyte of data it is important that only changed data be fetched and parsed +Looked at smbmapjoin query files and updated smb join queries +To simplify testing you should be able to use incremental s server components but s gems thats what Im doing on right now +Verified this on the latest build from ACS branch +By as it shown at demo Ive meant data model implementation. +Key is to traverse pojo by reachability to determine whether it has been accessed or not +bq +Ondra Please read my comments have already tried and clarified that suggested step does not work on EAP though it works in AS upstreamTo be more preciseonce again currently we cannot run from root in EAP without using that goes through the tedious process of checking each test even though it does not run themWe also cannot run single test from testsuite directory or subdirectories in EAP at all which is possible upstream as it ends up with dependency issues in subsequent pomsQE really needs this feature to investigate and test as its simply not possible to run entire testsuite everytime +We recommend that too. +I would like this in +Im testing it on Juno JEE RC build on linux bit +Could you please indicate the version that is affected by this and how to reproduce this problem +Great +Implemented proactively setting the content length in the resource metadata in Rev +If returning false does not cause the socket connection to be rejected then your suggestion is fineI think the four methods may occur in other classes for exampleIt may be best to grep the whole source tree for occurences +Unless there are any objections I will commit this later today +New file with the return type fixed +In any case I think the suggested change is an improvement as Im not very fond of prefixing fields with super unless there actually is an ambiguity +addresses that issue +Thats true but that also makes it more obvious that we could potentially have data flushed postmain truncate whose indexes get truncated so the index wont actually match the data file +Another suggestionrequest +added property isRunning testing for +Possibly due to some trunk changes not merged into those releases +ThanksPS I dont understand your reason for not having a +Patch attached for review +Yes I like that suggestion +Committed to trunk. +I have also found this to be a problem with some services I am trying to make use of +cordova version +Should be +Added a simple and also tested this on a local cluster +Closing issues after Configuration Admin release +Its still confusing because user is able to select different JRE using Open Lunch Configuration dialog but JRE is not user point of view its bug because UI allows to change JRE but it has no effect at JRE tab should be hidden or at least disabled because right now it has no functionality +The live site still needs to be fixed +Off the top I think it ignores the security aspect completely +Ive just committed this +In the branch used to have the following for the path for the empair model resourcesorgapachectakesrelationextractormodelsresourcesmodelsempairThat path had extra resourcesmodels within the path and was fixed in branch to be resourcesorgapachectakesrelationextractormodelsempairI see now this is not an issue in trunk and apparently never was from comparing history of branch and trunk +Ive committed this +Its easier to do it in this direction +That should be available in the maven logs directly if either e or X is enabled. +I am out of town right now +could be using it in lots of places including the unit testsAs far as the rest of it I think you are on the right track here +if simply removing the code that depends on it is an option that would be my preferred solution +Just to make sure Im clear on the usecase here for +Two thingsAre these about my comment above +too +Attaching screenshot of Tests results AFTER the part of the patch is fixed but BEFORE the part is included is created but duplicated elements are found +Plain jBPM performance tests implemented as ESB service +Ill try this out +Im not able to use the client credentials grant because my client is public and I need the implicit flow +Ive added code that allows to split the into the basic part and additional API implementations for +Yep unfortunately as Lewis put it they are horrible for searching +Martinwait for one minute then press the button on the page then after receiving the expired page press the browser back buttonAnderaI get this on +Reassign if needed +At some point we need to do a serious cleaning pass and not only update the readmes for all of these examples but consolidate them into a more consistent place in SVN to move them out of peoples personal workspaces +Committed to trunk. +Ive committed the patch. +Thanks Hoss +Wow that was quick +Workaround is fundamentally for EAP mentioned in are correct +I emailed you the project which is M the limit in JIRA is this is with jboss +What are adverse effects of not having the parameters specified +Im attaching a reduced test case that demonstrates the problem +Ill probably need to check with Manik on this + provides a patch on to handle large object size and also adds a new test case to the existing to test it +A bloom filter would help hereYep and this is exactly what I would expect +Ive assigned to you. +Well do the docs on this +The fix works for me so I finally close this issue. +This is not available at runtime it only produces the +Now that work is underway with AS all previous community releases are +Also those messages look like or we have a strange logging format +we can break api between major releases so its not necessary to keep it as deprecated +This change should be applied there +I cant view that link +Thanks for the quick review committed +It wants they in +but i download the final version and not work this patch +I have also moved the subscribers from the dev and commit lists into the maven lists +Going back to the reason why this issue was filed the wont fix or +By rolling version on demand I just mean there is an API to create a new version of an object and keep the old oneMy feeling is that thinking explicitly about versions rather than snapshots in the context of this design will simplify things and make more understandable for usersEspecially since the design is already based on versioning +Currently it says RB +ThanksWhat was the problem +Betting this is a duplicate of but without details I cant be sure +Since JNI is not possible in I triedto convert the code to Java +Oops SVN was slow and I didnt noticed a conflictIts really committed now +I dont want to build the release with a beta tool +Thanks this example is bit easier for me to check out although may take a whileFrom error message it is evident that deserializers understanding of what should be included is different from serializersAs to using same resolver this is mostly due to historical reasons whereas almost everything is separated between serializationdeserialization side and it is bit difficult to actually share things +OK pushed to master +Dear TimWe have increased the heap size for activemq +This is my current is my is my is my facelets Interface just to be a priority now that we have sfileUpload +For some reason the deserialized element qname differ from the expected element which make the data binging throw an exception +patch synced up to trunk +Thanks for keeping an eye on it. +I think it will be fine wout the conditional update it just means a bigger if the consumer dies just after rebalancing +I even tried updating my JDK version to the latest version of and it still worked +Ill add back in the versions this evening +If so why did you report this as a bug against Xalan? Or did this really happen with the old DOM? And we dont use either DOM for parsing stylesheetsIf you werent doing a transformation with Xalan please transfer this bug to Xerces or mark it invalid +for me +All the Datanodes have no problem talking to the Namenode in my cluster of mixed machines +It would also be nice to close all of the readers when the core is closed and not just on a reload tooThe whole spellchecker thing looks a bit fragile actually +I apologies for the inconvenience +Maybe parseContents insteadAnyways very minor issue go ahead as you see fit +Ive committed this to trunk and +Please rebase the patch against latest trunkThe changes to conf keys look good +Running it on is not supported. +For example a path of the form hbasetablenameregionnamewhen a master assigns a region to a regionserver it makes the change in zookeeper and then issues one rename call to hdfs to rename the top level region directory name to the appropriate regionserver name +Please could you verify that it has been applied as expected +sum file lists the tests in order of run the file looks like it reorders them so it is confusing +We need to add cglib to our list of libraries +Commit if you tested it +Looks like this breaks configuration parameter is a url? configLocation is never used if new URL succeeds +Spark text components default to using which has a different IME support logic +Uploading a patch which removes average aggregators +I committed this yesterday +This version incorporates feedback gathered over the past two weeks Introduces a new value for +Means somehow some classes exists twice in your classpath +Please check that it was applied as expected +patch looks good +Or is crawlparse using BLOCK compression too +Applied patches thanks to David Le Strat +I dont think well change s output to match in any case but Id like to hear from them whether this is an intentional change +The content of the new directory are new files that should be added to the project the modified contains already existed recently modified files +Go back and do that +It was giving to time out when the solicitation was done for the you +This bug appears on jvm but not on jvm +So maybe thats the problem here +Rob Cernich oh sure thanks for reminding me +Closing this. +Also ran findbugs and both pass +Committed revision Opened for the params via +If that works try upgrading from to +has the requisite headers +is it more expensivecomplicated to do it for an empty sliceempty result for entire row slice means it really will be gone when tombstone expires so the two are semantically equivalent +Is it possible to consider it as an enhancement and implement it after implementing HDFS Apends? Wont the system be simpler if we avoid this optimization at first go and then once HDFS Append is committed then we can work on this optimization +thanks for the patch +Ive just checked out the trunk with the updated and built a snapshot +I cant reproduce the problem +package resolves this issue. +Could just be an issue with winzip too Ill try another utility +But since the markup is cached the markup with langen is returned +also I lowered the priority. +The criticism of the patch was in my first comment it is a regression +Will be fixed on +The most annoying part is that everytime when the exception is thrown the lines are somehow different. +With the old full names format an arbitrary chunk of the file could be lost or corruptedOr even a bug in the codeMatts proposal reintroduces some of robustness of the old design +Going by the later one I should close this issue here as wont fix +Please see my comments about the attachment for details +prefix is used as an extension mapping? A string containing only the character indicates the default servlet ofthe application +Lets make sure we dont duplicate the functionality +Thanks Anil for your comment here It can be considered as business process according to which A reconciliationid cant reconciled if its any transaction have Canceled statusIs that right +And even has this problem +Is there anything the rejoining node itself can inspect locally so it can determine when it was last part of the cluster and use that instead +Committed bumping the last digit of the release id now that the release is out +Whats required is the decision on when a container is ready to be used by a different vertex and locality implications for thatI probably have an updated version of the patch somewhere from when I worked on this last let me see if I can find it +All SVN changes applied +I entered the Jira more for documentation than anything else +bq +The fix I am attaching is just for the types branch +Please review test in SVN trunkOleg +It did not apply to +My problem went away +Simon can you take a look? I looked at it closely and did run testcases again +Can you post the exact way that rpm is invoking cmake and make +Does it look good nowThanks +fixed tested so closed +Pushed to master. +I think it is needed to avoid competing assigning the same region at the same time +Hey Hairong +Let me know if you agree. +My only comment is that I think the test of and the test in this patch should be perhaps in the class for organization purposes +mark this as cannot reproduce because the reporter gave no more input and we was not able to reproduce it. +Updated module versions to removed dependency to removed stratos agent related modules from their parent filesChanges committed with revision upgrade was verified with a clean maven repo build in trunk. +Ive just committed this +cant be +This required setting the Karaf features validation plugin to Java +I tried to make a subtask of and got error Subtasks cannot be a parent issue +Is this created by mistake? This seems like a solr jira +Not sure of the proper fix for this yet +One other thing +Everytime will set and NM will get and use this interval +However the test is now there with Ignore removed +Thats unfortunate +Legacy +Now generated code can be compiled by both ARC and compilerI found the code for const initialization is so buggy and ARC compiler complains about it +the question is what domain should we use as an example is reserved for use in documentation by IETF perhaps thats best +changes are atSending actionsrcmainresourcestemplateajaxSending actionsrcmainresourcestemplatesimpleSending actionsrcmainresourcestemplatesimpleSending actionsrcmainresourcestemplatexhtmlSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspuiSending actionsrctestresourcesorgapachestrutsactionviewsjspui + +Marking PA for Andrew so that Jenkins runs +Looks like that wasnt the right log +Correct +Id like to see precisely what you didIm going to close this as it seems to be a compiler encoding issue with compiling the code not really CXF +Thank you Mayank. +OK Ive revised the resolution algorithm to disregard redundant mapping information for ambiguity resolution +Test with and tip of tree this is a status bar issue and might have been fixed with our viewcontroller on on retina and +Added an additional error check in to report the scenario +Attaching a patch to address this issue +the following sysinfo script +This sibling queue listen the same and good messages is not have the same problem +this patch removes many unnecessary methods from +For the crc file only the last bytes are allowed to be overwritten +After drilling into the code pack is called +Steve +As said yesterday Im gonna verify this on master +While the exception throws start server from outside the bin directory which is actually cause far as I understand the proposed patch it makes it mandatory to specify JBOSSHOME whereas currently that is optional +This is twice Ive seen this bug so I made false the default for safetyUsers will still need to configure a warming query though. +Juergen +for others following this issue what is the different way? +changing version. +done +So all issues not directly assigned to an AS release are being closed +I have added this to the Services Guide for. +For Leveled compaction we do this with the code +at r +Ive verified this fix by iterating through from Pointbase using Data Grids and is now working +However it should fail if is not present at all +file +My was that if any guys want to write another filter which has both filter and filter impl in future then it will potentially have a bumb maybe rightTo me i prefer the current fix in than the specific filter classes +Yeah it wont work for +It passes all tests in conf +Ive deleted my local repo tried to build showcase again and it works. +At least it works for me in the overridden Query class. + you should bundle them all in a jar and put it directly available in a web I wil provide a bundle later then +See progress view with running jobs +WowIll look at it more deeply this week but looks pretty good +Ill spend more time on this tomorrow maybe it will become clearer as I read more +Patch is ready for review +The only word of caution here is of provides are removed then Requires need to be aligned so meta packages like hue do not break on install or upgrade +Else shold we close +Assuming that the leader has enough followers supporting it any follower that gets stuck will eventually be upgraded +Contrib tests failed because Hudson couldnt pick up the task +But at that time we decided against that not only because it is more difficult to implement but also because it makes any performance increment in this area very hard to achieve +Another possibility is to rethrow a new with OOME wrapped as the cause +Ive also tested parsing of the new and elements works fine +It has never dereferencedPerhaps a bit of context might help +all data are stored in x including the mysql database +documents a new bug found when this bug was fixed. +Done as specified. +please verify in or in +Hi Hyungjun KimIve finished the review on this patch +Has Chrome been released yet? Ive tested in the latest Chrome and couldnt produce the errorWhich version of Opera are you using +please that the file has been renamed to +This was already available in WFK development believe this was already done in WFK +Please test again +Marking closed. +With the GC settings we have given it appears the JVM is allocating threads for GC +There is improvement FOR UPDATE required for updatable result set to work filed for the very same issue and it might be worth the efforts for someone to look into possibly not requiring it +This was caused by the changes in +Sorry I thought submit patch was the way to attach a patch +Thus if datanodes die during +I wonder what changed to make it stop working? Please keep us up to date on your findings +Patches applied to api and +No problem +Naming also adopted for latest releases. +If someone can point me at where in Shindig features get concatenated and minified Id be happy to produce a more general patch that applies the innocent code transformer on all such for +Patch to adds a link field parameter which allows highlight generation even when uniqueId is not defined in the schema +So please go ahead +Agreed +I suspect if can be fixed then this feature becomes available as part of that t work with macros only with Java method calls +File necessary for the junit test +Resolving as Cannot Reproduce +Closing this blocker that cannot be reproduced and reporter has not replied in over a year. +Ive repackaged the code in prep for Mark to submit it but waiting on Amazon to figure out my SNS access issue to confirm tests +How did that happen +More work +ok this day February th as +I consider it as my mistake for some reason I didnt wrap this method into try added extended catch for create methods to cover exceptions like in SVN. +I linked this issue to related jira +Hi PermaineIm not sure which testcases youre referring to +The attached diff files are the conversion of the Spillhash harness test to +When I remove signal event and converge gateway it works OKSo my guess is that in this particular scenario process persistence does not work properly as jBPM tries to serialize workflow process instance object +Fixed in revision this is fine for me. +Let me think through this a bit more +still seeing it in just a warning it can be +This patch will use that class +resync with trunk and include changes ofNeed to do svn rm manuallysvn rm srcjavaorgapachehcatalogclisvn rm srcjavaorgapachehcatalogcli +rar files has the directory structure +The previous patch conflicts with +Can you prepare a quickstart that reproduces it? Or at least explain what are you doing when this happens in your application +We might start with just the simple version +Add example of a not found that occured on a test run on trunk +A lot of code was broken out of the WMS and placed in modulemain in the package onto trunkcommits and +Starting with all integration concerns only have a dependency on +and unit tested! +To clarify the above comment is more in regards to the leader goes down case +All entries in the list refer to the same instance +Im having a hard time understanding the way to make use of this fix +I am seeing junit failures in this test when run on j see +But either way I think I will add them for you +Not sure about the property names as we dont receive anything from the broker telling us what it supports we could hard code them but not sure how useful that would be +If all intervals have availability data then all dots are displayed +Attached is the updated patch +I will write a mailto laterJDK is OS is RHEL update + +Everythings fine +I committed the above fix in trunk r check if the repo location is referenced in the system properties before deleting the repo in the file verify if the fix is sufficient enough to address your problem +Hi MirekCan you upload the mysql persistence config fileThanksHoward +This patch looks fine and Ill commit it +Please go ahead and apply the patches if you couldThanksSamisa +Try uninstalling and reinstalling +The results can be shown as a graph +Changed description to reflect actual do to workaround we could put the code at Bootstrap +see attached mvn output log +OK I confirm that Clover reports are now being generated +We thought the health service should not mix with child tasks reporting status and hence kept it differentThis is disappointing +Take +Yes I can reproduce the windows build failure with the original patch by Andrey and modified patch by the patch has been applied properly +though I would spell ESPCAPE differently Should I post updated patches +Is there any update on this issue? I am experience this as well that the integration tests are not rolling back the transaction for objects using the datasource defined in the datasources file +Gotcha +However these are bandaids to a larger problem which is multiple and how to swap them in and out +Yes both client and server should always use the later version +Do you see this issue with which the latest release +Just send mail to derby user list for questions dont file JIRA issues. +DavidI will have time to test this out next week +RMC is deprecated and I would use +is this hard coded +Please attach the full log not just the part that has the error +I think the use case there is a good idea and I dont see a better way to implement it +This works with all tests +ajaxSingletrue is equivalent to executethis in +patch supplied by Peter closing all issues that have been in resolved state for months or more without any feedback or update +it needed on average shorter timeWowIs this because this particular benchmark has many open named statementsfor the connection? I guess that if the connection has open named statementswe call on average times to find the one we want? Does thisindicate that we ought to have a hash table here rather than a sequential lookupIm a bit surprised that doesnt itself make this optimization +add the which is required in the jdk +Maybe null namespaces are treated different then empty namespaces somewhere in our meta data model +I am leaving this ticket open until i add the erlang version check to configure to help guarantee a known env for erlang +John I found the problem +Looks great +Others I think who know this quite well are Todd or Nicholas +New patch that works with the changes in +I need to test this particular scenario to confirm this thoughDo you have any sort of resources to reproduce this issue like the configuration and the synapse configuration? That would be really helpful to get this resolvedThanks be addressed in the release but this wont be available with the release given the fact that we need to get the transports release out soon +For this latest update I am just sending the data as it is presented from the data can be reformatted as required by the consumer +Thanks Yongqiang +The flush order is based on the descending order of region memstore usage +Sorry the project +Can you specify the complete requests that you are sending to Solr +Expiration cache is fixed not to return the expiration key as the client entered cache key +Fix for and +What do we have to deliver to support real virtualhosting +I committed this to trunk and branch +It is a different file +I didnt test this patch in RI + I suggest we change it to int for both methods since a length bigger than is not realistic anyway +for the WONTFIX +Alternatively instead of propagating Refresh buttons we could focus on determining why the current refresh isnt workingWhatever way you think makes more sense to pursueThanks +Thanks Amareshwari! +In the current revision of the documentation this is the caseWe should however check if my renaming wasnt too zealous + +there are still some Writables and These are wrappers and not actual writables sent across the wire +I also see several NPE in in the unit testsUploading a patch that delays the instantiation of +When youre just getting started bijection is explained in terms of injection and outjection something that I think should be reflected in the annotation +Also seam extends usal EL and we should use our validator even for EL in jsp +I havent checked but if trunk is indeed missing these tests and its not that they moved? then yes we need to port to trunk +Thexml andhtml are no longer defined for projects with empty sources +fixed +It should appear in the nightly +But if we do need to compact a higher level well first check and see if L is far enough behind that we should do an STCS round there as abq +Rossen did your example use subflows +Aaron those +This change should probably have bumped the minor rev +test but a manual test +Yes this is on the quad cores +We need to close this and documentation should state that requires +Please verify in EAP documentation on stage +Regarding branch issue it has to do with +Next time please try to keep tabs out they are set differently in different editors. +If any bundling happens it will be in or timeframe not. +Also one important thing is missing in the previous patch the schema dependency is needed in the axis config otherwise youll get CNF errorSorry Lasantha dont mean to step on your work but this prob is too serious and prevented me from testing my wsgen stuffclient invocation is still broken and can be tracked in a seperate jira +Made a start and checked into develop + +Any news about this issue? Please let me know if you need any more informationThanksEmily +We can add the new suite class at such a time as we have another test +Will depend on how complexinvolved the fix is +on patch +The comparisons here are only for CPU consumed by +Im using product build + removes the variable and the code that uses it +Fixed on trunk and in branch +My initial investigations were wrong +Currently its there in both +I think your previous comment needs to be mentioned under release notes +Tentatively marking as blocker per discussion on to track the Hadoop IO classIn the long term and should both be refactored to use a serialization method that doesnt add a hadoop io dependency +It looks to me like you can safely reapply the patch that generated r call this fixed +I tested but the issue is still present with the just released SharedPostponed for a next version +If the map output is already sorted this saves all of the costs associated with the shuffle and the distributed sort +I actually added this to the enterprise branch already but instead of a specific I just created a +Attachment has been added with description Example SVG file demonstrating the problem +Caching for UGI instances +Since this is all based on the clock at the master I don think this is really a problem +My first reaction is that it would be OK to allow Appenders Layouts etc as Spring beans but not Loggers +If this is not convenient please explain why otherwise Ill close this issue +What is the status of this ticket? Has it been back ported toThanks +With Lucene there is really no slowdown at all caused by multiple segments as each segment is searched on its own with no interaction and just the results added to same priority queueDo we have benchmarks for this in some issue would love to see some numbers +I will give it a go +TA I didnt know that beans will be always serialized to the session nowYes that how things are working now according to the new JSF spec +rename flat to dump +This somehow got added back in on trunk +Retry now build has settled +Yeah Otis +Im working on this. +Im passing the moved files collection into the +in trunk the impl constructor is public +Address for Duxbury wrote At the very least we should not assign a region to a region server if it is detected as no goodThat is an unfortunate wording of a log message in the Master +datanode warning param argument state is not a parameter nameThe reason failure is that the dataset needs to be updated after restart +To obtain mass removal a user can use findObjects to retrieve all the current managed pojos under the base name fqn +One reason were not seeing this issue more is that most tests are using the proxy which already imports the framework at version not version +Committed ref Thanks for the patch. +One thing that looks missing is when adding a class into +The current visual representation of HDFS heatmap +These releases are not off of Apache +See +Wont review board throw an error if it doesnt apply cleanly +The advantage here is that its pretty simple and already written whereas is going to be a much more involved +I already fixed the first twoThanks you are correct in all of the aboveSorry for a I need to tweak my settings of my specific Synapse workspace regarding line length +So it appears from conversation that +GA as well +to have logs per region with has not been active +Happened again with r +Hopefully I created this patch file correctly +Did you try it +I wouldnt have asked if this wasnt necessary +Tested the new projects for basic functionality +Ok I agree with you thats quite too specific to introduced this feature but what Im telling is that the problem of data migration between environment is a big problem that we have too face every time we design applicationsThe replicate feature does not hit the identifier generator during insert of new data therefore when you try to insert new data in a normal way it can happens a database constraint violation because you have previously insert new data during replication using an id which was generated in another environment +In addition to the provided test I also tested this manually on a cluster by hitting the jmx URL and observing the values shown there for the new metricsI implemented all the metrics above except for the followingbq + +I might have overlooked some code so we might reopen +Once to check that the document has terms in its vector and once to add it to the index +Ever +After all it can be set on the command line with D just as well it means that we dont have to change the code reading the params etc +If you finish it before we release beta please correct the version +Thanks Paul +project to for JBDS +We highly appologize to our users if specifying wrong authMethod or transportGuarantee field values resulted in endpoint now such wrongly configured endpoints will not be deployed to the AS and exception will be displayed at deployment time +Thanks +Added a pull request for the test case which clones and modifies binary content across workspaces in a disk into the master branch +You can contact me on my private email antoine at gmx dot de directly too +Its main purpose is basically to avoid passing a very large Store interface for things that only need to ask for a couple of numbersRemoved external visibility of compaction configuration +I have committed this +This fixes something broken by the patch Starting with the previous patch the loading of the toursdb tables stopped working because the script ran other scripts using file names relative to the current directory +Committed trunk revision Committed revision Please reopen if we should backport to or +Im not sure which implications this can have but any queries that reached that point in the code would have resulted in an exception being thrown anyway +Moving back to CR as the core of it is done and it should be in release notes +This is actually to do with a concept involved in deleting a user created by an admin +Yes access via intellij is now working. +After a few test there seems no problems here anymoreBut have to test it a bit to say that i use the latest Trunk with java for +So confuring an with alwaysIncludetrue doesnt seem like too much work? Or are you talking about defining more than one for mixed forwardinclude behaviorJuergne +Marius pull is merged +HelloThe patch fixes the exception handling convention problem and also removes the duplicated test codeWould you please have a try? Thanks a lot +The bug is not reproducible with the latest trunk build +So the is deprecated +We should also setup things so that either tests running on different branches hit different databasesschemas the tests do setup some lock to prevent other tests to run in parallel with them we change the hudson build so that no two builds run in is the rationale behind having different branches run on different databases? Since the test cases always tear down data before a test should not ensuring two builds do not run in parallel be enough +back ported to branch with minor conflict fixsibmsvn commitSending javaengineorgapachederbyimplstoreaccessbtreeSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawdataTransmitting file dataCommitted revision +It wouldve taken me a while to figure out that I should look there so thanks for the suggestion +Would it make sense to refer to that for info about the engine +I agree with Arnaud we should try to make those two code work Im still a bit new to MINA but I did a few searches and dont know what exactly you are referring to byIs there already a http codec that you are suggesting we insert this in to? I have currently implemented this in a way that it is +All tests pass +I dont like having two different floating around the code +r to make that changeSo now the CMIS connector seems to be working ok +bq +when i create the same name realm Realm page will go to previous page no any remind about error +Canceling patch +Did you verify that it was really a problem with the test and not the class being tested +I can confirm that your patch produces correct results from the script attached to repro and from the script attached to this and I have been talking about different possibilities for given the time critical nature of it +bq +Agree that we deprecate Text only +Currently the diagram does not support collapsed only expanded + +Fixed +WDYT? Nested sresource to go +Patch applied +has been released closing issues. +is it going to make? +I couldnt wait so I cloned master and built it myself +adjusted the thresholds for defrag and sleep further refinements to delay unnecessary defrag and to do a targeted truncationdefrag when memory is exhausted. +Yes it does +Unless we take all config properties and add them to be System properties this cant be done automatically +bq +So there is no need to explicitly print out clientid correlationid and versionid +Couse really I dont sure is it right way at all +I am unable to reproduce the failure on the current Harmony snapshot +Http request response and stacktrace +I just committed this thanks Sourygna. +The class proxying with cglib can lead to errors + for the change +A patch which fix the problem +Owen I think you just won the debate! In Ant you have to type Dofflinetrue +I hope it will be available as a point release soon and with that we can fix this issue for the on +kick off a new build and try again +This bugs happens only in command line +Hi AviAfter this ridiculous delay I finally found your ticket againIve your patch but it doesnt seem to cover all cases at least not anymore +The resulting error message may not be the most informative possible response but is correct operation +Please attach small Eclipse project that would allow us to reproduce your issue +Manually tested! +Obviously this is a type of issue at this point +The list of requests to filter should be the same for trunk and except for reconfig so if one has check I think the other one needs to have it too and trunk has it +Joseph as gratitude for the effort youve put into this I spent a chunk of my day giving it some eyes +change this issue to create the theme for the form widget I think they are the same waiting for Jacopo + +Youll also get a much wider audience that can help you +bulk close of all resolved issues prior to release. +Its certainly problems with the same error +To reproduce the bug run sample on the second xml I attached +Fixed formatting and removed the extra line for + +Changed Target Version to upon release of +It seems to me that you have to maintain a map btw SSTR and +bq +Thank you +Also comparators and you sure this doesnt work? Youd have to use the fully qualified name and youd have to use but it is supposed to work +I forgot to mention that this problem did not exist in but we started seeing it since minimal example +close subscription channel might be simple and straightforwardif we want to guarantee subscribe succeed if closeSubscription succeed we might need to adapt to same solution to kick off a closeSubscription request +go for find Andrew +Hope it makes it easier for others using the desc plugin down the track. + +Thanks +It did seem to pass the attached bug case and passed the existing online compress test +Thanks +The fix to removes this inconsistency +When I add a direct dependency between the and projects wont start because the EJB archive cannot be found in the local Maven repository +using nbsp in empty cells is not suggested +Trying to reuse them for testing other parts of the stack just complicates things with no specific benefitAlso instead of a having to explicitly specify a build profile it would be nice if the build automatically ran the integration tests also against whenever a local test database is available +Kelvin patch applied at r please verify thanks a lot +Yay Burr got this updated! Now all that left is getting the magnolia page updated +Posted reply in forum. +Assigning to Weinan to rebuild resteasy with a change to the pom to use the project for the +Since the patch doesnt involve generated protobuf files line length warning is legitThen there is the findbugs warning +I am not too familiar with Twisted but see how this looks +In case an application is expecting derby it will be aware that derby doesnt have any no standard method in unwrapped object so will not try to unwrapped object +x and trunk +Patch to allow global properties file +Actually will be released in. +The way it is the pool size grows to the configured and stays there +If you have some idea how to easily get line numbers for all instructions this would be really great. +I guess this issue can be closed then +It doesnt cause the whole app tofail if the connection doesnt ralf Im with Werner on this one and would like to not add anythingextra that would need to be configuredchecked as it would most likely just leadto more confusion + +As James Strachan says javac is dyingrant endI have used scala on trunk and I havent seen any errorswarns +The patch look ok +For information version +Nonetheless all preceding HADOOP tests are passing and this particular test in no way reflects on the functionality and compatibility of within HADOOP +Theres no xsddocumentation in the schema for this new attribute are you planning on documenting thisThanksI just updated the with documentation for +When you want an exclusion in an EAR you must repeat the exclusion for each relevant subdeployment within the EAR. +if I select a resource before launching the wizard it works if not it doesnt +So as long as you are adding new methods in interface I consider it to be backward compatible +looks good to for reviews ted and matteo +The finally block should be hit with or without an exception +The Gatekeeper only needs to look at this field if the status was manually set anyway +all tests passYesterdays break was due to not cleaning the build before compiling and running the testsCan you please have a look at this +Give this some time to stew in case anyone else wants to take a looksee +I am not opposed to including this in +I looked at this again and I cant figure out what is going on moving to future +The duplicate assembly tags arent necessary +Id very much like to move toward being able to have a single configuration point when desired +Version in the deps directory ok but the versions being deployed are notare you copying them in a fileset that gets filtered +That method was a quick hack for FOSSG +bq +Does that help +Can you please retry this operation and update the ticket +commited in rev +From this perspective however I like the original version better +any help is very appreciated +Part of a bulk update all resolutions changed to done please review history to original resolution type +In this case I didnt provide an example because I believed that the bugs existence could be determined by code reviewTo answer your question like you I cant imagine a case where someone would use the default name and just wanted make sure that the patch I submitted would not break someones code that was relying on this since it is possible +I think most file properties should be stored inside the index file and should have an option to tell whether we want to preserve the original file properties or assign new ones at creation time +fixed with rev I took out the into a dedicated issueMarked the failing tests as ignored with a reference to the mentioned issue. +Thanks for making the changes! feel free to commit if the test fix is hadoopqa again +Ive committed an error log message when a is declared as a Seam component until we have the SPI to actually do that +I guess i made a duplicate jira issue here +I see the same failure on when I run the test with locale set to nynorsk +The new version of the document has been uploaded to the website and should appear soon +For example we dont check if an extends type is indeed a class nor if a type referred to for an implements clause is an interface into Git as ID dfebbaabccdcbb +Any more news on thisAt the moment we cant store the name OToole via because it triggers this error +Thanks Nicholas The Changelog entries are not needed since we add the changelog as and when the committers commit +Ill start pushing in a second interval and Ill open a whole bunch of sessions +The patch lgtm +Thanks for the updateAlso for the comments about the standards associated with indicating a menu +Home objects are used in the conversation scope not the page scope +This assumes you have setup the eclipse settings for the module +more issues on s +I just found information about using backticks +I added a straight Servlet test identical to and it passes +Same could go with the toolBar +If you have any questions let me know +Groups will definitively be a good option to achieve that +I will have to look into this bit deeper there is generally a good reason why this is done and Jackson core functionality rarely ignores exceptions +This is fixed in SVN +As you can see regular queries work just fine but custom select statements dont behave in the same way as the ones run with mysql gem +Could we do the release with just the library +And what about changing columns and to type BIGINT UNSIGNED + and are not managed by JDF +If works that would be the easiest fix no idea if thats an option +As the issue has been marked as resolved already dont your other question did you compile the descriptors and add them to the classpath as well and did you include the +antlr all the classes are maven maven and are differentlogj see the attached file Why did you choose to report only those differencesI do not understand +I only backported the test case +This issue was solved inThanks +Please also try testsbinruntest testsbin and testsbin to make sure they work correctly from cygwin. +trivial patch was already fixed thanks for reporting Nate. +The patch for embedded driver is still waiting for commit +If an error occurs while setting that up it should fail the job with the job diagnostics indicating the exception message and stacktrace rather than simply exiting with no diagnostics +Morgan can clarify +This issue is closed now +It seams JIRA uncorrectly works with russian quotes ampersands in html entitiesampampamp while expected +What I would really like to do is change the underlying implementation of streams coming out of the store so that they can easily be cloned completely and not have their state shared +Subtasks have been done and this patch seems to be in +Also I changed the example solr config to make the defaultSearch instead of solr +The patch for this will be dependent on the patch to avoid a merge conflict later +Patch which fixes this bug depends on and tested on ubuntu Committed in rAlso commited the included tests +Once you have the changes working you can post the changes here to this ticket + +Fixed on the SVN trunk +Im watching r where theres no mention of a +So +Created to track the implementation of that idea. +That said it looks like for backwards compatibility and for ease of migration well need to include some level of support for existing Jackrabbit +This now depends on +Thanks for testing the patch outIve committed the fix for this issueTrunk Branch x close after is released +This fix works +DIH does not yet guarantee any state persistence between two imports + removes throws and fixes some code formatting +Implemented proposed modification to the unzip command line +and I understand that I should just build it instead of complaining about it +You may start from attached sample +I would refrain from enabling any setting since we dont know if users want to be doing reverse engineering or database generation +I downloaded the patch and compiled it with the source and it worked for me as well +So this appears to maybe be an issue with the latest then +Plus with the patch Im getting failures with custom I believe it is due to the changes not being through out the codescriptsconfig files +Should I keep them and mark them as deprecated for the backward compatibility purpose +This bug is a regression caused by the commit +as problem with the copying approach +Cant think of an easy way to unit test this sorry +QE will test applied r +See +README updated +Is there a workaround for this +Indeed just checked sorry for misdirecting Greg I didnt know js impl +I implemented additional button and export to this patch works for me against designer the patch from Maciek thanks! +bq +Assigning to Shelly to investigate or reassign +Moving to fix for just since we have gotten no feedback on this in terms of which dialects require which precess +windows build patch addressing comments from nslater and updated README based on starting this process again from scratch +At one stage our pom used to have a dependency on which I changed to as that was all that seemed to be required +My simple workaround mentioned in the bug description above also works +On trunk the will have to impleemnt some more methods as the JDBC interfaces evolved in Java Thanks a lot for the review +Thanks Tony! +Stu please reopen if you are still having issues +Closed. +Randall others what other conditions like emfile can happen where creating the file from scratch is not the right cause of action +and you can execute it on your eclipse to reproduce the reported bugRegardsThomas Modeneis +there was a broken version of trunk for a while last week +If we can fix this before the release we should +Hadoop moved the locations of its many jar files around +There are two fencing methods shipped out of the box RPC to the active NN to tell it to move to the standby state ssh to the active NN and kill the NN process +Fwiw I measure about slowdown or so +Burr there should be no problems to start JBDS with JRE could you attach file from installation that fails with +Is there any conclusion +did you provide details on how to run the algorithm the assumptions it makes file format behaviour if the output file exists already on the wiki? the class is named but if I read it correctly it looks like the entry point for both the parallel and sequential version +Im not sure whether this should be a HDFS or MR bug tho +war it assumes that directory is just for organizational purposes and will recurse into it looking for deployments +Adding to the top level pom +Upload a screen shot after the fixes +Xiaohong +also i want to use it with solr this instead i guess will not work completely +This does go away if I downgrade to checkstyle but its not the default plugin available from the eclipse update site +you very much for the contribution! +i TODOs this instead of cleaning up properly so that you get logging for the segment try at a patch with a simple test +After some investigation there does not seem to be a quick and efficient fix to this problem +When using the pJoin is relevance still applied to the results on the left side of the join +These changes only affect the list views on My Opened My Assigned My Subscribed and My Recent pages +This should be changed as well to a word +But the itself is so we need to introduce an additional field such as message length to suppress the interpretation until all data is received +updated code based on code reviewadded fix for into +JAXB seems smartcleverannoying enoughto map the String type to the standard XML namespace +Hi allIve deployed a and a containing the fixthanks a SanneThanks to both of your for fixing it and for publishing a snapshotAny plan to release a version anytime soon? We have to release a project soon and wed prefer to release it with a released version rather than a snapshotAnd this is definitely a pretty nasty bug which is quite hard to work around when you have encountered itThanks for your feedback +ear file find that this happens when you try to use a persistence archive with its own file +Im hoping security enableddisabled is in that list +if the user doesnt have these private props on opening the project hes prompted for it +I just attached a regedit export for the requested key +Hi Lionel now that has been implemented could you update your as described in that issue use a new snapshot and retry using the STOMP frame to the UDP port +Such as nothing creates parent dirs anymore +Items to be fixed in multiple releases should have one issue per release version or we cannot accurately filter for the release notesPlease edit the fix version field to only one version and clone this issue for any other versions that this issue should be fixed for +We hope to be not too difficult +I got some small conflict when trying to commit +also increase the checkpoint and cleanup interval +We depend on oVirt API to report this problem correctly instead of throwing an exception +So i fail to see what the standard way to get into the runtime classpath is? Any suggestions? the issue as duplicate of the newly created one in issuezilla +Chained APIS are dangerous +patch showing the first steps towards the subclassing approachI definitely think this is a much better way to go +I just wanted to explain why this isnt a make error. +This should apply to both and trunk +bq +fixed and the testcases run too +On my machine at this test fails too in +And now I can finally deploy the ask somebody else to verify this tomorrow +main +Thanks Owen! +Keeping explicit catch just for making the test fail will supress that exception traceAgreed thats an excellent point Uma +Pings have been sent out +If not then can you pl +Thats a whole different can o worms +still failed as timeout but cannot reproduce locally which is strange +Closing issues which have been released. +So Ive pushed one more commit on the branch to fix that +The latest patch for is already committed as +Next iteration of patch +Patch looks good to me +verified on last available build for linux platform +Similar to FILE api in libc the behavior of using the handle after close is called is undefined and might create problems +If something isnt working correctly it would be a JAXB bug not CXF +This error should be fixed now +Updated merge merge patch +regioninfo files? Im curious about what you did to need to use a full rebuild +Thanks for your review and comments +bq +That is essentially what the is though and there is no plan to revise the transports in at this point +Maybe i can make an additional patchRegards SaschaThanks +The problem here is that the site plugin needs one base url that serves as reference for relative links of all +fix included in +It improves on the previous patch so I will commit it but leave the issue openand see if I can make solution work cf Dans comment +The code that I am committing for this +Sounds like a good idea +Just saying itd be pretty damn cool +I think forcing users to put a from party is the easier solution +Thanks for the feedback AdrianI would normally have used screen widgets instead of freemarker but in this case much of the code was copied from Orders and Order replaced with Quote +bq +A configuration wont created until you call some remote debug session +Resolving as DONE +Removed that test since the mosaic offers no way to control +I only looked at the output robert posted so far but it seems that a gets lost somewhere hereI will look after can reproduce this easily and even if I set search threads to and index threads to +Do you really need features for the script in? That doesnt seem likely to me +Since this patch was part of it doesnt need to be applied. +The method was spotted in Hibernate as being missing and added there but Spring has not yet caught up +It seems to be changing many key points of the job and I wish to see that the functionality and optimizations are not being lostIm not fully understanding is handling of +accepts no liability for any damage caused by any virus transmitted by this email +menuadded component selection into new project wizard +addes del count commands to cassandra CLI +proposes an improvement in the following comments that was never implemented +Simply register a for your type +Let me know if you would like any part of it +a basic unit test is running. +Thus the Preferences +client retries a different primary datanode if the chosen primary datanode is deadIt seems to me that this fix does not degrade whats in the branch +its the +so I am not sure how to do this +so keeping the for snaps actually should fix that problem no? Anyways is this causing problems +The SCA spec follows the JAXWS to map +This is a new feature so will appear in only +If were speaking about table heavily written the regions should be flushed by the standard flushbq +Im not necessarily talking about compression here OK? Compression is an example of an alternate storage technique but it isnt the only way to solve this problem and as you point out it may not always be the best thing to do +With the reference in the last comment replacing the reference in the patch this looks OK to me +Adapters are a powerful concept but difficult to maintain +Thanks Clint! +Just verified that cookies are handled correctly if nothing passed in +Works well julien but could you Replace the column title Variation by New violation Not display a resource when the number of violations on this resource is in the log file replace HTML Report generated by HTML Issues Report generated +If that is done I think we can close this issue +Committing. +Raised to fix this in the compiler +I committed this +I left a couple short comments on reviewboard +Every customer of the portlet selector must be a user in addition to any other roles they are in +Created a Response object for Furnace Manager Maven is needed to display the exceptions to the output when needed +Yes it is an implementation detail of the for grouping +on. +Allows users to optionally include design documents in views +I checked in an addendum +Nicolas and I just chatted in hbase +batch transition to closed remaining resolved bugs +Still not a bug with XalanRegardsBrian specific control over the sequence would be an enhancement request +Attaching the patch addressing Hairongs comment +Ok Ill create a test that attempts to change the value from false to true +For inserts I was using a datastax java driver with cassandra single node on with SSDWrong result sets can be seen through java driver and cqlsh as well +Weve seen the workaround and a change like this would tend to break existing applications. +Lowering priority because it works now it just needs a cleanup +Does your custom shuffle handler have a port as the service data? Thats ingrained into the MR AM and tasks +Ill provide an overview of this issue for the developer meetup on Tuesday and Ill attach the bits here so that we can further discuss +Interesting doesnt seem very intuitive to me though especially because we dont ask for a path we ask for a string value of the context at the least it would still be an issue that its not documented and its not something Id want to require seeing in the resulting URL strings +Works the same except rules are pre installed in application environement when couchhttpd is startedPas tests with previous patch in couchhttpd changing the way config changes are handled +It currently has to ignore +Is it introducing a bug into Ivy? If theres a use case for a URL involving a backslash then I guess so +Fixed in revision +Id like to squeeze this one in for +This is pretty big change especially on a point release +I dont think JIRA being openclosed is the issue its more multiple commitsBut yeah as a separate note looks like there was no final comment and resolution after the commit +See. +Was the ruby runtime at beaver updated? It seems to me as if the problem does not occur anymore +Add the missing clause. +By way of example let us say that one of the keys denotes the namespace URI of XML nodes +The proposed arrangement looks good to meKelvin +Can someone provide a clear and correct explanation? If not I suggest removing them + the patch looks good to me +bulk close for the release +If you would like me to submit one large patch for both I can do that too +For my project which as about slc the memory thats used for storing that data is significant +Adding submit patch tag for more visibility +I have coded a patch and for and will migrate changes to and Westport +Im happy now that youve done the JAMA code +Therefore attaching a manually configured sample for initial testing purposes +bulk defer of open issues to +Deploy it and test it +Is this patch planned for any release +Another approach to this would be to have an admin command for the job tracker that would have the job tracker stop accepting new jobs and shut itself down when the last job completed +However if the request message is directly used as the response message there will be a recursive referenceSo if you are not using the request message directly as the response message this problem should not occurDanto be on the protected side we should change the code to not put the invocation context into the response contextIn addition should we even get rid of this log line or log only the keys? I dont know if it is okay to log the context properties from the security point of view even if the level is set to FINEST +Revision +Seems like we need some refactoring to the tests to take care of the change +write to hlog +This fixes the problem as Brian did basically and adds a test case +New patch for trunk! This also includes a change to the injector where injected fetchInterval is added to MD +What is to be done on this for +Patch applied +passing in the constructor allows for multiple slashes +same same underlying issue +Thanks Olaf +Committed revision Thanks Javier +better exception in +Leaving against CR for now +BTW theres a Jetty out +I just committed this +You could actually start with the XML editor in Tools and use the outline view so that a user can easily pick from a list of available providers listeners and actions to essentially drag drop file think the JPDL editor may also have this outline view for editing the file but Ive not tried it out +But ideally build and build should generate the same result +Agreed +Thanks Robert! +I reinstalled ALPHA tools with new eclipse and now I cant reproduce this issue +I fixed that in revision the colors of the image are wrong but I dont know why +Hey Suresh I think that was just because trunk was last merged into the branch by me on whereas was only merged to trunk on and was only committed to +Fixed under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Or if have an alternative solution that does not require that fix its betterBecause as I understand that patch disables for all IBM rightAnyway feel free to proceed with that patch because it at least enables running trunk tests with an IBM JRE +the current available constructor would then pass null and just forward the paramsi would prefer that over always returning null +Artifact dependencies must be resolvable through remote maven repository or local maven repository for maven eclipse plugin to generate +Having attached above patch I could definitely need some help with testing these new cache types +The indexing trick is something i remembered from back in the days when i did some applet development +I dont think this issue needs to be solved in the entity engine +I see anything from to failureson a typical run so it makes it hard to use to verify changes in the replication code +I verified that it works by removing the property from my file and checking that the log tab on the web UI of a job said Log streaming disabled +to the Nexus repo +Sharing the region state in ZK would be a simpler option as we would know if the region has moved or not +I added the property and verified in Web Console System InfoName Value HIGHIt didnt get passed to the job properties as either +OK Ive added a timeoutretry on the rename attempt +Since the latter is preferable I will work on having the scripts contributed to Apache +I just tried this on and I still get the Resource class interface has no valid constructor error +This sounds good +Colin youre unbeatable +OK finally saw that this is related to +Its useful to keep this issue around so others can either vote for it or add comments to request it but Im moving it out of to keep it separate from our release plan +You often take the example as its given and tweak itI also have very often found that its very useful having the war right there where I know it is rather than having to hunt in a deeply nested tmp dirGoal is nice though I work from Apaches Git mirror myself sometimes +bulk fixing the version info for and all affected issues have in approach for solving this problem +Would it be feasible to at least replace the versions in installed poms +This has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Parts of the proposed corrections were already made by some previous patches the remaining changes are done in this patchNo regression test is necessary for this issue +Can you check this with? I have changed it so that POM is not rewritten for deployment so whatever you have in your POM should be left alone +Deleting A from Dx is correct because that how NN handles excess replicas in general +My mistake +Popup is now only visible in the Script Explorer +inputrc behaves exactly the same as in bash. +should be Unknown +First I will start by just moving the core artifcats to commons +Fix made to the Execute tool and a uploaded to the distribution siute and the Maven repository +here is a patch and a dedicated test +Schema is created using build target +passed +Depends on a new event model in openshift +LOCALONE looks like a better candidate as a default CL than ONE +is Gb optimized Gb unoptimized +this processetctomcat stopsleep rm Rf optwebapps f optwebapps homemarc optwebappsetctomcat to hosting provider +oops sorry heres the attachment + +Thanks Mike! +Thanks for filing the bug against POI +done for done just have to find those FIXME and disabled to use Ignore on +So now if you put a component API in the WAR it wont be accessible from the EJB module +Pushing to the possible implementation path for this relies on a api I am thinking this should be pushed from +Id rather think of a simple lock to avoid such a deadlock without affecting much of the JBM functionalities that already work well +I dont see +Thanks Sanka +It should be jms component have too many options +Here I am unable to access the in the core project from eclipse but I am able to run it from the command case you are wondering why I do the above its because I use jibx on the classes and hence we consolidate all the classes into one folder and then run jibx binding task on it +IsuruThe log looks incomplete +Hi Daviddid you add styles to the page through a paragraph or through the editor by any chanceThose styles are conflicting with the styles I added to the project stylesheet and need to be removedIf not there are some mystery breadcrumb styles that I need to hunt downAlso I styled the breadcrumb to have a span inside the div +So or about Something is wrong if youre seeing +This log shows the connection errors then a server restart and everything is happy againIts not the called with null error that I was remembering but that might be because of one of the following +Have you noticed any long delays in the? It seems to finish doing the unpacking and then sit for minutes before handing onto the next plugin +It is due to my commit for had incorrect jira number +The change will ensure to use fresh UGI and its internals after cluster switching beside this any security concerns do you have? If any what kind of convenient support the UGI library can provide +Why dont we have a patch up while we wait for more comments on what folks think of this change +Hi Tim Verified +I will post an enhanced version of the unit test as requested by Tom +Close all issues now this version has been released +Done +Is it really necessary or it could be refactored away? Im confused but yes it seems its totally useless instead of figuring out what to do with initType as an int just make initType a that was inherited from an earlier version where I used a switch or some other construct to determine how to init the delegate Filter +Yes that was changed sometimes ago and was mentioned on the User List and even in Version Notes. +Thanks for the details PaulThis was still useful at least to find out that the full rev tree revisions field is being included in the credentials and its not necessary at all. +For subsequent recoveries choose a primary datanode which is not the same as the previous data node but is a recent heart beating data node +If the logj jar is outside the lib of the application then the logj initializers onStartup method will be invoked when every new application is started +got it roughly working but Im not sure about whether the pc also should get executed on the superclass +Thanks for the thoughtful comments everyone. +Added a bunch of type class to support configure loadBalance in Spring configuration file +Sounds very interesting can you comment a bit moreregards Berndfinally I got time to look at this solution What I likeit is really simpleWhat I dont like you have to explicitly define the signature in the what with more generic methods? Id rather configure this somewhere in the compositeregards this should be possible adding an attribute to declare this is an methodbinding or a different Tag like attribute. +I dont think that a change to the clientside representation of permissions should cause a change to the DFS representation of permissions +Will create new bug for this +Ive confirmed this issue with a unit test that exports to csv using the original export view a instance and list +The fix I attached changes constraint deletion to use a single exec that uses a cursor to walk through the constraints +Shall I close it as DuplicatesRegardsLukasz let me note if this workaround worked for youRegardsLukasz me this also looks like a duplicate of works fine this one can be closed I guess +the documentation is report is wrong and the issue should be closed as not a bug +Sounds good maybe you could create a new issue for sorting out updating unidentified widgets +The bug only affects downgrade to a specific set of versions it doesnt fail against the latest sources in the branches for the affected versions and workarounds existWorkaroundsIf you experience this after soft upgrade from one of the affected versions to or later and reverting back to the old version use one of the following workaroundsWorkaround Reboot the database using Derby or later +It is different from failing a task +This is probably related to my attempt at refactoring to be a query that performs an arbitrary function on its constituent subqueries +Ill be happy to contribute a patch for +Request for OS license sent +Would you mind opening jiras for the open comments? Also the test failure needs to be addressed +Fix is on master and targeted for is opened immediately but is empty for seconds and user doesnt know what is happeningAt least mouse cursor should be changed to waiting stateAndor in place where found will be displayed should be displayed text Searching for available while program is searching for. +This is what does +I have to add an additional comment the exception orgmetawidgetifaceImmutable is weird the missed class is in metawidget jars +Resolved by commit cbcdeefdacebfebc +Committed revision with the basics in place +bulk close of all resolved issues. +Im here to ask again if we can mark this as wont fix for the current trunk +should also take into account fieldresult +See the attached an example +Im assigned the task to myself until we have the basic svn structure for the site in place +Because for javascript numbers and dashs are different from ordinary chars +Chad can you please spin up a new patch and barring any objections Ill review and commit it +I will eventually handle server definitions launch configurations and other elements via an API to wrap handle all of our important issues +XSD and XML ARE NOT consistent +Most of these actually contain valid +Ashok could you confirm that the solution provided by Krister has fixed your issue? +Verified in JBDS +So I guess it still needs review + +One way to make the test handle the ambiguity around the DST switch is to force it to run in the GMT timezone which doesnt observe DST +Sigh I took way too long to take a look at this again +Patch applied +Previous value was by setting it to the as RI did does not render the border attribute anymore +Adds a bit of logging and enables for the test +Attaching a new patch which addresses all the comments +Deleted overly long list of offending statements +Thanks +This is a duplicate of and. +I meant the code would be simpler +Then we can remove it again +Im going to go ahead and resolve this issue as Wont Fix +So maybe lets just remove the support for locating java from PATH which is the current behavior on Windows platforms +In addition the cache is too tightly synchronized and does not allow access to the cache when a thread is loading data in case of a cache miss +Marking this critical for +Please let me know if any corrections are needed +I dont know exactly the definition of relation maybe my need is to have some like relationExplained from your perspective User has a collection of Roles +when there is an exception send error if the response is not committed otherwise leave output stream open and let the servlet container handle the exceptionWill try adding new tests +Data was written to Cassandra by a hadoop job with configured +Ill commit shortly +We should also remove samples sections as we dont have any samples we ever have it again we cam alway pull this section from the section was updated and now describes new structure as well as other details +besides the doc issue i think that something so trivial shouldnt prevent the whole domain from booting up properly +Isnt there a way to provide support for both version of mylyn in code? Such as the method that disappeared +open to making truefalse not require quoting but if its a pita im also fine with leaving this title +does it leave the origional files in place? or does it make a copy +works for me +I thought Ryan was all about the buffers like memBuffer +Fixed upstream +Need a solution that is also compatible with Mac Firefox +The point of not locking it is that you get updatesIf you ever cared about reproducible builds then I can only emphasize to follow Olivers suggestion and lock plugin versions +I think we can keep the same ticket Ill just change the name +Agree that this is wont fix +We also need to be careful to delete old task tags on a clean +If it is allowed by standard Derby should also support this +Didnt see this issue before I opened for Server +We just want to be short +Changed error code and error message +Likewise when looking for children of an EAR it wouldnt find Spring modules so it wouldnt show them in a list of children for that EAR +tested patch on a three node cluster that was using DSEs alternate secondary index implementation that allows for larger values for indexed column values +I think you forgot to commit the newly added classThe trunk core tests wont compile until you do this +Committed to and trunk +remove sslshutdown when closeSockets for mac os +In the abovementioned list the PDF should be +ie +start of attempt to make work with openejb offline +For now I added a big red bold warning and disabled the test +Will take a look at the failed tests in a couple of days +Thanks Todd! +Hi Brian dont suppose you are able to share the problematic workspace with me +I switched jetty to using the patched jsp servlet in rev accepted this change rev in tomcat we are using. +In such situation can separating writes from reads help +Minor tweak to log string caught by Suresh +Attaching +Result of +What about a new bundle called which holds these kinds of interfaces +Perhaps the navigator element could encompass the proposed functionality from and delegate http redirects to. +I have separated the patches into main areas the major execution engine changes the changes to the MR codebase to make it happen the changes to further abstract other statistics related stuff and the testing changes +Didnt see any thing related to my patch +Patch that lets these classes continue if they cant create a security manager +Commit the patch to trunk at revision thanks Gurkan +Committed to as well. +Because Buqi think that JIT should never remove the ckecknull before the vmrtmoniterenter it should be safe for the stub to remove the CHECKNULL instructions inside +this is contrib so lets add it in +the case where the slave is on gen it did just download the files for gen so it seems we are not picking up the latest commit point somehow +going to ask on the builds list for someone to look at this +Fixed writing of n now we need to process the Strings passed as parameters and change the line endings to the current system Napoleons snippet to +Just as an addition The same problem occurs when I use a composite component directly in a facelet +Fixed in revision . +This is job +Neil applied an equivalent patch but there still may be an issue depending on whether unescaped controls are allowed to appear in internal entities in the next draft of XML +Can a committer have a look at this and give a code review? DevarajFew miscelleneous points With this patch everytime a new task is launch its pid file is written +Ah okay that makes sense I didnt realize that the exception would be propagated to the +For me the workaround is to close all editors and restart eclipse. +Got it +Im making modifications of both pom and code and I wanted to do a test build +Any update +I have did changes to your original whiteboard to match with OM thing +will be generated if user configure it +If we can add dependency of only to test target it should be betterbut I am not so familiar with Mavens so I am not sure if it is possible +This would be a fantastic enhancement to the platform +Applied the patch to Subversion repository +Copying over an existing index without first removing all files in that index is not a supported use case for LuceneIe to restore from backup you should make an empty dir and copy back your index files +The current situation is true by defaultFor upgrade default owner fsOwner which is the user who run default group supergroup which is the value set in default permission So shell we mark this issue wont fix +create the in coming shipment and receive n units of X into facility the QOH n will be created +LGTM committed +Then change text files name to +I dont have a windows test machine at the please try to test it quickly we want to finish up this weekend +Attached zip contains a patch that Adds a topic explaining upgrades in the Developers GuideAdds subtopics to the new upgrade topic +Oddly enough I ran into this problem again today with +Or if you can tell me how to reproduce Ill try its a production environment that is able to reproduce this specific bug more or less reliably +Updated to use forkMode always +This is necessary to reconstruct mail threadsAlso please add or update the unit tests +One more comment it looks that you allow duration to be in the script file +It passes for me +deleteTable is to clean up after the test which is flaky +Best regards Nikolay Alexandrov +No problem for guys I was talking with Bob +If so then it throws a new and does not shutdown +I verified this works now +I moved MAVENOPTS from jenkins config box to a setting before the runs up on is another issue not because of this patch +So admin can see this logged event and act accordingly + +Im a happy little butterfly flapping my wings causing eddies in the air which ripple upward affecting incoming cosmic rays that deflect to your causing your code to magically be to my liking +The MERGING znode actually has the name of the merged region so my claim above that PENDINGMERGE having name of merging region being new is not totally correct +Committed revision Thanks Fergus +I still see this when setting up new Eclipse projects or regenerating them with mvn eclipseeclipse +Great +bq +Reasonable simple change +It is done as expected. +Im sort of thinking that if the server is validating data it should also store the data in the appropriate form converting from Thrifts string format to whatever native Java type is necessary +Cleaning filter closing Nick please reopen if needed. +I hope the change in package names and some class names is acceptable for a minor version +If you attach some diffs for the changes you think need to be made we can apply then to the doco page +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Had I any indication of how difficult this was going to be to work with you guys I would have submitted this refactoring a long time ago while it was fresh on his mind +This patch fixes some bugs in SWING module on which HTML is dependent +I tried with true and it works now! +correctionehcache had throw ehcache have without throw of great +I have fix the issues but am still using Findbugs dont think we should use a RC version since changes happen regularly with the Findbugs project +I will continue to look +Thanks Arpit +Committed patch to documentation trunk at revision +The others are fineJavadoc believe it or not is actually a Sun trademarked term that applies to the tool used to generate API documentation +Closing all tickets +Resolved +I dont know if it matters but I should have had said A extends Xto make parallel to my attached example where Organization extends appears to be similar to bug found the same problem when using Hibernate Annotations with GORM +I think we are done. +Thanks stack I will take a close look I dont know why I never received emails update for this jira +Intend to merge it into revision On branch +Fixed now for M +Fix in place +I also changed some of the docs accdg to whats in the the m in the memory settings is not required but I think it does make it clearer so I added it +meh +bq +The for included tests is ok since Im proposing this patch to fix tests +No to be clear this works for no situations at all which includes the example +But not for iBATIS +Added the setter +INFO map reduce INFO map reduce INFO Job complete job INFO Output dataoutput hadoop fs cat dataoutputpart +The issue is rescheduled to as it could not make it for the cutoff +Tomislav I have a similar problem see +This model is also present on annotations branch branchesvGACP and must be fixed +According to the last comment it seems this is not a bug in Axiom but with the underlying implementation +Note I created to track the impl ofarchetypeSync +Ive added Apache license headers to the code and checked it in +each entry should have a valid locale associated with it and the locale should be the weblogs default locale in any case where the user doesnt explicitly choose a localei fixed the xmlrpc places which were not setting locales but we probably need to add something to the upgrade process which fixes old entries with no locale +Ive increased logging level in case this happens again +It only make the oracle JRE hide the problem a bit better +We should watch what happens to that idea. +defer all issues to +Can you check against trunk againIt looks like its related to but that issue was resolved with so I have no idea what is wrong +Most people I have encountered just use the samples is pretty easy to assert the existence of a transaction in the Dao methods +Tried putting all the datasources into one file but no difference +I noticed this issue +Unfortunately the structure is out of my control and I am forced to work within it +Do you have a camel route or unit test etc +AmilaDo you have any idea about thisThanks just tried wstx and get the same result +Yes I know +Can this bug be closed then? +Hi Brianthank you for your feedback +Not sure the REST config exposes any way to also remove the data +For the Tree tag you must set the attribute rendertrue in order for it to generate the javascript support +Sorry about that +Yeah in my JR branch this is implemented with +I had the at the wrong position +With JCR it should no longer be necessary for the code to depend on Jackrabbit internals +The total set of annotations supported areTableColumnjoinColumn contains the change suggested by for Dianne under revision +lsof on the java process shows several hundred unix sockets open +My view of the class is mostly a brainless bean that stores the information that subclasses put into it +Added the reference docs and javadoc patches +Wo this fix and when there is a problem with the rack resolver script the will simply hang +Committed +It may be possible to do slashy string highlighting based on a full parse rather than a scan of the file + my updates to the Apache Extras FAQguidelines and docs that implements my proposal option +I wonder if we can come up with an approach that just introduces new plugins and doesnt add any hadoop code? The right thing is probably to use the compression codecs to encrypt on the way to diskthoughts +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +xulrunner error log +Thanks +I just tried it againAnyways Im attaching a zip containing the patched files +Better is to make OI transient any info they carry is likely available in different datastructures +further note reproduction possible with +Yeah I guess the code duplication and maintaining two sets of converts was mostly my issue +Verified +there must be some fairly major underlying factor that made the JVM developers decided to raise the minimum allowable from whatever it was to Doubt it +Removing or breaking that method would be a compatible change to the package export +and this is not a common case our users rarely use log archival! I am ok if you think it is better after the example as a seperate paragraph but as a bullet point under To encrypting a database with a new external key does not seem right to me +Examining this + +Anyhow if someone ever stumble over this issue I though it might be helpful to provide my code as starting point +As I talked with scroll docbook export provider they do not plan to implement custom macros export support finally +I assume the patch in will fix this +Made a small example and it works +ah +Updated patch moving the to rc and uses automake string substitutions. +Hi Michael what do you mean by backup? If you mean backuprestore of the global Sonar configuration I agree with you but if you mean backuprestore of each quality profile independently Im not expect to get the result that you mention +Frankly I dont think star imports are that evil +Shubham could you please reply to Alans question thanks. +Test case updated to handle regression +OK I think I have it +It shouldnt be a problem of +We now create the Interceptors and Decorators eagerly +Could someone test and confirm +Looking at code in hdfs this involves a trip to the namenode by my reading +This still doesnt fix the problem because the test for expiration is done using yet a different clock the consumers +RichDo you want to resolve this issue? this issueWe currently dont support attachments in REST mode but we have similar problems with attachment support and SOAP Dispatch +The first commit of this code did pretty much nothing and we were hoping to gradually build on that in the same way that the social api server is gradually coming along +Preliminary dev builds attached +see my comments along with the patchUntil thats fixed things are going to be unreliableIll commit that fix in a few first Ill Ignore the tests that now fail all of the time +unless I run locally then its fine +I test the commited version all seem ok +This works +Cezariusz is that what youre doing +I am attaching the package for a Sftp protocol handler +Here is an updated version +reopen to set fix version +patch applied +IT added +Installed and tested witk JBDSNo issues found worked without to set fix version. +The project ends up having code that knows how it works and is rarely touched by the core team +So when we can do this with Karaf then we can grab the commit and re apply it +This patch supercedes any previous ones +Gabe Thanks very much for putting this patch together it looks great only one suggestionIt looks like the file is one of your own configuration files that got included in this patch +no news from my sideShould we remove jsx and json for erlangIs this used somewhereIt came in with update your git version and the erlang client build should not have this theres a fix in trunk? Great i will give it a try +Please file it as enhancement at a temporary workaround I suggest renaming the project names of affected projects +The following changes were applied Using interface instead of providing own methods on and +Closing all resolved issues due to a successful release +My main concern is that right now its pretty hard to use it for scripts which may be launched repetitively +The is pool of threads that are doing the actual copy of data +The will be generated from that timestamp and the event will be stored in a file under that directoryIf there is already an open file in that directory the event will be appended to that file +I dont know why it has been created but in any case it will be a relief to get rid of it as it generates problems and does not help at all +Any pointers? +Aleksei is responsible for maven I assigned him + for the patch +not seeing this one easily +Verified that this fix works +But nothing prevents you from cloning and then archiving doesnt it +sorry about that +But indeed if you analyse a Maven project whose packaging is pom and which doesnt have any modules for me this is the expected behavior +We have seen this problem on a cluster that is purely used for archival purposes +Issue reopened due to users requests +The Mojarra guys have just relased a patch and the problem is fixed +Please fix it +Lets do this when I have finished porting color as of +Thanks ThejasI will try your work around lately +reopen to change CLOSED LATER status + +I think it is because of the changes done for +dont forget the one too +Dude this example is creepy! The office was wondering who was shushing! +This should be reopened and fixed +NET clients +See +Update patch to build with +Usecases involving slicing really need rangeslice metadata to apply this type of optimizationYes +Here it is +Resolved +that less of a Pair and more of a weighted Split +Changes committed resolving this issue. +I have reviwed the patch attached along with JIRA issue +Fixed in SVN AFAICT +Youri this ticket tracks the merge of the branch +Due to lack of developers having access to Oracle and Oracle going the way of the dodo at the same closing all resolved issue that have not gotten any more feedback or comment in the last error happens when you are using a spatial operator that cannot be evaluated without the useof the spatial index +Attaching the patch with the proposed changes +Werner thanks for looking +Regards the priority to optional +I think wed like to move toward removing the runtime from the tooling +Im using RC no difference +and +trunk Also fixed potential issue with restartingresuming a route which adds a job premature and causing the job to fire before the route has been fully started +Indeed +That is not a condition we need to generate a log message on +the problem has been solved +Once this gets committed I would be happy to open the issue to patch binhadoop and Ill post the patchAgain why cant we patch binhadoop as a part of this patch? That would be preferable as it would avoid duplication of logic + +Thanks for that feedback MyrnaKnut I think I need to chunk in the pieces of before we can start dividing up the product packages to remove the lint warnings +It seems the problem is still there in the client code +When I upgraded one of the sets I eliminated replication and began building them independently +True by default set to false to disable SASL on client +We will test it out and let you know how it goes +Do you have a WSDL which i can use to generate the webservice and reproduce the problemyou dont need a WSDL this will happen with any soap envelope +You probably dont need this patch but ill put it up anyways +Much appreciatedBest RegardsScott ES +Committed to trunk and. +on patch version Please use review board to get more feedback +Isnt it a user applications responsibility to figure it out whether its a fresh start of app or a recovery +Anyway Alejandro can confirm this +Im starting to think this is not directly related to Groovy but rather to a magic library which performs classpath scanning +Im not sure I follow this +Patch looks fine +fixed in again now properly recorded against Fix Version as well +Anton the workaround is to double click the server and use the Timeout poller instead of JMX +Attaching stack trace of a healthy node taken cca minute after other node unplugging +See +So there will need to be a way to tell the to reset these events +Deferring post console work +But wouldnt that just present as a corrupt replica when the block scanner catches up to it or a client tries to read it +Patch attached +May be we can find a better solution for camel with your help +Thanks a lot for fixing this so quickly Colin +Heres the simplest thing I could think of to try as an implementation for dump and load +I modified the file and it works +However the common case does not have a ton of anyhow as a distributed file system tends to have a smaller amount of very large files +RobbieWhen are you planning to apply the trunk patchRajith +Sure go ahead and commit Im about to leave for the day and I wont be able to until tomorrow +The changes made for should supply the functionality you want +If the database handle close operation fails during connection expiration and it fails by throwing anything other than a we could leak handlesr fixes this potential issue +This reflects changes made after the first round of reviews +In the meantime checked in fix for in rev +PrzemoThanks for the patch +If you finish it before we release beta please correct the version +This bug has been marked as a duplicate of +Do you want to manually run tests and as Hudson is not currently working +when screen is grey I can resize the editor part and immediately all the content is shown up +For thrift should have the patch to publish artifacts +All other files have been deleted +Hack which should be cleaned up as well +It is running fine for me locally +Committed to trunk +It solves the problem and does not require a major of the code +I just committed this +Thanks again +Of +Clarification patch for required code changes +Small patch +Just forgot to add the comment +Also all subtasks were verified +Are other people seeing this problem when running the test suite +Assigning back to Marek as requested. +So if you just deleted the plugin from your local repo it should get you the latest versionAs for the war plugin Jason wantesd us to play with it for a while before he deployed it +if you disagree. +I run several test which used a miniYarn cluster such as test test +Ill modify the the test and run it againI also tried run the test with HSQL and default modeshape configuration and the time of each loop wasnt rising +Thanks Paul! +Ive checked your patch applied to r and it solves this problem as well as +which is default inAt least its possible to disable this new feature +as long as continuum handles codes correctly thatll do +So the test will add about minutes to the suite +I merged it into a recent svn and I went through the ftl file deleting unnecessary markup and deprecated CSS classesThen the problems began +Exemple +I remember reading that anyone who has locktoken can unlock the node it does not have to be always lockowner +For trunk it was handled via +Committed to trunk +However I feel like we should fix the core issues that prevents this use case from stock rather than writing a fairly complex flush compaction policy just for +Thanks for resolving it +They should not and are not allowed in a compilerfor earlier versions of the langauge +It looks like the file at the top level needs to be updated to take into account each sub project +So I tired the solution you had suggested +Thanks +this is my question of the day how are we handling this which is really a backwards break in a way but honestly a bugfix because we should have supported Unicode in Lucene since thats the unicode version of java +fixed and confirmed +so lets fix this issue in +Patch committed +Fixed. +Committed to trunk +The panel is only visible when there is feedback registered for itstarget component +Why did you change the else ifs to ifs inAlso I have a better fix for +What I applied to +I simply have no idea what to do about itOleg +It contains mostly new files and only one small change to an existing file so it should apply cleanly against trunk or branches +So far I found followingdeprecated configuration file +Okay Ernst sorry for asking +Added meaningful exception message when Size and Column annotations do not have numeric attributes in Git ID cbcbfadbccdcea +in +Thanks Pierre! +Patch applied on and trunk thanks for the transition all resolved issue that did not see any further comment in the last month to closed status +A test case with a KB blob goes through quite a lot of code code that would otherwise not be testedSure but thats a completely different test case than the blob that is too large to fit in memory oneIm not against having also the kB test case but there should be some reason for it to exist +I just renamed the attribute transitive in onlydirectdep. +Actually for me the very same behavior can be seen with regular latest build from trunk +I think my solution works for good now I could not trigger any error anymore lets leave it the way it is. +Bulk assign all unscheduled issues with priority major or less to target +Patch applied successfully to my dev environment too +Thanks for the review Bobby +This is possible but not trivial +We now do not rely on other fieldTypes in schema and we can control precisionStep for the +ChristianAny updates on the patch attached +Ok this is the patch that is close to the final form +looking there now. +still trying to understand how the modules work and my first question is why there is not a module? +I dont think so I just moved the Fix For version to on the expectation that it would get scheduled for a fix +We need to scale the time and workload for them +We need some way of setting up the environment here in a sensible way same as how users automatically get HDFS on their classpath via the NM for exampleI havent looked into the confs etc yet so apologies if Im missing the obvious right answer +This patch replaces the internal on the with two similar ones on the +Provided test web application +Were sticking with +I dont think we really need this hereI will change the Whiteboard bundle to not include foreign contents and just provide its own stuff +Sending srcmainjavaorgapacheservicemixgshellfeaturesinternalTransmitting file dataCommitted revision +The distrubution default setting for is list +Bulk close for +bulk defer to defer to. +I tried the same with a local AS +the ordering of the region loads is importantUnderstoodThis involves passing starting to Cluster constructor and establishing indices corresponding to regionLoads +just reopening to set derbybackportreject flag +please take it up with the minigems maintainers. +My uninformed notion is that the way this HBASEINSTANCES stuff works is broke +How do run maintenance on it if necessary +Bulk close of resolved wont fix cant reproduce etc +Belated clicked Submit Patch +Currently this doesnt cause any problems +The implementation sources as eclipse have commited the three modules under revision rThanks a lot for this great contribution to Tuscany +nice comment +AS JBPAPP Branchx and trunk now using version of +Please try this patchThanks a lot +Thanks Navis! +I think I fixed this one +I did it on my eclipse which is configured with the trunk +No other failures other than the intermittent failure in lang +Misread your comment its what comes with the default project +The number of failures are down to with patch addresses Olavs comments about the vacuous test +Committed to trunk and. +Here is the updated patch +Anyone else seeing the same? if not what platform I wonder +This actually causes a major issue in a continuous build environment if you try to run the site goal as part of the build +Now this jira depend only on the move out of the from the for reworking the patches Matteo +Let me know what else +I plan to fix this +It looks like this still hasnt made it in +Im closing this issue as the poms from ibiblio are already updated +If the task fails at task tracker then failed time is set there and sent over to jobtracker +the route cause is there is no route for the mgmt cidr in the VRI will fix the issue +this was only in trunk +Thanks +Updated description section with release planning do not yet have any user documentation to provide along with this code +It sounds like you might have missed putting into your environment +It will be worth doing a full comparison before switching and see what the proscons of moving to Curator are +Seems save for me to do so +trunk rbranchx great now! Thanks for this nice improvement of overI did an ant with a fresh svn co of branchx. +We decided to go ahead with the earlier patch because it works correctly and we have gone through full QA cycle on it +Thanks very much +This is for new tests +It should be enough +This is very probably caused by arquillian container which bundlesI would open new issue in ARQ to cover this issue +from me +How much of a benefit do we get from the row cache being? It would be really nice not having to lock it +if the patch didnt work then reopen as Fabrizio has confirmed its not a dupliate and a real issue +It solves the problem +Did any junit xml files get generated in the target folder? They can contain more info about why the test failedTry adding emtpy brackets to your verifyImages calls Im not sure if groovy will resolve that as a property reference or a method call if you leave he brackets off +I applied the patch and things look better. +Implementing SNIFF is a big job and probably best done lower down the stack as part of or Ive moved this to resolve for +These tests passed for me as well against +Hi BrettActually I do have the Limit rows to option disabled so the DB is returning me all the rows +Venkata So looks like the GUI is sending the securitygroupsenabled flag as true If you look at your API call again youll find that GUI does NOT send securitygroupsenabled to the API callHere is your API call GET commandcreateZonenetworktypeBasicnamebzonednsinternaldnsdomainidresponsejsonsessionkeyigD There is NO securitygroupsenabled parameter in this API call at allThe basic zone you created is because the network offering you selectedYou selected so the zones securitygroupsenabled property is trueIf you select the zones securitygroupsenabled property will be false +Del Llibre destil de Vilawebencetar emprat abusivament per comenar en realitat significa fer el primer tall duna cosa intacta comenar a gastar +Yep that was what I thought als wellSo do you think we can assemble a patch and transfer that to the hama trunk? Would like to start with streaming in the next few days and need the protocol +Updated the description and subject to reflect that the only missing functionality is the ability to delete existing queues and topics +I checked with ajcommandLink and not ajhtml +Was there consideration given to bakcporting the fix to perhaps even with a different default forThe bug comments say Fixing this issue is but it introduces a behaviour change +Does it appear that pojo support will find its way into this component is that undetermined at this point +Perhaps this should be a REMJMX issue is there a way to move it to that JIRA project +Patch v does not include changes in v? is it intended +Changed to in revision . +In process of wrapping up +The expectation is to make the heartbeat interval dynamic we should remove the old config and not even add in the new config? May be hard code it to sec for now and immediately follow up in +On the and over thing even though it is included in the document legal was pretty clear that they want an explicit acknowledgment on dont understand how more explicit it can get? I doubt there anything more explicit than a legal contract with your signature on it saying your rightHow is an little web form click through going to get better than that? They do know I have no way of proving who actually clicked on the click through rightRegards some minor or their parents argued in court that they hadnt fully read the contract and and hadnt got to the bit about so by forcing a click we are covered against that +I dont think the error message is particularly helpful even though it is unlikely to be seen +comes back and is avail for a discussion +The patch containing the changes aboue and belowURL validation for Grant entry codeBase field keystoreURL and keystorefixed a caused by New menu on the graphical editor panel +That means we can delete our own comparator +Patch is identical and build pass so I guess it works +When a message appears the original queue is full and has just overflowed +If someone reproduces the issue with either no annotations Hibernate Annotations Please raise you handOtherwise we could consider this bug close by lack of fighters + +Ok +sorry forgot to add in environment in and commit f on repository. +If we could we would convert the site in question to an HTML based application +bulk defer of issues to +Thanks Regis. +The file attached follows zookeepers double barrier recipe for the issue that some processes go too fast deleting its znode before others are allowed to enter the barrier +Tim are you still looking into this? Or if you dont have the itchtime does anyone else want to add javadocs unit test for to move this forwards +Why do we need Ant tasksThis just seems like more code to maintain +I was only able to reproduce it locally with +Patch committed +Thanks Mahadev! +NTLM Proxies supported thanks to wagon +As requested by Jakob this issue can be closed. + +Made this blocker +If you finish it before we release beta please correct the version +In other words set more properties and upload a document to various nodes +Thats exactly the point +Not happening for every plugin for its current formI completely agree it should be done but right now not enough timeresources to make it happen. +Closing as WONTFIX +Ive run into this issue before so we have to upgrade to in this caseRegardsm gonna test now and release Pax Web tonight to be included in Karaf +I am happy to do this but in order to avoid a massive merge conflict go ahead and submit the above patch +I looked at and found that its scope is much larger thanI have the patch with rights to ASF for inclusionDo you mean that It would be better to check the mimetype of response also to figure out if there is any error in the response? +My concern is whether it is ok to make such change for +Logical plan related type conversion checking +Better memory management example +Maybe we should also rename to or something? Or split it in two since if I recall correctly some of the fault points are whereas others are just for the +So to be clear Forge doesnt have anything to do with it +This was caused by an exception from the workbench whilst trying to create an existing someone of you guys provide some steps to reproduce? I was trying to import several projects from samples +You do not want the orderid to be unique? But to make the mapping possible the orderid has to be unique +modified file +in svn r +Uploaded testcase bundle for use in step the first bad build I have contains revisions to +Can you summarize the impact of not being able to restore data contents and a recommended resolution for itThanks I merged the code into Acegis sandbox package for Ben to review +Please reopen if there are still problems +Paulex thanks for applying my patchIssue is fixed as I expected +Other things to consider for our updatesiteShouldnt the rd party dependencies be available on this site +Commit fcadfbefdafeea +Do you think it would be possible to go ahead with the simple bug fix now and look at the other idea as an improvement? If you could apply these small safe patches into the master then I can switch back to the master branch again and then maybe we can look at the riskier move of changing the id composition laterAt the moment we want our app to go live soon with a snapshot but without the patches in the master wed have to make a release based on our own branch with the fix which isnt ideal +Unfortunately Hudson seems stalled but since this is just a test change I will manually run the modified tests before commit +Thus any application that are presenting a simplifiedapplication specific view of Derbys management should be using Derbys to manage Derby +I was hoping you didThanksThanks for your patience +committed +Thanks JessiePatch applied to module at repo revision rPlease check it was applied as you expected +one might want to filter based on subject lineThat is the destination of the may be something other than a single addressThis could affect the design of the web application as well as the eventual LDAP entryIf only a single destination address is to be permitted this would simplify the design and these considerations wont applyIm not saying that the solution has to support anything other than a single address but I think it should be made clear upfront if any restrictions are going to be imposed +I dont think even Shalin has been able to to recreate it too often under the same environment +Patch to fix as Sylvains suggestionI think this is enough to fix when using MP +Hitesh! Did you work on this? Should I take it over +Which release do we fit this into? Seems like we can move this to after is ok + changes the test to not fail +changes made. +I just committed this to the branches and trunk +Your wish is my command MVSFTP is the entry pointCU is it brokenWhich version of are you usingHenrik +Ben per our conversation last night Im resolving this as fixed +postpone fix to the looks of the comments there has been no update on this issue in over a year +Patch that might fix the issue +Ugh +Patch to the Java broker to adapt to this change +I wont have a pg installed on my machine until the next Ubuntu LTS comes outOdd though pg seems to have been around for over a year how is it that we get a report just today? Ah because geometries are transferred are base encodedMaybe this is the cross version solution have also blobs be base comment on this one +This kind of logic does not really belong to as it has nothing to do with the request executionWhat kind of special use case do you have and what kind of API changes do you envisage? Im probably looking at the API in more of an way than it appropriate +the patch looks good to me +ERROR Not Found +defaultlib have zero Seam projects cant run in AS +Fixed with r +Created to test newRob can you review this change please +We should consider updating the start and end keys for the two regions on either side so this wont be a problemI agree that the custom mapfile implementation could encompass all of these little external files +I will make that change and commit +It is a step I didnt like to do but had to to be able to use the VM bytecodes on primitives and to avoid boxing costs +I started to fix this but the changes to started to get hairy +Uploaded patch for this improvement +It doesnt give me confidence by the late addition of what should have been there from day one +Good catch Scott +thoughts? backward compatibilityWe should do this but ill file a separate ticket to keep focusbq +Reopening and committing +Added a draft patch for SamThanks for the patch +update project to use common publisherNote project should be updated after suite integration +Committed in revision to the branch in revision . +Arun I know you dont like this because it increases the complexity +This is a duplicate of the bug I fixed yesterday +allows to filter applications based on type and state +is used in the unit test to compare the created DOM model with what it should be +No that is an artifact of a failed earlier attempt at fixing the declaration +a lot Ted for the patch +Projects depend on Velocity cant drop the old because of thisThanks in advance. + +Ryan could you assign this as appropriate +Pretty pictures for those who into these thingsPlease note the huge difference in GC Time of interest reached GIG of heap easily while is within mb or so +Its OK +The indication about the number of item displayed is always even if there are less item shownI tried to correct myself but Im sure Marco will be much more Firefox columns are not always too naroow +Resolved +patch was committed a while back +Please note that with the current version of JIRA is not guaranteed to be If this is an issue for you the only way to manage that is to move each issue one at a time +Our test assumes concurrent threads making search calls +Thanks David for revisiting this issue +Old Gen graph from Jconsole attached +Les thanks a lot for your feedbackMuch appreciated you take time to help us +So what do we do with this one? Its not a trivial change +Yes please +Thanks Laxman! Committed to +Committed to trunk +Sorry I must have been looking at an installation that Id already added the jars to +launch file and you can see it +There is missing in WAR added into WAR +Run the integration tests with Cobertura having the results being written to the file in mycomponenttargetcobertura +Thanks for the report. +This is a patch showing an initial step to typedef to stdstringI will apply this if we get consensus to do addition to the patch for Linux you need to delete the line from the cppsdoruntimecoresrccommonjsdo +Sorry for the false alarm +It becomes very long as ship groups are added +Another solution from belabelaBefore a TX acquires locks it should check if its owner is still in the current viewIf not the TX would be releasedThis way even of T runs after T T would commit suicidemmarkusbela yes +I greped the branch and couldnt find any more like this +Keith Yes + on this idea +Logged In YES useridPlease add your files Marek. +Given this change is useful lets backport using this jira +Workaround is to assign a to rel +cordovalibs then even when its overridden by a local path it uses the +Allowing multiple period before to trunk +know knownThe decorator text is a bit big suggest to make it say Key not found under SSH Preferences might be needed for login after launch +Seems we got the mechanics to do it but instead of unsigned content warning we now get signed content with unknown key dialog insteadStill worth it +simple add the back +Likely there will be shared ideas and implementation between these +Still waiting to see how we will inject the Validator Factory created by the deployer +for the trunk patch +xferring to Jerry for Agent implementation +BTW this seems like a good? for +Stuart Ive applied your patch with some alterations Fixed the weird code formatting Altered implementationtype Supported type on List type properties as well Used and by defaultThanks for the patch! Keepi them coming! +Trashing will be more efficentI think it is premature to optimize this especially if that involves complicating the namenode kernel +There are other bottlenecks we should look into however so well leave this open to remind us of an ugly parser edge case with performance +I also spotted a similar issue with streamIn in the same method its only ever set when were dealing with a blobAttaching a minor patch that addresses these two +May be fixed +These files should not be removed by the since that may exit while the job is still running +Wont mark Patch Available as isnt for trunkPlease review +Please review +Fix available on revision + +screenshot +I must have done somthing wrong with my request +ear +Thus it looks wrong to move this inside the createserveradapter logicIs there not a methodplace where all setup of the project is done no matter how it was created +just wanted to make sure anyone had a chance to complain. +I tested this new patch and I am able to buy and sell and digit numbers successfullyIm resolving this issue +In case youre waiting on comment from me as far as I am concerned you can submitI am snowed under and wont get to this for at least a week +Hence this bug +About Picture has no Version number +Skipped the checking of file format except in loading data +I am just trying to refactor class and use that in scanning +Good idea one less thing to copypaste +I am unable to reproduce the test failure even after merging up to latest on trunk +Patch modified according to Jons comments to use a simple +methods already exist I added a getIcon its not a bad idea to have implement and +Can you provide a new patch +You can closelajos +Theres nothing tying us to the current method of configuring jrrbsWhat are jrrbsbq + +The site docs have been updated as well +Theres no valid reason why the social data code should be forced to do this however +Its not optimal but hey unit testing in python isnt either. +Thanks for updating the ticket +I believe this to be complete at this point. +I say we bounce this to LATER for now and look at it again for the next release +Verify it in my latest local build yes its fixed +The spreadsheet lists what the complete download menu should look like along with which entitlements enabled access to which products +Upgraded to Maven but build fails at the same point test +I see then the change is required +A TFTP! +Gao v looks good to me too do all tests pass and how do we know it works? How did you test it +did you run javadoc look at the output? s value field isnt used and could be removed +reduce now uses a so that deletes are queued and sent to server in batches +CharlesI just want to say we should not change the Response in the producer and consumer +Use you please be a little more specific +nit flag was renamed but not the methodsOtherwise looks addressed the last comment +My mistake. +r +correct applies to version. +Committed revision NOTE IF YOU WANT TO BUILD WITH JDK YOU NOW NEED TO SET IN YOUR +OK I can reproduce this on trunk FYI +BTW is this any info in such a at all that we can use to get to know possible facets for the component? If so then I change this issue to a feature request to support CDK +Forget about this for the moment +Regression tests are runningTouches the following filesM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileThe full type information was being lost on the return values of functions +Revised for Spring will be available in tonights snapshot already +good to trunkThanks for the patch KaiThanks for the review Andy +Fixed +At first this appeared to not be working but then I realized we were having issues between maven and bnd both processing the resources +Really cool Rob! It wasnt obvious in the issue description but after looking at your patch I see that you bring in here some other implementations of cosine and sine that are not quite as accurate but run much faster +HiThis happens in work around of this? Any progress on the fixThanks! was also struck by this behaviour in +At some point I expect that we may throw back to the thrift client +It shows the operations involved in a delete table operation +or the starts should be TO starts +which is always going to be a list of size one +Many Thanks +Your team is great +Done Anders Thank you for your close watchingSamples is called that way because the core submodule is also called samplesIt can be called anything else feel free to update as you wish +For now for we can +We have just changed to and defined replication request handler +I noticed you are using would you please try with the current release? And if you can get the code from Git test against release fixed the issue with the wrong reference +The OOM happens because the PDF contains a large number of fonts and the font cache does not have a sane upper limit +I will go ahead and implement it without icons it will be useful but not as useful as it could be +Given I agree with you about what you say on POJO endpoints we could nevertheless do what the issue title says +Too awkward a bunch has changed between branches +Thats the job of IDE makers to show clean stacktraces or not we cant do much about that on our least we provide this utility class for the task +a Yes they are +from warn to infoinfo to warn you mean +As a result of the discussion over the past couple days I have come to the conclusion that constraints generated columns and triggers do not impose any privilege burdens on INSERTUPDATEDELETEMERGE statements +Project select PDE Tools Update Classpath +On second note the performance test was too big to attach +Its already there in +shall we consider a better practice using an exchange property instead of an header? +Ill attach a new patch to this problem +The unknown artifacts might be loaded first from the zip package +Reopening as it has been applied to the wrong need to apply a patch on top of this one +I am not logging a different bug for this issue since it is exactly the same stack trace as i saw earlierThanksI am not abe to repro the issue +Jimmy do you think this will affect rolling restart? StackWhat do you think before we integrate to +The code now never uses the default platform encoding on Mac OS X opting instead for as the default +Since infos is being modified in the scheduler code protected by the instance this leads to inconsistent access +Need to fix spec too +I believe the QJM test failure is unrelated +I just committed this +mvn deploy +Committed revision +I believe this can be closed +This file contains an example of the problem with suggested changes +This fixes the problem +So the question here is how much does this save +Cannot reproduce +The file cosnists of new classes alogn with changed classes +Alright +Its helpful to set the Component field to Doc for issues like this one +did not work as intended +This logic was likely changed at a different point than +My backend server will log and the additional leading space bother me when parsing the log. +Did the script keep going? Select a new location and move the region there? Was it moving the region to where the region was already sitting and that was why the exception +Patch to apply to messaging that I havent had time to create unit tests yet at least a couple of tests should be added for the properties and though I had was to create a default message expiration queue for people to an example +Add the queue dump into debug dump +Ive added the new exception +Attaching source to demonstrate the issue XHTML that contains the snippet above +Annoyance Axis does not use capabilities of can be configured using a method that causes to monitor its configuration file and automatically apply changes made to that file to its configuration +So if you have questions please go and start a thread on the dev mailing list and I am happy to help you +I dont see how this is uniqueThat said I also think the proposed use case is weak +If the master does not see a tablet server for a period of time it could ignore those entries in the metadata table and assign +Reproduced in +Committed as +Now that work is well underway with AS all previous community releases are +Adding a view of the hierarchy +deferring this for dont think this is up to date anymore +The optional flag is not passed to the war plugin period +To completely define FK relationships in XSD similar to database SDO Model xsds will need to support KEYREF +Sorry +NULL is not a identifier but a macro macro can not be scoped by namespaces in CCAnd yes since SUCCESS is not a predefined macro SUCCESS in two different enums is can be scoped by namespaces or classescurrently proto files are using javapackagexxxx to specify namespace only for java but the namespaces for all other languages are not specified +if we find any issues with it in production well submit bug reports +I can only attribute this issue to peculiarities of Mac OS Apple JVM +However if it is complex data type we need to cast inner bytearray to real type we still rely on caster +You can now skip dependency resolution withgrails true +So it would be very tempting to generate a String internally on the delimiter +Thanks Sumit and othersPatch is in trunk rev +Up to you in any case +fixed in +Thanks for the contribution! And yes documentation is the second part of any ticket too bad we dont have a wiki checkbox for that +Committed to trunk and x. +Had to set application attempt id on the created objects in tests +Now Im editing the JNI section +Sounds like this is something that should be marked Later? Sounds like its mostly done but not complete and perhaps not time to finish and maintain it? +BTW do we need the dependency for this? If possible it would be good if the parsing process didnt try to access external resources. +In this case we would need to process the class hierarchy when placing the proxy annotations and place it on the right class +Thanks Owen! +this needs to be done but not really for the is the what else do we need in mail lists part +Bug has been marked as a duplicate of this bug +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +QE please screenshots with broken fileUpload component after first upload performed +Also could you run the unit tests as Sam was unable to get them going thanks +yes + +this functionality should go in the db modulecommand to execute this script should be like thismvn Pupgrade clean installThe mvn build should call an ant script +Im not sure why you believe that someone can find out if the username X exists or not +Hi MarcYes thats what I meant +I applied the patch and could not reproduce the issue +on doing it in trunk for the next weeks or so. +The invokes this to get the schema for explain statement which is a single string column +AndreasTo use the old stream parser we would need to put more logic into it +We still dont have an XOXO and we havent been able to reproduce this on any other device +printdetailedinlinereport confirms that there is signficantly less inlining happening with the new compiler DNA +If you want a quick lesson on how references work just say +Regardless try using categoryBag or t with findservice +Bulk move issues to and +I addedchanged missing javadoc with see tag renamed isFinished to perform changed to use changed to with the messageI think that I still have not answered couple of questionsKathey On the test in addition to Myrnas comments do we lose anything by dropping from connections to Exactly the same functionality will be tested +Ah did not thougth about this was referring to another way of getting invalid fids insert new features into a feature type thats backed by a single varchar primary key which is handled by it uses the same algorithm as to generate an yeah we should really fix both places +It attaches the to the Get objectSuppose a user belongs to more than one group would multiple instances be attached to the Get objectbq +first picture shows second one my working Thomas wrote multi line support not working tested in and FF IE +This patch contains the annotations that are used for managed bean scopes +Patch With the license +Now we are considering to port it into I would resolve this issue as Partially Completed. +Using the types directly could break the mapping layer +Thanks Eirik +What I meant to ask is thatShould we move the call to Log Roll to the flush on the Region Server side so that even if the client times out the roll happens after the flush finishes +fixed in revision Svetlana please verify it resolves the a lot Mikhailall works fine +After we can compile and run tests on +The test cases are now documented well enough. +Attached is a patch that implements my proposal +Committed to the docs trunk as subversion revision . +I am guessing that the two dates were listed separately because there was no release in +After discussion with Simon from a technical point of view much simpler would be to show directories which included in instead of showing warning when directory doesnt exist +Manually tested +Fix fix init after had will be fixed by +This issue is duplicated by +has this approach been implemented in already +Patch provided bu Jsse Glick merged in fcecbbdcdffcbaedcf +This ones been passing for a long time on Hudson and local runs cant reproduce any failures +Gosh why dont they just report these issues themselves +And since it factor a few other methods too. +Thanks HemanthThanks also to Rahul Singh for earlier versions of the patch. +Then we should probably close this issue +So where is the code for the same? Any pointers +Source code and test cases for the featurePlease review and provide feedback at your earliest convenienceSubmitting an updateI did some code cleaning in this patchPlease review at your earliest conveniencePlease review + +Thats great to hear +Are you using HOD to create an HDFS cluster? Its not clear from your description that you areIf you allocate a mapreduce only cluster through HOD set for your HOD allocation then that setting will not affect a static external HDFS +Examples def abackslashb ab def aslashb abDollar slashy strings backslash escapes only EOL escapes interpolated variablesclosures and itself if required and slash if required use to have as last character or to have a before an identifier or curly brace or slash use to have a slash before a empty string not allowed +Hoping that log would give a good idea about the size of the buffer for large responses and based on that we could make additional tweaksIn the case of NN the log is is fine +OK we decided to commit this beforeIll wait another day and then commit this if nobody objects +system property controls what kind of exception a JMS throws when the prepare is called afterthe connection is broken +This will be in the next update +This is still a research subject +I know this file was sent by our client and it was generated by a software called Gumbo +log messages would have been better. +Doesnt mean hes wrong though +Addressed review comments Added CSS file to be referenced for the AllActiveDone Jobs selection added Custom Filter button for Bundle Jobs tab too +Any news on that +Thanks Arpit +Ive created the class and the corresponding test class +Resolving this issue the patch was applied a little while ago but the resolution was not set at that time +Well I think we can rule out this being a issue as it fails today with the same symptoms + +The same applies to the name of the new class +Use of custom namespace postponed until RC or later +is a similar issue copyField glob sources matching explicit fields stopped workaround is fine for now at least we can see whats new in now +Done forTicket +Seems like a good home +committed to trunk rev +a couple tabs made it into the last patch +Im low on bandwidth but if you had a suggested patch on the docs Id be happy to review and commit it +Maybe a WARNING log entry would be OK for this too? I know in another well used messaging product that I probably shouldnt mention here we used to report these kinda of things as WARNING level messages +I updated the sample to be more comprehensive there were a number of problems +Moving to because theres some uncertainty about the validity of the patch +Committed revision +The same resolution as for reproduce the issue with + +Minor comment weird formatting inThe line is too long +java source type in the project +I agree with you about that then +These sorts of patches dont age well +Closing all resolved tickets from or older +He is too busy to post the results +This may be a side effect of the changes +Huh? Surely GMT does not transition? Whereas GMTBST does transition +Works after making the changes suggested by RalphWill close the issueThanks a lot +See +Im using Cassandra and Datastax Java Driver +This is the side of which provides a framework for such +Nodes joined in the cluster no errors seen anywhere but will check once again +Can you review my patch? looks screenshot of the java docs with the Stefan +Fixed in rev +AFAIK we dont do that until we realize we need to replicate +I confess Im not familiar with which if those hosts is associated with +All I can think of is supporting System properties like +I think we should stick with the form for both expanded and forms +Most notably Id rather have the JMX classes depending on other parts of Jackrabbit instead of the other way aroundAlso to make it easier to collect and expose statistics over a longer period in revision I introduced a simple mechanism for recording a time series using nothing but an instance as the event counter +patch +thanks Freeman +cleanup at the time of startup +Yes please attach the logs +its just not used everywhere it should be used I would suggest leaving it in +So this request is better described as Allow access to the request object in validatedocupdate requestsI cant think of an objection that idea +I propose copying them into the embedded docs dir and not at the top level as they were for this give embedded and standalone the same suggest leaving the top level docs with the following structuredocs this will contain the eap top level docs exactly as laid out by EAP and the seam esb this will contain the updated documents that the docs team have been producingThis leaves us with the as this is superseded by the docs team version Id suggest deleting the with me +The problem with the test before was that the exclude file was overwritten with only the latest node to be decommissioned +bq +Attaching a patch for this issue +The first thing come to my mind is having another attribute such as overwritable at Annotation +Use spring Default Message listener container for consuming messages +Sorry for adding confusion to the matter +Set Fix version to +Thanks for reporting this bug +Waiting for feedback to commit on branchFixed on branch in a combination of commits by gangolli and snoopdave +A good idea with a more descriptive exception message +The problem with changesets is problem with initial retrieval of blame information is mitigated by tested +refuse to free any resources in the scheduler for a container that decreases from to mb +Moving to +Is something still broken related to this issue? Or just inefficient? If its an inefficiency we will definitely be looking closely at this for +Attaching an attempt of an import configuration file +classify and set fix close of issues +Bulk close for Solr +Allow for now a null proxy +Feel free to post an improved note +I think you are right Nick that bug is not strictly a duplicate if this bug +Thanks for your patience! Looking forward to seeing more contribution +Confused why would you delete a file that you just created +so you are saying that if you extract to classes then it breaks? The stacktrace you provide show the file being served from the jar +This guide has been created +Moved creating of to +I think the process of making the doc id absolute is simply adding up the previous segments num docs to be the baseRightbq +You uploaded wrong patch boss +In fact i forgot to put the new files in the patchyou can download the correct patch in attachmentOtherwise i nearly agree with you about jetty dependency +Patch looks good +Complete patch as discussed above +Marshall and Alex for working on the patch +Looks good to me +What version of Spring Batch are you using +One reason to keep this out of the subscription manager was to keep the two components decoupled the way they are right now +Just imagine JVM or Kernel crashes at a random time + +Please let me know if I can help in testing a patch any time +Patch applied thanks! +Added the missing files to functionTestsutil and committed revision +on to trunk thanks for the review Stack! +This prevents the use of as well since it is expected that there is no element in the Cache at some point a will then populate the Cache +uploading the patch to update the version string for release +Please wrong with Web CVS access +No I actually didnt include any of the concurrent structures because it seemed risky I figure their internal serialization mechanisms are already the best they can be +I would a strong on components inclusion as it leads to all sorts of unintended consequences +LGTM +I ran a quick job using and it loads the native libraries fine +task is done. +can guarantee that the Put wins over the Delete even though both happened at the same NOW timestampTrying to understand +Will be fixing these bugs in the next a couple of days +We need a prettier and more intuitive exception +I have been advised to remove the test since all of them are slated for removal due to mavenization +Including a JRE is all very well but if that is installed it should not be installed as the default as that may mess with the users settingsIn which case unless Studio is also updated to pick that JRE the original problem still standsIf you want to provide an additional installation package that includes a JRE fineBut I dont think users should be forced to download a JRE if they already have a suitable one installed +Run the full regression tests +Bulk closing after release +Excellent +I committed this +We already haveresolutionoptionalfor all the sun stuff and the felix plugin doesnt find that package since we dont actually reference it at all +We need to make sure this ability has been brought forward from the stable branch. +Slava do you think youd have time to address these two minor issues Libor has found +Im unable to make it work once I shake out the rest of the issues as its currently designed +Closing this. +There were events still processing setting up still going on when the setFocus is called +yes what I want is an error message between and so basically a validator merge when m basically in the same position as Natalia this limitation greatly restrict the usability of HV +As you point out asprintf was added to Solaris libc in +Thanks Amareshwari and Owen! +out files too +Submit to hudson +Thank you Kristian! I still cant run tests with jars because of the hang but happy to be building +Could any other experts have a review on this trivial issue? Thanks in advance +This can explain why reading a block could take so long +pushed to the dialog says what it should and behaves as it should +I am also using MasterSlave configuration for broker and faced the similar issue +Hi JeffAfter staring at this for several hours I finally realized that the import location is specified as +Added file mult test unit +Again note that one failure does not cause the whole job to exit +Guy can this issue be closed +and then nothing +This method expects tmpjars was populated with s default dependencies before it is invoked thats the extra step taken by the tool +Move it back in if you think differently +Hi Laurentattempting to talk to final release of continuum but we also have a +I dont understand how works but the code in dealing with finalizerQueue doesnt look right to me +Beyond that it is simply a less functional version of full mode +closing since fixes are in released code +Thanks for working on the patch Rob. +There is one outstanding bug remaining to be investigated +So youd like to see the example JSON in the types page as well as the element page +Im unsure Ill be able to backport onto +I took Hairongs suggestion and made file close adhere to the softMount period too +Could we put checksum as part of the image file name +If you dont want it for that is fine it can wait +New tag handler have been created +Hopefully that should solve the basics +The webbundle URL handler is heavily used by the Webapplications TCK +Fixed on trunk for release +Here is some end goal here is to have use the wfs xml support +Your comment leading to this behaviour was so the only thing the server needs to keep track of is deployProject +Duplicated code for additional attributes from the was removed +Mmmaybe +Ive gone as far back as and have not seen the kind of generated code that is discussed. +bq +Hello are there any plans for fixing this issue +Revisions and +It is the default caching library in Apache Cassandra and is the reference algorithm for Google Guava +If this is not a bug then the documentation should clearly indicate that a configuration overrides global exclude declarations +I confirm that now OCJ runs smoothly +Eclipse product +I believe it is a duplicate of +I just tested with and while logged in as Root on there is no Pragma header being set in response so suggested fix would not work as there is nothing to far as I can see there is now only being set in header for request andAre you saying that with Brians fix in it now works or without Brians fix it worksIf the latter then I suspect something has obviously changed in to handle the headers differently than the previous back to you as I think this is now safe to rejectclose based on Ivans last comment +There is no bad location exception +Adds tests for new methods improves one existing test + is using +I think this was fixed when we removed the admin credentials from the +Will correlation generally work when multiple client objects are talking to the same endpoint from different threads? If that is not the case we could also try to use the same client for every endpoint +Im resubmiting the path to hudson +I forgot to come back and close out the issue +So issue found on IE +Is that seconds? I assume its milliseconds right +But viewing on mapping of on the other hand it should be possible to find the a belongs to whileafter loading the mapping +were working on meant its a work in progress +This is a feature not a bug +Did you profile thatYes we saw this as one of the major causes of blocking in and a significant hotspot in the profiled runs + +This is what I did in my generated classes +The patches provided would make the samples services work when dispatching is required by clients. +This issue now occurs on a very slow and busy machine after consumer gets a dozen of messages it stops getting others +bq +Yeah I dont think it comes up often in the real code or the problem would have been reported earlier And the behavior of crypt is heavily platform dependent and the official are not precise enough to describe what should happen when bytes are passed as at the minimum should not fail with huge Java exception stack trace in this case +I considered it a bug that it wasnt starting automatically from +And one block is a by matrix +Deprecates the classes and +DevarajI found the following related to relogin The RPC client should try to when it detects that the TGT expired Relogin behavior for RPC clients could be improved Kerberos relogin interval in should be configurable Please create to port relevant ones over and link them to this JIRA +I added a initial version of the docs as xdocs sample and doc checked in +Hey Justin how about committing this one? How do we handle the conflict with that other patch of yours +I have checked and all repositories on reviewboard are up to date and current git branch master git show eafatal ambiguous argument ea unknown revision or path not in the working tree +Fixed with revision +Yes youre right we probably need to know the vendor in order to find out which jar files to include +I also would like to see just one table because all information in the guide should be related to production profileThis is not correctThe guide should give information for all the shipped profilesWhich descriptions are not rightAnd what ports are left out +Committed as r +Im basically fine with the patch +Updated the title and description to more closely reflect the problem +Nevermind about wrapping it as I see that it already is +Section clearly states that the parameters listed in that section can be added to the fragment component Im not sure where do you see that only the access token can be added to the fragment +Eric please test it and reopen if something wrong +Lucene gets the new Reader from and I would think Solr uses that on soft commit and not something else big and heavy +Moving to as these are not likely to be fixed by +I think the easiest solution would be to move into this is an accident is only supposed to include and but accidentally includes which means the TLD files for the JSP tag libs as well as the file +I suppose it is for +the Solr stopword loading code is more as it ignores BOM markers etcI think it would be good to only have one piece of code for this functionality and for it to be optimal +It was run on and the driver and persistence information is all set up to use the same +I dont see any merit of this config at all +verbose is fine or the old v for version V for one issue here is that the underlying cli scripts that are housed in each platform repository are silent by defaultWe would have to file enhancements to allow a verbose mode +Done +Reporters of these issues please verify the fixes and help close these against latest branch + to trunk and +The violation and measure drilldown pages dont work anymore with the Views plugin whereas everything seems to work well with the Dev Activity the Views or the Dev Activity plugins are executed for instance only once a day their leaves are not linked to the last snapshots of the technical projects and in that case the drilldown is broken +Committed to trunk +Yep + +canceling patch for +I was going to do the as +It should do this on any request when a previous request resulted in an error not just on commits +This issue was introduced with +I propose to use this issue because it seems that Victor and you have to synchronise your work dont you think soThanks for your help +BillCould you please help cleanup thesethanks necessary comments to all the public methods for +I dont think it will work in my case as multiple actions in the controller have the same model attribute up loud how this could be done the Spring MVC way is to add or more annotation to the method and that the init binder will only be used if the request is matched +This was fixed by +Please ask questions on the user mailing issue +Okay so the plan is to make a getter as Andrea suggested and add null checks to keep it from bombing out +Please download from here XXXX +If so this will be created with the filename ending in +With this change a command doesnt allow multiple occurrence of the same option +This will be fine for now +Sounds really good +whats the rationale for doing soI am thinking that the namenode has to spend considerable horsepower to keep forwarding block reports and blocksReceived to +Updated with README instructions for building with the Facebook Hadoop release +Ive refactored much of this and will be splitting some of it out to submit as separate smaller issues +message bundles from plugins now resolve +Im not testing with the data attached to this ticket +The iterator will either have a reference to A or not +Im in a crunch to make the call as to whether we go with NMS or not +Committed revision . +I agree with your wife Greg for the swiss knife logo +Marking incomplete +I would not call getting jetspeed to work with a default installation a specific environment +Can we get the missing images to retest +Sounds great +I think that I will save the discussion of the deployments for another thread thoughWe definitely envision and plan on a service registry component to HSSO which aligns well with your thoughts about layering HSSO over TAS we see the authenticating and service registry components as going hand in handWe are in the process of making arrangements for a security get together at Hadoop Summit next month +Is that still true today? If not can you elaborate on the key differences between the two systemsThanks has an internal metastore that has a similar relational table and partition model with Hives metastore +I just committed this. +Fix for this bug +We dont need new features at this point +I think the package should be under javadrda not javaengine +settings +Filed for the UI and metrics that summarizes what I was asking +Thanks Wolfgang! +JON has upgraded to the latest Quartz we may elect to utilize the Quartz clustering feature rather than limiting the Quartz scheduler from running only on the master node +Added a patch against trunk +can you please provide a patch +I also think it is a good comprise +Will fix +For some reason this causes hibernate to refer to the column in the WHERE clause using the alias instead of the column +The default fragmenter works in chunks of charactersAt this point Im thinking it might be best to leave it as it is +Then both bundles will get resolved when you do a start for example due to the station I explained in the email before +If it does not then i will be unable to use it for the large scale projects I am leading +If push comes to shove it could be moved to +Any committer can propose a project and check in code there +now I see your comments +There is however something I have been thinking about which may help here +This issue can be closed it is not a bug all that needs to be done is clearer documentation on the hazards of debug mode. +Matteo thanks for creating the JIRA uploading the patch +The problem is that we are using DK both for routing and for local key sorting partly because its very convenient to be able to use the natural compareTo to compare those two kinds of DKIf the only place we have DK with null key is for the routing case then the right thing is to convert those usages to raw Tokens and make key in DKi have a nagging feeling that there are more complications though +Henry can you take a look at this one +Committed to after fixing clienid in +Fixed showing latest Version for SNAPSHOT in Archiva GUI when more than is available for an artifact +at r Thanks Nathan. +That sounds like what we really needed +Attached patch proposing the use of to differentiate from other scenarios so that migration delay can be skipped for bootstrapping +If you point me to clover coverage report file I can configure that in Hudson for all builds in Hudson +Sharing is a different issue we can ensure that the MR related inputsoutputsprocessors all set up the proper MR properties in their respective job confs +Is this Project still alive +Thanks Gordon all working again +start namenodedatanoderesourcemanagernodemanager +It should be split and tests should be moved to each moduleWe may also want to have a mock runtime detector to test the runtime detector framework by itself independently of another component +Hi ChrisIm trying to reproduce this without avail +i havent seen anybody asking on the ml and it can easily go to trunk +Please review +Was able to build all successfully while stepping through the thread count from in increments of +Well as far as Im concerned is an extension point +update comments a Ill run the tests locally for the changed version and commit it if they pass +Verified that the test passes now +who changed this issue to Resolved status? +The patches needed some rebasing but after reflexion I also think that we should keep even with that +But this wont work in the sorted or distinct case +Good find +So it should be OK I guess +Yeah I just realized I am doing a total refactoring of bridge and caused me a lot trouble of even fixing my own bugs before really fixing thisIll take back this method and thinking of the simple way +I attached a sample project that reproduces this issue +Verified on revision +why not +see also I dont see what you mean in Comment +Can you look at logs if your continuum restart during the night +Well try and report back on the more involved first issue at a later date +is still ignored on setters for exception classes +When dfs cluster fails to start because of server binding exception log the error and skip the test +The batch is split up into batches one for each server +On some JVMs it looks like ms is not long enough to wait +Jody is right Stefan do you finally want commit rights +I guess should fail gracefully if not run in a environment but I agree with Danno that if there is no specific constraint for an older jvm should be preferred +bq +Do you want to say that corruption happens during running of applicationIt is all a guess at this point +A patch and a few test cases for null results based upon the patch as r in trunk. +can now be removed from the examples along with the reference to it in +Memory leak did not appear with it +I am also not sure to follow about HADOOPPREFIXetchadoop +Heres the backport i svn copied all the code and tests from trunkpatch shows the differences from the merge only mostly just java stuff +If we can settle on smarter rules thats of course a better outcome +thanks for the patch +bq +Thanks for explanation +Easy fix instead of returning null in fetch Ill return an empty list +I was trying to say dont care on the request and it didnt work which I felt was over strictFor assigning a set of containers close to each other Ive opened which is a more major piece of ok cool +Thanks Sanjay! +Fix on trunk and branch +only as phase during the build. +Only public API are considered for the backward compatibility +I have tested it one of tests fail from legacy package but probably not for this reason +Im still off work the rest of this week I hope to spend quite a bit of time on logj this week +Update for +And if there is or more you grab the first +If there is still interest I could post this codeYes of course +It looks like this enhancement hasnt been released yet +But in this particular case we dont really have the information and it doesnt break your first pointAbout your third point the only way to ensure this is to store the full dependency closure and disable transitivity and not only direct dependencies +Applied the patches Thanks +It will appear in the Featured Templates list +Unscheduling issue we cant fix for a released please make this issue describe the problem you want fixed now +the patch looks good to me +I have still another question How to invoke a failure of roll back in testing? I need to invoke failures of roll back to validate my checking +Is it correct that this artifact doesnt have a single dependency +HiI tested with the latest SNAPSHOT current behavior is already what you want a feature with installauto if you uninstall it from the console then restart Karaf the feature will be installed again +The wizard page shows only one warning +I merged the patch to. +Update the patch with exceptions thanks TijsFollowing your comment the current patch includes only the necessary changes and it has been updated with exceptions handling +Actually I think the workaround is pretty acceptable given that expiresin is actually explicit in the spec +looks good to me seeing how hudson handles ve just committed this. +Regardless let me know if there is anything I can do to help +I had it mentioned in the for the parameter you removedRemoving the terminal flag from kills anyideas I was pondering about how to make xx responses accessibleto applications that might need them +Confirmed resolved in resolved too on master +I also had to modify the test in order to match the new method signature and the result of tests +There was a WKB problem early in the RC series that also broke mapserver +I feel like reopening because the requirement seems genuine to me +These changes already went in agree with Enis lets fix it right since this is slated for anyway +here is an extract of what we use in our project now +Merged into the master branch +bq +Thanks for the patch Nikita +Committed as r +See dont want to style the datepicker widget +See +Changed the request check to just not returning a result set when expecting an update count +Need help from poster as explained in previous comment +Lets just leave that one for now then? Ill look at the nd patch soon +Removed cruft. +select and inplaceSelect components use common base popup functionality on M Autocomplete migration is planning on to major because functionality is working this would be an optimization +We have faced thisproblem earlier +The Console debugging and experimenting led me to the attribute supports which is set to true in the class +Im happy to do it just didnt want jump ahead and your commit your patchIf you have time now you could commit it otherwise Ill do it tomorrow morning +ping +Backported to branch as svn . +Ive added additional null check so you should not see such errors anymore +Or an enhancement that would add a switch to texttable to allow missing columns +by mistake I set the to be fixing this +running following tests failed +Whoops +You issue has been split up into Bill. +The user does not give focus to the component before clicking it +I think in my last patch I missed updating the classNo problem +It looks likeresolveEntity has the same problem +Not just authorization +Will change it and upload a new patch +I dont see that this patch achieves the stated objective +Ive got petstore images that fit these names but theyre not licensable to Apache because either Sun or a rd party image provider owns themWeve tried to get a license in a couple of ways but its not worked outAt this point the option seems to be to go find some that dont have the same license problems +patch applied please test and close tested by Michael Lipp +In environments this will still make use of all spindles equalizing disk use over time +Verified with JBT. +By making sure that both the view and the reference are in the double checked locking block +Implementation of spring web flow tags which we use in our project to render tags +the issue on the basis of the latest a minimum I need the compete stack trace +I have added more elaborate information about convergence criteria to the class javadoc of and added an additional constructor to override the epsilon valueAs your problem seems to work fine with an epsilon value of no further changes are required imhoAfter discussion on the mailinglist the behavior resulting in different results exceptions is actually good and should give the user a hint that the current convergence criteria may be too strictThanks for the report your problem has been added as a unit test +Here is the patch +Logs from day the error try having a look for the next patch release depending on how easily it can be back a bit as we already locked down Redback for the +srccontribzkfuse can we remove this file as well? The changelog does not match any svn revisions on apache or any revisions that are open source +The root had gotten full and had some rows with less than bytes and some more and the copy to leaf routine was incorrectly applying the byte minimum to all the copied rows and the copy did not work +please see amilas comment +cannot reproduce assume stale resolved issues +Committed at subversion revision . +All clear all problems got resolved thanks a lot Juergen. +do you think its a good idea to release with openjdk as default +Native files are properly removed + with the nit that the comment about that code being reached only if the isnt triggered isnt particularly helpful since it describes the obvious path through the code and wont be interesting once this patch goes in +Hope this helps +This corrects the defect +Committed revision +That is because of a drawback of Java +The camera code calls a method that checks to see if there is an external storage card mounted +Ive just forked the repos and put out a couple of pull requests instead of pushing to the apache repos directly + +sigh! just setting to null doesnt seem to make object getting garbage collected +No exceptions in and the performance is good +i just tested with current version there are still some issues with the social tab disappearing from the menu +Yes it seems that the blocks are gone after restarting the namenode +I recently migrated from Hibernate to and in the process I switched over to Javassist as well since thats the new default +Alright +Any plan to also add this toIm happy to prepare a patch for that maybe in a different jira +Hi MiguelI think the underscore helper youve added should go in next to the underscore function since its essentially the inverseOther than that patch seems good +Weka does similar things in their explorer GUI. +I ran into the same problem and spent half a day tracking it down +batch transition to closed remaining resolved bugs +Please add IT and hookup manifest mojo with DLL and EXE lifecycle +Applied patch into github master and +This is definitely bad news +Should have checked the console output fatal Authentication failed +batch transition to closed remaining resolved bugs +Ive tried several times to recreate this problem today and so far cantFor me the tests are running fine with or without revision One possible thing that might be causing it is the tests take longer to run with revision +This improvement will be included in the next release off the branch +Now the CMIS Repository Connector uses the GENERAL code for exception handling related to a +Len can you resize the dialog itself to change the size of the text box +Thanks Rohini! +I will review this patchSince the problem is in JPDA support Im changing prefix in summary and remove this issue from DRLVM component +patch committed +If we decide to remove it method begin need to take the assertion string as an argument +The existing one is for Guvnor core +At a minimum we should have a switch for this behavior +You can still add maven builder for this project +see +bq +I will close this issue and create a new one which focuses on bin folder customization only. +The failed test is not related to this patch +I fixed in ead +Fixed +The connector and update multiple in the current implementation and I dont think the bulk case is a good Teiid use case +Addendum upgrades maven integrated to and TRUNKThanks for the reminder N. +Refined this a bit more +Do you see anyproblem with thatYou are quite right +It shouldnt contain have any dependencies apart from standard wicket stuff +Yes to bothWe would need to maintain a implementation for platforms which do not the native implementationsHaving a runtime would let us access runtime information in a significantly more performant manner +Assigning to Arnaud +Alexs investigation shows that we cannot run JCR on database with default encoding UTF +It should be syncing automatically with respect to the contents of ivy files +should this be converted to a unit test +Add to vr patch +One more in +committed and now the parabuild test passed +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +After confirming the current behavior of Groovy I have applied the pull request in close to its original form and added a comment +it sure does +This happens if you read a post with only half your mind on +My patch program doesnt except the diff format that you +Base classDont think vertexId should be exposed in InputOutput +Check out the and classes in the package +Ill be out next week so wont get to this until Im back so feel free to clean it up and commitMike I will assign this to me and get it committable next week +Been reviewing so this will conflict with that patch mainly the ones in +Patch for the issue +This way I think we can keep it more flexible and fulfills more use casesTherefore Id like to change the instance from static member to instance member assuming its up to those who creates to decide whether the should be singleton or notWhat do you thinkRegardsWoonsan +Strike that +Confirming that the version this bug was found in was +Thanks Vladimir the patch was applied to CLASSLIB at rPlease check that the patch was applied as you expected +We have merged to trunk already +Its still a little more complicated than Id like the overhead of threading and detecting the deadlock expanded it a bit +I see it now +This would be a blocker if we were in production +It is pretty independent of BK client and Bookie noI was also wondering if we can code it in a way that does not require making methods public in the meta package +We forgot one important thing hereThe LAYOUTVERSION should be changed in order to force the upgrade for this changeCan we fix it really promptlyThe LV change should go with the code so the best way is to revert and +Closing all resolvedfixed issues of already released versions of Roller. +What exactly doesnt work there? The code in s get method should do the filtering there +tar zip jar etc + +Simple testing using the test programs that I attached to would seem to indicate that the patch for works +The approach looks good but can you move all the checks to compile time insteadI mean while generating a plan create a instead of if thedestination under consideration is on S there will be no move task etcThe explain work will show the correct planThe commit for will be a +Looks good +If you unzip an ear this way though youll get an ERROR message about the dir but its going to go ahead and try and deploy the modules in the ear since they are not located under +Version has been releasedMoving all related issues to the next version +This really seems like a simple thing to fix and its critical when running Zookeeper in Amazons cloud where internal IP addresses change each time you boot upPlease +Feel free to reopen the bug +Sending content CTrinidadTrunksrcmainresourcesconverterstrinidad Sending content CTrinidadTrunk Sending content CTrinidadTrunk Sending content CTrinidadTrunk Sending content CTrinidadTrunk Sending content CTrinidadTrunksrcmainresources Sending content CTrinidadTrunksrcmainwebappconvertValidate Sending content CTrinidadTrunk Sending content CTrinidadTrunk Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadcontext Sending content CTrinidadTrunk Sending content CTrinidadTrunk Sending content CTrinidadTrunk Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadinternalcontext Sending content CTrinidadTrunk Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadinternalcontext Sending content CTrinidadTrunksrcmainxrtsorgapachemyfacestrinidadresource Sending content CTrinidadTrunksrcmainjavaorgapachemyfacestrinidadconvert Completed At revision +Tests that the server when booted installs a security manager or not depending on other settingsM javatestingorgapachederbyTestingjunit Added some machinery for bringing the server up in a separate process +It may be okay to have NULL values in a map. +doesnt depend on Remoting and as far as I know uses netty as its default transport +And people dont know whyWe should INHO turn off the disk spooling in +Alright +Some minor issues in that you mayconsider for a later patch teardownNetwork I am not sure this is a good name for the method since it is actually doing more than just shutting down the network connection +Is that the one +Attaching patch that fixes the issue +The key issue for this problem is assume logical plan in a consistent state +Im not sure how to write a test for this +It just makes it complex without adding new functionality +Okay I managed to track down an issue that the reply manager should be disposed when the route is stopped so a new is when you add the nd route using the same reply to queue +That way you dont have to fail if the two names do not match +Since the data are written in different directories they no longer clobber each other +The second approach is parallel rendering on the client with the Jetspeed DesktopThe Desktop solution distributes the aggregation process onto clients +Im inclined to think the problem is caused by not Seam +Problem originally appeared at but after various source tree rearrangements this seems not to be a Gant problem but an Eclipse Groovy Plugin issue +Fixed by parent ticket. +there are some values +If any of both have to be accurate up to the latest millisecond then I believe its not possible to have a scalable microkernel implementation +Galder i think i have the issue fixed it is not reproducible on my computer so sorry in advance if it be not a real fix +Reviewed by Robbie +moving to doc version for better separation +Mirkothanks for your patchAs Jacques said we may wait for the lucene final release before upgradingCheers +Have a variation of the method that looks at result sets that does not close the result set +Could you please confirm +Derby Integration is complete +iirc its little more than running with XvmmeasureCompilation +I verified this in SVN +Unless I spin a new a RC itll go into +A CVS checkout has automatic linefeed conversions applied upon whereas the dist will depend on the line endings of the local sandbox of the guy building the release +You put me on the right trackMy URI Resolver generates XML from the database on the fly when an appropriateURI is called +The update mode solution seems fine for nowThe only case where items leaving memory would not be present would be if they fell off the disk LRU +This has not caused any problems since the CLASSPATH also contains the path pointing to the dir holding the PIG classes directly +This is just a visual perception in my tests nothing is deleted the Modeler simply collapses the tree on the left you can doubleclick on the tree and everything opens up again +Right now assignment manager is already too complex +Attached a patch setting interval to minute in order to minimize data losses on crushes for Nodes with NOSYNC applied. +The source code changes to enable this feature and the test cases for unidirectional lazy and eager cases +With leveled compaction the different tiers end up implicitly becoming different age groups +The pull request appears mirrored on my end +Eugene Looks like the sasl test failed +The test will reuse data in mailsdb database +Feel free to knock on mydoor if theres some interestCheers enough +Version is not maintained. + unscheduled issues to the next release +Or a true mock that just records any calls to it? This is what I would like nothing fancy +Thanks Knut AndersI do not anticipate any impact for customers who have coded their applications against our published APIExcellentAlso wanted to clarify that I have not reviewed the code change +It looks like something not covered in geotools and should be +Im guessing this is related to the special cases on minimum tokens should we just disallow using that in a live node token +Fixed typos in functional specs +Ill try to figure out if I can create a test for this +Test case added to the codebase anyway +final one +Thanks Shikhar. +Setting for consideration +Committed to trunk +r on trunk +Closing all resolved tickets from or older +I cannot mark the first patch as being invalid and neglected to use a version numberv is named +Closing at Clements request. +As I have been spending the last two days trying to migrate my project to slfj and to fix the transitive dependencies from maven I would dearly love to see this issue resolved +I understood at first that the transaction was created but then not committed I was wrong the transaction is not created at ve updated the issue description +I did not consider the permutations angle +Jesse could you attach your patch again +version is now been updated +Looks like Andreas is on PTO +Great work Garren! +The issue is now fixed with commit id eeefaececabcacNew are added +But that state is only in the class not in the you dont want to use an anonymous class create your own subclass +Im afraid raising the thread stack size limit is the best solution at +or +Verified at +We do not get the improvements I expected with the patch attached so Ill close this as wont fix. +Please ask ScottAdrian for further information +Alex once CSV is completed please review this and determine if CDK or other source for the tag lib items +will not create a symlink without the +Since nothings happened on this for a week or so I doubt itll be done for code freeze +patch fails to apply +Running test suite +I am actually not a student +Scheduled for jBPM +Can this be closed +Youre correct thats precisely the use case +I mean the changes in should be better than though it is not fully reviewed and tested +If the same issue occurs then we can fix it however we will not be adding to code to cater for deficiencies in Hibernate +Confirmed problem +Ive got a patch for it but it depends on one of my plugins being added to build +I will try to get one out this week +I have also checked that if you are starting a machine with more memory and bit OS version is used +Design selected and in production +Here is what I applied to +will upload by the end of the day or tomm +This type of mechanism was used in the sam demo driven by scenarios +is the ticket to improve this +Ive committed this to trunk and merged to +Or do you have a separate patch for giving the editor focusIn theory you cant have a datagrid in a UI and not have focusable objects +Basem ok! Thanks a lot for your help here +Sorry but I will not have time to look at that due to other duties +Committed to trunk and +Im going to work on a new patch for this once Ive cleared up a few more of the patches Im working on +Set correct fix version +Attaching patch +Very curious +This patch just passes the variables down to the subant call +Please let me know whether that makes a difference +the description is IMO pretty clear. +It is simply more strict in its dependency resolution and the Site plugin requires a bit more configuration +Are these shoots as you wish +did some changes on the tests provided and set version to +Also see investigate whether Remoting can be of use here +Because the same wrapper class instance is handed out multiple times it makes it possible for inadvertent sharing to occur +not sure this is relevant information but I had the same exception when running dbunit in the eclipse looking into the code the simple workaround is to pass an extra false argument to the prevents dbunit from trying to set the lexical it with the maven surefire plugin gives the same error so maybe surefires not the problem my guess is that its caused by a wrong version of an xml parser somewhere. +Be in touch soon regarding testing +Thanks Felix! fixed the message for now by removing the message until we have an appropriate page to link to. +The fact that you missed it indicates that the documentation can be improved. +I turned off coalescing for my test case because I wanted to test the serialization stuff +ctions ntusesorgoror uses text ions onuses +Sorry I forgot to flip the Priority from Major to Trivial +According to Anil we dont have an official link +Thanks David! Yeah agreed the class is not ideal +Hadoop handles all the race conditions that would otherwise happen as a result + close of all resolved issues prior to release. +I had an impression that java creates containers with only required features attached +cant we have an option for nuclear +This should load only numberIn the second test method an instance is queried with fetch groups and active +Confirmed +Thanks Andrea so can this one be marked as resolved then +The initial problem reported here is fixed +Committed to trunk. +It would be better to think of security outside of the handler +Ah thanks Andreas guess its better to ask this kind of thing on karaf users mailing list first rather than create spam jira tickets +Should I try with FF ? Or is it something else entirely +attaching a trivial patch +Tested in yesterdays nightly build of +Since Enunciate generates WADL with documentation references throughout it that come from the relevant javadoc levels that might turn into great fit for Enunciate since it already does it and also work for other languagesframeworks that either generate a WADL or people how create a WADL manuallyWell keep you posted but also interested in your thoughtspreferences +code is NOT part of the public platform API and should not be visible to platform users +I have no idea why the lock is static it should be instance based +I attached the patch for the tests +Please the ticket and update this with the model of device so that we can continue work on this issue +Solution works. +Thanks Randall! Followed the updated Building and it works great. +I can do that later this afternoonwhat I am seeing is +We have a measure in place thats in testing now so well close this issue for Maven and come back to the solution in +I can personally insert the date pattern directly in my java class but this will make my program less configurable +Also keep lines up to characters wide this is part of our convention. +DoneThanks for pointing that out +Back to the compatibility issue I think unless we decide to have a release with I still dont think it makes sense to get this one in +The documentation has been completely overhauled in the meantime +My suggestion would be to provide the option and let the user choose based on their needs +It uses a memory index and decomposes the original query to find the matches +OK sounds good +valuethen we can combine the results from client sideBack to this improvement the only advantage is combining the results from server side is better than from client side +are you positive that you have the latest versionsdo you define any custom setup for the +I think the code is already changed to do that in to do with the test setup in the particular AS setup +It includes one additional test case in think now that we also need to do something with the peerEpoch +and etcif fragment must have only one root which must have the clientid it will make the components extension very difficultAnd yes i dont like recursively method too +Looks to me like the failure was indeed unrelated +patch committed thanks shravan! +Fixed the merger updated +This is more use for quick search screen example in warehouse because the screen have just one text inputI propose to improve the actual function and move this on Product module +Ah +Interesting +This needs to be documented in the release notes +You need a bulk edit shipper? Something that allows you transfer MB of edits in one go has that intention else by default replication ships max MB of databq +Oops cut and pasted wrong section in JIRA +Ive read through and I have an idea about what Id like to say and where I think that it ought to ve looked at the code for the Maven Jelly tags and I have a partial understanding about the mechanisms by which the Jelly tag documentaion pages are Ive got a couple of m tempted to put some substantive comments into the and the files that explain the issues concerning the injectionretrival of values intofrom maven plugins via Jelly scripts +Please kindly leave this issue open in the meantimeThanks +Using can solve this problem in a flexible manner. +Wow +The replication is fundamentally a peer to peer protocol +Would a release in three weeks preceded with an rc next week work for you too? For me the time effort required for releasing is the same as for so why not pushing some of the good new stuff out as well. +If you see any other problems please let me know +Sanne per our discussions I just added the related follow up Jiras +To use requires integration of the teiid client jars into designer codebase as this class is not part of the teiid branch +The latest patch by Hoss can be committed +fixed +I think this can only be a problem if there is some kind of NAT between the two local hosts in which case I suspect the original connection would have to be made using a address and then the patch would applyIf anyone finds otherwise please with details! +Guess this whole thing is not really what I wanted +Also this method does not solve the problem that these are and need to be treated as such +From todays IRC discussion assigning to Ignacio +A beer is definately on its way to you if I get a chance in the future +It just has as location +Will be part of dont think this issue is resolved +Not sure howwhy other than that those redmond programmers dont know jack about programming +So the close may interrupt the sending of the end of block message and thus causes the error +Fixed +That would use hidden +Werner Attached is a complete set of xsd when I created I attached and eclipse environment with the complete set of files and with a class for generating all the necessary have tried to trim it down to a better test case but ran into to may dependancy issues +If the EAP releases are still being created from that branch then this is done +I am actually searching for an apropriate setting for XALAN to ignore the DTDHere is my calljava cp CProgram FilesJavalib IN XSL OUT +So yes we should show models in the data role editor +What version isGiven I cant reproduce this it might help it you past the stack trace +Updated patch which adds path separators as necessary + is a +Rather than adding new CQL keywords one alternative would be to take the pgsql approach of building this logic into the shell by querying system +Here is a sample Open Office document +This patch fixes a memory leak inAlso freeing of the allocator is moved from the clientserver code to axisenvfreePlease apply this patchThank you some more memory leaks in axiom svcclient echo client and utilsPlease review and apply this patchThank youDumindu +I cannot reproduce this with my UbuntuLinux either in the DV tried in debug and it appears the wizard is properly listening for text change and the folder location change events and executing validation and updating page else in your installation that may be different +Well strictly speaking needs Groovy as a runtime dependency +defer all issues to +The laptops obviously had different sets of usersgroups than what was on these clusters +Thanks Ravi +This isnt a blocker for any releaseThe issue is that you needed to specify JOBID MAP running instead of map +Also the existing get should be removedWe currently dont fsync the directory file itself and this is totally wrong on +applied however I cut back the duplicated image code +lopex what do you mean with broken codepoint? can you give me an example? +Rather it should just be a way to access configuration parameters +Were having the same kind of problem +bq +If there was a change the section needs to show the revision number as well as the other items for each of the changes +Committed to with revision Merged to trunk and committed revision . +If you beleive that should be please and contact the program team +I did but not with the current SVN +I can increase the memory but sooner or later its going to crash with OOME right +Please review this patch +Unrelated to JMX settingsIf using the in is picked up before the command line option as a result the file is used in conjunction with the specified file +See node for NPE. +Yeah we love contributionsSo just keep hacking on that github project and complete the missing pieces +Thanks Travis +In the meantime you can make the function an instance method and everything should work +Patch for with patch +I originally tried to fix the dependencies of s tests and came up with this configuration which failed for other reasons +Descoping items not being worked on for M into Unknown Fix Version for that this is only half a problem now +No the Filter approach is not an option it is critical to set to manual before view rendering to avoid changes being flushed from the view and potentially corrupting the markup if this happens after the view buffer has been flushed +Sorry I didnt realise that it was already assigned to youHere is the patch +Thanks Ben Ill look at this this weekend +This patch is related to + Username lqd +new patch fixes crawlId functionality for too +Information about it in VPE FAQ has been added so i am going to resolve this issue +The fix +We will fix +Fixed in +miroslav the logs you attached are about EAP but the issue is about EAP CRDid you attach the correct logs +I just filed as in addition to the potential race between the AM and its task there are also inconsistencies between how the RM and NM handle the preemption case +For Kerberos secured Hadoop Sun Java is required +this issue is to create a jira project no idea why several other components were listed +Appending seems innocuous but it is a behavior changeAccordingly Im marking this for the family +Closing issues solved in released versions +We can address the naming issue in another JIRA +Including some snapshot and logs during my test +You should catch this Exception and in this case fall back to extract no termsEDITOr simply use the passed in reader as delegate of then it will behave correctly for all methods + to trunk and branch. +JinchaoPlease generate the patch at root level not under src directory +I will create the issue right the and +committed to master. +I expanded the patch to include use of the existing jar file cacheThis shaved off our startup timePlease have a look + +Is it intentional or just inaccurateYour own patch from didnt include the in +rd time lucky I hope +To me leaving target unintepreted and intepreting it only when the link is accessed seems to comply more with posix +As I will be unavailable for the next couple of weeks here is an opportunity to thoroughly review and test it +There are unit test failures in test and test that may or may not be related Ill close this once those are fixed or proven unrelated. +Would like to get this backported to branch as depends on Delayable which is from this JIRA +stop a mapreduce task if enough values have been found +I will also have an eye on it and close this issueThanksChristian +Now that work is well underway with AS all previous community releases are +Hi Christiansounds really good +I like that idea sounds good to me +Tested with but still doesnt wouldnt it be cleaner to just pass gcBefore rather than the entire controller to constructor? otherwise +Deferring all the from above code looks like a good fix to me would you be able to submit a patch for this in diff format? You can do that with the SVN diff command for example in the folder where youve changed do svn diff and that will create a file named which we can apply automatically +Updating with an unreleased version as. +One other thing requires Java so we can use generics +Hi MarkCould you post your scriptI guess the output of FILTER is empty +BTW winboxes would not pickup ant shell script anyway because they willappend one of +Look like that this is Windows specific +So no need to use the slow method call was more to get around the fact that implementation is in Groovy so isnt compiled in current build in time for DGM but no doubt could be reworked. +Thank you. +Jesse can you please apply this thanks! I could use some more power its easier for me to make and check in few changes at a time +Obviously thats a little different +table a table name pattern must match the table name as it is stored in the databaseIf was addressed then it would be part of Derbys documentation setandrew I feel that a true universal test for references privilege is impossible sinceI didnt understand this paragraph not sure what a universal test is and this comment lost me grant and have references privilege on a column for which it has no meaning to have that privilege +Fixed on trunk module refactoring allows for lower dependency client and broker +Im posting the equivalent to the v patch for trunk +Ive been thinking about this too already Theres a bit too much hard coded in this directionId like to keep this open as improvement for a future versionRegards Florian +As said this API should either be included as bundles or as extension fragement bundles. +This is in subversion now log property added to all controllers dynamically +committed the patch to master branch. +I actually have run all yarn unit tests on patch v I can run another test on v +like it always is +fixed and added test +Sorry I miss that information +Closed after release. +The console log is quite long and there is no obvious way to pinpoint which test is causing the problem +Also the script looks for the manifest entry in the JAR and if it finds it it adds the dependencies listed there to the classpath as well +Marcelo needs more time to work on theAttached v that is just the code based on. +Again it comes down to information that the application already has to make it available for whatever purpose it wants +So a check before the conversion solves the problem +And I agree on your great work I love seeing simplifications like this +I see +I forgot to include cobertura for code coverage +Updated move script to include tests +Increases the data per a map to mb to guarantee a spill +Randall do you fancy updating your patch so it applies to trunk +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Now successfully installed on AS as Windows Service in Windows XP with as database +Lets reopen +Closed after release. + +Hmm I think this is more complicated +MONTHYEAR addremove a month or a year depending if we are after or before the middle of the monthyear +Jupp +Fix pedantic fixbugs warnings add license at top of fileNotes No tests supplied since theres no simple or painless way to run unit tests across multiple filesystems +No the tests shouldnt be running in parallel and Ive been able to reproduce this just by running the on its own +I meant that only one node is bootstrapping to one range at one time +Compute Driver Changes to the createnode function Added support for specifying a disk image for disk images available in the storage pool +The same problem with responsiveness have other new components as well +is there anything else on console before this exception +If I find the time Ill do a backport +Thanks Mike I think this info is essential to add to javadoc for startAnd Each user transaction can only nest a single readonly transaction and a single separate readwrite transaction +If theres an issue with the fix please reopen +It has been applied as expected closing jira +The broken use case is calling the constructor for for a type with null content +Can we add a test to catch the issue found reading the code? Yes it is troubling not found by test coverage +I need to recheck all the logs +I want to make sure we get some feedback on that idea before completing this +thanks. +Thanks Daryn I just committed this to trunk and +Hi JonI am comfortable with your patch +I hacked in a fix which remembers the order in which components are started and stops them in reverse order +Please justify or downgrade +Disabled text is using the property +Fixed in branch r merged to trunk +The projects source folder is specified with the sourceDirectory and not test element in the +No +Applied patch in revision Also merged into branch in revision +There may not be lot of overlap between this and other than indicating storage type as memory in block report +Not sure if adding dependency on is the right thing to do +close the issue as not a problem +If I understand this correctly it would be related to manual editing of the platforms which should no longer be necessaryIf this is not the case please with a more clear description +That is different then I was thinking as I always have a browser or telnet available to test out +rebased committed +I would rather keep the system operational with a warning +We are waiting for Sashas testcase for proceed on you get the testcase done SashaAlso +It may be more desirable to let users associate a table with an HDFS group? +Will not implement this solution +In my case I need to build and deploy the same artifacts with different artifactIds to different repositories +But we are using Cayennes pool +Trying to think through what the correctexpected behavior is hereScenario user configures DN to point to a single cluster which doesnt match its storagePossible results DN keeps running with no block pools DN shuts down once its service failsI think is fairly clearly correct hereScenario user configures DN to point to one NN +Going to fix in +I am not aware of any syntax change between b which the current codes are working to RC or nightly build +Ive just committed this +If you do not want to change parameters you might want to try the stax free parser that is attached to +only small change PM meeting agreed to put into next in CR +Trunk does not pass test case submitted for this issue +Is there a ready made reproducer around somewhere? +Oh cool +So I attached an update which runs without deadlocks +Added the examples to the trunk still need to verify they work and need to add Readmes + slfj is not compliant by default is just the API and the latest compliant is slfj contains fixes on the service not the I am migrating eclipse virgo to run it in karaf +Fixed thanks for the patch +Yes it does look like the site behaves differently when accessed using a common browser +The JVM analyses the bytecodes of the getter method and takes the correct action which it cannot currently do for the access call +Looking to upgrade our Nexus integration in the next +Yes I agree patch is actually enough +Patch applied +Did you look at the picture of what Im trying to achieve +After further analysis the crc in the corrupted entry is also only correct if you assume that there were originally unique names +Ill fix this for since almost all of these are straightforward incubator db replacements +Do you think that the issue is the same as described here +No newsI understand that theres no special reason that my problems should be your priority but for my own planning purposes Id really appreciate an insight into ifwhen you expect to look at themThanks +Slava any ideas on this one +I committed this +the idea is to add wicketpreview + +Uploading patch for commit +My first shot at cleaning up class javadocs +that is typed as a String value etc +Hiwhich in particular? We might test against older API versions but they are stable +this is a lot more work than I thought +Once we close +Try opening a DOS shell or use cygwin you probably can then delete the aar +i think leaves too little space for news +Jeff I tried reproducing this with but was unable to log in with either adminadmin or adminpassword login details +All done safe to close +I will have a look at it this week +Thanks for the help guysSo what is the process for getting this patch committed and and released +URL +CXF is setting to use to read the source file if you dont specify the it will use the by defaultAs the is encoded with MS the compile will complain about itIf you can let the JAXB to generate the file which is encoded with it will resolved the issue peacefully +triaged for. +But for now we can keep it simple and solve the problem at hand by using +This includes the new analyzers for different languages +I havent had a chance to look at this more +The Tika one is not designed for that kind of is not an intentional +Its really annoying for usTrying with this +This patch removes the API link from failsafes in r +So we want a conf variable to decide between distributed and single node splitlog +yeah self updating page doesnt make much senseIf you do that we could just send a redirect to that page so that it refreshes the complete page +But with this distributed setup each successive batch query takes slightly longer and by the time weve gotten to the batch queries are timing out and breaking anyway +This is a very trivial change so it would be nice if we could get it in the next release if its verified to work on other linux versions +ok I will need to get the actual nbms first and zip them up and upload +havent seen this in a long while. +What do you thinkI need tests for this behavior also +The problem would be a huge diff and losing comments and headers in the checked in property filesI must admit that I think our efforts would be better spent on adding the translations to and generate the translated versions just as we now generate rather than creating elaborate schemes for syncing the various property files +The following javadoc comment should be removed from the class A three level directory structure is used to avoid placing too many files in a single directory +It is correct for the holdability to change to CLOSECURSORSATCOMMIT in the global transaction +But it is not the case yet +I verified it works correctly after applying the change Andre suggested changing BINARY to BLOB in deploy +It adds formating changes to the code which makes it hard to identify the real changes +Perhaps I should have said cannot be configured. +RPC retry should be done in RPC level +This isnt in solder atm +Or similar +I thought that this plugins does the analysis on the flySo if I dont have project in Sonar instance yet it wont show me anything +Marked as blocker for so it doesnt fall again +Can someone please apply it +I committed this. +Default clause is required in order to add column name +Ok It didnt work +SumedhaAs I remember correct you wrote the EJB providers so could you please have a look at this. +HIGHLY could set a lower BREE like to JVM instead of removing it entirely +Im not sure if this is the best way to go with the limited usecase in mind +This is my proposed patch described above +Verified for first wookie release +Would you agreeTry fixing the EL evaluation and run this again +Will integrate patch v tomorrow +Fixed in the nightly build +Updated patch after feedback from Jan renamed uses of netloc to hostIve also updated the couchrep Host workaround to only set the Host header if its not already set +I use set the source and target compatibility on the JDK compiler +Still the question remains why single out this exception? If the region was assigned in the meanwhile were good regardless of what the exception +XSOM is Java only while Apache is Java and higher + +Can we provide methods to delete row and records in only? Will that be sufficientRegards think that I will try to work on this one +this became out of context after +Makes sense +Are you saying thats what youve done and you still dont see the DLL +please go ahead and add the jars to +this is class we using for attaching dynamic js is no longer maintained outside of issues please upgrade to +patch for trunk +There is class that configures Jetty Jersey combo it depends jersey classes +Again my recommendation would be to move to the latest release in order to catch up with all the fixes +For +I also changed the font on the small table headers to remove the italics based upon requests and to match the epiq more closely +Changed the call to so that the will not be used +Denis be tested on build includes revision after with JBT. +I agree + +Updated to move to as was issues after move to in +Anyone knows whether version is published to any public repo +too bad this is still not resolved with years behind and votes currently +This has been fixed by Emmanuel and Alex more than a month ago. +Perfect Thanks! +It does not use the Hadoop server daemonsThe mapreduce algorithms in Mahout can all be run in mode +Restults are still very unpredictable +bq +Im going to check this in +Only invalid functions are NOT loaded in the libraryEnded up making the aware of Function metamodel changes and operate on SQL cache as well as notify Transformation editorAlso wired the Query framework to add s as a SQL Alias so a model import dependency is added to model +Attached is patch against to and trunk +You are supposed to use Hibernates feature for handling ROWID +After suspending task immediately the task may not die it will still be visible when ps command is given +attached a graph extracted from +returning boxed types. +If I look at the code it should happen +bulk close of all resolved issues. +Please check it has been applied as expected +So I did a complete update on trunk made mvn clean install as at least my failure looked like the old code being executedAnd the surefire report now looks likeTest set Tests run Failures Errors Skipped Time elapsed it still an issue or can I close it +Lincoln can you use Issue Links please +A shared principal prevents direct communication with the HA because the client will use the actual hosts principal not the shared principal +Fixed thanks +By checking the values in the debugger with the latest code the issue has been fixed +Overall this is a great start + +Then you canadvertise the list using any capitalization you prefer as qmail willjust the name internally +For example calculating the order total from the related line items +for full pagination support +Harry could you put them also in the branch +Close. +Thanks Nicolas +as far as I know this is exactly the case the NFS drive has been soft mounted +I agree its not a critical issue +Agreed seems too trivial to add +As Maxim wrote the text selection in the visual and source parts is the same +Actually I could probably make this a target and do all modules at once +This bug has been marked as a duplicate of +so you wont see the above problem in +Log level is now settable through system property after release. +Is your vision for this to always be a separate goal or is the desire to actually merge this functionality into the original functionality somehow? Is this patch still up to date +Hi Mamta just a general questionWhy is it necessary to allow users to create a tables in the session schema when the schema is not created explicitlyThankssuresh + +Have no clue why the WS Tester would be interfering in this manner but am diving in now +Postponed to to we wont release a in schema refactoring branch as a side fix. +Heres the final one +The original problem I saw was just the executions defined in the parent got executed a bunch of times +Ideally Id like that team to be able to administer their profile but really dont want to grant a whole team Sonar Admin rights +Filling in the reviewer field +A big from meI already proposed something like this hereBut at that time I was not luckyMay be now we have a better chance! +bq +Looks good to me just some minor things +Is there any other way how to get thread dump +older issues for Apache since were no longer actively working on these at the moment. +I will run later the unit tests and give you more info +Thanks +SVN trunk or a nightly build +Patch applied +Thanks Jukka this solved the issue nicely however I just noticed that the javadoc for that method is now incorrect because it claims the method is synchronized and points to + + +push out to +What about readers? Those who can read different dbs they wont be able to read them +Please seeBut I cannot reproduce the same failures on my Mac which seems correct as this issue is +This is because the installer badge is embedded via iFrame so the get Flash Player opens inside the iFrame instead of opening in the if I remember correctly I did it that way to allow the Badge to be used on other sites like blogs +Check links to this document on Magnolia and SBS +Added simple project which allows to reproduce the issue +Patch applied +I think we will need to adjust the JS client lib to break up big STOMP frames into multiple WS framesThis patch fixes the issues +It doesnt tell anything about that particular implementation +Looks like Thomas rolled back the TC version back to to fix any work is done on this JIRA issue TC version should be brought back to is in blatant violation of the servlet specification +One reason for doing this is to address various deserialization errors that happen in EARWAR deployments in AS as determining the correct classloader may lead to incorrect results in a hierarchical scenario. +Bobby just stepped out but yes we are very interested in having this in +Francois your second comment is easily fixed in the map file but I need a bit of clarification on how to fix the first +I think it would be a very good thing to maintain support for multiple users agents as it provides flexibility to crawler operators to be lenient on how webmasters spell the crawler name in their +Of course that is the case w the MR implementation too so this is not a regression +Also update the title to reflect the real change +I cannot move forward with the work I need to accomplish without a fix for this issue +Please if I missed the point. +Did not review the agent changes +I believe the hudson failure is caused by the old jute files still being in the workspace at their old position in trunksrcjavamainorgapachejutecompilergenerated although they have been deleted in SVN +I use git and use git pull rebase to get a similar effect on my working branches +Ive changed the title because its important for this bug that there is a panel inside of a sheetVolker may you take a look at it? Im not sure about the support of laying out inside of sheets +Also sample does not include which is shipped with Tomcat but not Tomcat +Barry is the maximum retry count issue if you wanna track +Like all components does support the enabled flag +Completed +Ill try to reproduce +Or perhaps nodes shouldnt accept hints until they are up +VanI think the problem is only with JPA connector as Randall pointed out on IRC the problem might be that the connector thinks that the data was compressed so it tries to uncompress and thats the problem it tries to uncompress data which never have beem compressed +BTW please note that +I fixed the schema so that the dates are mapped to a TIMESTAMP column +this patch adds a limit parameter at the jobtracker and throws an exception on job submission if the exceeds this size +Can you refactor it slightly so that theres a single return statement at the end instead of one in the if statement? Otherwise looks good though there seems to be something wrong with Jenkins +We will use it after we get to a point where we can take a valid vote as a community on our first release +Thanks Hairong! +But regardless it shouldnt hurt to the in this situationThen how about this for the locking structure of snapshot operations Lock the FSNS lock in FSNS before calling into methods +I still need to have a look at the existing documentation and add a few test cases to our CTF +Rebased to integrate changes +I will post the patch after this goes through the review +IRC discussion suggests that we should prefer the Cache approach +Oh +Im +Would you mind submitting a patch? Many thanks in advance! +deployed oozie and run the distcp example successfully with Hadoop +Hi DagI had a look at the changes they look good to me +I am looking forward to receiving yourcontribution +Filed as a task to check it. +If you are on linux you can create the patch as svn diff on windows you can crete the diff using your svn client +This issue is complete. +This allows you to serialize every Configuration into a tailored version especially designed for serialization +Ill commit this shortly. +This could be due to which is being fixed in ER +If you find further issues please create a new jira and link it to this one. +In this case there is not necessarily a heuristic if the RBCOMMFAIL comes from the first resource we will attempt to call rollback on the subsequent since the first RM rolled back +Oops +Only hdfsxxx +A nit the code should be in package not +Class is missing in the submitted test +If you are saying that generating more code twiddling things around on the stack and generating more stalling Intel instructions is good for performance then Im clearly reading the wrong optimization manuals and going about clearing out long standing cruft in the wrong way +The patch looks good no comments. +the fact that currently picks the last one is just an artifact of how it was implemented +Fixed checkboxnull program in released in. +Not sure if this is an good ideaAny suggestions +I went ahead and modified the to use the tertiary operator to set the username local variable which is included in the code drop +Moving it out of test and into the example directory seems a bit counterproductive to me +Attaching the migrated patches from to am unable to build activemq as per the instructions in repositoryorgapacheactivemq +I think we could add methods to the hook as we see a need +Assigning to Brian M +It should be reading from all the intermediate map files that an earlier stage outputs to find column vectors +Ill do the initial version +Can someone please reviewcommit it +But it may have happened that since that time the domain expired and someone else than Louis grabbed it and owns it now +If i hit refresh button at View Repository Connection Status Page exceptions triggered again +Dont forget to remove profile from project +really no need to do all this heavyweight stuff for such a simple patch +The installation of CXF changed with EAP unforseen but it has a simple workaroundThis is not the blocker for CXF though +This feature will also be useful to avoid conflicts between plugins that generate at the moment +defines an interface for accessing a queues parent in the recursive structureComposite This provides functionality mentioned as the part of +Thanks Tucu! Here is the updated patch for including the new docs +IAC lets continue this discussion on +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Dont know this has anything to do with any test failures just seemed a decent place to note it without creating a whole new issue +The base tests should run times embedded client embedded xa client xaI am going to go ahead and commit this patch after I run some tests and then you can put in a followup patch to add client xaKathey +Fix the unit test case name +Wouldnt you mind updating also the documentation? We have a section in the manual which covers public dictionary properties +Patch committed to trunk +Sorry Pavol I missed that +Thanks +Added also default styles and committed in trunk rev +Nice catch Dmytro! Patch looks good to me +Fixed and renamed to are excessive test under jaxrpcxoprpclit +I understand none of them are ideal. +Actually it doesnt have to be JDK +That is precisely what I have been referring to +Not sure what this issue is about we already have svnkit +How about the other questions +Why do you need a new db column for this? Why do you not simply set the lock to the desired interval +I also updated the Delete test case not very used to writing test cases but I tried Hope I did not mess up anything +Closing this as duplicate as shows the latest misleading log messages +Does the dependent needs to be loaded with the same as the one that was used to load the? In other words do both and the needs to be in lib +verified that for EAPEWP ER the drools version in seam is still and the release notes docs status is Not Required. +Our plan is to only have openjdk as a backup in Whirr we are working on finding a reliable way of installing Oracle JDK +In the meantime the forums are a better place for el discussions. +Its only possible to have named module lib definitions when you fiddle with iml files directly. +Thats a valid issue and thanks for the screenshots but there has already been one opened for thisSo I am closing this one +Them should be discussed additionally +File endings or line endingsWhat solutions do you suggest +However that internal version provide new MR API and is compatible with both hadoop clients under the versions of and +My own take on things involves implementing proper receipt printing for requirements in Canada +Just a helps producing more expressive tests. +Thanks Jacques +Closing all resolved tickets from or older +I think this is intended behaviour what if you change the pluginManagement of child modules +Thanks +What to do you mean by the error occurs David +The node and domain code need to be fixed so that this sample no longer needs a tomcat or jetty dependency +fixed jdk with on Tomcat +This has been changed to just be an ssl cert and key for each instance +It was a quick fix to get my tests working again but it could be done without it Im sure +I will add back old method signatures with deprecated on them +I dont think this will be easysafe to do before GA but Ill raise it with AlexVictor who have been doing this for we validate tag can you attach some examples +So it would be good if I could download a complete build with this fix included soon +Just wanted to let you know that some substantive improvements are on their way that will accommodate your example +You only get the output redirection with Java java +Do you have an idea when this feature will make it into a release will it make it into branchDerby is one of the DBMS we use in our application and currently we are working on tools to help in the migration of our data model from one version to the next +In that sense memory overhead is same as deleting M files in one shot +Please review and accept +What could we do to resolve this? Use Servlet +should not generically try to lock and possibly wait for a resource to be available to log it +So based on this I do not believe there are other places in the code where we need to make any changes for missing stats for single column unique index. +So have sence move to XR skipping XR version + +Remy states this issue is fixed in AS +Ive committed this +This is another bug +Fix confirmed thanksColm. +Missing tasktype definition in MDL +Affected repos! Plugins are pushed versioned changelogedI did not test them all +With rev some app client persistence unit refs work. +Fixed the nit +can you please add tests for this +Sounds like a good plan +I added a to the EJB jar and it worked as expected +Few gotchas with the new jar naming but I think Ive got them stamped out now +bq +Fixed in +Overriding the receive side makes sense but why override the request side? You could just as easily just add the appropriate header +Your call +The changes made in make Accumulo use fully qualified urls all the time +This file differs between the one shipped with the zip distro and jar installerserverdefaultdeploy +I couldnt reproduce this with the latest remoting plugin +Ive got a fix for this just need to test it a bit +Committed to trunk with revision +IBM says its fixed +Ill backport the fix from Username djencks +Indeed with the proposed solution asynchronicity is only guaranteed when a connection is established +This new patch was required due to commit of +test hot test suite still has failures and error and they are different in Chunrongs report +This patch solve import bugs and noew the set work correctlyMade some changes to retrieve the contact if the party already existsNow the completeSale method of eBay works in the correct wayThanks corrects a typo for the EIAS token +Considerable changes have occurred in theinterim +just does an svn move while is the necessary build changesId appreciate a review +Patch applied on and all issues that have been in resolved state for more than one month without further comments to closed status +Hi Thejas M Nair and Jason Dere Thanks for your comments +But I do want to get the DIH bugs taken care of asap +Hi BrettThis is on +I used the profiler that comes with Java +The didnt comment the code I add strVal! null to avoid NPEThe while loop is here to concatene the string value until the ending string separator is found +It is implemented in ticket should be closed. +I used your added log info to help me find where the issue was occurring +uncorrelated scalar subquery planning allows the expression to be seen as basically a simple scalar +Upon doing a release build of Ivy all existing tests passed +Deadlock related issue +Thanks for the fix! Ive added you as a contributor btw so you can be assigned issues +Not a bug apparently. +Uhm guys even putting in line breaks after select from and where clause parts would be good enough for now +BobbyYou are right I should have done it that way +I found a bug in my patch when recoverLease throws exception I should have kept the original logic of triggering Here is addendum for trunk +Attaching a patch that explicitly supplies the directories to cd into when starting the service instead of just relying on the home directory +All projections the test scripts has been ported some of them modified and pass with Java update on Linux Gentoo +Looks good to me +v is the right diff +Hi I created a little sample based on the sample provided with see language selection dialog either select english or the selection no further dialogs in the Task manager there is a process running but on screen not happens +Failed test passed locally +The places where I havent used this service are get get get servicesThe Accounting Preferences screen now shows aggregated values and also if the value is inherited or notI have changed all forms that were making queries to entity now organization parties are filtered by INTERNALORGANIZATIO role +Does JMX apply only to Network Server mode not embedded mode? Embedded mode is mostly discussed in the Developers Guide though theres not an obvious place to put JMX info there +lgtm +I can move that somewhere else +Committed to trunk. +Thanks Raghu! +Look forward you fix this problem Thanks RegardsDevin seems to be used for properties elements and their parents containint the stylename +Dup of closing +Thanks for bearing with me +I just want to delete one of the modules +Plot showing memory consumption with onsubmit on poller +I am not familiar enough with these internals to write a test for this issue so Ill leave the issue open for a little while +Charlie does your fix fixes the disconnected? call? Right now it returns false +Things related to bootstraping and SCDL reading may not work yet +bq +Please share all the different items on the mailing list JIRA is not a replacement for a discussion forum like the mailing dont really think this is an issue since this is somewhat normal behavior +If I extract the somewhat off topic bitterness I seem to see +This issue asks you to provide an API for create an empty schema with a given target namespace add type definitions add global element definitions write the schema to an output streamI added corresponding to reason is that the generated schema must be cached by the otherwise we will have multiple File IO for every call that needs a generated at the javadoc API for we dont seem to have support for that +do you see an explanation? because if no code change for this i will only see the same error for temporary queue messages and persistent queue with the addConsumer fix is fineyour approach seems better if you can have more control on the slave side +After updating the local source files from svn just find the version for snapshot is changed toSo I wonder which version we will include in for I am working on the JMS portlet I need to the right version to work withThanks for giving a comment +This requires interceptor module +Great +Included in release. +Or remove the text from those pages and make them point to Libraries section in so that there is only one place to update in future + +Graphic shows previous support for this feature and suggested location on the store configuration form for reinstatement +Please describe your problem more specifically +I think that we should focus on just the none default analyzers at this point and we should think more about the +I think we may have been generating off of different versions of +Removed hard coded path +I do know there will be many reasons for new release +Because we are using the CTS to verify compliance instead of manually scanning the source +the bug is not in the plugin but in Maven Core see fixed in Maven about Maven? Is there a fix for that? +okay works now basically it was because pagerank input adjacent edges were marked as double instead of nullThis broke the serializationFixed build is fine +Hi again Simone same deal for arcgrid +And this patch introduces lot of changes and new interfaces for the purpose of future work which is yet to take off from POC stages +Some maps treat setting the value to null as being the same as removing the value +Meantime these additions are kinda sweet helps get numbers w how lzo changes the performance story +committed thanks! +I have not yet added any additional unit tests for this patch however all existing unit test pass which may not be that significant as that did not catch the DBQ issue +Closing this issues as Wont isnt support in EAP up to now so Im closing this issue. +Nitin +This property can be set on any deployer that extends. +Excellent! I have applied in Rev +Builds OK both locally and on Jenkins. +Moving to since this issue is not a bug fix +I just cant help imagining the future of this I want a non scoring collector which accepts docs out of order and keeps its next reader +thats more germane to than points out there is already a patch for this in +Need test cases especially with false eastingnorthing +Here is a new version of the patch No token filters no changes no invariant properties used Only that either returns spelling suggestions or rebuilds the spellchecker index is cmdrebuild is specified instance is no longer static kept example doc still using absolute path for index dirIm still recovering from a h timezone change so please shout if I missed anything +Ok will implement this in trunk and probably backport to as well +Kathey +volunteered +The TCK test for this would be heavily CDI dependent so I think it really belongs in CDI +Also ratcheted down the of calls in input split read ins so we dont call too often and clog network needlessly +In this issue we need to change the failover method in to throw a which is then handled by the handleException method in +Marking closed. +I think you are right Arvid +Looks good from a quick run through +We decided to remove for now +The issue is a systemic one concerning how floating point numbers are handled or not handled as the case may be in some cases in the AFP rendering systemCoincidentally r which Ive just applied also addresses one of many symptoms of this problem +Reopening for handling +I just committed this +I plan to add the Labels plugin when all have been merged +for now this patch just forces a login dialog to pop up +Though I has fixed this issue we should still warn customer not to do this +Fix the findbug the the changes Versioning scheme +The class not found mentioned in the stacktrace is in Xalan +libs +I traced the problem to needing to install the package +The one fact that bothers me is the use of swap without a negative impact on performance +Even better there should be a way for you to provide your own formatting rules +Must have been a refresh issue the wwaction code worked after removing the webapp cleaning and. +Hi Markus +Next I loaded up the checkout in intellij and applied the patch through the intellij svn client +if the ipalias deletion fails we log it saying that the ip alias cloud not be deleted +I concur w C Fletcher that some consideration to timezones and daylight savings time is needed +To paraphrase your proposalThanks for summarizing +Thanks for your commit Jeremy +Thanks +Asafs going to take this +On the other hand we do need to let them create a duplicate realm because they may need to switch from one to another +Without it Status logging will be unavailable +Hi AjithAs we can see the problem is in when we change in Axis RC everything works fine but once we use current snapshot we are getting this error +Sorry previous patch had the wrong name this one is correct +Hi Jacques +Patch applied ThanksPlease cross check and close. +The bug is solved for me with the file correctedI modified only this file to add the css renderThanksCordially. +Ouch! We have to develop the tests for this +Committed to trunk and to address the implementation for. +Corrected javadoc +merged into maintenance +were not developing Deltacloud any longer +I applied the patch and its working now +So how would check actually ever detect an overflowJuergen +you for your review +oh waitbq +SureIm not really advocating for jsp +Stefan can you please resubmit your patch? Please include only part related to the process status +Please note that is a for this patch +We can do this of course and fill it with the by default +closing issues for released versions +I will wait for feedback till just applied the patch above by cschneider Refactoring M cxfdosgitrunkdswM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerA cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerA cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerA cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerM cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerA cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerA cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerD cxfdosgitrunkdswsrcmainjavaorgapachecxfdosgitopologymanagerAnd more +Default value is admin +dfs its probably not linking Try my patch in + fixed a problem someone had now and that person doesnt buy much for the answer that well fix it sometime +So a in weeks should work +Nice find! pending jenkins +If your change does not break a test then its likely you didnt break anything +Ok good work +looks +there is no community release +There was a minor issue when invalid tarball is specified with tarball option then ringmaster failure occurs and return code is while for without patch return code is for the same scenarioI discussed this with Vinod and Suman offline +IT updated +This solution has solved the problem for the time bieng +jBPM will be added after the Web project +This is by far the biggest problem causing failures in the AMP Labs Jenkins setup so Im going to start working on it today but since I dont have any experience with cgroups any pointers I can get would be greatThus far it seems that the problem is that the cgroup is never being destroyed because the executor process thats associated with it is never being killed despite repeated attempts +Any chance of review on this +Youre right +the generated xml files in the src directory are putting variables in the wrong filesPeople who are not In the know will put configuration variables in the wrong file and not get the effect they desire +This mechanism is definitely +Gary has a proposal for a more general classloader design +Taking a look at this to make sure it will work for the SOAP binding as well after some feedback from binding looks AOK with this change +I will apply it shortly +The reason this does not work afiui is because when the client logs in a security context is set on the thread local and used during the threads lifecycle or until the client logs out +Thanks +Tested with your changes against and +Let me know if you see the same behavior that I am seeing +META +Corrected the logic to display headers on desktop browsersThe headers are now added only when the desktop site is loadedinstead of the mobile site +should be picking up the file in the resources directory +Committed the patch to trunk in revision +My guess is youre running out of memory +What happens is that we start Glassfish and our bundles get deployed +Hi TimothyWe encoutered a memory corruption while testing the latest snapshot in the lab +Yes it looks a bit like a problem +Works fine thx. +Hi Juraj The tests are intended to run unattended on Jenkins but they require DISPLAY to be set? Is there some dummy value we can try for this in order to get the tests running? Great! Do you mean that apart from the first issue the tests can be runThx +This is fixed now in the latest and +Thanks. +This got fixed in so resolving as a duplicate. +Going forward this should be less of an issue since the new numeric types all do validation on their input +I believe I have fixed are still some warnings +What should i do to resolve it? i tried to replace the port with but i still have the same error +I tried it back with my code and still get the Exception +Resolving as for now +Could you try with to confirm there is still the problem +is not cached anywhere or even put in a +I think the path is not correct or there is a conflictSure I will update again and check but I believe the change to from may cause the conflict since that occurs in the mergeAll method which no longer exists in the patched versionbq +Please look at is the status for this task +So any help resolving my issue in would be appreciated +At the meetup Ryan also suggested allowing a compound key mapping +Ill make a patch for this +Added jeespecs to the classpath as shark needed a and I didnt know where to get one so I stole this reference from the common The build succeeds now so I guess its right +Hi Miss MaricaAttaching patch for this issue +Quick update I havent had time to progress these issues because Ive been working at a customer site but I hope to devote some cycles when Im back in Malaysia next week +Moving out of EAP Post GA bucket +Can you give me a bit more Log +Both patches are committed to trunk +There are no differences to merge +Can you please try it out and give some feedbackMartin yes these fixes will be in Roo +Wow +I also added more error handling +The suggested patch is essentially correct except that Locale will be moved from to and additional Locale pluggability layer added so implementors can add custom policies for determining the locale +closing issues associated with existing releases +Those in the of the should be looked at by Daniel because I am not sure if that file is used we replace the ones in the testimonial +Re source formats for Debian packages we discussed this on the dev mailing list a while back I even volunteered to that a look but fail Ive not managed to spend any time other than teaching the build process to understand +I ran the axis tests for this patch and all the test cases were successful +What kind of things do you need as keys that cant be serialized into Texts +Everything fine just probably need to improve documentation. +New patch with test cases for testing with MO +Most likely will probably not be completed until after the branch is merged to trunk +bq +The difference between sessionCreated and sessionOpened is in how handles them +Yeah this code comes from Jaspell +I would prefer to keep the forgot password form as simple as possible so either ask for their username or their password but not both +There is nothing in LDAP specification that requires the data to be sorted in any way +Regenerating patch +Fixed in r +Ive turned the off now but I dont know any easy way to undo the formatting changes +It appeared later than the original patch here so there was nothing to look at when this was first submitted +Try downloading the latest source code from CVS and see if the problem has been fixed +If this really is the correct entry then something seems broken in the cartridge selection dialog +With that you will get a new configuration and hence a new Appender with each +lgtm if no one objects Ill get this committed +Reopen for backport analysis +Could someone pls try this patch +It will eat the uDig memory for breakfast if you make it cache a large to Andreas concern I was also surprised that this tutorial class was being used in production code without modification +LGTM +Moving fix version to ER +looks good to isnt an equivalent issue on Windows only for platforms that signal on failed socket sends +This has been is no class cannot be resolved to a typesrcorgjbosstoolsruntimecoreutilline Java Problem +I have corrected those and refactored the class a little +After many many hours of struggling with this I solved the issue by rolling back from to +I can try a few things to reproduce itIn the mean time Ill attach a patch to Deployer but if you can do something similar to for the that would be great +Nice +Can you try an update +This is a known issue with how we bootstrap the RVM +By default all EL errors are swallowed you dont need to set that flag at all +But this means a need report to a should handshake with Backup Node before it report to itAnd other blocks changes should also be forwarded to Backup Node such as block corrupt +Probably works for goodNeed to update the readmes after this change +Is this available in the console? I did not see anything that mentioned https in the drop downs around connectors +please review with this may have been fixed by the other processing really dont understand your description can you please attach a sample project +Committed as svn +Added provided test case to show that problem got fixed by closing stale resolved issues +Of course readLine should still swallow a r if it immediately followed by n +Patch applied thanks! +We should also define a description for the Apache LDAP API +Im nervous that in a connection a buffer associated with the socket will fill up and writes will block +Simon OK I will work on a patch that tries to maintain the package structureOther than package structure is there anything in the patch you are uncomfortable with? I can either try to unfix any small fixes you dont like or create more testcases whatever makes sense +When the request size exceeds limit the multipart request shouldnt be processed at all +Same is true for faceting replication queryparsers functionparser +morphed make Issue Tracker use the slave DB server for reporting and url should be configured in the needs to be added to the RE release notes so they know to update the file in +Good luck +The error reporting is bad because the exception is not handled properly this will be improved +Just did some basic tests +I agree a rewrite would be good +The original patch contains duplicate in all scripts +Text Areea doesnt work correctly now +Committed revision +And the first just has no other available yet thus it sets null inThis means my committed is totally ok because the second really can be null +Despite any comments above +I did not find it when I opened this oneAny plan to put it in the branch +Why is it still unresolved several years later if a working fix is known +Username cazzius +Is it at all possible to maintain a branch of this tool that is compatible with Eclipse +Im starting on it now +Any idea what the failed unit test was +Let me submit the files to the central repo +I just did a straight translation +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +OK +problem when using a relation field as key for a using a annotation +I tested with H and did not see any problem + +Can this patch also include changes to remove jboss repository from the main +Hi Rainerthere is no demand for a is working fineCheers need for a change. +for the to trunk. +Committed at subversion revision . +Also I have created instrumented jar files before and after the patchs application +Its on the right track but the test method stops as soon as it encounters a difference thus we cannot have a complete overview of all the differences with Octave +This looks to be inside javancss not the maven pluginAm I wrongIn the case this is true we should open a corresponding issue in javancss jira +UploadedMaybe we could include this in the main distribution +BTW please remember to send pull requests to upstream as well as downstream otherwise we will end up in an almighty mess +Can you create a phabricator review board entry? Also can you describe a bug a bit +The title doesnt reflect the change in the patchThe port was originally so I left the description the same so that it made sense for users reading the change logHow about we do another update in this JIRA to change the default port to as Vinod suggested +Dominiquethanks for the pointer that makes sense to me and fits other problems I am seeing where locks fail on non existent nodes during journal replayI will give it a go and get back to you +However where the specification is silent on a particular issue facing the implementor or the described behavior is clearly illogical then we follow the behavior of the reference implementation +Patched webwork to support standalone operation for builds from distribution +A global variable is a layer of indirection +push out to +Sorry for the late follow up on this +The surefire plugin was release today but now all my tests fail on the +Resolved by the addition of the new scripts in bin added as part of which will be a part of the release +Could you please reopen this issue or is it wont fix +CMS and staging area are setup +Revised uploaded to. +It is also good to surface these errors in the metrics +This is what I did Currently once it is decided that a becomes too large it is closed and a new one is open +names should not be aliased so i guess its because you have other columnsproperties mapped that you dont list +Wrap the long line +It can even have basic container to task assignment logic +OK thanks! I will fork the plugin and see what I can do there +Should be fixed now +Thanks for the report +Resolving for please verify and close +Fixed in acdeedebdcdcbaafac thanks for the patch and sort about the delay +Ported fix to +A possible solution for the problem +Hi Joseph Ill have a look and let you know just need to finish some other things first. +This is often thrown when you didnt build from Maven +Hey Leon it may be worth seeing if now CLOSED fixed this scenario +Redirecting JAVAHOME to JDK and rerunning resolves that problem +For decimal precisionscale support we decided the former +Improved code according to mentioned descriptionThanksAwdesh Awdesh Done at rAshish Vijaywargiya +Jarcec thanks for this! The scripts were not supported until recently so its possible the syntax in existing scripts is incorrectI left some comments on the review +It seems the latest stable version of railo uses Apache Jakarta Commons version +So the class should live in the shared ilbrary or in a shared library of its own along with the jar for everything to work correctly without altering the servicemix server classpathShould I add a ticket for this request or is it ok to reopen this one +Identity mapper join with concatenation would be the most straightforward scalable way to do it +json file as input to the input stream to test it the errors are identical +presumably the new regionserver has a new startcode so therefore the old one with the older start code cannot exist anymore that seems sound to me +Thanks +Ok the problem here is that when we have jars in jars we end up those contained jars repeatedly for every classloader resource we try to load from then on +I think the private fields factory and instance need to be made final to ensure safe publication between threads +Patch which adds specific apis instead of generic ones +Taking Nicks comment into consideration Id vote for and +Ive gotten the same problemPlease fix it and release the bug fix version. +Can you create a JIRA and a patch +Verified on EAP ER +CNF exception is born +Please open separate tickets for other bugs you are aware of +Good idea! Im not sure why it wasnt done this way from the start +Then when you bring it back up and do a failover to it it wouldnt have been listed for DT renewal +Attached the tested results +Carl are you still seeing tests failing +I clicked Close rather than Comment +ok great Ill have a look there +to be able to share the issue +Coded up the necessary changes need to add a couple tests to verify the uri options are set correctly and applied +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I dont know how I missed that +No need to apply the pull request. +Typical example Mac OS X Lion normally I would probably install the bit version but the universal installer automatically installs the bit version +This change has been released with version +So every ask request sends all outstanding requests to the RMAre you sure? looks like its treating it as a delta protocol to me as it is not clearing prior asks when a new ask arrives +Implemented via Integrator +For those playing along at home is where the desynch between JMX and is being addressed +Ping +Appearantly fileset mapper and entry are not merged as expected from the different type definitions +I concur with Grant on this the dependency from test to demo has caused me annoyance as well +Were using and for now until something better comes along +Unrelated +Ive seen every one of these stack traces +Steve Yours appears to me to be rendering right + +If this is not correct please feel free to toggle clustering off before importing andor if another file is required for clustering please dont hesitate to ask +Let us konw if you can still reproduce it +svn ci m Changing the current to a per +If the hardware is broken few guarantees can be given +patch v looks good to me +nologer with MOJO +Applied as is at r. +Tools like RHQ can probably do a better job of these over time period +Patch updated for latest trunk +as I wrote this problem still remains in Screenshot was not update but same problem in tested with it is still not we found that it is connected to the AJP Connector with Tomcat +Will wait a while +I am yet to figure out what needs to be done to snooze the warning but the tar did get created without any issues for me +But do we really need another jira to fix a wrong commit +If youre using then an example code would be s unfortunate that CDH chose to integrate Hadoop on top of the shell utilities as well since their behaviour is quite different in some cases and as the Hadoop keeps evolving well probably have to track down bugs like this for a while +Rest of the change looks good +It cannot be eliminated as it saves us from corrupting the logs +We can probably cleanup the even more but for now keeping simple +It looks that code for is missing inPlease check +Ok I can do and it would get ring into disagreement until upgrade to is finished with is a good thing in our case because users wont be able to mutate it +I actually looked into DBCPs implementation to see where such a check could go +I have raised for the problem of Roo deleting comments +In which case this document is ready for QE review and is available from the link in the description above +Refactor patchReplaces with an abstract class +Can we instead write directly to the underlying out of +NastyWith this issue indexing fails although some documents seem to be added +Please fix it and make changes public +There appears to be a fix but its not incorporated into a Hibernate build for some reason +I dont think there are any Acegi specifics apart from the DIT name +Thanks Owen! +Thanks so Ill close this +Also I am not sure about how long you can keep a socket open +This means that if I kick off a build of X in the background and switch to do email or whatnot fop will steal focus away from my email application when the build starts to produce its documentation +Tomcat handles it well internally Jetty has a filter which does it properly +I believe its definition error in this case +v attached that refactors unhinted writes for batchlog to not go through the entire business that full send does +Ill update its javadocs to that effect +Hi Freeman any update about this bundle? I would like to prepare the bundles release and submit to vote this week +As far as I checked there are no conflicts +Im not clear its still and we have one comment that it doesnt quite work with Mahout as of months ago +Activate inplaceSelect but dont select anything +Tagging jobs will require more thought +Ah hem guys how about we apply the patch on trunk now and see how it goes? My tests showed no harm done and we are still a few months away from the release which gives us time to see how this works in practice +Moving past M martin got any update +I rerun the tests manually and they all passed +Thanks Bruno! +Thank you in advance +It compares against the creation time of itself +Committed update to jdk master to trunk with revision . +Fixed a bug with calling destroy in the wrong place in All tests pass +Good point I see what you are saying + +Example template to cause the error and related SQL script to create test tables with a foreign key +Could you provide a patch for +As it is I just want with Nicholass option and duplicated the shared classes in every module +Is this OK to commit to trunk? Could someone ? Thanks +I suppose the byte code might be a little different but the error is a +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +added the information in the information from is done. +So it would need to probably update the global capacity value for those caches to appropriate values for rowscached some other limit for keyscached +It would be better to separate interoperability tests into different target as depends +However is someone does have a method that returns NULL as soon as no more data is available then the value can be effectively unlimited because a NULL return value would cause the loop to end +Hudson would have failed this +So Im not sure I understand the commentThe patch has a new GLRER at the end that uses the modular code to mimic the old classbq +For internals type codes is overkill but its hard to know how many user types will be wanted +to trunk +META +I think that the option of Click for detail suggested by Evgeny is the best option +So I guess that means we can close this issue. +Fortunately I do not see the DB repository as a complicated piece +Fixed in r. +you for your contribution! Committed to trunk +Took some time today on this and issued the attached patchI dont know if it is correct but it seems to solve the problemWill run the tests to check for regressions +Sorry Byron +I suggest to fill issues for the classes and make them as a subtask to this issue makes tracking more easy +Thanks +hit that one in about min then swapped to running all s a patch with the debugging code Ive been using +I wouldnt say this would solve a problem that cant be solved some other way it would just be nice +Please verify on the EAP documentation on stage +Ive attached a patch to s that fixes this +Try now ensure yourself and others use umask or youll come acrross this problem again +It doesnt break the functionality +I think proxies are definitely the planLonger term we should think about whether the manageable attribute store is the right solution if were going to try to run without proxies but I think thats a largerseparate issue +Committed and will be part of release +This is an update over the previous patch using the +If were to have a single snapshot for artifacts and attachments then it will have to be generated ahead of the installdeploy process which means its a to also include in the artifacts themselves somewhere under possibly +It does not seem to work in spring +Found one wrong key also in but problem with whole wrong key set is only in. +I think this has already been confirmIt seams to work now +Just for the record I cant replicate this on uname a myhost sunu sparc SUNW triedsucceeded in putting the whole Plexus without a single duplicate what is worth +Removed dependency on the +The is you will start faster but space cleanup will be delayed +Everyone have a great weekend +post jca merge only the credentials remain as meaningful property. +I mean all tests in the suite +All I can say something craps out inside Java security code either due to unexpected or invalid input or due to an incompatibility with the Suns Kerberos implementation +But okay on UNIX +Addendum uses of actually trimmed a few tests which passed when run individually from my list posted Aug I agree the tests in list might just be flaky +We need to investigate that more +Bob can you take a look at this when you get back +Please dont give up +New queries capabalities will not imply a change in the protocol for instance +This is done so that the thread can exit in a clean state +A stack trace would be quite helpful +ssh public is this aok? +this is fixed +Review board has been updated +This fixed the issue +Generally I do not think it is wise to have code behaving one way in multi threading and another way in single threading +I feel bad I didnt supply one with the prototype patchAbout the Solr integration this looks good! We can use a similar approach for autosuggest tooso this could configure the analyzer forI wonder if we should allow separate configuration of index and query analyzers? I knowI came up with some for that for autosuggest but Im not sure about spellcheckingI guess it wouldnt be overkill to allow it though +What is not clear to me is how we stop the internal events from reappearing as external events +Will not fix. +Spring added a with validationHints that we can build on for +patch committed +Running through Hudson +At least one test continues to fail for me +This patch cleans up mutability problems in this package cloning timestamps and arrays +Its is def a critical bug for users who are entering in data or checking rows believing that state hasnt changed when they are refreshed with the next set of rows +Logged In YES useridThe previous jboss instance vm must still be running and holding the ports open +should we add an api to enable the broadcast input to be notified by the processor whether fetching is necessary +Fix attached +Thanks David +even though in that test there is still a strong reference to the class It would be nice someone from commenting on this +Patch attached to bring compactions into compliance w the others +Drained commitlog with Cassandra then start Cassandra and works +Let me take a look +its fixed +no i only committed it to trunk +Works properly in current all issues that have been in resolved state for more than one month without further comments to closed status +should beMarkup id set on a component that is usually not rendered into markup +Thanks Bill +done the demo XTS guide and trail map +Yes all tests pass +Uh +Issue may already be resolved may be invalid or may never be fixed +Thanks TedHere an additional diff for a simple testPasses locally +Removing the jira component and versions patch was only for confluence +Attaching the patchThe fix is to change the test such that it uses Java libraries to create jarziptar archives instead of relying on shell tools +Im also having this issue but I went from +so freemarker cant trap the exception +Some of the directories produce io errors when read +Committed revision +Do you have snapshots available I could play around with +Why version do you have in your env? What is the JDK version also this may help? I dont know where to look for +Corrected the docRan ant docs on my machine +Interesting one! Thanks for the patchPatch applied to PORTLIB module at repo revision rPlease check it was applied as you expected +Ming Did you attach a profiler and see it consume a lot of CPU cycles around this method +Could you look at this and form an opinion whther the fix for is the underlying causeAs we are in release mode the sooner the better +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +We have this now +OlafPlease also consider submitting a test case for this issue +This patch merely prevents unhandled exceptions popping up in log files and screen rendering error texts presented to the user +Seems to work for me +This is one instanceof of the more general bug reported in close all issues in resolved state that have not received further comments in the last got the same error when try to add a Postgis changing web browser from Google Chrome to Firefox all works think it is a web bug in you! +Hi DyreIf subversion cant handle the merge then I think your solution will work as long as people remember to add the comment block every time they create a new message +I suspect you are going to need a lot of votes to get this happening unless you provide more information +In the one release headers are being stored in a with CASEINSENSITIVEORDER that headers were being saved noticed that we were using an for parameters some how though +You can set this jira as resolved +Thanks Steve it works! +Tom do you want somebody from QA to help or you can handleassign to you +Specific field readOnly info is saved separately and thus remembered +Fixed on trunk and +Marking as and Parser +In my case Jboss version is +Hi MarekI think what the issue is describing is the fact that they dont deal with termination handlers at all +Having reviewed the patch I think we need to pass the locale to the in the constructor instead for this to work due to when the subformat is created +We have a big object model simply creating two classes with the one relationship does not reproduce this problem +Just a question have you tried with? Not to say that it has been fixed but if the issue is still present in this version we will fix it on trunk as wont be fixedthanks +Correct release version after merge +Fixes the bug still need to improve the javadocs some +Performance should have been improved since your last comment if you can tell me if things are now ok for you with latest version I could mark the issue as resolved +for +Cant really do anything about this until either Kris or I have more time to dive into the code +But it would make sense to change both of them to use joinOperationThanks for the info about the probable history behind that reference to the DECIMAL scalar function +Ive decided not to do this for +We can try to remove it and we will see if somebody will complain +Thanks Konstantin +Since this would never happen in the real world this is not something to be worried aboutMake sure you have no more than identical selectors with compression and identical selectors without compression in your test case +Committed to trunk +test failures are unrelated were seeing them on other as well +This only affects +we didnt make the change at cisco because it would have just been a bunch of busy work +Applied to SVN thanks +I applied Mikes patch with a few small changes fixing an NPE and in dfsTheres still a bug where with only a single map and a single reduce are running on each node +And the empty dataset still warrants the dimension information in the response but no extent information is provided and reasonable defaults are provided +There has to be something unusual with your as there was in +According to Remy this is a straight forward port and will be in by Component code freeze +I need these and groovy must work with them +Let me commit tomorrow evening or so +Ok Ill defer this task for a next release +I missed the sql and the operator when I first did this +and Ill change some trivial code +If youre synchronizing on the and you still see threading issues by all means reopen this! +This patch also includes fix for a bug plus much needed Javadoc in some parts of recordIO +Is this commonPrefix wi the row only? Or is it row? It seems its row? Its not clear from the javadoc nor as I walk throughSorry for the confusion +iishost is actually the Web Site Description as seen in the IIS ll attach a new patch when its ready +Closing. +Thanks David! Rev fixes our databases exceptions. +simply take from the distribution replace s version with the old file rebuild the jars and see what happens when you use them with GrailsWe did for example change s inner and class to static simply because they are not expecting to operate on the surrounding objects state anymore +Did anything change since the last timeRegards this problem still exist and it would be good to have access to minerva and vesta to check test failures from time to time +I consider this to be a bug since it breaks Arquillians mechanism +Heres a patch for this +Marking as resolved +committed first iteration removed from min and demo targets following targets now supported in J main builddemominuiSVN commit configurations tested pass with last commit +Source streams first file with header containing info of all the other files +Ive tested this modification with Tomcat and Jetty and in both works properly +If you change your mind about this I hope you will feel free to ask me to help figure out whats going wrong for you +The update patch has been applied +Attaching the svn stat and svn diff files for the master ManjulaThis looks good to me except for the output +Thanks for looking into thisSijieI have ZK servers and configured likezkServersSurendra +out that there were issues in itself as well as inIve rephrased those two tests to expect equivalence between standard and there so if theyre running on a recent JDK build the tests are going to insist on the write methods being discovered in both cases nowAs for itself Ive relaxed the checks between read and write methods a bit accepting superclasssubclass relationships in both directions thereI intend to backport these changes to as well so that the behavior with running on JDK is equivalentJuergen +A new type of math transform is required for this one +I consider this fixed with since no more exception occurs +delivered but only as experimental in JBT at the moment. +This seems very high +Canceling the patch as it is over a year old and there are review comments to be addressed. +Thanks +Hi Aidan can you review this change for me please +For templateFilterfeatured or community we only list those templates which is in ready state that is in templatehostref table the template should show as DOWNLOADED +I think is not going to the API until +Yeah but weve already given up on i since that turns out to be a mess. +Also add some Apache license headers +Looks like an Infinispanclustering issue not a class loading issue +Fixed the validation errors. +The REST clients Server object was changed to know about the validated and original and whether it was validated +Fixed with svn revision applied in some problems here needs more work +Should be fixed in commit dbebccffdffebedacca please also needs specs to test for this so you could turn your test code into a spec for that +HTMLMESSAGE means that the message was in HTML which is often used by spammers and malwareTry sending in plain text instead +gc threads cause a crash +IO is currently compatible with the last release this can be considered ifwhen we decide to break backward compatibility +In current code base this property is set to false at That is already the case since r +That really solves my problem +Below is his justification +I am fine with committing the patch and openning a seprate JIRA for the remaining issues +Ive created +Every other transformer dealt with fixed named elements onlyIf we close this bug now it may come back to bite us when a future transformer is developed that handles configurable elementsI delve a bit deeper into the code base and see if I can find the real culpritDean +This change didnt do what it was supposed toTo be clear it did fix the reported problem of rows not being returned when they did exist +probably it was working for me because had common component built in back to testscommon references are removed +No it is not we consistently used the methods to account for this strangeness. +and the string version +Verified by Alexander. +Defer all open issues to +ThiloNeed your wsdl +Actually the solution duplicates shut down database cleanly before copying to the slaveClearly it fails in another way here +Reviewed by Robbie and Rob as indicated in the comments above +patches needed to get to work +closing the issue again. +I decided on this because when failoversucceeds the master database is shutdown at this juncture there might be number of clients connectedwhich need to be informed of this shutdown +The good point is that Enrique is going to commit all its SASL code as soon as will be out and this is also one of the reason we want to release this to be able to inject more features like SASL in this tagged versionsomeone can add or update a page on confluence where all these passwords issues can ba discussed +We do need to check for though +you may get inspired by how things are computed for the timers of computation to have an idea +Hi KojiI mean it only happened in certain combinations of the characters in filename +By the way do we have some docs documenting the vault functionality to a greater extent +thx +Namely that the connection should be using the real user of the current ugiA jira on fuse authentication brought to light that we cant assume the login user is the same as the real user +We cannot fix the case where imported dependencies lose the repository definition they need from the imported POM into the new one +How is this related toDoes it mean that Before annotated method can be used if whole test is marked as +When a remote connection is interruted the connection fails the bridge destroys its local connection to the local broker to avoid producing a slow consumer situationIt seems that the wrong connection id was referred to when the remove command was sent localy +Hiafter spending a lot of work AXIS C is working with BEA weblogc server +This is more problematic for resources like lucene +I recommend a strategic upgrade to that Spring generation in any case even from it should be a straightforward upgrade without any issuesJuergen +done +The Jangaroo team is currently working on a Maven plugin +patch looks goodStrictly speaking read should be read +Project once running can be accessed athttplocalhost have tried this in Wicket and it is still broken +The source lookup plugin requires all jars +A test case would help us validate the problem +r +Sorry for the noise earlier +Patch fixing all mapred findbugs test and test patch passed with this findbugs warnings reported for the previous patch did not show up in my local box where findbugs version is +Ive moved to so +if youre still around can you please answer Ians question? I have a similar problem figuring out how exactly I can provide the username and passwordThank you +Shame on me I had forgotten the import in mySorry for that. +Willem thanks for the clarification +that changes message parsing as described for client and server connections +Cheyenne could you please look at +Kannan That is pretty +MartinHave you figured out that problem with the local modeIs it still broken +We did make some changes last month in the way requests from are being proxied from the fronting web server to Nexus +Website up. +Can you take a quick look and see what you think about this and if you agree that the runtime should be raising an error can you create a JIRA +He proposed that the Aware interfaces might be deprecated in favor of passing the to Actions directly +Just committed to. +Dan there is also this ticket. +I think when this is fixed we can mark and as dupsJoe do you agree? Or do you think there is still merit in fixing the specific issues mentioned there +This has been fixed I resolved this issue but JIRA was bonkers this morning. +Candidate not committed for reopening for release note status. +Only one line but it ugly! cannot keep this in sync right +Nobody understand Danish anyway + +In trunk site is broke so cant really see this new doc +This is the one used by +but is the addin generating the correct groupID for the one specified in theIt should really validate the group ID and do something else if there is no company set +Retracting for now +That works OK in trunk but doesnt work in Studio I could reproduce this bug for new bug result of moving on eclipse and wtp +Then I am not certain of the workflow but make sure Shelly is aware of the info +This way continues to work exactly the same way for both and the other hand I had to make column index attribute not final on Declaration class in a way the can change the index when it adds the commited and test cases all green. +Fixed in rev +Im running out of time again +Updated the patch with missed files +Here is my latest patch that does the following Adds integration tests for all of the transaction strategies Modifies the behaviour of declarative transactions with respect to exceptions to be more in line with EJB Adds support for multiple transaction elements in a blueprint a newer version +Anyway Im back and can start looking at this now +Thanks for the reviews +will this suffice +Simple example project to reproduce the of the Eclipse console after reloading the pom of the submitted test project +Fixed +port file similar to a +Sorry but as I mentioned I now have a static IP and not a dynamic one +Vishal +JacopoUpdated patch is attached +Verified in ER +Right now we underestimate coverage +if it could break an app then surely this breaks the backward compatibility requirement for + +Should it carrying all of the language bindings or should it solely just be for packaging the C client and server code? +Next s the state of this +sucks that eclipse variable replacement doesnt have a context to use. +Fixed for the trunk and branch +Im attaching my proposed html and patch to produce do you have patches to your proposed pages? Please attach them so the maven dev team can choose what they want +Its apparently fixed in their trunk but they dont have a release available yet with the fix in it +The v patch doesnt apply cleanly on a fresh checkout of nnarkhed patch p i Projects patching file coresrcmainscalakafkaconsumerpatching file coresrcmainscalakafkautilspatching file coresrcmainscalakafkaserverpatching file coresrcmainscalakafkaserverHunk FAILED at +Unit tests pass on +Is this checked into the Spring trunk? If so Ill be happy to build and test this fix +I plan to commit in a day or two +Thanks +Wido can you check on this m picking this one up +Since Hive need the new methods to work with the interface should be extended +It seems there is some more that ismissing +Verified in ER +If I move the WSDL document to a path without spaces and change the targetDirectory to a path without spaces it will work perfectly +The Q shows it achieves the same performance as explicitly declaring double fields as double +Yes as indicated I needed to update the table tabbedForm and control css files as well as the and velocity macrosIt works fine now this was an problem thanks +Remember that the point of this is to support use cases +No patch yet Ill get to it on the weekend +It looks goodIt seems that we should separate test utils stuff into their own module +Great +I just tried today and it still points server library url to the serverdefaultlib instead of targetcargoconfigurationsjbossxlib like it should +It should have read this wayIMO should represent all columns and if a view creator does not want to inherit changes to base table schema then she can specify the exact columns instead of +QA bot picks the newest attachment even if the file is a pdf +Default it to false which would be how hive works now +Im using the debian installer +I checked it on Windows bit platform +When you detect EAP through runtime detection database connection using hsqldb from EAP is set up for you so when you create a Seam project with the wizard you dont have to create the database connection profile so its easierIf you add your server manually and seam runtime also manually then you have to create the connection profile manuallySo I assume this has to be added somehow during runtime detectionSo I thought we could do the same for EAP +You should check that code for any new properties to try to maintain consistencyIm think that SSL encrypting of the traffic between logj and a SMTP server is handled by and is transparent to logj but I could be wrongUse of certificates to digitally sign log emails seems wrong since it would be affixing someones identity to a email message without their involvement +gz files int eh patch +But the problem still here +Moving to to to LATER this will be nice to add in the next back to CR for now so we can reassess +Applied patch +This is the filter Im using +Thats what I get for not reading the patch +Punting from since its not fatal and some info is missing +Attaching found in +Hi Tyler do you happen to have a table dump that we can use to reproduce your problem +i am unble to open this attachment using either or windows native compressed folders support +So deprecating it in the series certainly makes senseHowever this would require major updates of the user guide which discusses in detail +Timeline Motion chart etc +untested patch to add json import to jaxrs yes todays works +So just to be clear big from me +can you attach a patch +Running full regression tests nowThis patch is just code cleanup but I plan to backport to as a test of post reformat merges. +Does your code ensure that the salary insert thread isexecutingcommitting before the employee insert thread? How are youcontrolling the execution order of the threads? Does the system time showthe salary commit occuring prior to the employee commitA common issue that arises with multiple threads is that if there is no code to synchronize them then there is no guarantee as to the order of execution of the code in them and you often seen the type of results you are noticing +bulk defer of unresolved bugs to +Which iBATIS version do you used +Is there still interest in thisThere is no work on it for months now +Why thatBottom A is accepted the left to do is not subtracted correctly but instead is added. + +well it seems different issuesconnection properties disappeared because we have missed these TAG write back code +It doesnt +I have no plans to work on this +Reopening for pullup to +was resolved but Im still seeing issues +Mistake +unfortunately in reality it is not so simple +So what I think now is that its not that hard to implement +Ok bumping to for now. +Thanks VlastislavIve fixed most those issues in the version of the Installation GuideI wasnt able to completely understand what you were asking me to add from the linked comment in EAP discussion +I exactly understand +Patch applied Andrew can you review please +Another reason for using a recent version is that extending test runners seems simpler since Junit +Why hurry now +If a user has explicitly set a checker it should be an error if the code cannot run that checker +Patch accept multiple arguments for s it for the micropatches +Attached for this issue +Hi JoergCould you clarify what you mean by this? Are you saying that your overlay is set with a packaging type of jar and an additional webapp artifiact is attached that contains the web content? How does the war that wants to use the overlay specify dependencies? Does it need to declare two dependencies for each overlay one for the jar and one for the webapp typeIs there any documentation or are there examples somewhere of thisThanksAlex +MINA checks session idleness every seconds and fires sessionIdle event if if the session is considered idle and if sessionIdle event is not fired yetSo if you didnt writeread any message since sessionIdle event is fired sessionIdle event is not fired again +v patch with tests that check the request for a certain representation returns a result with the correct content type +I think the larger issue here is that in this case theres a thread spun up for Swing which we dont shut down +Id like to know the idea from s original authorThanks between and series Exception type has changed from to as well as exception message has changed +Closed for good +Committed at revisions +Next patch set is against trunk +Thanks Andrew! +cccc +lgtm applying patch +I truly appreciate your help +Should this really be closed? The problem appears to be specific to IE +No thats fine +This lowers the risk of endless errors in the error log +NOTE I did change CONFDIR to be etchiveconf I understand the need for flexibility but lets have it as a separate patch. +icons are shown in the has a wrong icon in the CA proposals list +method overloading +Longest times inside in the are creation of new Writable instance doing readFields +Alexander would you mind creating a patch file instead of diffs +Attaching a release note about the new reserved keyword. +I see there has been a response. +Is this a known reason for using a repeater instead of a loop? Perhaps this is just user error +I tried on +This could be avoided by adding a timestamp component to local job ids? So the jobid would be something like joblocal instead of joblocal +forge directoryforge still has no shell and doing forge i core or forge i shell hangs is now resolved +has been marked as a duplicate of this bug +Thanks Navis +Tested +Fixed by +Maybe you can then also write the configured default in the database +Will clean it up later +Yes marking as fixed. +thanks patch to trunk with revision +Enjoy. +Also denotes the inconsistency filed in bug +I have observed this behavior with Spring but I dont have permissions to change the affects attribute +There is information in the comments on that issue +By the way this problem does not exist with the native launcher +Same for on also no event thrown in normal modus only in debug version the on is thrown in the client + +Patch addressing comment and fixing javadoc error +The problem I had was that I didnt know how to supply a parameter to a parameter and my intention was to make the wordnet synonym provider deployment as easy as possibleHow about this add a new parameter synonym Extend the interface with an init method that passes in the options string +Which version of windows was this test performed on +Also the more jars we include the more I think that we need to be offering a single jar containing all of our dependencies and building that jar based on do not go down that route +Perhaps when doing this work we will discover that there are other cases where it makes sense to remove some of the laziness of for the sake of scalability +Or if the reason youre specifying programatically is to verify that it can be done and is adhered to then you should create an alternate conf dir in trunktest that can be specified in place of conf +Can you please test and report back if it did work +This patch includes a few test cases for testing generics support +XdumpjavaeventsgpfuseraborttraceassertlabelCslavedumpsheapdump +Havent had time to look at this been on holidays code Vs Santorini no contest +This JIRA issue has a subject which indicates a general EAP problem and yet the examples presented only involve property substitution +my bad I havent had my required intake of chemicals to think clearly but I dont think this going to be completely doable directly in +If I change line of bincouchdb to backgroundstartargumentsnn works in background mode too +Closed upon release of. +fine with medo you want to commit it or should iand would it make sense to do the same for thewhat do yo think +But the rewrite of the should support this featureTherefure I created the issue and marked this one and other related issues to be superceded by the new issue +I made some changes +Im leaving for weeks holiday +Will submit a patch a little later +I suggest just moving it to its own +What about generating the license headers in the XSL scripts instead? Something along the lines of the attached patch +It looks like the root cause of this eclipse bug might predate native threads and just be being exposed more regularly with the increased levels of concurrency +I need the status of this issue confirmed ASAP if I am to meet the documentation deadline +Will try to get back on this tomorrowKris +fixed typo in JIRA title +Thanks Doug + +The test harness needs to be updated along with this change +Phew! I was worried when my build suddenly stopped +Thank you for your work to put the guide onlineAbout versioning I think that it would be better to create branchestags for different version as I also dont think that a migration guide should be part of the user guide +For some reason the schema generation plugin is not properly generating the objectClass for TOP +Legacy guide Web Development perspective guide has is guide is checked +This patch also makes sure that the does not read too far into the next split when its not necessary +Thanks that helped a lotActually neither the Directory Studio nor the Java GSSAPI is broken +findbugs warnings +Both and have subclasses that implements these methods +This patch depends onIm assuming that the permissive license used by bouncycastle is acceptible for use with the manual download target but if not then it should be removed +Running ant from any of the subproject directories is not supported +Attaching patch for trunkPlease review and provide your comments +sure it wasnt a regression due to but may be a regression in due to various or changes +Fixed in git closing stale resolved issues +Ive compared this bug to +That doesnt mean it is better +This capture shows ten consumers competing to evaluate an case with bigger XML old resolved issues +bulk close of resolved issues. +Thank you Alex! +Unparented this jira from since this feature is independent +i do not care either way but we should not be using EMPTYSET in various tokenstreams without this patch else we unnecessarily hash every input token for no good reason +Fixed coding style +We dont have converter for Generalized Time +Hi Vinay +I have created as a issue to add the missing commands +Then process classloader is used when processing jobs +Fixed in maintenance and +if foo is null an exception is thrown if foo is null then render nothingAnyway the prior behavior can always be achieved by adding an insertReference handler that handles nulls. + will have to be resolved first +We could just have minimal information in thos document and point it to the main document for info so info just like this dfsadmin go to the link for dfsadmin dfs commands go to link for dfsdommandthat we do not have to update twice in this document and the main one +IMO it will more depend on the release date +We dont support at this point +One other thought AndrewYou might want to use instead of adding your own version to +email thread discussing the issue +works for me +Marcus SanjeevCould you test this issue with the latest codesI tested on my environment it is ok nowI removed one host from the advanced zone and create a zone +The test now runs cleanly on both and in my environment +Jules suggested s shutdown handler was being invoked +Committed revision for method access +Are you behind proxy firewall or defined some private repositories in +We have been looking at the use of Factory classes to allow you to specify this kind of the connection pool is a great example of this kind of work note the uDig project has also resorted to the same solution you have presented more then a few was wondering if you knew much about JEE Application Servers and how they manage connections +It looks like Carsten already applied these patches in Rev +This is pretty much the realization of my previous comment +test for and that doesnt work because of this looked at the test and found that at the time the issue occurs we haveconnectionAttributes set to shutdowntrueshutdownDatabase shutdownandcreateDatabase createIf I clear connectionAttributes and shutdown then the database will create properly and the test will get further but ultimately fails with an exception that I havent investigated +Need to do testing and verification of results +Lots of people on this one last week. +Looks like a missing set on our side +I looked through the source you posted and nothing jumped out as wrong +Hi Alex +Removed invalid Java assert which assumed there must or objects +I committed the patch +I think that this is actually ideal bc getting rid of the Foreach and directly doing the insertion should be a nice speed boost +Ill be glad to investigate further if I get some instructions +I dont know which layer is throwing this error or where the actual error is getting lost +bulk defer of featurestaskswishes from to +now fixed +Ok Ive made fixes to all this on trunk and the original case passes as well as the other failures +This is experimental in trunkWe will mark this as experimental as well +bq +Initial patch +Should have checked the ticket upfront. +Looks like a simple and good idea +Supposed to incorporate fix for but seems like there was not any fix for in fact since I was told that they only changed docu for +Chapter +Credentials can be now edited in the dialog +Also I think this method could do with some javadocs cleanupfrom the javadocs it is practically begging you to call it if youever delete but doesnt already handle this well +I couldnt find the old commit but I remade it +Im working on it in fact I think I have it sorted now +Additional Notes For Mojarra be aware that is the same as but the mojarra team did not update in maven correctly +repo components setup gets difficult with this new scheduler in placeThe class should not need any knowledge of the scheduler +Thank you! +Linking in the jira where these were added back. +rpmnewThanks the issue based on above Abhinav for your workReopening the issue as the patchreview is still pending on master branchesSome clarificationWhile moving content from the to refers to copying configuration fields for exampleworkerscloudbrportresourcepodzoneguidcloudbrclusterscriptsnetworkdomrkvmhost +Please apply patch for the exclude list applied at r. +Policy intersection is supported by the Apache Policy library currently used in our implementation +Thank you David for the patch +If some error occurs in the transaction none of the records in the reduce should be inserted otherwise when the reduce is restarted some of the records might be duplicated +If somebody can test on other platforms and confirm the accuracy it would be a great help +Please try latest CVSthanks to be fine nowThanksYves +Thanks Bhallamudi +Hi Ben yes I am moving SASL processing out of +Can you pls check on this +same for metaLoads and blockLoads +I believe will fix it on the embedded side +Done just to do on another +Ill supply a patch shortlyMeanwhile is it possible to make those data members protected in anyway? Noticed as of now its not in the blocker list +In a little more detail if we separate the state from the atomic broadcast and return no zxid then the service layer on top of zab will have no such a notion of transaction identifiers +I committed some wording changes please review +But Im not on Windows +ok Reusing deleted template name +meaning it is not an infinite loop but just a long list of objects you are loading +I use Tools with Eclipse Galileo and with Eclipse Ganymede +NET platformBy the way the project file you are using have errors +WicketPlease attach a quickstart to this ticket that shows the problem without any modifications required by me +Committed at subversion revision +This appears to have regressed +Please review the patch added for this url issue +However this will cause much performance degrade even though CSV files are already well formatted +Fixed to dispatch the event type NNODEWASBLACKLISTED in and avoid preserving nodeIds in blacklist duplicatedly +I dont know whether it is default setting +Bart please do a quick search in jira before opening new tickets this one duplicates transition all resolved issue that did not see any further comment in the last month to closed status +Im happy to try and help with the solution +That situation is something that is both hard to diagnose and potentially expensive to recover from +The should be updated to catch such Broken pipe exceptions. +Just tried that and it makes trouble only in ChromeIn Firefox it worksAnyway you can also another way how to highligh text in chrome see workaround description +Thanks! +The command has been modified too to use the plugin architecture and that should be the last thing I thinkThanksJulien +Its a serious problem with meclipse since we could use a correct artifact or exclude a bad in project dependencies but are stuck when the artifact is defined in a maven plugin ie jaxws plugin Path to dependency jar jar jar jar activationactivationjar +You should remove or fix the class end comment in thoughAnd the author tag for both interface and implementationcheers Roland +Patch issue in not valid anymore +Assigning tickets to Toni as he handles BRMS issues AFAIK +Waiting for focus has been reworkedYet it hangs from time to time +If you feel like this should be in +There are a number of efforts that will require changes to the core container during the Spring timeline and its conceivable that we could do something like this +if support JSONXML how will you wrap up the logs in there? Is it worth +This patch assumes H is started in standalone mode on default port +configuring +Please update the tests with these images +Thanks for the patchFixed on and trunk +It is no bug but a library version change. +I believe we need just skip this error and let code callig load to try superclass +Draft release note text statesThe user had to manually select the login screens username field before typing his or her name +is no longer raisedI close this issueThanks Emmanuel. +Not sure what I did wrong the first time +A patch for review +The leader will use txnlog if it only needs to load txnlog fileI havent done extensive testing to figure out the optimal setting +Thanks for implementation for OS abstraction layer was implemented. +Corrected patch +Wouldnt be better +This is a long term task +Please verify it +Can you please tell us which testcases? I think we can solve this with some intelligent locking +another one for will be coming do we have patch for the same issue for both EAP and EAP ? I think we should push to use latest EAP and so only one patch should be processed the one for EAP +Move it back in if you think differently +batch transition resolvedfixed bugs to closedfixed +Change back the names of the three new methods in to be compatible with the current trunk +What Im thinking it is now is that the converters have already been defined by the JSF RI and that its redefining them from the provided by and hence the warnings +moving to doc version for better separation +Vladimir the null View might be caused by FLUSH can you take a look +In Rev +This is the problem +I remember seeing a recent PR against upstream which was merged +To run the version that I have attached just makes sure your BEEHIVEHOME is set to the root of your Beehive distribution +Can you try to reproduce it again? Im curious whats causing this +No need for release note +Thanks for your responseCould you please provide me information on Scanner functionality ishaving Blackberry or notShare me links and your valuable knowledge on Playbook creatingExtensions using Cordova API +Never mind I see youve opened committed this +older issues for Apache since were no longer actively working on these at the moment. +Thats what prompted this ticket +It is trying to modify maven +resulted in these fields being made protected. +From this information we will write the final release notes for all issues + +Hi LewisWe cast to the first because our defines more methods than from the api which we need internally +In which case would other components add additional parameters +This just catches the from generator +of the compares are exclusiveIt seems a safe action made by create or create wouldnt exist really so the equal condition couldnt be I will attach a new patch soon +As I wrote some time ago these modifications make two enhacements in the functionality Enable UUID parameter in addNode this allow to preserve the unique node identifier across a repository migration +removed +I have not made corresponding changes to CSV writer and will submit a patch in that regards soon to +I mistakenly requested to be applied against the +Attached new patch which deprecates get in and replaces it with a more intuitive buildFilter +Uploading again +Patch applied +Couple of minor points still has a few references to timestamp instead of clock +I havent found any time yet but hope to shortly here +What are your thoughts on setting the Exchange Header? Should it be a collection of Source or maybe a single Source from the header level? Ive implemented the latterIve modified the transformation to use a static +Maybe you patched the patch +Fixed on and closing all issues that have been in resolved state for more than one month without further comments +I also added a fix to avoid a NPE for the coords attribute as it can be null as proposed by Craig StiresBoth was added in revision Thanks for the contribution! +Updated patch addressed comments in review and on jira +I have seen the source code and this messages seem to be hardcoded in are +I think thats all you need +Changed the spec +This patch fixes a bunch of issues when you try to actually use a different bootstrap partition jar +Thanks Rohini +After a preliminary look into that it seams to me that this happens in for key generators generating keys before insert statement +In other words a TODO +Committed revision +See for a description of the underlying issue thats causing this +Patch applied thamks +By default its set to false for the moment Ill wait to get more feedback about it to see if it should be activated by default or notThanks for your contributionXavier +Nope that does not work either +should be in on EAP ER +i believe the disputed item at the time was behavior vs behaviour +Attached Example WSDL to which you can try create a CXF client from using JBDS +I can reproduce it on trunk +I am working on a patch for this much in the lines of the Solr benchmark patch inCurrently the direction is Add to Taxonomy Directory Taxonomy Writer Taxonomy Reader Add tasks for manipulating facets and taxonomies createopencommitclose Taxonomy Index openclose Taxonomy Reader facets will also build the categories into the document will bring back categories to be added to current doc will be extended to also support faceted search +You seem to have tested with an older version +Filtering datatable without the possibility for selecting a row does not make sense +I made the change to check we were in the keyspace we were deleting before setting it to null +The thought was that if any user just wanted to bring up falcon and play around with it he shouldnt bother too much about the hadoop dependencies +Work in progress +I tested them all out in in head and everything looks great +Why not just remove the surrounding trycatch and let the exceptions be thrown? Keeping the log messages doesnt seem to be all that important and the caller should define some catch block which does something more meaningful than just a log message +has this been done or do we have an extension that does not work in the downloads +Sande You can drop your docs information here and assign this back to Rebecca +Been trying to reproduce this but unsuccessfully so far +Currently I maintain a path to it mappingbut this is not en elegant solution and does not survives restarts +patch attached to update relative URL specified in I will also upload the document for reference +Should probably be in Atlassians JIRA not ours +Cannot recreate the error +Closing as requested. +Im guessing he means one run of the clustering algorithm not one iteration of the algorithm but Ill let him say for to use existing Lucene Analyzers and document steps to run Mahout clustering job in an EC cluster using the hadoopsrccontribec scriptscontains clusters with top terms +when we can find some time we will have a look at it +The issue doesnt appear any more +Hopefully this fixes hudson. +May be we should load index metadata in single thread rather than asynchronously +I have added a new option to generate only the top element data binding code Ewdc +Fix time for sure +Chris it looks like the MSSQL driver is reporting text columns as LONGVARCHAR and will treat LONGVARCHAR the same as other varchars which SQL server doesnt like +We cantshouldnt put any random file there +I included the jar with the release so it should deploy properly in Tomcat and Jetty and fix the java did you include it? I just downloaded the +david could you have a look at this +Committed patch with revision +Schedule and assign to Ales we need this to pass the CDI TCK +I have tried this issue with set load set to ALL as well as set to DEFAULT which is default in RF examples in this version of PBR but seems this doesnt impact this issue +Merged the development branch back to trunk in revision +Now that work is underway with AS all previous community releases are +Adding phase design document to solicit feedback +If it is also missing component please set both so that the issue can be properly triaged +Fixed by removing generic info from types of method close of resolved issues. +period +merged to branch +Hi I shall look into this problem today as well +The adds the start of a webapp with a JSP that uses the SCA taglib to have an SCA reference to a helloworld service that will then be wired to the service in the contribution +The server starts with address httplocalhostsomeservice and the client is trying to connect to that same address +This was a very dumb oversight by myself from moving too fast at end of day +What would be the difference between Spring Batch Commons and Spring Batch Infrastructure? Very little +Do we have do anything specific to make the broker available in these scenarios +The previous one wasnt all there +Emergency to yahoo hadoop distribution Ill upload one for trunk shortly +Thanks +Im happy to learn about any news there Spring will be supporting these features as they emerge +fixed to parse ids with such that groupId only gets the part UP TO the artifactId gets the whole thing with Did this both for dependency declarations as well as the poms idgroupIdartifactId +Test file +See +this is awesome stuffSending applicationsAdding applicationssrcjavaAdding applicationssrcjavaorgAdding applicationssrcjavaorgapacheAdding applicationssrcjavaorgapachegeronimoAdding applicationssrcjavaorgapachegeronimoconsoleAdding applicationssrcjavaorgapachegeronimoconsoleservletAdding applicationssrcjavaorgapachegeronimoconsoleservletSending applicationssrcwebappSending applicationssrcjavaorgapachegeronimoconsoledatabasemanagerwizardAdding applicationssrcjavaorgapachegeronimoconsoledatabasemanagerwizardAdding applicationssrcjavaorgapachegeronimoconsoledatabasemanagerwizardSending applicationssrcjavaorgapachegeronimoconsoledatabasemanagerwizardAdding applicationssrcjavaorgapachegeronimoconsoleservletAdding applicationssrcjavaorgapachegeronimoconsoleservletSending applicationssrcwebappSending applicationssrcwebappviewdbwizardSending applicationssrcwebappviewinfomanagerSending applicationssrcwebappSending configssrcplanSending configssrcplanTransmitting file dataCommitted revision . +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Attaching screenshot showing the works properly. +I must be missing somethingPerhaps you can provide a breif example for the docs that illustrate how you are using this. +Looks like this was done in Can it be marked as resolved +Will update my patch soon with the new test case +Please go back and read the ticket again +Attaching +Afaik we have good support for geometryless tables in most datastores and in the WFS protocol code +Odd +Kris can you look at this +problem still present in EAP +I will submit the test case later +Please take a look and tell me what you think +Hmmm +Patch doesnt apply +Werner Gregory and myself discussed to change and Persistent interfaces at London meeting and desided to use also there +Not a clue +A single template RC file is added +Assigning to +index in my repository root restart Archiva and run the indexer +verified on +I have the same problema but with Final +Moved some filter to have more coverage +Thanks +Jarek can you check +I think this is OK as a temporary fixMy guess is that we can change the routine to remove either the REQUESTHANDLER or the from request attributes before as well +The first comment on this JIRA statesI am trying to reduce the complexity of the client libraryMy question is wont designing a new protocol introduce more complexity risk and support overhead than using which has now had a reasonable degree of exposure +Have to revert the method return value check to handle the service if the filter is matched and to return immediately if the filter does NOT matchFixed in Rev +resources instead off +All calls to addColumn patched up +Perhaps we need to wait for that to materialize +Logged In YES useridFixed in CVS +Move the ones back that you actually have plans to do +Thanks Joel and Thanks Shikhar +Eric I think offering a macroscopic view of the trace data would be very useful in addition to the current microscopic view +Thanks Mark! +This issue was rejected in as it does not fit with the best practices for continuous integration +Hi RussellPlease incorporate the changes Marek Schmidt made to the release notes text field in the Release Notes documentCheers issues Release Notes entry has been added to the list of Fixed issues in the Seam section +Dont see this if running on oracle jvm +the current class should be renamed Do you mean to rename or? The current patch follows the design of. +The javadoc tool did not generate any warning messages +exhchange the template descriptors in the folder descriptors by new ones +using jconsole +I have it on my weekend TODO list so it will probably be ready on Monday +Committed in revision support for transaction isolation level url property +Ive seen early indications of major throughput and latency improvements with the version of this patch ported to CDHB with no indications of instability +Therefore it says that the environment is but I am actually using in my projects dependencies and installation packages +Put file +Committed at subversion revision +Closing all issues +That issue has been filed and fixed as so I am closing this bug Invalid. +I take a look at it +They simply treat thedrawings like an OCRd document and only extract the visible text +if yes WIIN! if not looose +The important thing for me is avoid changes that could cause problems and incompatibilities later +I have no idea why this is happening +The issue does not occur when I use JDK only with JDK +Esteban as a matter of fact they are JCR related thus the nameMarcel I was following the pattern in anything wrong with that +Good morning guys here is the first patch for today i removed the jQuery Class Plugin +I can follow your arguments but I do not intend to change the code for that reason. +It might be easiest to implement this as two buffers one and +Updated linkslabels to Indigo and Helios +As of r I moved to its own profile a full gt build will include it only if the temporal profile is added to the list of active profiles +included synapse config level changes to accomadate tempalates test cases for Synapse Observer API changePlease review and commit +I have a fix coming up +ThiloPlease try latest nightly and let us know if this class loading problem still exists then i can reopen the bugthanksdims +that takes care of some inner classes that should be static inner +Problems with mirror resolved thanks. +Perhaps I wasnt very clear the default sort column doesnt let me choose to sort ascending or descending and have that persisted all the time +blocks this +Adding patch with code and test fixes +in other words if one kills agent process by Control C it doesnt stop agents properly because lock objects are leftso if so one must call command +Moving to will not be a moving to for to Emanuel this will be fixed once we upgrade to a new version of the Microcontainer +fixed verify with JBT. +We could imagine to launch it once a day by default +Maybe not for tiny objects but possibly for bigger onesI also worry it would be hard to pick a single size to round all the blocks to because hbase allows configurable block size and encoding per table +This can be added to later releases so if it slips it is not a major problem +James Ive just noticed that launcher job will also get the same name as the real job this may be confusing for users and they will not know which one is their job and which one is the launcherIm thinking that should only set the job name of the real job the action +An updated test case shows for items are not evicted +Havent seen it since recent fixes +Add annotation on warning messages generated by javadoc tool and release audit are unrelated with this patch +lib +Changed my mind on checking site in under trunk +You might check and see if you can use code from Apache Harmony instead +See +Added panel Select JRE folder Select the JRE pathRed Hat Developer Studio needs a JRE + +But no gettersetter for inline +I am experiencing the same problem +This patch clears up the findbugs issue +looks like this is a bug in the documentation +JieshanIs it ok if i address these comments on commitIf not i can prepare another patch from home +Dave +Looks good +the reviewer +Ive attached a new patch to which seems to work +jbpm issues that have been slipping last few versions targeting John please correct if wrong +fixed except for stylecop where the reuse report is never for +Attaching a new patch that makes those implementation isnt beautiful as I commented on so thanks for taking this upA general suggestion Please add a little more information to the title than it isPlease fix the titleI dont see a point of multiple comma separated states the default result when no state is passed should return all active nodes +That patch addresses issues and +A patch on top of the latest codec branch that uses s support for finding zlib and uses to find lzmaOnce we get the patch file for the codec branch uploaded and committed Ill commit this on top of it before closing the ticket +Attaching the source +It is now resolved and merged to adding the fix version for the next release. +pushed +Merging cores is the part which is left +One question if a handler gets a timeout are you then still using a new thread to continue the sync dispatch of the event to the other handlers or are you just waiting for the handler to finally return and then blacklist the handler +The validator didnt resolve this property because it is not validator and CA must resolve private and protected members +lgtm +Verified by Richard +From our perspective patch seems fine since this particular change should not affect our users +Now registration is done Http server is started and the correct port is reported to +In the thresholds quality profile list in the column export can you add a link to get the findbugs configuration file as checkstyle or pmd +original task +This should solve the problem +Reviewed the patch and LGTM +Another workaround is to create another directory in SVN and add SVN externals to your flat modules +When one category is not included in the response RCI is assumed +Are there any more steps to replicate here? Which version are you testing on +Hope this provides some background on why we used an + Adds getterssetters for Dynamic mbean stuff with scheduling patterned off of changing the cache save period setcachekeystosave nodetool is against +I have a patch almost ready for the test + Fix +That just seems to be too chatty +Patch applied in trunk r +This seems to be a bit of a security red flag +INFO Running for rows took ms +The component works also fine to me +Thanks Karl AdrianTo be revisited +Ryan says this is a duplicate of is the snip of the hadoop dfs lsr hdfsapphbase qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongenderr qwapi supergroup hbasetransactionimpressiongender +Thanks +it is possible today to have a custom audit logger inside the without my patchNow I see your earlier point +The advantage though is that we wont have to mark up our code and we can make it to support +Some exclusions should be added to the pom to prevent bringing in transitive dependencies +I dont see anymore failures +All the fixes are committed +updated to a patched hibernate version. +I just resolved them and updated the patch again which is same as Ivan patch resolving conflictsAlso verfied basic flows with BK clusters +He replaces on But this class is absent in all dali it works ok +Just committed a fix in the trunk for this +Attached which contains logsdb from a failure in +Closing again as InvalidThere is no problem on Wicket side. +Committed. +Thanks for the comment +Havent uploaded formally into the JIRA since I havent actually test the code yet +This will still throw exceptions until we fully deprecate it in six months but we can ignore it for now since it shouldnt affect callbacks from Native to JS unless you go out of your way to use a broken mode +Not going to reopen this issue but the FIXMEs are still in the code and need to be done properlyThat work will be tracked in. +Same patch with fixes to forrest documentation +Applied the patch to Axis branch +Im happy to work on this or merge a patch +I too want to ensure we definitively find the root issue but that should be another jira andor offline investigation +This is marked resolved in the upstream project +tests included +What is the point of having the Hibernate Entitymanager product if hibernate specific annotations is the only way to use it +Its not clear what the cause was for this specific issue though I suspect it was related to a lack of explicit on those show in the top half. +After analyzing the management server log attached latestI found the real issue you encounter here is this you register a ISO or template to a secondary storage will create an entry in templatezoneref for this templateiso for each zone currently in the cloudstack system +We should also remove buffering in for the same reasonBuffering has changed as lot with and +FWIW I have strong concerns about the performance of Shale Tiger when used with large applications +We do however think it is reasonable to run two GB on a node +is a virus scanner runs on a Windows Server where you need those kind of things. +Fix was checked in +Ive fixed the problem but the issue still exists in CDK as I had to add attributes already defined in the parent class and annotated with Attribute +I am in the process of supporting the same processes and code in the Cargo project as well +One last comment hereWhould we limit the number of retries? Lets say the address is used but another process which is stuck or anything like that +The problem is that the consumers position in the queue is not replicated to joining brokersFixed in as it is not yet part of the in +Selection refresh is called on mouse move event now +It seems the fact that logins with characters or longer are truncated by to characters by some unix commands was a red herring +What problem was this intended to solve? Surely its the kind of thing that would better be solved by a Container subclass that implemented a new layout strategyJust curious +The upgraded of Giraffa to went well +Ive attached the patch +bq +Yes setting an on the instance that will throw a in the fatalError method was the simple that I meant. +Meanwhile it would be great to get whatever snapshots are available to include this code +Regardswhats the status of this issue? Will this be fixed in the next release +From a licensing point of view I think the answer is Yes +I also just hit this problem +Could you go a few levels down into newInstance and see what the heck is taking so long? My money is on get or registerFinalizer +New patch that includes missing is stuck + it to Hitesh as it is done already +Duplicate of bugs should not have a Fix Version +Doesnt the jdk parameter xxx achieve the same thingI seem to remember that I had an issue with this timeout +this patch has not been applied to any version of tomcat +thanks for your effortI just pulled the files into an editor that set the right character setyours is a lot better +Note that the test for this issue is pathLimit +yes I am +Correct this is an error when the case was entered +I did a sanity test and confirm it works +HiIve just run into this same issue +And we can have a single distribution which covers all platforms without special configuration steps +Not sure +This piece of code hasnt been changed since the code was donated to Apache so the commit log doesnt tell us anything about itThe synchronization statement that is commented out enclosed the entire if statement whereas the suggested fix only synchronizes the body of the if statement +Patch checked in +Now with looks accurate to me +Update it was user error +However I see no harm in them other than they affect more of the code than intended by the patch +This patch is Could you please send the code of Json support files of Axisc +correct has not been addressed. +Just trying to figure out list of Errai issues which will make it into WFK thanks Pavel +i think i have better testcase in a few moments +Contains minor changes and one new important feature nowtask breaks build when Castor fails to parse schema and generate java code +it seems to work +Good on you N +The reports should use the best possible connection. +online backing +Ed any progress on my little question? Id really like to get this featurecommitted for the upcoming release but without a test case this is notgoing to be achievable +Bobby it was not a javadoc bug but a tucu bug after allCommitted to trunk and +Should be no problem to publishit under the ASF license +Basically catches us up to the point where we can run the code and see what is working or not +I have uploaded a reworked would need your review Matthew as well as a clarification on the useless params maps in +d n nodesAgain alternate suggestions welcome +That would require a change in quite a number of classes +The tests with this patch hang on my mac and on linux box +Since I dont see consensus on this how about moving this out to +Hey HermannsCan you want me to help you with this issue? because there are alot of components to resolve this i can help u with some components +Ok patch commited snapshot for deploying free to send in other patches not sure Ill be able to have a closer look at this before a little while. +Please delete it as well thanks +Also can you log the variable in both of those calls? A hypothesis has been floated about the possibility of a diabolical malfunction in the key tree +Joshua any updates on this +The reason to use Tomcat is as far as I now the Portlet Bridge only runs on JSP because of JSF +Heres a second patch +I think directly depending on fsync call after creat also will have one issueSuppose if create sucess and trying to call fsync at the same time active NN failed and standyb becomes active +Updated patch completes unit tests for CS for all runtime resource limits reservations for apps etc +Thanks for your comments Achim Ill give it a try that way +we should include fix artixds and javaspace in the you are adding any new rd party dependencies let me know as Ill need to getShouldnt be any new shipped rd party dependencies +Will work on better documentation and supply that as a future patch +Moving to backlog +Logged In YES useridCan you use for sequence creation? Like this CREATE SEQUENCEtseq START WITH INCREMENT BY +Attached a patch to add the three missing packages to the MANIFEST to and trunk +for patch includes the Apache headers +Ok +When and where will the next one be? I am in Mountain ViewPalo Alto +Tangring confirmed on th Jan in IRC conversation on that Documentation should be the No +It would be nice to make a tool to migrate the double encoded files to single files when we make a patch to remove encoding from +Thanks for the patch Alex. +In this test you can use EAP or EAP or perhaps +Linking to the original issue for to provide a more dont rather create a new linked task or subtask to work on this +A user doesnt have to care about those internals though simply specify the defaults on the sender and will use do you think about that solution? really good and works just as Ill close this issue then at least for the time solution sounds good Juergen +This has been a problem in the past but I would have expected or to have fixed it +This is just my opinion but to me it looks we can get rid of all modules have one big one big classpath and finally have only one big JAR file for Lucene and Solr but thats real masochism especially for projects like ES +But is that a great issue + for the to trunk. +This is a rather serious gap +Hi William +Sounds like an interesting feature +Checked revision revision . +Hey Paul sorry about the delay I will ping aslak on this +There is a new method install in a class +Ive added the code that should perform paging when listitems is called to the branch +Whats more if diagnostics are enabled the diagnostic error is logged with most of the information missing +Thats the whole idea of the abstracting away how a single is stored in the engine and simplifying it when not specifying a the name is usedThe is an property +It seems like the karaf installation requires datatmp to be created +Thanks for the patch Gary +Closing no problems appears when JBDS starts for the first time and search runtimes +? Thoughts +Arpit there are lot of empty line changes in this patch +Does it make sense to create top level module and have and etc as child modules instead of cluttering them at the root levelAlternately we can add them all to module instead of adding a new module and rename the module to or some such thing + well Dinesh there is just a typo in the rule description replace was previously thrown in the try or a catch block by was previously thrown in the try or catch blocks +json simple and gson do a better job but they still allocate more buffers than is necessary +I have tested this fix and it does indeed resolve +This patch excludes the modifications for skipping the creation of module entries for the root directory of the project +Done. +thanks for the patch and test +What if userhive does not existWhat if user does not existMaybe it is better to let people make the directories themselves +Keep in mind this work has been under review here for months now and there are watchers on this JIRA so I dont anticipate needing a lengthy review period like we did for other branches +Ted I totally agree lets set up a skype call for everyone to discuss this and get it moving forward +The idea with soft upgrade is that a user should be able to run software in soft upgrade mode and then go back to running software if it is a database +Should have been +Again thanks for +EXAMPLE to write files to a different directory that the root context installation directory of the container +so it looks for matching file here icons fix this I just copied the filter +I think we should use classes to handle deamon functionality +just to eliminate another code path your databases are not encrypted right +The for reporting the issue and particularly for taking the time to put together a patch +I downloaded latest and it worked fine +I will fix it +Yeah but not all primary storage is able to do soSo I wouldnt be against handing this over to the primary storage but if somebody wants to use the cgroups for this they are more then welcome to do so arent theyBut first Id implement the actual polling of the and store them in the usage database so admins can see how much are being used by one Instance +Switches from selection sort to comb sort +Added the French translation +I didnt check this thoughTo be also verified for Span Queries +I have attatched an example one simple faces site one simple composite component with a facet +I suspect that they wont let it slip again so expect something after the weekend +AnkitOK I cant see a better way anyway your patch si in trunk at r Jacques +Most installer issues moved of to no longer present since we dont have p meta generation anymore. +Ill be around this weekend too although in a different time zone +I dont see that any changes were made directly to +Closing for +Sorry I didnt set the component +Minor change to use the TCKs suite rather than create a suite with the same tests +Thanks Vinod! +Isnt most markup XHTML nowadays? Forgive me if I am misunderstanding but I imagine XHTML and Dojo are not incompatible +Pinging to get an answer on my question +This issue has been fixed and released as part of release +Fixed application using suggestion inDeployed and smoke tested fine. +The issue is susceptible for possible memory leaks please check it +batch transition to closed remaining resolved bugs +additional patch added +The problem is that a failed task doesnt have counters and we assume that we always get counters which causes an NPE +Bulk close after release of +trunk Thanks for the patch +Ok I would like to work on this issue if none else is solving it +Thanks for reporting this Jamie! Sorry it took so long +Id agree +I think well can do it but ti wont be the default features will can be activated with a conf file or a. + to address issue discussed in comment +and reports Purchases by Product Report Purchases by Payment Method ReportOnly Open Order Items Report is not concerned I guess because there are no LESSTHANEQUALTO or kinda checks +Assigning to the EAP maintenance lead +Its a copy of the with some minor modifications to illustrate this bug +Ok I made some investigations and I think tomorrow Ill send correct scripts for ms sql generation. +Consider it done +For boolean properties it is nice to have an additional check for is and has getters +The findbugs warnings are all in which I didnt touch +if set to xceiverCount will reach soon as hbase data increasing then DN hangsset to min only useful in following situation hbase never readwrite more the xceiverCount mapfiles in the min otherwise the DN will stop serving datathe chains is more data more mapfiles more opening mapfiles more xceiver threads comprare all solutions we can have in the chain controlling opening mapfiles is the most efficientmy hbase installation regionserversnamenodes each handle about region G data +It appears this patch also fixes for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +Maybe problem are here? I cant get it why android project creation work from phonegap distribution and dont work from cordova +When then just nothing happens +We will not be able to help you here +This is an unused config file entry but it is not the same name as so it would never be loaded twice +But to be sure I would write to the mailing list devMost of the components there have already been updated to the new groupId but there are still some old legacy components that have not been released for several years +has a lot of new code encapsulated in new methods and it shouldnt be too difficult to make it fit this patch +Hi MaciejThis is a far better workaround than the one Oleg gave me thanksIt looks like the issue is something specific to either Resin or Erlends setup because does not work as it is supposed to there +Hi Henkallright im going to take so the sync will be done at and UTCregardsaccording your mirror guidelines you are expecting from the mirrors to use the Apache Webserver +Anything to make things easier +I think it would be nice to reduce the logging of hibernate statistics logging every HQL on INFO level doesnt seem very useful +see question above boss +so closing +I am also repeatably getting this exception with +Reduced overusage of and other minor used left untouched as improvement of readability. +I was just thinking about adding some functionality to help people testing down the road +I committed this +Thanks to Dora Rajappan for provide this patch +The group RPC for lookups is now in the HAJNDI class method lookupRemotely +Nice work nkeywal +But with this feature you cannot deploy the task trackers without shutting down the job tracker +If your server logs fill up because of INFO logs I suggest you raise the log level on the server to WARN or ERROR or at least adjust the log level for +But the only approach I see for addressing that is proxying the interface at RT when its available +I like this design very much +We need a better example eventually +projecttrunk +Im resolving this again the original issue with is resolvedThe issue John raises is an interesting one but it does not affect and should be dealt with in a separate ticket +fill try to see where are read during build process +Closed after release +Patch reattached from bugzilla +first viewsize and view index are the params used in particular purpose not all the form need them +Upload a revised patch +Hello I have same error this message is column Value is not a valid option +Waiting for corporate approval to contribute this patch +Will be closed once verified with JBT with JBDS +Since we only delete something that has a timestamp older than now interval +Removed initial patch and attached Fixed escaping of sample binary text +I am sure you will get the error when you start the ecommerce webpage +What kind of doesnt work are you seeing +lowering priority a bit +mimej determines that these fields are invalid however they contain useful information +Until this issue is resolved we cant move on this +With new implementation the route for managed connections should never be nullOleg +Applied attached patch +Also is this the very first error that happens +Patch added after deleting the test scenario which adds files in the done directory and then checking if the history files are still moved to the done directory +Yes looking at the code there is a problem with this is apparently mapped to a server wide property and it is set using a service specific property +Post new patch and Ill take it for another spin +this would be awesome +From now the plugin chain is constructed but not executed +Patch with test case +fails when using the HTTP Protocol +All users are created equal Which may be what you said all along If so forget my previous comment +When not in Daylight the DSTOFFSET is so no extra checking is I didnt use your patch Leif as I mentioned above I didnt want to use but as I mentioned in bug I certainly appreciate youtaking the time to submit the patch in the first place and alerting us to theproblem. +Why add to? Also dont we want to update all MR uses to use not just +No problem +You are the RM of +So you have to install ivy in your +Workaround committed in rev +With this code the specs pass both run and compiled +So checking if an inode is a directory should be performed before converting an inode to be +Already fixed +David would be good if you could create a patch or a branch for this +for the patch +Lots of changes +We should also add a test case where output is forced to HTML because the processor generates an HTML element as the first element child in the result tree +So this seems not to be a Cobertura plugin related its the way the build is triggered that make it the workaround is doing an mvn package before doing mvn site +I analyzed the numbers and there was no performance drop +Thx +Fixed in r +Some specs would be extra nice +Still some work to do +Solr however does introduce a schema and much more that assumes a flat modelIn we could add a as a field within a this would at least allow the output format to return a nested structureI have not looked this patch so this comment may be off base +There was a bogus token in the defaulttemplate +If were going to change the btree code it should be more major of a change +at this point this is not a bug so the jira can be closed and the conversation moved to the user mailing listI suggest you take the working copy of and copy it to a new folder then do +Excellent! Thank you Dierk +BTW just to clarify the above benchmarks are based on trunk with or without the federation patch +JSE package list excluding java +Patch for branch +Ill see if I can make a quick patch +RomainThere is no new version at the snapshot directory do you mean I should wait until tomorrows snapshot and retryAlex +This issue has been fixed and released as part of release +I use Windows +Ive committed this +As I wrote it is messy +I have reviewed this patch +Really great JulienTested +If not then they should work fine in a +Reproduced using JBM +And probably other places tooI really think this belongs in the default error handler +Thanks for the patch Matteo. +Because we are new to this what is our next step? Will you assign this issue to us or anyone can submit patch to it? JeffI have attached a patch we have worked on today +Just wondering whether the problem could be related to the fact that I am using a local repo which is not at the standard might be completely wrong but at least I have asked +shravan thanks! +fixed +This is the implementation used when running outside an OSGI environment +I would like to get as complete a set of tests submitted for that change before working on applying patches for code changes +Given the usage of evaluate it seems that the developer would want the option to specify locally scoped variablesAs to your performance questions the numbers are due to throughput so total time to execute the benchmark +Im sure youre busy! Thanks for the update +It is slightly different than suggested by Tim +Ive got the same problem +I modified the migration guide under Issues and Solutions +dml test group by clause of unique constraint +Thanks for pointing me to it +The epoch value is causing the comparison to fail but I think that the dont care values also have to be updated in the way I propose otherwise the comparison will fail too no? newEpoch sounds like a hack to me not a principled solution +The external code was changed when preference page was added for problems severity customization +Seems like a useful feature but I dont think its related to this +Instead a key word might be better if we are trying to avoid introducing another top level operator similar to the case of outer joins eg union onschema L LMore suggestionsopinions on the syntax for this feature are welcome +Have tried creating a test case and also installing source following documentation on site but have not been very successful with either +going to change rthe reason for closing to wont fix +Fixed +This will be fairly easy to code +New to be added into the start up the release build when ready +I suggest this patchIt adds a more detailed error message on class cast exceptionAlso adds two launch configurations one that will run the plutoinstall goal and one that will copy the packaged jar from into tomcat for you. +Will a Host be tied to one of those service names? For s getsetConfiguration do we have to restart services for changes to be in effect +Lets just do it +While I see that this class might be helpful to you I dont see this as a general usecase that should be addressed on the Struts framework levelAlso from an architectural standpoint mixing Struts Actions with database functionality might not be the best choice +on latest patch +This particular problem was caused by the fact that the service had two ports both with the same soapaddress location URL +The declares the use of and yet the more recent broken version is used +I like it +That should solve the problem +This essentially renders maven useless as a build mechanism if it cant run unittests +This issue has been fixed and released as part of release +If I invoke maven from the DLL project that is when it fails complaining about CProgram +for you John +Even though this patch touches many files it pretty much touch them all in a small way that should not be too much trouble for anyone to deal with +In this patch in now I am mimicking what we do for storing ie write the warning text in the readme file using a Properties object +Yes you are correct +KristianYes I know about +This patch covers all the cases mentioned in the description + +Im still seeing this issue in +I updated the copyrights for the Japanese and Portuguese books also +Is the idea to write it in a seperate WAL file? So that i can be picked up first + the patch looks good to me +These files are also in shared which seem confusing to me +Was this fix included in? It doesnt look like it +The doccomments on close explicitly mentions that it should be set to false on mobile +the way I understood there was styles is one for the standard tables the other one for the sortable tables +I think your keys are not usable by Fabric +Is there a chance that the log HHH had no statement associated with it but was not yet registered will be switched to a debug level at some point? Or is it abnormal to be seeing that log repeatedly? +I just updated the description of the bug +I havent found the exact same version in brew so I used the one which comes originally for hibernate tools +The right one +Also attached is a +looks like a good place to put this piece of information exactly because it is job specificI do not personally like much the idea of having metadata in a tuple with high replication it looks a bit hacky to meI dont understand your concern could you detail it a bit more? In any case the tuple is ephemeral as the and is not used for persisting data +Anyone following this issue if you get the chance it would be really great to give this Portlet support revision a try before the release so that we know that the issue is actually just tested with and the behaviour had not changed from at all meaning that even though session was set to be completed session attributes where not could provide a example portlet project for testing if that helps +results are now always ordered plust user can select the order in which they want to see the results. +Please add your detailed comments if you wish to make the case for reopening it. +Thanks for your input on this Uma. +Hi Heiko Thanks for reporting the bug and your interest in helping get it fixed +Hi John thanks for the patchUnfortunately I dont know how to build the release targets eitherbut I read through your diff and it looks fine to meI also read the release note and I think it is simple and clear +Thats another problem I found +But this will be different paragraph than the one currently used so well have to maintain two same paragraphs for some time +closed with release +mrepositoryUserstucu +Pls share your comments on this patch? If it is ok i can prepare for other versions also +timestamps +it +Looking at everything around facets and group collapse and the join trick the Solr query syntax looks like the database world right before SQL +gets shipped to task nodes we should not add anything in it which is not needed in backend +Thought a little bit more and I think this patch is fine +Thanks Erick! Ill close this and keep an eye on +is per field and is resolved by the codec +Fixed rev +Close out issues. +The fact that you could before was a bug that has now been fixed +We currently include rd party jars in our release as long as the license is compatible +Patch with tests applied to and trunk. +I wonder if this could be related to as well although that issue is more about schema locations it does reference configuration validation a management only mode may be a step closer to being able to validate configuration and show the resulting model for that configuration before an administrator applies it to their server +In advanced mapping mode the changes are written through to the node and all of them are tied to the transaction so on commit all the changes will be persisted +Ignore last comment its Camel kit deploy thats failing so this should be unrelated to to check in what triggers a problem with the ETL example which was caused by having it changed recently to work in +I think this should go intoSetting the extension to does not have a real use case at the moment +To reproduce run HY +according to the ri CharlesPatch applied at r please verify +By using this file the will generate an improved description for this bundle +I reproduced this bug locally on windows x on +This however will break if the second object is initialized after the connection is set +Finished applying updates to +java + +compare compares the outputs of runTest and generateBenchmark and decides if the test succeeded or failedTests for the harness are specified via a Perl data structure generally stored in a separate file +Considering this is quick fix that should be the case +So the conclusion is that the test fix is not required anywhere in productclient code and is a configuration step +Uh wrong patch previously +PetrYour feedback is quite appreciated +Some of the warnings here are Windows platform specific and need the attention of a developer who work on Windows am closing this issue because all the original warnings reported have been resolved plus a number of others that have appeared along the wayThere are inevitably new warnings appearing and it probably correct that these are handled in a new issue otherwise this will remain open forever appearing not to make progress +it is clear why they arent the same +Closed as requested +That worked great +Since the problem is no longer present in the latest version of. +please can you post a link to the reviewboard so I can provide feedback +The bug is still there +yes this is definitly a for input +From my investigation the current implementation does not kill the IPC threads until the max idle time has expired +So it can be removed safely anyway +I think that its best if we change the javadoc +I thought this would just make life a bit easier for us +project in +If the corrupt index is a result of adding one field with one word untokenized and all then thats a more serious corruption and how we handleinform it should be discussedAt any rate Ive seen apps preventing to run on old but not future potentially buggy +Bulk closing of issues that are old and have been released or marked invalid. +Making this a blocker because there have been multiple reports of user setups failing +I focused my attention on the proposed patch which suggested a change in +If it does youre best off using a request attribute even though its slower +Thank you Christian + +Makes complete sense ChrisI shall try to look into this on my own tooRegardsVarun +Current timestamp is probably so the expectation was the cell not getting deleted I tried option and it works fine for my project +Thanks for the patch! + +Im tempted to build a small test app because I cant seem to get any messages in response +Yes the attached configuration does have the correct settings here so thats m marking as closed for now +The stack trace is identical +Committed the patch to +Maybe I didnt understand the problem correctly but isnt this just a matter of catching in get? +a million Stever +BhushanCould you send a phabricator review request? I will take a lookThanksZhenxiao +We could approach it from the other end +patch with branch +Could you update the patch against the regarding the newYes of course +Instead I will post a patch for itself +I havent seen this issue in a while either +Bryan thank you so much +I also feel inits may have different signatures for different servicesActually we could break down inits to use configuration as parameters +in a sql element col names can now be defined as CDATA so you have a much richer choice of characters you can use +Thanks again +Closing +Feedback is welcome. +Problem fixed long ago by knuth algo patch no longer valid anyway + +I also think we should use the maven default folder layout +bq +Attaching new patch +If you are using a really large hbase set of data +Is this really fixed +The build infrastructure etc +Im not building anything +Since the connection has been logically closed by the protocol layer there is no point feeding decrypted data to the handler +but it looks like unrealistic to me +Committed to trunk and in +We need feedback from the community before deprecating but this is at least what I have in mindThe responder thread uses UDP currently and we have written this connection manager class to take care of the TCP connections we use for leader election +I will need this in the near future +Yes but the version is know in the parent for more simplicity for modify the roo not look the parent +I dont have a feel for how important this is for a next release and like Frank I dont have experience +I have the same issue as I am trying to use listners to write history as our DB admins are not willing to create triggers +All the unit tests seem to be passing on my box without any changes +They dont have a public instance +bq +they try to raise up a geoserver running on during the test if that port is busy they fail if the port is occupied by a geosrver that does not have toppstates they fail just have to kill them and replace them with proper testsOn my box the build is repeatably ok thought + +Nit declaration of File file can be moved to the line where the object is first instantiated +However +There are many other things that could be done to decrease the latency for small jobs however +this patch finds the port using strrchr instead of strchr to allow ipv addresses +Only an employee of Sun or Oracle who has signed an Apache CLA can contribute such codeThanks for your understandingHenry +Please make sure a full testsuite is run in any case +There is just one place in the code involved which knows how to talk to itself +excludedContextsThis should not contain any spaces +If application name is not present it falls back to project name +Groovy considers them by to be the same +The order in the list would change with each run because the list was a hash map +Hi MinI verified this on latest build from ACS branch +From next time I will create separate patches as you +we should just fix this bug in the meantimerobbie can you review this small fix? +If we can somehow do a mapping job which maps string or objects to int ids these issues go away +Will dig in more tomorrow +Perhaps it could go into howto but is there somewhere Im missing that already talks about bulkdocs functionality where I could just appendupdate the informationThanks! +Yes thats exactly the problem + patchThe memory maps before and after appliance +What if the test outcome under investigation itself takes low priority than some new failures? This calls for analysis on the newer test run at which time I have no easy way of knowing whether a test outcome for some test case is still under investigation +Jazz folks are you sure that Jazz also has such a special HEAD thingy? The HEAD in CVS is similar to the SVN trunk! It just denotes that you are on the CVS main route +may want to lock those projects if you have not done so already too + for the change +Fixed in trunk should not be backported to due to significant number of changes to validation API +Closing this issue +The addition of will be addressed in the timeframe +What I can say is that it is the classloader from the first deployment which is being referenced but have not yet worked out howwhy. +add test good thanks Jonathan +I figure if there will be disturbance in the build anyway I can do it right after to not have disturbance for too long +ericCan we close this +desconcertatsin +Modified the testsuite to use stack by default if no preference is specified +No further feedback assuming my previous suggestion is acceptable +Although I do not know why to use DOM event objects to dispatch such custom events because it may pop up elswhere +I dont have the before after on the same data any moreI will do a synthetic benchmark +for. +Could you please provide me with more details +Currently after the index changes are retrieved the writer is closed and reopened in to be aware of the changes just pulled in from the master +Ill work on committing this +Cheolsoo can you update this when the tests pass +I like it s how great ideaI want that Dila make other one too for meHow i can connect to u? +Any news about this really grueling bugI cant setup any custom handler also not for a compelxType with simpleContent +bq +truncate does not need to flush memtables to disk when snapshots are disabledIt still needs to clear out the memtables somehow though or truncate wont actually discard all the data its expected to. +This is a good feature to have for a release or sooner but not essential for M +Earlier it used url encoding to encode and decode strings +Can someone from the Web team take look please +Patch log and additional file for sample Ferdinand that is added nowI removed all tabs from the xsl we use indentAlso there was trouble with your patch file + Ben Ive even configured confluenza and thats working great. +If we can reproduce this with a minimal setting we could ask the Felix mailing list +Attached is the version +IT added +Yes please. +The another patch with directly modifying the CRT environment is attached +We can check the status of bundles upon receiving FRAMEWORKREFRESHED event +Ran this through the trunk version of Pdfbox and had no issues extracting the text +This issue has been fixed and released as part of release +The fix for the automatic runtime detection has been committed to beta +Truncated! I put and is returned +Please try to create a sample that shows the behaviour that you experience and attach it +Good catch +Not sure who is the guilty here but illegally accessing some internal classes of the JDT is probably a good suspect +Doesnt this already can put a file in works file for me +Originally it has rows +I fixed some of those warnings +its within the tools +Fix on trunk revision Fix on revision +Was a second bug created for the beans? Because thats still a problem. +I usually do that when its late at night or when commuting as it needs no brain power +backported retry on rename to and +Hi Hairong what bug in the OIV did this fix? and did the bug predate this JIRA? or was it introduced by one of the earlier patches +Not much information there other than things are bad really bad +Closed for. +If duplicate services have unique transfer the unique ones over to the duplicate service and remove any duplicates. +Simple but meaningful +You will see a new button Add in the Ingredient edit will add units to the amount property every time you click pending +Jean could this bundle eg flatpack be released and published to the central maven repo +Ah +This is happening again withPlease see my comments toThanks! +Ill work on stripping it down to get you a test case you can reproduce from but it may take some time +Thanks for reporting and the potential fix +pdebuild fails +Shreepadma More comments on RB +Check the compiler settings Preferences Java Compiler Classfile Generation Make sure that Add line number attributes is selected +The restarted regionserver complains about the ROOT location wth above exception +Anthony you are absolutely right about the value +Other issues w our current RZKJust to be clear Im not arguing to keep RZK or to ignore Curator code +Hmm +If all that is needed is a javadoc tag then a custom tag like could be added +We could only undo the change in or invent a new method there +The queries being truncated is of course very serious as userssee irrelevant results and assume that LuceneSolr is nogood +Yes I think your assesment may be correct +What about remote debugging +Is there a better way +Thanks Hong Fang Han for the patch +which performs this kind of calculation and can be used at different places +Thanks Navis +Perhaps Im missing something +I can continue to dig out jars and manually install them into my repo until all requirements are resolved but if you happen to be able to point me at a maven repo to do the job automatically that would be great thanks. + to trunk +m file is a proper xml file? can you build teh projects on teh command line +Its a regression issuePlease provide database dump +Thanks Mahadev for the interest +I have not very understood if you have found some thing that it does not go +Initial support for maven projects +I dont know whats the best way here +The problem with catching only at CFS level is that one cant get parameters other than key and cf +we set the level to debugtrace to avoid log flooding but when we do need to a production issue we can enable those messages using the JMX operations exposed by However if you see any log statements that can be improved please let us know. +Propagated dependencies from POM to POM +The connection with the correct port also fails because of +This does indeed appear to have gone +This doesnt happen often but is deadly I only test the file cache loader so I cant say for the rest +Committed to branch and trunk +Already fixed by blockers and critical issues are resolved but not verified +Lets close or make a BSP version out of it +See for comments +DR release in mid March +If somebody does a patch that first tries to delete and only if of dosent work and we are on windows and the bundle has a native library calls gc twice and delete again then I would be more in favour +Please choose another +Vlan id is a guest network lable is cloudbr +Ive added these changes with revision +fixed +Looks okay +Yes that code was already in before this patch was added +Sample application that shows the error +Hi us it is not relevant anymore as we consequently use panel ids +Please take a look +Is this the case here +Let me remind you we are all volunteers and open source projects such as this wouldnt exist without the valuable contributions other volunteer developers such as yourself can give +Reduce tasks are not they will fail in fetching the map output and retry the fetch will succeed once the succeeds +Previous patch seems to have wrong parents indicated in the ejb ejb and war modules +The Editor assumed I meant to bold the wordfoo + +I think we need to go back and remove the following filesr root root Dec r root root Dec r root root Dec r root root Dec Is there something I need to do when creating a patch to ensure that it actually deletes the files and does not simply zero them out? +I am wondering if is there a testcase for testing this issue +The local repository was moved over to labs earlier today. +Pardon me for also jumping in late but I also have a number of issues with this +The file system needs to support file names with characters which are in extended latin +As previously noted the behaviour described is violating the contract which is clearly set out in the javadocsIm not convinced this can be done in a generic way by the pool and am tempted to resolve this as WONTFIX +I forgot Lucene should compile with javac without warnings and this is working +I commented it out and built to see forsure if anything other than what I changed used it and I just never gotback to removing itAll the monitor info is gathered in the Stats object constructor +Fixed Fix Version +You need to set the fix version +Aims +The above could also use instead of in +I thought setting different timezones for updatescreates and reads should make it possible to test this functionality but to my surprise this is not the case +Source code from the article +Patch removing the chained I disagree +You also want to take a look +As long as index and search do it the same way its a coin tossFor this particular case was chosen as the underlying format as the majority of Japanese text and searches that we are seeing actually uses the versions of both the Katakana and Latin characters +from meRunning through test suite is desirable +Can you keep this open? I will give this another looks this week and let you know +on everything else +For RC there will be an additional JIRA open to resolve whether or not this fragment belongs in its own feature or not +I just committed this to hadoop branch +We have cite tests running nightly on the build box so i will want to verify there are no breakages once the patch is committed + +Youre exactly right Geoffrey +Patch already applied by Solomon. + the domain class Junit test the decompiled enhncedThis is quite common scenario in our project +As soon as we changed the double to an i or a string the java program would read all of the structs +I think some discussion on this topic could save us experiencing a substantial Gotcha issue in the future +Will we just punt this to +Actually the text of the comments is still referring to perform +Looks good +Thanks the fixes are waiting for the merge +Manually tested! +We cant actually add those listeners if we expect the templates to continue to be infrastructure +The schedule isMarch June September we will still have to report monthly for the first three monthsMarch April and May of this yearUnless the board is willing to take our February report as the first +Sanne yes you are right +With patch contrib compiles and tests run +The opening of the dialog after pressing Select file was connected somehow to the flash settings in optadobe +Ok I also think this wayWhat you think should be the next step +The limit node was being pushed too far which caused the projection logic to not correctly get the output elements. +I added up some comments on RB Jimmy +RegardsRay Chen + added in rev t close the issue until we update the documentation +This is indeed an assumption weve had for all the security work +Field problems ispublic so are we supposed to access it directly? Can we assume the field isinitialized by the compiler when we get the result? Ill look into it but inthe meantime I suggest applying the patch since it doesnt change the statusquo regarding +Sorry it appears that this doesnt fix it for reported issue is caused by Groovys previously inconsistent handling of bean properties with multiple setters +Use the below instead + I saw this earlier but just thought about it again when reviewing the RC +Marking closed. +Fixed in r +If there is any appetite for this I have a Docbook book version of the Accounting help that runs in webhelp +i didnt notice i added it to the bamboo profile +For all levels below that we can reuse the current +Swizzled the patch a little to only add a PU to the known only after the unit was successfully created +Done a while back +Adds a test to the example component that asserts the existence of the current stock seed demo userLogins +Applied patch into master with thanks to Stephan. +Oh I ran the tests that failed in the above locally and all passed +Hmmm +Would it be possible to scope the list down to just those classes that extend or whatever the base ESB action class is? That would make it easier to spot the supported classes and not have to go looking for a needle in a I agree with your recommendation as well +Bumping off +DONE +This was just not for old repositories +Committed +Ive added the project lib dir for now I personally havent found a usecase where Ive wanted to call application code from my webtest +Im here to for the current structure and svn repo +This should be fixed with the next MC snapshot is fixed +It should fix the problem +AlexThis problem affects M components +Can you please attach XML and XSLT to reproduce the problem? Also please explain what output you are expecting versus what you are seeingThanks testing shows that this problem only happens if you are using Xerces as your parser +That behaviour is probably already supported right? I was just surprised that I was unable to remove the id when it was an inner bean + +Hello GeorgeThe fix looks good +The DTD describing the hyphenation grammar XML files +revision . +here is the patch for solr +Other online schema changes should be fine +image Load the debug version to find the labelidhttplocalhostopenmeetingsswf Load that URL and you will see the debugger all labelids in the text fields +Maybe it is bad design but the reasoning was that the last runtime cannot be removed +bq +I moved your existing repo to and symlinked to point to it so the old links should work until someone nukes the symlink +Thanks Rodrigo! +The answer may very well be that its simply more expressive in terms of the process designers intent which is fine +The relative urls are not legit according to wc +Is there any specific reason to have the json reader and writer at http transport level folder +Youre right Galder I couldnt reproduce this with latest infinispan server nor with any JDG higher than ER +tested again in several setups could not reproduce it +Thanks Gopal! +On my perpetual list has been investigating turning our class into a JMX +Hmm I think Axiom perf check is just one part of thisWe need to run some perf test on things like round trip time and memory and try to improve them +We still suffer from some rather difficult to debug stack traces but tests all pass and IMO the stack traces can be addressed during the dev drive +Is there anything I can do to help resolve this issueAlso any idea why it works for everyone else with scope provided? Am I using a newer Maven version +I will do a more critical review in the next couple of days +Checkstyle fails complaining about tabs when I run cd langjava ant clean test +On the does support for composite in the column value make sense question Ill add that supporting them would potentially make for a nicer upgrade for people coming from thriftCQL that already use composites as the defaultValidator +However we havent heard from Dawid in quite a while +I have added a disabled test case to I will chase the EMF fix +Attached a patch for review +Still I have not tested it in a real deployment with security on +Its not super core like its easy to rip out out +There were much concern for these patches during last weekYou axis devs can decide to postpone this pactches until next releaseBut I think the Axis guys should be concerned over the conformance more than nowAlso in Axis release should be considered +Could you give an example of how it might be considered more useful +I looked through the patch +Move it back in if you think differently +patch for minor changes to Luc +Thanks Sascha for fixing itAnkit +Why do you report it here +That would not be an issue if only ruby was not displaying ugly warnings among the rspec outputStill I feel there must be a better way too +It returned to viewprofile on save click while all other edit forms required clicking on the done button +This was already implemented and is available in. +Actually there was already +launch file into your old app to see if it fixes the problemI want to establish if it is just an issue with the grails upgrade now upgrades IDE support files +Patch created from the trunkgshell patch created from root of trunk dir +Better comments based upon migrating all of core and contrib +There is now a constant you can override to accomplish this see +I will provide a patch +Checking mtime definitely is a good idea but it probably should be done in a separated issue +Ah ok I didnt realize this needs server access first +Thats doable yup +Strange thing was when i logged in again to Rave gadget started to work +Some networksystem +Can this be backported to +Robinson StOrlando FL +If you could apply this patch to the cassandra trunk that would be great +for reloading strategies +Closed after release. +Colin thanks for the clarification +Take a look at and +Pig uses javacc and it does not work with +Hi RishiShould we close? +Why else should we support itIf you want to just remove all of the Avro RPC stuff from Hadoop it should not be hard +As longer i think about i turns out that this behavior inst as bad as it looks this works i could very easy be defect my code because after the creation of the extension method my existing code calls the extension method with my change. +I dont think this issue is a blocker +Could you please also attach CBACwsassetsprocsentitymngmodulesims to this issue +I think that should not attempt to do any outage detetection buffering +will commit after tests pass +Workarounds exist +Both are up for review and comments +And just link. +Even if there is a it should imho not cause a +modifiedcomponentssrcmainresourcesschemablueprint doesnt affect as that version doesnt have a blueprint version of. +Closing stale resolved issues +I am doing it since I am fixing tests now anywayI the code so that the remote test cases use a Minicluster while the other test cases use local file systemOne question that I had was why test is in this test suite although its not a macro test case? I didnt move it but please let me know if you think that it should be moved to somewhere elseThanks +Dont have time right now to look at the other failure though +We can then keep the bundle for a specific source hostThis is really small overhead and would solve this problemBut lets solve this in you think we can combine the message buffering and checkpointing +Thanks Todd! +Affected files werepbcast +Will commit this this +What would that queue editor actually edit? which artifacts +In another console window from the bin dir +by Environment +Resolve and close +This patch loks good to me +Archive contains a quickstart and a someone can confirm this works in a production environment i will commit the patch +Because of this the in is used over my hope my theory turns out to be incorrect because this would mean wagon ssh external could not be updated without updating had particularly fixed an issue in to make this possible Ill had seen that exact exception even while using version of the +Unfortunately i didnt find the repo +Thanks + + +Have I misunderstood it +No change to actual code +Essentially the naming conventions and navigation used by make it impossible to select the same related entity to satisfy two different relationships +Fixed in the SVN +Linking the issues +I filed but the real reason seems very similar to what is being discussed here +Changed title to be explicit that what is wanted is very pauses and few retries +Do you want to give it a shot +Add back creating of files in +Works again now thanks! +in Tools +Im not familiar with it and have not had time ot look into it +Its not a Nexus plugin its a Maven plugn that is shipped as a regular wont affect behavior of Nexus +Sure I start working EmmanuelIm confused as to why you deleted all the resources with this patch? I cant see how the patch will function without html and js files +dont have good names for them +I have attached a patch file called that should correct this issue +they are not really there to configure the application +A Pull Request would help +Will commit the patch shortly +We would absolutely love this functionality +Here is patch fixing the bugI updated the testcase to parse such input +This is related to +The source code is already committed. +This jar contains the fix for an undeploy memory leak +I should not name the binary GA until is is actually we need an updated jbossxb release +Please ignore the javadoc for now and I promise Ill provide them later +indentation is spaces in places +patch applied +Yes it should be closed +will take a while before I can start reporting +Closed a wrong ticket +Closing all resolved tickets from or older +All tests in branch passed +Agreed it is a duplicate closing this +This way I would just add a transformer prior to the receiving channel of the so the payload would already be a Smack the Smack properties through message headers would be nice but either way I would probably be running the Spring Integration message through a transformer prior to letting the message pass through the outbound +I still think this is worth patching because the patch is so simple and it saves one in the provider factoryThe problem is that it wont work + +So all issues not directly assigned to an AS release are being closed +Thanks +Lets see then. +Feels like BD will avoid that potential confusion +how do you suggest to avoid the versioning dilemmaThis will be done by using a Maven repository without requiring users to use Maven +There is one new bug possibly related to this +Local to local there is none +out of worked and th failed +Thanks Zhong! +Changed as suggested after code review and pushed upstream. +Added license headers to many files and addressed missing licenses for imported jars +Im not tremendously happy though +The approach is actually very nearly implemented by whats in trunk +The nd and rd are really possible and so should be handled gracefully as well by proceeding to the delete of other filesdirs in the +Heres a patch +Assuming existing tests will verify errors in refactoring +If someone has a source code environment please leave a comment to tell us the result +These blockers and critical issues are resolved but not verified +Thanks Scott! +Please open a new issue if its not. +It occurs in when one of the attached methods is called +Thanks Rakesh for reminding me. +In the future we might add other backend items like Camel +Adding a test to to prove the problem and the fix +Validate all mandatory configurations before startup of the process +note this is for hibernate tools core in the +What needs to be is the PoolFactory itself and not the Configuration +define mix +For me it is showing that it is completing in seconds but that is impossible considering there are sleeps through the whole test +File structure generated by JBPM installed changed since the version moved patches need to be updated +Also cleans up the tombstone histograms from dont think changing API in minor release is good idea +Would move this to a util classFeedbacksuggestions welcome. +String literals not usedis used +If there are Maven Ant tasks that can do this that would be best +Only once I saw it running in the test environment +This patch looks good to me Laura +For reference I just tested with Oracle on Linux bit it doesnt throw any IAE +I like it much better than the one we have now +So the message is misleading +Attached see no errors in your logWhat is wrong with this recording +POM for cpaptf project +Verified Fixed +No response from reporter +Ive updated the forum post +Please verify +Verified and found to be working correctlyPlease reopen if necessary. +I just committed this patch to Apache HEAD branch +Attaching patch +Hence I feel updates to these must be done by JIP and not TIP +What is currently missing is the concept of an upper baseline vs +Not sure what caused this but its fixed now +adds tests for thaianalyzer token offsets and types both of which have bugstests for correct behavior are included but commented out +Is performance appreciably better than over HTTP? Id expect we might see a significant difference with messagesTo test this we might +Note The repro shows this using the embedded driver but I actually only saw this in the original test suite for the client driver probably because the embedded driver is always loaded as part of the call to used by defaultSuite +This is not turned off in CXF wsdl is getting generated in our home directory plz anyone let me know how to swtch it off or how to change the location of its generation other then the home directory? +It allows a producing a mimetype like applicationjson to be handled by the applicationjson providerIm not sure whether there are more places that should be adapted in order to support SSS + +BTW is not used +Verification failed withCounterTotalUNDEFINEDREFERENCEDUNREFERENCED +needs to create a instead of a when it sees the new configuration parameter +We already use for so yeah it is good to pick that up tooRegarding the implementation see +is already fixed +Wed have to do an instanceof test inside the main reason I dont want to do this is for consistency with the standard Java exception chaining mechanism +Closing. +Well definitely have another one alongside the GA release of the upcoming Babbage release train. +We definitely do not see this in our use of Castor +Hi JulianPatches are always welcomeWe currently work on improving Wicket Ajax code for +patch looks good +Remarking for since this probably wont happen before +Perhaps highlight the infinity in a different colour or something perhaps a glow yeah they all look really good especially on a dark prefs to V and V +As well your looks nice to be included in the distro To conclude the only real problem by now is +xsd schema defenition now +Hi Jan could you post this diff to? Thanks +This issue is part of a bulk update that required reopening the issue editing it and then resolving it +The problem is still present in version and we are also badly annoyed with this one +Suggested fix awaiting a testcase +Is there any fix for richfaces GAThanks +The documentation still needs updating which will be done as part of this issue +Well I suppose if there is nothing one can do you might as well mark this issue closed +Continuing with the import may result in unexpected behavior and errors +This is still being covered by the test +I dont need to read the entire Nice +Is it possible to fix it for too +Ratherthan check for newer modifiedDate I would just save the original modified dateand check for if it changed at all +Are there any plans when those methods for keykeyrefs will be available +I think it should be relocated at srctestapijava since it is a test for Java API not a specific implementation +I polished the patch a bit and made it work with non named parameters eg just using? as well. +Bulk update fixversion to not see this as relevant nor possible. +request to pull upgrade to Hibernate submitted +Old JIRA cleanup +I meant to make it Minor +may be you should expose generic interface for in your core + +And this may be the reason of not disabling of appearance of the same CSS class name +Committed at subversion revision +I felt like this feature split documentation into too many places +Cheers +What we need is a value stable enough not to be altered by changing the query used to access for the pk we use JDBC database metadata to figure out the pk and when there is one we get it +Updated to work correctly against tip of trunk +is limited to intHowever use for example in seems unnecessary +The JSON accessor methods arent really intended to be used with XML elements they are meant to manipulate structures which XML elements are not +sneaky glad that this stuff is single threaded in +Please if you disagreeThanks. +is this something that belongs in core or would it be better to make it available by the site plugin +A test is also included +file path configurations or routing appender configuration etc +The fix looks OK but I dont want to commit it just before the releasebeta is not in the dropdown because it is not released yet +It would have to be implemented slightly differently in this class to achieve a read but it wouldnt be too different from what is currently implemented in. +Pushing this out of this release +seeking branch in JBT nightly +It looks like you have already moved beyond that stage so the point is moot +Is there a better way to avoid the duplication of code +The previous patch failed to generate documentation due to Forrest compile errors +Patch to fix audit release warnings +But I think we can keep this code small +See the mailing list +User Admin in Rev +I will look into it +If two proxies talk to different servers I do not see how they can share connections by sharing the same IPC clients +Tsuyoshi and Andrey thank you for investigating this +Could also be bad ramIf scrub is erroring out we might be able to address that +Just forget to mention in commit log that last is a field on entityVikas +Heres an example +samplespetstoreWeb +Patch +Committed thanks did not take into account test cases that used Constants? Do you need me to fix that +Im new to apache so just getting up to speed +This issue is now closed +Here there are some undefined do I misunderstand the encoding and charset attributes +Karl just an ideagiven the issues you had with Derby instead of doing performance tuning or analysis maybe Marks idea is a good approach firstI know i originally suggested hsqldb i dont know anything about Derby except that its apache but ive done real work with hsqldb +One question for you though does loosening this check for the QJM case not unnecessarily weaken the check in the case +Commit the first step change to use file reference for the bundle installation +This is fixed for AS but as AS introduces Messaging a new conflict arises. +I intend to remove this attachment +In the next month get to where you can start writing codeI will send you Chapter of Mahout in Action which explains pretty well the structure of one current distributed recommender implementation +Verified by Mark +So its probably not a issue +Because it did not work we tried other optionsDo you really want me to open a new issue? I looked at the summery for this jira it does describe what we found Unable to receive any messages after message broker +Reporters of these issues please verify the fixes and help close these for this issue as per the previous comments. +we need many more tests that do this in Lucene +svn ciSending Sending srcmainjavaorgapachexbeanpropertyeditorAdding srctestjavaorgapachexbeanpropertyeditorAdding srctestjavaorgapachexbeanpropertyeditorTransmitting file dataCommitted revision . +I have this problem also if I invoke a target only on the main project +This one is doneThis patch is against the latest trunkFrom my last patch here are the major changes Changed to use the +thats odd +See +Theres no way then to make comparisons between projects +Correct me if Im wrong but is not officiallyreleased yet right? Do you know what changes were made to fix thisproblem? Is it something I can back port into easilyThanksBob +If you continue to observe this behavior please the issue including details of your browser environment +When updating a field of an embedded object the expectation is that no extra index processing will occur +Id be OK committing this is according release notes +So in order words cannot submit the response object to the underlying connection directly +Now we are using Tomcat I think we can close this issue +You get this exception when the EJB youre attempting to invoke isnt accessible from your clientPlease use the forums to ask questions +Better begut de cops de puny +bq +can you create a phabricator entry +Should delete memory allocated for the stub once the job is done is client code in samples fixed some of the tests before the final release +Let me check +It works for me if I turn off authentication and SQL authorization +Marking as resolved fixed in PermaineThe JIRA seems to be about two separate issuesThe httpd depends on the httpd package including the ssl certificates installed through the httpd package +Heres a patch that causes a fields default value to be used when no column is found corresponding to the field +We should think about solving this sort of issue once by maintain such lists in a plugable source of configuration and supporting the ability to hup the service +if the are they are ignored and all the superimplementation of the bridgemethods are also ignored +We can reopen if theres interestHey MikeGuessing this has been superseded by all the work thats gone on since you opened this one almost years ago. +Bela can you re attach the patches this time with the ASF license granted +You could instead set the balancer to run less often and have one big batch of regions moving at the same time while some region server may get really unbalanced +Therefore Ill go ahead and close this ticket as invalid given nothing is wrong here +in the meantime it would be interesting to see if just doing what you have in the patch makes your problem go away +Im not sure how this happened but this is not something that has been done yet +I was looking for some help there due to the javadoc involved +This is bigger than just adding filelocalhost to the path +Mailing lists moved +Thanks for patch mighty Chris +UPDATE I was wrong version behaves the same way +But both were marked as resolved +Thank you Jing. +Tim I tried your lock suggestion and it does seem to eliminate a great deal of even without asyncexclusive +I dont think we need to think of it as a Std +all fixed up now removed the existingPassword empty string check +Thanks for digging this out +I welcome this proposal as this should also solve the problem we discussed yesterday of getting people started with giraph without the need of many because we dont know on what graph theyre going to test it +See the discussions at +Looks good +Its only OS X on ppc that I would like to abandon. +Committed revision +Ill get some basic work committed today +data file should always be possible after shutdown +Oh I forgot that we restrict table namesIn that case we could just prefix all row keys with x except for the special end of region which would be x +vote for option as well for the same reasons you pointed out +And throw an exception if it doesnt exist +Will confer with Jurgen as to why will be thrown in corner cases instead of relying on singletonCache object locks to handle concurrency. +Hi Andrey I wonder if you are able to look in to the code +The table is defined to be boolean currently +Please do this. +With regard to pluggable I can try to extract the patch but it will be minority of functionality from this patch +Patch for test ignore previous patch was incorrect version + toThe image text you entered was not correct +ah it means that the rpm Thor is using was generated with an older spec file. +While testing I found another mistake +Ive just tested my project with your freshly built version and Im still having the problem +Annotate the filter UDF that sample is rewritten to accordingly +Thank you for the feedback Dan +runs fine +Fixed in this since has been released +Fix a findbugs warning +Failed when because header only had one character n +This is on purpose +Both and are added and removed dynamically very often so it is a good idea to make them Sharable and doing so fixes the reposted bug. +Any ideas on what could cause this +was broken by +Updated the patch +I dont see the correlation between the source and the binary +These pages are generated +A patch is attached +srcmainjavaorgapacheservicemixjbiframeworkTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixjbiframeworkTransmitting file dataCommitted revision . +We can actually set relative paths in the +Please provide a test WAR +With this attachment do you reproduce the problem? +Well work around it +See +I will go ahead and upload a handwritten so that we can publish it onto the release + was added in +Committed revision I will update the wiki pages accordingly +I believe I have fixed the issue with the local connection also + is untested +I will start working on this item +Also publish +Instead of using jms based queues the internal messaging is done via java queues +I think that this might be best handled in the type packaging ear includes its dependencies so dont pick them i missed this feature can i disable transitive dependency nowThanks for any advice. +Applied fix. +Thanks Arvind +it has a reference to the current and a list of historical contains one value that was in a variable at a given point in timethe package will also contain a history log session will have a map that maps process log classes to handlers +You probably have not seen the issue since you are not working with secure hbase +We need to commit this to and trunk +Its contructor takes a Model object for instance +Since this issue occurs in test code can you try running the mvn build with option? At least that may identify whether there is core code that is preventing a Windows buildAgreed that I dont think Windows was ever a planned target platform for Flume NG +I know that I definitely wouldnt have considered the importance of thread context in this situation +Of course if directories are used the matter is a bit different then this is neededLeft is only do we need and if all we want to do is reading services? Isnt it overdesigned +Manually tested +In adding new methods i was told to avoid using as they are deprecated +How should behave if you have many levels ofsubprojects? Update leafs only? +First one is +Thank you very much for this great pluginAny updates on this jira +We have been suffering a resource leak of this kind because we were seeing multiple Spring bean factory lifecycles within the scope of a single application classloader scope +Thats a good load test and thats why I had these errors I think + +Committing to trunk +Hence refers to description should be solved +Just out of interest which version of BRMS are you running? was done a long time back and there has been no further work to upgrade that integration +however should not be exceeded so you should not be seeing connections open from the pool with the default maxActive settingtest does not have to be turned on for excess idle connections to be removed when they are closed by clients +t is possible that the output stream created using caches data in some underlying layer +I just committed this +If you point me to some documentation i can do this myself +HowardIf this needs to be in CP you need to provide a tag immediately as component code freeze was more than a week ago +Try it again and It works thanks +The right way to get UTS to compile with the new Rail typedef needs this to be fixed +Doesnt this assume that the process reading the block is andor slow to read the entire block? That doesnt seem to jive with the vast majority of HDFS use cases +IMHO theres nothing we can do on our side +Both of these patches should be committed before +Closed without integration test. +bulk close of all resolved issues. +Available for reassignment +I am kind to work over the issue but would like to agree on the approach Chris and Andreas concerns seems the cached bounds not being that accurate does not seems to me to be that a hard problem more control than polling from time to time by the other hand could be too much a requirement that cant be addressed at all +These classes would have one method per instrumentation point and all the metrics computation causal logging Chukwa integration and so forth would be hidden in implementations +Another patch +String message Missing rows in expected rows nt expected actual result nt actualOtherwise the patch looks fine to me +It can be used as a Guide for BTI Developers +Hi this issue states Git pull has been sent but I cant see any URL attached +fixed +I committed this patch. +Closing issues in +content shows the correctly set header settings as defined in the previous comment +I have some comments about this patch so please dont commit it yet. +I think that Classpath container should be extendable by extension point defined in maven core wtp project or any other project that need handle its classpath in different way they will just implement extension point of maven +OlegUnfortunately am unable to apply this cleanly on branch +Hi BilginI hope to do a complete review in the following days +I guess I could potentially just implement the fixes for the simple case +It should be in addition to the which should help in my mind to overwrite these standard properties +Scrub should fix it. +I assume that Axis doesnt normally have a problem with this? If so is there some sort of interaction with Tomcat +How could I deal with it +Once the scratch file is guaranteed to be private then it can be safely closed +Hey Kris I got more work for you GoodI can reproduce the problem with searching for Redistributor +Hi SiIf you havent started looking at this yet could you hold fire please +Is this an actual issue or do you want to report this as a WONT FIX +Please confirm the current status of this issue and if its incorrect change it as required +ok +Version has been releaseClosing all related resolved issues. +Actually you are right with the current version we can make it an interface. +Any update on the status of thisI just found a way to greatly improve my projects design if I can just use this featureI am using Seam EJBIs there any way to use the hibernate mapping file only for the class that I need this functionality and still use the annotations for all other entities +verified this behaviour on the latest build and found that it works now +Id argue that if you are running in a cluster you are using a testcase as a tool to determine something +The only thing I dont like is the naming +If youre deploying inside an ear then the settings on the deployment tab of the server editor are irrelevent as those settings are ONLY applied for standalone deployments +Ill investigate the docs when I get a chance +It sounds like a good suggestion to change the note in the developers guide too +It would still be niceto get it working again though +If I understand your serverdatabase is on a remote host +I am anxiously awaiting RC and plan to verify everything there. +Ive attached a patch that works a bit differently than the current implementation +Fixed in issue The error handling is different due to the switch to the rbovirt gemCopying comments from mfojtik +It is not safe to use code from a JDK even when that code comes from Apache originally +by having an overview that has a systematic approach we can get a good appreciation of the coverage of the test suite for the job to multithreaded job executor tests to module +Unfortunately thats the best we can do from my years of experience +Once concern I have is with changes to since it adds a little bit of overhead to each record processing +Well sounds good Renato +So I should delete the fileWont it just return whenever there is an update +Is there some debug I can turn on that would shed some light on what is going on? Also with your knowledge of Infinispan could it be usefull to code a little snippet that would output usefull information like the cache and cache key in question when a NPE is about to happen? I could then insert this code to provide additional informationsThe hard thing is this thing happen really deep on the stack it cant be really any deeper and its almost impossible to know what really triggered it especially if its a concurrency issue somehowI was hoping that you would see in which scenario this would happen from the stacktrace maybe with the additional information I am talking about above we will be able to sort it out +Hi FredI think this will suffice in near term +Verified fixed in the ER build. +Thats the reason why CXFs pom has a special profile to make sure CXF is built against under and built against under jdk +Thanks +k is for and K is for with K doesnt appear to be a valid option + to trunk +If the script doesnt work correctly against the head of trunk I would suggest rolling back to the revisions indicated in the diff file then running the script and then doing an svn update to merge in the changes since the diff was made +So mvn releasebranch branchName https a feature request and not a bug +Ok this patch includes with the synchronization in rolled back +Committed with revision +Yellow Dog Linux on an Apple xserver with IBM JDK +Can you try the snapshot +The hard way +Patch attached +I checked in a patch to both the main branch which is slated for as wellas theKeith +Patch looks good but it needs a minor fix mime types may contain also +The technical reason for this is that annotation processing itself is implemented as +I think its ready +Please find attached an updated patch that solves what Bilgin has pointed out as Additionally I changed the template to show the Displaying out of z string even if the javascript is disabled +However like I said above I was hoping to take a crack using guice to do most of the factory stuff +support is still not complete as two nodes do not see each other when binding to addresses +for the right one +What we can do. +RESOLVEDTemplates for the following tags are createdmodified richdataScroller richlist richsub rich on +Necessary so that it can work +Committed changes to and +Sorry bout the overrides +Ill try to look at that +I like your point about being able to call du on the table in the trash but do you have any ideas surrounding any other means of a user being able differentiate between the tables in the client API? Are you thinking that trashed tables may still show up to some capacity in the monitor +Removed setCause method +Couple more iterms +Closing we released version. +See +Thanks for the report +The faq answer isnt incorrectERRORCODE is a dos env var equivalent to have test too the solution written in FAQ and it works +This allows logging exceptions from repeated tasks which is basically impossible with STPE +to attach journal +yes but any reason why we dont just link that up to the existing Verify provided by WTP +It was easy and is working +Linking this issue to because this problem was discovered as part of the work on that issue +Forge solved the plugin installation problem but unfortunately a bug fix that was implemented by Vineet and included in has a side effect causing the Forge console to wait for an input character +via Annotations or a specific implemented interface? I mean you can set a policy which says that you can change the username but if its physically impossible due to the fact that LDAP does not allow you to do so its going to be somewhat confusing +Issue fixed in. +Let me know if I can provide any additional information +General unit test creation is done. +Pushed tested fix to affected branches. +Fixed in r +Here is a zip +bq +Sorry for having forgotten this +Well see if JCA has some change that affects the behavior +verified on last build +I dont see any problems in +This maybe a windows issue +If youve not received seen a subscription verification then youre not subscribed +I cannot reproduce this can you attach a sample project that reproduces the problem +Or maybe a new would be more suited for what Im trying to achieve hereIt could deal with exploding archives and deploying srcDirectories and have a parameter overwrite or redeploy that defaults to falseJust thinking out loud here +This approach also makes it much easier for any community member to just take the Weld subsystem and give it a try with an existing AS install +this mean you did not add queryString +For any further changes we can file another issue +And will provide a patch soon +Attached is the removal of the explicit background color setting to white +The parameter hostname defines the IP address which will be inserted into endpoint URL. +Attachment has been added with description Updated sample has been added with description Updated sample PDF has been added with description Updated sample FO added some examples for has been added with description Updated sample PDF has been added with description updated patch against latest for your patch Andreas! One question +Added Punycode support +Weve shutdown the Gossipers timer on node A but node B will call gossip choose A and A will still reply +The series is now in maintenance mode +I committed your patches do you still want the ecore code moved to ecoresrc +please review this +Hooked up readlink support added in winutils +I havent updated the patch that Alejandro suggested +I have a tree also up on github at gitapurtell and Brian and I are tracking each others changes +All tests in passed for me +I will review I think I last worked in this area +Latest version attached +Another option is to keep p for geronimo but the port passed by p should override the vaule in +Comments were not patch +looks good to me +So shall we close it back +I think we really need it +issue seems to be a duplicate of +Wow if UPS can do this themselves is there really anything else to talk about? That leaves us with NO UI changes NO data model changes and we just have to rely on UPS to figure their own stuff out +with all the performance improvements and bug fixes im rather eager to get out there and not wait on these two remaining features +Reopening +Thanks! LGTM tables are exceedingly rare and commonly misunderstood +However I cannot exclude that other flows may show the same problem for as well +This was integrated on Jun and the status is OpenThats confusing +The same sort of mismatch will occur for aa a ao oe o +VerifiedThank you Alexey. +offers similar functionalityThe reason being if the index can be completely recreated from other sources you might not want to pay the performance hit instead recreate the index if corruptionhard failure occurs +seems fine again now +No code changes made only changes in information message +This is my first contribution so excuse me if I did something wrong +The patch does remove errors in fops junit tests in headless mode +I looked into this and the access to the decoder and encoder is synchronized multiple threads share the encoder and decoder but the access is synchronized +What do you thinkId like to help out with this +Hi Jeanso +The attach diffs contain tests and a fix that will resolve the issue of the missing persistent header from the received stomp fix was based off of source code +Verified. +I made the Etc modifications in r +It probably can be removed +Sounds fine to meColm +this test fails of the time and I improved the test to fail of the time or whateverBut really I think its more useful in general to have fastfail on by default +The patch obviously works +as both tests pass on local and the patch is still applicable +A patch that introduces a new pipeline type that implements the mechanism described above +Very interesting +Patch +But it would then be spec conform so Id rather say it would fix the code rather than break it with other EG folks +This patch make sure that js libs and include given using the options and are used in priority ie before the detection of the version + +A new index page need to be created +Just run ant and go the the Example main page +I am updating this issue to mark it as closed for future comments as it is now past the deadline for incorporating comments +Remove logging operations from +Changed Name to Service Name in the Contract control to correct this +Im not too familiar with Python packaging but ideally snappy and would be by +Fixed. +Well a is for the same reason not possible why there exists no works on a parser model while SAX is +Ive unassigned this from as no additional work is contemplated +The version of that I used was taken from RAD version +Ive just committed the patch to master and updated svnpubsub dir. +Thanks for all the help guys! I have the chapter to add Maven integration with FacetsD +Not complete by any means but its a start +This issue has been resolved for over a year with no further movement +This URL we get from +It is used internally to Sling +So we need to add an extra style which wont hide this calendarJacquesThis change was intentional as the navigation buttons are related to the calendar and not to the events +When performing incremental build we need to know which previously loaded sourcepaths are not valid in the latest state of project +with the concept +Looks like this one is already fixed. +on the patch +I applied the patch with some use do instead of doNotice verify that work too with your servertest if the connection is open before to use it +Per the docs the book field in the class must be of type String to match the primary key field type of Book +Except cant embed the version that reports +MQJMS invalid security authentication supplied for There is a problem with user permissions or Make sure the user accessing MQ Queue Manager is part of the mqm group +Is there a way to retreive ejb and war properties differently? +Question are there any good arguments against making based on a +It is just maybe redundantBut keep the way it it currently for to reduce the risk is prudentFeel free to close this one +When the user prefs are passed to the page properly encoded using encode or any other escaping mechanism that correctly converts all octets to the hex encoded form it works just fine in the Java build +Hmm +Does anyone have knowledge of this codeI am willing to put time into a patch but the documentation for the class is VERY sparse +uses the Aries artifacts and we dont it +Maybe it will have helped +The hang happened every time when I turned off the activemq service during the test +At we are trying to change the auth to use and authorization will continue to be via +no feedback so moving of +I dont quite understand this question + set fix ChrisIm a student at Oxford University in England studying for a masters degree in Software EngineeringIm very interested in the OODT project and Id like to get involvedIm hoping to participate in Google Summer of Code +Lets make this a priority for +I didnt see the bug when I worked with a VM with one ROOT and one DATA disk +Then I could run or package I then tried to run combined goals clean packageThis resulted in the original error again even though I had Maven configured via the meclipse the workaround apparently does require not running more than one goal in got me thinking what am I giving up by not using the embedded Maven? Apparently executing multiple goals must be a feature that requires Maven embedded capability +And prod should disable Seam debug mode. +Fix coming soon +patch applied thanks fix the files in source code and rebuild it +Do you have a proposed fix for this? Or why is treating the trailing +See the section Enabling Saxon in the attached PDF. +I do have the JMX Console enable so I am not sure why there is an issue with the JMX me know if there is something else you would like me to try +If this is correct I will begin work on a connection with JMX +I am worried that more than just this bug you found are affected by this +This has been fixed by the whole buffering has been removed withthe good to know +Not just minor when we generate code for large graphs +depth with argument is not POSIX for find. +Hi Jim Im not familar with DB wanted to fill in some additional details can you supply the SQL commands that will be required to set this +Yes please change the page to read Tools Guides + +No the change should be done at this plugin not then here we go +It cant be we release solr as an application on but separatelydifferently as an API over on I disagree +But for sure for Sonar +I prefer the second option +I assume you were doingGo to httplocalhostofbizsetupcan you show the steps you were on when this is throwing an error too but at the Create Customer screen see screenshot +Indeed may save space fro specific extrasoarse cases +Devaraj you can add a suppress warning for javac +Perhaps the correct result is not to return a visible length but rather to throw an exception forcing the client to retry or bail +See attach file +just need mcompatible flag to change the resolution status +I dont think this needs to be sanded down further +It only gives you a few days though +Ryan Sure I just wanted to iterate here a little bit and then test it out with the implementation before submitting it to +I was doing the new patch with only the FOP fonts if you can wait a bit you will have it before a button to print a FO file with all the default fonts +I created a subtask for the CLI bit which is not critical +Unfortunately however I was getting the exception only when I was changing the Spring version +Will wait for Owens new patches +Buildmagic isnt being developed anymore + NikolayThank you for the contribution +I would think that SelectDraw actions are also more common than delete +The reloading wont work in the parent controller +Im agree with Andreas +In general wrapping if needed should be handled by a higher up layer like the proxy layer that knows something about the signatures involved +Chuck please close this issue if you are happy with the resolution. +Its also quite late in the cycle so least change from seems quite appealing +Thanks for the patchesStill to do change someall of the interop and samples to use the new javawsdl tool + +Maybe include a config file that lists activator types to be mocked +Weve added a link to the original file in the README +I have added the line to wire up CLI docs +Note that you must specify DTD file for these entities to work +Sijie for Bookie format it would better to put it in to put all bookie commands togetherThanks Sijie for the suggestion +If this patch fixes what seems like a pretty nasty bug can we get dbcp pushed out to release? MaybeThanksRegardsJames +By all means +repository migration done for migration done for TYNAMO +Even with maximum possible streaming efficiency I think we still want +The patch is pretty simpleThe test copies the and adds the part where all images checksums are being compared +I have corrected this problem so that when a empty file is given it returns a empty stream +remove resource min from its pb and pro to +This should wait till is checked in +The goal is to get all of these apis in place for a longer stable api release +When it failed it was difficult to trace what each part of the method was testing +or is there any reason why it should belong to +That a call to get returning null means +Currently there is a file orgapachejdotckpcsinglefieldidentity only for application identity +And the fix has been ported into the TLV class +Its essentially a duplicate but tests a different code path so Im leaving it +I receive a lot of questions regarding rmls behavior and I point them to fs commands +Thomas we have to resolve this as unreproducable since it always starts on port for us when we start it with in the jndi will have to help us reproduce the issue if we need to fix it thanks +Layout can be changed +Thank you Arpit for porting all the related changes to +OK thanks for that +moving to BETA all to +Thank you Jing for the review +Moving it out of release. +We were using Tomcat when we found this issue +Attaching a modified patch adding this feature to +We use tree walker algorithm etc +was fixed quite some time ago. +for this JIRA we should do the inline splitting which means waiting on the distributed split to finish before going to the next item on the queueIt isnt just as simple as saying just make it parallel since there is an order to these TODO items and there are a number of different types of work items on the queue +Similar issues exist with bundles for and +NET vs similar to Java is a bigger conversation that we definitely should have +Thanks Thejas +If I run Solr branchx with this patch applied will I benefit? I can see from the commit log that unmmapping must disabled to benefit but I dont know if this is how Solr operates +Just to be clear I like the new ANT build successful message for some uses it is just that I have other uses of GANT scripts where that kind of message is inappropriateconfusing +Fixed by adding documentation on. +Moving to +The best way to do that will likely depend on the binary packaging form +Thanks I have used your solution +Test harness for Groovy quick fix tests including tests for Convert to Groovy and Add Missing Imports quick fixes +Closing as invalid +So if the xml doc validates against the XSD then it is sure to be good for the parser +After creating a Queue and sending message to the Queue Count and Message Count are both +We have upgraded from to +But what we have in the code base is a +Wish I would have caught it then because I would have RCSometimes I want missing columns to cause the row to be filtered sometimes notIll try out this patch and see if it can work for us +This is because stores options in an unsorted hashmap +and thanks for taking that one into consideration +I think its more of an issue of there always being multiple columns even for what should be a toOne relationship +I think the API you have now is fineI think its silly to add complexity to the code to support to the RPC protocol for replicas at this point +svn diff of changes i closing all issues that have been in resolved state for months or more without any feedback or update +Just a little misunderstanding of expressions. +In general you should call services instead of directly updating the entities +The should reference the upstream site so that if something in SOA Tooling depends on common tests as runtime usage etc +The address to which the secondary is bound is now the same as that to which the primary is bound +For some reason no exception is thrown because monitor exit succeeds on the object which wasnt used for monitor enterSo I am leaving this bug open +Great Ill close this as WONTFIX then +Rajesh my apologiesI had to untag the guide because it was pushed under instead of +Just checking in on this +I can try to email to you directlyIts bigger than the allowable gmail size too +I like this quote from the GuardianThe words Dylanesque and Pythonesque are already part of the lexicon and sure to survive the former is synonymous with notions of authentic rocknroll poetry and the latter refers to zany behaviour that is not nearly as funny as it believes itself to be +The next step is reproducing the issue with fundamental test +Added capability to accumulate positional information when creating higher level constructs +Can we prevent this null from appearing + +But the pointer implementation also has a method to return the whole attribute +We tested TCK on +Should probably just disallow this case +column families which are sorted by the index valueThis is the hard part I am having trouble wrapping my mind around a scheme that allows both different sort orders in different and always routes things correctly by node token no matter which CF you are talking aboutbq +lgtm +Was just about to review this but unfortunately trunk has moved on a bit and now the patch is stale +Validation does not filter out excluded java files we should implement that +I forgot to specify in the environment database JDBC driver mysql connector of thousands of people use Hibernate +Sorry +I will commit patch v if no to trunk andThanks for the review matt ted ram. +Thanks for the patch +FYI karaf distro from trunk still produces specs v a look in mn a SNAPSHOT is deploying +The patch looks good so farId really like to see some tests for this functionality so Ill check if its feasible to write some before committing the changes +Moving out some bugs to +Heres the fixed patch from my comments on reviewboard +HiAlso I have noticed that the Save Completion dialog which is given by the above coed is shown when singlewebappsrcbasecontentviewss Save button is clicked +SimonThank you for trying out the latest build +It WORKSJust downloaded and generated a test file in Eclipse Europa Version id I am stuck in OSGI land using eclipse is a must Awesome +The one thing I would ask is maybe a status on where its currently at +I will create a patch that you can use to reproduce the error + +no label is meant to display string values that can come from codeas a rule of thumb formcomponent labels should not come from code but rather in +Example of mutiref serialization with primitive types as sent by AXIS +Updated patch that actually works +Correct +Commit and Enjoy! +I guess the best approach would be to design the code with persistence in mind and initially work only on an implementation +Verified with modcluster +Initial patch for review +As promised. +In this case of this regression Lily has worked hard to fix the regression and made her plans clear over the course of development +Uploading a revised patch b which relative to the first patch contains a simplified test fixture Regressions passed +That is a good point +Ok I see +Im marking this as VERIFIED INVALID +Feel free to untag if you are happy to make another release that kills or deegree running in the same Tomcat instance +Hi Weinan do we hae a new resteasy build that gets the new? If we do please pass this onwards to Marek so he can do the same for Seam +I think Dheerajs point is good and the Java and C clients are already divergent enough in this manner that this shouldnt be an issue +Patch attached to clean out the vestiges +Konstantin writes This is the test problem +is LGPL the LZMA SDK is notLicenseLZMA SDK is placed in the public domain +This is what I applied +seems if I put a sleep after deployment and before I do anything everything works fine +Apparently its a cglib issue upgrading to the latest version solves the problem. +Created a new issue for the point about error messages +I would prefer a different column that can show the host that the AM is running on +Full script isecho offsetlocalrem dp is location of current script under REALPATHdpcd REALPATHcd +Thanks for prodding Ill create a wiki page with the docs +I did all the tests last night but for some reason they were consistently failing at +Im thinkin of opening a different jira for hdfsproxy test failureIts been long time since the streaming tests on trunk is failing so Im going to commit this and let trunk run the streaming contrib tests +The other difference is that the two jars detailed below were taken out of the serverdefaultlib folder so that they are picked up from the war lib version version same deployer thread should also be used so the issue should have only been causing a leak of the last deployment +This one is fine +I was a bit to hasty in the previous entry +Mine is failing setting the Wicket version to in the POM +bulk defer to +I will try to make tests work + +Thanks for your imme reply ThomasI have attached a file which contains captured logI changed in as you told still Im getting same problem +However the warning messages shouldnt be produced +Also when you raise the new issue please include your eclipse error log. +I will add these parameters in the config xml file +Thanks Chao +The gist of it is that we do not insert the updates to the consumer log in CSN order and this might cause updates to be skipped when sending them to consumer +The only change would be the shell script to do the late binding to select the right jsvc to use at runtimeRPMDEB packages should only package jsvc base on the building architectureDoes this sound reasonable +Closing the issue now that it has been resolved. +I guess this bug is probably still valid for Struts since you wouldnt want to ship with a bad +A must have for us +Ive fixed the relativePath logic now +This patch applies cleanly to revision trunk Thank patch for GORA branchRenato this seems to work fine for meOf the tests we currently have running within all pass locallyGreat work on this one Renato BTW +Ron I think were definitely speaking the same language now +Patch complete +I was initially thinking of creating a plugin for this but the schema generator class does not seem to be extensible to be able to create a plugin to intercept it before the generated schema is written down to the xsd documentthe other option here is to hack the jaxb ri and add this functionality +This issue provides regression tests for +I dont see an easy way to merge those settings into if that was what you were hinting toYes thats where I was goingWhat if we dropped the main index and just kept the sample index of every columns? Seems like wed trade a little more seq io to do less random io and being able to get rid of the index sampling phase on startup +Hi I have tried reproducing the error on Windows using wenc that should be doing the same API calls you writeBut I get the right result can you try running the same command line on AIX so that we can verify the issue is in the Xerces codeThanks AlbertoI test on my AIX runtime I also get the right result but the error ocurr in my application yetI think the issue is in my application code I will examine it againThanks given that works I am closing this issue +The bridge is meant to be the minimum amount of stuff from Logj to pass applications over to Logj +I just closed +Thanks +Also in the event we might not be able to pull this jira into separate patches will allow us to incorporate the OOM jira in +Attached as archive as I didnt want to create a patch before we know where in the source tree we want to put it +I like this simplification! I agree baking in BQs relationships into this Scorer API is not good and its better to let the caller decide howwhenwhat subs to visit +You mention that the enhancement wont take place +For clarification the tests that fail have nothing to do with correctness but rather that some error messages get changed by the changes in the parser +Patch Peter Kriens BND tool to allow processing of include directives in properties loaded programatically +Added the serial to the and cleaned up the import list of the in Rev +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Bulk move issues to and +I have committed this +If you still care about this I cant see how we could address this without a test case +Running grails upgrade didnt upgrade the file +I am not using the aop pointcut definition but springs declarative approach +The problem doesnt manifest anymore on EAP ER with and Arquillian +have tried out the proposed fix on the file in question and it is working +Sorry for the confusion +Did you find any errors? Is there anything I can do to help out testing with this issue? +Mike reportsbq +Review comments will come in the next JIRA comment +This should not be very difficult +Is it a guaranteed behavior +Perhaps you could just delete columns that you dont know what they are but that seems a little iffy to me +Something like java +If the server is stopped when the journal is gone then you should either verify it is or open a new JIRABTW we have already fixed +Could you please provide a quickstart or +Added fix for the case when Filesystem name in df output is too long and is split by several lines +Meh this is still hacking +How stupid of me I coded this and when looking for why the source file is not around anymore I did not see itAnyway there is of course a configuration option keepGenerated which does just that +I will close as Wont Fix for now +Arun chatted offline with me +Currenlty hlog is a sequencefile +Yes I will close it +Can you update it and Ill take a look +For the jboss release we want to bundle jbossweb as the servlet container +Ill keep you updated +also readBytes should not catch +Duplicate of +Please use this BETTER implementation and discard the other one also adds the to the regardsMaarten Hogendoornmhogendoorn also +This default isnt great as it breaks the toggling between l enabled and l disabled +That can wait post commit thoughThat leaves the following critical path items to deal with Get the tests to run without a hacked file +I think what Im missing is the connectioninclusioncall tree between these different classes +I created network with cidr and was able to add an ip range with to it wrong kind of behavior +Its just that you are trying to add a mapping file on top of the already generated descriptor classes where theres no need. +True you do have to for version vectors but thats a pretty low price compared to losing availability if a single node is down a large amount of code for one new op typeId very much like to explore the version vector approach in more before deciding that we need this extra complexity to achieve reasonable performance +I do not think the isJaxb method ever worked as planned +No worries mate +Dan on the test tag theres a force attribute +However the behavior that is there is still bad +Patch to set ar permissions for all the jars +Logged In YES useridIm apparently cronically unable to submit files trying again +Thanks Dan for your feedback +you AngelaThe fix worksThanks again for your great work +Otherwise it isnt. +The releases from here on should be bug fix only +fixed in to CR wont be fixed for Beta now +Oups prior patch contained some other stuff too by misstake +If a dependency of the the project is a continuum project and it change since last failure then well execute a new build of the project even if there are no changes in want to do that because a failure can be occurred because of a bad dependency +There is this stupid issue with copyright and publisher not granting rights for that kind of stuff +I have tried several things to make this process work +Yes the height works very well +bulk defer of unresolved bugs to +A simple patch to indicate the config directory for the hadoop thank you +Does Chukwa need to select a specific uid? How should we handle that uid being already in use? shell binbash +Also can you try cordova emulate android verbose +Anyone working on this issue +Does this mean hadoop would be released soon +Currently is still compatible to JDK +I dont mean we would have two different attributes on each annotation +defer all issues to. +At the moment this is still a manual procedure +Here is a patch for the table problems in +I think if you want to do that youll just have to fall back to conventional bean syntax where necessary +this is the right version +ok +Patch applied with some changes +unixdos converting file +cant be true either +Currently feature has no version but I found there was a jira for adding versions to feature the other day and I believed that features with version make sense to me since users likely to operate feature mostly instead of bundles directly +Paul Easy test is browse to jmx on regionserver +Id like to summarize my understanding of the patchIf you have an EXISTS nested inside an EXISTS the optimized inversion of the query execution plan does not occur +changed the as proposed sqlMap is now optional dtd on website should be updated in at least an hour +Allof these convenience functions including your proposed enhancement couldbe refactored into a separate class or a subclass +Ok the original poster on the forum has confirmed that with he has no still think get should ideally return a value other than null if it can as per the above but that would be more of an enhancement than a bug since null is of course legal +bq +It seems to conflict with other patches that were recently comitted +Thanks Mayank +RPC timeout occurs from a node that is not verifiable in the trace outputWell +This Jira should help with that +conf tests namespace and namespace have the same problem +yes since there is no reason to think keeps lots of files open +Im inclined to WONTFIX this issue since the workaround takes negligibly longer than what p ought to be doing wo the added assistance +A better fix would render the error response in its own window rather than using the log function because the response is in HTML +This is not the complete patch +Hi Derek could you attach to this ticket your log file to be sure to have the full stack trace? Thanks +Then this problem must go revision Thanks Ryuuichi +Its and yes its against TRUNK +It was due to a change in the WODEN API +Closing +what do I need to do to make the ant task copy the from red instead of using the old one? I still see instead of as Main password attribute is still an interactive option maybe if the password option is already given in the command we skip the interactive question? That would be the same behaviour like when you access mysql via console +Thanks for the issue and patch. +The patch was ready to commit only need a testcase +Then you compile hdfs and push the jar to mapreduce +Having said that Ive seen enough applications from a very wide set of users over the years that I can see many use cases where going beyond our simple storage mechanisms would be useful and giving users alternate tools for storage is a good thing especially since retrieving stored fields is almost always one of the biggest performance killers in real world applications +That would work thanks +Hi Eli and Todd it seems that this issue has introduced a deadlock see +Imported the code adding ASF headers and Airavata coding styles and changed package names +Is it possible to write tests for this? I dont know much about JMX but was thinking it still makes sense to have testsIf not what is the best way to test? FWIW the latest patch applies cleanly and all current tests pass + Which brings up the question If were going have the doctype set this behavior automatically should we have a separate output mode for XHTML? I guess a separate output method would still be desirable to support custom doctypes which use XHTML serialization conventionsThe serialization with the WC XHTML doctype actually works as expected so XHTML output has been implemented but is not triggered by the xhtml output method or the occurance of XHTML or xhtml in the public or system doctypes +Hi Claudioplease commit it! eventual modifications can be performed while working Im confident you know very well the topic Best I just committed! +The idea is that handle the serialization +I think Ive hit some issues before by leaving those unchanged though but they definitely shouldnt be handled by the CCSID manager +Can you try the trunk version? Ive fixed a bunch of startup script issues there see +Committed this to too. +The fix went out into the release. +In particular if a udf or input format uses jackson the target library needs to match the version +It turned out this is a problem in our application +Committed with revision . +Wed likely bump the version one point only though +This issue is duplicate of +closing +A possible solution is to look into copyBatch in the interface +By prioritizing on split load we should substantially increase the locality of most jobsWhen I was talking about the list I assume it is really just part of the data structure in from +Thanks now the test passes for me too +Kicking off hudson again to run through the patch again +Committed at subversion revision +Increasing to Blocker +Feel free to this issue once you can supply us with something that we can use to reproduce this. +Yeah I was hoping to remove the Simulator discoverer too at the time and have all postDiscovery operations reqd +Hmm +I thought that was what the resources directory contained +I am attaching the patch to fix this issueThis patch fixes the sql query used in the SYSCSBULKINSERT procedure to correctly use the optimizer hints using clause instead of PROPERTIES +Doesnt it only affect CREATEALTER KEYSPACETABLE statements that arent exactly large and arent usually prepared in the first place +Did you build the patch for the trunk or theIt would also be great if you could write something on how the patch works and if you expect any side effects +Let me just finish up this patch +Lee Surprenant posted a patch for that provided instructions for how to run the standalone calc sample +core tests +Apologies was a late night yesterday +The problem you are having is that a rank deficient matrix doesnt have an inverse because some of the singular values are zeroIn the Cholesky decomposition of a rank deficient matrix each step proceeds by dividing by the next remaining diagonal element +included in. +Ive updated the wiki pageCheck if what I wrote its correctThanksDenis +Merged patch with latest trunk +I would like but I am stuck with java +Go ahead +we should also report current block cache hit ratio it will save us the hassle of deriving it from the aggregate counts +from testing +Looks like this patch has the same locking issue as before +We have a property in our build scripts called version that can be overridden to specify the appropriate release +The high ram job processing is a much better model +a way to control the daemon programmatically perhaps via JMXI love JMX but everytime you use it a Ruby dev kills a kitten +Done at some unknown time in the past +The two first bullet points are according to the sources the target default value is and the source default value is for jdks and no source is set for older jdks +I dont think it will be particularly difficult +Ill assign this to him +This could be a tunable parameter if needed +as i know it is the dependency jar and the dependency ejb points to the same file and thats why the first one is enough +Great stuff! David can you please take a look at this JIRA and let us know whether you think your work will take of the use case it talks about +I will install today +Please note there are a number of optimizations to be made in this code +Can you tell Ive got some time for Mahout lately? bq +Latest patch looks good +Thanks for testing RayIve committed the fix in revision . +Override has removed +Lacks a test case or other notes on how to reproduce +My guess is since it has been so long there are surely a few issues +You are right Yonik +Fixed with the svn revision +The issue dont exist on single cpu systemOn hyperthreaded system if i bind java process to cpu there is no issueConclusion This occurs only on hyperthreaded system with no binded have exactly the same problem is there a fix or workaround avalaible since GBT what do you mean exactly by if I bind java process to cpu? Thank you. +Have you tried to see if anything bad happens after switching? If not +Updated patch for content component +The workflow submenu is closed for thanks for info I have closed it. +JPA processor currently ignores those Java methods that are annotated with +This is the html report file +we were running with this patch and had problems with streaming scripts being set to +I just upgrade it from to +Not sure whether to leave this as resolved or closed so Im leaving it resolved +This will mean that people who have not already installed will not have a problem but people who have already installed will have the same problem that you have +should help a lota new statement that lets renaming some of the fields would be helpfulthat said a new operator like that means everything else intuitively feels pretty confusing an error prone +after my testhive and also have this problemsso I create patchs for them +Attached a patch +qmx if needed we can talk with Wesley about the updates as well he said he hacked it for the video but we need to get it right for the launch +This should be fixed now +No it is not a duplicate +the homegitappname is already checked for I believe! if not then that is surely a bugThe check for project name on import is trickier since we cant predict the full nameproject clashes before actually doing the import +Thanks a lot Mark +I added you as a contributor to the POM +Design and Implementation details need updated on wiki for Java broker +Any idea +As part of formatting a NN with the newCluster option it will generate a unique and a unique which are persisted on the namenode +It seems like the prebuilt jars in the release are containing the file but it is missing from the PIGHOMEconf dir so that my custom built jars ended up missing the file and thus the tests failed +single file now +is compiled into the class tree but it isnt put into any of the Derby jarsThanks a new version of the release note which hopefully clarifies the point Kathey raised + +It throws an error off course but this error is caughtDenny thanks for filing this bug! Without it I wouldnt have known what was wrong with the admin +Mimicked the changes from and verified and pass +Thanks! +Backported change into codeline with revision +since deferred to any chance youll make it for +This one patches a lot more options than the other if this patch is +I think this can also happen when the scanner crosses a region boundary +Do you mean in terms of indentation? Also every now and then keeps failing for which there was another JIRA +Please reopen if this does not to the trick +But I think that we should look closer at using for accessing the Directory instance for the index +Attqached is the file workspacemavensrcmainjavaorgapachemavenrepositoryinternal Antoncould you attach your fix as a patchdiff fileif you could write a test too we would really appreciate it +Can you provide an example +We do not have this for all docs +OK nice +Otherwise +Sure Andrzej +Ivo can this JIRA be marked resolved? There exists for documenting this. +Is there any chance that can be addressed sooner than +his workstation is also windos +Hence link the related issues together to solve this bug once and for all +The test cases that fail is common in +here is a testcase showing the affects I agree we need to fix this +Attention MiroslavClebertI have written a Release Notes entry for this issue +please file a new upgrade RajeshSame issue is observed in both the setups +Will attach a patch for this in sec +Then I have another idea of building a on top of the normal to fasten it upIn a segment queue instead of storing messages in just one queue you store it in multiple segments then at the time of reading +Committed revision . +Just for the record describes how to fix this issue. +Here is a patch which addresses your fb +Everythink is working fine thanks! +The attached patch implements my suggestion +But specifying the default from address like zaratine worked for me +If nobody objects Ill commit it +In the attached patch for we check if the connection is closed every sec +According to the codeselect id from cloud +It seems the secure storage is not working properly +I havent read over your function tag patch yet maybe thatll help me understandFrom your unit test it seems like you want to create a new thats blank but contains the parents tag cache but no variables +Committed patch as svn resolving. +It needs to be consistent with the way Maven does it for version ranges thats why needs to query the metadata you are using a pom with two repositories you don t know in which one is the dependency +Also verified that the Clear Metrics and View Received Call Time Data operations are working. +Hi SergeyI will keep this issue open until we get the integration test committed +If I understand this correctly thats a duplicate of +Fixed in rev branches +Thanks for patch Chris. +I just need to see whether it breaks anything else +esbtests +Also if executing AltJG in a JAVA project and switching the user should be offered the option to add the groovy nature? thatd be cool +Can you check your project to see if the java files actually got installed +Just to be clear Im concerned only about the LOST updates which are sent by Mesos not by my framework +This action will now just get the execution plan without doing a full execution of the scrapbook query. +The tests do cover these cases since I had to fix them when I ran the suite +As indicated in the earlier comments this error means an attempt to access the coordinator failed because the corresponding CORBA object has been destroyed which happens when a transaction is terminated + +You will now need to select a subforum +The infrastructure is in place for supporting expressions in any persistent configuration item and Im satisfied with the places where weve chosen to make the model attributes support expressions +MAC OS is not supported with and there are no plans to do so. +So from this issues description should be changed to the following the classes in lib from the package also need to be made available in some jar under the client dir since they are mgmt client API classes +If you mean youd like to implement that I can leave this open to track this +Where does that parameter get put +I fixed this in revision +It looks as if the files wouldnt get picked up they should be automatically included by placing the into your webapp lib directoryCan anyone of our other users confirm this behaviour with the RIregards this is another problem the constructor of the extended dataTable didnt set the renderer type +tnx for is still happening under example above works fine under do you have a other example that example working fine in my test +Ill now debug into it +So if I remove the encryption introduction from the Guide I lose the concept chapter for the next security chapters +Is this repetitive or the metadata starts returning consistent data after some time? Since the metadata is communicated to the brokers by the controller it is possible that there is a time window after an event has happened and before all the brokers have learned of the event. +Tim BJThank you so much! +Fix a TODO from the previous patch which caused to fail +Hey Per since you probably still have this in a check out might be faster for you to update it to trunk than me unless Im confused all of the files have some small conflicts nowI can get to updating it if not but if I can get you to do it I will likely get to committing it faster +Everything on the same machineSo I can test the installer and see what it detects +Is there any directory that file can be locatedIm using GNOME version build date s what I did Right click on web choose Run AsEclipse Application +Ive added a fix for this issue in rev BTW if we are going to move to Java are we still going to use mxj or the jmx implementation of javaJMX supposedly has this query mechanism to perform better object name querying and I want to use that instead of the regex thing +Is it difficult that Hudson and Jira are integrated +Hi Olegi was hurried topost the patch and i havent revisited with more accuracy +Patch updated for unit test failures +My edits attached +Theparameter object may change its value asynchronously but such changeis never effected directly by the HTTP clientParams objects may be modified by the HTTP client namely by the settermethods that store the set values in the params object +Reupping patch with proper path and naming +Looking a bit closer at the class it seems pretty tied to +If distributed indexes are going to be exposed as column families specifying a name would be reasonable and youd need to indicate the indexed field somehowSince adding via the config file is deprecated anyway parameters to a Callback could be a stringbytes map to be passed in via Thrift +Need to set the default locale to anyway + Sorry +The mapping works iff a process is running when the ip changes +is going to fix the issue +This webdav lib is exactly Using instead which is the lib which gets downloaded from the maven repository if you build vfs locally solves the problemSo I would recommend to upload this newer webdavlib to ivyrep and make sure that when you build ivy this is the version which gets downloadedI will now test to see whether the source code change in the vfs repository in ivy itself is necessary or only good to haveRegards +I should also note that I changed STDEV to be a function of the TOTALKEYS +Still have not seen a test case reproducing this +Olivier Any update on this please +Struts added in the previous Struts group +In my opinion the way it is implement is good +I will upload new patch shortly +Code generation architecture in some sentences Hibernate tools are stated as a part of JBT +. +Please provide sample project that would allow to reproduce this issue +unit tests and webhcat ee tests in was also changed from to patch will need to be redone after patches for the patches Eugene +I think I screwed up a conflict resolution +There is no message or indication of any kind that that is what must be done +This is probably a problem with missing request parameter information +Verified status of this issue sure I am not sure about this issue maybe still need to be opened +Is there any way to do this even when the provider info is not set +If you look at the patch I basically redid some of it following the existing organization and content in addition to fixing small nits +Fixed in TrunkI changed the and to use a internally to preserve insertion order. +Do they have a use case out of function queries at the moment? If so then yeah Ill happily put them somewhere else +bq +Fix version was since I close tag Jira move all issues to +its highly suspicious +In workspace workspace Create a ESB project by ESB project wizard with server supplied ESB runtime and then there will be two classpath contianer one is jboss as server runtime cp and another is ESB runtiem +The Index is used to represent a not found tabu +Committed with revision +Sorry for MY ignorance for not remembering this fix +Thanks Carl for adding test suites +Any thoughts +defer all issues to +now saves the old and it when the is done +Still affects this error too with maven and beta In my case I run mvn releaseprepare in one run mvn releaseperform with Dtag and parametersIf I download the sources from repository and run mvn releaseperform DtagXXX XXX on level this error doesnt think this only happens when you run mvn releaseperform scmsvn +Both these issues look same to me +Do I need to submit a separate patch file? Its a fixyes +An assign activity usually runs just for miliseconds +Any other comments before commit +since not all are fixed yet. +Ive actually bothered implementing the semantic versioning server side so that when for instance when we have cql if the client asked for but the server is an old one that only support an error is returned +Took a crack at modifying to use Springs abstraction +But I think the overall idea of allowing the user to select the type of grouping is good +I suspect these test cases from the OASIS suite also demonstrate the same bugWhitespacesWhitespacesWhitespacesWhitespaces am using Xalan but on further investigation the bug appears to either be in my code or the parser +Ive attached a patch for review + +can create its block copysitsitsgOtherwise looks good and ready for commit +reclassify as New Feature since its a code contribution not a task +Hi EricThanks for your patch +Closed +Ultimately well be using LDAP as the primary for all users +As long as I get the ability to not go to ibiblio Im fine with this beeing a feature I do see the value +are you going to give a official fix about it? This problem could be encountered easily in our environment +the file tmp is a combined patch of and A small api change in the query string creator required a small code change +hmm after a closer look at the rendered source everything looks fineHave to be another issues somewhere +I usually prefer seeing the standard web page +Im going to still leave this open as I believe we need to make a pass over all javascript to make sure we arent losing jsessionidsill cover that bug under this could you please review the desktop for session tracking +See my last comment Notice problem occured on Tomcat on this platform too. +I am running the following file from the module project +Emconf file for reproducing the issue is attached +Reopening to write release note now that release note text has been written +I agree that seems like an overhead +Please nukeem from trunk +I have attached the patch for this including some testsI added the option as allow following a similar naming convention of a different componentFor documentation purposesName Default Descriptionallow false Used to specify if a null body is allowed during file writing +s comment updated in patch +I keep threatening to commit that patch only because i think its more legit to have this real release than from apache labs +Well we can not remove it is all over the code but we found a way to remove it from parts that need the serializationSince we have to interact with foreing code changing the serialization format is actually a for usBut for all the other people stumbling over this problem I have prepared a version of that has the order of the calls changed so it works with Application Server To verify the fix +Committed patch a to trunk with revision Ill leave the issue open for a while and backport it if no problems are detected +sounds for explanation Snjezana so the IU in mylyn connector are only feature IU I reviewed the discovery and for this specific part regarding im fine only gotcha is that connector id IU which I thought we would move away fromAs long as we keep the connector id stable no matter if atlassian changes their feature names im ll open separate jira with comments from my discovery review +But the idea is not for frameworks or apps to subclass Implementations of it +Im not sure whether to backport these changes +ack I reverted the change in revision . +Thanks for helping out Tony! +Heres my patch for website +Actually we are definitely going to move to Java in the next unstable branch +Shortly after posting this issue I created a class to do what I needed +bq +Thanks Kihwal +One thing I request is that specifying no prefix on a role name be supported at least syntactically +This looks like it is working for now but still isnt working for +Im not sure how this could be fixed in Surefire without changes to Maven core +The slowdown for smaller sizes is an issue and the modest improvement for larger ones is a surprise +in fact Im surprised you can configure that table at all +I fixed a few version numbers and removed the TODO on the amazon option after I added it +Bug was fixed during exception handling redesign in JIRA test works OK and this JIRA can be closed as fixed + +This issue not any more relevant for current trunk that was replaced by branch +good pointI will report to NN using errorReport interface +works with Nutch m closing this issue as the tool is not general enough to be included in Nutch +Set correct fix version +Thanks a lot. +I have no deeper understanding about the meaning of these batch files but it works now +Since the Development Phase is over please move the Jira to Resolved status +Commit introduced a bug +Great stuff Toddbq +Committed to trunkThank you Russell! +ah got it +If this issue has been rescheduled to it is because we currently focus of our effort on JAXWS +There is nothing special in the project +make sure you run the tests when you checkin +Ok Ill provide a patch in a day or two +This patch is a new for POJO SchemaApply is to airavata patched main providers is replaced with class from POJO Gfac Schema +Thanks Samuel! +Fixed at revision +what does mean? What results? what is the full stacktrace +you can find the bean methods on the old perisistence managersBR +For example if we have two portType elements with same namespace but with local names differing only in caseMy understanding is resolve works if the resolution is needed between different kinds of elements like between a service and a port type element AND their qnames are identicalThe test case fails to satisfy both the conditions +Hong will attach an updated version shortly +Issue may already be resolved may be invalid or may never be fixed +New packaging coming soon or you can download the source and build the integration module from that +Committed revision +So if we could set roleadmin for the special write operations that would be cool +patch looks good to me +Ill take it I was the last one in this code +But besides the unit tests we are also in the same boat as Christian our services are distributed as a JAR so we rely on being provided with a context that loads settings from a properties file in the classpath +Checking to see whether theres a cleaner way to implement this +Duplicate of progress +Hello Gregorywhat is your specific problem? I am using the plugin with Maven all the time and have no problems +Done! +The problem explaiend by aaime was that the GBIF datastore was lacking a primary index and thus could not be paginated. +Chapters are updated +We saw this all the way until early December last year always on Linuxvmware combinationsHowever I have not seen it since we moved the vmware machines on to faster vmware servers mid December +Hmm its been quite a while since I reported this so Im a little rusty as to the exact issue +They seem to do it the same way httppapers +Mass close all issues in resolved state that have not received further comments in the last month +Continued fractions are a very efficient way to approximate doubles +committed +Nicholas could this be due to the recent refactoring +Many thanks for the patches. +I would like to see Portlet Window states and modes persisted +This patch uses to get a list of corrupt files periodically and fixes sourceparity blocks automatically +The patch looks good to me +Daryn why are we trying to provide incomplete support for back slash based paths? Why not just remove support for it +We could continue discussion of that over on s do you feel about addressing these issues separately since request mapping and view resolution are independent concerns? This issue could be solved with a unless you prefer a different approach +other than that +Marking closed. +Manually tested +For some reason the fix does not show up in subversion commits but was committed at rev +contrib failures are in streaming not +When ODE is shutdown only one dialog will be displayed and on MMC there will be a notification regarding the shutdown +Here is the patch that incorporates your review comment +Are you sure that the job is not retired? I strongly feel this should not break because there are many unit tests calling this api +I dont know if we support solr urls +We would need to store the current role for the session +jars +After the subsequent test cases no longer fail +perhaps there should be a way of making the editor sticky with respect to some tab +In the jboss osgi runtime the webconsole works without the import +My goof up sorry +Closing all resolved tickets from or older +Allowing multiple deployed Application classes is going to have to wait though until Resteasy d backmerge it to AS but I dont see a branch to forkmerge to +Not after another commit not after restarting solrAll the old files are necessary for the index to work +Against trunk running the attached started withjava jdbcderbylocalhost I see the protocol error Dag described +Thanks passing now repeatedly on my old mac +Bug has been marked as a duplicate of this bug +Looks reasonable to me +Trunk version is also affectedOk lets wait for someone who is responsible for the last changes + closing +Whereever you would have expected itBut changes readme and javadocs sounds good. + for runs about m on my laptop +Testing can now be done through +Kelven can you mark this bug as Fixed once the patch has been pushed to branch +Theres also content sniffing if it looks like XMLDont forget content negotiation when you get applicationatomxml because you asked for it explicitly +Maybe add a constant prefix before QUALIFIERFor example htttpnsThis would allow adding pages to our website under ns to describe the namespaces later if needed by reserving this ns website path for this purpose +Therefore this operation should fail if we do not bind to all interfaces +As it happens I only use it once in this patch +In fact my temptation was to request multiple concurrent appenders +Can you maybe post some results from actual runs? Number of splits read by each worker and overall length of input superstep are the two obvious metricsRegarding +To test it open the commented out section in ecommercewidgetcategory screen on line +Assuming that the test cited by Bobby also works now +Patch looks good to me +Dont think we have example on this area for M release so will comment the configuration file for M release. +Committed r +a command message is a java object +If its identity then it needs the same object +Please note that the patch requires also the patches from the blocker tickets and +But I still believe that as long as ahandler is registered properly it will be called eventually +Decouple vertex implementation from edge storage Clean up configuration options read is always reporting vertices and edges Clean up directories created by test suite Remove Cruft Log message for aggregators has cryptic wordingIm not sure about Port Giraph to agree alessandroI think the port to yarn can wait + +Now the question is Why doesnt it work with bash? Anyway feel free to use my account for testing if you need it +Serializing Java exceptions only works for Java I think youve more or less got to coerce the exception into a string +A test case and a patch are attached +Sorry perhaps this should have been under the Shrinkwrap Resolvers Project +Im going to commit this very soon +I cannot reproduce it anyway +Got carried away with drop constraints because of work on another database that required it at one point +For example the tableName is already in the regionName of the +Changes committed in SVN revision +Alternatively we could prevent shut down of replicated databasesI dont think thats an option since that would break the shutdown server functionality +Yes what you are saying is correct +For the the standard ordering rules defined in will take place. +Ive just deployed a snapshot into the maven snapshot repo. +No need for integration tests. +i use spring. +In my recollection there is no NPE anymore but dependency is not resolved anyways but I need to check that again to be sure. +I thought that streams provided a pretty well understood mechanism to achieve this while being flexible enough for the some communicationI need to think about how we could structure generalized push support. +This made the warnings for the JDBC stubs go away +If all the instances are closed the desired implementation can be returned +Maybe that can be changed +The problem is that the way to bootstrap an sca domain has been completely changed by the looks of things so the initContributions method in doesnt work any more and looking at the Geronimo code it looks like it has the same problem +Correct me if I am wrong +Its a minor missing piece in +This normalizes the USERNAME arguments of the NATIVE procedures before using them to key into SYSUSERSI am also attaching a script which you can run after applying the patch +we can take a closer look at this for but its part of the original design of Maven that repositories are aggregated from dependency in order to help search for the next level of transitive dependencies +for +Ill have this check for name conflicts between attributes complex types and elements +Im attaching logs +too +not sure about the human readability but looks betterThanks Mate! +Backed out +Could you verify +Only the new Working With Derby doc fails to build on Linux the other Derby docs build successfully so perhaps theres a subtle difference in how the new files are set up +We should target with this +Two small changes +Ran system tests and all the tests got passedTests run Failures Errors Time elapsed secTests run Failures Errors Time elapsed secTests run Failures Errors Time elapsed secTests run Failures Errors Time elapsed sec +Initial patch for review +does not seem to be related +Patch for bnd OSGI metadata generation in for spotting this it fixed now with svn revision Well need to make a new minor release for these changes to take effectBTW +Thanks Julo! The comment looks useful to me +BJ Would you clarify what for is used your patch It looks like doesnt resolve the did you try? Is it working also for you +Attachment has been added with description Unified patch for and everyone sorry for the late response I was not aware that a ICLA was required for a patchI will gladly sign it so how should I do itA scan of the signed ICLA attached to this bug will doRegards everyoneIve just submitted the ICLA to secretaryThank you all +As per the linked jira if we could guarantee retries then we could hold off on the dialog until the second try + to trunk +The patch for this issue with test for it is attached +That was only for the summary I see now that you were talking about the actual description +applies to +Patch for +changed model color to resolved Release +Im not sure why Ant is more portable than Eclipse +I just noticed you had eliminated execute method from interface that takes as a parameter +so it gets assigned as part of the SSH bulk assignShould not cause a because that was our including in the bulk assign regions that could have been in OFFLINEPENDINGOPEN +Being the primary author of which added the priority compaction queue Im interested in knowing what the bug was! So what was the bug? And what was your fixI like the improvement and especially how it makes it much more granular in who should get priority. +fix reference +IW keeps its own reference +I have committed my changes to the trunk +The test cases were accidentally missing from my Eclipse include path when I did the search for that method +The upgraded version of being used in is more stable if no slfj implementation is provided +Done + +Bulk close for the release. +Of course different parameters might be required when preparing the call to different stacks tool inner implementation +I beileve this is invalid and that the problems are caused by not using the poms from the specs +We can weed out any issues we want to push to a later release from this set once weve done the consolidation +That was a bit of trickeryIt uses original INC and DEC rate values for without a in their or with unconfiguredPlease comment +Perhaps your Solaris box has less physical RAM and is swapping a lot? What JDK versions are you running across eachAFAIK is not run inside the dependency resolution code +Would you please help to verify it? Thanks a lotImplementation Details +Please let me know if anybody has any concerns about this getting committed as a contrib module +We are working on fixing these issues +I would certainly appreciate if the second issue would be handled as suggested by KnutDerby doesnt need the type code so ignoring it would be OK +So cast my vote for organizinggrouping files in a hierarchy under the demo folder +I think it was but not sure +We dont have any way to solve this issue on Sonar side so Im closing this ticket wont fix and add a link to. +Closing resolved Release +Committed revision Thanks AkshayI added an experimental warning to the output of the details command +We are looking for a mechanism to validate location provided by HDFS +Marking wont fix as part of triage. +fix ack +attached is a patch to fix the references on the solr site +Well I dont think we can guarantee an accurate result if we start pulling InetsSubnets from the new +Reopen if symptoms reoccur +update patch to reflect v of the patch now fails to apply cleanly to trunk can you cut a new patch for us Sylvain +I havent yet looked at what needs to be done for that +Ill take a closer look next week +when I added this in trunk I was not sure if there is a usecase +I tried bumping up the number of retries on this was only ms ms +indeed +layer is used by the WMS code as one of the artifacts represented in a WMS capabilities document +Done +Marius assigning to you as discussed previously to coordinate getting a fix done talk to ALR Carlo and Jaikrian about exactly how to implement this +I wil try these out +Currenlty only DB and Oracle will take advantage of this support +RTC work item close of resolved wont fix cant reproduce etc +The syntax for loading a database as has parentheses in it +Also updated eclipse version required to so that it is with the dependencies for the geronimo eclipse plugins in dev tools to prevent things from being overwritten with incorrect versions when working with both in the same maven repository +We will need to review and plan for portlet support in a near term future releaseNickWhat we need to get this working in portlets use module pattern in JS already done overwrite CSS class names add portlet namespaces to them affects renderer code CSS files overwrite CSS class names in this is I believe the trickiest partEither we overwrite or externalize class names into hash passed into module pattern constructorThese parts affects almost all components and I expected this to be covered by unified approach backed by please update this jira with the latest information from your investigation and resolve when you feel compatibility is good enough for +This is caused by a bug in Solaris basename uname rs basename math +We cant do much about this since groovyc itself does not support this and we do not want to break compatibility with the compiler +A start would be to acknowledge whither all Exceptions were being raised when the case is being satisfied +Please try to apply it and run the SSL tests it should show the test failure Im seeing +Now that work is well underway with AS all previous community releases are + +Forgot to add Xplanner works on same setup except with Tomcat at version +I assigned this issue to you +The renamed api was only available in so no deprecation work required. +Its fixed in +Im not familiar with but if Im reading it correctly it seems like this is more of an improvement request than a bug +Merged to the branch in revision . +Changes are to fill reverse relationships during server to client object conversion inside +Also fails to start on version of Eclipse for Java EE on on Windows and versions of on JBDS +This is out of date now. +request to be added as a contrib in Xircles and well assign the bug to you +So did you only plan to do it or is it a bug +sIn s decompose methodIn s decompose method +this issue to pmuir. +With in the chunk size is now configurable. +I was unable to update the issue immediately due to the server outages +Your fix does appear to resolve the issue with the in +skh targeting at the callablestatement logic to work with named parameters +Changed code to work as suggested +for more info on its usage. +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Also fixed the test to work with the latest change in trunk +should work with the latest Deltacloud server +There is an API calls for that now that should be used by the form rendering in Explorersee fixed in new Activiti Explorer +Thanks for telling me about you see any problem with jdk builds? They seem to me fine +We use such a class in our product as we are not on Java yet +Issue came up when i was making changes to for tab view but when i tested this after committing my tab code on fresh code from SVN i cant see this problem +rev trunk can the patch be applied? Thank you +The added foreign key indexes killed Sybase +Thanks you saved my time +Is it possible? I will begin tomorrow with the information that I have please send me all that you have related to this topicRegards plug in createdThis plugin reads the generated in and use it to create the rules and the violations +Applied to trunk and +It looks like the difference was http vs https which makes sense now. +patch needs more workBeside unit tests what work does it requiresThanksVadim +Is already in released! +Local variable for df +Removed the static nature of loading the system functions to enable the configuration controlled loading of the functions. +If the corruption happened to a checksum on a datanode then it does +I ran my test on and +This was fixed in version of to swallow any security exceptionscaused by this so the fix is in current nightly builds and will be included inthe next released version +Adds license headers to files that were missing it +I mark this as fixed with no fresh news on the fix +Back dating these changes notes would be extremely difficult and in my opinion not a good use of the developers time +If someone wants to fix this in please reopen but this has been fixed in since there are no more lazy static initalizations to screw things up +r for the basic mods to the ant build and the database driversinterfaces +in any event i think im going to leave it as a bug because its deprecated undocumented and a quick fix +As suggested code now checks that we are in a before casting +no longer relevant to trunk +Since the history information is recorded and used after the completion of an application IMHO we just need the after run log +the patch is really good +Thanks Jonathan very useful feedback +I just dont want to be stepping on any design toes if there was a reason to not delegate in the first place +New ones can be opened at any time closing +These patches are identical except that I added the synch around sendPacket to make it iterate over get which was not in the original change and Im not sure if it was an oversight or not +Working on it +Same patch different name +ping? it affect +ATM we cannot guarantee it will not change + +I realized theres a much cleaner and more efficient way to construct the parent dirs perms +Did you try start and stop out +If we decide not to do well change it back. +Ok I found the problem +Alternatively we could augment the generated with annotations which can be used during the SDOXSD generation phaseWhat I have put in place is the basic mechanism and framework to generate from SDO types +This fixes this special case and the test on trunk to also hit this I commit soon +Looking at that right now +For example +Thanks +The dist target will fail if they do not +Hi MikaWhich is the expected result +We would still check them for additvity +It is of corse related +Ok but I am curious why it works for a non static method and only shows the error on a static methodAlso could I write a DSLD to overcome this issue? I wrote a DSLD for the Slfj annotation which seems to work for methodsm experiencing this issue with all annotations including Logj +Thanks Alejandro! +We should go with an approach that generalizes for all schedulersQueue submission are enforced internally by the schedulers on addApplication +I dont like Exception because it is and will give a naive implementer the impression that they can throw anything up and itll magically be dealt with +bulk defer of unresolved bugs to +im utterly confused now The filter is there to make the redeployment sane! Without it deployment happens constantly even when file changes which should not trigger a far as I know this filter should not be removed it should rather be must be an error that it is not in AS +Great news +Not at this time no +Ok here is the few problems if no replication options where given NPE is thrown + +The document looks pretty Is there a need to creates something extra +OPTION HTTP will be use for that but that wont happens in this release +Rebased for changes in thanks. +LGTM +The documentation should note that the user must create a new for every +With that change I have no additional comments on the proposed checkin +Looks like the Java backend does not generate generic bridge methods for operators +notifications sounds great thanks much Wendy +The check is not working property for new also +I totally agree but we need to start somewhere +Typo Section Second sentence It contains a couple typos and I think it should be two sentences +It appears to be tied to the time the upload takes to stream from the browser to the server +Running test patch now +Like we should also remove all the version from the javadoc as they are pointless alsoSo I suggest to work on minor code cleanups for +Thank you for the review. +Fixed now +Thanks Zhenxiao +I will try to fix it by picking up some of the builder this bug into maven +Account locked +ah okthen Sonar should not inject target from compiler but source can now be analyzed without any problem +My delete without prompt preference was always on +Thanks Lukasz keep them coming. +If you want this changed for reopen this report and supply a tested backwards compatible patch file in cvs diff u format +Under facelets with jsf works of a bulk update all resolutions changed to done please review history to original resolution type +Daemon thread processing added in revision +ClintI believe that with this latest patch that your problem should be fixed +Point may involve more workIf you are not able to find the time to do so then I am more than willing to help here +Just a tiny suggestion I would extract an costfunction interface that can be supplied from outside and reflectively instantiated in your BSPSubclassing an algorithm is not so good for everybody + DONE +Since this is an Apache product this decision could have been done like for Apache Tomcat +Check nowFix confirmedColm. +There is doc on just using the dependency plugin but that leads to an extra copy if Im not more confused than usual. +I totally agree +This is the patch which will resolve this issue +Thanks Ill take a look now +Am I wrongWhat of datanodes do you think logged exception even once? If avg load at any time was able to cause this problem then we would see a large portion of datanodes to have this exception in their logs +merged with the latest in trunk +mvn Pshell clean packageWhen these Shell bundles are installed the Web Console has a Shell tab which allows access to the command shell through the web console +Me lets close this issue and keep the status quo +Resolving! +It all points to the fact that jaxws annotations are not imported +Ill write to the ML for this the problem is likely larger than just cleaning up +Ill check in this test with a name refering to this issue +Test case that shows the problem for both and +My guess is that it might be the svn tool you are using since none of those names should be breaking any Windows character limit +Thank Raymond for providing the patch +Added which is a very simple Aggregator implementation +what is current language Information the preferred langauge according to and possibly an overwriting cookie +The policy on this isnt really finalised yet +So far what I am doing is collecting logs from HDFS for future use +Yes +Therefor I can not later set the Xid +To put a fine point on that ensure that the refactored code is bw compatible with the original code +test error is unrelated and unreproducible +When fixing the earlier bug I have assume that pojo also maps the date to xsdatewe will revert the earlier change. +I am not saying you shouldnt have fun but I start to think I shouldnt have reported this +I am closing this as invalid but please reopen it if I misunderstood the issue. +of SFSB might become available pending resolution of. +I will wait for RC then. +I only renamed the list from parentDocs to ancestorDocs before I committed the patchsvn revision +META +Documentation has to be updatedI will create another issue forNow packaged as a single jar +Based on Nicks comment Im resolving this for Alpha. +I have delved deeply into your code and found things that need fixing +This has major problems we could miss a row and not be as live as we want +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +I tested this independently at homeIMO this is a bug that this is a bug fix +trunk uses only Could you please also upload a patch for +If we force clients to choose to push docs to a handler that is OOB from replication then we alleviate such concerns +Birt Plugin Reference Guide is checked +Patch looks OK to me +Now with for toString +no audiovideo option at interview room is useless maybe it should be remove + +Will let a buildbot run and commit if theres clean tests and no objections +Would you mind giving this a try and report back if you still see the issue appear +Column it does not work smooth in all cases + + +There are some integration tests based on Selenium which will start processes for the Selenium server as well as inline Jetty server instances +Ill leave the other issue for others to comment on +Controller is no longer maintained +Will attach the command code +Thanks for the patch. +Graham its been over three years since I put forward this idea +Thanks you Nicolas for your patience and your reviews. +Could we combine them into one utility? We can throw a generic with the right text +I did the import yesterday and just noticed that Im missing resources now I looked in yesterdays logs and got no exceptions like we saw before but the only services I have imported for my JBN Server is Statless Session EJB and Entity EJB +this is the minimal change that will avoid the problematic seam bruteforce conversion to be done on seam thus for this +Steps to reproduce needed +fixed in svn added a few performance benchmarks for this in revision +The settings are restored at startup +Bulk closing for +Step by step instructions how to reproduce issue would be also handy +With the approaches I proposed the bundle is the webapp and if it is installed by reference you can work on it and test the result of your work without needing to pack and restart all the time +and Castor dependencies upgraded +Closing all issues that were resolved for. +just infer the type cmon +The fix makes a group start or group end span accross the whole row rather than just the one attached my solution and differ fileshope it posted solutions would not work for me under +HarryIve just summarized the remaining issues +Thus while I believe most of the patch works the Solr tests will fail +The XML file with japanese unicode data is attached for can you please raise such issues with the user list first? On the user list theres plenty of users that work in international companies and thus can provide you with a lot of experience +Right after that i can view that pdf file +main question is whether this should exist within or least for now the plan is to add these adapters within in an integration +Version of the protocol will provide specific error codes for clients rather than rely on exception names +Typically this does not affect production so much but it is annoying for developers +Email is dan inside JIRA +We get away with it currently because of the way the tests are run but its best the test cleans that up will attach a second patch momentarily +Yes I see that this test is ok but if I start process from remote application on I get always value in process field of Task table +Re using setget as an extension point is this documented behaviour somewhere +Which I guess might be ok +Apparently test failures are unrelated they were exist before the patch application +The error file not exists continuously exist and backup command doesnt move database to different location +Could you do the following and retry +Without further information and since its working for us Im closing this issue with a wont fix status +I seems to work on JRUBY master on my machine +Ive disabled resource listener and reimplemented checker +Gurkan you closed this issue without committing the resource bundle patch +Then we saw this problem +patch applied +Made a commit NOT using the attached patch +Hi EvgenyThank you for your answerPlease move this post into SONAR sectionAdditional idea It will be convenient to go down in project by name and see dashboard by +you Peter +Ignoring the HTML problem because HTML relaxes the rules is not acceptable +I think the fix is to remove the smart in the brokers handling of a become follower request +I didnt realized that there wasnt any getter or setter until you told me +Committed at subversion revision Touches the following fileM javatestingorgapachederbyTestingfunctionTeststestslang +Committed to maintenance on. +avoiding optimized fp instructions and making sure things are always precision rounded by going through memory +I wasnt suggesting they be deleted yet just trying to understand how useful they are when the user comes to reinstall the same or a newer version of JBDS +Guillaume I committed a fix attempt at this in revision +Now works too although I ll simplify it so we dont need a different war for jboss web and +Seems the initial creation gets stuck looking into it +Is there any further context information that you could provide us with? that Juergenbuild appears to have solved the problems we were having with webflow in +Ive not looked at the code so this may be nonsense Perhaps you could make the processing conditional if it finds its running under JVM then use the JVM Method otherwise ignore the problem +I just committed this +Yup +This method is not used anywhere else in the code +attached patch to tidy up some aspects of chart patch to fix problems in previous applied +These changes may mean the parser results are differentbetter than before +JD requested that you verify that force close continues to function despite changesJD Andrew both requested that you run some performance tests to ensure that region assignment doesnt take substantially longer than +Attachment has been added with description Document properties per transition resolvedinvalid to closedinvalid +The only problem is that two threads may create the same at the same time because of cache miss but that is not a problem at all for correct behaviour +workaround steps disable extend observe call verify +Well there are a many users in Japan +Alieksey are you seeing this issue in the final CR JBDS build +This is one of the much required features +We could optionally at this point have JN copy over the segment from one of the other nodes but that copy can be asynchronous +But It would be nice if we could get this out of the box in the near. +Has this been uploaded now Wei Nan +Running run Failures Errors Skipped Time elapsed sec +bulk defer of issues to +If you dont do much damage to the testsuites yes +This is a known issue and we were planning on cleaning up the cache at the time of process undeployment +The applicable privileges shall include a privilege identifying O +I dont see any checked exceptions that can get in the way right now but if the code evolves to something more complex then it could be easy to forget adding the finally block later + +can make us commit buggy stuff because we assume the bug is in the testBy the way your in the same boat without the test you would commit the same buggy code you just would also have a possible false sense of comfort +Reopening as the upgrade to Surefire broke portlet tests See isnt relevant anymore +It looks from the archive that I wasnt the firstto miss +I think it should be pretty easyStop by IRC if you want to talk about that. +Keith +Change to supply Writable based implementation of Namenode +Victor Great! Thanks a lotRoman for +Mass close all issues in resolved state that have not received further comments in the last month +snjezana where is that version coming from +Committed revision +fixed a filename typo +Fixed on revision . +Incorporate the hemanths with related counters were removed from trunk as part of hence they are not in the fix for trunk For yahoo distribution metric is still in there in failedTaskimplemented suggestions by arun +Heres a patch that incorporates Mr Goldsteins OS X clock handling +Will facilitate migration from Slide +Test couldnt be compiled using common java language +Heiko simply says and even implemented it that the functionality is already there but its not yet included in any context menu action button bar +I like to commit the Test case part any concernssimplify unittest is not worth to commit there is an issue with some java libs and think it should be built with ivy or mavenand use a existing Java Testserver incombination with jetty or soany ideas here +Jim the issue said both mysql and cant reproduce on mysql since that just works for us as long as the proper metadataDialect gets hypersonic we need to write some additional custom code to identify the auto numbered fields +This issue has been outstanding for some time +This was included in but not shipped as part of the release. +Thats what I thought your answer is going to be +The simple template would be modified to insert the appropriate Javascript + +Thanks Pat! +Patch to make chukwa contrib pig compile +Did you run the scriptIt should create windows symbolic links that willpoint from tocatX VERSION +added unit test to check parsing of template jad patch and complete working copy added new goal to easiyl sync manifest and jad added parameter midletAttributes to specify JADmanifest attributes inline updated siteI guess I am happy now +Is anyone still interested in me writing a patch for or have you fixed this issue in already +Expanded ignoring uninitialised warning to all patchlevels of gcc on trunk in by me +The module isnt complete enough to be released in +In these cases the presence of a in a request is signaled by the inclusion of a or header field in the requests +it would be great if this could be configurable +Do you really think a framework should make life so hard for developers? Wouldnt it be nice if our would solve this issue once and for everyoneThere is RFC and there is real life +Do you mean the shipped theme has changed between releasesWe only ship one theme +Thanks +It does not work in the maven sonar plugin +bq +Thanks Zach +If the lack of having configured caused this then I would still consider that a bug +invitation more to do here. +I will subscribe to the dev list and will open a new thread on this topic +The test you want to run is +Please include the beta versions with XSD support as well +Is there news about this task +Hi Werneris your Cache any closer to being integrated with Castor JDOCan the code be checked into a pending folder of the Castor CVS tree for viewingI am also looking at how to integrate a distributed cache with Castor JDO and am keen to see how far others have +The failure to copy is only reported back in the web ui +Isnt following the explicit declaration less confusingNot a bug. +bq +Since Servlets can be mapped to multiple paths im wondering if it would be cleaner and more straight forwarded just to add an optional to to just like has that people can configure in exactly the same way +Hi Anton With the latest changes to in place the test case you supply does still fail +smile +You need to understand that the object is used by several components that is why you can have some component like the that doesnt usesupport some parameters that could be provided by the typically the workspace here only the support the workspace listMoreover all those components +on commit re related open at this point which may have pending changes in code code +Please provide your comments +is going to try to set fields in the readExternal method + +This patch adds preemption to the MR Fair Scheduler +So the data to which the permissions apply is only defined by an abstract label and does not have to be an instance of a specific class +I almost never find a patch eclipse wont apply for me +Perhaps just closing the intro screen when a map or layer is being added will suffice +Anton can you please take a look at thisTamas I wonder how did you managed to install Eclipse? Can you please attach your Eclipse configuration details from Help About Eclipse SDK Configuration Details +It is something we have added to BRANCH since the failing test passes prior to overlaying our changes with git checkout t origin the failing tests are test and testBasic +to sureshs proposalI suspect he has used enum rather than boolean precisely because we can extend it +The issues were that similar that I thought its not worth a new bugThe next issue will go into a new bug. +only startup requests +New profile diffmk was added to docs +Nate this patch doesnt remove the JNA dependency. +Martin responded to me by email that Apache doesnt have the paperwork +Thanks for the tip but at this point Im not interested in a workaround +We have open to do a scripted filter or use something like Groovy or other class compilers to load filters but had issues that resulted in OSGI discussions +After talking with development we are going to change the logic so that if you specify in the XML declaration of your XSD this will be used in the XML declaration returned to clients by +The inconsistency probably raises the priority of the issue +is there a patch +sync with the code committed this to +Create new package +None of the committers has access to an MVS system or any knowledge of it asyou apparently do +The patch for the TAN and CEIL function topics looks good to meThanks Laura +Thanks Mamta and Dan for clarifying what the SQL Standard says here and for spelling out the issues with our implementation + +What is the portlet bridge implementation and version? Doesnt look like its based on Apache +Eclipse Sun Microsystems constants OSwin ARCHx WSwin NL arguments os win ws win arch x IDE beta does not use the launch configurations present in jbosside +When you disable thread safety OR if you remove the dependency to devise the error goes awayReto + +There is a version of this in the JCA examples that could most likely be easily integrated to the JMS examples +Regarding your patch I think we should better not exclude those tests and should rather fix them You can also set the environment variable notestng to true before running buildr in order to skip the integration testsWhen you runbuildr test testnoall tests will be skippedbuildr test testallall tests will run but a failed test will not stop the buildrI typically runbuildr test junitreport testall +Is this also an issue with mm relationships +That way we can easily add different codes to Raid +This other approach seems like a slippery slope complicating already complex code with new concerns +The detailed description is suitable as a source change comment +I tried on Firefox the menu didnt appeared +Some modifications required because min got removed by +Checked in comment changes for SUBSTR with revision for main and for codeline +HiId resolution is part of the commons module for spring data project and therefore unlikely to change its behavior as you already mentioned all of the possibilities make sense close the issue if thats ok for youCheers Christoph +Well have to discuss with Adam and Lukasz to understand the intent here a little better +JDK fixed it instantlyThanks for everyones time and attention +I like rv because the uppercase R with blink star for me is so rich +Will improve the javadoc +Adrian All I was saying is that I believe this discussion has gotten and therefor should move to another forum where it would be on topic +Cause when I check out HEAD I have no java files in orghibernatetoolhbmjava +The reason this quickstart fails is because of the change from Invoker to Invoker class was hiding a specific flaw of the FileFTPSFTP one which appears to have been propagated through the is correct the other parts of the codebase are flawed +If you already have something do you want to share +Please be more specific which warnings you have seen +Awsome +Moving to will not be a moving to for of. +Committed to trunk +After committed please check if the following labels are still translated correctly after they are mergedagreementIdagreementTextbillingcontactcurrencyentryDatefixedinternalNameinventoryinvoiceIdinvoiceorderDatepartyIdpartypartypaymentIdpaymentpaymentpercentagephysicalproductproductIdproductrecurrenceremoveButtonroleroleroleserialNumbershipmentIdstatusIdsubjecttaxtextDatatotalworkThanks some Product labels and move some labels at common levelIf some framework committer can review and commit it I can then proceed with more cleanings on labelsThanks a lot lot of Product labels renamedmoved or merged +Verified by Oliver. +Runping is going to take a look at fixing the patch +Async over remote view invocations is not spec compliant. +Present on Sun JDK as well +Since the only one consuming the get value is the thread could the update be moved into that thread instead? Adding a new executor and new tasks seems like overkill and it looks like it would remove the potentially a flushed memtable being counted by jamm fuzziness +Hmm +I just dont like the extra hurdle this puts up for outsiders to build and contribute to drools +I can create a skeleton module if you like but when it comes to filling it with code I think you will have to spearhead this +Its been over a year since the last comment and Maven has improved since then +solutions Use select on the socket which return the answer which is what I do for consumer and just stop listening on this socket when you dont need it anymore or wait for each line and make a a new request each time an answer is coming which is basically longpolling +I would say that thats needed as the user gets absolute control on the top and bottom level clusteringAn alternative to let it be used in a very simple way would be to ask for bottom level cluster configs and a magnitude parameter which will just increase the magnitude of the bottom level clustering algorithm by magnitude times +Changes done +etc if any or you can provide the comments here +So I think it is fair to say that this change is not overtly harmful to performance +So if I understand the reasoning behind the iterator type it doesnt have a length and instead has a terminator + +DonePeople that require access to this space need to be added to the group +Works for meHave you checked the Include snapshot archetypes checkbox +I have therefore changed the nameartifactId of the archetype to +Users will have to ensure their tables have primary would have voted for this +Updated the fix versions. +I was thinking with this change that it would be good to check the problem url using the client with the server client with the serverJust to make sure it doesnt introduce a hang or anything worse than the protcol error we were getting beforeDeepa would you mind checking that out +Attachment FAQ has been added with description FAQ has been added with description has been added with description well be the xsl processor the location given is the location in the xslnot in the source xml +I suspect that updates to the files since this patch was submitted may have made this specific patch out of date +Thanks JarekRex +You are right +I agree there could be some refactoring but I think it would be best to do that in a separate JIRA +This is not unique to Linux + for the patch +Couldnt find a new snapshot so far +I committed the B patch to the branch +Am I right this is? +to commit +This book is scheduled for a QE review on Friday book is not slated for extensive editing but to allow time for conversion the source must be available by Wednesday th +Oracle has a similar problem +Hi Brian the wtp guy said he is in eclipse summit +Attaching screenshot +Adams working on the ant support +If it has been released since could you please point me to where I can find itEDIT Just noticed that this issue is still marked unresolved +This is fixed in the linked pull request. +So in the case a user specifies the could still process the job anyway as it does not look at this parameterWould it be possible to have a property? This was you could force Sqoop to use the or if you wanted to use the parameter you could specify Ken is currently experiencing a side effect of this parameter in that if you specify an invalid class here will still process the job but if it doesnt when the is used after that it will fail +This is especially important for the block thats reopen for append +Assign to during tidy up prior to +I could probably implement that in late summer +Thanks for the quick response +Found the reason and corrected the patch accordingly +But as it is this method is not returning anything useful for +Thanks for your ypu proposing file this at maven jira or rather consider this as a wont fix? The desired behavior does seem the most reasonable to I think that this issue should be moved to the core Maven project in JIRA +I am not sure it worths to do itso for this jira I would keep things simple and straightforward +I can explain the failure to see this soonerThe bugs are all in the code +i committed this +Small fix with tests +You are encouraged to put all localizable elements from the manifest into a bundle file designated by the entry +Stefan +There are some more work to be done to get this working the client is working with the Java service +Thanks for your comments Mike +This causes the route to be restarted before each test thus my call to stop it after the advice +Oh now I see your updateAnyhow I granted all the right to create plans +Yes because the alert interprets the backslashes +This causes leaks because onDetach is not called due to the exception in the +Fixed in trunk and branch Filipe +Committed at revision +I believe you are referring to the reference guide in which case Marek would need to make the changes +The method attribute only applies to the access constraint +This was on my todo for almost a year now +Nexus returned an error ERROR Server ve contacted Sonatype support to see if they can resolve +Will verify and then close out +HDFS has claimed that it follows POSIX semantics which as far as I can tell makes no allowance for whether the data is actually removed or moved to a trash directory +Anyway I tried that however its annotation processor support refused to operate with the circularity +Marcus lets get this done for CR and Branch and AS trunk +I believe the patch is a more complete solution to this +I admit it is confusing to have both disallowedRoles and allowedRoles configured +Committed to trunk +Can you show your complete route? Or code that uses polling consumer +One way to loosen this constraint and the Synchronized constraint might be to create a strict flag in component that could override to remove these checks +I am not sure whats left if this is complete Venkat could you please mark the issue resolved +Inside NFS gateway it checks if this is the only write if yes it drops the overlapped section and appends the new data section +For servers yes I agree it is a dangerous operation and I wouldnt propose people do any kind of rolling back lightly +Thanks alot +I think it should say Do not warn me when lauchingstopping an instance or similar. + is a model project set +Priority All other blocks which do not have required number of replicas +Need to upgrade the integration code and +The latest patch should resolve your exceptions around cache warming and eliminate the memory leak issuesThanks for helping track down these bugs +So seems Commands can change state Custom Commands cant and Actions cant +On pressing the button both functions are called twice to set a hidden input and afterwards all inputs are displayed +Thanks. +Just tried your code with your instanceschema files with code built from CVS no errors were reported +To reproduce the issue both preferences Mark occurences in file and Mark Attribute Values in page WebJSP FilesJSP Occurences must be selected +Can you provide a test case or a some sample docs that reproduce this behavior +Its a problem for not sure how it works on WindowsThanksBo +We can help those w filters do convertion to whenif they show up trying to figure how to convert +I think we should at least only do this if we know there are subscribersIm not sure thats actually useful as again bumping an unpublished version is free of side effects but yes that can be done do you want me to update the patch +should support this +And perhaps the refactoring of outlink filtering code +Fixed with revision +Thanks. +Probably should throw an error earlier if you see a negative number undersoft upgrade +Similarly we could combine the original cluster Id and write time to know replication latency from source to current cluster +My grey cells are not what they used to be Can you please help me grok the logic following statementIll add in some commentsbq +Committed revision . +It is not a Struts bug. +its easier to review now I think +Also added explicit value of indexDir element to to make it clearer how to change the value prior to initial startup. +So no new tests are added +fix committed +and are also discovered with includefalse +Yes you will have to edit them in an external folder +He uses which could be the cause +whilst a bean added via addBean is implemented any way the user desires. +Also added a test case for the patchThanks Uma for your off line help as well +It sounds like functionality like this is planned for M +See for the details. +Im going to look into it now +This is a small safe fix +patch that applies the easy changes noted above +more infos requiredany logs available +NET stuff +Should this be listed in the release notes? If so please provide a summary of the problem and any available workarounds +LimitationSince only the HM image URL is parameterised the port forward of HM only works with dashboard but not for grabbing new data for analysis +We can open an issue under Karaf for porting this feature +If we determine MAC is a viable option then allowing the data to be persisted between different runs would enable it +Make sure that your AxisC code have the patch applied +IMO its better to make the custom bindings actually useful instead of just keeping compatibility +When talking about caching items on construction time I would have said instead of tags +sounds like checking the reconditions before executing is a good ideaThe result of this could just be clearer exceptionsThe question is should this happen? If so it is not an exception +I disagree with the statement at no point does it make sense for a region server to say no to an assignment for one particular case If the region servers open queue gets too deep it will not be able to open the region before the master times out the assignment and assigns the region to another server anyway +On shutdown after has finished shutting down delete all connections inIs the last item safe to do? I think it should be +Well +Grant it looks like we have a good plan for going forwardCan youdo you want to submit a new patch or should I try to modify the original one +We have run into a few problems with code depending on iteration order failing in java +why not put the converted objects in the viewMap of the view root? They can be created when view root was created and be destroyed when the view root was destroyed. +Fixed at revision close for the release +support custom function call written in drools and parameterized function parameters and capture the values using cell editor +bq +Closing all resolved issues for. +Whenever a has a form matrix it will be took into amount within the Invoke operator by multiplying it with the current transformation matrixNow the attached example works like a charm with the +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Folks We have a release blocker and the most serious bug in to date any objections to checking this patch inIs it ok to commit this one? would like to say yes but honestly I have not had time to really look at thecode +I have ran latest patch against the spark package built from the master and all tests are passing considering that one has a running HDFS of course on the patch will commit it shortly +was fixed to correctly parse the PRIMARY KEY column constraint. +fix for committed +PTP am having the same issue here +I have not seen a single after this patch where passes +Fixed maven The patch does not match the description +add screenshot for style classes Sorting controlsLayout in Details of Usage in accordance with Table of footer facetadd top bottom up down attributes +Restore HADOOPCOMMONHOME and environment based control +Latest patch +Can you confirm the st attempt +Bug is +StackDo you think patch v can be integrated +For me the main purpose of is to manage use case where we have the data split in different db type +Other issues according to changes in model are still opened +It is ideally the only one that can interact with them for thread safety reasons +The activeThreads cannot be cleared before joining them +We are currently doing this by deleting these components by dereferencing them +Proposed patch +Patch integrated to trunk and +There is a new but we still need to and stabilize the entire codebase for serving our primary aim running Hadoop applications +This issue is fixed and released as part of release +does not close fieldsReader if open is false +this should also improve the performance as it avoids several unnecessary object creations every time we process a frame on a connection +Im not yet sure if this is the case? Maybe someone could comment +Well adding parameters is still a problem you cannot guarantee that the editlink is taking parameters in a particular way +Repository Registrar makes perfect sense to meThanks for providing the patch +with nit I think we would be should make naming consistent and change COMPACTIONOPTIONS to COMPACTIONPARAMETERS +If it was a it would be difficult to have the operate on itAs far as construction and initialization go we have many options to consider +Please let me know and Ill get you what I can +Thanks +updated patch attached +Ive seen some really nasty compaction behavior lately with whats currently the default +fixed. +Committed to trunk. +Scenario may be the following too +You just need to fix the AS is possible to change the version of libraries in. +Yeah this is out of date closing + +No +I checked the document but could not get any hint where they are and how to copy them +here is the patch which fix the issuepreviously the background is overriden by the in the css +You would need at least a long and an int to represent the date with the timezoneA String does use quite a lot of memory but its easier than writing a separate serialization +A quick solution on Linux to get Doxygen for DRLVM can look like follows cd DRLVMDIR mkdir doc cd doc cp CLASSLIBDIRdockernel Patch using attached DRLVMDOCDIR +push to +Maybe this makes sense IWC is high level +Not really related to this issue but no harm in being aware of it +Patch looks goodCan you submit a patch for TRUNK? for for your review +Im on the patch hopefully a committer can review and push it in +Patch has been applied +No docsWe are not using it anymore +Ill give this a more complete review over the weekend +I believe it sets the classpath to libaxis for theRichard and need to be adapted as well +Not sure if this is really a server issue or Designer issue +I will add javadoc soon +I cant see that being a problem +HiPlease review and apply the patchThanks applied with commit revision dcdaaabbdadbbceecThanks +It would be helpful to hear about exactly what pain this causes you +Added test case +Temporarily reopening to update release note info. +I couldnt get the current master at the time to build +The text after the first set of examples is incorrect and should be removed +Resolved +I think we will have to wait for the unit test to get fixed before accepting this patch +Check on one more time +Like you say it should be a small task to duplicate the tests even though it would be more elegant not to +Attached patch fixes this problem so now JRE is copied only if is not specified +Todd has said Would it make sense to add these functions directly to the S filesystems but not to the overall APII still see change in the new patch +Hmmm +That doesnt necessarily make required thoughIt would also be very nice if we could avoid breaking all of the existing by having a core that could also act as a default core or something +Attached WSDL file generated by Axis when service is deployed +If not I will pull it out of the docs until we can get it in +This time for properties +The ows dispatcher does not i believe you fixed that some time ago and the old dispatcher has long been sent out to pasture + find +BEFORE triggers do work but they have the restriction that the trigger action must be that is a SELECT statement or a VALUES claues +Sound like the split should use numRegions instead of numRegions +committed to trunk +Committed the a patch into trunk with revision +Then you should get the results you expect in the database +Map is accessed from different threads by build and validation +I like Kens suggestion that tells Hadoop the task is still alive +Added basic tests for reads and writes over JCR content graphs of various widths +Thanks Dag for fixing the bug and Yip for reviewing the patch +Hi KimI recommend reading the conversation which Dag and I had about behavior +I was the one who inquired about this on the mailing list +If we have further work to do we can set it for thanks for the screencast +for a managed case the XA case was made the general case +In my experience in apps that have to support a lot of languages there is a lot of redundancy in both the schema and the config +Marking as resolved as per comments +If you could test the next nightly build and confirm for your tests itd be much appreciated. +This will be available for release +Please if anyone on this thread who can reproduce the issue can spend minutes to take their app and distill it into a single +Suggestion on how to split spec and behavior to make it clearer what is required for spec behavior +looks goodIn the context of do the previous version that could release all my locks was nicer +Is this really still a critical issue? + +Also I dont think dist log splitting has anything do to with this failureTrue +Attaching michelle fixed in svn revision issue is resolved only when the patch is submitted to +Well Ive checked it in anyway +Arne do you still have it anywhere? +This doesnt compile +I believe we can close this one based on what we agreed to in but Ill leave resolution of this issue up to Jimmy if he is satisfied with the agreement I believe we have reached +This change would be immensely helpful +how shall rubygems handle such a situation +Thanks for that update Tsing +Attached a file that has the right formatting +The road to has been full of and changes and I think the way in front of us sees probably a couple of more things to complete the cycle +appears trivial well squeeze it into the next in trunk thanks for the patch +That might be a holdover from the ASAS integration or may just be a mistake +I agree a mojo parameter would be easier for activation +STM supports explicit definition of which state instances to save and restore but the default is to save everything +Its not very good to look at the message to get the exceptionOk +Do you have an groovy code example producing this +Seems like a reasonable approach to me +However it would probably still make sense to add a check for the presence of the context attribute in the session in case a new session has been created for some other reason +Pushed to master already +Committed changes into in trunk +Closing resolved issues. +pull request merged +Sorry I totally missed the updates on this one +given that even that the string concatenation has not shown to be significantDoug have you done any benchmark to support the above statement +Rather it would be better if the shared bits were separate projects or subprojects that are independently useful +It is obvious that has the same bug because it is a problem in the definition of the algorithm of +Added accessor for version +Attached the patch +Thanks. +Filed to address the precomit build issue +Removed whitespace tested the patch on the same configuration I used before +This is a trivial thing +If this is not configured somewhere else in one your parent your build will use different plugin versions depending on the Maven version +Yes I tested this and it appears to work great now +Updated the patch for trunkThis tests can create random failures +thank youI was not aware that using the Java methods is actually slower I actually thought that itd be more efficientthank you for clarifying and for your invaluable input +All integration components now are set up like Apache subprojects. +Upgraded the jar from version to +Build changes look okay and builds okay +Result will be a tutorial +No I mean a prototype of the solution +According to the trace it seems there is a network issue during the test that is why we see from to AS is not able to send the CONFIG that explains the I have tried something similar on yesterday I cant reproduce it +Fixed in is a tiny change to the docs to match thecurrent behavior of the code namely that MM and SS values arerequired for TIMESTAMP and are not optional +Looking forward to the patch +Back to the patch it would be nice to have the train file modified in a way it has a few short lineswhich contain a few sample records embedded into html tags +committed +The package was being exported by both monitor and +If you find further issues please create a new jira and link it to this one. +Were there any attempt to continue this effort else should we close +Patch in progress +Agree this is a bug and should be fixed +Emmanuelsince has been just resolved whats your plan for schedule for this task? Collection indexing support would be much more complete +I guess you would just have to set a when creating your? I suppose it could extend and for given fields do Trie unencoding uninverting and all other fields yeld to the super implAnd then if you get some other XXX encoding youll end up with XXXVS extends UVS extends UVS extends XXXVS or duplicate code from the other one +The current POM works fine for me with Maven and +I dont see many open source utilities that can open this formatThanks under r +this was changed to cannot reproduce so is the servlet responding for you? all the logs claim that it should be but theres still when trying to accessThis same app was working and responding fine without the manifest but when changed to use manifest it has stopped working +Can you confirm on your test project +Should be small fix +Pls provide your comments + This is a bug +Seems like we are using runtime OI only for and not for or +I can use an and find my entities it is only when I declare an Interface extending that the error happens +Ive posted my config files +It probably makes sense to add a parameter that specifies the earliest date of jobs they are interested in seeing and defaulting it to day ago +curious under which circumstances you would run a cluster of jackrabbit servers that have different definitionsanyways i am curious why you removed fix version does it mean you no longer plan to support this functionality in release +Thanks +Changing test cases to use which will always be present while running unit tests +Its a much used feature +If your files span multiple blocks the reads will be even slower +Closing issues +With regards to the adaptive failure detection algorithm Joys suggesting there was a ZK project to implement and evaluate different failure detection algorithms that may have some food for thought +partially implemented w currently fails if the method is not there but it still does not check for the actual type +Only one little problem is that is missing for Ive attached a patch for thisThere is another problem that Eclipse will pop up an error dialog when starting a remote server Ill open new JIRA for this +Please pull it again and we can close this issue after that +Now I see the httpd actually depends on the httpd package +Eric Newton has been experimenting with running goraci against +This is a dup and is fixed for Final +issue could be closed since pull request already merged +Moved this to next release since we still can explain the replication from the current GUI +already fixed it +It is recommended to use but I had so I figured I would try to upgrade first and if not downgrade +Seems like we should have this +The other changes are mostly a consequence of that or use cases in the file writer and readerIf we get that right then the API details around the edges will be easy to adjust +Committed revision +Applying the new functionality would couple the AMF module more tightly to Spring but without any noticeable benefit +Repro steps unknown happens occassinaly +Only the simplified edit drawing is drawn in place for those features +patch applied to trunk +Patch applied in rev trunk +Which file is this? PROJECT +reopening for backport +pluginsfeaturestestssite +Updated patch containing successful tests in fshdfs and examples packages +Attached is a straightforward patch that increases the number of rpc server handlers from to +Added a check similiar to the for trying to create a compilation unit if one doesnt already exist +Then again I am reading actual with quoted commas so its possible that the similarity in runtimes is due to the fact that sees the commas and allocates extra tuple fieldsD +Weve done enough of this requirement for our needs for now +This looks like what was originally intended by use of the mechanism but this was not working correctly +the issue is settled +While now all the faces configurations are added +This new patch has r and as option that need to be passed with value either true or false while running the script +Unit test could perhaps be simplified +Ive never seen this test fail because it runs across midnight so perhaps we should just remove the code rather than trying to fix it +As far as trigger frequency +Unfortunately this didnt get finished in time for the release but at least we have a workaroundLowered Priority and deferred to +A question How will MAX and MID be set? Do you intend for these to be hardcoded constants properties that can be decided by the user or something else +Gordon Chius patch has been committeds and s are now used in place of fixed length arrays so that users should not run into limitations based on array sizes +verified for +Isnt this the case with the Cliff Click Counters +In fact all the local links under Commons and General Information are brokenThats probably because the Lang URL was incorrect when the site was generatedThe URL has since been corrected but the site has not been +Withdrawing +this shouldnt be happening any more it also shouldnt be marked as major since the query is valid either way +Is there a way to those connection without stopping and restarting Tomcat? Thank you for your helpSimon +thats exactly what I had in mind +Ill fix and propose a new patch +fix verified. +If the contributor likes itwants to address it then it is up to him +But indeed all warnings I found in the javadoc refer to the API am I missing some other? +and could potentially generate forms for submitting procedures +Hi Julien looks like the parameter is not saved when updating the config I cannot test it +This looks like +Also change the comment following the change to say missingDeps null or empty means action can be Ready +I will commit this patch and then file another for the smaller fix +Do you encounter the same kind of this problem +Thanks for the tip and sorry for posting a duplicate. +So all issues not directly assigned to an AS release are being closed +Thanks boss +We can do the cleanup on this issue and then close +Thanks for this extensive patch Xander +Reviewed little remaining tomorrow will check this +Committed to the trunk as revision If no new problems show up Ill investigate merging this change to the branch +Id vote for simply as the name of repoGH desc Unified Push Server +Modified cxftrunktoolscommonsrcmainjavaorgapachecxftoolscommontoolspecparser cxftrunktoolscommonsrcmainjavaorgapachecxftoolscommontoolspecparser cxftrunktoolscommonsrcmainjavaorgapachecxftoolscommontoolspecparser +Now its fixed on the share branch tooCould someone please test itThanks! +I have test it also with and yet the problem remains +make a good point +It now makes pk optional for entities having + is limits the number of threads in the thread pool to be less than the number of splits +So when clicked this will goto the namespace map and show proper components only +The original forum posts problem lies elsewhere +let me commit this to fix the trunk build failure and file a jira for the fairscheduler test failure +Deployer now uses PMD from handle URL from EJB code getting it off its root in particular. +When we used newEpoch instead of the result is that an old leader still reports but the updated followers report newEpoch +Verify that you can create a new case for a product that requires a Red Hat You may need to create temporary subscriptions in the cache table +I just put the commented code into a patch for consideration +One more thing Id like is some common handler for socket setup connections that improve diags is it ok if we raise one subtask for this +Can we please rename the method and update the references to it in the compiler +That has some changes to that I think are from a different patch and it doesnt have the change I made to fix the tests +Recommended Practices and in section +This time created patch with Git +Thank you for reporting providing test case and helping on the analysis +Attached. +Thanks +Also confirmed this behaviour on +But is now STATISTIC as it does not have a done Yes Ive set the message to STATISTIC as it does not work if you specify a value for that the deployment fails so until this is not fixed it should be clustered is now readOnly but still available in the we close this jira then +Good work Rakesh +Thanks again for reporting this issue +The version used in EAP is have tried to update with from EAP +This will be a great opportunity to get together and talk about this rationalization layering and deployment usecases +With Safari and Mac Os this issue is no more present so I will close this issue +Im bad at autotools so I cant verify the patch but this bug really annoys me as well and I would love to see it committed +Ill upload a trunk patch later +As reported by Nahi this appears to be a couchrest problem not a problem. +I did overlook this so apologiesYour idea sounds a good one +Alex I have not been able to reproduce your problem +Patched Source to implement course this is a Bad Thing +Closing this issue as Duplicate of +committed +Each notification will contain the list of files changed +Hence when we call mark it doesnt go through the normal corrupt replica handling path instead it gets immediately enqueued for deletionThis makes me a little bit nervous if we had a bug for example which caused the NNs view of the gen stamp to get increased without the being increased we would issue deletions for all replicas +Plus anyone can it if they extend +Thanks Suresh Bernt and Kristian for the review +At the moment I have no timeinterest to dedicate to KML support +Timestamp +Rebased patch with fixes for unit test failures +And thanks to Ralf for starting this business of moving JDO functionality to a separate module. +Verified on jbdevstudio. +Made some minor changes added tests +This server left alone will persist across multiple debug sessions +Preliminary patch with some testing complete +Sorry I ended up hijacking this bug a bit since my issue turned out not to be really with the CLI +Are we ready for thisJPOX is ready and has been for a while +What do you think +tnx for input + +Path solves this issue +Yeah +Could it be related to issues like set to or or classpath settings +AndreiThat sounds great +It would be nice to consolidate thread pools a bit but this is ok for and well look at consolidation in along with continued work +Updated patch to remove conflict with +Egor these operations are bitwise +What the status here? I whould prefer to talk with spamd via socked +Then we can close of the issues as dup +fixed +Code has been imported for M. +bq +However I would like to firstly open discussion whether it make really sense to implement such channel and what other demandsuse cases would be +Anyway it would be nice to have it at least in another target language before we think on changing the IDLBy the way there is probably something a little clearer than structv perhaps external import or something like that? But thats just a detailCheers probably just dont see it but why do we need a new struct type? +This instructionGo to My Computer Properties Advanced Environment Variablesand addFORRESTHOME as CfullpathtoPATH as PATHFORRESTHOMEbin should be changed slightly for Windows XP to readGo to My Computer Properties Advanced Environment VariablesCreate a New variable with name FORRESTHOME value Cfullpathtothen Edit PATH by adding FORRESTHOMEbin to the end of the current value +I havent had a chance to review the latest patch if you want to take it on then by all means please do soThanks at rThanks Nicolas +This should be elevated to a blocker in my opinion +There are no new noteworthy changes for Beta vs the ones documented for the M milestone +Completed patch alongwith testcases +Patch applied thanks Ralph +Thinking about that one I had intended the to be a scratch pad for multiple threads working on the same transaction +going to close this for now +forkall would be nice to have but it can be bound through FFI if anyone really needs it. +Fixed +Logged In YES useridChristian is working on this now +looks good + +Should this be closed as well or is there progress on this issue +Trying again +I like the idea of to always give you the raw API +should actuallybe called since its unrelated to maven itself +The patch needs small reworking +support for this has now been added to BranchStable including a new persistence manager config file +You can reopen the issue as soon as you have such an example. +Please give me feedback on style guidelines and anything I am missing to comply with the standards +We can preserve backwards compatibility is the default impl with the new reader only passes to the deprecated old customScore functionI will provide a patch tomorrow +I was not able to reproduce on my local setup +I use multiple classpath containers for main dependencies and test dependencies so theyre visually separated in the package explorer +The local cache of svn logs was lost during the upgrade and sucking down k logs was exceeding a process limit on +Patchset applies on top of +Patch for the Example Generator +I split up my base testclass and ran it in a non transactional context without issues +What do we want to use the new logo +Forgot to add the new files +I hope these are okay +bq +Ill be sure to announce when its ready +Trunk patch straight forward port of Y version +bulk close of resolved issues. +We should mimic the behavior of ls when listing a single directory it emits no header and when listing multiple directories it emits a header +If you still see this please reopen it and provide some context but as far as I know this has been fixed already +We can open separate issues to migrate other implementations to the API +This is a patch incorporating Owens comment +At the moment that is not possible +At least we should test it both on hibernate and lower versionSo if it cost too much times I will suggest to delay this issue to next release +The only differences are that I exposed the replicate on write CF setting and added Super Counter Slice query and some counter range scan support to the api +what? it makes perfect sense to share these code generation with your team members +It seems reproducible enough if youre willing towait three hours for it to get to a point of failureWhen the exception is thrown itsaborting the entire thread and I have a copy of the index as it existed at thatpoint +assigning to is a duplicate of which was fixed after and the fix is present in Based on the dates I believe the version specified in the bug report is incorrect +Please reopen if Im closing all resolved issue that have not gotten any more feedback or comment in the last month +Moving to +This patch is simple enough that eyeballing should provide confidence +sounds good Ill wait for the ping +Mike since you have described only a workaround here I am moving this back to known issues +I added a class deletes record the ram used to DWPT and DW +Logged In YES useridhello sirwe are using and we are working on all youAmanpreet +However if you download Forge runtime and you use the ForgeInstalled Forge Runtimes preference page to add an select than your problem should go away +Often when I post to members Im directed to board which hates my emails and tells me to shut up altogether +Updated patch modified get to no longer check operation category which is fine its reasonly for printTopology and report to not failover +Also need updated for at least a new you need to apply the change while Im out next week the changes are in the attached file +cleaning up some old bugs no more work going to be done on release notes. +Im trying to make the patch +Looks good to me +More or less the same sentiment on this issue given the ongoing inputoutput format discussion +There havent been any Xalan releases since this patch was applied +And it didnt work before the patch +The revisions you say broke things for you were but I cant reproduce such errors locally +Im uploading a patch that trims the response to the maximum actual available content in the file based on the request offset and length parameters +Oh and I think that the spec needs revision so that the example actually works. +However according to the spec it should actually point to the class path root that the file has been found you by any chance already tried loading files that way? This seems to be the main reason why we need to pass a persistence root URL to the persistence provider in the first place +Close issue in preparation for release. +This effectively leaves with two issue tracking systems to worry about and is hardly a good way forward +Thanks Chris! +Andrei are you on a mac? Could you post the error plz? get no error the command just blocks +more details project tools were reverted to use for barName +Mass closing all issues that have been in resolved state for months or more without any feedback or update +does DTP have API for having dynamic connection profiles +I modified the above by introduce class for no wait I also implemented this usingit is equivalent to activitiasynctruesee and project is in the attachment. +New version of the that produces correct position increments with matchVersion +Regarding d I will removing the XSD implementation and keeping an implementation based on Xerces model +Well this might be a JIRA +Let me check again +But in that case keys are a tiny percent of whole file so any savings from VB will be insignificant +I think at the very least it should contain as much as did in +Your mirror was removed because it had not been updated in many days +patch for thanks. +Compiler fix and some refactoring landed on masterc and cbd +Suits me +Doug now that the patch to is out there do you see any issue with this approach? Today I rewrote to make it backward compatible and tested on my local machine +Can we just close this one +however when youredeploy your +This has been done thanks to Dominik for fixing the last couple of failures +There is missing source for some sample apps too +any particular reason you are reporting this in HBXMaybe we could fix it in higher levels of tooling code so it depends very much of what the underlying call stacktrace is +Committed with fixed nit and changes to tests added to the separate commit. +contains some changes from the earlier versions but the design is the sameI will commit this later today and use it as a baseline for further work unless someone want time to review it before it goes inI expect there will be some changes later and please note that the patch in partial and does not enable any of the statement pooling code +test failures are unrelated to this patch +Attached is a proposed svn diff patch for this issue +The examples for custom home mapper and locale first mapper are committed with r +As of JBPAPP testing RC this test is passing and failing for me on a regular basis on Windows x BEA JVM at least with the same error identified by should perhaps introduce some mechanism in JIRA to identify flaky tests so they can eventually be cleaned up across the board. +I meant Open with menu not Send to menuUsually I want to open +Consider removing unused importsShip it +Seems that the mapping file is still included +Shall we default on strict true or does that flag have other implications too? +Thought I could make my code a little cleaner +Thanks for filing the bugI have the following patch to change back to package private but your solution also looks good +trivial patch tested on cassandra it add any more libraries or just upgrade existing ones +now working with version submitted a zip of the changed project +Access grantedPlease upload your pub key to svn then let me know so I can pull it into the VM +and in that case you would have to be explicit just as you would have to be when you have multiple datasources and transaction managers +Fixed in Rev +I dont see an update in the last hours +Again as per the user mailing list discussion this appears to be DNS and logging related not Tapestry related. +Looks more like a new issue +sample connecting to Apache DSNote Because it is now possible to connect to other LDAP servers it would be nice for this portlet to be included in the Console as an additional easily accessible tool like the JMX viewer +all positive clauses are stopwords ok users were always confused about that +I think there is a wiki about it just cant find it maybe Uwe remembers where it is +Its working +Provided we do not decide to allow Dialect self configuring via db metadata which I am not really a fan of. +Simple Test that triggers the bug +Better formatted for reading same as first attachment that I do not have privileges to delete +I am on the patch +does nothing but invoke an evalution its like lines of code +Seems it didnt changing to ER +Steps for testing Go to Facility Stock Moves screen Fill up the To Location and Quantity needed to move in Quick Stock move form select the check box Click on Confirm selected move Products should be moved to specified you so much David for reviewing this patch +The deletion of in its entirety is a real problem because we lose the writeBinary methodLosing the JSON stuff is fine but writeBinary is the only way to serialize SGD models properly +Have attached a patch for a proposed that this bug is still not fixed +This is why I would not be any good in QA! Thanks for taking the time to look into this +Hopefully today +Reassign to J David Eisenberg for reviewing +Thanks for the info +This because thetestfile has multiple references to the same signed content +Previously this worked although not quite sure how +There is now a command line parameterto set the encoding for the output of the text renderer +However when I use like to test the same values as constants it correctly evaluates to trueij select from where a like escape IBMY row reviewed this patch and the changes look correct to me +sorry +Hendriks public preferred user name is hendriks +Reference and license added +This was creating a big problem for us so we needed to fix it asapNow we can think of more options +If there are no objections I am planning to commit this to +It might be a good idea to also consider when refactoring the input format +The code isnt working +One half of the issue is fixed at revision for nodes where jcrprimaryType is not null the node type defaults to ntunstructured +The process to add the m pom needs to be added to the JDO release process +I am worry about this test +I have reported a similar issue already which was fixed in +Manually tested! +Tested with and the issue appears to have resolved itself. +I saw in your replies to other bugs that you consider some of the cases similar to mine as misuses of the framework +Patch usable by +The documentation has been updated +I solved this problem by using exslt make sure not use libxml or create soft link for libexsltsudo ln s one seems to have been fixed by an upgrade to +changes commited to both trunk and the branch. +Nick the infoStream output looks healthy I dont see any exceptions +I think thats sufficient but let me know if this should be emphasized elsewhere +Cancelling patch so I can retry Looks like the hudson bot only triggers when you attach a new patch and mark the jira Patch Available +The patch wraps the redirect JSON under an extra tapestry key if on Tapestry for pushing compatibility forwards +Test case is not written because the existing tests will verify the correctness +Revision I enhanced the Felix Framework override using a more elegant way I override in the etc file in place of overriding the whole etc file +Moving to. +Great observationThe problem was with +Fixed so it allows multiple headers to be combined +Fixed by the new wrapper classes enabling of the file upload interceptor in the core +on patch v +Fixed in rev +I looked at the patch and I just noticed that the appender is added to the logging system but it is not removed when the test has completed +If the update log is in buffering mode then the request is buffered otherwise the version set by the source leader is stripped and leader logic is invoked +I use Windows XP Professional +Or are they getting caught somewhere? Whats the outcome that you see +I request for this patch to be considered for a commit +The changes look fine to me +a new patch addressed Nings comments +I think this will be a nice feature to have in the future as the same configuration perfectly works with collect +Stuart if you feel this is not a bug please close the issue +I guess the code assumes is equivalent to not set +Verified on EAP +Thanks! this looks good +this one is very trivial you can fix it yourself +Thanks! Lets commit it +Hoping this can be in the UI somewhere +EricProbably I was using some old version of. +Attaching here since the failed tests will eventually scroll off the jenkins historyThe failure is not in the JMX connection its in the provided to tests by ARQ and used by ARQ for deployments +Looks like problem was with the environment variables on my machine +Lets do this the simple way so we can roll rc then we can make it better for move migrations onto and use to avoid problems there +Can someone review the submitted patch +Will apply this tomorrow from tap computer +Patch looks great +I have tested concatenated bzip files with plus patch of and confirmed it could be correctly in MR job +bulk assign all unscheduled issues to next release +This patch passes mvn clean might be stale I couldnt patch it into trunk today +Switches to use bash to fix shell expansion used on line by patch submitted by Jake Farrell. +Perhaps we can track it separately? +Could you provide the example code you use to cause this +The commits for fix the reverse mapping example on branch and trunk +Note that I have a tentative code solution for this so there is no need for anyone to work on this right now +Yep JIRA by comparison is zippyTo be clear JIRA is operating as normal confluence is not usable at present +any idea which version of mozilla that runs + +Deferred again +I am attaching a patch which adds the property remove on the idempotent consumer +CommentsPlease remove private boolean has false from There are multiple lines that exceed chars in length +Committed revision +Patch that fixes the issue +Should be able to do this in that template thing in confluence Mike +Sorry forgot to set that +Thank you for your job +committed to branch and trunk +I fixed it but something has changed and I dont know what +varcatalinaworkvarcatalinaresourcesto customize other cache to delete when geronimo starts +there could be two transactions at precisely the same timestamp +Ive been following this thread and am actively trying to use +The only special case is the first request where you dont have a seqnr and get everything from your DB +Also Ive done some experimenting but the converted junit tests only need of the old releases but I cant get svnexternals give me just one file it seems to want an entire directorywhen I tried to have it point to just it gave me an error like soFetching external item into toolstestingderbysvn REPORT request failed on reposasfsvnvccdefaultsvnCannot replace a directory from withinSuggestions +Should probably have mentioned that the config file containsplugins s accesskey XXX secretaccesskey XXX bucket accountnumber XXX path images default certfile XXX keyfile XXX overwrite true +Nabeel Yes It was a mistake +Odd as I have the! in my local copy +table is waiting for timeout +Than why I didnt work on it so far +Basically the ifdef MSCVER wasnt working for me so I just hardcoded out the ifdefs to switch into Windows mode and I had to remove all of the defines for INTs and the four typedefs for ints from the diffHope this helps somebody! Cheers +Primary master running +Another option would be to switch to an essentially unlimited character type +These need to be fixed. +I think this would be preferred before we go optimizing synchronization because otherwise how do we know if its correct? is a downside to this whole issue of course +Looks like in the Big Gossip Refactor there was a slight regression +A new version has been tested based on opensso express +requests for search action in +patched thanks Florian +I turned it off and then back on and verified that the change is working and then set it back to tried it on the linked ticket provided as well as several others +Posting the JSP that is causing the error would really be helpful otherwise were just guessing +allows modifications +Once you fix the GREP and TR issue I am a for this +Real uploads this time +Ive made a small change to the as followsr added the Woden project root directory to the classpath defined in the validateWSDL taskdef so that when running the WSLD test suite the file will be located correctly by the resolver framework which uses it to resolve for the standard common schemas and to local copies +Duplicate Issue Issue +The test test fails now not sure how it should be fixed +Not a complete patch because its missing tests and such but for my use case Ive forked a copy of this class with something like this patch. +Not sure I understand +Committed to trunk. +Can those changes be made in the next dot release of CDH rather than in Apache Hadoop so that CDH is wire compatible with Apache Hadoop +A workaround to the problem where is rendered in the popup is to change additem request so that it does to orderentry +load if no directory is specified +patch committed +Im closing this bug for now for the reasons mentioned in earlier comments +All tests the norms bytes loading is good it seemed incorrect but Ididnt want to mess with it as I didnt fully understand it +Committed a fix +So checking in the relevant artifacts and releasing downstream versions is going to be on the forbidden list +if you find a way to do this automatically and attach it as a patch we will be happy to include itcheers +closed +Bulk close Solr issues +sorry I mean +Any chance youll be able to address Alejandros feedback yes +Please do the needful +Committed this. +I will proposing them a fix using your example document as test +So ifyou want to test sending receiving you need to emulate this for example via MockI updated the asf sources of the component +But it is not really necessary to enforce that condition instead this should be let to the developer of such validator but the jsf implementation should just allow itIf no objections Ill commit the proposed patch soon +Although I see in your sample you also have it turned on +So this is a as far as I can seeBut if there is a bug introduced by then Im glad fixed it. +Applied patch the issue is closed + OK Ill review the delta between and Will also in a jailed location and test in that location +WARN as orangeThe output should then be mostly default colour with the important stuff highlighted. +Couple comments Why are seq and the associated lock static? This makes it harder to unit test and seems unnecessary +i just used some jpeg editing tools to change it ot reads +At some point we should probably comment that code I had to think for a minute to remember what problem solves +The build seems to be ok locally I need to investigate what is causing this truncated issue in CI build. +the shell script exit becuase in we have a call to exit command +Fixed in SVN +We in drools now use role too but I d much rather use language +It enables NM to reuse failed diskes after they come back and tt also adds a new metric of the number of failed directories so people have clearer view from outsideFor unit tests I add a test to and mimic disk failure by chmod faileddir and mimic disk repairment by chmod faileddir +The oddness of publicfields is my fault at the time it was introduced and also mine again for not fixing things before was released +I think we have insertion tests as well +The test by itself is a good test to have +That is why we saw the error which said JACC manager gbean already present +As such it should not have anything removed +There are other places as well in hadoop where it was assumed that it is sufficient to use shortnames for example proxy user configuration in checking file permissions etc +Perhaps my x solution needs to be left in place and we need a little more discussion about the default valuebq +Ive attached a screenshot showing the Drools runtimes Ive defined on my machine +Please verify the docs from stage during CR test cycle +Fixed in revision close of resolved issues. +Were getting to the end of a release cycle at the moment so will look at what youve just submited in the next week or so after the release +Just committed this to trunk +Ive applied a modified version of your patch along with a change to a generated names for the threads based on RI behavior +Yes I believe that is one of the requiredderived properties when setting up a CP driver +I like the idea of sending this information via CountersThis data could be used by schedulers to make decisions on whatwhen to schedule new tasks or preempt existing tasks +This is especially useful when you have an HDFS instance on Amazon EB instance for example +duplicate of issues in hbx +patch +We have used procedures for other features like updatedrop statistics and it will be good to follow the same path for any work +We believe to have this same issue and worked on this with Robb G +tcsh and bash behaves differently for multiple globbingIf the behavior is not standardized I suggest to keep the current implementationMarking this as Incompatible Change +Adam is this fixed with your changes to ? Can we close this as duplicate? Thanks +the change may have no effect has a toString want to add toString to to make things more clear +Fixed in SVN now +Attempting to clean up JIRA a bit +Merged +I have attached a log fragment after region server killed +Thanks for spotting and reverting Angela +This excludes duplicates and the projects are used from the place were they are created +Its not ready and im not threatening to commit to trunkI created this branch to develop publicly +Attached is a patch would resolve this issue +Actually integration testing might be more complicated. +Can you provide some pointers on where this somewhere is? Do you have some some code samples you can shareThanks Gabriele +Is this subsumed by? If so we should resolve it as a duplicate +Thank you for the review +The most funny thing is that by mistake I then unregistered this postprocessor and everything worked even without it +Tests what happens when you delete all the columns in a family in various waysTest passes +ok +Release Notes text left as is but Release Notes Docs Status set to Not Required because this issue only affected an Engineering Release and was fixed in the same ER therefore no customer can have experience the issue +I was going to add some error handling to it before putting it up but just doing it now in light of most recent comments +Just to note We also need to change the Forrest index format documentation +should check for empty +i just committed this +The work also implements +At revision +This task limit feature isnt meant to solve the general resource allocation problem only to give you a way to limit resource consumption if you know that you have one job with very tasks and many jobs with less tasks + +In fact openConnection only declares to throw +Thanks Jake! +fix the issue using locks to ensure that does not get called between getting the List of store files and it getting to theBtw Ted also has a fix to this issue where in we use a reader lock and a writer lock in and the writerlock is usedwhen we update the files +Lets try to get this into b so we dont rip it out of a stable release. +For JSE there is an additional package in Felix file as shown belowssSahoobugs diff jse felixa committed this fix thanks for double checking. + +. +Its a big merge though with tons of conflictsHA! good to see you here! have fun with the +I think that this bug can be safely closed +out of hunks FAILED saving rejects to file applicationsmonitoringsrcmainwebappviews weird +here is a I like it +Otherwise we could build in a visitor style pattern for visiting java source files that might be useful +Okay here is the latest up to date patchThis uses for cloud tests I think we def want to use localhost in the end but I really dont know whats causing Roberts failures so Im doing a bit of straw grasping +for the patch. +So I suspect there is some intermittent bug with the bags rather than an issue with this code +Even with code the need here is great +Is this patch against or trunk +Double postback from firefox +Please verify that is has been applied as expected +But this seems like a low priority Im skeptical that those calls use enough CPU to matter +We also cannot apply the whole thing because of the issues with the name spaceDid you used a tool to add the license headers? If so you can maybe just the tool on a fresh checkout +Do you have any performance numbersIs the slowdown measurable in the worst case of a phrase query with both terms in every documentIll try and review it more soon +Was hoping the source of the bug would be obvious +Pavel can you add an example showing this issue? created a QS for reproducing the issue and attached itIt is used as a common QS ant deploy antcould it be fixed before GA? This issue blocks performance testing with CXFThanks +Moved to please move bugs back to which you will work on in the next weeks +best regards Dileepa Sisila ChandrasekeraDepartment of Computer ScienceUniversity of Moratuwa +First two parts done splintering the third part so the larger set can be in review while I experiment with some different designs on JMX +Log ends with instantiating compiling with Andrew the sleep is there to stop a race when building multithreaded iirc +Trunk patch +I reopen this one to fix a second reuse problem It does not reuse the like Lucene did +Paul I understand that showing complete is not possible but I do think it would help to log a message as each task completion event is retrieved so that the user sees progress is being made +There is no more a nor Invocation +Should I report a separate bugI am using to package itI also have added the battery feature in +Attached the additions to the plugin in UML form +It was also brought to my attention that is marked Private right now while it should be Public Evolving like PBCD +So I am inclined to check the output messages and use that namespace if there are no input messagesBut I sure would like some direction from the SOAPWSDL gurus out there +select the axis operation you like the module to be engaged +The Axiom code has been modified to output empty string for the empty namespace urior +This one shows it +sorry the environment is windows this still a problem with the latest release +Reopening because the spec needs updating +Therefore there is no meaningful way to share it except if we synchronize access to it which would be much more costly than creating a new one +Not sure if this is the same thing look for a duplicate process +I believe it would be useful to update fix this patch for +Also I couldnt get the patch to apply +Based on your comment I checked it further and I think I found the problem in beta is not found see line in log file of the attachment +I havent tried reproducing it on winxp I know that Denis only noticedtried it on ll try it on our Fedora +Hi KevinIf you check the link above there is a flag that means that your patch is not informational only and you allow ASF to use your contribution under its license +Hi MarcSorry to be such a pain but could you please also include documentation patches +But the raw input value is also serialized and then deserialized +I have read about RS codes and LDPC codes +How about this? parser dom util storage api memory based storage impl file based storage impl jdbc based storage impl think the separation of storage parser and dom could lead to a clearer design +In case somebody else misreads it too Id rephrase to say Field may be used to override the field name chosen by the you very much for implementing this Ill rest easier knowing document fields arent being silently overwritten because of a naming conflict +One last question now that is no longer available how do you go about setting properties programmaticallyRoger +Instead files are copied using standard java io API +Fixed by r +Committed to trunk and branches +Ill work on this again +Thanks +It is now removed in revision +Instead of just storing one Connection instance it should manage a Map with the as key +so the only place to store the information if the content was installed is the repositoryi have no problem uninstalling the bundle if the for that bundle should be avoided +Thanks for comments PavelThis issue is umbrella I put the composite patch for convenience purposes only please refer to dependent issues for individual issues +Has this already been fixed on +Please if its not fixed hard enough. +Instead some buttons are disabled and I cannot finish it I will try to discover what in my ear makes the problem. +Wow a lot of code removed there +the workaround is to increase the column length in your db +Excellent +Im on Linux and normally I can see whatever charset without due to lack of feedback for various transition all resolved issue that did not see any further comment in the last month to closed status +I think jboss need to improve internal communications to avoid situations like that +But I am sure it will be possible to use lower eclipse versions with newer jpa patch the required jars are from Eclipse except jpa +svn ci m Improving performance of get per s patch from Pino SilvaggioSending srcjavaorgapachecommonslangTransmitting file dataCommitted revision . +Add a private field superuser to and ensure it is set in all the constructorsLooks like not so useful hack to me and might not even work with kerberos etc +Documented as known issue with workaround in EAP release notes +Let me know Jacques +MoustafaThe idea is to st check that its still working with before trying with later version +No problems please try the latest release candidate and can you give more information about your browser and with which example you get this errorThanks also tried it with Tomcat and JDK and again no problems here +Any suggestions on where to look would be appreciated +What would be an alt impl +patch from RB +Workaround described in does +The structure of the Struts Tools reference guide is reorganized +You may only wish to plugin a and and +Thanks Jody +RescheduledIssue is related to internal DOJO problems +Shutdown process I already explained in previous posts and startup is simple we just start the dbs start the connections and start the normal work for application +If such functionality were to be provided where in the grand scheme of thingswould it be offered? Iow would it be made accessible +Now the runtimes sites are always shownCommitted to the trunkAttached is a patchMaxCould you please review it +Inclusion on is like a seal of approval +Automatically assigning not sure who would be handling naming +This patch was apparently never applied to the patch has been applied to the branch +What is the exact code you are using to register the backbutton event and anything wrapping? Try disabling other plugins as they could conflict +good to have in general + +It appears its not possible now +Patch attached +Just to let you know that the refactored test case you attached from bug are fully functional +User shameera Date Time AM To change this template use File Settings File Templates +We also need to investigate the metadata container +It will take more development time to add and maintain such tags +I will try to get this completed before or on Thanks is still a high priority for me but I have some personal repsonsibilities today and Sunday +releasing grammar patch will unblock others development +Ok so I got permission to hardcode certain things which will mean supporting shared JSTL and JSF +Removed the directories from SVN to resolve this as closed. +Will verify in next build. +Committed. +I just committed this to branches +log cant recreate this problem in the release branch perhaps if anyone else gets a chance they could try this on their setup +Thanks for the review Joydeep. +Thanks +It is a time problem +It is using the preivous +targetNamespace can be replaced with any local or other with the same result +that change wasnt really the point of this JIRA if Ivans ok with it Id be happy leaving it out and filing a JIRA to make that change later +Im sorry to reopen this issue as in my working on Ive found the patch can not work well +Can we please try the +I think the discomfort we ran into here is a good sign that this is something better managed outside of C. +Thanks. +New patch +rev +Since the replication factor is if both these brokers were the replicas for some partitions they go into the under replicated state and the script is never able to shut any other broker down after thatI think we should include this fix +Using commas is IMHO not intuitive if we go that far I think it would be better to refactor the parser and Query to support nested clauses if we really want to support more complex query syntax +configure it would stop with an error when a configure check that is required for java fails +This one bears a bit of scrutiny first attempt here +Do you want to take a crack at a patch Rakesh +Forgot to add the new test case on last patch +Sounds great! this issue has been solved with the commit of it can beCan you just confirm this was fixed in? If so Ill close this someone confirm my analysis and mark this issue as resolved +In the case of having the abortConnectiontrue this means that youd have consumer connection coming and going for no real reason which could impact performance of your application +Its not like getting +Modifying the accumulate expression as outlined in the drl file will switch to the correct expiration +I see +Committed as r +Closing in favour of +I didnt know about delayer +Therefore the value is returned in its persistent was able to reproduce this problem in possible explanation to Mauros observations is +The predefined classesloaders are not initialized +I cant reproduce it nor have you shown the error or stacktracelog +when i restated the repair it completed in minsI asked on IRC and someone said Oh by the way silly Oleg repair pr in DC will only handle the partitioner range for difference between tokens in DC and DC +Possibly with some of the new store updates madeIve got a test case to add as a regression case for this +I rewrote the implementation comment to fix spelling mistakesgrammar and hopefully improve the explanation +we have introduced these parameters as part of feature in itself but later we removed these parameters inPrior to these parameters are not there so I think no need to mark them as deprecated +being sent to Austin for us to redistribute +Changed fix version to match the version on the branch. +Looks like hudson is not picking up patches +I need these in by Wed +Im not sure +bq +If its just docs please to their team +ACS feature freeze date is November th +eos added to the rotation +The exception comes from the WMS specification that states if the request update sequence is equal the response should be Exception code +Patches updatedPatch against trunk adds Copyright headerPatch against also reverts the Attr String change that patch against trunk includes. +Thanks Colin +Closing with the release. +I just installed and had a look I can reproduce the issue +Merged fix back to +Please confirm it has been applied as expected +Ill file another bug on that issue +All wizards must have title +Sounds like a plan BenoyIt will be good to have a test plan laid out here +Thanks +Patch looks fine with me +You are probably working on a more recent file +This would also prevent possible in the because not all of them correctly check for +Okay there doesnt seem anyway to support the luke target going forward +Is this the attribute you are looking for? Probably not but I am pointing it outDo you know which method could have bufferSize set on +Thats definitely a pattern that seems to come up very regularlyI guess thats what is intended for but I never end up using it because feels wrong to extend from instead of just from +Thanks for this fix +EtcIm still on the fence about whether we need separate jars for each logical tool or whether we should just move all code into a single +What would be the next step in the process for moving it out of the sandbox and into Tomahawk? +If original posters use case is the same I believe this issue can be closed and solved +At the moment I dont understand how a object could possibly be of any use +If the geoip database does not match exactly with your firewall rules well wind up sending people to an inaccessible site +I also got better perf moving it +I have the same problem with PDF documents generated by on a MacCan provide samples if desired +ah ok thanks +the from +ok +It would be nice to add a comment that the test can not be run in client server mode due to the casting to internal Derby objects of some result set columns for dictionary tables +Username wonnekeysers +Integrated to as well +No documentation required +Tim thanks for explanations +It will be great if you could confirm it either way first +We can discuss for M depending on the timing. +Changing this from bug to improvement given that the main driver behind this ticket is to add more control to field management for this +Hi DennisIf there is a soap endpoint and an HTTP endpoint there will be two endpoint objects with different names in the endpoint list +While my job is running I notice a bug remove the old cache file push a new one to hdfs and then launch a new invocation of my job +Since every host knows the mapping from tokens host we could in fact use any token belonging to a host to identify it +patch committed +Is this a Linux specific failure +Ill update the page +testing +I ran all tests +I dont think thats worthwhile to take snapshot if your vm will be no response for few minutes +But it seems currently would want to know that there was a change immediately by requesting the content hash of the root node or the journal of the root node +Applied thanks +Added to which demonstrates the problem with relative file +Robert would you mind assisting us understand the root cause of this issue by loading the Roo shell typing development mode and pressing enter then running the script again +By the time if reviews are complete you can go ahead and merge the same +I noticed this also affects +That was over a year ago nothing has changed except the version numberand were now dealing with weekly OOMs which bring down every other serviceon brutusIf restarting jira alleviates the leaks then trading minutes of monthly scheduleddowntime is an major improvement over having a few random OOM outages for the same timeframe +If you havent signed the CLA could you please do so +If you wanted to take on this ticket that would be great +This is currently a forked effort for the release and well need to address this problem in directly +I already took overIll review the javadocs sometime this i already applied the patch +Do you think that this will be available for Solr? I would think that this is very similar to distributing regular facets +This is almost identical to Bill Aus version for so thanks Bill! +Thats a nice start! You seem to use the independent version of asio +Thanks Clment Koji! +I would suggest to this issueWhile fixing it it might be prudent to check the framework for more occurrences of similar code using +Therefore the getter and setter should NOT be synthetic but the private backing field should be +Works you very +Rob you dont need to build Lucene to try what Gilad suggested just modify your search code to disable complements +This issue was originally opened on my behalf but I dont think that the description really describes that problem that well +it it fixed +Attach a patch +The initial contribution is to affects fix version this was found in trunk while was being developed and fixed in +If we cannot fix them we should remove them from the setup +The xsitype is generated and abstract types are not treaded different to interfaces +If so I can fix it myself for my deploy but Im wondering if this could be added for +q there are more diffs are the last change +Which database platform do you use +in CSS Hyperlink when using Media rules is fixed + Robert and I already discussed about making one array out of it +btw with the progress does not show up when running with maven +Thus changing this may have a significant impact on existing code +thanks +This patch removes the text and only the PPMC member votes are binding +Fixed in revision +Perhaps as a JSON string? might be the place to address that if you need complex data types then perhaps each line should be a JSON value? Note that I just added a patch to that efficiently reads and writes JSON as Avro data that could be used for this +Yes I will plan on incorporating the test case into the patch +Its probably better to expose the bug than making this a daemon threadSo the only change in patch v is to address +Will I break anyone if I reject any paths that have +Verified +Could be that with we are using the JVM implementation of JMX Remoting rather than the one from MXJ +Closed the bug +The standard Seam WAR project also depends on the EJB project +Assigning for feedback +Any chance you could break out a test case and fix for this from that? Id like to see this fixed sooner rather than later and that patch is fairly large with a lot of unrelated changes in the other issue +This feature is much discussed and often requested +Set that was not the way to set it +Resulting in random strange exceptions +and expects it to fail +Scheduled for inclusion in it be included now that is out +I think we need to have the same type information for and! as we do for any other desugared binary operator +Zo all of these files are explicitly not granted the to ASF for use? Can I assume that to be an accident +Marking critical for and for +See the developers guide I had just addressed this a couple of days ago +can be used to find classes according to different rules etc +Will commit shortly. +All obsolete code has been removed +Thanks +Verified fixed in IR +Id suggest that you email dev list explain them the situation and see whether they can come up with other suggestionsPersonally any code that involves dealing or putting something in JNDI should really live in AS or in some kind of integration project +Patch attached +It also has a section on using the old JCC driver but mentions that the client driver is the preferable method and that the JCC driver will not be in the next release +I just to report that issue still exist in Solr I have tested in Windows with IE current trunk does not work with IE + I did the changes for locking +When implementing defaults functionality we realized that the hardcoded defaults did not make much sense and would almost always would be changed +changed request interceptor to setHeader + too bad weve missed this in previous reviews +The test can be found in +Interesting we do have feedback so we can drop things into the catalog but in the case of services that are already there we have a bit of a catalog view represents the local connections if you had a service you found in the search view you should be able to it into your catalog in order to connect +This was fixed when I rewrote the guide for only sucker refference Ive found has no code listing and otherwise its formatting seems to be fine. +Im not quite sure I understand what youre I am wondering if is compatible with the other spring great +Hey Guys I think we need too +Yes actually +I dont think that will work when clicking the group link thats how I ended up with originalTarget but apparently thats +This is an upstream but I have never had time to debug the issue you list here I have never seen +This is on the same lines as the in the package +possible fix for the for the patch submission RickFixed in trunk revision Fixed in branch revision in trunk and branch +It seems to me that simply removing synchronized should work +This has now been set back to the original soa defaults in which is included in the latest in in CR +Thanks +The change committed on requires change the way it is detected when we are updating a view +Its good as an example but not as code we have to maintaincheers Roland +Moving it out of release. +good to me +In this patch it will check whether the attribute is an alien attribute every time when the parser meet with the attribute if the attribute is not exist in schemaSo I suggest we introduce a Map type variable to to store all the alien attributesWhats your opinion SvanteMark this issue as reopened +I guess that is the answer to my previous comment +Fixed with can I close this defect? Have you tested it after Alena checked in the fixthankssudha +Ok thanks for that input +A test case was also addedMany thanks! +plain old field injection didnt seem to work when I first tried it but thats been ages ago now. +If there are still issues feel free to reopen it again. +instead of ignoring the whole layerBut that is just suggestions +The mapping in that was solved already in +When the Thrift was recompiled there were strange changes in some of the generated Thrift files +Thanks! Seems like you forgot to grant licence to the ASF +However I another test fails in HDFS that I am investigating it appears unrelated +Why not work up a patch and submit it +The doubling memory issue for should still be noted somewhere thoughAgreed if you can do it considering it unsupported without saying such is not very helpful to a user + +Martin Grigorov added a comment Last Friday Thank you for giving me a very cool tip Martin! The tip was very helpful and worked just fine for me +Additionally this test is causing build failures for me +Also its not a macro you should call it something else +I wrote to Jonathan and he will do something from their end if theres any issue +Hmm ok moving to then +I cant reproduce it in build do you use +Thanks Brock! +Works fine + +Thanks Kasun. +Integrated to trunkThanks for the patch AnoopThanks for the review Ram +Dans fix is in the file supplied by Dan assigning JIRA to him +It has been a while since we have done a release and I feel like people may be getting anxious +The Admin UI was already updated to reflect this in another issue so clarifying scope of summary to be specific about + provides similar functionality but at a finer granularity and obviates the need for this patch +An alternate solution which will involve a little keypressing is to simply set exception breakpoints for exception or Invocation target exception in the forked process to try to get to the source of the put a test case on a branch +but not a show stopper +Seftion does a good job of providing an example of how Adding a resource to the build should be modified for including and then +TicketclosedRegardsHPP +This would allow you to have a that would receive updates via config API and not get appended to some other random file that happened to sort last +Im not sure if they were looking for something other than xml though +Originalment la forma era regar fora de test per per all de lescatologia va passar a pixar fora de test +Please close this issue. +Also there was a possible +Openejb modified to use new jars in rev Ported to g +Sorry I posted too soonReading your comment again I guess that you are indeed going to support such queries +Verified with EAP +I think that the maps are relying on this contract +This issue was due to a testsuite setup error and is not a bug. +This bug has been marked as a duplicate of +My apologies for misunderstanding what was required +Andor update the functional spec accordinglyo Does this action need to invalidate statementsThanksRick +Here is a patch for your looks good to me +I feel like I have been using Hibernate for so long I dont tend to even think about Eclipse Link +This allows for the step to log out exceptions via an exception logger which is purposely separate from the repeat template exception handler +Travis yep Ill close it +yes it worked +Thanks MarkusTrunk Committed revision Committed revision +Heres a patch which use a and adds test resources before the normal +However it is a dodgy hack so you may want to inspect it. +all bulk edited issues have in a move issues to and +Adding some pom changes to this patch because RB wont let me add this to the Avro Sink patch because the file was modified in this patch +Hi Hairong Im interested in testing out your fix is there by any chance a patch against the seriesIve only been able to partially apply your patch to to the official release and it seems there was a major code reorg between and branches +You are right looks like it Publishing a new version is lots of effort and I doubt Ill have time to do that right away +sed s +I have logged this with Contegix theyre very quick resolving these if theres a fix from Atlassian +Someone already fixed it +I went ahead and checked in the fix since it was breaking my builds for other patches. +When Im done with the license I plan on change the package to wicket +Full response +Could you set up a build of that dumps stack inside and find out where its happening? It absolutely should not be getting called unless has been turned on +Hi BJIs it intended that ASF licence is not granted +Should be finished +Added missing piece of as resolved. +Hi thereI did a pull request with my changes on last week +The fix of this issue introduced the problems listed in issue +Thanks Rob for the contribution and thanks Sangjin and Daryn for reviews +Well if you apply the patch for the test case then the build will break because this problem isnt fixed +Couldnt you just compile against the library compatible with JDK and available from Sun? You can list it as an optional dependency in Maven +It so happens that the properties file contains version numbers but it could contain message strings or whateverThis is not a bug in HC +To give some idea +Shrank Summary +Thats what I used for the Markus +Content New +To explain the patch a bitThe bug was that making a file in deployable or undeployable would save a blank setting in etc +Other thing to investigate here is where the process information step could be moved from the intercepting thread to the batch reporting thread +has been comitted the previous patch wont apply +Commented out setting so not turned on by default +All the dependencies are core components in JDK +A lot has changed since I created this +Do we need the binary version of the travel sample distribution? It seems that people interested in the sample would want to look at the source code +I committed the patch to branch +Put another way if an administrator has specified a value for the tasktrackers memory limit we give it higher priority than ulimit setting +Is this for work to port it to branch +Also mvn test DtestTestNamespace passes +Committed to trunk. +GOOD FIND +Then once reenabled the replication workered would go about rereplicating everything +Todd Couple of questions What is the implication of ignoring RBW altogether at the standby? If editlog has a finalized record can we just ignore the RBW from the block report +Here is a javadoc patch for the reduce +Well see how that plays out +updated the install guide with the newest of new locations for the web console user properties file +this is about removing the need for trust between the user realm and the service realm +Not ciritcal enough to patch on B. +I fixed precedence issue and added more tests +Ive defined a new macro AXISGUTHTHILAENABLED and do the same deletion with different way which are suit to two memory leak is resolved and work fineregsI think you are referring to a compiler flag +The editor ate it +Moving out +License information is prominently displayed on the projects main page and also on the download pages. +Removing the helps too +Just that I forgot to remove the commented out code inBasically the code for both the launchers is same +I believe Jigneshs change is almost certainly the problem +It is pretty important that are stored against the step instance +Not being worked on +Not everyones quite as pedantic as I amJon +Im afraid listing as a supported media type isnt the right solution +This issue seems to be related to will close the bug as fixed for now if the problem is still there we will reopen the bug +I added an READMEUBUNTU file to cover issues +Only with a coprocessor would it be possible to exercise checkVersion and avoid the network IOAlso note that in your filter case youd still get the SEEKNEXTROWSEEKNEXTCOL in for each column that you includedWhen you get a chance could you check out the last patch on? Maybe you could run it through your micro test Id be curious how it comparesGenerally if that the column tracker code is not efficient we should fix that rather than circumventing it completely with a filter +Danper our discussion could you look into implementing support for namespace declarations across the WSDL Element model rather than just fixing the support thats currently only in +Its a style thing and Im not too fanatic about it +This is also fixed in the you resubmit the patch? Seems to be not there getting a HTML response from JIRA +this IS a real problem for us tooCan I assume everyone else using MavenSubversion has not upgraded to? Otherwise not using the release plugin +heres a patch i didnt add support for existing scorers thoughin my opinion all scorers that contain other scorers should implement this api so that you can traverse them +We tested this in production and its working good +would it make sense to add it to root for each component which branch its usingI dont think its a good practice to put the branch in the pom as it means that whenever we change branch we need to update the pom +Seeing this in the +Patch verified at r +As a result we could also optimize the the other joinpoint matching proccesses in the same way by matching only the pointcut expressions that can match that type of joinpoint +This fix is already in the branch that will probably be used to build the EAP version +as n increases the amount of generated code gets very large +fixed in JBDS and JBDS SOAThese features has wrong copyright yearThese features has no copyright informationThis feature has wrong provider and no copyright information blocking but lets get a basic test up and running to test for these instead of having it keep failing. +I believe that this is what Martin means by run the security checks themselves +Done in revision +Thanks Junping +uses the s TTL when providing the TTL in metadata reponses +Ideal case would be pass NULL and get the version from soap message +This patch fixes it +I added some notes in and committed the patch to master branch. +Tim Ehrmm +I changed the title as the old title was misleading +So in other words if you change Orders total to no discount on order because the total before Order is +Committed as discussed +Patch for Y +Im fudging here +If someone else wants to review it I am fine with that +works perfectlywill buy coffeelunch if you are around the bay area anytimeregardsMonish +Thanks Jerry I just committed this. +Since they act in a chain of responsability pattern and do not know a thing about each other I see no way to solve this differently without breaking encapsulationWhat do you think? turned out to be much more complicated than expected Jody jumped in to give me a hand and we got a little distracted organising some logic +Works for me +Stepanconfirm the patch was applied as expected +part for the entityresolver and additional mappings section. +Increments maps from CF map of Column longIf theres a good way to unify that Id love to do that +There is still remaining with old style finalName and context +see +Minor comment should probably be renamed as and also the class javadoc should be updated +changing this to a new feature +I would check which you are using +Maybe you can convince the chainsaw people to accept a patch that reads the data from an XML stream or out of a database +If I use I run into this issueI naively expected these filters to behave like you can do stuff you call the next in the chain and then you can do some more stuff +fixed +Im actually surprised that Rick was able to apply the patch without problems +Of cource we can rewrite all other consumers to use this new api +Applied to trunk and X branchIve modified your patch to flush the previousch characters not only inendElement but also before some other sax events like start element +As soon as Xwork will be moved into right place in Struts project Xwork is now part of StrutsRegardsLukasz +I dont really understand all of what is going on in here +If it were always that easyHave much fun with the new CocoonJoerg +We dont have any reports for this problem for earlier versions so Id rather see that the changes go through more testing before backportingFor the record it was the automatic index statistics update feature that triggered the bug +The original ticket was probably geared more toward Designer +Fixed in trunk in revision +Hmm +Fixed C There were tons of files that were duplicated even in the pressgang repository between and jbosscommunitystyles which made it impossible to tell where to fix a bug +Certainly we want additional coverage on the File Channel integration testsDenny I dont want to lose your thoughts here I agree that we need coverage on the failure case if we dont already have it +Juca please commit your fixes for the testsuite to the EWP branch and later merge it with the EAP note the changes support both EAP and EWP testsuites +pushed to and in Developer StudioVersion Build id Build date +Please also note that was the suggested workaround to solve. +Hi Ray could you provide more details on whenwhy a text bag property is useful? Thanks +Also involves a new config option to provide an executor pool size for async repl +I am closing the ticket. +ASV Scan Report Attestation of Scan More Actions after dropping it down there should be an Attach Files option +Ive uploaded a first stab of the migration code on RB and +Open build debug on an Android device and watch the trace output to see that the bogus events are being fired as a result of a simple orientation changeWeve tested this on two phones and with release builds as well and our app just unloads on every orientation change because we are listening for the to close sockets etc +It works with JSF RIbut it doesnt work with JSF RI or +tests were not producing such filesAdded logic for tests to generate +a Rule that executes testsI assume this is not intended for productive systems? I dont see a good need to have in productive setups +It doesnt work right +As you said just the esb plugin needs to be renamed and were ready to +But anyway I add a fix patch to support hbase startup script +to a contribution of Kite provided As with the other code we have taken there is a public open source repository available that is accepting contributions or issues a repo with issues enabled is a good example There are no dependencies on source +I will go thro the current patch also +The system properties for the keystore were missing supplied in a gbean in the corba config +Cannot reproduce this issue in EAP +This jira is about allowing the Hibernate attempt to bind to jndi to succeed +Im not sure what the best default upper limit should be so I have used +If it can be provided by a third party repository then thats better and more flexible +are these s transferable to other? +for the fix +Ill probably do a in a few weeks. +This has been tested with EAP CR that installation works with both and openJDK +Hello I was planning on undertaking this task as part of a project in my System Implementation course and was looking for a few tips possibly +It seems to work a lot better now +Just curious what is the use case for an empty Java file +patch for only +This looks fine +maxandersen applied PR in both Alphax and master branches +We have bounced JIRA which was the fix last time this happened +Modified to add suggestions by Daniel +classpath file from modules +Applied this work in rev +Fixes to play well with more styles +Hopefully itll help somebody to google it +Comments given by Ranjit against is affected to this patch +Thanks Brock +Dont think is worth the trouble is it? And it is a bit stuff. +Change Variable name from DISABLEWINDOWS to DISABLEWINDOWSTEMPORARILY this should be more obvious to yes in some cases we do need code +IT added for each new language. +In fact I think for dev +removed Disposable interface does not register for disposal +An error code is not a stack trace and IMHO it is an illusion that it will ever remotely function like a stack trace +Fixed will be in +will commit this momentarily +Hi ShaiI will try this out today also with Clover +Either way works for me +Username andreifsd +okBTW with not using this name sorting even for +Added in r thanks for the patch Rintcius +Sorry had to revert this fix as it caused BZ to crash +Youre also correct in that this is the correct behaviour for the Java runtime however the difference here is that the developer did not extend these JPA classes in their code nor did the developer override any of these class methods in their code +However it is done I dont think having throw an exception which kills the store scanner is the right thing to do +Thanks. +But let us say the upper limit sa a shouldnt it be simple to keep in ZK +I probably have missed some along the wayI like your approach to get a more complete picture and catch them all +yes this fails under juno Andrej dont you know whats the status of this issue +Resolved in branch at revision +It was generated for first patch is not very generic it focuses only on handling +Thanks Joydeep! +Thanks patch and +Setting to resolution state before reopening to change affects. +bq +Part of a bulk update all resolutions changed to done please review history to original resolution type +Addendum makes sense +Screenshot with updates for rules +Since the introspector in the JRE strictly adheres to that restriction one can assume that the spec writers intentionally excluded the Boolean wrapper class +Added default summary Patch for KAFKA +Thanks for verifying! +Ive ensured that all session logs are logged viaTake a look at for messages +Jackrabbit only works with LucenePlease check your classpath for jar files from old jackrabbit releases. +My guess is this issue will get closed in though unless someone brings forward a compelling reason to implement logical need to change when we use logical addresses say we have physical addrs A and B and logical address L whenever we receive a packet from L the physical sender could be A or B so we will not be able to compare the incoming packet to the localaddr! +by an external source +The global configurator will run again in the action phase and restore this object +it had discussed for two years +Ive just committed this. +Both PDF and PS render the same +I also did some tests with tibco ems +duplicate of +Im seeing the error right now on Developer id v date can you look into this and see if you can repro and fix +head for now +For my situation Id prefer not to use files +We will need to review this further and determine time required for this +Moved here from JBAS +I think youre rightWere probably copypasting the whole entry without removing operational attributesDefinitely a bug +Sounds reasonable +Fix unit test failures +Heres a more fleshed out design document including BN operation upgrade and some of the future direction design sketches +Resubmitting +Patch similar to Owens but with some improvements doesnt rely on static variables to configure the static user adds unit tests documents the new for the test case toddincorporated todds suggestions although im not wild about doing the username lookup for each requestThe default user is already in the code and doesnt need to be in the configuration file tooUsing a userid means the chance for false positives is much lower +This could be addressed when Portlet support is reduced to a plugin +Meanwhile we no longer need to skip this because the site wo bothing to or metadata has worked +On startup all beans get instantiated at once +Verified by Nathan. +updates the description of the joinStrategy property and updates the example for the joinStrategy property +Ive added a dependency on and used a configuration mapper +Ill close this issue +These are the improvements which need comments Where should we save the content how to input the saving location to whiteboard? Can we define the location when whiteboard initializes? More information related to old whiteboard to new whiteboard mapping +Done in one day under two years to mark this closedDYou shoulda waited one more day. +Lets bring it back now that were past the deadline +Attached a patch that contains specs and a better spacing for the fix +Can you please reduce the priority this is not a blocker for next relase +Please place the simple statement directly in the callers +alternative version using boolean flag from +If we want to limit network bandwidth consumption I think we can use no +Sure I am happy as long as there is some way to notify folks that they might be looking at stuff +I attached a patch adressing this Benedikthere is patch +Committed +localization is already committedI guess this issue can be marked +I have attached a guide for using viewfs with federation +Fixed +Also the following moves should be done modulessecuritysrccommonjavasrcjavaxcryptoto modulescryptosrcmainjavajavaxcrypto modulessecuritysrccommonjavasrcorgapacheharmonysecurityxcryptoto modulescryptosrcmainjavaorgapacheharmonysecurityxcrypto modulessecuritytestcommonunitjavaxcryptoto modulescryptosrctestjavajavaxcrypto modulessecuritytestcommonunitorgapacheharmonysecurityxcryptoto modulescryptosrctestjavaorgapacheharmonysecurityxcryptoPatch should be applied only after otherwise an additional update is required +Done as described +Committed and pushed to master +Implemented please +Ok +A little questionWhy did this patch not contain the following java files? mapreducesrccontribraidsrcjavaorgapachehadoopraid mapreducesrccontribraidsrcjavaorgapachehadoophdfsserverdatanodeThe class has also been imported in these files +The code in still uses the faulty DCL patternShould probably use IODH solution as per original issue was actually resolved but the introduced a new DCL bugSee for the DCL bug. +bq +I dont think this issue is relevant anymore. +The change is too small to have a unit test +I will update the stackoverflow discussion after is released +The container should be able to pass in a unique EMF name +Maybe this should be just an option somewhere in with the name faster +It does work in my current codeTry it out pleaseregards in current head +Not sure how to handle exceptions that do not have a standard counterpart in NMS +This meant that the policySet definitions from the contribution werent available for use by the compositeThe test case also needs to be modified to load a contribution containing a file with the required policySet definitions as Mike has pointed outFixed under revision r. +It turns out I cant this ticketThanks! +Actually I have since realized that I failed to observe the doc conventions outlined in the Getting Started book for some things +I think Emily is also working on this. +Ive split up all bugfixes as far as I was able +The bridge apparently expects to find something quite different than a bunch of empty strings in this list and therefore proceeds to remove everything related to the invalidation cache from the +Initial version committed in svn branch into trunk in rev examples are correct but proper documentation in the user guide is still TODO +Forge seems to work fine with it looks like it was added to the module but not the +to prepare raw content for use by Mahout and also to provide utilities that might be helpful in dealing with output +fixed in revision Vladimir please check that it fully resolves the +Message may get stuck when source node crashes and doesnt restart +Could you please provide more details and possible sample configuration which shows how you expect it to lookAlso i am going to downgrade this issue to a task since I am not sure at this point that anything is broken +though if it is used as part of the bootstrapping of the class libraries there may need to be some way to do to until we are ready to load ICU +Thanks Raghu! +The following are some more comments +Thanks +The bug was fixed as part of. +Shiro aims to be compatible and source compatible +I do not believe that this is directly related to but it may have some resemblance +Just saw you latest comment now +This issue is contained in the followup issue. +If the gc exceeds appropriate error is raisedAdded a script file in the contribbin folder +This issue was created for the first version of the. +this way you can keep it there in case you need to revert by removing the +How do we deprecate the current behavior? I mean how do we detect when to issue the warning +Dear allEnvironmentTomcat should be Tomcat Could someboday help me to update the Environment infoThanks a lotMichael Xiong +Dependency injection still sometimes fails on startup and requires one or more restarts to resolve +What Ive described in my preceding comment works today +just to be clear it does not has to be generated nothing breaks by this except when you communicate between two different JVMs and if you do that it has to be something else than just a single constant which users then ignore to create and thus breaks the serialization contract that has an annoying default warning for this which makes people think they must set it and thus set it but dont realize they then actually stated they will guarantee the serializability of the class +The relevant ticket for version number update is +Proposed fix and regressiontest patch Dmitry and Denis +No sorry I have no idea +trunk +Reverted but there are some later changes still to back out +I have commited a fix for this to CVS +That is how I have been developing up until now +Please review this patchThe signature file has been updated with all of the annotations and the signature verifier has been updated with new methods to check themStill undone update signature file forThe modified sources for the signature file and signature verifier have been provided so you dont need to apply the patch to review it +Woonsan Ive assigned this issue to you as I think it might be related to +Thanks for patch v +Hey Andrea did you make this type of performance improvement already? Its basically just making it so fids arent requested and we could do it as a hint for the wms +interfaces will simply be ignored now not causing proxy generation to will be available in tonights snapshot +Uploaded with the following changes +Thanks for the yet I just retried and sent you an email about the result. +It requires a test repository and an index in the attached +Id stillappreciate someones the javadocs though +I think if not set it takes it as +Ill commit this sooncould you help me w the thread dumps after the abort +The vote is on right now so heres your chance +was fixed +Please verify and close the issue or if the problem stil exists for works now but I have no permission to close this issue +See Maxs comment +Changes look good to me +I am going to close this issue +Eric could you possibly provide me a high level list of steps +Updated patch to follow this one breaks a few tests +okay then shouldnt the fix version be set to +So I am not seeing big issue here +Will keep you posted +HathawayXalan Documentation name commits works for us! Steven J +With the st option its not clear how to implement a protection from leaks +Is it scheduled to go into an upcoming Maven releaseThanks is working but it needs a lot of testing and wont coincide with the release +Benson I committed your patch to trunk in r and branchx in r +Is the idea to keep such configuration parameters in the corresponding log device implementation for +Any thoughts? Anybody +Could make savings here +Is a patch on still forthcoming +Got his log file he must have had the other logj categories set to INFO because the root category set to DEBUG still didnt produce DEBUG messages +Fix this test case and the multimodule case will be fixed as well +I am fine with the first go around the CPU number is in float and the scheduling unit is CPU +Removing Affects Version and Fix Version since this neither affects nor any prior release +I can open a separate jira if you plan to commit this patch as is +Good idea +Remove unused imports is not groovy aware and will not work in groovy imports will provide the same behavior for groovy bug here is that I need to disable Remove unused imports in groovy files +I anticipate having all the work committed in the next day or two and will call a merge immediately +I have no problem to limit it to sub set of that will be in a new configAvner +I updated the jira title to reflect the current discussion +HowardCan you work with Dimitris on this issue +This is +New patch addresses Rakeshs comments +I see now that this is a duplicate of +Thanks for the reviews +Old title was Overloading doesnt work with int and String setters +Every time we create un record in database a message Successfully created is added to would like those messages were taken from a properties fileMy motivastion we need those messages to appear in spanish Creado con exito +Per discussion above there is no more issue regarding this. +Applying a delete without deleteColumn or deleteColumns deletes the entire row +Each batch is then sent in its own thread while the main thread waits + Listener was missing in saveState and restoreState +Interesting looks like puts returns rather weird process id +Why do you think stack trace at the client would help +Thomass approach is better than mineThomas Can you update your patch so that the target depends on package instead of jar +Introduce pseudo jars and Implement +Michael thanks a lot for the patch +hello Im still having problems getting this to work where exactly do I need to place the Zip file? Im using Spring Hibernate Spring and Tomcat in advance +I just committed this to trunk and +Dont be so hard on yourself Afterall you did discover an issue that occurs between XML file formats which needed to be improved +A bit more work and I have found the missing class in the Spring dependencies directory in jee was looking in the Java SDK for the class since the package starts with working now +This is already happening +ThanksCan you please also add a test case that uses charactersOther than that Im for comitting this once we get out the door +No harm done in commenting these outIf someone prefer another patch please feel free to reopen the issue and attach the more complex fix +Resolving because it works for me now +Moving to this as issue where we should improve the documentation +The package is currently but it is in the srctestmisc directory which is even more bizarre +Closing the issue for upcoming release. +Hmm I noticed i left a in the previous patch for preflex but it wasnt being called fromThis is because we always wrap inside +se the asked a bunch of folks try and reproduce this none of us asked a bunch of folks try and reproduce this none of us could +Is it too soon for +I think its enough to have the fully resolved POM in the artifact and in the in ll start working in this direction +Only opinion so far is that is close enough to start asking whether we want to add new issues to +I havent seen it before +was the build for +Patch provides improvements to standard and logging of constructors and static initializers a new that is configurable at runtime etc + for pdfbox library +Committed +First pass at a patch that fixes we should just extend the existing to Exception +Too many routers can be confusing to users +guess is currently Derby seems to catch the IO exception and turn it into a readonly db when file lock fails with +New patch to conform to all findbugs warnings +The attached patch fixes the problem +Please also add doc detailslink as mentioned in comment th July issue out as part of JIRA cleanup +Because millisec iterations +The result of the first one that produces no error is taken +Look into the svn change seems now admin console plugin export portlet will filter those modules whose type is not car +A svn diff of the changes I made to use and test on Ingres build on linux community edition +The exception is likely to be thrown in deployment if there has been a deployment happened before and failed +the feature include all of bpel plugins I have not seperate deploy plugins from the feature because Denny said the deploy plugins contain a editor +Tryjava server Xmxm jar +Attribute State +I am not sure how this got past maven but it certainly makes current maven unusable for me +Once youve done this once then it takes about seconds to locate again every time +Thanks for the patch Elliott. +New patch attachedI think this is finally ready to commit! Ill wait a few days for any comments +Jim these are the JIRA issues discussed on todays IRC +Patch attached +I verified that the underlying surefire issue had been fixed. +Well I suppose thats really all technicalities and ways we can further improve the workflow +See? Now it sounds doable +for the proposal +Good point thanks! Applied the patch in revision PS +Thanks Ravi! +So how do I can make sure management bundle to start only after new propposed system bundle completes its initialization +but testcases that are XML related should be in the accept bucket +A Geoserver should be out soon that is based on just finished processing a tonne of highway data so Ill probably get to this in the next weeks +Very cool +Even if it would work I dont see any reason why someone want to use it +It is getting a little late in the development cycle for this enhancement aschanging +addressed in +Ive committed the code that should fix the get if doc was validated against XML SchemaPlease verifyThank youElena. +WSDL Client and log +However you cant access methodsattributes of these variables yet +Check please +Thanks Willem. +No problem Also make sure to select the right Affects Version because indeed it has been fixed between Beta and Beta m not saying that to bug you just hoping you will report more issues and giving you our way of working +As long as that behaviors documented I dont think it matters which it should do if the behavior is important and depends on calling add I can always set a manually so IMO it doesnt matter what the default behavior is so long as its always the same and its always correct +to future as it still isnt clear exactly what needs to be believe recent changes in the caching strategy for Freemarker have resolved this issue +We dont need to have a that solves everyones problems +This bug has been marked as a duplicate of +reopening to port to to +None of the functions you list have implementations in right now so theyd need to be added +In my current workspace it works for me +will address enabling it at the connection framework level. +having this at a warn level is probably too high +Thanks Vincent +And also there should be some mean to get the list of implementors along with a localized name of each +build is unstable and moving on isnt important for on later +Made changes to the initial patch and incorporated your commentsThe Hashmap change is to improve performance of keys method in cases where the initial capacity is huge with very few elements populated +Without a fix version this is not done +you mean applying filtering for the file with Maven sys props +Should this JIRA be to encompass a generic solution for all transports? Or should I create separate tickets for the jms transport work and the https transport work? What do you think? RonId prefer going with separate tasks +I am closing this bug due to lack of response +I prefer to postpone to in order to focus on others major issues +Ruth for providing the patch +and yes we will document this on but i dont think well promote its usage since its not a feature we really think people should use without knowing the possible consequences +The problem with the command line stuff right now is that I can build it but once I load it onto device it says Im missing some sort of library so I was kinda stuck on that for a bit +p +This is almost exactly what you want I would say +cc file generated compiles ok +From this point the logger remembers the generate file and delete them when their time has come +Thanks Jonathan. +maybe yes +This way it will at least attempt to inform the user of the problem +This will also be server side only change +John this means that youll need to to address this situation in older branches +Im seeing the same stacktrace when writing GB of data +This last method works fine and is a lot better than the other workarounds that I suggested +Closing all issues created in and before which are marked resolved +Thanks for tracking down the security manager issue with ijs exit command! I verified that with the latest trunk I no longer need the security manager test setup and have attached the resulting patch to asRegarding the maximumdisplaywidth issue it seems certainly cleaner to reset it at the end of the test +bulk close for the release +Its our bug +Have you made sure that it is available on the CLASSPATH +This looks like another reproducible ? added in the display is related to reading characters from the terminal and not displaying them correctly +Ill have to test some +Were run randomwriter sort with and without JVM reuse and also streaming and using distributed cacheThe results are showing a on release audit which Ive written to about +Ill add them to But Ill be interested to see how many people find it there. +Ashutosh What I fail to understand is why do we have to hardcode this value? Instead why cant we go for a better solution where array can be dynamically resized based on requirement +This would also save the time of having to copy the jars into the lib directory build the common jar point hdfs to it when running tests with no extra checking in or copying muss or fuss +Tapestry uses that constructor and expects it to invoke only the constructor of the parent class +thanks Brent! +Confirmed I see the same +this patch has it with invokeEvent I believe it might be better to reuse the instance instead of recreating them again for every notification since theI guess caching is pointless for a class which is used only twice for a full import +Via you could indeed change the used message template depending on the exclusive attribute +If you can figure out what changed come up with a fix or provide an argument to the contrary let me know +Moving out of +The redirect in same state behavior has been restored in Spring Web Flow +Lets just use most used prefixes +of course! this makes much more sense +Ive added your test process to the on trunkThe problem happens only when the the timer goes through an automatic step staright to the endWhen you put a receive task in between it doesnt happenI dont see a its pretty deep in the core of Activiti where this will needs to be fixedI will investigate it further +The attached patch file moves the loadClass method into a object and invokes it with a doPriveleged method +Its obviously guesswork all the way though because the user can add and remove listeners hosts and servers so the only way to be sure is explicit configuration +Patch looks fine +This jira is unassigned but contains a patch that needs to be reviewed for inclusion in + +I committed this without submitting to Hudson. +Reopen as Michael submitted a new patch +Lance can you update this patch and add a unit testSorry Grant this wasnt on my watch list +I think I didnt reintroduce the multiple events +Verified in ER. +You have +Please add the fix to branch as wellThanksIsuru +Drawback it need carefully hijack and also need control scope of the encoding Name contain that class changed still exists +Oozie needs to do the switch after Pig release +The site should be updated in a couple of hours +Currently these components are Message History writers extending them Message History writing functionality is some channel adapters still need to be researched in every SI module to find a common strategy or to create one which should be responsible for writing Message History +Patch attached +The patch +adds detail to logging messages so can tell more about whats happening +I didnt try type i would assume that should work as well though as they should return the same thing +I guess its just ognls fault and Im screwed +BravoI think including at least one serializer in the initial commit would be best to demonstrate the generality of the abstraction +I keep this bug open as the demo is misleading in this regard +stacktrace RossI cant reproduce this bug +believe really minor +failing test was which timed just committed this +Jeff can this issue be closed? +Is there any chance of having this fixed in? No at least not in the scope you envision see Tobias summary comment above +So it looks like i can the problem now +AFAIK a should always have an associated LM +I have created a branch and I have placed the classes into the package in the project +Suggestion to add the word String for ptBR following Julios comment and the error explanationincorrectString da formatao do CEP est incorreta +Thanks Vishal hopefully this is the same problem with the same solution +It is in the code the documentation does not say if it is supported or not then I suppose that WSDL spec is supported as whole +But I will change the above mentioned tests to this +Can you provide an example of a component which does that +thanks +This is because of limitations in PGAS header sizes and termination future we could do something like described in this issue in order to avoid some of the round trips +srcmainjavaorgapachehadoopmapreducevutil utilizes it to setup proper classpath structuresIf your comment meant some other form of relevancy please do clarify +It sounds like this is a problem that was there before and not one that breaks the functionality +It puts the importexport together in one class +Catching throwable with revision after release +Applied Drews patch svn rev docs are in but the site is waiting for new docs to be pushed to the server closing +I simply did create keyspace Foo use Foo create column family Foo describe Foo +We shouldnt be making any js changes to the package as part of portable contacts because it only changed the restful apis and not the js apisIf the type is undisclosed we can just set the key to null and the displayValue to undisclosed +Attaching patch +If no icons needed to be changed there the posted patch is final + +I suspect there are some funny issues when partitions are not the same format this is a newish feature and we may not have as much coverage as we need +Just found out that while using the default db the behaviour doesnt happen +Manually tested! +It does however fix the trunk build so that works against hadoop and hadoop profiles +This one was fixed in. + +Closing the issue now that it has been resolved. +This is required since the HTTP Service is now started to be tracked when the console is configured +Small patch to add link on screen +I think its all right ready to be commited +Patch checked in +Verified that attached ear works as expected when deployed in EAP Code verified and Release notes checked + +No +See +As a result it is now necessary to set the property in to +Followup patch as suggested by Kathey submitted +If the problem occurs only at shutdown we can also probably live with it for some time until the next framework release? That depends of course when that will be +When you say users true or false +I can reliably reproduce this error using the sample JSP provided +Thanks Lars for taking care of it. +All the suggested changes from our meeting this week have been made +WernerJustin can you provide testcasesI personally havent got access to the project which failed on this initially +So Logger Logger imported in Project and Logger imported in Project are different s in Eclipse JDT modelI think we can treat such classes as the same and then we will resolve such an injection even if projects are not configured absolutely correctly in design time and use different jars +Ill make a first pass +This has a few new classes and their tests +The issue is that I really just have no idea about the state of the attached patches +Duplicate use for QE of this will be used to track review comments changes. +All tests pass with this change +I am confused about what you are suggesting here +bulk move of all unresolved issues from to +Thanks for the writeup MamtaI logged to track some work I did on the same problem +Hello GuysIve been able to reproduce this error and also seem to have find a work error arrises when I used an Array of a DTO as return type for a Grails Service using grails witch uses groovy for compiling with Eclipse Groovy Plugin with uses Groovy the generated code is didnt try to update to in my grails installation dir but I guess I could you guys can fix this Issue it seems to be related to Groovy more than Grails doesnt itRegardsDiego a simple project that demonstrates the is a project that reproduces the error outside of Grails +tested! +Reopening to set the fix target +This is possibly due to the inputCalendars use of the displayhidden rather than displaynone styleThere is another jira issue requesting that displaynone be used +unit tests will not pass without fixing getPos +Thanks for the patchGuozhang +Am leaving this open until we get a bit of documentation in the wiki +Well done Julien this feature is really a big step forward for the Jenkins pluginCould you just do like for Maven when theres no Maven installation defined you get a warning message if you try to create a Maven job that redirects you to the config page +Are the ADF BC Wizards in maven? I dont understand how you would be using the plugin on ADF BC Wizards? It is not an project I dont believe +Previous verification issues were addressed +Need a checkstyle run to bring the code into conformance there +heartbeat response can pull the events for their task attempts from the corresponding tasks +Hey guys please delete this +I can see about getting this to our staging environment +But the problem with the approach before is that it has issues like what Santhosh mentioned bq +Patch against that fixes overflow inWith both patches I was succesfully able to load a bloomfilter +Planning to close this in the next hour or two +Any it was harder than we thought because of XYZ thoughts for posterity? +fixed in works now. +How does the test cover that the job is removed from the archives? Looks like it should pass even if we remove the call to remove in TT and the test +I have also incorporated the Callback support into the +Will be working to get this fixedthanksSupun Malinga +Ill see if I can find my work thought it may be outdated +Is the sorted by property always the same as bucketed by for a table? They are not the same +I dont want to change all these classes to create a patch unless there is some discussion on this topic and hopefully an agreement on this approach +Running through Hudson +Alan its out of the scope +This add the IN attributes so individual elements may be modified when necessary. +Ill wait until after the New Year to give the opportunity to others to reviewcomment the patch and design doc +This is more of multinode management server issue +In fact the version that is attached right now was what I wanted to give you initially +This method avoids having a another new file to hold the version and file description info put the info directly into the correct RC file and be done +This is to integrate changes for so that the user is automatically redirected to Add Nodes Progress popup if a deploy is in progress for added nodes +The ones should not nor should the disabled findbugs nightly siteAnd the JBDS Core site is missing too +Once youve finished Ill gladly review the patches +I had a look into the + removes one of the printouts of the database nameI ran and derbyall without failuresIll commit in a few days +So fingers crossed its fixed +Should be cleaned up now please reopen in case somethings still not as it should +ve just committed this. +Ignoring is a bad practice that we will need to clean up at some point +Attached patch +Patch applied in revision +I have also noticed this and It is importent for us that this functionality is working as it is now it is not possible for us to use this new betaversion of batik +Do you want more than the URL to change? Do you want all references to the word ibiblio changedI changed all occurrences of the old URL to the new URL but did not remove the word Ibiblio or change it to something else in text +I will organise a friend to test this onregards Malcolm MalcolmHave applied the patch and the +Thanks ChiragDone at rAshish Vijaywargiya +This is fine since many brokers might be creating the topic at the same time +I will fix itThanks +branch seems solid on top of on my test cluster and the high mixed readwrite load scenario +According to that we can accomplish our target using that +But then you have to be careful not to include +successIll attempt with master later but I think we need some more info on the specific test setup and for example is this network that is default for the VM verified working with another single attached vm +Its the is in null +Changing it in this patch is again introducing way too many changes +Its ok +Thanks Jan committed. +What is the status of this issue? Has the fix been rolled into a release yet? the first comment answered my question +Sounds very cool +if you need some hints just drop me a note +Okay will do. +See issue of block ids is addressed by the design proposed in +Correction of it is left as an exercise to the reader +Thanks +Thanks very much Rick +to branch and trunk thanks for the review Stack. +if there is unmapped past violations on different lines and unmapped new violations on different lines then we consider all unmapped violations as new +Fix test to query unallocated features design has been altered as part of +for history was the cause of this it seems +During the investigation of the related issues one more issue was found which affects the docs +It is not a usage issue +Now that work is underway with AS all previous community releases are +Same fix needs to be applied to the recently added and resources +Im closing the ticket but feel free to reopen it if you have something else for it +Did you try binding to +Im initially targeting this to and +thanks pradeep for contributing! +Im resubmitting the patch named with branch name +Cool +Could you confirm that there isnt problem with shaded ognlRegardsLukasz +Hi have documented the Release Notes text in Resolved Issues section of the Release Notes +Would you please help to try this patch +Chris looking at in GWT proper it seems that owner is a reserved word +will be reevaluated during the next iteration. +Thanks Shazron + to trunk. +This file has nothing to do with the actual build but does is used by the pluginad +I can now describe the condition that triggers this behavior but I suspect that its more of a JVM issue than a or Surefire issueSome native code called on the wrong ref and then tried to use it from within a native method from within an After method +Along the same lines as this is further locking in a client having one and only one identityIve often considered having subclasses of UGI that were specific +The problem does not seem to be limited to Excel export +Thanks for the work. +this helps +snjezana whats the status on this? you sent some emailschats but lets put the info here so doesnt get lost +Rewrote the description and added testcase +ive got the same problem with and +I see the images here without trouble on Eclipse and RAD +Rake runs a lot better for me now +Change the status to resolved. +Committed +flagging it as incompatible change because of a perspective this shouldnt be a big deal in practice because it doesnt break binary compatibility and its very straightforward to fix the source incompatibility for a recompileBasically it comes down to this the more we break the compatibility path from to the longer it will take users of to move to +Anyone know it +not only programmatically but the idea is to allow for annotations or any kind of configuration that you want to implement yourself as well +However in the upcoming Spring branch well build on make sure you use junit rather than junitjunit +is this always? or in a special case of using them +If there is an external configuration for JAVAHOME it will not be overwritten by the configuration within the wrapper configuration +Yes +Great thanks Chris! +DominicI tried to reproduce this with and with I am using OS X +It is modified procrunappsprunsrv +thx for clarification +trunk +Fixed in CVSThanks for reporting it and pointing out the way to fix it. +the JBDS entitlement front +Ill try the gshell commandBTW there is no varconfig in the exported serverThanks +Thank you I will ensure that I fix this +Fix has been commited only to trunk please commit in branch after GA +We should see if we can fix this sinces its going to helpful to see all your resources now weve gone to the trouble of supporting a large number of fixed List Set functionality across all parts of the site not just browse resources this took considerable investigation about how all the paging came together which tiles were involved and how postfix stuff worked removed some obsolete java code embedded in a jsp that tweaked url pagination parameters replaced by more elegant regex processing as part of the natural flow in the javascript onchange handler smoke tested browse resources and add resources to group pages since this was such a large change id appreciate it if someone with littleno bug fixing assigned to them would test this paging stuff across the parts of the site i missed +Updated documentation so that the violations are clearly marked +Verified in JBDS L +In I have also added additional filter to exclude the display of internal nodes. +That does look like a problem but I tested after changing block to have a synchronized block on runners around its body and it did not fix my issue +seperate jira as this is a pure build jars now added need to double in ER +Not critical to be fixed for release so moving it to release +ok +Added test for with minor changes in +Can you please attach the patch again and this time click the box? Thanks +I guess that should be Non +I checked +Now it use metadata +I have added a test case which returns reporting a FIXME that can be enabled one JAXB is upgraded with the fix +certainly fixed in last commits +Assigning all documentation related issues to final since they dont represent a change to the attribute is by now mentioned in so I guess this issue can be closed. +It caught my interest because I worked on +Thanks +Once I do ant clean the message go away +Stderr on the other hand works great +See also the admin guide for the system properties to emulate pg query semantics. +Thank you for catching and reporting this typo with such a complete JIRA issue +I would close this issue so you can mark it done and then just file new issues for the other tests +I think We will have to make huge modifications in Serializer deserializer and sdlws tool to fix this problem +the really belong in the Hibernate core project? Or should it be in the external OSGI container? Or perhaps someone may want to maintain an external project that generates the ready to be used Hibernate +It seems to me the schema location in this file is invalid +Found some time today so I added a unit test and applied the patch to the current trunk +Im not sure from looking at the log output if that is what is causing this particular failure +bulk defer of issues to +bq +can anyone else take this up? +As the docs would not be used yet during build I disagree to commit patch as you intend to also move the other source files for docs into this new locationWhere are the required changes to build filesIve asked myself if anything has been desided if we move whole docs over to confluenceAny news on the tool that generates static html pages out of the confluence ones? +Is there an ISSUE here? So far this just looks like a discussion which belongs on the mailing list +Changed the status from to NOT FOUND in Rev +Do you also have this problem with the or is it only with the +This might no longer be relevant +Hence the docs people and QAs can pick it up from the master branch tomorrow for their respective tasksRegards think Radhika has started work on this one +in fedora will pass soon Java +This should only apply to the patch until a decision on is that is the case summary saying Reduce the allowed Javadoc warnings from to would be a better title for the jira I think +Removed and renamed patch file as supposed to be +Verified in release notes +Please remove the dependence completely +I meant to close this bug but kept forgetting it +If there is a problem please open a new jira and link this one with that. +Lets continue this there for now +Affected users should upgrade to Karaf or replace with Pax Logging +Just because its the least risky change +Sort of similar to dbfile just using S instead of local filesystem +Please provide more specifics on your youre right +idle timeout handling was reimplemented in so open a JIRA if there still is a problem +Can be the same as still here but wo error message box. +Resolving +its already committed! Please add the addSuppressed because we should not swallow exceptions +Adds get and get helper methods to further simplify parsing code +we can add such text in goal documentation but Maven itself wont be able to give a nice message the minimum Maven version is checked at global plugin didnt precisely test to see the error message when it will run with Maven I suppose this will be something like NPE did you tryIf we get an NPE we probably can improve the code to report a nice message +updated with trunk +I guess I should have tried that but it is the wrong group id so didnt consider it +It appears to be useless complexity perhaps a remnant from the original patch against +Otherwise I had no easy way of demonstrating the current issueThe actual fix is described in the issue description +Also we are not asking for a specific intercepting clause just the spec being a bit less specific +Unless there are specific problems that where not addressed within the original JIRA I see no reason for keeping them in this state when most of them where fixed and resolved almost a year ago. +This update is just to make a tiny refactoring bringing all the handlers parsing classes under srcjavaorgapachesolrhandlerrichand all the testing classes under srctestorgapachesolrhandlerrichAll tests pass +hence Im not making it PA +Yeah this walks the tightrope between bug andenhancement +Yes this would be a useful issue so that users can use the same script to update passwordsNot sure if you are familiar with the script implementation but the utility is implemented as a set of states in the following class Any updates should maintain the use of states but there is already a state to check for duplicate users so that could be a useful starting pointAnother class that you may find useful is this contains a currently unused method to a properties file replacing the users details as it is currently unused it may need some additional testing but is a probably candidate to pull out somehow for +That was my fault. +table in one transactionTo clarify it is not one transaction but one batch RPC right +Also resetting the dataProvider may be a problem as well +The twitter screens are loading tweets per page +Otherwise patch looks great +Yes it holds references +Some confusion hereSample based means divide by N Population based means divide by N Is this our common understanding? DB stddev divides by N. +I tried what is recommended by this thread but it doesnt work well +Well that is possible to tell if the needed modcluster modules are loadedBut we cant tell about conflicts with modjk as the routing tables are received for the nodes +prefix in trunk and I missed itIndeed +Closing all issue was fixed a long time ago but the constructs linger in all and they indeed manipulate things underneath +What is more important than priority is the fix for tag which has been marked for +Maybe something has changed can someone confirm that with the latest versions of the plugins? +Thanks Otis +Thoughts +which server types do we set as requirement for the Java EE +Perhaps make the registeredEngines available as an unmodifiable Map +Do we have any problems with this? Eg our Camel releases should be compatible with and of Karaf +I will have a look at this issue +HiCan you give me any updates on this issue? +This got ignored for a while +Please make sure that Affects Version and Component is correctly set so the issue can be triaged +If there is enough demand I could cut a new JBC release with a newer version. +There is already hbaseregionserverShall we name the test as hbaseregionserver +Why does have a shutdown hook though? Its not something thats runnable standalone +reopening for the ll review this for the next version thanks! +Be sure to apply both the and + +Attributes will never be a parameter on an operation. +That could be just not showing it on the form but I suspect that cutting it out of the release entirely might break things? I could be wrong though I just thought I saw it referred to in the code base +Sbastien already pointed out his Polynomial implementation would fail with them and there is a high risk would also failI would suggest to just do a quick test on and if as expected it fails then drop the idea entirely +Thanks Owen for review +Committed to trunk at revision How about defining two different datasource one is transactional another is not +bq +Written the analysis feel free to review +If its absolutely necessary to support older clients a server patch would be required +It might be the case that multiple builders are acquired from a node state passed to the commit hook +I just update target version to +Anyway considering that we are near the release lets consider this in the next release if release seems we need to the issue of this jira reported any new comment? +Here is the patch for it +This seems strange for me because i only called the merge function which should not be influenced by the PERSIST flagno +I will try to implement one later +what are you looking at specifically? +Maybe the goals from failsafe can just be moved into surefire +All previous tests still pass of course +Looking forward to your proposal +A new type of failover support noop was added because in some operation occuring during the failover process itself the retying failover support is not appropriateThings do not cleanly in every case +Im hoping we can minimize the number of authentication calls we need to make +no problem to discuss but especially for tables such change was really important +I think write access for members of generama would be sufficient. +To make this simpler I would like to propose splitting validation logic from parser and make the validation extensibleI will come up with a proper proposal as soon as possible +having fun +Closing duplicate issue +After all all the projects relying on these tasks will use the corresponding tool right? So we can safely expect them to have the dependency defined and reuse it +And iOS is not yet using the logger for the console still using the old debug console plugin +Pass a column family name minus the creating HCD +Within the field myBreed is an enum type but in srcmainwebappsviewsdogs the corresponding form element is an input not a select + With this change I will be able to lower the log level in to INFO +This fixes it sorry about the delay I had not peeked at the JIRA this weekend As I sort of stated in the previous post this kind of error only shows up occasionally and in big jobs so if Sebastian could this same job on Giraph this patch we can confirm the problem is solved +YepI can do those things +Im moving all bugs out to and we can move some bugs out to to get some release action going +So the option is always there to cause some havoc +It shows that we no longer through an exception on this condition but it does not have a great need +This was not originally tested so i also added tests to verify the fix +Have you tried it against a few hlogs Kannan +So consider this work in progressAll tests pass and Im currently running them again with the integrationTesting profile +Thank you Raja for the review. + test case will follow in a separate you forgot commit it to trunk +We should be able to just point at the UIMA libs in under contribuimalib instead of having to copy them around +Attached patch contains a test reproducing the issue +in Admin Guide doc in CR for CP +This patch obsoletes +Ill also attach the simple case for the since that shows how to fill out the maven project stubI should also say the maven project stub is really dumb at this point and as functionality is needed on it we can implement that as we go +I have two versions of Railo running on my machine and I think that the test case was running on a version and then the other machine was a verision +multimodule project same git root multiple moduleswar datasource file same git root multiple modules +The Delete button seems to be gone +Verified with JBT. +Hi DagYes I believe that encryption is a defense against this attack +Here its throwing exception with err message and I also feel concatentation is fine +Other than that it looks good +Need to set up a test config and ant target to run youve got the open subtasks so Im assigning to you +By using a threaded model as I do in the youre not +I also need this for and my first draft also is quite a lot of code for basic infrastructure +Adding a new patch +Its divided into two discrete cases depend on whether user is anonymous or not +I just made a change for my local code +Sounds sensible +but i am failedThank youand i seeIt doesnt allow guest traffic on multiple networks +The state space is very large which causes underflow very easily +Also fixed a style error as pageCount was never initialised +Hopefully someone someday will track down where this message comes from and customize it +If it turns out to be quite a bit of work we can do that in too +BhushanThere are lint problems in the patch in +no but a patch test is awesome +Some are modal sheets using javascript and these provide a cancel options +All of the designs discussed so far discuss single client ordering of namespace operationsdata writes but not multiple clients +Attached another patch +Committed to trunk +Ive verified new build on Windows and Linux all tests passed +Patch to fix streaming interface on Windows without Cygwin works with and without Cygwin with hadoop core on Windows +same issue at my place +You can put the ldap specific properties in a separate and people can then configure their own LDAP server etcIs any other committer interested in helping with this issue? If not Ill assign it to me +the test passesfails at random +This patch adds this line to the buttons style pxto prevent overlappingI also noted the same problem in syle so I added pxTested on Firefox and IE on Windows +Changed to feature lower in the log when the ESB service is deployed +Thomas just to confirm I can reproduce using your uploaded repro thanks +more in rev +Reassigning to Marko in case its reopened +Can you rebase please +So if you add a file to spring and add it explicitly as a Spring configuration file it wont work +This has been fixed a long time ago I have just retested it with the current build +Yes I was not able to reproduce this too second time +might be a legitimate case am probably too tired right now to think straight +Seems like moving the test to jdbcapi and renaming the class would be good +Committed r to switch from HSQL and H to Derby in r to switch from HSQL to Derby in did you get them all fixed now +We are happy to do so but please let us know more precisely what to enhance +Moving out of +this could be handled along with other caching changes +Ah yes good observation +How we can set the identity file to use in configuration? +Kanna I added some more comments on reviewboard +C samples would hopefuly be available in or in this issue since description is incomplete. +classpath file like you will do inPlease fix it in once action too +Worked around by using a switch statement in +Committed to and HEAD. +Thanks Stefan for updating the patch I tried it with my test cases and they run fine as wellSo Ive applied your patch in revision . +Tim Yes looks like closing this was a mistake on my part +Fixes XA encryption and a couple of other tests on +Back when JEXL was first developed it may have been an EL superset but EL has had at least one release since then and includes functions and other stuff JEXL doesnt doThis would be a welcome addition +Anyone Ive talked to who has been concerned about over the wire has also raise the on disk issue +What everAt this point I dont have the time to go deeper into this problemWhen I get some more time Ill do this. +So we should just remvoe worked around the issue instead of documenting the licensing Ive been able to remove all of the artifact listed under the an unknown organization section Robert you say everything else in the distribution which isnt described in the LICENSE needs to be reviewed +I have updated the http wiki pageThere is a new table for component options as you need to set this on the componentFor example these options is component level only protected int max protected int connections +I dont get the OOME anymore allthough now I see different BTW move out Seam of snapshot! +The same thing happens with the beta But I took a stab and the IDE and now it works fine +Actually I have created a repository in my local system where has also been installed to check if it because the server is in a remote +Thanks +I think it looks good for any framework if it stays current on its dependencies regarding patch releases +Should we specify a parser loading policy or similar to? This could have values like Continue silently Log problems but continue Fail if parser class missing continue if others missing Fail if parser class missing or if its dependencies revision I modified the class to parse the available service provider files and instantiate the listed provider classes directly instead of using the imageio class for that +corrected this patch is for please proceed as suggested by MusachyId like to release tomorrow if nothing special happens until thenThanks committed only the change to +Also a ve test case which rejects date like as partitioning column will be good to include +tested with internal build and committed atRevision Author dantranDate AM Wednesday December Messageupgrade to +Changed the method name and committed to. +Do you have any expectations about the performance difference compared to visitorsThe new Event bus uses heavily the hierarchy traversal +xmlconfconfslavesbuildruntimelogs +I have looked at the patch and it looks good +Once creating a symLink is exposed through it should also support resolving symLinks and reading data through themThis could be either done solely for or universally exposed through + +That is a later NONPERSISTENT message may arrive ahead of an earlier PERSISTENT message however it will never arrive ahead of an earlier NONPERSISTENT message with the same priority +I did not restart the service after terminating the additional monitoring +Not for commit here. +Thanks Ill mark the bug as closed. +I would like to add in throw trace as well +Ive copied the artifacts to the live repo now so they should be available in an hour or so when that syncs up with the mirrors +Looks like this has been resolved in +Varun if you want move that back to feel free to start a new issue. +The patch is good for and too +The result is displayed in a dataTable but without any CSS +I had issues with other projects than James when code depends on xsddtd referenced as URI + +I find that there are some some keyspaces usingand then I drop these keyspacesthe decommission can work okbut there is another problem +Whenever a consumer crashes its message groups are reassigned but when this consumer comes back online it will never again receive any message +Testdata file to show the problem thanks to Julia Dobroschke could be used freely +bq +Why do you need to pass the full node to the? Wouldnt passing a be enough or even just calling a method to indicate that a node was scannedAFAIUI you only need the ratio of nodes scanned vs +In addition it will be better if there is a way of checking whether some file is already added in DC +This will cause an NPE when the route is actually called but the root cause is hidden +Otis Id recommend that you look at the patch for +Good work +Still investigating +Not sure why or what is different on my desktop +Alex since its your optimization feel free to if you dont like my solution. +After this patch what is the purpose of the type ids and names in C and C? Thanks +I see the problem but I dont see a way to solve itTry to set the first attribute of the datatable to a meaningfull value after you have changed the rowcount of your datamodel +It would also be good to ensure that an add plus a list delivered the correct result +It would really be helpful if you could provide some more information How did you create the configuration object which properties were set and which key did you pass inSince the last release development has been going on so the line numbers are no longer validThanks +I dont think it can help this feature only makes documents sorted so this cant be used to get documents globally sorted. +Ahh got it + +However the fix is a way smaller +Attaching patch while is down +A parent should not be aware of its children and its semantics on direct calls should not be altered by children +Still they are not meant to replace the native RPC implementationWhat am I missing +Thanks for the additional thoughts Dag +Could you please test with the latest svn and see if it worksAlso what was the exact original problem? Generating the code or compiling the source? +Initial commit in Git ID feedcebcbb + +Included in release. +Its possibly one of the default encoding is not changed by setting the jvm is made inconsistent internally due to setting it just happens to work in some vmsI dont think it has anything to do with running in a separate JVM is not for use in setting the jvms localeSee Suns java bug Btw I see three tests still being run under the old harness not just grep junit +I think it is ready to pursue +Take a look at the linked JBSEAM issue for an that is true I thought about that after I entered this issue but was not sure what additional requirements the dynamic resource work you did might play on this +sample includes use of configuring EJB security via annotation +Thanks Tommaso its compile and worksNevertheless its seems that you dont have commit the patch providing here +Please verify +Actually I need to rectify my statement from aboveExecuting DDL statements on an XA datasource is a violation of the XA contract as they imply an +Hi Eli and SandyI apologize for not having checked this issue carefullyThanks for taking this and writing a patch! +Try defining the type attribute for your id element and see if that changes I dont get any feedback Ill anticipate this is an hibernate usage error not a plugin error +of the calendar navigationI hope this helpsVincent MATHON +Thus Im very hesitant to change the default at this stage +I have left out jar and Spring jars out of the WEBINFlib directory so that the size remains under MB + +I just committed this +Obviously all the extra remoting features such as callbacks leasing etc +Am I missing something +The operation name is used to map to the Teiid procedure to execute +Do you know if there are any plans for that Pete +This issue is reported against a version which is no longer maintainedwhich means that it will not receive a fix +the fix for this issue duplicates resolved + + +He also has run all the unit testsJitendra and I have reviewed the patch +in in your error message you wrote documet instead of document +Ah it appears Tims issue is on so this could well already be fixed on a similar to memory leak scenario in the prefetch subscription. +Displays the message from the exception caught +Moving the file to the bin directory caused tests to fail +I think it makes sense I have no objections now but itll take me a few days to review it +Excellent +Sorry I forgot that children was an array your makes senseIm on the latest patch as well +Ramesh can you provide template examples that I can use in the designer wizard? Ive also asked Chuck for his input +still not fixed in +changed make to return! in +No IT +Patch applied +marked as optional are you still working on thisCurrently this is not a blocking issue for the Maven conversion but Im leaving it open in case you still intend to do this +Indeed there must be a way to inhibit server cleanup for diagnostic reasons +You can also debug on the line I mentioned previously to see what describesplits receives and then outputsJust wanted to facilitate reproducing the problem +The global configuration parameter is default value is set to secs +is not discussed +I just debugged this last night and this morning and found the same thing thatWerner found +Seems like this issue related to +Im leaning toward the first solution for the branch due to stability concerns and Ill investigate fixing it differently in trunk +instead of throwing exception could we just set the minimum of to the recordsMax in the +tests are almost complete now +The reason I want this is that I cannot find a good example within seam where processvariables are set +You need to provide a unit test that shows when this might be happening from a cursory search through the code it appears as though all calls to messageExpired do increment the expired message count so doing it again there would make the numbers incorrect +Hi AndersDo I understand correctly that this will only work for Maven ? +No per default all configurations are required so that if one cannot be founda will be thrown +So synchronizing on this or an object in the session wont work +It kind of negates the point of having a TTL setting in the Producer and is inconsistent with the other Message properties like priority and delivery mode which are always defaulted to the values set in the Producer +When I locked down the version for it I managed to build the site +Do you have key cache on +Is there another reason for keeping it +He can take a look on Monday +I doubt were going to get those warnings down without a more major overhaul +I left a comment on the reviewboard +The reason I was returning negative values was to allow builds to use positive values for their return codes +This isnt REST because you do abuse the HTTP status codes +the script will not work without first overlaying the Common and HDFS target directoriesYes this is because the scripts assume that there is a single HADOOPPREFIX directory containing bin conf etc +This is misleading +I hope everybody agree with the statement We can not detect http response codes based on responded html contentPrune the index is a good idea to solve the problem. +just a personal preferenceIll see if I can figure out how to get up and running +This information has been updated in the book thanks +Ive just gone through all the dozens of things i have installed in my Windows Program Files directory and not a single one has any licensing information at the top of the directory however i expect they all have licensing information in the Help About menu option which is a Windows standard +just committed this. +the current behavior +Hi Jervishow can bypass the auth popup +Alberto Did you get chance to look at this +That approach would imply scanning comments until ATTACHMENT ID is found right +And just to add ONE MORE comment here is a link to what seems to be the related bugfix on the WAS side httpsupportrsuid Ill give it a shot and get back to that suggestion of mine was nonsense +One idea we considered for future work is to define operator on bags to match SQL semantics +Then you just open it though the Open Project action +If you are refactoring can you disable it behind a flag say a strict is the source code of these patch +I plan to commit this soon +Ill hold this patch until the uddi fix the problem +Hi TruptiCan you provide more information may be a screen shot of the resource manager ui you are referring to +If it has an XSD clearly the user wants validation if not clearly the user does not want to specify oneNice idea Patrick +Look for a sentence The property in the above two files should point to two directories in your local file system +The patch also contains updated tests +It turned out to be pretty trivial with all the plumbing done already for the server side +No errors in the logThe simple history had result code for every link +This contains some jar files as well as a result we could not submit it through the normal patch method +Would you cross check the patch and close the issue? +test fixed +The latest patch Im using. +Now that work is underway with AS all previous community releases are +Remove handling for distributed log splitting +I find the module I want right click and do a publish +Marking LATER pending course of action +for the patchVerified results with and without your patch on Windows +Thanks for this patch +but more complicated expression would not run and it seems like there is a fair amount to get it all workingHas anyone taken a look at this recently or made any progress on this issue? Feedback greatly appreciated. +Filed issue +Create tier networks and private gateway +Missed a plan rev +This was fixed in JPOX builds m getting this with the current build downloaded an hour or two ago +This patch adds a NULL constant to Reporter interfaceAdded constant +Ive been there! +Will be part of the tonight +It would have also been better to fix the macro to take the arguments env and exc although I know it is already using env implicitly +Here is a test case +I am not sure how I can configure it to use local +This problem is not related to Commons Configuration +Thanks +Given the test failures are unrelated I committed this +I suggest to open a new issue if there is other problems with relocation. +Another solution might be inheritance although I hesitate to do that unless we have to +for the patch +The only downfall of this is that global configuration flags are usually not used for something so detailed +Please with the ESB team on vs processdefinition +putting into so that such an extension to the POM can be made so that this can be made to work didnt investigate enough originally and the attached IT isnt complete to test this +I commit the patch +By Java convention the names of static final variables are usually +Please provide update on your items as a comment +Dan this is a bizarro exception to be coming from +Connection refusedConnection refused connect +Committed Part patch of Phase +Patch +That should perhaps change to something else +Though that would involve some heavy refactoring and since you estimate this issue as minor then do we really need it +Fixed patch path +Patch is trivial and looks good anyways +Once is released the branch can be deleted and if any subsequent bug fix releases are needed they can be branched from the release tag. +Consider using the newly released logj instead of since it includes the long awaited trace loglevel +All I did was set JAVAHOMEcprogram filesjava and run a test +Good point LarsWhen users run cp in the linux file system against hard linked files it will copy the bytes rightI think we shall keep the same semantics here as well +Bulk close resolved issues +here is a updated patch with some i see the is for the output +wow that is a really dumb error! vishal could i ask you to modify your patch to avoid code duplicationwe should probably have Request si outside of the if and then just set si inside the if statement and then do the setOwner and submit after the if block to avoid code duplication +The nested javac task of Groovyc takes include into account +The only transport remains HTTP which is stateless +Closing +Fixed alter table syntax that was causing problems withDid not fix index creation errors they are harmless +Sounds like in step you deployed +I would expect the fields to not be empty when using from centralIs there any default handling now? it always seems to be blank for me +I believe this happens when you the tests without restarting the broker ude to the queue not being deleted +it is fully backward compatible with any persisted data that may be out there because it can still successfully receiveread the old format +Thanks to everybody who contributed to solving this issue. +So now not only do I really really not care whether this goes in I have noqualms about it eitherHaving users like Shai who are willing to recompile and regenerate to takeadvantage of experimental features is a big boon as it allows us to testdrive features before declaring them stable +Resolution will take too much time to make it for the to The latest changes to the security model will make it much easier to implement these but well not be able to do that before the release +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Done +Here is the process file and wsdl file +If you know an elegant solution patches are welcome +Howeverthanks for your help +did you add any specific plugin to the install? Map preview works just fine for me +This issue also exists in r of branch +Good point +This patch could benefit datanode decommission too +Done +Tests passed cleanly for me on on Java +On a related note am I correct in observing that this patch does not cover the deployment of libfb +Its traditional to use an arbitrarily assigned but small number for each zookeeper node +This obviously wasnt fixed in the past +isnt included into the is the only jbpm plugin included into the Runtime plugin isnt loaded +Currently we are using jdk to build +This bug has been marked as a duplicate of +issue too general to be actionable. +It doesnt do that now but it would be straightforward to add into the parser +I have not exactly applied the patch from Julien but used the idea of reacting in the expired event +Attaching a patch which will log the Engine shutdown message in +I verified it +Checking +Ok +This change was undone with due to a lack of consensus among the committers +resolving this +Attachment has been added with description Sample FO file that demonstrates the transition resolvedfixed bugs to closedfixed +I have applied it to trunk +Log aggregation is leaking which I fixed in this patch but something else is also leaking +Thanks +However Ive reworked to the previous status value at the end of the tag So in the case of nested the correct status object will be exposed at each scope level even after nested tags have would still argue for some sort of configurability because of a situation Im in where both springbind and a tag Ive gotten from a third party both expose a status variable +The classes are based on the classes perhaps some of them could be reused instead of creating new onesCurrently we have to give a list of urlsfiles the applet will loadIt could perhaps be enhanced in a way that only the sements file is loaded and then all other files are loadeddetected when requested by the lucene engine +What I really want is an easy way to create the source artifacts +I have seen several locations where the defaults are inlined in the call +Attaching full test results +Im closing this as wont fix due to the complexities mention in the description +You can go ahead and make it read whatever you want just so long as it includes Jeff Mesnil by name specifically +Should it be reopened for people running WAS? +Fixed in revision +Note Will work for Sonar +It is loaded whenever it is not found in the cache +if the response code is not do we need to return the bodyYes +I dont understand what the last comment is saying if you believe a patch is needed please attach it +go for it +How much of these plans is still relevantCreation of Cocoon blocks is handled by Maven archetypes and we have no in Cocoon yetCould someone more involved in this comment +This provides apis to create files based on the pair name the final output leaf file etc +See for database schema generated in trunk +Further improvements coming. +Refactoring and minor changes to move rss support to abdera extensions and make it part of regular build +Toshiya the support case appears to be for can you confirm that a fix for is also being requested +It does not help our users if we endlessly debate about the best solution and leave our users alone while we do that +That should now be fixedIt did uncover an MVEL bug though where its not enforcing type safety on the addAll method +svn hidden directories so please use the new zip file +Yes it is maybe you have mixed version of libraries +Ill try that patch too DidierVery good timingWe are discussing this issue inCan you please add a quickstart with your requirement to it +Thanks PierreI have reverted commit Jacopo +This means the content of a configuration will not change due to a reload while you work with it +Attaching a rebase for the branch +BULK EDIT These issues are open to be picked up +Brock Id like to get this into +Run mci with Weld to reproduce the issue +foo resolve to foo when on bar +Debug error update on this +Moving back to for now +But if you have more than one machine at your office maybe this will be possible for you? Its a challenge for me because I dont have the hardware at home +Patches applied in trunk r. +Patch was committed to CVS +ffdcaaabbde Fix clang build with exthashmapBuilds for me with clang Apple gcc gcc and gcc +i dont think it is unlikely that the treecontentprovider was calling the server during startup where jmx is not yet available +Should I open another issue for the optional creation? I might get to this one this week too +Ill wait for jenkins to run then push it in +Up to date with trunk also now reuses the instead of rolling my own +The build was added when was done +Hi ok will verify the solution +I dont know that Id say its acceptable for the transport to pass a null array to the when the client supplied a zero element array to the transport +fixes the issues mentioned in previous comment +Passes mvn verify etc +The encoding was not set to in my eclipse project when I applied the patchI changed it to reapplied and now it works fine +Esteveyes of course you can work with the API and construct any xml payload you want to send to the service +For the the user has to specify their name using the name parameter in the request or they will be greeted as null +Nice catch + +What would we get from not making them different classes? I mean if we were actually writing these classes and maintaining them in the codebase the gain is obvious +cant it be configured through +All works fine. +Code looks good in trunk +Closed. +I have successfully run the additional unit test and the ant on both trunk and branch +for reopen +Perhaps the installer or the classes on windows are not the same between linux and windows at the moment in the JRE folder or something else is going onI will be doing all future work in a centos linux environment since this seems to have better everything. +well +The current choices are disable or terminate +Im thinking of how Jetty prints out the location to its log when it is finished starting up +Applied patch in revision +My only fix so far is to just removed the unique constraint Hibernate creates but itd be nice to have that still in there +So it can be usable in your use case I have added an option to not show popup on resolve errors. +The reason I used MANAGE was that the type of operation is not determined by the object or the operation but by the name of the method so the operation token has no meaningThis would also interfere with the ability to add +Thats fine +Please use spaces per the Hadoop contribution code guidelines in future Removed an unnecessary import that carried on from a previous patchThanks again! Committing to and trunk +No additional memory or CPU requirement for the primary namenode +Restore old copy of patch +trunk +Which revision of are you using? Because it seems I cant reproduce could you give more details an URL would be perfect +Hi Patrickwith the client port may be correctly defined in the dynamic configuration file it doesnt need to be in the old config file the server will not detect that anything is wrong because everything is actually fine +All pieces in place +Thus a read where B C participates will see data that will be reverted once A B happens to be pickedNote to self Think before posting +Reopening this for the +This would be hard to fix as the message comes from the Xerces +Operational AttributesEntries may contain among others the following operational attributes +this is veeeeery strange error +Yes it is +Is this outdated now? It should display an error message if you dont have permission to do something +I agree totally the only reason I didnt do this was because I dont really know the code well enough to introduce semantic changes Ive never been a fan of swallowing exceptions But your changes to my patch look great +on the list is just a thread ongoing about the renamingrepackagingOnce done no adf jar is leftThanks +FreemanExcellent suggestion to add the repo location into the bundle definition RegardsCharles +Fixed in rNow component type of Java array will be serialized as needed +This is strange indeed it should not lowercase the URL at allMarked for the way you can still use solrindex command it does not lowercase +I will double check but I believe that capability already exists if the parameter is not one of the alias I believe we do fall back to being relative to the configuration folder so it should be possible to specify any other file in the history +I would be interested to see this +Dims had fixed it once the build broke +Uploading an initial draft of the patch that includes +Fix is now in CVS +Marked for keeping it on the radar. +The assumption is that different tables have different schemas +thats why we added the know there are valid scenarios where the warning is appearing but I dont think this is a valid I may be missing something +Could we change it totimestamp? also provides this capability beside from being confusing it also requires a valid SCM which we dont have sometimes specially when we need to build from source bundle deployed at maven repo +fixed in brew build with CR +This patch is not ready to be committed +Yes but in this particular case I thought it might be better to explicitly change the file name so other developers dont fix the media type for that file in the future +Other than writing javadocs we need to replace and I think +Sure the wrapped data gets reset when you invoke a method on the owning component but that means only the owning component can use it +If you do +Sync with BRANCH done +I havent looked in detail at the new annotations proposed by the draft but they look very similar to what we have in the Felix projectSo I think we will at least support them in the plugin that should be easy as we just have to map themHowever our annotations seem to be more powerful as we support metatype creation at the same time and seem to be a little bit more user friendly +the dumb of are completely to either filter them out find the magic derby voodoo to turn them off document as known +Attaching patch which implements this for updated patch that does not change the current encode behavour but adds a encode method that behaves has described in this issue. +Is there anything special on this project concerning unit tests? are you able to attach to this ticket the content of the target directory +This should save one allocation for most tokens +You do not need so many negative tests tests take a long time to run +checks the validity of each path and stops the job submission with a message to the user if any directories are invalid +once the Test annotation is in i think this is good to go +Something like that sounds reasonable +Otherwise isnt there danger youd overwrite other encodingsYou probably also want to pass back the code out of your new public add in case you need to add more than one methodclassHavent tested +Net but doesnt have the right defines +Can you attach the patch for the concurrent async delivery? Im not sure that we can get the you see as events have to be delivered in order if they are posted by the same thread +Jirka this is Spring yes +I will commit it shortly. +I dont doubt the performance of the SSE CRC +In my usage case we are using one session and creating consumer to retrieve one message from a topic before closing and throwing the consumer away +bq +Wow thats really odd +Creating with your customized file name filter for readers allows you to decide which files you want to use the caching layer forThe package includes some tests +then encode the content with +That would do. +When the methodbinding is executed in a following request this value cannot be accessed +No probs +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the fixed documentation wont be deployed to until Grails is released. +It sounds like the best solution would be to return an Agent instance with minimal capabilities in the unknown Agent case +Thanks Andreas I will check it +The patch is far from perfect and just uses the class to parse theURI passed in +I would prefer our objects to not require commons configuration +In the compatibility guideline wrote In the first instance we follow the description of each part of the class library given in the Java specification +in this case it is fine since the transformation gets applied but if someone were trying to write some client code against complex features that would probably be something that slipped them up +It makes sense to call that in +They are a concept on their own and it seems you want to use them instead of normal files +Moreover the signing certificate is expired from +I am not a maven guy if you have a suggestion Ill be more than happy to try it out +Marking as patch available so Hudson runs the tests +I didnt know method backing action didnt take an it +Thanks RatneshPatch is in rev . +As in all things though this is just my opinion and Im more than wiling to be correctedUsing the jndi name also has the additional benefit of leaving the well recognised url of tcplocalhost untouched as well as being tried and tested thanks to Edan +Can you attach this beans test to the bug +Once a consumer is closed the messages are redelivered to another consumer +Attachment has been added with description fo file illustrating the has been added with description Patch that fixes P open bugs to P pending further priority for bugs with a patch +Thanks for the test Korusaka +Great minds think alike +Company will close and I have just one day left +Ill commit when subversion is working again. +Can anyone shed a bit more light on this? Any example in ASThanksHoward +Im going to fix the is temporaly not available +Ill get get a interface to get that fixed +Test case succeeded during automation run todayVerified the same in local with latest build +The fixed in CVS +I think I will address this issue by removing the default URL from OBR completely +I think this is ready to commit +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be would be very helpful to allow for port setting in the build config depending on environment or even the db libraries +Okay thanks looks good +There will be Java and Java specific for EAP created by independent builds +checked in +Thanks for the additional info Kristian +Documentation needs to be improved +opened to clean up the backward compatibility stub from. +I have committed this +This details need to retrieved in the +hmmm +Things break down when you do a new install or upgrade +For now Ill go with the original request +Yet I think we should be able to provide instructions on how to use a different epsg provider they are pluggable by design +I pulled this into. +Jochen said this needed reopening on the mailing list so Im reopening since it is critical for Grails that this is fixed for good and hence shouldnt be forgotten on the for Jochen to provide the details of the issue he has seen with it +Final patch. +The Foreign key circularity +Its that were holding the namesystem lock to coordinate with an HA failover especially considering that there is no actual namesystem manipulation like Daryn pointed out +A suggested applied +The piece of code you have brought into attention would be gone +My badI run it in eclipse +But I have many objects on cache level +replaced by +In this case we maintain the invariant that all of the current batch of edits are written out or none of them are even when the disk starts getting close to fullPreallocation was introduced by Dhruba Borthakur in in +Updated the issue description to reflect thisObviously we want to make this less crappy but maybe its not worth making ZK a dependency anymore +My favorite one is with the grouping boxes which definitivly looks better than the auto generated graph. +Robert patch applied at r thanks a lot please thanks a lot! +hiI would like to mark this JIRA as patch available for the patch I uploaded on August th +That is a special case of an issue on edit layout +which helps but your build is hosed upI would like to resume my PERFORM A FRESH CHECKOUT AND LOOK +I have fixed that exception +if this JIRA is to capture work already completed on please close out with details of that work and raise a new one for any outstanding implementation required? Thanks +slfj from to as jboss AS supports only versions of slfj api and impl. +The link provided is an ASF link and it is the link we encourage participants to useI would greatly improve the user experience if we could add the following step to the instructions under Compile and DeployImport dependencies with the following commandmvn P deps +I think Ernestos patch should not be appliedWicket gives the control to the application code by giving it the notify function as argument +We open a bottle of wine +Please review +Attachment has been added with description Sample FO file +The method as proposed isnt general enough to be useful to more than this context +IN clause with a table subquery will be addressed by +Patch to fix all three scripts +Proposed up whitespace +Im running the tests and +Both approaches are valid for small clusters SSH is fine but for larger clusters a blobstore is better as it is not limited by the connection on the launching machine +Looking into this there are not only a bunch of methods that have been added to with Java they also refer to new classes of package +This has simplified the generated code +This is activemq configuration file that I use +and I can export patch successfullythere is one thing that might be improved for code generation +Let me know how to improve it +Otherwise I really have no way of this issue +Attached is +The call queue length per thread is with this setting we have a overall call queue length of +To do any more would completely break the current behavior with respect to fields however Im really not sure we are doing the right thing +I hope it is understandable even to those not knowing Scala +Committed revision . +HDFS isnt really an application +The spec says transitive should accept project and return all its packaging tasks and it passes +. +The issue could have been deleted it could have been moved or the sequence may have just skipped itIf you want to prevent people from moving or deleting you can remove that permission in your projects permission scheme. +Patch lgtmRather than add new API though should this just be the default behavior instead? On second thoughts maybe not +Ill let you change the component type if thats correct +I thought you were referring to a reply from this infrastructure issue which I seem to receive fine +Whats the exact version checkout +Just wanted to make sure folks were ok with splitting the command as it isI guess the main alternative would be to add more flags +CVS version of should be included already in final release +A temporary fix would be if the email sent out at the top or bottom included a direct link that would take you directly to the online web diff viewerNot sure if that is easier to achieve? +attach an example that reproduces the is symptomatic of a test not being the in the directory which matches its full name +Completed At revision Works for both and commentsLicense header added to signatures text file. +These adapters are capable of deployment and providing classpaths for projects +As Paul says loading the DTD is required for things like entity definitions or default attribute definitions +After resolving this bug our cluster is stable +This issue has been fixed and released as part of release +example IIE sticker example for ACCELERATE sticker design +Thanks for reporting +For details see +Mohammad on your clarification question yes only for logging +Ive checked and it is not documented there add it to the docs otherwise no one would ever find this feature + checked in for xdocs +Approved for +What are we buying hereIf the underlying need is to have an easy way for users to load on all user tables then I think it would be easy to add another configuration property to enable that +Patch applied. +How does this related to? Seems like if we implemented then we wouldnt need this but this is a much easier ticket right +With NFS as long as the client application is not performing random writes the linux NFS client does not appear attempt any random writes +The only questionfeedback I have is around the separation of regular args versus the kwargs in code +That issue was reported for the EWS +haha yes why not +Ok so that one stays in SVN then +Need more time to investigate potential ramifications so bumping to after maintenance release +Confirmed by test +This looks very dangerous! It does not obey transactional boundaries and I didnt see anything that couldnt be implemented safer and more scalable than the suggested patternAnd the merging sounds very dangerous in any other than the very simple cases where caching could be much better +The improvements Jacopo suggested were included in the patch found in which was committed and closedThe issue of where to keep a users time zone preference was addressed in +You are correct +This patch also fixes the class +the patch causes failures in new patch should fix the problemI attach also performance comparison before and after the patch for the gcd methodThanks +Common POM parents will be addressed in +Bulk closing stale resolved issues +This is the correct behaviourEnd users WILL go into production with the welcome context enabled this is the reason the redirect is only possible for simple scenarios where the same NIC is used by both and in those scenarios there is no guarantee that just because a user can connect to one they will be able to connect to the otherThe only other option would be that where the welcome context is enabled we actually add a config option to specify the redirect address to use but that will still be error prone +doc format for compatibility +You can always call flushtrue yourself +Closed upon release of. +Newer and cleaner should be done in a separate jira +Reverted back to original oid names +Instead I see it as a way to get a huge speed up on a very commonly used feature +Grid can you look into this one? It seems that for some reason its unable to find the implementation class package folder but Im not sure why its throwing a fit if the custom package name is an empty string +Attached is the logs for the day +I think I commented on another JIRA that I have implemented a dfs fuse module for the straight C fuse moduleYes you noted that in +Hmmm this is odd +I am testing with patch for and now +Exciting results +I will check this in and will be happy to incorporate any changes or updates you seekCheersAshwin +There is still room for optimization Im sure +wed been using the wrong sql scripts again +Reports new warnings +Created +I see +Ok I actually already did +Yup +updated +I guess this can be closed? +Ie and and both BE and LE for and +For ehcache you can do it globally by setting to +I agree that if this delays us getting a build it should be postponed but considered for the second Dev Support view this upgrade as in the best interests of our customers and a low risk to our timeline +Combined patch seems like good way to goThanks lads +The incredible tool you mention will help show that Java efforts on the rsync front have fallen short +You mention Ruby in the description but is what youre looking for really serialization cross language implementation +Im going to check this in without waiting for a review since the changes are trivial and the issue is breaking the build +I create messages and place them into the JMS queue +Compression must not damage any script +I have opened a new issue for the performance problem with the panel menu problem is also with richtree inside a richmodalPanel +but thanks for finding the bug! +Thanks Arpit for the new patch! Will try to review tonight +Sorry Ive been out with the flu for the past couple of daysBut I suspect you are right +Hibernate etc are provided because they provide implementations of EE specifications that must be integrated into the app server +Have you run it locally? Thomas and I run the tests on several different machines the result seems good +Closing this issue about samples builds depends in closing on the launcher this issue about samples builds depends in closing on the launcher question +does not pass retry number +Viewing a book produce the correct output +bumping again as there will likely be more and more items introduced that take advantage of java and or java +By What do we do for Windows I was referring to the native parts of this not the radix tree +Thanks a lot Prakash +Would be possible to rename users to be user? pctony +Bad data in import script which doesnt correspond to validation rule on Entity class +Ive added the JVM directives to MAVENOPTS in Jenkins +I am OK with applying this patch +So the improvement is total CPU for the job whereas I measured CPU improvement on the map phase only and saw +Agree with apurtell +Will wait on hadoopqa before committing +im using as my endpoint snitch they show up on the side as being down or not at all +Could someone else also verify that it works with the latestThanks +using the JSAll auto tests pass under iOS and All manual tests pass under iOS and . +Once I break it this into smaller patches we will review that piece separately +Issue closed with release. +Initial patch resolving references to and to local files within the +It is great to hear it from youPlease be aware that we are working on the version of OGNLAnyway it should not be difficult to apply your patch against itIt looks like your test is an test hence we cannot run without your app +As I understand it this bug exists in the release +The output of tests goes to files in target by default or set false +Committed derbyb as svn resolving this issueNote that the related issue must be fixed before all the bugs seen by the reporters repro are fixed + +Fixed. +Thanks Kim! Closing this. +Logged In YES useridplease try I have changed how noTx methods arelocked +Nice work +If you just change perform to execute that will resolve your issues unless you are calling perform from other thanSorry for implying the unrepeated name calling +Add +I just committed this. +The new implementation of is even more silly than the previous one and store a vint at the beginning of each block for recording the length of a block +but hey +Patch implementing my proposed solution to +I think we should fix them too in the same jiraAlso most of these templates assume that noNullsfalse and isRepeatingtrue means all values are null +Hi PatrickI agree and cant neither see why to calculate obsolete pricesSo I commited your st patch in trunk at r +Looks good +bridge the hole in +Those should be separated in a is a system properly help with Spring trying to classload classes that depend on others which are not available? has been historically able to start fine in mode by removing all web jars now its not able to do so anymore this is a regression + +Switching to tomcats own dependency doesnt needless to say jettyrun is unperturbed by this +How do you propose we deal with the external dependencies +Im curious if there is substantial cost to dealing with escaping +A static field might be a solution but is inadequate for EE environments in which multiple web applications are deployed +Makes test for EWS more stringent better to potentially name an EWS TC as Apache than the reverse +Im about to post another comment with a suggestion for the English docs and I suggest you hold off doing anymore until a decision gets made +For the record I published releases of pluginGen and pluginAnnotations to the Maven repo which contain the change mentioned in my previous comment +However libraries that need to dynamicallyload classes that have been provided as a part of the application need to use thecontext class loader to load the classes +This step happens before building the distributions +Note that one new file is added +I think Interceptor should be an interface that should reside in spi +Not an objection but a simple commentWhy not making a new and then and some fetch schedule plugins +Hi AndreasHere Ive the patch by adding the suggested java doc comments +This looks good to me ve committed this +Was checked once more. +This patch fixes this issueNice RegardsMartin Haimberger +This isnt really an error so I changed the code to not generate error codes for SIGPIPE signals +But since usually the skewed side is larger so this might be acceptableI find another issue however some rules such as depends on the right reducer and those rules are triggered before job launch +applied the patch sent via dev mailIm resolving this issueThanks! +to this patch + decoding of paths +When a new location is available then the admin can update the config and issue a refresh command to the NN +Is this a blocker for? I can backport if so +They should only be moved +Fix added to BPE Deployer and use the latest bpe jar +I have written the test case to reproduce the major compaction if all the entries in the store file is deleted we dont write any new store file and hence this problem happens +Not just being lazy but obviously it seems to be the simplest solution +To attach bug fix patch +But that starts to sound like a larger effort then moving a few classes around and splitting a launcher into two +Thanks for the patch RajeshThanks for the review Stack Jimmy and Ted +Naming the title relevant to the +Will commit when tests pass +Thanks Otis +patch looks good +yes items and are done already +Just to be sure I commented out all servlet filters and made the mapping above the only config in my +Until that is done I think this is much trickier to get right than it looks +Id like to recommend specifying your operations description with WSDD because Axis uses the information to make a response message as you desire +it wasnt so hard using hbm overriding be a solution +I did grep against the code base and didnt find any place that we expose the authentication password to logger +Will commit if the tests the tests are the change and running tests on my local +Thanks +I can reproduce this without using JSR annotations WAR file is attached note that I tried to put the WSDL file in both the wsdl and the wsdl directories in order to sneak past the warning no luck unfortunately +archive etc +Perhaps I was just being dense +If we moved the actual dispatch logic into the class in question it would work but theres still a larger java integration overhaul needed +So I rebuild my oracle jar and put it in geoserver libs +Do let me know ifhow I can help out with writing a bit of docs on the subject +Committed to and trunk +The initial issue has already been fixed in you have another problem create another JIRA with all the needed informations to prove that there is a real issue +Current patch is +This is an internal dojo issue we should consider raising IE to +If you want totrack the progress of a file transfer you can use store +can you please test with latest cvsthanks for the update +This design depends on the fact that a the client can detect the datanode that encountered an error in the pipeline +Hi Aaron thanks for your reviewAbout the description If no redundant resources are available we +An alternate implmentation which used a file with multiple tableindexes model would present different challengesI dont believe java gives a good way to track space used in the filesystem by directory so I think the only option here is for the storage engine to maintain the data itself +Out of Date +We have tables of which are used currently each with and only CF is ever written to for a given row key +Cheers +New Component struct Class with controller methods Principal Parameters Encoder to recover the objects to show all options Create a new component that will extend and use Scriptaculousajax to create a dinamic content +I like your idea of integration package feel inspired to create a patch for it? We have run with oracle recently and had to add the file so you can can control the name of sequence +Please note that in order for this functionality to work you need to install Central index feature +Thanks Chris! Updated patch attached +The way the list is used to by prepending the package name to the function name to see if the function exist +Ive attached a very simple engine which painlessly locates the revision number from the output of svn xml info the output was sanitized for my company name but is contained in attached think this demonstrates how easy it would be and less potentially bad to the checked out tree using the info subcommand is over the other mechanisms +bulk defer to +bq +Thats a really good idea +Allow JSF events like actionEvent valuechangeEvent etc to propogate to pageflow +This patch gives package protection preventing users from directly extending it +Thanks Roman! And thanks for attaching the example on that JIRA it really helps +Is this a different work or newer workBTWChecked the time stampsfor phonegap blackberry js file it is Aprilfor the one in your link it is April +Thanks for the patch Jrgen +Along with unit tests +I feel it would be a better solution to add this into the top level Apache httpd configuration though +its tomcat mean Ive commited to the tag just update. +Required version for Protobuf library is proto and above +Duplicate of +Maybe this can be better documented inside? Out of the box changing HADOOPHEAPSIZE adds a Xmx argument to the beginning of the binhadoop java command but the M heap size in HADOOPCLIENTOPTS overrides that option with a second Xmx +How about a parameter to set the globaldefault setting for the priority and then having it in the annotation for customization as desired +Its not related to +Should they just go here or is there a way to start another thread of comments +I think we should have a regression test that verifies the fix submitted with the fix either by enabling with JME or adding the case to another test that already runs with JME +The only question I have is with the service registration triggered in +I am thinking Docs should nevertheless reflect the new FQCN? That way old configs will continue to work but newer configs will use the new FQCN +Work Around After creating a domain object the Eclipse Launcher started in both Debug and Run mode +No I dont think so +As usual the rub is with Jute and enums +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Thanks +Ok I will try to send one more email to dev and move the discussion there +Jira freeze was on this is not possible for unforseen reason this must be escalated to the release manageer in a timely we dont stick to these simple rules our release planing becomes meaningless +Changes are not committed to branch +I would prefer ro do it in one pass +There is a bug with the getDeleted logic +Unresolved as of have not heard of anything like this recently can it still be reproduced? Note the window size changes should be communicated just fine to the rendering system +Youll see clearly in +Ive applied the patch you gave +Committed to trunk as r +Also though no one is in that +Thank you +Reason Missing Constraint version +Where it must be +typo Bilgin Committed in trunk rev +Related to this closed bug +verified at revision +Thanks Leonardo. +Patch +The links look valid to me and my RSS reader but they do hardcode values which shouldnt be hardcoded +I have tried the test times on the Windows box +I assume you have done the tests with libxml parserIf you use guththila parser you will get less leaksAnyway we need to fix the leaks for tried the same as mentioned above and below are the memory usage statsRequest count Cycle axiscserver started Kbrequest sending completed KbCycleaxiscserver started Kbrequest sending completed KbCycleaxiscserver started Kbfew seconds after requests sending started Kbrequest sending completed KbAs mentioned in the issue I did not experience the memory going up till about to test results by Evanthinka these problems have been solved. +Verified on revison + +The first patch contains a while the second patch does not +Thanks +zip file which is created when the test run is finishedIt should have been in exporthometmpostestingderbyall if it did exist at that time +Keyword final patch for lexer parser c generator java generator +Quickstart to reproduce attached +Conversations should only be active in JSF requests IMOWe could optionally provide an easy way to activate a conversation in other requests but really that should be supported by the underlying transport not some special Arquillian stuff +See if we can fix this for GA +I wouldnt be against just deactivating this parallelization for getpagedslice to start with +patch that addresses the unit test failure and javadoc warnings +I also found a subsequence recursion bug which was quite interesting +So far I have the following project structurejbossnetwork dao issuetracker utilEach of these projects correspond to that our current build produces +I think we need to focus on a good parser appropriate for server use and separate canonicalisation concerns from parsing ones +Except for the removal of the old test +This JIRA should be considered separately +Changes appear in Latest Alpha Manuals. +Thanks Brahma for reporting the bug and writing a test case and thanks Uma for the reviews and discussion above! +YY is indeed operationally much easier to accumulate and produce with MR in the same projection step +Update version attached +and now share the code located in a new utility problem came from Xerces and the implementation of the toURL method in the the JDK choses to handle path as fileabsolutepath instead of fileabsolutepaththis approach is not recommended by the IETF but it is not prevented. +I prefer the solution given in which would make this issue not required wouldnt it +added to +I would like to submit this patch over the weekend +I will fix this ASAPThanks issue is due to the fact that the Addressing information being copied while cloning the options and the client options gets a higher precedence on messages going out in the client side of the axis and this particular case in synapse acts as a axis client +Documentation has been updated. +If we do not support the current deployments that will be a problem for people when upgrading +lets try this to be added +As Paul told there should be at least two chars in shipmentmethod value +There is a semicolon on your jsp I will assume it is a typo +Code looks good tests pass +is one case of this usage. +Descoping items not being worked on for M into Unknown Fix Version for not issue out as part of JIRA cleanup +loosely related to +All the actions in those services should have been marked as in the release +Committed +Patch applied with svn revision +Because you would have to know all the record ids of the row before deleting it +Im not sure what you mean were using SVN head which makes perfect sense as were testing for regressions in behaviour but sure bring it up on the list if you patch to be applied on applied build server is running it + + +By the way this was on Server DB +I had a bug in my setup where the webapp was not getting rebuilt +This JIRA is marked as being fixed in CR the problem is still present in the CR build +I just tried the same scenario and everything worked correctly +That would allow things to keep working if parameters were changedLooked at the PIG code and thats exactly what it is doing so the tests passing is expectedIm not sure if Hive unit tests will work +Marking bugs with invalid fixed for release as fixed in. +Fixed by changes to +looks fine to me +Added a test each for create and appendTo +It should be investigated what exactly is causing the exception and if that is related to this issue +Thanks Marek +Verified +Can you please try this and let me know if you canreproduce the problemThanks activity in over a year so unassigning this issue +Running the Load test assumes that the database is clean of OE schema +Are these settings that should then be added to the ADMINISTRATION tab for the management console +This solves threading issue without decreasing performance +If its just it should be fixed +Thanks Pavel +The issue is localized to so lets write a unit test for that +Hmm +Hi Michel not yet but I will +Im not a wizard so I didnt know how to fix it +I have a temporary work around for this issue where we can make sure that a prefix is declared for the namespace +bq +If it does then its definitely the commitlog +Could you or a colleague please try this example project on some other machinesplatforms to investigate environmental issues +The issue is that it should not have worked to begin with inThough now I have a new problem with juddi +Committed to trunk +I rebuilt the distributable correctly and released it as +So with a it seems to be working now +Closing. +Besides Reuters is a pretty small set of data I dont even know how much it would drive a clusterIm running LDA right now and if it passes I think I should commit +Prashant has already fixed this issue. +Committed to fixed for release notes +Is there anything more to do on this ticket? Can we mark is as resolved and file a separate bug for bettydramits trafficshell issue +fajax and I think all regulare bahaviors have event attribute +Hithe idea implemented more or less correctlyI do not see any obvious lackHave you measure the performance effectDo I understand correctly that the optimization works at the first run of peephole that is before the regallocspillgen pass? +Required to enhance to include xsdtype details to response message when returning a Object or Object Array +was it ment to cover all four? Oups that was a typeo +other possible solution Allow only addresses in aliases and rely on people to keep their forwarder +Too stupid! copypaste error from the earlier days +Unfortunately we dont have workforce to support Eclipse. +Tom failed with Error +the new method in that patch +Attached updated patch that fixes readLink on OSX +Patch modifying to use new mapreduce api +Thanks Myrna +In a list that will lead to one or more null instances in the list +This is a suggestion for a patch. +I confirmed itDo you have any plan when will be released +I will double check the problem before closing the ticket +Could be that the problemRegards. +Im attaching a possible fix +In that context we should also rename to be +see Server base URL +And it is invoked whenever there is a change to a file in your project +This will facilate multipleprotocol adapters and also allow us to more easily add new serialization such as PB and AvroI have this partially done and will post a patch shortly +There is no a safe way to prevent double validation from different build threads +does this look ok? if so Ill commit itIf the ivars are not set on the the value from are used insteadI have not done any changes to the client +Would that work +You cannot take the d as an opener the quantifier works on all chars +The proposed changes look reasonable to meFor the drop attribute I think it would be OK to throw an exception rather than a warning if its set to an invalid value +Reason is I didnt know that such a tool exists That certainly looks far better approach. +Wow this is AWESOME added languages + now uses the install customisation mechanism to add the appropriate service deployer value into the file +Can they share code +someone already fix it? this is an old bug +Who is we? This is the first Im hearing about this +The semantics of flash scope are different than and are like that of other flash scope usages like in Ruby on Rails for example where the scope is cleared after view rendering following a flow execution redirect +I have removed the code which prints the password when log level is debugInstead now it prints XXXX irrespective of the log levelChecked in r on trunk. +a patch to fix the suggested problem +Fix has been committed. +Do we really want Hadoop RPC to evolve to the point where I can have a Hadoop RPC client that only understands Avro and you can have a server that only understands Protobuf and we cant talk to each other despite both implementing Hadoop RPC +Looks like problem with text with two words +Carl thanksNamit thanks +No response assuming the fix is ok. +Is this still the case? +We introduce a string image pool +You wrote that I could copy the libs from the nightly build into the plugins +Merged to branch as well since its a trivial fix +Id rather not tackle backporting immediately its going to be under a lot of flux +Ive changed the title to reflect the fact that we will upgrade to the next Spring version for +Thanks +Trying with mvn I still have the same problem +patch to for file for test case +That is fixed in trunk but this problem of what you have committed locally has not changed will still occur since that is how these things work +This is fixed in trunk +If you use that POM and remove all other dependencies you should be able to make the plugin work +why does hive test against derby and not hdfs thats a lot faster at the expense of some concurrency rules good for testing though not always what you want in production +Committed at subversion revision +I think its a nice idea in concept but wondering how much in practice there would be a benefit? Eli wont the cache only work with Immutable objects? etc are all mutable +a I dont think one giant feature is good here a set of smaller parts would be interesting +I merge this patch into +We might be close to a release so Im gonna hold off fixing the Strutscode as it would introduce a dependency on an unreleased validator jar +The same problem applies when using LONGFILEPOSIX +still a major issue though. +Trying again +I will keep watching to see ifwhen it does happen and post all of the information +Thanks Bengt +it was more a draft that Ive abandoned since Enis started working on the table lock and it was the proper way to go but still we dont have it +adjusted example with Yoniks example since mincount is a red herring here +adds a new clock for lack of activity +It seems strange to give not special treatment +This is code complete and ready for QA testing. +Can you post a snippet of the route that you are using to test this +MD sum for the tested fileabdfbccbafaacb +PeterSo can we close this bugthanks this bug still exists +Ive made the changes in trunk you are welcome to try it out and report whether it fixes things on Windows +Presently the zk apis usage are scattered in Bookie Cookie +This has been resolved with revision +v takes care of documentation nits +To answer both of you it is expected that the site will be branched with each release and the whole site should be included in the source distributionUsers are more important than developers +It seemed sensible to me because is just an interface and thus makes no assumptions on the implementation +no the patch is fine +For optimistic caches the optimisation cannot be performed as we only do the actual modification during prepare +Sadly I cant debug on this environment +Integrated +Works fine for me in tomcat steps work fine for me in Git head not sure what I can do to take this further +Patch applied with many thanks +These annotations cannot be placed on method parameters only methods and can move these to a annotation +If you want to do it in time instead of rendering time you have to make sure all annotations of ipojo are sticked to the original methods and fields which clearly brokes s documented behaviour +This also patches I believe this captures the main cases where vectors are created in the seqsparse output +This should not be the duplication of because on the left side the indexes is correctAm I right +OlegThe unit test seems to look a bit raw +All vectorization work is now happening directly on trunk +That would be a HUGE win +I think that the introduction of annotations on import statements solves this kind of a use case too so closing it makes sense to me. +Stated facts can now use identity and equality using your configuration settings +Thanks for the quick feedback Dan +Linking to issue Brain discovered yesterday +Working on this issue I created a Component and on startup I added as an to +bq +All the related sub issues has been closed. +Please confirm the current status of this issue and if its incorrect change it as required +No idea if it works +Hadoop +So it appears that this issue will only occur when the namespace prefix is +Looks good a few comments I thought partitions is not lru but keep a few partitions always in memory? Split the multiple disks and static graph part in two paragraphs +I found the issue +Hi Alexthanks for your support very nice to have an apache mirror in UgandaYour mirror will be added to the official list in a hour or soRegardsHenk noticed that the mirror is now about houirs old +pushed to in Developer StudioVersion Build id Build date +Doesnt look like a good approach +Feel free to remove the previous My original setup had Consumer threads +Verified by Sian. +Fixed that file to use sha directly +If not perhaps a replacement is enough +This adds two more test cases for signatures which mix types +KlausIn the code where you have Seq it should be Pos +Good catchI guess we should use a compound cache key core +xml results +Fix applied to TEXT module at repo revision rNot marking as resolved yet as regression test cases to followPlease verify this fixes the problem for you +If this can be fixed by changing the configuration only then no new release is ll take a is this resolved +This would work until we find a better solution +Thanks +Simple patch to disable the test along with comments explaining the issue +K I think I see how Why do we only call discardDropped if it has columns? is valid +Can you please post one that better illustrates the problem? Thanks +I suggest to revert r +If I restart the ESB itself then the service is running again if I hit httplocalhost the patch into +Even on a clean cluster with empty data directories we see stale index +Please be more specific +But this could cause a L update to miss if a conditional write occurs and fails since it would no longer update the L but held the lock acceptableAlso this would be a similar regression to how locking would work in that it would acquire the local lock first before the remote +thanks estebanfixed in svn r. +Let the transition be done by the master from SPLITTING to SPLITTING +Otherwise just define what happens f called with a null expressions we wrote a custom Voter which evaluates JEXL expressions and then uses the modified above to check them from +My fixes for this issue +Scheduled for release +Agreed +If we can get this in CR or GA that would be great +Can you let me know where the join operator for skewed keys is converted to a map join operator? Thanks +Ok I understandStill the work provided in this JIRA is alreay a nice UI improvement thanks +We need this event to track every change in a persistent object including removal of elements in a collection for a custom security policy +Attachment has been added with description FO file for reproduce the result of using execution is the same as previously upload PDF fileI suspect there must be some configuration different or my trunk version is different from yoursBTW i tried version and it return +Thanks Namit! +Great Im thinking about it a bit before I commit +Applied in r +The bug reporter didnt provide an example of a PDF document andor didnt provide more proof. +Thanks +Hi Michal +I should have said that a plain WAS install dont have Xalan in lib folder instead of saying it was empty +Thanks Rodrigo! +This is a side effect of fixing +Seems the version of the patch I put up is missing that todo sorry about that +receives the event updates its internal fields final and final and sends a new type event to the Scheduler allowing it to evict the app +The final release of Tools also has this problem +The cause of this issue is a scoping problem in the code generators element handling +Its very easy to fix +The javadoc for this new method probably should say One or more instead of An array of for the args parameterI am curious about whether this can be considered for inclusion in Solr +I would like to enhance this feature by allowing this flag to control the rendering of the widget title and the widget border +ThanksHow we planned release of this versions? +Its gonna be quite ugly though. +r the amqp field table does not support sufficent types to allow all JMS primatives to be preserved during transportation +Moving to +Closing resolved issues for already released to fix a ve always seen that to disable chunking encoding is always related to client configuration +Currently we send which means that can not delete more than a blocks every seconds +The method populate needs to be fixedI can provide you an Axis based that export a Web Service that accepts and returns a SOAP encoded long if you need it. +Too bad it seems nobody noticed that choosing as the separator in datainputs for KVP requests has been a rather poor choice +wait for unfreezing EAP +Well phooey on Java +If components implemented purely in Java can ever embed other components I would reevaluate things at that point. +Ive tried CXF snapshot it works fine +With a histogram of the results it should be easy to see whether the case matters at all in real scenarios +Specifically right after A is done flushing only the first C instances have valid object id +Testing Just setting the transaction mode to NONXA makes the bug go away +Works and looks great +Thanks yong yang for pointing those leaks. +Can you attach the troublesome ttf file +rUbuntu test case smoke kernel +Redo QA because of changes to thirdparty resolved issues. +thanks Ahmet and release bulk close +As suggested by the reporter this should be closed because no further advance can be done here +New code seems to have broken the fix +Thus it creates a potential future inconsistency +Richard I suggest you utilise users where you can get better interactive help +Changed the summary to reflect the current need. +reopened for fixVersion +An working example would be good as a guidelines. +Again it looked a bit tricky because of the multithreading and I would prefer to get something working in now and add that as part of a follow up JIRAI talked with Kihwal Lee about the extra HDFS load for an empty file vs a directory and he said about the only extra load is the extra PRC call to close it and because it is just two files per job I left it as is +Im not sure what the previous setup was since Ive inherited the server but someone may have messed up the cron job at some point and caused it to sync way too often like scheduling it as instead of +we can look for more speedups later but when testing every minute helps +added more tests as suggested +Thanks for taking the time to ensure thatwe have a common routine to handle the quote processing. +While I encountered this problem several times it disappeared after restarting the application serverSo I conclude that it is probably no issue of the maven plugin but some WTP issue +see +Closing as duplicate. +The process searches for an executable corresponding to the name that is supplied to the exec call +The patch should apply cleanly to trunk as well +Notice the package name is there twice in the path +merge is not trivial merge as some changes for went only into trunkWe may have to backport it separatelyIvan do you want to take a stab at itCouple of went into trunk they are not in are etc +property to customize timezone for all ID generations +I look at. +If you think it makes sense do you want to add support for limiting within segments +Dave are you planning to add the feature to the upcoming release? If yes lets do it asap bc we should update our docs for the feature +Fixed in me +Will do so and put some documentation of the command line parameters on the wiki while I go along +Should this be changed in as well +Im pretty happy with the patch +Please check that this fully resolves your issue +I have marked this issues Release Notes item Not Required as it doesnt affect a public build +as well this is +There is no way to say that a list member is requiredI think this is a red herring +We need to add test coverage for property nesting and abstract types before this can be considered truly resolved +Nits in always expects should read always checks since the response is optional +The implementation is quite simple and in the same way of usage as the excludes configurationCould you integrate it in a future version of the plugin pleaseThanks. +Issue may already be resolved may be invalid or may never be fixed +Looks like this requires special support in CGLIB + you are ONLY using property projections to limit the select clause then outter joins do not need to be turned off +From the description of the issue it looks like JRE is used and jar utility may not be bundled in JRE +I didnt recognize it at first +But the users have to deal with it on a daily basis +No not done +I ran tests with this patch and it did not run into the race +HiI mean that the custom config file take no effectionor in other wordjames server dont read param from that coursethe config file name is lowercase +This case is not handled correctly and well have to distinguish between streaming operations indeed +First cut at release question +When we have to move from KV to Cell based impl these are some areas that we may have to look for so +Checked this into trunk and +Tested this is a dev build and it worked great +This is closed as wont fix if this should be reconsidered please reopen with a comment +Was it fixed with the recent dependency +Stupid question but wont this drop the entire table? Do subsequent add operations work for you +I havent used regexp but when this was being discussed on the mailing list David Graham mentioned that ORO worked different from regexp in this respect perhaps he would confirm thisI tried out the javascript and without the anchors invalid data in the field passes validation +Tobias if you want to make sure that my new patch didnt break anything that would be great +But for toOnes with relationships we can detect cycles fairly easily +Bulk move issues to and attached patch added documentation +unfortunately its too late in the release cycle to fix this if it still exists +Its not word Tika can handle Word Word and newer files for metadata and text extraction +Thanks! Gives me the direction to test on +No response assuming ok. +Ill look into this issue +There was a specific bug in the original optimization that could cause stack overflow +Okay is fixing this nowCamel will now as before do by default regular attempt a total of times to process an exchange before moving to the dead letter channelCamel does this in or older just the counter was wrong when you was reading it from the dead letter channel after it was moved there +Thanks Antony! One thing to note hasnt been brought up to date with Nutch so I imagine well have to do that before we release Nutch +The attached patch is against but should apply to current CVS just fine +As Map tasks complete we update this +I had a lengthy discussion the other day with somebody I really appreciate and we talked at length about some of the possible refactorings wrt names of properties etc +In my environment this bug is related to the listed Maven Findbugs Plugin +Looks good +is duplicate of but contains example project data +Attached fix better exception handling in +Eric W. +It just forces n as EOL for the tests that failed due to newlines in the expected XML stringsThats all I have for now +It will also close the the connectionThis patch passed regressions for what its worth +You can switch node by node via jmx. + needs to be done before the strings can really be edited effectively +But I cant tell more what is the problem without JSPXHTML page +this freeking issue. +Also added extra check to avoid NPE +It seems more programmatic and I tried to make that change consistent throughoutIf we change node to empty string in then it conflicts with the way we have been displaying it in the UI as we have been displaying it as NA for inactive nodes +A Long in the input might be a part of Thrift union and a Long in the output may use Avro +Fixed in rserialize some transient fields for compatibility with java standard serialization +almost forgot about the Symbols explanation at the bottom +Resolved in trunk and branch. +Resolving for QE verification of in CR. +Marking it patch available to let the qa apply it and test it +It looks like somehow only contents of the form are being rendered +Thanks for the quick response Paul +Update follows shortly thanks so far +Yeah thats a blocker thanks for catching it +commit away +The name in the HELO command is a bit of a red herring +Hmm +Hi Pavel Separating the refactor and the saving index patches and tested basic functionalities +Sorry its the very right issue but for Hans its a good example of what I mean +I agree that we should limit the scope +The returned code is surrounded with exception handlersIn other words the right thing to do for default values is as follows The should invoke the internal SGs in order and let them convert the string +apologies thats what happens when I try to sound assertiveit was reported a year ago and applies to the current versionthe provided patch works and solves the problem of SAX and DOM having different namespace requirements +Here is a simple test class with such obfuscation code +Should be fixed in the next nightly +These issues seem to be the same +I believe the keys will get hashed to all reducers not just those used for sampling +I did verify that it compiles and passes commit tests which seems just committed this +Turned off major compactionsbq +I committed it toThank you Mike. +This is a bug in webkit on RHEL triggered by something in the example that have not yet been does not fail only in browersim it fails on any webkit browser on RHEL is not fixing their bug only possible workaround is to adjust the example somehow +Resolved in Git ID fafdadffdfeadff. +Why +srcmainjavaorgapachehadoophbaseclient +bq +How come the diff results for testorgapachepigimplutil were listed twice in +but this is the issue. +Correcting the tag nesting of the tagStack +It looks like maven is unable to resolve the coordinates of an artifact +If input recordEntry doesnt end with or +We need to change all the places that use and Configuration to accept a passed in Configuration instance to prevent the need from using static methods +I am facing a similar Ant Build Failed issue complaining that it does not know to set and using my environment +Thanks for working on this with the incomplete information +Added note to NOTICES about +Is this done? If so can we close this +A Session is wrapped within aThe XA Container will close your Session when he closes thePlease have a look at There we have several testcases about locking in xa environment +i doubt it +reopening for backport to backport to with revision +We should get this one fixed automatically if we upgrade to Pluto and adjust our registry schema appropriately +Here we use a static shims layer so for different hadoop version we need to recompileSince maven artifacts is not available so we need to simulate it by copying all libraries to lib +Thank you Jurgen it works +I have a couple of questionscomments +Ignore my commentWhen I thought of contrib I thought it would be like the Hadoop contrib directory where get their own folder and ant file +It deploys correctly and can be used from under +but in a different point where we used to see before. +Looks great candid you commit +well the meta class that calls the setter should do that +Oops! You are absolutely right +And please do not set all at major level as something that is related to a log can only be minor trivial +Here is a simple test case generating the error +Thats an interesting question about the JAAS context and the profile service calls +so I would like to make this part of the MOP +Just use caching when you operate against a native JMS where theres no pooling of Connections or Sessions done by the accessed it is possible to enlist one Session with one transaction and after commiting enlist it with another transaction +Works ok with current in and in seamApp test +Moved to James a contributor and assigned him this probably +Which is why the new REST endpoint could provide us with an URL per application to check its availability andor return the application state when querying for applications +This is not new in EWS it has been the same way since EWS +So looks like the jars were updated in the lib dir but not in theThe ramifications of this are that anything referencing cassandra on maven central will have the older versions of jackson brought in as dependencies +Until then please use the target version field to indicate where you intend to fix thisAlso please note that in order for the automated tests to run youll need to upload a patch that applies to trunk and mark this JIRA patch available +that may sound like a answer but its true +So if your ZIP file download gets corrupted Maven will fail forcing you to solve the problem +this patch works for me +As you can see in the exception above the whole trace is infrastractureThe infrastracture itself is the one that calls that exception from within the deserialization process +The releveant places in code are marked +Final patch for review +Thanks Kristian. +Please use the link provided in my previous comment to reach the correct Jira project for Axis +I dont know much about the bridge +Id try on our Samsung device but it doesnt seem to want to mount USB storage +In basically all cases throughout user code but this is probably not optimal +Bill +Since is solved now we can probably mark this as fixed when using the Tomcat said Im still wondering how deploys a persistence unit that is part of a WAR file +Closed. +old the changes are in so we can reslove the issue +When switching to EJB that was lost and the generated ddl defaults to a normal blob which has a max length of bytes +OK +Thanks Steve Brock +ok and the bug does not sound critical to me either +this will have to puntWorkaround for now is to use the management commands outline in the related jiraWe could add a action on our server that enablesdisables this flag on servers we have management access to +I have merged this to. +No worries +In both cases a single karaf instance of the remaining slaves acquired the lock and became the masterNote that those tests did not involve starting a karaf instance but instead I created a simple java application that followed the same sequences of calls that the karaf Main class does +Please see for test results and how to apply this patchThis patch should be applied after +Especially the annotation examples mention JDK all references except the eclipse documentation that probably should be removed as well. +Assume closed as resolved and released +Thanks Todd +Can I suggest adding an obvious download button version number to the homepageI know theres a download link in the main nav but it would be useful to have download link really clear on the homepage +Hi MartinIts not easy to pass a Message instance into an Spring version of DSL so I introduced the inBut I think I can add a Java DSL for setting the Message instance into the +Since cant get test environment to validate jme client will push off to next release +I put them under jca The console spits out a lot of log errors when you view the jndi tree +Removing target release from issues that fit any of the following criteria No known way to fix them Java integration enhancements out of scope for release Other out of scope issues for works ok in current trunk which will soon be. +I ran locally which exercises the Avro dependencies and it passes so I suggest we commit this now +It would be nice if someone could take a look before too many changes get in +The login module then can read the password hash the given user password and compare the two created an from the src which is doing the job +The purpose of the distribute system is high utilityIf you want to analysis running tasks performance you can just set one node with one map and no reduce. +You are right this was a clear hypothetical fixI will try to set up a test environment later +How deep does this assumption go? In JDBC all rows are supposed to have the same number of columns but nothing actually breaks if you quietly ignore thisThe reason I ask is all current scenarios will work fine with the take superset of all rows approach but UNION support in will make it easier to get resultsets of wildly different row sizes +bq +This Jira can be used to prevent the RM from accepting configurations where there is a conflict in the list of job submission queues +Do you remember any special change that occurred on your project version at that time +The second was to facilitate the use of an XML Data Binding framework +Which is +As far as i know not for the runtime +We could attempt to call the constructor we know about with nullfalse args but theres no guarantee such a constructor wouldnt throw an exception +This should fix it +addressed thanks for catching that +Seems that the script does not really work +bq +And they can add some more native UI to their own apps that display part of the pages and can use all the JSON stuff for just data requests +has been marked as a duplicate of this bug +the problem still on Len reported it as the fix should be included in the next build +test for itemClass item is added +This forced each invocation to recreate the and as a result of the presence of the Synchronized methods contend with other threadsThis patch replaces the original fix reintroducing the caching through the use of the class and removing the contention with other threads +I opened to correctly close for and reloaded +Thanks meant I verified that the tests run on +Sounds like your solution would solve this problem for us +With this patch the is enabled to determine the type of attributes from +If the original reporter can identify steps to reproduce please reopen the issue and attach a sample project or describe how to reproduce the problem +Revision trunk Revision +Back ported to closing after release +Patch applied in r. +And most likely when the decorator runs again the problem will be gone. +Oops +Ok Aaron I will do that +patch looks good +If you give me a process or a link Ill be happy to perform this task +insteadWhats the difference +The same question as for We use templates to generate code +bulk close of resolved issues. +Rename it to for running +to add dependency to ant jars and reuse the really think this is an urgent matter that needs to be addressed weve got a few tar files close to GB and this is a very annoying limitation +Wed have to change to point at which entries are added to the index cache also +I will address and at firstthen will take a look if it is still not resolved by then +Brent is also adding a test to validate that remains an immuntable objectI will be adding this patch later todayThanksscheu +This seems to be a bugproblem in the osgi installer core which does not translate factory pids correctly backThe jcr installer part ignoring those nodes is correctIll have to investigate a little bit this was a serious of smaller bugs in all three components +Radim what kind of workload do you want to use combiners without reducersWe consider whether the feature is really needed or not +would also rename to its private so does not matter but wozuld be nice to havedonebq +Alexey patch applied at revision r and I also merge the regression test into thanks a lot for this enhancementVictor please verify that the problem is fully fixed as you expected thank you +I added you so you will now be able to assign YARN to yourself +That gives us a hard goal to reach and a deadline for when well reach it +I assumed the fix would be applied in Geronimo and WAS CE would pick it up +Please be sure to subscribe to repository to follow any discussion regarding the repository such as upgrades etc +assigned to steffen +Enough done that declaring the umbrella issue resolved for +Ill include in the patch +OK +Bug has been marked as a duplicate of this bug +If there is no feedback with good reasons to allow setting prios and indexes on aggregates the corresponding method signatures will be removed from uimaFIT +the proper solution would be to unregister applications provider factory to the contextual data at undeploy +The first one is the aaa token +Sorry about the previously failing tests +Verified that the test case failed without thepatch applied +Fix good to me. + +Sorry for wasting everyones time +Please could you check that it has been applied as expected +The most recent DNS schema I could find +Thanks Alexei +Then when it goes to tag the code the parent pom has not been committed and it errors the issue is to determine why the parent pom is not getting committed and what I need to do to make this happen +Mark RichardThanks for the patches instructionsThe platform code has been moved from NIO to LUNI at repo revision Please check that the patches were applied as you expected +attaching server and test do not support the IBMs default policy provider +harald can you take a look? it needs to go into the branch +Can you please try running with this set and see if the problem goes away +It seems to fail in the dense vector case +Ok after sleeping on it I think the fix warrants a little deeper analysis than the patch I posted +No it doesnt havent tried that beforeAfter switching back to JNDI it works like it did beforeI get a nice string value like thanks for the confirmationThis is a known issue with the Apache Directory LDAP API not handling binary attributes correctly +Thank youHow does somebody get a Pootle account for me to add them to the projectIs it possible to add someone who is not an ASF committer +There are many ways users can shoot their own feet +applied patch +I mean at the very least we need some more specs that actually fail with to make sure that the patch is actually fixes them +Yes configuration for more memory will resolve the issue +Alrightie then old repos renamed will get the new ones set up today +Can you update review board with latest +Thanks +sure will do +Can you add the tld file to a new issue please +So we could fix this here but it will still break in more complex situations +I still have this issue with Spring and in Maven +Anyway the mentioned patches are already commitedRegardsDimuthu +close resolved issues +Completes the support for deletes fixes a spelling mistake +Removed disable loadingbar execbefore and execafter because they are covered by the event listeners +I tested it with configuration in and in +Actually all Extensions I know expect to not have to handle multiple classloaders but only are written for exactly one In this paragraph are a few things wrong +Sorry Tom I wasnt clear in my had intended for this to be a normal so that it could take part in the AS lifecycle sorry about ll make the changes +Attached a compact testcase that explains the error +How do you test it? with a fresh database or an old one +That way folks wont inadvertantly overwrite things +yes please test thisId like to push this release in the middle of next week. +I was planing to release Installer GA for AS my understanding is that we are not having released and all efforts are towards next release of Installer will be for AS GA +So do not worry I perfectly understand why you guys think this shouldnt be implemented. +Denny This is built into the architecture +I dont see a reason why it should be one and only one billing accountAlso the problem Im having is with one billing account and a credit cardIn the meantime the highlighted code is definitely causing a bug and Id like to remove it until you have something else +I took care of that but then all test scriptss loaded by the compiler when groovy compiles test files which means that their generics infois loaded by APP which was against the scenario needed to reproduce this issue as this scenario needed dependent files ast as not being loaded yet. +embedded console doesnt ship the Sigar native library since its GPL so as the message explains the console doesnt support this operation Cannot get the process table information without native support +I think this makes sense +Added under the new name +Patch b passes all tests +We already force the app to add the same field twice if +Please take to consideration that it from new you need new configuration details please take me know +The contract has been updated and works correctlyThe patch wont be applied in the skins. +Thanks. +I understand it to be implement this but would be good to specify more concretely how and where this idea might live +These are the steps that I was talking about the issue +Fine with me though it seems like Nathan meant that fields should be treated as optional in this scheme +the DNS record for should now be fixed but any outstanding uses of it should be repaired +Perhaps it was now fixed? +Patched applied please verify +This changes the semantics to throw when there are existingBut that aside Im not sure why switching from to improves things +Added license header to appears the assumption is that the attacker wont be able to get root privileges +Need to make some tweaks to the template to remove meta and make the pageinit binding more more update to the template to use jQuery which is the supported version with jQuery Mobile with. +Changes always take mins +trueI changed it. +The underlying problem here is that JAXB does not have a very strong API to determine what types are supported or not +Want to give it a try +There was an error in the original patch +This patch resolves the bug that the sets the propertyResolved flag to true also if the resolvation was not successful +Now I think this should work ok +Its committed now thanks +Yamashitas patch looks good to me +Let uns know +JDK and WILL BE INCLUDED IN DOC PATCH FOR Xalan Java release +Thanks +or should try to split the table without it +Please reopen if you still encounter problems +Here is the patch +Im still new to GrailsGroovy so I havent completely figured out where the lines get drawn +Pedro where are these tests coming from? Do you have a particular use case for thisTheres currently no support for versioned cache entries +tested and ready for +Ive been trying to come up with a unit test that will reproduce these problems but unfortunately I havent been able to do so +remember that it is common to have property replacements in these files so the value might not be true or false but +The patch was meant to apply to the branch +Can you post your configs as well +Patrick not everything is completed resolved the EJB issue that reopened the JIRA +The joor Library can handle that many calls so Id say that we should also be able to do that +Please close this issue if it works for you +No clue I suspect that it was checked in when permissions were added to theI emailed the list but no dice +Does this still happen if you put transactionaltrue on your annotation +In GillesIm a bit confused by both your comment and your fix +I dont think we should restrict request handlers from handling streams +Thanks for the commit RickGood idea to abstract the DRDA type checking +eos is back your site is up mirmon sees it fine. +Brett if there is a known issue with why is this issue closed? Im still experiencing it in +Marc I am aware of what needs to be fixed +Re AS logging thats a good question why its needed in the POM +And that is not true for Enums +bq +compactExpr in Pruner then compacts the nullsThe root problem is that doesnt distinguish partition and virtual columns so partition columns also get thereI will try to mitigate that in separate JIRA however ctor wis arg is called in many places so if they call cannot be easily replaced can be done before expression gets to the pruner +Theres also another somewhat related issue that I opened for +h file in the patch since you have modified the prototype of the functionUpdated headerThanks Chen +Right click on any tag error is displayed to console has been applied to branch to +Thanks +Anyway it must not be a release blocker +I guess you need to wait for the CVS repositories to be looks good +Patch that bumps hadoop version to is attached +doesnt transform and output the first and lastname of James +If you see these two methods are implemented in file for utilizing the other if it is task level we can implement in itself +This seems to be fixed in +Ive been thinking about it and Ill upload a patch soon +to fix with easy patch for for reporting this! +use u with appropriate user in curl commandexpu admin +it seems this problem happens if the test setup hits a or similarIn testError I get failure set toCould not create a new instance of class see cause +As people on the net are reporting this issue also occurs with the cdk plugin +With fixed Im marking this as resolved as well. +It might be useful to discuss what you currently do +My guess is its not impossible but might take some time to implement +This patch does not in any way interfere with normal operation of all tests pass +Abe You patch for make install is now in +bulk defer of featurestaskswishes from to +The last two commits for this were accidentally checked in under +log file in the +did I accidentally close this oopsThanks for checking on the dates of the change that will help a lot +sample sample application +To make thrift publish their artifacts to repositories +The issues title should be Edit user page uses XML store regardless of service that provided for the informationAfter changing to another role service its roles are appearing in the lists in the edit user form +That did the trick +Closing old resolved issues +cpp files but that doesnt get much of an improvement in build times +The Hudson are not externally accessible +Configuration issue broker was reaching its memory limits +This approach doesnt require any modification to Hadoop to achieve the same data reliabilityredundancy +finished for +Attached patch fixes the issue of capturing the first line of the backend error message when the backend message cannot be parsed by PIG +Confirmed fixed +You wont believe it I am currently working on thisThe fix is not as easy as I thought in the beginning thoughregards in nightly build image directory is found again both in IE and firefox +I will make sure ends up in the Enterprise Bundle Repository and also make sure that that version works in DM server +sdirectory permissionsfile pathsHavent had my coffee yet +In the current approach mergeIndexes is an admin command and the target core should be online +Fixed by changes made for have the same bug with was fixed for +Alan thanks for the heads up +updated patch as per Tucus native code per review by fix in two places dont log the same exception construct new exceptions with better context and use the previous one as the causethanks Tucu for pointing this out +Until the latest patch download csv button works well +Thanks Martti for the test case +Ive never seen a browser append the hash fragment to a or redirect +oops this has been closed for a long timeP +As such votes please +Hi PaulexThe fix looks good thanksBest regards by Andrew. +Removed fix version. +Thats the bug for just download the recent version of it to libpls see the discuss in Problems with tks pls correct the bug in futher distribution +Overall I think the changes now only affect only the new functionality so even if I find some related problems I am likely to commit to make it easier to address the remaining issues +It is possible that they already do but I dont know enough about mingw to be sure +Your code is slightly shorter so perhaps we can use thatYou can test it by running binwhirr from both a source directory and from an unpacked tarball built with mvn package assemblyassembly +We appreciate you taking the time to contribute back to the communityEither way you handle is fine wme +Yes I can also reproduce on and trunk +Joao I assume you mean for specific properties in editorThe editor currently follows what the DTD says what esb devs identifies as relevant fields to provide extra semantics assume these fields got specific semantics based on stuff not specified in the DTDXSD thus if you could list the ones you are looking to get support for then that would make our job much easier +if client endpoint is a valid cassandra node get the nodes DC and prune nodes outside of this DC if client endpoint is not a valid cassandra node try to infer the DC from its ip and prune dataendpoint nodes in a different DC +Sorry for the delay getting back to you +MB came from the rabbitmq producer I am attempting to replace so a small heap should work in theory +and TOPTERMSBOOSTBOOLEANREWRITE sounds confusing +We can exclude it for now in a hope to find a better solution later or just completely remove it +I found this problem by code inspection and have not reproduced it myself +For example all webapps inherit from the same base pom but these webapps may appear in completely different module hierarchies +Closing +I believe the original patch had an issue and a new patch was committed to the Trinidad trunkIf the issue still exists please reopen the original bug not this one. +Is there a commit in trunk that we could cherry pick and apply on Hive to fix this issue? +All except and fail with the known +You can also implement a wait behavior by allowing blocked exceptions to bubble up from the invocation +bq +Modified version of DMLC that implements dynamic shrinkinggrowing of consumers +Unfortunately I wasnt able to use the test that Andrew contributed since this feature happens in a static initialization block and depends on an env var +How do you mean active? mmap uses the same page cache as normal access does so the same page out applies I dont see how we dould be able to avoid using buffer for CLS if we to use SWEdit forgot to mention we should probably look into madvice direction if yoi want to control over CLS mappings instead of trying to use +Attached a new patch + Harmony VM was not yet adjusted to work as Java VM +These are useful beyond function queries +Patch applied in revision . +All unit tests are passing +Added junitee and ant target for war +They are basically working on open sourcing the API as well as the RI to make apache integration easier +In long term I want to move the functionality of setting permissions and groups from client to server so all these tests and code will be rewritten +sorry the name of the file should have been mapreduce and not zookeeper patch! Looks like I am not good at multi tasking + +You get what youve asked for +Ignore wrong about deb +Bulk close for release +Also added check for configuration in it a blocker for the check for configuration +Rebased merged into and cherry picked into master. +Thanks! +Changed to XML Schemas +Can ask someone to commit this once Hudson blesses it +in enhancements of release +Hence closing the jira. +ive found it much better to break out seperate modules for these dependencies +If it makes a HTTP call it will slow down the performance. +Will need to dig into this a little further a quick review of the code here suggests it may not be SSL ready Assumes the protocol is http and not https +Versions and higher of Suns let you override the locale by setting properties +Marking as enhancement till then +For hbase only way to get an hudson build is to do an actual commit +I then restarted the host and the instance stopped but hours later it hasnt restarted +Usersrjurneybinhadoop jar Usersrjurneypigbin +the created war misses several crucial libraries set in the build path thus deploying the war deploying the app from jboss ide to jboss directly includes these simple way visible how to force Project Archives into respecting the build path when creating an archive +Verified with id date reopening to update release note info. +This patch extends the behavior of to the Header interface and implementation +We recently upgraded to and were still seeing this problem +Committed svn revision +The verification of installation is incomplete since the injected Configuration object delegates to the User scoped configuration instead of the project configurationAlso given how the user scoped configuration instance is modified during the installation this is a bad idea to modify the configuration if were not sure whether it is user or project scoped +I dont think the source target compiler settings are sufficient +prefix to circumvent the special meaning of the command +See for example the patches I posted to which implement context sensitive serializations with no changes to +Thanks Sandy +Finally if there is no success then switch bookie to mode +This results in users typing data into the search fields then clicking create new which of course will not have the intended effectI think we may need to do something to break up those buttons somehow +Yuck +All is explained in the +Table name might be better but then what if you wanted differently configured connections to the same table? Would that even make sense? Thanks for looking into this +Thanks that makes it much clearer for me +Please feel free to reopen if you experience the same issue with the new build. +Which patch +guruofse this is NOT a support forum for solving users problems with how to USE hibernate +The Attachment is for issueThis patch was generated from Git +Server still seems to respond fine so Ive silenced it for now but would love to have it fixed +I tested this on the latest trunk version of and StudioEverything works as intendedThis is definitely a bug on Active Directory. +Image created using the current can confirm the described behaviour in +misguiding attachments th rd Jan has been removed +Modifying some configs and running it from by home dir +You can verify that by changing your ORM to where you will see all works just fine +Closed after releasing +I think youll see an improvement in a write heavy test +Thank you Andy +in both and HEAD now +Committed revision to trunk +danielsh helped me get svn access to the repo so I committed the changes back +Hi Manuel thank you! Another bug or a bug in my fix to the other bugIll investigate deeper in a bit +A boundary case is not working correctly selecting any node after selecting a duplicate results in incorrect rendering of the focus in latest patch +Feel free to reopen if I broke something. +Script should also check for existence of targetrepository folder to avoid loosing time unzipping +no its just broken +I am still digging +I will look at it but I have been swamped this week +This consensus on list is that this should be committed +Just to mark it as with. +I will rework the migration code +What would be a moreapproriate value? is fine with me +That as is compactions may be run prematurelyWe need to figure some means of bubbling up the result of flushCache up out of flushRegion in emergencies +this issue were closed with fix version LATER +Unfortunately my RCP skills do not suffice to perform the proposed change on my own +I will commit it shortly +Ill do that now +Closing all resolved tickets from or older +Defer all open issues to +Because this part of the code is rewritten by I dont want to make a patch for this against trunk +Harmcrest is required for the Extensions Development Support feature which you would install in order to write me configurators +PROJECT A uses ARTIFACT dep Mgt import ARTIFACT In my settings I have a profile which redefine the central repository to use the corporate proxyThere is no repository settings in my have no problem to build ARTIFACT build fails because it tries to download ARTIFACT from central repository instead of using our tests If in PROJECT A I add a repository entry for our corporate proxy it changes nothing If in ARTIFACT I add a repository entry for our corporate proxy I can build PROJECT A If in ARTIFACT I add a repository entry for our corporate proxy using a property I cannot find ARTIFACT because maven doesnt resolve this property If in my settings I add a mirror entry for the central repository to use our proxy I can build PROJECT issue affects and wasnt clear to me when I read this ticket the first time but there is a work depends on LIBB imports BillofMaterials LIBB I added the repositories section to that pom that points to our corporate proxy +I think this is something better fixed by the maven folks as i dont recall being the one to put that pom there +A patch that allows changing the name of the binary artifact not very elegant but it to change fix version + didnt compile at that time and Id to like to test changes before I commit them to make sure they dont break anything +this is actually an incompatibility with the ant tasks not the applied. +Attaching +Yes relogin wont cause exception from second request +Attach a patch that makes able to run the individual test case class +I wanted to get some initial feedbackThis time I could successfully request for a RB request +Ive just committed this +Hi Simonethanks for the notice +thats what is planned +Thoughts +FYI can you please look at these question that Len has raised +Is there anything else needed on this +Thanks for committing! +Alex Could you please have a look at this issue to offload Jochen a bit for the upcoming +Now I only lack the hour needed to create a patch +patch looks good to me +And i replaced all tabs with spaces in my Patch thats why you cant apply this patch +The doc needs review as a whole for the entire API anyway and it will be easier to review as such +Attachment patch has been added with description patch against source created files java classes the needed libraries from aXSL and has been added with description New patch has been added with description Updated has been added with description New files required for for convenience Ive put them all in one jararchive to be extracted at the root of Fop +I simply mentioned it to my view of the orchestration and this it is different from what is about +Thanks Juergen Hoeller +What is in the message payloadA pojo? The CMP entity bean? Some combination of this +I think they need to be merged into one or you better create new JIRA for thing +so if nothing else is done I think it would be worthwhile to always log why we are booting a db as readonly especially if it is not the standard place where we expect in getting the db lock +Can you attach your script +maven source project for patch +it will mislead the dispatching process +Sateesh This was fixed can you comment +Thanks +btw +Not working yet +That is correct +AS it was merging well I decided ot push it also in r r +Good stuff +If this is not a requirement for it can be moved to a next feature by background validation and validating connection factory implementation +What should RS do then +for one hasnt gotten around to look at this one +I put a comment saying the same thing in +OK thanks for looking into it. +I cannot find which docs to update for Unmarshaller debug +At any rate you have to makesure that the font youre using actually has all the necessary glyphs +So please feel free to adapt and add tests +Issue says this if for only but the patches against trunk +Thanks for the reminder to publish the bits Justin +Go ahead +bulk close for +Not sure if they have changed the JDBC adapter framework for or not but Im happy to share a howto if you want it +allows unit tests to depend on each other within a suite which your approach wont allow because of each unit test doing a rollback +Its the same as the page that works +I dont think its unreasonable to have an expectation that the application be configured out of the box to use the correct combination of these plugins +This patch is only valid if you operate with XML +Class from SVN with changes applied +No tests since this unblocks the existing tests +Heres some tests for if both are records the ordering of fields may be different fields are matched by name +Thanks Shai +svn ciAdding depsAdding depsTransmitting file dataCommitted revision + portlet does not assume the target web content is in encodedHTML documents can be in any encoding such as the default web encodingSo portlet detects the encoding of the web content based on the responses character encoding or meta tag of the content document itselfSo I think you should check if the response of the target web content has encoding header or the target web content document contains meta tagThe current implementation assumes that it is in encoding the web default encoding if the encoding is not detected by the ways mentioned aboveCould you check this in your target web contents +Patch va replaces patch v +Added example of a singleton method +Phase Make region server more event driven instead of server thread wake frequency +The javadoc issues are unrelated and are being covered by a separate JIRA +Happens in a single place execution so that rules out serialization related problems +for doing thisPatch looks good +Fixed by SVN revision +bq +Ugh this sounds like an ugly issue +Thanks for your help. +Will report back on if it fixes our bigtop ci issue +It seems that the JCR support in the Spring module used an old version of Jackrabbit which is not compatible with the OCM stuff +removed +metadata +Also adds debug log message for local read +A few small comments we should have a constant like METAMIGRATIONCURRENTVERSION which maps to the most recent we can use this when we have completed migration it seems to make more sense to have the constants just be ints and then convert the cell contents to an int where we do the comparison rather than have the constants be byte arrays and use byte array comparison +Really bad is in this case defined as Cost to maintain the driver is really high +Thanks. +Committed revision +That should do the trick +I was wrong sorry +We have embraced it with open minds and have benefited from itThe cost seems to outweigh the benefit though according to many devsbq +So the best we could do would be to allow this element and log a warning that it has no effect +Its likely youll find a file that contains that string +Now correct fix attached +Do you think this is relevant +There should be a command +This patch addresses Bens comments +Slfj api and implementation is still missing in in distribution zip available at address given above +I actually came across this because I was creating download tables +I agree that site deployment should be available probably via wagon +Does the controller method have the method parameter annotations? They need to be on the class +It is reasonable to track these twoWe have tried once with the limit +I tested the fix and it worked +Resolving as by design +But if its started then WAR is picked up before was deployed +Cocoon will stay with +For the case we read the various spill files back serially anyway so whether they are on one disk or many doesnt matter +I faintly remember that so I just wanted to make sure that the same problem is not there +I would really like to see this as well I usually create a first iteration called Project Backlog that I set the dates on to a past date +Uh oh it looks like this patch implements functionality that was part of? Do we still need then +Thanks +I want to know when this was introduced +bulk close of all resolved issues. +Committed changes as per last patch +If is implemented then at least this kind of error can be recovered from +But this shouldnt end up in a NPE in any case so the bug is valid +Do you have a specification of the method saying something different? +But we decided to move that feature to the release of Smooks +patch R also handles it +It was uploaded accidentally +Verified in SVN +for the patch +RS goes down and the region is assigned to RS +I am working on a backport patch for now +PeterI think that works as well +Sure we can definitely pull the user and pass from the Jira object +Hi DanYes the changed expectations in the unit tests is what bothers meAs I said I dont want to compromise that late in its lifetimeAnyway Im reopening the ticket so someone else can take a look and fix it if this is really broken at the momentIsnt it possible to extend this class and fix the problem in your class +Ive jut committed this +sure thing grab trunk +Merged arquillianmaster into arquilliannextRebased on top of arquilliannextRebased kpiwko on top of arquilliannextRebased kpiwko on top of arquilliannextPushed to arquillian to arquillianarquillianmaster +You might want to try splitting the method into parseThanks Justin that could be an idea if the internal computations are not more costly than the Array creation +I just called a vote +Thanks Colin. +makes the enableactivate and disabledeactivate happen per spec as enabledisable immediate and activatedeactivate asynchronous +The class is removed from the +Proposed patch moving the launchpad API to the launchpad base the patch in Rev +Like I said in a previous comment the dateFormat parameter is not used for formating but for parsing +Im checkpointing progress in case things break I can easily roll backWill this http access also allow a cluster with incrementally updated cores to replicate a core after a node failure +While I agree the slot concept is confusing in our users need some counter to track their utilization +class files +Trying running it in cygwin +Jar files have been reorganized since +I suppose I dont mind two sets of keypresses +Thanks for applying the changes +note Moved to the DIR project as its not a Directory Server specific issueWe might consider to proivide such tar to users +ulimit l always returns I gave the cassandra user a shell so I can su to that user +Sorry +This is indeed a serious problem +It adds a few missing new files and also includes changes to push the collation info from language down to store +Verified on +If you run commands as this sequencegeronimodeployconnect s remoteserver p remoteport u system w managergeronimodeployloginThen the remote connection credentials will be saved and a message prints out like this Saved login for deployergeronimojmxremoteipremoteport +Please commit and close the issueMany thanks +We are not going to provide any default css with the resource servlet originally we had done this but we changed our minds and have explicitly decided that the css needs to be provided by the note also that this control will eventually become deprecated in favour of an equivalent component and as such we dont want to expend a lot of additional resources to enhance it further at this time. +Looks like this doesnt use the authtolocal m opting to build a different filter for usage +This is still valid +Im assuming the following usage +Hurray! +Nevertheless full portability has not been achieved yet + +I will continue my investigation on Monday. +getting a bit ahead of myself I was looking at the global usage of and it looks like theres a lot of calls that come from and impls which could benefit from the reverse mapping introduced by the patchI would replace the map with a helper class that delegates the calls directly to the namespaces node state +No new unit test is necessary since this just impacts the buildBtw I filed and for mavenizing the build including it in the tarball and getting the test running +Feature completed. +In a history I see that feature was scheduled to implement for version but now its not scheduled at all +Have been monitoring this test for a while now and havent seen any failures +This obviously needs verification but Ive designed the patch to leave current functionality pretty much untouched +I bet this goes away in the next release +This was fixed in +Regression test case +The test is passing for me +I verified that the entries get removed from the Hashtable as part of the markClosed processing +Patch applied in r and a new SNAPSHOT has been you! +Also is there a reason why you chose hours? At present we are running CI checksbuilds every mins on the Newcastle systems +This issue is over years old with no progress +That seems like two functions into one but I see your point that these are both characteristics of the element of the vectorAnd I do think that labels on matrix rows should be inherited from the column labels of the matrix +With this patch this enhancement can be closed +Daryn the current patch looks good + +NOTE There is a silent thrown from Lucenes when the dictionaries are fed with readers with no documents +Perhaps but I cant say for certain without looking into it in more detailI did try at one time to fix this issue by trying to make the activator startstop its threads based on the framework state but when I looked into it I would have required a bit of work to get it to work like that +This has been identified as a blocker by Rajesh for the FP release +back port fix the fix +Committed to trunk. +I dont think there is an easy way to compile Hive in Eclipse for both hadoop and +assign to Daisy review +Hi David and AndreaGuess you have been pretty involved with FOSSGI have turned off the GZIP compression filter on our prodinstance now and its working OK with Apache doing the gzippingof all content types except images +Can you should me what I should put in my +Rob can you please have a look at this? just committed an update to so that you can access the local +When using IE then I CAN SEE the toolbar and the icons +Attachment as the proposed patchAttachment contains code snippets before and after the proposed patch + +Is it possible to address that problem? I understand it is not straightforward because of input streamsI believe it now handles all cases of the sync record being split in the buffer +Ok lets keep the size and just make the patch against types branch thanks +tests failing on my system cancelling patchSee the attached file for test outputnote to committer make sure to chmod x on the shell script +They should not clash because of namespace differences but the C backend compiler is inserting using namespace xutil into the headers of template classes +Hmm then again its a chicken and egg thing +Thanks Todd and Owen! +This new service is implemented in the package +They are desraptive for reasons the rearrangements that are done can cause build failures and need for manual cleanup of directories as happened last time they are also hard to test because we dont have a way to automate the testing +Duplicate of +Thanks Doug for reviewing itbq +Dont know about basic auth +Im not sure but I think that there might be a bug somewhere in grails code that leaves scanned groovy files open +The issue appears to be fixed in FOP Trunk and the fix will be made available in the next release +The following wiki pages have been integrated These still need to be integrated in some way These do not need to be integrated These are outdatedin a bad condition +I only had to fix some indention and replace tab characters introduced. +We can achieve the same affect as we have now in a more safe fashion by using a blocking queue +Why is it OK to remove the checksumming from ELFOSwrite +If you can suggest the correct jar to fix this issue it wiould be great helpThank you. +Correct on both counts right now revision numbers are just based on timestamps +In my case the problem is user induced +Ive just committed this to trunk andThanks a lot for the contribution Daisuke. +In the conf directory there is a file called +Just bring up any existing task and click on the Content tab +Wow in code no less +I have checked in the files the same way as before +We are ditching Apache client in Resteasy in favor of client so we can revisit this +I just committed this back to and moved the log in the +As soon as theyre posted Ill look into what the problem isThanks GaryGot it! I used to read in the xml files and for the and now the output is correct regardless of whether is specifiedMany ThanksTerenceBTW is on our system. +From looking at the source code it seems that create is invoked repeatedly on the not ordered flow elements from the parsed bpmn +JV can you add notes on about whats changed and why +The surprise is to the CJK users that get no results due to undocumented hacks that people refuse to let go of +Also can we store the recently chosen versions of the fields in the so that if you create containers in the cloud after one another it remembers the to use and remembers the group user hardware location image OS family OS version etc if specified and defaults them all +only those bits of the classpath in which there is a +Thanks RavinderI just wanted to check in on this issue +A patch is attached to the linked issue +OK I will commit and gasp merge to +Do you have any additional Maven properties in your file which could impact Sonar behavior? was having this problem then upgraded from Maven to and it went away +forgot to mention that the line number is without applying the patch +Fixed regressions with and at r +The test case you provide is not standalone +It doesnt seem to be to return a that isnt inline with the RFC standard and I cant help but feel that the number of users who will be negatively impacted by this change will be small +Section default table name should be the entitynamepropertyname +IsabelI find that keeping large patches up to date with only an SVN branch is infeasible +The difference between test and run is present in only this issue is now fixed with +They are separate processes launched by something other than the test framework +screenshot from eclipse also attached +Not only overload they just shouldnt be there +The new UI doesnt require preferences +Find attached the stack trace of portlet application +I can get it to hit very reliably on a dev machine here +Its a good idea +Thanks Roman. +Ive committed this. +There are newer versions of IE nowAndreas improvement will be included in +Here is the patch for thisPlease give me your comments to make it betterRegardsAbhijitAll interested parties in this component please review the patch +That would definitely solve the problem +This has already been fixed in. +The exampletest code could be used to verify the resolution of the problem +So Im going to cancel the request for tracing the cycle in favor of the other feature request of dynamically detecting deadlock of static initializers +Some experiments with quoting the value worked but not fixed the issue by making the operation do a POST of a hidden form avoiding all url check the code in after final verification by a customer. +Will a try +bulk defer of bugsminor issues from to +much better than the old one and popup pretty quicklyso ! Great +I believe that the component functionality could be extended +With groovyc it works too +The question now is what about the other products +Looking at the code again I am actually not sure anymore why building tailSet in and then calling next on the iterators would be scanning through so many more with newer memstore + based lifecycle supported on all components through annotation of a zero arg public void method with Initialize and Uninititialize lifecycles of all resources associated with a instance are deployed and undeployed with the deploy and undeply of the runtime instance +Moving items not being worked on for M afaik out of for this patch +But in my case I have them with nice default values which are sometimes overridden later +checked in +Will investigate a cleaner solution +Ill try to cut a new release of in the coming days weeks patch includes the following changes Bought up to date to reflect Jonathan Moores recent reorganization patch Reintroduction of interface and reintroductionreworking of Fixed name ProtcolProtocol typo the mispelled version should be deleted if it still exists Updated tests added test test and test Fixed update bugSome notes makes the assumption that the underlying Set implementation used to hold variants is serializable I feel like this is a pretty safe assumption It might be better practice to create a wrapper object as in intermediary for serialization in but the current version should only be a temporary fix until the new httpcore is released then we will be able to serialize directly Should be renamed to? I can update the patch to include this but I thought it might be easier to do on your end after the problems with the ProtcolProtocol typoAdding for failed updates is turning out to be a little more involved than I had expected so I was planning on adding this in a later patchPlease let me know if theres anything else you would like me to include or changeThis patch is submitted with the permission of my employerThank you checked in with some changes to the exception handling code +Thanks Junping and Jing +This is one of the closed legacy issues +I have confirmed that this problem is resolved in. +The test case for this bug is a system test not a unit test that requires considerable work and web deployment of the test case +committed revision +This makes it so that a ij will still work a Derby Client +I just committed this +Ive created with the fix since we have already committed a patch here +Thanks AmitSatheeshSending javaengineorgapachederbyiapitypesTransmitting file dataCommitted revision +I was not attempting to wave you off at all only trying to keep this discussion on topic +Agreed it is marked as a blocker for +Does consolidation include contribicu tooEspecially if we really go the route of individually packaging artifacts for releasing each component separately +bulk close of resolved issues. +Agree with Andreas users have to define the JAVA variable correctly +to set the desired object for the SOAP connection +Shouldnt that be fixedThanksNils +problem has been fixed in Guvnor code base +Tested. +I am using Eclipse version Build It is the +Having Hive own all the files and run all the jobs presents serious security issues since would be running code as root +Well need to give this some you suggesting that the context be passed on a basis? Or would it suffice to have a constructor for that took an instance of the context +I had a funny feeling that the would need some changes as I wasnt sure how to get the locale from the session +So I was going to take care of all these situations +It looks like your patch inserts tabs in a couple places +Please Create Subtasks off this for whatever doc youre working on +Looks alright +I think the original implementation did have a timer per request +Please change to if you intend to submit a fix for. +Thanks Chris! +Ah right not the first time Ive forgotten Fuzzy didnt get the new rewrite treatment +Could you the version of this plugin? could you provide the file targetsonar which is generated by Sonar in the Maven target directory? would if I could but the parent project is and doesnt produce a target directory +Resolving this is a duplicate of +no i havent got the anymoreIndeed it seems ES cannot load itself properly from the Nutch plugin which is a problem +Im just bumping this to the next release but my impression is this is not going to happen +In the latest code we get an error after illustrate Unexpected internal error +has always been the most tolerant and compatible JPA provider that I have used +I would either need to change those to be upper case strings or set the convert in the security configuration to false +Creating the package shouldnt automatically run tests first should it +Attaching which hopefully addresses Dans concern +I am not able to reproduce this it seems to be working just great for me. +This patch has been applied to the goalSending srcmainjavaorgapacheservicemixtoolingfeaturesTransmitting file dataCommitted revision . +please with grant ASF license checked +The failed test case does not exercise the modified code path +Coming soon +Having just forked muse internally I noted that Ive missed the WSRT stuff +I think the attempts to access the wsdl should also require authentication +The links are broken +Evaluate for +Sorry for my late responseIn the meantime i rebuilt the project with cordova CLI and installed the plugins on a fresh install seemed to make it workUnfortunately i cannot check what you requested anymoreCheers +Set Component to Test Harness +Not worth the effort rejecting. +is connector and is it has no sense but thats how sun called the I should close this issue then +This patch looks pretty good +Im not sure that will want this +At the datanode side the block receiver treats a heartbeat packet mostly like a regular data packet A heatbeat packet is put in the ack queue and its ack is also the same as that of a regular data packet +I believe Khaled fixed the issues with case insensitive matching today in SVN rev + AntonPatch applied to LUNI modulke at repo revision rPlease check it was applied as expected +I should update issue description on +In which case Koen will need to work on those though I intend to task him with porting the Hibernate Tools plugin to Forge in the immediate term +I dont have an obvious way to make feature unique across libraries +But we need tests to assure that responds in the same way as the current pig bash shell +plugins and whatnot should not be a big hassle though +Proposed POM for integrating Groovy supportLooking at this I wonder whether it would make more sense for to be a bundle in itself and thus be able to simply drop it into Sling +Assigned it to me by mistake! This patch is committed +I committed this to trunk. +Depending on what we figure out with the release perhaps we postpone this until +Hey BenSo if i understand this patch correctly it looks like this patch will affect gml output as well? I wonder is there any way we can enable this only for gml? I guess the new ids would be ok from a gml schema point of view +After I moved the documentation from the wiki to the source everyone refused to update it and it was a dead issue +This is a duplicate of +TAO has a similar implementation both in its Reply system and in its short object key system which mirrors s object +I can take it up if you are caught up in move issues to and is resolved as a part of release bulk close + +Im sorry it is not well commented code and the test coverage is so weak +Just verified youre sample project to with downloaded from Maven Central. +I started working on Hadoop integration and found the previous implementation of Hadoop provider in OGCE +Im marking this fixed because the test is not currently failing +After all not everybody immediately reads the updated reference documentation when doing a Spring upgradeJuergen +Once you know how to use it its much easier to write a multithread code with it +Can you check if this happens with too +Thanks for any help +Sorry but Im not familiar with the surefire internals +However tree icons and css classes differ +Done +I sitll need to put this on my dev server and try it out +For the second comment are you saying when is invoked ONLY with a sharelib option the WAR files are recreated? Because from my understanding shouldnt you create the WAR files everytime when you run the script? Or the command sharelib sharelibFile should merely just upload the sharelib to hdfs without doing anything else +I wasnt very happy with my solution either. +Dave could you give the fix a try? It should be available tomorrow. +All unit tests pass +Moved to upon release of +Fix version changed toIn JSP EL is allowed only in attrubutes and only message bundle is highlighted for openon there probably something similar toOpenon works perfectly in xhtml file except the EL in attributes where only message bundle is highlighted as in JSP case +Attach my test scenario code +Ive not yet had a chance to pass through and clean out this issue but since you brought it to my attention +Here the diff against the trunk tree without the patches for +Committed revision +create another level of indirection and then swap those units independently +In the latest nightly build this is fixed regards be precise builds from August on should not have this problem in themregardsMartin +bq +Patch that implements this +If there is a folder containing hot in the name it will be chosen as the action source folder +Yes it was a issueShould be fixed +Roger Ive updated the patch and attached it here +Minor update +Looks good +credentials file for sbt scala apache nexus uploadrealmSonatype Nexus Repository Managerhostuserpassword +node cluster with maps reducers with max mins sec Job took mins sec without the patch Job took mins sec with the patch +I am interested in developing this as part of the +This code in Rave looks like it hasnt changed since the initial days so weve never supported this SPI fully +Maybe someone who is more familiar with the parser or gshell as a whole could comment on that or anything else I just said for that matter +Reopening issue +Ok this patch protects the jsp from not finding the named task tracker by printing the name of the task tracker if it isnt found +Fix for layout +See also of this looks like meta data and a stylesheet modification to use it to create a title page for a PDF documentWe need to unify how we handle meta data there are a number of issues relating to thisIve removed the patch available flag to remove this from our patch queue filter and will instead create a control issue that links together all related issues including this one +This issue is no longer present in + added in the parent pom +Theres nothing functionally wrong with the application its just that it looks a little different from before +I appreciate any feedback on this patch +I think tccell should be independent for styles and such things +The access to it would have to be synchronized which would result in a noticeable performance degradation Oleg +I guess you need to install a Solr server to be able to use the +Good catch Ill try to verify that today and look for workaround +Attaching complete stacktrace for issues and +The compiler should optimize this outside the loop since it is final +PS +Please resolve if you are satisfied here +was intended for being pluggable from the start +Fixed in rev Found a tranql minor issue will open a different JIRA +I dont see a good way to fix this there +Contacted Chris O Brien as he does the packaging for EAP +Periodic verification uses that same RPC +Verified resolved in CP CR build. +NathanI am confused +Jacque the painbringerD hehe no problem im aware of it +The catalog is no longer empty +If you thisrm f RPMBUILDROOTusrlibmahoutlibhadoopinto rm f RPMBUILDROOTusrlibmahoutlibhadoop + added a filter to support case insensitive parameter names and change getOp putOp postOp and deleteOp to op +The mapred package is deprecated in meaning it should have been removed in so I dont think we need the fix there +Added the fox namespace to the read! The proposal is indirectly related to what Im doing and it would useful to add such features to FOP +I think create and create are methods from jsf world and from they are are simple predecessor of from unified ELBut if your statement is true then the difference must be in underlying EL implementation Ill investigate it more +However another problem emerged +to trunk and +Please verify using the select and autosuggest components in +These blockers and critical issues are resolved but not verified +that sounds like you changedremoved some previous feature of TLD reading +Seriously I dont want to have this discussion everything somebody comes up with a modules suggestion the is the way to go we agreed and voted on it and it passed +It is already possible two configure two executions of the surefire plugin but I cannot remove a dependency +Yes this is indeed a generics problem +Why do we care about excluding all of these dependenciesSeems like a lot for us to maintain +This will be fixed via that will track multiple documentation changes together in one big patch +Did you forget also +I changed to avoid type inference of the RHS of static fields +Working via added to AS itself +Committed revision +Snipped from the build logBUILD FAILEDhomejrottinghuisgitmapreduce The following error occurred while executing this linehomejrottinghuisgitmapreducesrccontrib The following error occurred while executing this linehomejrottinghuisgitmapreducesrccontribblockforensics The following error occurred while executing this linehomejrottinghuisgitmapreducesrccontrib Source resource does not exist homejrottinghuisgitmapreducesrccontribivy build fails on Ant setup before reaching this bug +This patch has one change to do with sync bytes generation from the previous patch +TTMR takes s and TTI s on my laptop +Agree about Hashset +If someone can take a look at v and v that would be greatI also included a sample which shows how to use Cocoon Forms without continuations from Flow +Please verify this issue is fixed as you expectedBest regardsRichard +I attached a zookeeper dump i took just moments before removing the data directory +Has this been obsoleted by the YCSB +Issue resolved on branch due to jline update. +Injection of service ids is provided for decorate and advise methods again. +Yes its a better way +So you mean the commands in the examples are supposed to run locally +expire after X seconds after the provided timeout will return an invalid do you think about the attached patch +I dont think it is Not A Problem +Ehh +Whilst issue creates a MA this goes a step further and creates a whole launcher but the two are related +Hi Micha this IS a pretty big difference and breaks existing setups +Basically done but the changes wont show up on the website yet +Marking closed. +So mapping pear phar is required to retrive all needed dependencies +Now deployable file is correctly downloaded to the server and successuflly hope this patch can be integrated soon in the main again for this patch +Some existing tests fail with the latest patch working on fixes +Uploading patch for branch along with ant log +close off release +Warning the user it aready exists does nothing to help them since it would be out of their control to rename the remote index +committed patch does this fix the encoding +Patch looks good other than thatWe no longer update it because with protobuf changes the protocols are always expected to be wire compatible +Fixed the regression test failure on Linux EMT +also cleanup some unused variablesm personally a little worried about checking for free memory +Would it be possible to continue using the app attempt object to store all the app attempt related information including the master key for the appAh good +I read the doc and wanted to know what you saw that needs to be changed +So the bug itself has not been resolvedAnyway it is not a bug of Tiles but of Struts in particular of classSee. +Thanks for the info Jesse +I was also wondering when this might be fixed? Thanks for input +I just tried it with Lingo and it works +Thanks for update +I will forward this JIRA to him +And regarding where to put it its not affecting particular EWS build or Tomcat version but general issuefeature of Windows which may happen no matter how much we try Thus I think it should be in docs in general +bq This is why I like the token attr based solutionAlthough I think its more general than An attribute that says these tokens go together or these tokens should be considered one unit seems like nice generic functionality and is unrelated to any specific language or search feature +I had slightly different results from you but I think those will fall under a different bug about gamma behavior etc +it could then have a radiating gradient or light orange to yellow +Ill have to look at it a bit further +Cant verify I didnt find EWS doc on stage +Thanks +It happens because event isnt reported for new IP locations when TI agent doesThis patch depends on previous implementation of and +So if you want this cursor to work you need to add only cursors to the underlying list not a list holding cursors or another list +This is a known problem with windows and capital letter env var names +Currently bundled into junits with small numbers of iterations +How can I add optional jars to the classpath? I want to run the ant FTP task from maven but I keep getting an error that says the task is not recognized +Arpana thanks will do that shortly +So it also tests if there is a deadlock +Personally I would like to support containers so I would like to fix it since we already have fixed some issues that made the use of Tiles under impossibleUsing the message of and exception is a bad practice and very dangerous a more specific can be a solution +updated the docs to reflect the sf issue detail add comments for zkinit +This one uses style cache loader config +but theres no good place for lets write a new moduleWill probably move to issues soon +At position load was removed from one of the nodes +What are the possibilities that the thread which is being the portlet is running in and the thread which is invoking the is the same reason I ask this is uses which has in it +okay +Please commit if tests pass +Agree verifying the distribution is difficultbq +This is great +Bulk close for close after release of +Hi SiddharthMy patch is ready for your review +Patch skips CRC on byte size files and when blocksize between source and target do not match +So I have changed the wiki documents +Thats awesome speedup Robert +Agree this is excessive +No tests included +It is relatively easy to write another implementation +Looks like you guys have done your due diligence in proving this no longer exists +In fact he was saying we should never play with it anymore +You are correct the s oneway call is obviously not handling that exception correctly +Stack trace of allocation that is never freed +I hadnt heard of Clirr before +but we could provide a prominent warning +No I dont think so maintained nexttoreceive which was set to the first seqno received if entry was null +Currently we include hbaselib +I should have looked at the FO more closely +Best RegardsSascha +They are used while generating urls with a product or category name and filtering specail characters in the name +Attached make patch removing javaxswingplafbasic from exclude list +error in postgres version +I guess Ill just close this and wait for a new report. +populated see also r and we are ok to go ahead and remove the incubator release dir for Onami and turn on pubsub for the new dirs +For config the quartz endpoint will remove the job data upon restartI also tried by given unique name but still same problemI would propose to look up the endpoint by trigerName and avoid to store the endpointUri into the jobDetail dataMap to avoid this kind of mismatched problem +xls files would you happen to know what to change for +Addendum to add category +It should be corrected to reflect the new packaging +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +So Im going to mark this as a Fix Version for as well and put up a batch for that based on the patch here +That is if you have a queue with small capacity say but its max capacity is say even if we had the configuration per queue for am and you set it really high it might only be allowed a couple of AMs when in reality if the cluster has no one else running it should be allowed more so it could use the max capacityWe might be better off leaving the max computation using maxCapacity but changing the max to use capacity and then allow the user limit factor to apply +Proposed design document for Dynamic discovery and aggregation of resource document version is the implementation of this new feature contributed by the IBM Team +This issue can be closed. +The configuration worked well without JNDI +I am using with default mean Jboss just confirmed this and Im sure this is something to do with the same issue we had with JSF somehow Seam loads the same component twice even though it should ignore cant fix this for GA +But the blur between persistence specifics and domain semantics is very subjective and I dont have a problem with some logic in the repository and keeping the transactional semantics as close to the data as possible +Logged In YES useridIt seems that the problem is more generic +why is that needed +Im inclined to think its caused by retransmitting the message to the joiner and the originator not waiting for the response too +This doesnt seem to be included in the maven release unless its under another project that Im missingIf I have time I could try using a snapshot of HS in my project to see whether that fixes the issue in my project as well or if theres something else interfering +Uploading version of the patch +A few nits In class and method names should we abbreviate fp or spell out fingerprint? FP means floating point to my eye +Issue is notNeed to redo all and make sure end up in right branches +I can understand supporting only the most popular but find it odd that Google isnt in that list +So its a concern and should be compiled as such putting it in core Hadoop is asking for trouble in the future since the Hadoop releases wont keep track with the union of PB Thrift and Avro releases +I could be wrong but is this duplicate of? +Ive done my testing with the latest nightly build and the transaction suspension appears to be working correctly +it will have better performance as is just a direct access rather then doing parsing or verifying headers +The underlaying problem is in the change of JVMTI implementation for generating VMDEATH event according to +Another idea might be music storage +Ill keep itbq +Resolved fixed +Unfortunately this has not made it into M +bulk close of all resolved issues. +yes +defer to +It runs just fine with my new code +I would hazard to say that with the current structure it is impossible to cheaply rename column families as it would require you to rewrite every mapfile in the entire column familyAs to the patch you misspelled +When changed things so that the empty string was treated the same as a null namespace this broke that schema +In progress verify that the goals are executed both in prepare and perform add form validationsNew do not prompt for tag base for scmDone verify that the goals are executed both in prepare and perform add form validationsIn Progress do not prompt for tag base for scm in branch +You removed couple of udfs in the patch +We can open a new issue if same thing happens in new contextAdded this issue to. +Resolve it was sufficient before +Bulk close for +JacopoSee if the attached patch helps +I think the problem is relate to the is never released +The issue can be closed now +We can weed out any issues we want to push to a later release from this set once weve done the consolidation. +Attaching testcase +I tried the approach of removing the user of and that seems to work ok in geronimo also see the attached patch +if I under stand you want to replace current coding with a Class the generatesBased on Davids past comments using Classes add extra conversionthough I like your Ideas I suggest if be web tool that generates the standard code from the input that could be pasted into say simple methods +Anyway I applied your patch and replaced the two problematic files with the attached ones +updating to keep status readable in main description of the issue +Right +Done +on provided that unit tests pass +Hey what OS you are using? Make sure you do not have a damaged memory chip and your java installation is not corrupt + Username fpiragibe +patch against that complete +build tarball rpmdeb +What is worse I currently do not see an easy fix for it +Sorry for the trouble. +Sorry +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Will post the performance numbers and then commit the change to based on that discussion +From what Ive seen most of the time different versions of a dcache are downwardly compatible +Please fix +updated site +the should be in the directory of srcm still getting the same pulled in your code from svn and ran the m install +i had seen this problem with one of our users earlier and the problem had boiled down to some ptoblem at the leader wherein it got some error while processing the disconnect and didnt send out the disconnect to all the followers +Looks like a path isnt been quoted thus the classic Documents and Settings is causing problems +Using separate thread for write in this point wont give any thing because read mutex lock is set on the file during write to avoid inconsistency +The verified result WARN Used bytes of buffered edits waiting for IO threads +The Portlet Libraries container contains the library +Talked with Sijie offline the concern is due to lastId not reliable in the pastI have updated the patch to address it +is still present but is being written by the glyph +configured redback xmlrpc r of of trunk configured redback xmlrpc authenticator upgrade redback to +Ankur do you have cycles to generate the patch which we will commit now so it makes into +fix for the patch +dwins can you run restconfig tests with the patch applied +Issue is not a Roo problem but looks like an AJDT issue +Can you attach it to this issue so we can commit it and credit the patch to you +One feature that is important is the link between records +I have a parent POM project that has child projects +Thats one option but would involve many changes in +This would also block any reads or other changes in the meantimeHeres the first approach that came to mind to handle not blocking readsupdates while publishing +The seemingly innocent action of enabling MTOM and CACHEATTACHMENTS in the client was all it took to take down the machine +Ah weve always used Ruby Enterprise Edition and Ive come to equate it to MRI but in this case there is a difference +Did you mean or? Comments in patch suggest but linked issue is . +other than that the patch looks good +Here are the files generates the client side stubs the global configuration axis config file the custom exception thrown by the serverCalculator implements the service on the server calls the server using the generated Thanks +Hi Jukkathanks for looking into this +Could you please run Maven with the e switch and attach the exception stacktrace to this bug? is an example multimodule with two modules where modulea depends upon is the maven log of mvn clean test ve applied a fix for this in revision which maven repo do you publish your snapshots to? thanks +Lets see if someone on the ml knows something about it +Also try setting the property or the WLHOME system property +Thank you Kristian for the clarification +Also fixed typo in simple method name save save +Fixed in + +This patch adds a few tests to to illustrate the problem +I could not start the server seems to be some jars are missing +Hi LarsYour patch looks great +Next nightly build will have the fix +Sorry for not being precise and thanks for asking for clarification. +If you find further issues please create a new jira and link it to this one. +added to fo directory altered +I agree with Owen lets postpone the shifting up a level at this point +Issue closed with release. +Just updated to cordova via npm and can confirm adding wp to a project works for me now +Good points +Reverted the caching behaviour introduced in in caaccfdbbedbdbcceadbdeee +Other events as needed can be added to the same interface in future if needed +Committed. +I guess it is kind of weird that has some static methods that take a Configuration and some that take a Job +bq +Good job thanks! +bq +Advise if otherwise. +I was cheated by the name also +WowCan I take this for granted or do I need to double checkregards have checked the list twice but who is perfect + +Mickael do you want me to create a pull request or will you simply remove it yourself +Running regressions +This one seems like a good place to start +I am not quite sure to what extent my code has the right quality for Giraph bud I will do my best +Checked in code changes to accommodate the following business logic A user with either an Application Platform or Developer Subscription entitlement also has access to Application Sever A user with an Application Server entitlement has access to neither Application Platform nor Developer Subscription +Thanks for the reminder +By default the Dbo service is injected by the activator and hence we cannot guarantee the existence of the service +Ill dig in tonight +Would you be able to give this a try from Roo sources and confirm if it worksThanks I got my hands on a Win Vista box and installed everything +Will try later +Verified it still passes on Linux and passes on Windows without Cygwin as well with and some unrelated test fixes +cannot reproduce +The blueprint annotation impl bundle should now use the same versions as everything else which is now +Please if this is still an issue. +Hope i did not forget something +Interesting +Since this list is meant to be a list of all current modules that COULD be deployed we have to respond to all new modules as theyre born not as theyre deployedSo without an event framework here the best I could do is offer a refresh hyperlink under the table +The problem is that we dont set for and it defaults to INFOconsole which is redirected to +This approach could be done with plugins so we wouldnt have to alter the core +Heres theWhile PUT methods should be idempotent in real life often they are not +Imtiazs fix works for me locally with JDBC +Committed at subversion revision Touches the following fileM javatestingorgapachederbyTestingfunctionTeststestslang +This issue has been fixed and released as part of release +Thanks Jitendra +You have to svn add the files before you can svn diff +Just tried it again on fresh machinesRestart still fails and now I see stop works sometimes if you wait some time after etccassandra stop returnsRight now stop followed immediately by start is not recommended +If someone could send a Pull Request via github i will get it into the upcoming beta release if not I will try and do it after the release is done +A simple patch that moves the jar loading code after checking if the CLI session is local mode +Ive revised s code to only obtain Log instances on the fly within the implementation methods not holding it in any kind of field +Ive made this fix in be but this still needs a spec +i thought one of the points of is that we record the mapping so guessing isnt necessary +Hey MichelIm checking on this today in the evening +This will be fixed though the build systemOnce we have moved to maven the minijar plugin will remove the dependencies. +oops +Please add it there too + This bug has been confirmed by popular vote +But yeah I agree the documentation should have stated that explicitly +Its too late to do it for Id say +Correct contains a potential change to fix the test failure +Added the Purple Gizmo to this category +I having second thoughts on having this patch in in and wondering whether we should revert this and keep it only in Tez branch +What would be a surrogate? the exposed ISPN attribute has a default of undefined +I think the consensus is that featuresrefreshUrl works as designed and should not update features +Committed w suggested changes. +on applying to all versions +Sounds like a good solution to me +This bug has been marked as a duplicate of +Ill be happy to finish the component for you +Standby NN JN should be able to get the journals +If not keep the current behavior hand load the item +Also I am not sure how the statistics would apply to this particular situation +This was a simple problem when the AS channel factory creates the package mbean wrapper object for the protocol its neglecting to call the attachProtocol method +Alex +The solution to this one should be the same as described in as of r by adding an additional check for the correct node in +For now heres a patch that takes the easier approach and simply skips the test if EUCJP isnt supported on the platform +If the extension resides in the EAR lib which classes will get scanned? Also the ones from the WAR files? In any case the for the shared EAR jars seems to get triggered before the gets created +Yes could work but I dont feel its the best way to do itSorry if you feel that Im spamming this Jira but Its the only forum I have available +I am quite busy at the moment +There are other problems with version see also and specially which was not backported to +I thought about this a while ago too +GWT supports these +Ive extended the existing authorize tag as described +Better for now to have a set of filters available for the client to choose from +Adds docid Field to the index for +See also for another HP aCC issue +Verified on revision +We already changed dependency for whole Castor to Java so we can close this issue as fixed. +Heres the icon for the first release of LDAP StudioIll try to improve it laterI have integrated it in the different formats and sizes for Mac OS X Linux and Windows +Committed. +Patches were committed at the revision r +Fellow riders please review the patch +Ive changed the implementation in revision a little bit instead of creating a node of type sling with properties for the configuration values a fileresource node is created containing the same output as the configuration admin is writing into the file systemThis is necessary as we need to keep property types like Integer Char or VectorCollection values which are all not supported by the repository. +that typically causes a lot of frustration and increases the treshold to get started with it in my opinion +Ate can you please find time to write it down? +Attaching patch which fixes the ptBR version +Thanks BJ your patch has been applied in the following revisionstrunk scott +Looks OK to me +Another patch for TRUNK +Please set component and affects version so it can be properly assigned and triaged +Im used to chars so this slipped throughRegarding b I changed the class and forgot to remove the private constructor +Please rest assured that this is default behavior by definition but it will happen occasionally that something slips our attention as it happened with this code area +Will be shipped in the next release +to disappear +Could some one from Dev please update on thisCould JAXB and MTOM be used together in AxisHi Naravanhas their been any updates on whether Axis will provide support for MTOM using JAXBThanks AllIm also looking for a definitive answer on whether Axis supports MTOM using JAXBThanksDom +Which means we are not done with the mirror API we need also the javac and source API +Patch implements this relaxationIt is working for me convenientlyTo do add a test update the package documentation about this Shai for the reviewFollowed all your suggestions it indeed made the code simpler +This attribute specifies which plugin controls the branding of the feature +Committed to with revision . +it looks reasonable but have you tried ratio proposed above +Not yet pulled up to branch +Do that then run ant again and you should be good to go again +I updated the patch so that it will pass the validations +I think we do need some way to force strict parsing of the files +Im not familiar with your version numbering and assumed this corresponded to version +Could you provide some more data? +OK I have submitted what I believe should be the final version +The advantagesdisadvantages of sorting is also shown +Is that the way its meant to be +Stephen The source code does not actually exist the line numbers generated in some cases are used only to locate where in the generation process something may have gone wrong +Ahh good news +It appears that the build last night resolved the issue by placing the projection file in the coverage directory with the image file +Instead of building a complex audit data management system I suggest making a log tap that sends audit trace to syslog either local or remoteI agree +Complete version of the +means basically reimplementing JCR +Added suggested comments to respective classes +Properties have the concept of defaults which are used but not written out +It may be needed +product fileTo adjust java options related to memory I can add them into +To see it working please see the comment in file ecommercewidget at line number +because we dont have much control over the jaxb code modelWe better generate the wrapper bean classes from the SEI directly the downside of this approach is well have to maintain more code and have to append the jaxb annotations by ourselfMaybe there are other ways? +Ill keep this issue around to see if theres interest to move it from the sandbox +Thanks Kevin +Now committed properly. +So a mvn clean or a rm target will force a of mvn package dependency but other than that the start up shouldnt change +Duplicate with. +Assuming fixed. +cfscfe and also not recurse into them until we fix and am fine with the patch +Thanks Tony +The new patch passed libhdfs test +I just checked out Hibernate and trunk and reproduced the problem +If not than i will go ahead with adding one for it +I committed this +Patch fixes the problem on my system +My original testing got a bit hacky so I want to clean it up before I share it +It may also be useful to deliver a connector registration script in the same place +now is referring to IR. +for the patch +This schedules all tasks immediately with their natural input dependent priority order +We also need to consider how we deal with replication of Seam managed java beans +The compiler is devided into several phases +Ben we could use your attention hereThe problem is that we queue NEWLEADER before we queue UPTODATE but inbetween these messages we send more sync packets to move us from SNAP to well UPTODATE +must be handled differently +Both of these steps are easily overridable +I could see implementing this though by using a coprocessor which intercepts all readswrites and for every column cffoo first checks a cfaclfoo before returning results or passing through the writeRegarding the multitenancy use case I imagine an deployment of would probably be going through some intermediary layer anyway to give users the illusion that they arent on a shared deployment +The client is offline in the beginning and comes online while sending message the presence is correctly reflected as available and am able to send the vice versa case where in the client is online first and goes offline during the process of sending message the presence still shows as available where actually the client is offlineThanksAruna +Documenting instead for. +It came up with only two buffers being leaked +We are managing projects with a single instance of continuum +I think these lines popped up after I started mcf +Seems youre not supporting Theme Scarborough anymore? Please have a look on the online demo with IE and +Hi all This is the third part of the jira which normalizes mantissa +It works now. +This probably needs a couple unit tests +A solution might be to use a Map of translated objects instead of a global Map +doesnt allow anonymous access that we use has anonymous access enabled +Confirmed on +Could you help review the patch and commit it into the trunk if no problemRegards cloning effect when draging a portlet +Try requesting a receipt for the unsubscribe action so that you are sure that the broker actually processes the request before breaking the socket connection +Thanks for the patch Kay Kay +Attachment has been added with description XSLFO that reproduces problem seems to be that the introduced by columns in the container table is correctly applied to the text in the contained table but not to the linksIve set the severity to Major because of the use of blind tables to control pagination or center tables +Hi MartinI was the one who filed the issue at tomcatIll try to create a small test case +The attached file is a patch +When CXF drops support for Java we can consider thisWeve dropped Java for CXF +The problem is that there is no Cocoon object unless you use Flowscript +Actually an unreasonable value should be rather in the area of a couple of minutes to make sure the test terminates at some point if a deadlock occurs +Apparently and are in the repo already +bq +I have found that this error can be created by allowing Eclipse to build the project +Is there a software limitation that is is Tapestry written with the assumption of a single instance of the filter per JVM? If not then why restrict it +I agree with Sergeys point above +Can you provide steps you followed in order to reproduce this issue +Closing this +Thank youBest regardsSpark will rise other to supply the implementation of methods ofBest regardsSpark said he will provide full implementation for by separate jirapatch so close this one for now. +OK +Thanks Pranay for Reporting this issue +This is perhaps because they want an algorithm that can be implemented in any language +Enhanced method selection for activatordeactivator methods to use the declaration document namespace to decide whether to only consider DS method signatures or to also consider DS signatures +is making all the difference in the way the command is ran by +The patch looks good +I dont think theres any obvious backwards compatibility issue and as theres no code Ive no urge to suggest we dig into this. +hans What other subjectsYou are talking about a problem which is actualy a wish +I am not sure but the covariants support might be affected too +Trying to understand the implication of Todds suggestion aboveCurrently each has reference to the it uses +But if you are removing methods then its not +One more for Alex +It does seem like it is possible to track down the original method body somewhere in the transformed AST but I dont know how to reliably get at it +it is usefull! If not then please tell us the reason why not to add this plugin. +Scratch the last clause on above +On second thought JNDI would expose the whole application outside of the context which is naturally out of question +Alternatively as Arron suggests take out the name attributes in the included JSP +Instead the service field will be the uri of the issuing DFS +The patch has been appliedPlease for meThank you Alexey. +That works fine if we have a but in the case of compactions and flushes we dont one so I have no way to protect the reading on behalf of a flush or a compaction +Yes but imagine you live and code in real lifeIf you almost never can have static method validation do you really want to offer itI might be wrong but it seems to be even in the Spring land only a small subset use the rest use proxy based technology +I also removed all the duplication and redundant and any luck with the patch? This patch should be ok +Thanks for your time. +This is what is weird +Did any one update the FAQ to indicate the instructions for embedding a portlet in a website using Pluto old Pluto embedded mechanism simply isnt feasible anymore with PlutoI doubt something similar is doable with Pluto unless you want to try something like but that is certainly to put it mildly. +Logged In YES useridmatter of fact if you change the very to anything else besides a or any of its subclasses the error messagewill i have fixed this in the Verifier +May be in the documentation i will add that either or needs to be updated depending on the build +Yes I think it might be easier to manage by closing this issue and creating new ones as new help patches are submitted +Thanks Paul the second patch was good +Nothing in that package should be used by any other plugin +These test programs test the ability to get data from the JDBC Connection for implementations that support the get and SQL optional featuresCurrently JPOX fails the tests because it passes the Connection object directly to the user instead of wrapping itNote that JPOX could add a feature to permit restricted Connection methods to be used by setting a property +not yet +And attaching the diff of the code that I wrote +Can you please have a look? Thanks +Whats an interest op +And will commit after adding a separate module for face recognition +The contribution was not applied exactly as expected the directory structure has been changedIve fixed the BTI integration scripts to comply with this new structurePlease find them in reattached archive +and I provide two testcases +The patch file with looks a little bit like Jon wdyt +suggest +pom only only Karl I just checked from Manage my attachments link it says all of my attachments are granted for inclusion +bulk defer to to +Verified by running framework test +Duplicate of which was fixed +thx for the cleanup please commit doneHave not the JIRA rights to resolve this issue. +Anybody opposed to the last patch? This is what I would like to commit to trunk +mmm +The first alternative would require the filters in Transaction and instead + +M might still be the default on that system +jb +Ill open another JIRA +AFAIK oozie lets you configure reties based on the error codes which requires change to conf +The jmx maintence release is actually version clear that the notification should be sent when the observed attribute is null +Canceling patch +Me giving them gsp does not help as either they or I would manually delete the compiled files +However every once in a while the CVS working directory gets out of sync and the I have to force a clean checkout to force the server to detect new changes and this requires me to check the box do a build and then the box again. +scrap the license request I have been able to reproduce with a simple test case +recreated workspace and everything seems to be ok +System test patch to reproduce +The page has been replaced by the page so you did the right thing linking to it for your Wiki +We dont use ant anymore +If a fix is required this need to be targeted toIs this still anb issue? +Sorry for this having dragged out +Thanks Martin +You are welcome to give the SNAPSHOT a try on your system +With files each with say buffers this would add MB of heap which should not be done lightly +Hmm +I am also experiencing this problem with a that has tagBase parameter defined +The patch changes fixed not to include the unused CSS file fixed to properly decode the command which is passed as parameter to it +closing with release +I am amazed I was able to get this put together at all +And we need a setting to work around the bug in Eclipse +Did some minimization for the attached file +FYI +You still wouldnt know the form id from outside the form but from within the form you could do whatever you want with it +Thank you for the doc Daisuke +There shouldnt probably be but will just wait for a little bit to see if other devs see any potential impact +Assigning a patch to include fix for as suggested by Brian +Todd Its the ivy cache that is stale +Maybe we should make the new getset methods deprecate the old ones and have the old ones simply call the new ones + is the improvement +additional comment to side note +It would require more significant changes for example the update handler should probably warn when the value for a field is truncated etc +Then later on quartz when we got it right +To be able to setup encoding to ensure for instance would also be very useful +Maybe Refresh button would be enough for this matterBecause there is no need to show all actions in vpe toolbar for Preview tab +If I have more time I will try to come up with a reproducable example but until I do I guess theres not much you can do +Try upgrading Ant and see if that helps +Ill also run derbyall +Try to avoid doing fancy arithmetic by counting the number of edits weve decoded etc +Find the logs under logs and the filename is +The one using the findbugs is always what the wrapped version of is using +Small addendum to fix the broke to fix the problem in a subsequent commit +See last comment. +my ubu clean install needed the following packagessudo install g gcc bison flex gperf patch ant junit junit epm we now have the Linux buildbot runningWe have RAT checking the files we have the install bits uploading we save the entire log setThis one is done +Checking correctness of passing edge profiler data from JET to OPTThere are still some problems with switches +Adrian So IIRC doesnt the first Xmx win so if you pass it in as an OPTS will that give you want you need? Else want to make a patch w what you want in it? Thanks. +So its doing what its supposed to do for now but doesnt mean it couldnt do something else +The result will be a bundle for each module project in the modules target batch mode will disable the prompt and cause Maven to accept all files beginning with finalName found in the outputDirectory for the project. +Noted as incompatible in + adds a quota +Used Tattletale and commits list to determine which +make the ssd as cache of sas diskssteal bits from the dir +bumping up to critical +I can let check the camelContext property before it decide to omit XML declaration by default +If we run into this again someday and have logs we can reopen itThanks all! +Fixed +Part of a bulk update all resolutions changed to done please review history to original resolution type +on hold until we have tests andor more patch to a config option to turn onoff reopening readers +Which forces all to rollback +If this is the case then does this entity manger follow the rules for persistence context propagation outline in JPA +pushed both upstream +I can provide more details if neededYes please do +The new patch does not require JDK +As an appropriate and are in the repository is going to have a problem +Ill diagnose the cause +There was a thread on the commons dev list about this not that long ago +Resubmitting +Changed Installing Apache Camel Components in section to include more info about the Camel components available. +Tree node events can be tested using nested input which triggers event that bubbles when reaches treeNode and its event handler +Does MANIFEST really need to be in the release tarball or can I remove it from the list? +Undo +And Im not even saying it wouldbe hard to add it probably wouldnt +Before the recent fix it was silently serialized incorrectly +Committed revision Jeffs patch that documents supporting JMECDCFoundation with JSR JDBC subset +Triaging this for the release +I am using the latest version Spring Batch +Provide the full class name of in the beginning of the documentSince this is a guide to users of distcp and not developers I left this outbq +At the level it is not possible to retry the connection because the method would have to be recycled which would loose all its internal settings +Im a little confused by this +the original patch +That can easily be added as a layer on top of this however those interfaces are really geared around a limited runtime model tied to Java reflection and dynamic proxies +fsck functionality is moved into namenode aand made available as a servlet +Assigning back to dev +Thank you for calling me out on this it was my bad +Please attach this for me to test +We have on Hibernate perspectivea Views Hibernate Query Result Hibernate Dynamic SQL Result Query Parameters Console Configurationsb New Wizard New XML Mapping file Hibernate Configuration File Hibernate Console Configuration Hibernate Filec Tool bar Run Code Generation see how well the multiversion hibernate works in M before adding it to jboss perspective +Moving the priority back down to the Seam Branch Ive finished with this task for DVD Store Booking JPA and +what is the resultant osgi package name for this packageHow are we going to satisfy their mention the souce license? is it enough to mention it in the plugin licensedescription +Thanks Tim +The user may want to enquire about the portlet in question which is not a good ideaFor example I have a page with portlets on it +In this case replace old file with the new one manually +Fixed + +Eventually well actually merge the docs and that paragraph will vanish but for now it probably needs keeping +If we could set back time on the db this would seem the natural fix +emm IMHO its not very necessary +Can I get an update on how to reproduce the problem? Will you confirm the error youre seeing? Is this still only applicable to Ant builds and not Maven builds +I believe this is a more general simpler and more approach to getting binary data in and out of Hadoop +Updated the test to A TestB Work +Should test with the and files from +If its indeed mevenide ill post a more detailed issue +Ive modified the code to use the identifier explicitly and run the tests against a Postgres database +the FFIR should check that the resource is not null but not whether the resource exists as you said +Theres no need to rush +bq +The test now fails without the patch and works with the patch +It seems to always go to the top of the page instead +ChanakaMany thanks for the logosasankha +However youd still need to decode it as a double then truncate it into the float +verified deb installation on to andThanks Eric + +will duplicate and fix +Ive tried different target attribs blank parent system whithout any effectMaybe theres a way to listen to an event when the tab is opened to prevent it and force the blank modeOr can I change the IAB default mode from self to blank? probably +I think any of the transports can experience this problem +Hence closing the bug +applied thanks! +look good just committed this thanks Terence. +Committed the patch to SVN trunk with r and to the branch with r +And a missed hunk to +Top margin done +For some reason it complains about the compareTo contract not being fully honoured with the original comparator and not with the patch +Tried on S and looks good +Start eclipse again +I will file a separate jira to fix the tests. +Besides the obvious option of adding components for that we could show the server with a much like works the examples project isnt clustered what good is it putting there? there is nothing needed in the core to do this it can be completely done by the user there are different ways to setup clusters do we need to build support for each one? nodename nodenamecontext name nodenameportcontext name etc etc etc + +If you think this patch preserves the original behaviour maybe we should get that committed and open another JIRA to figure out if that behaviour is actually wrong +r rfscholte for variable with generic typed method call +Please do not include changes to in this bug as any fixes there are completely unrelated to this bug +I dont think we can sync as there where some dependency issues fixed as part of M +These are warnings in JPOX CVSAs I posted on the mailing list a long time ago if you add falseto trunktcksrcconfyou wont get any of that +I would argue that we simply shouldnt support this in the first place +Yes it is regression against solution is to put in maven central +Added unit tests to be included in release +So in this case I am just making it work without a higher risk change that may break the implementationAs for the set field we are a little out of step but has the type conversion changed? I would have thought that value would be picked up by the so you would get a properly typed result which would then be converted to a String and assigned to the field +Ill look at this Stefan. +But in the Axiom case would the xopinclude node ever be seen +catch Exception in +fixed for with the release of. +doDecode will be called later when more data is receivedDo you understand whats going on? Please leave some comment +Youre running the warble command when you also have tasks defined in +Thank you! I must say I have some issues with your formulated answer +Second part of the donation +batch transition resolvedfixed bugs to closedfixed +When you run ee test you need to point to a pig location compiled with +Is it possible to get this fixed? Thanks +LDAP remove functionality has been merged to master in the form of a quick view support +I didnt notice that there was one +This defect also affects the currently implemented workaroundo I dont want to put any effort into the procedure or the SQL language approaches at this time +will still benefit from a simple proxy that does proper timeout retry on or connection reset +Im going to commit this momentarily +Has this gone stale? not sure of the status. +Attaching new patch with the suggested changes +update to in revision tried to build on Mac yesterday and had issues like you describe +All comments appear as requestedI didnt have to refactor anything minimal changes were madeIf this patch is acceptable let me know and I will flesh out the javadocs then I will submit a final version +Some NS and httpd work to do before completion +seems related to this issue and has been deferred to so Im suggesting that we do the same with this one as wellPlease feel free to jump in with whatever comments you might have steffkes +Marking as fixed because a few days ago Ive sent a notification on private so I think everybody is aware of this +Added a new test caseLet us defer this to +There are going to be lots of projects that want to tweak things just so and finding a solution everyone is happy with is going to be a pain +Also tested that properly displays the links by a working aroundSubmitting the patch +The issues are now closed. +The patch for to move this back into when theres a think this particular discussion ended up as a at least for purposes of this release. +Done needs is really hard to test cos I dont have EL available outside JSF container +I think it was mainly for technical reasons +To bad Java doesnt have intermediate access levels between public and package +Im not a spring developer but I work with apps that could benefit from this. + is out side of layout at all in place of image in place of is a bug in I think a manifestation of +Do you have an example project you can attach that I can use to try and reproduce this +Lets see how this does +Id imagine that since same build environment and build tools wed suffer from whatever hadoop is prone to +Passed tests locally +Fixed in r +HiI loaded your project and can confirm that the timings for batch inserts are very poor +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Committed patch a including the deletions to trunk with revision I fixed a few lines in the patch where I had used a mix of tabs and spaces for indentationThis finally concludes the work for this JIRA there is now only one version of the test and it is being run as part of. +is making changes such that image and edit logs will be named with the transaction ids +The attached patch seems to fix the issueSeems a bit too easy however +Freddy Ive just made a test and displaying the SQALE index works for me +I tested it manually +Perhaps it is time to update DNS? I see a number of improvements that can be made to the config starting with using threads to speed things up. +I was little bit wrong while porting upgradation +Thanks Greg +Dont think it is a BG issue. +Ill provide a patch for clouderas distribution shortly +In production I dont want to use SNAPSHOT builds but I will verify locally if it works +Thanks! hold pending resolution from Cache team +works fine with the sample web application are you sure you did set a different id for each table? If so can you give more info on the environment +After clearing browser cache everything is working okSorry for duplicate issue I search for something similar but did not found that onefunny thing is that earlier I tried refreshing page with SHIFT which should reload all resources from server +apply the patch to update the and file +Committed to trunk and branch +comments added on the upgrade page. +now takes care of everything +Since we are not logging much in the code and slfj jars works fine went with the easier change of downgrading slfj +Heres a patch file for the changes Im proposing to make strict work correctly +I agree that a way to specify metadata at the time an image is written is necessary but that should be done for all image formats in a uniform way and should allow you to write fields into any directory +What is the use case for rendering an empty string when someone has explicitly included a call to the formatDate tagCheers will try but I have some problems getting all the tests to succeed on my local PC +Lets come back to this when the docs are fully integrated +This is buried too deeply in type lookup to fix +Would you prefer to remove the methods +Claus try this patch instead +Hibernate mapping file of the second mapping file of the first in CVS +AaronThis is the patch to support automatic disconnection of clientsI have done functional testing with dummy clients usingMy test cases were treading into the category of functional tests so i did not attach them here todayPlease take a look at the patch and let me know your thoughtsRegardsVikrant +Sorry for the name misleading I did not shut down the broker but instead send a large message to it to trigger the +thinking about this some more Im not convinced that you should mix jetty servlet +Can you please provide an example of a hbm and a cfg file? + generate a random id directly +The problem is that the current exception doesnt really give any helpful information as to the cause of the NPE +new version after comment from applied to close all issues in resolved state that have not received further comments in the last month +I mostly just want to see logic for is since its easy to get wrong way around +Thought about it again +is not license for nonsense +Not sure Ill take a look at this today +Oh I see +I had no intention of closing this? Bizarre. +Any help is still greatly appreciated and thank you +Created reviewboard +Update Slovenian resource roller issues related to Roller release. +If you can provide script to reproduce thatd be most helpful +We will of course have a closer look at your issue +From what I can tell this has already been fixed +isnt an URL attribute +Thanks Richard! Ive made all changes to the EWP and EAP tags so that theyll be included in EWP and EAP and documentation +Committed to trunk +Related to in that the jUDDI server must be working first under AxisAdditional notesSince the move to Axis SAAJ sending indented xml as the console does via the input text box causes a type cast error +Thanks Lukasz +Where are we on this BTW? I can review the documentation tomorrow if need be +Douwe server generally has slower startup time and takes longer to reach its peak performance +Gianmarco since we have made no progress on mavenizing Pig +Flow defined in Java already support this in a usable way check the new for an example +Since these are deadnodes that have never tried to connect to the namenode would it be better to specify the port as ? This would mean that the namenode really does not know the port +my fault the patch works fine +Sorry typo there +Ok but can you give me an actual example of something that requires this functionalityWhat kind of information do you plan to make available to the hookI also think we need to be really careful about providing a way for people to mutate the resultset after it has been generated since this work will be done on the server node in a fashion. +Released. +I took an existing VDB and removed the JNDI name +The patch looks great for me +The changes are check for job after incremental changes to clock is changes to just once check two test cases min max check if the job survives min amount of time max min check if after max amount of time the job is expired removed some debug statements from jobtracker checkpoint renamed to snapshot Test description corrected and added some more comments +All of it IMO can go to usrlibFinally why usrlib out of curiosity? I was thinking usrshare made more sense +According to the spec a fallback child is needed for instantiated unknown xsl elements in forward compatibility mode or else an error is signaled +A third option would be a setter on +OK thanks for the update +We have class under which we can use for the reportCan you please update your concerns why we cant do thatThanks therere two existing choices and +Spring does not do so much code I wonder if we can improved the logic in the default and then be fully independent on Spring +return code all threads are killed no diagnostic files created +Just wowI cant tell you how impressed I am with the speed of your responses +In the bug report I mention that its method overloading in an class but I obviously shouldve said an interface +Thank Joachim +The file will be written to the same location as and currently it will overwrite any existing fileI have only tested it lightly but it seems to work fine for me +From RajeshVerified this on the current release +Brian Those are separate issues +Ill create similar one for assuming this one will get accepted +fixed in headbranch have verified this fix in QA +Here is a new version of implementing the two interface with the two added seems that there may be a better pattern for handing these situations +Adding more flakey tests is just going to cause more trouble down the line and it is better if we figure out and catch them before they get in +For symmetry it would make sense that the commandButton would allow this but of course we cannot go against the spec +With this patch Im successfully using thrift under with no problems +That makes sense to me +Thanks. +Changing to Major +There are some alternative build targets you can use which should solve the problem +Maybe my change did fix the problem? Or it was not a problem within the test casePlease if the problem still occurs. +Thanks Harish! +Can some kind soul please review this patch? This patch is essential to make type of jobs run on HDFS +I updated please test and close the issue +Attaching the patch +The VDB will be deployed but the VDB service will not be started until the translator is available. +I could not recreate above warnings +The endpoints is there in JMX as jconsole if its the latest code changes as it seems works +ThanksI reverted a couple of the changes because they broke other platform builds +We expected the packages property to only be set when actually exposing the default persistence unit in the given +Thanks Chanwit and Marcel +We may need to document this for as well +I cant see a way to upload file within the HQ portal framework while also utilizing Struts validation +The branch has the fix +First revision of the patch +update project to use common publisher increase wait timeouts +Happy to take this back to the user forum if the developer community feels this isnt a discussion about bugs +other things to think about w this issuehow do we reconcile this with the login to keyspace before doing stuff authentication change? should we require logging in to the system keyspace before doing allowing schema modificationsshould we prefix these methods with system or something to denote not normal data modification stuff +I get your point but cant it be interesting as an override? What if the context root or the security domain conflict with another package on the same server? In some situations there is very little information about the server where the application will be deployed and it is not always useful to change the source uses the principal of overrides so why cant do that +One trivial suggestion would be to move the codec stuff into a default case for the switch +When you submit any form on the page everything comes back to normal +Bobby have time for a quick review? Tx +Sample fixed in CP CR see some strangeness in the lab networks is still there. +One of them is the usage of the s afterExecute method +Yes the duplicate should be removed. +Patch committed to trunk. +Sure its possible +eg move all to the same classThe converter is well named however the class is also named converter +sjovt Rune check din hotmail konto +Seam or Hibernate OGM +branding provided in revision some reason this now fails when the GEP is linked into Eclipse but not when imported +Thanks Tony +I noticed two fundamental problems with the Cachetable of trees in the current patchset +I only see a failure in. +Is this all it would need +Rob there have not been in changes to the structure of ESB runtime releases in ages thus not understanding why you mention situation is the exact same as if AS guys comes out with an AS or even AS between our releases I dont expect our tools to error and prevent that from being used +But can you do that in another JIRA so that we can use this one for the split effort in futureI created for this +can you make this patch against trunk instead of? I do not think this is something we would want to add to +Good thanks! But it only works with maven right? If I have an adder Maven will ignore it +How does this look? This is a first stab kick it back to me with comments +Work complete and documented updated +Rejected jira +Attached is a patch to make the filtering more efficient by passing the filter all the way down to the +Thanks for the patch Amit! Kannan I internally reviewed this +original author of jsx here +from years back but turns out it became jscience so thats a library with a strong pedigree. +Emmanuel could you precise what s on your mind? I tried a few things by as lock of is on a Serializable id I failed to see where you want to go +Reset assignee so mails go to list +TBH if someone loses their password and this means they ultimately need to reinstall thats what you should have done instead of making a password plain text. +updated description and lowered priority this does not block a release +I am also aware of at least one BM implementation for Lucene which may or may not be what issue is about +This looks more or less related but definitely isnt the same either post your findings on the mailing list or open a new issue to discuss this! +this patch resolves my problem but breaks some unit tests +We dont have in here + +Are you talking about artifacts maybe +attributes will cause the current model to barf out +Attachment has been added with description understanding of what pt means that is implied by the bug report is incorrect +I do see all of the output artifacts which I expect including frame html monohtml and pdfDont know if this will fix the automated build problems though +completed. +Confirmed on trunk +So this particular issue is basically a deal breaker for us +Aaron please go ahead to commit this +JDBC tests run cleanly +This is a regression introduced by the fix for +interesting for sure. +Thanks for the update +authication dialog is reported on instances and images now as soon as you try to list the images implementation has to be refined though the current code is a POC +Ive copied the ajax and reslibs to RF there should be no issue in removing them from RF +RTs usage of doc stores ie reading from them while theyre still being written to will likely break thisTrue we need to cross that bridge w RTActually I think Michael I discussed it already on an issue but I cant remember where +Also it addresses the TODO in +Pull was merged. +Thanks Daryn! for taking a look +applied patch to +Im assigning back to David for review and any other points he may wish to add +yes by Vladimir. +Done +Attaching patches for all branches only difference is that in we cant check for the size of the executors mapThe NPE is now avoided by simply dropping the event and doing additional logging +I skimmed and didnt see anything major but I want to go over some of the build stuff and startup in a bit more detail before I can comment more thoroughly +to add it and adding it should not at all be that difficult considering how well isolated the code is for returning various attributes +Need to look carefully at each instance +Looks like wasnt included in the core SDK until Java +Ive made a start +Yes +Was a mistake on my side when I cloned a previous issue +We could also consider reintroducing something like a builtin role that could be assigned container roles and used across vdbs without the need to explicitly create the role for each +Other than that it looks good +Yes thats definitely a runtime issue +Simon this is exactly what I tried to mention +Moreover Added the logic to check whether it is the last retry +Provide a complete deployment and unit tests that runs against the jboss deployment so we can look into it +Logged In YES useridNot a bug. +Marking release notes in EAP and release notesHSQLDB has been upgraded to version to enable correct handling for let me know if this is not the official version name. +If there is a problem please open a new jira and link this one with that. +In particular what was happening wasclient conf job tracker conf task tracker conf +create +Ive thought of a possible solution +Reopening to account for fieldings changes in revisions files regenerated in revision seems that these are all apart from mxj which should be fixed now. +Here is an implementation for the Windows too +Adding two images one with acceptable result and one showing a bugged case +Should this be rolled back to or +Fixed in rUser can now configure the location of the either by configuring in the pom as property the location of the or using the command line switch + Applied patch with thanks to William issues +Right +Assume closed as resolved and released +Simple patch that addresses the issue +Still no one picked this up +This content was added to the upstream ESB Programmers Guide +Thank you very much for the review Nicholas +Attaching patch for earlier version of hadoop +Attaching a screenshot of the the Java Preferences window on Mac OS X showing that there is no Java +Added to team meeting for discussion create separate JIRA task for documentation update +AFAIK theyre not used anymore +Committed to trunk for +This inconsistency has been removed +There can be a link alongside the disabled link which can point to the wiki page +No code changes +I think that we need to rationalize these approaches in someway though Im not entirely sure that we can +a patch implementing this jira +handles the four cases for schema in WSDL and Schema and Imported Schema. +My suggestion was to make the affected value implementations abstract or similarpreferencesill leave this issue open until we have that fixed +No followup and many things have been fixed since then +Cheers. +For example when editing the propertiesattributes that have enumerated valid values the user will be provided only the valid values to choose from +Though which versions of erlang are we supposed to support for specs and comax age +Patch for +Take a look at the in the package I think it works better and we should have just one anyway. +Thanks +Perhaps testing that the message writer that is passed to the writeObject of is not closed when control is returned from the writeObjectIf I have understood the logic correctly the is indicates if the type class wants to write the element tag itself or use the default element tag which would be one defined by the name attribute of annotation in the mapped class or if that is not defined the simple class name +Verified by Mark +Done in revision +I wonder if we should just make streaming always fail the task for exit codes +Let us know if you experience ay further issues. +I dont know about the status of the but if youre trying to use the assembly plugin be sure to take a look at +I tried ant more that times and it worked perfectly while every attempt with failed +Closing all resolved tickets from or older +As Romain explained using old is an issue +Proposed patch to to set correct value of +I would suggest taking advantage of the inclusion of the Execution Environment enumeration thats included in the JRE management and point to the best matching Execution Environment rather than just the exampleIf the compiler is setup for then use this the compiler is setup for then use this the compiler is setup for then use this use this +no comments in a while. +Sorry patch without the patch inlined +Thank you for the contribution Mark +Thank the dependency fixed the issue +The only alternative is to spawn a thread to do the sends +Also thanks to others who provided comments on the jiras and participated in the discussions +Done in r at JBOSSTS branch +Jon will be in the office tomorrow and I will mention it to him Dave is here today and I will point him in the direction of this discussion +This issue has been fixed and released as part of release +It seems a work as there are dozes of test fails because of locale when I ran the complete suite +Attached management server log file and cloud DB +prashantI just added description For Linux in scale up VM dialog boxPlease see my attachments Jessica +A hunk was missing in the original patch submitted which added egressdefaultpolicy param to create +I also like Gregs idea and will try to go with that +You move Solr to a different server and everything breaks +Im fairly certain that this bug is a side effect of and has since been fixed +The member format has been added to the annotation +Ram are you still doing this or should we close +It is now possible to extend Seam to add this kind of functionality using . +It appears as though this also happens when in a managed server configurationwith Weblogic since in that configuration everything gets serialized +This patch is cooler than all heavy commiting during +Patch committed +to run them from eclipse system propertyhomejpeterkalibshould be added in eclipse launch configuration +JMM mockup proposed by Damon +Please review +I think this is a case where this is an error it still prints the error string but is harmless since it doesnt report Load Error in the location bar +I agree Todd thanks! Another simpler alternative to the patch I uploaded above can be to optionally only report these resource counters at the end instead of calculating and reporting them with every update +I have altered the RHEL installation notes and documented this issue in the EWS release notes +This error is most likely due to an incorrect classpathFop includes a runnable and all neccessary dependencies in thelib directory + +Patch committed in revision +New section Validator Preferences is added to the Seam Dev Tools Ref guide +If you can provide a sample and documentation that will be greatThansSupun +These are great additions +Dont commit it just yetNeed to double check and validate +The RPC proxy is created using block token +seems like we have not found a use for thisThanksDeepal +Sorry DevinThe problem might because of my networkI reviewed your patch and Im fine with itI push it and close this issue now +Thank you for fixing this Myrna. +But by the way in my opinion you should add your suggestion soon because Xalan is the default XSLT Processor for JAXP and this is used by JAXM +I had some stale stuff check into git locallywhich was messing things upThere is still a minor formatting change in DNS but that is all whitespacewhich indent hosed +Applied to TRUNK +The task is temporary delayedbecause now Smooks plugin is in the process of active development and modifying +The result is that my genericUDF failed to be instanciated and the rather than reporting the error returns null +Of course Alejandro has already implemented it In addition there are showButton parameters to override the default logic for showing or hiding the buttons. +the release note for this issue +I hope ill find time to write a few lines +fixed in beta is now named +since this is an update not a new plugin changing it from feature reuqest to bug. +I ran with weme and did not see the issue +No fix is in. +Because they make use of the response and the send Ive not added new tests for them +pid fileBrock +I also created a very simple example of what I now believe is the JDK bug at the root of all this +These work fine +Fixed with merge of feature branch +I havent looked carefully but do you all think this would affect the line as well +I just tried it +the sooner the better so then +Verified with ER and closed. +Patch applied +Are you planning them to be removed on next minor compactionyes + +You do not reallyexpect us to duplicate the whole API just to have methods that will notthrow questionable exceptions in addition to those that docheers Roland +Pushing to to to out of the release bucket as there has been no progress in some time +These plugins are part of the Eclipse JST Web Services project which is distinct from the Axis project +No additional information supplied an not enough to go on here. +No legitimate excuse +Let me try to fix it if youve not yet fixed it +This would work but itll mess up the byte order completely +Thanks! +The test HTML files extracted content will not contain the word footer if the is used +covers the update +Any news on this patch +BTW should be +Refactored packaging for the tests for the branch +I looked at your new test and have the following comments Can you change the test setup to not have assertions as recommended by Dan +tests pass system tests passThe patch looks good the block comment in describing the layout is out of sync + +Fix checked into trunk +Hi TimI based on revision +for new patch +Duplicate of +I will initiate a thread for the same +I cant understand your issue +Patch for review +To find out that the issue was fixed with the patch you need to look at CC runs stats for some period of time after patch is applied +Thoughts? +Willem do you mind digging into CXF from the given stacktrace and see if the NPE makes any sense to youNotice this test ran on AIX boxes and failed on both JDK and may not fail every time but we do have a few AIX test failures with CXF I have disabled some of the mtom tests to have the test continue and test all of Camel +I forgot to set the service fields when its not Serialized +When Avro specifies a secure etc +I should add that in the logs in the link above the timestamps on the AGENT node are UTC while the MASTER is EDT +See also +Ping! Could you please setup reviewboard for Tajo +We should also avoid the terms client and server since these are relative not universal +Additionally I created a and a that subclass +You can support it elsewhere but you need to be backwardly compatible +So it wont be for all future calls but it would be difficult to ensure you got the implementation you wantedbq +All set +So all issues not directly assigned to an AS release are being closed +support config interceptors in Since this patch depend on some changes from cxf so need mvn U install to buildIm a bit concerned about the change to the class +This is supposedly so to ease portabilityAs a result this patch should be applied for all Ruby implementations not just +I plan to commit this to trunk shortly +it was about sending results yes but this is done by transforming the docAlso you only see one use of the shows and lists in your text +Added patch which patches to test desired behaviour +To maintain backward compatibility with existing hooks this is enabled under a config setting that is disabled by default +Gavin is this being done in the location or are you applying it directly to each JIRA +The fix attached does attempt to cover all the threeAlso all the required canon files are modified to now test the fix with correct values +is only weaven into the route if you use the route builderFor manual code such as creating a consumer from an endpoint will not inject the got a situation set up the addremove synchronzation callbacks but they are lazy created just before the Exchange is processedSo in another componentconsumer componentproducer you can not add the synchronization before its processedBut you need to use the async processor and do it in the callback done method +for patch +RichardCould you please upload your WSDLthanks think you meant and not +As for Servers View DND Im working hard to figure out how to implement it within thhe CNF framework +nd edition added to acknowledge some pointers from the dev list +This patch removes one of the duplicates +Note that in FC a receiver sends credits before a sender runs out of them! This is governed by minthreshold minbytes +I tried connecting to the database with ij and compressing the table and it compressed ok +Ahh seems like a good find +Latest patch now committed to trunk. +Changing Expressions facade also +an SQL delete statement is now used instead of an SQL statement. +So this was my experiment Apply the patch to +NOTE this doesnt solve the whole issue of proper handling of redirected pages from the point of view of scoring and but it does solve the original issue described here. +Please provide more information for this issue to be investigated. +Heh chown blows up on Windows as should not break completely when some functions are not implementedsupported on particular platform but rather behave ase MRI raising. +Complete with comments and unit tests +attached retry sample job that demonstrates the behavior described closed as resolved and released +bq +In fact as the last comment in that thread says maybe voters are not the best approach in that situation +There are likely to be configuration options desired to wire in extensions +From the code there shouldnt be a thrid syncRequest call anywhere that I can see so its really confusing me +I think you are right regarding the clash just aplied the patch and it breaks +Ill do it as first thing on Monday! m sorry there was a few HTML lines inside the first version of the attached to changes at geotools level +bq +I believe I have resolved this issue +The table starts below the end of the menu in IE +The reason for discouraging them is because they break shit as demonstrated by all the people that wrote in to tell me I broke shit with the simple paren wrapping approach not to mention that this is the sort of engineering that leads to SQL injectionJust because its the easy answer for some folks doesnt necessarily mean its the right call +This means any changes to the IWC instance you passed to IW will have no effect on IW +bulk defer of feature items from to +Upon closer inspection Im really talking about Cruise Controls spamwhilebroken setting +The problem is that the generated url is missing the +This change will be available on for verification shortly +With pluggable web service stackes this cannot be done +SiI added this condition to clear out the paidDate field of the invoice entityThis condition will run when paidDate exists and we are changing the status from INVOICEPAID into INVOICEREADYIn the we are setting the status to INVOICEREADY and calling the service setIs there any better way in your mind to clear out paid dateRegardsAshishHere is the patch for which resolves all of your commentsRegardsAshish Vijaywargiya +Ok I think I am confused +Works fine in related to the one of . +if a class is modified then all its javadoc warnings must be cleared before committing. +Hi PaulexThe fix looks fine thanks a lotBest regards by Andrew. +Theres no reason for this to be configurable or reformatted its text +Resolved +So the lock issue is not really a problem but not sharing the dir is depending on the impl as I mentionedI first tried tackling this outside of the directory factory this quickly gets very difficult most due to the new that replication uses to avoid copying a full indexIm now exploring simply have the base directory factory cache all directories with the path as a key how the ram directory factory works now +It works +Guys if I understand your discussion the force should not be called when a is raised +Committed. +Tobias the regression is for Heikos first exception +Proposed committed in revision . +Attachment has been added with description patch Applied +avalon logging +Thanks TedThen lets wait forIf is not going in shortly will you be ok with this simple patch +Build failure is expected for this patch +Please provide update on your items as a comment +Thats not a bad idea but Im not sure how itd work entirely care to formulate a patch +you for fix +Accepted changes +If you have a design document that would be good if not just a quick overview of the changes being made +They are alread available on the web +I seem to remember that creating a is a fairly expensive operation so implementing this would make things a bit faster +Moving is critical there are comments about this in without it resetting doesnt actually work as some of the information from a previous build sticks aroundIll fix the whitespace this wasnt a final patch more of a proof of concept +secretary would be ideal +Actually need to register for both and so this patch is an improvement +bulk defer to to +I dont recall for sure which commit it was that touched all of these but if my memory serves me it was some change that made the compiler compile on win +handler name is removed in by Unreal Jiang +I just need to polish some last details up +patch looks good +Cannot destroy threadgroup until all threads have stopped. +This issue has been fixed and released as part of release +Ill commit the proposed addendum with the additional log messages you called out moved to TRACE level shortly unless there is further comment +May be we can expand on that +OK please let the various project leads know in time for the SOA Platform release +TP is generated using an optional get latest flag +Imagine that instead of a union type the list contains enums and you remove an enum value from the type +Maybe we can make the links to the nightly docs relnofollow +This sounds reasonable +ACKd on behalf of mark proctor drools project despot +this is available in the current beta or +Create a new JIRA on problem related to deferred to +Thanks for taking this up +could you provide a test case +Thanks! +Committed to branch. +also adds generics to the +We released archetypes which should be fine. +Matei is exactly right +Im interested +I finished the JDO part of the documentation +If you are a subscriber please forward this message to to get your new address includedSo I sent email to Thanks +It is just that the Namenode implementation of REST API returns fileId as an additional propertyI think your question is why returning fileId as an additional property in Namenode? The reason is that requires fileId +Regardless the specific issue of operators with is discussed in +The third int is the down skip level pointer +It does not know that the providers are authorizing the request +resolved +IT added +I couldnt say it depends on what you intended to implement for index aliases here +ConclusionId start with supporting the single table and discriminator approach +The above URL does not work for me +Patch applied thanks IvanCan you please cross check and closeCheersMarcus +But lets get this in first so we start from a base of compatibility with EAP behavior +I experience the same issue when attempting to delete a survery from CatalogStore Kojo Do you have a patch for this + updated the hive logo a bit just cleaning up the lines of the body and the of the ears +is it possible to create ajinclude programmatically in version? +Attaching a patch that for j checks for property and if not null add it to the command to be forkedThis works around the problem with IBM jOr should we check for always not just for j +Verified upgrade +Marking closed. +Thanks for scrubbing these sections Kim and Dag +looking some more +this issue the same as perhaps? Trying to perform a operation in Websphere +svn folder +If the field is some other type like a sorting on that field cannot possibly use the same amount of memory as sorting on a Java int field +Not intended to be applied in svn. +I think the unit tests passing show that this is currently a speculative scenario +Java! Groovy +We already have this dependency due to +Tried to write a Tycho plugin relying on p rather that to perform generation cannot consume p repo easily from TychoThen I realized that using anything else than would imply lot of difficult changes so +Revision same for Double property related tests +Plexus compiler fixed in ffcfaaabdabfefbfcf +This is not even a bug +Im not sure how easyfeasiblesensible would be to get rid of that dependency? the patch introduces huge diffs +would be good to get verification from sql server and db users since i remember having issues with this in another context than hibernate but that was a couple of years back +Oops! Im going to write some standalone slower tests that actually split larger amounts of data so we can get proper benchmarks and do a better job of finding these racesWill address your other comments in the next version as well +Im still rather unsure about the status of this patch +This leads the user to think the web pages havent seen any developer love in quite a while +Documentation +Should be fixed i locked down the dependency to a specific snapshot built specifically for java and rmeoved the mapfish repository from the printing +So in your case the region returned by get would have been null am I reading this right? Is it even possibleAlso do you want me to port my patch to your branch +Uwe ok thanks +In other words perhaps getProperty should do the check for null instead of is then it could convert the return value to whatever you would expect if it was say +But in this case it may just be good enough to store the estimate row count as both the row count and the number of rows. +bulk defer to defer to +Also this needs to be tested on Windows Probably by IndikaAnt under load +Like the problem is that the script is rendered before the component and the component is not yet present in the dom +There are actually products SOA Platform Data Services Platform and BRMS +Update test to and add Ignore to test +Attached patch with fix and a related test case +Ah great +Please the Jira to continue the discussion or better still open another Jira to address exception handling throughout the projectI do feel that putting logic into exceptions is a choice that causes more unpredictable behavior and is hard to implement +Kevin Cormier agreed to create the new testcase for and +Will test and commit +Right now we are holding up on the AS profile as it will not work out of the box until due to s a different issue +shows menu selection under metadata menu shows view for display of the SQL Reserved Words +Trying to debug +Try testing it again with a shorter path +Id have waited with this for release where we will also havecompiled docsFor installing docs and samples I would like to have at least separate and targets like we have a make samplestarget +Will check it inPlease file all the follow up tickets +Wed appreciate all the help we can get +the system test has been moved to the related jira +It would be great if you could test your use cases with the latest master and let me know +well even if it sintentional would be good to document it somewhere so we make sure library writers remove the doc content before adding deleted true agree doc content should be at least removed at doc compaction +I dont have time to dig in till tonight though thought you might shortcut me to the answer +Fixed in commit on trunk +action in JBDS to accommodate these two new variants +Notice the difference between Sat Mar UTC and Wed Mar UTC s fixed now +FOP embeds in document only used gryphs of font +Also still needs a test but Ive lost interest unless the code dupe can be resolved while maintaining the speed gain +I think a copy is now only created when it is needed so in cases where the map is modified to included more destinations +I must have overlooked some other change I made when changing to another serialization strategy. +I dont quite understand the solutionI have added the following to confI then restarted Sonar but the problem still persists +I plan to commit this later today +jar xvf +I just committed this thanks Terence. +And I just learned that we didnt have any integrity constraint in database while testing +Classlib patch applied at r +That one surprised me and I discovered that its not used outside of the package bc were return in the code instead of +and for attempts to move attachments between documents +I believe it is with the error I created this issue for +Does this work in? I still get errors +Thank you maybe we push this issue into and dont implement partial solution right nowAnd the more when we come to the were going to refactor this solution with to make it more KISS +Bug better solution will be to accept ldif files without a version as being version files +The maximum value that can be specified by a user via in MB +The patch is committed to the trunk +I am using the connection pool that is offered out of the box eg I refer to it using the osgi that works nicely +Feel free to reopen if you can provide a repeatable test case. +Can we have a set of severity levels that helps figure out how hard the error isThanksDanushka +Migrated image sequencer from to added new root folder for sequencersremoved in sequencer classfixed various inputoutput path issues +Ha I feel dumb +Attaching which makes the test case use a different assert method that checks the actual numeric values returned by the result set rather than just checking the string representation of the values +Simple but does the job +Checked into develop branch +The bug only happens if you add same field name as indexable and DV and then in another segment change its DV type + +Test case illustrating the issue +I made a patch from eclipse on linux +Updated the patch with log message for retry and corrected the typo +Thank you +Reviewed all these issues and marking them as closed. +In addition to the change has to be reset in to ensure that its instance has the correct clusterId set and therefore the correct token is selected when setting up the Connection +Marking as do it later +Patch committed to trunk! +Sorry for missing that bit of infoSo I guess this is acting as it should +Missing a from the previous this is the correct one +Adjust the to suit +Biggest key cache files are around MB +This woulkd be a critical bug since more than half of the tests in the TCK have this scenario +Its called here +I can not this issue on my machine +Theyll scratch their heads and say doggone it right here in my code Im setting it using connectionAttributes! What the heck am I doing wrong? Doth mine eyes deceive me? djd Probably not always clear but connectionAttributes sets JDBC attributes for the connection and here we have databaseName being an attribute in connectionAttributes and a Java bean property for the object +I used the existing JBUG favicon design to create an icon at pixels and pixels for the User Group channel on twitter +Also made stoppable +setup aeolus configured w vsphere support +We should open a jira on Could not read from stream problem for or +Any suggestions for a new nameWhat about the class? AFAIU delivers quite poor results so should we just replace the old class with the new one instead of adding a second oneWDYT +Didnt test it against a real project though +Thanks for your contribution +if reports Oracle for then that wont yes having jvm name reported somewhere would be it into Flash Version would make that field even harder to use does actually report for the various VMs is what need to find out and then where in GA request we could put it. +Updated patch +Rob any progress on this? Currently not even local connection to JMX on EAPER does not work +Is PE even still useful +As an option this is not needed +Can i suggest taking inspiration from proposed JSR +This patch adds some exception printing code to a utility method which wraps a around another Java exception +Thanks. +Should be fixed now +Not very urgent thoughIndeed lots of refactors are required +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +Showing the native with the result of the run of the Groovy class +Looks good to me! +It looks like some of the wiring is in place for this +Committed to trunk. +Yes this part of the build is now working for mePerhaps this should be tested regularly so it does not silently break again. +HenkWhere is IIM when you need itI fixed a problem with a crontab entry and corrected myThank you for reviewing the mirror and again Im looking forward to be part of the ASF communityJohn I just added your mirror it will appear in the list within an hour or so +One will have to fix the test so that the output sequence is deterministic +I committed this to trunk and +By the way because the fix will be to disable the effect of using the incremental feature through objects the obvious workaround is to not use that feature when you create objectsI know that might not seem desirable for your purposes but Igor and I went around and around discussing this problem and there doesnt seem to be another way to solve it. +Just read the javadoc its much better now +This patch had rejects in it for some reason +bq +Oh also theres a trivial patch reject regarding the clean target if you apply first +Actuallybecause were pausing the writesdeletes when get reader iscalled on the DWPT JMM concurrency should be OK +srcmainjavaorgapachewicketexamplescompref +I only removed couple of loging logging substitutionsPlease check that the patch was applied as you expected +Again I will finish that after the merge +But it helps anyway +How? A we change the property name so we can use the new one and still support the old logging some warning +Fixed in revision to the branch in revision . +addressed teds review commentsnew version is also available in reviewboard +Removing solves the problem + +We have to address this issue after the v release +Tested on eclipse with source from git +Although with ASM I guess it would blow up rather than the heap if there was a similar issue? +bq +Specific doc detail holes in the Controls and WSM tutorials have been addressed elsewhere. +I will update the patch in when I found one available +It sort of works but its a lot more fragile than just using subprocess and connecting pipes to stdin stdout and stderrI think it would be a good addition if the hbase shell had a option +as a +We seem to be going back and forth with this change I made +This work removes that source of conflict and no other +patch looks good to me +Sergei you are making so complicate here +Setting as RESOLVED if any doubts feel free to reopen this bug. +Committed with nits fixed thanks +AlexI am not very familiar with PNM but it seems that there are several formatsthat correspond PBM PGM and PPM +Did you verify that this doesnt break Sitemesh decorators that use Struts tags? The reason it isnt cleared directly is that Sitemesh decorators might contain Struts tags that require an +We have made good progress on Seam integration for the upcoming Seam release +Kind of like in Magnolia +released +Revised patch that works with trunk +Heres the same sort of patch to it avoids the String based conversion and thus will work better if any terms are true binaryI think were good to commitI will still do some performance tests but that shouldnt hold this up IMO +Doesnt actually fail anymore +Lets track outstanding stuff in separate issues. +change would basically load in the system classloader +the in the lang suite +Yes there are reads from a view at the same time as the compaction +Ill take a look at them and see if cleanup is needed there as well and if the new mechanism to prevent waiting could be used instead of the wait flagsThere is also one occurrence in the tests in lang if we dont commit the following insert will block due to fix of beetle commit The following insert statement shouldnt block for minutes! The compile of the trigger shouldnt wait for timeoutIll check if this still is a problem +You are totally right that there were merge conflicts +A release for M at all +Other appenders would fail if they were used with this layoutIn its place I have added the helper class from your github repository and updated the appender to use this helper instead of the layout class +Backported to docs branch at revision . +Problem node ip is problems introduced in the fatclient logic was changed a bit too much and the is check was assuming fat clients had dead state thus calling set to effectively not mark them as fat clients +Looks good +Doesnt seem like a great answer to me if you want to use please make sure you are constantly indexing and reopening so that you dont run into problems + +bq +testCase Please try this test when working on fixing this bug I want to make sure the test passes when this is fixed and let me know if there are other simple cases that I should add +Patch v reflects findbugs warning fixesI am fine with partitioning work by areaEach participant can list the names of classes he works onI will continue with the first two batches of reviews including renaming the coprocessor hooks +If we want to make the syntax for aop better we should introduce a little language say along the lines of aspectwerkzaspectj +Cleanup global connection Map when inactivity timer on a reply queue is not caching JMS connections properly leaving stale connections in a broker +New patch +As I commented earlier no response change either +It will be another day or so before we are able to +The shared key for this handshake is created by reusing Hadoops existing security infrastructure for which relies on a secret key shared between the NN and +Can I please have a list of confluence user names that will have initial admin rights for the new space + +So it not existing is a serious issue +glad I found out NOW how worthless you areOn to to report your issue that you claim is INVALID and make a bigger stink out there on you +It also added a test for paths +any idea Why are we not generating the wsawAction for output and fault in the dynamic wsdl +It could at a pinch process CSV tab separated or many other line orientated text formats and the power provided by nested entities andor custom transformers is considerable +This has been partial resolved with because the table properties in the tabledescriptor can add new fields types to the fieldmanager +The jobhistory component already saves the status of jobstasks on the localfs +Is there progress on this issue? It would be great to get rid of that redirects +Checked on JVM solarisx +If this is a paragraph we want to style we should move the styles into the CSS +I can address it in +applied the patch to trunk +applied to this record since is now released. +Fits in favicon and every place where you need an app icon +I dont have the time right now but in the beginning of march I can probably read up and do it. +I dont know what to make out of this +Updating a new version which replaces the first versionThis adds the comment Rick asked for plus fixes some whitespace issuesI will commit this soon if no objects arise +Added property Skip Duplicate and description to the Idempotent Consumer pattern in the Fues IDE EIP Reference Guide +It should not hang or loop indefinitely however +Ill investigate further temp workaround is to override this method and negate the super response just to play along with the buggy code +BTW Great to have a patch with tests and doco! +This log line from the trace has hostid specified making no checks on this host looks like admin test causes one to wonder if this new behaviour in ACS may be dependant on who is doing the VM deployment and how this user is authenticatedauthorized +Yes dropping the +Weve been struggling against the hang for weeks crawling through traces and we get no conclusion +Posting for review +Completed and verified for GA +You know these hadoop jiras are like death of a thousand paper cuts if I had known we were going to face so many issues i would have proposed making a larger shim layer +Fix and test case attached +If you add a default constructor as well constructor injection should work +instead of expanding it to individual imports? Keeps it briefI used in eclipse to fix the imports +ve just committed this +William Just to be safe which of the patched submitted should be reviewed and committed +If youd like me to give that a shot I can try it +ok I will do it in this weekend +I have a similar issue whereby we have domain objects already in the system which are uniquely identified already by a String +The library would export the symbol and one of the library headers would import it +Ill give it a try myself with some ADB code to try to determine whether this is or a general issue +Pretty much any time that it used to hang indefinitelyThat was fixed by it a failure if the producer or any of the consumers fail +Why pass a class to IW only to have IW invoke methods on that class to get the merges out? We do this for the natural MP because on events like flushing merge finishes we need to invoke it +now returns is the blob is not yet materialized +Eli +Hi RyanDid you get a chance to look at itThanks +to proceed +Please attach your tested patch so we can take a look +Since this will require deletion and of services Im marking this as affecting Documentation and Compatibility +Btw my patch also breaks current client to server compat +bulk move from M to to m from m going to close this one since weve been working on the decoder support for this release +Im really not in the mood to change it at this point +this be applied to the branch? +Thanks Kim! +Closing as by design +ve committed this +Especially the latter as the token is decided automatically and already before move command +Either one is fine with me +So resolution of this Jira is simply a matter of changing the trunk to refer to the version instead of version of the lib +patch applied with thanks r +Previously it was not set therefore the default is determined by cyrus libs themselves +hdfsTimeout is initalized after construction +To be precise we need to propagate every field from the AMQP method +Which mailing list would you like the review group associated withDo you have any signed up usernames I can add to the review group initially +IMHO we dont need to people who can just as well manually avoid configuring too large min values until the proper fix gets released +Just noticed thats thats why its using Hits it has been updated to use non deprecated since thenI have fixedchanged up some other little things though or at least improved including updating analysis section to be more up to dateIll look at the demos next to be sure they are good +Attaching the last few lines of the output +Applied. +I see no obvious file handle leaks in the code +for arguably the ability to hose a couch by a bunch of docs and calling the replicator can be considered a bug. +We agreed to simplify the solution and isolate it from the fix of +I should have goen thorugh the bugs before filing and fixing it +To get around it we temporarily moved our data into the public schema for testing but would prefer to use schemas when we go to production later +It would be nice to include in one of the test sample so people realize there is a feature of this +or do the names need to match the java class +NamitAre these javadocs enough or there should be one for each method? I guess since it is implementing an interface it is known what every method should do so I just added comments explaining details about how it is done specifically for this implementation +The package is also not necessary +Corrected patch uploaded +Youd need to keep that entire nested structure as a single unit for as long as possible I think +Hi HongSure that would be great +I know from the previous images that there was a problem with the double data points which has since been fixed +My suspicion is that if Hudson runs slow it could run out of time on this +But certainly there are features Id like to use as well +We should not have two methods for storing that information in the context +See. +Do you create a completely new for that one host or docan you override that one host directly +anybody who encountered the same issueif any would you post here on how to resolve this issuethanks in advance +Resolved at +Just to note this is a single threaded test but the problem appears to be thread and generational collector related +Ha one of authors is Kohsuke +On the subject of our position in the Matrix I was wondering if we could be promoted a bit differently +What is the best approach to fix this should the initialisation occur lazily on the onclick or should the date format somehow be passed to the initialisation +pushed Released Closing +bq +VerifiedClosing i dont seem to be able to reproduce this. +The rpc handlers thread that is processing the increment operation will block till the is complete before returning status code to the client +We are running web services in a closed environment and those services wont startup without internet connection +Thanks Amila for the contribution. +Closing this record since is now released. +part of release. +Hey Stack in the patch I uploaded I had taken into account your comments +I shouldve been clearer +Bulk move of M to the issue to next version since WTP still doesnt support EARlib folder properly +We should finish this though +Slight caveat to this updateAny interface that has or on it cant extend otherwise it wont workAt present I dont believe any of our interfaces do this so it should be I deployed the time and date example in in to AS There was a deployment error related to Faces and Pretty Faces but the Application Startup Bean that the app uses which also happens to contain the only Inject Logger worked fine and logged all the expected ll work then. +Thanks +I just committed +No luck just tried dev snapshot from Mar on my Mac same behavior +Thanks for the review Peter +Change delivered to tag +Any other commoning of classes for me is a deal breaker because it greatly changes the opportunity that more than one service will be dependent on a jar that you change on a web server and keep services from being independent of each other +I think this was a bad decision to mark this as NOTSUPPORTED given the multiple SELECT and INSERT statements in the code +It might be better to fix the profiles issue first and then fix this one +I make all tests using command line with FOP +Sounds good and clean to me +Looks like I had written it this way to differentiate between task being alivedead vs an error in executing the command +Added comparators and unit test for them in which of the three patch files is it that should be applied? Im a bit confused +If you need to use these characters in cookies use the following I dont like and in fact you described the way to set for Because there is a orgapachecatalinastartup in the and it should I take out the workaround then? Thanks +I am attaching a patch Ckmarsdenpatches +SvetlanaFixed in ARCHIVE module at repo revision Please check this fully resolves your problem +verified +Perhaps it should be +Thanks for the fix! when can we expect with this fix +Ill take a look this week +The unused imports were removed in r +A data node will not start up if its data dir is not readable or writable +btw no way to deleteedit a submitted comment? sounds inconvenient Yeah a long time ago it was decided that discussions were simpler to follow this way +Aaron and Tom for the review +Leaving this open for now for verification and further work +Why not execute them locally on another rackAlso in the interest of IO interleaving you want to do some amount of off rack reading early and not wait until the job is guarantied to be the long poleMy bet is that some amount of pre planning is going to prove necessary to get a semi optimal plan +Ive committed this +I think this subject for the issue is clearer but feel free to change if you dont like it +Fix Javadoc do not get warnings for these things with Sun Linux JDK +Hmm Im thinking we should relabel this as need findbugs support in ant and mark it resolved +I just committed this +thanks for the validation andy marking resolved +Hi Tom The patch looks good to me +Hopefully well find a workaround for those escaped partition names soon. +is the order of seqid and bulktime sorts reversed between an trunk? No idea +The issue only affected persistence unitsChecking in a fix shortly +Verified +We did already and no one responded so I think Facebook was probably the last holdout +I have attached a simple test maven project that reproduces this ve synchronised the access to templates with lazy creation as part of that +Since this is generally not a good idea lets close this as wont fix +This might also end up being the end solution for this ticket and then future tickets might want to improve on the solution +Thanks +I fixed some javadoc comments moved the directory and my changes to to test the JSON codeThiru Raymie tells me you still intend to do more cleanup here so Ive not committed this yet +sln extension indicates some version of Visual Studio +The implemented solution does not work for delayed and factory service components where the registers as a on behalf of the real component +update description +now its an enhancement and not a bug +The correct value for publicweburl is httpsomeserverpublic +See TP now includes me and a compatible. +Reverted the and temporary config changesI decided against doing the optimization for the checkSatisfied as I dont think it makes a significant difference as theavailable bytes computation is pretty much in memoryOne more comment on the usage of metrics it is possible to combineratehistogram metrics into a metrics Timer object +Site far as I can tell your assumption does not seem to be correctAttached a sample where all subprojects have distributionManagement urlsThe staging site navigation seems broken The navigation links do not point correctly to their parentI would recommend building the standard mvn clean site sitestage within the supplied project structureThen open the of the targetstaging directory in the root project and start navigating +That is cool Mike! I can put that to use in my FST text tagger work + into trunk and +Posted it but i can see some warning at top saying This post has NOT been accepted by the mailing list yet +Now its really fixed + +moved issue concerning system scoped artifact to. +This is a duplicate of which has a fix included in the framework release currently under vote +Ok I committed the patch and solve some bugs +Please check +The bug persist after commenting this line +Any more comments? Or it is ready to commit +Because they are separate projects jars they would each have their own POM +Sorry I just read the guideline for reporting issues Ill soon create another test project according to the guideline +Hurray! And by script you mean on Windows and jruby on nix? There is also a native jruby launcher on nix as well just in case. +Verified on revision . +ant test completed successfully +I just committed this +I guess thats my queue to go to bed +It always should reference +It requires neither deprecation exceptions nor instanceof +implements this. +The patchhere wont be applied +Stepping through the code sometime ago noticed upgrade does invoke this internal mechanism to update DBA as the owner of system schemas to authorizationID of the invoker of upgrade but probably because of wrong internal transaction semantics the change doesnt seem to get committedFor the second part of the full upgrade needs to add routine perm descriptors that allow of EXECUTE privilege to system routines that by default have execute access +that tracing a minor compaction could put enough entries in the trace table to cause a minor compaction + +We just have to embed the appropriate jar in our jar +Cant ship with a potential leak +have not looked at test yet +Ive added it as a dependency and will give it a quick test without it +Also writes mvcc +You must start the GWT development mode with mvn gwtrun +Hmm strange I looked at the manifest a couple of weeks ago and its full of stuff +Ill try to get another license key next week +unit s ok for me too on +This issue was identified as a duplicate of but remains OPEN and MAJOR +Could this be a VM issue +The problem also applies to sitedeploy +Ok +Committed +It is much better to see the newsHowever I think Antoine is right +mduerig yes you are right +Another item to add to this list is the initialization that takes place when you start an agents process +workarounds for the back compat test Changes for this one is kind of complicated +check in and the qpid +This patch calculates the stores the previous state and calculates the delta for requests +Changing that may not be so easy due to following reason +Does fix the problem +what do you mean run the revenge table per table +What exactly about that behavior is causing problems? Why does the proxy try to in the fourth request? Also is this a problem if preemptive authentication is used? patch should take care of the problem for +Well give the patch a try +This would get them in there for Beta and maybe out for CRFinal +I think the problem is in but the Valves that are called now by show a problem that existed for a long time +Lapog The error you reported is the one we can see in the UI +See for how to do this with logj implementing a version should be similarly feature is broken in Solr at least with certain files +I corrected this mistake in and it works fine +This solves that problem +The original bug is a duplicate of +Gavin youll have too look at that condition and decide how you want to handle it if you think it is a problem +Id want to dig deeper afterward though as a follow up JIRA +should be somewhat future proof but keep the amount of data per znode in the split children nodes fairly small +bq +Now I can see this feature is an ease of use enhancement +for the miller regression +this is awesome! thank you is broken +Ive committed this to trunk and merged to +Just a note Ant Code Generation is available from Export not from Hibernate Code Generation Configuration dialog as its shown on screenshot. +Works fine +I agree the logging level should be changedThe bug you hit is +Work on this issue has been suspended until the issue is resolved +If the community feels that it should be reverted in then I am fine with reverting it in as well +If I use the new consumer endpoint I will then see this sample project fail even on the that works with the old consumer endpoint +keep it open then +Would you mind also changing the description of this JIRA to more accurately reflect the increased scope +made it case class changed +I think you are right +Looking for similar bug I noticed that the same bug appears in and took web app from it +forgot bug still exists in the installer jar +Hi Tomasz +Additionally suspect the whole thing needs to work with the to be useful in the real world +Thanks for the patchFixed at ce +Thus it does not matter which reference you use they should all be the same +any idea? Is there any batch running to grab the keys? I had a RSA key instead of a DSA key +It worked with Guvnor up to a month ago not sure about the exact date as I tend not to swap out the war in our system on each build +We need an example of the mbean that is producing this exception +However I would argue that we provide a framework and do not make the final decision on what is an appropriate use of its capabilities +This makes sense! I hope Ill find time to implement that its so easy to do +Will take a look at pull request +Verified at r +Committed the patch to trunk with it. +forge directory and now it works +Unfortunately that is needed to mitigate some issues with the vfs classloader but ultimately it needs to be fixed at a larger scale which I assume would include rolling back out the changes in for a more robust solution that doesnt break classloading +It can be achieved in following way On set invoice status to cancel Accounting transaction entries associated to invoice should reverse +i dont think many people implement custom ones anyways +Could you write a quick test case? I just give it a simple go and for the case I used I didnt see the problem +The problem is that Smooks is overriding the context classloader with the one associated with its application context currently so has no visibility of anything accessed through the context classloader +A comment on my approach to the conversion I used the existing canons and converted them into corresponding Java result set tuples for the assertions to ensure the test semantics are retained +Weird +They prevent the codestyle alignment error +All the mock portal tests pass +But when the proxy that is injected in to the repository goes to retrieve the entity manager holder it cant find it due to the instance being a proxy +I created little test files named and +My hope is to have some basic sanity checks for compatibility which we can release with +Updated patch +My driver here was +i just want report this problem for future work your idea to make a patch looks interesting i think about it +theres just a bug there w HHattached fix to +Yeah that seems like a good idea! Ill give it a shot. +what is missing +Check out as an example +AFAIK the plan is to release around May changed it to the latest snapshot following the structure that you are using now +It was committed wo a reference to this issue or any equivalent +just committed this +Same as +Have the harness set these properties when running with a security manager using the values determined by the test utility +Can you share the code used to do the benchmarking by any chance +Committed the change in rev +Unfortunately needs a new API to support polling either +Did you run your tests against the release +Updated in revision of buildersjon buildersjon Commit message Add suffix to generated names +If all have been tried for a given txnid but none of them succeeded abort +with the patch it is in BBU and outside. +Are my assumptions correctfor the former yesfor the latter noimportXML has different sematics than a restore operation +Btw in the test case above the data is not really lost +I plan to commit this revised patch if the tests passIn the test there was a diff which this new patch addresses tooM javatestingorgapachederbyTestingfunctionTestsmasterAn error message changed because casts to BOOLEAN now pass the parser +Dates work modeler support is in so I am closing the taskTests run Failures Errors Skipped workcayenneframework grep l FAIL target +If you go to create a new Child for that parent error +I have considered implementing Evaluators as Logback does but I dont see the value over just using either a Filter or a Lookup. +For produce requests it is initialized on producer object creation +Leaving as resolved invalid. +not backward agree Max but Generics are really usefulWe should think hard on a way to allow them +resolve the issue by the easy way +It will be included in. +In the mean time please rebase v onto trunk and submit so that can do its thing +Protocol changes for adding metadataHow is the proposal to add I proposed in? The diffs also overlap with client publish changesSince the protocol changes for client publish is introduced for and why not put them together? Because the publish you changed could only work after and are inAs my previous comment I suggested you split you by features each feature is done as a jira +Then we can release remove the deprecation and update all +In order to get the source to level I just overwrote the plugin configuration of the maven compiler plugin for every componentMaybe an examples parent pom would be more appropriateRegards and thanks for reviewing looking there is a parent pom although it doesnt appear to be up to date +To my opinion the source code becomes a lot easier to understand if you make it explicit what the different modules are +I didnt hear anything back so Im calling this closed especially since I installed using the steps outlined here and was able to get the grid config going fine. +in r +No response assuming it is ok. +I do not think this failure it is related to this patch at all +from what I remember in maven dependency resolution happened before executing any plugin thats why anything that needs to be downloaded not just plugins will be downloaded befor starting to execute +No more feedback. +If all elements of your Set are from the same type then your approach is ok +Does it work with any other version +No response assuming ok. +Thank you guys +The transitions have always looked pretty good to me +This seems to fix I went through the changes again double checking all conditions +Thank you for your speedy response + +Patch added +Brent already tested this thanks man +for setting it to or for the graceful fence attempt +Code change occurred in svn revision +Please add your detailed comments if you wish to make the case for reopening it. +That would predictably restrict the rows returned work with procedural relational logic and be valid for use with source indexes +Define your UNIQUE constraints thereIf you want a trivial feature like this implement the annotation and patch Hibernate Annotations then submit the patch on this issueDowngrading +Patch applied at +page for details on caching of hashCodes +Sending to Hudson +Hmm looking at it again you are perfectly right +I have attached an image that will produce this image has a CMYK colorscheme that breaks every image tagfunction +target +Rather we had to use one of our test machines to reproduce it +Its already in the plugin but theres some bugs +This would get rid of all the garbage and I think deleting should not be too different than marking commit as failed in terms of performance +ll have to do it manually +Thanks for the fix +I set it to hour using reflection and the above configurationoverriding of classes I mentioned and that resolved all the problemsMaybe in addition to making it configurable we should also look at possibly using a different default setting +Should I commit +If the choice is using BC or Suns code thats an obvious choice so shortterm I think the fix is pretty simpleDoes using BC for this purpose assume youre using it as the JCE anyway +Core issue for from to just after merge of old branch into TRUNK. +Add code d prefer to avoid the name NIO introduces a new Path object to replace File so that could cause confusion. +batch transition resolvedfixed bugs to closedfixed +The patch can compile using + +Thanks Jakob +The schema name derived from +As suggested by Chris Pratt on the mailing list a good workaround is to add an emptyas a placeholderTested and it works fine +Yes +actionAt validators list link details points to in CVS +If we dont want to break anyone then we need extra explicit javadoc +Im definitely looking at replacing xerces with something like piccolo for parsing as this is undoubtedly the biggest bottleneck +Patch added +Shouldnt R on a table be enough to read schema and S for manipulating itI dont believe so +however there is the larger issue of usability +This involved a very large number of changes some of which changed from using commands to using the new Graph API +Thanks for catching thatI like your patch it fixes the issue just as well as my patch +Others are rendering fine with the same code +Thanks +Committed revision +Please include this patch as well +Can you code that +I need a couple of days +The build is very close to buildmagic +Attaching the source archive in order to donate the source code to Apache Sling. +comments on phabricator +Things seemed to work better then +Orginal source file in svn has many DOS characters which would caused a failure when appling this patch +thanks for reporting that back +Committed +thanks +Take care to use blank also and not tabs +removed comment about visible from tag doc. + +If no one objects Id like to commit this soon +Fix is checked into master but I dont have environment to verify +I think asm poms now have the correct dependencies so for next versions the findbugs poms can be smaller using the tested it today and it did not work +SHOULD This word or the adjective RECOMMENDED mean that there may exist valid reasons in particular circumstances to ignore a particular item but the full implications must be understood and carefully weighed before choosing a different course +The error messages refer to XCAS yet are coming from the XMI deserialization +I have summitted and it should meet all demands +Another thing to consider for should we silently skip any updated fields or should we somehow alert the userSilently skip as thats what the user is asking for right? But haiong saidthat if thats the case whats the difference between true true semantically +no the prototype has not been merged in the I was under the impression this was done and was not a s move it to since we have higher priority things to do. +I have just rerun the tests in IE and the results are the same as when run in firefox +Fixed +i havent benchmarked to see if this is any faster maybe even worsebut its no longer a linear did this cause Solr to slowdown? Did Solr previously have a more efficient impl and then they cutover to Lucenes +Is there anything now that prevents you from the server and modifying its launch configuration to add that argument yourself? I dont believe there is +Error to annotate a method with arguments with patch available +I think for starters since you guys have gotten FOR pretty nice we should just focus on that oneYeah I think we should do that +oaknamespaces. +Logged In YES useridSeams to me thats fixed +The previous test results remain valid +Or if persistent is the norm then we could reverse it and have a transient map that is not carried overThe persistenttransient map sounds like a good idea I will take a look at how that can be implemented +Is this considered blocked on ICU? Let me know if theres anything else I can contribute to this. + +Andreayes youve definitely got the spirit right Thanks a lotIf you could make one more tweak to it instead of passing the planner to the constructor of the implementations could you use the default instantiation strategy that Oleg has implemented for the other helper objectsIm about to wrap it up for today and will be offline tomorrow evening +Solr cutover to +Attaching patch for trunkCommitted test to about +Thanks anyway +Tuscany processes componentType files only from SCA contributions +Thanks for applying however it is a strict condition that Apache mirrors use the Apache httpd server not nginx unfortunatly +OK it looks like youre using an expressionIn your case Id have thought you could workaround by using a field constraint insert Fact Element and then select field type and the enum should simply followPlease confirm the problem is with expressions onlyThanksMike +when shall struts release +I dont think its important because it might happen even without this plugin if one replica goes down between two phases of the requestDo you have any thoughtsconclusions about this? +for my patch to compile you must change dependancy in for from to or +Just needs QE to sign off that is in Central in JBT is in Central in JBDS And that works sufficiently well w JBDS that it can remain there works sufficiently well w JBT that it can remain thereIf any of these last two turn out to be false reopen and I can remove the connector from Central +That was what the deprecated usedAs far as using all or extending the I wonder if there could be a cleaner way to do it if it were exposed better in hadoop to external projects +This was generated withgit diff trunk got it +Ill add more specific tests and create a matching patch +Could you reimport missing issues for? I noticed that more than werent importedThanks in advance +Is this necessary or can we ratchet down the visibility to protected? Whats the intended audience and stability annotations? Id rather not make this a supported class if we dont have to and it seems pretty internal to HDFS +That sounds like it could work +Committed at subversion revision +We have made good progresswith the maven artifacts since we released +Ive patched to capture the matched data inside the s a quick fix until this issue gets resolved +it that okV patch +Yeah this would be nice +User Workflow now is Click on GWT archetype in Central on the generated project and select Run As Maven Build +Ill be happy to test this with large datasets when i get back next week i think +This is something related to the jbosscmp customization I have used from the we use default files this works fine + to trunk +My error thensorry if I misunderstood how the code works +Tested the patch with the latest trunk +of the tests are not required or wanted could you please apply patch to remove +same problem +Uses proper exit codes for functionality with binnutch shortcuts using the suggested method +In the docs refer to force not reran all the unit tests for last night and could trace no failures to this change +Sets and Lists are supported in RC not appears to be no RC candidate for the are you talking about the RC release? +See for a request to implement a connector to other CMIS systems +Attached a patch that replaces the base class with static Log instances +Reattaching same patch to retrigger Jenkins +The change went as revision in and as revision in mainCommit commentsWe were incorrectly trying to set collation info on string types +Ive committed it +I have even set up the test so that it creates two separate tables now since theres a test that requires additional data to be addedWhen the said tests fail though I have confirmed that the sort is indeed still set to but the statistics just show that an internal sort was done rather than an external one +Yes sorry the test was run mistakenly from the wrong location +Ill do it in separate commits +The test case seems to work for me on both the X branch and trunk +We should only increment the counter when the final attempt fails and we are indeed going to deem this whole application as failed +a warning that this will be published in two days in the ebook +Found that my previous PR breaks opening the source in VPE +array and map are simpler we can work on that first before getting into struct +Once we move our communication to Netty we can use any serializers we want +Committed to trunk and x +UTCOverride Sorry didnt realize that Fixed +This should be now fixed +After making the change to I think it would be best to make the new interface as well +BryanDont know why the patch does not apply for you as I tried it on another source tree of mine and it still appliesPerhaps because it contains crlfsIll put up another patch sans rs and see if that applies for you +In addition to that I will check in a change that deletes corere welcome thank you all for verifying independently +zookeeper should be reserved for all the nodes +Would love to use forktrue but unfortunately the groovyc task seems to fail with long classpaths when forking so that doesnt work for our projectJust tried the latest snapshot and the error went away. +committed +Verified +This issue has been fixed and released as part of release +I feels redundant for us to provide an additional Validate menu if we can simply just be triggered when the normal validator of course make sure we dont run more often than before +Close the ticket +Has anyone besides Yip run derbyall on this patch yet? If it passes derbyall Id be happy to check it in +s initial size is set to number of replications for the file +The attached patch only adds field description text and does not make any other structural changes to the ofbiz code base +defer all X issues to + all resolved tickets from or older +To fix this we need to do the research on whether that bug has been fixed in newer versions and if so add it back to couchjsAlso JS tests should be added to test for this behaviour so that when people run the Futon tests they know whether their JS engine is properly sealing objects +You can reproduce this issue by + +The only XML required to deploy is itself +Ok Ive linked to a JBAS issue +Is there an actual patch for this or a Workaround? We see this same when attempting to start our nodes restarting the node usually works correctly however +Thanks Dave for contribute the patch +But after warning message also appeares an exception in error log +The test requires a running RM ask for the information +I get the error when running the with +The is deactivated right now but it was too slow anyway. +Please could you check that it has been carried out as expected +With the update the page goes back to View tab but not Attach tab +Simon Im going to get deletes working tests passing using maps in the RT branch then we can integrate +Its possible to do this through the API currently +Committed the implementation for this fix in the trunkThe next step is now to include the improvements made by Jago in. +Fixed on account for nested filters too +LeonThe ant task throws a lot of exceptions after applying the latest patch +No build failures using maven +And does that work using the RI +Once is fixed printing a good warning will be easier if using the default credentials. +all of it or just the minimum amount +Yes leave +Awesome! +Im excited that our defaults may no longer fill up our hard disks when stuff is awryThanks +Erlend for the new patch +Very cool! and important patch! Few comments +Can you please be a little more elaborate +This is why the folder on s old repo is not writable to prevent accidental releases +from constant was removed s +Verified in could you put here some example howto check this feature? Im not sure about what this feature should exactly provide. +If anyone watching this jira is still interested please start a discussion on . +Do you have a debug log from Ant to show the class loading + +Fixed up exception handling and removed in revision +Is your searchType set to substring +Santhosh BTW since the is not for Counter why having the reporter object in it +bq +will commit if tests pass +Will revert for in favor of on out and will be handled by in future versions +Replaced in r +Thanks for patch Jon. +Someone just needs to deploy the site +Seems to be a bit unusual to meIll try to debug this a little more today +Merged to with revision . +Closing as Not a problemThank you guys! +Attaching fix with regression test +But I feel its incorrect +Please verify and resolve +So close it as Not a problem. +However it goes deeper than that +After upgrading to a new problem occurs which has been raised in +Fixed with version dd Parent had I switched it to +Am I wrong? If I am correct then the timezone should not be a field but a schema attribute thats used by implementations when parsing and displaying values +We have that stresses the failover +pushed +Closing again. +This looks like a quorum peer was creting a new txn log file and was shutdown in the middle of that + FTP ones IIRC are for testing the Apache FTP server integration I did +Oh and for the read test you should explicitly disable minor we push this to +Theres not a JIRA for this yet so feel free to create one or I will later +svn ci m Fixing +Throughput Stddev Latency Stddev netty no netty So netty even gives a slight improvement +I have also a lot of tables and columns to be documentedIs there an roadmap when this feature will be available +are there any +Then you can use the extra stuff that Hibernate has which is not in JPA etc +Thanks anyway +Hi therewe changed the default behavoir of usersrepository in latest M to use virtualhosting by default +in that case it cannot reflect the name of the return value +Use FMPP? It is pretty nice and integrates well into mavenant builds +EJB services not part of TCK so pushing it to the fix to +Will try to attach a patch coming days depending the amount of time I can free for it +I had to add a few more quotes to get things to work correctly with spaces in directory names removed a spurious variable setting on the nohup line of and updated a comment +When the refreshQueues occurs the second time it creates new queue instances again since the old ones were discarded +The stack trace is not an so does not appear to be the same as the original reportAlso DBCP does not appear in the stack trace so this is probably a different errorCan you provide an example of the please +OK Ill try to diff what I had done before and those changesedit I also see bug fixes I had made that got lost +The initial bits are in place +This is an extended version of that supports GWT via resource path gwt +doesnt seem to be an official HTTP Header +This would allow developers to annotate once but be able to support both XML and JSON representations +With bean properties Spring tries to resolve them via a lets try our best approach but the initialization problem exists there too +If the replica is bad repair as how block scanner is now doing +So all issues not directly assigned to an AS release are being closed +Just double checking the picture you are referring to is from Cordova camera API + +It eventually seems to give up trying to provision it on that hypervisor and tries anotherI believe the VM image that is being used is located at usrsharevms +Could you please review and apply +Patch applies fine for me and tests all pass looks good thanks Gustavo! Ill open a JIRA for the other issues +Dan bumped to for now +Committed revision +Total nodes is the current number of active nodes rightYes now is the number of active nodes +Are we going to change JAVAOPTS during the platform build? Can this change apply to production config only +Wow this was a massive undertakingI removed the old workflow JSP webapp as part of r +There we display an empty string which in effect shows up as nothing +So the compression algorithm has to be configurableIf we compress the entire image file the challenge is to decide where to put the compression algorithm informationDhruba suggested to store this information in file VERSION +Change looks good to me +Add link to console +The reason that we use rather is we find that has out of memory problem when removing the attribute from the +Please remember to set the Component in future +JIRA wont let me close the issue though. +Have you tried? Im pretty sure this error doesnt happen anymore +an option to enable EPSV as default sounds fine to me +Alexander please close this bug if done +Its causing us some issues too +Applied branch and trunk. +This one has been fixed a couple of weeks ago as far as I closing all issues that have been in resolved state for more than a month without further comments +Patch updated. +If you could try it out sometime I would appreciate it +You will see the view simply refreshes and the save event isnt called as it was in the previous step based on the logs +This problem has been going for a long time and no response +ThanksSanka +look good to me +step from the last bytecode of the method which was called from native method +I considered briefly once keeping the REST scan cursor state in for transparent failover of scans upon REST process failure +Ive done a fair bit here but it could definitely be improved still +Please reopen if you are still having a problem so we can properly debug this issue. +Tim thank you I agree with your fixVerified on revision +Though this does not allow you to get a container if there is no capacity in the node +That can cause the backup to generate invalid position errors and shut down +probably setting the memtable to frozen while we hold the lock would fix this +Thanks LeoPatch applied to BEANS module at repo revision rPlease check that it was applied as you expected +Ive commited this example to cocoon blocks +lgtm on trying it +The chapters in the guide were rearranged +We are trying identify the cause for will take this issue as a separate thread +HiIs there a solution for this i can implement nowThanks +Make it optional per column family? Repair would still need to exist anyway so could fall back to that for cases like this +Kicking out of is an issue here but upon further investigation its not really a bugThe issue is around heavy concurrency high number of threads in +Did you you try to use the latest version? Did you ask on the User ML +Sorry for the confusion. +sorry +No referrals during bind are not supported +We could use jakarta commons logging which I think would still pick up the geotools logs +However if we empty then any client code that also extends would break +Added and. +Ive fixed the underlying issues with the scriptsWhen we branch for this coming Monday this fix will show up in. +Thanks +yes +That group should be able to read the submitted job files +In addition the SASL properties can now be configured for things like kerberos +The problem is due to reprocessing the block with the id by the column balancingcaused by the spanall block +maybe another way round will work +This will get rapidly harder to do once development resumes. +There is a short section on whats new thats worth to the package changes in Spring Security you cannot use Spring Security with Web Flow. +Adding an example to the JIRA description would be helpfulAlso the patch needs a unit test +update to alert to use socket endpoint +Committed revision . +Verified that the items are fixed in the v rc distribuition. +While I agree that making String sources cacheable by default might have been the wrong decision I dont like that map +FWIW sun +Wont be able to look at it today but will this week +Did a quick check unleashing wget on the workbench example +Sorry about that +I believe the documentation issue was solved in a different ticket. +People keep hitting this and are unhappy with the current behavior +A few comments line Is it intentional that fs is an alias for dfs? I think this differs from the sh version +Fixing that and adding some exception handling and logging +Thats running cat as the mapper on one file that fits in one block compared to cat on a local ext filesystem on the same machine +You probably say the script returns a class but there is no script just the class +Can you double check +Integrated v to trunkThanks for the patch DevarajThanks for the review Chunhui +OK shall I wait for a complete patch +See +Other packages are to be removedconsolidated +The RSS load test took minutes to run on my test machine +Jason any news on the release schedule with a fix for this issue +When trying to connect for second time it succeedes +Hi Dennis this should be fixed please try againsorry for the inconvenience +As per the comments on performance is a major requirement for any ACL implementation and I would therefore imagine direct JDBC would be preferred to access. +Davids comment suggests he reopened the issue so that when a future Apple or Tomcat fix becomes available we can resolve the issue and then close it +closing this issue then +A reply toJeff +Now that work is well underway with AS all previous community releases are +So we have a situation where all the reducers were complete but some maps are still running +Patch applied the following changes +OTHO Bertils examples could as well serve as a starting point for improved official examples and teasers +I moved the entry +I tried to apply Stephens patch to wagon trunk but it fails to apply +We now have get and get + +i dont think its worth introducing the minor performance penalty that all the other have to pay for this as we will not add it now +I committed this. +No other change can be applied to the current code. +Tech Preview and SOA Tooling are OPTIONAL sites not to be preloaded in JBDS to Max to confirm we dont want SOA Tooling site preloaded in JBDS +You are not suggesting another configurable parameter isnt it +And this is also an argument for kv stores without column families +Thanks Flavio. +Of course Flavio had the right answer in the first commentNow that is settled I have one more question out of curiosity +Diff of of Cajathe content must be copy to featuresbin or change the location on CAJOLINGTOOLSPATH +The answer to Dougs question is Not necessarily +I have tested JBDS and + + supports CDI so will support Seam +Ive just reindexed and the issues now appear with their correct statuses +Will commit if tests pass +When appender is triggered it only updates its counts field +if its type is Disk thats the expected result because snapshot does not contain memory +Bulk close for Solr +Since under concurrent relocations we have no guarantee of the order well receive messages in it was possible for the loser to send a normal first causing generation comparison which could ultimately make the token disappear from that host until it restarted +You can either select the builder or the builder +What data from the POM should be used though? I cant think any elements which fit this requirement +This was only introduced in and never reliably worked there either +svn ci m Applied modified patch for Nestor Boscans static method variants as reported in +can you try it out +hi jukka i would prefer a patch that is backwards compatible but dealing with missing +agreed scheduled for direcly appliciable to Forge +See +agree with you about removing methods from proxy interfaces +start from played around for a while last week with removing the containing folder from the deployment scanner and then deleting the project but this had pretty bad results +Also I didnt test the patch yetMy patch can be combined with Jukkas patch of course the advantage is the improved backward compatibility +Feel free to share the solution you came up with here for the benefit of anyone doing similar searches in the! +The fact that there are multiple znodes in the chroot is inconsequential +Did you try it? Does it work fine +I did try against no community schema stuff +Your call +Defer all open issues to +Above issue from to cols appears to only happen with caching code in place +Turns out some unit tests are failing +But if we are sure the +Thank you for your response Ive just added in file and it was working +This was fixed in. +Marking as critical +Committed revision +This is something I have already fixed locally I will commit it soon so it will be contained in the next release +Hadrian you are tight that about feature should only be offered for one single route +There is a new parsing hook via an override of the SQL that will be added to Teiid +Ill commit the patch momentarily +We are not going to add or support a CXF subsystem although there is nothing stopping anyone from writing a rd party one +My hypothesis is that RAM fragmentation is slowing down mallocfree +Thank you Michael. +I just committed this +Hijust checking this issue and I see only one entry attaching screenshot +especially since many Cloud users will be migrating from the world of masterslaveWe also need to be careful not to refer to leader and replicas which implies that a leader is not a replica +Fortunately fsimage does not depend on precision +With a union the restriction permits optimizations by not allowing two arrays in a union a writer using runtime types need only check whether an object is an array and does not have to check its elements similarly for maps +hey this OLDSNAPSHOT is no longer used +I am a little reluctant to add this feature since the name is only temporary lives under target dir its real name is fixed at maven repo +Recommend you try and create a test case so this can be validated + +Committed at subversion revision Touches the following fileM +Removing it from page code fixes the problem +Feel free to reassign this to a specific release but only if you have the time to do it +Verified by Vladimir. +Boriscould you please be more descriptive? What improvements have you made to Control? What part of missing api have you implemented? Or youve implemented all the package +However IMHO sometimes package properties are the best choice and I have to use a prefix for keys +If you look at the generated SOAP request when you try testing it youll see nco as the namespace prefix which is obviously not correct +Wont fix this for now. +is not well documented and is implemented differently by different file systems +Move minor issue to a future release +Especially because it is just a single line change that should not really impact the performance as it is only being called from something that is holding a bigger lock. +hasNext wont seek any more than a clunkier manual interfaceI guess it is actually the lazy object that might cause the problem I describedAlright +As you may have seen I have included a short description of the new property in srcmapred file +Verified by Nellya. +to Daisy review +Even the number of databases might be configurableHowever all the database resources should be cleaned just after the test finishes +trivial conflicts on due to whitespace and trunk having a few more of the special callIds moved into +Attached the patches for both andPlease review and commit them +Then I wanted to adapt the test case but I couldnt figure out how to test for a specific statement +Why would you expect an exception? Null is a legal value to bind isnt it +Well commit them to if will be fixed before release +This seems to be fixed in version of xalan. +Thanks to both of you for providing and testing the patches +retags some testsDoes not modify executed code +This is the copy of the database where the query should be run to see the performance db name is TPCDB user name TPC pwd whatever +Need to inject not +I didnt get an exception +I committed patch and Ill implement new issues as soon as possible. +Persistence is not done yet +Anyway thanks for your time it is ok now and I will close the issue. +They are because those are the default launch arguments +This issue has been fixed and released as part of release +Upstream issue in foundation linked +Also this is requiring changes to a core functionality of HOD +Verified on EAP ER CXF is usedorgjbosswsnative doesnt contain fastinfoset dependecy +Move to beta +Is private or public? We should add the appropriate annotationI think this is a public API +It looks to me that the issue might also exist for common application deployment +Applied patch with thanks to Nick also merged the patch into and added a unit test for it +has been marked as a duplicate of this bug +Added guard against by not calling the method if the target object is null in Rev +Yes you are correct +Just committed this change to trunk and branch +Thank you Ivan +Youll be much more likely to get help on the users list +Baffled +Replacement for the originally submitted patch +Im not especially interested in maintaining this navigator and indeed I think the ultimate plan is to remove all navigators except DOM from the core distribution. +which should be fine as long as we make sure the reconstruction log is replayed second time around if the current RS goes down +In the case the project is not configured locally and is relying on the global configuration having a entry to configure the order locally has no sense +maybe you could scan through this one too +Therefore I cant use the with an alternative atm as a factory xml system has many components and we are placing the spring config in a subfolder from the root application folder +Here is a test case that reproduces the problem +The list command check may be a scalability problem for the namenode with many collectorsagents +It seems theyre only really needed to keep the size of the CL header smaller + + +The test failures are unrelated because I only change the test case and not the core code or any of the flaking hadoopqa bot tests +Lucene infostream log file +I was thinking something along the lines of adding a tooltip to each value field as well that would display the valid ranges of values +missed first two lines in the think I am missing something here +The fix for this jira is already done but not yet merged to as master +We either need to implement them or have the reject them +Ill cry uncle +Please let me know your comments +but if is so I am sure there is a reason +simple test client +Just cant right now. +GaoThis problem occured in testcase +Ive not used because A Im targeting web browsers that all support natively and B Ive been informed that theres some sort of problem with the in Cloud Foundry +Problem still persist +Unfortunately this is happening on a production server so in the interest of getting it online asap weve switched back to compaction for the scrubWe may have other nodes affected as well and I may be able to try on one of them once this node has been fixed up +In order for the children modules to use the same property they must activate the same profile as their source folder indicates the artifactId declaring the profile multiple definitions if several projects declare the same profile id if the profile is defined in undefined if the profile is manually set in Projet properties Maven Active Maven Profiles and no matching profile has been found in the defined profiles +forcontributor +the webdav implementation represents a on top of a jsr repository present and not on top of jackrabbitand according to the specification its up to the jsr implementation if locks can timeout and when they timeout +I added the patch as its probably most useful for illustration purposes than for an final solution but maybe its useful for somebodyNeedless to say wed love to see official support for searches in +InterestingThanks. +Help Check for updates works fine +This is probably due to the fact that we are only checking for PENDING and INPROGRESS tasksWe should also check for QUEUED +Since jbds feature is included in richfaces component runtime cannot be included in product feature because it means cyclic dependency between components +Affected to we wont release a has been releasedMoving all related issues to the next version +Thanks Jon for reviewing it +Will provide a separate patch documentation on this featureThanks the patch +Well the problem with option is Axis have no idea about the behavior of the handler +When I call remove on an uninitialized map the object is not removed +Thanks + +Is there any additional change request for this patch other than the latest set of requests from Edward +I have completed my zh translation +It would definitely fit better in +This has several advantages The location of the text is determined by the template not the page editor +I may actually have noticed something about validation not called at workbench startup even when theres an existing project with errors in it +Attach a patch + +This issue has been resolved in stream because the solution is not that simple to apply +push to +Done the only notices are for the Oxford donation two SunOracle jars and the upstream Jetty. +Phillip has this been resolved for EAP CP ? If not can you please move this to the next CP +Applied the metastore upgrade scripts to the respective metastore +I pretty sure this is a duplicate bug and in that report there is test case +Attaching +Instead set it to something greater than your greatest term like maybe ZZZZZZZZZZZZZZ +that will make some very candy situation where when the destroy may cancel the flushinghere is the patch which will fix make destroy after the log file checking increase the time wait for flushing from s to s +One tiny nit the added Boolean variable definersRights in appears to be unused +If you agree I would like to participate in that work +I thought I saw this myself a couple times but when I tried to writea test program to reproduce this problem it seemed to start working againDo you have a test program which reproduces the problem? If so can you attach it? That would be very helpful thanks +No additional surprises beyond the headers mentioned above +If someone can review to make sure I didnt forget any permissions and take out the TODO where not required +I will fix it and add a new patch for the +Can do after patch goes in +Another question to follow immediately what target release? or? Since has already been officially released is supposed to be in the state of code freeze by now barring serious bugs only +Also schema sample could use a NTS example but not necessary +Making sure all Ps are at least marked Critical to make them standout compared to P and Ps +I will try to isolate it into a simple test case which then can be used for filing a bug against IBM JDK +I dont know how to make a reliable repro for this issue on a clean trunk but the attached adds some asserts that check if slot and slot are uninitialized +Mark +For IS and other tools besides the standard update site it would be nice if they were distributed in the same place with JBDS +IRC Brandon explained the changes in flowcontrol which will not allow us to stream data from other versions v removes those changes from m confused how does from differ from in this patch? It seems like a +Nick +Please try deleting your +Resolving this as a duplicate of. +Ok Ill dig later on +sorry for the latencyThis is not an error just a warning +Hi NamitSure I will change and upload the new patchThanks and RegardsSunil KumarExtn +Since connection factories are costly we can only have one instance of then +Having extend from seems like the best option to me +otherwise Id get little or no sleepIll fix this later since it really isnt a big dealJames +I think this request in its current form isnt valid anymore +The latest patch needs some rebasingtrunk tyu frejr tyu staff Jan +Ups wasnt supposed to be linked sorry +Itd be a nice step towards abstracting away from the file being the only way to configure a +I unassigned myself from this previously because if anything is done on it it will be done by the web subsystem developers +Trivial patch +interestingThis seems like it would have advantages in how we manage temporary storage and such +Hi Ted do you want me to make any changes to the patch? I am not clear on what the next steps are +Also not sure of how APR handles bit either +In addition I added a test treeimages to the +Lets remove the new one and encourage people to use the existing one +Thanks +If you can please provide the exactly requests being made via firebug +You have a few options the say I see it Directory contents change triggers an extension module scan Server triggers an extension module scanExtension Module Scan Server performs a scan on the directory to get a list of extension names The list is then compared to already loaded extension modules If directory list contains an entry that does not exist in DB extension modules load the missing extension from the directory If directory list contains an entry that already exists in the DB extension modules use a diff scheme to determine if the extension should be replacedThe diff scheme could be a simple file datetime compared to extension module modify datetime +Updated patch with a minor fix to rethrow an exception +Nothing else +Thanks +Stay tuned +Still on hold +Patch that uses the same property value for both searching and creating a configuration +Hello EmmanuelAre you working on this one? If you need help let me started few weeks ago but it wasnt a priority for ll be available in and if I need help I send you a mail for contribute on it or testcomment it +It may just be that the geometryless tests are not closing the readers used for the tests? Its a pretty common issue I introduced the above warning messages way after the tests were written +One other note the only reason you need these is because you are depending on +Yes in a YARN cluster the majority of the IO on a node for should be in the directories +splash says and About box says this +got NPE +If anyone wants to give this a stab for welcome to pick it up +I searched and couldnt find the issue anywhere in s issues +Also changed Client AM to use about the newly added files +Corrected the newline problem +I cannot reproduce this issue on WTP dynamic project with JSF and AS libsIt seems that some validator is invoked indirectly by or by some our builder which is not invoked outside of Tools projects for example while facelet core library is being loaded byThe issue needs further investigation +Fixed in r +Patch from Donald +Duh it was actually good +After some experimentation it appears that removing from the classpath circumvented the problem +If isnt out yet why should this deprecated method remain in? Why not schedule this fix for with deprecation and then remove it for +The test case is not working +Or push this out to and well do cleanup AFTER the GA release +New C keywords such as dynamic need to be colored in the source seems to work just fine except gendarme that show some false positives when using keyword dynamic +The fix is working on fedora but not on Ubuntu the difference is installed zip tool +This has similar requirment to for prefetch has the same result +Verified in dist built from SVN +Ah darn just saw your comment +Lets test this again in the next release candidate +I had tried and gotten the same thing +I applied it with some minor changes +Patch removing the unnecessary comment +So basic question +So I am going to close the ticket +Verified on latest build and it works fine. +Fixed with. +Forgot to mention an important detail from the same Ubuntu server running the deploy command from Maven works fine +Fixed in RC +jar think the general feeling from the discussion is that this wont be added to the core +We could either be setting the currency before the shopping list is saved or otherwise we can default the new carts currency to match the product store currency if nothing has been saved with the shopping list +and till we learn more. +I am running tests nowTouches the following filesM javaengineorgapachederbyiapisqldependM javaengineorgapachederbyiapisqldictionaryM javaengineorgapachederbyiapiservicescontextM javaengineorgapachederbyiapiservicescontextM javaengineorgapachederbyiapiservicesinfoM javaengineorgapachederbyiapiservicescacheM javaengineorgapachederbyiapiservicesloaderM javaengineorgapachederbyiapiservicesloaderM javaengineorgapachederbyiapiservicesioM javaengineorgapachederbyiapiservicesioM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapitypesM javaengineorgapachederbyiapistoreaccessM javaengineorgapachederbyiapistoreaccessM javaengineorgapachederbyiapistorerawM javaengineorgapachederbyiapiutilM javaengineorgapachederbyiapiutilM javaengineorgapachederbyimplservicesmonitorM javaengineorgapachederbyimplservicesmonitorM javaengineorgapachederbyimplservicesmonitorM javaengineorgapachederbyimplservicesmonitorM javaengineorgapachederbyimplservicesmonitorM javaengineorgapachederbyimplstoreaccess +Introduce new message with a new state and use older state only during soft upgrade runProblemThere will be impact of existing application if they are relying on sql state to detect the cause of failure in creating primary key +Released as a part of release. + +Can you please rebase? ThanksGiven this is not an issue with should we still commit this? Im happy to but not sure its useful +What you want is a mechanism to see the service state of Namenode so JT can go into safe mode +Thanks for that have the test case and it does reproduce the issue now have attached an update +Already implemented for +I committed it to +bq +Ive attached an updated patch which addresses this issue +Perhaps some more people will speak up on this issue during the night +Were an IBM partner so we escalated this issue up the IBM support chain until we got feedback from the Transaction Architect +In the line of the unit test there is a spelling error of subject +What happens to python and ruby clients? Will this be a option +Looks good. +Bringing into + +Now it works with both +I justed tested with and the issue seems to be resolved thanks +guess not + occured whenopen VPEclose eclipseopen eclipseassert VPE should be NPE isnt related to this issue +? Is there a way to make this work or any other ways to do this? solution is too heavy on the namenode so its suitable only for very low message volumes +Pavel I agree about +Sorry I didnt had the time to look into this issue yetBut it will get fixed before the final release +classes are envolved and for the patch! Applied in revision . +bq +Andrew can you review please? Thanks +It still has the same delay with under IE with of a bulk update all resolutions changed to done please review history to original resolution type +I dont my bad anyway I should have test thatThanks. +As he said mostly harmless stuff around checkstyle fixes. +Fixed +Thanks for reporting this issue and for the reviews. +Had missed in my previous patch also fixed a javadoc warning +So I would suggest that the leak should be fixed for not rc have some memory leak problem fixed in rc +This may be interesting for very very ugly very slow but simple and conservative conversion of to the new API +Thanks Raghu! +That shouldnt be included in nightly tests because it could be pretty heavy and long +Guys any updates on the patch? +Rebased on current branch +Cleaning the patch so that only has the required change and not the accidental reformatting +It looks like the intent is to only run Sonar against projects where artifacts have a certain naming convention +Ideally I would like for it to be a CMS site +When I stop the Windows service it waits for my Camel route to finish first +is now chapter of the user guide +So I think its a common improvement doesnt requires bump of Hudson version so can be simply scheduled on next version +Zheng and all have developed something similar at FB but that is just not open source able as that is intertwined with the FB site code and technologies +These particular tests run with a very small set of permissions. +See where Robert Muir advocates using an ICU collation filter instead of accent stripping +patch contained in solves the confirm solve my vote for your issue and i close this jirathanks a by patch +BTW the question remains about the possibility of using import and export in +But this means that you always have to expect such exceptions before the data you desire has even been read. +So the standalone Tomcat clustering has no dependency on AS except the some common code base and packaging details +it still held we need a test case for this one that shows the problem +Thats codegen not core +The request reader still uses the old style kvp reader but it delegates to the kvp reader which uses the new style +Had to update to fix an assertion for the new minimum +Thanks for the linkIn eclipse there is an would that solve your problemAlways depends on what it returns in which situation +There is no real way to start a remote server +The reason is that this file is now part of the web site and were relying on its name for other plugins to link to said as I knew some persons were using this property Ive still left the option to override m not too happy with this but I havent found a better solution yet +write into temporary file and save into Jobconf mapping temporary file real file if this an absolute path +Alexei Zakharov helped me to understand that the test is obviously related to the enabling of it crashes only on DEBUG DRLVM builds and passes RELEASE builds +Thanks tried to access this URL and after authenticating myself using my Google Account I get ForbiddenYour client does not have permission to get URL p from this server +What is the runtime detection basing the configuration onBefore it generated a +Currently I dont see another way +The patch wont patch as it complains about being malformed +The original approach where we wrote to the local log in process was pretty fool proof +I have looked but can find no such standard anywhere +If the difference is noticeable would you mind if we only used the Reader when the charset is actually present +Quickstart start the application open localhosturlwithequalssignparamval a fix for this +Ok Ive committed this fix and the simplest possible test that just confirms the following The main thread gets s classloader as its context classloader Threads spawned in Ruby code get s classloader as their context classloadersTesting that external threads dont get a context classloader is not easily possible from within a Ruby script since Java threads spawned by also naturally inherit s context classloader +The primary reason we are avoiding the patch above is to prevent cluster wide locking +I did not set those parameters and just picked up fromDBCPWill try the test with and no connection pooling and see if that makes a differenceIm curious why you run without connection pooling +But the widget should display this differential period for information +More generally the started documentations are not up to date tooAs soon as the build enhancements are completed I will update all documentations +To keep reference to valid stream have reference to resultset and blobclob also +Ill have something together in the next day or twobq +So im still in progress of customizing Eclipse to Apache standards +Certainly in this context it shouldnt matter +After having replaced failed disks or added new disks this patch can be used to them without restart datanodeThanksWang update document for this patch with updated trunk code and included test may remove failed volumes this patch enables add the volumes back after having disk replaced +Mention aspects of security configuration in documentation for archetype release notes +It looks right to me +But I think we should have a deterministic way of getting a new directory as opposed to a random method +This issue is closed now +Opening for banner +It echos the parameters blindly +Carston has done some work in this space +This kinda sucks because effectively we just instantly broke every existing file +Working on backporting this jira +But not sure how I can test it at the moment as I have no clue on how to uninstall that SDK +Fixed with r +The code is available in trunk and is enabled by default in the MVC namespace +Maybe someone else can give a hand wit hthat +Lets open separate issues for tracking defects and further improvements +The ATP is decreased but is the inventory item considered reserved? For example if you create a sales order for the same item I think the system would reserve the inventory item which should be transferred against the sales order +The sentence detector trainer needs to retrieve all sentences from an input CAS +The same issue occurs +Rama lets discuss how we can come up with a simple unit test +Passed all unit tests locally +While all privileges is lumped into one userrole the system administrator privileges can be specified one by one +rUbuntu smoke kernel test as specified in JIRA +Thanks Anis + +Another patch +I proposed this for the purpose of testing For publishing to staging repo we dont have to go for a voting +I mostly implemented this before realizing we had fish of varying sizes to fry attaching the work so far +Absolutely agree we should document semantics +Integrated to and trunkThanks for the patch Jie +Thats what was created for +Issue needs to be resolved at application side. +As per the referenced LGPL is precluded in a full dependency + +I feel that it might be better if we checkpoint more frequently than we expunge +already fixed +now returns null for large allocs and doesnt clone +Thanks for the patch Simon +Comment added during attchment upload remving not to cause confusion +i told them to open this +It reports problems with further dependenciesIm using VC express on Windows XPCan you add a README at langdotnet that explains how to setup for building the projectThanks Thiru Sorry about the missing dependencies +Please test now that is closed +Hi PeterI have the same problem for BC +Added test case to test invoking of Synapse Library functions +However when we went down this path before it seemed to cascade into larger changes +Yes you need to have a JSF project to use JSF Palette now +The WTP bugzilla is +Fixed in trunk +Fix checked in. +Let me know if you need more info. +Had a chat with the AMQ people +Need to update the implementation given the writer interface change +Proper fix required for +KoushikMay I know why cannot the User deploy a VM with local storage service offering with root disk on the Local storage and Data disks on the shared Storage +The Logging looks fine and works fine and since we are now handling the cleanshutdown of the jBPM background processing I can no longer get this error to occur. +sounds better +Merged changes into branch in revision +So yes a fragment bundle can be used to work around this issue it feels a bit clumsy though +compilePp in dumpSchema makes couple of tests fail returning empty schema +As you suggested the solr conf can live at the instanceDir wherever it is configured in the attribute +Patch looks good andrea +Now as a performance optimization we have stackless exception in case the method is not found +Their result sets will not get cleaned up until the connection ends because no CLSQRY will be sentThe reason for the change was to take the autocommit out of finalize and with that I fully agree but I wonder why do we have to take out the CLSQRY? write takes a boolean allowAutocommits +Committed for at revision +This looks fine to me but would appreciate a review before committing +Alessio Simone do you know +Thanks for the patch Olivier +I always thought this had a measurable win for apps like Rails which is a heavy module consumer? Like I said I believe you but this description is sparse +This has been fixed in JRE +And fiddling around with patches manually is not one of my preferred activities +thanks heshan for the contributionRegardsLahiru +What do you think +Investigate functionality to start container only start selenium server parameter to test demo from specified duplicated ftest add possibility to start container independently on t be is bad approach to scheduling issues closing as partially completed. +of fix has been applied to. +Thanks! +Regarding the excludeinclude of the props file in the POM is to do copy with filtering + for the patch +Reproducible under myfaces at +Created an class and junit for it +Thisll probably be bestJason I suggest you create a separate issue something like Integrate in Realtime Branch and I will take care of it + patch forSorry should have pointed out that the issue isnt that the proxied pojo must be serializable all but the direct test fail without the patch the reason for the references to classes etc were just my attempt to figure out what was going on +Hmmm thanks for the update on that +Going to fix it in releaseDoes everyone agree with my right type mapping suggestions +Ill try to reproduce Alexanders usecase +Will grab the branch +Thanks Indika +I think this JIRA needs to be on hold until I can get the other one working properly +Thanks ChiragYour patch is in trunk at r at r at r. +to upload a patch for a new version of the scripts to publish Derby test resultsThe new version goes into toolstestingreportingThe main difference since the previous version is the addition ofa simple failure analysis and the possibility to create plots of testsuite durationusing gnuplotThe patch contains README files explaining how to use the scripts +operator but doesnt work on the first root itentifier +Patch coming shortly based off of Suneels original patch +Thanks old resolved issues +ledger return OK with a null ledger handle rather than usingBesides that provide two simple test cases in rather than in +However this isnt a show stopper for me personally is a nice way of cutting down the config +Similar to this will allow to reduce size of plugin approximately by Mb. +Added new file +Thank you for the review. +Absolutely not critical +Done +Any other takers? Otherwise I will commit it tomorrow +Just need to make sure all the correct logic is in place for all widgets before committingreleasing +Thanks! I applied this patch +Less stupid patch. +Sorry Adrian +Fixed a mask calculation related issue +Sorry about that +You can use jps and jstack to get those dumps +Thanks for the patch TomI put this into trunk and +I was not able to repro however and secondary efforts to insert whitespace caused no problems. +meaning the features could diverge before ever being released +I just upgraded Nutch to Hadoop incorporating this patch +How will this help us in case a Kill happens when we are copying the incomplete chukwa file to the newly created +About the only thing to do would be to figure out how to gracefully quit mochiweb before calling initrestart +Found it is not checked into so filed. +BrilliantBut can I raise the question regarding static injection +Yeah the credential cache wasnt doing much for +Packages for this JIRA some of them may have built in individual for each package +Marking as resolved +Ill take care of this anyways + to store results minor at r + + +Committed to and trunk +I like this approach +links from inherited Java methods now supported +Interesting Ill add a null check Im surprised though though should be an exception above that or it seems difficult for this to happen +Can you try using the trunk +Exactly ideally into the extra lib folder of directly +checksums and licenses validate +Close all resolved issues for older releases. +Thanks Mike! +I do not recommend anybody to run these versions for mission critical workloads +Changed so that it logs this INFOURL +btw I didnt find a symbol which log url path with querystring before +Can you clarify what it is that the PMC needs to do other than release tarballs with version numbers +So Marcel I am to read your comment as a for putting back the ability to be able to replace the cache implementation + +This might make things simpler because I plan on moving it around if no one objects to the patch has details +we do however provide a hook so developers in these environments can implement a serialization process that works with their environment see +Integration tests were updated +Sorry for the late reply +Sergey I am using the following jarst reproduce the bug with provided pages +YMMV +This doesnt seem related to but it isnt obvious at first glance what is happening +I love JDBM local database +fixed it +I am sorry if that was taken in any form than to solve the problemHowever this Jira should stay open till it is fixed +Not a requirement of AS. +Closed after release. + +I got it working with cordova embedded webviewAs no phonegap code need to be changed just some code on my main activity Ill write it on the google seems the input file is broken on kitkat +See is a public class +Thanks a lot John youve been great help in testing and debugging this +As just suggested use SD JPA +If you finish it before we release beta please correct the version +This pull request Adds the domain security bits to the Switchyard EMF model Adds three new model validation tests Adds the domain security bits to the Domain page in the editor +Once the new jar is available for the build the method is available and all is well +Used test project and everything works fine +Interesting +Can you explain the use case for this change please +argh +Per spec the two policies should be merged together +I think this resolves the issue +showing errors is quite unspecific and not reallyhelpful +Please verify in EAP documentation on stage +Hi Richard thanks for this! Im going to open up a new issue and then link back to this one +Shall we add an unstable annotation for the sameAdded +Otherwise patch as Ravi Prakash for your advise +Already fixed in. +Patches updates on thatThanks for the patch +r changed attribute authenticationType to authenticationScheme +OK fixed ported to branch! +Based on +Committed to and +Thank you Bilginso now Jacques how should we proceedCan I commit the patch +I will +So why stick to the legacy in the default example +Ill try to read through it today +patch with a Robertthanks for opening this! I will take care. +This will still require a long timeout but that should be acceptable at phase zero the problem with a long timeout at prepare is that another participant might give up because of a timeout internal to its associated resource +However the artifact id does not have anything to do with the noticed a few other bundles with dashes is this a bug or do I have to rename my artifact +I will try the latest version +applied in revision Richard please check it was done as a lot MikhailThe fix looks goodRichard +Backported to the subset of which restores the original lock timeout on completion of and and see if that helps with the intermittent failure on codeline +I m thinking to jaxp stax related classes Source implementations activation related classes older issues for Apache since were no longer actively working on these at the moment. +This item is closed. +Enabled the test by introducing a direct endpoint before the +QA Verified for and plugins secured servers go green shortly after is terrific btw +Safari prompts several times for credentials when the console me know of you want me to also take a look I am off Thursday Friday so would need to look next week if not sooner +Justin do you have any direction about where the aliases should be resolved? In the wfs code or should the geotools stores handle them +Im running FOP on Windows XP SP and the batch file works just fine without the modifications you suggest +It can be at the root of the model project in the workspace +This appears to have been fixed in upstream +testcase attached +Attached patch by solving those three issues +JDK works fine and dandy +At the present time I dont see this as a requirement that the core API or runtime should support directly however +Should that fix be working inThanks version of have you defined in the plugin +This is patch +This patch includes documentation and unit tests solves one of the most voted for issues in Hibernate but still has not been applied +didnt see any usageevent generated for potable acquiredcommit id for generating acquire portable ip usage events was added in wrong place +The ticket you are commenting on is over a year old +Still have you ever used an Enumeration with Castor +and already do that +Thanks so much for submitting +Andrew what was the result of clearing the repo out +Ive already spent more hours than I care to trying to get Hadoop paths to work transparently across Windows and linux +It could be in Javas implementation +Logged In YES useridI assume this is simply a misunderstanding so I am closingthis bug +If you dont unzip it cannot see it +old editor was new smooks tools is focuse on smooks and user can create smooks config file too +I report your comment into it and closes this one +I method that would allow to clear this field without closing the factory would help in this case +However if we were to actually make this work for any router I think it would be too easy to abuse +Some engineering work is required per the note above who is the coremain engineer on this project +Fixed under r +added Stacktrace as attachment +was this actually implemented? well thats good +REQUIRESSUPPORTSMANDATORYNEVER! problems just start to arise when touching transaction address this with a subclass we need to figure out how to do suspendresume on without side effects +I can imagine that since the plugin is called emailserver it will fork a JVM and that the actual problem is there +Thanks Cheolsoo. +LSP Lucene spatial is in as a module actually as two modules spatialbase spatialstrategy +Is the address of your server +So all resource copying etc could be completed and the process may be launched in a suspended state ready to go +I will rewrite the patch if you think that adding a new class its too much +Perhaps it would be nice to add this to the hibernate documentation in Chapter +Committed patch to documentation trunk at revision +A project is created but the plugin stop to be executed due to the throwed thanks +lib +Thanks Vinod! +running thanks John! +Mapreduce patch looks correct to me but I am perhaps not the worlds foremost expert on Kerb +Jason reassigning to you +Running on my laptop +Sorry added in Git ID compile perfectly +Then sebastian fixed them and attached updated patch +I use connect jdbcderbyuseroeapppasswordassuredsysbootPasswordIve looked at too which one of my users incurred recently and am upgrading our DB for that reason +We shouldnt be having to do this if our accounting is correct +I have added guard against Committed revision According Upload complete converbng documentDo you have file you have uploaded in the file list of the roomAre building OM or using binary package +patch applied email sent to dev list no tests for as in and def +This is required in order to get link rewriting work properlyAffects module only +Unassigning +Hence having client side capability that can use is still needed +Thanks +Thanks the prepositional phrases look to my untrained eyes +Committed +Also i broke the layout lookup into a separate method for easier extending +Ill sleep on it and maybe some new ideas will occur +Yes sounds reasonable +I strongly agree would be really really really nice +PatchI did not go through all factories so maybe more need to be upgraded for matchVersion when switching to Lucene +Updated to latest trunk +The tldr in the section about metadata explains it the strict ordering of files by seqnum is no longer possible so if the default store picks up the files and sorts them by max seqnum it will have incorrect results for gets and potentially other issues +Project descriptor no longer going to be supported by Savara +redid the implementation in rr now using a workspace directive in the path entry +I merged the change out and reran the upgrade and metadata tests as a sanity check +These blockers and critical issues are resolved but not verified +Here is an updated prototype +Attached is the patch this starts a single thread that does progress reporting for the entire map phase +I think that should work +Think of the multi shapefile output case what fields are stored in one shapefile is of no concern for the next one +Updated the patch with a test that reproduces the issue without the fix and verifies that the fix fixes it +I added these settings +HI Jacques I have tested this in trunk and all works great! Many thanks +Applied fix +Fixed generated integration tests as Git commit bffbecaffaafbebcdcd. +This change seems appropriate to me +Closing resolved issues. +patch applied +Duplicated with. +Would be interested in Bens opinion firstWhat if we left this in and did your version in Jesse? Do they interfere +havent looked at the patch agree with the idea in principle but concerned about how we can make the syntax work safely +Added fixes required for Cpp Support not previously commited to master +removed the forced PIC for compilers +I think if a test takes seconds then it should be Nightly +We cant do because it doesnt allow optionsarguments with a in them +We will make the authorities and users protected instead of private +the WSRP screen has its own very odd issues on Java +Fixed in r +Its not a mandatory requirement but yes it is on the roadmap already so I think it would be a good idea to have this as a module eventually if proj lead agrees for this separation. +It is a maven project with the main class +Migrated the change into codeline + +An alternative is now agree that packages like need renaming to to make things less weird +you cannot link a +Attaching patch ported to branchTurning off Patch Available so Hudson doesnt try to run on patch +Regenerated against trunk +in gotchas section and FAQ not enough +I tried to create a simple test case but everything works in that +Please go ahead and make the changes you deem fit! Dont hesitate +Thus two general forms would be Thus class PSP class sessioncookiename class sessiondbm sessiondirectory class sessionfastcleanup sessionverifycleanup sessiongraceperiod sessioncleanuptimelimit sessiondirectory class Session session Documentation would state that any option of form modpython +Cmd or Ctrl works fine for me +There was discussion about JMX binding address on +This is an ugly hack required until the module cyclic dependency problem is resolved +Can we selectively turn it off for some deviations +Isnt it resolved already? Ive seen that JARsource JARJavadoc JAR was uploaded on th February +I agree with you +Implemetation changed from previous comments but issue is fixed using a new class which handles process output and logging and most importantly for this bug consumes all STDOUT and STDERR output from the launched processes. +We should be able to provide a first working version of this connector by the end of this week +It was aborting because of into +bq + +I dont see any other s try to sort this out for one way or the t we do both wrap checked exceptions internally in the stack an then unwrap and rethrow in the client API +We should explain why the benefit from having this feature justifies the overhead +We observed that this patch while reducing time for re execution of maps on failures is impacting performance negatively for normal runs on regular clusters +I do it more or have tested with trunk snapshot and with so I think it would affect every ones. +Please note that the patch supersedes the patch and includes support for specifying the classifier as part of the artifact string as well as using the classifier attribute +Attached memory usage hrs into test using Jason I have identified the minor leaks you have discovered from your h have fixed the thread pool reference and have pending fix for the backported I will a new SNAPSHOT build for you to give a test +Created to track not forcing test order +Okay I have a test using instad of and it works fine on trunk code +fixed with using scm +push out to +Java files that I changed in fource along with transition to closed remaining resolved bugs +So the mapping to the right sevlet doesnt work +The same patch goes well with branch also. +Upgraded Erlang and +It appears that it will be unlikely to get the creation of even the simplest stdbasicstream creation time down sufficiently to make the no logging case fast enough +Pass null from the client if no row lock was supplied or something like that +If all the copies are corrupt it may be necessary to use a hex editor on the edit log or a similar technique +Awesome! Thanks mate! +Braden also confirmed this was not previously working and that the patch fixed it +Attached fix patch with unit test and youre a committer +Reresolving +done upgraded to the latest plexus container have to uprev to in order to get the latest plexus +Yeah Ive been working on why our long running IT tests were failing +The latter is the option for the server bootstrap +seems like this patch will make Sandesha head diverge from Axis +Karl et alIf anyone is prepared to invest some time into a more systematic investigation of how common browsers behave with regards to cookie path attribute validation I would be more than happy to apply changes to the browser compatibility spec in based on the results of such analysis +As discussed this morning enterprise features repository could be the first part of the Karaf CaveKaraf Cave aim to be an OBREBR with extension for Karaf features supportWe can start by leveraging Felix and add Karaf enterprise features support. +John WilliamsSEO SpecialistCall E +Sure no problem +Basically Im suggesting that the grouping itself can be done separately by the Selection object since its completely generic it doesnt depend on which aggregate function youll execute. +basic cache implemented but only for the everyone principal +If you dont mind Id rather do both at the same time since thats what I tested +Thanks for fixing this! +just to be precise It should work for EL names which resolves to the same type that Seams standard messages component is +I should note that I have enabled Load and from projects classpath in Preferences Spring Beans Support seems that Spring Integration does not handle source attachements correctly as required by the tooling API +Tested all boundary conditions please go ahead +Thanks +Lets watch for it to be recorded +Did not dig in + +Marking invalid +was where the correlation id was added to the protocol +committed adding a finally block with revision +can we somehow make the regexp generation less evil +Now you understand my confusion + version of the multivalue double list widget +Replicated on JBDS with the attached bpmn files I am unable to reproduce this on Windows +Adding dev to CC for years without any objections I guess we can close the issue. +Fixed with revision +also +Verified able to create Autoscale policy within VPC +Hi Leventeit was a real pleasure meeting up with JNODE developers at FOSDEM and I have the utmost appreciation for everything that JNODE has achieved +Jean has provided me sample for this issue via email but the sample was quite complex to be able setup and reproduceJean could you please provide us simpler sample as I asked you in email +The file Ceclipseeclipseplugins does not existWhere can I get that jar +Fixes a concurrentmodificationexception +Thats definitely Forrest +The current parser stops reading when it reaches a startxrefEOF combination +Best ll resolve this as fixed then +NetThe calls do look straightforward so that is good +The attached test results was done in complete isolation to one anotherThat doesnt mean that cache drop is not required tho +Thanks Bruce +I will work on this commit today +First patch for vinays comments +XML translator can be used with WS or Text connection factories +Whats the problem with having an expanded WAR file running on a servlet container +Although I think the code is good theres no question that its very intricate and needs plenty of test time with large traffic before we should rely on it +Please Lus AlfredoUpdate before making your patch some Italians labels have been added and conflict with Jacques I think its ok now +This is the most voted bug in this project and it seems as if nobody cared +Basically the code retries the connect until the timeout has happened or success reestablish the interrupt flag on exit +Here is the upgraded patch according to your commentsI think Ashishs suggestion for giving patchs name seems superb +please provide me with a download link +Closing resolved issues for already released versions +Now that dependencies moved is anybody watching this issue or should i close it +to commit +Solve my problem +Thanks for the nice contribution Kevin. +Assigning this to Rick for as discussed on the list +It would be much more effective as part of the main build +It also changes the casting in the calculation of avg +Fixed a long time ago +I ignored the failure message of the patch commandand therefor I first thought everything was ok with the patchlittle bit confused now? but I think you can forget the branch when the the version is release soon! works with the new Avalon relaese wellAm I wrong here? youre right +This issue is fixed and released as part of release +Henk what snjezana showed will make it possible to locate the source of classes found while debugging based on the sourceclass path setup in the server launch configYour usecase of being able to access the classes based of the project classpath would not be affected by this meaning you would not be able to lookup the source of the internal classes unless you add it manually +While I have not had time to try out the sample that Lukasz provided above I have confirmed that this bug was introduced in because reverting back to makes it go away +Will checkin. +Much of this as well as potential solutions has already been described in +This would leave the tmp files in place +I am getting some errors when I try to kill a job that has of the mappers done +Oops apologies for the duplicate. +groupValue to check in on this issue have any commiters had a chance to look at the patch? What do I need to do to move this forward? Thanks +Tom another possible feature down the line +thanks for specifying the issue so clearly +Reopened to assign component +any chance you could generate an updated one +Dennis I think Florian is already working on it or am I missing something +Thanks +ill edit the issue summary to better reflect the pb +and approach is still confusing exec semantic +a new setter would need to indicate when to pay attention to Collation +No errors were logged either except on the guiIf I navigate back to the dashboard then all alerts are gone as expected so it really was deleted +Didnt enable it in any suite since it fails +I can submit a new patch after is commited +I just wanted to ensure that anyone stumbling onto this issue can see that theres a way to get the raw data +Does this patch fix the crasher problem in +passes ant +Works pretty well Julien but there is also an issue with the link on the total number of which is incorrect as soon as you start using the pagination mechanism +An example of what Im thinking would make sense from a backwards compatibility standpoint +A patch based on +No luck with +Ning is committed +OSNAMEs default value is Linux See the declaration of OSNAMEBut Yes your patch is much shorter +Everything timed out +Probably just before we produce jbossAS +So closing this as CANNOT REPRODUCE +jarExtracting systemorgapacheariesjmx systemorgapacheariesjmx +The patch deprecated listPath and globPath so it increases the number of javac warnings +Still an issue in Apache Flex changes to fixed this. +If no one else is working on this Ill take a stab at it +But without adding that extra set property definition +I originally thought that playbook lacked the functionality to record Audio but I played around with it some more and found a way +Ill delete duplicated empty lines in code samples +I see too many initiatives trying to talk at once and not listening and learning from each other +task is done. +The same link generation mechanism should be you could argue that since the action links are not showing up they are not broken so this issue is indeed resolved But I wont + +I stop my build right after the end of the build of my first project at build starting my build in my local repo I have Aug stopping my build my target directory Aug repo Aug repo Aug ll check to see why my artifact is not installed in my local repo + +Race condition between main and external checker thread +Source code used to state of my tree +But I think it is a good idea to move it to the callback just in case we have not covered scenarios when the check should be done +Where are we with this? Have you got a finalised proposal? Certainly bug squashing and javadoc fixes are good. +Because these are just a set of samples and there are lot of headings in the guide it just doesnt look elegant after applying the patch +Yes Im aware of this solution but we cannot use this because it means modification of hundreds views and of course we dont use only one uiparam per view but many more propagating variable in this manner means thousands of modifications in source code of views many of then are already deployed in productionI can do a global change from uiparam to cset but cset is unsupported inside decorate tag +I am OK with putting in a framework to filter the warnings more explicitly and on a per project basis +Assigning the correct priority +Patch against jbpm trunk by JBPM my comment in the forum post I linked to for how to reproduce in case this issue should arise again. +It should be pretty easy for me to extract this into a plugin and release it pretty quick +The attached patch contains a testing tool developed by the Yahoo Pig team to handle end to end testing +but there are some package cycles to fix before +I believe this issue can be marked as resolved +This is a little scary but there is a simple solution to the placement of the xml declaration specify the attributeyesin the xsloutput tag and put the xml declaration at the top of the prepended license file +have the same problem with the property named applicationThis problem has appeared since Struts version. +So the strength of the algorithm lies in this local processing without copying the whole data setI hope I have answered your doubt +It is a valid scenario to define an include that does not have an ID but does define internal since they may be used by internal script code +Be sure to think about when doing this +Per Andrews request here is my patchPardon me for possible white space issue +Really difficult to recreate +Fixed problem with failing tests +I did test this against and it has corrected the issue +Hi Amos is this closed? +The class which is used by has a method quoteArgument which could be used for this +what is the status? This bug has been untouched for months +putting on the backburner for +I think you can close this issue as resolved +The patch is attached +I dont understand this use the same proxy to access to the apache svn repo and no without proxy checkout for noise. +This should do the trick +Perhaps it makes sense to include the testing scripts in the test and jars +As we have a session for each connection we can add a counter to track such a number of renegociationWe can also add a flag that disable the renegociation on th SSL filter +The reason for us to get to wrong interpretations I believe have been due to situations like the following When using build a request processed for a topicqueue interaction would not complete +If you do an explain you will see filter still pushed up +And finally my debug logs are showing something! I still have to look deeper into whats causing this problem in after all the googling that I did +Here is what I have so far +Feature Validated +Unfortunately no time git pullCurrent branch trunk is up to datereal user sys I repeated it again to make sure it was cached with no change +resolving as fixed and resetting original owner. +And I recall they have done quite some work to get all the mail attachement stuff to work in Lotus Notes Outlook Thunderbird etc +Will look at that later +Discuss with Alan we agree to put get into +Let me know if it becomes too cumbersomebq +Java program that implements the schema and runs Derby +Any interest in generating these headers? Id like to see them in built jars. +Verified in JBDS B +Patch applied at r +The path does not represent an artifactI used a different PGP key to generate these bundles though so maybe that was the problem this timeAnyways any help would be very much appreciated +Pig has some fixes for memory leaksmanagement +Note that the is missing from your patch +I guess these interfaces should be in the project and Sindig impl elsewhere +Actually ignore that +As Andreas already mentioned there are many boxes +Michael Is there a reason for duplicating the Spring security files under Im on having copies of any file in the CVS +Agree WRITE is neededLooks good to me +Needs to work with importAll new issue for further tests. +I also migrated from lucene to Lucene +Otherwise it takes too many iterations +Also need to fix the test case issue with +now skips tombstoned rows properly wo leaving stubs in the new sstable +Would you please help to try this patch +Eclipse +This patch puts in exception handlers for the critical task job tracker threadsIt also gives names to a lot more of the threads +Oh no thank you! +Nice optimization +Just one last question Where do failure notices go? I would prefer that they go to the mailing list mailto +I agree that small steps are better and that the general Cache infrastructure does not depend nor block this issue +Patches applied at +will investigate because I actually didnt find why it was used. +I could probably make the change myself and send you a patch its pretty format do you like the patches in +probably not in the v the simple fix I was describing on the dev mailing list +Nobody is disputing what drools as a project does but this is an issue that needs to be discussedresolved at the platform level +Thanks for the update Alejandro +HelloI have a similar problem auditing a subclass which inherits from a abstract class and posted it on the envers discussions in which Adam pointed me to this JIRA issue +please make a sure of that +Kindly advice if the optimization is worthAlso this change has instantiated objects related to compression only if its a compressed fileInorder to ship the first line a readLine is called and this change would not create a new Text but use the available value for the method call +Hi It was a mistake on my part +Closing. +Thanks for the patches MartinHowever Ive only committed the changes +It is more important to spend time on the issues that we will support going forward such as the wise integration and the port assignments +Bjorn have you tried a realistic natural feature with a relatively large offset? Thats a situation where the raw offset curve generation will exhibit some really ugly artifacts. +CtrlF corrects such a mistake +there we have the opportunity to fix it right +Ill need to investigate first though. +Did the jar not exist? Was the representative class found in another jar? Applications can explicitly specify a jar file and that jar file WILL be used +I still have to check into why the site isnt being copied properly +REST will be the last thing we will implementSo far so good +could we just have a reference to in and check access based on the queue name there? is in package but is in package and looks like we can not add reference into +Resolved because subtask is resolved. +to and trunk +additional testing and inclusion in sample app are seperate jira issue for RC +Hey guys I havent seen much progress on this +no problems will commit if the tests +Andrew patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +Even if we are plitting in methods for and ones for we still have the problem that this would not compile on then +input +Im not sure I understand what you mean by pick up +patch for and patch for attached to exuecute patch build on hudson slaves +Hi KnutFair enough +Nonreproducible change to cp went out with Jon +Thanks! The other steps were only if the plugin did not work +Verified fixed with mrc distribution. +Assign to DevinDevin please fix the bug reported by tekni +It might be cause by inconsistent version of the plugin in repository +Check it out martin +We are able to use SSL via the gem without issue even though the SSLCERTDIR environment variable was unset +Check his hv fork if you are interested +bq +So is a duplicate of this JIRA +thanks dudes +is about upgrading to +Hmm good points +But it introduces some overhead into methods containing while loops since they must use variables +also is really promising as well and avoids thisConcur +The parsing code for config is in +Thanks Amareshwari +Good stuff +Would it be a big deal for you to rebuild it off the latest snapshot? OlegNo problem Ill regenerate it todayJon +We dont remove the server from list of servers until the processing has completed so as far as master is concerned there is still a RS out there and he wont go down till it does +Do you have any idea when the nightly build will be fixed? +We had a similar issue to this recently +See +Either Struts should store them and remove them or Struts should do neither +Instead it should read something likeCustom comparators should deal with the possibility of null values returned by bean properties or risk a The above is more to the point and implies that s default internal comparator deals with null bean property values which really ought to be the case especially in a library that is supposed to beBut then again after looking at the code does not deal with null values +Not sure why this issue was so Im closing it again. +the row in slot is a special row and not a normal key row +I would be ready to consider upgrade to domj +svnversion This patch addresses the relative path issue for base SSL cert paths +Verified that it is OK now. +Thanks so muchAnitha +Do you change that when pulling or should I change and commit it +I have verified that test no longer fails on the Monday after daylight savings time change +This also requires changing the panels that do add resources to contains the choice that was previously in the Rolando we agreed that the two commands that is add and remove selected should be on their own rationale is that the sentence composition in other languages may be different and thus wed have to resort to some more complex way of stating a full sentence with the two commands plus a way to determine where the add and remove links start and end in it +I spent some time but cant seem to figure anything out +This jira is unassigned but contains a patch that needs to be reviewed for inclusion in +I agree this issue can be rejected. +I no longer see tmptest after running this test +Patch I think its ready +Yes it can be a nuisance in the codeIn your unequal queries you seem to be generating the bNodesvariables at different points in the parse process because you end up with xy +There are few new files in there which looks like test logs +The attached attribute usage large +Hi Anty I have submitted a patch for to support a pluggable sort for MR +Thanks for the testcaseThe bug is fixed in trunk +synced to trunk +A fix is in SVN +I have too many uncommitted patches for this issue on my machine so I mixed them up +This also relates to the driverjob renaming proposed elsewhere +I think thats my change on +Better to clearly show that its not being handled. +For the implementation you should replace the sleep by a notify plus isClosed should be also be a simple boolean instead of an array tagged as volatile as its used between thread +However this workaround should make it possible to use felix when security is enabled and is used +RESOLVEDSee +Ive just committed this +Ok I do agree +Fix all the tests that are failing on native dbms like Mysql and DB +Thanks Bill! You used the right branch and my initial reaction is this looks good +will add links from sitemap and DRLVM navigatorcould somebody please verify that the latest README is packed with M? if yes we can close the issue +Adds configuration directive to to toggle this feature and added schema field numInlinks and fieldType int for storing the number of inlinks +I can rework it if neededWe run several instances of Solr with this patch for more than months now as we really needed fuzzy search withThanks +it wont be in the release timeframeIt is the default for JPA though +Added a system property which is documented in the admin guide to control the max string length setting +fixed and on the update fixed I mean I removed any ejb related code from mevenide +that sounds ok to me I might ask if any other infra members have an ve copied to the archives in the binary dir and linked to them in the same way as the rest of the jexl stuff in there. +Fixed in r +But my plugin was compiled against Maven but that shouldnt matter right +fixed in revision Mark please verify it fully resolves the good +If the users have it any other way then tools will fail but that is by into master resolvedI created for detectingspecifying of version +I verified that with this patch the unit tests take the same timeIn the process of committing the patch +found a bug the applied patch +Committed my patch +bulk close of resolved issues. +If needed I can commit something temporarily and then delete it afterwards +The problem is that we cant logout if we hide the adminbarIn WCM the adminbar is clearly not mandatory +For more robust error handling and recovery it may be good to strongly type these instead +Dont use Thin +But when focus is finally applied drawFocus should be called automatically +We should allow to provide localeversion query parametersThis is minor change. +Have you made sure to apply it before running your new tests? Either way I would appreciate if you could upload logs out of a run with the current patchThanks +damagecontrol +test package dependence +Ill create and close the Mac issue related to the new version of the codec +Still coding +Instantiates a single HBC in a static so we dont have to redo each time +pay attention +Theres a missing converter +failure from +Id been meaning to do this anyway +I reproduced it in all Client VM configurations from to mustang so its a previously unreported bug +css to update is plz give links to the updated page and resolve the issues if its done +TODO registration of node types is the only left task +bulk close of all resolved issues prior to release. +Attached couldnt get the patch to work on Mac +revision Also added code block with logic to test fields +I agree +I have a deadline of September for release notes so if I dont get the required information before then this issue will have to be excluded +on the part +We should just do +Hi RichardoI think this issue has been resolved in the RC or RC release +I believe that my latest patch addresses both performance for token auth and flexibility for mech negotiation +This change will require a release note as users my want to increase to accomodate that this value is now being used with BLOB and CLOB +Im going to work on this as part of a load balancer rewrite Im doing that is part of +I already removed the uidfoobar example configuration some time ago so I consider this issue fixed +And the current issue should be reproduced with Sun implementation +looks good +Updated to latest svn +Jeff will attach a new patch that makes the remaining changes that are required. +If you are getting your text from another application than you dont likely properly know you have latin stuff +I can see that the JDWP reply packet to the command contains nine entries in the table when there should only be five +With the above patch to only test case Collections ll try to find out whats the problem with this one problems with the old version was that it didnt create some of the tables required for test Ive posted the patch for maybe its better not to apply it +The above two comments were for +We do not use in this configuration for publishing results etc +There needs to be some minor changes in Tomcat for it to have any effect +Test failure is unrelated and failing for other tasksNo new tests are required for this change +You should always use varfoo if youre going to nest based on Graemes comments +Please make sure Xalan is on the classpath when using WSSJ +In that case we can add the retry limit in where is caught and retried forever +Removed from Affected VersionsNew features should be scheduled for minor releases +Just had a quick look at this issue actually this is defined in the war +Im on this approach +We should include this in the patch thats eventually committed for this issue +srctestorgapachecommonsconfigurationtestimport +when you are working with big company its very difficult to separate out repo team by team +This issue can be closed +Hey Taylor how does this relate to and? I think you are proposing something slightly different but I am not sure the exact relationship +failed because Configuration of it unexpectedly doesnt have setDouble +major +This affects tests only +I found another problemLink to in dita file is brokenI need to fix it too +This class isthen used by when the appropriate tar wasgiven +I have a fix for the test failure will commit today or tomorrow looking to see if there are more failures ATM +It would be good to aim to get this in the next release +It seems this really needs tuning with Java +Unsure if GUMP can build directly from Maven now though +Its a fix that will be corrected under. +Yes I checked out the code from the repository and build a snapshot build. +Unfortunately build log with reported error is not available anymore so Im not able to check used JRE. +Thank you Brandon for the review +Will do and and move the two classes intowhat is the reason that inherits while inheritsLet me figure out why and add some comment +It has this issue too related in Eduardo I will review the issue +There are many libraries which have favourableacceptable licensing terms which can be added to SolrThanks and Regards BruceI havent done any perf +Done + +Please confirm that Ive applied them correctly +Patch looks goodCould you please file test failures that dont have jiras yet +If trash is configured on the cluster pig will use it to store its intermediate results +New patch attached with jcrreferencePROPERTYNAME as the name +The only suggestion Id make would be to move the change type button so it lines up under the browse button +See screenshot +Hopefully will review and commit soon +and X have very similar low level patch for saslpasswd issue on trunk at r. +Ive darkened the grey to eee +Yes you are right +bq +They are just for method invocations not for properties +Yes the SECURITY schema comes from DB but the notes schema comes from a custom translator which does not support order by or the projection raising logic to handled aliased columns used in the order by +Duplicate +Is there any estimation for the release ofYou can trace the release plan via dev mailing list +Reopening issue +Can you give some more background +Ive just committed this +Thanks again +Thanks + +The originally code worked fine +You only need to change the interface of and impl of +It would likely work with future versions of Spring as well +Fixes event generation based on transitions +Some other we can removeI primarily relied on two pointers something is already deprecated in andor it has deprecated This will be removed in Lucene stanzaDont use the stanza if its not trueI also checked things required to support indexes and remember leaving some such code intact but could surely miss somethingPlease post corrected version of the patch or name exact things we should preserve +Uploading version to handle Knuts comments +Please let us know if there is any other information you are looking for to make this possible +The REST implementation does not comply with the wc spec +this one requires for its you mean +The reason is the antiCache parameter as we can see +The use case is chaining of jobs +Closing all issues that were resolved for. +That the code is stable is an argument against backporting this for two reasons + a patch to realize the resource bundle +Can you please do so +Thanks David +The SSE channel errors are misleading in this case +Ive just committed this +No as I said its not related to +If it times out seems like we cant access the output logs on hudson +Oops sorry reply should have been in Bug as thats where we are makingthe changes +attached factory with unit test Tommasopatch look good a few things I think the default should be to enable positionIncrements +Fixed +The funny side to it is that we spent a lot of time with Chris thinking how Jersey magically turns away xxxxxx with we thought initially Jersey blocked it even before dispatching but as it happens it was also passing it through I know! I got down the wrong rabbit hole thanks for the help both of you heh +If you pass a Mapping instance into you will again introduce a reference to a class which is specific to a mapping syntax which is what I diposed with my last refactoring of that you like to easy loading of mappings you have to search for another solution +However Iam keen to get an idea just how much stuff breaks if we do +Issue in PBR as well +give a try to the see comments above +Working on the fix +Each set of modules has an API module and a set of implementation modules +Based on posts to and this is a classpathconfiguration issue +Postponed to +If somebody written more especially one that shows schema metadata please share we could use while we test +Is the code available some where in the SVN? I have fixed the possible cause for this problem +Hi MarcIn that case why arent validated? They are either a header or the BodyColm. +HiId like to get all resolved defects verified before releasing the version of the bundleplugin so Id really appreciate it if you could check things are still ok with the current snapshot and if so close off the defectThanks in advance +There is currently no good way to interrogate the mapping closure in domain classes and pull out the default sort specifications +We may need to address this in the future +To implement this in Java we should be able to implement a and that wrap Jackson and instances +A generic file formatIn either case changes to the serialization framework seems like serious overkill +code was copied from type branch so dont need to in +IMO its better to failfast +yeah I was afraid youd say that +Sorry I didnt think about this being platform dependent +with the latest fixes for other issues it started to happen on +dita file to edit with jEdit almost all lines are displayed in underlined red with a message stating that the element type must be declared +the +And its less risk that you bring down Solr! I wonder if we should put something in the wiki that recommends that if you start having problems with Solr CELL then move to running Tika outside and maybe include some sample codeSolr Cell is an awesome feature but it can also cut you +Revision add convenience method to use it in leaving issue open to do the same with other test cases +are you going to share the code with us so we can consider putting it back into core +Yeah this can be resolved +The only new failure in the tests is which passes for me locally +bulk defer to +The problem is the fact that the editor is not even openedAdditionally describes a bug with the editor for prescriptiveACI under very specific conditions +Too early to start thinking of such feature +Hein next time please dont add Java features in your Groovy features should be working on when creating a new test you need to add a reference to it in one of our +Modified patch to include new tests +Plain ASCII results +Vikas. +Stupid me +end of output And the window will display This is a simple HTML file +This will take care of that sec delay +Adds a arbitrary map to Put and Delete commands +for the patch +Preparation for Gora release candidate +finalizeContext is a better place for handling on connection close events than +Unfortunately this seemed to help some things and failed to help others +If so it might made this issue into a +Closing resolved issues. +Since fixing this would be a compatibility break I feel it should be associated with a major version bumpThat said unless someone can demonstrate otherwise this inconsistency has been present forever I think its too late to change thisWho or what is actually impacted by this +Thanks +Attaching a simple fix that checks this buffer and if it is null close the connection +Theres no need to maintain the per BD instead I think its only useful per DW and for trunk we only have one DW being used at a time +This is from a time when I didnt know there was a difference between Jira close and resolve. +Some adjustification of description hereI deployed ejb to server server server and add server as server mulitpoint server via initialmulitpointerserver attribute in varconfig server as server mutlipoint servers server as server multipoint serverEJB client application is in another server totally sperated from server So the three problems figured out here dont exist any more +Its ok to not respond to new topicnew broker zookeeper watches since on controller failover the new controller will read the new topicsbrokers and put them in New state +extends +Done on the violation drilldown +It will not ask for you to upgrade +bulk move of all unresolved issues from to +Is this bug still open or can it be resolved + lexer grammar +Fix merged into trunk. +this seems a pretty good usecase for +Thanks for testing Shawn +Unsure why someone marked it wasntFix isnt critical for backport so resolving +if thats in the scope for what you guys have in mind for this module go aheadIts in the back of my head +svn ci m Vararging constructor methods much like the method ones were in +Current FOP Trunk compiles just fine on Java +Committed to trunk with revision and branch with . +Also manually fixing branch since plugman fails when dealing with that branch as well. +A typical example is the case I have reportedI have only been testing using the sun jdk and jre +Should be pretty strong reasons to break things out into modules +I finally had a chance to try this + +QA verified +More docs to patch documents options for agent and collector +problem +Deng can we close this one now +Further investigation suggests there is a second bug in and therefore does not need updating +In addition to this the does a few other useful there some reason this decision was takenI providedIf provided +Thanks for reviewingCommitted revision +You might want to mention in the javadoc what you mentioned here in the JIRA about what a large read operation may be +to trunk +Sounds attractive +I still need to finish the transactional logging and the test which Ill do at a later date +is kb and does not contain the binary as stated in the above comment +Comments on phabricator +So those are the three topics with changes in this patchHope this language is suitable? Please let me know if anything needs changing + +I dont know what to do with this bug and patch +There is a gap in time between commit and these calls is that safe? left TODOs in the code at places of concern +Triaged for +I dont even know where else we use the banner. +Patch for synapse applied to trunk and branch +Be awarethat if it depends on features of the JDK it may be a while before includedin the Pool distribution +war mentioned in this war has been removed. +Thanks GuoYour patchis in trunk at revision You guess what? I did not notice Its easy from French to English I dont know from Chinese to English seems harder +I attached as well the template file of checkboxlist that takes count of the html scripting events +Not a blocker also you could always do this yourself by overriding the redirect method in controllers or writing your own http protocol does not allow you to specify a target for redirects +The CLA is still in status Awaiting approval by project lead + +NetHow compatible do we want to be with Java +I think it could fail randomly today I had similar failure for the Page Design Options toolbar item +Patch is for class +It is modified procrunsrs +In that scenario I had several young generation every second +similar as +This file contains the queries which reproduce the bug +I havent tested in a container. +Merged +Version Released +seems to be the source of the divergence in RPC usage that necessitates the munge directives +minor Benthanks a lot for reviewing the patchI uploaded a new patch with a test for async reconfig as well as all your other comments addressedAlex +I agree with your mention +Peter were the changes you were referring to in or do you have more to add here +Id even classify this as a bug fix not improvement as this behavior makes much more sense than the original one +Code is in develop branch +On second thought I think we may want to mark this as INVALID depending on yourresponse +Closing per Bryants approval to do so. +Running patch through hudson +Alphabets are included in Japanese fonts so they should appear +launchestotal r pajbam staff avr r pajbam staff avr +I was also wondering about the synchronization of start +bug of a bulk update all resolutions changed to done please review history to original resolution type +You are correct Knut +I tried on both client and server no effect +Dont see any reason not to allow a symlink here trivial patch attached +Please use for patching the trunk +Just committed this to trunk and +This patch allow in dedicated thread +I guess this issue should be assigned to Svein Erik. +Complete patch +What this is proposing is logging the request object that is logging after the dispatcher has processed the entire request but before the service has executed it +The reality is that WTP doesnt support the latter approach and isnt going to in the near s more it isnt a hack to do it whether Maven dependencies are in a single classpath container or separated into different classpath containers by scope is largely a matter of preference unrelated to WTP and can be made optional +The Bean Validation API has a parser which eliminates the need for the server to parse the validation descriptors directly +Look at the import introduced to theimport Clearly this is specific +it has a timing loop at the end that you can enable by adding an annotation etc +There is only one parameter for the number of blocks the number of block moves per cluster total +Can you give me a more detailed example on where things have been going wrong +to an already existing issue +committed into rev +stub docbook in place moving this subtask under until complete +FixedCommitted revision +There were some other code simplifications unrelated to existence caching +Im reopening the issue so we dont forget to investigate this further +For serialization a type discovery has to happen on every key because theres no guarantee that every key is of the same type +fixed on TCK trunk and in impl +Thanks Nate! +Thanks +Could you check with? I can see the listener being called from the +Also startedstopped container message when running +Remove fix is still planning to release this featureI searched the artifact in maven repositories and svn project but not foundwould be very useful and easy if you could install through features mechanism. +Super minor nit Id prefer to rename to just Status! smile +Paolo I dont think there is any reason to not allow users to limit by scoreRegardless of the perceived value of Lucene scores its a useful feature to have whether or not the scores are normalized +One from JDT and one from us +Sorry I meant Id myself change the code as you suggested Ill do it this evening and try the tests +I would need this feature as well +Closed upon release of. +Changed variable name as per Ted suggestion +A new patch is attached +Reopen it to track this problem +Oops +This is working like a champion now for all kinds of users. +Thanks for reporting this! +There are several alternatives to this approach and all will work however this implementation had requirements it needed to clean for me Report where the cycle occurs and which components where involved and NOT report the constructor that failed to instantiate because we had problem +i vote for rm MC is no more +Creating an API just to abstract another API has no merit but there is performance penalty and more complexitySLFJ is first and foremost a logging API NOT a logging implementationSLFJ has pluggable implementations and supports Logback among others +duplicate of +Robert is Maven working for you +Its the completion of a node path as an argument value +This fragment contains a servlet trying to lookup this which should fail but it succeeds +Type this at the command line ant f ant see the error under Final +Hi ManojI see this is a libxml issue +Rado please verify and close + patch please ignore patch. +Fixed in version. +Can we add REST binding in addition to JMXREST has the advantage of only requiring a browser to invoke the operation +Did you change the property from validate to create after the reverse engineer? This has been reported before +Given tag is from Closing as invalidHOWEVER we should warn the users about the zipball tool is not reliable channel +Spot on Dan! Ive just fixed it +Do you think the EAR verifier is wrong in my caseAll other appservers I tested so far accept my original packaging even previous JBOSS versions +Trunk patch looks fine to me +fixed in r. +please have a look into the or attach yoursource code +looks good + +program that displays file descriptor info as seen by the vm +As it stands to get the behavior I need I had to subclass and change the way contexts are created by getContext to prevent the default strategy from handing out the wrong implementation +Spring templates are implemented already +Moving to M as the target is probably suitable for a plugin we dont supply installed in geronimo +Also rules with the same name being added to the same package will replace older rules keeping consistency and allowing for rule you for reporting and let us know please if you face any other problem on this. +For making HICC a generic data analytics and reporting system setting a default cluster in the user session should be optional +Note that is not a +preliminary support hacked in but dont want to commit until it can be avoided from the by work completed in Mercury. +Thanks +It seems that ant doesnt get error code like us in continuum + has been tagged in CVS +I know you only care about the side but one thing to note The console is merely a client to the operations exposed by the JCA subsystem +Users created and credentials privately sent +Feel free to close with wont fix +Maybe we could add a meaningful unit test +Let me recreate one +Release done +Thanks Nice to hear like you have it worked out if you want to submit a patch Im happy to take it +But I have just made the relevant corrections to the documentation source file +Dito occurs also with an Array in decorated object. +RC! +Cleaning up old if necessary +Attaching patches for and trunk +We dont trim those? Gross +Its unclear why Facebook is sending those messages but its likely either a quirk in Facebook or a quirk in the application code +If you like this feature patch is attached +One thing you still need to do is update the comments in TOBAG because it explicitly makes note of the fact that if the schema of the tuples does not agree that it will return a null schema +Patch looks great except that it needs to include the Apache License header in the new files +unit test and by Vladimir. +Server Version Played a lot with no changeI added a jsp page header to German and added jsp extension then loaded directly same problem +By the way could you take a look through the mean job in +The Value element property can be set by selecting Literal value and then the applicable enum from the dropdown +to the change. +Also in Andyjust wondering if you can create a test case that can reproduce the issue? Or this is something you encountered in productionRegards Andyno worries I created a test case and will commit a fix soon +you can use the over property outside the artifactItems and this will apply to everything that doesnt explicitly list an overwrite value +Will track the fix for speculative execution in a follow up JIRA +I use two helper classes to generate products and categories I dont know if thats the problem but I doubt itIm using Axis +added by Snjezana as fix for fixed this we still have problems with validation +So I will not consider this as a bug in Axis eitherThanks can safely mark this as Wont fix +Lets go ahead with current patch +Verified in EAP with James Livingstones example. +Hiranya Seems you forgot to use the java code for the patch creation +I would finish them in these few days +Found the issue +Some discussion on the Windows dependencies is going on the mailing list +Ill modify and accordingly +I just committed this +The patch has been appliedPlease verify +Thanks Ari +Removed Sun code relicts and license can you confirm that that class is original code and not copied from somewhereSorry for being a pain but we need to be very clean on such things +changing the subject of bug report to reflect the actual defect present within the codebase +Ted So you tested the patch that was submitted here? The random assignment? The balancer only does count of regions not load on the regions so yes Id imagine that its possible some regions would be taking no load +Test results from the earlier Hudson run are good as the patch only changes the name of the command line param and a variableVerified correctness of the param manually on my dev box and ran to confirm that distcp works as expected +The include? is called by rspecs include matcherTry it on a Rails app +should be fixed now. +wouldnt it be better to fix the renderer instead of buliding work arounds in the mosaic reader +Any chance of getting a small project that shows thisThis actually might be fixed with the latest snapshots +Tom Thanks for doing this +I think we need to simplify this example before adding this to the examples +disregard my last comment which i wrote while you wrote yours thanks for looking into this do you have a testcase too maybe + +I have added a for updating Sqoop docs for WindowsPlease let me know if you have any comments on this patchThanks is the link for review board for the to review comments Removed trailing spaces Made an early check for the JDK path and exiting with error if not the patch after addressing all comments from the review boardThanks failure in profile seems to be some sort of issue on Jenkins side will rerun the job. +Fixed as suggested in revision +is there a possibility of a jflex release in the future? It would be nice to regenerate it via IVY like javacc +Adding one more task while we are trying to fix problems with the existing ones might make things a tad more difficult to manage +Marking for +Looks good to me +Committers might be put of applying the fix if the patch fails so keep it to just what is relevant and create separate JIRA for the other issues +When Hibernate scans +But I guess not +Pavel I integrated the patch at revision with minor enhancementsremoved char support everythere dropped files which became bogusI still hope to get regression tests from you +Uploaded which has been tested in a local Hudson build +Fixed with new release of eclipse plugin + +Applied to r +If we find a project that is succesfully using the pluginI think is using it in their new maven buildthen its should be a matter of just doing it the same way they did with version +After loading the test configuration the Class Path property has been stripped after the first semicolonYour assumption is correct that the semicolon is interpreted as comment character and the remaining part of the line is ignored +I think it will be clearer to rename jdk suite to jdbc since these tests are for jdbc +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +And I had no problems +Heres a patch that adds +Another interesting thought If in addition to counting all the older than the readpoint I also count all versions of the seen so far we can use this for memstore compactions +I am attaching a patch which also adds the method under so that there is no dependency to hidden zookeeper packages +If it was on startup you could think of a timing issue or a permission issue with the folder +This has been discussed as part of the scope for the domain model work based on the meetings we had in Westford so I think we will have this covered but I will double check +It would still be easy to forget to use the right constructor since version works on most files +project +Once that patch goes in Ill upload a review to and post the link here +What is a group +For example VM can assert in exceptionjit on line Its duplication of andThe other possible source is evaluated +OK resolving as Wont FixI dont see how Tika can recover when the font itself is buggy +Any update on this or an ETA? This would still be a very useful feature for us to have +I created Diff file to solve this problem by checking the LASTVIEWNAME query string before append to the ajaxUrl ChatreeYour patch is in trunk at r +Ready for review +Is the current solution good enough or we still want to use third party +Thanks Nebojsa for addressing this issue +And I got this working by using the AXIS EJB Provider only by just deleting my installation directories serverdefaultdeployand serverdefaultdeployThank you once again for helping me in this issueRavi +Heiko Im assuming this is primarily a console issue feel free to assign back to me if it isnt +Nice +FerdyDaniel isnt working on +This issue had come up before on thelist and I knew it to be related to empty dirs + +Adding for both and since will actually be released first and may not incorporate all of the eventual issue fixes released in. +Could you please test this against the example in and then apply the patchregards Malcolm I will test soon +Beyond updating what was requested I also bumped the parent POM to build from beta instead of alpha as requested by Nick +This is now invalid after the update to Resolving. +Do you have a test case for this that can reproduce the issue +But as pointed out is often not enough whenyou want to know a little bit more or just get started +attach my patch +What do you think +ahh ok so loadMarkup is the method that is incorrectI will fix that that put is not called when cache key null +No more multiple slfj bindings in designer or console +Can not reproduce either using JBDS and Maven Seam Integration I am marking this jira as resolved +Committed revision Adding example for usage in revision . +Preliminary diff includes refactoring of Zookeeper lock classes to avoid package interdependency of CLIQL and Metastore +This is a task of unknown benefit and unknown time +Attaching a Ganglia screen capture showing an example of how this data would be used +I reopen the ticket +This question is related to Andrew +the skewed grammar will not only benefit list bucketing feature but also other feature like skewed join +I applied the patch and the unit tests continue to pass +Functions we still dont haveabsbinaryconcatwsfieldfindinsetmakesetquotesubstringindexunhex +Eldon are you planning to take another stab at this +In fact I think there are less than ten lines that these two classes have in common +As per discussion in the latest patch seems a better way of doing it +Snapshot is taken automatically by yourkit profiler and I see Gb of softreference generated by xmlbeans + +My plans are to substitute the standard html text area by gwts Rich Text area widget +Afterall this is not deep in axis +It looks to me as if the classstores font sizes in integers representing thousanths of a unit there are and conversions going on +I was skimming through the patch and it looks like its not there in the current trunk? Will be great if anyone can provide its current status or follow up jiras +I need to apply a few more changes which I found on the mailing list and the atinject issues +Ill close this issue once I verify that the code has made the jump to the branch +Just did +The suggestion is to this issues to next release. +You can only have a single lock on a given file at the filesystem level and locks are allowed to be upgraded to each other and by the same process +Please exclude these tests as well +Committed. +Thanks a lot Dudley Ive tested it with jruby S rake packageAnd it now creates the pkg and installed it successfully with jruby S gem install pkg Assaf could you please commit this to trunk +I intentionally did not put it into any of files +I can work on a patch in a couple days for that based on your comment +thanks I do not mind +Moving out of scope for since not enough background scoping done on this change and would be nice to get out + +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +please have a look at the comments in for some discussion about this patch +Patch is in trunk rev +for the patch +If the group information was stored on a table we wont be able to access the group information through normal means until after we assign ROOT META and the group table itselfRight so you could assign ROOT META and the group table randomly at first and then calculate placement and moves of the rest and perhaps also META and group after the group table becomes available right? Worst case that would be moves +Fixed in revision . +We might probably consider providing a test fixng this if resolution is coming later +Manually tested! +This is not applicable for trunkThanksMayank +This patch clears the threadList that holds the older thread references before adding newer threads to it +And no one seems interested on SO. +I dont see how it makes it harder to support more than one operation +changing priority to is more a state of mind than something that should be in the issue tracker +I will be perfectly happy if we keep the src directory too +Use the forums for support and note that Session is NOT thread safe +No this is basic java security stuff +bq +think there is an error in your code as mentioned on the previous copy of this issue +I looked at the formtest example and it seems to work with the old code +Fixed in revision Backported to branch in revision +Updated the ws connector logic to use cxf +Fixed in r +Most of my questions are about what will happen in the VV time frame when child will be running as separate processes in containers that are launched directly by the WF AM not as containers that are launched and monitored by the RM +a trivial comment in RB +Each of these will return false if the search key is in the other half but the coordinating class will only return false in the condition you mention +Thank you Arpit +Yes. +Feel free to open an enhancement request. + to new MR API to new MR Tejas thanks for your reminding +Mostly it shows the information of the previously selected task +I have had the rewrite in place for a while locally and have been waiting to release it until it passed our QA mainly due to above +The main reason was to allow developers to get started with remote EJBs quickly without having to add users and configure security +Let us know if it isnt to your satisfaction +Youll be thankful laterWith the file system based providers its fairly simple iconv or recode can do the job +It must have sneaked in while I was not watching along with the change in the rules youre mentioning +The is because this is a change to the commit build script and the commit build will not test changes outside of +This has been fixed under +either way I can no longer reproduce the problem + to trunk! +Send it to medirectly at . +I can just make it return a simple class and be done. +was working for me at time I did first build going to verify it on this issue about old product build or p enabled product build +Great thanks for that +Id rather leave as is since the follower might end up making progress in the case the leader is able to form a quorum of supportersI dont think I completely understand your comment +This behavior is actually a consequence of which Im leaving as a separate issue given Im not sure its gonna be fixed soon +Looks like Tika has an RC out +These bugs are a result of my misunderstanding of the schema agreements between Nutch and Solr +Updated patch +Updated and trunk patches to match the latest svn revisions. +The TOC in the PDF and html files use the actual topic titleThank you for reminding me to use the proper format for the JIRA issue in the message text +HelloWould you please try my patch? Thanks +Older clients may not see their debug logs if there is a result +The SNAPSHOT is being replaced now +fixed in revision Boris please verify it resloves the fine by Boris +I did use the source +Attached patch for this bug +Please apply your change again +Looks this to +Problems with this are often in the clients +bulk close of all resolved issues as part of closing items. +Thanks Lahiru Sandakith +For the moment I have had to convert my fields to varchar but that isnt a workable solution long term as my fields will actually Unicode text +Thanks for this passes my tests fine now. +I had to make a few tweaks to apply this patch to Thrift on Windows x with UAC turned off +Ioannis and I gonna review it +Hisure sorry was reporting all of these issues in a hurry before a demo to the bossI am using Mysql can provide hibernate connection into and config if required you reproduce this on HSQLDB as well? Anyway this seems to me like a plain jBPM bug which has probably nothing to do with the GPD +Fixed in and trunk +Im going to push this for cr but if I cant replicate it we may need to set up some or screencasting +I will investigate to see if I can figure out what is going on +If we can do it on the NM and not on the RM I think we might be able to conserve RM resources +Doesnt lose fields the previous versionSince Ive made this one disablable with strict mode and seems to help read corrupt files I am going to commit it +It does not say anythung about how many more copies are to be made +So all issues not directly assigned to an AS release are being closed +Rooted at qpidcppsrc this patch locates the template file using a full path name and it adds a common resource include directory for windows builds +so it just could not resolve the package +Thanks very much for the report your keen eye for issues and your patch contributions Clint! Hope to see more in future +please describe not works correctly for me +This patch requires the EMF driver I +In Beta message bundling is still disabled for all services relying on +Ill post the link to the run here when done. +Thomas I dont think this goes into the distributions so the root changesrelease notes havent been updated. +A fix was committed in revision +So then the proposed approach would be for it seems +Well the does overwrite the readwrite serialization methods to add its own fields +you need to ask them first if that doesnt work then ping Digester project is going to upload the sources attachment themselves. +will apply the patch if there are no more issues +I tried on CPU dual core AMD system +Committed to trunk with revision +some special get method like would be accepted Methodboth are incorrectFreeman +zip hotdeploy not really an error but the hint to replace deploy with hotdeploy is easily overreadHope this helps +Committed in svn rev Thanks for the patch Luciano! +you aint in sourceforge territory no more Bastiaan Wont mind jumping through the hoops at a later date though +Moving out of +instances are not supposed to be +Can you try with the versions of the plugin and archetype? Thanks +Apparently the call tree looks fine with though +Its already fixed for. +Sending to Hudson +Re +I have the same problem! In works fine + +Yes please +Here actually DN has removed from the pipeline while recovery +Okay Im going to resolve this we can make a new issue for the case where a replica comes up and is ahead somehow. +Unfortunately not all Maven plugins can properly work when workspace dependency resolution is enabled so I would recommend to submit bug reports to the Maven projects for such plugins +No +thoughts +I tried this on a new roo project against the same database it gives the same errors and outputAttribute type with value VARCHAR must have a value from the list BIT TINYINT SMALLINT INTEGER BIGINT FLOAT REAL NUMERIC DECIMAL CHAR VARCHAR LONGVARCHAR DATE TIME TIMESTAMP BINARY VARBINARY LONGVARBINARY NULL OTHER JAVAOBJECT DISTINCT STRUCT ARRAY BLOB CLOB REF BOOLEANINT BOOLEANCHAR DOUBLE mytestsrcmainresourcesline XML you please try with Roo and a clean project +Verified by Oliver. +IIRC then producer flow control will kick in +Added basic HTML rendering and test selection in revision I think tests should only executed on POST as they might be costly +Closing all tickets +Thanks Prashant! +Hugo fixed it in addbddfdfbbffcccefebd at around the time I submitted my patch closing this as resolved +I guess it would be good to add the hint above to Roos error reagards commit aabebbaecedfbfcfe the felix log stack trace is supressed for default Roo shell environments +I have attached my project jar file and the wsdl that it generates +With a option that would simplify things a lot +. +on patch +Maybe we can do it when finish current work if it is not very urgent +link to the FS a member of the community trying to create a implementation I view these proposed changes as significant deviations from the semantics that are being described as part of +I committed revision +Updated as well. +If we target the class as a plain value holder and keep the data members public Id suggest to call it struct +Thanks for additional infoIm working right now on fixing the issue +Tested the following Master fail over with and with out patch RS fail over RS with parital disable stateOne thing is this patch should get applied on the master for the patch to take into effect because enabling disabling of tables is started by the master +I worked around it by adding a char to the version BND tool has explicit code to drop the qualifier if it only contains numbers Ill have to ask why this is +Dima Im not sure howto verify this one +I think I ran into the PB vs issue +This exercise showed the design tensions so Im now wary of adding more features which subclasses must consider +I will inform the radiant team about a possible solution +Hi Jacopothis is a new patch containing last changes and minor labels that was not committedAfter you have commit this patch I think this issue can be definitely be closedThanks a lot for your workMarco Risaliti +Letsthink about this some moreI agree +This can likely be closed now that is the current release +I tested with and found the same issues +Im going to mark this as INCOMPLETE +The question is did you start with Cloudscape or Cloudscape +It appears that Plexus Archiver overwrites this MANIFEST when it packages the WAR so this will need to be fixed to allow custom +I was sort of surprised to see this code at this level I was expecting the check for synonym to be at a lower level like when the was first initialized maybe even totally hidden in the +Will try and assess what else is different in our setup and see if I can recreate with your bundle above +The problem is that item issuance is created when the shipment is first created and in the input state when order items are assigned to a shipment but item issuance is the only way right now to record items which have been placed onto a shipment +Nothing maybe a nice cold beer and a week of holiday +whatever struts you need to do a file upload +This proposes reshuffling the iteratorelement stuff which is marginally related in the course of actually addressing the issue +Its an existing bug where MR AM doesnt go away when told to +Sure that could work +Basic ganglia patch includes testcase for ganglia webfrontend and ganglia XMLI didnt know how to SSH to the hosts for getting the ganglia XML so I just opened up the port in via the firewall rules +Manually tested. +I see this with surefire plugin as well but cannot edit the issue to put in affected versions +Made some changes to log messages +Cant a bump to minor version enough to add pb stuff? Do we need v +View vs Modify for queues makes sense to me +Thanks for the patch! +in revision . +Like I mentioned earlier this is just a draft patch and will clearly miss some details like youve pointed out +Well that is very strange +Somehow this seems to contradict the fact that was properly resolved +Fixed in revision +Hi Howardthank you for looking into this so quicklySo I have to implement an for every application that switches schemesLooks a little to complicated to me in the first case there is now proxyfirewall involvedHmmm +The reason I wanted these changes in trunk and is so that we do not loose the information on how upgrade can be tested between the point releases to the current codeline of any particular release +I wouldnt be against separate generators in that case +I miss read the previous message hence removed it sorry plz find the updated +I unfortunately have not and dont think Ill have the time to do so in the near futureThe patch was updated to trunk not too long ago so may not be too difficult to revive assuming the original patch worked as fixing the version info for and all affected issues have in all issues not touched in a month to patch fixes some formatting in the latest patch and adds the base for some tests as well as one test its currently failing +Should have seen this when I translated it and made a ticket for both. +Assume tx has done work and locked a row then closed the connecton +can you please attach your wsdl or an test wsdl which could reproduce the problem +Thank you for reporting and let me know if you have any other problem. +BC is a difficult case +Please disregard my last comment about looks like Eli has merged the commit to almost concurrently although it seemed unnecessary. +Attaching another one +It seems that I need to checkout all sources codes and apply this patch then recompile all and jar them to +Itd be great if this could be fixed for EJB cyclic dependencies work for EJB beans and also work in Weblogic for both EJB and EJB beans +The stacktrace will contain all the information a user needs +Patch applied in r +So hibernate does not allow you to use a lazy fetching strategy with any object graph which contains a mapping where the one side may not exist? Basically this is not supported +Fixed in both and codebases by allowing defaultpageid to be null +I didnt know about the existence of +Btw if Im not mistaken the direct integration will use the etags by default but direct GWC links wont because etags are an optional feature +With the last grow up of user from to user the lock appears systematically after day of web app is used and i want to minimize the restart due to this is some hack out of the richfaces release flow that you can suggest you in +Slava is going to prepare some presentationdocs to show how it could look for users +Closing out as it has long ago gone stale +committed to and trunk +Should we add a new issue for the Ant Maven IDL tasksI dont think this should block +no running the test! ThanksGregor Click on the Subversion Commits link just above the comments +We timestamp each topology update that is sent but it is possible that if servers are started at the same time that these could cause problems +ve just committed this to trunk +I have a fix that I need to upload +Thanks for the review +This is the first zone in the setup +Seems to me the problem persists +defer all open issues to for the C backend committed in for Java backend is not as simple so not attempting +bq +Nicholas will post a test planDont want to complicate this issue +ve committed this +This was committed a while back. +Do you agree +Reopened to set the Fix Versions field to JDO beta. +It now handles that case correctly +Michal I would love to help but unfortunately I have not the slightest idea where I should lookI attach my test case as Xcode project +The documentation then goes on to explain what you should expect to see if you apply INTEGER to numbers and strings +great discussion above Bfost and Reesh +Im having the same issueI will attempt to create that database as a non utf +In this case the caller knows that something went wrong +Arnt you used to pushback on your code ideas by now +Working on it +is there any updated version of this patch? What do you want to addI was thinking we could reuse the same system used in couchindex and allows the user to pass any module in the settings that could handle attachment writing +Cosmetic change onlyRemoved unnecessary comment +Thanks for testing ll write up some documentation on this and then Ill restart the release vote +test is unrelated to the patch and is already tracked by +Typo in +Deprecate and add new methods +Will commit the patch shortly +Hi Markus as you mention in this issue is completely duplicated Ken has suggested some resources for tests which we could possibly port to Nutch +This has been ported to the x version of backstage but is untested at the moment +This is a specific use case that may be covered by as a general me mark this as DUP of +youre right +I have same problem + +Do we need something similar in Local stack info +I think you will end up with something more powerful if you write an implementation of ijResult which wraps a +waiting on approval for b +The effect is that the execution of the loop that iterates over all declarations of dependencies to the same groupidartifactid is not cut off after the first dependency was rewritten +Please verify vith your plugins and reopen if you can reproduce this but please provide project that would allow to reproduce issue +The TCK should probably have discovered this problem but obviously it doesnt +Im wondering if there may be a case in which not even the Long is big enough +Also this bug appears in IE so I would assume IE is not immune to it +I believe the current version of is working for all of the examples mentioned in this Joerg mentions you need to use either the short or long option name not the argName when accessing the option +Heres a new version including some new unit tests that set up a mock log reader and mock writers so as to test a lot of edits going through the splitter without actually taking any IO +Am waiting for to be checked inOnce new build is ready will run test suite and see what errors we get +Thanks for applying! +Thanks +I will review this and come up with a proper fix +Have communicated that to Ryan +What about other languages +I will open jira for that +This really should be our main focus in + +How would one application be able to release the containers of another application? Do you mean a malicious application +Any thoughts on it Mark in hindsight? Do you still use it from time to time or do you now think there are better ways of doing it +The rest of the document seems to be ok to meTo start debugging use RunDebug menu +It look like I cannot reproduce it either in my environment +Other iPhone image apps include metadata directly in the file as EXIF +The reviewed and updated guide has been posted to the fusesource wiki. +Otherwise do you think this approach will work +SamuelThe patch is not complete +In bf +Gautam that is a valid name for a dynamic vdb deployment but typically dynamic vdbs are not versioned deployments +Actually point from description is also based on rpc +Looking at the exports from the Windows I suspect that the AXISEXPORT declaration is required for the dynamic link +The attached patch fix this problem for me. +There was a problem in the unit test +Simply removing them is ok +Good cleaning job + +argh +Patch makes sense to me +This patch causes and anything based on or to fail as they no longer receive the value of the element +thanks a lot will make a try +Hi Dogacandid you finish to review my it possible to commit it +Have made it a debug statement in initialize and made output a warning if used with a zero recipient list. +This patch includes changes updated documentation and updated test case +This commit adds Camel FTP binding functionality to the editor and cleans up a bit of the Camel File binding functionality I checked in the other day +but will hopefully make it possible to call that method without adding a custom oneAnyway let me know if you need any other patches in related to this or if not Ill just close itThanks Chris +Contains unit tests and portlet can you please have a quick look at this? Its an old one so I suppose the problem might have gone away in the meantime I will try to look at this next week since it apparently isnt a crisis +ah got it +Well this is easy enough to changejust modify a string constant and negate a condition +Good idea to have some kind of config which is passed through to Tika if possible +it might make more sense to cycle through more delayed containers in that case +Unfortunately the earliest time this could be done would be +All current pending security issues will be children of this new task +OKin my implementatiion I use Company as the defualt and uses Defaultemail as the purpose +Its not ideal but practicalThe other thing we can see is to see why the tests are failing +Had EBS for a while verified working +I asked Julian for the final version of the script in found one small issue that has probably been fixed in an updated we have two options Have a script that can handle multiple jboss instances but requires Have a script that can handle only one jboss instance but needs only I would go with +Can you describe the situation in which fail? What is your OS encoding +The above patches work for me and passes the testsuiteI made updates to to verify this solutionGregory could you please verify if this works for you as wellIf yes I can commit the stuff +I filed for the ref count issue +As per Jochens comment. +this patch adds the sandesha policy to synapse we add this only to the distribution or also to the Web app +If you apply my changes then a number of files are no longer required +Do you have specific action items you want to carry out +This class is now part of the Common project +Updated the last patch does indeed fix the issue for me +A similar failure was reported in and fixed by disabling testing in the soft upgrade phase for the versions that suffer from +At this time the examples need junit so it needs to be in the docsexamples zip +Can you give concrete examples? It would be helpful to have concrete test cases +Can we refactor the repeating code out to method and call thatThis would produce a difference between the branch and TRUNK rightShould we defer this refactoring to another JIRAI agree with the comment about null pointer check +So we do have at least one use case +Does this need to be committed +Cheers for the on that link +Does someone conspire to prevent others from counting root categoriesHehe no no conspiracy +Also verify a few iterations of tests that all passed +Build runtime test with IBM JDK on runtime test with JDK on in kit release notes svn ciSending Transmitting file dataCommitted revision +Could be an issue on unix boxes that we acquire the read access to soon +am i misunderstanding the problem +I am running tests +Fixed javadoc problem introduced by previous submission revision +But the connector module worked from and from top level just the way it is using the version of +Editorial pass over all release notes prior to publication of +It also eliminated a problem with svn in trying to patch one of the golden files +i will add replace high bit chars with in next did but due to lack of time to test what current browsers do with spaces in query i will leave it as it is +This allows a secure client to work in a heterogenous security environment +The optimizer that will construct the point does not know beforehand that the convergence checker selected by the user will use or not the point so I dont think we can optimized away the cloningYou can go ahead with allowing null everywhere it is an improvement for the class that may be useful for some specific optimizers later on +Rob seems like this bug is affecting all renames which we dont yet have a user visible way of controlling what libraries are includedexcluded for a server I dont see any other way out than filter this specific jar out of the list of returned users need the jar they can add +Added email address to info added to all log the patch to be thread safe and applied +I am looking into this +Maybe theres a way to do this without analyzing all the included dependency jars? Or are the dependencies required to complete the analysis on the application code +Len is that more of what you were looking for +Shouldnt be you look at the comment in set the factory refs according to the destinations typesClick on that JIRA the regression occurred in that JIRA which happened in just needs to be reverted and . +The gettersetter for the endpoint option were more or less an afterthought your suggested solution is better +posting separately +This looks fine to me +Hi GiovanniThe intention is that credentials are only required when new dependencies are added it is a protection mechanism to stop abuse of +fixed with revison includes the switch to as replacement for +With that figured out I also pushed eecdcbdabcddaba which allows MS to be restarted byAll the tests now pass for me except which has failed on Windows for +For the moment with the back end functionality we just have to give the url only +defer all issues to +Fixed in trunkIve added a test a check the fix. +I tested on an old toolchain as well for sanity +What am I doing wrongWhat about the product directory tests? Can they run with set to external bits +Running patch through hudson +This was fixed short before. +Please try to reach Cassandra support in order to get helpMeantime I would recommend trying to override Sqoop default type mapping for example to string using StringJarcec +Hi Jonathan it fails for me too onI have never run this test since its in test +for the patchcommitted as r +Patch has already been applied. +We need to ensure that any other +Within a dictionary nothing but strings or streams should be decrypted + +Added samples to samples +Attached implements this feature and adds it to +diff with latest version of there is an error in the middle of the XML strem with the pull model I do not see the value of validating the end tag bacase after all the expected format is not there and it is going to fail anyway +fix on trunk +the patch is +Thanks go +Thank you for your description +Retry again +It should point to the FAQ +Refactored duplicate code for closingdisposing the objects into Connection +Yeah the clusters look good +You should double check what I have done probably +Martin As explained above the natural solution is to put this cache into a context or description that has the same lifecycle as the service class loader +I certainly have seen jobs that longer than that to work off the backlog +Bug will not be fixed upstream +Knut AndersCommitted patch a to trunk with revision +We should investigate why the change disappeared +Log Added advanced search increased input field sizeModified infrastructuretrunkprojectssitesearchPlease can this be propagated to the server? +Patch replaces the two instances of wrong class names with correct class names +As with the issue this is a duplicate of the resolution is WONTFIX use. +If it works I suppose it could be added in terms of a doc or a plugin or an onoff option if peiple here think it worth the effort +Im sorry for reopening this bug so quickly +I thought that because in other conversations we talked about customers still using the standalone configurations in production if they had their own scripted management environment from previous EAP releases +There is also a discussion on the cargo plugin +classpathTransmitting file dataCommitted revision +both is largely untested for a long time and I have no reason to disbelieve that youre seeing this issue +Brian thanks a lot for this detailed description +This upload credentials is the purpose of the plugin currently in buildSrc +Thanks Jacob I will apply though I think the event should be called was my first thought too actually I was just following the naming convention of beginConversation endConversation timeoutConversation +The other way already fixed in one is a problem related to myfaces core and it was fixed on +We may need to notify users in the release notes of the namespace uri change for the spring namespace handler +Changed adaptor interface to remember stream name +works as designed +Committed to trunk +This is the correct patch +How is it insufficient? We have an FTP implementation so we can import data from external systems that way +Sample application based on the portlet example application does not include the Spring and SWF jars +the problem is not paramname it is param railo get in confluct with the possibel param script tagexampleparam name susi defaultSusanne +Just to be clear the issue here is only with allocationSize when using the legacy generators. +This is the correct behavior +Also the DRM is apparently only used to find out about dead cluster members +It is very unlikely that a default EAR archive will create anything remotely usable out of a JSF Kickstart project +Did you check the source at all +Per Hairong request and considering that this patch has been already verified in y release as working I have just committed it to unblock further HDFS development in trunk. +We still need it of courseNotably because for now we cant call a layered lookup from a layered lookup +Making it a blocker forThanks Raghu for finding this bug +The ticket for will be used to keep track of this problem +Doesnt have to because I might need to debug the stuff. +Have test for memory leaks when editing in vpe leaks hasnt been founded +Depending on how the fresh test migration is done this script may need modifyingAlso migrating to tomcatbuild would remove the need for some fo the script actions in container +Looks like we should be point at released +on both patches +Please take a look the patch attrached and let me know if it resolves the problem to your satisfaction committed +Do you think its a good idea to also remove the classes that extend Service from each service? That code its not used +From the context I assume its just an oversight +This can lead to the situation where messages are captured by different Aggregator instances in the cluster and because they are unaware of each other the messages will never be of this you should avoid using the Aggregator action in a clustered environment +Realize this sets the DEFAULTTEMPDIR as a constant to data +Heres another take which leaves SOLINGER unchanged unless explicitly defined good to me +Small test maven project attached +Thanks for the analysis +thanks jukka. + +fixes the test failures and changes a test for the new feature +I increased it to sec and seems the error is gone for the time being +This sounds like a useful feature I will try to get patch reviewed soon +I guess ideally there should be no static variable which controls this for all applications and it should be done by methods on each of the classes +This patch provides basic support for declaring generated columns and for populating them at insertupdate timeo Basic tests written for CREATE TABLE Cooperation of INSERTUPDATE with constraints triggers and indexeso Known not to work Omitting the column datatype when declaring a generation clauseo Not tested at all Schema evolution and statement invalidation Permissions dblook UpgradeContents of the patchUTILITIES MESSAGES TESTSM javaengineorgapachederbyiapiutilM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceM javatestingorgapachederbyTestingfunctionTeststestslangCATALOG SUPPORTM javaengineorgapachederbycatalogM javaengineorgapachederbycatalogtypesM javaengineorgapachederbyiapisqldictionaryM javaengineorgapachederbyiapisqldictionaryDECLARING GENERATED COLUMNSM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileA javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyiapisqlcompileM javaengineorgapachederbyiapisqlcompileINSERTING GENERATED COLUMNSM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyiapisqlexecuteUPDATING GENERATED COLUMNSM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlM javaengineorgapachederbyiapisql +You can for instance specify a pattern organtrl but not src +Error message is not changed on Mac with Java bit and +I would recommend that you simply filter the collection before passing it to the iterate tag +Please choose another when I tried accessing an Archiva dav repo +But I did make some changes to the handlermapping functionality in the timeframe +I missed that +I would suggest missing explicit type parameterthat seems almost helpful +This handling is specific to dynamic vdbs. +POI is active and working towards a release +Could you review the commits for this issue +sorry +We dont need to double ship jars and make things complicated +I think this is a duplicate of +Also missing ivy directory +Queue timers apart from other job types to prevent timers from falling behind +Heres a patch for the easiest of the ideas +The derby starts fine in embedded mode. +We do not maintain documentation for previous It seems that its evident that actionlistener will not work for client mode +Currently Im trying to implement the composite key for my model class I hope I get it working now +What exactly do you suggest? Setting the default inclusion pattern to +Sounds fishy that normalizePath is called tens of thousands of times perhaps not necessary or could cache the result if its being called frequently on the same path +Thanks Carlus +I will push a new mavenized version of caja soon and will submit a separate patch then to update featurescaja +FIXME resolved issues for already released versions +That is the expected behavior for Maven +Those are all methods I added yesterday +These examples also show how to disable SLFJ at same time so thats why its valuable +is the final of this committed to yur branch now +will get the complete small table in the beginning whereas reads itrow by row +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +The workaround for this problem is to explicitly add JAVAHOMEbin to the PATH variable before running the binflume scriptAlthough its a simple fix it may not be an obvious one for new users or novice Java users +Verified closed. +Fixed +I just tested revision +Interesting that the Apache guys didnt notice it either +Fix still needs to be implemented +I will definetely provide the example sometime soon +Add conf option to allow adding cluster name to archiver output +Except for comments and +Is the observing session the same as the one doing the read? Alternatively this could be a case where for some reason the functionality is skipped +My simple approach is just execute Select ID wo unmarshall maybe existing stuff to decide of insertupdate should be usedBonus TODO Encapsulate the select and insertupdate into one transaction to avoid +bq +Test updated to cover all three. +Here are patches for both and trunkPlease remove the first patch +For fact it is working fine +class file +button label and use the proper wizard container instead of null which cause a NPE that was silently ignored +Log file from start up until the server went insane +However when I tried accessing the Properties using I encountered which is now this issue still relevant with the fix to +pointing at another cores data directory should only only cause the create of that new core to failLikewise if someone messes with a cores configuration such that it fails on startup it doesnt seem like this should stop the other correctly configured cores from loading +This is a blocker for us as consider releasing a patch version for this fix +review and commit this patch + to trunk. +Have we even built machinery to support multiple canons for our tests? Do we really want to supply canons for this test one for each of the localizations Derby providesIt seems to me that a better way to make pass in locale is to rewrite the test so that it is not and to check for rather than message strings as we do in our genuine testsSorry if I am but I am very confused about what is being attempted here +Michael this is exactly what i meantThis is what i thought about the API limitationThanks for your investigationRegards Samuel +After downloading and unzipping tptp into the eclipse the errors disappear +A summary of what the user of the change needs to know in terms of how to set it up manage it use it and its precise effects +As a result those still needs to put in applications classpath separately +It also differs by having a default locality group +to commit +Committed revision . +If basic authentication is present it will be honoured you will not be redirected to a form login +Thanks Pavel +because is in the sql standard is not +Santiago please verify that eclipse works for you +was added to provide minimal test coverage. +The root cause as far as I can see is the Mavenization of Hadoop +supported +There should be no NPE. +I am not convinced that the tool you need is a tool that others will use or whether it might be better to keep this in your application +Is there a service that actively polls for statistics to fill in the gap in the case that an event is missed +Closing per defect triage +Attached patch fix the pb by removing attachments from the ddoc we pass to the functionBut it only work if the log level is info with debug level it try to display the ddoc and then it hangs +Perfect harmony between simplicity andbeauty +What tells the scheduler not to assign both of the containers on node and node? These containers could be requested at different priorities but that would essentially be using priorities to do schedulingAre these use cases for YARN? Correct me if Im wrong but my understanding was that the primary reason that the resource scheduler is not a task scheduler is for performance reasons +indentation needs to be instead of I can changed that on commit +Resolved through a series of changes by multiple people culminating in the code generation for Java from Any a few days ago +I am not a developer on the project either +Patch with test case +Updated the sub projects of Jbossas so that their modules are excluded from the Maven deploy operation +bq +Going to update AS to and fix the compilation errors +Which values can you see on the security panel? + +Applied in r +Will fix +What did your look like? processApplication is done again in the servlet if the servlet has a Application +Updated patch +patch for the was applied to trunk +Ill poke around some more tomorrow but if you have any ideas let me know +Ship it +The of the loop should be identified +Patch looks good thanks Yasuhiro +In fact both Date and should be mapped to timestamp +If you finish it before we release beta please correct the version + with time spent +In the future we may need to expand the use of this flag +So for tool vendors this is not an absolute limitation +Are you taking about modifying all EJB and EE metadata via PAT eventsThis would be a very major change +I also dont like to have not used by anyone but I dont have any better solution so far. +There has been no interest on this for a long time +Thanks allows serialization implementations to be used both in transformers and where streaming is needed +Shelly FYI is a different issue with these tests +patch which includes a unit test and no committed to SVN trunk and branchOleg +Checking this in +I believe that the patch doesnt fix this AJAX problem +Hi ErandaYou could try running one of those queries interactively and then lookin your file for the stack trace information java connect jdbcderbybrydbcreatetrue describe quitAfter you run this you should have a in your current directorythanksbryan +Anyone else want to chime in with a aye or nay? If I dont get any complaints Ill commit this before rolling an release +If you still think that this could be an issue you can download a nightly build +If the project is a WTP project we would also add the CDI facet does this make sense for you +Improvements done in patch given by Raviendra Mandre I have implemented Agreement processing for and done some changes as per the detailsThanks to Rishi Solanki for reviewAnd Pranay Pandey for explaning work done by raviendraDone By Ratnesh Upadhyay +Andrey it seems still not working as expected in r +I created a session auth test which passes on trunk but it fails on the code by going into an infinite logon loop +Removed the changes related to forwardingsetDone which looked like the cause of the failures +For example TAB completion works fine forroo logging setup level FATAL package PROJECT Managed SRCMAINRESOURCESspring SRCMAINWEBAPP the deliberate extra in the you could expand on where you encountered this it would be good +Thanks Keith. +IT for in debug +Please refer to comments of for relevant details +Linking some numbers from the pastAccording to my benchmarks at the time saving image contributes to only of the cluster overall startup time +As were supposed to have more and more tests I would love to have more and more small tests +I havent been able to produce a simple test case yet either +Many people are using maven for many years and now are excited about seam but using this technologies together is real nightmare +It may be a thing +Thanks for the patch +Yes there is no guarantee that things wont degrade +Yup thats the oneI tried finding a jira issue i was involved in from years ago about this setting but couldnt dig itthem upWe worked hard to limit the problems it was causing lucene and solr users +Thank you Libor and team for implementing and pushing this to production in such a short time. +Would be nice to have at least process definition to have better startCan this be somehow simulated with test case +Good job for a snowy day +Added the Weld to the resource serialisation whitelistApplied to both the RF core and RF git repos +lgtm +We put efforts to reduce the size and able to reach to MB but it is still over the limit +can you dig into the code a bit more and see what that nil is supposed to be +u admin p admin invoke service trigger command will hang because of the deadlock see the for contains a suggested fix which when applied addresses this issue +Would you please take a look for it? ThanksRegardsIsaac +to ON +I added a comment there but as I mentioned weve only hit this once in months of testing and unfortunately I dont have the time to try to carefully reproduce it +Hi GavLooks like github is now synchronizing +Please attach the pom file you are trying to open and your Eclipse configuration details from Help About Eclipse SDK Configuration Details +Tapestry will not perform a directory listing if the trailing slash is left off +Patch v is ready for integration +Calling this fixed since there was a fix put in place and no. +Due to JIRA weirdness you ened to scroll on the code blocks to see them. +No configuration no tweeking nothing +It my not be worth the risk andor effort if a huge number of changes are required to support Reader especially given there hasnt been a request to support it on +Updates added in trunk +Thanks LevievPatch applied to NIO module at repo revision rPlease check that the patch was applied as you expected +Mistake in the git diff command used to generate the patch led to a busted patch +DerekIll add two protected template methods in order to make it easier to extend the default validation process as you already hope to fix this by the end of the day or at the latest on nextSergio B +overall +As of revision properties and child nodes are only iterable if the parent node is readable + Out of datanodes datanodes are in this state after days of uptimeDo you mean that there is no datanode restart during this uptime? If it is the case we should mark this a blocker +It appears that most DBs do want the escape character by default Oracle is an exception but probably not the only one +andFrom the Java javadocWhen waiting upon a Condition a spurious wakeup is permitted to occur in general as a concession to the underlying platform semantics +Some help from ASF INFRA would be greatly appreciated +i confused rowkey with column name i am so sory for that. +For the moment it demonstrates that this issue is resolved +From the document I get the general notion of cloud services having elastic nodes +All tests pass for me now +Im in the mid of another experiment but Ill keep you posted +Ill resolve this as duplicate and assuming no one else claims pick that up. +I think a regular browser refresh should do this +If there are problems with this Idea or the code submitted please comment +rewrites complex proximity searches into +Support for custom node types can be implemented using a simple file associating the name of the node type with the fully qualified class name of the class implementing the node is part of the element extension point to define new nodes +This is a good thing thanksI committed your patches in trunk revision revision revision About Ajax Pagination I see its well done and works without issues but please what does it brings to us in content cases? I searched to understand but found nothing clearThanks +Should we leave this open to knock out the additional cases you commented or open a separate bug for those and hope someone will get to them +Well see in the next week +I dont think it is appropriate to include rar gbeans in modules so I propose this requirement be handled by adding an external rar to an ear and creating a synthetic ear consisting only of external modules for the war and cases +If not I will address regards the new plagin not as a new featureThanx +Patch number is really tiny +Also I think a load balancing avro sink needs to be possible +Ok trying the second possible fix +If its still a problem please reopen +Not a feature of the Deployer but +Added fix for fix for the missing messages on first browse +If your use case is configuration just use the Java config +This is a parsing issue +Weird thing is that there is actually a unit test that is supposed to check that masking works properly +committed thanks! +Thanks Regis. +Hi supunI tested the patch with above two cases and it passes them correctly +Edited for readablity +Here are the sequence of events how this could happen +Thanks Raghu! +I didnt see much downside with this approach as it only affects the errorcode path with an added retry +Now each of the methods recreates the lock file based on the input method buildDirectory was not Thread Safe at the point of trying to create the directory +We are sure to be able to reproduce it if it is a code issue +Yea was looking at that +Seem to be duplicated +Ive changed the maven setup to allow test classes to be used in downstream modules allowing to be used in Cassandra etcAs for logging I forgot to remove the statements +java Test jdbcderbymemorydbcreatetrue false +fixed by adding parameters other than windowId to the use the +This has all changes before and in additionchanges of and and the re structuring of the UI for subscription of external albums +I could adapt my changes to use that for theAny thoughts +Its dependencies should be fixed in the same way +Jon You got what I mean? If it make sense I can put a patch +With the exception of the API signature tests we made it a priority to make sure all of the variations passed with level code to help ensure backward compatibility +This is what I call damn quick! Thank you guys you are gorgeous! +In which case I will plan on adding source tag at the start +but wanted to get that said in case it triggers some different thinking on this +This patch to Pages and Page uses a Boolean for loginRequired to allow TRUE FALSE or null with null meaning not explicitely set for that page or new way of deciding by most specific instead of any true will only be in effect if at pages element attribute true +The physical side has a separate notion of a visitor for the same +Committed these trivial additions to and trunk. +Patch applied +Two bash scripts to automate build config settings +Woud this be fixed for final or is it planned for a future release +Moved code to +This really is a misuse of the API +May somebody pls try this one +I agree but the current system is not coded that way +Uploading a patch which lifts this restriction for EXISTSNOT EXISTS and adds a test fixture to +remove assert of +Youll see it when you try to run maven +seamplusejb demo should be split into parts one with the setup and another with running the its necessary to use Run As +Would you please try my patchThanksBest regards Oops Ive attach the wrong patch +ATSROOT vs TSROOT TSROOT is the default +Attach a patch +With the help of the infrastructure team our new site is live +There is a discussion happening for publican at the moment that we could implement highlighting and revision markers using the role attributeI think well close this for now because the root fix is essentially done +The whole meta class layout is done to represent one getter and one setter +Since the background for this is to be able to browse the directories through the HDFS overview but to disallow downloading of the actual files maybe we can just use the FS default umask for created directories and only override for created files +Test deployment to apache shows were all set +passes davisp muster +Patch implementing the design discussed above +Committed revision with this change +This is fixed in the production config +This time wo that change +Assume closed as resolved and released +Thanks JonI put this into trunk and +The options tag provides exactly this functionality with the collectionattribute +Rebuilt patch from the trunk after svn update +Note If is in pig classpath found that in real cluster had to add JYTHONPATHLib to pick up the builtin modules as the jar gets extracted on the datanode and Lib is not in classpath +shipped with +This issues use case is handled by the changes in Resolving as duplicate. +I think the archetype is obsolete and should be removed and users redirected to the. +Just had a look of v + looks like most of the BCA cleanup came from merges from fixes +Ill check again tomorrow if whether its consistent +Looked at the specs all of them seemed to say that using same attribute name multiple times within the same section is not allowed +Its a broader issue with the whole BK design +There I always used as the file name +Moving out of +Sorry the previous patch contained few bugs +Just noticed +Can you try the latest version? A lot of things have been changed +If you want to avoid the overhead you can set a parameter and turn thi off and get the old behaviour +This is a low value feature +Not a of a bulk update all resolutions changed to done please review history to original resolution type +now runs against both jdk it just do Ted? Seems a bit much doing and given little real diff between the two +Well I see you have started making changes to the new space therefore it be best I remove the old one to avoid confusion so Ill do that shortly +Same project as except that it packages the war inside of an ear and places the libraries in the ear lib directory +Updated in revision +Another piece of information that I received was the database was possibly executed by a duplicate server at the same time +Derbyall runs cleanly modulo the wisconsin diff +Gatein its for eXo soThanks for answersThanks for reactivity +The only reason really is to compile against the as specified by the specifications both should be the same +Attachment has been added with description fo file and images tested in trunk and this is fixed in the trunk +BTW is DERBY a must? My samples are all using H +WorkaroundMark the derived attribute writable +This way well have an easier time merging potential bug fixes later on. +An exporter with a higher version is preferred over an exporter with a lower version +kidding me +If we can cut that out because those blocks are still in the cache it is a huge win for usWhat more testing do you want before you consider this acceptable +At this point the secondsize runs sees junk and craps out + +Attached the latest patch containing two unit tests +OK the scrub was successful with G of heap +I didnt commit my changes yet +Do you have an inconsistency between what you declare in geometrycolumns and the actual data in the tableCan we see the error message +Its easy to separate logical views using Apache rewrite rules or a repository manager +Applied patch with thanks to Ashwin +its good to use but it gives you the impression there is something wrong with this dance and there really isntI added in r you can swap this into some of these slow tests and see its definitely the problem +Hi Christian if the solution provided by Evgeny doesnt work feel free to reopen this ticket. +Correction +Thank this fixed? Or are you just suggesting the? Because Im using version and I have the same problem +It returned without actually renewing anything +Im essentially catching runtime exceptions and erroring out the respective key +Closing for clean up +I didnt change the sbt file at all +The patch file has the locations where these resources need to be committed +Thilo can you confirm I did the right thing here? If so you can close this +Attached with grants +ex rebase just works trivialex rename a file trivialex code modification that requires more than minutes of though +Thank you for brushing up the noteI think contents of Application Changes Required section was changed a little bitI read the description that TIME data type is completely impossible to useIn the truth application is not prohibitted from using TIME data type if it was appropriateIm thinking to modify as nextIt is needed to correct inappropriate use of TIME data type to use the TIMESTAMP data type insteadly +rOnly accepting the DRLVM patch +Please try it again and verify that you can replicate it +I still have a problem to solve thoughIn he current setup if I have script depending on Bean and script depending on it as well if I recompiled script and Bean because Bean changed then script will not be recompiled because the entries are all up to issue was marked as fixed for and but it seems due to the tests not always showing the real problem it was only partially fixed +Id recommend using Joda +this doesnt work on Windows yet this will be addressed as a separate improvement. +If an exception had been thrown when the test tried to delete A but could not find it the test would have exited and the server would be left in a Consistent state that could be recovered from +the new patch there is no so I will have to reassimilate it allIn the old patch there is a test case that compare two index readers enumerating all parts of an a priori reader and a test reader comparing the values +In merely minutes of running the broker the number of requests in the purgatory map shot upto millionThis can happen for very low volume topics since the replica fetcher requests keep entering this map and since there are no more produce requests coming for those topicspartitions no one ever removes those requests from the map +Hack added +I have a scenario in which the number on average over to each primary document +Ill also upgrade branch to trunk +This patch adds startup message for the RM NM +This issue is causing a lot of pain for many people +From testing the parsed lock mode seems to behive lock table blah sharedFailed with exception No enum const class FAILED Execution Error return code from If I use lock table blah SHARED it works +Committed revision +If the server is slow the time between ajax event and response will increase so as the time frame in which the error is likely to the flood error has same origin as you decrease time between to ajax calls youre more likely getting the error +let me check it out +Even checking out the parent directory might not be sufficient +OK can we get this issued reopened nowThe test case that Phil has provided certainly seemed OK and definitely results in an infinite loop on connect +When JMX is enabled as default this issue with conflicting urls should also be fixed so it works +You cant have it both ways +This is not a blocker for the EAP release but we need to document in the release notes that it is broken +This patch is against the head as of revision +I just committed this + +Marking closed. +If you bring over the same effect you put on glossy ike I think it would be a winner +Are you suggesting that the major version changes for any protocol change but the addition of an method name? Thats the only permitted compatible kind of change to a protocol? That seems reasonableNote that if a method is removed one can decide between incrementing the version or generating a runtime error when its called on a basis +Looks good to me +Apparently fixed in the current release +Not reallyYour best bet is to run the mvn verify command in the project +they seem to be loadedhooked up before the rest of the app +Link to deletion request for +Code committed on trunk and patch applied on in JBDS +yea lets do it +Hi Andrus Actually I have worked around this along the lines of discussion on the mailing list +Understood +I already did this +So the version has to be stored to a file somewhere +Compile is fine but if I try it never runs tests while without the patch it does +Attribute reRender was added to +Addressed Bikas comments +Manually need for integration tests. + +I guess there is nothing to do here for the tip +Just let me know +What platformHave you tried the standard maven untangling tricks like deleting +Thus getcapabilities is wrong but fortunately Geoserver is not that strict +Forgive my ignorance my only excuse is that my solution is cleaner and also works in a portlet environmentregardsMartin +bq +Performance fixed +Nialls fix seems to be working Passed tests in morning and evening which seemed to have some impact on the bugPlease reopen if its not working for you Li +However no Thrift types support virtual methods so this case does not apply to Thrift +It now works as it suppose to except it defaults to logj and not simplelogSomething I dont understand happens when is in the javalibendorsed directory +Again it passes with no issues +First iteration on this one +The main problem here is not with the MVC but rather a problem of metadata notification sequence +Thanks selckin! Keep feeding that awesome youve got over there +But please remember that I will have to address the problems with the test cases first +I suggest you add this to the Jira issue I will create for tracking improvemens and bug for the base classesUploaded which fixes +Thats a problem with your Maven setup not an issue +Verified with CR +confluence seems to be hanging when I try and edit the release page so Ill update that later with the changes relating to my last commit +Hi HyunsikNever mind that +yes what is needed here is to reap these builds that Continuum lost contact with the child process +A simple fix is adding the dependency to your project +In terms of dcmi we tend to take a pragmatic view on metadata standards +Just Roll back + to trunk +I am new to Ofbiz and i want to look code of ofbiz application with freemarker presentaionx in the user mailing list. +Thanks. +WSDL including the XSDs with no of the xsds with no targetNamespace its seems to be related to Axis more than to +This is partially a backport of but also fixes other files that are only relevant on the branch +And this is what Infinispan is doing and +gah i forgot that facet even existed m from before it was me +This is not considered a requirement of although may be revisited again in the mvn must be in the operating system search path +Improvement should be are now checked lazily +But I think we shouldnt add env to the functions where it is absolutely unnecessary and make the parameter list unnecessarily long +The error handling for wrong and unknown host are improved +looks like this is true for the as to max for his patchapplied to and +Trunk patch looks +verified on ESB Tools +Yes! Sorry for the mistake +Its that a parameter that was initially literal but was then changed by a mixin is now returning the updated valueDoes that feel better? +I just committed this +The patch also adds a forgotten test case to the default suiteCommitted to trunk with revision + FILTER +added with TRACE enabled for m rescheduling this since its clear IMO that Infinispan is not having issues here and so I dont see the rush to look into it further in the timeframe of adding cache managers is not needed any more since multiple instances running on a local machine can be found +Thanks for your speed applying the patch! +Have tried the times different combinations of and for both Roo and Social versions of Spring framework but no luck +Awesome +A test case would be nice though I can hack up a test case tonight +As other people took interest in thisI updated this with our most recent version +I guess we could reuse a or something I did wonder if we might want to support synchronous dispatch of the messages for easier transaction handling +The service source +I will update the wiki page with the steps I had to do to get it deployed +The directory is not open or used by the other applications +Changed the to in to properly load all java SQL typesVARCHAR is being treated properly now. +Ive made the dependency provided in many places and removed the dependencies for api and rtcore +Verified in issue was not implemented correctly the test app will be attached to test app was provided +Then I should be able to do the backport to shortly thereafterI still have to run all the tests yet again but I dont really expect much of a problemShould SOLR and all be closed as part of +If youre defining your own classes then you might be better off using Nullable for those fieldsThis is not to say theres not a bug here that we should fix +If we do add partParams we should have DDL statements to view and modify these +Im also encountering this issue +So you are building against a snapshot of the trunk? If so that must be a bit of a bumpy ride at the moment +I think this is a goodGenerally task can be implement in following ways a +Before apply it make sure that it comes after the th patch of +ThanksVincent +In other words the general idea is that the replyChannel header is used once a Message arrives at the end of a pipeline assuming there is a reply Message available at that this explain the behavior you are seeing or do you still think there is a bug +For example you should see something like this below svn stat src buildsiteM srcdocumentationcontentxdocsmanualsA srcdocumentationcontentxdocsmanualsM srcdocumentationcontentxdocsM buildsiteM buildsitemanualsM buildsitemanualsA buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsMore pages changed under buildsitemanuals than what you explicitly changed because of your update to +Address Ted comment and suggestion I not use progressfailed +Hey StackI could probably have a patch ready tomorrow mid next week at the latest +Correct +This might technically be an incompatible change since the output of the shell is considered a public API but I dont think thats relevant here given it only applies to the help output and that changes as we add flags +Thanks Jonathan! +Fix and test for this issue +Theres a report of this problem happening on too +Closing now that Hudson also approves +I have done some manual verification but nothing on a cluster over node +added filtering on actual jar path and c tests +Thats not something we do +It is unrelated to this work and I could add it back in if you thought it would be good +The point of using is twofold Firstly for performance whoever needed a threadsafe toString method and anybody who does can synchronize themselves +What do you think? +Bulk closing for also run into this issue when following the documentation on the syncope website Run Syncope in real environments +Sorry about that Carlos +Even the configuration is removed in the same jiraSo the bug is present in +Anyways now I seem to be ok. +Michael please update the spring please update groovybooking +Can you please revert commit bdafdeacddbaffacdb download the diff from the reviewboard and commit it +Hi ChristianSorry not yet I hope I can get chance soonRegardsFreeman +error reporting is informative about the causeThe wording could be improved thoughI move the issue to is a very old jira the text is referring to the initial implementation of our tooling +I see so Apache is actually responsible for the Github apache organization but technically the mirroring is done by Github +We probably dont wanna leave those properties like that +bulk defer of unresolved bugs to +Michael I assume you did read my blog post about resource handling in meclipse +The conversion between to is finally testedI added a simple testcase in +removing Hudson from the components list I think youre sorted on that count without our help +Andrea is the maintainer but yeah it looks like a typo +Currently ppr can execute serverside code and change html which is great +The test passes on all platforms +I did Run asJunit test on srctestjava which executes tests in classes +No problem +Since we dont ship TCP UDP transports with Synapse its may be not good to ship local transport with Synapse +The tests are excluded with jcc driver +The havent shown up on devcommons mailinglist +Assign to during tidy up prior to +They are not of +Would the plugin switching from using bashsh to binbash when Solaris is detected work in your environment? I dont have a Solaris box to test +It first checks if the current node is the same as the child node in the parent node at the index position equal to the row index +Note that the lines are commented out and not active for patch +Code changes were submitted as part of SVN checkins and +due to some change +Is there anything else I could check on it before I restart it +can you move the new code out of the for loop? And also pls confirm that mysql has the same semantic here +I still think it is useful to have this unless is going to tell me the offending row key +Theres now a cookieName property on the class. +result of running src on an Amazon EC +Can you estimate how much work it would be to port the improvement to I havent read the paper but it seems that the result is less accurate +Fix for bugTested against all current ipojo annotations and custom annotations to produce same MANIFEST after on both directory and +Changes to configure the DAG to generate splits in the AMThere was a check on numTasks! as a result of split generation which Ive removed +It should work fine without extending is just a convenience +Ill ask James to give us his option tooCheers too route planning version of the logo +Id just go ahead and build the generic distributed system keyspace in and well rebase this on top +Sorry for coming in late +what type of predicates can be pushed down and how are they applied? Thanks out of bounds +Hi GeorgeWithout that changeI get an error sayingscore and score should match Expected But was DIGY +The suggested approach would produce partially confusing result because there might be multiple instances for the same logDirs within second periodBut it would say that its aborted right? I actually prefer that +Im averaging about seconds per run with the original reference counts and about seconds with atomic reference counts +I have applied the patch but it doesnt help +It has been implemented in +patch looks good +additionally Im not sure how active rome is! Ill head over to user and see if I can get a resolution to the problem +Changing version +I debugged the execution and I was not able to find the bugWhen the regional setting is set to Turkish then something stops the exections of the Launch Configuration which wakes up our listener +In that case a user may want to start and stop the server without stopping the embedded engine +Ambud can you describe what the intent of this contrib is? Also can you describe what the problem is with more details +Still see the problem +covers the alternative of prompting for passwords +Using the following setup Apache Resin Hibernate For anyone using Resin I solved this issue using the JVM server option as well +If I understand this correctly the issue is that you observe calls to render the tabs when they are not the active tab? You are expecting the rendering to be +I prefer it to the where attribute +Cannot do much with this till we get a failing test case +What is the current status of this issue +Can we close this out? I dont think this is an issue now that and are in +sounds like the motivation is different but Im wondering if this is related to that was also just created +Sorry those changes got lost +This was done before I joined the Maven team I dont know the history +need to merge into trunk is probably fixed by r but I dont know what running means +much better Ubuntu smoke kernel +If they break other transports then we need to figure out another solution that works correctly for all transports +Regarding I still think this should be tunable since if a workload has column with very large data then they may want to increase the slab side to fit this otherwise they cant take advantage of this +As we will not risk introducing any last minute issue in the which is being cut today +Maybe a glitchReopen if someone else runs into the same issue. +Resolved conflicts with recent trunk changes +I need to commit stuff on JDBC and Oracle modules concerning and but actually I do not have commit rights on that modules and at all +Please verify +Would anyone want to try this +Good suggestion +alexis im missing the problem you are pointing out +I have changed my mind +a similar problem with Japanese +Thanks for reporting this changed to Unix line delimiters +When we publish to the we then only need an +lines char is originating from Twiki doc patch and so is exempt +patch to add relational operator optimization rules and logical plan migration a new another patch with functionality of loadfilterjoinstore and a few other expression operatorsThis patch is self sufficient and can be applied directly on SVN Trunk +enhtml the queue starts from the lowercase not Queue +the jar inside the bundle should be called my are corrected +The cover is px x px so it should be OK +A interactive subprocess is interactive and outputs its results automatically to the parents out +Reopening until we resolve my concerns +Will do on both counts +So Im guessing the confusion comes from the fact that I didnt manage to complete the thoughtThe Simple Web Service wizard provides a way to handle the case in Chapter as well as the WTP Web Service wizard +Laura please could you take a look at this and check if the Installation Guide is complete in native connectors on RHEL? +I thought about that but its pretty clear exception wise when you go over the frame size that it is a transport to think nice to have and was easy enough to add +More infoThe places where we through an broken pipe appear to be places where the channel has been closed unexpectedly +bulk close of all resolved issues prior to release. +Thanks Srimanth. +Is there a proper place to get help with build issues? Not a critical issue for me but I am curious about the process. +I bet of the users do not care about a wrongly named package when everything works +I do not know why but your test case works just fine and I do not see any exceptions +Committed revision +move the utility methods in to and then removeThe reason why I did not add it to is to avoid it having awareness of its subclasses +Unable to reproduce on the +Here is a patchIve moved the methods outside the class and changed how they are called inside the classBecauser these two functions are basically utility function which might get used somewhere else in the future it makes sense to put them in a separate file with other utility functions +My pleasureI will try to do it. +Thus it should be working on +Hi ClausPlease find attached a that fulfills this requirement +Commenting on a bug filed against a release for an +OK its time to go to sleep +Provided a patch that lets optionally output XML declarations with double quotes. +the Exception does not have default construtor or constructor +Thanks Janet! I will review the patch later +I see that it is not referenced in the config file I have attached +bq +Please update the JIRA explaining the stuff and the extent you have done this +It has some javadoc +Thoughts + unscheduled issues to the next release +This patch adds the Japanese manuals to the section of the documentation tab of the Derby website and removes the Japanese and Portuguese manuals from the alpha section +All of the new bits I could find added to the new site case anyone wants a quick peek +So as far as I am aware doesnt need any additional properties specified within the file however both the SQL Cassandra stores do +Committed revision test updated +I will address them in updated patch +There was a problem with groupId for fastinfoset jar so it was corrected and a new bundle placed on the download page of +So you basically just removed the skinclass allocation from your s as the default skin already is the correctOr did you specifically set the +I committed the attached patch to the trunk with svn has been committed +It sounds like is not on the classpath +So all issues not directly assigned to an AS release are being closed +get and get return defaults if not set +Additional note I dont know why but putting above in classpath order the project uses and work fine but if you put above in classpath order you have the only thing I can think of why would make any difference here is that it brings in a different version of mvel with it that gets loaded before the one in the classpath already +The attached patch is the least invasive +is meant for cases where an intervening save makes the underlying content incompatible with the changes that the current session is trying to save +It doesnt have to be a nested composite componentPutting the namespace declaration in the surrounding element works here too +I see this is handling the real user for proxy tokens the same way as for webhdfs +Has anybody succeeded in using Tomcat Realm authentication with Jboss Seam? +I just wish we could delete out the old mrv code so we dont have to maintain two copies +charles are you on windows or unix? if windows did you install tomcat using the +So the goal is that minimized the hacker can get once the hacker compromise one machineCompatibility is a issue but I minimized that by only use access control when it is necessary +please verify in has been verified by running the sample tiles webapp included in the distribution as of revision +Ok then is there a plan to add this functionality? If not that is fine +Note Ive deleted version and mergedmoved all its assigned issues to version After I did this I realized I probably better would have archived version instead +The code change is not that straight forward though +Cheers +Updating to instead of M to reflect commits +resubmitting against trunk +Thank you +I would like this fixed for to ease distribution mechanisms +In this injected interval overrides anything else +nulls are more universal then empty stringsThis is a straw man argument This feature request has no intention to change anything about handling type conversion andor null conversion of types +Closing as duplicate of +My first reaction would be that the linestring is not actually valid compared to the simple feature model so its the data being broken not the software +updated comments and committed +The correct thing to do here is when searching for method references and a potential match is found check the arguments +It was for C +yeah +Hi VenkatMy oracle G returns the date column as date type without hourminsec info +I dont know if he is a contributor or a Red Hat employee +Closed after release. +I can reproduce it with Tools XML Editor only +Im not sure how easy this will be to doThe issue appears to be in the logging the toString method of calls which calls the weblogic which is in turn calling the method +Cannot reproduce and I have not heard back from the original reporter +Do you have a jira account +Hello Ed Swierkthank you for your comment +Specific issue is because of the wrong message sent by the client +This should fix it but I havent been able to test it because Im having problems connecting to EC instances once they start +Patch looks good for me +Function to dynamically addupdate a pageflow as a string + +For this defect the issue is not seen after the workaround provided in +When I stop a process the Local Processes view updates in real time without any need to hit seems all is fine +Should we deprecate with this +Heres a nearly complete version of this +Here the patch with the suggested changes +This version will make work side by side with For +Committed revision +Initial patch +In theory coverage management and app schema should be completely unrelated turns out they are not +Will it be possible to have installed an Eclipse on the gump machine +This patch simply converts the property file containing a single version value to a class containing a single version value +The metrics are not computed anymore but its still possible to select them for instance in the Custom Measures are flagged as hidden from UI +Thanks Marek +Spring Framework is really planned as a direct to with no change in system requirements and no major changes in bootstrap and configuration so Id definitely recommend a straight upgrade to Spring Framework there even just for this that Spring Security will generally work fine in combination with Spring Framework +Using awk instead of tr now +Ive been using this plugin at work for the past couple months and it seems pretty solid +I wonder why this did not occur to me before +The new PR solves the problem on Linux +to +Confirmed that this works with the current trunk version of. +This issue was resolved in upstream and the fix is included in +So its possible theres another explanation all threads dont wake up but because some of them are still in selector logic the socket cant be set back to blocking +When I did the percentage changes recently I did put the non normal regions on my still to be checked list but havent got around to it yet +All the wsdl files are pasted in this are pasted in thisthis should be fixed in latest nightly +This issue is a part of actuallyJBT doesnt support custom tags for CSS +Thanks Ravi +Will commit as soon as jenkins does a +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Using Apache CMS was suggested for creating the podling site +Thanks Dan. +CommittedI looked again at all the comments here wow this issue was open for a long time and at lot of work was done herePaul thanks for your patience and all your hard work +Yes I think it is good to go +I am thinking the clone operation will be left up the user to give them the flexibility that you mentioned +I dont know if its a legal term if its fine to use nd party to imply another product from Apache or if that is only subcompaniescontractors +NET without source changes anyway? I suppose alternatively we can wrap it in conditionals +Its not urgent +remove current path patch works as expected with and +Just wanted to say you guys totally rock! Great teamwork here with all the work involved of getting the branch merged back +It tends to give more of a view of resource utilization it is hard to see the long tail tasks and what might have caused them +This is really an issue with leading white spaces +FYI Ive managed to fix the aforementioned CGLIB and have confirmed that it fixes the as mentionedYou are right waiting for the next stable release of CGLIB would be wise +Verified fixed in ER. +Would you be willing to submit a patch against latest cvsthanks bug as per clives remarksthanksdims +Patches to undo this particular change included in the Patch Collection issue. +It should be something related to richfaces in my humble opinion +i am not sure how a user will cope +I believe you likely have a permissions problem +Thx! +I believe the patch is only for +Im rejecting this Jira as it is incorrect as many failover tests are working properly on the profileLooking at Andrews patchesbranch I believe the intention of the Jira was to cast light on the problems with the firing of the exception listener during failover under +for configuration +More work needs to be done heretModels dont need a separate docDesc collection +retained for backwards compatibility but will now pass on to and well have just one script to kill the broker +I cant seem to find a way to delete the old one maybe someone who has access can delete the old one +Ill assume its fixed in trunk or if not its a rare enough issue that well file it again when we see it. +It was more of a DOMSAX model where Node were created on the fly +Im not particularly familiar with the code so its slow going +Windows users will likely submit zip files and not tar files since tar is not native to windows +I dont want to parameterize the mapping +Will if it comes back. +is embedded into the WAR +Ill commit shortly +It happens too +Attached patch changes this +Tested again The result of the clean is an empty PDF page + +I think that if you replace the listbox where you addremove fields with a simple field where users can type in fieldfieldfield I think thats fine for GA +Added a pluggable property from which the pool gets a CL to set as TCCL when creating a pool thread or returning one to the pool +Andrew and I had a chat on IRC about this +Patch adding a more complete testcase and a fix +Committing minor fix in config file that was missed in the original patch +Also one line in one file file solves all the problems without changing individula project buildsAnother problem is using in C context which leads to CC conflict +bq +I dont see how its even possible to remove those kinds of static references +Problem is now hidden with review by assuming receivedTime sentTime +as in the case of distinct values +Open ports in Ubuntu in page I think we dont need to manually configure libvirt and selinux in for kvm as will do all the same thing for you automatically when adding the host from UI +Did anybody else have a chance to review this contribution? Ok if I go ahead and commitTheres still the question if we should ask for a software grant +CamdenThank you for pasting this +This is because the insertion and the deletion can end up in different +Are you getting a file from get but then no data from texttable? Are the clob contents correct +For bucketized tables there is no guarantee that mapper will be able to hold all the data in memorySo we would always need a reducer +Matt if you dont mind Id like to merge this into since its been well +In the future we may want to get rid of the bat files and use exe files instead +GregoryWill you commit the patch or can I reassign this to me +A test case and a patch are attached +I attached a sample which can show the same error in the car maven plugin trunk version +Thanks for the patch! Applied and committedSince you seem to have a good testing ground for what mbstring has on php shindig did you run into any other issues with it +there is no maintenance mechism for the tests +I only decrement once in a while to keep the counter low and as indicated in my comments that may also not be necessary +Kurt Im targeting this for +This issue has been fixed since Spring alreadyJuergen +Thats interesting +I agree that the simplified solution is enough for the Beta but when you do not mention that in the Jira its quite hard to validate itAlso for the history purposes it is fine to have the new solution mentioned in Jira +This is the way Spring works in taglibs etc that cant be injected +I testedthe applications against the legacy RP and the new behavior degradesgracefully which seems sufficient for the distribution +It seems that problem occurs only when server is under load during the testing phase and it is not able to undeploydeploy queues after this phase +thomas looks like quite a few of the tests failed +Most of the suggested changed have been applied +The test succeeded yesterday for B records in totalExcellent! Lets get it in +Felix is embedded as POM +Please review and commit this patch +This is based on Davids Idea of using for keeping template of Preventive maintenance on +Could you tell me how did you conduct your test? what is the machine IP address of each machine? Also could you upload your test here so I can reproduce the issueThanksHoward +I havent tested this on latest trunk +Please feel free to reopen this if this is still a problem. +Looks good can you add a negative test where you are using regex columns in group by queriesAfter that I will run test and commit if the tests a negative test for Group By regexcol +Thanks +Are you sure the file is a resource to the project? Worked for mePlease post a sample that demonstrates the issue +Streaming mostly doesnt work across different versions anyway so I would be in favor of gossiping the Cassandra version and requiring matching versions to stream +Fixed. +Is there a corresponding JIRA ticket to add documentation for a Fuse library JMX Management guide +Running run Failures Errors Skipped Time elapsed all API and Pattern issues +It took more than expected time but finally it works fine. +The new and Array classes are using a pattern of abstract property methods in the abstract superclass that are overridden in subclasses +I think the behavior is dependent on the version of Scala used +It wasnt hard to modify the code to offer a point and point +This provides a massive page that can be searched in a browser it also provides a single page strongly associated with each module so developers feel comfortable answering common questions about their moduleMarking this as closed. +query facet etc and so a user could add their component at any point +Fix that only tries to use the manifiest from the archive being assembled if it exists +Attached Thanks for the review Ted +Done apologies for delay. +Eclipse project to demonstrate bugDatabase dump and cayenne project are included +However we would be particularly interested in any other scenarios which would cause to be updated under normal usage +bq +Piyush please drop your comment here so that I can see your JIRA id +This code needs cleanup but should give an idea of how the boot of startSlave could be done +weekly builds added for with OWB with OWB with OWB with Weld with Weld with Weld with Weld with Weld in +The attached test case which indeed fails with runs fine with so marking resolved. +Patch updates errors and addresses java warnings upon project import in the +But continuing to keep this open is precluding further work on +I think whatever we do needs to be incremental +bq +Thanks Arpit RishiDone at r +Patch applied at r +Its over the top but nice debuggingIt is only on if you enable debug +were the ui artifacts produced prior to the tests and do you find ajar in the current build tree +getUser API is for ROOT admin only by design +The will now automatically configure the descriptor path of additional engines +Pushed this to trunk and +Using TMPDIR is fairly standard and I think the right thing to do by default +As soon as weve implemented the were going to introduce it to the other contributers +Its backwards compatible and addresses the Designer issue +That way TTL and versions could be controlled from the outside +I think it should be logged once at startup and as info +I know you said it might be close but I dont think you ever said it was working correct +I committed this +Hence you need to add a new test or improve the existing one in such a way that the test fails without this bug fix and passes with it +This has fixed excalibur and all the other machines seem to be happy +It could be even better + in the build and found to be OK. +The build does now works without EPL licensed files +Hi Mike to this ideaDo you envisage instances being immutable or would you be able to change the Analyzer on a? If they are mutable would you see instances being shared across multiple Fields? Or would each Field have its own instance +Oh patch seems obsolete is enough +I just reverted the shell invocation line to be binshUsers should follow the build instructions and set FORRESTHOME +The jdk is not likely the issue if I too set my logging to DEBUG level I get the exception but the code runs properly ERROR level +SCRIPTFILENAMEKEY should be +Why? If we do not promote Extras site at all and only promote Central why should we test the Extras site? The installation from Extras site should be at your own risk +doesnt belong over in +Moving it into are we going to need this for +Ted committed this yesterday +Included in release. +I entered another ticket for the reporting issues +Unless I hear any objections Ill make sure that the message properties files get the license header as well +Multi has a couple of additional capabilitiesScan is limited to one table +Let me see if I can get the debug output and see if that helps +no longer has a reference to +From the Thrift download +Created follow up task for review steps so this can be closed as fixed. +Yes I tried PDF but a US Letter PDF is a pain to look at on a tablet especially with diagrams or pictures. +The stripped down is now called +There is in fact an issue with the default value it is setting +release bulk close +With this little patch +The above is just my c and might just add additional admin load rather than being helpful +Remotely thrown exceptions should appear locally identically to the same exception thrown locallyHeres a patch that takes the first approach +Sorry +The downloaded fine on our Jenkins server but did not work on my new local workstation +I corrected the OK button to be right aligned in Windows as it is on Linux +Jay we just need a consistent api to get the byteBuffer in the scala and the java version of +sample t this addressing almost the same issue as? Also I would like it if the would bind to the by default because that is what it does. +The compilation failure seems to be unrelated +whether the connection is coming from a host that the Kerberos key is supposed to be used on +This issue occurred a couple of years back and is similar to what I am seeing now +I will try to get the adjustment in soon +which are most used in portal applications now +Verified that and also work +If you want to upgrade to then we can do that +This causes quite a slow down with the distributedunit tests +You can do an svn add even if you cant commit as it is just a local change and then the added files will be included in future diffs covering the filesdirectoriesetc +Missing version informations it is most likely a mysql bug that it cannot handle comments correctly +See the above comment for explanation of the fix +I agree with your treatment of +I see this as well once my cluster gets under load this can happen on a regular basis +merged in the master +Can you update bugs field in the review board entry so that the updates will be seen here? Thanks +Will commit this soon unless there are Julien seems fine here +Verified in separator type in Properties correctly changes separators in Source. +jbellis Note that if repair didnt flush it would also be broken for another reason You could no longer rely on the start time of a successfully completed repair session when comparing against gc grace seconds. +The patch changes the dependency +Either way its been a long time since I worked with it +prior to generating release notes +Then we will get exact folder where to push the bits and then we can set it up on hudson slaves and in we dont have to wget the root of the server could wget a subset so itd go faster +Sorry I dropped the ball on this s happening right now +Ill go over it again and may add extra unit tests +Putting in too many hours perhaps +But as far as I know the section id is only used by the book code to resolve relative links between different source documents so slashes are necessary if you have I think the book descriptor is too complicated in general and should be replaced by something more usable +Possible someone test the attached patch +Can you please post your custom for completeness sake +Same here +Ill post that as a bigger for a later releaseCan we close this issue +not blocking Beta but upgrading the priority on MD sums means we got something wrong and if we dont fix it the md files have very little value. +If you have a low FD limit andor a lot of sstables this can cause problems +I think it does linear search on some collections which will be slow if large or complex +This has the same explanation as +Unless there are strong objections I can help replicate this patch for +The other verifies some assumptions +Thanks for testing the patch Urban and KatheyCommitted revision +I am closing this issue now and will open one after we release Cactus. +Happy to let you have a cleaner radar GuillaumeI have introduced the init parameter encoding to be used for template source files only in and not in as I didnt see any use of input encoding either in or +Yeah adds to the logs when NSRE thrown +Updating Fix For to Unknown on issues not targeted for this out the code leading to this JIRA no longer exists and as per we actually now need to implement heartbeating entirely. +this class has some serious issues but also had some scary dependencies +Do you think it worth it? I would be able to do thatThanksAndriy +seems to work and allows job to fail early in if specified types are not compatible +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +For example an applet jar with a manifest would be JNLP applications that are deployed via a war project +Might as well close it t seem like this will ever happen unless its a big huge itch for someone willing to share a solution +the parser doest really handle this yet. +So would be better than and as a general solution some variation of might be better +A patch was attached to issue +The patch shouldnt contain changes in its my mistake +The new jira test fails with the old code but passes with the new codeSuggestions about additional testing that may be needed are welcome +Could you give a small howto exampleThanks works! Thanks + because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This bug has been marked as a duplicate of +Cant they just increase the stack size of the JVM? Are you serious? Should this UIMA feature be documented as If a large list FS is created in the CAS and the is used be sure to increase the JVM stack size accordingly +Committed +Wizards should only show the important settings not all an editor for this instead or show it in the property sheet in eclipse +I have verified that you can use this new knob to test upgrade from custom patch releases +These issues depend upon community releases of AS that are still pending +Actually I dont think we should do this heres my line of reasoningInstead of dictacting data structures and maps in this abstract classreally we should just have abstract methodsBut we already have these! they are createOutputopenInputAnything else is unnecessary bloat here +Alex Liu the V patch is working fine on branch except one problem +This is probably because Maven is endorsing other xml libraries but those are not in the compilation classpath so you miss out on both truehelp +as sometimes its the customer requirement +I just committed this +This is a serious oversight +See related issues +Hi SebbIndeed it should I just tentatively wanted to make sure that IIOE was present pre which I didnt have the opportunity to check at the time +See if that fixes it? If so there is a problem w the table balancer it seems +JacquesI just tested the new patch with trunk +Mainly because that is the minimum required java version for Hadoop and Lucene +Thanks in advance +Updated patch gets service and jdbc tests passing as well +Setup CA by entering CA Name details etcStep +I set aspecific type of intance in it at runtime +Chances are high that your jsvc is busted in some way or you are running in secure mode or something similar +setting this to the LDAP Provider component. +html and +just comment everything regarding skins out +Is this ready for a commit to trunk +Patch for applied thanks! +i upload the code but not sure it had side effect or not +The EAP build repository tags currently inherit from the EAP tags +In we use jdiff to generate this +Added logic to look ahead for execsqlinstructions within a program and to initopen the processors this is the situation when there are sibling mappingclasses at the same document level +I ended up at this issue after looking for a way to have different configurations for dev notice that this issue has a low priority and I am surprised by this as most production ready systems support different configurations +In this case you can for now add a simple based where expression executes File rename you for the update +Well there are actually two unreleated things in this issue +Using async to save the states is also effective however the code can get complicated +doesnt exist yet so the bug doesnt exist there +What comments +You dont service factory to use axis with Spring +Borja I can confirm that after running mvn site the mvn will generate a report where styling is workingAfter running mvn site I get HOME +Thanks for any feedback +I dont want to rush anything out when we are so near to the release. +Ill fix that before I commit +If there will be more jiras like this there will be new build +This patch fixes a findbugs warning +Sorry my previous patch was overly verbose I did not notice that does not have svn set +If anyone watching this issue wants to check out the code and start testing it your feedback would be welcomed +Generally a bit fasterbutif you do a bunch of stream copying first and then use transfer its a dog the more copying first the worse it appears to beBut stream copying does not appear to degrade like thatIm using a gig heap for these tests +I did not know there is a Wiki just followed the Apache convention to get to the site +As Jiri noted several failing tests are related to database metadata acquisition +Verified fixed in Version Build id Build date +Addressed stacks comments +This fails under this patchThe patch doesnt actually provide any benefit unless it were setting the toolsversion or p command line argumentsWe should defer this until we have a separate concept of the tools version and the target version. +This issue can be solved by patch in in closing it. +Committed the patch to trunk in r +Let me know if you think we should play with the colors more and Id by happy to look into alternatives that lighten up the shapes m creating two reference nubs with smoother lines and will attach them shortly +The branch is created +fixed +This patch works +now passes I made sure the relevant selector is waken up. +Ive committed this and merged to and +putting the conf routine back allows history server to start +It turns out it was a misconfiguration in the +Ill chat with you on this +I added an unmarshal method which takes a +OK +I am working on a code fix for this in the branch +I did not bug it to eclipse because I am confident of this fact however if you insist that I bug this in eclipse I will +i ran the tests on windows it seems to pass +Thank you +Bruno sorry about that +We certainly need to improve that +But patch doesnt apply cleanly on fresh svn checkout +Q whats the reasoning behind creating an anonymous child context? Does it make sense at all? +rrd files in your tests with imagemagick too? Would it be possible that imagemagick was faster because it was reading the overviews +I was thinking that maybe there should be a tombstone deprecated comment in solrconfig for max so that somebody doing a file compare could know that they canshould clean up their solrconfig as opposed to some feature that may have simply disappeared or movedOtherwise for moving forward towards closure on this +Thanks for reviews lads +nsme ah ok +I just committed this +This is a common gotcha with XML since it is not obvious that there is this difference between attributes and elements. +See for further details. +I agree the problem is not the class hierarchy it is that outer class A is locking inner class B and in turn B is trying to lock A +When prepending or appending an ellipsis the code tries to make sure this happens on a word boundary +Verified to not cause problems in nightly testing. +As mentioned in the comments above resolving this as Cannot if this issue is still valid in the light of RC and your are better able to create a reproducible test case feel free to reopen the issue +Will commit in a day or two if theres no more comments +Verified with JBT. +Will users ever want to be able to grant execute permission to another user? I guess we are saying we dont allow such case +I like Todds idea of each mapper randomizing some fixed range of keys that buys a predictable number of rows and also a defined key space for doing random reads. +Thanks Jitendra! +this is a patch proposal against trunk rev +Improved with r. +Work is done +I think it is an important issue because some applications do store the file paths +Yes I will do that +I am not sure how these files have to be placed exactlywould be nice if you can have a look here Thilo +The pom being used to build the sample was missing this dependency and so the for the missing component was not being created +Justin please review this issue +I committed this +Looks good to me +Theres been no work on them currently and might be best to focus on them in a release instead of trying to squeeze them into the releaseHopefully this will keep the release scope pretty tight +Attached patch adds the two extra tests described above and addresses the dup import +I wrote this new feature so i assign it to i commit it? Any drawbacks +Cool thats the effect I was expecting it being slightly better to explain what is wrong than to error during the compilation +Assigning bakc for assessment +Why did you need to modify Ants bzip code? Could it not be used as isId hate to have to copy this code into Hadoop +We use visual selected node to do actions like toggle drag drop set selection and show resizer +Thus assigning to Koen for resolving when that is ready. +for your suggestion Luc +Verified for EAP Used war and jmx file from +Broker configuration used for the was actually looking at that code yesterday and noticed that the locking on the destinationsMutex could result in unnecessary waits +process except the return would be RTNREFUND Replace all the occurrence of service process under update eca with processWith these two changes everything would be working as usual +Corrected initial comment + +Ive checked it works for. +SijieI agree with you simple logic is to shutdown the bookie when thresold reaches and give the ctrl to the bookie recovery admin tool or restart the bookie +To get filenames try postscurl httplocalhostsolr F file FfileI like using nc l as the server first in these cases to see exactly whats coming across the wireAnd it seems like we should support multiple config files at once say you add a field that references to and want to upload the at the same timeideas optionally take a different confName in case youre prepping a config for a new collection that doesnt exist yet +In such a hole case master will wait for meta when assign user regions but SSH for meta server is prevented before initialized so one confliction will happen +On Ubuntu with the difference isnt as drastic but still significant time ruby server devnullrealusersys +Right! Looks good to me +This bug was inadvertently fixed by jdcasey in r when fixing +Thanks Jonathan! +Heres a heres the patchFor some reason the mini cluster was started with in but only inChanged it to in as wellTest runs for s on my machine +This needs to be reopened due to call Henry +My only question regarding the implementation does it really have to be so complex? I mean isnt it enough to call to obtain the instead of implementing the +We cant force older impls out there to implement what we need sure we can fix it in core easy enough but its a real hassle to do this in another way that doesnt require outside multitermquery impls to change we are going to have to fall back to this anyway with any current plans +perhaps we should a a specific error message that points into this direction in the method instead of swallowing the exception +My bad. +Attached is a test case that exposes the bug +So this feature has not been implemented yet +They should use the new bundle instead right +Yes the code was sent to me by harryh post in the gxp group they are usually pretty responsive +Hence deployment planner agreed to deploy but actual deployment failed on host due to insufficient resource availability +Fine with me +Heres a patch to avoid value where not needed in case of inner joinIn my microbenchmark where I was a big table with a small table this brought the task execution time down from seconds to seconds on about million records on the big table the second table being very small and the output is small too +attach a patch rebased on latest good +Resolved with updated versions of in +We then will be able to support our target development enviroment for say eclipseeclipse at a higher pom level +Let me know if anythings wrong. +Thanks again for the patch +And works if changes are applied i suggested in last you very much for your advice Michael! Ill take a look at this problem asap +verified +Mamta just curious why did the NPE occurs in transaction mode in but not in autocommit mode +May be filter console configurations basics on the db connection urlits tricky +fix priority fp per the issue meeting on May . +I dont think implementation will present a major problem I am just concerned about the user demand on this +This patch checks in the first rev of the release notes +OlegAgreed this is not the nicest solution +This issue is tentatively approved for SOA CP pending results of engineering investigation + for the patch +If for example an ajax script is added for the other event handlers the id is properly writtenThanks for the fast fix +Removed znode +Than I do not understand it sorry Can you post a test case showing the problem +I assume youre working with AS code +All of my concerns have been addressed +Im sorry to jump in late on this but I thought this was really important to considerId like to propose that we deprecate it really is a bogus value in many cases as this jira explains +Should the original problem remain then please +I havent been following can you please summarize the proposal here for discussion? Thanks +The fact that we dont support it is why we were able to make this change +Let me try to reproduce with latest IDE and GA release of Fuse the bug is still there +thxHi Matthias +this issue as invalid. +Personally Im not familiar with Storm and would not help maintain a Storm sink so I would not review this +While I was working in it I upgraded all of the tests to junit +This has been fixed way over a year ago. +Almost everything for this feature is in place except the actual implementation +It is also worthwhile to run the system and see the impact +Provided human translation The process definition is somewhat esoteric the kind you see in BPM demos to show how twisted your model can be but rarely in real d suggest a second fork be introduced after Fork but before Task and Task to correctly balance Join +let us know on the dev forum if you find how to fix dont support tomcat yet +changing component to common since problem is in new unit tests for in rev +stack There is only one and a Writer in a for at any time only one writer takes the lock and the other will wait +There is a bug unless there are also caching headers in place that suggest the client to keep the tiles for some time if thats the case well the client will simply not ask again for some time. +As it is mvn eclipseeclipse causes all of the projects to register as being out of sync because there are unversioned files in the directory +Please open a new issue and attach stack trace data and example code if possible for any new problems found during your testing +Otherwise it would have already worked since it should pick up the gems from your environmentAs far as testing thats tough +Can you please take a look into that +Will that mapping work after it is createdYes selected Classes Packages separated by the projects and files generated for each of the project separately +for the patchTest failure seems to be unrelated to this patch +All finished fine except of tests passed but AS wasnt stopped correctly there are two bugs No space left on device causing firefox not launching and causing tests failing +the issue in cygwin is a known issue to them though +The API is simpler and is consistent between java and scala +This is an expected behavior +We are now waiting for the artifacts to be replicated out +If Trie needs to be part of the new field cache implementation that would be a big red flag +It would be nice to have package level comments in as well +Whats your mgmt server ip +I did not understand the second part related to seconds +oh btw this was the wrong patchThanks Suraj who has seen this +If users want to deploy in another container thats perfectly acceptable +Proposed patch to branch r +MST is always supposed to be not ever +Thanks! +It should pass the original event +If the waits then it appears to work on my Mac the Cancel button does not appear to do anything of importance +Id love to see you guys support Java +I have attached a new patch that simplifies the makefilefilter function and adds some comments to help prevent confusion around the Python APIAlso I realized that my earlier patch did not include all of the files that I changed +For RST we can set the action at trustrst +Jaxen itself is just fine +Thanks Rex +Hi I have now fixed the additional failures reported with both forms of for loop +Thank you very much Saksham SrivastavaI saw also if the VM is connected to another NIC is not destroyed +Why dont we open an issue and track this downIs there an issue out there already that we can piggy back on? I seem to remember the idea of just moving to multicore coming up before did we not get to an issue for it thoughId really love to see the single core sutff get subsumed under the multicore a lot cleaner than the multiple code paths especially as zookeeper comes into play +Marking for hopefully Wayne can get it in before RC +added more description for nonInteractive option +Chris I am not if if we gain much by limiting the of fsync +Please update to a more recent version of CXF +Attached patch fixes jdktool build to resolve this problem and facilitate testing explicitly set execution permission for deployjdkjrebinjava file after copying JRE for testing dont clean and recompile tests each time testing starts remove copied JRE files while cleaning jdktools component +except the above question the patch lgtm +Initial patch for review +Thanks Chris the patch has been applied +t an example of new actsrc usage be added in s? +The test data generated in is not compatible with the pattern specified hence the failure + Bug has been marked as a duplicate of this bug +The business needs an incremental number for a piece of data so I need an incremental column +Thanks Saliya! necessary Maven fixes to work the CN moduleThanks patch provided by Saliya +I particularly like the generic propogation down to the nested record reader +I was looking for it in the UI pane that shows me the results +I can do this later on quick question first mind if I change your Issue Types from Default Issue Types to Apache Default Issue Types? this will remove unwanted issue types and align it with the rest of the projects also makes the field configurations for those issue types more inline with the Wiki Enabled Screen Scheme that we will switch to +shut down notification for super group Escalations +Fixed in CVS +In terms of API I think that Hemanth means something likeYes thats what I meant +In the use of a static final string is a good idea to remove the use of string constants +you have to remove the repositories section and make sure all dependencies are already in central is the status of this update? Dozer fixes a number of issues some of which I have encountered +Hey Konstantin should this JIRA be moved to the MAPREDUCE project +I could modify detect modified files as web archivesand repair all the aspects that fail when such archiveis redeployed +has been created to allay these concerns that have been repeatedly brought up to convey that it is goal of to have optimized snapshots +Try this patch outAnyways this does look like a bug in Android +DFS will immediately leave safe mode with when there are blocks +Patch is committed to trunk +asking the user for more off the rc path since I dont have any information to reproduce this +looks good for now +Thanks Myrna +bin +The problem is that in the current version of the patch is that there are two copies of each abacus class file one in and one in +has changed the area that i was mentioning above +Im closing this issue +Some minor comments on the test +When might this be completed so that we can announce the availability of the feature? This is the last bit of work and we have community members that are wanting to install it +Konstantin this is interesting! I will take a look at how causes the infinite loop after I come back from my vacation +I get this error messagecant find file to patch at input line Perhaps you used the wrong p or strip optionThe text leading up to this wasdiff git ajavaengineorgapachederbyiapisqldictionary bjavaengineorgapachederbyiapisqldictionaryindex +Using the constraint with other languages should be fairly easy though by registering other validators +That seems to have been the problem +Forgot to mention patches through addressed Sanjays feedback in the previous comment +Ted are you suggesting we should revert this one or just push for to be resolved and backported before the release +Oeps sorry my comment was about the maven jar plugin +One last thought then does it make any sense to have throw an rather than a? Then you wouldnt need any trycatch blocks in at all but could just allow the to escape upward to wherever they will be caught higher up in the stack +AttachedThis example is not read correctly +looks good if the tests +Of course modifying the comobox to optionally behave as a select would be great +but it should have been +If this approaches sounds good I will do a similar change for replication znodes handling too +Uploading a first patch for this feature which is intended to implement the specification as it currently standsPlease see detailed patch notes in +Sending Adding Sending Committed revision +m gonna try to work on this I have a perf improvement to spans and this class prevents it +this should be fixed when xulrunner supports gets here could you please verify when that is done and then resolve this bug +If I do import in Boo I might get everything? +bq +However how to set this up is Servlet container specific +This is already in Faces albeit untested and undocumented +But Id usually put workarounds as close to the defect as possible to avoid ruining unrelated parts of the code base + +So Im associating the two bugids with no condition that this be fixed before is marked resolved +Are we saying that the JMX connector is listening on an arbitrary port? That would be really bad as it sounds like it would mean you cant configure a firewall to allow remote management +Closing all tickets +Thank you for your contribution JeffTrunk rev work Jeff! Looks like this is missing from trunk + code modified +The url is from original thrift code +The user could define the a task with the same name as the assets task as we were lazily creating the task +This patch updates the link to point to the correct place and also moves to the proper place to have it be used +Switching to Blocker as this issue seems to be getting ignored and blocks running end to endKelven this issue was discovered days ago and days ago I gave the details of where it came from suggested a fix and assigned it to youCould you please take a look and comment +Alternatively maybe we should name them lastId and last +Deferred to EAP +Attached screenshot of new look +The component would be more useable for people if it was java code only as then you dont drag along scala dependency +In that case you can omit the +Perhaps we should explore creating a commondeploy folder and adding it as a deploy is a patch to copy the file just to the default deploy directory as Charles suggested +fixes +I suspect there are other edge cases that I may have missed maybe someone more familiar with the data model can suggest some additional checks +Ive poked around the code for a while to try to figure out whats wrong here +Should we put something in so that we have something as a provided dependency so that command line stuff still works +The updated controls chapter of the documentation should address this issue to a satisfactory level +Im not sure it is a good idea to remove it and I think the Executor should take steps to ensure that it is accurate +To use the test case unzip it to a directoryThen cd and update the file to update the value of that points to EAP home directory +for the patch +J has component +Thx for the patch mate. +These are low level data structure changes which are committed separate from user transactions btree split operations page allocation +Thanks +Further more we will enhance existing Natures as time goes by +All verified typos in original ticket have also been corrected in the EAP version of the doc +Correct. +This code mainly contains the test framework for JAXB Schema conversion +So I am testing with this enabled +However I would then suggest that perhaps there might be a more graceful failure available rather than an unqualified NPE +Joram Ill do that as soon as posible +Not sure how I did it before +How is this not a bugThe component is the last one to be shut down so there is no damage done +Committed to trunk and +The problem appears to be the following +Merged to the branch in revision . +Please could you implement this issueThank youJuraj +Were also looking into replacing our useage of commons el with the builtin implementation in Websphere +Current all UT passed +Committed to trunk and +I cant reproduce this on Android with the Nexus S or with the Galaxy Nexus +I verified this fixed the issue for me on Helios +Should create an component to assign this too +Given that latest comment from Mathias to this issue was added in Sept +Set to ms before using the socket in order to prevent a dead lock by waiting on the inputstream +Do you have any idea +in this case the first option element of the select boxThere is embedded within the page that use the pages forms +Thomas this good for you +well I opened it per recommendation in the forumbut still assumed there is a possibility that this is a bug +Havent figured it out yet +I will test it tomorrow +Hi MikeI wonder if youd want to follow through with this cause you have invested already some timeregards +Maybe we can refine itself to allow for both usage ve added a track flag to for gathering statistics without logging +JPA connection properties need to be set on airavata client also if you use JPA registry directly in the client +Closed with the release of. +However for a server this would be no good +Perhaps that work would resolve this JIRA too +NP +Fixes variable column label + Bulk move to ve added interfaces +Did you check the +Fix compile failure in latest patch fixes all the unit test failures +Committed to trunk with revision Committed to with revision Im marking the issue as resolved +Done. +we want the separate state so we can support removing nodes from the ring entirely as well as load balancing +Thanks Lohit! +Seeing this problem withwithout the ESB plugin +Configuring logj to DEBUG causes an additional log entryDEBUG send +Ralph can you comment on whis +reopen if still issue with the new configuration +Hey Ning any update on this issue is greatly appreciated +In other words I dont want to commit the time if theres no chance of it making it into the codebase +M window was extended so test cases and implementation are now committed in the tree +I see violation of other rules in source tree that are not checked now +Fixed +My additional commits address the nodes representing the logical operators +The EOF exception caused as follow compresses the input data +Simple patch now that is done +Devaraj are you OK with this patch after Karthiks clarifications +If we are going to add a attribute we may as well make it extensible for future attributes. +These are both linked on +These bugs must have been lurking the the Apache for a long timeI get a environment setup as test against this as well +Committed +Have you checked that the network proxy is configured on the archiva proxy connection +committed thanks +Is this intended? If so what credentials can we use? your Xircles credentials. +Was it really fixed +That would make verification automatic and there will be noneed to write verification code +and +Hmm thats too badIm still a bit confused what kind of libraries we may use for apache projects +I will following up on this. +Backported to the branch as svn closing. +Although the dependencies are always added in the same order in the bind phase when they are retrieved they may come back in a different order cf which retrieves an Enumeration from the hash table +I was too quick in ing itLets revert and do this for instead +It should come in as a utilty so you could do +Seems like a good idea +Marking as passed in testify +thanks Tom for tracking this down +So all issues not directly assigned to an AS release are being closed +Could either use an enum or have an extra deferredFlush flagIn I cant break the RPC compatibility so I have to use an attribute +So yeah +I thought there was a requirement for Spring to be visible in the application classloader +The latest version of the code on Github still has this error +I work on a maven project where the version is fixed so I cannot simply upgrade +Leaving this issue open but +The net saving of space is around . +line change that fixes this issue inRoot cause a discriminator object was created and assigned when it was used in a but would never be reset afterwards causing it to be applied to any parsed thereafterSolution In the end processing of a assign the found discriminator to the current and then reset it to null +Fixes this issue +Thomas minor comment http http +Hi LineshAre you using Axis C version? If so can you try this with versions or you can get latest codes from CVS and build and check with that +Execution start and duration at different Infinispan nodes. +Now eclipse launcher should work with +A temporary workaround would be to try to upload that pom to nexus directly +This can be avoided by use +I did not mean racks per block +But I think I can find that tomorrow out by peeking inputStreamIf the debugger shows any memory accessRead you tomorrow +On running the unit tests and measuring the cost of adding this information to I get the following partitions serialized String length Thats the length of the String equivalent of the serialized and how it changes as number of partitions added increase +That takes us back to the Bad Old Days though +Especially when the tradeoff is for giving better support to a usecase we should be discouraging +This issue is closed as out of date as per the middleware process rules for aging of bug reports +Will the final UG be a PDF +Even more so now that significant changes were committed to Nutch +The issues are now closed. +bump +for the patch +passedAlso passed tests fromThe patch with regression tests from the JIRA issues listed above to followThe patch may be committed to SVN +Yes I tested it and I get an now which is then handled in the component code which tries to register +Or it could take a Strength argument that is simply appended to the SHA prefix and passed as the algorithm to the +Its quite frustrating when editing long operations +Moved over from sandbox +I already fixed the test but accidentally submitted the wrong patch will reattach the correct one +You are bound to get better texts to put in the report than you would by parsing SCM messages +The better is that Castor only change the timestamps in a true commit +now I see +This request implies that somehow the Assembly Model in the Tuscany runtime got changed after it was loaded so that there is a need to save the changes back to files so that they are not lost +Updated to sslr and now use sslr toolkit instead of devkit +Therefore lets close this issue until I find a provable test case which is not very likely to happen based on your comment of buffers are associated with sockets not with events +test case with no external dependencies except for lucene and junit +bulk close of all resolved issues as part of closing items. +the mail archive private is incomplete Could you please elaborate? Are you saying that it doesnt currently functionbq +I found out that the IDXS metastore tables are generated when I create an index for the first time +updated patch following our discussion on irc +Thats because we currently dont store the delimiter when storing the schema! We just use it to delimit the header file +Attaching patch which adds a new section under the configuration subheading in +You dont need to override to set a different default translator +I guess you should commit then Ill make a patch of the diff between my modified patched dir and what you commit +Id rather close this issue and create a new one about the property thingWe are already mixing various issues here +Namit The latest patch addressed your comments +good point ill move the init into the try block +Sure Claus +Herman can you file a new jira for the and submit a new patch which is based of latest trunk +Not found this issue in latest runs +property dialog support has been added. +Jan maybe though I dont want the functionality to depend upon more obscure features of solr or xmlThe other issue is that this is not supported when loading config from zookeeper so we would probably have to create another example set without +And it can lead to very serious code andor system administration problems getting misdiagnosedThanks this is a valid point I agree +change the Outputs impl to suit their needs +JSR doesnt have any workaroundYour comment on classloader is interesting +The base implementation would be a string that matches the line prefix +Soon cleaning up code sorry for the delay +Hi Vinay the config is only used for tests +Please ignore forget about this issue +Thanks Daryn! Filed to track fixing the findbugs warnings. +Therefore there is an inconsistency +was able to execute the following without error in QUOTEDIDENTIFIER ONSELECT g +This one you have written is a generic oneThank you for your help +Any suggesstion? +The first draft patch +the issue was fixed for M for plugin rename I create separate task +m directory you should not need to use the file +Either way its a defect. +The third is different from the perspective of instantiation code but I cannot remember why it was made to be different +I am attaching a patch the fixes the typos +Some test coverage would also be nice Oleg +Marking all current RESOLUTION LATER issues to be against the Nightly Build +Fixed help text and added a message reporting number of rows imported and time elapsedUpdated the branch +check the snapshottable directories stored in fixes test failures +Can you give that a quick test to see +Thanks again Tom +Now I can resolve external JNDI references witout any trick +Mathiassame message than for you should have asked first on the mailing list before creating such tickets and we would have answered you that this is definitely not a this is probably related to the fact that some of your files cant be parsed yet +In my case I have an that is type and a Mask that is a JBIG and they do not go through the code +Extraction is completely broken now so I filed +The problem is on the due to some classpath issue +I hadnt checked that but will take a problem though is that Im writing a library that will be used by other applications so I have no control over how they handle their view rendering +Thanks in advance +This patch looks good +I try to be as fast as Contegix once a day +This would close the loop +Committed +Fix has been applied to branch for release +Is this ok for you +Im marking this as resolved since everything stable is now using bundle and package versions except the transaction wrapper parent which is blocked by +I see in this entry that the surefire integration is scheduled for but and I hope that youll forgive me I havent been able to figure out when that release is scheduled to take place +Id prefer the version not be a separate file but would not reject this design over that +I think that patch needs more + just committed this. +I have no idea if this is the only missing delegating call it just seemed wrong to me not to do it in +Steps to reproduceMake federated build in debug mode on Windows x ant debugin workingjdk directory run CLASSPATH +I will submit another patch to add support for encryption in temporary file +removed unnecessary change to applied the individual patches instead. +Shipping the artifact in the Maven repository implies that we support it and afaik we dont want to support GuiceAs for why we build it I guess its related to EAP productization policies +Unfortunately the methods are protectedI am proposing that we make the methods public to allow the convenience of using other packages besides +I used two Windows XP machines. +I can also confirm that it works perfectly without reencoding in JSP parts of the site +Patch to fix bug +Looks good thanks there accompanying documentation that should be updated with this? If we dont already it would be nice to enumerate the environment variables that are configurable through ACCUMULOCONFDIR it seems like ACCUMULOKILLCMD should be added to w a comment +Closing +Heres a patch that sets the output directory to classes and the javadoc directory to javadoc +oooh seems like we have another candidate for removal since it is duplicating an existing feature +just realized that this JIRA was for the support isnt due until so perhaps this might be better as a seperate JIRA + +This really needs to get fixed +For backwards compatibility it defaults to false +Ill make sure this gets prioritized behind for your patience +Thanks heres docs added along where pointed +Output on on would say that this is a workaround and we need a better fixBut this patch can be probably applied +Can we change the expansion of QPIDTSS to use boost on windows and continue using QPIDTSS as our platform neutral abstraction +Resolving this issue +coo thanks +I think this sounds fine +Writes need disk spaceLater reads may generate errors but no committed data is lost +Sorry for the possible confusion typo I meant Your patch though still includes the old ones not just once +It turns out that the combination of commands used to get the base directories also prepends a n to the output +Their current stable version does not work with Servicemix have put a fix into trunk for it and their next version will be compatible +Patch to hive service check to work with security enabled +solrI dont think thats possiblehttplocalhostsolr is a simple page that redirects you to each of the cores in a single core setting it has one link to hte admin page but in a setup it links to the admin page for each coreIt cant really use the because it doesnt have a to work with +getting rid of an unregistered stale target now requires first registering it only to delete it +Im still not convinced that all of this meta data belongs in the field set +The looks for conditions on partition columns in the filter following the load and extracts them out and passes it to the implementation in the loader + +Fixed under r +Please attach your jar and your configuration extension configuration attached +Could you call to in jdwp right before calling to +A patch for would be nice since this is pretty important fixup operation being able to close region +Updated as per Michals this issue as an age old bug +Mmmh actually I commented too fast +Note that this does not mean the platform code itself can be redistributed +Start of master fails if old masters znode is hanging source of the tests mentioned in this JIRA +I also fixed my self probably the same way as you did but had to extend the and the classes +The wrapper returns the hashcode of the delegate and the other one the filters hashCode +That is for it should be fixed in CP but Michal needs an build to confirm +The error no suitable renderer found are due by the offline generator requesting every uri with different to generate the different files +In the absence of parentheses are evaluated before EXCEPT or UNION operations but EXCEPT and UNION operations are evaluated from left to right +I was doing something else in my code there +Because on my Windows system w it was definitely happeningThanks +In order to minimize the index size I choosed to use this patch and set the default payload to f rather than storing f in the payloadHow and why I managed to get a majority of the payloads to be f and if I could avoid this by manageing boosts in another layer is a completely different story +JDK provided an initCause method and this method was to support both pre and post JDK use cases +Yep +bq +It currently includes the Environment which is not that relevant any more +QE please verify this issue +The patch has been appliedPlease verify +to in r +Do you think we should provide both +Guys can you please test this +As replicator is a separate database it would be very surprising for a completely separate event to cause a document deletion +Done for Beta and in JBT +Darn we should fix this asap +Even stuff is risky for instance if we bring up a brand new web server on a reused IP before its been completely configured if the IP was not removed from the upstream load balancer before this was the case it might begin to answer requests before it was ready to go +characters per line looks good to me too +I added a test in the resource reader if the same source is read +configuration addedBut is read only currently which fail the buildolamy hostnameolamy touch touch cannot touch file +Great thank you for merging the patch so fast! +This shouldnt necessarily be considered a blocker for it could be fixed in a point release for example +There are still some performance issues with the current namespace implementation but those are best handled in other issues like +With this patch a long running client will recover even from local being restarted +Youre right! is overkillHmm I think even volatile is overkill ensureOpen is only on a best effort basis it makes no guarantee that it will be immediate +A simple text editor IS a valid tool +In case anyone reads this issue after the fact the title and description need to be taken together to understand the issue because an OTS implementation is allowed to throw Unavailable from gettxcontext if it does not want to be XA compliantThe Unavailable exception is raised if the Transaction Service implementation chooses to restrict the availability of the +Anyways i will submit the patches to to the list for greater feedback but just wanted to get the feedback from you guys first about jdbc specific stuff +SSL tunneling usually implies that the client is expected to connect to the proxy with plain HTTP establish a tunnel to the target host with HTTP CONNECT and upgrade transport to SSL +A resolution issue for this has become critical so expect a merge of the fix into the trunk sometime around May st +Can another commiteradmin try +Paul isnt that something for you +The updated patch doesnt capture all the changes +should only use RAF +Thoughts +Committed revision +Thank you LauraI have no further suggestions to make here +Was the intended fix version rather than +IOW this is the equivalent behavior of in hadoop +Interim patch added to master +Implemented the changes in Rev +Unit test + applies to trunk and applies to the branch reviewed the patches and they look good. +The serial field was added too with the generated value +In reply to comment That would be nice +Marked Docs as this may require some update in this department +Patch is Updated also ant test results are attached Thanks for version case for hadoop version command is updated to be more generic for GIT and SVN log for version comment on that the version test case should be added back in as part of a patch for that Jira +Just tried it on the latest code and it is fixed +Let me know if Im missing something +Initially the proposal is to submit Owl as a Pig contrib project +Thanks Milind. +Forgot to say the performance improvement doesnt seem to come from where Joydeep expected +I will convert this into an issue that adds a builder for doing that +for this one I had to hack my way around you package private out the issus for the release. +definemethod appears to be special in this case so well go with my fix. +This is exactly the kind of testing that needs to be done to validate SMX as the deployment environment of choice for CXF endpoints +the new y patch +It works anyway +is adding system test support so hopefully this will get better as we move things that are really system tests into system test and out of unit test +Weve had several people see this already +If there is a problem then I suspect doing some dirty stuff +Thanks! +Ive joined myself and send mails to the others that should join the project +Java version fixed +Thanks for digging in on the failed test in spite of my and patch versions before commit +If you need to scala log down Id suggest on the client side and and on the server side +Id like it back if possible Ive recently been thoroughly digging around inside the messaging code so Ill look at what it will take to bring this back +Can you suggest to me a file to be tested to reproduce your issueDo you know if the not working samples are georectified or if they only contain Ground Control Points +Otherwise existing data is allowed to retain the current namesSo recreating schema from scratch will fail everywhere but upgrading existing schema should always be fine on Linux. +keep in mind that up until now Railo was not compatible with ACF on this issue so the proposed fix will bring us to compatibility of at least of the cases +Either part of or dependent upon I do not plan support for that +But I wonder if its easier for use to remove the custom headersbullets from our CSS file +I reviewed the available enhancment engines and would suggest to use the following distinction in stable unstableStableUnknownUnstable only one test for english +Filed and linkedIf the winutils fix by itself is sufficient to resolve this test failure I will resolve it as a duplicate +I have to look into the code in daytrader +not matches is working as expected. +You will see it in the next site deploy +Otherwise we wont be able to delete one with nodetool +Closing this since it was fixed by +Raghu if you want largefile support on your box you need to compile with the CFLAG DFILEOFFSETBITS +I will fix these and then commit to Apache +We define our specific dependencies via the pom +Thanks Lahiru for the patch. +But youre right many of these could be handled as wellAnd I agree that they should be handled to be able to claim to be doing a complete jobSo far Ive claimed that this class handles Latin and Latin Extended A which should cover most if not all europeanturkish languages using latin script and thus goes much farther than the in that respect +Please review +The patch failed core unit testsThere are other unit tests failing +Might even make sense for it to be a soft commit +regards too find its strange that they fire the event before the updatemodel phase +I just committed this +Consider how this would affect distributed builds where you may need to define a local repository cDocuments and Settings +Unit test have been added to demonstrate dedup of search results by shard order +Working fine +bulk defer open issues to defer to not essential to address for +The has a set method. +I think the code is so we can make sure by looking at it +Obviously this is not ideal but there have been quite a few cases where rd party javaagents have had this problem and it would be better if they just worked +Confirmed! Error happens with fop seems fixed in. +Address above review comments in v patch +Fixed by SVN revision +This is no longer an issue since there are no specifically for Java any more. +This is actually an EAP restriction and should be addressed there +Did not see an increase in speed though with respect to the old versions +Sample request +Verified with JBDS L +Now I see it in the editing screen +With those two options together youll have a bit of a performance impact but not awful and youll avoid data loss on complete power failure +Bill let me know how this works out for you + I very often end up moving the generated code elsewhere as part of a build d prefer for this to be separate from o if at all possible +Mark is it still enabled the same way as Joels original patches +From this information we will write the final release notes for all issues +Youre not the only one Plugging this into the CLFs is going to need some changes not least because I dont want to have a compile time dep on jboss logging +BryanCan I know for what sql commands this readPage and writePage methods in class are calledthanks BryanI put breakpoints at the beginning of the pageRead and pageWrite methods and execute those commands but it did not hit +srcorgjbosscache +What do you mean there is no intention to support Java? Spring Batch still supports a Java runtime right +RC should sink +same happened here on with a mysql +It works for me but someone need to check why that code was created in the first place +For example if the secondary master is paused sometimes we want to know this before it becomes the primary master +It can be reinstated if they are provided before the release. +I had to make it so the dist cache always was trying to delete things and it never actually deleted them I also had to modify it so that all dist cache entries take a long time to download and on top of that I had to hit it just right so that a job was finishing at the same time a dist cache entry was being deleted +What Id not like is different for secure and casesAgree idally the code path should be the sameInstead of threads this could also be long running Localizer processes if that makes the implementation easier +On a second thought Hadoop does not have a story for Writable backward compatibility yet + +Yes its affected in git as well +I should have a patch up shortly. +attach which change the load order load firstly keep it same as the scenario container +bq +Also if doesnt go in we will need to support flag in as it is going away in favour of flagI am just trying to see if we should remove the documentation of this feature for release and get the complete feature in for +To turn off aggregation you need to have w as command line option or set the property to false +I have to confess I dont drink alcohol anymore +TobiasCould you please comment on these changes and explain us why you think theywarrant inclusion into proper? Are there any particular speccompliance problems in s implementation of MultipartFormdata thatthey fixId like to see wirelogs of the request that poses a problem the request with your patch appliedOrtwin Glck +When mentioning maven meant to write the specified maven repo or the dependency groupartifactversion id is incorrect the compilation will fail instead of the specified maven repo or the dependency versionartifactrelease id is incorrect the compilation will fail +Added units to the logging Did not change to using +going with cannot reproduce for the moment +Fix is in queue +Why dont use Excalibur Sources for retrieving the contents? +Reams of rules exceptions to rules strange corner cases unexpected consequences +The file that led to this crash is the usual renamed to to avoid problems with the requesting do you have that file? Its around MB if you need it I can upload it it seems the above is happeninig only if the OS is in a low memory situation that is if the test is started with less than GB of free memory with GS already up and whatnot +i tried build on clean repo and its working fine +Thanks stack +Can we consider this patch and for the branch as well? If so I can submit patches against the branch +testDuplicates relied on the strange globbing behavior that existed prior to +Closing as discussed with Pete. +Also allows you to keep your infrastructure topology out side of application configuration +put in the homeofbiz am not able to give Apache rights there is nothing in the attach file dialog +And you are right the binde method should postpone actualy binding until after the component has been activated +from this end +I am really back to thinking this needs to point to external repos by default +If you were able to trigger it in a version before your patch Id say reclose the issue N +Im not sure what you think the behaviour should beThe set sets an upper bound on the number of calls that can be active on the stack but it doesnt cause the JVM to allocate more resources for the call stack you have to use mechanisms to do that +Release published closing. +still kicks in with sourceNote that the problem is beyond just the +Feature added in revision a object exists and it may be used to create rule bases with configurable behavior such as Assert behavior either identity or equals based Logical assert behavior either identity or equals based +This will be revised for addressed in Maven The Definitive Guide +Were also getting blocked by this bug as it effects the thrift gem +That sounds reasonable +Please be sure to also have this be part of the quickstarts. +Hive does not offer row level issue is fixed and released as part of release +Normally though youd expect to find it on slaves +Ill have a look at this as soon as I can +Logged In YES useridI am closing this upgrading the Linux kernal seems to have fixed the problem +Bob can you take a look at this +The former includes snapshot overhead the latter does not +runs the test in all Surefire executions +Ill commit this in a day or two if there are no objection +Only line needs to change +any help on this will be greatly appreciated as we are looking at moving to the new version of axis and this one problem is preventing us from doing so. +Reopening +See instructions you need to provide a bundlePlease next time put all the bundles you want to upload in only one jira it s easier for swixml site has no official bundle but I provided the URL of the closest thing to it +Three more required libs for reading configuration files +pushed to all streams on the strength of maxs verbal agreement during +If you check targetworkconfiguration file for test with error from subject I am sure you see where those plugins come from +This issue is specific for a problem in regarding the design around the which results in a hard reference to a from Weld +It no longer does +Closed. +Note also that the name is not +Thanks for the out when there was some german junk characters is encountered while deserializeThis was due to data problem and we fixed this close +I know we talked about this on IRC but where would an ordinal come into play for an extension? You wont have the same extension loaded twice +Is there anything I can do to help make this happen +I only did top level folders +Done all pruned out +Well actually I found we did unmarshal correctly however we can not get the from the what we can get is the ReaderThe problem is caused by the it check the charset before return the we seems always put the as as a default charsetIn order to fix this we have to unset the charset in mtom enabled deep investigation I found its nothing to do the the the reason actually is the Handlers in the chain consume the and then the is just goneWe have to cache it somehow +Reopening this bug as it still shows up in FUSE ESB which internally uses AMQ using the test case attached to this JIRA +Nigel we already have a test in that validates output of a command like does +A redirect via should work in ALL casesIf the specification does not allow this could this be fixed via +Please review + Fix file handle leak in +This is one of the reason why W rules +For a while we saw this type of error regularly but I have not seen it for quite a while. +Can you test if the ODBC code is compatible with the new thrift version +HiIt works fine with FirefoxBut it doesnt work with IE or IE +This jira is no longer a for the update Amol +should this be handled for or can it wait for +Yes! thats what i just found during the professionnalpersonnal system JDK but Tomcat version +The start time is set to zeroMarking this as major because removal of killed jobs in Capacity scheduler depends on correctness of the +I committed a patch which fixed this behavior +I met server error when I tried to attached the patch file +Can you add it here +Fixed in rev + was corrected with obviated by obviated by Compound issue value attribute usage obviated by onclick attribute should be addressed Compound issue value attribute usage obviated by direction attribute should be addressed This description is Ive reviewed the example description as per item and can see no problems +Resolved Fixed +Is there a workaround for this bugCant believe there is so little interest in this bug +is this easy to reproduce +testdeployvmmultipleaffgrps and testdeployvmbyaffgrpnameandid failed due to previous tests testdeployvmwithoutaffgrp and testdeployvmmultipleaffgrpsRemoving this dependency is another task which I will be working on along with Prasanna +I am curious about things Why did the error change? Is this error too specific regard with suggesting the user delete the directory? Perhaps this is a directory that just happens to exist with some other contents and the application just wants to warn that the directory exists so the user can put their database elsewhereThanksKathey +pass down the down to the filter factory when creating the bloom filteragainst we need to reschedule it to the instead because we have added murmur in trunk is that right +Attaching patch which does the job of setting the correct collation type and derivation for aggregates +Bogus issue + +I will look into moving the logic out of into +Take a look at +Are we happy to close this issue +mail changes made +the adapted directory was just used locally and is not deployed in the test environment +I will add some additional validation around that field +Spec not mention about viewstate handling for redirect +Is that configuration needed if its defined in the components themselves? +Thanks +I did wonder though if we should have a section on what happens if you try to read and system cant give you known correct data +Yeah lets see what the status is today and document this +I think Willem created it like that from the start +We should not make changes anymore with this issue but rather create a new one +The patch contains the fix to the particular patch is now updated to include a test recent patch now introduces conflicts +and the set to Resolved Fixed again to correct resolution which was lost in Bugzilla JIRA conversion +Talked offline with Daryn + is committed to trunk patch is created for on JBDS +This would substantially simplify tracking their progress +Fixed in trunk +LGTM +Hey folks +Closing upstream project fixed license. +The linked JIRA is resolved with jUDDI which is incorporated into + +If turning on security in config has no meaning to client we should at least update the documentation +Patch to fix AlbertThis issue is already fixed in Roo trunk since has raised the same issue +i am raising the Version to and changing the Componentfrom to Xalan because we see it beyond the useI have a set of testcases to add. +But wouldnt that also stop from logging? I think Im just going to end up writing a script for it +I created repaired with the correct were some troubles with multiple same ids in also application misses all these modifications example shows the welcome page but after registering new user application hang up and no error and result is printed +Oh I meant they are different methods internally in +So if the user constructs the blacklist the old way you will store that config and return it if he calls getBlocked +Seems like that must be the case although maybe something should be done to check for this and handle it +The patch break +Does or does not project work with last meclipse release? Please dont mix several issues in this bug report. +I think we should also rename the ujax prefix to sling for the request parametersBtw the sling client library is one of the greatest things on earth so it deserves the name catapult +but other then that it works great +We just need async functionality to test this + backport cancat +Thanks! +Itll be too slow +Patch that addresses the review failures are unrelated to this patch needs to be regenerated following the project split +I agree with your point I dont mind breaking backward compatibility either +python Fix mixing of oneway and regular requests in? I think having the language at the beginning makes it easier to skim logs +I was unable to replicate at first because i was a web project under the server +Closing issues. +bq +It looks like the issue went away with one of the recent changes +I think the OSGI deployment is mostly done and Im still think that you will understand why I did this once I can show you how it work in action +Bug has been marked as a duplicate of this bug +network issues is that ZK session timeout actually takes place so recovery is delayed by s +I couldnt find any problem so far and therefore I think this issue can be you ask Randall if he might be able to drop out a containing the current fixes? I would love to finally use modeshape in our app but cant really depend on a self compiled snapshot version +Id like to keep this issue open for a while Im going to test how it works with the new infra +Removed +So simply checking to see if expected files exist might not be sufficient and Im hoping theres a more useful solution than a preflight script that calls java version just to trigger this behavior +However I can not completely deny that something I am including is leading to this +We should investigate for +Ill be back likely in hrs later +It turns out that person is away but may check email +Having the source code is a requirement for an open source project and downloading a binary jar from some webserver is not an any Debian or Fedora distribution efforts are now using so the issue wit radeox is no longer with us. +fixed in +Carlo please let me know if you want this assigned outside the EJB team youve got a lot going on +The changes have been applied to branch remotingx and the unit test passes on cruisecontrol. +Stay tuned +Integrated to trunkThanks for the patch Sergey +Hi TorbenThe other day I also ran into this issue and applied a fix +On we got a thru put of where as on +It certainly helps to remove error logs +I think we shouldnt be creating Namesystem bean in the second constructor +xml but your test cases need to ignore those +Do you need any info from QA for this to fix +Marking as invalid based upon Mr +Give it a try +dotnet file is correct +Heres the result from forrest debug +It seems that upgrading to struts has fixed this issue so as of revision this should be working properly +Apply in class library rootIMPORTANT NOT TESTED +Integrated to Thanks for the patch ScreenivasuluThanks for the reviews +By default Cocoon is bundled with DOCTYPECocoon does not require Servlet it requires Servlet or higherPlease close this bug if my suggestion helps +So in this is fixed correct? you like to backport to and rightSo lets close this one then +Attached is the patch file for this change +srcorgapachecassandrahadoop Other than that looks good to me +As this is not standard behavior if someone needs this functionality they can easily add their own without impacting the rest of the I am linking the two because while this is not an exact duplicate of the mentioned JIRA it is very close +In Branch g +I think is is not related to the refactoring since it only moved the code to packages but not changing the logic +Ill take a look at the change and put it in and if it looks good. +VenkateshThe patch is not applying cleanly on trunk +This assumes that the next release is intended to be perhaps its betterto name it and then if the changes are not simply of patch nature thencall it +inputrc is a bash thing and youre rebinding up and down to bash commands and +Added a test case to verify this behavior as well +Given that there is a scricp most folks patches are managed automatically so in a sense the patches do not really break Im not sure what youre arguingThis patch will break every srcjava patch file no? Any patch generated before this is comitted will no longer apply after this is committed +What are the requirements these days? Which version of hadoop should the patch target? Ahad +So lets not close this +I didnt figure it warranted its own patch but I should have said it was attached +So please try this service combination where LB service is provided by +Why fork outside of ASF if we can keep it inside? Is a independent project justified? It seems to me there is a lot of infrastructure that needs to be duplicated and maintainedI agreed to starting a fork outside of ASF because I didnt think there was any possibility to bring code into +In any case Jay it boils down to personal preference so pick whichever way you like +There shouldnt be complexGateway in BPMN anyway +Attaching the rebased patch +would you please try my patch? thanks a lot +Ill add something to the migration page in the wiki next. +So not a bug +It has been showing to run for more than hours now +I have added tests where the given AT is injected without any problemCan you provide a unit test that demonstrates the problem +Thoughts +Added an openstack connection class test +Should i add and set to false +added just a default value +Can you please post the exact bean classes code or the attach the application? Just want to make sure that this issue is fixed correctly +Hold on Im investigating on a performance degradation in some benchmarks +bq +In Step you kill the app then relaunch but calling any previously authenticated calls dont work because the cookie was not saved +Thanks for review SidVinod +Having the master master as a Codehaus projectrepository makes it trivial to get artefacts into the Maven repository something not entirely easy from +Not a very good result when doing this see the log been about months since this issue was posted +As of this date releaseprepare appears not to run the install phase +removes the compile time type section from the topic +If it is going to commit it breaks +Default is true in other words in order to change the current behaviour set the value of the new property to false in a custom file +xsl is only allowed inside an xsl or an xsl +There may be challenges in how to trigger Bloodhound to update without direct but there should be a number of ways to do the jobDoes the NFS solution look good to Infra +Those may not work so well if people are doing huge inserts and the like +Ive essentially dealt with this by reading system properties specified in the test +This was fixed on Dec th +Patch for the issue +Patch replaces va +Stopping to investigate this because I cannot reproduce the same issue on the released just a patch that wasnt ported over? Thought Im not sure if this is or +Reopen for clarification of the term optional feature +This patch has your my Avro changes Gabriels dynamic initialization checks for Avro Matthias dependency cleanups some additional exclusions I added to make our avro stuff map to the same exclusions uses and some minor dependency scope changes +Dashes or dots in the translet class name are translatedinto underbars +Commit changes to trunk At revision and At revision I think we need to do more console refactoring work maybe in +Producer send always set TTL so no matter what you set in the Message before calling send the value will not stick +Alex could you review the changes please +Ah cool thanksCan this issue be resolved or closed now +KB Time ms +As you pointed out in the case of MSC this field is already set during the annotationxml parsing time +As the current engine does not use inference this will avoid clashes between what is stated by other systems and what is stated locallyNevertheless this does not seem like a good idea longer term +I mean mostly for backwards compatibility +Can we resolve as a duplicate then given that this patch contains a fix for that issue +Thanks Anton Ill close this bug. +for the latter +Yes please commit +patches applied +Rev introduces the mvcinterceptors element again reflecting changes in the latest Spring build snapshot. +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Patch rebased with trunk +Will have to start from scratch on it +Click NextERROR Dialog with no widget is displayed with error No platform is selected +Latest crash log from jbdevstudio have the same problem with Fedora there a confirmed workaround for this +Thanks +This is broken due to rev not being applied to the release +No response assuming ok. +Testing will occur on IE and IEAnyway our policy is to avoid or reduce as much as possible browser specific tricks +Bulk move issues to and +the implementation effort documentation effort potential new bugs ugliness of that approach in general etc +Im adding reasonable documentation too now +Tested and it worked +Submitting patch for fix to revision . +Spring Security is built against Spring so obviously if you want to use it in a maven build with Spring you will have to add the necessary exclusions +The patch looks fine to me +I do +adding at least one to match the exactly one of the other step and thus distinguishing the two steps +The artifacts and proper are also now up on the central Maven repositoryUpdated issue summary to upgrade to Lucene instead of +reopening for release note setting this to Not Required for the release notes for EAP +Ive attempted to reproduce it here and havent been able to +why? I can do s of thousands on RI +I think someone on mailing list or the build forum might know +It looks reasonable to me but someone should probably run tests on Windows before it is committed dont you thinkIf it is the simpler version of the patch it would be good run some tests on windows +Its not clear what we can do about this until OASIS actually release something +Couldnt reproduce the problem +Hi SteveI was looking at adding a set of base RPC services to the via thrift for exactly this reason +for this +Opened to track the custom partitioner multiquery optimization +Improved field descriptions in rev +Fix was to share cluster variable with its subclasses +The message ERROR XJ Conclusin del sistema Derbyis normalNot sure why dblook cant connect it works for me +failed but the test is not related +changed to use addPackage with like youve volunteered yourself to also make a test case for anonymous inner classes +Thanks for the review +r +pushed +Row spanning is implemented +with JBT. +Removed example and resubmitted +If you are measuring idle connections from the database it is possible that the number of idle connections from the database perspective will exceed the number of idle connections in the pool +You can now unexclude the test. +Will need to investigate thatfurtherDavid +corrected URL to point to article mentioning erroneous mixinOwner +in both testthriftserver and testthriftrandomserver +Third version of patch attached +Comments suggestions for better ways to do this welcomeLeaving open as we need to update the user guide to complete this +The error seems to be the result of someone in the Eclipse project deciding that backward compatibility is a bad idea and deleting instead of just deprecating it +Note I didmvn clean install in the wrapping folder first and that worked fine +I have a patch now that is installing the an and I am testing it simply by setting the heap size small and seeing what happensFor the second one I dont think the point is to try and prevent an OOM form happening +svn ciSending containersrcmainjavaorgapacheopenejbutilSending containersrcmainjavaorgapacheopenejbloaderTransmitting file dataCommitted revision . +Gerrit Sakthivel most likely has a corrupted edits log and his immediate need is get support on how to recover from his particular calamity +No I am using the same old accountStack I will have a look at it +subsumed by +Added reason to the end of the warning line to give a hint at why its failing + information is required in the for this to work + on Linux b +Thanks Denny! +btw I if youre ok with the approach I can propose a patch on +generate aggregate bundle via instruction + +Commited patch with revision Can you confirm that this fixes the issue +Ill build asmall test case we can put in the test suite +Scott Id really like to get this committed in the next week so that we can get a release out before the end of the yearWhat can I do to help? Do you want me to try to debug the tests that are failing after this is applied +Sorry Stefan was just trying to suggest in this ticket to completely remove all javascript from the WEB MVC JSP View Addon +A pull request has been attached +Whos working on this suddenlyThe patch that just went through seems to have major flaws +I tested and it applies cleanly for me +This issue is not urgent so Ill schedule this issue later again +sorry forgot to include my changes to +reopening as we need to rebuild +Closing based on Vincents comment that the cause was aggregator which has since been removed. +its been around forever and nobody seems to want toknow how to tackle it. +We need to make sure this is somewhere very obvious in the migration guide +So maybe the sort avoidance code is confusing the two columns for some reason +All of the changes are to test code and Jenkins gave it a so I give it a too +Thank you I did not know this paper +Patch for hadoop added. +Create issue for that in UNDERTOW jira also if you feel like it you can send PR with fix +Resolving +Can you update the documentation saying the array is unordered? Also patch needs to be re based to trunk +We need to force the Distribution build to happen after the core and extensionsant builds but somehow Maven keeps putting it before the extensionsant build +Today I wrote a test method for the and put that in to the +Bloody hell +Thanks +See the configs that take suffix patch fixes the generic conf code to handle NN as well as Nameservice. +Works fine. +release bulk close +If I can do it in one wizard why would we want to add levels of difficulty? I think if we fixed the issue where you have to expand everything in the java class in the wizard we should do that by default +Clint did you commit this in to the trunkNo its a patch that affects the core and has ramifications for all view layers so its being reviewed by more senior Seam committers +Not sure that module really has a designated maintainer so I dont think we need to wait on a review for anyone +has been committed +The one with the least score will be most likely not belong to the complement class +To me device parity is everything with its main purpose in fact +Alejandro was looking at whether needs changing in light of. +Found a small bug in the previous patch +bulk defer issues to +Change applied o everything is file Ill create an for stablebranch jobs and will get all stablebranch jobs to use it +Based on the vote i move it to release after the mojo parent version is out ill do that for instead of in r. +Hellothe committer accounts still need to be created +Thanks for the quick response Garvin! +Can you apply a link to Platform beta where I can download it? Because I can not reproduce warnings +If you try it let us know how it goes +Sorry for cleaning old attached files I wont do that next timeIll submit a new patch with doc functions in correct order +Those are the options yes +The Hive JDBC driver JAR currently depends on most of the other Hive +Closing this as a duplicate of +BTW I added the files loading to. +Affects scriptingjsp because the just assigns application to all files created and affactes commonsmime because the +Do you see an abnormal behavior with the existing code +btw its worth mentioning that this whole concept of index back compat wrt Unicode is already completely broken in Luceneif someone upgrades to lucene and uses java Character +Many thanks for the fast responses guys! +Setting it to Improvement as nobody perceive the issues mentioned as bugs +Can you upload those separately? should go in datafiles new version of the last patch that does not have the entry for the binary file +This patch improves the translatability of messages produced by the Derby tools +Thanks for the feedback ChunhuiI integrated double exception patch to and TRUNKThanks for initial patch ChunhuiThanks for the review Stack +All tests pass for me +Could you try it out? Id expect the error you saw to be very consistent so if its fixed it should be obvious +QE please check if the issue is still reproducible +does some related cleanup including moving the mbean method to +The original patch didnt apply cleanly +I have uploaded an example application that reproduces thet broken on trunk but added test anyway. +If you finish it before we release beta please correct the version +Ill fix that add a test and upload a new patch +Yes I suspected as much +Thanks +I will this once productized zips are available +Solved + is crashing my Eclipse on Mac OSX +Even if the leader is not alive it should depend on the controller to send it another to connect to some other leader or to become a leader itself +The solution is a long one and some more goals can be extracted out of it to make it cleanerMeanwhile please test and review the patch and provide feedbackThe patch has to be applied to +You cant add ssl or tls in front of the hostname due to the way streamselect works with +Damn +batch transition resolvedfixed bugs to closedfixed +At s suggestion onI made a patch for using Welfords method to calc the variances +This was regarding EAP with as JMS providerWhen configured so there was am empty node JMS Destinations in the nav tree and the destinations were shown in other nodeThis might be fixed now +I can confirm the issue +As far I see there no version to include this feature +Thanks +Changes constant evaluation to a long +As you said its possible only with debug info or th party like paranamer otherwise parameter names are not included in byte codeI could add a little fix for classes with debug info but It didnt seem to me a good thing to doYoure absolutely right about the error message Ill make it fixed ASAP. +Can you watch out the download page and then try it outBen +Thanks! +therefore an exception +Jimmy I think this is the biggest patch ever applied to +Attachted is a somewhat crude but effective Feb is replaced by Feb and after parsing the day is incremented +When these tests run the date Wed Nov EST is transformed as and hence comparison fails +Not sure of the fix looked at the patch at and it doesnt seem like the is exposed to the outside world anymore +This issue definitely relates to Clover so moved to correct project +Changed the title +No response assuming ok. +mike I committed your patch to the branch so we can iterateI just committed some fixes and added javadoc to several classes +The maximum chunk I tried was values +Taking dependencies into account causes the problems reported at +I havent kept up with the accumulo test conversions to know if that affected this but I believe I provided versions for trunk which applied to as well and the version that worked at the time +can someone review this please +I agree that the jar file name prefix would be better than a counter but only for colliding namesBetter yet the shortname? It wouldnt take much to sniff the first few kbs of the tld looking for the shortname +HiAs said by Richard iPOJO already supports private services +Thanks for your comments Claudio +Greg I have yet to encounter ANYONE who doesnt consider this a bug because all cited reference EXPECT concatenated files to work because the work in ALL OTHER cited instances including gnu tools web browsers etc +I moved data around and forgot to put the Apache content back in place +Introduce a configuration option to supportdisable hdfs appends +This issue is further complicated for jobs with speculative reduce tasks and reduce task failures since it can not be known at the time of the lost tracker event whether all the reduce tasks have copied their data yetThis patch addresses the symptom of map tasks running long after the reduce tasks are finished by allowing the restart of the map task +The SCM revision is a good addition but the build number is still useful +Constraint enabled +You cant expect a to connect to a remote FTP server traverse a directory structure download the desired file etc +Missed one file in previous patch +I guess the same will happen on the as well +also failed because of this please look at that test during the verification process on this JIRA +Mina sandbox version +fixed in be precise encoding problems appear only on installfile task not installinstall which is AFAIK the most usedthen normal Maven users who would use encoding in their when swithcing to Maven wont have any problem even with plugin version +The fix in is not in the trunk and there was some recent discussion onwhat might be the same bug so Im reopening this + +Attached patch for this feature +Doh +Issue is not reproducible on samplestool project +Its a very small job there are more fundamental improvements to be had for the same amount of effort I think +The dtdschema registration should really move out of the common project and be configurable +It blocks running EUT jdtdebug suite because of too many hung java processes hold system resources and this make testing almost impossible +Patch is missing a test case as well as the other one +Im trying to get the tests going + +It seems that the changes are not needed in your patch +Please reopen if you can provide a testcase +I fixed some problems I introduced in r +The security model for one is wildly incompatible with the rest of Hadoop securityIf the main reason for wanting JMX is so that other software can call these without the Hadoop jar Ill counter and say we should go even farther and allow other software to not even require a JVM +But this issue was fixed in JBT CR +Thanks Matt +Thanks for the tipcheers thats good +Im working on the patch too and it has some refactorings which may make merging two patches difficult +I dont like reshard so much it implies the data has been previously sharded +Are you still experiencing a similar problem with? Or were you just asking +as a duplicated of +I took a look at source and youre right +GUMP is no longer broken because of this +Restored ability to set columns on +If you use Twitter and you go and try and share a photo thats the same app that we use when we go and get a photo +If it compiles against the source then I think we should see the problem which is in my understanding the reason why we see it before calling clean +fixed and available on codehaus update site as xdt +I believe there is no way to fix this anymore +Hi Ziv can you try modifying the values of javacommonconfI belive that theres a conflict between and externalizetruecorerpcWith this setting the corerpc libraries load first and then the gadget specific features are loaded afterwards +Given the low priority huge amount of work risk of compatibility break and lack of patch this issue will not be fixedIf you really want it to be fixed please reopen it and provide a patch +if you do a bunch of stream copying first and then use transfertransferTo to a different file not the same one rightbq +Proposed patch forThis isnt substantially different code from what is in master + +The ability to call up a numerical key pad is a fundamental user interface feature on mobile devices +I was able to get the war to deploy +Dear MichaelFor the purpose of my study I just want to confirm if you have considered the four mentioned warnings relevantRegardsMarco Tulio +In cases like this would encourage in the addition of optional parameters to parent resource add methods to support creation of child resources +is changed to take into account the changes is now optimized to reduce synchronization contentions +This is the first time I ever hear about this +Have you verified that this is still a problem in M? Remy thinks it might be m getting close to checking in +Peter sorry for not appllying the patch earlier now it is done. +Im on a project in which were trying to integerate with iBATIS +Well in the short term we can open the palette in a particular state if that helps making certain drawers open or closed by default +false? Thats the default anyway +Other than creating this copypastesearchreplace version from Struts I dont think Ill be doing anything else with itI guess its like Gump it wasnt created here so no one wants to have to go off and learn something new just to maintain duplicate information for someone else +Ramification of maybe +Michael care to elaborate what is this about +Actually its not wrong +Im going to fix it after some discussion with the gt developers since that will invalidate a lot of existing SLD around +Ive updated the branch +So Ill close this as duplicate +Everything I came up with would almost certainly have resulted in a flaky test one way or another +I see +This looks like an issue with jQuery Mobile and the Android Browser since there is clearly some style elements added with that +I ran and it passed +Otherwise test will hang until this case splitting wount help to avoid timeouts +Deven the patch was applied to trunk at r please verify +Please create a jira and assign it to Sohil +bq +Anybody have access to the old tracker and can look up Beetle +I am about to cry +I guess failing is ok if its documented in the class javadoc +Thanks for the report please let the developer set the fix version +Investigating +bq +Forgot to include issue in commit comment +clearly explained with javadocWe need something along the lines of you cannot count on this to cover the whole index unless you control the use of your application to ensure that its possible that this will only correspond to one of multiple being used in the context of one large index +Moving to there will be no additional work on Ruby Fibers in this cycle + +Another general solution is to run itself on a separate federated NS and only balance the non NS which is supported today +The patch is complete and ready for regression testing +This issue was caused by a bug in the h database used for the jBPM service +Thanks Aleksander. +As per Sateesh commentit is clear that this issue is dependent on so closing it +Halley Pacheco de Oliveira has an ICLA coming in +I should have put the link to the review in the JIRA +Justin? Anything +yes its working fine +Couple of things can you pls port this to trunk and a unit test? Thanks +What bothers me is that other than logging it to the clienttrace on behalf of a possible downstream tool the parameter serves no functional purpose to mapreduce it is likely to be optimized away +I think with the patch in place the handover handling from the async queue can be removed as well as the new version ensures that the sync event is processed immediately +Thanks +Please ignore the last patch and use this +Fix committed to branches and the trunk +Yes I do need this on I tried to merge but it gave me this svn REPORT request failed on reposasfsvnvccdefault svn cannot replace a directory from withinAny suggestions +Committed to trunk and +If there arent any objections Ill be happy to commit itThanks Mike +I have a standalone repro of the problem which demonstrates the problem with local XA transaction +from Sandbox to SCXML +Juergen +Patch to reset the created local variable so write locks are not unnecessarily acquired +Please reopen if Im missing something. +too +Ive just tested and yes it works with appears to have issues with the new in scaffolding templates but that is a issue now one with core +Well post a more complete patch in the near futureExcellent +This is fixed by dims a while back ago +But Greks and my question is why do you iterate the whole callstack? Wouldnt it be enough to pass parameters from the parent only +Ugh thats instead +does seek to the next column it is interested in so even with many column this patch would make it worse +Fortunately i only screwed this up on trunk +And you also know the workaround +Manually tested! +I think I understand the scenario but is there any chance you could create a test case that reproduces the issueIll look in the code to see if I can find a bug related to workspaces in the reference file +Closing this issue. + for patch +However the method uses the view for several things when constructing the iFrame URL returned in the metadata request +Ill create a new patch +The PCLASS is not used as a condition in where clause in the generated SQL which obviously will cause problem when the POID for P and P are of the same value in Cs table +Screenshots and code snippets are checked and corrected for ajaxValidator graphValidator attributes are added to the components plz check that all of them have descriptions in the table of attributes if not add it +Committed +Attached the patch file +We should upgrade to at least but will defer this to given time constraints +Updated patch for +fixed in r but needs this issue is only a problem when using the for blobs on windows boxes i added a that opens the file not until the first file access is madefixed rev +Should be fixed by said its fixed so +I needed to have an X display environment around +ok I see I was sloppy and neglected to update the deploytool script to remove the augmenting of UIMACLASSPATH +That is a partial workaround but you still need to take carewhen preparing to commit that the are appropriatefor your OS and that the properties are set +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +the jdk is as the issue describes too +We cant do anything here for nowEach component library handle this case in a different way +Is it possible to just make sure everything is rooted under? Then we may not need to worry about how deep the home dir is in the directory structure +Thanks Dyre +Transitioning to closed. +I didnt verify fix in because the currently dont actually compile for me +Im concerned that this landed on trunk without a review once youd addressed Paul concerns +I see sorry maybe we can have logic in to infer the NH connection providerdriver from springs dbprovider as this would reduce the amount of configuration +Thanks for the info Brian +Changes committed +It does work for us +His argument is that hostnames are case insensitive and different may have the host with different case and this could create shadow groups if a process starts getting a different case DNS name +I made the language specific to each task +So it simply takes +When removing a datanode reset its scheduled invalidate blocksThere is also recent +Patch isnt applying cleanly +Im going to go with the assumption that this is not going to be done +Hmm +Some of these may be marked required using the new config introduced by this patch +I spent an hour going through the code and I just cant figure out where it loads in the config files +Lets tease them out for a bit +I rarely get a merge in this scenario just a list containing the right amount of views with generally one member with the duff name for exactly the reason you describe +Postponed to as any PERF issue +Committed patch a which improves the inaccurate comment pointed out by Dag to trunk with revision +I confirm that your address is still subscribed to both the user and dev list. +MichaelCould you please provide me with a copy of the version which you were able to get working withEven after playing with the EJB ref links configuration in multiple ways it still does not work with +So this ticket can be closed +Reverted the commit and put the sync one layer further out as suggestedAvoiding a deadlock or weird behaviour was going to be hard +This patch fixes the issue described and modifies a test to verify that it is working +Online resources also claim the same +Tez is designed to have a client side install +Marking resolved the maven files should now be available +I am going to sleep on this but I expect to essentially be backing out this fix tomorrow. +Will commitlater after I have banged it around a bit. +I have my Filter on because when I save the page I have everything rendered only in the sneak preview its not +This way there is less chance of corruption which is more important in these whitespacenewline sort of problemsI cant use your JSP it isnt a complete stylesheet it is a program which can be used to generate a stylesheet but Im missing all of the imports to your JSP and Im betting that they are proprietary for you company +Fixed +OK moving back to assigning to you Uwe +Yes its in lower case and it doesnt work +No Alan +Michael Im not gonna be able to work on this at all until monday at least thats what I am sayingSo I will check with you Monday and see how things are +We still got the same wrong messageThis issue is too troublesome +Here is a patch which breaks the materials out of into the separate file and adds the required text to LICENSE and NOTICE +When any entity outside isnt valid console cant be expanded and there is an error +This issue applies to codebase prior it actually fixed in? Im asking because issue is still unresolved +It sounds like the message that you are referencing is throw when the application is started + +I believe the costingshould be the same in this case whether you use ? or inlined values +Added unit tests that produce errors for getslice whena called with nonzero startb when data needs to be fetched from ssTables and ideaInstead of passing in a the getRow call resolves the entire row drops the first offset cols picks the next count cols and returns +Agree the owner should be the person executing the CREATE TABLE LIKE command +We consider some general reworking of the mail attachment support however itself is going to remain as it +package +from me +I am wondering whether a more complicated reaction is desired such as recirculating the event after a timeout +Yes Raghus comments make sense +Can you always assume that the predicate to be filtered will always be after the tablescan? Currently all optimizations are independent of each other and can be turned off this way you are dependent on a new feature may have stability issues +The literals would be as part of BOOLEAN +updated +All looks good +So we are not ruling out that there is not going to another system like model in the futureWhich should just mean another attribute for that model +The qpidxarm module is loaded by the MSDTC process via and the subsequent use of statically linked qpidclientboost falls squarely into undefined behaviour territory +I cant reproduce this in svn HEAD +Please open a support case through the support portal +it looks like due to http being async the bridge does not get torn down when the remote broker shutsdown as is the case with a tcp connection +So though there will be a place for JZOS and applications I think you should need to justify why not do it on LinuxWindows +Fix applied in trunk and branch. +This really is due to them changing the version of JAXB in Java from to something newer and the newer version generates different code +I verified that and instances are removed properly upon conversation timeout +I cringed when I saw us extending because the docs are quite clear but if nothings broken +Hi Wannemacher I have sxhead and i am using ajax in my page +Thanks Amareshwari! +No issuesAny chance to add some test scenarios to validate that number of tables is following as expected + +A generic class should parse out D conf fs etc +reopen to fix version +Any reason not to fix +Please use this updated one +I tried a simple test and I am not able to reproduce the problem on Suse +Also well have to add a few jars in to wssj Shall I add the +Ill have a look at to see how it impacts this patchRegards to version task this still an issue? +That means the only variables in the policy file now are the comments which explain how to customize the file +Fixed in revision . +ve fixed this in trunk of the Spring Maintenance tested this fix with Websphere and rollback now works properly +Added regression test for the fix described in +! for patch +add command line to to Priority. +You can change the vm arguments to a larger value like vmargs Xmxm mvargs can do this in your editor the resulting applicaiton will have these settings in the if I remember correctly +Please reopen if you want to work on this someday. +SLA Breach not calculating time right and SLA Breach using private events when calculating +Also hsqldb was not upgraded in trunk where the tests are failing as well with the same error +Resubmitting the patch with the given comments fixed +I merged to for inclusion in the release +gonna start a this be a docs ticket with Subtasks +Thanks a lot Marek! Really appreciate you and still taking the time to support Seam . +Igor do you mind looking into this one please? Cheers + +That was super fast +I dont think we have any at this point +Made a patch for the mentioned there is no difference indeed +It probably is but it would be harmless if you could make progress on a retry +Do you think it is possible us to have a better error message hereObviously the two points dont change the fact that the issue is resolved but I was curious if you had any thoughts about my questions? +Providing a fully featured groovy shell will take a significant amount of groovy console is not available +So the solution would be to only let Maven handle the resourcesSuch an approach would work but it is a little more complicated than that since bnd will only copy resources into the resulting JAR file if they are in a private or exported package +Avner would you please review the documentation for correctness? ArunAlejandro and I feel comfortable with the current patch as we explained abovePlease let me know if you still prefer that Ill create a patch without passing the entire toI have no problem with any interface that is accepted by you and AlejandroThanks Arun and Alejandro and anyone elseI would like to revive the patch in this JIRA issue and submit it to the updatedThis patch was already reviewed by Alejandro and got in FebPlease let me know what else is needed for closing this storyThanks Avner +Verified in. +The overall line coverage is and the line coverage of the remote transport is +Ill commit the patch later on so it will be included with when we decide to release it +We have docs +This would be handy +The changes that have been made to the and the are not is not just about me running local tests this is about users deploying samples that are also dependent on this domain +to insert the tableWho is that user? He also need to be able to write on +Hopefully youll be able to hunt it down +Since is added last the fallback could simply be to change its accept method to accept anything that has AVROSCHEMAKEY defined noOne other thing we should probably adopt a naming convention for metadata keys +David I applied the patch +integrated in master. +Never mind +Latest version of patch +OK so if is implemented then it should also cover this issue +I reviewed and approve the changes proposed by Brian for the original patch I proposed +to the branch +Ok previous patch file might not work right +If you delete your local repo and try to use your plugin in a maven project you will notice that it fails due to that missing can attach here the files but as its time consuming to put them by hand i cant guarantee doing it fast thats what we ask people to setup their own repository that well sync from our side into the central maven uploaded not sure if the plugins will work without the metadata though + +Attached a patch for an example servletThis patch includes a bugget throws an exception with as a parameter +Still needs a short test that channels have had their channel buffer sizes set correctly +We should restructure the Spring Social home page to make things more clear for our users +Committed +Attached document contains Test cases for wsdljava and wsdlmerge is the implementation of this new feature contributed by the IBM Team +Patchs for improving documentation after comments from mirrorbox on irc +these have been added so this can be closed +I dont have logs activated though something you might have +This will have to wait until a later version. +Integrating the domain aobjects and also cleaning up our package structure +Added new history columns to createdrop sql scriptsMade observe parameter class. +This is fetched instead of the live one for reproducibility of test resultsI made various other fairly minor improvements too +So all issues not directly assigned to an AS release are being closed +We need find to find a way of smart parsing to define which are what kind of data +Bringing back into +But hopefully its just reference twiddling and not any actual data copyingbq +Thanks NingAttaching a new patch against trunk +I wanted to create an launch configuration for my class project which will help fix this issue +Here is a patch with support for check +I definitely think we should get Ed and Roger to weigh in adding a message is no good idea +duplicate of +The patch to is rejected if the patch to is not applied before +added just moving the call to the request cycle listener out the utility method that dont have the RC set in the thread local variable +Ive been building momentum on coming back to this issue unless a commiter beats me to it +This issue can be easily resolved after his work is done +The fix will be available with the next release +However it can be compiled and run as C code under VS See +I just committed this thanks Terence. +If that should be fixed as RELOCATION than I would expect that in EAP maven repository +Patch +Original issues are now addressed +fix is to use truncate instead of drop recreate. +Thanks for the linksI think the point here is wether it is important or not to be fast on description I would say its not critical because I dont see critical use case for tests description like it would be in logging frameworkSo it is slower and fully standard against faster but not fully standardIm moving this discussion on the user mailing list to have more opinions +The main advantages in my opinion is that write dont have to acquire the flush read lock anymore +Still not tested thoroughly enough so setting and classifying for you ClaudiuDoes anyone have an issue committing this patchI have verified it on my side and I am for it +Going by the comments in the GF issue this appears to be resolved +The same issue was fixed in trunk and in in much the same way +Gianni fixed this by making the deployer run off a so you can specify exactly which module jsr support gbeans you want. +Trunk Committed revision +Looking forward to new version of HS with Lucene and updated required libs +This takes plase for Tools XML Editor as well +So whats the case here now? This item can be closed with wont fix because of peters explanation? Personally id agree him +Should I change this back to M? +Tapestry Logo as Illustrator SVG fixed with final dot after Code less do more +Im currently focused on WFK and tasks +This would basically work like processing transitive dependencies with the difference that we dont as we aggregate the lists +HTML in system browser Cannot understand when weinre works and when not +Thanks +Looks much better than the map version +Pls look into itregardsDavid Brainard patch +Perhaps change REV to REVISIONSure +Will also be used for several others +On the second thought throwing from setup is not a complete solution +there were little improvements latelycan you check if the problem is still here with latest snapshot please +ive committed some updates on the greek would be nice if someone could add the danish translations I supplied on the fifth +Thanks Dennisexcellent patch as always I liked your solution of still calling the but only in the special case falling back to not use itregardsMartin +Is this addressed in +As told by Serg its fixed at +But you can close this bug because after my latest tests with the current snapshot the Exception mentioned above doesnt occur anymore +Could you that it is really missing? Maybe attach the incomplete jar to this issue with the exact location where you got it from? is essentially supposed to contain everything but the MVC modules and the test module +Also untagged the specs. +MDC properties in the output would be useful too +I think this new issue could be related to as I have no Servlets in my classes but I do include libs that do +Sending srcmainjavaorgapachecamelcomponenthttpAdding srctestjavaorgapachecamelcomponenthttpTransmitting file dataCommitted revision +with JBT. +Can you point me to it +A fix has been committed as part of +It is requested in XML +By default the subject is +Ill get to it as soon as I can +Please remove if it is not reqd +We are aware of this problem +Patch has been committed in revision +Most the failed tests were due to an error in the build that it cannot remove a data directory +The use of deprecated code has also been removed +When youre ready please file an issue as a subtask of so we can get you setup correctly in the system + patch +If there is a problem please open a new jira and link this one with that. +After patch over Ran Sort twice and did not observe this issue anymoreThanks KaramSubmitting this to Jenkins +The diff between ydist and patch for branch +This would fix the ndThe first part may be related to but I dont know if it is since in my case it would give me an error one second but work great seconds later +I just committed this to trunk +This issue prevents the demo application in from starting. +My mvn ratcheck fails because my mojo plugin is not up to date I will be correcting that now but the patch should be good? Did not test on cluster btwThanks for you time this is my first patch on Apache hope its OK +New pid are correctly added into the local config admin instance but the existing one are not updated +Maybe this could be a Hook that can be implemented by the end user +Reverting the merge to fix the branch build +Dude dont undo what Ive done +Yes +Can you please post the list of headers you found that has this problem? in case we miss some during our search +I think I can upload them with a different name and that will work as a simple workaround +this looks to be complete given the bootstrap css and classes I see in the site +Setting to future pending a patch +This means that this patch wont be applied and thus the issue can be closed +Andrea it would be nice to have a third opinion on this one +Set version to for all issues since more true than no fix versionAdjust the fix version to match actual aerogear plan +I need to know whether there is any work around for this issue other than changing the namespace or the method name of the hosted services +Thanks +Try again +Max Is this issue same as which is fixed already in and Trunk +I clearly missed that portion of the JIRA +Although we could wire that functionality into the default somehow I thought it was best to keep it isolated +Removed extras added javadocs changed names +So all issues not directly assigned to an AS release are being closed +Reassign to a release that you feel comfortable resolving this issue in or leave as is. +Please reopen if you still have problems. +Its also needed for transacted redelivery or session recover where you want to redeliver the original unaltered message which you cant guarantee if you dont copy it +added changes to proxyuser keys +Im committing a fix for this +It will just be smart enough to generate a more unique ID when initializing a pageflows controls + Good point regarding people already using this via the current package +Thanks +do you have lot of build results by project +updated patch +Revision +I can send you a small example if you want +We do throw an exception indicating that failover has happened and the session is closedHowever I think we need to explicitly mention that the transaction is rollbackedWe also need to document +First cut at what this issue is attempting to attached patch doesnt work quite right and instead the larger radix archetype that will create a build for all components will better address this issue +Committed +Thats good newsIll test it monday +Test enabled for client with checkin +Including into API tables and please update me if will not agree +Im reopening this with as target +Mike maybe it helps that in trunk tries to adapt to the magic numberIt looks like it is time to convert into a constant in or elsewhere +I tried the test without the patch and it failed as expected +Question if I have time to contrubute to the project what do I have to doIf Id like to make this calendar into sdatetimepicker or something +Ill submit a patch tomorrow +Also contains serializable patch apply with care +I feel it is reasonable to expect a numeric major and minor versions as that is the way most version strings are built for any software no +Attaching merge patch from +Ok Ill modify it to be a bit nicer +No not yet +In any case the manual registration in the original code would mean that the would not get if a region closed on one RS and was reopened on another +Heres a very basic patch that allows ve factored out two methods in and made the loadClass method of protected this simply extends review and tell me what you think +Made patch for trunkRemoved tabs +Please stop creating issues for these well search for files with and delete s what I did +If a password is supplied then CLEARTEXTPASSWORD is used +Im not an Opera user however so it isnt high priority for meI have tested just with Opera if someone could test with Opera that would be nice +The machines postfix received the email and thinks it sent it out +Missed to mention one pointthe number of reducer doesnt impact the issueIn my test environment the Job had Reducers +backport to branch + +It acts as a standard source folder except that groovy scripts in here are not compiled but rather copied to the output directory +Thanks Zhijie! +Patch in to allow setting the listener to null if the session has been prior closed +Added unit testing for ssl and ipfilter on the same serverAlso cleaned up the test code by closingThe patch seems incomplete or corruptgit apply fatal corrupt patch at line version of Flume will this be in. +Cheers applied thank youTarget milestone +Required feature before modifing the pom as before but this time using to inline the right +Please help me. +Im afraid I cannot reproduce this it works for me +Great findI was wondering how you are going to reproduce this in a test +Just noticed that svn diff didnt include the image used in the unit test +buildivylibHadoopcommongriddevbin +Thank you very much for the clarifications StrongRajesh I am making as wont fix as it causes us no trouble +ResolvedThe algorithm of guessing follows If Internal Web Browser View is active than open the same URL in If Internal Web Browser Editor is active than open the same URL in If any other editor is active and opened file has extension +Any more detailed information you can provide will help us to see if there is a bug or to check whats happening in your case +ajax request target has a prependjavascript method +Now that work is underway with AS all previous community releases are +Added first draft onWe allow users to create a file called which can live either in their home dir or at a location specified via the system property calledContent look like as followscore default is max default is keep default is queueTypeDIRECT can be DIRECT for using a synchronous queue UNBOUNDED for a linked queue +I just merged this to. +Yea I wonder if we show a replication tab on master UI or somewhere which shows some kind of replication state +This is not fixed in JBM tested it today +The overall design will get simpler +Pushing issues to the next CR +Previous patch does not build +In the future we may have more this kinds of situation splitting one task into multiple tasks +I made some modifications so that default stylesheet is picked up +In this case there doesnt seem to be an obvious simple solution +One Issue should be marked as duplicate +I doubt that would work with an Infinispan CM since it controls shared resources like the Channel +Can you take a look? You can run the unit test by ant test Dtestcase Dqfile +Soak tests to detect memory leaks +Guys From recent conversations on the mailing list where Doug mentioned that this issue may now be rendered resolved by recent changes to Hadoop Im wondering if we can close this issue? Its currently listed as a critical priority bug and there are currently watchers of the issue +This issue was resolved and committed to from an HDFS perspective it should be marked as suchChanges to will be tracked in. +Is not a bug +While this bit is being looked at on page last paragraph saysYour mojo has captured the version of Java being used to execute the build a property that can have profoundeffects on binary we have configured the pom to report on not or etc +I had to see what I was doing while I was doing it +cant reproduce this in Grails and GAE plugin +yeah missed that my bad +I noticed this yesterday when looking at the DTD files +Username prudrakshala +This was done to make it easy to define globally unique flow ids from the path to the flow in the filesystem for example hotelsbooking and carbooking would produce unique ids of hotelsbooking and carbooking if the id was only determined by the filename youd have a collision. +Is it required that all tools be released at the same frequency as the core hadoop systemPragmatically yes +Thus an OK will be green PREP or RUNNING will be yellow KILLED ERROR or FAILED will be RED +Good point Emmanuel but according to the Javadoc a new Matcher is returned +Thanks Eugene +Ovidiu told me you are already working on this +Bulk resolution of old JMX Management Console issues +Ugh this is ugly +Already backported +Apache Directory Studio version has been released. +Handling the launched event at the killing state is effectively a as the container cleanup event is always handled after a container launch event +And it helpedI dont have any hard preferences anything that makes life simple is the way +So we could haveBut a property as you suggest in file will not be enough for LDAP JDBC +or perhaps set a property on the parent bindings to set the geometry ids +I agree it makes not much sense to overwrite intValue in such a manner but on the hand if the user wants that +Can you please address the findbugs commentAlso can we add some tests by configuring new test Trash policy? looks good +This is currently still an issue +Switching to Java Fix +Also this is my nd Wicket webapp so Im getting better but tons of room for improvement! Commentsupdates welcome! +it syncs every thread on every call to poll which is almost certain to be a performance hit +So since the compilation normally uses row locking it will have to wait for a timeout before it gives up compiling in the nested transaction and restarts in the main transactionAttached is a hack which changes the lock mode to table locking when compiling in the nested transaction +We will add a separate directory into CLASSPATH of the maven generated package to allow pluggable jars that contains customized idealstate generation now the user need to add their customized jars into CLASSPATHPREFIX environment variable then specify REBALANCERCLASSNAME in idealstate to use the customized idealstate calculation algorithm +I have fixed the TODO with the deprecated api from do you mind taking a look at? Maybe you changes already fixed that issue? +Simple as that +Sending srcjavaorgapachecommonslangtextSending srctestorgapachecommonslangtextTransmitting file dataCommitted revision . +used mem Sheet Time Worksheet save Time +Adding this as a subtask of +Ive received mixed information about interceptor but I believe that an interceptor can span multiple requests? In other words it isnt created anew with each request +Im sure in every single Hadoop ecosystem component theres a way to achieve this goal of cascading defaults +updated everywhere but needs to be checked for the last aggregated jboss tools update plugins verifiedOnce SOA tooling for will be available verify features below and then close this Jira +Many thanks for this contributionOleg +The latest patch looks good to me +Currently the configure is empty and properties are ignored when passed in +Like I said in my last commentIsnt the webappsaxis directory the equivalent of a war file? Also if embedding Axis everyones war file will not be different +There would have to be a way to enabledisable secret key management independent of token expirationIm a bit nervous about Getting the enablingdisabling of token expiration and secret key management right for the primary NN NN backup NN and standby +And on trunk you will just have to check those same metadata map parameters for backward compatibility +Many thanks to Mike Sample for the excellent ideaMike +Like removing the system ofrefs count +I added support to WFS as ignored wfsv as its schema location is wrong anyway +Closed all resolved issues related to our projects. +If not one could create a test case that handles an extend hierarchy with at least levels +Updated to revision +Also if you increase the frequence that you restart the broker you might be able to find the leak more often +The complexity of this interaction leads us to recommend the workaround at this time and proceed with full integration after +But we should make sure that these message do not appear in the namenode log at INFO log level +Good on you. +As part of my process I also removed savedcachessystem +Here is the updated patch with your suggestions included +If folks think that being able to tell a real from a missing value is not useful for Lucene wecould extend Ryans to include a factory method that creates populates etcThen all the bitset stuff could be kept in Solr only +Cool +No definitive test case +Theres a similar failure on trunk still but in the test testIf it still happens after I commit Ill track it down +I tried to remove this from each of the files but Railo adds them back in on start up it would appear +how can the jetty classloader then load the classes that arent in his pathsomehow you have to have it in your path +When the file is closed we write out a side +Ups! Sorry +Would you like to submit a patch as wellbq +Ok thanks Hemanth +Waitbq + +adds code and tests +Should I put this in the branch or orBTW Ive decided to only commit this to the security branches because a few people are likely to use new releases of and and b the security branches actually have security for the. +The interaction is very limited currently only creating ephemeral nodes is supported +I believe Apache only handles Unicode on filesystems that support itnatively namely NTFS +java jar ij ij version ij lets start the file with a nested comment and see what happens connect jdbcderbydcreatetrueIJ ERROR Unable to establish connectionij values hiIJ ERROR Unable to establish connectionijThis script works if the comment isnt nested +From this information we will write the final release notes for all issues +Some images have migrated away from their captionsI had trouble with header and footer text disappearing after some changes to the template +Could you comment on any overhead? Is it same as Simon stats +Check case when edit layout of are a lot of case like this +great! About future how usually the updates are applied? Do you let translators know before the release to update the translationsAlso about this encodings next time should I send you UTF file or translate it myself in ascii unicode? I can do that if you tell me how +I actually think that class is obsolete now +Verified in CP CR and CR +These should all be corrected before this patch is released +Ahoj Pete this should be assigned to somebody else as I do not have enough insight into these JBDS internals +Attached Looks like previous patch had slightly old generated proto files reattaching +Minimalistic project with for the patch +Steve is barely I am +Defer all unresolved issues from to +I have just contributed to get a few algorithms going with a somewhat descriptive interface +Thanks! +I shouldve also mentioned the javadoc warnings are unrelatedIm going to commit this momentarily +I think this is more an improvement for Ivy instead ofSee also +Noam is looking at this +Marking these as fixed against rather than against +yes am working on itthanks file with sample service and client code that i used for testing +It seems like something that would be a nice standard feature of the Hadoop toolbox +FYI Ive actually written a fix for this issue in an earlier version of the library for our particular implementation +We need to implement +Ill check the debug output on it tomorrow +This keeps breaking +Im sorry to bitch here but that bug getting unassigned doesnt actually increase my hopes to get this fixed in a timely pretty much everyone who is using Maven has some established setup like described on th March +There is a way to specify via xdoclet in Hibernate +But i do not know them and i prefer have no informations at all instead of erroneous you have a way to analyse a jar for its dependencies +Thanks RaviThe trunk patch isnt applying to branch +IT done +I will remove +However this issues is for to the policy is to not change files in the repository after they have been deployed +See for a corner case that has not been resolved by this bug. +One of the things that works really well in Solr is that any time some significant JIRA issue is undertaken a Wiki page is also generated that effectively documents the ideas in the patch as well as how to use it and thus results in the final page effectively becoming the documentation +Updated patch to work with the trunk +Only tested with tested on a fresh version of +Ill check that out +bq +The from Hadoop QA is due to the patch being html and not forrest xml +of and the properties of doesnt seem to be inconsistent +Can you take a look +I think this is a real improvement I have no problem with getting that in +I tried the version of the plugin and still got the same problem + test removed unnecessary and wrong method main +No interface would be best +I am opening a new jira id to fix all messagerelated failure in +Let me know if Im totally off baseI dont think youre off base +The problem does exist +And we should change the error message +I definitely see your point about the JNI code possibly locking or copying +The sort on node cluster shows that running sort with permission checking turning on takes s while sorting with trunk takes s +In release fix version +Thanks JornThey must have went to the span folderIll also have to make sure I use my gmail address for future posts as well +beautiful +When I originally applied your patch I didnt see errors. + to and trunk +The correct fix will be provided in +Thanks you are scaring me maybe I am doing something wrongHere are the steps to reproduce the problemcreate project top jpa provider HIBERNATE database HYPERSONIC persistent class jpa name +So I think were all set there +Ill look into it +This doesnt look quite as trivial as I thought initially so we should respect the freeze and put this in +Created the readme that summarizes all customer samples because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Nice +Closed for release +Hi Keith can you review please? Thanks +Also there will be no logs like No SCXML child state with ID null found thanks to the fix in +This a kind of philosophy to do things mostly automated but I prefer things like these to do it a bit more explicitly in some it is because we have a different view about how Groovy is used in the embedded mode +Done +I went through the logs youre right + +I am away now for days and theres quite some issues that I would want to include +I am willing to give more infosdetails about each of those issues if necessary +Alexey Look after your back first! Well survive meantime +Im new to camel development so I may be slow so if someone else is eager to add this functionality just say and Ill stop +I already started improving the outline view before this issue was createdThe result of that work is now checked in +I dont think well fix it +INFO chukwa +Correct me if Im wrong Im newbie in advanced Java enterprise stuffThanksWojtek +Need to make a small change in favor of already existing Regex support in +works for Double and Float too +Dale I understand that +Since some of the subtasks have already been deferred deferring the parent issue as well. +Patch applied thanks +as you mentioned this works on theso what is different between the and the svn +When will the full ibmjaxrs fvt tests be merged into truck +And another very interesting observation I found in effective POM the thing that makes a difference +StuartMany thanks for looking into the issue +But Id still like to see the behaviour for strings documented +Looks good +Given the example above if Policy is the Message content then your Drools DRL would have to reference the Driver as a nested data item from the Policy fact in the LHS of the the new functionality still keeping the Policy as the Message content you can specify a Mapping of and the Driver is inserted as a fact into Drools instead of the Policy thus the LHS of the rule is much simpler +If the escaping behavior is incorrect we should fix it rather than only providing users a way to work around the problem +I applied the patch locally ran mvn install and did tests with a sample app and it seems to me it fixes this issueI will make some more tests later today +Im also using Snow Leopard Safari +We already using this approach for Hbase for distributed log splitting +ChunhuiHadoop QA is not functioning +Im still not in love with the sarcasm or excessive negativity +Ive committed the changes for BIT BOOLEANINTChris could you test the current J trunk against and if indeed is resolved by this mark it as such +I am not sure why so I wrapped in trycatch and commented +Fixed in trunk. +The old projects have been removed from the Xalan Workspace and the documentation has been updated and corrected. +In other words Id start with doing everything in Statements and then optimize with under the scenes as necessary +Submitting the patch for review Ive now confirmed that this is a problem and that my patch corrects the behavior +the patch that corresponds to my comment +This is noted in the workaround +Fixed in r in trunk and r in +Things are still in flux and this is a very minor API changeI will just cap n at the number of users in the data model. +I have run into the issue where pauseresume where not firing which appears to be fixed in the commit just postedIs it possible to use the commit now without waiting until the next version of CordovaPhonegap is released +Once this change is in I would leave bug open but lower priority +The carried over state is only a reminder that the story was started in a previous a calculation perspective it should be considered as a new task When the storytask is carried over the original taskstory estimate should be wiped out +Definitely it makes sense +Should apply on trunk as well as after +Would you pls check it again? Thanks +needs guava and watchmaker is dependent on google collections which looks like the problem to me +Woo hoo closing one of the oldest geoserver wishes! +And Im just as confused as you I dont understand why this cast is nessesary +Uploaded test was moved earlier in the test only one heartbeat is done and the statuses are marked with the event transition stuff +With Forge around the corner and the start of phasing out the Forge integration this bug is not going to be fixed +There I might still change my mind about this +Client uses separate to avoid code generation into the same dir subtree +Thats a cleaner approach since the semantics of what a columnFamilies list is doesnt have to be special cased outside the deserialize which is the right place to deal with thisSpeaking of which dont we need some code in to force the cF list to all if target is an node? Otherwise what we send will not be what target expects +It should be working unless there is an additional case or you need to update a configuration file. +If that is fine the bit needs to be taken forward outside this JIRA issue +Thanks +m to define a repository +Thanks again for solving this issue + changes runtime binding logic to a service child of service container which is stopped by the owner undeploy processor thus always after the component lifecycle but be aware that the changes were made to fix a real issue by using an older subsystem you get exposed to it +In the future we will work toward making the framework components moregranular so every piece does not have to wait on the whole +Changes to for the target +Really weird Any explanations +IS this issue really respecting permissions in the fuse module or also the ability to set permissions? Maybe should be divided in since it seems the underlying needed in libhdfs may end up in themselves +So at the AM its hard to tell why the container exited +Working on trunk +Opening a new issue for this is definitely the right thing to do even though the title of this issue suggests that the problem I hit is related in a way +Michale Wyraz Your workaround of increasing batch size works for me too +But strangely this only seems to happen with Spring AOP involved +Upgraded to bsh +That said I cant give you a dedicated timeline for all of this I am doing this as a volunteer and am travelling extensively this week so Ill try and take care of it as I have time available +I just committed this +Will leave this open pending implementation of the more complex fix to the verify error +what is container in this db +it is possible to implement sort avoidance as implementation of plugins for and with a special implementation of Asokan as Ive mentioned before several times on other jiras I believe this is the wrong way to go about doing adding this feature +I have created a new type class which impliments Map with methods for a String prefix and URI namespaceThis required very few changes to the rest of the code but some parts could be simplified so Ive included these in this patch too +Under Runtime Operations Services clicking on the gateway view configuration the screen flickersAll above was tesed with Chrome +This can be configured via the property +builds are now running again. +Thanks! +And existing tests used for regressionthanks +Will this be fixed in the EAP rpms? I could not find an EAP Jira that addresses this +findbugs +I meant could you verify from the latest CVS +module containing a session bean for a web service +Also note that killing the build will likely not terminate the forked JVMIt takes less than K messages to consume the available memory +so could get byte arrays rather than streams +This has been fixed in +jbossws collides with jbossws not included in ejb with not included in ejbAll other services are available in the profile +Ive been following the issue and am satisfied with the outcome +This issue says create sample which has been done so Im marking as resoved lets create separate issues for the suggestions from comments +Mingjie Any chance of your carrying this one over the finish line? Your run of the miles on this one +I have only noticed this behavior with I have not tested with other result types +There are compilation errors in +in trunk supports iterator already +The How is used? is not useful to a user does not teach them how to do anything or provide any information a user could benefit from +I could not reproduce this bug +Attached a patch adding length attached patch adds length validation to class +Great thank you for looking into it so quickly! Hope the release is nearI tried making the cached stream transient today after your suggestions but that caused some other problems with some of my resources even though it might be working in the +Committed modified version of patchs and in rev +anyway it would be more consistent to allow a temporary queue here as wellCould this be done? +Not a bug but relates to a real bug +If it is OK then Ill back port it toYour quickstart app seems to work as desired now. +When the project examples xml files propagate up to the server through jenkins this should be resolved for all the examples where the text mentioned was included +TorstenThe patch that Amine provided had a block of LDAP code inserted in the file + +Moving to ll take a look at this if no one minds +should it be reopened +I couldnt find any newer version of the glyphlist containing the missing mappings +Ill merge this over the weekend +was released added to hbase was released without the native libraries +This is an area of improvement for the next maintenance release. +It is also the intent to open it up for external developers +I added more info logging since all the major parts of JCS log major events at infoI added a bit to some of the javadocsI added more unit tests and will add more to methods such as detect overlapI dont call defrag for the put list if it is emptyIll check in my initial changes that bring in the bulk of the patch +has been released closing issues. +bug and proposed fix correct +To clarify I understand that the spec is clear that must be treated +Unfortunately I can still see the failure with the latest trunk rev +Im not building a protocol Im building a platform on which protocols can be implemented +Sorry I did not recognized a problem in the first patch +Im not fluent in but the changes look reasonableI committed this +This actually affects +Issues resolved in +We could use all the help we can get! It takes some patience to become a good contributor +From what Ive seen this seems to be an issue with the Oracle drivers handling of and do not suffer from this problem +Not tested as I didnt know how to reproduce this issue with Sonar Eclipse +But I suspect tha replacing Municipality bu City will be enough isnit? I see +I will add additional tests to verify that the tables were created correctly including the nullability of the columns +Unless you prefer to log an empty string instead of null in that caseFair enough but since you offered +To be clear I do think we should fix this +This component is not intended to be modified from outside the portal so Id really would like to be ableto remove it and provide only an immutable interface to our portal developersRegards Applied +Its committedSebastian mvn compile and mvn package will succeed with violations +haha +Deploying the site once again with this patch should work +Isnt it true Todd +cant think of some sensible test case for this +I looked on the SVN repository and saw that the fix above is not committed want to confirm again that this fix works for me also in you have no objections it would be nice to get this included in the next Castor know that you have to care about a lot of other issues too I just want to make sure that it does not get forgotten +Paul attache is results from sstablekeys so there are definitely keys in your sstables +The list is iterated through and the listener is passed a object +Of course if you are actually using hopcount filtering you may not want to do that +Hi Timothy Ive attached the file +Tests passed cleanly for me on +Made phabricator entry for quick review +Closing as wont fix. +The test doesnt have access to the object used internally by to synchronize so I added a delay that double checks whether the watcher is closed if watcher was still open +This area is a big unknown currently +New patch but with unified context +Thanks. +The one piece we havent started working on is a connector to use existing Pig loadstore functions for a given format with as existed with storage drivers +Please review and commit if appropriate +I understand! Thanks againI will update this on RB too where comments on the last couple iterations of the patch contain good command lines for building and running it on the cluster +Modified version of patch applied. +This was fixed on rev +Its an alternative to it that should be a little easier than writing your own plugin +V of the patch + patch looks good +changing to synchronously call the handle routine when submitting an application +upgrade to I leave it open as we will have to upgrade to a released in rev +patch applied +And as a matter of fact if you do try stopping a compaction both v and v do log an error while v does not +We can fix a limit of temporary files for example every ten files we clean the folder just before the portraying process a test is launched to get the count number of files +Yet another patchThis modification fixes an issue where Exec was not properly handling primitivesNot posted to as the site appears to be down +added lazy loading of in to prevent opening the CFS file while opening +Hi David I think there are some jdk canons that need updating too +I cannot reproduce it locally +Will be covered in Jaroslaw I used the kb any chance do you have the different services running on different machines +on and still works on hadoop as well +He Pablo! No problem just wanted to see how things were going I would like to add it to CR but it can always be included in CR or Final! Let me know when you think you might be able to work on it! Hopefully its not too hard to get working but hard enough that its fun +Resolving Fernando and Clebert applied changes in ER tag and branch +Thanks! +I excluded regression test for this bug until it is officially fixed +Help in resolution is appreciated +We likely need something similar to +Unfortunately your pom refers to a parent pom which you have not included +Guys make a decision please +Now that work is well underway with AS all previous community releases are +Patch to for bug reproduction +Ready for review +Im a bit behind here +Slightly modified version +Ill resolve this issue as done feel free to reopen if needed. +Hi Petar this missed the cutoff Im afraid +Try now jd +Forgot to mention test is copied all but verbatim from +We would still need to put the nodes into a bootstrap state to get incoming writes forwarded to them otherwise you have to repair in the end anyway +project and +Pull request will be modified +Reopening this JIRA until Hibernate makes into EAP +Thanks for review stack +We are waiting on Infra to finish setting up the distribution areas for Flume +SSL may also be used at the socket layer though that would require work in the Hadoop RPC NIO socket code +verified in Revision . +See bug for opposite side of the coin +Stack thanks a lotI moved them to in and posted a review request +Reading blackdrags comment I guess this should be closed. +Tested under Mozilla and IE. +But in and later it overrides the getCause method from Throwable +better patch +Please double check and reopen if you still encounter problems. +Closed upon release of Hadoop. +try with the following settings maybe JAVAHOMEusrlibjvmjre export JAVAHOMEIt seems to work without any problems for me on my ubuntu server vm +I have also hot this bug +Edison can you check on this issue +Eric thanks for the info +Will look at it more closely +In addition the the class level javadoc perhaps we can expand the with some tables of Strategys vs +PS Do your tables have utf as collation might be the first thing to check +Fixed the type safety and unused imports issueAs to the concurrency point raised there was indeed a concurrency problem but the suggested solution wont work because it would only queue regions to be flushed eventually +Patch needs regenerating following the project split +Although we announced feature freeze for this is a easy change that should make it into so lets go for it +Yes Im interested in a release of this project +Testing best of three I get for Ruby down to with the simple autoload enhancements for down to with the enhancementsCorrection buildrautoload after buildr slows everything down but buildrautoload by its own gets us down to for Ruby for + +patch looks good +You need to change the url connection to OracleYou needn the following libraries to run this exampleOracle JDBC DriverHypersonic JDBC DriverJakarta Commons Lang and libs inside hibernate folder +Updated patch to address the comment from Nicholas +All is good now +task is done +Remove as fix version because you cant fix something in a release or else jiralint will complain +updated with all feedback just waiting clarification on one minor with feedback pushed to the book was updated in the release. +I attached a patch to fix this bug. +Id like it to be consistent everywhere and backwards compatible +Anyone have a strong opinion on itIt would make sense to me that if a user stopped the scanner we shouldnt automatically restart it without their consent +Marshall could you please verify wether this fix is for or +need to be handled as well as those could take an arbitrary amount of time as well +Will was going to enter another jira about what to do so it can support both and specs but got had to change the pom and switch the versions of and jasper would be willing to work on creating versions but dont know the best way to do it +Closing all resolved tickets from or older +The HTML files are included for review +Attaching the for activeclusterCouple things that you all should be aware of cause Im not positive of your desires here The version I marked it but perhaps would be more in line with your current version schedule? The groupId I set that to my understanding of the m guidelines is that groupIds should follow this general package type naming scheme I excluded some tests from the test suite please take a quick look and make sure I got it right and im not excluding something that should not be excluded +We need to normalize and abstract the configuration details specific to HA proxy F etc to notion of health monitor policy +Updated patch to reuse existing constant +maybe well go with something that is exactly the same as the picket link deltaspike api. +Fast fix to the bug +But maybe we will switch to java anyway soon +bulk defer to +Looks good makes it more readable with the refactorTwo comments DSS in loadEndpoints for dcTokens could a Collection be used that is naturally sorted so it doesnt have to iterate again through to sort them? should the comment on line really be before line instead where the loop w suggested comment changecleaning up loadEndpoints is a fine goal for another ticket +okay +Thanks GuillaumeWill update and repost +The main problem is that most people I know that experience this issue because some Xalan is dragged in via some transitive Maven dependency and they have no idea what to do +patch preventing the pagemap from getting into the dirty for page map max Thanks for the work on the test cases! Ive added them as well as a version of your enhancement patch to the code in r +This is still a pretty severe bug +Patch looks good but please make sure includes tests for the change + +Jon please dont forget to close this bug once youare satisfied with its resolution. +This issue is no longer relevant for it is fixed inChanging the affected version to +Attach a trace file + +Patch without reformating +Yes it is definitely not fixedMy Config Hello world project gave me this exact error +Verify that the listener is called +Thank you for your good jobI found a problem when render report usingThe works well when export report as pdf or word but the chart cant display if the is texthtmlHave you met this problem +Hmm I should probably tweak the Edge split logic slightly to reflect the fact that breakpoint materialize trying to decide whether we should just use the smallest materialized point or the first one or the last one what do you think +Well not exclusively with cql +Yes its worth doing +using s xml parsing methods +You guyes are implressively fastI will pass the info to the Wiquery team Thank youMarcus +Thanks Ivan +Russ patch applied please retest +Committed to trunk and. +Will provide a minimum mapping and test case shortly +Please attach a sample project to this issue +Ok if N N How do you specify NInfinity +Added Javadoc for Plugin brief how to use APT document +The new topic om trace looks good to me +You can fix this by changing the taskType instance to be a parameter that is passed when creating the mockTask and in the reduce tests create a new mockTask that has the correct task type +One thing that should at least be done is to remove the async is true on the fork and join +Updated patch some improvements +There wont be a release +I committed this to trunk and +Here my patch for and +I can declare the licensing of asl as a mistake and instead full gpl if we want to press the pointYou cant be on the PMC and play games like this if you ask me +Symlinks exist and their semantics are different than directories or filesThere may be an occasional program that needs a tiny change to be compatible with symlinks +Seems the likelihood of that being what the user wants to be lower but is easier to communicate +Can you provide an example please +Do other temporal databases need to be renamed togetherPossibly but I havent seen any failures other time so not sure + for the patch +Committed +Expecting reply on third point +You will find no event triggered by timer is inserted +Thanks +Response to AlanI agree that it makes more sense to do the overwrite in and I notice that theres a JIAR which is related with this +There is some more work to be done here is our data directory +Updated patch to remove the static keyword inthe enumeration will keep a handle to to avoid garbage should still close the in order to close the underlying properly. +the latest patch looks good to me +This has been fixed in the Trunk code +Existing codeapi is sufficient for the requirement. +Otherwise the patch looks good +In addition there has been a new geotools implementation completed on a geotools branch but it has not been merged with trunk as of yet +Updating the original patch to make a few more attributes required and adding a patch to add a page documenting the available tags +IT looks good now +Any help is appreciated +Verified thanks Tony. +Ive integrated the currently suggested impl at revision as it holds a train of big patches at the moment implying your promise to address the issue mentioned aboveI slightly reorganized the patch before committing to better match existing ifdefs order +Verified works fine with CR build. +Otherwise well scan the directory and return the first wsdl file we findone thing unclear is whether the servicename or portname specified in should win over what is specified in annotation +waiting for verification +Does anyone have some time to look at my patch +updated for M which now includes a jBPM Tools category w features in it +Logged In YES useridSure but its very specific to loom +Im pretty certain this is a duplicate of. +I have no workaround so far only fix +Good findsfixes Todd +I think that can be also thanks Strong. +in will automatically make sure that the conditions you mention above dont happen +Fix committed +Some folks recently had confusion about it in the Forums +patch. +Its cheaper than a dynamic field +Try the next SNAPSHOT for testing +Because all of tables in the jBPM is starts with JBPMSo I change Affects Version from to None +May be related to the Build Settings changes recently +See +Suppressing exception is fine I suppose +remove has occurred You should use care when retransmitting Messages to services +In your case you are downloading the sources from our SVN repository +Thanks +Bulk move issues to and I think scrolling would help in the case of Hadoop integration such as pulling a few hundred thousand docs based off of a query to the local node so that you can do an aggregated calculation with Pig or MR. +bq +requires the uploaded binary file and it passed in my local test +bq +You can install meclipse without Mylar integration or use meclipse +Attaching the Patch. +Please verify +where does meclipse create that? In output dir or in actual source directory? if in source directory it would get added to svnscm which would be wrong right? And how does this coexist with Maven command line builds? Wont things get very fastIm not against the suggestion just trying to figure out how it will actually work beyond the import +Can you answer Maxs question about versioning? Rob +For the top navigation on the project site I would change the Twitter JIRA and Github links to open in a new window +I only moved the two source files and left the chinese documentation in Common since mapreduce doesnt have any chinese documentation as of now! +Theres only one caller of get and its fine with the null return value +verified on +get will pass the to the parser and the bean inspector will use that class to set the properties +for now since all operations about the runtime installation are performed in jbossws facet installation so if you want change target jbossws runtime you should uninstall the jbossws facet and then reinstall it when reinstall it you can change it in jbossws facet installing page its really a little complicated +Hardy do you want to work on that? It looks like a nice one to dive safely into Hibernate OGM +Thank you Alexei. +Today one of them spent all day failing to respond to requests giving an exception for WMS +Hi AllCant we resolve this now? Thanks +Excellent thanks a lot +This is just like CFS which is disabled in +You think the problem is in WSSJ even though the version has not changed from to? So something change someplace else that causes a different usage code path +Isnt this how Bigtable does it +seek an update from the distributor of this software for a version +Pushed to master +I will make the changes suggested +Perhaps youve hit a timing issue with the index metadata load +Ill look at +Ok will open a bug and try to provide a patch to CXF +Plus the client has to dictate what is the server principal as we are enforcing dual authentication +RT monitoring works fine +We will also want to properly externalize the exception messages and select the right exceptions from the hierarchy +you for the tips it really help me +I committed this +Updated test program to demonstrate issue with xlink +Thanks Brandon! +But the same classpath scanning mechanism as could be used. +But in this case where we dont expect huge outputs that may be less of an issueIt might be worth developing a standalone benchmark that repeatedly spawns a process that writes to standard error andor output in order to test various approaches +looks good to me +What is an appropriate way to do it? Do we build the from Nutch code here locally and ship it as binary packagelibrary or as part of mvn install task ant task? Nutch Language package by itself depends on other libraries +Reopening as this is a broader than my earlier commit +client Xcvs buildoutputclient Xcvsjbos buildoutputclient r Xcvsbuildoutputclient Xcvsbuildoutput +Theres no point in you fixing my mistakes +which is ok assuming int assignments are atomicI havent had a chance to look through the patch but in general this is not a safe assumption +If libhdfs had a permissions API then we could make a test that would trigger EACCES +Looks like solves this issue tooThanks for the report +Not a problem is the correct way to describe this resolution as long as thoughtful and logical readers of this thread remember the purpose of this JIRA +methods especially this one +is internalDitto shoudl be private should be privateBelow should be private too? InternalBelow are superclasses of Put etc +This has gotten through significant testing by a user closing. +This looks like a duplicate of +Migrating bugs marked as resolved before and tagged as version as fixed in RC +I think registering a as an service should provide you the needed +A simple solution could be to just make these statements instead of +build break is due to the incorrect variable declarations in the patch attached includes the fixplease apply the patchThank youDumindu +Not a bug +Moving these all to a Doc release version. +looks good to me. +Heres a candidate patch which fixed the CPU spinning on my clusterWorth noting that this problem is more severe when the fetcher thread count is configured higher than number of nodes +resolving this with wont fix +Fix if possible for +Thanks Robert Ill add them in the Solr example too. +In addition this patch fixes a bug in the exception creation code in the Launcher and improves the check for in unit tests pass +Small initial patch that adds module and integrate it with build process +I do see that this patch is making changes in and though +Michael go ahead and close the issue if youre satisfied +with the patch its not easy to just sort on a NDV field with block joins but I feel like it should beBut as a start lets do this first and see if we can reorganize +back ported the to synchronize branded please update also the branded Seam documentation with the fix +Here is the for the patch but it is not a valid patch file and does not apply + push off to to in revision +and committed! +patch that resets the inputstream +I also had to include a small edit to to address the issue in which affects our system tests +sure Well do our best Bob but for the time being I cant promise anything +Linking this one and due to the fact that theyre both dependent on the WTP bug being tested again for this issue +JonSounds reasonable to me +I will see with the community if we should remove the original files we have in the repository +After five iterations jruby is not starting anymore due to described problem +This does not print out granted that we do not yet have the capability to register but I think it would be good to be able to list out as well considering that we are adding new ones soon +Filipe overhauled the replication code after branched +Applied in trunk at r and at r. +Missed a change +By looking at the patch however Im concerned about maintenance +patch file for the issue +The patch now passes all unit tests + +Surprised we dont have a util for Please add a test using to verify that if the client provides a type that is respected and if a client doesnt the default YARN is picked up +Ill send out a note to kick out the process soonThanks patch has been applied +Committed +All errors occur with both and in fops lib directory. +version is refered to explicitly by the class perhaps David you had copied the +IMO clientValidation doesnt describe the kind of client validation which is in this case scripting +should be merged in M merged to m branch +This implementation no longer needs Streams are closable and it revealed the problem in +Committed revision +Attaching configuration files and servlet for reproducing +This is with the generated code +Ive tracked down most of the issues so far but it seems cleaner to just do a single fixed patch than all these littleIll also kick off test builds on my own for future patches to this JIRA because apparently the common tests are lacking here +But if userschoose to do so they should be given that freedom I guess. +Do you already have a mail server running? Try and change the port thatdumbster is starting on from to something else +Is that almost the same thing as what we are discussing hereAlso I disagree with deprecating the current modemixed +even with your proposed change it seems dangerous to me that those NAME variables are mutableIf we fix them all to be final then the problem goes away wo needing the rest of your patch correctWe should also be able to whip up a simple test using reflection to fail the build if anyone ever adds a class to wo static final NAMEYou mentioned discovering this problem because you have custom query parsers extending some standard one where NAME isnt final Is there an aspect of this bug that youre seeing with your custom parser that wouldnt be fixed with the solution im suggesting +and doesnt extend from any more +Patch looks good +Left a word out of the new note in previous patch +I will take that to the list since I think it is important to have it discussed in the general context of versioning +appears to be a duplicate +Currently the vote for GEP is under going +Fixed in latest CVS. +When I switched to the accelerometer tests it reloads but references it as +The bug was fixed provided the provided in modifyFeatures were coming from the same schema provided by the datastore +Thanks Ugo for the patch +The patch that fixes the problem +Please test at your earliest convenience +The zipped file contains a self contained maven project that shows the bug +Patch with correct numbering. +Ive changed tat code not to meddled with signed and the we get for and TS are signedQE please verify if this is still happening on the EAP needs some fix still in the part that writes information to the I only fixed the part that was rewriting the as that was what was reported at the time +Seems like a reasonable request to me +I just committed this +On my opinion its more if we support old directory structure for existing files and let compaction do replace for us +I must overlook the log +How is this issue fixed? It doesnt work for suspect this issue is closed because there are alternatives albeit inferior IMO +It would be simplest to implement includes textually via a preprocessor like cpp or m since we want to include the transitive closure whenever we store schemas in files or send protocols over the wire +I did remember setting this value to zero so I started over with a fresh copy of Eclipse and new Tools plugin install +Thanks for the quick review + +Removing from movement in a long time on an exotic issue +I agree the code should pass +The versioned collection would be a straight subclass of so +This utility is certainly not a part of the Flex SDK +Is there a specific reason why you chose to use an external transaction manager here even for receiving the JMS message +New patch does something slightly ugly to make this check not be triggered for itself +Need new license agreement from License Agreement and attached to this JIRA these need to be included in kit +Thanks for trying this AmareshwariIve added your INSERT OVERWRITE DIRECTORY dir Select +Verified at r +Ive attached a trimmed patch thats depends on bug +Maybe we still need a little tweaking of the feature that are generated +This should take care of the deadlock +I think the permission for the server conf should be +Once the use of Chinese in the document document will be an error +Seems ok to have javadocs depend on +This is the patch in sync with SVNHEAD lets see what happens on hudson +Thanks Tom! +Closed upon release of Hadoop. +Duplicate of +A couple of other points that might be good for the docs This is an alternative to In +It looks like this code will still be accessed by the EPR registration process and the invocation which retrieves the ESB organisation details +Need to figure out how to do release builds from old doc from source tree +Testing it now +I would like to delete it +perhaps we should consolidateAh yes I think we should +The problem is that jvmti function was called without breakpoints list locked +Exactly the same error hyperlinks correctly on my linux machine +I have a screenshot from the performance tool +Lars Synced and run and +for to trunk. +Thanks atm +Daniel please verify +chunhui Could have a look at the suggested change inThe observation is that after the scanner is reset the top KV will in the vast majority of all case be changed that is only a problem when the top KVs row has changed in many case indicates more work than necessary to a callerEdit Misspelled chunhuis name. +Maybe we should reconsider the whole protocol package with respect to API vs +In order to accommodate this I have to move properties field from to and then have properties in as well +Resolved in with JBDS +pushed to with JBDS +Yes that exactly what I asked for +This will require documentation +Can we close this Erik? You implemented a bunch of long compares this way in +The shell script binactivemq that you cite is not in the latest binary distribution +trail version does not test and need fixed your comment +Nicholas do you know if such a JIRA already exists? If not Ill go ahead and file one +Id be grateful if you moved it for me otherwise Ill probably close it and create a new hdfs issueI moved this to HDFS and updated the description to reflect the discussion. +I applied this to head +What we plan to do is to extend the current backuprestore mechanism instead of adding a new export link on profile list +We should maybe have the doc team add some notes to the FAB documentation that it may take a longer time on first deployment +I think I didnt change anything else +is currently undergoing a complete redesign and we certainly could add support for proxy chaining to +Its meant for those who want extreme performance. +Looks good +We have upgraded to cxf but still get a instead of a when the service is not runningOK +Thanks for the response VenkateshWhy is this necessary in logs? You could enable debug with krb which gives a lot of information for debugging purposesWhere would that be to specify the debug to be enabled with krb? I hope its application specific conf setting +BTW I reported a separate issue for not working +Reopened to set the Fix Versions field to JDO beta. +I added the jar + I get some more info on this for Release Notes please +each record was too large to fit into memory then a paging strategy might make sense but if youre processing a list of records then this approach makes what was formerly granular enough for the framework to manage into an opaque blob +Using what code exactly? passes for me and has for the last few days latest Apache JDO and latest JPOX +Itll be a great selling point but it will be a lot of work too +Final. +same reason for New patch also added test case for disconnect and reconnect logic +I would be willing to manually populate level cache to have the benefits of bypassing managed objects in batch +I tried to reproduce this and couldnt +Sending trunksrcmainjavaorgapachesshdserversessionTransmitting file dataCommitted revision . +Yes I think the fix was in MVEL since this bug was raised shoudl be all cool now +Closing as Wont fix since this is not a bug. +Ill assign this to myself and start work on this +Thanks for pointing out. +svn ci m Making class final switching getMinMax usage to max usage +Will put up the patch on review board shortly and provide the link +classes as suggestion so no need to deprecate +The truth is the web app returns JSON for errors it raises itself but leaves things like to the servlet container +Thanks for catching that +I merged those as well +Attaching one more file of patch +Why is v touching compaction +We are basically waiting to solve the shutdown problem. +Hey Jan thanks for cleaning things upTo answer your questions Yes decimals should be encoded as USD +You seem to be missing from your patch +Committed at subversion revision +No I cannot delete the filesThe files are no longer present in the mirrored distribution site however they have been archived which means that some of the archive files are not accessibleThe following files need to be deleted +In one of my numerous attempts I added the possibility to start a miniCluster with a specific or class +Gerhard Please give the attached patch a try +Ive put a comment there +David could you explain me how to test this +They will have a completely blank environment until such time as the admin assigns them to a role +It was broken by a different change +xml and have no such problemCould you please provide a thread dump maybe we could see if really is the cause +Works for lucenesolrbranchx +This has helped a lot Thank by submitter. +Removing non ESB quickstarts from description +Whoops wrong diff +Have to do some basic UI designAgain like I mentioned was different as number of maps reduces doesnt change after job starts +patch applied i will get this incorporated in the next website thanks for addressing this so quickly +Since the JBWS jaxws tools doesnt support to generate web service code from WSDL using so please make sure is contained in your PATH environment variable +It is nice to expose it as an argument +Please check in your project file +Fixed failures caused due to low limits on activated appsFindbugs warnings seem unrelated tracked in +I think we can close this issue. +Soa loadb distinct ashould translate to map identitycombiner apply unique functiongroup allreducer apply unique functionAlan +Looks like its necessary for the given scenario +Bug has been marked as a duplicate of this bug +According to IRC this problem can be solved by new version of Ruby on Windows +It is basically the same there is no for the primary key +setting sql component dont know anything better for message changes +I think the caching we would want to enable by default using a check but Im open to going either way as to whether the viewID caching should be default in +We need to create a test case to test it sufficiently +Then youd see updates from that site too +CommittingModifying titleJust as a workaround for people not having this patch they may use Base codecs to get their delimiters in and out from the XML +Great catches Uwe Ill fold them in thanks! + +Other than that I dont really see too many issues +It unassigns the daughter regions when the RS checks in with the split message if the table is disableddisabling +As far as names it could possibly default to the name of the plugin jar or something else and first one wins second one fails +Thus the NOTICE would not be neededNow should the generate a separate DEPENDECIES file for the test that includes that? Maybe +I only upgraded my jackrabbit libs to and thats it +By in the patch I meant that I fixed in the previous patch +Ok +Our code doesnt close them by the looks of things +convinience vs bloat we will have to evaluate when the time comes +Vertically compressed thumbnail +exe version. +After further investigation this issue appears to be with the WSDL that generates +Thanks for your patch Narayanan +However clients shouldnt have to care about it must mean cqlsh displays reversed type properly rightWhat I mean is yes reversed part is a server side optimization but as clients we dont care unless we get the proper results +store all the applications in one map +Thanks for looking at the issue KatheyIll include the Jira number in the relevant comments when I post an updated patchI agree that the simple fix seems less risky +That sounds good +Yes please +doesnt even complile does it? +Reproducible if panel id! panel +Any load and store statements should indicate the name of file directory that is being processedThe previous version of Pig indicated the name of not only the files appearing in load and store statements but also the name of any temporary files created as wellThis functionality should work in local mode as well has HDFS mode +Failures cause by signing an already signed jar should not go the exception is thrown by default and a user flag is added to allow bypassing the failed the log message is a warning instead of info. +Patch to add +itself is done +I got this working +By crash I mean in the OTP way +I am not seeing the same behavior with RC the plugins and adding the plugins works as expected for plugins folder +Note that these are intentionally not targeted at because they dont belong in the release notes +assuming I remember to check for it +I checked your patch and it works on Tomcat +It seems it is not related to this issue +It does not do anything to prevent users from trying to set it though which seems wrong +Using a simple boolean allProps flag allows for an easy and straight forward implementation +Looks pertinent +Ill check it in +Thanks Liang for the patch and stack colin for review! +Would we force the client to know that ID then? Or do we have a good way to query running replication tasks +should be replicated too +I confirm there is a problem +I think the root problem is somewhere in nettyclient. +By this change will create new session id when failover happends so the request will stick to new node +Now I cant edit + +Im sorry +Really good question +Will know more once the reporters of confirm if that issue exists in the standalone version of Roo or only when used via STS +Closed as duplicate +Now its much less verbose because groovy is only loaded at the beginning and not in each moduleIt also removes the loadfile +logs gets deleted if option a or all is set to false or if option is not set +good to me +I massaged the options a bit trying to make them a bit easier to understand +The problem is not with container pooling but in the way that postback forms have been implemented in this tag the way the tags action attribute is modified is incorrectOn the issue of the modules it will only ever be the current module so iterating through other modules would not be necessaryClosing this is a mistake IMO the only way this issue is not going to bite people is if the postback forms feature is never used +Can this be resolved now since was released +I think I fixed this issue this morning +Was there anything outstanding as I think its good to go forRegards DavidId rather commit on both branches at the same time instead of having one functionality available in one version and not in the other +It was decided that we will not upgrade but backport a patch supplied by the reporter +Attaching +Ive fixed all the problems I can find related to this so Im closing this. +The problem can be reproduced in a test case +Hi Marek you cannot open the same JIRA for both EAP versions +Closed after release. +When using user preferences browser locale to resolve formats can result in different formats appearing in comments pages +How did you test this? If with the site plugin did you set the input and outputEncoding in the site plugin configuration? Can you provide a test case +Fix released on with JBT +If you find further issues please create a new jira and link it to this one. + Username objectwiz +Do we want a generic jaxme plugin as suggested here or do we want to make a plugin for the whole build process as suggested by mavenDo we want maybe pluginsnachoPD I am xhappy of my vacations +All other calls are NOT initializing there is no concurrent invocation of this method with init dont want to make access to the from a Message acquire a lock because this is on the critical path +Unless there are any objections Id like to commit this tomorrow +I verified the fix in trunk by running the repro after removing the code change in CPC +Will commit later on tonight. + +Actually Mike Charlie has just created a branch and all the critical fixes are it would probably make sense to try that branch +I created for refactoring +Please let me know how your testing goes +it is now a class offering the same trivial functionality but at over lines of code +Is this sufficient for reopening this bug +Confirmed that this is only a problem with DUPSOKACKNOWLEDGE and multiple a test case to help reproduce this reopen if its still an to set fix version to again with correct fix version +The class was was moved to and warnings about its standard added to the javadocs. +Committed to master. +the component was released in jackrabbit resolving issue. +I missed this one +closing as Undertow will be used for the servlet support as from +Depending on what I find I may simply the screen shots and stop there for eventually yes +has gone inShall we consider checking in this patch +Closing stale resolved issues +jsp mapping +Fixed by +added javadocs added getStatus and getVersion to thrift interface can now run test on stand alone server by setting in dataconf +I will create a jira for thatOnce the derbyall and junit run fine on I will commit the changes on that codeline too +No response +This does seem to fix it +If you provide an improved indexing transformer could you provide a patch against the existing transformer obtained using svn diff against a recent checkout of branch +cannot be cast to In class generated Version affects field named private String Version +I also noticed that the patch for test are out of date and cannot be applied easily +This was one of the reasons that I filed this issue so that the example code is compiled and we can ensure that it is not broken accidentally +HadrianIf that psgsign issue is still around I suggest to revert back the changes in the pom xml files to as there where for the M release so we can get outAnd its also very important that releases can be build on Max OS X as that is what allmost of all committers useAnd it should be possible for them usother than you to do releases as wellWe need a bit quicker pace on release cycles for Camel +Failure to do so for a consistently active entry in the cache leads to unbounded namespace leaks +Updated the patch for the followinga Cleared the findbugs noticed in testcodeb Updated the c Cleared javac warnings +Have you tried removing the spaces from the name and giving it a real Maven friendly version number like instead of +Hmm +Both Core and Client trunks now depend on LICENSE and NOTICE plugin version Oleg +Committed r +The fact that sequenceids are now clashing was probably always there but before hfile wed just move on one of the sequenceids a suspect practise and something we no longer can do now the meta is kept in hfile +okay thanks! +I dont believe this makes sense anymore in the context of Tapestry. +This ticket does not cover the src distribution +Bot the listener and the sender treats a file item lock as a file with the same name suffixed by the +Use a +HiAs per subject +I believe these were intended for future releases +The JIRA report isnt added +EJB support for this is missing. +build contains CDI feature +The patch adds a name to a Derby singleton timer thread created byThe thread name is passed to a Timer constructor as an additional parameter as it has been proposed +As per previous comment should be resolved in SVN head and go in release. +Attaching log +I didnt notice the patch and the solution to will solve this issue so I thought I could close this one as a duplicate +Hi Mikethis does also clear it up for me now +The failing tests are using and submit jobs +Thanks all for reviewing it +project file and upload the committed patch without it +All set +Thought that this was a different issue. +Committed +But thats not what RPC does today +Is there any situation in which using it isnt a good idea +The initial intention was to display outstanding issues blocking the release but I modified the filter to include all issues targeted at the next release milestoneOleg +Fixed in rev to use a Map. +i still have to convert atom info to image Luis its ok by me to have this going to the REST branch as we have plans to merge this sometime soon +Somebody needs to get behind this and nudge the committers to have it actually make it into +Patch committed in r in slightly adjusted form to trunk +Can I it to you or Marko? +It may also have problem on +Tony I do not seem to have the requisite karma to reopen a closed issue +I own it +Thanks DougAlso thanks to Giri and Bruno. +Im assuming this is finished please open additional jiras for further commits. + you! Sorry about that +Ill open an issueIt doesnt take much for us to accumulate many log files if the upload rate is high and flushing is taking a while or is heldup +Embedding such XML as a JSON field is certainly possible or it would even be possible to convert it to embedded or nested JSON +Seems like we can close the issue. +StefanThere was an issue about expanding the user account but after disconnecting and reconnecting the problem did not occurAs for user without domain I just tried that here but couldnt reproduce it +Closing this should implement a better logic and remove the required snapshot on restore allowing the restore of a non existent table. +Committed revision . +Expected results +Why is this needed? Because yes the exception was initiated on server but you wont know what piece of code on the client initiated the remote invocation +Ill try to dig up the info +Really we need to add to the handle uncaught exceptions log them at error link to for logging uncaught runtime exception was killing the +Are there folks who rely on the old behavior? Id guess there are +OK making responses OOB did the job +Solved while working on +is same as +Because of this server crashes +Its too bad that surefire arent picking up the before and after methods +I think it is the proverbial case of better late then neverPatch committed +It is just a cache and abstracts that sometimes the file will be in memory or a different physical location during application runtimefor your second question yes this does fix Nutch and it is then able to load all of the plugins properlyI understand your concerns and it was hasty of me to assume that this was a trivial issue to implement +We need to fix on the branch too that should be reasonable no +Fixed with r. +Thanks Sam thats greatIs that approval just for or can we treat other Category X modules in this way tooThanksEwan. + +I hava attempted in the Sandbox and this exception also existed + checkin fixes this problem for my test runningIncrease the maximum wait time for tests that start the server +Swap those into place as the live coresThe servlet page at this point contains the information from the import run at step +I vote for a shutdown hook in and then close +Its a significant change and its hard to justify the costs until we get more demand for it +Ive written a unit test that confirms this report +by renaming the unit test configuration file to this code gets executed +Committed revision . +Also verified that the same problem exists in as well +This issue seems to be inline with the discussion in +Feel free to give it an early try +Committed into tez branch +These disabled tests are not very stable unfortunately and they sometimes work sometimes not. +VBE +We should add a entry to notify users that we upgraded the versionAhh right will do. +Thanks. +would that alone bring on a GC? I suppose we should have a high friction on the tweaks +Interesting Ill dig into that tomorrow +I slept on it and before applying it is there really any ways to reproduce the issue before hidding it +I think its essential that the project remain attached to at the moment because I believe many people will want to use it with Amazon EMR which is on +We need to remove auto incremented ID s from Gateway table and Users table since same gateway name and user name can be replicated in those tables if we keep auto incrementing field as the primary key +More proof that the multiplefalse has a bug is that if its true it works perfectly as expected +Plain vanilla server +That wont work +Thus you have to have valid registered user at and use it to checkout the sources from repo +verified thanks +Can try to find out +This will not be done for if ever + +Trivial patch to fix the issue +In the master web ui I see the load as afdI think this is related +Previous patch was against src not +Doug would you like me to try and complete a fully working version of the solution? I was going to drop it because of the complications with unions and the recursive schemas issue but I might be able to fix it +Thanks for the comprehensive list of changes +Try giving them focus and see if the keyboard shows up +FYI this appeared to never get applied correctly to trunk so it remained broken in dev builds +I studied your implementation and learned a lot from it +The patch is reasonable +I will ask Paul +Applied branch and trunk +removed all references to +Thank you for the review ScottI agree would be nice to have but I was trying to keep the initial implementation simpleI like the XML structure you suggested I will update the file and create a schema +I will fix it when I have a chance to do so +Subsystems are outside the scope of this JIRA and should be completed by the subsystem owners +Thanks +Can you post your? Is it possible to post a reproducer +Trying out this changeset thing. +So I strongly believe that being able to always have the build run even if there were no changes is a core part of Continuum functionality +My company is evaluating using meclipse +bq +Instead of individual setters you should create newInstance methods in and + +No +They should be in too since python clients will croak +Which project in JBT core does require Graphiti? If none we should remove Graphiti from JBT JBDS target and make it part of the soa target extension suggested in +that might take a while +David this patch doesnt apply cleanly anymore +The use of is supported in cogroup and join statements as well so I am planning to keep it consistent and support this syntax in those statements as well +Thanks Richard +Does it make sense to be able to assign labels to the rows and columns and maybe even have it accessible as a map? For instance I think I could use these for the bayesian classifier implementation I am working on and it would make sense to be able to label the features and the labels +I think every other minutemoment is OK but it wouldnt idiomatically fit all of the examples I mentioned in the previous comment +Only reported against so I only fixed it there +Consistency check is no longer needed resolvedwont fix issues. +from the stacktraces in the description of this jira I think the main question is why the RPC is stuck + +Check is in +I will do a quick verification +assigned to release +They get highlighted in turqoise or something on my system which is different to a keyword not sure what it means though +Even better than the first which was also good +Lets give it a go! +Rationale implicit downcasts and implicit null return values let you shoot yourself in the foot whereas return type inference and access are features that are essential to boos +Nice find guysWe probably need an ee test for this +I will move this to the Jira project +Is this the right fix until we rethink the configuration location mechanismThis change implies that the file in the classpath overwrites the one in thelocal path but someone may be interested in the opposite behaviour thelocal file overwrites the one in the classpath +Since the fix for this issue overlaps with and since the fix and test for are so trival Ive included them in this patch +ok i will have to look into this closer tomorrow +Verified on +Hi Abmar Sorry for the delay in looking at this +This is a new version of the logo with the feather separated from the marmot so it doesnt look like a unicorn anymore in small sizes +When the attribute is processed the attribute is retrieved from the list of attributes and processed first outof order +kn and then read all keys when needed +Let me come up with some ideas and I will be back to you out of curiosity what tool is it you are using Castor for? + +Sorry to be late to this issue but should extend so wont pass through catches written against earlier versions +Tom since you wrote the Yarn service do you know which of the above dependencies can be replaced or upgraded to fix building in +Ive committed this and merged to +Let us know if it helps with your scenario +Over to Mark to remind us +It cant be backported any further because information used from the system table is not available prior to This will not be a problem though such cases of trigger will be handled by once it gets backported to older releasesI plan to investigate next +so what is wrong with disable that limit during optimize +Todd if you have a reproducible test case can you pl set the value of to more than minutes and then retry the test case to see if it encounters the same problem +Totally coolDo tests pass +Here is the webservicetransformer code +If you have a better solution to our problem I am more than interested tolisten to it +So if two users are in the same queue that queue can steal up to xqueue size etc +Eh good question +The Jetty component creates consumer endpoints not producer endpoints +It only happens because it is a field and there is a check missing +I am new to this but I dont think anonymous users are allowed to view their orders in the current ecommerce applicationI am eager to hear the opinion of more knowledgeable people +This is a major issue that is now causing us much grief +when there is an exception while adding one we dont try to all of them +Thank you Shreepadma +Thank you again! +It is already fixed in newer releases so new users will automatically get this fix +Attach a testcase to reproduce this issue +Logical is like before but it waits for failure fromadd rather than assumes failure +I think the store unit tests failed because my tests were built sane and I was running against insane jars +I need to go to Dell to perform those stress tests +btw IIRC MS SQL server is using go as commit transaction but given we shouldnt really need these statement in separate transactions should be working fine were at renaming directories lets rename to +My preference would actually be to wait some time until the details were sorted out but I can appreciate your eagerness to move forward +I feel more confident now +I suspect you have not assigned any group to the newly created userIll try to add checkwarning regarding revision +Fixed in +This makes navigation extremely confusing +Closed +Heres a first take on this one +Making immutable should be reasonably fast using this changing the PB mechanisms for other classes is a different beast though +This enhances the framework in you please provide us with a patch file which can be applied against the latest trunk version? The patch just contains the differences and doesnt force us to overwrite all the files +Great! +The Javadoc says that transferTo can be more efficient because the OS can transfer bytes directly from the filesystem cache to the target channel without actually copying themUnfortunately only for and other not forSounds like we just need to do some benchmarking but I have a bad feeling that all the checking overhead Sun added to NIO will cause it to be slower in the single threaded case +is FOAF Creative Commons +Greg Adding require jruby is the right workaround but its very likely that the use of happens before your target script if for example you are using Bundler and it is booting libraries for you +As we cannot use the logger to report a broken logging config it should at least write the messageto stderr additionally +did not make it +I may have messed up my build since i did not include the solr jar files in the class pathis there an enviorment variable that maven will use? such as CLASSPATH or a lib folder in the project being built +Postponing to the next believe that this is fixed by r on the trunk and r on the branch. +Thank you +As we discussed on continuum +Committed at subversion revision +for the patch +Are you sure about it? Which product? There are only two surveys from demo data with allowMultipleN and they are not from type CARTADD which means they are not supposed to appears while adding to cartI was testing Mouse Hand Poll with +Fixed in revision into branch in revision +I have verified that the works for me +Id suggest you work forwards from my simple test program towards your setup and backwards from your current broken setup towards a simple fresh install +did you build the target platform and parent pom firstif you mean for jboss tools trunk I doif you mean for jbds trunk I dontin my machine mvn install for the jboss tools trunk is ok +Would be nice if we could test this somehow +I am happy to put in a fix but I would like to better understand what has broken and how it was broken so we can fix the correct thing +We can draw well with the first color selection and width selection +I understand this may be more confusing for users who arent used to this but it is a lot more consistent across the whole product s easy to update the version in the just send a pull request to that project and we will do a release +In conversation Jesse Stack and I have agreed on calling this all instead of +Increase inputSize +So wats the conclusion +I originally had version installed but uninstalled it and installed the latest version to see if was in error +Whats the status of QA for ASWhich versions of it are going to be tested +I think we have already did it +BilginThank you for adding that for me! Hopefully Ill be on the list soonUnless Im missing something I believe both and are necessary dependencies +styling every screen for a nice print layout is very therefore ill add them only in little steps +Actually the seam faces should be deleted as it simply duplicates functionality already provided by JSF itself +What can I do with it +I like the idea of the script command +You have a patch +After some experiments I not sure that I am right and this is a real problem which causes incorrect behaviour +v patch Initializes all Types with null expect void and obj s the word on the latest rev of this patch? Is it ready to go +This patch updates the string for both the and sample wizards +Updated the bug to just be about the thread failures +Is there a reason this fix has not been incorporated into any release of Grails? Why is this marked as Major and still unresolved months later +Added a fix to address this specific case +Thanks all +BjoernUpdates in CVS +Ill try a manual merge tomorrow + for the patch +Another way around would be giving that event +i am using chrome +Would this issue affect jmsconduit the same way it affects httpconduit? Ive been having problems getting a client using Spring for configuration to contact a server listening via JMS +bq +mrepositoryognlognl java cp HOME +Changes have been committed to the branch for the ER release +It is the on the CP tag Vesion Control Flags GA I patch it in the build? Or should or Remy fix the tag +We should track that in a separate issue +Going through the screens I eventually seeNA Transition Issues NOTE You do not have permission to perform a bulk transition on the selected issues +Use Create a New AttachmentPlease state whether you agree to have these classes released under licenseWe are ready to start checking in your code into the official SVN repositoryTheres one important issue that needs to be cleared up before we can move on +Im a computer science third year student who willing to participate for GSOC and i am interested in this project idea +added support of dynamic settings of windowName that now it is OK. +If I understand correctly this is already being handled as David suggests +Please close this issue +As far as I understand the comments above you had tried to use enumerate in the past and it didnt work +actually this was a feature for me is there a way to have different rendered attributes for header and? +There is no build step when installing a +It did the same thing +Get current namespace from action mapper rather than from +Ill have a look at this and try clean it up +Anyone thats a parser expert is welcome to jump in and help. +Why it is needed at all +Now apparently were missing a file +The first is not a bug +Heres a couple of small questionsa Why do you need both parent and parentid properties in? b Is this a result of you trying to missing support for If thats not the case whats the genuine meaning behind these instancemembersd If it is how should and could be modified to be with supportfor polymorphism +This patch was tested on a local noticed the same issue in hdfs script +is usedTony please use this patchAlexei thanks for the commentsFixed failures on all unit tests +A simple test project that can reproduce the problem at least on a machine with slow disk +Only will augment the set of methods to include those for property accessors +Patch incorporating review comments +Hi JuriThat has now been done and will change to the new address within the next few hours. +Reopened until the discussion in is concluded +Ill try to get some of that in first before I send in the pull request +This patch solves the first half of this issue +opening new if need be +if we look at the log and see that there arent any proposals we can know that the server has been reinitialized +Once this has been fixed and updated will maven automatically download this jar and update itself? Or alternatively can we just replace with in our maven installation lib directoryWhere is the class path set to add +Correct there is a new entry to go in one descriptor to enable the interceptor but for the CP we will leave that commented out so customers can enable themselves +I still get the same exception I reported aboveIm thinking this is not supportable +Validation errors resulting from rules in the business domain model are collected in a Errors object wrapped in a +Ill release it today. +I have committed to trunk at subversion revision + now uses the same sessionId to store and get pages tofrom theKeeping sessionId with the actual container session id will be tracked in +Its a problem with the user requesting something thats not there +How about you Eric +I am going to fix in +Im on this type of URL correction being performed in Woden as I dont think Woden should get into the business of trying to correct invalid that it is given to parse +the this a patch against trunk? Do you also want to submit a patch against the branch +This patch solves the bug +Interesting patch but its been hanging around without activity for a while +Do you mean upgrade would not actually read the blocks if we are using old CRC +Fixed as per our discussion and committed to CVS HEAD +dims +Verified with CR and added to automated tests. +OK Ive placed the testcase in lunisrctestjavaorgapacheharmonylunitestsjavalang applied at r +I suggest we do it one instance at a time and write a test case that just verifies the engine starts. +This is a patch that introduces handling for an optional tracking URI plugin specified via a new config that should accept a given and return a valid URIThe plugin is used in cases where a proxy link is requested but the RM has already purged the given application +The projects INSTALL looks good at the moment URL and. +Thank this should be fixed in which is included in build +FDT has a bug it breaks if you write dynamic like that +This is kinda bad +I disagree with item +Oh thats great Thawan I will see about getting committed so we can get an easy test for this bug +It didnt work +The provided archive contains the same tentative patch as previously sent except it was made against the available TRUNK from subversion as of th Jan concerns classes and was generated using the WindowsCygwin port of the svn diff command patchsrcmainjavaorgexolabcastorxmlschema patchsrcmainjavaorgexolabcastorxmlschema patchsrcmainjavaorgexolabcastorxmlschemawriter WernerLet me know if the provided version suitsAlex +got a patch +The one thats a bit more scary is the notify but if you look at the Broker code the callbacks are merely waking up a task runner so its ok for these as well +The remainder of the changes are fairly straightforward whenever there is special processing for I did the same thing for +This is not all that easy to findIt can be updated using the CMS the page is at +For those functions that return understandable result its logic can be implemented +Committed at subversion revision Touches the following fileM STATUS +Ill upload a sample in a moment +Fixed wrongmissing headers in testcases +Take a look at the examples that I have modified also +This will cause two counters to be tracked for every UDF and in your script approxmicrosecs measures approximate time spent inside a UDF approxinvocations reports the approximate number of times the UDF was invoked to add the new file +Should we explain alternatives in the Javadoc +Upgrade to a newer broker version and you wont run into this any longer. +This prevents spinning resulting from requeuedequeue on the normal message queue and allows other threads to get the meta regions back on a fix for too +The full version of tomcat you tested should also be provided +Its close +Alexey please verify +Attach patch v that addresses those issues +Tested python cpp javaUntested php perl rubyAlso refactored the python example to match the cpp and java version +we cant know what methods on the interface should be there +This is where we need multiple toolmanagers if we have multiple views in a perspective things can get scary +It does require the fl field to get detailed info at any time thoughYour patch changed the way we request fields which made it possible to untangle the handler itselfTake a look and let me knowThis probably does NOT play nice with the old admin UI though we really need to decide whether to stop worrying about the old UI and just cut over to this one +Committed revision . + +So all issues not directly assigned to an AS release are being closed +attachment +Attaching a new version of the patch rebased on trunk and based off of the latest patch +Straightforward change that will also simplify jira for SPNEGO +Yes manual is correct +shouldnt we seperate editing facilities from testdeploydebug and monitorWhat does deploy mean for a single smook file +Thanks Tom! +Now that we write small meta data files through the directory heavy directory impls might be used because they are designed for index directories +Grant wrote it originally +On each property or property set it would be also very valuable to be able to inject a Validation object +Is it an include order thing? Let me play around with it for a bit and see if I can find something I like +I should have been clearer +Need a way for the master to send a CLOSE region from shellUI +Im experiencing a memory leak in Lucene that might be related to this problem +This would be in axishttptransportutilsprocesshttpgetrequest axishttptransportutilsprocesshttpheadrequest axishttptransportutilsprocesshttpdeleterequest +Issue was verified for tooling linked bz was closed this JIRA should have been too. +I target the issue for to leave it open for further improvements +the reason for a typedef is to have a type that is globally available +Hi DeepalThank you for the input +SumedhaCould you please take a crack at thisthanks is the test report with the failure +Please confirm the current status of this issue and if its incorrect change it as required +Fixed inAlso patched the which handles the and tries to unconditionally rollover pagecounts from the fromPage to the toPage which fails if the page has no pageviews at allthanks GlennregardsHarry + +User error +The RM will take into account available memory on a TT as well as a tasks specified limit when scheduling +Well hbase does use select so it doesnt consume threads to have idle connections +Never usedYes we definitively need to exchange more before starting using anything +Great job I must sayThanks Ren! +Can you attach RS logs +The issue is just the error message saying ulimit failed to set as limit +Only on my workspace +Using Messaging Scott has some changes for this already on his workspace +Also Id be happy to come up with some sample policies that implement such an interface +I have solved this issue and would like to contribute it back +Remoting is now available +Hey Sanjay +fixed at ea +to Hudson +Could you please verify whether this is the case? ThanksRegards to Didier this issue has been solved with the refactorings of the bundle over the last few monthTherefore I close this bug fixed. +fix bug in add all to cart +I tested to send the request one by one to call the webservice operations for times +Wont fix from an interface proxying perspective +all issues to the CR +Now it seems that it works +Yes looks like I can change the Bugzilla please +Now I have attached the debug output +Version has been releaseClosing all related resolved issues. +That link to jenkins doesnt work for me +I verified that it is doing the same on map task no file from are getting compressed +Steve looks like we need an addition to the scenario as the template does not include the runtime libraries +See has been rejected as this is not an issue. +One side affect is explain will see different number of reducer cuz intermediate file is not available at explain time +Would anyone try this +thanks +Confirmed that the problems lies in the current limitations of the using the +Reopening to move to as was issues after move to in +has been marked as a duplicate of this bug +After the merge the one page Checkout should work without problems +Will leave the issue open. +And even if it did not pointing to a cool feature in some other system doesnt immediately make clear why we should try to imitate it +Hmm i think we should keep the alternating gray shading of the repositories for consistency +Not approved for EAP +please adjust the priority to indicate that the Jira should be considered for CR +QA verified +Defer all open issues to +nits position is already assigned to false for the first test a few lines above where you added it the new test method shouldnt start with a capital T our method names are camelCase not One question on the logic if the last block is but you read from within that block then read will be false +Roman if you dont mind Id like to work on this +With hadoop its ok +i tested it again this AM +Verified that the Relational MED is added to the model retested ok. + +Are there any special items in your POS that require a split shipment +latest patch +Thatll just make it more difficult to use in scripts +Who invented the whole TeeSink originally for whatYonik and I did +We should have the following NN side NN has a newer genstamp but DN reported a replica with an older genstamp +That is what the test case does +Running the java project works fine +Thanks +bq +working on it right now +for the patch +to commit +Thanks issue is fixed and released as part of release +It can be removed if you think its not necessaryI am working on the to implement the modifications you suggested +The reason for the region stuck in transition in this case is to let OPS know it +on patchJon says hes running it too +Unfortunate for users but I guess they can upgrade to or just throw a restart response exception if they want this. +I have the same problem followed all the steps for the upgrade but unable to compile the app anymore +Having said that more dots gives a better idea of quite a few votes for d among the infinispan devs at the momentFeel free to join infinispan on to chat about this with the others as well BTW +Please test it out for yourselves and thanksCheers Chris +As such this change is not backwards compatible and existing applications will have to be adapted +Update takes care of javadoc warnings +Documentation of this new feature is part of the documentation update requested by after release +Not sure what the state is on this +The release notes should mention this tho +fileUpload fileName fileSize contentType valid size local local local local local progressInfo revision +The fix is very simple and will waste minutes of someone to apply with without side effects since the method isnt working the way it is now anyway +Offline Alejandro suggested renaming the state to STOPPING instead of STOPPED to capture that the HA service is not completely shutdown but about to +Heres a patch that clarifies these requirements in +Docfix patch attached for HTI +srcmainjavaorgapachehadoopyarnserverresourcemanager +The only question I have is philosophical as to whether it is ok to do so +Ill like to get some Idea on How we can make the Form to now show existing in Edit mode as a result of this error +patch for fixing the for separating out the patch into YARN and MAPREDUCE patches Ravi +I dont mind if you commit the patch but maybe if theres a simple solution we dont need to change the test + as bdddddbcecfecaa +Now all the files have thei ASF source headerAdded also and where needed and added their reference to filesChecked with RAT and every file has the correct license header +No further work needed + code looks good +fixed in latest commit +I just the fixCould you verify +Please give it a try and let me know how it goes +If the warning annoys you too much please open new issue +Added. +The approach would be a cleanly parallel implementation +The possible plan of action can be to first write test for those parts of the tools base that will be directly affected by the code changes and then use those test to verify the migration effort +Depending on the cardinality we could do the checks in the or rather in the top level collect method or in the Collector +Release Note ESB JON plugin no longer reports on a State version to CP fix will be verified when a new version of JON is available +Some commentsnits Logging object can be private doesnt need to be public +I also attached new +Ok Ill do it +Hi SimonI have a fix for this issue this one looks similar to the issue reported in +Failed test unrelated passes for me locally +El resolving doesnt return +Is it OK if the field is final? One of the two was final before and if they both can be then only a getter is necessary +Added temporary solution by moving the Convert action to the configureChanged fix version to LATER until we figure out how to solve this better +Sounds reasonable indeed +An additional consideration is that Tomcat never had a session id alphabet flag +PaulI havent had a chance to give it a try yet but I have a question whats Jelly current status? I mean are there any plans to release new version of the taglibs? Felipe +Merged for in revision . +I completed a fresh springsource folder vs +looks like +Assigning to Ben to document on the I read the comments correctly the attached patch seems to be moot +I forgot to mention that this piece of code needs to be to be more reusable +We got rid of svnexternals on joni sources + patch +Closing as wontfix will add user to the issue. +Why is a duplicate of +Then the user can open and modify what he wants but can only store the CAS if he uses the correct type system +Tutorials are very important for new comers +Verified + +Once you release youre in a sense offering a guarantee that the package structure is somewhat stable changing the package structure post release is going to cause a lot of problems for users +The tests described in the original bug behave correctlyHowever one noted issueWith some browser DOM manipulation when our tag output is inserted into an HTML document via the innerHTML property however preserves the newline in the textarea value +I just ended up removing the min function from the files which had it +And this fix for cqlsh is pretty simple tooIf you do want to change the interface so that doesnt get reported then probably is a dupe +modulesclientsrctestjavaorgapacheairavatawsmgclientutil +Attachment has been added with description Proposed patch to has been added with description Testcase renders text to file and AWT added new test case for fix in the regsvggentest suite. +I am very interested in this projectFrom what I understand my task would be to implementtest and compare or different which implement DBAASAs an IBM certified Database Associate and Academic Associate this project looks very promising and I believe that I will be able to make a contribution +Im sorry for not adding a would appreciate any help on this +I see a lot of conversions between urls paths files and uris +Some more comments on reviewboard +Die deadlock die! +I dont know if there is an easy and reliable way to know how much mappers were launched so each mapper can compute how many trees he must buildThe easiest solution when the user sets t to is to pass to each mapper so its the user responsibility to know how many mappers will be launched +I mean yes there is a new element in the style reference representation and we have no idea how a client is parsing that representation +Thanks. +You should be able to add a few calls inside code so that we dont send something thats supposed to just be family with a color or test to that checks whether Put does right thing when family has a colonAdds handler to old api getclosest to handle case where passed a on familyMost of rest of patch is like jgrays original +move resolved issues to so we can generate release notes +I cant reproduce with simpler code yet +Putting in limits on number of tasks is relatively fraught with errors and only works if you have exactly one job that ever pulls from the database +i can provide a repro environment if necessary +jbosssxsrcmainjavaorgjbosssecurity jbosssxsrcmainjavaorgjbosssecurity jbosssxsrcmainjavaorgjbosssecurityauthspi jbosssxsrcmainjavaorgjbosssecurityauthspi been corrected by adding the missing options found to your test principalClass was also missing so I added that as well +Juergen +Thanks Vinod! +My bad I committed just nowCan you regenerate the patch +Hi AndreFix for this issue is now available in our mailing list +Im quite new to roo and after reading documentation Im really excited about this frameworkBut I also feel like support is a must for me to give roo a real as I read somewhere voting is important for decision making on the further developement of this project so all you interested please votethanks +Fixed in +Patch implementing the Abdera instance as a attribute managed by aWDYT +contribs dont need to have the same back compat and I seriously doubt anyone is using the old way +Such a patch would include a patch for KB builder which loadsscans libs and KB model which should work with tagx files a patch for KB Page context factory which should be able to include related tagx files and make them available for CSS Open ons +Norioko please dont set the fix version this is for committers only +Proposed patch for this issue +Thanks Zheng! +If you just want to test the writes into then write into it and then test +Linking deprecations issue +Be aware that the default for on fopagesequence is auto which is responsible for this result +Assigning to Christine Li per the JIRA triage on November +I am able to reproduce this error doing the following steps now +ovr file and editing wont mess up this test +IMHO this is close enough and it will be the same +Ive removed the data corruption flag since all the data appears to be intact in the database +yes +Its the shutdown! People will get nasty logs but thats it! The runtime of the application will not be affected +I dont see a reason why visitor methods should be public +Can anybody see any issues with this change? I havent really been much involved in this area +Agree with Snjezana +Parents logs end up in the daughter entries because they are deleted from +cloud also available as varlibcloudmanagement +Note the scripts! script +Should this be fixed for +Committed revision in Nutch branch +Please confirm it has been applied as expected +slip all minor and below +Ok i dont tryed it with c because i dont estimate this behavior and i think it is a bad design decision +No integration tests +But then I dont agree with the requested functionality +The tests take about the same time they took a few days ago on my machine +I couldnt figure out what the initial sort order is it doesnt seem to be doing it by entity creation or thing is the applicationname property almost always needs to be changed after creating a project with roo I believe roos initial stab at applicationname is top with the first character capitalized +for the patch with some minor comments Some lines are over chars +The orginal error says that it doesnt work with EITHER windows svn OR cygwin svn +That looks pretty solid but its important to have tests for this stuff as otherwise theres nothing to keep it from breaking again +Ive committed this to trunk and. +This was only true if a was a single row result set +filipe I know it happen under heavy load but having timeout are here to prevent any unexpected long task +Fixed headers +In the future we should move these consumers to more general maintenance tasks in which case they can run and confirm all the metadata files are correct independent of artifact scanning. +The network server was started with assume this is an issue with the Network Server and not the standard DRDA and allsuites completed without committed in version . +the message id can not be null doesnt help +Most of the actual changes were in the pom dependencies I did have to change a few minor things in the Main and classes but nothing substantial +We found a few Transaction Manager issues along the EAP testing +I think that would be a trifle cleaner architecturally given that findFactory is already a private subroutineISSUE If someone programmatically changes the property the current code will track that change the proposal would break that behavior +marking it fixed in and a candidate for backport to resolved merged to branch with revision +For the time being it is easy enough to work around this issue through using the generic methods instead of the specific queryint +Avoiding endless loop by throwing an exception since the parser is failing due an unclosed comment patch was bad Im working on a new for +there are still rowKey and summary attributes in the table of attributes +This will ensure that only create the Spring bean factory once in the lifecycle of the application classloader and so do not suffer the slow leak +Done +OK I was wrong after looking at the changed I did touch that part of the code with the intention of making it static but I guess I never did it +Fixed at revision +In other words it can be rejected if no one else is as WONTFIX since the reporter is now trying to use . +Do we really need groups? I dont think implements groups The group is your posix login group +does not have an Infinispan connector as Infinispan storage is already supported natively by +need confirmation of jesse or to thanks much +Sorry the bug should refer to Utilities +Please ignore the other patches +Proposed patch merging provided stripMargin and adding stripIndent +Trivial doc patch +I only spent a few minutes looking at the problem so it doesnt surprise me that my speculation might not hold upKishore you never answered the questions I asked in my email namely all the other sessions that tried to create that node got NODEEXISTS exceptions? If they didnt it should be a big red flag +This is a rather annoying problem +Fixed failing unit test on Maven rev +chunhui looked at the patch again + +Patch that implements the binding lookup implementation +Thanks for your analysis Andy +The patch adds labels to the navigation icons +Also marked as Optional +reopening as I am seeing this again in common again some kind of CP error. +We dont want to put it directly in HDFS by extending the namenode since the namenode is already heavily loaded and a central contention point in the system +You can see them by running the android tool +Christian what OS are you using? I am on Windows XP +We wont copy the documentation +duh +Installation should contain in the green rectangle as specified in this JIRA +Sure filing the bug for the record makes sense +Sample be fixed in the latest RC. +and therefore the old service implemenation class will try to be used with the new service +I suggest to close the issueThomas. +Feel free to close out this ticket +An issue with using hex at all is that we cant represent the maximum GB column valueIm not worried about this +Thanks +If that can be captured it would be sufficient right +And since we dont retry sending statethe joiner will never finish the state transferThe fix is to wait for the local node to install the new view beforehandling any state request +I prefer my customer not to behead me +bulk defer to defer to. +Some comments on The primary keys for SYSTABLEPERMS SYSCOLPERMS and SYSROUTINEPERMS do not include the GRANTOR columnWont it make future upgrades easier if the GRANTOR column is included in the primary key? How does SYSREQUIREDPERM relate to the current dependency system +This bug commit went into master and branches only +Brackted an assertion in with a check subversion revision +Duplicate of. + +Rado the above changes have now been made with two exceptions the out of sync output are you able to provide the actual output? I do not have EAP running so have no access to the screenshot update can you provide a basic screenshot? Again I dont have EAP running so have no access to than that the changes can now be verified on downtown +Will fix the grammar while checkin +I still see the issue on the trunk even with empty patch +Please note that the source to cannot be compiled with JDK +Uploading the logs from my machine +One we have it in and test a bit hopefully we can cut a release now +Downing to critical +Not reproducible in latest JBT Beta +This looks like it will require a weld SPI +Claude A ordering dependency in tests around was fixed fairly recently +Patch applied +IDEA organizes them nicely +Thus it complains that it cannot find them and subsequently failsNick has created to fix this +It seems a little odd in the docs to tell people to look for the C source at srccsrc +Ed I dont think there is much overlap between this enhancement and what is described in +Thank you everybody for your feedback and reports saying the problem is solved. +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +That was quick! Cheers +Hosting at would help our SEO adding new keywords to find our website +The test suite ends without catching any problem. +I removed the LOGDEBUG line and committed itI talked to Volker offline the main reason here is having a clear code path and no surprises later. +Would a committer kindly consider checking this in? It would be nice to wean Oozie off of custom builds and depend on +was looking if there were some TRUNK builds available but couldnt find any +Thanks Alexander Vladimir the patch was applied to BTI at rPlease check that the patch was applied as you expected +even followed by colon +Here is the script and generated java files +My guess is that aaime is using pure java imageio and jai and this causes some errors +Thanks Arpit for the patch +WellInteresting I stumbled upon this issue after just posting a question in the same area in the jbpm forum today +The patch broke too many things +Oh +Good plan Ill have a think about it +Attached latest patch by addressing the Ranjit comments +PBUFTRAILERMINORVERSION +Well this stack looks suspiciously like a compiler bug +Are using a custom protocol handlerOleg +Fix will be in place for +Do you already know what is causing and how to avoid a repetition of this problem +Mass closing all issues that have been in resolved state for months or more without any feedback or update +DO NOT USE IT WITH LIVE DATA YETWhen youre ready you should be able to populate it with data +Fixed +That means there is only one iteration variable one condition and only one increment command +Caused by fix to +The question is if it makes sense to also close error and output streams through this method +Tried out and all tests I tried still pass +I believe this is caused by the behavior of reducers during the shuffle when they receive the shuffle header containing the size and then the decides thats too much data to receive right now +Breaks the symmetry of Va and if the coil in V had a few more twists perhaps asymmetric ones? like the idea perhaps even tilted at an angle perhaps towards the back left corner or something +Please reopen if you would like to clarify what you are looking forThanksRick +I have no problem closing this +I can upload the project and could you download it from a link? The single war weights MB please its really an important bug we must solve! would be preferable if you provided project files only not dependencies + +so i am going to leave them there +The attached patch fixes the issue +OK it was just a proposal +for the patch +I tried this XML in a modified that included classesIt leaks +Cancelling patch to address comments +I have committed this. +Thanks a lot Geoffrey for providing the! it was helpful. +The problem was that the test file was built for new version of junit +Sorry I missed the Dimitri +Method should block behavior if component is not enabled +Temporary drop location now uses stringb +You are now able to create validators again without it throwing a NPE under the covers +bulk close of all resolved issues. +backported to +I fixed NPE but we have to figure out why this doesnt work properly for JSP in jars +Thank you for giving me the way how to do asyncronous call back with java class and implement it +We should probably add the +Done +bq +Only comment applies to this one +This is not tool related the mappings doesnt get parsed correctly +As a first step can I ask you to come up with a UML diagram that depicts the current state as well as your intentions +Updated the trunk patch based on the YARN directory changes +that can be used as migration toolThis would seem like working against the goal to remove the old code +never worked on the and changed the page perhaps its a quick fix otherwise its on the list +Committed to At revision At revision Will keep this open until I commit to branch +Thanks +In the case of receiving traps specifying a host of makes even less sense since that would mean trap data could only be received from the machine where the component is runningThanks +I have debugged this issue for hours but theres so much indirection going on that I just simply dont know the codebase well enough to blindly make a change that could break other things without my even knowing it +Thanks. +I think we should keep the existing default folder name +Mike committed this last week. +btw the client retries number for this test is set to +What does first project mean? Why not just iterate until one is foundNow the Project Examples engine searches all created projects and shows the first cheatsheet it finds + AS name change is not finalized so well stick with the name for now and blog ifwhen it rest is minor enough to slip +Try latest cvs +Maybe that will fix your problem? You can download a snapshot off the download page and try it out +yml descriptors then booting TB I suspect if applications are deployed after TB boots a failed deployment wont undeploy the other applications +as far as setup having a couple of standard type templates of agreement with agreement items for leave would be niceonce I see you final version I will try to add to that from my templates +My project does not meet the restriction for promoting artefacts and it cant because it does not build with MavenCan these restrictions be relaxed on a per repository basis +No rush +Nice +Thanks Nicholas +I may even be able to rely on the HTTP Clients cookie parsing manually without having cookies stored in the stateYou can instantiate any cookie spec of your liking and use it to parse cookie headers manually +And another unrelated NPE fixed in r +All the regression tests passed + has been marked as a duplicate of this bug + +these comments belong on the list +Excellant thanks Andrew +Can u attach a patch for Trunk? Then we can run many thanks for reviewing +I fixed this bug in rev +Jiri could we close this issue since there is no exception +Should be fixed in SVN for both and in SVN test case extension included. +I took a look at rb +I think the root cause of this problem is the javadoc started forking lifecycles +This is now fixed +Although during replay the updates are never blocked if nonces collide +Verified on JBT. +yep maven doesnt prescribe the versions of the plugins you use that happens in the POM +We are trying to encapsulate the fact that the results in multiple fields +Yep looks good! +External Tools +Also you undid one of my changes to making conditional on +I will also be checking to see if there are other pages missing with character titleslinksThanks for the help! similar to the underlaycleanup script +Patch applied. +I am back on the topic again and hope to provide you with valuable metrics soon +For some reason I cant post to the relevant thread on that group any one else? +This patch removes the URL decoding done to class path elements in the +I think yes +Two is already too many +I could not get the build to work a couple of days back +It will be similar to the Completely Fair Scheduler in Linux +Incorporate review comments +Fixed +Those changes have already been part of DNS docstring patch and already merged. +Just wanted to check with you first +Not committing this until legal department can finish a little paperwork +I get same exception +We also need to make sure that the is complete and consistent even after the restoration +Since the build now works on the Mac OS X preview for JDK I am resolving this issue. +Let me see if can make it happen +HiAs per subject +I guess I wasnt just being knuckleheadItll be interesting to look at how youve refactoredThanks again. +I prefer reseek as it is easier to see what happens +Thanks for the help Arun and Amareshwari +The last failsafe plugin version that supported this wasI dont know about surefire plugin +Added this to the javadoc but I am not actually sure if this is a legal operation +Sorry to waste your time +Other references to this value seem unlikely to cause such widespread damage to other tables +I just committed this +The API needs to add the validation check and return an error when adding a nd VPC tier to a NIC +Example a user who watches movies is more interesting than a user who watches one on up to movies +bq +But I have patch to fix it for +Perhaps based on and Ooo dictionaries? What do you think of such a hybrid solution +as chris mentioned being able to set headers based on a makerequest is a security risk. +It would be great if there is already a clean way to accomplish our goal +No takers +Hi MartinIt looks like youve attached two patches and +That object created by the filter does not know about of the multipart request nature no portlet windowId or namespace parameters are available at the creation time +Is there some utility to giving the user the list of workspaces in this scenario +Currently I believe this is sun jvm nightly sun jvm tinderbox and ibm jvm nightly +Because the subscription now contains a message queue of messages it wishs to be resent to it when closing the consumer this queue must be emptied or messages will still be sent to the at at r +iv cache gets populated from the local +Thanks for the test project! Ive committed a patch in revision I think everything is working fine +Im in the middle of other things right now but Ill look at this asap +Hi Bryan + is reassigned to RSC during RSA shutdown although was just assigned to RSB by load balancer +Attaching Patch +I did notice a weird thing with the patch +Thanks for the patch +It has also been a requirement of many other jiras such as +Ill test it with partial classes later +So I agree with this implementation +I can build grails from the sources this was from the m distribution I downloaded but I have build several versions in the past so the ivy cache residual might be from those previous builds +Ross Thanks for your quick advice going to do it your way from now on +patch for hudson to rebuild +So a question +Currently we just have JSR on there +The patch attached includes all the latest changes in the trunk should apply cleanly against the rev +Definitely helps to be writing new features like this in my opinion +BTW It would help now to have separate issues in JIRA +Not going to run Hudson no point in burning the cycles on the shell script change +Then the wizard wont have to +Im going to call this fixed as of since its one of those nagging bugs +feel free to use however you like +Esteban was it you that worked on this? If not you can unassign +This attachment is the full patch against trunk +Hmm +Hi can this patch please be updated against the current trunk? thanks +Thanks Knut for catching thatI have attached a new patch with changes suggested by Knut +I think this is more intuitive the way it is +Im not sure I follow is there still an issue or not +You guys have a better understanding of Jackrabbits internalsI would suggest resolving this issue +The original names for the parameters are also confusing it is not clear that trainingUsers and testing refer to the same kinds of items +then we can try to reproduce ourselves +So one of the above hopefully fix that +the modules which are commented out in modules should be built using true +Can you please fill in the Affects Version field +We now include src dir and the shipped tgz is mvn buildable. +No geometryless layers should not be passed to GWC just like they are meaningless to WMS +Added attachment with the fix +I think that resolves this issue. +If I use the descriptor the problem sounds like removing the nodesc option fixed the problem +I have applied Vincents patches for and +Sandy +Fixing Affects and Fix in Version +you can live with the released version sure that would be fineThank you for sharing the background +So I havent seen a case where clean didnt work with the milestone +Trivial patch skipping rb +Licence and Notice doneJacques if you want put image in imagesiconsfamfamfam directory I change defautl icons location on globalDecoratorYes I think its better to isolate them +Also see for reference +is the issue covering the s failure + wasnt planning on changing the interface for the release +The plugin version for the ESB plugins should have been set to +Fixed by svn commit r +Now Ill close this issue as this is not a Woden issueThanksPing +An initial try at this failed trying to use just javac ran out of memory during the Classpath compile ant couldnt correctly specify the size of the heap to the JVM specifying the heap size for javac outside of ant as is imminent +Ill check again but yesterday it worked for me +I think this issue deserves a proper resolution. +We could either use standard or or use the Annotations or eventually use bothI still favour the aproach +lib +I havent run the code but I eyeballed it + st patch +the path matching behavior to use for JAR resources instead of reconstructing +It should therefore not retrieve the root class +Should we consider? Maybe too if I can figure out what is going wrong on it +Yes Im thinking about it +Is it +As well as the bug above here is a summary of my local fixes so far to devcloud +TODO this doesnt work with distributed searchIts not critical or anything since distributed search appears to still work as it always has but the new debug params are ignoredThe following request demonstrateshttplocalhostsolrselectqsolrdebugtimingshardslocalhostsolrIve reopened this issue to track since its hopefully easy enough that it doesnt warrant its own issue +Thus by default streaming tasks exiting with exit status are considered to be failed tasks +Marking this a duplicate of +I think we can fix this by introducing a new configuration property that controls whether DESCRIBE uses the standard syntax or the Hive syntax +If thats a minor change for EJB it would be nice to have it for AS Beta +We shouldnt be trapping +Hi AlexandreI will send a message to the user mailing list first sorry for that +I missed your comments +I just rebased and took care of the comments by reviewers +Isnt it setting the property to true that restores the old default behavior? Just a couple of housekeeping questionsCan the property be set on both a and a level? Im guessing only since it doesnt make much sense as a database propertyAlso is the property static or dynamic +bq +Ill try to have a look +OOM does not prevent from writing to the log +Lets get this into trunk +Looks good branch Committed revision trunk Committed revision . +With this patch you dont need to implement just is enough +Theres a naive caching implementation under contribwebplugins wich one might try out and improve +This patch adds new files for the new testingNote that the zip file also contains a gif image file that needs to be copied into place sincesvn diff does not pick up the binary files when creating a patch +What is the status on that? Im really missing this feature +OK what I take away from this is that I will store chunks not as simple byte arrays but as container objects containing those byte arrays as their payload +Thanks +in nowCommitted revision +scm is h +Return to the link dialog +i think we should remove the logging to avoid polluting the output +This issue is fixed in +result is revalidation in syncExec waits for UI thread availability and ui thread waits for to be unlocked +If a user wants literally a region name like x a user still able to do it by escaping the char with xCThe change will break the existing expectation which we need to document in release notes if we want the fix in the branch +Hi AaronYoure right changing this timeout wouldnt help +Shows that you can only choose custom as the template is correct behaviorActions will only be shown if weblog uses custom themeOnly actions that are not already in use will be shown. +They shouldnt see each other +Can not fix by upgrade +Its confusing to have marked against when its closed because transactional contrib was moved out. +Hey Tom is this the same as +JanneI have a fix on my local copy +Now both the portlet actions and the package actions execute successfully +Please verify with RC. +Destination Queuefoo Message id ID Transport tcp now in version for me so happy again. +every change might be spread over multiple projects which is a logistical nightmare for developers +as r +Verified on EAP. +Ill wait a bit for comments and then commit +the only danger is in applying patches generated by someone else you have to manually finesse applying the patch to get the right history bits flipped before committing +okey so lets mark this as duplicate. +Closing the ticketbq +Pushed to master in dce will be in. +But should not be specified for Java I think +It will be interesting to see how such a setting impacts performance +Im not quite done with this yet Ivo +move to works now that the all jar only contains Jetty +All the mailing list svn references were fixed +I have it as well although I am not using it +Current development is only performed on the your only option would be to implement your own custom which makes programmatic access to transaction management available to your test update the calls arent needed and can be removed from my workaround. +ScottIf I understand it correctly will you implement the change to the in an optional wayExisting war files infused in the old way preferably should still continue to work the old way if possible +Patch applied and upgrade jstl and standard libs to version Jetspeed working nicely now deployed under ROOT +Just like SQL HQLJPAQL make this same statement is not quite correct since Hibernate does convert column null to column is null when null is passed in as a literal instead of a parameterIn order to have consistent behaviour Hibernate should treat comparison with a null literal similar to comparison with a parameter with null value +In any case has tests should be there in all object iterations +xmlbeans sample patch +Attached a patchThe logo is now displayed correctly on the Tomcat Status page as seen in the attached screenshot +WB is integrated into and above so probably nothing much to update I terms of tests I would recommend copying a few out of the WB testsuite as a starting point +Fixed on trunk waiting to fix on gt when is AndreaIm afraid this +Not sure if this is the behavior we want simple enough to change +The root cause behind this problem is same as of +Im attaching a v patch that incorporates your suggestions +just committed this +I cannot duplicate this problem any more +Fixed one set of trace added new trace points into the removed some unused imports and preserved the original exceptions in throws the patchThanks to correct a couple of trace points and readded a trace point to indicate what message Sandesha is sendingAlso added trace points to to enable identification which sequences have been created +And this doesnt mean that we cant work with no released libraries in a branch other then master +Note also that misses import for +It happensxufeng Can you figure why the test fail? Test is probably doing silly depdendency on returned exceptionAlso was thinking later that you perhaps should check the Throwable +I pinged William a while back picked up the patch and fixed a bug or two but havent had the cycles recently will try to rebase on trunk and post soon +Lets chat about this on +Im curious what were the performance effects you saw with changing the frame size +reopening this one since its not supported anymore until we get that acegi provider integrated again in we get continuum updated to use readonly ldap support should exist in details +Lets try again with the then +The issue seems to be that the created for the Webconsole isnt getting shutdown all the time or at least not in a timely fashion when the broker is stopped +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +These classes appear to have been deleted so closing +This seems sensible enough +Thanks Maruan you are right +bulk defer of open issues to + +When reducegenericcast is set to false this fixed becomes disable +Dirk want to give a comment on this +Do you know if we have applied it already +This was generally considered a good idea +Checked into CVS +A new test folder is included with the first simple test code in +it is likely that this is same bug as is demonstrated on some machines by the test associated with +With that behind us did someone start an API migration guideNot yet I think? Go for it +Ignore my last patch in that case +Applied to trunk +Copying notes from mailing list +All our are and we are using generatornativeThis is readily apparent when attempting to save any entity +Pretend that was a driver class +Ill take care of this +Marc do you really have many more deep paths than short ones? That would likely significantly change the way well have to design the annotationI was hoping to avoid having a removePath or similar as that make it much more complex to understand +If you are willing to troubleshoot Kerberos related stuff I will happily help you with specific aspects as well as review and apply patches +Downside of doing what? Throwing clear exception will alert the user that the address is not configured and so the RM will not start +Sorry I should add the version we were using was the latest build these features were only checked in the code base at the end of January +Remoting based on findings +Closing JIRA. +Resolving and moving to fix version ER to check RN in next round +can you include a project setup that exposes the problem? Without some steps to reproduce this reliably its hard to even tell whether this has already been fixed in the impending code +Since nobody will see the warning this is less than ideal +Scheduled for +Fix for disapearing namespaces +Closing the resolved issue. +At present time I want it to not interfer with the regular File component as its used heavily in entire Camel for unit testing +Who is working on writing the test +Perhaps it should be in the documentation or help on the command line +With your suggestion one cant get a raw without getting it automatically cached +BrunoDid not find the time yet this weekend but I do not forget +Richard thanks for the patchPersonally I believe the dsl is too bloated the way it is and we should strive to remove apis not to add +I have as mentioned and discussed on irc offered to do this job +Im going to try something bigger to see if the difference gets largerThere are good reasons to stick with the wildcard approach having to do with documents that arent indexed using Manifold +Hi MikeThanks for looking into itThe current testcase attached just shows how it is handled in case of mapping fileAre you looking for a testcase that shows the differences between annotation handling and mapping filePlease confirm and I will start working on itRegardsRavi +We really should fix this across the board +this is easier but much since we have to iterate over all cacheStatus objects whenever we do the localizeCache +Checkpointing what I have so far Removed writing part of shared doc stores Reading part is kept for backwards compatibility +If this is not done then its the burden of the user to do it and if the user forgets to do it then YARN will probably not give it any containers and the job will be stuck +I dont know if this is a good idea but looks like you can get out by updating the database projectset stateoldstatewhere state think this happens when continuum is stopped and a build is in progress +Waiting for enlightenment. +Nice work +But if people are using this backdoor they will be affected by this patch +After applying the patch I compiled the trunk and it succeeded +listStatus does a list followed by a get for each entry +The change adds reference counters to connection so closing the connection here explicitly where needed will decrement the reference counter and allow for connection cleanup +Ive verified this patch works when applied to +Ignore the previous comment +i forgot one thing how maintained are the japanese versions of the filesi just realized ive only updated but not +I agree that it should be updated should be used only with externl one ds componentsm isnt so +Ok now it worksI happened when mail notification was enabledSolution When using tomcat have both and intomcatcommonlib and lib Dont use mail notification +the problem with the approach is that region move is the normal occurence so if users get these failures frequently they will probably retry in their own code insteadRight thats implied the tradeoff I see is basically between pushing retries up to the client or taking on a lot of corner cases +I dont see how thats a net win +Original issue fixed +If they use Java replication to poll continuously for updates it will leave vast amounts of junk files behind +Its getting very painful +Did you use port for the API testingThanks my mistake +attached test shows the problem very well +If an archive operation runs when a select query is already running the select may fail right? +As soon as I added the quartz jar to the classpath EclipseSTS starts to present the correct javadoc +The patch that adds and to maven repository +Added test +Unix group created website upLet infra know when you are readyfor the email move +patch like a charm! +I cannot reproduce it +After refactoring in html new page needs to hide loading image +and camelcasing thoses gives a big letterJust like HEREISAVARIABLE would be here +I expect that with patch committed you shouldnt see any more output from assert except for in cases where theres a real bug with not closing file handlesThe theory about the database shutdown now being complete is pretty much confirmed since the file deletion logic has failed deleting and tmp +No I havent +I hope IE behaves the same wayThen we add the improvement for Wicket +Thank you David. +Build and later has the fix +Another way to fix this is to change the documentation and maybe remove the returned list of deletes +Please read the javadoc for all methods and for the class for an explanation +We need a property language so end users can use it for predicate and expressions in the routes +I will include your patch also and do a full unit test suite run to see if there are any other issues +I will post a new comment once thats done +Sorry I should have been more verbose in the descriptionAnd yes my fix works with both myfaces and. +Alright Ive removed the check +module has been contributed as part of awtswing contribution +Attaching second patch which adds a test to jdbcapi suite for this issue +schemehostc +Otherwise we will end up with everything in the same package +To crete the examples package run mvn Pexamples that be related to Maven version? What version are you usingIve tested it successfully with To be sure it was not something related to my environment I used an empty local repository with the followingmvn clean install tmphamaand thenmvn Pexamples package tmp are a lot of things fixed and improved in Maven so Id go with itThe second point is correct I have a new patch which handles Maven minor version and handles the main class and dependencies for the examples artifact +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +I think my documentation might be out of date and the target name may now be supposed to be a JMX and there should be different fields used to map like youre trying to do +Uploaded +Yeah +This patch allows you to pass the sys prop boostrapconftrue to upload the configs found for each core in the local +Robert +Are we talking about dynamically changing execution plan based on the table size +This appears to work as expected. +It happens regardless of whether the user is a superuser or not. +Moved to the bundle. +I have added an iBatis driving query input source as well as a i +The external process that launched the job had kerberos auth to submit the job so it will have kerberos auth to query the JH server +our index contains over million documents already +thanks yeah i did add comments to the README about setting GOPATHJun can you verify commit please +Also work for mebut I have an additional patch for reserving semantics of the could you please review it +I see that trunk is versioned at but I dont know if that is because it should be or because it was moved to when was released and not moved backThanks for the comments Nottingham! That would be great that you could help refactor the codes +Moving out to for now which means unless someone moves it back to this will not go into. +Ruleset ignored due to bad in the generated html of the resulting page that element is right at its place and the id is correct +It is cool indeed +This is is now fixed and committed as of r however I am not sure how to get this in the hands of a tester without making another release +Ran the supplied test against the SNAPSHOT no unmatched ack exceptions were produced. + +Justinplease review the attached patch which fixes this issueKind regardsBen +Lets work out the implementation scheme and figure out the costs +Thank you for pointing out Bikas +local tests pass retrying for the third time +My Roller is installed on a machine behind a password protected proxy +Delete first skin DesktopThere is no reason to change skin on those two from user point of view but it change iPhone skin to first available skinAnother even more confusing scenario +Yes the issue still persists the same way as reported +x MOD results in zero for roughly one out of seven values of x +Yeah the nd patch is nicer. +This leads to appending the same all the time +Backwards compatible patch +resetting P open bugs to P pending further review +I think it sounds like a bug in JSF +From the way you phrase the error it seems this is the server sayign the port is in use and not the tooling +Reverting status to NEEDFORINFO as this was not imported from +If things are being edited in the pom for the release then theres something wrong with the release processI can see the check being part of the release process but using them for just doing a build seems excessive +Thankyou +Is there a recommended fix for this issue? Can they use different variables with the join process watching them all or does that even make sense +But users naturally expect that pair allows us task operation and process execution in one transaction +Patch looks good testing +ready for folks This patch is fairly benign any chance of getting a review? Cheers seems OK and jersey folks have good track of not breaking things between releases +No sorry thats my fault +although i should not matter much +A further detail I wrote a test program which verifies the following If I run with a security manager and the test policy then I get a connection but if I run without a security manager then I get a connection +Segons el GDLC s mascul per el DIEC no ho especifica jo tota la vida ho he sentit en femen i per tant ho deixaria en solitari ria +I dont remember the goal of this issue publish sonar javadoc or create a javadoc plugin +Please take a look at the patch when you get a chance +While reworking the tests for recoverable and simulated unrecoverable failures with the updated api I noticed that there are some problems in the test cases I previously wrote +this is actually different to fixed by Jason. +The diffs you see IMHO look like the optimizer is choosing different solutions for the subquery flattening +I understand that most devs are too occupied in other stuff but I think its time that be made more secure robust in this regard since some of the other MQ products already support these things +you uploadedPREHOOK Output filetmpnhuynFAILED Execution Error return code from Is that the failure you are getting when you run with ant testIf so you can find more info in one of the exec logs probably in tmpnhuyn. +We can probably add them now for branch and make an upgrade script from to for trunk +Please reviewThank you +This issue was logged against the drools stateful work which was recently added to the platform it is not connected to drools nor CEP. +I mentioned to Hans that I would be doing this +Committed to both trunk and Pig branch +Its needed to improve CAS client to get referer from request attribute and put it into CAS in sso plugin v +So I think we just need to check whether there is a separator at the end of the root class path provided by VM. +Im still able to reproduce with JBDS I try to create HTML Project from Central +add noformat tags and markup builder definition to allow the fragment to be run in trunk and branch +Great +Patch for branch +and were erroneously left on the list and have not been modified. +well put it into the queue +DaveSorry I left it unclear +After applying this patch I built Derby ran regression tests on Java cleanly and ran a sample regression test on +It seems reasonable to support only and newer in the integration module so thats one blocker gone +Attaching a patch for the test +Hmmm +Ill take care of this soon and upload an patch +Fixed in nightly build . +If a complex number has a component which is a why would you want to classify the complex as anything but? Isnt this going to effect just about every function which takes a complex operandHowever to maintain some consistency with the way java handles doubles it is necessary to comply with many parts of C +How about? The tabbing is off compared to the rest of the Thrift Java library +Sorry I meant at least one other environment but its OKI just committed this +But is a specific instant in timeFrankly Dates are completely broken in Java +This last client pipe that is created gets added to the threaded servers list of tasks and then the threaded server waits for that task to finishI will attach the toy program I used to test this +Did you run the test cases on this patch? Im working on combining and and the tests are failing +That explains itYes please go ahead and commit. +sounds good stack why do not not use ths then if we can do s make it a little less jumpy +Thanks Stepan +Could you please attach your modified filethanks Malcolm +Am i right? That part is yet to be implemented +Ok will wait for a GA by Ron +Test run will finish later today +I presume you mean Interceptors style interceptors not EJB +Writing to HDFS is not a good idea +If the patch interacts poorly with the capacity scheduler wouldnt it make more sense to tell users of the capacity scheduler not to use this feature or to improve so that it can take into account task limitsLike I said this patch introduced features limits and limits +Both block cache and prefetch cache associate with scan and user can adjust as per the need +This patch makes Derby raise an exception if the decryptDatabase attribute is set and it is set to some value other than true +Interesting +Attaching revised version of the patch which adds more tests +Closed. +can we close this +away for next few weeks +I ve checked this with versionStill i come accros this errorKindly help regarding this issue +I will close the issue +Rob could you provide some steps to reproduce because I dont have any idea or you could verify it by your own thanks +bq +Perhaps we should close this issue for now and then see if we can do anything more for a next release in a issueIt might be a good idea to add something like Wims script to our build so we can at least keep a steady eye on this and resolve duplication more proactively +Please let me know if this is a temporary issue that has been fixed +I tried it in STS and there I get no validation error as you say so maybe STS has built in support for the Spring namespacesStill gets validation errors in Eclipse though +the instructions are correct when they mention the XERCESCROOTsrcxerceschowever part where i say that the platform option freebsd isnt mentioned is still documentation about freebsd option has been fixed already since XercesC +This class was deleted +Appears like this problem is not reproducible. +This is a correction to the original patch +Now that we are on Lucene it seems like it would be useful to add these new capabilities to the factory +Is this right +Attached updated patch +It should consistently be tabs +Hi ScottYour patch break the old camel proxy behavior so I introduced an option using option to implement the feature that you wants the testWhen the producer working in proxy model it should build the request with the camel path info like the test just showed you patch just broke the behavior we have beforeHere is the patch that Im planing to apply I also update the code of and components which has the same issue as the +Assigning to Jeff as he is dealing with the associated issue +If they will do it I will make the test negative again if not I will modify the test +Im going to resolve this as WONTFIX since there doesnt appear to be a way to get HSQLDB to flush changes to disk in the server mode +change not to implement Writable since it is not used anywhereThe existing is already serializing proxySource as +But how would the response aggregation work in a scenarioUsing a protocol youd be no better off than if you iterated over a series of connections sending a request to each and then waiting until a reply is received for seems like youd want to utilize UDP broadcast or multicast for this sort of thing +However as Keith initially argued and as PR experience has shown the decision state is an extremely elegant and usefull construct so in this case it pays off even though its a bit of an exception +The second problem you encountered is due to the use of listtoexistingatom +Fixes bug when reading from cells that have been previously written from the same transaction +I just committed this +Also it seems that as a result newjoiner node may accept responsibility for token range it has nopartial data for +But of course +The dir in which it failed will retain the older image until a new one is successfully saved +you may have a current quota of items for a directory tree but the new quota only allows each directory to have itemsI share Boris concern over this +Oh good this is what I wanted the implementation to end up beingInI think there is a bug in the size tier candidate checks +this cyclic reference issue is maven issue +Is it wont fix +If it is kept in local dirs there would be a problem with two formatting unless something like metadata copy is done from local dirs of NN to NNThe intention is that when bootstrapping an HA cluster one must format one NN and then copy the metadata to the other NN +Fixed as part of because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Keeping a nd reference to the Flushable in question when it is only used to pass back to that Flushable for the write phase was unnecessary confusingPatch is the bug fix against and + +I am closing this out since this seems wls issue rather beehive issue +Otis yes we primarily focus development on trunk +With this fix the tasks will poll evry MINPOLLINTERVAL before querying the job tracker +Cannot reproduce neither with nor with of a bulk update all resolutions changed to done please review history to original resolution type +Btw committers feel free to defer committing this patchuntil the community has a chance to weigh in in thesynchronization issue I raised +to delay aggregation to +Fixed +it will get looked at whenever Al has time or someone comes up with a patch +It is not an issue but a feature +It seems to me the correct sequence is to first take down the server then the engine +I hope I picked the right status for it +Pushing to +Edit rule action should be available now +looks good to me +Applied. +Anyway its nothing too serious so not blocking Beta in any way +implemented inThe new package is +Turns out the get issue was just a misspelling I didnt catch think everything else is workingAlso this includes a slight refactor so the help text actually lives in the table class not remotely in +The good news is that look pretty stable +The extends the to support the access tokensecret for delegated oauth access +This bug could be a part of better error reporting in HOD +The third patch contains the comments above. +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Justin commit messages adjusted +Perhaps we can join our efforts to find a solution +finally finished +Im not sure if I understand the use case do you mean that some libs content is duplicated in both the JCR resource provider and another one? The performance impact of filtering all search results sounds worrying as well +It was the wrong patchAnyways my hack seems to be not very accurate for those things I think we should add aggregators in this issue and let it be done via a master task that does that aggregationSo the graph job has a list of aggregators and after compute the value of each vertex will be observed or aggregated and after all vertices has been iterated we are messaging a master task with the aggregated valuesIt applies the aggregation once again and then make it available to the vertices in the next superstep againThis way we could also handle the number of updated vertices it will then just be a specialThe big fail in this scenario is that we need an additional sync operation for the master task +So if we decide to do this +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Please use the target version field to indicate which branch you think an issue should be fixed on +less confusing and to use real binary type +Updated patch adds more tests +One could make the case of not loading a huge project file if that was the last thing worked on but that can be easily fixed by opening a new project +So the issue cannot be fixed for +But with the new public exposure of the field cache parsers this is a newly added trap? You could accidentally consume X the RAM +Ill apply the patch with the added license header for the release candidate tomorrow +Other than that its just maven ant and this question Bill the jars are from the directory of my EAP dist and are pullecd in via the maven support in my current simple project manifestation I did not install them manually as per the simple project page this time so its purely whats pulled in by maven at the moment +batch transition to closed remaining resolved bugs +Output of the mvn help task run. +Please file a new issue for your problem +In the case that the timeout expires we may still be left with stale files if some tasks are still alive and create the directory but I feel that is a bullet we have to biteIve incorporated all your comments +My suped up linux box could cause the previous deadlock you found pretty fast +So you need to change your code for the service lookup or if you cant do that make your bundle dependent to the required service the blueprint wont be fully resolved unless the service is up and running or a timeout happens +Further reduction is probably possible +That did the trick for me i can reproduce it with that last header +Moved because this is new functionality doesnt add to such a policy would be hard The assumption that the first member of a view is the coordinator is a tenet of as had been in place for a long time Therefore there is a lot of code which depends on this +In a word my system should be able to keep running without zookeeper providing the latest meta infoI would attach my patch soon +Committed revision +Doesnt seem to be critical as the retransmission will happen on the next xmitinterval in master and Branch +Could you attach the logs for the error youre getting? Thanks +HowdyI added some manual capture tests now +Please correct me if Im mistaken Vishal +I have not seen it happen since the original posting +From the server log this has nothing to do with WS integration code +Hi StefanTrue but the command option is not mandatory then is easy forget its definition +Thanks again +ignore my previous comment I confused this with locking pessimistic timeout which it is not +we can open up another jira for path validation on server side if you feel strongly abt itfor their should be some client c libarary that should catch unicode characters +Please do not forget to close this bug report once youve assessedits correctness +I am still using old POM to make thins work +Error is caused by attempt to change the data that were changed in same transaction +Added convenience methods to project In classes and tests for each to ensure they have no localization problems +A patch with a unit test +Thanks Olivier +To get the working I will add the values with a to the map as suggested in the java documentation about +i think that this has been resolved by the patch recently committed for +scrapped previous patch and started again with HarmonyI also added the unit tests but I could not get them to run +The best solution would be to see the true cause of the error appear in the Problem Occurred window +trunk +This seems to cause a lot of streaming tests to fail +Please check that it was applied as expected +Sylvain proposed making this the default which Im not convinced of I suspect the performance impact will not be negligible so this could be an unpleasant surprise for upgradersEither way though we need syntax to request batchlog either on or off depending on which we do make the default +Ben can you just put an Ignore on that test? The time generation code has not been touched compared to kml Im only guilty of adding a test that was not there now I have to go to the hospital yesterday a car crashed into my road bike and my neck is not too happy now +is renamed to UI is ready for integration into Contenthub. +If you just want a single entity make its id property always return a static value +Ill investigate thanks for the you sure were doing a select from table without any filter +Feeling somewhat embarassed at messing up something people count on Ive attached a patch which should solve the problem by restoring the behavior as it was in but adding an explicit test for the invalid case where dynamic javascript is requested but no form can be foundI will point out that the logic in that renderJavascript method is not the most straightforward and should probably be refactoredAs I have neither time to perform that refactoring nor to adequately test this patch I provide it in hopes that someone who actually uses static javascript can apply this patch to their CVS sandbox and verify it +No problemThe new version starts up ok but no data got migrated from the previous version so I have to recreate all the projects that I had on my earlier version of continuum instance +The point of this bug is that I doubt thisId like users to have the option of having an editor choice for each page edit + added +screen shot of renders correctly on Safari for patch fixes this bug and improves the layout behavior of the on FF +As long as still works with Spring Data I dont see a point in forcing users into +Marking as fixed Ive tested that server and cascading is now possible along with other wms servers Ive is however still an issue if GS tries to the images from the mesonet server in png it will fail there must be something odd in the way the are encoded to start with +Think about it +Let me check again +I fixed this is JTS +Commited plugin management for the in Rev +Ah yeah good catchbq +Hi Phil no comments on your patch +Tested with plugin. +Closing issues before next release. +HelloI would like to work on this project +Its a known issue +trunk +Nope not yet no hasteBut it might be prudent to move this to then +for Juns patchesIMO testconf is fine using the default value +The VDB sequencer has been converted to use the new sequencer API +Regards +Added duration logging for SQL statements +JVMs default encoding has been used to log output or consoleJVM settings it happens the effect is too large +Verified on JBDS. +When such update performed by a client there is a chance of concurrent update +both are seeing the same issue +we can provide that info in the readme and the script toadd the test usersCarl +For the index to be cross compatible sbyte must be used in +No I have not yet reported this problem to the Cyberduck devs as I didnt know if it was their bug +If you think your code is ready for testing by a larger community of users I could hold back the release for a week or so +But I find the solution to store the old log file names not good +The problem related to is also fixed by this jiraIf you use a newer Tools you wont need any workaround +Attached the diff for the backport I had to do a manual merge as the method doesnt exist on the and branches +So what is stopping the user from using that right nowIf you are just looking for metadata user can easily cutcopy the view definition from source model and copy that into a view model and hack it awayDoing something like that by default seems wrong +committed to trunk +Otherwise the patch looks good +updated the timestamp formatting toe not include the Z character +Verified. +Please review and commit if appropriate +Namenode actually communicates this enqueued recovery block to the primary datanode during heart beat +I just was able to get it to happen with the trunk branch +I have a WAR file project demonstrating the problem but its about MB and I cant seem to attach it to the bug +Again started created pyramid and +Ive modified to disable publication of by default +havent seen it but if the array isnt synchronized it was bound to happen eventually +but this is just a demo not meant for actual use +FYI CR is to to the GA +The ? comes from the fact that the German translation is not up to date unfortunately the Javascript code does not fall back to the default language +component should be compiler C backend +There should be a way to set this on the command line +In fact Im tempted to say that we could add the second release post release +The CVS as of works well. +Thanks again Ivan +Looks like I already fixed it on the branch +Attached fix and extra test +Need to apply same patch +Fixed the patch +Issue closed with release. +Hopefully will have something done this week. +there are some problems with my fix tho +This is nontrivial +Im currently on vacation now but will have a chance to look at this when I return +However since we do not enforce exclusive group ownership for the TT in the case we might end up opening the group permissions on localized files and make it less private than Amarsris proposalSo I am for the proposal above +The fix is in the patch just attached +Recommend that you add more test cases to the existing integration tests for expiration to reproduce what you are seeing +am glad the issue is resolved +I suppose just because I hadnt necessarily intended it to be extended but it is not out of the questionI dont mind just making it +Committed to and trunk +Here is the source code archive anyway +So the problem has been fixed somewhere on the way +He responds to queries on if you want to bring the issue up there +Could you describe the strange behaviour? Perhaps even submit a small unit test for it +Uma if you are busy I can do this for +so it didnt work with Hadoop in standalone mode right +Comments from JIRA bug Triage on Febrary In Santiaogo s hands already The patch to XSLTC was based on code and class was deleted recently by Christine Li as part of a security fix at the end of +If I get a chance I will look further into this and see if I can create a fix for this issue +I would like to commit what I have with JSON support in the current suboptimal IO model under contrib so we can iterate from there in a shared community manner +They want it all to just work +My understanding of the Namespaces in XML spec regarding thispoint has never been very clear +Assigning to Marios just so he gets the updates on issues discussed on chat today +fix verified. +Fixed with revision +Implemented test case to validate drools file productsautoscalermodulesdistributionsrcmainconf Changes committed with commit revision ebdeaafabaccbc +Done but the files in both templates still make references to the xslt method +The patch +Akira tx for taking this upThe correct fix is not to just make this counter but instead not even increment it for the first and then subsequently decrement for future attempts +Close after release +Verified the issue with latest build on upgraded fine +couldnt it be done simply simplifying the way a config provider is registeredAnother simple way could be to use system s already supported with annotation and but IMO to keep the great CDI loose coupling thats the integration of which should achieve it and not the injection point itself which shouldnt be aware of such an environment example +The Wrappable interface exists so that JDBC driver implementations can provide features beyond the JDBC API without the ugliness of casting +And I think as we have currentNode we should remove. +Given that gives if a segment is in progress one could return all the editlog segments to the client +But for example Serge should see an inprovement. +Now wizard notify the page about finish pressed and all validation skipped after that +Verified at +The difficulty of course is that you should not be using that one inside of the code only inside of testing code +But I think the best fix is to add a userInfo field for +its vey weird +Assigning to QE to verify that the change is in the CP docs +To be determined whether or not we should support other file loading mechanisms such as local file system resources. +I have successfully run on the following platforms Java on Mac OSX Java on XP Java on XP OJECTouches the following filesM javaengineorgapachederbyimpljdbcThe new check for empty credentialsM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceA new error messageM javatestingorgapachederbyTestingfunctionTeststestslangSome tests for this case +Removing extra Fix Version implied by earlier one. +The following components would be moved Checklist Colopicker The above all depend on prototype so it would be very important to have them isolated into a separate packageIn would remain only standalone componentsThan we could add later another package +Aargh +I fear I might end up rebasing incorrectly and volunteer for applying top of this rather than other way around +project files! The problem is that Eclipse will not change the project name anymore once you have imported the project refresh wil not help +Thats what I had in mind +Getting this in M is not critical +Trivial fix +Ship it +Patch for for hadoop QA +Yes sorry I was referring to issue ll add to the pull request the change from to was investigating about the field name defaulting to the for The code is actually not using this name as a Document field name to be used at indexing level its just used to match the correct Latitude and Longitude methodsIm tempted to refactor it a bit to make the code clearer but I dont think we need to change this default as it doesnt affect users +However if a subsequent REVOKE statement revokes one of the privileges that have been granted the ability to grant privileges to other users is implicitly revoked +For situations like Object we provide our own implementation in the common library interface +I want to revive this discussion +Ive just wrote some comments on the linked forum Anil still on vacation? I might go on digging this but dont know how much time this could take me may be someone with more knowledge on this could speed up the process +I think the reader could be smarter about it though and check the metadata to fail earlier and with a more meaningful exception +Fix the failed for Himanshu Vashishtha s reviewrename to get to get and add actually get method which contain current hlog fileFor there may be multi hlog writers we should make this metric contain current hlog files +Please review this +Correct me if I am wrongAlso as I understood n this particular issue there is a default room which has the title Video Only and coming with a white board +Ted pinged me that he is out alreadySince this is a Salesforce patch I should commit it anywayWill do so as soon as I get to itJesse do you have a feeling about how different a patch would be +I still want to write a test case +Theyd rather have to release a new version for Spring which is not what we typically want to force them to +I have not been able to follow this thread closely but I dont think I understand how Connection Pooling relates to this problem +Yes the question is why the bundle runtime attribute is undefined +I hit this as well a couple of weeks ago +We simply cant do this for backward compatibility reasons +This will not pass Hudson since it depends on jdiff files that are moved from docs to lib +There are no clean and nice solution out of the box with the the customer wants they can leverage CXF util classes or roll out their own custom string writer to handle the we should avoid using imports in our code +if you change the config file you will need to reindex for anything to change the search resultsCan you continue this discussion on the user list where you will likely get better results? +Please go ahead with this refactor +Not going to happen for +but your example of the acronym is not a good one +So is the reason that the local method is executed AND THEN a method with the same name but with that derived signature is called due to the undefined nature of the call or due to something else +yes this is releated +do you want me to take a look at that +Thus the note about reassigningI wouldnt report for a single appearance but It happened several times before and it breaks the code examples and when the document gets to such state it happens every time I save itI suspect it somehow relates to copy and paste IIRC I copied the code from another docspace page +Ask your system administrator for i then logged in gave the new guest user the Global Repository Observer role logged out and now browse and running searchs worksLooks like you are correct with the constant guest being set somewhere in the codegood stuff made some fixes in trunk r +It may have been some confusion with Spring namespace versions +Adding flag to the close is ugly but you keep the default of no arg so that should work +The diff should be the same in all of themsince your test doesnt depend on any particular network or JVM featuresTo see the problemjava Dframework langTo see all the files that need to be updated dofind trunkjava name +We shouldnt use because it creates a site stating no test coverage +Source side need to handle this exception otherwise it keeps on calling shipEdits without sleepI will submit a patch after verification. +CP fixed in revision fixed in revision . +Yes I mean +We have had EMF logging for WMS and WCS for a while +Dacord +is an unfinished version +Moving to TBD +also applied to trunk. +IMO the situation for the parent resource is minor and the alternative of having separate a separate classification for the children is worse. +Kindly look in to the matter and do the needful +in cr + contains a FOP screen renderer and changes to the Webtools component to enable outputting the entity reference chart to a PDF fileCommitters DO NOT commit this patch it is here for evaluation only +When you click the link provided you see a whole Java source file +well +I was also thinking about switching toShould we migrate if the Extended Session problem still exists? Does your patch solve the problem +Both tested OK +In recovery mode the recovery service assigns a success status to any task that has a FINISHED event reported +Triaged for. +This seems to be the issue from inside IDEA IDE most probably caused by incorrect dependency specification and some classlopader issue +The patch for the +Somebody can use a CSRs credentials to try to reset a password such as for an administrative user either by a scripted attack or URL hacking +There isenough there now but more examples are always welcomeThe examples are currently not being build with a normal build +I am using +This is a duplicate of +well here im just trying to prevent from being crazy classes where its confusing what the methods do +I am looking into thisBoth patches worked in Owens clusterDoug could you apply to the current trunk while I look more into the problem? The first patch is much more conservative and reverts whole of we could do this if we dont find more about the problem soon +thanks +That sounds right +So appending to a block simply moves the block to the rbw directory in the same volume +Wrong hashes for the commitsmaster faad! +Manually test added +I didnt see any log messages relating to failed validation +Commited changes to the branch and will be available in the next release. +batch transition to closed if someone wishes to restore one of these to resolved in order to perform a verification step then feel free to do so +Hi AndreasThe standalone configuration means that CARGO will create the configuration on the other hand assumes that the server distribution given as executable is the official oneWhat used to happen with CARGO is that the was actually overwritten by the one provided by CARGO +Thanks for all the attention to this Leo! +In my tests it reduces the CPU util by without any other change +I think Pontus needs to attach a +Dihao you can assign this jira to yourself +VOTE accepted Proceeding with IP Clearance process +Any ideas as to when a new version will be released with this fix? +Phenomena found in this test will be indeterminate +Consider a change to a package private class like and a test which references that API +Missed controller +Any or Jackson gurus out there +We need to add an interfaces of master and walker server as simple as possible +In the proposed patch Im simply disabling validation if the parser does not support this feature +the only reason this popped up was in cite +Im not sure I understand what you are saying +Attaching patch +It may contain legally privileged and confidential information and if you are not the intended recipient you must not use retain copy distribute or take any action in reliance on it +Strange +Then we need a plan for a jdk build of jboss do you want to reopen this task and talk to Ryan about JDK build requirementsIs retroweaver not an option for JDK? has nothing to do with build requirements +Done. +I think the pool creation and management commands should be part of and the cache commands should be part of +Yes I set to which is less than the number of objects in my file system +Should it not be to make it consistent with other MR configsI was trying to make it consistent with since changing the framework config requires changing the classpath config +so in that sense these are matching methods arent theyI think the request for a more precise warning should be file in a separate ticket + +Once again thanks +Ok this has been committed with one simple but significant change namely that the READVALUEWRITEVALUE perms have been removedThis has no effect at all on the only implemented authority but it does mean that its not possible to implement an authority that distinguishes between writing a column family versus writing the data contained within +Works well Linda could you just Replace ifStatement by IFSTATEMENT in the description as the rule names have changed Make the x parameter a +So this looks exactly like and we cant do much about it +same code can be run from any ide or just java cmd line +We can tell the issue if you post the exception that you gotThanks enough information regarding the issue. +the only thing holding us back is that its an API break +If you insist this change can be done for all methods by a separate jira +Thanks Rick! The compatibility tests completed successfully in the latest run of the tinderbox so the fix seems to have done the trick +or that actually shows the issue at hand +Sorry for the tardiness its done now with the exception that commit mail is still going to Ill put in a asking for that to move to TLP +Looks like there were no issues with the nightlies due to this change + +Guido You will not find source code for +Of course the copyright notices are required in the code for new files +patch which fixes to track the hostnames of the rather than the tracker name + removed such tests from the exclude list +this was committed to assume this is fixed. +TedChanged the javadoc as per the comment +If its going to be unsupported then it needs to be documented +Thanks for getting back to me. +Both mysql and redis packaging hardcode a user name as wel +This patch only contains changes from one of the files in +Rebased the merged version for the counter UUID changes +Seems to work well +Glad to see this may be useful to someoneCheers and thanks HenriCurrently we have a singleton instance in and the singleton is initialized in static blockThis seems to bring a limitation because the static singleton instance is not so flexible and controllableAlso Ive just realized that itself is and so someone can use it as singleton instead +I think some new JIRA also was raised for thisI think filters like or filter can be grouped to new filter type and may be if the filter of such filters return INCLUDE then the should be considering this and then decide on filtering +PranayI think you had swaped the entry for glPlease read the task details again and do the necessary changes in the next patchThe template service given in this task is create and in this the debit gl is COGSACCOUNT thats why i have swaped the entry for gl so i think that in the service create the debit gl will be FIXEDASSETMAINT +this issue is not broken byPrior to you just got a instead of a +srccorecontextskinscommonimages New file Small graphic for the POD link +I guess we can add a toggle instead of changing the behavior would be safer +Could you please provide a test case that I can test with +RFC has been combined with RFC into RFC +I was not able to reproduce this on a +Unfortunately punted once again +Ok Aaron i will upload the initial patch today +ColinI think this should be in s really simple and Ill address it ASAP +I volunteer to do the doc portion of this issue +Cheers Alex +For the benefit of anyone that comes across this closed issue a separate issue has been been raised to address the point raised by Frank Martin above namely adding support for configuring to use multiple attributes for the same class +ATM I would say no +Didnt sync so fixed by hand on repo as its easy to make the sigs look the same +Thanks Ill look further into thisWhere are the rules for the code format +So if we enable JMX for any writable interfaces we need to have a way to turn it off so that those of us that dont want to go through that pain and still have a secure system can stick with Hadoop RPCHTTP with GSSAPISPNEGO +We dont use CD in the classpath and the linux script does not add it +BTW tested on OSX as well +Thanks! +Narayanan can you mark this bug as RESOLVED? Then either yourself or myself can CLOSE the bug +Never mind +I thought coding style wanted spaces when completing a statement on the new line Actually it is probably my reasoning as mutations arent partitioned and messages are hence I use first the partitioning on the data that needs it +task is done partially needs to be moved to the next release +If the problem persists please reopen +Thanks just committed Rowans patch to trunk. +Excellent looks good +I see no reason to backport +But that theme setting is just the tip of the iceberg +Thanks ArtemPatch applied to LUNI module at repo revision rPlease check that it was applied as you expected +In my current module is an apklib and module is an apk +attached revised patchmvn clean install ran successful +Working on fix +will keep you postedIf this issue is not present in the latest code +The comment I just placed on should have gone hereThe first thing from the collections API that I would like to see is interactive CREATE functionality +Include a little example in the help just cause others do so and its nicer +Nah the problem is that its a nightmare to find the thing thats holding the reference to the class that should have been destroyed by a class loader that should have been unloaded by a war that got there were any way of finding the floating reference that would be the savior +I swear I saw this discussion in another issue +is integrated with r +Should I checkout the branch and make a separate patch or how is it done +This was generated against svn r which corresponds to modpythonIt does not add support for any new features found in Apache +Oops +Removed CLI from Components +Any docs on how to use this? +The above comments are mostly cosmetic in nature +The NPE is fixed in r +Just to clarify the infrastructure changes around NFSOur old NFS server has been having performance issues for some time and it was determined that the original etcsysconfignfs setting had not been updated since the days of us having hosts in the QA lab +thanks for the fix Bikas +Previously we are mainly focus on trunk failures +Im confused by the issue title +If we dont hear from Cos by end of day Ill commit this since we accepted his comment and passed Jenkins test and no one else has commented in four days +In fact I think that the mapped roles is a important case to be changed via plugin where admins can grant or revoke access without redeploying the VDB +Is this still and issue? Sounds pretty serious +The services use the new service permission styleI reorganized the methods in the two Java classes +We dont want to change the existing behavior correct? I believe the current behavior for Header would be equivalent to the mandatory false case +WayneFrom my reading I have come to conclude that for arrays the element name in array cannot be derived within an array value element names are not significant and can be anything +or have the latest xwork and webworkis this still an issue or can it be marked as resolved +In other words it is maybe not a real solution for you but on for the release +Test added +Attached another branch fixing the Illustrate +for the patch +Ted my comment has nothing to do with Henrys suggestion re refactoring I was pointing out that there is a similar JIRA to this that should probably be closed out please see have similar concerns as Mahadev re stability this is a big change to introduce into a fix release +The problem still can be reproduced in some in SVN revision +I think I can reuse the one from bug to reproduce the n error +Regardless Ive got my own implementation and Im happy with it +Changing fix version to +The German data seems to be in and the English data for CONLL seems to like both the ISO flavor and the flavor +did i miss the event type for the most recent migration? eventtypes was missed from yesterdays scripts are updated accordingly. +If not please reassign +Thank you very much for your contribution Edward! I have committed this to trunk. +Well throw it out of the finally +the command line arguments on windows will have slight changes due to powershell requirements +The number of data files is rising as is the aggregate of all time spent compacting +I agree with Arnaud the plugin shouldnt deviate from the behavior shown by Maven itself +Anything would be better than the blank said do we need to provide an option on the property page for someone to quickly override the interface? All they have to do now is drag an interface from the palette onto the composite service or reference and voila its we dont have a way to reset it back to an inherited state if they want to go that route +Thanks for the review ATM Ive committed this to +Hm on nd thought using in will mean people moving to a release will have to update their configs to defaultFS or this feature will break +Test passes in hudson. +Looks like it breaks after ant clean on root but the recipes will run if ivy downloads the files after ant test on root +Closed Issue +Obviously needs unit tests and even a run through in a real environment but I thought Id upload it to see what people think of the approach +Move the fix to +Shouldnt this issue also be assigned to Farid +Resolving again +Im marking this as a blocker because effectively the user configuration file is disregarded and a startup method becomes unusable +But it only works when validator is called as a hb eventEmmanuel we need to discuss whether we should keep the complex implementation of that stuff in the event or just ditch it and make the user specify Vaild. +Will you point to where the release notes are. +Still no fix for itPlease if anybody found a solution or at least a work around please post it hereAlso Steve when do you think youll get ot fixing itI use Criteria API all over the place in my code +If we set the allow options we are saying that its OK to not document parameters and return values for public and protected methods +Cancelling while Toms comments are addressed +It also does complicate IR beyond that unholy mess it already isI think its better to have an ability to add any kind of component to IR and then implement whateverCaches over it +Patch v from review boardSorry frequent update +Ive fixed this for now will be available in tonights for fixing the issue so quickly! +I have tried to build swt by myself but no luck +The Numeric type N without a following number should be treated as a N +The SVN is changed after last patch was submitted +What version of Maven did you use +ant example does not build additional libraries in dist and contrib directories +get the source code and edit it to generate what you need +using the wildcard in a schema with no tablesviews ij describe IJ ERROR No table exists with the name null Id say go ahead and commit these changes unless there are further comments +Need to incorporate an integration test for this issue to verify that its fixed. +seems reasonable to me +r on trunk. +Ive modified it to make the presentValue be true +Another solution might be to pass a flag to requestSeek to indicate that the Bloom filter optimization should also be disabled when doing the initial seekThat seems to be correct approach albeit a bit more involvedDo we only reseek to a row key during the initial seek? Can it also happen when a scanner crosses a region boundary +Thanks for the patch Doacan Gney +Part of the reason would be that ecommerce is now moved in specialpurpose directory +Ok I figure out my problem +On the other hand it is present on typical installation +we need to fix this line +Er never mind I see that you did specify for error message sorryHowever it still doesnt yield this message for me +new CTF test case changelog entries et alias + +committed the patch +ThanksAna +the original designed called for the build number to be linked would that be ok +updated Fixed For field +Thats the price to payThis will require to a metamodel at init time to be fast at runtimeAlso I think this should be an option disabled by default if we find that its consuming +I wrote a note to suggest this NOT be added to because its only basic and its apart from hbase so we shouldnt have to hold up hbase to get it in +I think arguments can be made for either approach +Cant reproduce the bug with your data +Taking for +OK Im wrong +Well we maybe could have both +The message is changed in Betad +Ok I see +Also for the record I also tested a Java Web Start demo which works in Mac Lion and this is not a Java issue +in +Debugging indicates that contains a managed datasource +No I meant that the fix as described in my first comment is in the attached +Patch committed to trunk +Here is a new attempt with the latest trunk on jruby and the jffi like tests dont pass because a launcher gem cant be installed +Now that work is underway with AS all previous community releases are +I think we can close this bug now as tracks the issue with Cross +LGTM Go ahead and add another ticket for nodetool work +I did not find sed command pattern which returns greedy resultsThanks blockers and critical issues are resolved but not verified +restore problem for inputfile is fixed is not included +Uli thanks for your quick response sorry it took so long to get back to you +Passing to Snjezana to see if she can within. +Draft documents to define Jar Archive names for Jini Service Developers and deployers +Max yes +I removed the surefire plugin as its no longer needed +Attached patch for this issue included patch instructions and zip file containing new ant task +But I think youre right that it has to log it as well +Branch name changed to this out. +Need to look into that +The spec has given preference to cif over ajax and can be corrected otherwise it will defeat the purpose of ajax +The patch you submitted also didnt catch all the elements after the last + integration module provides an anemic session feature which ends every HTTP session associated with a request for the sessions not to clog up a server +b +Then a flush is triggered but because the entity manager is closed a completely different em is opened to flush but since it is totally empty nothing happens +They should also handle Long unix times and will in a next patch +The only difference is that on it try to do the partial write to channel and here is to the file +I did one small thing the simple themes needed the datepicker stuff too. +patch to fix the extra sleep +However there is still no integration of Quartz or of jBPM business calendars +enums created by Groovy are implementing +added comment from matteos review +Thanks +oh are you talking about the Built project or the root director +Either a parsing issue or the fact that empty statements dont make it far enough through the transformations +I was just trying to see if we can get away with minimal changes for this patch alonebq +Reschedule if these happen to get fixed will not be a moving to for now +Well it is a and the error happens in with variables based on a hasMany +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Even with generatetrue the content does get marshalled but it gets placed in a protected field that can only be accessed via reflection +rebase on good to me + In FMC create a new profile that extends default drools Add the features repomvnxmlfeatures Add the feature Create a child container with the profile +It has been fixed for a while though +That is why I raised this Jira +Make sure session info is included in new request type and any modified request types +Most implementations currently have either of the two methods empty +New Teiid ConnectionDDL Importer should work for this import if is of type JDBCWell need an running Intersystems cache DB example to test this use case with Teiid DDL DB type is already supported on the Teiid side and is supported in Teiid Designer through the Teiid Connection DDL importer. +Will there be a patch available soon? Changing out domain model objects isnt a quick fix +As far as I can see this is platform independent so on Windows all sorts of assert dialogs may appear and stop execution of CC +I have made the required change thanks +we should probably move the encodeEntries method one level up so that its shared? +I just committed this +That proposal adds a diff per which has been snapshotted to keep track of changes made to that directory across different snapshots +closing as they were marked as OUT DATED +what we had before was that I did not even need the dependencies downloaded to build the code I guess that will still work +bq +GPE issue but again needs to be investigated +Need to be tested by David Blevins on freenode before close irc notifier works fine now on freenode or servers with SSL +I updated this based on RB comments will update RB shortly +Put all named templates in other stylesheets to be includedby the main maybe even make the main a mere shell with no templates +Aleksander Infinispan provides better concurrency than PL or OL +Are you otherwise +Attaching the latest patch +Works for me tested under high stress +We could just use the posted as currently done to cleanup the other tests without timeout +However from the common standpoint such widening of permissions isnt advisable +Chuong could you please tell Nikolay where he can download the product that you testNikolay try to reproduce on ecmdemo with the exact same product as Minh Nikolay any news about this issue +Patch applied. + +This is a configuration question rather than a code issue so Im going to make this as closed and invalid. +bq +Thanks for running tests Ted +Same here +Bumping to unscheduled +Hi Trustin I think is really a nice idea to provide such a serialization file so heres a quick implementation I made +Heres a patch that implements this with tests +fill out the form with a userid name and email address +Definitely a good idea +The columns would be poolname role +What I see as a problem is that the patch introduces a new configuration property that is not available on trunk +From what I can tell this should bring the doap up to date for the current status +If I understood him correctly he already implemented something which can be extended +Opinions +How are we for test coverage of the? Is this for or trunk +check in to trunk thanks Shenghao FangHi JeffPlease help verify in new geronimo build +Applied changes to and partially to +Sorry Im not willing to put this one off +Committed r +Please confirm +Whoops I spoke too soon +Ive committed and resolved it as fixed. +good idea the call inkfreelistnewinkfreelistfree can be significantly reduced by making the as cache of global freelist +I stand corrected +This affects support case should work with the new transformer techniques +Thanks for the update! +bq +What you checked in paves the way for all the functionality I wanted +Not seeing this in my installation will attempt to reproduce +Uploading version of Christophers patch that can apply +We could introduce shims into +Additionally make two fields final in theImplemented in Rev +Havent heard back from QA so close for now. +The title should have mentioned crash thanks +It should randomized to assure were really sorting the data and not merely passing it through +This has been fixed +applied the attached patch. +This patch fixes the conflicts from Nigels patch for moving test files out of tmp +So I proposed a compromise to handle that as a separate issue and continue but i got no agreement +And while you can do jarjar etc its a pain and one more chance to do things wrongEric +Should it do that +However I may keep a convention that separates the naming from the associated Accumulo java types +Build performance decrease should be minimal. +vin should be not protected no? I dont think this should appear in javadocScott do you agree with Thirus latest recommendations mine? If so will you be able to make a new patch today +If thats not possible then your proposal isnt going to help muchIm not opposed to starting the process of tightening up the UI so that it is secure in an open environment but please do understand that it is just the first of many stepsAs for the approach you propose I would like to think through a canonical implementation +closing as run panel now edit the no more and its not dependent on the assembly plugin configuration +The reason this caused a problem is because sometimes the map weight could be infinity if the job had no runnable maps and then we divided by this in update when calculating guaranteed shares for reduces leading to a in both the ceil and the floor as I said +it affects only trunk import now supports distributed as archives +Whats exactly the issueThe stuff +Can you rebase this all into one patch? Im ready to commit but hunks are failing now +Also trunk now throws a nice error message +Okay I finished the basic PST Tika parser +We did some intensive internal refactorings on that and its likely things have improved on that front already +Will be included in Wicket. +Only potential problem here is that if the connection drops while the session is open the user will have no way of reconnecting without restarting the client +Eric attached a document whose reasoning seems sound to me using the graphs you providedbq +and I can still see some mv javadoc jars +no problem +Also storing the assigned in the to avoid parsing the +Regression test +This way is usable by the widest audience Im sure the AS Class Loader can be setup that way +Note that these are intentionally not targeted at because they dont belong in the release notes +Trunk has bulk assignment +So what seems to be broken are the changes I made to HOW to allow null to be passed +This is just to decrease the number of clicks needed before the errors occur +But in that case if the user does enter next time heshe opens the dialog will not see the that was entered earlier +Here is the +Nick just looking at your amended patch now +Sure it might make it faster to leave this out but the first time someone happens to hit this problem theyll go out of their minds trying to track it downI think we can make some improvements that will make this patch a little faster +Ok in that case Ill strip out the path changes but leave in the description changes + +could comment here as wellplease find attached a patch for the outlined solution adding a maximum phoneme parameter to the engine that limits the number of phonemes processed returnedBy now I have assumed a default of if the user does not provide a value himself +fix at revision we wont start the fragment bundle when launch the frameworkFor I agree with you +Good on you +I never see the before i thought it was enough with JPA into the taskserviceI will try today this service and i confirm to you the resolution to close the ticketThanks a lot for your helpNicolas +Committed to and trunk +It looks like final fields are writable from any method which is called from constructor +committed +Closing issue Oozie is released +Super annoyingAlso an issue in IE happening in IE the cursor disappears when hovering over the text making it impossible to highlight anything +Yes +This way an invalid POM would be detected at the start and would avoid error messages generated by called plugin goals that arent always the current problem is that the XSD isnt final yet and apparently still needs some minor modifications +Can you ask Chris to do work on that on +I just committed this to branch +Hemmm +I started to work on a patch and will post it here for review before committing +Yes thats correct. +Ill email to discuss. +rev add for shared and modify its pom to run the pluginrev remove individual and from shared +Not a big deal as it does not make much sense in the real life but I will have a look +I dont have to test out with though but if you do it would be good to know if it resolves this issue. + +All changes have been merged to geolocation branch +As written we look at the inputs to the sampling job and estimating reducers for the successor based on those inputs +Though it would be a good start +This is bad +perhaps needs to be posted else whereIm working with Solr using the to store multiple locations for a documentif geofilt or geodist doesnt work with the is the only way to add this patch into solrIm using a tomcat and solr and jumping to might be a while even if its released soonIm not real clear how to apply the patch as I would need to create the file from the source +Saminda has fixed the issueThx Saminda +There doesnt seem to be much action around but it should make this a +Patch for new implementation +They said ther were seeing the same errors +Thanks +The problem is related to resolving artifacts in the same reactor +These problems are outside the scope of the documentation I am currently writing +Looks good +I guess Im not quite sure how to add it to the shell that runs the build +In fact Neha was not able to reproduce it in the branch according to the thread posted above +I prefer keeping information about it in the respective classes only +At any rate I am new to Solr development +You dont need to change the extension of the page filesThe Web tools in the Eclipse can manage +For that I use the option Keep native on layer projectionThanks +When this new compontend will be relased? Can i use this one in the currently version of tobago +Same as +You are right this is really weird +N RF +I think we havent published secure jars yet +it would skip past any File directives and go direct to the Directory contextThe only trick to solve is the best way of dealing with wildcards or regexes for directory +It will be released soon! Please check the dev list for progress. +Heres a new version that includes a test +Ill go ahead and port to the new API +Fixed in r + +Could you please upload it as an attachment +Thanks Fernanda for the testI will prefer to keep s scope at the package +Ill try to check it before the release +Access logic looks ok so Im suspecting it was the btree update logic that may be at fault +Just caught with all the discussion at seems like it isnt fixed completely +This is a Chromium bug on Android we may revisit this later when we fix other iFrame issues +Thanks Knut +I was not able to reproduce this error however I tried +If you dont know whothat is just use a name for the person who owns codehaus +Please ignore the fix for hcommandLink +I dont want to make many changes to the source generator while Arnaud is working on the binding stuff so Ill let him fix the bug so he can fix it in the CVS and his code without having to merge in changes from the attachment is a zip file +The check at the end of the tailMap branch of code in internal didnt make sure that the found row was still a match to the row were searching for meaning that in situations when the timestamp never matched for the found row but there were following rows if the first following rows timestamp was less than the search timestamp youd get that row +Actually this one is duplicate of the +The web source has not been modified to reflect the changes +Please let me know if you can reproduce the problem using this data +Using a custom serializer could potentially also increase performance as some later steps could be merged into the initial serializing dont see how this differs from which accoriding to the doc means no limit +Let me know if you need anything from my side +The only thing that makes me push less hard is the fact that everything is happening over stdio +Moved all code related to conflict resolution to a new package conflictresolution +Could be used as a for needing to modify a rd party category B dependencyIs this temporary until the patch is applied upstream +Patch to parameterize behavior when server closes connection instead of specifying trailing zero byte code line is no longer being developed beyond fixing critical bugs +OK Ill wait for those changes +Should have been recorded against in the trunk under revision r including enabling the tests and fixing the test comparison of the byte array contents +Thinking about this for versioning of RPM projects +should generate a new nonce whenever a new op is sent to new client retries on region move that would allow skipping the append of nonce to theI think that would reduce the complexity of the implementation +Besides that from a practical standpoint its much more difficult for a nodemanager to verify it doesnt need to localize anything when the item being localized is an arbitrary directory tree +Committed to trunk and +This is no bug either your database is offline or timed out +added tasklet job that deletes files in a given directory +Ended up being a dependency issue issuing the from the MANIFEST. +It looks like EWAH didnt work out in but perhaps its time to revisit our options +Going to commit to trunk and tomorrow if no objection +The configuration steps for also apply to IBM JDK + +As far as getting the schema in the output format that will be the job of your to store it somewhere that your can retrieve itThe path passed to the is the full path of the file to be written not just +Sorry took some time for me to get back to thisUpdated the patch to trunk +If we go that direction we might decide we dont need callRaw at all since this is the only use case so farEither way I can commit to cleaning this up +Thanks Ken! +Shouldnt be too hard to write a unit test in that wraps the container allocator in a service and verifies when the service is shutdown that the staging directory has not been deleted but is deleted once the application stopsAlso this JIRA really should be in MAPREDUCE +What do I need to insert to reproduce? I assume it doesnt repro just after CREATE TABLE +Thanks for digging into this +Hey Cam agreed but Im also fine for this to be the umbrella issue to get it working +I assume this is included in Web +We end up resolving again and the system property was not consistent I think for this particular issue the best workaround is the one I mention above explicitly pass the dataDir that was detected at core started rather than doing any re resolvingYou lose the ability to update the data dir with a core reload though +Thanks +I guess this process is broken by the patch +That way after clicking close button page expired error should be displayed. +Related to? trying to ve added some damage control here logging and npe protection +bq +Verified on +IMO it should be configurable at job levelPlease look more closely at the static methods Owen suggested +This issue also occurs on tomcat +on commit Todd +Cool Ill plan on checking it in tomorrow afternoon then +Im hoping that Jens can turn this back into a patch +I dont think it is a release note issue as it is being deprecated in the next release +Cannot reproduce in trunk +applied +Verified. +we will create poms with the old naming pattern. +here is the patch with changes to the +I wish to revamp the whole design +The replication handler calls force to ensure old index files are deleted after a pullin that case if you promote your slave to become a master it should have no problem +No problem Cameron! I guess I would have immediately noticed if you had specified the affected version as given that you marked all sorts of versions up until I just considered the current ll mark the issue as fixed in which is where Ive introduced the check in +Posted a patch to get this conversation started again +Please have a look at it +Please review +Unnecessary given the switch to maven. +Fixed in r +Cool I attached a rough patch with some logs to show how it allows lease recovery within seconds +WSRP admin portlet +added s ok with me +Note that the package refactoring occurred only in master +commited +Can we backport it +Hilarious thanks for the efforts. +I tried Bigtop rpms and it worked. +needs backport in for branch +Helper methods should help +FWIW this sounds like good behaviour +for the patch +Thanks very much! +Any resolution to this or at least an understanding of the cause? Were seeing the same issue with +How about a symlink +Imagine if we made it an exclude and then the line changed to something that could throw an exception +Ill edit the message above +a much faster UUID generator +But I REALLY like your testcase +oh well +Closed due to the release of Tiles. +Thanks committed to and trunk. +Looks good +Marking blocker since the smoke tests fail on windows +And I think in a application where you cant afford downtime to redeploy very important +Ive just committed this to the HA branch +I have also been waiting for to be resolved +Can you please suggest a workaround and sample code to work with +Will attach a patch momentarily +i had already tried this +In the past tests with derby will pass but behind the scenes and it will not work with mysql or will not work correctly with the pooling implementation +Implemented as explained above in revision the following properties to point to the launchpad directory in revision after release. +patch has run before runningthis means instead of failing only if you have forgotten to add a +For whatever reason it never dawned on me that we dont have to worry about a Span that has Terms with different field values +Print owner better since its an Option +HelloWould you please help to apply the patch? Thanks a lotBest regards Richard Changes committed in revision +Can you update the patch to just spit out a warning for mapred groups just like the case of mradmin? Tx +This is based on s the idea we need a with all the common methods from the specs being generated +These reformatting leads to huge patch and makes it hard to apply it cleanly not to mention that its just hard to review since important changes are mixed among formatting +The executes a of patch so Ill add a patch that just moves my check until after setup is called +Thanks Andrzej +Werner bug has been created for the cache synchronization with JMS messages +bq +Updated with latest comments as well as schemas and samples +I was able to get pass onWill integrate patch v if there is no objection +An alternative solution is select from tablename where is not nullps +Sorry for the delay looks good to trunk +It looks goodPlease upload a patch for trunk reverting the umask fix alone +Could you make sure the others are linked accordingly? These would include Eventing over Hot Rod Transactions over Hot Rod MapReduce over Hot Rod Querying over Hot RodWe have for at least some of the above +Do you have any benchmarking to measure the performance improvement? Maybe a that we can run in the unit tests +to be implemented since +The problem in basically is that added functions are lost over time due to garbage collection +Barry heres another one +Cool do you know when byteman will be releasedMany thanks Andrew +As always if someone wants to take on this work they are welcome to do so at any time and the target release can be revisited +I think you will get a lot of mileage out of that simple change +I committed the patch to trunk. +Does it update the bundled python thrift for cqlsh as well +Raphael I think we were commenting at exactly the same timeThat sounds like a fine workaround and if you want to move them back into the jar after RC that will work too +The queue and any messages it contains must be keptRegarding not showing in the console we must do this so administrators can deal with them +This happened in but not after of changes +It doesnt take much to have the point at the latest released upstream JBT update site and latest released TP +btw the problem is not caused by added vs +Thanks for the patch +The proposed patch provides an implementation for the proposed additionFeedbacks suggestions +Merged test into checked into trunk +Ok nice then lets remove the slack parameterYeah LBFGS we would need to implement first when it is in place we could start thinking about removing GISbut I guess having both for a while is also nice to compare themPlease open a jira to document your request for LBFGS training maybe with a link to a paper +Thanks +That will tell the person that the problem was with Quota +fixes the ORDER BY +Thanks. +Closing per Daves posted comments. +Fixed in the branch +Anyway this cannot be considered a bug anymore I guess +Try simply replacing with and close this bug if things work +This script contains the following changes +Its not only about the upper caselower case convention but also the case that the operationname happens to be a java keyword +I suppose itd have to be added as an AS module or something along the lines? Remember that such code would need to be available on startup so that cachemanagercaches can be started and so dont think this can go into a WAR deployment or similar. +The fields interface is indeed an improvement since iterating over maps is awkard in Java +The patch for Endpoints in is being ported to Java +Whats left is to use a with the sshd server +We have a total of secondary indexes on one column family +If you choose this approach then only must be patched and no API changes are required +very small patch +I think we should drop patch available status since we dont have a real patch here yet +that location wont be correct if the user have applied patches its fine for a first cut but just be aware these locations are not reliable +I dont see how this can be anything but a stateful object +Maybe you should produce a one pager with a well thought out plan that covers all the corner cases? Gather a couple of s and we can proceed +Shawn just pushed this work +patch applied at with. +Manually verified that the secure datanode logs are being written to the appropriate directoryTest failure is unrelated +I hope you get the gistSo within this task I just refactor out the messaging we currently have in ourThe overall design has to be revisted though +Im seeing this also in confounding our schema check verification +Although the source fix to this has been included in solr I needed it fixed in the earlier so Ive relocated the patch back to the earlier version +for the patch +Should be XML component affects view is that will provide a much more flexible solution to XML aggregation +Still unassignedunresolved +Committed. +By the way should I apply both patches or just +This argues for access time to be stored in the file system itself +Sergei Our experience with Maven and version ranges is substantially exponentially slower going from seconds to minutes depending on the depth and breadth of versions involved +reopening +Obviously we are the settings in +Increment the Counter in the sentMsg Method in Vertex If messages are combined it does not show up in the Counter +I dont see any need for two different implementations either +Verified +It was the only alias that ever existed and were never able to create their own +May also affect may also affect atfselectItems should refer to collection of objects not +So i simply removed the variable from and added it to Fieldthis is mostly cosmetic but coding bugs can arrise from these kind of variable hiding issues +I was just profiling some Kafka stuff and noticed the CRC eating up time and since I the pure java hadoop one was just about to file a JIRA hereOn a related note it would be nice to offload the CRC and decompression to a different thread than the users thread +Talked with Bikas offline still remove default zk address as its out of yarn controlAlso changed to use zk multi API to do multiple operations +Once I can get past the hurdles of creating my first patch from the git mirror Ill upload it +Can anyone confirm that? After confirmation ill close the issue +bq +added which contains improvements for the tail command +Restart Geronimo server +Assigning to phuong since hell know if a simple check to avoid NPE is the only thing needed. +Handing off to QE for verifying the resolution status +JaredI also volunteer to help with documentation maintenance +Tested and committed to all major branches. +Made the tests pass +Those encountering this issue can either fork the branch and fix it themselves or update to a supported version. +Indeed get needs to use the original property name there in order to cover indexed properties as well +The colliding bundle is queued and will be deployed once the bundle thats currently deployed with the given context path is undeployed +I can attach a patch for this +Joe but did you use theIf you use the return type it works but if you use return type I get unable to open content URL contenttested on nexus and nexus this is the same root cause as are no conditions for it to be a blocker +Got it sorry I misread the code there +He did setup a standalone test case and it workedboth uner That detail was in his original message on +There is the same behaviour in other Eclipse views +This patch changes the initialization of the attributes to avoid the problem +I cant think of a situation where having provides a clear advantage over the suggested by Hoss Man though may be I just havent come up with the use case +the issue was introduced in rev it seems +It is a central database used by all project teams and managed by a different team than oursCan anyone provide a solution to this bug +This issue IS in release +Termination hangs no longer recreatable with latest SNAPSHOT. +this looks good +you attach a test program that replicates the issueI also notice Netty exception in the logs I will ask Trustin what these mean +It is why Ichange this ticket as an enhancement request +Again I am on this change +Closing this issue. +but i dont want to write code that silently fails depending upon environment +Marking as resolved. +eliminate everything about FIELD support fromRemove FIELD support from +Rolling this into the next RC tag +If you decide that the current behavior is preferable to changing things then I think that the query UI needs a highly visible notice that requests to that shard will actually go to the entire collection unless they make distribfalse and there needs to be a checkbox for that parameterEither way I think its a good idea to make collections available as entities within the UI +Cool thanks a lot +Hadnt realized it was a still needed property for local execution +If I understand correctly this will be different from what you got above +Thank you for running the tests and for the subsequent pointers oystein +Proposed changes +That saidI could not reproduce this with master and +Joyce +Assign back to Rich. +Camille Gavin demonstrated that this is working you will need to adjust your patch paths +When I open up this other file the logo shows up +Thanks Arpit +Works fine! x +The only change I made was to ensure that all sections are closed by default since Im not quite in agreement yet as to what sections are or are not absolutely necessary +Let me know if youd like me to try something else +and also how long that took to look at! +Thanks Jing +Hi SupunThis will not require a machine to fix but will need a machine to reproduce +Bulk move issues to and +for send shutdown command with protocol level whether there is username and password or not +Updated to Drools +Heres a sampleI put in support for both i and j +Could you please get the IBM JDK from as we discussed +Whats the issue here +Hum indeed but they provide only half of the work +Same patch +The test measures the serialization and deserialization time for values from M to M +This is absolutely a bug in the JCA server integration +This issue is expected and the workaround is recommended way how to get rid off it +Looking to revive attention on this one +Ive filled some fields of the acroform using pdfbox and attached the file +When we migrate to Apache CMS this will be superseded. +The patch looks good as an initial start but I would have to dig into it more to really understand what is happeningIn do after creating the log dir if that fails we shut down the file system and send an event saying the app finished but we go on trying to aggregate the logs anyways +and writess for a node cluster is not really that much +This patch contains all previous changes plus the ability to store remote albums in local JCR +you want to keep new artifact Ids another idea is to propose a script to convert spring dependencies in +I think this is fair instead of relying solely on the TTL we can check against the current gcgrace and skip if necessary +Why dont you should open a new issue for that and attach a patch when youre ready +Ill need roughly hours but should have some results tomorrow +That seems wrong +Sure will add a +pete +Would love to get this committed in a next apache flex pulling my hair out on this one and trying the suggested patch without much luck on my custom item renderer I found another solution that works very well for me +I understand that +This is still an issue on +I added some more information to description of the bug and increased its majority +Patch integrated to trunk and +GWT support in Seam is deprecated Seam has much improved support for GWT via Errais CDI integration module +I just realized this is actually a dup of +We updated plenty of jQuery since +Is is worth trying that? I mean does this push the limit of the dimension posed by the type size +Even if it is fixed in is not something people like us will move to immediately after its out so we are stuck with for a good long while and basic functionality like this really needs to just work +I might look at making a few minor mods like hyperlinking SO numbers so you can go straight to the order to get the customer detailsCheers code I had worked on was a very initial attempt and on reflection I think I was going about things the wrong way +it is broke I cant use the in eclipse update site replacing the current rsync of latestbuild with sftp doesnt sound to me like a complexity its actually a simpliciation IMO +But do you agree that this patch is fixing a currnt bug in the code +I imagine a lot will change with the refactor of the scripts in broke the python think were really done this time. +One more patch for this test on Windows +Making this critical its an ugly regression not being able to browse to regionserver from master fix wont work if we start multiple region servers on the same m not able to allocate my time to work on this hope someone else can pick it up +Actually an initial version of such a maven plugin already exists in the Felix trunk in toolsmaven + + +This should always print zero but it still prints various values ve tested with binaries from HEAD cdaececebbdaaaandmaster HEAD thanks for clarifying +Now the Neural Network based MLP Linear Regression and Logistic Regression can all be trained in the same way +What is the problem at stakes that needs to be addressed through such a statement +Cheers +Can you please provide more infoWas this consistently reproducibleWhat was the setup likeDB dumpMS logs +It might just surprise the user that uses +Can you try with this may already be fixed as the callbacks for the stuff was adjusted quite a bit +I intend to commit today or tomorrow +Closing all tickets +Im pretty content with the latest patch too and I think it fits the current coding style best +Closing +This is a very simple fix that I believe helps to expose the skinnable feature of +I thought I had a really good solution but when attempting to implement I hit a wall +Levenstein is more compelling to me +Change UDF to take in its constructor an array that indicates ascending vs descending order so that the quantiles can be correclty computed +Committed to all branches +Please put back to the previous state after reopening for release notes +works finethanks +LinThanks for applying the patch +Added license header to the new files that didnt have it +If you believe this behavior would be useful then please provide a patch withyour suggested implementation +Closing all resolved tickets from or older +resubmitting the last patch without changes just to make hudson happy +Of course the user will need to implement a custom when his object model as a Collection that is not a standard +I only had the extra continue but it wasnt picked up +Too good to be buried out here as a comment on issue with such an unsexy subject +Simple agent to test if javaagent worksrunjava javaagent expected outputIts premainHello World +With the patch you suggest the no longer mention thepath of the file Also the input stream must be closed in a finally block Imnot sure the parser will close it on its ownWhat error did you get on loading a configuration from a jar initially +I will move the changes entry to bug fixes in the x branch. +Thanks for patch Kay Kay +However even when I first posted the patch to this JIRA issue I felt a little icky about the id field +I think the TODO is a bit trickier than what it writes +Theres some weirdness with negative numbers in the statistics and Im not sure that eviction and warming are working right but I am having trouble getting a test environment fully operational +I will find it out +Thanks for linking the request here +No luck +If were passed a stream someone elsecreated we should not close it since they may be wrappering out output inother data +I just added you as a contributor to JIRA +Applied patch with thanks to Balazs +TODO Update the CXF wiki Stephen +I think Chris means the use of sha in or thereabouts +resolved at revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Mass closing all issues that have been in resolved state for more than a month without further comments +The exception is related to an incorrect ref count of + +If you know any related information please let me know +This should potentially be in final +Marking closed. +UI modules become services +Fixed by the resolution for ASSEMBLY in OASIS updated in OASIS has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +Move to to to +Its an artifact of using the old bulk import code in FATEAlso its possible that a tablet server could copy something into the failed dir then a later import attempt by the master could succeed +Removing searching in the source will mean removing the implicit coercion from Cell and to T +That library allows aggregation of property files +We no longer do that silly walk code we used got rewritten in the process it got rid of all of those flags that determined whether we were walking or not because it was butt ugly +If you could test the next set of snapshots to double check that would be great +The NPE makes it seem like it is an error in the framework but technically it is not +You are right its easy to forget apply them +Geonames the place names in the dataset are not in random order but actually batched bycountry for example so you would have long streams of docdelta forcountryGermanys postings +This patch makes all tasks and works serializableCurrently there are no additional testsOnce the whole is serializable we will serialize the whole query plan after compilation and deserialize the whole plan before execution +Thanks for the bug report! +trying hudson +is not really a duplicate of this issue talks about relative path for include element in ivyconf file this issue talks about include of configurations in an ivy file +And i am pretty sure this code worked once in wicket but now i doesnt work anymore +Can we define these say in constant variables +Same as this issue +Because you did +Here is a question +What should I build my patch against? The latest svn or a particular release +Agreed +Thanks for committing the same +Is there any of this left to do CLI exclusions work great for me +Since it is my first time please let me know if Ive missed something +Ive committed this and merged to. +Please upload the complete WSDL +Already fixed in install issues related to Roller release. +I believe that the problem wont go away until one uses maven release a snapshot now and bared from problems I will release in the near future and bump the prerequisites to the use of the plugin to maven +File upload isnt required for core any more than dwr or is +Im going to have a closer look at what happens in the testcase areas should not have any adjustment as the text is not justifiedYour observation about trailing spaces is very interesting I thought their removal was ok but it seems there is something left to fixIf blocks have justify everything is ok if it is left right or center it seems that the text is correctly placed but the inline areas are somewhat larger than needed +Thank you thank you thank you +Mircea I thought NBST was implemented so that the data is available during the view change +LCM has been refactored +Looks like we still need to bump up the jBPM stuff. +Note that the unit tests have been revised so that each one creates a unique jar otherwise the results of previous tests can create false good +I agree the system account is the right way to go +Is it that big a deal to make this configurable though? Personally I will always choose a configurable value with a good default over a hardcoded constant simply because the kinds of issues Harsh is talking about can crop up in production and tweaking a conf option is far better than having to ship a custom buildHow about Yongjun revs his patch again based on your review feedback then we file another JIRA where we discuss changing the default regex for to be more accepting? This seems like a good compromise to me + +same as the last s the built before the fix made in +This change has also been made in the EAP version of the EM Ref GuideIt will appear in the next publication push +new admin for lmf core +Makes senseJust to make sure Are there any compatibility constraints related to upgrading from Jackrabbit and Jackrabbit JCR Server to the versions +Im not sure about as with potentially pointing to localhost such a mechanism may become unreliableGood point +But I feel get the right split instead should be the right way +This is doable but there is the issue about whether any of the components enclosed by the Grid need to find the Grid instance in a later action request +On top of that a usual implementation of equals returns false if the types are not equal enough +I havent done it initially to keep things simple for the first cut +you have to remove the repository section it has to build completely only with ve removed it and. + +By default Tapestry would put only one converter into this chain +Or using the signal name to appendprepend something +Im glad I could helpI think we could improve the gathering of file modification size details but we need to add specific operations to the and Sftpperations classes and surely to the that get this info by sending a specific command to the servers. +Here is a new clearer screen feature is committed Rev Saschanow there is a problem when you click on an example to edit it the decorator is no more present on the formto reproduceIn example component search all examples the click on an idCheers +The reason I suggested to create the struc the way I did was that it makes mutateRow less error prone +Should the also contain a asl statement +In addition the patch does have test coverage and doesnt influence any existing tests so unless there exist some fairly outlandish extensions by clients it would appear to be sound +HiI agree that we should eithera Throw an exception when the mapping wants to map ArrayMap to familycolumnorb Properly implement it so that reading does not result in trashWhat is your suggestion? You say you already have a workaround so that makes solution b not very difficult? +Committed recommended change +Revised an issue where the value of a parameter was not getting retrieved from the when it was null +That would be great! +Thanks Igor! Can you please put it on reviewboard I have a few comments +OK I will commit shortly! Another one downThanks for persisting on this Toru Koji Steven Mark Robert +this the general workflow +Sure will add those on commit +The execution itself is correct by throwing an OOME exception +new patch removed leftover assertion from +Patches are submitted. +That I got times out of repeated runs on my linux machine +Do you plan to provide a patch for that +Theres a lot of other things Id rather have in +Thanks for looking into this Fernando +Isnt this fixed in rev +fixed in solution! I removed a double occurrence of textjavascriptin r +Does the existing exhibit the problem +Committed to trunk +I made checksumLong be Deprecated since its no longer used by and with this patch it makes more sense to use to store multiple checksumsI created to track progress taking advantage of this feature for +Heres what you get nowCompile errors Template bar not defined in this stylesheet. +Is the config supposed to be loadfalseHow can I test if the car works +I actually mentioned this before but forgot to implement it +Should the class be renamed? or just Cache? OrOn second inspection is not shared between stores and regions +The rd party library is calling which just dumps the path of the URL and assumes it is a file +What went wrong with a that has an artificial limit on the data read? Even better would be one that was resettable to a new length for the next object +Hi Im ok your futher changes +That fix didnt do the trick +Updated in rev +Hiranya I have gone through the patch and seems completely OK to me +Please define the method at an appropriate place or try using a block instead +stupid problem that i introduced +Thanks for reporting it +As I dig into the Avro code for inspiration for this change Im finding some weird things +I could confirm the same from ivy and lib directory +builderseap r +The search now throws an immediately if something went wrong during the encoding instead of waiting up to the timeoutNote that the searchRequests filter is still considered as invalid +Iff we have some renewed interest in this we can revisit it. +Please provide some more information about environment used +Attached my local workaround for this the patch thanks Bram! +It works fine for me +agree +Ill take a look at the code soon and hopefully give you an estimate of when I could work on this +Hican you give it another try please +Timothy Chen is working on this already +And if we want to be distributing the code to Geronimo users we should be generating releases +in the params are no longer wrapped in defaults at all is that intentional? was that dead code before +We have a read only scheme ready for retired projects seems like we can just change the project to that +When this is complete we can do the component renamesSo after the upgrade we have the old netty based the new netty may take a while to do so I added a note on that its scheduled for upgrade so ppl use it with that in mind +I think the best way would be to just borrow as much as possible from a known system like Wordpress or Joomla or whatever +This is code and includes tests +See previous closing stale resolved issues +Ill keep testing this as I go along + +The commandline interface for fsck on the client now uses http to do fsck +Using my code above CF alerts undefined and then +ill also fix the querity revision . +The patch for ide based on the patch +This also failed using the libs to access IJIt did NOT fail using the libs to access IJPerhaps this is an incompatibility with a DB +Karsten Sperling was right +But Im less optimistic about how easy it would be to implement the approach I proposed. +I second Gregs usecase as I have a which contains no entities. +probably no help but heres another seed that fails with current trunkant test Dtestcase DtestmethodtestDMQ fbcafccdaace mike selckin for reporting these seeds. +Attaching the updated patch with all comment fixes +Can we close this then +defer all issues to +Other documentation is being pushed into the EAPEWP release bucket +committed to I added the label but forgot that jira doesnt leave a comment for that so here it is a belated is fixed +I cloned this because I know that this problem still exists in but Im not sure about trunk +Fixed in r see commit comments for more info +If that works it should be possible to add further languages this way +Currently the executors check is only run for main classes not tests +Resolved too soon +We could probably tag the compiled output with some version information so we know if its compatible or not +Autoexport sites have been removed and redirects to Confluence are in place +jBPM cascades all operations but deletes processLogs one at a time +Committed the regular FTP partTODO The SFTP consumer +Improve templates to for feature osgi based test cases +Wondering if this is because of? I applied but didnt apply easily to the branch so I didnt use it +Ive added the proposed changes with version +Theres an existing JIRA about builder patterns for client +done +small and medium tests pass locally +Thanks for quick response +SamindaBoth patches appliedThanks for the contributionRegards LahiruDoes this mean I dont have to the patches +Thanks for the comments DougCouple of replies kinda imply that other classes are in a sort of relation with this one which I believe isnt true +This is getting fixed tonight as well +Thanks +If you have validation mode CALLBACK but either the BV API or Impl is missing then youll get a error in the log and eventually a fatal exception once is resolved +Attached patch +Good suggestion +Initial patch to facilitate discussions +The patch itself is good but it does not fix the stated issue +roald wants to get a patch on this with tests in the next week or so just fyi +in the descriptionremoved srcadminguide and srcadminguide becaues they were no longer used +Sorry that findbugs was my issue due to an accidental commit I reverted it +Master is just asking the regionservers to close regionsSince were going to permit the deletion of multiple regions the above operation may take some time +First thing I will look at is the compress call +Looks good +Or FIX Hadoop cachingI wanted to fix the cache for the NM but after digging around I think the cache is working as designed +Michel is working on a patch so that all modules in srccontrib will be compiled and tested by the test target +We are on the same page about Entity +Minor comments +Now I can use the same schema across multiple services and not run out of should be so I dont think there are any threading issues +Just a thought +I commit this now and I can proceed with. +I dont see any docs links are all broken. +I have removed the XERCESSIZEOF macros from the headers which should help on all sensible platforms. +hopefully +The plan is to follow the release process as in the description +Leaving this issue open so we know the ones to test +Sidelined unless theres an update +That should be a vanilla impl of the spec from JSP presumably actively maintained +There is no hard limit on file size for + +shouldnt we opt to publish to the m repo and let the converter push it across instead of putting in this check? We talked about this at one time on IRC +A similar mechanism coudl be introduced to +But it is very hard to draw the inference the other way when you may have a run of bad jobs that are all expected to failCurrent blacklisting strategy looks at trackers blacklisted by Successful jobs +Otherwise Ill move this out of +All the commercial search engines based on nutch will definitely benefit from such an advanced feature +This was committed to branch on Mar just before was branched off it +Added new metrics and for them only the periods are displayed in the period select box +Also a bunch of whitespace changes +implemented please for temporary annotationsPlease add steps to verify and resolve the issue +Heres the patch for this issue including several test I have changed as little as possible to make the patch as clear as applying the patch would then allow for substantial refactoring +All lists have been created and are now active +again +I suggest to close this ticket +Works fine with provided sources for me +Fixed the issue for and removed another antcall in the processI am waiting for to be committed in order to rebase the patch as they will most likely step on each others toes +Once addressed we should revisit that issue and see if there are any remaining problems +Initially I just tried to change to and reduce the lock contention +Roshan Does fix this issue? I dont think a minor version upgrade is an issue +Thanks James +To simplify matters can you please respond on reviewboardThanks +Thanks very much Dag and Kathey +code changes are in branch +The issue there is that blacklists are hard to maintain +I didnt know how to do this. +Targeting to investigate that beforeA question here you are passing Artist not Painting to the expression +Rejected see above for other issue is a duplicate of +Thanks Jack! +The patch for the issueVerified on and new patch latest patch and regtest at revision The regtest hangs on Winx due to endless loop inlookupEntry will file a new JIRA +patch looks fine to me +This is Eclipse project with ant scriptTo build using Ant APACHEDSHOME variable needs to be set inTo build using Eclipse following jars need to be copied to +Dave thanks for this Ill take a look at it nowRe the current version I dont have good answers + so confusing +This was also reported on the list +Committed to trunk +This issue is fixed and released as part of release +Changes If command is and clean parameter is false or not specified if no documents were created and none were identified to be deleted then commit is not called +Closing. +Plugins has Red Hat Inc +Will commit momentarily +LDIF dn OUUser accountsDCdomainDCcomDirectory Active Directory with Domain compatibility level Windows Server R +This probably a similar problem to on +Its using a Syslog source so you may want to change the source +I guess I missed it. +Thanks for your review +The talk of settings up a database makes it sound like no unit tests are provided +I have amended the description provide the iOS crash log +I posted the patch over in +can you please upload a cvs diff u version +The fix has been applied to the project branch and will be included in the next merge to the SOA platform +My question becomes shall we introduce That makes sensebq +We can consider it for concentrating in to demonstrate the whole idea and moving in all performance improvement like this one +Closing the issue. +We keep watching it and will resolve it once we can test against a fixed instance +Thanks BryanCommitted revision +This doesnt look kosher +See the following patches and +Yes it is a separate issue to the one mentioned. +war that includes queue definitions so this issue can be seen without the ESB plugin +Please provide a method in the Environment class to access the Hibernate version at runtime +If you have then same problem using the APR autotools build that is you download APR or the SVN HEAD and follow their build instructions then the problem is an APR bug and you should reopen Bugzilla issue +It was just bad luck a problem elsewhere made it look as if my patches were the cause +These modifications look good to me +The Camel model does not preclude one from having consumers and producers sharing a session +Some adaptions to the CSS and integrated with main CSS Moved the menu setup to the template Created default category Main All plugins added to the Main category if no defined otherwise Web Console plugins categorizedThis now basically works but there are open issues The default category is Main should that be configurable or a different name? Categories are not translated yet do we want that? Using the bundle? The menu entries are sorted by label instead of by Title The localized label map is not a flat map but a tree of categories and menu entries The last issue basically constitutes a backwards incompatibility +bar and bar hdfssomeotherhostX the path will resolve to to hdfssomeotherhostX even if we normalize foo +Are you running a local instance at the same timeMaybe I was +box verified changes are in FP +Can we release extensions like JMS in a separate jar file? Or somehow make it modular so as not to preclude CDI from being used in EE environments +I removed the bad hunks by hand including those related to html tags +There was a hidden bug before the previous shcnage +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +It was an automated search replace that is easily verified visually and the previous changes of this nature have not caused any problems since they were made +It looks like its not quite getting everything that should be at the start of the constant pool right +Stat redirected stat output over my diff file attaching the correct diff here +Not yet +t find any samples where that feature is using +Looks like there was a timing issue which was thankfully solved by +Given the longevity of this issue Im going to close it and open a separate issue to be tracked independentlyThanks Rakesh! +its now fixed with transport revision +It is working as expected now. +I added automatic Range and support file response entities +Andy thanks for that and no problem we can pick this up in +Will try it again +Guy Ive just committed fix for this and deployed a new SNAPSHOT with the fix +Your was one of the points that went over dev and got lost in this ticketMy rebuttal was that while I agree that a config would be nice we dont have an infrastructure for that currently and that adding one is out of the scope of this ticket +Id like to go further to a design where the SPI client wouldnt necessarily need to parse the names and paths before passing them down to the SPI implementation but this is already a good step forward and its probably best not to overlay such design changes on top of the proposed restructuring +The ReadWrite classes was a similar idea of how the Hadoop code base handles this very issue +released +Update patch for trunk using class from Geotools to retrieve the fonts +Reopening the patch for it to be committed to and +I guess the use of the word client in my comment at the head of this JIRA may have been misinterpreted this was referring to the client of the interface +currently only unicast is being used by said server +Verified in GEP release process this problem is fixed +Yes +Now suppose that I come and replace C with a fresh server same id but empty state and I do it before A and B are able to elect a new leader and recover +Shouldnt it be closed inbugzilla? +Thanks! +updated patch followingbesides applying the patch remember to bump the lucene version in the parent toStill doesnt pass the tests If we run against a clean repo rm rf target mvn test Dtestit workson the next run mvn test Dtestthe test fails +Remind me that the file is called so I can find it +IMO if a user has a role in that is prohibited in the blacklist then it doesnt shouldnt matter whats in the whitelist +Thanks a lot Vinay. +Client to AM communication still needs to be looked at though +The behaviour is as expected but I Think if ARCHIVABASEconf exists then that will be written to instead. +Jan can you verify it on Windows installer and close the JIRA please +Well provide a minimal fix with that will give the user the details about the failure and allow him to find out how to solve the issue +Santuario is the Apache project name but I have no plans to change the C library name +Continuing with converting the code +Where is my mistake? I fixed this issue in the late April the jbds is created in +Adds the version when exception is caughtfor now perhaps it is enough to just let too new clients retry with lower version if they get an on lgtm +Its better to query than to navigate through lots of Java objects +Will this still allow out of transaction lazyloading? I remember you mentioned adding support for this using the proxyapproach in the email thread that started this fix +Confirmed m seeing this problem with cxf on tomcat with mapped to and cxf servlet mapped to services +are you still facing this issue +And that is the reason the state is not saved in execute +set to can be related to our doap file +did a quick grep looking for set in HTTPFS to make sure I was not missing from doing it there and didnt see realize it was on the client side +Its in the trunk +Nothing forbids us from supporting them for WFS as well +Thanks for the feedback +Is the attached dictionary your original work? Otherwise me might need to look into its license as well +would only appear in one Grouper instance +Agreed that get and get need improvement +We have a large variety of usecases that have different table characteristics +Sigh +txt is a test case showing the problempatch +Attach a batch ledger id implementation +It is part of a phrase being passes into the function +Also note that the JPA repository already offers save for have asked Oliver Gierke to comment on his view regarding the clear method +Changes to make Dofflinetrue for the patch Steve Ill check it out now +Arquillian supports assertions for deployment errors +has a few todos in the code which need further attention +Attaching patch bDecided to make the two new methods staticCommitted to trunk with revision +I will update the patch accordingly in +Point is why you are trying on purpose put obstacles in using other serialization frameworks +Yes Felix still has to be compatible with R bundles. +If nobody cares then Id say thats the end of the discussion +For the moment sticking with hibernate and annotation. +regardsThanks a lot +Both run from within Eclipse though so maybe the Eclipse compiler might be the culprit here +Fixed in commit abb thanks +We should soon make the next release if we want to pull this patch in we could mark it as experimental until we solved all the remaining issues +rename to and use and +The remaining work for this issue is getting tracked in following jira entries I will link the these issuess so we can keep track of the correlation between them. +thanks pradeep +How about and you could also use an rather than a single type parameter which would eliminate the need for the all type +Rebased and merged into the master branch. +is my exact situation too. +I suspect that cmdline has changed compared to when jsvc was still usedThe output from procpidcmdline looks like thisusrsharecassandralibvarruncassandraetccassandrausrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandralibusrsharecassandrausrsharecassandrausrsharecassandrausrsharecassandravarlibcassandravarlibcassandra that would be +Thats how it marks do you say thisLine of therefore returns falseI the lock manager saving and restoring state +I have retested this with and it works +It should be discuss changes needed for the other use cases once this has been done +Trying again +I checked it as I can see the issue is fixed now can you please check and using SVN seems to have gone away +False alarm. +It should be now impossible to create an instance with or +Yes that looks very nice +Fixed in trunk. +Thanks for the contribution Dvir +Intelligent use of these by the copy field mechanism MAY speed up things by significantly reducing the amount of redundant analysis work +But the patch does not what youdescribe +the fixcheck if they are duplicate and create only a single destination field +HiNot sure its still a problem with latest cxf any chance you can append a testcase which we can reproduce this issue? are generated the server and client classes using what was the Fix then when issue has raised earlierThis is a duplicate of which was resolved in +Easy to test for and easy to fixQuestion being do we want to fix this +I have successfully tested it locally in a runtime workspace but I also need to test this in a standalone workspace +Looks right to me it reflects the current implementation I think +This wasnt as straightforward to solve as expected due to the decoupling of the actual matching algorithm from the dependency resolution orchestration +I dont see any as it has both good collision resistance and distribution +I can see the motivationfrustration for this but it doesnt work very well with the Hadoop configuration model which is very open +Thank you for your ideasFixed by commit resolving fixed. +Alexey do you know about this +Apparantly tomcat was still using somehowThanks! +Sorry but I posted twice +Yes we should create one and Ill look to you for help on that +Asokan +I dont like putting in ledger storage itself since for other ledger storage implementation we had to same thing +That should remove all the firstScan logic +Leaving the issue open until we have a correct solution +Thanks +Imageedit is just one of the components +Changing affects from to so version can be removed +Then will never be called +Ms log and SSVM log +Just give a simple test case which can reproduced with a high probability +If you have an index with many segments even small ones I think only the larger segments should be separate threads all others should be handled sequentially +I understand now and agree thats betterI created with a patch reflecting my latest interpretation of the discussion +We havent had a huge amount of noise about this issue +Patch applied +Really want to see a services based solution that lets me add GDK methods in without EMC and potentially written in Java +As openvswitch is used need to make sure openvswitch itself works on QAs environment +You may launch instances of images that were not reported by deltacloud in its image listing +Yes in that case all is well +It sounds like at least a very similar problem to +I am gonna focus on trying to contribute with the open tickets instead +The resultdocmodifier is no longer null but the result is never modified +This was actually fixed by Adam see +Possibly related to all resolvedfixed issues of already released versions of Roller. +normalize it extract whats different from its original apply that to the latest version etcAlso picking an existing standard format is a good approach this may make it easier for translators as they may have tools that process the file +Ubuntu web account createdI dont have access to so I couldnt create an account for you there +Im planning to create a filter operator with unpushed predicates as a child of the operator through which the predicates could not be pushedFinally remove the original filter operators which are marked for deletionThoughts? Any suggestions +and Checkpointer are not affected +In particular the tops of paragraphs +It would be great to see this project implement SFTP support +Thanks Owen! +Took me a while +If turned on the base classes are generated along with the beans that extend them If such an option was provided then one only need to run middlegen then hbmjava and thats it +issue for that purpose thanks! +These issues should be able to be addressed with a comet timeline rewrite +Mass closing all resolved issues that have been closed for more than one month without further comments +Pavol mentioned in a comment in that appears to be fixed +Ive made the change over at +Given its not at Apache Im tempted to not care and leave it as it is +Heres a patch for this change +That last sentence is incorrect the connection options cant be set at all and are currently hardcoded +to trunk +This fact and other comments in the linked bug report make me optimistic that this is all that is needed but again still trying to figure out exactly what should get tested +Need to also add information about the capabilities override from the connector binding properties file +Verified in VPE works well +I tested this patch on a secure HA cluster where each NN is configured itself to bind to but is configured with an actual address for the other node +I like this plan +Looks good to me. +This patch was not committed to +I have to admit that I also restarted Tomcat and closed a second session of Guvnor that I overlooked on a minimized window +Please download and try again +Actually Jakobs ugly fix patch was not granted license for inclusion +date and time tag were basic checkbox file form img password reset row +added and tested new client didnt commit all for fixVersion +the collected works copyrightThe overview to the javadoc does contain a licence statement +That comment mentioned possible problems with boundary checks +We use a lot of simple components like input fields drop down boxes +imo only are able to understand rdf for all other users an rdf serialization is meaningless +I saw it at so older than your fix +Oops a couple more things +To test you need to update both JBM and remoting +Regressions passed with the patch +Moron alert Idea was not the project before launch +Could you also provide the source to Foo +Anil Patel wrote Scott I like your Idea Do you think you will do it Or should I send a updated Patch +Documentation change for patch provided +Missing are also another conversation with Jakob Ive bought his point that while the second test case is Ok example of a true unit test it isnt a good example of how to use Mockito frameworkIve removed it completely +In such as scenario pig would like a jobs input to be deleted when it succeeds +I have not run any tests yet other than verifying that the repro runs correctly on Java but Im posting this patch to see if the Store experts think this looks right +The code will try to group any remaining blocks and fail to produce a meaningful split +Looks pretty good +make a few changes for better backwards are no tests because its designed to fix tests in hdfs +This is obviously not correct +Patch applied +Readability helps in general wherever it is +Patch included +Lookgs very good to me +When you look at the diagram that shows proposed wizard page there are buttons to add or edit account +So closing the was quick! Thanks a lot +Not sure what to do with this the update site appears to be populated +The new patch nails this issue +Applied Jeff Levitts additions in patch committed revision +That might make sense +Could someone take a look at the issue and review the patch? Thank you very much +If you are going in a non solrcloud compatible direction with all of these massive core issues I think that is a mistake +Yes +definitly that previous code looks very broken when it comes to selections +We can but to make them useful in an runtime they need to have their context definitions in spring and some sensible server registry declarations in another directory +Please file a TLP request we cant create lists based on a TLP +patch mentioned in improvement appendum to this proposal +Same problem here is there any workaround available +Thanks Aleks +So it is problem +In case you were concernedCheersFor the adjust the CFLAGS and CXXFLAGS in your accvsizeofintp test block so that the flags are appended to the string instead of prefixing the stringIn this way you will then be able to let the package builder override the architecture values like so +No problem I will take care of those. +We are trying to deliver a replication framework where the basic replication itself is not happening as the triggers are not getting synced +Sorry for this but I have to update the previous patch it doesnt keeps the last maximized state in the user preferences Im attaching another patch +One is to do change directly in the other is to change all the affected subclasses +Other than the write access everything else looks good to us +Im sorry +Ill try it outThank you got a successful WSSJ build and my repo only has opensaml artifacts +as per email discussion the example app where this problem still occurs is linked from the previous found a workaround which was to move the code out of the GSP and into the controller +I dont think anyone is working on an approach towards batch exporting in a web environment yet +You seem to be size tiering across all the sstables and not the level ones +and then reject it +Yes this is exactly what this patch does including warningsWhere do we want the configuration converter method which class it should belong toI plan to use the same static converter in HDFS and MR and call it whenever appropriate +Hi Ankur I cant reproduce the bug with the latest code in trunk +Linked to the issue for documenting the process of producing Maven artifacts for Derby +Yonik I took a look at the interface and I think it serves a different purpose +I think this task should be reassigned back to Nick to add the corrected message into the components implementation +Stephan your patch is now commited to develop branch + +Note that we have a team of students from Harvey Mudd planning to work on this one +Of course not! The official patch will calculate the view id by using additional information like the mapping the DEFAULTSUFFIX etc +to trunk +Committed revision +The old import for far more sensible and only imported the users relevant to the project +Better add a new comment. +Big support for Adams comment too bq +Better late than never starting to review patch figure out why there are test failures with it +I wasnt sure about the best way to write a property editor for collections so I just returned an empty youd still like a test case let me know +I will send a separate mail to for that +Please let me know what youthinkChangelog no longer contain a instance +Fixed for in revision +This bumps the version numbers before I generate the beta distribution +Bulk close resolved issues +Thanks Andy and Sam this looks like a for being feasible and worth implementing +Verified by Sean at repo revision rThanks Tim. +if no force option just print a warning message +ve tried to install CR but it didnt worked for me +At somepoint this file is going to be just TOO configurable +More information needed for release notesClarify when no bean is present +Committed to and to. +cool +Revision added by Jerek Gawor from functionality from the commit in revision +I dont understand what you want changed +respect memtable thresholds when replaying commit log use normal write path but add a write flag so we dont out another CL entry for the clarify under low load you wont much exceed a single memtables worth of inserts since when a memtable flushes it marks the commitlog header as start replay from here +Is it possible to post the files for the run when it fails? And do you have the svn number for the first time this failure was seenInteresting observation about the disk cachethanks for bringing that up +Could we get the ICLA formatted more professionally i +Next step in the monologueTested it on a Windows machine and ran in to the same problems +Its a simple pojo +Hi Adrianthanks a lot Ive applied your issues before next release. +Oops +created new parent page for both existing pages. +So now the bug is fixed but after reading the comment I dont know if it is fixed or not was a translator added? there are no commits on the issueAnyway the value of the Select is not an index but a String which the application may interpret as it wantsTo make the interpretation much easier one could use a translator quite straightfarword then +TedWhat do you feel Ted? Chunhuis patch also makes sense but any way the idea is not to skip any file +Are you +As I said in my previous comment when these occur just search up through the log for and that will point to the initial cause +Resolving as Not A Problem as the definitions will always be available to things like commit hooks that work below the read access controls. +Sorry for the delays graduation processing ate my Apache time +This issue is because we cannot reinit a Service will file a separate ticket +Workaround explicitly set the location of the to be in the basedir instead of in target. +This issue has been resolved for over a year with no further movement +We hope this assists work around the driver distribution problem +Please consider this modified WSDL file +Please note it doesnt know anything about default ports for a given scheme +Its a huge unknown when you install BETA and your previous schema didnt have a version +Is there a patch for this yet +Ill resubmit +only key additions +Some subqueries which areindependent to each other can be executed at the same time if the clusterhas available resources +maybe a patch would reignite comment but I am not sure if there is support for this change +Attached is a patch that converts enums to strings +pete +I built on top of +Select the Feature Selection tool +and Im using Spring my case my implements directly +As Sangeetha mentioned all the reported issues in the bug still not fixedReopening the bugThankscan you try with new instructionsthanks priority to blocker +No intention of fixing the transient failures in Branch. +If handler throw exception when calling invoke method then will catch and create out of that +Attach one patch for the trunk which also has fixed the findbugs warning introduced by the previous trunk patch +Do we still need it or did we add this somewhere else +Adding JSF Portlet to a Dynamic Web Project +Im unsure about this +What makes you suspicious? On a side note I notice that in the Types class youre getting an instance of the Logger on +BergmanRoy T +Patch containing the part of that belong here +We were discussing something like this on the suite ml yesterday both me and Sophia think uploading shapefile via a web browser might not be the best idea unless we actually provide a java applet that can do FTP or some other serious protocol for volume data transfers with ability to restart an upload and no limits on the size to be am going to close this as there is now a community importer module. +Luc do you want to work on a patch? We should possible validate that the timeout option is a positive number and not allow or negative values +Insure proper use of. +Because you werent available for consultation P +MaxWhat do you suggest +Wouldnt it make more sense to remove the itests alltogether they use SMX as the container for integration testingThe proposed patch would inject a new for the example without any reference to SMX integration tests +the exclusion filter is not handling static consructors it seems +Had to revert back to for now since we had to distribute our app +revision Fixed Maven based issues on and modules +FYI I have a patch for that waiting on the contributors agreement +Will be fixed in build when the guys update the jar +The order does matter +Ive filed a support ticket with AWS and in the meantime will push a jclouds workaround +I will ensure that all tests pass before this goes inI was trying to find what do others have to say on thisThe warning is going to go in +Thus no change is needed +patch addresses Teds comments +For profiling +Im on commit but not critical. +Heres a new version of the patch that adds support for Snappy to Avros API and also updates the spec adding Snappy as an optional codec for now +Thanks for review. +Thanks for the clarification Andrew +Fixed in revision +dont use inverse on complex net relations. +Does this code only support RSA keys? Should we generalize to support DSA keys too? It supports only RSA keys but this limitation was already in the code base +here it is +If we do any test then maybe something that sets up a containing a groovy file and a GCL as child without any additional classpath and let the GCL load that file +Will be committing it to the trunk committed +Moving this from the alpha release to the +It would be better to say relation used as scalar instead of relation used in scalar context in above error messages +Trivial patch seems fine to me +doesnt communicate with the IW directly +Also add some documentation as part of this issue +Simple patch to disable delete functionality by default and enable it through a URL parameter +Can you check this one steve +Thanks Christopher +I need to do some more digging to determine exactly what is going on. +So are we ready to agree on wanted behavior and start codingIll try to formulate a suggestionPresence of explicit operators in the query should not totally disregard mmFor the pure OR case mm will be and work correctly as todayFor the pure AND case mm will be set to SHOULD clauses and work as expectedFor the mm +touches lines in the actual library code and adds a lot to the testpy test suite to expand latest patchs tests all pass +Thanks for correcting me +Dont know if that affects the phenomenon +In you could only see war modules individually not rar or ejb modules as they were all put into a single ear level pluginMy current thinking is that we should deploy ears as ebas with one bundle per ee module and possibly one bundle for the ear level +Nothing should happen if the flow has never been built to begin with +Again these could be options +I applied the patch after doing a partial fix yesterday +For users it could be I am getting heap errors what should I do etcThe list of questions can be discussed and then compiled +Yes but if the jcr api is installed a refresh packages should take place which restarts the auth +Defer to defer open issues to defer to not essential to address for +Hi RicardoCould you provide details about the paths youve installed Apollo to and the command that your trying to execute that fails +Is there a workaround I can implement in current that will not include the qt parameter? I suppose if theres no way in I can use directly and construct the URL myself +works with please reopen if you can reproduce +the first request should have the HTTP Headers configured to only ask for the kind of content on that answer we embed a browser which makes the second request +Yes this is much more work to do +I assume that you mean the logging should NOT cause the operation to fail +Canceling patch as a result +you mean to exclude +The convention is to check in only +batch transition resolvedfixed bugs to closedfixed +Seems reasonable to me we can do this on trunk +heres the correct debug log that shows a State transition Factory Deactivatingfor the component not sure why this patch Please do not apply this patch +Im not convinced this is a bug +The fact that sometimes returns false and returns true seems to be related please check the referenced forum topic +dataGrid page +Problem is that with the new tests added those fail +Should I apply it for branch +get it in for +This is incorrect +Attached diff withCould you please review the changes +I already thought of a Unit test for this but my code is so specific so I tried to avoid this in the first place +fixes deadlock in shutdown should be fixes deadlock in shutdown sequence +To fix this the main problem is that the selector is not aware of the context it is one context it would bePortlet Instance used by this window in the otherPortlet Instance to use for this window +patch looks goodWe should move this issue to Hadoop Common where the belongs to +If anything goes weird let me know as its just off my current development path but I dont think Ive done anything crazy. +Ill try to do I mention almost no unit tests i do not like name cause Tools has one Visual call every editor visual is not a right ll rename it +Resolved in ve experienced the same issues like Harry +Except that my test client which connects and subscribes but then sleeps didnt disconnect until after minutes +Its at INFO level not error +We could only do without maps by implementing an algorithm that creates the unique String key by Image object as decorator has no access to original key by which the image was created I think that map is a more simple solutionI agree that we should register all created images in so that they could be disposed by Eclipse when plugin is stopped +You can change the server name on commit +Looking at upping the default meg in memory part of the sort should be looked at this is incredibly smallfor a large number of the system we are running on +This bug surfaced for me when having built and using the regular jruby command to start scriptplugin inside a Rails think we should add a patched to the build in the meantime since it gets in the way of testing trunk +Hi Would you please try my patchBest Regards for the patch Jimmy its applied in revision Geir please check that it resolves the its OK +Its a blocker to generational mode GC collection +They have been applied and the new build succesfully passed all thes tests +Added a test in r +The test issue has been fixed +Do i need to commit in any other version also? +Patch attached +pig x localSystemLibraryFrameworksVersionsHomebinjava Xmxm UsersjoaDocumentsworkspacescompanyworkspacebin +Fair enough +A new issue will be opened to handle that task +Flavio +This patch has a couple of improvements to the build infrastructure + +Potentially this cost can be reduced it by caching directories created +I found that in Tomcat package is binary of procrunI could register as Windows service using next command +Close inactive window through context menuExpected inactive window closeActual active window close +slip +Be aware that the setting still seems not to be recognized! Only the parameters from SUN RI +DB connections in use typical figures when this issue occurs +no longer has the bogus combiner but still holds all clustered points so it really wont scale well +Thanks for the patch Jason! committed in rev +Never been much of a spec reader myself +Shouldnt be a big change +I think it is better to divide this issue into +Defer all unresolved issues from to +fixed rThanks! +bq +Hi Thanks for the patch Ive applied at r please verify +php files are new files which are getting generated +To confirm same Jenkins host deploying to same Sonar host everytime +nightly build runs fine for application and datastore id for this +OK now I understand itThe problem is that there is a bug in so watches the timestamp of the EAR directory not of the deployment descriptorSo we should try to touch the directory as well as the descriptor when we want to force a restart +Thanks for the review Brian +Im marking this as incomplete +With the implicit router we dont have any information about where to send the document +Attached is a patch for this issue +Would this case be handled by the Size Tiered fallback +We couldnt release because of the tabs broken functionality +This problem has not recurred in over a week +No tests are included since this just amounts to a logging change +I thought nested transactions were not supported in Teiid procedure language are they? If you are saying two separate requests then I could not see a good way to spin up an additional request based on the request for original materialization view selection the timer yes that was my intention to make management of materialization independent of the processing logic both for external and internal matviews but right now with this JIRA I am only targeting the external +found this did a fix use it if you like with +Following Andreas last edit I THINK that the problem lies in the chain of projectionsEPSG EPSG EPSGSo perhaps we should start with EPSG to ESRIs software latitudelongitudeI suggest trying out the projection chain and see what that gives +g +Closing. +scrubbed releasenote +I dont see a as existing so I presume this fix will appear in or and not. +The implementation of opensocial Group Rest request for PHP +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Min have you opened a jira for this! This is really a major bug. +Here we might ignore the failedkilled tasks and never call discard +Patch committed in r. +Hi WeinanI think that would be best if you didnt mindThanks a problem at all Then Ill rename the prefix to for branch. +Thank you Kathey! Thats very useful backgroundDo you think it would make sense to reclassify the issue under the Tools component instead of Documentation? Its now a tool command line issue evidently +Any additional functionality will need to be tracked as separate feature requests. +However is only marked as fixed for so it is not clear that the fix is appropriate or applicable for the older versionsThe bugfix probably should have been applied to and merged forward to +You can open another bug if there is a problem to run the sample cases with JDK. +btw it is a minor change and doesnt affect the functionalitybq +for the patch +IvanI believe the problem is that the instance is being reused here between calls and things like and are being shared +Also just to second what Andy said for well still want to somehow provide SASL auth for thrift clients with the thrift server acting as a proxy on their behalf but that seems a much bigger project +With this fixed the exception shouldnt happen even under high load +Added to track the fact that no documentation was added for these functions +One fix has been checked in and seems to have fixed one of the issues but it also sounds like at least one of the problems remains unsolvedMy reading of the comments on this bug report is that the issue Stefan saw could be reproduced and it has been fixed +Ramnivas I guess youve done quite a bit of testing here already +Chaining the exceptions to actions will enable the checking of levels of details to show to the resulting page +please move back if they will be worked on asap for +Thanks for the review JasonMinor update fixing the test +What are the steps to getting the Zookeeper test passing locally on Linux? I already applied the latest patch on trunk locally +Personally Id actually rather deprecate and eventually remove them if no one is going to take ownershipresponsibility for them +Fixing this still wont make them work together. +Tests are still failing +Logged In YES useridThanks. +This can be leveraged to create a suitably minimal repository for the clientHere are a couple additional implemented ideasThe app client builder should be examined to see if the app client dependencies are specified independently of the ear dependencies and that they are not all munged togetherIt would be best if the finished app client could be distributed as a single jar at least as an option +OK on trunk C doesnt get trapped so this is still outstanding +Have you build the docs with the patch +Both tests are failed as before my patch +Sorry for confusion +I just committed this. +Its just what I happen to have installed on the machine +Could you please test andor review it? If it works as expected Im going to commit it to and trunkWe decided to go for the cached solution each thread is now maintaining its own list of service clients +thanks +Rev +We dont want a Nagios box impersonating a server! +It seems that it never imports any schema associated with the jee prefix nor does the schema define locally any group named jeedescriptionGroup +A reason for putting it in a separate context which I think it is valid is that we do have our main application sitting on top of the ROOT context and there does not seem to be any reason for the need to merge shindig with our main application for things to work correctly +Please provide update on your items as a comment +Thanks for pointing out the typo Sijie +Thanks +Hopefully thissimplifies things +This patch cleans them up tooThanksThanks MattChamikara +They just feel good to me +Please ignore +This is a flume issue and will be moved to that jira accordingly +Submitting the updated patch +The public Hudson instance is a mirror of the internal instance and that mirroring is not real time +New build is available for and +Any word on when this might hit the SNAPSHOT code at least? Is this patch viable +Once these are included portal should work out of the box with jboss +I have asked in +Here is in trunk +Im also curious if they are looking for any specific issues that have fixes +Please check it +I ran only the regionsserver tests +Like removing the variables which are no not really needed and closing statements etc +Fixed with +Yes could be tweaked to save data in DFS instead of the local filesystem however data is not accessible via the API API is not accessible from a client information does not contain counter informationThe nice thing about the proposed patch is that for the client it is completely transparent if the job info is in memory or already persisted it is always accessible via the API +Previous versions were listing hidden files always +However if you are also using Embedded for your testing you will need to set this value +Attaching three files an XML file that will properly validate against an XML file that will should validate against but does not a schema definition defining based enumerated types +that code snippet shows that the dest dir perms are setup but doesnt touch the contents of the files itself +I just committed this + thank you for looking at this +Sounds like we need them for Weblogic compatibility so lets keep em and close this bug. +Ill take a look at this now +This new bug in our max versions logic might explain it +I have tested V patch it fixed this issue completely its time to merge it +Is there a way i can get a version of the plugin with this patch applied without having to compile +The code changes have been added to the upstream codebase +Not flying now +I would prefer to do it like for NRQFCRF with static factory methods +That issue was not checking on OS Linux Ubuntu SP clients IE IE Nautilus +Thanks for the patch +This seems similar to +Marked as fixed and asked for Devin to had been reviewed and pushed +But almost similar versions we have in bom +Im able to see all rules +He needs to go to logs folder to find it out +Oh +Attachment has been added with description Proposed Security on Thomas patcha +HiI have attached an improved version of the SAML validation patch +Also observed when deployed unpacked ear +yonik +New rev I think this ones ready to commitIt handles indexDivisor and pages the index term bytes allocations +Reopening for backport to +The third screenshot is the IE About box to show the exact version of IE I was using +This seems like an error introduced with the notify back to Lukas +However Im not sure this will ever make the same sense in batch +OK I think this is fixed the way I thought Id fix it in r +Thats a JON features +Fixed in r +Changing the Fix Version as I dont want any possibility of introducing an incompatible API change after in order to deal with this +Now that work is underway with AS all previous community releases are +though theyd have preferred no limit what are you talking about! I want them allSame here +Roger Ive added you to the project +Thank you ArpitPlease continue additional work needed in. +Ill post details there not a big deal +Its emergent to resolve duplicated path name +Hi JBHow does the be referenced in your routeWillem +The issue is not that you cant build on OSX but that the configure flag is not working properly +Looks good +Log in is a verb phrase so the buttons should read Log in and Log in with To keep capitalization the same lets change Create New Account to Create new account +Thank you Piotr for the sample +Committed with revision +app file on Windows +Committed to trunk. +Its also an opportunity to clear our designThe root patch should be or +see if I could come up a better solutionSure would be interesting to see +Thanks for the review Philip and Daryn +jaxrs jar needed to be removed from the GWT console server +Farm deployment is not enabled in EWP +Simplify the logic which require all relevant fields to be pruned before pruning +Please use a higher port range the ports suggested in would be fine and agree with previous comments that the best solution would be to use a random port if to OK? I mean are they open by default +Seems rather usefull feature +Sounds good to me +So +I think this is a copypaste error from the case for and it should have set tokenFound to instead +Feel free to reopen the issue if needed. +Applied a patch to fail installation of any bundles using the namespace in their generic capabilities andor requirements +Also should we really issue a warning log when falling back to a full page render +pushed upstream to to add fixVersionThis was included in via +That simply means that there are currently rows for that table +Ive tested locally using your setup and it works +Im seeing the same of the boot features are getting installed at startup +IE may tasks will use no more than I dont think a per job limit on RAM to map reduce works +What I mean by half is that the links work but the ampersand is not stripped from the display +New proposed patch +It may help if we start thinking like an external app developer who doesnt know YARN internals enough to find loopholes like these +works for me with this snapshot Thanks! +I see the same failure in my environmentHeres the statement in question select as idcolumn as idcolumn from t order by Without my patch it binds to the alias idcolumnbut with my patch it rejects the query because there is no columnidcolumn in table tThe behavior of the patch seems correct according to the rules we discussedBut it does cause a query which was formerly accepted to be rejectedMy feeling is that the query above deserves to be rejected but I wonder if anyoneknows what the SQL Standard says about the construct +The reason I decided to keep them together is that they work together to create something like a training set and something like a testing set +After updating the upload section of the struts the error no longer occursBruno Falardeau +To complete payment flowthe only automatic approval is one where the customer has either an account with funds or an approved credit line that is not maxedall others ETF Wire Transfers need to be completed before the order is changed to approved statusso no packing can be done till the order is approvedThen you have the case of dropship where packing is never done +I could do that +Ive created directories in dist added a there renamed the old to be and updated the Ruby build instructions. +Im using our institutionnal CAS implementation which was custom written by the staff here +The old version of the example was relying on some classes that are no longer available in EAP +Fair enough +for the patch +Hi Kirk +I believe adding this would be pointless. +Im not sure we would want to embed it in the name nodeIm not sure we would either +Committed to branch too +Once he ran it as a user all was to assign this to Hyperic to fix the problem of the installer not bombing out when it detects that the root user was running the installer +So that should be the reason thenI will try nowThanks +Fixed +Bulk closing stale resolved issues +bq +If you use Executor it takes care of pooling threads for you and you create without worries. +Hi PhilipI tend to think of this as an assertIn this case youre controlling both ends of the pipe the Thrift server pool and the pool on the client side +Please ask these types of questions on the Struts user mailing list. +I will get it fixed for fuse +Patch is incomplete and not implemented properly +see attached for compilation failure +New version of same patch but now including a new test case to test the new class +I think this must have been fixed cant seem to find the instance that is being reported +Wonder if this is related to +Hopefullly I fixed all of them at least now both well known marks and ttf markers seem to display with the proper orientation both when used alone and when used as part of a graphic fill +It looks like the input is used only to control the number of sleeps each mapper should observe so we can write a that emits the expected number of records to each map +But the problem can be reproduced at will just by changing the JDK to JRE and vice it does not create any other problem when working with the Im on so souldnt it be fixedRestarted many times with same hint +well to be even more grammar conscious its regular expression not regular express +I cant detect a problem so I suspect that it relates to some strange stringPlease show us the actual value retrieved then what does to itOur tests select string string and others test on avariable +Now Perf test shows that performance is about times better than the vanilla Avro version +Yes Ill recreate it and see what I can report +I dont have an issue with it in any sense of the word +I once created a maven archetype for component libraries which Id be happy to contribute +Also see if you are looking at optimizing +There are several reasons you can have an OME +The implements but the is not getting set properly when the default configuration is used +Maybe the page would include a default of seconds for the solr timeout +It seems that Im also unable to drop column families inside this keyspace +Fixes and test cases committed to SVN trunk +Weird +The problem last time was to perf test this and find a meaningful default +The channel has been closed so the bookie handle is in fact not available +The message needs to be pulled directly from the store when the pageSize is full and all subscriptions are full alsoThis patch is a tidier version of the original +bq +This test shows just a bug in but not a whole feature absence in projects +I checked and verified the list API on latest master and +I took the and Ive tried it locally to see if it is going to work but it didnt so it might be a problem with the source that I build or it might be misconfiguration on my side +Changes made in the plugin documentation Javadocs index usage examples required tags for docck in the plugins FAQ pageThe plugin docs are now ready for review +I dont know enough about Windows batch programming to fix the bat file +duplicates should have searched jira first +I would advocate for having the script take the passwords and pass them to keytool instead of using the interactive mode of keytool +Just a first step so people can reviewI havent had my hands in the fetchmail code so I only can do a fast review +Sigh attaching the original store function isnt enoughThe problem is that currently Pig the entire pipeline and doesnt use the existing results on HDFS for the JOIN in the above example +I notice in the code that all other comparator classes are made serializable except the class +Looks usefulHowever there are a few minor issues with the patch +this patch was only to evaluate +What about simply never calling the Step at all if the job determines that the step shouldnt be restarted? The removes the need for the call and any logic for handling empty arguments etc +After I created this Jira issue I later noticed the ant task Jira componenet. +Must have been fixed at some point +Updated screenshots for bpm and rules io mappings +Set Fix version to +In fact Ive ensured that even the doxygen generated docs go to buildlibhdfsdocs and can be packaged aptlyKonstantin can you please confirm that it works for you? thanksIve kept the target around in so that devs working on libhdfs can use it without needing to invoke the clean which nukes the build directory +Any progress on this? Seems like a hard won agreement +but whats the time scale of avgtime on these metrics? Averages that are since boot are kind of useless +A transient outage on the backbone is very different from the client laptop being suspended taken home and brought up again outside the VPN. +Is there some reason you especially want to do exception handling in your model +Looks like a schema resolution one +Instead of adding the attributevalueName we can overide value +Im going to commit it +This page was copied and slightly modified for the orderview page to replace the ship group section on that page + on screenshot I try to copy gf into buffer but get unexpected content in there +Nevertheless I still receive the same error message +OK cross checked. +Attaching a patch file +Hello any news about this bug? Will it be in +I have cribbed that text and put it into the overview section of this template +Bikas I think is more clear as yarn client today is a library used by other applications to write their clients +BTW this is of course related to +behavior now the same as native seems now with this change the global CLASSPATH is not being honored as my shell scripts that append the global CLASSPATH no longer work when upgrading from to how to I tell groovy of the global classpath now +Instead it would be preferable that they deal with any checked exceptions and for those they cant make the conscious choice of upgrading it to an unchecked exception rather than have that happen in the background +Applied to +Im not sure I have the authority to close it. +Defer to all open issues to with +Grouping module is unfixable without knowledge on the generics +It should error out and the help should be updated +Also it works when is providedmay be it is reading listen address or something fromDescribed this problem in more detail in +Im marking this as invalid as it seems another ticket has superceded this one +Fixed in revision +Looks just committed this. +Patch for documentation attached +ThanksIm going to observe it some days and delete them then. +Improve rate calculation for accumulated values +ok cross checked +Just an idea +This is my environmentJava Mac OS Xboth sane and insane Derby jarsThis is the ant command I runant see this problem as well on linux with both IBM and IBM jvms +What about rollbacks? Are they going to be supported too +Thanks +It makes more sense to fix them separatelyFor Follower Observer and Learner I have removed the synchronization blocks around pendingRevalidations +OK if the is modified its much better to be consistent. +It also has an implementation of wrap though it currently fails to exchange handshake Cath patch applied with some very small modifications at repo revision changes were tabs to spaces on just a few lines +Thanks for your comments +yea unfortunately jmx will still CNFE if you pass e as the initcause but i will change the error message +This bug just makes sure that we track this issue with the deltacloud server +Revision +The patch resolves the problem correctly +The attached class encapsulates a instance allowing it to be used as a instance +MaxIll take a stab at writing a test case for the patch submitted under what would be a good test case for prior Export patches for me to emulatethx of the reasons we dont do so much fixing in schemaupdate is because there doesnt exist proper testsuite +In it is clearer to change the variable pktLen to be payloadLen +Yeah just tried it out in other browsers +The idea of contrib doesnt really make sense as a result so this code would be more like an example or a working tutorial or something +Thanks +to trunk +syncclose can be set to true by the user to force closing the socket right away +Im using the most recent release +I dont know if it is the correct solution +Jena IRI library performs a detailed parsing of the string +Regardless of Spring classes you need to know API and WSDL spec +Ive remove some duplicated code in and relating to the translaction into a static method in. +May also be removed completely +I have tested it with s sleep +However if you have it fixed thats great to hear +I was hoping you might submit a doc patch +Sterling work +James already added a workaround for that +I built three times and everything seems alright +I believe as people build more complex applications they will begin to become disenchanted with what they eventually will see as a limitation of the Struts tags in not allowing custom attributes +close of resolved issues. +The scenario works for me if no further comments I would close this jira as invalid +I consider Constants part of the public API so I think that there needs to be a deprecation cycle involved +Implemented in trunk svn rev +samplesReadme Concurrently for the patch Amita +fixed +Railo use nowhere in the code new Boolean or new Integertnx for the awesome! +You should set MAVENOPTS environment variable to provide additional parameters for JVM with maven +Test case and fix are Li Jing QinPatch applied at r please verify it was applied as you expected +Can you confirm that you are ok to close it +But again your call +covers my scenario so Im ok with changing to given gives this to be derby internal naming +So all issues not directly assigned to an AS release are being closed +cd libjava ant clean compile ant publishshould publish the artifactsTo change the version change as appropriate in in libjava +I dont think youll be able to convince teams using Lucene that they need to switch to makingseparate calls instead +Yes thats correct +The change I alluded to is predicated on that statement being true +I am using betasorry for my carelessness +Thanks Eli! + +So this issue is fixed when we upgrade rider to the latest release having that fix +most that uses egit doesnt want to have svn installed or viceversaIts a balance to make +unit thanks +Is my patch failing because it only fixes of reported bugs and not all of them +r +I used this template for my latest Groovy presentation last week +The same app for +Resolved +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +too many other things on our plate +Denis Andreas please this issue if there are any new findings +However I cant reproduce your issue +As suggested we are now using a debug build of Derby so that we can provide more information if it happens again which it hasnt yet +Please and close +the scripts should live with the project locally +Please check your application classpath whether you have any XML parser or parser API jars lying around there +Tactically I think renaming column to cell in is a bit out of place w the rest of the code but the fix LGTM. +Added to +AdrianActually yes I was not clear + +Rich can you verify that what we currently have in the DR releases is correct +I added a link to this one so that we can keep it around and revisit the COW map based solution when we have the time to look at where and how we can use it throughout the framework. +bulk close of all resolved issues in preparation for release. +Alan this patch has never been Patch AvailableIAC this question doesnt belong on jira +I think an FXP is really a ZIP +Help in resolution is appreciated. +Thank you for the comments and guidance Oystein and JorgenPls find a patch attached containing the preliminary implementation of thelog shipperThe patch keeps to the design posted earlier however the code thatinvokes the log shipper from the is not present sinceI thought that should be done as a seperate integration issue or atleastas a patch in this issue +A virtual host is used to point to Tomcat in the file and according to my Apache logs the permissions werent being correctly routed +Not sure but you might want to change some other stuff I spotted where there is not int constructor +Now it is positioned as a bugfix tests to check the fix +I deleted the whole storage through Preferences +updated cookies and has different path +So this error will only occur if you try issue the same update request again following a request withIll fix this tomorrow by using a different approach to executing the +Ive checked out the surefire code and line says no need to add classestest classes directory here they are in the classpath elements alreadyIm having major problems digging out exactly what is putting these directories on the classpath in the first place it doesnt seem to be the responsibility of the plugin but perhaps passed in from Plexus? Can anyone point me in the right directionThis issue is blocking a major release for me and Im looking for as much help as possible +no longer uses the JPA connector so marking as WONTFIX. +So there is some bug in netbeans with adding platforms on windows? I could via autodetect platforms so I am not sure how autodetect works but it must be different from add +This also cause un marshalling errors once the cache fills +As you can see from the test results that I submitted the differences between are significantThere is no perfect algorithm which runs better on all but I think that algorithms that run better on most exist +As far as I can tell it doesnt affect the UI and you dont see errors in the AS console +This seems related to +HiCan you provide more details on it? I tried to run test case and can see that it is explicitly done by but not able to find any such instance with. +variable +We havent needed to do that yet and it isnt obviously necessary for the basic flows in the spec +fixed with revision catch +Reused crawldb code insteadThanks for opening this issue Sebastian +The files now in CVS look OK to me + is merged +Likely since the user is disabling the inactivity monitor their app is closing connections abruptly and they are then lingering on the broker side +Wed create a handler in Faces with a low priority so its one of the last things to handle the exception. +Im planning to commit it to as soon as I create a few test casesIf you plan to add collaboration between processes for replayer please describe your solution and then Ill create a branch on github and give you access so you can work on it +committed to I do not see any aditional preferences on first page compared to beta except remote deployment settings +create a new config for the retry wait intervalTest with HDFS HDFS is down while RM is running or before RM starts NN is in safe mode when it initially starts +Patch is against x branch +What kind of metadata do you have in mind +It checks for exported docs that are no older than one day. +Applied patch to Apache HEAD +Closing nothing more to do here. +IndeedI thought we had something like that but overlooked it when i went looking so i opened this issue. +This is still present in Maven beta +It is highly problematic to have to touch the context class loader at all +We now have no test for and reopen if there is still a problem +This may be the same bug as which is fixed in Derby +The sed method has been removed and now the tests just checks for the strings that it needs to check in the output without failing if the number of lines change. +Shouldnt this issue be moved to +Should I file a new issue against or the extender addressing the specific problem of how to engage this new support for resources in fragments specifically for loading application context definitions +I wasnt a PMC member at that time and then there was the git migrationAnyway thanks for the reminder +Its actually in no way associated with this case though +Thanks Amar! +If they fail again please share the outputThank you very much +Ralph Im starting to test d now +imlemented now +Resolved in svn revision . +Its maybe a fixed Bug +A block is not a valid inline viewquery expression. +Depending on whether is being used to track real time ideal time or your definition of Velocity might be confusing +Due to merge conflicts with latest master Im unable to upload the new patch +Unit tests pass for example attached +I see we are catching Throwable here without logging it straightway +started four nodes +All functional tests for container provided Event objects succeed +Im including some modifications in this patch but Im afraid they wont replicate the behavior we have currently +The generator selects exactly numSegments topN records to be consumed by the following numSegments partitioners +Without an interceptor to let you know when the invocation is finished how will you know when to destroy the bean +In any case the code for the toBytes method is written now so it should be a simple thing to move it to the implementation classes +Implemented awhile ago. +We cant just send every message synchronously and check responses from the server +Willem we need a new MR release in start of January it should be named or depending on whether we manage to do a Apache Camel release first at version is to be used in a new ESB release due in end of January +That one must be seriously out of date +That is all I can ask for If things firm up as to which release it will be targeted for please let me know. +I know this is late +So long as the versions of the libraries declared in and the version of Ivy declared in have not changed since an online build was done offline builds will succeed. +We didnt apply the Hibernate patch from yet +slipping to to short time to verify for but marked as patch +probably on? In any case resolving as duplicate solution is to upgrade to real. +Youre not wrong +Also found Web Studio LicenseRed Hat License +which AS do our test run on? This should be a runtime issue +what is it? And why do you need to deselect it? It looks all correct +ah didnt noticeShall we just drop support for? I see no point in testing OGM in different relational profiles we can then add a similar functionality to have tests executed only in some stores like it would have been handy to have to disable relational tests on Mongo +Sorry for the formatting mixup I adjusted my Eclipse formatter to match the droids code conventions +Will attach that produces thisbehavior +Hi Rajit is this a duplicate of +Check the bookkeeper project for an example +will revisit again for tests are no longer failing +thanks +We should start by at least documenting this behavior until we have a patch +Are not the +Sure +Fixes the unit test failure +Ive opened tickets for the other tests which Ive seen failures in. +It introduces a dependency which is not publicly available via maven and will break our ability to push this to maven central having a random github repo containing the library is not a viable long term option and should not be s only a few classes maybe we can reach out to the developer and see if they would mind them being included in the Thrift trunk? +The renders and + +That makes sense +This advanced rule has to access the +Do you know which bundle defines this EMF extension package? Is it part of the +Added issue to unscheduled release +It is not possible currently to configure passwordtypenull in and Usernametokenaction classesThis is used in BSP sample applications and it is also a sample usecase forThis allow to propagate identity and is for instance proposed in Websphere AS +Ok +See connection issue Riley mentions appears to be differences in how and MRI resolve localhost with the former sometimes using instead of resulting in the servers not actually using the same protocol +The looks good +Good page needs a new title +Now the required javascript is appended to the current onclick event +removed +In SVN. +Other tests confirm the fix +any luck? I have some big logs around here if you want me to try +Chris Thanks for comments +BTW Li Pi any progress on this? I would love to help you in this +Can you give me any way to diagnose? +Well clearly option a is what it is +That was not a problem for the java dialect that uses stateless expressions but MVEL dialect caches declarations and when wiring them each logical branch was overriding the wired declarations of the previous logical branchFixed. +Sorry Rob you are right this issue has been resolved +has a dep +For example the word its moves to the first line +Ive got the test hooked up +Still confirmed in should add that the example of the given in the screenshot is only visible in a scale range between and +A small modification had to be made for make +im fine with the name +The new patch seems to be better but it should be clearified if its for or +Unsure but I dont think we would need to do this anymore as we have RSS and Atom feeds as well as external tools like Google Reader. +Its not a effective to write the hdfs is we close the stream every timeMaybe we can consider to add a option to make sure the producer to close the stream once the producer is finished the processing +Maybe I assumed wrong +Moving to CR since we have less than weeks till Beta +Lets fix it separatelyHow do renames cause double leases +Thanks Elena and Mikhail +I ran a workflow using and workflow finished +Should this be moved to +Thanks for the fast fix works fine for me +The onus is therefore on the Validator client to ensure that result classes are serializable +I moved the vdb over to a teidd only env and this CNF still occurs so Ive changed this to a Teiid issue from a Designer issue +Changeset cdacaabadeeeAuthor marcosjunqueiraDate Message updated. +Original repro +Screen shot showing progress +can you try this patch nicholas? It fixes the issue for me +Where is it called I dont find it being called when rows are advanced does not seem to suggest a clear place where reset should be called +This will be done over the next two or three days +how to write and configure a user interceptor +We might want to get around to this eventually but there is not any demand for this feature currently. +Currently it has only the and implementations +is then making some influence there +Should this issue be reclassified as a bug +These are already captured in JBIDE jiras closing this as duplicate +Further the identity stores are in property files that if not present it will introduce another diff between installer and zips +applied! thanks. +Can you remind me +since this is production server +As soon we have the result I will update the ticket again +Arnaud can you try to duplicate this issue? If you are unable to duplicate it you can close it +This doesnt have unit test +If you need to get that stuff done right now you should be able to use on of the callback interfaces and the native Mongo API shouldnt you? +Would it make sense to correct this in +This has been addressed as part of the rewrite of. +Please go ahead with new improvement +We need this functionality to be always tested +any chance you can apply this +Pam Have you had a chance to work on this +I see +On that note where can I get pig that is not bundled with hadoop jars classes +If you just fixed thats fine Ill check there but in general attaching a bugfix rider on a big change like this is not good Derby politics +bulk defer of unresolved bugs to +updateLock is in the and if is it will block the whole regionserverOther the way set the flush handler as so it is the same as the beforeAnother way s lock time quite rapidOn the whole IMO taking a easy testing is more than analysing a lot +Is there a good discussion about it anywhere? +Attached another patch which changes the style of pagination similar to hadoop pages with First Previous Next and Last +The internet ways are impenetrable +The reason for this is incorrect configuration of the transformed AST +I added a patch that shows how SLFJ logs will work by embedding the groovy script in the pom +Agreed +Be my guest +indeed nethttps doesnt get required in some cases +Still reproducible +Verified +If we say that contrib in general does not need to maintain and were talking about classes that are in production environments then I dont think we have a real issue hereContrib does not necessarily have a back compat but its up to each contrib to determine what its back compat policy is +Thanks! +So well simply add on now +The OVF technique is the only way to skip validation rather than stop processing +If thats the case we can rename it to Exceptional mappings +fixed now +This development is rather a POC using UDTF hive function but it worksI will focus on removing the need to the DUAL table and improve the implementation +Best to read spec to validate and then check with code in refimpl for methods with check or init to find look for Broken to find examples of source files and can be used as a Test case for the validator showing errors +Actually it is tests that got enabled +Could the fixer of this JIRA either create a new or incorporate into an existing test a test that would create a for this function so that we do not regress +The patch broke a sorted pref unit test +I would say Derby will not close rather than Derby will fail to close because the second phrase suggests to me that Derby is doing something wrongThanks Rick for catching those items +We are working on this issue and for the same we have done the following we have created a view entity named using and Entities +ThanksLooks OK but I have to verify how that behaves with various startstop method combinations +I see thanks +Hes right Max we actually removed functionality compared to the beta +How often does it fail for you +Verified at r. +But that shouldnt be executing for you either +Any fresh thoughts on this issue? How can we build groovy without relying on logic +on patch for +Thanks Nencho +have committed this to and above. +I wanted to specify which artifacts to scan basically because I have use cases where I depend on without wanting to run the tests in them +Could you please try to be a little more systematic +I will take a look at it and get back at this +Thanks for this patch Manjula +There is already a limit operator we can work on pushing it up as well but both these approaches seem independent +Before reporting this issue I investigated for some hours but could not solve it myself +I think that is sufficientDont think its worth the change +That is the default +tcksvn commit m Improve error handling add query string to unexpected exception srcjavaorgapachejdotckquery Sending srcjavaorgapachejdotckqueryTransmitting file dataCommitted revision svn commit m Improve error handling add query string to unexpected exception srcjavaorgapachejdotckquerySending srcjavaorgapachejdotckqueryTransmitting file dataCommitted revision . +can you please allow a smooth transition over to your code +Hopefully less failures of the testHundreds of tables in +I dont have deep knowledge about changes made to airavata code so attaching a patch of my changed before committing +instead of changing we can just put in a Map we are already storing the requestParams Map no need to changeAdded a test over Nobles patch +Committed to r +Should be against component +Closing resolved issues. +uwe here is one i found that got skipped in I didnt check contrib +Trying against hudson +There is an EXTERNAL TABLE feature where you can give the location of the table such as file during the creation of the table +Ben There is a max parameter in the constructor +Elena please to verify that if this issue still exists or not thank you +Finn who is our properties guru wrote the class +I didnt get a deadlock on the original Hadoop release +The conditional writer is an inefficient way to write mutations relative to the batch writer. +That is +Is that so? If not let use an H database +In ideal conditions on a node cluster Ive been able to run through the ruby shell script in minus the enabledisable calls and got the expected count of rows back +Im not sure its documented +I believe this was already established a long time ago but opening this Jira as a formality and completenessWill close it in hrs with lazy consensus if there are no objections +Does it sound like a proper solution + +Verified in SOA ER. +Anyone know the status on this or how it will impact the need for hbase to implement thisDo you have an issue id where this is discussed BruceThat youd get the bloom filter exception on table that doesnt have it enabled or that never had it enabled in the past is odd +We changed the on our own but we would prefer to change it in the core itself instead of customizing it ourselves +Fixed in svn revision thanksdims +Changed visibility of constructor to protectedsvn revision to the branch in revision . +Still more testing to do +am okay adding length to blocks but it should be done in a new issue. +Solved by copying the client directory to the directory and by adding them in thearchive attribute in the appliet tag +I need to do a more detailed test but for right now I just renamed it to and dropped it into my Maven repo and commented out my temporary and everything worksNothing breaks anyway so I say ship it! Thanks so much Sergey +FYI I created an issue on legal to categorize the IPADIC license that we have some feedback on what is the next step we need to do to move forward with this feature +Reviewed by Barry L +Thanks PierreYour patch is in trunk at r at r at r +server can now be stopped using or stop without will be in build +Im not sure how much of that design is predicated on what was done in WTP before but its definitely odd +Thanks for your prompt response on this +If that is true then why are you using a file? Download better the zip fileIn short workaround Download the +jar etc but never search for +Attached a would be really nice if create returned a public type since the comment in the source code indicates that it is a hook would also be nice if a custom could be plugged into +jar file and are found in the classpath +If you did send a cancel call message youd probably want to remove the call from the servers work queue if it is still waiting to be processedIt is tempting to try a strategy where you check the age of a call when you start processing it on the server and reject messages that are too old but the problem is that it is not seconds from the start of the call but rather seconds with no data received on the socket which is hard for the server to estimate +Perhaps worthwhile investigating +fix javac warnings +Thanks Jake! Well add you as a contributor so we can assign this issue to you. +does this work for people who already have tables with replication scope set to +In this patch I do not think that is the case as a new RAF and channel instance will be created for each new lock instance +Defer all unresolved issues from to +In fact I consider it to the the killer app for Maven and for a long time was the one thing that kept me on Maven rather than moving to Maven +bulk close for release +This issue is resolved from a docs perspective +This removes the use of the factory name deferring to a simple Instance mechanism where the is a simple counter +Rather than increasing it blindly a better way may be to estimate how many rows can be fit into memory based on the row size and available memory and adjusting this parameter automatically +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +Bug has been marked as a duplicate of this bug +All went well +Closing the issue +Please could you check that your changes have been applied as expected +can you manage to post a patch today or tomorrow so that we can include it in the release +Thanks for the pointer I will investigate whether the libffi bindings get loaded properly +The editor is usually coming from the Errors object that is exposed alongside the target object in the model +Add the TTL to the delivery from broker to consumer if present +fixed in HEAD +Thought it is a good idea to have you guys review before I check this in +Can you confirm the issues is fixed with this combination +This used to work in but since we are unable to connect successfully +As encapsulate all Exception types we should log this information +plist in Xcode +It should be since that is the version that will consume EAP +Done +Yup +Should we do something about that as wellWould be a nice improvement but it cant cause a deadlock since its not being done from an IPC handler thread +It publishes to JMX by default +I described the problem and attached a patch to that issue +The way things were in the existing code does not seem correct to me but I need more time to determine what to do about itAlso we should probably discuss whether the required changes should go into or be included only in the next releaseFor now Ive tried to disturb as little as possible +After I converted the file to run under Derbys test framework the disk activity LED on my computer would go wild every time I ran it +Hmmm +we have found a solution for this is the problem on Linux or Windows +jdk supports the option XXWith this option the JVM will create a dump it goes out of memory +Attached you will find the updated Jar file for the implementationChanges include that string values are now quoted correctly when base encoded +Bilgin you will be able to apply this patch yourself soon +Right now like you said there is only one API that updatesreads log and it synchronizes those accesses +Haha! Are you crying about the choice of defaults or about my first patch being submitted? +Just to clarify those are not standard Eclipse tools and we are short handed right now to work on support of those tools unless tool owners or someone from community would want to implement that +I have merged this to. +I guess we have to have a bake off and decide what the value of ahuman readable files is +I will try a more recent one when I get time but you may want to check it made it overAs to the duplicate element issue +This causes a hang when we try to shutdownThe obvious workaround was to unfreeze and then throw the exception +Flush by RAM usage by default +Rich Please review these Javadoc changes for VThanks +I tried to apply this patch for to trunk but got an error I applied patch successfully first but this one failed +I think the reason is that the eager load group specified is not defined +The fix looks good for me regarding the issue +It seems like you dont have to do step +Perhaps Stephanie +Or it could live on a wiki page too +Unscheduling due to inactivity +Denis reported yesterday that everythings working in his setup +It is still up to the client to display this issue +I think we should do this refactoring in another jira +Some more cleanup Todd I addressed your commentsAs regards to repeated code even though that part of the code is changed and looks different it is still there +bq +bq +MingI havent looked into this but makes sense +Thanks Paul! +This happens only with some images like one Ive attached +Have you tried to replicate this bug against the latest version of Castor +If we advertise that file permissions are implemented in this release then we ought to attempt to make sure that theyre usable noyes of course should be usable +Can someone direct me to where in the source this is hardcoded so we can manually fix the issue and recompile in versions less than? +Not a strong opinion but alternatively the build system could duplicate the file during packaging and we dont end up with a weird empty fileThe problem with this approach is that unit tests run in might not see that +Attached is a new patch to integrate the functionality provided by Mark into the class for the version of the Sling TaglibsI did not reproduce the use of as the referenced issue has been closed +They also are saying nothing useful about regularization whatsoever +are logging frameworks +Why assigning as seqNum aboveTo have some sort of valid sequence number +Seems like it shouldve been removedBTW do these files need headers? I believe they support commentsbuildsharehcataloghbasesrctestbuildsharehcataloghbasesrctestbuildsrctestbuildsrctest + you post a test program and your config? Then someone can take a look and try to replicate it +Here we go +This bug has been fixed in Xalan Java +Ok thanks a lot for both of your help on this! Would either of you like or is interested to submit a patch for the fix +It knows how to look at files start jobs etc +This assumes that is also applied as it fixes two javadoc issues as well +What aboutoffering this by way of a property in the file? There isreally no easy answer to this issue +Can we close this Mike? It seems to be committed +Im looking at line of but I honestly can see anything that can cause a NPE +To get around this you will probably have to play around with the endorsed directory or set some system properties to configure Axis as the SAAJ implementation +EvgenyI do not think this a blocker for +I am attaching a patch with the javadoc clarification on the runtimecomplexity of the method +Closing jira. +Ok +It works perfectly here but it may be a windows you please provide a testcase I can run here? +Everything passes now +Moving unassigned issues over to the pool where they can be worked on for the in all probability patch +jon is implementation of interface +More is needed +The problem with choosing the first message is that the integrity of the index can only be validated using deep iteration on the log and the index then points to a message which doesnt have the index entry given in the index which is very oddThis code looks very confusing to me +moving to M +Kristian first of all thanks for verifying my changesI was planning to backport to next week after the changes have sat successsfully on trunk for a week or so +I will commit patch to trunk tomorrow if no objection +Looks good to me Doacan +Removed silly commentsAdded to the list +Verified on jbdevstudio +High difficulty setting up a Deployment method seems a separate concern to this issue +as a decimal separator for numbers that is a bug +Also removed cobertura GPL jars +Anyway people can always copy your code and tweak it to their specific needs +Request commit to by Darryl +Also is it necessary to move to +Could it be a problem with the state? Are you using server or client statemyfaces currently only holds the last state if using server state +Applied patch from Jasvir +Alex great work! Let me know if there is still a problem with +Ok so it explains the change +If the truly problem is the JDBC drivers is possible to use a saved schema for updates detection +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them. +Thanks +I think just change that to say that itll produce the sum of the total group counts which is an upper bound on the true valueIf documents of a group do reside on more than one shard then the count will always be higher then it would be if all documents of a group stay on one shard +Closing old resolved issues +Fixed by forcing session creation +Promised +Implementation based on similar is the code in patch sorry for not noticing youd already attached the Andres +Happens inAlso includes resetting the Enable direct integration with WMS setting at every restart +I ended up falling back on adding it to my XML Catalog. +This fix will be available in the next snapshot release that I will put out in a few hours. +I verified that I get the out of memory error out of even with trunk and that passes fine when run independently +finalizeJob locks the s instance after taking the lock on the itself +I was in two minds myself where to put that +The issue here is that the stateful bean lookup is triggered in a TCCL which doesnt point to the deployments classloader +Look at issue closed. +I will post an alternate version of patch following your suggestion +I still havent resolved the problem for that one workspace but I am going to mark this issue resolved here +Thank you for the reply +greatBTW I wonder if its worth considering JAXB support too using JAXB to turn XML into supports the xstream wireformat. +It makes sense when youre using a different interface for listenrpc addressI should also note that for weve added gossip states and we have padding states in so we dont really have room to add anymore in this release +Please create a new Jira for the docs now that this is resolved +Take a closer look at your code the second message send is setting the string property on the original message msg instead of my fault +Im going to wait to talk to Randall before committing +A possible implementation +Sorry I cant its confidential +and ROOT editsI wonder what effect this will have on performance? If it slows things down thats probably ok since catalog edits are relatively rare but I wonder if it would bring the server hosting catalog regions to a crawl +When we serialise we can override the serialize method of the so that it will first serialise the Header and then the body +master assigns region to a region server +Provides detailed stats on webui so that we can know if there any blocks that are not verified in a log time +If you can point me in the direction of where this might be happening Id love to run this down and get a fix out to you ASAP +Please reviewThank you +Minor changes changed region count to make sure theres split instead of preventing it +The changes should go live in an hour or twoResolving as Fixed. +A preliminary analysis does point to a race condition between a jvm just cleaning up and a new task trying to kill this old jvm for a free slot +AFAIK Hardy now talks about the root entity in the doc which clear the superclass doubt +see and +Delete from javatestingorgapachederbyTestingfunctionTeststestslang +fixed in r thanks Gabe! +Will change to pull parser and more robust checks next +I suspect this is problem is like +Minor nit Id request you to add comments about the locking assumptionsorder to some more methods which need them +Go to the client directory +Both tests fail before the patch and pass after the patch +I already have a working prototype +Perhaps the newer releases of the resources plugin have resolved this? +removing fix version we cannot break api that badly in that Nabble removed this mail thread from the archives and now the description and discussion are gone +In that case it is probably good idea to remove the stack trace even in development mode since there is no exceptional situation provide another command say entity to change the current entity +should what each have a different tableThere is a separate table created for the mapping column a many to many results in tables +how about you can validate different things? I forget the specifics its been a long time since I dabbled in XSD +Shouldnt this have a Fix version of? Last status published to the lists seemed to indicate this was the only thing holding up though I see other issues assigned to now +Verified in because the test was failing this test is temporary disabled. +The only way this can work is if the rowid is required in the query which will require us to parse the query at the controller +if we need to keep the current impl then at least we must refactor so that all those extra dependencies are optional and only used in case a user actually specifies a wsdl importthe reason is that in the setup im adding the capability to build and plain java projects +But is there a way to suppress the warning on the console +Cool so the short term solution seems to be Switch to a soft hash map for the cache Make the cache size configurable by the adminperhaps upping the default size of the cache as hopefully that should is the first cut at a patch for this one +Ann in a perfect word when you manually review a piece of code you should not have to create a manual rule and to attach the manual issue to this rule this is the case just due to some technical limitations +I decompress the of by and I dont see theI think you may miss this class file when making the +The project administration page Exclusions is going to be removed and replaced by a Settings Exclusions tab +This is why anything editable by the server should be stored in the server and vice versa +It will save many times for usFor example BEPL againP now I have to create another component to store partnerlink definitions +Thanks Erich. +I have found a way around this in my code base however I still think that this would be a useful feature to have in Janino +It only occurs so far on mode +I think one way to resolve this is to change the logcxx threads so that each thread has an APR condition variable associated with it +So why should one throw a and the other return? Other similarity implementations that compute the similarities from a can throw as they know all the preference data but not these +The faultString of the marshal exception is now compliant with chap +Sorry about that +it was on a bad node the RM will restart it +Thanks Rick +arc patch will work with an updated Arcanist if you need any other particular workflow to work without a certificate and it doesnt right now drop me an email I can probably fix it +Thnx for catching +committed to branch and attached patch for it +Dont bother submitting anything to upstreamIf the source folder can be empty then thats certainly better than building an actual source feature +This sounds great SHOULD be used with languages +Also available on in homeniclaspivot if that helps +returns a non +We can then further expose the option as needed but at least it will be there if desired by a user. +Mike thanks again for the great work! Thank you for the sudden burst of effort to make this happenSo I think this is ready to commit +But it would be nice if you could add some javadoc comments for the interface just a quick note about what its purpose is +Im not sure this would be a blocker +Patch cancels and replaces completely PatchIncludes modified and added sources and a svn diff fileRemoves dependency to Correct the value normalization +I get the message Comment authentication failed! on my blog but only on the actual post +The provided for spring comes from the RPM and is the spring upstream jstl richfaces and struts should use from the is by design +I submit the patch +Updated title and description to more accurately describe problem +I believe this now fixed +We now log a warning in Development and if a Dependent scoped bean gets created via +I could not find anything in the spec which states the default value for name must be the unqualified class name? Did I just miss it? Or is this is a hibernate specific decisionDo you know how it is recommended to handle the case when a product is including two thirdparty jars with conflicting unqualified Entitys where changing the source is not an option? This is the case that I am concerned about wrt this issue +Yes that is the plan +The code looks good +not ready to comment on +I committed some work last night very late and got tired and went to sleep +However I am not able to reproduce the failure +Eclipse reports a warning on the cast of the Map passed into the method but without the cast it fails to compile +Hirankit Can you show us whats not working as a test case? You have shown a lot of output but not shown whats actually wrong with the output +While its a shame that the interface is not workable as originally designed the workaround is fine for my purposes +Changed to actually check in connectors folder for existing jars +Patch in a modified form is in CVS +Was this fixed in that version? If not can someone tell me what I need to do to fix it? I have a deadline in my project and I need this resolved as soon as possibleThank you +To further my point I will say that these example tests failed last night on both branches so if they are really unit tests as you claim then the example config sucks because it fails oftenThis is why i think the two concerns should be separate or it isnt a damn unit test at all +Patch attached +It should have fixed too +The Servlet specification does not support your view on the matter. +Fix available as of revision +I have committed this to trunk and +Patch enables drlvm run with launcher on and tested on Ubuntu Committed in r +Its also reported as records are read which should be concurrent to writing +At this time we do not know any workaround for it +Marking for so that it can tracked accordingly +I would suggest one thing +let me know if this is not upgraded in the beta and Ill edit the release notes. +Virtually everything has changed because of the jalopy reformatting so its unlikely we would be able to use the svn merge +Great point +Im going to ensure that this is the case and if so resolve this +Move the fix to +attachment removed by users request +Added UI MockupNew project +New styles +I am running some further verbose tests and may have a better idea shortly +I have done very simple testing with a quickstart but not extensivePS the patch is against the wicket folder or submodule itself +Sorry my bad +Committed to trunk as well. +This is fixed in SVN. +Thanks. +has some code for creating images in Rackspace +Patch with code to apply patch and run svn add commands on new ve made some changes to the parser +Looking at the test failure it doesnt look like it is related to this +Thanks Todd and Scott +Do we need to update to require the timestamp for ops now that Thrift isnt enforcing that for us +So that at the time of writing to the Container we can record the locations in the file where the individual Blocks start +It is recommended to not call any method on these instances after the factory got closed +Done for need to port the fix to to head as well +Being as youve marked one as removed it will then remove it +This was also a in contribqueryParserI think this is ready to commit +Sorry for the delay package tangle index is not part of Total Quality but sonar core +Sheesh. +Not bug +Guess only clue was the fact that I did put label on +Take a look at svn diff +Thanks Elena +Getting started on JDK greatYes +Sorry +Username ktnagel +Yes Ive checked that I did tracing for length argument in characters methods and length was different +Carl thanks for the review comments +Thanks for explanation +nit can be addressed on commit +A temp reply queue exists in the same broker that manages service input queue +srcjavaorgapachecommonslangexception +Note two classes in the it seems the problem is with transacted and CLIENTACK +Thank you Shane! +Any notion whats wrong with my attempt to run binhadoop jar write +Committed revision Thanks Benson +at least we should fix the testsNOTE reproduce with ant test Dtestcase test FABAC true esPY CanadaMountain +There cannot be a case where it is more correct to store a sharedptr to a newed object before it has been initializedNihil Obstat +No updates other than I have created another issue in JBPAPP +Thanks guys! So let us bring over only whats needed and no more +Patch file for changes to web controls section in patch to remove formatting differences causing large diff +But its frequent enough that Im convinced something here is causing an issue +Patch committed on and no further commentsissues have come up so resolvingclosing this Jira entry. +Attachment has been added with description Patch to to use a fallback if the user directory is not transition resolvedfixed bugs to closedfixed +This is fixed in which will be used for the EAP +Committed as +Redoing things like rename and length are not future proof because folks can simply use File apis and get incorrect results +I reverted back to the synchronized version which yields better performance and therefore no more race condition. +Added to the set of Affected Versions and set Severity to Incorrect Behavior +Lets see if this occurs in too +getrangeslices is thanks for the help! +It seems that Lucene should have ways to incorporate new bitset implementations in the future using interfaces and things +Is this a real source package you mean a jar dependency what +This would be an issue for the launchpad base checked adding +This issue subsumed by Jons work on. +But the core code now compiles without any unchecked warningRevision + for the patch +Included locations of and stated root element of in trunk r +cache misses +Hi where do you get the s from? Do you convert it from the document +in live cluster works fine +if the user specifies a type system then do at all cost use this oneIf its rather a type system then we should move the search strategy code +Since Im not a native English speaker Im not sure whether I made clear when the issue occursa NOT when the web app is loadedb The issue occurs at runtime when an object shall be persisted which belongs to a concrete class in an inheritance Hibernates entity classes +I tried debugging but it seems like the request doesnt even reach the servlet itself +Committed branch and +Yes its in STS + for the to trunk. +In such case the version String can contain any char even not a valid can solve this issue for m requests but will make the artifact not available to m until m artifact version is splitted a more robust way +Forget this is solved correctly i forget to set encoding and had output xml first shot on this everybodyI had some spare time and started working on this one +and that set of records are passed to much more complicated javascript functions +In this case the object is pulled from the session cache in the same thread +I added a feature a quick documentationAdded on cellar trunk revision on revision added a restriction to the logging of outbound events so that the log isnt spamed with warnings +as i mentioned before this is a likely error if your using unidirectional mappings and start from theHope this helpsGreezMarc +Moving checks before we perform a zookeeper action +Next up Make progressbar choppiness go away Use appropriate skins for background and buttons Hide yes and no buttons after download starts +It works for simple case but lead to problems in complex looked through Java EE spec and found that deployment descriptor is optional for all kind of modulesStarting with version of the Java EE platform a web application module an enterprise bean module or an application client module need not contain a deploymentdescriptor +Due to the raising adoption of portal servers like liferay in for example Germany I see a definitive need for portlet support in the major frameworks +We still need safe closures +Im not sure why with Adobe Flex you arent seeing this same problem but it could be that I still dont have your test case set up properly +patch looks good here is one that fixes that Cancel actually cancles and does not show the portforward dialog and attempt to fix widget disposed when put in background + for the fixWe agreed on the list that we will NOT do any dirty tricks for container bugs if they are not really necessaryIn that case I frankly dont care if the shutdown fails in a single weld version +Patch looks good on read through +So I am closing the Jira. +It has not been possible to reproduce the NPE seen above and Rob is unsure it was on a clean trunk following some changes to persistence etc +Committed r trunk r branchxThanks David +bulk close of all resolved issues. +Fixed on trunk and branch +Patch will save skipped mutations out to separate files in case user wishes to perform forensics +At this point I need to see the behavior in order to debug it and it is not a trivial task to get your application into working you spin up a simple test case and try to reproduce this? Ideally the only dependencies would be on Spring and possibly Spring Security +An updated revision of the standard RFC section example follows the same reasoning +For the dialog to be shown must be equal to true which is the case by default +Done +I have tried this with Geocities servers but I cant reproduce the issue with RC? Can you provide some more info +In that period you wont have console access +Removed check for existence of max +Jakobs changes in have resolved the problems with this sample +I already did this +Fixed in to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +just committed this. +My current patch can be made more performant Ill work on that before submitting it +Can you please provide us some feedback otherwise I am not sure what feature do you want and therefore we may just resolve this issue as Not like we can explore to any child folder if we have a parent folder in case of file integration support need the same for ftp +However this is different from the POSIX standard +the backup node can just any blocks it thinks are when it becomes the primaryPreciselyHowever when we fail over from NN to BN there could be blocks in the BN that we should quickly find if there are any blocks in the BN that do not have any valid replicas and then query all datanodes to send a block report for those blocks +bq +Moving out of +Hazem thanks for your pointing out the issueUnfortunately I had not had time since mid August and I had not noticed your important comment until Lewis mentioned today as aboveI think should fix the issue +However I do understand that the loader not failing nor being able to give any indication of failure is a problem and now that we have we can address that +Closing this as not a problem +Ok about messages processed without any errorThanks. +but this would seem to only get us halfway there +Thanks James +Rather than synchronizing maybe it would be better to simply provide a means to turn the EDT check off + +abandonFile makes sense or abandon or something like that +If you see anything else let me know +Since Im generally paranoid about legal matters it would be good to get a copy of that class with the Apache license already at the top of it since just adding it myself seems a little dodgy +defer all issues to +Moving all issues to the CR +Thanks Ku Jun Guo for these comprehensive testsAny failures against these spec tests should be opened via individual Jiras +Thanks for the infoSo theres no way to apply this one in Eclipse? Id like to see what happens ina convenient way. +I shall rename this package +Verified closing bug +JD is going to have a look + +And to tell the truth I cant think of a reason that it would get in the way +But i havent looked into the most updated code to check if this really applicable so its more like food for thought +Note that Tim Williams found the bug in the file originally +The patch doenst break any tests so i am going to apply it +for the patch +Ill document this when I am addressing +Thanks +I surcharge method to remove the problem +Can we change it to subtractAlso it might be useful to add two additional methods to just check for operation safety and not throw an exception +Can you attach them please? Thanks +Has this been applied to? Or do I need to apply it manually? I tested it on and it does not appear to be in there. +The same fix for youre willing to make sure the patch applies against test it Iwould be happy to commit it +In this cause no automatic splitting will be done in +Patch to create a more specific error message +To clarify the comment about testing I meant I could not figure out how to make a test to test the fix as the test would have to be able to see if a row was inserted +I should add a quick note here about as well +I successfully issued a kill request that actually kills the application +Updating patch for earlier hadoop version on behalf of Amar +Not sure whether this will deploy on its own as a +Mina version Committed revision +But seems that it does not fire in any of my cases + +Zhenxiao Patch needs to be rebased against trunk +rebased fixed thrift and cql tests committed +I will continue to testreview +just a quick update +Updated corresponding patch over in as well +Thanks enis +My bad +Thanks Houx! This patch looks fine to me +r on trunk with suggested fix by Caleb Land +Ive just committed it +From the Designers standpoint were moving in the direction of removing any notion of Default Username and passwords +This patch adds exception handling with checked exceptions for namepath handling +Any update on this? Is this ready to go? Thanks +decode we now have a basic parser extracting all entered values and their keys +Is it possible for you to tag new source which depends onThanks +Ill be as happy as a tornado in a trailer park to take this on +A workaround for the lack of this feature would for EJB clients to retry JNDI lookup on a catch section +If it cant be cleanly applied let me know +PR this issue as it seems to be a bug in Forge + +Thanks Scott +Anyway the behavior was wrong in and and is now right in +Readme file includes all relevant infoVerified on CR. +It was signed by Comodo? When were the certs issued? Its possible that the old Comodo certs were rightfully pulled from Android +Especially given it is the same as instanceof Cloneable +Basically just to have some test code that runs against it as a quick sanity check to prevent breakages +Hi TimTed To build GEP what is the recommended version of Maven and subversion +Im going to assign this to Ben because it might lead to a release and he will be the one that makes that decision +It includes an update to tests but unfortunately theres a problem +This allows to reproduce the reported a look in +Updated for new implementation +If the server has been restarted it will have no knowledge of that old session. +attaching a first iteration of the patchthis is still work in progressmost of the tests pass +in JBDS Version Linux JDK +sql file altogetherThe storetests are merely a group of store tests that could run after one another while using the same database there is functionality that sets them apart from other store tests so there should be no objection to moving them over +Ill upload a new doc with more details and incorporating your changes. +Rather than changing the default behavior of the in code it seems more appropriate to use the configuration option to toggle it for your applicationIt sounds like youre reluctant to do this since you dont have easy access to modify the files +Diff for the aforementioned changes +Very odd timing type issue +Hi Robertyour patch is interesting but its pretty hard to understand which changes you made compared to the original one as the in Infinispan is today likely quite different than the one you started fromIf youre not familiar or interested in Git could you send a patch or diff file from the original file you edited and add some commentsIf you could apply your changes to the latest version of that would be ideal. +Only info error and debug +I believe that all slfjj jars even should be removed from the war +Thanks Bela for all your answers +So whats the issue about exactly? We expect region server to time out opening AFAIK so is the problem more about the idempotent nature of opening a region and then failing at doing it when its assigned somewhere else +Thank you so much +I guess it depends on how smart the compiler is when compiling for jsr +Andrea I dont think this will mandate sorting for those data sources that cannot +OK +Anyone want to weigh in on whether its worth the effort +Here is the patch for the class loader it modifies the way deals with invalid urls it does not set the invalid urls to null in the internal list but put them into internal excluded list instead +However this is my own strategyRegardsClement +Withdrawing patch +I suspect this is a an incompatible change and b will therefore require a release note +there are some small adjustments to the patch I would like to see +Ive probably missed something +Steven if you have timehardware can you retry your with in trunk +Looks like that library is somehow the output ofls l optwwwjbnhqdblibAnd compare it to your installation +I get that env is used to make sure we invoke the sh command specified by the users environment but is it some sort of POSIX specification that env itself is always located in usrbin +also can you add a tryFcntl method that does at least rudimentary error handling the way tryMlockall does +Someone please close this JIRAIt is no longer true The server will now start with Java without doing anything special +I added the test Philip suggested and committed this. +Fix committed on trunk + +The problem here is what PUBLIC in configuration properties this is not true and this is not editable value I try to search project catalog but do not find a way to edit it +Please check it out and let me know +They accurately describe the new ALTER TABLE behavior for changing the nullability and default value of a column +Idea to prevent loading of properties at every construction of Marshaller through +Thanks TobyBTW I wonder why we are doing the job of the jscalendar guys +Both changes have now been doneNote that the only Wiki Enabled Screen Scheme that use has fields enabled for wiki markup Comment Description and Environment +I seem to recall a different JIRA about separate Riftsaw vs +I can CdricWhile reviewing this patch it occured to me that it may be better to add the listeners for sessions at a upper level so that the listeners could be informed when sessions are created +How big is the WFGEN WAR? why not just include the WAR in the TARBALL and quickstart instructions indicate copying it to webapps? this seems a trivial step no +It must have been public when the release was cut because the class turns up in the published api javadoc for that releaseTouches the following fileM javaclientorgapachederbyjdbc +Ive added two new integration tests to verify this one for windows and one for unix line endings. +The other concern is we only change not byte +Fixed in revision See the Subversion Commits for a link to the changes and a patch that you can apply also to the Jackrabbit sources +I created to deal with getting the server version displayed in the report +out find v +objections to checking this one in? from me +Ask to hudson +Please verify +I agree that we should delete old split logs on master startup split +I believe its ready +xml works damn near anywhere. +Reverted r due to regression see linked forum +A screenshot +It is negativePlease let me know if you need any specific detailsThanksRaghu +I am however now getting some odd results in terms of what is considered a relevant line while im using +BULK EDIT These issues are open to be picked up +It just needs to be used with Spring which will hopefully go into Grails. +When you feel like you are at a good stopping point lets go ahead and get the html status code in the repo so others can help you +thanks Ivan and Sijie +Do we really need to have core toolbox and plugins? Cant and be plugins? If not why? Do we need to change our approach to plugins so that they can include client toolsTed +There are currently no plans to include this feature into +If this is a CREATE or ALTER TABLE statement then is not a column in the target tableThe problem with the query is that a column in the A table is referenced in an ON clause even if A is not in any of the operands of that JOIN operator +This patch fixes the issue now for all casesThis just removes anything after from the version string before it compares +Ive forgotten to complete the last update +Tijscan you please look into this one +if we take that off you will have to make sure that access to all shared values such as fields of a component is properly synchronized yourself ie by using synchronized keyword locks blocking queues whatever +No response assuming it is ok. +Logged In YES useridJust to make clear +I built the jars from thrift revision from apache +I was wrong about logging the local user in with digest from the local browser +Code implementing the proposal +changed splittingmerging so that it can start only after master knows about it at first +HTH +Attached patch fixes a bug in decrementing the metrics +There is the case where the RM no longer knows about the application so it will not be able to forward it automatically to the history server +OK lets hold off on the public APIJason can you give an example of how youd use this in? I dont see the connection +Patch checked in +Can this be committed? Thanks +Pushed in +how has this been sorted? I cant see any changes +Supplying a patch to add the output +Attached patch for trunk with the comments addressed +So make sure to take this into account when doing a mergeRegards closing of issues that are old and have been released or marked invalid. +This should fix the test failures. +One easy improvement The class of should be Set instead Map +My CVS client would not embed the +Could you not support the old params and issue a WARN on the logs for deprecation +Thanks Manish +Native and CXF stacks are called web servers there but they should be called web services or something like that +bq +Ive added a new stemmer in trunk called +Ifwhen we need we can simply make it support a list of filesMultiple properties is an uncommon case +I thought that the ant build scripts were supposed to download the automatically using the Maven repositoriesCan you thoroughly check the results of your build steps and see if you can determine whether this step failed or notI think the target is called installjunit +Looks like a good catch +Duplicate of I think +The access AT has a type which means you cant modify its contentThe only solution would be to delete it and recreate itThe former error is a different beast +If you use bits from the time but only take from the th bit then you will only be able to create a maximum of ids per millisecond before overlap +well be bundling FS clients but we have to test themSo we have to see how to modularize that +Volunteers to provide the translations +But switching back to Tab A and then again back to Tab B fails to render Content B +Support for overriding in geronimo plans was already implemented for and see We need to check this works for openejb as noted thereHowever as noted here we need something even more flexible +Committed revision +Postponing for now as it involves a bit more work then we currently have time for in the timeline +for Java and dropping Java support for believe I made all necessary changes +The form as is +This JIRA is reopened to address that +Session handling needs to be addressed in more depth at some point as there are now several places where flags are set indicating whether new sessions may be created and whether existing sessions should be overwritten +Prototype of codegen +very minor comments on phabricator +Where do you change the logging to use a url key +this is a pretty tedious way to deal with it +The new is now the default +Thanks Tyler committed +Its a bit of work and its a minor optimization +The datasource is not being deployed which is the cause of the test failures +If not on the server the script will not wait for the to stop before proceeding but should properly shutdown the system +Where would you put it? contribmisc maybe +The recent suspected incidence of this problem showed that NN failed to find a node on some racks only around attempts +if for no other reason than porting to GraffitoRandy +This is something I intend to complete when I get back +I must have done that when submitting to hadoop qa +Andreas thanks for reporting thisThis is a fairly easy fix and we love contributions +With this you can also change the toolbar without the need of specifying any menubar +Ive fixed the related jbws issue tests running on hudson this night +produces javadoc warnings +Ran test locally and it passedIntegrated to trunkThanks for the patch GustavoThanks for the review Nick +Then try to call any other method of bean in same threadServer and client threads hang infinitely as it was mentioned aboveServer becomes inaccessible through remotingTCP connection is in established state +When parent pom gets fixed and build is back to stable you should also configure CI job to remove SVN entry for requirements folder +Verified in runtime has been changed in server properties runtime need to be selected again +Ive sort of got it working here but not quite ready for prime timeThanks Tom +So if we do commit this we should list it as an incompatible change and suggest that folks who are short on RAM might use the prior value +Please express your wishes about how or what should the hibernate example application doThank youAdrian +Closing this issue if theres a specific test to add randomization to lets handle it in separate issues. +Done +Heres a first pass +Treating them as if they are ALWAYS used in that ONE CONTEXT is a FLAW +Im going to upload a another patch to replace the first one +commit aaebcbedcd + +Never mind Google has an internal maven repository thats synced well just deploy artifacts there. +Some of the comments there might be useful for working on this JIRA +So we may close this JIRA probably +All works +This issue is no longer there on the trunk +if its not a break why do we have the method there throwing +Attach a new patch addressing Ivans comments +Is there a way to get more details about the state of job retries? Being able to find out how many times the job has already been retried how many retries are left and for how long weve been trying to execute the job come to mind +Fixed in HEAD. +I also fixed the missing space. +Forgive me but this software seems to have sucked all my whitespace indentation into oblivion +Verified in +But if we do we will still be comptable with only half that audience folks who want the namespace URI compared as a string would be very annoyed if we compared it as absolutized and vice versaGiven that sitation I believe that rejecting them is no worse than either of those other choices +Looking at the title what would the benefit be of adding configuration to enable silent authentication for cluster nodes? Once you move into clustered deployments I would expect multiple remote machines to be involved which would mean silent authentication is not an option anyway +Id rather not expose additional functionality to Thrift when were trying to start people moving to CQL +I did a little digging on this here is what I foundI am currently running with the following optionsXX XX XX XX verbosegcAt the start of this I had copies of the when doing a ls l procxxxxxfdI pulled a heap dump like jmap dumpformatbfile and wasnt able to glean much from thatI then did a jmap dumpliveformatbfile and after I ran that the file count was So it appears triggering a full GC cleans up the open files +bq +Closing the bug based on Abhinandans comment +Within the message receiver i added code to throw an +Will you feel more comfortable if we move this just inside fairscheduler +Fixed as discussed on forum +Change historian column family so it times out after a week rather than lives for everNow if you select on catalog tables will probably need to narrow select to info else get loads of historian noise youll not usually be interested in +Although opening the project that had the page open might cause problems +Is that the required functionality +Yes it looks like it should and was about to ask if it does +Even at DEBUG I would still put some kind of threshold on it as you dont want EVERY read triggering the log message though at DEBUG it could be small like or +I managed to coerce Java into cooperating with int sized variables and its significantly faster now +Mass transition all resolved issue that did not see any further comment in the last month to closed status +Jaikiran yes for sure I will try it +Bruno youre right +Attached is a patch that fixes this by SVN revision +I think we could consider getting Pax Web upgraded to that version doesnt use Jasper and might get us around this problem +Fixed This issue and added the FAQ list created by Spencer to the svn head with some additions and modifications +Im not sure you should have reopened this issue just based upon that +We either need to figure out how to have structs satisfy the haszero property or take out the method guard for +Ok so this issue was resolved some time ago +Delegated to forthcoming dependency management. +Some code in dependent projects has to be changed due to dependencies +This problem has to do with the implementation of the app not directly with the iOS VersionIf the from the old app saves the localstore under the Caches folder this is independent from the iOS VersionIf the old app has not the backup phonegap implementationThe localstore is only saved in CachesIf you Update this old app with an app that uses cordovaThe backup restore method override the existing localstore +So I added two related issues to this one +The patch is for trunk +So all issues not directly assigned to an AS release are being closed +The query I modified is may find the logs above thats all I get +Here are the jars in this directoryjruby colindow jar tvf buildlib grep Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb Fri Apr EDT orgjvyamlb did a complete checkout of and I was able to build it successfully +committed to both trunk and the branch. +Test failing is help help for ls +Hey I checked on latest stable win and still the same error occurred +Will this be on time for the Geotools release +Initial patch for review +We already use the beginthreadex and endthreadex for this reason +So there is some problem here +The right solution depends on how we handle +Im about to update the patch to work with the current version but wanted to know if youd already done this before I spent the time +for its clean and simple +Attachment has been added with description Patch that should solve the perfect Vincent +We are putting our main efford into JAXWSJAXRPC tools issues may handled by community include it in +A workaround could be to set MAVENOPTS +Please verify they were applied as expected. +The new patch looks good +Thanks for the catch JD! +This is a transient condition and we can just make sure that we dont consider such a queue for assignment of tasks +TCK tests TBD +Should this avoid all stack traces and only print the message or are we just looking to add a more meaningful message to the exceptions? It seems to me that just printing the message would be the more complete solution since the stack trace is no longer needed in that instance but was looking for some feedbackThis is still the exception in also +I think it would work if it was moved yes but I doubt we will know until we try +Replaced non ASCII characters in repositoryconf. +I tested though building and running +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen +was updated to use iteration plus some code adjustments were made +Adding as a dependency does helpHowever both and generate erroneous code +nice catch mike! Here is a patch fixing fill this is a general trap and should be fixed in this method IMO +On closer examination it waits for either seconds or JMX Test Runner whichever comes first +The patch for trunk is applicable to +I can not agree I used database and the same error +Patch as described +Applied something along these lines +resetting P open bugs to P pending further review +Do you understand why the changes in are needed? The last part of the statement name is based on the section number so the patch seems to have increased the number of sections needed in this test which sounds counterintuitive to me +Thanks Abhijeet +Ilya any updates on this +Forgot to mention that I committed the patch +Modified description +How about committing this? Would be great to have it with the next release +I cant do anything with it +Korhan perhaps its a good idea to discuss things on this on the dev list +I attached a patch that does resouce filtering on sources in ear +OK I just thought could be build the Maven way +I rely on nodetool ring command to determine what nodes are up and their load distributionLet me know if I have not answered your questionThanksThe latest trunk works for you now though correct +not verified yetwas closed by script doesnt support custom namespace attributes in Safari thus broking Dn for beta +I think the status quo makes sense for Bigtop +Consider processes running in admin sessions trying to synchronizereserve using locks +Thanks +I made a number of changes last year to try to tighten up the detection of syntax errors in expressions it looks like this is one that slipped through and I think that ought to be regarded as a bug +AFS is taken by Andrew File SystemFiled sometime ago for renaming the existing classes +That said I hope we wont have to the release candidate +What do you mean by dropping all oracle objects? Is the equivalent to dropping and recreating the schema +I wont resolve this until you comment further in case you can offer some idea of what we can do with an element with no type other than to complain +There is no need for checking empty ackQueue because the packet responder is the only thread that removes a packet from the queue +As far as I can tell done. +Attached mapping files attached to Ryans original email +Other places where fparam is used still have to be changed +Calendar screenshot is updated. +would you please try it thanks a lot +I committed patch a to trunk with revision Ill give it a week or so before I look at backporting this to +In this case if we have not drained the stream before thenit is not guaranteed to still be valid in the store +but the problem is if you have some customized feature interface which extends Feature directly and also some implementation of this new feature interface +Thanks Ryan! When do you plan to release it? The version doesnt have this fix yet right +Reply received by emailI saw your comment on this issue +All Aries committers should be able to publish the snapshots +I build the karaf snapshot and reproduce +The extra files do not have content changes but were updated to remove a tag that my editor adds to files that I open +Just before I was about to start a full rebuild I happened to glance at my file and noticed the new default had been added +Was version released? Hown can I download it? +If it shows up again try to capture as much information as you can and reopen the bug +Is that what youre referring to +We will defn incorporate itJohn +Bulk move of M to move to CR +Its fine for bit machines running bit OSlibI will post my final comment today after some testing +Tests passing in both master and branch. +The should not be resolved at all it can only be loaded if you are either using or if you are running +Patch to fix this IssueIt contains fixes for both the specific TEST and for the function in Specific ReaderThe Test was producing a false positive by having the constructor of the record populate the object with data during the READ part of the test +So are you saying we could close this as not being a issue +So that package now has subpackages named io file util and mapreduceKeeping the package would be more consistent with Hadoop which has the mapredmapreduce distinction +I dont think we ought to hold this up for +So solution seems to publish the result of to maven after doing a clean first + +see the javadocs for ve seen the javadocs for numDocs and I still dont get why is it that when I delete de document with field patha the document with field pathc is removed from the indexDo you care to explain? +For a problem may be that after we upgrade the version when we find a filedirectory named +After many hour trying to find a way I could not find a way to generate this properly +No performance changes as wellIts good to hear that +Those interested in this direction can contribute patches to the branch and once we feel its functionally good enough to replace the current custom we can merge the code to trunk +If the prior AM attempt was backed up in event processing and launched speculative task attempts after a task attempt completed then it ends up waiting on them but they were never launched +When using an unidentified sender it checks the queue passed in to the method otherwise uses the queue which was passed at sender applied svn revision +What about adding a throw if the delimiter is the equal character? +Unfortunately due to licensing issues we cannot ship or even import JCIFS is a purely licensing issue and it has nothing to do with +Timothe information was only meant as you wrote +I committed it to as well so this will be inI cant think of any reason why youd want to wrap another with NRTCD? We can fix the docs to state this +Yeah not so trivial for an amateur like me if only we had an expert to help us +The only problem I can think of is if someone has say batch jobs in production and wants to write all their new ones in +Ill add a note as proposed +These replies would be nice to save in a storage for later reference when needed +The server must always respond with either a failure or missing dependencies message +reattaching the file +To be more specific Pig puts ivy cache into a text file will search for this file and load dependent jars +Thanks +Ed +Today I got the official APAR from IBMThis issue has been addressed with PK see httpsupportrscontextSSEQTPdcDBuidswgPKlocenUScslangenrssctwebsphere for more information. +prestore and store are invoked to put the object into the database +The worst case for on java strings is bytes per java character in a strings +Cleaned up +Were currently passing oauth tokens and verifying them upon receipt +This will be in the next update +Jeff since this is first case of using for clustering using can you please review it +We need Find Seam +bq +Will run it through tests tomorrow and commit if no one else does before me +In this case value objects will have to be generated into a separate Eclipsemaven module so as to be packaged in a separate jar file accessible fordistributable to clients +If there are only one join key or two join keys it will use or instead of +I have just updated the mentioned wiki pagePlease write in mailing list for support before opening Blocker issues thanks +Please limit the length of a line to chars +The previous patch tripped some errors in the regression tests +Build of sandboxes at r +I propose following +I am going to do the first QA build for on Monday +Categorizing bugs related primarily to stylesheet elements etc +For CXFs version of AegisSimple frontend not having them on the wire is probably correct +Its possible it should be added as a configuration if problems arise +Delivered fixes to solve this issue for REST calls +Comparision to show the layout problems when using had the same problem and fixed it with the following patch +apply from the root one simple change in configs +Patch looks fine to me +Although The other I think I can rewrite the test because the ordering is ok just doesnt fully match that of the original test. +Committed revision +Paul With the patch supplied for the is now saved rather than just the action +upon review this probably could use Expression support as well +there is a file a to in the folder frameworksrcmainwebapp +again thanks for the tip +Any plans on this? +If someone preferred not to store the distances at all for memory sake maybe then they have no control over this +a +the problem was that my path has old java in it and the server would not come up since it was compiled with java +tested! +is locked up to the outside +Removed unused imports one last time +closed shipped in +So there is no reason for the producers to hang as eventually the producers and consumers should achieve equilibrium message rate +Pull it back in if you think different. +Hi TatyanaIve integrated iterative classlib tests to CC +is alivehindessm and bodewig have accts +fix pending resolution of to roll this back if is rolled back for good. +Apparently I found it a while back see + +You need not to do any injection +A patch with the changes described aboveRelative to +Committed! +Looks to be very similar to hopefully a fix for one will fix its the same reason +Resolved by commit on out that half of these data members werent being used at all. +This is working now but well keep an eye on it +reviewed yesterday and should get to was committed can we close this Nicolas +Thanks +fine just in since its implemented in in JBT +It became a local decision at the namenode +Since xml works on Windows and Linux I removed the ifdef so that xml is always used +Thanks again David! Resolving this one. +I understand the specification in that way that you can not lock a node till it has pending changesIn the context of a transaction it is a little bit difference +Its possible we could make some progress on compaction in scenarios where a single sstable is corrupt +Patch with fixes attached +Pig keep on overwriting static variable each time we launch a new mapreduce job +There are few pros and cons with both the approaches +reopen if you can provide more information +project files but could not get reliable results +resetting P open bugs to P pending further file no longer present in transition resolvedwontfix to closedwontfix +Scheduling for +No its fresh standalone installation and my first experience with for ever + +I dont think shading will work as the Sink interface is passed between core and doing this there will be an implicit contract that the sink API can not have anything removed from it over time though thats probably fair enough +It was just the docs that were lagging +Wait +ThanksHoward +The User Interface does not present ANY avenue for installing is what I the toggle boxes are checked off for the mylar +They look like they are consistently affecting a number of the publically reported regression suite runs +closing since is now released +failure seems unrelated +Add Contributions section plus some minor edits +I decided to implement this feature by making it part of the new implementation +IMHO the default timeout needs to be override only when the expected execution time of a test may be longer than the default one +Patch with changes which solve the problems observed with Saxon on the classpath +It seems that I cant edit some values anymore after the handlers are created +about singleton viewsThe issue is not so much that an singleton exist but that the model code has dependencies to the view instead of the view being the one that listens to changes +Thanks Ben! It is working indeed very nicely! +so might just takes microsecond +I use this plugin on windows and problem seem to appear only on linux + for this since it is the right thing to do for CDI enablement +Closing for +Do you still have more to commit? I have some trouble understanding how it works currently as there seems to be no references to the in the code +Asked for review on the ml since Im no CSS on and transition all resolved issue that did not see any further comment in the last month to closed status +There might be several versions of a jdbc driver in our server repo so a selection dialog is needed I think +Later DN marks these files as corrupt since there is no metadataThe above is one of the biggest source of corruptions +I think perhaps reboots are more up for discussion but if I dpkg i I dont want it to start +maybe its related with multiple graph examples jar files when job submission is done +Patch to the to add a failing test due to this issue + for patch +The method implementations are very simple as the resource adapter is a front for a singletonStill get the same error +The Stomp Connector has the same issue although it is much more unlikely to happen +The change to broke this patch +Well it does export AND import those packages so in theory it should be possible for it to use your exports as long as they meet its requirements and are resolved first then it should work + +Ill add the checkstyle plugin and work on cleaning everything up a bit moreThe rat integration could take me a little longer to troubleshoot +Ill see what we can do about this in has plenty of protected template methods so you could try to implement there I suppose +hangs based on patch +You meant unless objection I assume +I opened to resolve the new bugs. +Id forgotten about that spec thanks for the reminder. +HADOOPOPTSserver XX XX XXG XXG +What is there is good to go but I did not cover all of our +this should not be done via dependency properites but instead by a configuration filter if correct output path would have the project prepended eg target +More to follow. +Healthy mailing lists some material already on main site +I agree with Anurag that it is best to use the database temp directory for temporary storage of +They both look the same to me +looks good will commit if the tests out some unit tests failed +big the more we get to pure bytes the better IMO +Yup the attached is the commit lock issue +Thanks for getting this set up timely. +Sorry for the delay in responding +Got you for your suggestion. +Used entity and work effort module calendar screens for implementing Training Calendar +btw the above patch sequences after the following +kont and it wouldnt matter if the URLkeeps getting kontkont +Sorry should have made this a minor for the suggestion Ive added the service class to the error message +WORKSFORME +I changed references from to +ill take a look at this as soon as the beanutils release is completed +Fixed now +I have absolutely no idea how or when I ended up working on this but I think the attachment nearly addresses this issue +A binary chop on the milestone builds shows that this seems to have been fixed between M and M +The bug was found in the FUSE distribution of but can also be applied to the official Apache distribution +Better distinguished between remote and local errors +Im tempted to classify this one as not a bug +in trunk and rev +I didnt much tests in the so I didnt write any +Ive attached the patch +And I have clean jboss as install +Wao thanks Savas for the prompt response and fix! Great support +Please find updated patch attachedThanks a lot. +Lets take an example testpool and jdbctestpool will get the same artifact id testpool I will append a random string if the testpool exists already +mentioned patch attached for posterity Im going to give a shot at just updating the analyzer now +We have not been able to catch the issue on any development system +I am usingI have open and click on a link on a web page +If there arent any other issues then can we go ahead and commit these +In fact this is the main value add of buffered flows that allows every agent to queue up events waiting for the destination to get unblocked +It will be simple stuff like version and some more portal related stuff +Closed after the testcase has tested majority of possible refer to the forum thread for more information. +This will solve a broader type of issues beside the join +Overall looks good +Review please +r +I think forcing it to be a string literal match would probably go against the general understanding of +Even if we manage to bump the version we should still not export the dependency via the Maven pom +Ning I had the same problem downloading the patch so it must be corrupted +What project file are you using? bcc or bcb? What source file was compiledwhen error occured +Thanks Jesse +Ive verified mvn clean install +Here is an example TXT file FOP produced from the example FO file already attached +Im a little unsure as to which specific race youre speaking of or whether youre talking about a generic exception that can indicate any type of race condition + resolved Release +Corrections are Getting Started with Developer Studio Guide ? Tools Installation The currAnt version of Tools +One thing is confusing me there is a file at both the trunk level and the individual projects level +Logged In YES useridI have the same problem with this bug can someonefix it? or I must use Integer Class +I am using Windows VistaThis problem is fixed in Eclipse SRThanks +Thanks for contribution and It would be great to have sample to demonstrate this improvement +A patch to patch to Sergey Pavel +et alias +dmp files are minifull dumps they can be opened by windbg or Visual Studio and give a snapshot of the crashed application. +Attaching patch for issue +Thanks. +How do I check out the trunk from the Subversion repository +Is this what you mean by sparse representation +The data xceiver threads protocols etc +Latest patch still does not apply for me removal of merge method and one of the imports are rejected +I have integrated this change +Im so sorry for wasting your time Ill be more careful next time +In some cases we have seen this last from several hours up to a day or two +for the a new patch to address the review comments. +Ive created an issue so that we dont forget implementing that including updating the docs and examples extension dependencies is not enough for jars of type +There are plenty of things that can be improved but changing stuff that aint broke induces errors and unnecessary risk +Set and Lewis Can you kindly upload the changes that you had done? That way your changes would be in the jira and people can look into it +bq +Its getting kinda busy +mex sample patch +When reviewing the patch please test the performance and make sure we dont the slowness observed atThis is a mustEven before that I dont think this approach will work at all when we consider streaming and pipes jobs also +Thats probably why you see a slightly different result when you build it yourself from head of trunk which includes both fixes +So i will do it module by module +Hi Shane No need for verification +Moving to open until the patch is changed per the comments by Santhosh and Pradeep +We could ship with a config XXkill p or whatever that trick is +Had to make minor changes to regression test code and some other minor changes +As this comes up in mailing list frequently lets revive this +looks good +Thanks +I need to think more carefully but Im thinking that we might need a different replication protocol +I can understand your arguments for expected behavior +It contaminates wrong setting +Aidan can you review this pleaseTest case +I committed this to trunk and +This results in a page displaying User Registration is disabled even with the first Global Admin userPlease let me know if you are aware of any work around for the same +svn +If theres no other idea or patch to improve this Im going to close this as being superseded by was too quick with my comment but contains a part of the solution +object and database entities as well as attributes and relationships of all kinds were included +Yes +I think the change and introduction of is unnecessary +Very few changes to +Actually this is in and was a old comment +What is wrong with using and the associated? These were added expressly for maintaining the whilst rendering an error view +This patch is on top on commit breaks the current patch +I believe this is fixed with having the same problem with this fix applied +I just saw you resolved the issueI tested the scenario with and i dont see the problem +Modified the checkserverversion patch slightly +The security API does not handle thats the issue +This causes issues with the java profile the negotiation seems to break down for some reason +It appears that the issue is related to the reuse of the message object in the sender +By Looking at the code looks like it is the intended behaviour in case of +Should we change this then in HV together with the other BV changes +I assume the same thing will happen if requirevaliduser is trueSo I dont believe the original problem of this issue exists it was just being exposed based on other settings +Werner the load testing cycles have completed and the bug that was there related to the memory leak has definitely been fixed +Makes sense tagging for +As far as I can tell the deployment planner returned the proper thing and then the deploy was given an alternate plan +Rakesh could you generate a separate patch for +I believe the Community team can assist with theseI dont believe this file is part of documentation +Committed v of patches as revision +Instead of the prints could you add a failing assertion to the test caseIt would be great if the attached test case could be modified into a patch against srctestjava +This path needs to be make sense to make as inner class of. +As per you suggestion will it be orgrichfacesrenderkithtmlscripts +Uploading patch for trunkI couldnt think of a way to test the patch +In Lucene there are a lot of local variable and unused import warnings +However there are timeouts reported for builds on the EC jenkins this is why I want to get the patch in on today +Modified to assign better name from Payment Group Member to Payment Group Members Edit Payment Group httpslocalhostarcontrolpayment Enhanced UI +converted all core tests to use Version looks good I will commit this with in a day or two +Just retried with Xalan after closing down some other apps and time improved to +Can you let us know what failed to run with these settings and what were the values neededSettings for each OS are not necessarily the same and if we raise values with no need some people will complain so we rather do it on an informed basis +Not a bug +At the same time I can imagine that there are applications which need functions like COUNTMINMAX to be performed faster then by scanning through all rows +The current patch choses to introduce two new CQL only executecqlquery and executepreparedcqlquery methods and start refusing CQL queries in the old methods +This problem is known and is actually under heavy investigationThanks for the report +If nothing is specified part is used +javadoc javadoc just committed this +to be clear what I get now isselect items Open Mapping Diagram editor with everything in it +So its supposed to cleanup all the files +However I have a preference for removing it all together since it currently adds no value +One problem I did spot was that the topic manager was completing the operation without error sometimes even when it couldnt acquire the topic +reopened just to set the resolution to with resolution fixed +Tested this on Solaris again and it seems to workCommitted patch b to trunk with revision There may be improvements to be made maybe tuning the output or removing some inputs +Updated patch uses the simpler equation +But now the builds are consistent with the published artifacts +Yep the issue persists with as well +It should end in a proper compiler error message +Hi Dag +are we going to include the yahoo translator in the kit +Ahmed +Okay heres what I will Fix the issue as reported by Michael as part of a separate Address the unbounded choice issue as part of this issue +I used the pom dependency which is not entirely recommend but comfortable for me right know +reopened these issues so that I can move them to Future and remove wishes +Ok patch applied please check +to this defect +Status message changes to rendering complete while nothing happens with appears to be fixed. +batch transition to closed remaining resolved bugs +Already some fixes by Bruno at r r r and Bruno +I reviewed the patch and applied the fixes for +case mm will be calculated based on SHOULD clauses and work as expected +Just some to understand whats happening +Since read has no schema location I think that your fix has to be it +Theres separate doc in parent jira for that we can add it to comments or book +After testing the noted above only works for simple cases +There are some distributed tests already that you can build from + looks good +The compliance tests generate the same number of errors and warnings before and after the patch +Im closing this issue as out of date +Maxcould you please verify this issue we could close it +Theres an old JIRA where we threw around some ideas similar to this maybe last August or so but cant seem to find it at the moment +I do not included it into commit because of problems with it on EMT +OK this is the way to check children presence in component tree and do not add it again +NicolasIs there any update on this issue? We have a production wherein of data goes to one CF and remaining goes to two other +I change the something Required is a lightweight version without database and http stuff +I would expect this problem to affect as well +If I get time Ill try and code a patch that doesnt affect the status quo and then we can start to seriously think about tacking webmachine onto +The application acls include the job acls +All attempts get killed as soon as when the request gets acceptedRan the system tests that were failing a few times and they passedAlso the ran the full unit tests and there were no failures in the mapred tests +Hi EmmanuelSeems like there is a fair amount of refactoring done since reckon we will have to wait till things are more settled before we work on embedding apacheds into our app again +Could you please provide a sequence of actions to reproduce this error or perhaps a unit testEvent types must be registered first to take part in the variant stream +Committed rThanks for the patchWe will close this issue after the Smack bundle is released +are you saying that if you do and without it behaves differently +Maybe sometimes its good for master to die +was assigned to Rob as the default assignee of ServersFor raw server usage then manual configuration of the classpath is the only way around thisLets focus on the mvn setup for now +Ive just committed this patch. +Im down with atjvmexit +Fixed on main branch of Apache CVS +Eliminating NAE in this case will result in that currently check for nulls and throw a NAE with localized message to instead throw with the same message +fastforward and fastrewind are still create tomahawk test app and check with t a bug +JensFrom the ticket in TomcatThe URL that is being encoded is +Thanks +The patch fixes the problem but it does not pretend to be a final fix as done offhand +The exception in this case is expected. +libs grep SONAME SONAME +I also understand your concerns that a reload may lose page content +please help +I dont want to hold things up so if youre really close Ill make an extra effort to squeeze this in +The average search time was around ms +I noticed that also put the network error filtering logic into and so Ive added patches for those branches too +resetting P open bugs to P pending further priority for bugs with a patch +Thanks + +Freeman why should we depend on a SNAPSHOT release of CXFShouldnt we wait until is GAWell the ticket should be assigned to Willem as he is our CXF man +Thanks Ning +it will work up though +True +Pls is passing in my env before and after code changes +It seems there is an issue in prepare there is no lock there +This is addressed in by the switch to the library to perform asynchronous loading of libraries and modules. +integration tests for the usermanager post esp page with usage for supplying these patches after having applied the last path for the of I am now going to apply these patches +Thanks! Its working great +Attaching Patch file +Thanks Sarvesh! +It makes the nested wrapper class public versus private +Will it apply the patch in the new version +and are bug fix release branches but in my opinion some of theissues surrounding using in a Map could be considered bugsAPI changes are not required to fix the bug so Im not sure if I should be adding and deprecating API methods on and. +I think this is a Tomcat config issue +I am confused +Im looking forward to this +This is related to container as well +This has been happening a lot lately +Initial patch for review +Moreover if the problem will be identified as described by Phillip it is then serious issue with JBM internal implementation +Forces a reload of the language on a configuration change. +BTW the reaper thread how does it differentiate old tmp content that can be removed from new content that might be created during the startup +Username lsanders +Said commit added a static method to which was apparently enough to change its serial +Preparation for Gora release candidate +Committed json and new jar separately +Usually the number of facets will be very small and sorting the list will be very quick +Paul +Ill put this patch here just in case somebody will some day want to use it +attaching a repro app to give us something to work is the current WSDL +It only tests if Clob experiences the bug +Can make it a part of +Closed with the release of Sling JSP Taglib +I think its reasonable to assume that exit within an embedded script will simply bubble out the exit status as the scripts result +I checked a simple case with the only proxy server setup I have and it worked +pending JenkinsPlease unset the fix version until its committed and set the target version appropriately +Unit test for duplicate messages is +I would even play safer and not setting ODF but as well not removing invalid attributesWhenever an attribute was removed due to invalid value I have now added warning log with this patchThere is still an invalid attribute in one of our documentsWARNUNG Attribute chart was removed from element style was removed due to its invalid valueit seems to come from the slide test therefore could one of your team investitigate in this problem +Ill close this issue as it isnt the best place for answers +realized that userName is not marked +Is it case sensitive? makes more sense than Is it case +Test should only take a minute +The suggested patch resolves this problem +Uses new features in CGLIB to skip interception during construction for all methods +Sorry Stefan but PAM was faster in fixing and me in closing even when you where indeed +Attached the latest patch with already converted was migrated to pegdown was very easy +I just noticed the Coldfusion is case sensitive with query of queries and Railo isnt. +Agreed +If a jar registers a service for some obscure image format but does not bundle all dependencies it would mean that the application itself can no longer handle any image format at all +This happened in my test setup also and was quite a pain +so ill assume this is a blocker +committed. +Committed. +Thanks Amareshwari! +trunk Thanks a lot Sagara for contribution of this new component to the Camel family +A best effort is all that we ask of youYour review will help us that EAP s release notes are as accurate and complete as possibleThankyouEngineering Content issue is out of date and will not be fixed +Oh and by the way did I mention we would happily accept a patch? ll commit this patch in a few days if no one complains loudly checked inOleg +Will be committing the patch committed thanks Aniket! +Seems to be fixed in. +Do you think the bug report can be closed or there is still something that you would like to do about it. +check the date of the isoif still issue is there then definitely there is some issue with the way we serializerDeserializer +Need to merge to branch when its is any workaround for this bug? We arent able to test every route where we use. +That is what we needed to run fine in a standard the extra indirection layer between quartz and jBPM would have added extra unnecessary did get inspired by reading their code and i have developed a unified job scheduler that is used for both asynchronous messaging and timers in jbpm +They include commented out jmxremote property hooks in bin as well as a blurb in +Closing all issues +We can live with this in the unstable release cycle +Namit I think its fine to take an incremental approach with this but then its importantto spell out what the known security holes are so users and administratorsknow what theyre getting +Added an archive parameter to support signing and verifying a single archive instead of project artifacts +I agree but in that case we should revert the initial commit +While Seam is integration framework +I fixed this issue during the hackathonPlease chcek with the latest nighly buildsThanksDeepal +Looks reasonable to me +Im working on rebasing +Please do not JIRAs against released versions +I tested this with a custom built Thrift library with changes will retest with and update +Then it suggest the options when chars typed in. +If the job fails in lots of records then there should be some problem with the implementation itself +not start dbAdmin and +java file what I think is good reason +Error is disappeared +Fixed in Git ID edbcfdfcadcefefeed +rd times a charm +Thanks Simon Kitching! +It is even possible to have multiples global transactions for a given session so having an independent panel still makes some sense +The Birt servlet was ignoring the locale order to fix the existing project you need to replace in your lib directory with the attached in the trunk +So it would seem that it was versionitis +Danny this is all expected behavior +is for branch and is for trunk code +A workaround exists After creating the publish area reselect either the workspace or repository combo selections +Another detail it doesnt appear to be the or syntax that results in not working but rather the presence of the annotation +Thanks Babak +This would be a very appreciated improvement +Open source revision for this fix is +this issue cannot be fixed at the moment since we do not support XSD validation via DOM API. +FixedIncreased the column size of DOMAINNAME of SECURITYDOMAIN table to which is the same size of NAME of SSOSITEThe service component has allowed to create a domain with the same name of sso site for the principalSo sso related portlets uses the url for the name in this implicit additionTherefore the column size of DOMAINNAME must have been the same size of NAME of SSOSITEAlso added the detailed message of the root exception cause on unexpected errors in the sso portlets. +Resolving final comments addressed assigning to Erlend as he did most of the work marking fixed in but most of this did go into +Target to +RalfWhat Martin said still stands +We fixed this particular problem for JSF composite component +Because the body can be encrypted +I will go ahead and make the ZK change only for now +Investigating +I just committed this. +Quick braindump would be cool to be able to create a Table object and then to print that to multiple formats +However +Thanks Alan for review +Sure +The objective here is to only cover the cleanup during startup? If so then this issue should be a subtask of as it covers cleaning up containers before shutdown and also during startup +OK I agree Ill put the +Lars G +Fixed for EAP EWP +Downgrading this to minor as theres a +I was able to access the wsdl without any issues +Verified in CR build. +Thanks Namit! +same for +i was able to get security working with cxf +The concern is not the failed to delete error we know why that happens +I dont seem to be the admin yet tho +Dave Meikles fix for said unit test is also applied +Verified on EAP CR. +It still did not work after commenting out javassist in serverproductiondeploy had to change javassist to cglib in serverproductiondeploy so that hibernate would use cglib as the byte code provider +Anyone know of anything +Max +Discussed on IRC wido will put the rule in post install script to create for both Ubuntu andFor the agent directory a symlink is created to the common directory +Fair enough +Alexey I see that behavior is expected I just wanted to note it here to maybe save someone in future some time on realizing what goes on +ER is not released yet I could verify this issue once ER will be out +As there are no negative comments regarding the overloading of a configuration loaded twice I proceed with commiting the patch +Ready to go +Hopefully thats it +Sorry for that +Rebased with trunkDuring my testing I discovered a NPE when converting from thrift avro migration thats fixed +Ill backport to once Uwe opens it up +Assigning to James for consideration +i think this is a regression victoralexey +looks good to me +Is this a problem? The namespace modelcommontypes for getEmployees and empno is correct so the bug that previously caused the namespace modelcommon to be used has been fixed +verified that it is ok now +We need some investigation +Accidentally upload wrong patch on creation of issue +In my understanding such bookies should not create a ledger descriptor upon an attempt to read from the ledger + will commit if tests you please rebase? Tx +I dont think this is a bug or dwinslow has already fixed it + +The recent code to support more elaborate error processing counted warnings and debug messages as errors +Are some some known issues with the ajstatus component +Why does extend reverting to semantics when this also employs the annotations +Thanks Nick +Committed a version change for the tranql wrappers in Sending Transmitting file dataCommitted revision Tested the creation of a mysql database pool the url prompt is gone now. +To learn more about that check the JDK API docsAnd no this isnt a bug at all +but rather maxOccurs +Indeed the correct default values for height and width of modalpanel were px and px respectively +cd home before you svn diff +What should happen is this javac checking runs +I have no problem launching Apache Geronimo and Tomcat +no patch exists yet +needs! +Sounds like more complicated task to me +This fixed now + +I had originally been going to wait for the Groovy release before doing the actual Gant think I need to extend the build system of Gant to better target specific releases of Groovy there should be an idiomatic way of doing this anyone any ideas +Moved to HADOOP since this is an issue in +Creates the directory into which the website is generated +Is replacing with will help? Please advice +To top things up nicely there is also an extra bytes added per RPC for the clientVersion that this is always set to zero because as a comment in the code says is exempted from version checkI dont understand whats the point of any of this stuff so I thought maybe someone could shed some light on the reasons why things are the way they are. +Doh I was looking for thanks Niall +Committed to + resolved issues for already released versions +I am OK for the new patch +and the whole unit tests are just bypassing attempt transition logic and just manually sending the app event to trigger the app transition I think this can be fineI put some comments for describing the test purposesbq +Whats the state of this issue +Hi MarcusI used RPM packed version you should be able to created package from +So there is no need to update any indexWith this patch I tested that after upgrade the authorization id for the system schemas get changed to that of the user invoking upgrade +Internally it would just ignore it +This is just routine +Thanks Greg! +I have added a KERBEROS logger and added the associated logs in the and +Archetype synchronization could be part of is working and part of +Now if a JMS client wants to force the relaxed selector evaluation it must prefix the selector with convertstringexpressions +Works great +Tag para +Previous patch is bad and came from an old checkout +Its even presenting as a concurrency issue it happens rarely and sporadically +Initial patch for review +Id always assumed it was just an opaque number so we dont get cyclic revision trees but there appears to be more to it than that +Cheng Even with this patch were still updating an each time we get a cache hit right? I had assumed that that was the slow part +committed to trunk +Fixed +in the environment the version is stated as which likely doesnt exist anymoreIn any case this seems fixed if it was ever a problem in a released version. +Patch applied thanks. +Thats why it was not catched by th suites +Beyond the console issues this raises the question of whether we should allow a user to remove its own role from an admin group +let me try it here +They are not unter control of the application and they are not only used in HQL of the attributes are used to implement capabilities some are used to map legacy data so there is no possibility for the application +Thanks for the fix Justin. +Isnt that supported in Identity server tooThanksPablo +Updated patches to handle unavailable versions for a given range +which is moving the multipleoutputs directory outside of the main output directory +Needs to be committed after committing patches of +Or maybe not The only alternative is for me to assume that the component is being resolved to null which of course makes no sense +This is more a feature though I dont think it is a bug +Sorry about this Ill try to solve the problems with these tests ASAP +As its not exactly our bug weve made a Future Request to modify behavior of some templates. +Please reopen if it exists on and above and if you wish the fix to be backported. +I had the same issue +it too late to correct the title on this one? Whats in this ticket was different from whats in the version and neither was quite right from my standpoint +I would have checked it in directly but there were some aspects that I didnt care for but couldnt seem to get around +I believe that Sachins commit in gets the same basic functionality +But it took a lot of time to identify the problemAfter day only i found that there is a port change that the ZK was using and the one in the error msgMay be i was not aware of this existing thing +Thanks +But I see the Metrics instance grow steadily +Add additional comments +Patch works +Sorry revision in revision +Ill provide different patches for and trunk +Im not a fan of using abbreviations for +I commented out the bad test and also added a warning if jlong isnt bytes long hope thats ok +Do the same fix on as well when implement this ticket +Bill I dont understand the need for the dummy +Joydeep + message TedCould you please attach the java class as well as service archive fileThanks file containing classes required to reproduce the issue +It does work in that the correct result is returned +Changed the kitting code such that connector jars are placed in the lib directory and no longer references these jars in the file +And nice addition to the wiki thanks I have closed this as wont fix as it is basically you as the end user that uses a custom object type on the bodyThe resolution is to provide your own conversion using the annotation tags etc. +patchBtw I couldnt apply your patch not sure why. +Ill add it to the review board tomorrow +Except for a couple extra long lines and some redundant white spaces at the end of the lines patch looks ve just committed this +Would it be possible for you to provide a svn diff of your changesThat would make applying them a much simpler task +It might be a good idea to be able to set some global properties in your IDE configuration too or some other embedded scenario. +Thansk +Committed to and trunk +Patch to handle passwordless usertokens and configurable usernames in +I just committed this +If you specify a cache location then it works the same +Closing the issue is already in the integration notes so nothing additional needs to be added for it +Still no luck can we try domain username +Actually I am using Hibernate Tools +I talked to Dhrubha and he advised that we just upload the tgz fileIn order to get this into the source treecopy to srccontriband thentar xvf to get all the sources into srccontribThis compiles and tests with svn revision for hadoop trunk +Is it not enough +The patch has been appliedPlease verify +As I said analysis made by Simon was incomplete everything is fine with packaging +I just tried it in R and it doesnt work there eitherThis has nothing to do with converter code changes +Updating summary to be more descriptive for the release notes. +Alright here is a new version of the patch that removes the layer of indirection that was in the previous patch +Denny wouldnt be ok for you +Correct typo also in branchesSending srcorgapachepivotwtkskinterraTransmitting file dataCommitted revision And corrected some missing imports in the version in trunkSending srcorgapachepivotwtkskinterraTransmitting file dataCommitted revision +Ran sort again with this patch and +Could we use the badness detector in dynamic switch +Patch assigns result of serialisation back to member var +it includes NOTICE checking and a bunch of other additions +Thanks +Could you make it clear +Unfortuneately I dont have a solution in mind +Thanks Joel +maybe a side effect ofCan you try again with fresh version? +If you agree then at this stage throw the exception +According to amc this should be closed +I cannot think of a good way to give a meaningful error messages without making update aware of OBR which doesnt seem right +Committed the to the. +is already provided via mapred configuration and admins will have two places to chose depending on the in use +Attaching the test which fails now but is fixed with +Since embedded runtime is delegating the correct class will be loaded each and every time +This feature is a modular addition +To apply the changes expand the file in the qpid directory +should no longer be an issue in the release plugin +The latest patch set against svn patch set agains svn latest patch after the huge formatting change +Looks good +applied script to both trunk and of unclosed issues. +Should I file another JIRA for it +Apparently I should submit another JIRA against +I see the naming is a bit confusing +stack it is pretty safte to commit +As we are in the middle of camel release +Is there a possibility someone could provide a tgz of patches that applies to the branch? Thanks much +access is not needed for that a user account can perform theseWere only able to provide that level of permissions to committers +if you pass an empty string to the wsdllocation flag of wsdljava it doesnt generate teh wsdlLocation attribute on the annotation +The good thing I suppose is that my change to a for dummy scopes has caught a number of additional bugs where we actually needed a real scope to be present +The API does not provide a method to change the heading +Ill look at this further +Ill get in touch during Wednesday +Tests are successful +The patch looks good +Thinking there is just one normal out we then think we can do more optimizations to bb +Thanks Gurkan updated + +state correction +If its really fading out for you though I can brighten up the saturation +i did make the changes to the superclass so you can move forward with your own implementation. +But if we have it Id say it should be in common +Hive has an which I believe can read schema on the fly like this +some new configuration or fix would be picked up when node manager restarts would that cause any issue during the statework recovering process +what a dumb a ImLet me fix that +depends on which is not loaded in the default profile +If needed this patch could be backported. +Updated patch to Fix exnraised assertion during finalization start upFix consist in save exception before start java method in ObjectsToFinalizerunfinalizers and restore it after java method +fixed. +Thanks for fix +At least either that or add more control if the use cases justify thatAttaching an updated diff which implements the flag as a port option +also added to the release notes as resolved The ACCESSPOINTURL column of the BINDINGTEMPLATE table in the jUDDI database has been extended from characters to to allow for very long +This has a high chance of breaking existing portable extensions +Checked in to set fixversion +Which is not possible without more writes +Verified tests and validated the feature in platformIs a patch required? Just need to change the parent pom to the new versionNote there are some new features that may be of use but the component works as is with +Could you please verify that it is addressed in your patch +Linking to which appears to have the same root cause +For what this proposal was trying to achieve it addresses the concerns +patch in revision Denis please check the patch was applied as thanks! +We are using the CVS source seems to be at so we havent done CVS diffThe one method changed in this class is an overloaded null and is tagged with author connorbto make it easy to find +This new issue would be about trying your suggestion and breaking out compaction so it standalonetestable +For example where before I could write box +Thanks for testing and reporting back +Will try to spin up more time for this issue in the near future get those tests written +Moved to as this issue relates to functions which were moved from Grails core to Quartz plugin. +Im not sure if the runtime classloading will like it +This changes the wording for the error condition + +Thanks Todd! +Since xlC has its own STL I suspect that this was indeed a problem with the libraryWe are going to deprecate support for xlC since its no longer supported by IBM +Attachment has been added with description patch transition resolvedfixed bugs to closedfixed +Removing patch flag since the patch can no longer be applied +Perhaps some captcha + +Can you check if it is ok for you too +I want to backport this to releases to facilitate things on Buildbot +applied +create aar file from the schema specified above and make sure your aar file contains original schema +This should work for both secure and unsecure +As a side note the groovy extract method does not respect these settings either +I committed the patch to trunk +Hi DavideIf I read the patch comments correctly once youve applied the patch and recompiled you need to add entries of the forminto an XML configuration file that specifies which field should be included and what boost they should have +Okay tested it with the latest M build and the problem seems to have been resolved +Looks like duplicate of this is a duplicate. +My bad +Thanks but you should be able to initiate download of new template you can give the same URL of the original template and it should use the proxy +Of course the whole apartment concept was introduced to eliminate locks +Yes +I have now run my version of the test to create large files and use methods getLength seek and get on some variety of vms and oss +Added unit test to demonstrate the bug +It generally fails in loop or for me +Linking this to since it would make sense for dealing with multiple async running tasks +If you find further issues please create a new jira and link it to this one. +Major refactorings are completed +It should work now +But to plug in the mocked implementation you have to explicitly set the reference after the object is constructedIt is cleaner to use the mocked implementation from the as you avoid executing a lot of code and that will make your unit tests more focused and +Compiling Suite with target would only work if the rest of the jdbc classes have been built +Dont you think this alone would be an issueAnyway if you think this is not an issue I can close itEnrico +The patch looks good to me +Hence the significance of this patch is only to the extent that it got working for me and no real impact anywhere else. +What would we use as a replacement? TheMike I apologize for not making myself clear enough +Thank you for the contribution Xuefu! I have committed this to trunk +Good point fixed for RC and now +eliminates the use of static fields for configuration in the clustering code in all cases where it was present canopy kmeans fuzzykmeans and meanshift +Thank Tom for reviewing it +Removed the spurious code fragment in rCreated and fixed also to port a bugfix from to CM. +and It appears to work on you suggestion it will be put through some rigorous testing now +Could you explain the use case more clearly please +This package thus has two files with duplicated lines +just committed this +Closing test adjusted regarding comment. +I followed up with Murali and this issue has been resolved +Did you build it from source +Thank you Darek +Committed to trunk +Hi MaciejYou currently need to run ant before running +Defer to defer open issues to defer to not essential to address for +Related it would be really nice if you could introduce different icons for qualifiers stereotypes etc +Once a package goes on RHN it cannot be removed from the channel thats how RHN worksSince these subpackages Requires the same version of we need to continue to ship these otherwise if a customer has already installed these they wont be able to upgrade justAlso modcluster will be part of EWS the tomcat subpackage cannot be avoided +Seems like a really beneficial improvement +It does not stop on not found so this is secureIll commit soon and merge flex +a go at the latest svn and heres the problems +Im not sure about change how this validator works because any change in behavior must be done at spec level +Attached class which can dump http session contents into a file in the form of object tree +They must be trying to pull lots of stuff from Wikipedia or something +for the patch +This provider would also do the corresponding type parsingThis work will simplify the implementation of delegation tokens support extra warnings are all because of the reason explained in the prev comment +Any chance to see it applied to the nightly builds +tests javadoc findbugs javac warnings were run manually +With applied to the current head of trunk Iget a failure in the upgrade testI think that maybe the mode dbsql is still required when runningin the portion of the upgrade test or something like that +I think clearing the values would generally be useful but I think copying them follows the principle of least surprise +Thanks for the pointers! +Updated example +Committed patch in revision +Patch v combines my patch with Ramkrishnas fix +in +Also I added a new test for strings that end in this was a bug that is now fixedThe patchfile does not include the extra makefile stuff +upgraded to use woodstox andRevision Im checking other +You really have something that should work here +Eli this was contained in. +Patch is in +But I went through some code design and felt that the data structures and manipulations were quite complicated +FAST had that capabilityHow does this impl compare with the Tika one for short texts? And wouldnt it make more sense to add this on the Tika level letting the detection method be configurable? Then all Tika users would benefit from it +The more I think about it the more I think that the character in the passed to the should be treated as a wildcard +Maybe we could start with switching trunk to default to throwing exceptions +I fixed it the hard way The dependency to is no longer needed. +Closed upon release of Hadoop. +However it might be advisable to wait until at least all development are completed +So closing this defect as duplicate of the above +This means that there different versions of floating around and instanceof tests can fail +we need more input to investigate +We should just be testing it as part of security basic unit tests +Attached the complete sample application +Patches are very welcome +can take this one if required. +This is part of simple theme design +I havent verified it but I think the patch for wont fix this because it only avoids opening an SSTR after closing the writer +SM code is refactored and now this issue no longer exists. +This will work for containers that support pure JAAS login for ejbs like you only specify the and youll have to specify an initial context factory that accepts and like for example +I rebased this on top of the current approach + +Patch committed to trunk. +You mean after changing the source? No I didnt do it explicitly +jar where does it come from +I neglected to mention a probably important detail Im running an embedded Cassandra instance using and do my writes via the API to avoid thrift serialization. +Ill shift the ticket to +The attached design doc is about having Tags in the KV byte arrayWe had tried both the options and the attached patch is with inmemory Tags rather than appending to the KV byte array +Patch for trunk attached +Revision . +Jasypt bundle created and deployed on the maven snapshot repository +Assuming you actually meant to include the link to Hama and not just the incubator I applied the patch. +This patch fixes the warning found by patch incorporates changes suggested by Chris Douglas +Let me know if you disagree and we can reopen. +Absolutely +Trivial patch that does just that +Thanks Scott +The G isntSanjay I was mentioning that we already have a sequentially increasing persistent id that we store on disk +Thanks TimPatch applied as expected +As discussed in there were issues of old versions of the same scripts popping up +Im going to close this issue +This is only needed for the final query API in Infinispan which is targeted for Infinispan +Now that I got clovercoberura and Maven to work I can better pinpoint missing holes in our test suite +Running a little test the getFulls all return in a fairly constant couple of milliseconds +It should work as described +Your matrix is simply a proof that Maven works fine +Can you check the patch? It doesnt apply cleanly but neither file in question has changed since you posted the patch +Ive applied the patch and tested it outThis patch toggles the disabled state for both the SSH username and the JDK path when clicking on the checkbox for JDK path +For simpler this may suffice which I am assuming is the goal here +Moving right now sorry I didnt catch this sooner +Nevertheless maybe they can give us some idea on how to avoid and bypass it +Thanks +Thanks David +Patch for an earlier version of Hadoop +Add more tests to cover all cases +Most of these display issues are resolved however the site is designed for a fixed size font there is nothing we can really do about this +Hi Jeff Thanks for the quick response +The attachment should fix both problems +So this just is a compatibilty issue with? +By the way Do we have a TS note for +docs +This plugin should probably be merged with the or optinally be named +Since these will need to be multiplexed at a lower level why not keep the code path clean and obvious? again why? If were going to do that we can move most of the methods like that +bulk defer of issues to +I added the docs but the patch was worthless +In my opinion retry limit can be delegated to hdfs layer +Here is the class with the problem would find a Long and a more appropriate if an Integer does not have enough bits +I think you are correct the message should be more user friendly +lol nps +Whats the status of this +This fix is wrong +Any reason why this cant be put into the regular Cargo svn +Patch applied at +We can table this as low priority as the highest number of partitions we have seen for a dataset so far is just +Otherwise every client operation will get the warning msg if the NN is active. +Ah wait I just read what Olga wrote again +I attached a sample project +Applied +btw nice work DanielI believe we need to fix the stack walk but we can postpone this issue by just doing the manual inline +Also Becky I should ask why dont we like to extend the Camera API +The persons interested can do that themselves +Simple plain doclit wsdl without external xsd +Adriangot it thanksAs a side note Ive noticed that the font of the applications screen is slightly different from the one of the other applications my guess is that there are some conflicts with the special css stylesheet that is loaded by the screens to render the calendar pagesJacopo +in keyup event the javascript can dectect whether the text is actually changed if its not then no need to notify the server +I will apply all needed fixes to trunk as soon as I have some spare time +fixed as of r and tested on postgres and +only issue outstanding atm is that wsdls that include and xsd and are downloaded and processed locally blow up since it expects the xsd to be local as ll spend a bit more time on it soon just and workflow appears to be broken on this jira ticket ie +I just committed this +Once the CC test run happens can get final confirmation. +Thoughts +It should have been iterating through all of the delta children regardless of how many children there are +trivial to master. +It makes an assumption that all post requests are form posts +mouseover dragoverIn the case of over events the toggling is unusable because each movement of the mouse causes toggling the tree +Updated patch for trunk with fixed +FYI Xalan and libxslt do now agree on the output from this test +It should be added to the next release if there is any. +I propose to clean up remove this attempt at admin portlet with JSF +Estimated histograms for row size and column counts and replay positions will also be available via this class +make still fails trying to build the ruby lib +Now moved in ve found a few other places that has the int cast problem so Ill apply them all at once +Ill wait a couple more weeks at least before trying out the new branch code +Fixed in compiler api +Because of compile errors in java files and Object mapping as id test couldnt find a getter for class good for remaining fail see +Yes +Oracle Database g Enterprise Edition Release bit Production With the Partitioning option +For me the only useful default is the path that was chosen in the SCM +modified renderIcon so that if the has the icon field set that icon is displayed instead of the default icon +They are independednt on form key +Closed upon release of Hadoop. +Proposed fix to enable POST requests of WFS file to fix issue on branch +thanks +Also is required to create the uber jars if this is not possible please can you raise a separate issue and we can keep this one as the findbugs issue which I can resolve assuming the tests pass +Instead of having multiple rules in one asset the feature allows you to view multiple asset editors in one can save one at time or all of them at once. +And this is specific to the file being in a Guvnor repository and does not happen when just double clicking on that file present on the filesystem +Took out the and delayed the argument setting to later +Massive cosmetic patches will break pretty much all existing patches unless those patches are applied with l or Match patterns loosely in case tabs or spaces have been munged in your files +Marking the bug verified is a good way of signalling that its no longer an issue so thats what Ive done +Thanks for the patch! I will take a look into this issue +Any other opinions +Rebased and merged into the master branch +Leaving this open until it can be into the branch. +Logged In YES useridEverything seems to be working properly now as far as my tests! +In future if we enable append by default again we need to fix this Test Code +Fixed in I suspect there is something fishy about the patch since it breaks our unit tests +But if we start the bundle those unwanted blueprint service is started +We are seeing this also and are using is in fact a problem with interpolation and the issue is already raised in the resources plugin +Thanks Anil your patch is in with rev +One thing one could do is to let Identity implement Iterable which would make the code more elegant +Please verify +Happy path typical app use +Just one comment you can use the session timeout from it will give you the value that is being used at runtime +For the first pass Im going to add a crowd group and just cahck against this group and the employee group when adding a new member +rebased on latest import is broken again can you please update? Tx +But we could do with more information +I removed the hacky part to make only MTQ single pass for now +is this still relevant? The last error seems +This purports to improve the style yet it does not conform to Hadoops stated style guidelines +The second column is the number of tasks that run on each of those nodes +We have AIX support now +Ive updated it to include the latest target that automatically copies release artifacts to the releases directory in svn +I made those corrections to the pages for Second and Minute and to confirm +As you can see though the test failures are just the absence of quote can put out a Gant if that would be helpful. +I believe you are mistaken. +I will try to configure it better +Pls review +Hi RalphI understand you +QPID for +I know its kind of clumsy but I dont know of any better way to do this +Attached patch implements the feature detailed in the description + +Component should extend +The script can be compiled the window opened the script changed to be illegal and the refresh button clicked +With older version of Castor a Java type has been used that didnt support the value range as mandated by the XML schema specification +ok will make SNAPSHOT version force SNAPSHOT to commit +Right click and select Show advanced options +Yes I was referring to the opera problemI saw your email and did a diff on the codeWhen I looked at the diff the behaviour of the tag seems to have changedHowever when I just ran the example code it all looked goodSorry for the false alarm! the example app works but mine doesnt +Not much +and test failures are definitely unrelated +You could cache it on the Similarity instance +I also commented on that abovementioned handlers should be rendered in another way +Thanks +Resolved in you mind trying out the latest? This is also scheduled for backporting +Close as this was deployed in June per Gregs comment +Patch applied +If you have a chance to make a patch ill be happy to review and commit +Logged In YES useridI am no no longer a committer so I am unable to fixany bugs +Did you edit the file by hand? Also the gpd version and eclipse version are probably not the ones intended current gpd is build and tested against Eclipse +You upgraded the Spring version to recently and no longer exists in so you should remove that dependency +See attached picture +As mentioned I havent run into this issue since reporting it so I dont need a workaround +By this way we can add new features quickly +Alexander sounds like we are on the same wavelength +fixes all balancer related testsRunning Tests run Failures Errors Skipped Time elapsed sec in Running Tests run Failures Errors Skipped Time elapsed sec in Running Tests run Failures Errors Skipped Time elapsed sec in Running Tests run Failures Errors Skipped Time elapsed sec in Running Tests run Failures Errors Skipped Time elapsed sec in Running Tests run Failures Errors Skipped Time elapsed sec in +The patch looks good and fixes quite some test are two important failures left testcancreatesocketserverandacceptnonblocking just hangstestunixsocketpeeraddr reports wrong errnoHmm it seems we have a very bad situation with errnos since they have different values on different example ENOTCONN on is but is on our Errno values are hardcoded to be in sync with which is problematic for all other platforms +My guess and this is just a guess is that maybe is doing something to the document after it reports domready and that is messing with jQMs initialization +reopened just to set the resolution to with resolution fixed +The became and the current was missing this. +The problem is that it in my case sends emails every minute even though there is no status happends in my case is this +Addingskipping mapper test is your wish +Were you able to verify this on your end +We can set this to a more reasonable value than but production users are likely to want to modify the DDL from the DDL generator utility to tune the size of this column for their use caseOracle SQL Server Sybase Informix and Teradata do not have this restriction +Looks good to me +Not seeing the issue in the latest GA build +Power Vodka all the way +I think the plan is that SDG will upgrade to Spring btw +I just checked the code the is the member object of outputStream that is the key of NPE +Yeah we will have to move the query parsing before this is doneIll move the code out to a method and keep it commented out for a while +This patch provides a proof of concept for integration of spatial search in the solr handlerPlease note that tests and documentation are omitted for readabilitys sake +When is release this will become clear +Sadly I agree with the assessment by Hiroshi +Thank you Glen +might be able to help here +other to comment about itWhat do the URI spec mean in such case? Not that we are URI spec compliant but we shouldnt implement new stuff which is totally against it +Is there a JIRA for your task? If not you can use this one else I will resolve itThanksUpdate to exclude lz header and source files in audit checkingUpdate copyright notice in lz to patch lgtm +Please let me know if you can still see them +add validation for this and errorwarning if JAVA CLASS andor JAVA METHOD are nullThats where there is some grey area in design +Thanks StuartPhillip could you please attach the application which reproduces this issue +Sorry for the noise +I prefer to isolate the change in and dont touchI will comment their as well +Consider upgrading to +Delegation token was added in +The attached class is the one I implemented When I have to handle System exception during ideas suggestion +Feel free to close the issue. +I guess you could do it if you used a vertical text field but with a font that rendered each character lying on its side +GertI tried your suggestions +It seems addReference is called for each application schema to includeimport +This has moved to the devel mailing list for anyone whos interested +Thanks Uma and Thomas for looking into the patchI have updated patch +done. +Reopened +Is there any chance of additional beta releases coming out prior to with this fix in it? +Can you please attach your attachment archive in a standard recognised format like +With the latest update to fix the printing of double and float values this section of code is complete. +You are right that the PWC need not be filed and thats actually not the problematic partThe issue is that the Workbench doesnt let you access any objects +More patch for older version of hadoop + +I agree that these calls should probably go through static methods via rather than + +Currently leave handling as it wasjust adding WARN log when cl equals null. +Brian why would need a int property and a string propertyJBM just needs an int are these other uses +Do you see any incompatibility issues with older clients +One more thing should we make the handler threads a percentage out of the total handler count +Im finally getting around to looking at this and I cant seem to find whats being reported +Those hierarchical fields were removed from Roller +Thanks Rahul +I am just wondering since you put as a fix version +bulk close of resolved issues for release. +Not fully there yet +We have changed all of the headers for the subprojects that are going to be included in the release so I will close this bug +I have been trying to find the source file for to adjust it but I can only seem to find +This has changed to being CDI for JBDS +This may mean that a Motorola update will fix it or it may mean its an issue that is confined to your deviceCan you provide a more generic test case perhaps one that works on the Motorola Emulation Suite +Im not going to make that argument +Larry if you have an opportunity to verify this in your environment that would be great +We dont really have a proper mechanism for that in place right now +If folks agree I will make that change in the next patch +Okay I have this working now +Thanks Billow +issues +minus ownergroup +Oleg I applied the patch in +to Ted to finish XML file creation and making sure Security works +Im ok with your suggested fix though it isUnfortunately I dont think there is a standard JAXP property that does thesame thing +Please dont bump here +Seems ok at least I can not reproduce any longer. +But is there a way to avoid broken of the route and create an exception so with a retry i can accomplish my task without broken all my services +The administrator can be notified of the change through a zookeeper watcher +Removing target release from issues that fit any of the following criteria No known way to fix them Java integration enhancements out of scope for release Other out of scope issues for for amorphous bugs +cannot reproduce with ASF CS master environment VPC dropdown list is already hidden from the cannot reproduce the dialog works correctly where the VPC is hidden when a network offering is selected and visible when it is. +It seems to me that a simple replication of configuration is way too simplistic and Im not sure to understand the use of the topic in that designIf you want simple replication why not simply using a with a shared database underneath +I will incorporate your comments and upload a new patch +Changelog Each method derived from now has a instance The instance associated with an executed method contains allthe parameters that were in effect at the time of methods executionPlease let me know what you think OlegI think the idea is correct but Im not terribly happy about copying the params from one host config to another in copyDefaults +Other minor cleanup +This patch let the user specify its own +Fixed in revision . +GarimaWhen you say Now I hit submit on the +After a user successfully completes the new user page redirect them to their weblog page or to their edit weblog page. +I have confirmed that this is still not fixed. +done +Need an event model to respond to changes in not have a hard requirement for an eventmodel should just have a button to sync the settinsworkaround is possible right now manually by typing in domain but the other parts like deploy project are also not fixable manually +The reason why compression is not supported is because is meant to be content agnostic by the project charter +at first I did not undertand exactly why you recommended this +The issue will be falling back when the priority url comes alive again +I have made the changes to remove the compilation warnings +If so did anything come of it +Both trunk and patches look good to me +Verified by Mark. +Ill look into it in a few +Not a bug. +I am sorry this affects you +SteveI applied the part solely dealing with encosng for +Hi ChrisThanks for finding that precedent +Let me put up a patch +Committed in r +Bulk defer all frontend issues to +I just am not sure quite how that conversation ended +Patch looks good to me +Attached the patch +The new testcase does removes and moves for cases with three SNS child nodes and with three child nodes +bq +AXISPRINTFFORMATUINTRegards update the templates to use these constants +Closing this issue. +If its windows and its a local filesystem URI do the subbingYou cant always tell whether its a local filesystem URI +I hope its not doing it for mevenide projects as also triggers refresh when the application gets focus after working with other apps +Oops +JIRA triage not a blocker for GA +However this is a very minor break in bw compat drop in jar replacement +It seems you want to have a central configuration point and all compilers should behave like that +CheersJohn +most of codes are already there branch i will start trying merging it to master resolve conflicts providing test scenarios +Fixed in for now but removing the newly added test for hbase in the dependency jars +Sorry its a mistake +When I add to Gemfile it runs fine +I am working on this right now and got some moving parts working where we can trace to some destination of choice for further processing +Thx +Verified at +Does that mean that we can close this +Attached test GVN and test from DRLVM Validation Test suite pass now +This problem could now be considered resolved +I couldnt recommend a system launching multiple jobs using those as the only safeguards or with this proposal +It looks like initialize bails out without initializing the p field that triggers this bug whenever theres no arguments passed +This work was completed. +Could you include a test case that actually has some assertions? Or otherwise add a little bit more detail +Ah the binsetenv Was directly run with binjames without running it first +jMock and are now by default and only in +So for me this seems to be no bug +Can you please provide an example configuration or some screenshot +I would commit it +Thanks SianPatch applied to LUNI module at repo revision rPlease check that it was applied as you expected +If it is there should not be any other combinable in the s inner plan +It depends on the implementation of the callback interface +Moving to since this will likely require some kind of dedicated service we deploy to intercept requests and pass them to the appropriate ruby runtime +Please let me know if you think we should adjust itAlex +I also need to write up a little page on the wiki explaining how it works and how to add new projects that can also support targeting. +were not fixed in +With each time the quality profile is changed a new event is created on the project. +I agree we can handle that in a separate JIRA +changed was pushed to master in commit cefcbebbafeddaad +Can I help out somehow +if nobody objects i will commit it by end of this week +Committed trivial patch to trunk +Closing resolved issues. +Nevertheless the paths is reporting do not me this can be closed as everyhing works as expected if one studies documentation accurately +license issues now all issues that have been in resolved state for more than one month without further comments to closed status +If you are going to fix this then lets fix this for three years Synapse has lived without this +I tested this manually by running with both and AIX clients and confirming that it works as expected both with and without Kerberos enabled +is my proposed fix for this issue + +That stuff isnt really conducive to js manipulation anyway +This is not longer relevant because MRV is deprecated +commited the patch in reversion +With this change ran successfully in my environment +Yes I think we should get rid of CR and CR tests and add a new one for +Updated in ESB codebase will be in next merge +Closing this as duplicateAppreciate help validating the fix though. +FYI since this code was placed directly in it wasnt detected by during compatibility check. +If we got rid of delete family perhaps we could skip more +Does that make sense +Hi Eoghan Adding the config for now and well see what we get +ear would be now have a war builld too +I dont have a windows machine to test so I rely on others to help in making sure unit tests work on windows +It gets blocked at as well +Thanks Libor +Lucene badly needs good ootb spatial searchI think a branch makes complete sense +Closing this one +I am not sure about other branches +Separator display in IE broke down +Set to fix version as we absolutely need to resolve this issue cleanly before a release +When I say pretty sure +After talking to the JAXB guys it seems this is a JAXB bug +Can you clarify what you mean that should no longer be necessary? You mean that its not necessary since we already have applied to branch? I agree with that +Problem solved in newest build +Siegfried thanks for offering your will have a look at it. +Do we have a testcase for this issue? or can we provide more specific information on which feature we not implemented in tools? also should include the specific soap version like is it for orAs i remember we should support soap header and soap fault for both and so im kind of confusing about this enty is it a but feature enhancement? or something elseand currently we are under refactoring so i guess this fix should go into the new tools directly +Oh I wasnt expecting that answerI think Plugman is set up to honour proxies using the npm proxy config +Im trying to develop a simple webapp which renders huge data which should be able to demonstrate the memory leak in +The debug info of the maven shows that groupId as tomcat instead of +Closing this record since is now released. +The patch is a simple build change and the failures doesnt look related +This test just hangs on OS fixed +Not able to reproduce it +On doc in general I came across this recent quote Documentation is like sex when it is good it is very very good and when it is bad it is better than nothing +Thanks. +Just had a chance to look at the patch +Patch is for patch works for TRUNK also +I think we also need to add a link to the current audio podcast as well +Right now I use a to accomplish the rotated labels but the problems with that approach is that the decorator doesnt affect the labels preferred size so I the height of a table pane row to accommodate the rotated labels +In Data Base semantics NULL is a special value so for example NULL does not match NULL +Im putting this in the next without a test will write test later for was commited as part of r as well as r +Only +Ive been there too many times with web service protocols whose adoption of XML was meant to handle versioning well but turned out not toId be worried about running multiple datanode versions in a single cluster for example +Let me know if you have any platform are you building on that doesnt have INTMAX and +So this is done now right? uses now +Currently Whirr does not support Eucalyptus based clouds +It looks like the issue is caused by the reconnection logic being removed from the admin proxy +Please reopen if any disagreement +Screen shots of groovy shell script running a Runnable and the other shows Eclipse failing to should be possible by altering the way that searches for scripts. +Pls review +A few nits which I will cleanup on checkin package should be to be consistent with other lz source unused imports after the latest Vadim! I committed this to trunk and. +I will submit a patch for removing it from HDFS +The test has no problems in my setup +Made some further changes for the patch + +It looks like the compression thing does not in the code require a compile time dependency on the LGPL library +small maven project with does running the Unit Test via Maven from the yieldFirst impression is that the defining the JNP Server is not getting deployed +I can confirm that the problem still existsWAS Oracle iHibernate First time the session got queried we got the unsupported exception +This is a part of hflushHow does the patch look to you +I should say proxying class in Spring is possible +Yes Gregory you are right +If we change the design of users mail boxes so that a mail box is only accessed via a User then the User could have two attributes mail box capacity and mail box size. +Turns out the previous thread model was too simplistic +FYI here are the top screenresolutions for users running yesterdayAs you can see large majority of users has a resolution lower than x +Keys values of the map are wrapper java objects if we can use better data structures that will be further winI am just putting up as thoughts which came to my mind in mins perusal of that class +testAutomaticallyentity class +Postgresql supports IIRC lowercase and mixed case quoted identifiers +A delete follow by a create can confuse a reader +ISTM that we could get of the benefit here wo too much complexity by decreeing that column value sizes go from +New d patch that also has the changes to +On Eclipse menu presentsthumbnail +Sorry please assign this Jira to yourself and continue +Im pretty sure we DO want to forbid counter mutations in batches entirely +The is for branch. +Java lists are not. +Otherwise WSSJ is fineColm +Raising priority to blocker on branch +I tried to implement a Validators for java bean validation w +this does not seem like a minor issue to me +Link all related jiras in this umbrella jira +added thanks +Nice! Thanks for adding nice debug logging too +Closed after no further comment has been given. +params from the collections api call to the solrcore api subcalls +has this problem +nrm file should not even existSo we need to figure out who is truncating these files +Uh really! Damn +First I will not consider this as a blocker +Passes close resolved issues +Pointed to another jdk and everything worked +Might be caused by the same reason +Verified in JBDS B +I can probably ask if this project starts getting underway +Wrote it and used it at home for a month and there havent been any issues +Its very similar to distributing Java software that needs a JDK to run +looks good +Two litter questions why declaration is different fromThat is to say why removes shell injectionif these commands can be executed outside karaf or env +Duplicate of + +applies atop and +And also for events and timersBut now I think why Ive never seen for sequence flows +Instead Master just uses the the RS volunteeredSo what happens if region server passes +Thank you +Patch for the you Boris +This is the reason why this check is in WSSJ +Replacing on the rule names for in the generated code as a way to work around JANINO compilation failures +The decision will be delayed to a future release. +This is being tracked in +This was caused by an incorrect ordering of generated parameters in Designer +I added it on purpose so we know to convert these places when we remove LUCENECURRENT +Patch looks good +Point of clarification why did you choose to use the context based configuration over the value that is explicitly set? If you arent really expecting them both to be set then I can understand that it doesnt matter but if there is a case where they both could be set then Id expect the set value to take precedence +Please if seen again +Yes you are welcome +So the mock implementation is moved out from test component +The extension point must be used to decorate the section test added +Hi Todd I worked on this as part of Balancer issue +Unfortunately porting the groovy shell into Eclipse just doesnt work right +This issue is really closed there arent any new Java features that will be added to and even if there were it would make more sense to track them in separate issues. +I need to add Im not sure about this and havent tested the behaviour +yep +Re opening to review for AS as demand from at least one user is identified +Thanks Chris +I was approached by one of the submitters to look into this issue and sanity test the solution on all platformsId like to have some more contextual information on this prior to having this submitted into source control +did you measured CPU time used during md vs murmur tests? Not only wall clock time +This problem was caused by the submit template is not. +Could also bump the ceiling via set +One use case would be when moving a service to a different machine first deploy on the new machine and then gracefully undeploy on the old machine so that no work is lost +This appears to be related to two problems race conditions around refcount private caches are never updated to reflect their true size +Do we need to add every field to uid instead of just field + on Horias suggestion +Thats not possible without a major refactoring of sonar core +JasonIf this the issue of moving all reporting plugins configuration into the Site Plugin? Is this a requirement for Maven ? Would it be possible to make this work for Maven as well +How does commondeploy work? Do files in there get deployed for all config sets? That directory doesnt even exist in my AS build +However the default class loading is delegating +Project created. +if so + +The problem there is that you have to make a call on each individual node to get the Stat object +I agree +Waiting on test and reference documentation +test +That was my conclusion also but isnt that just absolutely foul +For some reason this issue ended up being priority major it is really a trivial issue +Ok I think I got it now and will try it out the include filter files feature should serve for my needs but anyway I would need the prios exported +A client side javascript API would be required for this. +I dont think any Facet module needs to be concerned with SchemasRight they should be field type aware +not seen anything since +We need a deployable example and the actions necessary to demonstrate the problem +My configurations tested on Windows Vista on Windows Vista on a LinuxFor these configurations blame output is split on tabulator characterI test on projects +talked to Pete is completely in the SEAM codebase +Is this fixed +I intend to make this happen before GA +bndlib navigates recursively in the classpath of each module +Thanks SaschaYour patch is in trunk at +Can someone restart it again +use a servlet mapping to map the path to replicationI guesswe should let people to have both the the options bq +patch is updated and headers are moved to vmcoreinclude together with stack iteration implementation moved with LIL +Passes local tests. +Frank +taking care of unset as well +DavidNo I dont think so +The issue is resolved. +add another test which invokes the NN checkpointing code latest patch looks almost perfect to me and good test cases ColinOne small nit theres no need for enabling the DT config in the new test case in +Here is the patch for fixing the problem +Thanks! +Brijesh please change your rsync to as requested aboveI have changed the rsync to +thanks for the report +Heres the pull request that addresses the issue +If not maybe this can get into CR +Ill raise up my question in the redback dev list +Running patch through hudson +This patch needs a little bit more work before its really applicable +Ill update it accordingly +I didnt have time to investigate as soon as I found this issue so I just shoved it in JIRA to keep track of it +Test case added along with a all tickets +Unfortunately it has been committed as is +Do you think that can happen +I also manually tested this on a secure cluster +Username ehillman +The important aspect is release versioning +It seems that John has the same pb but filed the jira a minute before me +So this problem is not very important for I created this issue It was for your information +This has been added to Hibernate a while ago you are welcome to backport it if you need it in Hibernate. +Closing old issues +But in the paper they boil this down to nice little i param tables extracting some sort of pattern from that process +bq +v patch addressing last comment from Luke +Fixed +Significantly reduced the review board reviews +Ill do this for the branches as well +JDBC defines SOUNDEX as an escape function so I think this is a reasonable feature request +Committed to branch and trunk + +Good candidate for merging to branch +I checked the doc of DB Sybase Oracle PGSQL and MSSQL and none of them mention the fact that it is not supported so the fix should only be applied in case of +This patch fixes some additionally bugIve verified mvn clean install +But I could concede that its the worst when you have lines and that we could do a bit bigger with the polygons +And the test in the second patch seems to be referencing a diferctory where is thisAnd please add license headers to the new files +This is actually against on and trunk +In fact it looks like there are many methods in the replications tests Utils class that could be replaced with methods in +Sorry I did not mean to Im facing a very similar problemBest regardsMadhan +Tried and experiencing the same issue +Im actually seeing it with a newempty repo +The patch applied + +Clearly this is not at all what plugin developers would want to the patches +Deleting the +WPS does not it uses special fake insteadIf someone sponsored such development we cannot remove it but proxy base url handling is not working for Michael +The end is having a project for experimenting with CDI JPA JSF Bean Validation and Arquillian +Hum +This feature requires validation in real clusterJonathanAre you able to help Li in this regardFrom my experience in the past three weeks development involves coding running test suite discovering defect through failed unit tests bug fixing validation through ycsb +If a struct has a required field of type enum and a value goes away again it will be treated as an unset field +Could you be more specific about theproblem you found? +This has been working for me but I only just realized its because Im using rather than +But I really dont feel strongly one way or another +Id need some more time to write the +Ill also look into being able to build a binary release of the files +Im not sure if this is still relevant +This issue can be if we find problems with these localizationsor bugs can be filed. +Let me find the exact test case that is causing problems +We automatically add about default filesets for things like and other javaee api jars but we always get new requests by a user +your method needs a throws or a try catchcurrently breaks build +Im starting now the process to get it into the code base +In my case I just used one rule per with a in the RHS to show that the gets executed +actually we dont have to do it transitively since the plugin itself does not do that am also not involved much with this plugin except for some simple projects that i currently up the include files is not hard s can do that with hard part is getting the mojo to pull that down may be have checked the initial code to deploy +A new patch that does not screw up the formatting and that is up to date with have been working on this exact same problem Have you created an tests for itI am attaching my version +Cannot reproduce this using trunk. +Fixed thanks +Thanks a lot for the review and the suggestion Todd +Make the move from the end to the beginning more smooth +will try to reproduce and see if a more accurate error message can be produced +Some better Filter API is required which lets us know of the context +on by Tom. +The combination of patch looks goodI have try it and it works for meThanks to Timo Filippo +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This patch is not enough also testFailed and testError needs to be modified +Lars the dependency on is more to make the code simpler +I installed the settings following the instructions but still have to do some formatting manually like indenting throws statements at spaces instead of +BTW I just fixed the issue with API context getting appended to the RESTURLPOSTFIX +The Natures contain basic JDO attributes like table and column name and a few more +So we may do an with a release +Uploading a new repro showing another incarnation of this problem this time with a UNION query which relies on the +My setup using for Logj and a beta version all work fine from most of my webapps but my Appfuse spring based application seems to have this problem using exactly the same activemq and logj versions and jar files + +Clearing proposed Target Release value from to affects product branch +Create a Sales order +Im not sure just voicing a hunch here +I reverted the current for committed the change sorry I missed it +docx format +Im looking forward to commit this to the trunk as soon as possible +I dislike this idea +Also validate theattached patch against the CVS HEAD +thanks Simon I revision +Closing the issue now that it has been resolved. +Thanks Chuan +Changing this is doomed to be questioned I close this issue now +Our scenerion is that we have hod installed on a set of machines from which the cluster can be accessed +Updated the patch to throw when the ledger doesnt exist +An example of how the Offer Identifier from could be saved for later use +Since this issue associated with Active Directory here is a new issue about creation of users was resolved so behaviour in case of Active Directory has changed as followingLets assume that we already have user tester and now user tries to authenticate using Active Directory and username Tester which equal to tester according to Active Directory Sonar createUser property set to false we will see message Authentication failed Sonar createUser property set to true exception Validation failed as was reported in this issue Sonar createUser property set to false message Authentication failed Sonar createUser property set to true new account created because Sonar treats tester and Tester as a different usersSo we decided to add new option which will perform downcase of username during authentication and creation so Active Directory user will be always treated as same user with downcase characters +This is part of WS so and dependency on Java are ok right? If so lets just use that +So basically I am still in agreement with your former self on this + has been marked as a duplicate of this bug +Thanx for the commit Rick +Is that insufficient control for your needs +The bin package will be revamped or replaced +He went the extra mile that I didnt think was necessary +Sorry for the delay +The timeout behavior was intended to be since the usage counts went to zero so I think were generally talking about the same idea +the extra logrethrow is just there because sun messed up on malformedobjectnameexceptin they dont include the name that was malformed which makes debugging hardSo I log then rethrow the calling code will actually log the exception and then handle in some way this log is just for context when reviewing the logI could have wrapped the exception with another but I didnt want to do that for this patch +However neither in the next release most probably +I have android command on my root account +Will see if I can give some more useful information +Checked in new Books Parts project sets +Contained in s +Test it still pending +There was a temporary name resolution issue on the server but that has now been solved and is now up to date +Koen arent these tests supposed to be run from mvn so they get executed during buildsif not what are they +Elements td have width less than defined in elements cols +Someday I would like a script to do a set of builds and run a test to see what change introduced a bug +My guess is that it might be the svn tool you are using since none of those names should be breaking any Windows character limit +Added unit testsFixed all existing unit testsAdded some javadoc for the new methods +updated Fix version +Modified the logic to check the L for the embeddable oid +Did not notice the issue was already discussed +I didnt want it to get forgotten this close to the +Clean up the warnings in and fix the broken test cases in it and +dav is still at level as such leaving request open +based up some investigation and analysis of internal testing Myrna found that svn revision for is the causeI have a fix backing out some of the changes for that fix but I want to investigate more as the fix really indicates that the node is doing more that its name suggests +Ill need to write a couple of user types for these but it should not be considered as cross JDBC DriverDB since Blobclob are messy esp in Oracle. +Will commit shortly +I would also add can Johan publish the benchmark being used? Perhaps that sort of thing should be in source control so we can have it to test things later +Must improve the patch to be data so there is still a bit work left +Good thanks +Either way I m sure the Boo team will keep up the exemplary work +However the downside is that remotetea has a LGPL license that means it is incompatible with the Apache license +However the code section emanating the exception is same +I would like to look at it but I cant at the moment +For the other validators it means that disabled fields are not validatedSo your statement not supposed to validate fields that are set as disabled is true for all the validators except required +I think this deserves a reopen the default behavior should be building without bundling additional jar files +Committed revision +Providing the wicket artifact but not providing the sources needlessly imposes a penalty on Maven users who choose to reference that artifact from their files their will be unable to find the sources +Thanks Dag that helps a lot +Its now able to deal with this tested several different ciphers and it seems to work fine with them all +That seems pretty significant +Attaching latest patch v which addresses review comments and contains more contains changes to and modules +It would be better at lease investigate it during codebase to be sure our API can easily support different impementation than current reflection based one +other opinions +Added new question How to ExpandCollapse Tree Nodes from code + +See implemented validation rules on the screenshot +It works also remotelyIm not able to reproduce I think that the issue comes from the custom distribution +Thanks Arun! +one of the query methods could be used to force a list of Serializables to be returned instead of using the so it all seems superfluous to me unless youre allowing whole objects to be returned +So i am gonna close this issuethanksdims +I do not see why not just do not force a fluent API on users IMO +The result is the is saved independently of the main GORM transaction and you cant accidentally invalidate the GORM is checked in to SVN overwriting the previous file +parameters get appended to export links as well as sorting links so we need more details if you still have a problem +Patch for coresite +Push to Westport per RFE meeting on Teiid Data Source importer provides a flexible way to import new sources +products can use any version. +The behavior on Windows on Cygwin and without Cygwin could be very differentI dont think Cygwin has much to do with it +As Jesse points out this is not a legitimate bug +For the record my comment about was for improved readability only its less verbose and people know what the method does +I adapted and ran the patch on trunk with machine created regions did some enablingdisabling +Shouldnt this have been Nullable defined in common +Merged into and in master. +things like the not using hard checks is one reason why I want to go through it a second time +File +the revision Thanks Noble +Here is a proposed patch to add includeexclude functionality +timing differences due to different RPC stack maybeChecked a Hadoop build and saw this too which lends some support to this hypothesis +well default is mentioned when you check out sources from SVN and it should work out of the box +You are right about closures changing the behaviour +Then we will override the first servlet +Screen of in with debugging information showing the error in Javascript part of +Is there a planned release date for the latest version set already +Is the default content charset statically available anywhere elseIts defined in the these days +jbosstoolslogorv +Committed to trunk +Since is not used I propose to remove this component +But I cant use the naming trick because a configuration does not have a defined lifecycle +bq +This one works much better +In case anybody wants to try it. +How do I get the original alias that was used? For example if I map UpdateItem to the operation and the actual operation that comes in is can I get the or the Foo string from inside my server codeDave. +Im almost sure we should harcode the in all Serizable classes and backport the Changes to to issue fixed in released version. +Commited at revision +I would rebase it to latest trunk to make it compiled +Ive included testng and spring yesterday in my local build +Tried which seems not to be compatible +This functionality would be useful in for example +Essentially we special case the situation where we cannot set the parent class loader to null and in that case set it to the system class loader by default instead of the boot class loader +Therefore the name of the namespace itself must be all word characters while the variable can be anything +This give the container an opportunity to easily track what mail objects are aroundAlso cant be static so you would create a new Mail instance starting from another Mail instance dont like this +Then the solution could be to set a default timestamp naming strategy in Application and you can override it through a +this is the patch to fix the the rolled up patch in fixes this. +Hello luhainingAFAIK nobody in the team speaks Chinese but I hope that Google Translate is giving enough hints for resolving the problemBest regards Florian +If you manually run the identity command on the same image you uploaded to receive this error does it output the expected resultIf so then mind sending me the entire stacktrace of the error and try opening with Preview or some other program the image referenced in the error to see if it looks right or got corrupt the Ill look at creating a simple Ruby application using to manually reproduce the entire scenario +Glen you are a gentleman! I like this idea and very much appreciate your offer to work on it +I would expect to be able to successfully interact with both server implementations by configuring my provider with falseShouldnt the flag on the provider control how it sends attachments in the request to the server and the flag on the consumer control how it sends attachments to the client in the response? Both the provider and consumer should always gracefully receive mtom or attachments +Thanks again +Merged into master branch. +To ensure that only the aggregate document is schema validated the handler was setting the schema feature to false to keep the child pipeline it creates from containing a schema validator +I think the file I attached is an RSA key +Looks like it might be an issue with your datasource config mind sharing yourSure here you godevelopment adapter jdbcmysql encoding utf database gdfdashboarddevelopment username root passwordeventdirectordevelopment adapter jdbcmysql encoding utf database username root passwordresourcedirectordevelopment adapter jdbcmysql encoding utf database dcat username root passwordideasdirectordevelopment adapter jdbcmysql encoding utf database ideaschest username root passwordsdrvdevelopment adapter jdbcmysql encoding utf database sdrv username root passworddefectdevelopment adapter jdbcmysql encoding utf database defect username root passwordtest adapter jdbcmysql encoding utf database gdfdashboardtest username root passwordeventdirectortest adapter jdbcmysql encoding utf database test username root passwordresourcedirectortest adapter jdbcmysql encoding utf database dcattest username root passwordideasdirectortest adapter jdbcmysql encoding utf database ideaschesttest username root passwordsdrvtest adapter jdbcmysql encoding utf database sdrvtest username root passworddefecttest adapter jdbcmysql encoding utf database defecttest username root passwordproduction adapter jdbcmysql encoding utf database gdfdashboardproduction username password foobar host db port eventdirectorproduction adapter jdbcmysql encoding utf database username foo password bar host resourcedirectorproduction adapter jdbcmysql encoding utf database dcat username foo password bar host ideasdirectorproduction adapter jdbcmysql encoding utf database ideaschest username foo password bar host sdrvproduction adapter jdbcmysql encoding utf database sdrv username foo password bar host defectproduction adapter jdbcmysql encoding utf database defect username foo password bar host +Thats something normally done by a moderator of the mailing list no need to involve infraA moderator should send an toand reply to the confirmation request +was the reason that EOI check was needed +Use this version instead it includes some error handlingOtherwise the window will auto close with an unhandled exception +This also means doing some integration tests and really supporting this feature in the long term +The problem was while submission of the job job client was not using the reference of the UGI which is been created while creation of the client object however it was defaulting UGI of the proxy user +bulk close of resolved issues. +OkSo if you invoke a out side service what is the binding that server uses for the epr you specified +Hi Anoop Ram Ive attached log files one of the master and one of the region server that died +exe and didnt see it +Tracked down to an issue with my env migrated off that box and next release will not have this issue. +Resolved +Thanks for spotting that bug Fred The problem was due to having the project configured with support but no lib in the classpath resulting in type not found giving a null type hierarchy +This is due to some svn mv and I created the patch with svn diffI can provide a patch with the complete diff but you will loose the svn mv infos so the svn history of the file will be lostAny advise is welcomed +Also I just replicated the same issue in Grails +version +Slava can you take a look +Defer to +If not I suppose we can look at using the +Code Assist for tag attribute has to check tag to know if css paths have to be added to Code AssistWhats thatCould you provide steps to reproduce? Im not sure I understand what it means +In fully Arabic paragraphs this could be rl order but then the breakpoints must be interpreted as breaks at the lhs +Someone might appreciate a comparable deserialization feature +Patch attached to pass checkstyle test +We had to hack the jar with different version +It was there since long ago but now it reproduces more frequently because there was a harmful optimization in that was hiding this error and this was removed when fixing +push out to +sbutton with propagationnone does the conversation job fine but without submission +Added support for Integer Float Double and fields in Git ID ddeafbbbfdabfb +Could you please indicate which version of Axis has this fix +Tomek update on this? Is this working now +IIRC this was the rationale used to define the current implementation of Complex equals +As soon as well have defined the final position for this documents well update the urls in config files +This problem also occurs in other parts +jnlp fileI would like to add a wiki page how to get Camel working with webstart and how to generate the jnlp using mavenMaybe if you have the ideas how such a wiki page could be please feel free to post it hereI will test the patch a bit more before commiting +For some time we have two schema objects hanging around and as cores reload theyd get the new oneBut that just proves I should look through the code first +Sounds good to JacquesI have checked the point raised by you and yes you are right I didnt found any hidden field with h +Undone automatic updates for +Works like charm now I get a nice messageYour default time zone has been set to XYZThanks. +If we dont expire previous assignments then the Report on Who uses what shows wrong dataIll like to expire old assignment before creating a new one Is this OKRegardsAnil +Yes but next mvn compile will make Eclipse happy again +Verified on JBDS +Awesome thanks! I will test today and post the result +This is necessary for circular references to work correctly so two EJBs can reference each other via JNDI without causing a circular dependency error +The configuration name for the acls enabled should be following the nomenclature in branch and trunk +Good work +How did you get the filter through the assert statement without final? Strange +But in there is a that extends in +henry i realized that didnt answer your question about the diff +Thanks again. +My earlier tests showed this provided a large performance increaseAlso looks like you cleaned up the which is good +I am unable to reproduce issue with trunk or +I dont have a nice simple unit test repro +You can replicate the issue by using the in the attached + +I will write the notes in a comment here so you can copy it to the wiki +This quick start will not considered at this time because of the complexity of dependencies needed in AS +Attached new patch for services API fix +This attachment has full trace logs +Moved back to does not rely on the jgroups jboss config already supports that property so the issue is about making the testsuiteuse it +This simple patch seems to solve the problem +I had to do that to reduce contention and required synchronization on the bag contents +Need to do same inPatrick its ERROR level when its run out of retries usually +How come this didnt make it into beta +Now that settings are validated after being parsed its no longer possible to reference a non existing cache thus this NPE is not possible anymore with a validated resolver. +Calendar works as expected in revision for version +Bikas is this needed? Are a combination of total and completedfailed counts enough to detect progress +Would it make sense to add a topic on the EXTERNAL NAME clause for this purpose +I will investigate merging this fix to +This issue occurs when performing a normal mvn clean install rather than releaseprepare +This patch would validate if the message is null +domain does not equal container as per the mailing listI will now proceed with the supportedFields changesIf this proves good well merge down to + +That would be suboptimal +Switching to an improvement +However the way escaped chars are handled is problematic as we remove the during theWe have to get rid of those problems and this has to be done when we will refactor the full DN parsing which has gone totally wild +It has been doing for openejb release and release +The patch has been appliedAlexey please check that everything is OK +Could you give it a try if possible? + +Blocks on individual datanodes may even need to change their idsOR one could copymove the entire filesystem to a new clean oneThe selected method allows for a seamless upgrade requiring no conversion and would work equally well with a id address space. +Attached the new patch with explicit null initializer change +As part of the upgrade an IP type for logging was added and the format made binary +These description added to components but CDK ignores them should be processed for additional elements facets listeners etc. +BULK EDIT These issues are open to be picked up +Although it seems that it is not updated but it will somehow help you +Yes a flag can be added but could you tell what is so particular about the constructors? Groovy compiler also adds getterssetters for properties and for them also I dont think there is any way to distinguish whether they were user provided or groovy generated +Thanks for looking into this +well see +Do not hesitate to use it +So it doesnt seem to be suggested to actually customize those +Fix Versions Do you want this backported to the branch as well + blocks Tier db testing on jBPM blocks cluster config is a GA blocker can be worked around for link to as that JIRA is closed +Please review and comment +I am going to rebase this branch since I believe it is now with trunk +The current code in has this part commented since it was failing back in M +I think the only way this would fly is if we can get the console bits which add connectorsetc to update the configuration file so that one can still use the console to add muck and it will retain them over a restartPerhaps also the console might need a simple editor thingy to edit the Additionally we may want to consider including the +I will do that. +I get from to random duplicates on my laptop when sending to blocker +Please check whether you have mixed up jars from diffrent Axis versions +I agree with Florians comment +This is probably a spec issue and not an implementation issue +It was problem on my local filesystem as was not getting copied on Filesystem +I think that would work for the fellow who wants it +I tried doing a little bit of work to resolve the typedef to the underlying type but that didnt do it +I close it in order to clean up. +is refactored to provide a utility method that serializes and then deserializes +Im looking at the other test +The test not fail repeatedly at the same point because there is shuffed which is romoving fromIf is in same order as List which we are removing than there is no error reported +Closing all resolved tickets from or older +Im actually pretty happy its used in ADS and was really just pointing out the licensing issue +Also add the documentation for as part of this JIRA +u got me +Normally iPOJO uses the pattern for such kind of injection +In revision updated the code to use the interface name specified in osgiservice jndi name for service proxy +For some reason I thought this was committed so it fell off my radar +Patch introducing test which tests both and In addition to the patch the part of the test requires that someone create a file srctestresourcesmrepomaven this is not part of the patch because the patch format does not seem to be able to represent the creation of a file properly +I dont know how to delete an attached fileNot really an issue with CXF itself but may be some issue with how or maybe the route itself is setup +Is there a similar location for automatic syncing of their sourcesI can create a new bundle but there is risk of me somehow accidentially submitting with the wrong pom or something +Thanks +But the throttler itself could be overloadedThis only applies if asyncDelayedtrue was enabled on the throttler +If you ps on the console the old version is printed +Thanks very much! +What still needs to be done here make blocking of released artifacts configurable per repository implement functionality to web upload selenium test for this feature once this is implemented to the web upload form. + solves the issue of leaving tmp file created by the mapjoin undeleted when the job is killed +I found that my code is not being executed and my deploys are taking roundabout minutes so thats why Im still encountering the problem ocassionally +It feels strange that we have to register the consumer here +I added fix version to SBS + +Sure enough when I removed the require from railsadmins gemspec my app booted fine +marking as fixed. +All tests passed committed to trunksibmsvn commitSending javaengineorgapachederbyiapierrorSending javaengineorgapachederbyiapisqldictionarySending javaengineorgapachederbyimplservicesdaemonSending javaengineorgapachederbyimplsqlcatalogSending javaengineorgapachederbyimplsqlexecuteSending javaengineorgapachederbyimplsqlexecuteSending javaengineorgapachederbyimplstoreaccessbtreeSending javaengineorgapachederbyimplstoreaccessheapSending javaengineorgapachederbyimplstorerawdataSending javaengineorgapachederbyimplstorerawxactSending javaengineorgapachederbyjdbcSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javaengineorgapachederbylocSending javasharedorgapachederbysharedcommonreferenceSending javatestingorgapachederbyTestingfunctionTeststestslangSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +I still see those warnings when Sessions concurrently read a property while another session removes it and adds it again with a single save call +Now that has been integrated into can we resolve this JIRA +I think we could do one of the following Improve the error message to note that the admin should make sure the other NN is active before proceeding Have it automatically transition it to active if this is the caseWhat do you think? I think makes more sense since the admin has to make an explicit decision +Im morally opposed to this as it wont be type safe and each component that uses this will either have to include casts or its own usage of thePerhaps we could come up with a better informal communication channel +Yes bad habit +I was just able to complete lucene checkouts successfully on so Im cautiously optimistic thanks JoeThe IO on the box is still shockingly slow +Any objections +Still sstableloader throws NPE on the execution path at +Hi Radoslav are you following the branding discussion on the dev list? This seems quite related +Awesome +Ill look into those when I get a chance +I can confirm that the same issue exists with that works in MRI fails in thus reproducing the +have doubts myself +Patch for attached +Working as designed. +Attached patch that adds the XML examples as per the defect +Thanks for logging this issue Kim +Proposed diff against trunk +I tested the performance issues and it fixes them +uploaded rev cool +You can refer to the intermediate v patch in +HiIm also affected by the same issue +but also touches the same part of the code +So this should guarantee its in the list and alive before ruby Threadnew returns +To be clear Im proposing the big patch for trunk small patch for changing the Adaptor interface is a biggish change for this stage of release cycle +This in turn calls my own class which implements the generated interface +Updated autocomplete html to disable the owners drop down if no one is selected. +Alignment of REST services to Stanbol web Albertoyour patch was goodThanks +It is actually neither +Since previous patch doesnt check in +looks goodi will commit it if the tests pass +Sent from my iPhone +Great +We have tests for all +We usually recommend placing any Java +So this remains usable if the certificate is within the limit consider making the day checkwarning configurable and disable it if it is +Dont hardcode the name and comments +In my opinion KVPs are silly to use anyways there are already limitations +Neat +Attachment The DTDNote there are now third party XML schemas for citations we should probably use one of those but I dont know if any have become a defacto standard yet +Hey Harsh J How did mainipulate this batch? Did it have any problems +I am sure this will come up again for other people trying to port existing portlets +The timestamp error is likely due to the difference in our timezones +I agree I had the same thought when I wrote it down above +Limitted and unlimmeted query plans +Cool will change the error messages as suggested +Wido can you fill in description for have requested the libvirt guys to make a new release of since the code is in their master branch already just not in a releaseHopefully this can be fixed in For now the docs have been fixed to instruct users to add it to was released +Native SOAP stack doesnt support JAXWS any more +Applied and the itests now look like they work fine so ill add them to the build as well + as near as I can tell +zip files in the Customer Portal +Alexandre good idea! I like boostCan you please provide option for this featureThanksroger +or other +Thanks Tony +is slated for Friday release +Bulk close resolved issues +So I vote to commit +Add as a new plugin in repo for +Thats a good idea but you cant null because its part of the primary key and a not null field +I was just writing these as an exercise so they wont do me much goodThe licence is the apache licence +If we get specific information about other we can add it to the Developers Guide topic +However in most cases nested tables wont help +Waiting for a review on this one +Ted alternatively we could just shutdown the datanode instead of disabling heartbeat in leaving rest of the test as is +The problem is that object is not just objects they are thought to have behaviors and the Principal class has a behavior around its name see the JAAS Reference Guide and JAAS and Java Tutorials authorization in Java as I understand it is based on Principal names and personal I have only seen the name property used when two Principals are advice you again to use the name property to compare the two Principals as that is as I understand it the natural behavior of the Principal you look at the code I suggest it takes into account that the Principal might have a special behavior around its name and for this reason use the equals method +I see that makes sense +Move code from converter to Seam +No need for IT. +Please comment any further information to that issue +select +Olegwhile there is no strict need to have a COOKIEPARSINGENABLED flagI still feel it is more elegant to have one +Could not reproduce this ve attached a patch to the and copied some of your code you could your code or patch the and modify it in a way to proove this issue against the current version of +The file was removed to +In short this behavior is NOT documented +I assume parsing messages from AS exception is not a nice way to get some information out of itPerhaps there could be some other way to get an object or XML snippet with the error info and render itBut that would probably need too substantial effort in AS plugin. +Do you mean more test cases +And I think it should be turned on by default +Hey BenHere is a patch for this one +Thanks for bringing this to my attention and for supplying the am currently working my way through testing the JMX features in Spring for a proof of concept I am working on + +If thats incorrect please do it and retag the issue as. +Hi Reimer could you try to set a fake jdbc url in please +The complexity may be the result of me leaving in the existing painful instructions +The attached patch removes the redundancySomeone with authority should probably apply this patch since it effects HEAD etcIn the meantime IRIX users can use the sproc threading option in runConfigurewhich will produce a library useable with pthreads applications +PaulexApplied patch to ARCHIVE module at repo revision Please check that it was applied as you expected +Reopening issue until it is incorporated into the nightlies and is resolved +As per the core spec add new package dependencies to the header +HadrianClaus I understand Ive attached the code from my repo and granted inclusion rights +see comments +A review of the commit is welcome +Hi allI have added two patches which improve those portlet pagesThe old pages of this portlet contain many table and make the page too long to read for the usersThe new pages put different table into different which makes the page not too long and the user can easily close any which he does not want to checkThe default monitoring page has there big tables which makes the page too long so put them into different to improve the usability +Shifted off the top of the hour next sync should be at GMT let me know if you need anything else changedKarl +I think so to +No integration tests + +I think Groovy should not go the path of using custom hidden type transformations. +Its not possible in normal SQL +we should try to control some order here that we can not control +Nice catch! +If you were to simply use the Logger interface you would not be able to properly use the +Ill try that revision out later today but I dont think thats the issue + Something broke in the tomcat builder the context path should never be null +I second that as I had to scratch my head for minutes to find what to doI am under Mac OSX already fixed but not released and there was not JIRA for it +Thanks for the reviews and discussions +Was released in. +An alternative would be to allow the user to provide an instance himself instead of configuring with a classname in order to do some configuration at runtime perhaps +Marking as fix version but leaving open as I intend to merge this to eventually +Ideally a cleaned up patch would not delete old constructors but rather deprecate them +Patched against revision thanks for that! +The WAR generated from a process definition is deployed outside the EAR yet it uses classes from the EAR +Its an awkward implementation not the fault of the author but because no mechanism for passing extras rpcing +Ive just committed this. +Im going to check the stacktrace this saturday +Link to define FS like the get Delete approach for reasons already mentioned +Unfortunately this is still not resolved in +Tables should have TH tags consistently +just fixing some more bugs +Probably the first one should refer to Initialisation Initialization +Please read my posting from April th as it explains the problems I am experiencing +How we can reproduce this issue +Please let me know if you think I can go ahead and commit the changeCheersAshwin +So more useful woud be successful count and failed count on a per session basis or something like thatThat is perhaps best decided in a different ticket though +I have handled and seperately +Unfortunately updating to changes some that will require more changes and testing +It is the same bug +The and fixed compilation errorswarnings in revision . +A full publish override is provided for each module +Spring is not released yet +Thank You Rahul! I will try and take a look at it +rescheduling +The port number was picked up as although I had configured Apache to run on +Running the tests results in errors but those errors also happened before applying the patch +Have to set priority to blocker because of this blogs from making a release +Thanks a lot for the patch +Active master is published in ZK +I think we should keep it open until the patch has been applied and committedThanks Veresh for the patch +Added factory and static utils class +Is more work on this forthcoming +they are read fine using +associated WSDL +You have to configure N max maps and N max reduces per node if you want to be able to execute some jobs in parallel +Part of mobile showcase skin work that has not been completed +interestingly this becomes much less of a problem if gets dealt with because it significantly decreases the chance of failure in mark and reset +zip distro +Please reviewAmit +Closing this again since the patch was committed +In this case rethrowing socket timeout as a recoverable exception sounds wrong at least from the conceptual standpointI agree + code has been merged into trunk +Id like to extend the property expression language quite a bit more but keep it concise and type safe +Yes im using from xalan because xalan no longer supports the class am i rite? So is that causing the problem +Running the quickstart using the picketlink binaries seems to address this issue +Will we get a sync storm when we have a lot of column familiesWe do sync per region not per store as far as I can read the patch +Resolving issue as the patch was committed by Satheesh with svn +hmmm +Ohhh I did not notice the problem in first place and we agreed that we are not going to add any more new features +fwiw Im running this on r of the +Correct demo data present on specialpurpose folder and add and missed on after ant with resultTestsFailuresErrorsSuccess rateTimeThanks Jacques and Scott for your observation +this is the simple test application to there any way I can help in solving this problem? I can try do do more debugging on my own provided I get at least some hints on where to start looking +mothballed component use wicket viewer instead. +I have fixed caused by validation job +I think I was wrong in assuming that switching to REQUIRED helps in any way the users are still getting these concurrency errors and we havent gone to UAT yet meaning that theres at most users active at the same switching to REQUIRED makes the test case queries to run about x slower so this would be a definite in production +Rebased to current trunk and no JMX +I think we should use to detect the version and use some fancy reflection magic to work around this +are at the moment connected to the form +Sorry Joseph I misunderstood your proposal +push out to +I agree with this issue +Sangeetha can you paste the result of the below sql execution to the bugselect From host where type +deployed on bsc all resolvedfixed issues of already released versions of Roller. +If they are the patch must be reverted +If you cant open a sstable because of a bug in cassandra then its almost certain the actual culprit was in the flush or compact code so knowing the extra parameters at the time of getRow helps not at all +This makes no sense to me +HEAD is in a poor state at the moment though and has problems building +With next build you should be good to go +Closing old issues +I forgot to updatecommitted to trunk +Im not sure to have time to investigate on that but Ill give you administration rights so that you can see if you find a way to enable such issues linking. +As discussed in other alternatives could be to drop such tuples altogether or to assume nulls in all the fields of tuple +Action now calls the on Tools +This is solid on windows I need to test on linux with the MSR +This latest patch removes ALOT of the unrelated code +The rest of revising and new only remaining TODO items in the github docs are deploying to jbosseap +Fair comment but its largely irrelevant to this patch +I will post a patch later today + The login links are used only in developing be verified in gatein master and reopened if still valid +Hopefully something will come out of it. +Efficiency improvement committed in revision However I did not yet include the speed test which I dont quite understand +That said a patch would be welcome and i will happily review itThanks +Bill Im not sure about the log output of the DIH with threads it would appear to be specific to the DIH based on your description and unrelated to the files involved in this patch +sorry i meant if you could have a look at the +Should be easy to fix +is an internal class specific to Hadoop so we shouldnt expose thatAt a higher level there is something else I would like to be able to do as well multi file splits +Patch checked in. +This feature will be dropped in +We also need to double check the ErraiGWT archetypes if they use enterprise dependencies +I think is appropriate for this case +A release whick fix this is about to be released +Vasilythe default style sheet can be substituted by user one +Added wiki documentation as well +We used a separate directory on jenkins before +Until we receive this notice we will continue to commit to CVSCheers for the delay MarkIll aiming to spend some of tonight hooking your scripts into the migration script and repeating the test migrationOnce that looks good we can plan a real migration +Please add them back +If a file is found those requirement builders fire locally fetch whats needed and the tests proceedSounds like a complex and not immediate workflowframework +the resource and the permissions for that +Its better to separate the actual UI from ATS C code +This patch uses the block to understand this am confident that I am reading it correctly at the least because of the but need your help to make sure I am not wrongSo In the usual case of NM killing AM while it is running and before AM unregisters with RM the error is visible on UI +set the default timeout of to second timeout is too short for some of the types of requests that may be run +when there is network congestion the Routers will start to drop the packets and which will cause the write on the socket to hang +Marking closed. +Separation of concern and clearly laid out package structures are both very important of course +Nope +I dont know the Hudson API at all so cannot help you on execution +Let me close this issue +If you are not redirected within a few seconds click herethen on the next pageRedirecting to the login page +Upgrading to the official release of resolved the problem. +An easy workaround is to build it with Maven +Ok +Ive added audio mime magic test in r we just need +by setting +Upgrading prototype is not possible in final is going to be final JSF release +Further if we do see them getting printed multiple times we know we got something to fix +You mentioned tomcat crashed when more than one person accessed the app and I can tell you that grails powered app can definitely run with multiple user access even with client jvm and tiny memory allocation +Id like for their still to be a limit +The exploded EAR is zipped +Will add a unit test and do more testing +Removed some of the attachment that are just confusing +Not able to run ant because the trunk does not compile +In general can we get as many of the committed while not introducing any regression? I dont think we need to solve the issue completely until we can commit anything +Forgot to mention +The last batch got added because thisll be an M release and we were blocked on the unreleased pom +So it would berequest Aa Bb Cc timeoutmsimmediately get dataBb Ccrequest Aa Bb Cc timeoutmsno data for any topics request blocksmessage on Cget data for Ccrequest Aa Bb Cc timeoutmsEtcI think this multifetch approach simplifies things for the consumer implementation +Id like to commit this in a few days if there are no objectionsI dont want to conflict with work inI think its pretty straightforward and all the backwards compat is preserved in any new code +These are google apps domains registered with godaddy +Fixed in SVN. +To address Prasads comment Ive created +That is what Ive experienced as well +I agree with Emmanuel that it shouldnt because avoiding race conditions would require locking between check and insert and that should really be up to the developer +It works fine in Drools +dear I didnt spot this getting added yesterday Manjula was already working on converting this test seeOops my bad! I always search JIRA before I create a new issue I dont know how I missed that oneWell no matter I didnt spend very long at it so Manjulas is probably better anywayIll just close this then. + +This is a side effect of a bug in Jenkins where the target platform for Kepler M is being built from the branch in git instead of the one +The bug is fixed isnit +Not sure if its helpful or not but just in case +It is possible for some task to try and use a random number or something to produce a unique file name this would result in multiple copies of the tasks output showing up +I searched the forum but found nothing that would have help me with this situationI also tried posting my problem on the forum but got no reply +Looks good +Actually looking at the javadoc this behavior is correct +The client can then decide to look for a replica with a CRC emit a warning fail etc +the new patch looks good +Preliminary patch for resolving this issue +When posting patches over HBASE in future Ill make sure to number them every revision +Punting to +Seems that the issue reported by me just above is fixed +I am kinda late on this but I would appreciate if someone can provide brief description of how this patch improves the memory layout and alleviates the spill problem +We should still keep open to see if we can do better about not violating . +Code that depended on deletes being instantly visible across threads would no longer be guaranteed +See SHA cdfdFix also includes Recompile workspace after workspace script folders change Recompile project after project script folders change Cache the script folder checker for each looks like the transform is being properly run on incremental compiles but not full compiles or at least something is being done incorrectly on full compiles +My guess is youre running Eclipse off of JRE +Closed upon release of version. +I definitely agree this isnt a jruby bug but since I saw arjdbc bugs filed here I thought this site served both code bases +Can you try that with your changes +I tried downloading just the pieces of guava but that has checksum issues too +Patch that applies the described changes +Why would the test jars need that unless you are actually copying code into your test tree? You arent shipping the jar or any junit code + +It should not have been put in program args but rather in the vm args +By the time I was ready for that was out +bq +patch as svn . +Simple test case raising the NPE +So all issues not directly assigned to an AS release are being closed +Even though this practice is not recommended we should fix this to maintain compatibility +Magic guessing removed for xml +full test attached +number to see what i was looking for to show last two distinct txttitlu and for them latest ep +All JTS jars are in captitals now with the move to +This patch implements the needed functionality although perhaps not in the cleanest way possible which would require more refactoring than I felt license to do in a patch submitted to a bug tracker +Will need to dig in on next jenkins fail +I cant promise real progress for the next weeks since my job is keeping me very busy +Is it really the same session if the first request was never completed +so it would seem to be something external that is causing odd output hopefully will fix that +I dont want to make you generate another patch just to fix the majority check on that message so if prefer not to do it could you make sure there is a jira to fix it? I didnt quite understand why you moved all that code between and +Shadow works ok +Please investigate on the forum and assign it properly when its done +Ah +This problem has to do with the gzip compression and occurs when allow compression is enabled +and you think x null works in sql? please think again +I added the default value for the K value in revision Please try again with the HEAD version +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Ive run into situations like this also +jgplugins was using instead of +This should be the last one blocking +Does not seem to be an essential feature + on Todds comment +Sorry for the late responsePatch looks goodOne minor thing as we decided to take care of the version stuff in things related to that we can remove +As an example aba depends on b cb depends on d qq depends on zThis currently results in a b c d q zThis item will change it so we have a b c d q zYou can thus tell why a bundle is in the deployment metadata quickly and easily and this can be used by the when installing the application to determine how and where to install the bundlesIn my mental model this separates into isolated shared but it is down to the to put this into practiseAlasdair +Note this is for +proto file comment to make class names agree w what is now generated +so I think it could be useful to update arquillian version in txbridgetests +Jan what specifically did you in mind given s lack of command line testsAs it currently stands the entire source is built for distribution each night by my systems and published as a nightly build +Hi Olga lets open a new issue and start discussing that with the community. +I didnt understand what had You want to t this dialog allow to add new css classThis dialog allows to edit different files and saving during activity with the wizard is required +For instance only has less server configured compared to. +Hi RaymondI identified the issue here this is because of the implicit services imposed by spring beans +the new patch looks good +Updated some tests +Andrew if we go this route wont we break all existing MR jobs? You think that OK? Or should we try and leave the MR as is deprecating it and come up with new MR classes that implement new stuff +patch to allow Eclipse workspace artifact order to use the patch you need to use a newer version of the embedder jar file +added integration test under has been obsoleted by the fix for +Okay just changed the implementation of the iterator so it follows the correct convention +There was an existence test buried in some old code which created objects +I guess it was missed intially due to confusion around typo in Intel Architecture manual which does not list this opcode but AMD manual and practice proves its OK +Showing available versions for EAP +some work still needs to be done slip to GA release +As mentioned before putting the WSDL does not change the message serialization formatThanks you attach your patch please +Willem gert is on vacation +had local changes on adding the form comments back in +From this view it makes perfect sense to let c return an int +According to this it is excluded from test execution for other tests should work without failures +So heres the summary Fixed Not a bug I think its reasonable we pass on generated ids in the metadata I dont think its really relevant as its an edge case +We have not been able to attend this in or was fixed while I was fixing memory leaks for +Fix committed with regression tests. +Now I have serverdefaultdeploy as the result so I close the issue +Sorry its taken me so long to reply +and passwordField of course +I think I can see how this might happen now +But throwing an IAE is a contract change and the NPE happens right away when null is tried to agree with Bertrand +Thanks Lars and Jimmy for reviews. +Merged to merged cleanly for the engine code test needed to be edited slightly due to no identity columns in +Its also worth noting that keeps open until an GC completes so in the meantime you have both the old and new ones around +Still quite a bit of commented out code which will hopefully be tidied away this wee as we get the and Endpoint serialization story sorted out and we can finally remove all of the extensions from these classes +Thank you. +There is no any descriptions how to reproduce this issue +this is not really a critical issue +Are you sure you arent seeing both the event from the filter and the one from the +Applied. +Its such a huge patchHowever whatever is easiest for you +Fixed on trunk +good to me +Theoretically if we guarantee full backwardforward compatibility I could use my old application with the new client for some period of time but that would be a burdon because I would have to test my old application with the new client and who wants to spend time on thatSo how would I upgrade in this scenario +srcmainjavaorgapachehadoopyarnservernodemanagercontainermanager +All the hadoop slots on Apache Jenkins are upI think we should attach the final patch for TRUNK and resubmit for build +That was supposed to go on +sync hlog +In any event this should be part of the server classpath as this is a required component of the javaee specification +Please use when generating patch using git. +So +Worked fine in unit tests but gets a when running on the Hadoop cluster +Now with operator lifecycle events in place Im experimenting with a more generic way to monitor the state of operators and their input channels +this would be a change +Id like to see what youve done there +Jeremy yes a GBK would ensure the single partition + Most changes are simply replace with +I agree with Steven the standard directory layout should include the test directory and so on +Once again sorry that I have not gotten feedback on this to you Ralf +Jirka is this a single instance or a cluster +Hi I revised the test a bit +I suspect this was fixed by something like +fixed in svn rthanks +for the new patch +The UGI is a critical class +Thanks HyunsikI missed them +This patch does not requeue flushes compactions or splits if they are already in the queue +thats because we havent made a release since we fixed these issues. +If a user purges some docs compacts the DB then queries an view the pointer will be nil and the view updater will probably just crash +done but i guess i should use clogging instead of jboss it is probably not the appropriate thing in this case +Apparently MX records with IP addresses are not in accord with the +Verified in EAP +and verified mvn clean install +What about having a special Geronimo class that is designated as by Geronimo as primary then have a convention where a special login module would insert it into the subject? This way people could have a variety of schemes to insert the primary principal by simply writing their own login module that followed the convention and we wouldnt have to have an uber metadata and code to handle all the different possibilities +for the patch +Ill take another look at it and respond +Mention that the term should have been indexed for to work +Thanks Sreekanth +I have tested manually that page renders and that sorts by last sort correctly across month and year boundaries +Thanks for doing all this everything looks good now as far as I can see +We should probably file a different jira if you want the feature any way +Will look into this after the new release unless you have a fix +I just committed this thanks Bill. +Even with set to true Im seeing two MR jobs getting launched +Hey Carlos does the above address your issue +and un reviewed for months +File to be used as source for the header binary field in JBMMSG table +Futures sounds good to me +Such page would have to set the archetype properties since it is the only mechanism supported by archetype core component for passing parameters to the archetype +Perhaps we can muster the votes and effort for a release +Please have a look at the patchThere are several questions open to this issue is called after validate but impersonation is now done after validation before getting the resource resolver +I realized that penny must be removed from too +For I realized that there are a few more changes we will need to make to do preemption based on multiple resources so I would prefer to make the change in a separate patch +Thanks +Nothing we can really do about this atm +My original thinking was that this would be useful for allowing applications to be moved between Fair Scheduler queues from the command line +Im leaning more and more towards we should implement ary indexes querying first then later add full view support since the former we can do wo opening the whole user defined functions box which is a pretty big deal +The chunked encoding is the HTTP feature +Sorry to take up valuable time +bq +Flume is still read only +Sure +Here you have registered the metadata so we need to see the XML that is being delivered by the and the metadata that is being used to generate the static classes +Heres a patch +a lighter hammer would be useful imho +Hey guysAre you hoping to try to get this into for beta? The target version is currently v +Tested it via grunt dev and grunt couchappdeploy both worked +The other part where the client waits Ill have to check +Attached is a test case and a patch for the issue +Doug you want to take a stab at this +We will pick up the fix from the groovy compiler. +jruby java +I think the site should display both and we would certainly benefit from some very explicit traversable documentation of both styles of pom relationships as they both have a very significant meaning in terms of logical and physical component and package I dont think they should necessarily be related +As far as I can see that is not a valid WFS request the filter is still encoded using GMLPlus you dont say whats wrong do you get an error do you get an unexpected result +Eclipse is more strict about these things +Strictly speaking this is not a change suitable for a patch release since it changes the API of a core class +Hi JimI have tested this issue in Linux +Please test I explicitly specified the classloader to file a new issue +pushed both upstream +HDFS controls all of its components +patch attached +bulk defer of unresolved bugs to +However the above comment doesnt apply to test failure which still have an unknown cause +fix blank file in is resolved for Solrcas +now +How about a Util that basically has the same kind of stringify method +In this patch the tests in will pass +bq +I think we should avoid changing to +This patch goes on after the previous one to change the raw doc to the unique ID defined in the Scott +I assume you were wanting a patch file +Ill be checking it out during the weekend +I just created to address this. +Patch updated as header file name have changed from to at rIlya please use unified loaction for producing patches normally it should be project directory +So you might be able to reuse it as is +You probably want to make Cleberts suggested changes and then assign to the CC team for review + problem remains in the version +One is only for leasechecker and anotther is only for close +Burr could you check if you were using bit java to run installer and try it with bit java? Im moving it to next version for now +wtf is a Kmeleon browser? +If no further objections I am taking this on trunk +Moving it out of isnt a bad idea though but the API should say on and call whatever the new class is +Is my understanding rightSo could you please answer my questionBest RegardTang Yong +However if we want to keep the current order I think we should make some changes in the implementation +buildr instead +Havent seen this in a while. +HTTP chunked encoding works both on MRI and for me +Thanks for the prompt reply Luke! The above curl was from a remote JHS +META +Copyright holders can choose to license things pretty much however they want +I think Henryk and other in the community wants to maintain more activity and ensure releases gets promoted to maven central etc +I was unable to retrieve the correct ones +This happens because dependency arent currently candidates for profile it seems we will have to trigger profiles for these +start browser +but fwiw we generally dont run snapshots or milestones in our production environment +Anyone still working on this? +Thank you for your commentWe do not care about the message but about all the prefetched messages for the sleeping consumer which needs to be redispatched delivered to other consumers +The previous patch had a few stale changes +The jetty server starts for me if +Yes Handler is the best name +but +Also there is no support for foreference by rtf output +This is coupledwith the fact that objects in the dictionary cache dont have atable descriptor field when they are first loaded the table descriptor value gets setby +this is an attempt at a patch for this issue on the plexis couldnt locate the mechanism that plexus would use for a cleaner way of obtaining url username and password from the https url string so I added a couple of scrapeX methods to the builds in just fine +That was what I tried to tell you +So this issue itself is fixed closing +When I try to implement the plugin in xCode Im getting warnings similar to Incompatible types assigning struct expected struct +Ill have to come up with a fix for the test to show the problem +Ive seen this problem when running with Suns JRE version although I havent seen the problem when running with IBMs version JRE IBM Windows build +In our case its quite important and in fact a showstopper for using the plugin in our use case is as follows I need to work with some complex XML schemas +moving open issues in X DT to X DT as discussed with Bob +patch file for +I am concerned about the performance of Trevni +What if the implementor actually wants to return DIRECT in some circumstances +The current code is pretty simple and straightforward +i dont know why but the th option in attracts me +And Cagatay mentions thatIt should be an implementation issue as delegate CDATA calls to Impl using and end api PF uses standard apis for this +isnt yet a stable release but a feature release and it isnt supported yet by date format is totally different in cvs and we need to allow new cant change the actual format by HHmmssZ for all cvs client because lot of them doesnt support this format +If I were you Id talk to the maven people abouthow to get more content into your CMS source tree +Commit to Trunk Commit to +Attached patch fixes the HTML BUT clicking the link generates a NPE message +Better to add an error message to these checksI have not changed this part in the new patch yet +Correct except that collapse result is only used as filter to the final result to hide collapsed documents +okey so on Java with any SOAP service it fails +for the noise +And would also affect EAP +The web based client should use instead of instantiating the Service classThe standalone client should should list jars for CXF or at least mention that the jars are different for CXF +Now it compiles correctly and I started testing +Fixed in r +On the one hand right now the top level generated directory has only subdirectories in it and no ordinary files +Committed to trunk and +Thank you. +Perhaps it will be overridable with config +Committed revision +This has the benefit of releasing all of the key collation from being bound to a single core in a fairly intuitive wayThis would also allow a mapping onto the current database delayedcommits with a write causing the nodes in RAM to all be flushed or not as well as saving on the upper node duplicationsOr something to that effect. +Opened for that +Almost there +If everything looks ok well commit the code and itll get picked up in the next release +Gregory Shimansky reproduced the bug +ok you guys have choices +Compactions are very important to get fast and with optimizations in this issue we could potentially get rid of almost all KV disassembly +Changed fix version to match the version on the branch. +applied the patch +Still needs Release Notes thoughWho would provide the More Info for RN? +Can you either move the existing list or create a new one +Definitely will look into that +Can we run commands in setup before execution of the binary +Voted and released. +Moving under +Im rolling the unrelated changes that were checked in on this ticket from both and +Hey Christiancan you confirm this to be working in? Attached I have a project that contains a package protected Implementation +There was talk but no concrete proposalpatches put was some discussion on the scope when moving to JDK but since noone has done anything concrete Im closing this as fixed +Hmmm so this doesnt work with trunk due to all the changes +As the original reporter I recommend this JIRA to be closed + target findbugs warnings are suppressed +The functionality was split into different classes +Revert the patch so that test continues to work but the lang suite goes back to choking in certain environment +Chris do you manage the patch review or do you want that I take it +Hi Peter I am not saying that your patch is breaking anything +Also there is some risk to making such a change. +Making the endpoint an instance variable instead of a static member in fixes it apparently. +I have it on my list to try to set up like I did for the Maven build removing circular deps including the new one introduced by the codec moves +Patch which allows to specify the options via the Context object +Should have all the files now +I also moved the new tests from to since they were getting big and to be consistent with which I modified to be a junit style test +Looking forward to +Once that is done I will be creating an to be able to persistload a release descriptor tofrom an XML document + Alternatively we could just set the initial size to the default as the resize logic shouldnt really affect performance to any notable degree +So the spec wont be broken +SVN patch for the +Ill give a patch latter +The reduction in the maximum amount of heap used for messaging during the life of an application is quite large +I think its actually due to the way the test uses JPA and +Not exactly a NPE but seems to be the same problem +I will try with this one first +Assigning to myself for M as there is a possibility JRMP and pooled invokers will be removed +Committed some interim work but still needs to handle nominals across files in a directory and needs to output the mappings etc +Committed to trunk and branch +The uses ascii hex to encode long int float double I like this for debuggingThe mapping file is similar to hbase except for table setup +Go ahead and commit +thanks Bobby! +Thanks again. +Patch to support trafficserver startstoprestartstatus under OSX +Yes JacqesIts working fine +Added JS to handle buttoncount sizing instead of reconfiguring how button count function which has the handy side effect of preventing the button from being substantially smaller than the like count text above it +If having two sub modules is causing too much difficulty for no benefit then to reducing them to one +its better to let the regionsize just grow. +Just need to add proper unit tests! +Separate the fix for the specific problem of and that of to make the other issue independent though the two issues share the same root cause +v replaces v Add with its implementation +Fixed by svn revision . +Trying again +For convenience heres the file that its expecting +Hi KenSure any patches would be appreciated there has been a TODO to improve NNTP client performance for a very long time +Stuti Awasthi on the list confirms this fixed rowcounter for him +Another suggestion is that we need to document the requirements for this test +It looks like a very nice improvementWhats the purpose of failed flagsI am going to to set it up QA servers for in deep forgot to comment on the failed attribute here is what its aboutfailed true when a resource failed recoveryThis is just the exposure of yet another internal state of the connection pool which may be useful for monitoring +OK please do +I reran multiple timesAs for the error messages they are like that since the path validation is done on destination first +is the same as +Ugh my bad! Thanks Shai +Yes it is and its very like a duplicate of to me +I took the first step towards a solution + +The out variable is also setPlease close if this suits your needs. +I can confirm that the problem reported by me in this issue is not present in the latest CVS version +Closing the bugcommanddeleteidresponsejsonsessionkeyXCqbQ +Rebased and modified original patch to use minor versioning of sstable metadata +Proposed patch for this issue +Hi David I see what you mean but this is really a special use case so Im not sure that well take time to cover it +I changed the listenaddress and rpcaddress from to localhost +Ill add docs in Confluence if the example is integrated in Camels codebase +Thats great thanks +This is not product issue its test machine issueNSAPI is working well with SJWS instructions for SJWS will be sent to. +Kurt we are still waiting for the jBPM fixes to the hibernate issue as well +If we do not fix this issue I think we should close themMarshall what do you think +The parser code is much more readable now that you have abstracted the LOOKAHEAD logic into a separate method +Veriefied on build CP CR + the patch +Easy to fixIssue Ill provide the proper svn diff filesIll try to simplify the algorithm so that it uses more VFS classes without using extra classes unnecessarily +Should we convert this to a doc bug + Not sure how to deal with this nd patch +Sorry Im not proposing we commit the scores +My company embeds Trinidad and would like this fix so that we can operate on Firefox +For lowering to in the test configurationThe discussion about raising the memory limit in seemed to trail off into ambivalence + to classloader stuff and behavior refactoring without tests arent prove of concept revert my change with thread class loader backhere is fragment from our discussionvy +But it is explicitly defined in similar issue shows up in an ear that that includes an unpackaged har archive since the jar entry name has a trailing slash because it is a directory +First the documentation should describe why only the binomial case makes sense for ranking and secondly the code should detect and complain +Remove fix version in preparation for release. +Allow as long as not at start of name + +which is a line added by my editor +some code has been checked in so this should be moved to In theres more work to be done for this issue please or open a sub task for the remaining work +Rebased for trunk took +Not really sure what to do about the no test +However the framework provides alternate means to the same ends by simply specifying the view resolvers like and do expose additional properties beyond simply providing a default for the required view class +A test case that can be used to explore the issues raised here +Ill add a page to the wiki shortlyAs for testing this is actually something Id like some direction on +I think that the patch should fix the failures +If documentation needs updating or something I can probably apply a patch somewhere? +Made changes suggested in review +adds checks and fixes conflicts with latest trunkIm rerunning tests now +Sample applications are to be used with Jrmes observations are based on the results after applying the +It will be removed once we agree on what the code should look likeWouldnt it be sufficient to use the existing journal and cluster node interface to get changes from a given revisionIm not aware of any such interface +I think the appassembler plugin has another generator for to to +My vote for the label would be Total excluding tax or Total Id prefer to see the exTax total before the grand total +Any interest in working on themOf course +for the hope this gets marked for +It might be a good idea to fix them and check them in if they are broken +rebasing the patch with current trunkThanks Sid for the review +If there are multiple created at contiguous timestamp collision would be happen +The test waits seconds but in this test run the recovery manager never invoked recover on the blacktie resource +Why do you need? You can add the configuration in as and +BTW is it possible to implement functionality that would start jobs that are lagging on nodes that have completed tasks like google does? for example if your done and the last jobs are hung because of bad hardware slow response or failure and have the ability to redo the long running jobs in parallel on alternate nodes and complete the first one that finishes? this way if you have a huge crawl and certain nodes slow or fail those jobs can be alternated on completed nodes to try and wrap up and terminate any dead jobs when donehope that makes sense +Is the password is literally +I think the splitting up into and makes sense +Changed from Bug to New is also about +Yourewelcome to pick it up and port the patch to trunk it shouldnt be too muchworkMichael +I can appreciate using URIs but seem like a strange choice +This problem is a project issue and a development standard +Native of course +We can find the inMeanwhile I checked that the prefix constants in and are already marked Private +The patch looks good but the API change should be reverted and the indentation fixed +Can you explain what different behavior you needI think that if relying on equals for cri and subject is not sufficient to assure matching then we need another pooling implementation that supplies all connections in the pool to the match method +The file is an implementation of Eugene Myers difference algorithm developed at CS Systmes dInformation +Interface +I saw this again today +I wonder if simpler check is possible when selecting +misunderstood +bumping to Beta since that is where these should be verifiedresolved +Yea something like rsync or Sqoop for file systems seems more appropriate +I just tried and could reproduce this +You can use those from m cant you +Any ideas +JACC module added that implements JACC for Tomcat +CarlosTheres an explicit reference to junit in the stax pom +These dependencies were already being identified but the information was being thrown on the floorM javaengineorgapachederbyimplsqlexecuteM javaengineorgapachederbyimplsqlcompileM javaengineorgapachederbyimplsqlcompileAdds logic to hammer the dependencies into +On exceptions such as this that clearly fail repeatedly an exception should be thrown and then logged +We shouldnt see this any more +Are we sure that is always available on all platforms for all versions of curl even the old ones? I would prefer the test to use if it is available and fall back to the old way if it is not +Users will need to register in Xircles and set their JIRA usernameThen you just add them to the project in Xircles and this will be reflected in JIRA +This patch fixes the issue and the server appears to work fine but the tomcat fails +like the matrix following STEP explaining combinations of and valuesFor the license header formatting is offFor license header is missing +Committed trunk revision Committed revision +Rather twisted +I just and had the fail as well +Attachment has been added with description patch has been added with description file transition resolvedfixed bugs to closedfixed +MarnieAll though we discussed I do not think we agreed on a concrete proposal yetI believe there are still some open questions that didnt quite get resolvedI am also working on a proposal which I am hoping to get out by tomorrowSo Id appreciate if we keep this JIRA open until we get some sort of agreementRegardsRajith +Back after a clean and a new shell +I was going to open a new bug but found this that I believe is the same bug here is a simple test case that shows that the suite method isnt properly test will wont execute any tests with this project +I am looking into the test failures +Please give it a try with. +This is a duplicate of am experiencing this issue too and I agree that either The code should be changed to agree with the documentation +Id rather leave to the application to decide because of the following argument +Let me know if you have more questions if you have some docs to review feel free to send a link. +Instead it is using user settings in home dir as I said in my last comment +I suspect it has been fixed by. +Thanks for feedback +Just completed the implementation of the Element Traversal specification +After some reverse engineering I found how to solve the problem +This PR attempts to fix that as well +integration is done as part of release +so they can be done in the platform integration code of jbpm +Ill and confirm either way +Can you open another issue please? +Committed patch to and trunk +Assign to during tidy up prior to because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Since Harmony has moved to the new charsets implementation this issue is not reproducible anymoreLess memory leak thats cool +trunk all tickets +and deployed SNAPSHOT versionThe modification concerns the import of the useradmin package which I marked as resolutionoptional and moving the inner class to the end of the class filePlease close this issue if this is ok for you +Let me know if these assumptions are not correct i will back to previous semanticsI will fix the doc and tests +Make sure it includes in all commit messages +This issue was raised from comments from my team and I did not check this bug myself until nowWhen analyzing the issue I have realized there is no such a problem and there was a misunderstanding with the schemasMy apologies +Attached WSDL that has a SOAP Header my suggestion to Barry last week was not to generate the separate request and response procedures that will get in way when the message vs payload is selected but generate one single procedure that contains the code for all the request and response and wrapper +Oh think I was clicking wrong thing meant to just link them +Could you guys give a short description on what the input and output is? Seems this is all based on text and mahout is using document vectors as the input and a configurable distance measure +Duplicate of +with +MSVC compiler support it very sensitive to header order and forward declarations need small rewriteThere is a patch that includes previous patch and fixes those two issuesUnfortunately I have problems with porting of buildmakefile to current trunk too many changes against version for now I leave it partially patched it may not work +The upadated patch is attached +Please also see my last mail in the hadoop dev list +Ive fixed the original issue but I cant use the code attached without the author signing an agreement +I think we need to move away from checking in any of the autoconfautomake generated files +Can you check trunk? I shouldve looked for a GA tag when I did it +This will be fixed and a rolled out ASAP +This test uses nested predicates +Much of the work has been in the cleanup of and friendsThanksEli +The latest patch looks good to me +Thanks for the patch! +First is the and second is +META +Fixed in trunk +Hi JinwonSince that is an internal JPL issue that question should be directed to +Added a test case for this expression. +Hi Jody I had sent Tyler Mitchell a couple notes asking him to clarify the need for copyright assignment but never heard back from him +also to +In short iptables is not smart enough to allow access to and block webhdfsv +Using this configuration file for Tomcat makes sense since it has retained compatibility across major releases of Tomcat whereas the container classpath hierarchystructures have significantly evolved +Duplicate of the issue +Registering a Timer to commit the transaction is probably not the right way to tackle this problem +I can start publishing the php client as to and then reference this on for composer usage +Just some general sanity review would be good +The attached patch removes the code that initializes the field JAVASQLTYPESBOOLEAN in since its always initialized to +the dstPath is in the testdir and no one set the rootDir as far as I can see +Karsten thank you for your help +Apparently it happens after timeout is reached +Patch is committed. +Please fix these things +Hi RobinAbout month ago Ive sent the patch for Dacapo to Steve +Im going to do more testing tomorrowAnd I need to remove my obsolete comments +Patch looks good +Vertical inheritance tests patch and failed tests output are attached +OlegI have much respect for the bold move to do a rewrite +I wasnt able to merge it to easily please merge it in if youd like it in that branch +If the upgrade drops column it is necessary to create all the depending views again +Attaching first rev of a release note for this feature +the newest patch does some minor cleanup +Hrm +Cleanup of unclosed issues. +See the target in Julies comment +This bug relates to the old Chemistry implementation that has been deprecated. +can confirm that this is a bug +My fixed. +The original file needs deleting +Verified JBDS Beta +Thanks BrentResolved in at revision +Great patch +Many thanksI gave a first overview and it looks good +I guess its best to rename to MinimalDefault as you guys suggestBut this problem should be solved +Sure on single patch +What about pages which are instantiated but not rendered immediately +Ive also checked the patch and I do not have any comments or objections +disabled the file protocol in r until someone implements Local protocol +Change to wait until client actually observes the disconnection +Ive committed the fix +Good idea patch attached +Ill be back in the office on Monday Dec +Ok +The manifest info is now created in an appropriate place +Allowing resolveresult for naming contexts obtained through managed ref note the fix allows this use case to go through yet it does not introduce any change wrt msc dependencies which are globally under discussion in the context of +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsernodes +Something is wrong with datadir instructions +So if we reuse the same API for this functionality it is as good as registering a fresh ISO +settings srcdocsbuild libexcluded etc +I have already attached detailed trace for this +Should be applied to coretest but again its an Eclipse patch so may need a little tweaking +Thats simply just not meAlso this ticket is about an issue on the Mac platform +Regardless its a matter whether we want to correct or continue to follow the incorrect behavior +Thanks for your work on this and the patch it helped +how necessary this functionality really is for my use cost if the functionality is absolutely required in order to have correctness then I think the cost is OK +Seems to work +Would it be good enough to simply iterate over the lines in the trace file counting occurrences of RDBCMM and RDBRLLBCKI modified your repro adding some more commits and rollbacks +Default behavior continues to remain as earlier wherein no properties are inherited +An informative message will also be logged. +Shouldnt be used simultaneously with binding or type attributes +Please do not issues if they are be closed without providing proper justification and agreement to +Btw in future if you mark Patch submitted as Yes in the JIRA were more likely to spot it +In that case shouldnt the current patch throw an exception when the current volumes full rather than calling get +Thanks for the review Tom! +Thanks Andrey and Denis +Please seeBrock +This should be a case of extending the module so that the client selects an appropriate mechanism from those advertised by the server from the httplocalhostrestsasl +Ludger thanks for your contribution +Thanks Juergen and tested the snapshot it types like Resource work in Collections and Maps as well latest fix will be available in tonights snapshot +merged. +A big to leaving this in library code rather than framework +should be fixed in svn but Im having some trouble deploying a new snapshot +rebased patch +It however cannot be distributed as part of the drools distribution as well as that would result in a cyclic dependency in the build. +Before we commit this we should run some actual workloads +Committed to trunk with revision +I updated it according to your comments hope I didnt forget anything +bq +These issues were fixed in and have been released w Lucene +As preemption is a YARN feature wouldnt make sense to have Preemptable as a YARN annotation and have utils classes that help an AM to do implement such logic? By doing this we could use this in the AM itself to implement AM failover recovery +Im good on keeping resolved. +Fixed in a series of recent commitsTo summarize I added a more generic class for tracking all Closeable resources associated with a +I have been studying the code related to and +Pretty strange +Now you can try with adb +The forks must use different variables to remove the contention with actions after the join responsible for merging the data +Otherwise it simply makes the tool look poorly designed +the maximum size that will be stored for a given task +Rebecca Newton incorporated these changes as part of the Security Guide for +I need a bit of clarification are you adapting the WAL to log to HDFS via appends or are you working on a mechanism to shove the logs into HDFS once they are complete +how about submitting patch to allow the plugin to override the default javah default location? with test done +It makes cobertura reports pretty unusable at the moment +looks good +I had to destroy two failing nodes and only the first one has been terminated +The content type is changed according to the SOAP version in use just in OUT messages +I think Exception name in rename to +included the Dependency manifest entries in the core richfaces jar + +nochecking openssl usability +Really appreciate you for the help +Youre wrong edit this file then execute the script to create the WAR file +Nothing obvious in the logs why these tests fails +Sorry the second sentence in the above should read This test starts a set of Quorum Peers and then shuts them down one at a time and starts replacements for the ones that were shut down +If we could also put from first and to second that would be grand +Attach a patch +It seems to me naming a Vector is very very common +Commented back +Not properly understanding both and their relationship made this appear like a bug with no help from my use of randomly generated dates. +Not making +Valid point at this point its just a thought exercise and this ticket canbe closed as a dupeSent from my phone please pardon the typos and brevity. +Committed. +This is a language specification issue +Will close after weve seen a representative set of successful builds. +Grant do you resolve or do I +same for tags you will be able to use the overwritten tag inside a new perhaps i was not clear the example code above i have named the function whatever but the template where the function is stored is called so the name of the function does not matter only the name of the template the template i can use whatever but outside i can the function only call with the name len +I believe Im done with this panel +New patch include test release audit warning is caused by a new golden file +close off release +rebased to trunk +Closing all resolved issues due to a successful release +I mean keep purging that cache asset +What do you think +Current trunk does not enforce the rule that the same consumer cannot receive the messages againResolved at r +Check the comments on re documentation of +is probably more correct but that describes the implementation rather than the intent though in this case that is a fuzzy distinctionFinal implies that the value doesnt change which is subtely different to not being able to set the value from outside +If you are up to it could you create a that would accept lines of input and provide lines of output to test this? It would make testing the shell a unit test which would be much fasterfrequent +should i apply them too +I changed the description and title to reflect that there are ways that this condition can arise +Committed to trunk! + +We have applications that create and teardown their own Flume configuration at startup and shutdown which seems to exacerbate the issue +If not then the upgrade procedure needs to be documented in the release notesIf there is no immediate demand for the upgrade I would rather wait until Jackrabbit is branched before doing this +Please reopen if reproducible case becomes available. +I just committed this. +Looks good the issue with the modal windows is gone and I havent noticed any other side effects so far +Could you verify this by simply restarting the Vaadin UI +This asymmetry seems to bug me for some reason +Need to come up with the basic requirements for the property + dont want to change the default but want to get rid of the default value at allconsumers of the Oak class should explicitly indicate which security provider theywant to use +But the problem is still there +Committed revision . +Ive merge the into and gateway has been released withThis JIRA was continually used for additional bugfixes and code refactoring +seems done +Reorganize the Geronimo console have been completed with +fix in +Also persisted after reconnecting +Uwe Thanks for creating the jira issueCan you add some simple query examplesWhat would be the lucene Query objects for those queries if it was produce by a QP that supported that featureAlso elaborate what is the current expect behavior for those queriesIf you can write a junit with one or indexed docs and a lucene Query that retrives just one of those docs and not the otherwithout using the queryparser that would be helpful +Verified by Mark +Patch for earlier file had some extra junk in it please use this file +If someone wishes to address it in the future they can open a new issue with the more appropriate solution +Therefore from that point of view the soap fault I got is also acceptable. +I will use your patch while checking in +You could have done also by using a common parent and declare the dependencies in a pluginManagement section once. +activates per default +Giraph is about to graduate to TLP status would be great to see this available in the next release +Can I suggest you transfer your attention to that issue +checkin broke some unit tests and cause others to hang +Using the output committer will mean that the temporary output is in the output dirs subtree and hence the rename works +I think there is only one way create new configuration and delete old one I do not sure is it a right decision for the least should be a way to setupedit such hiden properties for configurations. +I really dont have time to look into it right now but Im sure whatever the problem is it could be solved with a minimal amount of debugging +It seems this has been fixed in trunk +Ill make this change when I next update the patch +Currently each sub searcher was its own context now we are just using IS as an executor of one leaf but still being aware of the entire contextAll tests pass +I created for this. +Will fix in +works fine +This is something I really believe we should address for +to demand the a proposed capability token youve described +Ill create a task to emphasise this in the release notes +Ole I can you handle this +promote this to just committed this +Please have a look at my latest patch which includes an addition to the Castor test suite taking some of your original tests and moving them to the required format +bulk defer of open issues to +Being a new feature this should go to +is closed now and as per the latest update the committers of Thrift should have access to publish to the snapshot repository +Itd be nice if error messages could contain the version of the code they belong to But as specified in the Issues environment description I use maven as packaged in Ubuntu also just verified that this Ubuntu did not gamble too much +Now that work is underway with AS all previous community releases are +Could you please post the error that you saw +Is there general interest in this +So this wont be available until hiveThanks +How on earth did this get unnoticed for so long +and have been made to change that map +This seems to be a problem with Array not invoking on the subclass for some reason +This sounds like a good thing to clean upThe attached patch changes the constructors so that they take a instead of a +Ah I forgot doing it +an array of Scorers that match the disjunctionAt the same time the score computations of the subscorers could be delayed until an actual score value is neededThat is a rather drastic approach but the specialized queueheap of might also help performance of +bulk close of all resolved issues as part of closing items. +add jsf capabilities at least need to be possible. +The title of this JIRA used to be Potential deadlock with Renaming to be more in line with the solution which is that and should not attempt to deal with concurrent access to Fqns in the cche loader the cache loader implementations themselves should be thread safe +Looks like even with the patch we are running into this issue +Ported the change from to trunk +Yes the aim of this benchmark was to see how large jobs with small tasks would be affected +Proposed SOA TP Release NoteMSSQL Initialisation of example data has been removed +Shall we change the Parent class or just modify the? I dont know why writes a tmp file for every tupleIn our own code base I just modified the because I dont want know the how the subclasses of gonna behave if I modify the whole spill logicI want to contribute to thisthanks +I assume the patch fixes the problem +Ive never seen this technique used in real code so I doubt it will be missed +Aside from understanding this trivial detail everything else looks great to me +I see so the problem was when there was no security config specified and you tried to add one? This is fixed in r +As we know currently HDFS trunk can not fit for requirement of +As for the datanode directories they were already empty +Think about files which will be uploaded in chunks where the total count of bytes is unknown and expect files +extends and extends updateSequence I need is the one included in the request and the WPS object does not have it +One small nit +I am working on it +I ran the test cases and they pass +fieldName description notNull sizeMax add field string class +The attached patch is the trunk rev +Patches have been applied and appears to work ZararYou are right I havent tested it with the spring AOP scenario at that time it might have some problems in this situation well as youve reported another issue we will track this problem thru that JIRA. +Thank you for reporting this. +Heres a patch +Right Ive looked over the JAXB default binding and that works well if your objects are already annotated as such +It looks up html META tag and stored the pairs into metaData map then you can index the info +Closing +Maybe a good start would be to just delete the deprecated styles from the style sheets which will break the layout and hopefully compel everyone to use the correct styles +Thanks CWAlso thanks John for the review. +its still existing in release and snapshot +Do not delete this issue +As long as they are unsatisfied without the required API key they do not really hurt and if users do have the required key it would be much easier for them to start using themIn any case we need to validate that the Enhancements created by such engines are in conformance with the Enhancement Structure because we can not aspect from users to adapt there clients only because the are activating a different set of Engines +orgapachehadoophbasereplication +rev +If it turns out its fixed we can certainly add the test after the fact and backversion this but for now it remains open +Thanks again Nick +to and trunk +Retested and +So far offsets are returned in ascending order +test passed +was trying to see if reassigning would bring the close button back +That is not rational that describes why cleanup is no longer requiredThat is true but it also doesnt say that +Could you please attach the generated wsdl +While fetching CMR collection we want be the first and be the second +With the attached patch the performace comes down to ms to perform parses and ms for to perform evaluations +A mismatch occurs when running with JDK serial error for current Im going to add this to the badPackages for now but should we consider having +Added Javadoc +if we want to continue to cache fully rendered pages then we will have to rewrite the authenticator to use something like AJAX so that we can make the html generic and have it call back to roller to get the authenticator info +If we move trie into core what do we needwant to fold in from Solrs +bq +The id in needs to change form long to String +Tests passing in hudson. +Any chance that can get bumped up quicker? +Instead it should be loaded from the bundleAs a temporary workaround I added to +Thomas sorry this has gone stale +is finished +This is attached to relaying +Attached patch has a testcase that tests the Hive client to server communication with SASL on delegation tokens +Instead I have imported all dependencies form +This will simply mean that only a user who has locked the node can update or checkin the node +I just committed this +Hi JessicaI added this part into install guide commit id bacecebdddbaeedcOnly one question Setup Best Practices refers to Linux NFS on Local Disk and DAS and Linux NFS on iSCSI which originally as a example configuration in previous document but now missed in the lastest install guide +Second and third remain the sameCan you figure out what is happening? Ill try to do the sample as soon as I can +Need a flag or attribute to set whether can select directory +Explicit handling of the Exceptions is needed from the XB layer +However the AS based runtimes does not work for me as well investigating +They are not even showing up in +find other mechanisms to time out operations that allow the thread to gracefully kill itself rather than be killed from outside +Thanks a lot +Verified this with latest build +fixed in only and there are conflicts with cherry pick the fix into master + +fixed +Username sreich +In fact we should also relax the ticketreview requirement particular for simple issues that have no bearing on any of the discussed semantics +Without proper use case. +No problems have been reported with the committed patch so closing +Attachment has been added with description Big table spanning has been added with description Example FO using latest Trunk on Apache FOP Version was originally attachment ID that was lost as part of the data loss on Note that this attachment ID has since beenThe original attachment comment wasBugfix for has been added with description Bugfix for has come back in apache fop version cant reproduce the issue with the attached example +The new escaping syntax introduced in now helps with this issue. +Let me know if you think we need more. +on removing the comment about allowing nulls +With the system dir this would not be possible +Created account and sent email to James. +Please explain how to reproduce +looks goodThanks Cosmin +depsFiles? cpipesMakefile? cpipes? cpipes? cpipeslibtool? cpipesimpl? cpipesimpl? cpipesimpl +Just change it to and it will compile +only controled by timethe edits file is very smallwhen the system is this caseit produce a good many of edits filer root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr r root root Apr +New auth patch that correct paths now that everything is going under common +What is your ram set toWhat happens if you commit periodicallyI suggest you take this so lucene first +DOMAIN elastic load balancer hostname echo ip agent +PR merged +fixed in revision +hey Jesse thanks reattaching EvanThe failures are suspicious +Another remark is that Freemarker already has static methods to escape HTML Javascript Java +Thanks Kevin +The exact same parameters works fine with HudsonLast but not least After the exception the smpt host is set to a empty fieldFor information im behind a proxySo this in my configurationbut my smtp host is and doesnt need a proxy to be accessibleI have remove all proxy information but i have still the same exception +Ill rerun with the latest patch +Is this patch still is any shape to be committed? +It still takes them time to analyze the issue and research a workaround +Resolved in trunk and as part of. +Create another JIRA about this and lets close this one if its ok +These issues have all been resolved and just need to be closed. +can you provide more info or close if you dont remember? Im guessing since trunk has moved to a new release manager version it will need to be +We started a Wiki page for figuring out the best practices for releasing using m +Im uploading the latest version of the patch +Heres my stab at it +That does not mean that fire and forget means using a registration as the local instance is illusory the resources it depends on remain global and it does not solve the resource collision problem described above therefore even if one wanted to solve the problem through the database object it fails to do so and represents a poor contract as a re strange users of it that continues to be a thorn in my side +I want to emphasize this issue again +One change between and was the of the release profile in the release pluginConfiguring source and javadoc Maven plugins should do the trick +Just a blank mail is enough an it will add you mail id to the users list +Changed the SubjectI guess now allows to change the file name depending on input from user +I committed the patch to trunk andThank you Arpit! +Havent forgotten +It is hurting my brain because it would be best to simply have a shared context object that represents all shared data between but this probably will not work as it would move variables such as deletedDocs norms and the writeLock +Here is the patch with fix for all the links on this ftl +Closing +Mike I want to commit this patch and let is bake in a bit on trunk and x any objections? I will remove the DV use in x on trunk this is not a problem +This one is somewhat interesting because Flickrs authentication is but not exactly +In my opinion a natural next step in the direction of improving the css support in is that of changing the code in the class so that instead of rendering the single form definitions by html tables just use divbr etc +Also FYI here are the configs to add to a server to get csv reporting enabled on startup +I am also able to reproduce it with two distinct nobody can reproduce this issue with ER it should be closed +We did the same change on and ran it for a day under heavy write and read load on the test clusterThe unittest included just reruns the with multiple number of reader threads +Nevermind I have fond it and added it +Expect a new patch fairly soon +Im thinking virtualbox snapshotting revert to clean state easy fast +Pls grab the latest version of this from HEAD or Branch +Ive been doing a lot of manual resetting of data in Kafka and one thing I noticed is that clients dont always behave so well when I do that +This is an old one +commited in rev +LeonardoI will not push existing and potential validation users in a wrong path +Please dont use Bugzilla for asking questions +Although I guess if the words are returned sequentially you can easily reform the phrase so that works tooAll you see from Googs is their frontend so who knows what their spell checker does + for patch +In order for Eclipse to cleanly recognize a revised SNAPSHOT version of a Maven plugin in regular builds or incremental builds Eclipse has to be restarted +Thanks Knut +If anyone can give an insight to why this might have arised might help others in future to get an idea of what they should avoid or how to get a permanant fixRgds LarsThanks for ur inputActually we dont really have a Master slave set up as such +Works fine +Otherwise you need to dig into the build script to find out where and how the files are being copied or moved around +Attaching patch for Y! distribution +You get a from me +Hi Davidhere I need your help when you have at disposal a minuteFor me this is a configuration bug in case clientAuth in is set to want doesnt work with SafariSo probably its a default value set for go directly in production where a SSL certificate is installed on the server and probably its correctly to set it to wantIn a test environment probably it has to be set to false and a warning has to be inserted into production guide to set it to want +It also seems awfully recursive otherwise +Heres a patch that implements the first two suggestions +It was either fixed by something else or user error +Now that I think about it it makes perfect sense that the fix solves your problem +Ideally I would rather that none of the code has to consider invalid cases +Issue may already be resolved may be invalid or may never be fixed +Deepesh + issue since we will be Redback for additional changes to CSRF and XSS vulnerability fixes +but full versioning not yet implemented in all spi impls +Shouldnt do +I had the same thing in mind +Replaced the patch version +Is this Issue a Blocker or Critical priority for EAP GA? If yes +oh +I attached the surefire historical chart for one of my the parent projects +can reproduce +If an annotation is good enough for you then I suggest you fill a new issue this time an improvement and describe there exactly what you think is a good way to implement it +The solution is correct but there is a performance hit +Apart from the service implementation no java client code needs this method thats why I removed it +HiActually Ive not mentioned it but we also use Spring and this is the issue +Attached Maven project required for the reproduction steps +asc +If you finish it before we release beta please correct the versionThanks +This issue exists in but I am unable to reproduce it in and +I dont imagine that it will need updates very often +The same about I cannot see any commits in tests +The font is fairly closely matched with the Apache DB Project Font +The advantage of this design is that it makes leasepipeline recovery much cleaner +I did not +Checked in changes to use and let me know if this works for you +Removed the commented out unsupported sequencers from the configuration has its own configuration file in the build repository that is deployed for the product +Moreover it is relative to +to trunk +The patch applies on top of +Currently custom HTTP headers which start with a lower case x will have an additional X example will be converted to when sent through the outbound gateway +New bench is testbenchlanguage +Will commit if tests pass +After job tracker restarts all the information of the connected tasktracker goes off + +I dont think we need to do that in +Only the field is packaged into an aggregate document +Attached UML for style interfaces. +Attaching +I should also add that this would be the same for any error at deployment time and not just validation. +Fixed with a test +Hia wouldt it be better to fix the jsf apis instead on both implementationsWindows mobile is a problematic issue generally due to the browser being somewhere on ie level but as far as I know the xhr Object is there so it should be fixable by someone who has the resources to do the winmobile testingIt probably just boils down to a few fixes and special hooks in the codebase to get it up and running +Do you think we should fix only these issues as blockers or is it OK to leave them as known issues? +Thanks +Ive committed this to trunk and +Last couple redirects were created and checked before the beta release. +Hi Im facing the same issue in project +This may just be a case of ignorance on my part and not a bug at all +Trivial patch verified in some unit test logs that the exception wasnt logged anymore +Attribute width works fine. +I havent tried the head as Im stuck on for the time being +Is this still the case Norman +Looks to me like what is needed here is just +I just committed this +Committed patch to documentation trunk at revision +I think it is very important to determine the proper number of fragments assigned to one worker at a time and relationship between those fragments + +No answer +Update JVM was GB not MB +The attached patch adds getting of a read lock to get +This kinda already existed as does that do the same thing? could these be merged then pulling the class into a common location and combining aspects of bothKinda existed but takes a sample rate which cant be known if you dont know the total size +Created reviewboard +From what Sun told us the last method being compiled by the JIT are the one we mentionned in our previous comments and coming from xalan +Should I merge this to branch as wellYes please I dont see a reason to not do so +Ill submit a patch for this tomorrow +It would be good to get someone with more language expertise to look at that part +Serialization between VM and VM of class A is ok just that A will not get the new fields +ensure in clients appears to have the same issue as ensure in +There does not appear to be an ESB issue yet linking in the SOA issue + added tweek for main tweeked Alejandro +But it would also easy to setboth I guess I you know whereregardsChristoph +lining up for any chance you could produce a simple junit test case that just exercises and multiple threads +Reduced priority of non critical issues which have current is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Essentially the same as earlier but supports ant mvn forrest now +if it gets started at all +Do you think this will happen +Bulk move issues to and +Im suspicious of that those errors are from this patch +It actually not depend on if you import second project +Run the maven tests +Im going to mark this bug resolved and you should open a new bug for the new now that were pulling the strings in as the source encoding we still need to get them to Java as the correct encoding when Java needs a unicode String +Please take a look at Deravajs patch +In our CI environment there is defined so thats why it is working there +Unscheduling +ant will see if he gets a clean build against the current trunk +Removed explicit dependency on xmlsec +Patch appliedDIGY +Logged In YES useridThe same applies to Username pilhuhn +Integrated to trunk. +API refactoring didnt touch this listTemplate api. +Yes there is a test in as user sam try to do on table that is owned by SWIPER +thoughts +The client code is very similar to the one Ive submitted to JIRA service is quite complicated +for +Have you used another version? This cant be solved without your active investigation +In the end seemed to be mostly because of missing Exception classes and reformattingIve applied the to M will open another jira for trunk to align +Related to perhaps +If the senders are consistently sending messages faster than the receivers are accepting them then you will inevitably hit the limit at some point +perhaps the submitter of the issue is using the published schema beta +Certain combinations have special meaning +I was asked to split this patch +So I guess this change is ok and does not break +This is a problem with indexing +So the end result of all of this is that in the example above all three requests end up ultimately making the same last request +updated for data +Upon further investigation updates appear to go through even though the API does not report success +Looks like this issue is making the resulting feeds invalid so its quite important to be fixed soon +No I havent got it working but I didnt try much more as I hadno idea what I might tryThe X server is the one delivered with Solaris more or less outof the box +Added proper queue property for cluster config +Upgrade to to match the ASF Jira instances +Thanks for the patch Johnny +Hi FabriceCan you point me to the source of next version please +Cool thanks +The only database I know of that is not allowing null values in unique constaints is DB on windows +Sorry I havent noticed youve started on thisMy apology for the duplicationWould you mind if I ask you to review the change +Would anyone want to try the patch +However I hope that in a horizon supporting the Java formatter options will not be forgotten I like Eclipse precisely because it is uber configurable +LONGNVARCHAR is as so. +Patch v corrects logging +there was some problem with the that I was using +Jan please verify +Also renamed the template to fsglusterdefaultname so that it will not be used when HCFS is not being installed +The recipes depend on the libraries being in srcc which only happens if you run make in that directory +Attached test case to reproduce the problem +Ive just committed this +Ive just quickly tested the component over the and it correctly generates the model +I think it does require one though since trying to specify the encoding via the XML prolog has the undesired side effect of causing IE to go into quirks modeIm not against the suggested change +Latest stable or latest unstable +Well pick up the better wording if we need to build a second release candidateThe JDBC version level when running on Java has caused confusion before +Solved since the core handle arrays of event listeners +However the bean component is for invoking beans which in case they have Exchange as parameter they get the real exchange in use +I figured that much what I was suggesting was that in the information on the FAQ on the website include that line as well as the line about the serializer that it already has but I failed to communicate that clearly +This looks ok on trunkflatheadtrafficserver jpeach ls l r root nobody Mar flatheadtrafficserver jpeach ls ld bodyfactory nobody nobody Mar bodyfactoryflatheadtrafficserver jpeach ls ld bodyfactorydefault nobody nobody Mar bodyfactorydefaultflatheadtrafficserver jpeach ls l bodyfactorydefaultr nobody nobody Mar bodyfactorydefaultREADMEr nobody nobody Mar bodyfactorydefaultaccessdeniedr nobody nobody Mar bodyfactorydefaultaccessproxyauthrequiredr nobody nobody Mar bodyfactorydefaultaccessredirecturlr nobody nobody Mar bodyfactorydefaultaccesssslforbiddenr nobody nobody Mar bodyfactorydefaultcachenotincacher nobody nobody Mar bodyfactorydefaultcachereaderrorr nobody nobody Mar bodyfactorydefaultcongestionretryAfterr nobody nobody Mar bodyfactorydefaultconnectdnsfailedr nobody nobody Mar bodyfactorydefaultconnectfailedconnectr nobody nobody Mar bodyfactorydefaultconnecthangupr nobody nobody Mar bodyfactorydefaultdefaultr nobody nobody Mar bodyfactorydefaultinterceptionnohostr nobody nobody Mar bodyfactorydefaultredirectmovedtemporarilyr nobody nobody Mar bodyfactorydefaultrequestcycledetectedr nobody nobody Mar bodyfactorydefaultrequestnocontentlengthr nobody nobody Mar bodyfactorydefaultrequestnohostr nobody nobody Mar bodyfactorydefaultrequestschemeunsupportedr nobody nobody Mar bodyfactorydefaultrequestsyntaxerrorr nobody nobody Mar bodyfactorydefaultresponsebadresponser nobody nobody Mar bodyfactorydefaultresponsebadversionr nobody nobody Mar bodyfactorydefaulttimeoutactivityr nobody nobody Mar bodyfactorydefaulttimeoutinactivityr nobody nobody Mar bodyfactorydefaulttranscodingunsupportedr nobody nobody Mar bodyfactorydefaulturlroutingnomapping +be also a duplicate of? +Reverted commit +Oh instead of resolving once the patch is available flip the JIRA to me for inclusion +Let me close this issue then. +March should work properly nowFor some reason I was still getting proper backups locally even with the wrong patch I think that when you set the ios cloud backup user default it may automatically restore from the old location but the way I am testing is by moving folders around across ios simulator version app directories not by actually upgrading +is a birthmark that has be globaly unique it is fixed unless one is merging clusters + +This patch adds a timeallowed parameter which takes a time out value in milliseconds +This issue was fixed under +Reworked with review comments +my patch to this to branch and trunk +m repository and try to build? Wondering if you have some file versions in your repo that are no longer available +And that probbaly would be mean index does not exist as wellFor now this patch should definitely work +However most of the privileges in are not sufficient and the fact that dbs tables and partitions can specify custom locations means that we cannot use pure Hives enforcement of auth provider implementationI dont follow +This been postponed for many releases now +But there are more important things HDFS needs so I rather compromise than debateI am testing the updated patch and will post it shortly +As far as I know most of the test failure is caused by this reason +This way they should work regardless of the order in which they run +Brian there are three project example xml files has been updated in the zip file could you check it and commit it to repository just talked to Max it should be move to download site ASAP if it works for you thanks +Do you have the and also in your properties file in the web app? You may have to implement the to pass the right properties +I was going to update the but wanted to wait for some feedback +I was about to ask thisthis brings us two major features Id like to use parallel testing configuration on demand +I think it would be good to put this fix into the branch +Hi RobertThanks for the patch but I am not sure about updating the support class since compression is deprecated and will not be available in We shouldnt encourage people to use it with a more performant versionBut leaving this issue open can be an alternative for those who really wants to use itDIGY +Hi DamianI cloned your test project ran grails compile then grails and all the tests passCan you attach a stacktrace or something to give me some more informationthanks Lee +This was committed in under srcmainjava but if this is used only for testing then it should probably be moved under srctestjava + for the patch +William thanks again for all your hard work on the JMX stuff +Many thanks for your voteI have added a couple of comments to the classes and I hope that they are understandable +But OTOH I also like the new Generics stuffSo I think the best thing is to merge the two implementations somehow and distinguish between them like it is already implemented for python +So with a suitable default value nobody will ever know the difference unless they start changing the parameters +Is that normal +Close all resolved issues for. +Should we add docs in source package +Thanks +These are images rendered using the known portlet prefer if these images are published somewhere on the jetspeed site so that they can be linked to from the wiki. +No tests added since this is a minor enhanacement localized to the rpc metrics code path +The unit test is likely to fail one cant assume that disk space wont change between calls +And the property name in the document should be accurateI attach the patch which I changed to a current property name. +So replacing the grails jars with ones makes this issue go away. +This would be easier to predict if the total was not displayed in the Journal +Paul this issue relates to software that isnt used anymore which is why we wont fix it +MVEL has been released with a fix for this. +I am still getting wildly different tile content depending on whether line is to left or right of map center + +Then it does a system catalog insert which will fail if the matching table name exists +This has been done +Updated patch to also include some fixes to andUpdated patch to work with the latest shindig revision that is subversion revision Ian any idea when you might be looking at this? I dont want to add any more pressure but would be great if we could have an inkling on when this might be resolvedAlso if you let me know when you start I can update the patch with whatever the current revision of shindig will be since at the moment you are likely to get conflicts +This is against because this really only touch stress but Im happy committing to trunk if anyone object for +rather all the things are addedremoved by handlers on subststemmodcluster +bq +This restructures the security a lot to make the mapping much more independent of the application +Committed revision . + +If not its probably safe to ignore its unlikely to cause problems other than the noise in the logs +Here is a patch +Unfortunately not all encoders support partitions at the moment so for now tests that rely on partitions do not draw CLP at random +Attached jbpm queue config and messaging config files for both servers +The poms are incorrectThis issue needs to be addressed in the future releases in the release process. +The changes to the upgrade test definitely showed the bug and were fixed by the checkin +Updated the patch +Ive not seen any reports of this peer not authenticated error related to Nexus or higher so Im leaning more towards this being infrastructure you attach some detailed logs from your build when this you experience and strange issues when using the Nexus web UI +Feel free to switch reschedule if you feel appropriateThis is really a tuscany issue if they want a CXF binding or not. +Greg is this issue still relevant today? +My comment about the file was to bring out attention to it and upon testing decide what to do with itDoes it mean that the file is not requiredused by M either +fixed with +A new option sequenceName has been added to entity jpa command and a corresponding sequenceName attribute was added to the and annotations +It addresses the issue by making extend now which provides access to the manifest attributes and certificates +thanks. +Patch to fix issue on branch +Thanks Anty and Ravi +Added message to exception in readMapping method +The real solution to this problem is in fact the latter +bq +Converted utils to but some unit tests are still broken so Im not committing this until I sort out why +I just pushed some plugin updates would you mind trying again and see if that resolves your problem installing the plugin +Please try RC and report behaviour changes if any +George thanks for the code +Two comments Bravo for Atlassian for fixing the XSS vulnerability Mauro isnt this fixed in SVN now +Where do I fine the user listHowever as I told in the first message the generated HTML when saved as afile displays in the right way the +Resolving issue w no assignee and no dont like the hardcoding of the value workaround because I dont quite understand why hfds causes this test to fail +Patch to remove unused variable +r improves a number of things in the implementation but does not fix the problem which seems to stem from the fact that writing data to zookeeper doesnt actually write data to zookeeper +Such issues can always be solved with recompilation of your source code +All three committer accounts are now done +libs +I had removed the target and this is still in so resolving is still valid +Thanks for the patch Andrew. +Build and tested it. +If this is still a problem for you please reopen the bug report and provide additional information. +The patched file names need to have relative path +a very very easy work around indeed +That path is +I would leave classify as it stands +We set the tis in the ctor or if we load a NRT reader +In fact this would ask special support for the thrift client to not screw up the setting +Since they can be implemented as pure plugins developers wouldnt have to worry about getting their library committed +Updating affects version and assigning to Tiago as he fixed which also fixes this issue +Almost i got Domain Class Ordering for them +You did right to remove the colon from within the single quote but the punctuation mark at the end should be a semicolonconnect jdbcderbymyDBterritorynoNOcollationTERRITORYBASEDcreatetrueThanks +Can you confirm if this was actually fixed +PortalCRHA tag should include a fix. +to your suggestion +A similar problem are optional transitive dependencies the plugin now requires them to be project depends on acegi which has a lot of optional dependencies that I dont use dont need and dont want to hear aboutsuch as javamail activationmvn ideaidea +Right now this place is not exist so it cannot find from that locationI am finding file is packaged under these new places nowusrsharewebappsbridgeclassesusrsharewebappsbridgeclassesI think you have to modify the location of accordinglyChandan to unblock you please modify the file usrbin with either of the above location and test it +oh I realized you might not be able to pull dependency versions out to top level if they are not generic across all profiles in that case it is fine to keep them just add a variable for the same version used everywhere +Otherwise it will have race condition with client updated pipeline +I Will provide a complete patch +Here is a scenario I have testedStep +Verified updated javadoc +The EXPLAIN ANALYZE times are still quite short and therefore do not demonstrate problematic behavior at the momentWhat is the job doing? Does it still appear to be hung +Christoph this is one of the corner cases that cant be solved just in meclipse and is actually much easier to handle in the plugin. +May be a proxy server issue +Patch applied thanks +Id rather not open it up to all users +Updated patch uses git log instead of git whatchanged +I only made methods virtual that absolutely needed to be virtual and only made data members protected that absolutely needed to be protected +I absolutely agree with you sum up your ideaAll procedural operations must be put into the Task chapter +Close all resolved issues for older releases. +I think we need add the bpel example into both features as you said +Done +Patches also resolves dependency patch +Else you have to adopt the method used in echo sample where you recreate the payload with different parameters +done for calendarcontextmenudatafiltersliderdatascrollercolumnscolumngroupcolumn for dataList data data dataTable subTable dndParam dragIndicator for drop dropSupport dragListener dropListener menuGroup menuItem menuSeparator effect +The command line launcher could deserialize these into properties to pass into the launcher +Given that the context of any given component is its container ancestry this provides a pretty powerful way for any componentcontainer in the context to hook into the menuing system +Yes if you could post a Seam version that would be very helpful +It is just a svn mv away +Im marking this as having existing application impactI do plan to commit the patch shortly +Hi JohnQuick question where is the src table created for the +Apply AOP to internal classes but also to classes that implement internal interfaces +The biggest blocking issues are done +Can you try this with jrockit instead of sun? Also I see you are using MP +Its quite transparent to the user and admin. +Oops leaving in was a mistake +Will do it sometime today. +a is no longer relevant since a similar fix was committed underCommitted to trunk with revision +The plugin doesnt depend on so only the dependencies on and need to be updated +Can we organize the security related tests under a package to facilitate this +The cache is cleared on calls to add and remove +I think we shouldnt care +Additionally making an alias for should be thought of. +bq +Army You have a section for the JDBC perspective +Sorry for all the confusion I caused +These tests are very old +So this has been fixed as part of +Applied patch at revision because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Apologies. +Does the Copyright statement apply to the work as a whole or the file in which it is containedGiven that the NOTICE file will contain exactly the same Copyright the line seems entirely superfluousHowever I do think it would be worth included the URL for the AL pageTheres perhaps an issue withApache and Apache Foo are trademarks of The Apache Software Foundation +Assigning this to Marius along with svn access to jbossastrunk +Resolving. +This is fixed in version of the xsd. +Not sure how it was working before your changes but one problem is related toAs the terminal parameters are not correctly sent to the ssh server the jline incorrectly set the flag for switching backspace and delete which leads to backspace not working on some environments. +We will really appreciate any feedback about it pbs of course but also usage statistic number of entries and of course a testimonywe will add to the site Running I didnt get it the first time +This way am thinking that we can actually know who performed the revert operationwdyt +It appears to generally be a simple ordering issue +Darn +Complete w revision +Ive played with the support and I decided not to include it inside the main distribution since its confusing and dangerous +After the latest tests the client is still slower than Memcached reopening +to the suggested approach +Ive been looking for a way to do this for some time +will test when we get new code fix in the code drop +I was wrong please close this issue +Lowering not applicable on this FTP resource! +patchset attached +Since the underlying channels are transactional can also be easily made transactional +It would for the tamames example but not I feel for the other +Actually I know why it sends an ERROR saying unknown command +Im not able to reproduce this +Everything else looks fine. +patch applied please of Enders patch was applied. +Noble looks goodJay give it a try and comment back +Good job JingI think theres some refactoring we can do to improve it further +Then please edit the FS +ZhengI want to laugh now but I spent about hours trying to figure out what problem Hive was having +Tested +Not really a requirement to graduate from is the patch to add support for primitives typesThe patch was created by running svn diff command in an updated working copy of the repositoryAs part of this patch I created a new class defined in file +I got into this because I wanted really to update a field value +Please select Grant licence to ASF +Ill keep the tests in mind for future patches +I am not sure I understand the issue. +What help do you need? I might have a look at it. +Second patch search functionality implemented +On IA expression stack is at the end of the stackframe so there is nothing to get smashed by pushes +If it passes EOF it returns less number of bytes +This is so there are no hostport collisions between the different node managers running on the same box +This transformation only works byte by byte +Committed +The change is in thread block reuse +Added patch Fix for the issue +fully substituteI certainly dont think its a substitute at best its a nice alternative +to master as cc +I also agree to move this out of as this issue needs more discussion to finalize on approach. +Needs to satisfy this the related defect +Setting Fix Version to future for issues without a set fix version +Therefore reducing the sleep period to a small value doesnt improve the run time of the test +and a comment with characters current behaviour is a kind of channel will never consume input +If we ever see that this actually is a problem we can reopen. +It passes +The AS has no way to distinguish the case of a property never being used vs one that was used before usage tracking began +The fix for is already committed so Im marking this one as the duplicate issue +Agree with the first three items and will include in next version of the patch +As long as you choose the BSD license I see nothing wrong with checking this code into your projects svnYou will need to update the project NOTICE and LICENSE files to incorporate the legal requirements of the BSD license +Looks like current capacity was never being set +Performance improvements are not considered for inclusion in branches prior to trunk +However if the value is still checked on the server side then the validator will return a error message telling the hackerwhat the length is anywaySo option is NOT an optionIf users dont want the maxmin length of a fields revealed then developersshould not use the validator to inforce minmax +Thanks it works when I declare the namespace in my instance +We can only do +An argument could be made for consistency in look and feel however if all select boxes in the UI are going in this direction though +Were can we get the src files that match the latest jars? +testAutomaticallynew persistent class jpa name +Need some works +Please note that I downgraded the +Bulk close of old resolved issues. +I suppose this is a wont fix. +We should probably investigate why this happens when you restart a node but Im fine with ignoring it for the trash sent to storage port case since people like to monitor it for some reason +fixing a is due to the change that axis did in the release to append the endpoint binding to the service name and hence our listeners has to be modified in order to present the correct in the get methodsFixed in the Mail VFS and JMS listenersNo need to worry about the HTTP and HTTPS listeners because they return the value passed as the service name appended to the EPR prefixNeed to look at the FIX and UDP listeners +Up Down Top getItemsimplemented +That can easily be resurrectedLet me look at this version as well +We should wait with this until we have implemented +output for the systemvm usrsbinxen just applied the fix to the master branch +Sounds good +Wasnt my intent to go all process on you just to say I wouldnt mind commits like on this jira without review +Yes this is a good point +git Jenkins so it will test the patch +I dont like to put long text into the ticket because this is copied in every mail which is sent for the issue +Using the same in a standalone instance I see the same stack trace but the server does start +No the JSR specification is not licensed to the ASF exceptunder the terms of the Specification License +As someone who uses key X with random partitioner all the time to walk through results Im concerned about this ticket +I see the commits +fixed in subversion repository as of rthanks for the resolved issue for release +Marking as closed. +leave for it to after all after the latest fix by Juergen. +committed to in build +Closing since its an issue +Dominik Do you have a patch for the bug? If not would you please post the code to show the bug? The code shall +Sorry Andrea I just sent the pull requests on github. +mvn clean eclipseclean eclipseeclipse installmkdir tmp tmp script mvn gwtrunCreate created with no user nameDid I miss some cleanup or something +changed and tests fixed. +Simply turn amazonTarget into a and specify amazonTarget as further interceptor name on the amazon rather than as target +Ill aim on preparing a release this weekend then the voting can startWould be nice if you could verify the fixes +Committed the patch after verifying that all unit and integration tests pass +Worked for meIt looks like the same patch applies cleanly to +Added option warn to the Failover transport to log a warning after every N retries where N +Yes for some reason the automatic update didnt work for that commit but the code is inI think the remaining bit in is what to do regarding the empty file that gets created but I dont think that should hold up the update to the documentation +Yes cql support by stress is funky +This issue is a duplicate of all issues that have been in resolved state for more than one month without further comments to closed status +Where would I do thatThanksTodd +Fixed for and mean that it will be in once it gets released +Reviewed on stage +still need to modify some test results +Thanks for patch Jon. +Thanks for the explanation +I do not know how to include the jar file in a patch +perfect +Verified. +If you add an for a class in then this will get used +Tests against AS from branch and CR tag passed on JVM which had problems with zipped CR see +Hi MaxI am not able to apply the patch looks like some changes were made the +Hi ScottThanks for the reply +Like a small snippet I provided +I dont see any jakarta listWould you like me to create a commitsturbine andschedule svn commits to go thereI do see a jakarta list whichI will move to turbine +Theres an import missing +If you blow away your +One line fix +Custom blog aggregators like Google Reader take care of much of this need a user goes ahead and just subscribes to those blogs he likes +Want to take a look at this Pete +Everything is at its proper place +This is not supported as we have already provided remote access to JMX including authentication integrated with the security realms for managementIt is still possible to connect to a local process +Everywhere here means only two annotations and Length eveywhere else the field is either not a string or the blank input doesnt fit the pattern +The algorithm as copied from the method javadoc is as follows If dealing with character string types then make sure to set the collation info on the dominant type +I do not know about the advantages except that it is the latest version +already had ANT there is no ant file in exampleI will add one for it +META +bump to +just updated the patch to current revision +Basically you need to quote the names so the replace works +The code stream is what you should be looking at and it the latest release +With only a boolean it could be hard to know precisely when to clear it nobq +Unit tests for the port are part of none included in this patch +Might make a difference too +Most should pass but anything related to backupnode and secondary will fail for the moment + +Looks good +Closing. +Ah thats a different story +Thats great that it works in M but will it work in RC? Am I missing something here? an update +Several things the rc seq file were in the original patch +Should be done at some point +The default behavior for is to use the constructor with the most parameters +i am gonna commit the patch thanks! +It accepts core names to which split index should be merged +Do you have a test case patch? The developers of this library dont have much time to devote to it these days but Id be willing to tackle it if you give me a test case that I need to make work +Ill have a look at whether we can do anything about would you actually implement check? The problem as I see it is that the overflow will have happened before so how do we know whether there was an overflow versus a deliberate specification of the resulting number +The headerFile parameter is deprecated +Thanks Sushanth! +Youre right for some reason I was thinking of JNA but the mmaping cleaner dont depend on it +Len still post a trace if you can get one just in case these are unrelated +The log in is intentional to let the user know one of their xml files is malformed +Between memstore and blockcache we cant just look at JVM heap occupancy we will be keeping it pretty full +Really I did fix it. +Correction this account just needs to publish snapshot artifacts releases will not be handled by an automated job +There is another idea for this but this works only for the serialization +Reopening to reintroduce the amended test case provided by Fuhwei +At the row level getslicefrom only cares about exact row match +It modifies the reference counter in a persistable message to provide a callback that indicates whether the completion is occurring in the context of the receive pathAdditionally it updates to pend completion on outstanding asynchronous messagesIssues I have with this patch theres no clean way to pass the completion status from the layer back to the for a given command + is a source releaseIn order to install a template you will need to build the templateSimply open templatesfull in Visual Studio and from the file menu choose Export TemplateI recommend you name the exported template with the version number so you easily identify it +Applied patch to trunk +This test passes for me when I do it for a class but fails in case of a script on versions +Plus I will take a look at the unit tests and see which of them and generate patches if needed too +Its thrown in for example the implementation +Patch to to implement this feature request +Please change to if you intend to submit a fix for. +At that point it can decide which of the cases it is in and react properly +Resolved from docs perspective until further information is available +Batik bundle Chartree for your quick answerI will deep in it and do whats neededcheers +Something very similar to this has been testing on node cluster for several hours +Applied to TRUNK +All itests greenThanks Christian! +Its much cleaner IMHO +Increasing priority +Yes my calling thread is set to MAXPRIORITY + +Moving fix version back to since we havent decided to release without yet +Otherwise it would pick the closest version which in this case would be +All in all I dont think this is a Groovy problem +Unfortunately I think we need to fix this +There might be something else at play here +Note for the above I added the attachment +We can look at updating the readme and project metadata but the runtime detection for the project examples should have taken care of this +It reads the configuration files for the brokers in a cluster starting at index +The grammar uses square brackets for quoting identifiers node names and node paths +Switch the product feature group applications form to a so that all sequenceNums can be updated in one go +But theres still a numLeafes in bloddy dyslexic german Ill go ahead and patch fixed the leafes problem and added entry +But whenever data is persisted its schema should be stored with it. +Another scenario that I experienced while working on importexport was importing XSD from different contribution in a WSDL in this case even tough the XSD was being resolved ok in the Contribution layer the WSDL processor didnt have any knowledge of the Contribution importexport on its internal resolution mechanism +Patch adding +The test failures are most likely unrelated but the javac warning seems to be due to this patch +Yes all of this is because of the compiler bug +The regressions showed that one more test test contained code to work around this issue for the client case +I am aware of the +In just a few years even it has come a long way for sure and being able to embed scripts within our Java programs is just amazingIve never been involved in an open source project though Ive been feeling the urge to contribute +solved but forgotten to am using and ran into this exact problem +I have changed it for Camel issues +Yes this works good idea and a good improve the its possible not only to improve the docs but also to support the mygroupsyntax for to have a more transparent view to the blackwhitelists +Properties are propagated back to supersub flows +on to trunkThanks for the review Stack +As discussed were not going to apply this rule against array initializers and annotations but only to code blocks +Well for now the workaround is to set the following in conf can still receive wiring debug messages +Found it thx. +No problem Sergey. +The interfaces on the classes have been cleaned up by moving the parsing of the xml document out of the and Schema classes and updated calling classes +Thinking about it I do not particularly like eitherHow about or or or +bq +After looking at the patch again patch seems to solve the concurrency issue but synchronized on the filename intern does put a burden when requesting the same artifact +the old lrumruworking set model was specifically designed for tuplebuffers which have a linear insertion and forward iteration model +New version of theI have just attached a new version of the files for the one line is changed in the XSD for allowing custom tags inside the parser uses the new parserContext and parserHelper to delegate the parsing of custom the refactoring you did on the parser the fix is much cleaner than before +Resolved for insert and remove operations on Accordion. +Not sure if I should be validating your changes but they look ok to meThanks +We must be following a different code path somewhere +Yes I am aware of that +Please verify. +tmp files will be deleted when their associated marker objects are reclaimed by the garbage collector +I cant talk to their IRC channel from work so this will have to wait until this evening +Fix is committed and will be available in the next build +Committed revision Thanks Ahmed +fixed +I am a applicant and I find this project interesting +Chris is your feature part of file in the bootFeaturesAs the fix will certainly include some refactoring I prefer to postpone to at least +Similar to this issue must solve similar edge cases such as the one highlighted in + +Is the issue still here using STS +Dupe and fixed. +I have looked into the error handlers as well +What can I do +it had a project which had to calculate and draw the according to calculated data +I cant remember why I changed that. +Derbys TIMESTAMP precision is so the string format of that should support the complete resolution +With in place the VM creation and installation of the bit is likely to be fully automated +I dont think JIRA is the best place to handle this. +Sure Todd +I am not sure I understand why we need to add a test case for this to Derby +Thank you Plamen. +Bulk close for +Since currently its an eitheror decision between LRU and for data blocks a slowdown does not seem to be acceptable +This seems to have resolved my issue thanks +Bulk move issues to and +Both the artifacts are in fact the same jars bit by bit and different names +I prefer approach +The code to fully qualify the Body id attribute happened after the begin tag was written out so only the unqualified tagId value was written outI added the BVT test to verify the propert behavior +Closing issue it has been resolved for a long time. +Jim this is a potential duplicate of which has been resolved in +Fixed +Looks like we are not including the skin for this +seems ok and the biggest problem with is the lack of artifacts compiled against Hadoop +If you read the previous comments on the issue carefully you will see that it is about a process started via system in +Thanks for spotting and reporting the problem. +Michal and why do you think there are duplicated classes? Unles those classes blong to completely independent jars and loaded by the different classloaders or applications in the runtime you going to have issues +on CSPand in detail! +Application action is implemented as +Need to upgrade KVM host to at least RHEL before upgrade to cloudstack +Hi Suresh Sure +Also removed myself as assigned so that others could investigate potential fixes +It would be sufficient to leave off autocompleteonRichard +I think this works on some platforms onlyDanushka +Problem due to new not being a submodule of the uimaj pom +That seems a bug to mealso since the string is preappended as VP would it not use it own counterDoes every counter need to be tied to a entity +I dont think soIve removed the version as its the newer of the pairsvn ci m Removing as its already found in +Changed to incompatible change since it increments the client protocol version +Where the dependency is a jar from the maven local repositoryin parent directorymvn clean ideacleanmvn installcd webappmvn ideaideaLoad the webapp in IDEA and view the module settingsThe packaging method will be listed as Copy files to +There appear to be a number of issues in the juddi RMI transport so it is easier for us to provide our own version in a similar way to the have tested out the new RMI transport using two servers the slave of which has had the juddi registry service removed +I didnt look carefully but my guess is all are likely supposed to be private +The error message has been cleaned up too +Torn as where to put this one but Ill put it here for the points the user to section of the for info relating to jndi settings +Updated code to handle empty classifiers not just null classifiers. +Hi I was thinking about the proper waiting time +They icons would show up dynamically good for demosBut it would require some coding around the methods to not call them until after login +Updated patch to also fix scaling and glyph positioning issues for text using a font +moving it off until its really needed +The latest comment looks like a possible fixIs anyone doing this +Ive asked Noel to make the requisite request + after release +AFAIK JNA is LGPL and thus incompatible with Apache license +Stroooong I wanted to do that long time but some tests were made me afraid +Why +Ive been able to reproduce this with older versions of the JDK +Heres an svn patch that makes the required change +Also fixed another problem that if the new classs package fragment did not exist an NPE was checks to make sure package fragment exists before creating new groovy class +Should we just disable factoring in requests in a region servers load for the moment? It might lead to a worse distribution of regions but it might also make no difference if on average all regions are equally busy which would be the case if youre running big time mapreduce jobs +When building the bitmaparray offsets the stored image pixelsIm not sure how to solve problem though +John In +The new patch is not able to detect the case that NN first initiates a block recovery and then client initiates a block recovery on the same block but uses a different primary datanodeThe jira seems to get into the details of detecting concurrent block recoveries +Fixed in commit +Patch will commit now +it will simulate being external +It was caused by we upgrade the default version of spring +also update defaults and docs. +Patches according to issue description to modulescodegen to modulesadbsrc per issue description +I took a brief look and I see some reported issues on memory leaks that date a few years back +However the consumer tests are a bit tricky +Applied we close this issue? +IMO we shoulndt sacrifice performance for a use case which isnt properly described in the specification +Also later I will upload patch for branch and trunk +I remember the day I wrote program in C before to switch to we will want Franks confirmation but its look like we are heading toward closing this issue as Not a bug since Frank replies suggest that the Linux output is the right one and Geotools seems to be consistent with it +Need to add the property to +Roman interesting +Actually I simply applied this for use with +Taking a stab on client blind initiation based on offline discussion with DarynThe patch optimizes for token auth and leaves door open for future optimization of other auth types +Sergey could you help them with that info +Are you submitting results from multiple threads for the same test run? I think that could explain the locking problem +Added form to modal window itself. +Looks like this is already fixedthanksdims +Another version without dependency to Neko +for the patchbq +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Lets keep a watch on situations where the communication gets stuck due to an OOM +I mean think about it in this simple message flow connections max are kept open how can that be a good thingThanks Freeman thanks! +failing to resolve multi threading with Lucene +I just committed this +I dont see how in theory the seek can be a problem when tailing a log from the start since we read the whole file +im skeptical that youd be able to balance fast enough to deal w hot spots significantly better than balance by disk space but i suppose its worth a loadbalance feels like a dead end. +out of date +Gavin WDYT of allowing that in the case +Alessio what about to be forced in +I quoted a line from the patch +Merged into +Ill make it internal to the server +ran tests manually ant test passed +Since this isnt very intrusive maybe you can squeeze this into. +If you dont mind I would like to take a look at this +I didnt realize Spring Data Commons was a separate projects +Tests against ran cleanly for me except for the heisenbugs in +Marco do you consider to provide a patch and an unit test which shows this issue +This content can be expressed as a range of versions and in the resolve process the content is locked down to a specific version within that rangeThe resolve process also identifies a set of bundles that are needed to support the application but are not part of the core content +for the to trunk. +If you finish it before we release beta please correct the version +This patch includes core domain utilities and the tld plugin but excludes the changes in and +not a bug in Archetype but MPH and that is fixed +Do I need java to run this +Youve pointed it out yourself +Alternatively we can use a property for ignoring set so we are compatible with legacy applications that dont rely on the ID being set on the connection URL +I see the problem +My clients system is down +Applied the query patch changed test to use it +Instead have a showDetails option to enabledisable detailed stats in each thread probably with default set to false +Once this set of run is complete I will start running my test with trunk and with jboss think this issue has been resolved +Heres a patch that implements this +Im getting this error for group by both with and without the debug parameter +Suggestion Tools Knowledge Base collects information from projects to provide info to features such as expression language validation and content assist +Patches are preferred in diff format +Do you know if sql standard says anything for it +Revised with Florent Guillaume comments get should return a list could you please add a unit test to srctests? you should in particular check the call of get on the root in which case it returns an empty list because the root has no parent +The patch tuscanypatchupdatejira seems to be generated from a old trunk version somehow my tortoiseSVN is reporting The patch seems outdatedRegardsAdriano CrestaniUploading corrected patch made from the latest trunk versionI think enough patches have been applied to warrant resolving this issueAny additional issues related to the ant build system should be dealt with in newBrady JohnsonLead Software Developer Rogue Wave Software +This is similar to bootstrapping other HA systems with persistent data + trunk +Ive looked though what youve done and it seems like it might fix the issue we encountered +Is this a known flapper +Patch fell out of date +Its ok with me +Applied +So I guess there are warnings of instance field in serializable class +passed on my box +Thanks Gregory the patch is applied as expected. +Note that this is a change +Testing for should shake out any bugs +Reported issue is not observed now +bulk move of all unresolved issues from to +I just committed this +Dag saidNo dual boot is really happeningThank goodness! I really didnt sleep all that well thinking this was back +Ill make a wiki page like for the other components +IIRC this issue has been fixed in the refactored template generator available in the template block now + +See my email on the users list +I also prefer the simplicity of patch +Then youd need the timer to track the mouse movements because there wont be any mouse events fired right +Exception more rundomly throwed exception for this test is NPE +Your patch is in CVS +I think thatll workActually tried the import of that file and it appeared to be working fine and I could selected Elements on next page to create columns and proceed with import +Shifting to are not logged under Production. +Good we do have partial classes so maybe we could select different file sets for different framework versions? ugly but a related note well probably be shipping only +If you want to use relative paths change directories first. +to fix up resolution and fix as wont fix since indexed is moved out +Thanks for bearing with me here +I was thinking that the connection timeouts are unrelated toYou are right only runs maps so there is no chance to uncover this issueAre these socket timeout exceptions? I remember running into those with gridmix but never got around to the bottom of that because of more pressing concerns +Still we could reduce this to a single call if the dir does not exists well get a FNEWhat sort of transient failuresIm looking at this from the point of scalability issues weve seen on large clusters +IE version works for +Please let us know if the problem persists in later builds andor if you got any clues on how we can reproduce it +Assigned to Randall as the teiid sequencer seems to be the only remainig decided to resolve this since the only outstanding subtask is which will be tracked on its own lifecycle. +When that is combined with the fact that datatype handling is much more strict now but also extensible makes it difficult to parse dbpedia with the huge number of custom datatypes that they useWill need to write a custom implementation of for I think before completing this. +Seems the inverse bug is fixed in! +Tests test test and test work incorrectly due to bug on EclipseThe similar problems was obtained on Sun and on BEA Bug was submitted against EclipseI think that this issue can be closed +This exploits the fact that the EJB container calls quite carefully depending on the lifecyle of the underlying EJB +Looks like there may be an issue with the Windows code for the Mutex class not sure what it is yet though +Hopefully you guys can setup a build server soon +So what would be a nice patch that goes around this tomcat issue but doesnt depend on tomcat or breaks any other container +Yeah we should document that the hosts wont be serialized +Actually persistence is the correct term to use for messages since that is what JMS uses +I have already solved the problem +Closed as it was fixed some time ago. +They all use FF +Labels in Libsvm data sets will be separated to files with suffix +So a release note will have to be done for the AIX kits +I suspect that most of the default was actually used by tests and often not by intentionSo the code checked in for x is slightly different than the patch for in that it puts back the default and includes an explicit test for constructing the default since the point of the other changes is to remove the dependency on theSo going forward when people write tests theyll have to either reference a solrhome that contains a fileor copy the file to the right place +patch for branch corner case is now handled but the new tests added dont test for it +Builds are the normAlso I was looking through the code that was checked in and I still see lots of values for versions that shouldnt be there +What is the issue number of the primary report if this is a duplicate I would like to follow that? Or should I file a new issue +Mmm what u mean under clutteringWhen i press Add Spring nature from context menu all works +Fixed with revision +sounds like quoting is not happening chance you could provide the actual schema and exact DB and Driver version you are having problems with? I couldnt reproduce it +Attached is a protocol subclass of TCPSSL which should in theory set up the SSL Context from properties of the configuration string for the stack +It would be deseriable to have it settable in +I think a sample app and some more information regarding what you are trying to do here would be helpful + +Net will do the same conversion of its hash value which will get us around the encoding problem +Might be because I only have nodes each with datanode and regionservers +moving it back to mean at least for the arcsde extension an old geotools arcsde jar is being included with the extension zip file among with the actual one grabbed from the m old arcsde jar is in our geoserver svn under the releaseextensionsarcsde directory and seems to be from geotools +Can you describe the motivation and purpose of each of your changes? it will make it easier to review +Are there any errors in the logs? Did you test on your local machine or deployed it to GAEIve done most of my tests on Im upgrading to right now to see if there are issueskind regards I havent had a chance to upgrade Click JPA to the latest Click jars +The fact that is making instances smellsAnd on your approach + just does git commitpush right +You mean to say dependency classes should be loaded from any path containing a lib prefix within the jar +We have onyl changed this from globally open to localhost by default in for security reasonsMy opinion is to leave it at thatI could imagine though that we extend the remote shell to support configuration for easier configuration +Another thing is if you could add an apt file for the report mojo patch adds a test and updates the apt +Nico work +It should also be legal for the NM to clean up application state given a richer set of semantics for the intermediate output specifying retention +Note that explicitely calling mvn clean sourcejarworks with and creates a correct option is to just the default location into the war plugin +Provenance is clear but dont want to modify +Exclusively for +please has been committed. +The secure processing feature was not being read from the in the constructor of +So is that a subclass of KV? Or a bunch of conditional code wherever are handled +Please check my attachment. +MattThanks for the updates on the Exception handling +Not sure how that happened +No problem I was just going through some old issues Ive had an interest in and thought I would a patch for that one +Reading documentation helps you not waste our time +bq +for patch +also applies to the branch reviewed the patches +Applied patch +Updated the environment details +get is a privileged operation +This should resolve the case of its use within the broker side code but will not resolve this issue within the client side issue +Thanks for all the work! Ive tested the and results do indeed seem to be faster although I had an issue Im not sure is due to my code regarding reading results twiceAlso concerning limits do they currently work with sorting? If I limit to results will it be the sorted first results? Or is sorting done after limitingLet me know how I can helpBest regards Serge Huber +The same is true for pending merges etcThis simplifies the locking a lot here since we basically cant deadlock anymore from DW or DWPT since we dont even have a reference to IW anymore +Thanks patch with your suggestions +Hi GuysWould you want to provide a patch for this +It is exactly equivalent to the old code in the only case of +Nice +Im hoping this fix addresses Andys original problem +pete +This is not a bug +This is broken again +Robert can you please update this +What is currently the workaround to have it running without the class not found +I attached a patch to fix it added a test too +Martin can you review this change please thanks. +I will check and fix as part of +Here is the patch for the same +Im kind of stumped could this be a classpath issue? I guess Id still like to try it with a version compiled with JDK since I dont know what else to try +Issue may already be resolved may be invalid or may never be fixed +can you put it on +This drift idea was abandoned by me after realizing the performance was no good +Nicholas if you are working on this this feel free to upload a very early version of the patch so that we can review it earlier +Thanks for this st course Scotto +Committed to main trunk with revision +Last but not the least we should eliminate in memory buffering of the request content body +I have the same problem which is very annoying for could work on that +JeffIt looks tome like we dont want to share across threads +Trunk patch tested locally with lets see if picks it up +Initial commit ceadfdfdabedcaddf adds refactors Roo classes to use the new commons classes and removes Roos Assert and +Looking into adding some tests for writing for both composite column types and composite keys so they both can work +Ill add a fifth patch to add that syntax soonish +Just trying to make sure there is no this would break the SI assembly encapsulation +I even tried adding document at a time +There is no workaround + +Regarding atomic ops on nonold gcc systemsWe could import portions of libatomicops which fall under an MIT licenseIf we want to do that lets file a separate jira issue specifically to do that +I added one more change to the balancer +what were the motivations purpose and technical challenges encountered whilst working on it +Same happens in +After completing the wizard and attempting to generate the zone there are errors defining the cluster for +If someone can provide a good one one alternative may be to provide a proxy implementation +How close to release is +bulk close of all resolved issues in preparation for release. +In fact Owens point is to say that you can put jars in the jarIs there any reason for this hack to unpack in and look for a lib directory containing jars? Isnt that too much of magicWould it be a viable option to allow tasks to take a list of jars instead? Or is there a compelling reason to only give a single Jar + +This means querying the metastore converting the table schema into an hcat schema then converting the hcat schema into a pig schema +Attaching the patch which changes the CREATE IF NOT EXISTS to simply CREATE +Im not sure a new writable needs to be introduced here +Implemented following changes Ivy Ant tasks are now compatible with Ant +The HMS will createdelete HDFS dir use the server side UGIIs there a way to use client side ugi to let HMS manipulate HDFS without +The nd patch looks great Ill commit shortly! Thanks HanIll also go improve the javadocs in List outputs impls +Not a blocker for any release +heres another additional patch to remove modulessuggests analyzers dependency again it doesnt actually analyze anything so I think we should remove this dependency for simplicity +Attaching a patch incorporating Ravis comments +Code has been imported for M. +Patches go in when they are readybq +Added fixVer of Amos +Different expected exception was happen in negative test in blobclobBLOB +Maybe we can do by including only in the testsuite +So tmpmypath which is already properly encoded was again encoded so the turned into resulting in tmpmypath +properties to theThe specific issue of configuring the retry handler has been addressed in +Committed to JBESBGAFP with revisonCommitted to main trunk with revision +I understand that by default the polls the target website but how do I make it work in a http streaming scenario? For instance say you are connected to a twitter stream which keeps pushing chunked data on a http connection that you would keep open how would I handle this scenario using camel +pulling in these changes fixed the problem +marking this issue to blockWith Dans review of the first checkin he noted that there was still work needed in the area of casting and comparisons for +Might be good for this to be independent of hdfs +The above is clear and solves the problem +I will mark the new jira entry for npe as not a bug +Cant verify at the moment however +This could be related to priority to major with version the perform succeeds even though I still see the errors. +this looks really goodyou build an ack String but dont use it +Gmaven changed from its own stub generator to use the one from Groovy thus loosing the javadoc in the process +This patch came in as anonymous therefore we cannot accept it as its provenance cannot be proven +This patch has fallen out of date can you please regenerate it? +A du with recursion implemented on the namenode took less than seconds +query was SELECT foo FROM WHERE KEY bar and KEY barNothing a sane person would do I was just using the randomly generated keys for the stress test the same way the thrift based one does +Many of the CLI tools expect output to be piped to a file or another tool directlyMaybe a better way would be a version CLI parameter +Having ZK remember all servers that connected and the admin command to list the servers showing their status would make it easier to see whats happening +Now that work is well underway with AS all previous community releases are +Todd does it look OK to you as well? If so Ill go ahead and commit this rebased patch once Jenkins come back clean +Per Sureshs suggestion closing this JIRA as invalid +I just committed this +Moving back to to have it removed +This patch fixes this +All tests in branchxsolr are passing +Any progress on the verification +According to EXSLT spec +If I add an empty class X and an empty interface Y and if I put a def in front of the methods then I see no i didnt add x and y for will email you all the files possible attach them to this issue +Thanks for your work +The pagination still does not work on Oracle Batch Admin Batch reported a new issue regarding SQL queries for Oracle. +Similar issue exists with other dfs commands especially cat ls lsr mkdir rm rmr and du +Also added more tests which verify the targets work with multiple outputs the source can read from the same path as the target and fixed other similar tests which were not working in previous patch +I would rather see us learn and use it vs +verified +Hi need more information before I can resolve this issueFor A +Sorry for the confusion +The IP resolution is done by the which explains why it is slowing down the number of being available for fetchingI dont think that the oldFetcher allows to group by IP for politeness in which case why not making default to false in the new fetcher +Made change in to make a check on the syntax of select do a code reviewThe attached screen shot shows how the select query works after the patch done in my local +Thanks Alexander Ilya the patch was applied to BTI branch at rPlease check that the patch was applied as you expected +The problem here is that we dont do anything with type parameters on methods +Now that work is well underway with AS all previous community releases are +Thanks +get the SRPM from Cloudera and rpm i it +the resolved issueLahiru +Actually it was just fine +added or corrected descriptions of these components calendar columns comboBox componentControl contextMenu datascroller columnGroup column dataGrid dataList dataTable subTable dragSupport dropSupport drop effect extended hotKey inplaceInput inplaceSelect insert jQuery listShuttle menuItem message messages modalPanel orderingList pickList propgressBar scrollable tree treeNode recursive tree virtualEarth +Pls apply this patch to trunk +leaving the properties on the target bean +Ioannis I also attached the documentation patch +What do you consider a short timeFew secs milisecs nano. +View of the eclipse maven central browser failing to locate +Hi Joelthe concept here was that the Table rowList could be very large and we didnt want it being serialized as it could really blow out the size of the session on the serverHowever it is possible that we should think of making this behaviour configurable with a table property clear or we could just set the rowList instance variable to nullregards Malcolm clear property +As Cassandra grows I could see Tools or Core splitting as other people have them above +Nevertheless what you see happens actually from time to time and theres not much we can about it but to increase the timeout +Thread safety in is now achieved by the use of atomic references and booleans in critical sections rather than through synchonisation on method levelPlease reviewOleg +i didnt realize this were just in the test project not in the actual test plugin +There are several methods in to help herecopy copies a very tiny schema and config file +I think you are in a better position to do it since you have committed the first patch +This is the addendum +I just filed on the proposed feature +Filing a separate issue. +Looks great +Pure java cvs provider is in development in sandbox of +Thanks +The patch was applied at the revision +Its still reproducible +I would implement it myself but honestly the bean factory is a complicated piece of software and I currently dont have the time to dig deeper into it +Can this fix be reviewed and pushed +Reopening since I doubt this was fixed in +Please specify what behavior should be seen +yes it is the right patch +This is the work that is being done for any case please avoid making small changes like these without confirming that we are putting in a similar patch in our codebase +Updated exclude list patch + +I probably could have gotten away without the query string parser but it made the test cases simpler to write +The current file is correct patch filePlease give me your feedback on this fileRegardsAshish for looking into that Ashish +used to didnt happen again +can you give us a working unit test case we can start our investigation from +I am reopening the issue in order to apply the follow up patch +No +closing in favor of Stus more recent findbugs tickets +It should be the package of Widget A +OK just making sure were using the same bits +I have just committed this +Rejecting because this is not a problem +Riley would you mind putting this up on? This is a pretty big change to read through as a patch +I tried to kill server before ack is stored and everything is ok +Closing in favor of +why are those lines removed Benjamin +As far as I can see we are using +Very good catch KlemenThere are other places in the decoder where we have the same issue +Heres an implementation of the task monitoring code plus instrumentation of some processes like master startup compactions etcI verified that this works reasonably well on a clusterThe UI for viewing the info will come in a separate ticket + of release. +I committed this +The logic for is unchanged only faster I hope I guess for the we need to think a bit more about what to do +jar file +Looks like the server log is now getting appended between every server that the macro which looks for server shutdown message wont work as it would find the messages from earlier shutdowns in the same server log +Thanks langPatch applied to imageio module at repo revision rPlease check it was applied as you expected +I just committed this +The multipart handling in Struts was known to be more than a little buggy which is why Struts now uses CommonsIf there is still a problem with Struts andor Commons RC please file a NEW bug report against the relevant component with as much detail as possible so that we can fix it before a Final release. +RegardsParimal Sumit it was fixed in rev +Verified by Spark. +looks like it would be relevant to the use case mentioned above + +It also would be possible to add a interface that can be used to mark messages that contain such a timestamp +It seems better to make the restriction explicit rather then wait for things to go wrongeh +I can commit this after the weekend +Committed revision Thanks Apostolos! +The documentation for allthe essential classes should be factually correct and consistent +Same here +Was this ever fixed and if so in what release ofWe have the same problem in our environment verified with your with Jbossweb +But do you have any testing data to support your case +I made a few modifications which I think makes things clearer +Thanks for the heads up +attached patch contains fixes for this issue +If you are finding this not to be the case we can and I will take a look +Chris how do I run the integration tests +Disabled in r +This bug is still valid +Is the FS being closed before this finishes or what +Yes what a luxury it is. +no +Trunk will be updated when mergesresolving. +This problem was caused by a duplicate from RC present in classpath when using netbeans +what does procmeminfo look like +Fixed by XSLT. +How about this as a use case? project top foo persistence setup provider DATANUCLEUS database HYPERSONICINMEMORY gwt setup persistence setup provider DATANUCLEUS database GOOGLEAPPENGINE applicationId foo gwt setupHere youre stuck with no way to generate the GAE client server classes used in the scaffold all because you changed the database type after running gwt setup +This patch is for +What I need to do to had french language in my springroo project? Have you created an external +One line change to the integration test to pass after the fix in the depends on +Done +Note that this is the same assertion as in which is not related to the python client +The API on OS X refers to it as escaping +committed +Chained implicits I have taken a closer look at this +Same problem applies to Map tests now work as well +OK I just closed the issue since it looks like the Framework approach is under ways +As part of this task I will have to modify tests that experience changed behavior with locators +I fixed the incorrect XML handling in solr a year ago and did the same thing to pass the charset to the XML parser as hint. +original patches incorrect +I have very little of a linguistic background so Im not sure if there is a solid for this or how would it work +This issue has been resolved for over a year with no further movement +New qpidAddress splitted out of qpidUrl qpidsys implementations for linux and solaris New methods in qpidsys to deal with the interface addresses Usage of the new qpidsys methods in qpidUrl Inclusion of qpidAddress in qpidsys and qpidUrl +This is so that you dont end up with a query that reads +I fixed this in revision Thanks for the report! +See also and for the same problem in the library +move reporter from Steve to to +Logged In YES useridIn the current implementation literally follows thespec +Specifically and the base load in mapside cogroup need to be excluded from the split combination +As Kan reported this has been fixed by another patch +So there isnt support for in is it +It introduced some API changes a version matcher being necessary to know if a revision is accepted or not some methods now need such an instance +x reads are again very consistent but with little difference between PS and BBh +NO MORE release is the only maintained branch in the community release currently the most resources are target to right now +Applied patch and delivered. +Glad youre using the library and willing to report things you find! +You can check screen menu +is ok for me +Just sent the document to Apache cced Claus for reference +The fetch plan algorithm crosses detach tests makeTransient and fetch planPlease comment on the attachment that shows what the would look like for +In the case youve lost it I can of course make it come backIn the patch itself I have seen that the checkstyle rules are not respected +However +ported and applied patch to and applied to +In this revision filter parsing works well +The patch was applied at r +The fix has been applied to branch remotingSPCP +Lot of conflict while merging +This was an issue for me for testing FUSE is installed to match the kernel so on a bit kernel I have to use bit java and bit libhdfs +I found it +Will publish to docs stage for review once the problems are resolved +Im very curious as well +Unfortunately there is nothing publicly available whichhas comparable qualities in terms of usability +Recharacterizing this as a bug +Fixed in CVS +I have checked organizer in the FF IE IE with jsf and it has similar you please attach the screenshots +It also takes care of removing parent shard information when a shard is switched from recovery to active states +Charles are you using the to convert a String into a Query? If sothats the problem +was verified to adhere to the new policy +Committed revision thanx pat +Ok I guess this is my faulty understanding on how works +committed to and trunk +committed to trunk +Is this patch ready to commit +New version of same patch removed some unnecessary comments +Marking bugs with invalid fixed for release as fixed in. +I suggest to create a new subclass of for thisissue where we add the malformed URL as a field +Torsten please check the seems to be a typo incontraints should be constraints +With the patch the user cant reliably control the writing or verification of checksums wo explicitly resolving paths and creating a new context for the resolved path which kind of defeats the purpose of the contextDo you have another solution in mind to control the flags +I have attached a patch for this +If it helps I can actually use the web service and I can list the service using +It creates a port forward makes a connection through it and immediately severs the client connectionThe test looks through the stack traces of the threads to look for this specific condition and times out after seconds +The solution will be tricky +Thanks RichardPatch was applied to LUNI module at repo revision Please check that the patch was applied as you expected +Previous was the wrong patch +Fixed in nightly build +can you debug in weblogic to see if the second request can find the first http sessionIn other words is a http session created +Seems to be ok now. +If we want to have it in GA then yes +May I have a +But to provide an unprotected version of replaceNode for the call in apply should be better +RamThe analysis is correctHowever I think your solution is not use the optimization made inIn my I just change the file name of edit log to In this issue with the patch it will not skip file because we shouldnt however in other case for example we first put many data to RS then move regon to RS and put many data again +are correct +I mistyped in my previous comment +Allthough we can make cleanup to also create files at the same size +I wont comment on viral changeBut I think this is a totally fair thing to do for since its a new major release +In your mapping you should be using instead of columnPlease let us know if this resolved your issue +Having cargo skip deploying to a remote container because someone passed in seems to be confusing mixing concerns +Attaching an initial pass at thisAlso mixed into this diff is a change to drastically reduce the number of times we need to iterate over a large databag to estimate memory +An enlightening comment +Done + +Do you have a simple standalone example that could show this behavior +Thinner XML for program for to use against +I think this is because the patch creates far less garbage for GC to periodically reclaim +So alldocs behaviour for keys isnt applicable to user viewsWhy alldocs returns error information on missed key? Probably to help make stateless applications when after request you dont need to know payload to process response data +I did not want to see a good patch get too stale +passes mvn verify etc +I think that bundles should be released independently +If we try to make the jdbc act the same as sql then why have both? I am happy to help +A full report and solution will come in few days as I am finalizing the fixes +Does this sound reasonable +However old java code that is final classes will not work +While readability is important Solrs goal is not to make a product that a CS grad can read its too build a blazing fast search server +Hi Andrew thanks for looking at thisYour suggestion is fine with me but I think we also have to take the link to the demo out of frameworks +So I agree changing this +Thanks for the test case and investigating the problem. +This may still be a valid bug but please provide a test case +Ill try to provide a couple of patches for most of the trivial cases +Also the chinese docs have references in the following filessrcdocscnsrcdocumentationcontentxdocssrcdocscnsrcdocumentationcontentxdocssrcdocscnsrcdocumentationcontentxdocssrcdocscnsrcdocumentationcontentxdocs +Some bugfixes from the epic patch were also done in other JIRA meanwhile +Theyre processed into svg png and gif +The fixes for appear to have fixed this as well. +As such it is good to add not only versioning information to the original request but a very stable reply that indicates that there is an irreconciliable version mismatch +Theres not a lot more information in there though +that would be fantastic +Hi this was added to the last release +I committed this without submitting to Hudson since Hudson is not able to test this +Fixed in r +Thanks Leonardo +Closing were assuming it was a jvm issue of some sort and fixed somehow by the next release. +nice catch and nice fixsometimes it takes an alert user to find a subtle use case that exposes a more general underlying problem +Preparing for release +Are you still working on this project and if so do you have anything to share with the community or any further advice which we can provide? RickI am still working on this although I havent had much progress for the past couple of months +Sorry been snowed have resubscribed you to the list and verified that you are showing up as an allowed should be good to go +Updated patch with some improvements +From the Teiid side you can already use JTDS and tell the import not to use the catalog name turn off quoting on import altogether configure the connection pool to use an initialization statement to enable last point should be captured in documentation related to the Sybase translator +A chooser which would never choose the last volume would eventually decommission it +Also disables short circuit checksum for the no +I really feel more efficient with consuming directly without installing them locally on each change +Verified in. +Thanks Emanuel for porting from trunk. +Committed revision . +Attached is a patch which backports this change to +The way the python implementation works with Twisted should be a better approachIt would be extremely useful to be able to make multiple asynchronous calls in parallel through the same client and use a callback mechanism without having to use threads but do you think this should be part of Thrift or something that sits on top of Thrift +I was looking at it from performance point of view +the works well with async hbase +We need to rewrite one from scratch if we need it. +If you read the chapter in the reference manual it explains this +In my experience maintaining separate components only makes sense if they need to be deployable separately +Sorry I meant to say RHS in the prior comment not LHS +Hey DanielThis is great work! It looks like the original intent of this JIRA is out of date with Richards new scaffold +Revised pull request submitted as per comments from Gary +which encoding are we dealing with here +Im relative new to this project and have not made any branches or tagsI cant tell which ones can go I personally only use the trunk sometimes the BRANCH and a sneak preview of the Stripesbranch +Andrew Please add changes to at the top level to make sure we make the right attributions in it +Does anyone have suggestions to get this sorted? Its this kind of stuff we need to deal with if we want to shift? NoAlso where exactly is this dependency managed or even specified? I cant find it +Assigning to Laura as she is the new lead for EAP docs +Fixed via +Thanks for your help +identifier is null for a identifier so will never matchAttached patch sets explicit if all of a component ids properties are specified +I just figured that there would be enough disk space here in the default configuration as the logs would be placed here also +Yup +I have corrected the issues +There is a workaround so its not a major issue. +As I understand it the security is between client and server across HTTP +As the other times a set of smallish commits doing different things just one diff but renames are not showed as deleteadd for robots this one will actually apply one diff with deleteadd I am ready for commit either as commits or as one jumbo commitRegards changes look goodCan we just get this in now such a simple change yet it seems to have generated so much workThanks Santiago for shepherding this through +Resubmitting it +Can someone from Facebook comment on Rushs last post? He raises some good questions about having two versions and if the patch is approved and committed to the trunk +will avoid exception in case server does not need to authenticate client +This patch is based on the first version of the patch for which is also included in this patch +Thank you for your contribution. +No problem bro +User error resolved by updating docs. +I did see this tag being added in the release notes of but in the jar files shipped with the it is not there +to go ahead with the changes in and start using this +Documentation is easier to understand nowI also noticed that you renamed the include attribute of the war element to include which makes its function much job and have a nice weekend +trunk Available from SNAPSHOT repo through version + +Very odd +Im not sure how much further we can go than should probably mention that we are using spring +can you verifythisGreets changed the build to point the document to a child directoryYou will either have to get a snapshot of the build back when the bug wasopen or create your own testBasically if a document atdeveloping +Applied! +Is it a better idea that checks the root cause of the and retries if the cause is +perfect to me +So +The serialization method was two orders of magnitude slower than the copy constructor +Patch for and for and +If I have that right can you please attach one such problematic BMP written by Sanselan + +Note that start and end are invalidvalues for +To fix this go into your HelpSoftware UpdatesManage Configuration +I am having an issue with +Looks good to me thanks Shalin +I can successfully execute HTTP GET againsthttpsapps using and JRE and get the content +Best I can tell you is that it was a Java project with no maven know its a bit vague was hoping the stack trace would be enough but I can understand if its not +Theres some overlap between and Ill discuss on list if and how we should merge these contributions see upcoming Resource merging and symlinks two similar patches do we unify? thread on dev +no it isnt related +Well I think it is a good idea as long as it doesnt impact the standard code +Not sure what the multiple aspects are but thanks for fixing this. +It isnt a bug +Prepositional Phrase Attachment Dataset from Ratnaparkhi Reynar Roukos +Looks like the generated files were checked in on purpose for the build fails without them +Committed revision . +I hope to have a review up by Thursday evening +batch transition resolvedfixed bugs to closedfixed +sorry about that attached thanks! +Any updates on thisI dont have a machine to repro on +The implementation isnt +Are you willing to help us with testing +So It looks like things did not fundamentally change in +let me check today may be bad code in a sense that we expect these things to be set and they are correctly not set for views +New version of the patch +Fixed in svn revision . +Im with David in thinking that this is simply a schema resolution issue but I think its reasonable to give Tom a chance to dig on this and see what comes up +the grand total composed by the sum of abDone changes on reconciliation changedThanks to Rishi Solanki and Chirag Manocha +and remove themIf you call Debug AsRun Configurations +I agree that improving the general testing is a worthy goal +Got any other suggestions for thatAbout the lines then yes I would also prefer not to have any lines go through the boxes but sometimes we cant avoid it will take a week to create a alternate color when drawing on top of a box then we should definitly just stay with what we have +grailsdownload and try to unzip it manually just to ensure it is not corrupted +Such in this context is only referring to +Sorry +Jesper Pedersen will need to comment on this but I wonder whether validating the ability to connect should be part of starting the datasources MSC services themselves +The issue is the same as the failure of the smoke test +Hey JackThe last activity on this bug is dated more than a year ago +Sorry for the delay here +Opened a separate issue to get some better error messages for this case +This fix should solve this test failure also +Everything is in order and committed now +I just was testing the updated on the blog assembly and hit a NPE because of the removal of the comment service dont think thats supposed to happen +So we may have to go with something like block timestamps but I still dont see your point about timestamps substantially reducing corruptionThey enable detection of stale replicas thereby eliminating that source of corruption +Closing this record since is now released. +i have moved your comments to ticket i have also tried to cover all your inputs in ticket it was not my intention to snub yousorry for that! +I am uploading a new patch +Seems to be a leak in either maven or logj that this test exposes if I set the root logger to DEBUG it gets an OOM pretty quickly have changed the to log to a file instead that seems to have taken care of that problem now just waiting to see if theres any other issues +Method add is deprecated now in favor of using unified methods pair addTo and removeFrom so this issue seems to be unreasonable +The main problem was we didnt return scanner to the previous state even if our rule didnt recognize the token +only exist for Java Projects with a Groovy +to be applied inside the core script that tests whether the log output contains passwordsrequires and ant +I messed around in the widget tests and didnt see any problem. +Is your point that we should figure out some other way to make the underlying FS allocate blocksThat will be great but I suspect that the jvm does not give you this feature +So I took liberty of refactoringthings quite a bit +Attaching an initial implementation +The will then use that return value to update the value of the Holder object +Templates are optional will create VM based on disk image using default values +Then bulk inserts could enjoy the performance advantages of the IMPORT procedures and be driven by arbitrary table functions +I dont know if folks think this is generally useful +I hope that it is possible to replace the marked dependencies it works with now with something like all the compile scope dependencies +I looked through the changes and they seem ok +Thanks for your commentThe attached screenshot looks really greatIll test it +Issue could be resolved but may not be see for further details. +Same error +Best thing to do for now is to mark each copied section with a comment +Marking closed. +But Robert put that code there and hes been right about everything else on that issue and that code +Reworked patch committed +pending jenkins run with the combo in trunk +The operational meaning of these GPS coordinates will be a matter of business implementation and need not be reflected in the data model +hiram I think the best way to solve this is to use the DNS provided by the registrar that you got the domains from and point those at the IP address of ve used this mechanism before on other sites like dont think bob wants to do DNS if he can help it +I use this feature and it is one of the main reasons I am not switching to groovy or the like for the time being +If you only have www changes I would suggest creating a new project and copying your old www into the new project and also your plugins +Hi Konstantin! Could you please point me to documentation tell me how to actually run this tool in order to test the changes +Thanks a lot for taking a lookok for me to merge the test with existing one +Awesome thanks Shane thats greatWhat kind of features are missing +Rename property in docs to match object +Original patch reworked as unified patch +Simple fix +Use for indexing instead +I was just thinking about removing the old queries when has been changed so that the new queries get transformed correctlyWould it be ok if I go with option for now but create a new JIRA for the change to? I might even try to fix it myself at some point but right now I would like to limit the scope of this patch if at all possible +Thanks Deva for the patchAnyway null can not be registered +There are other possibilities also we could run the enhancer or the mappingtool or the schematool or some thingMaybe if we did java jar version we should output the version +Here are my findings using it together with Maven Java and Win XP SP +Sorry for the noise +modulessrcmainjavaorgapacheairavataxbayaregistrybrowsermodel +Minimal install JBT Central into Juno Java SR +There are still warnings and examining the warnings prompts errors +Tom C pushed this for me +Do you still experience this problem +It is a recurring situation for us we keep getting this situation two to three times in a day +So in the case of ICU I would suggest to create a bundle that embeds the lucenesolr ICU modules +Yep it was a jaxb version issue +All regression tests passed with the patch +Added some more detail on and the key generation strategy in the stateful retry +The NM would give it a configurable amount of time say seconds to do the cleanup and then if it has not already exited it will shoot itThe RM would need a new variable when the AM is submitted to indicate that this should happen and then if that is set it would turn on cleanup on kill for the AM when it is launched and it would turn on cleanup on bad exit when it is launching the AM for the last retryThe MR AM would have to be modified to look for the environment variable and only do cleanup if it sees it +This is extremely confusing when you see entity definitions and config screens and you go to a lot of work to configure them all then find that all your work isnt used +Save To mode is used to select a directory and Save As mode is used to select a file +No more packages are arriving to fulfill the request so it fails +Is this issue covering only the numberguess example +Id also like to see side effect files jobs secondary sort keys allowable failure percents +I will fix other scripts as they come +I am seeing the exact same failure and have attached my own log +Sorry I misread copying schemaTool is fine +I know you said a user can move them but it should be visually easier not something the user has to do +Ill follow that workaround. +HeinieThanks for your contribution +patch addressing review comments +I was originally concerned about itself and you confused meI was only thinking to change the install to signal to the people that previously needed the with vars for on linux but will no longer need that +Torsten et al +Stopping work on this +Thanks for addressing my feedback Bikas +It appears as though the implementation requires Java or higher correct +Then place this file on the classpath in the root of the classes directory is usually a good can extract this file using jar command likejar xvf Downloadsdistmodules orgspringframeworkjdbcsupport +binds a set of variables and generates a object which then runs the bound pipeline +This could be caused by some problems in Eclipses PDE +As I understand there is no problem right now? +Dan +I think we should go ahead with this change for +Test case simulated the inaccessible you apply the fix I did above is it now solving your problem? If not have you managed to find a solution +IIRC a will set the flag for all added Configurations automatically +All right back to the drawing board I will try to figure it out +I think this is out of date +I updated the content for the beta release +As you can figure out from reading this issue there is a lot of work to make this code readable and maintainableThe missing break is most probably intended but the code should not be written that way nevertheless so I would refrain adding comments on constructs that should anyways be refactored +thanks pat. +Sorry to bother you guyes +This patch goes against srcmainresourcesorgapachetikamime is the change to the parser module which recognizes the ANPA parserThis patch goes against srcmainresourcesservices file which parses and categorizes the ANPA wire feedsThis gets added to srcmainjavaorgapachetikaparseriptc for the patch Ive applied it in rHowever we do still need a unit test for this +Fix typo in ticket tile. +But you could still define the exact size if you wanted for performance purposes +Well probably want to revisit the overall threading landscape after. +Wont change the name of the file in a point release. +It might be nice eventually to add a hook for client code to track these in some way since we display the number of errored objects on the UI +Ive had good luck emailing when Ive run into issues with in the past +The signature is like a unique key and modifying it seems like a rare +Heres an updated patch with a bit of cleanup +Can you provide a more detailed explanation of how you want this to behave it is too vague understand the problem I still have no idea what youre suggestion as the solution but patches welcome +Can you please document all the changes you have had to make to get this to work? The SOA guys will almost certainly have to go through the same process when they build the next SOA platform and will appreciate any help +Closing all tickets +Ahhh that looks great Steve +lots of people will uwe will but I dont see why this is any issue for nice catch Martijn +Are you sure the download page is a Confluence page +Looking at portlets and its not clear what has been added that makes the installation more stable +Make implement Executor no longer implement Wire the logic for removing the oldest segment is exposed in bytes now +I use this technique to store the users selections even if she disables the +I would need it in my application too and now I need to do ugly workarounds because the lack of thisAlso it is very strange that only collections are supported with the mechanism this should be a generic facility working with any kind of relation +Yeah I am finding an easy way to test from an IDE +Scenario will regress if we move processing of to connection time +or am I wrong about that +mgryszkoGmail sends emails with an envelope sender address set which is probably not subscribed in the mailing software ezmlmIf you have an account you can subscribe all addresses and then just set the gmail account to allowed posting so your mail will go through +Ive made it critical we dont really do the whole blocker thing critical is fine itll get the attention with this. +On other servers the specific home interface for the target SLSB might be ve just moved the homeInterface property and the narrow call down to +Now the Maven JSF configurator doesnt change the file by defaultThe user can change this behavior by checking a preference shown in the screenshotCommitted to the trunk and +I have just seen that this issue duplicates an older issue +Treat unscheduled issues as Native SOAP stack doesnt support JAXWS any more +We have tested on latin in such create your database in latin encoding and use dialect for JCR. +Patch for applied +Its interesting to know that the error changes if the enum is inside an interface instead of being the direct parameter +That means that roo RC which ships with the unpatched jline jar wont work properly with the jline jar on Windows after adding the jansi and jna ve attached my patched version of the source and an updated version of the jar file that users can use to replace their copy with that will enable properly functioning color support +When live nodes change the cluster state is sanitized and leader flags for shards that live in nodes other than the ones listed in live nodes are removed +It is a SAX parser +This makes it at least possible to fix +bug +localhostin the etchosts file +If we assume then ASCII strings are decoded improperly +And yes I think the best solution should be to have register a listener on the catalog that handles purging of the cache +I would be concerned about the performance hit for such a feature that while only slight will probably never be used +Its ok when there is option skipTests +A couple of recursive calls were added in and they all seem to cause that kind of problem +The problem is in PMD itself and we are using the latest available release +I didnt add too much logic in the collector itself and it merely collects error messages and present in the end after the logical plan is materializedPlease review the pull request and let me know what you think +kind of bad +This is temp version of AS as it includes a bug fix which causes tests to failAtm you need to build your own AS instance and it will workThis will be fixed as soon as next AS release is out +Yeah as I said before you will have to use conditions to check if theres user input or just show the think I have to add some documentation about that and at least remote the when the UI is first presented line +m repo and rebuilding should help +I committed the patch +So again one can bind with the users credentialsTomcat supports this pattern see the user attribute dont know if Geronimo does +Marking this patch PA +Since this information is effectively fixed for the duration of the session and NAMESPACE is only available after login this will be addded to +Or perhaps Im confused +same to trunk +As you can see the edit box is behind the logo so I can type in there but have to scroll left and right to actually see what I type +So it cannot be removed until. +Attachment has been added with description transition to closed if someone wishes to restore one of these to resolved in order to perform a verification step then feel free to do so +FrankI tried the Reset Password in QA and I received an email with a new password +No IT +Committed in rev +Mikhail thank you very much for the investigation and the test +were seeing success with this in the field +Roo uses the hint command to offer instructions a little like a tutorial +Not sure how yet. +Even if its all deleted immediately afterwards its still useful to have this fix in place for comparison against s implementation +Fixed by moving files to resource directories in the test source tree. +It is my understanding that there is no translation tool needed when working with XML filesThe latest revision supports both the custom XML file format and the standard java XML file formatWhile having multiple translations in the same file seems odd at first when you look at one you can see the advantage missing property translations become obvious +Consequently the notifications should be treated separately +pushed to cr +Thanks +I forgot to complete that part of the issue formthanks +HI jacques Sascha Atuli found that when we open a layered lookup from the simple lookup at the time when when we select the value from the find list of the layered lookup the value get set in the main screen field from where the simple lookup called instead of simple lookup field both the lookup get closedIt is because the simple lookup uses the set value function of the not of the then the same function is used by the layered lookup which is inside the simple lookup when we select the value from layered lookup its target got replaced by the simple lookup target and both the lookup get closedIf we use the js function then its giving error and i am working on itThanks +See that JIRA for details. +I am hoping that all serious blockers would get fixed by tomorrow and I can get an RC out by friday +I believe that this has been solved by +I think this is indeed a duplicate of +in now generates unique message id differently that was file unfriendly +IOW we need to warn people to be aware of this +bq +Test run looks good +Fixed in trunk +I think Ill push this off until the next release so we can get out the door pronto +Fixes the problem warning user of weak password when provided a password less than digits +option implementedI have set as deprecatedShould we delete the file from CVS? Delete Nobody is dependent on it in the other projects using +Also in the future can we have this discussion on the list and not in a JIRA ticket? I didnt see it because JIRA issue mails come in by the hundreds +I added the license to the beginning of the logj and other properties files +reopened changes were reverted fix breaks other functionality +Patch looks good but it looks like we still return null in the build the builder section +Since this is a PDF specific issue however lets track it ineverything else here seem to be done. +The goal of this class is to centralize the activity of decoding encoded streams +Netbeans project with jms client +I dont believe this adds anything +is also not in BRANCH +I am not sure how I missed that beforePlease let me know if you find any more issues with thatArunI believe the issues were in terms of recovering the jobs from the point they crashed +Again without presenting credentials this seems like the wrong behavior to me +Ill check it in shortly +Why I did not have problems on one machine and had them on other I have no answer until now +Did you have anything else in mind +I thought was only a dependency because used it +We need to check if theres adequate documentation covering this +There are cases when the makes sense as well as other cases when a is more appropriate +Thanks for reporting turned out to be an issue in the Pattern implementation +I used those to test the Replayer one all automatize and the second requires the response of a financial analystThe first one works fine with the Replayer since there is no correlation +But it would be cleaner if we just stopped the unexpected read lockers of the system catalogsI am coming up with escalate situations for sequences are there more? Here they are parent transaction did the create sequence and possibly in future if we do alter sequence we are incorrectly escalating and should be waitingtiming out ie +Could you please attach the patch tp the issue rather have it inside your comment? Thanks +The problem is actually in the resolution of dependencies +Thanks AtulYour patch is in trunk r r r Jacques. +Do you see any problem other then the warning +If you can think of a way that would be awesome +I should remove this line out +Just a note this patch was made against +Created for adding fault injection tests for the new pipeline close. + +added new FAILAUTH constant to allow for providing feedback about failed credentials extraction this is of course not the same as no credentials in the request +The process image etc +However it is a real world issue +file path +Attaching a patch that saves xsloutput properties in a more genericflexible wayand which knows if the property is a default or value +The only skin that we really support is pelt and even then we are not bothering too much +Fixed with revision +preparing release +Then the point would only be emitted to the closest cluster and many distance calculations could be avoidedId still like to understand the changes you are proposing to the existing algorithms +whatever +see attached +I dont really have an opinion on the top end of the range + The maximum number of simultaneous connections The maximum throughput of messageReceived events The maximum number of unprocessed events which are waiting to be processed in queue +Yay! Great work Nikola Mike +This patch fixed the javadoc warning +Reopening to fix the resolution +Sending srcmainjavaorgapacheservicemixsoapbindingssoapimplAdding srcmainjavaorgapacheservicemixsoapbindingssoapinterceptorsTransmitting file dataCommitted revision +Thats the reason why there are errors in the project examples +This is not a Torquebox issue but an underlying AS issue +Leaving cell values empty is a valid schema choice +If the init method is used with the deprecated arguments the same String is used to set the default IV importantly I updated the patch because I realized I didnt need the second condition in the if statement +Moved code from old into +submitting +If you remove the rich drag indicator from the application it will function normally +Ok +when you say +batch transition resolvedfixed bugs to closedfixed +committed revision Thank you Chattree for the contribution. +for it could be a separated jira to do it +If youre saying it terminated abnormally then you need to provide more specific information like logs and stack traces +It really hurts +Patches regularly go through a vetting process so dont assume that yours would be rejected as long as you are willing to accept guidance concerning the not fitting +I wont have time to look at this in the next few days I consider BEEHIVE higher priroty than +Any objections to just io? If not lets update the patch for that +Does it finish in the end or never? I would guess itll finish but I dont know +Merged to the branch in revision . +Agreed +Is that a phase or a goal? Either way its undocumented AFAICT +pushed upstream +Ill also study the build instructions and set up that environment before submitting future patches +Hi GustavThe problem is that a lot of people will use the modeler for working with their data model and never use it to connect to a DB +Pls ignore the Environment closing stale resolved issues +Ive lost a bit of context here but can someone tell me why that is not enough? Is it the MR AM that is sending the and then doing some cleanup after that? If so may be we should fix that instead +I am creating the project in the and will post the code there +Is disabled why uncompressed went from the highest latency to the lowest +Too late for CP given that we still havent even proven the bug in a unit test much less fixed it +Ian Emmons patch works well for me on Mac and Java +I depends on used proxy technique CGLIB or JDK more convenient way is needed +If the tx updates the key after it will be contained in the tx context and eventually overwrite the L cached value and if it is rolled back then we still have a consistent L cache. +I think the main reason for using the filesystem is that it will work on all servers with and without db +If the build works on that version we could add it to the system requirements +Thanks for doing this +Attached it now +Ill c ontinue investigating +This patch should fix the problem with Tomcat +Here is the patch to implement the new design +ThanksL. +oops my bad +Please post the build error if you think it is related to the patch +I also get No service named +I have added an instruction to do this to section Upgrading an existing SOA Platform installation +Fixed in revision +Thanks a lot +Ive been using the sed awk grep tools for merging logs files to see trends across time +Apologies +Waiting for some feedback on the destiny before adding it to the community release profile as closing all resolved issue that have not gotten any more feedback or comment in the last month +Hi ClausThanks for your inputs much appreciatedCleaned up the Test and added an updateCount to the +Reviewed by Gordon Sim and Alan Conway. +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Easier to see this way than as an actual patch +v adds a log line before running any scheduled task +Ive committed this thanks Colin +Please take a look at in the unit tests +Also if we tweak the header buffer size we should doing it in a configurable way +We will run the full suite of tests on the RC anyway so the patch will be tested even more at the RC +We did that tooCurrently we have no more SSVM and we can not add any secondary storage + +Will need to bump up the openejb version to to fix this +Patch applied thanks Dan +We should not have files for any of our standard services +Attached patch contains the notes +Moving to +performs the functions of a sender if the Spring sender isnt availableThe modified attempts to instantiate and on failure instead instantiatesChanges to the other classes tie them to the local interfaceThese should be accompanied by a change to the Maven configuration so that the dependency on is optional +oops forgot a check in the surrogate case +addFirst This method appears to be only called from test code +What are the Confluence user names for those that will be needing to edit the wiki +The test ran with insane jars so only the most severe errors are logged by default +If there are less than future iterations backfill with past this in the too awkward basket for get a reasonably clean solution we need to change the paging behaviour of the table perhaps using a new parameter something like initally +Youre right +Ive just committed this to trunk andNow Id like to get opinions on how could we do this for pre releasesMy thoughts would be to create a branch with shell maven project where we can add the and for previous releasesThoughts +Hi ThomasI have created an app using the script from your attached sample application and have not been able to reproduce the issue +forced to show in the same folder as Palette and Outline views perspective needs to be reseted before testingthumbnail +limit and offset clause date time and timestamp literals as modulo operator at arithmetictermWill address this one by one soon +Itd be nice to have a shell command too +However if the single attempt fails we cancel the attempt and that is the end of it +The class has been created +Any objections to committing +Although I didnt compile it and test itIll give it a more thorough view in few days. +Pushing into Maven Central is perfectly acceptable +Filesystem has changed but no migration script yet to bring over data +This includes the EncoderDecoder improvement as in the patch and the Consumer and Producer console options to set character encoding manually and place it into the relevant Properties objects for consumption inside Kafka +Please close if satisfied +Makes assumption that common has already been built using a separate this target seems to be serving purposes kinda unrelated to the component in question Id suggest to at least remove its description to hide it from targets list +Added some more construction options to the class +the usage info is perfect goes to stderr and exits with just like any other unix tool g +I have manually tested that the fix resolves the have left the JIRA as unresolved since we need a test for it +From screenshots Im not quite sure I prefer the new version +I dont think we must retain it at this stage +So we should get a new fixed client jar pretty soon +Merged to the branch in revision . +Jarek recently put a patch in for that was committed to fix injection but they have not published it to their repo yetSo in order to run this I had to build locally then build Geronimo trunk +If you click dont show me again you wont be bothered by the dialog +The sooner it is integrated with the better +No functional changes are included and therefore no tests are included +nice code +bq +Also add a note to the reference a solution we will expose the source as a dedicated bean using the bean name +Thank you for your commentI will try to upload files during the next regards login Michael Furman +cr or GA +Introduction of does not make any difference +I added logic to make the and returned by pluggable +I agree that we need to test performance impact of BLOCK compression before committing such a change +As long as the JSON coming back is valid Im a happy camper and can deal with accidentally indexing invalid data +Just because you say install from the mirror doesnt mean you will if p decides it can pull from a really slow mirror in China instead of the fast one in the US +Test application which triggers described issue +Lars good stuff +patch showing the initial good to me two comments Maybe you could add some dynamic stuff to to verify that it gets the correct scripting context not sure if its needed at this stage Id remove the at the end of the test for test ok I didnt realize you were nuking the whole apps that should initial test in rev +Vlado can you attache the screenshotI cant reproduce you have CDI validator enabled for your workspaceproject +It says The key for the Message ID +Consecutively run test timesAll of which passed +If not please come back to me and reopen or drop a line at the mailinglist +The console does not fail silently +Heres a patch +committed thanks! +Thanks Stack Ted for reviewing the patch +Fails QA see comment above +Thats one problem +Attached patch with article userctx and request objects description with few examples +So all issues not directly assigned to an AS release are being closed +There had only been one method that once had been copied into it from some Sun source together with the misleading license text +Thank you GoranIve managed to reproduce the issue by creating a new app and copying the necessary filesWith Jochen weve narrowed down the changes to Groovy that triggered this problem by bisecting till finding the right commit that was causing the issueInterestingly the commit which was fixing a problem Luke had with his remote closures project but which covered some other issues that youve been facingWe found a fix for that already well release another Groovy release next week with the fix +I have just joined the mailing list and will post here again regarding this issue +We want to apply these convenience patterns to US since they are causing confusion with some other locales like ja chinese +Spyros Chris has agreed to write some unit tests for Hadoop +Unchecking the crash box +I have no idea what your whitelist looks like and this is an issue related to build not Apache Cordova +Yep youre right I didnt realize the use of ns above was meant literally as a for a full namespace that would have to map to two different schemas and thats certainly wrong +Hi Frederic any chance you could attach your WSDL to this issue with your configuration? Or you could also send it to me privately dan AT envoisolutions dot com +Good catch! I was passing in the test explicitly like so mvn Dtest +Probably makes sense to do it +Sample differently in ACF then Railo +It should be up to the user to clone the objects added to the cache. +has conflicts with in trunk +Please review +Any opinions whether using the aspect for job interruption logic is the way we want to go and how it fits into the release schedule +Unassigning myself not actively working on this +I have not yet tried myself +I receive the document and then I extract the imageIll send you an example +Will commit once tests pass +Thanks Brandon! +after release +rebased the patch on current trunk +New patch that updates HADOOPCLASSPATH and doesnt change tests except adding new tests and +this fix consists only on patching the website +So I just ran into a problem running rolling restart where users were creating tables against the cluster +The purpose for the lock file is to hold a lock on that file while the is running +There shouldnt be any server configurationsince the labels are already on the left of the chart I missing something +Completed in +Regardless your fix looks good +We should fix that bug for and address the larger issue of improved progress reporting during sort for +make a minor change in test class by calling setUp and tearDown on each test case to follow the convention +Can you please create a sample bundle that I can try in our distrothanks +The latest patch is still not applied to the code base +fixes both to trunk and to +Proposing for +So I think we should stick with http but either add comment and alternative ftp url in comment or preferrably kick the server admin so the misconfiguration is removed +Close all issues. +Worked with an inconsistent build taken from Hudson. +Id like to take a close look at curator before I weigh in on this +We could also provide a path to access a raw for Power Users. +Part of a bulk update all resolutions changed to done please review history to original resolution type +If you finish it before we release beta please correct the versionThanks +NOTE Im running this from my personal build of Solr from my IDE so the classpath will be unresolved +I am guessing src is the one to test with. +BTW +But these are just examples +Committed patch in r +Jonathan any update on this +Verified by Andrew. +This IT passes +Attaching patch a which modifies a link pointing to one of the symlinksCommitted to trunk with revision +The attached is the initial proposal +pulling this off PA given that Thomas is working on another approach to this in a separate set of jiras +I made an issue for JuhaniSorry that I did not yet provide any feedback +We had another web app with on older version +Patch committed to branch +full esb has a fuse mq standalone broker configured etc is responsible for thebroker called you want a fabric controlled container then you need to remove this entry from the etc or mv to a differentunknown pid eg mv etc etc we have this in the docs anywhere? Couldnt find it +This may be appropriate for simple actions while a design is finalized for async execution similar to the agents +You can ignore the core tests failure as Hudson is having some problem with the tests +So you could create a script that runs after the page loads to move DOM elements around +Just not in junit running from maven build +The MR gets from the AM again its up to this to enforce them +I am experiencing this problem as well +Patch committed +Thanks +I totally understand the issue at hand but did we investigate how hard it is do the UI change to not load all the applications? May be that is the first issue to fix? Once we can fix that increasing apps is an obvious solutionCompleted applications are explicitly managed by +fixed in this since has been released +But it appears for fresh installation +Yes its my mistake Im overlooked this property +the link is dangling +This is a unix specific bean so dont forget to defang the test if the bean isnt available +Asynchronous is fire and forget +Let me know if you make any progress Eric Id like this to make +Is there a way to get source with this patch and test it immediately +Thomas It is difficult for me to understand from your description exactly what the problem is +If it is RI bug then we need to move this JIRA to the category +Canceling for now +dataTable subTable dataGrid onclick onmousemove onmouseover etc +Marking as resolved. +So that apparently has nothing to do with thisSuggestionsThanks Guy +Upload unit tests to reproduce the issue while clearing the blocks +Johns explanation makes sense +With the fix the correct rows are returned as the flattening is skippedMy approach is possibly too brute and might be relaxed but Ill post a proper patch tomorrow morning +Standalone CXF test case that works +We are working on the new Hibernate Tools core and plugins to fix the issues. +Thanks would you please do an svn update and verify that the accented character is okay +Eugene the test patch commit was rolled back due to the breaks in all other platforms than Win +I think you will have todo more research to see if theory is good +Attaching a patch that sets everything up +We could probably stand to do that for all the eventsAs an added bonus the gui seems much snappier and responsive +My work did break Grant work heresHere is a patch to correct thisI put the logic into the instead of the as there was allready code to handle Tier level borders there +bq +When I post a patch I usually run and derbyall using the following scriptjava client true XXM Dcderbyoldreleasejars org +since we have merged lucene and solr and chris has fixed analyzer to have a performant api not by experts but by default I think we can mark this issue resolved +If possible lets schedule to didnt test this patch but we can simply add combiner function like thisThen the communication can be minimized since only the summarized values are exchanged among the peers in each superstep +In this case we are talking about Hibernate mapping files and I am not sure whether we considered overriding there +There is content under view and help I didnt really go into the portlet menu stuff to much. +Closing as cant reproduce +Also checking for null and executing the operation are not atomic so it can still break +Thanks patch applied. +I am not sure this is exactly right +Can you please post the complete stack trace +I should have mentioned that I tested against current versions not. +reopen to set proper fix proper fix for fix version change closing. +Can you post the code in +Thanks +I just created which depends on this issue +At least I like to add version number in my war files like +But what if you are not connected working locally no internets +If you find further issues please create a new jira and link it to this one. +Just cleaning house and pinging this issue +Okay I see it in nodetool netstats +Changes to get and import checkpoint workingIvan +ok so there is a functionisbgfillnecessary that lets TS know if it should do a background fill +This patch adds support for accessMode and dirtyCheck +are you still working on this Jonas or could someone else dive in +Marked as iixed as Christian has updated the issue. +Hi HiranyaI added the based configuration deployment for the DLC +They may survive but Rails has shamed them +Attaching +The only drawbacks like we mentioned above is that youll need to add yet another request parameter +I am still investigating +In the datasource definition for version an incorrect URL was being serverinstances now deploy correctly and the tests now pass +I will think over simplifying evolving this file and get back again +Sorry I was wrong if GPL is not Apache compliant LGPL can be wrapped. +Attached is an updated patch that updates the bnd dependency to the fixed version along with the original fix to transform the XML friendly prefix to It would be great to get this patch in soon as there are a number of projects Id like to move over to the bundle plugin but they require this fix to import external manifest files +The test breaks when checking for a contiguous chinese string my unfortunately breaks it +Will open a new Jira to deprecate the List methods in Chunker +Thanks Nicholas for the Work +It should be good enough to be used in New an Noteworthy commit the changes until is resolved + +Nicolasthanks for the patch committed at r +When there is a connection problem any pending packets get set to connectionloss but the does not exist +This patch fixes the issue +Sample applications to demonstrate that between a war and an work +I will run the hcat tests manually +This effort may be probably better suited to the code base +Its all there somewhereProbably I can submit an example for configuring SSL on javafirst web service using embedded Jetty server if needed +Example web service +In addition the doc saysAndroid Quirks The Android simulators will not return a geolocation result unless the enable option is set to true +This is great for merging +Check Apache issue for more fix version to reflect back port to +The warning output should no longer happen in since the class now closes the directly via close and does not use the Batcher methods which no longer exists +Good idea would that be best put on here or on the wiki at bitbucket + this is very confusing +The only part still missing is documentation but otherwise it should be fine +Do you know if is part of the class library for the VM? +The code is a short now +See Comment have committed fix for Service Category as +So i would recommend we consider adding some sugar to indexreaderThats unrelated BUT yes please! And you already noted in the signature one is very important This method must be FINAL in IR +Brian I think that is exactly what we should do +Tests pass happily need to dig into the implementation code +Did you remove the old one from your classpath? Run java verbose and see what your classpath actually looks like I think that should tell you +fix fix Rich Faces Components in VPE +I am running tests nowI also added a login timeout to the +This is deja vu of +I think its possible to do it from Jenkins +What testing did you carry out on thisAgree the documentation is excellent +Java hasnt been released yet so I hope the return value will be fixed by then +Well change the API for commit only +You might want to run litmus with the k switch so that it keeps going even it some tests fail +As soon as i figure out how to do that i will post the code here +Fixed +Im hoping this can be done in itself since its a very quick fix +Richard you can revert the patch from in your local build and try to get the Spring WS upgraded +I have the file follow by your suggestion change sign to unsubscribe wordRegardsChatree created a email template for contact lists communication event +So this bug is fixed and can probably be closed with this works in as you can see in AS well have a look in the future to see if we can improve it to be accurate +Cancel Upload button doesnt clear the listBTW I agree after upload files must be cleaned +I am working on meclipse plugin and we have extensible project configuration framework that could be integrated with Spring IDE to generate any required projects configuration either using information from Mavens or any other metadata +pushed to In your opinion is this blocking the release? If not lets change the priority to major +I cant remember the particular JIRA either but it seems to me that the regionserver shouldnt even get to the point of doing recovery if the logs havent been completely recovered +Fixed in revision to add the right Fix Version +Actually the version in Romans extra patch should be rather than to trunk and branches +is an enhancement +Tom so whats the best way to encodeescape these bad characters in SOAP message while ensuring the message on the wire is interoperable with other SOAP stacks +I think this bug is fixed +The initial docs were already added. +Please stay tuned +Also note that having the WSDL based ws translator allows for CXF timeouts to be specified at a port level rather than at a service level which minimizes the need to have an explicit timeout parameterinvokeHTTP does not use any timeout now +Of course I have no access to that so I will get that part as complete as I can before submitting a patch +I have ruby plaforms registered in netbeans builtin RC and another +Please add a release notes flag if this upgrade is complete for +Closing all resolved tickets from or older +Im happy with any solution you can come up with for it +Its not about source of truth +Patch fileIm marking this as fix for for the backwards compatibility reasons outlined +A patch creating a test for the case described by Bryan +Go to Preferences Tools Tools Runtime Detection none of the locations will be in the list FAIL Fixed +Theyre in the jar +What about the esb plugin update? Doesnt have the user to the when updating the esb plugin? This means that all historic data is lost right +I think its a timing issue +I will committ it shortly +It absolutely does not make sense to set params to null but Ill fix ths anyway +Different versions have different expected values which does not seem to make sense +I verified mvn clean install +So now Im running a test where I insert extra whitespace one character at a time in front up to of them to see if I can tickle the bug +What a pickle +Good job tracking down these threading issues +The safest thing to do is to add more methods to the class representing the clearer API you propose and mark as deprecated the old ones +We should postpone this as long as possible because it will make reviewing patches created before very difficult +I closed this rubbish issue once already + +For example activemqtcpactivemqhost + Bug has been marked as a duplicate of this bug +Attachment has been added with description SVN patch P open bugs to P pending further priority for bugs with a patch +Fix version should be provided while creating jira issue +End streclaimlongcol storeallstoretests +All unit tests pass +I have no clue why I cant reproduce this today +This is not an EE issue but a groovy and groovy issue +Now that is in I can start looking into implementing this. +What browserOS you on? Do you have HTML refresh turned off +Assigning to Jonathan as a proxy for Cal +Theres no major difference in between and mode as far as execution goes and rcov has been kept working for a long time nowI will poke at it a bit right now +Second issue is even there is constraint not to use for ws tester manifest eclipse uses it +We dont want to createdrop the test database for every +HelloIs this for Camel SJMS or Camel JMS? I totally agree that SJMS must add JTA support +Do you use a publishWSDL element on your proxy +Thanks VenkatThanks for good explanationsI just believed that namespace of type used by service is same as schema namespaceAnd I didnt know package of class was used for namespace +I created an other issue for TrongNo +Then I run into the issue reported by +Ie +Synced with the latest from +Hi MichelThank you for the contribution +adding the full stack trace of isequivalent when the statement got seems the code may not be reachable so no need to backport. +If it needs to be public for other use cases fine but generally speaking I dont think variables etc +Both are collections with a replicaIf thats the case you should be able to get it to work using the actual core names +But I have code that go right at the whose default strategy is to new the default +I just updated the patch based on your commentsTom +I will look into getting a snapshot released +Hi OysteinI looked at jdbc and it looks like a different testHow about changing jdbc to jdbcThanks am not sure what you mean by looks like a different test +Ive attached a logo press kit containing the logo variants as separate imagesNote logos are large size I assume you can scale them down to a smaller size yourself +And there is no way to specify which encoding is usedThe next issue is fastbinary +Otherwise we get the symptom explained in the trace JrgenIll try it and Ill tell you if it De La JrgenI have tested it and it Works +The fix is indeed trivial +Hello AswathThanks for your keen interest in our patchHere is the answer to your queryHUMANRESAPPROVER security group contains HUMANRESAPPROVE permission +Seems to be in our build now last I checked +Lowering the priority since it occurs on only page +But is that the complete fix? For any othernew jni based function we will be back to this situation because on Windows the native code is always loaded. +In version the issue should be fixed +Exactly the same IIS logNet +So Id commit this as it is a blocker for and we can consider the issue later +patch looks goodBTW if youll keep the same name of an attachment file JIRA will show a older versions in gray font and the newer one in more bright black one thus youd be able to tell which version of the attachment is the newest +Verified the RN text fix. +opps forgot to set to resolved when I s +can we check this in? i meant we could test the setting of the correct serde from a java test program +Yes I really like the idea of having timeout in wait when and connect to a server +So then what do your role and user files look like precisely +Do they get a new ID +As Navis pointed out if reduction ratio is not high enough mappers may run out of memory and than we suggest users to aggregation +ThanksTorben +Sorry about the confusion +I cant reproduce this +Lets pretend it was never uploaded +Updated +Fixed on and trunk +As a consequence we have no plans to reintroduce or any of our former LOB user types all of which were classes to work around issues that Hibernate should really solve itself since they have nothing to do with Spring and caused repeated maintenance headaches on Springs re free to create your own version based on Springs old of course +Thanks for fixing this +on windoze and had no problem +bq +Jeff We have some notion already of priority where +How about an enum parameter to add that determines the lifecycle of the object +I too have figured out lately that for these type of locks mode is much better performance +Thanks for the patch +updated the jfoundation canon with revision +Needed CSS changes have been made +But try to change deploy path +Sorry again! Not having a good time! Unnecessary debug print introduced in previous patch + +License Headers are Not Mandatory in Every Source FileLicense Headers are required for source files in most casesA short License Header can be used instead of the full License Header for small files with significant creativity where the size of the License Header would detract from the usability of the fileNo License Header is required for small files that have little creativity or little value and a License Header would detract from the usability of the fileWordsmithing the above will need a more focused discussion in a different forumTo Larrys See above for my recommendationTo Larrys Im not suggesting dropping copyright notices from distributions +Its back +Snjezana what you found I would say is a different issue than this since the server respondsworks for me it is the eclipse UI that is slowing down for meJim +Tested the patch in my local cluster and the duplicated error msg is gone +Can we stick it in the wiki or generate a PDF from it and post it as a standalone guide? +Step changes to support bloom filters +made suggested changes +Committed branch and trunk +Is this being included in changes +Hiding it may show up as problems later +Attached is a new patchFirst you have to apply then BTWEAP ER and EAP ER include the version of this artifact +also tags each message with an application you have a perf test that reproduces the issueAre you sure youre not sending with multiple threads or tag your messages as OOBIf NAKACK delivered messages in an incorrect order that would be a big bug and Im interested in getting to the bottom of it +Verified in EAP +checkPeriod probably doesnt need to be changed at all in insert since the checkingPeriod method coped with dynamic vs normal anyway +Thanks I applied the patch Michael + contains morepublic static String HOME public static String WORKSPACE public static String WORKSPACENAME public static String MODELVERSION version private static String ECOMHOME private static String ECOMDATA and these constants are unfortunately used many places without actually referring to +I think it might have to do with bitbit +Seems reasonable to me +should mean zero or more NOT at least? +I dont think it makes much sense for the parser to be checking on the filesystem and this allows the creation of Slicers that donthave any files on the filesystem like the one I made for the test +ACTIVEMQHOME was pointing to the old M releaseCould not reproduce it anymore with M release and correct +Sorry for not looking at this so long +claims to have been fixed on and but the test is not there in the branch +Onkar Im assuming you figured this out and the issue is resolvedDo let me know if youve still got problems and we can reopen this issue +I donwgraded the issue from Major to Minor and set a due date to +In this case I think that we can summarize what Derby implements of the interface +Thanks Namit! +Does this patch work if localhost is specified as the host to listen onJust wondering because the security documentation has the permission listen which is only used for localhost so I wonder if accept works when listening on localhost +Hi there Sreekanth The barcode scanner api is not apart of the cordova api so I dont think filing a ticket here is the right place for this +Closing +Let me provide you with some details +Documentation for CLI and stress wasnt updated to list +The Project preferences section has been updated and Seam Generate Entities as well. +Next we move on to Tomcat which by the way needs a separate issue report +I can see where Daves coming from but Shinyas patch doesnt seem to make the code very much more complicated to meWhat Im most curious about is why this method is using the global newdelete and not the pluggable memory mechanisms? Dave Id be particulary curious about your take on that +Resolved in revision +Though this does not seem to be related to as i do not have any large views which timeout while updating +So itd be good to make that clear in the new issue +there may be two situtions cause mkdir failedtoo deep we should tag this dir as a final dir do not try to create subdirspace full do nothing +IndeedWe will investigate thisThanks for reopening the issue and providing the code to reproduce it +ThanksAaron +did enough of this. +Updated patch fixing the remaining warnings +The site has really bad caching +Thanks! Look at New Server wizard screenshot +Ill be more careful next time +I hope you can figure it out but I really dont know what more to say +Target +I am not an expert on the scheduler code so I have not done an in depth review of the patch +The wiki information is great and should be very helpful when I have time to transfer it to the docs +We really should be compact no doubt but can we suffer the pain of a somewhat weird way of reaching that? What if we just switched the default leaning on the RFC wording documented it clearly as a breaking change released and waited for the explosions? +Are there any plans to also publish the new release to the Maven repository on? We at Jackrabbit still use Maven as our build tool. +Its patch set in codereviewThanks +I can only imagine because knows about it or how to use it +Internally theres an event dispatcher which starts when starts + +I cant see the linux option in GCC the references I find still refer to m +Bulk move plan to delete this old defunct page and commit shortly +I have a patch for this ready but still wonder where to put the implementations directly in make a subpackage for each space +feedback again would like to see this more documented +Let me keep that open until I have the patch +Cause Missing maps +Bulk close for Solr +Hence closing this as cannot be reproduced. +modified sample thing to note the wsdl was derived as I took the java first approach +The full fix requires the following fix from repo ce +So the error obviously doesnt occur anymore +Im not really sure what purpose that combo serves or why theres two categories in the drop down +Ive made a test with your idea but it does not work very smoothly +Done using JSF. +We need to have a mechanism for the newly added journal service to get the missing edit logs from the active journal service daemons +Javadoc for and as well as miscellaneous changes for correctness including removed some throws clauses that were spurious or misleading +I have attached a newer patch that contains fixes for +Check if Jenkins like it +bcall You taking this one +on the idea +Missing file names for attachementsID XERCESCROOTsrcxercescutilPlatformsID XERCESCROOTsrcxercescutilPlatformsID XERCESCROOTsrcxercescutilPlatforms samples works fine out of the box +For example it requires +Closed Issue +No comment means no strong opinion against this? +The current title sounds much too harmlessImplemented a workaround to lock the attachments and cache them. +While you are fine with dependency on the Hadoop jar could you provide insight as to whether your users would be fine with it tooThanks +Can I be granted permission to work on it + to the site! Thank you for doing this +Please assign this back to me when is resolved issues for already released versions +On my side there are two things that should be changed in this file change the first linebinsh ebybinshin fact i tested Linux Fedora and Linux Mandriva and all not succeed to start with the e flagIf i do with the original init scriptoptcouchdbetccouchdb startNothing is startedif i do with the original scriptbinsh optcouchdbetccouchdb startI get aStarting database server couchdbAnd couch is indeed startedIf i edit the init script and remove the e flag and launchoptcouchdbetccouchdb startI get aStarting database server couchdbAnd couch is indeed startedI look in all the scripts in etc and none have the e flag it should be necesssary to add COUCHDBUSERcouchdbin the init script to reflect the security consideration in the README file in order to not start couch as a root user e line should be POSIX compatibleDebian mandates the use of POSIX compatible shell scripts so that the shell canbe changed from bash sh ash or dash +So a contribution to the DTA could check for names ending in password etc +Fixed as requested in last comment +I dont know of any way to make Struts ignore certain requests. +For the completed capabilities theyre all based on scheduled polling +Moving back to for now. +I have attached a logo which might be a good candidate +I take it back history is a mess wo topgit +There is now a component with docs and osgi features file etc +If you still want this please submit a Pull Request in +bq +Up to youThanks for addressing my feedback in. +From the description of the problem its not clear to me where the fix needs to be applied +This test demonstrates the dont know if there is a more elegant way to do it but the applied patch fixes the is a general problem of the docbook parser see it should be fixed there. +However if you change the quality width or height the code tries to figure out the real path to the image +See comment by Bharat in the related doc bug +IMO there is no way to do this as it is done by the target OS +Might be useful to somebody else tearing their hair outRgdsAssaf +Committed revision in HEADCommitted revision in Trunk codeThank you zm for contribution to Nutch. +I have tested on the slightly modified helloworld locating the reference directly +Upping priority some +How the content url can be configured? For the seo perspective I feel the url should be DomainNamecontentcontentname +IMO a ary index CF should inherit the compression setting for its parent +Trunk and maintenance +Where a shipping address isnt available the billing address should be used and when that is not available no tax should be applied +Fixed. +Have you looked at the newest design doc in ? Its changed so that has a smaller role and is only ever used to create input and output streams +Thanks! +On the other hand the would access such property filesThere is a chancedanger of collision though which would be resolved as follows A Branding service always has more precedence as branding properties If multiple bundles are providing branding properties the properties from the bundle with the lowest ID are taken Branding services are ordered according to the spec by their and the service is more flexible I think we should just pick one option for all extensibility issues for now therefore I would favour the optional import stuff with property files +Ill do some tests today +Sorry it took so long +but it doesnt work in our program +Added comment to code explaining how to avoid this ie escape with . +Catalog was updated +Alfonso are you happy with this issue? Are you willing to resolve +The test failures are not related to this patch +the library choice is a better solution especially as we move to a common t use +The patch looks component also uses this plugin +Regarding the version perhaps it should be? just to leave the incubating suffix for releases +Which Jackson version is this in +I can review really quickly and commit your patch after thatThanks +thanks +can you please verify if you can reproduce this problem with the new code? note that meclipse requires wtp +scenarios log into QA as fmerenda redhat create a new case validate status is New log into issue tracker as fmerenda redhat flip the status to Waiting on Customer go back to the CSP and comment on the case validate that the status is changed to Waiting on Red Hat log into QA as fmerenda redhat create a new case validate status is New comment on the case as the customer validate that status is still New add an attachment to the case validate that the status is still New log into issue tracker as fmerenda redhat flip the status to Waiting on Customer go back to the CSP and comment on the case validate that the status is changed to Waiting on Red Hat log into QA as fmerenda redhat create a new case validate status is New add an attachment to the case validate that the status is still New log into issue tracker as fmerenda redhat flip the status to Waiting on Customer go back to the CSP and comment on the case validate that the status is changed to Waiting on Red Hat log into QA as fmerenda redhat create a new case validate status is New add an attachment to the case validate that the status is still New comment on the case as the customer validate that status is still New log into issue tracker as fmerenda redhat flip the status to Waiting on Customer go back to the CSP and comment on the case validate that the status is changed to Waiting on Red Hat +Fixed +was the same issue as this but I assumed you closed the issue down as it contained some wrong information. +Close as duplicate of +orgrhqThe and are the most problematic the rest have Nexus in a separate profile which can be deactivated +I still dont understand how create in my patch is different from other operations marked as idempotent +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Reassigning to Thomas +I just noticed that we have a that tests just the localized output of Sysinfo +one multiple word tag second multiple word tag singlewordtagwe had debated this in the team and found this to be the most intuitive way +Thanks Corinne! +The patch has been appliedPlease verify +Can I simple update the JAR file in Final? Where I can download Weld or what is the Maven dependency for thatYou can close this issue if its working with the last Weld version +No bzip really is block oriented so it may require reading more data than is actually available from the underlying stream +This should be discussed on the list +Yes we have multiple machines being able to write to the same index on the SAN +Closing as duplicate of +Attaching a patch that introduces edge managers to manage the data movement and edge connections +Crashes due to cext are now Wont Fix since cext is deprecated +This work needs to be done as part of +Ignore above comment +I suspect its a rounding error +rev correctly reads the JNP URL from SERVERHOMEdata if it is present +Committed revision +Hellowe were using hibernate with the above patch apllied for a while now +All code has been brought into master. + +This was implemented for Drools +this is test eclipse you test this patch +fixed and an IT was added to the plugin +Raised to deal with review changes. +In that case cant we find only components that uploaded a file? I thought we could +Weve now moved away from using +Josh are you able to apply this patch and try it out? Its not ready to be committed but it adds a bit of defensive programming that should hopefully fix the error youre seeing +attached a wrong file +Its you guys call +Trunk version +Except in both cases usingreflection invalidates the protection +So all issues not directly assigned to an AS release are being closed +However most tests should clean up as much as possible I think +Thank you so much for thisNestor +for now +svn ci Sending Transmitting file dataCommitted revision Thanks Anita! +I was going to release today + contains the updated code +A possible patch for A problem with this patch is that if the script is invoked from a location other than the the bin subdirectory of the build directory the script will fail to call the locale and localedef utilities as it expects them in the current working directoryThis will likely be a problem when manually calling any locale generation wrapper script from the command line but the behavior of makerunall is that the wrapper scripts will be called with a current working directory being the correct locationChangelog Alter calls to locale and localedef to call +no unfortunately as of right now there will not be a backport +Does not apply to trunk due to records property to avoid calling target twice +We prefer this way +Mina version +sorry for the spam. +This seems to be an unfortunate side effect of Spring s aggressive caching of constructor arguments +This should be a trivial This is fine I have notified Citrix engineers to hold off commits until IP clearance issues are resolved +On both this problem happens +is covered by the nd ignore +patch looks good +commited +Hi GlenCould you confirm that this is working with CXFColm +It used to hang around the PUT and it is a positive sign if this one succeedsLet me know your findings +Will be back if still an issue +Thanks for your report! This will be fixed in next patch +If its OK I would like to let this lie fallow a bit longer until some of that code is in then try to get this patch up to date at that point when some stuff it touches have stabilized +There are a ton of operations that return paths +Not respecting final here is a you are right +Randomly makings disk via mounting +It is unfortunate that this change introduce a binary incompatibility between between the two different versions. +pending tests +fixed at with the release of. +Can you try to clean your plugin cache directoryYou have several releases registred for each plugin and it is anormal +If the uberbuilder was replaced by something that took the output of the aggregate build and wrapped an installer around it could we stop supportingupdating the uberbuilder +The code I used to test +There are no maintainers left +FixedThank you for the patch. +A filter is a better fit for wrapping the request. +Fix positions on Query Scorer would help you know I think this is ready I will commit that soon +If there is more than one provider for a service the application should continue and maybe display a warning about not being able to load a certain service provider +I am going to have a look at this myself +I can manually enter the goals into the runner dialog anyway +Anyway thanks for your help on this ill let you know if I succeed to isolate a reproducible +Thanks this is a good find. +Jared assigning to you for EAP +Hi Yves and Magne I do confirm that Sonar doesnt support parallel analysis on the same project see and +No tests because Im only posting for community comment at this point +I moved the former into its own file since it was used by two other classes and the latter into an inner class in and that seemed to fix it +Works fine now +Lets reopen to reconsider the test case so that this jira can focus on the bug uncovered rhevm driver shows createimage action when its running while it should only do that when its stopped +HiCan you use the following cli to generate the patch svn diff the root of the continuum +If true its another manifestation of the issue in the DN filtering methods +Hmm having the channel and pool thread groups in TP is possibly as the transport can be shared by multiple maybe moving the channel thread group into and implementing ref counting for the pool thread groups would be the solution +Pull request updated +The issue is that the java parameters are stored in an environment variable PIGOPTS +The first is when I line up my nodes etc into a logical order and save it +Moving out of till we prove we have this condition for sure +Sorry my oversight use did not work as I expected +Pushing issues to the next CR +If we do apply the patch we should also update relevant parts of the API and adjust the documentation accordingly +The provided test case passes +Refactored from to +I will start a discussion in our mailing list +So fix them together +I checked this and worked in pseudo distributed mode +I can only reproduce this if the restart does not run cleanly +Ted YuOK I attached the patch againand also provide the test suite results +Just had a look and it doesnt seem like the reducer that fails gets any more data to it compared to the others +Also cdouglas niq issue depends on their being a repository to point to let me know when there is one thanks +Ticket creator confirms that metadata works for him +Resolving as fixed +Thanks Ravi. +Id be happy with just iowait to be honest which can be easily read from proc +This should fix it I did a similar version of this patch for our branch and it fixed the users issue +Closing. +Proposed fix +The long encoding needs to be variable long encoded to reduce the size of the resulting serialized bytesIf it looks ok I will implement Externalizable in other classes +sure when will you review it +And itsOverlay feature is not useful without targetpath +So issue can be +The webapp correctly validates against a max of +Ive updated that chapter added the default class information to the namespace appendix and added another mention of the parameter names to the Javadocfor the strategy class which implements that particular logic +Should I check it in +I have updated Grant Revoke specification +new code +Applied the addtional patch to branch closing it again +thanks for catching it +If you ahead of time create this directory with the wrong permissions that is your mistake to correct not the packages to solve with brute force. +It seems the issue has something to do with the data dir +seems to be fixed +merged with patch +When the analysis fails I am stuck +There has not been any other interest on the issue and since there is another way of doing it I am going to resolve this +Reopening to perform some modifications to the ID generation process to make it more consistent and ease future changes to the configuration model. +I am still having the same problem +This actually would have been better in the EJBCLIENT JIRA project but I think having it here will help it track for +Therefore this issue can be closed after applying the attached patch +It is too big to attach to the JIRA +Should be system agnostic now +All of the failed unit tests have been recurrent failures over the last many builds and do not seem related to this patch +The configuration changed between and. +WhyI suppose you are going to subsume tag within the row as that makes serialization and deserialization of tag easy +A call will be very useful in this context as it would return an immediately which the client can use to track query progress +The earlier errors with was because I tried to build with so false alarm +Heres the patch +Patch to change the rampart implementation to use the same encrypted key in both request and response and also to support encrypted header +The vote passed the artifacts have been pushed to the repo +If you finish it before we release beta please correct the version +Thanks Professor Roger! +Please provide update on your items as a comment +Committed to JBESBGACP with revison +Assign to J David +ant is the legacy build file +Committed the changes into SVN trunk with r and into the branch with r. +So the writes to this stream should not be using default encoding +Ill open another issue for that +Netstat output is attached with Reset exception stack trace is attached with this +Patch includes a set method for and and a test case +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +hrr hmmm +ProbablyAt any rate syslog is not available on Windows in any shape or form so Im removing from FFIs platform and adding a regression for the proper way to generate the platform files on Windows see the discussion in. +However you should now be able to see the attachment I have just uploaded +Wonder if this did work in the past +patch is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +Test be fixed now please test +problem is with FF Mozilla Firefox Yahoo! EditionI do see it on my local machine with FF It is also visible to me in IE Likely something with the Yahoo! Edition +Rob please good +And it depends on +is making use of Undertow which is built with support for pairing SPNEGO with other authentication mechanisms. +This is not a clean battle +Checking and testing the patch +There are two separate implementations and +Looks like a useful addition! +If you post a full patch set I can suggest +You should also javadoc that variable +Perhaps we can incorporate that in this issue as well +I did just rebuild jbpm trunk with the most recent console and the problems seems to be gone +Tried it succesfully on a windows platformIt does not work on a Ubuntu hadoop configuration with two mappers running in parallel? +What I am asking is what is the correct way to discover what sequences exist in DB +To also make this work we would need some machanism to transfer the change in terminal size to the ssh server +Another problem that effects reproducability of the bug is that the current SQL query by default has no order clause +Attached initial version of patch Which is basically merge of and added the code to try from other when getting the length for last block +done via in a featureimprovement request for itsome flexibililty have been lost for better testability but its the goal to bring good flexibility back beforethe the beta release. +So if you use it please verify if the patch works for youbq +Thanks Lohit Mac Devraj for providing the comments +Try applying Latest patch +assigning to Rob since Max apparently forgot to assing it back +Please be back today +It doesnt really give equal footing its just alphabetical discrimination +Furthermore I didnt even realize that I was supposed to close the zero copy cursor since it just came in from closableMy complaint stands +Were planning to get a vote out for in the next day and that will be a great time to loop back and reviewhelp out +Which perspective and view are you right clickingthe file new is dependent on those two parameters +Although the pattern will prevent files with name build I think we can live with that +This is fixed in SVN +Iainrecently we improved the inventory receive services +Ted Ive gone big on this and collapsed a lot more iterator related things into this package and tried to unify a little +Fix committed by to mark correct fixin version for this Jira entry +Thx. +Note weld version will be picked up on the rebase for ER +Looks good to me thx +Lets backport this to +These changes look good to me +adding specification patch to address windows compilation issue +A bare cordova create should show at least an error explaining the correct syntax +Dependencies are notTry your test again with commas for what I imagine will be a better result +Hi Andrew +So when data is on disk were pretty close to what we can read from +Moreover it migrate immediate component management code from the instance manager to the lifecycle callback handlerThe patch deletes classes namedThe patch creates classes named It modifies the file to export and packages +Its really just exposed so this thing can be tested in some way +Bill can you try it +Can you be a little more specific? What exactly is missing where +That would then need to be +Which version did you test? Please set it as fix version +I just committed this +This was completed in December! Thanks for the catch. +I mean dynamic queries with parameters +Implemented in rHi MarcHow about making it possible to modify the list in in code? This would allow us to implement an for the tests something that is required for some of the interop tests to pass as they reference resources that have changed onlineColm +When the number is larger than maxQsize the earliest are deleted +Strictly speaking does not need escaping because it should always be clear from the context +How does that sound to youIt sounds good +Please ignore my previous comment it was intended for a different jira +Bulk move to chance of one xmlrunner coming out for Galileo bits +Attached patch nulls out the Fieldable reference +Thanks for doing all of this Adrien +the first one? Alternatively you may send such private pdfs directly to the devs +on backporting +Thank you for thisI was not aware of the show featureI am already caching results and in some cases i am storing the view results as documents to speed up responseI am currently using the jcouchdb driver to connect to couchdb and so will have to check back and let you if this works +See for further details. +I have found this bug also +log and configuration details file for reference +JBDS is based on the WTP maintenance stream specifically WTP still present on JBT +Scope most probably would not go beyond them +Output images from old and new scenarios +We should not include functionality in the stable launcher that is than missing in a releaseMetaxa is quite useful and it would be a good to have it in the stable launcher +Im not sure about Javacc but Ive seen reorder cases in switch statements even when there are no +What about plan names like Hi Im a plan name +Support is now in place for +So you could add a web fragment module via and whatever is in the fragment gets picked up +There are interesting levels +Thanks DagCommitted patch to documentation trunk at revision +Instead of throwing an IOE a should be used +All cache documentation has been moved to the new location there are only a few API files still to be merged will be done a meeting with Joshua Wulf it was decided that moving everything over was unncessary and needless work as after migration all other languages would need to be migrated as well and at present there is no time to spend on something with such low priority +If the constructor throws an exception then children wont be initialized +This was fixed March th in trunk +What do you think Lars err on the side of reflection here +This is the exact opposite of what we want to achieve with replicationThe functional spec of states that resuming replication after it has been stopped is a good candidate for extending the functionality +Revised patch using parameter map and unit applied on and trunk +Here I have extended the scope of interface where it acts in a more generic way +It seems that before my patches the would fail to enlarge an image at all when was set to true and only one dimension was supplied +Fixed by +These are patches and script for integration of awtswingtestsStep for intergation tests +Thanks Eric +Filed as +Ah OK we established that this is not a VFS issueNor is deployers since deployers only understand what VFS tells them to doPotentially there could be remote impl of Deploymentbut my guess is like I already mentioned that this should be handled by PS +What should fuse return when the group is supergroup then? the current code will end of returning nobody +Static bloom filter implementation for +No special protocol change is not needed for session recovery If connection can be +c file rampartsrcutilCheers apply this patchCheers include +Henry I visited the URL thats itBTW Gav sorry about the usernames skipped right over the part that said they were moin usernames will get it right next timeThanks Gav! +Can you submit a testcase +Do you want me to give a try to the version +Unit messages for the unit Not a bug +I will create a new JIRA and allow null keys and values except when testing keys out of the mapper with the rightThe bug in the nd comment is that we try to create copies of the output to add to the list but dont try to handle nulls correctly +Moved to is closed +Although its not in the patch I am sure you have play with it +Is there a good way to lower the danger of accidentally doing this thats currently supported by Hibernate? I could consider this in the patch +I first triggered a successful run by modifying the port number in the pom and committing +So I would adapt the Installing CXF section and make it relevant to +Agreed the failure is unrelated +Ah nice catch +Revolving on on +This issue not only creates a lot of overhead but causes interoperability problems with +Because I am trying to match revision nnnn and rnnnn whilst not matching rnnnnn and rnnnnn that often appear in svn urls pasted into bugzilla coupled with the fact that my regex skills are pretty basicI think I have a solution now +This does not add overhead the trycatch blocks are already there +Mass closing all issues that have been in resolved state for months or more without any feedback or update +Our software is quite complex piece of software for a JEE app tens of tables hundred session beans hundreds of thousands of lines of code etc and ironically the complex parts work just fine +Looks just committed this. +Fixed in rev and thanks for reporting +for the reviews everyoneCommitted to and trunk. +At the init method it calls init for all transport listners +With version they are both integrated in the trunk of pdfbox including an update of the ant and maven related files. +It should be able to disable tests in the configuration +Im talking about the shell command line executor thats shipped with mesos +html components like panelGroup will not work in pdf documents +Testcases should explain what exactly it does +maybe that should be proposed for the Lucene query parser to side step that particular rationale for wanting strings to be unanalyzed provide a map of analyzersAt the Solr schema level we could simply have string be a with only and then users can copy andor customize as they wishThis begs the question of how or whether the Solr schema side of the house will rename the string field type or keep it as string and simply change the class name. +Thanks for all the contributions Asanka. +Not all of them work +This bug has been marked as a duplicate of +I wanted to write a tokenizer to support anyways +Removed all hard coded mapReduce task number limit from code +Issue may already be resolved may be invalid or may never be fixed +Some please doublecheck that my expected is right in the unit test I dont know the scanner semantics that well +A test case reproducing the problem would help in getting someone to look into this +Okay so I noticed that the current code actually simply dynamically creates destinations if the endpoint doesnt have a resolver set +Request a review for this doc update +Ill look more into the code Brett posted above +If the name was ebfore scriptdef only then removingthat part will lead to being used as name +Hi Jacob thanks for your annotation +Hadoop is not allowing wildcard for so I wasnt able to test itI couldnt assign the JIRA to myself can you assign this JIRA to me +With it true there is a chance of deadlock given specific circumstances. + +Attaching an initial patch +patch has been changedpatch for patch for and trunk +I can still make itfail with the IBM Linux VM but it does seem to work with the Sun Linux VM. +patch removing all nocommits with additional testsI think its ready to commit +I actually borrowed the code from our initial contribution since it met some project requirements that I think we no longer need +Fixed Test Case is attached +JIRA isnt for tech support try the User mailing list +My implementation is slightly different from yours to ensure that circular dependencies are logged again in each resolveOTOH I didnt took care of duplicate mrids in the same circular dependency Ive thus applied your modification to the with just one modification I break the loop if the same mrid is found twice just in case. +Attached a patch for the trunk +I have tried to upgrade to but fails so it needs further investigation +Since this is a static link the symbols in are not exported by and hence the symbol uuidgenerate +Otherwise the automated build wont work +In case the broken behaviour is still needed it can be emulated by providing as the Lucene match version +schema +This is done in a backwards compatible way old files wont have this and so will all be visible to everyone on JHS but new files will only be visible to authorized usersWebapp only patch no tests +Prepare delete wont work with table storage +The remaining problems can be flushed out after checkin when the build breaks on someone elses machineI think that risk is acceptably small +the document is good +Ill probably apply it in our distro just for the sake of the unit test since it does add yet an extra layer of verification on the tricky code +But as long as URI classcorrectly implements the spec it should be ok +Some of this work is already done in +If you finish it before we release beta please correct the versionThanks +Re row groups Ill admit that I dont immediately see how this would help to address the issue but assuming that it does we could potentially use the class to support it +Invalid is still present in +In the shell if you do enable table and immediately follow that with a disable table some regions may still be because they were in transition during the disable operation +deleted and substituted it with into project deleted all calls to JNDI +had a bug in handling of times in writes were expecting milliseconds when the API really calls for seconds since a tTime is a timet which is typically seconds +Are you sure its a good behaviour for Alpha +when ant is called from those subdirectories so I think its reasonable that classpath exists in both +Committed. +Another quicker solution would be to create a sandbox page where users can play around with dnd and other features without having an impact on the rest of the Portal +Is this issue resolved then +Scheduling for ER as the upstream issue seems resolved +Added a check for HADOOPCONFDIR too +Since the patch is already applied I vote for leaving it as it isThanks for fixing the regression tests +That doesnt look good for the rest of the world +. +I offered this solution to the latest thread about this +With applied I cannot reproduce the varying optimize timesThe times are all in a tight range after nearly x as many tests as producedthe variable optimize times before +Is this issue considered a blocker for? Perhaps it can go into +RPC remains on time out it seems that this issue has to waited for +Created and related OR and registered it as a new activity Compensate class and related OR compensation BPEL XML stylesheet +See last comment +To be clear the same stack trace as Sergey. +However this is not that issue +Reassigning to myself for required manual merge +Its important to also provide the option to send the full text of the post in the email +It is for +Ok I am not objecting to that +Look forward to your findings +So one way to get around the bug is to store the transformation result as a DOM and then we write our own HTML serializer to serialise it +Great thanks for the confirmation! +Thanks to Mike Joyce for working on this with me. +The last change that we needed was to perform an ANALYZE before every time the query was run +Then you compile mapreduce test and push it to hdfs +Thanks again Bilgin! +It did not capture the move of the module +We have been on hadoop for a while now. +Ive fixed compilation errors but someone have to implement new methods and verify an affected could you please check the Dali integrationDenis did you commit? Cant see them on the fisheyesvn tabs but maybe that is down at the moment +would continue to not build on Windows but would be the alternative +I agree for TCP transports +Not approved for EAP +Some implementation in now can be moved to but since we will remove and replace it a Snapshot feature currently I just leave these code in +Its been fixed on trunk now +Luke when committing please add both the contributors name +Resolving as fixed as discussed above. +This probably isnt a bad idea +Alex the problem is that collection is so we dont have many options to find the current selected object and in my opinion the combobox shouldnt reset the current selected item just because there was a refresh in its dataProvider +Just marking for inclusion in the release +This issue was observed in of runs executed on on all platforms but not with other database +New patch version following Andrew Robinsons request to impact instead of sobryan linesChanged paths M myfacestrinidadtrunksrcmainresourcescomponentstrinidad M myfacestrinidadtrunksrcmainresourcescomponentstrinidadcore M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidaddemotable M myfacestrinidadtrunksrcmainwebappcomponents M myfacestrinidadtrunksrcmainwebappcomponents M myfacestrinidadtrunksrcmainjavaorgapachemyfacestrinidadinternalrenderkitcoredesktop trtable scrolling of content Thanks for the patch Cedric +How does that work with beans of the same type? In our case we dont have qualifier to discriminate them but use autowiring by name +Unassigning myself from this issue for now +reopened to verify +Thanks Ruth the patch was applied to SECURITY module at rPlease check that the patch was applied as you expected +Duplicate of which has been fixed +to Pranay Panday and others for providing implementation +adds a parallel compaction feature and makes some improvements to the algorithm +A new snapshot will be available in the afternoon +is OK but at this point +Regarding AVRO again I think we Cassandra developers need to decide in general if we are sticking with Avro or not +Not sure I understand this entry +and are not supported by windows in a filename +Any update on this one +for Jessica to triage the child issues and see how many are actually fixed +Found the issue +I just cant make everyone happy can I? Ill try to revert the changes to rewriteURI method without completely breaking the fix for +Closing old resolved issues +Todd new comment looks good +Sorry! Can you please repair it? Thanks +Most of the discussion has been about the v patch but the last comment is to commit v as wellAre both patches required? v seems to be an updated version of v not an additional patch +At the moment we dont change the default behavour maybe we come back to that later. +add compile line to the rest until. +I am fine with implementing Builder pattern later +Sending trunkassemblySending trunkassemblysrcmainSending trunkDeleting trunktransactionTransmitting file dataCommitted revision +Hi Evgeny Well unfortunately we have like pom files and our codebase is kind of intertwined so Id have to send you all pom files ugh +Javadoc for public methods + nice cleanup +checking if it is fresh heartbeatbq +Thanks for the review Todd. +Ill keep checking and post a comment once its available +Testcase fails without the patch and passes with the patch +The ZK failures seem to be caused by the CI host independently from this patch +Verified that the type is now enumerated and only accepts the following values and +it would map to BoilerplaterTypeInteger +Unless this is a regression lets fix for for trunk not a problem in the branch +yeah I think now it is much will create a pull request +Imported code based on instructions in dev list. +I feel bad about no automated testing but convincing the etap suite to run multiple local instances doesnt look simplepatch applied in r. +This is a result of see my comment from nd Dec +Do you want to update patch as well +Maybe we can tackle it some other time but at this point Im feeling pretty comfortable about the current plan and dont want to have to go through a bunch more testingJukka do you do IRC? Im usually idling in asfinfra as agreed to take over this issue and coordinate with Jukka +Also if possible smaller examples are usually much easier to debug +This no longer causes a NPE +Allen regarding using bundled in Linux +optbinjruby grep s +Yesterday it took me hours to run the tests on trunk. +I will try it on a linux machine when I have the time +Need to figure out the column order for the extention though +bulk close of all resolved issues. +Zookeeper doesnt say you have the wrong session id password it says the session is expired +Follower gets some additional tranactions forwarded from leader applies these directly to data tree +Please provide a example is way past nowIll try to find the time to do one. +This work was committed commit ecddaafabcdfedad +However as part of commit aedcdeefdcfebe Edison reorganized code from to and some code from commit ccefdfeedceaddfde was missed out resulting in resize volume not getting kicked off for vmwareIll log a separate bug to track this and provide a patch containing the missing code +I also reverted this patch completely and reran the test case to be sure that this patch was not trigger any strange hidden behaviorAmir the test case has been tested on power platform on ppc Jenkins server +I committed this some days ago +I think the resulting artifacts that the build process produces should never be released so to speak only the mechanism for a user to download these +Committed +In future refactoring code or doing things like rewriting NM in C will be least intrusive with the current module structureIf no of jars is the problem can we just merge the jars at build time the way we want +Committing soon +String parameters is the path of greatest abstraction eliminating a proven vector for bugs and it keeps our query interface as friendly as possible +serving logs and open a separate MAPREDUCE ticket forFor the YARN changes Remove the comment above the code which talks about I think we should separate the exception message to clearly say whether this was an or something else +Revision for Revision for trunk +Still playing with a couple details so this is a placeholder but this is almost ready +is really just adding a workaround for one specific case +Updated the SQL statements based on the comments from Mamta and Yip +Removed the splitRange recursion and replaced by a simple loop +Initially supported export formats are CSV TSV and XML. +A Japanese translator might provide files in EUC or SJIS +Can you first test this with or master and if you still get it well have a probable blocker somewhere in the system +Im using Groovy +All that said Im reopening the issue theres no real harm in it sitting open. +Thanks again Leigh for the repro project Ive just taken a look and indeed it demonstrates the issue quite clearly. +issue fixed in iPOJO +I dont have strong feelings either way +How can we do that +Do you this I should close this isse for now +Its almost like the final URL is being URL Encoded twice? Ive been trying to compile trunk but some tests are failing when I try building with Ant +One git mirror coming up +Khaled Thanks a lotI have just now committed these changes to SVN +Then I go to the web console click on Queues active consumers and I see the connections for the consumers just fine +Ive attached as the latest and possibly final fileset for +Ill try to throw up some better tests today +To simulate custom scenarios a interface names Simulator can be implemented with just one method +Reopen to attach patch that was attached at the wrong issue. +Is there any way to log this as a change that may impact backwards compatibility? Or should we just not bother since it is? precisely has changed? +But no change same exceptions on both cluster nodes +Patch +Ryan it was working fine for me +Check existing in repository tenant of created workspace in step but see only system workspace +Reverting the commit and the issue +Hi AntonioI can take this one if nobody is working on it +Closing as outdated documentation ideas are now gathered for m. +Has been working on that for more than months and still no RC release yet +not going to support domain mode in timeframe as it is now. +Any idea +I failed to run the unit tests on the original the new patch fixes the exceptions thrown by the first applied the patch in revision as proposed with some reformatingThanks for the contribution +Just a tweakI included a couple of extra tests for the on the basis that more tests cant hurt +I just think that we should standardize this across all sinkssources +Now I suppose that this manual will be built automatically together with the others Latest Alpha Manuals and the only thing yet to be done is to add a link in page Apache Derby Documentation for this manual with the title Guia do Servidor e Administrao do Derbyhalley +closing +Committed to trunk +bq +fixed the state change to interested +Patch applied +I did look at the other test cases but I didnt see a reason to extend +And can we start running perceptualdiff on Hudson +Consequently I have today added support for method overloadingWalter Id really hope that you finish your good work by adding some docs +Works fine +It looks like it is possible Ill check windows it can be fixed only by switching to the new version of because registry operations required to run privileged to get registry access +I wrong to select Components area please correct it +is not more flexible since the mixin does not have to be an inner class but ANY pojo just has to be defined within the scope of the aspect +Hi Jorge are you using the Sonar exclusion mechanism or have you configured your file to specify the exclusion file to be used by Findbugs +However this process is not executed within a Spring To make it works we have previously stored the and are established in the new thread just after start processing the remote response +Max this was seen only in seems not to be fixed +Classes Tests and Documentation updated +The java file I use to run it is nothing more than the sample java code on the apache web page +BTW I dont even know why formform adds a querystring to the action url +Will close this issue when is released +On the other hand AcegiSpring security is what most Java developers are familiar with +Please find the patch attached for Randomized Bipartite Matching Algorithm with HamaHere is the command to run it +I dont were likely to ever include such a thing in the core engine +Regarding Storage Engine BoundaryObviously +This was resolved somewhere along the line with an undocumented config parameter +Thanks Sravya! +Or is it same as +I must thank again for your scope attribute in snapshot works fine in my application and junit test programs +To find where it is currently running you need to use recentTasks +So all issues not directly assigned to an AS release are being closed +Moved all the parameter renamings out to this issue +Id also like to say thanks for the patch Sim and welcome to Apache River +I think there is a bug here but the bug is not in the Pattern class and does not revolve around the use or of the constant +Tested with maven site and maven +Ive also attached sample certificates and keys +from output of git diff trunk +I tested that no is run if baseport is set and that all the tests are run as normal if no baseport is set +MatteoOne kind of late question for the comment unless you need all the history put by put you can take a snapshot every hour and the files not compacted are shared between the old one and the new one and maybe the tableFor disaster recovery we need to export snapshot to another cluster if following the snapshot way right? In this case can files shared between local snapshots still be shared between the image? Or do we support incrementally export snapshot? IMHO we need this incremental feature to reduce copy cost for both exportbackup and restore whats your opinion? +My apologies +Wrapped the SQL in a single line +Marking patch available +Attached sample project +Its a great first step towards getting more efficient ledger storage into trunk. +and to TRUNK. +To achieve this I would like to get committed +it seems I didnt mark the DEBUG level for I am going to check this again +It looks cool +Presumably those systems are familiar with the problems of managing multiple Derby databases in a single JVMo I do not understand your comment about plugability +Id say its no more relevant only if we have othr issues to track its afaik you already working on a workspaces page but the new layer page is still not there and I can take it right +Info marker does not work here +Maybe check with Aileen +I noticed in Java that uuuuuuuu is legal +Should work now +This copies less and never passes large unused buffers over JNI +But it is easier to do so than to write new code +Since and coprocessors remain after this merge my point was that we need to keep security profile for running the unit tests related to these coprocessors +Actually change return to return rc +The only thing missing in my opinion is the Maven script modification +Could you please also add the ignores to +reopen it for more element and divide this jira into parts and coverage resolved +Id like to but not sure how to turn this into a patch +Thanks Jason! +Image validation exposes the errorCurrently NN uses roll to trigger storage directory recovery +Thanks to Dyre for the workflow reminder today. +Im not familiar with the HR application but I suspect that the separation information is already kept somewhere else +Need to retest with CR as this build has the correct version a basic testautodiscoverystartstop AS +If were going to accept null as an option here we shouldnt use option at all and should just check nullity of the value +Closing all resolved issues due to a successful release +Please see +Whats the exception stacktrace that you get on commitIn general this looks like gross misbehavior on the side of +Does the storage engine only optimizes logical plan or part of the logical plan physical plan resolution or merely executes a physical plan? +If you see a grey dot on the availability lights then that means no data has been received for that metric collection data point +i have no idea why people waste time submitting jira issues without tests +File Menu +I kindly request you to guide me to some useful resourcesMy email Best regardsKevin Wu +So I closese the issue +Im good with the last version of the patch of just adding a dclocal read repair chance +Tested on tomcat only +Should we update the catalog database after this patch +doesnt need to be documented as a SOA issue +Attached patch to bundle along with sources for Maven applied +Btw there only workaround for is to restart the server +However Im just pointing out another part of the spec that needs clarification +I have updated the Server running +Issues resolved in +Agreed covers the same ground. +I cloned this issue from a similar issue but dont have the option to edit to update all of the other fields +I have updated the patch due to numerous changes on the trunk +I wouldnt want this to be a breaking change +Patch for +I think its worth committing this now even with the so we can address the in parallel +We cannot treat this as a duplicate issue +can you provide a patch for documentation of? +Try refresh the page we have a ajax login for editing which isnt updating the profile bit i guess that could be regarded as a bug +Proposing INVALID if no further info is given +If you really want to go all out you could use one of those fancy buttons that behaves as a popup menu so something like Change to Java Interface WSDL Interface ESB Interface +I am montioring the store availability at CI side by using heartbeattime in symnode table +Uploading the trunk patch again unchanged to Jenkins can run the precommit build on itThe changes look good to me +I had started looking into how ASMs was dealing with signaturesparameters passed to it +Thanks Nicholas +Move to +Withdrawing is not fixed +fixing the links now +Please also remove the empty redundant directorywwwcontent the following directories are notAFAIK the owners cannot fix this because the parent directory is not group incubatorChecking wwwcontent slemesle incubator Feb +yes ive got a copy working with some changes that ill be supplying a patch for soon +Ill try to commit a fix next week or so +I think the Enable style would works the best for such addition could you please kindly revise the comment in and mentioned that even using within Configuration still not working as expected +Thanks Patrick! +Hi Davidcan you look at this when you have time a disposal to review this new patchThanks in advanceMarco +It works +To handle ranges I thinking we can iterate over the owned token range boundaries for the local node binary search for the approximate index position with the and then check the index to see if the nextprevious key is outside of the owned ranges +Verified on +Thanks Jarek! Next I will focus on the comments in codes so that it can generate a better javadocRex +Related or duplicated issue seems solved by fix to +Small inprovements +That means that this patch is the last one modifying the table in the Derby properties topic +the feature is tamed +I updated the Jira with a test to verify the fix +fixes build issue on trunkTested on and find that its not needed although that might be a qualified yet it might become necessary if the hive patch that introduced this change is backported toCommitting on trunk +New implementation of for the contribution! Ive added the code. +Good to know thats great +Awesome thanks! +Both can likely be corrected +I hope that this patch can fix such a bug +That versionll implement the delete pool per segment only for those segments created after the last flushcommit +At that time we may want to handle this situation more gracefully + +Im not sure about crawl performance with more than rulesIts easier for me to just manage a boolean value in the nutch confI know this is not the natural way of crawl with Nutch but it could be interested for somes nutchs userThe most important problem scoring from external links is affected by this patch +r Sagaras patchesr merged in some further changes to files from trunkr Sagaras OM serialization patchesIll do some testing of these latest patches and respond to Sagaras comments +If you dont plan to update the version constantly its still a lot better to keep it on SNAPSHOT for a svn versionbecause otherwise its possible to have a version in a local repo which is actually a snapshot of a totally different version +Cool +You dont have to have a custom file +that is right this exist lines in trunk as I expect this should be already in branch too +sure Ill make sure I move my current code into that stub if thats ok for you +Additionally every bugfix alters behavior But your word is last here +Instead schema includes will be generatedPlease let me know your opinions on thisThanks did not include a testcase for demonstrating the specification of schema locations externally +I suggested a solution that I think will work the generated are kept in an entity where they are part of the primary key and you write a loop to keep trying incremental ID values until a store is successful +It sounds like there are strong points both for and against adding and we need guidance from folks on whats reasonable to do +i assume for betais it a regression +Its probably better to create a separate jira for moving the requesthandler out of socket server +Will be fixed in Geb which we are working on right now +A similar problem appears in the plugin +With regard to can you point out what would change after i remove that +Also added a dummy checkstyle target +Fred close as +I would love to see this added as well +I will give it a shot today +Can you please add some comments as to what the regex is supposed to do? Comments for each of the capturing groups what are they planning to comparematch themselves against would be good enoughComments addedbq +A Map does not imply any order on the values +I thought that because there is resourcemanager now and I did not touch the applicationmaster or the mr app client and the mrv code is in trunk that I did not need it +ReopenedSeems like we actually tried on Hudson +Sorry for the typo +Send email notificatin to customer +release bulk close +On a related note I think we need a way for updating existing associations in case the setting has changed. +Please write requirements and provide screen shots of this dialog +So any further updates would be masked by these deletes +Justin what do you think should the geometry actually inherit the srsName from QueryImho it should not +Id rather see this as a part of a package +re outdated requests i now see that in you are eliminating already queued files from consideration so requested files will never have disappeared between when a compaction is requested vs executedLet me take another stab at explaining the problem +Please make sure you update to the latest code from CVS before producing a patch + + patch is on +Agreed +Thanks +Cheers for the excellent plugin +bq +Was released with closing. +Plugin authors will have to update eventually how much work are we prepared to do to buy them two months of wiggle room for a change to their files +I would be grateful If I can have more detailssuggestions on the approach and further material to referThanksMalith +Thanks very much for your help +Russian localization is added +However in the test we do not track what is flushed and what is not +Marcel wrote As already mentioned in I think you should be able to simply replace the with in an existing Jackrabbit installation +I was referring to create sane a +Heres a revised patch for class that I created after reviewing JSR specification and its implementation for +committed in in efceaeebbbee +Like what the displaytag used to do with frameworks like webwork and do not need to use the dataform +Without it the last repeated attribute wins +Changed status to major because it prevents running wsdljava againstwsdl URL from a deployed webservice if using a bindings file +Thanks KevinPatch applied at r please verify it was applied as you expected +Configure css to maybe use our portal css as per the +sorry I did not have much timeI was also thinking if it is just the matter of setting this env variable rather we require Apache Forrest to be installed on building machine and this is another request to Jenkins administrators +This was instead of closed +Unfortunately we are unable to remove this action at present but we do need to address this at a later time +Patch was applied at +The unit test failed is unrelated but a umask error on directory creation and I already file a jira to fix it +I see it now +Dupe of +No core dump and SNI is working even with chain cert +Changed payment field as non editable +Part of that could be to peek into the second level cache as a first option before physically querying the database +Looking forward to Spring final +They are not aimed at us but at other components and we cannot force them to look or to use our tools +Unassigning due to inactivity +I dont think giving this responsibility to is a good idea +Submitted a simple patch +Option b also implicitly solves the problem of what to do about fsimages in the standby as well as just seeming overall safer +is it not? +Please check the build succeeds for you now +Ive got conflicts updating sources over my patch but couldnt find changes +The timing between the Rack error and the are a ways apart so maybe they arent related +Would it be a good idea to generate distrib by a simple filter of to avoid code duplication +What advatange does pausing the transaction processing give? +All tests passed succsesfully +Can you please elaborate how this would work +Heres an updated patch which takes advantage of by no longer needing to check for concurrent in the event of concurrent method invocation failures +Activate M facet +Done +Hi NicolasI havent reviewed your patch but it sounds like a good idea this feature already exists for the menu widget + was added +unfortunately trying to create a simple scenario works as expected +Its ok now perhaps The targetdirectory is used to house all output of the be corrected withThe target directory is used to house all output of the build +We have certainly made some progress on some process related things but the site and documentation are still terrible +Dev mailing listother existing tests can help to tell other ways to write tests +Works fine in Maven +Closing issues that I resolved in already released versions +was resolved as wont fix +Tested and passed basically moved the DC comparison logic up in +Lukasz thanks for the patch +Hi this is the current state of my code +Let me think about it more +Thomas would it be possible to attach the projects to this issue that demonstration your problem +That work is already done in the +Youre looking at the wrong files + I just updated the web site +Recently I found that has great support for Maven +There is no extra round trip for PLAIN and Digest while you can have up to n round trips with your proposal if client doesnt support the server mechanisms +This issue has been addressed in M as part of the Blueprint integration which requires services to be unregistered during shutdown +This will keep a HOD user from creating leavings all over the HDFS but allow for alternate uses likeWe have use cases we need to think about separately A cluster with one permanent JT +It is starting to annoy me as its getting late so I am going to call it a day at this stage amybe pick it up tomorrow or some other time +Very impressed with the turnaround time with a response from the team. +No problem Marek +So I think modification is the one to support +Yes +Sorry for the delay +Dhruba can we instead use the generationStamp that the NN maintains persistently on disk? Isnt the txId different it is increment per transaction +simple process definition with mail node and template needed by the to deploy the process process +Adds driver reuses Pair not integrated into yet +The quickstart works OK on my Google Chrome on Linux +Sorry for the late response was having some issue with my env + disallows null in also fixes some typos in +FYI ant test passes for me with the patch +We will reference the existing work in odfdommore information please reference We will reference the existing work in odfdommore information please reference Bug Chart convenientAny guys who has suggestion or requirement for Chart API please write your comment here +So to me this is more than weird I suspect inconsitency in the service +You should try using an or the latest release which incorporates many of the latest fixes to see if the problem is still present +Candidate file for inclusion in Spring I found an error in the file I submitted The problem is that no extension can be in the file more than once +Patch looks good +Sorry ORM is no longer Brett for your answer +Fix attached +The problem is that issue to too complicated to be reproduced while in debug mode +The failure occurs in new code added in so linking to that issue +Patch applied +and +A magnifying glass? +Hi SaschaThanks for the thanks Ankit + +Changed warning message fromResources are low on NN +Still need to test this on a couple vms but this should be right +libmozjsd I believe this to be a release +This version fixes some nix specific issues +Your help is much needed. +Given we cannot patch them it is better to not ship them than ship broken scripts We already have cluster startups and orchestration +But what dose this have to do with posgis? I got the the picture shown above via Layer Experimental data source on trunk does not have this for the screen snap a picture is worth a thousand words +Thank you for putting work into +applications dont need the finalizer +My opinion is that we should just drop the sync and remove the deprecated methods in +You can use the to add source maven hack will not be supported by the plugin +I had thought about using svn externals but since the code will be moving to git that wouldnt work +Will discuss this on the list +Yeah I think per workspace servies are the final step to actually completing virtual services +So the privilege will need to be passed to the authorization calls as a string. +Simple fix just check if current segments file is null before trying to open it +Will close this one +looks good on my end now +Retired process is still running and not terminates +Carl let me check +Ive now updated trunk to use Jetty by default +We could make a single instance of that all of the Targets shared in a synchronized fashion that would probably be safer than what we have now since even the scenario you outline in your example would be subject to a race condition in the current code +Thanks ClausCan I do that in XML to route messagesNone the less my needs are to be able to open a tcp socket and have two way async message flow over that same socket +Ansi +Are you running into +Steps for reproduce needed exactly +application you can commit it to the sandbox and close the application for being a committer for this project is pending too +All the old now become redirects to the new URL locations +As pointed out by Doug the user details need to be uncommented in serverdefaultconfprops once this was done the adminadmin username password details worked fine so could not reproduce the issue reported by Ivo +The reason is that we always put the operators to the reducer of the former job instead of the mapper of the latter job +for the patch +The NPE occured due to a with an attribute that had no path set in the XML +It is quite strange that the static type used by the the caller of the bean makes any difference at all +I tried a few times locally and its passing fine +Thanks Zhijie! +See the attachment +Added to wireTap as well +CXF currently can not do continuations overt HTTPSRon is supporting continuations over HTTPS a priority +We already have this functionality built in to the enum so its just a matter of applying it to the places where it is not yet used +It requires and to be present in working tree. +This patch looks good thanks Ning! I made a few small changesincluding how RAM usage of buffered delete term is calculatedAttached takeI plan to commit in a day or two +we dont offer any consistency guarantees after partition healing +I implemented this in by creation of the Validator Factory inside +Bug report is invalid +Minor rewording and example code typo fixes for trunk +Feel free to remove the +The exclusions are there for a reason +My vote is for b create a page defining tested and supported and changing the link to point to that +Many thanks for this enhancement +hack sounds a little hacky +I thought this was better as it would keep all the changes to the physical connection at one place +demonstrate this +ll try it will use this by default right +Hit this issue +for the to trunk. +Since there is no good way to specify the execution order I the test code a little +I dunno +HiI dont want to add in the users can have the rigth one and it will the multi modules you can certainly try using N when using the continuum plugin +I had another issue with my JAI that seems to be related to this +Hmm +This is what Im seeing on the latest marvin when doing associateaccountdomainidzoneidapiKeycommandassociatesignatureDresponsejsonThe credential verification is happening correctly and the API call succeeds +will commit if the tests +KB Time ms +Tried the suggested recreate and the move up to Axis has fixed the problem +batch transition resolvedfixed bugs to closedfixed +Would you please try this patch? Since this patch will also modify native code you may need a rebuildBest patch applied at revision r with minor changes thanks a lot for this enhancement please verify that the problem is fully resolved as you expected +Working on RHEL it works for most platforms without maven settings for xulrunne then it doesnt worth to keep xulrunner configuration through system properties as default +sorrycontinuingsomeone may come along at a later date and replace the int with a long thinking its a trivial change that couldnt possibly break anything +bq +And it would be better if you can provide test project +Thanks Kim +Of course we could add these packages but then we would soon end with the whole core in there +Yes I omit that thanks for suggestion. +I have made the change in my sandbox and I am currently testingIf successful I will commit this change in about an hourThanks Revision +This was not fixed +Help in resolution is appreciated. +If you want these typos to be fixed please post a patchThanks +By default we could set map to true so that backward incompatibility wouldnt be introduced +Im really swamped at work right now so Ill get to it as soon as I can. +Dear Edgar and Others interested in and LDAPI have spent the last couple of days thinking about the needs of our team here at Dataline +yep sorry about the typo I meant false +mean that a task should have a pair as acceptor as well? Ive split this into two new and +bq +refresh becomes a big issue when trying to debug with large number of tasksattempts +Postponed +I stepped through the debug and added all the +sets up a reader under the covers and it is that reader that has an encoding +If you need to count use integers and combiners +Patch that makes the shell do binary as hex +No separate buffer on the this case wouldnt be clearer to have only one parameter to tune this single buffer? Otherwise we may get bogus configs with the buffer size set in two different places to two different valuesIf the only difference between the two is the I would prefer an explicit parameter for it +in +This has been fixed as a part of another commit. +It is looking extremely unlikely that Greenhopper will start without massive downtime to diagnose and fix the issue +Formatting broke your patch +Todd we can do that +Attach a patch and Ill apply it +From ESB project perspective its not a issue assign this to Rob for tracing if there is server adaptor side issue at least it works for deploying a compressed archive +Looked at the patch +Elliotts patch only w the logo in srcresources dir so can be found when run region server metrics into a table +Property Sheet tab has color attributes +Attached PATCH along with testcase +It also makes use directly instead of the field from +These headers are easy to spoof and could become a security issue +I have a user with a trigger referencing a bad conglomerate number +We still need visitor interfacesaccept methods for menu models and form models +Some confusion between comandament and comandant here? +That was a hugh one +Thank you for reporting. +Keep up the good work. +Yes Prashant +I am running regression tests nowTouches the following filesM javaengineorgapachederbyiapierrorM javaengineorgapachederbyimpljdbcauthenticationM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceNew message and changed messages +If nothing is set the default is used +I think thats fine but why do you still have the lists? Having two copies of the data seems dangerous as youre likely to get out of sync + attached patch +It appears to have fixed some of the problems with the WSDL + +Anyone can commit to both lucene and solr so we can always put things in the correct place +Best regards George +Currently it is not producing a negative effect in the server so it should be ok +Steve Rowe sir Im going to need your assistance I think Ill take a look +Having not looked into the patch what about some sort of event listener idea? The Directory can register that its interested in listening to merge events and the event can be fired when a merge is started +The change in takes into account aggregate expression in the null expression and datetime expression +But Im not sure why this happens yet +Once it fails it is pretty consistentI put a sort into test and all seems well +Ill take a look it the issue and if it doesnt break anything as stated above try to include it with +Committed revision +It would be worth examining the syslog on that node to see if any disk or memory errors are reportedI assume the reduce was rescheduled and completed? If so then I will resolve this issueYes the task was rescheduled and finished successfully +Ill reapply this to the current code and will be back if any of these issues still apply +patch looks good and does what is expected the gpl codes jar is pulled into the build directory +I think it would be sufficient to only set the persistent header if the message was indeed marked as persistent otherwise the absence of the header should be enough to indicate that the message was not persistent and reduce the amount of headers needing to be transmitted on the wire +Ok I figured out a workaround +This implies that any lockable node must also be referenceable +We like to use ajax ability of richdataTable by moving from tdataTabe +Please verify and close. +I am beginning to see interest in using Extended Attributes on HDFS +Unfortunately CI job is in error +Keith thats exactly what I was hoping to hear +Presuming this is sorted for you +Note the previous test failure isnt related to this jira see Anupam! I committed this to trunk and the +Looks good +Robert is interested +This sidesteps the problem seen +Actually this doesnt resolve the format issue +running against a hadoop clusterHADOOPHOME is setHADOOPCONFDIR is not +Sure will do thanks for the reminder Jakob +But lets open another JIRA since that file is for other as well +Removed patch text from description body +If there is nothing logically wrong with my last comment then can this issue be reopened +I can post a patch unless someone else wants to take it +applied the workaround and move to workaround is enough to resolve this issue +The MEAD build is being run in a chroot environment with a minimal set of packages installed +Please optimize the import list +If you agree I can upload a patch for that jira when is committed +I just committed this. +There are enough projects who can release without a problem +Job finishes fine but Distp client still throws this exception. +Hi DawidIf tagsoup is not going to release a new version soon then IMHO the best approach is to follow your suggestion and patch tagsoup locally +Indeed seems like a reasonable addition +Could you please look at the latest as r +If not should probably go in a release or LATER +Thanks for reporting this +Heres the patch with the real version +I think we should hold off on resolving this issue until the test is incorporated into the nightlies +In the end the many different projects that get integrated into the AS are what open sockets the configuration is just metadata we make available to those projects via our integration layers +When the is touched the sar reloads the are stopped which call unregister on the listener and gateways +Simplified the patch +I will closereject this issue soon if can be bothered providing a test case +Thanks for the patch Andy +Eugen +hmm +I think for now we can start w removing the interface +bq +Done +Ive just committed this +The patch has a lot of r characters in it +Nothing in the EJB spec describe the expectation of scale and precision regarding the schema Hibernate uses them as a hint for some databases and some hibernate for +But this is a workaround +This affects patches that reproduce and help to debug problem +uses buildmagic for building +I am on the whole thing and was just trying to point out alternative approachesI describe all this stuff in detail so that everyone can read it even the people who dont have years of experience like Kihwal and yourself +which all need to be updated to display CPU cores as wellWhile obviously it would be easier to read this patch if it were split into pieces the new tests for CPU as a scheduable resource require the updated queue metrics and accounting creating an +Also the frameworkcommonwebcommonincludes file has already been fixed so you only need to include a patch for the eCommerce component +The problem is still present in latest trunk +is famous for its rather ehm special behavior in terms of class loaders so the issue could easily have its root cause is packaged at the same level as the jar file that contains the appplication classes +I expect that installing solr and then replacing index with solrindex in +Recreated problem as described and it works for me. +Next build of EAP meaning CR? Or next release +Looks goodI just made two edits to change the default to be to match the default at RM +It may appear with different conditions and is a feature commonly found in clustering implementationsThis issue can be marked as resolved if the patch has been applied and worksThanks to Nate for reporting and fixing the issue thanks to Gilles for reviewing and applying the patch. +In the meantime Ill commit the attached patch for this issue okSounds good to me +CommentsSuggestions +Will slip to WFK app I created was tasks based +Enhanced the to improve the performance of bulk logging when buffering is disable and not in lossy mode +Closed +done experimentally +worked for me on Mac +Although these are not they do give us a way to easily run all the WC test cases and generate the results required by the WC to report our progress +I was referring to the things that inherit fromYou are correct that Bayes isnt online or updatable as it stands +Thanks DagSomething along the lines of patch aBTW this code must be CDCJME compatible +Thanks +It contains updated script for the scenario run on Windows +It is not possible to add a comment until all the attributes have been estabilished +Dimitry sorry if I missed your intention +I can create the repos mentioned to but Im not sure what you want me to do about the reviewboard issue +The nd now only supports TomcatFixing tomcat would require decoding the usernamepassword Strings in the corresponding callback handlers +When this error happened? When you launch your app from Eclipse +Jay Im assigning to you as this is related to you investigate when you have any time +Im taking a quick look now but should have more time tomorrowThanks MichaelSeems we can drop it to me thenI dont know that the performance would be that bad you are only extracting the terms from an index with a single document so there is not likely to be that many terms +I will commit the patch so far but leave this issue open because of the missing fix +Sanjay also gave a on the parent jira +Oops this fell off my radar +Once the stuff is ready I think we can create a new feature branch for commit +I will go back to this JIRA after review the patch for with GEP so I close this JIRA. +Please file it under MNG for else to do in this issue +Yes lets do that +Closed upon release of Hadoop. +Should be fixed +In section and are not discussed +Added sortable tables to actions tokens and tracks +Basically this is just experimental to highlight the approach I have in mind +remove unused constant from test always add padding to the very end of the file not the current offset reset position of the padding to before each call to writeFully +which I think is not goodId like it to have it fine grainedJust to make it a little bit more complicated to find a conclusion I think we can drop the word STRICT SUCCESSIVESPACEENCODING would be sufficient. +this is the node file +For a standalone machine you can just set the values to nothing +That is the most logical +if is set with false the deployer will ignore all jaxws related processing +The weird thing is I have yet to get this error when running jboss on my laptop but get it every time on the dev server +Out of calls of them were true +My bad +readline char interpretation is a related problem +simple string parsing and concatenation + +Here it is unsuccesfully trying to access a jndi URL to parse +Per your suggestion we are trying to use instead of +Yes Im sure I did and it didnt work. +Transactional interface may need to indicate if abort is supported possible more styles of transaction than just everything supported and nothing supported +well have to re submit all the PAs so trigger hudson. +XSL file +Forget about xml format for the time being +If you think that is necessary put in a feature request and well consider it for a feature pack release +Can I suggest we leave it open for now which will provide a bit more time to think this one through +It works +Can you explain benefits for end user to have intentionally delivered software with memory leaks +Username lqd +Leifyou wrote The site SVN was always done which I took to mean The site SVN has been doneAnyway Gavin has now sorted SVN and updated the site workspace on minotaur. +Patch updated per Chris suggestion +Using the correct already solve this +It would be nice if we could come up with something cleaner but for now I think its OK. +Edward I committed just now +Therefore I have recieved over hours +So some Groovy change between and must have caused it +Does that help in your usage of custom exception translators +This is not a bug +Something changed I dont know what +works for me under both backends at +I have committed this change and trunk +I recommend consulting the PHP bug tracker or whatever is responsible for the server side code +The patch was diffed against the tag in the projects base directory +Committed to branchx and trunk +When its ready + +Only task is to release +I can submit a patch otherwise +Willem you can easily add a property on and have xstream builder method in accepting an encoding as well +The join methods on did seem to return and allow the main thread to become runnable slightly faster than on my Fedora system which played a part in many of the errors we initially saw and fixed +This patch is added to the topic branch +Committed revision +You can participate in the discussion on and Authorization frameworks happening there +The new junit test from the earlier patch finished successfully with Mikes patch +its irrelevant +If you would like to submit it as a pull request well need to cover cases where the group by is not needed and verify that it works on all platforms +The jetty descriptor is supported for a quite long time +It has been considered a security breach at all log levels +Listing directories is expensive storing filenames is cheap +Sleep was a good hack for of all cases +david can you make a snapshot of available? thx +Refactor of code took care of this instead. +My fault +Jake please take a look at this one commitclose as code is based on the old LDA impl we had which has since been removed entirely +Method descriptions updated committed +In the case that there are not that many accents it is much faster and in either case the replacement of is much faster +Changing assignee for review +Assigning it to you to do more investigation +This should work if you provide a relative path for a check for directories to exist. +Well Im happy to accept patches +Thanks for your help +a In all but one fields is declared transient +Since no one objected to my plan Im marking this issue for only +Closed see linked bug +Seems silly not to do this for a performance release +isnt something provided by sun? why do we need to patch those +Can ipsec realistically solve this in a small cluster? Is there a software product that can be installed on frontends and Solr servers to set it up transparently to Solr +By default Accumulo tables use a versioning iterator that keeps the latest version +It is if they choose to use itbut this is for when they want the additional guarantees that a transport such as JMS provides +I can do it today or tomorrow and post the resultHowever I can only post some jobtracker logs to show that job initialization fails before apply the patch and every thing works fine after apply the patch +There are two failures that are not related to this patch +Shutting down data node and restarting would change its internal state +This looks good checking it in. +defer all issues to +I would be willing to draft a change proposal to the SQL Standard for this area + role workflow adapter seems not particularly difficult to implement however building would take some time while benefits are hard to justify such effort without a real use case +The XSLT function is there merely for convenience +Server dise can be fixed very soon +This is in the sandbox for now under +r +Resolved in trunk in commit . +Patch with a fix +Committed at subversion revision +Tried local and shared storage with the same result +Great +Not found this issue in latest runs +A few fixes have gone into the regex stuff since +Thanks for being persistent! +Click InstallInstalling GWTGWT installed +From the short description this sounds like a dup of +looks for the unique key property and appears to do the right thing +Bachl for isolating and reporting this bug +lgtm +Thanks for your patience +Sorry about not providing enough info +If Im reading the code right TOOLPATH isnt defined in trunk +For the meantime Ill point to dev manually and then later a cvs email address can be inserted +I didnt see any odg files +Does this need to be in? It seems like something that can be developed unrelated to release +is fixed so before installing jrebel everything works fine +It seems that the connection to is not closed after used I will keep eyes on this +Deferred to defer of unresolved bugs to +Thanks to Jim for reporting the issue and providing the patch +Fix for this issue +Added this by adding special class that will close out zk when all is doneAlso removes from the running of shutdown hook to do clean up of filesystem +need fix tests +Attaching new patchUse the define for +Thats David Lloyd and Flavia Rainone +Lets see if the full set runs on Hudson and if so how much faster it runs there +Got all the patches sorted now +Changs in revision Removed the installable resources factory again and converted the installable resource to a class this makes it easier for clients of the osgi installer +Note that in two cases the param privType actually needed to be priv to match the name of the parameter +Verified by simple test that lead to the discovery +This is a public API change so changed the categoryNote Public API changes arent viable until we start a new JSR +Thanks for reviews Bobby +However I agree with you that should be modified to pass the default values +Like getv suggestionNow we can leverage the wireTap and add some sugar so you can pass in an expression for the bodyThen we can have a fire DSL like send that wont affect the original exchange +Its curious to note that the trochaic stress pattern of crtel is a since the word in fact has iambic stress in English as it has in French and German + +Updated patch due to changes in + +I didnt realize there was another issue open about removing those +I meant completely independent from the contents of the container +Once I am settled down I will give a try and let you know what I findIf you want to reproduce this you can just use any malformed xml in a jspx as a template and observe that the original exception will not be available in the log +Looks good to meAdamYou need to mark the issue as Patch Available so that it shows up in patch queue for commit +The fid in this case is completely generated and thus not persistent so their FID is already sort of wrong +Use words like Macys or Cs of diamonds +Marking it a blocker so this is not dont see the instructions yet +We must use +Hi Eduardo which JVM do you use please +Thanks Ill try this class to check if it fixes our the suggested workaround solve your problem? Then I would close this issue +Mhmm still catching up from my week in Atlante +I made this patch to add list support to sample code was not compiling as is so I modified it a bit added include +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +This patch the tree local into a interface and provides default implementation for the methods in +Since patch to fix this issuefixes files in filing a new JIRA and duping this one +To fix this issue we should process every template in list so to next version +I think we should update the localhost with the actual NM hostname +The original support for fragments was very conservative in how it determined conflicts among requirements it didnt allow any overlap at all +Removed deprected +A explanation is required pointing to docs t servlet be scope provided +A refactoring of the class runAutomated now throw a meaningful instead of returning false if something went wrong +Can you do it right now? So I can do rc tonight +Thanks Venkata Guddanti for the patch +Also Im going to doubt the stuff will be able to take any endorsing stuff +Approved for if done by due date +This patch needs the patch from the linked issue for Oracle to be applied first +update the patch +This bug depends on and +It was a problem of my bad svn commit via subclipse +Hmm +Thanks for let me know +Hi ErickI am not sure but I think the package structure has changes in that what is causing the problem +moving to BETA release +As I can see buttons are centered horizontally for now +Maybe this is not correct way to express that +My very first patch +Would prefer not to inline mainly due to symmetry with add and clear +The matches is the same for each commandI will add this to this will be backported to +Closing all tickets +If you get a chance to send a patch please open a new issue +Thanks Tom! +I think this issue is more of a subset of than a duplicate +It may be worth fixing if this is the only show stopper on mac but if there is others it might be something WONT FIX since hive expects a case sensitive file system. +Therefore closing. +Hi VamshiI dont understand exactly what the issue is could you please be more specificAlso take in account that we dont offer the Eclipse integration module anymore +Fixing that with patch and are referring to the new subclass of by its old name which is not correct and results into build errors +the applications name cant be set againIt seems we have to properly destroy the old application to make sure everything is reloadedWhich version of Wicket did this work for you +Was this a +Reassign to Doug who has worked on a fix for this +Otherwise a bad configuration will be Please test this with and without script prefix +It sounds like we should remove this option from the user interface if it is not making sense +I just wanted to round out the patch to include everything in the subject of this issue +Tomcat is not listed in search runtimes +Fix applied to release branch +Committed to trunk +the above static in video driver is actually reproducible in google chrome alsowhich likely means that it is a feature of pivot killing nsplugin via volatile memory accesssince I cant remember any other applets frameworks with this feature probably this can be fixed in pivot +will of course be updated accordingly so Im going to keep this bug open for now until we get final word +Closed after release. +What version is the from and to +I dont think a dependency on a Common workspace is the very plausible idea +I also see repainting issues when I navigate using the browser back and forward buttons +Committed revisionU javaengineorgapachederbyimplioU javaengineorgapachederbyimplioA javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTeststestslangU javatestingorgapachederbyTestingfunctionTeststestslangA javatestingorgapachederbyTestingfunctionTestsmasterU javatestingorgapachederbyTestingfunctionTestssuitesUpdated to revision +We need first to release the keyytol shared component I will try to do this soon +yup lets keep the generated projects clean +having apps install on the fly during boot adds one more failure mode when dealing with reasonably large clusters +Need to allow to accept external Lzo codec implementation +I had to back out a previous restriction fix due to the fact the fix was not comprehensive and also introduced a bug that prevented code from being generated in certain scenarios +Committed revision . +The problem with was in the test should be fixed now. +while the actual jar files in the distribution are etc +Jacek please verify if this Jira is fixed and close it if so +As I understand it now GRANT NOACCESS is meant to be a way to strip all permissions for the resource from a user +v patch is not useful for anything except splitting large sstables +Merged patch with latest trunk +I had to make some changes to the code as well to make it working +Just committed to +Merge done. +There was a question of the tests of still pass so there is some evidence of correctness +I was able to reselect the repository but it still gets into that state +I will make binary target depends on targets in +Will incorporate them in this issue itself +OK +Are you envisioning checkpointing as reactive to a preemption warning or something that the task just does periodically +A much simpler to visualize this is for every table there is a primary cluster and a list of secondary clusters +just map your simple type to xsstring +Gertv good catchPatch looks goodI know running tests takes a long time +I agree +I really wonder why the people havent created bundles for their hibernate validator? Are they not on the wagon +The problem as described is now fixed +Some panels are loaded but others never load eventually leading to an exception +Found some problems with the patch uploading a new ignore the previous ones +There was no obvious place to add a patch in the Add Patch workflow so I am uploading it here +Thanks VenkatCould you check and take a look at please? is service class and is java beans classThanks +bulk close for the release +I have linked these two together +Next time please copy all bundle urls in only one thank you for the quick time +I just committed this +There wont be a release +So when verifier calls it is possible that exception is already raised and well have assertion failedSo I generally like the ideas of but VM components are to be tuned for the new realities of classloader +However one of the points Jakob made in his comment was that the name is pretty generic and theres the chance that other may use the same key for other purposes +Just a couple of minor comments +I assumed you were using NIO which is not the case +FYI testcase for still attached +Change patch reference so that it can be applied on openejb parent module +Either way I will have already implemented this patch in my own deployments and I have already moved on to more important tasks +To address your comments I made two replicas corrupt and checked if the pending replica size is in the new testcase +IMO have la should be enough for it its just me +I think Chansler opened a jira for handling corrupt fsedits in SNN a while ago +Spring IDE is now in Central TP version and so resolving. +It also should not be necessary to move the features folder somewhere because it already is in the right placeAs for we are currently using the PHP version revision of the Shindig trunk +This patch may not apply without if we see any obscure issues +ChengIm aware of and the first lines of my patch are exactly for this +Oracle NG scans both the user specific table and the all table so should find also tables in other schemas +and copy the generated launch and then edit all the details in there and run it directly via hibernate tools instead of via the wizard +As per Rafal Krzewskis suggestion +So are we deciding to tag this for LATER and try it postTed +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Obviously it was within the though I dont believe it does anything special to try to read the data file off diskI do not and have not seen this exception with a single local directory configuration +with my thoughts it might be low cost to do a rippling lockAnother alternative is to use a data structure that prevents from being read by scannersThis is a dangerous area we can and should do better in imho +So it looks like some initialization is not done when it should +Do you know the cause for this +Marking closed. +I am confused what the problem is here but it seems like you dont want to remove stopwordsIf you dont want to remove stopwords just create a passing as the stopwords parameter +Ive committed a note to the getFactory method javadocs which shouldbe suitablePlease check and if not satisfactory the please reopen and attach a patch forthe documentationRobert +Actually I think it should interpret int as unix time in seconds and long as unix time in miliseconds +Hi Dragoljub Spasojevic I added rs serbia to the list of country codesPage is ok now +Ah OK So I still see a potential problem with thatService propertys have no unique constraint in so there could be a a hundred services registered with the property namehello of all different types +Fixed +Yup Adobe deprecated SVG as part of the promotion for FXG +Maybe there was a concurrent running of another junit test +Also useaxutilstrcmp instead of in the trunk +Updated patch including the required properties file +The plugin works when the persistence classes are annotated +Closing as wont fix +Bulk closing after release +Please verify the test passes for you now. +Lets file a jira for it and fix it +Can we just have whoever does The Great Reformat to take care of this as well +there are other files than just versioning and just because you have +I saw the fix in CVS but you also need to change bgclolor to bgcolor +If not set an exception should be result should be per database in the target directory a script with the upgrade ddl a schema diff issue for which this is needed probably will get an extra column this really belong in the user guide? I dont think we want to support upgrading between arbitrary versions +Hi HuguesI can help with testing and creating examples +There is still something horribly broken +apparently the fix was for instead of are committed to HEAD and branch +Thanks benoitc for advice on how to handle the Makefile +Please close if satisfied +I have asked a question on on how the changes in that might affect the semantics of the operation in this jira and +Part of a bulk update all resolutions changed to done please review history to original resolution type +I just changed the priority to blocker because if I cant send symbols in GET parameters I cant use Axis +DaveI was looking through this issue and Im a little confused on the difference between stop and end? +This ticket is stalledThe driver behind it was being able to support a feature that Aperture has +Added +Manually tested +thanks you couldve just provided a patch from the get go +Also try using something like wireshark to capture the actual network traffic and seeing if the response is bogus there as wellIs this still an issue? I have no idea what to do with the above to test it. +Not yet +Hi Aleksandar do you have some examples of issues or comments where wiki syntax is a problem? I dont know yet about any example of this problem +My application is partitioned into different modules each top level module owns a different part of the url for example my app has portfolio network +This problem will be addressed by +Verified at +looks good to me +I actually have some modifications to where you can make some number of tasks hang +If you want to integrate source and Javadoc into Eclipse use meeclipse check the checkbox and meclipse pull sources and javadoc integrates into your this point of view this has nothing to do with maven and is a wont fix +Sure we can punt on it for now but at some point itd have to be dealt with no +We use Thrift on very small systems and do not like to pull in additional dependenciesWe need just need a little subset of APR +This is why I do not mark it as a bug +Implemented in. +As soon as I get a certain MTOM issue resolved Ill be publishing a new beta with it in it though. +Methods to be implemented can be found in the and attachments to +Fix in openejb rev +Thanks +I put java for the and it works +The reason behind the constructor and initialize was the class needs to be class loaded during configuration because the specific instance is provided via configuration +Id suggest building and running Zookeeper with JDK and see if that works with its own ant clean testEugene +Moved. +Ok +I have changed this issues Release Notes Required field to Not Required as this issue is a release notes entry has already been prepared in another JIRA issue +Also now requires more ram than previous to build the site +Also note that while we have one global cache internally theres nothing stopping us from splitting out the different to different save files +srctestjavaorgapachehadoophbaseipcprotobuf did look at the patch +bq +Attached the patch for the issue +Closing the bug based on my previous comment +Should tell me all I need to know +Joao do you have any comments +Next stop is to try building the tag myself and see if I still hit it there. +Steve Can you please answer my query +Now I got the right click menu but Im still not able to create new archives +Attaching +multiproject does not exist in this time secondsFinished at Fri Aug EDT pm mhusbygallium workareasequence +attached patch should fix +And this operation will be probably slower that just iterate over all the continuations +The requirement is to have fuelType selection be dependent on engineType selection +Please verify in ER and text in release notes is fine. +Need to backport to have committed the patch to branch +Apply to trunks serdesrcjavaorgapachehadoophiveserdeobjectinspectorprimitive. +maybe call the tab Edited properties and have it show a No properties edited +When I apply the patch I still cannot deploy the war to the minimal Tomcat assembly using deploy o deploy due to the having a dependency on the openjpa config Caused by Unable to resolve dependency openjpacar +This issue has been fixed and released as part of release +My guess here is that we need to find a way to make the two resolution mechanism to cooperate among themselves but Id have to investigate it further +remove hdfs from the value field for +This is currently commented out I believe the test is invalid and should be changed removed due to reliance on GC timing +We have upcoming usecases that will likely use OPP and this was a blocker for that work +I think what we have in mind there is more in the area of millions of child nodes +Test illustrating the implementation of the above along with a fleshed out unit test +If websphere is doing the right check and we are doing something wrong I would like to know +Renamed to and it also creates and in the root of your project when run +Confirmed fixed after Hyperic drop fixed after Hyperic drop +A well tested patch +Please close if you are satisfied +Committed at subversion revision +Fixed in commit eafdfedfffbdae now depends on +Hi Leon Im ready to commit your patch but could you please explain a bit more what it does in term of functionality or use of this functionality +Another concern is that this design has more steps than paxos which is generally considered complicated to get right +I cant contribute time to the release since Im concentrating on and release lines. +So in other words this issue is that the schema version should have been bumped right +One indirect way for namenode to recognize such nodes is to check if latest stored version on datanode is later than namenodes +Hi Mickael it seems the issue is fixed +The MR execution code depends too closely on the stuff that went from class to interface to ever let people run against hadoop wo a recompile unless theres some trick to doing it that I havent figured out +Including Replication related code +GalderHas the change to make the method return a instead dropped? Should I log an improvement for that instead? Adding that change makes it a lot easier to extend the class and change the lookup returned +I can perhaps put out a special release for you that contains the fix +Its easy to promise interoperability over HTTP today +A real bug existed that cost us performance +Create the simplest app that demonstrates the behavior and run grails in that project +Attached the patch +I tend to close this issue soon if nobody the problem can be solved by giving more memory I close this issue +Ive put there some more comments but from my tests there is no bug +Messages dont get dropped for other operations +The test failure in capacity scheduler is not related to this patch +Comments welcome +Suboptimal for sure +With the query I describe I certain fallback logic +Preparation for Gora release candidate +One of the two paths listed in the editor is the ultimate deploy folder the other is the temporary deploy folder + +correct the forum link and set priority +However I dont think my change will affect these code Ill check it +This is designed to allow users to ship dependencies inside of their corprocessor jar +Looks just like my patch thanks! +If the value is explicitly set to false which would normally only be done in a slave the reader is always initialized and from a directory instance and not a writer instanceThere is logic in to open a new reader rather than should the underlying directory instance in the current reader not match that of the new index writer as that means that a full copy of the index was downloaded into a new directory as would happen during replication if the slaves version was ahead of the mastersThe patch was created on the lucenesolr branch with all tests passing I can create versions for other branches if needed +QA requested promotionRequires a patches suitable for both and +But I dont like deleting snapshots on all those machinesAfaik also schema updates create snapshots +What revision are you running? This should have been fixed recently +I think the resource path should already contain the context path and then used as is +A draft implementation of using FST would make more concreteMikhail LiyinPlease comment on how may be shaped to support indexing +Added testcase confimpincl. +This mac osx specific thing +what we maybe could do is to test if GROOVYHOME exists as file if set and to give an error message if it does not exist as file +I looked at the patch and need clarifications on what its trying to doThe patch does not set the SRID if we are using OCI version on windows +Patch committed in r. +meta microsecondsread block microsecondsSo just looking from system calls perspective it is however it does not show up in the end to end test because in general there are a bunch of other contention issues inside HDFS which hog CPU resources +Hi ClementThanks again for your helpWhen do you plan to release a version with the fixRegardsDavid +I have a junitreport in the Gant build itself now so if a problem arises with Gant and this task it will be somewhat noticableIf a new example can be found that shows the problem then this issue can be reopened or a new one opened. +Patch for the which processes the rollingPolicy and the triggerPolicy options of a in a +Thank you DevinI reviewed the patch +I was able to load the indexed polygon my adjusting the data near the errorI also attached a new test and I was not able to reproduce the problem you report even with the default dist +Confirmed +Perhaps if the exception named in the is a class thats loaded on the client and is permitted by the method signature then RPC should automatically try to construct an instance and throw it +Will this be applied to branch +Merged into with revision . +Berkeley has licensing issues associated with it and doubt well be shipping it +Please rename other short variablemethodfunction in other classes in the branch as well if possible +Guo please submit a review request on reviewboard or phabricator +nasty +I now replaced the spring dm loading of the by simple java code +Also see looks like a good improvement to me +Thank you Chris. +Thanks for the help +is moved from each to and now it will try to compact all if there are no compactions +this is the updated patch file with a option to show all files +container for etcWhy not container for as they are very closely related to namespaceShould belong to +Sending trunkfilemonitorSending trunkfilemonitorsrcmainjavaorgapacheservicemixruntimefilemonitorSending trunkfilemonitorsrcmainjavaorgapacheservicemixruntimefilemonitorTransmitting file dataCommitted revision . +Consider implementing adaptive flow control where the producer rate is automatically limited to match the consumer rate +thanks +no longer exists so the issue is not related to repository anymore +In Groovy you can always define untype variable +Verified it on PROD which was freshly deployed today +JerekI dont see the class in the patch +Attaching a patch rebased on trunk changes +It will require writing a bit of code that arguably replicates some logic in the compiler but thats ok for nowId like to keep the Jira open as a request for additional API but we dont need this for +The patch looks good to me +I think that more work could be done on clarifying which infrastructure classes are allowed to run on small platforms and which arent +afaik is still used for test suite +This has been measured with a rougly about million URLS +Previous title was confusing considering +Would the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Please review for me fine +I havent test it with trunk build yet +I ran the changed test multiple times on jdk jdk and j vm successfully +Also this issue with witeUTF would simply not exist +It looks like this is only an issue when booting using torquebox run which is of course the easiest way to boot things +I want to test performance relative to and and also consider switching from lookahead chaining to single regular expression per term type to improve performance +You can do that from Window Preferences +The fix version is the patch is ready but not picked up in +See +Ill close this shortly. +seems that the owner of domain and java net author are the same +Im working on this one +apply patch on behalf of Han Hong Fang with thanks +Default value is true +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +again erm. +There are no annotation hence events it self are not generated +DB temporal and join tests are almost ready +You should also update the example to include info about the new setting + for addingHowever I think you should also replace with +I am also trying to phase out the flags because they do not help very much IE they cause more problems then it seems to be worth +Latin db with weird chars +Verified in +probably a lateral effect on other parts +Might be the same thing but need to know whether its just Hypersonic where this behaviour appears +Instead it keeps them in an in memory store +Committed to trunk branchx and close after release +This is improving performance for some more logging improvements and cleanups +htaccess file that should make go to eclipse +Thanks for the patch +My pom a work in progress as I am setting up a new app to eventually use for demo trace Im sadly on windows now which I use for courseware delivery +With fresh clean workspace this issue seams to not be present +Is this patch validatedCan you please tell us when it will be integrated +This is a pretty important feature whats the status for? If none a quick explanation would be nice so we have an idea why it cannot be done +New SNAPSHOT bundle uploaded with a fix for this check it and resolve the issue if this fixes the problem +Disabling JSP precompilation in the module reduced startup time for my admin console in jetty from s to s +that patch didnt introduce any real changes I am going to commit tomorrow late morning unless there are objections +attache my progress +Bill Noirin already did via the infra list. +And somehow I dont think an Overview page for theOverview page would work all too well +It is not yet clear to me how this interrelates with JACC though +Ive added some code in order to reduce the likelihood of this problem occurring +The patch has been appliedPlease verify +Looks like theres a minor naming issue with the new from Products and the productsymbols table is imported with the SYMBOL column resulting in a SYMBOL is another column SYMBOLTYPE that gets created first and for some reason the new logic is thinking that SYMBOL already is a column. +Committed the patch to +Amend wait so the master ready state will wait for all SSH handing are done to match the old behavior that existing test cases are expecting + +Added to trunk and X you rock +Perhaps a temporary or committer access might make this easier for everyone? I know people are busy and such +bq +No longer think this is an issue. +My mistake +All tests pass +i agree failing a build is definitely in enforcer agree that having it in enforcer is a better idea then having it in versions chance that it can be included in roadmap for +Then a parent container can stop all of its children that are started regardless of how they were started +r caused a performance regression on jbb r is designed to reduce the amount of synchronization and correct the performance regression +Guillaume will see to it +I tested on dev as well with the latest EAP version of JBM +Never mind + +I agree that the value for true is subject to debate but the JDK IO methods use so I think it makes sense to add those methods and use as the value for true +If only i could figure out how to map a keyboard shortcut to restart the deployment for when that is needed +Code reviewedA small point in the constructor it would be preferable to initialize the workerThreads field before the shuffle field since the to could otherwise occur before workerThreads is initialized +Thanks a lotOh! the attached patch doesnt have the license +Hey JoelAfter reviewing the functionality as an alternative to using ngroups for performance reasons I have a use case that needs sorting specified in the search to be respected before collapsing +It seems the translation works as I inteded it to +It was released with the fix in +Thanks. +As Xavier noted I cannot reproduce this issue with the latest nightly build +Perhaps a different property should be used for the actual number of tasks allocated +Please see the design document in for more details related to upgrade +and are the two relevant patches +Admin users can currently modify protected items +As soon as you hit Create you get redirected to a page that tells you that you requested a page that does not exist +regarding the last comment I had checked out the tag rather than the trunk which is why the patch didnt apply smoothly +This bug has been sitting in limbo for months now +To build it its only necessary to add client dir to libraries from eap distribution with installed hq +The VERSIONSOAP variable has a value of +Did you notice this +We basically need to enhance the to allow configuration of the flow definition attributes +We need to use karaf activators to solve this +I will check when new build is H no exceptionIn log I see cannot obtain entry for path from jar DJavacommonlib +Here is the patch to resolve this issue +Assign back to Anton to verify in the next good build +A srctestorgapachehadoopexamplesA srctestorgapachehadoopexamplesA srctestorgapachehadoopexamplesA srctestorgapachehadoopexamples as ticket except things have been moved into srctestorgapachehadoopexamplesand the abstract helper class has been renamed to ironic +You cant because Weights produced from same Query are different for different indexesYou can probably modify Query inplace for a given index produce some scorers do scoring then modify Query for another index produce scorers etcBut now your Query is no longer and I cant reuse it from different threadsSo with all its strange looks the trio of Q W S is still the best approach if you ask me +The reason they had to be removed is described in this issue +All tests pass +which is no problem as DIH is contrib and experimental move t fix for in we should move DIH from Reader to fixing the version info for and all affected issues have in all issues not touched in a month to Does your concern here entirely have to do with when DIH is indexing XML files +Well the linked issue is fixed thats been the pattern for all EDG jiras +Will do ASAP +At least AS should be would tend to include it to Embeddable too +Minor update to add a set method to the async client +the problem happens before twix transformation +Ive added code to ensure the current get method instance releases its connection when an error occurs but Im not sure of the wisdom of releasing the connection before reading from the methods response body +Thanks for review +Ill commit this against head once the release tag has beencreated +Can you give a more concrete example +I rewrite the loop such that it can assign only one task at a time +Thanks! Testcase applied +I dont know how far the is away but it can eventually replace whatever is going to be implemented here in regards to that flag +Committed revision +I also plan to backport the fix to. +Hi David +Looks good to me +Thanks for fixing this Rick! The hidden error message is indeed much more helpful +Rebased ll have to use deprecated minicluster classes for MR for now because of +But please just make this contribution higher quality +The FOP parser thinks your are providing him some option when using hyphens in your file nameFop d xsl xml pdf hope this helps in a within filenames are of no concern for FOP except for filenamesstarting with a dash which are supposed to be optionsThe error messages is caused by a third argument consistingonly of characters which are invisible if printed on a consile likecarriage returnsIt seems that the combination of WinME and some JREcauses this kind of garbage to be passed to the Java code +If developers have different sets of dependency projects checked out then they will need to have different +If You feel that this decision was wrong please reopen the issue and provide more information on how lognet causes the reported problem +one more questionwhat is the recommended nodetype for the jcrcontent nodeand why will it be changed in the + +ini control for thisIn regards to Antonys comments I definitely agree having a more dynamic timeout mechanism would be best but would require me more than the expected or so lines of codeAntony also when you say do you mean a heart beat to or all the way through to the HTTP client? Id assume just to in terms of a Back off mofo Im still working! sort of way which we dont have yetIn another slightly unlreated context I contemplated the other day passing a UUID with all requests to osprocesss +Fix by Ill take a look at itDoes the aforementioned configuration make sense +Sure I will regularly add here updated tests +hope to help out asap +They are not available from within scripts +Sorry bout the slow review crazy couple of days +The code in is tries to identify the implementation and to choose the right strategy +Since I implemented this component I guess my responsibility to fix it as well +With simple testcase +That will leave only the case of commit in which is a simpler fix +Thanks Espen +Current set of code if its of use to anyone +Need to update Remoting Guide +QA Verified. +Applied your patch in Rev +Just curious as to whether you have some empirical evidence that this is an issue or if you are just guessing? It would be nice to see a comparison +Having them as test artefacts will not serve the purpose as we want to test actual config files shipped with the workaround to have the test running properlly using product config files is to execute similar script prior executing TScd p buildtargetstandaloneconfigurationcp JBOSSDISTstandaloneconfiguration +So they added to the composite because a performance contract violation has been raised +oops is the full patch a is a diff from your patch +I am going to defer this problem until after the next Sanselan release. + you shouldnt be touching or any other EMF things directly +That will obviate the need of Out class which is just a wrapper +The current functionality in adwice with and whatnot can already do this +unlocked the guest user +Thanks Jitendra for the review +This issue is duplicated by. +Using the falsemade no difference whatsoever +codes look Nicholas +There is a global metric that accumulates the size of all read by a region server so its not what you meant I guess. +orgmortbayutilCredential matchesBinary file +Thank you MarkMaybe we should document this issue in something like Harmony Build Guide to remind other people to use the latest make +The JPA spec itself isnt perfect but not persisting the discriminator value is even more pointless than just waiting for the JPA spec to be updated +Closing + +Sure +What do folks this of this? Other thoughts? Suggestions? straight to subproject path is only available if the code base is from a single organization +If everything goes OK Ill check it in +fix on javadoc +Want to take a looksee at Matteos latest Ryan +Thank for reply +Use the Felix framework constructor +Thanks for the review PavolYou are right we have missed this instructions in have reopened +Herewith I have attached patch which provides a workaround for this issue +these they needed the license header and indenting +So Ill have to do some other work tofigure out what changed on the trunk and how my project is side note Im now compiling with JDK +Thanks a bunch JukkaIve spoke briefly with danielsh and may look into helping create this Web Form +Thanks for contributing +If user would like null support he need to set a system property like +Thanks +Thanks Martin! +Two new Main classes have been addedwhich test and plain Jexl respectivelyThese are very basic at present and need enhancing +Applied Toms patch rebuilt and ran all tests. +This is kinda messy +Thanks for taking a look Lars +Switched trunk to in revision +Of curse +The server class seems very robust +Throws if authority isnt set to This way upgrading users who didnt care about authorization in the first place wont have to update anythingThis is friendlier than just exiting but Im not sure if we should be friendly +Going through the Libraries tab on the Java Build Path screen shows the configuration screen with a Resolve Dependencies from Workspace project checkbox which does exactly what I was looking for +Tom The latest patch on RB looks good +Hiya are four properties that you need to configure for ftp access +new version of the patch +Those are only thrown when another thread is trying to do a clean shutdown +Hi Erwan thanks for your contrubutionYour patch is party implementedCommitted revision Regards seems that something changed in Buziness Time rendering after patch has been appliedAlso bullets are not renderend in in Buziness Time I have created for firefox and ubuntu working fine herecan you give a hand fixing thisYou use firefox under windows +This bug has been marked as a duplicate of + +I just committed this +Patch checked in +I do not understand how it makes sense to allow distributed execution for local cache? We allow MapReduce on DIST cache and distributed execution on REPL and DIST +But if you only have argumentsshould be read as But if you only have argument +tested with jclouds recently synced to maven centralcommitted as revision Thanks Andrei Cosmin Andrew for writing this +javatestingorgapachederbyTestingfunctionTeststestsjdbcTest case for embedded blob set methodsjavatestingorgapachederbyTestingfunctionTeststestsjdbcTest case for blob stream +Then when responses are received it should be possible to optionally handle them as they come in or alternately all at once to merge the results +If you can still reproduce this problem please attach a test case xml and xsl +Interceptor called twice for each test +Now it only clones the term that it receives from the input streamWould be good if someone could take a look at the ngram changes +Patch looks fine +Attaching a new version of the test incorporating the following changeso Cleaned up the output script +Jeff this is actually result of bug in management model definition but you have this already fixed in your RD conversion branchalso this exact behavior is present in and messaging subsystem is not the only problematic one +Because I think was the correct change the default Lucene queryparser is OR and there are many goodreasons for this +Want to do a bit more testing before I commit +Again as I recall + handles this elegantly +The WAR attached to will reproduce in M +Hm would it maybe make sense to distinguish between plugin artifacts and regular project dependencies here? For the case reported here the warning is apparently oversized since the plugin is not actually used in the build + +Thanks I applied the patch +Will be doing the integration next week +Otherwise there are too many cases where the user can still hose their cluster +Closing issue. +Will add Spring tests to the component as wellCheersAshwin +Hi Michael hi Jacquesi committed a modified version of the patchThe part Jacques recommended is used by the condition list +Im surprised to see the pages in the document base directory are based on +This means that any flow control on the address will still be adhered too +If its set to I got an error similar to the one you posted +Why download the models from a separate place? The models are not Apache licensed +to TRUNK +Postponed to +Log of the decomposition over a nested join was not appropriately creating staged group by with the needed grouping columns + and related classesFixes regexpmatch support regexptest support regexpreplace support +Under Mail Configuration someone changed the from address for notifications to note the extra j +to and trunk +The java files are not compiled because they are not located in srcmainjava and srctestjava +If we want to get our own RPC this should be. +I tried this myself +META +So we need to deprecatedremove this feature and rely only on people adding refs to itAlso this feature is very questionable as end users want to have control of which routes is used +Thanks change from directory to really solved the issue in the simplest make it default behavior in next versions of maven plugin for netbeans +Thanks for the contribution Ioannis +Camel slots have been removed so the only remaining slot definitions is +Knut thanks for your comments +While working on PLEAC and I also noticed a couple of other inconsistencies which I have addedcount and join on iteratorhead tail first last push on listsreverse on arrays +Every time we create un record in database a message Successfully created is added to we want to replace it by the equivalent in spanish Creado con exito +After some more work Ive come to cancel my both assertions which say NPE and there is no WSA and WSSE headersIve noticed that I called stub of serviceA not serviceB I didnt add and parameters to the ofBy the way Ive just seen a strange situation +Closing the issue +Parts of this bug have been completed and other parts are unfixable +The Bean Validation will actually validate +Otherwise it looks like the committer is stealing credits from the contributor +Some additional background similar request on hadoop core with list of procon +The method is protected so potentially could override this but I wonder if a defect should be raised against Spring +A patch that refactors code to define sender and receiver queue +Instead we pass a new context to the formatter duringthe rewriting +I have the right to sublicense an unmodified copy of the JAR +each coprocessor is represented by a string the potential savings can be considerable especially if many regions are hosted on each region server +Assigning to +yes go ahead +Answer from developersThis does not appear to be a bug +As per your exception stack RM is failing to start +The fix should be in the next snapshot +Created branches to work on this +Fixed test. +So its not a bad option to have +but testing on as well might be helpful +Maybe it doesnt always execute first? It could possibly make a difference if you use instead of with that servlet having a lower value than all other whats the full stacktrace for that get call with the thrown? Who calls that method at which point of the bootstrapping? Thanks for quick response +I changed the interceptors to be serializable +Looks good to me +Attached updated version of the patchChanged isUnicode to include characters not from but from upwards +There is an alternative syntax for declaring parameters in the Criteria API and there may be performance gains by using literals in some cases vs using bind variables +That way this is private to the codec plus recored in the index in via the codecs name +I suspect this issue has gone away in newer builds +Pushing issues to the next CR +I need to write a simple testcase to check +But the error is very very confusing +Ill see if I can come up with something +Yes +Pageflow parsing tests belong to +In Java we can then either implement this from scratch perhaps stealing code from or we can try to implement it using a framework + +Defer to +ChristophChuck do you have an authoritative answer on if supports +The ability to create server based on another one has been removed +Either we make this a real visitor pattern and add accept methods to Query which I dont think is necessary or we should make the name specific for the task +pushed to works great +My understanding is that the type information is only stored at the level as the type alias and could possibly be cached to aide in deserialization but that seems like a bit of a hack +Perhaps if it is to be moved back that could be handled as a separate Jira +However I view unit tests as fundamental to ensuring that the code has been built correctly so I dont think they should be viewed as optionalAs you can see there are a lot of personal views above so Im afraid youll have to wait for a more official answer from someone more knowledgeable than I +I will pull the request in this weekendthanks for pulled in +It might be time to retire poolMap as well +If you want to test it look for the plugin in your plugins folder and replace the with the attached oneTell me if it worked for you +possible issues +This patch will fix the overview +The neighbor region would have new delimiting key +I agree that the parallelism determination should be to enable other logic to reuse existing codeYes +yes Properties is probably in the top of the worse classes in the JDKThese methods are useful beyond debugging +A word of warning I had to add some trickery to be able to test this change without adding a couple steps to the buildtest process +This bug is only seen on branches where has been fixed +Attach version of the functional spec incorporating Dans feedback +Start cluster with the upgrade option +Only problem is with update tool +ATM there is some interest on our side to merge on as well which would mean gs s leave it up a bit and remove the fix for since the merge on is ready +Thanks Dheeraj! +r for the rest of it. +NamitSince this bug silently results in wrong results I would like to have this fixed before is released +This issue was fixed more than months ago as of r +since the patch is not working absolutely correctly +Please note that update went without described issues +If you are unable to change the classes output directory for the plugin please reopen +Imports of have different Plugin dependencies in my setup +I would expect the changes to be based on when I click the buttonI think the way to solve our issue is for the diff request to return the things that did not change as well +fixed some overrides to compile with java prev patch did not apply cleanlyyou need to run svn mv as ben last stated adn then apply the patch +patch to fix this issue for mysql +Spring must either autowire the untype variable to any defined bean int the application context or not autowire any defined bean but both solutions have unexpected side effect that will surprise the programmers +Manual available on page Apache Derby Documentation for anyone to access. +Im marking this as not a problem because the output is the expected behavior when youve got set and falseIll remove the setting from the nightly environment + running on tomcat +Please test with new with +If this isnt related to the first issue let me know and Ill open a new go ahead and file a separate bug for the SYNONYM thing +This reverts a change that Patrick Wendell made in when adding RPC tracing +the console log for the bundles is set to auto then the bundle is also started by the cannot decide whether this is a behaviour compliant to specification + patch not possible so I made two +Fixed similar inconsistencies at progress as with other database CAHNGELOG and entries. +If just felt that output optimization based on the existing deficient architecture was not worth the trouble +What do you think +My pleasure thank you! +Adds +bq +New patch addressing Alejandros feedback +Can you please add a test caseShould this be a blocker? It could break Configurable Writable implementations that depend on a configuration +If you finish it before we release beta please correct the version +Just a thought sorry if you think Im jumping the gun +This issue was migrated to +Resolving +Tests pass +There are no tests because the output of the page has not changed +Attached is the patch +Isnt the directory found in conf? i can remove the RTE and make it log if not found +Also the method is free to build whatever Envelope it feels like it just guarantees it will return a new Envelope +this fix to master. +Will be done by integrating you wont fixed this issue +Yes +On those old files are going to get +also qa bot found some findbugs issues +So to disable it just add +I am attaching patch here +I have the same error am use Groovy and issue is really annoyingid really like to use but this issue is a no go for me +It might even be nice to make the code common between this and the resource manager +on current proposal +javadoc warnings are unrelated +Would Hadoop require recompilationUnfortunately it would +DOH! Thanks! I was staring at that Admin screen and for some reason my eyes didnt see the Maintain featured projects linkAll sorted now thanks +If they run into an exception they can turn the variable on to get out of the problem +Once that gets resolved I would say a couple of hours from my side to resolve the mdb security context and when any issues are brought to my attention I promptly fix them such that the tck work proceeds +The current commit adds responseHeader twice +However it might still be desirable to suppress the completely and handle the response without s +my quick guess is that should output +looks to issue as it has been released as part of release +So what is the status of this +Some cleanup and creating new diff wrt trunk since the previous one was for the testing I just want to be sure that nothing catastrophically bad happens in these cases +Patch that cleans up all the warnings +Wow Jon youre on a roll +This certainly isnt an exhaustive list +No IT for that issue. +That poses problem then because what they do goes against the spirit of the specThe service file is a public API so are the implementation and the Configuration subinterface +How would it run under Tomcat only? I dont care for the arguemnt formatting in Anyway theres plenty of good stuff here too +Hence the key must be serializable +nobody had noticed it yet +Agreed thats how I see it too +Verified that when and are truethe reserved value is SOoverprovisioning factorIf and rae set to false nothing gets reserved +Useful for small clusters agreed +in branch +Small change to Command to describe the arguments +I can do a with just this patch +for the to trunk. +both m and surefire where from svn m beta snapshot and surefire is most likely fixed then +Hoever if this starts turning into another debate Ill defer the issue to someone else Im more interesting in cleaning up issues that can and will be done quickly at this point +Actually I can just automatically call describekeyspace and use that to setup automatic value conversion +crap Ill look into it today +This one isa feature request not a bug entry and I wasnt sure that scabanovich would understand from the titile this is what we had spoken about +Works for me without a problem +Awesome! +Marking fixed in ER +My initial reaction was it was not worth for this either +Jerry Is there a patch attached? I dont see it +Previous patch was invalid upload +The original purpose of this issue and the attached patch was a simple consolidation of two css files nothing more +I tried with and all problems persistShould I try with snapshot? fix is supposed to be applied at but it still does not work for me +If you diff it against the previous patch you will see that this is the only difference +Committed all the changes +Possibly +Fixes and deprecation done in r +Works all like a charm +Sebb comments please I spent quite some time on this +pretty weird +commited new websphere reference guide chapter and changes to examples +Btw I am assuming that you are on Java or higher as I have taken the SVN tag for Castor as the root for this SNAPSHOT release +Note we should make client cache configurable but that is fine to come later +Call of returning empty of returning the output of wanted to comment on this issue from Team as we want to say that is officially supported on Railo +Can you provide me with instructions of how to establish a JDBC connection to Pointbase download it install it configure it add a new database etc +Simple half line fix +fixed and backported to and +Andrzej did you have the time to review my new patch +Closing issues in RC. +Yes you are right +verified and pushed to with JBDS +I checked another project of mine and it doesnt have that setting either neither does the Cordova or projects when I open them in +The idea seems OK to me. +Actually the Future support was added without depending on +Thanks +So if we treat all these as resources slots memory disk space etc maybe we should look at a uniform way of describing these +This implementation is very simple and it basically demonstrates how sorting can be avoided to solve query problem described in and +Marking as WONTFIX as per Matts comments and the lack of reply since. +If versioning the codecs like then could we have accessors for tags in the Cell interface but would throw while a would support it? Andor a method in the interface that a user can interrogate for capabilities +Apologies for the noise +Merged to doc branch at revision +were backported to as well adding the latter fix version. +It suggests that something isnt being notified correctly when failover occurs +Maven project to build Spring integration for the patch but I decided to go my own way with it +Feel free to should the issue persist +Thanks! +Jared passing to you for EAP +Instead of using a pool kept in a static global variable now the dedicated pool is returned for VM from a component manager +bq +Voters watchers vote for watch that issue. +bulk close of resolved issues. +Checked in for close for +will pull back to V if this works +Same patch but it applies at the root directory of the guys Ive added a test +Its easy to implement both in the server and in the client +How does that sound? +Stefan Si think were done with this issueI have searched for image creation in the whole project and they all seem to be using the nowI think we can close this issue +They are useful when you have to do with some legacy code which cause naming conflicts otherwise I find them harmful +The only thing left out is the dependency on Saxon which isnt in the repository so I left that out and I think this may be optional if you dont want Saxons support +attach a patch to provide killadmin subscription mode as description +Thanks Jyotsna Suryajs Patch is in trunk rev +I dont think we need an orderable variant because this relates to the child nodes right +Ive run into a challenge in creating a patch for this issue +Defer all unresolved issues from to +Thanks +the file actually shows up only the validators behaving badly and wTP bug +The skeleton project I sent you doesnt have any code that references the Spring Security packages so changing it to runtime works. +Moving the other annotations support to another issue +Is reflections tricks the answer +Fixed +Thanks for pointing it outPlease make sure that whatever gets to also gets to +Committed to trunk +Tested on geotidy using concurrent threads each one asking CRS and transforming a point with them +Ron thanks for the heads up +Thanks everyone +I did that but then it is not compiling +Ive checked in a patch to both the main branch and branch +Doing more testing with the release and having EJB team code review changes to classes +Just to have a look at the log tonight +That should be it +Check for updated original JBDS was installed with the universal installer and then I followed the steps above +Looks good to me. +especially the resolve add method relies on the new get and changing the others to use this method also ensures that they will be consistent with one another +Yes Im on Linux +Serge it would also be helpful if youd post the test class you used to repro this +Marking as resolved to mobe JIRA out of open bug list +Patch attached +Yes thats think this is as good as we can make it given browser this issue +Thanks Chris! +Ill debug more tomorrow and try to see if I can turn up anything +Pavel since Im unable to reproduce the problem Im closing the issue +I started looking at this but the configuration doesnt even start properly +does not match there is no in our repoAnyway before you make a new patch +Attaching a small fix to handle the case of the thread stopping the daemon being interrupted while doing the join introduced in this issue note the interrupt and retry the join +Hi GuillaumeGood point! So reopening this issue +Should we mark as a dupe of this one? Im marking this for too +I think should add releaseautorelease mojo to do both releaseprepare and releaseperform +only in a few deployments this feature will be used +Problem solved when +Could you please describe a bit better the issue? What concrete bits is this issue about and what does your pull request addressProperly describing issues and the expectations of the fix are important to understand what is going on and will help us generate better release notesThx +Im unable to replicate +I believe that we are planning on supporting this for though +Updated the DD of the sample +not a problem +Heres a preliminary patch +Added diff file with correct suffix +Sorry for the screwy english +It covers Bean attributes Collection attributes. +Close +All unit tests are passed +Sorry +Matthias can you send a pull request for this +and I did attempt unsuccessfully to resolve by refreshing all packages and by restarting classloader support the reinstalled bundle and the bundle. +Perhaps you could have setters for these +This is actually true +bq +Reproduced in both trunk and versions +This must be checked in after good This is hcat only change so I will make sure it doesnt break hive build and run the hcat unit tests before committing +Point being this could work but it cant because Hadoop is bundled in the jar +Edwin you can decide if you want to track problems here or not since Crimson doesnt have its own bugzilla category yet does it + +Again the xml file size was MB +Maybe this should be discussed on the dev mailing list before any work is done +works well with upwardsnot a on +Thismay be ignored +Tests are different story +Agreed that is client side mount but that was a way of getting this solved +Best to wait for JDK source level +Youre right it does appear to be thread safe! Hmm +One is to generate random timestamps +Attachment has been added with description Contains images sources to transition resolvedfixed bugs to closedfixed +Ill definitely wait for that to get ironed out and committed +The plugin will then start +Unfortunately I dont have any thoughts since my last comments +We are generally very cautious about adding syntax unless there is significant demand for it +will address supporting ROCG at the connection framework level +If you desperately need to do something like that wrap it yourself in a of some kind + +Kicking change looks good to me assuming that Jenkins comes back with success +This is Coprocessors dangerous is implied +the patch has been committed thanks snathosh +Ive prototyped a fix for this and it took the from seconds on restarting one VM down to ms +Committed in trunk +heres a major bug that we have a solution for can this make +Could somebody give this issue a more accurate and descriptive summary line? I mean its not really Guice OR Solr is it? More like Guice AND Solr right? Thanks. +This bug is a defect in the committed version ofThis fix is clearly important +you are rightThis issue can be closed +Ill try to provide a patch the next days +Wait and see +If the tool supports an exclusion filter and it does then the BC Wizards also need to support the exclusion filter in order to be good citizens IMO +Then provide a no op rmi marshallerunmarshaller by default so no extra overhead of marshallingunmarshalling twice for marshallers +not show Termination type and reason fields on Create Employment forms +attached testsuite changes for patch review of this would be good +Isnt more clear with subresources? See and let me know what other informations you need +Fixed by Sheng +I guess changing it to an INFO might be appropriate +I would like normalize to be and to not change whether there is a final line ending +Im thinking more in terms of parsing configuration. +user with name zun seems duplicate in DB and causes problem in JIRA users +KevinI reviewed the problem +I may be wrong but doesnt the padding only occur with SQL columns of type CHAR +Approved for. +I have a static final boolean that is set to false +I think this can proceed now that is done +I guess Olivier already created a ticket in Tomcats issue tracker +if you are waiting then you should assign it to someone Anyway patch makes sense and should be made +This is the correct patch +We minimized the number of methods for now +It explains how to get dependencies included in the sar archives. +so is this going to be fixed? we need to use this in our project +enhanced header file inclusions for breaks the build in linux therefore I made some small changes and applied the patch +The class that I added the test to doesnt really run the tests if youre on Windows so watch out for that +No further objections were raised so Im proceeding with the suggestion to resolve this as wont fixwith the outcome being that we expect and prefer documentation submissions to be provided aspatches to the DITA source. +Thanks for the patch JamesI fixed it almost like you did in the patch just I preserved the method signaturesNo API breaks no new classes used. +I really doubt the timeout interceptors are the cause of the problem +Thanks +No problem I just wanted to ensure no mistakes got turned into factsIt seems the debate is lively and alive on the mailing list so I guess the best bet is to wait for things to iterate to a conclusion there and then post a summary item here +It seems that all of the tasks are done in this issue +its in there +Thank you +The problem is even worse in current SVN trunk both your tests gets into an infinite loopIve tried to commit a fix into SVN trunk could you please give it a try and post your feedback quickly so I can add it to RC tomorrow +Ill try if Batik Beta behaves differently and Ill try to provide a test caseIm not sure if is affected since I dont use it yet but I can give it a try +httpdiv appear to be hosting a bundle repository using the above and I would suggest that we follow their example +The above is a good description +I could check this is but Im not quite sure to understand what it does precisely +Ive provide steps to reproduce as I think Daniel can reproduce it +Is there at least some kind of workaround until it gets fixed +Do you have a timeframe for uploading the new bundle +ok committed a you very much! +I changed this recently so that if type conversion fails a catchable exception is thrown +Why anyone would return a raw address as a hashcode I can only guess +Seems Eclipse got something wrong here +The changes look good to me and I dont know how this could impact the failing tests +Since it is the JTA API which is blocking and thats the info is in the log i close it please reopen if you dont manage to get the info +It sounds like characters is the correct limit but we need to fix our logic where we determine the actual length +Actually I have got an EC tax number so we could do it +Regardless of the actual solution I think passing paths is a bad idea. +can AP include references to logback in its build scriptsYes but only if such references are optional and not enabled by default +looks like it though theres no version tags in Ill track that down with I am reopening this +You are missing the avro generated files +But I would not merge it to since I had to refactor the code we use in the palette for loading an swt browser and move it to to be able to share this code +Closing resolved issues. +Modifications on how to calculate the order of the days to show into the calendar +It does not support NPNchecking whether to enable TLS extension support +Is it for the shell in the Eclipse plugin? Or do you mean the Groovy Console that we launch from theIf its for the latter Groovy s console will feature line numbers in the source area +Youre right I will try on hdfs list +Benjamin Could you check the Gant release and see if the problem has gone away +I got that point +The problem I ran into was when the out handler read back out the rampartcontext and then went to use a certificate from it to sign a response there was a segmentation fault when accessing rampartcontextkeymgrThis segmentation fault doesnt occur all the time otherwise Eucalyptus would always have segfaults +Just +Heres a patch which is a bit bigger but should be deadlock free and cleans up some encapsulation issues as well in the process +It can be reproduced with or +Same issue with +Cleanup patch so that we dont close statements when resultsets from those statements are still active +Same patch as above but with granted license to ASF +This seems to be a simple and good solution as I see +defer all issues to +Not sure HC and can coexist Webdav provider files inherit from Http provider +resolved in +The same problem still exists with IE BTWThanks AngeloFixed at revision +for the time being since there doesnt seem to be a compelling use case for child threads to inherit transactional resources while keeping resource +Hi JingAre you working on it currently? I would like to try branch on but I need as well the write path without most of the time is spent on the write errorsThanks +Could you convert them to spaces? It looks like modifying srccsrchashtable is not necessary +I do not end with catalog suggestion schema suggestion +I attach an incremental phase patch so you can review only phase patch by itself +Same issue exists in +checkFile now checks file replication along with block replication +For me Spring IDE works well on IDE +Issue fixed for and in commit on trunk +Renamed multival multiValuedAny comments on functionality naming or conventions before I prepare for commit +So my patching is fake in this case +Will try to get to this for M or at least mentioned on the pull request its probably fine for us to mark this issue as resolved in that weve done a big bulk of the work and remaining efforts will probably be incremental and as time allows +Trivial patch to fix this +Attachin inital patch to do the same +This is certainly specific to this think that it is due to the fact that in some plugins tests we call the clean goal after the test and not windows is a little bit slow an handle isnt remove and the delete step fails +Thanks Mike! +See +This patch resolve some dependencis in and resolve MULTIPARTFORMDATA problems in in rev +Ive committed this +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues hossnofix +Looks good +Checkin number merely changed the nature of the failure from an NPE to a failure of a test case +Maybe the logic should be put in so that all can inherit it +on patch as well. +This should still be dealt with for repositories m assuming Mercury will be the facade for those. +And checked Copy projects into worspace +Thank you for the excellent patch Bassam! I changed it slightly to refactor out the get method into a common class called +Sorry I cant entertain this nonsenseLucene and Solr merged as equal projects +Will commit soon +AshishI agree that adding two lookup buttons there its a good idea but please be patientJacopo +But the way you have suggested use java anotations which is specific +HaikalDid you mean to state update jpa ormProvider HIBERNATE jndi jdbc so you can use the JNDI provided datasource with your which is supplied by the specified? optional jndi has been introduced as of rev +Yeah looks like the fallback code is always used instead +in those lines I dont quickly see thread creationwhere are the extra threads coming from I wonder +We really would like to avoid guessing how your projects look like and can directly create test cases from your projects +Locally its fine but I still see no updates at the web +You mean Resolve? Only QE can Close a JIRA +So we will need a new repo name without incubator +I think adding schema hashes needs a bit more work and discussion so Ive opened for that +pending Jenkins + +No strong opinion either i guess +The above patch includes fixes for and as well +Hi RaviIt looks like the that the new test depends on the file created by test +Not opposed in principle +The issue reported here seems to be an enhancement and not a will change the category of the bug +There are countless times we need to be able to share session data between controllers +Any idea why +Thanks ApritAlso thanks Chris for reviewing this +But there are also sql scripts for Oracle XE and DB you want to use another database system you just have to change in the EJB module and in the root of the exploded EAR +Point is not perfecting the test cases +See +You know a belts and suspenders kind of thing +I have added javadoc notes to the API to document how to properly control the lifecycle of the routes eg use the API from +dependencies should use the groupIdartifactId sintax and not simply id +Richard can you check the quickstart into the product branch? Trevor has the branch name if you dont +Well just map cssjsgifpngjpg requests to servlet which will read it from resources directory in maven project +It was only intended as place holder for a good start to a solutionGary +The following itemizes the resources that have config tracking and what properties are currently supportedResource JBN Agent server version Agent Java Java Linux platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default Solaris platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default HPUX platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default AIX platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default Win platform arch Architectureversion OS Versionvendor Vendor Versionram CPU Primary Secondary Default File service md Message Interface service mtu Maximum Transmission Unitflags Interface Process service user Usergroup Groupexe Executablecwd Current Working Windows Service service path Path to executablestartupType Startup typedisplayName Display nameResource Apache server version Apache Versionexe Server Server Tokensbuilt Server Server Apache server version Apache Versionexe Server Server Tokensbuilt Server Server Adminmpm Server server version Apache Versionexe Server Server Tokensbuilt Server Server Adminmpm Server server version Apache Versionexe Server Server Tokensbuilt Server Server Adminmpm Server server version Apache Versionexe Server Server Tokensbuilt Server Server Adminmpm Server server version Java Java Build Version JCA Connection Pool service Criteria Connection Pool server version Java Java Build Version JCA Connection Pool service Criteria Connection Pool IIS server version IIS IIS service docroot Document IIS server version IIS IIS service docroot Document IIS server version IIS IIS service docroot Document resolved as per meeting on +Let us know whether youre able to resolve this afterward +Not yet tested under a proper security patch needs to be I will do this +m not sure I agree that theres anything to fix in the servlet +only makes sense as a port to bind to and this config is also used by clients to connect to the resourcemanager +Select Customize View +Thanks Amareshwari! +You may have seen from the other JIRA issue that other people also did not want a minilang implementation of the checkout process either +As far as I am concerned I would expect a fix on both side +attached patch serves as proof of conceptadditional jar files are not included in frameworkwicketlib +Absolutely +Just for the record this is now fixed in jclouds and will be part of the release +For the second question the default value can be no size limit to keep it consistent with the old behavior +This should be pulled into the branch +Submitting the third patch +If so can you close this issue +Thanks Nicholas +An easy way to reproduce the issue is to send x number of messages to Queue A in a transacted session where the for Queue A is +Whenever ad new default resource is added all configuration objects are reloadedTODOmove the default to src fix all the documentation related fix the code in contrib which uses we want to get this into lets file a separate issue for the documentation as documentation changes can be made after the feature freeze tomorrow +This patch needs to be applid as well +However the patch has been delivered to the customer that has confirmed that the problem is solved +Resolving. +for the patch +But I thought for maybe they can be converted by extending in just one time so I havent created subtasks and would do so if necessary +We should only see this issue after a reset or an undeploy but the issue will be documented in the JON guide +Do you use tiles or some conditional tags? I had the same effect and resolved the problem by defined the id for each JSF tags inside the conditional tag +Please verify +I had assumed this was about deployments which would have made a lot more sense +but I guess that should be pretty easy to fix now we got the xcf +I still need to investigate it a little more but so far so good +PS the code above currently only builds with skipping the tests +I just noticed that the pdf version doesnt state Beta but html versions do I will fix this before release +Log warning instead of exception +Look at screenshot. +This patch looks for the checkstyle jar in libexcluded and also excludes this directory from the build classpath +Resolving because the redirect is working now. +to be reopened if we find an actual problem in Springs iBATIS integration +Marco I was looking at the problem for a parallelconcurrent logins happening in the system and did not visualize the issue from sequential logins perspective +nick there was a discussion lately about extending the server view +Attaching version of the patch which adds Java and XML DSL elements for the new +I will run regression tests tonight +The initial login screen + +Yes I think its the most we can do at this moment +If I did it right then it went from INSTALLED to INSTALLED with a stop in between stating that the bundle could not be found +Use this verion of the patch +is referenced by SDN since the problem goes away if I change that +Do you have approximate values for x and d mentioned in my comment above? When you have threads you could be running out of different kinds of resources memory kernel memory or ability create any new threads etc +I also want to make sure this isnt also an issue in the broker due out +jar should already be on the classpath +Fix applied +Would be great if someone can check and time Tijs Rademakers explained it clearly +This class expects an URL as described in the associated forum reference +Youre right +It can now be invoked by the toolbarmenu item Interrupt Script that get enabled only while the script is running +so it works but I had hoped this would have fixed the bug +I dont think I have permission to check code in +Very nice Cheyenne +So when sending some VK event is sent it needs to check if it needs to convert if not use the a key is typed it needs to check if the key is not on the reference keyboard and then what key are used to access I dont know in detail all your code maybe I am missing some crucial issue or some specialOlivier +Thanks for your review and suggestion Konstantin! Ive updated the patch to use the taskID to seed the RNG +Looks like my IPS was blocking it +Its all rightThank you Nathan. +nothing changed functions verified +As I prepared to remove the deprecated method from trunk I saw get +That must have been it the test works for me now +Attachment has been added with description glyph warn transition resolvedfixed bugs to closedfixed +Fixed on trunk waiting to backport on until the freeze was closed by fixed the problem DB works +is the bound class but its recommended to use +refactor a bit +Sorry I thought was +Closed upon release of Hadoop. +Hi Stevenplease update this ticket ifwhen you copied stufffrom distxml to distxalan to close this ticketThe Apache XALAN PMC done say the participants. +new release +bq +Saw this on master last night with mvn P developersystemvm cleaninstall +I have been svn mving them my end first before applying them however your patches include diffs of the files in their original location as well as the moved location + +Duplicate of docs are updated when a a plugin is deployed or when the target is run. +Hence this situation should be handled synchronously to the service unregistration notice +Yeah I was afraid you were going to say that +and we cant reproduce the specific failure without further to add support to what Juergen said the jsf implementation that was packaged with Exadel seemed to have this problem for me +Ive emailed Tom cos I dunno how to debug this further +Done ages ago +this patch only fixes this minor issue of null checksdead code it doesnt add full validation to +Patch for upgrading to Scala Michael for your patch its applied in revision Crasten works great +The project build system is Maven full stop +Im pulling parts out so that we can make progress +Applied branch and trunk +The GRACEPERIOD event is broadcast whenever the container is waiting for the initial set of services +That is the expected default behavior see you look at the other issue? Embedded connections will wait for their vdb to become active before proceeding +that list needs to exist before I can include it +WOW amazing thanks! Please dont forget to explain me which kind of mistake I did Maurizio great achievement! CongratsThank you Simone but it too soon to declare victory +Patch coming this afternoon +patch looks goodThe failed tests does not seem related +for patch +John do we need this in? If so it needs to happen soon +dll files especiallyif one package can contain many of them and modules alreadyprovides a mechanism for using the transparently from userspoint of view +Curious what the impact of this would have been? Has streaming effectively been broken since since every cluster has wrapping ranges? +Thanks Luke for the review +Updated patch with reentrant locking and exception rethrowI also added catching exceptions in the register method but not sure of this is an error +Fixed in afaadceddfdadf +Sorry missed a in . +Attachment has been added with description FO to reproduce has been added with description FO reproduce missing Guillaumeare you sure you tested with the latest version of the Trunk? I didnt manage toreproduce the problem +the role of the expression being evaluated +Looking into thank you +Please Matt +can you try w this patch? it adds additional asserts to narrow down where the error here is the Assertion Error the null is on metadata +Thank you +Jira will get updated with the right version number for the release version of. +Here is patch for this issuein this patch I have replaced anchor tag with submit type input RohitYour patch is in trunk at r +But I dont have a good answer to thatI will post comments to the new issue about the new interface we talked about +The issue there is that our config obviously would not be in affect unless you set the property to tell it to use the normal sun classes +fixed the revision +Im glad were on the same page +have just committed it. +But my comment from Nov contains everything related to +This isnt really meant to enable gathering of status information +Uses a retry strategy +And if the Xmx too bigand using the shuffle in memory logic in reduce task right now meanwhile there are several reduce task running on the tasktracker that will makes the situation of tasktracker get lost because of memory of the slave machine overused. +config files checksum in the scanner but have hardcoded how the scanner is retrieved +See If we have to remove the nullification in serialization to avoid NPE we should nullify the split either after the task completion or after the first update from the map taskThoughts +srcmainjavaorgapachemavenartifacttest be changed to rely on the default from the settings builder +Sorry I was between releases and should have deleted this build as I resolved it at the time +Why +You can certainly extend but in any case there might be the possibility that signatures change in renderersregardsMartin +Do we have a single sbt command that can compile for allJunthe other two null comparisons are for entries not offsets so should be valid there +For DIST it will only show local contents +Seems to me we should be able to do a GB or so of data in a row so long as were picking and choosing columns rather than asking for the whole thing! +We know now that weve fixed the bug in this jira so lets proceed with thatTo the committer just to clarify what needs to be committed please commit to trunk and to +. +Ted your patch looks good as far as configurability + +verified at revision +Ive linked as its related +Thanks Cedric! Ive applied the patch with minor changes +Thanks. + for the patch +My only guess is somehow perhaps some axis flipping legacy importer uses when it reads the lat lon bounding box +OlegQuite right on point one Ive now changed passing the host in to passing the connection inAlso right on point two Im just finishing off cleaning up the class you sent but have to go to work so will finish it off and post an updated patch tonight +Ill commit this minor improvement tonight EST +It only filters characters that pose a security risk +To address your other point Derby never automatically creates a Credentials DB +Thanks for the review Chris! Attached is the updated patch +Partial state saving was not being done by default as is necessary for apps using JSF +I see the branch now +The patch attached has been created relative to the contribution directory +Should be resolved +Therefore nothing needs to be explicitly coded for this anymore. +Self contained test case reproducing as per Joachims comment I can also verify that having implement Cloneable solves the problem +The seems like the most likely culprit +Im going to create a new issue for the changes in. +Patch for +I will check into this +The checking would be done after pressing submit but before its actually submitted thereby Jetty and the errorIm a firm believe that the user shouldnt have to google an error message +Your call +Jordans new patch looks good to me +Replace by +If found these will be used otherwise default templates within grailsHome will be patch also contains a convienient command line script for copying the templates to the current files scripts command line script for installing the templates testgroovyorgcodehausgroovygrailscli test class for install templates script srcgrailstemplatesscaffolding externalized scaffolding template for controller srcgrailstemplatesscaffolding externalized scaffolding template for create view srcgrailstemplatesscaffolding externalized scaffolding template for edit view srcgrailstemplatesscaffolding externalized scaffolding template for list view srcgrailstemplatesscaffolding externalized scaffolding template for show view Changed files scripts checks for presence of templates within application when creating artifacts srcscaffoldingorgcodehausgroovygrailsscaffolding scaffolding templates are now externalized checks for presence of templates within application when generating scaffoldingSee for detailed changes the attached everything is OK let met know then I can write some documentation about this on the Wiki to finialize it +I just tried the test case with the latest code and its passing without any problems +Is that enoughI think it is +Trivial patch as discussed +If the issue still exists with the current AS CR can you create a JIRA there since it occurs when using the web component of AS? Thanks. +I have rebased to the recent trunk version +Either way a single task failure should never cause a job to fail right +Will merge into codeline shortly +Fixed on which branches? Shouldnt it be fixed on X trunk but also RC +If it turns out that there are major browsers lacking this collection it can be modified Ive verified consistent behavior on firefox and safari +Adding a requirement that you call some method to determine if the claim of implements is actually true or not is missing the point of the type see and I agree with you generally speaking +New issue created for the Velocity and other dependencies fix in. +Because thebyte array length for a segment is always the same they can bepooled +I did create my own test in themeantime and spent some time on it yesterday +We felt that this was cleanerAnyway thanks for the time and effort you spend on webwork just donating this back on the chance that it might improve webwork +The work going on to separate messages from vertices will help quite a bit with regard to thisIn the spirit of this thread I could definitely refactor to be more friendly towards subclassing like this patch shows +Thanks Jakub this will help a lot +Thanks for the review SushDilip Namit et al would you guys please comment? Itd be great if a committer could review too +Helmut is this still a problem for you? If so feel free to reopen this issue. +Thanks Renaud Uwe! +I dont see that failures in earlier test runs that include this patch +I am wondering if you have found out what the issue is yet? Thanks for your help +I thought that this was where the socket connection was being made +Moving to Messaging +You are right lets get it going! I will commit this soonsimon +Note that in any of my suggestions for adding header footer prefix andsuffix attributes to either of these tags it should be clear that these areobviously OPTIONAL attributes +The default out of the box behavior is to use producer flow control and stallblock the producer +But the failure details differ from run to run so I have not made it an official test yet +As an example heres some runs I did prior to the optimizations INFO startSuperstep Superstep after prepare totalMem maxMem freeMem INFO finishSuperstep before flush Superstep totalMem maxMem freeMem INFO flush starting for superstep totalMem maxMem freeMem INFO flush ended for superstep totalMem maxMem freeMem INFO finishSuperstep Superstep totalMem maxMem freeMem INFO prepareSuperstep totalMem maxMem freeMem INFO prepareSuperstep Superstep after inMessage assignmnt totalMem maxMem freeMem Note how the free memory would dip to MB at times +committed to trunk +Thanks for the prompt response +Oh Ram and Me commenting at the same time with the same question +Ive setup another cluster with agent and acks seems to be working perfectly fine with multi masters +Should this method be removed from the protocol +Updating patch +This problem is not resolved yet + +Any suggestions are difficulty here is that not supporting classes such as in JDK labels Groovy as a legacy product +Thanks Pavel + addresses issues for this jira in branch +Is this the only way that I could get an editable version of a relationship? What does we loose if we defer it? Extra to grab an editable entity +Hi Army I am working on updating the dev guide now and will include these comments for the ref manualI should have a patch out soon for you to review +Although the code is now broken down into logically organized files there are rather extreme levels of indentation that makes the code hard to read +This issue has been fixed and released as part of release +But this is in my list targeted on this weekend +Is this bug now closed re M? Does M compilerun on SolarisManuel can you elucidate +No tests etc +Hive users should be able to specify in their queries just as INNEROUTER joins +Attached log file +If there is problem with time and scheduling then move it into another release but dont just say WONT FIX because it is a special +Thanks Santhosh for contributing! +Look Ok thanks. +The solution is to apply the same technique of creating an overridable scope factory method that was used in + +The patch is to applied on the iPOJO core projectIt allow external handlers to participate to the architecture +Patch created to remove depends on and were applied I was able to build Rev on with Sun JDK +I have added screenshot just to show that Target is connected +Very professional No links to the duplicated issue ID no reference to the FAQ URL +I need to know where the ZIP can be downloaded from and what it is called the full name and location of the HQ switching script mentioned above the exact command required to run the HQ switching script on all supported platformsCXF where the CXF ZIP can be downloaded from and what it is called the location from which ant should be run install instructions for ant for all supported platformsNatives ZIP how to install the natives ZIP for all supported platformsI cant get this book done until I have this information and Rajesh cant QE until the book is done so speedy responses are appreciated +Commit changes to branch at rev and trunk r +Its all yours +If this content cant be provided by the end of week th December then we should move it to to need some idea of what is happening with this JIRA so I can schedule docs work for volume of material +Merged now +If no one comes up with a better name a new patch would be nice +It was released with the fix in +At a minimum I like the refactoring done here +I was hoping to get an RC out for this weekend and was planning to add only any blocker bug fixes to the branch until then +the way to go is to implement this feature for the alsoWill do this in the next week +We should fix this before next problem is with the unix generation of the docsOn Windows the comes out but on Unix it doesntI tried modifying the text to not use code but that didnt help. +I added that code so I suppose I should take this +An empty map may or may not mean that the row doesnt exist +I think the trigger is that when we have a midkey for a half file which does not correspond to the block boundary both the top and bottom regions will try to read the whole block +Pushing this to trunk and based on +Thanks for pointing this out +CraigAre you sure that a mention in NOTICE is necessaryWould it not be enough to mention the files in LICENSE? Or even READMENOTICE files are for required notices only or copyright moved from source filesIm not sure that either applies hereThe NOTICE file must not contain anything that is not strictly necessary +But to do this you need to present valid credentials +I got that piece of logic from the example in the Hibernate in Action book Chapter page Thanks again for your time. +I woudlnt mind having a which gives the ability to resolve entities using all sites +Fix and test committed +Dan is producing beta this afternoon that will resolve this +The latest Sonar plugin for Hudson doesnt appear to have any way to set an auxiliary classpath for +Thanks for looking into thisDavid + in needs to include from the client lib directory +bulk defer of open issues to + eliminates the need to maintain a separate patch for basic authentication since this plugin takes care of NTLM as well as basic authentication schemes +Thanks Owen! +app server thread pools? Should threads spun up for s internal mode also get TC classloaders setClocks ticking on +OK finally got this one ready +Sorry I have seen know you have desribed under how repository should look like +Youre right I had thought it was already done +this patch fixes the build failureSumanCould you pls review this patchtnx +suggested which fixes the issue has been corrected +Committed revision +I havent looked at all of them yet +if a client has written some data to a set of replicas for that block and then all the replicas go bad then the client gets an IO error and stops writing any more data to that filewhat is ur proposed fix? can you pl explain thanks +Hi MariappanIve created a patch which just unconditionally creates a because I confirmed that instantiation of is not expensive and implementation cost is also lowIll make this tickets status Patch subimitted after finishing subtasks +They are there for backwards compatibility reasonsJason and I agree though that there is no need to support these legacy configs in the console +Sorry Ill fix it +Changed section name to be parsers +Steve I think this is a mistake +Patch applied at +Sorry that was pretty unclear on my part +Linking a pull request which is probably not intended to be mergedPlease see my inline comments in that PR +I will close this bug +The way Stefan fixed seems to be perfectly fine to meIt is widely known that thread safety does not come for free +Problem fixed +Thanks RayPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +Ill duplicate this in bugzilla and assign to Charles +CPRR is only part of +I tested and applied this patch in r +Feel free to comment especially about constants in removed file +Sorry Im having some challenges with that +Ill try to to this at package phase independently from indeed this is not a bug but a feature suggestion +Started a little streamlining +So whatever changes you want to introduce it must be able to run the unit test without problems +This bug has been fixed in the latest trunk +If Lite is enabled we could reopen it on every url change +an alternative approach This is still v once you guys agree on this i will make a fresh versionLedger is mainly a look up map to invalidate the filters when there is new changesThings to do Make ledger an object and attach it to the cache instead of extending ASC More fine grain locks on the t the s ledger also need to have entries removed when entries are evicted from the primary cacheWould it be reasonable to have a tunning knob so only rows with more than n columns are filter cached to avoid paying the penalty of storing the row key twice +Please test and review havent had a chance to test this but it looks good to me +MartinAh that is a good idea +Thanks for creating the new email listsThe archives for dev and commits look a bit dodgy +So it is not possible to resolve forward references for the parser +is a medium testMeaning lock manager tests were not run +As long as the marker and quick fix clearly states that its changing Project A and not Project B then im fineagreeing with that +AS has moved to JBC +The only thing I regret is the need of running open office as a service +Guess there arent too many people who have their host names with a mix of upper and lower case. +Failed to import LDIF into backing SimonI think that the attached ldif file is not the one which cause the pbI cant found any extensibleObject in itOtherwise do you have modified the schema or added your own schema? Of so could you also attach the modification or the file +David why do we have in one of the builds? seems to be the norm in the others? If this was not some kind of typo could you quick fix it +Doh! I had the tab open so didnt notice the existing comments +Looks good to me +I just committed this to TRUNK. +Converted to resolve to online by default +RM Side of this is fixed passing it back to Lou +Committed rev on trunk which replaces the cargo specific logj file with standard system property for logging threshold +This difference in this version from the previous is the formatting +Andy sounds ok to for release ARQ +Retargeting this to the future +Its a bit different from an API change as it is the data that would need to be fixed rather than code +This patch solves the bug in UNION ALL two SCRIPT operators but only show results from one operator +Why is the diff so weirdCan you include my changes into the patch +Maybe I can think of a better solution +Hi Dmitriy sorry for going a little off topic here but could you elaborate on this? Ive been experimenting with using either cosine or tanimoto distance on the output of ssvd with pca true +It was probably mistakenly copied from the grant statement pageOther than that the changes look very good to me +This patch attempts tominimize changes but because it moves responsibility for setting upsome variables into callers it will need good testing +just found a good bug in this while doing some fault testing +I think the has to save the step executions before handing them back to the +It does not use a webapp +I put that dodgy local variable in to fix your issue back in May +This is considered a item for Java EE who subclasses Embeddable will run into the issue as GF ships with Weaving turned on by default for looks like a bug in Eclipselink not Weld. +I created a banner based on the project banner +Right the script is ll fix that so it posix compliant and itll be either part of new build or rm optredhatewssharetpmcatlogs try actually now deleting again it is as opened since I dont have an official release to have it verified the echo n works using usr on my there is some problem with CATALINAPID file and its probably your varrun odd + reproduces the issue +Seems like were generating a lot of code that looks very similar +includes a launch config but not the Tapestry jars or its dependencies +Thanks for fixing the error ScottThis changes the specific compiler output which causes to fail +Thanks for reporting this +Not part of the spec but Apache Harmonys javadoc says Answers the major version number of the database software +bin packages now have in conf directory and scripts are configured to use ituse mvn package assemblysingle to generate packages +This issue havent fixed yet +Switching to tests +Attachment has been added with description Reproducing environment and error P open bugs to P pending further review +this patch includes the patch and the test for trunk +One more comment this should be in the incompatible section in +It would be super duper nice to have it and passing the FFI specs so let me know if theres anything I can help to move it along +Have a look at the interface and see if you are happy with the idea and I will sort things out +Are you asking if the plugin can include a test for this behavior once the issue is fixed? +Oops sorry Aaron +Rampart is not engaged to the service +AFAIU not +I have added that to known issues + +Would we have to setup individual subnetworks so they can all run jboss cluster wkith the same partition name? +Eddie can you please take a look at the patch and check it in for me? Thanks +Real time implementation of the above mentioned is crucial +As for those task logs they unfortunately have already been deletedDeveraj Dont have the job id anymore so wasnt able to run that command +Ah I forgot to mention that I ran client and server in the same machine +So if you add foobar to the sa tag it will be passed up to the output +Vicky Im assigning this to you +NET and Java can parse a datetime seem minor but are just as important in my opinionThe problem is that +I just found a workaround for this bug +The code has been committed to master +is a new topic thats titled Overriding the default optimizer behaviour +Was just a suggestionbq +bq + +Thanks for the info +? I was confused at first as to the nature of this bug then I realized its just that is putting a config in the wrong place +Seems to be duplicate of +This is good news any use in Tika +Sorry this is too late to make +Postponed to +does this problem still exist +Thanks Ive updated the docs +I dont believe is relevant as the property being unset in that change needs to be set by the job submitter and thats not happening in this caseI believe the keys are dropped by tasks because the constructors take a collection of tokens instead of credentials +Personally Im not convinced this really is a bug at all +Implemented to be included in release +Implemented the requested featureI started from patch from Mike Patnode but used the new maxTotal pool parameterto limit the. +Patch attachedI think sometime in the future we should just register a help object with help messages for each command and shell should automatically list appropriate help message in appropriate places +please review update to the patch to format the lines to be with in columns +Please dont commit this patch +This was very helpful to us in tracking down issues with particular orders or visitors on the site so were contributing it back +I was running F before. +Bulk defer all frontend issues to +I believe accounts need a valid email address so either way we would need some help Ill create an INFRA taskJIRA saves descriptions as HTML hopefully some simple replacements will be enough to convert them to plain text +I dont see an issue with taking the path of recovery rather than bomb out +committed thanks! +Attaching patch for trunk +I tested both with and without security enabled and both scenarios worked as expected +e +Attaching derby knight with sans serif font separately +So until we clarify that +The second initialization was apparently broken again by r it works for me with r +The plugin would need to know about the A notation anyway so why not make it explicitSean Patrick why does the plugin need to know the A notation? Isnt it just an argument like verbosetrue but with Averbosetrue thats sent to the compiler or what am I missingIs this the correct forum for these kinds of questions? Or should this be discussed in another venue? newbie codehaus third patch seems flexible and interesting +Rob please Snjezana If I remember correctly you were in charge of adding profiling to the AS instances +Thank you for providing a test program +In we now have which replace this stuff +Released +It has been tested on SP SP SP SP +still doesnt follow groupId conventions in groupId was intentionally kept to fit to match the package name +In revision I committed a change that should allow such SNS paths without errors. +Supported +Dont have my oracle and sql server vms to test against at the moment +Pig Latin contains information about comments +This is a current issue not a retrospective have a naive fix for this locally that I can share +I think there is something about the way that the imports are specified that is causing this bug but without more information there is not much I can the source of this problem is the same as. +Should we expect a version +changed files for attached the changed files +Hmm this seems to be a confusion of and +I like that idea Im going to work on this more next week when I get into the guts Ill let you know. +The compression could be independent for the feature structures and the sofa +Yes I agree with you +Enhance the role mapping configuration with a default role to be assigned to all authenticated callers if no other roles are assigned +Hi Stefan you are correct +Felix gotcha was confused by the bar launcher thought it had something to do with how this all made far more sense after a few beers +I just committed this. +I keep this ticket open for a decision merge to? As I said before it might be delayed until another user requests it. +Thanks Hitesh! +Why do you copy the jsp files into the build directory? That way they wind up in the jar but only the +I guess this should be reconsidered with the new file cache store implementation +Committed to trunk +Forgot to mention in the comments that Ive attached the logs for both scenarios +I have changed the codes and remove the reference of dom +I am working on vand shall revert back with a new patch and corresponding test runs in afew hours +That was exactly along the lines of my thinking but I needed to get the first step done first get the TSR available +The idea is that edit with selects an editor at start of editing so there is nothing in the editor that can be lostI think it is better to give a user the option to select an editor each time she starts editing +It is likely that object map keys will win out because Long is too short for some ids in TDB in the long term +Lets see what Hudson says +Im working on session storage now for +Make sense +When confirmed this JIRA should be closed +Reopening against EWP +Use it more than onceI think these differences become important in the following scenariosIn I dont think you can cache Matchers instead of bitsets because Matchers dont have features and and in contrib dont work with Matchers because there is no support for Is there something obvious Ive missedCheers forgot to mention that boolean logic on Matchers is already in present inThis is because each Scorer is a Matcher +Thanks a lot for your response Im gonna try this +under +T and should be threated as such +utf files are in fact the source files for the files in which case we should probably keep the +Id like to propose to revert back to Bootstrap because +Legacy index files actually have for the cardinality +Not a fan of supporting another web container out of hte boxbq +Changed file read write to bytes from String +removing fix version +The empty exception gets its stack trace filled in when its instantiated +Stale +Since we are coming up on year I wonder if anyone would like to comment on the release is there a particular time frame you all are shooting forThanks +It seems that some internal provisional stuff is used which may have moved +The issue exposed with the repro has been filed as +We are actively moving it forward to GA although there is not a fixed date as yet +Ill take a look at the poms when my machine comes back +Ugh +Committed revision to trunkVerified this fixes the startup failure on and verified it doesnt break LinuxFrank thanks for the patch. +Test failures are not related to the patch +Sounds like this is just fallout from then you can delete the hints and see if it continues +HelloIve attached an SVN patch based on the trunk as of that implements this feature +top left etc +Marking patch available to get tests to patch is failing to compile +Could you docblog your experience going apache maven release process? This is very handy for any new commiter what would like to sponsor a release +Please remove ovm folder +IgorThis is going to be several weeks away +Patches for this issue should be applied after patches for otherwise it may break other pending JDWP patches +AttachedSeems to work fine +Good luck Robert! +I have the patch ready +That would probably reduce the job jar size too and probably avoid some problemsSo just that one change to core makes it that much smaller? Surely and a few transitive dependencies is pretty OK to bring in to anything +This doesnt preclude us throwing our own unchecked exceptions at all +I am really hesitant we should commit to Hadoop sources from another project +Great catch! Now Fuse IDE will open the file but switch to the XML view so the errors will be highlighted in Red in the XML cut later today. +Id rather do a implementation that knows about all common persistence providers +This issue is not reproducible in ODE branch +Agree on Marcs comment +I just tested this out and it appears to work for me +trunk r x r Thanks Shawn +My local test case always fails +Attaching proposal patch for +Replace this file in specialpurposeposscreensdefaultmenu +Implemented in to make the two above patches work properly and nonintrusively +fixed rev Thanks! +Works great thanks! +bulk close of all resolved issues as part of closing items. + +Closing solution per Radu +and fyi as says that line never did anything hence why we removed it to avoid users from thinking it had any effect +patch available +Updated RIE to return sane values when there are no columns instead of This was a bit dicey given the hackish nature of the serializer +The same for should not poll the file before it has access +Ive gone in and fixed all of the issues to get a clean build but Id like to take a step back and take what I know now and make sure that the design of this change is based on some consistent modelIf youd like me to commit what I have anyhow I can do that but just giving this heads up first +Patch also committed to +scheduling so this gets tracked +Is that still what you are thinking +If you look at subversion this was merged into was included in but not shipped as part of the release. +However If you feel that a few more months should be enough to finalize the API we could probably delay the BETA phase a little and do one more ALPHA meanwhileWhat do you thinkHow about trying to find the middle ground? I made and package private as these classes are more likely to require major changes than any other +Users cant just change their from address from gmail and expect our ezmlm to believe it +Subclass implementers will appreciate it +I have added two additional ORDER BY columns to and +Except of two COPYRIGHT files and runtest scripts +And I also tried the version which also crashes like if this is a feature two questions Is there a way to make sure that what we pass to the XML Writer are only valid strings containing no control Could the error message be more explicit +For instancestart node node is masterstart node node gives up being master node becomes masterstop node node stays masterstart node node gives up being master node becomes masterWhile I am not sure if this is the intended behavior right now I do not expect any issues with that +Ran into this problem tooIt seems pretty serious since the documentation implies that scroll and fetching should work +Could you possibly provide some setup information that may cause this error? I currently have the exact same stack trace and would like to identify how to correct it +Thank you +It was custom security code in Solr shipping code and a major problem with this patch +Hello Would you please try my patchThanksBest regards patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +I have tried running just the upgrade suite as well as the entire junit suite and none of them have reproduced this intermittent failure on my Windows XP +Will include it for in trunk +Nice catch guys +The stack indicates readExternal is looking for collation information but in a soft upgraded table there should not be any +Setting to minor since now we know which are the required props we can update our metadata +Current default font size is possible to use font with the size it is readable on the paper and reduce number of print to reduce print diagram size is necessary to reduce icons size other idea is create toggle button to switch onoff icons +Once Java is a minimum requirement this is basically a good idea +Committed this to trunk andThanks for the review SidThanks to Roman Jarek and Mark for filing this and help testing it +As much as Id love to have two spaces everywhere Im worried about the adverse impact on svn annotateThat is indeed the biggest cost that I can see +Committed trunk revision +I will try to do it next weekend and since you want it for version I also need to see if there is something different in this part from the version CheersNuno +it would be great if someone could contribute itIm of course willing to provide needed pointers and help integrating it +Weve actually moved away from having an environment variable +Thanks +Ill have a look at it in a little while and see if we cant get the ball moving +Feel free to commit before Stack cut the release since I cant +You do an optimize and the old files will go old files did go away after an optimize + uses an unbuffered stream +That worked +Turns out this patch breaks the new specs submitted in +ClaudioThank for a nice suggestion +This patch tightens up the encapsulation of the array wrapped by +srccontribhbasesrcjavaorgapachehadoophbase Update hbase getting started so it now refers to new scripts +That the implementation you propose is not as efficient as the current solution +Resolved by +seam is not a valid view id +closeRegion get get etc +The class column is added by Grails in the inheritance scheme +Fixed +committed this to +If you need that by specific data please tell otherwise it will be prepared at end of this week +Unsetting Fix Version for unassigned issues +Which one +Closing as a duplicate of. +Commit it soon +Unfortunately Im really busy lately so mimej has a really low priority for me now +Should support using the old method as well so older management console will still work +This is probably actually a mac bug trying to trace through the maze of symlinks is a JAVAHOME to the full SystemLibraryFrameworksVersionsHome instead of LibraryJavaHome and using the mac symlinks is a think the workaround is good enough here especially with most Mac devs on Snow Leopard now only having! +But thats misleading +Disabled formattingSet Fix Versions to +Hah thats kind of obvious post morning caffeine intakeIts possible theyve configured the AWS keys in the client also +bulk move of all unresolved issues from to +sql files are waiting until after +When I used a servlet implementation of the same logic the thread count rises and falls as threads are GCd. +also it is comment to start a commit message with the bug +But exception must be thrown before method invocation +Can you add a negative test also? The partition does not existCan you file a jira for the same functionality for a tableAlso can you confirm that no table internally is using UPDATE as a column name +Vote on the source release staging maven repo artifacts +Issue is resolved +Im staring at this screen and I dont see how to attach the zip file with the minimal example +Fixed initialization ran integration tests pushed to master +Patch to fix the two and passes tests locally +Since this is probably not the last refactoring of serialization I think it would be more convenient if the serialization methods would not be part of the CAS interface but rather to have separate utility methodsclasses for serialization such as there already are etc +Yongqiang thanks for the pointer +The second chunk is different than what is actually in +fixed in efffacfafffedebd on zone seem toh ave worked in automated env +I created threeattachment the source xml the xsl file and the java file that runs the test +We already have a nice arbitrary definition +I will close this issue again +I can verify it now passes the tests +I wish I could reproduce this more easily +Alexander Ive reworked your test and added it as regression test forCould you review and comment it? See modulessrctestimpljavaorgapacheharmonyxnettestsproviderjsseThanksStepan +Those two test cases need about s each to execute and scaling them down does not make sense +Perhaps I should clarify my whole use case here +Thanks for review anyways +Im for this changeIIRC building service model is one time thing when create endpoint you wont see it per requestMoreover that log told you building service model from WSDL or from class this is an important INFO IMO and I believe lots of users already get used to this INFOBTW I do think this kind of issue should be discussed on mailing list first. +Committed and resolved +The script should check for URL to handle this case +Please do not open issues against +I have closed this issue as the information requested is already documented appropriately +Could you please set eol style to native +Looking forward to thoughts +Thanks for the patch Li and to all others who hacked on this patch and reviewed it +Can you please double check the sdeexport contained the desired attribute +still failing as in Sijies patch +If there is no redirect there is no persisted + +Assign it back as necessary whether to approve the patch and have me commit it or to have me make additional +Now that work is well underway with AS all previous community releases are +How do you handle compatibilitymigration issue +It duplicates the underlying array on each write +Scheduling for but I think this may already be fixed +The more granular version allows a user to implement pairwise deletion +Build had no failures with all tests executed +Hi Tiho no idea +Can you add some examples to that exercises these corner cases +Thanks +Thanks Chris for the commit and indeed big thanks Arpit for jumping thru all the hoops to get issue fixed +This variable will be set by if that task finds a valid jdk installation on the machineB If is not set then set it to be along with the JDBC stubs as was done previously +The existing message logging for each individual connection close which happen because of engine shutdown will still be logged but they will now print the name of the database being shutdown +test failure not related to this patch +This is already done +I assume were looking at powerpoint keynote and openoffice +If u press Finish right after entering projects name package will have name with dummy +Its a nit though whether this is changed or not +Ive just done what trunk does todayI see thanks for the pointer +Thank you very much +Abhijit Meshram +I have the same issue +This seems to be a significant hole in the REST support in Spring and over years without any plan to resolve +im a bit concerned that i may not be testing searching quite properly right now though since i think most of my indexes being tested only have one segmentYup went through this in Solr +Added new types weakref softref and phantomref +Closing issue which has been resolved for some time + Username malkimi +Was verified by Ilya. +The directory modtime check looks good to me +This comparator will then be used to order items in the backlog +Patches applied and tested +Shouldnt be instead of generating an xdoc we generate an HTML report? Currently it generates html report about the goal but not about the parameter +Closing out the issus for the release. +Nope it is still bad first submit is outside of the conversation because it claims that conversation is over but I still have conversation Please use the forum for support JIRA is for tracking issues with Seam Please do not set the fix version +Changing to major because I think that getting rid of the trailing would be very very nice +Among the one that do I doubt if there is going to be a noticeable improvement in runtime +Tinderbox runs show no errors running after this change +If there is a better solution I have no problem with reverting this change +I was thinking about moving the into samples but when I saw the which had the same interface I already needed I changed to use it +mrepositoryjbossjavassist +I believe is the best default option available +Sorry for that +Robert Mathieu +If you get a chance could you open a new issue w a patch that use use to print messages +Yes please attach or describe in more details how the Java model looks like and how it is bound +fixed syntax error +If reads are enabled and cluster has good HDFS locality and data is compacted you will get dick blocks for free +But these tests are critically useful regardless +We do that so we can recognize the type based on a naming convention +If not then the examples need to be updated +Verified on Hudson stable branch. +Im going to try to get this in since its so well tested +Splits bulk import and table copying all create references to files which remain shared +Thanks for the reviews Hitesh and MahadevCommitted to trunk and. +just FYI +Sunny please add some description and a meaningful title +Thx for all the hard work on this release. +Ok I ran our site against the new code and everything looks good +ISTM that a simple approach here would be to define a maximum amount of time to hint for since tracking a nodes downtime is easy and then you dont have to go poke a JMX on nodes to disable HH for a dead host +Attachment has been added with description transition to closed for remaining resolved bugs +I didnt check if there were other stuff like thisThanks Nicolas +Great suggestion! I would also like to see the capability to specify the operation in the URIhazelcastqueuefooput +Applied to branch in SVN rev +Grant are you going to backport to +Hi Aidan Fixed the issue with the test so this is ready for review now +I set the default external encoding unconditionally to and clearly this gets picked up by some other means +Simplicity tends to help portability +See head revision +this command leads to mvn output shown in attacmentIf I rename project grop to Hopsurf build became also mvn versionusrjavadefaultMaven version version name linux version arch i Family unixIf this command running on Fedora build successful even project group name is The backend +pushed in Developer StudioVersion Build id Build date +They can just put them in +this patch is dependent on the +Will try to patch this for +There seems to be a problem generating the corbabinding from a wrapped wsdl +Thx that explains it +Candidate fix also addresses. +It would also be nice to change it so that when an invalid transition happens in the node that more then just an error is logged but I am not really sure what else it should do +I think Ill try to start implementing some bootstrap code it will make it simple for new users to get a cluster up and running in addition to making further development easier +Found some reasons +The values should be between and thanks for the info +The bug is fixed +The tests should use otherwise looks great +Ok I see what happened +I have found out that the property can be set to unknown +ran without failures +Comments +After some investigation it looks like the actual exception is logged but after its logged some number of times the summary is logged instead +The +Ok +for previous version not for commit +r for doing the same thing to the rss connector +Test passed on my cluster +For now I changed Configuration to only deal with Strings since that is all it is used for anyways +It should also support only headers attachmentAnd more now that we created util for ENHANCEDSTATUSCODE support we could remove the innerclass and use the common one +released issues. +I used instead PREVIOUSPARAMCommitted in revision . +Tested times without any failureThe advantage of for me is that it eliminates a quite tricky behavior but the fix would then be outside +we should be able to revert the workoptimization done and avoid this bug +Ok if I commit this +Can you please take a look + +is failing consistently now after I the tests to check the failure of each bookie +Thanks I will try it out +Great work Pierre this looks good to me +This bug should affect you guys as well you havent been seeing occasionally missing tables when you issue metastore queries +Lets keep it although its not really maintainedI had an issue with the HTTP Client upgraded and wanted to get rid of atom as it caused problem +How about we just add a name attribute to and a getter and setter thereof? It would automatically be serialized with the vector state and can be further integrated with the Matrix stuff next +Fix committed to trunk rRequest review and merge +But we can set the cipher suits and protocols manually and it is working +The is attaching to the s deployment unit and as a result the jar is then deployed by the which results in a duplicate service exception as it attempts to install the same security realm +Is this ever fixed in any releases? I m having same issueThanks +Please verify and marked CLOSED +I can send the jenkins projects if you need David Ive started to launch several times sonarcontained sonarcontainer and sonarcontained but I dont manage to reproduce this issue +The previous patch was generated in error +Its with the version bundled with EPP so. +I was not sure whether TDB should be supported just in the webservice module or it should be supported within a more core part of the code +Sorry I often confuse the new Close and Cancel buttonsI reopen because it needs to be at r at r. +Shared project poms were also supposed to change but they didnt +RolandGenerally I feel this is very similar to the mutable immutable HTTP message interface story we had in the early days of +Retry to see if failures are because of this +i will wait for the patches +Do you still have it +At least linking to Vimeo and not directly to a specific vendor would be a good sign +Seam text requires either full URI with protocol definition or link starting with +I have noted some issues with PPNL before with regard to abstraction namely Pig provides the to the listeners which is not relevant in a differen execution engine +pushed upstream +I would appreciate it if someone could delete monitoring +Switch over to the RI to have this problemMartin +More information since I think that there is still a problem related to this in the product +This problem can now be avoided with the option to rename the timers table andor specify a particular schema name see. +So if HDFS does s it has the right state when starts the recovery +Closing all ResolvedFixed issues lacking a Fixed Version +Done. +it also has about MB of text files that dont appear to be in use at all +Was not able to create the fancy table using HTML and CSS created a default table style +This works fine in most cases but not for the Palette +Please open a new one for if this still INVALID FIX to limitations in JIRA the resolution on some issues had to be changed as part of removing the fix release from the issue +Yeah its without of course +I think Jonathan is referring to the server side impl as the application +please note has been added that you need to create a file that already includes the default settings +If you are watching this issue please feel free to participate in the discussion regarding the proposed Deliverables and the corresponding Pseudocode Examples listed in this issues! issue depends on in order to support common context hierarchies in Spring MVC web applications. +OK I see you are right +If no one objects Ill apply this patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Hi HarshSuppose someone does not understand what does it mean by files and directories blocks total +Hi JukkaThanks for responding to my issue +Would you want the java formatting standards to be applied on every build? Or just a check for compliance? Ive never seen the former done +Verified on Build +I hear a lot of concernts about HDFS from folks and it would be great to gather up the concerns in a single place so we can start tackling them one by one +Do we need a test that checks that the request fails for a cookie that is bigger than the header buffer +new attributes supported on the docs module for. +This bug is not reproducible on M snapshot antui suite has failures +Although I already know that you are working to fix this issueIm working in a because I need this new feature +Good catch StevenI think where you put the tryfinally is good +Please dont use but plexus logger instead +If you guys are happy with the output directory now coming as the first argument then we can commit +I really just needed the stream but thanks for the code. +In the v patch I make a single method that set the flag and then did the call to expire +Just have committed to +Ill look into it as soon as I get a chance. +Verified in JBDS +applied the patch into trunk and branch +This is a simple patch to make the works correctely with an alias when using the createCriteria method +I dont have environment now so I cannot test thisAs you mentioned the should not be included in the patch file so I removed the ldapbuild and then created the patchI saw the new in xml files in trunk +See fixed by code is much better but still assumes two values retrieved from the sequence to be consecutive +The fix is however not supported by jboss because it includes old ironjacamar +for this investigation Santiago. +I have a working patch for this but it requires to fixed to allow the test files I am working with to function correctly +Mike is this still a problem w the refactoring done recently to addIndexes +the first merge would create a single +The overlay system is broken as well there is yet another dir that contains specific filesls toolsconsolesrcmainresourcesoracletoolsconsolesrcmainresourcesoracle toolsconsolesrcmainresourcesoracle in CR +Actually we should create a JPA version +bq +We just pick the xmlsec jars from the Apache releases +Hi EricThanks for will try with nightly build +Committed +I have not been able to build a WPF application at all due to this issue +As long as you have the zip code Geo setup with the ID in the STZIPCO format it will find it and then as long as you have the rollup between the zip code and the county defined it will find it when it expands the Geo records to include all parent records +The only logic for redelivery in that happens on the client side is for Transacted sessions +Sometimes the suggestion box appears with the previous suggestion ve tried various changes to my code to fix these problems +Closing all resolved tickets from or older +Logs from second run with XXM +to generate the patch its easy for us to review and apply the patchThanks for contributionThe diff patch is OKWhen you append the attachment you should see a radio button option like grant apache license for your attachmentBut I just found the radio button is gone now there must be some change I dont know for the jira instance will try to get it back +we just need to make sure that a server that returns doesnt get its vote countedIm adding the problematic scenario to the description +it must be a resin bug +Thanks Tom. +Adding a field on the write side is the same as removing a field on the read side +on +thus unlock the node without interaction with an administrator or the original lock holderas far as i know that was not possible in the original implementation as a different sessionwas not able to become lock holder unless the original holder dropped the token from thesession +Trying again with build +Fixed +Other information the space in the transaction has been fixed too. +since this is an improvement and not a bug and i dont want to delay the release for this tom +check if is available before enabling without removed whitespace be this commited +Checked it back in now so it wont break exisitng functionality. +Closing old issues +We could either update the to not send the event twice or we could update the to handle getting it twice +Im afraid there could be some subtle memory leaks if you forcibly terminate the javascript engine without going through the operationcallback which could be used to gracefully terminate the engine +Just using the maps and zookeeper states available +thanks again +I think it would be OK to move to the base class since support for java is not there anymore +Done +Not from me +Restart all the running after upgrade to the issue should be resolved +Hi Zevthanks for the feedbackCould you attach a log of the mentioned test failureThanks the test should be fixed the sleep was larger than the test condition which could lead to a race condition +Well just need to write some notes documenting the pitfalls of executing commands and how to avoid them using the new methods +Its not a truncate issue since that float itself doesnt matchEither a parser or a rounding issue +Ive reassigned it back to you +Oh brilliant yet another refactoring blew away the trunk patch here +Thanks +Applied to TRUNK +Weve been running this patch and things are working fine +More docs changes needed +Same applies to read by the way +Checked in by James. +Oh no +Thanks Arun for your contributionYour changes are in trunk now at revision no +I think the whole point of dismax is to support a very limited subset of the query syntax right? I think you want to add a new queryparser not modify dismax +I will check with Gavin why he designed it this way +Marking closed. +Corrected logo on titles in ESB guide +Great workI havent actually tested the patch with Mustang and JDBC butderbyall passes on Sun JVM Solaris x +Patch looks good like the javadoc we get lots of users asking what those exceptions mean + +So from a conforming perspective then it should raise the syntax exception which is what I tried to say in my previous commentBut I have to admit that the current behavior is somewhat awkward since the implementation has no way of deducing the resultant data type returned by such an expresson which will force the implementation to default to some default datatype +Committed plus a change toto subversion as revision James thanks for the patch and for the tests and for the good comments +New patch without extends +From the information I could find via google Ive come to the conclusion that is always set to regardless of platform +Ill check it in ER test cycle +Sounds like a plan +Hi Alessio do you know the native test name? I try to find but not found +For those were very much looking forward to how this feature takes shape +It seems as better place for such code as the sink should be just responsible for sending events not to cache themInder what do you think about thatcurrent path implementation was the old thought however after chatting with Arvind and Sharad we had finally decided to implement at channel level something like aAs Sharad stated we can work together on this and get it reviewed by Mike team +generic jms xa pool module package packageexisting impl remains but now extends the generic impl module package +Ive started the vote on Friday so these should be available Monday or Tuesday +Please check that the patch was applied as you expected +There is a system property called you would need to enablePlease use the forums to ask questions in future +It would be possible but perhaps hard to generalise. +The best way is for Charlie to apply it to the branch +Committed to TRUNK. +These blockers and critical issues are resolved but not verified +We discussed this with Rob and Nicholas +Sanjaythe patch looks fine but i want to raise one question which might be irrelevant to this patch should we throw or for any limit cross case +I think it would be good to backport to +Im surprised we never found this before +The reason it escaped me when testing is that in this implementation the cores api delegates to the cpus api if not found +Any thoughts on which suite would be the most appropriate for such case and if there is any existing test which I can extend to include this case instead +The work on this really should be done on +Attached patch that modifies the FS connector to fallback to a copydelete strategy if calls to renameTo fail +Hmm Mike CODECFILEPATTERN is package access only +Exceptserveralldeployserveralldeployserveralldeploy Everything else is ALPHA the problem with the jars should be fixed +Please ignore my last patch +Im good with this +Moving post adaption is corrected with +Is this a problem withOr is it a problem withShould this issue be moved to the plexus projects issue tracker +Its not clear what problem this change is supposed to be addressing +bq +This project is linked to on the resources page Struts projects on +Im still working on the Agent and Hypervisor docs +I will revert those changes +We are technically able to do this for a but not for a or unless they happen to be expanded to the file system +For the log message for I think its better to change the log to Debug level rather than change the guard +As you mention we need to figure out how to get the tests working locally +Oh and I can remember for such an error that can be be solved by upgrading JDK +Fixed the unit test to reset the sent message flags +investigating now +AS per the comments the issue was fixed with commit cfaaaaaaccfaefcd +I reproduced the out of memory on my machine in a clean codeline so do not believe it is caused by thispatch +Unless theres a common usecase where such abstraction is badly needed I suggest we close this one +The bug is fixed. +A description of overall approach will help understand this better +Your script logic should just work when I deploy the new snapshot. +That related to Todds comment in point +patch which fix the issue +Patch applied with thanks fingers crossed the nightly build will kick in soon +Adds the static Configuration in +Does this mean we will be able to merge non consecutive segments +It looks to me like whats happening is connections from the previously application may not get closed until the next garbage collection after the application is undeployed + + +Could somebody reset that? Or better yet leave it alone and upload a patch +Then we keep the APP constant around for months past and deprecate itWriting to the jail seems to be what the users want plus JoeFil are on board as well +We need a more pluggable architecture for the commands context +Fix committed to trunk w accompanying test cases +Patch adding rack map incrementally +no blocking CR afaics +Changed component from Test to JDBC since this appears to be a problem with the product and not with the test +Benoit wroteNote that its hard to compare the memory allocation ios desktop due to really poor optimizations when compiled for iosSo lets focus on speed optimization which is more a concern on mobile than memory rightRegarding the benchmarks I suggest you compare typical tasks take make heavy use of layout constraints such as Form display or anything else that you find appropriatefrom the end user perspectiveeg +This still doesnt work correctly as of Beta +Completely forgot about that discussion and spent time on rediscovering what the problem is +Andun I also prefer deleteI believe element reusing will be not very activeIn current action every element of WB is located on its own level so elements are not connected +It has more to do with Seam EAR structure and deploying it on EAP +The launching scripts are not really any different and just launch the same main +I expect that those printlns can be removed completely one day but since Im still developing the feature being tested it is very useful to have some more info about what the test is doing so thats why I kept them +I observed on the training data that each sentence should be separated by an empty lineI was not aware of this last point +FYI fails with OOM because is not reading the size of the map correctly and is trying to create a huge map +whats the holdup? I believe we can commit it no +Thanks Sergey for initial investigation +In any case not a blocker for having this go in and can be addressed later if the eventual consensus is to change current behavior +This feature is essential +Looking at the UI for one of the jobs that is stuck in this state and a heap dump for that AM I can see that the JOB is in KILLWAIT and so are many of its tasks +in +Maybe part of trunk at some distant time +Perhaps you just tried running test method? If thats the case then the test methods in that testcases are perhaps using some data left around by other test methods in that same fix you provided is correct +We are using findbugs and IDEAs integrated code analysis as documented in the release instructions +So therefore also check your personal mail for the password. +Patch applied under r +The patch looks fine to me +Or I would suggest putting them back unless there is reasoning to removeI agree that I dont think the build environment qualifies it as an incompatible change +Thanks ChrisFor reference the site is set up to give all committers write access so a change like this that has been agreed on the lists can be added straight to the staging site. +I will get back to you if I have any questions +This would mean being able to manipulate which is private. +This is more of an issue than just not being able to set an optional value +PR open a new one and link it to this. +When I went to apply the patch most of it failed to apply +That would make sure the crypto providers that xmlsec provides are using the same classloader and other resources as the rest of the systems +Otherwise itll just get pushed to M +Done! the only change between the v is change from cipherSuites to ciphersuites +Part of a bulk update all resolutions changed to done please review history to original resolution type +Seems like a DN bug to me? Have you tried the latest version of DN +NM cannot work with a RM. +But not I have added them to trunk +Well if it is just a documentation bug even better +This looked neater than a bunch of checks for nullUnless by default maven injects empty lists +verified on YahorIm all set +Judging by infrastructure mails nagoyas Tomcat was giving trouble last week and its likely all Java apps were slow +Ill commit this +What do you mean by that? Just download the latest snapshot and check +is not intended to benchmark typical MR usage pattern +Is there any way to put the implementation in pluginspluto and get to use it +Any thoughtsThere was a long period of inactivity till Oct when Stack picked this upI dont see a second since OctMaybe I missed something +This is actually really hard +This ability to pass system properties provides testing opportunity if one needs to quickly checkreproduce somethingIll provide config file option shortly +I have already overridden the class to set the cache level to CACHENONE and this works with Websphere +added the link to the If you use sendRobust then you do not need to have sleep statement +The shutdown scripts have been separated for domain and standalone mode and works for both as in build +Patch already applied. +This does not allow for optional dependencies +Closing. +Abstract classes one needs to go up the hierarchy in order to find these methodspeter +I think we should too +for the change +I like to think that in the future other sorts of changes might be allowed to surface in that feed +Any suggestion on this +Ill pop my head back up next week when I get a chance to breathe. +No real work but some more necessary cleanup +hm +Sorry for the rant but I had to say it +youd like Avro to be the one and only serialization format that Hadoop supportsNot quite +Sorry my bad for closing this +Closing resolved issues. +It may turn out looking exactly the sameThanks for the info +Confirmed fixed in drop from Hyperic +Attachment has been added with description Few suggestions for improving current code for SVG fixes a stupid merge mistake +Fix applied at apache with new test to validate +i dont think we can do anything else except document this in the wiki andor docs +Tested and closed! +I wonder if we could add list of regions to end of hlog as metadata? There might be other useful metadata wed want to add to an hlog +Cleaner less convoluted refactoring +It may still need some work +Let me know if you are happy for me to commit to trunk and +OK I think Ive got the necessary libs just need to get the tests to revision +To achieve HA each application would need at least Zookeepers running +Ioan +Second commit now merged to the release branch. +I have attached a solution as a be validated +I committed this. +After again updated to latest trunk the problem is resolved +Should i try to test the compiler way or should we go for the approach +Test committed to closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +They all pass now +The writing thread will succeed but the reading thread will fail +Since all subtask have been resolved this task is resolved too +Might be interesting to aggregate a given time periods log events before sendi g the mail out +Dima if you select several items you open several diagrams in several editors I mean if several items selected open all diagrams in one editor +Fixed with revision +This is not an event that has happened in a vacuum +The examples are just that examples of normal wicket pages +Thanks to Dyre for the workflow reminder today. +And one more thing we will probably have to implement breaking down OR query if it gets too long +describe table +even mvn validate fails the extension fixed the problem but fernando probably put it in for some reason so assign this issue to is working fine for me with a clean build and an empty +I ran out of time and had to run to a meeting +I use a bean to drop the database and recreate it before running a seems to work well however is not well documented in the reference documentation +Im fine with including the fixes along with the new tests as part of this JIRA +This is nice +bq +This issue has been fixed and released as part of release +Full and configuration +I see this is an overkill +But this is all I am doing for nowAlso the atsintegration folder seems to consist of useful code for testers but probably not appropriate for a quickstart so I think it needs moving to a more appropriate home +Fixed in nightly builds? +Otherwise it is good +how should I do this? Some initial attempts to alter the and in the applicationstest directory failed +closed +fixed in leaving open until this has gotten merged into +sstringspace +Thanks Rogerio +I committed the latest patch in revision I keep this open because of further improvements with license and maybe ship it with lucene +I could help to modify the after multiple Application are implemented +Len when you get back from vacation can you look at these and provide some feedback? See my last comment for details on what to test +We do not have any deployment code in Tuscany SCA at this point so we dont need to make code extensible We can reopen that issue at a later point ifwhen we start introducing a deployment story but so far weve been able to run off the development artifacts without intermediate deployment steps so Im not sure that well ever need to +But all of that can and should be done in a separate JIRA +Hi VijayPlease use the for such questions especially when its not a bug and when the message is really clearDatabase must be upgraded +Patch welcome +Current chain of patches is as follows to revision from +messaggi tradotti messaggi non tradotti +Agree with Dhruba +In your current implementation the iterator method call must appear last otherwise the settings from the previous method calls would be lost +drl files directly in the project? The drools tooling should work just as fine in there +Closing all resolved tickets from or older +Some changes necessary as part of broader refactoring were introduced in Git commit effbceabdabbafed +Im a volunteer to investigate this issue +This is not yet well tested +The problem is that the integer is assigned a in the +Good point about extents masking each other I bet I just went through and gathered the extents in the wrong order! Ie we should gather from the root parent first and then down to the leaf child replacing any extent information we findI will review your patch as it sounds like it fixes returning null see above for why I suspect I screwed up. +Attaching an alternative patch that handles nulls and returns a boolean if the filedirectory is didnt see your comment think I was uploading my suggested patch at the time Id prefer returning the boolean +It does though behave substantially more +I would not want to fork this entire code just for the sake of being able to reuse these statistics etc +But we are also running z and i cannot reproduce that issue +thanks I think that did it +Ive double checked both and and they do not contain enum. +Ill make that refactor no timelines promisedSo are you saying to use the javascript client side tests pointed at a nodejs httpserver? Is that how other implementations are tested? Would would then be the right way to test the http client if I finish it +no I dont do the actual AS releaseThis was setupassigned to Jason for CR and Beta too part of his final days release setup to catch any blockerurgent issues +It is related and its what prompted the dont know why its just started happening but by Id like to guess it was caused by Update to Maven and the last successful CI run of the quickstarts that change +Search for in the attached file for fixes +This patch copies the output values of INOUT args to the corresponding slots in the input arrayM javatestingorgapachederbyTestingfunctionTeststestsjdbcapiAdds a test case to verify that INOUT values are preserved across invocations for all datatypes +I think it does not matter a lot for this kind of issue +Closing old resolved issues +My bad +Investigate remaining test failures and fix errors +will succeed but will return is a bug when you add more than one query +I am fine with both of these suggestions +We could just count calls to read but there could be some issues +Its still openunresolved +I think your change looks fine +also remember to check if we actually should include or remove the in the EAR +Moving to CR +there is a problem but not w the spec file +I was speaking about the inner would be a tabbed pane with only a single tab Values +Suggestions and fixes are more than welcome +socketmultihomehomesconnecthomeswill cause a server to bind to and +Tested +Drew +There was a missing case that fell through the cracks causing extended or certain other types to generate uncompilable code when used with default values. +Ok Ill try to reproduce it in a simple setup +This would be a VERY nice feature +We only do it in one place now in +IOW it would work for ntfile content as well as simply properties on other nodes +AddedThank +We should use assertNull instead +The intent of was to have server side request metrics to match up with client side metrics +You can try moving the root +If I can gain access to one NM host and its keytab I believe I can secretly launch a malicious NM? currently share a global key container token secrets but there is a jira to move to secrets so sharing a global AM secret would be another step backwardsExploring alternate avenues to avoid global trust is passing the allowed am token allowed to get status and stop the container with the launch request not feasible +My queries are working now +Yes Erik I plan to provide patches that depend on Mockito Ill submit them today +However we should be aware that is the case and hopefully get some idea of the websphere time frame for fixing this so that we can remove the duplicate annotations late +this is the server that contains the bridge and produces the messages +Also the last comment references a SVN repository please attach the patch as a svn diff and resubmit for review +Tests are passing in hudson +After adding a few classes I got the plain ejb tests from the TCK working locally +This is pretty much a for me to use tools or middlegen +I see this when adding a shell project +Changed master file to match current output as Army points out the change is a valid optimizationWill still research why the change occurred but wanted to avoid failing tests for now +I created some kind of trivial anomal client behaviour detection system based on Kohonen maps algorithms as practical part for one of my university worksMain use caseLearning for usual client queries and then yelling when unusual query discovered +Alpha is released and I got a BUILD SUCCESS notification from the Jenkins JBT Forge build so we should be good to go +The fix works for me +So closed. +Your diff does not contain the filename diff +Ive tested the patched OJB version and it works great +Opened to fix Streaming UDF for Hadoop +This patch should enable researchers to experiment with various modes of HDFS block placement without changing code +Ive set up a configurable size for the LOB buffer in +You may be able to increase random read performance a little with caching but I think youd be taking a pretty noticeable hit by aggregating your entire clusters traffic through a single machine +Is this the same behavior you are seeing on the iPad +Also table is created and populated is there hskaEJB is a stripped down EJB project containing Entity class and enum any info on why this fails just for enums? +I have tested and cross check the process the implementation matches with the description and all the entries in acctgTrans and acctg reflected also the reconcileBalance of changes accordingly +At this point T is a successful attempt that has been recorded in success +Build is clean and execution module has all implementation details +Probably we need to do it for the m core too but we have a JVZs comment to use core doesnt have any dependency on Doxia. +Daniel can you please take this to the forums so that we can see what you are trying to doYou are not providing any information for us. +Im good too w moving env variable to HBASEJMXOPTS +seems to differ from trunk quite a bit +Forte Tool actually rolled back instance variables as well as transactions but Ive never seen another language that supports that feature +The difference from running in geronimo is the client app streamer and wsappclient does not work in jboss +A few missing license headers were found but nothing seems serious +Log archiving will also be problematic +Pig commiters could you review the latest when you get a chance +Please commit to the dist repo +Alternatively you can run Maven update project configuration action after importing Maven projects +Can I get a status on this +Please attach all templates css problem seems to be the sdecorateI will attach the also see this issue with a generated to enter a edit page and you will see +SergeIs there anything we could do to get this one moving? If the wholesale content migration from Bugzilla subproject is not technically feasible let us know what other options we may have what do you need from Atlassian to get this moving? An importer that can import only a single component from Bugzilla +I guess this is everything we need here so resolving as Fixed. +I dont care anymore +Erics commit is in master +I asked Paul Ferraro to take a look those issues +Sure Mike chuck it in see that it does more that the +The problem is caused by member please close this issueThanks +java reduces the compile time to near zeroSure this comes not very close to my real project having this compile problem but hopefully it helps to spot the bug +I think this is the same asYes it is. +bulk defer of bugsminor issues from to +pushed both upstream +Changes have appeared in Latest Alpha Manuals. +That is not the purpose of the using statement and is consistent with the behavior of C +Howard can you take a look please +There will be a bit of a delay to post repro because I need to change the case as it doesnt actually throw the exception anymore with the fix for +Alexy I guess this is because we dont know where the web root isYes Max +Taking for wed love any and all help you can provide to support the retroweaved version +Regarding it turns out with a path filter also uses and traverses the complete treeUnfortunately yes +The spacing should be done in tabs to match the existing libraryI do think SSL is something sorely lacking in the C implementation and would like to see this further developed +I have double checked the Quick Tutorial and it is indeed still relevant +xml and + skh targeting to columns were only introduced to represent procedure input columns since procedural relational logic no supports projection of inputs there isnt really a need to continue with the concept of columns +Let me make it a blocker since itd be kinda silly going out w them +Funny we have a patch weve been working on for similar thingsAttached patch only does dropped messages but it also includes a recent variant in JMX which we need for our monitoring +This issue is closed now +Our connection tracking stuff means we also need to be able to register a synchronization up to the point at which we try to complete the transaction. +bq +Yuan Yun yes we should expose and leverage Nutch REST and extend them using. +Talked with Chris offline and he is okay with this patch getting in now +Also please keep the name of the patch same as the last one +I mispoke about the risidual statements not getting cleaned up +Any updates on this ticket? Getting ready to create release m working on it +If the proposed fix breaks backwards compatibility for existing uses such asTomcat and Struts it will never be accepted +Hi JanIm no Microsoft SQL Server on Azure expert however from the error message Cannot open server nsct requested by the login +I hope thats not too far back. +Verified the attached patch through from +Willem can you also align the SUN Mail version at Apache Camel to be same version as SMX eg so they are in sync +Just to add to that side effects are similar to what was observed in when testing the SEQ source +Observers are registered in different session then the one that manipulates the node +Closing released issues. +In the fisrt column end you have a new idintification number remove one and add new or edit +Give it back to you for so the workaround is that the user enable generics priority and will close it unless someone shows up with a really good reason for not keeping it as is since you should allow jdk constructs when you are using ejb +All tests pass ill commit shortly +Hi Robbie please can you review the updated patch +Try also a mvn clean and then rebuild the whole project +Thanks SaschaYour patch is in trunk at r at r. +This needs to be tested a little bit more perhaps. +But all tests continue to passLet me know if this is ok or if it needs more workthanksCarl +Attach the patch and end result conversion for DB in various SQL functions +Without this it is impossibleIn fact I would like to compose types with compound type as a subfield +Patch attached +Fixed in check in +Other minor change the decision on whether to use the stopwords filter or not is extracted to a new method so that it can be overridden +Joney patch looks good but how about a unit test if only to demo how useful your fancy new doohickey is? I can help +So all issues not directly assigned to an AS release are being closed +Henri has done the move and we are already busy using the SVN +bq +Otherwise conversation functionality would not function correctly +I remember you mentioned this before +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Extra requirements needs to unit test itand we ll need for which the Solution and the entity is extended +It would be awesome if somebody could provide a patch for then the upcomming Debian Squezze will get this patch and work together with the version of Ganglia in Debian Squeeze +Please use the new cumulative patch available in +I uploaded the and releases to and added instructions on how to do this to the wiki for future releases. +Well have to come up with a workaround in that project +Updated patch with support for similarity by query +These issues should be discussed separately to this graphics too big issue +There are lot of discussion going around dynamic imports in the web +java files that could be found inside resource folders configured as source folders +Ill move the xml to a testresources and have it in the classpath for tests per you recommendationbq +So its about to create or extend new tag library with such to suggestion is just to generic +There are some issues on how to implement itIm assuming were interested in doing a full chain build OS plugin platform plugin delivery plugin +Sorry it took so long been busy! Please commit +dbmigratein insertrulecategoriesfrom optlocalsharejavatomcatwebappssonarconfig +Verified that sysinfo now produces the correct Java and JDBC version info when running with jdk +Hi RichardWhat do you mean by Avoid data bloat by not simply appending the pathsDmitriy what do you think about my comments above +Committed trunk revision Committed revision Committed revision +If you want more you should open a separate one otherwise we just end up with balooning requests and specs in a single jira that gets never closedOne task per jira if you think my request is out of scope then this issue should be resolved +Lets mark this as a blocker +Removing target release from issues that fit any of the following criteria No known way to fix them Java integration enhancements out of scope for release Other out of scope issues for should go into along with the other support +Alas I wont make GR US Ill be flying the Groovy flag at in Tokyo and Beijing. +However in order to deliver this functionality in a most timely manner this effort will be carried on separately +I dont understand how this can be happeningJustin any idea about what might be going wrong +Yes +lets move this to +Its good to remind ourselves that Chukwa agents might run on a machine without Hadoop +Infrastructure would have to hold the certificate and do the signing +It should remain the users responsibility to throw away the database and try againTouches the following filesM javaengineorgapachederbyimplservicesmonitorRaise a special error if the database directory exists but it is missingM javaengineorgapachederbylocM javasharedorgapachederbysharedcommonreferenceSpecial error message suggesting that Derby may have crashed during database creationA javatestingorgapachederbyTestingfunctionTeststestslangM javatestingorgapachederbyTestingfunctionTeststestslangTest for this error condition +The scripts were built to run in a specific environment to make it easy for a black magic grown system +I think you forgot to attach the is this commitable? Looks so +Ah ok I did not know this but absolutely makes sense +Sure I remember we agreed on working in CTR some time ago +That sounds correct yes! For some reason the deploy dir is not set correctly +Sorry Eric still dont see the change +Rather than trying to shoehorn the cases we can I think that it is best to accept that data types will be and respect the expectations of programmers in each language +If I use in the then the Add Data dialog is moved behind the main udig window +This is not an issue for G but for Maven +Rajeshdisable the tables again if something wrong happened in the middleImprove the comments for this class +It would be nice to test the other prefix that is in the schemes that are in the test viewfs to assert that works +Im going to work on a patch that limits the use of this class to +I moved it to common since it involves IPC server changes +I think we should make it a standard procedure to run newly added tests on all of the agree +Will upload a patch that does this +NET CLR and the results are slightly different Passed Failed Errors There are two tests that fail on Mono but pass on +This appears to be a Hibernate issue as I have tested a simple script as well as a schema with composite primary keys using the JPA reference implementation and it works OK +No thats not itWhen you delete two wikipages one after the other thats when the name of the first one you deleted still remains in the breadcrumbs +Fix ported from branch to trunkSending javaengineorgapachederbyimplsqlcompileSending javaengineorgapachederbyimplsqlcompileSending javaengineorgapachederbyimplsqlcompileSending javaengineorgapachederbyimplsqlcompileSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +PR Merged. + +Closing issues +This should be fineThe change in setup fixes the issue of trying to delete workDir which is the current working dir +Possiblyi just applied the existing strategy that hasbut this could be changed +I will add something regardless. +Patch committed +Linking just to keep connection between them discussion is going under + +This removes the memory leaks which can prevail in case of failures like subscription failure +Whenthe user enter hours in the timeSheet can also add comments with the help of text areawhich is easier to handleThe user will be enable to enter comments at the time when heshe enter their hours in that text area +just checking if this is a showstopper to the release +This is simplified because the rest of the schema isnt reloaded +Its a pathological implementation wouldnt it be better to provide a more useful example rather than one thats explicitly not meant to be usedUntil we start jaring up things separately currently the Hadoop unit test is broken when the is not found +committed to branch and on Tools +That seems smartAgain Im not working on this actively but want to make sure it is easy for someone to pick up + +Are you actually starting tomcat up and passing the correct file encoding system property? For example was able to reproduce the problem by creating the war file with grails war +Yes please close this on request of the reporter. +See for details but basically since handles xlink and associations much nicer than the current hacks i think it makes sense to remove it and just +should the in folder support putting our own message bundle +See comment from previous message +I was wondering this on my comment from Dec as well +Thats essentially what Gary was getting at above I dont think its really worth it +The attached file is a patch +empty queue +You can use or failover test from issue +Thanks +Sorry in the middle of the last paragraph it should be but there was no ONLY so +We plan to tackle this issue first and would just like to with you beforehand to see if this is ok +Im planning to commit this tomorrow +The subdirectory should not have been in the original requestThe modules are located directly under svnwebservicesand the projects have their own mailing lists under. +fixed by checking for to the branch in revision +its really important to resolve this! Can you please guide me to where your answer is? for the responseThe can be found in two location in my webservice here Cwebappsaxisservicessrcand here Cwebappsaxisservicesbinthe same applies for the rest of the webservices I use +Thanks +Please also check that you dont have multiple rememberMe cookies otherwise it may imply another issue with Shiro +However I would like to leave the patch here should someone need it +However the Configuration Handler triggers a Config Changed Event using ASYNC delivery +Why is this an issue? The static method is called only once with inittrue by the +The regression test passes for me without applying the fix +I like the fail fast approach +And Ive found that also doesnt check method return type +can we fix this in with a patch +Wonder what kind of license it should be sincethat file also bundle other jar files include a GNU license +New patch which does things Adds a require thrift line to the and files to be consistent with which was generating the thrift require already +Closed as duplicate of +Patch that deprecates the contents ofIve gone down this road instead of straight out nuking since theyre in core +Submit it to and report back here please +thats fine +Could you try two thingsa remove DERBYHOMEjdbcclasses and try again +uses Java the tag stuff hasnt been introduced until Java +attached the wrong file +Seems like JIRA upgrade made this issue unusable for reopening +Did you try it +Is there a requirement to patch this on hadoop as well +all issues to the CR +To elaborate +I expect that the changes to v and v wont be difficult either +I think a cleaner solution is possible +Would love it if you can post a patch for that as well +I can also verify the trick with disabledtrue +Thanks for the bug report +Would be nice if the gwc gui could be somehow integrated +If the problem is that its putting unrelated links into the page that take up space then Id suggest asking Suggestimate to implement a dropdown arrow so that they are collapsed by default +EdAre you opposed to it or just noting that the community didnt accept it before? If you are opposed Im interested to hear why +the compile goal has a default phase so does not reproduce the error +Nikola changed exception handling in branch alsoNow s stacktrace wont be logged +Please reopen if that doesnt resolve your issue +META +and Im marking this resolved +licenses are in the liblicenses and licenses folders accordingly. +Removed support for running HSQL in mode +Fixed by adding an new method in Folder API +Before After added +out some actions were not defined because the wrong version of was not using the configuration provided be working fine now +Simple patch for Lucene modules only since they already support +Had to fix it again due to smooks upgrade from to revision . +Manually tested! +Not +First step preliminary cleanup of the codebase so I can see whats going on +ASF license +Marking this as invalid due to lack of response +Btw reflection is likely to break in AS environments where is not used. +This is only a javadoc change for authenticateUser method +Ive included CN related files under +Ilya thanks for the simple work around +Since the factory bean supports any being used with keyName this seems kind of overkill +So this is what we you want me to raise a new Jira issue over this? Do you require more details beyond what I have stated +No movement +I dont actually know when javaXML is preferable in reality +Arggh +If the projects are either no publishing will occurrA project that we import from gets a +looks great! Thanks Rob +added to for trackingjiangwen good suggestions +Try to verify with the updated patch for +how about +But with patch client driver will commit everything executed before the call to set +oops previous patch had a simple error causing the conflict errors +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Wont this break all other indexes +i know this is a minor issue but i just want to give an update that im working on it and should have a patch ready for review within a week +Its not possible to scan a JAR +I think this might be due to having in the patch +Zhihong Yu I think Phabricator already attached the patch automatically +If you dont mind can you send the code to me +Also having a and a check per app to be more resilient does not seem heavy +nd variant of modified and even more Testcases fr sublist +Yes that is essentially what Im proposing +To move the listeners field from to this issue must be +I cant reproduce this issue +Yeah I think so but I have it on the same branch as a bunch of others so Im waiting to do it in batch +If you add a library which has a offering optional CDI support then the behaviour of the app would change as the container would try to wire up the Inject injection points in app using CDI and probably failA more popular option is to offer the ability to enable CDI for an entire application explicitly +Implemented +You can reach me by email or mobile +fixes a bug +Please fix +I applied the patch +HiI think we should add this to instead of Modello use it on a when it generate regards +Thomas could you please share your experience with us +I like the idea of having it in the script via set command +This isnt a simple issue +OK +grep pipe sed and remove the references +I have tested this with a dot net service and now the actual response is sent to the client after neglecting the Continue response from the serverNeed to test server side of synapse if the client sends a Http Expect header but the fix for the is final because if Synapse has to send Continue it should send it while receiving the request and not after receiving the full request again according to the HTTP rfc +Code looks good upon review +Well do what we can to add features as theyre solidified. +Yes please open a pull request to discuss your work in progress +Hi have you been able to reproduce this at all +Now I understand the code better +patch committed +VPE Selection Tag something like that +You tease me with asking me to commit it I cant +Looks ok Ill just wait for somebody from the core team to take a look at the compiler changes +NN might have aborted due to loss of quorum and unless there is manual intervention NN will not be restarted +Feature +Installer sources in cvs have been updated to the latest from the tag +Patch v addresses the two issues Lars raisedThanks to Lars who verified that this feature is helpful in the described scenarioFor the regular path what we can do is to optimize away all the new code based on disabling on demand column family loading +actually only seem to be reproducable when doing it on the folder +Any updates on this one +CSShover should be used +could you verify and close this one? Thanks +I now have very flexable and powerfull functionality that is IMHO miles better than the equivilent functionality in the commercial search engines I have usedHowever I tried deleting some documents that dont exist and was wondering if I the full trace back was required +Please Verify +Christian do you work on thisIf not maybe someone from the community want to help work on this? Using stored procedures with IN OUT and cursor types would be great +Thanks for digging into this Shai +I dont know Im not using the booking example +Major thanks +Thank you for your improvement +I am reopening for further testing and evaluation +Once there the plan is to branch them in order to have a Community branch and an Enterprise branch +Currently Text is the default clas for mapreduce text input files in which records are deliminated by and a line may contain binary data +IT already information Cobertura generates a lot of warning when analysing java project as the embedded javancss library cant manage to parse source code +reopened for fixVersion +Starting a DN with security enabled and none of the env vars set +is very different +ExampleWar A is a template project +Also the wait for replication can make what the problem is a bit more clear if it fails if there is a write pipeline problem during the test file creation. +You can select this option when upload attachmentThanksGert V +we merge and into one? +The patch to fix the problem +Is this Issue a Blocker or Critical priority for EAP GA? If yes +I am attaching the patch for this review and provide your comments +Unit tests in place to verify that handling of +You should drop and start overOn the shamd yes you should deploy those as well + to trunk. +Committed trunk revision Committed revision after release. +on latest patch +htaccess file +I will restore this value after we enforce that a replica can not be finalized until receiving the pipeline close notification +Updated generation to miss out targetssources so now loanapproval generates correctly +Good suggestion +I dont expect this to get done in but we should look at starting the migration in this release +Seeing that the push system creates enormous amounts of objects it looks like the expired sessions are still sent a push event that then is never destroyed +There is a junit test in class called test +Yes I did suggest that in an earlier commentHowever turns out its quite a bit of extra work to do so which I was hoping to avoidAlso there is already a unit test which compares the accuracy with Dfp +Could you please attach it to the JIRA issue and reopen it? Thanks +Alexey what is this menu +While it would be nice to have I dont think this is big enough an issue for anyone to work out a patch after almost years. +Is the plan to have that patch for as well? Im asking because in another jira the preference was for parameters set through the configuration file +Do you know of others +Itd be tough to do +Closed as a duplicate of. +Vincent being quite new with Clirr I cant see how it supports exclusion of single methods +Currently similar specs exist but only for mathn we are now talking about behavior without mathn library right +We can discuss this in the EG however this would be a radical change from the design both in API and in spirit +It is an improvement +If you see a problem create a new issue for that +anybody has any ideas on why libhdfs was left in the subproject and not in the hdfs subproject +Specify a different file name +after dropcreate everything should work as expected +fixed it should be in +Unfortunately havent found a better way to detect edits yet +My tentative doesnt provide look for duplicate but splits the Description in parts to guide to better also wrote a preference page to store the credentials of this dialog is targetting JBTJBDS users I think the Priority and Components fields are too much to require from a +This patch updated with the support to subscribe picasa albums with credentials +Ive just got hacky batch files so far I dont have the nice shell script system you guys developed +Yeah if you do compensation logic in the component then you need to handle that yourself +They should be in synch as of today +we should also ask all shards to sync to us initially not just active shards better that than rely on cluster state which can be slightly stale +It would really help if you would tell us what wherecondition is and show your collection mapping +log that shows bug is still there independencies? +Daisy and Tony since you both already supported me with this bug whats your opinion +Thank you forthe detailed bug report it was much appreciated +Cloned this issue into for tracking the corresponding feature in and deferring this one to +We would need to identify an alternative xpath implementation that can replace the functionality provided by Jaxen or identify which parts of Jaxen are causing the IP issues and extra those out of our default dependency +Hi SumadhurIm unsetting the target version from since that release is already under way +No no no! This is reverting a fix for file transfers + thanks +Only answered questions were ported to docbook +CMSSession does not have a recover CMS and CMS do not have a close method CMSSession does not have a createConsumer method that takes noLocal as an argumentIve implemented both of these in trunkI also fixed some constness issues that you created in the VC build +Committed to trunk and branch +It is almost like the expected behavior of prototype and session scopes has been swapped +If it increases the size we need to give a separate zip file for javadoc +I think that this section is a bit misunderstanding point of the specification +You must define source and target parameter to your comment solved the problem on my machine thanks for this +Sorry the name of the Groovlet file is not reopen if its still an issue. +I also included a simple batch file to start the have just noticed a mistake in the protocol string used by this test application +Failing tests fixed +It would be nice to get traction on this bug +Thanks +Great its good to hear thats already been the attached example project I didnt observe any Maven version to output the warning more than once during a build +The code in the branch is complete just the documentation needs updating +Also reopening this to reactivate the bugs that are currently commented out +OK hopefully the right this time +Tiho any ideas +More info If the primary key constraint isnt configurable you cant configure two different to use two different tables and have the loader create the tables +On further thought I think its inevitable that the extra thread is needed for a truly robust solution +its in our CLASSPATH +Ill do anything to help +modulessrcmainjavaorgapacheairavataxbayacomponentregistryjackrabbituser +will wait on hadoopqa before cutting next have an updated javadoc for parseColumn +You will have to switch to the Sun policy provider on IBM JDK +In that case the title of this JIRA should signify offline snapshots right +As it stands the Coprocessor could at best function as a post filter +May I proceed with a new issue +How are these permissions granted at create database time +Thanks Devaraj! +modeshaperesourcesruntimedeploylib +This comment should be forThis is fixed in the axiomelementcreate method. +is broken up to package so be careful when rebasing +the test doesnt wait for all of the threads to terminate gracefully before doing so +I belive the from was wrong +Thanks Tom. +Theres no big problem here +Thanks for the investigation work +This could be a more general loader issue +When could that be integrated to the base of NutchI think it could be very usefull for further developments of different parser null causes NPE +The maven problem is fixed for me the new module wasnt added to +Once both are approved for trunk and I can put the full patch into +For example roles gets turned into rol and states get turned into stat +Committed to trunk. +Has this stuff gone into ZK yet perhaps we could reuse some of that code +The JDBC functionality is dependent upon adding support for these in s query language +The has been improved to use a temporary to attempt to resolve JSF managed beans that have not yet been initialized. +integration tests cannot verify that authztokens work so devs wo access to a secure cluster may unexpectedly break tokensThe change to allow tokens with simple auth must be backwards compatible +patch looks good +Would need to run the same version of continuous ingest against both and not the one that ships with it +HelloAny possible date for starting working on this? I consider it quite important because without it the redundancy provided by the use of Message Groups is partially broken +BTW +Looks fine but since we add a new interface we can just as well make the yaml loading just a special loader that is used by default and simplify things slightly +Fleshing out looks like the future to me +The only subclass was not written to work as an argument but only as a program +So either we have a hardcoded xaconnection factory Edans fix or the xatrue jndi env property needs to be embedded in the external provider url like tcplocalhost? websphere there are options in and to add what are called custom properties but these do not get translated into jndi environment properties unfortulatelyIn terms of which route to take I would say that the provider url should be left intact +Editorial pass over content. +Same patch attached +works fine. +Here is the patch for the sameRegardsAmit at trunk r Thanks AmitAshish AshishRegardsAmit Sharma +Does this need to be documented as a Known Issue for EAP? Thanks for your help +Spring now +This is fixed with the patch for by the patch for +Issue seems similar to +bq +After that I was able to start the default profile with as the JMS provider with out any problems +I have tested it with MSVC and gcc Linux. +to resolve it again because jira things it is open +Issue ported to GORA branch patch Committed revision in GORA branch HEAD. +dims +Please verify by marking CLOSED. +Fixed in revision +Committed revision Merged trunk revision Thanks Robert for investigating and Yonik for reporting this hairy one +Create a return for a Sales Order and select the return type Replacement +Revision is probably because the logger and related properties have been removed in trunk but the example configurations still contain the property +Myself and Dmitri Chebotarov have start to work on this +jar +is unavailable whereas this issue is about fullScan of +v is fixing most of what Stack pointed out well keep the messaging via EOFE because I tested it a lot +I use the XML format all the time to be able to connect to indexes of various versions of Solr +via Web Service call +This WSDL with have such a redefine construction +Patch against trunk to support escaping newlines in syslog messages +Its a tad overkill to make users specify a config option in the case when we dont really need itYes for clusters etc we will need a unique identifier for identifying which EMF +I noticed at the logger in the class is not static +Eranda I wonder if you are seeing +This was fixed by token renewal changes in +I will get the tested and comment in due course +Thanks. +Same happens with Topics +or time range +to apply this patch looks great bfost +Group the module by using a parent pom and a directory structure +No exceptions found +This will be used to calculate cluster usage over a period of time +Patch contains new subsections for the Open in New Window and External Link Icon also took liberty to add subsections to all prior info to make the page easier to read +The code drop for this JIRA is checked into +Sean can you please do me a favour and move this issue to a separate bug report and attach the relevant files as well +Thanks Thejas! +Ive made a workaround to this problem I added empty sisetcallback function which aborts +thrift file and test code which will failAlso I have a feeling that the real fix it to get rid of undefined in the struct definition +A setting exist to do that but default it is off +What is the exact error that youre getting? It seems to me that this is a versioning problem most likely unrelated to the endorsed mechanism +The issue can be the bug +So all issues not directly assigned to an AS release are being closed +If a Avro implementation cant be found Ill through an exception +sshconfig nothing more should be needed afaik +I am thinking this shouldnt hold up the release since the test passes on other platformsjvms it is likely a jvm issue +Marking closed. +Patch +removed unused imports +people might start using this feature because they expect better performance +The database is definitely corrupt after deletion of any files under the seg or log directory and behavior can be totally unpredictable +I have verified that the registration procedure now runs cleanly +If that doesnt match the cyrus default then an upgrade will no longer point at the correct path +Hi hairong does your comment about aggregate refer to the the use case when there are multiple simultaneous readers of the same file +Yeah when I asked for tests I was thinking of test cases! +I still see this issue on latest jbossastrunk as well as the issue reported in +The patch is attached +Reclassified to be fixed in the t get any more responses about the current state of skins so the available skins output was changed to belowtimForrest provides the following default skins which should meet most needsCurrent pelt supported tigris available common availableDevelopment available the patch +The previous patch lost some comments from my originalAlso I added a trigramma function + Designer were still able to reproduce the there any fix version or the bug should be reopened? +I am not sure exactly how much else I can offer +Eclipse these steps Start Eclipse +If we are extending the namespace in this area it should be to provide more general support for strategies rather than individual attributes for role prefixes case changes etc. +patch looks good +Seems reasonable +breed yet more debt for those looking for things to complain about +This is not a bug you need to deploy an osgify version of jta API since Hibernate or one of your classes depends on it +I coded up a solution this problem that works nicely +Ive just committed this to trunk and. +Please ignore my comment i was inadvertently looking at the wrong data +snapshot build +Trivial patch skipping s quite simple and clean change Im . +Can you think of a better way to make the new interfaces optionalMake the take it as an option and antmaven can pass in a flag to enable it? Can this option leverage templatesbq +Pleaes verify +However there is no getter for in interfaceYeah +I just committed this +Some documentation needs to be added +Hey Stian I have added simple naive workaround for this issue +However I think its unlikely apps are using this rather expert API and we already dont have jar ability for anyway and as you said if you attempt to pass a youll hit a cast exception anwyayI plan to commit shortly +Also the checksum validation is most probably due to networking issues which were the reason the deploy failed more than once yesterday +Ate Jasha can you verify the fix works for you as well +Passed on Sybase. +Hopefully we get these resolved soon and get out there +Yes of course. +Patch applied. +Will verify against DR once it gets released +Thanks Maxim. +I got the following errorCdomcount eusertempFatal Error at file eusertemp line char Message Invalid document structureErrors occured no output available +Code from the student is being reviewed +Or better yet just attach your WAR +Mukul checked the code in a few days ago. +But easy to fixBy the way This was also not different without in the older patches +So if maxbytes is K with members it is K with members it is MB this does change the time to send a STABILITY message and to be in line with what was suggested above we probably need to remove the scaling too. +only a little different +We would need to wait for the hcat xdocs to wiki migration for that + +The problem is that both and block on construction on these cache locks +As such Ill close it for now as cannot reproduce +bq +This test fails when the data directory of the server enters a particular state based on previous activity of the server +Adds annotations to exclude and from managed request scope +Check what is in SONARHOMEwardeploy +Attaching +Search components in reRendered set changed to same algorithm as in houtputLabel hmessage components absolute paths started from used as is search all enclosed components for a given descendand search from +Im confused since it does not looked like it is related to default localetimezoneI tried on RI and RI both of them printMon Aug CST MST FAILEDEST FAILED +Thanks for looking into this +but its good that you triggered me to verify this as i also have a test set up with hsqldb and this seems to be pretty stable in memory +Thanks Shawn for pointing me to the list +alex what do you need from the jboss jbpm integrationafaik the BPEL extension does not use the jbpm service archive rightwhat integration between jbpm and jboss must work in order for you to get started with the integrationregards tom +That should have been fixed by the fix of to my test it already fixed. +Anyone had a chance to look at the v patch +I think the advice and advice should be renamed to select and select +libs +It looks like bit installer started with bit java and thats the reason for this error +This patch adds a fixture to make the aggregate tests run both with and without territory based collations +So I just disabled it explicitly +Committed +thanks pat. +Complete +I think this should be marked as fixed an a new issue related to supporting the WITH DATA option addedMarking this as fixed means that it will appear in the list of fixed items for the next releaseEven without the WITH DATA option it is a useful feature +Assiging to Brian M +I thought I took the latest patch from jira +Also it would be good to modify the JIRA title accordingly +is the updated patchNow this functionality is available for current New Customer page +RomanCos I spent a few hours searching docs and poking around Hive code +As I am saying in it exists a problem when the cascaded entity has idWhat happens is that when persisting or merging an entity with a to an entity with auto generated ID the id is not updated in the object +we need another patch +Although many email clients have those features in the case of it makes sense to have many of those features in server side +The issue is supposed to be fixed in jbsosesb +Setting the priority to low +Charles Thanks for the commit infoI would have expected that the backquote situation would be a special case of the more general methods +Thanks for reformatting +SELECT FROM tickets t where gid product SELECT FROM ticketgroups t where tid LIMIT gid and product id hibernategroup gid +This seems to be a result of some domain URI checking that was added for OASIS compliance +Sure I will commit it once you make these patch significantly different from the previous one +Thanks! Leif +Please make sure to update the javadoc and the implementations +I am moving this to because thatts the next release available in the dropdown +Realistically if you leave it the way it is its not going to get changed by anyone else in the foreseeable future +Again the bug no longer exists in +I still dont understand why using without a tx is undefined behavior +Fixing regular expression for wicketmessage attribute +Im not sure if I got them right though so Rick Id be very grateful for your review +The linking behavior is slightly less aggressive than in the past +Wildcardd that +this should be an refactor listener I guess that allow users to undeploy the original I have a feeling this would be better done at the WTP level not by us + +The javadoc warnings are unrelated +The interface is just too useful in terms of debugging and getting a snapshot of site health that I couldnt give it up +To make this work need to fix merge to instead of appending to the new store file the passed src store files instead it interweaves as scanners do currently +Thanks +You have a message with bit content and without bit this is not RFC compliantFurthermore we probably could even reject it on the SMTP side as bit content is not allowed when the server does not expose BITMIME extensionSo if I understood correctly the scenario your client is violating rfc rules and we should close this issue because it is an issue in the client +Could you try trunk +Rspec script to validate fixI hope this patch helpful +This is actually a bug in javacapt +You are right its not so logic Im removing from the filtered properties. +Verified fixed in IR +Heres an updated patch for. +Doesnt this mean that just about any Rails app deployed with has a vulnerability if you request a URL with a very long integerIf it could be determined what input is troublesome could an exception be thrown as a basic countermeasure +Heres a patch which addresses the issue by doing a sort of locking on the NN HA state +Ive added httpclient to maven distribution and its working for me now +This needs to be done for PHP too +Thanks Sami! +not ready for the issue is valid it seems to be just part of a bigger task which is setup properly security on undertow subsystem and once that one is worked out this particular issue may not make sense anymore thus resolving it with wont fix +bulk close of resolved issues. +Have a setting that would allows people to bypass this settingI will proposea patch that does that in coming hours + +Hi AbhijeetTo answer your questions True is not out yet but its going to be backported to the next release of CDH that is coming pretty soon +so it is an often used pattern +Sure no problem +This was not downloading the file at all instead it was getting a error as Eric describedCommitted revision Thanks Eric +Im pursuing to put a uber patch in the uber jira +Ive confirmed that optionaltrue works as expected when the associated entities do not share the same ID +Is there any absolute requirement for this +Thank you very much Andrew. +The test error is unrelated +Then FUSE can create a instance with this kerberos ticket cacheIn the future it would also be good to use privilege separation to contain the power of a fusedfs instance running as root +I should really clean it up and post it +Ah now I understand what youre trying to do a little bit better +Have a look at +Verified JBDS CR +For instance the current approach to row based security is applied based upon the assigned roles each time a query is planned +Somehow youve got an empty hints row there that hasnt gotten compacted away +This sounds like a good plan +Im not able to reproduce it on andI close this issue please reopen it if you reproduce it on one of the listed versions above +To be specific the annotation is giving a problemWill create a JIRA issue and will follow it up BTW why is Jettison jar file needed in andThe standalone application complains about a when the jar file is not presentI will do the needful thanks once againCheersSathish +Shouldnt the method be called to? Its not quite obvious whats going on by just looking at the name +Great! Sound like a fun! Will do that soon +well new releases shopuld be properly changed and csv file issue as project deploiement will be moved to waiting for an account +Verified in JBDS Beta +Use the above function for decommissioning failure handling and the unload admin commands so we get similar behavior and single implementation across the threeWhat do you guys thinkdhrubaAgreed I will leave the configs as is for this jira +Hi LeifCan you attach a diff u so that anyone can see the changes a bit easier without needing to perform the diff themselves +For instance a string with homejgordonhadoop libusrlib would also have problems with surrounding quotes because the would be interpreted literallytouch this fileis different fromtouch this fileIn that case the more ideal solution would be to not pass javaopts as one string but as an array of options or a hash to improve the ability to reason about them at all layers in the stack +Looking at the source this would affect the behavior of Cluster CLI getJobsget methods +Attaching the v patch rebased against trunk +Hi Amar I have a couple of questions about the history server +Regression tests added to test. +It doesnt make sense to say that and then not specify any nodes or racks +Checking on this +Maxim please could you verify and close this Jira or provide steps how to verify it +as though Id not made a fix though fix is in the code +Closing as requested. +Commented out the reference to deployment plugin it doesnt exist yet +Attaching which introduces the unsigned arithmetic optimization discussed on earlier +bulk close of all resolved issues. +OK I see +Patch is availablesomeone can help review and commit this patch? thanks in advance +This was triaged for SOA +But for now this will give you a chance to review code and provide your valuable feedbackIt features options +ant test says all the tests fail where as only one two tests fail without the patchDoes Hudson actual run Zebra tests +Ill update the patch +Additionally I think ODE Scheduler is a core component and it needs to be very customizable flexible +Merged the pull request thanks. +weve had a few issues with jline and pulled it out. +Apparently this can be improved upon and made totally transparent including of YARN daemons +Gridmix unit tests and passed on my local machineI just committed this patch to trunk +Null check applied. +Added +OK I seeBut We havent decided the architecture or the architecture is not fully shared +Why dont you just create your own implementation of and set that on the Connection object +To ensure that everything happens in the correct order each event type has its own variable and each variable change is separated by its own timing event +The description can say what is null +Sorry for the mistake +I just committed! Thanks John! +If we are shipping winutils with the distribution it should do the right thing which is to either fail the creation of unusable symlinks or handle the path conversion +Im not sure which API youre looking at Shane there is no registerSynchronization on +Why dont we open a jira for adding an ee test case for this feature and let this inSounds reasonable? Let me know if anyone has any concernsThanks +So I wonder why it could have worked for you before +Or maybe or instead +User can see buttons but can not click them +I think it would make sense to be in the branch so that it breaks fewer jobs migrating to +JNDI pools usage has been added along the datastores and its standard part of the GS closing all issues that have been in resolved state for months or more without any feedback or update +Also after this is implemented evaluate whether is still an issue +Sure thanks for pointing this out +For some odd reason there was no notifier for CocoonI added the notifier and observed Continuum for couple days +I am going to close this issue now since there is no bug here whatsoever +I have showed this in the sample app +Now also supply user and passwordsinA javatoolsorgapachederbyimpltoolsdblookThe new class that handles output of role definitions and grantsM javatoolsorgapachederbyimpltoolsdblookI noticed that authorization ids were not properly quoted on outputI added that since userrole names can also contain quotes and weirdcharacters +Add srctest is a dup of +Thanks for reviewingYoure right we dont have a reference implementation or corpus that we can validate against so forking the Swedish was perhaps a bit optimistic +Committed. +The task will be closed because its no longer datethe editor was completely restructured. +Bulk move issues to and +Great +Thanks for the reportIm going to check this +This should be easy to do and is not something wed want to release note +Has this been fixed? Also this patch makes no sense to me it appears to just remove and put in the same thing +for token string literal replaced with constant +For kicks you might log both memcache size in bytes and count and resultant log file size +yep you were beaten to itThis is a good test case since we dont handle inner classes particularly well either +hi Jon I am of the Solr cache implementation +Check for the party method +Hi new to yarn where do I look in the code base for this +does not break existing cli scriptsBy cli do you mean cqlsh? If not how will a CQL shell avoid breaking scripts that use the clis syntax +I spent sometime refactoring the camera code so that it uses less memory +Do you get the same error with even a trivial Groovy hello world example or only with Grails +The patch looks good and simple +catching and logging Exception instead of just after release +See what Jenkins says +Ive just checked using and its fine there +Patch in r +But maybe the file to be referenced should be referred to in this config file +m not sure what the protocols are here with regards to when things get committed so I apologize for my possible ignorance +But if we end up taking that route I think we should leave the scripts for the sake of consistency +I have moved it ot the othe category +I have added these to the images directory and Im rebrewing the document now +Closing issue as version has been released on. +The attached file contains the release properties used by the release plugin +Thanks for pointing this out! There was a check missing for objects from A relies to operate on a object so we need to invoke it only for a object from a +This property contains a list of test classes to be excluded from the run +those comments were in regard to when the issue was opened +Ismael can you post a brief example demonstrating the issue +which is a copy of s welcome text +In general serializability of objects is not a container concern +the offset in the file to start at c +We can not alter the table to add a PK or create a view and point to this +Mahadev verified that the classpath on both the AM and tasks containers are the same by looking at the CLASSPATH after passing in the true +Ok +I think the problem described in this issue is not completely fixed yet +Committed + +This holds for all the issues you have raised +Please can you provide more details how to check and close this issue or could you close it +Fix committed +Attached patch with integration tests +Merged to the branch in revision . +Now we have several individual bundles for the Blueprint container +Find attached a test case that illustrates this issue +Im running trunk svn from Friday James is running Any other information you need? Any fix I can try? Or logging I can put in to help moreThanks cases perform a direct rev lookup and both return empty when they shouldnt probably the same cause +This has nothing to do with local support and everything to do with where on ZK the file isThat should not be configurable +I dont understand the following query seems ungrammatical and them seems ambiguousQ +Add the no option to the java command +is done +The Javadocs on Error states that these should not be caught +Code for the Axis isolation layer this is fairly small its main tasks being the mapping of actions to the Axis service and conversion of Axiom to DOM +Is this Issue a Blocker or Critical priority for EAP GA? If yes +This seems unrelated to +Sorry for misinterpreting your comment Dag +This doesnt work on linux because separate arguments split by in are not glued back together into one command line for execve or execvpThe approach to build command line for in should be different +based on CP testsuite distrojust replaceadd files into +Just wanted to point it out so we could clean it up when possible as I was working to port this change to WAS and saw it was still there +Thanks Amir for contributing +Thanks Namit! +Closing this because Ive added a new issue to be fixed in +To me the fact that it works in maven but not in STS leads me to believe its a ROO problem +Sending assemblysrcmainSending jbicommandsSending jbicommandssrcmainjavaorgapacheservicemixjbicommandsSending jbicommandssrcmainjavaorgapacheservicemixjbicommandsSending jbicommandssrcmainjavaorgapacheservicemixjbicommandsSending jbicommandssrcmainjavaorgapacheservicemixjbicommandsSending jbicommandssrcmainjavaorgapacheservicemixjbicommandsSending jbicommandssrcmainresourcesspringAdding jbicommandssrcmainresourcesorgAdding jbicommandssrcmainresourcesorgapacheAdding jbicommandssrcmainresourcesorgapacheservicemixAdding jbicommandssrcmainresourcesorgapacheservicemixjbiAdding jbicommandssrcmainresourcesorgapacheservicemixjbicommandsAdding jbicommandssrcmainresourcesorgapacheservicemixjbicommandsAdding jbicommandssrcmainresourcesorgapacheservicemixjbicommandsAdding jbicommandssrcmainresourcesorgapacheservicemixjbicommandsAdding jbicommandssrcmainresourcesorgapacheservicemixjbicommandsTransmitting file dataCommitted revision Waiting for the kernel release to close this close this issue once Kernel is released +Can you please look into this? +I dont have time to figure out and redo any math right now but I hope in the future that we can have random tests to verify this stuff +Bulk close after is released +RightThen this patch is also worth I think +In when we move to Java we can make Resolution an enum and then use a switch statement on passed in Resolution +Got rid of +It reduce memory usage and side effects in times +Patch committed to trunk + +Were now ready to close the lists down +Verified +Going to have to close this one as Cannot reproduce +All we have to do +Ah that sounds about right +postponed until CP +The client code however was acting as if they came back in DN DN DN +And this is a function may block +I think the root of the problem is that file modes are not so Java doesnt have an API to access themAs a can you instead of directly invoking your scripts invoke binbash ? That way you dont require things to be executable +The persistence unit gbeans are supposed to be using their gbean priority to get started before the ejb module gbean or any ejb gbeans +That sounds like a good the wiki page for the Java DSL and Spring configuration example. +Yes turning off the workspace resolver fixed the transitive dependency issue +Hi Permaine Im sorry but I didnt manage to commit in time +Help in resolution is appreciated +It is legitimate that we use syncLimit at the part of the Leader code because when we start that while loop the leader has already received enough acks for the newleader proposal +Waiting for a test that can be uncommented as soon as is ready +I want to test on cluster before applying +Other then that a big ! Great work Niels +I have just found it here both parts of code that requested to open JSP editor were in our code and it was easy to figure out how to prevent the conflict +there is still a potential issue with corruption from a reboot that should be addressed under this jiraIn my understanding the partial writes during syncs that Bill is referring to would be captured by the crc checks +Weird when I setup the same type of userrole on my windows development box it works fine +Katheys guess was close but has its own policy file so the extra permission must go into instead of +Please file jiras to addresses the issues Bikas has raised if needed +Added missing jdom dependency to console configurations as required by use in information portletPatch was created on windows xp from geronimo root +My only fear is to the logics behind the restrictions +If thats not the case can you attach the you are using to reproduce this +Theoretically it is necessary for optional fields to be omitted but we can skip it until someone needs it +The new PMC is raring to go +Hi NaokiThis is a very nice and useful control +It should be deleted from the branch too + Is it reproducing always? Yes Is it with the fresh setup or before creating VM did you destroyed some Yes its fresh setup before running automation we server clean up primary and sec storage Attached Created VM In UI i can see ip address allocated in NIC tab but i cannot access access with this IP adress Access this vm from console proxy ip not configured for this vm +RayIve fixed the inventory item problem following a different approach the screen now accept a null facilityIve also committed the viewprofile patchSvn rev + +Dan can you give me some pointers here? +I just committed this +Thanks Lily I will pick up the patch and commit it Monday so this useful enhancement will make +Right +dist +Its configurable so clients can set to whatever value if they want toIts retries on trunk btw which is min +Maybe the cause is an unsynchronized method in the scheduler which results in a broken store structure +the attached is a patch patch has been appliedPlease verify +In implsrcmainjavaorgapachejamesprotocolsimpl only the already existing methods are implemented all others are auto generated stubs +hmm read this again you are saying parent pom includes dependencies by default but where is that default since Tycho normally dont include optionalDependencies does it +Resolved +One simple solution would be that when the NM tells the RM about reduced resources then the RM only reduces its internal by the amount until is reached +Patch to add ideax task for creating +In other words if we come to already marked subroutine we need to follow the path till the entrance and on return examine ret local variable indexIf we come to the top level code than no analysis is required +I changed the href to link back to and I also removed the ifexists because the variable was already tested going into the if block +So now we have passed the DST hell the issue still occur for you GavinI cant reproduce this on any of the platforms +I dont understand why in patch you downgrade shell and to be version +No worries at all I gonna fix. +However Java is quite a pain to invoke especially with multiple JVM switches such as Xmx +The rest of the code should work as is +Seems ok so far thanks Adrian +I dont believe its possible for this bug to be valid anymore +Yes +Id say Tomas is right immediates only use is to pass on conversion validation update model and execute invoke application directly after decoderegards the wished. +Fixed by GF team +project and +Thank you! Ive changed my version of surefire to and it resolved the problem. +MickaelDo you think the close on rollback also has something to do withIve create bug for the new its just a lot ofsorting out stuff at the moment +automatically build add to the package target able to compute large timePartition now able to work on now able to work on +Please give me some workaround so i can finish my task +Sorry for the confusion I tried onUpgrading JAXB fixes the issue in the endorsed directory fixes the issueDumb question where can I download snapshot build +I have been seeing similar problems with MSLAB enabled +The Sotto Custom stylesheet now has the new +I meant to say +Provides a way to run on Eclipse using bit JVM selected in preferences +A patch for the code in svn revision thanksdims +Integration test added +Perhaps there was a misunderstanding from my side +Committed to trunk and. +Test case +I dont see how a pooled connection factory would stop the from closing the sessionconnection resources unless it is using resources that filter the calls on the close methods for the objects similar to the way the and filters requests to the stop and close methods +seems that the post killed the indentations in the first example but they are actually there in the output +Running through some final testing to see all if the builds are all fine +In any case that wont help with the OOM problem since we expect to be able to eliminate negligible numbers of L sstables that way which is the problem here +I can provide an updated patch for branch if theres still interest in this +The second one is a bit different and I wasnt sure if it should be in there or not +Heres a new version that changes a few more tests from to and also shortens s in a couple of tests +Ive opened the for the resources usecase +Attached is a version of that could be added to the Lucene core as submitted by Yura Smolsky a user +Thanks for your contribution. +I reviewed it and especially compared Bokml endings to the previous version and its completely backwards compatible +Potential patch for think you have attached the wrong patch hereSuggested patch attached +I suggest removing prototype from the commit test configurations on the basis thata it takes most of the timeb errors in the baseline compiler will most likely be picked up in the other to note that building on rvmppclnx is about times slower than Intel +All uncompleted issues are now moved to the new To be determined for future releases category +The disadvantage is that it is yet another reason not to get an Apache release out +Failed test is capacity scheduler +Hi Guenther Thanks for your helpful reply as a matter of fact im afraid i can able to move from Jboss Tree cache to Infinispan since we are using Jboss AS GA +Using Stream semantics you can avoid this +I also changed the package from toCould someone please review this +Filed and for the pending work +on s suggestion +It is true but if it is really needed we can implement it as a single big catch block combined with a series of statements for each exception type +That was indeed two comments Not sure about these +How about if we have a dirty bit that gets set when you addmodifydelete a pooldirective +But still +fixed. +Could you elaborate a bit on what exactly is supposed to work and doesnt work for you? Here is the problem I am having with +My upgrade to Blocker was based on my comments from May +This removes the downloadRuntimes from the extension point used and instead uses the download extension point which pulls data from stacks +I have imported the live data from July st in my system +For super wide rows we are going to be working on the slice filter patch this is for time series data +The puts line in your example works correctly +Not a problem All help is appreciated +Matthias Id like to integrate this new functionality step by step as this makes things in the long run easier and more traceable +A config should be usable will all types of transports +Patch looks good thanks Uwe +When set the mallowDetach flag to false for catching cases set the flag to the contained instead of the containing +The Wrapper would be a Java Proxy that exposes all the interfaces of the orginal message and simply routes calls to those methods to the wrapped object +Furthermore I dont see why something being a business method invocation requires something to be a CDI beanOk +Please add more details and reopen the issue. +This caused NPE on bit jvm preventing CPC reply from being sent +Ruby not yet +The fixes completed for got us most of the way through an build of using JDK but not all of the way +However for now we will just leave it open since it is a known issue +There are two separate issues warning message is being handled by of which this is not a duplicate +The repro worked as expected with this patchI shall run tests and revert back +Patch in r. +There is nothing to stop the implementation from a cookie in autoLogin +Since this is multithreaded I am finding some difficulty in getting to the actual cause +javax +If the command line tool runs locally very little functionality beyond calling the interface methods would be needed +site is a build plugin +Im not an expert so I dont know if these are the right numbersThanks for the fast fix +This patch here has no problems in that regard as we just move the methods into new superclasses +It also resolves constants referenced by name +Patch looks good. +fixed with the xmlSchema +already fixed it again + +The attached contains the implementation of C support for XMI serialization +If there is no clean solution for this now lets use the old for this test and we update it when is properly done +Thanks for the patch +One of the plugins you are using probably relies of xerces which is no longer shipped with turned out one of the installed plugins required xerces +Raw is a good term too +It did not dont know what a protected tag is but above are the steps I executed +Ill need to write lots of tests before I can commit +The attached patch changes the +Lets get whats here in independent of that fixup +Ah +However rules can continue to be true even after they have fired and be removed from the Agenda +I dont see a harm in supporting it +How about making udf match? Wouldnt that be a better approach? We already have definedFunctions map we just have to make it +Fixed will be in CR. +I would recommend to create an Help umbrella task and to put existing issues as subtasks under +Thanks! Resolving this issue as Fixed. +How was this implemented? I need the exact same in EAP. +Please always post a stacktrace with exceptions like that +The Runtime Detection plugin will recognize EAP only if you add EAPHOME or its parent as a runtime path +The rest will be fixed in. +Changing the way the Vshpere server is started fixes the error +Just confirming that it is a regression between and +Pepjin if youd attach the code in the description as another patch and check the checkbox to contribute the code to Apache Ill be able to use it +Thanks for packaging and trying Ryans patch Tim +The running app context includes the reporter etcGuice injection for Tasks is set up to use instead ofIn a follow up jira could be included in the with guice injection setup for +We will consider some custom caching of Expression objects for Spring +Please comment if I have this wrong. +fyi im referring to length of the property NAME not the property value +Youre saying that youd rather have any application supporting Avro to have that support hosted in Avro rather than treating it as a library to be included? This seems like an odd approach to me essentially inverting the domain knowledge of each application to Avro rather than the application itself where its developers frolic and work +This is a real mismatch between DOM and +I see yes it is the foreach inserted to enforce the number of columns thats much easier +Oh? Okay Abhi will take a look and dogfood with tomorrow +patch v fixed most of the issues but was failing because it was passing an incorrect HTD to append +The Chrome browser on Linux creates transfer data with the type that Eclipse recognizes as which is the cause transfer data to not be rebuiltI have fixed the issue as follows added a new transfer type which correctly recognizes the url created by Chrome ensured that the new transfer type is used only on LinuxI will create a bug against Eclipse +I think the allow attribute is great +Similar validation checks would be in register new widget and it works there +Moving to +It seems like a very reasonable price to pay for ease of administration +Yes if abstract isnt possible an would at least get closer to the source of the problemFrom my perspective at least backwards compatibility is already broken since Lucene doesnt work with that dont implement getbut I understand yall will have different requirements in this regard +Great catch are you saying that doesnt work on Linux? What if I just lobbed off the beginning if running on LinuxThanks for your patience + for the patch + +Proposed patches for respective revisions of the file +A couple of points +Fixing concerns will fix this also +Not hard really +bumping to +Here is better one +Ill work on this issueThanks to Caroline and Pete for the info +Assigning this to myself let me investigate this. +Closing. +I did consider this path but didnt want to submit a nasty looking patch before discussing the issue +From my understanding this jira when I install JBDS with bundled EAP and open up a new workspace I should have server configured to deploy to metadata +if running on a windows machine you may need to runmvn execjava into the branch and into the master branch. +But often in combination with an join and cancelling the remaining tasks +Jitendra The previous code would not set the service if the user called which is broken +Committed revision +Added patch to fix the problem locally to +Ive copied your comment towhich is going to contain the performance improvements. +Fixed as part of changes in. +I dont know if thats new or not but I was used to just seeing that with adding docs so when it still jumped by when i added docs it appeared a bit weird when I lowered the step to I saw what I expected but I didnt realize that the odd numbers I was seeing was the searching task logging by +Draft patch attached +Please note that I also made changes to the docs comments and packaging +This is a very simple change and easy to merge but I dont know how to test it +Blocker for and +This has been fixed in the Portlet Bridge codeline but hasnt yet been fixed in the codeline hence the bug is left open as a reminder to merge the work +No problem thanks for the help +Patch committed at revision +is empty does it make sense to define an empty proto message to be used for all void returning rpcs +We need to fix this in the Annotation Editor +Patch attached +Is it OK to close this issueOleg +Affects Array Array Array Array to be fixed as per nightly test run. +will commit if there are no to and trunk +but its exactly the opposite apologies for wasting your timeStan +But Im not sure whether this is worth the effort +The main problem is that the really needs the scorer to be present to work properly +fixed +will come out a month later and will be included in the update for JBDS +Add database part +The only includes the topp namespace in the wfsowsURL attributes if it is also present in the request +Thanks for reviewing AdrianoIll commit tomorrow unless anyone objects +Hudsons is my fault +If the system being queried was not a SOA system for example SOA then the bean and the functionality does not exist +snapshot +has been released closing issues. +Ill check tests as well +If this is running on Ubuntu nodes then I think we need to ask for a Qpid lock to be added to Jenkins so that we can prevent it running at the same time on the same node as part of the Java test matrix otherwise they could fall over each other +patches are anybody cares about committing these please let me know and I will produce updated patches if there is a chance to see something committed I can these enhancement to the M eclipse plugin as you make this optional +Verified on build +Were shipping in lib +Draft text statesWhen users created new Rules projects with JBDS and SOA Platform they hadto add these JAR files to the As of SOA Platform these files are now being included in each server profile +User has a collection of Roles +How can I map the from uri to uri to a Java DSL file without writing a stupid parserCan you show me some hitsThanks JamesIt is much clearer for meI just have another question how can I get the camel context from the gshellIs there any registry in the SMX where I can look it up +Is it correctI added into imported project Seam Support but it doesnt add Seam Builder and Seam facet file in +This patch shuffles the list of brokers used by both producer and consumer to fetch metadata +I know that this is not yet the case for NER engine but we should keep this as a design goalOther than that I am ok with using the taxonomy for the default classes in the default configuration of the of packaged launchers. +The patch looks good +bulk defer of issues to +Seems fixed +Ill see what I can m marking this issue as done feel free to reopen it if you can reproduce it +maximizing Solr Cloud indexing performanceI think until you are using more than a single thread per server you will be very very far from easily maxing performance which is why the other issues interest me a lot more when considering a concurrency improvement +Can you attach a simpler example +Let me know when you want me to look at something or do some testing +this has the same bug as all strings are assumed to be binary +I am testing a tool for testing concurrent code +Which rpm install are you talking of +On the last platform generate the tar ball and you will have a single build with all of the native bits compiled on multiple platforms +hmm +Yes I think its gone by fixing recent or something +handle wasnt actually removing the endpoint from gossip beforeright that was a bugbq +Committed revision +Stupid patch submission dialog has no place to actually put the friggin patch +requires trunkexamplesresources +Fixed using instead of + +before image when I enter the when the page returns after I need htmlEscapetrue to prevent scripting but not all characters are unescaped as expected becomes however more common characters like and are no m using and seems to be a bit odd it looks like youre seeing the raw HTML code there not properly rendered by the browser +Similar can you rename to C API libhdfsSome of the docs need to be updated to mrv still but that is a separate jira +But you do the change I let you decide on that one +You give me too much credit Just ran out of time +Ive applied it in revision Please cross check and close this bug +The patch adding a single debug line +Also the chance of getting into an infinite loop with equals is very small +It was a detailed patch that must have been time consuming to create so the least I could do was spend a little time checking it when applying it +not sure why I didnt close this already. +New patch tightened up more testing +Network Server used to be a separate product so traditionally had a proper name but probably doesnt make sense for the network client driver +It doesnt have to do with groovy in the subflow I had to specify explicitly the view in each view state +Sorry for misleading +I also need a sortdistancePatrick I cannot get you current localsolr to work with locallucene trunk +To me that makes perfect sense as well +By the way do we have a windows build box? It would be great if we can run buildunit tests on a windows box automatically +I think they used this method for variable substitution +I cant reproduce this +Actually phrase queries would be really interesting too since they hit the term positionsdid you compare it against MMAP? I +The vectors can be wrapped around with id as the name to trace them back which solves the problem explainedSo I am not sure whether this fix should be applied or not +Do we need additional test cases for this issue +This should not happen any longer. +I looked at performance vs +I have updated on trunk +Hence I was waiting for committing this to branch +This makes them easier to parseIt is the time to rename all the classes since the pubic API is going to be changed a lot in anyway +Adam any updates on this issue? I am working on creating the Sqoop website +previous attachment was not correct. +This issue is duplicate of but I have changed fix for version to got it fixed for EAPEWP +At the moment I have no any way to update this file to git +From the trace I can tell the method is not called +I can confirm all the tests passed +Added timeout to unit test +The way Exceptions are handled in this patch is that each thread With a new impl we should consider how we want to do this carefully +your right with that i forgot but i still think avoiding space is better as using this tagFYI using is slow because railo use reflection to access this method this will change in railoits better to do simple string concatinationFYIin this case is better than no synchronisation needed +Can Apache projects rely on components under prohibited licensesApache PROJECTS cannot distribute any such componentsAs with the previous question on platforms the component can be relied on if the components licence terms do not affect the Apache products licensingFor example using a GPLed tool during the build is OKQ +Currently it only displays everything! More to come +Ill check +for the latest patch +I will be happy to mentor this project with Catherine Hope as my +I had to adjust a bit to the refactoring +Its a bit more complicated than it seems +After all +Then taking a snapshot is just rolling the version of a subtree +Fixed in r +Closing old issues +SvenThe attribute Im talking about is value +defer all X issues to +disable the for that fragment +Jonathan Thanks for the comments +Thank you very much +Should also be fixed in recent versions of Maven where bigger extensions are loaded into an extension container and not the core container. +Applied to branches and trunk +Thanks to Martin Koci for provide this patch. +It fixes but it breaks tests that are being run by +Hi StephenThis cant hurt and is I suppose the right solution +Thanks Jitendra! +attaching a new patch removing some more javac warnings +Please could somebody with more Karma change thisthx +This is what I wrote +stack I was hoping we could address concerns in since this JIRA was riding on thatHere are my responses to your questions for this JIRA We seem to be removing the body of the class +Thanks Sandy! +Please take this entire conversation over to the list +Looks like this issue has already been addressed with a separate patch. +yes but a isnt that great either I think that JCIcould easily detect if the classloader should be able to find it and attemptto look if it hasnt tried already +Excaping would also work but Id like to find a way to automate it or something which I need to for the commentUnfortunately that patch didnt fully fix the problem +the output of ls laR or some equivalent +Jarek I found the problem both ESB and EAP are using an old mvel version when drools depends of latest is Fixed should receive the configuration as a constructor parameter To test against eclipse janino doesnt support JSE Enhanced FOR I found a problem in janino when we use variable janino tries to tokenize as an innerClass will open a ticket in janinos I think so too +This should be done in one of the next iteration before +Not sure if you read the discussion in section snapshot of being written files +Merged into in revision after releasing +OK its actually on my other machine +Each wizard page from the project example wizard initialized its own version of archetypeModel when in fact it should have used the one set in the requirement page which is the only place where its infered from the stacks information +I think this is a bug in dali validation +Please have a look at wicket +Im not available todo this now +Add factory class to return appropriate instances of and based on the configured type +command is missing and was replaced by command in +Committed addendum to and trunk. +Thanks +closed +I believe this is now fixed +Frank I think we are talking past each other +I just committed this +It seems that the problem is the null characterGoogling a bit I found that Postgres could have some problem to manage strings with embedded zero bytesMy proposal is to replace always all chars before to store a task execution message +Assigning to Ian but if he doesnt comment on it soon Ill close as not a problem +Ready for QE Rajesh +Some files seem to have tab characters the indentation is longer than two spaces +Stefan yes as you noted this was fix for EAP ERgiven that I am closing this issue. +my mistakeWhy dont you want to check for initialization of the persistence managerHere is the patch with Stefans suggestion and tested it +Incorporated the comments +In particular whats boostfilesystem? +No objection to putting it into trunkI dont see any need to put it into or +Thanks. +I appologize for wasting your time +close to jira issues were updated the last days +Thats actually how I found it +HiIn the functional spec it does not state if this feature will work with the richscrollable also +BTW maybe should be considered when reviewing this issue +trunk +JanWhy is the from being used to test RF? That demo was never made to work with RF properly and was replaced by was curious how this example appears here since Ive never found how this demo appears here bud hadnt time investigate more about itProbably legacy bits from earlier testing and not managedremoved correctly by gitClosing as not reproducible. +ben can you take a look at it? We should commit this ASAP since is waiting on it +I didntd see this issue in latest automation run +Bruce the runtime error is a separate problem +This issue seems to be duplicated by issue also. +Attaching EJB which Failed to connect to server localhostwhen shutting down +Is this fromIf not you can try to add the d flag to the command to start scriptserver? That should give more information about the reason for this +Changing the existing behaviour would therefore not be acceptablePerhaps for Betwixt it should default to true instead? Or is it too late already to change existing behaviourEither way it would be nice if the javadoc indicated what the default behaviour is +We have to introduce web login for solution +This is the project I encountered the problem in +Thank youMohammed either event I would say that we probably need to remove this contribution in the short term so as not to hold up the overall release of the next version of the client +Gary assigning this to you +You can just removeouter xslotherwise and xslchooseAnd whoever applies this patch please keep Javascript and Python versions in sync +Could we please make the failure less catastrophic than an exception +I have finished looks much better now not just sizing of the will be releasing in a couple days +It turns out I need to handle a special case for StoresThanks Alan! Im looking into it +Unfortunately from my investigations there appears to be a bug somewhere at the language level +does not apply +An update needs to be made to the spec regarding default constructors. +I had attached the patch +bulk defer of open issues to +Thanks Glen! +I think that should be fixed +CPU bound processes would definitely benefit from having the VM stay around for the length of the job +I meant this +Here are some smaller issues to would be nice to see a total document count across the shards +from invites a where Hive history filetmpTotal jobs Launching Job out of Number of reduce tasks is set to since theres no reduce operatorStarting Job job Tracking URL httplocalhostjobidjobKill Command Ccygwinhomebinhadoop job localhost kill job map reduce map reduce Ended Job job with errorsFAILED Execution Error return code from So what is the problem and how to solve itPls suggest meThanks. +I agree with David That said Im not sure that inventing something totally separate from is necessary or particularly useful +I have attached my comments to the propsal in the wiki +On a current mac book pro with firefox the graph makes my browser pretty unresponsiveWhat version of Firefox? Its pretty zippy on Ubuntu with Firefox +patch looks good +Its our use of boolean as the target type in cast +This patch should allow the diagnostics to report the appropriate pointer address instead of a garbage pointer address +RM appears to be throwing NPE in response to the NM heartbeat +Ochoa do you have any update? I would recommend that this issue be closed because it is a usage problem and not necessarily a specific bug the best way to continue solving this issue is probably on the. +Thanks to the Torquebox and Rubymine teams +Thanks Omer! +An quick and simple example of a that I think it could just replace in click examples + is in so ARQ should not be loaded you cant just fix the test by allowing either because tests should have no dependency on ARQI assume the scenario that is affecting you is that you are trying to test a custom model implementation in a context where you do have ARQ loadedIf this is the case then I suggest maybe the appropriate refactor is to extract an abstract superclass from with an abstract protected method that returns the Class for the expected reader for the format in the given context +same is true for the +Thank you to Snjeza for setting me straight +Will keep an eye on it +The nic will hot plug but you either have toreboot or manually trigger a dhcp query +thanks tom. +So all issues not directly assigned to an AS release are being closed +makes sensePlease correct grammar in javadocs +I had a look today and am happy to take it on +So does that mean it is fixed in trunk? I thought was an older version that +Appears to be fixed +im using ASA and ASA . +Arun what is your takeLast while we all agree to commit this to +Server maintains a table keeping track of the recent operation result history per client +I think the simplest solution would be to remove the Sonar project url from the Free style job +where the object instantiation cost was just too much +Heres a simple patch that limits it to on the main page and indicates if there are more that are not shown +Of course we could do more +I ran the test over times without a single failure I run the full test suite times without any issues +I added the plugins to not sure more is required for it +As it is +Could you please create it against the latest from trunk please? I am getting errors applying the patch as it isGaryThanks Thomas +fixed. +Im resolving this ticket. +Unassigning due to inactivity +Am I missing something +done a first pass in rstill need to check if it is in r +LGTM . +The second retry fails because mistakenly assumes the URI has already been visitedWhile Tinyproxys behavior is clearly broken should also be able to deal with failed redirects and retry those correctly +A second fix following Ians comment +The patch which fixes the issue +There are still instances that could be shared between tests and thats not even considering instances which could also be shared +Thanks +At least I believe that you strongly suggested that and I have no better alternativeProceeding with your steps then +Dear MaintainersWe have released recently and cannot synchronize +Let me know if my modifications work for you +Attaching a patch and a screenshot of what it looks like +With Im going to move all the analysis factories into the analysis module so we will need a way that we can share across multiple Factory instances not just Solr cores. +If no one disagrees can you close this issue Charles? +But I am sure my code worked with one of the beta releases of +i would add an additional step if POST url contains and extension use this url +I added a hidden input field to the form and an to the that submits the form +I did not install a database in glassfish I used what ever Sonar has embedded +The while finalizing a block sends the to the +James can you look into this? Thanks +I applied your patch +The cool thing about it is that all generated code is outside of files editable by users +I guess I will wait until a user asks for this or runs into this problem +Marking this as resolved as per Alans comments. +it makes sense that empty strings are converted to nulls this is what of users expect +running tests +Committed with the ticket at svn revision ThanksDimuthuCan this issue be marked as fixed and closed +Resolved while woring at integration into Castor main project. +We may want to consider moving some of them to core to allow wider use +Thanks Hasitha for highlighting this usability issue +Red herring. +thanks paying attention for this issue +Finally I found the reason is the propagation on MEM operands +attached example that works with plugin. +Updated the patchThanks +To have one or the other just put the value you want +So argument of SOAP doesnt have is null and void +Removing from NEEDINFO which means moving to ASSIGNED in Bugzilla +Committed trunk revision Committed revision +Paddy thanks for being patient and reporting and testing this issue +Enis +With the changes in we still found crashes on the lab system but we did not catch any in our development environment yet +Updated patch to improve javadoc +Im not talking about combining them +The patch works great only watch out for Java next time like the autoboxing of that boolean +This is the one shipped with jboss and the one with jboss only differs by the entries in this +attaching patch +The fix for this issue was incorporated in the changes made for. +Nice +setting fix version for a fix +pushed in Developer StudioVersion Build id Build date +and I deloy them in same jboss +Doesnt really matter as the solution isnt going to be accepted as is +Oh thats funny indeed +added to book to Andrew for reviewcheck off review items from informal review content of documentation +FYI Dawids introduced use of Guava outside of the clustering contrib +Leaving this open until I commit the jars to the affected projects +nd patch to fix cleanuphelpers confusion in the face of snapshot files +Oh and the repository id can be of the form which will be created ashomedomainsrepositorycodehausorgpublichtmlorgcodehauspicocontainer picocontainerrepositoryIs that the plan +Modified files in diff u files and unit ve commited this to the repo after no objections +Thank you! That exactly what I want to hear +We generally do not honor HTTP request headers because an object does not represent a request from an end userLastly this patch doesnt actually do what the description claims it does +For comparison our cluster also runs job frequently with new files across nodes and we havent experienced this issue with Hadoop +Probably it is better not to break it +Fixed in final +bq +Feel free to bring it back if you think this needs to be fixed for what it is worth this bug cost me a large amount of time +You got a range error since you entered a negative font size. +v attached for branchIt basically catches on both sides and lets sessions closedI also implemented onFailure to make sure latches count down to avoid process hang +This is related to since this was encountered when attempting to work around +Thanks for the report! +Spoke with Tim +Attach final version of the Patch other diffs are old +Odd Ryans first patch was giving me issues without the jars in core but perhaps I just hadnt refreshed ivy right or something +installationNeed to dig somewhat deeperdirk +I agree with Chad does update at the end of a +I think cannot assume anything about the token it gets as input and once it returned a token it cannot assume anything about how that token is used +While debugging I have removed the true entry from the exchange targets filter properties and the messageExchange was routed to the cxf jbiclusterenginesrcmainjavaorgapacheservicemixjbiclusterengineSending jbiruntimesrcmainjavaorgapacheservicemixjbiruntimeimplSending jbiruntimesrcmainjavaorgapacheservicemixjbiruntimeimplSending jbiruntimesrctestjavaorgapacheservicemixjbiruntimeSending nmrapisrcmainjavaorgapacheservicemixnmrapiSending nmrcoresrcmainjavaorgapacheservicemixnmrcoreSending nmrcoresrcmainjavaorgapacheservicemixnmrcoreSending nmrcoresrcmainjavaorgapacheservicemixnmrcoreSending nmrcoresrcmainjavaorgapacheservicemixnmrcoreSending nmrcoresrcmainjavaorgapacheservicemixnmrcoreSending nmrcoresrctestjavaorgapacheservicemixnmrcoreTransmitting file dataCommitted revision . +My expectation in looking at the patch is that by setting exectype in the config file Main should then respond by starting with that exectype +It is a pretty simple fix +FergalRoute got simplified a while back the ASF now assumes that any patch submitted via JIRAmailing list is implicitly contributed to Apache unless the submitter explicitly states otherwise +Thanks Uwe +Our initial root tablet doesnt have its replication set to its the system default +Manually tested with Chrome and IE +make sure that you list the files which need to be executable in svn +Fixed under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Thanks Amar! +Do You think we might reopen it to get it fixed in? Id really like to use some of the new features that were introduced in Maven but this issue prevents me from upgrading +Pushed to to progress? We have some time until is released but if this is not resolved by Im going to close it +I managed to reproduce the first NPE error by not applying the serviceClass attribute on the I will commit a quick fix for it shortly +If we are breaking BC better we do it as a part of I say + +I think this is actually similar to the problem thats been resolved for management server +I have looked at the source codes for Hive and Pig +sorry missed your comment you can disregardthanks for your helpBilgin +For we may have a that it first chooses and then chooses storages +If it applies cleanly and the tests still pass sounds good to me +This is related to number +Ok Im now +The server should be able to detect this and will not send locators +It does not say that there are violations on those rules +the documentation has been updated +This is fixed. +push to +When the Leader sees the new version it sends the LE epoch and zxid in +Can you pl point me to an sample xml document for Hadoop that uses Forrest? I can then use those XML tags in my document +That probably doesnt change your reasoning for wanting it in but I figured Id point it outOne question why is it that this function is needed if nothing is currently calling it in +Code was committed in revision Issue +Different approach I found that playing in the lower levels could mess when a client would close a connection while sharing it with a region server and other stuff like thatIn this patch I created a new inner class called in HCM that manages a connection to ZK +In windows console BOM appears as ? in stack trace. +Firstly i wanted a notag possibility too however i figured out that that would be very hard to make since decorate needs a tag that ajsupport will rerender if used +So problems like above do not occur anymore +When user wants to externalize same value the key from the properties files will be usedLogic and validation have been updatedResolved +The Programmers Guide and javadoc were updated as well as removing XML comments of the unrecommended wsdl scheme from the quickstarts +Fixed by + +The only reason why I have not started working on asynchronous version of is because it do not want to end up the only person maintaining itCode duplication can be as bad as abuse of inheritance +In case we cannot find a solution now lets just wait until the problem blocks something significant +build with mvn installyou need to build from svn Thanks for valuable input! I will try this on Monday when I get back to work +In thanks for the heads up! +In those cases an empty collection seems a much better choice for clients +The first spot declaration of showCaptcha makes indeed no sense because its never used most likely a relict when portletPreferences could not be resolved +optSubversionliblibstdc +But as a handler author he knows what he is doing that is why I mentioned at the runtime handler can check the message context to check whether it is need to invoke or just ignore the messageHmm I dont think we are in a position to make these changes for even if we really need to take this into consideration +Changed the description of the issue to reflect that this is a documentation bug and not a code bug +Even if keep deleted cells is false +To get a better understanding of the use case and what I should document I assume youre using a singleton instance of the +Ah since we have the snapshot dependency on the CXF we need to add the snapshot maven repository in our pomI will did a quick fix it +Bulk close for +LGTM +Scheduling the dependency upgrade to Quartz for the Spring timeline +This issue was resolved after applying the patch and increasing the stack size using the S option +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The file never got applied in this patch +pushed to beta for in JBDS B +This is needed to register the CDI Extension +to commit this +I will resolve the conflicts and regenerate it as soon as possible +So I this issue is resolvedThe issue that the server should return date values always in UTC should be addressed in its own issue if we see this as a problem +to reopenMost experts agree that is bad code +End user picture looks like this +has been deprecated and is not supported at this point we recommend users move to the in the project. +Non locations will make sense when affinity based preemptions occur +Oops this needs to be visible after all +Without testcase or your Seam component source code the reporting issue is useless +it does not reproducable with I get this exception only if start all jbds plugins from my eclipse. +Project has no Archives feature enabled in project preferences but Project Archives view still work as it is looks like Project Archives view works for any project even if this feature is not enabled for the project +Rebase the patch after checked in +Like filling out forms be it directly of via fdfxdf and using overlay in forms alse requires a not a +Merged the patch into branch. +This should be checked for for all examples as well + +Oups sorry Jacques I trying actually netbeans I return to my svn line client to make a better patchI add small documentation and for iterator no special comment I correct the code version of patch with documentation and remove iterator +A readMe already exits we just need to update it +I think a simpler solution would be that the Job Client does NOT load the instead it only loads the if its presentThis way only the default values on Server side take effect +First draft of generic WAL mechanism which builds on the changes +Forced appClass parameter to be specified in sr command +Thanks for looking into this. +Thanks Andy! +Thanks +Isnt this just a matter of setting the limits as you like? To be clear any set smaller than a given limit is not sampled +to test the hudson patch to trunk +Logged In YES useridProvide a file that allows for this to be also appears that class is missing +Thank you when I find solution I will add appears to be an issue with either or configuration not a bug with geoserver +Fixed in revision In addition to the changes in the patch I added the utility class which can be used for importing +looks good to me +is shown instead of the ofbiz error logSure this is a great way to contributeBTW do you mean that you clicked on an existing link or have changed value in URL manually has you mentionned in your st comment? +Save it to buildlib in the jruby source tree rebuild jruby and let me know if the verify error goes away +Im marking this as backportreject because the test was not in junit format with so this failure would not occur. +Thanks patch applied. +This patch fixing the issue +We can resolve this issue unless somebody disagrees with that solution +It is necessary to add some attribute description into the documentation. +Submitted a fix to trunk and also added test cases +especially since we are looking at a replacement of this suggestion document it as a limitation +Laura answers to your questions Yes the example below is accurate +initial proof on concept patch no extensive tests run yet +Why is the still exists in contrib directory? Isnt it necessary to remove the file from contrib +Im going to file a different jira and put this file in location exctly as hadoop doestnx +I would like to add that I am experiencing the same issue the same stacktrace down to the line number +OK I finished the first draft of the patch +This will all typedefs that are instanceof +Svata if you can reproduce this problem reliably it would be great if you could run with true and true and post the full contents of +Or a random one wins to make sure we dont depend on coincidental but repeatable behavior +From the RFC The mechanism consists of a single message from the client to the serverBefore the first request the client sends the SASL anonymous message but it does not block waiting for a response instead it directly proceeds to write the request +looks great +SNAPSHOT isnt and btw we dont use this naming anymore +Done +I did observe the problem at one pint but this seems to be fixed on the master branch +Katta is more like an distributed index serveryes we do not need to fit katta into Hive +New patch restores the last line as expected +I racked my brain trying to figure it out +Also it was discussed that the constainer resource capabilities could be resized +Id be happy with an admin page that lists all the roles greys out the ones that are attached to users and lets me delete the reference counting I see it counting up as you add groups to servers and then losing track of reality when servers are just shut down without first deleting the projects +Test case +Committed to branch and trunk thanks for reviewing Tom. +Verified status of this issueBase on devins comment this issue maybe can be closed +Yes is correct I changed name from because is better handled by github discussed the next Magnolia release with and it will take place on January th +Better would be to not use one action for multiple form beans but use multiple actions which can then call into common code if necessary. +Latest patch applied. +Why dont we put this in a new file +There needs to be better exception handling as well as diagnostics +Originally this was not a problem on a small test tar file that I created but now it is having the same issue. +This strategy will only mimic that already does it +The bug is in and is triggered by order in theThe following description discusses of HibernateIn line call is searched in the sqlString and if it is found after ? and it is assumed thatthere is a procdurefunction with output parameterWhen the first ordinal style parameter is found it is further assumed that this parameter is thetarget of the result of the procdurefunctionThis implementation is a bit naive as seen in because call can amongst other things be contained in the name of java bean propertiesI will attach a Testcase and a patch shortly +I think having xml files with rn as line separator is very common +But I am looking for a url parameter that works well with the page parameter defined in the Seam application +Thanks for explaining NicolasI wonder if a good default would be some fraction of the flushsize +No its because you are a moderator and therefore already are a subscriber +let users to configure that +Patch committed to trunk +Unscheduling due to inactivity +An ugly relic of the past that wouldnt get through the door today +Committed to trunk +Not much else we can do +Now available in build snapshots +Hello VladimirCould you attach your blocks as zipped archive pleaseIt would make it easier to take a look at your problem +Uh +st pass of st agree with that but I didnt know what the best alternative is +Actually this patch was to update the options available +Also I think that documentation should be part of project wiki is optional. +A quick check for this would be to see whether the corresponding jar is includednot in the hadoop classpath In the current case the wildcards wont be of much help unless the user is ready to recursively check every directory +Thanks Devaraj +Thanks for reporting and for the patchCheers +Thanks! +Fixed in jbossweb by r +has been released closing issues. +Closing as pull request no longer applies to master +From JBOSSHOMEserverdefaultlib copy contains instructions on how to integrate portal into the ejb mode to be done post the based jems installer +Im also in favor of renaming it +Mixed up with Anoop based on your comment does that mean the configuration of is irrelevant? Please advice +Fix +Although Id suggest maybe adding a ctor that takes a filesystem instance +Thanks indeedI dont know how you guys accomplish so much +Text extraction works fine using the currebnt trunk +Could you clarify whats meant by the server shutdown abnormally? Was it a power failure an application crash or something else? Thanks +No its not +How about we commit this to and above? Any thought +Sorry I should have noted that this assumes you have a property configured in conf +was closed without fixing it but this error is really annoying. +Or perhaps its okay to delete it as long as it doesnt leave any Template with no other Build Defs +I just verified that this problem was fixed in M. +No comments? I would like to finish and commit this +weird actually reproduced this when installed +I guess I usually have it installed in my local repo or Nexus +We can always changr the name again but it is imperative we produce a release without the name npath in it +I have reviewed the changes made in code +Another class B which is mapped to table TB +Please close this issue if you are satisfied +committed to trunk nice catch +batch transition to closed remaining resolved bugs +Best for the reply +For the record metadata cache store . +Please check that it was applied as expected +The patch removes the extend clause for +A better patch +I also have to add some more tests to this +fixed the issue with revision +Thanks Kim +But it didnt +Update with a small intro section on JSF and and then point to the wiki pages for more +We should see significant gains by doing IO +README file updated to include missing remark concerning stopstart of you follow the remarks added in the README file it works NOW +Note that I attached the patch to this ticket because Im unable to publish a ticket + +No single culprit. + make serializeable make by Eike KettnerThank you very much and well done Eike +Good news Jasper! Do you plan to stay at or to join Codehaus or another more powerful forge to host the project? +The other Aries Modules are compoents but is not and it seemed like a good idea at the time +This feature is not something that should be included into the core but rather as an extension +Have verified it in geronimo +Ive done the fix and written the tests +Is that what you meant? Im fine either way just want to make sure this is what you meant +Plus it makes it impossible to transparently upgrade from current structuresbq +Wish you luck +Thank you for your patch first of all Norbert +I will rename the processor I am of course very interested in your patch +Ah could be +I just need to test things a bit +Replaced the use of Xerces by JAXP which causes the following changes constructor throws a if JAXP is not able to provide a transform +So Im reopening this JIRA as we need to find a proper place to document this behaviour +I just committed this +Fixed in rev +I dont think thats the solution hereI agree even if we prohibit recursive deletion at the Namenode itll get implemented in a library and called nonetheless +here is a patch +That works +Go through the emma reports and add tests where we are particularly bad +Added check that a schema exists before trying to print it +didnt notice that +Hi SergeyI am interested in asynchronous processing in general including the following transports in priority order jms http https +Also the AIOOBE exception that I saw dumped on console several times when running EUT jdtdebug no longer appears so it seems like the same bug with index is fixed somewhere else too +Yep it will skip detection if the field defined in is not emtpty and updated to fit new directory structure updated comments to point to Wiki docAlso optimized regex now patterns instead of using directly +Thanks for reporting this it should be fixed in CVS +The redirect now works using +Thanks +Looks good +After approval by Van Ive merged the proposed changes from onto the master and branches and marked as resolved +The change that you may not like is that all exceptions are still wrapped +s the filter definition syntax for collections +They enable durable sync +for wont fix +Fixed +This changes some visibilities of code thats already been released +Update mvel version to +Patch provided by Mete Atamel to address the above mentioned limitation +I havent tested the patch for trunk +The patch creates the following new filessrcmainjavajavaximageiometadatasrcmainjavajavaximageiometadatasrcmainjavajavaximageiometadata patch has been appliedOleg please verified. +Committed revision +bulk defer of featurestaskswishes from to +has been marked as a duplicate of this bug +Seems this patch break tests +Alessio go ahead and commit +Fixed in revision +I confirmed docs by patching it +I tested using both random and ordered partitioning +very much RickCommitted patch to documentation trunk at revision +Still getting same issueVerified the server is now running and not wssj +And also unlike a distribution package this is not going to be used over and over again +If I have time Ill try to get hold of the branch and test that. +Ill commit this +we should check in the source for the docs +Thanks Mike +Thanks Gary for all your support I will wait for you guys to final release before I start using Spring INT forSachin +The class will copy these properties to the object when the is created +See the new it planned that HQL would honor the? I ask since currently HQL is not designed to honor the fetchjoin in hbm this is true in all cases for attempts to fetch multiple bags +This bug has also been fixed +Can we update the bulkloader itself to also not use gossipHave you looked at the first patch? I specifically did that first because I knew streaming was the hard part and changing the loader would be way easier than dealing with hadoop and troubleshooting streaming at the same time +You will need to add the axis jar files to lib and change where the BEEHIVEHOME is at in the two ant files which are in srcEnsure that there is a blank +Attaching patch from to let Hudson run it with MRI realized running it with HDFS Hudson was not useful since the patch does not change hdfsThis patch is the exact copy of Joshuas patch except for a couple of cosmetic changes and removal of public from two methods +Added a test and committed. +Diffs for version I didnt see your request from last November +way better than before +boxgrinder config +Thanks +I need branch as well +Nothing to do with +Streaming support can be provided with a far less complexity by leveraging the streaming work to a new jira since it will be based on a different design. +compatible patchA couple people mentioned that it would be much easier for them to benchmark if I produced an compatible patch +Thanks for the explanation + have no problem with making implement Iterable +Attached is a proposal to modify the Getting Started guide and anHTML version of the proposed new patchPlease let me know what you think +But maybe with a read workload it doesnt respond the same way +Lets change the order on this How would you like this to be configuredI can always do the necessary coding and preconstruction in order to follow the configuration requests you can come up with and build any necessary reaper methods into it in order to have it reap in the appropriate making the cache LRUtimeetc +Skimmed this very quickly looks nice and clean to meWhy is this not in contrib yet? I didnt spot any dependencies +looks like a works +One major difference is that an error in the token persistence layer is not fatal as it is for the RM +Ill make more comments on that issue in the jira thats opened. +one other comment however why did you not use the performFind service to search +Seems to be OK with CP +Committed at revision in +The fix was added by Rafael in Qpid trunk at I have reviewed and tested it. +Also make sure that CATALINAHOME is set to the full path to the directory +Verified in buildtest moduleThanks for your help! +Ill take a deeper look at that and update you on the status later today +from me +Please notice the difference +The latest included runtime is so this issue has solved itself. +Lars I mean replace the in withYes we need the as it maintains Soft refs +I just committed this +The length of a string is limited to a short not a int +What is interesting to see is that in the Spring framework they dont use it probably because they came to the conclusion that it doesnt work in a servlet container +Did you check the log for errors +Patch works +Moving remaining bugs to +This should be discussed +Yeah Im not saying it always work but interrupting a thread will only work if the thread can be interrupted +Hi everyone just wondering is there any movement on this +Missed changes to in previous commitCommitted revision . +patch to remove the extra semi +Im not sure if we can make gossip state transactional so we can rollback cluster state to its original +There are new versions available I will try the latest first parser was upgraded Jenkins build passes but it doesnt solve +Findbugs are related to this patch +has few plugins that in some cases will fail to initialize +Please assign the bug to me if this indeed regression due to portable IP feature merge +That prevents the mouse updown events from targetting those components and creating the problem +If they do span multiple projects and are hard to split up you can leave them under HADOOP +If you resolve a module using the ivy file of that module you wont have the published artifacts of that module only the dependenciesHowever you can use the inline mode on the resolve task to accomplish what you want +also refine the buffer size of could you create a separate jira to track this +is passed when invoking ant +Blocker for and +I noticed that there were some unused imports and variables declared in some of the classes +Downloaded the Axis binary zip from one of the mirrorsThe documentation in that also says it is for +Ideally should be resolved within an assigned release but somehow it just occured to me that we dont live in an ideal world do we? Or is it just me +Looking into it +It would break backward compatibility if encoding isntThats a nonsensical statement +The other case when this limits is set specially by ulimit s command +In memory buffering of request content pretty much eliminates all the benefits of using in the first eliminate in memory buffering do you mean to use something other than the in the on method? The uses the older +Please let me know what needs to be modified and further enhancements to this +Ive shut off the network adapter Ive attempted to wireshark sniffer to see traffic etc +Not sure I wrapped my head around the issue completely +Questionp is a persisted object thus persisted in one slice +Ill be finished over the weekend and will either try to incorparate this in that or work with you afterwards to get a new patch +By the way shouldnt we contribute this to +This is this. +I am also not a fan of of isSupported methods though +Im sending a patch showing some implementation for it +In the ETL EACH of our greenplum slaves reaches out individually to the derby instance to post back result sets and does NOT use any resources on the master +Bulk close for release +The entity class may have other constructors as wellThe constructor must be public or protected +Please make it a list of strings so that messages arent lost when the next one comes in +Since delayqueue is processed first all we did was try to process the close even though in the other more live todo queue we might have the process of the regionserver shutdown waiting the process of the regionserver that was carrying +I doubt it will be exactly the same style as before but it should be a bit better +Justin sorry I missed your comments this morning +For RHEL fresh install we need to add to the list of packages to install +Didnt have a DB ready +covers most of the output but likely doesnt have as many permutations +Note too the above page lists EAP but not +AM +Here is a scaled down service that generates the same error +Get the latest Rhino sources from and compile them +Ive deployed new archetypes with the patch it should be working now +Thanks Konstantin +At this stage of the discussion is there any intention that the replacement for string fields will permit analysis at least analysis? After all that is one of the main reasons people get confused +set fix for on the dev list consensus on moving to JDK I have updated the trunk and M branch build instructions accordingly +v at plus RPC options works for meThe Thrift HTTPJSON server is an interesting possibility here it could coexist with the regular Thrift service and still serve the same API +Suggestion tested and patch attachedIll commit it when my ASF account has been established +Logged In YES useridone more. +subclass the core test to run things on a mini cluster +Branchx is now using JBC +BTW I am developoing with this config +This is not a bug +Is RC is the build I should try +Should I update the patch to log a warning +view is that will go away and it will just be SolrYes it will +Patch is committed +The test sends messages +So we started based on your good work and implemented some morecomponents to support our portlet servlet module defines the interface that is very usefulto handle web requests environment api documentation of the says Mainly intended for generic web request interceptors giving them access to general request metadata not for actual handling of the request +The test passes on the RI + +I suggest committing in place of +Theoretically we shouldnt be hitting the DB at all for this Entity since we have a and we know which fields are dirty +Dont think the test results are relatedThis doesnt require any new tests +Patch to add internet address parsing to javamail including unit tests for the package +Moving out of +It now corrects for its parents width when using RTL +In that case each createfile rpc could have created multiple components in the specified path +Fixed the sql to make it a valid statement and changed the logic to expose the exception it was previously falling thru +They have unique path in one Sink to distinguish prefix no matter repeated path suffix +Please switch to the juddi mailing lists rather then jira for this +Joydeep Agreed although currently the only use case for that is inside text file format +It is the same key here too +yet why does this block did you mean to link instead +We will consider backporting the fix to if it is reasonable to do so. +late minor items +The Apache snapshot repo should now contain the latest code if anyone wants to give it a tryFeedback welcome +People can write their own to implement this kind of behaviour if required. +But frankly I would like to know more about your use case because you are doing things here that are a bit out of spec +or do I have to build the railo myself? +It definitely sounds fun +For now start two or more remote processes and call Debug AsRemote Java Application and you will get the Remote Java Application dialogHave maven dependencies now been added correctly to the source attachment +thanks for working this +Applied patch from Dan. +Even on freshly imported GWT Project archetype +simple app with one class that exposes the localessimply deploy and navigate to localhost the class that exposes the locales +Additionally it is optional to allow properties in the SELECT WHERE and ORDER BY clauses that are not explicitly defined in the node types listed in the FROM clause but which are defined in mixin node types that may be assigned to node instances of the types that are mentioned in the SELECT clauseBut again the JCR specification does not define the behavior when properties are used within the SELECT WHERE and ORDER BY clauses that are not defined in the node types listed in the FROM clause or subtypes of those node typesTherefore I will conclude that does not need to enforce which properties can appear within the SELECT WHERE and ORDER BY clauses +Sorry I must have clicked Resolve instead of Close +for my self +Actually this only matters for local mode +More importantly it would be nice if the seperate modules were only dependent of seperate modules in spring itself +Could you give it a shotThanks Lincoln the sample application was not working for me without this part of patch but Im not sure about proper fixI reported another issue addressing this problem +Selecting the version was a mistake I meant to leave it unassigned because it is a new feature +for the approach as the current one of full fidelity replay while great when it works is subject to being brittle +There isnt a bug report for that one and trunk behaves like tc +Is it possible to delete a plan after creating it? I can configure it but I dont see an option to delete it! Actually all I want to do is modify the unique ID on it because I gave it a stupid name but I doubt you can modify them after creation. +Thanks for the prompt and courteous tried setting the cache parameter to CACHENONE and saw no obvious change in that we investigated our Weblogic configuration and found that we had not configured any session pools for the JMS we tried configuring a new session pool WLS requires a parameter named Listener Class which it defines as The name of the listener class for this JMS session pool which is used to receive and process messages concurrently +RafiI added a check to see if both routingKey and bindingKey is specified and will throw a Regards have incorporated all suggestions and this has been testedTherefore I am going to resolve the s a patch thats the start of attempting to bring some sanity to this +Resolving as is resolved for ER +Unfortunately after been made redundant from back in February I dont have access to be able to fix Roobot now and there is no one in who can or is willing to try +ve tested again after svn update and the problem seems to be fixed somehowThanks Dims. +Please close this issue +greatBTW can you share the tablets distribution of matrix a matrix b and result matrix +There are a few things that need to be changed howeverI dont believe we need a default setting in a properties file +Yes the last one is v and its the one in RB as well +Committed to trunk +In one suggested by me earlier on this JIRA and by Bikas in his previous comment if I want a container only at node I submit my request for node with the flag turned on +Ive merged it into branch to test it let me know if you want me to commit or youre going ahead +You can configure a or in Spring a +Assigning to Justin as he implemented the joining support +It allows the webResources to be filtered. +Initial patch for review +Can we close this issue or is there remaining work that needs to be done +is not targeted for ER +Cant see the issue Sebb +Thanks Anurag VivekAnurag your patch is in trunk rev +Committed revision optimization for and +As I want to use method from the wizard I need to change method from private to protected +Cheers +Looks good +Ben +Patches a and a should be committed together I split them up to ease the review +Ill see if I can create patches independently for any other issues created as well +SergeyCan you make sure the jbi endpoint that camel from used can be access from the JBIs quartz timerAnd an example can help us the find the issue can you submit the camel route for this +This seems already fixed in trunk +ThanksChatra +released with correction +Then looks like we do have to keep SD so what is the advantage of putting additional columns in Table instead of SD? I just feel that code and schema needs to be duplicated down the road +There we flush based on two triggers An upper limit on unflushed records and a trigger +So in the worse case we make two passes of the column index but not the column data +Committed to trunk and Thanks Vinod for the fix and John for verifying it. +I guess you could ask a user to select the appropriate account and connect that specific danger there is if several users share a common provider account +Sad? Maybe + +However the patch resolves all modules in Lucene and I usually also create an umbrella bundle which has a dependency on all the Lucene bundles with such headers +bq +Patch incorporating suggested fix for test suite on +The file should be renamed to make the file naming consistent +Ive applied attached and the BPEL compilation error has disappeared +The file is my diff in order to get all destinations including temporary destinationsThe is my jUNIT test which show the problem and validate my diffs +its got trace messages at the same EIP to backport to to backport to branch as well +If you finish it before we release beta please correct the version +Instead suggest users move to the wicket viewer +And this is the reason why these builds work on RHELx but our builds do not +Updated patch I beefed up tests for both nrt caseThis is ready + +Looks like this issue is probably ready for closing with the result being the added unit test to show that Codec cant do much here. +the resolved issue. +Another workaround is to use the XSLTC as needed +Source copied from release +PR applied in in JBDS Alpha +Sorry again for dropping that +I suspect these issues are related +Oh I shouldve mentioned the two test cases which will fail without this patch are and +settings +Committed in branches and you Sathwik Ill be compiling from sources then +The handles this the same SOAP binding thing is going away dont worry about that +released as +classpath I know what I am doing +Let us know +Noone is arguing with the importance of declarative transaction definitions or of AOP itself but the strategy of declaring these in annotations +modified jchris said I should just attach the whole file instead of a diff +DprojectcameltestsDprojectcameltestssrctestresourcesDprojectcamelcomponentssrctestjavaorgapachecamelcomponentjmstxDprojectcameltestsDprojectcameltestssrctestresourcesDprojectcamelcomponentssrctestresourcesorgapachecamelcomponentjmstxDprojectcamelcomponentsDprojectcamelcomponentssrctestresourcesorgapachecamelcomponentjmstxDprojectcamelDprojectcamelcomponentssrcmainjavaorgapachecamelcomponentjmsAt revision is not upgraded to on upgrade patch has been merged in the branch + contains tests that fails regarding this issue +Its an interesting one +We fixed this by use only +I dont want to see full classname in class property +Although it is hard to imagine a case where I wanted to build Hadoop and not have svn installed +Ill try it out +Not sure then what is going wrong +Works for me on +In the interim we need to workaround this issue since this is a problem for consuming +We should look into this once trunk is built with this issue with Maven Build project +No just added the link to commit to keep track +When you are comparing the results this can be annoying but its not incorrect +Perhaps we shouldnt log such events +Please hold off till then +XHTML seems to have as large a following in the web development world as a dead sheep despite its obvious the jspdirective tag work as outlined +Any tips on how I can check if the temporary destination is deleted +I believe this is addressed with the move to autotools in. +I looked into that +Added the proposed move issues to and +Replaced the previous with a new one that has a simpler way to expose the encoding and decoding methods +have a test to test file copy between journal this JIRA into two for better issue management is used for creating JN http server and this JIRA is to enable the edit log transfer +Jarek pointed that the previous fix will break if there is no in the bin folder +Please close if satisfied thanks +So wed have to synchronize on them or maintain a pool of them +SOAP Arrays are no longer supported in +Happens to me all the time and Im never sure seems the stack trace here is related to shutting down eclipse trying to stop the debug plugin aka shut down eclipse +Could you be more specific? Maybe change to +Is it okay for me to commit my patch +Ive updated the ticket with a more complete stack trace +This causes the table name alias to be appended to the column name but puts doubles quotes around the column name as wellselect as EXPIRY geoentity +I will file a separate bug for that. +Thanks AkashYour patch with minor typo fix to the form name is in rev +The problem is that its building a sparcv libjffi even on sparc +Bulk move issues to and +Ill see can I locate that info +patch applied +Then we add a dependency from to the new component +Bulk close for Solr +One warning though if you do something like delete of the total size of HDFS at once it can cause some rather long starvation on the though since I havent seen it much past secondsOn configurations under high load youll need to be careful as well but this is the case for the hbase timeouts as well so should be okay +Patch for and Vladimir Patch committed in revision +I applied the patch and changed a few things to private and protected in two commits +Thanks for the feedbacks +In fact I think there is no need to throw since master will assign to the same regionserverLogging a message in regionserver is enough +Were using and still have this issueWe believe its unrelated to as we see all associated params being sent through to Struts upon a form submission + + +on trying it +As a bonus this would avoid the need to kick the CRM thread after every addmodifyetc +Verified and closed. +old +It helps reproducing intermittent problems +OK if expiring is a common use case then log level for expired messages should stay trace level +This may lead to SVN head showing compile errors until the whole task got +I think the license is actually not OK +doesnt support Eclipse +Requires component upgrade release notes +As in offerAccepted cound check for existence of bean with the Seq Id for both requests before one of the session inserts itI have not been able to track down why the deadlocks appear but I might have found a temporary workaroundIn Sandesha findUnique is called where it is expected that either or is an unique keyIf we can make this constraint on the we could check if key is already present before insertingProposal to follow +This is clearly intended for development do you have any idea regarding option or ? will be part of release further work might occur for has options now lookup cacheHome refresh +When implementing a system where each end user has one or more databases this is an essential feature +But why should the exception come at all? why is it coming as soon as the management server is started? this was not seen during in previous releases +Hes working on porting to the new assembly build. +This will be a requirement anyway for future patches so its best to delegate this to the sys into the system classpath +Perhaps just class? Id keep Jobs as higher level constructs which orchestrate multiple driver or sequential operations to achieve their goals +The current API is not compatible with multiple wars in an ear I think that going forward we should try and remove the static application field if possible and have all code that relies on it responsible for passing in the application map otherwise it will continue to be a source of bugs in multi war applications +Patch applied +At this point the connector may be complete +Thanks Jimmy +I think I agree with Jonathan +I generate a new ledger for each of the runs so the value is different in each +please let me know if theres a better way to do this these js files need to go in a new dir named jsLibs intrinidadsrcmain apache license to version of patch which also updates the client validation dev guide patch thanks Gabrielle +To the best of my ability to look at this the absolutely only answer I can come up with is that there is some other project in your workspace that has an incorrectly formatted fileset somewhere in its +Hi GlennThanks for your patch +These failures seem unrelated to this patch and the same patch already passed system test framework yesterday +With our current geronimo classloaders there is no way to write a dependency plugin without marking dependencies as is done with the m pluginIf we have a directed acyclic graph classloader in which every car file or jar gets a classloader of its own AND we modify the m build to intersperse configurations and modules so that a module such as for example the connector builder gets its connector classes from the connector configuration rather than the connector module jar then we could write a dependency plugin that collected the dependencies +Whats the process for getting the RHQ documentation thats in the related admin ref +They were intendend forSorry for the inconvenience +Thanks. +Ive added target to all build scripts in the Signavio Core Components that call javac +Hi AdamYeah I noticed this when I looked into it +The docs dont sayyou have to alter your bean properties to get multipage validation +ChunhuiBut deleting the node what improvment we get here? Still the scenario that i was mentioning may happen right? Correct me if am wrong +I think it was the last checkin on github before the cutoverIt is important that decompression always happen with the codec used for compression so it cant just be the case that there is some property in the config because a mismatch on producer and consumer would lead to unreadable data and if two people send messages with different codecs you would be totally screwed +It was finished before I committed my latest changes +I removed the enclosing modal panel tag and the selection parameter for the suggestion box is populated again +Resolved +Thanks for pointing this out doc should be updatedIn fact this is not a strong dependency supposed you should be able to build with standard javac as well +At the same time in +close off release +bq +Thanks Bryan +I applied this patch +Decided to use BCEL for pack pack implementation so patch has been applied at r. +This import file came from the same repository as the previous import file and we just need the best way to stitch the two import files together so they look like they are oneI assume you have all of the necessary privileges to do that +I dont agree that Asset Maint should emulate Project Manager instead of Work Effort +To mitigate the backward compatibility problems we could phase in this change over the course of a couple releases warning people in that the old usage will be deprecated in +Looks like that it supports computing CRC +Subscribed Farid myself. +This means that the first parsers service file is taken the Tika one and the Vorbis Parser one is ignoredWelll need someone who knows how that maven plugin works to fix this by having the service files appended together rather than using the first one wins approach which is ignoring the additional files +PR for master branchcontains only test to validate that the properties are supported by when the MDB is merged in master branch +I am currently testing his patch +Current latest versions are hibernate and +trafficline s v trafficline xtrafficline mingzym +Already Fixed +An employee could abuse the system by shippingsomething personal and then deleting the transaction +Also it would be nice to have unit tests for your implementation to track regressions +my weekend was busier then i expected so i went ahead and committed the patch to fix the bug and spun off for future improvements. +reflect changes in grammar for statements + looks good to me +Attaching management server setup has cluster wide primary storage and zone wide storage poolsWhile migrating VM cloudstack is allowing VMs volume to migrate from zone wide primary storage pool to cluster wide storage poolfind is being called when user attempts to move a VM to other hostfind is trying to find list of suitable storage pools that are attached to each of the suitable hosts for the VMCurrently the selection of target storage pools for each volume of the VM is left to storage pool allocators +great thanks. +Resolved in Branch and HEAD +Committed to trunk and. +Thanks Tim +If you can get a strong reference in a cleanup routine then its ok to use it because the fact that you have a strong reference will prevent garbage collection +Thats a good suggestion + +Snjezana can it be removed without causing problems for existing installations +Fixed incorrect sync block +See for further details. +Please check that the build works as you expected +Though it is possible to add to them in the preferences page +Version of doxia sitetools which uses version of the xsd is the released version +This mirror will be in the next update +Max allowed age I Current disk usage +in +Sorry my mistake +Chapter The warning at the end of the chapter in the guide should be added to the that this option should be used with care because the entire service will be removed including all references +Are you sure the cause isnt something else thats just with having applied the patch? the OOM may not be related while Im not sure about the NPE now checking again +At least the relevant section matches my recollection of the nice description and justification for the shuffle algorithm in the original link +What the hell were already labeling this ExperimentalShip it +Right but CLI is not up to date on NPM because I dont have the proper credentials +Oh I see that the changes to can be reverted as now picks that codec at random +I just committed this +bulk cleanup of Jira versioning +depends on reducer of the MR Job + +Ive removed the logger field from +As Adriano asked me here is the first patch ready to be committed +Looks good +I should have also mentioned that your change works for me +The patch looks very good +Remove leading more duplication +See comments in this ticket. +Obviously this is a very strange state for a user to put the new action class in but I can see it happening by accident fairly regularly +Would you be filing a patch changing that in the maven conf +I also like the idea of deploying adaptors as S applications and therefore benefiting from the programming model and distributed nature of S appsBut I wonder how to partition and inject the data to be injected in that design +ummmm? Yeah +So for Edisons test case after we detach a data disk and then attach the same data disk out of the many data disks it is due to the hotplug is not supported on the VM +Didnt apply the chinese one as its not a patch +Then simply type antThe app runs at httplocalhostkitapdemo are not going to correct your bugs for you every time you post a zip of your entire application and mark it critical +great work Uwe +And please pretty please can you createnew bug reports for the last three bugspatches +Thanks! +Yet on main well either need a dependency on or copy over the Base class from the postgis module +the patch lacks at least one class patch as b missing d like to use this patch but looks like it doesnt apply for which we are a quick glance Im quite surprise that doesnt reuse element so it doesnt benefit from the can you shed some light into why this is the case +I suggest removing them because three is of no practical use and unlike zero may give the impression that has been localized for Frisian when in fact it has notWhat am I missing? How are they usefulCheersPaul agree with Paul its better to remove support then have a partial one especially when the partial is very close to zero +I will commit it +Looks fine to me go a head and apply it +I take that back +patch to is the to trunk +Ill log at least one new issue tracking the final piece of the functionality required to insert the required meta data for Clobs inserted with one of the JDBC overridesI dont think this issue requires a release note but Ill comment on a few things before I close the issue +Adam Hmmm this seems to be some fallout from the ANTLR hack we used in to avoid making ROLE into a keyword +yes thats what I was thanking we could still queue up the compaction request on the region server but only start the compactions once the master leaves safe mode +I will be committing this change to in a few in current SVN +Attaching another patch taking Daryns comments and also enabling position based testing in synced up to the patch to the trunk again +The configuration change was made in the Javascript part of the sample application. +Do you know the real use case for thatThe header is set according to the information in the configured WSDL so I dont see any reason to even allow overriding it +There may be others that the application I was running did not exercise +If I had to guess I think the problem is in the way the TLD is written? I only say this because Robs tests are working +We cant have a bug over something not implemented +We keep as default for the time being and push the RC with this includedIt is a real nice addition to the release candidate +Resolved in commit ecdecadeefdfffccaefcd +The test failure has been reported in and +oldlogs is filled with logs that are month old +Patch applied in r thanks Nelson +Chris adding you as watcher to this issue Let me know if there are any plans to have time to make a more embeddable rendering + +Attaching a reproducer and configuration file +This sounds like a good compromise to me +The locking issue will be solved separately as. +What URL can we use with? Thx +fixed with the patch in will have to help us here because as I said earlier the line of logs above doesnt really says much Do you have steps to reproduce? Are you sure it really hangs and its not just slow? Would you have the logs from all the nodes when a hanging happen maybe +Im attaching the test case + is a generic API where we take input of various types +Im not sure how to help people avoid this problem when deploying other apps +There has been a bit of a build up of JIRA tickets but we should get round to reviewing this soon although it may not be by your March th deadline +It is not a very frequent operation +Patch applied at r thanks for your contribution please verify +Wed have to uncomment the deps in the POM uncomment code in or different classes and then rebuild the plugin +Good luck! Leo +Michael Can you take a look at this? Is it a GUI issue? was in Guvnor +But I really have no idea about how to do this +the workaround works but the changelog plugin is quite useless if I restructure SVN changelog needs to be resetted that sounds bad and ugly. +Where is the attached screenshot? +I think the issue can be closed +Great work TriThis monolingual English speaker is very impressed by your efforts +I will review your patch +We just need to find out what is causing it +Resolved +The bean source service generated WSDL +Brandon sorry was just trying to say theres got to be some mechanism to do that and hadnt looked deeply into whether macros could do it yet +rebased patch +I would wait until the situation settles down before adopting it +Ill have to check that it wont cause ambiguities with the for loop though +Moved from SEC after discussion with Juergen. +Attached patch +you can use the characters ofas an sure the case is sent with correct characters +OK ignore that +Looks like it references notification instead of capture +I would expect the Seam CRUD generator to generate whose fields are in the same order as the related columns in the related database tableHowever if you start with annotated EJB entities and generate a database schema the columns in the generated tables are not in the same order as the fields in the annotated EJB entities and therefore if you generate a Seam CRUD application from the schema the JSP fields are in the wrong order +The new patch was my by svn diffI made the change to the if we code it the other way round as it is faster to understand +This feature has been implemented inPete if its still doesnt work then please reopen it as a bug with step to reproduce the issue +lets see how far I get +Also I dont assume that this crazy expert syntax is a termination stage for the this issue +The approach is detailed in. +Ive made copycut operation according to the screenshot the operation passed fineThus I have two questions for you What content did you get? I can see on the screenshot that in source you have all the text in span is selected but in visual only two letters +Stacktrace from error log is pasted in the description section +As I understand it a possible disadvantage is increased network traffic and bandwidth +However the patch looks as if it covers the corner cases +forgot to mention we should push this one one and then do MR as a separate JIRA +This is no longer possible +Did you adjust something in the config +Reproduced the example above and the hostname passed to the is picked up as expected +I dont see much of a reason to change anything at this point except to rip it out and amybe make the +Andrey Im sorry but it seems that there should be done more work hereThe file modulesawtsrctestimplbootorgapacheharmonyawttestsimage was intended to be at modulesawtsrctestapijavaorgapacheharmonyawttestsimage since it was implementation independent +irrespective of whether Add button is clicked or Edit button is clicked +Ive been testing with branch not TRUNKRetesting with a fresh checkout plus only applied looks good Ted + the same patch with ASF licence +Adding jiras for hadoop +Cant think of any good specific testcase for testing this +from me +Is it worth having a separatetest case which fires the trigger by way of an UPDATE statement ratherthan a DELETE statement to demonstrate the code path via +Hopefully the next patch should be the final one +And now for another episode of damned if I do damned if I dont +Please close this issue if satisfied thanks. +I understand that it counts as a failure after attempts +bq +Hi Todd do you mind if i ask your opinion on the potential impact of this issue and possible mitigation for it? So far weve not been able to reproduce the issue either in the nightly regression testing or in specific tests +Can this one be closed now +I guess this is just a convention +Ill close this report please if it. +Committed this to trunk and +Ok +wars this is the case whether or not the ESB plugin is installed in the JON is with the embedded server in IR +This is just as bad or even worse as adding the methods directly to +Should we rename it to EAP instead? Problem is most community users dont know what EAP means +It seems refactored the code +Its of a bulk update all resolutions changed to done please review history to original resolution type +Rev . +Manik I have noticed that you have changed the synch on loader only so for the get that is will return right away +sorry I should have thought of this +Fix attached +For the new test could you add ORDER BY for determinism? I know that the GROUP BY does it implicitly currently but we shouldnt rely on that +I cant see a reason why this would be controversial so Ill start with a patch for that +A new task will be created to apply level change in which the algorithm editor will be replaced by generic algorithm control viewer +and Eriks as wellWhats Eriks reason? Hes for includingbq +Great! Ill close the issue then +Please click edit this issue and fill in the necessary information especially desired JIRA key notifications email address and JIRA admin for this project +I am not sure is it possible to cause some channel is still leaking without close then block Nettyrelease again +I think scanner caching would take care of that +Okay I have added this to the Release NoteNoteField has become property to avoid direct field injection as Red Hat prefers you touse setter method invocation instead +I have addressed comments from Jitendra in the new patch + attached patch +the patch looks good to me +Can you take a look at the use case in? It looks related maybe it or one of these can be closed as dup +closed now +I understand its not clean but the actual behavior of nutch isnt correct either +I dont think the bug is in the comparator but possibly in the implementation +Thanks Milosz I hadnt checked trunk yet and I missed Jeremys change for +A proposal for the changes will be prepared soon +Ive tried to split out the stuff for the first part and attached it in this +Are there any plans of supporting non http operations? If not I dont see any benefit of having a webhdfs scheme +Updated the patch to trunk +Did you set the view definition via set +I dont think that mounting sitemap from locations other than filesystem is not the critical functionality of CocoonWhat do you think +its a little tough for me to follow the patch youve attached since it appears to both introduce tabs instead of spaces and because it appears to make whitespace changes on lines that are otherwise unchanged +patch committed +To make it very clear I am on removing as a part of this jira +Additional infothe infinite loop of testIOLoopUTF is due to the fact that the buffer size is set to while for encoding each input character requires at least bytes +The generated REST WAR now support query parameters in addition to URI parameters and XMLJSON inputs +Thanks for the patch Karthik and for the review Daryn. +what do you mean it does not work? The property was never removed +Fixed the unit tests that were failing due to the switch over to using s init method +Thanks Owen! + Should jbpm be in BOTH categories connectorsjbpm should be only in the brms category +I see where youre coming from +Isnt this something that we could do in +The real server adapter is implemented in on +patch uploaded to +The Ant Plugin should create a that requires only Ant to run. +Ive made you a contributor so you can be assigned to issues +Forgot to add YARN jars for the sake of JHS +Verified +jTDS was faster but did not returnall the tables +Ok thanks for the update +Never mind Hugo +a create a packageb create a package c move to the package created in step b +The does not include and it is not possible to configure this parameter +Which test framework and version are you using +with Doug comments Remove Xnutch to constants names Add prefix to constants values Move constants definitions to a interfaceI take it +I confirmed that this gets rid of the compiler warningIm going to commit this momentarily +Hopefully splatted by something else that works. +Marking closed. +It would be better to create a separate Jira for that and to discuss the patch there +I see +This is one is costly in maintenanceIn the spec file There is a macro to handle prepost un +I have also the same problem with Times New Roman +start with the simple one then I know that your are so busy with Sandesha and sorry for asking about the prograss +I think the latter is enough +class file content in the jar and not just the list of files in the jar +bq +Confirming this is no longer an issue +Unfortunately yes +I was surprised to see that this high priority is still not committed for +thanks for the information +Also if each action in the chain do their own addMessageaddError things might need to be handle differentlyIMHO we might want to give further thought on this issue before we implement the action chaining mechanism +HTML file included for review +It is in a negative q +I used MVEL version. +If not Ill reassign it to someone else +Solved with Its not solved it is still there +Ann and Jacob as I dont manage to reproduce this issue with Sonar here is my proposal Im closing this ticket with resolution Cannot reproduce and were going to fix in Sonar +Fixed +Verified by Oliver. +Now working on gettting it all to compile +Use this one. +Steps to repro shouldnt have ambiguity in them +The exception was noticed once again with IBM Diff file derbyallderbynetclientmatsderbynetmatsderbynetmats Start datanet derbynetmatsderbynetmats a Exception in thread Test Failed +Thanks Yi Xiao for the patches +Hi Andrea as just commented on the mailing list Id prefer to hack on the current wfs module and taking care of both geoserver and udig trunk builds continually rather than at the end of the process +And this in turn is used by the implementation to save in page context +This is a very logical and common use case forms are constantly getting reRendered based on an Ajax response +NEW patch added more transitions in and +Attached the patch for the details changed hashmap to return rather than null if key is not found added a sequence to invoke scan when repo is added or modified create the directory if it does not exist to have an empty statistics during thanks! +But Im sure I can figure out a solution to thatThe anchor approach will work with all connectors but it sounds like the web file system and CIFS connector are of the most interest +this is not limited to create server same goes for remote box in the openshift wizard when its an existing projectshuoldnt have to be implemented twicedifferently should it +Please contact the administrator +Sorry I dont know how I got to the xmlbeans forum +Thanks Vinod +Tests passed cleanly for me on +Added a target to create build directories and then made compile depend on it +Now config files are definedIm reviewing the unit tests and deployment on J +Checked in a partial solution to this issue +Using this decorator will allow you to overload decorateSQL method in which you can put all the create tableviewfunction etc sqls +This means they should be avoided in cases where the exclusion logic can change throughout the request lifecycleIn fact appropriate use of c tags can be a real performance benefit +Thanks a lot Emmanuel +m puzzled +Committed to trunk. +Putting this aside for a while until finish moving coprocessors over to protobuf +Sorry I was busy last few days and did not got a chance to update hereSo here is the concern about thisLets imagine your have a row with an ID describing a unique identifier for your use case +can delete the directory when it is shutdown +Tested it on nodes the UI graphs and metrics show up all fine +The patch fixes disabling local execution of TI to be disabling event reportingThe patch also adds check if TI needs to report particular events before calling to jvmtisend event reporting functions +This patch is for +Sure +Disallowing map joins on hive server seems bad because of the performance penalty +Maria this is one of the biggest features I think for an upcoming release +Assume closed as resolved and released +Manuel do you still have the issue +fixed in trunk +Sorrythe above three commits should be associated with +corefIn other parts we have good junit test coverage there it would be safe to apply your changes +bq +The failover LB in Camel now supports round robin +Mind if I hack on your site to fix your download script +removed patched files from my view and did cvs update to getcurrent version and did make distclean then all the stuff to buildeverything including make install and rebuilt my test app +I reopened +I am conducting a few tests in comming days +Can we show a webpage pointing to the adobe licensing page in a HTML component? +Should VAT be separated from Sales Tax or should handling of indirect taxes be more general and flexible? If in the end the problem is that Sales Tax entities and accounts are poorly named but they do everything that is required the name is not to big a problem. +this is new thing +The patch is simple enough if Jenkins comes back OK I am a on it +Anyway thanks! Can you link the PR please +I think its a pretty often used case that the form calls a page with the values of the form as parameters +i tested the quickstart on trunk and it worked. +Tested a number of different wsdls with and it works perfectly +I have revised the description to apply more accurately to the statement of are a few SPI related issues at the moment +Updated the patch with missing files +Thanks Sharad! +Upon reopening the connection it tries once and if that fails then the call fails +Closing. +The issue appears to be that the tests expect the initial app event type to be and since with jdk the order of tests in random the last test test is sometimes being called before the others and it doesnt reset the type back to +Actually is about the server modules test client so not really relevant here +Fixed up my patch +Thanks. +That is the right command +Hi Jrg Would you please do a uniform diff against current CVS version instead your local copy? I have a problem applying this patch +Looking at the current version of in the trunk does this fix need to be applied to and as well? +bq +Committed revision +But it might be an initial symptomI believe ulimit n is per user not per process +marking this bug as invalid until the submitter decides to respond +Change made +Should be fixed with snapshot deployed +If there is a problem please open a new jira and link this one with that +Sorry for the delay +However the users who are requesting connections over Admin channel are subjected to same domain authentication as the JOPR does thus Teiid enforces the same security as JOPR connection +Marking closed. +Allow me to confirmreject this bug by running target in the distribution install script +If you need any further info from me on it or testing let me know +I had assumed Derby already supported this behavior and only discovered yesterday that its is a new feature and has only made it into the branch +But I fail to see where it gets put into itDoing the fix now +My proposal is still that we remove the call to takeSnapshot in loadData +My patch was at level and it seems that its difficult to apply since the svn location has been moved +The javadoc tool appears to have generated warning messagesThis is not this patch +I like have a couple proposals +Sorry I meant to say I filed +The suggested patch file +I would also like to get this committed into the trunk so that that we can work on the secondary JIRAS related to this patch +Committed +Welcome back JonI do not think it is question about fair vs +I guess that there a redirect that doesnt work +Im not sure why we have the old release notes files hanging around +Hope this helps +Thanks for reviewing +Let us not document itWe can add the config property with the default value reporter +I have some insider information on one thing that you can get wrong in reviewing code +Hi EvaI have removed this out of your queue because Amazon EC may not make it into EAPCheers we describe and profiles in getting started guide +Had forgotten changes to +The two are exclusive of course +This was done as a part of the Google Summer of Code program +It would be a request for enhancement to or a seperate bug to restore this is not a violation of any jee spec to not include the ejb interfaces in the war +In fact those deleted source files are correctly purged from the Sonar DB. +thanks again for the reveiw markkonstantin +The Yarn could be disabled by default as well +Moving these out to please move back if you plan on working them in the next few weeks +The integration test should be run from the root of the project with the following commandant f integration optjbossassrcBranchxbuildoutput to main with revision . +Attaching patch a which is a suggestion on how to fix the problemThe test helper class now adjusts the expected counts for statistics entries based on the version of the old database + +Todds patch is not smallNow that interface has been introduced rebase is needed +IMO should return a Table object for the results set instead of this new object +I have patches ready for trunk also +dealt with on trunk. + +Fixed typo +and found it to be quite an enjoyable exercise +Thanks Mike +I am trying to access this SOAP interface with a grails application hence using +ThxBest regards solution works for me as well +My mistake yes I had in mind +Sorry left off on the git one more time +I have to creat xmlns attributes in the root element otherwise they are all over the XML file enlarging the file unnecessaryWill fix this soon and add regression test so this will not happen again +What you are doing sounds fine +bq +assign back to me if there is still a bug +All the regions can use the same zk namespace managerWhat does it do in the RS? Why is it neededAlso asked How often do we do this getThanks +Ill close this issue as resolved if nobody complains in incomming days. +AFAIK users should not set the fix version this is left for the developer +textDump of file File Type DLL Image has the following dependencies Summary +Committed +were for? Have you tried those +please adjust the priority to indicate that the Jira should be considered for CR +This would allow writes to be collected and flushed simultaneously thus saving rewrites of the upper nodes as well as having transparent edits to the tree that will be flushed some time in the futureThen for parallel writers we allow them to traverse the tree and prepare writes that can be merged into the current memory state and eventually flushed +I fixed this in trunk w a hack of overriding the problematic executions from the pom with executions +That way we prevent the AM from job output or changing the final job status after notifying the client +used internally in Struts Core renamed toPlease review the changes but I think it should be postponed till as it will break backward compatibilityAnyway support for the Convention Plugin should be introduced ASAP +I think I agree you provide a link to the original discussion? I just cannot see how this behavior can be tolerated +Thanks Eric I just committed this. +Thanks for pointing out +Do you have any ideas which can help me add some quick tests for this? The current test executes though the code checking that it does not result in any Exception or Error +It only removes the rootdir from the path +a single line of code that made the difference Everybody putting an effort on improving the project is a contributor so YES you ARE a contributorLooking forward to apply more patches from you +See srcitprojectsmultimodule for an example. +r contains the fix +This will fail if you do not apply this patch +is not a fork of both evolved independently and are quite similar because the spec is the sameSometimes someone update a javax +As the data log segment naming style has changed from +studioeclipseplugins +The Request which is called for showing the variants is a normal controller call +It seems that the test specifically verifies that during a restart the configuration is only stopped and started not unloaded loaded or failed +It looks like I do not have the karma to close the issue myself yet +Then verifies that a client can automatically accesses the remaining valid replica inspite of the following types of simulated errors +Here is a screen shot of the error that appears in the browser when attempting to view the WSDL +It seemed to me when reading that there is some difficulty in making a single implementation work well in both situations +Attached is which I think is the correct patchfor the branchMy build environment doesnt seem to be working rightnow Im getting dozens of failures in the test suites +committed to trunk andThanks for pointing this out Ill piggyback on the next commit I do to fix the typo +bq +Otherwise changes will apply to the next versionId be also grateful if you post your ideas on the ranged versions +Rebased into two patches +ant test passed after the patch has been applied so I think this change wont affect existing code +Would anyone be willing to beta test when its complete + +We have a similar data model to what Koen indicated we have objects that represent a tree that in some cases becomes obsolete and must be +Closed after release. +I think he means he ran the new tests without the fix to and it failed as expected but passes with the patched as desired +harsh the timeout should be an oozie server config property not a job property +Removed the double from xdocsreferencedevelopers that I have found but have not been able to fix In the section Remote Repository Layout of xdocsreference the directory tree example is no longer accurate because jsse is not available at ibiblio +Basically the Project Examples framework takes an exported Eclipse project and wraps it up with some XML that exposes it to the users +However it maybe better to express our dependency for package management system to install Java for novice user +In addition contains the same file as with the same incorrect claims as mentioned in above +Its compilation with a different groovy version seems like the issue to me now +looks good to me +latest version tested with to with some of my suggestions partially to Steves commentsfixed test casesfixed formatting issuesfixed issuesfixedIve pulled this down and am running the tests +or just provide appropriate controls to the user with reasonable defaultsIN any event I dont think it is a given that compression of hot data will always be inefficient in all codecs for all hardware for all users at all times +Ill finish this and send a pull request tomorrow +You had the test for trueLikewise I adjusted the unit test in to extend the unit test from +Moving to wont fix as its superceded by namespaces. +If a getterMethod is found for the property it uses the decorated value for sorting and not the underlying object +Any preference whether the benchmark results are from a machine with SSD vs not? How much heap size do you typically configure for +then we can make a tool for doing this job as a work flow +I have applied the changes to trunk at svn revision +I cant demonstrate the problem to myself +Tom is right r newlines are also used by recent versions of Excel for MacI am attaching here a simplified version of the patch I have tested it and it works fine +Attached patch replaces the incorrect example with a working one +from me is wrong corrected the group +Could you provide the log when this issue happened +With XML Spring config does notexist in the Java process since there are no references to it +After task executions are done all the attempt directories are changed to and owned by +Ive not checked +This patch simply replaces maven site plugin with maven site plugin +Id like to find out the maximum value of the score and of a couple separate function queries and then after use those metrics in a relevancy formula to sort the results on +reopen to be able to target correct fix version. + +Thanks JihoonIve committed now. +It avoids browsing all the plugins to get status of purchased licenses +I propose making the providers load from the current threads context class loader as described in Peter Jones comment above +Ive released and pushed it to central +This is the fixMeta option in the hbck improvements. +Most of simple logic are refactoredFolders docs itest tutorialdo not contain warningsIn framework subsystem we have issue only with Synthetic PKAnd in modeler we need to think about common logic of UI changing +unzip the +Though its kind of awesome to have such a broad ticket +Fixed in CVSThanks +However I dont think we agree on the fix abdera thing so Ill comment on that next +It is not clear to me why you think the framework is restarting a bundle +Then well remove the old one in time +WSDL Java is performed using Suns resolved issues for already released versions +You dont care if the implementation is pure Python a C library wrapper or whatever you just care about what parameters to give it +I dont want to force it this is an internal hook so you know what you are passing you can cast it when you get it +Will update wiki after it is committed +has deprecated in favor of and the latter one doesnt come with a +i didnt realize it would create the exact same issue without allowing me to edit it firstyou can ignore the description and mark this as a duplicate of you can also use my patch for its really trivial. +bulk move of all unresolved issues from to +still may conflict with something certainly understand and respect what youre trying to accomplish but the sad fact is that if this suggested fix is implemented it will break of the artifacts in my company and will mean we have to do on the inhouse things weve done that depend on artifact id cvs module name and project name all being the same +committed +If the change requires a SNAPSHOT version of the plugin then we have to check if we want it in already +switched to lein and its working ok here +Thanks for the confirmation. +Thanks! +looks good will commit if the tests + +Any workaround +of has a first try of switching to conjunction mode see advance and requiredDisis in thereI have not even compiled it yet just showing the basic idea no more time at the moment +I have a working fix for this I was just exploring different possibilities because my solution would wind up calling the default method at least once any render it was accessed even if the underlying value didnt change and I wasnt happy with that +Preliminary patch without agent changes + +I tried to reproduce this in to environment but i dont find the issue in that setup +So if I resolve things on the fly I thought it will take so long considering there should be a lot of active process do you thinkI can share the model code if needed +We can introduce and default to hbase +I believe what you are asking for is an alternative classloader that fetches from some other type of storage +It is not very clean but it is the best we can do without impacting wagon and you look at this code you will see it is currently not easy for a plugin to use the proxy +Changed Target Version to upon release of +The and are only trivially modified from Andruss pet store example +You are correct +A simple and straightforward combination of Ant and Ivy will do fine for that +I should mention that passes of the Acid TestsMojarra passes but they have recently done some fixes in their nightly builds so that now they pass Id be hugely grateful if you could download it and at least confirm you can reproduce the problem +testAutomaticallynew persistent class jpa name +bulk close for the release +As I said before Im somewhat reluctant to do this through a forward because it would change filter logic or whatnot +I believe that needs to be What shouldnt matter is the ownership of those dirs once all of the permissions are +It seems that he property can be defined after its usage +Thanks I also had someone else test on Android as well. +ThilinaCan we close this issue now? is it still a problemthanksIIRC I fixed this issue some time back +There are patches here for and to correct the order of the calls to getWriter and a script I havent had a problem overriding the contentType as long its done before the script prints anything to out +Do we really want it to be able to be set at the composite level as well? One place to set this so its not too confusing would be my recommendation +Postponed to next release +has been marked as a duplicate of this bug +Rini this might be included with Niels changes +makes it impossible to use a local +That helped +I think checking the length of the file is important especially when you copy a big file and that the write isnt atomic enough +Resolved in cbf tags are updated as well. +Given this Id say drop interpreter +Good to have forgotten password feature with email password guysI really like this feature request and really like Roo how could I help with development of this feature? Or is it only developed internally +Attachment has been added with description This is the base has been added with description This is the to parse the xml to there been any findings or updates to this bug? I am running into the same problem and was wondering if there will be a fix for thisThanks transition resolvedfixed bugs to closedfixed +I shouldnt remove all data file and recreate a new node with minor version upgrade +up to date patch with Robins most recent Vectorizer commits merged in +And corresponding test cases +Ok understood I disabled the cluster event broadcasting in the as it creates an infinite loop in the which consume a lot of CPUI have to find a way to check where the update comes from in the event handler +Hey Rick Im not seeing any +Youre saying +Like XMLBEANS or OPENJPA +I will commit this shortly +These sorts of discussions are a waste of time effort and community good will and only serve to confuse and alienate people +Theres not much added value here since instances and images are only loaded when requested by expanding tree items +Is it possible for you to attach the WSDL file that you are using hereIt would be helpful to solve the problem +OK I get it +Im concerned about the costs introduced for building family maps just for audit logging +I am not able to try to reproduce the issue because I dont know where EPP is placed +Fixed in +I think I actually hit the same issue again +The mapred contrib tests passed with this patch and with a jar that avoids problems causing +However today when you mentioned you wanted to commit it it surprised me so i spoke upFair enough +I do not know if spec is incorrect or the runtimeP +Since these changes were in trunkdocs directory thought I should include them in the patch +Create new Entity through the in JBT +They are all too long +This was a piece of code contributed by an external committer so if theres any blatant duplication of GPL code we should get it out of there +With this patch both and pass cleanly +Ignore my last comment +Julian whats the point to remove it now if it will be needed later onWhats the point in keep unused code in the project when it can be when its needed +This issue can be reopened should we wish to include it within a future release of Nutch which includes a web application +Thanks Amareshwari! +Ahmad your patch once again contains TAB characters and there are quite some lines with wrong indention +I think that is good enough +Our CI server will keep running them against b whereas they wouldnt be run on locally installed JDK builds in order to improve the contributor experience for the time for the Eclipse compiler thats just a build dependency for the compiler +Import with profile! +I agree that its better to add the properties setting into the constructor of +In the future could you please add a little note telling me what part changed? It took me minutes and the help of Tareq to spot the diff between the first and second snippet +Thank you guys +I believe Snappy has slightly higher compression but I dont see a reason to use it +To determine what the issue was I marshalled the Generated Object to a xml file +Added annotation to testThanx Ted +Will there be a fix for the problem anytime soon +fix version needs to be changed +bq +Initially we provide a for we plan to provide a corresponding adapter later specific reasons for postponing support? seems quite light on new features and is certainly in wide use these days +It can stuff into that field whatever combination of fields it added to the document already +Attaching an alternative patch which puts the trademark verbiage in the footer for all site pages +Does this change still work on Solaris? I noticed the makefile being removed might be removing a issue too +Patch is not complete +Is there a reason we have to list them in an additional JIRA comment as well +revision +Both regular test suite and security test suite are green for me +This is pretty critical for us in Cargo land and I guess for other projects as we dont have any safeguards anymore +Second comment whatever we do I think the most important is that it is not hidden from the user in any way shape or formLike swallowing it and carrying on as if nothing happened +Ill look into this issue +What is it that is performing the grouped aggregation? Does the aggregate itself do this as the syntax suggests? Or is it a query that does it or something else +According to the most recent comment this issue appears to be solved on the user end +This is pathetic however everything that goes wrong with an operation is returned in the LDAP response as a stack trace so that helps at least +Ok I can adjust the fix for that +All set +Thanks Arun Devaraj! +All tests passed. +add can be called at start of page when no HTML has loaded thus get returns null and the event will not be registeredBest option seems which will be solved as part of issue with option and have rolledback partial fix +corrected patch +Found the reason and solution for the above linker error in +From the description it does not seem like it causes bad things to happen +I suppose if I had a job where I expected it to have nothing to do frequently Id want to be able to catch this exception and return success +I just did svn up patch svn commit and got merge conflicts from you applying the same patch +The latest patch works good +Assigned to Remy for schema not sure who is in charge of the schema though +May close at end of cycle if still unreproducible +My goal of removing all the loops from my previous attempt at a rewrite make me believe that theyre all interdependent and that getting rid of loopwalking is going to need some pretty fundamental strategic changes to the way we examine and process status information + +Merged to internal build +sounds a bit too loose +At least no more NPE will be thrown. +Final patch +And yes Hoop supports all readwritemetadata FS operations +Fixed in our latest jbossas builds marking this as resolved +I cannot do that since such a property does not exist in components +Akash I guess you have the +I dont see any problems with the regular expressions that generate a match on the provided example +It prints all the field names of the offending class. +tableinfo created in HDFS +Overlap removed and Apache license to +Verified on +This archive contains subdirectories as per project layout for scratch in attached patch. +I dont think its safe for s move it to +Chatra Could you please find a spot in the documentation for the configuration mentioned in the commentsthanksdims +Reopened this issue +A receiver callback in +This is going into the code base in one form or another so why not tag or branch trunk and apply thisIf thats too radical how about committing this to a branch so those who want to work with it can do so +Ive just committed this. +Both Symphony and Open Office cant know left and right so the default value start is set +Hi Anil whats the status of this +Im providing this because I got permission to it and people want this capability +I can confirm this is also happening in one of our production apps which has a file download servlet taking advantage of web we have sessions out there right now that were supposed to expire over days ago +Acctually I have a docs patch for this +Exposed methods to register one or multiple as per spec +I dont mean to be rude but I think the number of successes indicates that this is not a Thrift bug +I was surprised to see a blank page when upgrading from that works with Tiles to that doesnt work with Tiles +What you gainlearn from this can be applies to as well + eaffbfeafbafbeeb commit in +No tests because its just adding log messages +So I will not commit your change which is specific as using an EAR into the Application server rather than using the embedded Tomcat like most people do +Ive already started tests with my merge of so Ill keep with that unless theres added benefit to patch +Closing stale resolved issues +Im not sure what is supposed to do +Pls check the lacks a test case +Applied. +Are they not exposed either +Thanks Gunther +Please provide more details +No worries Michal +I distinctly remember having this confusion and Ive seen it go round multiple times on the users list +This broke since its now getting trying to close the proxy +Otherwise looks goodThanks +Output and modification of the Ivy dependence example project +In principal any EE archive can contain descriptors targeted at different application serversThe urnjboss namespace is clearly targeted for AS +As I recall delegated everything to its right? So if the is working was just supposed to echo what it saw +Thats it +It also causes a new warning from javadoc which needs to be fixed +Write hidden flow field render links with key eventId render submit button with the correct eventId name and optional hidden flow interested I can share the code +thanks! +Code looks good +Hum Ive tried to upgrade the jetty dep and plugin to and while its working its also quite a lot slower than rc jsp compilation is visible again +Too late for pushing for +I think the right syntax is a little different and Ive attached a modified version of your starting file +If you see the exact same problem again please reopen and add the details. +In other words I am forced to manually configure the classpath for a new project before using Update source folders +bq +Thanks for the fix +Will wait before closing to see if this failure mode is actually fixed by this patch +fixes this issues by supported array types +A quick comment on whether or not these poms will be uploaded or if there is something I can do to help speed up the process would be much + tested! +is it possible? If not I think I need the config xml +The issue here was that the maxValue constraint for the annotation was not in the proper date format +Sorting can be disabled by setting the number of reduce tasks to zero so that map outputs are written directly +Other fields are not +Included in release. +I like that approach +Do you agree with this approach or do you have another suggestion +Could you also close this issue? +Thanks +If nothing else we should see a regular Ruby exception now which may give us more information +Uwe is right EFC has been around since we should not delete it but deprecate itEFCs extra stuff should be moved to +Is this issue resolved +can you upload a new patch +LarsYes i was just checking that part of the issue that you have saidI was first thinking not to throw an error if empty key is passed but considering the other case where an empty key is passed along with other keys then ideally that should be the choice +Please +RTC work item defer of items from to +Ive discussed the issue with Bill and analyzed the bugs in TJWS and then sent the issue to TJWS community hope they are still active and will response soon +patch is clear from Im gonna commit this one today unless objections +Amazingly the Xalan conf test harness kept going +You can also do the same layout of homepage but at the bottom of a projet page. +Sorry this should be against of the Code coverage not +ah so then its treating your options as I want replicas in datacenter replicationfactor but all your nodes are in datacenterI thought we had logic to disallow naming a dc that to prevent this confusion but apparently not +Bug has been marked as a duplicate of this bug +Patch for the described applied in the current SVN sources +Attachment has been added with description FO to reproduce bug +Do you see any areas where there is lack of consensus +Sorry for duplication. +Patch looks good +Good one Lars +I applied the patch with minor modifications please close if satisfied +This is a major change and I need this before doing minor +I have made the patch based on rev dd +Thatd be useful +r thanks! +This was the runtime gets unaffected if thats the problem with introduction of new logging and legacy controllers would be useful too +With v patch derbyall ran with no new failures +Note that this patch can only be applied to after the is applied. +Hi LukeWhat is this parent JIRA currently waiting on post Metrics? Can we resolve it? If not let us what it is waiting on so work can continue? Is it just the deprecation of Metrics that awaits +Theyre sooooooo slow +If we run w single thread the build will take forever +I believe that both parts of this bug have been addressed now +It works with nightlyThanks. +Also dont forget migration the version property must be set to the current id constraints moeten gedropped JBPMPROPERTIES and schema version property used to be a separate issue +This change is to mainly to get feedback +We are not certifying and Anil actually needs our CR build to create a policy +Btw it would be useful if you could run jstack against the blocked so that we can see the status of all the with all threads +Issue mentioned in the bug is tracked part of. +Thanks Prasad +Ignore my last comment as well +I still see old error messages +attach my ve just committed this. +Fixed +This is the only thing we needDo you think we can arrage a packaging of Hama with CDH +couldnt we just include our own mock implementation of the interface in question? As long as it has the same methods it will work just as well +Incorporated comments please review +Agreed +Would still allow rolling upgrades from a prio version of +Ill try out the FACESSERVLETMAPPINGINDEX and FACESSERVLETNAME to test them out from the descriptions it sounds like both will also solve thisThanksKen +Committed revision +Tested with with patch applied +Now that Struts is out whats the best way to advance this? Im happy to take a stab at an XML schema for to get the ball rolling but dont want to waste my time if the committers are planning to go in a different direction or already have thoughts on what such a schema should look like +Hi Milosz I remember the conversation +Thank you for reporting the issue but this is being closed as since the smooks component is not currently being developed and have been removed for year +I agree but only to a point +I will commit this soon +Fixed the soapenvException problem +I checked the th patch above just to see if it fixed my problems but it did not here a little update on what i am seeingThe flushing happens correctly the first flush of the server once it startedthen it flushes as fast as it can over and overbut if I stop the job the flushes stop at the same time with in a secBut once a job starts again the flushes start almost at the same time its like the flush size has been reset to or something after the first flush +Hmm I dont think I have the code anymore that caused the will try to find if I can reproduce the issue using the provided project +Is the New Server Wizard intended to work as it did before? just encountered on the list of fixed issues using version that I just installed +Add the following features new command aliases that prints the list of defined aliases +Would need support the tooAlso it would be nice to get some numbers about much latency is increased +Bulk close for +Is it a common practice to store data under tmp for unit tests? Might lead to problems if different developers share the same physical machine for testingNo tmp is used if is undefined +fixed in rev +Requirement is changed to control single part or multi part upload based on object size in the meantime support or +I volunteer to help get patches applied +Maybe add an are you sure prompt or somethingSounds good +Mikhail You see Todds comments +Thx. +mine defaults to JDK at the moment and a little later to +Patch to make JCS source problem is still reproducible +Using a GPLd database is OK if it is accessed through an abstraction that is compatible with the projects license such as a JDBC driverOTOH Q says if the components licensing DOES affect the product then it can only be used in the specified mannerThere is some overlap as the JDBC driver example illustrates +The fix for that issue will address this one as well +What about an optional cluster nameIt can be omitted for performance reason if the user can guaranty separation of clusters +Im not sure if this patch is free but at least it works for me +The patch contains fixes to the bugs listed below found while implementing the XMI serialization a removed assert statements which were preventing proper reporting of exceptions when loading an annotator dll which is not found in the path and on call to getIndex with an invalid index name +In my particular case Im fine with disabling it for all applications as I dont want any new lines inserted in any of the DOM created by +I mean add the initParams in addition to the keyAlias +Maybe we can pick this back up when most platforms will have a compatible version. +Patch has numerous irrelevant changes this should be a diff +Is there any problem preventing this from being fixedAFAICS it is just a simple job of regenerating the +Resolving +Seems to not throw this message +When a directory is marked as being snapshottable check only then that that directory does not currently contain a filedirectory named +Sure created. +Satoor wrote The upgrade test Upgrade is not included in this patch because the patch submitted for first needs to go in and then the upgrade test will give the correct master diff +IT added +tested on trunk verified on revision +Thanks for the patch Matt +Wiped out the dir per suggestions which is cleaner +fails consistently for me +Mass closing all issues that have been in the resolved state for the last month without anyone commenting or reopening them +Bug has been marked as a duplicate of this bug +Done +Feedback welcome +In this way we are able to slow down client writing when dfs is not able to handle the amount of data in time +And speaking of which how would one update these feeds as clearly these documents are more live than the rest of the s a good question and one that Ive had since you mentioned that youd justoffer a single file +On the other hand I do want to see work on really start to settle and focus solely on bug fixesI agree that it would be a shame to not include a feature that we had code sitting around for such a long time +Thanks a lot! Maybe we should notlabel it performance optimization guide but rather something like bestpractices as not all issues deal with performance onlyBTW setting up a Wiki would be a nice thing so anybody could help improvingdocumentation +code adds additional checks but it is correct and is a good change +Just renamed issued to make it clearer. +Yes the URI is handcrafted at present which means that no illegal characters are being escaped +I would like to port this to +I think it may be OK to hardcode what we are verifying rather than writing generically +format this one looks ok +Yes +somedays. +Hey Benn I updated the when I made the switch +Not a requirement at allThink about this architectureMachine A User running Web client Machine B The Web server with JMX Bridge also where the UI resides +Happens since +Integrated to trunkThanks for the review +Done +Reopening as needs release note text +Is this a Spec feature or just an extra Weld feature? I think this is bad either way +This can affect performance during simulations because created a whole new local interpolation table every time it needed an interpolation even if its identical to the one that would have been created for another interpolation in the same job +I can confirm that the patch fixes the problem +yes thanks Im now member of incubator and it seems to work +Please let us know when you think this is ready for reviewmerge +Here is the patch the test works now with the DRLVMThis is really good newsIm going to try running Quake on Harmony this weekend using lwjglBTW AFAIK another popular driver JOGL uses to for direct access to some native AWT data +This can remain +In Slashy String improvement Proposal Option for Groovy is still not implemented yet and trailing backslashes make sense for windows paths and regular expressions +And same should be shown in Sonar see +This should be moved to the local logical to physical translator +Hi is it simple to apply the same to C builds? It might make sense to use the same build systems for both +You can update the existing savan sample totest the has been applies to Savan svn head +done but theres a bug in the streams svn auth group +Any chance you can test either or better trunk? I dont know that this has been fixed but I know theres been a number of fixes on the freelist code since +I think archiva creates an for every file that is being added in the repository +I applied the changes in your patch except for the changes in and which didnt seem to be related to the Visitor pattern changeThanks for contributing to Tuscany! +Probably will not fix implement this. +Thanks for fixing this +Youre right using the current mode of the source file for is betterIf the source file is created with no explicit mode the OS will use the users umask to set this files permissions and these default values will be kept when moving the file around whether by file sftp or packagingIf the source file is created with an explicit mode it will be kept even when moving it aroundAnd this behavior can be generalized for all transports and packaging mechanismsIt sounds very consistent to me +Most of the time during interop testing we had to adjust our clocks to the standard timesBut I guess we can provide a switch where IFF a timeskew is specified in the assertion the timestamp validation will take it into considerationItll be great if you can provide a patch +Resolved in both trunk and the branch. +Some other basic issues that would need to be addressed are the services API spec read and making sure the and other resource files are being read correctly +If sorting is required for an field then it shouldnt be mutliValued +I think solrj has a bug +I would name the utils package common or something instead since util might imply its for usersShould we also move the input output format implementations to an io package or to the lib package +Thanks Todd +Definitely agree +The problem is that creates a brand new instance for each started cache +Added appropriate error codes to all +Even if we had introduced some significant confusion in some topic I think the harm caused by that would be dwarfed by the harm caused by not being able to read the docs at all +fixed +sorry for the fix sounds reasonableThanks for fixing +Good thinking the only show when editor open I think is fine the palette expandingcollapsing might be annoying +Why we getting? Whats being thrown that is not part of the interface? Connection refused? Can we add thatAny chance of a test Ted? A unit test where we do some function on master with HBA and then we kill and start new master? See if HBA keeps going +Commented out dependencies to be more descriptive ans specify that these are tools dependencies to work in the beta after this gets into ibiblio +somehow missed that exceptionthere is an issue when the eventhandler is registered will take care of it + +comments on ve misunderstood join conditions +Initial patch will put it on review chance of posting that RB link +But does it need to be configurableShane +I committed the last patch plus some formatting +After applying the patch the issue is fixed +I dont believe it ever supported landscape assigning to Becky for it never supported landscape orientation but now it turns to landscape orientation if iPad is in landscape mode +attached new patch a smaller diff file without any of the generated code or include files +Ideally these classes will have all fields included in the specification on +test uses and includes a patch to it +Code that uses ispn to store indexes in a generic fashion so they can be queried via hibernatelucene integration into ispn +No it should stay in the first place +fixed upstream +theres no reason to touch +Changes Added Tests for Sorting on all trie type fields Return a for trie date types Added Tests for function queries on all trie type fields Upgraded Lucene jars to r Created new for through ant All tests passThis patch also contains changes for and +The above patch fixes the problem for me +For example in my workspace about Warnings +The challenge in this problem is to find a generic solution acceptable to all applications +Also someone choosing to upgrade from to is going to pick up new changes and even new features and most of them dont have a flag to disable them +I have to find some time to look into this +The data mapping objects could be serialized in a proper way and added to the installer even as XML +Any thoughts on this fix +I need to trigger the log to get a breakpoint in the broker +Thanks +Hii was on a trip for a few days and returned home only yestrday +Attached ZIP file contains entire code including test coderegardsSuneel +Lets just defer this issue to a later time since it has no impact +The findbugs warning is not related to this patch +This is a serious issue for users that utilize the tool +Erick can you take a look? Do you see some obvious problem + +Could you apply this PR? I will add the required features within +Done. +Given the recent fixes to oracle data store we have been unable to reproduce this problem +Bulk close for Solr +Requested test case +Sorry if this spams things however its unlikely that Ill work on these. +We no longer use the geronimo interop module. +And something that takes a List will accept a Vector anyway right +Different story for end usersm not flaming +This depends on their distance and other factors +Will post something tomorrowThanks Timothy +So this could be structured as a contrib module whose when a build option is specified downloads patches and compiles +Thank you Michael! Ill make some further small adjustments and add you as my mentor +Suresh are you satisfied with Marcelos responses to your comments? If so Ill go ahead and review this patch for commit +Candidate for SOA verify in SOA Platform docs +I think the best thing to do is just search the path +I think there were added a test for it last night but i cant find it add the moment +Thread local cache is used for string and byte buffers +The other formats work produce very little whitespace and now even have indenting in the HTML source +bundleKey went to bundleName and name went to bundlePath +Code looks good to me +Thank you Tim the patch was applied correctly +I managed to release a couple of plugins but not reliably +We can try and take the table description but not the column description +Its not actually a positive Linux detection though since its the defaultbq +I should have mentioned that I believe this will be an issue for servers as well +All other shortcuts that fail to be created in all langpacks have some unicode characters in their names +I made this change to indicate a desire to get this into if possible +Doh! Sorry I didnt realize normalization occurred prior to validation +Closed. +I started with a clean uped working copy ran mvn and got what I reported +I cant think of any finders in my app that are useful if unsorted so I will have to refactor all of them just to add an clause +Hi Jarek yes you are correct +I have committed this +Final improvements now done in. +shouldnt show up in release notes +We have a tier structure where each layer is built on top of the last +Hi Dogacanwe need this setting as we have the situation where we have a single host which has millions of files and it is impossible for a single machine to crawl it in a adequate amount of timeIn this case web politeness isnt an issue as we also own the site in question and we know it can handle the loadWe thought that other large sites might also run into this issue so we might it into a config you tried playing with option instead? If you set it to a fetcher can open more than connection to a machine +Id like to let you propose a patch with what you think is the best way to go as you seem to have some good knowledge of this featureThank you havent had a chance to review this +Its more conservative than what we currently have +h files to rampartinclude and +Only degradation in performance to due notification of a large list of listeners +I fixed the metadata in the Apache M repo a while back +The patch looks good +Well we need to tag in Ben or Rini on this one not many developers have a background in this code +Great writeup on the processing for ORDER BY +Adam +Ive encountered this kind of problem previously as well with my eclipse setup differs from mvn by not using +Then I started digging into why and what could be done about it +We could certainly fetch the value from the conf my point was not to insert this configuration parameter in the file +Unified all minimum version support to +I tried this again with the version as well as the trunk and I still cannot get the PDF to print properlyI attached a PDF document generated from Pentaho from which I still get the same resultAny ideasCheers fine after resolving +Thanks Jose +I hope it helps +Generally view handling aspects are out of scope of so closing as Wont fix +Any idea which java versions have this bug +Anyway +I wonder what the working group was thinking? I suspect they were trying to make it easier to distinguish from authorities and allow user info and port to be specified for registry based authoritiesThis particular issue is not listed in Appendix D of Modifications so I wonder if the working group noticed it +It is the serializers responsibility to ensure that when the DOM is written out to XML syntax any implicit namespace declarations are notices and written out as genuine attributesNote too that the serializer can take advantage of namespace declaration inheritance which is whats happening in your case +Wow that is fast! Works perfectly now +and thats interesting result +srctestjavaorgapachewssecuritystaxtestsaml +Theres a relatively simple system when you pull a SNAPSHOT build download the source from SVN immediately +Attachment has been added with description a sample file for handling missing transition resolvedfixed bugs to closedfixed +Ok thanks for looking into this +You shouldnt really be using the Expires header its soooo HTTP +collapse counters if not updated after a given time period +for the patch I committed it to branch +Thanks JonathanThe new fix looks good and I applied it to the ARCHIVE module together with a regression test at rPlease verify it works as you expected +Ive been playing around a bit and it seems like I have not broken anything bymoving the classes related to the source generator to a separate JAR file +Committed! Uwe I think I fixed all the places where we were making a placeholder term just to hold a field. +Regression! Tests outputerr and relating to PI names just give the unknown XSLT error in +guess you should reopen this pending that +Indeed Ive reverted your patch on my working copy and ran the new unit test and modpython behaved as you want without changing anything to applyfsdata +I think that this is a duplicate of which was fixed and included in +Do you what in cobertura is running then or how it is conflicting +just in +so that would seem to be a sporadic test failure also +Thank you so much for posting this and looking into this +resolved implementing our own classloader forcing to load tols related provider +Much appreciated +In general Im in favor of allowing the editingcreation of any extension properties to prevent there from being any hassle with MED unless that is what the user wants +I think we should put this into +implemented with the final fix from Benjamin in rthanks for the ran into a case today where the was failing because I wasnt allowing for error when doing comparisons +Thanks so much for the fix finally we are able to see all our Integration Unit test reports with the Sonar analysis now +Sorry Roshan I actually spent a while to decide which name should I call you +Could you please reattach? Thanks +shipped in +for the latter classpath for the test scope will include dependencies with declared compile test provided and other scopes +I am as well +I think we can put some rules in place and probably should about build breakage +This sort of closelogic always seems to be a bit finicky but I think youve got itnicely packaged now +As nobody commented on this issue in over years I guess it should be closed +HiI will looking to this for the feedback! I was not aware of having multiple filesI will try it +Thanks for Willen Voogd for reporting this +No objections raised to closing +Thank you for looking at the problemTo answer Ricks question I checked against Sybase Adaptive Server Anywhere Sybase Adaptive Server Enterprise Microsoft SQL Server and Hypersonic SQL just because that are the data base systems Im currently working with +oups I mistakenly tried to close that issue thinking I was on the one I am involve in I removed the comment and reopened the issue +I suggest to change this to only grab it if there is exactly +However as far as I can tell this method doesnt work unless youre able to annotate the entity source and access the main +Ill try to coordinate that next week +I forgot the panelPopup +Typo +testcomment to verify if my comments are always added twice sorry forany inconveniences +use the getType method to resolve the. +this is the same as but with the correct class name +Todd I tried to run with your patch +Sorry I missed one file in my patch so this is the updated patchThanksAnkit Jain +I will add a hive option to control the number of retries +the patch that was added here is the same as the patch in. +Cool thing! to add this check +So the line is causing this +Fixed with improvements. +JinchaoThe above case will fail because we dont do expire which is done in TrunkI am tring to make a new patch for introducing trunks expire logic when configure the is s +When I stepped through my test pages I found that the page load events are not firing for the child controls +Can you try testing this against master +Prototype Maven is looking really good I like that the interface captures the same basic logic of the that will make it pretty easy to integrate I reckon +Issue is not reproducible on the specified environments +Fixed in too. +This is now waiting for the new mgmt server to be built and ready +Have to do some cleanup and testing but Im quite confident by now +Moving issues that arent going to make it into to is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be by. +seems like there could be some weird situations +Attached the patch +Also to with revision . +Hi Can some one please give me the solution in in? That would be great. +Attaching a client side file and a server side file from running the test targets and +Ok Ill try +The test runs fine on trunk build right now though. +Could you give a concrete example project and what result should be? thanks +I have attached the modified from the testsuite directory +Hmm and I dont see any failures with jruby and on Ubuntu +due to the presence of W in the join order I think the results would still end up out of orderIm not seeing that seems like the data from W is irrelevant to the final ordering but I think its good to be cautious and verify all of this with some actual tables and some actual queriesMamta you might want to mark this issue as assigned to you +I took the liberty to rebase this to current branch +And those with tests even more +Ken can you make here an additional statement that you donate this code for the project to be published under BSD style license? Thanks +Ill check out the solr code and see what I can find +Im also including default formats in this task as it involes the same added to +I will combine two patches unless it conflicts with the original design +Danushka go ahead and work on it and ping me as when there are patches you want reviewed and committed +Same prob and experiencing excessive build times which is a big problem for our project +Yes Gao +The Hibernate test suite runs just fine on innodb +Patch looks goodSo if cluster has only one master and master loses its lease in zk previous it just stuck around as a zombie +And if you had to do a visit pattern afterwards that might be to late as you have already aggregated the incoming exchanges +I am only now noticing that my util file added in patch never made it in with the +war in the and replication is working as intendedThanks a all for the help +The only change is to add the alternative for situations to use instead and in such cases reduce the number of dependencies that are required +Its made against the tag not the ve also attached a sample subclass that leverages both features to do dynamic reloading of scripts +yes it doesnt really make sense to do anything with the written resource anymore +Ok will fix both +Committed to trunk and x. +Bulk closing stale resolved issues +JacquesI agree the layout of the Theme Selection lookup could be improved +We can just remove the reflection and keep Users interface as it is +I think the current implementation is correct +Sorry I dont get it a is destroyed when it is not empty +Stackdababcfcc Did this region get opened up correctly? Then the second set of logs is not a problem +The patch looks good +We are aiming for a release in winter +or at least if it ever has the option its off by I do a schema update I do not expect data to vanish in fact that would make ting difficult during development because I wouldnt be able to migrate the old data to the new you want to tables removed use and youll get exactly what you expect. +Vinithra the patch needs to be rebased to trunk and submitted for review +still there in +Have you tried to contact Fabrizio? I have planned this for a long time but stalled due to time problems +Hey DavidI just posted that link because I was currently previewing the page +Therefore this issue is rejected. +uses +trunk +Otherwise I will try ur other workaround +Thanks Owen. +We could add a note there on this difference? +Let me know if this isnt what you were referring to. + to trunk and +I have a similar problem with Rails and +Isabel Sure I will make those changesKarl I couldnt get +Ill have to download RFT but if I understand correctly the class path organization has the Logj classes in a parent class loader and the logj plugin classes in a child class loader +Updates +My wrong +If the same problem recurs in SVN HEAD please reopen +for bugfix release +Solved this Issue with a +You should not use Notifiers for communicating with endpoints +Done +Matei Ive had to do this in one of my branches too +You can repush the corrected changes the pull request will be automatically refreshed +Factoring out the common string might make it even better +My initial instinct is that its probably not very useful + from me +Note See also the descriptions and comments in the linked tickets +Thanks! Patch Applied +With fixes for findbugs warning in the related code +Is there a chance that this issue will be resolved in the near future +This is separate issues +Thanks +Thanks Senthil. +Confirmed that EAP CP contains version +Changed from a blocker as it is internal and doesnt change the functionality +Im away right now for the winter break but I will get to this again when I have timewhen I get back to work +Unknown + +Hi Allesio +Applied with changes +New patch fixed javadoc +submitting +Im hesitant to include it into if we are not going to include it in the release candidates +Hi Vasiliy thanks for patches +Suppose the machine on which my JT was running dies and I want to make the running start communicating with the new JT on a different machine +provide the . +Closing +I committed that change +This is going to be hard to make work for a while since jsf is a bit of a strange spec because the api and implementation jars are very strongly linked and wont work without one another +I think youll see the problem if you make sure the keyvalue entries in each row of your src table are uncorrelated +is part of the service interface contract but has a constructor that can include a generic Throwable object so that Throwable may or may not be an exception thats part of the service interface contract +I had tried to reproduce it +Thanks Emil. +Should be good to go +Duplicate of resolved issues. +Looks like both issues are review the sense adding listener more aggressively solves both issues and listner is cleaned up on disposal in existing codethus for the fix +Big for no feature and only bundles +Yes now works fine in application as well as in tests. +This looks goodOne problem though the patch assumes is being used for the UI only +A fixed version of the script by ips +Upload a new version soon with some complex think that this alternative +Yes I can see this too on latest branch and on master on Windows +New functionality should go in the next release +it took minutes all together. +I think this is related to this bug +The latter has been committed to the trunk and branches +Fixed +Not that Im aware of anyway +Same goes for a top level registered in and loads delegate filters based on the security configuration +The service ref for looks like this nameservicescore interface +Just removing the trailing spaces +I have added a note to the page you suggested. +I included Teds fix +I have an example that introduces though it is a bit involved because it uses some modified custom collection classes for business objects from +emptied Tomcat work directories etc +We can definitely leave the files from uncommitted if it worksThanks for having a look anyways +Since we fetch all commits from branch root until latest commit it should contain all the relevant affected paths +I see no +I have fixed the plugin to use properly +two additional thingsHow many nodes do you have and what replication factor you are using? If replication factor is smaller than total amount of nodes could you please verify that this happens on all nodes +al is on holiday for a bit so stay tuned +Thanks Brent +Attaching patch for fixing the args issue +MR not released. +the patch needs to be recut slightlystand by +solrexamplelibHave you actually tried that? I was talking to someone the other day that said they had to package this in the war to get it to work but Im pretty sure I remember getting it to work by putting it int he jetty lib dir in the past +I think we need to pass in the properties in the producer and consumer right +I am not sure we should add new error message here +Attached additional tests for this issue +Its the same effort a single command +Attach right patchIgnore previouse one version of the patch +Revision +So the solution is to decrease VPE update time +to your updated patch pending Hudson +Committed patch to documentation trunk at revision +See also security enhancements from clients to the release +you should use the latest code in SVN since is very old +Daniel thank you for your answers they brought interesting information to me. +Hiren Id suggest using the trunk version is most likely to be released in less than weeks from now so youll be back in safe position using official release reasonably soon +should be available only for warnings +However it only has one of them even though there are pages +Absolutely agree with Erik and Yonik I cant think of a realistic case in which this would present a security risk. +fixed +This caused subsequent loaders to ignore any components after was called the first time loading a composite file +Above comment should read enable not enabled +Bumping this patch up Daniel did your ever get a chance to run these tests +The demotemplates is evidently taken from the file javadrdaorgapachederbydrda in the code tree +When i updated my client I saw had changed so took a look +I only changed comparison when submenu is processed in original version doesnt matter if submenu isnt rendered or child isnt menu item and ul tag is put +bulk close of all resolved issues. +Having a new API for passing a list of robots names would be a clean solution +I cant proceed with running the other tests with a collated database until this issue is resolved +Ive reused the codefix for +Also we cant just assume that this class wont be used in production many applications run the first few months of production under extended QA and it is often useful to be able to turn this kind of functionality on in production without it having too much of a cant say for sure that the regex approach will be faster than using but I would assume it will be +The last patch fixes the warnings for me +So the next release of as well as wont print out warning +Will attach a test case in a few +The mappers are for the input formatsThe code could include a rule for how to decide that the first line is a header and skip over it +Exception indicated that some preview VDB jobs were running when the associated resource didnt exist +Thanks for the quick feedback Ralf +So resolving as wont fix as discussed. +Im gonna make a try to see how the JBI endpoint is behaving +The looks to be an interesting exercise to make it run in a real cluster +Need to produce HTML and PDF with Josh Wolf on Friday trying to see if the docs guys can help us out on providing some docbook versions of the current docs +Done +Reviewed by Gordon Sim +Verified fixed in CR build. +I executed the following command to build okxuepan sudo install ThanksDave +On ML Andrew asked about what the folder points toMight need to bring up app again hereP +Work completed +However I understand that this is not small issue +Resolving this as duplicate as has been fixed +Then keeping both deprecated and new API should be supported easily +Tim Changes look good +Another one thats not really Tez specific +Thanks +did an incremental merge with the branch to create the +HiGeronimo has still this problem +Patch applied under r +fixed thanksIm a little puzzled by this though +For remaining smaller parts and bug new issues should be created +The new patch looks good +Patch removes the test for and is intended to be applied from Tuscanyjava directory +Clearly one estimator does not fit all datasets +Bulk close for Solr +Lets remove the project naming context +We also confirmed that the hang does not occur with the changed test +Moving forward +which files should I download +Sounds good to meThe option to limit to joined text also sounds very useful +Ran both ant clean test Dtestcase ant Dhadoopversion clean test Dtestcase to trunk +Verified by Paulex. +I had to do the same +As I said I dont think there are any issues with it +Probably not although cluster will copy the file name around it wont copy the actual file +Also I read your email message but Ive posted my message about building Xerces +Oh ok +I am not even familiar with this kind of php mongo driver error +jira cleanup +Changing the component to Server +Same test as still applies and fails without it and this +Hi Hiro thanks for the support and suggestion +thanks see the enclosed test for large data set serialization +This is a patch for trunk +Might need to be rebased +the icons where the baseball is overlayed needs to be fixed too +Attaching a new review package because few files have changed since the time the last review package was generated +the new test was an issue where the set command would not modify the configuration of the Hive instance when multiple tests were run sequentially +Any advice + focuses on xml syntax and model structure is related but focuses on ensuring that the subsystems that are likely to need to extend their static config expose a generic properties element to make this possible +By default only projects have this tag +I believe that this issue is fixed. +in the above list of isnt possible to use AOP like +Instead of doing ist right in the parser maybe we could use callbacks in such situations to correct the parsed data +Good article Jonathan +Addressing Jitendras comments +Yes absolutelyI came across this wiki page too and was hmm a bit surprised +Thanks for reporting +No text appeared in the headers for these columns if they were enabled and the header pulldown menu displayed the text undefined for those items +Thanks for the contribution Attila! +all the update does etra is make sure the creation and writing of the settings file is make as atomic operation as far as the IDE is concerned +must be run with +Sam thanks I added a few comments to the gist. +The version banner could appear as part of the usage message +It looks like someone forgot to follow the release process +OK thats my fault Im sorry +This appears to be the same as the problem described in Jira issue +I would prefer to see documentation of how to set up a reproduceable development workflow using maven and eclipse along the lines of the constructing a server using maven +I think this issue can be resolved +conf should just bebin to set fix for version. +Patch extracted from usrucbwhoami directly is definitely the way to go IMO +Yeah both ideas are worth separate Jira issues +I like the ideas here but the design needs some discussion and the new classes can be added without compat breaks so pushing to patch is quite old +Need o get checked in to untangle the test harness so I can invoke it cleanly with no cores defined in discovery mode +By the way I was able to see enough contents of Method objects to identify my method +Being explicit is never messy imho +Adding Release Notes flag and docs status +Looks good at first sight +Im planning to review and commit this patch +r actually just made the password box on the index page disabledIt has not fixed the Given name issueBy the way is the Change password? needed? Can that be hiddenremoved too +Looks good to me. +Im buffering up a bunch of comments in review board but wont publish for a while since itll take me some time to go through all the code +Targeting revision . +The comment above was meant to be for another bug I posted. +Ive committed my changes to add the Redelivery Policy to NMS and use it in everything appears to be working am planning on more testing tomorrow +Hi Chetanthanks for the information and sample project +I dont agree with that sentiment but I accept that it requires an expert to understand when it is or isnt ok to refreshThus as a default I think it makes more sense to refresh since that will generally get you back in a reasonable state even if it ends up refreshing multiple times. +If it could be made pluggable it would be great +New patch attached to this issue as file +Fixed in SVN +This has been fixed as well. +So need to be updated in HDFS project +when removing a mixin M declaring a property P the current implementation first triesmigrate P to a definition declared by the resulting effective node type +Thats just great! My congrats youve been so fast and helpful! Thank you a lot +If we are to permit use of a native libhadoop there is at least one more thing that could be optimized the annoying issue of getting the amount of free disk space which currently requires running a Unix cygwin utility and parsing its text output +It also affects Oracle i and Oracle affects FP +BenAw nuts! I dont know how I missed that +Once the new VDB is deployed the old connections and cache is cleared. +Committed rev eebedcebafbfa +The fix could not be verified because of +Patch against SVN HEAD to treat multi valued fields like single valued fields when highlighting by looping over the field values and accumulating the highlighted snippetsThis corrects the behaviour Ive described and simplifies the code +I also added some more Deprecation notices and one log message +Please verify after is committed +Let me know if it needs to be cancelled Or maybe this should be in the release notes +Ive tested it in several scenerios +have created a patch about this add a method validate in a methode is called in connect method in +Thanks +Fixed the tabs and will commit shortly +I refactored the logic so Camel now aggregates the parallel tasksThis makes a tremendous difference +Anybody want to jump in +Tests pass +Otherwise having some test classes would be good as wellThis will make our code a bit lighter and give us more flexibility as well have more choice in the strategies to adopt +i was suffering from im pretty sure now +Ill fix all tabs in the next update +Can you create a phabricator entry +There is no way to tell using the existing API +If you want to use a inner class weare very evil and make you implement a custom for it +Were using jline which should support that on windows and unix platforms +See attached screenshot +Done in the linked issue. +Maybe the alternative crash should be just that +I think Igors point was that there was a need for arrays to support more complex page activation contexts but theres no components currently in Tapestry that require a literal map syntax +This will definitely get done by Dec +This led to Eclipse not being able to compile the project because the same library was in there more than once +Adds some test code +for the patch +Ah sorry I had already deleted it before I saw your latest comment +You dont have to test the Add button as well as the Browse button in the Edit dialog using +SPRINGCLEANING we can reopen if necessary +reopening so I can modify fix target to +IMO its less harmful to say you need to set this to a sensible location and reproduce the dump than you need to deal with all the problems a full root fs causes then set this to a sensible location and reproduce the dump +However I dont follow the scenario you mentionedbq +Could you please related issues +Ashwin can you take a look at your component in trunk and reduce the INFO logging to DEBUG andor TRACE to have a more gentle logging when using ClausI have made the appropriate changes to facilitate reduced logging +Will you apply my pull request +It would be great if we can fold it in execute though not sure how +there is no rule that there can only be files in a partition doesnt need to know the subdir name +Moving to +So I will create an account named +Verified by Nathan. +The doc must reflect changes in release of Axis +But explicitly creating objects one from one JAXP implementation and one from the other and mixing them looks like a user error. +dependencies changed Java now required Logj now requiredRuntime dependencies changed None +It has been a couple of years since I looked at this so my recollection might not be accurate +The account still shows the applications +Patched against the trunk +a Early Access page under Other Downloads like we have for source zip and JBDS update site zipb CSP as new downloadable zip to point about zip being a snapshot will add that. +Good stuff E. +Or else would have uploaded a patch +Closing. +Local scoping such items as local variables I believe makes the code much clearer see where how the variable is created inline with the test +Can you apply that fix +Murali created the event framework with this commit id commiteafcaacddefcfbc and this should have resolved this issueI verified in my setup and all the events are getting populated properly also all the annotations are in placePlease let me know if you still finds this issue +Is this possible? BrechtI did not actually test this with https so it is technically possible theres still an issue there +so whats the verdict +The code has changed drastically since the last patch is cut +We run Derby in embedded mode and often have multiple different databases open simultaneouslyHere are the stats you requested page cache size page size is JVM max heap is anywhere from GB to GBThe most recent times Ive seen the error has been at a customer site +Injection works fine on glassfish see jee example for detail use the forum for support +Hopeful fix in r the name of profile options werent renamed when they were as part of +Check config of Sitemap portlet +etc +Once I did that it stopped failing +I changed the log statement to CONFIG +It would be nice if the update to was also done on s side +Yah this stack trace is a result of trying to close the database when it has already been closed so its harmless +Carl will you work on this +Thanks in also +Thanks Andrew patch applied at r +I think this is done but I really dont know how the installer works could you check it a bit? I put variables in the and instances of them in the client and server corba plans +Well probably change the default persistence to classpath and also determine the persistence root URL according to the directoryjar that we found the file the default to classpath right away +Additional messsages should be in message and message formats +I agree that the data returned by CURL is correct but its different with any browser +Also added an unit test for it +Bulk closing stale resolved issues +Moved to of rejected or a dupe +Heres a rebased version for trunk +A page dump on Nospc error would be helpful in debugging patch adds the ability to dump a page in an insane build and adds calls to do so in outstanding nospc error cases +Patch for it will create a new table without compression +yeah positive +I hope that helps. +Submitting patch for a run. +NielsProblem was solved by introducing attribute that allows to turn this feature off +But the test expects the fields to default to persistentThe test case needs to be changed to identify these fields as not persistent +Whats strange is that these types of warnings errors only fail on Ubuntu +What value should we use? Any suggestions? +Yeah no problem +Modifed module to use port as well +I believe you are missing the point +This covers more cases that the issue identifies I ran into a similar problem months back where I could see the repository names in my code but not the contents themselves +I was getting error only after redeployment of the as cannot reproduced +I think sctemme already did it see rGav +The other two servers on both chains have processorsI suspect that most of the potential gains that the LFU algorithm might be able to provide are swallowed by the very inefficient implementationIf anyone has some thoughts for me to pursue I will be happy to do so but I am out of my own ideas +Marking closed. +Moving to as it a new featureimprovement +Closing until someone wants to this issue and do more work. +Removed findbugs warning +re right +Review it and we can try to design its behavior from here +Hi Bryan I have change to get the correct result in and +arquillian service is now registered lazily with the layer +Is it safe to assume that the dependency is within the files generated by the source generator +just committed this to top level site and all subprojects. +Bumping out to +I just saw there are readUTF and writeUTF removing readString and writeString I added +Added new improved version to get some feedback if anyone else is interested in a feature like implementation will when spring starts check all beans in the beanfactory looking for properties annotated with +Please provide proper Fix version thanks +Also Swakesh dont forget to add dependancy into your as the classes are in that Stuart DouglasTravis De Silva and Francisco Hernandez issue was resolved +It does support it just enable WindowCustomize perspectiveCommandsJava NavigationYou could argue we should see if we could make it enabled by default +Please note this is not the forum to discuss the behavior of any one implementation of the specification +I guess for this one we have to roll a custom filter that reads from the configuration which output formats have been excluded. +Now the reproducing condition is clear and its an encoding related issuethe issue can be reproduced with jruby cp e puts and chcp doesnt help +wow now that is a dumb mistake +Im essentially keeping the vote prior to the update to send in the case we find messages with older versions +Attaching a diff against the s version per Devarajs comment +Added the Using Apache Camel in section +the diff was not generated from the root directory +If someone else is able to reproduce this issue I think this needs to be fixed in as it is a regression +Tests pass so Ive moved it into to remove this dependency issue +I would recommend seeing if this reproduces with stress +Closing. +Sorry to nag +Does the supplied patch work +Thanks Jacques for committing patchYes you are right Jacques there is so much in its really hard to test all at a time +We should address that in new issues +phantomgalileoworkspace haikal find vote name applic +me too sorry! Eclipse left me blind for some funny reasonwaiting for test to complete before I commit again +I guess we need to wait for results from QA team. +A should be on the order of or so +In my opinion make install shouldnt be putting in usrlocallib in the first place and you should be able to run the java tutorial without installingFor the Erlang tutorial I assumed that youd already built the liberl directory and then pull in the dependencies using a relative path +The doesnt contain the +Can you take a look and comment there is a slight problem even with this patch as Ferdy explains at the end of the issue there is work to be done with Gora trunkI think you may also have uploaded the incorrect patch for this issue? Thanks +DmitryI just created a JIRA in maybe you can take a look at the patch and publish a new bundleIn this way we dont need to update the camel feature for it +As you may know you control jcifs by way of a D switch which sets its lmcompatibility level emulation +We cant make the replay itself multithreaded since we need the right order but we could certainly multithread the deserialization in which case were just soaking up previously unused CPU resources thats what youre getting at +A minor fix in the patch change FIXED to MAPFIXED in line +Is this warning something that we should work around or silence? If Im reading correctly it appears that you are planning on working around the bogus warnings in that case +Attaching with full stack trace +Nick +This wont be necessary for maintenance and snjezanas patch can be used directly +Instead we need to create an intermediate layer that translates JPA criteria API queries into our query objects. +I thought the bug fixed in? Apparently not +Actually I think the smart way to do this would be to just put the userBasic in geoserver and then to point the param thing at it or better yet have it be the last default in the class +I wrote down the replace by to my exercise bookBy I meant that the same changes should be also done for RHELs rpm e command in the Procedure +LICENSE NOTICE and modified version of the binary jars MANIFEST all placed in the jar +The following reserved words were added based on the Derby docs COALESCE CURRENTROLE GETCURRENTCONNECTION NONE OVER ROWNUMBER above testcases fail for both Derby and +Wouldnt setting the default values of copyErrors and copyMessages to false break the current behaviour +I cannot figure out how to get this databinding extension loaded +What not Listenable? But then this annotation must be on a List field so is perhaps best +minor tweak to remove Used Capacity from the hover over each bar since its already at the end of each bar +I resolved this by installing the optional plugin Maven issue here +Here is a patch doing the change +Builds are passing tests again +I should also mention Ive removed the dependency on odict +Dmytro the patch does not compiled +Minor comments I prefer to retain using instead of simple class name while building the client name +Can you be more specific with an example of what youd generate your schema for and where youd use it in a larger project? I just need some context +Any feedback on the designstrategy would be goodAlso Im working on adding some more documentation on the general strategy somewhere in the codepackage and improving the javadoc for the same as well +With this patch a Browser connecting to a Trinidad application will generally function though there may be some minor issues +will add +Attaching patch to switch using indexSummary instead of where affected +Do you have a patch +None of the Trinidad Committers have the time to check this and integrate the patch? +ive disable the test for now +I just looked through the commit history and confirmed that a version of running each RS as a different user was present prior toIt looks like the run as different user stuff goes back to +Nit change LASTROW to EMPTRYSTOPKEY +More specifically I used the config d command to delete an iterator + +Can you post the dump got byjavap c verbose +This should patch from to the revision with the patchWednesday July PM you wrote +tokens are discarded in a background thread +I removed the synchronization from +Just added few more logs and ensured that the proto file is not fully changedMade some changes to the enums +Let me figure why before changing default from to something like or M srccontribhbaseconf Add M srccontribhbasesrcjavaorgapachehadoophbase Add an +Last line should read then launch second MR job to do step +Unnecessary classes should be removed +Thanks again +Marking closed. +I tried deleting the temp odfold dir but I dont appear to have sufficient rights +Thanks for the patch I applied verify the fix in the latest in CR and in the and release notes. +Hmmm I must misunderstand about Attr interface +I would merge agentadaptor and sender +Resolving this +Thanks Bobby and Kihwal +Im also still looking for much greater specificity as to the use casesIll provide some of that for each individual ticket +Tom assigning back to you so we look at this as another possible feature for the next release +This patch is provided by giri +As long as test class inherits from all test class fields references will be nulled +In the servlet can you get the items from the trash folder and delete them rather than having me pass them +I dont see the exception on trunk so I think it shouldbe OK to close this issue unless someone wants to port the fix to +fixedThanks for reporting +I just committed this +Easiest to apply to trunk and cant make any promises it will apply cleanly to code +I suggest we dont restrict user and group names at all but provide a way to escape the special characters in the commands for example raghu +Updated to better support containers +and it seemed to work alright on Linux +Im also concerned that we would do this extract every time and it might not always be requiredIm looking at an alternative solution which is a little hacky but I think will solve this problem more efficiently. + +Patch uploaded +Sorry I cannot change the priority +No I dont want to change UGI +Fixed along with. +Pushed to github +would be super cool +Id say commit +This test passed against the trunk as Dan supposed the problem is fixed there +Updated patch +The reason there isnt a stage is that Thrift and Avro independently manage their own client threadsselectors inside of a black box +Yes Ive tried but doesnt workIs that because of or something else? Otherwise you would need a different Avro schema with a different sorting for each kind of grouping you want to do in the reducerIs that so bad? Any ideas about how to solve thatWould help? +Postponed again +I do appreciate it! +Bug appears in current CVS version as of all these issues should have fixed by +spin off issue for the NPE found during one of the iterations +Enforcing final equals and hashcode in a base class is not a good idea since objects which differed in their subclass state would then appear identical +Worth committing to +The priority message order in JMS is best attempt and not a guarantee that all messages will be ordered +these calls should work I think +Remove that test if you cannot make sense of it +Im assuming you can close it unless you want the list of javadocs somewhere +In we should send no ack on since the producer is not reading the response yet +The duplicate config files are a cause of many problems +Heres a better patch that calculates the exact size of the +Now that work is well underway with AS all previous community releases are +As Aaron suggested I have change the title +Correction The problem DOES also appear on JRE can confirm that this issue is solved in the beta from September +No idea why I didnt do it that way in the first place +Do I need to add all of them camel platform features or only fop bunlde +To check this I ran the test by specifying Dtimeout on the command line +Please take into account that JIT must work in generational mode when implementing the optimizationIn the simplest case it means that arraycopy should not be optimized for object arrays when write barriers are required by GC +Only in one case does the transaction fail to do the right thing and then we get an unrecoverable exception +Thanks for finding this elliot. +These issues have been resolved but never closed +bulk defer of open issues to +Please correct me if I am wrong I thought icons needed to be bundled in UI as they are displayed in eclipse perspectives +The ejb security work does not affect the other work that is happening in the ejb world +Ill fix this up +JBPAPPGACP branch switched to CP againCommitted revision . +Thanks to Christian Kaltepoth for provide this patch +The patch for that issue also included the patch attached here +Thanks for spotting the anomaly. +Marking this as resolved +Code changed but leaving issue open until we provide some info to the user about how to connect to their DB +Fixed. +Id rather see solved through the naming strategy support I still need to work on +Reordered the columns to Name Value +The code changes all look good now +Gobble up the SYNTAXRM and just throw the conversion exception so it looks like there was no SYNTXRM on the client and the client behavior matches the current behaviorI am a bit wary of throwing away the exception so I am planning to go with option which would be less compatible but have a cleaner implementation and wouldnt cause the SYNTAXRM to mysteriously disappear between server and client +Thanks Stack +Please also see below the two questions I haveIn sandboxctrezzo there is another app right at src folder +After I merge this one could you add a test case with a script element and issue a new PR for thatLooks good otherwise +now that we dont key files by their sequence id in this is simpler just allow duplicate sequence ids +Added InstallUninstall. +However if were stuck with at the moment were stuck with itRe the the then would tie Velocity to something I dont think we want to do +I dont know if there is api to check the facet +Eldon you need to run all the tests and modify to add a corresponding udfYou also add to add unit tests for these functionsAlso it would be simpler to add a patch containing all the changes instead of different files separately +ThanksCould we have a new release please +Thank you for the commentsI have fixed all the issues pointed out in the previous patch I had submittedI think you could use see a bit more instead of repeating the javadoc +Could you check if and works +Need to revisit. +I just add some new methods and comments on these methods +Would you please try the patch? Thanks very much +I just figured out that the problem only occurs on files that have breakpoints +All initial committers have their on file. +Here is the unit test that regresses this fix +Oh you are right all packages are set as optional +See the mailing listWhat does it mean? What should I look for? Is it related to undeprecate? If so can you give me the short answer yesno +Yes Id probably want to use Netty +goes in xercescparsers +Taking a picture from the camera and accepting returns to the app but never calls either success or failure. +The dependencies resolve correctly but the install fails basically saying it cant find a repository for any pluginsfeatures contributed by a referenced repository including core JBT plugins +Thanks for your patience +Thanks Sharad and all those who reviewed +Is this a bug against Rave or a new feature that is now requiredWhat Im trying to figure out is why this is showing up now and never before +The code is imported and cleaned up +I love stupid little things like this so I went ahead and fixed it +See and move +Grid once Libor gets a bit more info here can you take a look at this +Fixed in branch weird that it happend in first place +Update basic install on RHEL seems functional now +Hi Scott I believe the release note draft is fine +Updated patch to fix findbug warnings +projectonly for that particular use case +Duplicated with generated patch fot this issue +although its hard to prove this will preventcheckpoint patch +Im eager to see the Jenkins results although it seems the Jenkins build is stuck right nowAfter a closer look Im wondering if theres one more subtle difference between the old and new versions this time when the user specifies a filter +You can handle the message change if another jira if needed +hmm now i got it to listen to it if I override the setting explicitly +for the. +Darrin in future can you open CLOSED bugs and then just immediately resolve them with the same status they were closed with +It seams to me that there is something missing that allows me to reproduce the problem +The last item is I believe the most significant +mac os x +Yes Ill do that +Is the Seam Catch integration sufficient to cover what was available in Seam ? Or do we need something more +Seen my problem +Update with a comment explaining that is wrong but were doing it anyway +c Would it make sense to also run this with the taking a length argumentadded thisSee new patch +arrayunique doesnt actually work on complex types +Please check that it was applied as expected +although not sure whether it handles multiple splits to the same range or mergesNick pointed out that we are actually splitting those files by those files +bq +Its about lines of code each +Alexey I assign the issue to you to pay your attention to take a look the issue is Hibernate Tools related but not only cause look like we saw the similar issues before I remember the issue with the zip file and the issue is reproducible only in JBDS not in separate Hibernate Tools +Thanks +mrepositoryjavaxinjectUserstucu +Just a simple revert +context for the last comment I had Airplane Mode update +This should be fixed in CVS in the next few days +I have added from the provided patch +applied the changes and committed +Modified every test method to create a different table for testing +Patch for the same to make hudson pick up this file for trunk +This is configured by the properties +Hopefully you can get an idea from this for a better testI modified the composer to selectively throw a Error during the processing in this case it attempts to break each thread in it succeeds then the processing will eventually stop if it has been fixed then you will see messages such as the followingcompose a message on while looking for where the running worker thread number is lower than the one being searched for +Though Im just wondering if there are some clients which make some assumptions based on the implementation of Trash when they use it? This might break things for themClearly articulating what Trash is and what it is not is a good start here +Additional changes to documentation related to. +I was wrong about trunk I had not updated to the latest trunk and once I did I could reproduce this consistently +I think that we should remove all references to password substitution +Marking this as duplicated. +It turns out my workaround only allows the spring context to load but does not actually work +Backported to docs branch at revision . +Would suggest running hadoopqa a few times +Please dont use absolute pathes in patches +The visibility is not as strong condition for me as the presenceBut now I am not sure with introducing incompatibilities between Selenium and Graphene if there is an option to use then I think it is not worth doing that +Shutdown processor threads they will enqueue the currently selected keys requests in the request queue +Changes have been made to store the configuration in a and consistent access to that configuration +This patch updates the wsdl binding schemas import adding a schemaLocation attribute +One minor nit the body of the new try block isnt indented which Ill fix on checkin +Attached an updated patch with all the bugs review and provide your commentssuggestion +Remember that this bug was introduced in and my patch adapted from solved this. +Could you provide pointer to Metadatas SCM repo? Ive been trying to find it but cant find it yetStephen suggest me file another ticket about the diagnostics exception message +unfortunately I do not understand how to fix it but I am willing to help in any way I can doing releases or whatever +Ive spent some time to resolve the issue I was havingIn my case it was caused by a CAS filter thats used for single sign outThe filter mapping was set to I reduced this so that it does not pick up the webservice urlsEverything is working fine again in my caseMany thanks for all the help +The bat file shipped in EAP marking as resolved nowThe release notes for this version were not updated? +Im resolving the issue now and will close it in a couple of days if nothing further comes up +Ill keep thinking about how we can improve as it relates to modularity and security +We have recently converted an ant based build system with about projects to a Maven based build system with about projects +fixed in for better decoupling of from components +Last chunk upload does the actual creation which returns +added url to related the page should be ok once the site refreshes +Heh +Patch for which exposes the attachment name instead of the temporary file name +If i do a delete by entity referencean exception was thrown +Joram can you check if we should delegate this to Signavio +Thanks for looking +Additionally It is necessary to separate some catalog store implementation involving many dependencies from module +That the switch to Nexus didnt resolve the issue it was intended to is a separate issue. +Does the sql spec say anything about limits on these types? What would be the soft upgrade behavior? I think there may be more complications regarding implementing this for clientserver than embedded +As long as I can keep the overhead cycles to a minimum I am fine +Attaching the first rev of a release note for this issue +load is not a required step in resolution. +Oh definitely agreed but the immediate problem here wasnt the internet connection it was that kept crapping out during downloads while I had no problems downloading from anywhere else +Ill attach my example my example project +Ive also refactored more LoaderStorer tests to run against orc as well +Would this be a sufficient test? Should I expect a higher speedup from the optimization? Can anyone tell me why the if block would be a significant optimization? It appears as though the ifcompare is very similar to the two binary ops +I know that there is a workaroundI think this is an issue you might not agree and i respect that +Change the default value of from to + +When I apply the handlers from WSSJ +I think this is related to where iOS has a quirk where you need to wrap stuff in timeouts if they occur during the resume stepThat said this is more than just a quirk with the resume event +Although it does not state the solution it talks about predefined filter nodes for True False and Undefined +Right specifically the discussion revolved around the default environment should be something that just works +This should fix it but need to write tests +This issue is closed as out of date as per the middleware process rules for aging of bug reportsIf you believe this issues should be fixed please reopen and add reason for reopening +Yep this is a regression +The issue is that the clean command was issued prior to the testsuite clean command therefore there is no longer a built image in output +Looked great. +on eclipse related changes +Im kinda dissapointed +This feature is aimed at solving a certain class of problem alone +There must be a proxying problem then +Bulk move issues to and +The fix is to be done in the just like for +Heres the trunk patch. +This one had totally slipped my mind +Backported to + +Testing and checking it in tonight +I verified the changes in trunk and so am closing this issue +sorry forgot to change new Can you help Alex over in with how to get the thrift were using? Maybe you know +I think we need to get that expression tsanyType out of the XSD +In Lucene we had the same problems with our own impl that uses the Abstract classinterfaces classloader to loads its own implementations +If deprecated in then remove forOtherwise set the fix version for this issue to +Thanks Rob +In my testing I had a different place where an alert might fire if the app doesnt have a connection and that would run as the app was starting up and is not in the resume event + for the patch +Try this set rpcmaxthreads to something reasonable like and see if it repros +This patch adds an additional assert check for output path existence and removes the problematic mapred argument to from the unit test +for the patch +Its only if you specify without any URL qualifier that you get an exception if the file cant be located on the classpath with todays implementation +the last snapshot is OK for me +I just updated my tree to r +I believe Arron has committed the fixes for this now. +A zero timeout on the failover transport doesnt really make any sense every very small ones dont really make much sense since you are essentially making it fail before it ever has a chance to reconnect in which case why use failover +the patch looks goods +changed the buttons panel from with Actions to simple Buttons in Composite +do we support resultsetmetadata for mostly static +Patch looks good mike I think we should backport and add a changes entry too no? +Looks like Eclipse may have made a whole bunch of formatting changes to +Do not reject it +sorry Max it seems that its ot duplicate thought you wrote its a duplicate but I guess you mean its not a duplicate +Manually integration tests +Thanks to Divesh Dutta for the discussionThanks RegardsAmit Amit DiveshAmit your patch is in trunk rev +If the jstack contained then it possible it contains data from a table +This patch add a new method writevDirect to and only implement native code on Linux Windows side code will implement soon +postponed to CR since we review thePatch works great! Do you want to commit itThanks you already did it in +This patch proposal includesimproved comments in the code as well as additional testsThe grant test is enhanced as Mamta suggestedto verify that the trigger is operating as we expect it to and toverify that it is revoked as we expect +Thanks Tim the fix looks fine +Naveen please check that the issue was resolved as you expected +Since groovy seems not to be something to consider anymore we can go and remove that operator again completely if we want the bug error message is no longer reported instead a containing the +They can rather only be accessed indirectly through their containing might revisit this at some point but Im afraid that using beans for such purposes is exactly the recommended solution at +If either the majority of the instances are of Foo or if there are runs of Foo then runs of Bar wed like to be able to instances +Thanks for review and Jon. +But your point about reproducing the configuration that produced the problem is salient +It would be nice to do the directory game once +r to fix a problem with above +Im looking into it +I am happy with that. +This issue is already addressed byFeel free to do everything to improve it but I think the work has been already done +Perfect will test as soon as possibleThanksBengt +Tim regarding the setting of the that can be set at any time +And Im lacking tools to investigate because my tests on the linux box are only from the command line and Im lost without an IDE +Really fix this timeThe failure seems unrelated to my change +Not a tier supported JDK. +How is disableall supposed to work? Its supposed to take a regex pattern +This is a known bug and is fixed in and latest jci we will update those for forgot to say that the error occurs anytime there is more than one function in a drl with janino. +This was a bonus feature anyway for which we unfortunately didnt think the consequences +possible workaroundI changed the parameters of the method from int thumbnailSize to Integer thumbnailSize and it works +Ive heard different things when the dev is a committer but they tend to err on the side of caution +Now that work is well underway with AS all previous community releases are +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +Attaching a Test case in review it +This is the only seriaizlable class in jaspic +Increasing the junit test heap size from m to m +Also Im a bit concerned about the upgrade code on terms of testing +It may need some other tweaks to run with but it ought to be the same as yous as we grab the classloader which createdtrunk +Will resubmit after is committed +Based on the comments it seems its a for now. +Okay Im going to use the other issue just to revert the Searcher more of a taskThis issue can then be used to track the new work for this bug here +In are in release +Therefore it isnt really logical if you reused the value of the existing enabled parameter because then the call would do nothing +What caught my eye was that the advertiser value is a class name which means it isnt a plugin +Reopen hopefully someone will send me a complete test case and a patch +I think the delayed callbacks were so that they could all happen in a transaction that was going to commit +Thanks TriYour slightly modified patch is in trunk at revision The following labels used en instead of vi I fixed them The following labels have duplicated value in Vietnamese +Problem exists when to many connections +Should be noted that we would want classloader isolation for this packaging +I think that behavior have to be corrected! For tests I have changed the code a bit to erase elements with its dependencies +Since this JIRA has been created mail transport sender has been gone through drastic changes and we removed replyTo from mail sender and user MUST use the header +This issue is a bit similar to where we also keep track on number of consumers +May be these issues are the fix to this too seem to be solved +have updated the version to feel free to drop it from this version if you want to but I want to kill as a this bug resolved in +address the assert nit +closed +Lets do that then +I tried Martins suggestion of a simple example but I cannot reproduce +Correct +that doesnt work let me fix that before you check the code +Should we target it at as well +This will be called for each document that is scored in the result set +PROPERTYNAME explicitly refers to that description whilst for FEATUREID this is not as clear +An declared in the method signature of the ejb should not be getting wrapped +Patch for +and committed +Hihere is my possible contribute to the issue +Please set the requested fields so that the issue can be properly triaged or it will be rejected due to lack of information +Yes this is fixed in the M release +Interestng question +Fixed thanks for the nice patch! +Looks like this hasnt been picked up +Im happy for you to close the issue +Attach a new patch to fix the test mistake +This is a issue from +Patch that fixes the issue +There is a workaround from Oleg Tikhonov in this ticket. +Bulk close for Solr +How hard would it be to let each filter choose whether to work on prefix terms or not and run them through analysisA use case is for the Nordic characters +In this patch manage puppet manifests using git works for all nodesNext add support for multiple install configure scripts for services migrate zookeeper or cassandraI will get back to this in a few days +Committed revision +Hitesh can you pls rebase? Cant apply the patchAlso could you please run mvn install to make sure the integration tests pass? Thanks +But since theres a bit more discussion of the topic there Im closing this one as a duplicate. +Thanks passing remote objects as parameters. +Thanks for reviewing Sijie. +Hi replication by doc was designed to not use the changes API therefore no support for continuous replicationIf you want such kind of replication use a filtered replication +The test should be fixed not to be dependent on localhost at a minimum +Id highly appreciate if the Myfaces team could port all components on to portlet container ASAP +If you first install the http feature and then the webconsole it works +hudson test failures seem to be due to some temporary env +Alexey I have modified New JSF Project wizard to add facet to the generated project +Previously one underloaded server would be filled up before the next underloaded server is considered +Custom +Many thanks for this contribution and for putting up with my grumblingOleg +Attribute parameter is optional so it is not required in the file +If you were on EE you could use from the container +Since has some serious bugs Im going to chalk it up to that. +I did fixes wrt JSF and later have been a missing point for JSF Restoring view issue but I can reproduce this issue +Description of this patch Add description for under Setting Network Server properties +Though I agree that it is still useful for the individual pages after filtering out based on state +The problem of not returning a zxid is that currently the broadcast layer and the application layer are tightly coupled and separating them creates a potential problem for generating snapshots that we use to recover or bring a new follower up to speed +Closing resolved issues for already released versions +Have already been fixed at fix will expand the height on demand instead of showing the scroll on IE Firefox and in ER. +If the response not served from cached there is some extra processing in which case the share of marshallingunmarshalling is lower +Patch for JDBC Message Store attached +Thanks Ted and Jeffrey for the reviews +It triggers in these tests because pageCount is set to so pages are getting flushed very often +I dont think modifyThread should be a required permission +class entries from jars +Cocoon is already using a CVS version and if we use a beta of the httpclientthis will be even better +bq +Presuming youre talking about testNow fixed in JPOX CVS build dated or later +Please review and apply the attached patches +Will commit later today +This seems to be a global problem using and time sync +Thanks for the review Sergey Jimmy and LarsCommitted to trunk and. +Changes in the and things such as menu bar integration have not been done +Im going to change s to use and possibly break earlier Eclipse versions +Even in this case all outputs are copied out because the template specifies all of themThe patch is okay +Should confirm for and if true add a separate issue to add more functionality to the new portlet. +Yep this is it +Honestly I think the servlet spec is not real world focused especially when a large companies Internet web and app server network infrastructure comes into play +doesnt know if the region is opened yet +This removes nocommits from and also makes porting Automaton one step easier +Text Box Added For The Entity Name Search Will Perform All The Related Search For The Entity +So if youre listing a directory while someone is deleting it the permissions will be displayed incorrectly but the listing program wont crash +Please let me know if any more info becomes available before release +We have many other things to do +Patch submitted to make changes discussed abovethis breaks the build methods are removed from ARS but not added to CRS +Initial patch for merge to +how external of a client are we talking here? running on the same host or remoteAnything that doesnt use s own configuration +Grabbed from an node +Ill do it next time +Created two new classes Constant which extends from and +No apology necessary for the noise +fix version needs to be changed +can we try to preserve that somehow? thanks +railo is more flexible in the path input we will first check how many of the time is used to find the right have added request based template cache to railo +patch looks good +Ideally we could layer this as a user library +However adding fixed name for jdbc socket configuration is not possibleIMO the JOPR code is looking up these components +Do you mean VFS as in the Linux virtual file system kernel API or some other VFS +Ah +Tom just making sure this patch isnt waiting on me youre still investigatingThanks it doesnt look like this issue is past EAP QE yet +HiIve attached the maven project for camel base data format +Yeah this doesnt look too hot +Okay +Hi PeteFeel free to close this issue +The client simply has the option of requesting and using a token but the client can still continue to run with no token as it does todayId suggest we debate the pros and cons of adding another conf setting for whether clients will always use tokens with simple auth on the other jiras linked to the parent of this jira +Test case based on the demo that ships with the patch are in the apache trunk and be merged into fuse repository. +You might want to the fix which has been checked into the trunk and backported to the and branches +Thanks Burr Ill close this as a duplicate in that case +Hi Even after the datanode has sent the block report many times the overreplicated block is not deletedEven if i execute fsck for that particular file after sometime the block remains overreplicated +Lukas would you please review the atmosphere release notes and determine if an update would be beneficial at this time +Defer all open issues to +Uhm I attached a diff that wasnt enoughHeres a patch as well +Ill close the issue and adjust the Fix Version. +The Xerces dependency is necessary because jaxen works with Java and later +Is there a way to run standard SQL queries against Jira? The querying ability in Jira seems to have improved with JQL but I dont see a standard SQL interface yet? Still i would be limited to the results so probably not practical to use +well see dunno latter plan looks good to me +Hi Folks A patch is available on this issue +Can you provide more detail about what is and what its for +But I could be wrong and it was just omitted +If a data set is too big to fit into memory one can already put it in a relational database +Ok sounds good +the feature do deploy from designer to runtime engine is targetted for easy development +src and made sure the configuration names are correct +Just for the sake of easier testing I created a new archive with everything included and broker security disabled you try it outThanks DejanUsing that config all seems to work okay +As youve said you already have the ability to customise the command line. +attached patch +Ill file a revised patch for this +Since this code iscould be I think it wouldbe nice to add a code comment including a link to the URL you mention +Werner this is a new version of the XML that conforms to the +Colin Patrick Please look at this patch and that addresses the related comments +i have discussed this with Sean Corfield the committee chair of the CFML Advisory agree that this is not good because this is not a type it is a object reference but we both can life with self +The use of RPCs effectively restricts the impact of the scan on the FSA datanode operation that streams through a block without transferring it over the wire wont correctly check checksums using our existing mechanism +Ive tried to do it for but havent quite got it nailed down +Requesting separate approval for the libraries on which Spring depends would be an extremely lengthy process +patch released Released for trunk although not much differences +Ill regenerate it and attach the new patch +After further considering I think this is the wrong I changed the solution so that instead if no belongsTo is specified than the relationship is treated as unidirectional for each side +Third try to get the license comments all right +The patch looks okay +Verified. +Nicely done feature thanks! +Thanks for the last patch applied it to trunk +Good question personally Id say theres very little chance of other projects depending on this particular one but Id love to hear what rules apply to statements of individual Whirr services +upgrade hasnt fixed it +The erroneous predicate is coming fromcalled by called by +I dont think this bug will ever cause any problems seen by the users +Please attach a sample Graemetest bug report contain the application generate with SVN this version the error shows to SVN the application works Ill look at it this you so much GraemeI tested my real application with SVN an everything worked againstefano +Heres a revised patch +A deliberate travesty sounds sinnims que he posat sn els que apareixen en els diccionaris anglesos +This needs more thought +! usrbinenv jrubyBut in fact just outputs a shebang of! jruby +Thanks Joydeep +The project I attached works on OS X +Kevin C says in the duplicate bug filed The fix shouldnt be in the listener though it needs to be in +Consider closing this issue +WorkaroundOnly Hibernate uses the and there is already an option to NOT use +The only service not explicitly unregistered is the +It includes the following changes Infinispan upgrade upgrade Marshalling upgrade Removal of STATETRANSFER FLUSH from default protocol stacks Web sessions are now store in a cache per application +the search parameters form should start out expanded +Patch is committed in SVN . +I run mvn test +Vinod wanted to make different patches for the different projects +I can reproduce this s important to note that when connecting to the localhost IE defaults to compatibility mode which will not trigger the bug +For how a simple fix is required to address this issue it would make sense to do it to prevent this kind of risk +Test is passing in hudson + does not look for any file anywhere +This is not possible for is no way this is a maintenance task +Lets move the row cache discussion to a separate ticket and keep this one focused on key cache as the title says +Will take care of snapshot stuff in the next patch +Build start the webapp with mvn jettyrun and go to httplocalhost can use internal as the repositoryId +Verified +Attached validators and patch +The description clearly states that the config of the file is not used while the test case is using the annotations +Thanks JacquesI correct my patch with your patch to use already file that dont useI also add all field form person +Hi BrunoI like it seems to work well at st glanceThanks +Hi Brian Ive currently no way to reproduce this issue +Great I will test and commit in few hours +However it would be great to see in the output that the exception is due to the noimport setting +Heres an updated version that handles default values and enums +moving to future as not urgent for this release +If you want to edit the Connection profile you created then click the edit on import Database via JDBC dialog box. +The purpose of forcePurge is to try to clean out dead memory +Merged post +Consistent with service createPersonAny other parts of that does not respect this mandatory requirement should be reviewed +Upgrade tests passed +The report of store is correct +Tested in +This bug will have been fixed by tomorrow evening +Shut er down! Ive pushed the described change to master in the commit below +I did strenuously test on a secure cluster with combinations of normal users proxy users tokens proxy tokens both with and without a TGT present +Regarding Gavins comment unfortunately it is not possible to suppress this logger category it is a solution that is too extreme because the message is logged at level ERROR +The other one is activated on mousemove click or keypress events and removes the message in secondsMy patch comments out both timer and removes the message with delay if the user clicks on the message box. +provides unit tests that fail before the patchThe logic is definitely a bit hairy here so any thoughts on how to simplify it are welcome +I misread the patch +Applying patch of has fixed all issue as per description +Alexey Im afraid you cant execute Java during bootstrap +I have the same problem as Adrian using RAD and HibernateCould you advise where I could source Hibernate from so that I can run under RAD +Could you please confirm you are licensing the path tothe ASF? Thanks +Tested it locally with a single node job running maps and reduces and it passed with the changes but failed without them +If not Id say it is better to disable this example rather than shipping a broken oneAgreed +No longer applicable +actually restart is required in grails but is not required in so theres something different in and that is causing this problem which was carried on to the this is definitely not the case and its likely something to do with your environment +We may simply pass received blocks and deleted blocks in separated arrays +Next up the off flag +The configuration is now showing Subversion commits on our issues +What about giving Cygwin a try? Its a pretty cool env under windows and Im pretty sure you will be able to install and make ADS running on it +Is there any progress on this? The workaround gives problems with secured directories. +Thanks! +Why did this occur and what was the fix for the issue +Can bump it later if we have to +I was doing something wickedI also got a pointer to a near duplicate of this issue +I dont think you can have it both ways either misbehaving loggers are removed or theyre always calledI think you are missing my point +Can you attach the fix for this issue. +Please check that the patch was applied as you expected +Hi Sorry for the inconvenience occurred +Fixed some indentation issues along the way +Vladimir please add useful information from the following cookbook chaptersHow to use jsFunction with to ve added useful information fromHow to use jsFunction with JSON How to FAQ +here is a patch +If anyone finds a counter example please reopen or create a new issue +done +Dennis could you reviewed it +I see a lot of locks in preparing the +My assumption was that it is missing by mistake +Works fine for the first item but shows a problem with the forced caching on subsequent item adds +Verified with JBDS. +When compiled with JDK itll run fine against JDKThe caveat is that there is no guarantee that the JDK compiled will run against JDK +The patch looks good! Will commit shortly +Please see the nits on review board and upload a new patch if you agree +However it is not the same as proposed by this issue +looks at masterFUL puppetry at masterLESS idea was there any follow up on this? someone created shall I create a new ticket specifically for the standard master driven topology? as I can see all that has been implemented is towards the masterless design +DOH! and I looked into the brix source to find the protection +I just downloaded fresh copy of AS and installed latest java and still got same result +was just committed +Ive an question here +Ah you need to pass E to the patch command when applying the patch so that the emptied file is removed +Annual +to tell junit tests that runs with the upcoming embedded +The test also requires an updated commoncommonsrctestjavaorgapachecxfresourceresources file +Candidate for SOA is not an issue because different documentation has been created and therefore this problem is no longer applicable. +I found i filled a email address and correct it now +How to reproduce +no intended to have this uploaded to the Maven repository and figured that no would have been required in that case +Throw error until this new feature supported to let user know about incompatible input +It still knows nothing about the Request type hence any knowledge about Kafka +The bad news is that it hasnt done anything for my shield issue +Committed to +Let me ping the dev list and make sure a quorum of committers is on board since its something we will live with for a while +Expanding on my last commentThe +I can take on the fair scheduler for the reviews Bikas +The patch has been applied. +I didnt modify any javadoc so I dont believe this patch was responsible for that warning +bq +Closing the bug. +I dont yet see anything which clearly implicates erroneous as the cause but we should investigate further +The default implementation isnt correctly populating the actual location +If it is then it will appear only in java +Patch for for XSD to generate very large startElement some time yesterday to inspect it +Great thanks so much +Thanks Johan! +Im very interested in proposals for how this can be done +Uwe Please feel free to go ahead and do it and commit it that I can pick up the diff +A valid configuration must never yield duplicate services +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. + +Lets break this out into multiple issues +Committed to trunk with revision +Im also looking for this functionality +How did you implement this? Do you parse the config in your code? I am hesitating doing that because changes in config will make your code not usable +You probably can ask JCA guys for some helpHoward +Being able to add stuff to the target map makes it easier to carry stuff around with the target but I suppose there are alternatives for that +securityversionversion Seems not accurate anyway if youve tested it and its OK for you I can remove the one from smx specs +But if I zoom to some period where I dont have information between the entire time interval I get +Should this go into branch too Nicolas? +Fixed? new error +Can you elaborate more what you mean with revised support for JSONadoesnt cover XML? bfewer libraries to be usedcPoorly documentedwhere do you mean the documentation is poor? And if so can you help improve it +Im expecting that wiki will probably contain only one documentation for most newer version +Has fixes for bad startstop in hbase. +I will check out the patch +The added cost is surely minor while keeping the code simple +Ill try to look at these parts of code base for better memory management +Hi Leonardoclearly you will need to reset the cached instance when a behaviour is added but apart from this the caching is certainly goodbest regards MartinOk I committed the missing lineregardsLeonardo +Problem shown as a for a solution +If kerberos is enabled on the client but theres no TGT it doesntshouldnt throw an exception +Oh I see +This is because introspector now detects more properties in several Swing AWT classes than it was able to detect before +Or maybe leave the current implementation and make it possible to grow more trees than expected when the number of mappers happens to be different +Ill do that from now on. +I think the right solution is to just close this bug without any action +Closing issue. +Attach a patch for enabling web deployment +Please change to if you intend to submit a fix for. +Hi Flavio sounds right +Most likely another on the default portAs I mentioned before this patch does not reverse port rolling behavior just provides support for the deprecated config parametersAnd deprecated configuration parameter message does not appear because your configuration does not contain old namesThe application you use should explicitly specify port if the actual port is meant to be arbitraryDoug I agree this would be a better solution +I can see what is happening here after the bean is removed the association interceptor attempts to release it and does not find it in the cache as it has been removed +Where will this functionality be embedded +Any update on this +Fixed now Closed. +Ive not applied the patch to remove the test from the windows exclude list because I dont have a windows machine handy and I still get two failures on Linux +Patch that moves as many tests as possible from junit to ee +Andrea is this issue really just +rtf with correctwrong result +I think it is not necessary to call from this goal since maven dist takes care of compiling all the sourcesIt also adds a postGoal to the maven dist goal that copies the distribution and the files for the remote maven repositories to the releases directory parallel to trunk and branches +Id say if works even though we have to move a good few tests to medium lets use it so we can undo our dependency on custom surefire +resolved in of a misunderstanding here which is my fault for not being clear +Forgot to mention I dont see any javac warnings with the patch either +This was recently fixed by perhaps the change hasnt propagated to the git mirrors you may be using yet +Testing this was probably more work than the patch +Thanks Justin I will close this one. +I have sent you and email as well +Please confirm +ok to move skip design in Java +Username jacyg +Workspace repository now filters package exports during compile time resolution to ensure that projects are not supplied as providers of capabilities that are not available from the source code avoids messy compile time errors when a project is added as a provider of classes which it doesnt contain until after build issues raised by me as closed +Crash issues should always be majorCan you provide some sample code that demonstrates the crash? Thanks. +gossip is for internal cluster state we definitely shouldnt be using it for rpcaddress and rpcport +Either solution is ok with me +that stuff rarely goes away +Just wanted feedback +This can easily be a factor of or more +My question is whether this new API should handle recursive traversal and symbolic resolutionMy intuition is that recursive file listings for open should follow symbolic links since open follows symbolic links +Also could you please describe exactly what you are doing? Does the file appear in the attachment directory? How about in your workDir? Does the upload succeed and then you get the error or does it appear when you just go toThere are no errors in the log files so therefore this is almost impossible to debug +sorry +enough votes to make this a +Nice catch Eugene +This method is only documented by the JDBC javadoc not by Derbys documentationIf the application was fetching columns by name then the application sees no changeIf the application was fetching the SPECIFICNAME column by number then they would have realized that there was a bug in Derby and reported it +For an update I need to insert more data +If not is there a timeline bywhen this will be fixed +I should have read Mithuns original description bit more carefully +Can this issue be gone? EAP branch deploys and runs +I think it would be better to file a separate issue for it would make it easier to track revisions and issues +If the enum definition is moved to a Java file then Javadoc does appearIf the enum definition is moved to a different Groovy file there is still no Javadoc +Updates to the patch responding to comments should also be indicated in Jira with a comment on the issueI did attach a patch to JIRA before it was committed +Attach another patch to make it work +So +This gets complicated because the rows boundary is not the chunk boundary I tried to map more blocks than needed for a given boundary but there is a possibility of running out of memory hence i am planning to do the followingMake boundaries for the chunk instead of the row positionChunks between the boundaries will + branch no more under active development +v doesnt apply +Here I am attaching the patch which has included code changes required in both task and when using quartzAs a summarization of code changes done Earlier new trigger was created using and now it was done using which introduced through DSL +This should solve the Clob error messages +So if thats the only thing thats stopping you feel free to break it +patch attached +Lets see what we can do here + for the patch +to put the ACL stuff in it +Hiding implementation is a good idea +testcase for me with version see testcase attached +Is the project running ok on command line +needs new WS and may need extra config +Is this something that we would need to have plugin writers take into account or would this be one of those ME plugins that shows up when you import something new +Oh yes it can be +I have a lot of web services and this is the first one to produce this situation +If this patch is ok I will prepare patch for also +What release should i use? i think i will need all the help i can getdthis is a serious bug though the kind of bugs that just makes potential users dismiss the whole thing! Anyway i guess the effort will be best put into the right release +Unfortunately we dont have choice thats the reason why there is a reference implementation +I also dont see the problem with the +Made the private methods to now be note this patch also provides the fix for the bug reported in issue think we can fix during the refactoring for +Thanks Ning +Part of a bulk update all resolutions changed to done please review history to original resolution type +Marking this as a bloker as the resolution must be either Resolved if we can use Vladimirs workaround Make it a Documentation isse for the release notes with a Solaris warning about this added am starting my Solaris install +My current guess is +this looks good one comment could we remove the redundant Links for this job bullet point +Thanks Peeyush +Attaching version of the patch +The english data appears to always have a blank line after DOCSTART tag causing the to not get cleared properly on the start of a new documentIm also fixing this bug with this issue and submitting a very small sample from the data set +Opened a separate issue around cleaning up the Hibernate Search build hacks +It looks good at my you have a chance to give this a spin with the latest code would you let me know how that goesThanks. +This is a problematic issue if we dont fix it we cant reach the STANDARD certification +Onemay be better off adding notes within the documentation of the latest versionon the web site where the mptest example is given highlighting that it is onlyvalid for and later and that people should check which version they areusing and refer to older documentation if they are using an older version +A better patch +The reason we did not include it there originally was that we thought we might implement distributions for which we could not define inverses +the problem is now on the server so its in but adding center fix the problem at least a track to follow now +bulk close for release +Resource used in the experiments +With the new ctor introduced Bill had to make the namenode instance nonfinal +I think I was too eager filing ticketsYep we should be able to completely do away withIn the same way you suggested to pass the readpoint to the consructor we should be able to pass the same to all +Great work Prakash +Sending to David for upstream push +Done as Kiran Kirancan this issue can be closed +This breaks native code loading in all deployments +Let me know +It would actually make sense to add a menu bar to an app similar to how it works on iOS or to add an action barI tested this with the Samsung Galaxy Note running Android since all other devices just dont have a menu button to press. +Thanks Benoit for the portSince the code of the two classes are identical I am checking whether we could simply remove this class in favor of the other one +Fixed in rev +Closing tickets +However currently its just a stupid type structure doing nothing else but wrapping two values +Closing this as duplicate. +and will be required attribute for the service so optionality for them should be false and no need to have facilityId as a In attribute in the service defPlease do the needful changes and upload the updated patch +Those shown in the will appear if all previous actions have been executed +looks good will commit if the tests in both trunk and +I think we need a new RC which includes the patch for this issue with abort as default behavior until this is sorted for +Thanks Jie and Jimmy. +The documentation is wrong there +A release notice should be enough +LGTM to trunk and +yeah the python docs say that repr should be used if there is no str and every other case I have seen follows that rule but the exception printing apparently does not +up +I cant run the client due to unsatisfied dependenciescxfjarjarjar I got this working + Lets add a new optionThere is a little problem that sync in and hflush in are different +for the time being should work when using the appropriate means. +Im using Hibernate Beta and keep getting the exception Unknown wrap conversion requested to is an enum type +Thank you Ruyue MaDoes not look the test failure has anything to do with the patchI ran it several times with the patch locally no failures. +Fixed. +I think it makes sense to be able to replay the sax recorder for desig verification just after deserializationHowever if the any element of the message is modified later then it should be optimized during serialization which also involves in removal of ignorable whitespace whithin THIS element +this is still a the is wrong here and probably the cause of all your problems +The next step We need a new that includes the new method Ive added to the plugin +So the requests sent by the individual fetchers from the consumer will appear with the read name in the Kafka brokers request log +Im thinking as more of a utility method since otherwise callers have to implement the logic to figure this out +Which incidentally makes this whole issue a moot pointIll play a bit more tomorrow and probably close as Invalid +on to WONTFIX this issue +Verified all components needed to be verified manually Push Data Scroller EDT Drag Support Tooltip Color Picker Editor File Upload Inplace Select Ordering List all of the Rich Miscellanous group +Committed patch as revision +but no crash for me +Committed to trunk. +not fixed rather its a duplicate +I believe Ive now resolved this problem and this jira can be closed +Please go ahead and commit +Moving new nextjbosstools to LATER purely for not cluttering incoming unassigned issues +can you include full STit sounds like a thrift bug not ours +the patch with an improvement for the you feel this is the way to go I will change the other classes that have a main method as well +probably i just tried and the code worked fine +Internally this constructor invokes build which iterates over the sorted map and builds the internal structures of the CSLM without iterating +I think they should be ignored in this JIRA and new tests should be added to regression test suite +In the file you have place holders for registering queues and topcis + patch +Test failure is +Please ignore my comments aboveAfter checking with the reporter I found the issue is that while we list the RAR modules in the portlets we only check that whether the configuration gbean is running and list their statesAlthough one of the is occurpted for it will not affect the configuration gbean +After fixing errors are gone. +New patch fixing all nocommits +Not a bug you need HAJNDI for this +This fix is very important for us in our use of the plugin in when will get pushed? + +I have also update the tutorial doc as well +I plan to do this next week using the header above +Let me know if it helped. +Any update on this +Resolving as cantrepro though I suspect may have solved it. +Hi DavidThis patch fixes all of your previous comments +Ive got some work Im in the middle of and perhaps some more to come on further review +There is a known issue when the number of maps is high Ill try to solve it in the next patchThe main limitation of this implementation is that each mapper loads a copy of the data in memory and because I dont know how to actually share the data between the mappers of the same slave node this means that is you launch simultaneously N maps per cluster node youll get N copies of the data in each nodes memory +Ive never tried though as the GXP tags are meant to be a replacement for the Struts tags +Bumping off +Fix applied +issue cd +Hi LionelIve applied the patch you submitted on the mailing list so that should be take care of in the next release. +I added this patch to the Trinidad for testing in the portal environment + looks good to me +Same as trunk just some lines didnt merge cleanly. +Changed function name to move +How would we be compacting with with such ratio? Unless major is forced it will compact s together in various combinations until they reach right +This means you can specify query analyzersbq +I think its a good idea to replace the deprecated APT classes from the packages +I ran locally and they passed +Renaming the patch to make it clear it adds to the previous rather than replaces +Sharath Please fix this in Branch as wellThanks should be fixed for scheduled for Wednesday +Closed upon release of. +does this one ring a bell to you? Could restconfig have issues in this caseDavid any news on reproducing this one? Otherwise I guess its better to close as cannot reproduce +without doing any benchmark with modern JDK I believe my approach that uses indexOf and avoid regex is slightly more efficient +Can you add the new parameter in the file +This approach has long served HTTP SMTP and unix enviroment variables as a means of specifying parameters +If anyone else wants to develop it further they are welcome to do so +ok +We need to ask if we can have access to the CSS for the site and update ourselves +Fixed script will run from any directory +I will pick up the next snapshot buildIn the meantime I have been updating from SVN and building locally as it seems Hudson is really backed up with jobs + to trunk +Resolved thanks for reporting and patch +remaining issues are covered by the VM crash issue +We could have handled stopping the service in prerun management +I used to post questions on the forum +This is what I see in the target directories for this quickstart jtstarget jtstarget jtstarget Should it be looking forjarclient when doing an undeploy from server ? Maybe Im backwards +A bit had to restore core +Verified on JBT +There is a design for static initialization that will fix this +I committed this one too +fixed r +As I understand it the additional list setup information is sponsoring PMC comcom initial moderators hindessm zoe list is original concept for the Apache Retreat was something along the lines ofHey guys +works now great +Will commit if tests pass +I think our dependency is from back when we used to package hive along with hcatalog and can be removed +Though we may need to add some tests to the plugin suite to cover this scenario. +excellent diagnosis of the problem Yangpatch attached to push the flush calls off of the CL executor +see. +Can I have logs from around the above exception if you still have themMind if I move this out of? It looks rare and its not going to happen in because the memcache datastructure now can be iterated without throwing CME +Thoughts please +Some additional infoWe have been testing this some more and it seems like the issue is only reproducible on a bit have tried several bit systems here and cant seem to reproduce +Theyve already been corrected during the changes but Ive also edited to fix the link. +trigger the build +The attached zip has the Java src but not the rest of the files any more could you attach a zip of the complete application including all the files src +Any plan to solve this issue? this bug was reported a long time ago. +I would assume that most JDBC drivers keep the autocommit state on the client side and send it as a commit this statement flag along with a statement to be executed in which case the performance hit of the set call is insignificantNote also that with we now guard the set calls with get and only change the mode if needed +Henry what do you think? slating for HenryAny commentsThanks +Committed to trunk +It works for the same reason this code doesnt work if say isnt well formed + failed on contrib +. +Closes issue after bundle release. +With Davids suggetion +However I still get the hang intermittently on my machine +Hi ClausPlease find attached a patch for this issue +Would it really be necessary to maintain a list of exception monitors? Then you would still have to make sure that an exception is passed to the correct exception monitor associated with the the session was originally created by? Or am I missing something here +A notice was sent to list +At IRC channel David Lloyd and I agreed the best approach is blocking the problematic migrate before the exception is thrown thus holding the thread until select is invoked by the worker threadThis would prevent the exception from occurring +Uploaded a new patch which changes to log NA as a principal name in cases when principal is not available rather than using JMX SERVER as a principal look good patch applied. +Committed revision Thank you Pierre however next time please provide a diff file which is generated from the root directory +Guess its time I stop staying we should and start saying Ive doneIll work on refactoring the build process and the hadoop package to get things rolling +New patch fixing the test cases +I have merged this to. +Thanks Knut for your suggestionsIterating over Activations vector only when we know that at least one activation is marked unused will save some cycles in case of activations +I downloaded source code and see that in last hibernate code in class connectionProvider always will be null in case! multiCould you please help me to solve this problem? Are you going to fix this issue +Touches the following fileM +Thanks Alok Agnihotri for your valuable patchSpecial thanks to Rishi SolankiPranay PandeyBrajesh Patel and others in helping AlokChanges are in rev Ashish to go off topic but BJ I didnt show you that must have been someone else +Thanks again +I agree with you that this will be much more efficient because this will only be executed once at the time of buildling the statement +Creating archetype based project example now let the user choose a different project name AND keep existing projects based on the same example +Good idea will get it in Camel pretty soon +Thanks +Patch is for Java Cath Are you happy for us to close this issue now or do you still want to apply a change to the +Postponed. +Superseded by and related. +How can I do it + that demonstrates the behaviour +No such book exists for. +We will need to document the issue +Resolved. +Bulk close for +This is all to I dont want to risk Solrs good standing by merging this to a stable branch +With Jason to begin initial you try with +Removed an unrelated change +the version was released +Im not sure we should stop deployment if there any error markers in a project +I have seeing the javadoc but its still +I have labelled the jars as releases in line with the main release +We now have a providing the basis for method resolution and invocation and we have extensions for various method support like activate deactivate modified bind and unbind +Closed +Try the forums. +Its a very tricky thing to test since it depends on either or one call reading a backref and one call writing a backref in ways that are not immediately apparent to the compiler +HADOOPOPTSXmxg HADOOPCLASSPATHhometoddgitbuildhometoddgitbuild +I cannot replicate this in JBDS but neither could I in JBDSFeel free to reopen if you can still see the problem after the fixClosing +For example I dont know what go to internal presence probes would meanCould you please give a complete and precise report of what works and what doesnt? +Fixed +Closed as a duplicate of. +Thanks Jing! +libcassandrabin +Still needs work for spanned cells +start with queries then warm things with the first and use the second for the benchmark +I have to concur with Javier +What I want to avoid is having to complete tasks for the JDG before I can release the styles or vice versa +Here is a sample of the generating an instance XML +You can change the version of Felix it tests by editing the dependency + +It seems this is rather done by the Maven Archiver itself +actually this patch might be better +Should also include the transfer rate the speedup from rsync etc +if someone is happy to implement it +Ive incorporated this into docbook format +Obviously we cant ensure that future writes beyond the taking of the snapshot end up in the snapshot +by default its in localstatelibdirIt save the uti httpIPPORT on one line when couchdb start +Patch applied. +Nice catch Joel +help review thanks! did not know how the monitor portlet works but Is it possible for us to delete all the graph belong to that server? maybe we could give a confirm box which writes that all the will be delete +bq +The is in a separate jar inside an EAR +I am now creating and together to make sure that we always have both created +GertIk ken zelf wel het n en ander van DBs en kan inderdaad anders wel op een DBA beroep om een de DB te kunnen optimaliseren is het vooral belangrijk om een beeld te krijgen van de statements die worden lijkt het mij in de eerste plaats handig om via SQL Logging te kijken welke statements het meeste worden uitgevoerd op de de hand daarvan kunnen we dan bekijken of we beter bepaalde indexen +Thanks Stack for the review +Thank you Jonathan for the patch +resolveCyclesResolves cycles by identifying a nullable foreign key and then calling either add or add depending on the operation causing the dependency haschecks for local columns kernel added messages used in +That wont cost much and you open the project to a broader audience. +addressing any further limitations to the memory usage tracking as it pertains to actual jvm usage will be punted to +As this is a duplicate report Im closing this one +Is this still an issue? Can we close the ticket? +This is really an issue with the xml format in pmd +I dont do this every day but maybe two or three times a year and then run extensive queries locally using the exported data certainly much less pressure on Jira than run those queries against Jira itself +you cant get to any hibernate related data +Applied +I now register a validation event handler with JAXB +Ok +It really only helps apps that dont use Seam as their primary framework which is quite rare currently +Todd updated to use hashlib not realizing that it broke compatibility +Fix javadoc a core and +Previously there was occasional glitching when selecting a new Connection which caused it to hang on Broker Disconnected +Multi Update had changed most of the codebase +Then you can configure just the base config class as in the IDE and use the three beans config sets to distinguish between the profiles +samplesReadme updated to include good to me applied thanks for the patches +I guess I should have RTFM +merged fix from trunk to codeline +Rebased and merged into the master branch. +Yes +Ive applied Matts patch thanks seems that this bug is not fixed yet in grails +I could not determine what the javadoc errors were +From our side the patches resolve all review comments and pass unit testWe can further discuss the issue about supporting string data id separately +I am using jboss Solaris and JDK +We are missing the references to Index and Unique on We have Indexes Index Uniques Unique that are specifable at class level + +I noticed came but nothing has happened on this ticket +Has no test cases yet +Suggested fix sounds good to me +should we commit this +So if anyone thinks a refactor here is helpful I can rebase it +Shouldnt the change be that avalon doesnt claim about null componentsI remember long ago that it was decided to allow releasing null components inorder to avoid hundreds of tests against null which are also done by the CM remember the same +I just committed this +Isnt this done already? Or we have it in the docs +Assigned for release + +If the response contains the requested number of rows then youll have a next page +Attaching patchThanks you please provide more detail +closed after releasing version +We want to be able to treat the node list as POJO but not everythingIm loving the comment explaining whats going on right up until the end of this sentence +apply the native patch on the directory +So if you want a singleton bean you have to use for EE apps and for non EE apps which I dont think is that great +Feel free to reopen if theres still an issue here. +create files in tmp directory atomic put regions to +Maybe the description of the modified BNF syntax should be moviedcopied to the Reference Guide +That would cause the clients to violate JMS spec behavior which we dont want to do. +I just file for tracking this issueMeanwhile as the need to fix this issue is greater I am going ahead with the basic approach and changing the error handling +Proper test infrastructure +IMHO its a showstopper to use Archiva in real a good day best regards +I have updated the page with your comments +THANK YOU for the patch +What does it say on the master at this time? Why did the master decide to send an open rpc +Updated +At least Hibernates custom Entity annotation should allow the Hibernate name to be specified +This resolves several test failuresThis patch is intended for +I think this should work if the cycle does not pose a problem +janneNot much done on thisThe value of the selected editor is apparently both stored in a session parameter and in a cookie parameterIt should actually only be stored in the cookies no need for a session parameterNo time yet to dig into this +This is clearly wrong for exception variables because their scope never outlives a single method invocation +Im not able to test the patch on windows +The root cause is the file system of router vm is in state +In my environment the tool is correctly configured after the installer was run +So for that strange cases there could be a method to specifiy a route from a message to an operation it doesnt matter if its more or less complicated as its going to be more the exception than the norm but having a convention could lead to Muse not being integrable with other WS specifications even if they can interact with WSDM perfectly +I made a deep path in windows and a repository in svn +However on Cordova the app actually will force quit +I remember we were able to cover all +The zip distribution is dedicated for windows platformOn unix systems you should use the distribution. +How can we decide what is the constant we should use as Format of inside the Subject element of the assertion? For an example in the attached patch I have set Format type as email +Reassign to a release that you feel comfortable resolving this issue in or leave as is. +You must try build after you have a chance to test with +Im ok with something like that probably widthheight to be consistent with other items but it does mean you probably should do a bit more error checking to see that sensible combinations are suppliedI guess I was pairing them up like objects +bq +And it looks similar to what I have with Technology Preview text +Sorry in what kind of configuration is the WPS community module in the build +Item is not our problem or even not a problem at all +the patch looks good +If no objections I will check into also +What do you mean by that? I was using the CR update site +Ack! Looks like this hardcoded check for api level in the cli did not get fixed in not even updated to look for api level +Im happy to contribute if youd like +I do +I printed out the stack and found found that the traver method had been invoked for many times +Hi +This fix will be included in the release +By default they report only first problems +Apparently Im forgetful in my old age +let me know if you have any questions +Editorial pass over all release notes prior to publication of. +Username jboynes +The first thread is our sender thread and it hangsThanks +Specifically on windows I see a failure for and +I think we need to spend some time on this to do it right +committed both patches for Cale +Committed revision in trunk +eg Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB Up Normal KB In the above is used is DC and is DC +I cannot reproduce this in trunk +I will be porting these changes to trunk soon +Patch consisting of changes to expand escape handling in CSV have attached a new test for that shows how the handling of escapes could workEDIT got a bit confused with the escape handling so removing my patch +Marking patch hardcoding of port numbers in the test look wrong Subbu +Committed revision in +No extra infodetail is logged in debug +I have reopened and attached a stack trace +The getPassword method returns the boolean true or false if the text box is a password field or not. +Please verify +Its not really documented and Im not sure if anyones using it +Integration tests were updated. +Really forward slashes +full have added a progress dialog +We used to have a profile and property in the parent pom for filepathtocachedrequirementsfolder Im guessing those are now obsoleted by this change? Not really +JochenWhy dont you do it yourself you have karma I might have the necessary permissions in terms of SVN and Unix +Thanks regards Dragoljub Spasojevicok added ftp url will show in an hour or soRegards HPP +Should be fixed now. +But when i run again the error is not there +Stack trace seems to indicate this is reproducible +Rebased to latest to branch. +Any progress in fixing this issueIt is really blocking +Please create an attachment that shows the problem +It seems that the issue is not having the jar in the classpath +So if git svn works I will go that way +With Greg to +I had to switch my application to use the Filter +But be aware that tab is hidden by default +For example is converts enum to String type field or pid to String pid field +Thanks +So is +So +Hi LahiruI was just wondering if you have managed to find out any more information on this issueThanksCan someone please provide me with a response regarding the following questions Why the default plugin download is not working under JDK? Can you tell me what the latest build I should take and where I can locate it? I can download this and attempt to do a standard ANT build +Actually I have a pretty good idea what the problem is here +I will add read and readFully methods to incorporating review advantage of having the checksum internal to the data file is that this mechanism could be used outside of too +You could also try to employ interpreter +If you want to disable cache loaders simply comment out the relevant block in XML or dont set a in the Configuration object +I wanted just to say that i had doubts that removing javabeans or changing exceptions into simple classes should fully resolve the question +Please file a new bug if you can still reproduce this. +Im hoping to submit a draft tomorrow to begin soliciting feedback from interested parties +Marking closed. +Thanks +I will mark this issue as cannot reproduce for now. +Any word on the tests you were looking into +See attached file +I guess the question is assuming something goes wrong attaching the patch to the issue and it ends up corrupt with your approach the merge will still succeed +It was fixed by earlier today. +Id like to bring across the content from our existing moinmoin instance +Fixed along with ConverterFormatter lookup had an issue in the +Defer to +You are right sorryDeleted duplicate patch and uploaded the good one +This is a key use case for Instance though so Im not sure +Documenting this seems enough for now I think we could add some server number checks as well + I restarted the namenode with no successDo you still see the blocks on the after restartingIn there was no so called lease recovery for files that were not closed properlyDo you know whether the files were closed before becoming empty +Reopen when this information is available +Same here Ive the exact same problemCan you please give us the good reason to not use the inherited properties in association +We have same concerns about the maintainability of +Fixed checked into CVS +my opinion filter for Database schema tree is a better solution + However it makes me wonder how ss worked beforeEdit reread the code properlyThe way its supposed to work is that jabberserver namespaced stanzas are rewritten into jabberclient in but that clearly comes too late +Did we do any performance benchmark and see any improvements +Ive started digging this problem +Just changed the bits at the top of +ll run zkpython tests and post the results herehenry you might want to start adding comments to the file so that its easier for other developers to work on it in case someone wants to pick it up +RaviCan you please check and for similar cases +moving to BETA this is fixed nowreopen if needed +New patch attached which also cleans up the unit test section +Still is there a test which tests a simple entity change like this +MREPOgroupIdartifactId of MREPOgroupIdartifactId bug is quite a blocker especially since ant doesnt complain about some files missing when copying a fileset resulting in incomplete dependencies being copied +Verified in the release notes +Perhaps somebody with more power than me can mark this accordingly. +svn patch for the put only version +Patch. +PS Willem I try to make next changes using git if needed +I agree the best would be to split the tests but for now the only chance is to raise permgen to MB on HotspotI will prepare a patch that rauses permgen for Solr tests only have to think how to combine that in a good way also with the clover special case +This looks like the issue that got fixed in by providing a phase to processresolve the contribution dependencies +apply to scaruntime to scasamples the patch Thanks! I guess we can close this one well create many other for build integration further improvements etc. +An existing output directory will be silently ignoredThe existing output directory will be removed +All of the unit tests now pass with mimej parsing quoted printable base and in encoded text +Thanks Chris +I changed the reassign partitions admin command to issue the partition reassignment request if that path doesnt already exist +This is a Java problem not an one +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +if you are already decorating the docs you get back from your master index before adding them to your specialized indexes then nothing i described would prevent that from continuing to work +Do you have any special servlet configuration for ? Cause it seems to me that your page isnt rendered but straight send to your browser as if a servlet is waiting to provide everything from +This transformer has multiple settings to handle indenting line breaks etc to improve the readability of the generated XML document +thanks ben +All set +Patch committed as svn +Delivered to +In some cases you might not want to have the defaults and by adding them we would make it impossible to replace them +Ah yes a loopback fs should do the trick will try that later +Im still new to Magnolia but I guessing theres already an image library for things like downloads docs etc +Was waiting for the initial feedback on approachThanks the patch reflecting to patch addressed your comments ready for review mostly good +If it does not pass the vote we should merge this toYou are right that this only changes the default value +out of hunks FAILED saving rejects to file testcommonpatching file testgadgetsHunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at Hunk FAILED at +Looking forward to your pulls +I could see exposing the implementation version of the Session. +Closing JIRA. +This patch teaches people why both installations are important and how to use the installers and deal with their gotchas +Right now all the commits for mapreduce are going to common +please review the we should add this view in of Tools plugin rather than in +You have do some refactor to remove the dependence +Sync patch to latest trunk which keep consist with doc just attached +part of this fix for should include the problem in namely dfsread not returning the entire of bytes requested when EOF not hit +Can you get us instructions to reproduce along these lines? Start C CREATE TABLE +I cant think of any special reason to keep it interesting to see if either compiler framework is faster than the other +Username pedrosalazar +patch committed thanks santhosh! +String comparison works for me on with the latest trunkNhan Can you include the output of select from staff and describe extended staff +Just noticed were actually returning a byte array here so I changed this to a long in bfaefaecccbcfabbc +Ill make sure to track this down for key factor here is whether the interface has already been loaded in the main at the time that the class gets temporarily loaded for type comparison purposes +Thanks ATM for reviewing earlier as well +I have set the flag +Good idea +It does more than just String +Looking at the original code it probably got worse too +so yeah +patch to save the yaml file in a configuration column family in the system keyspace +So I have no concerns if you want to close it +File containing svn diff of srcjava directory of package +why check it twiceSame is for null fields these are checked twice +Attaching patch which appears to provide the correct durability guarantees for compressed output +Adding Sylvains patch from as a reference +Hi MauricioYes actually in my current implementation I also try to resolve the variables that are assigned to Human Task by looking at the Process Variable resolving things on the fly it surely can be done but the problem is the user UI is a grid that list all the process instances which heshe may be involved in the future +Will apply after trunk is branched for hbase +Added recursion detection and reporting of recursive references. +Ive reproduced the bug in a unit test and fixed the problem +Its never a good idea for important builds to depend on snapshots that are out of your control +This doesnt need to be fixed in +I have talked to the IP department of the university and they have no objections against the contributionI am going to prepare a patch soon and upload it here for review +Will try to integrate different services in conf +Closing as a wontfixI think this is complex enough that its worth using regexs instead of code +None of the +To confirm this I run the sample program in httpKindly update on this? +The will pass when run as part of the profileservice tests but when using the booking application deployed to will encounter issue SEAM examples to +Applied patch thanks +Thanks to Andy Schwartz for his patch. +I had modified the trunk version of to simply use +Heres a new patch with Aruns patch that removes the group check in because group info isnt shown on the UI in +Sorry I dropped the ball on this one +Moving to +Could you please confirm +Configuration files required slight modification to enable validation +Thanks Isabel! +Note Setting the DMLC cacheLevel doesnt seem to work at all and seems to confuse the problem further +Unless Im reading that wrong it looks to be counting the number of nodes where the last element is in conflict +fot the patch +Jaroslav can you link the JIRA to the issue you created for this? We never did get the two hooked up so far as I can tell + +I have never seen such a case in a production setting. +Invoking the provider via a HTTP GET is fine however using POST does not work +pushing out to Validate functionalityDocument +now that worst case scenario is a documentation issue it is not existing locations of the files and readme are acceptable +Committed revision +Hmm +Scheduling this for next version However if someone can manage it for upcoming that would be better +only outputs names with default type +Rather than attaching src code could you pls fork the project on create a new topic branch on your fork and commit your changes there? That makes it easy for me to do a diff and pull in commits into the main project +Dirk +Its a pain that this only happens on Oracle so hard to debug without having it installed +I guess the only way is to change the interface +If the user doesnt uncheck a box the installer can automatically upload the log stats to a url or even straight to Google Base so that we can later run stats on it +If you go to the main url that the app is on the will redirect to the correct URL automatically +Im closing this as wont fix because it seems that this feature might better belong in the rpm maven plugin. +After oneboth of us gets the delete working lets compare code and post a common solution and jUnits +Looks like this is a patch for +I committed this to trunk and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I did not get why it does not work with partition schema update +Also there is nothing preventing you from replacing it with some other boilerplateI think we need some kind of instructions about this +Let me take a look +I have upgraded the that is in the integration project but it looks dead +May be you are trying to run new test before the patchIn this case you get this error message +Please ignore it +Ill be able to review the patch fully tomorrow and provide some feedback +We should probably avoid having Gossiper inject application states directly if for nothing else than to not make life harder for not sure why i decided to do that +Here is a patch that I had worked on +So without an explicit dependency on the cache service just getting a cache from the container would automatically start a defaultly configured cacheSo now there are explicit service dependencies and we use the embedded manager to ensure that we dont auto create an inappropriately configured cache +Forgot to mention this happens when rendering to PDF +What is next? What URL shall I put in the browser and what class to put the breakpoint in +Patch that fixes +Without multiple class loaders or with multiple JVMs derby uses file locking to prevent thisIt looks like file locking from the same jvm and class loaders does not work does anyone have any idea how to implement a locking scheme that will work from the same jvm with class loaders +the referenced patches +Im not sure if im making sense here but if i am maybe this is the cause of the problem and i guess the solution would be for Tapestry to hold the default value in a serialized form instead of the object itself +Thanks Daniel +Resolving +You seem to have installed RSE over a base eclipse and in my experience this should always work without a problem +Built and tested on XP with MSVC but not elsewhere +My guess is that powermock is doing something quite similar +French and English we add package specific properties in extensions client applications will no longer be able to override these messages globally in their +see also and Olegs +Fixed Toolstrip in Ballroom +If webservice on example is called from two client at same time service became broken +I am attaching another patchthat I think addresses your comments +Any updates? Moving to +We have to fire Metro bug to allow this need to verify whether this is still true with update Metro artifacts Metro stack related jiras previously marked as for community contribution due to lack of interest no action. +It may not all be relevant anymore +Thanks AshishI would love to improve this patch as you guided will update it soon +I will provide another patch to fix The test widget however was still configured to use the old versions and so it failed +Lets see how long it takes now +looks relevant Looking +It is also already in branch the CP branch for EAP. +This would be huge + seems to be the culprit +decided to not supporting the setTimes on directories +bulk close of resolved issues. +Filippo youre conflating indexing and querying +updated patch based on davisp conversations in IRC +Lets get this patch committed. +Please review +That doesnt work in stackedproxy filesystems or anything that wants to do custom filtering +A failing for the bug report + +Repeated the manual test with patch it works as expected now +Thanks +The disambiguation phase work is tracked by +I imagine most users are dealing with quickstarts one at a time but this is good suggestion +Thanks for the patch Clint. +The change gives users a place to specify their own JAVAHOME if they so desire +I seem to be getting the test failing locally as well +Im checking why +When I tried that Maven failed since because wants to compile and ends up creating an empty jarHow are you supposed to create the bundle in this case +Would you please try this patchBest regardsSpark patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully fixed as you expected +This issue was migrated to +Can I please have a few days to review this code and the updated design doc? ThanksAlso I assume that this will be checked into the branch? If so I think we should create an fix version and set that as the targetfix version for the of +I made it a child of +Thanks RobertPatch applied to LUNI module at repo revision rPlease check it was applied as you expected +but now you got the info you needed write a. +Not sure when objectid was removed but it is no longer there on the master or the branch. +Just add an assert to check the return code in the new test and Ill commit +Here is a second patch with code cleaned up a bit +IMHO downgrade to minor is unwarranted WSDL and XSD are often supplies from external entities leaving a significant user base without a workaround option +We had to set the environment variables and +Applied in the branch needs also the patch for contains the fix for. +Proposed fix +Recent requests plugin display in revision the web console includes a new plugin at systemconsolerequests see screenshotThe information thats displayed comes from the if we need more details we just have to add them there +Thanks Stefan for the feedback and suggestions +I thought it would be a problem with LuceneThank you for your help +Committed to. +Nobody can find the way to become registered in Xircles +Using console handlers in that case leads to batch script waiting for Terminate batch job? YN +This bug also causes a x performance regression in due to lost opportunities to batch +However I really think of not matches as being another keyword like excludeseven if its not a new keyword in the the risks of adding a new keyword theres not really a standard antonym for matches +Looks good to me +We need to fix them properly +The fourth consumer should of course be included but its lack of inclusion should affect the test +Can you submit an example form you cant validate +Why cant it use the regular Hadoop RPCAltho at this point I just dont care about this jira +Please verify and pay attention to make sure that commit notices are going to the developer mailing list +Ah scratch that we need to check if the leaf of the conflict edit branch has been deleted of course +Thanks Krishna! +This patch has full readwrite functionality +for code reviewtest suite ran clean +Dmitry Im using hsqldb and using core configuration type +Menashe adding an example +Jochen may have additional mods in mind so leaving open for now +Here is the updated patch which has srcmain code exactly like Omkar suggestedIve tested it by using a pendrive to simulate drive failure and the file is indeed localized again +My favorite is the first one but I left it up to you to decide if any of them is worth committing +Thanks for explaining Grant +Assume closed as resolved and released +I manually installed the jars according to the README but that did not resolve this problemInterestingly there is a dfference between executing mvn install in the stanbol directory and the kres subdirectory +I agree with Kim here I think if you want to step on doubles with doubles the number on which the step method applies should be a floating point number too shoudlnt it +I may have screwed up the xerces dependency when I completed will fix this for was able to reproduce by using the standalone jetty bundle I wonder whats going on here? in +Ill run the tests and let you know what I see +Not sure about enough feedback but I am exclusively using Eclipse and would be really missing Spring IDE in my configuration +See tag reference for details. +Another optimization is their anyone interested in optimization streaming renderer that can pick up these jira items +Ill include him on this ticket +Committed trunk revision revision after release. +committed fix in revision in with in with r. +Attaching which fixes this bug +for going back to the original behavior +do we really intend for these to not be reusable? i think the only problem we should protect against is using them for two concurrent calls as opposed to making them +Can you please whats happening there and submit a reproducible test case if the problem remains? was never able to reproduce the issue with a simple example +Bruno can you please review this +Im happy to see this committed +Thank you for the review +As mentioned by Siarhei Dudzin maven constructs the classpath differently however the dependencies still appear towards the end of the am trying to understand why this is the case +Thanks +One parting shot hereI just took a look and it appears that there are three uses for lang in Mahouta as a way of getting atomic primitive types +comments on phabricator +Fix added in revsion +If I am not mistaken it is not getting as far as the original error +I believe this is a duplicate of please verify +Do we need to sort datanodes by distance? Why not just do a linear scan for the on node and on rack instances put them at the front of the pipeline and leave the rest in random orderAnother option would be to do a linear scan and bucket nodes by distance sorting just seems unnecessary here +This is fixed in +table +The root cause is by in revision by granting anonymous read access on as per. +So I reworded it +silly me +Will test with larger collections and String Reader +This patch fixes the issue +Spatial on a class without means that the Lucene document field prefix is the fully qualified class name +So close the issue +The test report web site is not available at the moment +Thanks Cheolsoo for committing +Closing this issue now since I have put in all the changes I intended +Minor updates +Perhaps clicking the reset my password link made things worse but whats done is done +Alright then lets get Todds patch into +Attached patch a +Templates registered failed in one of post processor step so virtualsize is not put in causing template sync cannot find the template in secondary storage leading to another download +QA please verify. +Initial refactoring from will be done during current issue +thats exactly what I figured out +Specifying on the made the filter solution work + +This has just reoccured on a different client site +All integration tests pass on +ant instead +Ive created fixed. +DeletionsD mathsrctestjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathfunctionD mathsrcmainjavaorgapachemahoutmathfunctionD mathsrcmainjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathmapD mathsrcmainjavaorgapachemahoutmathmap +And maybe a different name than I might be biased but it sounds similar to where we have a potential set of providers and not just a single one +Now I have created the patch against branch XBRANCH +Yes you are right if there are such jars +This ticket is reporting that the empty directories stick around +Fixed now and committed with tests. +This class will also be largely deprecated in favour of an approach in Spring Security +This looks good +me sweet! +I think this is done +Upgrade libraries to latest releases +Igniterealtime has just released a new version of their Smack library with support for pubsub +In any event the test case for that issues does not reproduce the reported issue with the current fix level. +This only affects version +Not sure what more I can do here +You can already avoid automatic population of a form bean by simply notassociating a form bean with the action mapping +Is there some way to force Hibernate Console to scan its configuration classpath for entities +removed the return with revision . +Fixed in SVN rev +It seems to be related to Hibernate very aggressively closes the connection that has the open resultset connected to the iterator +You can also just replace the dist directory with the. +If one is found an invocation object will be created without any argument information +You patch is in trunk rev +Brilliant thanks Dan Just tried it and everything works. +DIR is for general issues not relating to any particular project at Apache Directory like the website and PMC related issues +If the assignee thinks it can be finished before the release please note here and notify the team +Thanks for the review Andrea the patch is committed as of r I will leave this bug report open and ask for feedback on the user and are indeed only deprecated at this point in time +have decided after all to reclose this and instead to create corresponding subtask. +I noticed some schema validation errors on the component model output and will open a JIRA for them. +Not only from time to time +Two attempts to start in attached log excerpt +However it is not as convincing as suggested to change a feature affecting user behavior simply because someone has reported on the mailing list +verifyTimeout? There is no such option for but there is one in +I will look at committing this patch +This will of course expose any connection leaking in underlying connectionprovides if you have enough tablesThe bug is fixed in svn +Will add a specific Default Gate editor which can show all outgoing sequence flows of a gateway and allow users to select the default one thanks for reporting +if anyone has a chance to peek at it over the next few days I will be available to respond quickly to reviews for now +the Java Home must point to the root of the Java folder +Can you possibly attach git status output so that renames are easier to track +The contains the same module as offered by the other patch but depends on the patch from being applied firstThe recommendation is to use this one instead of stated in the comment it is recommended to use the patch which depends on to resolve this issue +Great then just let the end user use the ROME api and seed that ROME if the object is eg simple and have simple setters we could support the map also and use to set the values from the map on the object +The big savings would be all of the ASs deps but getting the versions of all of them is trickier +So for example if I create a Topic with the following valuesTopic Name topic Binding topicBindingthen after the topic is created if you go back and look at the configuration it will beTopic Name topic Binding topic applies to both and Queues and this jira explains some of what is going on appears to be deliberate because the JNDI Name attribute was unset +No link was given for Approve order as order was already created in Approve order status +So I looked into my browser bookmarks and found this very old one +Perhaps it was inadvertently scooped into the patch by your patch creation tool? Many thanks for this enhancement +Those failed tests passed on my local box +Krishna Im resolving this as a duplicate of +Also please complete toXml and fromXml methods +Is this for the Camel REST binding or the binding? The referenced forum thread seems to suggest the latter +passes for me +The commit comment was as follows Wrong join column returned by right outer join with NATURAL or USING and collationAdding comments to some of the tests stating that Derby is allowing incorrect behavior because of +Please reopen if otherwiseThanks! +It has been made against to master branch in +Agreed the cause is the sameClosing. +Pete do you think the current weldmaster is compatible with Glassfish? I mean + +But now that you are at it you can have it fixed +Love the idea +The Provider stuff in puts the developer completely in control of pretty much everythingHowever you can turn on schema validation and it would reject a message if it cannot validate it against the schema +Sorry I have overlooked your changes to CHANGELOG +Copying the bins to the endorsed folder did the trick. +I just committed this +Had forgotten to commit changes for last patch +This issue is no longer relevant for it is fixed inChanging the affected version to. +Could I ask you to attach a complete stylesheet that demonstrates the problem? Otherwise if Santiagos suggestion resolved the problem please let us know so that we can close the bug report. +Xavier can you look into this +Should be fixed in revision can someone verify under Windows that mvn clean install works in extensionsaptparserPlease close this issue if that works +We could try writing our own very simple setsid program or ask Mac users to install such a binary from elsewhere +out of of a bulk update all resolutions changed to done please review history to original resolution type +Martin you actually cannot verify this +Hi BenOh well I guess this is why such testing is so important +udpated math R code to match working notes more closely +I have committed this +Hi Paul no it doesnt really affect the status of +r. +See srctestgenavroinput cleaned up the +This solves the above issue +M trunkcomponentssrctestjavaorgapachecamelcomponentcxfM trunkcomponentssrcmainjavaorgapachecamelcomponentcxf +Like in a GORM for the clarification +We use version ranges in out projects and thus cannot easily reproduce a build +The change to is just for optimization +Nice that all artifacts is in maven repos +Committed to trunk +Its unlikely to be PPC weak memory problems since is running just fine with multiple virtual processors +Upgrated on trunk as of revision using snapshot +Thats greatThanks Stefan +temporarily disble the test +They are just questions + +but Im going to see if theres any way we can make at least backticks interruptible +Thanks for reviewing Donald +I suggest I commit the patch as it so all the folks out there could test thelatest fixes +A database UUID wont work +Patch applied at +Its great to see a prompt response! ThanksFor now would it fail also if I use CLI scripts +So far all our configurations have bag semantics +A alternative would be to ignore the negate option manually in this iterator and document heavily that it is a poor solution +has been resolved +Add some base test cases for Sequence Versioning +I think we should just leave this whole issue alone +fixed in adfefbbaadcbfa +patch the patch thx Upul for the patchDJ +Thanks Arthur for your valuable comments +The reports in this case are biased IMO wed get a much better picture by going through logs +itll likely only take you minutes and you will never get a ticket like this again +Another point a distributed search across embedded cores should be serial not parallel. +Bulk close for release +The category attribute must be the id of an existing preference is fix is to remove it +Implemented in rev +Each release of JBT will require a new client jar so retargeting for +Now fixed for me since is also fixed for me. +You are right in that exists only to catch exception scenarios +Thanks for your patience + for patchneed this. +All the memory is eaten in seconds +its seem that when i got a select item with a content like kdjakfjda fkdjsfkjdsa jkfdajkdsjfjdsakfj ajfkdsjfkdsjfka fdsfkdsafasdfkdjakfjda fkdjsfkjdsa jkfdajkdsjfjdsakfj ajfkdsjfkdsj i get a error that make it disabled but with a strange comportament that sometimes work and sometimes not +So an can return false from recoverTask to indicate it doesnt support it +I think driftx on irc had a newer version +When I will have commited I will migrate it to jQuery +But doing this will not include file +Looks good +Another option for us would be to set the result as bytes or some other value type +committed in +Thanks for logging the issueNo Enunciate only recognizes annotations right now +This data is not available at time its only available after the restores its offsets and registers with +Ha or rather than with your latest code we can the webapp example in birt component right +Log files for both scenarios are included +If the test works in direct mode Ill tweak it until it fails in direct mode +Third batch of graphics +You must not set it to WORKSFORME otherwise the bug wont be fixed in Cocoons CVS +Ill attach one below +comments that its invalid +committed to trunk +is not launchable from Java editors only and xhtml right? why not create a command context or whatever eclipse calls is it and only have it enabled when that context exists +The status code not being does NOT indicate the body is empty examples being s jedi page and Twitters fail whale page +according to what iunderstood is We have to create a tool to get the daily information aboutworkflow of the application +I did some more testing with this and everything works fineApparently there are no tests for importexport anywhere so I am adding a test for existing functionality as well +But the approach of doing db versions checks in readwrite methods I believe isnt a good optionI shall move onto extending to cover this caseI am currently running derbyall with my patch and shall attach it soon +Before I was only looking at commons investigation +For larger problems additional tricks will be necessary +We do not support multiple count variables or anything like that +Found two typos causing the blob references to be compared instead of the blob bytes +just delete commented code if it useless or make safe code in comment +This is a patch for +Wow! That does sound significant +Thanks +Some just require a dependency on which I guess requires an update dependency update in the pomclasspath Ill look into thatOne that requires more effort to fix is the import of tuscany classes into the generic project +Attaching final patch. +I also got this you will find the patch to fix it +Thanks +Thanks Dushshantha for helping fix this +I just committed this +Its fixed with revision at trunk + modified in revision +Can you please confirm +I have added Jsch JAR and it am able to solve the issue the issue because of special characters in password like +Today I cleaned out my local m repo updated the svn source code and ran the tests just inside the security module +If you dont mind +I tried to verify this but am having the same problem as with +please could you review this code +All of these properties should be deprecated in and dropped in This interface and implementing classes are postNo need to wait for to clean up +Can you put more detail in descriptionIf you can attach what was done in branch that would be nice +Thanks Namit +JBDS says it needs a specific featurepluginUser decides to install specific plugin that is higher than what JBDS says this errorHow do we ensure that doesnt happen for our SOA dependencies? Here its just taken the latest from a build instead of actually using what is required from JBDS feature perspective +I hope to have a chance to close this up by Tuesday evening +Those of you that voted for this issue and are not happy that Im not fixing it comment here or email me with details on what you were hoping to get out of this issue being fixed and I will do my best to help you out +JAXWS is the natural replacement for JAXRPC +With this patch they would practically become part of the API +If that is the case then I will really have to step back and think about my approach for graceful degradation on the main navigation since the key design principle for this section is based on the fact that the list falls into columnsI am curious as to why you feel this should be a key objective +Thanks Chris for the backports and Nicholas for the additional review +what do you mean by current working directoryWorking directory of the task local file system +Can we just set them to null insteadbq +has been marked as a duplicate of this bug +It is needed only for JIT interpreter doesnt need any prediction because it just executes bytecodes one by oneYes doesnt work for JIT yet +Please verify +Andrei I guess you try to edit TLD form jar +Hi AmarIn Hadoop in the Job History file the task tracker information of unfinished map task is stored in TASKSTARTED and MAPATTEMPTSTARTEDRegardsMitesh +Without in play +Dont forget to add line to when and closed at revision +Why including the wrapping tuple in the schema used to work is still a mystery +initial version complete +For those of us who dont care about ivy in fact lose something a easy to build hbaseIvy definitely is not a silver bullet to dependency management issues but it makes a honest attempt towards that +Existing tests apply +Patch for trunk and +Do we still want to apply this? The WAP extension was deprecated in +ok thanks for the clarifications I will take your suggestions into account and provide a new patch +this be marked resolved now that the code has been committed? +All the parameters have default values +Do we agree that something is broken? I never said that adding the block curlies as a workaround is a strange fix +Do I need to use this as a repo in or just download the rpm files individuallyIf I try to list the directory I get permission deniedIll have a look next week +Patch added some changes that were lost btw and broke the patch in an unrecoverable way +Moving this to TBD until we are able to confirm the issue still exists +AndrewCan you take a look at patch v please +Eugene it would be better to create a new JIRA ticket for the compression as it will show up better in the release notes +dims +Attaching an updated patch where Ive added the ASF header to all files except a few ones in the clients directory +This issue hasnt been reported during the past years so Im closing this ticket wont fix. +Reducing priority +This month and next month are a little busy for me +but the configuration you cited would still work without the XSD definition. +Any chance you could test this with? The compiler stuff was completely redone for so this may already be fixed +It affected both version +You can loose that lib only if you are not doing anything that Groovy does or by inlining +An Array instead of an iterator sounds just fine to me +We have to decide when to bring the parameterized version of Safegroovy back into the tree +Same error with assume this test was run from the package of which I have mentioned on the QA qwiki? It seems to be running fine on my machine though +Reopening to add javadoc at least +Eclipse will sometimes keep uninstalled plugins around but they will be inactive in the running may be the case that by manually deleting the folders you have corrupted your eclipse installation +Let us fix it in the +The values entire length must be parsed and the resulting number be between the max and min values for that type for the number to be deemed as a suitable number of that type +Thanks for the comment +The patch doesnt apply clean on other branches will handle it in a different Jira +I think its a good idea to make collections available as entities within the UII think the right first step is to add a UI for the collections API similar to the one for the APIFor UI stuff though Ive been depending on the kindness of one stranger mostly +As well with SSD so on we might expect the object creation to be more more important so starting to use likely a step in the right direction +projects that want to become Apache projects need to go through the incubator +ETA underlying problem here is that when using a client pool size with Remoting invocations can arrive in a different order to which they were cannot just set max pool size to since the current remoting pool implementation sleeps for second if a socket is not available p this would cripple current workaround for this was flawed and causing the have implemented a better workaround and am now running the tests +Rebecca Ill give you as detailed desription as I can but I need to consult it with someone whos not here right now +From what I can see he problem originates from get +Jozef did you write this test +Added a new testcase to the extended test suite to force initialization failure of a Fix Version +Thanks Sanjay! +What exactly is it going onRe For the Map change couldnt we update the testcases in the branch? Well I have already! At the beginning the testsuite from EAP was being run where weak hash map was still used but of course I updated the test in the The QA lab has just been taken down for hours +attaching osgified Maciej! +Has already been implemented in the release +This is only partially implemented for the +OK I see that this only happens when Ive established a session in the appcontext +we want the resource bundles to be loaded from a content management system using HTTP +As only one method in the web service will have that name this would beter satisfy the requirement in Basic wouldnt it? +Show problems with jira has been closed as part of a review and archiving of the aj project +This is not critical +Who proposed it I will kill himher +Attaching the project proposal +It continues to do so +Hi Jaikiran ive replied your answer at the thread could you please answer? thank you +Its also not released yet but should be functional +Yeah this is awful +Checked in fixes for +Alexey is your a custom implementation done by you or you using the Xerces implementation of +I realize this issue is a bit old but maybe worth noting prior to v DB really wouldnt let you put ORDER BY in a subselect +in this test inside run method it is throwing the exception This is happening some timesI added some wait to complete all the threads before assertionThis is happenning randomly +The only difference I saw is when after I applied the patch VM started to support Xinvisible and printed out invisible methods +Insane it is +Rodney if you are still looking for something to do on this maybe you could try to figure out how to enable ECHO of input when executing a blocking background process from Forge +If were storing doc ids alongside the SRthe docs correspond to theres a lot less to worry about andjust seems clearer +Will run full build and commit +As of now one has to manually set these filters to correct maven generated +The code changes also look good +Issue closed with release. +r improves the message +Meanwhile Ill look at the parts that fix compilation warningsThanks get javadocs warnings whereas you get none +Displaying what the user has entered is the principle of least i do not like the idea of adding a condition to my three dozen of variables but id be more than happy if we had this documented my +Removing Alpha from fix version so this can be properly verified in Alpha +Ill wait for Knuts tests as well then commit it +Lets go for the refocus this solely on the WF use case +The only thing I can think of is that it may be related to the libraries Im using +Hence Id prefer the functionality to stay in m unless there is some compelling reason to move it to mavenproxy +Ive removed the deprecated method that could have caused problems +couldnt get the original patch to work on the trunk or branchx +Let me assign this tomyself + can be applied to the other branches +Unfortunately I have no time to dig into this further +Verified with JBDS Alphac +Removing Solr is a bit confusing is a module in the morphlines project this is a morphlines module with stuff for interacting with Solr perhaps we just call it morphlinesmr Seems we should rename this +It looks like someone else took care of it. +Thats what I did in +The fix for this must have been outside the parser +In this benchmark for each vertex the compute function sends a meaningless message into all edges of the vertex +gig +as it wont be clear and would conflict with other namespaces +This was the problem +Yes we could easily add a number of additional overloaded methods here +TODO need to update the tickets +So looks like we need some distrib tests for this to start though it looks like the single instance test fails with the latest patch +As David has suggested that there are not so imp scenarios in which this feature can be used I would suggest to use this with a cluster setup to get the best out of this feature +related to +the code referenced in the patch above was added to fix of the only reason for the it seems is to make the hibernate annotation work consistently across different +Please hold on +bulk defer of open issues to +I dont think this is a Groovy bug +It sounds like a regression to me +Im a former of theoriginal submitter +I believe that either the scheme or the authority includes the user name allowing it to be multiplexed on that +This should not happen with the v patch for since we would serialize the requests from a client +If you dont the dependency from B in the cache is used +this already seems to be the case in the current shell script +Thanks. +you want to run it +Yep +Certainly my project didnt have build result +Click on Finish and Eclipse will import the projects from svn. +I havent created a unit test and patch for because migrating the test scripts from Junit didnt seem worth the effort +This should work right +Apache JIRA is sometimes quite slowAny chance you feel like attaching a test case to? Doesnt have to be a formal junit test just something like I create a node with these properties then I do XYZ then FAIL +Retrying same patch +A master cluster could synchronously update its peer clusters when addingdeleting a peer and when setting the replication scope on a column family +The NMS clients are documented separately from as they are not entirely the same +Block report request is changed to send report per storage +I made the Activity class contain only an Actor Target and Verb +otherwise +i need to know field and text for each matched leaf scorer Why? For scoring purposes? has a configurable to control if you want the max avg or sum of the child matches rolled into the combined parent score +Thanks Sean I see the change +Looks like an Arquillian issue +Hi Jason this is dynamite +The suggested fixes involving XML Schema may work +Moving to a tmp file in the middle just creates another failure scenario where the DN crashes in between the renamesCan we just rename directly from the old genstamp meta to the new genstamp +I originally used multiple trycatch blocks to get a different log message for each possible failure but just including the Exception object in the log is sufficientHeres an update with unit test +The idea is to reuse the same endpoint but being able to send it to a destination of choice +The feature is not currently on our roadmap +And this effectivelly disables using Sonar for php +Need to create a test caseapp and see if its Eclipse or not. +Pushed specs to trunk in ca +Rather than adding this null check and ignoring the error silently it would be better if you added back a template so that that file gets built if this is possible after the project split +We should make sure also gets this fix if its broken there +Must admit that I dont find that the most natural name +That sounds about right +Fixed with JANINO. +Prob not necessary as most deployments will prob be on server but future versions might be betterSeems like a little win on nix systems anyway just from the CPU savings +I figured wed just check in the source in some side folder in our SVN and only post the +Sorry to come in late the patch has gone stale +Ive run gwt client applications multiple times only on Ngnix or Apache Httpd without any JAR +Thanks +Added the user manager as bundlesjcr in Rev +eclipseplugins +has been filed for this +I think that as a critical bug we should try to solve it +When you are doing Java Schema you are only working on a subset of schema features +Always works fine in my IE +xls format thats compatible with all versions of Excel and the newer excel XML based format thats compatible with Excel +Victor please add tests for this new resolver +because it will result in a browser redirect and a request +Thanks Pierre for the feedbackSo I close this issue now. +If I finish it before you Ill submit the patch if it does not bother you +With the patch casting is only required when a variable needs to be eligible for implicit conversion within the script +Seems I cant editrename sorry but yes a generic way of merging artifacts would be fine +A few good things we could get in related to this +ThanksThis is a property that can take values other than test but currently we only support test but in future will be used for other durability modesTherefore I dont think we should say This property can be used under testing purposes to improve performance but maybe instead to say something similar toThis property can be used to change the default durability of derby to improve performance at the expense of consistency and durability of the databaseCurrently the only valid supported case insensitive value is test +Try the latest snapshot +Hah +When I was working under this bug I found interesting thingwhen used xml mapping discriminator is required for such hierarchy mappingbut it is not required when annotation mapping annotation mapping hibernate itself creates discriminator value with string type and DTYPE column when Discriminator is not we have such inconsistency +If this still fails in the rebased versions we can reopen +Instead of selecting random nodes give priority to the ones with lots of free space and few is an important problemBut sorting for each block allocation seems too expensive even if done n log nShould we just permanently store a or a of nodes prioritized by their remaining spaceThis would require updating the map when changebut will be less expensive for allocations +Thanks Colin. +Ive double checked +I will fix the test output and commit +This makes adding new fields problematic and we need to add new fields to both stats and quota nodes for throughput quotas and we need to add new fields to quota nodes for hard limit quotasFortunately the parsing code in the constructor leaves an opening to add new fields if we dont use commas and string them after an extra sign after the bytes portion of the string +HelloIt works great now thanks +If so add the string and object to the Map if not push a new Map to the stack and put the String and object in itThe only minor wrinkle is that we might want to put our own custom Map in place of say a though our own Map would have no methods +Selected should read checked +So all issues not directly assigned to an AS release are being closed +Verified in JBDS Build id to add fixVersionThis was included in via +has a lot going for it given that groovy is used despite the Sun naming convention +create production +I see that I was missing your point with the common base thing +This was an oversite +Regarding the status of the issue I thought this is resolved +Can someone provide an example of how a Solr response in any of these formats could be problematic? All strings are escaped so code within a string value of a field would still eval only to a string and would only execute if that string got evald +For anything else it is quite normal to have it in srctestdata +I was thinking of debugging this issue but I cant reproduce the upgrade test failures on my machine any more +Im guessing that this may not work in oracle at least with the current version of and local install shows these changes as fixing the problem +Revised and posted +Maybe I can be convinced its not needed but if you look at a tarball youll notice all the srcs are included and the build file so if prob +add Maven project sources in +are not in a secure place +Isnt there some way to qualify that the script needs the groovy class instead of a variable +But via arc patch is not applying cleanly +It seems to me that the test has little value and should be removed +the tests does not need to fail if the tests want to test the examples working +Older versions of the protocol have weaknesses but we can enforce TLS Kyle +not sure if its the right place +What exactly are environment proxy settings here native OS settings? Also the are usermachine specific so why is there a need to edit the files every time +Remove it to reenable the test +I think two audit logs may be okay +It was just running slow and no noticeable run Failures Errors Skipped Time elapsed sec +Thanks +path ownersThanks +no need to mention clustering key? is that considerered part of the column +I have applied the patch Michael +Map is a natural fit for that and it then occurred to me that expression support for maps would be useful +HiRebooting the database doesnt change anythingCorrupt means that the statement above worked for several times until the first exception and then never againIll attach the corrupt database to the issue +Reopening since resource resolution is broken after last change +Are you working on this? if not can i take over +i +Glad to hear Rohit +Ive found a bug when I benchmark with large scale data +The test cannot pass issue and are not test cannot pass issue and are not removed because other issues need to be corrected and someone else could take this issue to test blocking issues have been resolved and all tests now work. +Added this ability in r +Thanks Shawn +Fixed in +Fantastic +We have working tests for similar circumstances + +Does anything in this report ring a bell for you? Like knowing more or less where the issue could be in the code +I recommend we resolve this as wont fix or the handling for sf over to procedures and not tables since soql lacks inline view support +I should go through doxia and look for blockers as well +We can add more constraints later onOk will address that in a follow up jiraThanks for this enhancement +Id suggest to declare methods parameters final it will improve readability and make testing easier +jUnit test case for this issue +I just had another look at the Core doc and saw that the CURIES are defined without spaces +waiting on software to be completed +Go for it its a private impl class what should we do else +add more info about editor +Len please see the associated issue which IMO duplicates this one +Added snapshot restore clone time fixes the metricsMaster vs masterMetrics and vs metricsMaster m planning to commit this one today any other comments on the patch +For me it is very important that the check is disabled in Jackrabbit because it is a major performance problem in my environment. +on addendum v +Fixed with revision of buildersesb message Comment out logging service in jBPM configurations +Ok so my idea doesnt make things any more stableWith my change things start to lock up around K iterations with the existing setup I am already well past that with no sign of slowing down so I will abandon my changes +Now I merged the previous patches into two patches + remove the lease if the corresponding is not found +thanks danielsh done +It could save the history on the dfs and on top of that you would need a wrapper around it to return you objects +On second thoughts Im going to rename to +I applied the patch and verified that the output on my server looks fine in IE too +Though Im not sure what the best way is to improve the situation +unclear what this is about. +Can you review and let me know if I can close this off +I figure a is generally useful beyond clustering +corrected links three broken links scenarios and published it to the webpage with revision +for just committed this +I am not getting a stack trace rather some command not found errors +I want to add an optional module a class file reader and other things +Rev has a better way to figure out if the source object was custom marshalled +The Wiki page of Pig provides quite limited informationBTW I assume you would be the potential mentor for this issue rightRegards Wiki page of Pig does provide guidelines how to write a application +And it is a fixed preconfigured stack of searches it cant do a dynamic number of searches +Ill take a look at this as Im working on related area +Teiid doesnt support EAP AS yet +The purpose of my request was simply to let you know that it could be useful to increase the default I cant reopen an issue btw +Must do before RC for inclusion in Windows release +Dont see how to progress without more info. +Thanks for the test case thats a big help +Closed by mistake sorry +Yes +Lets assume predecessor nodes taking the action of replication +I am not sure but I will try to create a quickstart that reproduces the error +I have the code its just not in junit form +messes w get near flushes +The changes look fine to me +Based on the commit date setting fix version to am not seeing Use File System hints in ER build +He said we actually dont have a choice the fix has to be changing s MANIFEST +to break the ice rompre el gla +Ive just updated the committers list its fine now +This is a well known issue with WFS all outputs go out with latlon order because the transformation happens before reaching the output stage whilst in theory only GML should have flipped axis +Correct the typo +Thanks Bill +This patch has been applied in revision thanks Fred +Ramnivas this seems to be a JDK compatibility problem in +This was resolved by a previous pull request Ive lost track of the specific one +I believe this is a duplicate of issue +And when it is pulling the contents out like we do will cause the datamodel to be unusable +Reopening to mark release note exclusion since this was reverted +If the rename fails the will instead try to copy the file to the renamed file and then delete the original file +I think it makes better sense to handle this concern at the step level +Description of patch Remove processing to expand data from of blobclob to memory before sending to client +I am trying to use element method now to generate the fpml schema the automatic conflict resolution does not seem to are thousands of binding elements +thanks steven for figuring out how to fix this! Simon for figuring out what the problem was +This is verifie but why would you exclude a dependency that youre adding after? the version will be picked from the one in your dont need to exclude it if youre adding it again +So it must be something about the tutorial +Will update the a new patch which adds test case for ledger id generation +I should have been more specific +implemented as add +Revision Revision and revision add the root tree to the the issue +as I need to rollback part of this patch to prepare for a new release of the I see the shared components are now released so closing this once more +I will commit them as soon as I go through them +There might be something else wrong with your repository +by setting the qname using an +Would really like to see a checkin on transaction work first +There are two places to be fixed +It only allows base and specific exclusionsThe new scheme only applies to subtreeSpecification attribute of subentries +Incidentally I know the JDK is old +Ill try itand give you feedback. +jongjinPlease apply you changes to CVS directly and add a test case as well +I think we should go for streams or we pass in a lucene directory while I think unless really needed we should stick with a stream +NA post mavenization +and give up on having support for long lived sessions in the frameworkYou can have long lived sessions but not the ones created through the authentication mechanism +Resolving as fixed +Lets finalize the patch and continue building upon it +can not recreate it +This file is the patched up version for package +Notice that we are using the you fix this theres a typo in the error message should be retrieval not retrieaval +It includes a unit test fixes the error message in and updates some Javadocs that I had previously missed +We cannot close the Database ever +everything rsync Pv gives you +weve certainly oscillated back and forth here quite a bit +Will look into this just have to have consideration for +Had to make lockunlock responsibility of individual Migration classes since lock is no longer global +Hi WillCould you let me know if you have a patch for this yet? I guess it is feasible for now to use head n and note that some gc options can cause problems +Attaching +Should this become a c issue +Let me check this +The directory javasrc is empty after this change I could be removed +Attached is the +Just tried again on my local machine using a nightly I just downloaded from gridlock +This issue doesnt exist in current Axis trunk +TIA. +Actually too late anyhow as you have already committed this +Find a revised patch that appeases +PR for master merged in master +Thanks! +The birt addition is not yet in the trunk that will onlyhappen if the community agrees with it because it rather a largeaddition +Here is the just committed this +Thanks Nicholas. + + +The problem is that the superclass proxy is being modified at the same time a subclass wants to add all entries from it +Include this in. +The toString may not be the nicest but it definitely worksThe only test I get failing if I use it is the +When I assigned and scheduled this it wasnt my intent to require anything complicated +Thanks for the contribution +committed +If the problem happens again Ill add pooled true to the s dataSource sections +However I need a little help +Thanks +I can do thatI have a java class which can serializedeserialize a usingwhich I used to send response and deserialize it backI can post it as wellNoble Noble Paul +Please confirm it is applied as expected +Looks like the result of an overzealous merge from +Fix will be in IJ has been updated to use IJ Affects and Fix in Version +Im going to take this issue Im applying a variation of Alekseys third patch +Ray JacooI finally commited it in revision Thanks +Attaching patch that makes distcp to check if the destination is a file and is expected to be a dir and emit a meaningful error message from setup phase itselfPlease review and provide your comments +Improving the services would be another issue +If you create a command that does just that and a script that calls Initialize and that I think that would be fine + for the patch +That definetely can you please take a look at this +It might work I havent tested it that much yet +Note that import package is not enough since the interface pulls in a lot of classes to name just a this helps +We should perhaps even configure this to be used by the example server by default given that the performance impact should be negligible if NRT isnt being used +Chris Lalancette just pushed a patch +I just committed this +Also removed the spring version number from karaf running clean install before push +the workaround is mavenupdate project configuration but this resets all project customizationsI feel this bug as blocker because the developer dont recognize the classpath problem he writes code without to know that maybe the classpath is schedule the issue or give me an hint where I should look for fix it in the meclipse sourcethanks +Ahamdis modules field a dropdownis environment field a dropdownThanks now just output the text and do not make them environment time to service breach time are all been displayed for checked into cvs for both reopen if you find issuesproblem with the change +Can you please try with the new nightly build it has a number of fixes for +No we wont be reverting +This is nice feature +I think as Roo becomes more popular this issue is going to keep coming up +The only remaining feature that seems useful would be a system procedure or a system table to get pathtype information for a given xml document model element +It is pretty complex and hard to understand +What should the default property values be for destination and destinationType +My suggestion still stands though +One possibility would be to fetch these checkrequirements scripts from the tagged version in git rather than bundling them into CLI +I still get the message so I must have a dll version crossed up + +So all issues not directly assigned to an AS release are being closed +issue should be resolved now +will take a possibly check isInterrupted in addition to active in the if care of as ll resolve this after merging the fix into the SI branch +fixes this issue +The ticket has been updated to remove the question +Reset only inner component state not model +Ok thank you +It looks like the synchronized block was already removed in +Besides that I can suggest log messages and debugging +one minor issue is that get is abstract so we need to implement it for both and DFS +Hi Micheal There are many lists and i am supposed to join The Xerces Java developers list which i am unable to join +for the patch +the only way I can reproduce this is if I type into the project field +I downloaded the and it has +Look at my comments in with regard to how the proposed patch fixes this issue +minor an extra space in indentation inOtherwise the patch looks good +There appears to be an inherited justify +first time started at finished at this is the project cache +Wicket is not assuming that everything is writable +The smooks editor is no longer being developerd so moving to LATER as agreed with +Fixed with patch applied for +It is a follow up to this issue +good to me +Isnt this a critical problem for Geronimo given that the Orb is shutdown if ANY EJB based apps are stopped or undeployedSeems that this would be a for any developer or Eclipse user who tries to do a redeploy while developingtesting their app +This will not make it into the first alpha without some discussion +Viliam is doing additional applying the patch im unable to reproduce the issue +Thanks Jimmy +JDWP agent uses internal daemon threads for handling JVMTI events including VMDEATH event +Henrik the way you describing it look like we wont be able to do anything about it +Fixed problem with DA +Manually testing was performed and found to be working as listed above +Closing this issue since someone already brought javassist into WF to resolve a JDK build issue. +is probably better for time series collection +Oops +I no longer see these in the logs with the latest build. +I guess the code needs be be modified for requirement C. +I have tested It passThere is some reason for add attribute +Youre almost as helpful as Gavin +Rewritten as per comments +For example the application is used to lookup the on the node that we to +We should have committed this one to the branch +Closing assuming all work has been done for this issue now. +Please verify that its as expectedThanks +There were actually more issues with and Comparator than met the eyes The Comparator did not work at all correctly with the order flag The property was used for comparison whereas the property should have been used The fallback check for the property did not work as the service ID is a Long not an Integer which was assumed The count was never updated locally leading to too much service reacquisitions And of course the class cast issue reported by TomFixed in Rev +Next question is why are required? Afaik this wasis not the case with online +bq +If youre able to reproduce this with or later please reopen. +im still not getting this +patch v all the tests pass +but it has decent autoconf comments and readme files and has been working in production for quite a while + +Added clarification to the javadoc in rThanks for the report. +I created the bug to capture this. +looks good +bq +Not sure that this patch addresses the questions in this bug + +Will file another JIRA for that +Note that there will still be failures when running underComments Note the method duplication in +Most of these other cases are trivial an example being KWELEMTYPE KWKEYTYPE KWVALUETYPE identifier and they are all correctly handled by ANTLRI have created a followup JIRA. +I believe the core storeless engine has been added +Uploading a patch and a jar file for this +We need to make sure we have a proper state after broker restart and work well with mix of and durable subs on the same topic +Leaving this Jira as there are some other potential changes that will be made based on review +files in +Hi AndreyIm not happy with the solution +Hi HairongActually i was also started looking into this +This patch updates the attached to apply without error to revision +here is a first patch for review +Thanks for looking at the patch DagCommitted revision . +Please file a separate JIRA if you still want that to be addressed +Tiles released. +In particular threads could have blocked on a read if the socket timeout was +LGTM +Looking good! Minor nit Considering Community is one page it feels weird to have a drop down of links that all point to the same page +Should have been closed in ! +bq +We believe this is fixed +Making opcodes universally uniquely is unnecessary +Updated +I dont think it needs a separate issue +again if you dont have a small failing test case then i cant do much. +Thanks Michael! +simple patch +Ive moved to. +Forgot to mention thanks for the patch +Is this still an issue +En tots dos sentits damunt davall +Patch Commited closing bug +arable land +We are in fact monitoring disk space with dedicated monitoring tools but apparently required at least two times the current datasize for compaction and that can be a high number for large databases which could be compacted just fine with less free diskspace availableAnyway were just suspecting this is the threshold were hitting in production since the message is currently logged at info level and the instance is only logging at error level +These is stored in the file which should probably be renamed +Bulk close for release +Updated in ESB docs +I have a fix locally but dont think this will make +This should fix the problem in the CI test although it already worked correctly for me on OS X. +Assigning to you Ben +Yes it is to access the app we do not use any asf accountBecause Hupa is an it is necesary that the user provides a valid userpassword in the imap serverFor demoing we have Hupa configured to use gmail imap servers via ssl so the users who test the app have to provide a valid imap enabled gmail account this is the reason why we have Hupa configured to listen in a secure port +If not should probably go in a release or LATER +Just made the change +I only implemented the detection of faces config from because if exists in targetclasses upon deployment AS doesnt detect it unless the default exists under +can you point me at a test case where this error occurs +Once users can copy a shared page to a private page they will be able to use +I cannot reproduce it so I am closing this issue. +Jim If AVRO is JSON based then why not just use JSON? Pulls in less external dependencies +I like this enough that I committed it. +Very very simple workaroundIf enclosed type is a String return it otherwise return nullThis change works for me but probably thats not the perfect solution for this issueMaybe shouldnt be null and the right value returnedI cannot respond to this as I dont know well this part of Axis +Passing the patch through hadoopqa +Nice work Jon Auditor Hsieh +I will close the issue unless you can provide a test case for that bug +Ive fixed the compilation issues +This is a dupe of and later discussion lead to something similar to +Ive upgrade my project to Grails and same issue exists runs fine in still +Need to fix for this release +Did not change our build system +using a similar workaround as inFor we need to cover the files that are deleted through its ancestral directory +still need to delete to make lock workThanks in advanceRegards for me locksnapshots stops working if I ever run mvn install in my base project deleting it from the local repo clears up that issue +here is a screen shot of the error I am that this is in a sample and would appear when the database hasnt been Im going to go ahead and fix for the next release unless anyone has concerns about this +The version of hsqldb used in EAP is and has been patched the patch used for version the jar and uploaded to repository as hsqldb +The correct output is abc. +Conversion from stringbytearray boolean only true ignore case will return true otherwise return false +applied should we bring this up for discussion on +Please disregard the previous patch and use this one +this should be fixed in rev +Thats indeed easy after the refactoring +I just committed this. +Feature is already done waiting for documents completion to close the bug +Do you need anything doneto the JIRA config +Attached patch renames fixes and to by default disregard deletions when checking the cache +I do not believe that should be a blocker for for issue Jim +I think we should be consistent with EAP +Already present with useLock in the. +bq +Committed to trunk and to branch. +No new IT as its covered by existing +Ill review all issues before to release the eclipse plugin and Ill begin with the ones having a patch +AFAIK split combination is used to improve performance +I guess that in order to get what you want you should also increase the towards or thats what I did to get the results in the screenshots guess that in order to get what you want you should also increase the towards or thats what I did to get the results in the screenshots yes I had +added ui support and file writinglocking for artifactIdgroupdIdcurrentVersion changes for the edited project the currently opened projects are checked +hmm +Closed all tickets +updated ErichJust a line to let you know that your solution worked for me Kind regardsCdric Vidal +I also removed the geronimo API spec from the features assembly +the point of transforming expired columns to tombstone is to quickly the disk space used by the column valueRight which is why using a constant expireBefore value that we pass to the serializer instead of computing it locally each time seems like a better solution +The additional is installed by the http featureThe failure only happpens if the http test is executed before the aries test +Done. +Patch which demonstrates and resolves an issue where the fails when reading a instance +Committed fix in svn revision +so two rules +If this gets backported to well keep the oldAPI call to maintain compatibility +If your Child class above would not contain any method declaration then it should compile in Java +Attaching the patch which fixes the issue in +I did the test again and the custom exceptions are now handled correctly. +Im seeing this issue also +Thank you for the patch Juergen +I added the weblogic x flags for the latest snapshot +Closing the issue as version has been released now. +resolved this is resolved and needs a Release Note please provide Cause Consequence Fix and Result +Like calloc +Thx for your confirmationI had put the hack to svn head but leaving this issue opened + attached patches +Isnt the short term fix for this to copy the list of dependencies into the pom that creates the aggregation +To isolate the problem just do a test with host put the host to maintenance mode and Check if it works +d on RB will wait for to resolve conflicts and commit +would like to see another example before committing to include it in the inputoutput api +As I said earlier a callback interface of some sort called from inside the enumeration method should probably be the easiest approachOleg +Marcel please test it out and if it works close this issue +Post each filename you have reviewed so we dont duplicate efforts or miss docs are generally ok +I have downloaded the source files for different projects +It also defines an which currently does nothing the intention is that a will get change data from the actions in theThanks am setting as a fix version because it was actually fixed and released in +Dont worry if you cant +Removing them fixes the issue. +YesI issued the following commandmvn clean test DtesttestI kept monitoring searching for above mentioned exception message +I didnt consider this because we have the notion of internal packages +Thanks Yongqiang +Ive update that branch with my latest fixes +If you agree then please close it +Patch applied at +Attaching a new patch that merges with trunk and incorporates Vinods review comments +The problem happens in both the Windows version and Linux version +This is still a problem using Hibernate GA +This is the associated patchThe must still set multitenant to Y in in frameworkcommon JacquesI somehow provided the wrong patch +We are generally trying to avoid further dependencies in the core framework distribution since even if they are entirely optional they can turn into a maintenance headache for us very such a simple class you could arguably simply create your own version and maintain it yourself in some util package of your codebase +So for now Im marking it as a documentation enhancementI believe that the expectation would be that if a format is given in a resourcebundle then by default it should be localized +agree with your response as wellthe latest patch seems to include more changes than just the API +Can you provide a simple file to reproduce the error? And can youhave a look on your serialized to XML file whether there is an ambigious styledefinition? I looked my xml and there was ambigious style definitionThanks this now a bug in the serializer or in POI itself? +patch comes the patch! Some strings were corrected on in rev you test +If it is an ajax request i dont return the normal checkLogin request i return a a special ajax request +RTC all remaining issues so I can release in JIRA +The cool thing is that in contrast to original SVN svnkit can handle all types of checkout so once Jenkins upgrades to later version of svnkit the build still runsThe last build was triggered like thathomehudsontoolsantbinant Dversion homehudsontoolssvnsvnkitbinjsvnversion homehudsontoolssvnsvnkitbinjsvn clean I think thats a good solution for now +Do I need to use something other than ant test +as per tinderbox this occurs on trunk and on the fix patch on is now commited to trunk and merged to fixed on trunk and. + has been marked as a duplicate of this bug +Moved to The is only an +But this issue remains +Arquillianprivate Deployer deployer +reason for not liking it is that JBT does not need this D in and if we set it JBT plugins are stuck forever and furthermore if jvmArg cannot handle duplicates how does this avoid duplicating Xms and Xmx instructions? +No code changes +The src has a mistake directly trying to link against libCrun +After study +To be explicit the following indicate files that do not need a header +Added profile to the to change component +Ive made the change +DOTNETCLIENT connects to PHPBS and PHPCLIENT connects to DOTNETBS etc +Adding tests for Clints new filters +gem install pre still think it is a poor default for a class called to completely stop working because of one client but I admit that MRI does also appear to have this issue +Userss generally do not want to download the sources and build +Fine by me +We are using the from the servicemix apache snapshot repository it is bundled up using the Maven SAR plugin for deployment into +It compiles and displays the first page of the wizard but doesnt actually perform the refactor yet +Ill look at the code again from that angle and add my comment +Can you verify if theres a jar in mac os x JDK that conatins +Patch attachedAll tests pass except and Im not even sure that is due to my changes +If I then go to my blog I see that comments are not allowed +One suggestion is that should not only validate the meta information but also the integrity of block content before participating in a recovery processCreatedThanks Hairong +I think there exists only a few itests in camel itself but none on the componentBenjamin +yupIf any pull requests come we will be happy about receive an email to dev which is mirrored at apacheEnjoy. +This should allow us to make changes in the nutch code to ignore work and temp directories +Thanks Lars +Good job Thanks you all for thisThough we have spent about a day to change our project back to Lucene to avoid the bug now I think it is time to change it a test to tease out more corner cases I temporarily dropped s chunk size from to and ran all tests +For the same reasons like Darwin said I strongly discourage any state +Im happy to note this in NEWS +Patch committed in revision +And what about other projects +Closing based on Nathans comment. +CtrlH mnemonic is EI checked and none of the accelerators or mnemonics conflict with existing items +YexiCoolI am going to test your small MLP and work with you on the large versionBest +Is there any reason why is not applied? It seems to me that this patch is only a subset of. +same effect with couldnt reproduce the issue +Normally I am against commenting out code since it is in version control anyway but in this case those tests are actually useful and not in version control so probably makes sense to leave them. +Also add bimargulies as a moderator +I forgot to say this problem remains with the build + for the patch +FWIW i actually had to debug tar once eons ago +is currently failing on trunk +Having spoken to Rossen last night about what hes done with config in MVC I think the plan here should be to improve the docs and tests around Aggregator and Resequencer and then deal with fully as part of patch could still be considered as a stepping stone but only with better testing IMO +bundled in the summer of code zip! I used that jar and everything workedI have attached it to this issue in case anyone else needs itthanks +Trying the option now +Log file since problem started +Apparently there is a permissions problem mule mule Mar mule mule Feb mule mule Mar devdrwx mule mule Mar mule mule Mar mule mule Mar userNote the permissions for the issues list differ from the rest +OlegThe wire log was chopped off from the top +service commands are now available +fair enough and usefalse solves my you put putting your valulable time for this issue. +Well +I have just committed the rudimentary manifest support +After the is fixed I refactor the and check it again +moving to tooling delivered by +Completed initial feature involving importing from VDB one source model at at time +Ok fixed +The patch looks good +OK cool +update modification to get to simplify allowing the runtime to provide the bus and destination from the MapAdded Destination as a a property to to make it available to the runtime +This issue should be kept open until we have a fix from the JBAS side +Also can the property be removed from the file when an external key is used +Please use Ivy instead of +bulk close of resolved issues. +Well seeMarking for +New patch available +Ill create test cases to ensure it works. +This looks like the same issue as +Please provide details on what exactly is the exception from Hive that you getReverted change from trunk stream. +Changes look good to me +In seeking a resolution of this issue I created the class +Close this as wont fix. +If so flipping to that workflow is very quick and easy +Then it would work fine with Lob with or without a given length +A single formatting file would be a good idea although Im not sure if the checkstyle config XML to eclipse settings mapping would be trivial +So we finally tested without invokedynamic and that worked without errorRunning again with invokedynamic enabled caused the crash againDo you have a link to the bug report +If my interpretation is correct that is +I was often met with exceptions by GNU java +Suddently theres no clear notion of a bad entry in an edits file opening the door to all sorts of nonsenseWhile not incorrect in any way this approach would IMO make the dfs more vulnerable to problems that are inherently hard to debug +Merged to in revision . +Using tasks there should be fine +Reservation is the mechanism used by the scheduler to accumulate resources on a machine till a container request can be satisfied +Apache Directory studio version has been released. +Its a little more complex be we can do a least loaded thing +And of course the value should have been lower anyway +Additionally changing the number dynamically is a big jump from the current model +Backported to for the fix Carl +This is being caused by the parsing logic in admin shell +If both the JT and a JIP must be locked the JT is locked first +im going to close this as duplicate since it is not only related +we should really be using the recent history to do the comparisonas a side note if we were to chose not to take the maximum zxid during recovery we need to make sure that we still cover all committed messages +Could you try with mvn +Committed to trunk. +Here is the updated listselect list IIFor select listsselectselect data table sort icons except case when listShuttle +I mean reset the datascroller +Verified in beta build. +In the future it would be good to have comments explaining the logic behind the magic numbers in the code +A rerun of passed without failures so did my run of derbyall +Certainly an arbitrary Double cannot fit in a Float variable without loss of precision +Gant definitely submits a negative number to +Milo I cant see the fix right now I need first reproduce itI have assigned to Juraj to try reproduce it using the steps you have providedOnce we will know how to reproduce we will know it has been fixed +Thanks for reportingGood Eastern +Sorry for nitpicking +Bulk closing for +GIT is a distributed SCM +Maurice very weird does it happen for all your bpel files? could you attach one that crashes always? Does it happen if you use Java ? Bob can you take a look +Did you already begin the rework mentioned? It would be great to test a newly available snapshot +there are user cases when there is no sence to have loaded Hibernate Console Config +I didnt know Java enums were nullable +How easy is this to reproduce? Do I just need to start running mapreduce jobs and it starts to show up eventually? Do these need to be long running jobs? From what I have seen some of the renewal code has hard coded values to renew once a day but I could be wrong +first I need a unit test to show what is really happening here +Please just regenerate your and it should work +Ah +We can email the mirrors in question and ask them to fix the problem it will take a while to go through all the mirrors perhaps sebb will script a check for us +Thanks Xuefu +Looks good to me +We warn for a reason +I think then makes sense +Indeed +rebased +this means we detect problems earlier rather than when someone asks for the field the first time +The index recovery point is the of valid index entries in the segment to which the recovery offset belongs +It doesnt seem right for to do any purging +The system tables of Every node is the sameThe cluster run well after I delete the system tables in each nodeThks Williams +If the statement is a prepared statement would it reset to on subsequent executions + this be backported to series? It seems a trivial enough fix. +The message exchange back to the IL should also indicate a FaultFailure +At least no embedded params but i cannot be enabled either right now +Added new RFC constant +I restored the directory temporarily before heading to bed last night so that people dont find on Monday that their automated buildtest systems spent the entire weekend failing it will be removed again on Monday I actually think the changes made to the module in this JIRA and need adjusted such that directory is not considered a module any longer and merely a container directory eg like the management directory containing the various sub modules +This would allow generic pre and post processing +No more needing to use properties +No other change is neededThe only reason for keeping them would be to distinguish as a different kind of number which seems unnecessary +Other filesystem implementations would also benefit from using this test +For now is far enough +These wrappers are not needed +I think we should not rename it unless there is a significant change of it +Ive created a patch to solve this problem and it seems work fine +for the patch +So I assume it is added unit and integration tests to the codebase anyway +Reuploading the patch as it dint apply cleanly with trunk +I mirrored the tests that were already there in for a simple consumer +If possible can you please attach a simple application which reproduces this +Without a specific example exhibiting the problem there is nothing that can be done so closing this bug +Thanks for the pointer on a possible problem with the path +Im sorry this fix is more likely to miss CP the reason isThe problem i have mentioned doesnt exist up to this problem occurs on CP and onwards due to a fix on CP +Could your barrier not query the currently deployed objects on startup? It could then determine whether it has already been deployed and trigger the barrier allowing the esb to only depend on the barrier and the war to deploy normally +It runs for EVER but passed w the patch appliedThanks Matteo for the persistence +should be fine +Verified against HQ build EAP CR. +Mass closing all issues that have been in resolved state for months or more without any feedback or update +I checked in it contains the same fix from master +Fortunately this is for +We asked Andrew to contribute this to. +Can someone change the priority to high please +Updated patch attached +open project right click on node under project root node +Modified descriptions on the configuration of leader election +It appears it is not possible to make scrollbars to be the same as in AndroidiPhone +Linux EM +Hence added my custom methodThe master inside the token should not be changed even though the token is renewed +And what about the problems mentioned by Krzysztof and Burhan? Are they resolvedPS I would consider updating the RF on the livedemo to nightly so it does not scare Opera users away +This is a regression caused by the fix to +sense to me +HiI just want to share that I managed to get this IBM classes working on my applicationI used the jvm inside which I believe is java version My attachmentsCredits to original authors are still in the comment sections +What isthe differenceIf you start reformatting refactoring please submit smalldocumented patches +Indeed we dont manage to reproduce this issue. +Or we can postpone +Looking at the source for the fix it seems as though it was around parameterized types however joda are not parameterized +as it turns out +fixes with revision +fbricon reports this is now fixed +The other things mentioned should be reported as separate bugs and feature requests. +The lock change would be good in as wellYou want a subset of the patch in +Thanks +regression tests passed +Final patch after review +Its not just between ORACLE and MSSQL we need to support any database for our application +Could you please report the number of warnings after this patch is applied and if possible the other types of warnings that the compiler generates such as deprecated code or unsafe operations +ivycache as the lib directory? Not sure that that came with this patch but cant that directory be anywhere? +Such solution still changes the numbering if there is a value conflict in the original output but +patch committed thanks Corinne +libcassandrabin +Functions was not going to work because the payload and resource were needed so I created a tuple containing both +Updated patch in branch fixed the original problem thanks a lot! from my catch +Thank you Nicholas! Please commit to and for + was integrated on July thAt that time was still being discussed +Attributes were moved to separate I didnt find any possibility to determine menu item size in pixels dynamically +Rob this is bad is it still true +If its a platform problem then we have to report it against the platform +Smells like one of the oldest standing bugs +tml files protected other context assets available classpath assets protected +I am afraid current patch doesnt work with I couldnt force it to execute before method +Could you please supply another patch so that these tests succeed +Heres an alternative patch that brings those same files into the packages +Excellent work +Thanks for patch v +To resolve install Blackberry MDS +Just checked the svn repo of branch the patch was synced so we can close it. +establish until I meant to request that the redirect should be established by at least on the DNS changes +Mass closing all issues that have been in resolved state for more than one month without further comments +nothing heard from OP should not be an issue just a misconfigured application +Ive got the changes ready for this +I just open two tabs showing the same pageWhen the model in the first tab changed and if I collapseexpand the panel in the second tab it will throw NPEYou can run the quickstart to verify it +Attaching a patch for trunk with minor fix +Did the committed patch address the last set of comments? If so can you please attach that patch to the jira for the jira record? If not can we please have a jira to address the comments +Thanksrev +This fixes it +I mean it is still not clear for me if it would be possible specify interceptors for example in for whole classes graph in web application or notRight now as described in my post which I have linked we have an issue in jboss +Feel free to reopen with more details if you still have the problem or a similar one with trunk version. +If it becomes a problem well it. +I dont know anything about is this perhaps something specific? If that is the case this ticket should not be in the cloudstack issues listI see Frank already fixed this issue with commit dbadcabfcfaaeceefded and abdfbfbbafbeb in master +Im merging into this issue whatever solution we arrive at must be the same across all scenariosFurthermore its important to offer both SPI support for this so alternative approaches to ordering can be created by the community as well as API support as this is a core concern +The patch to resolve the issue +ResultsFile MB textPack javajfastlz s cfastlz sUnpack javajfastlz s cfastlz sFile MB binaryPack javajfastlz s cfasltz sUnpack javajfastlz s cfastlz s +We replace them after the build +If so I agree with Denis that option two is better +here we set appropiate http values by getting those values from the +I want to ask a favor while you are in there +Patching Sbastiens fixes into my workspace fixes the problem for me +Those changes do not look like they have anything to do with this JIRA +Hard to figure out the reasoning behind this change but I guess well have to accommodate this somehow +Libc functions on the client do all the resolution. +on D patch vNice work Mikhail +you ARE good to go now +First cleanup of +I think this is not the problem of the parser or the underlying parsingprovider itself but rather the problem of maintaining bnodes mapping inAlthough one cause of the memory leak in the bnodes mapping has been resolved in this issue requires another solution since the map will keep growing with increasing number of bnodes in the graph +Having said that I still believe that JDO offers a far cleaner interface interms of configuration and acquiring an instance of JDO +I checked in a fix and tested directly with +I just committed this to and +Please consider providing details how to reproduce this issue +Cool +The tests passed cleanly except for known instabilities in test and testPing +That copy is now out of the mutex thus will not trigger contention but is still much heavier than doing a hash lookupAlso since you are returning a reference of the map I believe the read lock in get is likely useless +There is an additional parameter which can be set to true to use the legacy block readerI think we should start adding description of what a change is doing so that it is easy to understand the change being done without having to look at the patchWhat is the need for config parameter +Minor conflict with the current patch fails on applying with attaching a new one which fixes the conflict +Yes updating both end of the association would make sense when both entities are in session +Ill add additional SQL injection filter +Screenshots of correct and incorrect behaviour of CDI Refactor attached +Jerry from the forum discussion etc +Does that test pass w the patch in? Should figure a way to get a test in there if possible +Thanks +Despite having this on our TODO list from ages back I dont think theres enough justification to add this. + +Are you ok with itThanks is already available via the various binomial coefficient methods +This looks like a weld bug to me +Download links are all down pointing to the previous dist location +I committed this. +Setting Henry to be the reviewer +modelimpl but I cant see an artefact in the patch to that effect so if not deleted this file will show lots of compiler errors +committed +This file can be used to test this patchsave to xdocsadd to siteHave description to be concise +Added freemarker component and added wiki document as well +Mike are we going to postpone actual deletion of these classes for +And a version with test fixed +what do you need it for +This message is being referred to by many classes and its probably not a good idea to remove it +There must be some missing detail in our autoloading logic whereby we should not look at toplevel constantsHiro You may be right if timecop uses autoloading to pull in missing featuresI still find it rather odd that you can define an autoload for a constant that would resolve even if that resolution happens at the top Object level +Heres a new still preliminary version that uses the Nullable annotation from and that passes a few more tests when is enabled +Shalin the new patch looks good +PS +That all said this did find a couple other bugs that Ill log related to the wsdl validator +Would this have any impact on other projects using generama +Most likely when we get to that point we will probably locate those files inside AMQ and CXF themselves t be maintained and deployed with those releases +putting in http +Hope following issue could be sameAbove are the Hebrew +Another approach is to reset the interest ops to not select READ or WRITE after a request is completely read +I have on my development machine an automated corefile backtrace script attached is the backtrace I get from a crashed I stand corrected class is not a jclass handle but in fact a pointer to the javalangClass wrapper for the wrapped objects class +I would agree with Halley that if you click the button when putting files in a JIRA issue that its OK to use the code that a ICLA shouldnt be needed +This issue can be closed as has been applied +Arggggh! Starteam is a bad you post a sample stcmd output +Patch for +Release is currently undergoing QA and will be released shortly. +Users can type a report location in a text box then press a view button +has been released closing issues. +AFAIK its not something maintained by the Geronimo project +thanks Samuel. +If you make the onclickevent bCanceltrue no Javascript validation will be performed +Fix committed in revision . +That is incorrect +Were seeing an increased number of community contributions of integrations of other stores so Ill bring the issue to the teams table +Thank you +Customer requested fix version ready for James commented on the apache jira to indicate he would validate the fix against his expectations as I worked from a test case I produced +You have to set up your own project and add the dependencies yourself +If we create the map in the constructor and never replace it then threads racing can end up with a map that is X Y which matches neither threads view and thus is inconsistent relative to the two views the threads wanted to publish +ok the issue returned after a quite some time +Also seeing this issue +The program that when run causes the error +I did not write and do not have control of the server code +LGTM +Thanks for responding Wendy +Any feedback on this +However this seems to be a common requirement for Java EE development +The unit test on my dev VM box always exit with unable to create new native thread at the end of the unit test suite even when my patch is not there +I do not see this change in +It makes delegation possible +Hence the idea to always use merge and simply copy the id over +I dug around a bit and it was with the Enum code and not the Builder code where we were doing such things +See attached heap dump screenshot +The automated tests have now run for close to executions against this patch without any problems this time so it appears to fix the issue +Thanks +Obviously this slows down the write but thats an acceptable loss +The distinction is that New KSCF are restricted to characters everywhere Cassandra will detect paths on Windows and fail startup if existing schema definitions are too long +the relevant issue to track is. +The problem appears also if you print the special character directly from a servlet +Is there a way to add them apart from gwt setup that Im missing? +Would be great if exists a special parameter like true +Seems in each patch that is the hardest partHow about updates and updates +private final Object is rather a general best practice than something lucene or module specificsimon +I just wrote a unit test to confirm that the value gets stored and is able to be retrieved so the bug seems to be at a higher level +Current code does not work well when the last Feature is reserved for the older release since the last Feature determines the current versionAdding a notion of reserved for older release in Feature so that appropriate layout version is chosen as the current version +Hi Vincentdid you restart SMX or did you remove the data directory after the stopRegardsSorry I was in hollidayI just restart SMX I do not remove the data directory +This is also fixed in this revisionAgain thanks looking forward to more +Not sure why the backtrace isnt generating +Fixed in r +Hehe it would have helped had I put this bloody issue under the right +mvn test Dtest CbeaWLSserverlib jdbcoraclethinhostportdbname user psswd falsewill result in assertion issuesc +Because of this those features have not yet been pushed into +Could you be a bit more precise pleaseWhich components do you use? Do you use clusteringDoes it happen under load +This file doesnt look a jar archive did you use a zip utility? please use jar +Any such tightly coupled modules are placed usually as sub modules which I guess would be atypical with open source development scenario +Issue may already be resolved may be invalid or may never be fixed +The strps to be followed are given below +Should I move that method to somewhere common or just copy it +Closing since there isnt going to be a release and this doesnt exist in +for code review pending local results +Seems as a pretty simple patch so Im giving my here. +wont removing the active SOM bork things if another stream to that target is going on +The patch is committed to both trunk and branch. +but with respecttrue the exceptions will be shown after about times restart container +My email id isnirajcognizant +One you mean you want the empty strings in column C to be loaded as nulls? Actually Hive interprets N instead of empty string as null string +Looks like we need to give the backup broker uris in case of failure +Hi AlexYes the keyboard layout is in thats what I need Ill help you gladly with the testing you +Thanks looking into it +Also why dont we add the count of superstep to and as a part of report +Lots has changed since than +AS uses versionAttached are a patched binary and source jarThe patch isnt official. +Quick verification is welcome +I was trying to compile JAXB generated code and started getting the compiler error that source requires target +In java this bug seems to be fixed so I tend to close this bug with Wont fix + on patch +Investigation in progress +Sorry Omkar I dont have logs for this now I have faced it once during long runs when the load was high on the cluster +Ill try to run the tests today and get back to youThe funny thing was that Java wasnt giving an errorit would abruptly stop reading the binary format +EAR with embedded based WAR to demonstrate the problem with the endpoint address URL in the jbosswsservices console +updates on this +Not a bug +That will resolve this issue and allow removal of the src dir +I dont have time right now to debug the issueany chance someone with a clean test env can run these through +Compiles with following tests failed I them again individually and they again failed +svn repo frozen pending migration to git +Ill look into it further +Thanks Harish! +This is intentional as there is an attempt to the JNDI context and factory problem are you seeing with this +Under what circumstances this exception is thrown? Looking at the source code of I can see that validation results can be null +While putting this together I noticed that the XPATH test does not work correctly only when processing the first child element at least in this test case not sure that will helpThe test case also runs the version with a JRE so that you can see the difference in output filesThanks the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +The only thing that comes to my mind is that there may be collections which wrongly implement the methods in question and do not correctly return the change statusSo the way it is implemented right now is a bit slower but will work in all cases +Does it work +technically i dont see any objectionsGreetings Freddy +If a regionserver tries to open a region and the assignment message has already expired it can just skip it +The equals method is here as a bit of a hack so that related events in a queue can be removed without having to write code to walk the queue themselves + +Hopefully that solved it if not we can reopen. +Do not remove the dead nodes channels from JBMPOSTOFFICE when occurs +I am a Mono contributor who uses solely under MonoLinux +committed +Console log +Release note content for known issues following do not work with the minimal configurationaggregator requires smookswiretap requires requires webserviceproducer requires following have been successfully tested with the minimal configuration added to release notesa reference to this should also be included in the Getting Started Guide +The ear contains a webapp with a that only defines a servlet without any and isInside the webapp is a fragment that defines a +Markus what do you think +Sure why not? Please go to the forum if you want to extend this discussion + +We cannot fix that +downgrading priorityany chance you could providea simple testcase your literal text doesnt say me much +Since I havent been working on the project I encountered the behavior in for months this issue might as well be closed. +Snjezana and that includes not showing the weird dialog up front or not +Once you have a solidified a proposal for the different connection types can you submit it to the community for vote +I tried modifying the +It sounds like thats a matter of setting usefalse in the VDB and adding a bit of caching logicFor a couple of reasons Im planning on using a delegating translator to wrap all of our real translators so it make sense to centralize the caching logic there +Will close +There are many minimizers but not many are present as maven plugin and not many are easy to configure +This will do nothing harmful and wont be interpreted as a class to load +If it works then all is well +bq +If has been postponed should we back out the user visible changes affecting existing applications on the branch? Based on my short experience I think that the changes are going to be painful for Network Server usersIf we are going to be making more changes for startup etc I wouldrather hit them only hit users once with all the changes once they areready for +Could a committer please review it comment on it commit it or +webdav bundle updated in r +But I just noticed that the add column didnt apply to all regions just to most +You have to take care when comparing to sstables sizeOtherwise it looks ok +Once Chukwa is its own subproject I was thinking to just do a big automated reformat locally then commit +Also since you are looking into it please mark it In progress +the patch for test will be created after test patches for JIRA is applied +In this case returning of null would make sense +It turns out that examples jar is included in the classpath when hadoop jar command is run +Do any of the agents have GUI capabilities? Such as a buffer mode X driver? or are they all totally headless +in the username and he couldnt login to JIRAThe second account he created does not have a +for do not use a generator strategy on the mappedBy side of the unless it is a foreign generatorfor you are using a bag semantic read the reference documentation on how not to use a bag semantic +but the suggested is not an option for me in our production environment +Older versions have been left as is +Ive committed the delta to trunk and. +The property is supposed to have the current directory as a default regardless of the security policy setting +did you find time to review the patch +Thanks for the digging Tom +this is umbrella task and some of sub tasks are beyond so changing the fix Emanuel +Data point I see this in +I just moved all the code to an abstract class and created a new mojo with the exact same code as the except for the execute annotation which I got rid this patch is not applied on top oif the trunkIt is applied on top of trunk we have to reorder these issues all patches will change +This is a duplicate of +Nothing was changed +OK works well here +Will go into the release. +Reverted from at revision Note to myself a non functional change to had slipped in at r +Yes since we need only db adapter to generate DDL its not necessary to connect to a live datasource +Thank you for the detailed suggestions they were incredibly helpful +bulk close of resolved issues from releases prior to +I will send a mail +Can you create an Arquillian test for this +Thanks for going to the trouble of your survey Dominik +related issues. +from me +Max would you please take a look at this thanks +This issue stayed in the In Progess status too long +I think so did have a test for it which failed under Davids security changes hence I know it should be there +srctestjavaorgapachehadoopmapreducesecurity + +in +ok i will change it to postgres all old Resolved issues +Edit srcmainwebapp and remove the in front of jspsThis has no effect in me know if it fixes the problem +This is consistent with compilation dependencies +Hi Charlie thanks for the update! Is there any way for you to verify whether the patchattached to this issue affects your symptoms? Can you build a modified version of with this change and try it out in your test environment? That would be very helpful +I prefer to deal with them individually +Ill switch back and make the changes if camel is still on +Uploaded v onto review boardAfter WAL version metadata design is finalized will add that +I think it is better to keep separate base classes for keyed and pools as the core pool methods have different signatures and semantics +I was able to reproduce this issue +Distributions may be reclassified or withdrawn by majority vote but the release number may not be reused by another distribution +These are the things that I would expect would be bloating your +An example project that demonstrates the +Is this really fixed I get the same error for mentioned above for +Like I say I dont know of a use case but that probably speaks more to the limits of my knowledge than anything else +It makes perfect sense to commit it as it is and make incremental updates later +Task logs look fine too +I havent tried the workaround deserializer but this patch appears to fix the core issue. +I had heavy assistance from Mohammad Shahangian on this one +The fix is neat +Also since the image is compressed the extra disk over head shouls be miniscule isnt it +Lisen can you please add Javadoc comments to specifically note that the timeout doesnt actually cancel the query +Just got bitten by this bug +Thanks +xmap svnnative +logWhen it failed on you anything in the +Passing to Fernando for incorporation of release into EAP CP +GEP trunk works fine with it +Cannot repro I just tried it with a new project with the command line tools in the cordova subfolder for the new projectWhat does return for you +We will revisit this for but I must say that by design a is there just to support this particular adapter +So from my side +All pass with the following revisions revision TCK revision resolved issues for already released versions +Please verify. +to on. +dfs already has a tool called that allows to run multiple datanodes in a node +Attaching an updated patch after some large scale tests +New version of the and patchesThese patches are needed for the replication test framework and replication testsI plan to submit new patches for replication test framework and replication tests in a couple of days +I think that the simplest solution then is to have something like bulk scanner in instead of single row call that we use right now +The only hypothesis that may make some sense is that were seeing a reset due to either an unhandled database exception or due to an error condition + put in and checked in +Another area that could be generated is correlation sets although this may need to be part of a second phase BPEL generation +Thus you wont see itNot sure on the second case +If every call to setMapping caused an unmarshaling then this never did what the user wanted +It really works! ThanksWant me to close the issue +Wrt +Fine idea but We should continue to keep velocity buildable wo javacc so we checking the artifacts of javacc +The problem we face is that one badly written query can bring down an entire server +String literals do not belong to any schema and hence its seems reasonable to have them take the collation for the current compilation schema +Is there anywhere a available? I am happy to test it myself +require users to set up JAVAHOME environment in Windows OS +So our will load and we just need to make sure the appropriate methods are loaded in initCore by any hackers interested on a project this is pretty simple +The trunk patches will be available when the plugin is consumed by trunk +Checked into develop branch +Committed at subversion revision +Thanks for the patch. +Applied patch with thanks. +Fixed in which we can include in Central +You can find that files are not in the correct path while named in the form like classes +This will avoid backward incompatibility as well as need to create symlink on Windows +So I should back this change out then until we have the extra ts in hfile +bq +Ok Ill commit this evening or max tomorrowThanks for the infoBye +I will commit this +Or is the goal to see if we can make it to three years before we get a new solid release out the door +Folders have the codes for each vercion +If I get ambitious later I may create an issue and patch that allows configuration of the max token length on all tokenizersThis is part of an effort to close old issues that I have reported +It should be online in a few hours. +Its relative to the moduleskernel directory within the axis tree +Specifically if the project does not yet exist in the workspace the import link is not present and there are no error messages in the wizard at all +Duplicate of +I know this is closed but just to add that Scott Carey was incorrect in saying that repositories in dependent projects arent picked up +First of all thank you very much for your contribution! For now we dont have any committer who knows Scala so importing your patch immediately might make it unmaintainable module without your help +Thanks for the reviews ATM and Tom Ive committed this to +Jira project group createdjacmorel added to itjacmorel also a so can add other users to the group as for all jira messages to scm is XPR +Click on it and it will insert the content of the entry to which you are replying in a quoted block of in the end we just need a button that does that automatically when you reply hence the Quote button +If you dont want multipe rows in your select box then leave out the multiple attribute +Linking to as this change will help diagnose that issue +I guess that wasnt the best decission as one thing thats not good about my fix is that the messages are done for the top component while you probably typically want to have messages done via the components that should receive inputSo I can roll back those changes and make setRequired instead +And replaced the bad code with the usage of those properties +Reopening issue thanks to Stefan catching the bug +no reply +Ive also searched other references to find general guidelines without any luck +For this reason its postponed to next sprint + +Added basic support for files currently it is hardcoded but it will be based the indexing config in the short future rev +simple test this needed for +Fair enoughI agree with you all on your commentsI will add methods in corresponding class against a new JIRA issueWe are good to close this issueThanks for comments +This seems to be fixed and the relevant method is available in this was fixed in r but only on the trunk +Tried to applied the patch but got again some errors even after translating the end of linesRyan could you try to produce the patch again with cvs diff u Reading a bit of the source code it looks like theres more potential of that could be used thereIn particular I would believe it would be very helpful for maven to provide to the the list of possible completions from the set of goals +Verified on JBT +HelloArchive contains reorganized improved and documented version of BT InfraIt brings many useful features into BT Framework allowing easy configuring the test run easy integrating of new test suites easy to setup and launch the test run +NET type this has been written in a similar fashion the the Java code +If you dont mind Id like to try out the new unpack filtering before you release +bq +With large scale tests we are seeing unexplainable issues with running on a cluster although the fetcher still uses the old API +Removing the line fixes the problem +I am trying to replicate this I added an orders dir under webapp in but I dont get the bad anywhere where did you see them +an extra attribute would do the trick +So picking an invalid case doesnt mean it failed +By my opinion even now validation message enough correct +comments from erikThis approach seems fairly brittle +Those characters arent exited properly +I just committed this to trunk +This has frustrated me for awhile too +If Lucene had a different policy on then this could change +Let me mark it as Wont Fix. +The problem also goes away if I uninstall mewtp integration +committed as r in trunk r in branchThanks Rakesh. +If you guys are using it for the training then that should be something ye manage yourselves +Eclipse WTP Its very odd +We will have to cut a release asap +Ah that is a windows problemD I guess we have to change to +I was looking at the avro schemas and just though I would double check +We abandoned Infiniband network after an discussion +Initial docs added +Need to verify that the method described by Charles in is correct if it is then this RFE is already in the UI +Committed +So you want a symmetric hook when job success? does not help +Blocker for +The spec says names have to be unique or its an error so I think we should fail validation in this case +Heres the version I tried +Opened since ! Any progress? +Before binding to the port the receive buffer size on the is always set to value from the which previously would have been the OS default but is now bytes +This has been incorporated into. +Man I reverted the snowball partLets change to a zip file as the tests will never change +I think requires registration of functions in command for getting functions argumentsVasanth JarcecI have one user experience suggestion on your approach +Many thanks for your work and dedication all of you atThomas +Im intending to submit fairly soon +Secondly the error is causing a null pointer exception because the configIds is expecting a string but is now being passed an array in light of the new multi option plugin installs we can do +Fixed for the seen such issue on. +As long as youre doing reads for data in any CF from some node you will have latency info for it +By parsing this text again were just wasting cycles reparsing something thats supposed to be output properly to begin withMichael is this an immediate problem for you? If not itll wait a few weeks while we fix the XML tags properly. +i barely started to scratch the surface looking at this patch while checking something for but before i forget i wanted to raise one red flagthis patch removes the call to set from and moves it to with the assumption that can take care of it +pushed +symlink problem is + WARN rint Extender +I am running your sample against and have enabled jmx in activemqI have sent more than messages and got this JMX outputI am running on what is the status now? Do you have open connections when you have sent messages +is fixed for HSQLDB version introduced new table name pattern +In addition I have added acatch clause for as discussed +thanks Daryn +How do you propose one get to know if the feature is of ordinal type if we use subclasses? Wouldnt that mean lots of instanceof or a visitor? To me it seems much easier to bake it all up in one and the same class +On Yahoo Mail if there is a problem loading a message I get a nice looking dhtml popup that gives me an error message +This issue relates to more general issue that proposes adding GZIP Data Format to Camel +I do not see where this is coming from? Can you please guide me hereCheersAshwin +As a workaround if you are not using Groovy or supplying a user modified set of extensions you can probably just hackremove the services file from the groovy jar +Manually add the VDB archive file to the deploy directory of the AS instanceThe fix for this needs to be implemented in a manner such that it fixes the Teiid plugin in JON without breaking the on JBAS +Inline macro namespace was being dumped during parsing of the source to be evaluated. +Hi JoeIs it in any specification that normalizing line ending to n at input time and to rn at output time? If it is in spec I need to fix our DOM parser side +I am sorry it is possible to download attached artifacts with the help of classifiers +I think this issue is related toPS By mistake I linked to another issue first can somebody with the appropriate rights remove this link please +Doesnt work with XB issues resolved in released versions +Note that the annotation for message bundles has changed so you need to update it +Please verify the fixThanksJyoti Sharma +No +Im thinking of getting done and out +Updated patch to use and as index cost values +Thanks Navis +Thanks Dave +Im working on that right now +toI found only in a folder called tomahawktargetmainjavaSo where can I patch this classClass is also in a target folder namely tomahawktargetsharedsourcesMoreover I was not able to add a entity for to the patchI added all these files separately into the zip fileCould someone explain these problems and the project structure to meIs there any documentation available +Patch as requested by someone to create a update to to reference the on subversion revision until someone sorts out reactorforked lifecycle issues +Thanks a lot for the quick review Todd. +Ooh I think I fixed it as well +I think it would be good to resolve this before landing on trunkwhat exactly did you expect? I mean its basically the same thing but reusing the name sucks +Vaibhav rather than just to simplify things I think the value in removing the concept of regions from the is more about defining the concept of a Table contract without imposing the notion that the table is split into regions +Discussed this feature with Nileema Stack KarthickDesired characteristics include +Im sure other applications do too +There was a minor issue when invalid tarball is specified with tarball option then ringmaster failure occurs and return code is while for without patch return code is for the same scenario +I talked with Austin about this on IRC and came up with a really simple way to do what Austin wants have our launchers add the or flag to JRUBYOPTS env var immediately before launching the JVM +Test added +I appologize for not having checked the facts myselfJuergen +Its not because tools like nexus exist that this should not be fixed. +Added define statements into the header files that allow the shared library open failure messages to be displayed if available +It is not working +Ive applied some changes that may address this issue +I just pushed this +No more complaintsproblems with this reported +Attention Take care that this patch includes a new file in the path tomahawkcoresrcmainresourcesorgapachemyfacescustomdojoextensionsresourcetemplates thx to Stefan for providing this useful patch. +popup was fixed on the level +I had to make a few wfs tweaks as well to get the change to actually be picked up +Im all for the complete version here +is more widely used than +b on master branch +first draft versionJust made static methods so far +Relaximation uses a fixed number of writers and readers and fires off a new request whenever one finishesThat means that a decrease in write latency means more writes overall and if any resource was saturated before it likely means a commensurate impact elsewhere since theres no free lunchJan if you can find the time I think running the writes only test would be interesting +Thanks a lot +Why didnt the user just stop the server and edit the with a familar text editor directly? So I strongly object the idea to add a big edit box for tomcat +Trove is LGPL I believe so I dont think we can use it thoughIll look around and see if I can find another library that does what Trove does +Instead of having script as part of the source code its better to have it at the top level in a diff directory +Travis Agreed and also locally tested on my end +Did you run ant with Dresolversinternal options? When I do a clean build I dont get any of the duplicates you mention +This is buggy behavior example +Hi Shane works great for Authentication but I have the same requirement as Gerry for progress on this issueThanks! support in Seam Security onwards will now be provided by +If you need more info please let me know +Implementation based on and Tomcat +This is not a bug report it is a feature request +a converter should be used whenever there is a different type and if not no need for a converter that does not allow nulls +Please let meknow how it is shaping up +need more assume this is according to the eclipse error I had the same error on my machine same stack trace in the eclipse log same records in the Server view event additional observation Before the error happened I couldnt hot deploy as expected because the required JEE module dependencies were partly lost +I committed this test only patch into the trunk +Update to the latest axiom code +Someone should assign this to me +Closing as the has this feature already implemented through similarity pacakage classes +Was merged in +Resolving based on this attachment for this issue +Attached is the modified +Marcus evaluate the leakage at your convenience + +I think content encoding decoding streams MUST NOT close the underlying stream EVER as the encoded content may be just a part of a larger message +So please rename the issue as I am not allowed to do that +I think it is caused by recent changes for EJC complier support see and corresponding discussion +See for steps how to start the Camel using right click on file +what does you query look like +I have read base document about Hadoop and SIS +Ill double check your test schema against the CVS version. +I should have mentioned this as we are not going to do any further release on the code branch. +Yes two places we need to handle for this null +I hope that some day Ill find a non confidential segment somewhereThe file still has its background problem but this isnt related to the regression so Ill create a separate issue when I have the time +The file you are looking for in is in HADOOPHOMElib It is not ver but worked for me +Added dialog with text area for error +no idea +The slightly longer answer to Bryans question is +I cant replicate this +This jira has been closed as part of a review and archiving of the aj project +Please revert this patch +Cool +So it overrides the logic and does the float comparison +Ill let Paul decide what to do with this along with +Before I arrived to this problem I needed to use this patch +I applied just the Logger class but not the test case as I think this class is trivial enough to be correct without a test case +Im sure theyd both be able to help out if you have questions +some overlap between these two +Bumping out to would like to address this issue in Karaf +Unfortunately we got caught up in time crunch here on this release +There are no errors in the log +Youre right I was confusing DU and DF +Alternate implementation based on instead of a queue +There are problems with this suggestion +For the trunk the patch is fine +Thank you you please patch against trunk which should be revision or greater? You patch is from revision which sounds completely wrong +Nope I have my own now +Just make it a static final variable compiler will inline it in the referencing class +When I said the class is now in the I should have been more version I want is in the but the SDK version is still in the +Just for illustration to show what worked for me for while tinkering around +Could you please advise as to which patch was applied to? I am unable to upgrade because is incompatible with Compass due to repackaging of a Hibernate class +Ive checked in changes that should fix all the relevant issues now so proton should now compile under OS X +It is up to date +Please verify and if the problem still exists +See +Ill create another patch this evening after work with everything merged and well see +You are obviously more familiar with the code than me +Reassigned to Max for patch is for other issue +I think Rupert did already some improvements on this +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +I suspect this will be resolved with RC +We made a deliberate decision to keep up with the latest and Unicode versions so recognize that we depart from the RI in terms of compatibility hereAs you say trying to match a particular version chosen by another implementation probably isnt a productive use of our time +Patch that sets it false by default in both mapred and mapreduce +Myrnasvnexternals expects only directories to be mapped +You patch is not correct +I see +I think a better solution would be to add a callProcedure method to that returns an Object +Now it must be used with the validation interceptor +repository jpa interface +Im going to double check the r with +Sorry was my first commit hereI found the place to add the files wait on tiles work after +I understand the problem the reason while it is this way is to let people know that the name is missing which is a bad thing most of the time except in the use case that you described +Verified in JBT. +Propose to close this as not a bug? +Thanks Kanna +I guess what I really want is to be able to annotate the exception tests that stand out from the others and its probable those that I want to be able to disable +Highly likely is related to this one +Thanks Scott +The Properties view always contains the properties of the current provider +Thanks Sandy +Hi Mikecan you add the whole config you used here? And what exactly happended +thanks for supporting regardsHenk Penning +You have to rewrite the read routine +Philip could you please close the related issues that this patch fixes +The logic hasnt changed see my explanation in phabricator +If it is positive Ill go back and update javadocs to better document the feature +But it would be good to know whenThanks a remark on +One of my test case which makes different length data compresses and decompresses at different split points depends on +I have to look at whats causing the issue with your file Thomas +I would use the interceptor to set the scope and display parameters for Facebook sign in instead of specifying them like hidden fields on the page +Now that work is underway with AS all previous community releases are +Currently we are hitting when trying to run Oozie against still need to write tests etc +Thanks for reporting +aha so check columns timestamp values I bet they are from the future +This will be used by the datanode to connect back to the namenode on the right RPC server +Thanks NicholasSo I will mark as duplicated and will help to review the patch when it is ready +It will not work with jo therefore I thought Jox would not be appropriate +Im rejecting this since the Memcached spec is not specific enough in terms of whether the casversion number should be unsigned or not +I just uploaded a new draft index page +Thanks for the patch +For currently we use current for getting diagnostics info from the AM for setting status on UI so we cant set it to nullFor the tranistion in actually its valid +I agree that packageasgem should be made a plugin it would be easier to test it too +Version of Accumulo already includes that fix. +Daniel if you happen to have a stack trace on those three threads Id love to see it would like to ensure there arent any other points where this sort of thing can happen +The attached patch fixes the bug on windows XP and does not adversely affect other an alternative solution + +I checked it and even if no viewRoot on facesContext the algorithm should calculate the render and get a valid because after all if this happens on an ajax request the render could be on requestMap or in the worst case a valid response should be send +Closing issues that I resolved in already released versions. +About the bugs I talked about see trunk r r r +Using a form in a popup dialog seemed like the natural approach not an edge case +To your point if controller failover is happening the rebalance attempt will fail +Need to reopen to refine this somewhat we dont want to log the error message if the class is being as a result of an inheritance relationship +It still needs lots of documentationIm now leaning towards adding this to Avro rather than +For reviewers For the convenience Ive uploaded entire big patch to review board and also slice it into three more or less independent pieces for easier review +Testcase patch for can you please make this part of the main test server and client so there are not a bunch of servers for each case added to existing Test serverclientModified created against do you want to implement any of the TODO items listed in the patch prior to this going +bulk close of all resolved issues prior to release. +patch against rThere was a problem in my last patch related to the test classpath this is fixed now +This issue has not been picked up please review and see if you can help fix the can you provide a sample log which we have s secretkey printed out in the log +Fix pulled into repository. +Done in trunk +Patch against the git repo +Unassigning since I dont work on this right now +Replacing with cross join causes a lot of issues +Committed revision +dbmigratein upfrom optlocalsharejavatomcatwebappssonargemsgemslibactiverecordin migratefrom fileoptlocalsharejavatomcatwebappssonarlibin measureI hope this helps +Figured it out myself +Thanks Nicholas! +Sorry for the slow response +LGTM +Im marking this as resolved +Stand by +Hey andrePlease externalize those strings in for both trunk and branch +Thanks LionelIve applied a fix +Chatting w Benoit its kinda crazy having to look at start of rowstart of family before trusting found cell +Verified for closure by Julie ZhuoClosing on her behalf as she yet has no rights in Jira. +Without looking at the code I think that this is not possible due to the XML parser +works fine in manual testing but im not really sure how to write an automated test for this +Assume closed as resolved and released +so I dont think this works +Problem is we need it to not happen at all +The only thing to add is if you want to create a on the fly using java code instead of providing an external but I guess this can be a separate issue +Yingda thank you for the patch +would you pls try the patch? thanks a lot +ok closing then! thanks for info. +Committed to snapshots branch +Compilation errors occur if compiled with JDK to Denis for patch review looks reasonable +Sorry James my fault +However because of issue I am stuck using for now +What if you try INTEGER instead of VARCHAR? Its a related issue +It is probably also worth pointing out that and higher fix this mistake and require the security pieces to be there in order to be RFC compliant +What is the status of this? If there is development work to be done I volunteer to help +Need to use for consideration one more time in a Future Release +Thats the only way to alleviate this +The problem still persistsWhat should I try next +Then when I attempted to do more work this morning on it my test server ran out of space on I had just did merge and suddenly there was a new commit from youOh well I guess if its there thats all that matters. +In the future this could be modified as the capabilities of the Portlet container increase +Patch fixes problem for py lib + +Committed patch to resolve issue +Rob please review my patch for is go +Thanks for reporting the bug I will fix the issue sometime today as my time permits +Sam from a brief analysis of the stack traces and given that this works against and not against M this could be due to recent changes in +Excellent debugging Sergej! Thanks for following up on this well try to get this into +This has been implemented in the experimental version of Logj +This is screen shot after applying my attached patch +In order to reconstruct the original object you need just a subset of those chunks +I dont have a quick solution at hand thanks +Not to be committed +Let me take a look +is not a very intuitive structure at all which I remarked onCheersChris +I will be patching the component to make this parameterized +Absolutely the header was there only because I used Carrots grammar to start with and forgot to remove it +Exactly +Invalid JIRA shouldnt have a target release +Forgot to set the Fix version set to in +Will the user ever get to see any benefit? moreover it is going to be a compat thought one of the bugsfeatures was to make DIH more API friendly so I figured this would be a good way to do so +Final version of fix for trunk patch + +It is a custom macro so autoconf didnt automatically took care of everything +Agree with +libcassandrabin +AmitaYou dont need to worry about test case for CTS because this is Tuscany SDO specific featurePlease open another JIRA to work on supporting unknown features because it may require different discussionI will take a look at your patch today and provide comments in this thread +Frame sources in the wicket examples were hardcoded in the HTML +Can this be started plz +What about double types + this patch looks beautiful + Server log file for a single thread single run scenrio of loggin in posting and replying on the is a matter of configuring logj properly +Display information on primary keys and nullability +nice work +I remember seeing a JIRA a while back about Configuration dropping names or values with the empty string or something like that but it worked fine so that must have been fixed +out of hunk FAILED saving rejects to file javatestingorgapachederbyTestingfunctionTestsmasterjgroupBy +Attached are a few suggestions +Thanks +Addressing Teds review test case and javadoc chance to put this up on RB for a final check before commitI would also like solicit opinions +very good +JohnI agree + +This is IMO not the case here as not many people use the TUNNEL combo so Im downgrading the severity to major +Furthermore can you elaborate on what you mean by does not render correctly +Can we use JAVA specifics for the trunk? yes +This is a patch for this issue but you also have to upgrade javassist to version + fixes null issues linked in this test +Its really helpful to know if anything is running right now isnt it? Say total completed +For example how many disabled would you add to a security realm configuration to allow fo future additions? For another why is it better to have a bunch of empty unused stuff that will be modified in the future rather than just adding new stuff as its needed? +Ok I have committed the solution +By the way for HTML content that you want Forrest to process as raw include the file with the extension +Was missing from +Issue is in +On the other hand is targeting java making it lighter typesafe easier to maintain library and one wants such dependencies over legacy onesIve attached a patch to which replaces with +Works well thanks +I can confirm the patch works for our dataset +Please see messages above for more info. +Could you be more specific about what isnt working +apparently weblogic does not have a free developer license +You might have to check the CDH +but the error is not in the fact that EMC is used instead of +I also dont understand +Is that a possibility +comments on lot of tests are failing can you debug +Maciej So far we have been unable to reproduce this using either trunk or +Does mvn test pass on trunk with this patch +Attached committed +Okay do mind +IDL order keeps that option open +This looks like a problem with uDig +LGTY Michael +pending jenkins +It should really allow for multiplefalse +The not and ternary operators now suppress and throw instead +This is fixed in JBT but not in within +I was looking at this issue yesterday for a customer case and could see part of it was missing from the svn update +Be careful out there +Such a file can then be read in via s configLocation property with all local settings overriding the ones in the config you can keep everything else defined on with just the cache strategies defined in +Yes we couldshould add build to the SVN ignore list +So Ill update the CP branch with the messsageID fix no problem there +Sorry for the would be a big step forward if we could at least turn if off since I could then use the pingURL again +We should just make it final everywhereI think you should be able to set it while you are useing IW +Updating a new patch to fix it +Im a bit confused by your description +Code was tidied up a bit +Yeah Ive seen those statements too +But I think TMB should give just to be consistent +Probably the approach is better for the general case +The first attachment is relative to trunk the patch is relative to the first patch +I am trying to work around this problem by putting a limit attribute to my SB but then the SB does not work at this a bug or am I missing something? of a bulk update all resolutions changed to done please review history to original resolution type +Thanks Joydeep! +Now the default language file is assigned to the DEFAULTLOCALE and thus default translationsPreviously when opening the Configuration status page the servlet used to generate code for ALL the configuration printers and this caused the page to be loaded slowWith this patch the configuration data is loaded on demand when the corresponding tab is selected +Im not sure why sonar has this class of bugs since the rollup metrics should all be available +This bug was fixed with bug +Ben is there a JBAS JIRA for this +There is streaming support available via in the latest builds +Judging from the dev mailing list Id say that there is enough activity to get the release done this weekend +added patch for +I applied this patch w p +I think the general idea is that we wanted to be cautious considering that a method might never return NULL +project and +So people can turn this off if used in rare cases. +New patch now uses Guava API and MR Job extends +Will attach the the XMLXSD example +I will reopen it if I have more information on that +Done +Do you have a spec reference that indicates that entirely relationships should be supportedThanks I see the purpose of these two optional elements +For example if I cut and paste from a +Um + for the patch the failing test case is unrelatedShould this go into +This patch modifies the form tag to write the saved Token as scoped hidden input + +let me take it and go through it for M to verify everything is addressed +Check the files in the target directories to see which file it is +Lets go on for other components +Bah +GeorgePlease try this version instead I fixed the bug in as well as +Konstantine can you take a look and update the component if necessary +this bug exists in rc too +done +This also helps with debugging jstacks so that if the user never uses handlers but configured then its more pain to read through +If the user selects days and if the violation build is not executed then the system may not be able to show these results +Hello MiloszIn case patternvalue do not accepts stringexpression values I should make this bug invalid right +tested manually since no test framework for the web ui +Committed to branch +Not sure why this feature is needed +Theres going to be some kind of performance penalty and you can accept it because the functionality is worth it +Since we now have that save these assumptions across separate invocations of cli by storing in a +I would not do any work on the Materialization stuff as this is due for overhaul which we left out in only connection pool that was monitored was the infrastructure connection pooling the connectors connector framework never exposed the connector connection pools +Garrett Wu tells me that if one uses the new mapreduce in then one can use s directlySo once is committted all that we might need add here are some tests to illustrate this +Maybe it only happens on a windows platform +However I am currently not able to test it sorry. +Will let you know +JonathanThe client code should not need to connect to the master directly so all the functions should be available from or +I think theres no JIRA issue yet about it but the idea would be to get rid of it and just have a message in the status window for instance +Is this true for databases created with or is it only true for databases created with which are then upgraded to and then to? Thanks +Overriding the names of the descriptors would be a bit tricky at this pointbecause of how Castor currently looks for the descriptor files + and containing patch which allows for a transitive flag to be set on a Configuration +I added new patch that is rebased to the current trunk +Does this mean the is now ignored? If so how to we set other JPOX variables? +Is here any resolution whatsoever to this issue or not yet? Presently I have to stick to Qualifier but this adds unnecessary code and makes my code base fragile thus I will be glad to see and test any resolution from your side. +I just check in the patches for Could you review that? Jongjin +Thanks for all the great suggestions RickI will certainly add links to the collation material as you suggest +Working on the remix +We had to upload it using the same groupid of this other versionIt wont fix older releases but once the trunk as well as future releases uses the JWNL from central repo we wont have problems with Source Forge quota anymoreYou can find the fixed in the SVN trunk +What type would you annotate? Would you annotate each field thats a short? Would we implement mixin annotations +Jap Jacques good point i like the ideaIm a little bit in trouble the next two days i will have a look on friday and fixes the bugs you posted + is not a stable release compared to +I have also left the Solaris section in the Installation Guide since it will be available for GAAll other changes have been actioned as requested +This event fires when the app goes to the foreground which happens before the deviceready event +Once I got the JMX connection for the remote working I am consistently not able to shutdown the server the shutdown process gets started but when its finished the AS is still running on the remote system +Attached as for your consideration is a minimal set of changes that put back level request metrics +You got what you needed from Ted? Let us know +So for the release which HTTP verbs are supported with which parameters? Well want to havebuild tests to cover what were supporting +What are you using to validate it +messages mean +Ive taken a look at the failing tests and it seems to me that there is no deployment problem with the endpoint servlet unless it is generating metadata that requires authorization checks +Further discussion can be had in. +While I think it would be nice to have support for native SQL insertupdatedelete this is not the way to do itBy the way already has this functionality in HQL +No response assuming ok. +This is now redundant given that lets you allow to use arbitrary Pig to be used as storage driver +adjusting as Wont Fix due to inactivity +Its a maven activation that chooses the dependency set +Sorry Steve but no +This issues found by Mark Struberg should be investigated and then if javassist is released with the fixes it should be upgraded to the release +Patch to correct the issueI have removed the use of subst which was mapping the P this is unneccessary on newer windows releasesI will give it a go on some test machines at work so far this tests fine on windows +remove and methods +All complete. +I would think you would see the same problem from the downloaded from the web site but I havent tried yet I can try it top level parent +Fixed +Created Hudson job configured and started waiting for results. +Sorry had a bad commit day hereI did not include a test here because this was revealed through watching log files during NN startup which indicates to me that the condition is easily traceable +Committed! Thanks Jarcec +Thanks for the report +Perhaps you or someone else on the list has some time to dig into this problem and figure out where the real bug is. +for simplicity +Prior to this was independent of the being used +I dont think the patch does whats intended +Ive noticed no other errors thus far +It seems to have resolved itself +Does the current suite your needs +Bulk close for +No objection this needs to go in +Patch for for the test was run manually and passed +Graphs are clickable +Thanks Carl +I was completely new to OPIE and I entered the wrong password the first time therefore Im not able to get sudo permissions again because the password I entered was not computed starting from my secret passphrase +Anil this is initially assigned to you because the xsModel seems empty you will not have to implement the anonymous types assign this back to me once we have a valid xsModel in +Jacopo we dont have the resource to run another Jira on the machine +with jetty Im left seeing the contents of the web application instead of going to +You removed the indexVar whereas I left it aloneagreed better to leave it +Upgrading the console should follow automatigically with +I have a fix for this +I attached the logs. +ThanksJim. +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Paul the same applies to community then too +Permissions are just another spot of state from the inode so we cannot combine these in any meaningful way +Youre correct I was thinking +Something that greps forward from the JAVAHOMEjrelib directory for the to determine the path would be quite elegant and much simpler than trying to divine the OS Architecture signature and build the right path +Make sure we honor their request and that we report back ll also have to look in to the response strategies this may only work well with the buffer ones speed will likely screw up since the error occurs when streaming out +We should do your first suggestion too +Forgot to push +Revisit for a point release +Why would that not be valuable even in a standalone agent +Hello CraigThanks for the excellent writeup +I cant reproduce the issue with is not related +Closing tickets +Also there are some stylistic issues like star imports and the use of type casts that make the patch a bit troublesome +I will take a crack at a does the same thing of removing the Date object +This filter canbe used as action or render the servlet environment I extended the by implementing the that puts the into the extends the analog to the setup allows us to configure the conversation framework indepentently from the attached the sources mentioned above and hope they will make it into the spring Linkon behalf of Allianz RenThis sounds very useful Im going to look into it more the way by the end of March Im going to visit Jrgen Hller to finalize the Window and Conversation management so it will make it into the Spring M release finally +Reopening for EWP for EWP +So am presuming this is now spurious and can be closed. +If some of them dont then there is a bug +You probably want to set the listener in both places instead of just one +This are annotated with Jackson typing Annotations +Can you add a bootcamp task for thatI will commit this one as soon as it passes the test +Verified on of a bulk update all resolutions changed to done please review history to original resolution type +I dont think it is a good idea to sleep for seconds every time +This will solve the problem +It looks like File Install uses the Log Service if its present but if not then it always logs to stdout +I will review these changes more carefully in next few days this comment is just based on your comment describing the changes +It has a field totalNodes which appears to really be active nodes +As nothing is marked as final in everything should be fine +Whoops there shouldnt actually be a brach in thats just in +Status? Has it been fixed +I just recompiled the porject again and was surprised to found that the porject passed the compilation! But I swear it was not got compiled yeasterdayI remember that the only thing I have done is to remove all my +First I spawn of x number of competing threadstransactions +This patch works against +I will commit Friday or the weekend! revision forgot Thanks Yonik for the good ideas and discussions about API and help with coding this new trie implementation +This sounds to me like the issue of identifiers with non jdbc based datastores does it not? Any identifier you get back from a transaction against a shapefile is unfortunately unreliable +DevsThis seems to be an easy fix +Streaming needs to bypass that logic since it will create a copy of the stream +I assume this is now resolved +I will most certainly not do that +Can you please modify the test attached for us to replay things as we wont be able to fix the issue without being able to replay had been no response on my questions on how to reproduce the issue. + +Can we close +Need to add a css and mess around w more to get stuff to show in right place +batch transition resolvedfixed bugs to closedfixed +Ran derbyall with using Sun on Windows XP +Thanks +Trivial patch introducing a new config param for the timeout check +should be of the classes are subclassable on top of the public interfaces +Fix must be coordinated with new runtime +v +I was affected by this issue pretty severely with the freezes happening very frequently and I can verify that this patch fixes the issue +Committed revision . +Hi Todd ATM Before moving i wanted to know your opinions on this proposal +and dont see any problems with the current implementationIll need to get the a wrapper though for the String compares to be case sensitiveinsensitive as set by the flag +Please reviewI used as an inspiration +Patch includes the following changes Updates to test the new Modified to only register the clickdblClick events if the appropriate buttons are created via svn diff This changes the behavior of the component +Presently we are relying on people to email links whenever they update which isnt ideal +Some page load properly others appear to be missing data as if a packet was lost +Hello MariosBad typing Arubacloud shall be the new nameIf its not a problem wed like to change the name everywhere because we are going live with new services in Europe with this new brandLet me know if you need other requirementsThank you +this avi is the bug avi does not work for me +Committed after all local tests pass. +pass in actual +Resolve +In particular the open benchmark with threads and K files is as followspatch trunk opssec opssec +I think a soltuion would be to skip generating of new external login key for ajax requestsAjax requests can be identified by presens of http header +GERONIMOHOMEvar folder must exist otherwise deployment cannot succeed +Fixed please please create tests for all the test cases from the description +Applying the actual decorator pattern just needs some extra lines in theCheck for the right constructor to delegateIf no such constructor is there we just use the class provided in the file wo decoration +be +I have applied all suggested changes and a new RC is available +andrew how about if we put this info in the webI as well dfsadmin report? will that satisfy you +Marking as fixed since master appears to revert back to performance for the given benchmark +Im closing this issue. +The replication commit reservation code definitely has a bug +Fix on javadoc new patch is Bikas continues to review the patch I just wanted to say that the patch and the code overall is so much cleaner now thanksWill look at the patch again once these comments are addressed +as you pointed out in I think it would be good to have both the old CLI and sqlline supported +We have seen this issue before but we did not reproduce it so far +Scratch that last comment +fixed in svn revision thanksdims +Changed the resolution to test +Flat File Delimited Columns Parser layout to expose buttonThe Texttable dialog includes the inlcude HEADER optionAdded source and view model names with source and view where is the name of the selected flat file minus the file extension +Ive lowered the priority to Minor. +Closing as fixed since the main problem is solved and the comment problem is fixed by +Awaiting respinLocal spin from M branch completed successfully no more deltacloud features or plugins in the site +OK I found out whats causing it +I am willing to cut a new release if you have the test +please reopen if you still see the issue after updating on fine with new system vm template +Fixed in ESBtcunninglocalhostwebserviceproxysecurity svn commitSending webserviceproxysecurity file data +now looks more than okThank you has been added with description SVG will reopen the bug as long as it seem to be still there! Ive created asvg file with lot of images in its body encoded into and some svg circleswith fill opacity the circles and images are each of them groups +This was not a popular suggestion and would be a major slash and burnoperation +Its an attempt at doing a cluster that has no RPC +Only the formatted dir is inside varrunhadoophdfsnamenode +commit awayAs an aside my favorite statistic about my first year at Google was that I deleted more lines of code than I added +Committed a first version deprectaing the +Verified artifacts and output with jdk build +Actually we are experiencing the problem while loading the data where it fails to connect if the host the MR job is running on is dualstack +fixed in trunk will this fix be available in a build? If not soon can someone post a few steps on how to patch my build? closing stale resolved issues + +Error message which are going to users are non descriptive they should explain what the error is in simple terms without going to technical terms +which log file +An Eclipse project to reproduce the issue +Looks fine! Thank you +Bobby could you please upload another patch without the unit test +Also no ajax request was fired +Im having the same issue has this ever been resolved +Attaching the patch that closes all the streams in the finally clause +Ill compare our changes of with revision and see if changes will be necessary to the new classes. +In your case the returned just doesnt contain any +Trigger build on last version +Most tests are passed except for and tests that are effected by +Eugene your slightly different issue is tracked in +This is actually a pain +Grails will still retain your changes when +although the petStore app probably doesnt have one i can verify this next week +Actually I see now those are runtime deps not covered by the pom +That works even better +at gitxiaoyang at commit in Nicos patch for file channel bug and now use commit Im glad to hear its solvedResolving as Invalid since this wasnt a Lucene problem +I believe this is fixed +I agree with you completely Cocoon cannot find the configuration fileThe actual problem is that the data file over the address filelocalhostCaloneservicesservletjspworkjspTempcocoonrootf is attainable! That is the point which confuses me +That is about times faster thanWith the patch for majority of the data the path changes from file direct buffer Java buffer direct buffer socket to file socket which mostly explains slightly better than times less cpuThe main remaining issue is with sockets as mentioned in the previous comment +Nigel Could you please review this patchtnx +A new distribution has been uploaded to Codehaus +test twitter has been applied thanks +Hmm +Attaching a patch which is the result of svn merge c for but needed manual adjustment +Patch applied at r please verify +Patch applied at +Ill run tests again and commit +The attached patch file incorporates the changes as mentioned in the earlier commentThe key change was to determine processes older than a certain time +The overflow property is specified on or after +last one has a wrong configured directory for upload a corrected patch +I missed one file in the previous patch +Committed to branch and trunk talked to Stack offline and he thought he saw that it was calling current for every edit but its only for the last one we ship so he was ok with it +I think this should be closed as wont fix +good to go +THANKS +will commit once tests pass +A link from the releases page makes sense +This is not a regression +Thank you Jacopo looking forward to your comments +Oh and Ill take care about wiki update +Applications will be initialized up to the INITIALIZE phase while running on integration mode +Fixed +Marking as done +So feel free to commit it but leave this issue open +Thank you Bryan for taking the time to create the patch +This was merged onto the master and branches +This wont make it for +I didnt look nearly close enough at the patch for to realize it was different than whats in trunk +This is done via the configured in deploycluster +Thanks Stefanyour patch is in rev +Trivial patch attached +This breaks test and test +It is also an opensource +This is working in the itext example +changed eols to osgi unit test as well +So if I make a mistake in the thrift definition package and release the rpm to a yum repo then yum install and build the binding it fails +make the stem filter final and add explicit test for the mobile e +Closing for now reopen ifwhen we decide to backport to +is always created but never closed +changed to +Can this be addressed +If you have dual installations of you can use the nunitHome configuration parameter within the test plugin configuration to point to the correct version +validate methods now throw exception instead of in the latter casePlease review and let me know if that fixes the problem for youOleg +I removed four types from one which was not used anywhere and three that were copies of classes. +What is the scenario for this? If it need not be strict then I wonder how its going to be useful + +Reopening to modify release note text will set back to Closed Done momentarily +and we should document this feature as well as the pig one +test project isnt renamed but this is another problem +Perhaps the next step should be to publish this as an initial website we can always create more content later? +Builds fine +Will submit new patches in which the set is correctly set +passes in a row so far so it definitely looks like these failures are related to the compound file format +I agree with Max ant only means the other hand its quite common to say sabates dant instead of sabates de vellut or sabates de pell or sabates de pell avellutada. +Thanks JD this should help the compiler be more optimal and it should be an easily recognizable structure for programmers who might recognize the manual tail call optimization +sure thanks for spotting +Thanks for all your attention to this Aki +Okay I see now how this is a sufficient fix +Sometimes user will use a dynamic web project that has some servlets +Calling the after a deploy and upload from the UI should fix this +I know confluence cannot manage versions but maybe you could have used different global spaces one for branch and one for branch generating the static documentation from the correct space each time you need to create a versionI dont want to bother you it was just a minor question to know if there was a way to download a correct documentation for the versions +Setter sets xads classname field class loading and creation of instance with xads classname are only done if the xa datasource instance has not been setThanks for the patch! +bq +patch with the is on as part of the recent merge up of directory as first class issue. +There is a trivial thing +Users can install it from Central + +Sorry I did not read the list last Sunday Let me know what the options are and why do you consider Hupa demo dangerous running in the asf infrastructureThanks ve turned off the gmail instance and keep the demo one so we do not need the certificate anymore since the demo instance has a public usernamepasswordBTW the demo instance does not send nor receive messages it uses a set of mock messages and a stub smtp implementationPlease close the issueThanks Manolo +The original issue was closed while the problem is still reproducable upto version +If its hard to fix Id vote for removing it or replacing rows loaded with bytes loaded I really dont care if the latter is wrong and this way I still know that some data ended up in the tableIdeally though this number would be correct +No significant difference with G. +Ok ok I pull my final wild card the name of the enumeration is already +samples should come from random positions in input files since input files are frequently themselves sorted +If not is there a workaround available +shows doc changes +Xbaya is able to list all the remote jackrabbit object but not delete them +Closing all tickets +Change of issue + server is not actively maintained +Committed at subversion revision +Because there is no TASKSTARTED event in jBPM +From the crash log it looks like a vhost problem +I checked what Windows does and it seems to give you the option to restore a recycle bin file to its original location or copy it to somewhere else +This should help to kick start our data dictionary effort +Depends on +Rebased with latest would like to include support for in Oozie +Is there a reasonable situation where a +Maybe its some bad AntJVM interaction +Well besides this of course +A simple backward compatible improvement that no one has to turn on +That is due to recent updates on UI component +Fix has been verified. +I more or less rewrote the relevant area afterI honestly dont know if it works with in I dont suppose you could trya backport to is fairly likelybenson +But eventually yes +This patch works for me and does not generate the errorsERROR Unable to create Timed out trying to locate root regionand it restores the strictness of which is necessary for Please verify that this patch works for youThanks +Stupid me. +the fix is part of moduler classloading probably but this really only affects hibernate spring AS and since AS now is moving to so if anyone wants to do that im okay with itmaster already has this fix see +bulk close of resolved issues. +you can reproduce this behaviour +The ansi handling for a global temporary table is that it is a proper schema object and is implicitly bound as a new instance to each session upon the first use +Please review the code +test that shows that the reported problem is not an issue anymore as all of the test methods using plain OQL or SQL CALL statements complete successfully +for the patch +Timers survive containercrashes and the activationpassivation and loadstore cycles of the enterprise beans that are registeredwith them +I didnt realize this +Nice catch Todd! +I was thinking of that when I was refactoring +trunk should be compiling and all tests passing as of about hours dont fully understand the changes made in already +The fix for the target host null issue is not in trunk yet +not a good idea to change +Your help would be greatly appreciated +great I agree that the tests are at fault here +Please review +This issue looks like a duplicate of +Patch which fixes crash on interpreter is applied at +Fixed in FOP and probably earlier transition resolvedfixed bugs to closedfixed +So this JIRA would break that functionality plus probably others. +Now this variable bytes was removed from +How many beers did you need for that +I think there is a similar problem with doc values vs +Jay I verified the patch works on my system too +The Plugin works fine with Maven +The hardening part of this task is done. +Let me work on an updated patch +I didnt test this expose primary keyoption when I wrote the fix for multiple joins +Sorry I got distracted by other things +Heres a quick patch that adds which overrides to add sorted property +What is the probability that this fix will be in a EAP patch? We are using CP and would really like to consider moving to the line but cant due to this issue +and when resolve is called for something other then this information they aquire the real resource and then delegate to is assumed a catalog REPLACE event goes out at this time causing layers etc to subst the fake wrapper for the real thing +Thanks Gail really helpful +Closing the issue +Parent issue has been superseded by the AS domain model work but Im leaving this subtask open as its a concept we havent dealt with in the AS work +Ill get back to this next week probably +I am for less magic +it was for the wrong bug +Deferred to Beta +defer to ve added the necessary code to build BDWGC on Cygwin +It doesnt +Out of curiosity because I already started experimenting in this directionHas anybody benchmarked the option to keep bloom filters on disk +We need to complete this with all the code and configuration files as well as convert it to an archetype so it can take on different application names and package structures +Brett this pull request is not correct see my previous comment +New patch generated on branch with unit test +I could take a crack at it you havnt already started +It will help a lot if you take care of these issues thanksOK will do +I have updated and pushed the changes in the Forge to the Nazgul pluginimplying testing should be simpler when the release is out + the code was copied elsewhere? If so were the tests also copied +The existing lock wrapping is done only around REST request and wicket requests to make a test that only involves the catalog one would have to grab the locks manually +Thanks Namit +Have tried to mvn run build locally build is broken so cant reproduce the issue +Paging Hudson for new review +bq +Can we see a +Thanks Vikas for commitDeepak Dixit +These classes could do with a general revamp +Oops! This should not have been unassigned from bug was caused by an incorrect initial assignment in getPipeline which may lead a null writer +Apply from SVNROOTcomponents +Of course well need to benchmark to verify +If someone wanted to do it I would not it +An int of bytes will be read and be casted to a short value +If you are working on this feel free to ask anything that could helpThx A quick first version +I also added a so we can override it +I have reproduced the problem on serveral different boxes in the QA lab and it consistently you or your team look at this for the release +The webapp implementation details really documented +Their capitalization should be enough to distinguish their purpose in the textcgsinstallingderby I prefer just distribution +You can still access the old web UI using the same until the are removed. +Remove from svn page mail page downloads front page +Stefan Just on a code review point of view this looks exactly like the way I would have dreamed it +I said I agreeThough they do not assume not null +Some new +Yep I dont like that we have to put passwd and password in the property files but have no time to spend on that particular fix +A bunch more to go +I just fixed this. +Any thoughts on implementing with velocity templatesI dont want to slow this down since any effort is great! but long term it would be great to drop JSP! I tested +The do should propagate a so that the web appss configured exception handling can take over and display an error page +Assigning to Julian to drive this +Is the autoconf change related +Cant reproduce in develop on OSX +Is the deprecation truly necessary here +Not a svn expert you may have to svn update to see changes not sure +I dont think you really want to attempt to fix this in. +Ive committed this to +makes and use to set deadlock timeout and wait timeout + +Adding a new requirement should not be done lightly but only after broad consensus is reached through patient discussion +not all test passing with that patch yet +Linking to existing issue wrt other question is whether to support a Java type mode where Override wouldnt be allowed for interfaces but just super classes +Jack su You are probably in the wrong directory +It seems like only controls inside richsimple are richsimple was applied work workaround style of one inner div was changed from overflow auto to overflowhidden +May be able to steal and adapt test cases from JTS +Ill take this one and submit a patch +The needs to provide legacy support +I can do a custom Wicket build and patch by removing the fix for but I would rather not do that if this is going to be addressed soon +du shc features +jar should be ignored when building the job file because they are only needed for +Fixed in rev +This bug is solved when moving to maven +Thomas the patch looks good +Ill make the general statement that I dont think its really fruitful to keep changes to the series +Simon thanks please commit this +Message delivery appears to work as expected when using the connection factory at java but not when using java +Yes for now +Proposed issue is the complement of good to me +Otis Ive used it just once and noticed the problem +Without the patch it fails with RCSo for committing it to trunk branchIll build RC once this is committed +The decription file for patch for the libjava file to support creating the manifest using current bnd Davidwe have to options here I will change the build instructions and add a new target in which the current bnd file will be diwnloaded +Thanks for this contribution Xander +refactored patch to Owens suggestion as the functionality is specific to File +In the v spec of Grant and Revoke under the database owner section it states the followingUser creating a database is referred to as Database Owner +files are not being updated properly +The end state is that we want to only recover the portion of the log segment from the recovery point instead of recovering the whole log segment +Leaving this JIRA open until that has been resolved +davis what is the status of couchqueryserrvers how do you expect it? I find it odd to have it in couchdb folder while most of the code of views are in couchindex in couchmrview now +I can reproduce it on AS too +Uploading latest patch for Yahoo distribution security branch and this patch requires the and +Committing so can put up a RC +If there are no objections Im going to apply the patch attached to this issue +You should probably remove the README for +Bela thanks for the tip +We also intend to reopen this issue in further discussions but this wont happen in this JIRA issue +May be a simple method to be subclassed if you want change the quote +Yes also easy to add constructors that take arrays +reproduced in please fix version of as it has been already +Since changing the join type may broke some other part of the code so lets be safe here and only push filter in front of a branch not generating null +Did not make EAP need to keep open to jar integrated into for ER +Also which driver are you on? There were on that day and one of them had compile errors +bI dont know if it should be to the branch for m resolving this for now +It all works for me the Repository Purge by Retention Count was set to so snapshots remained after a scan +If I run them local they are fine +committed in r +Which breaks the dont make me think rule and then some! We can easily make updating the published version number part of the release processFor similar reasons Id also support putting the prominent download link back in +original fixed the and added a test case in revision +Started a nodes cluster with +committed as r. +Test cases and the error message are attached here +FROM DESTquery SELECT DEST +committed first attempt to the LANGPOST branch. +Do we need to introduce a new patter for the Ids I mean the capital letters? I think we should continue to use camelCase but using shorter names that would fit in user field or even better change the field type from to in entity where all the ids would fit wo any change +This requires more version tolerant rpc +Defer to defer open issues to defer to not essential to address for +Will do +I tried on EPP with JBDS and I didnt hit this problem either +Ah didnt know that existed +Thanks is the patch ready for review +Forget to include examples when this issue was closed +I just dont see a lot of benefit from making +Closing as this is ancient and reported never followed up. +If a groovy file is already open and settings are imported settings applied properly If settings are imported when no editors are open and then there is a restart settings applied properly on restartThe only way that I could reproduce is to close all editors import settings open groovy editorSo I can now have a deeper look but it does seem that the problem is slightly less severe than before +There might be other ZZs that dont get fixed but this will ensure the items in work and cover another set of formats +SureIve attached the simple patch it adds empty definition to target defines this for kernel classes build +The error window +I totally agree with you that we need to examine the syntax tree rather than the execution plan for things like sort by order by etc +Also he tested with Python so it may be changing the default buffering optionsSkipping it is reasonable +defer all issues to +I am observing this issue again with releaseI and can say that is fine +I would suspect that the actual test failure is related to some locale setting or a releated issue +Thanks Nicholas +This patch removes the dependency by introducing a so that instances can be created from their byte array representations without the client having to know the subtype +That right there is a recipe for a broken keyboard and monitor so Im gonna let a better web dev figure that one out +btw +The wrapping itself doesnt cause the entry to be removed that is done through a command +I do not reproduce locally with rJacques +This patch adds the path change but doesnt seem to improve much the performance compared to patch looks fine +Job directory cleanup upon submission failure can be addressed in another jira +I think this should be a Chore that retries until the ID is found then exits +This patch can be applied on top of +Fixed build by adding forgotten files +So I think +Ok +We can just use the logger and let those who care redirect to JDKs logging +Might be better to have a startindexsupplied flag in the Query object to trigger sorting rather than smuggling it as a magic value +I was actually not sure how to accomplish this +There are also some places where an absolute redirect is usefulThe same problem is with pathInfo while it is available from the you cant always rely on one being availableMy chief concern is introducing a new configuration item which most of the time replicates information in some other parameter +I believe this is fixed. +The whole point is that where are conversation ids stored is not always available in case of before processing is initialized directly in RESTOREVIEWI have tried to use to start initialization of conversation and the contexts bu this event certainly is not fired in Postback request where it is recycled from previousI saw similar issues in WELD like or +You should be changing dont understand why differing local repos per project is useful though? +So I used an external database pool connection with max active connections set to +The problem is that the message isnt to appear in next builds of and master streams +Conceptually this could be a side output of whoever rolls up the imageedits files into a new image yes? potentially it can be implemented this way however it depends on the requirements from the end consumer of this change log +I will provide a test case on the weekend +Thanks Mark +Thanks Jody +Thanks. +Please review the attached test case +We need the ecj dependency because all JDK cant compile Classpath due to a bug in javac in particular our set ups are relying on JDK awaiting a more stable VM release the ecj that is installed on Fedora and by default is the broken version that creates a broken Classpath +Hi Richard Ive tested the new code and it works perfectly thanks! +reproduced in the latest build of JBN +Fixed for trunk +Please test it +This is indeed a duplicate +I will check but I believe with the receipt of your CLA we can setup your account +Approved for +fixed with migration to new transformer api which handles all this cases. +Thats correct +Could someone post the correct link here? +Either use the jar which includes all the dependencies but is as Claude says not the ideal practice or else add all the dependencies independently being sure to upgrade to +delete keep it would be nice if we could add redirects from and to and respectively +An application wants two containers the first on only node or node and the second on only node or node +I know K of ram job a man to the moon but why are you running such a low default? +just committed this. +If it is what I suspect we need to perf test the effect of having region based marshalling enabled by default + +I think it will likely be the case that we work it out for trunk and then backport +VR spefic issues are tracked by so closing this bug as fixed +fixed closing all issues that have been in resolved state for months or more without any feedback or update +It might be good to at least do a test on a trimmed destination value before assuming its a queue and getting into this scenario +could you add some explanatory comments to these changes +JulianPlease wrap the long line manuallyIf you use Eclipse to work on you can easily see the long lines in your code by turning on Show print margin option in Preferences General Editors Text Editors and specifying as Print margin column +I verified the tarball produced by ant tar includes both a core jar that is just pig core and a pig jar that has everything +In mine only the serviceClass and the address attributes are supplied +Lukasz im afraid i have tested this use case for +phew! +please merge the patch to merge to to +This patch moves the iterator method from to and changes the classes that use the iterator method +proper spellcheck of repo also contains which was uploaded in the same timeframe and also appears to contain said signature change +Implemented changes added test cases and updated documentation +In I changed to and then it worked fine +If the exceptions are mostly at debug and stop shortly after the server goes down then Im fine with resolving this +Also added a test for a situation where a base URI with parameters on the path in instantiated +depMan is now used in the trunk. +Sorry I apparently dont know how to generate patches with git +Closing all old Resolved issues +cool +Some of what Josh pointed out earlier have been fixed +Patch applied with thanks to Christian +OK +Dont set the fix versions +The test looks much better now thanks! Theres still one little issue thought it seems that you need to create tmptestcli somewhere in the Before method +Thanks William to report this +There is work to support both SOAP and SOAP over JMS in the related Jira. +Closing old resolved issues +Well the pooling is pretty useless anyway particularly for this case +basic fix is jira is giving errors but still performing actions +Patch updated against latest trunk +It should be issue +EmmanuelHardy what does Immutable mean in terms of inheritance? I had it on a by accident +Duplicated with +My first choice would be what I originally proposed because all the elements are the same and hold essentially the same chunks of information as I understand it +Well need to generalize this a little +thanks. +is in the dist now but and are not +Patch against current against current workaournd is nice +Much better thanks AndrewPatch applied to PACK module at repo revision rPlease check it was applied as you expected +Ive checked it in for RC +Im looking at the patch +at the very least its a that appears to work around the problem +The agent should basically be able to run arbitrary commands monitor the process that it starts at the very least +Ive been working on this over last few days +Resolving. +i realize textile is the way to go +I will keep watching the area +but in production cluster it would be better to avoid this situationAs you told removing the check only can solve the issue +Closing this issue again as the problem seems to be fixed. +I think there may be comments on the usage of the error handler in the just wasnt sure about getting the primary exception and subsequent failures +added back max in criteria size for all translators and updated the docs +Ill think some more on it +Let me know if you need anything or have any questionsThanks patch is being submitted by David Mays on behalf of Comcast Corporation +Our project has encountered this bug as well when attempting to upgrade from Axis to Axis +Open a forum thread in the future if you have questions +Regression fixed thanks for the test Simon +Patch is applied at +I didnt want to change the semantics of the existing Merlin implementationThanksColm +Updated patch that works with full screen and regular desktop application to go with the most recent fix from Chris is great and doesnt require api changes so moved toIn the commit I will change only a little some details +It includes the code change that I proposed in the previous comment on this issue and a simple set of test cases to be included inderbyall and passed +It also fixes the guava thing so if you apply it I think that all issues will be solved +If you could it would be much appreciated +Merging the changes took a while +If a workspace becomes confused and folks are seeing unexpected problems ant clean should always repair things +I would not go through and remove the use of interrupt +Request for review and comments +So your contention is that only is retried and that theres no way a server could deliver a near a bounce? Can you elaborate this +Marking as closed because as far as I can tell this works correctly withAs a side note one cause for problems would be having classes available from more than one class loader and not beingcareful to ensure that all the classes are loaded from the same classloader +bulk close of resolved issues for release. +Daniel I thought that issue reporters could close the issue that they created +Will add the pom file changes for maven dependencies once is committed and the hcat artifacts are with an example that exercises Hcat credentials +also breaks applications running on server and using +You could add to this testGood stuff Eugene +In the datanode process should die on a failed read or write +I wasnt planning on documenting it +push to get ready for RC . +Do a google search with geotools and not recognised as a supported file format and you will find this associated with gdal +Initial subsystem with a child repo node and build changes to allow for creating a kit deployable to an AS into the branch +This issue seems to be fixedIll close this jira +I think this seems a duplicate of which should be fixed nowPlease reopen if Im mistaken hereMaarten + +I think that in most use case it will work +Thanks for the reportWhat we want to do is as you said delegation to the wrapped contextIf you can please post a patch +closed +can you review this JIRA and close if complete or move to the appropriate version +but then how will the client know how what to filter on to actually constrain the query by foo since it wont actually ko what query that wasWell its optional +Had a quick look to this and not sure if youre workaround is valid or not +Please attach an example we can use to reproduce or a patch against the jpa m closing this because theres not enough information here +Maybe in contrib to help indicate its not fully supported and can use more input testing +Exclude tests failing during a full build +what I saw was that the progress monitor is still reporting files are being deployed when +All data feeding by UDF should be Pig compatible data types +updated issues for released versions +Patch applied had to make changes in the openwire classes as well to support this +I only added adequate unit tests to cover table storage updatesAll unit and integration tests pass +Please include this feature + +okay +Applied patch test now succeeds on my machine. +I see still returns the sqrt value and other calculations are moved to +Happy to pull it forward to if ready in time +Please report back results of your testing +Updated dashboard and sidebar components to use the filtered event handler instead of data provider listener. +Can you try out the latest code +I separated the three issues into different Jira tasks per Kathy Marsdens suggestion +Whoops +When the indexed shapefile datastore was merged the +OK Im cool with that +Yeah thats a philosophical question I was always under the impression that srcmainconfig contained runtime configuration but I could be missing something +I added a test on Sep +In either case I think it will be good to open up the +Im unclear on how I can help hereA java class can have many depending on what Roo annotations are present +It is no longer replicatable after this patch +I can move nodes around and save without now I cant install any plugin I had before wheter GEF nor JEM nor EMF nor uploaded thanks Koenthis will mean that if I want to work with the GPD I habe to set up a completely new eclipse workspace and each plugin that I want to install I always have to extract it to the eclipse plugin folder +This has been fixed on trunk the changes are so extensive that no backport has been wrong fixed for release +The second import will fail exactly as described An empty project is created within the workspace +Ill attach the log which I reproduced because the above link became old +Most of the compile error were due to wrong values in +There was never an intent for it to be run more than once +The question is whether it is worth at this point if we dont have a good business case for it? I know for sure that the solution I suggest will work in my particular case and it is not very intrusive to the current source code +Dupe of. +Thanks Carl! +For the moment the Hudson build fails because of another change +closing this as of now will reopen if there is regression in future +Hi paulex the patch works fine +Attched which passes the local test with applying +we can support all of those notations using synonyms in the ANTLR grammarThe drawbacks here outweigh the positives IMO + +Please develop against this repo and frequently contribute patches +The previous problem was that this was after the latest SNAPSHOT was build +This might be slightly different +Resolving as invalid no description of abstraction layer provided. +According to me this should be and not we dont know +This looks good +I should have mentioned that the code in just needs to be dropped into libbuildr +Here are the files I have +problem be solved. +All the properties I have looked at seem to be null first name last name gid +Body contents are not serializable. +Merged +I will close the JIRA and will open a new track for and +already applied to Kevin for committing changes +database model change has been committed in rev implementation will be done in +The issue was that in rewriting the join to the cogroup the parser was incorrectly connecting the resulting foreach to the load instead of the cogroup and it was incorrectly setting the projects to star instead of two the appropriate columns +Im using but Im not able to select this version in JIRA +I dont know what the A business is +Each node N periodically computes a Merkle tree for rows in each new generated through compaction +This is just about our last chance to fix it. +Is that still the plan +Our goal is to release next week +All tests passWe shouldnt push this intoI still need to test across more platforms and see what performanceimpact is +Thank you for letting me know +thanks for the updates Aaron +theres nothing wrong with the +From Dr Wes MunsilFor me adding this to my file seemed to make everything happy again true +if leftShift calls the dynamic toString method then there is no problem because of course you can replace the toString method you just need code that respects the dynamic aspect +I removed the local drive referencesI used the create patch save in workspaceusing eclipse problem getting this committed +I will create new ticket for that. +Patch to curious why would you cron snapshooter instead of using the postCommitpostOptimize hook +Attaching Ivy dependency reports generated on the branch +Easy way to reproduce this +This is reproducable and occurs on every eclipse install we have +java +Attaching some thoughtsframework for comparing approaches I wrote last month +Anyway we definitely should fix the NPE +For the record see for the issue where we switched from nekohtml to tagsoup. +for both and with limited timeout issues is now fully functional. +Ive committed my suggestionNow if there are no in keys that match Wicket will use the messageBut I think in your case there is some generic key that matches no? +This is very important +Either that or each region server can put a watcher on the master address node so that when it is deletedrecreated they update their stored master address +Can this be moved to +Fix annoying concurrency issues +Does the patch include +thanks +M is really old M is old +Any protected methods you need to be able to extend and override capabilities that are EPP specific will be fine +There was a bit of discrepancies between how the handled null norms compared to a +Seems like a reasonable thing to do and I cant see how doing it would have any adverse effect so Im willing to go with it. +Initialize before thrift port is needed +built from r works for me +I believe this justifies raising the priority of this bug +I am not sure the logic is correct here as patched +SiI looked through the class and it appears you might have a configuration problem +In this case wssj reads the values only and does not check it against anything +Thanks for your reportClosing the issue reopen is something is still wrong. +Is tomcat ready to have ALL releases under to be deployed to +OK I narrow down the scope + +To be discussed at the code sprint +The interruptor will use tryLock and if it fails it knows the thread is working and does not need to be interrupted because it is doing something and when it finishes it checks the stop flag and exitsIf the tryLock succeeds it interrupts the thread to wake it from its sleepNothing else needs to be done because the join at the end of will take care of coordinating the region server exit +Can you publish the application you used somewhere so I can look at how it works +It needs a configuration parameter +Or was this about injection and not about the bean itself +May also require +Still missing not supported userId is not available from events Instead of move events adddelete events are reported No support for the noLocal flag Filtering on node types not yet implemented No support for user implementation for node type filtering in revision contrast to my previous comment this was a missing check for hidden child nodes when a whole sub tree is addedremoved +Ugh thats pretty bad +Im using the custom dialect workaround but I hope this will actually get fixed at some point +I hoped the forum postings gave clear scenario for the problem but if more clarification is needed Ill gladly supply it +Michael if you do this can you mark as resolvedAhh yes indeed +Thanks Owen +Ok we can remove these three files that were created by the Alfresco Maven Archetype +for the quick review +I think a will be the next step here +I was just wondering +This should give a more helpful error message +There are a few similar issues with zkpython that are surely related I must make sure to address them soonCan you let me know what version of Python and what operating system you are running +I asked this question in but raising it again here because its related +Can we see some examplesdocs on this +All set. +For example every hour to avoid flooding the serversThe feedback stream gives informations about inactive devices +This issue is fixed as. +you can pass in the core on which you wish to fire the queries +to risky to attempt a pure performance change like this for +The JDO specification needs to be updated to reflect the addition of the methods get and get to the class +closed with release of +After authentication server can query the established GSS context to get client ID as which can be converted to a String +Can someone just remove it and Ill make sure id doesnt happen anymore +Logged In YES useridFrom my perspective I felt it should be done by the container because updating the data directly leaves the cached entities in an inconsistent state +Ive attached a patch for and which should do this but it seems header values may need to be filtered or serialized somehow as field throws an if it receives an object value that is not one of its expected types +It is specific to snapshots which were created before the log files and db dumps are in new error is caused by a different error from before snapshots upgraded have a wrong installpath format +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Lets remove JMS from the showcase +Hi NathanI just put this in locally and realized something +I still dont buy it +Thanks Nigel +Here is the patch for the issue for trunk and for release branch +Committed changes described in to +It would support only JNDI connections so that we can determine the VDB and other connection information via a segment in the can leverage Randall and teams DNA experience with to use for the infrastructure +I am not familiar with the reasons why other run into this problem but for us it is that the indices correspond to userid that changed over the course of time +Thanks everyoneChristian I created the specific jdbc patch by so something must have gone wrong there Will upload new one soonJustin is good +The code in the patch looks fine +No unit test added since it only fix regressions. +The fields youve listed shouldnt allow whitespace at all +Is there more to do on this issue? Or do we just need to fix +To my mind having a callbacknotification mechanism implemented in the base classes offers a better flexibility +Changed it so that if an ACCRDBis sent and a connection has already been established it will resetThere is still a fair amount of cleanup to do for xa in the client +I applied patch to TRUNK +Chris Do we still need this patch? Can you comment on whether this problem still exists in trunk? And if so can Igor supply a unit test too +At the time this was filed Embedded was a prototyping playground that grew into the new Boostrap impl. +Your of the issue helped set things in the right directionThe final fix though needs to take into account all the operations that need to be performed atomically to createupdate the original problem was still what was stated that during archive upload it always assumed that the files did not already exist resuling in just the Create New File now falls to a File Update operation if the file exists and creates a new live version corresponding to the version that is contained in the archive +Note I just grabbed Henrys code from an internal branch for the above patch to give credit where credit is due +Petecan you paste a minimal to be createdRegardsGeorge ll love to have a shot on request included +content will be corrected +Closing all resolved tickets from or older +fixed in rwe now move up to the parent directories if the local checkout fails until we hit the root or the checkout succeeds +I agree that has the information but why should the interceptor have less information than the has to create the object + +Spring Social needs to work against Spring Security providers and also support the latest drafts +Test was fixed by Bharat +The maximum testOutput size is bytes +I dont see any easy way to unit test this so no test attached +I have a similar patch for the branchI believe the crux has to do with which configuration object gets passed to the MR jobs in +take that back the fact that is broker ensures that once a broker listensbinds to an ssl address the brokers will be initialised and used for subsequent Ssl commsSo long as all listening is done before client side connecting networks etc it will work nicelythanks hiram. +and The changes applied Moved implementation in the scr and packages down into and packages Created the Management API in the scr package Export the scr package to make the management API usable Added a Felix Shell command to leverage the management API from the Dependency Manager as a component again +Sorry for the poor Summary +separate file alongside This makes sense +bq +I am using Groovy latest release and runtime you need antlr as well +I guess it makes sense to me to use the standard property names since they are the official properties used to configure the port and the service properties just reflect that configuration right +Sounds good +Let me know please if you have any comments re the way the code can be improved for the cases like yours be easier to deal withCheers Sergey +I fixed the original problem +It is helpful point me which issue number that is a duplicate of this one +And sure having it in order tems screen cant hurt +Hey StanMy understanding was the requirement was Fuse CDC was to use repositories by default +one in clientpositive and the other in clientnegativeThanks Zheng +Slipping too big a change too fixed! +As far as the master is concerned its accounting shows the region properly assigned though no regionserver is now carrying the region +How do I join the developer meeting btw +Debian packaging does not have a feature for conditional restarts? +Note this patch is actually removing the author tags not adding them please reviewcommit asap +Have you verified these leaks? I did some work in or to address this when Tapestry removes content from the DOM it walks the removed DOM elements and clears the attached Tapestry object and nulls any function properties it sees +Or someone else can take the lead +I have another issue with the leaking memory +Your configuration states that every module has its own copy of the filter file +This should stay open until jbossws cxf is released upstream and updated in AS +Thanks for spotting this +For those users that do want to do that then theyll eventually hit this and tune appropriately +In the meantime I will commit the patch to the branch because this patch fixes a large number of tests. +Much of the work in the community happens on a volunteer basis and whatever time anyone puts on the project is a bonus and should be respected by all +Do you have any thoughts on this +I attached the working version only for comparison purpose if needed +Added reference to discussion about the location of schemas +Id be nice if some one with better understanding of than me could verify the patch +Alright v looks good +I can code generate with xmlbeans manually and get this sample working +I think these tests are now on vmware systems so we have even less control on the external effects +Can you please try it again +See interfaceOne of the implementations of transparently scrambles and unscrambles the data as it is written and read back +What do you mean when you say that the classloader is weakly reachable? is only showing strong references +I restarted Eclipse and it helped +For each child if thecolumn component has a rowHeader attribute with a value oftrue output a th element with a scope attribute with the valueof row +Fixing a typo I think I tested this by building which was silly +And I am sure that we will add more and more checkstyle rules in the near future. +I could imagine that if somebody uses CFS this could be a perf win too +Thanks for those remarks +Just deploy and hit httplocalhostThe problem is that the legacy Facelets scans the whole classpath looking for files with +Im building with Visual Studio +Do I understand correctly that removal of this parameter will be affecting downstream tools like CM and Ambari but we dont care +Everytime I request this I get one step closer to an account creation request +great job pat +I talked to Todd about this and I dont think its that simple +Modify the class so that cleanup is called in the shutdown methodWe should also investigate why the scripting feature is implemented as a module with a deployment engine and not just as a simple deployer +a very simple patch that fixes this test the snapshot and make sure the code is working well and is actually removing the hotspot +Added requested debug +Excited to see a fix version assigned to this +Was a great solution why is this issue closed with the hardcoded and no support to change the classname within? Any other workaround missing in documentation for use my own impl? versions are always future versions and set by the core team +can live without it +Still dont know the root cause +Will not get to for the release +Good point + +Yes seems like it is fixed in trunk +The naming convention for environment variables changed from something to SOMETHINGLIKETHIS +Will verify with RC +Will commit if tests pass +I think we should just pull the master and not the transient deps +Andy can you create a sample of your problem and attach as zip +Hi Folks Attached is a candidate patch for committal prepared by Doacan and myself +It seems plausible that the opt compiler should not be applying final field chasing optimizations to stdin out and err +i dont think were done with this yet +Ive had that same problem with password validation +Fixed in +Many thanksHopefully this should satify everyone. +All of my comments have been addressed +batch transition resolvedfixed bugs to closedfixed +You should never be rendering only a single radio button because its a huge usability issue +It just add a bunch of new adjustments which bring up the issue mentioned in the jira issue +Does anyone +I did not want to remove files during unload which I cannot later add to a host through the admin interfaces +actions should not come into play here and the parameters recieved from the url will all be render parameters +I was just trying write a controller to insert testdata +StackThe is fixed in +Update patch as comments above +Attaching a build of the gem with just this change and whatevers on master +So if SAP is not yet ready to officially submit this patch you just need to make a comment to that effect for the time being +Rajesh are you ok with the latest patch on RB for trunkLars I am fine with letting it go with and +We need to investigate why isnt injected could you please look at this +You can verify it when building trunk. +As its difficult to masquerade as a different IP and doing so would bypass existing controls faked should not be considered a blocker for this request +Hi Dheeraj The patch contains tab characters +Should I generate a pull request for this kind of changes or just attach the modified files here? Ive just created a simple script and some modifications in the file +Or so it seems to mecheers this is certainly a different angle from which the problem can beapproached +review board with fix on current branch +Can do it on commitDo you think the migration would just be the rewriting of the HTD component of regioninfo in +The standard Eclipse and Netbeans work fine +I dont see this profile type in if you call this dialog in Data Source Explorer either +On second thoughts a simpler solution why dont we use a simplified version of Merger as the interfaceThis way Syncsort can just implement that why bother with trying to deal with memory reservation etc +This patch should be better +Please verify in ER on ER +Forget it +There are platform related problems with the expected output +pushed + +Does not seem to be picking up recent updates +Note that I do not think this problem is restricted to +Why are we talking about regexes instead of taking Tatus advice +Thanks for the tip +Patch attached +rubuntu smoke and most kernel +Do you feel after this patch the logs are clean enough to be ready for the official release or is this just one installment +Didnt realize that Jenkins complained about it before I went ahead with the commitThanks Sid! +HiI think this is a duplicate of +I assume that this change will cause a change in the disk format and with it various upgrade issues? Perhaps also a major version bump +Change the code and test framework so it can work to accept urlBase as jdbcderbylocalhost +I guess one of us will get to it but it isnt quite as straightforward as you might like since there isnt a unique last step execution any more now that we support splits +Maybe some attributes specified in the descriptors are skipped if there is an issue on this sideMay you tell me what we need to pass to the connector to get such XA objectAlso I am wondering why there is no that can expose directly anThanks DavidQ why would one choose the approachA In fact I am comparing to Tibco or IBM MQ +Considering the new contract of is that returning null means the item should be filtered I changed the to return null if any processor in the chain does +Otherwise we have to disambiguate expressions such as which could mean or +I see my latest change broke existing core C testsIll upload a new version +That makes sense thanks for clarifying it. +The hforms that go around the active steps needed to have padding and margin set to +Basically it looks more like the request channel idea from the socket server +Yes does appear to resolve the issue +java xargs grep author cut d f xargs perl pi e s author +Or we can just treat noacks the same as act for now +Im still having the issue tested +Ill reassign this issue when the docs are available for verification on +Correction Exception is not visible when trying to retrieve empty release result for a project group +Issue closed with release. +Ill wait on commit till we figure answer to your question aboveOn the agree we dont need it +I am working on the and too close to complete +If you give me an updated patch with those corrected Ill run through testBut one question do you have any test that verifies that constant folding is actually taking place as expected? EXPLAIN doesnt tell so it would be good to have a regression test in place that fails if someone accidentally breaks the optimization +Under the amount of bytecode is increased quite a bit due to some perf fixes in the compiler and so the JVM bytecoded methods generated for this file no longer fit under the JVMs thresholdI have done some work on master to shrink the size of large literal array bytecode I will attempt to merge that across. +Move to +for the package change +Committed revision Thanks Wojciech +projectbootsbt +Please report the issue of EAP packaging in JBDS +JessicaRadhika tested the patch on master and and it applies cleanly can you confirm changes so patch can be submitted? this record since is now released. +gz files if the hadoop has the feature checked inAnother use case for it is Hives which may point to too many +That makes sense to meResolving as invalid unless someone can reproduce with the Thrift API. +Some issues Spark seems to have added Hadoop support but this is only CDH with MR support really +Attached a patch +Could someone post here the link to this page when it is created + +If you can patch up I will happily commit +Hi Benson is it not working for you with? This is the version I tried +see back +Someone should takeover this +URL lookups get new instance in rev +We had a lengthy discussion about this path over on the commons dev main objection is that the site is not necessarily built with the same JDK as the artifact itself or at the same time +spinner displays when you click the cell and goes away when the network request is finished. +The previous pqs will not compare right they wont index into the ord array for the current Reader they are indexes into the array for the previous Reader +It looks like these may be separate issues with the CXF integrationworking around the first one leads to the following on the console This would suggest that this code is not using the context classloader associated with the war to locate the classresource +This is how I use it myself most of the time but it takes some configuration files +Agree FreemanI reopen the bug to upgrade to the latest CXF version on SMX trunk +This is a straightforward merge of the trunk patch +Made suggested fix of inverting if condition in r. +Awesome Ill update ASAP and let you know if everythings ok in Solr now too +la gota que fa vessar el gotin English this is the straw that broke the camels back +Blindly finalizing them explicitly introduces polluted blocks to the system and it completely depends on NN to clean them up +async Producer queue c +If you dont want it to wait then you can use the connection property wait +Dominik this issue should be fixed with the beta build which includes verify the issue and close it + +I will update the subject. +on committing to branch as well +On a given request you use either one or the other +I personally would go that rought +Ive used the build +We are no longer doing work on. +Were upgrading Bamboo now to overcome this lack of polling issue +Fixed in revision +Patched files and think it is a very useful feature +But it seems like this compression flag would be a better fit on the message set and then use that to encapsulate all messages +Patch committedOleg +Tony I might help Lukasz a little bit +This cache will have to be aged in some manner because it will not be distributed until a J wide cache is implemented +Targetting this for rather than further pending virtual synchrony enhancements in more sense to focus on the new state transfer mechanism in +Thanks for the reviews +trunk Committed revision +There were two things that I noticed that I was wondering about In the version every creates their own executor +Fixed in + for the patch +Added documentation to show that the registrar is not thread safe and should only be used in a single threaded application. +Updated patch with a testcase for the new code is attached +I think we have sufficient coverage at the moment and I will simply keep adding to this over time. +I am really sorry +Trying to get the setting from lib +Corrected indentationCorrected coding style +I think I can work on this issue +There is an improvement of about seconds +If its critical I can email it to someone else with access for trimming the file down +Stack RamPlease comment on latest patch +Attaching httpd conf +Duh typo Thats a reduction not +The modified test fails with the patch +Many thanks for reporting it +attach my through hudson +Attaching +I have committed the fix to the core and added tests for all the dbs i am capable of testing +Thanks +Patches cut against current and the code I have no reason to believe the problem is fixed in or trunk +The locker the table if its missing in the same manner as the +The fix would be to put Collection Iterator in the request and access it with local variables only +thanks +It seems to me that the changes really help but I still see occasional failures from running in a loop Id say it happen roughly of the time +I have not yet fully studied the API +Im not familiar enough with the changes from to to identify the change that caused this +I was not able to reproduce the thread dump that is associated with this JIRA on Linux OSReassigning to Jim because he has Windows machine and could reproduce the problem +See next exception for details +Ive been thinking about this for a while now and Im honestly not sure +Ill take a crack at it andthen throw up a new patch done with the annotation change +Use the same testcase from +When the region is later activated state will be pushed +im not sure how they are related +The issue is that the current which is stored on the session is assumed to be a but it is not + to trunk +Thanks for the good suggestion! +Pat is not going to be happy about not having +Here is patch that fixes the issue in Button class plus patch for test +Ok Lars +Bug resolved as invalid. +I do not see any updates to the schema nor do i see generated and stored to the file +Here the patch that simply fixes the Schindler VM unroll bug +Deferred to defer of featurestaskswishes from to +Please credit these patches to Thejas Nair the original authorThank youChris +Juergen can you review the patch when you have timeIf you dont have time to work on it then just return it in the pool with your comments +Since user cannot select project without proper facets on the first wizards page he she will never get to the situation where file is not present in the selected project +you also remove from line +When the user expands the tree for the first time or refreshes the server a connection is then made +Closing issue as fixed +However since the current default is not to checkwe normally wouldnt change the default from what Struts releasedoes +Not urgent +Issues in it are to Nick for feedback +provides a mechanism for binding job parameters etc. +David let me know when you will start working on this maybe we can split some of the tasks +questions do you want to make this new behavior available by default or by using an option? why to only batch the update association operation +So here is the JAR that worked for me +Ive committed this to trunk and. +I dont have access to the project right now +What is the for needing to access the nested class +Bulk close for +I can still connect to the repo in my service and test my rights against a JCR nodeAlex interesting approach will look into thatok looks like this bug can be closed +thanks thats something to look at +Manually tested +Yes +Before its included convert LOCALONE to s just leave old and around and add new classes that use +RightTitle should be gml parser not working in special installation once the issue is created it can be edited only by a can confirm that the problem is resolved in +It is also possible to add failfalse on line and to package as its previous state +If youre busy I can finish this +Thank you very much for the patch +Here is second version of the patch that moves the class from the generated code to package +If the datasource is actually enabled and you still have this problem please. +The patch is a sample and we can improve it furtherThanks RegardsAnkit Jain +Fixed by r +In a scenario where you want to have a correlation id through several intermediates ws addressing can be used +fixed on branch +The exception indicates that one of the archives has an empty name +Lets keep this closed and manage the archiving of things elsewherebayard is a file containing the SVN repositories to migrate +It replaces the property dollar by the value +But if you introduce filters some parts are lost and when the intermediary is gonna send those parts are missing in the out message + +How those are used in the component is up to the components designerJacopoThank you for the illustration that helps +applied +I tweaked your patch a little bit so that the two tests maintain their initial intent to test both MAPPED and UNMAPPED roles +The generated artifacts from mvngenerate +I fixed that bug in rev +Updated version +This patch makes work well enough for my purposes in triplesec +Is this addressed with together with +Ill commit shortly +Fixed in trunk just deployed a new snapshot for testing purposes has now been released to the maven repository +Please use the spilling buffer implementation from patch for performance +Andreas VeithenIm not sure is there any issue in the client code +Arrays can be easily pooled further reducing garbage collection +May I ask which version of was running prior to being deployed? the row was manually deleted twice I guess there was some client hanging around because code base would be using only one format +Hi KimThanks for making these changes +Thanks +replaced by issue numbers +Please set line in last uploaded log you code around the missing the script also refers to bin which also doesnt exist +Server is started and module is deployed editing disabled fine +Thanks for doing this Owen + +The test still failed the mocks is not satisified i did a try for that before committing +Thats completely broken we should make that abstract +OSX is not identifed as a required platform in the EAP PRD Operating System and Architecture combinations +Only issues that are actually fixed should have a fix release. +Already fixed with commit +We should resolve it for as such with a link to the issue it duplicates +I had a quick look at these methods +Kenney I just dont see how can you do that +Thanks Erwan +This will allow us change the role when the NN failover occurs +How about implementing using jflex to be minus NUM and ALPHANUM +Ive been able to reproduce the error and my preliminary investigation shows that calls made by remote clients directly to the are being secured as expected +Hiwe have the same problem when using the following configuration in our POM +Verified at +The patch fixes the on startup +moving this to Beta since we already have too many high prio items for Beta +Do we want another method which takes a journal IDnamespaceID for journal related to a specific namenode +Because it works on my ubuntu too but not on a win machine +I suppose what you need is the existing java bean property on +pep compliance and docstrings plus fix about kwargs Tomaz I attach a patch +I took that part and changed a bunch of encoding stuff +fixed in headSending etcSending pluginsSending pluginssrcjavaorgapachegeronimopluginassemblySending pluginssrcjavaorgapachegeronimopluginassemblySending pluginssrcjavaorgapachegeronimopluginassemblyTransmitting file dataCommitted revision +Resolving as per pskopeks assessment and changing the fix version to so that it can be verified in the next round of QE +This was a config file issue so closing the ticket +Its not a Java Spring service its not a class library +applies the fix to the jar indexing codepath too +patch patch applied at revision r thanks a lot for this enhancement please verify that the problem is fully resoved as you expected +Committed revision have updated the writeup so that it matches the committed patch +No the intent is that it ends up in Unscheduled +The only remaining question doesnt call validateJVM I dont see why not +Resolving again +Now the test passes for me +Thanks Yi Xiao for the patch. +Is this a case ofThe core test failed is not related to this JIRA +Somehow somebody who needs it eventually will figure it out how to make it whatever +Closing this issue as resolved +Could we have some unit test about +Do you expect a significant performance difference hereGood point +There was a bug in the +Please try it outIf I hear no complaints Ill commit the patch and close the issue in a day or twoOleg +Build to see if the bug still exists +Thanks for the information +Hiroshi any luck or progress with this one +is often really not that slow for many many applications +try it again now so you also get snapshot contained fix too so its all good +Here is the final patch +One comment the isnt running as expected +The order in which the children are added in the array and then extracted from the array is the same +Ill commit the patch after the milestone lock down is over +is partial fix for this bug +No objection? This is the last call +I think you want to use it simply to build an index using benchmark which is perfectly fine +Initial patch on trunk for review +I guess we need to update +Not sure about any usecase for getting some advantage by making it configurableDo you have any usecase where we may get some advatages by tuning that parameter? If yes feel free to file a JIRA we can discuss about it a lot Ted for taking a look on itThanks in upon release of. +Made it a blocker +Support for reading composite row keys and add integration integration tests and support for writing composite values back into ve removed the original patch and added two new ones +It only supports switching checksum algorithm number of checksum bytes has to be the same since that determines packet sizes etc and would be a more complicated patch +bq +So the issue is resolved +Youre right +fixed all generators are stale resolved issues +Its fine as long as it starts with servicemix and clearly identifies the component +Arjen since we discussed those but do not seem to have committed an actual fix yet Could you fix this in trunk since you know exactly what youre doing? Ill care for the backporting to then + doesnt use annotations which is touted as its strength +I recommend that people use a single geronimo plan with all the stuff embedded in it so having a separate way to specify directly in the seems to me like overkill and too confusing. +on caching this +but using it mightcause a problem because it is now a separator between a nested class name and its enclosingclass name +While the desire to limit duplicate data makes me favor leaving off the execution id I dont think its a deal breaker and would be open to either solution +This issue can be closed +Any application role to security domain role mapping should be done at the security domain configuration level as most users complain about having to manipulate deployment descriptors for this type of task since it couples the deployment to the security domain being used +bulk defer of featurestaskswishes from to +I added code to safe guard against the potential of nulls being used in the hashCode andor equals methods thus avoiding any potentialThanksHeath +The basic control bus is in place +As long as you do not provide a unit test where this problem can be verified all we can do is to close this issue as not reproduceable +It was the same problem +Doug committed this. +Solved by applying the patch in revision r +If there are no concerns with the overall implementation Ill submit a rebased version for trunkThis looks good to me I like the abstraction +Fixed in transition all resolved issue that did not see any further comment in the last month to closed status +Fixed with since is now released +OK so it works with curl as wellThe trick is to use the curl G d cmdupdatewloadbalancervlrvlrivltvleevlxvlsonvlmvllmimetxtjkstatuscmdupdateResult typeOK messageAction finishedSo the option G and d must be used +When run from the eclipse jUnit runner with or without the meclipse plugin as there is not integration with the filters provided by maven the test case fails +Setting this to review for that I have only checked it with JavaI have now updated the pull request with a second commit to restore Java compatibilityI have used reflection instead of extending +Attaching patch a which cleans up a few minor issues o converts docs to Javadoc o remove stale doc o correctimprove some comments o makes physicalConnection o makes check finalPatch ready for review +Simple patch to fix this problem +It looks good to me +Will post it shortly +I see your point about the inability to satisfy both contracts +Ill see what I can do in ourdevelopment environment though +Ill take a look at that +for the first patch +Also I believe there are subsequent issues been raised against command line anomalies but if you have a particular specific easily reproducible example that needs fixing feel free to raise an Jira +HiCan you attach a thread dump? This would be more useful than truss output if you can get it +I was able to get clean local and server based builds today +Please verify +So should still be able to run with Hessian as long as youre not activating that debug exactly breaks for you? Do you get when running against Hessian even without that debug mode? Please provide the exception stacktrace in that case +HiExcuse me for hijacking but analog CUSS is really strict in case of any errors it gives up and throws them back to clients +Note can be applied to the branch as is +Fixed in I also fixed the password for the cacerts keystore on OSX Snow Leopard which has been changed from changeit to changeme in a Java update +Please reopen if you continue to have problems with this +Forgot to mention that this dependency between test classes across different projects works when test is enabled without testno +What if there are different Vector classes and some empty slotsUsing Matrix is the default policy in other places so lets go with that +Do you have no concern about adding a jsr concept to +If someone can give me a version that fails I can do some further investigation +Hi Brandon please run and all the unit tests on your patch. +Pushing to as this does not appear to have been completed +Todd currently has a pretty solid use case and after talking it through I think other applications could easily benefit from it as well +Cheers thanks for the patches +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +So must be something to do with +Its true that when is not available it will be slower +I can deploy several of my own jar files multipletimes just fine +maxIve updated copyrights to state by Red Hat but ref Jeff Mesnil explicitly + +It happens whenever is updated +messaggi non tradotti +Verified on JDSB to set fix version. +v looks good to to snapshots branch +we can reuse the compressor by having one DeflaterInflater object per WriterReader instead of the pool right? Even then creating the stream with the above compressor will probably be quite significant for where we would need to do this per record? Appreciate any further details on this one +As far as I can tell and Joni look good +It is simple enough to push a manual refresh button if I wan an update +The normal semantics for invokeMethod is to be invoked when no method was found +In return can you take a look at +Not all machines but it did on my MacI flipped the schemaCache to using the same keys as the definitionsCache +You may give it a try +is this the final patch version? im unclear from irc if goffinet had more changes suggest the attached skipping the nd nextToken call when unnecessary should be slightly faster +They touch a lot of the same code and we should be able to use the same counter for both +Ill update it shortly +Could someone take a look at this and let me know if this test will be excluded? Thanks +This improves the previous patch by including tests of the new return a map functionality for activate modify and deactivate methods +This issue is no longer relevant for it is fixed inChanging the affected version to +Thanks a lot +Id suggest adding javadoc to the load balancer explaining how it worksI tried the code +Closing resolved issues. +Please see my comments on phabricator + +I dont think RPC should dictate what exception should throw +Thanks for pointing out this +The term heavy committing is just a jokeI know I wasnt using it very seriously myself bq +Cool Thanks Ashutosh! I will try that +Yes this does sound like a bug +I work with are seeing similar behaviour also when using splitteraggregator on a direct channel +Oops I missed this jira +Vitali I remember this as being a duplicate but cant find it +The contains the following +The default on Fedora certainly doesnt +Attached patch fixes the HTML link to simple example +Merkle tree calculation failure Data streaming failureThose have been addressed in so lets revisit that issueFor point one we need to change message to report back failure so we have to do that in next major releaseBut point two more likely to happen than point one fix that alone should be in minor release +I am running JPDA tests to see if this patch doesnt break anything +Just to be clear does this just affect and works fine for you or is it happening for both +Moving a check for the timer name during endpoint creation should solve the issue without ugly hacks +bq +I created for that purpose and added Roberts latest patch to itI will like those two in a secondbq +Unassigning as Im not explicitly looking at +For large we probably dont want to enqueue too many of them not to waste memory or even get out of memory +thank you +Can you please try the release +Thanks Knut! Ill start working on a diagnostic patch to collect some info so we can know more when it happens +removing locking +The same patch is applicable to +Merge all changes to the trunk +If anything I feel like the burden should be on Hudson to run the clean first by modifying +From Pete just need to be on one of the +Ill open a new issue to refactor the dispatch clients. +Passing object to API should be fine +Dave is wrangling among your skills +It seems that the original purpose for the definitions has been lost because both SIFDIR and symlink are defined and declared respectively on modern Linux +Dave Thank you so much for being super responsive and for putting this module available for the m really enjoying using your spring oauth m going to work on a blog post talking about my experience with this moduleand how Im customizing it for my company business needs for example using a custom grant cutting edge +Though I wouldnt be opposed to adding a flag to nodetool to display token as hex if that is useful to some +Actually its not so easy as that the offending hardcoded JNDI name is in not in +The class had a private helper method that did proper quoting of SQL identifiers +The patch looks good +We use Nexus at work for the same technique and couldnt be happier with how easy it is +Markus thanks for catching thisi would definitely consider this a bug there shouldnt have ever been any restrictions on the type of legal characters in the key +Attachment has been added with description Patch P open bugs to P pending further do you plan to commit this patch +I dont think anyone objected but no one thought it was important either +Running the Daily rolling for the same day twice was failing because the output dir was already there +It could be an environmental issue but would be nice to know what is causing it. +Thanks for adding the example I did not manage to complete mine +Looks very good but IHMO it would look even better by removing the center logo +That sounds like a reasonable path and a sound workaround +This patch adjusts the release notes changing the release id from to +Are you using? Can you try with latest +I think that the correct behavior is that is should throw asyntax error +This issue has been resolved in IIRC where the RPC handlers are started at the end of all initialization +As I said by email the will be used even if you dont pass the cookie param to the JS function +So should I implement my own Reducer to set them instead ofIn my opinion should be able to used easily in any program +replaced tab with unless their is a way to do tests with shell script changes + +Here are some tests which fail due to this defect +I also think that people are more likely to check for changes when they switch to than when upgrading to +Actually it relates to +A workaround is to delete the from the generated lib directory of the should the jar simply be excluded from the d project +Hi SebastianIn the normal case when you deploy a widget with no id Wookie should generate one +The real fix will be to update the POM file for hadoop core thoughI tested with latest released hadoop and I am able to compile fine +Many users may have the impression that they have to put their one by one +There are three files that are stored directly under the folder sandboxtest and +The log snippet were not complete though so it is possible that the indexing task for the xmlbeans artifacts just wasnt executed yet +Thats a far cry from everything you start with is an exampleOthers with pre typed json dont want to have to redefine types +updated patch with all of lucenesolr and including uwes stuffall tests pass + for the to trunk +I planned to work on with integration of Jetty which will be in version of +Hadnt thought about unique values but in principal its pretty straightforward as we hold on to all the values as we need them to count thingsMaybe create a new ticket for thatI will also think a bit more because often what is wanted is not always a stat but the ability to transformreduce the set of values coming back +Ive reviewed the patch added some comments and committed it at the revision +Happy to spend a little time with you offline to discuss the design if you like. +What is the correct way for a former incubator site to redirect its website? Surely everyone needs to do this +Via auto discovery by placing the tag library discriptor file within a jar on the web application classpath naming thefile so that it ends with +The basic idea of the patch is to separate Locale and libraries from the common library +Otherwise Ill commit it on Monday. +Please clarify what you mean +Better to do a pull request +Ive published a report server +I can look into doing the work required to use a full if you think that would be more usefulThis feature does seem quite useful but it could also be implemented on the protocol handler level without having to change the existing IO session API +Bulk closing stale resolved issues +As far as I know this issue is resolved except for the possibility of ambigious EJB references which will be resolved under +The class cast exception in was fixed +in practice learn messages wont be lost randomly as in your exampleThis is really not an assumption I would make + correct Fix Version to one without x like depending on where it was ok now closing +tested in a cluster using cgroups and works as expected both the delete and the timeouts +the correctionapplying your patch is making the right column of the screen disappearing +Fixed and added to the unlockable accounts list +The operation result is shown in the GUI +although Im stumped as to why I wouldve strayed from that Ive found the code for story and story tags laying around on an old machine and will marge it in soon +The goal here is to discover if this is a tooling error or a runtime error +This has been resolved in Apache and will be in the Release +renamed currentMark variable to trackingMark +Updated ESB docs describing can you get these changes merged in and staged today please +Replication doesnt work before because a second parameter of putremove methods which was kind of ignore listeners was set to trueWhat was fixed now you can replicate your cache across nodes using listenersOf course it was tested in real life and using standard test not granted license to ASF before so uploading againNote that end lines are unix for the patch. +And indeed we need such a facility especially for logging +Okay Ive modified my code to now take instead +If we time out we should probably cease further delivery attempts to avoid hammering a node that is behind +Thats what I learned since creating this issue +patch added fixing the proxy user authorization +bulk defer to +it is good for me +holdqueue requests cannot be used in all the problemsTake the case where there is recursive use of a function and the recursive function is called through the Remote Interface which again goes through Layer and so thus this holdqueue will not there has to be way by which some beans have to be allowed to make concurrent callsBye for think an extra interceptor could work out nicely any hints on how to implement this interceptor? Is there an interceptor we could use or use as a good base? We would need some kind of waitingsignaling to implement this +I agree that they are confusing +Ive just fixed this issue by modifying to reallocate its internal buffer when buffer derivation is detected +Closed. +I saw that RTE should work in SBS but tooltips will be in Thanks +The patch looks good to me Dag +Check whether the failed fault injection tests can fail the build +Confirmed bug in Jettison +See for related discussion +Debug output when executing example target that triggers the error +In that implementation once a job has launched a task it can keep launching tasks right away without further waits +Added few integration tests more need to be added. +i just confirmed it no jsession id is appended to the url not even on the first Ive been testing without cookies like crazy and never ran into this +It sounds like you are vetoing removing the docs from the build and I would veto their removal from the web site so we are going to be stuck with finding a process that keeps them in synch +Still a problem as of forgot to say I use Hibernate and Hibernate +Horizontal offcet take place under Opera only. +Just run the plugin as you did before without the new encoding parameter that I added and see if that works for you +The testcase ran locally when I ran it with the patchBased on this I think this is good to go +Im still having trouble seeing this +So while the scans running before close would see no delay scans started after close will still see a delaybq +Created separate issues for other plugins +You should see the new please wait for page to load message instead of the incorrect behavior. +The code will retry the transfer multiple times and still timeout. +Applied the Lite transport specific changes as well +Closing because scope is broad +What exactly is not working for you? Could you please give us more details stack traces log messages sample code or the likeThanks plugin doesnt support but has upgraded sitemesh to mark this task as not resolved +November th is the Red Hat Government Users Conference +It is fixed now +I think the improvements to the for loop optimizer cover this. +It would seem its some kind of Groovy level Premature end of file thing is a Firefox related issue theres a JIRA issue for it and a fix +trunk +Using system getproperty and if windows machine changing initial path slash value to correct format and passing that all the way through to the absolutepath displayed at the confirmation pageJoseph patch on a windows deployed AG to verify before posting patchJoseph Available! At minimum the user should provide an artifact id otherwise assembly named will be created +I committed a first iteration I have more refinementimprovement coming +Assigned to Dennis as he is the conditions specialist +now some of the things in the api also reside in the impl package which may look weird but thats the way it was before and can not change for + et resourceloaderphaselistener have been removed +I dont have a much awareness about the issue and how to resolve it +Can open new issuees for more fixes. +Its possible for root location to be null +Test case +Javadoc external links should not destroy a build +Can you attach a screen shot to avoid testing the patch itself +Great thanks so much! When is the next official release containing these fixes roughly slated for? +This lets us eliminate canons for the blobclobBLOB test +Also tested with regular spot +Docs will be ready for release at GA and not before +Note that I already removed the getters and setters in the latest uploaded zip +As the resolution is deployment related I am closing this bug +OK let me see if I can express this more coherently than I did in the previous responseIf you reserve a byte for then you will not need the code to and copy the string as your patch does +Im not working on that place anymore +Provided patch +It should work now +Marks and Tomas non default field patches are combined +Please conduct discussion and contributions on the mailinglists or via the issue tracker +Attaching the screenshot of the error +Ill add them to the unit testsCreating a new entity is not supported when the session is being flushed +if a write lock outstanding on a table then the balancer does not balance regions in the locked tableI like the above approach +Fixed by upgrading to a later version of CVSNT +I doI intentionally worked on this patch to get it ready for in advance because the discussion was concluded and everybody seemed to be in agreement +downgrading to minor as agreed during bug please verify fix with when available +Thanks +now onto the other modules +Committed revision hopefully we can take some more steps here soon to make SRs norms handling less scary +There was no NPE when there was only version of the artifact +I am extremely sure this was the problem that caused the error he posted here. +I see some code merged into the branch +I will post more when I find out +Correct and this is not a bug +by r +Simple fix is to change to use springBut we should really be using Maven dependency management at the root level to avoid this problemIn general there should not be any version numbers in any of the +icons this i get when and try to open its Outline view tree +Im sorry for the delayAlberto can you proceed to test the fix and see if its solving your problem? In my tests the connections are immediately closed so you should expect no issues here though +I have checked the server part of the generator for C Python C PHP and Java and they all correctly return the seqid sent in the requestThis change only checks that value and throws an exception when they do not matchThe current behavior will either fail with a protocol error or return an incorrect result +Really a nasty bug +in the meantime i use the package explorer +Also the new test cases in lang run correctly in both codelines +Thanks a lot +Good catch fixed +Thanks Pritam for the work on the fb branch from which the trunk patch was derived +But thats not the right fix +Have you tried overwriting the sax parser dependencies to a higher versionPatches welcome. +Committed branch and trunk +The wiki page looks great to me +Will create a separate Jira for that +So would be a good placeA thing to explore would be if this can be defined in the proto definition of the API +Just to let you know Ive got the confirmation of being on the list now +Java QMF API for clarity the update is the K dated th Jan should probably have given it a different nameAlso to confirm its all released under the Apache licence Im not sure if I was meant to tick something when I did the file upload to confirm this? But it IS all Apache update of the QMF libraryContains fixes to cope with the changes made to how amqplist is exposed to JMS in QpidNow allows disabling QMF Events +Does a mutation even allow more than one identical puts +Verified with JBDS. +patch changes the rpmdeb name as mentioned in the comment +We havent been actively maintaining I quickly ran the test case but did not look deep enough +I dont think that is needed at all +Thanks Aaron +Verified with GA +Maybe it was though that plugins could have their own files but doesnt support any of that +This should work if you use the included as part of the plugin instead of the provided by +defer all issues to +Thanks Raghu for his effort on investigating this issue +Created for root cause management in the future +Ive added NPE guard against IS the spec says that an empty content in case of JAXB should result in otherwise a corresponding method parameter has to be set to null +Corresponds to attached rm audit log +How about modelling this on the package structure for our build tools? We could use the following namespace for Derby demos rather than +I agree we could make the dialog text more explicit +Patch removed dependencies on webhcat +The feature also very crucial for me since it allows me to runmultiple of the same using different system paralelle saving lots of can work on the patch if needed +This issue has been resolved for over a year with no further movement +All issues should be resolved in build +Committed to trunk +uname aLinux studio SMP Wed Jan UTC x GNULinux fop versionFOP Version rewrite this bug report in terms of content +Its horrible with those null checks +This wires the release into the website +Updating User guide +Hello AlexeyIs your admin password shorter than characters +Resolving. +now i hope only in your help! thank you very much. +Closed due to the release of Tiles +Is it okay to port the doc fix to the branches as well +Attaching and with these changesM srcrefD srcadminguideIm removing the Error message differences topic which Id removed from the map file but inadvertently left in the workspacePlease let me know if more fixes to the Autogenerated keys topic are needed +Perhaps they can just be out back +Thanks for the patch! +Feel free to this issue with the requested details. +Mina version +Ive checked that this works with the geronimo directory plugin +Using this the user can see all actions take place +Going with WONTFIX +The patch looks good +The way I understand this oneOption When a JT machine fails another machine can be brought up with the same IP address as the original JT +At present i need to bundle the relevant Spring +fixed trunk rev +bulk close of all resolved issues. +we import Maven project and should be able to deploy it without tweaking trough WTP can you please open a bug report on WTP about that warningerror issue +will take a look +I have not the problem now +So it is not a critical bug but it is a +The extra commitlog write is going to dominate the extra work anyway +check to see if it is the right exception as thrown by the Noop Operation if so drop in to the block section +I committed the patch to trunk and +Also Im not sure how to cause the error +Its not uncommon for people to store objects serialized to JSON or other some other serialization format into columns +instance the changes to srcwcssrcmainjava dont actually seem to change anything unless i am missing something +Committed revision . +Thanks Marshall! the failing test seems to be the one addressed in +Arun what do you think of tests verifying lock ordering on all JT methods? We can verify that we hold JT lock before holding any other lock +This was committed in and should be in the merge please test with SOA with +This usecase can also be applied when working with webservices as they tend to work with errorcodes as wel +My usecase is something like I described +Wow that was dumb on my part +Certainly I will look into itIf somebody could stick the patched onto gist and paste the link here just so I can be sure Im looking at the same s just the vanilla from the is passing in Windows now on trunk +Keep up the good work. +bulk move of all unresolved issues from to +Assigning Backlog as Fix Version as that means this is a qualified issue but we dont know when it will be done +The Service and Shell classes seem usable +If parallel to sequential was straightforward we could have done it here but as it is not we should have another JIRA to look at it +Patch looks great I like the new unit testIm going to try to get some of our tests to actively test the other codecs so we catch this sort of thing sooner +Rather than go through all the work to move them better would be to remove them from source control and make the build process generate them +Merged revision to from to branch +Making this a feature something we should consider in later releases +Closed the wrong issue +Even after trying many trials still struck in the problemAny workarounds pleaseraghu +Since I am not a contributor to logj I am just going to attach the source files here +Ive applied your patch thanks a lot for your contribution +Proposal create a that defines the version constant create a VERSION file at the root that contains the version number should edit automatically to copy value from VERSION and svn rev number to add handler for version switch to print version would change this in a few ways +Altered the regexp and what should replace the the matched text in commit abfdcfbddbd +I think we should mention the upgrade issue in the release note for NATIVE authentication +Same comment as in +should be fine +Ah I read sorry +The code for this in needs to be cleaned up a bit +Committed as r +So I would assume wed simply turn that off if running on +Whats the best practice in these situations +that would indeed be a nice to have feature but its not in our current priorities +ant testreport in hive uses junitreport task whichexpects each test to produce an TEST +Ill keep this jira open until we move have to update some of the requirements in order to verify this works as expected +This is more of a performance issue where users do not detect till they see that reducer job has failed in the sort phase +Ive committed this +jar though +Updated the wiki page for it +Attaching the patch file for earlier mentioned change +I need them for development +Im using the default implementation and the only problem is a space in a file nameTheres no difference if I use tmp Americas ortmp Americas in your second check programm +Patch is applied at +Ill work on this for +bq +Probably it tries to solve issues in special environmentsI can live with the provided fixWhen will there be a new release +Heres a patch that implements it +In addition to answering Deepas specific questions It would be good if someone with some Store knowledge could comment on this assertion and estimate how serious it is +Thanks +. +Of course youre welcome to propose a different project structure +Camille we should remove ready but it is still used in ZabTest which uses it to check that the leader is ready and able to receive connections +They may or may not appear in the commonsnightly location +Added a tutorials section to the articles page like in the wiki +More info the target usecase is that parent pom is released as a separate project +I think making them stand out more would be better darker font bolder + +Just uploaded a patch for +Patch includes the two implementations and good idea! Didnt pass my mind Ill go ahead and create a patch for this +I also didnt do any tests so far and I didnt move any of the new stuff in the old file as per Ryans latest commentId favor keeping it separate instead of mangling names etc +Single quoting the whole password is easier that checking for special chars and quoting them individually +Fixed on trunk revision . +patch b to trunk with revision Will backport before closing +Patch added +Were going to hold of until need to put the security in place before doing this +Looks like a duplicate +Lukas can you test it on linux and close this issue if it +Thanks Chris for the Hudson snapshotI checked in this snapshot and the statement in the manifest looks good. +Im not sure if you are suggesting an automated way of detecting if the jars are available before using JAXWS function or simply that the user would include and its dependencies if they knew jaxws processing was required +Ill try to verify the fix using integration test +is stuck using metrics until all of our users move to since we need to support both and and the metrics API is incompatibleOK you can add support in along with the older API thatd be deprecated +Can you attach the unit test which fails for youThe attached test case works just fine! In fact it doesnt use any thread pool at all +I wouldnt think the mere existence of a being throwncaught would signify a dead persistence any case starting a tx then marking that tx as and then not rolling it back seems orthogonal to the question of whether the persistence context is dead + is being deprecated in EE so I would ignore it +Code is safe in that nothing happens if no is registered and exceptions in the CP wont kill the master +Consider for full well be able to just check requires just after parse +The method not is cache not cacheable in addition if there is an issue then its in the Hibernate layer and you should raise a JIRA in the Hibernate reading from various messages in grails mailing list it seems that cacheable i widely used +Any ideas on that one +Arun are you suggesting that the job submission process first generate the input splits then determines if the number of map tasks is smaller than a certain value and if so then set to zero +Some solutionGeoserver +We just need to do the following for these casesOld final new final Old +Moving this issue for now +Other projects that have attempted to move away from will be a low priority and best effort +Have reopened +IT added +HiIm the PMC chair for the Pig project where Aniket is commiterI can confirm this is him +Is there another issue somewhere I should vote for? It would be an immensely useful feature and I cant really see a good reason why it shouldnt work that way +Unit Tests that expose the defect +sounds good so i guess weve come to a conclusion to only include the Intellij and eclipses ide related configuration files to the template project +I did some tests on by using CXF and CXF after revert the change of all the tests were passedJust as Dan said can run with CXF now the can still support CXF and CXF at the same timeI will committed the change of into the trunk after we fork Camel branch +Backport to checked inNils please review +Stack should we open another JIRA to deal with your TODO +SPRINGCLEANING +Maybe you can it +here is the snapshot of the just to set the fix version +It includes many new tests and increases coverage generally +Committed to trunk +ping subtasks complete. +Verified in ER +More of a font already included in iOS thing with workaround by Brennan +If so Id like to commit it in and move on to infix suggestions support maybe +As Workaround setting a seems to solve the problem +Thanks Brandon! +Let me change this once more +Thanks Brock! +Fix in place +Committed patch to SVN to resolve defect. +Ive applied the patch with Bryant and Nick suggestions. +Please do not create blocker issues without asking for help on the Forget the hint with the derby database +If it were a question of making backwards incompatible change to navigator or deleting the pattern package i would favor deleting the pattern package +Verified that this is still an issue added to the Server component +One for the old schoolers who dont like executing everything from ant and another for the Future the people who like their XML frontends. +I created a testcase in which I push an event every two seconds +What do you think? +Unless there are specific problems that where not addressed within the original JIRA I see no reason for keeping them in this state when most of them where fixed and resolved almost a year ago. +The deployment works fine with each module as snapshot and I have cleaned my repository before +doesnt apply cleanly anymore +It is a backup for the whole JIRA though +HiCan you please let me know the solution to this issueThanks +Could you please check if the attached patch fixed the problem +Thank you +It seems like this also affects you guys close this issue? +Thats odd +So for this to the trunk and branches. +First patch adds our own Progressable class so we can dodge most of the mapredmapreduce mess and adds it to BOF in BRW where we report progress anytime the loader reports output +I tested this with Wildfly which contains Hibernate and Infinispan and it reproduces there now +This is a candidate for SOA +bq +and dont provide the actual number of runningcompleted tasks they only seem to provide the progress as a float +But sometimes it doesnt work +This issue is basically complete and is just lacking some documentationThe name of the setting was never changed and is stillAs stated before to enable use of new importer for all Python interpreter instances in other words for everything will be necessary to specify To enable use of the new importer for selected Python interpreter instances instead of a comma separated list of interpreter names can be specified +Yes use my v patch and JDs initial patch +its not reproducible because the fix was added directly to the checkstyle think we resolved to fix this in the core eventually and to work around it in individual plugins for now +So there is one advantage to making this a k think this issue is an important one so it should be living somewhere until done. + +Furthermore I addeddocumentation for itDo you need the integrated part trying to handle every file as image as MP andas XML file? just to set the resolution to fixed +If you do the same operation in the generate it works +But adding An override should work since it will not change codes? Dont all overrides have same code +is the base class of and +This support is described section of the Java EE specification +Removed in Git ID faedfcec +I added the deprecation tags for the release +I tested with the patch and got the following results +You have to understand that issues that cover it does not work are uselessI do not see any reference to problems withAs covered in the migration guide is not meant supported for use yet +One concern with implementation is if some of this logging is happening under the jobtracker lock it might impact performance adversely +Attaching a utility target to display all system properties +Oups sorry got it now it just doesnt happen for the bug report well look at it +Note that the Ant Task does not support listeners but the Ant Task does +Adapted Revision Joachim can you please document your changes at and set the fix version at the jira issue. +Sorry for the duplicate message still figuring out the patch submission processAnyway the patch is uploaded all existing tests pass and I added new cases for this specific bug +The inconsistency seems to be due to the font names not being retrieved properly +Uploaded a patch that should fix the findbugs warning +Oh yeah I stopped route before I copied the file to I guess stopping the route may cause the JMS connection to stop and force a failover +were backported to as well adding the latter fix version. +Does this sound like a reasonable approachMy initial attempts dont seem to make the query use the index +EAPEWP contains of Security. +Although I imagine the code inside get and layout would be fairly uglyAt any rate looks like anything we do will be nowAdding complexity is not an argument against adding a new feature +Any way to test this case +Please feel free to any of those when required. +So we should only use it if we think it will make writing Hadoop easier if it consists of code that wed need to mostly need to write anywayA of URI is that it provides filename syntax +looks good +It also only worked after applying it would be good to commit them both togetherIf no one else expresses a concern with this patch Ill commit it along with +So you have to declare it in your +RobertWhy do we need to write a special tool for it and not just use the maven assembly plugin to generate this kind of stuff +Removed the catch statement and changed condition to check maven two executor id +The README +So you mean treat the phase for the mem pipeline as if there was exactly one observation per? But that one observation would still be processed by the combiner so like if the combiner did something like take the values sum them and add to the result I think that logic would get executed twice once in the combiner and once in the reducer +Make patch available for test of patch +Simply unzip the archive cd into the logisticx directory and run mvn install +Actually there is a hack which is to call doGet on ALL HTTP methods out of service which is a backdoor way of calling serviceWrapper given the current implementation +Now the real cause is that the task is not fully ended when the process reaches the end state +trunk Thanks Stan I have applied the patch. +No I mean we read both and then merge the contents if both exist +More recent releases of Shindig have replaced juel with so I dont think this issue requires any further action +Attached patch +The code hasnt been committed to neither the trunk nor the branchMy opinion is that it is a little early to declare victory but your feedback on the patch is very welcome and may speed up the processThank you for taking the time to report bugs and testing out patches +and needs to be undeprecated +Patch file that changes the and API +Pedro what MQTT client are you using to connect to the Broker with +Hi Jean the issues cant be caused by resource optimizationAbout the packaging of custom skins resources there is already JIRA open +Thanks Jason! +First the should trigger replication and remove it only when the block is fully replicatedAll this logic would work fine with my patch if the issue with keys was resolved +Update to avoid the bug of determining the transit SAVING state +Some time ago we decided to expose the PVM apis and only provide an extra convenience methods in +When there are larger version differencesOn the other side you already have dublicate libraries on your class path + +Applied +Youd have to use placementstrategy +This has been fixed. +fixed in revision +Thanks Eli I got backed up this weekend +Am not sure it is safer because the filter introduces a change of behaviour to the normal exception case +Pete this validation case is implemented please attach an example with which it does not work +why cant this be resolved by just using the seam jars in the classpath +on patch +Getting this far in the configuration process depends on solving the libm problem discussed in +There is a bunch of relevant configuration around the cluster update feature filters etc and it is configurable per transport connector +Changed the predicates filter and also done some modification to one of the servers content types handler that did not process the expiresIn rThanks! +Please set omit to true if you dont want them to appear in your markup. +The user said that it works in local mode but not in production +If you can run for a few days and let us know if you have any issues that would be great +The question is what is level of risk to the NN and is it acceptable? Lets say the db driver segfaults or scribbles on memory +assuming this is for nick +Also seen with anyone tell me if this can be circumvented some way? Im migrating from a to database soon and cant afford my system going downAnd if someone knows a fix please tell me +Ok +Heres a patch +Normally i didnt commit anything without testing the whole test framework more than once +If anything we could have extend Collection but I really dont think theres a ton of value +I checked current seems moving all properties is straightforward +Verify that everything works and then close it +Its a bug and we will not do it for SOA same fix as for and added a link the which invalidates the session and uddi token +Failover properly works even though standby nodes do not publish information in Zookeeper +changed locally yes my next commit will include s fix is in GCL now. +Good if we could easily fix it to have it in but otherwise for next rev +I feel like adding unnecessary levels often gets annoying to both write and read. +If you would like to polish it go ahead or send me a list of things and reassign to me +Please write detailed instructions the patch does not work +It certainly should not run you agree +yeah you are right moved to RC +tnx for iplemented +Reassign to a release that you feel comfortable resolving this issue in or leave as is. +Fixing to merge cleanly against changes made in r +Thanks a lot for the review Todd +Dupe of or +this is a patch with doing a rcp call and parses the remoteexception +In that chain Yi is AB and theres no way to compute that on the fly +So Scott did add a workaround to hide that problem in AS in +The default target does the following +Added DR vs +There is the additional issue of whether or not the person posting this request actually does own the copyright +Cloned ticket and will cover docs impact in this issue +Looking forward to the new patch +Committed to TRUNKI moved the binary file to srctestresourcesorgapachehadoophbaseeabdcfdafcdfacThanks for the great work Mikhail +I created a logo from the image Nigel mailed around a while ago +Ok I removed some bean fields for simplicity Ill bring them back and will also try in web console +Do a test with that new option enabled using a you notice any errors on or consumer +In taglib was at the same level as servlet etc +It sure would be nice to be able to specify under which transaction manager a bean should run +I just moved the snapshot information of download page to the developer section +is not guaranteed to work +FixedId be glad if you could update to the latest build and give it a again for your contributionCheersSergio B. +So if I get you correctly event using a timestamp later than the first timestamp should result in an undefined behavior? I did try step as given above wherein the second insert was with a timestamp much later than the first timestamp +We use groovy and the to add functionality like wizards and to our Eclipse RCP application +In Maven the id used in the server setting must match the id in the repository in order for it to know which credentials to pass +Please confirm they have been applied as expected +Integrated to Ted. +Thanks +Is anyone working or looking in to this? Or is there a simple quickfixalternate method +Reassigning to you +This is a core sorts of a patch +The data one pops up a dialog which lets the users pick fields to export along with a delimiter to a +You could also try the NMS openwire impl as Apollo supports open wire +For clarity the SLDs were inputted using the text box form rather than as stand alone files +We hide the model behind our interfaces anyway so it doesnt matter how ugly it is +Really the is choking on a bad attribute that is not recognized +A simple solution is to treat the high bit of the bit value just the same as in vByte and allow a vByte to follow in the bit case +Agreed it should be more informative at least Downloading examplewizard metadata +That should be fixed in the meantime already +txt +Thanks for review stackAdded logs in this patch +Fixed as per see screenshot there for confirmation. +HiThis issue was consistent in my setup when it was reported +Do you want to say that corruption happens during running of application +These pinned resources would accumulate overtime as the were recreated and then a workaround that would help in our case is to install a web module with a reference to the bean factory group that contains the bean factory so that even if the disappear the bean factory group does not +I am good with deprecating them just didnt want to break anything with the current patch +Makes sense +I dont think that we should loose time on tests to ensure that this parameter correctly handled by because even if doesnt work this will not be an issue on our side So this ticket can be simply verified by checking that agent was executed and was configured properly. +This should not be a sub task anymore +Daniel are you going to commit this +Please can you add an estimate for this issue +check in svn backports the fix from the trunk to the branch +to adding the licences +Reschedule if these happen to get fixed sooner +Moving the filtering via down to the candidate item selection phase where applicableCan you have a look Sean? Dont want to commit without your OK +Yes unbounded by defaultLet me look atAgree it is a band aid +Is that acceptable +Sorry I didnt include the file names in the comments for the attachments I expected them to be displayed anyway not just when the All tab is selected. +As explained when this was injected into another ticketWhat was this other ticket? +Perhaps Pallavi can at some point resubmit the patch + resolved issues for already released versions +Despite being marked fixed the return value is still not documented +Committed patch along with additional comments. +We can turn it off by default +Namit will take over the development +Its currently in the sandbox +ping for Dave to add any has never been a unit test for all the catch blocks that refer to in Spring AOP +These blockers and critical issues are resolved but not verified +A small change needs to be made to the mvn sample output for to remove the logj warnings as these are no longer being displayed +Does anybody know how to use this +tested with fresh release problem fixed +I verified it works fine for HADOOP as wellCan someone please check why hadoop jar prints the warning message on using the deprecated file system URI format even though is set to hdfshostport +Im tired today but I have taken a holiday on friday and going to check this completelyJust to know Do we want to open a reviewboard entry for that +Use GettersSetters where possible in FVH +differences for for for for a bit snowed at the moment fancy applying this one +Can you please attach a repository file containing the problem and explain any steps needed to replicate the problem +any thoughtsEdit finnally splitted in +Downgraded since both of the components are outside the core +BTW I know is a great lib on concurrent area +Here the workbench +That does sound scary +Please look at the graph attached +a tmp fix which slightly changed and +Reopened to assign version +Showing failures on Linux build +Good Point Incorporated thatThanks the patch does not apply +If there are no reviews and no users Im not really eager to try and get this committed +The script is documentedIm having trouble running due to a security exception signer information does not match signer information of other classes in the same package. +I am working on that +The sandbox should have most of Amitas updates merged and it has fixes for classnotfound and unrecognized available in trunk now +Committed. +Attaching new patches for trunk and branch +Verified on running cluster as well +Great work Pranav +Substituties previous file +Fixed in trunk r. +correct reference is the copyrighted version with a version typed in by hand from the jsf specification +Is it possible to move the checking of the file size down to Commons Validatorwith a fileSize validation? +updated in both and there no longer going to be a release +We have essentially optimized for what we believed to be the common case +As I know CVS client converts the sources to the target platform and developers tared them from nix +Thanks for the patch +Thanks +I assume this is still an issue? Bryan Call have you looked at this +I forgot to mention I would suggest that the appropriate behavior for this situation is that the tool should either refuse to map the LONG field as part of the PK and return an error or perhaps instead return a warning before proceeding +Any datapoint is helpfulParallel vs +The error you mention above was resolved in +Totally missed this ticketis this patch PR ready to be reviewed +It also depends on the Directory impl +For the Aslak notes or for mass operations where overwriting is desired N times its a great fit +New version uploaded which resolves also +Under Section we have warning that reads The compatible interface for the EC Query API and the S API are Work In Progress +Will merge to branch before release. +Do you agree Benoy +Log when reload events are inserted and how long it took to insert them +A simple application showing the problemThere are two schemas one including the other +This issue requires that patch in to be applied first +Not following +Ive just committed this. +googling says mvn false test +Everthing looks good +Hi RobertHaving only one mojo needs to deal with arguments according to the selected output type +I had forgotten compaction of multiple sstables before creating JSON data with sstablejson +We can try that approach as a followup +Added forms on overview screen +Only kind of +Verified with CDE +My only complaint is that get and get moved into Id prefer that View be exposed from so that those methods could stay where they were +I forgot to set the Affects Version its +upgraded to Thrift in ticket +Kannan Yes +GlytonIve tried your issue on most recent release and everything seems to be working finePlease do not hesitate to open this ticket in case that you will see issues onJarcec +the part about If we simply use the old schema directly Cassandra will give cell names and values autogenerated CQL names column column and so forth +No problem sometimes we simply need one to talk to +If yes it throws an exception +It needs to be a service running somewhere that any hive committer can access and administer +Will submit the patch for trunk soon +Sorry to be picky about this but we dont want Configuration to become a of application state +Looks like this is fixed now. +We should move to once that is available on maven +Duc Hi I am newbee to resteasy i didnt understood the workaround u mentioned in the earlier comment as i am facing the similar issueI tried specifying the dependency in my application menifest and it didnt worked for u help me out +Paul if you want to work on it feel free to reopen and implement on trunk with and all that +The test case in the patch is closely targeted at the method which takes all the time when you have large numbers of matching terms in a single document +Fixed with revision of buildersesb message Reverse revision and set includetrue when copying the quickstarts +Here is an illustration of how the class handles heap size settings when is appliedInput the value of jvmflags when passed to Output heap flags set when starting the Network Server via Old +I added it in the first place to avoid questions about where I got the magical number fromCommitted patch b to trunk with revision . +attached is a patch cutting over a lot more tests +Logged In YES useridTomcat default installation is not affected thats whyno release has been made +The main issue is it might be too broad when removing byproduct bundles which could result in more work being done than is actually necessary +ccl pom has wrong bundles are updated with good maven versions. +cancelling patch to address Toms comments +That also worked fine +is in core now so you can use that rather than a copyRemoved and global one referencedbq +Users can then cache the menu in whatever scope they want. +looks like trunk compilation was broken last time +bootstrap was run and prompt the user to +Im going to have Karthik to too +fixed in r +Although I dont think adding snapshot verification makes sense here since on some cases snapshot will fail and we do not want that to cause test failures +I will work on it and upload the patch soon +We only get when theyre deployed +Therefore I dont see much value in keeping this issue open unless there are others who have encountered the same problem +No good idea if I think about big feature sets +Should help w a test that has been failing pretty frequently of late +Great to see you have a look at this problem +Created +Then use a fallback loader or auxiliary local loader dependency for the lazy packages +It can be streamed back into a later if they want to access it again laterOr do we want to try to define an Entity representing a +I would like to see the new numbers to be convinced +Thanks Dhruba. +Something to investigate for implementation of this how exactly do apps like svn cache and store it +Attaching +New patch generated to include update to unit test +If no one opposes the change of the sentenceThese scripts are located in the DERBYINSTALL directory +Attached patch which addresses the issue the fix is in to make the comparator of group keys treat two keys with null value as different if they are from different inputs +bq +attached is a patch the synonyms and shingles tests still fail +Now the same image is added with addition to the image name every time +Cannot reproduce again probably some bugerror on my side +Thanks paull +The thing is its not that trivial and will affect how we process transactions so it requires some work +Should be fixed now +Yea the main intention was to have a fail fast behaviorToday I tend to change this though we have to think about first as I dont want to change it again later +It functions properly with no side effects +Not ideal for sure +I do feel like our unit tests should NOT get in the way of our delivering code but instead support it +changed fix versions cleaning up bucket +Use a dedicated server for formatting the job history +Committed to trunk and. +This is a candidate for SOA as the underlying issue has been closed and resolved already +I created my own based on md hash but im not ready to suggest it now +patch to fix the issue +It works correctly for two places +I propose that we resolve this because because a versioning scheme is part of. +assuming tests are few more +The site docs have become a bit out of date +Ill restate the workaround +Yes +You are welcome to grab the latest NMS and source and test it out +Component is in active state when only Controller service is available +fixed in for the thorough explanationI followed your steps to achieve different scanners and it works as expectedVerified in +i was just testing with submit and linksubmit to prove a point +Since its almost a year old Id prefer a new jira + +in other projects changes show is the locale used by svn? +As this undermines the of as multi tenancy solution +Please apply the attached patch +Awesome stuff Michael! This looks like it should speed up most phrase seaches +so or +ok + +Missed a minor change in where we should refer to instead of for the retry logs +Will need to be investigated +Closing issues. +all resolved issues due to a successful release +I have just done a complete clean build against Spring Roo master with an empty maven repo and it worked without problems +I cannot think of anything better and I clearly am not as creativeOthers what do you think +Maybe after zookeeper is released +DEVICETYPE PIPETYPE SOCKETTYPE and the like? Id guess that an FTP client can never make much use of these kinds of filesystem nodes anyways +Good catch Alex +you mean it should be on second left outer I agree it should be on the second outer join +Those options is named delete unseenIn Camel all resolved tickets from or older +so I have read some thing about it and learn first +I think it is ok to include for now +when i am analysing the i have verified the sql syntax it is not allowing functions for orderbysortby so i have added a function check validation +The findbug problem is not a issue as local variable is used in controlling +bulk close for the release +Like Ive mentioned above I need a good example to demonstrate the capability so Ill look at your link to see if it fits the need here +I dont see any relation to I however have an slightly different usecaseGiven a type someType with an attribute someAttribute where the someType is nillableIf the element of someType is nil and has an someAttribute value of xyz the attribute wont be retrievable as the the element being the parent is null +Sure thing +we dont have dependency changes littered all over the subversion logs +I dont know if the pdf version is widely used +bq +Patch for removing Cloudera repository +Now that work is well underway with AS all previous community releases are +The patch itself affets only the parts for there are also classes like involved which we cannot change or replace in a fine are supported by mapping the patches submitted are real patches +Need another couple of hours. +replacing with or somesuch +bq +Actually what was happening was that the French descriptions were coming up even on the English website +I have performed the runmessagecheck target with both CLASSPATH set and with CLASSPATH unset +usemvn pl am tomcatrun +Thanks +We could make that the standard if windows wants to handle converting that +Defer all open issues to +deploylib +Updated in revision . +add more logging if you think it is necessary +Also I have removed the configuration key conditional for creating inode map +Let me know if this assumption is unfounded +appliedThanks! +Andreas does that fix and +But the ambigous interface treated NULL returned by the parsing function as an error regardless of whether the local variable table was incorrect or just absent +Developers that have participated in developing this drop areNick GallardoRich ScheuerleJeff BarrettNikhil ThakerMike RheinheirmerBrent UlbrichtDan SedovLizet ErnandSamuel IsokpunwuPlease review the changes +Thanks James +Thanks for sharing your experienceI found a solution now I need to find time to implement it My plan is to finish the Ant testing script early next week +Feel free to reopen +Its up to the to enforce these +Added a patch to forum is for a different issue in do you know that will work for anything other than? We should get a fix driven into for a start and also see if we can bracket this test in our code with something that ensures it is only triggered when using a resource manager +Somewhat new system +This particular feature was introduced in of Maven Archiver and is undocumented in the reference section on the site +The uses the URL of the Properties file as the cache key +Improvement Use instead of in the Zookeeper and +add lets commit this as it you for the patch Andy +Chris Could you attach the class files from which you disassembled and? Also any other information that you can provide would beuseful +Feel free to go through the tests and remove the defaults and please verify whether it still works +Yeah Im about rpc these times so good having differing perspectives on this stuff +After the first installStart iOS version Printing list of backup infoLabel localStorage databaseshouldBackup NO shouldRestore NOOriginal varmobileApplicationsLibraryCaches Exists? NOBackup varmobileApplicationsDocumentsBackups Exists? NOLabel websql main databaseshouldBackup YES shouldRestore NOOriginal varmobileApplicationsLibraryCaches Exists? YESBackup varmobileApplicationsDocumentsBackups Exists? NOLabel websql databasesshouldBackup YES shouldRestore NOOriginal varmobileApplicationsLibraryCachesfile Exists? YESBackup varmobileApplicationsDocumentsBackups Exists? NOPrinting folder locations check for existencePath varmobileApplicationsLibraryCaches exists? YESPath varmobileApplicationsLibrary exists? NOChecking plist varmobileApplicationsLibraryPreferences for keysKey Value varmobileApplicationsLibraryCachesKey Value varmobileApplicationsLibraryCachesEnd After app updateStart iOS version Printing list of backup infoLabel localStorage databaseshouldBackup NO shouldRestore NOOriginal varmobileApplicationsLibraryCaches Exists? YESBackup varmobileApplicationsDocumentsBackups Exists? YESLabel websql main databaseshouldBackup NO shouldRestore NOOriginal varmobileApplicationsLibraryCaches Exists? YESBackup varmobileApplicationsDocumentsBackups Exists? YESLabel websql databasesshouldBackup NO shouldRestore NOOriginal varmobileApplicationsLibraryCachesfile Exists? YESBackup varmobileApplicationsDocumentsBackups Exists? YESPrinting folder locations check for existencePath varmobileApplicationsLibraryCaches exists? YESPath varmobileApplicationsLibrary exists? NOChecking plist varmobileApplicationsLibraryPreferences for keysKey Value varmobileApplicationsLibraryCachesKey Value varmobileApplicationsLibraryCachesEnd And the error message deny privatevarmobileApplications code exactly did you grab for the plugin when you said you applied this to +The issues are not really related though +We could lower to eg since FST uses so much less RAM than packed intsbytes but for now Im just leaving it at to be safeThe let FST builder pick the indexed terms turns out not to be very easy to do w this API +Unless someone objects the patch Ill go ahead and apply it and the test caselater on today +Core happens after some requests not immediately +I dont think that can be done so easily +current xmlrpc is not bundle we need to create wrap bundle for itBTW I will add a wiki page today +PMD is now at version +Bug converted to a of +Is there some other alternative +Ill be checking this out later today +field reference jpa class +I see that this is assigned to the backlog for fix version +Fixed in Rev +I can pull it and test it Please do so thanks +Fields are still switched +bq +So these lifecycle XML files would have default States and Stages for a workflow that a user can reuse for every workflow right? A user would need to only edit these files if heshe wants to add or delete statesstages or to change priorities and mappings +Doing something for this will be useful but lets not hold up What needs to be done for will now be handled over in testing on shows we are more than okay +Version has been releasedClosing this issue. +I havent tried with the latest X branch yet +Would you please help to try it +Incidentally there are a few other local variables that are never referenced and could be safely removed from this class +Yes needs to be resolved before +As Julien notes the patch simply catches the error +Try D +Merged the fix to the +Heres the patch file +This can be done after an initial release +Figure of Terabyte sort on Hadoop looks like an improved version of what we had originally +s remove a bit of magic on how lines are tested! +The following patch extends the test cases to demonstrate the root problem +I did not know those libraries were actually pulled out into gemsIts probably possible for us to move our version of the libs into gems but its something we have not previously attempted +Added anchors to certain lists +To clearly separate the work that has been completed as part of this issue from what has not I have created two new Jira reports related to running tests with EMMA to cover work that I think would be useful to do and +So if the user does not know about it then the CAS Editor does behave as before +Its just that I noticed this during final testing. +Thanks for catching this Hitesh +what should i do? fixing and commiting patch to svn? how about integration test rule +good point something to look for +It would only be used with help and the process exits after that completes so its probably not a problem to add the dependency but Im still leery of doing it +Cool +Missed modifications of and files in SVN rev Also added missing webappfragmentsadminpage directory and files in SVN rev . +didnt find any problem with this test +Very soon were working on it see and is commited in trunk +Committed with revision so we can see where the is thrown the next time it happens +Just to be clear I gather youre suggesting that I fix rrefsqlj and rrefsqljrevoke now rather than wait for the final decision on the behaviorIll get to these items pretty soon +from last GA is a bug in all minor and now with proper fix version +Creation of an symlink log shows up this way in the audit logugidhrubausersengineerssvnuseripcmdcreateSymlinksrcuserdhrubafoodst +The fix intended to make the situation not happen but does not fix it after it hashappened +Hopefully this patch is formatted correctly +Thanks Joy +Im fixing this one now +Looks like the libDir variable contains the value of the sharedLib attribute +In a remote container this is no problem since you would be executing in a different located in a different jvmSo while they will be technically contained in different scopes you still dont have control over which to use +And if were going to change GUIDS we might as well fix permalinks as desribed in. +This is more of a reporting issue than an actual but +OK closing this issue then +So currently this tool is meant to check the locality of files or regions to a given single host however it could be modified such that if the h option is supplied then the report looks only at the given host otherwise it will enumerate the live hosts in and return results for all +bq +Yes that looks like a good place to add the new functionality +Apparently fixed +I have the fix ready but due to lack of extra hosts I am waiting for hosts to be available by tomorrow after which Ill verify and commit my fix to master +Recently there were some test case failures +I cannot see any other Assert? occurrences inside of a block in the IJ test suite +The Generics policeman will visit them and will hopefully help fixing +I see the rest of your commentI still like iterating through the vector or matrix without a method +The solution suggested in should address also this issue +Am I missing something? Please comment +Setting JAVA HOME is covered in the appendix of this guide +I tried to use specific exceptions as much as possible and only resort to if there is no good match +Attached is the run of when I completed this testing +Patch available in consolidating patch +Revisit Jira assignments +File delete tests performs reseved name processing and only inode map deletion +The seems to always create a +If you have a better idea for solving this installation issue let me know +Attaching updated patch +attach my ve committed this +Fixed in r +Assigning to Vlado to get feedback about the Check for Updates +Please ask this on +And Im saying that if this class includes an observer method it should be treated as if it were injected somewhere +Can you create a separate diff with just the code changes for review +xml from a client +Yeah Id probably just improve the task tab so that it has a table for each type of task and then each table can have the columns and formatting it needs +Suresh I just posted the patch its in the description +Paradigm is still too different +resolving +Of course we need to continue enhancement of join order algorithms. +Attached patch file that covers both and which is query everything search +thats why they should be in the forum and not here. +Of course Valgrind or Purify wont ever find all the leaks but it better than nothing code reviews by our users are also very important so we appreciate your efforts +So far issue and have been resolved +Fixed +With Hadoop trunk these fail and the web UI is brought up on the default nodes +does it have a custom format for the plugin jar? some kind of a along with the jar +I think the test is actually working correctly and showing the bug +what about committing it? Shall I go ahead? too +jstplugins +that would mean the single resource is a remote tm something that we dont support directly ourselves at the moment +I suppose it could be done with new optional fields in the connection header and response instead but this would need to supportYes adding new fields should work +Added inputStream overload removed String path overload +The is used to pass parameters at query time to the collectorsAlso formalized the parameter passing to delegating collectors +bq +Also makes the changes suggested above +I cant find in this +is a release note for this issue +Patch is the one to apply +The downside is that this breaks mvn install true unless you can get the from some repository +Already fixed in version. +Aiming for +Ive tested master on Mac OS X and Ubuntu and they both have no problems withIm incline to mark it cant reproduce +Ill have a look some time this week. +But thats an implementation detail the approach is sound +Ive change the code of now it doesnt add aditional namespace to response xml +There is the same situation if user enters invalid information in input fields or doesnt select sex +Sounds like a better solution +Might seem like not that big a deal but here it something to consider +Tom +See and the backport of this for SOA at +I am planning to commit this today +my simple installation file used on ubuntu +reopen to attach a doc patch to match the change in doc patch for doc patch looks good to me to commit +That could be the problem I used svn Isnt the svn repo syncedupdated from the git repo +Test case to demonstrate the to be fixed for can fix the NPE but then I have another issueLets say we have the entry E and the alias A pointing on E +Hi alWhat is the status of this issue? Will it be integrated into a future Spring release? Yonatan +Moving out some bugs to please rearrange if you plan on fixing any of these in the next few weeks +ok will fix that +Which version of Derby are you using? I saw you mentioned that you were using version in +What is the expected values for bitfld +The one in distexample is a exampleIt sounds like we are in agreement +Good catch Thomas +solution proposed in is probably the best way there is no need to add the parent attribute if you have made your own stereotypes +But this method calls other methods and its not clear to me whether this would skip some essential activity +Please verify and close in the EAP CR on EAP CP and EAP CP. +Now there are only two warnings in eclipse. +java file headers as well as more dynamic generation of checksums in case that was necessary additionally now the classes that are compiled for the test jars are kept clear of the normal compilation happening around themso aside from the apparently largely orthogonal test behaviour that is basically the compilation of the test jarsBest for the patch Mark +Bulk close for +I understand that the patch changed behavior for instantiation +we can take advantage to to do the filtering for us +Ive just committed this to the HA branch. +Is there a workaround for this one since is not released yet. +Hey Andrew thanks for the ops perspectiveThe idea of embedding these logger daemons inside others is definitely something Im considering +Shanyu can you please rebase on latest trunk and upload the new patch +Another option could be work on the rest api +Right now M is the only milestone shown in JIRA +So Im for moving to a target +When it prints sleep sec stop server now before I wake up stop the network server +Patch applied +Note if you refresh the page with F editor will be rendered and all the reRendering after refresh will be successfull +For example in the file the text appears in IE asThis figure shows one connectionwith multiple execution of the samewhich uses the samestatement execution planSo for text that appears on one line in Notepad is truncated if it istoo long for the default size of the text box in IE +Yup I specifically did not wish to mess with because changing it would mean breaking everyones rememberMe cookies and because there might be other uses for as well +I wonder if we see major issues with current features today that does proper caching and works across jobs too. +This patch changes the cjava code to not send setwatches message if there are no pending watchesI dont see how to neg test this we are already verifying the positive cases so failure for hadoopqa should be expected but I think thats ok +Yes did request that and will update here as I get details +This has actually been fixed time ago when fid validation was introduced +Just a placeholder so I dont forget itBottom line people shouldnt remove core or other packages because this can break the build and we cannot guarantee it wont +Reopening to write release note issue now that release note text has been written +Committed revision in branch +Closing all bugs. +An alternative approach would be cache this information in the Mailbox +Must be a failure in leader election +Review and run unit tests please +It seems this issue has been solved in + +I still believe limiting the task launch time is helpful particularly in the case of hardware issues failing disks slow networks etc +I held off making those changes because it would be a big change for users upgrading from +Needed to put setIO permission back for codeclasses +The only small thing we lose is if a disk full is going to strike +looks good to me +Reproduced the errorThe config file wasnt being initialized +Since they all use the same wsdl this STRONGLY implies that the wsdl caching mechanism in isnt working on the IBM JDK +There is too like I mentioned in the description +Its not that I dont think this JIRA is important +Thanks for watching into this Craig! Really appreciate your work +Btw +Can you please post the result +extractionlib +Please keep in mind that many of these are new to me and right now I am a bit overwhelmed by that +That looks very usefulI have looked at the source a bit is there any good way to add debug information to +Gets the local business interfaces of the EJBThis is IMO a clear requirement that the transitive closure of local business interfaces must be returned if you feel this isnt clear I can update the Javadoc to make it more explicit +Testing the precommit build with these attachments +Just to make that stack trace clearer the first output is from the working test case it ran successfully the first time without problems +We just have to hope now that someone using Oracle will read this. + +The builder launch configuration could not be ProperiesBuilders have invalid external tool builder instead of the java this helps +last version was missing aIs there stuff worth keeping here? Doc improvement is really worthwhile but this has lingered for a looong time +needed to give the license +Hmm +Typo I think this shouldnt be a problem since +Just an FYI also see which we use for tests +Laura the doc changes look good to me +I will review again few initial comments +I think that we should make a name for the default status if we intend to keep it so that its a clear state +Patch committed to trunk. +bulk close of all resolved issues. +Go for it +Any input will be appreciated +I dont have time to review and comment on this right now but I just want to say that I dont want to see this committed as is +Closing issue +commit the fix for branch and the fix for branch in revision commit the fix for branch in revision +I certainly want to find time to apply you get time please file a zip of the code diffsnew files and point out which should be applied and which should would be great to create a MAVENUPLOAD request for the testng and link it to this +I have tested it with the same kernel and java as you used +What does deprecate mean? Will the API call be available in but marked as deprecated +If the question is can we do it more nicely? then Id say the best hope is Guice and the minimum effort required to use Guice is to rewrite every class that uses a RNG to take it as a constructor param and add Guice annotations and a config file and Guice dependency +Change applied in r +transactional +Also formatted the help message to be displayed in character lines +Today orhigher is recommended for RSA keys as fewer amount of bits isconsider insecure or to be insecure pretty soon +Let me know if you need this in +The fixed version is marked asThe affecteced version was marked as which is clearly wrongChanging the affected version to which is better though perhaps not fully accurate. +So it makes sense then for every segment with few terms in it all of the unique terms for the whole index get checked against it +bq +Awesome +Merged patch to doc branch at revision +If not already present then only it will copy +DavidThe current implementation uses as separator for outputs FIXNAMEVARNAME this means that the use of as part of the FIXNAME or VARNAME would break things if you allow AllenRegarding allowing other characters even if HDFS supports UTF Im not sure what will happen if you have and in the names of a file and you use that file as an input for a MR job +Shouldnt the aggregator wait until all messages arrive and then create one big message and pass it to the next action item? It looks like the aggregator sends different messages +Added wth pached pom +Closing issue as it was included in version which has been released +Thanks. +This problem has to do with enums and how the enum class uses generics that has itself as an upper bound +bq +This patch is not ready for commit +That may be the reason that the brokers logs are emptyThanksJohn +One question is the problem only with epsg or all the coordinate systemsThanksAshraf +Did want to do something different? +I am renaming this jira to be specific +ok not sure stil what this means this is generated by maven on the build server what settings can be changed for this and what does it mean? +Here are two simple archetypes +Attached the latest patch by including changes also +i think youre right that a different name is in order +In case using ajhtml instead of ajcommandLink sent successfully but isnt display in mail body +increased the waiting time to seconds since +Snjezana any ideas +To compile the code the is required against the ows is a summary of the changes +Hi allI encountered the same problem in Tomcat and JDK +for going ahead with the fix +Update the patch +It works after removing the +It happens on Tomcat tooIf not RF bug im sorry +SORBS is a bad BL to use its not maintained +There is no compression one row data only +Slava is this going to be dealt with in JBT or should we push it to the next release +There is a slight API change so this is going into the change is so little so we can backport to +Thank you for comment +Will it go into? If so when that happens we have to do this patch there +Today I have fixed some bugs of Hama PipesNo problem I will transfer the documentation to the wiki +I will mark this issue as wont fix +But +This version of the patch beefs up with test cases from the repro +committed the patch! +Would it be more prudent to ensure that copies of the that are needed are made in rather than copying everything wholesale up front +Please let me know when fixed so I can ensure Guvnors CEP support handles and +Weird as Im using it for months now could you please give some concrete details +ResolvedAlso tested on Windows Java Linux Java Committed in both Beta branch and trunk. +Sounds reasonable will take a look +Close to avoid running out of zookeeper connection +The attached demo project was created from your description but does not reproduce the s possible that some other module of the EAR erroneously depends on the seam JAR with typejar instead of typeejb +Patch for trunk +Thanks +If the cache was set up for Delta updates then closed adding documents would cause an NPE +Downgrading to trivial because this issue does not seem that important +Patch looks good other than what Enis said +if you uncomment the duplicate field in class B should print Woo it worked +What is the problem with that way if the jbpm runtime team will make sure it works +fixed r +Perhaps a specialized app then +Fixed in trunk since quite some time +Verified on CR. +Adding a Portlet to the JSF Project +Marking as closed. +The question remains though whose responsibility it is to get the subject capturing the info about the authenticated user or client identity? IMHO it is out of scope for the data provider otherwise where is the limit between what the runtime does and what the provider does? For your custom provider it may make sense for others could be an extra implementation issueI may be wrong of course +This patch uses a table of random number selected based on the size of the disk partition and selects the closest to the center of each bucket as the bucket owner +Do you + +This patch definitely improves the condition on trunk to prevent a from baling out prematurely +I could not get it to detect from netbeans then I deleted Ruby from platform and tried to via add platform +Is there any solution to this issue or is it already fixedwont fix +Ill give it a a is still a better idea because when using you need to specify the name of the default servlet which is an implementation detail +I set svn and svnkeywords for these files. +patch to include filename in io fix findbug updated to include latest file changes in the current patch swallows exceptions for which is not the desired behaviour +What is lightweight Python and supported by ASF? Something like Flask would be perfect and Im looking to avoid something heavier like Django or anything that rhymes with Qwisted +Tried with a and it gives me a null +Roman Im not sure of the release timeline +Fixed in r. +Useful when they are completely unneeded is a bad thing + +The patch also contains a simple fix for enable split by file in the load statement in this case pig should not try to split the input file by block size but process the entire file in a map +AS is compiling and running with JDK +This was committed some time ago +Set hadoop version to regarding the version bump is the latest stable build for the branch and addresses alot of bugs and issues + of contents of +The reliable technique is examining the instance metadata to see in which availability zones the nodes are +Ok heres a patch that changes the exceptions into logged warnings +Although at the very end the implementation of s Logger property returns null because I couldnt think of anything else to do +Thanks for notifying Jon +Ill work out the patch +Other packaging schemes can be added when folks contribute them +Adrian could you make the issue more descriptive +If the oncomplete is the only chance to do this right i think the best solution is that the table itself has an oncomplete attribute +Here is a patch with possible solutionIt cannot be applied to Wicket because it breaks the API of +added has been converted to use HTML elements in +Yes I have that option so I checked the box and no change +Im going to change the ticket accordingly +I think the fix for would address itself and this jira +fixed and added tests +When I first tried to run itit did not want to start even thoughI sud from my accountFrom what I understand su give me root priviligesbut does not make me the root userSo I tried again from the root user shelland as long as I set JAVAHOME it runs fineWhen the RPM installer gets done will run under the user apachedsso we may want to wait on this unless its really urgent for someoneCheers Ole +Is that ok +Digester has since been moved to its own project in the Jakarta Commons +Closing this issue as it will be resolved with the upcoming Atmosphere component upgrade. +fixed lapsus +Its now applied as a post query filter mechanism +If Welcome screen is maximized on start of Eclipse Central will be corrupted unresizable and squeezed into a tiny part of the editor space +Marking as cannot reproduce for now. +For the commons log the JDK level is already to and despite that animal sniffer is giving problem +I just want to provide a data point regarding the number of blocks to processiteration +latest ideacd tmpgit clone openshiftrepo cp rn openshiftrepo fullpathtouserproject copy but do not overwrite existing contentrm rf openshiftrepocd fullpathtouserprojectgit add agit commit m setup to openshiftMaybe the cp could be reduced to just be first level directories to capture +Added mention of H to the release note. +Attached the first cut patch +svn revision Added checks when the generated class files format is created to ensure the structure is not corrupted by writing truncated values +Issue was resolved several months ago so closing. +Log file with X e options +Fixed +It may result in a rewrite that splits one job producing both Q and Bt into several though +Terradata specifically comes to mind +When the handler is called I check if the element is the last element that the marshaller had created an attachment for +But it does not mean locking itself is not needed +I updated to get access to a field inside a mulipartformdata form so maybe to test character encoding problem +I does not seem to cause the sameregressions as the previous fix + +This is blocking our teams use of CXF as well +Rick +Yes ive got it +Looks like a clean implementation +Good points Eric I think we can use memcached for that as suggested in the Guava docs +Youre right +Fixed to use constants intread of raw strings in both and +more commentsEven we could let and do the right thing we could not guarantee and agrees on same topic ownership at the same time +Unfortunately we have chosen Axis for our production system and we get out of memory every now then +Attachment has been added with description Patch to Thanks! transition resolvedfixed bugs to closedfixed +will commit after tests pass +There are reported problems using with +or happy to if you want to do the config option in this one +adds a test +Thanks Viknes. +Reverted +I consider the patch complete by nowplease reviewAll regression tests passed +There are three reasons +For example if you set minincrmax to GBMBGB you never run into the unusable memory +This was fixed by Fil Maj as part of unified JS +Keith can you please look into as I remember a recent commit changing source sode in this Ill clean that up +Please share your concerns about this patch if you have any +Each client will have to do likewise and the broker can enforce the grouping as well to be safe +patch lgtm +ids will be to VL if unique clusters are there +I have opened RFPL issue for further investigations +Yea +This issue has been fixed and released as part of release +have attached a that maps to a Danish keyboard +I can certainly imagine this being very nice and cleanInheritance is going to be brittle over the long term so the cleaner we can make this now the easier it will be to attract a user base confident in investing in the new platform +I stopped as I was stuck with the tricky issue of data to keep between pages + +I dont think it needs a new test case +Cannot reproduce locally Windows Eclipse +Fixed with checkin . +Thanks Ari +Patch attached +bq + +Should probably name the metric more appropriatelyHow about memstore and memstore +It seems to me everything is working fine now waiting for feedback from also Windows platform users +Confluence is back +Ive reviewed the patch +I dont support configuration addition to the module is OK by me +Ive updated regression test to cover this issue + committed the patch. +Rather go small iterations here instead of a large patch +Jason could you give it a try +on havent run related tests yetI give tentative we either have commit the addendum or back out the already committed portionI dont want a half feature in an official release +I dont believe users will see this as the test plugins are not distributed +I would assume this to be fine since the server was started from cli so I wouldnt expect eclipse to stop it. +Correct +Here is a patch to fix this issue +The table only need columns +management behavior noticed in the also +Regarding I am undecided +Can this be closed now that we have a first release upcoming +merged. +There is on TODO in the that could probably be addressed +Running tests +If you click on any of the links on the staging site it takes you to the current license +lets bump the default for AWAITREADY to s +Now with the fix System are spinning up if S storage configuration is not part of zone creation wizard +With the update of the Maven Site plugin to the problem disappeared +I can do it if someone gives me committer rights +A couchconfig lookup takes only a microsecond or two +Could you make the patch with ? Thanks +I changed the doc +Will iron out and post a diff early next week +This error was seen on Fedora Core with version of running on the command line +This is probably counter to what most people would like to see printedIf you the byte with xFF first itll get converted to an unsigned integer +The file name remains etc however the content of xml file is actually for database that you select +addcommitter userianm projectboo some other futzing about with skel and permissions +has been marked as a duplicate of this bug +I forgot TLF wasnt transitioned. +Resolving this bug as wont fix since I cant do much about the underlying issue +The passed authentication result should be respected and utilized to access other services like Zookeeper +but every time I see that diff I panicI think we just need a release note that describes the new message that users might see explains that the message was not displaying properly previously and how to grant the permission if they want sysinfo to be able to read the classpathKathey +It turns out that the problem was actually that the project in question was configured without a source pathHowever we should probably add some sanity checks to the source path to detect that +can you attach a sample FO file so that we can run some performance measurements? also and even though you patch is simple a svn patch is always desirable +Also added a brief comment for each message so translators know what theyre used for +Hive uses Ivy for of its dependency resolution needs and I suspect that the other where we use were introduced by a developer who was more familiar with Maven than Ivy +Its not really testing anything meaningful bc of the platform dependencyHere is a patch that simply removes it +Ive just committed this to trunkHowever the patch doesnt apply cleanly to +Thats a better design in my opinion +I saw the same conditional there and didnt look more to verify +Its similar to +Attached the process definition which produced the +Why? Synchronization is need to ensure consistency of the object while its state being mutated from multiple concurrent threads +protectedprivate field but its setter is public there should be no injection tooAns No the property should be injected in this case +Some of the subtasks are done for M but the overall list wont be +And abort the patch so +ant clean ant +Attached is the correct patch +Can you post the +I was just trying to change as little as possible from the example + svn commit m add sample logj snippet that can be added to the Adding confTransmitting file dataCommitted revision +But if youd still prefer addremove I dont have a problem with it +Trigger the fault tolerant recovery logic when a connection exception occurs +Verified on EAP for RN text is fine. +The same is already possible with Architecture Rules which is is contrary to your stuff free +ok just verified that is already a bundle also with the stuff so we neednt recreate bundle for itBut need import jaxb api version so we need create bundle for and +Thanks Mayank. +As Igor mentioned you may it and provide a quikcstart or add the java code as well. +Ill sort out the disposition of this and the other two patches with Gary next week +This is not replicatable in replicate in +wow +We do consider supporting such annotations in Spring core though provided that we got enough time left +Instead of just being able to assign a key stroke you could assign a gesture that will trigger the javascript calkl for touch screens +Sounds like a good idea maybe we can support both ways of FrancisI am quite surprised that you still monitor this issue nice to see +The downside is on the theoretical side and for me its a hold the line type of position +Addressed Sijies comments +Things look hopeful for now that this is a workaroundStill sounds like there is a bug in repair or something wrt bootstrapping a node +This patch corrects the dependency tracking to look for in the generated tree rather than in the source treeTouches the following fileM +On the plus side I just verified that this fix works in AS . +already fixed in SVNthanksdims +Thanks +Since the total heartbeat load is proportional to the number of nodes scaling this way accomplishes itAmareshwari is a patch with proposed design +ClausThanks for clarifying that completely missed that changeRegards all resolved tickets from or older +MINA team has been busy with day job +Also looks like the icon on the name is wrong it could just be names shown are wrong any way for now we should show empty properties dialog +I made some changes to to prevent editor from crash in case of a failure of text editor +Guys Can we now close this please +Hi please find attached a patch that provides a workaround that makes pagination working with the testing data inThe patch also adds a sorting of results by productId and product and removes three import declarations that seem useless +Discussed this with Wilfried and we agreed to close this issue +tps +I can open a new bug and submit a patch that way if its preferred +Within the reactor you need some caching +Last submission failed because of the unrelated +What prevents this issue to follow this standard approach +So we are still working on it +Attaching the second +Select and input visions created +I am still emeritus +If a required privilege on one of the objects referenced by the view owner is revoked then the view cant exist anymore and will be droppedLOCK TABLE STATEMENT page has following lineTo lock a table you must either be the database owner or have authorization to the schema that the table uses + for the to trunk. +This will allow this patch to simplify its handling of maps to match avro maps since Pig maps now only allow strings as keys +This does not work for me when i take a fresh checkout from the trunk +bq +Agreed this seems solid and very complete +Ship it +Bug was in demo application +Submitted a patch to resolve the problem +In a column out of scan the performance improvement was big becomes with the new code +Hey Johannes thanks for the additional details +It seems like we should have for the generic counters for the job specific ones +Id suggest we checkin random for now then experiment thereafter +This behavior continues until a consumer connects and begins consuming messages +Good to go Todd +Im stopping progress temporarily +You could also set a breakpoint in s check if it still fails for you +Id like to see this one fixed as wellI have built a hacked local version with just the file modified to use the char lengths rather than char lengths which is easy to do and seems to work + page has been added as template +The top level all inclusive spring POM can not be used as modules in a pom are not treated as dependencies +Thanks +The package required for CentosRHELFedora isAll tests pass for me with this patch +Note I had to remove the CFLAGS setting in as that fails native compilation on Ruby. +which caused a conflict on my sideso +Thank you +I just created a normal context diffIve no idea what the format is supposed to be I just use diff and patch commands +Should I remove the fix version or do we already have some official backwards document available somewhere to resolve this +You can use Mockito to make a mock that returns the you want for the sake of the test if you need a hand with this you can catch me on hbase IRC Probably a good idea to put in a constant and maybe change it to be more obviously scoped ie something like When you upload a patch you should take the diff from the root of the SVN trunk so that it applies with patch p from inside trunkThanks +Patch committed to trunk +Because optionalconfigurations are an additional feature a developer must turn it on byspecifying optionaltrue +Committing +If there are no objections for the next couple days or so I would like to commit this patch +Weve got to stabilize +No functional change +Not sure what else you need +! for the patch +Hi Jon let me pick this up again I just read through this issue again and let me +Were trying out the experimental within. +Ill add the Store enum to the numeric fields as well then fix all tests to use it instead of +Youre right about that +case insensitive sort support trunk version + push to +Closing as wont fix for the reasons explained by Eric months ago. +Did I overlook anything there? I dont have a coverage test tool here +I plan to commit in a day or two +Latest patch by incorporating the offline comments from Amarrk +it also forces the web context of the war itself to the same derived name ignoring there right now any quick workaround? +I also changed a to a critical security fix for so moving to put a fix in trunk in rev David Blevins suggested putting the schema somewhere in pluginsopenejb so I put it in pluginsopenejbopenejbsrcmainresources +Ive fixed this in SVN +bat out of common to some etc folder +under IE I couldnt force footer locate at the bottom of the page as under IE FF Safari +Assume closed as resolved and released +Or they can set theirMETADATA replication to and be done with it +No additional information given so closing. +I think this patch makes great progress! I think it would be good to tease out a single deployment structure in the future but this works perfect for now +on the patch that is basically the same as trunk +Now we dont get the NPE +These files get copied to etchadoopconf in the dist step +Couple other stats that are also part of this are the number of reads and writes into soft cache based store +PR applied in master +Just found similar issue with job listeners that are required to be declared after steps +As long as there are no incompatible changes between and to as cac +Now that fileinstall support +Keeping the event model abstract would be bestPresumably if I would then be running Grails tests with other frameworks and assuming the same events are being published and assuming I implement the Eclipse viewer test listener based on those events then the Eclipse view would also simply work for these other test frameworks +Hi ColmAny thoughts on the latest patch that I providedSeumasHi SeumasSorry I havent had time yet to look at it as I wanted to finish the JSR port on trunk +It would be nice to mix the path based retrieval of the current extension with a mechanism for getting a more raw form of data out of the present this extension seems only useful if theres an serializable entity returned I have a need to test the JSON response in a more way and there seems to be no way to get that raw response body with this otherwise extension +Resolved using a specially named hidden field to detected unchecked checkboxes +Any comments or suggestions for improvements please let me know +was applied as expected +fixed +New patches for trunk for for +I dont know if this information is relevant but I hit the same problem with RF and I resolved it by updating the Flash Player to the latest available version +Good point. +I reran the testsuite excluding the classloader leak tests and using the default memory values +moving to +I appreciateyour commentsreviewThanks Mamta for your review +Were the failures actually related to my changeI ran through and everything passed +If the fix is already there thats great +This can also be a problem when the map task itself forks off some external script +Then we will be able to create a number of subtasks for rewrite rules and join enumeration algorithmsHowever the design of the interfaces may take a long time +Loenko +This is how it worksProcess in the cached process list is Active and Current state of the process is Retired and the state change requested for the process is Active then do not allow this operation + +The API will actually fail due to the default impl of handle thats a little odd +Closing issues for beta . +Its not a Mac specific thing as I first assumed +The fix makes sense as well as the workaround +Committed. +Committed. +We will do hibernate in a new branch whenif the time comes +I suspect the execute phase in the Mojo is responsible for this behaviour +Looks like the translators have fixed the issue +Ill try to have a look at these issues and get a release out next week +Thanks for starting this +Yep I can confirm the fix. +You cant go back and forth between these techniques +Just doing nothingId looked at Zimbra and it seems what they do +Heres a patch that actually passes the testsNote theres still a little oddity with the Snowball program that needs to be worked out thus I dont recommend running this patch in production yet +Ill fix this too +I am attaching the file to this jira +As you have showed in your patch this behavior can be customized by registering different converters +button in the Assigned Users section and add the user that to Javier to add a fix +To keep the old behavior I had to modify this was missing in patch aPatch ready from reviewThe test passed but since I made one small modification Im rerunning it now +Moving to +I tried looking for you in the jbossqa channel but you are not online at the moment +Can we resolve this as a dupe Roman +Would suggest we start a tool package soShould implement Tool and be run using + +Can you show us some sample code so that we can reproduce your problem? I amusing Lucene the same way and it works fine for me +This is actually a Bundle Repository issue not a framework issue +change in bugfix as +cut paste from the comment does not work well so I attach the with rv Thanks very muchJust had to count array vars cause I would like to support as many PHP as possible so I didnt count on return of false instead of empty arrayI also had to make test fit like the other testmethod you send by patchHowever thanks very much! +The attached patch solves the problem but I wouldnt mind another pair of eyes on it as Im quite green in C +Are there any plans to release hbase +All Nunit tests passDIGY +My restart was an improper shutdown of tomcat +The file defines POM version +Point me to an exampletest that shows what types of resources are failing and Ill look at it +Attached patch limits the memory used while initializing the hierarchy cache to MBInitializing the hierarchy cache of an index segment with two million nodes is now three times slower but uses five times less memory +Replaced all the assertTrues that I could find in the file +This looks good to me +Priyo and Konstantin thanks for taking care of this! +should be relatively easy if the toolbars implemented and vetoed processing of their children if the toolbar was not the one submitted. +Trying to get the setting from +Firefox renders HTML and it fails to render tables properly with an empty tbodySander In theory browsers can repeat thead and tfoot sections for long tbody sections +in hive has hardly any dependencies it only needs the +Resolving. +Hi Waleri why this issue still marked as in progress whereas code was committed +However I am quite busy at this time and can only do it in a few months or so +So all issues not directly assigned to an AS release are being closed +Thanks BJ for the patch +Thanks for giving feedback on the overview DanI was not aware of what you state about the cache in the embedded driver and this was useful information +Storing and Infinity is tracked by +Reopened to set the Fix Versions field to JDO beta. +I could not reproduce the problem specified +Scott how do you envision the export functionality working +And in a couple years everyone will be on SSD and we can make BCL the default +No I dont think different versions in the same cluster are supported Re the prefix this is the task of the guy who configures SPING by default prefix is null +change in bugfix release +Will the first version support users collectionindexactionQuery and a later version support users namespace collection index shard action Query? It will be useful to put together a table of which operations require which privileges along with how privileges are inherited +Its not ready to be comitted yet but since Ive been working on it for a couple of weeks I thought I would share where Ive got to so far in case anyone wants to commentThe new classes are which represents a persistent connection which manages persistent connections and which contains the details of an such as host name port etcStill to do is ensure that it works for Https connections write tests for persistent Https connections and write some JavadocAny comments are welcome +at is simply a property method defined in the interface which is implemented by all x values +Sure Ill close it +Can we get this fixed +Im sure that we will forget to specify triggers when creating new tables +Test cases included +This looks a result of not creating a directory in the webapp directory and unzipping the war thereDid you just drop the war file in the webapps directory maybe which is not supported as mentioned in the READMEIt looks you get stuck at the same point as in your other issue +I THINK the feature is used as a way of providing preference values but I am not sure of the advantage to using templated preferences +Should the new property go into the test conf instead +Punting to for now +Rather it just wait watch +You should not be putting anything for plugins in the core lib directory. +Can this existing mechanism be used forAt one point I thought the JDBC spec had a static method for returning the correct type from a like your code in was this removed or did I imagine it +Good catch on the bug Jenkins +The issue is still linked to an open issue in AXIS +Ive attached my implementation to this JIRA I didnt know how to create a patch but if someone has those details Ill do so +Otherwise every channel needs its own multicast port +If there is an alternative issue I would suggest a new test case to be attached here or a new Jira with an accurate description should be opened +Looks like the fix hasnt been applied into or it doesnt address the problem Im having +Thanks this works fine now! +Maven repository is the determining factor and I think is the last release for think this problem was fixed for a while ago by making srcRef transient +It still needs a little work but the refactoring is done and could use a review so we could coordinate +Will use separate issues to track any new problems +Thanks Knut +Ive merged the relevant changes to my working copy and theyll go up to shortly +Jake I think we can do a polymorphic base class +Ive added Christians patch with version Thanks to Christian for the thanks for your quick reaction! +If this bug is confirmed it clearly explains some of the problems we see in +Also try deleting your plugin cache +If someone wants to do that one they can open a new issue +It breaks any possibility of running in environments like applets or other environments that cant support code generation and dynamic classloading like mobile devices +no one else seems able to recreate this hudson constantly runs trunk and all tests pass +Attached is the test output +The problem was that the new filters add the last action mapping to the request this makes it possible to use Prepare and Execute as filters but the should always lookup the action mapping +bulk close of resolved issues. +Especially in SE there is no clear single way how it must be done but really depending on the use case +We were trying to use HDFS instead of DFS at least in new documents +done in for fixingSome comments Can access to the field be made method access? Exposing and accessing fields IMHO is bad practice +so we can let end user decide if its mandatory or Both simple and bean language now supports a OGNL like notationStill to do is the shorthanded notation for getter style also work with Header and Body annotations +We support both JSON and Binary protocols but they are incompatible +Patch that makes and protectedalso makes weights on like the make Boolean protected and note in the javadocs these can suddenly change approach best +Fix attached for company header version that uses the URL does this happen in? Would you mind pasting a screenshot +There were some code changes last week that broke it +You are right +If we still insist on checking this at startup we should do it for any admin tool +But i find id difficult to fix set a version of Junit in the plugin +Updated patch with other case handled +Yes this is a horrible assumption and one of many many things that have beenbegging to be refactored forever WRT and +Tom do you want a review on the latest patch? Do you expect more changes +I just dont see how it can possibly be done with bootstrap markupother that what I envision +This issue is fixed and released as part of release +Looks good +No response assuming ok. +Can you please test with snapshot versions and update here +Just checked it in. +committed revision in trunk +Changes tests in and trunk +Weve developed a solution that provides more linear performance +Change these two lines in and +go back to this page +I have the same problem see my test project in attachmentYou can see in the logs that it crashes because the lombok javadoc resources are not found. +Fixed in SVN +Anupam if you havnet started on this can I ask Hitesh to fix this as a part of? Thanks +Any more details you can provide will be much appreciated +Passing taskType as String should not break any +bug is in agent obsolete flag has to be initialized to false in case the function returns an error +And to show HTML as pretty printed all you have to do is click the world button to show it in a Browser +The datanode heartbeat seems to recover automatically +We ran extensively this week with our encryption hardware because we wanted to confirm that didnt cause an issue +Otherwise same patch +Thanks Jrgen +Hi DmitryWe can still see the type in trunk so Ive created a patch +This will require a new generation of image output streams that can switch from memory to disk buffering dynamically to work properly +This is a to solving this problem by having a module which is designed to be different in every distribution +I would prefer they use the apache repository for this +Brian and I can work on top of it. +Thanks for the patch +Yeah had thought abt that +To secure the console in the standalone version Copy the file toolsresources to serverdefaultdeploy +If not and its already stored in the job configuration do nothing +This was fixed as part of changed value was has been released closing issues. +if no one objects to the latest patch id like to commit in a day or two +Closed after release. +The cookie clearly carries Version attribute which implies it must comply with the requirements of RFC at the very least +At best this can be used to suggest better error reporting +Wrong JIRA +Fees will be calculated offline +I am rerunnign test againTed the seems to me to be a heuristic it can for sure tell if the key does not belong in the file +Syslog file for Windows couldnt be possibly be created on Windows +Thanks Mark! +Its funny to give new names to existing things and see the resulting confusion +duplicate of. +Previously the value A B C D E F would print as a b c d e f is a nice improvement +The fact that the rd coordinate is null is evidently not this discussion is sterile as I told you this should be fixed in current nightlies +Moving this out to release. +Hi BrianYou know if we use a to separate the two parts we need add a Palette item before the +we can still create an new enhancement issue if we have a brilliant idea +Patch to properly add comma after element inclusion in a mixed content text +They should be reused +Wrong default value for stub class includes in goal to bug in the meantime +Thanks. +As Jon suggests out on the list lets kick this issue and get this effort going again +Ive made a fix and am just running the tests now +Adding to identify missing CR issues +The main issue with the patch is that it only narrows the possibility of a deadlockUnfortunately yes +Updated the docs on the Similarity to indicate what offset and length are used for +Therefore no snapshots were there +It seems that reddeer libraries cannot be found for some reason +In that case I think well push this one for now +Applying the next patch I getThe XA Error isThe XA Error message isA resource manager error has occured in the transaction branchThe Oracle Error code isThe Oracle SQL Error code is cause +OK there are usecases where it is usable to run incontainer +Hi Howard how is this coming along +Committed as svn +For example the wiki plugin works however the does not +Closed due to lack of response from reporter. +String reader based on the value attribute string is created to parse this text +Hoss I had one major motivation in filing thisIf you use set to send a request to the handler named adminping the qt parameter also goes along for the ride +Changed version to CP fix will be verified when a new version of JON is available +Logging into an app is not sufficient +Then I think this is just an usability issue +Thanks for extended logs I thinks Jasper guys should better update their logs +Yes +Hopefully we get this in the DocsAlso the Registering a template Method is not commonly known to users and would require quite explicit documentation assuming no knowledge of the file structure and +Thanks all for your continued effortsBTW I also frequently get proxy errors when trying to edit wiki pages +In the future when we support arithmetic operations on tuples the output will be a tuple with a field schema having type and schema schema of the outputThanks in PPVSet patch +Duplicate of +The stack trace issue is still unresolved and there might be other comments +I am not quite sure what Exception you mention above +Remaining tasks +looks good to me. +Fixed +Rebased committed +which is referred to by has been dormant for more than half year +orig files from the merge +Changing short summary from adds copyright header in the form of Copyright Apache Software Foundation to Cleanup copyrights in the DITA source and generated docs +General Keys +if you need the missing parts of JSR in jcrspi and the SPI layer you could +Could you try with version just to verifyAre you on Java +The example command made thing that would also be nice +As we have seen with the lease recovery feature all these things can cause serious problems and should be tested in detailsSo as stated in a test plan should be both designed and implemented before such a big feature can be committedImo it should include a lot more unit testsBriefly looked at the code noticed that you forgot to increment the although the comment is there +this issue is not reproducible under linux x +Should we update commands manual doc as well +So I just exchanged the hostname with an IP address someone please review and commit to +That happens +It disables nearly all of the editor functionality syntax highlighting folding outlining hover help etc +Only one of these two patches should likely be approved though this second patch would be best approved in conjunction with my patch on +Implementation is already in trunk and the workaround was reverted +All tests passed except the known failures on trunk +dependancy librariesThe assembly file is currently including all of these as fileSets rather than including the module source and binary +Closed upon release of +New patch addressing the comments in +I just committed this +Proposed patch +if you still have this problem please open a new case with simple process xml that shows the regards tom. +From looking at the legal clearance page it sounds like the vote is simply accepting the donation as opposed to voting that the branch is ready to commit to trunk correctI did a Jira search and found no previous references to vote from Google search I got the impression it was more related to podlings rather than simple code module contributions +One way might be to simply choose each disk with probability relative to the amount of available space on it +I always thought that a JVM is a matter of a split second. +jBPM is DeadNo any one assigned for this issue for weeksit seems that the rumor is true +The error is because theres no XID on the Session +But looks like Lars mentioned that he had run the tests so that should be okI will try to fix the environment but dont wait on me +factoring out the CLASSPATH creation code in binhadoop into something like bin Sounds reasonable to me. +Alright good to hear that it works for youJuergen +Thanks for review Ted +rdfcat schemagen because a logger is created before the call to set the logging configuration to the cmd line setupThese fail there may be others rdfcat schemagen sparql update tdbqueryReproduce take an distribution remove run command with JENAROOT set to the directory. +Lets face a very simple fact we do not have enough people to maintain even with its current feature set +How about something like this patch +Since Ive added comments to modpython src code to mark the areas which may need attention will now give many false positivesThe attached patch contains a small modification which will skip any comment that starts with PYTHONeg PYTHON New uses Pyssizet for input parameters The addition of these comments will allow us to keep track of the code areas which will need attention for full python support +I have fixed our local issue the URL being used for the needed to be changed +It uses null instead of raw as the name of the codec +HDFS cluster is initialized when the very first namespace volume of the cluster is created +Its fixed for Fabric and ESB distros as well. +please help me i have the same problem and i saw that the bug is not solved yet i have the same trouble but with geoserver and postgis and i used the last nightly build and it is not working neither please to the jira a dump of your database the style and the wms request that makes it fail +basically i am not in favor of inventing additional repository exceptions within the scope of the SPIsee also for discussion on more detailed exceptions with jsr +Ok can you at least tell the name of your timezone and how it differs from GMT? date command should normally display the name of the timezone +I tried to test on Ubuntu but I get Exceptions when run Portal +Done Done +This has the same root cause as but it doesnt pick up the fix because it doesnt use to start the server +I cannot reproduce this +just committed this to trunk and +I just committed this +Not sure this is a blocker +If you get a number of exceptions from different servers now youd only see one +This implements option as described aboveI deviated slightly from what I initially described regarding the configuration parameters +you should know the environment you are deploying into and configure the application accordingly +Ive tested pretty carefully when testing the import on JIRA and this appears fixed +What I question is whether we want to support authentication and logging extensibility with such a universal function or whether it might be better to provide more specific extensibility mechanisms for these cases +For the purposes of this exercise extend it for implementation of JTS and wrote to him documentation +I cant reproduce the failure but if it happens again itll be good to have somewhere to track it +First off we need to reproduce this on a developers machine +looks fine to me +BTW now its safe to resolve this issue +I am working with the JEE Solution of SAP and they are currently not so far to support the requirements of the newer version of myFaces +It only works if the body of the message is a object +Thanks. +Each datanode in the pipeline has the same timeout of minute +I could never get to work +However Im wondering if this could lead to corner cases during KILLWAIT like those reported inIt looks like will report TATTEMPTKILLED after it succeeded but only for map tasks +It works for me though +Hibernate Code Generation Configurations refers too Hibernate Console Configuration +Dont bother with since were not talking about doing anotherrelease on that branch +If not what is the easier way to do it? Thank you. +This patch grabs the inet address when the connection is formed so that we dont get a null pointer exception when trying to print it out after the connection is closed +I have just checked Programmers Guides and +I believe the best solution would be to replace the with a VectorPlease suggest me if this one is the right approach +On Mac all works fine but i have reproduced this bug on fedora linux x +Could you please clarify which issue addressed thisCheck existing issues it should not be closed yet as its not done yet +I dont think the serialization capabilities exposed by Protos need to be exposed to users either +I fixed one of the broken unit tests + must set the system property true if they want equals method to compare only +Im not familiar with the development history of itRegardless even if its not its purpose I do think its a useful interface for users outside of the project to be able to extend +Committed +The index merger thread forces initialization of the before the index segments are replaced +OK Ill keep it with subclassing in mind then adding a corresponding comment to s default encodeUrl +Try upgrading to Surefire and see if the problem goes away +Maven and meclipse are superior at this point +Frans did you have time to integrate Jochens suggestions in a new patch +You would have to check in or earlier to see if this is a regression or just an issue with what elements the criteria builder is exposing +Meh +Separated properties patch from +Are you saying you have high cpu usage after start has completed and with no activity on the app server +And removing logj is not always feasible +Nothing indicates problem in our tests +Both of them generate N selects for the unidirectional case So performs better here +Looking at this issue again I noticed the title is incorrect +I will bring that topic up in the dev mailing list +i attached a polished version of my git sandbox code +Adrain in your very example there is no benefit sure +Open pending final edit by have made this correction. +Not in the registered users but in the system just installed using an administrator account login you please login to OM as admin go to AdministrationUsers select admin user change language reloginwill it solve the problemIll double check the installer +But we could add a specific comment +please provide a quickstart to replicate and the test case I noticed that the state method at tests the parseResponse property to avoid the described problem. +The root problem here is the bit solaris inI have invalidated that binary in jruby and will need to be rebuilt for bit sparc. +All Done +It seems to be working fine for meIm getting the New Context Entry menu item when I on the Root DSE and DIT nodesI tried with an old connection but also with a freshly created connection +This makes it possible for most analyzersfieldstokenizersetc to have no dependencies a few to need and a couple to need +change to change the file name altogetherc include this new file in the docs +There is another jira that is looking at making the history server generic that would help with this issue see perhaps this should be a documentation change at least for the time being +Attachment has been added with description Changes to use the new class in transition resolvedwontfix to closedwontfix +Might make more sense +No need to say that if somebody beats me on it I will not patch to latest trunk version due to changes in the that caused conflicts fixed to relative paths also did a little tidy of source in placesTesting wise it worked fine +Patch patch was and closed. +It will be nice to finalize the code and close this bug +Lets fix in +Well thus probably limit to client RPCs so that a given client can talk to multiple clusters that are not running identical versions of Hadoop +I also verified that this works with JT HA + +The conclusion was also that setting headers in makeRequest was undesirable and a security risk +RTC defect close of all resolved issues prior to release. +If not thats ok tooThanks Andrew +Ideally should utelize the when making the we need a dedicated resolved issues for already released versions +Mike is working on a support case where customer is streaming binary content and for some content the server seems to slow down a lot due to excessive memory not fully confirmed that this is the source of the issue +Im reopening the issue to address I didnt sync up this morning +annotation was rewritten injboss have been metadata declare others done +What is the scale that you expect this bayesian classifier to handle? How many training documents does it need +But very few files need to be replicated +Certainly an easy way to configure this currently as you say would be to pass the number of worker threads to be employed at construction time of theIve got some free time today so Ill try and tidy up the prototype and submit a like the idea with a +Its not correct +Updated patch fixed the parameter name typo and formattingThis patch doesnt define a type for the container signature +Based on the trunk codes cant reproduce this issue +The exact same problem occurs again in the release +Im going to fix that. +First glance seems related to may have been resolved by that. +If folks are OK with this direction Ill file those and address them there +I think including it in SP would be good +Very likely barking mad +This is why you have to open your own session and transaction on some transactional databases and not solution to this bug is to NOT use GORM to save the object and instead use a separate hibernate session and separate transaction with its own commit and close +Instead recursively process each Control +Closing as this works for me now with M. +Jeremy Ive tested the new snapshot and I cant reproduce the problem anymore +In general no application should really care for such details as they are deployment administration details and not developers details +yep +Yes it seems to be teenage slang so any of those things teenagers dont know whether its restricted to No em ratllis! or whether you can say for exampleLa mare mha ratllat tot el dia +I have to say that I disagree with the decision to WONTFIX +committed to and trunk +Incorporated Prasads comment +This is because STREAMINGSTATETRANSFER is missing +This issue does not have a technical solution +For the other case I think that the specific field names are not being carried through properly and get is not returning results for themIll add debugging output so we can see what is going on in more detail +The original test case would fail under sun jdk and pass under sun jdk +It uses from the JDK which we know works. +hi I finish the patch against head today but somehow it breaks one of the unit test case +Bulk closing all resolved issue that have not gotten any more feedback or comment in the last month +So Im reverting my change to that test +Added a couple of tests to test this as well. +Fixed +I have committed this +Please note this is still experimental code +Thanks DeepeshYour patch is in trunk rev +updated weblogic JPA example so that is set for servlet. +This is a duplicate of issue is no longer relevant for it is fixed inChanging the affected version to. +passed jenkins right before the commit +Attaching text file showing Bundle ids start command showing hibernate issue +bq +AS trunk now has +If that does not help I would then rebuild the indexes one more time with the cache empty +This jira has been closed as part of a review and archiving of the aj project +Its strange because it is very common scenario as stated +Ive committed this and merged to. +BTW I would close this for you but I dont seem to have permission. +Fixed this issue but found some other related issues that will be addressed in another JiraJim can you please check if it works nowThank it works just fine now +It produces and consumes the application mime type and it processes only Serializable objects +If a reference is active then the shouldnt be considered inactive since some Session is open and could be in use +The user would also need to add a dependency to the plugin so it gets added to the classpath +You think though this will add up to the seconds +Closed +From Maniks comments I get that components might be cleaned up so this is not an issue in Infinispan +I think that can be closed. +No code changes +Fixed debugQuery on a query with shards that returns results will NPEThere are still some issues with debugQuerytrue but its not critical since it is just debugging +What is Slive +So instead of being identified by their originally generated correct ids the new incorrect generated id for the topic was being inserted into the final generated HTML Book +see test +Err that should have been removed bindaddr attribute +Patch for trunk +Bulk close for +New patch addressing the comments +Orginally this was part ofIs this still happening in latest? +Later I restructured them but this was again just an exercise on the tool as I wanted to get familiar with git as a toolAll in all Im very satistied with it I did not use eclipse refactoring facilities or similar just git grep and a simple text editor +Super Derek +It is more a feature request in this respect +sure ive verifications this one and a classpath issue about EJB interfaces +Resolving this issue now that has been removed from derbyall +On second thought Im not as worried about performance since it hasnt been released yet and therefore has no established baseline +One possibility is that its related to the direct use of in the travel sample +Hi any action on this patch? All checkpoints green +Please I couldnt currentlytest it myself +All the reported violations you see in the screenshots when reported as are reported as violations again after next rebuildRegardsI have noticed that the flag disappears for the file after that file has been refactored and moved to another package +Im testing with and +One of users came across this bug while trying to upload a file with nameBy sniffing the traffic generated by IE filename is encoded with page charset as Dolf has kindly suggestedHowever IE does not send the charset after the namefieldname filenameC textplainSo to exactly mimic this behaviour appreciate if part charset is separated from the charsetMany thanks +Fixed the in Rev +Heres a first take from one of our Amentra employees +Verified by Dmitry +project +We might also deprecate the class and rewrite it under a new name for example. +The problem mentioned is not evident there +out of time not a serious issue but annoyance +When I checked with Gary that folder is supposed to always be bpelContent just as ESB has its consistent esbcontent folder +Added reference to forum has been fixed now that Adrian implemented a to be used by the. +Ive migrated all of my functionality touse the code and seen no seems to a good piece of work I encourage you to get this intoSpring it something I know that I want something in the Spring JIRAand something I know a few other JEE people for a fine piece of workDean +I like Dougs idea ofharportdirfileinharand we assume any directory ending with +Adding it here is overstepping +Sorry forgot to mention this +This is the test application developed to compare the performance between nSpring and +What does nodetool ring look like +Change applied to branches and. +It has nothing to do with scaffolding +I just committed this +Is JIRA a mind readerYes I would like to close the issue I already said that in my Oct postI would like to do this myself should I close or resolve the issueor perhaps first resolve and then close? I think I read somewhere the lifecycle of an issue but I cannot find it +Press down key +Patch can be applied at Martin! +Per but somehow under MC umbrella. +Sleeping with a lock held is never good and should be explained on a case by case basis +At theres around a repeat rate on reads +Looks good to me passes all internal and public tests +All patches updated with changes from review and all include tests +Agreement needed on API before doc can be finalised draft will cover API as it currently stands +I think I have fixed both issues but Im not sure +that apparently the bootstrap code tries to access a which effectively cant be in place at that time as usually the Txmanager works with the in turnI dont follow your comment? The exception occurs at creation not +Recreated patch after downloading fix for Kiran thanks for your work if this patch contains could we link the to issues its easier to folow up + +Ive attached an updated patch +You probably ran with java after my initial that disabled the test if Xalan was not available +Since only one adapter can be used at a time it doesnt make much sense to have both +Thx +I jut checked in a fix + code review +I am hopeful that I wont run into any roadblocks and this is my top priority right now +Message from the mailing list by Raj SainiI do not see this working for Credit Card payment +Thanks Tim. +Thanks Arun! +the loop body is unchanged except for removing the redundant foundCount variables. +To construct a response we run solr queries one for results one for facets +We have had many releases of since then and Rails for that matter +We will apply it on release planned for beginning of March +Thanks all. +Add More JMX info in replaces v +Anyway thanks for reporting this +If the MR AM scans staging directory followed by intermediate directory and the done directory before starting the new AMs jobs we should be good +I have implemented your patch and I am preparing a new release of the plugin to go along with Grails +Manually tested with provided sample project! Works well + +This could be added now as it wont break api compatability as at the moment there is no way to retrieval globals at all +If I edit the files with subclipse I cant use the command line anymore because it just gives a warning that a later version of SVN has been used +select a project and have it deployed? do you mean run it on server Right click on the project and select Run As Run on Server and deploy it using the popped up wizard +Thanks Todd +I guess making small limits would make users more aware of them but with the existing values most users can simply ignore them +See the linked task its already complete in the branch +AFAIK our latest test was run against the very first release +Simple check to see the ip validity and change to the broadcast ip +Pushed +As for Im not really sure what you mean +Of all the pending bug fixesfeatures this is the one that keeps my group from switching from RAD to +This is trappy! This should be changed +add attached into the sample and rebuild it +I would like to make sure we are in concurrence on the overall class and the methodsMaybe a better name would be JVM +Is that right? I thought some of the work that Thomas did on web services was also exposed via jmx +But maven wants to run them in the opposite order which doesnt work very ve also wanted to run the to generate my DDL followed by an antrun plugin to load some sample data but I cant be assured that the plugins will run in that order +Both new test cases failed without the bug fixesThis applies on top of and though I think only is necessary +Move issues to next release +This is because the object sealing was broken in Spidermonkey for Array objects +Please update it on the +I can take care of that one +The phrase really important was a poor choice on my part +at linked above. +If there are some more complex situations we could discuss on it then wdyt +Trying hudson +Just as some follow up the Java compiler converts all use of to with target is set to +at least for the time beingJuergen +Ported to +If the TODO is a blocker the jira needs to be marked as such +Not included in the patch +RichYou should have karma +The SVN revision is looks like the current version snapshot in SVN is though so the version in this issue is going to be incorrect +I use the maven ant plugin to do the copies +Not sure how to submit it for builds +The output would be cleaner without wrapping the multipartalternative portion of the email inside a multipartrelated email but it seems to render fine +I disagree there are too many source dirs already and the works perfectly +Unless the variable is declared as volatile rightI just checked the JDK implementation of didnt declare the size variable volatile +Its complicated +I think the problem is with Date and Integer when using indexed properties and leaving one of the indexed properties empty on the form +That sentence above should read So we need an elegant way of configuring register this flow in the registry with this id and these attributes +Either way we should address this remove the unnecessary sort and make sure to document that the results are unsorted +So I do not see how your example produces that result +This issue is resolved as the core implementation is committed and working +This is the packet that fails and since it doesnt require a response then none is sent back to the caller +I think I must not have been clear enough +Test case + +Attaching a patch which resolves the raised when TT is +You dont want to close y because its been registered and the will close it +Ill let you know how things go in an hour or so +website updated in revision +opening to add another patchwarn and convert the following providers +I didnt open it so I shouldnt close it. +Damn I would close this out from the fix to except the quartz and seampay examples use a custom XML namespace and the XML schema validator is not happy with that +Thanks +SergeyI just applied the patch into the trunk it could be great if you can provide an integration test to show the whole story +I didnt feel like dealing with any other so I checked all the possible ones for that line +We dont think this is a good application of filters +This fix results in a graceful failure rather than a stack overflow. +JSP onlyTo +This issue is fixed and released as part of release +Once its assigned to the model then you can select the model and right mouse click select Modeling Create Data Source to push your data connection changes to the server +Going once going twice Im going to add Overrides too as part of the change +I ran against this patch now that its committed and it works fine +It is done in the Theme +if there is a bug in type renaming youre probably fixing it in the wrong place? +It has been running for hours now +Until then it affects a small minority of applications +I am attaching a patch without the add change just for the sake of comparision +affects multiple release find the reason that mentioned in my previous comments recreate the patch +Thanks Alok Agnihori and others for providing the patch +metadatapluginsconfig into Embeddeds config directory to fix the issue +It used commas instead of periods +But I will see if I can find a toolthat shows theI didnt change that bit of the scriptI do not think the results of the script were checked in withoutchange before for it wasnt working right +So for you to test Build a war that doesnt contain any spring specific jars in its lib folder +I was wonder if any one is actually using the babe and since it is so close to release I will way for that instead +Things are a bit tricky on +Then when a update or generate job is submitted there would be available task slots on the task trackers for these jobs +Hi WernerIll make one last patch as you ask later this day I Xerces this is due to the tool I debugged and that relied onto a specific suite of components Castor Xerces J JDK SAAJ Axis some reason this tool uses to parse the result of a schema serialization made with imagine this is to provide Xerces with a schema that contains all definitions for a given schema the definitions ones defined into the importedAlex +Were now so this can be fixed directly +Sounds good +EugeneImplementing the same syntax in will be neither possible nor desirable as doesnt know anything about the Spring world + to trunk. +The patch depends onUbuntu smoke kernel +Will be back +Patch applied with revision +but OK. +Fixed most of the errors reported from QA automatic build javadoc findBugs audit +Fixed violations boolean complexity magic number unused parameter +Folks that are running something in through should really upgrade anyway due to the severity of some of the bugs never mind the security holes that have since been found +Im not clear what you mean by supporting FTP proxiesThe only references I can find to FTP proxies are to servers that act as FTP proxiesAFAIK these dont need any specific client changesNET is a client library so an FTP proxy server is out of scope +with me +I counted by nowI have or more jiras to add to thisThe changes are being done in small increments +Possibly +Marco sorry I just realized it was only the multicast part that was fixed I need to revisit it for the recipient list as well +Also have you considered the column alias is c not col we can possibly determine this from +note select of as should also work. +Were missing that datatype +I wasnt sure did the problem occur because of Groovy Eclipse plugin or because of JDT +Would someone pls try this patch? someone pls try this new patchThanks +I could start a discussion at the group easytesting but I dont think that would be the best place +Checked in trunk for release +Im not gonna do this myself because Im still convinced that we should be doing this differently I just dont know how exactly +I had hoped that wiki markup would work in the above comment instead the s show where the code was modified +A Service and Client with the rampart scene without rampart is correct +Fixed in latest +Ive experienced myself this problem three days can you confirm this is no longer an issue in trunk? I want to close this issue +Key initiated from the client is now handled by the server +Does it include valid checksums +My boss doesnt need promotional items feature +Is it possible to put this functionality in an optional target for now that would allow others to test it and add support for other vms before enabling it as the default +my mistake +just try to write tests when you write new of unclosed issues. +If my understanding is correct we should file a JIRA to add a test case to verify this +A different fix would be to iterate on codepoints instead of characters +I believe this jira should be considered invalid +For now I have modified Woden by commenting out the calls for testing +It throws an invoke +JIRA is done +So far I havent been able to reproduce this using the properties login module and a simple jsp page that prints the principals +I agree that we should use zero compression everywhere +I have reviewed this patch and I approve it. +Sounds good to me. +Restart is the most domain name +Ive updated the bug to be correct +patch committed. +jQuery wants the datatype attribute as dataType +The Xalan bug has fixed this thanksWe still need to update Cocoons Xalan +bothCommitted geotools acdaaebebbcCommitted geoserver close all issues in resolved state that have not received further comments in the last month +This one has been fixed today along with another issue with graphic transition all resolved issue that did not see any further comment in the last month to closed status +it was using type instead of bundle to test for supported project types. +I agree with this route +Attachment of the logo files +Includes first draft of the docs and a cleaner needs work on Synchronization handling crash recovery and automated test suite +So we might +At this stage we have not scheduled a release but I am marking it for so it does not get lost +See example in examplesquartz and doc in docreferenceenmodules. +I understood +Postpone until the metadata repository is fleshed out +architecture document +AgainI made some notes in the Jira run book on how I did this last time +Please go ahead and file a JIRA +We need to make the classes in the tools module available outside of the shaded jar +See do not see anything in that addresses this issue +Looking into this +fixed in alphamaster put branch fix into cloned jira +A lot of these problems may not exist if some of our currently disabled optimization phases were in particular SSA. +BPMN model validator is complaining about them I believe +SolutionUse INVSYNC and if a remote eviction fails with a data version mismatch on a remote node propagate the exception back to the caller fail any tx on the callers side +looks the same and the bug is still there +It may make sense to validate tck again just to be sure to be sure +not sure ask on that issue and estherbin should answer +Try and hack this patch into our TRUNK +If you need some assistance in testing a fix let me know and Id be happy to help to try and install worked +Heres a list of all findbug reports across the hadoop project in HTML and preserved in their own dirs archived +It would be nice if it worked automagically +Gary assigning to you +Wed like to move to hsqldb but people are seeing some deadlocks on This is not a supported AR it is just provided so people can bring te server up out of the box and perhaps have a working server for development in a desktop +Closing tickets +Attached test passes nowPatch applied at +Nice jobAll try this plugin as soon as possible +it is a different name +Setting to future pending a patch +Hardy will try and get it in contact him if you need to start the release before he is done +After reading a record the cursor must be moved to the end of record +Problem is in file logger. +CekiThanks for chiming in here +Issuing a second disable table should disable the remaining regionsFrom the API you can use or to determine a tables state +if you change the resourceType filter the pageIndex will be reset to automatically. +Fixed the typo +Hudson pulls the latest patch from the attachments and is pulling eugenes java patch +There is also a number of Blueprint Extender threads wonder if we need that many +Mike that looks great +The patch looks good +IvoOf course I forgot these tests would do this also +The is a child of the panelFor technical reasons Im not able to fix this behavior +we can do repairs of specific ranges now +this two file are put into root of project so we have a only resend the patch +Creating JIRA to make user profile consistant with Admin old resolved issues +I will claim mostly ignorance shell escaping issues and I just need convincingexplanation of this one thing before I apply the patchHow can we be sure that were not loosing side effects of the shell startup if we move from wrapping the process in a shell an just using runtime exec directly +Can you look at that to see if we can do something about the caching and if there are changes to the project or workspace +As mentioned in the previous comment the script could configure the XML appropriately but then when shared in cvs svn it wouldnt be portable for developers on different suppose the answer is to create two launch scripts one with no suffix and one with a +in prior version Railo just has returned false perhaps this has not throw a exception in transfer but it was wrong as well because this function never was working +Ping +If I can be of any help let me knowRegards project can now be build with all submodules +no problem there from our point of view the value is nullLeaving the data in the UI tree is kinda useless since if you navigate to a diff page it is gone +Resolved fixed! +There is no ia +In trunk and on close after is released +patch looks +Its been very inconvenient for users +I started a bit of work on an archetype for a webapp but need to do a lot of cleaning up on itFor a bundle as a starting point it might be sufficient to drop into one of the examples and run archetypeThe main hinderance I saw is that most inherit from a sling parent which you may or may not want more generally +initial version of Japanese resource bundle based on Revision +I will ping Wayne about this. +So in most cases meta edits will be interleaved with other edits in a deployment +I think Flavio is saying it works for himAlan any idea whats different btw the versions that are running fine and the one where it fails? Java VM version? CPU core count networking setup is there something that stands out +Yes makes sense +Err changes to reference tmp instead of +Theres a default set of providers a implementation needs to provide thus having to specify all of them would be tricky +I see +Hopefully we can begin to determine the best default solrconfig settings for NutchSolr integration. +Doesnt change any behavior. +Ill try to reproduce in our test environment tomorrow +What can prove it? What do you need for proof? It seems that under some conditions separated stateless invocations are working partially. +Verified for +Thank You very muchI will check the first case once again and makea sample page +Ok that makes sense +Issue happens because component in JSF has an implicit value +May someone check if this license would allow distribution with Forrest +I see where the hibernate does the work Im just not sure how to access that +Typical bootstrap problemMatthias Im curious whether this makes the problem you were having +classpath file and the file in a central location that is easy to edit +This behaviour was added deliberately +Its likecommit option B with +Should it be optional +Committed to trunk +Test Adrian I will take another look at this now +I reverted the workaround described in for compilation on EMT +Support added in and interface handling in support of callbacks adjusted in and has been in RESOLVED state for over one year so now closing if it turns out to not be fixed please. +This is because this demo depends on mechanism and I cannot quite figure out now how the two canThe bigbank demo will run successfully just that the authorization will not be functional +First cut of the feature there are some issues with new working on it will rectify the issues and will upload the new patch +so we are not extracting items from master slides +I upgraded the plugin to use the latest client jar +The demo Support of Custom Facelets components in the Visual Page Editor is created and commited to svn +Downgrading this issue to minor +I agree in general with this approach but lets take discussion of these issues onto the mailing lists +I tend to think that convenience for the end user is more important than convenience for the developer +Ill take care of making things work wo bringing over no worriesOtherwise looks good +Thanks patch applied +resolving for now file new issues if bugs are found +This will bring in the behavior where the channels recovery is manual? I suspect you are going to hit many more issues when you enable dual checkpoints and fixing that is going to beCheersHari Roshan Naik Fri May at PMWould it make sense for default config setting for the to be set to false then? Hari Shreedharan Fri May at PMRoshanNo that would break all config files from Flume and Flume +But in kind? No +snapshots are hard links making it configurable is a bad idea because I guarantee someone would configure it to be on a different volume and then stuff breaks +People can try to apply patches themselves. +I agree +Note that label and log are apparently misaligned +Thanks for the contributionCompleted At revision + in reopening to update release note info. +Dennis asks me to make the modularity better +Im on the v patch +Are there plans to turn optimized murmur hash into standalone library +Anyone feel like providing some scripts +Ive created a new patch in Eclipse and changed the extension from +I suspect someproblem with so +The problem is caused by which should use a correct way to rewire operators +Otherwise lets just close it as Not Fixed and move on +So it relies on ugly exception catching to deal with it +looks fine +As per IRC discussion +has been acting up recently and choosing an alternate plan is also typically memory relatedI think until the problem is understood we should be very careful about what we let in the trunkCan stressmulti be changed to use jvmFlags to reduce the heap to compare the and heap needed for this test +Committed at subversion revision +Deploying mvn jettyrun works fine but when I run it standalone it seems to load up but I get a when I try to access the admin pages +Closing this is fixed on both and trunk. +Thanks for looking in +Hey Gary I think this is just an error +That makes sense +Good job Todd! Lets just fix those test failures +Might be a bug in beta +In addition this adds an Issues section which is identical to the Issues section for +interceptors debugging profiling and scoped are removed between chain and modelDrivenThese interceptor should be applied only once +Hey JulienHow are we going to replace the Nutch webapp? Cheersm not sure I agree The Nutch webapp is just a set of web pages that let someone know that Search is working +I applied the addendum +Please keep this issue in and open a new one for possibly as subtask of +Committed +This problem isnt only with wstools deploying a service with a similar wsdl file also fails +Jonathan can u run mvn test and and post results + thanks +I took liberty to see whats happening and looks like in would need to check if its default portal +In your latest patch load already loads conditionally so if we simply dont clear the database in loadData then we should avoid having the leader loading it twice +has been added with description A testcast image has been added with description Second attempt at attaching thanks +There would need to be a change to Maven maybe a new Mojo Exception that would tell Maven that there was a failure but that its ok to continue building the module +I thought I had installed based on what I saw in my Eclipse config +Closing the issueThanks Jayaram! +The problem being that the java code serializes the NULL buffer as and the c code does it with serializing as and does not handle on deserialization or serialization +Can someone review it and if its ok patched it into once the code in Apache inbator stabilized +reports an error for this kind of code when there are two Process types found +Juraj please verify +I think it does not matter as long as it comes after The current format is tableName +If it does please let us know so we can close the Jira +So the dsw code is now independent of spring dm +Can you mock up a new WSDL that doesnt rely on a deployed service that I can use to see the differences between the namespaces +Ah I was curious about the millis vs micros +To be reviewed for the parent issue behaviour. +I hope that I have described the situation accurately + +It produced an even distribution and took about mins lesser +Write our own xml parser to read this and get rid of this dependency +Resolving as no further issues have been presented. +I just applied the Emmanuels suggestion see that seems better than throwing NPE +Just need to figure a way how to detect those subselects automatically via a unit testOf course the first best thing is to have it on the console already +The idea is that a client can only connect to a broker when it is started and has exclusive access to the storeSo any plugin will not be able to do much bc the broker cant do muchYou could init your plugin as a spring bean and have it created when the context is loaded but maybe a plugin is not the way to go +was intended to have an effect on the serialization order but our solution to this problem inadvertently also had an effect on things like constructors and service method parameter ordering +Then we can do another round of discussion API checking and modifying +andThis will be mostly refactoring work from and this translators +Right now I have no way to check if the current status is ok or not as I do not have enough info +I dont htink just symbolik links would solve the problem +The problem is a bug in your JAR file which does not have the codecs listed in its metadata +Suggest to keep this open for the time being and list it as candidate fordirk +Robin just see your response +Teiid wants the storedproc info to have a callable name that is the fully qualified form. +is talking about replacing our histogram implementation to make it more accurate +This is implemented in the Parser however in the Server the author info is defined as properties of a Widget rather than its own bean interface so the next step would be to change the Widget model in the Server to support this +Ill double check the status next time when I am taking an issuebut I am worried what will happen when two people push Assgin to mebutton around the same time whether the second person will get somewarning or error message and not getting the issue assignedWell part of the problem is that CXF is returning a partial response when it should be returning an empty one +thoughts pleaseDushshantha +Once is fixed make sure that whatever fix is checked in for works in release too +Upgraded Commons Codec dependecy to +Also rebase +I am at this point just cant see how this patch could cause the test failure +Seems some filter condition cannot push to the loader +Patch revision +JIRA cleanup +It may also be because the original request which I and edited just slightly is missing BBOX +That was caused by the wrong etchosts file +TODO If Remoting ends up propagating back IE or throwing a different Exception create a JIRA to investigate whetherthe should react differently to such situation +Sent the patch to fix this +The zip file does not contain libraries +I am marking it Blocker because it seems like such a likely user hit +add implementation note to spec noting difference between Java and C implementations of structs in current X implementations. +This looks like a classpath issue +Actually its ok if we keep the not left rounded guysmay be you could find interesting this patchIt fixes the rounded corner issueThe only think that does not work is that with IE the mouse hover only changes the right edge of the tabThis is due to one of the many IE bugs but I think the effect is still OK. +Oops this blocker on an exception is thrown +should be there but bash may not be +Okay now Im a bit confused +Is there a create that works even if node already existsWhat is this param str String to amend +Ive integrated your patch thanks a lot for your contribution +Im also experiencing this issue with maven with a weird on the base Spring test class which is instead declared with test scope in the a mvn clean saved the day +Im moving all bugs out to and we can move some bugs out to to get some release action going +Committed to branch +For a stable branch it is a good idea to keep the number of changes to a minimum +The field turns red but I never see a tooltip +I agree +Updated patch necessary after SVN move of all XML artifacts to the xml module. +I also dont understand why this patch was only committed toIt is because only has the UNCHECKEDREGIONLENGTH stuff as Colin mentioned earlier +Snjezana is this because we are missing a quickfix for fixing ESB runtime or something else +For example using a sample with for example tasks does not take up so much memory. +Did I your patch? Was there a mistake in your original patch? Have you tried testing the trunk? What exactly would you like me to do +Also wondering if its something I should put on every Managed Bean method or let the users use themselves +Any thoughts on how youd like me to proceed Phil +Got the patch and tried it on the test project attached to this issue +If there is any interest I can build it out more +What could be the possible solution to over come this situvationThank you +class seems to not appear in any recent WSTJST jar and why wed be deploying an Eclipse jar into an appserver is beyond me +The patch attached +I like the async plus proper communication to user that they have to poll for the status some sort of queued or async task list +Do you have a device attached? If not this is most definitely a duplicate +Can anyone else reproduce +I am also facing similar issue +is tricky because is apparently relied on the old exception behavior that I removed +Yes I have noticed this too but as I have stated several times this is not reproducable locally nor apparently on any other host than ubuntu +One needs to be added +I am ok updating this either with wont fix or not an issue +Ill review the proposed fix and commit it to +Its incomplete and setting up the context for the balancer is awkward but not too badI gave up on this tack though for now I was trying to repro because balancer is only one half of the balancing story +Jetty has been upgraded. +class should be named +retry thisI add the dependency of javadoc to use javacc to generate the java files first +Hi AnthonyI was starting to look at this issue +It looks like theres an initialization errorno runnable method error is there a way to exclude files from being run in the exec folder +Bug has been marked as a duplicate of this bug +Can you verify this on the EAP AMI that Chris provided +Good point on the documentation Buji is fairly new and was being vetted so it wasnt fully materialized yet +Commited the test with revision The commit comments are as followsWe do not have many tests for triggers on tables with LOB columns +Now it appears to be working after rerunning all the be available in next snapshot. +I think for the release we should just leave things as they are +Working as designed +Attachment has been added with description A unified patch against HEAD +There are known issues with the XML synchronization when adding and removing nodes by typing in the source editor +feedback bug servlet TOC is created on each chapter of guides +Yes the attached patch resolves the RPM build issues for me +But I guess that ROWNUMBER will be quite difficult to implement +How about file a Hadoop Jira to get and make this Jira depends on that one +actually leaks the idea of segments in different parts of the code and with another name +Could you leave open for a week whilst I replace and do a quick test that all is well then Id be happy closing the very much +The test must have been commented when we rewrote RM a long time backCan you please try uncommenting one or all of these tests and fixem to make them work and thus illustrate how to use this class? Tx +Closing to use sync events in preference +This is scheduled for beta what still needs doing +So a quick fix might be to add a retry or add a longer wait on the lock +Trevor if you have some time to look at it further it would be greatly appreciated +Im a bit concerned how the resetore got dropped but I did do a bunch of manual massaging of the APT markup and perhaps I just deleted it by accident +Happy New Year +The Android API wont let me increment a quota past what the user defines +All examples work correctly now +securitysrcmainorgjbosssecurityjndi securitysrcmainorgjbosssecurityjndi file data +Here is a screenshot of what the revisions should look like +Thanks for your commentI also think that this implementation might increase the complexitySo Im going to close this issue and open a new issue to resolve the asymmetric deserialization problem. +So your blanks in the interface will have a value of null or depending on if it is a primitive type or not +Being merged in in +Resolved all of these warnings except two on the query plan parser +Ill muster up some help with the implementation +Uploaded new Trunk patch on Oct +Tina Ive traced the code and find the GEPs shutdown method invokes the JMX remoting method to shutdown server finally! So Im not sure whether its the servers problem +I think that is the case here so a is in principle OK +They already are +in +The test case also makes sure there is not null string in theresponsePlease review +Please let me know if you would like the for the business that originally reported this bug +Attached is a new patch where I have addressed the comments made by ystein G +Then everything should work +Could you show me further information like a query and data volume +Applied fix to both trunk and branch +But it is and I can go on with the my cents on why it should not end the request +Adrian OK I will review thanks +Well I think I can add test for case of a +I was suprised too +A patch for to enable the new service wrapper functionality +bulk defer of bugsminor issues from to +on and issues being marked closed in conjunction with the release of. +Bisect says the winner is but I never encountered the corrupt sstable exception +The dictionaries in Apache are using XML tags to separate the tokens and there is also an UIMA integration for it. +A sane schema does not have null values in a map +So its scope is the one XSL transformation +it was a stupid hardcoded bit i forgot to change +The design started saying it is ZAB like then added paxos and since then has gone back and forth +Feel free to reopen if there is a need. +Ill try to incorporate those as wellThere are still remants of the old process in place for building the examples that i need to update before i check this in +But sometimes my transaction works and sometimes it throws Fetch Out of Sequence when EJB access Oracle DB due to which transaction rolls this is intermittent. +Hmm I fear I cannot influence DLTK to do something differently and I see yet nothing that the CAS Editor could do better +Lets drop this for and since. +i have checked this now with the current trunki also have seen spaces inside wordsexampledas T agging deutsc her T extcorp ora +I have tried this scenario in my cluster but not able to the same finally I just tried keeping debug points and has +updated the web +Thank you for checkingI have recreated the patch for the latest codebaseThis patch becomes a little big as I applied to sources according to How To Contribute +Ill have a look +Thanks for your prompt response! +we cannot change the behavior in but in trunk i changed the default to work like Id like to propose the class be extended to include all new HTML elements that are defined to allow a content +Huh? I dont get it +We could even propose to return a list of missing labels for one language etcGreat stuff +Replace ivory with falcon and incorporate review comments +Runping what version of Hadoop did you see this on +Bumping to so as to remove version +Having Form properties to override defaults could be useful but an intuitive default will be goodAccording to the spec the disabled attribute was meant to be used in the context of so catering for this use case make sense +Got the servlet test fixed +Ive committed this with the suggestions that Tom suggested. +Ill figure it out +but yours has more options i see +This seems to be a failed operation but I do not see any indication that data has been corrupted on disk +I agree +Sample added to Metamer +This is the result of different builds of Seam in EAP and EWP build +I am not sure if the patch fixes any issue +Thank you for your patience and contribution! +just some additional javadocs +Can anybody post international characters to form with secure web application on CR + +But unit tests are run against only the shared versionIll check in after receiving your feedback +Weld upgraded to in EAP so this should be fixed +The attached file contains hotfix for the issue it also enable jUDDI from attachment incorporated in revision of resource message Rename messaging datasource file +Additionally the installer can have additional non selectable labels below to indicate to the user what the type of server and versions instead of selectable Type radio buttons and Version list box I recommend adding the following labels +I seldom on my remote linux server java jar myProject is more convenient for trying out software locally excellent +After thinking this through I recommend closing this issue as wont fix +Im going to first try the approach I suggested in my previous comment if you still think it would be better to go back to the old approach I can go back to that way + This bug has been marked as a duplicate of +Rob +looks fine to me +Something else still seems to be leaking sockets +Sorry dont have docs to hand +Thanks Martina +Thanks Regis +Repository Tests Patch +The code changes are working and committed but we must have a unit test for this which is missing. +fieldName telephone type sizeMax add field number class +Please update once the doc generator can be invoked to host the HTML locally +This is being caused by your codehaus address not being on the allow list for your scm could set your codehaus address as allowable or you could wait for us to resolve it for ll look into it in the next day or two +Added a test case and a fix for that scenario +So what currently stable highlighter implementations would work correctly in the indicated use cases? +Patch applied +I copypasted +I tried the test case with and without my patch for +Definitely can do it when the ivyconf location is not set and it falls back to the default configurationMaartenI think Xavier already explained it I find all projects that are open and have an Ivy container and grab the being used to find out all the details about the module its configurations etc +Committed into cvs will be in the next drop +I know now to put all my responses directly into bugzilla +Sure +This name attribute is optional +That way we expose a clear API and we avoid the risk of incompatible types or of a class cast exception. +Thanks Denis +Most of my code was written generation so there was some significant benefit for the two propositions +Looking at the image it doesnt look like youre getting the string from toString anyway +Like Gianny I had problems using Safari +Im going to try to do a bit more to get the ratio at a more usable level +Therefore all userland code which depends on Lucy and by proxy on JSONXS will be Perl code +Nevertheless sollution should be similar to yours but in other class +ESB archive for server +Attaching the complete patch +I never said that this patch includes work that belongs to other issue +this is the wsdl generate +cryptography is involved in Metaxa or +They should be considered together +Builds under Java +Is there any plan to integrate this improvement into Struts with no need to apply the patch? Thank you +I will canvas the rest of the Mojo Maven developers for their opinions +Userguide updated for shell task +I was just wondering about the Learn More link and if it should be a button or a simple link? At the moment we have both +It was discussed on harmony that threading code has to be merged between classlib and VM +I have some responsability in this misunderstandingI was the one who quickly suggested to open a Jira issue from the dev ML +Lucas As previously stated there is no sufficient grounds which you have stated for which we would consider editing +Yes was released in Oct +But neither nor rpc callback functions will be affected at all +Is there any other documentation improvement you people can think of that we are missing? If not we can close this issue and we are ready to produce a release candidate for +Linked to where the changes for this went in +I cancelled the patch +Are there any problems using this method to preempt containers +In this ticket were talking about the dir for the configuration +Removed the dependency on the packages +So yes it is specifically import? Not to seem incredulous but there is almost no reason for that to occur +Thank you Gopal +it definitely wouldnt be something Id publish to a publicly available site +Why does immediatetrue not work? That will allow you to trigger a back action without applying the values +and integration tests relied on these methods as well as the MVC and JSF +Attaching backported fix for if needed please apply this +please to TRUNK patch for CR to be the patch now all parameter on since needs to be committed to BRANCH reopened sorry +Adding a more realistic graph generator based on this paper jurepubsThe input takes two argumentsK which represents the number of Kronecker powerM which represents the initial matrix in matlab notationfor example the argumentsK M will give a matrix of size with randomly generated edges with power law distributions and all the goodies +Sorry for the delay mate Ive been pretty busy at work I am having a look at it tonight +Ive tracked down the test failures to the fact that the class we were trying to to a delegation proxy for had no constructor +I would advise you to start a thread in dev to discuss this and get some input as for myself I think the whole idea goes against some of the Struts core patterns +See +Thanks committed. +Richard wait a bit on this one as we should take a bit care to not add to much stuff into our components +change applied r +Improvement now Note heading shown on pdf only when then select Internal Note to No when Add Note to a sales order +Attached patch for patch for. +My work covers a more generic system to add parameters and save settings in cookies +The SBM docs are only relevant to the AS domain model approach to the same issue wont use the SBM +In this particular case the client receives the same number back +Looks like this is mostly due to interface defined to handle access as well as it being +moving to B since we already have plenty of high prio items for to LATER since it is unlikely that well have the time to implement this in the near future there are many more items with higher priorities. +If the PMC votes to amend our policies and docs and the vote results in my being invalid by the new rules then it should certainly be considered invalid but at this point were just in a pissing match over our two different viewpoints neither of which has been solidified in any vote or any discussion which came to a clear consensus +So some support for paging is required +I am not sure why but a second try of the command line and has no errors and the script is sucessful +Ive committed the test changes and new master filesDan if you could please confirm that the tests are now passing for you and then close this issue Id be grateful +Completed all mods to get working with LDAP. +Maybe Im blind and it is being used somewhere but if it is Id expect it to be a bit more obvious +The pools using them need to be threadsafe but I dont immediately see why the config classes need to be threadsafe +A patch is s really important for m just notice maven plugins in maven repo now have the +this patch fixes the testDelete failure that people have been seeing +I have lost track of how many times we have needed to change and bounce the jobtracker in the past weekAdditionally HTTP LDAP etc have real proven HA solutions +Is this still an issue? Does it relate to the other Range requests bugs that we have either fixed or working on +Heres a demonstrating the issue +Reopening because the spec for this is failing on current trunk +to this approach +Thanks Ivan for the help on this one and thanks Viola for finding the issue +Done for the Seam Tools and Hibernate Tools guides +Either this is obsolete or will be addressed by . +Thanks Arun! +This is part of the work done for the effort +Fixed ages all issues that have been in resolved state for more than one month without further comments to closed status +Commmitted branch and trunk. +Patch with nits like the current patch only applies against +Thanks Ramya +trunk +Its up to s implementation whether it has to look into inner plans and in such cases helper methods are provided in the base classShravan From Alans sample code after you pop a walker back you will have w and currentWalker which hold inner plan state and output plan state respectively +The SOA doc has the invalid reference its correct in the project doc +I think we should hold off on this for now. +Currently the and tests in Capacity scheduler are broken because of Class not found exception of Commons cli +Here is the patch for Inventory Valuation ReportCurrently this patch works only for average costThanks to SumitRishi and AwdeshSurya Kusumakar +So it seemed like a good command to have. +Please if you find other issues +Navigation is now properly synchronized so its not a matter of making the request run in the portal context +See duplicate with workaround of setting password to no greater than characters +Could you please retest if this bug still exists +btw i think this was a good patch breakdown +dims +But I guess eventually the client will find out when closing or getting next block +Applied patch submitted by David Keaveny +This patch fixes this issue +Committed to trunk for CR +Even the Archive Manager on Ubuntu didnt like it +Please apply +not reproducible in r +I think the issue is over then +When you added the was the still in there +It looks like one of the numRows outputs is saying zero rows instead of some number before in an explain in each of these tests +bq +Assigning back to Permaine to determine how to proceed +I fixed the dependencies problem and everything works now so I upgraded Remoting in to again. +as r +Brian what is the status of this one? Should I wait until it is complete before releasing +The platform Im using downloaded as as default datasource JDK Windows VISTA with all fixesDo you think there is an issue with JDK? Im using the distribution for JDK because it is announced as more tested +Ive given Jonny commit rights on the code + +javalicenses +And yet another weird behaviour what about add? When that would be used to add an which always returns result from its post method only those beans which were not created yet would be affected for sure because for the others there could be set already +classpath files +And when not serializable discard the exception for the message and +I apologize for your comments. +This is a really annoying issue. +On restart of JBDS its fine +Thanks +For contribs imho the idea is much more valuable than actual implementation details +Thanks Enis + would provide a workaround for the metadata to be saved without doing a checkin +I was hoping to merge all that stuff back by now but have been waiting for it to bake a bit +This seems to be so that we can set the charset of the response to be +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Im going to do more work around class in +Committed revision +Trying to maintain this in synchronous form across nodes is overkill I think +I agree about the scroller but dont like the drag and drop just because it is too hard for users to discover +I want to know how about this issueI want to get it HA future inbut I want to know how to integration with can tell me how about Jobtrackers with Jgroup or others +OK thats sound good for me +I committed the changes +On the other hand nodes closer to the root will probably not change that often but read more frequently +Also the itself has undergone a bunch of changes sinceI cannot reproduce this with or any other versionIve attached a modified version of the javafirstjaxws demo that adds the logging interceptors on the server side and creates a K string on the client side and echos it +Would you pleas fix themThanks updated patch +Jochens patch is refactored so that subclasses can monitor events for purposes of illustration +Werner Ill see what I can do about the documentation update and possibly the Confluence page +We should consider having the SASL server always return success when authentication is complete +fixed with commit thing happens on trunkObviously the http servlet runtime scope is not sufficient we need scope +feeling a bit of a fool now +It flags if we hit the condition +Closing issue fixed in released version. +The problem is when you want to navigate in the results of this list for examplehttpslocalhostcontentcontrolWhen you click on the next the URL will be httpslocalhostcontentcontrolthenhttpslocalhostcontentcontrolAnd I totally agree when you said that is not a valid field name but it is what is written in the URLHope that can explain the patch +Attaching my presentation on Java in the Database given at USA in Austing +This jira is unassigned but contains a patch that needs to be reviewed for inclusion in +Looks like a bad classloader usage pattern in xalan +Theres quite a lot of code here and I dont want to commit it to the core so soon before the release +Neither on Windows nor on Linux +That is the goalGreat Im glad you agree +And indeed I dont see it in the source I have subfolders but only at that level and I am missing etc +java cp target text +Hi AnkitMy question is that why does come this type of error? is my local want reason behind this +Aries bundle is now available on Central +Im returning it to dev so they can look itThanks is a +If no one votes against to have this removed then here is a patch for review +As a workaround one may use a long property where the full value range is supported +Thanks Cheng. +svn wcstill gives some lines mentioning authors in the srcsite and srcjsp directoriesI dont know whether thats okThere is also some code in srctest using a query with an author field but thats no problem. + +And yes somethings are still hardcoded and somethings still dont work because its still in a state of fluctuation +Done on all three issues. +The correct location should be consoleWelcome since the page title is changed to Welcome +changing log from warn to debug +I have also increased some of the internal delays in some of the tests that might also help with this +I have the same problem as Ryan +Trunk revision +Package renaming is just whitewash. +though that particular gist is a little encumbered since its had so many authors and all of them would have to sign off on the fact that they are cool with an apache license + +Im planning to commit the patch tomorrow +Thanks Bobby +BrunoI removed links that are not in because there is no guarantee they will existLets stay focused on the task that is defined in this issue collapsing the masthead + push to +which I mute to in my configuration filesCheers stale resolved issues +So this excludes option above +Sorry I didnt think of doing this sooner +Manual tests were performed to verify the fix +also the command to run cpp tests via ant is ant Dcppunittrueand not as mentioned above +Then then the proper logs should be generated from the execution so that the history database can be filled with it +Everything seems to check out fine for me +Rolling restart is an essential building block +Postponing to TBDSquawk if there are any questions +Would be a timely opportunity to verify whether the Windows procedure is actually correct +Oozies not yet in the code source and theres no reason documentation readers would be familiar with that project +Hmm it looks like the best way to avoid that deadlock is to have the push out events to the instead of it pulling them +Attached patch which fixes this issue by adding an addAll method to which allows the same CF to be explicitly added to a mutation more than once +Fixed for speed +Problem is more related to +yes I dont see why not but it seems to me we have many implementation options all easy to try out but we need some good tests to make a conclusion out of it +Thanks for the review +Please close this out onceyoure satisfied everythingis complete +sure +It does not fail now after running mvn install on project root +With a client will still readwrite through thrift +this could be riskySeems like we should only do this when the expression has a single statement +patch I am betting on there generally being one date format in a given log so I am switching to the speedier version as Alan moved some javadoc down and now on parsing failures we just return which results in an empty string in out +Thanks StackJimmyRamTed for review +I willstart on it in within a couple of days if there is no push back +Fixed in the trunk under r and r. +varbinary in mysql seems to be more closer to a string than to a number +Ok I get that +Updated patch based on review feedback +Please use or as the repository +I agree with Sven that Wicket should provide just the baseThe details should stay in the applicationsThank you for your contribution +v adds a method in per Teds patch integrated into s reran with latest v patch on my box its OKRunning Tests run Failures Errors Skipped Time elapsed secResultsTests run Failures Errors Skipped +this same customer wants me to do this in which were in the process of doing +You can write ampuuml in your +So with some lag whatever you set for the replication will eventually be matched no matter the failures or random state of the cluster when the collection is created or its replication factor changed +Then when all packets will be received the stream will beWhat do you think +Adrian please upgrade the component to use in both the and branches +Good stuff Yes its probably wise to relocate to a subpackage to avoid version conflicts. +But what do you think will be able to support javaapp +Sureshbq +Please file another issue for another release that would contain the remaining problems then this can be closed +Ok thats good and bad news +Thank you Santhosh for the review! +This is the defined behavior of the +Per comments in integration tests are now failing because of this change +tccell is used instead of tcpanel inside gridlayout if you need spanXY or scrollbarsprior to this changes the tccell was rendered by this enables the use of and clientId in javascriptnow we need to put a panel inside the cell to get this back +the attached tif I was having problems with has data of type float +Can you give more details? Have you installed the snapshots in the local maven repository before attempting to install them +Passing it in would also allow a way to get rid of the hack for in the future +First of all the code is safer +Issue was a memory problem of my machineRegardsHerwig +Andy find me on IRC to discuss +We can synchronize all operations using an interceptor but this is too inefficient +Calling folks names probably wont help your agenda +This either needs to be completed in CR or shift to +I merged the patch to. +Everythings fineThank you Mark. +Why do we have in a completely different location than + unscheduled issues to the next release +bq +Shall I say this jira can be closed nowTrevor It looks like its in the trunk but not in the platform branch +doing another run of thanks Syed! +But fixing that issue is a separate concern from what this ticket tries to address +Hithere were a lot of java processes from past builds running on given slave consuming almost all memory +Yeah the problem is known unfortunately the author of the javascript component we use to do syntax hightled editing is not respondingWe might have to just give it up and use a plain text minor improvements to the toolbar +So you want us to delete the directory created in case of an error during import +I think we either need to put the schedulers into packages or put them into contribI would vote to commit this into contrib +Therefore I take over this issue again +Updated patch with new file +In that file is generated into the source tree +Hey Jay is this something you wanted to take on +For content you can supply the charset in the +As suggested above I recommend to upgrade to instead for the new circular reference support in +post method? Or is it sufficient to wait until they release the final version +extensions and bootLibs are optional +See for more details +There are two occurences of typeof fuctionis that intentional or should it be function instead? +the need for a stable release of in the central mavenivy repositories +inparticular i completely dropped the distinction between methods used for workspacevs session importin addition i modified the security related configurations that they can exposetheir implementation specific importer functionalitycurrently this code is not used in the which IMO needs some majorrefactoring in order to really provide support for the full JCR import functionalityin a next step i would verify if the initial draft was suitable to coveruser informationfeel free to adjust or modify it if there is something missing or not suitable +Thanks Guys syncing up with the latest code fixed the problem for me. +Hit tab in the last field and it will go to the connect button where you can hit enterIts an odd problem i only see it ocassionally +committed to trunk since its a straightforward patch +Additional Informationgoing to httplocalhosttesttest redirects back to httplocalhosttestalso if I add an id parameter to the redirect function in the index action it adds a id to the issue no longer exists with Grails +which dont exist anymore of course +Is there any way to inject an alternative Deserializer into the oauth rest template +A shell script has been included which compiles and runs the tester +The only real solution I can think of in order to avoid deadlocks is to add add each key of the intimidate cache in its own transaction. +This is a rather common problem and a good example of why one should avoid holding a lock when calling into the framework +This patch fixes two minor bugs of the first patch is already applied +Defer all unresolved issues from to +In the std webtools the header row and data rows are part of the same table so their width is automatically set to the largest cell in the columnUse the Firefox extension firebug to debug it along of course with your log filessupposed to meaning the extent of the artistic use that has been applied in the patch +Thanks Tom +messaggi tradotti messaggi non tradotti +Just so you know I wasnt reviewing the entire patch +Patch to trunk +Committed to x in r thanks GiladNow porting to trunk it is more involved than anticipated because of contribmodules differencesManaged to make the tests pass and the benchmark alg of choice to runHowever I noticed that in x that alg when indexing reuters added the entire collection that is docs while in trunk it only added about docs +Deferring until this requires an upstream Infinispan fix + +Added tests and pass after applied atop s patch +DSA +Here is a patch +For what it is worth if you think there are other file leaks this is the change I made locally to determine the allocation stack trace for +POM test exclusions have been removed so this issue is done +The idea of share all the libs on a multiple cocoon instalation is great +How about my attached patch? It just checks to see if the result was for java +Please follow my instructions to build daytrader branchIf you dont want to make geronimo tag you can modify daytrader root pom and change geronimoVersion property to that let you avoid build geronimo code and make daytrader build work directlyGood luck +However it breaks some behaviour on windows so another patch will be needed +Looks good +ok i will stop my tests then +It was the biggest challenge of my life but I managed in the end to get php in the list as well! +For example if the sentence said either If this is a base type this value is not present +As expected minutes later there is a sequence of calls to attributeRemoved one for each attribute in the session +RightThanks +patch looks goodI guess it was meant to be marked against trunk right? Currently it Affects +Reopening bc this caused massive failures in the TCK +wsdl i used for testing +Seems its solved in MRI trunk +Pushing packets too quickly? No exception handling? Not knowing Mina its hard to say what the problem might be +Please verify the property setting +Was the set in the case you noticedRegardless we could probably do something better +Attachment has been added with description Patch has been added with description Test transition resolvedfixed bugs to closedfixed +This approach will also work with tables partitioned by time +In any case there should be a timeout for Jackrabbit locks + +Opening again for some further improvements and simplifications +Postgis does provide a function that can be called to spot the sequence name bound to a column is db able to do the sameAs for what you propose is good only if we can retrieve the generated value back +Thanks to your test case. +You are correct that you can still query even after the exception happens however its a slowly degrades to being a vegetable +Oops +Committed revision +There are already two tickets for those great! +I must have looked at the +Ill leave this issue open in case you have more test cases coming alongThanksHiranya +will conflict with this patchIm happy to do the merge regardless of which is committed firstSince is a blocker my preference would be to commit this issue first then that one since it is not subject to todays deadlineI still need a committer to this +Bulk assign is not to be copied in my opinion its lax absent critical checks +Best regards looks fine +tests pass +This should be fixed after goes in +Closed with the release of Sling JSP Taglib +for the committed the patch to and trunkThank you Chris! +derbyall passes +classscalatoolsnscastTreesempty +My personal reaction If I werent an Apache Rat project member somehow enamored of the name rat and if I were interested in achieving commercial success for a famous software product Id pick something different as my name +My understanding is that the split goes to the region server when the region is closed so it is not served +Note that this bug is mislabelled but I cant edit it now that its been submitted +The problem is that multiple when loaded altogether in the same context doesnt appear to work. +all primitives would be the fastest and best on the GC +If you dont have versioning enabled you will get both versions in indeterminate order +Really the only exceptions I can see most userswanting are a bad addresss some sort of smtp error +commit someone asked if the function return value no longer has a purpose we should change it to void +Trying to wrap up this week but we should get as much as possible in +Weve allowed that in every major release and I dont remember a single instance of someone blowing his leg off +Committed revision . +Could not start a at the log file and noticed its using +I didnt find the last release hard to be honest +It would yield significant performance gains if I did implement them as many users were complaining about lack of pushdown for these functions which are often used in the WHERE the capabilities interface doesnt allow you to add support for functions in where clauses unless you support those functions everywhere +The iff means if and only if and the Windows lapsus calami has been fixed in the CVS repository +bq +Not sure which variant to prefer but TLS looks a bit more elegant +The view depends on the PUBLIC level privilege that user harry has on t since user harry does not have privileges on table t +So theres been a lot of code developed between then and nowThanks for any clarifications you can provide +Attached a patch and html zip file for these doc updatesSeveral questions When is it appropriate to have classpath in uppercase +in one callMakes senseMay be an additional thing to explore is bulk api on oozie side instead of making calls to each entity in the list +Changes the table name used for collecting intermediate statistics by removes an empty +I tested with suns implementation and Axis and was not able to reproduce the problem +the latest patch looks good to me +Are the styles two +DougYes a sandbox is still a Hadoop +Yura please report it to eclipse bugzilla. +i will open an jsr issue for that +I also see the command logging shows Loaded materialized view table from cached contents from another clustered node +moved to ve updated commonjst and jsf features and passing it to Gary to update psoa and savara +So in that regard I do not see a big advantage +Bug has been marked as a duplicate of this bug +Yes I agree +Will we break anything by changing xoauthsignaturepublickey to xoauthpublickey? If so can we just pass both for now without causing other issuesAdding opensocialappurl is a and is essential for simplifying the verification process +Attention to such details is highly appreciated +Has been fixed. +You mean we have to keep local machine dedicated to write all entries in it? or you are expecting just to choose local Bookies node mandatorily to ensemble if available in good state +My problem was more with all these small files like segmentsXXXX and or +Verified by Tony. +Adds a port field to andAlso fixes which has been failing silentlyThe Rumen changes are not tested +Committed revision +Tested on iHavent tested on x +Ive created a patxh which you might apply and test +Channel no longer maintains deliveries making this race conditions impossible to happen now +Probably what most people have been using anyway and I dont think its quite yet worth making configurable +thanksK +You may give the head revision a try +Note that fixing this results in coming up earlier hiding until that is addressed +Instead of reopening for closed issues with fix version we create new issue But could you please switch to mailing list to discuss your issue before? When it would be confirmed as a bug we can create new ticket. +Solr has given our application new functionality that we could only dream of before so this has been very good +We need the website to be updated +Thanks for the report m now going to port the patch to the trunk. +Im having trouble seeing how this is used +I merged it all in +I just figured I should report it in case its a sign of a bug +actionSo should be able to reproduce this in Servers View should be the only one that should exist other one needs to go away +Dupe of this issue was fixed +For embed mode usehttphostnamehiccembed +Not something we can fix without forking ADT +The on the other hand does not such an method +Thanks Alexey +Translated missing zh labels for party pos and product components +When doing a relation with and the parent on the child does not get set during deserialisation +Sorry but I am really not seeing the link between my proposal and your proposalCould you give more explanationThanks bit of code looks so complex and I think this can be improved a lot but just to get it fixed for this release I have taken the quick approach of fixing it. +Even though its stored by token it would be nice to immediately see the ip in the log without having to look it upIm also unsure about the reasoning behind the last patch +committed +yes +closing issues for released versions +batch transition resolvedfixed bugs to closedfixed +Reopening to attach release note +Jan can you verify this issue in the ER build +If the list doesnt exist do +No it wasnt there but the dependency wasnt there either +Thanks +Ill therefore apply yours and if nothing breaks commit and move on to and the rest of the chain +Patch applied to CVS +Note that named windows provide a subset of this feature as a new statement can be initialized from a named window simply by referring to the named fix version +Unit test passed and I committed the changes +Sorry for late +See my first comment. +Ive committed a workaround for now so this no longer is a BlockerIm therefore downgrading this to priority Major insteadAnd as we intend to release end of next week and I expect well need more time to fix this properly Im bumping the Fix Version to as well +Updated this to focus on the standalone use case testing +Potser podriem afegir a la entrada dretdret right law standing up estar dret to stand up dret to get thinkestar dret to be standing dret to get up to stand up +Ended up recreating my database instead of adding some salt +Since the reference counting is eliminated so who is going to delete the IDOM document that is used internally by the wrapper? How does the memory management work? made the changes +Attaching a test project that reproduces this issue +Interesting +I just committed this +Good point +If you feel this is good enuf please go ahead and commit it +This is coupled with the fact that the bundle must be a singleton inside of Eclipse means that we cannot support multiple levels unless there are some fundamental changesIt is theoretically possible to provide a very large abstraction that will allow multiple Groovy compilers to run while at the same time providing the proper compiler level for each project +Thanks +Added fix for debugger not adding at end of statements that needed it +Ill run it a couple more it looks like you havent updated your working copy since I committed the fix? For branchx you need at least r. +this was a case problem dont use AND use And +I have most of this done on a branch off + includes a potential fix for this defect +Ill make the correction +Thanks JD! +JanIm working on it I will upload a new patch shortly +Assigning to him to either close as dup or pull across the finish HBCK is handling this case now right? This issue can be closed +So extra steps need to be done to avoid duplicating data in a distributed env and therefore this needs to be done in a separate JiraJerome +I got confused while putting together the example +is available on both jira and phabricator +I fear there is something different in tomcat +Not really that diffrent than care to review +It looks like this patch contains unrelated changes make sure you rebase your development branch + +ThanksIn fact Im now afraid it wont work I dont know java HTTP coding ver well +Also the support for validating will have to be think over +With this patch the shell will use the new method to fetch the list of online controllers the functionality remains the same as it used to before +If the assignee thinks it can be finished before the release please note here and notify the team +It does make hashCode a little more efficient and equals a bitmore readable but should have no effect on lookups +CorrectAssigning to you +Lets see if I get this now +I like the new listI have one very minor question in the list we have twiceI wanted to check with you is this by accident or did you have some other spellings in mindIf it is by accident let me know I can just remove the duplicates before committing +it must be something with your server setup +Yes thanks for explicitly mentioning forgot to mention that point +There is a workaround available and unfortunately there is no way to resolve for all browser combinations +The issues with Eclipse artifacts still exists as you pointed in your comment and this is not related to PDE Maven Plugin +However I think we would like to know why a ping wouldbe affected when no disk access is expectedCan we discuss access to these instances vieKind regardsAlexandre +Open Session in View does work for the typical case issues are usually a misunderstanding of how it works or a known limitation encountered rather than an actual +Thanks +Please note that this diff is big mainly because I changed tabs to spaces and refactored some code into a function +I did a grep for and usage in mapreduce directory and only found to use it +Its not going to be easy to fix as the client client functionality have never worked together beforeColm +This was a sub task of and SOA which have been closed so Im closing this one now. +Is it enough to rely on an initial character as a selector? Could the selector be a column valueAlternatively would you expect texttable to do something as general as apply a regular expression to the line before parsing the columns +I didnt see any fields that had to be added to context classes that wouldnt be there if they were interfaces +This results in improved performance at the cost of slightly longer GUIDs. +Android support would be great but it might become another subproject within Apache Chemistry +owen taskmOwen sure this makes sense +Attached are not needed since its a simple standard file check +Attach second rev of a functional spec for this feature +reviewed in branch in repo then released issues. +I got the from the old Issue tracker +Reopening this one again because I now have two test cases that show a problem in DOOM that may be related to the present issuetask. +Fair enough +We dont even deal with isNull in +Sorting the underlying list would be done in the database so would be faster we really need this thingWhy should we have tags for sorting selectItems when we dont have such a feature for dataTable etcSeems like bloat +Since the Sonar +A small project should confirm that +Use this one instead +I believe this is done +You did not see them because you probably were still running with java but use java and youll see a different iteration orderAs far as I know there is quite a bit of them around the code base if you want to hunt them down and fix them feel free to +close in svn for b +A complete patch would include changes which build this file and the Derby logo into the documentation tree and the distributions in the appropriate places +Please try attached +Attached patch changes the defaults and removes these keys from the documentation +Daniel Please take a look on the file of the project +has been released closing issues. +update seems to work for me with h when use rather than issue should be fixedAll that is necessary is an option to force creation of the schemas if none are present +And thank you very much for helping me on the patch AmareshwariI think the quality of the patch has been improved a lot compare to the original one +Making it patch available +Bumping priority because at last check this affects of all devices +Openedthanks! I committed this to trunk and backported to nice iterations +Fixed code in this fileThanks for pointint it outNext time please send a svn diff this makes it much easier for me to handleThanks! +Alternatively what about our wiki +Its not completely obvious and Im still scratching my head on this. +Ive done some checking and I can run rpmsign on the three hudson nodes on which we build soavmg soavmg +Feel free to submit a patch for review but it doesnt sound like a requirement many users will have +The will require a special client to testconfirm correct operation and the needs to be properly designed +Maybe someone is messing up in the lab? Strangely I saw it locally too +In the meantime I will test the basic usages and the advanced training capability on top of the attached patches +A background thread that periodically does the sanity check would be nicerThe patch should also clean up the code that does the error handling +Can you describe how the files should look like? Can we simply copy the etc and etc from the root +Ive left internalCode as is for nowThanks for the bug report Philip +Sorry I have not see itThanks a lot +Am removing EAP from the Fix versions +a second patch that adds a method for figuring out the charset of the server response from the header +As you said if a user add its console plugin to geronimo he have to also manually maintain the +is making use of joining to A? Looks perfectly valid to me joining to a FK of the primary table of the statement and indeed works on all other RDBMS Ive usedWhere do I get hold of derby bearing in mind it isnt yet released? Nightly builds +My bad documentation claims only that schema tool can be run only once +no worries! Good to hear that it works for you now +That way we can clean up the package structure and put off having to shim a little longer +Im wondering if this is a problem of unintentionally shared Spring WS configuration across multiple endpoints +Automating running AS webservices testsuite +But when this is done the can not be clearedThats why the patch attached did comment out this the call to get does already check for null and creates one which after this can be cleared by anyone who does use the OptimizerAm i wrong +I can confirm this issue simply doesnt see any nvarchar fields on my SQL Server database tables +I see it is possible to use the PMD and configuration rule files instead of using sonar profiles definitions in order to get this goal +I know that by default uses case insensitive collation but also case sensitive collations are legalso please please no comments like You are using in a inappropriate way or similar +lock keystore availability successfullyScenario step +My thinking is that key constraints on the destination table should protect most cases of erroneous data by simply failing on a duplicate key +With your advice I resolved my problem already +I added a patch to make the variables I think are causing the problem intt. +Can the postmaster check why our emails bounced in the qmail logs +bulk close of resolved issues. +Two data points Disabling LOJ reordeable has no effect so it is probably not related to Disabling the call to transfor in preprocess makes the query see that in the table number for t in the predicate is However so is also the table number for t in the clause This is the immediate reason why the bad predicate is added by the search logic t is assumed to be the same as tt and t both have the same underlying base table +WSDL validation with Apache Woden +I find this confusing butknow you prefer this methodology +Sure I can help +Thanks +jar +Make this ticket as wont fix +The outline view is now improved and is nearly like Philip Ogren described +jar do gpg passphrase passphrase output armor i donefor i in +I cannot find any license notice on the pubsub extension so we cannot put it into SVN or distribute it until we clarify thatNevertheless we could be testing against it and leave it to the user to aquire the jars +Committed to trunk and branch +Scheduled for +This is a new failure after the branch you describe why this fixed the problem + for the patch +Same behavior is happening with modcluster module as well +the applied patch looks OK but im not able to close any issues even if im the reporter of the issue +This change has made Tomahawk incompatible with File Update now +Hope this answers the question +Didnt you intend to remove it and only keep +This issue has been fixed in trunk long back +This is information independent of schedulers +Hi AnthoninSounds like a path class resolving issue dont you think so? example now worksThere was some problems with the jar generation which caused problems to the panel detection +The upload worked about same as usual so this patch doesnt seem to change basic workings +Sample checkstyle +Now fixing it +As I mentioned I didnt have the problem at home so I cant debug or fix it there +I will add something to the javadoc. +Let us know plz if you face any other problem +Great this is something to continue working on +Lets make it actually compile and we can apply them much more I think this PR could be simplified by removing the explicit duplication of the base url +The XA error code is also printed out to verify that we get the expected exception +Brian who should be assigned to this +Added fix version see Make sure all fixes on branch are applied to branch +If production sites are using edismax this seems like a very critical issue +Probably we can win percent more however I suggest to pause this work and devote some more time to another important issues +Feel free to reopen if you can provide requested info. +We should at least document it in the java docssimon +Avner sounds good +Though the nice thing about the former is it just works already +With the latest CVS I cant reproduce this +Coloring works in now so the issue can be closed +Ill file a patch for thisI wonder if this started showing up due to some new speed up or slow down in starting a andor a perhaps introduced by or +No both the and allow iteration over in the repeater mode +I am currently using maven and and it is still failing to properly create the file in my repository +Makes sense to disable it now unitll we upgrade it part of and upgraded Guava to in the changes with revision + should be added to consolesrcmainjavaorgapachekarafshellconsole are now expanded in tab completion +Backported to docs branch at revision . +Changing in revision the problem was due to the attempt of stopping a fragment before the actual updateIve also refactored the code a little bit to make it cleaner and change the package refresh handling to a minimum +on the directionAgree with Dhruba that this should be explicitly enabled until we have more experienceNot sure we should mandate the serialization of task execution maybe we can use multiple slots if available on a particular TT? We probably should defer such optimizations after JT refactor is done though +Thanks for the patch Michael! +The dispatch clients now use most of the same infrastructure and thus things like security should work +So Ive left this as is +fix When I rename an Embeddable all referenced embedded attributes are not renamed + +patch looks good +Suresh seemed to think doing it on a branch would be to code sharingThere is a branch already created for +We can add another subtopic in the JDBC section on the interface +In this jira lets focus on the protection and we can have a separate jira to track the resource containment +its in the aspectcompile task that it blows ve attached a sample project that shows what Im faced you run mvn clean compile test you will see what blows up here in the test but thats because this is just a sample like what Im trying to do in my code in another I remove aspectj from running in the file this builds and runs just doesnt use the aspect +I cant explain wonder how the the improvements on the write path would cause this read impact +I dont see the need for initializing it to and then setting it to some other value based on that +I was fearful that any functional change would break the code so Ill toss up another variant of the original patch that keeps the two names separateWe really need this dns fix so I think well need to break the unified and proper handling of the dn hostnames to another jira +Fixed by +ChrisI think we will keep the and code as it is in JavaThere is no need to convert this code into +However I think that you also need a class library build for working with the class libraries if you are going to work on individual modules even if its one you have built yourself so Im not sure if it makes sense to remove point from the original +bq +Thanks the fix works for me. +Thanks Eric! +Please reopen when you supply more information +a scripting request handler is beyond the scope of this issue +Tests passed cleanly for me with this patch except for a ping heisenbug in the network environment +Closing released issues. +I have been playing around with strategies that do not require additional but have not been successful in constructing a key that would sort to the right spot if we got a result back that said that the region was splitoffline +As much as we would like to it is not feasible for us to purchase every single Android device on the market to test with therefore we currently try to pick devices that are representative of their manufacturers +I ran native build on Linux +Modifying a launch configuration is just a launch and is a local file whether youre launching a local or remote instance +I agree this is ugly if you have to have add a to the call +The fix applied here is a fallback +Tried this on a XS setup and not able to repro +It seems now it is bit too late for a Rampart +Agreed +fixed at r +Big thanksIs there somewhere in nightly builds of the SanselanIf no Ill get current version from SVN and compile selfBut ready package is better and easier +Is SLAVEDB a good name here? It seems to me that weare talking about the database at the master hereChanged it to must admit I do not have deep knowledge about Socket communicationin Java +The failover state tracker was not aware of some simulated remove commands that ocurr while the transport is disconnected +As soon as I uploaded that I realized that the source file under srcjava is actually Groovy source file not a Java source file +verified okVersion Build id to set fix version +All comments address +Axis does however instantiate the class without Spring +Maybe in future other parts of lucene may start using it +I have to believe this is related to if not the same thing since no other module should be able to have null wires +As mentioned on dev Im actually thinking of solving this by bringing and back to the main Jackrabbit trunk +Lets get it in +which fails on only +I have committed a fix by using as a hardcoded value so the bundle entry can be resolved +Also the for the hornetq default security manager uses plain textAlthough users can use other security managers who doesnt use plain text passwords we could still provide masking support for users choice +committed to BRANCH and on +First cut +The is a bit misleading since it returns the oid of the owner not the embeddable +I got the idea +Apprently this is sample data from a vendor +This doesnt seem to work in but does inCreate simple grails app add a domain class edit add two methods testA testB +BTW any idea of psuedo element selectors should followdocument element selectors as to case sensitivity orshould they be considered CSS and hence case insensitive? +I attached the same patch as before but with Unicode escaping +I need to use the admin panel +This probably needs to be tested on a newer behavior on EAP can you take a look at this +Fixed +Closing all resolved tickets from or older +We are not doing any validation checks for the vSwitch Labels in MS as these names are generated by vmware internally while adding a network adapter in web sphereThanks RegardsDamoder +bulk move of all unresolved issues from to +Later research by Graham showed that the problem is not Python butthat the script isnt a portable Bourne Shell script it seems torequire bash ATMIve modified the line in trunk but the real solution shouldbe a proper binsh script +I will try to provide a fix in the next few daysThanks for reporting this +Yea I had the same experience but wasnt sure if the body onload is still brokenFor what I was is that you also can execute a javascript using setTimeout with a timout of seconds +They worked for me the other day but not now but I was told on to submit the patch for review anyway +Sounds like is the root cause grab the latest nightly SNAPSHOT and give that a try +Patch is submitted for branch +The one small complaint is that syntax highlighting is not turned on by default you have to hit a button +Giri looks good to me +Shouldnt be a problem anyway +Or if using maven try the version instead +Marek Ive incorporated your comment +Closing issues. +I tried applying the patch locally but I still saw the same test failure +Operators performing recovery do not expect it would cause data lost +This latest fix didnt work either so Im reopening so we can look into this again +but some nits Dont set the class attributes to null +Sorry for the trouble. +I believe that has solved this +My question is Can the Interceptor class automatically get a reference to the without having to manually use a setter or use Spring to wire this together? It would be convienent to get a reference to the from within the Interceptor without having to wire this together at all +but the reported case is working now +Specifically having both and is a bit confusing because a casual user wont know which to import and why theyre different +I didnt come up a straight forward way to refactor to be a factory and make it clear +are already always logging the original URI so adding localizedPath shouldnt be a big deal +I reverted the fix in b and added a temporary hack to in ecbae +However these are relatively minor and could probably be checked manually every now and then +At least mappedBypainter does not reference a but a String +Actually Im using nightly snapshots of the cvs because firewall is blocking direct connections to the cvs +Could you do me a favor and test it with +Moving to later can look at if theres anything that can be done +bulk move of all unresolved issues from to +It was very easy +Erick did you look at the related? I already have a patch fixing the issue of SHF construction for the new xml style there +Program that shows the might throw an exception if connection couldnt succeed +I understand that the specified solution isnt implemented but but as far as I can see it isnt needed as auto as well as +By the way this of course passes Mikes testcase where he tests this with a dummy Analyzer that overrides the old method +Perhaps +The set method of the is used to change the default timeout associated with the current thread and not the global default +So if we include it by default we theoretically create conflicts for people that may be packaging their own libraries as part of a UDF or something +This was marked as closed in the first beta is there a unit test to check this +Can you commit it to as well? I will provide the rebased patch once is committed for branch +and +Committed to trunk +for moving all existing external fs client support to separate jars +CR becomes GA so for now no issues to be put there +order and move are replaced withontrols attributeontrols works on CR +I guess the Q is do we really need this extra abstractionAt the same time by adding extra abstraction with hadoop modules in this case turn into smth like and modules +The API of FOP trunk has changed since beta +Others feel free to reopen if you think there really should be something addedchanged in Pivot. +resubmit +But I dont understand the intricacies of logging to say if this is a good test or not Ill leave that to others +disagree its useful for people to know about once the RT bug is fixedFair enough +The old log files are purged when checkpoints are saved see the discussion on at least for the time being the Standby NN is deliberately only going to read finalized log segments +See patch in revision +The attached patch fixes a concurrency issue when running with the maven MNG branch +committed to trunk +Having done so i had to create VO versions of my business logic annoying indeed +Great thanksRodrigo +If you do not need Scala support I suggest the simple variationI will post the simple variation patch again after rebasing to the current head +The network sever part of this bug is fixedThe issue of the having null arguments has been moved to +We could say the type default to string if the user doesnt declare a typeNothing prevents us from eventually supporting other declared types for the key with string as the default +Dec have tried the following the problem is probably related to on linuxJust talking to Jirka apparently the this problem is version specific +rsync to is unreliable so not really practical to do this +Thats another story though +You provided +Changed license to the pull request into the new the into the master branch of the new repository +Increase the sevirity as it greatly impact our ability to use maven repository +push out to +Thanks Bryan +It seems is not documented anywhere and seems to be not useful +The IBM know how comes from more than year JEE experience on websphere products +If jstack is available on Z series machines it could be used see what the process is up to +I will test in release branch +Verified on CR +Something to consider if work goes on later +Its finished +Though there is some talk about making the next version of Groovy require Java which may end up coming out before Maven moves to Java +Attach a new patchIn the new patch I just remove code sending consume request when closing subscription +Fixes library issues for me in Eclipse although I still have to use the and test targets to make sure ivy brings in all the libraries which we should fix at some point +Attached is the component class +In the mean time other tests were written using first of all the exception test case in the framework Ive added a test case to cover just the last missing thing +In addition the coreWeb BVT is passing. +Let me know your thoughts +Could you please use the mailing list first to discuss the issue I think this is more a question then a Bug Report +Close all resolved issues for Engine release. +Adrian you are rightbut setting encode attribute to false for external links should solve this issue? +FLUSH forbids sending to all members +in +Supports Sequence File InputNote Please download and save it to examplessrctestresourcessvmdataset after patching the new patch +Here is the patch that fixes it for Oracle seeing all other databases in the have before the database name and oracle has +Thanks a lot Nathan +maybe we can use for this +Im also still experiencing this same behavior on just tried again and I cant reproduce it on +Properties file translated to ptBR +SDO project doesnt have a WSDL to Java mapping +Thanks Alex Im looking +It turns out that the version of IBM ships contains a private method called get +Moving this out to the next release. +We have a hard time to reproduce the issue I would suggest to check in the patch and see peoples feedback +Please Verify + +The present Maven API does not provide any context what requested artifact is needed for so while workspace resolution is working fine for needs other plugins including the fail when they get reference to an exploded classpath folder instead of a jar +What does your POM look like? Do you have ANT as a dependencyI see it working with all the plugin tests and with a number of projects at work without issueGarvin +As I am not experienced I am looking to get some feedback on whether I am headed in the right direction and if so will add tests and make sure everything else is completeThanks in advance looks like a good start + marks the start of server side replication related code changes +Bumping up the priority +Also fixes the offending test related issues +Any chance of this making it into? Ive attached a patch which I think captures the final decision on this bug +So we need to enable it in Default Field Configuration too +Hi MattI recently tried out and made a fix for this in the source +The builder uses code to build the requests so there is no need to have the WCS module around except that the requests will fail if its not +In fact the JSP code refers to an bean named which has a property named list which provides the collection for the options tag +This issue is fixed by. +That would give us flags instead of the current flags +That shows you how to define a family of table functions using annotations +This solves the problem +Just to clarify +I refactored PMS again and incorporated your idears Edward +The default setting is to run Mahout on a cluster +I understand and this change wont happen before the tools like the support this other waythis is how I see itproject nameutilgroupIdutilwar eclipse +ping +related not the same +I had mostly been looking at with Fix Versions and since this JIRA hadnt specified it I missed it out +I am not sure if that makes sense +Thanks for the commit access +Committed with revision +Regression tests passed cleanly for meMy plan going forward is to address Mikes concerns I will add a header tuple to SYSGHOST so that the expense of garbage collection is reduced to one page read in most situations +I have read some more about the different logging engines +Ive committed this and updated the site +A millisecond timestamp is too heavy to carry +Not retry on +Since Ivan was also involved in the discussion about this issue previously it would be nice if he could give it a look +arguments gsiftpscratchusjobsinput gsiftpscratchusjobsinput +This means that we cant read off the annotations on the target class in order to setup the transactionAlso as we dont have an instance we cant do the injection +The last commit on the version that I tested is there any known workaround to make remote calls to stateless session beans work with +I honestly do not think it is a bug +I was not asking why it is a good idea I was wondering why the existing functionality should be kept around +Close it now. +this patchadds set method to andWill appreciate if someone familiar with can review the changes +The underlying code was converting a byte array to a String using the default locale versus a +The following are block transition events +is interesting and I would like to discuss more about this on the mailing list +With Hadoop I simply dont know +I have not yet analyzed exactly which change hit the nail +It looks like the current behavior of Buildr is actually correct but there is no spec I could quickly find that checks it +Since Vladimir is too timid to commit this Ive pushed it in dd +Verified by Alexander. +One way this query is different from others is it has desc +I suspect its taking the entire data node and then reserving not by volume wise +I need to talk to my jobtracker to give it tasks and I need to talk to my namenode to retrieve the results of my jobsMy hadoop cluster lives on a different network than I do so I cant talk to it directly +The console handler is deleted from ER onwards +Thanks Vasily the patch was applied to BTI branch at r +Ten runs and the ASSERT has not hit +The current implementation treats this as if there were no window counting total errors. +In svn you need to svn add files +It doesnt matter what I enter as a param for commit when the finish method executes always equals trueUsing curl httplocalhostsolrindexerbookscommandcommitfalseThe commit is still occurring +From a quick look ignoring the event for Reduce tasks should be sufficient. +btw getting rid of the hardcoded parallel is a suggestion feel free to push back +We need more than a stack trace to work with +We now check to make sure if the name is set first we dont mess with it +on +For example in those checks are in the form of assertions +Reduced user acceptance may be the result +I dont want this to be a public interface at all evolving or otherwise +I know it is only workaround till this issue is fixed +Or maybe in case we dont discover we should check lnspr to add it to the cflags too +Close +Looks good to me +Any chance you could release an updated patch once is released +If it isnt fixed on trunk someone should this bug +turns out that the protected variables warning didnt come from findbugs but from checkstyleSo I only needed to tweak to our already existing checkstyle config in ourI also published a fresh OWB shipped in +Future enhancement would be to make the Lucene version configurable and avoid this frequenct Version updates in code +Look good +This is to ensure that generated classes from different do not clash with each other and with applications classes +Will fix it the test and update the patch +screenshot from google Tomahawk Team It will be great if you could tell me in which version I could found this improvement +Olga there was a missing change to to ensure the right import for +Ill issue a pull request. +Which renders the imported base table your conclusion is that user should import with spaces and fix the source model +We can have a comment such as We used to have a at each level delegating to parent +Ill post a patch for it in any case and we can discuss it then. +Im on it +It shouldnt matter that they arent pooled either as long as only a number of them gets all many people use Quartz in a JEE environment which even manages its own thread pool +Is there any update for this feature +The interceptor that can be used in non web the priority was set as major +add test for +cancelling patch for flavio to update the logging +Eric Woods Adam Clarke and myself have already started working on this +Harsh do you need a help on this? I can backport the patch into unless youre planning to do it +Building on top of is necessary for this patch. +Reopen for backport to to branch +Seems like an array splatting issue with the keys passed to the block +Ah +JVM are able to resolve such base packages at runtime for further traversal down the package hierarchy but do not resolve classpath roots in jar files unfortunately which is why specifying the base package makes a difference +Yes +passing to of +Moving to version +nevertheless I do agree that extending this patch to handle some other overflow situations is a good idea and I will look into thatfor testing purposes then it would be helpful to have a sample FO file that causes all kinds of overflows messages meant to be handled by the new patch +Oleg In your last comment you wrote about using the strategy pattern +Temporarily assign yourself and add a comment that you are working on it +Out of date +Plus we already have Features and that allow adding specific features and interceptors to the endpoint +Yes definitly copy paste errorPatch solves this +Submissions welcome +I agree the introduction of a special character simplifies the identification of a and allows to remove the handling from the +Thanks for reporting +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Please check the patch is applied as expected. +Feature is in master +Slip to I thought we were putting these in under a Source category via composite site so installable via updatesite +Verified in ER +It will be available next week on the CVS +I doubt that we will get to it in +Simply clicking away after making your selection is not quite good enough you need to watch the cell to see if it worked or not +DammitIve just realized that I forgot to add g to my CFLAGS CXXFLAGS and therefore this log has call stack and no optimizations but no local variable information for apr and +A similar feature worth bringing up in this contextId like the ability to replicate a selection of documents by id +Hi ToshiakiIve already start processing this ticket +The actual donation had to be split in two files to circumvent the maximum upload size of MB per file +The fix was to track in the thread what object its waiting on so that when an interrupt is needed it can be fired appropriately +found a bug in the fastpath handling +The implementation of is untouched + +Im surprised velocity became a core dependency but nonetheless I think it should be possible to use Solr in an embedded fashion without pulling in extraneous dependencies like velocity and others +definemethod is one of the few places Ive been able to see the effect +Attached patch a which brings the invalid host in accordance with RFC I dont expect that this will make the failure go away since that didnt happen forCommitted to trunk with revision +This should fix it +Fixing the propagation of the security context from on to the executing thread and clearing it at the end of the thread run +Thanks Ashish. +Issue resolved +Ill try to reorganize such that we avoid this dependency +as per previous commentDavid +Description of This patch add new test lang +the new patch is perfect +on patch +Help in resolution is appreciated. +From Maxs email on the dev listbtw +Can someone please confirm if the latest module in trunk resolves this issue +Fixed +However I doubt that implementing some form of further global optimization would be a good idea since uses constraints already make the algorithm slow and complicated enough +the patch for review +Normally we just run the EAP testsuite then apply the patch and run the testsuite again and check for regressions +directing for you to look at to verify if its fixed or plugin was added to manifest to fix this on last nightly build from trunk +Dominik I can set up a Hudson job to do an rsync from qaRHDSbuildsnightly to youserverpathtosync if you tell me how to connect and where to put files. +For example in arithmetic operations and as literals +Thanks for reporting Ill make sure a proper is thrown instead +Does implement Comparable +As an end user you should just have a way to configure said have you tried? +Also I attempted to merge in Patricks recent docbooksdoc changes +Duplicate issue duplicate issue +I wasnt aware of the precise functionality of JIRA there +Patch looks good except for the identified weirdness where we write to WAL after the update +Vertical space Horizontal space seems to be OK +Michael Drzal what can I do to help on this issueIf you want to close this one because there is it is OK for meAnyway if you need any informationoperation let me know +It seems to only break when this code is called from a Controller and not from a standard service defined in the applicationContext +Patch file +I took the patch a bit farther to store the headers in both a map and a list +We allow setting it all over our API but we only internally support stamps +bulk defer of issues to +The drop zones were defined by the operations team and included additional behavior like renaming the files to temp names while processing and then naming them to final names when processing was completed +I havent worked with snapshots before +My guess is that Ivy doesnt find the ivy file of your virtual module which means Ivy will assume a default module descriptor defining publication and no dependencies +would be added as environmentVariables +assigning back to doug for him to comment on this +Feel free to change the if this is wrong +should this be a hard vs +Will this be possible in coming up versions? +I think this should be a big priority for so we dont release the code again +I bet I didnt notice the http redirect when using a browser Good catch +The general solution to this problem is to not reuse your objects to begin with! Im nevertheless considering to make this work in Spring +added components and affected version +Ill even get to use some of the testng parallelizing features to test! +The key is sorted on row then column fam column qual then column vis then timestamp then delete flag +new patchallI upload new patch file fix the problems mentioned by Robin except testsI will upload tests later after I figure out how to write it +If its not too late could you please set my CVS user name to be fraz insteadThanks +Use legal dom instead of deprecated dom for xerces major version +Due to feedback from the community I have split the AS and AS projects and reopened all unscheduled AS issues that are a year or less old +Please take a look and raise JIRA for any issues you find +Otherwise our only alternative is to shutdown the VM +Patch incorporating Aruns offline review comments on Y! patch +CSS for Hyperic Web mutilation of the Hyperic Web UI through css of the web ui page with a new header +Nice! thanks for the warm comments +this analysis is correct its the same basic issue as Robert Ill update the name of this issue to address the replication slave case as addresses the NRT case +patch was committed +bq +See +Many thanks for contributing it Dmitry +The state of each of these helpers is not large until you start registering typesso i think there would only be significant cost if there were duplication in storage of metadata across typehelpersso the problem with multiple would be when loading XMLAlso please let know if the issue mentioned in the JIRA still a requirement +Patch with Olgas comments all the changes had left out the statement to set the found matching spec as the new func spec inOne extraneous change is in srcorgapachepigpen which has an unused import which was causing an error in eclipse +Resolving issue now that this has been committed. +This issue has already been committed to the CP branch. +too +Even though the fix version is set to it isnt merged into at all +Depends on the availability of a depends on EJB support +then add symencsigencsig folder and content to svn +Adding link to related issue +Thank you for everybodyI am very glad for this excellent library to become furthermore good. +Max please review the patch +Its not a change +One line addition to the docs making note of the SQL Injection issue +Thanks for pointing this out Ill wrap it up shortly +There is a new snapshot with the fixes referenced in +Waiting on input from people who needed this so well leave the version in place and address a native option for post +About many child nodes if the hash or id property is different and the node has millions of child nodes then all of those nodes would need to be read with approach right +I am pretty sure this is no syntax but a classloader problem +Are you training with cbayes option +So it is not intuitive to look for a table functionWe might also find out that a spreadsheet table is not the same as a word processor table or a presentation table +Xiaowei could you retry this with applied? + +for the patch +I even replicate when the system was overloaded but cant reproduce this issue +Fix looks fine +is on purpose to allow downstream projects to continue using the same API for testing for all their legacy testcases with noneminimal disruptionFor new stuff folks should use viaIm planning to commit this now are we OK +These blockers and critical issues are resolved but not verified +Can you avoid this somehow? This is just an extra variable which does not help with identifying the CXF issue if any + +If it is mentioned in the refguide refguide should be amended +it is importing when it should be +to the approach +buildivylibcommongriddevbin +It should be okay as several other UDFs have a member result variable to reduce object creation see for example +Too late for for handling +I am marking this as a blocker for release +Added additional javadoc comments to mergeZipsLet me know if you think this needs any more refinement otherwise Id like to commit it soon +Fix is commited in trunk +If tests pass will commit since we need to burn this puppy in before release +Reopen if necessary +I guess it was too late and I thought FOP would be a test case +I dont get the error if I create input from versions of and using the older Hadoop API +tar contains a simple test case +This patch fixes for this case +Im going to commit this momentarily +As it is the extension sample would compile but at runtime the user would get a +Please review +It can be quite difficult to read such a patch but Ill give it a try +Thanks Carl! +Patch that uses Guavas +This issue is fixed and released as part of release +didnt have an issue with extracting the text +The sample output file for queue +Apply patch as normalUnzip into FORRESTHOMEmainwebappskins +Committed to. +Also getting problems with jquery +Thanks Tomas! Great patch! Ive committed it. +Please checkThank you so much for doing the review +This will still lead to accumulation of conf files in the logdir of all those jobs which were completed but not yet retired when the JT was restarted +Thanks for the fix Jon. +attaching current patch +This will make it work with bytearrays but not +Necessary changes to a couple of unit tests +Doesnt it mean seconds pass before we actually go do somethingWouldnt suggestion of looking into zk and using number found there be better for +Thanks +No longer reproduces on trunk for me +This patch doesnt apply cleanly anymore. +Patch for +Then it would wait forA thought would be If the parser would be able to just read the whole thing and scan for before newlines and then decide if multiline execution is needed or not then it would work backwards compatible enough I would think +This is already part of please checkout the trunk +Javas implementation of message parameters looks like it may have other issues and needs more tests +Could potentially impact console the majority of and blocker issues to the next point release +it +I agree its also nice to remind the client users in the release notes. +Applied to trunk and branch +sample configure script is attached at generated using patched acinclude fix release to well address it there +configure need not yes +Thanks. +committed thanks! +After the third one anything +Merged patch with trunk +Committed to trunk. +session should not be a big issue +Open a file +If so it is ok. +I see thanks Jiri +Im really really surprised by this response +See properties for your project +As far as my research showed this issue only comes up on IBM JDK where putAll indeed does not delegate to put but rather calls a putImpl method +Allows the junit output format to be set by the property +I would like this because I run sonar from a top level agregator pom but it appears this does not work and I would have to add the configuration file to each sub project individually and the property in each s pom file +I added ignoring Strings and required invalid +Is CXF also used on the service sideNo response to query for more information in over a month +The unit tests in were failing after I applied the patch +Well update solution patterns +Is it normal +Same error on Mac OS X JDK Eclipse +Therell be plenty of time to review the changes +rev . +It looks like youve got an infinite loop in your +Hmm +I just learn the hard way how maven is working compared to maven +It returns not found +The MIME type on this patch says its plain text but that doesnt seem to be the case +Ok got it Dinyar youve got good eyesThx for the explanation I will commit your patch Thx Dinyar. +classy! I havent thought of going with Listeners with seems now a much cleaner solution +Updated the pull to add Parameter andor Result to the process definition if no custom input andor output names are specified in the wizard +I should be able to get to them soon +My argument being that from a perspective +Undo changes in thrift generated file +Are they actively working on it or just got halfway through and moved on to other projects +Test failure is the normal port out of range shenanigans nothing related to this patch +It is not often that you need to use them on a panel usually these events are used on more granular components +Will keep looking +Futher investigation shows that of Webflow removes the attribute +Im not a class loader expert but please keep in mind that passing the proxy via a remote interface also should work in a distributed environment with the remote instance running on a different server than the local instance +Did it work this way in? It was implied on the user list that things used to work this way and then were changed in +I agree its a little annoying but does the output to standard out actually harm anythingAs you say its not an issue with Mahout +Committed. +In this was not a problem +Patch attached to make that change +when i add a training where can i approve it +Ive committed this and merged. +If the problem does remain with a recent nightly build could you please trigger a thread dump stack trace to work out where the code is getting stuck? +Im not sure where the fix is though +is the initial coordinator for the cluster +This isnt actually a bug +The original issue has been implemented so you can use sqlType as a column attribute +Thanks Laura +One proposed workaround is to have a custom writeObject method that writes a null resolvedObj if it can be determined that it is not serializable +They are two different things +This is the modification from Eric Lapierre in patch thanks +Also given that compress isa very heavy weight thought the extra work for the trigger stuff would be ok +Fixed in the trunk. +I was noticing the same thing +Of course you can always change it in your own project. +Ultimately do you plan on always reenabling the interrupt after retrying or getting to a safe placeWe should definitely throw an error in the case of an interrupt during a lock wait this seems like theone valid case to send an interrupt to derby if a user thinks it is waiting too long +That previous patch should throw exception for a normal job +if you want no guarantee do not make them Serializable +You know who is looking for regressions in the product very closely + +oh yes great optimizationthe hybrid approach is more and more convincingPerhaps we can use Apache httpd rewrite rules can help too to assemble part of the whole siteanother point from the component builds are current versions which will change over time the other are versionned compoennt builds that are published as history and will never changethen updating these ones is more a waste than something usefulperhaps we could assemble the whole site from sources TLP site with CMS and immediate publish currentlatest components which change on release components history which never change after the slow Herve the file is to be namedcontentresources +Ill commit the patch around GMT if nobody objects Massive cleanup of deprecated methods class changed to take advantage of the new preference architecture class changed to take advantage of the new preference architectureLet me know what you think good to me +Postponed to +The mgmt server side works fine and offer the correct DNS serversDowngrade the bug +Further transaction recovery can determine which transactions were originated by the server doing the recovery. +But text under codelisting speaks about attribute lbfactor which isnt in the listiingI suggest to add new lines as in documentation can you add the suggested changes +I think I can even update Spring server to pick up CXF interceptors in a similar fashionHow CXF interceptors though can be picked up individually without the features +Im marking the ticket as unassigned as I likely wont have cycles to complete this any time soon and I am just holding it up by keeping it assigned to me +Asolution would be to write an alternate implementation of and build aspecial package for JME +Not a bug expression string returned is correct according to template code +I may reconsider using your approach Christian seems more clean Sounds good! As soon as i have more time to work on the proposal i will attach a patch file with more tests and behaviour +I hope you had a nice vacation +The attached log of changes that need to be made should work for M to M and should be distributed as part of the release +I had considered the solution Carl mentioned but I didnt go that route because it could result in clutter in the directory specified in particular we cant guarantee the file will be deleted in the presence of catastrophic failures +jws file +It can be the abs is created and not installed +Nice catch +which for the first time includes a for the as well +is also present in but if that is to be resolved I guess it should be a new issue +the source file that gives the error etc +It is just a comment +Hello sorry for the slow turn around +A case when validation does not work yet +This patch shouldnt drag in support for ruleflow +The problem is that most of the content on the site is in the page hierarchies under learn so its helpful to users to have a persistent visual reminder of where they are and which other content categories are available to them + Username scoy +Converted Allans description into a patch +Is there anything to do for this issue +I assume in the original description inserting the same value really means using the same SQL statement ie +This isnt clear from Stuarts report I know +Also i dont see the point in having the Batcher keeping track of Statements when it only does it in some cases +Please add the apache license header to the new fileApart from that the patch looks good +This war demonstrates the problem with relative paths in URL rewriting +OK good idea that would be nice + looks fine to me +Yeah might be a good idea to have will be nice to be able to play test suites as we should also check since theres some support for it in Groovy already to check suites run too +Fixed the issue for Hibernate versions by excluding the hibernate dependency from the wsrp consumer to be a more generic task +If others with more experience in this area have ideas please pass them on or maybe you can even run the tests +I agree with AndreyIn our application we need to support showing in the GUI with the requirement that the user can sort on any column by clicking on the column headerSince the list is paginated we cannot sort in memory on the frontend so we need to the database adding the proper ORDER BY clause +Ill make the review changes and merge my style check fixes into the new code locations +An introductory comment in that regard would be helpful +As I said this now works as I expect but I am not totally sure about other combinations +Uploading a patch that does the latter +Updated the distros so is now available +Did you test this patch with different usernames? If you didnt Ill test this patch with different os usernames +Thanks Tim +Typically trying to do that especially through these sorts of changes is very difficult +There is no reason why this should not work unless the transaction manager doesnt set the Transaction on the preparing thread +Then atsome point a new bug was fixed but as a side effect could no longer deal with buggy system catalogs on drop +My Cient programm that I used to test the web Cient programm that I used to test the web you have use your own client +Please test and verify this is OK has been fixed some time back +Done please test by committing and check the website immediately after +will look at that shortly +Can I mark this issue as Wont fix? +from me +Thanks again Serge for the contribution. +Closing this as duplicate +I believe the existing tests already covering it +For now I wanted to get this in +This is the updated on +Closing resolved Release +The only part missing is the inclusion mechanism for the property +The idea is that all nodes would mark their sstables repaired no matter who the repair coordinator was +For testing provide capability to inject client callback delays +Although the could override the commit implementation most of them dont need to +transportNote that this is the complete thrift root +For the fix to make it to the snapshot it would need to be +There will allways be one SDK preinstalled so why wouldnt you check for that one instead of hardcoding the API limitAnyway as I think a fix for this wont be available in a few days how can I do platform add android without downloading a new SDK? When you have the platform installed you can change the number in the android manifest file inside the platformsandroid folder but this is not possible right now +I see no reason to not include that in the response +Commited on trunk as of revision . +Triaged Sept +Thanks for the review Steve. +Its basically a result of the refactoring and I will check in a fix shortly +But what qualifies as a long timeAnd Id rather it take a while and WORK for large complicated than be quick and fail miserably to generate correct SOAP messages +Sadhu wait already fixed and sent a patch for review +made a blocker for in version +To me its the same argument as not having a grayed out Delete Database button when viewing the list of databases page +The core test failure is being fixed by +Im on the patch on the branch revision thanks. +I went to apply the patch but had a few problems with xdocs and +also jquerry +Just clarify the summary this bug gets listed by itself anywhere +so what about the issue? The fix is done +I guess this is a classic singleton bit so our solution needs to be to apply the correct singleton pattern here +setting patch available revision +Since repositories etc are maintained by Codehaus Ill try to reassign this ticket as a Haus chore. +Ill try and build with mvn and see what errors I get +made roadmap a link to the cwiki as well I think this make more sense +added patch +Thanks Jimmy. +I dont see to access the menu as a huge issue i do and thats the whole point with this got significantly more situations where i have to publish fully which takes significantly more time because of the keyclick sequence necessary and from a usability point of view it is incomprehensible why incremental publish deserves its own button full does not +Ive just reached the same conclusion and tried to change the source almost the same way you suggestHowever I do not have access to VC but only VC and Im not able to make a successful build with that version +Ill create a separate JIRA for that as its irrelevant to DN upgrade +Looks good thanks Tim +So we cant wait until we reach the Handler to check that the DN is correct or notRrrrrrahhhhhh! We have to find a solution which does not implies that we dont rewrite all the snickezrs and twix messages +Thanks for the explanation Filip. +When this validation is disabled warning is not present +Sure +Ed any update +have as little custom processing inside it as possible +This will take care of all cases Java Streaming and Pipes in a generic manner +Actually just looking a little further the public nature of methods and classes without further modifiers originates in +This test does the following Creates roles +I realized after this upate that the thing to add to SVN is just the zip files of the docbook stuff if anything +Default is zero meaning a connection will be kept open indefinitely or until the OS decides to close it +Thanks to Carl for the contribution! +Committed to trunk. + +Otherwise it looks like a straightforward fix for the problemcheers Roland +Until the patch is committed it cannot be be marked resolved +Vladimir yes this patch fixes the VM crash of Tatyanas test case above but has no effect on the tests introduced by +Link to the issue that explains why this expression is deprecated +Thanks Jing +cant make a release with verbose errors enabled by default +Yes I agree using Location and letting it get masked would be good in general +This is an test profile that uses a sync store +attached a patch that should fix the vmem problems +First draft of adding highlighter support to benchmarker +thanks +Heres another patch that improves the logging of schema creation +Please note that Ripple works with any Cordova do you know where should be placed? I mean is the locationname of stored anywhere in the native project settingsAlso I expect that presence of would be useful for tools for code completion +AgreedGregory Wanna take a stab at it +Just in case I added all xsl files secondary xml sources +attachedNOTE this is just a workaround for the current requirement +Attached output from Maven list of steps to reproduce and lists of dependencies +As soon as people start mixing it gets messy +output from ls alR +Specifically I only changedCheers file contains both the plain +I think making a check that is not the first argument works +attaching test content and added more info +The code in looks goodI am a little worried about the test case though +latest patch changes back from to removes the information under collectionscollectionshards +BTW + +imo this is very important for applications that are a little more complex than CRUDplease reconsider to implement earlier. +please update teh commited the code I put in one of my comments +Committed revision +The should not work +Thanks Andrzej and Doacan for helping get this updated +Hi ChristianThanks for the patch +We should simplify the code by removing the needless implementation at the start +Applied to trunk as open a new Jira for this +However afaik the loader delegates part of the job to the GWC one so maybe the GWC patch is sufficient +Yes I think what may have happened is I uploaded an older patchThis time I did a revert applied and just edited those files and created the patch +This appears to be an issue with autoloadAutoload only appears to be safe in MRI +With the second patch I have the same speedup for pmd but nothing for jython +If I remember correctly there is a reason for the flush marker it ensures that the RS hasnt been fenced on HDFS that it hasnt lost its connection to ZK and already had its log splitting startedThe reason this is important is that otherwise it could move on to delete old log segments which would potentially break the log split processIt may be that the locking can be more lax though +OK Ive coded up a DOS batch file that exactly replicates the shell script using only Windows commands so this fixes the original problem I identified removes the svnant dependency adds no new dependencies and will work on Windows natively +on JIRA +Please reopen if you continue to see thisproblem +For now feel free to add release notes link in Description In the newest version we already have download counter that is integrated with our Google Analytics account which enables us to do very nice reports on those +Hm it works for standalone for me +Will integrate patch v in hours if there is no objection +Alex thanks for the clarification +Attaching it early to illustrate the design to get feedback +Committed to trunk. +I tried invoking content assist in the file that you sent me but this error didnt appear +Shows how two tags could overlap +There are many instances where other actions hinge on the result of an ajax request +If you are updating the data every months or so then pull the data in month blocks and then concatenate locally +This was probably resolved at some point long ago +You might be able to get around this by using ant or ant to compiletest the java code respectivelyI dont know about maven naming conventions is there a pointer to docs on whether to use sources vs src +a bit of modification to the adjustments to the mentioned in a previous comment by Gerald I have created an patch by placing all the files into their respective paths and creating a patch from the root +The DDL page was updated +Also improved javadocs +snapshotbinstartvncWhile it is tremendous benefit that users can easily get to their files as noticed in this example commands like find will also hit the snapshots +Yeah I noticed the exact same too during Nutchgora debugging +Slightly cleaned up code +Everything seems fine +I have reverted to old version before and test +Sorry about the patch not going in ok + local variable lock was unused removed +Do you have enough infoattachments to reproduce this? If not please let me know what else I can do to get this resolved +I just committed this to trunk and merged into +I think most of our access restriction problems occur when the test and the class being accessed are in different packages +I agree with Alexis on that the code has already become pretty messy around the this job details mapThanks for the refactoring effort +The checking for data primitive classes needs to be tweaked a bit but will handle with a separate Jira. +Niklas what do your entity definitions look like +The intial automated build test for Linux x should be up in about a platform coverage for the following platforms will take the remainder of the week estimateLinux x ia xSolaris Sparc ia xWindows x ia xThere is a risk of delay on the Windows boxes we will try to address this with IT ASAP +Thanks Devaraj! +The latest downloadable JPOX nightly build has an error in the scope of some of the methods which is fixed in CVS and will be in next build or later +. +I agree with you +patch for Hive QA to kick in +In any regard Im just throwing it out here to get kicked around and hardened +Thanks for looking at this +is correctly dealing with methods and method parameters +I will patch this issue separately +It shouldnt take me that long if I can get some uninterrupted cycles +Ive committed this to trunk and. +Just tested that on RHEL only etctomcat regards the SECURITYMANAGER variable +I will make them enable in patch +Great thanks! +Saw it didnt make hoping for +Hi Aaron I updated the patch by removing MB from the s default value it should fix the problem +Thanks for the patch Ben +my local repository didnt include files even though ti contained multiple versions of +The reason I called it tree faceting was based on the same thing Hoss just said +I will reuse the same issue to patch and HEAD +Im closing this as this issue has been solved. +See or for details +Included patchregards already fixed in the features file missing combined in I take a look on the bundle listener +This same problem was reported as bug +All is all rightThx +Filesystem +on +Hi DanieleShould this issue be marked as resolved and have it closed then? Thanks for you work for adding this to the plugin +Added forum thread shows a thread dump of the serverconsole app that has almost thread trying to load the profile via the profile service? This was during an attempt to add a new jms queue destiation that originally failed with an interrupted exception +Submitting for tests + +Now its both possible if I remember correctly Tom told me that didnt work well and was more like an experiment maybe we should validate this statement and if it turn out to be true it should be removed one day +Since changing this may break existing code well need a new admin setting +By setting the concurrency lower than the cache size items are evicted in error +mapreduce FIF is subject to this same bug with directories only mapred FIF checks if terminal paths are directories +So Im optimistically marking this resolved +Hi Dennis thanks for providing feedback on this issuethe allows overwritting username and password thru plugin configuration and command line +Ill report back as soon as I have done that +Okay sometimes it pays off to read things more carefully +I havent seen any documentation for the fragmentsBuilder element +It does seem to be an omission because the version closes the wrapped stream as do the that I checked +to branchWorking on trunk patch now +So it will be easy for the reviewer to pick the most recent oneIf you will keep your mouse on the previous or current patch then it will show a pop up window in which Date and the name of person who has uploaded patchI will review your patch in a day or so and will let you know my thoughts on thisFYI I am uploading your patch keeping name as that of previous one without any changeAshish Vijaywargiya +Any insight would be appreciated. +No test cases are generated when generating only the server side code +Also the same is done for other dir options and components like hadoop +trunk and branch respectively +thanks German +Resolving as fixed +I confirmed with the ASF Secretary who says all is in order +Ill make an effort to get one in by the end of the week +This is what really takes a long time +Feedback The development team has reported that using the null entry for allows tasks to complete +An important observation is that there will be a transition from standalone to ensemble which I think wont be difficult to deal with in the code but we have to make sure that this observation is correct +I think I spotted a bad map name in the return issuance service one of the set operations is using v instead of what the others are usingThanks Thanks for looking at the patch +Works great for me +Do you have multiple data directories configured +Yep +why are even trying to fix this? was it not the decision long ago to not try support the newer me eclipse versions open ranged +It would be better to keep the command running as long as the container is running and also the other way around kill the container when the command is ended +I can have a look at it if you want +We should be able to catch related to NN unavailability and handle them by deferring the workAlso I can see a useful release which includes some backport of the fix for this issue +This problem will not be fixed as the developer has moved on we will need to address this using the GTXML parser that Justin has worked on making this an expensive bug to will need a volunteer or funding to fix this one +. +I tried this yesterday under and it didnt work +Yes this is an Incompatible change. +So it depends a bit on how exceptional you think this use case is +Maksim the error you get should be same as +when create a web service with the wsdl file it can not generate implementation old date issues. +Hi Justin Daniele +Works on my computer +Will integrate later this afternoon if there is no objection +We believe this is fixed please check in the and if the problem is still there reopen this issue. +this is a special case on my computer as it seems with the following installdir and userdir combinationnetbeansprojecttestbinnetbeans userdir as wont fix +Note the profiling was performed on a machine running for h +Thanks BrockRev bbcedaacebbaeef +It works you can close it +I dont see the scroll bar always even with a new page on dashboard +Im on Firefox on Ubuntu +There are a couple or a few failures that I think its not related to the change +Not very good +usability issues are hard to argue and usually get closed with that is the way we do it so seriously a HUGE thanks! We did get it to work finally but the effort was quite a bit more than but the other features like EL completion rock! glad we switched +Assigned to requires a change that is not backwards compatible version may rename to total and datapoints +Backported to and Uwe you can go ahead and respin +Thus resolving. +Thanks Kim +Karel would you prefer to have this implemented in Arquillian Drone +Possible patch for this as this issue is valid for I propose instead that the Name feature be promoted to the Vector interface? After it matches the Bindings feature in MatrixAs with Matrix Bindings would implement Name and very few Vector implementations will have to do anything about itAnd the Vector classes that wrap a Matrix row or column could then wrap Matrix rowcolumn Bindings as Names +Hi ClausI have made the changes you recommended modified the tests run sourcecheck and verified that the fix is ready forPlease let me know if you have any further comments +Also Im including for now the complete shard response in the track section +Just deploy to AS upstream access httplocalhosttest and check the server console to see the exception +We have to manually adjust our treestate when we make changes to nodes leafnonleafstatus +Once that is done I will the patch max made to the plugins +Can anyone suggest me what do i need to do in order to print the in seperate lines within aThanksve attached the zip file +Ill look into supporting data objects starting tomorrow +ok I think I will do something +Please suggest if the JIRA can be I go ahead and close this JIRA? I guess there are no further improvements required for the tutorial. +thank you a lot once again Carlos! +Will retest it +Im using Borland C Build Windows XP Build Service Pack on a Dell Precision MT Xeon processor RAMIm using +I was thinking something like toolsmaven and tools knowing the background of this request youre not trying to duplicate the from Codehaus are you +DO you mean just the information of an ls l to be shown on the screen + +Works without issue on +Tara youre absolutely right +All the things in my list is closing +Any committers fell free to assign the task and do it +Yes +Since it was never an issue until now we never noticed I guess +nsid and nnid start from zero +This requires an upgrade to JPA and the engineer indicated that he hoped to get to it next week so this will not make it into ER +If you say its a usability issue please provide proper documentation of what am I doing wrong +Added struts and bundles revision +It would have been preferable in the abstract to split this into two patches but it would have been hard to tell if the new connector architecture really worked without trying it with a real connector +The annotation now has a deepSerialize flag so do the json all and json add commands +Tagged webOS for integration +Looks good +He he I just happened to have my IDE open at the right place and saw your patch The second part has been applied as well +I like the simplicity of that +nadya +Thats what the mailing list is forCould one of the committers please close this issue as WONT FIX? Thanks +This is likely not related to this reported issue and maybe just a special case for my setup +This issue is actually related to already applied fix to other optimizers +tombstone buffering on the next attempt could cause similar to as well +Agree with you Chris and Arun +Are the jars in supposed to go into srccontribhdfsproxylib +zip of srcmyfaces directory +th +verified in RF +Users may often change runtime or other such things and want the majority of the urls and system paths updated but they may not want to have to the changed system args every time +Then we can lookup this service to check for the quota +I just cant figure out why it isnt a win could be a bad test or that theres some misunderstood interaction with Hadoop but I wont commit until its sorted +Note that there must be at least a single space between the last param and the closing paren or the work around fails with the same exception +Not sure about this issue +The real change of the patch is hidden by to much changes to unimportant the only advantages that code of gets reduced by moveing parts to a factory and separating rsponsibilities I dont think we should burden our users with a change of the main interface of castor persistence would only consider such a change if there will be a valueable amount of improvements also for the user +Could you include more of the file +Everything is fine thanks Alexey! +Using Commons CLI seems to me like an overkill +This is demonstrated in a unittest called the attribute is added to bpmorchestration and it is documented in the j +Can you force the inlining using the +I think that having optional error reporting built into in the concurrent class is important even if it cannot tell you which of your recent updates failed +Setting bigger reserve space effectively means less of the page available for insert and so even shorter keys will be disallowed +Rev +I think you should add a entry to honor Stefans contribution this is a huge change and a good one too +related to this we should eliminate all the command line options that are obsolete andor semantically unsound +for +The protocol of looking at at local file is file +Once a feature has been installed its hidden from the software page +Patch submitted +The problem doesnt apear when using Castor at the command line level +The latest patch uses ls on windows and binls everywhere else +Hi Martin did you try to clear your browser cache +As said the problem wasnt with my code but with Quartz for which I have proposed a fix already +This feature has been added and will appear in Beta +Agreed +It can also be a bug in GC +trivial patch +I attached another patch that chops off the second pom. +Test case and fix included +I will commit the patch for now +Took a while to remember but here it is +Probably good thought +Indention problem occured by mistake +Attached jstack supplied by Ted Yu on this is all related to relate to +Id also like to see this functionality ideally considering the following Move extension tag into transformationSet to allow different extensions per XSL Maybe assume the default output file extension from the XSL filename such that would imply a default output file extension of html +One could proovide a byteStream class in C that provides these interfaces +How do you find stop efficientlyAlso what if a filter seek hint jumps you past stopIn any case I think we have good consensus already + +We have tried a custom backup MR job +Thank you anyway +After solving there was still an encoding issue +Thanks Jon +patch redone against more recentYour patch seems good and I wanted to apply it +In routine entitizeContent in the first loop which counts the number of chars that need escaping does not cater for strings +I think that the serialization failure is a real error the serialization happens before we ever start talking to the collector if it throws an exception thats a bug +I agree that it could be improved along the lines you suggest but we also need to work out what our plan is +Resolved layouts templates were added to VPE +Note this would be a fix to MINA +Could you try this with? I cant seem to duplicate the error with it +Thanks for the review Army! Committed revision +Committed. +If it does not effect this then yes I agree. +interface facilitates this but is useful on its own +The error is just a classpath problem I think not really a Spring Batch problem +Well thats fine and dandy if you have permission to change the security manager + + +Please check and thanks! +Inside get get and get really dont need to be synchronized because get is already synchronized +You should try to enable also those tests from TC to TC commented out yet as now all required tables are available +We need to be a bit more careful about adding new files without the version Rev Date +Yes I think the PR is ok to just created another JIRA for maximized component in Portal layout +ive added a lot of descriptions in the xsd +Committed webbeanjpa test to trunk at revision +Here we are dealing with String in memoryThen its normal that the character is escaped +Gonna email Kabir and see if he attached is a hack workaround it worked for me this morning anyway +is closely related but its about another issue however using this idea may help to solve the issue as well +Thank you for the review Jorgen I will go by your suggestion for the exception information I have not run tests on this one I will run tests and post the results +I dont think I have seen this in a while is this part of the test suite? +I have to refactor some changes in Add Target Cluster as well +Both the pages seems to be page because I can see sessionid are embedded in the pagesand I guess we have a template but Im not upload the translated csp help page later +I am closing this one. +I have seen emails back and forth about moving to Corenerstone to incubation +Eric recently asked if there are any tables to doc for this +IJ checks the URL and warns about unknown attributes but thats only implemented for the embedded driver for now +Thanks Philippe for proposal +More details please refer to the comments on help review +This would make me more confident that our impl compresses data efficiently and that it does not have bugs +now maintains order of parameters by their order parsed from a query string or in order of addition +I just committed this +images related to the have applied the changes to trunk at svn revision you +Java itself has moved towards such a goal by deprecating methods that needed to be enhanced +Test case showing this happens with both the if and the plugged in foreach. +There is another ticket having to do with the not getting generated under Maven +Ive had a chance to do a bit more work on this +But since not necessary all Linux builds will use gnu transcoder will be used only if is specified in runConfigure if nothing is specified or if native is specifed the original Iconv is still used instead +But Im attaching the stress test that I used to discover this issue +The doc is updated +I think we should work on it +Fixed in branchx and trunk +Security for Java routines is what it was in and Derby uses +I also search the trunk code and theres no such exception with the given message +fixed in test has been in +When a is created we must determine the outcome message charset +Marked Cannot Reproduce +Since the checkin only updated a comment I believe the failures are caused by +There may be some work to do +Ill clean it up run on my machine once and post it here to be readily checked into the codebase +The compare function sorts the output of each and calculates an md checksum to see if the results are the sameWork remaining to be done is mainly in the area of deployment +Change the test to have space indentation instead of tabs and fix indentation to be consistent There is a typo Bug instead of +Being a grails user starting from I did some problems but nothing like what you mentioned here +It was discussed in offline discussion with Hemanth and Amareshwari that right place would be to do in where slot metrics is captured +I got your sample creation wizard is a to create a sample web service for user +affects version as well +Im resolving this as NOTABUG +You should have all of the following +lives in IMP and the fix required a change in the constructor so I copied it in XDT and did the proper filtering in a class called +If this is to serve the newcomer I guess the usual initial workflow is add a store add a layer from the store style it preview itMumble +Mircea I know but my point is that for Alpha which included I didnt need to explicitly enable extras and source jars were generated just fine +Ok locally +If this pointer is zero then it means that exception didnt occur in java because java cannot be executed in such condition +Ah ok then the text selection event would be nice to have +cant you say what the dependencies are? you should know better than me how they relate to each do we need dependencies in here? I suppose the mxj team would know much better how they relate +The latest patch looks good to me +It is no surprise that you get better recall with OR and thus find certain documents related to one of the terms which do not contain all terms +We can backport later if the need arisesScott what do you think +No need to make things any more complicated +Not sure what I want +Fuse plugin documentation will mention that the server plugin starts disabled +Can you please check out SVN trunk build the yourself and provide me with feedback? +Have just had to grant Sonar Admin rights to a developer as I want him to take responsibility for the profile for his project +The remaining issue here is whetherhow to support addingremoving source metadata +BTW you can remove the just leave the file along we can still get a instance after unmarshalling + for for for reduce antlr loading custom expractions when a jar is already in the policy custom actions jars list +Next two weeks Im going to write draft test cases for content assist and I will write here all that can be a bug or may be just obscure for me functionality +backported to in rev +from me +So I try to change it to false and we will see +Committed to trunk +Alright well what am I supposed to link against? I tried FFILibraryLIBC but that failsYou can also use FFIPlatformLIBC its deprecated since it logically belongs in FFILibrary but if you need to support older versions of use it since it wont be going away anytime soonffilib c should also do the correct thingOn linux ffilib libc wont work because linux just has to be different from every other unix and is a linker script not an actual dynamic lib. +This is fixed now in the head revision +Committed initial display of MR service page. +A only dependency +Rebased and merged into the master branch +thanks Pradeeban +actioning review comments actioned so ready for another did you get a chance to review the latest patch for this +We can decide at that point whether should follow redirectsto the same URL or not +Cheers dont get pull requests are you talking about github +I would love if you helped maintain this jira for jbpm too im sure I or Koen will react if something gets edited wrongly +Im sorry +It tries to recreate a graph by tracking producer addremove events but this is problematic +We should keep the unit test pull out the shell code I added and get the thrift issue fixed +See above my setup is probably not appropriate for measuring small differences +The endpoint registry depends on the endpoint patch enables an endpoint registry so that the itestnodes runs successfully +Alexandre can you post your sample code for this issue? I am having trouble running jQuery using the build and am wondering how you got this working in the first place +bq +mvn grails works and mvn grails deploys the app +Will be able to provide a patch for that soon possibly in next weekendthanksCharith +So no further action is needed. +fixed in trunk wasnt clearing the cache after an import +Applied at rThanks! +One option is to create a jira and think some more about it after the release +Navis Is this ready for a final review +Seems like we created the JBPAPP jira at the same time +we have removed all method call for and but not calls of the public constants like because this is still faster +Gavin has provided a very prompt response and confirms that yes the CDI EG definitely intended to be and NOT confirming my assumption that the note about standard scopes defined by Servlet spec was just an oversight +Update to reflect review my patch before updating Gurkan +cecddbce +Under Tomcat I cant see any error nor warning +Committed to TRUNK +One less issue to track +needs verification please +Nop please see the attached screenshot +Would be very useful if the error message could also contain the source directories used to search for source files +Can you post logshow you found about major compaction +bulk move of all unresolved issues from to +backward compatibility etc +I assume I only need to update one file +The changes to abd look good +I was going to ask if and when we can update the jar to be included as part of the release and if it can be included with the latest snapshot of +Thanks so I set this to resolved +on commit w above two edits +Thanks again Rolf for this great contributionYou should soon be able to see them used +As suggested by Thomas and Angela this can be solved by keeping track of the of newly created nodes using a mapping in +We may want to do a patch release with this and perhaps some other fixes +Provided update +Look OK +Else I will take a look in the next few days if nobody else beats me to it +Patch looks fine +Im still testing this a bit +fixed moved debian folder add debianREADMEcommitted! +Has anyone run any numbers against it to see that savings or performance of it? Im at work now but I may try it out in my tests when I get home +Let me sleep on the can use Java dependency +Ive added one sentence that I hope also fixes an issue filed a long time ago +I like how you generate also the! So it looks like maybe the best solution is override deploy to skip deploying but leave the assembly artifacts attached +yeah thats the case with this seedBUILD SUCCESSFULTotal time minutes seconds +should possible fix this later +bulk move of all unresolved issues from to +was unassigned assigning to you bc its in +Please reschedule if necessary +However it is very unlikely there will ever be another release off that branchPlease consider migrating to Oleg +As a workaround added the ant dependencies as provided dependencies and referring to them as reference in Rev +Issue a create rest call and send overwritetrue for the above file +There are setter and getter methods on for fileid +If one has been changed as if the changes should be lost or is too stupid +Closed at request of Kelvin +bug was only present on incremental deploy of zipped archives +Patch committed as revision +Tried enable operation twice +Removed a typo +That will be easier to review +Thanks Amareshwari! +This issue may be better considered as a way to let beans manage their own lifecycle. +all bulk edited issues have in a in +removed dump status CM action which is no so helpfulId prefer this is kept around it can be useful when debugging HA tests where master and RS are killed +Checked in as incorporated into codebase +This relates to the JAXRPC binding layer + +Thanks a lot for your reviewIris FreemanThanks for your test patch and reviewThe does not follow javabean specification +New test fails on builds +Please help me to commit this patch many thanks +It is obviously not very important and we have many things to doI will reject unless a test case is forthcoming +I dont want to introduce separate page variables for each entity +Not yet +Please let us know if it fixes it on your side and then I can commit +I have yes +Committed revision +shouldnt the groupId be +Please verify +Also you have EAP EWP as both the affects version and the fix version +The vote has passed as soon as the artifacts will be on central I will test and update on Karaf. +That level of control isnt quite there yet +Ok nevermind request respectfully withdrawn +The attribute schemaValid referred to in the test assertion is an attribute on the element in the that surrounds the actual message and not an intrinsic part of the tested message +no wonder Commons Math has some problemsSeems to me that the Math test cases need enhancing this should have been discovered when the patch was first tried +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be all documentation related issues to final since they dont represent a change to the codebase +Bulk close for Solr +Os Windows XP SP +Patch applied under svn revision + +Used with permission. +This is already fixed in the latest code +close issues for release. +Also not sure if itself is yet serializabledeserializable should be +If you worry about the amount of methods on this could be a way +Also Nick youre an admin magic away +patch to add a simple server and artifacts +The same issue is with registerTemplate which also shows Task completed in the UI whenever any template with valid URL format is attempted to register +And what about getting rid of this file and assembling the container using pure Java? It should be easy and much faster and this is importand as container is started many times during unit tests +Attaching rebased patch +On branches without the fix for a is seen instead +However the solution is really for the specification to be updated to support +Rayees you should change your test suite +Tests fail without the fix and pass with itIll commit this tomorrow unless there are objections +Solution was to reset the writing boolean in a finally block +and then recreatedThe trivial patch is ascrsrcmainjavaorgapachefelixscrimplmanager bscrsrcmainjavaorgapachefelixscrimplmanager public class extends mcomponentContext null mproperties null mpropertiesOverwrite null mconfigurationProperties null +The ODF doc should go in testresources and the test case in testjava probably inThis will make it easier to review your patch and ensure that this use case doesnt accidentally break in the future since well always be testing itThanks! a patch for and a test documentIs the test patch sufficient? It just tests that a list in a section is included but nothing more +It shouldnt be but it does not look like we have a test +This seems to fail pretty consistently on the branch as well and will produce some noise in the release candidate test results unless a fix is committed pretty soon +Patch for for the previous patches being named instead of +I guess that this approach would have been sufficient because if I understood the code correctly there is an internal communication protocol that surefire uses to pass along test results and logging informationI guess it would be enough if could be with a plugin dependency so that knowledge of runtime weaving would not be required +Patch v adds Override +bq +Opened to deal with remaining ODE issue preventing processing of headers defined using WSDL binding to message part +Thank you +Without it there are no headers generated + +Please consult the cxf documentation the cxf mailing list or file a bug report with cxf +Attached a testcase to reproduce the exceptionTypemvn clean testto run the I found a typo in my test +Since its a straight forward patch will commit unless some one has an objectionHive committers are not allowed to commit their own patches right +Am I right? If right I think we should filter these messages to avoid displaying them in the Choose type of the variable wizardAny ideas +The is now exported and the unwrapped request is used to call the +bulk close of all resolved issues prior to release. +Yes elimination of GC pauses for a lot of applications +can also generate an with the differences but it looks like youre talking about another fileCould you attach an example +but I assure you that once ACG is reached the implicitThis is changed! I confirmed that multiple times using the debugger +Done +Test is passing in jenkins at least the runs that dont crash because of external failures. +Some issues around the img tag in IE fixed and verified that works fine in both IE and firefox +Committed revision Please confirm otherwise well have to reopen this one +Snappy is not officially in Bigtop theres a quirk in Hadoop build system that could make snappy be packages as part of Hadoop but I believe were not enabling it +picking the first match logging +Finally adding the Arabic support for core componentsNow If the browser language is ar then All the core component messages will be displayed in arabicCheck the screenshots +Porting to HEAD done +Committed +The reason for the passwords to display the next time you visit the page is because I have set the option to remember the user name and the password +always preferred detecting and correcting collector congestion on the collector side +to trunk txn +on trunk it would be pretty easy to create a test +Alright +Create hive job history directory if the directory does not exists +After applying the patch my tests using Java pass +duplicate of browser error on submission caused this to be entered twice +This item in the spec is obviously not covered in the tck and we are focusing on integration tests that target all SPEC requirements +to create the wsdl filethe elementprocess bug does not occur in the generated wsdl but does occur in the wsdl axis returns for the created web service created with +I attach a simple code modifications on update when provider is Hibernate When include non portable attributes validate When NO include non portable attributes noneRegards +Thanks Adrian! I appreciate the will have a look at this after world +Marking all current RESOLUTION LATER issues to be against the Nightly Build +I agree that this is very confusing and makes it almost impossible to do parallel builds on Jenkins boxes +It can be set as bean property +Laura I looked through the text change for the examples and it looks good to me +I just added a little bit more property definition to make sure they get created as varchar but the underlying issue of is what you will see by running the test case +Marking this as a duplicate of. +Also it populates a DB with two tables that will later be used to store data because the goal now is to use a DB to store the snapshot information instead of an XML document as XML tends to be too slow for querying +Also on a separate question do you have any idea how to set the jaxb provider to be the default without explicitly setting in bean or web client +Closing I am happy with what we have nowFor Clustered Stateless please check. +It is not a fix +contrib testsThe failure is because of +Issue fixed included in the iPOJO release. +Given there are test cases in maybe the test runs for minutes +Awesome thanks Tom +I went ahead and applied it to and +Hi MarkOK +It looks like is a git patch +In addition upgrading to Hibernate is both for us and Grails users as it has several breaking changes to the package structure +The same spec lack for such a +Should we change to use instead ofI can file a jira to commit +fixed in branch Modification to suggested fixes As dicussed if a exception occurs raise an error message and exit the JVM +In most cases the total size to be copied can be determined up front before the copying begins noYes you are right that we can lists of files being copied and impose whatever constraints +The default template which you cant change in the free version simply shows to much +Patch rebased with the latest trunk +I post results of latter +this is common for subprojects framework module which causes cannot exclude these common modules from build +Some minor changes +from to in order to develop the bugfix. +Thus avoiding feature reqests such as this +Closing all tickets +Oops forgot the most important part +There will be cases to run the common joinOne is when the resolver of the conditional task returns the common joinAnother is when the map join local task failsIf not reset the tag during the getting the backup taskhow to distinguish these cases +This topology runs in spawned processes +Thank patch for trunk which takes on Sebs comments and also implements correct logging mechanisms +The other are thrown when there are inconsistencies in the user inputs and operation cannot succees for example creating a file that already exists +I just committed this +This is now available in as well. +The mechanism for requesting this feature could stand some polishing but to do so would require some API changes I dont want to make this close to a release +It doesnt say when they can be garbage collected nor does it place restrictions on implementations to garbage collect after executing SQL +What is the use case that you require this for? It shouldnt normally have any impact outside the security configuration +reopen to change resolution status +Thats the idea +Thanks. +Credit for the idea goes to Doug Grove by the way +Note that this event is only triggered if the view scope has been created +Ill look into this issue +validation does not fail for the later +Then having a single folder for all the results which is outside of the module build dirs is probably safest +Unfortunately turning off the cache didnt it occurred again after about messages sent +New patch w fixes to not overallocate states +This is just a theory that I cant confirm now as I cannot log in the runs +Ill add a specifialized subclass for JR to deal with this case to make sure the module remains dependent only on the spec API +I am not sure if there is any traction on this +If there is an easy to use Ant based mechanism I would use that +Ill hold of on resolving this until then +Steve Im mainly after the fact that FOR UPDATE is broken since Hibernate so if that can be fixed in it would be much appreciated +Thanks Sean! +patch applied and tested +Well this actually seems to be more of a bug with xfire error reporting +Thanks Gunther! +Instead of removing pool from the log statement the log statement is fix has been applied to branch remotingGA +Static? Werent you against that? He meant a static final! It is just to make the analyzers that depend on core stuff fix to a specific version +Can you check if this is still a problem with MINA +ThanksGav +This is deferred until final +Documented as a known issue in the and release notes +Ive managed to compile everything under jdk +Im not sure why +Do you have multiple plugins each with their own Logj jars? The default is the +submitting patch +Then you have to handle this in the service because each checkboxe will return a Y or N value and you have to handle it +ISSUE RESOLVEDMessages are tossed if you do NOT use client acksOnce I set and used client acks to the broker no messages are lost +Between the file versions and the scope of the lock statements in the Get Set and methods were changed +Daniel you need to add GITHOMEcmd to your path on WindowsBurkay Im looking into it right now +Please add this patch +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +WDYT +I will start with it next week but I will probably need some help since I still not familiar with the CAS Editor code +Ive committed the rest of my work +Reopening workflow issues that are still marked as resolved but were part of the release. +quickstart project change here broke httplocalhostajaxtreetableeditable and Topicus which is an to wrap the generated columncell component and be able to use parents markup. +Thanks Sreekanth +I see there is no activity hereIf you have any code or ideas then Ill reopen it. +The pull request that where the management context returns a is merged +Tim the fix looks good to me +Patch applied to headSending configsSending configssrcplanSending configssrcplanTransmitting file dataCommitted revision +Emmanuel had renamed a couple of methods when applying but I have that under control now +Perhaps we need to create a list of possible choices +If we change matches to find then we might not be able to handle the cases where we need to do a exact matching +Submitted Jira to Codehaus +hi Jeff I pass this jira onto + +Thank you Mayank. +Looking deeper thru the code I dont understand why my patch worked +Close this bug for now +And I have no one willing to assist from a committer perspective +Hi JoeIve talked over the exporttrue issue with the rest of the Resteasy team and the belief is that exporting httpcomponents is appropriate as the default +Sure +Wow Ive been busy recently and havent had time to work on this muchIm going to open a subtask of this issue to track the contribution of a smallbenchmark to measure GROUP BY performance +Its unclear from the docs whether the queryresult size is for the entire result set or for just the batch that the queryresult represents +Added equals and hashCode for so that equality checks evaluate correctly. +attached need the Java class to help reproduce and fix please +Assignee is now removed when resolving as tested! +I did rebase it +Lets move on + should test for tickets +reviewed and committed +Please pardon my ignorance but I have never applied a patchHow do I apply it? and to what version should it be applied +It was applied with minor changes +The sm stays the default +I got it that the token will contain both the main principal and also the group membership based on the discussion on other Jiras +sigh needed anymore marking Resolved +In my experience a pool of connections always tends to perform better than the same number of connections despite the overhead of thread contention for the pool lock +thats the console of the junit jvm not the console of the jvm it forked to run the bet having ECST extend will clear that +Should we keep this issue open + +env entries rev rev allows subclasses to be +Users expect their jobs to complete even if a TT running one of their jobs fails right? So were clear on terminology what I identified is that a single TT failure can cause the job to fail +Given that it breaks the console the value of the whole thing is very questionable +The NPE is caused by a null ivy local file in at line +Shouldnt the title be Generated Method Names are NOT using Customized Class Names in xsdconfig File +So it does look like the Broker bean was destroyed and hence the +Can you refresh me what do you envision the startstop hooks for +defineCONFIGDebugDEBUGTRACEWindowsPLATFORM +Reversed patch +On Windows those files stayed there after running tests in suggest this one I think i know the problem on WindowsSee my changes to +Hi Juergen somehow I missed the email notifications from jira hence the delay in responding to you +I didnt include any unit tests with this patch +And Ive created ticket to prevent removing a filter which is used by a filter +Then when I ran the application and loaded the form that uses scheckboxlist tag I got the NPE +Ben I dont think I got the comment +Are these things on the road map? issue is related to and affects XCP as wellThe patch to address the split brain issue isnt a fix as much as a workaround +Its commented out both in the eclipse workspaces file and the one included in the deployed jar file +right +Ran test derbynet with ibmibmibmibm as well as the sun jvms +Now I have to check if it is the case of a single json object and do all the manual parsing +Just try to use with a long +Open the schema types in the debugger and see what information they contain and then look up the Xerces API if anything makes sense +Im going to resolve this as done for now since it works for the normal use case +The latter is probably related to the +Patch v adds read lock to is modified to relax split policy constraintThis requires in local cluster +Closing with the release. +bulk close of resolved issues. +Well JAXB works for HTTP GET invocations when you have a SOAP binding set up +Linking to latest issue that revealed problem +It has the unfortunate characteristic that it will perform very badly on most file systems and even in some cases of HDFS +Will do that +Give me a URL and Ill include it +I also incorporated the reprotest Kathey has writtenShould we try to say something about what went wrong in the error message such that the user has a chance to understand what went wrongPatch ready for review +Mr +Thanks Aditya +Thanks IlyaPatch applied to ARCHIVE module at repo revision rPlease check that it was applied as you expected +The first comment refers to and the second comment refers to are new samples attached so lets retry to reproduce for applied pass +This is very crucial piece of namenode code +probably needs to be closed +I added it but got the same problem +Thanks for explaining + pending jenkins +Closing issue which has been resolved for some time + support just added +Thanks Varun! +but its giving me a +thanks GabrielRe when the switch was made I think it was but let me double check that +Closing out patch applied. +They are located in qlsrcjavaorgapachehadoophiveqlexecvectorexpressions +I may try to push out a beta release soon but i cant promise anything +Hi BrianIndeed this is an interesting infoBTW do you know reasons why it cant be uploaded to Maven Central repository +Do you want me to test? Do you test this problem? why this happend? Originally Our Company want to put into our production environmentBut I am very worried Even a basic decommission operation can not work ok in +FYI this was fixed about a week ago in the nightly buildsSo download any build after and it should comments were not descriptive enough sorryThe Javascript tag is always loading the from the and using them to build the dynamic javascript +Set +Committed trunk revision Backported revision +push out to +Can we bump up the priority? I am looking for a implementation with full spring support +Thanks for bringing this up! Fixed in revision with better handling of embedded OLE objects +Fix using find instead of cp R all the infrastructure of build +We can this if more information becomes available. +Thanks Luke! A very trivial bug related in would you help to review it +The and have tesed +Proposed SOA TP Release NoteSupport has been added for local configuration of +Let the workers use their and append the log file path +I think I have found the cause of the failures +Attaching a patch which adds a blurb to the top level download page +The rebased patch looks good the shutdown changes I was referring to were in v +Thank you +Can we have confirmation that fixes are in place for +AIUI using a static logger may cause problems in some environmentsMaybe the logger could be cached instead +The zip file has a bin directory +Thats not legal. +It arise because the event listener is not copied when copying the page to should probably also copy the listener that are on the portal objects too probably +comment in getState is incorrect +already done with flying colors. +final version +Reopen to investigate message +The check instrument ready was only implemented for Wombat +so we wont read garbage last data is better now otherwise +Sample files would be appreciated as well +Shawn upgraded jaxb to so this issue should be resolved +changes only +Good idea but discuss these things on list so everybody is involved +implementation this use +Same exceptions +to Rado based on Pauls previous comment +Ive also faced this issue and it blocks to default behaviour of the releaseperform goal +Thanks +To handle youd need to add another such data structure it would probably be best to introduce some new abstraction to represent all of theseWould you like to try this onebq +Why was this issue marked resolved? I dont see any resolution thats essentially what Im doing as thing I dont understand however is why the file continues to upload after the exception is raised +Otherwise theres a more serious bug that we need to investigate +Mark can you remove from? I got a lot of inc token log messages when running +Thanks for catching the parallel structure problem Knut! Im attaching yet another patch and which makes just that one change to the bullet itemKathey I would be very grateful for your feedback on the overall patch +Thats not correct for Tomcat the Commons components in Tomcat have a different namespace to avoid classpath conflictsNot in the ones I downloaded +However in the JMS message listener container case there are no such local JNDI settings +To test this patch see the use of emphasis in the thanks. +section of the still describes opcode as being bytes and gives the possible opcode values as hexFixed too +After that it will be released into probably in February +It is really desirable in such cases that the large chunks of static html be relegated to include files and that the stylesheet should just be able to include them +Fixed in. +I think that Railo is a good opportunity to upgrade from to I can look into it if given some guidelines about the classes in play etc +Is the cardinality set wrong to begin with or is our estimation logic in errorNext you can definitely set option nocache from the user level query +Hence I would argue that there is no need for special proxy handling usage scenario do you have in mind there +Merged +Great work! So it seems like we are affected by something else and I have no idea what we need more debugging info +So its closer to transport than to command provider imho +I was able to stop the infinite looping by removing the mm attribute from the first block in the document +Attached console log with tracecompilerstart stack trace is triggered as soon as I open the Groovy preferences switching the compiler and restarting this is not another restart does not help +Sir I tried removing the space but still it does not work +Thanks SaschaI think its minor but may be useful I commited in trunk at r +Committed the revised links. +Instead of using a Map I use a Set as I want case sensitivity +Just committed with a simple unit test +I think the downloaded snapshot is working fine with geronimo +By this I mean that for each version a user wishes to have in the xml you can either use the new way or the old way but not both +But I noticed this same problem and I thought that a quickstart couldnt hurt +Youll want to look at the input and output of couchkeytreemerge as thats a likely place where this sort of thing would occur +These issues have been resolved but never closed +Now we filter results out in java. +bulk move of all unresolved issues from to +Attachment has been added with description testcase resize window to see the affects only the rightmost and bottommost pixels of the whole imageAfter I added the single line to the testcase the right edge of the gridseems to be intact. +Making it a new component also allows us to skin it differently and I imagine that a tray could easily look different than a sheet +Added +Tested only in the simulatorHowever what i did was to test the command manually in the terminal when asymlink using the SRCDIR from the script which gives me the followingios cristobaldabed ls l www cristobaldabed staff Mar www +Looks good +i dont see any subversion commits for it? did you fix it +But Ive not taken the time to investigate any of them yetThis JIRA should remain open +I think we should create before asking for resource +Yes the race is intended because allocedsize is only used as a hint +See lines +Hi Harsh can you have a look into the updated patch when you find some time +It only failed on which is not related to the patch +Cheers +Fixed with auto as the default for backward compatibility +I think this would require the ability to unload or update the assembly to a newer version without restarting +Thanks! +I can run the program with rows in about minutesI also took a look at the query plan and it seems that the query isusing a reasonable query plan scan MessagedUsers for each Messageduser which matched the account ID query Users for that user idLastly I looked at the program and it seems that half the rows in theMessagedusers table are for account and the other half arefor other accounts +Just committed this to trunk and +hmm +Might be that the old datastore is still inside the resource pool +Hi attached patch for fix of the if isScanned value a before being scheduled to be scanned +Rather than permanently disabling the cluster it should be able to come up +I reverted and updated my svn checkout of +I will commit that if there are no votes against it within the next hours +It seems to me that you shouldnt need to add a URI variable for path parameters to be ignored +Any ideas why +mark as regression as this worked before +Actually the reserved names are not only groups and users but there are some others like private download + +This seemed to shake things up +Smooks tools only pass the EDI Mapping file path to smooks core engine so the file path was loaded by smooks core engine from the config I think we need not to parse the EDI Mapping file to get the resource path and update it when pass it to smooks api +Sure +Until then users can learn from examples +We havent heard from you so Im going to mark this invalid +By default Axis dont use the Mapping so xsdunsignedShort are mapped to internal Axis typesTo avoid this you have to specify to the generation tool that you want a special Type Mapping versionfor command line add T +Provided to Michael Kurz for this patch +however I have a comment on one of your update +Darguments helps but I think its for parameters eg +Add dependency entry to MC file +Resolving this issue because it is was pushed to master only and going to be available in in JBDS build. +Doubles can possibly be varint compressed as well + +one bug is that this generates overly short words since the max we pass in is really a maxbut we would want that to be the exact number of elements +Closing as resolved. +If the original problem was fixed and this issue is unrelated it should be done under a different bug report yo keep things straightThe second screen shot seems to have a totally distinct stack trace from the first +This could potentially help that. +So yes I think it is just that simple +Sorry didnt notice Alberts update until now +I thought someone told me there was some issue with the Spring DM testing framework but in the end if it would work then it is fine by me +if it applies cleanly enough seems like a reasonable line in the sand. +I assign this to Lucene as we may support this with flexible indexing when we enable custom segments files +Fix committed to branch +notes have been committed to the branch so this must be ready for verification +Not relevant. +I cannot reproduce this could you send me a test case to reproduce +The example is for the +Maybe just commit this patch since it seems like a problem +Tld discovery should be working now +And in my test when Compare family only scan +Fixed and rebased for trunk +must be in one linemvn archetype +But I dont really care if it stays or changes +Is there any chance we can get this featurefix in for release +The new property is and defined inI did not change the default behavior meaning by default nutch will crawl the parents for the file protocol +done from my side Seam needs to integrate beta now we just gotta wait for next EJB RC. +Correction this should be the now that the build is back in a decent shape I think this is basically just a matter of properly configuring the classpath correctly +The XSD files have been written later by translating the jibx validation to assist developers while they create new configuration files +I think thats a reasonable solution it pushes the responsibility into the hands of the developer and is more generic +The list of filters would influence the design of the because it would need info from the child filter optimizers +fixed. +Dump is a dead link? dump is not working for RESTThrift server +I just upgraded from sonar to and Im facing the issue as wellIm going to try out the sql statements mentioned above now +I think this is probably OK but dont want to commit it right now too much going on and I havent thought about it enough to risk having it break AS in some obscure way +If possible please get the latest version of and verify +fix and checking +bulk fixing the version info for and all affected issues have in appears to have been committed and included in +There are Scorers which cannot implement that efficientlyWhich Scorers do this slowly +please cannot reproduce works fine with both and +I did not create two separate patches anymore since they will be in conflict when merging in svn and the fix is pretty straightforwardThe fix basically does not render the popupCalendar button if the role is not in enabled or is readonly for inputCalendar to be consistent with the behavior of inputDateThe patch works fine for both tomahawk and tomahawk +Ill take a look at this patchCheersChris +There are other differences of formatting that arent essential to fix but fixing these as well makes life a lot simpler because it makes it much easier to see the important differencesTo resolve this issue I am making the following changes +Just add a serial equal to the previous value as this is never used outside of the server +Committed to and trunk +When appender is triggered it only updates its counts field +Closing as cannot reproduce +basic selector into as there was not sufficient time to get into +Our script is not up to date with the application +Maybe it is a good idea for to optimize as much as possible for developers especially for those migrating from Glassfish +Could you attach your screenshotAnd the quickfixes are not implemented yet +Thanks will try that and update you on the result +But though maybe we can simply add a comment that the default value shown here could be safely ignored and instead the one in xxx is valid +Patch to make Gant behave more Ant for sure if the way I have implemented the right way but I hope I did +The goal here is to develop a archive format +Would you be ok with doing that optimization in separate JIRA +Previous changes assumed that once user was connected it could be kept as is +Essentially I am looking at an easy way to reproduce the bug with the +The cancel action is not thread safe see need to find where this event be canceled If its not performed when continuation called back thats the bug +move on to the major compaction after the flush happens + as I have seen this need elsewhere +When you update you need to use the d option so that it checks out new update yes +It should be fixed now +In my opinion this is not an explicit behavior and I would like at this you. +Applied to trunk r +The connection is only a partial connection +The issue can be if necessary but Im not sure the suggestion is feasible. +So I think we should fix thisWe found the issue while profiling a service in our application +not a bug +Will attach Victor I think we can push it to Alpha as fix version +If I can come up with the time I will gladly take your generously contributed maven build files convert them to maven and start using much effort would this be given that I know very little about it + +Personally Im ok with it as I said its slow but at least correct we cannot say the same for the old code +TomGreat thanks for trying it out +stuff + +The description in guide is corrected +Marc thank you for the patch and sorry it took me this long to get to thisMichi +For now I think doing the patch is fine and I agree with Dies that we cant magically delete FWs unless we can be dead sure they were created by DC to begin withBTW the solution can fail simply because the DC server is shut down before we ever get to issue the +If this isnt working tell me and Ill just zip up the files +backported to +You down w my committing Ted +Rick can you please check with Rob how to resolve this? original value for messages is now restored after the tag has completed processing. +I am attaching a really early draft of the implementation for anyone thats eager to try it out +The standalone ESB server should be no different from AS deployment as it is just another profile of the the documentation or to clarify this rather than support two deployment options is Which quickstarts do not work in the standalone ESB +Can you link to this bug report where you mentioned pspy for the first afair i thought we said that its pspy whos misbehaving +javadoc warnings are related to Suns security packages being imported and are unrelated to this patch +Heres a directory listing of the rsynced backup of today under homebdelacretazdaisyr bdelacretaz bdelacretaz K Nov r bdelacretaz bdelacretaz Nov r bdelacretaz bdelacretaz Nov r bdelacretaz bdelacretaz K Nov r bdelacretaz bdelacretaz K Nov r bdelacretaz bdelacretaz Nov r bdelacretaz bdelacretaz M Nov r bdelacretaz bdelacretaz B Nov r bdelacretaz bdelacretaz B Nov Can someone confirm that this is all we need to recreate the Daisy setup if needed? +I was originally concerned about defaulting the code version levels and thought it would be better set in Attached is a patch that sets the versions in and reapplies the changes +Thats the one Im looking for +This would seem to be a bad thing to me +if you would the patch id like to get it committed +We should retain backwards compatibility support for ora if possibleIm tempted to use the tri prefix +Im not sure whether the documentation is added to the correct section or miss great +Fixed in rev +Changed to a new feature request +Its very strange +What happens when you set true +I would graph ingest performance vs tablets +does ur scenario match with kalyan? as tim says a unit test or project to reproduce would greatly help +I have changes in progress in this area for and its easy enough for me to just roll this change in with my patch +Committed to and trunk +bq +There was some discussion on including this in on the dev list +Still confused Why resolved and not just left open +So rather than revert the change Id vote for implementing this behavior +Review and applyThanks +UIMA based NLP pipelines can use components like Solrcas or Lucas to write their results to an index +Heres the Jetty libs +This seems a resonable request are you comfortable working on this issue Tore? It sounds like the solution is going to cover a few area of the code +Since I use eclipse I offer to maintain the eclipse build for webservices +Maybe go a step further and make a symlink etccloudstackagentplugins easier for admins to find +Thanks for the donation. +I do some testing then i commit the patch +This will help in understanding the test better +to trunk +Yes youre right I ran the tests w code coverage analysis enabled and the is fully covered +Toby contributed it to Open Symphony +Fixed that. +Attached patch changes to use name for equalshashCode +Make it a static nested class +It belongs in startsrctest in the trival change from my last comment into a new patch +I think this change may be worth posting a news item now and after the switchover is complete +This went in as r but I neglected to include this JIRA in the commit message. +currently this cannot be tested as the corresponding features are still missingapart from that i would consider this issue fixed +hmmm it does work on roster security examples but not in new projects created with the secure archetype +Im assuming you didnt notice that it was assigned to me a while ago +bq +Works for me too +Im with eric here +Your one is in notificator +Lets fix the js generator +I am almost finished just a nasty issue in with the commandLinks which I still need to figure outif anyone knows how this renderMessages javascript works I could attach a patch containing what I haveso far +Confirmed on XP with latest Chrome the popup displays a loading +Im reluctant to commit to another project at that fundamental of a position in our definitions. +Heres a corrected one +Ping If I can help somehow maybe do it myself according to some guidelines document I have found please let me know +in M +A chronological mess +Weve already migrated most of the user code away from +The patch is perfect thanks! I will apply it ASAP +However I have limited context here so if anyone else knows more about how users are using this stuff I hope they will speak up +Anyway it shows up correctly on my system. +Ill see if I can get the error messaging to fail or perhaps dump a warning to the screen when we are unable to load the openssl gem +I just committed this +Fixed on trunk. +Validator validwhen now trims the string before comparing to null. +I will need to DEBUG my tests to see the source of the NPEs +At the bottom have buttons Add User Done Editing +This issue in concerned with the usage of and is not specific to will update the description accordingly +With the latest source it works but not with the official releaseWe need a workaround or another fix to release our app with Cordova Thanks and greetingsKlaus +I would appreciate if someone can look at this patch too +Im observing the same when trying to compile it with VS +Ok Ill commit the patch before the end of this week +Kurt could you check to see if jBPM has a way to setupdate the description of a task? Im not even sure if description is a field on a jbpm task +Committed. +in XS Xen PV drivers is required for attachvolume +True +Whoops remove errant semicolon +Hi Nick thanks for reporting the bug I will look into it the following daysWerner +Luckily I saw it during code review +my changes to the abstract maven report mean it gets rendered to a string not a file +This is a set up problem the permission gets denied to create the directory for the templatenfsexporthomeautomationsecondary error msg mkdir cannot create directory varcloudstackmntVMtemplatetmpl Permission denied on the host do we have the correct permission +So I was treating it as a known issue which should go into the release notes +Geotools provides the version and geotools works with the version +Glaring deficiencies of the spec are no longer opportunities to make a product better than everyone elses +Sounds good attach the feedback here and we will address it. +OKSo can we get patches reviewed and committed? I can bring them up to date on whatever branch anybody would like +This issue is not approved for SOA CP +Mahadev can you try thisIf it works greatPlease dont commit this to trunk though it isnt needed there +Minor comments Typo differnt The newly added comment is not very clear +Error in log which appear when try to deploy Seam Portlet generated by like missing Seam libraries +Yeah known problem +This fixes the problem and the testsI used s extension for creating a fixture which spans testcases allowing us to avoid bouncing Zookeeper in between each testcase +Indeed switching back from Xvfb to Xvnc when necessary is possibleBut we also make husge efforts to get build behave locally almost the same way they behave on CI so I hope thet in case a build is blocked on CI it is always possible to reproduce the issue locally which is more convenient for debugging. +committed to works fine on +patch root is trunkcontrib merci Karl. +People who care about proper build targets should file a separate issue +Test case +Yes that sounds like a good understanding of the issue Seam is making a wrong assumption that servlet mappings always end in +I have encountered this issue alsoand I think this maybe is environment issueI can build success on a virtual machinebut in a real machine I encountered the same issue. +is the new to trunk +Spooky it is as if you were looking into the patch I did not post here +DarranHas this improved with EAP? Or is it still an issue? +Shouldnt be source dependent +So i could set tempDir to dirtemp while the JVM working dirs are dirJ? But I still have no chance to get the JVMs own directory from inside the policy file without a sysprop. +Retain support for the Phoenix container +Triaged for +The tests all pass on trunk without the patch +Will attach a patch that allows user to supply a local maven repo and override the groupid and version for DB and Oracle profiles +Therefore probably the patch may accommodate to after its updated +Maybe its because I am using functions which dont work on them yet +excellent idea +With this driver all test pass +META +Id expect a how to document that we should test +I just committed this +Assigning to the maintenance issue will be due to sect tags being used instead of section +All uncompleted issues are now moved to the new To be determined for future releases category +Thanks Milind! +Marking patch available to tests run +To view the process details and package content you have to patch the RESP API using the patch provided in this JIRA +Looks basic to me + +bq +After creating new auction service there should be a conversation ID generated and inserted into Conversation ID field which doesnt happen with IE on Windows +Perhaps instead the format script you guys have should be checking the size of the imageIm just not buying this is a change that needs to be made especially compared to other file system technologies +Thomas thanks for the report +I spent a lot of more time looking at this +And magically everything worked +Anyone know how to parse the contents of the release audit warning diff? Applying the patch to a fresh top of trunk tree runs the raid test successfully +Over the weekend Ill setup an area in skunk then copy the source files from trunk to preserve svn history then add the new modular build filesIve downloaded and built the maven zip +I am facing this same issue +More to the point why are we shipping +It was a very minor thing to adapt the test cases to fail on old code and pass in new code so I added thatCommitted in Thanks Jason! +Regarding my previous commentUsing HQL that is exactly like the example in the issue description +Heres a patch that replaces the classes weve copied from Ant with a proper dependency toThere is still no release of so Im not applying this yet in trunk but Im planning to create a feature branch where we can add support for more formats from +Please verify +Committed thanks Raintung Li +Whichever you think will be easier to integrate and maintain +Just checking in on the progress of this document +My workaround is awkward and Im not sure how usable the Command interface is without it +I looked at you approach and it looks it be possible to provide a patch against current SVN HEAD that introduces this behaviour to? The main issue with handling is when are being used to do things like decorate a page +The policy runtime is slightly different in when compared to +Is there a better way +WillemThis far its clear to me but how do I determine the to use? Is the Proxy sending the method name? If yes I could use it together with the service interface to determine the soap action to use +I was creating the new Type with the same URI as the generated Types +So Doacan will take a adds better with basic examples +As long as makes it to this is not a huge issue for meWe can keep this Jira open for further discussion or close it as duplicate ofIn addition to the question of how we should handle tmp pid volume dir errors maybe additional feature for the datanode to decommission when it decides to kill itself +Is there any documentation on how this should work exactlyThe iOS version of debug seem to be able to build any project via args is this a requirement? or because the template includes the scripts can it just build the project that is inie +Im fine with that +I have committed this to the trunk and branch. +This has been committed for me by Alan +Attached patch fixes this problem and implements regression testThe unit test also includes the check on correctness of encodingdecoding work on extensionAlso the patch slightly improves the API of class +However the problem I was seeing would also apply to a standard queued think the problem stems from me not appreciating the need to tune the pollers when dealing with queues +If set to false will not use the bytes ID of the foreign message +It is better then before but needs a lot better test coverage before we commit +Ill verify it when its included in productised build +Igor + +Made the issue more generic so to involve more changes +So all other will be idle for that time +Specifically well be clear about the fact the sets its order property to and that this can interfere with when flow and Controller paths overlap +As I mentioned in my comment I am reluctant to patch this myself as it is not jbossts code per se that is at fault rather the transactions integration code in AS appears to be electing to use a different impl and this is not my code so there could be a reason for this +Ill create a separate JIRA for the handling in which just seems wrong +Proposed patch +btw a minor comment here I think the fix version is not correct +thanks the patch looks good ill commit thanks for thatStrange that we did not notice it sooner. +really just waiting for build relation to the JMS problems and we can confirm to you the problems with JMS can be considered closed +This is my implementation of this feature +GreatIt was because the setter did not look like the getter ie not a single argument +How about introducing a annotation that takes an array of values? +We may encounter some encoding andor line wrapping issues with this approach but at least the basic functionality works again +Opened +Here is the documentation +But is this issue really closeable yet? I can still reproduce the error behaviour even with the new +punting to pending karaf to be resolved. +As this issue has not seen any updates for over four years Im closing it +cherry picked commit from master and pushed to with JBDS +Can you confirm that this happens with? Also have you tried creating the document type via the factories in the document rather than the implementation? Does that make any differenceCheers What you see there is not a memory leak +Does it work for youFYI we do characters a line but looks like the character lines are generated is that right? So the patch includes the thrift generated code it seemsIs this comment right? It doesnt seem to apply to your new class A Mutation object is used to either update or delete a +Metrics is apache licensed so I think Im OK to just copy classes overYes just mention this after the copyright header in the source file that the code was forked +Oh sorry I think I got confused with the warnings for Xvfb and gjslint +Could someone add the Testcase included flagThanks +Attached patch +Fixed its foot. +Im simply trying to see if I find out what is going wrong and help him fix I am the one who requested and implemented I think I know what I requested +If the alternate patch is fixed and deemed the right way to go this issue should be renamedset remove qt parameter from final request when it starts with slash +Thanks WillRev adccabcffffeefedde +Listen to a different event such as select +In that long call chain it looks like there are multiple places where we can get an NPE from +My eclipse is not behaving +Defer all unresolved issues from to +the part causing the problem when allowing not to haveAmila. +Thanks Matthieu! +Ill take another look at it tomorrowdont knock it its still better then the docs we had about beforeCommitted revision Committed revision . +I have converted the variables from primitives to atomics +The fix in in revision of file srcorgjacorborbiiop got a patchfile with this change he will generate a new version of the library and upload it to +I have a small comments +bulk close for the release +In theory this is not true but in practice my concern was that once a value was fixed and known thered be a significantly higher probability that someone would include it in some data +I am unable to appraise this until Friday if we cant get any further comments for the time being Im sure we can wait until then +bulk defer of featurestaskswishes from to +I still feel that is a thing but its not a blocker for meIt sounds like youve tested under the conditions that concerned me though your findings contradict my own +Otherwise you have a memory leakIt would be really simple to come up with a patch that does counters +but dont if not +bq +is the latest patch +But I see your point +Where should shared utility methods go +thanks Matt +I inherited this issue from the previous intern here +No +I commented out all junit annotations +For example do you move a region because it gets hot for seconds? minute? minutes? minutes even? Im not sure where the line is but it seems like the goal should be to move regions for persistent high long term load not transient spikesThoughts +rename it +Please test to verify it works +committed to SVN trunk branchOleg +Assigning to Pradeep for review. + doesnt have getOperator but it probably should +Not sure about this but will cause an exception to be thrown? If yes can you check if thats the intended usage where you are using it? are stored inconsistently between Deprecated and hierarchical configuration +Hi Manik +I dont normally see that much difference between runs +Revision is a showstopper as currently trunk Java clients cannot connect to the trunk Java broker on Windows +I was hoping to get it into Milestone which releases Monday +The Seam Servlet project offers this functionality + is dead +Suresh any updates on this +Fixes lazy initialization of the components Cas Pool +Im also not sure if it would be ok to park the requesting thread anyway some more discussion would be required +Steve would you please review the fix I committed and close this issue if you concur +EPP eclipse packaging projectbut any kind of aggregating p feature which have too limiting versions on them will prevent installation of these wont they +You really dont need change at all +Looks like someone else checked in a fix for something that was hitting the same bug that my patch was hitting +It will definitely clean up a lot of code I have seen +Closing this issue as VS seems to be will be out of office todayPlease use the following distribution lists for supportMarket data issues OBJ support issues JOLT issues Discovery issues MX issues issues Rgb support Thanks Dheeraj +Patches are welcome +I tend to agree at this point +Please provide proper patches and the bug. +Thanks +JBESBGACP +A good question is how do I reproduce it? running simple compilation or are there some special steps? +sstablesplit should only be used offline +Thanks to Thomas for noticing +Good spot + +Archetypes are actual +Shall we target that instead +Fixed in resolved issues +Usage of conversation scope seems to be another workaround need checking to confirm +Minor set is changed to set although it still takes only one token +Ive updated to be an issue that we will address with the JSR. +Looks like this issue is only in +Yeah its there in +add casesPlease check the patch again +intersecting sorted int dicts should be faster in the general case +This doesnt have a bloom filter due to reasons I outlined in my previous comment +bulk defer to defer to +Use username for that right now +This is because JIRA is often unavailable these days +plugins +For Whirr I think it makes sense to use a logging facade like slfj rather than tightly binding to jclouds logging system +If it is not this simple we probably need to push to +allows for constructors annotated with Inject and I believe only restricts this further +esb archive described necessary screenshots provided +A possible fix to this problem is to copy the class to rename the classname and remove the Action interface with its associated methodsNext have extend and implement Action and its methodsThis at least solved my problem where Im using the restful plugin and needed and as well +We should probably also make setquota accept both n and b when called from the client instead of barfing +We cannot commit patches to the plugins yet since they wont build if we upgrade to because of a broken transitive dependency on +class files +You have to wait until the website is updated to see the changes. +The version Im working with +Its because we have some more discussiondocumenation before there is consensus and this does not break API +I am reopening this issue as a result +updated with trunk +this is implemented for packaging but there are problems not implemented for type implementation currently uses which is configured which is a reproducibility problemNeeds to use the plugins defined in the POM for finding them but this introduces more issues is this too onerous for a dependency type definition? verifying all of the plugins in the POM before building is inefficient so it would be good to be able to point directly to the plugin that has the type handlers will need to be loaded into the extension classloader so Ill do solution is a little hacky might want to take another loook at it with the artifact refactoring +the new patch looks good +updated description to what it is about +None of this code has changed in over months +John do you have any code for this or is it a feature request +I can get rid of the part if we want +Added to where ever it was neededused instead of printing stack just committed this +Log space seems to increase MB each time the feature is installed and uninstalled +Make sure that you do not have old copies of around anywhere in your classpath +HTH Ted +Well stick with the workaround until we get JBR or use the AMQP transport +this to be this fix make it into EAP +here is the message that failed to post last night i took a look at simply disabling the tests that dont work +The problem is solved via a userlog setting or turn up the level +Hi Kris can you give me some some sample code on how I can get a handle on the underlying command service and how to configure the lock? Thanks +Oh youre right +The client class is are the client and EJB source files +Scratch that after the first choice the ranges can I guess you could store ranges in header and then have each KV contain few bits for which range its coming from but that now becomes complex +The bug tracking is not such a good forum tool +Either approach certainly has its merits +Sorry should work now +Thanks for the patch Hari! +Can you provide me with a detailed setup and step by step instructions so that I can try to reproduce your issueCheersKoen +And users could disable them once the migration is done +Patch applied with thanks to Rich BolenCommitted revision +We need to think of a way around this but I dont think its critical for +I have just committed your latest patch +Can you specify these as well +I was just looking for a summary of all of the columns on the page that shows all of the groups not the the page that shows all of the projects within a group +Yes Max we added this limitation and and new option in preferences to disable revalidating unresolved Els at all +Yeah taking failover out of the picture makes this work as does taking ssl out of the picture +Please close if its no longer relevantneeded or bring it up to date if you intend to work on it +Listbox for choosing a region list box when country is region list box when country is is this now targeted to wishesThe combobox is quite useless when the height of the list does not show the entries if there are fewer than is a bug for sure +Attachment has been added with description Patch with a unit test to reproduce the is a proposed patch that fixes this issue with minimal changes +Fix applied to branch remotingx for release +Hi MirekThe seems not a problem +Some changes need to be security review +Dynamic Web Project implies structure project rootweb content +I didnt add any unit tests because they would end up being flaky as they would be checking a thing +So all issues not directly assigned to an AS release are being closed +Hi ChristianSorry the patch was not meant to be totally complete I just wanted to show the original point i was making +bq +Im getting this issue as well same stack trace as posted in the original description +Assigning this to jliu to decide the best action for this JIRA close integrate reject etc +At your comment Insert as the first result and truncate at num + +Its not expected if youre not authenticating +But you are using so you can set every necessary parameter by a method +Closing old resolved that dont have a fix version and therefore cloud queries for M release +a unified diff to patch Axis +I did some debugging in AS and the implementing class is which is in the same module +Hi MayankThanks for the contribution +But I am in the middle of completing binary distribution +runs in but fails from ant apparently the file paths arent quite done the same inside and out +this was a hard one to find. +If this is desired new enhancement tickets canbe created. +Test creates a listening socket with a backlog of +Thanks Bobby! +Thanks. +Added Serialized Transient +ResultsTests run Failures Errors Skipped +A lower level plugin could explore the tree to find specific details if it so wished however a lower level user is SOL if its all hard coded above them +This include patches forrampartrampartINSTALLrampartNEWSPlease somone apply these patchesThanks +damn but this mean that for linux users that might also trouble with the visual part will also not have codecompletionI dont understand why visual dom is relevant for the content assistant which is only for the source dom +Indeed thanks dave +At least for the package option I can tab complete it with the last package used saving the user from typing it again +The passing test has wicket before bookmarkableMozilla Firefox for Ubuntu Passes testingDEBUG session sessionsengmxzmzcwrzlxqDEBUG servlet wicketbookmarkablenull defaultDEBUG chaindefaultDEBUG call filter DEBUG loaded class from beIE Failes testingDEBUG session sessionuidhztspxahdqkpbqmhDEBUG servlet bookmarkablenull defaultDEBUG chaindefaultDEBUG call filter DEBUG call servlet defaultDEBUG Resource bookmarkablenullDEBUG Resource bookmarkablenullDEBUG uribookmarkable resourcenullDEBUG Server RESPONSE bookmarkable was under impression that the urls were like wicketpage +However I tried to see if the patch applies to branch as well since this is a blocker for that version +apps within trusted network that does not need to be paranoid about thisI disagree +Verified in ER +You missed the overlay concept though it is not meant for dev but for easy use of packaged retail libraries +Mistake sneaked into the previous patch +Assign unscheduled issues with Blocker or Critical priority to +This is now handled properly +Thanks Knut! Ive linked to to have that info handyIt is true that the issues in the two topics are different thanks for pointing that out +The ma runs a a single bundle without requiring any other +Runs sonar on slaves you Jakub thats a great news +Note truncated lines as well as garbled subject and body page in Chrome same with the introduction of Sams regexps. +Thank you for suggesting the option and reviewing the request +But since we dont use when compiling yet thats not a problem for now +the problem was the trailing slash that is present with the root node butnever with regular node paths. +looks good and tested fine. + +Moving back to for now. +Then its also consistent accross all RQ parts +thanks! +But of course follwing the standard also is a good thing +Attachment has been added with description Upgraded collections in has been added with description Updated update to missed a variable in use this patch youve come across a conundrum I faced I dont know why someone used URL for these since these are pointing at local files +are other places where weve allowed Java exceptions to propagate through Ruby for the same reasons you have Java frameworks expecting to see those exceptions come out the other side +Also would provide a nice example how to provide authenticationauthorization in SOA apps +This is not a problem and as Markus mentioned the tool does a pretty good job of this already +I think dealing with config and what not in zk needs to advance quite a bit before we can push zk into non solrcloud mode or push non solrcloud mode into zk +Reporters of these issues please verify the fixes and help close theseDoesnt look like a dup +Please reopen if more such cases are found +thanks for letting us know Jay +Taking a look at this now +On second thought this might be +unrar has been m using maven and i still get the same error when trying to unpack a rar dependencyEmbedded error No such archiver rarIs this bug really fixed +Updated issue description +Added Etag support +So please the tag so that I can commit and you can test + +Hi JukkaI found this thread because I also had a bad memory leak withUpgrading to has improved matters a lot but there still seems tobe a small leak +i would leave the doctype references in the configs as guidance for xml editors. +No unit tests since this is fixing a testI manually tested it and verified that +Thanks Shai +Not a big deal thoughSo what are your plans for this code once its at Apache? Will you continue to support and develop this code +I just committed this to branch. +Awesome! Thanks Rafal +Thanks for the review +This guide was never meant to be correct for CR it has been updated for CR to inlcude the changes present in the link given and these docs are avaliable now +Hi everybodyWhat about this one anybody interested +Heres a new patch which fixes the problem +Under concurrency wouldnt you effectively end up with batch mode +Cannot reproduce bug either +mavenrepositoryplexusjars SUCCESS Offset Length PM READ CHomeCaleb +Reopened to add a left margin for trend icons so that theyre not too close to the text +it is an issue for both the c and java clients. +Hi All this is a very exciting feature and Im trying to apply it on our systemIve tried patching on and on the trunk version but both give me build errorsAny suggestions on how I can build or latest with this patchMany Thanks updated patch that works with the latest trunk +Need to check the configuration because the cannot be used for sequencing its only for file check in storageAlso there is already a DB jira for this issue +We just need to check if that is enough and we should decide where to put it in the distro and how to name it +Fixed in the current code base please check whether it is working for you. +changed title to reflect the actual problem + +Alexey please verify +Tested the migration from a +On the contrary Ive read every single thing youve written some with puzzlement and some with the thoughts of Oh its undocumented +Bulk close for Solr +The verbosity level is held separately in the as a global static and also in the Ant project associated with the instance +Happy to help them debug once they start down that path +As we were currently already hosting the library I updated our repository to include the jar +Attaching a version of with changes incorporated +Please create a separate ticket to implement using the Apache CLI library +I was going to cherry pick certain bits of your pull request to apply however if you could put the test case into a separate pull request that would save me a lot of trouble. +wickets component hierarchy is fully dynamic with or without the component resolver because you can swap components on the fly something you cannot do in tapestrywhat you are talking about is letting the markup drive the component creation instead of java which is not really something that we advocate +Its not a critical fix either so we could probably do wo it in +We should check whether the component should benefit from this as its likely dont need to retain messages + +In current patch listeners are also added at level +To clarify this is already fixed in +Easy fix I wonder why havent seen this before +Sean Brent etcWith signature gets verified +The only means of filtering is by including or excluding the feed itselfSo it sounds like what you need for this case is NOT to understand a file but rather to permit discovered to be filtered in some way +Actually AS should be default in jboss tools and EAP in JBDS +At least the documentation should be updated to inform the users to expose the backing engine as a service +please committed to trunk. +The following test case shows the bug +looks good to me +And I cant see any really bad side effects to having extra types that have been deleted around +The use of the operator for appending strings is inefficient +The GOP fix has been ported to GKOP. +Linked to and slating for the general backlog then +As noted above there is no default value for this +too on moving this out to because it runs too long +Paul Im confused +Closing all resolved tickets from or older +Please open a new jira linked to this onebq +Committed fix in trunk revision Committed fix in revision close after release of +We can apply that to branch and trunk +Added a jndi entry for a connection factory and add the and to the excludes lists +If you can the urls such that all modules get deployed to of the top project everything should workAs a fix for the site plugin the only solution I can see for this particular problem is to extract the actual url by looping over all modules before processing anything +If this behavior can not be generalized then we can close this task item +Correction There are still two important parameters in the policy file viz + +Its a bug +Thanks David your code looks much better +srcjavaorgapachesolrrequest +The fixes for and should fix this problem too +Attached the patch please review +Georgeis this issue still relevant +Added the repro attached flag to indicate that the steps to reproduce are provided in the bug description +OK +If you wipe the geronimo plugins out of your local maven repo you need to remove the +Thanks for providing this patch and the interesting analysis +I can confirm the latest patch fixes the Solr issue that prompted this +This happens when two sessions addmodify the same reference property at the same timeYour patch effectively removes the patch for and the does not fail +OK youre probably right +Correct the name of the is a fix but may break existing uses of this component because it will change the generated name to have the suffix in this defaulting case +Here is one reply from Randy ONeilI like the exception +I also replaced Iterators by for loops by and tabs by spaces +Sounds fine to me +Oleg and Sebb for your valued input +Patch showing first pass at the changes + Roman +Im told this is a cluster and have suggested an upgrade +So I attached the jar here +The fix will cache the metrics and flush every calls or the closed +This will be fixed in the code new behavior of the AI portlet is to not remove platforms from the queue +This was done to avoid an if branch in the earlier Mapper class +With instance and concurrent clients it happened times +I tested against trunk which is of course fixed +Do we have enough for Alpha +Thanks for the exposition +OK Ive managed to reproduce this in a straight lucene testcaseIm doing further verification and will open up a Lucene bug shortly +Rejecting for CP +will upload another one soon +Heres what the site looks like for me +FYI the is moved to mina in Camel we should have different policy to deal with this kind of issue +See for all progress related to this issue +Portal support is a priority for JSF and so should be covered by that +assign to Daisy review +TP issues now fixed +Shouldnt these matchnit the build target source should be called or something +Thank youIll work on proper fix for and supply necessary changes for this issue as well. +Thats what I thougth +Sending mail to jira works +if you agree with this can you please undertake this +This should be mentioned in documentation +Oops committed missing folderCould you also integrate the example i committed while youre at it +This was addressed by is already part of +This patch solves the problem for me +Patch v adds check for state against null +I think it all makes sense +already does that and all the others should follow +This patch causes to fail +We are having the same exact issue +Interesting +Simple nullempty check required +Right now we are not using tabs in our application so I cant spare the time +Other samples also work fine +Sergey was the patch applied? +I dont know is it really worth keeping the old cache entire rows behavior around if we have something more sophisticated +Why cant you do this in the tomcat integration +Do you identify a notion of user roles or groups in your requiremnets +committed to trunk and +Good stuff +This is a nice to have but for at the earliest +Hi Joy Xu can you check your configuration files like or? We can help you if you can share these configuration files + +The Antlr plugin is pretty simple and just works +John See +Translated missing zh labels for trunk +Ah right abort is new in I think the tests you have for trunk are fine for now +Markup generated by the Pelt theme could look like this capture. +the last block recovery id is greater than the new reported recovery id commit exits without making any change to the last block + Will run tests +Also adds flags field which is needed for fencing +My expectation was that I would be able to use theoutput of the generator for further processing in a Cocoon pipelineit was the fact that I could serialize it as XML but couldnt transform it thatmade it seems like a bug +I can see the diff in the doc tree but I cannot see the change in the link that you mention +Tyronne please read our previous comments EJB project will not be replaced as its only purpose was as a placeholder until WTP created an appropriate project type +It looks like this issue is still present in the version included in with ER +Reviewed the git commit history and that looks complete +Created a version that had the missing equals and hashCode methods validated the rar with the script and it validated OK +See +should be free of UI elements and UI dependencies +Hi David in fact if my understanding is correct if we manage to fix then nothing else will have to be done to get? So perhaps we can consider as a duplication of +Ive just checked this in +Without the patch mtom fails +the test fails because is hangs when trying to read from more than one stream at a time +Incorporated all comments +Also fixed the remaining TODOWith this code we break the already broken code but make it similar to trunk so migration is easier again +v patch attached with the typo fixed will commit shortly +testing now to commit backport to x +Could you describe how to reproduce this +The defaults for entity and collection can be overridden on a per entitycollection basis using the following format above configuration indicates that for Person and Address entities and collection a cache named with configuration will be used +The driver has been upgraded +Additional cleanup of objects by setting them to null to see during the execution of more memory is freed by garbage collection +rev no more NPE when messing around with large lists of users roles for alert notifications +Could you please show the intermediate interface you came up with to diagnose the error message in a bi more detail +Pull request applied thanks Jozef +Once that is done we can think of as well + +Ill try to take a look soon +Attached with Teds comment +The jar files I propose to put in directory libendorsed are also added to the classpath +But I couldnt manage inheriting a base bundle list nor creating partialbundlelist +Hi SimonIs there a target release for this jira +Thanks Thomas! +Exposing this tuning parameter could really allow someone to damage their instance +If the errors come back Ill open a new ticketSorry for any confusion +The attached patch removes client which behave differently on Windows and systems from the test jdbcapi +Arnaud its a hit +resolving this issue as uses INNODB for all create table statements. +This emulates the automatic scaling with heartbeat interval that was suggested +rebased and committed +The package should probably be public evolving +If you have a watch on a node it will only be triggered once so it was being triggered with a event in the test case most of the time because of how the threads overlap +Looks good +I tidied the new test a bit putting it into the current class and using UTF as the foreign class instead of a example WritableHopefully Hudson will find this patch more digestible I followed the submission guidelines except for building the tar which proved cumbersome on +Sorry my fault +I gather the branch is again open for checkinsCommitted patch to doc branch at revision +This can be done by copy the property files from standaloneconfiguration or domainconfiguration +Currently there is no working implementation of distributed userbased collaborative filtering in any version of mahout unfortunately. +We are silently loosing messages if a queue is deleted via the console and the application continues to run +Fixingminor errors both is scripts readme updated expected failures updated version of patch with fixed new lines in At revision to buildtestbranchesPlease verify it is applied as expected. +By moving it out to a separated file your life will be much easier +Triaged July Assigned normal urgency marked as Regression Test Failure. +Thanks Thejas +writercache packages? for moving both LRU and CLO under the +I can follow directions +fixed in r of branchr of of does the fix work what config file is affected? Will people who keep their configuration files separate need to do anything when they upgrade +A separate bug please preferably with a patch attached +In old days when everyone subclassed we could have set the property in there but its junit world nowIf you think this ok Ill commit it +Even better if we could somehow capture the header of HTTP resources but Im not sure if Javas standard libraries allow that. +Porting to general like reopen get should default to return null in +Amended to add model management and validation support for sections are clearly marked with an amendment header like this one Custom Addition and an amendment footer like this one End of Custom Addition + +I applied the latest patch OK but there is no associated test case +Apologies for bringing this up again but at least it confirms that the fix worksThanks very much for the help +the patch looks good +Committed patch with revision +JS Import will pick up from the targetclasses folderJetty is automatically configured to pick up from the targetclasses folder as changes are made +Install the attached to this jira into LOCALHOME + +something like makes it very explicit that a user is using these classes +I am also uploading an updated patch in case you need it to revert back to the original one +Look at +Ill take a look at thoseI forgot to say +Robin is fixing some other things w NB so hes going to take this. +Added the deprecated patch for am now testing for the parent query handler index +So it can be removed from the private constructor you are introducingbq +Tested on both trunk and branches patch fixes and in both cases code looks good but maybe move to the to make it reusable +Hi TedI ran sh but no failures +So this kind of comparison is bound to fail whenever JSF is upgraded +Then the struts poms that refer to this would have to change as well +Do you think this is a timing issue? Maybe the keys are not installed on the remote machine when jclouds tries to connect using SSH +I suggest to remove the code duplication in the various launchers resolving before resolving this issue +The first patch for this issue reduced the amount of generated code needed for maintaining the execution count and row count information used to determine whether its time to check if the plan is stale +Backported in at r +Rob reading this JIRA again its not clear to me whether this is something that needs to be changed in the plugin or the meclipse integration +So any ops coming in over the http interface would not be able to acquire the lock and change anything anyway +so we should close this. +This is greatI tried to apply your patch but it failed +Removed the demo code from this patch +It is correct generated +Roller works fine with +Thank you guys +In treat the RTF variable much more like a template that has to count its variables at and link and unlink when the lazy variable is resolved +Patch including fix and test case attached +It is still a bug +But It is not working properly +Committed with revision Committed with revision +Right now I dont see any reasons to not add the access rules and as you have noticed the issue with the access rules is being brought up in the mailing list quite if is using compilerId eclipse it should be possible to implement mapping of settings from Maven plugin configuration declared in compilerArguments element back to JDT so we can replicate the exact configuration and can consistently disable or enable the access rules in both Eclipse and Maven CLI +So I reverted a change again in a subsequent commit. +Please close this issue Nick +Remove the backup file for the hsqldb in the db directoryand restart tomcat. +I am able to do this +Could create the XB test case for this +Attaching the result of the decompile and verify step +The reason that these methods exist is to be able to handle and treat them as equal +Closing as wont fix +Oh sorry I forgot the patch is for is a minor problem in the test method setup constructs a virtual table using a sql stringselect +Most got closed +I believe I found a way for me to POST many messages on the same connection but with varying correlation +Parsing and formatting date values is now done manually and works on JDK too. +Thanks for catching that. +Roman would it make sense to use a separate hcatalog user thats a part of the hive group instead +a patch which contains the modifications and the test file It overlaps with my patch to but I wanted to make both of themThe test email contains a newsletter from CNET +Please have a look +The test infrastructure has had a huge upgrade since years ago +Where can I access the snapshot releaseThanks does in eba meanCould you give me a simplified example of the bundle manifests and a description of the steps needed to reproduce the issue? If so I will look into it +Same thing should be done if Delete has not checked the delete from hard drive +Ill work on getting design document together +JAVAOPTS can also be added there +FYI anyone interested in this project should think about watching which is the jira well use to actually execute the merge +Closing issues which have been released. +Committed to trunk +Hi Rick If you still have the test directory for this failure could you please attach it the to the jira entry +Thanks Arup +This svn in is too much dependent on your local installation of svn tools +Let me knowAgain thanks for reviewing +This issue is not about replacing and its configuration files +I found no issues with that +So it will be better if this number doesnt changeI did some code diving into CLHM it looks like it annotates the cache entry with the and uses that in remove or replace +These are probably all obsolete +It is affected by my file which I am attaching +All tests pass +So in SCM its going to have the same name +Here is the patch for the same +All true tree dont working with RI and state saving +master pull requests merged +Hi Jones Ahk I dont have any good documentation for cluster configuration +so closing the bug +Patch for handling failover +Im tempted to mark this resolved as invalid +Im planning to disable it and do all the indexing manually not really a good solutionOh I should add this is with Grails and Searchable the workaround listed above in place objects are never added to the index except when reloading the app from scratch even with mirroring enabled +We always have the option to back out +Dont know why its there probably introduced by new systemvm templateWould fix the to trim that character +Re downside quite right +The reason for the strange test failures could be due to two issues +Looks good will commit +The exception still there to be fixed in jaffl +I moved the class from the to the project some days ago +Attached a patch that just does this +I tested with Mono VS Win and OSX +committed the latest archetypes into so now everything apart from the karaf archetypes have Ioannis! Assigning to me so I can include the new archetypes in IDE when they are built +Thank you for comment +jiri what is the usecase for this? you want to install them during some integration testing or +Kevin heres the boot log for the same instance +it works fine! +the latest patch looks good to me +Ive just fixed this feel free to give a current snapshot a try! tested from CVS and its working now! Thanks a lot! Hey arent you supposed to be in London and Rome? I hope youre not solving Spring bugs on holiday! +So I think repair was blocking minorsrepair should be on its own executor now so it shouldnt block minorsbq +Good point about OBR I will also remove the lines since they seem to cause more pain than relief +Added new plugin to plugin repo in +I see that at rc the caps doc defines the element in the internal dtd +a +crossing fingers +Thats just the same as to not really +This must be something elseno this is a misconception +patch applied +fixed in the new release +I can do the latter if needed. +to reproduce the failure +Perhaps related toUnfortunately that one is only fixable in trunk +Obviously there is a limit on the cacheThe query plan cache is indeed hogging my heap space but most likely the queriesthemselves are very large +is used only on a temporary basis and as such no leaks should occur +Please check that it was applied as expected +Can you provide some additional context as to what your test was doing and when this exception occurred +The patch fixing the problem with debug info +This sounds very great +Id like to discussreview the test framework for replication specifically for scenarios +Burr please raise a JBNADM issue and link it to this one +can it be divided in ant projects as well? about managedutil as far as i can see it is jmx stuff +This should be addressed +bulk close of resolved issues. +Verified in +Okay here is an updated patch which uses your interfaces +repatched it +This issue is closed now +with webflow and Spring RELEASE works checkData and checkData +Hi guys can someone explain this to me in simple terms for the release notes please? Thanks. +Ive fixed up the and wiki articles to address this +Okay one more patch for review +Is it not public +compareTo in +The Jasper dependencies seem to be gone but not the Castor in in rev issue was in fact completed in previously rev broke the console +You emailed the community for input we created a wiki explaining the options +This patch removes Invocation too +Committed to sqoop +I tried multiple times and did not received any error +Committed to trunk. +It looks right but I never seem to remember all the details without it breaking at least once +Which file are messages supposed to be added to +been offered chriss solr one +in the end is probably not the right place since it is created to frequently +by new CMS site. +Verified on EAP for for release notes MarekWhat was done to prevent the deadlock? +However this is not the case for schema +Initial patch containing a implementation for bytes sorted +There is no claim to cover all protocols +I have changed the to point at the internal URL +hi felix In this case I would strongly suggest to modifiy the patch to comply with +If they think that the change doesnt belong there then well need to use a catalog here +What value does the add here? It would be simpler to just put keysvalues in a buffer and iterate over them in place +Tens ra Gill tamb funciona en laltra direcci. +Ive relocated the admonition I originally put into the Bridge Config section into each of the and mbean sections in EAP trunk +Thanks for takingthe time to include that information Dag +The version is + More documentation added about +close after release of +Committed to as well +Please review your items if not ready for please move them out to future release +is a link to the Allocator Implementation Guide from the Developers Guide +The implementation presented here could be improved +Removes the need for in multicore and still allows for the singlecoremulticore merge in the manner explored aboveI dont think the back compat part of it would be very difficult at all either +Applied the change in Rev +I tried compiling with the following command nant tHowever it continued to build all target frameworks +I see thanks for the hint +Since we require bash we can cheat a bit +Lets see if Werner prefers to move this issue to XML component or create a seperate issue for +Console logger will remain removed in domain config for EAP. +Thats the reason why page breaks are not correctKind regardsPeter Herweg +Thinking more of the solution earlier suggested I feel convinced that it is a correct solution +Ive made changes so that deserializersobject inspectors are initialized with the union of the table properties and partition properties with partition properties overriding table properties +And btw how do you plan to handle suites that also use the mechanism? You need to be able to includeexclude suites from the Buildr build all the same +Much appreciated +it would be good to log an exception when such memory problems occur. +Issues with deploying Mongo +I guess that also means that the column in the table is okay +is still strange because the whole change was one single commit +It seems that for some out of band headers on the output side we are processing them even without the flag set +Please reopen if this is not the case +good +Patch has been applied +Git doesnt like empty folders +What is the branch for +and rerun the testThank you +Username juanmartinez +Updated patch for +The second task would be to add the and implementation as per your original patch to implement the current parsing behaviour and make it pluggableColm +Presumably we can do that now by putting the delegate key twice in the flow constraints +bq +already in docs +done in aabfaedfbefcbad +Now that I have finished the Citations Plugin I think there is a bit of misunderstanding on the use of the see element and citeIt looks to me like the see element is in fact to be used that is provide a href link to an anchor in the same Glossary document +We cant have and out of lineMy preferred fix would be to revert the invalid change to +If this file exists it contains information about an emergency login +Would the scripts all work on windows including hard links +any pmc that didnt comply didnt get the new committee group on changeover +I can reproduce the error in under I found that most of the fds left open were sockets rather than files +So most of what you suggest about log formats I think should be couched in the terms by default right +bq +and contain fixes to the javamail spec files that are required for the NNTP support to function properly. +I also created a new test case to suit which is restricted to justPlease have a look +Committed at subversion revision Touches the following fileM javatestingorgapachederbyTestingjunit +what is the status of this +I agree it is fixed by +Hi danI thought about that but than again i figured there might be places where we wouldnt want to take a hit of parsing dom elements to Objects +okidoki +I retested with and the problem is still thereShould this be reopened? for investigation +While weve released a few independent modules as teasers getting the site documentation done is the main reasoning holding back our general release and any formal announcements +That way the failure screen can print out why it failed for each type of service +Thanks. +I think this is just an issue that the log runs out of a new Journal abstraction API that is implemented by the HOWL logger and a custom logger +I think we solved this issue +I dont see Technorati playing the major role it used to when this item was opened in . +lgtm +It wouldnt be particularly helpful honestly as some of our parameters are Longs +EUT jdtdebug passes with an exception for test test which fails with an interactive message +Other signatures with similar objects appear to work +I did not see any of them on the portal page +Just marking resolved as this one is in the +patch for the Single Row Select +Fixed trunk change is restricted to the HA module +Removed need to use the new CL enum instead of ints +Weve done that mistake before with the aws maybe its time to refactor it somehow +please review and the patch +Just checkpointing so others can see the tests we plan to add +Brings up to the trainer to be an brings up to trunk +The attachment fixes the issue +Thanks for fixing this quickly! We do not have access to the snapshot repositories through our nexus servers +Updated with RC and comparison against +Nulls wouldnt be passed for regular maybe some improvements are possible for this use case +yup that does it +It was released with the fix in +I am unassigning myself from this bug in case someone else wants to add JDBC delete support for other resultset types +I have no immediate patch for this but Andy wrote at some point However it would be easy to change the global to being a lookup in the context and set a variable to control abort behavior +He only wanted it registered here so he would not forget it +Yes I was looking for this option the other day +This may also be a good opportunity to add some Tool Options to the Axios tools I am also available to help with this +Apologies if i am +That seems minor enough to wait for but Gavin can put it back on the roadmap if he wants +Recreate or close +The error is due to usage of collection classes +Do you like better than throwing a +We could do a third for the common proto definitions if necessaryI believe that YARN does memory enforcement in the same way whether or not the LCE is used +bump the fix version since all my doc work is on trunk adrian +Ill try to spend some time on this next week +I could induce by lowering the heap size in a manner similar to issue +I love it when patches apply wno conflicts +Scott if I dont hear otherwise Ill change the test back to use a random seed fix the checkstyle warning and commit this +This updated version of the patch also allows quotes in the Ajax response +I will start testing this as soon as I can figure out how to implement your fix +No such problems on bit icc on Windows +Sorry I forgot to mention your first questionI have never seen the on Linux +Ill just wait til things are in place +What we actually need in was a way to create and handle streams based on memory buffers +Thanks +In any case this is too complicatedPlease attach a much simplified complete stylesheet so that we can investigate +Applied in commit dceefbebdabdfca. +This patch passes all the tests has javadoc etc +This issue is ugly +this is the code but it includes a and a if you need it too please let me know investigate for is the ear of the proyect i hope yo can install it and prove itthe url should be httplocalhost only have to click in actualizar button and it renders the uploaded to reproduce bugregards +Discussion moved to +With this patch it passes always +they all work exception +I tried to reproduce it but couldnt so far +First we should be careful in our language +Download the poms from this issue into the appropriate directory structure +Yes I tested this and confirm that it works +No need to fix in not critical +The same behavior is reproduced for ordinary hdataTable +This code is missing +Committed revision . +Attaching a new patch anyway +Its against trunk but it looks like mine was out of sync +But certainly the point is taken that bad things can happen with abuse +I ran a couple simple tests with the patch and I am no longer seeing the exceptions +some jsp pages still uses webwork in it +We have copyright notices in the NOTICE file I think we are okay here so I will close this issue. +Wouldnt make sense for the classifier to be h as will eventually be the stable version +This issue is related to copypaste within highlighted block which is improved in SBS and will be in production tomorrow. +The multiplexer sar has been added +bq +Nothing this design seems better than limiting physical interface name above commit +I was able to successfully deploy to WLS using nightly build +from me +Thanks Aaron I pulled this into. +It worked +simple patch for my projectmay could help +v fixed the javadoc warnings and avoid an unnecessary byte array new +I am not sure how the build system check for ant tar failed +This patch adressed only enhancer classpath issues +I think we should introduce the variables +thing I would appreciate feedback on was regarding the classloader to use which I am picking up from at the moment +Can this be closed +rhotrod p cacheconfig +This is fixed long time ago in. +I have no idea whatsoever how to test thisAdding springAdding springAdding springAdding springAdding springsrcAdding springsrcjavaAdding springsrcjavaorgAdding springsrcjavaorgapacheAdding springsrcjavaorgapachegeronimoAdding springsrcjavaorgapachegeronimospringAdding springsrcjavaorgapachegeronimospringtransactionAdding springsrcjavaorgapachegeronimospringtransactionTransmitting file dataCommitted revision +This is clearly not a pig issue +Ive also modified the Geotools projects permissions so that all can see the svn commit tab +Fixed in CVSPlease verify and close the bug +Roald the proxy servlet always worked perfect for me +I have prepared a patch for branch of be fetcher instead of parser for component field of this issue +filter is pushed to inner branch +Wow it was really hard workPlease try this patch +Did somebody have a chance to take a look at thatThanks +Assigning to myself temporarily until I can figure out how to assign to aamonten +Trigger QA run on v +Im not sure how common our use case for the date picker and date dsl will be +Review maven supportand see where we are running into issues +Why not get rid of the trycatch completely and just let the exception propagate out of main? That would produce the correct error code as well and remove lines of code +When we do we will send another comment to ask you to test the fix +v +Hi i would like to know if this plugin will be updated to Aspectj Jars in order to compile Java Generics +Attached is a slightly simpler method to accomplish this +Its fixed. + +looks good to me. +Since the Development Phase is over please move the Jira to Resolved status +Another issue that we have to tackle before this can provide hot standby is thisWhen we close a file or add another block to a file we write OPCLOSE or OPADD in the txn log +Hmm I tried again with a total fresh ESB and this time it installed this was a glitch +Gerhard Isnt issue with global settings fixed? I didnt look at the Open button issue which should havent been a separate JIRA +I think it can be replicated by deleting the log files as they are being tailed +I cannot seem to the problem when I simplify the Java XSL and XML +Could you please elaborate +Rev fixes the immediate problemDain could you check this out +made non static +I guess its not a viable solution for most users +When running normally we add +I think better approach is to make that behavior consistent +Mass closing all issues that have been in resolved state for months or more without any feedback or update +i will commit this monster soon if no one objects +also has its own implementation of skip which doesnt use readOnly using read directly can result in loosing optimizations for some stream implementationsAlso note that the contracts are different for andFor reference heres an excerpt from Java SE Subclasses are encouraged to provide a more efficient implementation of this method +IIRC we have discussed the situation when RI throws nothing +committed rThanks Sean Timm +Given that the Alias service is gone is this still a problem? +With smaller ranges and the same input split size theres a higher chance that the split will be outside of a single virtual token range +Sample files Word document with some embedded pdfs +Yes but not for this release now unfortunately +Thanks to Paul Spencer for this fix. +This is again user error + +we need something for Beta marking as blocker since many are relying on some kind of sane container classpath being returned +Fixed on and transition all resolved issue that did not see any further comment in the last month to closed status +Im sorryIt seems to be a problem within Eclipse +A couple of last minute translations for +I though I had fixed that already but apparently not +Hi JohnThere was a little problem when the httpLocation String contained a +Hi Nick the mfInternal is an internal data structure which should not be touched from the outside should be present as the spec specifies it +My mistake Adrian thanks for catching that +However I has asked one of my team member to test and he confirmed that it is working +Jeff has already implemented web sockets support not sure if this is needed now +Whats the point of having a soft limit if it can be by the users? Would it simply trigger a warning message? I dont know if that use case is compelling enough to merit a change +Holger looking at the package names can I assume that you want this to be added to the Castor project and not the Logj project? cause in that case we need to change the package names +Believe this is fixed +The documentation for this is not present anywhere +Hey Robert +Have a look and inform me if it works for you +Even the uppermost parent pom. +What do you think +Hi WeinanI can confirm the bug fix also on my side +so pig expects anyone running a job to keep a local copy of the cluster config in sync with the real one +cqlsrcjavaPersonally Id go for number in the short term so we can add that whole directory to +Im reopening because navigation issue is still present on Struts +Then it might be a reasonable default behavior to simply leave the compounded tablecolumn name without quotes +AS trunk is now at for +Guillaume any thoughts on this +Updated patch to fix the test failure in +Cool thanks Konstantin +can you update docsexamples to use the new command +Though reading your comment makes me wonder why on earth it took you two days? Id appreciate an honest opinion here. +What about just checking the return code of the start script with binary if not successful dont create the pid file logic? +I looked at the pdf report and IMO we can do following Instead of type we should show Shipment Method description +Christian Thank you tooThis one was a big PITA for us at guvnor our workaround triggered loads of other issues + only takes seconds for me but is too slow +Thanks +Will take a look at that and open a jira if required +Jira cleanup +WS CXF the correct name should be either or +Sending srcjavaorgapacheluceneutilAdding srctestorgapacheluceneindexTransmitting file dataCommitted revision +Please add to the classpath in front of +It doesnt happen on our Linux doesnt happen in every one and notice its sec not also tested your calculation for next scheduling time which works fine although scheduling itself is wrong +Are you getting any exceptions or is it silently failing +However it would also make sense to be able to provide configurable guidance on when a rollover must be forced +Batching is disabled by default +Lets do the docs later +What are you searching for? Its not clear that the URL is correctly encodedIf you are running with something other than the included jetty server you may want to check that your servlet container is correctly set up too try running exampleexampledocs to see +Resolved as not reproducible +Browsersim is working. +b There is no +If test passes for you for commit +Added facter and df info to the host registeration info +Using curl with the basic auth workscurl basic u smxsmx httplocalhostcxfsecurecrm? REST security example just uses the basic authentication to setup the security token for cxf to think we just need to update file for it +My bad +I can switch to the Orderings tab and it will tell me that no query root has been selected yet +but i think the encoding seems not to be the right one +So I think we can resolve this issue fixed. +updated to seems to work no we cant compile with JDK anymore +Ok now I understand +Attached Maven example project that illustrates the problem +Its slightly uglier but will at least support ipv properly on Windows which is still the most common case +I can only comment on Cordova here this is not a Build issue trackerCordova has no preference starting with Cordova the preference is. +and +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +If you had two artifacts which shared a common pom the upload task would call upload of each artifact which in turn would both call upload of the pom artifactTo resolve this I modified the uploading of an artifact to be a task itself which ensures it will only be done once +All metrics are there +removed urls from wsdl file +The in the sample application doesnt have the registered +distribution +For M at least get to the bottom of why Peter cant even get to the point of being able to select a test then for M look at replacing selection with a mode that runs all of them +Resolved with commits against +What is the output that you getIm running on Windows withThe only error I get is from ECDSA it says Only named supportedbut this predates my fix +Otherwise why not do the same for SUM MIN and MAX +committed +patch to go from datanucleus to +The configure command should be +is fixed in which is what continuum trunk is using. +I actually had it this way first but then when writing the code in it felt kinda weird to get the whole only from the Merger SPI without even touching the Provider SPIFor me it made more sence to get the Provider first and then to pass it to the Merger so that you see on that there are acutally two involvedHowever I am also ok to do it the other way +gitpatchWhen you have resolved this problem run git am resolvedIf you prefer to skip this patch run git am skip insteadTo restore the original branch and stop patching run git am abort +push out to +One way to do this would be to code logic directly into the modeler that checks for a structure and guesses at srcmainjava +Provider name should be a division of Red Hat not Inc +This worked for me at least. +Sounds good to me +Just switched to the incref rather than Reader ownership change +If we cant complete the write of the snapshot then perhaps we should delete it write after learning it +All but some nits about null checking equals and speeding up the cellcodec accommodatedFix javadoc warnings too +the connections from client still dont fire the lifecycle callbacks. +I dont think this should be considered an incompatible change +Have only been able to test in Camel +Thanks for suggestions Tom! Revised patch with the unit test is attached +Likely this should be the effected version and the component should probably be plugin or something +please confirm my understanding AnimeshI went wider than that Im in the COUCHDB project and received three large summary messages from the CLOUDSTACK project + +Fixed +I guess I mixed up things here +to trunk in rev +So it seems only by calling to I can convert the string format back to the real byte and find the right region +classpathM pluginseclipse +Have you tried to create a snapshot as package developer? The buttons are there but you should see an insufficient permissions error once you try to createoverwrite a snapshot +To facilitate the evaluation +Forgot to say that pre this fix if have multiple SDK compiled for different locales or FP versions you couldnt easily tell them apart in the Flash Builder IDE +Heres a patch that implements step allowing the to login from a keytab file on startup +Thanks for explanation Chunrong +That was quickThank youSylvain +Where is this mentioned on the forums +Housecleaning. +The unit test failed seems to be unrelated so will address in separated JIRA +This is a bit of side issue but I noticed that the patch uses the same mechanism of early property evaluation as for +Only if the hlog file is corrupted cant handle it. +We should not make a codec the default thats not implemented in every language +I see you already updated the comment +We caught one of the nodes in this state in last couple of minutes +Will incorporate them soon +Hi Paulthanks for the Patch its committed in Trunk Rev Dont forget to click on the Provide Patch button easier for searches later +Ill try to take a +Problem is finally resolved +Standalone tooltip div needed at the DOMs top level +Now that the blocking bug is fixed committed at subversion revision +Now that work is well underway with AS all previous community releases are +Its that started passing that address in the it used to be that we were just passing the hostname that we were resolving before that +Updated the test classes to ensure that no temporary files are kept in the project patch still misses a single file Ill provide an updated one soon. +Not sure if its the same or not but I added your test to out test suite and it passed so I assume fixed it +Re Yep having it mature in hcat was the plan +for the patch +Lets run this by QA to see what it thinksThis basically replaces by iterating over the +I think it would be sufficient to declare the repository in the related pom no need to have a separate bundle +We can discuss whether we want to adopt the various corrections that R makes +StackSorryI attach a wrong patchI updated problem XufengI reverted the v patch from all places and then applied everywhere your v patch +HelloI refactored the code by introducing new classes to process fields withwithout annotation mapping +Both aggregatewordcount and aggregatewordhist work without exceptions +Currently I cannot supply more information but i will walk it through as i strongly want to see the packages in Alpha +However the configuration is statically generated and will cause problems as itconflicts with the dynamically generated configuration in to juddiscout is available through ESB in the SOA platform sowe will not ship the in SOA +i can be OCD too +I did it to fix the problem quickly +There should already be a change in the doc replacing the with the but there isnt anything explaining it +should be fixed +I just committed an edit that adds example config +applied in revision Anton please check it was done as patch is applied as expected thank you +You choose however to climb in your ivory tower and rule how the world should I suggest you take a look at how the world functions including all its inefficiencies and bad practices + is the release targeted for has not scheduled the ejb spi for that release +Reassigning all XJava debugger issues +How could Dparameter be used exactly +Looks good +Yes Sir! Ill take care of it +I will mark this as for bringing this up to date + +Patch with release patch is about moving get initialization in the constructor and making rest of the code to access it using class attribute +Thanks for reporting this +Generated mixes schemas and +I have committed this +This is no longer a blocker for me but if I cant find a reliable solution to this problem I might have to move the update site off of codehaus +Installing from the downloaded version in GSG updated +Pipes and Streaming support the lazyOutput flag Added test cases for both mapred and mapreduce the patch with and ant test passed on my local box with this test that timed out passed on my local machine and this test is unrelated to the patch +You must build from to see this problem +Btw theres package private new method on which allows you to create a stream to write to wa fixed size and create a bytestring without copies +Patch to fix test +Patch added +If this is really a concern what we can do is to limit the total number of calls in the server including unserved ones being served ones and pending responding ones +Hi CrhisHave you had a chance to look at this issue +Added a test +Thanks Permaine I created + +I verified that it is working on the HEAD +Thank you for your contribution +Instead if you use Sun to compile xalan everything worksfineCarlo +Attaching +When the update has been made through the there is no interception point in the Teiid code that can do needed modifications to VDB +Please correct me if I am wrong +Please verify that this patch fixes all of your problems +GEOS said +Having the website with the project seems nice but I havent seen a projects doing it this way with success +Ill also correct the comment so its formatted right +is another small issue related with this bug +Can you test the patch +Lukas You have already created. +Looks good +This was resolved in the process of updating +Reopening +Great catch Martin! This is fixed in master and the bugfix branch +Committed to trunk +patch for to. +Complete log showing threadstate when a JMS route is shut down +Logged In YES useridOld version with problem happening outside of and no more news +changed the default limit to MB. +Ill see if I can do one of these in that codeErick +This should be a blocker +Could someone please provide information so that this can get documented? Thanks +Would be nice to have a more groovy implementation of this. +Sorry JIRA button issues +You can then connect via Eclipse or Netbeans remote debug option +An excellent patch BTWI have some minor comments +We could do set but I cant imagine it will really make much difference given the small number of items that we typically would expect to be elevated +I would like to bump this issue +Press the wrong button sorry +Id like to know why this is correct behavior +Not sure if well be able to fix this or when as we dont have access to this particular platform anymore +This configuration provides lots of flexibility for a project +Currently moves the checksum down to fill the gap when a partial packet needs to be sent +Heres the output using the snapshot +Ive replicated the issue in my environment under Windows XP SP +It should beFor beta we will just have a Update settings from button or similar so users at least can work around it +Changes look good to meWith respect to downgrade issue I wonder if it is necesary +Is the quick fix for Guvnor +Im not sure where documentation for the new parameters belongs +Alejandro I just pick up the task recently +I have attached patch to fix this +Rewrote components for better components for better components for better components for better managed bean resource for the sample managed bean example for sample comments regarding managed zipped file was attached last make it easier for others have packaged things into two a fix to the jsp sample fix better fix with better implementation later on +At the recent contributor meet up the topic of making the assignment manager pluggable was discussed +Marek can you create a new issue to fix this properly +Ive attached new test with flow is the idea how Ive fixed the defect +This shows that close was not being done so wasmodified to call close if Rails didnt do so +it would be more clear if observe took a object instead of a CF to get a serializer fromnit would be more accurately named SSTM deserialize versioning logic would be clearer if it were all in instead of split between that and open +Sorry for the typo the bug was +Please review this patch + +Hi +Looks like from though that OSX doesnt have this option think the same workaround is acceptable here +Oh I mixed it up with +I happened to fall on the code in JDT that does this and thought I would port it over to +This means the token the node is moving to will need to be included in the state +It should be a simple request param to make it async +if a sub query doesnt want to count as a match it shouldnt matchhaving the notion of a wrapper query like BQ that cares about a minScore from subqueries in order for it to count a doc as a match would be something new. +Ok Ill ask around +I think this was fixed in? +bq +This is a documented limitation of the current versions +Unfortunately I have at present neither the time nor probably the expertise to implement itBruce +I believe the SCA Spring Specs will address this issueIn reality the SCA properties does not support the direct use of primitive types like int short boolean etcThe SCA properties are designed to deal with JAXB data types which always points to a classimplementationFor example SCA property for integer should point to class +done +It might be worth auditing DFS to make sure we didnt miss anything elseHi AndrewI am sorry to hear that your patch causes a serious bug in which is a very important client API +committed on steffkes behalf since a release seems imminentCommitted revision Committed revision +patch looks good and it is great to have a testcaseOne thought I wonder if it makes sense to have the originalExpiration property omitted from a message copy +Rebased and merged into the master branch. +I would definitely agree with Benson to use the wsdl and generate jaxwsjaxb clients for this +oh I think this particular thing is different +could you provide additional input showing that the proposed fix is adequate to address this problem doesnt introduce regressions etcThe short answer is I cantAt the time I looked at the source found a spot that seemedreasonable to fix the bug and gave you the patch + fix unit test failures +Trivial patch attached +However it seems the shutdown of the plugin tries to shutdown terminate all other processes that are running and so it probably is related to the shutdown server bug youre seeingEither way Ive seen it dozens of times but the process always seems to terminate afterwards so Im always confused by it +Any comments on patch other than missing Javadoc? If not I can write a summary for now +in which case I am suggesting we put together a package for that purpose +So on addendum +The hope is usually small number of row table performance does not matter +This wont solve the problem for things like you want to call a method or access a property on the you still think that we need a feature that makes all references implicitly quiet lets discuss on the lists +In fact maybe thatd be a good syntax to support for all parameters too +will do +One should be allowed to have two coverages with the same name in different workspaces +BillDo you remember which version of the patch applies to? Ive tried the last few versions on github and Im getting bad code +bulk defer of open issues to + could be a task too +And as always upgrade your Camel version to see if its fixed eg Camel is an old release +Once a new release is out Ill update the bundled version in lib +Pull it back in if you think different. +committed thanks Todd! +Should I delete that +Checked the change in so I can scour the wiki for required changes +This is CLI related +Merged +All the commits for this issue went into +My original issue is fixed so marking it as closed. +I have changed the libraries in order the project example to work on EAP EAP and EAP as described in the previous comment +This is fixed by introducing sync blades status API in +This seems to work but +Fix against branchPatch not intended for Trunk +otherwise it might be affected by the slow down of the bookies. +for antrun +Any objections +Values in the super class are copied downThe s resolveInheritance methods do the work +You can verify that using the or Adobe Acrobat +I am inclined to close this issue baring no objections +what announcement should jira generate +I will commit your results after a quickyou are right there is something not working in the script +Committed to and to trunk +Considering this resolved as it is consistent with upstream +LDAP groups created and all members added to the unix and committee groups +Rebased against like exampleshadoopwordcount needs to be updated as well +Just curious as why we would keep it. +to deliver a configuration solution that supported local JMS transactions +Attach patch for branch +Also add a class comment saying what your new class does +Admittedly there are tradeoffs to the approach used in as outlined in previous comments +Aaron you are right +Please find the attached patch after incorporating the review comments +I think it should stay open until a release containing the fix is made +I have attached the patches for Message Queue Component for Nuvem which includes the Nuvem API and Nuvem Amazon SQS that according to Jeans comment and also include the implementation of the interface to GAE Task Queue and the test cases +Ive integrated your junit tests thanks a lot +Jboss AS works fine +The new class and all its public members should have informative javadoc comments +This is true for New Internal Note and New Email as errors do not appear anymore +hod allocate d homeusertest n t homeusertempINFO Cluster Id INFO Cleaning up cluster id as cluster could not be allocatedCRITICAL Cannot allocate cluster homeusertestThe specified tarball here does not exist +In fact on looking at it a little more we should copy the format that the old one had just incase someone had a script screen scraping the master ui web pageIll get that patch up in a little bit +It is a helper tag for layout management only +a second thought is that do you think will it be better if we update the ID to the in the second screen of the wizard? Sorry it sounds like I am a bit picky here +Aleda do you have a full stack trace for this one? Or can you enable verbose exceptions on that server +Stefan reports it is still not working for him but has now attached a sample project +Can you please elaborate further +fixed in subversion repository as of r for trunk and r for branch issue as it was included in version which has been released +Duplicate of +just remember that the timeout you set may not be the timeout you get and sometimes you may have to wait longer +If someone can review the patch it would be greatThanks +adding a separate if for type BINARY and creating an Stream should do the trick. +looks more fault tolerant approach +SVN revision . +True overlooked that I guess you can close the issue then +The test case is insufficient and should be changed to perform that checking +Should be expended +Thanks for the contribution Roman +I thought it would be better to also handle it maybe relative +For manufacturing maybe a warehouse +not yet +Ill try to test your solution tomorrow at the end of the day as well +These are licensed versions +Tbh I think the idea was to include something like this into lang as there might be a use for it when doing simple calculations +Syncd with trunk after has been committedThe first test case is failing intermittently because all datanodes are bad diagnostics +Add more tests +Fixed in Rev +Have a rd for another example we are adding files which can be used as an alternative tothe mvn pom files +Added applicationscontenttemplatedocbookwebhelptemplatecommonimages +Closing released. +If you invoke the welcome page the javascript error shows up +So the current patch only handles +Mumble yet this is happening only on GS with and the same jdk it seems to be fine +We will +I also couldnt reproduce +I also know that in time all libraries have a tendency to grow larger so Kb saved today will translate into even larger savings as the years progress +Ive managed to trigger the Comparison Method violates its general contract on Java by adding random items but only once +Annotations andor taglets may be a great tools to help implement this +Tested with snapshot version of Metamer +on moving to anyway +Reads and writes are handled gracefully when operations come in after session expiration +A quick question if you dont mind +is this a duplicate of +The Hadoop QA bot cant test this out since its on a branch +Attaching correct version of basicrefactor +Assuming lack of comments means the change is OK. +Reworded by Dennis comment pls update put on stage resolve and assign to me +Fixed for the stuffWaiting for content to make it to the Portal before pushing the on the fusesource site +Ive added some debug logging to and to provide debug logging of how components are resolved and to show which component creates endpoint +Done +also happens on a local installation of jira +Oh this one is probably the first issue of the lot that I need to be fixed +I fixed this for its already fixed in trunk +It is relatively small patch appreciate if someone can review it +Where is the XSL document to match with the provided XML? Brian serializer code has been since this bug was reportedCould you please verify that it is still a problem withRegardsBrian this bug down +can you do it before the release? I dont think this patch will break api hope to have some time this you set a date for yet +Removing the plugin doesnt help and I dont see how it would since the packaging being pom it would never copy test +Thank you very much for fixing this +Interesting to read back through stuff from the past +It basically compares the maxPath value with focusRow value in most of the cases +Not seen in our regression tests since CEST +What is the position with this one? Im in the middle of updating the wiki and there was reference to this issue +Nope the tests you mention are not affected by the extensive tests flag the builds work for me the tests you mention pass in Eclipse too regardless of me running them with Java or Java Is the build failure OSX specific? Windows specific +Works within ant task. + +Btw the Forge Demos project will be restarted some time soon and that we should all capture those kind of demos in this project for future reference +The consumer has no easy way to know what happened at this time hence the rather crappy error message. +You did however begin to make a point of sorts Markus any thoughts now that this one has had a bit of time to settle in. +and probably and +Ive just committed this to trunk +good ve committed this to trunk and branch +thx Rob for the feedbackAgreed with your comment will change before commits reviewing this +This should also include support for based distributed coordination again without the tie to WS +Fixes bug in the new excess imports +Seems like the content you refer to appears after the document root which is a well formness error +the patch looks good +As a related point it would also be good if we could set up another Hudson config running tests against +Yea its unrelated +This bug is fixed in the Cocoon CVS +on me for not even seeing that v wasnt collection memtables +Testing in progress +This has already been fixed in trunk and will be part of the release but thanks for reporting it +Remove the TODO added in the ZabTest too please +I am noticing some other issues +But even it you have not checked Publish Web service in earlier screen in the wizard it still forces you to start the server in Eclipse +Icon and the other stuff added and tested in devel environment. +i will provide +This patch is an example patch +I have committed the rest of the changes with revision +Attached patch with test that fails on old code when run alone and runs OK with the proposed code change +This issue should now really be about converting and into coprocessors at some point +This is based on the old M runtime +Attached patch overloads all buffer and toBuffered methods with a size parameter to specify the desired buffer size +The answer is yes +It took me a long time to realize that the solution to most maven problems is to clear the local cache. +Other than a few tweaks to enable testing its a change for trunk +It looks like its a convention that makes sense to follow +Suggested solution +In order to implement the support for wrapper style mapping of method parameters in JAXWS jaxbri needs to support unwrappingIs there any plans to implement this feature for jaxbri +Id prefer to keep BJ comments but now the issues are not related anymore +Thanks a lot for the explanation Arun +Thanks +Since overloaded constructures where used I decided to stick with that but generally agree that properly named static copy methods actually would be better +thats not too automatedWhat do you think +The CENTRAL question is do we provide the greatest portability by a lowest common denominator bash script or use Perl script? When Ive built Perl in the past it goes through an extensive check to see the capabilities of the underlying system +I committed the patch toThank you Arpit +This way of notation is bad to read but have some performance advanteages +So is the workspace resolver setting the revision asHowever turning this off is not an acceptable workaround because I also have dependent projects in my workspace that need to be resolved as well +This patch makes changes to the public interface +example test using mockito +Ill let you know once Im officially added as a have added the initialization in the create method of the rather than in the that is used by the palette +However I find it hard to do totally clean The is used throughout Lucene but is hardwired for specific types of Terms +Committed to branch as r +Committed revision +Closing for +Btw Im taking a tokenstream break today +Our QE team has manually verified the patch by running tests on a cluster with old memory configuration and verifying that the number of WARN messages has come down drastically +I have committed this to which contains libhdfs +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Removed dead links from the original bug report +Note that this is only for JDK only +Single line fix +BTW how hard it is to support CREATE EXTERNAL TABLE? If not we might want to add that feature in as well +When do you expect to be released +Hope that helps +I started to use couchevent from davisp but stopped at some point since I wasnt sure it was the correct solution +example in rThanks Gilad +Votes please +Just to clarify Im not thrilled with the idea of a general feature that requires people to enter internal IDs to do what they want +The problem is in we skip the entire tuple in the value if we use one field of the tuple as join key +Add functionality to add NLP annotations also to the part so that other engines that prefer to use the NLP API can consume them without accessing the RDF enhancement are now added to the content part +Ive investigated carefully the two problem cases indeed work differently on Windows and Linux both on RI and HarmonySo I suggest to remove the two problem lines altogether and close the issue as fixedTim do you agree +Low risk thing obviously +Commited Your patch as revision thanks! Would You please verify that the latest svn trunk of lognet now compiles with mono +OR the storage state as NOTFORMATTED when this problem comes but this will be a problem if it is corrupted manually or through some program then this solution will not be appropriate as this will format the datadir +wsdl and +to commit this then +serverdefaultdeploypluginslibis unsigned +Ive also tried server OS those as before plus OS ve tested against and I can confirm that the problem has been again for the swift resolution +Committed patch a to trunk with revision Backported to and with revisions and This concludes my work on this issue +Database corruption can result anytime different systems boot the same Derby database +NoI see that this issue is closed +PS Thanks muchly for the submission though Siamak +Ive implemented that so please ignore the current patch +This patch includes in the distribution +Deferring to +i just had to comment out the guideClasse beanmappingthanksdims +So this task is finally complete +Patch blessed with blessed with ASL +This issue is related to this one is sorted out instead of running sonar via hudsonjenkins sonar plugin Ive configured m extra step to run maven sonar plugin sonar goal with alternate file passed with s parameter by referencing WORKSPACE build variable in the path to build specific file +Thanks Hairong! +Applied the patch +I committed the addendum +I updated the list of the projects where I get the problem +To which printerdriver are you printing +Ok I can reproduce thatI dont really know whats going on though what is the difference with the former case +Proceeding without anyone else out there in Spring land have this working and or know of a work around? Any help would be greatly in advance +Lets try not to be too scary to users trying to report bugs eh? Concerned committers should properly update fixaffects versions as they can but I dont know we need to lock them down to committersThanks for the report Renan +Personally I would use because Im familiar with the API but the approach that you propose is certainly more user friendly especially for novice users +Ive tested the simple and the jaxws frontend +I understand that is the intention but that is not what the installer the version combo is waaay longer than the list of specific jboss adapters we EAP is differently versioned but they map to the same AS AS AS AS this list is fullI should correct version lists +Until few seconds ago this bug was not fixed. +We have rollback your patch because in some case it was silently s pehaps a problem on your system +Additional comments on the documentation can be found on the Wiki for doc review +Applied the patch +The patch computes implicit vertex creation and vertex mutations grouped by partition +Fixed the code to not hit the cache when user did not turn on caching or did not provide a cache hint +Thanks Xi for the patch looks good to me +I was thinking the same thing but for now I cannot reproduce the failure on my machine +Its a bug in the Stax parser built into the JDK +I think most are going to use just Solr AD File DB and Web withThats fine although is not yet been certified to scale with unlikebq +The test +it is unlikely that Directory can be excised from Apache but there are outgrowths that might make a good call did not happen +Strategy class should implement interface +Please note that this wont break any default behaviors if you continue to use this improvement is not related to you +BUG exists in as well! +Can this ticket be closed now that has been merged intro tunk +The reason being I would like to try to fix this issue once and for all +Given no follow up has happened on this jira for several days +Any fixesideas about this? Otherwise the service seems to be doing what I want it toAny unrelated ideas about Whirr best practices that Im not using? For instance why am I having trouble accessing the global scripts in my service scripts such as installtarball? Also I am still finding out from the Spark people whether theres any point to delivering a spark conf file directly to the users local +That is the issue I intend on fixing today if anyone sees a problem with that please let me know +One more issue From I start the slave until I start the master the slave is filled with the message Replication slave database testdb listens for connections from master on localhostIt would be nice if this was written only once +Error reported incorrectly. +The attached junit test verifies behavior of the getValue method with arraylistset and index below or above upload a version of the test submitted by Stefan with added tests for negative indexes +So youre saying the configuration is not loaded when the coprocessors are loaded via HDFS +Committed to trunk and change should be put intomibmsvn commitSending javaengineorgapachederbyimplsqlcompileSending javatestingorgapachederbyTestingfunctionTestsmasterSending javatestingorgapachederbyTestingfunctionTeststestslangTransmitting file dataCommitted revision +HiHave we established whether a possible patchfix will be provided for this issue +This is an improvement in terms of more robust error handling +this one as we dont need this. +Its just checking that two calls to score return the same things but if it returns NAN they are not considered equalSo I guess the question is does it make sense that the score is NAN +Ideally the executable would be located using the same mechanism as for other executables like csc and gmcsThis would also address the problem with having an absolute path for nunitCommand on platforms where leading in the absolute path is confused with the used for command flagsAt present the best known workable solution for platforms is to explicitly put the directory containing on the environment path which is not desirable given all the hard work already put in to making sure global paths were not required for any other executables +Could you rebase? git apply Downloads error patch failed coresrcmainscalakafkatoolserror coresrcmainscalakafkatools patch does not apply +Could you live with this approach? Also the just quietly shutsdown with out saying why +Depends on you have some documentation piece to suggest +bq +Done +Patch I think its readyI added expert protected method score and fixed to sometimes return true hereI added an that then verifies that all are positioned on the same docID being collected +fixed in SVN thanks for the patch +I believe that directory paths are understood without saying that they are names +I just performed the following steps on my Windows XP machine Removed from eclipseplugins Copied from a clean unzip of into that directory +Hi RenatThanks for the patch +subscriptions do not move across regions +Awesome thanks for your attention +It is only intended tostart a discussionProblempseudo collections do not have any operations + +it is possible to attach patch and hope that developers will apply it +I think I can fix this soon and attach a patch +to decide when to refuse a connection +Works fine for me +bulk close of resolved issues. +Thanks for fixing the bug so quickly +Still trying to run it +attaching suggested patch in a diff format. +If its not acceptable please let me know the right way to do it +Patch that never returns dnanodeDefinition property from session cache +added an annotation for a deprecated just committed this +Alan same behaviour for both STS and roo shell +Isnt this optionalWhen an instance is loaded using get a Query is executed or an Extentis iterated the implementation may choose to use the active fetch groups to prefetch data +The + s a patch for +Note that there may be a balance to strike between syntactic sugar and API stability +doesnt have any Japanese characters but there still are question marks in the canon +Agreed about fixed +Yet given that patch I think we can safely move the CFS related settings from and into or add an intermediate class +Looks good overall +This is likely a dup of of +Reopening to add label +But now I got thousands of compiler errors +Thanks DmytroIf we need this for branch can somebody pl upload a patch for the branch as well +Validation of passivation capable beans and the is not a passivating Section A managed bean is passivation capable if and only if the bean class is serializable and all interceptors and decorators ofthe bean are serializable +Any suggestions on were to look to make that happen +hoverClass attribute is removed andhover is used in CSS +This is very bad also for small areas expressed in degrees where getting at the very least decimals is important +This was done in Alpha. +This is a problem with the new resourceExists method +He said it looked like its required because it gets the host address from the seed thats in the configuration file +Done +Reporters of these issues please verify the fixes and help close these on Build. +SankaIts your callThanks DeepalI need to be able to change the effective policy without interruptingthe service + on patch +Hopefully this should solve several other cases of similar problems on Windows +for Chriss patchI did ant tarcd srcbenchmarksgridmixantThis created With Giris patch though I was able to build jar following the steps described by Giri when I run I was facing problems with runtime +Verified +Real fix for this one is critical rather than blocker since dont have cycle and hacky workaround +Simon could you provide the list of available parametersWe must also add a project node to easily retrieve the project containing the issue +Got a basic properties component up and running +I think it could be merged on Karaf branch +Incorporating Devarajs comments +Since we probably shouldnt be adding things to RCs +Similar IE have problem with style fixed +Ah got it thanks! Will add to release branch. +Dave I assume its OK if I use this contribution also in Apache Tika? Please do thanks! this as a duplicate of +But it would be nice if there was just a switch to turn on + tests pass +Added examples and improved description for ajaxSingle and limit attributes +This was a hibernate jpa entity manager related issue +Attached patch +Since this looks to be fixed in is there any reason to keep this open? Can we close it out as fixed in +With the in the build the directive in the manifest looks likemypackageversionWhen the is removed the manifest containsmypackagevers ionversionIt doenst matter which sourcetarget version is specified either +Im not saying there isnt a problem or that it couldnt be improved but I cant see any concrete failures right now +It has been sandboxed btw. +Given that other than some documentation and a hack no XHTML output has been included in Xalan I tend to agree this is an enhancement and Id be happy to contribute an XHTML serializerSome considerations Should xhtml be a separate method or triggered by some parameters for html and xml output options such as the doctypeI tend to prefer a separate output method xhtml relying on doctypes somewhat limits the options Should be a separate class or can the XHTML serialization get merged into the or classesThe separate class is easy to implement in fact I have created on based on +This count is updated every time the process tree is refreshed which is once every monitoring iteration +Also hadoop record IO does not allow this character to be part of a record definition +Agreed with Yonik would like to see this wired purely into Solr where it retrieves stored values +Since people dont use Permanent Link and rename articles happens very rarely we could provide permanent redirect for old HF URL to the newest version HF URL +The idea is to use the to construct that match whatever pattern we use for and forward +And test the REST senarios as well as soap senarios +Hey Koen this is expected behavior +Tom Fennellytfennelly cygdrivehJbossESBtrunkproductbuild find name mysql +We would like to keep it so that if you are not running secure hbase then you do not need a ensemble +If you are using and it isnt working then perhaps you might try using directly +As in geronimo the datasource was configured through console dialog +Status please? By reading the forum thread I can figure out whether something was implementedfixed or not +Issue closed due to the release of Tiles +Let us think about how we can enhance this to have a which can be the base for all +can be still reproduced in Trinidad +Im uploading new patches that also update the list of valid characters displayed by command usage +Committed as is. +Please verify by marking CLOSED +expand the reference guide and where relevant +Is it possible to get it included as wellSomething along the lines of +Did you ever add suppoert for entries in ZK? I believe it will be usefull in a lot of common scenarios +and investigation of internationalization would be a nice thing +utilsrun because of the use of ERLLIBS now which is incompatible with how our source directory is setup +A first version of the implementation which can be forgot to upload the api module i designed it to be put under deps but it can be tried to simplify it and adapt it a bit to the trunk +bq +Revision is revision Ive rewritten of it +Writing a test for this hope to have the fix available by clean up small bitsWill rename is +This is now fixed in upstream +made it java compatible +That is an apmail function that we have to do on hermes for you +Verified in ER. +Ok no problem +released closing. +Ill add that having this pluggability in place it would have allowed to have a fix for avail for jobs hitting the problem until the fix is delivered in a Hadoop release +the JEE descriptor file such as server cant decide which deployer can be used to deploy the zip package +Moving out to please move back if you think it will be done in the very near future +Start of a patch for discussion +JaroslavCould you sent your configuration details and the error log +Thanks Ole +Hi FreemanI made the suggested changes and everything works! Thanks very much +Thanks Kevin and Aaron! +So maybe we do need to go to RF to make sure we have another copyAlternatively maybe we could special case this and write a hint elsewhere if we detect that the DCL replica would be local +oky +Yes refers to a assotiation with Subject +It has been marked as fixed todayIts hard for me to create a small test case as I need code generation from the wsdl with a custom binding and then unmarshal a soap message with the generated code +Shall we reopen the same to fix it for +I understand your point and totally agree that the current implementation isnt correct +A flag mallowDetach should set to false for all of a union expression +Outline toggles the RM node to connect to +Ill put it on hold temporarilyNice catch +Our motivation to use the Select for was to be able to define the for the in an which is not possible using the Wicket +Second thought about the stuff related to after chatting with Owen +I have run with and without the sysproperty feature and the only difference I see with forrest debug is the fact that it adds the property valuesStill dunno what exactly breaks investigation continues + has been fixed and it should take care of checking bundle signatures when the framework restarts +Heres one for the default also if you want +But using a threshold to force large distance measures to product zero affinities instead of just small affinities would make the A matrix sparse again and allow subsequent processing to scale better +Is it too late at this time to present the changes to the Coverarge interface to OGC? +Latest patch changes look good +classpath is what I can think of +If you want to retain lots of logs then just bump the retention size in the config to whatever the partition your logs reside on can support and youre good +HiIm very sorry +Ratnesh ChandanHere I am uploading patch for the secure url issue in adding updating and deleting adjustmentsPlease dont include them in your patchAlso this issue should remain open till we change all the urls +Will update result later +Exception thrown if the topic is froma different sessionUnit tests svn revision +Perhaps it was my mistake for including an example at all +If there is a problem determining the installed version of Java it must only be in the groovy fromthe installer +I am going to add more debug statements to my code to see if I can get more details on the exception that is occurring other than the that is occurring in the Util class +See the commit log under the Source tab for the details of this refactoring +This issue has been fixed and released as part of release +This for example means that though I scrolled to the end of the list the scroll position sometimes moves back to some position in the middle of the list where it again only sometimes remains +Ill take and next +In the past only committers could get that because we used it to generate our website +Gump is happy with the trunk of Batik and XML Commons. +But i see you only targeting empty directories which cause little confusion at allIts what had my attention at the time due to combination of writing about them and working with git +folder +Thanks Jarcec! +Verified Thank you +Retargeting to fix in since has been shortened and made more focused + +s ok but in IE under JSF it is still a problem +new version +Patch applied +Any more word from Perforce +Is there an implementation of the special support? I have a similar problem with dutch word decomposition at query time i would decompound words like hulparbeider in hulparbeider OR hulp OR arbeider but the parsed query contains only one word group hulparbeider hulp arbeiderCan you give me some tipThank you issue is very old and got never any response +Same environment same error +committed the patch. +since this wasnt fixed +Could you please attach your reproducer +The patch needed a lot of effort to upmerge again +This is a bug on ri and myfaces behavior is correct +I just committed this to trunk +by catching the null case and providing a reasonable reason you very much for your fast response! Works now! +Yes richselect and hselect do behave different in your case +I have tested version and the same behavior has been observed +Cool +Maybe for now we create a module in +Assigning to Rich for review and check in +Committed to as well. +Dont know +Looks great patch! it as order is not the problem the problem was to support multiple messages in a bundleNo one actually should write the size and vertices and then expect to come elements in that ordering +Jukkaa small patch adding the oneliner as lines into but I am not sure if this is the right placeBest some experimentation I figured that removing the old wont harm existing clones so I went ahead and removed them from all the mirrorsThat should cover everything for this issue so Im resolving this as fixed +Trivial patch attachedAny concerns +I guess also what I am unsure of is if you select certain options when setting up the developer account does it affect what you can do with the features available in Spring Social +Probably good to synch this to avoid weird merge +Closing released issues. +I thought about making it a coercion but decided itd be better to just let people pick their own poison +Let me know +You are probably safe returning texthtml everytime until your renderkit is fully xhtml compliant for all components +The issue was an incorrect check to discover whether the module was a flexible project or not +We always should decapitalize +Ill post a trimmed patch +Please that dynamically created trees can be run and now get named. +By default it is kept disabled +This defect is reopened in order to undeprecate the +What would be useful also would be the ability to add a features URL via a similar argumentSomething likeadmincreate furl mvnxmlfeatures f or +Generally speaking native is not a good choice for all the reasons that led me to write the enhanced generators to replace native in the first place +With the commit of as r this issue is resolved. +Im resolving this as cannot reproduce for the moment +Barry I debugged through this one when cleaning up my and its not a hardcoded type issue with int and datetime +Committed to trunk +Its in iOS as well + +The patch is attached +Malcolm it would be great if you could report these two bugs on a different issuethread +Why add another requirement to the mess +Can you check the pom inside the archetypecommandtargetclasses folder? It seems the maven resource plugin did not do its job correctlyIm using maven and and both work problem only happen with maven Ive fixed the issue which was caused by a wrong version of the to be used. +However the namespace read lock is highly undesirable and the repeated dns lookups are questionable +Changed to awk +hey dude there is one on peate too which allows you to plug in any Extern Metadata Extractor +Total pages total working days average pages done per day is about a little tough for me +I linked in related should look at this in the context of performance and whether I redesign of is therefore needed +Currently the implementation of starts one listener per Conduit +Bulk update to fix version +backport patch for +JPA recognizes in the srcmodel directory and creates in this directory +Committed fix version to that we can keep track of this btw shouldnt we update our own in assemblysrcmain? +That makes the initialization behavior explicit +Its the same issue but it does not seem to be fixed +Committed to trunk now +Fixed the issue with only loading one payload per term and added unit test for it +The website will change to reflect the current code and not the distribution +changing closed later to as WONTFIX +How about posting the snip of the jsp causing the exception and the stack trace youre getting +With a sliding window you are in trouble if you cant compact fast enough while you still remember that you flushed a lot to catch up +Thanks Norman +iocorecluster +I will try to get to that soon. +forgot a fileuse the second +I see the value of and am interested in working on this +This is part of resolved issues. +Verified with build of trunk at rThanks for the quick resolution. +But Accessors arent cache they hold reference between a type and an utility classes +make sure that we are using the same instance everywhere to avoid more memory being used than necessary +is the right place to be defensive now +I think msec is way too long I was thinking more like msec +after addressing Teds comments except for the hashcode multiplication thing +I am not sure whether we ignore NULL in join but that seems easy enough to doDo you have a specific example that the current Hive code breaks +Commit more changes to branch at also merge all the changes to trunk at rev be fixed. +Applied patch in revision Thank you for providing the patch. +Almost much changed but just putting what I have up so that it is captured +This should be treated as any other dependency +Patch to address that patch does not apply cleanly to current trunk at r can you please recreate patch against latest patch against +Well get right on this +Attached is a new patch for this issue +Jrn I believe the framework should make it as simple as possible for the ML implementers +Fix for the hostinfo deserilization +Search for Start WW patch till End WW patch +I synchronized the patch with the current trunk and reversed indentation changes +Okie Im reopening the ticket +In Svn +This is definitely a problem with Validator since it is requiring that thedepends attribute be included in the fileThis is a duplicate bug to +Fixes validation issue where portsocket not +see previous comment +I ran once locally and it passedThe test failure appeared in build as well so unrelated to my patchI plan to integrate to trunk first +Available in patch of +Asking user to map data point first and specify rowcolumn id is inconvenientWe response the comments and wait for the further discussion +Looks like the reason of this failure is the same as failure of one of the test from +As for the reformatting of the sql it seems to be a feature of the utility +Gah. +cool thanks for the review! Ideally a unit test should run with as little extra code as possible only in integration tests should we have all those extra components enabledI fully agree we can start trimming the test dependencies as soon as this patch is in +Revision adds a tool for manually marking a store clean after a total cluster crash where there is no clear last member +It already is an issue for you comment what could be wrong and that it needs to be assigned to the specialist +I also removed the inner class it made a lot of sense when mark and reset also needed to be tracked but was less useful now that the need is gone +thanx ivanCommitted revision . +Since it cant be sent when container finishes we need to send it when the application finishes +hey any progress on getting a repro together for thisThanks +Please with JBT. +Paul this looks like a Spring problem rather then a Spring DM specific one +But it seems many other commands just output error msg and succeed like describe +Agree +Patch to fix catgs and and feature descs +Slip +At this point lets consider the initial new UI layout work complete and move on to new issues +Hi BJ would you review and test Ashishs patch to see if it fixes the problem you reported? If so I would like to include this fix in the upcoming release +Now uses a blocking queue to limit the backlog of pending connections for readers +tested on Ubuntu and Fedora Core. +addressed some of Ted Yu comments on reviewboard +I still need to refractor the code with regards to logging exception tracing checkstyles etc +just moved the site src Im building the for taking a lead on this Simone +Committed revision in trunkThanks Markus for review. +acquiring a portable IP in networkZone and enabling static nat with a VM that exists in and disassociating IP works fine +Ill resolve that and if there will be any requests for clarifications them will be separate tasks +Auth is the same as servlet +Triaged for checking repro attached and setting normal urgency +I have committed this +looks good to me +That is fine then +For those watching this issue I pushed a couple more changes to the github repo linked above +Everything checks outReviewed by Alan and me +We need to be able to figure out which principal class is primary and should be returned from get +So I did some ant package ant docs and then ran +Id rather save h for human readable for whenever someone commits code that changes numbers from to TB +Ops teams are expected to have multiple different configurations for services served via their configuration management systems +My bad +Michael there is that is a straight Layer implementation is actually just a wrapper around itI think we just need to deprecated thats all +I believe this is now fixed in the latest +By a list for every vertex we are going to use memory whether the vertex will receive a message or notI thought you might be looking into a higher level concurrency by using or something like that for transient + In fact printing out the request line of a PUT request appears as HEAD HTTPMaybe the server is indeed getting a HEAD request which explains the fact why it does not implement dont you think? Is this functionality supportedYes it isPlease provide an isolated test case that reproduces the problem otherwise Ill close this issue as invalids the test code +If that isnt a code smell I dont know what is +I compiled it successfully but cant figure out how to test it +This would allow for each remap rule to carry one of these as well + +However please file a new jira and supply the patch there +The file linked by the QA bot doesnt load but I am guessing it is warning that the datak file has no Apache license +The first patch had an extraneous change set that shouldnt have been included +Not sure how you failed to reproduce this + +Users should migrate to AS as soon as possible. +therefore the standard java rules and restrictions apply nothing special +For example in the class the default encoding is explicitly set to by default +Attaching +thanks for review +Changes look good +l peeked at and it looks like the private field instance var should be set in the ctor using the field from the first clause no? How come your field is null in the first place +Closing issues that I resolved in already released versions +Id be grateful if someone could review and apply the fix +No more feedback so resolving for now +Thanks +Its still open no patches have been proposed +Increasing resource is harder because we have to wait for resources to free up +I dont fully understand the semantics of PERIODICAL though +compile compile runtimebq +It is an all all Apache effortBy that you mean it uses open source technology from the Apache Software Foundation rightIt was not developed as an open source project and did not follow the Apache WayI think the other demos that were code dumped actually demonstrate my point they have not grown the community its basically dead code sitting in Derby +v no longer applies heres a v +I guess some thought should also be put into how much should be done to our current implementations and how much should be done on in the way of replacing our implementations with colt impls +Nevertheless the delay really helped for some time +If I receive positive feedback I will work it into the tomcat component as patch also apparently doesnt work when loading the overlays from the reactorOlivier given this issue how are you developping Archiva with so many WAR overlays? Are you installing the overlays and restarting tomcat each timeI see you have some configuration of the but its with skiptrue +If someone has a concern with this please let me know +v only sends timestamp int +I hadnt thought of but it should be easy to write a pipeline stage that stores everything passes it through and that also has a worker thread that does the write to HDFSWhat do you mean by removing the Hadoop dependency? I assume you dont literally mean breaking all dependence on +PeterIve tried but couldnt reproduce the problem using trunk and Solr example +We may need merge the ranges there are still many small ranges returned by describelocalring if other DC is configured with vnodes +Asokan +Title and issue type changed +Is this for building or a runtime problem +I have made some progress concerning profile diff feature +Pete do you want me to send pull request for such a changeI can see no other Quickstart upgraded yet +Did you look at changes in the pull request I made +on JBT. +I will try to follow your recommendations in order to prepare a working patch to in a few days +If you have some ideas we could discuss them on the mailing list +The documentation has been updated to specify the SupplementaryExtra channels required by Enterprise Application Platform which corrects the dependency issue and results in a valid installation +Done +Let me know if you run into problems +If theres an obvious place to expose those constants so they can be set by the application that would be the simplest compromise but Ill take what I can get +Shouldnt we make and tooThose ones arent currently causing deadlock potential though perhaps they have potential +Thanks for the quick patchbq + for allowing use of Lucene Analyzer to do the delimitingthis could then let people control stuff like accentcase normalization synonyms etc +Id like to succinctly reassert that its unexpected behavior on OS X for a java app to not honor the users preference +This needs a documentation update to describe this new functionality +Since hudson has some issues ran the unit tests and testpatch +The work so far looks great and I would be fine seeing it checked in as is and backported +OK my colleagues in Beijing were able to reproduce the problem +Closing +uses the new percent suffix made available by the plugin change. +I know I know its a +To run the script include the jar in the class pathjava Xmxm cp HADOOPCONFDIR to write a unit test for that since this issue only happens in real cluster attach testing script I am using for manual verify +I doubt we would want to get into class renaming but I suspect the code would become seriously unreadable if we were to fully qualify every generated object just so clients can have services named String Service List etc +Otherwise cluster restart times could potentially take much much longer +Doesnt the file manager deal with element validation? And does the CAS curator also modify the same xml file to add element validation informationTotally understandableSounds good +From a quick read through the code it look like they were not intended to be extension points +Only session beans with a remote interface can call this method + +Yahor Ive fixed this in M by using commandactionSets see +This would also allow non Groovy code to use it +and atleast should not be driven by Configuration +After creating several thousand AFP File you soon find that the Operating system has exhausted all its Open File Handles. +Mostly FindReplaces and a few manual corrections +I am going to check this in +Many thanks for reporting the bug and submitting a very precise and detailed description of itOleg +Can you give me a test case that demonstrates it? Thanks +The wiki will need to be updated for the next release since it when release artifacts were still in the build +This is going to be removed correct +Epic patch committed +Ashwin James is already working on this +The failure was because we now check the connection attributes in the call to set itself +bulk defer open issues to defer to not essential to address for +and derbyall completed without errors +bulk defer of unresolved bugs to +We are uingWe are uing Velocity Servlets Struts Hibernate Jboss +also added test for mismatched cluster id. +Is there any plan to replace this with a testThanks for the explanation Dag +yes I dont want to remove it completely since I think we may have it at a later point maybe +From the log retries took at least hour +This looks like a duplicate for meWill work on both now +you need to remove the repositories and pluginrepositories sections from the done +I assume that will need to be placed in the s projects +Pat I have to argue with your last comment +Where are you storing the Database? Can you please provide that information +Patch for upgrade of from to works as fine +Probably after review commit Ill see if its feasible reasonable to propose a backport +How was the read performance impacted during your test and after the fix +slip to next version +Rebuilt to make sure versions are removed from WAR plugins uploaded again +I yesterday after updating from and are thrown for each search +Now that work is well underway with AS all previous community releases are +Looks like for some reason the instance was null +What I was seeing in console was notthe true story +People searching on a job site dont really care for top vacanciesresumes they want eeeeeverything! that matched their requirementsJohn on the other hand we here already have code to merge results coming from different shards with stripes primitives and whistles to appease GC +Two more corner cases not yet fixed if rows we get an NPE if and are both then the counts for the resulting doclists are all zero +I believe your example above describes what changes we need to make to the API and I agreed with you on the various flags in the table descriptor and column definitionsThanks I can do anything to help please let me know +It could lead to exception to be raised +Issue fixed in release of Jacob +If you are using the official IBM zOS toolkit you should work through your official IBM support channel +have a look at this issueThanksSanka +Maybe a good example here is where we ran into a Tomcat buggie with the package nameThe other project that we will probably be related to is since we started using JCR as the repository +The tmpLoader isnt strictly necessary for the xml format but its a related code change +Tomo filed another JIRA sometime back with a similar idea +The patch is in CVS +Committed +Could you attach file that you use in your project +I think this is ultimately a duplicate of actually. +Resolved along the lines described in the description in required the fixing of a couple of testcases in addition to the mainline code since the initialization sequence is changed to require a Registry object to be supplied used when creating the +The geronimo code is assuming everything is bound under javacomp +Version id It also seems that the file created with Create POM +In any case this would take not too much time to switch back to the when another argument appears +It is also probably worth noting that the equivalent of the b command line switch is +Patch to replace most of with an embedded +A new approach is being elaborated. +Simply try mvn sitestage absolutepath +In revision I modified the portelet not to load and unload disabled modules +to the patch +To run the tests was added to the pom +The src dir has all of the tests in it +that are quite greedy when polling +Launch configuration Newconfiguration at Dprojectssbsbtmpworkspace +Teerific! Thanks! +Were experiencing the same error in Surefire for a missing spring dependencyReverting back to the test appropriately fails and produces the expected output +Id really like to drop the need for multiple bundles and have groovy properly support sourcetarget levels +I will try to rewrite the issue on accessors rather than Reflection API +Right map side cogroup needs the sortness of the input but just the side inputs need the feature to be able to seek on a key the base input will only need presence of all duplicate keys in a mapper +To me integrated into master later on +I just committed this +So here is a new patch to deal with an entity which has two or more attributes with the same entity as type +Now when compiling the jute files javac finds the old classes and the new ones under buildDid I mention that this issue would not happen with GerritJenkins? Id really love to help getting Gerrit into the ASF but would need some Apache people to recommend it. +Thanks Igor +Thanks again Marc +wrong jira. +I can get sftp deployment to work when Im connecting to a Solaris system but not Solaris +tested +Jarkko please test and close the issue if all is ok for you +I removed the duplicate +We dont want to risk regression and given the scenario deployundeploy application times it would be unlikely to hit this problem in a production environment any time soon +Fixing the title + +Ok to recap there are some unclear design decisions made regarding the exception throwing in checking visibilities against authorizations +So I propose to postpone this for the time being +I think the tag logs it as debug because these errors should be cleared up before you release the application +Ive included a test that would fail as part of to make sure its covered in the future thoughThanks again for the reviews and attention +I just committed this +That sounds reasonable to me +I just committed and deployed a new snapshot because I found a small bug that resulted in the bundle class path not being calculated correctly in all cases +Hi Dan your fix looks good +removed some stray log on sense re points and +This issue was not fixed in time to be included in the release +This blocker critcal was created before July please review and resolve we are approaching code freeze in days +Bulk close for Solr +This is the functionality that we would actually like to add +Mark I have implemented this workaroundCan you test that work in windauze +what do you want in that file +Ie +Although I am new to I do have interest in security and web services xml in d be happy to take this task +I will submit a parent pom patch soon +Hi Michaelyes I saw that it was used during Json serialization but not for xml so I recommended the +How are you using the Dependency PluginIf possible please provide a complete POM that we can use to test your issue +Thats the only way I can see it +A client that was not patched was able to talk to this server +There is not an easy way to conditionally register that listener +Hi Bertrand Ive applied your patch in revision +I dont really know about the testng +However it looks that you try to post some documents here by coping our design and then claim that you are one of the authors of the designThis certainly was not my intention +we may simply use the same lookup mechanism for but search for files with the suffix +bq +Thanks Vikas +Matthias for the suggested changes I will update the javadocs and other required ones +Moving fix version to +Committed. +This patch depends upon have this addressed in the updated patch on +This patch passed mvn clean verify +Attach a patch +Looks like Suresh filed with a fix for this. +reopening for classpath part +Alex do you still see this happening on the trunk? If so please provide a sample configuration that will reproduce this +Committed revision +CCFLAGSmarchpentium CCFLAGS CXXFLAGSmarchpentium CXXFLAGS Finally I did a test with marchi with the v patch and it does not work +HiCan you attach the WSDLThanks by r in because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +Added the flag flag help and support in both the proxy and server synthersizers. +Yeah good idea the same as I concluded as wellChristopher can you try with updated code from trunk? And report your findingstrunk +In case that theyve added new column to their table this process will not be straightforward as user might need extra step for some temporary tables or some other workaround +I tried to remove redundant code as far as I could with abstract implementations and already modified the existing trainer to extend it +We do our best Patric. +But I think is probably easier so we can customize it to our needs and not have to pull in lots of transitive Todd thanks for the clarification +Weird it works now +We probably dont want to just use ARM since based on the info you provided Windows RT also has ARM in the user agent +Yeah sounds like the right direction I did not pay attention to +If you have any ideas about the design please write it down here +These new ones include the following +with this change gc test cases cannot run concurrently +It has been in the nighly downloads for a long time now and I would say that you would be better off using a recent nightly build than RC +Cbayes has scores which is decreasing with increase in class affinity +fixed with patch for +Verified on JBDS +patch +Since I cant figure out how to include files with Radeox we need to postpone the issue +Why not get a random available port? However this only will be a fix for this particular test +is already quite crowded with massive changes including parser and logical plan switch +Besides sometimes we do want to serialize a this class can be used thenBut this is a very simple fix that we feel is immediately needed in so if you think it is of no use in trunk I am happy to leave it as +But this can definitely live as a patch and anyone who needs it would benefit from it +I asked Santhosh to review that part +Thanks +git bisect implicates +What do you think +Just noticed the patch contains changes to the import section +Thanks to you too this was quick +Please verifiy that it solves the issue +Bug has been marked as a duplicate of this bug +The procedure for JDK installation is comprehensive and works. +Thanks Jacques for the quick response Ive updated the Documentation page with the release number +Looks good to me +Add the testcase about this +bq +Values for fields annotated with Size are padded out to meet the min size requirement if specified and truncated to meet the max size requirement if support for Min and Max +It doesnt kill tasks beyond yourWe can discuss this more. +You would then have to play catchup shoving all that L data into the other levels +I committed to trunk given from Aniket in RB +I just commited my changes +Now the build is no further from working on WindowsRalph since the build is no longer any worse for wear on Windows I see no technical reason for not withdrawing your Regards accepted +Im not asking to publicize it +Then you would only have to check one time for the interface and cast to the interface +Switched to RHT since it violates JPA deployment root and some tests use this exploit. +passed tests in my local +We can use to replace urls in Test methods but not in static methods such as Deploymentwhere it is often necessary to have it in order to configure server via management apiWe should decide how to solve this +The error message above is on debug build release one says +The sample app works great however I am unable to obtain the meta data +This is fundamentally an Oozie problem +I am going to code the patch so it does the same checks when determining whether or not to do the lib deploy folder +The datasource Config took the Spring jar from the Nightly Snap shot and I have added the property you have provided to the Spring config file +Some of these items are inherent issues with the target languages +bq +Could you please get the latest snapshot off the branch and your application with itOleg +This would be a nice addition for configuration option +Works with through +Right after that a fresh Struts build will follow. +Sick thanks dude! +If p fixes the problem I would vote for having p in and +Is this satisfactory +A little change through this issue only validation during marshalling should be addressed +Yep I still like a full replacement or a fully pluggable solution more but if thats what the community wants I will go along + for patch +I will take a look at the current pull request +This happens with BOTH offline and server deployment +Closing as wont fix. +See can confirm the same behaviour after upgrading Phonegap from to +grammar close for release +Also includes the testcase +Please ignore +and failures are caused by +So what do we do about value? Passing as idle timeout seconds to Netty disables idle timeout checkout which is the same as saying that idle timeout is unlimited +Thanks +Patch applied in the branch under r +Applied with a small change +This would work for both full Accumulo installs AND for the MAC +samplesblockslinkrewriterwelcome +Im thinking that this could either be done automatically by checking the creation time of the data that is used to create the checkpointed data or by having some kind of run mode to force overwriting checkpointed data +Mark suggested we investigate it a little bit further so I reopen this issue +What steps accomplish that goal etc +I could not run through ODBC test yet +I think it would be implemented with mechanism of resolving EL expression in whole source text not only in JSPJSF attributes +This patch fixes the issue +So can we resolve this issue? Do we need to doc +Sammy is the first AFAIK +Other platforms display Fuse even Win +Are you sure you followed NOTE Dont forget to run mvn o N install in GERONIMOSOURCESHOME as the parent pom has changed? Please run it and your troubles go away +let me know if this issue goes away when you applied the in m being affected by this again +Any thoughts +and just upgrade the jar +Tests pass all Java and the cpp profileI left test owning to the comment Currently the cluster does not support transactions that span a failover and the fact that I have no way to run the clustered cpp tests +check latest cvs javawsdl has a deploy option that can create the wsdds as well +Thanks Konstantin +Note that I also saw syntax errors after running the script and I also manually corrected the generated java class +So I dont think its a duplication at all +I forgot to mention that Im having this problem with the latest version +ts been a while since I looked into upgrade and stored procedures +Already changed our copy of the source code +Ive changed the tests to always expect a on the end which isnt the nicest solution but it at least makes them pass +Closing resolved issues for already released versions +Igor please review +bq +Fixed in r +We always prefer to keep the namespace as simple as possible and have it cater for the most common use cases which people encounter and those where the gain in terms of reduced configuration is substantial. +Fixed +Ill take a look at this +Still not works in with Mojarra and Weblogic ServerBut the workaround works fine +Currently it runs the test through maven via the Maven test menu but there are some issues with that +to shutdown +More fixes and tests +I think this is some premature optimizationI am not sure if that is premature +Ive checked out a clean version of applied the patch and it works fine +Check is neither supported on the propertycollection level nor is it part of Column +Committed to trunk +It is a bit scary if generation can be changed by clients and datanodes +Thanks a lot for the contribution Todd and thanks to all for the reviews. +Username vharcq +Also this looks like we are changing the protocol between the client and the server +Closing the issue now that it has been resolved. +for the branch only. +The third set are tests which I have found significant dependencies in +I am waiting for Kishan to merge networkacl branch to master after which I will merge the UI changes for this featureThanks support for this feature has been merged with Master +If we include it in any fashion then its content is going to be at the mercy of the transformation engine being would be better to make it consistent for all usersengines +Looks good Ben +Sorry misstyping When creating a new in the construtor it calls the method initMembers which iterates over all the method of the component and its superClass and calls scanMethod which will register the Observer +I cant reproduce it now +Jboss AS works fine. +Ill keep poking around +Already reviewed +Patch committed r +Besides CXF depends on JAXRS while I can see a JAXRS bundleAs a side note to the dev we probably should submit a JAXRS bundle pom to the servicemix bundles project because I suspect does not deal with OSGI Locator +Although Im in favour of this change does it really classify as Critical? Ill downgrade it to major although it might even be regarded as minor considering the workaround +Denis or Daniel can you please add something about the GWT configurator in +or simply use the seam wizard and try and deploy works with default jboss but not s at least the bhevior I see her for the default adapter +I will attach an additional zip containing just the test classes +Do you have some pointers about what goes wrong and in which environment? the above CI failure is actually a different issue see Unicos post on dev +No tests just trivial text changes +You use variadic macros with logmessage and a GCC extension of them at that +I am running with sun jvm now +Feel free to continue applying thread safety or unsafety comments if desired +As noted previously the fact that this showed up in a module deployment was probably a coincidence instead the Chain class was just incorrectly paying attention to the value of which I believe only applies in. +Again I refer to the question WHY is it trying to create a folder at ? I havnt told Maven anywhere that I want something created there +There is a solution +Does this mean that is able to be configured at runtime +Done in trunk +This wont be fixed for +The temp file will remain in disk if the target file already exists. +Yes I did this intentionally when I discovered that the Apache server had been tweaked to allow BOTH a to be shown and a generated indexWhats the problem w this new look +As per Andrews feedback a Config Set solves this issue. +Only a select fewI dont seem to get this problem when building locally on my Windows XP Machine only with Hudson on the Linux ServerIs there something I can do on the database or server side of things to workaround thisIve looked through every class and cant seem to find any funny characters that might throw the error +Also remember that mock model is loaded once and never updated in that Eclipse session +openejb tests are being skipped +agree with jarek The aim is to prevent TP technology being mistaken for a supported technology in a SOA context. + +I ll look into it +Ok Ive duplicated the problem +Im getting a lot of failures when applying this patch +Closing resolved issues for already released versions +target version update +plugins +Lets take care of this in +Heres a patch that does the following adds JVM parameter which can be used to set a port other than changed the logging to only print the stack trace if trace logging is enabledI did not add a test because of the difficulty of properly unit testing things that happen in static initializers and logs +Then it somehow renders the error page in the popup +Fixed in SVN rev transition resolvedfixed bugs to closedfixed +Local variables and parameters are not supported +Adjusted the title to reflect this and to be able to fix it. +Thanks +probably related issueI added the Maven bundle plugin to resolve the dependencies +Thanks for looking into thisI still think this it a least critical +the week end i will surely check this +The problem and solution is similar to +I have MOCKITO a path to a jar defined in different waysMOCKITO +The rest of the code in the contribution remains unchanged and the contribution remains a jar contribution +Feel free to reopen it when you can provide more details. +Maybe not +patch for looks good Brandon +This is the patch for holder support and new Method Marshaller code +Not to be too picky but the directory you added was called as opposed to +URL attributes that begin with a are automatically and intentionally prefixed with the webapp directory hgraphicImage does the same thingTrinidad includes additional functionality where that begin with two slashes just have the first removed +If the numbers remained consistent across restarts it wouldnt be a problem +the thing is +It seems like these changes will work +Hi Heiko Do you have an update on what your roadmap and status are for these functions +Is there any reason for this? I would like to see this continue to be included so that we arent forced to update our connector right now +Broke submitted patch file into three separate files per Craigs request +If not already done so Ill apply the patch today +The attached patch fixes the problem and adds a test case +Resolving as a duplicate of. +Ive now fixed that and the updates should be up in a few hours +It also fixes a small bug in when an empty object is written to a +Fixed in r +I think can benefit from this by using the as the key for the session cacheFurthermore are there plans on adding to the resolution process? I think it raises some other questions +Many users haveasked for this and it adds to our standards implementation so seems like a good feature to add to DerbyMy opinion is that we should not change the existing behavior for unique indexes +bq +Looking at that code we already build up an XML Document object before we output it so theres no need to hold the Configuration lock while outputting +The patch generates the again +closing the bug. +See comments in for a use of the node factory +Tom whats the status of this +PS is this going to make code freeze +I tried applying this change but it breaks the unit test +Otherwise if the hash is still in the old format we keep it that way +Keywords worth considering Java authentication and authorization AuthenticationAuthorizationCryptographySession h titles on page and they now all use Apache Shiro +I believe these changes are stable. + +My understanding is that most commmitters of Incubator projects go on to be PMC members when the project graduates +Reviewers should use IDE enabled highlighting and change tracking with MercurialIf you like to continue to work with FIXME during coding +To demonstrate the problem Compile Run through ij +Closing. +Will commit to trunk first. +So closing this JIRA as duplicate +I committed this with a slight modification to the patch to have know the number of disks on various EC instance types better +This is not necessary +The byte object size multiplier is fixed so the method is fineI dont think its fixed see the XX above +Verified by Mark. +No action required Strong +Are you saying that I dont need to get the s to change the effective policy or that I dontneed to change the effective policy but can achieve the same effectthrough the interfaceIll have a look at and the contentThanks for looking at my JIRA entryJackson Wynn +IE does not invoke ATG handler when pressing Enter to submit the form that has a single text input field only +We could put a there +Thatd be sweet if we had that +For the time being you can unregister that interceptor or look to debug it and patch it for your needs +I think the best way to do this is to actually augment to allow creating a cluster with multiple instead of duplicating the code in another class +Bulk of the change is indentation +Had another patch named that I didnt notice +changed return type void to getting back into work after returning from Agile +is in the process of being replaced by +Ok Olivier This bug should be fixed with Sonar +Here is the proposed fix +Committed +Please review the this patch and test case +We could do this without sorting +It should be placed insrctestresourcesprojects CdricYes were having some troubles with the embedder +I reverted this patch +Service for signing up for a mailing list done as of r this be closed +Maybe illustrate what you think the syntax should look like so we can better understand what you mean +distributions respectivelyJeremiah would you mind sanity checking this at some point? Everything seems to run fine but a second pair of eyes would help hereThanks to Beehive test lead for closure +patch needs some additional work. +I copied the source of this test from some other class +Ted do you know the status of this? I dont think the build error pasted here is related +I created the issue for that +Ill look into the problem tomorrow +Made changes to mark getunderlying apis as private as suugested in +It also possible that we should pass a resources map to the createinstance method because it could contain initialization data however mythinking there was that pivot level initialization should use its normalprocess. +Also if it is binls it should be bash c binls + just committed this +Kiran the new error message yoire seeing seems unconnected to the original issue which was addressed with the fix I put in +Thanks Jerome. +Im reluctant to close this bug on the basis of this fix because it only fixes the symptom in this particular component +Change to +There are some quirks that arent at all apparent like distance has a max of and when distance you actually get a generic syntax error because only two digits are allowed in the token +Patch that removes the recursive calls with a combination of continue and +Josef Thanks for explaining that further +I checked archive from your url and it also only have plugin so where the hell did you get that asm fromCan you try to get Eclipse for Java EE bundle? It has all dependencies required for POM editor and WTP support and after you install meclipse on that version you can try to install your other stuff +Thanks Daryn Bobby and Kihwal! +Fixed in longer. +localize classloading code tosingle class +Im going to look at this and fix mailing lists etc at the same time +Also need to consider which added a webdav interface that currently does respond to anyone who connects +Yeah that might be the case +Thanks Tim. +Im trying to use HDFS on a system where each data node also works as a computing machine with multiple large mounts +A database owner has more privileges than a normal user of a database +Its only if you have a different CI server or you want releases that require additional action at this point +I didnt just make up a convention when I named it +No problem +Reason was unable to load class gives this error after Eclipse M +Im looking forward what you come up withAbout the image its not a second one its a replacement of the old one +Patch MSI build with patch passed test successfully +Yes and no +Thus it is very easy for features to creep into these scripts +Ill take a look at it as soon as I can +I think that means someone has to give you JIRA power and hasnt yet cant remember who to bug on that Hoss or Grant I think? Perhaps the right person is watching +I also totally agree with having a helper function that takes the inputs for all the versionengine checking +Moved to xPlanner +In other words the generated select only locks table CLASSB and NOT table CLASSA in my test caseHope it helps +Updating the patch with the latest version posted on the review board +It hoses +Please let me know if you find something +You might want something like triplesec +Patch for yahoo branch +The more I think about it the more I agree with BJ and Adrian st is easier and less error prone +Thanx for the commit knutNarayanan +agreed with comment second one needs to be proved +Committed revision . +I think some flags would be appropriate to determine the exact behavior +Some tests are failing with the latest patch +Jacopono problemBest regardsFabian +for the patch Alexanderapplied in revision Vladimir please check that it works +Is this required by the or by the mappers reducers themselves +I could give you write access if youre interested +Bulk closing stale resolved issues +I think its ready. +the following note to the documentation also allows you to override corresponding values +That particular example is nasty albeit somewhat benign +Thanks Stepan +Just caught up with this there are several issues being debated here simultaneouslyIt is really pointless to start arguing about them until we have a clear and thorough design doc along with a preliminary discussion of technical merit +Robbie could you review please +I agree that last set of screenshots above is good +bq +A separate acceptable variants section sounds good. +Changed the +Hello RushDo you think this patch will be applicable for thriftOr could you provide patched files to build runtime c thrift libs with VS Thanks +Now It is fixedthanks Lahiru Sandakith +This thing seems to be needed for usThank you upstream +Fixed in r Thanks for the report. +ChunhuiIs this issue applicable for? If so can you prepare a patch for also + +Added a check for the component state before checking the bundle state in Rev +any way to share some stuff wYup lives on so this can all share codebq +I guess one thing to make the launchers launch method less visible is I can make that protected +This seems to be fixed in Maven +Also can you explain how you solve the problem? It would be good add a brief description of the solution you are proposing +IMHO theres not much we can do about it its an issue in VJET where they show context menu items regardless of the view were in if nothing is selected +Nothing more to be done worry +Gradients are still missing but hashbars are present with and PBR +OlaWell be thrilled with any improvements you make to reading files from a jar +Patch to fix missing pygmentize +Since we no longer rely on tarballs anymore this is a now +Perhaps it would be an idea to extend the sdk with the syntax how to do things declarativeGreetingsHenri +We cannot make this kind of backwards incompatible change for users +How can the consequence bind to some data matched on the LHS? how does it know which pattern or which field +Can you please take a look? Thanks +Attachment has been added with description diff against transition resolvedfixed bugs to closedfixed +What you think Mubarak? Id think that if we started a new issue itd be called online merge and would first work out the designMubarakbq +Failed to start please check log files for better information +This done as part of work for but will provide not only the URI building capability but also general improvements to the way URI templates are processed +crap missed this one +OK heres a prototype patchIll add some tests next +it +We now need to decide if we want to support that form +I made a separate feature request for the time range +resubmitting +defer to so I can release in JIRA and get the release notes generated +Daniel I am taking ownership of this until I can sort out the issue +Added test cases for casting of null constants in comparison binary expressions +Resolving +Ideally as markdown tooI will commit that now as a start +Moved from unsupported to plugin on trunk as of revision . +Not likely to happen for CR or even GA because the missing piece is inside Eclipse it self not something we can fix in due time +Hope this works for youI do intend to provide subsequent patches with tests and a unix version of the script +Manik Let us work on this +Wrote a simple integration style unit test +Rebased +It will only ship with BRMS. +When has an error doing IO then it reports an error against its to +Any way we can ensure its done safely? Is there anything I should be on the lookout for other than a simple svn delete +Juca if this is covered in a different issue then please close this issue as a duplicate and link this issue to the correct issue +The features were changed in a web client and communicated to via +Thanks for reviewing patch vc ystein +It is too complex for what its testing and doesnt even test what the assertion saysFor now I propose adding the three test cases to the exclude list and fix this later +Attached is my proposed solution to this +Reducing the left margin of the page master seems to proportionally decrease the degree of cropping +the userdata you use is invalid encodedtryuserdataa which is encoded from hadfaadfadsf +Looks like Seam is not usable with Jboss Tools for war deployment type due to the workaround is to use Seam for War deployment +It was patched in XBRANCH but it was never merged into the trunk which seems to be the basis of all subsequent releases of +Thanks for your interest +Please have the mail sent from jira orUsing the name of the list makes it too easy for spamto get through +will create an patch later on ofc is already including will require to be comitted first otherwise the Cloud Tab will not work +The reason I sort directories by name is because directory modification dates dont always update when their contents change +I agree it would be good to do +Thanks Ravi! +this is very very unintuitive and hard to find for joe average developers +But we will support it in for DOM L load and save +since there seems to be a lack of discussion im closing it +I think it is possible a more generic still robust solution but need to study further the sources +AFAIS not all of the prototype code is removed +Patch for revision on trunk and on branchxThanks James +Are you sure you need File Install? It sounds more that you want a method to just install and start bundles from a directory and get going +This should take care of the memory leakFixed in trunk + +Ill backport the fixes i made in master +Committed revision +Did you know you can give the transport registry a url and it will return the approriate transportHave you looked at the class at all? I think if you combined your ease of use with some of the stuff thats in the xfire tree it would make for a more coherent client do you think +Combine input records and Combine output records are shown in the countersMay be it would be better to have counter for combine output bytes also because that is the size of the intermediate file generated by map task +I have done a minor improvement in ebay storeIn the ebay account added a screen to add catalog to store so one can add the catalogs to ebay store +Help requests belong in the forums. +EstimationIt is not a high priority task +In the case of error message increases content width and height that changes place of other component and shows scrollbars +Forgot to mention Im using the quartz plugin dont know if thats causing any issue or not but I need it +Thanks for tracking this down John +We do not keep the original configuration around +I think each Node object takes up bytes right? So the short lived linked list would be instead of MB which is likely to push it out of the young generation +patch at rThanks Olivier +And user can write a plan to import packages required +I dont think we should be arbitrarily keeping classes together like that thats what namespaces and directories are for +I think that documentation will help but it would also be nice if an old server could be detected in either to alert the user or to silently fall back to XML format +Mancinelli for uimaFIT +I have fixed enumeration namedordinal type discovery in Envers code. +Resloved on HEAD User vlada Date Modified srcorgjgroupsmux Log Revision Changes Path srcorgjgroupsmux Index relevant files to branch. +Added a sequenced date to the a patch that provides enum and annotation support +Tested in msvc with warning level at a few issues on os problems fixed in trunk +If we are not using it here I think we should open another issue for that +I dont see this as useful for our usersIf we are really worried about them using libraries and headers from different configurations we could generate a symbol name based on the configuration +Next issue for reference manual was resolved during translation +Doh +However C explicitely states that a complex is considered to be an infinity as soon as one part is infinity even if the other one isSo it appears that the handling of special values like and infinity in the Complex class is not compliant with C at several places +If I understand the current implementation correctly a should be added under srcmainjavaorgapachehadoophbaseregionserver and implementations would be added under +With the create topic ddl some of the broker configs like probably dont make sense anymoreI will create a jira for each item to follow up +This issue has been superseded by the linked issue of +Well done Evgeny this is no extra fast even with big program files. +x +NOI certainly do NOT want a MB max size +But then maybe Im insane +Hudson is successful closing. +workspace under cant reproduce under without runnign into so please remove as an affected version +and the typedef for endl should be removed as it is no longer necessary +Sending srcmainjavaorgapacheservicemixjmsendpointsSending srcmainjavaorgapacheservicemixjmsendpointsTransmitting file dataCommitted revision +Hence the patch. +Will open another to see if we can handle the stale cache in +Ive also checked via fisheye for and am not seeing any changes or anything that would indicate this was fixed +JimmyYour suggestion makes sense +Can you provide this +Text and HTML being two obvious ones +Investigate for SOA should be a simple change +Im sure Im logged in +Making the parallel builds work gave us the performance increase +Uhh well since you havent revealed who the person is Im not entirely clear how you expect anyone to do anything about it +Lets try this again for trunk at least compiles locally this time +So rather a cosmetic issue +Found a typo above should read where getData wasnt handled in patch v +Scheduled a meeting this Friday to take a look at this with some other FB folks and get you more feedback +The in commons codecperforms over x faster than the Sun implementation +It wasnt committed before Hudson ran but it passes on my machine +As it is not yet possible to unregister node types this test is just for a one time use +more optimized version with more local variables +And given that it appears connections arent reaped until garbage collection occurs would mean that theres always a potential this component could run the ESB out of lets face it a lot of components establish connections to the outside world and each of the Jar files loaded by the ESB also consume +For example I wonder who should host this code +Any ideas on this one? Thanks +Yeah lets shoot for px with it still looking good shrunk down to around or px wide on mobile devicesYes I was referring to the HTML logos +all done +Attaching the stylesheet in the description +Are we ok to close this ticket? I see there has been a lot of work done to related to sync +HiThe issue is +Ill send you the contribution to be added via emailAgain thanks a lot for your helpDaniel C +Error logAlso I wonder if there are other projects in the workspace that are causing this +Support for allowing custom deserialisers to define behavior for deserialsing nulls would be great too +Done +JIRA Triage meeting Tuesday March agreed to modify the behavior to NEVER output a newline after a PIAssigned to Brian M +Shouldnt this happen in not +Anyway I will keep this JIRA issue open until I have written a regression test +Yes thats what I was thinking about +Bela please correct me if Im wrong but I believe the purpose of this test is to show a class in has a bug preventing our using it for READCOMMITTED semantics +This is reasonably clean now and passing the CT so I am closing it. +The modclustermanager pages are now using instead Domain. +That is indeed a useful link and probably should be linked to from this documentation pageIt would be great if the page either had a Prerequisites section or there was a lead thereIn fact I am not able to find the iPOJO Factories Principles from the navigation menu + test suite passed with patch +Assigning temporarily to me during the backport +The problem is in the plugin name +Regression test case +Mylyn +Greg I think the issue is still relevant as I can imagine there are folks who want to process data with +Committed to trunk +to understand the example +Also I reenabled the disabled test in that tests for master session expiryAll the tests are passing on my local +I had to adjust the unit tests to Junit but otherwise the changes are fineThanksChris +Attached patch for talking to secure hbase server +varloghadoopdir mr is rootroot dir hdfs is hdfshadoop dir root is hdfshdfs and are hdfshdfs I have some time to work on this if the current +No I dont like this kind of solution its too restrictive +because patch has been provided +Libraries from EAP works +A couple of comments +I cant find it either +Sorry user error on my part the files are different I managed to mistakenly wget the same file twice +Dan Murphy mentioned to me that there already is a very similar test in the CTS donated by RobbieI think the three of you need to sync up and figure out how best to leverage the CTS for TuscanySDO BVTThanksFrank + feature +Users tend to never enter a query the way I want them to +Broken on trunk after commit of +I believe whats happening here is that you need to have the guavahbasezk jars on your classpath when you start pig +As a result we were only able to apply part of it. +Updated preferences page +Allows a preceding backslash to escape a comma in a file name +Attached is a patch +is for javadoc +deployed since the deployment can happen afterwardsIm not sure on your meaning here +There are a few problems Maven local repository directory is hard coded so does not work with custom Maven local repository directories The custom Maven local repository directory was not being passed into the Maven commands for installing the artefacts +Is there anything people are prepared to compromise on here +Its a hint for the renderer actually +The plugin should honor inclusions exclusions +Modal icons also disappear if +Maybe does not fix this special configuration +This is now implemented for. +I have committed the fix to the trunk with change and backported the change into the line with change +Looks a bit odd that commons net and JSCH is needed +move the top level and classes into a subpackage named server +For start maybe you could tell us what do you expect to have from the Mahout side for what and how you are going to use neural networks +moving to update from submitter +My test code is checked into the issues repo under +Cheers now also committed the JWNL and Morfologik based implementations +bq +You are right we should use and +Fine I will wait for a fixed version and will then provide a patch. +Can you regenerate it +the class extensions and new classes summarized in a short class implemented physical xplain schema which is manifested in the system logical schema of the xplain catalogs in a summarized extended protocol which was file documents the new system procedures and their parameters +I hope were able to release it in the next days +In case of the error reported in the client does not need to retry at all and can simply declare a failure +It was released with the fix in +Here is a patch +Will update shortly +I think the plugin should stay with jfex to get proper attention +Any and all fixes to these are welcome +I will let this patch sit until he returns or we get answers to Knuts questions +One other question though can you possibly expand a little regarding what type of licence headers we would aim for users to be able to configure +We should try it out +There is another bug in the Installation Guide in the section Choose which optional components to install after the Procedure +The cluster has been up for months nowI think a combination of large heartbeat interval and busts of deletions trigger this +I agree with Grant +Thanks Harsh I resumed the comment +Hi EriksIn rev +Hi Harries Would you please provide the code you are using to add port which would be easier for us to debug and testThanks very much +Oh I submitted the corrigendum Carl Reed said it looks good +Move it out of next release +Please review it +Hurrah! +Review ok had a look around some of the commits +Add the option to your should beAdd the option to your not +However I need to deploy my application to a specific servlet context and apparently that cannot be done even with the unmodified +The code change is simple but the impact is huge! I would like somebody to review this patch but I will also post performance numbers on it in about a week +The missing test was about deleting a resource for which a working copy is available +ok this was my bad i applied the patch wrong +For most web application it could be easier to rewrite the interrupt near statement than session perspective +As Jakob wrote The status for the Adventure Builder is that the application is running without errors +By user reports with this patch applied the issue does not seem to be happening is actually related to the thread local used there might have been using another factory that actually generated subclasses of the base JTS geometries which would not have been working with the GML encoder without this patch +As discussed +and versions of are now available +Carl thank you for giving me the opportunities to work on it +Rebased and merged the into the master branch. +Another option would be to use Java think I came up with a real use case for this It would be useful for Development or Quality Systems +I also put an assert into get to assert that nobody pulls a toplevel context from the schizo ISAll tests pass with looks goodBut can you change the new assert to say something like cannot access top context when IS is a leaf reader or something? Right now if you trip that assert its like not clear whats gone wrongAnd I think either remove the jdoc on that method or clarify that its only sugar when IS is not based on an leaf readerI really dont like this schitzo IS though Sometimes its top reader sometimes its leaf reader +We are using release you mean that I should use the build fresh option? +Ill add some testsRegarding your third commentSorry about that I did not run derbyall previously Ill update those tests soon thanks for running tests KnutThanks again for quick reviews +Thanks Oleg! I agree this should be an improvement sorry about that +I further notice that it sounds like wiki or wik it +looking into them +I propose that we name it excludeVersion instead of ignore +I tried using different options for startup and shutdown poller with EAP and all seems to work +Committed to trunk and. +this works in at least +It already exists as +Thanks Rodrigo! +is not committed yet but when it is it should solve the problem. +I put a comment on when we upgrade after the next thrift release well probably add an option to support constructor order in IDL order +I see your point so lets leave it like it it and close this issue. +IMO nothing this should be why it should not be an issue with driver? I am using the driver and it leaves the padding spaces +The cause of the write lock timeouts I fixes for most of the little problems are in +Casting and instanceof are indications of poor object modeling +By the way for the patch pending Jenkins +Im in the process of clearing this up to find out whats intended and whether this extension point should even be deprecated at all or if we should use it +I understand it is one that we cant really fix but dont agree with reclassifying the issue as a task and unmarking the regression checkbox +Joram make sure you make a test for this +Please give that version a try to verify that this issue has been properly resolved +yes sorry for not closing it bad habit in other processes we have whoever takes the pull request close the issue. +Patch with fix and test +Cool Shaz +Any news about that +Attaching patch named Ignore I forgot to make the test class I added as part of it +committed thanks Krishna Kumar +note that it is not just the select tag link that is broken +Closing JIRA. +I can also login to Hudson in UI +Press tab focus moves to Delete Selected +I just noticed that the inner join ca wasnt being used so I removed it +Waiting +We seem to be missing the lib with revision of trunksrcopenssoCommit message Update version to here too +Missing was the ab in changes no fixed +What do people feel is the next step? Should this jira be resolved as wont fix since Alexander seems to be eliminating support for windows native paths +Mike Also for some reason I dont have access in JIRA to select myself to work on this fixed version to +This is where the JVM resources would normally live +I tried all these after did not work thinking that it may be because of the version and it is the same error all round +Jonathan I think you are still assuming that the string is a Unicode string or a string of Unicode code points +An estimated time line for a fix would also be greatly appreciatedRegards +Unassigning myself from this issue and reducing to minor +It doesnt start the countdown so to speak until the message is actually sent +Add implicit imports depending on the MXML dialect +for the quick turnaround! Im in the middle of something else now so Ill try it tomorrow morning +for the patch + +It assigned default values for sides and mediaSize attributes if they are not specified as part of the endpoint configuration +Could not reproduce. +I see a substantial amount of flushing in my logs and looks like flushing happens until it should stop +There are no current critical issues relating to signal handling or JVMs hooking of these signals so Im bumping to post +This patch also blocks redirect cycles that include a fragment identifier +This is different from the scheduler way the gateway handles missing files has also to be addressed whether removed by another gatewayESB or by hand +Zhihong add some logic about when to wait and pass Jinchaos testPlease take a reviewThanks +Ive ported my code to the new API and earlier patch and Im not going to port that back so cannot test +I hope i dont have to this on a regular basis otherwise i will report backreopenupdate rules SET description None +Due to some recent changes in Synapse every time a new endpoint is created it has to be explicitly initialized +I think we can move to Xerces c without any major changes to the system +Seems like the value attribute of uiradio could be used though Ive not tested +Proxy services sequences endpoints etc can now be hot deployed and hot updated by editing the relevant configuration files while the ESB is running. +This ticket contains a longer discussion about the topic of backslash handling +Verified +The changes are in SVN rev for the trunk and Ill throw it into the branch too. +Is this still affecting AS +increase priority for bugs with a and fixed in trunk. +Do you need more information on this problem +After reverting your one commit I rebuilt and everything worked +Yeah me too +eh yes that is what the exception states Null value was assigned to a property of primitive type +Fixed in SVNWe will log an error if we detect overloaded setters +Some lingering links to clean up stillCouple issues here +Here is a better patch file +batch transition resolvedlater to closedlater if you have new information or a patch that addresses the issue you may reopen +Looks like Hossman has also seen this same issue with +The upgrade was what motivated me to make this ticket but the tickets focus should be on ensuring the upgrade from goes smoothly +rebranching from trunk has eliminated the need for this closing all issues that have been in resolved state for months or more without any feedback or update +This issue also causes just a duplicate of and therefore should just be closed as such +its fixed +If its good to go it should be in for sure +needs fixing for just do the tab setting the split screen option is minoroptional +GHOMEvarN Nth server instanceand if someone wants to use something besides varN then thats their prerogativeThis way we dont have to update all the TCK automation code or the docs and users of will be use to the layout + +databaseTypeh +Hi Ken correct +bq +Please take a look +Although it was hard to find MDS as it is not listed in the download section of Blackberry +The revised patch handles corrupt rows with a warning +New patch incorporated the comments from Doug +This has been fixed on trunk by implementing Cipher Strings more correctly in. +Ill load them into the wiki instead +Snjezana is there any progress on this issue? Could it be fixed for M +Hi IvanI cannot test it +This issue is no longer relevant for it is fixed inChanging the affected version to +here is a stab at adding give up support for stopping a route +Ive only seen this happen once +If I understand that well the proxied instance is not actually serialized at allso we are safe because does not invoke any code which wouldnt be part of CDI context before the readResolve method is called +It looks similar to the stack trace for +If this is the approach you are suggesting I would suggest dropping support for Derby and simply replace it with Hsqldb +Patch attached +is the actually on the classpath of your project? why is it not inside src so it will end up in is the structure of my project +Bulk move issues to and +cool thanks Luke +Dont forget to send out an email to everybody so theyre aware +I dont have much of a preference one way or the otherJust that we have need for controlling what key should we split at to be able to certain parts of the data that belong to the same tenantNote that already determines the split point that part was not just not extensible and did not work correctly forThis is a small change to make the existing functionality more usefulNot opposed separating the two +We should try harder here to get the real message +A configuration management system that needs a configuration management system to install sort of defeats the purpose +Attached which enables the package private test to be run as part of the ant targetCommitted the patch to trunk with revision Note that the previously attached patch a is still uncommitted and up for review +I believe we found a different solution using the. +With version it doesnt work +in r created get Im not sure what your actual concern is about this method +But currently were in sort of a limboAnyway I dont think its related to that issue and can be handled in a separate issue +I should probably clarify there that the behavior applies to embedded mode tooNext week Ill file another patch +Thats why the controller and are generated in your application so you can use them as a starter but customize as necessary. +Instead of I chose though +Louis has suggested moving to joda time +Hi DagThanks for your continued attention to this issue +Verified on +Added additional Solr factory tests to test parameters +Check Appendix which shows how should be handled when is used +The only workaround is to add the server url to the list of not authenticated +Edit release note for publication tests not. +This issue also leads to make the Undocumented API rule generate some issues. +However thats not the reason for my commentPlease remember to click the Grant License to the ASF for your contribution +Thanks Reto +The one functional change is that extract now call extract rather than having duplicate logic in each +Updated proposal is to clarify that BBD should occur after all classes are discovered and before beans are discovered +Merged to the branch in revision . +Youre trying to optimize for a test and not a case +Example to demonstrate the issue +Heres a patch which fixes the observed race by queueing up events in INITTING and NEW states bothI was able to get through the map phase of a terasort on nodes with this patch +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +The test methods related to byte array are maked with Ignore +I see attachment in both messages +in rRajika +This may have been a problem with version +Granted apache license to patch +The fix will at least report the problem clearly until the mapping rule is agreed +As least the main thread is not deleting the file while the sstables are opened. +Though noisy its harmless. +All calls has to be plexus independent for sure thats why we have the pojo in the first for the API I dont think the constructor should take the scm url as a argument because then you would have to restart plexus every time you want to change the scm the API and implementations is stateless it shouldnt be a problem with having a stateless POJO said that he wanted to generate this POJO facade but I dont think there is any code for that yet so if we want to try out the scm code in any of the plugin we can make a temporary facade +Thanks for the very fast response +Ill give it another try with derived queries from interface the next days or so +Will mark the issue invalid. +Yeah Im worried about the performance slow down as well +Julien Jon what do you guys think about the above proposal +Yes I think so +I cant figure out why your set of media type mappings is empty +reproduced in branch +Mimej now requires so the consensus is that should upgrade +Aaron I agree with your comments that it would be more user friendly to not force HTTPS on the user and to instead rework some portlets or doc pages to solve thisIll cancel this issue since it is no longer applicable +Tested on EAP ER classes still presentunzip l modulescomsunxmlbindmain grep +Im amazed by your understanding of the server +Marking this as resolved +If you can find a way for me to definitively this Ill backport it to for youI just cant get it to happen +Maven is currently adding the proper LICENSE and NOTICE files to the directory +Can you pl post the for this one +fetching the reference xml is now done right after loading the initial pageeeep +Defer to +No I dont object +Thank you very much for yor a is not posible in our application +I am not removing it yet as I feel full automated testing should be done before taking that step +Move it back in if you think differently +Opened to optimize this +groovyBuilder is present and will always attempt to solution is either to check if the project is a Groovy project or just associate the groovyBuilder with the groovyNature as part of the extension point +To reduce this I plan to load the blame information in parallel +Now we add a link to the FAQ and that seem to have made these questions go away and users now know what to do to shift between themThus im fine on improving this but making it even harder to figure out what went wrong is IMO not a good wayWhat do you suggest as a nice way +Attached screenshot of fixed contract JSP application +add a restriction that ORDER BY could be only used in combination with WHERE clauseChanged to use comparisons instead of using comparatorFixed nit +Date Sender euxxLogged In YES useridIt seems that problem is still in +What do I need to do to have this patch reviewed by lognet contributers and integrated +CarlThe main reason that Yongqiang and I decided to disable this feature by default first is that we have not got a chance to test this optimizer heavily +Looks great! Will commit after testing +I dont mind including code for it but I think as a matter of policy this isnt a good idea +Verified in CR +Attaching patch +Ondrej did you have a chance to read email discussion in attached link? I think your raised questions are a long time ago answered +Can you receive it +Verified in +They reflect different requirements and the reason the JPA spec is part of the EJB JSR was purely political and in the future may not be better name for is +I will create a pull request from github today +Thanks +Ive put a fix in trunk and also added the test case you contributed to itest +Clarify restrictions on usage of NEXT VALUE FOR expressions + +Work for this item is done. +all issues to the CR +Otherwise build time enhancement of dependent entity super class requires the build process to expand the depending jar file content run the enhancer and zip it back up +its not a console problem hence to on the slave it fails for me as well. +The patch allows dojo to serialise the form data +This is how a patch could look like +JuhaniDo you mind to post the heap dump +I would like to see a proposal on how to share a piece of config data between multiple config admin config ids which seems to be what you are thinking of +everything else is fine passes mvn verify on all other counts check style +If you are going to include the paragraph about setting the property to then I would add a comment to so that we will remember to remove that paragraph after we fix that issuectundepthSimilar comments The real point of adjusting is to reduce the amount of time spent preparing statements +I had a similar idea +But Im starting to think that we should postpone this to for example +KoenThe RHDS doc team could go over the tutorial steps screen shots in your existing docs and make sure that they are all accurate as of RHDS GA +I fixed the issues mentioned above +Heres a patch which disables the testsI emailed the dev list a few days ago and this test has been failing for over a month +I guess and have been cleaned up after was filed but the issue was not updated +There is a mix of things in this JIRA so Im openign a new one for Alexs report with this specific test case. +Martin do you mean that you are no longer in need of a fix for this issue? It sounds to me like your new structure with separate modules one for each artifact follows the Maven way of doing things +txt and no unfortunately making it harder debug + +i guess in afew hours you will have access to the tagged files and i will keep an eye on +JacopoNot at this time +Committed revision Thanks Noble +After some other issues I completed intrunk r r r r rI better understand Patricks answer to the issue now +Aparently we have problems in detecting the platform +Or vice versa +Pull request on the way +The more general we make the underlying code the easier it should be to add more template types +Once they finish Ill submit assuming they all pass +Hi Ivan would you please run that patch against hdfs mapreduce and a few other projects? Im still not able to run unit tests even WITHOUT this patch I got the VM crash error here and there so I even couldnt finish one project +Changed fix version to +See the attached patch for one possible solution to this issue +Simon thanks for the effort on this! I have taken a look and updated the patch with a test case and a change to use value convention +on the grounds that it is not guaranteed that +triage for derby +I havent managed to come up with any specific set of modifications that reliably breaks itMy gut feel is that it is much more likely to happen if Ive been writing some groovy code that for some period of time does not compileWhat is the best address to send it to +Finally I think this proxy problem is not related specifically to cargo but to the way the proxy manages local address +As best as I can tell this only occurs when for whatever reason the gets undeployed by before the application does in which case this exception is expected. +no +the metrics for mock looks super awesome thanks for contributing this +I exposed it as a different api but if we wish we could have it be conf driven +Basically Ive done three things Delay the addition to the mapping containers until a map rules is succesfully initialized +This would mean a big change in the design of the compiler to realize this +When using as a source for Cascading we would benefit from having as a comparable so that it can be a part of the input tuple and notneed to flatten it before looking at the data +looking forward to the fixes and perhaps a white paper conf presentation on how you use this +What do we have to gain by mixing state and protocol logic in one class? am a strong believer in keeping state and behaviour protocol processing decoupled even if they seem to be closely related +Probably nothing this year and I am sure they want a definitive answer from Kirill first +request was not matching the wsdl +If the returned type had been some flavour of GML we would of asked for the content and placed it in a feature viewer etc +Just to clarify thingsIf you say boundary event as in then it can only be attached to Activities +this did only occur if there were unrecognized properties both before and after creator parameters code was skipping handling of trailing properties if there were buffered leading parameters +I ran your test case on and it works as expected. +In this case though I think the service group is crying to be tossed and as a +Patch applied on and trunk thanks +cells and r this patch shouldnt include both WAL and are good enough for me +I did a fix based on provided patch test added too +Will investigate further +How do you usually buildf Cliff? Its not referenced from +csantana did this a while ago were using which doesnt buffer its output. +Included a workaround that were using until a permanent fix is created +It looks to me like remains to be updated +Instead it uses with the default implementatinos throwing UOE +Sample EAR app to reproduce the trace from the console of as the patch was applied. +mvn and +No error +We should also document our expectations of projects in this regardI think requiring AL for project websites is appropriate +Finally filtering understands source encoding as target encoding as this is the encoding it converts files into if it understood it as source encoding then it would not have to change the encoding. +Code looks good +However if you have tons of regions in transition it is possible there will be more events for more than regionsIn your test how many regions are in transition? How many regions are there in total +When its ready Ill attach it to this issuebq +I was thinking that our updated script should also create the branch with these removed files and tag it +For examplepathtoandroidsdkYes it is normal for the +Attaching trivial patch that moves the call to update at the end of mergeSchema +reassigning to myself to revert the change in dialectStrong please resolve the hibernate jirawe dont need to revert the dialect as it has been fixed in code +Hi Maximillian thanks +Thanks +Not worth carrying issue forward as a feature request. +In particular we may need to remove the SOAP stack from local calls +patch to final build +This kind of info will be really useful in debuggingIf there are no more comments Ill commit in a day or two +TODO Remains implementing focus at application level for multiple application for Testing +Thanks I just committed the patchIt seems like the admin core thing is a weakness of the current design +Im pretty comfortable with this solution and would like to commit in the coming day or two if others want to review it +openejb is exported by and imported by +I volunteer to work on this issue +a new diff for reviewUpdateBased on some offline discussions this patch takes a virtual column approach to build indexAdded virtual columns in this patch one is input filename and the other is blockoffsetAn example of querying virtual columnsselect INPUTFILENAME key BLOCKOFFSETINSIDEFILE from src +This has been committed to trunk Ill reopen if we see issues. +If this were a scientific calculation then it would be the least precise operand that would dictate the precision +Thanks for the review +However it should still work like the original patch and catch explicit singleton resolvesBut what it sounds like you are saying now is that this partial functionality is actually not sufficient for you afterall is that correct? If so then yes we will have to investigate how to check selected candidatesOne thing I also thought about is that we will probably have to make this check more efficient since it will get performed often if you are resolving a big set of bundles +Passed all tests +I suggest simply remove it since no one is maintaining it +Here I have attached the modified patch please apply from modulescore +Sure youre right but for example to generate a release notes and include it +The APT sources are available in Subversion checkoutsI checked my copy of and it has the sitebuilding directory +Ive linked Remoting shouldnt require addresses to be surrounded by brackets +Committed to trunk +When moving to it may be good if we can simplify this +Attaching patch of changes +Committed the patch to add clustering support for the +I couldt reproduce the failure on my linux platform either +I have partially applied your patch with modifications +The caller checks if the returned value is the same as the default if so it assumes it is not set which may not be right because the value could happen to be the same as the default +Thanks Ted Kirby for this patch +I see the problem is on the delete only and not during create +I am new in Sonar mailing list +It turns out that we cannot piggyback on the so I am making an class to pass the context from regionserver region store store file layers +Rafal Can you please add the subversion revision number of your actual commit here? Thanks Karthick +Do we already have some choosen favorite way to commit an action after a timer +Theyre not perfect but theyre out there +I asked about the throw change because I wondered if it changed API contract for any callers +Problem is now solved as a of implementing and having a proper auth check at form submission time +I meant the definition of the table Projects +It seems the bug is intermittent and it may depend on something odd like object identities as hash keys +Then there wont be work done in parallel with the execution of the test and I think it will still expose the bug that it was supposed to expose but I would have to test that to be sureI commented out the fix for and turned off inthe test and didnt see a failure so if we turn off thetest will no longer expose the bug that its supposed to expose +it is moved and made and intended for monitoring above and beyond scheduling +ErwanThe exception you reported was expected because no Blob converters were developed +I have committed this. +I still see this issue in RF supposed to be fixed or should I use suggested work aroundsThanks update on this issue I do see that it is closed now however still I do see this problem with I am still witnessing the same problem +Here is a patch that restores of OK to apply +For attribute we have always checked only that it starts either with or +Please check +Can you add an exampletest project? The will add an import for any package which is not exported or listed in so if it is generating an import then that suggests the package has not been included in the bundle +This test hasnt failed for a while +See above +Click Upload button and then Stop button +why removing the home icon? I would an other icon but not removing it +Request merged please verify and Ok us this issue is you veryChema. +Re compiler error messagesI get the same error with Sun Java looks like the Eclipse compiler is more forgiving than it should beIll see if I can create a better patch +Thankssuresh +select col from table union all select col from table but it will not be insertable unless it is a partitioned union +Thanks Ravi PrakashI am sure there are many others who will thank you for this +this patch moves the to the new trunk projectit currently contains only one goal file attached to this issue explains how to integrate the plugin into the test phase of the xmlctf am getting exceptions namely Failed to find compile method +If you create the directorysrctestsxmlsourcegeneratorExtension and add the three testfiles to it you will see the test fail with the current you then apply the patch the test will pass +Logged In YES useridI did a global change for all xml files including jms +I am using on +Turned out to be a more complicated patchAfter a brief discussion with Owen I decided to change JobCluster to share the same which is copied as early as possible +Style wise this patch looks good however you must change the indention tabs to spacesI recommend instead of adding the longpollingchanges call in the code just reuse the keepsendingchanges call and after the call to sendchanges add a check if and the long poll option is on stopAlso Im not sure about calling it longpoll but I dont have a better name myself +KateWhats the scenario you are dealing withthansk DimsI am trying to make axis generated SOAP soap request work with Apaches AJPconnector with IIS server +In the future can you please omit CDocuments and SettingsmwerlitMy Documentsworkspacemyfaces current from your patches? It makes it a little more trouble to apply +It doesnt have tests etc and so is not complete +Committed patch in rev and deployed snapshot. +Its not relevant to this just reattached the patch to trigger the test. +As Jason said the user is using an invalid use case +Can you attach the server log +OK I see that inList now replaces in but that in is still supported +Made a trivial patch against which should also apply against the trunk +I hit homeindex and after chewing for a second it returned +Reopening to adjust fix version +Sorry weve been slow to get to themSo do we still have this issue in trunk? If so Ill take the time to mess w your patch +moving to R +This style could be rolled back to the standalone server so as not to have duplicationOr maybe put the configuration almost entirely in leaving only the dataset declaration part for each service in the config file and let the webapp container manage it have two configuration mechanisms +Thanks Daniel I will put up a patch with test cases soon +If you can produce a test case that demonstrates how you are getting into this situation I will look at it again +Is the patch attached to this ticket official can it be applied and then used to the stubs so the infinite loop no longer occurs +I think this was a misunderstanding of the architecture +It seems like this issue has been fixed in +Implementation of basic result type +Unlike normal job submissions oozie requires a JHS token to query job status if the AM is no longer running +Ahh ok Ipv ddff +Vasily Yes you are right the bug only mentioned the copy constructor and the original patch dealt with copy constructor only +We would forcibly set this variable and leave aloneTo summarize In we would change to +See attached file to issue than work will continue on that issue +txt file is a patch for and the +Im surprised it used to work +Im currently studying about Synapse and trying out the existing samples +This patch adds the routing slip EIP to camel +Closing this +The first version of the patch renamed +Carlos please close if all is well +Thanks for quick fix +Where should I see the exceptionThe content of target isTest set Tests run Failures Errors Skipped Time elapsed secBtw I am working on a linux workstation +Sorry I forgot to git add +I need you help on how to run the single test case +some thoughts AM decides the Tasks +minOccures to as of Deepals suggestion +Seems like general Embedder issue. +This sounds like something we could document around for this release +One specific use case is finding the distribution of a directorys data blocks +However this patch also needs some updates now +I noticed that I could simplyfile the implementation and use instead +Attaching a patch for this issue +Commonly used utility classes are exactly the kind of thing you can test without mocking +Cancelling patch since it no longer appliesAlex it seems to me that this use case might be better suited by analyzing history logs but Im not entirely against itI noticed lots of watchers on the ticket do others have some thoughts about this idea? +Exception is thrown in afterPhase so another afterPhase isnt called in which to flush the conversation context so faces messages are lost +aOne is to make sure that atleast one operand has collation derivation +Applied the patch. +compressing the fsimage avoid downloading the fsimage based on checksums storing only the last component of the pathname in the fsimage etc. +Not sure what could cause this and so far its been hard to reproduce +was integrated into +Thrift was designed to make it possible for things like this to work properly even when the core is completely unaware of them +I just changed the donated plugins to use the dojo plugin bundled with Geronimo +The same is true for +By fork do you mean or even running system jruby +Visual Studio does not support C so had to fix a couple of declarationsLukeBinglin if the addendum can be rolled into Binglins v patch it will fix the Windows build breakWith this fix I verified that passes on WindowsAlso reviewed the v patch +I believe prev curr is intentional but not a bug +Will submit after deps are checked in +I had a discussion with Hairong about this one +The solution for all Oracle LOB problems is to buy the JDBC driver for Oracle to use plain JDBC LOB handling and to send the bill to Oracle. +cleared patch available flag since this patch is not yet ready for no longer deal with compression +Another thing we should check is family existence +It still not complete yet +Asked Jive support what would be the problem +An alternative could be to only document JPA XML in the documentation +I run it locally and the exit code is new addendum patch +is super old +Can you try now? Sorry dont have a windows machine around to test ve committed a change to the way Grails loads classes this afternoon that was causing some oddities could you update and try again +Renamed testcases to Added testcase intention doc Added testcase assertion intent doc now use encoding Added missing ve applied the nd patch and in a separate commit changed some minor details like APL headers etc. +Hi AndrewI work in Twist dev team and we have twist scenarios which are mapped to backing java or groovy code the requirement is such that both needs to be in sync so if scenario is refactored then corresponding groovy code needs to be modified and for this purpose i needed groovy plugin to support rename participant callbacks to include scenario changes in the same for your time let me know once you commit the changes +to trunk thanks Kelvin Kakugawa for the original code. +Could you please clearly state what features of the testsuite harness you want in and which not +If you dont do it I will +I cant reproduce this issueI am not sure what you mean by Browser is not opened for JBDS Project examplesDoes it mean the browser is not opened the server is not started the browser is opened the server is started but the browser doesnt contain a correct contentsThe latest issue can happen when Eclipse WTP opens a page in the browser before starting the server +is a HUGE issue +This is going to be on the best effort basis since the Free Cas Queue is not known until the first repy is received +My version of karaf for the update Im gonna fix it asap +This patch is intended for version +Sorry again +No +Bikas doneHitesh Ill track the MRTez translation as +The fix was committed as part ofElena couls you please verify that the problem is not reproducible any more? +MS found in latest runs +The changes make sense still a little complicated by sasl wrapping but we still get a win I believe +Is this patch applied to by chance? We are experiencing this issue with regularity on a node cluster attempting to load GB of data +patch p should handle the Mathias +Resolved with new documentation +cannot reproduce this issue even with your page +Any idea how to accomplish this +OK so you would rather that fixes both this issue and the List issue +loads lazily but needs to be able to get the actual IO object out of to see if I can come up with something +Will commit if tests pass +how about adding a known issues and possible workaroundslinks to workaroundsAlso just to clarify Are these readme files be consistent across life cycle +Thanks Jarcec! +A long time ago we had a history column in +I dont see areason for a driver to complain unless its spec +I just committed this +I havent encountered this on windows +done works reasonably well now and the problem of the double ided elements for a fraction of a second now is gone +I have updated my hibernate version from to not sure if this made a differenceI used luke to view the lucene documents and only one per Address Phone and Contact appeared +Id like to take this issue +this persistence service file will work only when you patch as described +Verified in and ve also attached the file for the execution +As all config delivery is async what happens when the configadmin gets rewiredrestarted during a config update? That may trigger the NPE +An updated patch that also changes the exclusion list to remove all slfj jars +Upload +Thanks Kishan Sorry for the ping but Im trying to bring home now +Good point Andrew + +The api call to DHL to void the shipment is all in place and working +Yes not persisting between restarts is expected it currently gets reset when at every flush +JesseIve committed the patch with minor modifications to the PREFS module at repo revision rPlease check it is ok +Was the test failing of the time before any of these patches? Was it occationally failing before was checked in +The nd and rd patches for need more work +The attached patch wont work except for single column properties +The issue defines the repository part of this issues to be fixed at next issue is no longer can you please create a new issue with your particular problem if this is still a problem +Im confused as to what youre trying to do here +In particular we leverage schema namespaces for providing convenient tags in Spring bean configuration files for example for aspect definitions and transaction +It describes the correct usage and structure the must follow for builds +Much +This is a feature that should still be implemented in the abstractions in package regardless of its not high priority but it is relatively easy to pull off +Please note that I prefixed the method names scandir and readdirr also with axis to make them consistent with the others +Thanks +Trivial patch not backward compatible +Setting the versions correct sorry about that +What u sayI am attaching the patch here +Please could you verify that it has been applied as expected +I have the same think this bug report is the same as the bug wrote a new class that solve the problem of the Cookie class that is not serializable +Ive tried to reconstruct one that reproduces the problem based on the description in the thread but with no luck so farKnut Im unchecking Repro attached since the repro appears toKnut depend on one particular database thats not attached +This is by design we only show the logo on the front page and the login page +I used inner classes because the Revsion class is already inner class and I think those classes are useless outside +Changing the priority from Critical to Medium +And wht is the fix? where can I find the files modified? Thanks +In addition max is set in the environment when launching AM +Now there are JBOSSBASEDIR JBOSSLOGDIR JBOSSCONFIGDIR environment variables to configure the log and configuration dir +Closing all issues +And a empty line in is removed to keep the test pass +Yeah cant cache KVCan we have something for one server first +We have sensitive information and we protect our our Solr nodes with minimum basic authentication SSL +The stacktrace was generated through a debugger it was not seen in the like we definitely need a issue for this now thanks for creating the test +Actually one additional pointI can imagine an extension which really does want to yield the same datum the same Node twice in a row +Its if you trash the platformsios then rerun cordova platform add ios cordova build iosThen the plugins are never recopied +Bulk move changing fixVersion to for any open issues that are unassigned and have not been updated since March Email spam suppressed for this bulk edit search for hossnofix to identify all issues edited +Alexey thank you Everything is OK +Added a link to my blog post when I tried to explain the different ways to connect to AS +Updated patch with similar changes to set the reference to null in +Thank you for the update. +Here is my suggestion +I had added a few fields to the User domain to support some of our other functionality +Thanks Daryn +A patch with an implementation and test is RobbieDo you mind to review and commit the patch attached +When using panel with Trinidad input components you should always set simpletrue on the input components +If you meant the dependency on the corpus cant Ant download this too as a dependencyRe Project Gutenberg good point this is a good source for documents +Modified javadoc descriptions according to the suggestions in this jira +I agree its a nice medium +Ok +I think it is caused by trailing space from the previous arguments which can be fixed defensively by trimming the arguments in Java code and trimming the input at data entry +The synchronization isnt entirely clear within the class looks like some things are both using the rwlock and concurrent collections which always makes me suspicious it probably makes more sense for this to be a cache rather than not clear why we have a and mark instead of just a single limit where we close exactly one channel if we exceed the limit +Attaching a patch with a testcase update +Hmm as long as the JDK implementation of String does not change it would not be a bugI assume this optimization is quite important the fact the the optimization actually is not used is lostwould it be worth to add a test in a static initialization of this class to test the existence and functionality of the private constructor and log a warning that optimization is not usedcheers! +I tried to merge things manually but I am not sure if the resulting code is correctCharlie could you regenerate one more time +ryan not sure if I understand +and still would like to see this cruft dropped +Even if the commision problem remains +Some minor changes +To reproduce you need to use core mapping with files +The attached patch should fix the issue by removing the invalid +From your description it is difficult to see what might be the problem +I can create a process which demonstrates this +Maybe that can help catching erroneous builds +I have successfully run this patch on a production system +In that case it would be all the more useful to have a clean option to take care of all these additional directories +is commented has been the guide added +o print a warning if we are interrupted while waiting for the output from the process to be gatheredRegression tests passed on SolarisPatch ready for review +Thanks for the update Bharath +Thats strange indeed I cant reproduce it anymore +Thoughts +I am nearly sure that all the these version have a working cluster stack + +Interesting syncfilerange is getting EINVAL on the build machines +The main purpose of this JIRA is to support an API to retrieve per query log +Updates to supported and planned features build instructions transaction configuration settings and known issues +I though this issue was long closed +Its what Im doing for now +Ive committed this +RickIs moving the offending classes to different packages is the only solution or are there other options +Rename field to doc +Thanks +Haha no problem Neha +I tried using open jdk but passes +Looks good to me +Reviewed by Alan +Patch file Committed revision Patch file Committed revision +condition checked and exception added +So could you write an helper class that will be reused cross that will do the correct escaping for both standard logj event and the ones that contain some exceptionThanks +This will likely be the internal jboss CI servers at first although we can certainly discuss alternatives as well +Updated patch for Jason! Thanks for your review again! Ive updated the patch with your suggestions +Initial code for this will be in the next code drop +The only other potential issue is that the hdfs loader drops the jars in with an arbitrary identifier included in the name so if things arent cleaned up I wonder if we could potentially hit a collision causing an error similar to what were seeing now +There is another process that takes care of servingdisplaying completed job info +Does anyone have a or a patch for this? Its a blocker for my project +Since there is no filename and no explicit name for the sink nothing will be created +Michele can you look into these failures please? interface sounds better +for the patch +Any chance of getting this committed? Id like to use it in a project +Solution implemented +Add your project to the wiki when you have a chance +That issue was already reported here +It might be nice if we could include a description of the problem as well Ill look at it more to see what I can come up with + +It worked was my faultThere are some more issues mkdirMDKIR +Node wants to read node A +Do we really need a meaningful progress bar for the garbage collection process? Whats the use caseThis sounds like added complexity with little real benefit +we cannot reproduce this at the moment so we are kinda shooting in the dark here you are saying you have other issues pleasereport them separately with as much details as you can provide on it + +Must be fixed for +The other solution would be to copy the storage dir to all the standbys before starting the first NN +fixed for testcase if overriding default property with lower priority +Sorry +Closed Issue +The need to be modified in order to provide data for consistent display + mvn clean compile testto activate the select invoke as mvn clean compile invoke as mvn clean compile test +De sobte apareix loncle fatxa i lamfitri pensa Vaja el que faltava per al duro! Exemple +for the patch +The jar is the standalone version that would need to be deployed to a repo somewhere +Ive stopped counting the numbers Ive told people that they had all this memory thats not being used because they were keeping the configs default +supposedly the filter query was going to be faster for the negative form field at least thats what Uwe originally led me to believe +I dont think we have any Romanians on the team +Subordinate to the primary dialect class +Closed +I tested this patch with +Probably will still get due to javadoc but no Java code was introduced in this patch +More cleanup and made more fields privatefinalI will commit this now +I looked through the svn logs and this sleep has been there since Nutch and I cant think of any particular reason why its usefulIf someone can think of a time when it actually is a good idea to sleep Ill upload a patch that makes it configurable but otherwise I think we should just remove it +Zip file required for musetrunksamples in order to run Manager samples +Priority in bugzilla was minor +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This patch is really just a workaround It disables the closing of statements +The difference is that a SIMPLE auth connection may also use a token whereas today an insecure cluster will kick a client back into SIMPLE auth if it attempts to negotiate with a token +Please if you still see it +Awesome +I agree this is not really a bugThough when a user includes the customizer first and the service reference second this will result in an errorcelixinclude error expected declaration specifiers or +Updated patch on trunk +Unit for release +Could you try testing your image with uDig? The image handling in uDig has been rewritten +I am going to roll up my sleeves and do this rightJonathan Im using these opportunities to learn pig +If method closing the fails it checkes for open closes them and closes the PMF again +Im going to commit this momentarily +Heres a fixed take +Let me know +Can we please move this thread to as it is much convenient there to talk about this +Marshall knows it is a tough problem and weve been unable to find a solution for fix for listener so view is awake after changes +Marking closed. +No longer an issue with the new structure the snapshots of dependencies are included in the local repository +Thanks for the catch Kan +Will try to propose a better solution after the Xmas break +All +Any solutionsworkarounds available yet? +It seems that in this case Eclipse does not see jrebindefault library and include it into list of system libraries and this leads to a compilation problem +Snjezana plz provide additional info to verify this issue +is an updated take on this idea. +Committed. +has shipped moving to +Looks good to me +Attached but its not debug enabled logLet me know you want that i can enable and automation agent log please search for in both agent and MSCould you attach the around +patch for +New patch with a test case + patch looks good +If Oleg doesnt check it in Ill do it later this weekThis issue should remain open though as there are a fews things left to figure out +So it wires the endpoints to the servlet that cxf is startingSo I think we simply need a more refined strategy to choose between jetty and servlet +Sounds reasonable +Well this if you need to but for now +The user could always disable caching if they didnt want to use ehcache or plug in their own implementation based on something else +Attachment has been added with description Patch to revision of branch Temp +Using it in webservices integrating into is another one +ie +Now that work is well underway with AS all previous community releases are + this from me thanks! +For though its fair game +When we want to show the active activity instances of a process instance a custom process image is created and rendered +In general it will help us to review the patches if you can specify the location +Is this confusing? +appears fine Id like to remove the to fix for version as it doesnt apply +I can not run version as an NT service +And its targeted for June th checked into CVS +I think this should be THE general pattern in all the services but guess fixing thats a superset of this issue +Will take this outbq +See for Jira issue to RECODE ME +This is superceded by by +That did the trick +But Im not strong on this opinion so really feel free to rename it. +Thanks Mark the patch was applied as expected +Thanks John Uwe! +But the DFS works in distributed modeI also come to think its machine misconfiguration +If this helps I dont use Zepto +does that help +should natively pick this test up and run it once this is fixed +Thanks both FTP and HTTP added will appear after the next server update +We reproduced the same problem on the admin are investigating the problem and will let you know our findings +Hi HarmeetSeems that you forget the attachmentBTW could you do something to wrap your comment? we close this issue? +Anyway if you think its a good idea to use the registry for these two options I will submit the patchActually the call to set does introduce as it will try to resolve from registry every option value with a leading +In addition I recognized that package structur of source generated with current codegen are different to the mapping sources of current SVN HEAD +I changed textxmlcharset to match the files +I think Ive been polluted by exposure to too many different varieties of English over the past few years +We just have to install the right module and let submit patch call a URL +We now wish to use that gem in a environment where thin is not available so Im forced to either fork the thrift gem or add a dummy to my gem +Cloned from which is for EAP +suggested patch for with more files +Thanks Jie +To run the cellcounter with default rowcfqualifier separator for table planetjava cp +I really do appreciate all the help +FYI I was able to generate a WAR with your VDB using the latest Designer +Could someone commit the patch into trunk? Thanks +Ill dig one up +Here is an updated patch for the dao split +Makes sense +But that means that util may not reference support as we should not have both good about F because if you rename in svn on a windows system such as F f then that fails +This issue ir related to the following ones I think +In a situation where the Store is replaced or a different Store is used this may not be desirableNo thats not a problem +I cant reproduce it +Im not sure I understand the request +After all that proxy variant is really just meant to be a convenience class +eclipseplugins +Since this is a breaking change well put it into the record the patch I submitted allows you to use either or explicitly so that it wont be a breaking ve got bigger fish to fry though so postponing this to a release in order to get out sooner is OK by me +Thanks for reporting the issue and providing a patch AlexeyCould you provide the patch as file attachment with the grant license to ASF so that we can apply your patch with no licensing issue + +Catching up on Closing my reported issues +is required for tests to pass version applies on and trunk +This works cleanly because of virtual methods +Forgot to modify the unit test sorry about test fixed in revision . +Hi Marco Ill do my best to get this checked into CVS this you have a test case or test cases you can attach that would help speedthings upThanksKeith +Thanks +The code for collation checking for operator is in method. +Regression issue support of HTTP status was changed. +Joins should work work on cloud mode with collection name +applied all patches +Closed upon release of +Patch committed +However it should pass after is fixed. +Depending upon stakeholder feedback well need to expand the hint and the options supported at the translator level. +We had a very strict test for when it issafe to send attributes to the output handler +AndersI know we still need to add a mechanism to failed IO worker threads but this functionality can be added without affecting existing that is after the API freeze +So introducing the new metrics library should not be a big change +Ilya I have a few observations on the suggested patch Please extend documentation comments in the new header file explaining rationale and semantics behind it +I agree with Todd lets just fix the regression rather than adding zlib as a dependency +Ill try to rewrite it +The MR AM is already tracking such things but I dont beleive theres a feedback mechanism from the AM to the RM to help the RM figure out which nodes are bad from an AMs perspective +it would be nice to not lose it and just add a new method with the new return that does the correct delegate method calling inOk let me see what I can do +Well to comment further as I work for a private company we want to enforce the good practice to include on the installation medium untouched original archives of the opensource software we embed in our solutions hence the ZIP files +I assume its the connection to the SMTP server causing it in that case +Id have to look to see if that directory location is configurable +We rely on thrift to do that +pushed +No problemBut you can use the users mailing lists or our IRC channel to ask questions and receive help faster. +Agree +Moving it to the next cycle +Candidate for SOA jdbc driver is inclusive of the rest client +bq +Attaching a patch +yes if you want to upload stuff into the users homePOST uploadwithout need to know the exact location of the user home and the userid on the client +It happens if you first add docs with no enabled fields and then later add at least one doc with term vectorsAll tests pass +I dont understand your comment +problem of jetty not of again +Thanks Hairong! Thanks for review stack +Is the following scenario handled cleanly +Thanks for the job +to the concern about being smart +I have this error under Windows so in this case Windows in no less forgiving than Linux +Manjesh Any chance you can make a patch for trunk? Your v patch has numerious FAILED patchsets that prevents easy testing and review +You can check performance improvement using this micro benchmark or test from +Updated patch +This requires that n be special and therefore quoted when in the boundary case +It would be nice if the submitter of this issue could determine if the reduced durability mode provided for fulfills this request +This was committed too quickly. +only this patch +I think using checked exceptions is a good solutionI would use unchecked exceptions only if Jackrabbit as a whole also uses unchecked exceptions +We did some rough calculations internally to see what effect an uneven distribution of data into column families was having on our network IO +Didnt look for attachmentTest contrasts Strings native with Texts and then construction of either from bytes +Thank you very much for this +escape or some such +Uploading the up to date patch for the current revision of the Rampart trunkThanks the patch in revision Thank you Thilina for the patch +As you can see Ive refactored this ticket to use this capability and. +Seeing this on my projects with Grails by +Id rather get a patch that modify the code myself although I can do that also +Thanks +into the indexes of Javadoc +Fixed in r +DavidCan you advise can we apply this patch even when theres no harmony nativeascii component? In my opinion nativeascii is widely used in JAVA world +The advantage of JSP is that it allows for longer or more dynamic outputs which is not the case here +I left the memory checking in this reworked patch but I think this feature should probably be removed +It was again broken as again changed +Bulk closing stale resolved issues +Graeme how exactly are you overriding a bean definition in the registry there? I assume youre calling register with a new instance? Or are you rather modifying the existing object? Can you maybe provide a simple test that reproduces the problemWe did introduce aggressive caching of bean metadata in Spring in order to reduce prototype creation overhead so you might find Grails still working with Spring but not with any higher version +probably +Can you file a bug more details will be great to have +If this work out we can think of a Flume Agents current state in Zk probably to aid some kind of Monitoring UIs +But my main point was that theres not much precedent and its not very useful to package classes by what kind of class they are +No worries +Yeah this needs to go +Im now logging directly inside auto to gather URLs to reproduce +Thanks +Could you just let me know if what I want to do is possible? possible +Will update this ticket when I figure out whats going on +Passes all collate tests +Committed a bunch more today +I no longer thing we need to involve ZK for this +Closing JIRA +Apologies +Here is the MDB I used for testing +Yeah looking at the patch it was just the producer side not using the get +Maybe a better option wold be to show them with a status of Excluded +I added that in so that I could overload s constructor withoutclashing with an existing constructor signature +For instance The Meridio Connector The ConnectorI think wed be well served by going through each connector a second time to insure completeness before going further +Steve AFAIK thats the recommendation from Basic Profile not the soap spec +Fixed in SVN trunk +In subsection Limitations for Database Manager values some of the rows in first column has most in it +th patch Committed revision in goraamazon branch +Moved to next milestone release +Lemme clean up this mess +Done +thanks +Sorry for the confusion +This patch also makes the message format consistent with the other serialization formats for both key and value +Fixed priority down to fix in main after release +The fixed patch is well in keeping with the spirit of the method and would result in a an easier to use feature +What did this class provide? Is this one of your classesThomas +only me thinks its not a network framework role to do that? only something more high level can manage that the right way +All tests pass except which is unrelated and also fails without this patch +I kicked this off again +I already ed on the latest patch will commit shortly +Looking at LCE it seems to handle this case already since createcontainerdirectories uses mkdirs which will create parent directories as neededIll create another patch which removes the comment and temp files +The quickstart fails after a clean install of IR or IR +But honestly this gives more leverage to force the client to upgrade to +Im I wrong to believe the is only called when the Indexes are been created? The Indexes are only being Read but the issue arises throughout the dayI am sorry if this problem I am having is within the wrong section I wasnt sure where to rise it +for the doc writer to pick up +Hence moving to post from client side using based transport IPV support would be available +The first is that Lucene now has animplementation of PFOR and that shoudl provide much betterperformance and slightly better compression +I dont like your patch but i like the ideaP +Hey ColinScript looks good I had to update it slightly to reflect the latest libhdfs changesLooks like now compiles by default whereas it previously only built when Pfuse was passed we need to preserve that behavior since we support the native build on systemsThanks cannot be built it is currently silently ignored +This is because the patchin was already having some of the changes from but with one more bodyAlso the patch was not taking for some reason +Updated patch looks goodIt will be better you add a testcase to test decommissioning with openfiles +Marks advised me that porcelli and jliu have almost completed the new design and we can start to integrate Guvnor soon +Also lease related structure should store names and paths as String instead of + Fixed problem with DA +I just applied it +Will look look into your comment +Just in case anyone does a quick google putting it in etcprofile didnt seem to work +Logisticx Demo overview is complete. +Resolved in upstream for tag +No more dealing with snapshot changes +I presume you dont want this any more +This is now trimmed. +I guess a batch of K or MB should be much efficientNote that this can also work for uncommitted transactions +Thats now fixed +I committed the patch to trunk andThank you Andrew. +Yes this should go on trunk +Koji you didnt react to my comment that an app would be incredibly unlikely to actually depend on the unlimited behavior thus there is no +I managed to get tests to work at derby by using scripts form old test framework and using unique names for tables primary keys and foreign more hint you have to close eclipse for derby to forget the old tables +It works fine in Tomcat and Jetty +LEFT or RIGHT is a really good idea but the enter key is the enter key sends the form which doesnt work in textareas as the enter key types a unlike in textareas the enter key doesnt provide functionality that cannot be of the enter key I d use the space bar just like in Id like to propose to commit this patch but with the space bar instead of the enter +Short live +Oops I misread the year of Henris last commentAs far as I can tell the original question is not blocking to any current project and thus I would suggest that we resolve this as Wont Fix until we have an actual case like thisHenris proposed addition looks good to me though shouldnt the note rather be in the LICENSE file? +We decide to change the default link in portlet by +I really would like to see this go into as well +Mark could you please take a look at this? Seems like you just got a run on the hive smokes and it should be still fresh in your memory +Committed to trunk at +No tests attached since the changes should have been part of +We are getting the error message in the log file of an active production site running on +Problem resolved with TOC at end +with trunk +This issue is fixed by the following checkin +Choices include a counter sequenceid or the actual timestamp +I second that +The first one is for the etc folder and the second one for the deploy are three configured Instances running of the File Installer configuration for the deploy folder +If there is time i will take a look at these logs later today in the eve +Im curious how common it is to only get in the pipeline when you request +I know that the thing didnt work well so maybe this isnt possible +Note to self is it possible to retrieve the txm from Environment +Thanks Bryan for proposing that Ive overwritten the patch with your suggestionSorry for the inconvenience +btw +I think its ok to CDI support after Maven Project if a cdi dependency was added then CDI support will be enabled CDI support will be enabled only when user invokes Maven Project in masterFor projects CDI support will be reenabled only after a Maven Project Update not automatically after adding some CDI dependency +Will be available in AS upstream somewhere this week +Seems like it might be a good idea in to sometimes not add docvaluesyeah I think we should make this per RIW session not per document though since we already have random Types so some docs might get docvaluesintxyz and some might get docvaluesfloatxyz fields +Start Eclipse +trunk old resolved issues +So all issues not directly assigned to an AS release are being closed + with corrected test +Assigning to Rich to review as an additional test for V +We have no class header nor any explanation about what those classes are doingThe only alternative I see is to remove completely this filter from the release +Well include them in the overnight test suite +But they were always hackyIf youre saying people are actually using the default then maybe its worth keeping +verified with me +Thanks Rick that helps a lot and Im making progress +Before I apply it could you please explain to me why itis okay to ignore upperlower case characters for a German language stemmerNouns are in German so wouldnt the case have a special meaning toconsider before stemming a wordFurthermore would you happen to know whether this is superior ordifferent than the Snowball stemmers for GermanThanks +But where should it appear in the UI? Im leaning towards adding these as additional fields on the Inclusions and Exclusions tab + +Patch committed thanks for the fix Santhosh. +Since both of those components are pending their releases the issue remains Later for Struts +Tracking the namespace of schemas enables this +for less fs lookup +Its not mandatory but will improve the over code coverage +cancelling patch until Henrys comments are addressed +I dont see enough information here or in the case to say what may be going on +and it doesnt hurt if you complain in the netbeans issue +Frederic +Thanks Rick +Simple maven project for a Cocoon portlet with working and other portlet goodies + is going away once types branch is merged back to the trunk +into a database or another high performance system optimized for writesMy primary focus though is to make it possible to stream records into a secondary so that the latter could always keep an up todate namespace state with a perspective of building an HA solution based on that +Without the original source files the artwork cant be considered fully donated to CordovaASF because only the original artist can reproduceedit the source +Push to next Beta release +Found out that the actual purpose of outputFormat is for advertising and in some situations its value may differ from example the output format imagepng stated in the capabilities produces an bit png image but the actual mime type written down as an HTTP header keeps being imagepng hence the usage of both outputFormat and said I keep thinking it shouldnt be like that and would be happier with a single property where the parametrization of the content type could be done through mime type parameters + +content change verified in CP CR version of the doc. +I have applied it in Rev +Waiting for changes in POM configuration on the server +Same fix also applied to +created should prolly close this one +Good point so the actual bug is that is does not detect that encryption is already setup +Thus something has changed between and which means that in effect Groovy is think the code acts as a test +Verified on Build +Jira admins you are more than welcome to remove all old attachments except imagesoh i had no clue my status was upgraded +I was holding off on committing because I saw that is in the process of being pushed out and I didnt want to do my first commit in the middle of that process in case I did something wrong +Anirban indicated that this change should go together with dimas stuff +Any help would be greatly appreciated? +relaxinnoisccchecking usability +im comfortable w closing it as is. +Fixed in after release. +Hi Brian I have add the comment in for those missing samplesAnd Bob I have tested them in jboss tools codes base then I will download the night build to verify these samples +It would be very good if the watchersvoters could test it again because groovy has changed much and there was for sure improvements in jrockit as well as in groovy +Verified with JBT +r +Actually s current behavior corresponds to Ants behavior matches all files in subdirectories there too +Benjamin I used your test case and couldnt reproduce my problem every compiler error was caught +Patch adds a check to to alert on the use of and as well as documentation in CHANGESall benchmark tests pass +Updated in revision . +Candidate for SOA t get back to whatever state I was in that caused this error +rev +patch file for file for dev branch intended for this includes the LATENT result type +Closing resolved issues. +Well Im thinking of enabling this feature exclusively for a verifier extension +Jiri is right some from feature are missing in AS rse requirements +META +jar files to lib folder of Solr instance home +Thanks +Add a inline header to both alternative parts +does not exist in the repo +calls Tomcat Instancemanager for creation of MB instance and injection of annotations +Hi Michael Are you still working on this? we can open a new issue. +This wont enforce anything because if they dont rename it it wont be read it all and system wont complain and it will fail later with other problems +it would change the natural of help requests from why did i get exception to why did i get exception is this a bugIt would remove even more of those silly why did I get exception questions that tons can still get at or whatever it is +Its a standalone class that wont be able to see other members in the dangling part of the script +see subversion commit tab for test cases for this +target files +Which client did you use exactly +The patch worked and for master Ive added in an additional commit to remove the two unused methods and remove some deprecated usages +Chipneed some infoWhat do you mean by version? The one downloaded and and is available on the local Web server +I see was addedits a inner class of Weight in that patch but we might move it out +Perhaps AUTHOR is superfluous? Furhtermore I figured out that some mails from the enron data set do not contain email addresses in the adress lists +Thanks +Integrated to as well. +Im just going to rename the folder to misc and make everything the short version unless anyone objects +The almost silent catch was introduced here +The obvious fix is to add the to the permitted characters +Mongrel works fine +Will try that out tomorrow +I agree that this is not a very general solution and may introduce race condition if multiple threads in one zk client try to update the same znode with different version number then current code will hide the BADVERSION exception +For sun jvm it is in which is added to classpath in but Im not sure it is enoughTo verify it is hudson problem Im running build the same way hudson does but with and I guess it is going to be fine then Ill try sun jvm to see if it is an actual problem +display a number +Actually the You have encountered an unknown error running Maven message is shown only when a goal calls another that doesnt exist +done +We have +I think we can probably close this issue now though we might want to wait till the content of the website actually rebuilds with the fixed links to be sure +Is there a way to make it work with older brokersWe also still see messages delivered twice +A server restart deploys the +Again I was never able to reproduce sonever was able to debug itThese page counts were never meant to be exact to the page and no application should be counting on them +committed +Yes of course +sync on +Changed title to reflect the issue being fixed +BikasThe attached patch is originally created for which is marked as a duplicated task of this ticketThe patch may be a reference so I attached it to this ticket +just realised this is already released! +what is the output of the following commands for youuname suname puname mCan you run the start command using bash x +Closing issues that are still marked as resolved but were part of the release. +icons problem that looks better. +Issues resolved in +Of course this can be done manually but maybe could be done better in the helper class +What immediately jumps to mind are use cases like trivial soap messages complex user types array types I would assume that we are not the first who think about measuring ws performance +Patch is availablesomeone can help commit this patch? thanks in advance +procname is used on all systems as name for the system logThus it cannot be you opted out procname it fails on now +And this not only fails with select sysdate from dual but also for any view or table that has date information +That is a mistake you are correct +pytx doesnt get it by virtue of building on py +Assigning the jira to Jeff status of this issue is patch available how can I submit the patch to hudson +Thanks for the info +Im comfortable with that resolution though I might use aword like discover or determine rather than guessSince backing out reveals no new problems and since itresolves the scenarios raised by this issue it definitely seems the safestYour new test case seems good to me +I am going to treat this as fixed unless we can get a reproducer running or later. +was attached on september th +This was already remove from +the obvious thing would be to just discard buffers that are above a certain size threshold +When the processing starts we traverse over the tainted actions and if they are clean they get added to the active cleanup queue and removed from the tainted list +Im updating the servicemix script +Posting a preliminary version +But I tried with them +Perhaps you should add it to +MuraliPlease provide a link to the functional spec and the name of the feature branch you will be using for this featureThanks +We ought to fix this at some point for the Java backend as well +Fixed +I committed in r . +I think this problem could be fixed in SVN trunk +Except for the hanging of the command running with the flag does exactly what is needed +Attache a patch to resolve the issue +fixing small typo +Hemanth we rm each file identified by svn stat before we call svn update +Wildcarding can be accomplished as John and Martin have described so there is no need for the special or PUBLIC syntax +Looks fixed +I think I see what the problem is +Addressing Mahadevs modifies so patch apply failure is expected +To your question Yes a deny all all rule prevents creation of a federation link by anyone not explicitly granted that permissionWhat we are trying to prevent is users innocently allowing link creation in the absence of a CREATE LINK rule granting that permission +Griffons autogenerated +A slightly simplified looks exactly like I proposed itThe only thing For the approach is not needed? You can check for null so the Bits interface is not needed +As Todd hinted this isnt quite as clear cut as it looks thoughDid some scan testing with various KV sizes and scanner caching settings and in some scenarios leaving Nagles on is actually faster +Then create a snapshot of dr with the same name +Committed to trunk and +Does the build script give you a more specific error above +So +I am closing the issue +Assign to Svante. +I upgraded gpg to it may work or it may not. +Hi looks interesting feature and would like to take this ahead +The guarantees are such that filesdirectories that were created before the snapshot was created will be part of the snapshot +Thanks Jing! +We need to support max too so that users who do a lot commits can reliably replicate without storing too many generations +Hi Bikas beside resending resource request through hearbeat do we think some other way like putting resource requests to ZK store +If thats the case whats our real level of tolerance when it comes to platform availability If the platform is down for longer than X minutes then it will begin to affect the availability ? What is X in our caseAlso presumably when it comes to platform availability we dont differentiate between a platform being down and the agent on that platform not running +Sorry I didnt get to this sooner had a priority issue +I plan to commit shortly +This bug is fixed +sh +all the above mathematical formulae can we then say that the policy listed below can be a good first stepThe first replica is on a node on rack r +guys! if no one noticed +Refactored bootstrapper test to add nodes in bootstrap state while checking for srctarget ve been testing growing a cluster to five +But what if a developer wants to connect to a running remote server for memory analysis or just to get a heap dump by +That makes sense about and Mockito +Either one works for me +Delivered. +Can this be closed or is someone waiting for a JVM fix +Also for the record i think its garbage that some stats such as docFreq just silently return here but for things like sum is a hassleIts already annoying we have to deal with the preflex case here too +Hey there PeterId like to comment on changing the scope of this JIRA and its priority +Patch looks fine +Lets fix it in the patch only add static fields that is not used for anything how can that changefix anything +I am using Groovy you setting Xmx too or just Xms? I am not sure what happens if you only set one of these one would hope it adjusts both but I havent checked the doco +I performed some debugging regarding the rd problem reported here +Some sort of IE caching issueI dont intend to fix this for maybe +updated patch +Awsome thanks +Will check it after installing Fedora Core +You should attach a debugger into your tomcat to see whats wrong +It is probably during failover retry +I think this bug can be closed +Looks good +Ive just tried with GML and it also doesnt round to desired number of decimalshttpservergeoserverwfsrequestversiontypeNametoppstatesBBOXoutputFormatgmlRegards even sure GML is supposed to +Ill try to get one up very soon +David J says it may be possible to support saving these properties to via a andor something like you please take a look at the attached patch? Thanks +Thanks for filing a Jira and taking this up Eric we are kinda waiting for this fix +Same intent different methodology. +If we dont know the issues yet then yeah cant have it do fixup +Per my comments elsewhere I dont think we can afford to break wire compatibility in +Errors on read. +Ilya thanks +Resolving issue. +the new patch should be rewritten to eliminate that need +Addressed comment about NULL handling from Eugene +unit in revision Svetlana I have not apply your patch Ive fixed it in a different way please verify that is resolves your problem +Keith any updates from your side +i guess if what ive described really happens then it will fail +Stupid stupid stupid. +Sending srcmainjavaorgapacheservicemixjmsendpointsSending srcmainjavaorgapacheservicemixjmsendpointsSending srctestjavaorgapacheservicemixjmsTransmitting file dataCommitted revision Sending srcmainjavaorgapacheservicemixjmsendpointsSending srcmainjavaorgapacheservicemixjmsendpointsSending srctestjavaorgapacheservicemixjmsTransmitting file dataCommitted revision +q file and use +Hmm but I need these properties files in the EJB AND in the web module +If you set a slower consumer at one of the nodes the messages are going to build up in one of the nodes +Kurt I tried the juddi jar and reverted my changes in scout +Even the message isnt descriptive enough for now this could change in near future +Assigning to Tom to see if he has some notes on the migration strategy +I am running it on my local system and wonder about how to test this +I also changed some properties Im still hesitant to apply the patch due to the projectId url and filter properties are they really necessary? Couldnt you use the information on the POMGenerally speaking we try to avoid adding new properties to the plugins unless they are strictly necessary +Shelly a tracker on the JBPAPP side to pick up EJBCLIENT for requires an upgrade of to +Due to some loose coupling I had to initialize through a few other objects in order to clear the cacheTested on my second environment and it worked +It seems like it would be nicer if the solution worked for repeaters as well but that could solve things for the slave case the likely side effects are probably confusion if a user see the numbers are different and they should align i dont like that so much +Certainly a too long vacation +Ive committed to the solrcloud branchThanks again +One question though You have several attributes without a type specification +Classes must be generated for these. +PING! Please this is the most voted and watched JIRA for Checkstyle +use the escapeSql function +I patched couple record types recently Ps +If it passes all tests the next step is probably to merge into trunk where the UI changes can be readily done on the infrastructure thats already present +Bottom line wed like to make it work on windows but we need someone to take that on +what do you guys think about a which is similar to but centralized for twiddle shutdown and any scripts that follow +Im assuming that you could close it again as wont fix +It keeps coming up as an issue +Good one Jimmy +That said I dont see how that ticket is related here +The user doesnt exist so it complains about an Invalid argument +created. +This patch extends the to get a statistics object that it can use for update the number of bytes read and written by the client +Thank you +Closing +please commit this to also since its a pretty severe Committed revision +I prefer to personally solve the problem just to make sure I maintain the code standards. +Would not submitting it as a patch pending repackaging it as a maven build w the proper packages be ok +Heres a patch that implements Wolfgangs suggestion +when enough compactions queued up it would reject additional ones instead of waiting for a thread to free up +OK thanks for your +I dont think we have many options but to make defensive iterators for +Hmmm +Closing all bugs. +I am interested in bring this branch up to date with the current trunk and contributing it to Lucene +Pigmix results before and after converting strings to textPigmix queryTrunkModified codeL L L L L L L L L L L L +Resolved by introducing a divider string on both java and javascript level to fix yes thanks I have missed that method +Committed revision Committed revision +Also after you post a patch if it is close to being readytrivial you could also submit the patch +I think this will solved the problemHenrik can you check this work for you +If a subclass does not use the instance vars the class could be threadsafe +Ill get to comment on this +Set now insert schemacatalog if different from one to test since only hsqldb with java supports this notion test case added but not added into overall test because of this limitation +for the patch +Im a bit irritated that I have to waste DB space with a dummy column and pollute my to get around this buggy behavior. +Wicket will use CGLIB ASM and Spring +OK I named it and left as it was +it should be in the clean phase for the war packaging so that it gets called as part of mvn clean +Ive seen this one +I have not tested this exhaustively but Ive tested several components on several platforms and they are all working correctly +I would like to ask for fixing this seems not to be a big the other side there is a difference between working with and with is laborious and tedious +Validation of test cases against R or some other package is also desirable +The cipher suite in play can result in the packets containing the openwire commands to be broken up and the transport wasnt able to properly reassemble them +Probably another manifestation of the bug related to multipart forms that I already reported and is now fixed. +Applied to trunk and branches +Yeah in the YS patch i had added the user who starts the NNDN to the admin ACL +bq +This will lead to problems down the road with inconsistent data users are going to encounter two different results for the same datetime criteria +Dupe of +Looks like we can mark this as resolved as the main bottleneck is gone +bq +This error doesnt occur if either the names match or sparse is set to true on unique indexes +hope someone can helpthanks in advance +v lgtm rebased and committed +By default that cache can keep up to states for views per session +Changing to just be +Good questionIf the one set in is invalid the default JVM locale will be used +classes which I am investigating + thanks for the patch Randy +to the expected usage for mv +Proposed fix whenever the name returned from the register method differs from the name we passed to the server store the new name so we can look it up from the original name and use the name when unregistering +I think probably the latter +cannot reproduce on newer tika version so probably it is not an issue any more +This is still not as will be null if a particular Attribute has not previously been rendered thus we have no choice but to add that attribute to the map +Test failures seem unrelated resubmitting for hudson +bulk move of all unresolved issues from to +The profilerOn x JDT unpacking scenario the output is following bytecode patch at r +You can get it by logging into Probe and then type the following in your browsers address barhttplocalhostproxymanagementtables +chris the run was from the trunk +Patch looks good +This kind of deals with this issueI think real fix is for RPC server to properly handlereject any requests when its in a weird state between construction and starting + +I would love to be able to use for our swing app using remoteExcellent is the likelyhood of bakporting this to EAPmany thanks in advance +See pluginssrcmainjavaorgapachegeronimoconnectorwrapper + +Until then Im resolving this issue. +I think Ive already done this in which hasnt been committed yet I need to verify + +Thanks Sijie +tested a jar made w this patch in place so going to commit and put up new RC +That makes no sense at all since its the same code +According to Dan we can now mark this issue resolved and close it. +This mostly works +I do not see what the merits of this would be +Pushed to trunk and branches +You could fix them immediately without filing a bug about it +A few notes as I discover themThis error occurs every time the nodetool runs repairrepair effectively does not work on this node now +This occurs if you a table then issue an explicit split before loading data +Committed patch to doc branch at revision +Currently I am not sure how to patch this +Hi Laxman can you take a look at this patch and see if its what you had in mind? Thanks patch with tests and comments +all jobs created and run successfullythere is a bug in CDK Subversion structure workaround was needed +i guess i should probably just use T huhstill im curious what you would do if you already had a T +Eric Im not sure why you want to define the serialization format and the transport protocol at the same time? How you serialized the data and how you send them its thingsIf your idea is to lock everyone to one single transport protocol HTTP then Ill put a on it since theres no reason to lock the transport protocol and HTTP is not the one Im using for exampleRegarding the serialization format this should applied only to the datasink file format the final output +Touches the following filesSending CHANGESSending +I think faster is fine since it is a dv codec parameter now +Is there anything here that wasnt addressed in +Fix committed on trunk in r +I just unzipped the bin kitAs stated in the README for the binary kit I didcd frameworksantThe files templatesswfobject and exist +Release Notes field marked Not Required as its policy that component upgrades are not included +Ill buy you a beer if you fix it +Thats my usual environment we dont have any problemI have updated several hibernate libraries in serverdefaultlibWhat exactly isnt working +This is just one small exampleIf an application doesnt want to support byte files it should use a validator or actionListener method to check the lengthThis is different from a blank string in +For this and as well the integration effort goes far enough beyond running the testsuite that the RD developers will need to handle it +So this is like saying someone in the future could write a virus into Solr we better not run code +I guess Ill file a JIRA for that +After all it couldnt be the coordinator if it still had the old coordinator in its I missed this it looks like our test suite always gives the coordinator enough time to install the new topology before shutting down +Master host was marked as Down and the slave became new masterLet me know if any additional steps were performed +Attaching patch that resolves the problem +As for the sorting order since this is a console change it will be deferred +I was bit hesitant to change that main logic +The reasoning behind a list of detected languages was that a more advanced detector could go sentence by sentence and tag multi lingual documents correctly +Not really +make sstablejson error out on corrupt data instead of generating invalid json +Thanks Giri that fixed the problem +OK I think that this solves that problem +Committed +set permissions in without adding exception handling to ignore this exception +Jason I cant apply the v patch +In minimun rack requirement per block should be the minimun of the blocks replicattion factor and +Attaching rev of the functional spec +allowing a URL that changed the host would be a security issue + +On Ubuntu and AIX the cases named above sometimes run for seconds and sometimes can run even longer or just hang if theres no timeout +defer to +Great find Eli +Patch for trunk that stores the token that can be used to find the current views state rather than the state itself to avoid duplicate for and in Trunkrevision Fixed in revision Fixed in revision Switched from storing the active page state object in the session to storing the token for the active page state object +Yes its the same as except a file name changeI will commit to and mark this one as duplicate +Youre right sorry fix will show on the server in a few is now fixed. +If I can summarize this issue we do not package apr as part of RHEL EAPEWS natives but use apr from base RHELThe apr binaries which slipped into the ER build have been removed now in CR which is the right thing to do and we only have symlinks pointing to the base RHEL librariesHowever to have valid symlinks we need the following packages from base RHEL installed +Pingmvn help lists the flags but that is no substitute for proper documentation +So I think Ive got a work around for the test failures on windows something to get us started +I dont think onoff switch is necessaryI think the assertion here is not overhead is insignificant +Please commit to both Lars +For JBM we can allow to specify a quorum size in the cluster a connection fails over the node will try to contact up to quorum number of members if it succeeds then it will assume responsibility as live otherwise it will doom itself as a network partition occurred. +Im planning to have a look over the coming few days +but then perhaps the javadocs I linked to need updating +The other thing it does is give affinity of pthreads to particular processors + +Attaching v patch with license year for the quick turn aroundWill wait for days to see if Mikhail has further comments +This is not a complete patch +Thanks Giri +This changes the regex to allow double and single quotesedit tested and it works +Lets wait untilfirst at all sorry for my bad English +Thank you for confirming the issue +Both the ones still attached to the branch and the ones in wasnt working for me at the time I started this so I did it against +set the property +fixed and made sure the patch compiles with all hadoop +Please provide runnable code to reproduce this problem +thanks +if JSF comes out then old servers could still be able to run it with the right libraries in your projectjust like I can run jsf on tomcat even though tomcat doesnt supports it +Hi SebbMy Mistake i had a foward slash in front of theIve changed it back to the example Kind RegardsSascha Saschathanks for your support your site will be added to the listin an hour or soPlease rsync once or twice a day four times a day is maxPlease subscribe the contact address to mirrors see the for help the mailing list is veryRegardsHenk added +I didnt know this is possible +Hi ChrisIll definitely apply it to the branch so we can see about getting a release outThanksd also be curious if theres another way to handle this relationship in OSGI where bundle A needs access to bundle B but may not know anything about its existence until runtime +Patch for Parser with Junit test +Ive verified that it is really in the bundle but not in my application bundle to which it needs to be added manually +Can you please explain the use case youre targeting +One thing Ive noticed from GAE is that get does return a value but trying to deploy the files leads to exceptions +If you find further issues please create a new jira and link it to this one. +But now I am concerned how new schemas would work with existing and +I am still internally dreading having two separate code paths for every UDF +Both Fedora and +I have added a dependency to this issueThanks allIve read and I was waiting for some feedback from you about how to reorganize my codeI agree that some basic functionnalities should be added in the odf layer and others in the doc layerMay be you already have defined some criteria to make the decisionFor example the functionalities which act only on one odf element should be part of the odf layer? This could apply to the following methods get change get loadImage findI might submit one or two patches for these methods in the current issue or in other ones if you preferRegardsPhilippe +What is it and is it only available locally +please patch fixes existing code which is tested in other tests hence no tests were includedPatch committed. +Committed all the work that needs to get done for this including some regression is much more sane now + for patch +The id I log in with is simply anew and the email address associated with that is anew +Do you mind if I assign it back to me +Pushed to master. +and the dependency to jaxp is commented and it will be dealt later onBesides that functionality as written above. +current impl all partitions equally across the new set of tasks +So this third version moves the worker routines into where they fit a little better +WMS cant be exported to kmlDo you have an example WFS to try on? Or at least a shapefile that does not export properly +Can you edit the description depends on this one +Not yet +Liam helped me understand the point of this on the list +We did some profiling tests with the recently released version of CXF and it seems like we dont have memory leaks anymore as we had with version after a few webapp redeployments +This need to be investigated and might have a deadlock somewhere else in your code +Oh duh the conversion from Unicode UTF and the determinization are in +Thus we mimic svn and cvs like operations +We spent a long discussion last winter over this topic of Job and Step + seems like this problem only occurs with autoCommittrue +Thanks +Build and promote jobs should work even if we decide we want to have both artifacts after all +Thanks Cath I dont think Ill ever get those exports right first time Patch applied at repo revision r +INFO Running for rows took msjmspaggihbasetesthbase cat grep took grep INFO Running for rows took ms +remove the shale default hidden then fine the question is how uptodate the templates are? any idea +Patch looks good +Please help how to get rid of this error +You are on the code right now so Im fine with your opinion to win against mine but Ill keep my opinion +Verified in JBDS L +verified +I close it and leave it only for historic purposes. +Is there a way a user can remove or replace icons in the current implementation even if just using CSS? If not then Im not sure this can wait for post documentation work required +Do you really need a Lucene constant to tell you that a divisor of is no divisoreg Lucene is not using that default in a way that you guys are arguing users would want to its using it to set the thing as off +That was a mistake in my reading of his patch code +Your symptoms are unclear +This issue is superseded by changes proposed here have been implemented in the new class The package will be entirely removed in +Something like a tooltip on the title tab or something +Well need a new issue for this +Sorry for coming late to thisI read the design document and it looked like it is quite a complex feature to implement and maintain going forward +We need to move the files as suggested inDo you want to do that? Or should I +But I think it is okay as a default value. +What project structure you are talking about and what kind of wizards you need to use on top of that +Ive talked to IT and Mark Newton about this and was told to wait until they have more servers m currently looking into an alternative hosting solution and will provide more info when its available +The attached patch fixes the for elements in stacktrace that do not have a line number by setting it to According to javadoclineNumber the line number of the source line containing the execution point represented by this stack trace element or a negative number if this information is unavailable +I mainly pointed this out because I stumbled upon it when working on the reference implementation +Closing all ResolvedFixed issues lacking a Fixed Version +Release published closing. +Actually the patch addresses also the order manager UI +The comment is wrong +u t i l +Clean up shell scripts as agreed in this issue +This is something that needs to be communicated on that subject disabling that setting in the debug settings does not affect the execution immediately +I suggest to use the SSH command line to test this +bq +broken by fix for +NitinI have unassigned this ticket as it is related to your feature Scaling up Memory of running and the fix for this bug is disabling the scaling up functionalityI have updated the ticket with my findingsCan you please assign it to yourself as you have the better idea about the functionalitySanjay +The junit report now generates a separate report for the test results are now also used in clover analysis +Its been linked as being part of so it will get closed when does +I just learned about and was excited that it would just work but then I got my and realized it was because of my missing transitive test dependencies. +This patch adds one more parameter to method which specifies whether to exclude crc files +OK Ive got it +I agree that this is the sort of implementation detail which the Derby user docs generally avoid +Eclipse Smooks editor quick start there only PDF versions of the guide no sources in XML +Latest patch with support for deleting old revs +While the wire was null before the attempt to load the class the get resolved the dynamic import and wired the requester to the providerWhile I dont expect this is necessarily the correct way to address the problem it did seem to work +Yes HAEAPSIZE is not part of this jira +This bug has been marked as a duplicate of +To patch it user must login into and Save changes to his profile to trigger synchronization into nukes table. +Im closing this bug as a duplicate for +This isnt as simple as I originally thought since hints store the original serialized mutation which can cross multiple causing us to have to see which ones actually need to be removed and then serialize the remainder in a new mutation +Access the CA Helper Application at httplocalhost through a web browser that supports KEYGEN tag +Patch is stale at this point +with clean shutdown completely clears them down to Mb +Committed revision . +The book contains an appendix of current state of. +Closing since embjopr testing on this should be done too +Bulk close of resolved issues of +Now you are PMC member for Jira too +Removing from +I appreciate the help +This is not clear to me have you a test case to provide +It may be overly complex having a default package name +May be you need to fix it in the tool to have the correct package name for the driver +Made changes in places +According to the comments in this this is done +We will keep trying to recreate the scenario as the upgrading does not eliminate the freezes and it is very serious issue +Attaching clipped partial stack trace of the alfresco version of the error +This is a bit late in the game but looks worth it +The same should hold on trunkArpit please wait for me to review and test the patch +Perhaps it shouldnt be part of the distribution +Please do because I absolutely dont understand where the problem isWe cant not escape the when rendering the markup +Verified JBDS Betae +fix applied. +is the relevant JIRA issue +Ping +Ok +If not then its faster you adapt the patch to the actual sources for yourselfI will put a comment again here about the stateRegardsLars +specific paths in your file with envReplace environment variablesThanks +You should have handled the case of qualifiers in version such as beta or must be rename by and not by reported this as a new bug +Joseki server can only read RDF datastores and so I am not able to access the content of VCL MYSQL database without replicating it into an RDF store +This because it has also some advantages +Committed revision . +My tests all work without it +This could become an issue +IMO can be marked DEPRECATED +The patches for hdfs and mr are trivial +Whats the semantic difference between and +This has been committed and your content should start to appear shortly +so the identifier is still backward compatible with old version clients +Bulk move of fixVersion fixVersion for issues that have no assignee and have not been updated recentlyemail notification suppressed to prevent token identifying these issues move issues to and +on patch +Or more accurately Cant Fix +Just to be clear that is related to this and there may be some conflicts with it as it changes the name of a C test file under srcc +My guess is that the answer would be no but we should check with usersdevelopers have commit this patch this is my fist time of committing patch hope I didnt make any mistake +The solution is use another buffer and add it to the final chain +This class will be available in and clients would be able to unit test their code using it +johan dont think this should be blocking moving fix version to unassigned +Just noticed this was first discussedreported in the mailing list +It saved us hours upon hours of not needing to write directly to Cassandra. +checking to see if hbase is still planning to have hadoop binaries published for +Checked release notes closing. +class created in revision +Is it looking for both extensionsCan you give me an idea of where to look for the Java source folderjars selector? please look at +I applied the fix by Jukka Zitting to version recompiled and deployed to servers that are having this issue +Thanks +Making this part of which tracks differences between embedded and client has been implemented as never noticed this issue. +If the AJAX solution works well for the authenticators then we may consider using id for the referrer info as well +Then its a new feature thats why i felt we should do it in +Thanks Suresh for your commentsI got a little further with the additional Class approach and as you mentioned I had to create another child of that will behave just like the original one was until now +Committed +Ok +I raise this issue because numerous tests from both testsuites are now failing because of this +Can I go ahead and do the changes +Closed all tickets +Also please name the feature branch that youll be using in the asf repo for a future MERGE requestThanks +I like Roberts workaround +Fixed as part of the work in. +nd patch did not work due to missing implements Comparable of +Thanks RichardPatch was applied to LUNI module at repo revision Please check that the patch was applied as you expected +Fair enough +docsupdated +The only scenario we do anything remotely different is with inflow transations +I tried with the workaround suggested by Devdeep and was able to launch the Management server UI +The situation only happens when the type itself has no default ctor +The problem is not detected on windowsBut the file shows that its there also on windows +This issue is fixed by the patch of issue +Caused by bug in activemq we wont be using this. +I guess in the typical case split load is better because almost all of the splits are the block size +Preparing for release +I like that solution Robert +Looks related to now you will have to disable the crypted url coding strategy during your tests to get it to work +JBDS H L +Here is the patch +However the attachment contains which follows the maven repository installed by the installer the has no version number +lgtm please update the jira to reflect that were also making webhdfs more restrictive which I agree is the right thing to do +Looks like we wont fix this just for eclipseSorry +The combination of initialising the filesthe etc +Couple notes so farIm dont care much either way but I wouldve just parsed proplists from Erlang terms from the config file like we do for other various options instead of creating the keyval syntax goopNever register anonymous config change functions +I think the special empty string value is +The fix will be in EWS CR. +If this requires changes to the code within project svns then make sure you create reciprocal JIRA issues with those projects +It samples values to a tolerance of +bulk defer of features from to +The software on this Linux installation is of different origin no environment pure serverThe version is as on the Windows computerI would not like to restrict the test case fail criterion to failure on pressing the browser back button +Please review +in +As you are right cant know how to use my objects +Downcasting is probably good enoughMy idea was that it may be more convenient to get a number that still resembles the original valueFor example If a long contains billions converting that to MAXINTEGER it is still a large positive number +What do you think about it + image for s not bug but feature request +Greg could we close it +For AMQ apply the patch which fixes the first part of this issue +Alternative patch to not tested with the attached test case from patch releases the lock on the base class and acquires a new one for the extended class +never passes the version info to master +I believe the request is for an composer that handles the conversion automatically so that XSLT andor Smooks can then be applied with zero Java coding +Set it to x +Committed Todds fixup of Ashishs patch +In this case calling it would be accurate unambiguous and tells the user generally how the adapter works. +When I tried extract interface on a class with much less references it worked +Brock so both payloadColumn and are required when using is that correct +They both need to be addressed to fully support the intended templating +This patch fails on both jdk and jdk +Added also it have some trivial changes +Would definitely still be nice so I can pass around the schema object for various usages. +Ubuntu Guest are missing from guestoshypervisor tables too +The problem is that both and need to be notified when there is a new suspect that I should be using instead of +I dont thing this should fix the problem youre seeing +without the taglist plugin it works think I will try to use the sonarsonar goal as a post goal in Bamboo +When we bounce the servers the message beans then consume the persisted messages because once the servers come back up they arent in the being delivered state anymore +ah +I dont either +Next time it would be better if youd opened an additional issueregards phase of the password strength component is completed. +Closed for. +The patch was applied +Applied to SVN thanks! +bq +HelloIm a searcher for an european project and Im trying to use a geoRSS model with a timer to relaod it in orderto build an application for real time tracking of order to do this I started from the demo shipTracks found in mapbuilder read the discusion about memory leak when you call loadModel on a geoRSS modeland since I have the same problem I would like to know if this problem have been someone has some clue +The rest not so important but did it anyway +However I do not currently know of any Java implementations +The downside is that I had to disable the build since theyre still doing their own thing +For me it seems that if you can create the database structure you understood the hierarchy and youre normally able to CRUD itThis bug really blocks me +See to run +The datasource used in the MDB needs access to the instance in order provide feedback in failure looks like this will not affect EAP as far as I can tell +Moving to. +The patch fixes the bug but does this page really belong in? Should we just have the page in the user guide? +Patch adds the line numbers of the offending lines in the patch to the report +I agree with Pat on keeping the changes minimal for line +The test jar is the same as the base jar +Patch files generated using diff u basefile modifiedfileGrails was used as base +it is computed just in time +So changes will be present in EAP to be present in EAP they should be commited to JBPAPP branch too. +This was also the reason why I created a patch in the first place +This exists back to fixing affects version +Also makes a copy of the server address before stuffing it aside into the queue of meta regions to scan +Marking as resolved since all issue has been addressed. +in favor of +when is this going to be fixed? +I dont know if this is already there or not but the is a must have feature in atleast +I should have thought about them and yes it lis the correct fix +Ill commit shortly +The message is that Weld is not getting along very well when theresFORM based authentication enabled +Ill and commit once some basic unit tests are included +Jimmy this ready for reviewcommit? If so please submit +Attaching new patch merging changes in the trunk. +Will do this in a next couple of days +JMS Specifications Section Closing a connection must roll back the transactions in progress on its transacted sessions +However there is a slight nuance +The test is using nativesDist qahomehudsonstaticbuildenveapwhich is exactly the same as you wanted Maybe a productization problem +Not a bad idea +I think would be a good choice to implement that feature +Sounds like a reasonable idea +I like you propositions +Thanks a lot Josh for the quick fix +Verified thanks! +Just missed to mention have also done the following +Since the same comment appears before the second equals method the second method will need to be modified to reflect the same business rule +This would be a great addition +OK +Nice catch in close the changes there look great +Also changing the translator to use implicit widening numeric conversions fixed +That was me committing a change that I made for reverted itSo far there are no Driver programs yet This is only the sequential version +Have made changes to as per Bens suggestions + change looks good to me +I did attach a new bundle collection that reproduces the problemAgain use zip s npe O full to recombine the Ive gotten to the bottom of this second issue +I changed the setting in it doesnt make sense to overwrite the while using a Rolling appender. +Doesnt that still have the downside of DN disks? Is it worth all DN drives to save +You mean there is nothing selected by default or that the list of connection profiles are completely empty +Thanks I will try with your suggestions +Maybe add an signature index to alias file params so that the signature can still be easily to understand for debugging +Patch root isBest RegardsStephane Landelle +I dont even know that I can come up with good values for those numbers much less ask a user to do soWould tmp be better? I want to avoid tmp since size may imply memory size and we may also want to use that name in the future for +Hopefully the last iteration of renaming +note that the fix for this requires function added in to srctestjavaorgapachehadoopyarnwebapp but requires function so will wait to put this into patch available until goes in +Or simply use the passed in reader as delegate of then it will behave correctly for all methodsI agree with the exception but which reader are you talking about +Are they valid tableNames +looks good to me +Ill update the example project showing the workaround +If fails then the bean is not attached to the transaction and thus should not take part in any transaction semantics +Looks like the same issue as in +Ok got it missed this class +This seems to still be an issue with and included in the runtime + +Probably just you discovering the EMMA plugin and ignoring the fast that if we opened for all segments that are not covered we would now be on issue numbers? +I can reproduce could you take a look +Just a hint +Im reassigning to Julien who may reassign further +I added this feature in trunk +First attempt at localization for Redback demonstrated in the example default resource bundles are registered default messages provided in included in overrides the failure messageEach app can register the default resource bundle or supply its own and also register an additional one that is not included in the app but can be provided by the user if desired +Hah the confused me +I pushed your fix to the trunk in fafc +Thanks Yuri +The new rev number is +With only one no problem +If we didthrow an Exception Id vote for +Are you on target for July th +Patch to validate column paths in a slice predicate +FF and IE do not assume a charset of on the XMLHTTP request +patch against r fixes artifact handling during patch against r included missing patch fixed this bug for me +Id be really surprised to learn that anyone actually depends on the current behavior +I am going to empty the open folder in order to avoid sending out all those mails as soon as this is resolved +I looked at the patch it looks pretty simple +I think that you need to create the corrupt storage file in all cases not just upgrade +Moving these all to a Doc release version. +Also I hope this fix will be propagated to the Seam stream as am seeing this problem with Firefox and and Seam am temporarily resolving this issue by putting sleeps on the server side methods that are too fastbut thats a pretty bad hack +The PR should be sufficient +Snickers is dead. +Added the dependency to the photoalbum verified that this particular issue was resolved +Can we close this or do we need to resolve the subtasks first +Otherwise the patch looks good to me +youre quite welcome Brunomarc +I am not sure as to whether this could lead to any security related considerations but is one nasty bugger +This counter helps to identify the context where this EL expression belongs +set the transactions to isolated level and set the statements to concurrency level to stop synchronization issuesthese two patches fix the problem. +Can I close this ticket Evgeny +committed to to reproduce +worse yet if the double quotes where in there i got a hanging of the shell when doing the exampleI think for json the basic rule is double quotes for the data valuedata stuff goes here and if there is json for the stuff use single quotes +I made a research here +I can imagine a workflow solution where setting all parameters on every step get onerous +Committed at subversion revision +Unless rules files dont work with I think we should be distributing both and registering both locally +I see no attachement +Definitely possible just something to be cautious about since I dont think a regression here would be easy to spot in dtests. +Patch +please reopen this issue if it still happens to you in the upcoming build +not in Tools +But it is nice that we can get the correct total numberAnd there is no much cost to use long integers hereWhat do you think? Im curious have you guys run into a situation where an int was not sufficient +Rebased on HA branch with following changes some places used to persist blocks only if append was enabled now it also persists them in those cases if the configuration is set the config is forced to true if HA is enabled improved the unit test so the writer keeps writing after the NN restarts and verifies that both the data written before and after the restart can be read +I just merged the to master and maint. +I would like Nunit for testing +That is concerning +That is very strange +For if the test above passes it should prove that detection properly worksCheersChris +Amos Has this been done for the docs yet +Thanks sounds good +did you try what was described in the blogAnd not sure why you say it is quite impossible to be a memory leak when the error says Java Heap Spacebtw +applied to M so resolving. +I checked the stack trace and found +I did a full clean and rebuild and dbsetup after importing +why an option +QA needs to specify JVM property to setup proper versionThis file name can be investigated by entering the following folderEAPHOMEmodulesorgjbosswsnativemainFor example I used this commandline to execute CXF core suite against EAPmvn Ptestsuitehudsonjboss homeopalkasvnjbossastags +Unless I applied the patch from the left table looks wrong now +Im attaching a slightly modified patch that adds the same feature to the jjtree mojo +Thanks for early review Todd +Hi updated patch to delete repository and directory purge configuration +Patch to Nikolay Petrov for this fixI wonder should the new be prefixed by an underscore to really make sure that this is not part of the public API? Its already so its ok to add it but maybe we should have an additional indicator hereregards patch due to JSF +Any chance you can include the new files too? See the top of the patch +I attached a patch with a fix to which introduced the problem +Changed the to the token and ignore the entity referencesAlso regenerated and committed the fileFixed in revision +Did you intend to attach the patch +when accessing the admin portlet to add a new user i got thishttpwxpastebinshow +Also take a look at what can be closed here please +Thanks for making the changes looks better +And soon we should ship with h which will also help +with CVS +bq +This bug has been fixed in Xalan Java +Id like to create a patch on but that Jalopy crap has caused some files on my machine to be styled and others on the CVS server to be styled +Prototype of synchronous API to remove watches +Added to portletswap but not the distro until we have themed page groupings. +Plugin will use sources that exists in the local repository. +To be clear the issue had been tested with release in combination with Spring now upgraded to Srping and Spring Data Mongo +The problem is that get returns null in newer versions of android we should probably test for null where we use it +Thanks Nicholas! +on patch +The basic idea is to be able to store state local to processing of a particular previously Hibernate had no such store +All fixed up +Sure any feedback on the issue +Bilginyes you are right +The balancer change is in +sounds like to a good place to start looking for an answer +never mind I got it reproduced now +Patch for site reference to +Any more details +canceled patch since its not been reviewed yet +We already have parsers for JPEG PNG and TIFF +Rerunning regressions +The jar is what I ended up publishing rather than the minimal netcdf jar +Sorry about thatThe fix is small enough +Oops +Were now experimenting with multiple parallel per server to stripe across more disks getting another or so from that in some tests +Fixed +In draft the prefix is changed to +This issue does not affect Teiid but Unit Test should be migrated +pending outcome of discussion of infringement claim download of tarball has been disabled. +Search for the where id couldnt find a doc nor tombstone +How does it look under mac? If unusable we would just need to live without ctrls working for we know why this ctrls is affected by using a different combobox control? any eclipse bug id +Verified +fixed +Because if the textbox is disabled user should not be able to perform any action with this even scrolling +Thanks Stack for your commentsI am working on the patch +Bug has been marked as a duplicate of this bug +So lets wait to make use of that api once in +So I vote for leaving things as it is now +Even for this simplistic initial system something better is called for +Colors for methods and fields are in the Java syntax coloring page +Fixedplease test itif its OK now change this issues on this issue I found +IIRC it doesnt compress references by default +Direct support would probably be better than relying on a rd party module +Can you give it a try against Solaris and see if that fixes this issue +Thanks Dmytro +Take a look at let me know if they resolve your issues +Wasnt sure about that so I tried removing orgsljfj again and made sure to restart sonar after removing the dependency and got the same error +No its not a transfer issue please download the file and see +I am wondering if there is a way to keep the old way of specifying timestamps as an alternativeOf course its fine that if someone wants to set operation attributes that the old way wont work for that case +Fixed +Test to simulate the stack traces +What state saving method do you use server or client + +Good to know +You could also add support for custom skin defined by a properties file +We can close this as soon as we fix the two remaining tasks +Thanks Chris! +Closing as per Garys comment +JRE classpathentry first then dependent JARS? +As a first step for this issueI have implemented the UI support for the album subscription for flicker and picasa and store album meta data in a java collection as class variables separately +After researching and thinking on this topic I am beginning to question the use of the GAC with NMS +No functional difference +Well revisit this when we add the support for enablingdisabling XHTML output +If this were to happen again on a slow machine it should be run with set to a higher value. +reverts the changes +it works fine with jute +Oh actually we have to do the refactoring other we could satisfy a read request from an uncommitted transaction +Lets keep the discussion thereBased on these explanations can I resolve this bug +Patch validates if already exist a schedule with same name when update or create a schedule +If this needs to be reconsidered please raise the issue with the SOA PM team +I have also tried it with true which certainly didnt help the speed issue but I cant comment on the hung threads as I didnt bother letting the fetch complete +Cant close yet +linked to optimistic locking issue because they both involve the hibernate mappings +I have attached two screen shots to show the problem +Provides a standard means for obtaining Configuration instances using a configurable provider +I tried making the dependency not optional but this is not enough to solve the problem +This must be respected +Did this fix really get in to +So this uri syntax is mainly the same +The issue arises with fb make not the thrift make +Please reopen if the failure is still seen. +cqlsh does the work for you +Fix is in CVS +galder it does sound irrelevant nowThanks +Youre rightThere are checkstyle errors +Why did you gave it a different nameThe names are up for debate +If we get IT tests for these before is final we can take a look +This broke some user codeFixed problem by adding support for. +dfs +Closing the issue now that it has been resolved. +Thanks a lot Bob. +It may also help to provide separate patches to address them + Having immutable value objects as part of the domain object graph is the essential design property of good OO and DDD agree +In fact there is no EL stuff specified at all +Committed to and Trunk branches +Changes added unittest minor tweaks for getting the encoding of the XML files version new dumb decomposition filter uses a approach by generating substrings and checking them against the dictionary seems to work better for languages that have no patterns file with a lot of special cases Is roughly times slower than the decomposition filter using hyphenation patterns No licensing problems because of the hyphenation pattern files Refactoring to have all methods used by both decomposition filters in one place Minor performance patch according to Otis suggestions for longest matchNext step move to the compound word tokenfilter stuff to contrib +Note this bug isnt present when you use the Tomcat plugin +I tested it under java will be out of office tomorrow will try to produce the benchmark when get back on Monday +Pushed to master +Well ship the fix once its available +Thinking about it Im leaning towards using a Spring bean name rather than a class as this would allow you to use injected dependencies in the resolver rather than having to rely on static access to any stuff you need +So it would only assemble the jars of your modules +So the merge finished and it tried to warm the segment by running searches through IS that was using the ES +I read over it and it looks ok +The retrieved number appears correctly. +I believe it was the Accumulo project who noted that +This is a problem +This involves changes to the jbossesb schema so I would suggest that it be included in the FP rather than CP sounds like a slightly different issue can we have a second JIRA with examplelogs attached +David note that we cannot fix this issue for Resource as this annotation ships with the JDK +Fix has been integrated in head in branch version released. +I wonder if the second example in Reference manual failover should say failover instead of stop slave thoughI think it would be good to wait a little bit before writing more because the functionality still changes +This is not needed for since the webservices stuff has been removed from M +The flat hierarchy at ibiblio doesnt scale too well +unzip geronimo in a short name directory for example cI can run g with the same jdk on Vista without any problem +should be considered for too lets see how big a changefix it is +This would defeat the main purpose of which is to make it so your build cannot depend on any geronimoopenejbwhatever artifacts that it did not generate +Can you open the chrome console and see what javascript errors you are getting +And I dont know anything about FDT +Just ran into this problem as well + +I didnt finish that one bullet +simple patch +It is already released +This is the screenshoot of the purchase order created +Samson could you please set up the directory structure for this +simulate this guy going down and that guy and not deal with timing issues and so onJust like lucene doesnt actually write continuously to your disk until its really full to it mocks the disk fullSure these mock techniques arent perfect but they are much easier to debugfor real integration tests maybe junit isnt even the best tool for the job anyway so i would prefer if these were separate from the unit testsThese integration tests are especially frustrating for lucene developers who seriously dont want to break solr when they change things underneath it +fixed to return full FQDN in GSLB rule response +If the release audit warnings are all in generated code then dont worry about them +Thanks +Anyone +Try it Presumably this is because those annotations have runtime retentionI think it would be ideal if consumers of didnt need this jar +This JIRA is a for Filter based on TTL +Bisect points at specfically the ninja commit in like merge to trunk from was syntactically correct but semantically broken +Fernando can you confirm that this shouldnt go to EAP Release Notes +Does this make sense +ThanksGo ahead and upload your patch and I will be glad to review itThanks have updated the first level patch for review +Attaching patch a which adds a toI intend to commit this tomorrow +Hello EspenI propose this patch +seems like it is impossible for us to reproduce the problem +committed to trunk +As before all the alternatives symlinks are removed but all the actual configuration files remain +Since this is quite a blocker for me I added a test case +file for custom properties for the mysql setup +I did find a few examples of that while running the unit tests and corrected them before committing +need to move blueprint to use the to do this as util was introduced in compendium +Want me to apply +Need the script fix before the packaging can be said to work +However I just posted an email to dev and I can see it via Nabble +Moving to I am nearly sure that code block I added to solve a NPE +Thus it no longer makes sense to keep these issues aroundIf you feel this issue was closed for the wrong reason then please open a new issue. +For the GA release of GPE and STS well move this configurator over to problem that hes seeing with the error is due to GPE Im currently fixing that issue +With this patch applied and with reverted up and down as well as enabledisable is much smootherLet me now take a look at safe mode +The is now also wonky because of LSH +Need to add tests thoughPS These are my first steps with Erlang so bare with me +Its a pretty trivial change +You have a large cluster havent youThe guys from nutch wanted a BSP implementation so it would be nice if you could ask them to provide a sequencefile with some data Im going to write a parser then +I was also wondering where this error was coming from +Either something is broken or we need to update the Integration section on the project site +So if we think it will considerably simplify implementation I dont see a problem with adding this restrictionYes I agree that this restriction would greatly simplify this issue +Comments in havent looked at the patch in detail +I have tried apply both patches though some tests are failing now +Closing issue fixed in released version. +Koen who will this issue affect +Is there an estimate on which version will have this bug fixed + +Yes it is +Another problemIf you press Create New Quilifier and then accidentally click on the Quickfix dialog again you may lose the new quilifier wizard if the quickfix dialog is bigger then the wizard window +Already fixed. +done +How about +Can you find it +That last point is interesting +Fixed +Ill try to look into this +it looks like the same what is the change +r in r in trunk +lgtm +To me this definitely is an edge case +It will show a message for unsupported loggers +at incubator dot apache dot org created +Marking closed. +Ive refrained from looking into this issue so far because it was making my head swim but now that Janne has proposed a fix I have GOT to see the patch +How are you running these tests? What were the failures? Could you attach the unit test reports for these two classesIn the mean time Im specting the issue reported by the customer when testing this patch +I like all the tests +I will create another bug that addresses cache cleaning +Thanks +Any takers? now I see a problem here +GreatSo we fix it as LATER and please reopen it if your new version works. +BrettHints to the root cause would go a long way towards debugging the issues +Thank youByeSandro +I just wanted to make sure were not mixing into this +fix in review it on latest build from branch and works fine. +I know the new Lucene In Action book uses TIKA does anyone have some sample code that I could look atI presume I should bring this up in the lucene mailing rather than adding to the JIRA +It still occurred as of two days ago so it looks like it is still an seem to have had a run of bad regression tests and RC is known to have problems +Having a java client that reads from an old server and writes to a new server does not seem like a crazy thing to supportCorrect +Fernando taking whatever SRPMs are made from and create a out of it is not an option +You appear to have a transparency that is turned on in Android +This issue is closed why? I dont see the solution and Ive got the same problem +Submitted a tested patch with a workaround Ian please commit +Thanks useful to knowIt would be interesting to know the times for the second invocation as well +Jervis I did see a QS named bpmorchestration in kit built by myself using trunk +I think the suggestion to use the initialCapacity parameter set to or can still be useful +If Im worried about someone finding it Ill just chown it to fred and let him get in trouble +Closing tickets +I think the should have a way of obtaining the reason the service was not obtainable +Note that a simple workaround for this problem is to set your Mac OS X terminal character set encoding to match your systems local system script encoding +When we do commit this patch lets leave the JIRA open to remind us to the after mustang fixes its bug +So far with a strategy I havent experience one single issue in production +Given we know how many filesblockswhatevers we have in the namespace before we even start the operation maybe we can just make the dot frequency relative to that? eg print a dot for every or so you just get a line or two but still have some indication of progress +Slip this one the code is frozen +What the code is doing is initializing a session in the thread but then starting the query from a thread insideCurrently there is a as a private member ofI think the complication of my design is that the class should extend and implement RunnableHopefully this can be done without modified the +We will add some kind of version information to compiled scripts +Working on triplesec I found that the maven dependency between core and the standard jar makes it very hard to swap bootstrap partitions +Charlie is taking this one +pathes to source folders have been corrected +As a simple fix the call to the remove method can be surrounded in a block +This change was not in the patch for or but is present in +Lot of discussion on stuff in +The policy that causes the is the SOAP request that causes the Cannot find Reference in Manifest Parts are checked to see whether there are empty in of the above mentioned scenarios +Alex if this is true please resolve when it is complete +This issue is duplicated and resolved by +This should be refactored as well +Closing resolved issues for already released versions +Lets focus on getting this into though any further comments on latest patch anyone +I just checked it out and tried it open the solution file with Visual Studio Premium version Update and I am getting the following error +Add routing slip test case to +Committed to branch as well +Thanks for the new patch! Ive done some work on splitting out the code already +Otherwise I think this is the correct code for the endpoint IMHO +it should ALWAYS be Id use an OR and Id invert the order to check first for queueroot +From there are already calls to get the bootstrap properties and the cluster key +Rejected bugs should not have a Fix Version +the tricky thing is that we dont want a hard dependency on JPA +Though I thought we were discouraging subclassing these days in favor ofYes +which does sprintln there a writeup that describes how this is implemented kind of hard to review the patch otherwise +the lost heartbeat response to the TT +Sorry for confusing you +Since this issue causes threading issues I think it should be treated as a bug +Attaching the patch +I am not sure I like the fact that getParameter handles NULL values differently from its typed siblings +Also whats the unit of measure for these results? Seconds milliseconds +Whats wrong with generating it every time we publish like we used to +I believe we can just ask a mentor to set it up for us +Ah that sounds reasonable +I have started to fix th eissue but it will take longer than expected as it impact a hell of code all the classes using the Attributes have to be replaced to methods and much moreHalf of it is already done but there are still errors to be fixed all over the code. +ssl is too generic IMO +I think this has nothing to do with but throw it back if proven otherwise. +I think its up to your judgment and Teds as to what that entails but if the patch attached now isnt quite ready well think of this as still something with a +On the other hand we dont want to make the changes too small or this will drag on foreverGo ahead and add the accept method to the models +AFAIR this patch was backward compatible +Achieve this thru entityI personally like all the ways of achieving minimum order quantity all has some pros and cons + +WSDL in a more convenient form +It is just replacing the list member with an unmodifiable reference +Should be fairly easy to change this back +Why not implement an actual +Attaching a patch that Pass the conf when compressor is created One test case to make sure no compression is done when NOCOMPRESSION is being set +I will commit this soon +note patch needs to be applied while in the +Would you be interested in giving a try at fixing this? A patch would be very welcome +Sorry missed your replySome committers are quite inactives but we havent talked about trimming this list down so Im quite pleased with keeping the list as it is +I wonder what the hold up is on fixing this one seeing as youve had a solution since the bug was submitted +For Netbeans this means that you first have to compile the module containing Node and and then your module +It reads better now but is still not perfect this is a known issue with tables in to you for any other issues +If omitted the esb name is used +Before this gets committed i imagine its going to need a unittest or two +If stubbing is all you want we can certainly do that for the next releases +Based on Knuts comment +Sweet! Thanks for looking so deeply into that +is currently a large patch and I dont want to add more complexity to it +Id rather not +That should solve both needs +Itd be great if someone could review! Thanks +Thanks to Chirs for the review +Im still not able to reproduce this +Issue was fixed for is course thank you +I ate burgers and dogs at the post match drink up coffee on the way little English those of us who speak English know +HehePlease change the line tobinsh eThis should be the case for every single shell script in our codebase +Candidate for though the fixes are likely to be in JON. +I think this would be a great addition +Also I think we maybe want that ouput in another colour or something like that +Corrected the xml schema syntax and removed the Unexpected character Unexpected character encountered from the xsd file +That is not legal +the patch looks good +Obviously it wont do that for those lacking the url +Also consolidated type fetching within the container deserialize functions +or Dhruba I think this is ready to be committed as it passes QA and Zheng reviewed itthanks pete +Please reopen if there are any new patches. +Its correctusernetdevvarmavenrepo mkdir p ukconetdevntfntf does read but if you are root you probably didnt have it or you had the ids misconfigured +I have different implementation of file browser +fixed and committed +Since we wont be officially supporting this in the platform I recommend pushing this out to +NM starts connects to RM does its work and then say mins later the RM diesThanks for the quick turnaround on the patch comments +Look at the tab for the mod I had to make to work around this for the release +I think this means that there is a bug in npandayver which causes the generated library to be a byte file. +and all later versions support fsutil while earlier versions do not +Here again unit tests are key to understanding +has been eliminated from branch +still no progress +I have problems using it with Eclipse and I am rather annoyed of EclipseI could upgrade my Eclipse +Attach the patch which add a workaround to get the real ip rather than the loopback address +Messages that would be of use to spring users trying to figure out configuration problems should be at least INFO +The issue regarding the broken builds of the examples should be fixed already by a patch related to. +When you addDoc DW has but when you only addIndexes DW doesntmaybe we should adopt what trunk does checking all the FI if one of the stores vectors unless you is readonlybq +patch is goodBTW Im not sure whether the size of all blocks should be equals +This patch should do it +A test with lines resulting in MB size worked without problems +OK thanks +Excellent +Ok I dont mind a new create just not remove the independent create sure +Thanks Simon +Can we all agree? Seems like will need to be changed +Profiles affect your builds and change the behavior of what you are building +Can you look in the Simple History report to see whether any ingesting has been done? The null output connector will log all ingestion attemptsbq +Patch with the new task and a test +I have updated reviewboard with a comment +Patch to remove final from all public static final class in Pivot sourcesTell me if its Ok to apply +It looks like this change has been lost during the genericsbranch mergeI think we can keep it as is and highlight it in the release notes. +attached is the DNs thread one line change +I must say I forgot that when closing and backporting because in the meantime Pierre provided EN complementPierre please use and create new subtasks if necessary +Further enhancements in commit afcbddfcafab. +Patches for +Table partitioning requires a timestamp column that is set at insert time and not changed later +lgtm overall +Will keep you updated. +You need to have jythonLib directory in the classpath +I will go to dwr mailing list to ask for thatand currently I think there is not any new update codes against this jira for the warning message I prefer not to take a look into it immediately because I am working on the improvement Migrate code to base on the dojo which will the portlets +I believe this is dup with +If it actually gets resolved in XDT for go ahead and change back the fix target when you resolve it +bulk defer minortrivial issues to +bq +this record since is now released. +What do other folks think? Would it be confusing to change that now? If so then lets not do that +Sorry didnt mean to ignore the expand embedded entities within tables aspect of this issue +The balance plugin is useless without a lot more Yahoo infrastructure +Can anybody please help on this +The fact that regionOnline is called async from the moment it happened makes it harder to manage if a RS is goneIt seems we should only add the server when its actually started and remove it when its deadAlso we should consider clearing the queues that have references to something thats now stale +Open question How the can know about the ledger and password while reading the entries as part of replication +Uploading version of the specification for this feature fixed typos found by Krisitian thanks! added an item for doc additon to REVOKE statement +group +We will take care of this issue +Use package protocol since we dont expect to break this on any future release +i committed it at revision +Regarding the workaround I would recommend disabling the combiner only if other steps such as increasing the heap size or increasing the number of reducers do not help +Ive used object serialization in for simplicity +Sorry Pete I have no idea what I meant by ambiguous back then +Message id IDbut I dont think this could result in dangling references to files in the store +Hi Franck I cant currently provide any information on a date of availability +If good Ill make a new issue for to make splits faster +That said David documented originally it is possible for it to have comma separated values users have read into it and IMHO it just has to be maintained +And for what it is worth it helped improve similar handling for Jackson JSON parser a lot as well +Can you try it out yourself +Ted Im working on addressing your concerns raised above +is this related to license +all of the servers should converge on some standard semantics addresses this noThis patch is now unfortunately stale +Sample pattern +This was really more about adding the translator for the DDL importer +If something is listed it is most likely a field name and then only if you explicitly make it a value would it be a valueother people have opinionswhat about a field name ? does that need special escaping just because it is also a number? How would this handleflidfoo when foo is not a real field name? is foo a literal or a field name that does not exist +Are we that sure it wont cause problems +It probably should buffer deletes in a similar fashion sending many thousands of separate deletes over the wire is not that efficient +Fixed in revision +First project is the reference project +Committed to trunk +The correct patch has been attached +or with equality constaints on primary key does not use indexI dont think the optmizer will do as currently designed +Yes +Brilliant +So it will need to declare an implementation using the factory added thenplace a facade over it +Im still not satisfied by how we handle exceptions around that part of the code +Is this something thats just included as standard with an android app or would you have to deliberately add it? Where would I find it if I DO have it in my projectIf Im not using this code what else could the problem be with localStorage +I tested it in Firefox +The l port parameter causes the shell to listen on the specified port accept connection requests and process lines of groovy sent to this do what you want +Sounds good VikasFYI I reached to that page by entering manageAddress in the browserTested the screen then committed the code +Adding more fields to the actual class also has the that we need to make sure that serialization still works correctly so I would prefer to create an inner static View class which is returned when calling subListI will further play with your patch and give you more feedbackThanks +Youve missed at least fingerprint generation +Thank Matt for his great effort in pushing this jira forward +Do you want it back ported? +Hi Alejando I followed your suggestions on naming +We probably need to do some stuff in a servlet filter to unwrap the thrown out of faces and get the underlying exception +yes so if it hits a devstudio +I added some tests in and all passed + +Closing because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +OK +AFAIK no is next release +HiThis patch was tested against release +Thanks +looks good +I am thinking of merging the fix into tomorrow just to make sure that tests run fine on the tinderbox +Resolved by of a bulk update all resolutions changed to done please review history to original resolution type +OK thanksSeems to me that it would make more sense for the attachment number to be relative to the attachments for the current Bugzilla issue which is what the OP was trying to doMaybe Bugzilla could do this instead if the number was sufficiently small +Simple patch +So its actually a bit more complicated than at first sightWe could omit all this and just add the processor but then tracing interceptors error handlingwill not kick in. +Ive looked into this a lot more over the past few days and I think I can make this work at the cost of creating new classloaders for each execution +OK Im making them optional for now +Thanks again +Also do we perform a check right before the rpc handler process the request whether the client already abandoned the request +Temporary solution is to disable tests as part of build just describes manifestations of +Updated diff +Because the ack end piece is not present an ack is never completed remotely so no ack is generated and thus the data in that particular log gets retried every retransmit time +Proposed changes +fixed still does not work for me too if I am trying use placeholders in spring +I now see the proxy error but the repository closes on refresh. +Nope because I need the font elements to set our custom font triplet names +Do it the other way round +How much more work needs to be done on the patch? We have a client who is very interested in getting updateable fields working for document tagging and could probably be persuaded to support working on this if its not too far from viability +The bulk of change is required initially by Lucene with a simple fix for Solr to follow +Thanks Glen for the patchYeah patch rename does not go so well +Youre probably right that skipTo is not likely any faster +Please note that the attached files contains the patch for this JIRA + +I remember mentioning that when I refactored the messaging subsystem transformersThe new attributes that I added were reaching the legacy version with the default value set +Commited a working version of the new packaging Not block other developers in case of commiting a new packaging directory and using exopackage using exopackage and having +I suggest that you repackage your application as a WAR I tried that myself and your application deployed fine +push out to +Google! +I have just submitted the test +Fixed by Dave and Avi in r and r +user questions user question not a bug +Moved to +the user continues to see the previous state until its safe to see the final termination state +Already fixed in current snapshot added to keep track of the will fail with next release of proper mechanism using file will be appropriate way to accomplish so that we can track the removal of this hack for the next cobertua release + +Attached is an example that reproduces the issue +Thanks Dhruba +Okay this definitely happens on Windows but not Mac OS X +Ignore the patch for a few minutes +The project has no control over +The first time I saw it I couldnt figure what it was all about +Thanks +wont apply to trunk there is no cluster there but it should also apply to patch as hudson wont take it plan to apply to then to upon release of +finderName find Contactadd field string class +Patch v is rebased on trunk +hope it will work +Also added detection of symbolic links and behavior which will either traverse the symlink while deleting or else simply delete the link depending on the value of the follow flag +is a test output +What dont we do for our windows users +File was attached +It also moves to make and deletes +as of Beta but more engineering efforts to come between Beta and for JBDS +I would say this is a duplicate of +Going to commit soon +VikashYour file is in revision + to trunk +I did the following and it worked for me +In the case of the metadata table want all writes to have max durability do not want to rely on all code writing to metadata table to set this correctlyWhats a use case for specifiy this at the batch writer level instead of the table level +Hudson is idle +I think right now the preview page makes direct calls to the config to get the bounds +How about as a step toward this ideal we bundle the master into the regionserver such that any regionserver can become master on current master fail? Then wed have one server type only rather than two as we currently have +looks like nobody needs it +Just added the files patch +Need to check with the other OSes to see if they can follow suit +Tiago offered to try to reproduce as well +Perhaps that software caused the problem +Thank you! +Please check that the issue is solved for you. +Patch extracted from the one in +Rohini +Similar to all other plugins the life cycle hook should be pluggable in a sense that you can deploy them +Thank Cheolsoo! Its working now +This includes adding additional bytes if you overflow your array size +Just curious +Of course we should do this +However the size should be on the same order of magnitude +After restart I can recreate the table successfully via cqlsh +So if it is just a performance problem why is line affected and line isnt? Also I dont experience performance problems scrolling through but Eclipse does use MB when its open +Attribute names get Key suffix in generated code. +This JIRA will add some context about this I meant +Hoping someone can try my patch and run junit tests to see if everything runs fine on their machine +Regression tests passed on andCommitted to with revision Committed to with revision . +Added one more test for validating this +so this isnt fixable anymore. +The other two are as documented in my previous comment in this issue the ESB one has been fixed +invokestage and invokestage dont follow the generally accepted naming conventions for Java methods +Committing. +We can reopen if there is a new and concrete patch available which addresses the issue and previous comments from committers. +Sorry for the wild goose chase +Let me do it perhaps tomorrowIf anyone has objections please chime in +Attaching project with failing test case +One the beauties of Mylar is working completely offline which is not possible with external reference to the actual issues +If this remains a problem feel free to reopen +Is this ready to be run through the tests and possibly committed? It looks ready to go but its never been marked patch available +the code has zero NPE checks for servers that does not have runtime downloads and it seems this has zero connection to the download runtime list we provide via runtim detection +I observed the same bug with Check valid sql field in the Data Source Validation tab +Very likely more since we have our own clone of openssh that may or may not be working right here +I have the same problem when use javawsdl toolJust create any web service that has method with long param and you will see that after javawsdl execution wsdl file will have nilable long params +for the patch +fixed code s a matter of taste but since theres already a separate method for Im inclined to agree with you. +Problem is I believe that the before closures are invoked after the validation takes place +If you find an issue which seems to be related to this one please create a new jira and link this one with new jira. +Hi Jacques it seems cart view slipped my mind +Its an incomplete patch to this bug +A new test has NOT been written if anyone would like to contribute one. +Can we please remove this part of the change +crc visible like this and were not doing anything special in the rest of the code to delete them +was reported as as well +fails for fmerendatesting for jt +If the patch can help provide an implementation of evaluate that would be great +Most if not all fetched pages are well below the truncating threshold +not for bug in eclipse egit +This enhancement will be considered for a release after Struts +I propose to remove visualOptions attribute as it is not implemented and it does not serve any purpose anymoreand it has its replacement +We can still have the protocol determine which encoding is the most efficient for the array by looking at all the values though so well still probably get an efficiency boost +OK I added it to wish list just because we might not have enough time to include this feature in release +Closing no response from contributor assuming OK +Christina thanks for filing the paper work I just called out a vote on dev hope to get this done soon! I see a couple of files in the resource folders that dont have a license header we need to make sure that all files do have an ASL license header before we can finish the IP clearance process +The trunk patch looks good +Just remember to svn update before you add the javadoc +Do we need to keep this issue open +should take care of this +you for your contribution! I have committed this to trunk +Looks good to me +the SCM information and Apache specific information in that POM by using import scope to import all the managed dependencies from this POM into your own project the benefit is that this allows the information to be mixed into any existing POM but this approach does not allow you to access the properties that have been Gert would it be possible to also move the external dependencies and properties to the parent file as well? +Ok this is confusing +A separate JIRA will be raised to ensure crash recovery is tested. +Replacement with the bug fixed +Kiirellisiss tapauksissa ota yhteys tuotetukeemmeIm on vacation from Jul th to Aug th +Oh I forgot the reset lost password use case +The aborted merge event is now generated and tested for +I think this should want until after True +Ioriginally thought release was supposed to be called oneach . +Even if speculative execution is ON there is no conflictoverriding because the competing speculative tasks are using different temporary directories and only one will be promotedThe assumption for this to works is that named outputs are not paths +I dont think it has a way to do so without either using a static class something we try to avoid or propagating those settings down everywhere which does not make sense eitherSo Robert I think youre right bufferSize should not exist on +This has been proposed in the past and rejected +Derby runs in embedded mode so that wont work +Revised Patch that uses Context to enable adding anThere might be a better place and name for the Symbol constant used as key but it didnt really know where to put it +i find that weirdIMHO an alternative mk implementation should only require a dependency on the APItherefore for for +When you select the editor the toolbars for that editor are still presentLong term I would like to make use of the a Palette to handle tools which can be either the palette view or made directly part of the editor. +So the full scenario is that Y remains up Z shuts down +I am currently doing a camera refactor +How do we proceedI havent looked at the code but I guess the first thing we need to consider is legal +Fixed in revision . +The get service should do the locking inside the simple method and only if the strict sequencing is used +A separate section may be enoughI am glad that several week of misery seem to have resulted insomething useful +Very interested in you views on the plugin commands +The assert happened because the NPE prevented the istat tx from committing and the cleanup code expected the tx to have no uncommitted operations +er +ok the parent module seems to try to be backward compatible to jdk +Im assuming this one is yours Uwe +That sounds great +This may still be related +Bulk close for + for import is very nice +Any interest +Would be happy about some feedback +Thats how Eclipse PDE product build works +created target platform for geronimo server when geronimo runtime is created +However unwrapped exceptions have only the first line from the stack trace without the exception name +Attaching the patch with fixTested the patch with thread marking the task FAILED and TT reporting COMMITPENDING FAILEDUNCLEAN KILLEDUNCLEAN FAILED KILLED and SUCCEEDED states for the task +the following appear to be the list of jars required for our Spring integrationr tfennelly staff Jul r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Jul r tfennelly staff Jul r tfennelly staff May r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Apr r tfennelly staff Jun r tfennelly staff Jun r tfennelly staff Jun +So this could be set to a low priority or closed although I will attach the patch here regardless +just a caveat to my last comment +has been marked as a duplicate of this bug +The feature is enabled by default +does detect and does avoid infinite redirection loops in its presentimplementation +The correct fix for the nio problem is for the to include properties that indicate that module requires those classes +sorry some unused imports in previous patch +It should have a small syncd at the end where it inserts the new into the segments and computes the new base for remapping the deletes +The second version of built without trouble in my original build environment and has also past the and the without hitting any failuresThe test run is still going but its looking good so far +the changes to the files based on the comments Ive recently MarkIve cleaned up the code added javadoc and organized the tests +Ive started seeing some peculiarities in the sitemesh decorator today +Hello JustinI agree + +like double would belet me know if you want that for +This issue has been created as a result of the analysis in +Have you considered creating a with this? I was thinking in something like but instead of using an array of ints for storing the docIds using an compressed bit streamThe Vigna paper is all about posting formatsBecause of this I first implemented an encoder and a decoder in a long format and then used these here for a that works on intFor a postings format the encoder would need an additional constructor from index data +Will do it now +will review patch for since it has a test and bad its just the test no fix yet +So you wont have to send a separate pull request for this +But the description we obviously do control +If someone is trying to do that they can invest the time to implement it +I think this is not a problem anymore +Closing again. +Closed upon release of +Added to the list of affected versions and estimated effort. +Patch applied at r +svn ciSending srcmainjavaorgapachecommonsvalidatorSending srcmainresourcesorgapachecommonsvalidatorresourcesTransmitting file dataCommitted revision . +If CGLIB cannot override those final methods it cannot delegate its execution to the correct instance thats why the methods are called directly on the proxy which is possibly shared by different agree that it would be much nicer to make all those methods final but I dont see much of a choice here if you want to support would appreciate it if this issue could be solved in a soon to be released because I really would like to use the features +Changes code so that it submits the full classpath of jars to +Thanks Mridul Done at rAshish +Its an arbitrary number +this patchReplaces the with and Keeps the registry of configuration objects in Configuration +There is no need for it anymoreResources come from forrest core or will be resolved via the +but not much we can do +Thanks He! +The fix should go to +We hope to make this available shortly +It should work fine if you on a project prior to selecting the wizard but with the patch it now doesnt stop you dont go that route +EverAs for the spring issue try concentrate on one thing at a time? See comments on the other issue +Set both Affects Versions and Fix Versions to +Carrollweijinit is not only affect the logging but also make remap incorrects patch can fix have attached a patch for testing +Thats going to take some work I wont be done for a couple of days most likely +Sami does my comment above make any sense to you +it wouldnt affect productionIve changed the reset to wait a completion of the executors in case of slow consumer +Can you confirm whether the tags will allow a wont fix or accepted debt type use case that will filter those issues out of the overall counts on the main project dashboards +We could put also the version for this agent to unknown +good call mark I will give that a shot +Thanks for that perspective Rick +Its phurty +Ive created patches for the branches +Other platforms report back milliseconds +Fixed please verify +Hi guys I tried to start a cluster on today but find a lot of info with uname a printed to console but I wish to get a more clean one as before +Any ideasBestIt looks like does not have the fix yet +But I respect your decisionThanks again. +This javadoc patch is almost identical tothe javadoc added for and +stateless forms auto redirect nowAlso multiply other fixes for the different url coding stategies. + test failures seem to be related to +But if you try coberturaclean and coberturainstrument they will be successfull with a message saying there is no source folder +Sorry to ask for more edits but the quickstart is complex and I think that the changes will make it easier to attached files +Is it possible for you to provide a sample that reproduces this? Im working backward from the point of the exception to try to find the bug but a reproduction would help make it go more quickly +the groupid is doesnt exist or has any relationship with retroweaver that i knowI suggested to create an issue with three bundles instead of three issues with one bundle each +I moved back to using because we cannot release pointing to the SNAPSHOT version +connection session pool max +To do this we have to fix or fork the code base as it is defined in the class +defer all issues to +Please verify the fix in the latest for future reference it would be nice to have a diff in unified format it will make patching easier. +feel free to close this issue +As you can see its really just the example client from the OA website including the reference js +This patch fixes the problemAfter that fails in +I just committed this +this has been fixed for a little while now +Seem someone else moved it to +I have asked infra about this and Ill report back with what they say +from that bug I did learn that it is important to run CI over different versions of jruby combined with a set of different versions of javaI am sure to provide a consistent preloaded API across is very helpful for lazy programmers like me if works I would never get the idea to maybe require something more since it might be not there in some other context +The majority of Spring users is running on Java after allIn any case we will choose the safer reflective part wherever we can in the Spring m using a Mac with the Sun JVM +frontend has been updated to pull in core Abdera modules +Thanks BikasDo you mind sharing one example that was not working for better Jira history +I assigned this to myself since Jonathan is seeing the result of the component model interchange format output + +Just to give you some information about the performances youll get on a decent LDAP serveryoull be able to proceed roughly authentications per second on a LDAP server running on a Laptop way more on a serverIts very unlikely that the thousands users per minute youll get on your will ever notice the performance gain youll get with a password cacheThey most certainly be pissed off if their password is exposed because it resides in memory on a machine exposed to the external worldReally caching passwords is a bad idea +There were ways to remove extension properties from a model object +Please attach a quickstart applicationThe browser is not important +Forgot to mention that all tests passed +wrong patch first time +Is anyone picking this up? It sounds like is pretty imminent +Go ahead and attach the rest of the patch +smileWe will submit patch for so that we can verify that hudson builds are working fine +The test case passes fine +Compromised by using a slower but simpler implementation +bq +Seems like we would need the ability to create a scanner +Thanks Ilya +Well Ill be thankful if you could give me a link to your code to patch with so I could recompile long could it take to build jboss ejb profile on athlon xThanks +Start server with IBM JDK +Fixed on trunk. +I think we agreed on your solution. +I hope those people know they arent actually secure anymore +by Ruth. +Ability to define variables and set existing variables is working to define closures is not available +Fixed by adding support for synthetic ears with no ear file just external modules in the vendor plan + +My humble bet is Eric Evans has hardly ever seen a repository as capable as our +Fixed for upcoming release. +Please apply this patch in Airavata directoryThanks profiles are in place +Patch attached +Adds support for IF on strings and related tests +The patch covers both and +added a test case to that demonstrates this +Log message Allow multiple reconfigure macros to be applied to the same file +Thanks Nuwan for the contribution. +test case relative to srcbugs. +Please use to grep changes quickly. +After that I will create a patch for this +The modeller bundle necessitates which in turn requiresRemoving these bundles from the runtime and replacing them with some other implementation of that provides a service capable of delivering service capabilities and requirements from a source other than Aries Blueprint should work with the current subsystems implementation +Change committed. +I leave it to the committer to decide if that should be fixed +Shawn i hadnt noticed bad performance for distributed MLT before until just now +Im having no issues running the package targets and building recipes +This is a patch to Spring that allows all the in a bean factory to share the same instance and hence same +This should be documented as a known issue in the release notes +Thanks Tim +Feel free to open a report for the lack of SLD operations in the capabilities document +Thanks for the contribution +So were probably in the wrong being this restrictiveIll look into whether its easy or hard to fix +Ok an endless loop of feedback Here is the opportunity for someone to close a old issue! +repository is now working but not fully tested I dont want to start the rsync process until I am truly happy with the setup +I was able to reproduce this outside the automated test suite with a WAR +TedIf you are ok with v i can commit it tomorrow +Only issues are with huge records +Forgot that I deleted all exclamation marks and replaced them with points +TP approach sounds better to me easier to maintain but I assume much more diskspace or is it the same since we aggregated from the composites anyway at release time +Dalis implementation libraries are just normal user libraries not something we should look into if we can provide something meaningfull later on +Updated the test case +Resolved in Teiid Designer This build of Teiid Designer will be included in m still able to reproduce this +Assigning this to we reopen this with more clarity about what there is to be done and if this is superseded by future issues? Otherwise I dont see reason to suspect action on this. +Cancelled the patch earlier since I was seing some task failures after applying this patch +Whoops Ill take a look +We actually dont need it of care about its eventual presenceSo what type of service are you invoking? Do you know it uses a specific web service layer? If theres a real interoperability issue because of this missing element then we can add it but if its just one partly broken implementation then adding it could break othersThanks + +Tests pass on my machine with it +Thanks for applying the patch Holger +I dont know if this will cause a conflict with these autocompleter changes or not but I fixed two bugs in the current autocompleter stuff in the trunk in SVN rev +pico invoker plexus invoker spring invoker are changed to use this new +Closing all resolved tickets from or older +batch transition to closed if someone wishes to restore one of these to resolved in order to perform a verification step then feel free to do so +We install Apache Xerces C parser in DWorktempxercesOur project is in DWorktempapplication +Pull request merged thanks. +bq +Raise priority for release +The issues you highlighted above should all be fixed too +needs to forward the properties to the task +yes that is the intent +What do you guys think about moving these utility jobs out of the mapreduce namespace and into a tools namespace? By these utilities I mean all of Export Import +Testcasepackage public class resolved issues. +That will be a release for Nutch but will probably contain a or release for Hadoop +New patch addressing the review comments +Add exception handlers inside the while loop of the map output copier threads and the map output lease thread +I normally end up using google would like to see a lot more options on the search funtionality including +On the define I think intent is clearer than doing SOMETHING +In this test generate random variable expression and then compare the mvel excute result with Rhino +Delay could help but youre assuming that the selected VM template has hotplug support and that youre waiting long enough thats a shot in the dark though Im guessing if the delay is minutes that would be sufficientWhat we really need is a way to interrogate for hotplug support before trying to do so +I fixed this last week +This patch fixes the problem by invoking target before +So all issues not directly assigned to an AS release are being closed +YARN also shows the app as KILLED +Checking the spec isnt listed as a valid connection close code +Thats why the spring plugin didnt pickup the dependent context for looking at this closed the ticket prematurelyStill facing the problem +as in build clangtrunk in a vagrant Linux box +Hi AmosInitially you can use a standard socket in the java implementation I would say +Committed to branch +Marking closed. +Thanks Rahul! +I dunno if theres anything more to do than adding the appropriate facet to the runtime component +also is really promising as well and avoids this +I am hoping to get some pointers on how to recognize which resultsets should be closed and which should be left often +Will look more closely at it in the morning but would like to hear from too before I go making changes might hold off on actually making any changes until then +Jeremy was right I had forgotten to check in the update to etc +In addition can you please start using the since annotation for Javadocs to make it clear when a class has been added +keep this issue open until the path JP LeibundguthCould you please append the patch with apache license grantedAlso generally we use svn diff +This patch includes Uwes last patch changes related to andThe fails with Lucene trunk which Ill look into +Please open an new issue if this one creates a problem for you +Thanks for the feedback! I rebased to avoid backing out the earlier change and then I committed to resolve the issue. +Its certainly not a perfect situation but this is not a bug according to the spec +Not sure if this already contains the proper setup fix version to in preparation for release. +good to knowThanks for sharing the information +We will submit the code for it very soonWhat we are planning to implement in Handling of Order Notification message coming from Google +in but then I suppose that would preclude standalone operationThanks for pointer to the doc +Committed in trunk at revision +are not marked with anymore +svn dirs had the same s applied and so svn diff did not see a changeMy bad +So it is still a valid successful eventT is removed from success and its T task entry is marked obsoleteNow T status is checked and if successful T is added to successThis means while retry T was running T was considered successful because it was in success +It may not be okay to active threads without closingpoisoning them +I grepped through the configs and found it set in +Reclassified as a documentation issue +So can this issue be closed +Can you attach a full thread dump? Are you sure your code is closing all connections +For CXF and were going to keep it useJmsfalse for compatibility +It shows when not all scores are identical when you use multiple terms +I would normally report this to Chromium but I have no clue how to reproduce it without +Thanks +Running Pdev in trunk console the is okI have the same problem with my overlaySorry to be a PITA but Im seeing the above when launching Syncope in embedded modeColm +Both runs now compile after this fix +Ramabq +Boost is now used if builtin tr headers are not found +This must be changed on prod git because ER was already rebased +I think there is something wrong with the test +No Im just trying to make sense of whats doing what +Closed after release. +Im sorry I didnt double check the change by running the test. +I now routinely debug the process +That can take time and the server would somehow have to what its fetching as the client consumes little by little at the transport level +Could we create a common target for them +Added comments in the property file about the places in code where the wrong property was used +Still looking into this assigned to releaseNote that in the example provided order is a reserved keyword versus a which is listed explicitly in the EPL doc + +Im working on the fix looks simple enough +the decoupled validators allow all this validation logic to be reused across these tiers which for larger apps is very important +committed to branch and trunk +The workaround foran application broken by the new enforcement in such a case would beto a create a database user called APP to match the system schemauser name effectively creating the owner and b always close downthe database connecting as APPOne possible way to balance the concerns is to enforce only forreadOnly access users thus effectively allowing fullAccess usersdatabase owner powers for and +but its practically impossible if you setup the listener on the same host as cassandraJMX serverin the rare event that data is lost we do use this in conjunction with complete snapshot backup so we could just cut a new clean snapshot and add the incrementals from there on +I think that it may be best to extend the configuration option of the to have a switch and in the second case simply delegate to the +Defer to all open issues to is no longer reproducible +This might causeThis will causeWe know there will be build problems if that specific artifact is missing +This bug has been marked as a duplicate of +resolving as per latest commit +backported to in revision . +Reopening to make sure we get to the bottom of the perf loss +Ok +Right now I still have to provide a minimal working set of source files laying out the proposed changesIf the idea is accepted there will be a lot of modifications and your help will come in handy thank you for the offer +Ive tried to reproduce a crash in Windows XP with VS and Windows bit with VS as well as on Linux and I cant see to make it happen +Resolved. +If this proposal moves forward I would hope we store the keywords in a metadata field labeled since thats the appropriate field in Dublin Core for keywords as used internationally +uses in this case while the iterator uses Invoker which doesnt handle failover +bulk defer to defer to +I have already mailed this issue to the user mailing list and Jukka Zitting does know the problem +This limitation might be legacy from the time when all endpoints were accessed through the wsee context +test ran into this problem as well +I have applied it in Rev +PrashantThanks for working on this +I think this is not a bug it is a consequence of the interaction possibilities from Groovy and in V of the plugin. +The is forced +Fixed +In particular setting the property to the dir where the file is located and point to another dir in the JDBC connect URL shows the effect that the log file is created in the dir referenced by even if the property is set in +The attached patch treats all filenameIRI with a scheme of length as a windows absolute filenameNote even relative filenames with a specified drive letter will be treated as if they were absolute +No Im fine to merge it now just wanted someone else to sign off +You definitely ran into a problem but given the information thus far I cant seem to reproduce it +Youve convinced me that adding the different methods with explicit types is indeed useful and safer +so delete the hsqldb and then run it again to verify its really failing +The content type is not sent by Safary there isnt much that we can do. +Trunk version attached +I added my PR which fixes all the stray threads I could find +Adding as this is in master as well +The recompile alone would be helpful as these packages are currently compiled to be Java compatible +Oops last patch was from the trunk +Please look at the attached file. +I have also noticed some memory leak problems in clientOur symptom is that the memory footprint will increase as time +It looks right in code +Thanks Stein +May take up to hours for the patch to reach central repo. +So it might be a smoother way to allow this and log a warning instead +Just to make sure there is nothing done related this jira yet yes +If I compile and run it with Java it works fine in all platforms but if I use Java got these error in windows as in we could simply add a dummy element to the zip +Thanks for your contribution Gilles and thanks for your bug report Pavel! +Wont fix because its it sync with the ejb server coming from the build. +So the behavior has changed since I last tested +We can revisit the latter later +The first patch was not you actually try the patch or not? news hence wont fix +Thank you Van when execting all SQL modeshape seems to be working fine +Could any user check in your formatting config +If the code being compiled is the same there should be no need for different jar file namesversions +remove one location +patch doesnt apply anymore +Jon Gray can reliably reproduce +After the master rewrite gets merged in it will be possible to retain assignment information in META during a shutdown and to then use it on the startup for the initial region assignment +Would it be enough? I can produce a patch +Todd this is a good question +Thanks Hitesh +Since it is decided not to implement this Il close this +batch update of issues with fix instead of the correct +We need to address this +We can always change the implementation in a future version targeting +bq +Closed with released +I still really want this one to be done soon so Im punting to +Were you testing on Windows ? Can you be sure that your cursor was in the text editor of the Transformation Editor? reason is that UI focus is important to managing keyboard I have updated environment yes Im sure cursor is in the right place +what kind of errors did you get +Commit Rework to use the changed API and to be compatible with trunk Commit and have different goals +I took a quick look and your approach indeed does look far nicer and more general than my approach +Im closing this we can continue the exchange of views on the incompatible change on the dev list +patch looks good +Thanks SuranSanka +Okay so are you cool wme editing the code in the packages to explicitly turn on sorting where it is necessary +It isnt like issue is assigned to you anywaysIf you happen to have time to work on some issues please let us know. +Kinda convenient +Id also suggest wrapping our use of everywhere with something that manages +bulk change to is obviously build problem +attached is a patch written by Uwe +Fixed in r +These files are generated by the application and are used for setting the defaults to use for connection settings and for the nodeviewers used on first start up +But I must admit that they were not very deep +I think it was around early so Im fine with +I did had some problems with filtered resourced in eclipse but with release they got resolved automatically +Fixed +Logged In YES useridYou are doing a find in ejbStore right? Did itwork in +Added get API +Thanks Craig +Ive found an error in my test here is the fixed and improved versionThe test still shows difference between Harmony and RI behavior Harmony actually does not destroy groups containing threads +I agree this is something we need to clean up +fixed +Removed trace printout as the dbeug printout showed all that was in svn revision +The path for the issue +Thanks for taking care of my comments +Changing affected version to and fixed version to the originator of this issue please verify that this issue is fixed in the release by adding a comment to this issue so that we can close this issue +Attach two patch files +Now you can argue that if youre setting values from the result of a form submission then the object really isnt immutableBut in that case Id like the data binder to use a constructor instead of setters + ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun is zero prefix in the file name is the dir name from which the server was started ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun ldimaggi ldimaggi Jun r ldimaggi ldimaggi Jun r ldimaggi ldimaggi Jun is a new problem in CP CPGA did not create these files +Probably this was done as part of a housekeeping job to delete all the files which were produced by registrationsI thought we agreed that nobody owns any code so please if you do not like this code go ahead and change it? As I already mentioned we are extremely busy and currently we just contribute the parts which customers ask us to do and which do not need much extra timeRegardsHans +I was hesitating to store the exchange itself due to the runtime implications of doing thisAt runtimethis could require a copy of the exchange to be saved in the Map until the exchange was delivered and could lead to significant memory growth if the exchange payloads were quite large +In the same way couldnt it do the same thing for source jars by just adding +Patch with the CDI field producer +The method is assumed to make a best effort attempt to close resources hiding exceptions from the present it only catches opening up the possibility for Error to force early exit and impact callers +checked in code sets query cache off by default so closing this issue so itll be included in the release notes for the beta +Attaching could do this in base transport listeners handle method but my personal opinion is that this should be handled by the individual transports +If we do that we could also expose loading Packed Fields via and have one API for our users +remember the life of a dictionary is the lifetime of a DFS not merely a particular run or a DFS file +Pushing to to and targeting at could be bumped up to based upon using the existing invoke method with the action parameter as the port name to assist in situations where the WSDL specifies static soap header information +Ive committed this to and +Renamed target to following Owens suggestionThe java builder is only invoked to build the record io files for tests so everything is complied for Eclipse you can still use ant +by a sentence splitter analysis engine adding an analysis engine +Apparently this was a VM bug in it doesnt happen any longer in everything works as expected now +Most rollback to be safe but Oracle at least likes to commit the transaction +This diff includes those includes all of the previous diffs updates so it is the only one that needs to be applied +Given some of the discussion on recent merge thread I am asking for any reviewers interested in reviewing this patch to start reviewing this patch +The release of this feature will have to wait until a decision about Java compatibility for is taken +This should now throw a +Missing update to Select +Committed to. +Thanks Bryant Luk for commiting this code. +This issue is not related to Hudson +But if its unnecessary I could remove it +Ok Release out which includes Clauss fixes +I have tested the server and the API injecting entries within levels and everything works fine assuming the server is started with enough memoryIll profile the code special when profiling +On my cluster I go from about k threads to only +works now though. +That works! I didnt think you could log in as a user that hadnt been created yet but apparently you canIll start working on topics for this material now +Thanks Tomer! +This will take care of creating output files +The remaining question then would be why only do this for specific files? Do you forsee a significant speed hit to just logging all allocateBlocks calls to the edit log? Is the throughput of such calls significant enough that we need to bother treating important files differently +The issue is actually solved by. +nl redirects to +fixed on on +It works fine for me now. +For the time being I just made the streaming execute method completely bypass the caching layer +Could I get you to add some way of configuring that +I would recommend it be addressed in another JIRA if it is an issue +But this is for the first time Im hearing something like thisIf you agree that this not something to do with then I can openup a different jira and deal with hudson issue or we can file a bug with hudson dev +Thanks Wing Yew! +Solved separating headers from netty client props +Java Code for workaround +Also can you mention if its a master branch or stable branch cut from master +What do you mean by picked up? When I restarted Eclipse nothing changed +revision should fix test this and close the issue. +I still dont know exactly what the problem is this should only happen when changing the cipher key or if the data serialized is an older serialization format than what the runtime environment reflects +Applied patch with thanks to James +I have problems with reconnecting my backedup DATADIR there was permission problemsI was discussing with in this regard and when I changed to the ownership to tomcattomcat My DATADIR starts displaying but when i click on layer preview I got this error +Yes we do not need to initCredentials since we are not talking to any hbase serverSo would this completely bypass security? I also want this functionality for certain use cases we should just be clear on this caveat +We can just try to download it and if its not there we would get an exception +Is this meant to be a patch? I see that its been assigned to for a fixPamela Fox Can you write a ticket indicating what device you ran into this on? Ive run into this on very Android devices +The problem is that the debian side has a separate init script thats missing a call to +This patch makes the old overloads behave like the new overloads If the jdbc type is not a legal Derby data type then a is raised +Go go go +Vendor in German is most likely a Verkufer or Anbieter +Im thinking the MTOM options can move to the bottom so the endpoint address endpoint proxy group are closer together but other than that Im not sure +Here try this out +This was committed marking as fixed. +Steve just open it it works that way +Just because I had to touch up the code a littleconcerning our code conventions doesnt make your patch bad +So a bit more on this issue If other are present in the SLD then the line rendering is not correctHowever if all are present and I movethe in the same rule that has the to be below it then it works fineFor me this issue is resolved however not sure if this is a bug somewhere +It should be an explicit decision to use external JAAS as opposed to internal configuration +right sorry missed committed the patch +Ported to x +I think the current behavior is correct as far as the spec is concerned +No spec for data in management directory +Just forgot to update the issue +Fixed with a modified version of the patch +Sure +Fix contributed and committed by Knut so Im closing the issue. +are those changes backward compatible +When name nodes starts up it initially creates all the blocks in Indode while reading the image file +I still think that most user will use this class in a context +Since that is a breaking change the only scope this can be done is Groovy +Otherwise youll have to release the response yourself +You can apply if you want but I withdraw this patch +That way we could get more feedback +What do we want to do here guys push to? Id really like to see this one get closed out esp after all the work youve done +But I have two questions Would there be other situations where tying two begin events together would be useful? Seems like that might be a possibility +Why at all we need KB builder hereI dont think that to show path proposals is necessary build my opinion its can be done easy and without builder +Merged into in revision after releasing +Updated file name in Release Note text as per advice from Honza +Attached sample eclipse project containing an application that demonstrates this problem +we can do this if theres ever a real abundance of repetition or people get really annoyed about fat fingers +We can give you a portion of the infra tree to use as your live site if youprefer not to litter your source tree with website content +Sure I cant test under a mac but Ive just replaced by +I will continue this discussion shortlyBased on that discussion we can decide what to put in a release note +Chris will you have time to take a look at this today +You cant have a Java class start with a numericso your stylesheet name shouldnt begin with a numeric +LDAP viewer webapp for GTomcat build +Im not completely sure as I havent been very active with Webwork for the last year or soI was using the solution I posted to handle the issue in my own webapp thoughPeter +Yes I think it is ok to make the classes unaccessible from the outside of the projectExceptions are different they are returned to a client in the form of right now but if the client wants to distinguish between them then we need to keep them public. +And remove it when the last dataset using it is removed +I can do a release to include this fix now if you need it ASAP +The only thing I can think of is if you create a composite servicereference first then create the service component and want to wire the two up +For Japanese documentation I would imagine a parallel set of documents +That should read +This is one of the things that will get documented in +The uses the systems default charset encoding but the database tries to store the data using patch that will use the database encoding instead of the systems +Sorry for the typo in the first line + +I have no idea why this wasnt caught by the CLI +this is a first pass patch that replaces the functionality of specifying identites by file objects with specifying them with VFS urls +Ill try to find some time to look at this for +I cant tell from the release note exactly what the incompatible change is so its hard to comment on if it should be backported +In Lucene and get subclassed to support RAM indexesIn Lucy its the lower level that gets subclassed analogous toa file descriptor class +However since this may not be easy to do we can skip that part for now +I dont have further comments +See comments. +endTime startTime +release is now optional will only require the user to set the release if heshe clicked the release button +in which case i wouldnt qualify this issue as major bug but rather as an improvement +When I tried setting session timeout to a lower value I found a bug in Undertow you try to replicate my results on your end +just validated +I dont really know quite how to contribute such code to the project +the value gets persisted in +I just realized that ognl is shaded also +svn subdirectory would definitely improve the startup time of the applicationWould like to get some opinions about this +Applied the patch to trunk with r. +This was a more difficult one because of the partially parameterized generics in the condition evaluation +You should close this issue +tinymce created with zip r to include all files recursively +applied thanksI added a comment you might like to address later use a mime type of texthtml to trigger the changes investigate html parsers to more accurately pull out the content +Thanks Amareshwari! +Adds a force to force ajc compilation even if no sources have been modified and the ajc options havent changed +I havent seen this for a while now but I think its because were joining stdin and stderr +First steps in rev in genesis trunk +Also its not guaranteed that this will happen only during start up this might happen any time a pNode is created which is either at start up or when another node dies and a standby pick up the taskAs Flavio mentioned read after write is guaranteed and we should be fine with checking for pNode existence +Since most job is already done by Nutchfor example you want to check certain rule and based on this rule to add some information into the index so you can later search your index based on that tagThe way to go about it would be to develop a parse filter plugin +However based on your feedback Ive added back in and tried to filter out anything weird in decomposePredicateIm testing a patch which allows a single equals predicate or a single or double ended range scan against the rowkey +It looks like you are getting back a list of Integers +I see no harm making these methods protected +Container is either not yet running or has already completed or doesnt belong to this node at all +As queues are added they are registered as services which expose the statistics +See this invoice that isnt scaled when upgrading to and we need this version to fix other issues +This is a logging configuration issue best dealt with on the mailing list +The reason being that there is no RSS feed file +Patch includes unit tests for new IN expressions and translation to in +In JET mode DRL VM works OK +Works brilliantly well +Fixed in revision +nm guess thisll have to wait. +Here is my patched file +Switched this to ensure they get logged only if logging is enabled +This seems to be a classpath conflict +Pull across revision +It looks like Derby returned rows for the query up to version +Committed to and . +See for further details +I committed this +Thanks Nicholas for committing and Chris for the review! +Can we close this issue +Especially since the schemaBase is now not available +Or Is this a limitation to OGNL? Yes this is a new OGNL limitation +issue is resolved with solving. +Net and CXF +Ill try and monkey up a bare bones example of the problem +Cannot create and QE images before release +Closing issue as it was included in version which has been released +Yes I dont think we need this. +Brent do you know more about this +Uma thanks for the reviewI have committed this. +Committed revision . +Yes current bundle lock holders can finish but they wont impact the refresh since they will have to finish before the refresh can proceed +Patch reverted unit tests are fineTrying with any sieve script gives some issues however will prolly open another jira for this. +Thank you +I think we should call the class and put it in the package +New patch attached +Please discard my last comment. +r +release audit was because of the wrong format in which Idea describes deleted files +Removing the fix version for this issue +Ill say I think this is importantIt would allow dynamic queries +Please combine it then follow the comment dated In YES useridUnfortunately there are problems unpacking testcase archives you submitted +the point is about howwherewhen knows about the path to use when dealing with the static resourcesmy point is that instead of trying to guess where to find those resources based on the servlet path of the request we should just make it an explicit configuration the same way its explicit in that way people can use anything they want even if they choose to bind to multiple paths +Fixed by adding an empty space when theres no code source +So i did it test case attached +this is a dupe of +bulk close for +Thanks +This should apply cleanly Eclipse did not want to behaveTested this reverts back to what we had +Does the current behavior actually cause any problems? Or is the message just disconcerting +If OK with you Ill add the serve piece to the CLI doc +Fix a bug when join value is nullAlso fix the all the output conflicts in this patch clears the conflicts of test output file completely +This was fixed a few days ago are you on the latest revision +This is awesome Jarkko! ThanksBefore i resolve this does anyone know of any code that couldwouldshoulddoesdidmight reference the or classes? It seems unlikely to me so im tempted to just remove them rather than merely deprecate them +Couch would ignore this header while would use it to decide whether to send the extra infoJust a thought +Push to +Decided not to backport support + for local www as well +i got the same problems with Tomcat and Icefaces in every browser +Tag with the same attribute declaration are reused from the container and the result is this mixed will look if there can be any bug introduced in latest snapshots and try to test it in different containers +Only a new registration doesnt allow me to use the same Im not even sure if unique email is a criteria +I can try +Its clear the change was not intended so we wont mimic the broken behavior. +This is somewhat limiting. +Closing. +Subsequently adding another project wouldnt display an error message or show the in progress icon but it wouldnt be added +Checked it out it seems to work with all the test code +verified on Developer Studio +Now that work is well underway with AS all previous community releases are +The user will not have to know what the Javaclass name is if heshe is using JAXP as the Java class is wrappedinside the Templates object +The only test I left in the tools project itself was one that looked at package names and for that I bypassed the plugin +Aaron is committed to trunk when you get some time if you can review and commit the doc changes if it looks good +By relation i mean that one server in the top view corresponds to one tree in the lower I have two issues with that The top view is just wasting precious white pixels on nothing The two top nodes represents the same thing but does not have the same actions available in the right click is not intuitive +Added to services +for committing it +Patch needs to be reworked to not cause a build failure in. +It is automatically added to the correct place +app is the app actually submitted while appState is the state retrieved from the store +closed as WONT FIX +What Data integrity validation +The patch was submitted to show the starting point of the final patch since I couldt decide on whether cookies should be systematically escapedIn I couldnt find +This one would also detect invalid trailing elements even if a choice was satisfied. +I understand it wont work but perhaps throwing an exception again or simply logging a message might do the job +I was concerned about breaking DROP TABLEs idempotence so rather than throwing an exception I just print the error to the console +The transient I saw was on a static fieldnot me then I am not sure if i use fields in Writables in Mahout at all at this point +I cant usefully comment on the suitability of the release in the cheeseshop but would it even be to upload with the attached patch applied +Bulk closing stale resolved issues +You have described implementation +Attachment has been added with description New patch now avoiding static variables for min and max patch doesnt make use of the static variables in anymoreUnfortunately it still uses the static variable but Iwill be removing that soonId rather update frequently even if there are upcoming changesCheers you very much for the pointers AdreasI will do my best to follow them as well as the requests and I will post back here +Jerry unless Im missing something from your description this is a duplicate of +I see advantages to having an autonomous process doing the health checking +It is fine that we now run which calls a sql file but I think it would be less confusing if went away so folks would not try to run it is a standalone testCould you submit another patch for thisThanksKathey +Found the instructions +trunk Is enabled just like the tracer +I suspect a classloader issue as Ant works in a different context to usual +I copied this comment from a different email discussion involving somepossible interactions between and this bug There are a number of code paths in the server that use the baitswitch method +Thats the correct MD matching what Igenerated on my laptop +dict to +HiI regenerated my via git diff . +Next I need to look at the thread safety of under the new locking system +The test failure is not related to this patch +I close this one because no issue was found +Hi Joe I just realized the zip file includes many +but if I remember correctly the plugin descriptions are displayed only after the user explicitly clicks or selects a given plugin +Also I apologize if I jumped the gun a little early I became irrationally exuberant when I thought about moving to a generic configuration system +From the Facelets code it looks like it just logs an error and goes on +Right +Bumping to possible next maintenance release as nobody seems capable or interested to further investigate this one for nowWell need feedbackpatches from based developers as none of the current committers are +bulk close of all resolved issues. +Will watch out for a Rampart project implementation of access to validators in the future and come back to this issue +I am no optimizer expert so I suggest at least one more reviewer before we consider moving this patch from the trunk to codeline +Some notes on the implementationocj must be configured for directory must be pointed out by the property +The geotiff file can be found in the sample data attached to this to Simone for cooking it up + +Want to rerun your patch +This is a result of the changes in +deploying file before is finished +On restart there is nothing in the cache so it is necessary to read pages from disk to the cache +Patches applied thanks for the contribution. +Thanks Andrs +Sorry it seems I forgot a few references to in +You may need to clean poms from your local found this specificaly after cleaning all our repos +We have enough problems with flaky tests and Id rather not have to consider hadoop code changes shifting under us as another source of bugs +Verified that the code generated for the supplied test case now compiles correctly. +Youre talking about timers in your description and we use timers for failure detection +On AS calling httplocalhostctxrsc will hit the servlet +Thats what I meant +i didnt realize it was going to produce a +Thanks for the suggestion Ive refined that section a diagram hopefully making this distinction clear. +we have in a test that uses a primary node operation to add an id field +Javascript test case broken fixing for js client test in the js test cases folder to now use. +Widening Generics for +Lets discuss to double check +. +Would be even happier if no recursive lookup +extends Tokenizer but its constructor takes a rather than a Reader +So you need to configure your website in the entity the file or both +Committed patch at svn closing +I approveThanks! applied to the patch to the code base +I will push this patch soonDevin +added heres the deal I am going to commit unless this is being votedagainst +Thanks for the fix Dave +Have attached the patch with the necesary synchronizations +Are you satisfied that this is not an issue for scanners? If so Im okay with closing this +This issue is being resolved as part of and +Hopefully itll be helpful I havent contributed before +Would anyone want to try this +Ive made code adjustment for richtree treeNode and related classestree renders in the same way as and templates +Peter Ive reviewed and committed your patch +Yes we can improve the message on the we should do more to document the huge hole that was in Seam and strongly urge people not to try and do this kind of thing +Thanks for the reportThere has already been a similar report recorded as +Committed +Simplifying the test case quite a committed this +Also I changed transform loader so it has to be a +We have it yay! +Pascal Geronimi from Puzzle promised to create Usability design that would improve the Editor +should it even have a writer instance +All tests pass on my laptop if jenkins says its ok Ill close the ticket +latest build should now include the extra debug +Thanks for the report was there an associated PR +Changes on branch prj +Patch to support Node annotations +Currently test depends on which runs fi tests as well +SQL has no notion of read only so can you elaborateI am no expert on DB though from what I have heard its locking and isolation is really a thing of absolute horror so maybe thats coming into play here +comments have been cleaned up + +I have it partially working already +I ran with this one. +Each site might have different requirements +Therefore it is not a big risk to return null for this methodSummary of a discussion with Nathan should be fixed +I havent really thought through all the edge cases yet +So at FOSSG youll have enough to get drunkp +Looks like moved ahead +Committed to TRUNK +could have a set method +For some reason it did not occur to me that the defines the interfaces as well +The default tabs label of dashboard is localized at the first time it is created +Nag +I have discovered a similar problem with the Smart Chinese toolkit +I ran it with connections +You can checkout the latest version from Struts repo and install them locally and they will work for youRegards archetypes are going to be released +The issue is fixed from my for confirming you have the JIRA you mentioned above? +I went down this path initially to prevent a lot of chatter on RPC and edit log ops but that might be a premature optimization +It is a patch on top of your patch +Just another though +It seems that there was no error on ER when talking about this particular JIRA +Ok is thread safe now but is still not +I took a look at the projects causing the +Michael once you fixed the property method did the cache store get instantiated and does it work as expectedBased on my testing the way configuration was built for Spring meant that cache stores would be started at all not even constructedAnyway Im fixing it so that parsing is done using the new parser which will show such errors +There is a test for this in the Weld testsuite underlying issue has been fixed in the meantime in +Looking at the test profile the wrong dialect is used instead of ve merged the pull request +for the patch +Can you give me concrete information about this so that I can convey it to customer +No you dont disagree we agree this depends on how the resource provider is mounted I thought of the situation where a provider claims to own the sub tree while youre thinking about the shadowing caseSo the search should not return resources which are not accessible via got it +Patch applied in r +If needed that method could be promoted to a property utility in the admin moduleFor a similar approach could be taken for local connections but the code will need adapted as the logic also refines the to states and does not inhibit connections to Inactive + +Or perhaps just hunt down feature iterators that implement both and separate them think the former is a more robust solution +how about with mutiple app servers in a cluster? +not fixed +Continued list discussion is probably necessary to figure out our preferred approach +Im fixing this for M +and are backward incompatible with and we really dont want to break compatibility in this release +Do you have a for this issueColmThanks Michael +Please confirm whether or not this is specific to by running this on vanilla EAP +Run in it dir +In the SVN branch the n option works correctly interactively and fails to reset the chain in batch mode +This appears to be an endemic problem in much of the JS code +This may be related to or Eli +ThanksAddresses latest comments +It sounds like an MVEL issue we will isolate the issue and contact the MVEL developer about it +It seems like there are several issues here +The are not in libendorsed because of the commit Thomas which comments out the copying of them in Im reassiging this to Stan to assess the impact of the not being in libendorsed and why the conflict with a CXF client +This exception from was always there +It should cover most cases +patch has been submitted and committedthanx fix version +After solving the attached pdf looks different there seems to be an encoding issue quite perfect after solving the rendered image misses the X after Telefono and at the right of Solicitado porPossibly same problem as in +So should be still seeing this issue for plugins in version +Thanks a lot Jihoon! Ive just committed it. +trunkCommitted revision +Patch is against master +Heres a simple fix attached +My point here is that operations like garbage collection are IMHO not something that a JCR client should be doing especially when the implementation still doesnt have any access controls and it relies on the client doing the right thing in the caseNothing that we expose through a API should be able to break the internal consistency of the repositoryNote that the garbage collection operation is in principle a safe operation so Im not totally opposed to putting it in a API but until we either declare the data store configuration as unsupported or solve the obvious safety issue with this API Im quite reluctant to expose such functionality to all clients +Not resolved until the patch is applied +Please see the attached patch +Can we review these on Monday and either open up new as needed or decide to not press for fixes +The old perl script reads in a kind of configuration file and than construct the parameter that are used to call the MainI suggest instead as hadoop that pig has a conf folder and a default configuration file +Moved tests as well +I had some bad downloadsSo it seems OK I have visited all the default update centers an installed all available updates and everything is working for the help +Responding in the correct place I had responded on the mailing list overlooking that this was a ticketI think the only use for is to configure aggregators which are deprecated in favor of combiners +A well behaved app would do C tricks like naming them method and method +That was a on the patch for +Hi is there any progress planned +The make file did not exist yet to run clean from +I dont know +In fact when an artifact cant be provided when it wasnt downloaded because of bash hash with a fail policy maybe this would be better to return something else than a +Why we didnt think about this before +Thanks Travis! +Once we get that resolved Ill check this in +included a switch on the tweet form to include the users location or not +Committed to trunk +Applied to master +Committed to trunk! Thanks Mark for your contribution +Testing seems to indicate the latest Thread class patches resolve this issue +I presume this issue is for relations and the other one raised was for option A is preferred +To DirkPlease look at the picture the smooks can load the classes of required projects if you add the required projects like the picture show the class finder can load them +It incorporates changes that are a result of fixing issues reported by checkstyle +The current implementation uses Jettison +Sorry forget the issue I realize what was the error +Dont we want the src though? Could we have a project with the source and have some instructions on how to build it? The releases can then have the jar which people can install and run the command on +It was only incremented if sharedCount but always decremented which could result in strange behavior +Same problem here it wouldnt even honor an overridden central repository in +When benchmark this on a single hfile region the overall table scan throughput can be improved by +May somebody pls try this +It may be possible that the C client has a bug in the marshallingunmarshalling code on +We already do include ORO in our SVN tree +Gleb I fully understand your frustration but behind this javascript plugin and all Sonar plugins there are thousands of integration tests and if we merge all patches as soon as they are provided maintaining those will become a nightmare and our own internal productivity will fall +already has a set real issue here was needed low coupling with Serialization +Do you want to add a separate JIRA requestWould you consider contributing the new view to the project that would be great as it would benefit everybody and would be the first test for theThe workarounds would certainly work but would be specific to your application +Ive updated the patch which avoids searching the parent and also cleans up the new version which does even more code ve applied a slightly modified patch now which fixes a b and c by directly using and using the same algorithm to scan the tree as beforeIll leave this open for some days to gather potential feedback +Correction in previous comment The combined patch for hadoop is for jiras and. +Once I get Sakai safely up to Pluto and checked in I will jump to and refactor the Sakai code to use the new approach for. +We need to read superclasses properties as well +Thanks +I recommend you provide some sort of unit test here to further clarify what you believe to be wrong in the current implementation as the description and use case isnt entirely clear +New files +My hunch is that the processedunprocessed flag for the artifact in the db needs to be set to unprocessed when the project model was not added in order for the changes to be reflected +would change that to never ever skipping rowOnly if you actually did have a counter in the columnmetadata right +Xavier Max are there any steps to reproduce it +Moreover given youre using cxf I assume youre not using the cxf thats included in AS +We think this looks like an old version of code and this is already fixed +Bulk move we add this the behavior should probably be parameter driven +This issue was fixed in jBPM but it was not included in SOA because of performance degradation +in c directoryREADME file content is this to myself +While these comments Im quite confused +Moving out till a patch shows up +I had mentioned the same thing to Alex off line and he is going to try it out. +I am actually close to proposing a branch for that since its truely huge and it would be easier to keep up with trunk really +Ill try again today +The Nexus index has been removed from the Maven core plugin in and trunk in order to prevent these crashes +Hmm I see fix version is would it be possible to add this fix to the next stable release? Thanks +The fix looks good +The solution would be the same +I think it will be good if we can have a which works with both modes +But one other thing that happens that is a bit of a surprise is that TWO copies of the class are created +I just committed this +Did you try setting the timeout with options +Could you provide tests and documentation John? +John thanks for patch v +Thanks Andrew for feedback +This is part of some authorization check lookup with bypass set +Also we would need to handle Serialisation which is only implementable in a performant way if you know the underlying logging framework +Another option is to change all instances of to in the sample and add URI specification where necessary to avoid the default port to be used by Tuscany +Fixed in trunk with the update to Bnd however note that Im still testing this against various projects to check for any build breaks +will have to verify if this doesnt affect that update to this huge patch +Sorry about that +Attachment has been added with description Patch for the test case which shows the has been added with description The patched Glenn for applying the patch everything seems fine after reviewIm closing the bug +One comment theres an unused UNPACKDEFAULT constant that should be removed +The add method if is removed +I talked offline with Owen about this and he makes the case that a interface to the aforementioned types would probably be more than sufficient for practical uses more readable than the partitioner handling differentlayered length encodings and a more general abstraction than this isThe only remaining reason would be the aforementioned spacetime tradeoff saving an int per record while adding a call to the partitioner for each compare in the sort +Im afraid this wont help +If you installed EAP using the graphical installer and you want to access the JMX console you can use the username and password you provided at the time of installation +as svn on trunk +there is no design document about this feature anywhere freely explained above thanks for great work + +Just for my own curiosity are you keeping classified ears as attachments or making it so that the classified ear could be the only artifact +Files I modified to provide archetype functionality +See attachedFirst stab at a quickstart cheat sheet contributing guide +i tested my initial replication procedures and its working now after the latest code drop merge. +Woud be ok +Need to fix +Bruce why would we need this +Yepp verified it on windows +Submitting a new patch +Currently this is possible using the which can also be configured at the global installation level +should have imported first +Certainly if downtime is acceptable that would be the right response +Works for me in the booking example +Not very likely no +but i think the bits of the code you are talking about wouldnt change +This little change makes fail most of the time without the rest of the patch and never with this patchLars I dont quite understand your above comment +Patch that includes the fix and test cases +For Weblogic I was able to solve the issue by specifing another It may have side effects though for existing that rely on the default weblogic +proposed patch +Initial patch w test case fix +I still was unable to see it working on my configuration but since you are reporting it as working I am leaving it in an experimental state into the distribution +instead +Notice that youre using the bean taglib not the html taglib +This patch sets minOccurs to for choice members thus allowing the coder to make a choice +Test case +The insatllation directory can be provided with prefix flag for the runConfigure or my setting an environmental variable DESTDIR +OoopsThe current word is not expression but statement +This issue and its counterpart are about sandboxing the jBPM scripts and action handlers +Please provide more details in other case I will close the issueRegards flag is only used for development purposeDo you mean devMode or throwIf throw I could not disagree moreWhen a method we had been building on is no longer present through some incomplete rafactoringthis is a grave bugand we want the application to fail as cleanly as possiblePreferably at compile or deploy time +So errorsexceptions during shuffle phase of the task will mark the attempt FAILEDINTERNAL +The existing connections will continue to work until the next VDB with same name and version is deployed +Could just flip the configuration to turn off prefetch in? When the prefetch was added i thought there was also added a facility for reading in all of the meta reading from HDFS IIRC? but I cant find it so probably a pigment of my emancipation +Looks this has not been fixed +Committed to trunk +I guess I dont have perms to close this out but things are operational after voodoo incantations with the blood of an ox over the keyboard +To support class resolver configuration both the old and new configuration will be enhanced to have a class resolver configured. +Hi Gail is this fix in the Hibernate CP? Otherwise we need a CP asap +I did not investigate whether the ones already defined in other projects can be moved easily +Have you ever reproduced this with persistence enabled using this test case +After looking at the code again and tracing some history I have to correct myself +Patch break +Closed after release. +thanks! +I tried the patch together with andThey eiminate recent classlib tests pass rate degradationdrlvm smoke cunit kernel pass +I just resolved the link issue by adding the preprocessor definition DECAFDLL to the project +Change your protocol to t or take the default and you will be ok +Yeah I experienced the same zip file closed error +Hi Glen yes I was thinking about types but assumed that convention came from the fact that in WSDL a types container tag is used to groupschemas in WADL its a grammar tag so for the moment Id like to name it as grammars just to align it with WADLs applicationgrammarWe can easily switch it to types +You could also try that +Patch to fix the alter table issue +Liketestprojecttestprojecttestprojectsrctestprojectsrcmain +It also keeps track of the overall count of the Transformation Chain +Come on Roger what do you expect us to do based on the issue description you provided? Please take care of that or well close the issues with Cannot reproduce +I saw Release in version so I wondered a bit +So I would document what to expect +This will help us better match you to someone with knowledge about your objectives +CUsersismailseyfiDesktoptrunksrc +Yes it definitely isnt a clean merge in its current form +JDT and janino seem to allow but I am not sure if it is really problem is due to the being unescaped when we process the I would rather people did NOT use + +This required refactoring of +Please note that the version that has already been pushed to Maven Central was not pushed by me having the plugin and thrift core under one umbrella means changes between the two can be kept in sync. +Just change the order of a line can solve the deadlockAlthough the status of some logical nodes could be wrong while mapping physical node and logical nodeAnywayit will be correct afer mappingThis is acceptableisnt it +Closing it now +We ran into the same issue +Did you perform stress tests with many concurrent active transactionsIf you dont mind I would like to post a message to the seam to get the opinion of more advanced Seam experts of this component confirming that this change is harmlessI really dont want to be impolite or rude here and I dont make any assumption of what youve done here +For long string inside progressBar user can increase width in the style attribute +Added to node +good catch Fangmin +Instead of thinking of it as more differences to review think of it as fewer lines of code to review +Patch not quite ready to commit theres a big block in +Will pending from Hadoop have submitted a refreshed patch +Thanks for fixingJust a side note Where does the image come from? Is there a license attached to it? Thanks +Tell CDI tools to ignore validation of injection points annotated with Reference or Property or of types or +David thank you for the quick action please send me the whole jar +patch adapted for t know how to make Hudson apply the last patch on for testing +Fix versions should be and since they will be the next available releases of those two branchesThanks for the fix Hamlet! +Assign to myself +Switching our Thrift dependency to fixed the issue +Can this ticket be closed now +I reopened because I didnt know if you would get my last update +Resolved +A unit test that demonstrates the bug is forthcoming +See. +can you try this +This is just a dev convenience intended for apps +This causes to fail +closing since fixes are in released code +ping +please committed +Jan can I have some more details here please? Not quite sure how to read your issueThe style is saved with the map the postgis georesource does not know about style at all +Replacing this patch with an updated one with a proper unit test +The client is responsible for setting the service to where it got the token +There will be no type checking +Thanks Bill +Please revert this patch +VankataCan you be more specific about which components are inside and outside your firewall? Does your firewall limit outgoing connections to the httphttps portsAre you trying to host AMQP servicesqueues on a private network that are accessible from outsideThanksTed +Sorry bout that I think there were only or files that got changed particularly if you delete the enum patchGlad to see that development seems to be moving again +Same problem with Hibernate and EDIT +Moving out +however i have some recommends for anyone who attempts the sameBuild Recommendations check out code mvn compile if downloading of any dependency jars stall kill mvn compile and just rerun it mvn install export MAVENOPTSXmxm mvn true sitesite +Hrm Jira doesnt have a NEEDINFO status +fixed and added a junit test to avoid regressions +Out of date +The various INFO log messaged should probably be switched to DEBUGNo problem +am not sure about this change +Works for me +In cases where the module does not need to be scanned to find the class to be introspected I wonder if we should have an option to index the class reflectively instead somehow +Im closing this JIRA please reopen it if its still an issue in. +By additionally adding the Encoded annotation the value of these params will be provided in encoded formSo Encoded must be specified when we want to handle decoding ourselvesThanks for your time +Will help if we see this issue again +Its easier to add the code to escape than it would be to undo itThe escaping high characters issue will be fixed to not escape. +Youre saying that these problems are caused by an issue the external Scaffold Plugin that you are developing? Not by Forge core itselfIm having difficulty understanding what you mean sorry +Will work on a patch tomorrow +When extracting into new workspace after extracting run mvn eclipseeclipse and it now turns the Eclipse plugin projects into Eclipse plugins +The above comment is incorrect +In hindsight it would have more sense to put it as critical as it was a universal issue + +JohnThats very cool +By the way I changed the method name public void set to public void set +Im not sure what the deal is with this test +Thanks it was trunk indeed +forgot the images sorry +When I tested the download it didnt show an extension in the Save As dialog but it had one when the download was completed +the same code applied to cxf trunk thanks +requirementstarget does not move requirements component from trunk to trunkbuild and update path to requirements in parent pom +Let me know if this causes you an concern +think this is not a duplicate of because that fix was about notI will try to my code with the latest snapshot tomorrow +bulk close of all resolved issues. +for the patch +I am a bit puzzled how the thread can manage not to inserted the correct number of rows here if is completes normally as it seems to have done hereI note that again we see instability with but this time not on Linux. +Would it be difficult to allow the user to declare a class for the key when creating a table? I think we should be able to get enough performance and gain considerable flexibility + +Logging and view log in CVD work regardless so this is still minor imho +So Id appreciate this change being made in trunk +only store a large number of nodes into ZK the data associated with each node is very little +IIRC uses and an usesIt wouldnt be the first time that a bug was fixed in one but forgotten about in the other +Attachment has been added with description the class illustrating the to work out a patch +Reviewed and pushed. +maven may have fixed it could someone confirm +Any way changing origin of MZKREGIONOFFLINE does not effect much +Given that keywords and tags might be considered similar the approach that weve taken is to use the to provide Lucene with information about the declared tags for a given wiki page +For the special case when min max we want to shed the extra regions to those RS what has regionCount of +Let me know if you still cant edit +Findbugs not related to this patch +I am sure these tests were succeeding for me on my Mac as recently as Wednesday night +Also exceptions thrown by handle called from the try block will be caught and passed to hande by the catch block +If you like we can start working on revamping her together rather than focusing on these minor issues +A bundle marked as +Close +a lot for the review Todd +msec transferTo + +Can you please keep an eye out for the hadoop artifacts from? We can come back to this when ready +Thank you Ted +cordova src contains platform specific code to retrieve file however plugman util src only does this for android +Ok lets close the issue again I provided a patch for myfaced to enable it and an external way to fix the viewstate which works on both implsThe final fix will be which will get a protocol change regarding the viewstates or a viewroot detection method +See Claus Am I right in assuming you will no longer want to attempt to lookup the localhost value and instead will always set the value to Thanks keep it as is so the out of the box is more secure +to commit +Marked as out of date as fixed upstream as per Darrins comment +I completely agree with Farhad Dehghani it is silly to freese the complete message delivery when some messages shoot in redeliveryIn a few weeks we are again smashing our head to the wall because of this problem +I thought javaw tests can be run with IBM JDK SUN JDK and then on a Harmony build for a comparitive analysis +How do you know that a key will exceed a threshold until after youve passed it +Works locally waiting to test on QA before +I think someone mentioned that it happens on corporate networks when people with the same IP address log in +Havent had the time to put it through its paces but theres not a lot to it so +I will check and get back to s need this backported to the branch +I am working on section of the above requirements Modification to Create Account screen +upgraded to blocker since its probably a migration issueproblem of runtime mismatches +The patch makes sure any existing jar file is removed at first which could be incompatible and lead to +Bubble errors up to the caller immediately instead of trying to recover because only the caller typically has the semantic context to perform the appropriate recovery actionBump it up to then +Fixed text formatproperties for files corrected typo in notice and added appropriate content for attributions +postponing to ensure more users know that the Forge Console exists how about adding the Forge Console tab to the Perspective? This could remove the need to add a button to the toolbar although it obviously increases the number of tabs open and this might be viewed as a negative +in which case they should be legitimate +And youve got a point for the major version upgrade +there is also a trainings assignment where can i do that +The values could also be shown using editable Components such as validated or Spinners to provide alternative methods of setting or modifying the color. +The current situation is just confusing for the user +If no how do we reliably know that audit logs have been writtenThe RPC proxies implement synchronous calls +Errr too quick +I think that the workaround is good enough for this release and am deferring the longer term cleaner resolution to the next release +Patch to fix above +Verified by Mark. +Some lines are in English +Having spoken with Myat closing this until a better host is found. +I can reproduce it with the test case provided as attachment + +The one thats actually a WAR should the one named differently +This patch only for test purpose +I agree its very verbose and only went down that route as I saw them used in a few other places in the codebaseSorry about thatbq +Some databases support specifying collations on a per query basis +Merged to the branch in revision +All these hbck comments are probably for not this issue +Thank you for fixing the bug I might check the fix later. +Ill fix the tool tomorrow +Adds provider stanza for virtualbox and configures GB RAM for virtual machine +The downside is that the code is less readable +I did not use the provided patch please verify if the fix solves your issue. +Lets keep this one open some more cleanup is needed +If this is not the behavior you want you should not use this component +You should then check the Camel documentation for which options is available +We need to support both MTOM and with SOAP +we have optimized error message but for security reasion we leave the default security level like it is at the dont have an issue with this being the new default +See SVN rev for the fix +A possible solution I see for this is to keep nodes in the just map in Gossiper until we can verify that replication has completed +It was due to a problem in the initialization of the gallery +committed +Looks like a simple and safe fix +everything after that is from the will be files Im attaching are from new runs btw +Ive checked in webserviceconsumer +Committed in rev +Per changes in fix committed to and trunk +Improved caching +This fix also elminated all difficulties we have had withgenerating th XSLT and spec +All tests complete fine +Hi Pavel Stuart has fixed the code in AS please help to verify whether that solved the problem thank you very much +But here we opearete all restarts in same JVMI feel that should be the difficult here +bq +Is there a change in degree? Yes +Gary can you verify close +Work required to support updates to models is substantial and would have to be for to Future and making a known issue. +Assuming you are coming from a JSF page I think you should see it hit twice +Thanks bettydramit for responding to all the bugs so quickly much appreciated! +Sample project which has theIs there anybody who can help me on thisThanks in advanceRegardsChema +But I cant find the rows on the existing node anymore where have they gone to +We know it does not scale properly from performance tests we have done previously +Castor doesnt use for anything internally +If applied to trunk the test should provoke the NPE +mrepositoryorooroUserstucu +Automated publication have been triggered +Im attaching the example files +Note that the patch is bigger because indentation has didnt notice the revision so the warning its not the older patch +is different it uses a new file for every logBen was talking about changing this at one point +So this ended up being a major change instead of a minor oneExisting recurrences should still work the same with this patch +resolved in version and +Can you try this again with the QE testsuite using the CP branch +Ill be able to review your patch on Monday at the earliest so no hurry +EAP is our supported enterprise version of AS and you can file your issues against any release during its very long support window via CSPRHNIf you believe your issue is still relevant to AS then please verify it and reopen. +Main thing is update of doc +Yepp of course +Thanks for the quick response Im glad we got it into RC. +Wodens treats both documents and any other XML documents as invalid documents +works for me +I think the framework is fine and I guess this only fails some of the time? Do you use or other characters in your? If so you probably need to set the max in your configuration andor change the schema to make those colums longer +I bet that people put the ZK servers on the same machines as the RS servers which is a cheap way to make it work but not really a good way +But for this element the parents setting is inherited and appended to +We can do it either in this jira or some later ones +This may be the time to do those changes as theres no reason not to improve the codebase except when it makes patching worse across branches +WSDL demonstrating the issue +I added a flag to the streams generated by writers so that the readers can know how to parse the stream +If the problem here is that defaults to listing standard for the qt param then the fix is to change and leave that box blankNo problem for me to fix +I think it sounds like a great idea although I cant say how easy it will be to do it until I really look into it +Now VPE doesnt support seam resource bundles +My concern is Im not using options at all the is generated by my code +Okay let me test it +yes it is intended that if the connection attempt fails we log the error and stop +Cleanup the system VM templates from the templatehostref table +Id guess that youre running Java +The patch is wrong because of the the core and max number of threads needs to be the same +bulk defer open issues to to +hemera is called ubuntu in Jenkins +Fixed with commit with commit with commit fabbdfceffccbcbffafa +That makes sense +With the slab approach performance will suffer when number of ledger grows higher than the number of slabs +I think the HTTP Service is waiting for start level to release the bundle lock +Ive just committed it. +Oh sorry I didnt notice this was related to +Do you have any suggested list of proposal filters? What should be filtered and what should be sorted? How much of a requirement is that this should be an extension point rather than a hardcoded set of filters and sorters + ve committed this and merged to +Is it possible to have two separated tasks? one will be used to periodically export documents from Confluence and put them under docs second will be used after new release comes out will generate and update everything under releaseAnd there be still dedicated task to update the main Struts websiteDo we need two different repos for that to hold different per each task? +Closing resolved and verified issue. +Can you incorporate them too and post an updated patch in reviewboard +Good that sounds like a good approachI looked through the patch and it looks fine to me +I just all the +Benchmark shows that the change has no affect on performance +There is no need of any changes in that part +Looks like you need to import +Thats the job of the maven war plugin with well defined lifecycle and configuration elements +Project dependencies for one of the failing projects +I think you should simply enhance the logformatter to a degree that makes sense and any additional tooling either make a new ticket or perhaps a github project for the work +But im going to if the implementation exactly matches the algo +Also they are not always exactly in sync with messages by their design +Trademark by its very nature is an absolute monopoly on use of the name in that categoryFor an incubator release this permission can be as simple as an email message to the dev list saying that Yahoo! wants the ASF to distribute the software as Apache Traffic ServerThe graduation requirement is that Yahoo! transfer the authority to make that decision to the ASF. +The text This textshould be on the next page in the sampleis of course not correctThank you +Thank you +bq +defer all X issues to +Heres my guess Applications wont implement +Remove project accounts after copying mysql delete from account where type Closing this issue. +I tested this patch by ensuring that the KSSL principals were commented out when testing checkpointing with SPNEGO and likewise that the SPNEGO principals were commented out when testing checkpointing with KSSL +If you would prefer to let the crash than I understand +methods added +I have exactly the same issue with maven +Committed +Waiting for David J to get the poms updatedNothing I can do about it now unfortunatelyWe will need to resolve this post several other changes for m support +Sorry there has been a lot to do lately +Console is now defunct +Lucene can now check if norms are present without looking at the directorythis patch actually doesnt even write norms if not absolutely needed +Will update the ticket when its done +My suggestion is that we drop Fraction from Lang and point people towards Maths more in depth implementation +Committed revision +Tweaked javadocs Removed some dead code nocommits Besides the java ghosts for which we will close our eyes and hopethey disappear I think this is ready to go inIll way a few days and then commit +When standby switching to Active standby first goes to recover of inprogressx +This version does not automatically do the detection it would need to be manually called + +Can I just add that in my opinion the best solution would allow configuration of the server before deploymentFailing that it needs to be done with the upload or between upload and enabling the application +I tend to adopt Jeffs change since it requires less change +Obviously with concurrency it gets more complicated +I think this would be extremely confusing with multiple fields +Can that JIRA be marked as resolved +covered by the buildxlib functionality of xc implemented by Dave C +This appears to be a duplicate of +Working on this patch now +I verified that without the code changes the new lowmem test case fails +Dhruba can we instead use the generationStamp that the NN maintains persistently on disk? Isnt the txId different it is increment per transaction +We do not provide full application transparency such as restarting transactionsqueries in the event of a failover +Should we create a branch +I will test this on Windows and see if I can reproduce it thereA couple of thingsYour shell transcript shows grails Baz +I dont think this is a good idea +The patch can be checked in now + +Committed revision +Marking as WONTFIX I dont see anything being done here + +We may improve it later on +Ok Ive committed some changes around this feature requestIn the View menu and on the popup menu when you on a node in the view youll see a new menu Freeze Node on +Might be related to old version of we faced this issue primary on. +Thanks Mike! +Fixed trailing spaces +for the patch +Test comment +fa master masterThe commits are fabcadeabefeec and progress on this issue? Does the problem still persist +This file is also supplied by JDK +Back ported to guess this should fix my issue as well? I have to check it out +Having the docs hosted on Apache infra would look more professional in my opinion +A comment on RB +The test case +What is the id? Any ideas when it will be implementedreleased +On postback the client sends the view ID and JSF retrieves the view from the session +ul rule for displaying it correctly in bizznesstime themeThanks Deepak Dixit +And we may still decide to keep it as is if there isnt a need for this anyway? +You cant maintain aspects as they will be overwritten by Roo +Please provide steps how to verify this Jira +Created a patch to decouple the update method + +DanielHow should this be done? Can you explain in more detail what you mean +Simply take that attachment call it and plop it into the package andyoure done +Yes if we can do this in slfj then we should close this issue. +Change committed to trunk +srcjavaorgapachecommonsconfigurationbeanutilsimport +In most cases the API was intentionally keep the same as the Java client in order to make future porting easierIve added a second patch that adds the code generation necessary to compile to +Do I understand correctlyOf course we could say that since the Codec has to peek into thereby making it it should not encode stuff in a different format but is not enough to enforce thatI dont think that gening FIS solves the problem of obtaining the right DVF in the first place +In my setup the error is raised even if there are no spaces in the tomcat installation path opttomcat the webapplication context path opttomcatwebappsvomstestvomysqlSo it could be that the presence of spaces isnt the only reason behind this bug. +To be pluggable it sure needs to be modular first and this requires removing the relevant global static variables in the UGI class +I plan to commit in a few days +The synchronization on the collection of replicants has been removed from so one thread trying to registerunregister a listener while another is sending a notification while no longer cause a is only one of the deadlock issues related to see the parent issue for the other +Currently the only way to destroy destinations using the management API fails when there are customers bound to the destinationThis means that using the destroyQueuedestroyTopic methods from theremove management operation will fail if there are any consumers bound to the serverOnce allows to destroy server resources we can fix theremove operations and make sure JMS destinations are properly removed. +Does that sound reasonable to people +Assume closed as resolved and released +Manually tested! +Ive open new issue with the test +the parsingWhat happens when the same format is used to create a new recordThe format should be symmetric +Verified during EAP ER test cycle. +Thank you! +Keeping the list of bound values tightly coupled with each prepared statement token greatly simplifies the binding alignment when the number of operations is large +since EAP is code freeze now im setting the fix version to on EAP +As a aworkaround this should be sufficient but I agree that we should look at this loop more carefully +Btw Pete refresh my memory if necessary but isnt gonna enable such API? If so maybe worth waiting for and then simply implement it rather than thinking about the SPI itself here +The PHP SDO implementation makes use of this behaviour +I am not familiar with JSON so I didnt realize the original format would cause problems +Wim you can use the command line interface for Eclipse update manager to create a local mirror of any update site +Committed patch to resolve issue. +I couldnt find that augment section +Hi Gavin Ive sent my mail over to the guys for this one +This is partially completed +Its generally recommended to run nodetool upgradesstables between any two major versions prior to doing any kind of streaming operation like move bootstrap or repair +Discard erases the field buffer or the file +Heres a new version that fixes a bug Patrick Wendell identified offline +I think it can be reproduced using Resin as the JSP container +looks like a duplicate +Its a way to indicate that a post is aspecial announcementthat should be pinned to the top of the of a weblog site +all is with a scale +If we do so for solr what do we do if other such extensions are requested? Wouldnt some kind of plugin mechanism be a better solution for such cases? Having said that I havent thought how such a plugin mechanism could look like nor how difficult this would be +resultput writables to use protobufs +This is not really required as gives the access to any char in the string +If someone is only iterested in some of them he can implement his own filter to such events. +Upgrading to or later makes the warning disappear +This means that in the code that Ive submitted doesnt solve the problem +added ruby support to microsling in the whiteboard repository +Hi Brett I cant seem to replicate this +HiI verified this issue +Assign to Daisy review +Good jobThis is really necessary +Btw for curiosity sake what is the correct way according to Oracle to quote the username in terms of opening a Connection? +I agree what I meant with my first comment is that just adding the annotations is not enough we first have to change the and to use annotations instead of theLet me try to get some things working here +by failing builds because of it +A user reported that the server became unstable when they set this property and we could no longer update the because of a missing parser exception +How does issuing a point release work? Ill remove this from the code base but if we want to release that update and update the package do we have to go through the whole release process +when I apply your second patch I get the testPredicate failing only messages received instead of applied with thanks! +Thanks Mark I have closed this issue as you suggested +Fixed +Thanks. +Sorry i forgot to mention that i need to render this highwayroad with different colors which represent the traffic intensity +I think Andreass fix may be to a different problem in a similar vein? is not on the stack for the I reported. +Integrates Prasads comment +Is it possible on JBOSS? I made a search on the documentation and didnt found nothing +not said earlier we would make sure me was upgradable but now we keep a feature instead of listing the maven features? +Checked in as provided above. +So how would you fix the hole where we set region to closed state when we return to the regionserver the CLOSE message but if we clear the closing flag because surely a region cant be closed and closing then the region becomes assignable though it hasnt been closed on the regionserver +I figured this out by modifying the to take pageSize as an argument and to count how many iterations of insert and interrupt it took to fail +Of course error message in this case could be more friendlyAs I reported previously in Shopping cart still shows the price without VAT +Where does this YAML come from? Im ok with fixing it and MRI doesnt blow up on it but its not really valid YAML of any kind +At any given time I may be concurrently working on multiple versions of foo +Could that work +Simply pasting in animage might work but how would others in the community edit thatimage in the future +was a bug in the previous patch which this fixes +I am suspicious of that to cause the issue +Please provide also the patch +Pretty sweet tests +I started with one of the projects objective restructure and its related classes to support number and text range queriesI created some querynode interfaces such as that abstract the value a leaf node may hold since now leaf nodes do not only hold text anymore but also number valuesLet me know if you have any questions or any suggestions about the codeI expect I created the patch correctly as its the first time I play with UweThanks for quickly reviewing the patch +Ive switched the dependencies so they depend on the recently released versions so hope this should now be fixed. +Patch for tabbed form example add Javascript option for form validation +This looks like Bundle and Bundle are seeing different versions of the class exception Z +There is lots of code ready to be imported +Thanks Bobby +At some point in time we used to pass entity class loaders to the CL helper class so that the CL loading the domain model would be used to load user provided analyzers fields etcThis code might have been removed but this was added explicitly based on user deployments issues. +Ill add this +Unit test showing the issue problem and its configuration attached +Good catch My desire is to ensure that checkpoints are flushed prior to taking new writes +May be I dont understand it properly but I belief we can do that already +I do believe it could be that simple just remove the sites but wanted to get it confirmed +apply +We look forward to using in Giraph +Is large number supposed to be hundreds thousands tens of thousands hundreds of thousands millionsIll be surprised if it ever crosses a few s +Ill redo it and attach another patch later on +In the current design if someone wants to change the capacity of the memory channel they will have to shutdown the current agent and create a new one losing all the data in the memory channel +are they working +Sscoping items for early release of Java Eclipse Mgt Console +A b d and e are doing it for me cant say Im a fan of sets of a few random dots could be highlighted in some way to depict activity of sorts? But yeah as a general direction I think were winning here +I was not sure how to modify the properties file for chinese and japanese +The whole point of releases is to move away from snapshots and rely on solid releases +Validated with test branchesenginesrctestjavaorgteiidcommonbuffer +Requiring any user data on the server for anonymous users is a defect +Released. +Makes sense +Unknown elements and attributes are handled as out of typesystem data only if a handle to the is passed in to the call to deserialize +As such wont it be better to keep it in +I like that gists dsl much better +Committed to trunk. +Patch applied with many thanks +v +HOME is not set is not correct +Use columns attribute for this to work +Reported to affect proposed release candidate +Thanks for catching it +At start of task do rough calculation using representation of start and end keys to figure and done +Adding keys dynamicallywont really keep the tree balanced in my implementation as in myimplementation the tree is balanced by ordered insertion of keys +Vasilythanks for the patchIt has been applied +test attached +Still exists using Hibernate stopped throwing exceptions ages ago +Thanks patch applied. +Apparently they were republished just very recentlyYes supports this as the forms code is shared by the x branch and trunk +I think the FST created in could be used to compress the rowids in the while simultaneously enabling fast lookup +Godspeed sir + update instructions to use public repository URLgit clone gitjruby I clarified the instructions a bit +We have missed colors so need to add it processing for vpe with default values +The produced Maven artifacts will just use whatever version is in the POM +Seam Team Great work! I love seam +But only for and not releases +Thank You. +Grep on that +Just setting it up so that the mass move of unfixed issues to will pick this up when is another test file with danish Andrea text in the map is correct! Thanks +Is this good to go for? Thanks +Alright closing the JIRA because it is resolved inIn this case the resolution probably should be Duplicate. +Is Isaac doing this +It may be a coincidencered herring +This could be solved by setting the resolve order to CLOSUREFIRST for the closures the config slurper works on +Solution is to escape the delimiters for the output table +It uses for this task +I was able to scroll an entire OU entries at a timeSo it might be a limit on Active Directory or the user youre using to browse entries +Version Released +Every char referenced in the antlr grammer is implicitly added to the charVocabulary and when you say f it means every char other than f in the charVocabulary +Tagged version as since thats the latest version we have +I am leaning toward the second option of indexing all links but removing words +If you grab the latest then that includes the Apache Geronimo Validation API and Apache Bean Validation implementation +Strangely the commandline interface does work with the xsl file and xml file for me too +A red X appears in the POM beside that jar files name and the error says the jar file cannot be hope this has been more informative. +Jira site is in Java doesnt take security lightly unlike some browsers out there +Im changing any tests extending or its subclasses regular tests can be converted to as they are touched later on +can you please revertreopen and do this +I will commit after Hudson returns +Id consider immediately setting the dst to be the srcName when dst is the cwd +This is a error not a programming error +rApplied fix to remove the method from API support it in implementation via default access and remove redundant calls in source code and test cases +Or build xwork from source locally though a couple of tests are failing for me when I try +It would be very much appreciated if you could help out with that Phiilip +As no longer allows specification of the view server via a MIME type this request makes no sense anymore +With a centralized XML file that Seam uses you know all the will be stored in one place and itll be located under a particular view definition which is easily navigated with basic editors using search +Thanks +Good call +This should be done before Ill address that in this ticket +off to +Anyhow can you please elaborate on whatyou are trying to do? I am very much interested to improve Castors transactionhandling come whatever +jcrmixinTypes a +I dont know whyEdward Capriolo Added +Cannot reproduce it anymore +The main problamatic uses are caused by the use of gethostbyname and inetntoa functionsThose can be avoided by using libcurl and httpd transports +For safemode we should count only complete blocks and finalized replicas +linking to related should be a lot easier if we take care of first +Click the all tab +Allow one or more commas within arrayrank brackets in the type node of the expression grammer +The removal of code related to which was added in Maven fixes +It might just be too late in the dayWhen the distinct issue is fixed the Army reported is avoided since is not used +I dont think we have the right to include +Please close this issue and file a new one. +We should be fixing the util function that parses the UTC date which youd notice under that method + we do this this is going to be an incompatible change +I favor chosen by JT over chosen by for reasons +I do however think that it needs more unit tests to validate that the serialization works in all instances and it needs the same technique to be used for atom serializationThere is now an based implementation with unit tests that worksIf this issue is still a blocker I can complete the integration of this converter and port the atom converter to use the same codeSo +The RI caches Long instances as well i believe the javadoc even says this +Yes exactly +Its in rev +Hi Martin could you attach some screen shots so I can see what is going on with the prompt +The extension framework part of the spec was broken for some time +There are a bunch of tests in the tree that do client expandcontract so I didnt create an explicit test case + +This takes about minutes +Ive reviewed the patch and all looks good with it +It now throws an illegal argument exception. +Committed development branch revision Thanks Christian +and so whatever Apache APR API does we should do +the Exchange object shared in the inboundoutbound message will guarantee the mapping between request and reponse +Closing this issues as Wont Fix +This is the initial version +complete +This works fine under version of the plugin +Ah damn I thought it was fixedGuillaume +Changes in blank node label handling have reintroduced this bug + +changed priority because it is an optical Problem is resolved in the +Bulk closing of issues that are old and have been released or marked invalid. +This is still failing if the parent of the new node happens to already have a child with the same name as the child node thats supposed to be autocreated +I have committed a fix +Hmm +Interesting +resolving this since hod is checked into the source code. +There is definitely a Download Template buttonlink when I upload the template using Cloudian S server +I will post a separate enhancement request that should address this +Lets not change the NOTE back to TODO Wrap the long lines +The first time I choose the sharedraw widget after starting the server I get this behaviour +Verified in EAP. +I read in the readme about number of map reduce jobs hive require to handle group by +Returning different error codes might help rather than return for all types of errors +Yeah I can make that top +Ive written a whole new partitioner and refactored the examples to use it +The same changes must be ported there +Unfortunately I may not be able to look on this weekend as came out of station +Committed the patch to trunk +ya please commit +If the original session does respond within the timeout the broker can respond w the session attach error +Hi Marcel I applied your change but it also fails with +Also with permission I might be able to fix this +Hbase can add last if else with this public API +Thanks Ashutosh +regardless of whether youre current most definitely made his changespatch before and so may have run into conflicts there +The edit log showed that the file that caused the problem was moved to the trash right before it was open for read and thus reset its access time +NET compatibility will stay compromised +Thanks Jukka! My change in revision is a bit hacky because I duplicated some code +Installation it to Radhika to take care of documenting additionals teps needed for installing CP on it unassigned +It was an issue with the test which use the old blueprint bundle version +Not for inclusion +NEEDDATA will be translated into return false and the loop ends +This patch adds a sixth factor variable per site +Dymitriy can you check that latest patch? and are the same +If folks think the bug is serious enough we can certainly include the fix in +Sending Martins patch again but moving the unescape call unescaping attribute at level because uses its API directly improving the test case to cover the escaped attribute value add in with r. +no alpha is out being tested +Hey Rahul thanks for the patch +Hi Leoni have problems importing your patch zh looks really strange in my eclipseD +I used XSDanyURI service for this because we can test new changes of wsdd files at the same time. +the guava Optional wrapperThere is also a feature request to provide a generic version of Frequency. +I meant something like this +Its better to throw an exception or replace characters or whatever when generating the XMI than making the user believe everythings fine and later they cant get at their results +failed or +As for including the protocol nameNormally a server implementation should hash only those methods in the interface named by the client + patch depends on POI so well need to wait for before applying it. +These are more or less notes so that I have something to make me remember where I was at with this +so my change will simplify this testing by providing automatic another port binding +Attached different repository providers to track this problem +Fixed problem in which was causing this error no delimiter was set after each of the row paged part +If we choose to remove snapshot mode interval this kind of exception can be reliably propagated to the request +which is clearly undesirable +Can anyone else verify this workaround +Good idea +Were trying to build at least some sense of consensus here before moving forward full steamken gave a +fixed in and to and resolve to both versions of a bulk update all resolutions changed to done please review history to original resolution type +It wouldnt run by just doing npm install g cordovaYesterday I tried copying the Android SDK folder to Applications +lock available in the repositoryI guess we need to improve the JCR initialization pieces +asplugins +I committed this +Arun could you set to a ? It might gets around the deadlock +I think I found the cause of this +To fix the failing test I took out some of the slots based code in which doesnt really make sense in the context of YARN + also affects other notificationquestion cases. +Unfortunately I will not be at work in a few hours I will try tomorrow +crc filesblockCrc upgrade does not throttle deletes explicitly +The event handler can detect the end of the message when outputReady event is fired and update the context accordingly +Running mvn test locally doesnt really show this issue +This was a Maven Issue +Marking as fixed. +Made the access the vdb modification synchronized +Will investigate +Curious why? Is that an arbitrarily large number? Is there some better value we can use to assure no conflicts here with other use cases? FLTMAX +Whups typo in the topic name +great work +But if someone can help us to createmaintain it then it can go for it +in response to allens commentswe expect that the admins will have some common subtree for usersappssomething like appsapp so in that case it would just add one node for a new user and in zookeeepr node corresponds to bytes or soso usage should not be a problemalso regarding administration we will expose the setting and getting of these quotas using jmx and will have command line tools to do that and get stats +Exactly +Applied to r. +Attaching the final diff for future +Does that work from the command line +And updating the namespace prefix might not be something unaccepected to clients as a result +This appears to have been fixed for other changes possibly during the refactoring I did. +In fact I doubt whether I can do any development work in this month +this patch does not include test replaces removed API with new one +Made a workaround for this issueDoes all the folders in and srcgroovy and srcjavaIf someone could point me in the direction of passing variables to this script I could modify this to accept additional directories +Passes fine +Looks like you like the idea for system +yeap +As mentioned while it is binary compatible its not source compatible as source that compiled with the old apis may no longer compile with this change +Implemented for Java Backend in change committed in +Thanks for doing such a detailed performance analysis on this Patrick +Could you retest with version or newer +should be If a constaint is violated in a date box the error message is displayed under the field as in above +The latest patch looks good to meIn my previous comments I agreed with the decision to put TIMESTAMPADD and TIMESTAMPDIFF in the builtin function section +test case for the symptoms different implementation +Its just a different opinions on the definition of the applicationTypeSometimes users want to have both information the computation framework and the source +Fixed patch +Otherwise the problem you describe seems hard to believe given the frequency that people have been using relative for the DTD over the years +looks goodIve committed this to trunk and merged to +I am ok with doing this change now but would be more comfortable if a test enhancement request was filed before closing the issue to write a test thatcould check that the correct XA global xid is returned and did not suffer from the current ordering problem +Fix merged into trunk. +IMO if this can be enforced during development we should +Looking at the socket cleanup now +But honestly Im not sure there is a point in disallowing tracing for some query especially since thats more work than doing it for all queries +Could not reproduce this using the unit test to check the values +its valid url +I loaded up a bunch of regions then shut it down +Committed revision . +Issue may already be resolved may be invalid or may never be fixed +Logged In YES useridI do not know and I guess it is really up to the core developers +As a result we are closingthis bugIf you can reproduce this bug against a currently maintained version of EJB please feel free to reopen this bug against that versionThank you for reporting this bug and we are sorry it could not be fixed. +All I guarantee is that the code compiles and the runs from inside +you can put properties file into the classpath as well as load from an external link +Patch for this you MartinDo you see any other improvements? Do you have data that can improve the test coverage? released version. +are mutable so if two different requests share the same cached UGI then tokens from one request will be shared with another request +I committed the patch to trunk +Adding an entry to a full queue causes a thread to block and removing an entry from an empty queue also blocks the thread +Hi RohitI dont have the permission to assign case please refer to. +Is there a timeframe for when this might be available? or is there any early access to the code? We need to implement timers and Id rather look a consistent solution so when this becomes available we can remove our copy. +Oops yea I missed the fact that the caching was removed again in the later patches +support is delayed to should address the SVD issue +Thanks Nicholas +I would not check for log as it can be located on another disk and figuring that out requires access to the file +Ah I debugged a bit and discovered the issue is likely because your http proxy has scheme http +Yes Ted + +Just wanted to get it going so the quickest waywas to removechange signatures +I actually got a user request complaining about a similar issue lately +Is this in the works or has it been taken care of already? Shall we push to +The old release uses runConfigure which is compatible with the old release ofThe current makes more direct use of the autoconf and automake tools +I saw the same errors on hudson in the results of other patches +clarification i did not restart the server for each hotdeploy ofcourse only once after switching from to have a yourkit cpu trace log available for both versions +Thanks Thejas! +The new one for is maven compliant since the artifact id is +We wrote a workaround for this as we didnt want to modify the hibernate source +This way the user never actually has to specify or know about the vector or raster +for GUI +Please supply more info or this needs to be closed +pwdhomedwardreposjbossesbbranchesJBESBCPproductsrcmainwebapp svn commitSending webapp webapp file data +patch +Username radl +Im OK with the current approach with one suggestion Id prefer the sizecollection pair implementation detail abstracted into a local utility classMakes total sense +Patch to fix thread for the issue contributionis there a particular reason to use thread locals instead of singletons +This affects to version too +In rev I changed this so you get to specify the bootstrap partition you want to use when you are assembling the serverI dont know how to see if this breaks the installations but the integration tests pass +I can now actively connect using ssl and my local certificatesI cannot provide these certificates they are corporate internal but they should be pretty standard so I would expect this to be easy to reproduce with other certificates +Also be aware that now test coverage can report which tests cover what parts of the code +Thanks +Sent latest concepts to Randall via two rv and rv +Its not a bug and there is no patch Postponed +I not sure if you mean me with Charles but I switched to the latest snapshot release from the apache snapshot repository and tested the attached test case and my project with the snapshot +If user provides signed contributor agreement we will try to apply his updates to fix +You can find the additions be searching for include this class into ojb because it depends on oracle classeswill make set +verified on all rhels +I looked over whirr for deprecation issues as it uses things in guava that have replacements +Note Lang does not escape above xf for HTML +We can also include massive Shapes and tiny Shapes circles points and whatever else we end up supportingHaving this standardized Shape suite would be a big benefit to testing all the Strategys +Please let me know why you felt it might be a problem to print null as the label if the label was null? I think I may be missing something +Hmm +Hello Petr I believe that your problem is that your admin role user is not also in the role of user +The issue is more complex than I thought +a test case and I commitr +on Benoits structural change this will also get rid of the prerewriteuserctx dictionary itemIve pushed a patch with a test that resets the count when a new request begins +No need for IT +But it still needed to be added as a in so that it can be mapped and handled by Castor +The work for this task is complete +I started to do so but there are some unanswered questions to deal with Where does watchdog trunk go? Do you want to be able to tell the two sets of branches and two sets of tags apart? I dont see any watchdog tags +I just committed this +Id be wary of changing the default especially if this is means to go into a branch +Adding a fix for this issue with a slight tweak on Bjorns unit test +This issue has been fixed by correcting how WSDL is generated and the correct namespaces prefixes are now usedPlease let me know if this is an accurate descriptionThanks! +Attachment has been added with description Proposed patch has been added with description Revised patch no more idfirsts attribute in area tree XML +I have filed where you may want to contribute the patch + +Attached a patch to show new usage when hadoop fs mkdir with no arguments +We need more people like you to help +r on trunk. +I agree that this does not need unit testing because we are just changing the log level of the items but we do need to test have tests for log entries when we can +bq +Tabs should have the same height + +The patch is for the code as it currently exists in SVN not for when the first patch is applied +implemented in java need JSPHP implementations +The problem was that the layers was keys in a +Is your project an open source projectIts an order to have a project to test this can you provide a simple project in order to reproduce thisThanks +To +Duplicate of +It doesnt matter that much if the background is black but it wouldnt hurt to fix it just in case we end up using it with some dark background that is not black +Closing resolved issues for already released versions +If you can do that from Java you should be able to +In particular the footer processing in still calls a static method +EAP has switched to the convert at build approach so raising to Blocker to ensure this goes into the next release and the change is not lost +Because the Administrator determines that the user Organization +Not sure how I messed up eclipseThe data members can be package private now that they are in the package +Includes new etap tests as well as updated the JS tests +Defer all unresolved issues from to + full log all jar use by application axis conf file +or something like that seems better +I forgot to commit it to M branch +build and deployed and the configs are generated +that method should be would be incredibly useful if indeed it takes a Collection instead of a List otherwise I dont see the pointVoting for it if the title is changed + connecting to a different cluster but the old one notices it is back and starts gossiping to it againActually this cannot happen since gossip SYN carries cluster ID and only messages from the same cluster are handledI think removing fat clients automatically is OK +Can you please close the bug once youve assessed its correctness +One more thing Does your patch work with too +and branchx rev + ER it is called jUDDI Registry +to the new changes +The second patch implements comments + +Reviewed all these issues and marking them as closed. +The warnings were added in to help with cases like this so you wouldnt have seen them before even if youd run from the console +There was a recovery issue with the LDAP server that has now been fixed by Contegix +It is reproducible without Hibernate installed as well +This is a Spring facility it can be useful to make anything that uses Spring a little more easier to configure +Jsp file with example was attached +Resolved with fix +Ill ask Alan to give the third vote +If so follow the link to the to see if the assigned thinks it is still running +Jukka +Exactly +Did you try to test with the above Perl code instead of using Java Stomp Client +This is not the correct place to report issues with Hibernate +I think for your particular case a custom phase listener that set this param to true will work better +Iirc any version changes for the quickstarts was going to happen in the jbossas repo +OK now I understand your comment +CDK bug was fixed in Jun for +I will continue to evolve these examples but I think these should live in a wiki +I HAVE t but ts say student license and limit +Instead XSLTC selects the child element of doc whose position equals the number of attributes of doc +If youanyone else has better wording please suggest +I dont see any folding available at all when I enable you show a screenshot of what you actually getset +Here is the module dependency with a classifier which is used for testing +What about the commit emails? Did you find something wrong? again now +You can close the issue thanks +Hopefully will find its way to a maven repo soon. +Committed to with revision . +Heres a draft please review and correctI wonder if we can now remove the reference from our mission statement at the index page +This is great +I am not sure I understand + +Im starting to come around on the idea of having our Enumerator thread pretend to be the same as another thread +Im planning to merge it to +Brought up ui and links are present and working for task counters job counters links there and working when appropriate and no link when no mapreduce task affects +Doh I should have tested the patch a bit the +The patch needs to be applied against the root directory of project +Are you sure this an issue with Lucene and not your code? Any hints on recreating it +Passes mvn verify etc +Committed to trunk and +The fix is in the repository pending message files regeneration by Alberto + +thanks Jiannan +This would require though someone to invest a few hours figuring out how Firefox Chrome and IE go about generating multipart content these days and whether or they behavior is consistent enough +JD punt to +Marking closed. +This parameter is required by the Ruby On Rails stack to correctly perform redirection +Committed in rev +this should make things a little more readible as well as improve performance a little +Updating the bug summary and description to make it relevant only for trunk +Feel free to open up new issues +Inactive for too long +I integrated them into the integration test I wrote to verify the bug fix +release bulk close +we need an agreement about the before its possible to start with this topic +I may have more time this weekend +I havent yet tried wil this patch +Patch checked inOleg +I mistakenly thought that my commit is IPF specific and therefore didnt check on other platforms +Reopen per user definitely has to do with the xdoclet plugin using antrun directly and then the build later trying to use antrun +Well work on it a separate ticket +I will take a look +Im attaching a preliminary patch for the and part of this issue +Bringing a comment from the last time that we reviewed scheduled issues Last time Ross said OK to go into but we will need a FAQ entry explaining how to turn caching off for development since with it on we are forced to reboot Forrest each time the LM is changed +The hbase master process normally called hmaster is often mistyped or misread has hamsterIt is very distressing to see messages like Hamster out of memory leads to crash. +Im uploading a patch that does that +But it is no matter the problem is in debugging this project +Verified by Mark. +Communication between bit Windows and bit Linux worked without you want me to look into issue with bit Windows and bit Linux you have to provide me either access to a bit Windows box or a image of it so I can test this here +Further to this our configuration is that all slaves run both and processes so they are doing double duty +Patch part introduces a for instances in +Fixed with Kens SVN +Our application connects to a remote server so we are using async functionality to process the requests more efficiently +I think there are just two one line changes actually required to set the Schemas in the parse and parse methods so I have just made those changes to trunk and the woden branch +All subtasks have been already verified +As to why this can only reproduced with winXP when using AJP connector remains a miracle +This has the same changes for and as the patch but the README has fewer changes mostly because the README has had fewer overal changes than theOnce these two patches are committed this bug can be closed +as we need to implement this in our deliverable soon +Please check that it was applied as expected +Not a bug I think. +But in Java spec +Ive roller +I see the problem now +Passed all to trunk and +I believe these can be dev playground. +I assumed the string was parsed on the RS instead. +Commit changes to trunk at rev are many openejb document on how to configure the deploymentId with modueId +The list is in description of the JIRAI got the original list from another team +Actually I dont know of any issue that is caused except of some warnings of a library Im using that does custom classloading and gives warnings about thisIt my be true that Commons Configuration works with both versions +as for patches Im trying to see if there is a good way to do paging queries on Sybase havent come up with one yet +It will keep trying to to the SNN and will succeed when the SNN comes back to life +It works for me adding and on clean tomcat have tried to deploy registration example and works fine also messages example but it needs to install and into tomcat shared lib think Ive worked out the problem +I believe Jozef replaced the struts JAR not the groovy JAR file in our build hasAs long as it is in the war lib that error will go away +It would look better if there is no New menu item but just Component +This will make it easier community members to find and work on themFuture releases beyond can be done provided a community member steps up to coordinate them +pushed to in JBDS Build id +Not sure how we can better test the stability +The trick is that only one source folder is the default and its output folder is set to null in the classpath variable +scheduling for depending on testThanks for doing the unit tests that saves me a job I was planning on doing today or tomorrowIll test them with my patch and see what the results are +Having read about your system for demo projects just now Ive added one no JSF based on +Intervention is required +To get them under one version we would need to convert the web service client code of the XML and Salesforce connectors to Axis +This is still happening with the final release +Committed the patch for M +RegardsWerner A definite bug methinks +This problem has been mostly solved by making the global data internal to Axis and putting the typedefsenums inside the axiscpp namespace +It is my intention to retain these assertions for a few more releases while the new code is exercised more thoroughly +Afsal could you describe your environment with sqoop command that youre trying to executeJarcec +Unfortunately if the block of code within the loop throws an exception then the file isnt closed +Software grants where already faxed to the asf office +I think it has something to do with linux at least in my default java installed on both Fedora Core and Ubuntu seems to be the JRE not get the Sun JDK you have to install it and then configure it as an alternative using the linux alternatives is confusing to say the there a way to tell if eclipse is running the correct sun jdk from eclipse itself? If so how? Then we can supply you withbetter data to your questions above +I would like to go ahead and commit this patch soon +The code isnot yet commited however Ill commit it later if no one objects +Vladimir yes +Need Hbase binaries built against Hadoop for this +All I can say right now is that adding the to the classpath triggers the version of Linux are you usingIs that a version of Java which came with your distro or one which you downloaded directly from Sun + +Attaching a patch the should fix these exceptions +So really that slowdown we see right off the bat for the updates case is the added cost of having to check theSo the test was unfair +Harsh In discussion with Kevin this issue appears to still be a problem +Dont remember that thread and I cant seem to find it +Currently the code is specific to fields having name fromDate and thruDate further we can improve the code and make it generic +Yes +SighEnclose c in a square bracket and then enclose b +The problem with this is that we have about a x ratio so recompressing on each read is now recompressing the same stuff times on average +Submit patch is available again now that Im assignee +Cool! This is an illustration of why one big example is better people test itThe convention in solr is to add and collectionsolrsolrcollectionsolrcollectionconf +technically this has been a problem for a long time but we probably havent seen much about it because our page cache efficiency is so low +Never used in be +ok get is goodOnly thing ill miss is what the specific sql does +Unscheduled for CR will be investigated after CR is released +Could be empty string or +Committedrev bcf +reattached with change included +Any rubyists want to review this patch? Its minor so if I dont get any feedback Ill just commit it in a day or so +I dont see a on the apache website but they have a which they claim is binary compatible with checked the manifest in the binary jars they have the proper version in there +fixed in BACKPORT IN downloaded the final version tried it in WSAD for WAS +I saw this with other tests too It looks like it may happen with all tests that extend +Still a problem in revision +Thanks i will give it a try +Better to work on hosting on ZK now in that case +bq +namespaces in the check like you are doing +Wow this is the fastest turn around Ive ever seen. +Trivial patch implementing the suggested fix +Built acounting webhelp web apps +I can attempt to edit the XSL files +Pete showedAmazon All Images shows imagesDeltacloud http access gives a few s and eclipse tooling shows the sameSo obviously Image view is not really usable should just be a Find Image search dialogwizard +Ill be happy to reopen the issue if more changes are needed +When is getting released +Eclipse JSP Editor works in the same manner so this is not vpe my experiments sometimes vpe save very quickly for the same page sometimes not as I think this depends on free system memory. +I still think there is something that Im not seeing because theres a plugin that you have installed that is behaving unexpectedly +committed the additional files from in rev +The error does not occur when using ssl connections +If you wish to change the contract for if you wish to remove or however you wish to address it it should be addressed before you commit rather than by committing a patch that causes existing tests to fail +Nathaniel could you see if the following patch fixes the issue +Both give same perfect matching results but first requires one more condition +If you dont start this issue I would like to resolve this issue as soon as possible +this issue fixed for the moment +Earlier versions of this patch did use genservers and they were neither less complicated nor did they give the desired performance improvements thus we went past them +Add sasl PLAIN security value +test build output for the record cd xulrunner export MOZUSERDIRmnthudsonworkspaceworkspacexulrunner MOZUSERDIRmnthudsonworkspaceworkspacexulrunner chmod +update fix version +Thanks N +bq +it depends on the use case +This patch depends on and this issue to subversion as version . +It should enter safe mode in this case +Pushing to to out of any release bucket for now as progress has been slow +closing the issue due to missing information possibly not an issue as can be used. +All tests ran cleanly +Successfully implemented +It would be great to see this project implement SFTP support +Party solved with the topology view +Patch for adding clean attribute for wagon it is added the ability to delete stuff would be a great feature +Not sure if this is an issue related to Open JDK +fixed in r +Can you generate patch for trunk for Hadoop QA +Hi OlegWould you be able to review this onethanksHow exactly does the problem manifest itselfUsually if a channel does not have any input it should return not +to display proceduresfunctions indexes views synonyms +Actually well copy and to and +Could please someone take a look at this issue? Theres a test case plus patch attached +Yes I tried to remove +And in the all complete paragraphs but only references to for green field projects I will not model my objects this way +No it is meant to be like that +I think it would be more correct to use jlong in VM native code as well +Reopening this issue to provide complete solution for both ways of specifying which subprocess should be started +this patch causes some unit test failurs +Changes made +bulk defer open issues to problem is that the only check if sources have been generated +but for this case I do not understand why any tests would trigger this since they dont require usernamepasswords afaik +I deleted dutch french and germanSo for now this is my workaraoundYET I still find it strange that I get this errorSelecting the english as language as user in both cases the installer uninstaller quits with an error when other languages are possible +ok +Thanks for the review Vinod! Attached is the patch with the suggested changes +I believe this issue is resolved +Patch for Hadoop +We may need to revert some changes to make things work +If someone decides to pick this up for a later release it might be good to create a new Jira +It would be preferable if a development howto about custom formatters for storing UDT Rick! Im working on these +Added a new patch which addresses Marcels comment +So we need to add support for single field think Ill do the following create dedicated project at public repository and push the fix with the test case as well +But the internal implementation does not support them +attached is Franks patch with a testcase that fails with as he describes and passeswith the patchwill commit soon +The only addition needed would then be skinning support and maybe PPR +Ok! +Heres a pretty provisional patch for this functionality +Will commit it now +Also the attached is no more needed Ive found a more Spring style of integration +Part of Solaris of FFI stuff +Add CPU and available memory calculation in and the subclasses +class files and jars on the classpathThat is the behavior Im experiencing with a build from source as of yesterday although I still need to simpler test +Since does not provide a static scope it blows up hereThere are a couple ways we could fix it Special case the logic in to avoid attempting to get a static scope from a Modify to always return a dummy static scopeIll explore the latter since it seems a bit cleaner +OWB now provides the necessary parts +thought I had posted this +if nobody objects within the next couple of days i will apply michaels patch +Just my +Closing all ResolvedFixed issues lacking a Fixed Version +The fails to compile because the JDK generics declarations are different on the invokeAll and invokeAny methods +We have the same issue on the which have the schemas in its bundle +Logged In YES useridPlease add you comments to the bug to avoid duplicates. +Since this jira seems to be stalled Ill have a first cut +There were also other parts of the UI that I realise now that I wasnt seeing correctly that dashboard and the java properties views +I believe the recently patched class addresses and has fixed this issue +I can confirm that the problem is solved using AS and the jBPM version as of August ndRegardsMartin +I am happy to convert it to use conf location but this would make conf a template +Patch applied I realize my previous question really didnt make much sense +Tested with +Thanks for the detailed report Miles. +This is working for +Max please review my looks safe +Thanks Nigel! +patches welcome +After thinking about it for a while Im not convinced that we should throw a securityexception in that case +Switching all issues that have been in resolved state for more than one month without further comments to closed status +Lets create a main OSGI task and within it various that capture the things to do. +Use the suggested workaround. +Thanks for your help with this one Kathey +otherwise we loose too much time chasing usage problems +Here is an updated patch for X new test and typo fixed + +Our EAP Maven repository Jenkins job also generates a duplicates report which lists artifacts with more than one version +OK lets leave this issue focused on Lucene only since changing Solr is contentiousbq +Thanks for the issue +Is there another bugreport that I should watch to follow the status of the more general case +Im not sure how best to test the new functionality +This has not failed in ages +I need to survey to make sure callback context can change a Configuration instance that is used in all the important places wed want to change +Thanks for the patch Helen! + patch looks good +Updating affects have finished backporting +I think this issue is superceded by which suggests a better approach than using the bare User Admin service +Committed a change on trunk thought Im not too enthusiastic of the both on and trunk +Editorial and committed +The tests failed without the patch also +Ok I am seeing the same errors here +Thank you for answersI use a typesystem stored in a separate XML file and import it by name in annotator descriptors +Looks like jsommer also needs to be added to. +This functionality can be satisfied by. +Thanks Chinna! +I propose to close this as wont fix +So I agree to resolve this as proposed. +There are no complaints in Jira and mailing lists +Eugene thanks for your time looking at this +Err +There are few comments I would suggest we dont check only for existence of workDirs +Please close this issue if my fix based on your patch works Gaston +bq +rej fileThanks and good monday SaschaYour patch is commited at r +If we document this step well it should be no big problem for people +By each sub directories I actually mean each cache URIThe change will be substantially bigger than the attached one but that should solve the problem in a better way +full rebuild should be ether stays uninterrupted by new events or should clear all events from line except one requires full rebuild to avoid loosing events +Fitting into seleted page size +Move the to package +Hi Leonardoin this case its not clear to me why you have an API and IMPL bundle at all +A B and C form an ensemble and reach up to zxid xb +Please see javadoc This class is not thread safe +bq +Client and server to webdav functionality should work once the migration to Jackrabbits webdav API is complete +Please let me know if you prefer the old format or would prefer that helper method to be elsewhere +Fixed bug in detection of pluggable process tree really need this to get commited +We should document it somewhere as you wrote but Id also propose a field names validation at Solr startup using the rules so that Solr accepts only allowed field names and can guarantee the proper behaviour with all allowed field namesWhat do you think +Well if people would stop requesting documentation from me maybe I could work on it! jk +in +resolved issue on his end +Hi MarekI meant to say ER build which probably didnt exist at the momentBut didnt realised it was fixed in in this case Ill try to download the and checkThanks it on Dev and its still a no go +Another point this node doesnt serve as request coordinator until much LATER than the exceptions +Thanks Amila +We will need to at before releasing any other JPA artefacts which will be a good test of our semantic versioning +bq +Committed the changes +To go further with this patch would need to change our zking to be asyncThough giving up on this bulk assign will reuse the most of this patch in a new issue as it improves general bulk assign +Thanks Skye! +I have been through both SOA and SOA ER releases and can only find the class within one jar specific to the seam installationseamlibWill take a look at the current jboss cache release +Chunked encoding doesnt seem to be used +the patch looks good +If two threads are making requests one of which changes the protocol it can close the conduit used by the other thread prematurely +Whats the purpose of the resource field +Perhaps it would make sense to move the preferences that the refactoring plugin requires into the refactoring plugin itself +I havent considered its randomness +Not yet sure if this is a configuration error or a bug +Screenshot of artifacts that may be added to a package or +Feature is working +Although now we do data exchange through ports on the same machine I think we could optimize it by a disk copy later on +We cannot break the API as there are rd party type converter loaders that depend on the current APISo if you can find a way without breaking the API we may consider the patch +Reporters of these issues please verify the fixes and help close these dicussed with Sanjay on the IE still reproPlease help to confirm if it is supporting on IE or not +A hack we use to speed up running all the tests is to explicitly set running false line +A tool to induce fragmentation in a heap to trigger a full heap compaction +Im going to commit its a one line addendum +heres a cumulative patch so that jenkins has something to +Ill come up with a patch for fixing them all +I dont agree with you that the application should be tolerant to this situation +Reading the log it seems as if your server is scanning deployable folders in the middle of the folders being deployed which would indicate calls to your server via JMX to suspend deployment scanning are not working +Attaching because it was finished already before other comments and in case we want to go that way +could get us into hot water as Pete mentioned so we might want to just drop that one +Please resolve dependent resolved in +As retrieved on my machine and are identical +A Powered by logo +If we introduce additional offers for a slave considered revokable via a boolean flag existing schedulers will not be checking this flag and will schedule on the offer on the assumption that the offer semantics havent changed +Posted from Bugbox for updated cordova via npm as it was a newer version than what I had +Its my job and you helped to Seam +Since this patch is pretty self contained i am ok with committing and running cite to verify later +Patch contains a newly added file +Returns message DIRECTORYCOULDNOTBECREATED if a failure to create the message to reflect error for creation of directory +Committed to trunk with revision +Bulk close for +I do receive the notification of my own comment +If you provide a test case then its easier for us to fix +I attached a very simple web application to illustrate this problem it was created in RAD but should be very simple to convert to any eclipse based app +does not rely on a flat namespaceadded a test to check for is not a connection of added john as a friend of jane in the tests +it is resolved but I assume you just ended up saving changs to r the root project which for is the same project but for WAR and EAR projects it will be a different project +Class that could be used as an example for implementing the orion deployer. + +Added a dependency for bug as without the additions mentioned in bug one cannot successfully test this new behaviour +for the patch +Hi MattDid you get a chance to work on this +more over they can not implement the class +Tested +Marking Out of Date Issues closed +reason not to commit this patch +Updated fix versions for details see +Is it possible to not use Holder +Only other comment is you should generate patches in the directory not above as not everyone has trunk +Thanks +Here a patch that the assertion status of the actual class is used +Committed fix for failing test to and trunk +Thanks applied in r to trunk. +tests are now passing with the above changes +Yet another poorly designed wizard framework +Thanks for committing Nicholas! +Upon my own comment I shouldve left out certain nonfactual keywords and worded things a bit differentlyI agree that breakage at runtime is not ideal +In the case of a union this might not do the right thingAh yea +Ill be happy to commit the patches. +Throwable patch committed with I understand what you are doing with the renderer check but left the todo in +If you manually drop them and let them be then things should work +Delete war file and the exploded directory in webapps +Not being tied to Kerberos tickets allows us to accommodate other authentication mechanisms if needed +as they are still a major means of transport for integrations +I have this error too so i hope it will be resolved in the next version +All app transitions go through waiting for final state to be stored and then do corresponding final state transition in Similarly All atttempt transitions go through waiting for final attempt state to be stored and then go through Changed the recovery logic to not start terminated appsattempts and populate the final state +All in all this still made it a lot easier to switch to port which unlike was available on the Bamboo build server. +this patch adds tests to the above featuersalso changed the limits to group and counters in each group +Missed symbols in the path of Home Directory after examples +Fix for this issue +This works now +Tested successfully with for testing but please dont close it yet. +Strangely many people encountered this bug in + +Closing as duplicate +Documentation for the CONLL converter done +Lets solve first and after it lets close this JIRA. +If you think they are critical for CR please them +That way you can throw a simple email at it and get the text but the rest of the parts are available if you want themI hadnt noticed that youve added seems an elegant way of doing +we will change the order in +based technique Where the amount of time to wait before sending the progress is determined by Currently seems like a better technique +Submitting patch for HudsonNo tests were added +I have no idea which package needs to be moved from hive to hcatalog so Im closing this +Would be great if it could be published then the pom can also be updated +and each inherit from Address +Will look into that as time permits +Looking into metrics I havent found how String metrics should be represented +You should be using at least as described in +So Im closing this Jira +META +Manually verified that can now create a in another package and that the new param is seen +Todd is this similar towill the problem goes awayif we dont take the replica at the reboot nodein the recovery target +Wayne please try with the latest nightly SVN source and let me know how things work out +Wayne this is a very good argument +Fixed +Since binary fields in are backed by which are not serializable but Thrift structs are declared Serializable your suggestion would be that we provide a switch to generate buggy structs this does not sound right especially since the moment you will add a binary field to one of your struct your Java Serialization will break +Committed to trunk and. +The plugin is currently being voted on for release so I attach the fix as a patch to not disturb the process though I hope the release manager will consider its inclusion +thanks shravan +Verified that with no names table expected and no legacy lookup expected it outputs the right thing +so my view is that we work on versions would need to process the current reactor +Marking as closed. + +To addcould be disabled modes attributes names itemMode and attribute for mode just mode attributes for icons icon icon and so onTo add could be disabled icons iconLeft icon +Borrowed the cloning code from Xerces +Hmm I think this UML diagram here is very misleading anyway +Im unable to create any applications without having a domain name +Verified +Applied a modified version of the patch provided by Oleg GorobetsThanks Oleg +We could find out the buffer size after examining the size distribution +Attached log for both tomcat and tomcat. +Ive opened for the defaulted fields issue. +I have tried the current cvs HEAD sorry svn trunk of the this plugin back and forth now with and without the new property and it works as expected +Andrzejare you still working with this or should I proceed as I originally planned +Improved the error handling with rNow if the file is existing but cannot be opened for some reason Wicket will just log the problems and continue without saving the pages bytesIn your case it seems Windows Explorer or any editor has locked the file +Looks good. +Fernando AOP and MC are ready to be updated when youre ready for them +Ive done that +have you looked at the existing test cases for native sql query usageIsnt there a test for this exact thing that actually passes? +Of course the really interesting way to do it would be to have a project to analyse apache log files use the site as a skin with Google analytics tracking code in it +? Patch +In general has a lot more native JAXB support since comes with the JDK now +As headius suggested in IRC we could also look into MVM on the C level and maybe see if C extensions might declare so we can decide whether to raise an on a basis +and thanks to both of you +PiI agree with +I went with sessionList synchronization for the sake of minimal +Patch looks good +done +Please reopen if anybody has further concerns. +Committed for version +now throws an exception if it hasnt been started tell the end user that it must be started before it can be all resolved tickets from or older +Hi PaulexThe fix looks good +Im inclined to move this issue to github as they seem to be better at than JIRA +When creating a Seam Web Project in JBDS GA the is not included as a reference can attempt to manually bring this in but my experience was that the gets wiped out in the process +The debug documents list needs to be synchronized +I have committed this +the wrapper for all things related inThe code that parses the was a bit out of sync between the two after the variable substitution patch went in and so it was causing some differences in what each thought was in the config fileThis fixes it by having call s static parse method to do its parsing +Requires some reflection stuff to work nicely +Read the installation guide for more details +Thanks a lot for filing this issue RomanIn the future please only set the fix version field once the patch has been committed +Patch to handler to check if rampart is engaged and immediately return if not +When you say client identity do you mean client id? Strangely the spec seems to omit any detailed description of this check but I can see why it makes sense +compactions were discussed there to some extendIn general I like the idea and for some use cases guaranteeing an upper bound on the number of store files touched on the read side will definitely help +Thanks Jeremy rev . +Have you been able to look closer at solution optionsThanksKathey +It also has a LOG so that the log level can be changed +This is a side effect of +Committed in revision +rename to +an be initiated creation a new build in jenkins or do I have to create the build by himselfI can not find a newer version for RHEL +output etc? It doesnt seem like a very flexible way of making portable tests which can be executed on different platforms with potentially different temp +Im not sure if this is the driver or the server itself +I did encounter some other issues with your sample app but it should not be related to the original issue. +In trunk look at the usage of the map and reduce variables +Attached is a patch and generated html file +David thank you for taking the time to put together D support there was a lot of work that went behind this and it showed +Weve always talked about the map daos really being for testing and I was having issues with some new methods in the interface +Closing this as no longer reproducibleIf you feel this is still an issue please retry with the latest codebases and if it still fails reopen this Jira +This issue is still present at ER where PBR WSRP are used +This will need to be refined later if we do want to offer remote continuous or just want to allow use false such that results buffering is needed +are something that could use the improvements that you are talking about +Breadcrumbs are added to the installing and configuration Common client side articles and changes to FAQ. +One important decision we need to make about this feature is whether we support simultaneous schema changes for more than one tableIf it is supported master needs to maintain a set of region servers per table at the time schema change for that table starts + +Added collation sort test added collation key debug printing +If you any issues with the front end code you can always ask me about it + +I fixed the duplicate element issue under r in the branch +applied thanks +Is this open source? Whats the license? Can we include it in the default windows install +See +Retrive hibernate Query object +actually nothing from jboss tools is loading +update test output file +is not application it is and Java doesnt know anything about Cygwin so behaves exactly like it would behave on pure Windows environment without Cygwin +Can the design be moved to the Kafka wiki instead of a pdf attached here? It will make it much easier to discuss missing details in the current design +Hi RyanLooks like this problem is being caused by the caching service of Equinox Aspects +Ive also moved the descriptor bits completely to core and removed its dependency on plexus +BTW do we have any tests covering the type conversions all those corner cases etc? Would be nice to get a regression test here +patch conflicted with so here is the proposed updated patch +I might miss the point here but why dont you maintain list of addons in git directly? so that all the people beeing able to commit to git could change that? Moreover as you guys can change the code you could change the location to fetch the addons I guess +s correct sorry +s ran into similar problem wherethe local username and svn username are not the same +The in HDFS in addition to its primary role serving client requests can alternatively execute either of two other roles either a or a +Theres a bug in the current version of Tortoise SVN create patch +resetting P open bugs to P pending further provide a minimal input FO file that corresponds to the attached PCL output file also provide a PDF output file that demonstrates lack of has been added with description PCL FO for Space there a way to remove me from this bug and give ownership to find the attached minimal FO and corresponding PCL and PDF documents +Improvements by recommendation of more thing +Investigating the reasons of my unit tests failures Ive found out that when is delivered source component doesnt have parent component yet set +gotchya +I think it is safe to downgrade the priority as minor. +User error was building with commandlineonly +I tried SIGQUIT on several main java processes but still did not see any thread dump in the or does it go to the console logThanks In YES useridHi ChrisCan you attachtestsuiteoutputlog the hanging In YES useridThats a bit test does nothing unless you haveexternal contexts do you get inoutputreportsm beginning to suspect a bug in antjunitespecially since you are seeing a hang rather than output do you get from pstree can you forking javaDo you have any jars in jrelibextRegards In YES useridI noticed a typo for haltonfailure +Thanks John +unit test for this issue was added to isuue patch applied together with s at revision r thanks a lot for this enhancement please verify the problem is fully resolved as you expected +The Connection class ensures that all exception thrown from a failed send are s so a client that is using the only should add a try catch around calls to catch the error and shutdown or manually reconnectAs for the Failover transport issues those have I hope all been resolved by other recent issue resolutions +Otherwise Writersdont use encodings. +Thanks +this is solved in +Yes it seems that the latest JVM has fixed the issue +It would be awesome if I can replace the sideline fork Im currently use now with the mainline thrift compiler and library +Here it is +PeterI think the should be a transaction of its own but each service should be run in its own transaction using its own time out which should be defined in the services xmlWhy dont you take a look at this patch? Peter I noticed that when you uploaded your file you did not click on Grant license +There are two issues +Of particular interest are those which have been resolved whether the cause is known but not yet fixed or the cause is known AND has been fixedFor issues which are resolved please provide appropriate Release Notes text following the format of providing the Cause Consequence Fix and Result +commit to feature is also required for G and trunk +Removed from planed release because will be solved by FE update to later after available from rolled out some time ago problem is patched here +New patch against a relatively recent branchx +A lot of work needs to be done to fix the current state of +I was just about to give Archiva a shot after working with a modified for two years now +Attached the link file we used for installing plugin +I am uploading patches for and +Working on a fix now +Info level log messages will occur if a local db shuts down during replication but there wont be any nasty backtraces about closing a db thats not open +Will discuss further there +Could you rephrase +In your change you close the no matter what +pushed +I saw you named the new interface as well +this was part of so I didnt wait for hudson. +Ive applied Mikhails patches at revisions and +It seems you are not a watcher for the ticketSee. +Hi GuysWe are in the process of setting this up now but dont have port access we will use SSH port forwarding for the time being until the SSL proxy is to bloodhoundssh l mpoole L +I ran unit tests with the change I recommend in the description +The getting started wiki page does not cover the details of the implementation that is still required +One of the possible solutions is to add JAVAOPTSJAVAOPTS tcp into JBOSSHOMEbinThis works on all platforms and makes all servers run in the TCP mode +Resolving this issue as the patch is committed by DIGY +Please help me verify the fix +Patch in r +Thanks Jon I put this into trunk and +The patch is surfacing issues that were already there +This ensures that the producer is not shutdown before the mirroring threads have finished their work thereby avoiding data loss caused due to +Added createCategory deleteCategory and list all categoriessee in for more information +Ive seen many project splits implemented in ways that are difficult to work with maven +is in folder pointed by AXISCPPDEPLOY +Can you remove it locally from the jar and see if that fixes it +There have been no new ones added for +The NTLM concern feedback from Apache Legal seemed adequate to me perhaps because I was not expecting a definitive ruling though. +using customized cas client still present today +Thanks Jason It works as expected. +Without caching ear i get the same result without caching +could you please explain like Im +Not sure if I understood what you are asking for? +the test case assumes that localhost is reachable with patch should fix the issue +This is used to resolve class names and eventually enqueue new sources + +David you can check the integration test test +ICLA sent +I dont think we should I dont understand the use case behind these static methodsIs it really faster? Do you have any benchmarks showing the gain +We will need a client wrapper like so that the controller can use to send the commands through RPC +This does not mean that we shouldnt look at a better way of tracking what happens with the To and because those are visible to everyone that is receiving the email +This patch file now replaces the conditional test with the more robust comparison methods +Thanks for those +no this is only relate to the wrong test case +Please provide the full stack traceincluding the wrapped exception +maybe instead of rediret to buffer we can handle this with redirect to render. +Because introduced fixes into leader election that added stuff to the ZabTest that I missed +Patch to remove it +This is kind of implemented with the latest changes to the httpd module and the ongoing work with MTOM parser +has but JAMES can not headers if they contain no information about encodingDatabase and IMAP show question marks POP hangs on this headers +So that i could avaoid two times get invokation +We should probably do that in a separate jira as it isnt directlyrelated to this oneBTW it seems bytesSent is not updated in seq +Maybe this causes the issue +pushed +It also seems useful to add this capability to all the applicable methodsrows execute firstRow eachRow query executeInsert executeUpdatebut probably leave it off could perhaps have a look at this issue again if the patch still applies +No +Patch looks good to me +thats still not quite right +Patch to support values to be case committed +Im pretty sure it works by just looking at it because I used very similar code as a workaround. +Without knowing the implementation you couldnt know whether a component isthread safe or not +Thanks Benjamin and Stefan for contributing this patch +Ok I see your comments about the test +And I dont think should be there as well as its part of installing Camel in Apache Karaf +Keith I dont think that works +We figured once we got the local map reduce mode working we could remove the extra code as we have time. +Thanks Sanjay +my bad +Package support is not quite complete +That being said even if you dont allow to mix regular and counter you still have the fact that BEGIN BATCH will be atomic for regular but not for counters + +Thanks for cleaning up my dangling importsI wouldnt mind if you felt like being ambitious and cleaned up that TODO in by either pushing code down into or subclassing it for more specific behavior +We should definitely be setting the and not the an interesting note +Now that work is well underway with AS all previous community releases are +Then the patch is good +I have it fixed and will commit this evening. +I think the current unit tests will test that this code refactoring does not break existing functionality +Setting to for now if I can reproduce this and create a fix quickly itll be included otherwise it will get bumped +How can we sort vertices before read vertex +Thanks for confirming that mvn up fixed your issue +I have trouble reproducing this +Exactly PaulIll be changing the type field to be easier to handle with for example for view compaction it is the string View Group Compaction something like viewcompaction would be easier for applications +I dont see any value in wrapping the exception here and in an interceptor it could possibly confuse the callers +i have fixed this issuecan i submit a patch +jsp demonstrating s default page to include link to to demonstrate the s Ricks Brocks code essentially unchanged +Doesnt follow the URL to the tracker +This seems like it would cause an issue regardless of what length is being used +It seems like were talking about different xml descriptors +Closing all tickets +Supporting TCP only is sufficient IMHO +Ive uploaded a new SNAPSHOT so you can easily test this out +I think that this work is the most written yarn client code that I have seen so far +Thus will hold on until a solid fix is found +Removing a barrier to someone getting started is always a good thingIve added a note on the How To Contribute page under setup about this FWIW +I went into so much detail because others might run into something similar this may help them fix it. +Gary has fixed this as part of +Tech previews are things that are expected to be supported eventually +bq +I do have the same behaviorWhat seems to happen is that MINA is repetitively trying to send the BAD DN message ad nauseam +Looks good to me. +If it is simple then I think we could move it to +Your patches are most welcome +Somewhat related is which will likely be revisited in as DI container gives us more flexibility to implement it +I agree with your analysis +Issue should be closed +in the example deadlock you have above possibly there is a bug in what information is being displayed maybe fixing the bug would be a better approach than drowning the log with too much information +A couple of questionscomments on this patch +Please reopen if the updated examples do not match your needs thanks! +Both should have stabilzed after so is a more realistic as out of date because of month no activity +Yep thats a fair point I will this as an enhancement in our server implementation. +I committed this +Hi Steven the vdb was created by teiid designer +Since youre working on this can you drive the related discussion on the forum around spawning new listeners +first shot at a committable patch +In my experience finalName is set to something shorter than artifactId +thats a great idea +Anyways as specified in pom of the example attached the plugins version was already enforced with +Im saying that even before you start talking about drift were correlating millisnanos and saying This Is The Reference Point For Zero Micros but we could actually be micros into the current ms we could be we have no way to tell +Looks like a good fix +pending addresses comments +The maven management with possible exclusion should be great +I can do it +Thanks Bikas! And Arpit for reporting and verifying the fix +x +Ive been using for term expansion +Ive been thinking about this more and I dont think implementing secondary indexes is worth it distributed or otherwise +So I would recommend introducing a test that kills a queued job and makes sure it is not there in the list +Added comments from Milos Kleint for suppose we sometimes move over to maven so my suggestion is toa +Song title wasStill would stand all time +On the other hand Windows wants to show international charcters as one byte characters +hi fergus if the document is empty it is not tried to be added +from has some code that can be used for this. +The comment spam Ive received recently had tons of links in the body of the comment +Will work on getting to your fork somehow to send pull requests +For one thing I think strongly implies that we are connecting to a bean in the service layer as opposed to a gateway +This might just have to go into a branch. +The tag itself is all thats required in this need to change the default behaviour just make the push optional +Thanks a lot. +See if this works on your mac nowThanks as this appears to have been completed. +Thanks for patch Riley. +Thx. +Hello Sebbyou were too fast and added a comment before I could fill in the description +Taking a break +Judging by comments here Im assuming that that is expected so Ive marked the two tests as Ignore +Wish it offer you some value +Not seeing any demand for this so wontfixing for now. +this should be done now but i want to sanity check all the packaging before resolving +Not a bug in Derby +bq +Camel should auto look in spring registry for any beans of type and use this instanace instead of default oneLike we do with Tracer etc +In this case I would prefer separate methods over if statements in code +But as I understand replies always come back in the oder DN DN and DN if the pipeline is ClientDNDNDNThe patch mostly looks good +Project property page is added in rev +This is a pdf of internal documentation and contains query dml and ddl examples as well as the overview of the system +Bryan thanks for your checking. +test is added for the issue +I think the branch would also benefit from these changes +If you prefer your implementation go ahead and check it in +This appears to be running in a version of Tomcat embedded in +The static type part would be a Groovy thing while could be fixed in a minor release as bug +Plus we need to add unit tests +im saying it was fine the way it was +The only way to get a quick fix is to propose a patch with a unit test yourself this normally gets the patch applied within a few weeks otherwise youll have to wait for a volounteer to have spare time and decide to fix this particular issue among the others contained in this ticket tracker +These blockers and critical issues are resolved but not verified +The reason why you get the deprecation warning is because of you IDE linking to the current version of instead of the depending version of the +pushing to future XML Message Structure metamodel has been removed from Teiid Designer as a creatable model +Deploying this service you get the WSDL malformed with the import missing +Hi Ravi it will be great to see these improvements in +This one was too quick I think +Bryan Can you provide a test patch which shows the issue of looking up the keys +I am guessing all the tests should be easily modifiable to deal with the new code? Or would those need to change substantially as well +Theres no need to calculate exact size +Bulk closing stale resolved issues +Jonathan Do you want to give a shot along with this? I was bold enough to pick it up but cannot find much time to do it as of now +Odd +I am for committing this to trunk +Can this be the problem? We thought that network state was enough +We had discussions on the mailing list a few months back about support on the ORM side +The before does publish such an event + +Please review the patch +Thanks. +I dont think is Oozie responsibility to proxy logs of systems where actions run +Lin this is fixed now +You should create a transcoder for the encoding of the input string and usetranscode only for data coming from stdin or being printed to per my understanding database manager uses the following items in the order presented to determine the active code page LCALL LCCTYPE LANGIs there anyway to determine which code page is in use right now on Linux system +I did not add the strong userid password option there because I wasnt sure if this was going to be documented +I also updated our poms to reflect bnd as a build dependency +Troy Poppe is working on a full provider for Maven SCM as an alternative to using the Ant we create a place in SVN for this to reside? possibly in sandbox? +bulk move from M to only do this in trunk not in maintanence. +I have deployed a new that includes this new +Checked in first phase which is adding a method to the database abstraction and support code specifically for +David in chat you mentioned SPI files but I have no idea what you mean +hmmm +released with +Next monday is fine +Query plans for second query +Screenshot of negative messages reproducible in after purging the message queue STOMP clients +Correction the thing I added does not apply here but tackles a similar issue in other parts of the code +Committed revision +org AS vs the EAP +Im pretty new to GWT and didnt realize the shared convention used here + +The one real issue I can see with this is that the spec identifies that a datasource can be deployed as and transactional +Do you mean I have to take the latest build of the server +Moving these issues to as I dont think well get them for +Fixed with SVN revision The revision to the netuitextArea tag modifies the logic so that we only force text content to start on the line following the close of the rendered begin textarea tag if the content begins with a new line +I ran it first on the which shouldnt be affected by the patch to and it showed about difference +Paul we currently do in the test to avoid collisions in case other tests fail or if a negative test happens to use the same table name since theres currently no way for a negative test to clean up after itself + thanks. +The err prefixes are misleading. +Iwein Im assigning this to you now since you are most familiar with the issue and the refactored aggregation code +langjavascript support is possible if you apply the attached patch file to the and files +for removing old LDA as I dont know how much it was used anyway due to the scalability issues +Will try to update on Review Board as well +I have attached a patch there +Can you try again with the latest release beta and let us know if the issue is still there +ll get this done I promise +Sorry about that Daniel +Thanks Tom +All uncompleted issues are now moved to the new To be determined for future releases category +Actually updated to fit on top of revision of branch lucenesolr but I believe this is the same as Solr. +Apparently it doesnt run these tests +Ok the unit tests have been passing on OSX and Linux for about a week now +I cant reproduce this problem with my projects and I cant open attached tgz fileI have noticed that all Violation in sample file are not activated by default +patch for changes that needs to be applied to code for required changes in jbpm code shots just to illustrate how does it look like in the +The logs for the linked builds have been auto removed by jenkins can you attach a copy of the logs from a build when this occurs or change the configuration for the build when this happens to not remove previous builds +First the person requesting this was either a machine or was overly ambitious in trying to get an exact value +Happy review +In addition I think the error message should make things a little bit clearer +Also I think maybe the right thing is for these classes to actually go away moving forwardHow about I submit this fix to provide the user some relief and then file another for the general encoding cleanup +Seems running the tests sequentially has solved the error +Patches for and attachedIve tested by building the assembly plugin with Does anything else use? Any reason not to just remove the local metadata generation? +Add Whole Column Family Iterator patch +Can we do the type system search strategy extension point in the CAS Editor thing as soon as possible? I was selecting type systems today for at least half an hour a bit annoying +We also fixed the call duration calculation +This JIRA is for release +If you installed into using then everything you need to run exists under the servers modules directory structure +For a maven project this would be srcmainjava +The test case just reads from thisservlet using a POST +Use styles and attributes of fileuploads parent tags +Attachment has been added with description List of updatednew has been added with description Patch has been added with description List of updatednew has been added with description Patch has been added with description PatchJust to keep you up to date Im almost done here +When there is no match for the last optional rule element the target span for the new annotation is not correctly calculated +file will be deleted when the snapshot loaded during startup and at that time currentEpoch will also be updated before deletion +I tested plugin locally today and realized its documentation is rather outdated +Handle anchors in links and check for various link in EXTERNALLINKURL type to distinguish the first token of an external link from the other tokensAlso adds in a POM template so this project can be mavenizedAre we there yet +Commited on Branch will wait to check it works properly before commiting to other brances +Therefore this can only happen with an unscheduled outage of the storage serverWe can add a few things to make this happen less often +this is ready to commit +I will commit it +Hm that mysterious piece of code goes back to the first release of Derby so there are no clues available in our code archaeology +You should have let us know about it +Sure I can add the same but can someone confirm this is the right way to do things? I just tried a workaround. +Just checked I cant find any more issues with this work +in all current user repositoriesWe could have optional repoId tooWDYT +Thanks Mithun +Is it for the string argument? Couldnt we just make it stdstring and avoid the use statement? Just seems like we dont need it since the rest of the code explicitly uses the namespaceI agree fixture data should be regenerated but isnt currently +This is available as an option for now too +bulk move of all unresolved issues from to +Look at the docsexampleconfigs configuration file +My productivity is down the was able to fix by running update source folders on all related projects +bq +Based on the discussion at the contributors meetup a few months back didnt a lot of us anticipate the use case of a snapshottable user and then allowing each user or application to then make snapshots of their own userfoo? Or a snapshottable root directory and then snapshottable subdirs +The methods in can be used in other classes too +JeffI was wrong in my initial assumption because my test case was flawed +Im working on a patch +Consider that +Thanks RegisPatch applied to SQL module at repo revision rPlease check it was applied as you expected +this library is from the same company as an explanation for the groupId was provided there +bulk defer of open issues to +The owner is persisted along with two rate objects +In other words if there are no bindingtemplates this array needs to be null +Patch applied at +I dont see the necessity its just that implementing equalshashCode is never a bad idea +For the test cases I used the same data as in +The attached are the implementation of and unit tests +Committed the deserialization itself looks good to me +Changed cron expression so job runs every minutes rather than seconds +All suggested corrections are done +Fixed in revision +Should we close this issue nowMubarak even if we have some option to limit the value size we also limit the number of concurrent accesses +But we should still consider the fact that will be coming with and weigh the benefits of fixing this issue against fixing other issues that dont touch queriescolumn iteration +The two new lines are now in the new patch attached +Patch submitted based on trunk revision +for the patch +you can revert qlsrcjavaorgapachehadoophiveql +Could you set the following property in your and post the diagnostics that it prints outprinttrueThanks +Please resubmitWhat does your patch do exactly? There are already numerous result tags +Could you also update the How to Release wiki page +I did effectively try it but I could not compile after it +However the new block entity reset the second location list +Bulk close of resolved issues of +And therefore I cant imagine what against mvn repositorybundle should be executed +This should now be resolved on the branch as well +I think this is the right directionThe AM should request for disk in addition to memory viaIn terms of timelines I think its a little early to right now +Please read the comment in the source and in this bug reportthere are two classes andThe first one is deprecated and only there for compatibility andthe second one is the new implementation extending the excaliburstoreAfter you read this please close this bug again +When you log a bug like this you need to include either the actual file or a test file which demonstrates the problem +So +In case a result is selected the project tree is scrolled to the selected result +Can I assume that its under controlThanks check the forum +Are we running two TTs on each node that talk to different JTs and migrating slots from one to the otherYes +Just as an example this is how we implemented a feature similar to Elis suggested part on our internal branch +I have not thought about rate limiting at this point +Reopening to commit changes that will allow SOA builds to work in addition to regular AS builds +Well need to work with web services team to get on the roadmap alligned with the jbossAS schedule +It is safer to make independent from existing environment to avoid recursively append of environment variables +Max yes +Cleaning up this would likely delay the release for quite a while so lets defer it till +Some of the units are not passing for me on the branch and therefore Im currently unable to run this against all the tests +Please close this as Will not fix since this is not supported in the EAP version +I think that your feedback has helped improve this tool greatly +All set +On the other hand users do sometimes contain target +Logged In YES useridAccessing JMX over a remote channel as twiddle requireswill only work with types that are serializable +resolving because this was opened for tomcat +Limit the amount of memory used in the call queue +Thanks stackOk +I have committed this +Im not familiarized with Maven yet but definitely im going to learn how it works as soon as possibleThanks Ivan no probs at all thanks for attaching the Context setters to the dedicated interface fixes the SergeyIve noticed that if other class extends from the is not injected! However both and are correctly injected +minor nit probably a typo the version is bumped by +It looks like youre trying to build from the binary distribution not the source distribution +This is an annotation change and doesnt need unit testThe test failures are not introduced by this patch +I suspect it is a issue of configuration +but it doesnt make sense to duplicate it in and its confusing when debgging. + +Screenshot looks goodAbout version numbers I guess all will ever use same if not other variations could be addedThanks +Not surecomputational formula is optimal though +file hdfstmp length file hdfstmp length Since we are printing all splits can we avoid the duplication of in each line and combine them into one statement to save log space +Franki imported the latest snapshot from PROD into +Deleting records eliminates your ability to maintain a paper trail +I will review and commit this patch +Log attached +assembly depends on all other modules +Many of the other localizations have drifted with the English versions and in my opinion those skews are more significant than the missing Polish variants for sysinfo +Will check if the changes do not influence XML test suite under Windows +Hi Mikethe last Hudson build failed +removes unused in r. +Yes this patch is for only +think the purpose of this JIRA is to generate WTP API doc +I dont really know what to say regarding the random other failures are really testcase bugs the tests are implicitly assuming a given execution order while that clearly changes on different platforms +this issue fix will be back was back in the EAP and release notes. +This issue is resolved in +But I was not able to test this +The auto test cant guarantee that we have this. +Updated version of the patch +Moving to release date +Also there seems to be extra spacetab after the column metadata is printedThanks for fixing dailed to failed +Pom files were not in the orginal patch +Specified the scope of the upgrade +commit it nowEsp for a bug fix unless one thinks there is something likely controversial about it or one is unsure about something and is thus requesting feedback +Ill tamper with the stylesheet to make it prettier if youd like +Now we supports variable number of arguments in the hash function +Attach my dirty patch before leave officeNow Im satisfied the performanceIf anyone can optimize or refactor this patch Please feel free to upload your patch +Is there any particular reason for the finer granularity for system privileges +I pushed up to mavosphere +The tool does not allow this +Should this be closed? Theres much better error handling in place now +Its back +It did not even require any refreshing +Ill comment on both of them before Monday +Looking at the html there are still some newlines coming through without BR tags between themFYI PDF parser has the same behavior +See the for a scenario that places a select node under the sourceWill look at refining to not need a new node type however the initial fix is the most localizedappropriate to be +Sorry I overseen the license option on the first upload +Currently Run as +an extremely large interval is recorded which is the time the node was left dead the first timeWhat if we reset the intervals when we get a node +I have no strong position on this as long as the newer versions dont break things thats why I created this issue as a minor wish +Checking the order value from the object is a workaround only +We dont afaik +Done +Looks good to me +This is a dup of right +Thats not to say we couldnt rewrite the Contacts code to use Ericas approach of copying from an into our own type of Person class +Pig has no retries beyond what hadoop provides +Bad IP address unplug the cable disable network services etc +yes right now testing on the latest version planning to patch them +In case of DN failure you need to wait until the rbw block is recovered to see its data + +Yes there is no need to collect push the scripts during node creation +Can we push this into config? Should we +It will always log the number of matches not the number of groupsAre you logging it in your latest patchIf so we can close this +Im using Sun JDK on Windows This is actually a bug in the Schema parser built into the JDK +You might find MINAs build useful it automatically copies related license files into the artifact +Ill close the bug +For extending the files a script is written +seems to be a really bad name +Aaron could you please describe the solution along with the patchSure +On return from the send method the messages header field contains this value +Agreed I will check and update the Jenkins builds to add itAlso itd be good if someone could try this patch in a high load scenario to see if it brings the desired performance improvements +Closing +hard coded text in jsp files are extracted to external resource bundles and translation are added +In the new model Ive made it an optional argument in +The attached patch removes the section on isolation level from the Differences between running Derby in embedded mode and using the network server section since those differences are no longer valid +I totally agree +Could you compare this code to as well? Both are about +I also amended the javadoc for UTF to indicate that it encodes patch includes the test sequence Colin provided above +I split up as follows will have the same interface as today and will know which strategy to use and how to instantiate the interface that describes what information a strategy has to provide to it is more or less the same as cdr had been before but now cdr has another which implements and is the implementation of todays Castor strategy to resolve class descriptors +I understand it doesnt implement JPA but it is such a popular API that it should be the one exception to the the Roo JPA rule +Peters patch with the divided interface +Thanks Gilad added a comment and committed +I confirmed that the test fails without the fix and passes with the fix +Hi DougNoticed that neither this one nor is not marked as blocker for +This patch applies Dags suggestion putting the new trademark language in the footer of all Derby website pagesTouches the following filesX buildsiteM srcdocumentationM srcdocumentationcontentxdocsPerforming status on external item at buildsiteM buildsiteM buildsiteM buildsiteM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteintegrateM buildsiteM buildsiteM buildsiteM buildsiteM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsitepapersM buildsiteM buildsiteM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsitemanualsM buildsiteM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsitereleasesM buildsiteM buildsiteM buildsiteM buildsitedev +It may be a strange setting that I have with my products +I believe Boolean is the only one that needs special handling other than all the standard subclasses of Comparable +The situation is not different here +N Probably should post the patch up on reviewboard +JakeI should say that aggregateRows and aggregateColumns would probably be very easy to get working on the DRM +This patch fixes the above problemsinlines the example source codevailidates correctlyAlso this patch catches in its net a couple of instances of converting computerout sdoc tags to programlisting goes herezookeepertrunksrcdocssrcdocumentationresourcesimages are some link issues that are causing the pdf generation to fail +the index is translated into the model which is in turn translated into the PK +The notion that conflicts should not be fixed at all is confusing to me the distinction between we isnt clear this is talking about a simpler way for users to resolve their conflicts how they choose +New archetype stuff is much more flexible so we can have a srcmaingroovy dir +Do you have a reproducer for this +The class found there has a cancel feature. +We should probably standardize the config knob names to We cant remove the public members from and +Seems like adding the offset and length is a good enhancement in itselfThere is another unrelated enhancement which is to take directly in case of NIO clients that read it from somewhere and want to log it to Bookkeeper without copying +fails but far less from the mark eg when it does and actually passes some of the time +patch for flex also is undeprecated as it was accidentally deprecated in favor of itself +Attached is and +I also Like your Idea We can also do thatRegards +Attaching patch which modifies and as previously describedHenry Zongaro has gratiously agreed to review my patch +Has this issue been fixed in? Because I still seem to have this problem that the directory does not show up on my project source directories +I will make a test based on that +I also agree with Mahadev that we should make it backwards compatible +Hmm Im going to have to think about this a bit more +Its in the first patch but it messes up how scripting engines and other frameworks accesses the Map +Removed this outdated test +bq +Russel Yes Groovy is in the PATH +As advertised +Any news on this issue? I run into this issue when trying to use the Sonar plugin for certain projects +Hi the last two builds do not have a jar +The structure of the plans for the deployers and is much more flexible here +didnt get into either but now I am even more certain that the way to do this is with plugins +trying to get Hudson to pick up the third patch +Sorry for the delay on thisWe have worked to get some changes pushed upstream but I have not reviewed them in a while +in buildprevents build error messages when just deploying the producer sidepeople who want to use the proxyportlet in the pluto environment need toadd the include file in order to allow the proxyportlet to be deployed. +If the JSP source is stored in but there is no other indication of the contrary the JSP compiler will read the file with and thus output wrong charactersThe only and correct solution to your problem is to explicitly state that the JSP contains characters +This test reproduces the problem +bpel or +I have one doubt for analyze configuration +when xalan is tested with heavy load +Jochen attaching patch V +The idea of this issues was almost same as that in +I still dont know what the root cause is but it happens pretty reliably on a cluster +And give it seems even more useful +Essentially we have to parse the results grouping by expression and then picking out each range in the order we want to illustrate it +Heres my environment java version maven macosx snow leopard +Ive fixed the basic issues +Thank you Chuan. +patch looks good +Really the copyField thing in todays example schema is an anti pattern since we teach people to duplicate all their content while most people would be better off using +Corrected file a bug has been PEPCAKd +Attaching patch aThis patch shouldnt cause any behavioral changesRunning regressions +Looks good +Given the error message the dependency needs to be in moduleThanks +Either the configuration printer implements the interface or has a method with the same signatureAdded in revision after release +Can do that elsewhereRemoves because has little left in it for test makingReview please +Please review it +Perhaps a which implements Cloneable with a default implementation that does a shallow clone? If its not a SPO then I think we just ignore itThis doesnt help people who setAttribute directly on the session does anyone care +Thanks for reporting and analyzing the problem +Therefore whenever BOTHIDLE is fired READERIDLE is fired immediately causing an exception +Test this functionality by adding a Jetty container target to the integration module test suite +Anything can be put on the right hand side of the and match exactly as it would previously +the new schema would superset of the config of the old schemaHowever I confess to not realising that itself currenlty has the same limitation so it does seem unfair to raise this JIRA against CAMEL CXF without also addressing it in CXF coreAs an aside when folks are developing spring based schemas for CXF or other products they probably want to take into account that folks will want to support refactoring of config and declare the types concerned at a level that supports itGlancing through some CXF schemas this issue seems pervasive and as a result a lot of the flexability that spring config can give is lost which will make administration more difficultAnyway I am happy to leave this for now +Rolling over to reduce workload for pre orphan control yet new ORM docs much etter. +If you finish it before we release beta please correct the versionThanks +There is no duplicate funstion scripts anymore +Further IF someone wants to add Maven build then Maven allows for the directory structure to be different from the default so even then a src could be used. + +Closing all resolved tickets from or older +Replaced by LMAX Disruptor based solution from +Could that be the case here +We can use two approches for thisOne approach is to provide the needed getStreams and other neccessary methodes to theOther solution is to introduce and make it a place holder for attachments with neccessary attachment handling API +Applied the most recent patch. +Create new Mobile Hybrid Project +Tested against enAU enUS deDE +yes lets do this now since its trivial +The issue is in the shell install command +Assigining to you since the two remaining tickets are assigned to you. +Because swapping beans only partially might be hard to do right +The thread is started when the first object is attached to the instance +Its dependent on which is dependent on for native IO on Windows +Also the proof of concept patch for included a toy mapreduce framework +Yes its worksThanks. +Yes rebooting the router does work +The problem persists +Alex the test code you checked in to the svn repository has a problem that it stops sending messages when more than one messages are contained in one packet + +It causes regression on IBM Sx running zLinux + +And then it would also be nice todo away with password since were already breaking back compat +Thank you now it appeared working as expectedClosing this issue. +Whoops left an unused variable in the previous patch +If regions were unreasonably being offlined over and over its possible that youd get because an expected region would never be online +Sources of the most simpe test application and screenshot are added +the tags arent populating the pageproperly +Right now if you make the start method sync you end up with more issues coming from the discovery bits +A sign that its getting too late +So why argue so passionately about not fixing it when its a simple you need a good reason to fix it then It is bizarre that there is no predictable order for the classpath even alphabetical would be better than +But with installation using GUI mode and skipped cluster properties field MSI should check if HDP is available and show warning message if not +I reviewed you test case and noticed that your consumer uses the same connection as the producer +Fix the inputoutput for create table drop table create table like create table as create view drop view +I have created the Castor OQL BNF for different CocoR with corresponding Junit test for their performace testing +patch looks good +The code supports tcp http as well as failover in the broker url +is a version number really what you want here? Shouldnt it be done more like modern protocols such as protobufs to introduce a mechanism of optional fields? Strict versioning of protocols is very unpopular any more because of the brittleness introduced into protocols +It is starting to work even if the files are not there +Fix applied at r. +The property placeholder is implemented properly except the update relatively small functionality is quite hard to achieve since it requires changing existing bean definitions and singleton instances in a fast and thread safe manner +It seems it is a problem of my own testing need to it on a RHEL x from RHTS and then I will close this issue with Cannot reproduce bug +We execute this call to the from a Bean in a JSP +There are dozenz of examples in the docs already I dont know how this one would help someone who doesnt understand that an interface is a class is an entity type in the context of Hibernate +Ill have to get a workspace from him as well +Can you please upload a new patch +This patch is another version of conversion inspired byUpdated +just moving the assert inside the if that actually uses the cached root arcs +Closing all tickets +I just noticed that Solr depends upon some methods in +Marking this as an issue to backport to to trunk +Updated to last trunk changes +Please confirm +bulk move of all unresolved issues from to +He wants to edit a document so he just has edit permission on it +I agree that this needs to be fixed only keep the identity tests intact +fixed integrated in branches with rev +With your patch it is located at the center of in the leftmost position +You are correct about those flags basically being required so perhaps were already halfway there +Same trunk patch just uploaded with grant license toggle flipped +Please post your additional comments to the new issue. +I will redownload the early access and see if it installs and if this problem persists +After further discussions were likely to provide this via since it enables the possibility of defining these via XML too at least for external modules such as nd level cache provider +I both the +Then when I start zk on that node everything works as expected +Jeff I take it auto bootstrap is off? When autobootstrap is on the schema will always arrive within the seconds of mandatory waiting +Thank you Binglin. +Ran the test many times against did not see the error described +Hey RobinOn the where did that come from? Is it something that can be checked in +So was committed to trunk +bug set to fixed +Well it is not very common to log as root to the graphical desktop environment +Got fixed in +If fsync is true any hflush call will force the data to be synced to disk and any block will be synced to disk upon close as well +Are there any chances to be fixed in? We use grails and maven on project and want to update to recent version but with it is impossible with this bug +Attaching another patch with the changes required to fix the one failing testcase +It maybe also part of D and prior but I didnt run on the case with my system +Bulk close for release +that way we have still backwards compatibility +Committed svn r +I just committed this +Fixed +The LOB stored procedure code did not detect this as it only checked if the reference to the was null or notI fixed this by also checking if the is open on the client +changes are bit more involved than mapper changes +patch attached +Patch adds to handle the caching and uses it +Closing issue Oozie is released +The docs need to be fixed +I think this is useful +All the other elements are in nis and autofs schemas +That would make this work as long as you access the row data through the instead of retrieving the and using that directly. +I dont have an OSX machine to test that +I will file a separate jiraThis has been a long standing issue +It uses a discovery process to see which of the four possible logging backends to use based on its visible class path +core tests +My clever cloning plot didnt quite work because Throwable doesnt permit you to overwrite the initial cause once it is set however I submitted something that is reasonably equivalent. +Now just grows horizontally not horizontally and vertically +Please try my patch according to the spec of Java +Same hereSonar PDF Plugin Plugin disabled everthings finePlugin enabled Same error as mention on top of this pageThe requested URL contains problem Sonar PDF Plugin My requested URL contains about byte errorMy requested URL contains about byte no error +Thanks guys +Marking this as resolved since with AS which contains we now have timerservice functionalityAny specific issues will now be tracked in separate. +Committed to +Depending on how literally you mean pristinesetUp andtearDown might help you clean up any state left from previous tests tooSorry I thought these were Hadoop tests that were taking longer +LGTM Nice catch to branch +What do I think? This is a big change in AM internals +Similar code can be found multiple times within wicket core and it is working +Thanks for patch David and review Uma. +Also the MEAD maven repo already uses the new artifact and things have been changed to refer to it +I should add some justification for this request It is provided my jUDDI customers expect to provide integration with JAAS Apache jUDDI web site specifically states j is not for production use If customers want to use jUDDI for registration of their web services they will expect to have some control over which users can publish these services to the registry +Update applied to move the authorisation check outside the protected code block +Please free to reopen if you see it is still a problem. +I didnt realize Lily had posted a patch but I think I will use this one +to what Grant said +Unless of course you want to change the shutdown server functionality as well +Need to check if there are no other IDM issues to push there as well +has used invoke method from the release what breaks seek on the components with rendered attribute set to false +Fixed patch with a version using a private method and no +Yes they are the same version +Hi LeonCould you please confirm if this is an still an issue? There was a commit in early Jan which probably resolved this see rev +The numEntries stuff which would not let the code kick in if there were less thatn NUMROWESTIMATESIZE of rowsWe had used as a heuristic for unknown types. +Error log as in syslog +ISNT but then neither is! So uh never mind +The real bug is that the compiler picks the generic version over the integer overload +bq +Cant find a way to resolve the wtp issue +Nice find there MistyI think adding this info to the clustering section will be useful for the User GuideMDB Message Driven can incorporate this information once the main Clustering section is written and once Elspeth has finished her major rebasing +Quite not figured it out +Setting this issue to Resolved but not closing it until Rick merges it into the branchChanged Fix versions to I hope that is correct +Heres a modified examples that shows the error it adds enough records to the file to spill over into a new block +A sample that shows the jsf? should provide a feature to guarantee the compatibility across frameworks +It looks like its somewhere around the code where install directory is chosen +Thanks Richard +Moved to to this functionality as part of a Spring Integration port to +I found a way to reproduce the issue and then the way to fix different files in Tools XML switching between tabs in each editor and between editors +Thanks Amareshwari! +depend is Ant slangprereqs is the appropriate term for Maven +Mass closing all issues that have been in resolved state for months or more without any feedback or update +The gettimeofday change is fine though that function has been deprecated in favor of clockgettimeId prefer if we could come up with a solution that wouldnt require further forking of axboostbase +Ive updated the junit version recommended by +Not a UUID happens on nodes so I thought that it would be appropriate to fix it there also +All right this is legit +Thank you for the patch Jolly. +Unless you can provide more information to reproduce your issue like the output of mvn v that shows the actual platform details we cant help you +So we cant use Jetspeed because we have use cases when admin users configures public pages +No test case +patch works on trunk +From someone who recently integrated snappy and support for it into an internal build this would have saved that time +Create a unit test that can reproduce the issue if possible +Assigned this bug to Rick since he fixed itUnchecked the regression flag since this is a bug in new functionality +Please check latest andor master branchesThanks! +Browser sends every request with JSESSIONID parameter +Looks good to me +The problem with the test for was fixed under because this has been in RESOLVED state for over one year if it turns out to not be fixed please reopen. +This work item is intended to track the implementation of Policy +The problem has been fixed. +Awesome! Curious how this is implemented +Extensions were introduce to deal with already signed dependencies but I thnik this make the usage of the pack a bit messy we should have a better look to this pointAnyway once this done can I close this issue or is there someone that has something to propose? +Im not sure if we should do anything +It doesnt look like a problem of these particular tests +ClausI just didnt some test about camelContext name within the container I dont find the issue of static atomic counter which causes all the camel id end up withBut I agree to add the bundle id in to the camelContext which will help user to find their camel context easily +Interfaces overlap +The underlying cause of the failure is retained as a nested clause in case whoever using the client library is curious as to why the connection is failing +I havent tried Hibernate yet +for the patch +Some time ago I had suggested a related approach Query parser should generate for pure wildcard terms to boost query performanceThere were objections from the Lucene guys but now that the Solr query parser is divorced from Lucene maybe it could be reconsideredI couldnt testify as to the relative merits of using the filter cache vs +I seem to recall some discussion of this that it was not possible with Dita to have a cross manual reference +Would a suffice here +We are not changing the notation for it has its purpose +Above I say If successful it will replace the original file +Stefano your worthwhile Mail refactoring broke the patchNevertheless I expect that it will ease unit testing a lotIll refactor and resubmit the patch +I updated the pull request based on Pascals comments +patch looks good +We really need to revisit this trailblazer at some point hopefully before the next have tidied up the jar usage in the trailblazer and it now works from within the +The c broker currently strips the domain to accommodate the java clientOnce the java client is fixed the C broker should also be modified items unlikely to be completed for out to have checked in a fix at rev in Qpid trunk. +has been marked as a duplicate of this bug +Commit Changed function to to be more consistent. +I am trying to work out how to replicate your issue +If you already have JAAS configured manually you will have remove that configuration +It requires JVM to run +I just wanted to point out that even with this patch there will still be spurious failures if the task notifies the AM before the AM sees the container status from the RM +so different messages would be better +Logged In YES useridThis is the expected potential behavior when independentdeployments use the same classes +Attaching with more changes from hbutanis review fold into and into +If you cant let me know +This does not show in the currently deployed site +Yes theyve been applied correctly +Release note added +Martin I have attached a smale example +My argument is that the token is already there and its been parsed by good working code its just being masked by the process +BrunoIt looks like one of the Hindu labels was changed accidentally +This is an old task which might be already fixed +Committed at subversion revision +Mirrors should allowsymbolic links to be followed or they should be removed from themirror list +Id vote for option +Thats right +Still looking some minor comments +Thanks +Archives zip doesnt contain tools which means you cannot create a division of grom AS runtime +You need to remove qpiddxdispatch as well +Thanks for the patch Bruce +Thats a regression plus it makes other efforts to ensure things are working correctly in trunk more suspect +We should be spec compliant and not compatible with most implementations I dont care how wrong IE implements this +It fully fulfills my requirementWhen will the fix be available in the official maven repository +On a large project with a quite diverse group of weve found it a lot of work to keep creating and maintaining dozens of custom tagx while the library of components appears to contain much of what we had to build +That was quick! Thanks Galder +Needs modified which is is a modified class +Thank you +Of course i dont need it and wont use it as i do still see cases where java devs might want something to be public but not accessible in a template but i also think it should be easier to override that default than it is +Refactoring to notiifer Jesse Can you please patch this +There are two things that might be worth considering here Why not access the pojo directly by its id? Perhaps this would be best handled by an addition to the we only have a barebones history implementation in place right now but one of the things we should be able to do with the history is add information deeper in the stack without violating encapsulation +So i have not used the new system templates which are for +In a model where the partitions end up in different Shards databases etc +This patch file changes the behaviour around this method No patch is provided for is it possible to merge this in or has there been too much change in already +Unable to reproduce with JBT and Osgi standard plugin deployed to EAP +Closing. +But that said I know practically nothing about optimization and I dont know why that particular property was added for the old test +The submitted value is altered by and it is not in our hands +Absolutely it makes sense +this is not exactly the same that I mean java code editor similar navigation +For my purpose i wrote a plugin that gives jcoverage the list of directories to also wrote a report that works in conjunction with the one generated by the current plugin but which is an xml that is put in the directory and which integrates better in the of maven. +I am guessing I have to keep the name of the patch same so tests can be run +I request for this patch to be consideredfor reviewscomments +Patch looks fine to me! Weird I swear this worked on my mac before pushed a patch +This is still not quite as good as it should be as we are still buffering the results just not sorting them +Theerror has more of the feel of a bad invalidation of whatever the second query you are running +Also affects on Solaris +Perhaps you could use the strategy design pattern for the algorithm? That way it could be pluggable +Thanks. +I suppose it is best to have Axis default to a safe behavior +What issue it is solving? Also already commented on your gist but it should be made switchable for the following reason This patch has a severe impact on couchapps too since its restrain the ability to browse multiple dbs and such +Closing as it seems to be a bad use of the onclick attribute. +Thanks BruceI will look into it +Looks good +Just found that was already filed for the main issueIll edit this ticket to fix the cases where we are not consuming diagnostics at all +bq +I wrote to the list but I do not expect to get any reply seems to be dead +Fixed in rev +Most other implementations use a wrapped stream approach so Im not sure that a straight replacement would be available +The new screens are in the places suggested by Jacopo CappellatoComments and suggestions welcome +The hbase +Aparently going through the mail thread again you are not using a WSDLPlease attach the code the file as well as the command that you used to compile the service to this issue +You must first rightclick at bin and then click Build Path Add to Build Path +Its committed at r in trunk and r inIn the same commit I have also reverted the temporary window presentation workaround for product lookups which I did at Weird variant lookup for +MB is per right? Does seem a bit small +Didnt Bil has fix the issue in trunkIf he has then we can close the issue +Reopening since i dont understand how it can be resolved when you say it does not work +Closing as resolved +I dont follow why flushes are foreground but compactions are background? We can flush while serving reads and taking writesThe only difference would be during region movement in which case double flush prevents unavailability due to flushesOtherwise its actually the compactions that we wait for when we put up the update wall not flushes unless youre talking about the first flush not finishing before its time for the second flush +It still doesnt explain why the validation worked when I did it in the custom handler but not in the interceptor but removing those makes the interceptor work properly +Alberto this looks perfect I just have one last question will run on JME? If not we need to either include it as a separate module or document it very clearly +You changed mSsql in HEAD +We need to move out of SNAPSHOT asap +The spill percentage still controls the soft limit which resets the equator into the unused part of the buffer at an offset proportional to the average record size +According to spec rows unbounded preceding legal rows unbounded following illegal rows between current row and unbounded following which fixes grammar to handle these cases +Setting Fix Version to +Not sure this change is worth it +Thanks for the review TedCommitted to +Does v have versionning now? If not we will have the same problem we have with spidermonkey +Thanks Ted +Attaching the merged patch +This is very useful info to have thanks BryanI have a couple of language suggestions Being kind of a stickler about differently than I would suggest changing the sentence to eitherNote that these scripts behave slightly differently on UNIX systems and on Windows systems +or vMy suggestion is that we add a command handler attached to the RCPT chain. +In actuality they do behave differently for example when called without a block the Enumerator returned must call the original method name. +Same with and other charsets +Added setter in to be able to update the value from +So you can apply your patch. +What is the situation with this on SP with Maven? I have set MAVENTERMINATECMDon and regardless of which command I run I am always getting ERRORLEVEL as . +Hi GuillaumeThanks for the quick response +I am very familiar with Java and Android development +I modified the test to be more permissive of keys leftover in the threadlocals. +We can revisit this problem when we have more time +This will become important as we start changing the DSLD language itself +But before running this patch first apply the patch from this issue +The following patch should allow the EJB Binding to work with both EJB and EJB fantastic thanks a ton +I can however implement a class implementing directly and make this mutable +Cool I will try to get a patch against trunk ready +believe that this problem is fixed by +It would probably be better to file a bug report against J and close this issue +fixed in head +Jetty is sanitizing both the HTTP response line and the response body so it is OK +Nicholas Konstantin or Hairong any comments on this +Patch applied to both and trunk +Allowing full metadata updates is significantly more work and requires more metadata validation logic that we current have but thats the direction were heading +st Qu +from me +code for xwork annotation JeckiI added these annotations as unsupported feature forFor we will rework them to fully support the configuration features +Thanks Kevin! +Works fine now +Please review your items if not ready for please move them out to future release +Ive seen and seen reported that the constant score version scales much better +But our user wiki is not translated hence the link doesnt work +The EAP build now uses the component versions inherited from the new BOM structure created in +Latest patch should fix the failures and address latest comments on RB +Thanks for looking at this Sandy +Many thanks for contributing it JamesOleg +is also used in +Fixed in and is no longer under development. +Thanks Matthieu. +Need to test a bit more but quite likely this is related to repair on with secondary indexes +fixed in +All the patches for this issue have been committed +Im sorry my description is so vague but its hard to tell whats actually causing the freeze Ive added logging throughout the application and havent really turned up anythingPlease let me know what sort of information or description would be useful to youEdit for reference Reproduced on Galaxy S Motorola Milestone I can test S if that would be useful +Solution is not as simple as allowing the category user to view all the infoFor example We have a model that has Person fact in it +Thanks +From what I see now in code I see that this method will return null if no or Subject is registered in or as so probably this information should be put in javadoc +How To link removedtutorial link explains its only in JSP with a pointer to the wiki +Mistake was at my side +in preparation for our migration to the current version of shindig and opensocial im revising all patches and feature additions we did over the last year and will be looking at which are worth comitting +Further appears to provide these classes it just needs applied to the and branches. +closing with release +So a few things are broken +This will simplify running the example +Committed revision +Hi Sijie I have edited it a bit please make sure I didnt break anything in this new patch +There are many projection examples in the documentation of course +As soon as UI is started itll override the by its which does not pop up the auth dialog +Migrating bugs marked as resolved before and tagged as version as fixed in RC +Working on it +Thanks for the continued feedback Knut +This are my test case files +Ill check that tomorrow with as they are connected +My current plan is to leave the server side as it is and upload a new patch that hides the client side SASL code into a custom socket +Submitting +Codehaus png logo for Nexus attached! +A few comments on the patch Using a static could potentially cause problems if the plugin is used more than once in a Maven project since only one parser instance is used and it would retain references to earlier schemas +Trivial patch attached +And Im sure people finding this issue would also find it valuableThanks a lot +This is not a native platform issue but rather a issue +LiangThanks for the referenceShould be a bug instead of improvement +All fixed enjoy! +during deserialization the Loggers that werent static caused an exception to be thrown +If you finish it before we release beta please correct the version +the definition of primarykey is that it is and unique You can use that argument for any length hence just keeping the default is the you want to add a Default string length which when specified inserts colum lenght and if not does not putany annotation +Sure your review will complete more my first step +results on openbsd docs created in s docs fetched in s docs deleted in si dont understand why its worst than on the mac mini +I ported this to the branch as part of issue using a named query is covered by + +Its a hack when it depends on implementation detailsAnd I dont necessarily think hacks are a bad thing +But indeed it would be fine to have a XHTML serializer in Apache XMLs based on discussion in +Its problem of seam templates in vpe but not of seam +I have added release notes to svn with rev r and updated them a bit for and put them in the two +Please also provide a trunk patch that can go in first so we maintain feature parity +No tools or patches necessary here and definitely no need to break backwards compatibility +Ok +has garnered some more discussion that should be considered by anyone interested in this issue +wouldnt it be better to open a new ticket for specing becomejava! and just resolving this +one should check again whether the HDFS or MR builds break since the signature of a public method is changingThis method is not being used in HDFS or MR in trunk. +Revisit issues that have bugs older than a year as Out of Date +So its safe to keep this close +Ben in our conversation you said you were going to handle this so Im assigning to you +Does that sound reasonable to you Konstantin +Having a SOAP message with a BOM is quite uncommon anyway so I dont think this will bother many users if any at all +This issue is fixed and released as part of release +RM completely forgets about all the applications todaybq +Thanks again + of a bulk update all resolutions changed to done please review history to original resolution type +Im not sure how to get it to EAP distribution +MikhailThe commit related updates that usually comes up once a commit is done was appearing in this JIRA +Does anyone have a test case that illustrates this issue? Im having trouble explainingunderstanding under what conditions this defect shows itself. +Looks straight fwd +Hmm I dont know if this completely solves the problem even though the code seems like it is right +Simple patch to avoid duplicate stop +on Linux buildbot were running into undefined reference to + file shows the sample output retrieved from Geronimo ruuning from eclipse in trunk at revision +I just committed this +Thanks looks good Ill add it soon waiting for a couple of others +I am inclined to commit this assuming no one wants to run the old junitTests anymore +is a warningerror that some text couldnt fit within its container +Youre right +Thanks! +This problem was probably caused by use of facelets instead of the one we dist with Seam +To quickly see where the problems are look for instance for +I am using Rampart and I could not get the alias as you specified +I havent heard from Bigtop but I hope this will meet their requirements +Therefore it created a lot of local files and failed with Too many open filesDoes this mean theres a fd leak? Or at least a leak during the create request? If so is the test at fault +Yes could be moved +yes +Sorry I dont know I have new job now and Im not working with jPDL anymore. +My suggestion is that this be added in phase after Henris initial changes are committed +To pickup the hama fixes the hama version will have to be bumped to +Didnt attach file +Ive just recommitted this using svn mv +Perhaps ant for all of them and ant for these fast ones +Galder how are you testing that remote transactions arent being shut down? Can you package this as a stress test +It didnt have anything to do with the email structure +Thanks very much for all the +Thanks +I added a number of comments to the document. +Patch the reading code to show the counting error +OK that interesting Thanks Simon +They dont need an extra dependency on hsql +Could you raise a Jira? Maybe if you learn about it you could kickstart a little documentation for it? +Improved patch that splits collecting of files between lucene and solr so extra folders in root dir are ignored also exclude solrdist and solrpackage +I think ko is more suitable for the name of the korean anzlyzer package +moved to CR since we have to focus on critical bugs in duplicate issue +Please make a judgement if that change will impact the whole logic or not +Will post it for review once derbyall passes +Patch ready for review +My bad it does not work well in trunkIm going to check whats going on +All the recent builds have the same +Deferring to +Runtimes cannot be and exported or stored separately +This fixes the issue for most cases. +This will not get added +Is there a way I could use the local installation of ant without the root having to update the ant installation +If the first bytes match the zip header signature it does zip checking +Docs for this feature were committed under +So I guess my next question is does this have more to do with the max size of the buffer? For instance if I made a K Buffer would that be just as fast as a creating one of arbitrary size UP TO K +Elements are sorta like a Java anonymous inner class +Adrian is working on faults this still an issue do we have a test that proves this case +Hi Tatsuya +As this isnt the case for Word files everything seems to be fineThat being said the description of the property in the specification probably could be improved +This will also avoid the values again after a rehash which is another problem I had in the past and really slows down rehashing as it relies on the peroformanceIm really talking about the shared cache stores only so Im not sure how this could be applied to the use case +This bug is very difficult to find +wow thats a huge perf gainThanks +bad idea on my part in general dont knowWhile you rewrote the logic with some enhancementfeature addition youaddressed the existing issue with a better logiccode +So passing that via file does not work +This issue has been resolved for over a year with no further movement +Thank you Meria for including my patch as wellThats Jason for have a closer look and merging itPlease guide about howt to categorize the bugAs this was a issue faced in MAP REDUCEAnd was supposed to be raised in the right JIRA project is a straightforward mapping from the projects in the code base Anything under maps to Hadoop Common Anything under maps to Hadoop HDFS Anything under maps to Hadoop Map with test case +EAP on AIX is now working with patch +Of course there are still a few other public static which need to get cleaned up +Instead these classes should composes the fields required by the user. +Nothing changed with port usage +patch looks good +Forget about pluggable RPC implementations the complexity doesnt deliver benefits +Added support for DEPENDENCIES files in Rev +The application contains a static resource which has the hash ABC appended as its version +I was wrong when I said that the issue was OS dependent +Konstantin if you care to review it once more please do so +Was there a patch for this? If so Ill commit it +bq +This bug introduces problems further on in the dependency chain when maven randomly picks between and to be included in the classpath as a transitive dependency from +Sounds good but I have no idea how to implement that! Can you give it a try +I added the patch as proposed in revision Thanks fro the contribution! +Let us fix this +Pushed an update to the snapshot repository +The transaction synchronizer capability has been removed as the benefits it provides are not clear given what will be addressed by +I tested this problem again to create pull request for the test +This patch generates a bit field translator based upon what Hadoop thinks is true in vs +Patch that fixes the problem and adds tests +I guess the solution is to either update the README to state more clearly that the is required or update the file to also check for in the available task which is checked prior to running the cameldot macro to decide whether to run the macro +Now that work is well underway with AS all previous community releases are +So it looks like a new image incorporating the TM logo is required +resolving based on commit log from management server log file and cloud DB +Attached is a sample implementation +Sorry that last long list is not correct +Sounds reasonable and quite easy to implement +Or are there any exceptions? Does the interceptor get called +Alex +Thanks Kevin patch applied at r +Straightforward patch +The API is versioned so we can always add that in as the use cases come up +Ah my mistake +db version +Though it needs someone who wants to take care of it +Uhm thisslightly? slipped for a number of reasons +The original Collector code is identical +Ill have to review the code but I took your Jdbc stuff as a starting point so I think well be ok there +Lets have a joint look at this in the timeframe +Sorry I found it use return address now but it only accept is a enough? eg the s this for M. +You shouldnt be rebuilding it +But unfortunately logSync throws so now logEdit has to throw + +Ill have to look into the overWritefalse issue though +Please proceed for QA process +Improve the if condition in +You should use this as your servername +. +Can we close this now Ted +If a commiter deletes the attachment I will upload it again with the licence permissions +Many of the changes such as fixing the openejb groupId may be duplicated in other patches +SDN didnt fallback on an exception in a check method. +AlefAny chance to address this before Sunday evening? no problem +the patch with the there are no objections can you go ahead and commit this fix now that we are past the beta? My personal opinion after reading through the javadocs is that we should support this +I dont particularly like that so Im OK with the new dependency +See attached file +It might be a bug I cant see a reason why it gets incremented for when went toIf this is the case then the can be reverted and the test can go back to catching +Now it has come up again I will go andchange the behaviour unless someone voices an opinion against this +It is listed under Maven Project Configurators update site category and has to be explicitly selected during installation +Once I have tests passing Ill have another look at some of the short scan performance implications the change addresses +Thanks Aleksey your patch worked for me too so I checked it in at rAndrew in the original patch did you change from using Iterator to because its faster or for a different reason +Unless they could get the from the or something of the sort +HiOk thanks for the info and the fixYour IP is now on the rsync serversYour mirror is added to the list this will showup in an hour or soRegardsHenk Penning +I reverted the pre interface change part +Fixed this in trunk please close +I believe this is a dupe of There is a patch associated to it have a look. +Putting the definition into the child xml doesnt work with Spring +Can you try now +I cant see why Axis C should generate the code differently than Java +Sorry Mickael this is not planned for Sonar but already for Sonar +To build from the root of an SVN extract issue the following commands +So in patch if rpcClient was still null after reconnect is executed you are simply setting up the Event without using it +Is that required for hbase or just something you took care of while you were in there? Im not sure what we want to do about minor hadoop version compatibility but just thought Id bring that up +Verified in ER Service Entity is displayed in case the description of a service is empty. +I assume you tested it because I didn brings us to what will it take to release? +Else Id burry it +I would have tried it but I found the fix before reading your comments +This tree printer can be poked into the compiler using the previously attached database procedure +Converted to groovy into trunk rev +Its a big change +The class itself could have a default list Oracle that mimics the existing functionalityIf you agree that this is a worthwhile interim solution I will pursue providing a patch +No longer reproducable. +Attached the AS server log +Closing issue out as part of JIRA cleanup +Thanks for the patch! +It may depend on the connector being used +Hopefully we can see a performance improvement in the quarantine regression tests. +Be careful eliminating the bracketing step +Ill try to do better in the future +Will post on review board and after it make it through there I plan to backport to +It still does NOT work +This is just a bug fix after allThis happens only to columns just after those that the scan actually selects +Do you mean you want a way to add a global ast transformation from command line +There was a reason that there is a shaded version of asm used in Geronimo Spring etc +Give it a try +Since it can be replicated in stream it is not new issue and should be fixed in next version +Also I want to save space for filters and dashboards so this is another reason to not put the logo on top +For Junit tests the system properties added in the XML report is a dump of the classloader classpath used for tests and not the value of. +Too many processings +It is called in SWT code from native method +I never have had or seen anyone else have this issue unless too much was included in the web container +Yes I did set JAVAHOME var +through hudson +If theres a way to automatically make this work that would be great + and are seems to be the cases when shared dirs not available +ie +Java class attached +Committed to trunk. +I thought it was will t ry to reproduce to see if I can install it faster +Simon for me this is already the case but could you just confirm this point +Considerable rewriting happened by the release including most of this +So the old codes cant find the service classIt is same to +it builds successfully when i ran android update project p +a quickstart to reproduce the problem +Note that with this commit trunk is now tracking Hive trunk +If has kb nature and builder then everything worksThe only case not implemented is when is a bare Java project with no kb nature and builderCurrently kb builder builds only project to which it is attached it does not look into projects on which it depends if those projects have kb builders they are built separately and results of build are shared with dependent projectWe can compare this situation with the following +Cdric can you test if it is arlready fixed in the version? +Youre quite right the patch is wrong I shall rewrite it to properly combine the includes and excludes left the includes aggregation also who knows why someone might want to include an external resource into the report +I will investigate +Looks like the dialog checking for the existence of the iOS SDK can still trip people up +This should reference to instead +Can you provide your new patch as an SVN diff file please +Passed in Xemopt mode +Six of one +that npe is odd though +Tests for that parser will be valid for all transportsWhat is the point of having the header continuation parsing implemented in three different placescheers Roland +is it persistant +Please let me know if anyone tried and these patches are working fine +Thought +So the case of changed somehow? Id be interested to know for future reference +Sounds pretty serious but easy to resolve +This metric would only really be used for a system that performs a significant amount of work +I will work with Viraj to make sure this makes it in +And add the xsd files in to resourcesschema folderthanksRex +Thanks Anatoli. +Would this make more sense in JSP Tags +Could you pull and resolve the conflicts? Then I can apply and help with the test +I have a feeling this might be a bug in the those of you that have commented filed this ticket can I be safe in assuming that you are attempting to download from a repository layout of default +Hi jacquesI found a minor bug in the bizznesstime theme when we open a layered lookup inside a layered lookup the image of the calendar is replaced by the fieldlookup imageId fix it attached the patch for it +Im sorry to say that I dont have access to that source code anymore and so Im not able to test this new version +However if I only put one consumer pointer in the that also mean I can only update one consumers position property when requeue the message +We can maybe only remove the new but the other one is identical to current code +Attaching the failing +If a user doesnt have an LDAP entry we should still look him up against the ofbiz user loginsAlso if a user is defined only in ldap but not in ofbiz how do we define what ofbiz application permissions he has +bq +has been added with description Sample as the error message tells you there is a cell overflow +We might need to revert the extends changes and copypasteedit the company classes +This configuration is overwritten by in dir where the retries property is missing this causes the retries value in to be zero +Not sure so Im askingIs this primarily to distinguish between the edge and inner? If so would it make more sense to just make use of Token type variable instead +I didnt notice the license on +Reverting because breaks building against hadoop +This issue is a duplicate of that is fixed in JBT JBDS +Is this issue been fixed? If not please provide your inputs when this exception is comingRegards be honest Im not sure +As discussed now with UTF keys and the text based output format +All tests pass in trunk and +Hmm +Here are answers why does it need eclipse compilerThe test to reproduce the bug must be compiled by Eclipse compiler +Arno looked again for a possibility to close the issue as per your request but have not been able to find an appropriate +Very minor patch to disable the pom checking +Deveraj So that seems like a decent workaroundGoing back to original comment sounds like we shouldnt be assigning regions for tables that dont exist +If no one objects Ill commit this in a bitThis is NOT A COMPLETE SOLUTIONRemaining work separate app client stuff into more modules consider separating connectortm into more modules FIX THE DEPENDENCIES +Also changed the timestamp nano precision to since mysql does not store fractional seconds +Im not sure why we would use or though over standard logging solutions + +I think we should get more input before closing this issue +Verified in +Can you describe what other issues you are facing +It would allow a user to step through a guided path to create the service gateway listeners actions etc +By eliminating the primary category selection in it seemed to then default to the desired category in my cases +You may use mvn clean jettyrun and check on httplocalhost +hi Grzegorzin the article there was something about Lack of a Standard Library +A patch for JDOMs Huffmans patch seems to have resolved this for JDOM +Paul you should definitely review it also before it gets inI am reviewing it right at it as well +This could break existing applications in a bad way +Assigning to Lorin. +fix run twice issue +Just for completeness +I dont think we should start down the road of bundling external libraries with Thrift though +but that well want a banner image logo and maybe favicon +shutdown doesnt not source there removed the debuging options from and updated those ones in +It does NOT generate redundant prefixes +Hmm the way I understand the test is you do need write access to the parent but not necessarly read access + +The best actually would be to provide two so that Cocoon release engineering team can pickup the one without securitystuffSo Cocoon users could use the latter when wishing to execute on SVG +OK I am resolving this as cant reproduce +See. +This is by design +We have done a better job then usual producing a wiki page with a design spec for table links +my god I need this feature! High concurrency not work for me +If I access an action directly that does nothing more than go to success and forward to a tile this error shows up too +The problem is a compatibility issue +Thanks for the quick review Knut +This is now removed in OSX Android iOS and more importantly in the JSSubtasks created for blackberry windows +Added David Van Couverings entry to list in the description +Is it possible to provide a for +feel free to commit it +Thanks for working on this BJ the patch works great +Closed after release. +This will also create a new connection poolSo the ideal solution would be to have a Spring test in which the beans are injected into your java delegateBesides that I also believe my fix on the branch above would fix the connection issue +Just a note +Therefore resolving the issue +The patches Laura attaches to Jira are in Windows format +Why was this not caught by any tests? Should we add oneJan I have opened a separate issue for that so we can close this one for x +Steve Dynamic fetches are always to be considered a hint +Let me try to modify the test so that it can still test the Streams +Given that it is not core functionality it could be a wiki page or a readme file +integrationmaventargetclasses and not for +Sending srcmainjavaorgcodehausmojowebstart file data +Committed +We need to schedule our workload ourselves +Ill do that instead +Hey Otto we got the patch landed in trunk thanks for your help +We still use it for as well if you request a hashCode +I do see what your saying +Certainly easier to merge changes to that thanaccept an entire new file system upstreamLet me know how I get can started helping outNoah +Richard please this if you disagree +Issue reopened to change Release Notes field. +The files even have a mix of unix line endings and windows line endings +Thanks +MILYN +Attached is a patch that fixes the warnings +When the C code has not been compiled the crash does not happen + and are the only classes that have set set incorrectly in will have the linger timeout set to instead of will receive the full backport where timeouts are configurable +Thanks Todd for the review will commit this momentarily +Ops! Sorry about that the duplicate is not CheersMarcus +Nokogiri installs no errors. +Ill just go ahead and close these. +I believe it isnt cuz i had to patch the latest cvs to get crlf working on my own win xalan build +Thanks for reverting your veto +nothing critical but annoying +Im working on this now +Please confirm and close. +Can we make use of rather than hard coding the list of special chars +sorry Maurice +We are splitting out the repository generation into its own Mojo and custimazable repolib namelocation is now supported +Otherwise the indexing would then be exactly the same +My proposed solution is to remove the code in the filter that skips those and instead map the filter to the grails and gsp servlets +Actually Im seeing crashes on Raspberry Pi because of compression on internode commsIm not necessarily against using lz for internode compression but its a slightly different issue so Id suggest opening a separate ticket +At the risk of pointing out the obvious I tested this on the RC using ij and the issue does not ariseInterestingly enough the folder that is created for the database seems to have the Unicode representation of the character and is just called ueIm just wondering whether it isnt somehow assuming the ue as a string literal rather than a Unicode character +So we can close the the issue +I cant think of a case where this kind of setup would be of any useCheers Olegyoure right that the cookie policy becomes pointless if cookie parsingis disabled +th? Currently all projects are visible in the you cant see the projects like Fuseforge that have already been marked private in the reason the other projects are still visible is because we have not yet made them private +grails does this with glink tag +Its fixed now +We could change this to a decimal giving a max value of +The failures left now are due to connection refused exceptions +Thanks Sean. +patch for GORA and trunk revision in GORACommitted revision in trunk +If you still see an issue a test case that shows the problem would be helpful in debugging this. +index folder in the repository during startup +I am not too familiar with so I dont want to commit this without some review + belongs in the client Jira spaceIm new to this is there a way to move the issue or do I itSteve +time of last retrieval connection security etc +Thank you +Agreed that commit does provide a directory with all the libraries by using dist so I think we can close this + +I see that datafiles is missing from the commit +Thanks for your help can close the issue. +Bulk close for the release +About the only thing more we could do here is change the value as typed in +m almost done w a conversion +It says it will use this object not a clone of it +Makes sense to me +same patch but with cosmetic fixes the writeFreqwriteProx etc in was imo confusing especially as in flush its assigned to readFreqreadProx etc +This issue is marked as resolved but that does not appear to be the case +Is there any further news on this patch? From what I can read above the tests now all pass and it appears that the performance improvement is quite significant. +Committed at subversion revision . +Can you please provide more information complete pagebeans code +From the sounds of this it looks like there is an impasse here +get is the scan a lot of rows method +fixed in thanks and that was fastChristianthat was the same URL I consulted to determine that it specified an equals sign for the table metadata +Hem +the basic step +Every time I think we are closesomething else pops up +Running test +Thanks! +Youll see a resubmission from me with these items within hrsThanks for working with me on this + +car deployment of the rarIf we werent copying jee artifacts into car files wed have less of an issue although I would still prefer a minimal AMQ rar to be available somehow and for us to use it +Im willing to look at this again if someone is having a problem with RAD or later but so far no one has mentioned there being any problems with installs on later versions of RAD +Now name property of could be changed in Properties view but the changes is not saved +We cannot do anything at forrestIfwhen we move to a project we would not need to change these again because we would still be affiliated with the XML projectDave the current experimental +OK Ive retested with a complete build +Am on commit too +That should be generated at build time +I can still reproduce the problem in +The query version of the search component which fails the test +This might save others some work in the futureThanks for your help +pluginstotalWorks locally but Im not sure its working on remote +skipping rb attaching simple patch +I cant reproduce it here Ive let this test run for hours +I havent changed the commit messages over because people are not yet subscribed +This is the screen shot that goes with the patch +At some point there is an overflow and numbers even show negativeLooks like we have to first turn on logging for the FS so that we see more messages and figure out what is going onCode looked like this may be when reserved containers go into allocated but the metric is not updated or there is some synchronization issue with that update or something +Waiting on hadoopqa before committing +outstanding requests seems low though +I have attached a quick start that reproduces the error in questionNote i have also include my suggested fix to Form and +A more basic case is booleans as existence tests on keys +This is certainly an HP compiler bug but the workaround is not a big deal +and the logj and app configuration files and classpath. +This allows you to set the Created Time for both Timestamp Created values on the outbound side +It is a Connection problemIf I change the request connection settings from Connection to Connection Closeeverything works fine +Ive tried this again on the ESB and it is not an issue there anyway +Well the patch allow WITHOUT brackets but disallow WITH brackets there is a patch of the patch +I think we can close this since we felt there werent issues here since the code is just modified Apache code. +This issue is currently consistently observed on my Mac OS under Java with installed +this is how it is supposed to work for beta feel free to open bugfeature request for better +Isnt it simpler to set the user header or something similar header with the userId info after verification so that anyone can do the filtering later if they want noregards aki +This issue is related to and I committed a quick fix for it just after CXF was released +So Id like to break it into small pieces and tag them one by one +Another is to have the track the active nodesracks and filter requests for nodesracks that arent in that list when we normalize the ask list from the AM +The procedure to migrate the resulting web pages into the project documentation web pages is also describedThe API Reference Manual is properly targeted to the builddocsapiDocshtml directory +It would be better to encapsulate these concerns in a layer atop whose we already make promises about no +I would agree that the problem is most likely in the VM but wouldnt rule out a core class library allowing the VM to exit before it should have +Certainly the results for a given query would depend greatly on the parameters +Also it only has the add method as you suggested +The problem is precisely that and thats because I put an after +this patch is based on the previous one +It cant be beat +Thank you paulex +committed to trunk +comments +Bit tricky as I only have woodstox set up on one of machines I have access to +This is strange +Dhruba say CRC inlining is done and reduces number of by half +Ill put up another soon +In the shipTracks example the is stored in a model and rendered by after an xsl transformation to wfs +is something new in Wicket +The fun of infrastructure with non unified authz. +Observed errors in firebug +Let us know if the setting works. +Many other enterprise applications like Collabnet SAP are connecting to our AD instance and working happily +I hope this one is better please let me know if you need any more details +Yeah so there is not much more we can do in the meantime +yep looks good cheers. +I do agree with Chris though we strive for visibilities to be human readable in whatever language that human uses so lets make that the only option +An alternative approach would be a pool of those editors like it is done in the class in +Steve Id agree these differences are intentional and while I labeled them suspicious initially each appears to check out as expectedHowever for one Id like to nuke the var expansion anyway and two we should check with Roman about what the issue is in regards to big top integration and see what we can do to smooth things out if possible +ChristianYeah we can make this nicer so I logged so we have at our fingertips +But an attacker doesnt have the seed and thus is not able to generate the next numberThe hit ratio would be threadcount chanceofotheruserrequestinbetweenand a new game with even better ratio on each subsequent retryI think even unencoded this is superior to our current sequencer viewidhash encoding macThe reason is that any encoding which only generates a low number of bytes can be cracked +Committed to trunk. +Everything looks good +Then what about agregation with Views? Default values stored in the metric table would be used? +Maybe configuration yes + +Id also suggest to be careful about potential impact on performance +The st is not a issue as children is already checked before +Revisit issues that have bugs older than a year as Out of Date +After going to trunk then we can make this JIRA patch available +I talked to Max about it and he decided to skip that fix and enter a bug into the webtools are some more memory leaks that arent related to this jira and arent so extensive +Do you mind taking a look at the patch and have a run with that +Committed to trunk as revision . +Weve thought about it but have no definitive plans +Set the fix version to an early release if you resolve it by due to inactivity +The only one I can see that it wouldnt work for is if you joined by key fields but wanted to include a field as the additional property +If were going to validate configurations more thoroughly as suggests I thought it made sense to enumerate all the possible API properties and bail early if they are invalid +Here is another +I think this is ready to go in ill wait a bitI didnt make any changes re though I still feel we should fix this but it meansdealing with backwards compatibility etcThe changes in this patch are backwards compatible in the sense that consumers are already correcting initial posInc to posinc anyway +Ok lets let this change for the next release cycle +This is needed to make sure the same score is not distributed within different buckets +Now integration uses +Im wondering if maybe it somehow gets back around to select a second time +svn ci m Removing the unused variable as mentioned in by caching the object srcSending srcjavaorgapachecommonscollectionsbidimapTransmitting file dataCommitted revision +for the information. +I do not have performance problems anymore +Thanks for making the changes Richard +I made additional TestsIt works with and But it doesnt work with Can implement a workaround for this JDK BugThe Documentation says it should support JDK releases are binary compatible with version and require a minimum of JDK +batch transition to closed remaining resolved bugs +bulk defer to +this probably can be closed because its committed to and +Hi for your response +Here is a patch fixing the NPE +Elranbq Why do you say the assumption is the type is restricted to single valueParts of the discussion mentioned sorting which is undefined on multivalued fields +New look and feel pushed to server +In addition XMPP recognizes an HTML element in the message body but the endpoint only supplies the text elementId like to suggest the followingNew endpoint parameterspresence boolean Accept presence packets on input default is falsepubsub boolean Accept pubsub packets on input default is falsedoc boolean Set a doc header on the IN message containing a Document form of the incoming packet default is true if presence or pubsub are true otherwise falseNew input headersfromtypedoc Document containing the inbound packet if presence pubsub or doc is set message body still set with text element if anyNew output headerstodoc Document containing the outbound packet overrides the message ignores the message to sendreceive pubsubpresence packets +ClausThanks for pointing that out I just changed the constant name to DISABLEHTTPSTREAMCACHE +Code became simpler as addresses specifically the case of defining a fixed set of additional outputs during the job configurationThe patch as before allows the different outputs to have different and KeyValue classes from the job output configuration and among themIncluded in the Javadoc sample usage +ImprovementsRemoves remaining in and Eugene any luck getting this implemented as a +See lines of the class +I agree that is archaic but I dont want to change the behavior across the board for all applications since I have little idea of the impact of such a change +Yes intentional +FixedThis issue will be also fixed. +Redone the patch on the latest version +Is there an easy way for me to try this out +Thanks. +Proposed patch incorporatesSuccessfully passes testsGives on unpacking scenario +It provides easy access to all sorts of information about the current mojo execution and plugin +Hiya very much for this +Paul do you agree with the Critical level +If that produces the same behavior lets try a few a sanity check type and look for anything with Groovy in it +Its just a document update though so no new tests are expected +In other words it has owner +Let dlopen use the default search path for ran afoul of this too and Im dropping back to jsvc for now as a resultId suggest one minor change to the patch check for first before the names +If you find further issues please create a new jira and link it to this one. +Thanks AlanI have logged a ticket with STS for this +Patch for the test +Committed at subversion revision +However it is not guaranteed that the message will get through since the destination might drop the connection while the message is in transit +As knut has pointed out there are some supportedvarchars that we cant index but with current limit it is possible we could fix that without a major rewrite of index code by looking atpage sizes +Thanks +Does not seem to fly las I recalled +Committed at revision +NamitCan you please review this changeThanks on Namit Li has updated the diff with a unit test could one of you take a look at the diff +Blocker for what do you think about the current fix for now and if we see any issues we can redesign things as described aboveAs it stands this bug is responsible for a lot of hung hudson servers +Siamak is the rest of the patch still relevant +Modified the Query creation to use the cached names +It duplicates that and the related issues they are caused by the same issue was closed as duplicate please leave your comments in +Tried the same result +Resolving as duplicate of +Im not sure if this would cause problems but Ill try it out +If you plan to rename the mailing list we just need to remember that theres a special section for your case +I modified the test content class Main otherwise it doesnt compileLet me know if something is wrong +Im starting up tests now and will commit it if they pass +This lets us specify numShards per core though which is kind of what you need to support it inBecause the numShard will stay in the corestate Im not yet sure how that will work if you end up being able to reshard by starting a core with a different number + +Closing this one as WONTFIX for nowOleg +Committed revision +Attaching server log +To me they are distinct and seperate methodologies +Has any of this been repeatable using the trunk version of? The problems you are seeing shouldnt occur there as transacted messages are correctly without hitting the requestmap +It looks like it was broken when the task directories moved around +on patch v +Not all are so bad +Unfinished sentence +One of the big benefits of webhdfs is that you can update the server components without changing the software on the clients +They said the bug is their on a particular version of Informix +How to achieve this if the alternative sources also dont specify any package versions and user does not know the bundle symbolic names of the alternative bundles? They have all the normal mechanism available in to differentiate and can always disable the system bundle exports if they chose +I need to remember that I was simply copying the file to SNAPSHOT not chanelling all downloads to one file. +So the number can vary depending on what state the cacheis in before the test is runNote the counts I am talking about are the raw counts on the individual tables where the counts you are looking atcome from doing additional math on those numbers that math is in the optimizerThere are some points in the code where the sql layer of the system knows that it has just done a scan of everyrow in the table so it knows the exact count of the table and it updates the stores estimate through a store interfaceI think this is triggered by a full index or base table scan +If you think this is a bug please provide a testcaseI appreciate the consderation on raising issues but if you dont provide an evidence or testcase you are just wasting someone elses time. +Addresses Kan Zhangs commentsadds class for both Zookeeper client and serverremoves addcred command that was only useful for development and testing of with SASL unnecessary and unsafe to add passwords by especially without authentication and encryption +ok Ill take a look +removed from the trunk +Are there any workarounds for this issueLinking this issue with the related Dispatch client issue +Attaching new patch with minor changes to the comments and assertion failure messages in testcases +Perhaps you can provide more detailsThanks this is a hibernate tools related question only issue is that Sadi is referring to a releated version a netbeans plugin +Closing. +sorry need to reopen to remove the fix version field. +You can also change Fedora to since it also works there and is the latest Fedora version +So is this harmlessexpectedI will take a look at the master logs to see if there is anything interesting going one with these regions + +Hello Mikethank you for your suggestion +Since last is already a field in can we use those instead of adding a field to parameters +David Can you have a look with the new and provide a full test case if it does not work +Can you answer Knuts last question Do you still see problems after rebooting the database? Thanks +Removed JSPC precompilation from modulesweb +I am sure I didnt introduce any new release audit warnings since there are no new files in the patch + This issue has already been fixed +I will try but dont expect it in less than one week +for the patch +This is there in and so in +Has anyone else tried this latelyI just now tried it using Tomcat I first packaged a war file using forrest war no errors all went fineI then loaded it into Tomcat loaded fine viewed it in httplocalhostjsessionidECBDECBBBBand it all works fine + can you please attach your from location shown at Window Preferences +Thanks its really appreciated +That last change fixed the problem +The same needs to be done with bash jruby script +applied patch and test itApplied patch on the trunk r +Move to +Ill go ahead and commit this +The problem should be due to the following bug in RHEL systemBug not installed when upgrading from is not related to so close the bug now. +Ranier can you verify that the showcase default action works? I dont have an install at the moment +If you add one more node to the cluster the problem will go away +Thanks Travis. +Fixed in the branch under revision r and in the branch under revision r. +We may want to update it with the message in when that code has been written +You are correct +Closing released issues. +Thanks Binglin +I tried the same stepsRemoving the name of field makes the pencil work againBut giving a name to the fact doesnt make the pencil work +This patch also cleans up a few minor issues with the and POM +Ive gotten my snapshot rig up and working with hadoop and now +If they cant do this minor tweak why should they be allowed to participate +Ill look into this issue later today and see where I can get with it. +bq +This should meet most needs +Edited for readability +This is good news and it would be great having it working inside karaf +Should I restore the fox prefix +after fixing whitespace in your operators +All right! Heres a first draft of Admin Guide topics based on John Embretsens wiki material with some changes in in that book and the Developers Guide +one thing to keep in mind is we do not offer atomic guarantees for multiple rows or API calls +Sounds good +JimmyHave you run test suite over the new patch + Add basic tests to test grantrevoke feature works when upgraded with set to true owner of system schemas is changed from pseudo user DBA to the authorizationID of the user invoking upgrade Does this step happen regardless of the value of the property +Although implements Serializable it appears this it isnt really serializable +Thanks Vinod for the comments +Since we recently resolved several memory corruption bugs I retried to fix this issue and finally succeededFix applied at revision . +If you use the method it will return false indicating that this is not composite +the specification defines that and exceptions are thrown in case the item does not exist OR is not accessible +oh okay +is one possible implementation +Applied on the branch. +Simplify the test cases +It then gets reported by with a failure automaticallyThe question is the wordset is initialized to be empty if missing +The same code works fine in +Just to keep you posted I pretty much have completed this issue just running tests +Try on Windows then will close cannot be because of the unswizzled for taking a moment to look at this Brett +Proposed patchPatch is pretty trivial but gives on MT on Core +As the simple themes own documentation states it aims to be a HTML implementation +Changed the relevance API to accept a list of context types obtained from the context compilation unit +on new config properties +This is a useful feature +If you tried before the official release then please delete your Ivy cache and retryYour Ivy cache is located at USERHOME +CVD is launched using same classpath that PEAR Installer was itself launched with +Not that i expect you to apply it but because last one didnt meet my personal quality requirements +removing entries from attempt needs to happen after the callhandle removes the task attempt from attempt before it attempts to remove the attempt from +Currently looking for the referred specification documentation to see what it says +Cassandra with this patch only loads the migrated keyspaces after nd restart after has applies the migration +Please verify and close +Thanks Andrey and Karthik for getting this patch ready! +Ill put some instructions on how to configure development environment on the the other hand wouldnt it make sense to fix the archertypes you are using to generate those projects so you wont have to use such hack +Thilo Done! +Patch away +And theres some interesting stuff in there like rewriting methods to and such +I have fixed the logging to facilitate this +Ugh I just did a cvs annotate and then remembered that I had changed this last week in my working copy and forgot all about it +Why upgrading the such as explicitely? Arent they pulled transitively along with GORA +Subclassed to include my converters too +Its trunk version excactly manufactured a +apt performs complete checks with error reporting +Carl oops thats what I get for only reading the tail of the threadOK besides the interpolation question are there other unresolved issues from the discussion between you and Ed? Lets try to drive this to closure +What environment is this in? How are others testing if the mgmt server isnt able to start? Specifics would help! Thanks +Thanks for these nice new metrics Matt +Thanks stack +There are some security fixes for RHEL and Fedora tomcats that need to be done first +both dont work for with a trailing it works but only after a in archivatrunk revision a redirect to the proper URL in the specific situation of a GET request to an existing resource that is a directory but doesnt contain a slash at the end of the URL. +I cant guess enough to figure it out though so Ive gone for a largely pattern matching approach insteadCan you try with all your files and report back if the matching rules are too strict not strict enough +We wont be introducing further autowire modes mainly because we have the general Autowired facility in Spring supporting the notion of qualifiers +Youll have to at least check if the namingContext suffix has and only namingContext before returning it +Removed the related section from as its present in +Other Cellar bundles may be affected by the same behaviour +Ill retest my scenario I think that I did not have the query component configured at the time so hits for each collationQuery were zero even though the terms were in the dictionary field +vDone move categories tests to common add a dependency from to common for the tests put the checkers in common restructure to allow to add new checker add a checker specific to Still needed migrate the localTests to a newer version of surefire remove the stuff check the impact of hbase integrationBut globally everything is there +trunk closing stale resolved issues +Updated test case to play nice with +And if the donate check box cant be restored then there needs to be some mechanism for a donor to explicitly cede ownership to at least confirm the donation even if its legal status may vary from jurisdiction to jurisdiction +Specifically we need to move and to the API library +svn ci m Committing work in progress version of a new Range class to replace the math +I just retried this and was able to reproduce with +Thanks for the patch will apply something similar for the as well. +Not sure what went wrong but I will fix it +I need to explore deeper whats happening +will contain the fix for EAP +This issue is a cforms issue. +The headers compile cleanly now +i think this should be released as an knows more ill ask him to comment on how we deal best with this believe that the aguizar branch contains some parts that can be leveraged as we already did some of the rollback work on the day of the release +Now we seem to have to root cause +I have been playing around with this in my local workspace and this is working quite well +Do you still see this failure after the fix for went in +I added a test for this + +Spring introduces a Marshalling framework including Marshaller and Unmarshaller interfaces +Roles do have a stand alone existance they have to be attached to some user to actually work + +Now and pass on DRLVM without any problems as well +Im also going to check whether its possible to commit it to +setPPP moved from IWC to IW +If a MR job consumes too much memory you want to fail its component tasks and eventually fail the jobMy concern is that the current implementation of this feature is not doing thatTo be clear Im not against tracking physical memory used by the process +Please stop submitting test suite failures to jiraWe know +the extra stuff should be in addition +I see the warning if I build without an file but I dont see it if I build with jlib pointing to a proper Java runtime. +Wouldnt and suffice for you? Lets figure out what we need more on top of +I also tested other archetypeng goals and they fails or succeed the what makes this fail or succeeda project fails it defines archetypeng as an entension which is mandatory for using packaging +MaryamCould you please upload a new zip with svn diff against latest SVN? Looks like some of the code in this patch is common wiht the other patch which i already checked inthanksSorry for the delay I had problem updating Heres the updated patchThanksIs this a patch against latest SVN? Some of the changes for example with are already checked in as part of the prev patch that you submitteed +The workaround would remove the distributed cache settings from the action configurationThis workaround behavior would only be activated if a flag is set in +I think we should look at moving deletes to codec after you commit +Hi guys! I added supports generated keys to Ingres +I mixed two changes in last one +since alpha is already released this is alpha +Added a little modification to deal with at unit test possibilities now this updated patch looks great will try to help get this committed t forgotten about this Sean moving back right now but will get to it when I get home have a safe trip back to school yourself +Either way is fine +Actually I was wrong + to trunk +Id be happy to apply a patch if one were provided +Hey Denis Was wondering if you knew where to fix this issue in the installer +Arpit thanks for taking a look +OK committed without that one bit. +The repro showedup as described when running the test twice +Another solution is to just replace the Cocoa import with Foundation +My two cents +I suggest we are not going to do more work on handlers until more people require them +r adds debug logging output for situations where content itself says it should not be indexed. +Im looking into it +Fixed in Rev +Committed. +Is this still an issue in +I just tried on Redhat Enterprise Linux with the IBM JDK and I cannot reproduce this error anymore +Glad you mentioned the here Henrique +Applied to branch At revision +Note that this only impacts some of our projects most projects work fine +its not actually a parameter in the request and maintaining it as a necessary requires future versions to preserve itI remain opposed to adding a string to the query to be logged on the remote side +Patch for the described catch thanks for the patch Ive applied a slightly modified sure if wrapping is really needed it seems that at least newer slfj versions support array types however Im pretty sure that older didnt +If theres something you dont understand in my pull request please ask +Applied via Jira +If there are more use cases in the feature we can think about a general mixin or what ever +HiI am in China until June nd and will only have intermittent accessto email until thenThanksBo Bo Shi +Replace check for init parameter with check for are still seeing this issue when view root caching is enabled +bq +You would not be able to reference until the release is out unless you did a local release on your PC +Committed everything except these two files +Assuming an is required for this type of data source it should be considered down and the VDB made inactive no? It would save the user a lot of time +So unless it gets added we need a copy in libjars and a reference to it in +Hiya there any way to configure Hudson so that it will always send an email? One of the benefits of receiving the emails is that we are aware of changes shortly after they happen +Thanks +Can Transifex support this approach? By the sounds of it its a manual step so we can massage the filesystem to workIf I understand correctly we are not actually buying into the Transifex service +Closing this JIRA issue as per Michal Fojtiks comments and test output copied above. +Could you try it with more iterations +applied fix to trunk and merged to branch. +Patch for removal of the facet entries +Please review and apply it if its acceptable +Sound okay? +Declarative services no longer needed +Should be part of the server information already +I updated Pax Web features XML with the features namespace and the usage of the OBR resolverI will update in Karaf now +POJO src files +patch for early commit of this issue +Reto +Thanks a lot for your help +Definitely looks like a bug to me +Lets see if our users need more convenience functions. +Add the to applied to trunk in Rev +The builder uses to preserver the order of children so it is when we implement say resolver or another Java resolver this is something to watch will leave this open so that I can permanently fix it. +Im struggling to find where this info should link to in our enterprise docs +Thanks! Committed to and trunk. +the crappy test script i wrote to confirm the lowercasing behaviormaybe we need it again if we go to java in this decade or maybe i screwed something up +Committed to trunk. +ok it seems it is possible to test it this way +? Using the last item in the glob expansion or search in reverse order should generally get the latest versionFrom googling it looked like was a stable path on Solaris ie no glob necessary +My fix for introduced this issue thinking on it +An optional boolean init parameter httpCaching has been added to +I made up for it in revision stencil mask looks lots better! Awesome +Setting back to Closed Done having modified release note text. +Hibernate and Spring ORM and have not had any problems +Ill look into them +I reverted back to the original patch and added a new test to to check that the output from the XML processor includes the fields labeled by DELEGATIONKEYID DELEGATIONKEYEXPIRYDATE and DELEGATIONKEYBYTESAlso while the OEV tool does have a binary processor to create the edits file from an XML textual dump the OIV tool does not so for my understanding no other changes are required +Updated fb to look for and netinet +You are the authority on aa so I will trust your judgment on this +Patch in fixes this +Hi SagaraI have tried it in and it behaves the same way +I also changed the field to be called +believed fixed in branch but needs verification if still present or in JBDS +Hey Jeremy! Looks like a misconfigurationOops that was actually my dumbness actually contained data from nodes log +There were some comments in the removed code This actually happens +fixed for inclusion in version rev +Committed a patch to trunk please close if satisfied +In many support scenarios in the past I often rely on offline compress to fixup problems in the database +Rebased v +Sorry I had the window with Attach file opened and I didnt see your commentAnd sorry again yes it is a duplicate +Not too much differences from the previous but I did it so I think it is better to talk about the latest +This patch is against the source tag +I committed a patch that works when I run the test individually up on hudson but as part of the general suite it continues to fail +Hmmm closing this. +Note that the scheduling mode can be fairshare or fifo mode still +Thank you for your contribution! + site cone be move in Dungeon labs for sure +If the slightly misleading name of the is to be changed than that should be done with a dedicated IssueOf course. +Ive updated the patch +the latest patch looks good +JDK users will now no longer see failure however they will now encounter +f +Unfortunately the pmd plugin project released a new version that breaks my integration +I think with the changes the modification has sense +Apparently it can be worked around with you Joe +Hopefully a fix for this issue +It doesnt used the right APR +Changing attribute visibility in will only affect applications that extend this class and use the same attribute namesIf I am not mistaken Java allows of attribute names in derived classes and scoping will make sure that the behavior remains the same as it is nowIf there are problems with name clashes the workaround is to rename the attributes in the derived classesBased on the on existing applications I am in favor of this changecheers Roland +This has been committed as a part of +Oops +I dont think the administrator should be required to clean up the dir before being able to set the quota. +It seems the implementation is not very hard we simply just add more helper functions that could handle the request to snapshot a subset of columns since snapshot is snapshotting columns one by one alreadyOne things worth discussing is how we support specifying both and simultaneously via in CLIWhat i propose is we add a c option which will accept an argument as column names +The package commands have been superceded by bundlecapabilities and bundlerequirementsGiven those are more general and can display generic constraints rather than package constraints only I dont think we need to add back package commands +It would be better solution +tests pass +Its in thanks +for the patch +It will be good if you can provide a patch that remove all duplicated you mean duplicated code in the new files that I created or in the files that are already a part of +So please ingore the testcase otherwise it will fail on linux +Hmm I shouldnt have stated an absolute like that +There are many things that can cause a CPU to go to utilization +replication is not yet supported hence the behavior of what I like to call replication event cycle +Fixed with revision The test case for the repro that is attached to this issue can be used to test this fix +Hey Markus I am closing this off as I cant replicate it +Potential youre right mislead by the name synchronized we assumed that the listener is not called concurrentlyIf applied a potential fix which creates the instance on demand and also synchronizes the set in the bundle listenerCan you please give this a try + +librecordio requires xerces +RegardsJohn JohnI have verified that samples build and runs fine on WindowsI believe you need to address Sarahs comments about location of the filesPatch looks good to meYash patches have been applied to cvs +The dist target in langcsharp creates a binary tarball in the dist directory and also uses doxygen to create API documentation in the build directory for inclusion in the documentation tarball +The committed change is going to break all the context filtering logic in +Thanks a lot Thejas +There are still a few broken links left that I havent been able to patch is for +Now you can use BASEDIR or REPO +Tags +It hangs when it reaches this test very often +Youll see patches put up with nocommit in comments patches that dont compile etc +reopen for backport to maintenance branches +Fixed in revision . +Manually tested! Thanks Alpar +Also I noticed theres a lot of ugliness in which I dont want to clean up in this patch but probably should be done at a later point +The point is to make patches maximally easy to review so that reviewers can easily focus on the changes of functionIf there are lots of tabs in the codebase then someone should file an issue for that and then fix it in one independent sweep +pushing it to a blocker +In the build output twhen running here is something a little strange though I have no idea if its a problem or notThe target produces the message EMMA collecting runtime coverage databut the target does notJust seems strange +This may be related to +The only reason I left it open is a lack of real testing in a clustered environment +Thanks for the reviews Stack and Ted. +The proxy would have to know the instance secret for it to be able to write its location to zookeeper +Although a bit late Id like to checkin a fix for the upcoming release +I have reviewed and approve Dave Brosius patch +how do you create multiple transaction states per connection? I agree that is plain wrong a connection is meant to be used by one and only one user at a time +merged changes with into all issues that have been in resolved state for more than one month without further comments to closed status +We can address this during the next SOA platform release as appropriate +Yes its also a possibility but by this way its just generic for not for other +Lets commit it +Hi Ed +This improves performance and paves the way to implement Avro container filesThis patch has been tested on Ubuntu RHEL revision after adjusting fot. + +Closing. +I will fix it +Sure added the header +Thanks for the patch Doacan +An issue I can think of is that the compilation unit source is displayed to the user in a view or editor but I dont think this would happen as I believe Documents are created from Files not other issue is if Compilation units are created in more then one place but again I think theyre just done in the other benefit of this is that I can have our own decorator to display the easyb files with different images in the package explorer +The proposal can beimplemented any time now +Help in resolution is appreciated +Thanks Karl this one missed my radar +is this meant to be a Groovy issue +NET SP ve just changed laptops wasnt installed on this ll update the various to indicate that requires +Does the documentation of this component claimpromise any? I think the best could be to ask for other commiters opinion as well including Richard Kettelerij how denoted this component to the community +The attached patch enables client for +to commit +Not very nice if you ask me +Application dont see any EJB involvement in the stack trace. +Sean is actually an error that is the result of specifying dimensions when they are indexed in the input file as when the program is expecting +Also at this point its more of a design specification anyhow +Fixed by synchronizing DOOMs addChild method with that of LLOM. +Ive just committed this +Sample project that works for me using Maven and the latest of the Site plugin. +However I think many useful CP use cases must have a table access scope wider than the single regionWhat we need is a lightweight server to server IPC? Would present but lack the fat client semantics +Sorry forgot to include the revision within which this patch has beed commited revision . +If working on the backport for this issue +I dont remember the number but we already has it somewhere in JIRA +Added solr qparserplugin +See the commit log for more details +Can be replaced in demo by a basic can be deleted I think too +Im not seeing any emails from the runs if the new CI is in place +Please RC is very old and and have all matured greatly +Parsing nested imports from css files was added +Hi RemyI dont understand why this was closed as Out of Date +makes the hashcode cache final and to be a concurrent hashmap +Updated and etc to use new name +When one of failed namenode didnt close editlog and unlock the storageAre you saying for the failed directory namenode didnt close editlog and unlock the storage or is it for all the storage directoriesIf it is for the failed storage if failure is due to a disk going bad in most of the cases we may not be able to do these cleanups +For this change would be costly for my company as our standard is to use dots in all our artifact names instead of dashes +FYI a single document ends up on a single disk +Committed +Fixed. +Percentiles that match included test case +Anton Nadya thanks for improvement new doc +So Ive decided to go with a reflection +Ive just committed this +Solution file to add +Any update on this? It looks like it didnt make it into +Changed file now appears in latest alpha docs so closing issue. +that would work provided the security subsystem is not hiding attributes otherwise we have a security the way it works now is that when the user has a custom the underlying feature type is never accessed so more or less ignores security +Attaching a patch for version +The documentation for has the incorrect order of arguments +Nice! Lets add it to the mix Im guessing Jasons is quite a bit faster for splitting segs and this one nicer in that it can split indivd segs +Committed in rev together with a test case +ok but how can I test this JS to be sure everything is fine? Because groupId in mavens repositories was changed from pdfbox to and this should be tested. +While the current set of active committers may not be looking at these sources I am sure that there are people working on it +t know why but the is not being garbage has a reference to which has a reference to the entity classes +Can you attach a command line client that exhibits this behaviour +Hold for new patch +I think this would be a great way to shrink the connector code have more consistent behaviors and share much more code +do you have a complete unit test case for us +The patch with the fix plus a few unit tests than should send the correct MEP when starting the in trunkSending srcmainjavaorgapachecamelutilSending srctestjavaorgapachecamelprocessorAdding srctestjavaorgapachecamelprocessorSending srctestjavaorgapachecamelprocessorSending srctestjavaorgapachecamelprocessorAdding srctestjavaorgapachecamelprocessoronexceptionTransmitting file dataCommitted revision +Is the ping thread not aNo +closed as of release. +I looked into your code and found you were using and hi not and hl +PLEASE DO NOT REMOVE THIS VARIABLEOozie relies on this variable to be able to dispatch mapreducestreamingpipespighivesqoop actions +Basically if we search segments created from NRT updates with this approach and search merged segments with the existing approach then we should have performance pretty close to what it is today with the benefit of not having to reindex the row for every record mutateThis approach has main problems to be solvedThe first is the ability to do merges and colocate the records for a given row during the merge +Im not sure why this doesnt work can you take a look at it +thanks for the further information +bq +does not populate the row key for columns after the first +Would be nice to have this done transparently by spring social +Will try to look into this to see if the problem is still present +Hudson will bomb without the new unit directory +this is consistent with the m is a classifier coming after the version +I removed that method in this new patch +An INVALIDDNSYNTAX must be thrown instead +Attached diff from top level +Hey Andrzej see super on this +The updated can you submit this as a patchdiff please? I could use this myself so Id love to commit thissvn diff bBt +To clarify modification this was done two ways +v updates QRH arguments to use responseCount as well even though its. +The reader scans from the start of the file until it finds a valid sync marker and then reads in the schema +This issue have been fixed +Closing again after addding screenshot for Burr. +I am not sure if the state information is loaded from the cache or is persisted to db and loaded from db when failover occurs +Since this limitation has now been documented I intend to resolve this as a not a bug in a day or twoFeel free to file an improvement request for this particular query form since Knut has established that it might work +Thanks +Good idea just add a note to the method javadocs that you have to override this if you change the contents by the filter +Java classes scripts and sample xml files and schema to reproduce problem +Committed revision Committed revision +Im not too happy with the direction you took in the latest patch + adds +Its been broken in previous milestones as well +Bill Thanks for looking into it +No big deal but worth to mentionGeronimo is being rebuilding +Good point LarsA discussion on devhbase +Thanks ErwanYour patch is in trunk at r +This is intended behavior +reopen no proper fix version nor fix comment +Closing this record since is now released. +Clients can catch this retrieve the Errors object analyze the errors and presumably show them to the userOther exceptions that may be caught by the presentation layer are mostly that are caused by programming errors +looks like this issue does not happen for current +Heres an early patch which shows what a more fluent interface looks like +And for the user side the easiest way is differentiate the error based on the different exist code if we set different exist code for different types of error instead of just simply throwing the exceptions IMHO the most common case for this API is to make sure the application is no longer running and the caller isnt so much worried about the exact final state as long as its a terminal state +The new patch should solve this problem. +this is a patch file that for run on Hadoop. +Please confirm it has been applied as expected +It batches the state changes for all partitions to be able to batch the leader and isr requests sent to the brokers +srtemppidstatusresolutionsorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornoneGET srtemppidstatusresolutionsorterfieldcreatedsorterorderDESCsorterfieldprioritysorterorderDESCtempMaxresettruedecoratornone HTTPHTTP message The value PROJECTKEY does not exist for the field projectNot quite sure how to continue now +Im still unable to reproduce this issue +Ahhhh +That works Eugene great job! Installation now works fine and the plugin operates as expected +Whats your external verifier please? Ive recently switched to groovy and dont want to come to killing anybody +Thanks +I made some changes in this area related to datanode exit +Sorry I have to close this as not reproducible +so on bit os instead using the property directly change the rpm name to x +Ill try to fix it +I cannot reproduce the issue +I will remove it and submit a patch +The compiled them automatically their result ending up in the dependant projects output path. +that looks better Im working on applying ittesting it +Now with a last small fix preventing an NPE when trying to map the root path in Rev +This is not quite as elegant however and I think the requirement to run code generated by with is reasonable +Fair enough +I still havent been able to test your patch yet but as long as you removed EMPTYDOC Im guessing it will work +Fixed the unit tests +The only way to survive is to cancel that background job If I access connection again it asks only once as it should +Thanks Tim +Looks good to me +Once you verify that the pull was successful can you please close the requestThanks +Verified status of this issue or fixed or no response maybe can be closed +Andrey Ill aply the patch for later today +Also applied to trunk +deps +Incoming commit to lay the ground structure +Because of the low frequency and complexity of this as well as the the the amount of design and development involved pushing this to a future release until changed by a community or product priority. +Any updates on this issueRegards Any updates on this? This affect Mina and I am going to proceed with the above fixRegards TrustinAny updates on when this will be merged to and code baseI can get this fix in if I have the permissions to change the Mina code baseDo you agree the fix is correct for this issueRegardsve just checked in the fix +However we may have better design for that +I changed the JIRA title +For CR can we you manually update the fileFor I hope we have some time to devote to tooling related to creating configurations +I understand now +Committed at subversion revision Touches the following filesM javaengineorgapachederbyimplsqlcatalogTurns off the istat daemon by defaultM javatestingorgapachederbyTestingfunctionTeststestsstoreDisables the istat tests +mistakenly forgot that sanity check back in previous patchYes that will be required if n is negative then underlying stream can throw IOE +Nothing to do. +Thanks for addressing the feedback Sid. +I didnt have to drop and recreate the index for it to work +There is no option in a standard XML parser to not load the DTD specified bythe XML file since it might be used for entities and the like even ifvalidation is in use +I notice that some would have it as far away from Sling as possible but not me and I think its not for Slings sake +Falko but my assumption was that the modeler would not be updated a lot +There are two equals methods in one takes only a table name argument and the other takes both table name and schema name arguments +features yeah seems like an easy fix +Rumen should be packaged with Hadoop +Wow +A native hadoop way of doing that is required +Thanks. +Thats the missing link! Your patch makes sense for that case thanks for it +Bulk Close for all issues before incubation +This is going to be a know bug in +looks for attached +committed fix in revision transition resolvedfixed bugs to closedfixed +Good stuff +Sure could be another type of coprocessor +Passes SBT tests etc +I just committed this +I believe this is fixed +Id be tempted to as for and wait for that +This patch leaves the default behaviour as it currently is but is a boon for a paranoid administrator because heshe can rest assured that binhadoop namenode format will not work against a NN that is configured with this option +test dependencies to am using Eclipse with the Mobile Tools for Java plugin and have the same problem with my compile scoped dependencies not beeing marked exported in the + +Thanks +JamesIs there progress on a reduced case? Does the problem persist with and the latest version of servolux +Releasing takes ages because of this problem +Youll get a dialog that allows you to keep waiting or to close the dialog +Not believed to be a regression relative to GACP removed CP in fixed in field +Heres a patch without whitespace problems +BUILD SUCCESSFULTotal time minutes seconds +cd +Implemented the changes Ivan suggested and replaced Unsafe with in the new API calls +patches for each of the for comment from author and will push these patches +Fixed the problem with we needed to disable the nested map and list support for the used in the test now we actually can +Hi Justin it seems to me the attached patch fixes the issue +I always get a lock timeout instead +crazy formattings +this would be super +We should rather move all error reporting back to the native english logs +Yes I noticed the latest patch is reffering to which isnt present +The other part of this is to make the approach also agnostic to object layout +I would prefer to see this one left in even in its unfnished state +Should apply cleanly to the current trunk for +We may expand this property to turn on tracing in general in the future +Birt Plugin Reference Guide is checked +I did try to disable the integrationTests in extensionsjcrinstall but that only took me too new failed tests in wheretheres seems to be trouble with again +a Party Relationship Employment between Party Roles FROM EMPLOYEE TO INTERNAL ORGANIZATION values of these two fields are defaulted to these values and didnot show these fields on form +I agree with Marcin I just posted the second test case to show the exact behavior of MRI +Whenever you set new Website ignoreWords or referer spam words Roller will clear out any existing referers that match. +Ive also found that a hint needs to be given for the color space of these images +allow users to see only related plugins on the custom server assembly portlet +I guess this issue is happening since returns all the functions from all the registered python files +Martin Thanks for the quick reply +I think the idea has a lot of merit do you have any examples of utility classes out thereMy experience was that every project finds a better parsing tool Domj jdom xom xstream etc etc +bulk close of resolved issues. +I now believe much of the existing JVMTI code will work just fine with the new thread lifecycle design +Verified by Ilya. +HiWould you have an example of the JSON document containing the class field and datadavid +svn ci Sending srcmainjavaorgapachexbeanfinderTransmitting file dataCommitted revision Didnt get this applied for the binaries up for vote now + tsuna on our staff found that in a string intensive processing he was able to speed his app up by a factor of by not using that API +Actually rather than its +on +I thought it shoud be related to the screenletI will wait for a solution to the issue you raised before committing my changes enhanced by your tip and something for Shipper Account screenlet too +Thanks +Rejecting as the behaviour is correct +really tried but cannot reproduce like this was already fixed recently. +older issues for Apache since were no longer actively working on these at the moment. +Im using exactly JSF +Probably not going to get a fix in for so Im bumping to +I rebuilt with the latest trunk version and everything works as expected +So I would go for option +It may be possible to do this without any change to wise andor smooks +sorry I can not preserve code formatting for several classes cause I change much in where +Martin you were right about they being different but still rightBut my example was not goodThis new one shows whats really wrong +Committed revision for branch + provides the basic infrastructure supporting validation rules +Sounds good although this is certainly not critical +Is that a different feature or bundle? I am trying to install the ode feature from mvnmxlfeatures +Im resolving the parent one unless there is extra work to it. +I remember ssh failing if it finds a world writable +I have recreated it +Yes everything is dynamic from here on out +This containsuseful sample table functions for making a table out of a property file making a table out of an XML file making tables out of subversion logs apache web server logs andJIRA reportsIn addition this demo provides machinery for declaring and refreshingparamaterized subscriptions to data from foreign SQL databasesTouches the following filesA javademovtisjavaorgapachederbyDemovtiscoreA javademovtisjavaorgapachederbyDemovtiscoreAncestor classes of most of the table functions in this demo +Havent looked through bugzilla for this yet but I will +but we no longer close the writer on commit +James C +Allen any updates here +A patch is attached +PRD has been updated +Chris my comment was based upon the fact that were talking about having people move over to Firefox instead of working on fixing the real problems which are on IE sorry to not put enough context in there to make it clear +svn mv coresrcmainjavaorgapachewhirrservice coresrcmainjavaorgapachewhirrservice +Ive annotated the two patterns on the MBs graph +Thanks Erick +A new version of this patch that builds on rather than ignoresWhat Id like to see is a code review from stack Billy Pearson Bryan Duxbury and Stack run this patch through their bulk load tests to ensure that this patch is better than the current code +Reopening just to edit the release notes text +However I think we should do a measurement of the performance overhead experimentally to validate the theory +Theres no reason to split the optimizer over it when the validator can handle it +if you cant scan the whole table then this patch gets more complicated I believe +Uploaded the last patch to a new review request +I have committed this. +Sorry to be a pain but the names of the new figuresFigure +And Integer can not hold that long values +Version has been releaseLets close this issue. + hand off tests passing for ER +not configurable via a setter +by r the change is applied to vbeta treeJulie are you sure in netui package you want to build coreWeb again +I dont think these should go in the normal bin directory but I think including them in the distribution tarfile might be good +We have the Orders quickstart running as part of the release build in AS so I dont think this is an issue any longer +Some useful log observation from INFO Removed completed container container INFO Getting for container Seems like after container is removed from tries to fetch the status for the same +Nothing to verify here. +see comments above. +Here completing a task means ending a task andDo you think that also fixed this one +Just have filed for testing +No response assuming ok. +Guessing in which jar the specified is included in is a trickything to do +All unit tests and tests pass +I want no dependencies on commons jars when I am done and I want to reduce our footprintThere are also other issues like having too many projects all over but this is another issue +I think this is closeThanks +Thoughts +Many thanks for contributing a test case for a complex problem +In addition to the virtual EAR functionality that Marc described we are also interested in the functionality currently described by this feature request whereby modules arent necessarily classified by their extension +Cant reproduce this using latest installer version +Again Im fine with posting a version of the patch if that helpsThe currently attached patch is against +Its odd but this catch Exception block is necessary since reflectively called constructors in a implementation may leak a catched Exception even through a signature that doesnt declare it +You can work around this by explicitly calling zkupdateAliases +Applying Matteos patch all tests passthanks Edell +Having it as data member could confuse? Looking at it it could be hard NOT having it as a data member + findbugsThese warnings should be ignored +The has been updated to no longer user and it uses +Nicolas the online doc will be updated in the next days +I will be dealing with this through the child tasks. +This needs more thoughts. +oops confused with crawling downloads +One way of doing that is by implementing client quotas and shedding requests coming from the faulty clients +User functions written in like this issue has been dormant for a while +Closing. +Thus it should be a configurable +The issue seems to be related to the way in which defaults in node type definition are handled + +Proposal of patch partially fixing this issue +it now supports reading avro files with different schemas so long that a meaningful union schema can be derived +Todd which varint are you referring to here +should be encoded with byte other characters should be encoded with sequences of or bytes each byte in sequence belongs to range +bincassandra needs to be tweaked to supply the right class name but that is trivial +Updating patch Adding unit test case to verify totalStored Semaphore usage is correct +Ill commit after the unit tests run +May I suggest you make a reviewboard entry for your patch? If you havent used reviewboard before you need to create a login and choose for the repository +If DFS wants to ensure that the instances that it generates always have a Path then it should construct them with a Path no? Until is fixed this will make mapreduce slow +Attaching a patch +lets bring it in +You think we need to fix bloomfilters for? Im thinking not +The problem described in this Jira issue has been resolved +As I already said from the example supplied it is only apparent that the SELECTED event might not have been fired +Somehow I did not think this would make any differenceThanks Amareshwari for looking into it. +This issue is essentially a duplicate of so I am closing it to help organize issues. +And My mother Language is not English so if I wrote something wrong please forgive meI downloaded the updated makefile but its still not work for me! My configuration is os x Xcode I compiled the source code it reported was not found +Its odd that its not in the version that was created on th August and at that point the empty check was in but it was not in the first version of +See attached screenshot +The div node unneseery becaise insert an extra div +Sounds like there are othersI agree with the boss who wanted you to use deprecated machines +Hmm not sure who relies on this behavior but it may break some code might be better if the code path was changed to check of the parameter exists before overriding should have thought about like I dont have access to the request or to the parameters from the +I think this change doesnt need tests +Thanks for spotting this Brian. +It has problems with stability +If a data node w a partial data loss can be restored wo copying in the data it has not lost then avoiding striped LVM is a huge win +Attaching patch for +Committed to trunk +yepas a result of Yoniks work we now have a close for Solr +I opened to track adding Subclipse to Central in a future JBDS release. +usrbin +This patch fixes them +Exact table name breaks functionality in embedded client +thread dump of frozen Eclipse this test used to work before migration to Luna M? I mean are we sure its related to target platform +My reasoning for this is that it is a lot easier for me to keep up with other peoples changes and a small patchApply before any of the following +Changed the code so that the jcruuid property is stored +OK this is solved in SNAPSHOT soon to be M +There used to be places that filtered anything with a score however +This reflects the following changes which were introduced during the last couple weeks Authorization is no longer required when the server installs a security manager The default has been changed to not have a wildcarded port spec +txt from the file name to using methods to remove extension +Just an idea should we default to this implementation in trunkIt seems to be a significant reduction in RAM +Could you test against when it comes out and I can schedule the bug fix for before + +Would be interested to hear from the Yahoo folks on this one + +Ill be at the contributors meeting +Also its to have zero errors on Android LINT since everyones Eclipse setup is slightly different +We need a more sophisticated classloader to sum up + issue caused regressions in test cases +I think its ready to be committed +The branch does not +should handle that exception and retry after sleeping for a short while +Looks like it was that changed it although I doubt it was appreciated then that was broken and this fixed it +We should now work on that branch to fix all the remaining issues +Reopening the issue since I forgot to add the test to the +moved target version to upon publishing RC +applies to batch processing onlyFor single action individual exception is used + Test is added +Can you add a link to an example for ? I dont think I would like to introduce Deltaspike as dependency so we would have do something similar or go for the simple property based solution +given how many unit tests are in the same package as the internals a lot of tests will fail and someone would need to go through the results to determine which were false positives +Tested manually that it works +Here is the new patch +The pdfbox project is suggesting that because this is a bad font we should try to precheck the font with +This patch contains the test case for this JIRA +I think thats a good idea +Differences from memory You are creating a new class each time through the loop I was reusing a single instance +this an overview of the raster +I am n this patch too Thanks Daryn Ill check this in for you +Thanks for the review guys +Attached diff contains the necessary plugin changes to understand both single and split kernels +Modify the file for the addressing module to include the new Change all files that I caould find to remove the old from the default configuration +was told that the config changes will take an hour or two to reach the mail serverthanks +There is a bunch of code about memoryaccounting in the regionserver to make this not failWouldnt we just be centralizing and possibly bottlenecking the sortof the hlog to machine when previously it was done by all machines? +sobq +Now that work is underway with AS all previous community releases are +The use of to quickly test a questionable service is associated with jira This issue is just dealing with the current undesirable connection betweencas pool size and timeout valueYour description of increasing the timeout by the number of outstanding requestssounds reasonable for setting the time value for the request message +Ill take a +This compiles for me +Big data does not need shared NFS file systems for unit tests +Added service in +See for how to add the attributes +Haha yes +is already open for the inability to use special characters in passwords. +Attached a patch that solves the above described problem +Patch committed. +The anonymous emails would go via the mailing list moderation facility so it shouldnt casue any problems should it +thanks. +Fixed with r in and r in trunk. +In the patch you write top Creates a that requires in order docs did you mean? Because is abstract +If you can reproduce this please elevate to critical +Bug status payloads added by this filter do not get written to the index! status forbidden api checks fail +Not sure that this was exactly the correct thing to do +You cant really make assumptions about what kind of data will come in on the query stringI agree that the bug can be closed as invalid if the preferred configuration method will result in these exceptions being caught by my exception handlers +But I checked hrs ago for Rajesh and on a system where the install didnt proceed the owner was rootroot and on another one after rpmtest the owner was jbossjboss +JoergHow do I submit the patch to Forrest do you know? Its not listed in Bugzillas Product list +This is lucky as I cant find a way of importingexporting the codestyle +If it isnt doing anything useful right now Id say move it out we can always move it back once someone wants to get serious with feature schemas +Thanks ChristianI could not reproduce the two errors you have +No IT +Both of these new events will cause logs to be garbage collected + for the patch +derbyall passescleanly for meOne question though I see that there is code in the patch which runs during the upgradepath +doesnt compile? Exceptions are moved to +Is this committed and fixed on trunk? Or are we pending more changes? Im trying to get a handle on release so we can get a release out soon +I need to alias the field name for my class to package upon marshalling and I do not want to directly use the instance of +Attaching patch that limits the string size of the Pig counter name to less than characters +Using the new generator works for me +Im working on the feature that the user can specify when this reduce to one line should be you only like to get the Indentation to be fixed use the specific function for this +whoops that was the same patch as before here is the new one +Ive fixed some bugs +Odi are you planning tohave a resolution on this one before? for me +Do not reopen this unless you are using and still have the problem +from me +Tagging features we could implement in the near term +Ive tried a few times now to rewrite Range to support inverted and thus be a replacement +I have requested accounts for those with recorded iCLAS the following remain outstanding Paolo Castagna + +The internal model of the project is listening to workspace and updates but it takes time the update is going as a background job +Seems like the automated tools are able to upload patches +I wasnt expecting people to use this stuff while its still unreleased +By the way we are looking into this currently +Great stuff +mountMap does not seem to be used anything other than for counting capacity and bytes remainingAnother small patch to include Tb and Pb for bytes reported by dfs report +But it is reasonable to have a message key that translates into the version string +This is caused by people reconfiguring their space permissions to allow anonymous comments +You mean like a PDF document that can be downloaded through +Thomas one thing that bothers meShould we return some sort of error indication back to the AM in the case that it has been stopped? Currently your patch just returns EMPTYALLOCATION +The first is a security provider the second is a model +Matej can you also do a quick look on this classand maybe add the js functioni now build up in the javacode into improving it so that it walks recursivly through the childs of the radio or check containerIf this all pretty much work we could integrate it into believe you are working on this now +Its OK in the current AS +run +the search extension into its own module is the best we can do right now +Is there more I can do to facilitate this? I know I updated several times I apologize for not providing cleaner packagingShould I reopen this ticket or open a new one +I dither I preferred both unique and parent prefixes for so I renamed it to be +for user settings opens this with maven in my eclipse logthere seems to be no connection to the preference issueattached log fragment messages in the maven console for setting or changing the user settings file fix to trunk +I think solving the issue in the +Okay Ill just adjust the documentation for now +This patch solved the issue I just posted +If you can repro with a new Cordova project please +If the only implementation will be in and the only caller will be in then why not just include this logic in? Have I missed something +This seems more appropriate a suggestion for perhaps as an addition to the class +We wouldnt see the warning because the exclude would hide it +Weve got a case with a custom session registry implemented using the database to store session information for a stateless load balanced deployment +I mean I could write to my documentation for the users of my web service just do not change the default in Axis for the dot +hibernate tools plugins have now been updated to use hibernate and related entitymanager and annotations. +then the other loaders cant seeAnd he only reloads those classes +Verified resolved in CP CR +I got the exact is still more to fix +I will first test and commit this to the branchThe patch also applied cleanly on branch however right now I am short on testing machines so I am not able to test on yet +If it is it might make sense to commit this patch and then make additional fixes in smaller patches +bq +That would at least unblock you and give us a place to plug in the Commons Math code later +It should now report the proper error message +Still though I think we should document these dependencies inYeah +When the user attempts to save the server and if the user does not have valid selections in the combos the editor section will reply to the server editor framework that the selections are invalid via an Error status object +How about I take s patch and mark this as duplicate +My comment to Poul was based on comparing this line to whats in the Java version +The referenced images are corrupted +I will commit it later today +attaching proposed the issue and the patch are based on the assumption that the jcrcontent node cannot have an alias + +The emitter can always chose the right emitter based on the listening mode of the listener node +We had a holiday in Germany yesterday + has been backport to the apache branch and should be available in the FUSE ESB branch very soon +Ostensibly doing this would make managing poms easier in a few ways be it tearing down and rebuilding a prototype from a or file also it would make usage examples more easily +vs +Hi AidanCan you review this change +Alexey Seam Beta is available but only in community productized Beta will be available hopefully this week +Does any kind soul out there want to translate those curl commands into a JS test? Im not sure when Ill have more time to devote to this in the near future +I spoke with Tom over the weekend and he reported that on micro benchmarks the performance impact was only a few percent +For example we could fire an event to signal that a tooltip should be displayed +Patch is attached for an early solution +Your latest patch includes the deprecated styles I eliminated in my patch +Running a right now will post results shortly +Do you have this patch adapted for PIG apparently it does not apply have tried PIG trunk patched with without success same behaviorThe patch does not apply on PIG did you adapt it +This is the version that will be used by upcoming is going to be a part of Apache POI +Vladimir fyi this is with an installer built from the src days ago +Looking forward to getting this +txt and +I did a trial run with and nothing blew up +Did you try? I am almost sure I fixed this in trunk +Should we add set to? If so we need to make honour it +Please verify the fix in the latest dist contains updates +Committed. +Closed with an Invalid status as it seems more like to be a user issue than a lib issueThanks for the feedback! +What do you think +I have the same issue +Ok the sub is gone as of revision +Committed revision not reproducible with +is an example +Nop we didnt set the delivery mode attibute on the producer which means according to the JMS API it is default toEvery message is successfullly persistented in the local broker on the producer side if the network is down before producer sending activity beginsBut if network is down when message is transfering between two brokers there is always message loss +Thanks Ahmad +Manuel thanks for the patch +But im not sure if we should raise the count +But I dont see errors related to spawning server +Ok +OK so it will be fixed on the release +Everything compiles fine for me nowThank you too! Greetings is ok but isnt +Add it to at session scope with the name tableTest +any objections +Thanks +Moving feature tasks to LATER and unassigning them +Until we somehow manage to unify the entire tab group situation which we are nowhere near doing yet this is working as intended and is the chosen impl for the suggested future it works as expected then +Thanks +Maybe if there is quorum on this we could add the support to the extended tdataTable +yes please close. +solved in both trunk and beta in JBT. +We do anyway so its just a matter of making it always +And Durable Subscriptions are singleton in JMSThis would be an issue if using directly multiple threads +if no objections i will commit it tomorrowThanks! I already used it in two places here +Updated patch with better tests +It this fix being backported to the future? It seems pretty major since theres no workaround if you want to test Ajax form components +Havent been able to make progress on this +Thanks +The pull request for AS is in to incorporate and upgrade to SASL this issue can be scheduled for the first EAP version that will include SASL or later +It seems the code changed a lot because now in trunk post throws an IOE that kills the server +The latter produces a warning but still generates the correct code +Attached patch that overloads set to take the class name of the factory as well as an instance of the factory +Tom can you add Sorin? +Updated Jelly SWT image var and resource patch +There is no scope time to resolve these remaining lower priority issues for so moving to final only issues considered blocking will now be fixed +It says explicitely that the map must be synchronized +In other cases people might want to do it in the or even the +Thats a question for the mailing list not the Struts bug tracking system. +There was one map id for each split of the data so fork splits there were k map ids to know about +Basic proposal replaced with proposal as submitted toAny feedback would be greatly appreciated so that I can continue to tweak to proposalMany Thanks like this was accept to nice +Yes +I need to spend more time on this do not commit this patch or run it in production +shows and hides very quickly is by designIt is necessary to block the whole form until changes are savedThere is nothing to change from my point of view +I vote for this improvement +The fix Id like to do is to improve the reflection stuff in these operations to be able to pick out the primitive types and support them +Actually took bit longer to get to actual reported issue code was assuming bean or container serializer but must also accept +This was committed to trunk a while back +The file was not checked in for some reason +Task completed by Mads Toftum. +Addendum patch +hmm apparently trunk moves faster than I thought +I investigated the dependency manager to see how it handles columnpermissions +Ian do you think we can get this applied to the branch +changes avro version to in fails with even without this patch +video options are working fine but screen sharing I am getting only mouse pointer movements on black background +This is a reasonably straightforward port of the analogous Lucene class +Issue closed due to the release of Tiles +And thanks to this the component seems to be the last component that gets shut down +Seems to be a duplicate of +Mahadev those tests should likely work if JAVAHOME is set in your environment +I would assume we want to include the latest community build which would be +Catching and in would work as well +The can be a different bugPorting wise the reverse will probably happen this implementation will get ported to. +Simple patch to cache gems during ant build +Might be related + The diff file the diff file and the new manual page +Will test and commit +Thanks for looking +should the following be true? quotient quotient +And all the issues quoted above seem to be characteristic of the job in question +I had a look at the taglib reference and there is no mention of this +Thanks +Perhaps Jack you can look at the Aperture code and post exactly how they do all thatSecond its not going to be terribly efficient to download and unpack an archive repeatedly to extract its contents one item at a time +See for my usecase and a patch against that allows this +Please find attached a patch for this noticed this issue actually happens in a few more files +ant passes +fixed in revision close of resolved issues. +Due to the Alfresco roadmap this improvement is actually stoppedWe are waiting some informations from Alfresco to understand how to use their new client REST. +is used inside of to track what tracespan is currently being worked on +The new sendEvent method would generate an if the type does not match the configured type or is not a subtype +The thread count is memory have attached screenshots from jconsole with your example using Camel with the thread count is stable between live the memory is also GC when we hit about the mb will get backported to the Fuse Camel branch and try with that as well +Also for this issue there exists a trivial patch +Assigning to Adrian as the original author of +This may have changed of course +Some integration into Hadoops documentation would also be good +I agree with Andrew here +Added a test for it in r. +whats the status of this one should it be addressed in or can we move it ot +Personally I feel that it is simpler without the inner classes +Anyone else +Schema and instance file demonstrating the behaviour +Agreed that its not committal but if its his deadline and he agrees that pushing it back is reasonableNo need to be paranoid though +EAP in ER +during that the testcase try to get the online region provided it will return which lead to the case failture +Now can you please retract your veto on this issue? Thanks +You can send a signed copy to secretary +A first think that we are together working on Flex technology which is not OSS and FDT is the only existing Eclipse plugin for FlexBut I understand your point of view + +Will this get fixed please +Also can you provide details on the problem this caused you? Thanks +Maybe youneed to try a different font +I would prefer stack traces within comments though descriptions are fine +Looks good +I think this problem has been resolved now unit tests can not reproduce it any more +Maybe its possible to implement it in a Behavior so that it can be added to all Links? If this is a good idea I dont mind having a go at it myself +Thanks Stephen Eli and Luke +Closing all resolved tickets from or older +But when I thought more about it I think that it would be unnecessarily complex without adding significant value so Jamess simpler solution sounds better to meCommitted the patch to trunk with revision and to with revision I intend to commit the fix on the branch as well as requested by Kristian +Ill have another look +I see no changes that are obvious to me for without further advice or instruction +For the record my interrupt is better than polling alldbs thing is derived from operational experience +Keep them for now +Sorry dont understand what you mean +Ive seen this +Applying the patch solved the problem +As she is not a committer I will support her by and taking responsibility for committing supplied patches for this project +If it is not too much trouble could you provide a patch that corrects all of the errors introduced in rev +see also +Trying to think about this from a users standpoint +Attaching a patch which might be a solution to the issue +If there are no objections Ill check it in +Burr the reason for JBDS working might be because JBDS had a hack that marked the server Started if startup could not be determined in looks like the as startup poller doesnt work for your server +Patch which is up for review +This is with mvn versionMaven version version name linux version arch i Family unix +Can you upload Tomcat also? Thanks +Have to reopen this issue because it breaks the integration build by introducing a dependency to commons codec which is not available in Sling +Applied Douglas version in rddc +Or even better update the source and create a patchThanks sorry for the RTFM +nkeywal What you need me to do here? Should I try and see if it still has issue Andrew identified +We must review the in order to allow such name collisions +In any event I ran times and saw the problem in of the test runs +For future reference the wiki can be edited by anyone so you could have just updated the link yourself +ScottI looked at the patch and only saw changes to +I must be really blind +Under high load parallelized state saving actually brings the system down faster because the thread manager cant keep up. +Target findbug warnings suppressed +I removed them as the osgi container was complaining it couldnt resolve the referencesThe libraries you attached didnt work for me as they were compiled against libc and Ive only got on RHELGood to see someone taking an interest in the MC +on thinking about splittable META right now +resolved +Im a on the pseudo vote here +Since I am working on the version your fix in the trunk does not there is no Class and no package +Ive removed the profile and the line excluding the test I also removed the extra instructions from the docs +XML file with sample rules attached +IT Same problem with Eclipse local mode secure instance +Actually that would be the Startup Startup event in master branch +But then again you have HQL to worry about +The latest patch looks good thanks Suresh for merging it with trunk +I forgot to close it because I was waiting for any eventual bugs +committed +I would love to get but we cant upgrade to until its more mature and part of DSE +Ignore this comment +Removed it for now to be consistent with other +Hi MirekWhen you set maxretry to and after the retry the node A gives up +that was ignored +Hi Daryn I raised for handling the similar problems in +Hasnt been tested Ive ported over the code added some comments and some basic docs and am looking for initial review before going any further +Same problem for me +Patch looks good Shai +Thanks Billy for all the hard work and endless benchmarking so nice to have a block codec that is simple and clean and reuses our packed ints optimizations +Hmm yeah sounds more flexible for the future and I see the point to store related infos +Thats a hack according to my beliefPlease also check into the way the base decode length is returned +Im closing this bug +If instead it were an array or a class type then you would see a string there +Ive implemented getuid and getgid for the backend in r of via shelling out to id u and id g respectively +It should be linked to not +Ive committed my proposed patch with Nadavs suggested changes +Got busy with some other works +Can you check on thatThanksMark +We have stylesheets for the latter task +Unfortunately the master is not standards compliant here +When we create a new object we automatically generate the uuid so that we can identify the new object with its generated PK when it comes back from the server to the clientI voted for making the uuid our pk but users seem to hate having to do select from table where id rather than id go figureSo your partial fix would most likely be a full fix for us +Looked at my busy but healthy cluster +Updated patch +a trailing slash is one thing but a double slash in between +If we remove the context completely then how do we pass ACLs around in the unmanaged case? If this is correct then perhaps instead of removing context from the launcher we might want to improve it to accept ACLs and add it to the context +SeeImproving serialization would be nice but wed have to wreck our serialization mechanism and have something custom on the wire +Add decode logic to to recursively display message structure values and type codesRooted at qpidcppbindingsqpidexamples patch provides a nearly +Usage that was the case +Closing resolved issues for already released versions +In certain circumstances like a classroom we have to assume that we will always be disconnected +I implemented the above and it works very well +If the underlying transport is not destroyed when the is destroyed this must mean another location in your code is still holding on to a sharedptr for the underlying transport +from my side +As I commented in Im not really certain sbinservice is the way to test this +language spec should be complete for +The more users we get the more I think this is going to be an issue for us +Honestly OOME are common in many situations and not necessarily related to commons +If there are improvements to do on Springs side beyond those basics Id rather only do them for +This is a first pass at dealing with this +It seems that we have the exact same problem here with SonarI checked that we do not perform parallel runs and we dont +Recommend you switch to as there are many fixes in there +Also created to add method to get address from client proxy instead config. +Patch to take care of in the also added a for cache specs +When recursively listing under a directory if a subdirectory is deleted the more appropriate response is to ignore for that directory and return the remaining results +If theres a specific action Jaxen can take short of moving to Maven and the Maven folks tell us what that is then we can do term I think well move to Maven after and presumably that will fix any remaining issues +There is a new version of will it do the speed ups that you were mentioning +is off the hook +I believe the fix was to change the bpel facet ID so that its consistent +The submodule would haveto be built first even if the user had no need for it which is not what we want +I checked just now +I am sure things like this and others related to speed will be included in ruby like optional typing +Backported to as well. +Hmm did you expect a SVG? Its serialized as JPEG so a SVG browser cant showit +Patch that containes updated Dutch je Roland ik was te lui om dit zelf te doen +Entities are generated also tested the Scaffold UI from entities but it doesnt work +never mind my last comment I just traced the problem it was on my end +usrlibsparkorg rootroot +proposing to move this to next releaes as this is too late to fix the UI part +will create a new issue for the next part +Moving this to because these pages should actually be moved somewhere else +I may try to put up a reproducable app soon +in +However I dont think the test file is laid out correctly +HiHere is the missing fileIm sorry for the delayRegardsve applied the patch with version Thanks to Eric for his contribution. +Ill start the VM tomorrow +pom and I instructed Archiva to generate a +Thanks charith for the contribution +shouldnt be burdened with coding for a specific schedulerOK making this required sounds reasonable to me +IR was a respin of IR and therefore did not pull across this authentication error with JMS is a configuration issue +Please verify the docs from stage during CR test cycle +Thanks Daniel +As far as I know the spec does not mention a header called so why is this a Felix bugUse if you want to query versions of packages instead +A repro for this problem is attached to +Works great thanks. +This has been resolved on trunk with the attached patch and the helpful comments from the review board +Any additional test in the main iteration loop is undesirable +Thanks +Try adding set x to the start of your postinstall script to get more debugging info from the shell +was fixed +ResolvedThe patch committed in both Beta branch and trunkNow the action set is enabled in the following perspectives CSS Editing Web Development +as well +Its now around times faster on my Oracle db +We need additional tests for serialization and for foreign characters both with andthanksdims +Looks good to me +Thanks +Thanks +Lets dig out this years old issue with last comment from years agoMaybe this was actually implemented by now +class ever goes missing again +How fast do you think you can take a look at this? The main reason Im asking is because well probably release CARGO by end of July so was wondering if it would make it +So in builds previous to the refactor the bug was not visible because the codes workflow was slightly different and the bug was never hit +Thus I cant verify it right now will do laterThe patch also contains new test and instrumentation but it wont be verified until AOP is fixed in trunk +Do you know if code alsoexists as a standard java extension for older? only implementation of that I know of is unfortunately suns +I want to make corrupted PDF as another issue from this \ No newline at end of file